diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,40374 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 3.0,
+  "eval_steps": 500,
+  "global_step": 5763,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0005205622071837585,
+      "grad_norm": 6.376532092079771,
+      "learning_rate": 2.890173410404624e-07,
+      "loss": 0.5233,
+      "step": 1
+    },
+    {
+      "epoch": 0.001041124414367517,
+      "grad_norm": 6.4124152514036075,
+      "learning_rate": 5.780346820809248e-07,
+      "loss": 0.5204,
+      "step": 2
+    },
+    {
+      "epoch": 0.0015616866215512754,
+      "grad_norm": 6.458779162822453,
+      "learning_rate": 8.670520231213873e-07,
+      "loss": 0.529,
+      "step": 3
+    },
+    {
+      "epoch": 0.002082248828735034,
+      "grad_norm": 6.285519837284632,
+      "learning_rate": 1.1560693641618497e-06,
+      "loss": 0.5214,
+      "step": 4
+    },
+    {
+      "epoch": 0.0026028110359187923,
+      "grad_norm": 5.796645692951652,
+      "learning_rate": 1.4450867052023122e-06,
+      "loss": 0.5185,
+      "step": 5
+    },
+    {
+      "epoch": 0.0031233732431025507,
+      "grad_norm": 3.9879840763701693,
+      "learning_rate": 1.7341040462427746e-06,
+      "loss": 0.4823,
+      "step": 6
+    },
+    {
+      "epoch": 0.003643935450286309,
+      "grad_norm": 3.6346238149034646,
+      "learning_rate": 2.023121387283237e-06,
+      "loss": 0.4531,
+      "step": 7
+    },
+    {
+      "epoch": 0.004164497657470068,
+      "grad_norm": 3.3536127160309777,
+      "learning_rate": 2.3121387283236993e-06,
+      "loss": 0.427,
+      "step": 8
+    },
+    {
+      "epoch": 0.004685059864653826,
+      "grad_norm": 2.7703984048593324,
+      "learning_rate": 2.601156069364162e-06,
+      "loss": 0.3954,
+      "step": 9
+    },
+    {
+      "epoch": 0.0052056220718375845,
+      "grad_norm": 2.6269808449331644,
+      "learning_rate": 2.8901734104046244e-06,
+      "loss": 0.3784,
+      "step": 10
+    },
+    {
+      "epoch": 0.005726184279021343,
+      "grad_norm": 1.7055430768828044,
+      "learning_rate": 3.1791907514450866e-06,
+      "loss": 0.3593,
+      "step": 11
+    },
+    {
+      "epoch": 0.006246746486205101,
+      "grad_norm": 2.517105495712176,
+      "learning_rate": 3.468208092485549e-06,
+      "loss": 0.3521,
+      "step": 12
+    },
+    {
+      "epoch": 0.00676730869338886,
+      "grad_norm": 3.195572501235165,
+      "learning_rate": 3.757225433526012e-06,
+      "loss": 0.3539,
+      "step": 13
+    },
+    {
+      "epoch": 0.007287870900572618,
+      "grad_norm": 2.323008206835536,
+      "learning_rate": 4.046242774566474e-06,
+      "loss": 0.3425,
+      "step": 14
+    },
+    {
+      "epoch": 0.007808433107756377,
+      "grad_norm": 1.5310630307639206,
+      "learning_rate": 4.3352601156069365e-06,
+      "loss": 0.3317,
+      "step": 15
+    },
+    {
+      "epoch": 0.008328995314940135,
+      "grad_norm": 1.365492849675652,
+      "learning_rate": 4.624277456647399e-06,
+      "loss": 0.3148,
+      "step": 16
+    },
+    {
+      "epoch": 0.008849557522123894,
+      "grad_norm": 1.4095650207655759,
+      "learning_rate": 4.913294797687862e-06,
+      "loss": 0.3157,
+      "step": 17
+    },
+    {
+      "epoch": 0.009370119729307652,
+      "grad_norm": 1.5215580885689424,
+      "learning_rate": 5.202312138728324e-06,
+      "loss": 0.2976,
+      "step": 18
+    },
+    {
+      "epoch": 0.00989068193649141,
+      "grad_norm": 1.2318061379322545,
+      "learning_rate": 5.491329479768787e-06,
+      "loss": 0.3052,
+      "step": 19
+    },
+    {
+      "epoch": 0.010411244143675169,
+      "grad_norm": 1.1045087369383553,
+      "learning_rate": 5.780346820809249e-06,
+      "loss": 0.2965,
+      "step": 20
+    },
+    {
+      "epoch": 0.010931806350858927,
+      "grad_norm": 1.1036987329271282,
+      "learning_rate": 6.069364161849711e-06,
+      "loss": 0.2897,
+      "step": 21
+    },
+    {
+      "epoch": 0.011452368558042686,
+      "grad_norm": 1.0423572850985787,
+      "learning_rate": 6.358381502890173e-06,
+      "loss": 0.2845,
+      "step": 22
+    },
+    {
+      "epoch": 0.011972930765226444,
+      "grad_norm": 0.9845462775484162,
+      "learning_rate": 6.647398843930635e-06,
+      "loss": 0.268,
+      "step": 23
+    },
+    {
+      "epoch": 0.012493492972410203,
+      "grad_norm": 0.9205758761394418,
+      "learning_rate": 6.936416184971098e-06,
+      "loss": 0.2728,
+      "step": 24
+    },
+    {
+      "epoch": 0.013014055179593961,
+      "grad_norm": 0.936483616234927,
+      "learning_rate": 7.225433526011561e-06,
+      "loss": 0.2594,
+      "step": 25
+    },
+    {
+      "epoch": 0.01353461738677772,
+      "grad_norm": 0.8503591857964459,
+      "learning_rate": 7.514450867052024e-06,
+      "loss": 0.2628,
+      "step": 26
+    },
+    {
+      "epoch": 0.014055179593961478,
+      "grad_norm": 0.8672336072786627,
+      "learning_rate": 7.803468208092486e-06,
+      "loss": 0.2555,
+      "step": 27
+    },
+    {
+      "epoch": 0.014575741801145237,
+      "grad_norm": 0.9212778120486387,
+      "learning_rate": 8.092485549132949e-06,
+      "loss": 0.2547,
+      "step": 28
+    },
+    {
+      "epoch": 0.015096304008328995,
+      "grad_norm": 1.0623750925591287,
+      "learning_rate": 8.38150289017341e-06,
+      "loss": 0.2518,
+      "step": 29
+    },
+    {
+      "epoch": 0.015616866215512754,
+      "grad_norm": 1.0990772755341645,
+      "learning_rate": 8.670520231213873e-06,
+      "loss": 0.2542,
+      "step": 30
+    },
+    {
+      "epoch": 0.016137428422696512,
+      "grad_norm": 0.7201508676237249,
+      "learning_rate": 8.959537572254335e-06,
+      "loss": 0.2411,
+      "step": 31
+    },
+    {
+      "epoch": 0.01665799062988027,
+      "grad_norm": 0.8464593786312,
+      "learning_rate": 9.248554913294797e-06,
+      "loss": 0.2453,
+      "step": 32
+    },
+    {
+      "epoch": 0.01717855283706403,
+      "grad_norm": 0.8038877087534331,
+      "learning_rate": 9.53757225433526e-06,
+      "loss": 0.2358,
+      "step": 33
+    },
+    {
+      "epoch": 0.017699115044247787,
+      "grad_norm": 1.0702814348205445,
+      "learning_rate": 9.826589595375723e-06,
+      "loss": 0.2375,
+      "step": 34
+    },
+    {
+      "epoch": 0.018219677251431546,
+      "grad_norm": 1.425305836752578,
+      "learning_rate": 1.0115606936416185e-05,
+      "loss": 0.2344,
+      "step": 35
+    },
+    {
+      "epoch": 0.018740239458615304,
+      "grad_norm": 0.9940426565737253,
+      "learning_rate": 1.0404624277456647e-05,
+      "loss": 0.2388,
+      "step": 36
+    },
+    {
+      "epoch": 0.019260801665799063,
+      "grad_norm": 0.7934472038238055,
+      "learning_rate": 1.0693641618497111e-05,
+      "loss": 0.2352,
+      "step": 37
+    },
+    {
+      "epoch": 0.01978136387298282,
+      "grad_norm": 1.06277101439209,
+      "learning_rate": 1.0982658959537573e-05,
+      "loss": 0.2382,
+      "step": 38
+    },
+    {
+      "epoch": 0.02030192608016658,
+      "grad_norm": 1.2617518893593866,
+      "learning_rate": 1.1271676300578036e-05,
+      "loss": 0.2299,
+      "step": 39
+    },
+    {
+      "epoch": 0.020822488287350338,
+      "grad_norm": 1.0241590169095665,
+      "learning_rate": 1.1560693641618498e-05,
+      "loss": 0.2341,
+      "step": 40
+    },
+    {
+      "epoch": 0.021343050494534097,
+      "grad_norm": 0.7725404532271828,
+      "learning_rate": 1.184971098265896e-05,
+      "loss": 0.2361,
+      "step": 41
+    },
+    {
+      "epoch": 0.021863612701717855,
+      "grad_norm": 0.8376692345275107,
+      "learning_rate": 1.2138728323699422e-05,
+      "loss": 0.2362,
+      "step": 42
+    },
+    {
+      "epoch": 0.022384174908901613,
+      "grad_norm": 1.0428393775559166,
+      "learning_rate": 1.2427745664739884e-05,
+      "loss": 0.2338,
+      "step": 43
+    },
+    {
+      "epoch": 0.022904737116085372,
+      "grad_norm": 0.7784715293287426,
+      "learning_rate": 1.2716763005780346e-05,
+      "loss": 0.2316,
+      "step": 44
+    },
+    {
+      "epoch": 0.02342529932326913,
+      "grad_norm": 0.9048629801576277,
+      "learning_rate": 1.3005780346820809e-05,
+      "loss": 0.2253,
+      "step": 45
+    },
+    {
+      "epoch": 0.02394586153045289,
+      "grad_norm": 1.1048673864810445,
+      "learning_rate": 1.329479768786127e-05,
+      "loss": 0.229,
+      "step": 46
+    },
+    {
+      "epoch": 0.024466423737636647,
+      "grad_norm": 0.7942075601745159,
+      "learning_rate": 1.3583815028901733e-05,
+      "loss": 0.2285,
+      "step": 47
+    },
+    {
+      "epoch": 0.024986985944820406,
+      "grad_norm": 0.9826389811914579,
+      "learning_rate": 1.3872832369942197e-05,
+      "loss": 0.2276,
+      "step": 48
+    },
+    {
+      "epoch": 0.025507548152004164,
+      "grad_norm": 1.6396773861852498,
+      "learning_rate": 1.416184971098266e-05,
+      "loss": 0.2242,
+      "step": 49
+    },
+    {
+      "epoch": 0.026028110359187923,
+      "grad_norm": 1.0452409902447666,
+      "learning_rate": 1.4450867052023123e-05,
+      "loss": 0.2281,
+      "step": 50
+    },
+    {
+      "epoch": 0.02654867256637168,
+      "grad_norm": 1.0756275987207078,
+      "learning_rate": 1.4739884393063585e-05,
+      "loss": 0.2277,
+      "step": 51
+    },
+    {
+      "epoch": 0.02706923477355544,
+      "grad_norm": 0.8991098023314276,
+      "learning_rate": 1.5028901734104049e-05,
+      "loss": 0.2172,
+      "step": 52
+    },
+    {
+      "epoch": 0.027589796980739198,
+      "grad_norm": 0.9582658807979682,
+      "learning_rate": 1.531791907514451e-05,
+      "loss": 0.2178,
+      "step": 53
+    },
+    {
+      "epoch": 0.028110359187922956,
+      "grad_norm": 0.9911148408927335,
+      "learning_rate": 1.5606936416184973e-05,
+      "loss": 0.2341,
+      "step": 54
+    },
+    {
+      "epoch": 0.028630921395106715,
+      "grad_norm": 1.3180692766328292,
+      "learning_rate": 1.5895953757225435e-05,
+      "loss": 0.2195,
+      "step": 55
+    },
+    {
+      "epoch": 0.029151483602290473,
+      "grad_norm": 1.486871730799441,
+      "learning_rate": 1.6184971098265897e-05,
+      "loss": 0.2247,
+      "step": 56
+    },
+    {
+      "epoch": 0.029672045809474232,
+      "grad_norm": 0.8943418358679089,
+      "learning_rate": 1.647398843930636e-05,
+      "loss": 0.2208,
+      "step": 57
+    },
+    {
+      "epoch": 0.03019260801665799,
+      "grad_norm": 0.9549840398600492,
+      "learning_rate": 1.676300578034682e-05,
+      "loss": 0.2303,
+      "step": 58
+    },
+    {
+      "epoch": 0.03071317022384175,
+      "grad_norm": 1.6440802970143786,
+      "learning_rate": 1.7052023121387284e-05,
+      "loss": 0.2189,
+      "step": 59
+    },
+    {
+      "epoch": 0.031233732431025507,
+      "grad_norm": 0.8692681682580999,
+      "learning_rate": 1.7341040462427746e-05,
+      "loss": 0.2242,
+      "step": 60
+    },
+    {
+      "epoch": 0.031754294638209266,
+      "grad_norm": 1.0685256597249488,
+      "learning_rate": 1.7630057803468208e-05,
+      "loss": 0.2124,
+      "step": 61
+    },
+    {
+      "epoch": 0.032274856845393024,
+      "grad_norm": 1.401339524265495,
+      "learning_rate": 1.791907514450867e-05,
+      "loss": 0.222,
+      "step": 62
+    },
+    {
+      "epoch": 0.03279541905257678,
+      "grad_norm": 0.8600585202663065,
+      "learning_rate": 1.8208092485549132e-05,
+      "loss": 0.2138,
+      "step": 63
+    },
+    {
+      "epoch": 0.03331598125976054,
+      "grad_norm": 1.2477374949025173,
+      "learning_rate": 1.8497109826589594e-05,
+      "loss": 0.2282,
+      "step": 64
+    },
+    {
+      "epoch": 0.0338365434669443,
+      "grad_norm": 1.097660347997359,
+      "learning_rate": 1.8786127167630057e-05,
+      "loss": 0.217,
+      "step": 65
+    },
+    {
+      "epoch": 0.03435710567412806,
+      "grad_norm": 1.2951250009085689,
+      "learning_rate": 1.907514450867052e-05,
+      "loss": 0.2199,
+      "step": 66
+    },
+    {
+      "epoch": 0.034877667881311816,
+      "grad_norm": 0.826872263396979,
+      "learning_rate": 1.936416184971098e-05,
+      "loss": 0.2134,
+      "step": 67
+    },
+    {
+      "epoch": 0.035398230088495575,
+      "grad_norm": 1.326379381993655,
+      "learning_rate": 1.9653179190751446e-05,
+      "loss": 0.2253,
+      "step": 68
+    },
+    {
+      "epoch": 0.03591879229567933,
+      "grad_norm": 0.8549714338599687,
+      "learning_rate": 1.994219653179191e-05,
+      "loss": 0.2118,
+      "step": 69
+    },
+    {
+      "epoch": 0.03643935450286309,
+      "grad_norm": 1.1468222920831441,
+      "learning_rate": 2.023121387283237e-05,
+      "loss": 0.225,
+      "step": 70
+    },
+    {
+      "epoch": 0.03695991671004685,
+      "grad_norm": 0.8454115307212339,
+      "learning_rate": 2.0520231213872833e-05,
+      "loss": 0.2165,
+      "step": 71
+    },
+    {
+      "epoch": 0.03748047891723061,
+      "grad_norm": 0.8498885949939733,
+      "learning_rate": 2.0809248554913295e-05,
+      "loss": 0.2098,
+      "step": 72
+    },
+    {
+      "epoch": 0.03800104112441437,
+      "grad_norm": 0.8214398654369537,
+      "learning_rate": 2.1098265895953757e-05,
+      "loss": 0.2183,
+      "step": 73
+    },
+    {
+      "epoch": 0.038521603331598125,
+      "grad_norm": 1.0466964698294483,
+      "learning_rate": 2.1387283236994223e-05,
+      "loss": 0.2193,
+      "step": 74
+    },
+    {
+      "epoch": 0.039042165538781884,
+      "grad_norm": 1.1184070427085229,
+      "learning_rate": 2.1676300578034685e-05,
+      "loss": 0.2187,
+      "step": 75
+    },
+    {
+      "epoch": 0.03956272774596564,
+      "grad_norm": 1.040775059918436,
+      "learning_rate": 2.1965317919075147e-05,
+      "loss": 0.2194,
+      "step": 76
+    },
+    {
+      "epoch": 0.0400832899531494,
+      "grad_norm": 1.2038697374050564,
+      "learning_rate": 2.225433526011561e-05,
+      "loss": 0.2184,
+      "step": 77
+    },
+    {
+      "epoch": 0.04060385216033316,
+      "grad_norm": 0.8663131100942328,
+      "learning_rate": 2.254335260115607e-05,
+      "loss": 0.2107,
+      "step": 78
+    },
+    {
+      "epoch": 0.04112441436751692,
+      "grad_norm": 0.7771080010057934,
+      "learning_rate": 2.2832369942196533e-05,
+      "loss": 0.2091,
+      "step": 79
+    },
+    {
+      "epoch": 0.041644976574700676,
+      "grad_norm": 0.9075557788861287,
+      "learning_rate": 2.3121387283236996e-05,
+      "loss": 0.213,
+      "step": 80
+    },
+    {
+      "epoch": 0.042165538781884435,
+      "grad_norm": 1.1814000831436555,
+      "learning_rate": 2.3410404624277458e-05,
+      "loss": 0.2111,
+      "step": 81
+    },
+    {
+      "epoch": 0.04268610098906819,
+      "grad_norm": 0.808697178051325,
+      "learning_rate": 2.369942196531792e-05,
+      "loss": 0.2079,
+      "step": 82
+    },
+    {
+      "epoch": 0.04320666319625195,
+      "grad_norm": 0.9213395387224003,
+      "learning_rate": 2.3988439306358382e-05,
+      "loss": 0.213,
+      "step": 83
+    },
+    {
+      "epoch": 0.04372722540343571,
+      "grad_norm": 0.8345054388332048,
+      "learning_rate": 2.4277456647398844e-05,
+      "loss": 0.2099,
+      "step": 84
+    },
+    {
+      "epoch": 0.04424778761061947,
+      "grad_norm": 0.8849593943677236,
+      "learning_rate": 2.4566473988439306e-05,
+      "loss": 0.222,
+      "step": 85
+    },
+    {
+      "epoch": 0.04476834981780323,
+      "grad_norm": 0.8478338478744712,
+      "learning_rate": 2.485549132947977e-05,
+      "loss": 0.2203,
+      "step": 86
+    },
+    {
+      "epoch": 0.045288912024986985,
+      "grad_norm": 0.9085767079232848,
+      "learning_rate": 2.5144508670520234e-05,
+      "loss": 0.2134,
+      "step": 87
+    },
+    {
+      "epoch": 0.045809474232170744,
+      "grad_norm": 0.9152260021422175,
+      "learning_rate": 2.5433526011560693e-05,
+      "loss": 0.2212,
+      "step": 88
+    },
+    {
+      "epoch": 0.0463300364393545,
+      "grad_norm": 0.7878287625575658,
+      "learning_rate": 2.5722543352601158e-05,
+      "loss": 0.2088,
+      "step": 89
+    },
+    {
+      "epoch": 0.04685059864653826,
+      "grad_norm": 0.9443429001103073,
+      "learning_rate": 2.6011560693641617e-05,
+      "loss": 0.2126,
+      "step": 90
+    },
+    {
+      "epoch": 0.04737116085372202,
+      "grad_norm": 1.0473598485121471,
+      "learning_rate": 2.6300578034682083e-05,
+      "loss": 0.2211,
+      "step": 91
+    },
+    {
+      "epoch": 0.04789172306090578,
+      "grad_norm": 0.8361641199343858,
+      "learning_rate": 2.658959537572254e-05,
+      "loss": 0.2128,
+      "step": 92
+    },
+    {
+      "epoch": 0.048412285268089536,
+      "grad_norm": 0.7931786259368718,
+      "learning_rate": 2.6878612716763007e-05,
+      "loss": 0.2166,
+      "step": 93
+    },
+    {
+      "epoch": 0.048932847475273294,
+      "grad_norm": 0.8124807154897088,
+      "learning_rate": 2.7167630057803466e-05,
+      "loss": 0.2122,
+      "step": 94
+    },
+    {
+      "epoch": 0.04945340968245705,
+      "grad_norm": 0.7637460807323695,
+      "learning_rate": 2.745664739884393e-05,
+      "loss": 0.2157,
+      "step": 95
+    },
+    {
+      "epoch": 0.04997397188964081,
+      "grad_norm": 1.0197675734431848,
+      "learning_rate": 2.7745664739884393e-05,
+      "loss": 0.2202,
+      "step": 96
+    },
+    {
+      "epoch": 0.05049453409682457,
+      "grad_norm": 0.7718672139219506,
+      "learning_rate": 2.8034682080924855e-05,
+      "loss": 0.2157,
+      "step": 97
+    },
+    {
+      "epoch": 0.05101509630400833,
+      "grad_norm": 0.6967370402605544,
+      "learning_rate": 2.832369942196532e-05,
+      "loss": 0.2115,
+      "step": 98
+    },
+    {
+      "epoch": 0.05153565851119209,
+      "grad_norm": 0.8860212093442391,
+      "learning_rate": 2.861271676300578e-05,
+      "loss": 0.2164,
+      "step": 99
+    },
+    {
+      "epoch": 0.052056220718375845,
+      "grad_norm": 0.806336638090273,
+      "learning_rate": 2.8901734104046245e-05,
+      "loss": 0.2066,
+      "step": 100
+    },
+    {
+      "epoch": 0.052576782925559604,
+      "grad_norm": 0.8839202966628119,
+      "learning_rate": 2.9190751445086707e-05,
+      "loss": 0.2133,
+      "step": 101
+    },
+    {
+      "epoch": 0.05309734513274336,
+      "grad_norm": 1.0715053048576961,
+      "learning_rate": 2.947976878612717e-05,
+      "loss": 0.2094,
+      "step": 102
+    },
+    {
+      "epoch": 0.05361790733992712,
+      "grad_norm": 0.9923778641494126,
+      "learning_rate": 2.9768786127167632e-05,
+      "loss": 0.2147,
+      "step": 103
+    },
+    {
+      "epoch": 0.05413846954711088,
+      "grad_norm": 0.996319311639693,
+      "learning_rate": 3.0057803468208097e-05,
+      "loss": 0.2119,
+      "step": 104
+    },
+    {
+      "epoch": 0.05465903175429464,
+      "grad_norm": 0.8155226884859121,
+      "learning_rate": 3.0346820809248556e-05,
+      "loss": 0.2078,
+      "step": 105
+    },
+    {
+      "epoch": 0.055179593961478396,
+      "grad_norm": 0.7905094789778886,
+      "learning_rate": 3.063583815028902e-05,
+      "loss": 0.2096,
+      "step": 106
+    },
+    {
+      "epoch": 0.055700156168662154,
+      "grad_norm": 0.8773860436858838,
+      "learning_rate": 3.092485549132948e-05,
+      "loss": 0.2167,
+      "step": 107
+    },
+    {
+      "epoch": 0.05622071837584591,
+      "grad_norm": 0.8873612182478008,
+      "learning_rate": 3.1213872832369946e-05,
+      "loss": 0.2147,
+      "step": 108
+    },
+    {
+      "epoch": 0.05674128058302967,
+      "grad_norm": 0.8093261489985791,
+      "learning_rate": 3.1502890173410405e-05,
+      "loss": 0.2147,
+      "step": 109
+    },
+    {
+      "epoch": 0.05726184279021343,
+      "grad_norm": 0.7069615535722936,
+      "learning_rate": 3.179190751445087e-05,
+      "loss": 0.2175,
+      "step": 110
+    },
+    {
+      "epoch": 0.05778240499739719,
+      "grad_norm": 0.7496643468031777,
+      "learning_rate": 3.208092485549133e-05,
+      "loss": 0.2133,
+      "step": 111
+    },
+    {
+      "epoch": 0.05830296720458095,
+      "grad_norm": 0.6778780641394953,
+      "learning_rate": 3.2369942196531794e-05,
+      "loss": 0.2142,
+      "step": 112
+    },
+    {
+      "epoch": 0.058823529411764705,
+      "grad_norm": 0.8038377581955503,
+      "learning_rate": 3.265895953757225e-05,
+      "loss": 0.2081,
+      "step": 113
+    },
+    {
+      "epoch": 0.059344091618948464,
+      "grad_norm": 0.8262906806241153,
+      "learning_rate": 3.294797687861272e-05,
+      "loss": 0.2198,
+      "step": 114
+    },
+    {
+      "epoch": 0.05986465382613222,
+      "grad_norm": 0.8225060449744769,
+      "learning_rate": 3.323699421965318e-05,
+      "loss": 0.2085,
+      "step": 115
+    },
+    {
+      "epoch": 0.06038521603331598,
+      "grad_norm": 0.9474684268193776,
+      "learning_rate": 3.352601156069364e-05,
+      "loss": 0.2141,
+      "step": 116
+    },
+    {
+      "epoch": 0.06090577824049974,
+      "grad_norm": 0.9490625592067119,
+      "learning_rate": 3.381502890173411e-05,
+      "loss": 0.2116,
+      "step": 117
+    },
+    {
+      "epoch": 0.0614263404476835,
+      "grad_norm": 1.0111096680389842,
+      "learning_rate": 3.410404624277457e-05,
+      "loss": 0.2199,
+      "step": 118
+    },
+    {
+      "epoch": 0.061946902654867256,
+      "grad_norm": 0.9650260633005836,
+      "learning_rate": 3.439306358381503e-05,
+      "loss": 0.2126,
+      "step": 119
+    },
+    {
+      "epoch": 0.062467464862051014,
+      "grad_norm": 0.6980785653915276,
+      "learning_rate": 3.468208092485549e-05,
+      "loss": 0.2099,
+      "step": 120
+    },
+    {
+      "epoch": 0.06298802706923477,
+      "grad_norm": 0.8534007635131398,
+      "learning_rate": 3.497109826589596e-05,
+      "loss": 0.2193,
+      "step": 121
+    },
+    {
+      "epoch": 0.06350858927641853,
+      "grad_norm": 0.8531206959278465,
+      "learning_rate": 3.5260115606936416e-05,
+      "loss": 0.2121,
+      "step": 122
+    },
+    {
+      "epoch": 0.06402915148360229,
+      "grad_norm": 0.8548017488318986,
+      "learning_rate": 3.554913294797688e-05,
+      "loss": 0.2228,
+      "step": 123
+    },
+    {
+      "epoch": 0.06454971369078605,
+      "grad_norm": 0.6612943150792601,
+      "learning_rate": 3.583815028901734e-05,
+      "loss": 0.2103,
+      "step": 124
+    },
+    {
+      "epoch": 0.0650702758979698,
+      "grad_norm": 0.7715488990577469,
+      "learning_rate": 3.6127167630057806e-05,
+      "loss": 0.2035,
+      "step": 125
+    },
+    {
+      "epoch": 0.06559083810515356,
+      "grad_norm": 0.8671638801555337,
+      "learning_rate": 3.6416184971098265e-05,
+      "loss": 0.2152,
+      "step": 126
+    },
+    {
+      "epoch": 0.06611140031233732,
+      "grad_norm": 0.9914635629797826,
+      "learning_rate": 3.670520231213873e-05,
+      "loss": 0.2162,
+      "step": 127
+    },
+    {
+      "epoch": 0.06663196251952108,
+      "grad_norm": 0.979896089844123,
+      "learning_rate": 3.699421965317919e-05,
+      "loss": 0.2161,
+      "step": 128
+    },
+    {
+      "epoch": 0.06715252472670484,
+      "grad_norm": 0.6633992741028703,
+      "learning_rate": 3.7283236994219654e-05,
+      "loss": 0.2121,
+      "step": 129
+    },
+    {
+      "epoch": 0.0676730869338886,
+      "grad_norm": 0.7005104328408702,
+      "learning_rate": 3.757225433526011e-05,
+      "loss": 0.2068,
+      "step": 130
+    },
+    {
+      "epoch": 0.06819364914107236,
+      "grad_norm": 0.7770874448510542,
+      "learning_rate": 3.786127167630058e-05,
+      "loss": 0.2148,
+      "step": 131
+    },
+    {
+      "epoch": 0.06871421134825612,
+      "grad_norm": 0.8338865623089665,
+      "learning_rate": 3.815028901734104e-05,
+      "loss": 0.2143,
+      "step": 132
+    },
+    {
+      "epoch": 0.06923477355543987,
+      "grad_norm": 0.7816413255198202,
+      "learning_rate": 3.84393063583815e-05,
+      "loss": 0.2189,
+      "step": 133
+    },
+    {
+      "epoch": 0.06975533576262363,
+      "grad_norm": 0.65556462363938,
+      "learning_rate": 3.872832369942196e-05,
+      "loss": 0.2079,
+      "step": 134
+    },
+    {
+      "epoch": 0.07027589796980739,
+      "grad_norm": 0.7701568466848703,
+      "learning_rate": 3.901734104046243e-05,
+      "loss": 0.2093,
+      "step": 135
+    },
+    {
+      "epoch": 0.07079646017699115,
+      "grad_norm": 0.832197605955844,
+      "learning_rate": 3.930635838150289e-05,
+      "loss": 0.2131,
+      "step": 136
+    },
+    {
+      "epoch": 0.07131702238417491,
+      "grad_norm": 0.8404782199225694,
+      "learning_rate": 3.959537572254335e-05,
+      "loss": 0.2183,
+      "step": 137
+    },
+    {
+      "epoch": 0.07183758459135867,
+      "grad_norm": 0.6621631106005148,
+      "learning_rate": 3.988439306358382e-05,
+      "loss": 0.2176,
+      "step": 138
+    },
+    {
+      "epoch": 0.07235814679854242,
+      "grad_norm": 0.800987879800019,
+      "learning_rate": 4.0173410404624276e-05,
+      "loss": 0.2068,
+      "step": 139
+    },
+    {
+      "epoch": 0.07287870900572618,
+      "grad_norm": 0.8086309231367268,
+      "learning_rate": 4.046242774566474e-05,
+      "loss": 0.2149,
+      "step": 140
+    },
+    {
+      "epoch": 0.07339927121290994,
+      "grad_norm": 0.6390918137014469,
+      "learning_rate": 4.07514450867052e-05,
+      "loss": 0.2067,
+      "step": 141
+    },
+    {
+      "epoch": 0.0739198334200937,
+      "grad_norm": 0.6208193264026112,
+      "learning_rate": 4.1040462427745666e-05,
+      "loss": 0.2146,
+      "step": 142
+    },
+    {
+      "epoch": 0.07444039562727746,
+      "grad_norm": 0.6102728010873583,
+      "learning_rate": 4.132947976878613e-05,
+      "loss": 0.2136,
+      "step": 143
+    },
+    {
+      "epoch": 0.07496095783446122,
+      "grad_norm": 0.7098434352977124,
+      "learning_rate": 4.161849710982659e-05,
+      "loss": 0.2124,
+      "step": 144
+    },
+    {
+      "epoch": 0.07548152004164498,
+      "grad_norm": 0.7136938829382273,
+      "learning_rate": 4.1907514450867055e-05,
+      "loss": 0.2176,
+      "step": 145
+    },
+    {
+      "epoch": 0.07600208224882873,
+      "grad_norm": 0.6206135947118576,
+      "learning_rate": 4.2196531791907514e-05,
+      "loss": 0.2128,
+      "step": 146
+    },
+    {
+      "epoch": 0.07652264445601249,
+      "grad_norm": 0.6662245815894154,
+      "learning_rate": 4.248554913294798e-05,
+      "loss": 0.2096,
+      "step": 147
+    },
+    {
+      "epoch": 0.07704320666319625,
+      "grad_norm": 0.6047404577832085,
+      "learning_rate": 4.2774566473988445e-05,
+      "loss": 0.2067,
+      "step": 148
+    },
+    {
+      "epoch": 0.07756376887038001,
+      "grad_norm": 0.6072140153823105,
+      "learning_rate": 4.3063583815028904e-05,
+      "loss": 0.2173,
+      "step": 149
+    },
+    {
+      "epoch": 0.07808433107756377,
+      "grad_norm": 0.6556745346884483,
+      "learning_rate": 4.335260115606937e-05,
+      "loss": 0.214,
+      "step": 150
+    },
+    {
+      "epoch": 0.07860489328474753,
+      "grad_norm": 0.593639230209489,
+      "learning_rate": 4.364161849710983e-05,
+      "loss": 0.2088,
+      "step": 151
+    },
+    {
+      "epoch": 0.07912545549193128,
+      "grad_norm": 0.6371994241596302,
+      "learning_rate": 4.3930635838150294e-05,
+      "loss": 0.2046,
+      "step": 152
+    },
+    {
+      "epoch": 0.07964601769911504,
+      "grad_norm": 0.638999242172759,
+      "learning_rate": 4.421965317919075e-05,
+      "loss": 0.2161,
+      "step": 153
+    },
+    {
+      "epoch": 0.0801665799062988,
+      "grad_norm": 0.5754915699358542,
+      "learning_rate": 4.450867052023122e-05,
+      "loss": 0.2057,
+      "step": 154
+    },
+    {
+      "epoch": 0.08068714211348256,
+      "grad_norm": 0.5766300270893668,
+      "learning_rate": 4.4797687861271684e-05,
+      "loss": 0.2087,
+      "step": 155
+    },
+    {
+      "epoch": 0.08120770432066632,
+      "grad_norm": 0.6651018920554423,
+      "learning_rate": 4.508670520231214e-05,
+      "loss": 0.2111,
+      "step": 156
+    },
+    {
+      "epoch": 0.08172826652785008,
+      "grad_norm": 0.6875773025621368,
+      "learning_rate": 4.537572254335261e-05,
+      "loss": 0.2168,
+      "step": 157
+    },
+    {
+      "epoch": 0.08224882873503384,
+      "grad_norm": 0.6150201790012484,
+      "learning_rate": 4.566473988439307e-05,
+      "loss": 0.2141,
+      "step": 158
+    },
+    {
+      "epoch": 0.0827693909422176,
+      "grad_norm": 0.6599147042410466,
+      "learning_rate": 4.595375722543353e-05,
+      "loss": 0.2153,
+      "step": 159
+    },
+    {
+      "epoch": 0.08328995314940135,
+      "grad_norm": 0.7633380008679602,
+      "learning_rate": 4.624277456647399e-05,
+      "loss": 0.2118,
+      "step": 160
+    },
+    {
+      "epoch": 0.08381051535658511,
+      "grad_norm": 0.7353603486570877,
+      "learning_rate": 4.653179190751446e-05,
+      "loss": 0.2125,
+      "step": 161
+    },
+    {
+      "epoch": 0.08433107756376887,
+      "grad_norm": 0.7963914788392694,
+      "learning_rate": 4.6820809248554915e-05,
+      "loss": 0.2177,
+      "step": 162
+    },
+    {
+      "epoch": 0.08485163977095263,
+      "grad_norm": 0.7816123652355957,
+      "learning_rate": 4.710982658959538e-05,
+      "loss": 0.2113,
+      "step": 163
+    },
+    {
+      "epoch": 0.08537220197813639,
+      "grad_norm": 0.673743870011103,
+      "learning_rate": 4.739884393063584e-05,
+      "loss": 0.215,
+      "step": 164
+    },
+    {
+      "epoch": 0.08589276418532014,
+      "grad_norm": 0.5227320089849515,
+      "learning_rate": 4.7687861271676305e-05,
+      "loss": 0.2103,
+      "step": 165
+    },
+    {
+      "epoch": 0.0864133263925039,
+      "grad_norm": 0.678198278022982,
+      "learning_rate": 4.7976878612716764e-05,
+      "loss": 0.2193,
+      "step": 166
+    },
+    {
+      "epoch": 0.08693388859968766,
+      "grad_norm": 0.6374039379559228,
+      "learning_rate": 4.826589595375723e-05,
+      "loss": 0.2167,
+      "step": 167
+    },
+    {
+      "epoch": 0.08745445080687142,
+      "grad_norm": 0.5577509103478314,
+      "learning_rate": 4.855491329479769e-05,
+      "loss": 0.2047,
+      "step": 168
+    },
+    {
+      "epoch": 0.08797501301405518,
+      "grad_norm": 0.6071594590963697,
+      "learning_rate": 4.8843930635838154e-05,
+      "loss": 0.2129,
+      "step": 169
+    },
+    {
+      "epoch": 0.08849557522123894,
+      "grad_norm": 0.5535849448502077,
+      "learning_rate": 4.913294797687861e-05,
+      "loss": 0.2112,
+      "step": 170
+    },
+    {
+      "epoch": 0.0890161374284227,
+      "grad_norm": 0.5570160624184318,
+      "learning_rate": 4.942196531791908e-05,
+      "loss": 0.2249,
+      "step": 171
+    },
+    {
+      "epoch": 0.08953669963560645,
+      "grad_norm": 0.5926801003230638,
+      "learning_rate": 4.971098265895954e-05,
+      "loss": 0.2183,
+      "step": 172
+    },
+    {
+      "epoch": 0.09005726184279021,
+      "grad_norm": 0.6340827920564475,
+      "learning_rate": 5e-05,
+      "loss": 0.2134,
+      "step": 173
+    },
+    {
+      "epoch": 0.09057782404997397,
+      "grad_norm": 0.5566326852792648,
+      "learning_rate": 4.99999960519183e-05,
+      "loss": 0.2109,
+      "step": 174
+    },
+    {
+      "epoch": 0.09109838625715773,
+      "grad_norm": 0.6364150749071997,
+      "learning_rate": 4.999998420767445e-05,
+      "loss": 0.2207,
+      "step": 175
+    },
+    {
+      "epoch": 0.09161894846434149,
+      "grad_norm": 0.5946977297441665,
+      "learning_rate": 4.999996446727219e-05,
+      "loss": 0.22,
+      "step": 176
+    },
+    {
+      "epoch": 0.09213951067152525,
+      "grad_norm": 0.5756761019081571,
+      "learning_rate": 4.9999936830717745e-05,
+      "loss": 0.2157,
+      "step": 177
+    },
+    {
+      "epoch": 0.092660072878709,
+      "grad_norm": 0.755144779893014,
+      "learning_rate": 4.999990129801986e-05,
+      "loss": 0.2149,
+      "step": 178
+    },
+    {
+      "epoch": 0.09318063508589276,
+      "grad_norm": 0.8329885805863328,
+      "learning_rate": 4.9999857869189735e-05,
+      "loss": 0.2104,
+      "step": 179
+    },
+    {
+      "epoch": 0.09370119729307652,
+      "grad_norm": 0.7690781702453137,
+      "learning_rate": 4.999980654424112e-05,
+      "loss": 0.2078,
+      "step": 180
+    },
+    {
+      "epoch": 0.09422175950026028,
+      "grad_norm": 0.6308760503667642,
+      "learning_rate": 4.9999747323190195e-05,
+      "loss": 0.2197,
+      "step": 181
+    },
+    {
+      "epoch": 0.09474232170744404,
+      "grad_norm": 0.6148255176499823,
+      "learning_rate": 4.9999680206055686e-05,
+      "loss": 0.213,
+      "step": 182
+    },
+    {
+      "epoch": 0.0952628839146278,
+      "grad_norm": 0.6734184929964855,
+      "learning_rate": 4.999960519285878e-05,
+      "loss": 0.203,
+      "step": 183
+    },
+    {
+      "epoch": 0.09578344612181156,
+      "grad_norm": 0.5406620690992346,
+      "learning_rate": 4.999952228362317e-05,
+      "loss": 0.2051,
+      "step": 184
+    },
+    {
+      "epoch": 0.09630400832899531,
+      "grad_norm": 0.5805819703164237,
+      "learning_rate": 4.999943147837505e-05,
+      "loss": 0.2243,
+      "step": 185
+    },
+    {
+      "epoch": 0.09682457053617907,
+      "grad_norm": 0.6790309432200481,
+      "learning_rate": 4.999933277714309e-05,
+      "loss": 0.223,
+      "step": 186
+    },
+    {
+      "epoch": 0.09734513274336283,
+      "grad_norm": 0.5725871196289053,
+      "learning_rate": 4.999922617995847e-05,
+      "loss": 0.2137,
+      "step": 187
+    },
+    {
+      "epoch": 0.09786569495054659,
+      "grad_norm": 0.5475251087174847,
+      "learning_rate": 4.999911168685486e-05,
+      "loss": 0.2151,
+      "step": 188
+    },
+    {
+      "epoch": 0.09838625715773035,
+      "grad_norm": 0.6035996664978388,
+      "learning_rate": 4.999898929786842e-05,
+      "loss": 0.2073,
+      "step": 189
+    },
+    {
+      "epoch": 0.0989068193649141,
+      "grad_norm": 0.6086080086431157,
+      "learning_rate": 4.999885901303781e-05,
+      "loss": 0.2183,
+      "step": 190
+    },
+    {
+      "epoch": 0.09942738157209786,
+      "grad_norm": 0.4937337226206942,
+      "learning_rate": 4.999872083240418e-05,
+      "loss": 0.2102,
+      "step": 191
+    },
+    {
+      "epoch": 0.09994794377928162,
+      "grad_norm": 0.725349945989928,
+      "learning_rate": 4.999857475601117e-05,
+      "loss": 0.2089,
+      "step": 192
+    },
+    {
+      "epoch": 0.10046850598646538,
+      "grad_norm": 0.7612664949980464,
+      "learning_rate": 4.999842078390492e-05,
+      "loss": 0.1991,
+      "step": 193
+    },
+    {
+      "epoch": 0.10098906819364914,
+      "grad_norm": 0.5707299858159967,
+      "learning_rate": 4.9998258916134055e-05,
+      "loss": 0.2126,
+      "step": 194
+    },
+    {
+      "epoch": 0.1015096304008329,
+      "grad_norm": 0.5643415112854695,
+      "learning_rate": 4.999808915274971e-05,
+      "loss": 0.2069,
+      "step": 195
+    },
+    {
+      "epoch": 0.10203019260801666,
+      "grad_norm": 0.5633442640544284,
+      "learning_rate": 4.999791149380549e-05,
+      "loss": 0.2132,
+      "step": 196
+    },
+    {
+      "epoch": 0.10255075481520042,
+      "grad_norm": 0.5614129286121543,
+      "learning_rate": 4.999772593935752e-05,
+      "loss": 0.2054,
+      "step": 197
+    },
+    {
+      "epoch": 0.10307131702238417,
+      "grad_norm": 0.5646241630980633,
+      "learning_rate": 4.999753248946441e-05,
+      "loss": 0.21,
+      "step": 198
+    },
+    {
+      "epoch": 0.10359187922956793,
+      "grad_norm": 0.5425474957654737,
+      "learning_rate": 4.9997331144187255e-05,
+      "loss": 0.2187,
+      "step": 199
+    },
+    {
+      "epoch": 0.10411244143675169,
+      "grad_norm": 0.6177258752993052,
+      "learning_rate": 4.999712190358965e-05,
+      "loss": 0.2082,
+      "step": 200
+    },
+    {
+      "epoch": 0.10463300364393545,
+      "grad_norm": 0.6029817403963842,
+      "learning_rate": 4.999690476773767e-05,
+      "loss": 0.212,
+      "step": 201
+    },
+    {
+      "epoch": 0.10515356585111921,
+      "grad_norm": 0.48878469751460535,
+      "learning_rate": 4.9996679736699924e-05,
+      "loss": 0.2024,
+      "step": 202
+    },
+    {
+      "epoch": 0.10567412805830297,
+      "grad_norm": 0.5654845375393931,
+      "learning_rate": 4.9996446810547464e-05,
+      "loss": 0.2107,
+      "step": 203
+    },
+    {
+      "epoch": 0.10619469026548672,
+      "grad_norm": 0.5149953607057386,
+      "learning_rate": 4.999620598935387e-05,
+      "loss": 0.1983,
+      "step": 204
+    },
+    {
+      "epoch": 0.10671525247267048,
+      "grad_norm": 0.4962347495654175,
+      "learning_rate": 4.9995957273195206e-05,
+      "loss": 0.2095,
+      "step": 205
+    },
+    {
+      "epoch": 0.10723581467985424,
+      "grad_norm": 0.48397875587172107,
+      "learning_rate": 4.9995700662150015e-05,
+      "loss": 0.2062,
+      "step": 206
+    },
+    {
+      "epoch": 0.107756376887038,
+      "grad_norm": 0.49662456882149275,
+      "learning_rate": 4.9995436156299355e-05,
+      "loss": 0.2186,
+      "step": 207
+    },
+    {
+      "epoch": 0.10827693909422176,
+      "grad_norm": 0.5132623063501726,
+      "learning_rate": 4.999516375572677e-05,
+      "loss": 0.2106,
+      "step": 208
+    },
+    {
+      "epoch": 0.10879750130140552,
+      "grad_norm": 0.48217335031801317,
+      "learning_rate": 4.99948834605183e-05,
+      "loss": 0.2196,
+      "step": 209
+    },
+    {
+      "epoch": 0.10931806350858927,
+      "grad_norm": 0.5083613448407168,
+      "learning_rate": 4.999459527076247e-05,
+      "loss": 0.2092,
+      "step": 210
+    },
+    {
+      "epoch": 0.10983862571577303,
+      "grad_norm": 0.45169614742277503,
+      "learning_rate": 4.99942991865503e-05,
+      "loss": 0.1936,
+      "step": 211
+    },
+    {
+      "epoch": 0.11035918792295679,
+      "grad_norm": 0.5475430570698664,
+      "learning_rate": 4.999399520797532e-05,
+      "loss": 0.2089,
+      "step": 212
+    },
+    {
+      "epoch": 0.11087975013014055,
+      "grad_norm": 0.4605893449333839,
+      "learning_rate": 4.9993683335133535e-05,
+      "loss": 0.2023,
+      "step": 213
+    },
+    {
+      "epoch": 0.11140031233732431,
+      "grad_norm": 0.48657373710725793,
+      "learning_rate": 4.999336356812344e-05,
+      "loss": 0.2141,
+      "step": 214
+    },
+    {
+      "epoch": 0.11192087454450807,
+      "grad_norm": 0.5120193381886232,
+      "learning_rate": 4.9993035907046034e-05,
+      "loss": 0.2026,
+      "step": 215
+    },
+    {
+      "epoch": 0.11244143675169183,
+      "grad_norm": 0.49871162631911875,
+      "learning_rate": 4.999270035200483e-05,
+      "loss": 0.2207,
+      "step": 216
+    },
+    {
+      "epoch": 0.11296199895887558,
+      "grad_norm": 0.4917375628104701,
+      "learning_rate": 4.999235690310578e-05,
+      "loss": 0.2045,
+      "step": 217
+    },
+    {
+      "epoch": 0.11348256116605934,
+      "grad_norm": 0.5254314690487704,
+      "learning_rate": 4.999200556045739e-05,
+      "loss": 0.2068,
+      "step": 218
+    },
+    {
+      "epoch": 0.1140031233732431,
+      "grad_norm": 0.5379406712902184,
+      "learning_rate": 4.99916463241706e-05,
+      "loss": 0.1999,
+      "step": 219
+    },
+    {
+      "epoch": 0.11452368558042686,
+      "grad_norm": 0.5497570174948546,
+      "learning_rate": 4.99912791943589e-05,
+      "loss": 0.2174,
+      "step": 220
+    },
+    {
+      "epoch": 0.11504424778761062,
+      "grad_norm": 0.5159970185644596,
+      "learning_rate": 4.999090417113823e-05,
+      "loss": 0.2121,
+      "step": 221
+    },
+    {
+      "epoch": 0.11556480999479438,
+      "grad_norm": 0.5255410280335246,
+      "learning_rate": 4.999052125462705e-05,
+      "loss": 0.2005,
+      "step": 222
+    },
+    {
+      "epoch": 0.11608537220197813,
+      "grad_norm": 0.4370433291491418,
+      "learning_rate": 4.9990130444946295e-05,
+      "loss": 0.2062,
+      "step": 223
+    },
+    {
+      "epoch": 0.1166059344091619,
+      "grad_norm": 0.5368330251522332,
+      "learning_rate": 4.9989731742219415e-05,
+      "loss": 0.2053,
+      "step": 224
+    },
+    {
+      "epoch": 0.11712649661634565,
+      "grad_norm": 0.4997030944367229,
+      "learning_rate": 4.998932514657232e-05,
+      "loss": 0.2035,
+      "step": 225
+    },
+    {
+      "epoch": 0.11764705882352941,
+      "grad_norm": 0.48475596745975486,
+      "learning_rate": 4.9988910658133445e-05,
+      "loss": 0.2066,
+      "step": 226
+    },
+    {
+      "epoch": 0.11816762103071317,
+      "grad_norm": 0.4549147383812301,
+      "learning_rate": 4.99884882770337e-05,
+      "loss": 0.1947,
+      "step": 227
+    },
+    {
+      "epoch": 0.11868818323789693,
+      "grad_norm": 0.4829688354526709,
+      "learning_rate": 4.998805800340649e-05,
+      "loss": 0.2043,
+      "step": 228
+    },
+    {
+      "epoch": 0.11920874544508069,
+      "grad_norm": 0.49782219243743464,
+      "learning_rate": 4.998761983738772e-05,
+      "loss": 0.2051,
+      "step": 229
+    },
+    {
+      "epoch": 0.11972930765226444,
+      "grad_norm": 0.502253063495323,
+      "learning_rate": 4.998717377911578e-05,
+      "loss": 0.2169,
+      "step": 230
+    },
+    {
+      "epoch": 0.1202498698594482,
+      "grad_norm": 0.5061752506717145,
+      "learning_rate": 4.998671982873156e-05,
+      "loss": 0.2035,
+      "step": 231
+    },
+    {
+      "epoch": 0.12077043206663196,
+      "grad_norm": 0.511314601070742,
+      "learning_rate": 4.9986257986378434e-05,
+      "loss": 0.2034,
+      "step": 232
+    },
+    {
+      "epoch": 0.12129099427381572,
+      "grad_norm": 0.4988500109798501,
+      "learning_rate": 4.9985788252202284e-05,
+      "loss": 0.2007,
+      "step": 233
+    },
+    {
+      "epoch": 0.12181155648099948,
+      "grad_norm": 0.47625297037289466,
+      "learning_rate": 4.9985310626351453e-05,
+      "loss": 0.2065,
+      "step": 234
+    },
+    {
+      "epoch": 0.12233211868818324,
+      "grad_norm": 0.5281298227193022,
+      "learning_rate": 4.998482510897682e-05,
+      "loss": 0.2063,
+      "step": 235
+    },
+    {
+      "epoch": 0.122852680895367,
+      "grad_norm": 0.4762927778180656,
+      "learning_rate": 4.9984331700231716e-05,
+      "loss": 0.2091,
+      "step": 236
+    },
+    {
+      "epoch": 0.12337324310255075,
+      "grad_norm": 0.5474311482218877,
+      "learning_rate": 4.9983830400271995e-05,
+      "loss": 0.2061,
+      "step": 237
+    },
+    {
+      "epoch": 0.12389380530973451,
+      "grad_norm": 0.4774773588042615,
+      "learning_rate": 4.998332120925598e-05,
+      "loss": 0.2002,
+      "step": 238
+    },
+    {
+      "epoch": 0.12441436751691827,
+      "grad_norm": 0.48763373747314165,
+      "learning_rate": 4.9982804127344515e-05,
+      "loss": 0.2012,
+      "step": 239
+    },
+    {
+      "epoch": 0.12493492972410203,
+      "grad_norm": 0.5185578830596879,
+      "learning_rate": 4.9982279154700905e-05,
+      "loss": 0.2003,
+      "step": 240
+    },
+    {
+      "epoch": 0.1254554919312858,
+      "grad_norm": 0.5024445332518108,
+      "learning_rate": 4.9981746291490955e-05,
+      "loss": 0.1951,
+      "step": 241
+    },
+    {
+      "epoch": 0.12597605413846955,
+      "grad_norm": 0.5449525262128915,
+      "learning_rate": 4.998120553788298e-05,
+      "loss": 0.2032,
+      "step": 242
+    },
+    {
+      "epoch": 0.12649661634565332,
+      "grad_norm": 0.4665771668363775,
+      "learning_rate": 4.9980656894047776e-05,
+      "loss": 0.2039,
+      "step": 243
+    },
+    {
+      "epoch": 0.12701717855283706,
+      "grad_norm": 0.4372817573132417,
+      "learning_rate": 4.998010036015862e-05,
+      "loss": 0.2051,
+      "step": 244
+    },
+    {
+      "epoch": 0.12753774076002083,
+      "grad_norm": 0.43781674957374894,
+      "learning_rate": 4.997953593639129e-05,
+      "loss": 0.2054,
+      "step": 245
+    },
+    {
+      "epoch": 0.12805830296720458,
+      "grad_norm": 0.4541311939721181,
+      "learning_rate": 4.997896362292407e-05,
+      "loss": 0.1971,
+      "step": 246
+    },
+    {
+      "epoch": 0.12857886517438835,
+      "grad_norm": 0.4384524770223584,
+      "learning_rate": 4.997838341993772e-05,
+      "loss": 0.198,
+      "step": 247
+    },
+    {
+      "epoch": 0.1290994273815721,
+      "grad_norm": 0.4471256807532643,
+      "learning_rate": 4.997779532761549e-05,
+      "loss": 0.1986,
+      "step": 248
+    },
+    {
+      "epoch": 0.12961998958875587,
+      "grad_norm": 0.5163351668297698,
+      "learning_rate": 4.997719934614313e-05,
+      "loss": 0.2006,
+      "step": 249
+    },
+    {
+      "epoch": 0.1301405517959396,
+      "grad_norm": 0.5102778772697563,
+      "learning_rate": 4.9976595475708873e-05,
+      "loss": 0.2016,
+      "step": 250
+    },
+    {
+      "epoch": 0.13066111400312339,
+      "grad_norm": 0.4253852174219019,
+      "learning_rate": 4.997598371650346e-05,
+      "loss": 0.1965,
+      "step": 251
+    },
+    {
+      "epoch": 0.13118167621030713,
+      "grad_norm": 0.46645746563329804,
+      "learning_rate": 4.9975364068720106e-05,
+      "loss": 0.2031,
+      "step": 252
+    },
+    {
+      "epoch": 0.1317022384174909,
+      "grad_norm": 0.43965296000744936,
+      "learning_rate": 4.997473653255452e-05,
+      "loss": 0.1995,
+      "step": 253
+    },
+    {
+      "epoch": 0.13222280062467465,
+      "grad_norm": 0.44221288800503233,
+      "learning_rate": 4.997410110820492e-05,
+      "loss": 0.1982,
+      "step": 254
+    },
+    {
+      "epoch": 0.13274336283185842,
+      "grad_norm": 0.4370059916055007,
+      "learning_rate": 4.997345779587199e-05,
+      "loss": 0.1939,
+      "step": 255
+    },
+    {
+      "epoch": 0.13326392503904216,
+      "grad_norm": 0.5211107612902853,
+      "learning_rate": 4.997280659575892e-05,
+      "loss": 0.2026,
+      "step": 256
+    },
+    {
+      "epoch": 0.13378448724622594,
+      "grad_norm": 0.488605855839782,
+      "learning_rate": 4.997214750807141e-05,
+      "loss": 0.2005,
+      "step": 257
+    },
+    {
+      "epoch": 0.13430504945340968,
+      "grad_norm": 0.4932959365760645,
+      "learning_rate": 4.99714805330176e-05,
+      "loss": 0.2019,
+      "step": 258
+    },
+    {
+      "epoch": 0.13482561166059345,
+      "grad_norm": 0.5546717733269304,
+      "learning_rate": 4.997080567080817e-05,
+      "loss": 0.1948,
+      "step": 259
+    },
+    {
+      "epoch": 0.1353461738677772,
+      "grad_norm": 0.48295394900788036,
+      "learning_rate": 4.9970122921656257e-05,
+      "loss": 0.1955,
+      "step": 260
+    },
+    {
+      "epoch": 0.13586673607496097,
+      "grad_norm": 0.5105683280432859,
+      "learning_rate": 4.996943228577753e-05,
+      "loss": 0.2075,
+      "step": 261
+    },
+    {
+      "epoch": 0.13638729828214471,
+      "grad_norm": 0.5759006373431579,
+      "learning_rate": 4.996873376339011e-05,
+      "loss": 0.201,
+      "step": 262
+    },
+    {
+      "epoch": 0.1369078604893285,
+      "grad_norm": 0.483482317837608,
+      "learning_rate": 4.996802735471461e-05,
+      "loss": 0.1938,
+      "step": 263
+    },
+    {
+      "epoch": 0.13742842269651223,
+      "grad_norm": 0.48756843106965886,
+      "learning_rate": 4.996731305997416e-05,
+      "loss": 0.2,
+      "step": 264
+    },
+    {
+      "epoch": 0.137948984903696,
+      "grad_norm": 0.5231891365741146,
+      "learning_rate": 4.996659087939438e-05,
+      "loss": 0.2009,
+      "step": 265
+    },
+    {
+      "epoch": 0.13846954711087975,
+      "grad_norm": 0.518441009753121,
+      "learning_rate": 4.9965860813203345e-05,
+      "loss": 0.1831,
+      "step": 266
+    },
+    {
+      "epoch": 0.13899010931806352,
+      "grad_norm": 0.45224708236227923,
+      "learning_rate": 4.996512286163166e-05,
+      "loss": 0.2059,
+      "step": 267
+    },
+    {
+      "epoch": 0.13951067152524727,
+      "grad_norm": 0.5456839005168297,
+      "learning_rate": 4.99643770249124e-05,
+      "loss": 0.205,
+      "step": 268
+    },
+    {
+      "epoch": 0.14003123373243104,
+      "grad_norm": 0.4891910966189773,
+      "learning_rate": 4.996362330328113e-05,
+      "loss": 0.2,
+      "step": 269
+    },
+    {
+      "epoch": 0.14055179593961478,
+      "grad_norm": 0.5478174066545909,
+      "learning_rate": 4.996286169697591e-05,
+      "loss": 0.2004,
+      "step": 270
+    },
+    {
+      "epoch": 0.14107235814679855,
+      "grad_norm": 0.5582267970906812,
+      "learning_rate": 4.99620922062373e-05,
+      "loss": 0.203,
+      "step": 271
+    },
+    {
+      "epoch": 0.1415929203539823,
+      "grad_norm": 0.5105647459053501,
+      "learning_rate": 4.996131483130833e-05,
+      "loss": 0.2029,
+      "step": 272
+    },
+    {
+      "epoch": 0.14211348256116607,
+      "grad_norm": 0.48249484447771235,
+      "learning_rate": 4.9960529572434545e-05,
+      "loss": 0.2015,
+      "step": 273
+    },
+    {
+      "epoch": 0.14263404476834982,
+      "grad_norm": 0.5235774115905943,
+      "learning_rate": 4.995973642986395e-05,
+      "loss": 0.2036,
+      "step": 274
+    },
+    {
+      "epoch": 0.1431546069755336,
+      "grad_norm": 0.47923742023520904,
+      "learning_rate": 4.995893540384707e-05,
+      "loss": 0.2017,
+      "step": 275
+    },
+    {
+      "epoch": 0.14367516918271733,
+      "grad_norm": 0.43341131129494787,
+      "learning_rate": 4.99581264946369e-05,
+      "loss": 0.1995,
+      "step": 276
+    },
+    {
+      "epoch": 0.1441957313899011,
+      "grad_norm": 0.5535433768591606,
+      "learning_rate": 4.995730970248893e-05,
+      "loss": 0.2048,
+      "step": 277
+    },
+    {
+      "epoch": 0.14471629359708485,
+      "grad_norm": 0.3972052460404149,
+      "learning_rate": 4.9956485027661136e-05,
+      "loss": 0.2039,
+      "step": 278
+    },
+    {
+      "epoch": 0.14523685580426862,
+      "grad_norm": 0.594647280576585,
+      "learning_rate": 4.995565247041401e-05,
+      "loss": 0.1998,
+      "step": 279
+    },
+    {
+      "epoch": 0.14575741801145237,
+      "grad_norm": 0.4962198874500988,
+      "learning_rate": 4.995481203101049e-05,
+      "loss": 0.2018,
+      "step": 280
+    },
+    {
+      "epoch": 0.14627798021863614,
+      "grad_norm": 0.4301303739409537,
+      "learning_rate": 4.9953963709716034e-05,
+      "loss": 0.2053,
+      "step": 281
+    },
+    {
+      "epoch": 0.14679854242581988,
+      "grad_norm": 0.5069059225942892,
+      "learning_rate": 4.995310750679858e-05,
+      "loss": 0.1979,
+      "step": 282
+    },
+    {
+      "epoch": 0.14731910463300366,
+      "grad_norm": 0.5279838955451033,
+      "learning_rate": 4.995224342252855e-05,
+      "loss": 0.2017,
+      "step": 283
+    },
+    {
+      "epoch": 0.1478396668401874,
+      "grad_norm": 0.3970535952201932,
+      "learning_rate": 4.995137145717889e-05,
+      "loss": 0.2025,
+      "step": 284
+    },
+    {
+      "epoch": 0.14836022904737117,
+      "grad_norm": 0.4049155457791142,
+      "learning_rate": 4.9950491611024975e-05,
+      "loss": 0.1893,
+      "step": 285
+    },
+    {
+      "epoch": 0.14888079125455492,
+      "grad_norm": 0.4413501454399157,
+      "learning_rate": 4.994960388434471e-05,
+      "loss": 0.2012,
+      "step": 286
+    },
+    {
+      "epoch": 0.1494013534617387,
+      "grad_norm": 0.39118558952416227,
+      "learning_rate": 4.994870827741849e-05,
+      "loss": 0.194,
+      "step": 287
+    },
+    {
+      "epoch": 0.14992191566892243,
+      "grad_norm": 0.4209076336573591,
+      "learning_rate": 4.9947804790529176e-05,
+      "loss": 0.1923,
+      "step": 288
+    },
+    {
+      "epoch": 0.1504424778761062,
+      "grad_norm": 0.4149330323573577,
+      "learning_rate": 4.994689342396215e-05,
+      "loss": 0.1944,
+      "step": 289
+    },
+    {
+      "epoch": 0.15096304008328995,
+      "grad_norm": 0.41004710305075126,
+      "learning_rate": 4.994597417800524e-05,
+      "loss": 0.1993,
+      "step": 290
+    },
+    {
+      "epoch": 0.15148360229047372,
+      "grad_norm": 0.41295205841010674,
+      "learning_rate": 4.994504705294881e-05,
+      "loss": 0.2048,
+      "step": 291
+    },
+    {
+      "epoch": 0.15200416449765747,
+      "grad_norm": 0.3865464642966178,
+      "learning_rate": 4.994411204908567e-05,
+      "loss": 0.1952,
+      "step": 292
+    },
+    {
+      "epoch": 0.15252472670484124,
+      "grad_norm": 0.4033991442815395,
+      "learning_rate": 4.994316916671114e-05,
+      "loss": 0.1945,
+      "step": 293
+    },
+    {
+      "epoch": 0.15304528891202498,
+      "grad_norm": 0.37826488837927863,
+      "learning_rate": 4.9942218406123045e-05,
+      "loss": 0.1989,
+      "step": 294
+    },
+    {
+      "epoch": 0.15356585111920876,
+      "grad_norm": 0.3925013641735211,
+      "learning_rate": 4.994125976762167e-05,
+      "loss": 0.1948,
+      "step": 295
+    },
+    {
+      "epoch": 0.1540864133263925,
+      "grad_norm": 0.4297598640191415,
+      "learning_rate": 4.9940293251509786e-05,
+      "loss": 0.199,
+      "step": 296
+    },
+    {
+      "epoch": 0.15460697553357627,
+      "grad_norm": 0.39427507343176355,
+      "learning_rate": 4.9939318858092664e-05,
+      "loss": 0.2038,
+      "step": 297
+    },
+    {
+      "epoch": 0.15512753774076002,
+      "grad_norm": 0.41295108211994785,
+      "learning_rate": 4.993833658767808e-05,
+      "loss": 0.1855,
+      "step": 298
+    },
+    {
+      "epoch": 0.1556480999479438,
+      "grad_norm": 0.428783102563254,
+      "learning_rate": 4.993734644057627e-05,
+      "loss": 0.1925,
+      "step": 299
+    },
+    {
+      "epoch": 0.15616866215512754,
+      "grad_norm": 0.39643845709605496,
+      "learning_rate": 4.993634841709998e-05,
+      "loss": 0.1953,
+      "step": 300
+    },
+    {
+      "epoch": 0.1566892243623113,
+      "grad_norm": 0.40315780437097026,
+      "learning_rate": 4.993534251756441e-05,
+      "loss": 0.197,
+      "step": 301
+    },
+    {
+      "epoch": 0.15720978656949505,
+      "grad_norm": 0.42325725150557625,
+      "learning_rate": 4.9934328742287285e-05,
+      "loss": 0.1993,
+      "step": 302
+    },
+    {
+      "epoch": 0.15773034877667882,
+      "grad_norm": 0.4080909081022836,
+      "learning_rate": 4.9933307091588796e-05,
+      "loss": 0.1872,
+      "step": 303
+    },
+    {
+      "epoch": 0.15825091098386257,
+      "grad_norm": 0.4403757079324371,
+      "learning_rate": 4.993227756579163e-05,
+      "loss": 0.196,
+      "step": 304
+    },
+    {
+      "epoch": 0.15877147319104634,
+      "grad_norm": 0.4315054835172639,
+      "learning_rate": 4.993124016522097e-05,
+      "loss": 0.1988,
+      "step": 305
+    },
+    {
+      "epoch": 0.1592920353982301,
+      "grad_norm": 0.4181529747933808,
+      "learning_rate": 4.993019489020446e-05,
+      "loss": 0.1874,
+      "step": 306
+    },
+    {
+      "epoch": 0.15981259760541386,
+      "grad_norm": 0.38805878488629814,
+      "learning_rate": 4.992914174107225e-05,
+      "loss": 0.1885,
+      "step": 307
+    },
+    {
+      "epoch": 0.1603331598125976,
+      "grad_norm": 0.4442375624813517,
+      "learning_rate": 4.992808071815698e-05,
+      "loss": 0.1918,
+      "step": 308
+    },
+    {
+      "epoch": 0.16085372201978138,
+      "grad_norm": 0.3809208737314787,
+      "learning_rate": 4.9927011821793766e-05,
+      "loss": 0.1997,
+      "step": 309
+    },
+    {
+      "epoch": 0.16137428422696512,
+      "grad_norm": 0.4138404674084266,
+      "learning_rate": 4.9925935052320214e-05,
+      "loss": 0.1959,
+      "step": 310
+    },
+    {
+      "epoch": 0.1618948464341489,
+      "grad_norm": 0.41341096380173886,
+      "learning_rate": 4.9924850410076416e-05,
+      "loss": 0.2007,
+      "step": 311
+    },
+    {
+      "epoch": 0.16241540864133264,
+      "grad_norm": 0.43053394989910876,
+      "learning_rate": 4.9923757895404966e-05,
+      "loss": 0.1898,
+      "step": 312
+    },
+    {
+      "epoch": 0.1629359708485164,
+      "grad_norm": 0.40770200417265084,
+      "learning_rate": 4.992265750865091e-05,
+      "loss": 0.1961,
+      "step": 313
+    },
+    {
+      "epoch": 0.16345653305570015,
+      "grad_norm": 0.44481929282099114,
+      "learning_rate": 4.9921549250161817e-05,
+      "loss": 0.1903,
+      "step": 314
+    },
+    {
+      "epoch": 0.16397709526288393,
+      "grad_norm": 0.4214318597046138,
+      "learning_rate": 4.9920433120287726e-05,
+      "loss": 0.1978,
+      "step": 315
+    },
+    {
+      "epoch": 0.16449765747006767,
+      "grad_norm": 0.42351713283633147,
+      "learning_rate": 4.9919309119381155e-05,
+      "loss": 0.1913,
+      "step": 316
+    },
+    {
+      "epoch": 0.16501821967725144,
+      "grad_norm": 0.40073423145965437,
+      "learning_rate": 4.991817724779711e-05,
+      "loss": 0.201,
+      "step": 317
+    },
+    {
+      "epoch": 0.1655387818844352,
+      "grad_norm": 0.46802701541194947,
+      "learning_rate": 4.99170375058931e-05,
+      "loss": 0.1984,
+      "step": 318
+    },
+    {
+      "epoch": 0.16605934409161896,
+      "grad_norm": 0.4391832742508613,
+      "learning_rate": 4.9915889894029124e-05,
+      "loss": 0.1896,
+      "step": 319
+    },
+    {
+      "epoch": 0.1665799062988027,
+      "grad_norm": 0.4276877940360142,
+      "learning_rate": 4.991473441256762e-05,
+      "loss": 0.1883,
+      "step": 320
+    },
+    {
+      "epoch": 0.16710046850598648,
+      "grad_norm": 0.4783858273158986,
+      "learning_rate": 4.991357106187356e-05,
+      "loss": 0.1914,
+      "step": 321
+    },
+    {
+      "epoch": 0.16762103071317022,
+      "grad_norm": 0.37335690024180623,
+      "learning_rate": 4.991239984231438e-05,
+      "loss": 0.1861,
+      "step": 322
+    },
+    {
+      "epoch": 0.168141592920354,
+      "grad_norm": 0.5571380480881387,
+      "learning_rate": 4.991122075426001e-05,
+      "loss": 0.1938,
+      "step": 323
+    },
+    {
+      "epoch": 0.16866215512753774,
+      "grad_norm": 0.4673856871456706,
+      "learning_rate": 4.991003379808286e-05,
+      "loss": 0.1956,
+      "step": 324
+    },
+    {
+      "epoch": 0.1691827173347215,
+      "grad_norm": 0.6907331185852513,
+      "learning_rate": 4.990883897415781e-05,
+      "loss": 0.1946,
+      "step": 325
+    },
+    {
+      "epoch": 0.16970327954190526,
+      "grad_norm": 0.4420336300041284,
+      "learning_rate": 4.9907636282862256e-05,
+      "loss": 0.1917,
+      "step": 326
+    },
+    {
+      "epoch": 0.17022384174908903,
+      "grad_norm": 0.4673927210202748,
+      "learning_rate": 4.9906425724576075e-05,
+      "loss": 0.1868,
+      "step": 327
+    },
+    {
+      "epoch": 0.17074440395627277,
+      "grad_norm": 0.39164253998754844,
+      "learning_rate": 4.99052072996816e-05,
+      "loss": 0.1811,
+      "step": 328
+    },
+    {
+      "epoch": 0.17126496616345654,
+      "grad_norm": 0.44534231109604294,
+      "learning_rate": 4.990398100856367e-05,
+      "loss": 0.1938,
+      "step": 329
+    },
+    {
+      "epoch": 0.1717855283706403,
+      "grad_norm": 0.46887801231015935,
+      "learning_rate": 4.990274685160961e-05,
+      "loss": 0.1903,
+      "step": 330
+    },
+    {
+      "epoch": 0.17230609057782406,
+      "grad_norm": 0.3968509774573516,
+      "learning_rate": 4.990150482920921e-05,
+      "loss": 0.1908,
+      "step": 331
+    },
+    {
+      "epoch": 0.1728266527850078,
+      "grad_norm": 0.5263341947271295,
+      "learning_rate": 4.990025494175477e-05,
+      "loss": 0.1922,
+      "step": 332
+    },
+    {
+      "epoch": 0.17334721499219158,
+      "grad_norm": 0.4775537352423801,
+      "learning_rate": 4.989899718964107e-05,
+      "loss": 0.1841,
+      "step": 333
+    },
+    {
+      "epoch": 0.17386777719937532,
+      "grad_norm": 0.41924991923989763,
+      "learning_rate": 4.989773157326535e-05,
+      "loss": 0.2015,
+      "step": 334
+    },
+    {
+      "epoch": 0.1743883394065591,
+      "grad_norm": 0.5252465757639925,
+      "learning_rate": 4.989645809302736e-05,
+      "loss": 0.1917,
+      "step": 335
+    },
+    {
+      "epoch": 0.17490890161374284,
+      "grad_norm": 0.5169440105481551,
+      "learning_rate": 4.9895176749329334e-05,
+      "loss": 0.1917,
+      "step": 336
+    },
+    {
+      "epoch": 0.1754294638209266,
+      "grad_norm": 0.42246102114852874,
+      "learning_rate": 4.989388754257596e-05,
+      "loss": 0.2005,
+      "step": 337
+    },
+    {
+      "epoch": 0.17595002602811036,
+      "grad_norm": 0.5029332963068417,
+      "learning_rate": 4.989259047317444e-05,
+      "loss": 0.1903,
+      "step": 338
+    },
+    {
+      "epoch": 0.17647058823529413,
+      "grad_norm": 0.41379461152486186,
+      "learning_rate": 4.989128554153444e-05,
+      "loss": 0.1844,
+      "step": 339
+    },
+    {
+      "epoch": 0.17699115044247787,
+      "grad_norm": 0.45322329232876984,
+      "learning_rate": 4.9889972748068134e-05,
+      "loss": 0.188,
+      "step": 340
+    },
+    {
+      "epoch": 0.17751171264966165,
+      "grad_norm": 0.48168712385017154,
+      "learning_rate": 4.988865209319015e-05,
+      "loss": 0.1981,
+      "step": 341
+    },
+    {
+      "epoch": 0.1780322748568454,
+      "grad_norm": 0.41086331407455057,
+      "learning_rate": 4.988732357731762e-05,
+      "loss": 0.1942,
+      "step": 342
+    },
+    {
+      "epoch": 0.17855283706402916,
+      "grad_norm": 0.540497000624346,
+      "learning_rate": 4.988598720087015e-05,
+      "loss": 0.1907,
+      "step": 343
+    },
+    {
+      "epoch": 0.1790733992712129,
+      "grad_norm": 0.38794214622006457,
+      "learning_rate": 4.9884642964269824e-05,
+      "loss": 0.1874,
+      "step": 344
+    },
+    {
+      "epoch": 0.17959396147839668,
+      "grad_norm": 0.5258434187192316,
+      "learning_rate": 4.988329086794122e-05,
+      "loss": 0.1877,
+      "step": 345
+    },
+    {
+      "epoch": 0.18011452368558042,
+      "grad_norm": 0.5353658539595278,
+      "learning_rate": 4.9881930912311394e-05,
+      "loss": 0.194,
+      "step": 346
+    },
+    {
+      "epoch": 0.1806350858927642,
+      "grad_norm": 0.4392015964672634,
+      "learning_rate": 4.988056309780987e-05,
+      "loss": 0.1904,
+      "step": 347
+    },
+    {
+      "epoch": 0.18115564809994794,
+      "grad_norm": 0.5430979821127188,
+      "learning_rate": 4.987918742486869e-05,
+      "loss": 0.2045,
+      "step": 348
+    },
+    {
+      "epoch": 0.1816762103071317,
+      "grad_norm": 0.47746127042266084,
+      "learning_rate": 4.987780389392234e-05,
+      "loss": 0.1877,
+      "step": 349
+    },
+    {
+      "epoch": 0.18219677251431546,
+      "grad_norm": 0.48616627754776087,
+      "learning_rate": 4.98764125054078e-05,
+      "loss": 0.1856,
+      "step": 350
+    },
+    {
+      "epoch": 0.18271733472149923,
+      "grad_norm": 0.4137932189199551,
+      "learning_rate": 4.987501325976455e-05,
+      "loss": 0.1882,
+      "step": 351
+    },
+    {
+      "epoch": 0.18323789692868298,
+      "grad_norm": 0.48916635581424217,
+      "learning_rate": 4.987360615743453e-05,
+      "loss": 0.1872,
+      "step": 352
+    },
+    {
+      "epoch": 0.18375845913586675,
+      "grad_norm": 0.45140194045768417,
+      "learning_rate": 4.987219119886216e-05,
+      "loss": 0.1974,
+      "step": 353
+    },
+    {
+      "epoch": 0.1842790213430505,
+      "grad_norm": 0.4683554765097198,
+      "learning_rate": 4.987076838449436e-05,
+      "loss": 0.1887,
+      "step": 354
+    },
+    {
+      "epoch": 0.18479958355023426,
+      "grad_norm": 0.45043420134623535,
+      "learning_rate": 4.986933771478052e-05,
+      "loss": 0.1953,
+      "step": 355
+    },
+    {
+      "epoch": 0.185320145757418,
+      "grad_norm": 0.38632079781772455,
+      "learning_rate": 4.9867899190172505e-05,
+      "loss": 0.1944,
+      "step": 356
+    },
+    {
+      "epoch": 0.18584070796460178,
+      "grad_norm": 0.4502927446904294,
+      "learning_rate": 4.986645281112469e-05,
+      "loss": 0.1993,
+      "step": 357
+    },
+    {
+      "epoch": 0.18636127017178553,
+      "grad_norm": 0.35109988684784443,
+      "learning_rate": 4.986499857809387e-05,
+      "loss": 0.1832,
+      "step": 358
+    },
+    {
+      "epoch": 0.1868818323789693,
+      "grad_norm": 0.4855353635745786,
+      "learning_rate": 4.98635364915394e-05,
+      "loss": 0.1928,
+      "step": 359
+    },
+    {
+      "epoch": 0.18740239458615304,
+      "grad_norm": 0.3907280285159689,
+      "learning_rate": 4.986206655192305e-05,
+      "loss": 0.1876,
+      "step": 360
+    },
+    {
+      "epoch": 0.18792295679333682,
+      "grad_norm": 0.42227390242070895,
+      "learning_rate": 4.98605887597091e-05,
+      "loss": 0.193,
+      "step": 361
+    },
+    {
+      "epoch": 0.18844351900052056,
+      "grad_norm": 0.39584572126339823,
+      "learning_rate": 4.985910311536431e-05,
+      "loss": 0.184,
+      "step": 362
+    },
+    {
+      "epoch": 0.18896408120770433,
+      "grad_norm": 0.3990052626089986,
+      "learning_rate": 4.985760961935791e-05,
+      "loss": 0.1881,
+      "step": 363
+    },
+    {
+      "epoch": 0.18948464341488808,
+      "grad_norm": 0.42636340213787094,
+      "learning_rate": 4.9856108272161614e-05,
+      "loss": 0.1926,
+      "step": 364
+    },
+    {
+      "epoch": 0.19000520562207185,
+      "grad_norm": 0.42577564707501814,
+      "learning_rate": 4.9854599074249633e-05,
+      "loss": 0.195,
+      "step": 365
+    },
+    {
+      "epoch": 0.1905257678292556,
+      "grad_norm": 0.41616380256414365,
+      "learning_rate": 4.985308202609863e-05,
+      "loss": 0.1885,
+      "step": 366
+    },
+    {
+      "epoch": 0.19104633003643937,
+      "grad_norm": 0.46495409518003894,
+      "learning_rate": 4.9851557128187755e-05,
+      "loss": 0.1908,
+      "step": 367
+    },
+    {
+      "epoch": 0.1915668922436231,
+      "grad_norm": 0.41526376329857984,
+      "learning_rate": 4.985002438099865e-05,
+      "loss": 0.188,
+      "step": 368
+    },
+    {
+      "epoch": 0.19208745445080688,
+      "grad_norm": 0.3935066236122138,
+      "learning_rate": 4.984848378501542e-05,
+      "loss": 0.1885,
+      "step": 369
+    },
+    {
+      "epoch": 0.19260801665799063,
+      "grad_norm": 0.4088622746642285,
+      "learning_rate": 4.984693534072467e-05,
+      "loss": 0.1918,
+      "step": 370
+    },
+    {
+      "epoch": 0.1931285788651744,
+      "grad_norm": 0.3809983710951146,
+      "learning_rate": 4.984537904861546e-05,
+      "loss": 0.1826,
+      "step": 371
+    },
+    {
+      "epoch": 0.19364914107235814,
+      "grad_norm": 0.40658651643232196,
+      "learning_rate": 4.9843814909179345e-05,
+      "loss": 0.19,
+      "step": 372
+    },
+    {
+      "epoch": 0.19416970327954192,
+      "grad_norm": 0.43969698475310975,
+      "learning_rate": 4.9842242922910345e-05,
+      "loss": 0.1894,
+      "step": 373
+    },
+    {
+      "epoch": 0.19469026548672566,
+      "grad_norm": 0.3862559796395953,
+      "learning_rate": 4.9840663090304965e-05,
+      "loss": 0.1914,
+      "step": 374
+    },
+    {
+      "epoch": 0.19521082769390943,
+      "grad_norm": 0.5347847427366343,
+      "learning_rate": 4.983907541186221e-05,
+      "loss": 0.1917,
+      "step": 375
+    },
+    {
+      "epoch": 0.19573138990109318,
+      "grad_norm": 0.3974569413531495,
+      "learning_rate": 4.983747988808352e-05,
+      "loss": 0.184,
+      "step": 376
+    },
+    {
+      "epoch": 0.19625195210827695,
+      "grad_norm": 0.4412640683782702,
+      "learning_rate": 4.983587651947285e-05,
+      "loss": 0.1855,
+      "step": 377
+    },
+    {
+      "epoch": 0.1967725143154607,
+      "grad_norm": 0.4423096342665371,
+      "learning_rate": 4.983426530653661e-05,
+      "loss": 0.1923,
+      "step": 378
+    },
+    {
+      "epoch": 0.19729307652264447,
+      "grad_norm": 0.43821702301793497,
+      "learning_rate": 4.9832646249783694e-05,
+      "loss": 0.189,
+      "step": 379
+    },
+    {
+      "epoch": 0.1978136387298282,
+      "grad_norm": 0.3924545030570308,
+      "learning_rate": 4.983101934972548e-05,
+      "loss": 0.1891,
+      "step": 380
+    },
+    {
+      "epoch": 0.19833420093701198,
+      "grad_norm": 0.42299618371561903,
+      "learning_rate": 4.982938460687583e-05,
+      "loss": 0.1903,
+      "step": 381
+    },
+    {
+      "epoch": 0.19885476314419573,
+      "grad_norm": 0.4001882492331783,
+      "learning_rate": 4.982774202175105e-05,
+      "loss": 0.1803,
+      "step": 382
+    },
+    {
+      "epoch": 0.1993753253513795,
+      "grad_norm": 0.49011114003395145,
+      "learning_rate": 4.9826091594869974e-05,
+      "loss": 0.1813,
+      "step": 383
+    },
+    {
+      "epoch": 0.19989588755856325,
+      "grad_norm": 0.37442432473888976,
+      "learning_rate": 4.982443332675385e-05,
+      "loss": 0.1802,
+      "step": 384
+    },
+    {
+      "epoch": 0.20041644976574702,
+      "grad_norm": 0.48051842908306147,
+      "learning_rate": 4.9822767217926456e-05,
+      "loss": 0.1947,
+      "step": 385
+    },
+    {
+      "epoch": 0.20093701197293076,
+      "grad_norm": 0.4257153401498972,
+      "learning_rate": 4.982109326891402e-05,
+      "loss": 0.1884,
+      "step": 386
+    },
+    {
+      "epoch": 0.20145757418011453,
+      "grad_norm": 0.47752005156773103,
+      "learning_rate": 4.981941148024526e-05,
+      "loss": 0.1857,
+      "step": 387
+    },
+    {
+      "epoch": 0.20197813638729828,
+      "grad_norm": 0.40745960509516727,
+      "learning_rate": 4.981772185245135e-05,
+      "loss": 0.187,
+      "step": 388
+    },
+    {
+      "epoch": 0.20249869859448205,
+      "grad_norm": 0.4359497459483064,
+      "learning_rate": 4.9816024386065973e-05,
+      "loss": 0.1871,
+      "step": 389
+    },
+    {
+      "epoch": 0.2030192608016658,
+      "grad_norm": 0.3892740872530919,
+      "learning_rate": 4.981431908162525e-05,
+      "loss": 0.188,
+      "step": 390
+    },
+    {
+      "epoch": 0.20353982300884957,
+      "grad_norm": 0.420465792965565,
+      "learning_rate": 4.98126059396678e-05,
+      "loss": 0.1936,
+      "step": 391
+    },
+    {
+      "epoch": 0.2040603852160333,
+      "grad_norm": 0.3773127318291861,
+      "learning_rate": 4.981088496073472e-05,
+      "loss": 0.1875,
+      "step": 392
+    },
+    {
+      "epoch": 0.20458094742321709,
+      "grad_norm": 0.4041436379650743,
+      "learning_rate": 4.980915614536957e-05,
+      "loss": 0.1915,
+      "step": 393
+    },
+    {
+      "epoch": 0.20510150963040083,
+      "grad_norm": 0.3808004928284413,
+      "learning_rate": 4.980741949411839e-05,
+      "loss": 0.1886,
+      "step": 394
+    },
+    {
+      "epoch": 0.2056220718375846,
+      "grad_norm": 0.39329042034751993,
+      "learning_rate": 4.98056750075297e-05,
+      "loss": 0.189,
+      "step": 395
+    },
+    {
+      "epoch": 0.20614263404476835,
+      "grad_norm": 0.3950097377987076,
+      "learning_rate": 4.980392268615447e-05,
+      "loss": 0.1922,
+      "step": 396
+    },
+    {
+      "epoch": 0.20666319625195212,
+      "grad_norm": 0.3768193415771947,
+      "learning_rate": 4.980216253054619e-05,
+      "loss": 0.1863,
+      "step": 397
+    },
+    {
+      "epoch": 0.20718375845913586,
+      "grad_norm": 0.40683290447145076,
+      "learning_rate": 4.98003945412608e-05,
+      "loss": 0.1888,
+      "step": 398
+    },
+    {
+      "epoch": 0.20770432066631964,
+      "grad_norm": 0.3547903108580832,
+      "learning_rate": 4.979861871885669e-05,
+      "loss": 0.1916,
+      "step": 399
+    },
+    {
+      "epoch": 0.20822488287350338,
+      "grad_norm": 0.36494431710070346,
+      "learning_rate": 4.9796835063894765e-05,
+      "loss": 0.1806,
+      "step": 400
+    },
+    {
+      "epoch": 0.20874544508068715,
+      "grad_norm": 0.36485565181417856,
+      "learning_rate": 4.9795043576938384e-05,
+      "loss": 0.1833,
+      "step": 401
+    },
+    {
+      "epoch": 0.2092660072878709,
+      "grad_norm": 0.3632899653463087,
+      "learning_rate": 4.9793244258553375e-05,
+      "loss": 0.185,
+      "step": 402
+    },
+    {
+      "epoch": 0.20978656949505467,
+      "grad_norm": 0.35512187990656785,
+      "learning_rate": 4.979143710930805e-05,
+      "loss": 0.1817,
+      "step": 403
+    },
+    {
+      "epoch": 0.21030713170223841,
+      "grad_norm": 0.4054160927254296,
+      "learning_rate": 4.9789622129773195e-05,
+      "loss": 0.1807,
+      "step": 404
+    },
+    {
+      "epoch": 0.2108276939094222,
+      "grad_norm": 0.348222334286123,
+      "learning_rate": 4.978779932052206e-05,
+      "loss": 0.1802,
+      "step": 405
+    },
+    {
+      "epoch": 0.21134825611660593,
+      "grad_norm": 0.35217898427995553,
+      "learning_rate": 4.978596868213037e-05,
+      "loss": 0.1876,
+      "step": 406
+    },
+    {
+      "epoch": 0.2118688183237897,
+      "grad_norm": 0.3399990298458504,
+      "learning_rate": 4.978413021517634e-05,
+      "loss": 0.1824,
+      "step": 407
+    },
+    {
+      "epoch": 0.21238938053097345,
+      "grad_norm": 0.3914421983501113,
+      "learning_rate": 4.978228392024063e-05,
+      "loss": 0.1832,
+      "step": 408
+    },
+    {
+      "epoch": 0.21290994273815722,
+      "grad_norm": 0.3599097944411457,
+      "learning_rate": 4.978042979790639e-05,
+      "loss": 0.1857,
+      "step": 409
+    },
+    {
+      "epoch": 0.21343050494534097,
+      "grad_norm": 0.3764468760970812,
+      "learning_rate": 4.977856784875924e-05,
+      "loss": 0.1874,
+      "step": 410
+    },
+    {
+      "epoch": 0.21395106715252474,
+      "grad_norm": 0.36370332252391396,
+      "learning_rate": 4.977669807338726e-05,
+      "loss": 0.1909,
+      "step": 411
+    },
+    {
+      "epoch": 0.21447162935970848,
+      "grad_norm": 0.3590343415362341,
+      "learning_rate": 4.9774820472381037e-05,
+      "loss": 0.1808,
+      "step": 412
+    },
+    {
+      "epoch": 0.21499219156689225,
+      "grad_norm": 0.3964803489327254,
+      "learning_rate": 4.977293504633357e-05,
+      "loss": 0.1993,
+      "step": 413
+    },
+    {
+      "epoch": 0.215512753774076,
+      "grad_norm": 0.409647084566613,
+      "learning_rate": 4.977104179584039e-05,
+      "loss": 0.1828,
+      "step": 414
+    },
+    {
+      "epoch": 0.21603331598125977,
+      "grad_norm": 0.37684217160913236,
+      "learning_rate": 4.9769140721499466e-05,
+      "loss": 0.1903,
+      "step": 415
+    },
+    {
+      "epoch": 0.21655387818844352,
+      "grad_norm": 0.40627101878375,
+      "learning_rate": 4.976723182391124e-05,
+      "loss": 0.1947,
+      "step": 416
+    },
+    {
+      "epoch": 0.2170744403956273,
+      "grad_norm": 0.4293268173755893,
+      "learning_rate": 4.9765315103678646e-05,
+      "loss": 0.1928,
+      "step": 417
+    },
+    {
+      "epoch": 0.21759500260281103,
+      "grad_norm": 0.363143103690957,
+      "learning_rate": 4.976339056140706e-05,
+      "loss": 0.1859,
+      "step": 418
+    },
+    {
+      "epoch": 0.2181155648099948,
+      "grad_norm": 0.3801901908980379,
+      "learning_rate": 4.976145819770435e-05,
+      "loss": 0.1839,
+      "step": 419
+    },
+    {
+      "epoch": 0.21863612701717855,
+      "grad_norm": 0.3728199137683034,
+      "learning_rate": 4.975951801318083e-05,
+      "loss": 0.1836,
+      "step": 420
+    },
+    {
+      "epoch": 0.21915668922436232,
+      "grad_norm": 0.39399419625104315,
+      "learning_rate": 4.975757000844932e-05,
+      "loss": 0.1869,
+      "step": 421
+    },
+    {
+      "epoch": 0.21967725143154607,
+      "grad_norm": 0.352068817370931,
+      "learning_rate": 4.975561418412509e-05,
+      "loss": 0.1836,
+      "step": 422
+    },
+    {
+      "epoch": 0.22019781363872984,
+      "grad_norm": 0.38356796463195586,
+      "learning_rate": 4.9753650540825855e-05,
+      "loss": 0.1813,
+      "step": 423
+    },
+    {
+      "epoch": 0.22071837584591358,
+      "grad_norm": 0.3610859855606606,
+      "learning_rate": 4.975167907917187e-05,
+      "loss": 0.1862,
+      "step": 424
+    },
+    {
+      "epoch": 0.22123893805309736,
+      "grad_norm": 0.35330258565992684,
+      "learning_rate": 4.974969979978577e-05,
+      "loss": 0.1769,
+      "step": 425
+    },
+    {
+      "epoch": 0.2217595002602811,
+      "grad_norm": 0.384329198269658,
+      "learning_rate": 4.9747712703292714e-05,
+      "loss": 0.1807,
+      "step": 426
+    },
+    {
+      "epoch": 0.22228006246746487,
+      "grad_norm": 0.34746642781914455,
+      "learning_rate": 4.9745717790320344e-05,
+      "loss": 0.1809,
+      "step": 427
+    },
+    {
+      "epoch": 0.22280062467464862,
+      "grad_norm": 0.3880165680244598,
+      "learning_rate": 4.9743715061498716e-05,
+      "loss": 0.1867,
+      "step": 428
+    },
+    {
+      "epoch": 0.2233211868818324,
+      "grad_norm": 0.351470867292313,
+      "learning_rate": 4.9741704517460406e-05,
+      "loss": 0.1894,
+      "step": 429
+    },
+    {
+      "epoch": 0.22384174908901613,
+      "grad_norm": 0.3645151488001888,
+      "learning_rate": 4.973968615884043e-05,
+      "loss": 0.1788,
+      "step": 430
+    },
+    {
+      "epoch": 0.2243623112961999,
+      "grad_norm": 0.3625527069125457,
+      "learning_rate": 4.973765998627628e-05,
+      "loss": 0.1914,
+      "step": 431
+    },
+    {
+      "epoch": 0.22488287350338365,
+      "grad_norm": 0.343361514535999,
+      "learning_rate": 4.973562600040791e-05,
+      "loss": 0.1848,
+      "step": 432
+    },
+    {
+      "epoch": 0.22540343571056742,
+      "grad_norm": 0.37503966422043744,
+      "learning_rate": 4.973358420187776e-05,
+      "loss": 0.1879,
+      "step": 433
+    },
+    {
+      "epoch": 0.22592399791775117,
+      "grad_norm": 0.3474973215881361,
+      "learning_rate": 4.973153459133071e-05,
+      "loss": 0.1894,
+      "step": 434
+    },
+    {
+      "epoch": 0.22644456012493494,
+      "grad_norm": 0.3508809916689539,
+      "learning_rate": 4.972947716941413e-05,
+      "loss": 0.1895,
+      "step": 435
+    },
+    {
+      "epoch": 0.22696512233211869,
+      "grad_norm": 0.3643384265149809,
+      "learning_rate": 4.9727411936777854e-05,
+      "loss": 0.1829,
+      "step": 436
+    },
+    {
+      "epoch": 0.22748568453930246,
+      "grad_norm": 0.36350332295792714,
+      "learning_rate": 4.972533889407417e-05,
+      "loss": 0.1813,
+      "step": 437
+    },
+    {
+      "epoch": 0.2280062467464862,
+      "grad_norm": 0.3539689125746054,
+      "learning_rate": 4.972325804195784e-05,
+      "loss": 0.1842,
+      "step": 438
+    },
+    {
+      "epoch": 0.22852680895366997,
+      "grad_norm": 0.3822911141420241,
+      "learning_rate": 4.972116938108611e-05,
+      "loss": 0.193,
+      "step": 439
+    },
+    {
+      "epoch": 0.22904737116085372,
+      "grad_norm": 0.38773043508499666,
+      "learning_rate": 4.971907291211866e-05,
+      "loss": 0.188,
+      "step": 440
+    },
+    {
+      "epoch": 0.2295679333680375,
+      "grad_norm": 0.35091644126465377,
+      "learning_rate": 4.971696863571765e-05,
+      "loss": 0.1824,
+      "step": 441
+    },
+    {
+      "epoch": 0.23008849557522124,
+      "grad_norm": 0.34427733124831933,
+      "learning_rate": 4.971485655254773e-05,
+      "loss": 0.175,
+      "step": 442
+    },
+    {
+      "epoch": 0.230609057782405,
+      "grad_norm": 0.3543495011979818,
+      "learning_rate": 4.9712736663275974e-05,
+      "loss": 0.1849,
+      "step": 443
+    },
+    {
+      "epoch": 0.23112961998958875,
+      "grad_norm": 0.36206158086227636,
+      "learning_rate": 4.971060896857195e-05,
+      "loss": 0.1891,
+      "step": 444
+    },
+    {
+      "epoch": 0.23165018219677252,
+      "grad_norm": 0.3599367526208678,
+      "learning_rate": 4.9708473469107676e-05,
+      "loss": 0.1853,
+      "step": 445
+    },
+    {
+      "epoch": 0.23217074440395627,
+      "grad_norm": 0.34979470112428146,
+      "learning_rate": 4.970633016555765e-05,
+      "loss": 0.1761,
+      "step": 446
+    },
+    {
+      "epoch": 0.23269130661114004,
+      "grad_norm": 0.4092021184005037,
+      "learning_rate": 4.9704179058598824e-05,
+      "loss": 0.1842,
+      "step": 447
+    },
+    {
+      "epoch": 0.2332118688183238,
+      "grad_norm": 0.3745830889783029,
+      "learning_rate": 4.970202014891062e-05,
+      "loss": 0.1854,
+      "step": 448
+    },
+    {
+      "epoch": 0.23373243102550756,
+      "grad_norm": 0.3787251823910667,
+      "learning_rate": 4.969985343717492e-05,
+      "loss": 0.1882,
+      "step": 449
+    },
+    {
+      "epoch": 0.2342529932326913,
+      "grad_norm": 0.3986619843101879,
+      "learning_rate": 4.9697678924076066e-05,
+      "loss": 0.1856,
+      "step": 450
+    },
+    {
+      "epoch": 0.23477355543987508,
+      "grad_norm": 0.3430314984382981,
+      "learning_rate": 4.969549661030089e-05,
+      "loss": 0.1839,
+      "step": 451
+    },
+    {
+      "epoch": 0.23529411764705882,
+      "grad_norm": 0.3621352775817525,
+      "learning_rate": 4.969330649653864e-05,
+      "loss": 0.1848,
+      "step": 452
+    },
+    {
+      "epoch": 0.2358146798542426,
+      "grad_norm": 0.35888862276629535,
+      "learning_rate": 4.969110858348108e-05,
+      "loss": 0.1874,
+      "step": 453
+    },
+    {
+      "epoch": 0.23633524206142634,
+      "grad_norm": 0.3942881061832858,
+      "learning_rate": 4.96889028718224e-05,
+      "loss": 0.178,
+      "step": 454
+    },
+    {
+      "epoch": 0.2368558042686101,
+      "grad_norm": 0.33338571503996683,
+      "learning_rate": 4.968668936225928e-05,
+      "loss": 0.1915,
+      "step": 455
+    },
+    {
+      "epoch": 0.23737636647579385,
+      "grad_norm": 0.3789691197396563,
+      "learning_rate": 4.968446805549082e-05,
+      "loss": 0.1814,
+      "step": 456
+    },
+    {
+      "epoch": 0.23789692868297763,
+      "grad_norm": 0.36459554389496945,
+      "learning_rate": 4.968223895221865e-05,
+      "loss": 0.1874,
+      "step": 457
+    },
+    {
+      "epoch": 0.23841749089016137,
+      "grad_norm": 0.40802558988494547,
+      "learning_rate": 4.96800020531468e-05,
+      "loss": 0.1876,
+      "step": 458
+    },
+    {
+      "epoch": 0.23893805309734514,
+      "grad_norm": 0.35893156732538567,
+      "learning_rate": 4.967775735898179e-05,
+      "loss": 0.1825,
+      "step": 459
+    },
+    {
+      "epoch": 0.2394586153045289,
+      "grad_norm": 0.3651439723126297,
+      "learning_rate": 4.967550487043261e-05,
+      "loss": 0.1853,
+      "step": 460
+    },
+    {
+      "epoch": 0.23997917751171266,
+      "grad_norm": 0.3624747101403952,
+      "learning_rate": 4.9673244588210684e-05,
+      "loss": 0.1783,
+      "step": 461
+    },
+    {
+      "epoch": 0.2404997397188964,
+      "grad_norm": 0.3360578895946992,
+      "learning_rate": 4.967097651302993e-05,
+      "loss": 0.1786,
+      "step": 462
+    },
+    {
+      "epoch": 0.24102030192608018,
+      "grad_norm": 0.3929754036367695,
+      "learning_rate": 4.9668700645606704e-05,
+      "loss": 0.188,
+      "step": 463
+    },
+    {
+      "epoch": 0.24154086413326392,
+      "grad_norm": 0.34507860442461075,
+      "learning_rate": 4.966641698665982e-05,
+      "loss": 0.183,
+      "step": 464
+    },
+    {
+      "epoch": 0.2420614263404477,
+      "grad_norm": 0.363947787934868,
+      "learning_rate": 4.9664125536910597e-05,
+      "loss": 0.179,
+      "step": 465
+    },
+    {
+      "epoch": 0.24258198854763144,
+      "grad_norm": 0.3444630195014257,
+      "learning_rate": 4.966182629708275e-05,
+      "loss": 0.1815,
+      "step": 466
+    },
+    {
+      "epoch": 0.2431025507548152,
+      "grad_norm": 0.35943929861594537,
+      "learning_rate": 4.965951926790249e-05,
+      "loss": 0.1873,
+      "step": 467
+    },
+    {
+      "epoch": 0.24362311296199896,
+      "grad_norm": 0.3629355287385724,
+      "learning_rate": 4.9657204450098496e-05,
+      "loss": 0.1857,
+      "step": 468
+    },
+    {
+      "epoch": 0.24414367516918273,
+      "grad_norm": 0.32513099997931477,
+      "learning_rate": 4.9654881844401886e-05,
+      "loss": 0.1845,
+      "step": 469
+    },
+    {
+      "epoch": 0.24466423737636647,
+      "grad_norm": 0.375672815309898,
+      "learning_rate": 4.965255145154625e-05,
+      "loss": 0.1825,
+      "step": 470
+    },
+    {
+      "epoch": 0.24518479958355024,
+      "grad_norm": 0.32902747038638014,
+      "learning_rate": 4.965021327226764e-05,
+      "loss": 0.1831,
+      "step": 471
+    },
+    {
+      "epoch": 0.245705361790734,
+      "grad_norm": 0.3723575912403685,
+      "learning_rate": 4.964786730730455e-05,
+      "loss": 0.1828,
+      "step": 472
+    },
+    {
+      "epoch": 0.24622592399791776,
+      "grad_norm": 0.3291696338939864,
+      "learning_rate": 4.964551355739795e-05,
+      "loss": 0.1825,
+      "step": 473
+    },
+    {
+      "epoch": 0.2467464862051015,
+      "grad_norm": 0.40037826390407993,
+      "learning_rate": 4.964315202329127e-05,
+      "loss": 0.1948,
+      "step": 474
+    },
+    {
+      "epoch": 0.24726704841228528,
+      "grad_norm": 0.3980798213979412,
+      "learning_rate": 4.9640782705730394e-05,
+      "loss": 0.1818,
+      "step": 475
+    },
+    {
+      "epoch": 0.24778761061946902,
+      "grad_norm": 0.3307799216619339,
+      "learning_rate": 4.9638405605463646e-05,
+      "loss": 0.177,
+      "step": 476
+    },
+    {
+      "epoch": 0.2483081728266528,
+      "grad_norm": 0.37858138177738815,
+      "learning_rate": 4.963602072324184e-05,
+      "loss": 0.1829,
+      "step": 477
+    },
+    {
+      "epoch": 0.24882873503383654,
+      "grad_norm": 0.34060237598449605,
+      "learning_rate": 4.963362805981823e-05,
+      "loss": 0.1872,
+      "step": 478
+    },
+    {
+      "epoch": 0.2493492972410203,
+      "grad_norm": 0.3491413889386904,
+      "learning_rate": 4.9631227615948516e-05,
+      "loss": 0.1719,
+      "step": 479
+    },
+    {
+      "epoch": 0.24986985944820406,
+      "grad_norm": 0.3693858175983402,
+      "learning_rate": 4.962881939239089e-05,
+      "loss": 0.1855,
+      "step": 480
+    },
+    {
+      "epoch": 0.25039042165538783,
+      "grad_norm": 0.329979856271684,
+      "learning_rate": 4.962640338990598e-05,
+      "loss": 0.1885,
+      "step": 481
+    },
+    {
+      "epoch": 0.2509109838625716,
+      "grad_norm": 0.3796178436862751,
+      "learning_rate": 4.962397960925686e-05,
+      "loss": 0.1896,
+      "step": 482
+    },
+    {
+      "epoch": 0.2514315460697553,
+      "grad_norm": 0.34129494053229703,
+      "learning_rate": 4.9621548051209075e-05,
+      "loss": 0.1772,
+      "step": 483
+    },
+    {
+      "epoch": 0.2519521082769391,
+      "grad_norm": 0.33923486612711795,
+      "learning_rate": 4.961910871653063e-05,
+      "loss": 0.1809,
+      "step": 484
+    },
+    {
+      "epoch": 0.25247267048412286,
+      "grad_norm": 0.32895841536481185,
+      "learning_rate": 4.961666160599198e-05,
+      "loss": 0.1801,
+      "step": 485
+    },
+    {
+      "epoch": 0.25299323269130664,
+      "grad_norm": 0.34829345184997373,
+      "learning_rate": 4.961420672036603e-05,
+      "loss": 0.1783,
+      "step": 486
+    },
+    {
+      "epoch": 0.25351379489849035,
+      "grad_norm": 0.36383868089193916,
+      "learning_rate": 4.9611744060428156e-05,
+      "loss": 0.1822,
+      "step": 487
+    },
+    {
+      "epoch": 0.2540343571056741,
+      "grad_norm": 0.36477630270026584,
+      "learning_rate": 4.960927362695617e-05,
+      "loss": 0.1808,
+      "step": 488
+    },
+    {
+      "epoch": 0.2545549193128579,
+      "grad_norm": 0.39127337500191006,
+      "learning_rate": 4.960679542073036e-05,
+      "loss": 0.1889,
+      "step": 489
+    },
+    {
+      "epoch": 0.25507548152004167,
+      "grad_norm": 0.364605809032773,
+      "learning_rate": 4.9604309442533454e-05,
+      "loss": 0.1837,
+      "step": 490
+    },
+    {
+      "epoch": 0.2555960437272254,
+      "grad_norm": 0.387563172322994,
+      "learning_rate": 4.960181569315064e-05,
+      "loss": 0.1813,
+      "step": 491
+    },
+    {
+      "epoch": 0.25611660593440916,
+      "grad_norm": 0.3574697533499239,
+      "learning_rate": 4.959931417336956e-05,
+      "loss": 0.1778,
+      "step": 492
+    },
+    {
+      "epoch": 0.25663716814159293,
+      "grad_norm": 0.3772487104771915,
+      "learning_rate": 4.959680488398031e-05,
+      "loss": 0.1852,
+      "step": 493
+    },
+    {
+      "epoch": 0.2571577303487767,
+      "grad_norm": 0.37856588039150824,
+      "learning_rate": 4.959428782577544e-05,
+      "loss": 0.1852,
+      "step": 494
+    },
+    {
+      "epoch": 0.2576782925559604,
+      "grad_norm": 0.3473512333204632,
+      "learning_rate": 4.959176299954995e-05,
+      "loss": 0.1883,
+      "step": 495
+    },
+    {
+      "epoch": 0.2581988547631442,
+      "grad_norm": 0.3717884159305744,
+      "learning_rate": 4.958923040610132e-05,
+      "loss": 0.1901,
+      "step": 496
+    },
+    {
+      "epoch": 0.25871941697032796,
+      "grad_norm": 0.3333259893749479,
+      "learning_rate": 4.958669004622942e-05,
+      "loss": 0.1755,
+      "step": 497
+    },
+    {
+      "epoch": 0.25923997917751174,
+      "grad_norm": 0.352187312912872,
+      "learning_rate": 4.9584141920736656e-05,
+      "loss": 0.1814,
+      "step": 498
+    },
+    {
+      "epoch": 0.25976054138469545,
+      "grad_norm": 0.35402911428193723,
+      "learning_rate": 4.958158603042782e-05,
+      "loss": 0.1862,
+      "step": 499
+    },
+    {
+      "epoch": 0.2602811035918792,
+      "grad_norm": 0.351609210599415,
+      "learning_rate": 4.957902237611018e-05,
+      "loss": 0.1853,
+      "step": 500
+    },
+    {
+      "epoch": 0.260801665799063,
+      "grad_norm": 0.3961682023708728,
+      "learning_rate": 4.957645095859348e-05,
+      "loss": 0.1804,
+      "step": 501
+    },
+    {
+      "epoch": 0.26132222800624677,
+      "grad_norm": 0.3432093967728205,
+      "learning_rate": 4.957387177868986e-05,
+      "loss": 0.1777,
+      "step": 502
+    },
+    {
+      "epoch": 0.2618427902134305,
+      "grad_norm": 0.43841871784060643,
+      "learning_rate": 4.957128483721398e-05,
+      "loss": 0.1967,
+      "step": 503
+    },
+    {
+      "epoch": 0.26236335242061426,
+      "grad_norm": 0.32944557567419036,
+      "learning_rate": 4.9568690134982884e-05,
+      "loss": 0.1847,
+      "step": 504
+    },
+    {
+      "epoch": 0.26288391462779803,
+      "grad_norm": 0.4072278216406123,
+      "learning_rate": 4.956608767281612e-05,
+      "loss": 0.1916,
+      "step": 505
+    },
+    {
+      "epoch": 0.2634044768349818,
+      "grad_norm": 0.3470914106501254,
+      "learning_rate": 4.9563477451535664e-05,
+      "loss": 0.1686,
+      "step": 506
+    },
+    {
+      "epoch": 0.2639250390421655,
+      "grad_norm": 0.3431768907003393,
+      "learning_rate": 4.956085947196595e-05,
+      "loss": 0.1844,
+      "step": 507
+    },
+    {
+      "epoch": 0.2644456012493493,
+      "grad_norm": 0.373990259509472,
+      "learning_rate": 4.955823373493385e-05,
+      "loss": 0.1866,
+      "step": 508
+    },
+    {
+      "epoch": 0.26496616345653307,
+      "grad_norm": 0.36035756528381147,
+      "learning_rate": 4.955560024126868e-05,
+      "loss": 0.1828,
+      "step": 509
+    },
+    {
+      "epoch": 0.26548672566371684,
+      "grad_norm": 0.37277974712871015,
+      "learning_rate": 4.9552958991802245e-05,
+      "loss": 0.1877,
+      "step": 510
+    },
+    {
+      "epoch": 0.26600728787090056,
+      "grad_norm": 0.3791475893537772,
+      "learning_rate": 4.955030998736876e-05,
+      "loss": 0.1787,
+      "step": 511
+    },
+    {
+      "epoch": 0.2665278500780843,
+      "grad_norm": 0.34601195974476934,
+      "learning_rate": 4.9547653228804915e-05,
+      "loss": 0.178,
+      "step": 512
+    },
+    {
+      "epoch": 0.2670484122852681,
+      "grad_norm": 0.3860891277749901,
+      "learning_rate": 4.954498871694982e-05,
+      "loss": 0.1786,
+      "step": 513
+    },
+    {
+      "epoch": 0.26756897449245187,
+      "grad_norm": 0.36197045930962224,
+      "learning_rate": 4.954231645264507e-05,
+      "loss": 0.181,
+      "step": 514
+    },
+    {
+      "epoch": 0.2680895366996356,
+      "grad_norm": 0.35660823996500324,
+      "learning_rate": 4.953963643673468e-05,
+      "loss": 0.1816,
+      "step": 515
+    },
+    {
+      "epoch": 0.26861009890681936,
+      "grad_norm": 0.37899933398158164,
+      "learning_rate": 4.953694867006513e-05,
+      "loss": 0.1794,
+      "step": 516
+    },
+    {
+      "epoch": 0.26913066111400313,
+      "grad_norm": 0.37134644471067046,
+      "learning_rate": 4.953425315348534e-05,
+      "loss": 0.1829,
+      "step": 517
+    },
+    {
+      "epoch": 0.2696512233211869,
+      "grad_norm": 0.35629118982864444,
+      "learning_rate": 4.953154988784667e-05,
+      "loss": 0.1825,
+      "step": 518
+    },
+    {
+      "epoch": 0.2701717855283706,
+      "grad_norm": 0.3569372189926499,
+      "learning_rate": 4.952883887400296e-05,
+      "loss": 0.1852,
+      "step": 519
+    },
+    {
+      "epoch": 0.2706923477355544,
+      "grad_norm": 0.3784207348633246,
+      "learning_rate": 4.9526120112810445e-05,
+      "loss": 0.1792,
+      "step": 520
+    },
+    {
+      "epoch": 0.27121290994273817,
+      "grad_norm": 0.3571363618294585,
+      "learning_rate": 4.952339360512786e-05,
+      "loss": 0.1829,
+      "step": 521
+    },
+    {
+      "epoch": 0.27173347214992194,
+      "grad_norm": 0.37323568055590195,
+      "learning_rate": 4.952065935181635e-05,
+      "loss": 0.1844,
+      "step": 522
+    },
+    {
+      "epoch": 0.27225403435710566,
+      "grad_norm": 0.3405402620718736,
+      "learning_rate": 4.951791735373953e-05,
+      "loss": 0.1817,
+      "step": 523
+    },
+    {
+      "epoch": 0.27277459656428943,
+      "grad_norm": 0.3604521476507971,
+      "learning_rate": 4.9515167611763434e-05,
+      "loss": 0.1897,
+      "step": 524
+    },
+    {
+      "epoch": 0.2732951587714732,
+      "grad_norm": 0.33208621053766485,
+      "learning_rate": 4.951241012675657e-05,
+      "loss": 0.1792,
+      "step": 525
+    },
+    {
+      "epoch": 0.273815720978657,
+      "grad_norm": 0.3839270762469873,
+      "learning_rate": 4.950964489958988e-05,
+      "loss": 0.1841,
+      "step": 526
+    },
+    {
+      "epoch": 0.2743362831858407,
+      "grad_norm": 0.31535848820876206,
+      "learning_rate": 4.950687193113676e-05,
+      "loss": 0.1743,
+      "step": 527
+    },
+    {
+      "epoch": 0.27485684539302446,
+      "grad_norm": 0.37827478667026426,
+      "learning_rate": 4.950409122227302e-05,
+      "loss": 0.1883,
+      "step": 528
+    },
+    {
+      "epoch": 0.27537740760020823,
+      "grad_norm": 0.3582790920870559,
+      "learning_rate": 4.950130277387695e-05,
+      "loss": 0.1772,
+      "step": 529
+    },
+    {
+      "epoch": 0.275897969807392,
+      "grad_norm": 0.3584180526860497,
+      "learning_rate": 4.949850658682929e-05,
+      "loss": 0.1876,
+      "step": 530
+    },
+    {
+      "epoch": 0.2764185320145757,
+      "grad_norm": 0.5738445813183063,
+      "learning_rate": 4.949570266201317e-05,
+      "loss": 0.178,
+      "step": 531
+    },
+    {
+      "epoch": 0.2769390942217595,
+      "grad_norm": 0.3744948980693089,
+      "learning_rate": 4.949289100031422e-05,
+      "loss": 0.1823,
+      "step": 532
+    },
+    {
+      "epoch": 0.27745965642894327,
+      "grad_norm": 0.3379331374106325,
+      "learning_rate": 4.949007160262049e-05,
+      "loss": 0.1768,
+      "step": 533
+    },
+    {
+      "epoch": 0.27798021863612704,
+      "grad_norm": 0.3480919801415185,
+      "learning_rate": 4.948724446982248e-05,
+      "loss": 0.1859,
+      "step": 534
+    },
+    {
+      "epoch": 0.27850078084331076,
+      "grad_norm": 0.3842609731913717,
+      "learning_rate": 4.948440960281313e-05,
+      "loss": 0.1753,
+      "step": 535
+    },
+    {
+      "epoch": 0.27902134305049453,
+      "grad_norm": 0.37532802640166946,
+      "learning_rate": 4.948156700248782e-05,
+      "loss": 0.1837,
+      "step": 536
+    },
+    {
+      "epoch": 0.2795419052576783,
+      "grad_norm": 0.37959061595323806,
+      "learning_rate": 4.947871666974437e-05,
+      "loss": 0.1833,
+      "step": 537
+    },
+    {
+      "epoch": 0.2800624674648621,
+      "grad_norm": 0.4342019133871026,
+      "learning_rate": 4.9475858605483074e-05,
+      "loss": 0.1765,
+      "step": 538
+    },
+    {
+      "epoch": 0.2805830296720458,
+      "grad_norm": 0.3600696888138904,
+      "learning_rate": 4.94729928106066e-05,
+      "loss": 0.1839,
+      "step": 539
+    },
+    {
+      "epoch": 0.28110359187922956,
+      "grad_norm": 0.4075659366581729,
+      "learning_rate": 4.9470119286020134e-05,
+      "loss": 0.178,
+      "step": 540
+    },
+    {
+      "epoch": 0.28162415408641334,
+      "grad_norm": 0.3288846470475826,
+      "learning_rate": 4.946723803263125e-05,
+      "loss": 0.1822,
+      "step": 541
+    },
+    {
+      "epoch": 0.2821447162935971,
+      "grad_norm": 0.4265164240412001,
+      "learning_rate": 4.946434905134999e-05,
+      "loss": 0.1857,
+      "step": 542
+    },
+    {
+      "epoch": 0.2826652785007808,
+      "grad_norm": 0.3490790807245664,
+      "learning_rate": 4.9461452343088835e-05,
+      "loss": 0.1792,
+      "step": 543
+    },
+    {
+      "epoch": 0.2831858407079646,
+      "grad_norm": 0.3906850709137392,
+      "learning_rate": 4.945854790876268e-05,
+      "loss": 0.183,
+      "step": 544
+    },
+    {
+      "epoch": 0.28370640291514837,
+      "grad_norm": 0.3673184770495506,
+      "learning_rate": 4.94556357492889e-05,
+      "loss": 0.1791,
+      "step": 545
+    },
+    {
+      "epoch": 0.28422696512233214,
+      "grad_norm": 0.3435991187690188,
+      "learning_rate": 4.9452715865587274e-05,
+      "loss": 0.1799,
+      "step": 546
+    },
+    {
+      "epoch": 0.28474752732951586,
+      "grad_norm": 0.33557383372158994,
+      "learning_rate": 4.944978825858005e-05,
+      "loss": 0.1754,
+      "step": 547
+    },
+    {
+      "epoch": 0.28526808953669963,
+      "grad_norm": 0.3472232615973859,
+      "learning_rate": 4.944685292919191e-05,
+      "loss": 0.1829,
+      "step": 548
+    },
+    {
+      "epoch": 0.2857886517438834,
+      "grad_norm": 0.34437324485625154,
+      "learning_rate": 4.9443909878349945e-05,
+      "loss": 0.1849,
+      "step": 549
+    },
+    {
+      "epoch": 0.2863092139510672,
+      "grad_norm": 0.3366544790820249,
+      "learning_rate": 4.944095910698372e-05,
+      "loss": 0.1732,
+      "step": 550
+    },
+    {
+      "epoch": 0.2868297761582509,
+      "grad_norm": 0.3249404378916035,
+      "learning_rate": 4.9438000616025226e-05,
+      "loss": 0.1756,
+      "step": 551
+    },
+    {
+      "epoch": 0.28735033836543467,
+      "grad_norm": 0.3849346781758215,
+      "learning_rate": 4.94350344064089e-05,
+      "loss": 0.1856,
+      "step": 552
+    },
+    {
+      "epoch": 0.28787090057261844,
+      "grad_norm": 0.3243739567095206,
+      "learning_rate": 4.9432060479071584e-05,
+      "loss": 0.183,
+      "step": 553
+    },
+    {
+      "epoch": 0.2883914627798022,
+      "grad_norm": 0.3612277561023643,
+      "learning_rate": 4.942907883495261e-05,
+      "loss": 0.1786,
+      "step": 554
+    },
+    {
+      "epoch": 0.2889120249869859,
+      "grad_norm": 0.3323067953459406,
+      "learning_rate": 4.9426089474993696e-05,
+      "loss": 0.1816,
+      "step": 555
+    },
+    {
+      "epoch": 0.2894325871941697,
+      "grad_norm": 0.3316679874226504,
+      "learning_rate": 4.942309240013905e-05,
+      "loss": 0.1787,
+      "step": 556
+    },
+    {
+      "epoch": 0.28995314940135347,
+      "grad_norm": 0.36297638930718357,
+      "learning_rate": 4.9420087611335265e-05,
+      "loss": 0.1895,
+      "step": 557
+    },
+    {
+      "epoch": 0.29047371160853724,
+      "grad_norm": 0.34447907177875525,
+      "learning_rate": 4.94170751095314e-05,
+      "loss": 0.1831,
+      "step": 558
+    },
+    {
+      "epoch": 0.29099427381572096,
+      "grad_norm": 0.37466310329246094,
+      "learning_rate": 4.941405489567893e-05,
+      "loss": 0.18,
+      "step": 559
+    },
+    {
+      "epoch": 0.29151483602290473,
+      "grad_norm": 0.3240041218848788,
+      "learning_rate": 4.9411026970731805e-05,
+      "loss": 0.1695,
+      "step": 560
+    },
+    {
+      "epoch": 0.2920353982300885,
+      "grad_norm": 0.34612245370701383,
+      "learning_rate": 4.940799133564637e-05,
+      "loss": 0.1722,
+      "step": 561
+    },
+    {
+      "epoch": 0.2925559604372723,
+      "grad_norm": 0.35368918014173173,
+      "learning_rate": 4.9404947991381416e-05,
+      "loss": 0.1845,
+      "step": 562
+    },
+    {
+      "epoch": 0.293076522644456,
+      "grad_norm": 0.3048992546896397,
+      "learning_rate": 4.9401896938898185e-05,
+      "loss": 0.1764,
+      "step": 563
+    },
+    {
+      "epoch": 0.29359708485163977,
+      "grad_norm": 0.3780353682612039,
+      "learning_rate": 4.9398838179160326e-05,
+      "loss": 0.1763,
+      "step": 564
+    },
+    {
+      "epoch": 0.29411764705882354,
+      "grad_norm": 0.33431520426027345,
+      "learning_rate": 4.939577171313395e-05,
+      "loss": 0.1738,
+      "step": 565
+    },
+    {
+      "epoch": 0.2946382092660073,
+      "grad_norm": 0.3574740739801916,
+      "learning_rate": 4.9392697541787587e-05,
+      "loss": 0.1864,
+      "step": 566
+    },
+    {
+      "epoch": 0.29515877147319103,
+      "grad_norm": 0.3080198068295583,
+      "learning_rate": 4.93896156660922e-05,
+      "loss": 0.1758,
+      "step": 567
+    },
+    {
+      "epoch": 0.2956793336803748,
+      "grad_norm": 0.34221401560536524,
+      "learning_rate": 4.938652608702119e-05,
+      "loss": 0.1787,
+      "step": 568
+    },
+    {
+      "epoch": 0.2961998958875586,
+      "grad_norm": 0.3422289356320483,
+      "learning_rate": 4.938342880555039e-05,
+      "loss": 0.1791,
+      "step": 569
+    },
+    {
+      "epoch": 0.29672045809474235,
+      "grad_norm": 0.32294407935462033,
+      "learning_rate": 4.938032382265807e-05,
+      "loss": 0.1804,
+      "step": 570
+    },
+    {
+      "epoch": 0.29724102030192606,
+      "grad_norm": 0.32293386743920566,
+      "learning_rate": 4.937721113932493e-05,
+      "loss": 0.1745,
+      "step": 571
+    },
+    {
+      "epoch": 0.29776158250910983,
+      "grad_norm": 0.3339982925147738,
+      "learning_rate": 4.937409075653408e-05,
+      "loss": 0.1855,
+      "step": 572
+    },
+    {
+      "epoch": 0.2982821447162936,
+      "grad_norm": 0.33201821976883394,
+      "learning_rate": 4.9370962675271106e-05,
+      "loss": 0.1867,
+      "step": 573
+    },
+    {
+      "epoch": 0.2988027069234774,
+      "grad_norm": 0.36333200193928655,
+      "learning_rate": 4.936782689652399e-05,
+      "loss": 0.1798,
+      "step": 574
+    },
+    {
+      "epoch": 0.2993232691306611,
+      "grad_norm": 0.3170978099956422,
+      "learning_rate": 4.936468342128315e-05,
+      "loss": 0.1734,
+      "step": 575
+    },
+    {
+      "epoch": 0.29984383133784487,
+      "grad_norm": 0.3809200990521368,
+      "learning_rate": 4.936153225054146e-05,
+      "loss": 0.1731,
+      "step": 576
+    },
+    {
+      "epoch": 0.30036439354502864,
+      "grad_norm": 0.32675460808837287,
+      "learning_rate": 4.93583733852942e-05,
+      "loss": 0.1796,
+      "step": 577
+    },
+    {
+      "epoch": 0.3008849557522124,
+      "grad_norm": 0.34942233525996,
+      "learning_rate": 4.935520682653908e-05,
+      "loss": 0.182,
+      "step": 578
+    },
+    {
+      "epoch": 0.30140551795939613,
+      "grad_norm": 0.34564982764778135,
+      "learning_rate": 4.9352032575276255e-05,
+      "loss": 0.1778,
+      "step": 579
+    },
+    {
+      "epoch": 0.3019260801665799,
+      "grad_norm": 0.3716440462553411,
+      "learning_rate": 4.9348850632508295e-05,
+      "loss": 0.1793,
+      "step": 580
+    },
+    {
+      "epoch": 0.3024466423737637,
+      "grad_norm": 0.38121896546844525,
+      "learning_rate": 4.934566099924021e-05,
+      "loss": 0.1846,
+      "step": 581
+    },
+    {
+      "epoch": 0.30296720458094745,
+      "grad_norm": 0.34320281278816944,
+      "learning_rate": 4.9342463676479424e-05,
+      "loss": 0.1877,
+      "step": 582
+    },
+    {
+      "epoch": 0.30348776678813116,
+      "grad_norm": 0.356611150010106,
+      "learning_rate": 4.933925866523581e-05,
+      "loss": 0.1787,
+      "step": 583
+    },
+    {
+      "epoch": 0.30400832899531494,
+      "grad_norm": 0.3535551866752405,
+      "learning_rate": 4.933604596652166e-05,
+      "loss": 0.1753,
+      "step": 584
+    },
+    {
+      "epoch": 0.3045288912024987,
+      "grad_norm": 0.3587966333569922,
+      "learning_rate": 4.933282558135169e-05,
+      "loss": 0.1767,
+      "step": 585
+    },
+    {
+      "epoch": 0.3050494534096825,
+      "grad_norm": 0.32820492797103323,
+      "learning_rate": 4.932959751074305e-05,
+      "loss": 0.1765,
+      "step": 586
+    },
+    {
+      "epoch": 0.3055700156168662,
+      "grad_norm": 0.3668795974443606,
+      "learning_rate": 4.932636175571531e-05,
+      "loss": 0.1847,
+      "step": 587
+    },
+    {
+      "epoch": 0.30609057782404997,
+      "grad_norm": 0.3893815535302477,
+      "learning_rate": 4.932311831729048e-05,
+      "loss": 0.1769,
+      "step": 588
+    },
+    {
+      "epoch": 0.30661114003123374,
+      "grad_norm": 0.32745455373903526,
+      "learning_rate": 4.931986719649299e-05,
+      "loss": 0.1803,
+      "step": 589
+    },
+    {
+      "epoch": 0.3071317022384175,
+      "grad_norm": 0.33500065348948266,
+      "learning_rate": 4.9316608394349684e-05,
+      "loss": 0.1759,
+      "step": 590
+    },
+    {
+      "epoch": 0.30765226444560123,
+      "grad_norm": 0.34040682897700947,
+      "learning_rate": 4.931334191188985e-05,
+      "loss": 0.1809,
+      "step": 591
+    },
+    {
+      "epoch": 0.308172826652785,
+      "grad_norm": 0.30393987030284975,
+      "learning_rate": 4.93100677501452e-05,
+      "loss": 0.1707,
+      "step": 592
+    },
+    {
+      "epoch": 0.3086933888599688,
+      "grad_norm": 0.3351048511454525,
+      "learning_rate": 4.930678591014986e-05,
+      "loss": 0.1792,
+      "step": 593
+    },
+    {
+      "epoch": 0.30921395106715255,
+      "grad_norm": 0.36537436869160295,
+      "learning_rate": 4.930349639294038e-05,
+      "loss": 0.1808,
+      "step": 594
+    },
+    {
+      "epoch": 0.30973451327433627,
+      "grad_norm": 0.32634199146147447,
+      "learning_rate": 4.930019919955576e-05,
+      "loss": 0.1736,
+      "step": 595
+    },
+    {
+      "epoch": 0.31025507548152004,
+      "grad_norm": 0.3299845870559719,
+      "learning_rate": 4.9296894331037405e-05,
+      "loss": 0.1793,
+      "step": 596
+    },
+    {
+      "epoch": 0.3107756376887038,
+      "grad_norm": 0.34165209773973343,
+      "learning_rate": 4.9293581788429136e-05,
+      "loss": 0.1814,
+      "step": 597
+    },
+    {
+      "epoch": 0.3112961998958876,
+      "grad_norm": 0.3453422147378798,
+      "learning_rate": 4.92902615727772e-05,
+      "loss": 0.177,
+      "step": 598
+    },
+    {
+      "epoch": 0.3118167621030713,
+      "grad_norm": 0.32274993522205253,
+      "learning_rate": 4.92869336851303e-05,
+      "loss": 0.1708,
+      "step": 599
+    },
+    {
+      "epoch": 0.31233732431025507,
+      "grad_norm": 0.32030522939856365,
+      "learning_rate": 4.9283598126539524e-05,
+      "loss": 0.173,
+      "step": 600
+    },
+    {
+      "epoch": 0.31285788651743884,
+      "grad_norm": 0.3482608585554554,
+      "learning_rate": 4.92802548980584e-05,
+      "loss": 0.179,
+      "step": 601
+    },
+    {
+      "epoch": 0.3133784487246226,
+      "grad_norm": 0.3692046388241526,
+      "learning_rate": 4.927690400074286e-05,
+      "loss": 0.1812,
+      "step": 602
+    },
+    {
+      "epoch": 0.31389901093180633,
+      "grad_norm": 0.33322086716168214,
+      "learning_rate": 4.92735454356513e-05,
+      "loss": 0.1852,
+      "step": 603
+    },
+    {
+      "epoch": 0.3144195731389901,
+      "grad_norm": 0.3715077169273792,
+      "learning_rate": 4.92701792038445e-05,
+      "loss": 0.1767,
+      "step": 604
+    },
+    {
+      "epoch": 0.3149401353461739,
+      "grad_norm": 0.33348425102156853,
+      "learning_rate": 4.926680530638567e-05,
+      "loss": 0.179,
+      "step": 605
+    },
+    {
+      "epoch": 0.31546069755335765,
+      "grad_norm": 0.34957898455987335,
+      "learning_rate": 4.926342374434043e-05,
+      "loss": 0.1824,
+      "step": 606
+    },
+    {
+      "epoch": 0.31598125976054137,
+      "grad_norm": 0.35273594539974945,
+      "learning_rate": 4.926003451877687e-05,
+      "loss": 0.1784,
+      "step": 607
+    },
+    {
+      "epoch": 0.31650182196772514,
+      "grad_norm": 0.32859387991060474,
+      "learning_rate": 4.9256637630765425e-05,
+      "loss": 0.1761,
+      "step": 608
+    },
+    {
+      "epoch": 0.3170223841749089,
+      "grad_norm": 0.3364281217569474,
+      "learning_rate": 4.9253233081379024e-05,
+      "loss": 0.1757,
+      "step": 609
+    },
+    {
+      "epoch": 0.3175429463820927,
+      "grad_norm": 0.3180815183990718,
+      "learning_rate": 4.924982087169296e-05,
+      "loss": 0.1688,
+      "step": 610
+    },
+    {
+      "epoch": 0.3180635085892764,
+      "grad_norm": 0.3265120523959636,
+      "learning_rate": 4.9246401002784976e-05,
+      "loss": 0.1794,
+      "step": 611
+    },
+    {
+      "epoch": 0.3185840707964602,
+      "grad_norm": 0.3435491116451335,
+      "learning_rate": 4.9242973475735224e-05,
+      "loss": 0.1807,
+      "step": 612
+    },
+    {
+      "epoch": 0.31910463300364394,
+      "grad_norm": 0.3601397832581144,
+      "learning_rate": 4.923953829162628e-05,
+      "loss": 0.1844,
+      "step": 613
+    },
+    {
+      "epoch": 0.3196251952108277,
+      "grad_norm": 0.33676918035319736,
+      "learning_rate": 4.923609545154313e-05,
+      "loss": 0.1794,
+      "step": 614
+    },
+    {
+      "epoch": 0.32014575741801143,
+      "grad_norm": 0.35847840458223534,
+      "learning_rate": 4.923264495657319e-05,
+      "loss": 0.1796,
+      "step": 615
+    },
+    {
+      "epoch": 0.3206663196251952,
+      "grad_norm": 0.3517065818342989,
+      "learning_rate": 4.9229186807806284e-05,
+      "loss": 0.1772,
+      "step": 616
+    },
+    {
+      "epoch": 0.321186881832379,
+      "grad_norm": 0.32840575469268135,
+      "learning_rate": 4.9225721006334644e-05,
+      "loss": 0.183,
+      "step": 617
+    },
+    {
+      "epoch": 0.32170744403956275,
+      "grad_norm": 0.3304656690924349,
+      "learning_rate": 4.922224755325295e-05,
+      "loss": 0.1826,
+      "step": 618
+    },
+    {
+      "epoch": 0.32222800624674647,
+      "grad_norm": 0.3365642578033948,
+      "learning_rate": 4.921876644965827e-05,
+      "loss": 0.1849,
+      "step": 619
+    },
+    {
+      "epoch": 0.32274856845393024,
+      "grad_norm": 0.321873293657072,
+      "learning_rate": 4.921527769665011e-05,
+      "loss": 0.1842,
+      "step": 620
+    },
+    {
+      "epoch": 0.323269130661114,
+      "grad_norm": 0.3304218237484672,
+      "learning_rate": 4.921178129533036e-05,
+      "loss": 0.1766,
+      "step": 621
+    },
+    {
+      "epoch": 0.3237896928682978,
+      "grad_norm": 0.36159820669310355,
+      "learning_rate": 4.920827724680336e-05,
+      "loss": 0.1833,
+      "step": 622
+    },
+    {
+      "epoch": 0.3243102550754815,
+      "grad_norm": 0.3125115981551096,
+      "learning_rate": 4.9204765552175857e-05,
+      "loss": 0.1687,
+      "step": 623
+    },
+    {
+      "epoch": 0.3248308172826653,
+      "grad_norm": 0.355646049290069,
+      "learning_rate": 4.920124621255699e-05,
+      "loss": 0.1831,
+      "step": 624
+    },
+    {
+      "epoch": 0.32535137948984905,
+      "grad_norm": 0.31743405014825066,
+      "learning_rate": 4.9197719229058346e-05,
+      "loss": 0.1719,
+      "step": 625
+    },
+    {
+      "epoch": 0.3258719416970328,
+      "grad_norm": 0.3320164493908055,
+      "learning_rate": 4.9194184602793904e-05,
+      "loss": 0.1779,
+      "step": 626
+    },
+    {
+      "epoch": 0.32639250390421654,
+      "grad_norm": 0.3569014079597437,
+      "learning_rate": 4.919064233488006e-05,
+      "loss": 0.1792,
+      "step": 627
+    },
+    {
+      "epoch": 0.3269130661114003,
+      "grad_norm": 0.3422394188289934,
+      "learning_rate": 4.9187092426435634e-05,
+      "loss": 0.1853,
+      "step": 628
+    },
+    {
+      "epoch": 0.3274336283185841,
+      "grad_norm": 0.45779658269843443,
+      "learning_rate": 4.918353487858185e-05,
+      "loss": 0.1865,
+      "step": 629
+    },
+    {
+      "epoch": 0.32795419052576785,
+      "grad_norm": 0.36967762282280137,
+      "learning_rate": 4.917996969244235e-05,
+      "loss": 0.1842,
+      "step": 630
+    },
+    {
+      "epoch": 0.32847475273295157,
+      "grad_norm": 0.3383443636578056,
+      "learning_rate": 4.917639686914317e-05,
+      "loss": 0.177,
+      "step": 631
+    },
+    {
+      "epoch": 0.32899531494013534,
+      "grad_norm": 0.34576867722134114,
+      "learning_rate": 4.91728164098128e-05,
+      "loss": 0.1741,
+      "step": 632
+    },
+    {
+      "epoch": 0.3295158771473191,
+      "grad_norm": 0.33889725319396574,
+      "learning_rate": 4.9169228315582094e-05,
+      "loss": 0.1733,
+      "step": 633
+    },
+    {
+      "epoch": 0.3300364393545029,
+      "grad_norm": 0.3181574991844868,
+      "learning_rate": 4.9165632587584346e-05,
+      "loss": 0.1712,
+      "step": 634
+    },
+    {
+      "epoch": 0.3305570015616866,
+      "grad_norm": 0.3150178248046848,
+      "learning_rate": 4.916202922695526e-05,
+      "loss": 0.1769,
+      "step": 635
+    },
+    {
+      "epoch": 0.3310775637688704,
+      "grad_norm": 0.3231916056988664,
+      "learning_rate": 4.9158418234832935e-05,
+      "loss": 0.1711,
+      "step": 636
+    },
+    {
+      "epoch": 0.33159812597605415,
+      "grad_norm": 0.3219246488562882,
+      "learning_rate": 4.9154799612357905e-05,
+      "loss": 0.1742,
+      "step": 637
+    },
+    {
+      "epoch": 0.3321186881832379,
+      "grad_norm": 0.3490019970784047,
+      "learning_rate": 4.915117336067308e-05,
+      "loss": 0.187,
+      "step": 638
+    },
+    {
+      "epoch": 0.33263925039042164,
+      "grad_norm": 0.3404922017535072,
+      "learning_rate": 4.914753948092381e-05,
+      "loss": 0.1814,
+      "step": 639
+    },
+    {
+      "epoch": 0.3331598125976054,
+      "grad_norm": 0.33038580290518343,
+      "learning_rate": 4.9143897974257845e-05,
+      "loss": 0.1836,
+      "step": 640
+    },
+    {
+      "epoch": 0.3336803748047892,
+      "grad_norm": 0.48447675580050537,
+      "learning_rate": 4.914024884182534e-05,
+      "loss": 0.1723,
+      "step": 641
+    },
+    {
+      "epoch": 0.33420093701197295,
+      "grad_norm": 0.31792440208506284,
+      "learning_rate": 4.913659208477886e-05,
+      "loss": 0.1744,
+      "step": 642
+    },
+    {
+      "epoch": 0.33472149921915667,
+      "grad_norm": 1.1187782722485253,
+      "learning_rate": 4.9132927704273376e-05,
+      "loss": 0.1782,
+      "step": 643
+    },
+    {
+      "epoch": 0.33524206142634044,
+      "grad_norm": 0.46784530940515706,
+      "learning_rate": 4.9129255701466284e-05,
+      "loss": 0.1785,
+      "step": 644
+    },
+    {
+      "epoch": 0.3357626236335242,
+      "grad_norm": 3.7910679408191483,
+      "learning_rate": 4.9125576077517356e-05,
+      "loss": 0.19,
+      "step": 645
+    },
+    {
+      "epoch": 0.336283185840708,
+      "grad_norm": 0.5244754308415,
+      "learning_rate": 4.9121888833588795e-05,
+      "loss": 0.179,
+      "step": 646
+    },
+    {
+      "epoch": 0.3368037480478917,
+      "grad_norm": 1.3332280542365016,
+      "learning_rate": 4.91181939708452e-05,
+      "loss": 0.1785,
+      "step": 647
+    },
+    {
+      "epoch": 0.3373243102550755,
+      "grad_norm": 0.6152204952369942,
+      "learning_rate": 4.9114491490453585e-05,
+      "loss": 0.1817,
+      "step": 648
+    },
+    {
+      "epoch": 0.33784487246225925,
+      "grad_norm": 0.6322972902919656,
+      "learning_rate": 4.911078139358337e-05,
+      "loss": 0.1834,
+      "step": 649
+    },
+    {
+      "epoch": 0.338365434669443,
+      "grad_norm": 0.4481338546785872,
+      "learning_rate": 4.9107063681406376e-05,
+      "loss": 0.1825,
+      "step": 650
+    },
+    {
+      "epoch": 0.33888599687662674,
+      "grad_norm": 0.9859952621710755,
+      "learning_rate": 4.910333835509682e-05,
+      "loss": 0.1809,
+      "step": 651
+    },
+    {
+      "epoch": 0.3394065590838105,
+      "grad_norm": 0.4410398257105147,
+      "learning_rate": 4.909960541583133e-05,
+      "loss": 0.1758,
+      "step": 652
+    },
+    {
+      "epoch": 0.3399271212909943,
+      "grad_norm": 0.46192330205640203,
+      "learning_rate": 4.909586486478897e-05,
+      "loss": 0.1745,
+      "step": 653
+    },
+    {
+      "epoch": 0.34044768349817806,
+      "grad_norm": 0.4347198500546606,
+      "learning_rate": 4.909211670315114e-05,
+      "loss": 0.1763,
+      "step": 654
+    },
+    {
+      "epoch": 0.34096824570536177,
+      "grad_norm": 0.47923777050652683,
+      "learning_rate": 4.908836093210172e-05,
+      "loss": 0.186,
+      "step": 655
+    },
+    {
+      "epoch": 0.34148880791254554,
+      "grad_norm": 0.3778493845758958,
+      "learning_rate": 4.9084597552826935e-05,
+      "loss": 0.1778,
+      "step": 656
+    },
+    {
+      "epoch": 0.3420093701197293,
+      "grad_norm": 0.4629741259881899,
+      "learning_rate": 4.908082656651544e-05,
+      "loss": 0.1835,
+      "step": 657
+    },
+    {
+      "epoch": 0.3425299323269131,
+      "grad_norm": 0.43536744466295096,
+      "learning_rate": 4.907704797435829e-05,
+      "loss": 0.1836,
+      "step": 658
+    },
+    {
+      "epoch": 0.3430504945340968,
+      "grad_norm": 0.390115416460036,
+      "learning_rate": 4.9073261777548954e-05,
+      "loss": 0.1835,
+      "step": 659
+    },
+    {
+      "epoch": 0.3435710567412806,
+      "grad_norm": 0.46457793969633604,
+      "learning_rate": 4.9069467977283255e-05,
+      "loss": 0.1771,
+      "step": 660
+    },
+    {
+      "epoch": 0.34409161894846435,
+      "grad_norm": 0.3963393707607232,
+      "learning_rate": 4.906566657475949e-05,
+      "loss": 0.1769,
+      "step": 661
+    },
+    {
+      "epoch": 0.3446121811556481,
+      "grad_norm": 0.36870322720760557,
+      "learning_rate": 4.906185757117829e-05,
+      "loss": 0.1781,
+      "step": 662
+    },
+    {
+      "epoch": 0.34513274336283184,
+      "grad_norm": 0.3979707755147091,
+      "learning_rate": 4.905804096774274e-05,
+      "loss": 0.1801,
+      "step": 663
+    },
+    {
+      "epoch": 0.3456533055700156,
+      "grad_norm": 0.36701302139429726,
+      "learning_rate": 4.905421676565827e-05,
+      "loss": 0.181,
+      "step": 664
+    },
+    {
+      "epoch": 0.3461738677771994,
+      "grad_norm": 0.3698838374588819,
+      "learning_rate": 4.905038496613277e-05,
+      "loss": 0.1877,
+      "step": 665
+    },
+    {
+      "epoch": 0.34669442998438316,
+      "grad_norm": 0.352717338392521,
+      "learning_rate": 4.9046545570376484e-05,
+      "loss": 0.1789,
+      "step": 666
+    },
+    {
+      "epoch": 0.3472149921915669,
+      "grad_norm": 0.3809906085196877,
+      "learning_rate": 4.904269857960207e-05,
+      "loss": 0.1779,
+      "step": 667
+    },
+    {
+      "epoch": 0.34773555439875065,
+      "grad_norm": 0.3717632611173711,
+      "learning_rate": 4.9038843995024606e-05,
+      "loss": 0.1841,
+      "step": 668
+    },
+    {
+      "epoch": 0.3482561166059344,
+      "grad_norm": 0.3616057869533888,
+      "learning_rate": 4.9034981817861534e-05,
+      "loss": 0.1832,
+      "step": 669
+    },
+    {
+      "epoch": 0.3487766788131182,
+      "grad_norm": 0.3720170377297904,
+      "learning_rate": 4.9031112049332715e-05,
+      "loss": 0.1773,
+      "step": 670
+    },
+    {
+      "epoch": 0.3492972410203019,
+      "grad_norm": 0.38176483972868525,
+      "learning_rate": 4.9027234690660396e-05,
+      "loss": 0.1795,
+      "step": 671
+    },
+    {
+      "epoch": 0.3498178032274857,
+      "grad_norm": 0.3971475485254792,
+      "learning_rate": 4.902334974306924e-05,
+      "loss": 0.187,
+      "step": 672
+    },
+    {
+      "epoch": 0.35033836543466945,
+      "grad_norm": 0.33041713668617295,
+      "learning_rate": 4.9019457207786265e-05,
+      "loss": 0.168,
+      "step": 673
+    },
+    {
+      "epoch": 0.3508589276418532,
+      "grad_norm": 0.3660144363430606,
+      "learning_rate": 4.901555708604095e-05,
+      "loss": 0.1856,
+      "step": 674
+    },
+    {
+      "epoch": 0.35137948984903694,
+      "grad_norm": 0.3766530978916359,
+      "learning_rate": 4.901164937906511e-05,
+      "loss": 0.1794,
+      "step": 675
+    },
+    {
+      "epoch": 0.3519000520562207,
+      "grad_norm": 0.3378578696974802,
+      "learning_rate": 4.900773408809299e-05,
+      "loss": 0.1757,
+      "step": 676
+    },
+    {
+      "epoch": 0.3524206142634045,
+      "grad_norm": 0.421166687268464,
+      "learning_rate": 4.900381121436123e-05,
+      "loss": 0.1824,
+      "step": 677
+    },
+    {
+      "epoch": 0.35294117647058826,
+      "grad_norm": 0.36026982081841086,
+      "learning_rate": 4.8999880759108844e-05,
+      "loss": 0.1811,
+      "step": 678
+    },
+    {
+      "epoch": 0.353461738677772,
+      "grad_norm": 0.3655859893200682,
+      "learning_rate": 4.899594272357726e-05,
+      "loss": 0.1739,
+      "step": 679
+    },
+    {
+      "epoch": 0.35398230088495575,
+      "grad_norm": 0.3821275209491641,
+      "learning_rate": 4.899199710901028e-05,
+      "loss": 0.1783,
+      "step": 680
+    },
+    {
+      "epoch": 0.3545028630921395,
+      "grad_norm": 0.3409239426235484,
+      "learning_rate": 4.8988043916654126e-05,
+      "loss": 0.1782,
+      "step": 681
+    },
+    {
+      "epoch": 0.3550234252993233,
+      "grad_norm": 0.37076066161212423,
+      "learning_rate": 4.89840831477574e-05,
+      "loss": 0.1732,
+      "step": 682
+    },
+    {
+      "epoch": 0.355543987506507,
+      "grad_norm": 0.3742846141003235,
+      "learning_rate": 4.8980114803571084e-05,
+      "loss": 0.1874,
+      "step": 683
+    },
+    {
+      "epoch": 0.3560645497136908,
+      "grad_norm": 0.3537697722706192,
+      "learning_rate": 4.8976138885348575e-05,
+      "loss": 0.1813,
+      "step": 684
+    },
+    {
+      "epoch": 0.35658511192087455,
+      "grad_norm": 0.34743989109554174,
+      "learning_rate": 4.897215539434566e-05,
+      "loss": 0.1759,
+      "step": 685
+    },
+    {
+      "epoch": 0.3571056741280583,
+      "grad_norm": 0.3591186392260258,
+      "learning_rate": 4.89681643318205e-05,
+      "loss": 0.1677,
+      "step": 686
+    },
+    {
+      "epoch": 0.35762623633524204,
+      "grad_norm": 0.35400264664937536,
+      "learning_rate": 4.896416569903366e-05,
+      "loss": 0.1755,
+      "step": 687
+    },
+    {
+      "epoch": 0.3581467985424258,
+      "grad_norm": 0.36086365972267287,
+      "learning_rate": 4.89601594972481e-05,
+      "loss": 0.1738,
+      "step": 688
+    },
+    {
+      "epoch": 0.3586673607496096,
+      "grad_norm": 0.34105586937400784,
+      "learning_rate": 4.8956145727729156e-05,
+      "loss": 0.1743,
+      "step": 689
+    },
+    {
+      "epoch": 0.35918792295679336,
+      "grad_norm": 0.3373349436933188,
+      "learning_rate": 4.895212439174457e-05,
+      "loss": 0.1733,
+      "step": 690
+    },
+    {
+      "epoch": 0.3597084851639771,
+      "grad_norm": 0.3503292696755782,
+      "learning_rate": 4.894809549056447e-05,
+      "loss": 0.1727,
+      "step": 691
+    },
+    {
+      "epoch": 0.36022904737116085,
+      "grad_norm": 0.35086053297097963,
+      "learning_rate": 4.894405902546136e-05,
+      "loss": 0.1863,
+      "step": 692
+    },
+    {
+      "epoch": 0.3607496095783446,
+      "grad_norm": 0.3839353193339784,
+      "learning_rate": 4.894001499771015e-05,
+      "loss": 0.183,
+      "step": 693
+    },
+    {
+      "epoch": 0.3612701717855284,
+      "grad_norm": 0.32399392427850626,
+      "learning_rate": 4.8935963408588134e-05,
+      "loss": 0.1748,
+      "step": 694
+    },
+    {
+      "epoch": 0.3617907339927121,
+      "grad_norm": 0.3585859771864591,
+      "learning_rate": 4.893190425937499e-05,
+      "loss": 0.1811,
+      "step": 695
+    },
+    {
+      "epoch": 0.3623112961998959,
+      "grad_norm": 0.3227714956193789,
+      "learning_rate": 4.8927837551352784e-05,
+      "loss": 0.1782,
+      "step": 696
+    },
+    {
+      "epoch": 0.36283185840707965,
+      "grad_norm": 0.3474440141200088,
+      "learning_rate": 4.8923763285805965e-05,
+      "loss": 0.1754,
+      "step": 697
+    },
+    {
+      "epoch": 0.3633524206142634,
+      "grad_norm": 0.3322694368519195,
+      "learning_rate": 4.89196814640214e-05,
+      "loss": 0.1774,
+      "step": 698
+    },
+    {
+      "epoch": 0.36387298282144714,
+      "grad_norm": 0.34820159808484086,
+      "learning_rate": 4.89155920872883e-05,
+      "loss": 0.177,
+      "step": 699
+    },
+    {
+      "epoch": 0.3643935450286309,
+      "grad_norm": 0.3031746341954232,
+      "learning_rate": 4.891149515689827e-05,
+      "loss": 0.1736,
+      "step": 700
+    },
+    {
+      "epoch": 0.3649141072358147,
+      "grad_norm": 0.36173867640457913,
+      "learning_rate": 4.8907390674145335e-05,
+      "loss": 0.1861,
+      "step": 701
+    },
+    {
+      "epoch": 0.36543466944299846,
+      "grad_norm": 0.30797603475460056,
+      "learning_rate": 4.890327864032587e-05,
+      "loss": 0.1766,
+      "step": 702
+    },
+    {
+      "epoch": 0.3659552316501822,
+      "grad_norm": 0.4370440494274063,
+      "learning_rate": 4.8899159056738646e-05,
+      "loss": 0.1791,
+      "step": 703
+    },
+    {
+      "epoch": 0.36647579385736595,
+      "grad_norm": 0.33055435550054857,
+      "learning_rate": 4.889503192468482e-05,
+      "loss": 0.1794,
+      "step": 704
+    },
+    {
+      "epoch": 0.3669963560645497,
+      "grad_norm": 0.3232689437227807,
+      "learning_rate": 4.8890897245467934e-05,
+      "loss": 0.1692,
+      "step": 705
+    },
+    {
+      "epoch": 0.3675169182717335,
+      "grad_norm": 0.32294570761168684,
+      "learning_rate": 4.8886755020393915e-05,
+      "loss": 0.1815,
+      "step": 706
+    },
+    {
+      "epoch": 0.3680374804789172,
+      "grad_norm": 0.34062024576611194,
+      "learning_rate": 4.888260525077106e-05,
+      "loss": 0.1702,
+      "step": 707
+    },
+    {
+      "epoch": 0.368558042686101,
+      "grad_norm": 0.30465620374884,
+      "learning_rate": 4.887844793791008e-05,
+      "loss": 0.1751,
+      "step": 708
+    },
+    {
+      "epoch": 0.36907860489328476,
+      "grad_norm": 0.3420067292282057,
+      "learning_rate": 4.887428308312402e-05,
+      "loss": 0.1787,
+      "step": 709
+    },
+    {
+      "epoch": 0.36959916710046853,
+      "grad_norm": 0.32157102458891257,
+      "learning_rate": 4.887011068772835e-05,
+      "loss": 0.1757,
+      "step": 710
+    },
+    {
+      "epoch": 0.37011972930765225,
+      "grad_norm": 0.3128376282076131,
+      "learning_rate": 4.886593075304091e-05,
+      "loss": 0.1743,
+      "step": 711
+    },
+    {
+      "epoch": 0.370640291514836,
+      "grad_norm": 0.3401322277124052,
+      "learning_rate": 4.886174328038191e-05,
+      "loss": 0.1777,
+      "step": 712
+    },
+    {
+      "epoch": 0.3711608537220198,
+      "grad_norm": 0.33579339844499495,
+      "learning_rate": 4.885754827107395e-05,
+      "loss": 0.1772,
+      "step": 713
+    },
+    {
+      "epoch": 0.37168141592920356,
+      "grad_norm": 0.3153726122024893,
+      "learning_rate": 4.885334572644202e-05,
+      "loss": 0.1789,
+      "step": 714
+    },
+    {
+      "epoch": 0.3722019781363873,
+      "grad_norm": 0.32118752532405626,
+      "learning_rate": 4.884913564781346e-05,
+      "loss": 0.1719,
+      "step": 715
+    },
+    {
+      "epoch": 0.37272254034357105,
+      "grad_norm": 0.3213935784665974,
+      "learning_rate": 4.884491803651803e-05,
+      "loss": 0.169,
+      "step": 716
+    },
+    {
+      "epoch": 0.3732431025507548,
+      "grad_norm": 0.33814274890520907,
+      "learning_rate": 4.884069289388783e-05,
+      "loss": 0.1781,
+      "step": 717
+    },
+    {
+      "epoch": 0.3737636647579386,
+      "grad_norm": 0.30359841482721706,
+      "learning_rate": 4.883646022125736e-05,
+      "loss": 0.1733,
+      "step": 718
+    },
+    {
+      "epoch": 0.3742842269651223,
+      "grad_norm": 0.30429885008494284,
+      "learning_rate": 4.8832220019963514e-05,
+      "loss": 0.1693,
+      "step": 719
+    },
+    {
+      "epoch": 0.3748047891723061,
+      "grad_norm": 0.331016348005072,
+      "learning_rate": 4.882797229134551e-05,
+      "loss": 0.1771,
+      "step": 720
+    },
+    {
+      "epoch": 0.37532535137948986,
+      "grad_norm": 0.3339222828334093,
+      "learning_rate": 4.882371703674501e-05,
+      "loss": 0.1782,
+      "step": 721
+    },
+    {
+      "epoch": 0.37584591358667363,
+      "grad_norm": 0.2991710011140988,
+      "learning_rate": 4.8819454257506015e-05,
+      "loss": 0.1672,
+      "step": 722
+    },
+    {
+      "epoch": 0.37636647579385735,
+      "grad_norm": 0.36396331898025386,
+      "learning_rate": 4.8815183954974896e-05,
+      "loss": 0.1715,
+      "step": 723
+    },
+    {
+      "epoch": 0.3768870380010411,
+      "grad_norm": 0.33416943073891897,
+      "learning_rate": 4.881090613050042e-05,
+      "loss": 0.1771,
+      "step": 724
+    },
+    {
+      "epoch": 0.3774076002082249,
+      "grad_norm": 0.36523650639258104,
+      "learning_rate": 4.8806620785433726e-05,
+      "loss": 0.164,
+      "step": 725
+    },
+    {
+      "epoch": 0.37792816241540866,
+      "grad_norm": 0.32073950251793465,
+      "learning_rate": 4.880232792112832e-05,
+      "loss": 0.171,
+      "step": 726
+    },
+    {
+      "epoch": 0.3784487246225924,
+      "grad_norm": 0.3484933060375919,
+      "learning_rate": 4.879802753894009e-05,
+      "loss": 0.1803,
+      "step": 727
+    },
+    {
+      "epoch": 0.37896928682977615,
+      "grad_norm": 0.33643384179749536,
+      "learning_rate": 4.879371964022731e-05,
+      "loss": 0.1762,
+      "step": 728
+    },
+    {
+      "epoch": 0.3794898490369599,
+      "grad_norm": 0.3844044688078634,
+      "learning_rate": 4.878940422635059e-05,
+      "loss": 0.1757,
+      "step": 729
+    },
+    {
+      "epoch": 0.3800104112441437,
+      "grad_norm": 0.31212059732422176,
+      "learning_rate": 4.878508129867296e-05,
+      "loss": 0.1744,
+      "step": 730
+    },
+    {
+      "epoch": 0.3805309734513274,
+      "grad_norm": 0.37952982057783435,
+      "learning_rate": 4.8780750858559794e-05,
+      "loss": 0.1873,
+      "step": 731
+    },
+    {
+      "epoch": 0.3810515356585112,
+      "grad_norm": 0.36193360687044024,
+      "learning_rate": 4.877641290737884e-05,
+      "loss": 0.1746,
+      "step": 732
+    },
+    {
+      "epoch": 0.38157209786569496,
+      "grad_norm": 0.3415108925924172,
+      "learning_rate": 4.8772067446500235e-05,
+      "loss": 0.176,
+      "step": 733
+    },
+    {
+      "epoch": 0.38209266007287873,
+      "grad_norm": 0.35739113924155885,
+      "learning_rate": 4.8767714477296475e-05,
+      "loss": 0.1773,
+      "step": 734
+    },
+    {
+      "epoch": 0.38261322228006245,
+      "grad_norm": 0.299683012812736,
+      "learning_rate": 4.8763354001142426e-05,
+      "loss": 0.1742,
+      "step": 735
+    },
+    {
+      "epoch": 0.3831337844872462,
+      "grad_norm": 0.3281156150608031,
+      "learning_rate": 4.875898601941533e-05,
+      "loss": 0.1773,
+      "step": 736
+    },
+    {
+      "epoch": 0.38365434669443,
+      "grad_norm": 0.29779263356766955,
+      "learning_rate": 4.875461053349481e-05,
+      "loss": 0.1754,
+      "step": 737
+    },
+    {
+      "epoch": 0.38417490890161377,
+      "grad_norm": 0.36739569269258815,
+      "learning_rate": 4.875022754476283e-05,
+      "loss": 0.1768,
+      "step": 738
+    },
+    {
+      "epoch": 0.3846954711087975,
+      "grad_norm": 0.3305403555222801,
+      "learning_rate": 4.8745837054603746e-05,
+      "loss": 0.1782,
+      "step": 739
+    },
+    {
+      "epoch": 0.38521603331598125,
+      "grad_norm": 0.3461843255219805,
+      "learning_rate": 4.87414390644043e-05,
+      "loss": 0.1761,
+      "step": 740
+    },
+    {
+      "epoch": 0.385736595523165,
+      "grad_norm": 0.34024915037224157,
+      "learning_rate": 4.8737033575553556e-05,
+      "loss": 0.1783,
+      "step": 741
+    },
+    {
+      "epoch": 0.3862571577303488,
+      "grad_norm": 0.36955073863687815,
+      "learning_rate": 4.873262058944299e-05,
+      "loss": 0.1801,
+      "step": 742
+    },
+    {
+      "epoch": 0.3867777199375325,
+      "grad_norm": 0.33810479957816336,
+      "learning_rate": 4.872820010746641e-05,
+      "loss": 0.1781,
+      "step": 743
+    },
+    {
+      "epoch": 0.3872982821447163,
+      "grad_norm": 0.31816896200979733,
+      "learning_rate": 4.872377213102003e-05,
+      "loss": 0.1701,
+      "step": 744
+    },
+    {
+      "epoch": 0.38781884435190006,
+      "grad_norm": 0.2994338543173188,
+      "learning_rate": 4.871933666150239e-05,
+      "loss": 0.1768,
+      "step": 745
+    },
+    {
+      "epoch": 0.38833940655908383,
+      "grad_norm": 0.3143457501312435,
+      "learning_rate": 4.8714893700314445e-05,
+      "loss": 0.1745,
+      "step": 746
+    },
+    {
+      "epoch": 0.38885996876626755,
+      "grad_norm": 0.3323503389224882,
+      "learning_rate": 4.8710443248859464e-05,
+      "loss": 0.1747,
+      "step": 747
+    },
+    {
+      "epoch": 0.3893805309734513,
+      "grad_norm": 0.3144112661589175,
+      "learning_rate": 4.870598530854312e-05,
+      "loss": 0.1722,
+      "step": 748
+    },
+    {
+      "epoch": 0.3899010931806351,
+      "grad_norm": 0.3299614127535381,
+      "learning_rate": 4.870151988077343e-05,
+      "loss": 0.177,
+      "step": 749
+    },
+    {
+      "epoch": 0.39042165538781887,
+      "grad_norm": 0.3303857038262982,
+      "learning_rate": 4.869704696696079e-05,
+      "loss": 0.1686,
+      "step": 750
+    },
+    {
+      "epoch": 0.3909422175950026,
+      "grad_norm": 0.30943696088141864,
+      "learning_rate": 4.869256656851795e-05,
+      "loss": 0.1748,
+      "step": 751
+    },
+    {
+      "epoch": 0.39146277980218636,
+      "grad_norm": 0.3347597384412751,
+      "learning_rate": 4.8688078686860025e-05,
+      "loss": 0.1848,
+      "step": 752
+    },
+    {
+      "epoch": 0.39198334200937013,
+      "grad_norm": 0.3357985035932315,
+      "learning_rate": 4.8683583323404514e-05,
+      "loss": 0.1722,
+      "step": 753
+    },
+    {
+      "epoch": 0.3925039042165539,
+      "grad_norm": 0.3229921181970955,
+      "learning_rate": 4.867908047957125e-05,
+      "loss": 0.1716,
+      "step": 754
+    },
+    {
+      "epoch": 0.3930244664237376,
+      "grad_norm": 0.4022384622310971,
+      "learning_rate": 4.867457015678244e-05,
+      "loss": 0.1786,
+      "step": 755
+    },
+    {
+      "epoch": 0.3935450286309214,
+      "grad_norm": 0.3049771099599678,
+      "learning_rate": 4.867005235646265e-05,
+      "loss": 0.1793,
+      "step": 756
+    },
+    {
+      "epoch": 0.39406559083810516,
+      "grad_norm": 0.4173130163682277,
+      "learning_rate": 4.866552708003882e-05,
+      "loss": 0.1778,
+      "step": 757
+    },
+    {
+      "epoch": 0.39458615304528893,
+      "grad_norm": 0.3313304796655795,
+      "learning_rate": 4.8660994328940235e-05,
+      "loss": 0.179,
+      "step": 758
+    },
+    {
+      "epoch": 0.39510671525247265,
+      "grad_norm": 0.3218402610994695,
+      "learning_rate": 4.865645410459856e-05,
+      "loss": 0.1628,
+      "step": 759
+    },
+    {
+      "epoch": 0.3956272774596564,
+      "grad_norm": 0.32361763518417075,
+      "learning_rate": 4.8651906408447795e-05,
+      "loss": 0.1749,
+      "step": 760
+    },
+    {
+      "epoch": 0.3961478396668402,
+      "grad_norm": 0.33838851288907523,
+      "learning_rate": 4.864735124192432e-05,
+      "loss": 0.1731,
+      "step": 761
+    },
+    {
+      "epoch": 0.39666840187402397,
+      "grad_norm": 0.30076334190054943,
+      "learning_rate": 4.8642788606466884e-05,
+      "loss": 0.1697,
+      "step": 762
+    },
+    {
+      "epoch": 0.3971889640812077,
+      "grad_norm": 0.2972937365361002,
+      "learning_rate": 4.8638218503516554e-05,
+      "loss": 0.1732,
+      "step": 763
+    },
+    {
+      "epoch": 0.39770952628839146,
+      "grad_norm": 0.2930332941508139,
+      "learning_rate": 4.863364093451679e-05,
+      "loss": 0.1767,
+      "step": 764
+    },
+    {
+      "epoch": 0.39823008849557523,
+      "grad_norm": 0.29879453433882425,
+      "learning_rate": 4.862905590091341e-05,
+      "loss": 0.1649,
+      "step": 765
+    },
+    {
+      "epoch": 0.398750650702759,
+      "grad_norm": 0.30522662679820955,
+      "learning_rate": 4.8624463404154575e-05,
+      "loss": 0.1779,
+      "step": 766
+    },
+    {
+      "epoch": 0.3992712129099427,
+      "grad_norm": 0.29429201663178467,
+      "learning_rate": 4.8619863445690804e-05,
+      "loss": 0.1742,
+      "step": 767
+    },
+    {
+      "epoch": 0.3997917751171265,
+      "grad_norm": 0.2828872044198585,
+      "learning_rate": 4.861525602697499e-05,
+      "loss": 0.169,
+      "step": 768
+    },
+    {
+      "epoch": 0.40031233732431026,
+      "grad_norm": 0.31056313318997847,
+      "learning_rate": 4.861064114946236e-05,
+      "loss": 0.1728,
+      "step": 769
+    },
+    {
+      "epoch": 0.40083289953149404,
+      "grad_norm": 0.28343310391927706,
+      "learning_rate": 4.860601881461051e-05,
+      "loss": 0.171,
+      "step": 770
+    },
+    {
+      "epoch": 0.40135346173867775,
+      "grad_norm": 0.31511791024068336,
+      "learning_rate": 4.86013890238794e-05,
+      "loss": 0.1754,
+      "step": 771
+    },
+    {
+      "epoch": 0.4018740239458615,
+      "grad_norm": 0.3106130269067802,
+      "learning_rate": 4.85967517787313e-05,
+      "loss": 0.1755,
+      "step": 772
+    },
+    {
+      "epoch": 0.4023945861530453,
+      "grad_norm": 0.3216761568260499,
+      "learning_rate": 4.859210708063091e-05,
+      "loss": 0.1709,
+      "step": 773
+    },
+    {
+      "epoch": 0.40291514836022907,
+      "grad_norm": 0.3169473466081924,
+      "learning_rate": 4.858745493104521e-05,
+      "loss": 0.1677,
+      "step": 774
+    },
+    {
+      "epoch": 0.4034357105674128,
+      "grad_norm": 0.2858364547750457,
+      "learning_rate": 4.858279533144358e-05,
+      "loss": 0.173,
+      "step": 775
+    },
+    {
+      "epoch": 0.40395627277459656,
+      "grad_norm": 0.31534006666184045,
+      "learning_rate": 4.8578128283297736e-05,
+      "loss": 0.1701,
+      "step": 776
+    },
+    {
+      "epoch": 0.40447683498178033,
+      "grad_norm": 0.2978277533533381,
+      "learning_rate": 4.857345378808175e-05,
+      "loss": 0.1642,
+      "step": 777
+    },
+    {
+      "epoch": 0.4049973971889641,
+      "grad_norm": 0.30248488777067317,
+      "learning_rate": 4.856877184727204e-05,
+      "loss": 0.1763,
+      "step": 778
+    },
+    {
+      "epoch": 0.4055179593961478,
+      "grad_norm": 0.31081161780013244,
+      "learning_rate": 4.856408246234739e-05,
+      "loss": 0.1672,
+      "step": 779
+    },
+    {
+      "epoch": 0.4060385216033316,
+      "grad_norm": 0.33291363782085454,
+      "learning_rate": 4.855938563478892e-05,
+      "loss": 0.1773,
+      "step": 780
+    },
+    {
+      "epoch": 0.40655908381051536,
+      "grad_norm": 0.3108957897270069,
+      "learning_rate": 4.8554681366080104e-05,
+      "loss": 0.1693,
+      "step": 781
+    },
+    {
+      "epoch": 0.40707964601769914,
+      "grad_norm": 0.32908383574112354,
+      "learning_rate": 4.8549969657706775e-05,
+      "loss": 0.1708,
+      "step": 782
+    },
+    {
+      "epoch": 0.40760020822488285,
+      "grad_norm": 0.3107460244869553,
+      "learning_rate": 4.85452505111571e-05,
+      "loss": 0.1683,
+      "step": 783
+    },
+    {
+      "epoch": 0.4081207704320666,
+      "grad_norm": 0.31853260828349084,
+      "learning_rate": 4.8540523927921616e-05,
+      "loss": 0.1687,
+      "step": 784
+    },
+    {
+      "epoch": 0.4086413326392504,
+      "grad_norm": 0.3316073002641256,
+      "learning_rate": 4.85357899094932e-05,
+      "loss": 0.1684,
+      "step": 785
+    },
+    {
+      "epoch": 0.40916189484643417,
+      "grad_norm": 0.3027467331091613,
+      "learning_rate": 4.853104845736706e-05,
+      "loss": 0.1715,
+      "step": 786
+    },
+    {
+      "epoch": 0.4096824570536179,
+      "grad_norm": 0.3278257429615449,
+      "learning_rate": 4.852629957304078e-05,
+      "loss": 0.1773,
+      "step": 787
+    },
+    {
+      "epoch": 0.41020301926080166,
+      "grad_norm": 0.291153268830071,
+      "learning_rate": 4.8521543258014276e-05,
+      "loss": 0.1681,
+      "step": 788
+    },
+    {
+      "epoch": 0.41072358146798543,
+      "grad_norm": 0.3559444910615816,
+      "learning_rate": 4.8516779513789815e-05,
+      "loss": 0.1807,
+      "step": 789
+    },
+    {
+      "epoch": 0.4112441436751692,
+      "grad_norm": 0.3085670500497958,
+      "learning_rate": 4.8512008341872e-05,
+      "loss": 0.1719,
+      "step": 790
+    },
+    {
+      "epoch": 0.4117647058823529,
+      "grad_norm": 0.35264383435574276,
+      "learning_rate": 4.85072297437678e-05,
+      "loss": 0.1794,
+      "step": 791
+    },
+    {
+      "epoch": 0.4122852680895367,
+      "grad_norm": 0.3356960894753313,
+      "learning_rate": 4.850244372098651e-05,
+      "loss": 0.1769,
+      "step": 792
+    },
+    {
+      "epoch": 0.41280583029672047,
+      "grad_norm": 0.32740161142053137,
+      "learning_rate": 4.8497650275039795e-05,
+      "loss": 0.1783,
+      "step": 793
+    },
+    {
+      "epoch": 0.41332639250390424,
+      "grad_norm": 0.3351198703316088,
+      "learning_rate": 4.849284940744163e-05,
+      "loss": 0.1733,
+      "step": 794
+    },
+    {
+      "epoch": 0.41384695471108796,
+      "grad_norm": 0.312070876021951,
+      "learning_rate": 4.848804111970836e-05,
+      "loss": 0.1866,
+      "step": 795
+    },
+    {
+      "epoch": 0.4143675169182717,
+      "grad_norm": 0.3282848492692099,
+      "learning_rate": 4.8483225413358663e-05,
+      "loss": 0.1789,
+      "step": 796
+    },
+    {
+      "epoch": 0.4148880791254555,
+      "grad_norm": 0.286266024247938,
+      "learning_rate": 4.8478402289913566e-05,
+      "loss": 0.1698,
+      "step": 797
+    },
+    {
+      "epoch": 0.4154086413326393,
+      "grad_norm": 0.33058577040283377,
+      "learning_rate": 4.847357175089643e-05,
+      "loss": 0.1684,
+      "step": 798
+    },
+    {
+      "epoch": 0.415929203539823,
+      "grad_norm": 0.2934521521206386,
+      "learning_rate": 4.846873379783298e-05,
+      "loss": 0.1753,
+      "step": 799
+    },
+    {
+      "epoch": 0.41644976574700676,
+      "grad_norm": 0.31483778450446187,
+      "learning_rate": 4.846388843225125e-05,
+      "loss": 0.1665,
+      "step": 800
+    },
+    {
+      "epoch": 0.41697032795419053,
+      "grad_norm": 0.30494366672931933,
+      "learning_rate": 4.845903565568164e-05,
+      "loss": 0.1791,
+      "step": 801
+    },
+    {
+      "epoch": 0.4174908901613743,
+      "grad_norm": 0.2737424192822608,
+      "learning_rate": 4.845417546965688e-05,
+      "loss": 0.1689,
+      "step": 802
+    },
+    {
+      "epoch": 0.418011452368558,
+      "grad_norm": 0.3176283439520924,
+      "learning_rate": 4.844930787571204e-05,
+      "loss": 0.1728,
+      "step": 803
+    },
+    {
+      "epoch": 0.4185320145757418,
+      "grad_norm": 0.3013277659758763,
+      "learning_rate": 4.844443287538454e-05,
+      "loss": 0.1793,
+      "step": 804
+    },
+    {
+      "epoch": 0.41905257678292557,
+      "grad_norm": 0.31841896666187924,
+      "learning_rate": 4.8439550470214124e-05,
+      "loss": 0.1725,
+      "step": 805
+    },
+    {
+      "epoch": 0.41957313899010934,
+      "grad_norm": 0.32537384441510847,
+      "learning_rate": 4.8434660661742894e-05,
+      "loss": 0.1726,
+      "step": 806
+    },
+    {
+      "epoch": 0.42009370119729306,
+      "grad_norm": 0.2918667124673871,
+      "learning_rate": 4.8429763451515263e-05,
+      "loss": 0.1713,
+      "step": 807
+    },
+    {
+      "epoch": 0.42061426340447683,
+      "grad_norm": 0.3068100516789931,
+      "learning_rate": 4.842485884107801e-05,
+      "loss": 0.177,
+      "step": 808
+    },
+    {
+      "epoch": 0.4211348256116606,
+      "grad_norm": 0.32460789829684644,
+      "learning_rate": 4.8419946831980236e-05,
+      "loss": 0.1742,
+      "step": 809
+    },
+    {
+      "epoch": 0.4216553878188444,
+      "grad_norm": 0.2946290268979509,
+      "learning_rate": 4.8415027425773386e-05,
+      "loss": 0.1681,
+      "step": 810
+    },
+    {
+      "epoch": 0.4221759500260281,
+      "grad_norm": 0.3258905754569639,
+      "learning_rate": 4.841010062401123e-05,
+      "loss": 0.1741,
+      "step": 811
+    },
+    {
+      "epoch": 0.42269651223321186,
+      "grad_norm": 0.29790984995596925,
+      "learning_rate": 4.840516642824988e-05,
+      "loss": 0.1694,
+      "step": 812
+    },
+    {
+      "epoch": 0.42321707444039564,
+      "grad_norm": 0.31430235551801683,
+      "learning_rate": 4.8400224840047795e-05,
+      "loss": 0.1692,
+      "step": 813
+    },
+    {
+      "epoch": 0.4237376366475794,
+      "grad_norm": 0.28497704460198,
+      "learning_rate": 4.839527586096575e-05,
+      "loss": 0.1692,
+      "step": 814
+    },
+    {
+      "epoch": 0.4242581988547631,
+      "grad_norm": 0.3251548281082162,
+      "learning_rate": 4.839031949256687e-05,
+      "loss": 0.1786,
+      "step": 815
+    },
+    {
+      "epoch": 0.4247787610619469,
+      "grad_norm": 0.3037172282531429,
+      "learning_rate": 4.838535573641661e-05,
+      "loss": 0.1784,
+      "step": 816
+    },
+    {
+      "epoch": 0.42529932326913067,
+      "grad_norm": 0.2921332397976667,
+      "learning_rate": 4.838038459408273e-05,
+      "loss": 0.1627,
+      "step": 817
+    },
+    {
+      "epoch": 0.42581988547631444,
+      "grad_norm": 0.2871659618387349,
+      "learning_rate": 4.837540606713538e-05,
+      "loss": 0.1669,
+      "step": 818
+    },
+    {
+      "epoch": 0.42634044768349816,
+      "grad_norm": 0.3022037954071766,
+      "learning_rate": 4.837042015714698e-05,
+      "loss": 0.1737,
+      "step": 819
+    },
+    {
+      "epoch": 0.42686100989068193,
+      "grad_norm": 0.31565270517728566,
+      "learning_rate": 4.8365426865692345e-05,
+      "loss": 0.1787,
+      "step": 820
+    },
+    {
+      "epoch": 0.4273815720978657,
+      "grad_norm": 0.32793352347608645,
+      "learning_rate": 4.836042619434856e-05,
+      "loss": 0.1748,
+      "step": 821
+    },
+    {
+      "epoch": 0.4279021343050495,
+      "grad_norm": 0.317161732106535,
+      "learning_rate": 4.835541814469509e-05,
+      "loss": 0.1782,
+      "step": 822
+    },
+    {
+      "epoch": 0.4284226965122332,
+      "grad_norm": 0.3049128671196971,
+      "learning_rate": 4.83504027183137e-05,
+      "loss": 0.1816,
+      "step": 823
+    },
+    {
+      "epoch": 0.42894325871941696,
+      "grad_norm": 0.28919765868015046,
+      "learning_rate": 4.8345379916788505e-05,
+      "loss": 0.1672,
+      "step": 824
+    },
+    {
+      "epoch": 0.42946382092660074,
+      "grad_norm": 0.33488488955581175,
+      "learning_rate": 4.834034974170592e-05,
+      "loss": 0.1688,
+      "step": 825
+    },
+    {
+      "epoch": 0.4299843831337845,
+      "grad_norm": 0.31612472215220894,
+      "learning_rate": 4.833531219465473e-05,
+      "loss": 0.1639,
+      "step": 826
+    },
+    {
+      "epoch": 0.4305049453409682,
+      "grad_norm": 0.30731831334383447,
+      "learning_rate": 4.8330267277226006e-05,
+      "loss": 0.1712,
+      "step": 827
+    },
+    {
+      "epoch": 0.431025507548152,
+      "grad_norm": 0.31217223215929757,
+      "learning_rate": 4.832521499101319e-05,
+      "loss": 0.176,
+      "step": 828
+    },
+    {
+      "epoch": 0.43154606975533577,
+      "grad_norm": 0.2795778724288852,
+      "learning_rate": 4.8320155337612014e-05,
+      "loss": 0.1703,
+      "step": 829
+    },
+    {
+      "epoch": 0.43206663196251954,
+      "grad_norm": 0.2970259430025653,
+      "learning_rate": 4.831508831862055e-05,
+      "loss": 0.1693,
+      "step": 830
+    },
+    {
+      "epoch": 0.43258719416970326,
+      "grad_norm": 0.31169918235532607,
+      "learning_rate": 4.8310013935639206e-05,
+      "loss": 0.1758,
+      "step": 831
+    },
+    {
+      "epoch": 0.43310775637688703,
+      "grad_norm": 0.3282580950489392,
+      "learning_rate": 4.830493219027071e-05,
+      "loss": 0.1701,
+      "step": 832
+    },
+    {
+      "epoch": 0.4336283185840708,
+      "grad_norm": 0.3035542216950865,
+      "learning_rate": 4.829984308412011e-05,
+      "loss": 0.1759,
+      "step": 833
+    },
+    {
+      "epoch": 0.4341488807912546,
+      "grad_norm": 0.3170042392985621,
+      "learning_rate": 4.8294746618794786e-05,
+      "loss": 0.1714,
+      "step": 834
+    },
+    {
+      "epoch": 0.4346694429984383,
+      "grad_norm": 0.30492935178098757,
+      "learning_rate": 4.8289642795904433e-05,
+      "loss": 0.1769,
+      "step": 835
+    },
+    {
+      "epoch": 0.43519000520562207,
+      "grad_norm": 0.3023169478849347,
+      "learning_rate": 4.828453161706108e-05,
+      "loss": 0.1684,
+      "step": 836
+    },
+    {
+      "epoch": 0.43571056741280584,
+      "grad_norm": 0.2934309140121328,
+      "learning_rate": 4.8279413083879063e-05,
+      "loss": 0.1726,
+      "step": 837
+    },
+    {
+      "epoch": 0.4362311296199896,
+      "grad_norm": 0.33964067234560835,
+      "learning_rate": 4.827428719797508e-05,
+      "loss": 0.1811,
+      "step": 838
+    },
+    {
+      "epoch": 0.4367516918271733,
+      "grad_norm": 0.3239469204735507,
+      "learning_rate": 4.8269153960968094e-05,
+      "loss": 0.1801,
+      "step": 839
+    },
+    {
+      "epoch": 0.4372722540343571,
+      "grad_norm": 0.34044634387274025,
+      "learning_rate": 4.8264013374479446e-05,
+      "loss": 0.1734,
+      "step": 840
+    },
+    {
+      "epoch": 0.43779281624154087,
+      "grad_norm": 0.3143167652108177,
+      "learning_rate": 4.825886544013275e-05,
+      "loss": 0.1742,
+      "step": 841
+    },
+    {
+      "epoch": 0.43831337844872464,
+      "grad_norm": 0.2944577153853259,
+      "learning_rate": 4.825371015955398e-05,
+      "loss": 0.1653,
+      "step": 842
+    },
+    {
+      "epoch": 0.43883394065590836,
+      "grad_norm": 0.3052896712979636,
+      "learning_rate": 4.82485475343714e-05,
+      "loss": 0.1704,
+      "step": 843
+    },
+    {
+      "epoch": 0.43935450286309213,
+      "grad_norm": 0.3177119837657539,
+      "learning_rate": 4.8243377566215616e-05,
+      "loss": 0.1719,
+      "step": 844
+    },
+    {
+      "epoch": 0.4398750650702759,
+      "grad_norm": 0.30859779939404325,
+      "learning_rate": 4.8238200256719554e-05,
+      "loss": 0.1779,
+      "step": 845
+    },
+    {
+      "epoch": 0.4403956272774597,
+      "grad_norm": 0.3226603688721493,
+      "learning_rate": 4.823301560751843e-05,
+      "loss": 0.1686,
+      "step": 846
+    },
+    {
+      "epoch": 0.4409161894846434,
+      "grad_norm": 0.29317543800324647,
+      "learning_rate": 4.82278236202498e-05,
+      "loss": 0.1673,
+      "step": 847
+    },
+    {
+      "epoch": 0.44143675169182717,
+      "grad_norm": 0.31460096487006833,
+      "learning_rate": 4.8222624296553554e-05,
+      "loss": 0.1753,
+      "step": 848
+    },
+    {
+      "epoch": 0.44195731389901094,
+      "grad_norm": 0.2919013066978429,
+      "learning_rate": 4.821741763807186e-05,
+      "loss": 0.1694,
+      "step": 849
+    },
+    {
+      "epoch": 0.4424778761061947,
+      "grad_norm": 0.33306663050219115,
+      "learning_rate": 4.821220364644923e-05,
+      "loss": 0.1725,
+      "step": 850
+    },
+    {
+      "epoch": 0.44299843831337843,
+      "grad_norm": 0.32508594947874425,
+      "learning_rate": 4.8206982323332485e-05,
+      "loss": 0.1703,
+      "step": 851
+    },
+    {
+      "epoch": 0.4435190005205622,
+      "grad_norm": 0.312739185199514,
+      "learning_rate": 4.820175367037076e-05,
+      "loss": 0.1747,
+      "step": 852
+    },
+    {
+      "epoch": 0.444039562727746,
+      "grad_norm": 0.3139934324616478,
+      "learning_rate": 4.8196517689215515e-05,
+      "loss": 0.1697,
+      "step": 853
+    },
+    {
+      "epoch": 0.44456012493492975,
+      "grad_norm": 0.28334748849023433,
+      "learning_rate": 4.8191274381520515e-05,
+      "loss": 0.1701,
+      "step": 854
+    },
+    {
+      "epoch": 0.44508068714211346,
+      "grad_norm": 0.31522836280773053,
+      "learning_rate": 4.818602374894182e-05,
+      "loss": 0.1724,
+      "step": 855
+    },
+    {
+      "epoch": 0.44560124934929723,
+      "grad_norm": 0.28857799909498244,
+      "learning_rate": 4.8180765793137856e-05,
+      "loss": 0.1748,
+      "step": 856
+    },
+    {
+      "epoch": 0.446121811556481,
+      "grad_norm": 0.2972500213315476,
+      "learning_rate": 4.817550051576931e-05,
+      "loss": 0.1653,
+      "step": 857
+    },
+    {
+      "epoch": 0.4466423737636648,
+      "grad_norm": 0.29255400412368143,
+      "learning_rate": 4.81702279184992e-05,
+      "loss": 0.1615,
+      "step": 858
+    },
+    {
+      "epoch": 0.4471629359708485,
+      "grad_norm": 0.2886655749254805,
+      "learning_rate": 4.8164948002992874e-05,
+      "loss": 0.1634,
+      "step": 859
+    },
+    {
+      "epoch": 0.44768349817803227,
+      "grad_norm": 0.31437191386746677,
+      "learning_rate": 4.815966077091796e-05,
+      "loss": 0.1749,
+      "step": 860
+    },
+    {
+      "epoch": 0.44820406038521604,
+      "grad_norm": 0.29304891186896304,
+      "learning_rate": 4.815436622394441e-05,
+      "loss": 0.1708,
+      "step": 861
+    },
+    {
+      "epoch": 0.4487246225923998,
+      "grad_norm": 0.31745839082325317,
+      "learning_rate": 4.814906436374451e-05,
+      "loss": 0.1686,
+      "step": 862
+    },
+    {
+      "epoch": 0.44924518479958353,
+      "grad_norm": 0.3046376488171113,
+      "learning_rate": 4.814375519199281e-05,
+      "loss": 0.1654,
+      "step": 863
+    },
+    {
+      "epoch": 0.4497657470067673,
+      "grad_norm": 0.30408750750055136,
+      "learning_rate": 4.8138438710366204e-05,
+      "loss": 0.1692,
+      "step": 864
+    },
+    {
+      "epoch": 0.4502863092139511,
+      "grad_norm": 0.2900405154788453,
+      "learning_rate": 4.813311492054388e-05,
+      "loss": 0.1719,
+      "step": 865
+    },
+    {
+      "epoch": 0.45080687142113485,
+      "grad_norm": 0.307651302837432,
+      "learning_rate": 4.8127783824207344e-05,
+      "loss": 0.1768,
+      "step": 866
+    },
+    {
+      "epoch": 0.45132743362831856,
+      "grad_norm": 0.3003033631697678,
+      "learning_rate": 4.812244542304041e-05,
+      "loss": 0.1687,
+      "step": 867
+    },
+    {
+      "epoch": 0.45184799583550234,
+      "grad_norm": 0.3226224029065988,
+      "learning_rate": 4.811709971872918e-05,
+      "loss": 0.1737,
+      "step": 868
+    },
+    {
+      "epoch": 0.4523685580426861,
+      "grad_norm": 0.3431087957838699,
+      "learning_rate": 4.8111746712962066e-05,
+      "loss": 0.1792,
+      "step": 869
+    },
+    {
+      "epoch": 0.4528891202498699,
+      "grad_norm": 0.3241044954870311,
+      "learning_rate": 4.810638640742983e-05,
+      "loss": 0.1747,
+      "step": 870
+    },
+    {
+      "epoch": 0.4534096824570536,
+      "grad_norm": 0.29640273607618955,
+      "learning_rate": 4.810101880382548e-05,
+      "loss": 0.165,
+      "step": 871
+    },
+    {
+      "epoch": 0.45393024466423737,
+      "grad_norm": 0.2992144530417321,
+      "learning_rate": 4.809564390384437e-05,
+      "loss": 0.1734,
+      "step": 872
+    },
+    {
+      "epoch": 0.45445080687142114,
+      "grad_norm": 0.33183641077540893,
+      "learning_rate": 4.809026170918414e-05,
+      "loss": 0.1737,
+      "step": 873
+    },
+    {
+      "epoch": 0.4549713690786049,
+      "grad_norm": 0.28653542372832375,
+      "learning_rate": 4.808487222154472e-05,
+      "loss": 0.1634,
+      "step": 874
+    },
+    {
+      "epoch": 0.45549193128578863,
+      "grad_norm": 0.32334785309979097,
+      "learning_rate": 4.807947544262839e-05,
+      "loss": 0.1733,
+      "step": 875
+    },
+    {
+      "epoch": 0.4560124934929724,
+      "grad_norm": 0.3030305396087403,
+      "learning_rate": 4.807407137413967e-05,
+      "loss": 0.1658,
+      "step": 876
+    },
+    {
+      "epoch": 0.4565330557001562,
+      "grad_norm": 0.296769998908498,
+      "learning_rate": 4.806866001778545e-05,
+      "loss": 0.1767,
+      "step": 877
+    },
+    {
+      "epoch": 0.45705361790733995,
+      "grad_norm": 0.31049567683022117,
+      "learning_rate": 4.806324137527487e-05,
+      "loss": 0.1797,
+      "step": 878
+    },
+    {
+      "epoch": 0.45757418011452367,
+      "grad_norm": 0.30534013515879393,
+      "learning_rate": 4.8057815448319394e-05,
+      "loss": 0.1691,
+      "step": 879
+    },
+    {
+      "epoch": 0.45809474232170744,
+      "grad_norm": 0.31938583186414876,
+      "learning_rate": 4.8052382238632774e-05,
+      "loss": 0.1729,
+      "step": 880
+    },
+    {
+      "epoch": 0.4586153045288912,
+      "grad_norm": 0.3046756327864025,
+      "learning_rate": 4.804694174793108e-05,
+      "loss": 0.1623,
+      "step": 881
+    },
+    {
+      "epoch": 0.459135866736075,
+      "grad_norm": 0.30819124791549257,
+      "learning_rate": 4.8041493977932685e-05,
+      "loss": 0.1739,
+      "step": 882
+    },
+    {
+      "epoch": 0.4596564289432587,
+      "grad_norm": 0.29113272544105995,
+      "learning_rate": 4.803603893035822e-05,
+      "loss": 0.1635,
+      "step": 883
+    },
+    {
+      "epoch": 0.46017699115044247,
+      "grad_norm": 0.2908354554872441,
+      "learning_rate": 4.803057660693065e-05,
+      "loss": 0.1642,
+      "step": 884
+    },
+    {
+      "epoch": 0.46069755335762624,
+      "grad_norm": 0.3228709959167957,
+      "learning_rate": 4.8025107009375246e-05,
+      "loss": 0.1687,
+      "step": 885
+    },
+    {
+      "epoch": 0.46121811556481,
+      "grad_norm": 0.2875173089594343,
+      "learning_rate": 4.8019630139419555e-05,
+      "loss": 0.1779,
+      "step": 886
+    },
+    {
+      "epoch": 0.46173867777199373,
+      "grad_norm": 0.33584341705384174,
+      "learning_rate": 4.8014145998793416e-05,
+      "loss": 0.1733,
+      "step": 887
+    },
+    {
+      "epoch": 0.4622592399791775,
+      "grad_norm": 0.2896189188284826,
+      "learning_rate": 4.8008654589228984e-05,
+      "loss": 0.1684,
+      "step": 888
+    },
+    {
+      "epoch": 0.4627798021863613,
+      "grad_norm": 0.3260162748091208,
+      "learning_rate": 4.800315591246071e-05,
+      "loss": 0.1679,
+      "step": 889
+    },
+    {
+      "epoch": 0.46330036439354505,
+      "grad_norm": 0.28906298710463635,
+      "learning_rate": 4.799764997022532e-05,
+      "loss": 0.1748,
+      "step": 890
+    },
+    {
+      "epoch": 0.46382092660072877,
+      "grad_norm": 0.28621158344042275,
+      "learning_rate": 4.799213676426185e-05,
+      "loss": 0.1745,
+      "step": 891
+    },
+    {
+      "epoch": 0.46434148880791254,
+      "grad_norm": 0.309418801701727,
+      "learning_rate": 4.798661629631163e-05,
+      "loss": 0.168,
+      "step": 892
+    },
+    {
+      "epoch": 0.4648620510150963,
+      "grad_norm": 0.2872933837593369,
+      "learning_rate": 4.798108856811828e-05,
+      "loss": 0.1695,
+      "step": 893
+    },
+    {
+      "epoch": 0.4653826132222801,
+      "grad_norm": 0.31032860424563535,
+      "learning_rate": 4.7975553581427715e-05,
+      "loss": 0.164,
+      "step": 894
+    },
+    {
+      "epoch": 0.4659031754294638,
+      "grad_norm": 0.2944281461376994,
+      "learning_rate": 4.797001133798813e-05,
+      "loss": 0.1715,
+      "step": 895
+    },
+    {
+      "epoch": 0.4664237376366476,
+      "grad_norm": 0.3185115115776022,
+      "learning_rate": 4.796446183955003e-05,
+      "loss": 0.1664,
+      "step": 896
+    },
+    {
+      "epoch": 0.46694429984383135,
+      "grad_norm": 0.2898044563512308,
+      "learning_rate": 4.795890508786622e-05,
+      "loss": 0.1681,
+      "step": 897
+    },
+    {
+      "epoch": 0.4674648620510151,
+      "grad_norm": 0.30483790574973163,
+      "learning_rate": 4.795334108469176e-05,
+      "loss": 0.1687,
+      "step": 898
+    },
+    {
+      "epoch": 0.46798542425819883,
+      "grad_norm": 0.30599381056013947,
+      "learning_rate": 4.794776983178403e-05,
+      "loss": 0.1711,
+      "step": 899
+    },
+    {
+      "epoch": 0.4685059864653826,
+      "grad_norm": 0.3093671697842366,
+      "learning_rate": 4.794219133090269e-05,
+      "loss": 0.1696,
+      "step": 900
+    },
+    {
+      "epoch": 0.4690265486725664,
+      "grad_norm": 0.3000416412774675,
+      "learning_rate": 4.793660558380969e-05,
+      "loss": 0.1748,
+      "step": 901
+    },
+    {
+      "epoch": 0.46954711087975015,
+      "grad_norm": 0.3242303350202561,
+      "learning_rate": 4.793101259226927e-05,
+      "loss": 0.1775,
+      "step": 902
+    },
+    {
+      "epoch": 0.47006767308693387,
+      "grad_norm": 0.31647313130174715,
+      "learning_rate": 4.792541235804796e-05,
+      "loss": 0.167,
+      "step": 903
+    },
+    {
+      "epoch": 0.47058823529411764,
+      "grad_norm": 0.32293924392398315,
+      "learning_rate": 4.791980488291456e-05,
+      "loss": 0.1763,
+      "step": 904
+    },
+    {
+      "epoch": 0.4711087975013014,
+      "grad_norm": 0.29224165341381086,
+      "learning_rate": 4.7914190168640196e-05,
+      "loss": 0.1678,
+      "step": 905
+    },
+    {
+      "epoch": 0.4716293597084852,
+      "grad_norm": 0.30657161857494225,
+      "learning_rate": 4.790856821699823e-05,
+      "loss": 0.1739,
+      "step": 906
+    },
+    {
+      "epoch": 0.4721499219156689,
+      "grad_norm": 0.29572827932510726,
+      "learning_rate": 4.790293902976435e-05,
+      "loss": 0.1635,
+      "step": 907
+    },
+    {
+      "epoch": 0.4726704841228527,
+      "grad_norm": 0.3140759869076264,
+      "learning_rate": 4.789730260871651e-05,
+      "loss": 0.169,
+      "step": 908
+    },
+    {
+      "epoch": 0.47319104633003645,
+      "grad_norm": 0.30710825268098163,
+      "learning_rate": 4.7891658955634964e-05,
+      "loss": 0.1658,
+      "step": 909
+    },
+    {
+      "epoch": 0.4737116085372202,
+      "grad_norm": 0.31793790159265933,
+      "learning_rate": 4.7886008072302235e-05,
+      "loss": 0.1753,
+      "step": 910
+    },
+    {
+      "epoch": 0.47423217074440394,
+      "grad_norm": 0.3014852750545304,
+      "learning_rate": 4.788034996050314e-05,
+      "loss": 0.1709,
+      "step": 911
+    },
+    {
+      "epoch": 0.4747527329515877,
+      "grad_norm": 0.30747571574129023,
+      "learning_rate": 4.787468462202476e-05,
+      "loss": 0.1612,
+      "step": 912
+    },
+    {
+      "epoch": 0.4752732951587715,
+      "grad_norm": 0.3065142379561827,
+      "learning_rate": 4.786901205865647e-05,
+      "loss": 0.1621,
+      "step": 913
+    },
+    {
+      "epoch": 0.47579385736595525,
+      "grad_norm": 0.3017055320367006,
+      "learning_rate": 4.786333227218995e-05,
+      "loss": 0.1759,
+      "step": 914
+    },
+    {
+      "epoch": 0.47631441957313897,
+      "grad_norm": 0.3063810493732478,
+      "learning_rate": 4.785764526441913e-05,
+      "loss": 0.1766,
+      "step": 915
+    },
+    {
+      "epoch": 0.47683498178032274,
+      "grad_norm": 0.3139212066831126,
+      "learning_rate": 4.7851951037140234e-05,
+      "loss": 0.1686,
+      "step": 916
+    },
+    {
+      "epoch": 0.4773555439875065,
+      "grad_norm": 0.2976317597214064,
+      "learning_rate": 4.784624959215176e-05,
+      "loss": 0.169,
+      "step": 917
+    },
+    {
+      "epoch": 0.4778761061946903,
+      "grad_norm": 0.28710602424260284,
+      "learning_rate": 4.78405409312545e-05,
+      "loss": 0.1653,
+      "step": 918
+    },
+    {
+      "epoch": 0.478396668401874,
+      "grad_norm": 0.30826508978927336,
+      "learning_rate": 4.783482505625149e-05,
+      "loss": 0.1716,
+      "step": 919
+    },
+    {
+      "epoch": 0.4789172306090578,
+      "grad_norm": 0.31724938862767077,
+      "learning_rate": 4.7829101968948095e-05,
+      "loss": 0.1682,
+      "step": 920
+    },
+    {
+      "epoch": 0.47943779281624155,
+      "grad_norm": 0.30073475133813604,
+      "learning_rate": 4.782337167115193e-05,
+      "loss": 0.1666,
+      "step": 921
+    },
+    {
+      "epoch": 0.4799583550234253,
+      "grad_norm": 0.2941398432084064,
+      "learning_rate": 4.7817634164672875e-05,
+      "loss": 0.1654,
+      "step": 922
+    },
+    {
+      "epoch": 0.48047891723060904,
+      "grad_norm": 0.30183359482377575,
+      "learning_rate": 4.7811889451323114e-05,
+      "loss": 0.1704,
+      "step": 923
+    },
+    {
+      "epoch": 0.4809994794377928,
+      "grad_norm": 0.27257120459980766,
+      "learning_rate": 4.7806137532917085e-05,
+      "loss": 0.1588,
+      "step": 924
+    },
+    {
+      "epoch": 0.4815200416449766,
+      "grad_norm": 0.3017092170804171,
+      "learning_rate": 4.780037841127152e-05,
+      "loss": 0.1632,
+      "step": 925
+    },
+    {
+      "epoch": 0.48204060385216035,
+      "grad_norm": 0.31115577416098184,
+      "learning_rate": 4.779461208820541e-05,
+      "loss": 0.1723,
+      "step": 926
+    },
+    {
+      "epoch": 0.48256116605934407,
+      "grad_norm": 0.294393697239707,
+      "learning_rate": 4.778883856554004e-05,
+      "loss": 0.1709,
+      "step": 927
+    },
+    {
+      "epoch": 0.48308172826652784,
+      "grad_norm": 0.31179094440237864,
+      "learning_rate": 4.778305784509894e-05,
+      "loss": 0.1837,
+      "step": 928
+    },
+    {
+      "epoch": 0.4836022904737116,
+      "grad_norm": 0.29561645952616644,
+      "learning_rate": 4.7777269928707946e-05,
+      "loss": 0.1679,
+      "step": 929
+    },
+    {
+      "epoch": 0.4841228526808954,
+      "grad_norm": 0.31643861362801334,
+      "learning_rate": 4.777147481819515e-05,
+      "loss": 0.1652,
+      "step": 930
+    },
+    {
+      "epoch": 0.4846434148880791,
+      "grad_norm": 0.2990057285412248,
+      "learning_rate": 4.776567251539091e-05,
+      "loss": 0.1652,
+      "step": 931
+    },
+    {
+      "epoch": 0.4851639770952629,
+      "grad_norm": 0.31236323289513435,
+      "learning_rate": 4.7759863022127864e-05,
+      "loss": 0.1704,
+      "step": 932
+    },
+    {
+      "epoch": 0.48568453930244665,
+      "grad_norm": 0.3453944960772348,
+      "learning_rate": 4.775404634024093e-05,
+      "loss": 0.1748,
+      "step": 933
+    },
+    {
+      "epoch": 0.4862051015096304,
+      "grad_norm": 0.3629879836622649,
+      "learning_rate": 4.7748222471567275e-05,
+      "loss": 0.177,
+      "step": 934
+    },
+    {
+      "epoch": 0.48672566371681414,
+      "grad_norm": 0.3120022773445736,
+      "learning_rate": 4.7742391417946345e-05,
+      "loss": 0.1741,
+      "step": 935
+    },
+    {
+      "epoch": 0.4872462259239979,
+      "grad_norm": 0.3335724998214946,
+      "learning_rate": 4.773655318121987e-05,
+      "loss": 0.1748,
+      "step": 936
+    },
+    {
+      "epoch": 0.4877667881311817,
+      "grad_norm": 0.34481706739223045,
+      "learning_rate": 4.7730707763231844e-05,
+      "loss": 0.1764,
+      "step": 937
+    },
+    {
+      "epoch": 0.48828735033836546,
+      "grad_norm": 0.32529182532452,
+      "learning_rate": 4.7724855165828497e-05,
+      "loss": 0.1689,
+      "step": 938
+    },
+    {
+      "epoch": 0.4888079125455492,
+      "grad_norm": 0.32960723240412076,
+      "learning_rate": 4.771899539085837e-05,
+      "loss": 0.1706,
+      "step": 939
+    },
+    {
+      "epoch": 0.48932847475273294,
+      "grad_norm": 0.331766989087599,
+      "learning_rate": 4.7713128440172244e-05,
+      "loss": 0.1691,
+      "step": 940
+    },
+    {
+      "epoch": 0.4898490369599167,
+      "grad_norm": 0.32802357802208776,
+      "learning_rate": 4.770725431562318e-05,
+      "loss": 0.1753,
+      "step": 941
+    },
+    {
+      "epoch": 0.4903695991671005,
+      "grad_norm": 0.30860493564486086,
+      "learning_rate": 4.77013730190665e-05,
+      "loss": 0.1728,
+      "step": 942
+    },
+    {
+      "epoch": 0.4908901613742842,
+      "grad_norm": 0.3022595357659953,
+      "learning_rate": 4.7695484552359794e-05,
+      "loss": 0.1722,
+      "step": 943
+    },
+    {
+      "epoch": 0.491410723581468,
+      "grad_norm": 0.2934783342140039,
+      "learning_rate": 4.7689588917362905e-05,
+      "loss": 0.1608,
+      "step": 944
+    },
+    {
+      "epoch": 0.49193128578865175,
+      "grad_norm": 0.28487899547455736,
+      "learning_rate": 4.768368611593795e-05,
+      "loss": 0.167,
+      "step": 945
+    },
+    {
+      "epoch": 0.4924518479958355,
+      "grad_norm": 0.2990203794907809,
+      "learning_rate": 4.7677776149949315e-05,
+      "loss": 0.1651,
+      "step": 946
+    },
+    {
+      "epoch": 0.49297241020301924,
+      "grad_norm": 0.32061252229163467,
+      "learning_rate": 4.767185902126364e-05,
+      "loss": 0.1626,
+      "step": 947
+    },
+    {
+      "epoch": 0.493492972410203,
+      "grad_norm": 0.2807651296414107,
+      "learning_rate": 4.7665934731749825e-05,
+      "loss": 0.1668,
+      "step": 948
+    },
+    {
+      "epoch": 0.4940135346173868,
+      "grad_norm": 0.2967222786779605,
+      "learning_rate": 4.7660003283279045e-05,
+      "loss": 0.1721,
+      "step": 949
+    },
+    {
+      "epoch": 0.49453409682457056,
+      "grad_norm": 0.29814117868358586,
+      "learning_rate": 4.765406467772472e-05,
+      "loss": 0.1778,
+      "step": 950
+    },
+    {
+      "epoch": 0.4950546590317543,
+      "grad_norm": 0.2788975714416808,
+      "learning_rate": 4.7648118916962535e-05,
+      "loss": 0.1681,
+      "step": 951
+    },
+    {
+      "epoch": 0.49557522123893805,
+      "grad_norm": 0.28418700831999905,
+      "learning_rate": 4.7642166002870455e-05,
+      "loss": 0.1727,
+      "step": 952
+    },
+    {
+      "epoch": 0.4960957834461218,
+      "grad_norm": 0.29906644459887316,
+      "learning_rate": 4.763620593732867e-05,
+      "loss": 0.1623,
+      "step": 953
+    },
+    {
+      "epoch": 0.4966163456533056,
+      "grad_norm": 0.26491394611771796,
+      "learning_rate": 4.763023872221965e-05,
+      "loss": 0.1677,
+      "step": 954
+    },
+    {
+      "epoch": 0.4971369078604893,
+      "grad_norm": 0.27853554570105976,
+      "learning_rate": 4.762426435942812e-05,
+      "loss": 0.1738,
+      "step": 955
+    },
+    {
+      "epoch": 0.4976574700676731,
+      "grad_norm": 0.3177673300895612,
+      "learning_rate": 4.761828285084107e-05,
+      "loss": 0.1758,
+      "step": 956
+    },
+    {
+      "epoch": 0.49817803227485685,
+      "grad_norm": 0.2822714144980152,
+      "learning_rate": 4.761229419834772e-05,
+      "loss": 0.1688,
+      "step": 957
+    },
+    {
+      "epoch": 0.4986985944820406,
+      "grad_norm": 0.2931606790760609,
+      "learning_rate": 4.7606298403839586e-05,
+      "loss": 0.1714,
+      "step": 958
+    },
+    {
+      "epoch": 0.49921915668922434,
+      "grad_norm": 0.3067169511093501,
+      "learning_rate": 4.760029546921041e-05,
+      "loss": 0.1707,
+      "step": 959
+    },
+    {
+      "epoch": 0.4997397188964081,
+      "grad_norm": 0.28137140614635753,
+      "learning_rate": 4.7594285396356184e-05,
+      "loss": 0.1761,
+      "step": 960
+    },
+    {
+      "epoch": 0.5002602811035919,
+      "grad_norm": 0.28844189796145525,
+      "learning_rate": 4.75882681871752e-05,
+      "loss": 0.1666,
+      "step": 961
+    },
+    {
+      "epoch": 0.5007808433107757,
+      "grad_norm": 0.2878826071675348,
+      "learning_rate": 4.758224384356795e-05,
+      "loss": 0.1683,
+      "step": 962
+    },
+    {
+      "epoch": 0.5013014055179594,
+      "grad_norm": 0.2998487365957084,
+      "learning_rate": 4.75762123674372e-05,
+      "loss": 0.1686,
+      "step": 963
+    },
+    {
+      "epoch": 0.5018219677251432,
+      "grad_norm": 0.27860322662131287,
+      "learning_rate": 4.757017376068799e-05,
+      "loss": 0.1729,
+      "step": 964
+    },
+    {
+      "epoch": 0.5023425299323269,
+      "grad_norm": 0.27665779300643556,
+      "learning_rate": 4.7564128025227566e-05,
+      "loss": 0.1669,
+      "step": 965
+    },
+    {
+      "epoch": 0.5028630921395106,
+      "grad_norm": 0.2961362102520933,
+      "learning_rate": 4.755807516296548e-05,
+      "loss": 0.1731,
+      "step": 966
+    },
+    {
+      "epoch": 0.5033836543466944,
+      "grad_norm": 0.28373948335406174,
+      "learning_rate": 4.755201517581349e-05,
+      "loss": 0.1647,
+      "step": 967
+    },
+    {
+      "epoch": 0.5039042165538782,
+      "grad_norm": 0.3027972283519984,
+      "learning_rate": 4.754594806568562e-05,
+      "loss": 0.1744,
+      "step": 968
+    },
+    {
+      "epoch": 0.504424778761062,
+      "grad_norm": 0.32929413135285035,
+      "learning_rate": 4.753987383449816e-05,
+      "loss": 0.1677,
+      "step": 969
+    },
+    {
+      "epoch": 0.5049453409682457,
+      "grad_norm": 0.29541171416091216,
+      "learning_rate": 4.753379248416963e-05,
+      "loss": 0.1718,
+      "step": 970
+    },
+    {
+      "epoch": 0.5054659031754295,
+      "grad_norm": 0.3445502102948572,
+      "learning_rate": 4.75277040166208e-05,
+      "loss": 0.1753,
+      "step": 971
+    },
+    {
+      "epoch": 0.5059864653826133,
+      "grad_norm": 0.2832624623321996,
+      "learning_rate": 4.752160843377469e-05,
+      "loss": 0.1674,
+      "step": 972
+    },
+    {
+      "epoch": 0.5065070275897969,
+      "grad_norm": 0.2837250649227678,
+      "learning_rate": 4.751550573755658e-05,
+      "loss": 0.1648,
+      "step": 973
+    },
+    {
+      "epoch": 0.5070275897969807,
+      "grad_norm": 0.30259169690721716,
+      "learning_rate": 4.750939592989396e-05,
+      "loss": 0.1715,
+      "step": 974
+    },
+    {
+      "epoch": 0.5075481520041645,
+      "grad_norm": 0.29554573173203236,
+      "learning_rate": 4.750327901271662e-05,
+      "loss": 0.1767,
+      "step": 975
+    },
+    {
+      "epoch": 0.5080687142113482,
+      "grad_norm": 0.3048549802014968,
+      "learning_rate": 4.7497154987956554e-05,
+      "loss": 0.1736,
+      "step": 976
+    },
+    {
+      "epoch": 0.508589276418532,
+      "grad_norm": 0.28859818824566424,
+      "learning_rate": 4.749102385754802e-05,
+      "loss": 0.1762,
+      "step": 977
+    },
+    {
+      "epoch": 0.5091098386257158,
+      "grad_norm": 0.2907661277669279,
+      "learning_rate": 4.74848856234275e-05,
+      "loss": 0.169,
+      "step": 978
+    },
+    {
+      "epoch": 0.5096304008328996,
+      "grad_norm": 0.3020314402188709,
+      "learning_rate": 4.747874028753375e-05,
+      "loss": 0.1705,
+      "step": 979
+    },
+    {
+      "epoch": 0.5101509630400833,
+      "grad_norm": 0.28916453332411907,
+      "learning_rate": 4.747258785180774e-05,
+      "loss": 0.169,
+      "step": 980
+    },
+    {
+      "epoch": 0.510671525247267,
+      "grad_norm": 0.27306730871455454,
+      "learning_rate": 4.746642831819271e-05,
+      "loss": 0.1712,
+      "step": 981
+    },
+    {
+      "epoch": 0.5111920874544508,
+      "grad_norm": 0.3577975221682602,
+      "learning_rate": 4.746026168863412e-05,
+      "loss": 0.1714,
+      "step": 982
+    },
+    {
+      "epoch": 0.5117126496616345,
+      "grad_norm": 0.28463660820354614,
+      "learning_rate": 4.7454087965079675e-05,
+      "loss": 0.1622,
+      "step": 983
+    },
+    {
+      "epoch": 0.5122332118688183,
+      "grad_norm": 0.31003657876962476,
+      "learning_rate": 4.744790714947932e-05,
+      "loss": 0.1695,
+      "step": 984
+    },
+    {
+      "epoch": 0.5127537740760021,
+      "grad_norm": 0.27961769726505104,
+      "learning_rate": 4.744171924378526e-05,
+      "loss": 0.163,
+      "step": 985
+    },
+    {
+      "epoch": 0.5132743362831859,
+      "grad_norm": 0.33088943157458217,
+      "learning_rate": 4.743552424995191e-05,
+      "loss": 0.17,
+      "step": 986
+    },
+    {
+      "epoch": 0.5137948984903696,
+      "grad_norm": 0.3056859101086278,
+      "learning_rate": 4.7429322169935955e-05,
+      "loss": 0.1697,
+      "step": 987
+    },
+    {
+      "epoch": 0.5143154606975534,
+      "grad_norm": 0.31119710051950294,
+      "learning_rate": 4.7423113005696275e-05,
+      "loss": 0.1671,
+      "step": 988
+    },
+    {
+      "epoch": 0.5148360229047371,
+      "grad_norm": 0.3143711776010868,
+      "learning_rate": 4.741689675919403e-05,
+      "loss": 0.1698,
+      "step": 989
+    },
+    {
+      "epoch": 0.5153565851119208,
+      "grad_norm": 0.3061616967117495,
+      "learning_rate": 4.7410673432392596e-05,
+      "loss": 0.1842,
+      "step": 990
+    },
+    {
+      "epoch": 0.5158771473191046,
+      "grad_norm": 0.29278315778967423,
+      "learning_rate": 4.740444302725759e-05,
+      "loss": 0.1678,
+      "step": 991
+    },
+    {
+      "epoch": 0.5163977095262884,
+      "grad_norm": 0.2957634913771008,
+      "learning_rate": 4.7398205545756863e-05,
+      "loss": 0.1751,
+      "step": 992
+    },
+    {
+      "epoch": 0.5169182717334722,
+      "grad_norm": 0.279576531903814,
+      "learning_rate": 4.7391960989860504e-05,
+      "loss": 0.1633,
+      "step": 993
+    },
+    {
+      "epoch": 0.5174388339406559,
+      "grad_norm": 0.3073973561502015,
+      "learning_rate": 4.738570936154083e-05,
+      "loss": 0.1645,
+      "step": 994
+    },
+    {
+      "epoch": 0.5179593961478397,
+      "grad_norm": 0.29041369434808123,
+      "learning_rate": 4.7379450662772394e-05,
+      "loss": 0.1626,
+      "step": 995
+    },
+    {
+      "epoch": 0.5184799583550235,
+      "grad_norm": 0.30177127364477313,
+      "learning_rate": 4.737318489553199e-05,
+      "loss": 0.175,
+      "step": 996
+    },
+    {
+      "epoch": 0.5190005205622071,
+      "grad_norm": 0.29152000130899824,
+      "learning_rate": 4.736691206179864e-05,
+      "loss": 0.1713,
+      "step": 997
+    },
+    {
+      "epoch": 0.5195210827693909,
+      "grad_norm": 0.3104231023579747,
+      "learning_rate": 4.7360632163553595e-05,
+      "loss": 0.1703,
+      "step": 998
+    },
+    {
+      "epoch": 0.5200416449765747,
+      "grad_norm": 0.28598295804809104,
+      "learning_rate": 4.735434520278034e-05,
+      "loss": 0.1685,
+      "step": 999
+    },
+    {
+      "epoch": 0.5205622071837585,
+      "grad_norm": 0.31603631518913927,
+      "learning_rate": 4.734805118146459e-05,
+      "loss": 0.1786,
+      "step": 1000
+    },
+    {
+      "epoch": 0.5210827693909422,
+      "grad_norm": 0.3129741436841726,
+      "learning_rate": 4.734175010159428e-05,
+      "loss": 0.1732,
+      "step": 1001
+    },
+    {
+      "epoch": 0.521603331598126,
+      "grad_norm": 0.3176913349121328,
+      "learning_rate": 4.73354419651596e-05,
+      "loss": 0.1678,
+      "step": 1002
+    },
+    {
+      "epoch": 0.5221238938053098,
+      "grad_norm": 0.33041896990240904,
+      "learning_rate": 4.732912677415294e-05,
+      "loss": 0.1645,
+      "step": 1003
+    },
+    {
+      "epoch": 0.5226444560124935,
+      "grad_norm": 0.2830876561876967,
+      "learning_rate": 4.732280453056894e-05,
+      "loss": 0.1679,
+      "step": 1004
+    },
+    {
+      "epoch": 0.5231650182196772,
+      "grad_norm": 0.34261613836366955,
+      "learning_rate": 4.7316475236404454e-05,
+      "loss": 0.1737,
+      "step": 1005
+    },
+    {
+      "epoch": 0.523685580426861,
+      "grad_norm": 0.2760763443773014,
+      "learning_rate": 4.731013889365857e-05,
+      "loss": 0.1709,
+      "step": 1006
+    },
+    {
+      "epoch": 0.5242061426340447,
+      "grad_norm": 0.28543357092624544,
+      "learning_rate": 4.7303795504332604e-05,
+      "loss": 0.1694,
+      "step": 1007
+    },
+    {
+      "epoch": 0.5247267048412285,
+      "grad_norm": 0.2968860076469517,
+      "learning_rate": 4.729744507043008e-05,
+      "loss": 0.1686,
+      "step": 1008
+    },
+    {
+      "epoch": 0.5252472670484123,
+      "grad_norm": 0.2955425625928012,
+      "learning_rate": 4.729108759395677e-05,
+      "loss": 0.1668,
+      "step": 1009
+    },
+    {
+      "epoch": 0.5257678292555961,
+      "grad_norm": 0.3064862906090259,
+      "learning_rate": 4.728472307692067e-05,
+      "loss": 0.1717,
+      "step": 1010
+    },
+    {
+      "epoch": 0.5262883914627798,
+      "grad_norm": 0.2944863868003504,
+      "learning_rate": 4.727835152133197e-05,
+      "loss": 0.1658,
+      "step": 1011
+    },
+    {
+      "epoch": 0.5268089536699636,
+      "grad_norm": 0.29581781869973445,
+      "learning_rate": 4.727197292920312e-05,
+      "loss": 0.1664,
+      "step": 1012
+    },
+    {
+      "epoch": 0.5273295158771473,
+      "grad_norm": 0.2655727030333062,
+      "learning_rate": 4.7265587302548766e-05,
+      "loss": 0.1588,
+      "step": 1013
+    },
+    {
+      "epoch": 0.527850078084331,
+      "grad_norm": 0.32401890632087926,
+      "learning_rate": 4.7259194643385796e-05,
+      "loss": 0.1689,
+      "step": 1014
+    },
+    {
+      "epoch": 0.5283706402915148,
+      "grad_norm": 0.31117831297374177,
+      "learning_rate": 4.7252794953733294e-05,
+      "loss": 0.1776,
+      "step": 1015
+    },
+    {
+      "epoch": 0.5288912024986986,
+      "grad_norm": 0.3060852587762897,
+      "learning_rate": 4.72463882356126e-05,
+      "loss": 0.1671,
+      "step": 1016
+    },
+    {
+      "epoch": 0.5294117647058824,
+      "grad_norm": 0.32556487990025834,
+      "learning_rate": 4.7239974491047236e-05,
+      "loss": 0.1775,
+      "step": 1017
+    },
+    {
+      "epoch": 0.5299323269130661,
+      "grad_norm": 0.2799847281876232,
+      "learning_rate": 4.723355372206297e-05,
+      "loss": 0.1675,
+      "step": 1018
+    },
+    {
+      "epoch": 0.5304528891202499,
+      "grad_norm": 0.2821768720138456,
+      "learning_rate": 4.722712593068779e-05,
+      "loss": 0.1664,
+      "step": 1019
+    },
+    {
+      "epoch": 0.5309734513274337,
+      "grad_norm": 0.3123419119344323,
+      "learning_rate": 4.722069111895187e-05,
+      "loss": 0.1733,
+      "step": 1020
+    },
+    {
+      "epoch": 0.5314940135346173,
+      "grad_norm": 0.30697486370866406,
+      "learning_rate": 4.721424928888763e-05,
+      "loss": 0.1722,
+      "step": 1021
+    },
+    {
+      "epoch": 0.5320145757418011,
+      "grad_norm": 0.2895747411066602,
+      "learning_rate": 4.7207800442529706e-05,
+      "loss": 0.1647,
+      "step": 1022
+    },
+    {
+      "epoch": 0.5325351379489849,
+      "grad_norm": 0.28336131313295926,
+      "learning_rate": 4.720134458191494e-05,
+      "loss": 0.1657,
+      "step": 1023
+    },
+    {
+      "epoch": 0.5330557001561687,
+      "grad_norm": 0.2902463741761351,
+      "learning_rate": 4.719488170908239e-05,
+      "loss": 0.1579,
+      "step": 1024
+    },
+    {
+      "epoch": 0.5335762623633524,
+      "grad_norm": 0.295514891496485,
+      "learning_rate": 4.718841182607334e-05,
+      "loss": 0.1716,
+      "step": 1025
+    },
+    {
+      "epoch": 0.5340968245705362,
+      "grad_norm": 0.32915905626702163,
+      "learning_rate": 4.718193493493127e-05,
+      "loss": 0.1671,
+      "step": 1026
+    },
+    {
+      "epoch": 0.53461738677772,
+      "grad_norm": 0.3026674578937171,
+      "learning_rate": 4.717545103770189e-05,
+      "loss": 0.158,
+      "step": 1027
+    },
+    {
+      "epoch": 0.5351379489849037,
+      "grad_norm": 0.28381405365121987,
+      "learning_rate": 4.716896013643313e-05,
+      "loss": 0.1677,
+      "step": 1028
+    },
+    {
+      "epoch": 0.5356585111920874,
+      "grad_norm": 0.29824689910486607,
+      "learning_rate": 4.716246223317509e-05,
+      "loss": 0.1703,
+      "step": 1029
+    },
+    {
+      "epoch": 0.5361790733992712,
+      "grad_norm": 0.33329704323409864,
+      "learning_rate": 4.7155957329980126e-05,
+      "loss": 0.1698,
+      "step": 1030
+    },
+    {
+      "epoch": 0.536699635606455,
+      "grad_norm": 0.3004672079802441,
+      "learning_rate": 4.7149445428902786e-05,
+      "loss": 0.1589,
+      "step": 1031
+    },
+    {
+      "epoch": 0.5372201978136387,
+      "grad_norm": 0.31860987261597545,
+      "learning_rate": 4.714292653199984e-05,
+      "loss": 0.1672,
+      "step": 1032
+    },
+    {
+      "epoch": 0.5377407600208225,
+      "grad_norm": 0.3442296126297135,
+      "learning_rate": 4.713640064133025e-05,
+      "loss": 0.1632,
+      "step": 1033
+    },
+    {
+      "epoch": 0.5382613222280063,
+      "grad_norm": 0.3064387531445744,
+      "learning_rate": 4.7129867758955196e-05,
+      "loss": 0.1635,
+      "step": 1034
+    },
+    {
+      "epoch": 0.53878188443519,
+      "grad_norm": 0.3123991398030029,
+      "learning_rate": 4.7123327886938076e-05,
+      "loss": 0.1664,
+      "step": 1035
+    },
+    {
+      "epoch": 0.5393024466423738,
+      "grad_norm": 0.2900379032500118,
+      "learning_rate": 4.711678102734447e-05,
+      "loss": 0.1699,
+      "step": 1036
+    },
+    {
+      "epoch": 0.5398230088495575,
+      "grad_norm": 0.3101402740827743,
+      "learning_rate": 4.711022718224218e-05,
+      "loss": 0.1674,
+      "step": 1037
+    },
+    {
+      "epoch": 0.5403435710567412,
+      "grad_norm": 0.2978308638116788,
+      "learning_rate": 4.710366635370124e-05,
+      "loss": 0.1629,
+      "step": 1038
+    },
+    {
+      "epoch": 0.540864133263925,
+      "grad_norm": 0.2934026273877542,
+      "learning_rate": 4.709709854379385e-05,
+      "loss": 0.1557,
+      "step": 1039
+    },
+    {
+      "epoch": 0.5413846954711088,
+      "grad_norm": 0.29754336522736463,
+      "learning_rate": 4.709052375459442e-05,
+      "loss": 0.1708,
+      "step": 1040
+    },
+    {
+      "epoch": 0.5419052576782926,
+      "grad_norm": 0.2941537759903196,
+      "learning_rate": 4.7083941988179594e-05,
+      "loss": 0.1688,
+      "step": 1041
+    },
+    {
+      "epoch": 0.5424258198854763,
+      "grad_norm": 0.3197429755584853,
+      "learning_rate": 4.707735324662818e-05,
+      "loss": 0.1711,
+      "step": 1042
+    },
+    {
+      "epoch": 0.5429463820926601,
+      "grad_norm": 0.29878401691495904,
+      "learning_rate": 4.7070757532021224e-05,
+      "loss": 0.1667,
+      "step": 1043
+    },
+    {
+      "epoch": 0.5434669442998439,
+      "grad_norm": 0.34624169271494265,
+      "learning_rate": 4.706415484644195e-05,
+      "loss": 0.1719,
+      "step": 1044
+    },
+    {
+      "epoch": 0.5439875065070275,
+      "grad_norm": 0.2825166282554898,
+      "learning_rate": 4.705754519197581e-05,
+      "loss": 0.163,
+      "step": 1045
+    },
+    {
+      "epoch": 0.5445080687142113,
+      "grad_norm": 0.3328465767542314,
+      "learning_rate": 4.705092857071042e-05,
+      "loss": 0.1655,
+      "step": 1046
+    },
+    {
+      "epoch": 0.5450286309213951,
+      "grad_norm": 0.3231665017996849,
+      "learning_rate": 4.704430498473562e-05,
+      "loss": 0.1666,
+      "step": 1047
+    },
+    {
+      "epoch": 0.5455491931285789,
+      "grad_norm": 0.33401260004138217,
+      "learning_rate": 4.7037674436143466e-05,
+      "loss": 0.1704,
+      "step": 1048
+    },
+    {
+      "epoch": 0.5460697553357626,
+      "grad_norm": 0.30948523577299875,
+      "learning_rate": 4.703103692702817e-05,
+      "loss": 0.1682,
+      "step": 1049
+    },
+    {
+      "epoch": 0.5465903175429464,
+      "grad_norm": 0.33513343731742634,
+      "learning_rate": 4.7024392459486176e-05,
+      "loss": 0.1719,
+      "step": 1050
+    },
+    {
+      "epoch": 0.5471108797501302,
+      "grad_norm": 0.29920676443214966,
+      "learning_rate": 4.7017741035616124e-05,
+      "loss": 0.1624,
+      "step": 1051
+    },
+    {
+      "epoch": 0.547631441957314,
+      "grad_norm": 0.3063741016132584,
+      "learning_rate": 4.701108265751884e-05,
+      "loss": 0.1651,
+      "step": 1052
+    },
+    {
+      "epoch": 0.5481520041644976,
+      "grad_norm": 0.2934234878431351,
+      "learning_rate": 4.7004417327297325e-05,
+      "loss": 0.1648,
+      "step": 1053
+    },
+    {
+      "epoch": 0.5486725663716814,
+      "grad_norm": 0.30182853291753425,
+      "learning_rate": 4.6997745047056836e-05,
+      "loss": 0.1626,
+      "step": 1054
+    },
+    {
+      "epoch": 0.5491931285788652,
+      "grad_norm": 0.3175333083074021,
+      "learning_rate": 4.699106581890477e-05,
+      "loss": 0.1719,
+      "step": 1055
+    },
+    {
+      "epoch": 0.5497136907860489,
+      "grad_norm": 0.31314464288948607,
+      "learning_rate": 4.698437964495074e-05,
+      "loss": 0.1598,
+      "step": 1056
+    },
+    {
+      "epoch": 0.5502342529932327,
+      "grad_norm": 0.32599179752881036,
+      "learning_rate": 4.6977686527306556e-05,
+      "loss": 0.1631,
+      "step": 1057
+    },
+    {
+      "epoch": 0.5507548152004165,
+      "grad_norm": 0.3308531331011867,
+      "learning_rate": 4.697098646808621e-05,
+      "loss": 0.17,
+      "step": 1058
+    },
+    {
+      "epoch": 0.5512753774076002,
+      "grad_norm": 0.2982979929430078,
+      "learning_rate": 4.69642794694059e-05,
+      "loss": 0.1708,
+      "step": 1059
+    },
+    {
+      "epoch": 0.551795939614784,
+      "grad_norm": 0.32955914457005925,
+      "learning_rate": 4.695756553338401e-05,
+      "loss": 0.1671,
+      "step": 1060
+    },
+    {
+      "epoch": 0.5523165018219677,
+      "grad_norm": 0.3070588254800077,
+      "learning_rate": 4.6950844662141096e-05,
+      "loss": 0.171,
+      "step": 1061
+    },
+    {
+      "epoch": 0.5528370640291514,
+      "grad_norm": 0.31217947398655915,
+      "learning_rate": 4.6944116857799936e-05,
+      "loss": 0.1697,
+      "step": 1062
+    },
+    {
+      "epoch": 0.5533576262363352,
+      "grad_norm": 0.2934020560226863,
+      "learning_rate": 4.6937382122485484e-05,
+      "loss": 0.1744,
+      "step": 1063
+    },
+    {
+      "epoch": 0.553878188443519,
+      "grad_norm": 0.2854842827185457,
+      "learning_rate": 4.693064045832488e-05,
+      "loss": 0.161,
+      "step": 1064
+    },
+    {
+      "epoch": 0.5543987506507028,
+      "grad_norm": 0.2815405598740039,
+      "learning_rate": 4.692389186744745e-05,
+      "loss": 0.1621,
+      "step": 1065
+    },
+    {
+      "epoch": 0.5549193128578865,
+      "grad_norm": 0.28832423549074593,
+      "learning_rate": 4.691713635198473e-05,
+      "loss": 0.1621,
+      "step": 1066
+    },
+    {
+      "epoch": 0.5554398750650703,
+      "grad_norm": 0.29233220575342456,
+      "learning_rate": 4.6910373914070404e-05,
+      "loss": 0.1664,
+      "step": 1067
+    },
+    {
+      "epoch": 0.5559604372722541,
+      "grad_norm": 0.2904867981234466,
+      "learning_rate": 4.6903604555840374e-05,
+      "loss": 0.1709,
+      "step": 1068
+    },
+    {
+      "epoch": 0.5564809994794377,
+      "grad_norm": 0.33583477749715884,
+      "learning_rate": 4.6896828279432725e-05,
+      "loss": 0.1592,
+      "step": 1069
+    },
+    {
+      "epoch": 0.5570015616866215,
+      "grad_norm": 0.2809477801971266,
+      "learning_rate": 4.689004508698771e-05,
+      "loss": 0.1676,
+      "step": 1070
+    },
+    {
+      "epoch": 0.5575221238938053,
+      "grad_norm": 0.32697745725327254,
+      "learning_rate": 4.6883254980647787e-05,
+      "loss": 0.1671,
+      "step": 1071
+    },
+    {
+      "epoch": 0.5580426861009891,
+      "grad_norm": 0.28644207450920894,
+      "learning_rate": 4.6876457962557575e-05,
+      "loss": 0.1704,
+      "step": 1072
+    },
+    {
+      "epoch": 0.5585632483081728,
+      "grad_norm": 0.3293642883398715,
+      "learning_rate": 4.68696540348639e-05,
+      "loss": 0.1722,
+      "step": 1073
+    },
+    {
+      "epoch": 0.5590838105153566,
+      "grad_norm": 0.2844767472247252,
+      "learning_rate": 4.686284319971576e-05,
+      "loss": 0.166,
+      "step": 1074
+    },
+    {
+      "epoch": 0.5596043727225404,
+      "grad_norm": 0.30371406025926573,
+      "learning_rate": 4.685602545926432e-05,
+      "loss": 0.1718,
+      "step": 1075
+    },
+    {
+      "epoch": 0.5601249349297241,
+      "grad_norm": 0.303036966019252,
+      "learning_rate": 4.684920081566295e-05,
+      "loss": 0.1696,
+      "step": 1076
+    },
+    {
+      "epoch": 0.5606454971369078,
+      "grad_norm": 0.2771701255154809,
+      "learning_rate": 4.6842369271067185e-05,
+      "loss": 0.1606,
+      "step": 1077
+    },
+    {
+      "epoch": 0.5611660593440916,
+      "grad_norm": 0.3233636024191239,
+      "learning_rate": 4.683553082763475e-05,
+      "loss": 0.1699,
+      "step": 1078
+    },
+    {
+      "epoch": 0.5616866215512754,
+      "grad_norm": 0.28548453667968193,
+      "learning_rate": 4.6828685487525554e-05,
+      "loss": 0.1621,
+      "step": 1079
+    },
+    {
+      "epoch": 0.5622071837584591,
+      "grad_norm": 0.33431595584526136,
+      "learning_rate": 4.6821833252901646e-05,
+      "loss": 0.1579,
+      "step": 1080
+    },
+    {
+      "epoch": 0.5627277459656429,
+      "grad_norm": 0.29278248350889036,
+      "learning_rate": 4.6814974125927304e-05,
+      "loss": 0.1622,
+      "step": 1081
+    },
+    {
+      "epoch": 0.5632483081728267,
+      "grad_norm": 0.31082966905234066,
+      "learning_rate": 4.680810810876895e-05,
+      "loss": 0.1745,
+      "step": 1082
+    },
+    {
+      "epoch": 0.5637688703800104,
+      "grad_norm": 0.304493159369903,
+      "learning_rate": 4.6801235203595195e-05,
+      "loss": 0.1657,
+      "step": 1083
+    },
+    {
+      "epoch": 0.5642894325871942,
+      "grad_norm": 0.2949622274273174,
+      "learning_rate": 4.679435541257682e-05,
+      "loss": 0.1608,
+      "step": 1084
+    },
+    {
+      "epoch": 0.5648099947943779,
+      "grad_norm": 0.27199415580975606,
+      "learning_rate": 4.678746873788677e-05,
+      "loss": 0.1626,
+      "step": 1085
+    },
+    {
+      "epoch": 0.5653305570015617,
+      "grad_norm": 0.30641660626180384,
+      "learning_rate": 4.678057518170021e-05,
+      "loss": 0.1705,
+      "step": 1086
+    },
+    {
+      "epoch": 0.5658511192087454,
+      "grad_norm": 0.3045178599347487,
+      "learning_rate": 4.677367474619442e-05,
+      "loss": 0.164,
+      "step": 1087
+    },
+    {
+      "epoch": 0.5663716814159292,
+      "grad_norm": 0.2942177088983781,
+      "learning_rate": 4.6766767433548885e-05,
+      "loss": 0.1702,
+      "step": 1088
+    },
+    {
+      "epoch": 0.566892243623113,
+      "grad_norm": 0.3112050068584472,
+      "learning_rate": 4.6759853245945256e-05,
+      "loss": 0.1727,
+      "step": 1089
+    },
+    {
+      "epoch": 0.5674128058302967,
+      "grad_norm": 0.2968014199642348,
+      "learning_rate": 4.675293218556735e-05,
+      "loss": 0.1605,
+      "step": 1090
+    },
+    {
+      "epoch": 0.5679333680374805,
+      "grad_norm": 0.30379202567617314,
+      "learning_rate": 4.6746004254601184e-05,
+      "loss": 0.1665,
+      "step": 1091
+    },
+    {
+      "epoch": 0.5684539302446643,
+      "grad_norm": 0.4356850886928418,
+      "learning_rate": 4.6739069455234886e-05,
+      "loss": 0.1737,
+      "step": 1092
+    },
+    {
+      "epoch": 0.568974492451848,
+      "grad_norm": 0.2873286927122915,
+      "learning_rate": 4.673212778965881e-05,
+      "loss": 0.1671,
+      "step": 1093
+    },
+    {
+      "epoch": 0.5694950546590317,
+      "grad_norm": 0.307373163176704,
+      "learning_rate": 4.672517926006545e-05,
+      "loss": 0.1696,
+      "step": 1094
+    },
+    {
+      "epoch": 0.5700156168662155,
+      "grad_norm": 0.284743433783671,
+      "learning_rate": 4.671822386864948e-05,
+      "loss": 0.1603,
+      "step": 1095
+    },
+    {
+      "epoch": 0.5705361790733993,
+      "grad_norm": 0.3019989838556474,
+      "learning_rate": 4.6711261617607725e-05,
+      "loss": 0.1735,
+      "step": 1096
+    },
+    {
+      "epoch": 0.571056741280583,
+      "grad_norm": 0.3033519694233394,
+      "learning_rate": 4.670429250913921e-05,
+      "loss": 0.1639,
+      "step": 1097
+    },
+    {
+      "epoch": 0.5715773034877668,
+      "grad_norm": 0.31363574509134545,
+      "learning_rate": 4.669731654544508e-05,
+      "loss": 0.161,
+      "step": 1098
+    },
+    {
+      "epoch": 0.5720978656949506,
+      "grad_norm": 0.28285834283513167,
+      "learning_rate": 4.669033372872868e-05,
+      "loss": 0.1665,
+      "step": 1099
+    },
+    {
+      "epoch": 0.5726184279021344,
+      "grad_norm": 0.2628857284685492,
+      "learning_rate": 4.668334406119551e-05,
+      "loss": 0.1566,
+      "step": 1100
+    },
+    {
+      "epoch": 0.573138990109318,
+      "grad_norm": 0.2655054843532405,
+      "learning_rate": 4.667634754505323e-05,
+      "loss": 0.1604,
+      "step": 1101
+    },
+    {
+      "epoch": 0.5736595523165018,
+      "grad_norm": 0.26626184972682054,
+      "learning_rate": 4.666934418251166e-05,
+      "loss": 0.1554,
+      "step": 1102
+    },
+    {
+      "epoch": 0.5741801145236856,
+      "grad_norm": 0.28305815343523066,
+      "learning_rate": 4.6662333975782795e-05,
+      "loss": 0.1698,
+      "step": 1103
+    },
+    {
+      "epoch": 0.5747006767308693,
+      "grad_norm": 0.2707887337606054,
+      "learning_rate": 4.6655316927080784e-05,
+      "loss": 0.165,
+      "step": 1104
+    },
+    {
+      "epoch": 0.5752212389380531,
+      "grad_norm": 0.2917568726342999,
+      "learning_rate": 4.664829303862194e-05,
+      "loss": 0.1686,
+      "step": 1105
+    },
+    {
+      "epoch": 0.5757418011452369,
+      "grad_norm": 0.26069329637196703,
+      "learning_rate": 4.664126231262472e-05,
+      "loss": 0.1625,
+      "step": 1106
+    },
+    {
+      "epoch": 0.5762623633524206,
+      "grad_norm": 0.27239030458009167,
+      "learning_rate": 4.663422475130977e-05,
+      "loss": 0.1537,
+      "step": 1107
+    },
+    {
+      "epoch": 0.5767829255596044,
+      "grad_norm": 0.2575395498508633,
+      "learning_rate": 4.662718035689987e-05,
+      "loss": 0.1617,
+      "step": 1108
+    },
+    {
+      "epoch": 0.5773034877667881,
+      "grad_norm": 0.27545374757966323,
+      "learning_rate": 4.662012913161997e-05,
+      "loss": 0.1616,
+      "step": 1109
+    },
+    {
+      "epoch": 0.5778240499739719,
+      "grad_norm": 0.28448600250229067,
+      "learning_rate": 4.661307107769718e-05,
+      "loss": 0.1646,
+      "step": 1110
+    },
+    {
+      "epoch": 0.5783446121811556,
+      "grad_norm": 0.27298548291298,
+      "learning_rate": 4.660600619736076e-05,
+      "loss": 0.1628,
+      "step": 1111
+    },
+    {
+      "epoch": 0.5788651743883394,
+      "grad_norm": 0.3145003842578059,
+      "learning_rate": 4.6598934492842114e-05,
+      "loss": 0.1737,
+      "step": 1112
+    },
+    {
+      "epoch": 0.5793857365955232,
+      "grad_norm": 0.27194978692066746,
+      "learning_rate": 4.659185596637484e-05,
+      "loss": 0.1621,
+      "step": 1113
+    },
+    {
+      "epoch": 0.5799062988027069,
+      "grad_norm": 0.28956670097111964,
+      "learning_rate": 4.658477062019465e-05,
+      "loss": 0.1722,
+      "step": 1114
+    },
+    {
+      "epoch": 0.5804268610098907,
+      "grad_norm": 0.2826807429961728,
+      "learning_rate": 4.657767845653943e-05,
+      "loss": 0.1688,
+      "step": 1115
+    },
+    {
+      "epoch": 0.5809474232170745,
+      "grad_norm": 0.28708843741623624,
+      "learning_rate": 4.657057947764922e-05,
+      "loss": 0.1649,
+      "step": 1116
+    },
+    {
+      "epoch": 0.5814679854242581,
+      "grad_norm": 0.29098772804179435,
+      "learning_rate": 4.656347368576619e-05,
+      "loss": 0.169,
+      "step": 1117
+    },
+    {
+      "epoch": 0.5819885476314419,
+      "grad_norm": 0.3050742875550255,
+      "learning_rate": 4.6556361083134705e-05,
+      "loss": 0.1736,
+      "step": 1118
+    },
+    {
+      "epoch": 0.5825091098386257,
+      "grad_norm": 0.2883430860003268,
+      "learning_rate": 4.654924167200123e-05,
+      "loss": 0.1636,
+      "step": 1119
+    },
+    {
+      "epoch": 0.5830296720458095,
+      "grad_norm": 0.2694576550703567,
+      "learning_rate": 4.654211545461443e-05,
+      "loss": 0.1596,
+      "step": 1120
+    },
+    {
+      "epoch": 0.5835502342529932,
+      "grad_norm": 0.29077979216048633,
+      "learning_rate": 4.653498243322508e-05,
+      "loss": 0.1614,
+      "step": 1121
+    },
+    {
+      "epoch": 0.584070796460177,
+      "grad_norm": 0.30454032706720835,
+      "learning_rate": 4.652784261008613e-05,
+      "loss": 0.1657,
+      "step": 1122
+    },
+    {
+      "epoch": 0.5845913586673608,
+      "grad_norm": 0.2809170140826117,
+      "learning_rate": 4.652069598745267e-05,
+      "loss": 0.1674,
+      "step": 1123
+    },
+    {
+      "epoch": 0.5851119208745446,
+      "grad_norm": 0.2998875317771076,
+      "learning_rate": 4.6513542567581914e-05,
+      "loss": 0.1745,
+      "step": 1124
+    },
+    {
+      "epoch": 0.5856324830817282,
+      "grad_norm": 0.2907341448806623,
+      "learning_rate": 4.650638235273327e-05,
+      "loss": 0.1722,
+      "step": 1125
+    },
+    {
+      "epoch": 0.586153045288912,
+      "grad_norm": 0.3240717556080497,
+      "learning_rate": 4.6499215345168255e-05,
+      "loss": 0.1715,
+      "step": 1126
+    },
+    {
+      "epoch": 0.5866736074960958,
+      "grad_norm": 0.2856383542388814,
+      "learning_rate": 4.6492041547150555e-05,
+      "loss": 0.1704,
+      "step": 1127
+    },
+    {
+      "epoch": 0.5871941697032795,
+      "grad_norm": 0.29395504573057224,
+      "learning_rate": 4.648486096094597e-05,
+      "loss": 0.1644,
+      "step": 1128
+    },
+    {
+      "epoch": 0.5877147319104633,
+      "grad_norm": 0.29455509913506334,
+      "learning_rate": 4.647767358882249e-05,
+      "loss": 0.1637,
+      "step": 1129
+    },
+    {
+      "epoch": 0.5882352941176471,
+      "grad_norm": 0.27422990925352025,
+      "learning_rate": 4.647047943305019e-05,
+      "loss": 0.1581,
+      "step": 1130
+    },
+    {
+      "epoch": 0.5887558563248309,
+      "grad_norm": 0.28624654561233115,
+      "learning_rate": 4.646327849590134e-05,
+      "loss": 0.1659,
+      "step": 1131
+    },
+    {
+      "epoch": 0.5892764185320146,
+      "grad_norm": 0.2589094816834805,
+      "learning_rate": 4.6456070779650326e-05,
+      "loss": 0.1545,
+      "step": 1132
+    },
+    {
+      "epoch": 0.5897969807391983,
+      "grad_norm": 0.28489485708163403,
+      "learning_rate": 4.6448856286573684e-05,
+      "loss": 0.1669,
+      "step": 1133
+    },
+    {
+      "epoch": 0.5903175429463821,
+      "grad_norm": 0.27991494765414887,
+      "learning_rate": 4.644163501895008e-05,
+      "loss": 0.1692,
+      "step": 1134
+    },
+    {
+      "epoch": 0.5908381051535658,
+      "grad_norm": 0.271863121129334,
+      "learning_rate": 4.643440697906033e-05,
+      "loss": 0.1703,
+      "step": 1135
+    },
+    {
+      "epoch": 0.5913586673607496,
+      "grad_norm": 0.261056414536902,
+      "learning_rate": 4.642717216918738e-05,
+      "loss": 0.1621,
+      "step": 1136
+    },
+    {
+      "epoch": 0.5918792295679334,
+      "grad_norm": 0.2929328942515898,
+      "learning_rate": 4.6419930591616336e-05,
+      "loss": 0.1629,
+      "step": 1137
+    },
+    {
+      "epoch": 0.5923997917751171,
+      "grad_norm": 0.2606767168924215,
+      "learning_rate": 4.641268224863441e-05,
+      "loss": 0.1654,
+      "step": 1138
+    },
+    {
+      "epoch": 0.5929203539823009,
+      "grad_norm": 0.2908040086603664,
+      "learning_rate": 4.6405427142530954e-05,
+      "loss": 0.1588,
+      "step": 1139
+    },
+    {
+      "epoch": 0.5934409161894847,
+      "grad_norm": 0.2611032398127685,
+      "learning_rate": 4.6398165275597494e-05,
+      "loss": 0.1616,
+      "step": 1140
+    },
+    {
+      "epoch": 0.5939614783966684,
+      "grad_norm": 0.27150334105117246,
+      "learning_rate": 4.6390896650127656e-05,
+      "loss": 0.1693,
+      "step": 1141
+    },
+    {
+      "epoch": 0.5944820406038521,
+      "grad_norm": 0.27540568978983837,
+      "learning_rate": 4.638362126841721e-05,
+      "loss": 0.167,
+      "step": 1142
+    },
+    {
+      "epoch": 0.5950026028110359,
+      "grad_norm": 0.2676573123405451,
+      "learning_rate": 4.637633913276406e-05,
+      "loss": 0.1643,
+      "step": 1143
+    },
+    {
+      "epoch": 0.5955231650182197,
+      "grad_norm": 0.2607951523773393,
+      "learning_rate": 4.6369050245468243e-05,
+      "loss": 0.1672,
+      "step": 1144
+    },
+    {
+      "epoch": 0.5960437272254034,
+      "grad_norm": 0.28522127225344346,
+      "learning_rate": 4.636175460883193e-05,
+      "loss": 0.1689,
+      "step": 1145
+    },
+    {
+      "epoch": 0.5965642894325872,
+      "grad_norm": 0.28313291478789127,
+      "learning_rate": 4.6354452225159416e-05,
+      "loss": 0.1678,
+      "step": 1146
+    },
+    {
+      "epoch": 0.597084851639771,
+      "grad_norm": 0.2781466635540155,
+      "learning_rate": 4.634714309675714e-05,
+      "loss": 0.1693,
+      "step": 1147
+    },
+    {
+      "epoch": 0.5976054138469548,
+      "grad_norm": 0.2993701779876083,
+      "learning_rate": 4.6339827225933665e-05,
+      "loss": 0.178,
+      "step": 1148
+    },
+    {
+      "epoch": 0.5981259760541384,
+      "grad_norm": 0.2757353899839654,
+      "learning_rate": 4.6332504614999684e-05,
+      "loss": 0.1686,
+      "step": 1149
+    },
+    {
+      "epoch": 0.5986465382613222,
+      "grad_norm": 0.2893083043239989,
+      "learning_rate": 4.6325175266268005e-05,
+      "loss": 0.1613,
+      "step": 1150
+    },
+    {
+      "epoch": 0.599167100468506,
+      "grad_norm": 0.28169522960512217,
+      "learning_rate": 4.6317839182053603e-05,
+      "loss": 0.1612,
+      "step": 1151
+    },
+    {
+      "epoch": 0.5996876626756897,
+      "grad_norm": 0.32089924116837154,
+      "learning_rate": 4.6310496364673534e-05,
+      "loss": 0.1688,
+      "step": 1152
+    },
+    {
+      "epoch": 0.6002082248828735,
+      "grad_norm": 0.259177259755011,
+      "learning_rate": 4.630314681644701e-05,
+      "loss": 0.159,
+      "step": 1153
+    },
+    {
+      "epoch": 0.6007287870900573,
+      "grad_norm": 0.3124787935064602,
+      "learning_rate": 4.6295790539695354e-05,
+      "loss": 0.1642,
+      "step": 1154
+    },
+    {
+      "epoch": 0.601249349297241,
+      "grad_norm": 0.297935731727392,
+      "learning_rate": 4.628842753674203e-05,
+      "loss": 0.1592,
+      "step": 1155
+    },
+    {
+      "epoch": 0.6017699115044248,
+      "grad_norm": 0.2917605841000676,
+      "learning_rate": 4.628105780991261e-05,
+      "loss": 0.1738,
+      "step": 1156
+    },
+    {
+      "epoch": 0.6022904737116085,
+      "grad_norm": 0.2742189116346638,
+      "learning_rate": 4.6273681361534796e-05,
+      "loss": 0.1676,
+      "step": 1157
+    },
+    {
+      "epoch": 0.6028110359187923,
+      "grad_norm": 0.2831911499474275,
+      "learning_rate": 4.626629819393842e-05,
+      "loss": 0.1659,
+      "step": 1158
+    },
+    {
+      "epoch": 0.603331598125976,
+      "grad_norm": 0.30018675422590946,
+      "learning_rate": 4.6258908309455424e-05,
+      "loss": 0.1624,
+      "step": 1159
+    },
+    {
+      "epoch": 0.6038521603331598,
+      "grad_norm": 0.29631188813134685,
+      "learning_rate": 4.625151171041988e-05,
+      "loss": 0.1671,
+      "step": 1160
+    },
+    {
+      "epoch": 0.6043727225403436,
+      "grad_norm": 0.29206096012142957,
+      "learning_rate": 4.624410839916798e-05,
+      "loss": 0.1755,
+      "step": 1161
+    },
+    {
+      "epoch": 0.6048932847475273,
+      "grad_norm": 0.29805577625196694,
+      "learning_rate": 4.6236698378038026e-05,
+      "loss": 0.1674,
+      "step": 1162
+    },
+    {
+      "epoch": 0.6054138469547111,
+      "grad_norm": 0.284263114326214,
+      "learning_rate": 4.622928164937046e-05,
+      "loss": 0.1611,
+      "step": 1163
+    },
+    {
+      "epoch": 0.6059344091618949,
+      "grad_norm": 0.29763872316389595,
+      "learning_rate": 4.622185821550782e-05,
+      "loss": 0.1753,
+      "step": 1164
+    },
+    {
+      "epoch": 0.6064549713690786,
+      "grad_norm": 0.2869777890373756,
+      "learning_rate": 4.621442807879477e-05,
+      "loss": 0.1554,
+      "step": 1165
+    },
+    {
+      "epoch": 0.6069755335762623,
+      "grad_norm": 0.2743988330225703,
+      "learning_rate": 4.6206991241578115e-05,
+      "loss": 0.1662,
+      "step": 1166
+    },
+    {
+      "epoch": 0.6074960957834461,
+      "grad_norm": 0.27540743834473974,
+      "learning_rate": 4.6199547706206726e-05,
+      "loss": 0.1632,
+      "step": 1167
+    },
+    {
+      "epoch": 0.6080166579906299,
+      "grad_norm": 0.28312699949690634,
+      "learning_rate": 4.619209747503163e-05,
+      "loss": 0.1636,
+      "step": 1168
+    },
+    {
+      "epoch": 0.6085372201978136,
+      "grad_norm": 0.2637175185952636,
+      "learning_rate": 4.618464055040595e-05,
+      "loss": 0.1649,
+      "step": 1169
+    },
+    {
+      "epoch": 0.6090577824049974,
+      "grad_norm": 0.2907917728207116,
+      "learning_rate": 4.617717693468494e-05,
+      "loss": 0.1688,
+      "step": 1170
+    },
+    {
+      "epoch": 0.6095783446121812,
+      "grad_norm": 0.2726681176203265,
+      "learning_rate": 4.616970663022596e-05,
+      "loss": 0.1695,
+      "step": 1171
+    },
+    {
+      "epoch": 0.610098906819365,
+      "grad_norm": 0.3021128819540319,
+      "learning_rate": 4.616222963938847e-05,
+      "loss": 0.1689,
+      "step": 1172
+    },
+    {
+      "epoch": 0.6106194690265486,
+      "grad_norm": 0.2611672657701669,
+      "learning_rate": 4.615474596453405e-05,
+      "loss": 0.1601,
+      "step": 1173
+    },
+    {
+      "epoch": 0.6111400312337324,
+      "grad_norm": 0.27250524535283555,
+      "learning_rate": 4.6147255608026394e-05,
+      "loss": 0.1595,
+      "step": 1174
+    },
+    {
+      "epoch": 0.6116605934409162,
+      "grad_norm": 0.2780403650376052,
+      "learning_rate": 4.6139758572231315e-05,
+      "loss": 0.1613,
+      "step": 1175
+    },
+    {
+      "epoch": 0.6121811556480999,
+      "grad_norm": 0.28819496612165146,
+      "learning_rate": 4.613225485951672e-05,
+      "loss": 0.1701,
+      "step": 1176
+    },
+    {
+      "epoch": 0.6127017178552837,
+      "grad_norm": 0.289358937511583,
+      "learning_rate": 4.612474447225263e-05,
+      "loss": 0.1663,
+      "step": 1177
+    },
+    {
+      "epoch": 0.6132222800624675,
+      "grad_norm": 0.2720168641397797,
+      "learning_rate": 4.611722741281118e-05,
+      "loss": 0.1637,
+      "step": 1178
+    },
+    {
+      "epoch": 0.6137428422696513,
+      "grad_norm": 0.278359207879303,
+      "learning_rate": 4.610970368356659e-05,
+      "loss": 0.1644,
+      "step": 1179
+    },
+    {
+      "epoch": 0.614263404476835,
+      "grad_norm": 0.26805244317170895,
+      "learning_rate": 4.610217328689522e-05,
+      "loss": 0.1681,
+      "step": 1180
+    },
+    {
+      "epoch": 0.6147839666840187,
+      "grad_norm": 0.28205230211004056,
+      "learning_rate": 4.609463622517551e-05,
+      "loss": 0.1669,
+      "step": 1181
+    },
+    {
+      "epoch": 0.6153045288912025,
+      "grad_norm": 0.269236517123811,
+      "learning_rate": 4.608709250078803e-05,
+      "loss": 0.1619,
+      "step": 1182
+    },
+    {
+      "epoch": 0.6158250910983862,
+      "grad_norm": 0.29722253174417096,
+      "learning_rate": 4.607954211611543e-05,
+      "loss": 0.1677,
+      "step": 1183
+    },
+    {
+      "epoch": 0.61634565330557,
+      "grad_norm": 0.2726195453326547,
+      "learning_rate": 4.6071985073542464e-05,
+      "loss": 0.1656,
+      "step": 1184
+    },
+    {
+      "epoch": 0.6168662155127538,
+      "grad_norm": 0.2945066289976058,
+      "learning_rate": 4.606442137545602e-05,
+      "loss": 0.1577,
+      "step": 1185
+    },
+    {
+      "epoch": 0.6173867777199376,
+      "grad_norm": 0.3225752803253219,
+      "learning_rate": 4.605685102424504e-05,
+      "loss": 0.1711,
+      "step": 1186
+    },
+    {
+      "epoch": 0.6179073399271213,
+      "grad_norm": 0.3155533738497284,
+      "learning_rate": 4.6049274022300604e-05,
+      "loss": 0.1711,
+      "step": 1187
+    },
+    {
+      "epoch": 0.6184279021343051,
+      "grad_norm": 0.29404463902874495,
+      "learning_rate": 4.604169037201589e-05,
+      "loss": 0.1644,
+      "step": 1188
+    },
+    {
+      "epoch": 0.6189484643414888,
+      "grad_norm": 0.28530333471749375,
+      "learning_rate": 4.603410007578616e-05,
+      "loss": 0.1611,
+      "step": 1189
+    },
+    {
+      "epoch": 0.6194690265486725,
+      "grad_norm": 0.2841758773756127,
+      "learning_rate": 4.602650313600878e-05,
+      "loss": 0.1655,
+      "step": 1190
+    },
+    {
+      "epoch": 0.6199895887558563,
+      "grad_norm": 0.29874212816673795,
+      "learning_rate": 4.601889955508322e-05,
+      "loss": 0.1678,
+      "step": 1191
+    },
+    {
+      "epoch": 0.6205101509630401,
+      "grad_norm": 0.2790940324283885,
+      "learning_rate": 4.601128933541105e-05,
+      "loss": 0.1669,
+      "step": 1192
+    },
+    {
+      "epoch": 0.6210307131702238,
+      "grad_norm": 0.2819144485571506,
+      "learning_rate": 4.600367247939591e-05,
+      "loss": 0.1594,
+      "step": 1193
+    },
+    {
+      "epoch": 0.6215512753774076,
+      "grad_norm": 0.2789287724673938,
+      "learning_rate": 4.5996048989443597e-05,
+      "loss": 0.163,
+      "step": 1194
+    },
+    {
+      "epoch": 0.6220718375845914,
+      "grad_norm": 0.2862188445057931,
+      "learning_rate": 4.598841886796192e-05,
+      "loss": 0.1752,
+      "step": 1195
+    },
+    {
+      "epoch": 0.6225923997917752,
+      "grad_norm": 0.3007115626630004,
+      "learning_rate": 4.598078211736086e-05,
+      "loss": 0.1795,
+      "step": 1196
+    },
+    {
+      "epoch": 0.6231129619989588,
+      "grad_norm": 0.2655000681844505,
+      "learning_rate": 4.5973138740052455e-05,
+      "loss": 0.1652,
+      "step": 1197
+    },
+    {
+      "epoch": 0.6236335242061426,
+      "grad_norm": 0.3051733991758238,
+      "learning_rate": 4.596548873845081e-05,
+      "loss": 0.1638,
+      "step": 1198
+    },
+    {
+      "epoch": 0.6241540864133264,
+      "grad_norm": 0.2928160917194925,
+      "learning_rate": 4.595783211497219e-05,
+      "loss": 0.169,
+      "step": 1199
+    },
+    {
+      "epoch": 0.6246746486205101,
+      "grad_norm": 0.29112623504997737,
+      "learning_rate": 4.5950168872034885e-05,
+      "loss": 0.163,
+      "step": 1200
+    },
+    {
+      "epoch": 0.6251952108276939,
+      "grad_norm": 0.2675061785401877,
+      "learning_rate": 4.5942499012059316e-05,
+      "loss": 0.1657,
+      "step": 1201
+    },
+    {
+      "epoch": 0.6257157730348777,
+      "grad_norm": 0.3040715761688371,
+      "learning_rate": 4.593482253746798e-05,
+      "loss": 0.1623,
+      "step": 1202
+    },
+    {
+      "epoch": 0.6262363352420615,
+      "grad_norm": 0.29614925372633366,
+      "learning_rate": 4.592713945068545e-05,
+      "loss": 0.1643,
+      "step": 1203
+    },
+    {
+      "epoch": 0.6267568974492452,
+      "grad_norm": 0.2638537701916355,
+      "learning_rate": 4.591944975413843e-05,
+      "loss": 0.1618,
+      "step": 1204
+    },
+    {
+      "epoch": 0.6272774596564289,
+      "grad_norm": 0.3005257646399674,
+      "learning_rate": 4.5911753450255665e-05,
+      "loss": 0.1598,
+      "step": 1205
+    },
+    {
+      "epoch": 0.6277980218636127,
+      "grad_norm": 0.29494144677802137,
+      "learning_rate": 4.590405054146802e-05,
+      "loss": 0.1718,
+      "step": 1206
+    },
+    {
+      "epoch": 0.6283185840707964,
+      "grad_norm": 0.29636762114195014,
+      "learning_rate": 4.5896341030208415e-05,
+      "loss": 0.1637,
+      "step": 1207
+    },
+    {
+      "epoch": 0.6288391462779802,
+      "grad_norm": 0.2934192989066668,
+      "learning_rate": 4.5888624918911884e-05,
+      "loss": 0.158,
+      "step": 1208
+    },
+    {
+      "epoch": 0.629359708485164,
+      "grad_norm": 0.3124230206823834,
+      "learning_rate": 4.588090221001553e-05,
+      "loss": 0.1672,
+      "step": 1209
+    },
+    {
+      "epoch": 0.6298802706923478,
+      "grad_norm": 0.30615588299221824,
+      "learning_rate": 4.587317290595855e-05,
+      "loss": 0.1629,
+      "step": 1210
+    },
+    {
+      "epoch": 0.6304008328995315,
+      "grad_norm": 0.2988825912755345,
+      "learning_rate": 4.586543700918221e-05,
+      "loss": 0.1723,
+      "step": 1211
+    },
+    {
+      "epoch": 0.6309213951067153,
+      "grad_norm": 0.29738980460994413,
+      "learning_rate": 4.5857694522129855e-05,
+      "loss": 0.1573,
+      "step": 1212
+    },
+    {
+      "epoch": 0.631441957313899,
+      "grad_norm": 0.2797032464087098,
+      "learning_rate": 4.584994544724695e-05,
+      "loss": 0.1678,
+      "step": 1213
+    },
+    {
+      "epoch": 0.6319625195210827,
+      "grad_norm": 0.26491065606515474,
+      "learning_rate": 4.584218978698099e-05,
+      "loss": 0.1626,
+      "step": 1214
+    },
+    {
+      "epoch": 0.6324830817282665,
+      "grad_norm": 0.2652322894153915,
+      "learning_rate": 4.5834427543781596e-05,
+      "loss": 0.1666,
+      "step": 1215
+    },
+    {
+      "epoch": 0.6330036439354503,
+      "grad_norm": 0.264807354619328,
+      "learning_rate": 4.582665872010043e-05,
+      "loss": 0.1641,
+      "step": 1216
+    },
+    {
+      "epoch": 0.633524206142634,
+      "grad_norm": 0.2738204236926736,
+      "learning_rate": 4.581888331839125e-05,
+      "loss": 0.1703,
+      "step": 1217
+    },
+    {
+      "epoch": 0.6340447683498178,
+      "grad_norm": 0.299863744726911,
+      "learning_rate": 4.581110134110989e-05,
+      "loss": 0.1667,
+      "step": 1218
+    },
+    {
+      "epoch": 0.6345653305570016,
+      "grad_norm": 0.2956315906472553,
+      "learning_rate": 4.580331279071426e-05,
+      "loss": 0.1685,
+      "step": 1219
+    },
+    {
+      "epoch": 0.6350858927641854,
+      "grad_norm": 0.28753743007877214,
+      "learning_rate": 4.579551766966435e-05,
+      "loss": 0.1608,
+      "step": 1220
+    },
+    {
+      "epoch": 0.635606454971369,
+      "grad_norm": 0.28657496675188177,
+      "learning_rate": 4.578771598042221e-05,
+      "loss": 0.1692,
+      "step": 1221
+    },
+    {
+      "epoch": 0.6361270171785528,
+      "grad_norm": 0.27201718233205613,
+      "learning_rate": 4.5779907725452e-05,
+      "loss": 0.1592,
+      "step": 1222
+    },
+    {
+      "epoch": 0.6366475793857366,
+      "grad_norm": 0.26136498905384203,
+      "learning_rate": 4.577209290721991e-05,
+      "loss": 0.1625,
+      "step": 1223
+    },
+    {
+      "epoch": 0.6371681415929203,
+      "grad_norm": 0.2898833505811805,
+      "learning_rate": 4.576427152819423e-05,
+      "loss": 0.1699,
+      "step": 1224
+    },
+    {
+      "epoch": 0.6376887038001041,
+      "grad_norm": 0.2668251937477724,
+      "learning_rate": 4.575644359084532e-05,
+      "loss": 0.1591,
+      "step": 1225
+    },
+    {
+      "epoch": 0.6382092660072879,
+      "grad_norm": 0.3003314921218274,
+      "learning_rate": 4.5748609097645595e-05,
+      "loss": 0.1673,
+      "step": 1226
+    },
+    {
+      "epoch": 0.6387298282144717,
+      "grad_norm": 0.2684194237776865,
+      "learning_rate": 4.574076805106956e-05,
+      "loss": 0.1587,
+      "step": 1227
+    },
+    {
+      "epoch": 0.6392503904216554,
+      "grad_norm": 0.2845756731306716,
+      "learning_rate": 4.5732920453593785e-05,
+      "loss": 0.1642,
+      "step": 1228
+    },
+    {
+      "epoch": 0.6397709526288391,
+      "grad_norm": 0.27739696831325483,
+      "learning_rate": 4.572506630769691e-05,
+      "loss": 0.1623,
+      "step": 1229
+    },
+    {
+      "epoch": 0.6402915148360229,
+      "grad_norm": 0.27961191274533287,
+      "learning_rate": 4.571720561585963e-05,
+      "loss": 0.155,
+      "step": 1230
+    },
+    {
+      "epoch": 0.6408120770432066,
+      "grad_norm": 0.2807787449921395,
+      "learning_rate": 4.570933838056472e-05,
+      "loss": 0.1569,
+      "step": 1231
+    },
+    {
+      "epoch": 0.6413326392503904,
+      "grad_norm": 0.27535967325114985,
+      "learning_rate": 4.570146460429701e-05,
+      "loss": 0.1534,
+      "step": 1232
+    },
+    {
+      "epoch": 0.6418532014575742,
+      "grad_norm": 0.31053243876551606,
+      "learning_rate": 4.569358428954343e-05,
+      "loss": 0.1687,
+      "step": 1233
+    },
+    {
+      "epoch": 0.642373763664758,
+      "grad_norm": 0.2735613339054094,
+      "learning_rate": 4.568569743879293e-05,
+      "loss": 0.1619,
+      "step": 1234
+    },
+    {
+      "epoch": 0.6428943258719417,
+      "grad_norm": 0.31025137855515883,
+      "learning_rate": 4.5677804054536544e-05,
+      "loss": 0.1669,
+      "step": 1235
+    },
+    {
+      "epoch": 0.6434148880791255,
+      "grad_norm": 0.3108370328370114,
+      "learning_rate": 4.566990413926738e-05,
+      "loss": 0.1616,
+      "step": 1236
+    },
+    {
+      "epoch": 0.6439354502863092,
+      "grad_norm": 0.3339548204834231,
+      "learning_rate": 4.5661997695480595e-05,
+      "loss": 0.1616,
+      "step": 1237
+    },
+    {
+      "epoch": 0.6444560124934929,
+      "grad_norm": 0.3056231219128003,
+      "learning_rate": 4.5654084725673404e-05,
+      "loss": 0.1654,
+      "step": 1238
+    },
+    {
+      "epoch": 0.6449765747006767,
+      "grad_norm": 0.29210355340648536,
+      "learning_rate": 4.564616523234511e-05,
+      "loss": 0.1686,
+      "step": 1239
+    },
+    {
+      "epoch": 0.6454971369078605,
+      "grad_norm": 0.29213232167646325,
+      "learning_rate": 4.5638239217997034e-05,
+      "loss": 0.1621,
+      "step": 1240
+    },
+    {
+      "epoch": 0.6460176991150443,
+      "grad_norm": 0.28705338754520415,
+      "learning_rate": 4.56303066851326e-05,
+      "loss": 0.166,
+      "step": 1241
+    },
+    {
+      "epoch": 0.646538261322228,
+      "grad_norm": 0.3057168972516095,
+      "learning_rate": 4.5622367636257264e-05,
+      "loss": 0.161,
+      "step": 1242
+    },
+    {
+      "epoch": 0.6470588235294118,
+      "grad_norm": 0.2716038210905968,
+      "learning_rate": 4.561442207387854e-05,
+      "loss": 0.1628,
+      "step": 1243
+    },
+    {
+      "epoch": 0.6475793857365956,
+      "grad_norm": 0.3359502419591334,
+      "learning_rate": 4.560647000050602e-05,
+      "loss": 0.1641,
+      "step": 1244
+    },
+    {
+      "epoch": 0.6480999479437792,
+      "grad_norm": 0.2677924387144485,
+      "learning_rate": 4.5598511418651324e-05,
+      "loss": 0.1692,
+      "step": 1245
+    },
+    {
+      "epoch": 0.648620510150963,
+      "grad_norm": 0.2687659769423282,
+      "learning_rate": 4.5590546330828154e-05,
+      "loss": 0.1652,
+      "step": 1246
+    },
+    {
+      "epoch": 0.6491410723581468,
+      "grad_norm": 0.27911537151113053,
+      "learning_rate": 4.5582574739552254e-05,
+      "loss": 0.1681,
+      "step": 1247
+    },
+    {
+      "epoch": 0.6496616345653305,
+      "grad_norm": 0.258050298522539,
+      "learning_rate": 4.557459664734141e-05,
+      "loss": 0.1632,
+      "step": 1248
+    },
+    {
+      "epoch": 0.6501821967725143,
+      "grad_norm": 0.2714387670557213,
+      "learning_rate": 4.5566612056715494e-05,
+      "loss": 0.1602,
+      "step": 1249
+    },
+    {
+      "epoch": 0.6507027589796981,
+      "grad_norm": 0.3007289110925111,
+      "learning_rate": 4.5558620970196406e-05,
+      "loss": 0.1603,
+      "step": 1250
+    },
+    {
+      "epoch": 0.6512233211868819,
+      "grad_norm": 0.2751562123537327,
+      "learning_rate": 4.5550623390308086e-05,
+      "loss": 0.1637,
+      "step": 1251
+    },
+    {
+      "epoch": 0.6517438833940656,
+      "grad_norm": 0.26206723907101465,
+      "learning_rate": 4.554261931957657e-05,
+      "loss": 0.1602,
+      "step": 1252
+    },
+    {
+      "epoch": 0.6522644456012493,
+      "grad_norm": 0.29168906928554644,
+      "learning_rate": 4.5534608760529895e-05,
+      "loss": 0.1655,
+      "step": 1253
+    },
+    {
+      "epoch": 0.6527850078084331,
+      "grad_norm": 0.272148724671491,
+      "learning_rate": 4.552659171569817e-05,
+      "loss": 0.1567,
+      "step": 1254
+    },
+    {
+      "epoch": 0.6533055700156168,
+      "grad_norm": 0.2918024931656026,
+      "learning_rate": 4.551856818761357e-05,
+      "loss": 0.1656,
+      "step": 1255
+    },
+    {
+      "epoch": 0.6538261322228006,
+      "grad_norm": 0.2897642359961369,
+      "learning_rate": 4.551053817881028e-05,
+      "loss": 0.1635,
+      "step": 1256
+    },
+    {
+      "epoch": 0.6543466944299844,
+      "grad_norm": 0.2675950487049928,
+      "learning_rate": 4.550250169182455e-05,
+      "loss": 0.1571,
+      "step": 1257
+    },
+    {
+      "epoch": 0.6548672566371682,
+      "grad_norm": 0.28801786916053895,
+      "learning_rate": 4.549445872919468e-05,
+      "loss": 0.1616,
+      "step": 1258
+    },
+    {
+      "epoch": 0.6553878188443519,
+      "grad_norm": 0.2602745738457694,
+      "learning_rate": 4.548640929346102e-05,
+      "loss": 0.1629,
+      "step": 1259
+    },
+    {
+      "epoch": 0.6559083810515357,
+      "grad_norm": 0.2876893050906943,
+      "learning_rate": 4.5478353387165946e-05,
+      "loss": 0.1575,
+      "step": 1260
+    },
+    {
+      "epoch": 0.6564289432587194,
+      "grad_norm": 0.2881481837684052,
+      "learning_rate": 4.547029101285389e-05,
+      "loss": 0.1625,
+      "step": 1261
+    },
+    {
+      "epoch": 0.6569495054659031,
+      "grad_norm": 0.2612428197323997,
+      "learning_rate": 4.5462222173071335e-05,
+      "loss": 0.1652,
+      "step": 1262
+    },
+    {
+      "epoch": 0.6574700676730869,
+      "grad_norm": 0.2701617714724092,
+      "learning_rate": 4.5454146870366775e-05,
+      "loss": 0.1652,
+      "step": 1263
+    },
+    {
+      "epoch": 0.6579906298802707,
+      "grad_norm": 0.28509359219297975,
+      "learning_rate": 4.5446065107290786e-05,
+      "loss": 0.164,
+      "step": 1264
+    },
+    {
+      "epoch": 0.6585111920874545,
+      "grad_norm": 0.26256618912001867,
+      "learning_rate": 4.543797688639596e-05,
+      "loss": 0.1554,
+      "step": 1265
+    },
+    {
+      "epoch": 0.6590317542946382,
+      "grad_norm": 0.27155835804114575,
+      "learning_rate": 4.5429882210236926e-05,
+      "loss": 0.1613,
+      "step": 1266
+    },
+    {
+      "epoch": 0.659552316501822,
+      "grad_norm": 0.2918933060230762,
+      "learning_rate": 4.542178108137038e-05,
+      "loss": 0.1656,
+      "step": 1267
+    },
+    {
+      "epoch": 0.6600728787090058,
+      "grad_norm": 0.26850912684301664,
+      "learning_rate": 4.5413673502355e-05,
+      "loss": 0.1658,
+      "step": 1268
+    },
+    {
+      "epoch": 0.6605934409161894,
+      "grad_norm": 0.2555610557403195,
+      "learning_rate": 4.540555947575157e-05,
+      "loss": 0.1555,
+      "step": 1269
+    },
+    {
+      "epoch": 0.6611140031233732,
+      "grad_norm": 0.492044807765271,
+      "learning_rate": 4.539743900412287e-05,
+      "loss": 0.1668,
+      "step": 1270
+    },
+    {
+      "epoch": 0.661634565330557,
+      "grad_norm": 0.2749406988327086,
+      "learning_rate": 4.53893120900337e-05,
+      "loss": 0.1571,
+      "step": 1271
+    },
+    {
+      "epoch": 0.6621551275377408,
+      "grad_norm": 0.2755114222031435,
+      "learning_rate": 4.538117873605094e-05,
+      "loss": 0.1604,
+      "step": 1272
+    },
+    {
+      "epoch": 0.6626756897449245,
+      "grad_norm": 0.27347742130970076,
+      "learning_rate": 4.537303894474349e-05,
+      "loss": 0.1628,
+      "step": 1273
+    },
+    {
+      "epoch": 0.6631962519521083,
+      "grad_norm": 0.2859200379145871,
+      "learning_rate": 4.536489271868225e-05,
+      "loss": 0.1663,
+      "step": 1274
+    },
+    {
+      "epoch": 0.6637168141592921,
+      "grad_norm": 0.2855335291443677,
+      "learning_rate": 4.5356740060440194e-05,
+      "loss": 0.1699,
+      "step": 1275
+    },
+    {
+      "epoch": 0.6642373763664758,
+      "grad_norm": 0.26154237753790416,
+      "learning_rate": 4.53485809725923e-05,
+      "loss": 0.159,
+      "step": 1276
+    },
+    {
+      "epoch": 0.6647579385736595,
+      "grad_norm": 0.2993939894006488,
+      "learning_rate": 4.53404154577156e-05,
+      "loss": 0.166,
+      "step": 1277
+    },
+    {
+      "epoch": 0.6652785007808433,
+      "grad_norm": 0.2747912605662342,
+      "learning_rate": 4.533224351838914e-05,
+      "loss": 0.1515,
+      "step": 1278
+    },
+    {
+      "epoch": 0.665799062988027,
+      "grad_norm": 0.2736057262584086,
+      "learning_rate": 4.532406515719399e-05,
+      "loss": 0.1642,
+      "step": 1279
+    },
+    {
+      "epoch": 0.6663196251952108,
+      "grad_norm": 0.2616489484367182,
+      "learning_rate": 4.531588037671326e-05,
+      "loss": 0.161,
+      "step": 1280
+    },
+    {
+      "epoch": 0.6668401874023946,
+      "grad_norm": 0.26322733292018735,
+      "learning_rate": 4.5307689179532085e-05,
+      "loss": 0.1574,
+      "step": 1281
+    },
+    {
+      "epoch": 0.6673607496095784,
+      "grad_norm": 0.2522456597339631,
+      "learning_rate": 4.529949156823764e-05,
+      "loss": 0.1615,
+      "step": 1282
+    },
+    {
+      "epoch": 0.6678813118167621,
+      "grad_norm": 0.2588976061704534,
+      "learning_rate": 4.529128754541909e-05,
+      "loss": 0.155,
+      "step": 1283
+    },
+    {
+      "epoch": 0.6684018740239459,
+      "grad_norm": 0.27256034050552175,
+      "learning_rate": 4.528307711366767e-05,
+      "loss": 0.171,
+      "step": 1284
+    },
+    {
+      "epoch": 0.6689224362311296,
+      "grad_norm": 0.2538123767756482,
+      "learning_rate": 4.527486027557659e-05,
+      "loss": 0.1617,
+      "step": 1285
+    },
+    {
+      "epoch": 0.6694429984383133,
+      "grad_norm": 0.27020857968681916,
+      "learning_rate": 4.526663703374113e-05,
+      "loss": 0.1648,
+      "step": 1286
+    },
+    {
+      "epoch": 0.6699635606454971,
+      "grad_norm": 0.2772645799194081,
+      "learning_rate": 4.525840739075857e-05,
+      "loss": 0.1679,
+      "step": 1287
+    },
+    {
+      "epoch": 0.6704841228526809,
+      "grad_norm": 0.2842658573717009,
+      "learning_rate": 4.525017134922821e-05,
+      "loss": 0.1704,
+      "step": 1288
+    },
+    {
+      "epoch": 0.6710046850598647,
+      "grad_norm": 0.2695068859454936,
+      "learning_rate": 4.524192891175138e-05,
+      "loss": 0.1685,
+      "step": 1289
+    },
+    {
+      "epoch": 0.6715252472670484,
+      "grad_norm": 0.2563249875383298,
+      "learning_rate": 4.5233680080931415e-05,
+      "loss": 0.1575,
+      "step": 1290
+    },
+    {
+      "epoch": 0.6720458094742322,
+      "grad_norm": 0.25282111526223566,
+      "learning_rate": 4.522542485937369e-05,
+      "loss": 0.1581,
+      "step": 1291
+    },
+    {
+      "epoch": 0.672566371681416,
+      "grad_norm": 0.27360763527857895,
+      "learning_rate": 4.521716324968558e-05,
+      "loss": 0.1611,
+      "step": 1292
+    },
+    {
+      "epoch": 0.6730869338885996,
+      "grad_norm": 0.2635465269224253,
+      "learning_rate": 4.520889525447649e-05,
+      "loss": 0.1613,
+      "step": 1293
+    },
+    {
+      "epoch": 0.6736074960957834,
+      "grad_norm": 0.2826648530069749,
+      "learning_rate": 4.520062087635784e-05,
+      "loss": 0.1597,
+      "step": 1294
+    },
+    {
+      "epoch": 0.6741280583029672,
+      "grad_norm": 0.27562512183656485,
+      "learning_rate": 4.5192340117943063e-05,
+      "loss": 0.1606,
+      "step": 1295
+    },
+    {
+      "epoch": 0.674648620510151,
+      "grad_norm": 0.25422173470452253,
+      "learning_rate": 4.51840529818476e-05,
+      "loss": 0.1591,
+      "step": 1296
+    },
+    {
+      "epoch": 0.6751691827173347,
+      "grad_norm": 0.28058709543239363,
+      "learning_rate": 4.517575947068893e-05,
+      "loss": 0.1595,
+      "step": 1297
+    },
+    {
+      "epoch": 0.6756897449245185,
+      "grad_norm": 0.26381435173704143,
+      "learning_rate": 4.516745958708652e-05,
+      "loss": 0.1613,
+      "step": 1298
+    },
+    {
+      "epoch": 0.6762103071317023,
+      "grad_norm": 0.28533524054694864,
+      "learning_rate": 4.5159153333661854e-05,
+      "loss": 0.1611,
+      "step": 1299
+    },
+    {
+      "epoch": 0.676730869338886,
+      "grad_norm": 0.26541389475587734,
+      "learning_rate": 4.515084071303843e-05,
+      "loss": 0.1638,
+      "step": 1300
+    },
+    {
+      "epoch": 0.6772514315460697,
+      "grad_norm": 0.2695669437026599,
+      "learning_rate": 4.514252172784178e-05,
+      "loss": 0.1654,
+      "step": 1301
+    },
+    {
+      "epoch": 0.6777719937532535,
+      "grad_norm": 0.252500452369742,
+      "learning_rate": 4.513419638069942e-05,
+      "loss": 0.1612,
+      "step": 1302
+    },
+    {
+      "epoch": 0.6782925559604372,
+      "grad_norm": 0.2593569283907477,
+      "learning_rate": 4.512586467424087e-05,
+      "loss": 0.1573,
+      "step": 1303
+    },
+    {
+      "epoch": 0.678813118167621,
+      "grad_norm": 0.2783028111560752,
+      "learning_rate": 4.511752661109768e-05,
+      "loss": 0.1734,
+      "step": 1304
+    },
+    {
+      "epoch": 0.6793336803748048,
+      "grad_norm": 0.26185446584580446,
+      "learning_rate": 4.51091821939034e-05,
+      "loss": 0.1634,
+      "step": 1305
+    },
+    {
+      "epoch": 0.6798542425819886,
+      "grad_norm": 0.2713826383452847,
+      "learning_rate": 4.510083142529359e-05,
+      "loss": 0.1627,
+      "step": 1306
+    },
+    {
+      "epoch": 0.6803748047891723,
+      "grad_norm": 0.31074372566311925,
+      "learning_rate": 4.5092474307905785e-05,
+      "loss": 0.1587,
+      "step": 1307
+    },
+    {
+      "epoch": 0.6808953669963561,
+      "grad_norm": 0.2487669262126004,
+      "learning_rate": 4.5084110844379584e-05,
+      "loss": 0.1637,
+      "step": 1308
+    },
+    {
+      "epoch": 0.6814159292035398,
+      "grad_norm": 0.2869857982196813,
+      "learning_rate": 4.507574103735654e-05,
+      "loss": 0.1698,
+      "step": 1309
+    },
+    {
+      "epoch": 0.6819364914107235,
+      "grad_norm": 0.2790644846811937,
+      "learning_rate": 4.506736488948024e-05,
+      "loss": 0.1653,
+      "step": 1310
+    },
+    {
+      "epoch": 0.6824570536179073,
+      "grad_norm": 0.27483161848797866,
+      "learning_rate": 4.5058982403396244e-05,
+      "loss": 0.1572,
+      "step": 1311
+    },
+    {
+      "epoch": 0.6829776158250911,
+      "grad_norm": 0.26249570703461655,
+      "learning_rate": 4.505059358175214e-05,
+      "loss": 0.1545,
+      "step": 1312
+    },
+    {
+      "epoch": 0.6834981780322749,
+      "grad_norm": 0.27003457349130416,
+      "learning_rate": 4.504219842719751e-05,
+      "loss": 0.1548,
+      "step": 1313
+    },
+    {
+      "epoch": 0.6840187402394586,
+      "grad_norm": 0.27502243535142196,
+      "learning_rate": 4.503379694238394e-05,
+      "loss": 0.1626,
+      "step": 1314
+    },
+    {
+      "epoch": 0.6845393024466424,
+      "grad_norm": 0.3145544670289913,
+      "learning_rate": 4.502538912996499e-05,
+      "loss": 0.1674,
+      "step": 1315
+    },
+    {
+      "epoch": 0.6850598646538262,
+      "grad_norm": 0.2662239312987116,
+      "learning_rate": 4.501697499259626e-05,
+      "loss": 0.1597,
+      "step": 1316
+    },
+    {
+      "epoch": 0.6855804268610098,
+      "grad_norm": 0.2745260554615778,
+      "learning_rate": 4.500855453293532e-05,
+      "loss": 0.163,
+      "step": 1317
+    },
+    {
+      "epoch": 0.6861009890681936,
+      "grad_norm": 0.2657154339820778,
+      "learning_rate": 4.500012775364173e-05,
+      "loss": 0.1635,
+      "step": 1318
+    },
+    {
+      "epoch": 0.6866215512753774,
+      "grad_norm": 0.2832483197266812,
+      "learning_rate": 4.499169465737708e-05,
+      "loss": 0.169,
+      "step": 1319
+    },
+    {
+      "epoch": 0.6871421134825612,
+      "grad_norm": 0.2893358983294465,
+      "learning_rate": 4.498325524680492e-05,
+      "loss": 0.1657,
+      "step": 1320
+    },
+    {
+      "epoch": 0.6876626756897449,
+      "grad_norm": 0.2716434135511077,
+      "learning_rate": 4.4974809524590814e-05,
+      "loss": 0.1682,
+      "step": 1321
+    },
+    {
+      "epoch": 0.6881832378969287,
+      "grad_norm": 0.31689018683469905,
+      "learning_rate": 4.496635749340231e-05,
+      "loss": 0.1634,
+      "step": 1322
+    },
+    {
+      "epoch": 0.6887038001041125,
+      "grad_norm": 0.27519106860257103,
+      "learning_rate": 4.495789915590895e-05,
+      "loss": 0.1571,
+      "step": 1323
+    },
+    {
+      "epoch": 0.6892243623112962,
+      "grad_norm": 0.3070166511158656,
+      "learning_rate": 4.494943451478229e-05,
+      "loss": 0.1625,
+      "step": 1324
+    },
+    {
+      "epoch": 0.6897449245184799,
+      "grad_norm": 0.26270345276600776,
+      "learning_rate": 4.4940963572695836e-05,
+      "loss": 0.1594,
+      "step": 1325
+    },
+    {
+      "epoch": 0.6902654867256637,
+      "grad_norm": 0.2885003932017069,
+      "learning_rate": 4.4932486332325115e-05,
+      "loss": 0.1595,
+      "step": 1326
+    },
+    {
+      "epoch": 0.6907860489328475,
+      "grad_norm": 0.2571626418463362,
+      "learning_rate": 4.492400279634763e-05,
+      "loss": 0.1616,
+      "step": 1327
+    },
+    {
+      "epoch": 0.6913066111400312,
+      "grad_norm": 0.27955455405062174,
+      "learning_rate": 4.491551296744288e-05,
+      "loss": 0.172,
+      "step": 1328
+    },
+    {
+      "epoch": 0.691827173347215,
+      "grad_norm": 0.2691060387491177,
+      "learning_rate": 4.490701684829235e-05,
+      "loss": 0.1613,
+      "step": 1329
+    },
+    {
+      "epoch": 0.6923477355543988,
+      "grad_norm": 0.2739293243057916,
+      "learning_rate": 4.48985144415795e-05,
+      "loss": 0.1665,
+      "step": 1330
+    },
+    {
+      "epoch": 0.6928682977615825,
+      "grad_norm": 0.29877047824582253,
+      "learning_rate": 4.489000574998979e-05,
+      "loss": 0.1644,
+      "step": 1331
+    },
+    {
+      "epoch": 0.6933888599687663,
+      "grad_norm": 0.2707072638145847,
+      "learning_rate": 4.488149077621067e-05,
+      "loss": 0.1614,
+      "step": 1332
+    },
+    {
+      "epoch": 0.69390942217595,
+      "grad_norm": 0.2782882849164749,
+      "learning_rate": 4.4872969522931556e-05,
+      "loss": 0.1646,
+      "step": 1333
+    },
+    {
+      "epoch": 0.6944299843831337,
+      "grad_norm": 0.27327557400755875,
+      "learning_rate": 4.486444199284386e-05,
+      "loss": 0.1659,
+      "step": 1334
+    },
+    {
+      "epoch": 0.6949505465903175,
+      "grad_norm": 0.2806747248809842,
+      "learning_rate": 4.4855908188640973e-05,
+      "loss": 0.1622,
+      "step": 1335
+    },
+    {
+      "epoch": 0.6954711087975013,
+      "grad_norm": 0.2690587883243696,
+      "learning_rate": 4.484736811301826e-05,
+      "loss": 0.1573,
+      "step": 1336
+    },
+    {
+      "epoch": 0.6959916710046851,
+      "grad_norm": 0.27243238739702624,
+      "learning_rate": 4.483882176867308e-05,
+      "loss": 0.1686,
+      "step": 1337
+    },
+    {
+      "epoch": 0.6965122332118688,
+      "grad_norm": 0.28204045583043863,
+      "learning_rate": 4.483026915830477e-05,
+      "loss": 0.163,
+      "step": 1338
+    },
+    {
+      "epoch": 0.6970327954190526,
+      "grad_norm": 0.2772739553002315,
+      "learning_rate": 4.4821710284614636e-05,
+      "loss": 0.1624,
+      "step": 1339
+    },
+    {
+      "epoch": 0.6975533576262364,
+      "grad_norm": 0.29231455388507804,
+      "learning_rate": 4.4813145150305965e-05,
+      "loss": 0.1616,
+      "step": 1340
+    },
+    {
+      "epoch": 0.69807391983342,
+      "grad_norm": 0.2640575107760888,
+      "learning_rate": 4.4804573758084046e-05,
+      "loss": 0.1585,
+      "step": 1341
+    },
+    {
+      "epoch": 0.6985944820406038,
+      "grad_norm": 0.3071872152649576,
+      "learning_rate": 4.4795996110656105e-05,
+      "loss": 0.1673,
+      "step": 1342
+    },
+    {
+      "epoch": 0.6991150442477876,
+      "grad_norm": 0.2749143884079691,
+      "learning_rate": 4.478741221073136e-05,
+      "loss": 0.1609,
+      "step": 1343
+    },
+    {
+      "epoch": 0.6996356064549714,
+      "grad_norm": 0.2754640339891129,
+      "learning_rate": 4.477882206102101e-05,
+      "loss": 0.1606,
+      "step": 1344
+    },
+    {
+      "epoch": 0.7001561686621551,
+      "grad_norm": 0.2639768649579088,
+      "learning_rate": 4.477022566423823e-05,
+      "loss": 0.162,
+      "step": 1345
+    },
+    {
+      "epoch": 0.7006767308693389,
+      "grad_norm": 0.30604022298071126,
+      "learning_rate": 4.476162302309815e-05,
+      "loss": 0.1628,
+      "step": 1346
+    },
+    {
+      "epoch": 0.7011972930765227,
+      "grad_norm": 0.26151484803979685,
+      "learning_rate": 4.475301414031791e-05,
+      "loss": 0.1561,
+      "step": 1347
+    },
+    {
+      "epoch": 0.7017178552837064,
+      "grad_norm": 0.3069542520801269,
+      "learning_rate": 4.4744399018616566e-05,
+      "loss": 0.1717,
+      "step": 1348
+    },
+    {
+      "epoch": 0.7022384174908901,
+      "grad_norm": 0.26118236966791997,
+      "learning_rate": 4.4735777660715186e-05,
+      "loss": 0.1594,
+      "step": 1349
+    },
+    {
+      "epoch": 0.7027589796980739,
+      "grad_norm": 0.2821690722820644,
+      "learning_rate": 4.472715006933681e-05,
+      "loss": 0.17,
+      "step": 1350
+    },
+    {
+      "epoch": 0.7032795419052577,
+      "grad_norm": 0.28425332393518327,
+      "learning_rate": 4.47185162472064e-05,
+      "loss": 0.1656,
+      "step": 1351
+    },
+    {
+      "epoch": 0.7038001041124414,
+      "grad_norm": 0.2895241985407839,
+      "learning_rate": 4.470987619705095e-05,
+      "loss": 0.172,
+      "step": 1352
+    },
+    {
+      "epoch": 0.7043206663196252,
+      "grad_norm": 0.2907907968206427,
+      "learning_rate": 4.470122992159938e-05,
+      "loss": 0.1577,
+      "step": 1353
+    },
+    {
+      "epoch": 0.704841228526809,
+      "grad_norm": 0.26360172964644457,
+      "learning_rate": 4.469257742358258e-05,
+      "loss": 0.1641,
+      "step": 1354
+    },
+    {
+      "epoch": 0.7053617907339927,
+      "grad_norm": 0.30825700427174524,
+      "learning_rate": 4.468391870573342e-05,
+      "loss": 0.1613,
+      "step": 1355
+    },
+    {
+      "epoch": 0.7058823529411765,
+      "grad_norm": 0.2832274278276974,
+      "learning_rate": 4.467525377078672e-05,
+      "loss": 0.1616,
+      "step": 1356
+    },
+    {
+      "epoch": 0.7064029151483602,
+      "grad_norm": 0.2918275203692509,
+      "learning_rate": 4.466658262147927e-05,
+      "loss": 0.1576,
+      "step": 1357
+    },
+    {
+      "epoch": 0.706923477355544,
+      "grad_norm": 0.2623509716876954,
+      "learning_rate": 4.465790526054983e-05,
+      "loss": 0.1548,
+      "step": 1358
+    },
+    {
+      "epoch": 0.7074440395627277,
+      "grad_norm": 0.2830506524726103,
+      "learning_rate": 4.4649221690739095e-05,
+      "loss": 0.1577,
+      "step": 1359
+    },
+    {
+      "epoch": 0.7079646017699115,
+      "grad_norm": 0.2668706458778001,
+      "learning_rate": 4.464053191478976e-05,
+      "loss": 0.1693,
+      "step": 1360
+    },
+    {
+      "epoch": 0.7084851639770953,
+      "grad_norm": 0.2529153576757616,
+      "learning_rate": 4.463183593544647e-05,
+      "loss": 0.1637,
+      "step": 1361
+    },
+    {
+      "epoch": 0.709005726184279,
+      "grad_norm": 0.2624787559358483,
+      "learning_rate": 4.462313375545579e-05,
+      "loss": 0.1612,
+      "step": 1362
+    },
+    {
+      "epoch": 0.7095262883914628,
+      "grad_norm": 0.2751465060223429,
+      "learning_rate": 4.461442537756629e-05,
+      "loss": 0.1549,
+      "step": 1363
+    },
+    {
+      "epoch": 0.7100468505986466,
+      "grad_norm": 0.24739529719217984,
+      "learning_rate": 4.4605710804528474e-05,
+      "loss": 0.162,
+      "step": 1364
+    },
+    {
+      "epoch": 0.7105674128058302,
+      "grad_norm": 0.26495855736115675,
+      "learning_rate": 4.459699003909482e-05,
+      "loss": 0.1659,
+      "step": 1365
+    },
+    {
+      "epoch": 0.711087975013014,
+      "grad_norm": 0.27568214342295516,
+      "learning_rate": 4.4588263084019746e-05,
+      "loss": 0.1635,
+      "step": 1366
+    },
+    {
+      "epoch": 0.7116085372201978,
+      "grad_norm": 0.25595986364541223,
+      "learning_rate": 4.457952994205963e-05,
+      "loss": 0.1569,
+      "step": 1367
+    },
+    {
+      "epoch": 0.7121290994273816,
+      "grad_norm": 0.2731708239567679,
+      "learning_rate": 4.457079061597281e-05,
+      "loss": 0.1603,
+      "step": 1368
+    },
+    {
+      "epoch": 0.7126496616345653,
+      "grad_norm": 0.2587032809571712,
+      "learning_rate": 4.4562045108519565e-05,
+      "loss": 0.1609,
+      "step": 1369
+    },
+    {
+      "epoch": 0.7131702238417491,
+      "grad_norm": 0.26523949719681184,
+      "learning_rate": 4.4553293422462134e-05,
+      "loss": 0.1578,
+      "step": 1370
+    },
+    {
+      "epoch": 0.7136907860489329,
+      "grad_norm": 0.2594662248421297,
+      "learning_rate": 4.454453556056471e-05,
+      "loss": 0.1585,
+      "step": 1371
+    },
+    {
+      "epoch": 0.7142113482561167,
+      "grad_norm": 0.2557915985298669,
+      "learning_rate": 4.4535771525593426e-05,
+      "loss": 0.1536,
+      "step": 1372
+    },
+    {
+      "epoch": 0.7147319104633003,
+      "grad_norm": 0.2742705642572711,
+      "learning_rate": 4.452700132031638e-05,
+      "loss": 0.1673,
+      "step": 1373
+    },
+    {
+      "epoch": 0.7152524726704841,
+      "grad_norm": 0.2806193185281088,
+      "learning_rate": 4.451822494750362e-05,
+      "loss": 0.1648,
+      "step": 1374
+    },
+    {
+      "epoch": 0.7157730348776679,
+      "grad_norm": 0.2809327056175134,
+      "learning_rate": 4.450944240992711e-05,
+      "loss": 0.1656,
+      "step": 1375
+    },
+    {
+      "epoch": 0.7162935970848516,
+      "grad_norm": 0.2759622382555981,
+      "learning_rate": 4.45006537103608e-05,
+      "loss": 0.1612,
+      "step": 1376
+    },
+    {
+      "epoch": 0.7168141592920354,
+      "grad_norm": 0.2624638331201247,
+      "learning_rate": 4.449185885158056e-05,
+      "loss": 0.1668,
+      "step": 1377
+    },
+    {
+      "epoch": 0.7173347214992192,
+      "grad_norm": 0.2689887733722017,
+      "learning_rate": 4.4483057836364225e-05,
+      "loss": 0.1608,
+      "step": 1378
+    },
+    {
+      "epoch": 0.717855283706403,
+      "grad_norm": 0.26933355871425685,
+      "learning_rate": 4.4474250667491567e-05,
+      "loss": 0.1629,
+      "step": 1379
+    },
+    {
+      "epoch": 0.7183758459135867,
+      "grad_norm": 0.2735573234930184,
+      "learning_rate": 4.4465437347744285e-05,
+      "loss": 0.1647,
+      "step": 1380
+    },
+    {
+      "epoch": 0.7188964081207704,
+      "grad_norm": 0.2680034708829736,
+      "learning_rate": 4.4456617879906056e-05,
+      "loss": 0.1678,
+      "step": 1381
+    },
+    {
+      "epoch": 0.7194169703279542,
+      "grad_norm": 0.2519063095093893,
+      "learning_rate": 4.444779226676246e-05,
+      "loss": 0.1592,
+      "step": 1382
+    },
+    {
+      "epoch": 0.7199375325351379,
+      "grad_norm": 0.2602912105474266,
+      "learning_rate": 4.4438960511101046e-05,
+      "loss": 0.1575,
+      "step": 1383
+    },
+    {
+      "epoch": 0.7204580947423217,
+      "grad_norm": 0.2726477230959709,
+      "learning_rate": 4.443012261571129e-05,
+      "loss": 0.1604,
+      "step": 1384
+    },
+    {
+      "epoch": 0.7209786569495055,
+      "grad_norm": 0.27905940371356286,
+      "learning_rate": 4.442127858338462e-05,
+      "loss": 0.1654,
+      "step": 1385
+    },
+    {
+      "epoch": 0.7214992191566892,
+      "grad_norm": 0.2583560743210376,
+      "learning_rate": 4.441242841691438e-05,
+      "loss": 0.1528,
+      "step": 1386
+    },
+    {
+      "epoch": 0.722019781363873,
+      "grad_norm": 0.27099777286648374,
+      "learning_rate": 4.440357211909586e-05,
+      "loss": 0.1667,
+      "step": 1387
+    },
+    {
+      "epoch": 0.7225403435710568,
+      "grad_norm": 0.2894163802959934,
+      "learning_rate": 4.439470969272631e-05,
+      "loss": 0.1645,
+      "step": 1388
+    },
+    {
+      "epoch": 0.7230609057782404,
+      "grad_norm": 0.2544924261078653,
+      "learning_rate": 4.4385841140604884e-05,
+      "loss": 0.1585,
+      "step": 1389
+    },
+    {
+      "epoch": 0.7235814679854242,
+      "grad_norm": 0.27636086548700217,
+      "learning_rate": 4.437696646553269e-05,
+      "loss": 0.1523,
+      "step": 1390
+    },
+    {
+      "epoch": 0.724102030192608,
+      "grad_norm": 0.2713258603157374,
+      "learning_rate": 4.4368085670312755e-05,
+      "loss": 0.1645,
+      "step": 1391
+    },
+    {
+      "epoch": 0.7246225923997918,
+      "grad_norm": 0.2764758232297857,
+      "learning_rate": 4.435919875775005e-05,
+      "loss": 0.1526,
+      "step": 1392
+    },
+    {
+      "epoch": 0.7251431546069755,
+      "grad_norm": 0.26060022386382026,
+      "learning_rate": 4.435030573065149e-05,
+      "loss": 0.1524,
+      "step": 1393
+    },
+    {
+      "epoch": 0.7256637168141593,
+      "grad_norm": 0.29792250327932235,
+      "learning_rate": 4.434140659182588e-05,
+      "loss": 0.1666,
+      "step": 1394
+    },
+    {
+      "epoch": 0.7261842790213431,
+      "grad_norm": 0.26037287596745995,
+      "learning_rate": 4.433250134408401e-05,
+      "loss": 0.1586,
+      "step": 1395
+    },
+    {
+      "epoch": 0.7267048412285269,
+      "grad_norm": 0.26727896558958614,
+      "learning_rate": 4.4323589990238545e-05,
+      "loss": 0.164,
+      "step": 1396
+    },
+    {
+      "epoch": 0.7272254034357105,
+      "grad_norm": 0.2754571354466698,
+      "learning_rate": 4.431467253310413e-05,
+      "loss": 0.1659,
+      "step": 1397
+    },
+    {
+      "epoch": 0.7277459656428943,
+      "grad_norm": 0.26954157638947546,
+      "learning_rate": 4.4305748975497294e-05,
+      "loss": 0.1591,
+      "step": 1398
+    },
+    {
+      "epoch": 0.7282665278500781,
+      "grad_norm": 0.262164138500046,
+      "learning_rate": 4.4296819320236524e-05,
+      "loss": 0.1605,
+      "step": 1399
+    },
+    {
+      "epoch": 0.7287870900572618,
+      "grad_norm": 0.27456638161498215,
+      "learning_rate": 4.428788357014222e-05,
+      "loss": 0.1572,
+      "step": 1400
+    },
+    {
+      "epoch": 0.7293076522644456,
+      "grad_norm": 0.2720040860336858,
+      "learning_rate": 4.4278941728036696e-05,
+      "loss": 0.1667,
+      "step": 1401
+    },
+    {
+      "epoch": 0.7298282144716294,
+      "grad_norm": 0.2636935543695637,
+      "learning_rate": 4.426999379674421e-05,
+      "loss": 0.1678,
+      "step": 1402
+    },
+    {
+      "epoch": 0.7303487766788131,
+      "grad_norm": 0.27172824164676596,
+      "learning_rate": 4.426103977909094e-05,
+      "loss": 0.1654,
+      "step": 1403
+    },
+    {
+      "epoch": 0.7308693388859969,
+      "grad_norm": 0.2557001119750677,
+      "learning_rate": 4.425207967790497e-05,
+      "loss": 0.1598,
+      "step": 1404
+    },
+    {
+      "epoch": 0.7313899010931806,
+      "grad_norm": 0.26111748434009874,
+      "learning_rate": 4.4243113496016326e-05,
+      "loss": 0.1587,
+      "step": 1405
+    },
+    {
+      "epoch": 0.7319104633003644,
+      "grad_norm": 0.2833129354173884,
+      "learning_rate": 4.423414123625694e-05,
+      "loss": 0.162,
+      "step": 1406
+    },
+    {
+      "epoch": 0.7324310255075481,
+      "grad_norm": 0.2575834825018457,
+      "learning_rate": 4.4225162901460676e-05,
+      "loss": 0.1597,
+      "step": 1407
+    },
+    {
+      "epoch": 0.7329515877147319,
+      "grad_norm": 0.2753343459846711,
+      "learning_rate": 4.42161784944633e-05,
+      "loss": 0.17,
+      "step": 1408
+    },
+    {
+      "epoch": 0.7334721499219157,
+      "grad_norm": 0.2971924494704701,
+      "learning_rate": 4.420718801810252e-05,
+      "loss": 0.1563,
+      "step": 1409
+    },
+    {
+      "epoch": 0.7339927121290994,
+      "grad_norm": 0.2579981593383931,
+      "learning_rate": 4.419819147521793e-05,
+      "loss": 0.1586,
+      "step": 1410
+    },
+    {
+      "epoch": 0.7345132743362832,
+      "grad_norm": 0.2763781873570336,
+      "learning_rate": 4.418918886865108e-05,
+      "loss": 0.1679,
+      "step": 1411
+    },
+    {
+      "epoch": 0.735033836543467,
+      "grad_norm": 0.27888896994369716,
+      "learning_rate": 4.418018020124538e-05,
+      "loss": 0.158,
+      "step": 1412
+    },
+    {
+      "epoch": 0.7355543987506507,
+      "grad_norm": 0.27992425465308046,
+      "learning_rate": 4.417116547584621e-05,
+      "loss": 0.1582,
+      "step": 1413
+    },
+    {
+      "epoch": 0.7360749609578344,
+      "grad_norm": 0.28643991156800214,
+      "learning_rate": 4.4162144695300834e-05,
+      "loss": 0.1674,
+      "step": 1414
+    },
+    {
+      "epoch": 0.7365955231650182,
+      "grad_norm": 0.2642558460300512,
+      "learning_rate": 4.415311786245843e-05,
+      "loss": 0.1645,
+      "step": 1415
+    },
+    {
+      "epoch": 0.737116085372202,
+      "grad_norm": 0.27570896762545,
+      "learning_rate": 4.41440849801701e-05,
+      "loss": 0.1634,
+      "step": 1416
+    },
+    {
+      "epoch": 0.7376366475793857,
+      "grad_norm": 0.26823860392672444,
+      "learning_rate": 4.413504605128885e-05,
+      "loss": 0.1615,
+      "step": 1417
+    },
+    {
+      "epoch": 0.7381572097865695,
+      "grad_norm": 0.27413381020561156,
+      "learning_rate": 4.4126001078669574e-05,
+      "loss": 0.1638,
+      "step": 1418
+    },
+    {
+      "epoch": 0.7386777719937533,
+      "grad_norm": 0.2682256952523028,
+      "learning_rate": 4.4116950065169124e-05,
+      "loss": 0.1604,
+      "step": 1419
+    },
+    {
+      "epoch": 0.7391983342009371,
+      "grad_norm": 0.2934261761780093,
+      "learning_rate": 4.410789301364621e-05,
+      "loss": 0.1665,
+      "step": 1420
+    },
+    {
+      "epoch": 0.7397188964081207,
+      "grad_norm": 0.26702280332753814,
+      "learning_rate": 4.409882992696148e-05,
+      "loss": 0.1584,
+      "step": 1421
+    },
+    {
+      "epoch": 0.7402394586153045,
+      "grad_norm": 0.27945453060552594,
+      "learning_rate": 4.4089760807977474e-05,
+      "loss": 0.153,
+      "step": 1422
+    },
+    {
+      "epoch": 0.7407600208224883,
+      "grad_norm": 0.26111156671410596,
+      "learning_rate": 4.408068565955865e-05,
+      "loss": 0.1633,
+      "step": 1423
+    },
+    {
+      "epoch": 0.741280583029672,
+      "grad_norm": 0.2618834889479115,
+      "learning_rate": 4.407160448457135e-05,
+      "loss": 0.1633,
+      "step": 1424
+    },
+    {
+      "epoch": 0.7418011452368558,
+      "grad_norm": 0.2548316686323859,
+      "learning_rate": 4.406251728588384e-05,
+      "loss": 0.1611,
+      "step": 1425
+    },
+    {
+      "epoch": 0.7423217074440396,
+      "grad_norm": 0.27198720127436277,
+      "learning_rate": 4.405342406636627e-05,
+      "loss": 0.1518,
+      "step": 1426
+    },
+    {
+      "epoch": 0.7428422696512234,
+      "grad_norm": 0.25780399068453724,
+      "learning_rate": 4.4044324828890715e-05,
+      "loss": 0.1539,
+      "step": 1427
+    },
+    {
+      "epoch": 0.7433628318584071,
+      "grad_norm": 0.28708604768264273,
+      "learning_rate": 4.403521957633113e-05,
+      "loss": 0.1677,
+      "step": 1428
+    },
+    {
+      "epoch": 0.7438833940655908,
+      "grad_norm": 0.2854990852578142,
+      "learning_rate": 4.4026108311563394e-05,
+      "loss": 0.1591,
+      "step": 1429
+    },
+    {
+      "epoch": 0.7444039562727746,
+      "grad_norm": 0.2601422292209927,
+      "learning_rate": 4.401699103746524e-05,
+      "loss": 0.1628,
+      "step": 1430
+    },
+    {
+      "epoch": 0.7449245184799583,
+      "grad_norm": 0.26792776342524677,
+      "learning_rate": 4.4007867756916345e-05,
+      "loss": 0.1576,
+      "step": 1431
+    },
+    {
+      "epoch": 0.7454450806871421,
+      "grad_norm": 0.2553899632837398,
+      "learning_rate": 4.399873847279827e-05,
+      "loss": 0.1689,
+      "step": 1432
+    },
+    {
+      "epoch": 0.7459656428943259,
+      "grad_norm": 0.2631912627429382,
+      "learning_rate": 4.3989603187994454e-05,
+      "loss": 0.1544,
+      "step": 1433
+    },
+    {
+      "epoch": 0.7464862051015096,
+      "grad_norm": 0.2719680455212801,
+      "learning_rate": 4.398046190539025e-05,
+      "loss": 0.1648,
+      "step": 1434
+    },
+    {
+      "epoch": 0.7470067673086934,
+      "grad_norm": 0.2623299175088546,
+      "learning_rate": 4.39713146278729e-05,
+      "loss": 0.1685,
+      "step": 1435
+    },
+    {
+      "epoch": 0.7475273295158772,
+      "grad_norm": 0.27789094278392973,
+      "learning_rate": 4.3962161358331546e-05,
+      "loss": 0.1588,
+      "step": 1436
+    },
+    {
+      "epoch": 0.7480478917230609,
+      "grad_norm": 0.2625407847886398,
+      "learning_rate": 4.395300209965721e-05,
+      "loss": 0.1616,
+      "step": 1437
+    },
+    {
+      "epoch": 0.7485684539302446,
+      "grad_norm": 0.2506574404285458,
+      "learning_rate": 4.394383685474281e-05,
+      "loss": 0.1522,
+      "step": 1438
+    },
+    {
+      "epoch": 0.7490890161374284,
+      "grad_norm": 0.26122428518589536,
+      "learning_rate": 4.3934665626483175e-05,
+      "loss": 0.1654,
+      "step": 1439
+    },
+    {
+      "epoch": 0.7496095783446122,
+      "grad_norm": 0.25983335544974423,
+      "learning_rate": 4.392548841777497e-05,
+      "loss": 0.1592,
+      "step": 1440
+    },
+    {
+      "epoch": 0.7501301405517959,
+      "grad_norm": 0.26244778657616386,
+      "learning_rate": 4.391630523151683e-05,
+      "loss": 0.163,
+      "step": 1441
+    },
+    {
+      "epoch": 0.7506507027589797,
+      "grad_norm": 0.24661772307545587,
+      "learning_rate": 4.390711607060919e-05,
+      "loss": 0.156,
+      "step": 1442
+    },
+    {
+      "epoch": 0.7511712649661635,
+      "grad_norm": 0.2681126416443138,
+      "learning_rate": 4.389792093795444e-05,
+      "loss": 0.1658,
+      "step": 1443
+    },
+    {
+      "epoch": 0.7516918271733473,
+      "grad_norm": 0.28319213581654684,
+      "learning_rate": 4.3888719836456823e-05,
+      "loss": 0.1603,
+      "step": 1444
+    },
+    {
+      "epoch": 0.7522123893805309,
+      "grad_norm": 0.2717403925825211,
+      "learning_rate": 4.3879512769022485e-05,
+      "loss": 0.1618,
+      "step": 1445
+    },
+    {
+      "epoch": 0.7527329515877147,
+      "grad_norm": 0.2672807713995564,
+      "learning_rate": 4.387029973855943e-05,
+      "loss": 0.164,
+      "step": 1446
+    },
+    {
+      "epoch": 0.7532535137948985,
+      "grad_norm": 0.29074179763902774,
+      "learning_rate": 4.3861080747977565e-05,
+      "loss": 0.159,
+      "step": 1447
+    },
+    {
+      "epoch": 0.7537740760020822,
+      "grad_norm": 0.2644544725678007,
+      "learning_rate": 4.385185580018869e-05,
+      "loss": 0.1608,
+      "step": 1448
+    },
+    {
+      "epoch": 0.754294638209266,
+      "grad_norm": 0.2786228789121159,
+      "learning_rate": 4.3842624898106464e-05,
+      "loss": 0.1663,
+      "step": 1449
+    },
+    {
+      "epoch": 0.7548152004164498,
+      "grad_norm": 0.2568411909485992,
+      "learning_rate": 4.383338804464643e-05,
+      "loss": 0.1632,
+      "step": 1450
+    },
+    {
+      "epoch": 0.7553357626236336,
+      "grad_norm": 0.2704063306890479,
+      "learning_rate": 4.382414524272602e-05,
+      "loss": 0.1559,
+      "step": 1451
+    },
+    {
+      "epoch": 0.7558563248308173,
+      "grad_norm": 0.3052629008083044,
+      "learning_rate": 4.3814896495264544e-05,
+      "loss": 0.1667,
+      "step": 1452
+    },
+    {
+      "epoch": 0.756376887038001,
+      "grad_norm": 0.25388945533730933,
+      "learning_rate": 4.380564180518318e-05,
+      "loss": 0.1596,
+      "step": 1453
+    },
+    {
+      "epoch": 0.7568974492451848,
+      "grad_norm": 0.2876303341079924,
+      "learning_rate": 4.3796381175405014e-05,
+      "loss": 0.1571,
+      "step": 1454
+    },
+    {
+      "epoch": 0.7574180114523685,
+      "grad_norm": 0.26037804849757196,
+      "learning_rate": 4.378711460885494e-05,
+      "loss": 0.157,
+      "step": 1455
+    },
+    {
+      "epoch": 0.7579385736595523,
+      "grad_norm": 0.281072192425641,
+      "learning_rate": 4.377784210845981e-05,
+      "loss": 0.1605,
+      "step": 1456
+    },
+    {
+      "epoch": 0.7584591358667361,
+      "grad_norm": 0.28910101581498615,
+      "learning_rate": 4.376856367714829e-05,
+      "loss": 0.1627,
+      "step": 1457
+    },
+    {
+      "epoch": 0.7589796980739199,
+      "grad_norm": 0.27247204660508845,
+      "learning_rate": 4.375927931785095e-05,
+      "loss": 0.16,
+      "step": 1458
+    },
+    {
+      "epoch": 0.7595002602811036,
+      "grad_norm": 0.24843057073483057,
+      "learning_rate": 4.3749989033500224e-05,
+      "loss": 0.1552,
+      "step": 1459
+    },
+    {
+      "epoch": 0.7600208224882874,
+      "grad_norm": 0.27297992132744714,
+      "learning_rate": 4.3740692827030404e-05,
+      "loss": 0.1545,
+      "step": 1460
+    },
+    {
+      "epoch": 0.7605413846954711,
+      "grad_norm": 0.25731476919380347,
+      "learning_rate": 4.3731390701377675e-05,
+      "loss": 0.1547,
+      "step": 1461
+    },
+    {
+      "epoch": 0.7610619469026548,
+      "grad_norm": 0.27094205358391144,
+      "learning_rate": 4.3722082659480076e-05,
+      "loss": 0.1509,
+      "step": 1462
+    },
+    {
+      "epoch": 0.7615825091098386,
+      "grad_norm": 0.27051817962373237,
+      "learning_rate": 4.371276870427753e-05,
+      "loss": 0.1514,
+      "step": 1463
+    },
+    {
+      "epoch": 0.7621030713170224,
+      "grad_norm": 0.2776605729402276,
+      "learning_rate": 4.37034488387118e-05,
+      "loss": 0.1648,
+      "step": 1464
+    },
+    {
+      "epoch": 0.7626236335242061,
+      "grad_norm": 0.2632398573808106,
+      "learning_rate": 4.3694123065726553e-05,
+      "loss": 0.1625,
+      "step": 1465
+    },
+    {
+      "epoch": 0.7631441957313899,
+      "grad_norm": 0.2631443834365562,
+      "learning_rate": 4.3684791388267287e-05,
+      "loss": 0.1588,
+      "step": 1466
+    },
+    {
+      "epoch": 0.7636647579385737,
+      "grad_norm": 0.26943148019545626,
+      "learning_rate": 4.367545380928139e-05,
+      "loss": 0.1552,
+      "step": 1467
+    },
+    {
+      "epoch": 0.7641853201457575,
+      "grad_norm": 0.2649233188506122,
+      "learning_rate": 4.36661103317181e-05,
+      "loss": 0.1615,
+      "step": 1468
+    },
+    {
+      "epoch": 0.7647058823529411,
+      "grad_norm": 0.26835305552005984,
+      "learning_rate": 4.3656760958528506e-05,
+      "loss": 0.1637,
+      "step": 1469
+    },
+    {
+      "epoch": 0.7652264445601249,
+      "grad_norm": 0.2578891624619545,
+      "learning_rate": 4.364740569266561e-05,
+      "loss": 0.152,
+      "step": 1470
+    },
+    {
+      "epoch": 0.7657470067673087,
+      "grad_norm": 0.26524236019881625,
+      "learning_rate": 4.363804453708421e-05,
+      "loss": 0.1676,
+      "step": 1471
+    },
+    {
+      "epoch": 0.7662675689744924,
+      "grad_norm": 0.2583254156227514,
+      "learning_rate": 4.362867749474101e-05,
+      "loss": 0.1543,
+      "step": 1472
+    },
+    {
+      "epoch": 0.7667881311816762,
+      "grad_norm": 0.2577114988061728,
+      "learning_rate": 4.361930456859455e-05,
+      "loss": 0.1546,
+      "step": 1473
+    },
+    {
+      "epoch": 0.76730869338886,
+      "grad_norm": 0.2655743069628856,
+      "learning_rate": 4.360992576160524e-05,
+      "loss": 0.1563,
+      "step": 1474
+    },
+    {
+      "epoch": 0.7678292555960438,
+      "grad_norm": 0.25925048966807795,
+      "learning_rate": 4.3600541076735346e-05,
+      "loss": 0.1562,
+      "step": 1475
+    },
+    {
+      "epoch": 0.7683498178032275,
+      "grad_norm": 0.26321769667524203,
+      "learning_rate": 4.359115051694898e-05,
+      "loss": 0.1587,
+      "step": 1476
+    },
+    {
+      "epoch": 0.7688703800104112,
+      "grad_norm": 0.2792326144746915,
+      "learning_rate": 4.358175408521212e-05,
+      "loss": 0.1555,
+      "step": 1477
+    },
+    {
+      "epoch": 0.769390942217595,
+      "grad_norm": 0.26756537678776365,
+      "learning_rate": 4.357235178449261e-05,
+      "loss": 0.1666,
+      "step": 1478
+    },
+    {
+      "epoch": 0.7699115044247787,
+      "grad_norm": 0.27412529165548094,
+      "learning_rate": 4.356294361776012e-05,
+      "loss": 0.164,
+      "step": 1479
+    },
+    {
+      "epoch": 0.7704320666319625,
+      "grad_norm": 0.2632617066780282,
+      "learning_rate": 4.3553529587986184e-05,
+      "loss": 0.1554,
+      "step": 1480
+    },
+    {
+      "epoch": 0.7709526288391463,
+      "grad_norm": 0.2665209273532276,
+      "learning_rate": 4.3544109698144206e-05,
+      "loss": 0.157,
+      "step": 1481
+    },
+    {
+      "epoch": 0.77147319104633,
+      "grad_norm": 0.26609079034706307,
+      "learning_rate": 4.3534683951209416e-05,
+      "loss": 0.1561,
+      "step": 1482
+    },
+    {
+      "epoch": 0.7719937532535138,
+      "grad_norm": 0.2782722665482425,
+      "learning_rate": 4.3525252350158904e-05,
+      "loss": 0.1633,
+      "step": 1483
+    },
+    {
+      "epoch": 0.7725143154606976,
+      "grad_norm": 0.24586603055391748,
+      "learning_rate": 4.351581489797161e-05,
+      "loss": 0.1625,
+      "step": 1484
+    },
+    {
+      "epoch": 0.7730348776678813,
+      "grad_norm": 0.27660175665693065,
+      "learning_rate": 4.350637159762831e-05,
+      "loss": 0.166,
+      "step": 1485
+    },
+    {
+      "epoch": 0.773555439875065,
+      "grad_norm": 0.2718655540992913,
+      "learning_rate": 4.3496922452111656e-05,
+      "loss": 0.162,
+      "step": 1486
+    },
+    {
+      "epoch": 0.7740760020822488,
+      "grad_norm": 0.2583348630376606,
+      "learning_rate": 4.348746746440612e-05,
+      "loss": 0.1632,
+      "step": 1487
+    },
+    {
+      "epoch": 0.7745965642894326,
+      "grad_norm": 0.2692273301860101,
+      "learning_rate": 4.347800663749801e-05,
+      "loss": 0.1513,
+      "step": 1488
+    },
+    {
+      "epoch": 0.7751171264966163,
+      "grad_norm": 0.26291032421691457,
+      "learning_rate": 4.3468539974375534e-05,
+      "loss": 0.1545,
+      "step": 1489
+    },
+    {
+      "epoch": 0.7756376887038001,
+      "grad_norm": 0.27809527086476604,
+      "learning_rate": 4.345906747802867e-05,
+      "loss": 0.1593,
+      "step": 1490
+    },
+    {
+      "epoch": 0.7761582509109839,
+      "grad_norm": 0.25920104592248266,
+      "learning_rate": 4.344958915144929e-05,
+      "loss": 0.1609,
+      "step": 1491
+    },
+    {
+      "epoch": 0.7766788131181677,
+      "grad_norm": 0.29330224785096815,
+      "learning_rate": 4.3440104997631084e-05,
+      "loss": 0.1686,
+      "step": 1492
+    },
+    {
+      "epoch": 0.7771993753253513,
+      "grad_norm": 0.2727948495509996,
+      "learning_rate": 4.343061501956959e-05,
+      "loss": 0.1592,
+      "step": 1493
+    },
+    {
+      "epoch": 0.7777199375325351,
+      "grad_norm": 0.25156183079449285,
+      "learning_rate": 4.3421119220262185e-05,
+      "loss": 0.161,
+      "step": 1494
+    },
+    {
+      "epoch": 0.7782404997397189,
+      "grad_norm": 0.26655483033933713,
+      "learning_rate": 4.3411617602708085e-05,
+      "loss": 0.1556,
+      "step": 1495
+    },
+    {
+      "epoch": 0.7787610619469026,
+      "grad_norm": 0.25869405345561675,
+      "learning_rate": 4.340211016990834e-05,
+      "loss": 0.1518,
+      "step": 1496
+    },
+    {
+      "epoch": 0.7792816241540864,
+      "grad_norm": 0.25895490167143587,
+      "learning_rate": 4.3392596924865854e-05,
+      "loss": 0.1583,
+      "step": 1497
+    },
+    {
+      "epoch": 0.7798021863612702,
+      "grad_norm": 0.26930038307801346,
+      "learning_rate": 4.3383077870585334e-05,
+      "loss": 0.1578,
+      "step": 1498
+    },
+    {
+      "epoch": 0.780322748568454,
+      "grad_norm": 0.2655149958667704,
+      "learning_rate": 4.3373553010073355e-05,
+      "loss": 0.1666,
+      "step": 1499
+    },
+    {
+      "epoch": 0.7808433107756377,
+      "grad_norm": 0.2700875671699295,
+      "learning_rate": 4.3364022346338295e-05,
+      "loss": 0.1499,
+      "step": 1500
+    },
+    {
+      "epoch": 0.7813638729828214,
+      "grad_norm": 0.26841277303835387,
+      "learning_rate": 4.335448588239039e-05,
+      "loss": 0.16,
+      "step": 1501
+    },
+    {
+      "epoch": 0.7818844351900052,
+      "grad_norm": 0.2664241191896805,
+      "learning_rate": 4.33449436212417e-05,
+      "loss": 0.1566,
+      "step": 1502
+    },
+    {
+      "epoch": 0.7824049973971889,
+      "grad_norm": 0.2748558495343656,
+      "learning_rate": 4.333539556590612e-05,
+      "loss": 0.1561,
+      "step": 1503
+    },
+    {
+      "epoch": 0.7829255596043727,
+      "grad_norm": 0.260891571763646,
+      "learning_rate": 4.332584171939936e-05,
+      "loss": 0.1639,
+      "step": 1504
+    },
+    {
+      "epoch": 0.7834461218115565,
+      "grad_norm": 0.26117170009910184,
+      "learning_rate": 4.331628208473897e-05,
+      "loss": 0.1535,
+      "step": 1505
+    },
+    {
+      "epoch": 0.7839666840187403,
+      "grad_norm": 0.2501439526366261,
+      "learning_rate": 4.3306716664944344e-05,
+      "loss": 0.1558,
+      "step": 1506
+    },
+    {
+      "epoch": 0.784487246225924,
+      "grad_norm": 0.24457951812109338,
+      "learning_rate": 4.329714546303666e-05,
+      "loss": 0.1539,
+      "step": 1507
+    },
+    {
+      "epoch": 0.7850078084331078,
+      "grad_norm": 0.25274194830382657,
+      "learning_rate": 4.328756848203897e-05,
+      "loss": 0.1564,
+      "step": 1508
+    },
+    {
+      "epoch": 0.7855283706402915,
+      "grad_norm": 0.24212505466279638,
+      "learning_rate": 4.327798572497612e-05,
+      "loss": 0.1585,
+      "step": 1509
+    },
+    {
+      "epoch": 0.7860489328474752,
+      "grad_norm": 0.26756424368954107,
+      "learning_rate": 4.3268397194874796e-05,
+      "loss": 0.1617,
+      "step": 1510
+    },
+    {
+      "epoch": 0.786569495054659,
+      "grad_norm": 0.24184787477349412,
+      "learning_rate": 4.32588028947635e-05,
+      "loss": 0.1619,
+      "step": 1511
+    },
+    {
+      "epoch": 0.7870900572618428,
+      "grad_norm": 0.25168399053293283,
+      "learning_rate": 4.3249202827672564e-05,
+      "loss": 0.1602,
+      "step": 1512
+    },
+    {
+      "epoch": 0.7876106194690266,
+      "grad_norm": 0.26161299334783633,
+      "learning_rate": 4.3239596996634125e-05,
+      "loss": 0.1576,
+      "step": 1513
+    },
+    {
+      "epoch": 0.7881311816762103,
+      "grad_norm": 0.23097070104546066,
+      "learning_rate": 4.322998540468216e-05,
+      "loss": 0.149,
+      "step": 1514
+    },
+    {
+      "epoch": 0.7886517438833941,
+      "grad_norm": 0.26176672102662274,
+      "learning_rate": 4.322036805485245e-05,
+      "loss": 0.1604,
+      "step": 1515
+    },
+    {
+      "epoch": 0.7891723060905779,
+      "grad_norm": 0.2509842627379737,
+      "learning_rate": 4.3210744950182603e-05,
+      "loss": 0.1609,
+      "step": 1516
+    },
+    {
+      "epoch": 0.7896928682977615,
+      "grad_norm": 0.25286215696526987,
+      "learning_rate": 4.3201116093712045e-05,
+      "loss": 0.1519,
+      "step": 1517
+    },
+    {
+      "epoch": 0.7902134305049453,
+      "grad_norm": 0.2554977617373658,
+      "learning_rate": 4.319148148848202e-05,
+      "loss": 0.1507,
+      "step": 1518
+    },
+    {
+      "epoch": 0.7907339927121291,
+      "grad_norm": 0.2610030848646668,
+      "learning_rate": 4.3181841137535585e-05,
+      "loss": 0.1565,
+      "step": 1519
+    },
+    {
+      "epoch": 0.7912545549193128,
+      "grad_norm": 0.25847858497390264,
+      "learning_rate": 4.317219504391761e-05,
+      "loss": 0.1513,
+      "step": 1520
+    },
+    {
+      "epoch": 0.7917751171264966,
+      "grad_norm": 0.25023349404312234,
+      "learning_rate": 4.316254321067477e-05,
+      "loss": 0.1632,
+      "step": 1521
+    },
+    {
+      "epoch": 0.7922956793336804,
+      "grad_norm": 0.2641861638565703,
+      "learning_rate": 4.315288564085558e-05,
+      "loss": 0.1618,
+      "step": 1522
+    },
+    {
+      "epoch": 0.7928162415408642,
+      "grad_norm": 0.2421590693178086,
+      "learning_rate": 4.314322233751034e-05,
+      "loss": 0.1558,
+      "step": 1523
+    },
+    {
+      "epoch": 0.7933368037480479,
+      "grad_norm": 0.26085418193866505,
+      "learning_rate": 4.313355330369117e-05,
+      "loss": 0.1595,
+      "step": 1524
+    },
+    {
+      "epoch": 0.7938573659552316,
+      "grad_norm": 0.2439161457038632,
+      "learning_rate": 4.312387854245201e-05,
+      "loss": 0.1466,
+      "step": 1525
+    },
+    {
+      "epoch": 0.7943779281624154,
+      "grad_norm": 0.2713532358303694,
+      "learning_rate": 4.3114198056848585e-05,
+      "loss": 0.1608,
+      "step": 1526
+    },
+    {
+      "epoch": 0.7948984903695991,
+      "grad_norm": 0.25922505436788407,
+      "learning_rate": 4.3104511849938464e-05,
+      "loss": 0.1557,
+      "step": 1527
+    },
+    {
+      "epoch": 0.7954190525767829,
+      "grad_norm": 0.2886790825671015,
+      "learning_rate": 4.309481992478098e-05,
+      "loss": 0.1644,
+      "step": 1528
+    },
+    {
+      "epoch": 0.7959396147839667,
+      "grad_norm": 0.24108201278459665,
+      "learning_rate": 4.308512228443731e-05,
+      "loss": 0.1596,
+      "step": 1529
+    },
+    {
+      "epoch": 0.7964601769911505,
+      "grad_norm": 0.27215617885447624,
+      "learning_rate": 4.30754189319704e-05,
+      "loss": 0.1573,
+      "step": 1530
+    },
+    {
+      "epoch": 0.7969807391983342,
+      "grad_norm": 0.28694200101027506,
+      "learning_rate": 4.306570987044505e-05,
+      "loss": 0.1686,
+      "step": 1531
+    },
+    {
+      "epoch": 0.797501301405518,
+      "grad_norm": 0.24652959196034746,
+      "learning_rate": 4.305599510292781e-05,
+      "loss": 0.159,
+      "step": 1532
+    },
+    {
+      "epoch": 0.7980218636127017,
+      "grad_norm": 0.2819197614862714,
+      "learning_rate": 4.304627463248706e-05,
+      "loss": 0.1588,
+      "step": 1533
+    },
+    {
+      "epoch": 0.7985424258198854,
+      "grad_norm": 0.2909511370680243,
+      "learning_rate": 4.3036548462192986e-05,
+      "loss": 0.1606,
+      "step": 1534
+    },
+    {
+      "epoch": 0.7990629880270692,
+      "grad_norm": 0.2530278943329065,
+      "learning_rate": 4.302681659511755e-05,
+      "loss": 0.1544,
+      "step": 1535
+    },
+    {
+      "epoch": 0.799583550234253,
+      "grad_norm": 0.282363644399392,
+      "learning_rate": 4.301707903433454e-05,
+      "loss": 0.1556,
+      "step": 1536
+    },
+    {
+      "epoch": 0.8001041124414368,
+      "grad_norm": 0.2676419532705495,
+      "learning_rate": 4.300733578291953e-05,
+      "loss": 0.1637,
+      "step": 1537
+    },
+    {
+      "epoch": 0.8006246746486205,
+      "grad_norm": 0.25796205124564936,
+      "learning_rate": 4.29975868439499e-05,
+      "loss": 0.1609,
+      "step": 1538
+    },
+    {
+      "epoch": 0.8011452368558043,
+      "grad_norm": 0.2819765513723248,
+      "learning_rate": 4.29878322205048e-05,
+      "loss": 0.1624,
+      "step": 1539
+    },
+    {
+      "epoch": 0.8016657990629881,
+      "grad_norm": 0.27897029452667205,
+      "learning_rate": 4.297807191566521e-05,
+      "loss": 0.1572,
+      "step": 1540
+    },
+    {
+      "epoch": 0.8021863612701717,
+      "grad_norm": 0.27824823196428977,
+      "learning_rate": 4.2968305932513866e-05,
+      "loss": 0.1622,
+      "step": 1541
+    },
+    {
+      "epoch": 0.8027069234773555,
+      "grad_norm": 0.2738961827219555,
+      "learning_rate": 4.295853427413535e-05,
+      "loss": 0.1601,
+      "step": 1542
+    },
+    {
+      "epoch": 0.8032274856845393,
+      "grad_norm": 0.2665241074273365,
+      "learning_rate": 4.2948756943615985e-05,
+      "loss": 0.1534,
+      "step": 1543
+    },
+    {
+      "epoch": 0.803748047891723,
+      "grad_norm": 0.27182571850263426,
+      "learning_rate": 4.293897394404392e-05,
+      "loss": 0.1664,
+      "step": 1544
+    },
+    {
+      "epoch": 0.8042686100989068,
+      "grad_norm": 0.2553846747993963,
+      "learning_rate": 4.292918527850907e-05,
+      "loss": 0.1543,
+      "step": 1545
+    },
+    {
+      "epoch": 0.8047891723060906,
+      "grad_norm": 0.25252091088164785,
+      "learning_rate": 4.291939095010316e-05,
+      "loss": 0.1569,
+      "step": 1546
+    },
+    {
+      "epoch": 0.8053097345132744,
+      "grad_norm": 0.3033445790518775,
+      "learning_rate": 4.290959096191969e-05,
+      "loss": 0.156,
+      "step": 1547
+    },
+    {
+      "epoch": 0.8058302967204581,
+      "grad_norm": 0.26780210520674946,
+      "learning_rate": 4.289978531705395e-05,
+      "loss": 0.1629,
+      "step": 1548
+    },
+    {
+      "epoch": 0.8063508589276418,
+      "grad_norm": 0.25340358566449683,
+      "learning_rate": 4.288997401860303e-05,
+      "loss": 0.1676,
+      "step": 1549
+    },
+    {
+      "epoch": 0.8068714211348256,
+      "grad_norm": 0.26032260808346724,
+      "learning_rate": 4.288015706966578e-05,
+      "loss": 0.1599,
+      "step": 1550
+    },
+    {
+      "epoch": 0.8073919833420093,
+      "grad_norm": 0.253837369944548,
+      "learning_rate": 4.287033447334286e-05,
+      "loss": 0.1584,
+      "step": 1551
+    },
+    {
+      "epoch": 0.8079125455491931,
+      "grad_norm": 0.26639544259069964,
+      "learning_rate": 4.2860506232736706e-05,
+      "loss": 0.165,
+      "step": 1552
+    },
+    {
+      "epoch": 0.8084331077563769,
+      "grad_norm": 0.2374985977664141,
+      "learning_rate": 4.2850672350951516e-05,
+      "loss": 0.1545,
+      "step": 1553
+    },
+    {
+      "epoch": 0.8089536699635607,
+      "grad_norm": 0.2528600110588083,
+      "learning_rate": 4.284083283109331e-05,
+      "loss": 0.1579,
+      "step": 1554
+    },
+    {
+      "epoch": 0.8094742321707444,
+      "grad_norm": 0.25899495284687546,
+      "learning_rate": 4.283098767626984e-05,
+      "loss": 0.1533,
+      "step": 1555
+    },
+    {
+      "epoch": 0.8099947943779282,
+      "grad_norm": 0.2596540439723637,
+      "learning_rate": 4.2821136889590696e-05,
+      "loss": 0.1623,
+      "step": 1556
+    },
+    {
+      "epoch": 0.8105153565851119,
+      "grad_norm": 0.2593899738580702,
+      "learning_rate": 4.281128047416719e-05,
+      "loss": 0.1622,
+      "step": 1557
+    },
+    {
+      "epoch": 0.8110359187922956,
+      "grad_norm": 0.27318043091362676,
+      "learning_rate": 4.280141843311244e-05,
+      "loss": 0.1614,
+      "step": 1558
+    },
+    {
+      "epoch": 0.8115564809994794,
+      "grad_norm": 0.2650371042158417,
+      "learning_rate": 4.279155076954135e-05,
+      "loss": 0.1551,
+      "step": 1559
+    },
+    {
+      "epoch": 0.8120770432066632,
+      "grad_norm": 0.2554530943228691,
+      "learning_rate": 4.2781677486570576e-05,
+      "loss": 0.1622,
+      "step": 1560
+    },
+    {
+      "epoch": 0.812597605413847,
+      "grad_norm": 0.2742519335456577,
+      "learning_rate": 4.277179858731857e-05,
+      "loss": 0.1619,
+      "step": 1561
+    },
+    {
+      "epoch": 0.8131181676210307,
+      "grad_norm": 0.2613073326722418,
+      "learning_rate": 4.276191407490553e-05,
+      "loss": 0.1603,
+      "step": 1562
+    },
+    {
+      "epoch": 0.8136387298282145,
+      "grad_norm": 0.26817158410006026,
+      "learning_rate": 4.2752023952453465e-05,
+      "loss": 0.1571,
+      "step": 1563
+    },
+    {
+      "epoch": 0.8141592920353983,
+      "grad_norm": 0.26632267357937267,
+      "learning_rate": 4.274212822308612e-05,
+      "loss": 0.1599,
+      "step": 1564
+    },
+    {
+      "epoch": 0.8146798542425819,
+      "grad_norm": 0.2596167990888178,
+      "learning_rate": 4.273222688992904e-05,
+      "loss": 0.1588,
+      "step": 1565
+    },
+    {
+      "epoch": 0.8152004164497657,
+      "grad_norm": 0.27568862315736015,
+      "learning_rate": 4.272231995610952e-05,
+      "loss": 0.1542,
+      "step": 1566
+    },
+    {
+      "epoch": 0.8157209786569495,
+      "grad_norm": 0.27114645918535957,
+      "learning_rate": 4.271240742475664e-05,
+      "loss": 0.1576,
+      "step": 1567
+    },
+    {
+      "epoch": 0.8162415408641333,
+      "grad_norm": 0.24591250210741347,
+      "learning_rate": 4.2702489299001224e-05,
+      "loss": 0.1542,
+      "step": 1568
+    },
+    {
+      "epoch": 0.816762103071317,
+      "grad_norm": 0.25998302637843507,
+      "learning_rate": 4.269256558197588e-05,
+      "loss": 0.1597,
+      "step": 1569
+    },
+    {
+      "epoch": 0.8172826652785008,
+      "grad_norm": 0.277349113622191,
+      "learning_rate": 4.2682636276815e-05,
+      "loss": 0.1656,
+      "step": 1570
+    },
+    {
+      "epoch": 0.8178032274856846,
+      "grad_norm": 0.2591716268174999,
+      "learning_rate": 4.267270138665469e-05,
+      "loss": 0.1636,
+      "step": 1571
+    },
+    {
+      "epoch": 0.8183237896928683,
+      "grad_norm": 0.2626931448102875,
+      "learning_rate": 4.266276091463286e-05,
+      "loss": 0.1582,
+      "step": 1572
+    },
+    {
+      "epoch": 0.818844351900052,
+      "grad_norm": 0.2533409992404636,
+      "learning_rate": 4.26528148638892e-05,
+      "loss": 0.1559,
+      "step": 1573
+    },
+    {
+      "epoch": 0.8193649141072358,
+      "grad_norm": 0.2635890206268693,
+      "learning_rate": 4.26428632375651e-05,
+      "loss": 0.1587,
+      "step": 1574
+    },
+    {
+      "epoch": 0.8198854763144195,
+      "grad_norm": 0.2781221733340093,
+      "learning_rate": 4.2632906038803765e-05,
+      "loss": 0.1661,
+      "step": 1575
+    },
+    {
+      "epoch": 0.8204060385216033,
+      "grad_norm": 0.2466422831727149,
+      "learning_rate": 4.262294327075014e-05,
+      "loss": 0.1609,
+      "step": 1576
+    },
+    {
+      "epoch": 0.8209266007287871,
+      "grad_norm": 0.26214772989894536,
+      "learning_rate": 4.261297493655092e-05,
+      "loss": 0.156,
+      "step": 1577
+    },
+    {
+      "epoch": 0.8214471629359709,
+      "grad_norm": 0.25723469064882787,
+      "learning_rate": 4.260300103935459e-05,
+      "loss": 0.1562,
+      "step": 1578
+    },
+    {
+      "epoch": 0.8219677251431546,
+      "grad_norm": 0.2597832572362618,
+      "learning_rate": 4.2593021582311354e-05,
+      "loss": 0.163,
+      "step": 1579
+    },
+    {
+      "epoch": 0.8224882873503384,
+      "grad_norm": 0.25814069220724356,
+      "learning_rate": 4.2583036568573184e-05,
+      "loss": 0.1583,
+      "step": 1580
+    },
+    {
+      "epoch": 0.8230088495575221,
+      "grad_norm": 0.2680515018520899,
+      "learning_rate": 4.257304600129384e-05,
+      "loss": 0.1544,
+      "step": 1581
+    },
+    {
+      "epoch": 0.8235294117647058,
+      "grad_norm": 0.26943464813073753,
+      "learning_rate": 4.256304988362878e-05,
+      "loss": 0.1579,
+      "step": 1582
+    },
+    {
+      "epoch": 0.8240499739718896,
+      "grad_norm": 0.2536098084394136,
+      "learning_rate": 4.2553048218735256e-05,
+      "loss": 0.155,
+      "step": 1583
+    },
+    {
+      "epoch": 0.8245705361790734,
+      "grad_norm": 0.2757591782017878,
+      "learning_rate": 4.254304100977225e-05,
+      "loss": 0.1606,
+      "step": 1584
+    },
+    {
+      "epoch": 0.8250910983862572,
+      "grad_norm": 0.24717329055472512,
+      "learning_rate": 4.253302825990051e-05,
+      "loss": 0.1456,
+      "step": 1585
+    },
+    {
+      "epoch": 0.8256116605934409,
+      "grad_norm": 0.27846929929536657,
+      "learning_rate": 4.2523009972282534e-05,
+      "loss": 0.1619,
+      "step": 1586
+    },
+    {
+      "epoch": 0.8261322228006247,
+      "grad_norm": 0.2456486639173495,
+      "learning_rate": 4.2512986150082555e-05,
+      "loss": 0.163,
+      "step": 1587
+    },
+    {
+      "epoch": 0.8266527850078085,
+      "grad_norm": 0.28099219066923725,
+      "learning_rate": 4.250295679646657e-05,
+      "loss": 0.1581,
+      "step": 1588
+    },
+    {
+      "epoch": 0.8271733472149921,
+      "grad_norm": 0.28012381846944834,
+      "learning_rate": 4.24929219146023e-05,
+      "loss": 0.1618,
+      "step": 1589
+    },
+    {
+      "epoch": 0.8276939094221759,
+      "grad_norm": 0.2514914295217897,
+      "learning_rate": 4.248288150765925e-05,
+      "loss": 0.1565,
+      "step": 1590
+    },
+    {
+      "epoch": 0.8282144716293597,
+      "grad_norm": 0.2716012154641059,
+      "learning_rate": 4.2472835578808635e-05,
+      "loss": 0.1589,
+      "step": 1591
+    },
+    {
+      "epoch": 0.8287350338365435,
+      "grad_norm": 0.2815268850024281,
+      "learning_rate": 4.2462784131223434e-05,
+      "loss": 0.15,
+      "step": 1592
+    },
+    {
+      "epoch": 0.8292555960437272,
+      "grad_norm": 0.26832812464875594,
+      "learning_rate": 4.245272716807834e-05,
+      "loss": 0.1516,
+      "step": 1593
+    },
+    {
+      "epoch": 0.829776158250911,
+      "grad_norm": 0.255986260897498,
+      "learning_rate": 4.244266469254984e-05,
+      "loss": 0.1557,
+      "step": 1594
+    },
+    {
+      "epoch": 0.8302967204580948,
+      "grad_norm": 0.27756007774738645,
+      "learning_rate": 4.243259670781611e-05,
+      "loss": 0.1582,
+      "step": 1595
+    },
+    {
+      "epoch": 0.8308172826652785,
+      "grad_norm": 0.2788845280857182,
+      "learning_rate": 4.2422523217057104e-05,
+      "loss": 0.1611,
+      "step": 1596
+    },
+    {
+      "epoch": 0.8313378448724622,
+      "grad_norm": 0.265204765669698,
+      "learning_rate": 4.241244422345448e-05,
+      "loss": 0.1654,
+      "step": 1597
+    },
+    {
+      "epoch": 0.831858407079646,
+      "grad_norm": 0.29862671442559635,
+      "learning_rate": 4.240235973019168e-05,
+      "loss": 0.1628,
+      "step": 1598
+    },
+    {
+      "epoch": 0.8323789692868298,
+      "grad_norm": 0.25068737790327783,
+      "learning_rate": 4.239226974045383e-05,
+      "loss": 0.1612,
+      "step": 1599
+    },
+    {
+      "epoch": 0.8328995314940135,
+      "grad_norm": 0.26109969950747847,
+      "learning_rate": 4.2382174257427845e-05,
+      "loss": 0.1526,
+      "step": 1600
+    },
+    {
+      "epoch": 0.8334200937011973,
+      "grad_norm": 0.2708108326367472,
+      "learning_rate": 4.237207328430232e-05,
+      "loss": 0.1574,
+      "step": 1601
+    },
+    {
+      "epoch": 0.8339406559083811,
+      "grad_norm": 0.2603788294937712,
+      "learning_rate": 4.236196682426762e-05,
+      "loss": 0.1603,
+      "step": 1602
+    },
+    {
+      "epoch": 0.8344612181155648,
+      "grad_norm": 0.28645615014842474,
+      "learning_rate": 4.235185488051585e-05,
+      "loss": 0.164,
+      "step": 1603
+    },
+    {
+      "epoch": 0.8349817803227486,
+      "grad_norm": 0.25263310940554584,
+      "learning_rate": 4.2341737456240815e-05,
+      "loss": 0.152,
+      "step": 1604
+    },
+    {
+      "epoch": 0.8355023425299323,
+      "grad_norm": 0.29024837078909443,
+      "learning_rate": 4.233161455463809e-05,
+      "loss": 0.1584,
+      "step": 1605
+    },
+    {
+      "epoch": 0.836022904737116,
+      "grad_norm": 0.2698645950290543,
+      "learning_rate": 4.232148617890493e-05,
+      "loss": 0.1603,
+      "step": 1606
+    },
+    {
+      "epoch": 0.8365434669442998,
+      "grad_norm": 0.28134226221542846,
+      "learning_rate": 4.231135233224037e-05,
+      "loss": 0.1492,
+      "step": 1607
+    },
+    {
+      "epoch": 0.8370640291514836,
+      "grad_norm": 0.2436518446483368,
+      "learning_rate": 4.2301213017845144e-05,
+      "loss": 0.1543,
+      "step": 1608
+    },
+    {
+      "epoch": 0.8375845913586674,
+      "grad_norm": 0.28214111662149993,
+      "learning_rate": 4.2291068238921714e-05,
+      "loss": 0.1689,
+      "step": 1609
+    },
+    {
+      "epoch": 0.8381051535658511,
+      "grad_norm": 0.2863837316615571,
+      "learning_rate": 4.228091799867427e-05,
+      "loss": 0.1578,
+      "step": 1610
+    },
+    {
+      "epoch": 0.8386257157730349,
+      "grad_norm": 0.28091970790308574,
+      "learning_rate": 4.227076230030875e-05,
+      "loss": 0.1577,
+      "step": 1611
+    },
+    {
+      "epoch": 0.8391462779802187,
+      "grad_norm": 0.2593788595939422,
+      "learning_rate": 4.226060114703278e-05,
+      "loss": 0.1591,
+      "step": 1612
+    },
+    {
+      "epoch": 0.8396668401874023,
+      "grad_norm": 0.270786976366135,
+      "learning_rate": 4.225043454205573e-05,
+      "loss": 0.1646,
+      "step": 1613
+    },
+    {
+      "epoch": 0.8401874023945861,
+      "grad_norm": 0.2689847000833549,
+      "learning_rate": 4.224026248858868e-05,
+      "loss": 0.1574,
+      "step": 1614
+    },
+    {
+      "epoch": 0.8407079646017699,
+      "grad_norm": 0.2577066953840738,
+      "learning_rate": 4.2230084989844454e-05,
+      "loss": 0.155,
+      "step": 1615
+    },
+    {
+      "epoch": 0.8412285268089537,
+      "grad_norm": 0.2550154489290601,
+      "learning_rate": 4.221990204903756e-05,
+      "loss": 0.1564,
+      "step": 1616
+    },
+    {
+      "epoch": 0.8417490890161374,
+      "grad_norm": 0.273560448413933,
+      "learning_rate": 4.220971366938425e-05,
+      "loss": 0.1578,
+      "step": 1617
+    },
+    {
+      "epoch": 0.8422696512233212,
+      "grad_norm": 0.24371532149607608,
+      "learning_rate": 4.21995198541025e-05,
+      "loss": 0.1531,
+      "step": 1618
+    },
+    {
+      "epoch": 0.842790213430505,
+      "grad_norm": 0.28921374195808397,
+      "learning_rate": 4.218932060641198e-05,
+      "loss": 0.1561,
+      "step": 1619
+    },
+    {
+      "epoch": 0.8433107756376887,
+      "grad_norm": 0.2829460232916761,
+      "learning_rate": 4.217911592953409e-05,
+      "loss": 0.1621,
+      "step": 1620
+    },
+    {
+      "epoch": 0.8438313378448724,
+      "grad_norm": 0.26061758786155376,
+      "learning_rate": 4.216890582669194e-05,
+      "loss": 0.1484,
+      "step": 1621
+    },
+    {
+      "epoch": 0.8443519000520562,
+      "grad_norm": 0.2619069964371061,
+      "learning_rate": 4.2158690301110366e-05,
+      "loss": 0.1622,
+      "step": 1622
+    },
+    {
+      "epoch": 0.84487246225924,
+      "grad_norm": 0.27728638024745567,
+      "learning_rate": 4.2148469356015896e-05,
+      "loss": 0.1573,
+      "step": 1623
+    },
+    {
+      "epoch": 0.8453930244664237,
+      "grad_norm": 0.2660951427495475,
+      "learning_rate": 4.213824299463678e-05,
+      "loss": 0.1653,
+      "step": 1624
+    },
+    {
+      "epoch": 0.8459135866736075,
+      "grad_norm": 0.2617492948395115,
+      "learning_rate": 4.2128011220202976e-05,
+      "loss": 0.1633,
+      "step": 1625
+    },
+    {
+      "epoch": 0.8464341488807913,
+      "grad_norm": 0.29138627520898736,
+      "learning_rate": 4.211777403594617e-05,
+      "loss": 0.1596,
+      "step": 1626
+    },
+    {
+      "epoch": 0.846954711087975,
+      "grad_norm": 0.24888350370609963,
+      "learning_rate": 4.210753144509972e-05,
+      "loss": 0.1606,
+      "step": 1627
+    },
+    {
+      "epoch": 0.8474752732951588,
+      "grad_norm": 0.2505367676114268,
+      "learning_rate": 4.209728345089873e-05,
+      "loss": 0.1629,
+      "step": 1628
+    },
+    {
+      "epoch": 0.8479958355023425,
+      "grad_norm": 0.25895259510640567,
+      "learning_rate": 4.208703005657999e-05,
+      "loss": 0.1591,
+      "step": 1629
+    },
+    {
+      "epoch": 0.8485163977095262,
+      "grad_norm": 0.25678447658914816,
+      "learning_rate": 4.207677126538199e-05,
+      "loss": 0.1606,
+      "step": 1630
+    },
+    {
+      "epoch": 0.84903695991671,
+      "grad_norm": 0.26827714044896583,
+      "learning_rate": 4.206650708054494e-05,
+      "loss": 0.1457,
+      "step": 1631
+    },
+    {
+      "epoch": 0.8495575221238938,
+      "grad_norm": 0.2722481858513276,
+      "learning_rate": 4.205623750531076e-05,
+      "loss": 0.1575,
+      "step": 1632
+    },
+    {
+      "epoch": 0.8500780843310776,
+      "grad_norm": 0.27881803489843254,
+      "learning_rate": 4.204596254292303e-05,
+      "loss": 0.1581,
+      "step": 1633
+    },
+    {
+      "epoch": 0.8505986465382613,
+      "grad_norm": 0.2519593987498123,
+      "learning_rate": 4.203568219662709e-05,
+      "loss": 0.1565,
+      "step": 1634
+    },
+    {
+      "epoch": 0.8511192087454451,
+      "grad_norm": 0.26920318449984887,
+      "learning_rate": 4.202539646966993e-05,
+      "loss": 0.1525,
+      "step": 1635
+    },
+    {
+      "epoch": 0.8516397709526289,
+      "grad_norm": 0.26653026491152526,
+      "learning_rate": 4.2015105365300276e-05,
+      "loss": 0.1609,
+      "step": 1636
+    },
+    {
+      "epoch": 0.8521603331598125,
+      "grad_norm": 0.2827093407759518,
+      "learning_rate": 4.200480888676853e-05,
+      "loss": 0.1611,
+      "step": 1637
+    },
+    {
+      "epoch": 0.8526808953669963,
+      "grad_norm": 0.2652398385704472,
+      "learning_rate": 4.199450703732681e-05,
+      "loss": 0.1513,
+      "step": 1638
+    },
+    {
+      "epoch": 0.8532014575741801,
+      "grad_norm": 0.2696612536923782,
+      "learning_rate": 4.19841998202289e-05,
+      "loss": 0.1566,
+      "step": 1639
+    },
+    {
+      "epoch": 0.8537220197813639,
+      "grad_norm": 0.28633933577187565,
+      "learning_rate": 4.197388723873032e-05,
+      "loss": 0.1578,
+      "step": 1640
+    },
+    {
+      "epoch": 0.8542425819885476,
+      "grad_norm": 0.2690705270209936,
+      "learning_rate": 4.196356929608825e-05,
+      "loss": 0.1558,
+      "step": 1641
+    },
+    {
+      "epoch": 0.8547631441957314,
+      "grad_norm": 0.24483032302626148,
+      "learning_rate": 4.195324599556158e-05,
+      "loss": 0.1494,
+      "step": 1642
+    },
+    {
+      "epoch": 0.8552837064029152,
+      "grad_norm": 0.25311938346196466,
+      "learning_rate": 4.194291734041089e-05,
+      "loss": 0.1585,
+      "step": 1643
+    },
+    {
+      "epoch": 0.855804268610099,
+      "grad_norm": 0.2687437228508878,
+      "learning_rate": 4.193258333389844e-05,
+      "loss": 0.1585,
+      "step": 1644
+    },
+    {
+      "epoch": 0.8563248308172826,
+      "grad_norm": 0.23790698337821162,
+      "learning_rate": 4.1922243979288205e-05,
+      "loss": 0.1591,
+      "step": 1645
+    },
+    {
+      "epoch": 0.8568453930244664,
+      "grad_norm": 0.2720940365664535,
+      "learning_rate": 4.191189927984583e-05,
+      "loss": 0.1591,
+      "step": 1646
+    },
+    {
+      "epoch": 0.8573659552316502,
+      "grad_norm": 0.23733908382168187,
+      "learning_rate": 4.190154923883865e-05,
+      "loss": 0.1497,
+      "step": 1647
+    },
+    {
+      "epoch": 0.8578865174388339,
+      "grad_norm": 0.2767062941528966,
+      "learning_rate": 4.1891193859535686e-05,
+      "loss": 0.1606,
+      "step": 1648
+    },
+    {
+      "epoch": 0.8584070796460177,
+      "grad_norm": 0.23093654286620274,
+      "learning_rate": 4.1880833145207655e-05,
+      "loss": 0.15,
+      "step": 1649
+    },
+    {
+      "epoch": 0.8589276418532015,
+      "grad_norm": 0.24409975896473038,
+      "learning_rate": 4.187046709912695e-05,
+      "loss": 0.1608,
+      "step": 1650
+    },
+    {
+      "epoch": 0.8594482040603852,
+      "grad_norm": 0.2641900349527886,
+      "learning_rate": 4.186009572456765e-05,
+      "loss": 0.1583,
+      "step": 1651
+    },
+    {
+      "epoch": 0.859968766267569,
+      "grad_norm": 0.25565972911270796,
+      "learning_rate": 4.184971902480552e-05,
+      "loss": 0.1543,
+      "step": 1652
+    },
+    {
+      "epoch": 0.8604893284747527,
+      "grad_norm": 0.26824705481294847,
+      "learning_rate": 4.183933700311801e-05,
+      "loss": 0.1673,
+      "step": 1653
+    },
+    {
+      "epoch": 0.8610098906819365,
+      "grad_norm": 0.2430609538785938,
+      "learning_rate": 4.1828949662784236e-05,
+      "loss": 0.1462,
+      "step": 1654
+    },
+    {
+      "epoch": 0.8615304528891202,
+      "grad_norm": 0.2846978931791095,
+      "learning_rate": 4.1818557007085e-05,
+      "loss": 0.163,
+      "step": 1655
+    },
+    {
+      "epoch": 0.862051015096304,
+      "grad_norm": 0.25943378302982295,
+      "learning_rate": 4.1808159039302795e-05,
+      "loss": 0.1561,
+      "step": 1656
+    },
+    {
+      "epoch": 0.8625715773034878,
+      "grad_norm": 0.26198173198500524,
+      "learning_rate": 4.1797755762721787e-05,
+      "loss": 0.1552,
+      "step": 1657
+    },
+    {
+      "epoch": 0.8630921395106715,
+      "grad_norm": 0.24923720021291665,
+      "learning_rate": 4.17873471806278e-05,
+      "loss": 0.1586,
+      "step": 1658
+    },
+    {
+      "epoch": 0.8636127017178553,
+      "grad_norm": 0.2974201345925281,
+      "learning_rate": 4.177693329630837e-05,
+      "loss": 0.1578,
+      "step": 1659
+    },
+    {
+      "epoch": 0.8641332639250391,
+      "grad_norm": 0.24813334349056498,
+      "learning_rate": 4.176651411305266e-05,
+      "loss": 0.1585,
+      "step": 1660
+    },
+    {
+      "epoch": 0.8646538261322227,
+      "grad_norm": 0.2688434182973993,
+      "learning_rate": 4.175608963415155e-05,
+      "loss": 0.1627,
+      "step": 1661
+    },
+    {
+      "epoch": 0.8651743883394065,
+      "grad_norm": 0.2525026939637232,
+      "learning_rate": 4.174565986289758e-05,
+      "loss": 0.1588,
+      "step": 1662
+    },
+    {
+      "epoch": 0.8656949505465903,
+      "grad_norm": 0.2734677290901909,
+      "learning_rate": 4.1735224802584946e-05,
+      "loss": 0.1566,
+      "step": 1663
+    },
+    {
+      "epoch": 0.8662155127537741,
+      "grad_norm": 0.24890893238886064,
+      "learning_rate": 4.172478445650953e-05,
+      "loss": 0.1605,
+      "step": 1664
+    },
+    {
+      "epoch": 0.8667360749609578,
+      "grad_norm": 0.2705875269181692,
+      "learning_rate": 4.171433882796888e-05,
+      "loss": 0.1629,
+      "step": 1665
+    },
+    {
+      "epoch": 0.8672566371681416,
+      "grad_norm": 0.2576695174793914,
+      "learning_rate": 4.1703887920262195e-05,
+      "loss": 0.153,
+      "step": 1666
+    },
+    {
+      "epoch": 0.8677771993753254,
+      "grad_norm": 0.2566955998695762,
+      "learning_rate": 4.1693431736690386e-05,
+      "loss": 0.1538,
+      "step": 1667
+    },
+    {
+      "epoch": 0.8682977615825092,
+      "grad_norm": 0.2795710749963426,
+      "learning_rate": 4.1682970280555986e-05,
+      "loss": 0.1593,
+      "step": 1668
+    },
+    {
+      "epoch": 0.8688183237896928,
+      "grad_norm": 0.2565679446639193,
+      "learning_rate": 4.1672503555163215e-05,
+      "loss": 0.167,
+      "step": 1669
+    },
+    {
+      "epoch": 0.8693388859968766,
+      "grad_norm": 0.26408964383887445,
+      "learning_rate": 4.166203156381795e-05,
+      "loss": 0.1518,
+      "step": 1670
+    },
+    {
+      "epoch": 0.8698594482040604,
+      "grad_norm": 0.24825446479368055,
+      "learning_rate": 4.1651554309827725e-05,
+      "loss": 0.1589,
+      "step": 1671
+    },
+    {
+      "epoch": 0.8703800104112441,
+      "grad_norm": 0.2674815525271392,
+      "learning_rate": 4.1641071796501764e-05,
+      "loss": 0.1579,
+      "step": 1672
+    },
+    {
+      "epoch": 0.8709005726184279,
+      "grad_norm": 0.26322901088276796,
+      "learning_rate": 4.163058402715091e-05,
+      "loss": 0.1587,
+      "step": 1673
+    },
+    {
+      "epoch": 0.8714211348256117,
+      "grad_norm": 0.27350744005632355,
+      "learning_rate": 4.1620091005087714e-05,
+      "loss": 0.165,
+      "step": 1674
+    },
+    {
+      "epoch": 0.8719416970327954,
+      "grad_norm": 0.263352010398878,
+      "learning_rate": 4.1609592733626335e-05,
+      "loss": 0.155,
+      "step": 1675
+    },
+    {
+      "epoch": 0.8724622592399792,
+      "grad_norm": 0.2693710231228204,
+      "learning_rate": 4.159908921608263e-05,
+      "loss": 0.1589,
+      "step": 1676
+    },
+    {
+      "epoch": 0.8729828214471629,
+      "grad_norm": 0.2846789455817984,
+      "learning_rate": 4.158858045577409e-05,
+      "loss": 0.1557,
+      "step": 1677
+    },
+    {
+      "epoch": 0.8735033836543467,
+      "grad_norm": 0.25566156183607636,
+      "learning_rate": 4.157806645601988e-05,
+      "loss": 0.1645,
+      "step": 1678
+    },
+    {
+      "epoch": 0.8740239458615304,
+      "grad_norm": 0.2817627765773898,
+      "learning_rate": 4.1567547220140814e-05,
+      "loss": 0.1601,
+      "step": 1679
+    },
+    {
+      "epoch": 0.8745445080687142,
+      "grad_norm": 0.2505616741147018,
+      "learning_rate": 4.155702275145934e-05,
+      "loss": 0.1547,
+      "step": 1680
+    },
+    {
+      "epoch": 0.875065070275898,
+      "grad_norm": 0.2553052002789369,
+      "learning_rate": 4.154649305329958e-05,
+      "loss": 0.1591,
+      "step": 1681
+    },
+    {
+      "epoch": 0.8755856324830817,
+      "grad_norm": 0.25048785354273917,
+      "learning_rate": 4.153595812898732e-05,
+      "loss": 0.1597,
+      "step": 1682
+    },
+    {
+      "epoch": 0.8761061946902655,
+      "grad_norm": 0.25427482435094684,
+      "learning_rate": 4.152541798184995e-05,
+      "loss": 0.161,
+      "step": 1683
+    },
+    {
+      "epoch": 0.8766267568974493,
+      "grad_norm": 0.24530895241431033,
+      "learning_rate": 4.151487261521656e-05,
+      "loss": 0.1581,
+      "step": 1684
+    },
+    {
+      "epoch": 0.877147319104633,
+      "grad_norm": 0.25440752591956056,
+      "learning_rate": 4.1504322032417864e-05,
+      "loss": 0.1638,
+      "step": 1685
+    },
+    {
+      "epoch": 0.8776678813118167,
+      "grad_norm": 0.2587793514600556,
+      "learning_rate": 4.149376623678623e-05,
+      "loss": 0.1587,
+      "step": 1686
+    },
+    {
+      "epoch": 0.8781884435190005,
+      "grad_norm": 0.24239699887208244,
+      "learning_rate": 4.148320523165566e-05,
+      "loss": 0.1514,
+      "step": 1687
+    },
+    {
+      "epoch": 0.8787090057261843,
+      "grad_norm": 0.24553152349960053,
+      "learning_rate": 4.147263902036181e-05,
+      "loss": 0.1538,
+      "step": 1688
+    },
+    {
+      "epoch": 0.879229567933368,
+      "grad_norm": 0.2499331408281633,
+      "learning_rate": 4.146206760624199e-05,
+      "loss": 0.1603,
+      "step": 1689
+    },
+    {
+      "epoch": 0.8797501301405518,
+      "grad_norm": 0.24703228194514043,
+      "learning_rate": 4.145149099263515e-05,
+      "loss": 0.1577,
+      "step": 1690
+    },
+    {
+      "epoch": 0.8802706923477356,
+      "grad_norm": 0.2723118994778378,
+      "learning_rate": 4.1440909182881857e-05,
+      "loss": 0.1587,
+      "step": 1691
+    },
+    {
+      "epoch": 0.8807912545549194,
+      "grad_norm": 0.2754696773958591,
+      "learning_rate": 4.143032218032435e-05,
+      "loss": 0.1562,
+      "step": 1692
+    },
+    {
+      "epoch": 0.881311816762103,
+      "grad_norm": 0.2617623912854439,
+      "learning_rate": 4.141972998830651e-05,
+      "loss": 0.1585,
+      "step": 1693
+    },
+    {
+      "epoch": 0.8818323789692868,
+      "grad_norm": 0.30890725644539296,
+      "learning_rate": 4.140913261017382e-05,
+      "loss": 0.1624,
+      "step": 1694
+    },
+    {
+      "epoch": 0.8823529411764706,
+      "grad_norm": 0.24884174691256375,
+      "learning_rate": 4.139853004927344e-05,
+      "loss": 0.1559,
+      "step": 1695
+    },
+    {
+      "epoch": 0.8828735033836543,
+      "grad_norm": 0.27439772981131394,
+      "learning_rate": 4.1387922308954154e-05,
+      "loss": 0.1562,
+      "step": 1696
+    },
+    {
+      "epoch": 0.8833940655908381,
+      "grad_norm": 0.2542935955914035,
+      "learning_rate": 4.137730939256636e-05,
+      "loss": 0.1619,
+      "step": 1697
+    },
+    {
+      "epoch": 0.8839146277980219,
+      "grad_norm": 0.28136889554044375,
+      "learning_rate": 4.1366691303462144e-05,
+      "loss": 0.1625,
+      "step": 1698
+    },
+    {
+      "epoch": 0.8844351900052057,
+      "grad_norm": 0.247375024118648,
+      "learning_rate": 4.135606804499516e-05,
+      "loss": 0.1599,
+      "step": 1699
+    },
+    {
+      "epoch": 0.8849557522123894,
+      "grad_norm": 0.2608875547949653,
+      "learning_rate": 4.1345439620520744e-05,
+      "loss": 0.1639,
+      "step": 1700
+    },
+    {
+      "epoch": 0.8854763144195731,
+      "grad_norm": 0.2556809971776182,
+      "learning_rate": 4.1334806033395845e-05,
+      "loss": 0.1596,
+      "step": 1701
+    },
+    {
+      "epoch": 0.8859968766267569,
+      "grad_norm": 0.25036073808421017,
+      "learning_rate": 4.132416728697905e-05,
+      "loss": 0.1567,
+      "step": 1702
+    },
+    {
+      "epoch": 0.8865174388339406,
+      "grad_norm": 0.24789643869487482,
+      "learning_rate": 4.131352338463056e-05,
+      "loss": 0.1551,
+      "step": 1703
+    },
+    {
+      "epoch": 0.8870380010411244,
+      "grad_norm": 0.2597436931442242,
+      "learning_rate": 4.130287432971222e-05,
+      "loss": 0.1635,
+      "step": 1704
+    },
+    {
+      "epoch": 0.8875585632483082,
+      "grad_norm": 0.25176527895305045,
+      "learning_rate": 4.1292220125587494e-05,
+      "loss": 0.1529,
+      "step": 1705
+    },
+    {
+      "epoch": 0.888079125455492,
+      "grad_norm": 0.22936720495484114,
+      "learning_rate": 4.1281560775621475e-05,
+      "loss": 0.1536,
+      "step": 1706
+    },
+    {
+      "epoch": 0.8885996876626757,
+      "grad_norm": 0.29208711846524826,
+      "learning_rate": 4.1270896283180896e-05,
+      "loss": 0.1669,
+      "step": 1707
+    },
+    {
+      "epoch": 0.8891202498698595,
+      "grad_norm": 0.24185973845450207,
+      "learning_rate": 4.1260226651634074e-05,
+      "loss": 0.153,
+      "step": 1708
+    },
+    {
+      "epoch": 0.8896408120770432,
+      "grad_norm": 0.2405834957600981,
+      "learning_rate": 4.1249551884351e-05,
+      "loss": 0.1548,
+      "step": 1709
+    },
+    {
+      "epoch": 0.8901613742842269,
+      "grad_norm": 0.2640002798637536,
+      "learning_rate": 4.1238871984703255e-05,
+      "loss": 0.152,
+      "step": 1710
+    },
+    {
+      "epoch": 0.8906819364914107,
+      "grad_norm": 0.3051529300584874,
+      "learning_rate": 4.122818695606403e-05,
+      "loss": 0.161,
+      "step": 1711
+    },
+    {
+      "epoch": 0.8912024986985945,
+      "grad_norm": 0.2700106707452003,
+      "learning_rate": 4.121749680180818e-05,
+      "loss": 0.1564,
+      "step": 1712
+    },
+    {
+      "epoch": 0.8917230609057782,
+      "grad_norm": 0.26880881799117995,
+      "learning_rate": 4.1206801525312144e-05,
+      "loss": 0.1584,
+      "step": 1713
+    },
+    {
+      "epoch": 0.892243623112962,
+      "grad_norm": 0.2658890951024931,
+      "learning_rate": 4.119610112995398e-05,
+      "loss": 0.1567,
+      "step": 1714
+    },
+    {
+      "epoch": 0.8927641853201458,
+      "grad_norm": 0.2546814903680581,
+      "learning_rate": 4.118539561911339e-05,
+      "loss": 0.1554,
+      "step": 1715
+    },
+    {
+      "epoch": 0.8932847475273296,
+      "grad_norm": 0.24712755281104856,
+      "learning_rate": 4.1174684996171644e-05,
+      "loss": 0.1535,
+      "step": 1716
+    },
+    {
+      "epoch": 0.8938053097345132,
+      "grad_norm": 0.2806241667002411,
+      "learning_rate": 4.116396926451168e-05,
+      "loss": 0.1563,
+      "step": 1717
+    },
+    {
+      "epoch": 0.894325871941697,
+      "grad_norm": 0.27342193998178105,
+      "learning_rate": 4.115324842751802e-05,
+      "loss": 0.1592,
+      "step": 1718
+    },
+    {
+      "epoch": 0.8948464341488808,
+      "grad_norm": 0.23962746337276375,
+      "learning_rate": 4.114252248857679e-05,
+      "loss": 0.1472,
+      "step": 1719
+    },
+    {
+      "epoch": 0.8953669963560645,
+      "grad_norm": 0.26324287304353916,
+      "learning_rate": 4.1131791451075755e-05,
+      "loss": 0.159,
+      "step": 1720
+    },
+    {
+      "epoch": 0.8958875585632483,
+      "grad_norm": 0.26150443069954205,
+      "learning_rate": 4.1121055318404264e-05,
+      "loss": 0.1539,
+      "step": 1721
+    },
+    {
+      "epoch": 0.8964081207704321,
+      "grad_norm": 0.24755890951530543,
+      "learning_rate": 4.1110314093953305e-05,
+      "loss": 0.1582,
+      "step": 1722
+    },
+    {
+      "epoch": 0.8969286829776159,
+      "grad_norm": 0.27523785393031364,
+      "learning_rate": 4.109956778111544e-05,
+      "loss": 0.1629,
+      "step": 1723
+    },
+    {
+      "epoch": 0.8974492451847996,
+      "grad_norm": 0.2591994387571103,
+      "learning_rate": 4.108881638328486e-05,
+      "loss": 0.1478,
+      "step": 1724
+    },
+    {
+      "epoch": 0.8979698073919833,
+      "grad_norm": 0.29966605704015586,
+      "learning_rate": 4.1078059903857355e-05,
+      "loss": 0.1601,
+      "step": 1725
+    },
+    {
+      "epoch": 0.8984903695991671,
+      "grad_norm": 0.25390075851127014,
+      "learning_rate": 4.1067298346230335e-05,
+      "loss": 0.1501,
+      "step": 1726
+    },
+    {
+      "epoch": 0.8990109318063508,
+      "grad_norm": 0.2687980166473991,
+      "learning_rate": 4.105653171380278e-05,
+      "loss": 0.1498,
+      "step": 1727
+    },
+    {
+      "epoch": 0.8995314940135346,
+      "grad_norm": 0.28950396723952554,
+      "learning_rate": 4.10457600099753e-05,
+      "loss": 0.1549,
+      "step": 1728
+    },
+    {
+      "epoch": 0.9000520562207184,
+      "grad_norm": 0.25514115726981085,
+      "learning_rate": 4.103498323815011e-05,
+      "loss": 0.1647,
+      "step": 1729
+    },
+    {
+      "epoch": 0.9005726184279021,
+      "grad_norm": 0.2587811609001319,
+      "learning_rate": 4.1024201401731005e-05,
+      "loss": 0.1583,
+      "step": 1730
+    },
+    {
+      "epoch": 0.9010931806350859,
+      "grad_norm": 0.2604547864469705,
+      "learning_rate": 4.1013414504123396e-05,
+      "loss": 0.1561,
+      "step": 1731
+    },
+    {
+      "epoch": 0.9016137428422697,
+      "grad_norm": 0.24850657249226776,
+      "learning_rate": 4.1002622548734296e-05,
+      "loss": 0.1522,
+      "step": 1732
+    },
+    {
+      "epoch": 0.9021343050494534,
+      "grad_norm": 0.2625089988338353,
+      "learning_rate": 4.099182553897229e-05,
+      "loss": 0.1618,
+      "step": 1733
+    },
+    {
+      "epoch": 0.9026548672566371,
+      "grad_norm": 0.2632632791671295,
+      "learning_rate": 4.098102347824758e-05,
+      "loss": 0.1555,
+      "step": 1734
+    },
+    {
+      "epoch": 0.9031754294638209,
+      "grad_norm": 0.23666054780572104,
+      "learning_rate": 4.097021636997196e-05,
+      "loss": 0.1595,
+      "step": 1735
+    },
+    {
+      "epoch": 0.9036959916710047,
+      "grad_norm": 0.2591388444201557,
+      "learning_rate": 4.095940421755883e-05,
+      "loss": 0.1656,
+      "step": 1736
+    },
+    {
+      "epoch": 0.9042165538781884,
+      "grad_norm": 0.2591712441056336,
+      "learning_rate": 4.094858702442316e-05,
+      "loss": 0.1526,
+      "step": 1737
+    },
+    {
+      "epoch": 0.9047371160853722,
+      "grad_norm": 0.24844682559914494,
+      "learning_rate": 4.093776479398151e-05,
+      "loss": 0.157,
+      "step": 1738
+    },
+    {
+      "epoch": 0.905257678292556,
+      "grad_norm": 0.2559696276702533,
+      "learning_rate": 4.092693752965208e-05,
+      "loss": 0.155,
+      "step": 1739
+    },
+    {
+      "epoch": 0.9057782404997398,
+      "grad_norm": 0.2568180726661143,
+      "learning_rate": 4.091610523485458e-05,
+      "loss": 0.1567,
+      "step": 1740
+    },
+    {
+      "epoch": 0.9062988027069234,
+      "grad_norm": 0.27176199980897775,
+      "learning_rate": 4.09052679130104e-05,
+      "loss": 0.1473,
+      "step": 1741
+    },
+    {
+      "epoch": 0.9068193649141072,
+      "grad_norm": 0.26678573538868655,
+      "learning_rate": 4.089442556754243e-05,
+      "loss": 0.1591,
+      "step": 1742
+    },
+    {
+      "epoch": 0.907339927121291,
+      "grad_norm": 0.2568264736060921,
+      "learning_rate": 4.088357820187521e-05,
+      "loss": 0.1624,
+      "step": 1743
+    },
+    {
+      "epoch": 0.9078604893284747,
+      "grad_norm": 0.2551009665433581,
+      "learning_rate": 4.087272581943483e-05,
+      "loss": 0.164,
+      "step": 1744
+    },
+    {
+      "epoch": 0.9083810515356585,
+      "grad_norm": 0.2596099419866776,
+      "learning_rate": 4.0861868423648985e-05,
+      "loss": 0.1593,
+      "step": 1745
+    },
+    {
+      "epoch": 0.9089016137428423,
+      "grad_norm": 0.25573551965558855,
+      "learning_rate": 4.085100601794695e-05,
+      "loss": 0.1598,
+      "step": 1746
+    },
+    {
+      "epoch": 0.9094221759500261,
+      "grad_norm": 0.25024824627264436,
+      "learning_rate": 4.084013860575956e-05,
+      "loss": 0.1499,
+      "step": 1747
+    },
+    {
+      "epoch": 0.9099427381572098,
+      "grad_norm": 0.24563669418483275,
+      "learning_rate": 4.0829266190519264e-05,
+      "loss": 0.1533,
+      "step": 1748
+    },
+    {
+      "epoch": 0.9104633003643935,
+      "grad_norm": 0.25857714800578796,
+      "learning_rate": 4.0818388775660083e-05,
+      "loss": 0.1616,
+      "step": 1749
+    },
+    {
+      "epoch": 0.9109838625715773,
+      "grad_norm": 0.25995854352713227,
+      "learning_rate": 4.08075063646176e-05,
+      "loss": 0.1607,
+      "step": 1750
+    },
+    {
+      "epoch": 0.911504424778761,
+      "grad_norm": 0.24199870728320377,
+      "learning_rate": 4.079661896082899e-05,
+      "loss": 0.155,
+      "step": 1751
+    },
+    {
+      "epoch": 0.9120249869859448,
+      "grad_norm": 0.2585528070362897,
+      "learning_rate": 4.0785726567733e-05,
+      "loss": 0.1695,
+      "step": 1752
+    },
+    {
+      "epoch": 0.9125455491931286,
+      "grad_norm": 0.2535260703066128,
+      "learning_rate": 4.0774829188769946e-05,
+      "loss": 0.1553,
+      "step": 1753
+    },
+    {
+      "epoch": 0.9130661114003124,
+      "grad_norm": 0.24975570980491604,
+      "learning_rate": 4.076392682738175e-05,
+      "loss": 0.155,
+      "step": 1754
+    },
+    {
+      "epoch": 0.9135866736074961,
+      "grad_norm": 0.23729277290872783,
+      "learning_rate": 4.075301948701186e-05,
+      "loss": 0.1559,
+      "step": 1755
+    },
+    {
+      "epoch": 0.9141072358146799,
+      "grad_norm": 0.2576836288112198,
+      "learning_rate": 4.074210717110534e-05,
+      "loss": 0.1508,
+      "step": 1756
+    },
+    {
+      "epoch": 0.9146277980218636,
+      "grad_norm": 0.23905943214710218,
+      "learning_rate": 4.07311898831088e-05,
+      "loss": 0.1595,
+      "step": 1757
+    },
+    {
+      "epoch": 0.9151483602290473,
+      "grad_norm": 0.2623063356166442,
+      "learning_rate": 4.072026762647043e-05,
+      "loss": 0.155,
+      "step": 1758
+    },
+    {
+      "epoch": 0.9156689224362311,
+      "grad_norm": 0.25004028247757853,
+      "learning_rate": 4.070934040463998e-05,
+      "loss": 0.151,
+      "step": 1759
+    },
+    {
+      "epoch": 0.9161894846434149,
+      "grad_norm": 0.24742792524698254,
+      "learning_rate": 4.069840822106879e-05,
+      "loss": 0.158,
+      "step": 1760
+    },
+    {
+      "epoch": 0.9167100468505986,
+      "grad_norm": 0.2536956962460334,
+      "learning_rate": 4.068747107920974e-05,
+      "loss": 0.1525,
+      "step": 1761
+    },
+    {
+      "epoch": 0.9172306090577824,
+      "grad_norm": 0.2483073074101714,
+      "learning_rate": 4.067652898251729e-05,
+      "loss": 0.1516,
+      "step": 1762
+    },
+    {
+      "epoch": 0.9177511712649662,
+      "grad_norm": 0.24230992836435572,
+      "learning_rate": 4.066558193444746e-05,
+      "loss": 0.1521,
+      "step": 1763
+    },
+    {
+      "epoch": 0.91827173347215,
+      "grad_norm": 0.26776759518460846,
+      "learning_rate": 4.065462993845784e-05,
+      "loss": 0.1615,
+      "step": 1764
+    },
+    {
+      "epoch": 0.9187922956793336,
+      "grad_norm": 0.247701252468107,
+      "learning_rate": 4.0643672998007593e-05,
+      "loss": 0.156,
+      "step": 1765
+    },
+    {
+      "epoch": 0.9193128578865174,
+      "grad_norm": 0.2525494661335194,
+      "learning_rate": 4.063271111655741e-05,
+      "loss": 0.1544,
+      "step": 1766
+    },
+    {
+      "epoch": 0.9198334200937012,
+      "grad_norm": 0.24152047171275015,
+      "learning_rate": 4.062174429756958e-05,
+      "loss": 0.1582,
+      "step": 1767
+    },
+    {
+      "epoch": 0.9203539823008849,
+      "grad_norm": 0.24113773563979826,
+      "learning_rate": 4.0610772544507925e-05,
+      "loss": 0.157,
+      "step": 1768
+    },
+    {
+      "epoch": 0.9208745445080687,
+      "grad_norm": 0.24499303443994122,
+      "learning_rate": 4.059979586083783e-05,
+      "loss": 0.1589,
+      "step": 1769
+    },
+    {
+      "epoch": 0.9213951067152525,
+      "grad_norm": 0.2441852531096595,
+      "learning_rate": 4.0588814250026255e-05,
+      "loss": 0.1514,
+      "step": 1770
+    },
+    {
+      "epoch": 0.9219156689224363,
+      "grad_norm": 0.2618027069982354,
+      "learning_rate": 4.05778277155417e-05,
+      "loss": 0.1581,
+      "step": 1771
+    },
+    {
+      "epoch": 0.92243623112962,
+      "grad_norm": 0.2604906720770046,
+      "learning_rate": 4.056683626085422e-05,
+      "loss": 0.161,
+      "step": 1772
+    },
+    {
+      "epoch": 0.9229567933368037,
+      "grad_norm": 0.23267751863286024,
+      "learning_rate": 4.0555839889435446e-05,
+      "loss": 0.1482,
+      "step": 1773
+    },
+    {
+      "epoch": 0.9234773555439875,
+      "grad_norm": 0.25641954777935727,
+      "learning_rate": 4.054483860475851e-05,
+      "loss": 0.1541,
+      "step": 1774
+    },
+    {
+      "epoch": 0.9239979177511712,
+      "grad_norm": 0.2596515890918495,
+      "learning_rate": 4.053383241029815e-05,
+      "loss": 0.1583,
+      "step": 1775
+    },
+    {
+      "epoch": 0.924518479958355,
+      "grad_norm": 0.23481778190953714,
+      "learning_rate": 4.0522821309530635e-05,
+      "loss": 0.1548,
+      "step": 1776
+    },
+    {
+      "epoch": 0.9250390421655388,
+      "grad_norm": 0.24362441004243998,
+      "learning_rate": 4.051180530593379e-05,
+      "loss": 0.1584,
+      "step": 1777
+    },
+    {
+      "epoch": 0.9255596043727226,
+      "grad_norm": 0.2520010539214911,
+      "learning_rate": 4.0500784402986956e-05,
+      "loss": 0.1551,
+      "step": 1778
+    },
+    {
+      "epoch": 0.9260801665799063,
+      "grad_norm": 0.26110890320468977,
+      "learning_rate": 4.0489758604171076e-05,
+      "loss": 0.1531,
+      "step": 1779
+    },
+    {
+      "epoch": 0.9266007287870901,
+      "grad_norm": 0.27497633290869694,
+      "learning_rate": 4.047872791296859e-05,
+      "loss": 0.1618,
+      "step": 1780
+    },
+    {
+      "epoch": 0.9271212909942738,
+      "grad_norm": 0.26975908040013474,
+      "learning_rate": 4.0467692332863515e-05,
+      "loss": 0.1487,
+      "step": 1781
+    },
+    {
+      "epoch": 0.9276418532014575,
+      "grad_norm": 0.26862955487856527,
+      "learning_rate": 4.04566518673414e-05,
+      "loss": 0.1561,
+      "step": 1782
+    },
+    {
+      "epoch": 0.9281624154086413,
+      "grad_norm": 0.25511699277211003,
+      "learning_rate": 4.044560651988933e-05,
+      "loss": 0.1615,
+      "step": 1783
+    },
+    {
+      "epoch": 0.9286829776158251,
+      "grad_norm": 0.2667422054425098,
+      "learning_rate": 4.043455629399594e-05,
+      "loss": 0.1582,
+      "step": 1784
+    },
+    {
+      "epoch": 0.9292035398230089,
+      "grad_norm": 0.2706278456881864,
+      "learning_rate": 4.0423501193151416e-05,
+      "loss": 0.1537,
+      "step": 1785
+    },
+    {
+      "epoch": 0.9297241020301926,
+      "grad_norm": 0.2540911793302258,
+      "learning_rate": 4.041244122084747e-05,
+      "loss": 0.1513,
+      "step": 1786
+    },
+    {
+      "epoch": 0.9302446642373764,
+      "grad_norm": 0.25585354118375886,
+      "learning_rate": 4.040137638057735e-05,
+      "loss": 0.1558,
+      "step": 1787
+    },
+    {
+      "epoch": 0.9307652264445602,
+      "grad_norm": 0.23363718441591247,
+      "learning_rate": 4.039030667583585e-05,
+      "loss": 0.1468,
+      "step": 1788
+    },
+    {
+      "epoch": 0.9312857886517438,
+      "grad_norm": 0.2779496142875094,
+      "learning_rate": 4.037923211011929e-05,
+      "loss": 0.1513,
+      "step": 1789
+    },
+    {
+      "epoch": 0.9318063508589276,
+      "grad_norm": 0.235742116241504,
+      "learning_rate": 4.036815268692556e-05,
+      "loss": 0.1501,
+      "step": 1790
+    },
+    {
+      "epoch": 0.9323269130661114,
+      "grad_norm": 0.2561251442991573,
+      "learning_rate": 4.035706840975403e-05,
+      "loss": 0.1611,
+      "step": 1791
+    },
+    {
+      "epoch": 0.9328474752732951,
+      "grad_norm": 0.2564452558632798,
+      "learning_rate": 4.0345979282105637e-05,
+      "loss": 0.1573,
+      "step": 1792
+    },
+    {
+      "epoch": 0.9333680374804789,
+      "grad_norm": 0.2545473340743804,
+      "learning_rate": 4.033488530748285e-05,
+      "loss": 0.1543,
+      "step": 1793
+    },
+    {
+      "epoch": 0.9338885996876627,
+      "grad_norm": 0.25319840729825216,
+      "learning_rate": 4.032378648938966e-05,
+      "loss": 0.1591,
+      "step": 1794
+    },
+    {
+      "epoch": 0.9344091618948465,
+      "grad_norm": 0.24462206700432024,
+      "learning_rate": 4.031268283133158e-05,
+      "loss": 0.1492,
+      "step": 1795
+    },
+    {
+      "epoch": 0.9349297241020302,
+      "grad_norm": 0.2442034089324051,
+      "learning_rate": 4.030157433681568e-05,
+      "loss": 0.1564,
+      "step": 1796
+    },
+    {
+      "epoch": 0.9354502863092139,
+      "grad_norm": 0.23610337924487748,
+      "learning_rate": 4.0290461009350535e-05,
+      "loss": 0.1446,
+      "step": 1797
+    },
+    {
+      "epoch": 0.9359708485163977,
+      "grad_norm": 0.7555257854589889,
+      "learning_rate": 4.0279342852446234e-05,
+      "loss": 0.154,
+      "step": 1798
+    },
+    {
+      "epoch": 0.9364914107235814,
+      "grad_norm": 0.2730083918906193,
+      "learning_rate": 4.026821986961443e-05,
+      "loss": 0.1557,
+      "step": 1799
+    },
+    {
+      "epoch": 0.9370119729307652,
+      "grad_norm": 0.24844427506768513,
+      "learning_rate": 4.0257092064368266e-05,
+      "loss": 0.1556,
+      "step": 1800
+    },
+    {
+      "epoch": 0.937532535137949,
+      "grad_norm": 0.2550146011348462,
+      "learning_rate": 4.0245959440222425e-05,
+      "loss": 0.1526,
+      "step": 1801
+    },
+    {
+      "epoch": 0.9380530973451328,
+      "grad_norm": 0.2643186732751096,
+      "learning_rate": 4.023482200069311e-05,
+      "loss": 0.1648,
+      "step": 1802
+    },
+    {
+      "epoch": 0.9385736595523165,
+      "grad_norm": 0.2375368349668477,
+      "learning_rate": 4.0223679749298025e-05,
+      "loss": 0.1558,
+      "step": 1803
+    },
+    {
+      "epoch": 0.9390942217595003,
+      "grad_norm": 0.24780404748279455,
+      "learning_rate": 4.021253268955644e-05,
+      "loss": 0.1558,
+      "step": 1804
+    },
+    {
+      "epoch": 0.939614783966684,
+      "grad_norm": 0.24937088179496925,
+      "learning_rate": 4.02013808249891e-05,
+      "loss": 0.1525,
+      "step": 1805
+    },
+    {
+      "epoch": 0.9401353461738677,
+      "grad_norm": 0.247516280623822,
+      "learning_rate": 4.019022415911828e-05,
+      "loss": 0.1497,
+      "step": 1806
+    },
+    {
+      "epoch": 0.9406559083810515,
+      "grad_norm": 0.2565387438992933,
+      "learning_rate": 4.0179062695467784e-05,
+      "loss": 0.154,
+      "step": 1807
+    },
+    {
+      "epoch": 0.9411764705882353,
+      "grad_norm": 0.25957570816875336,
+      "learning_rate": 4.016789643756291e-05,
+      "loss": 0.1575,
+      "step": 1808
+    },
+    {
+      "epoch": 0.941697032795419,
+      "grad_norm": 0.25711864420404396,
+      "learning_rate": 4.0156725388930495e-05,
+      "loss": 0.1583,
+      "step": 1809
+    },
+    {
+      "epoch": 0.9422175950026028,
+      "grad_norm": 0.2439002536135534,
+      "learning_rate": 4.014554955309886e-05,
+      "loss": 0.1538,
+      "step": 1810
+    },
+    {
+      "epoch": 0.9427381572097866,
+      "grad_norm": 0.2734062021761085,
+      "learning_rate": 4.0134368933597863e-05,
+      "loss": 0.1587,
+      "step": 1811
+    },
+    {
+      "epoch": 0.9432587194169704,
+      "grad_norm": 0.24078383939318468,
+      "learning_rate": 4.012318353395887e-05,
+      "loss": 0.1541,
+      "step": 1812
+    },
+    {
+      "epoch": 0.943779281624154,
+      "grad_norm": 0.25527670699498406,
+      "learning_rate": 4.011199335771475e-05,
+      "loss": 0.1641,
+      "step": 1813
+    },
+    {
+      "epoch": 0.9442998438313378,
+      "grad_norm": 0.26131896589926484,
+      "learning_rate": 4.010079840839987e-05,
+      "loss": 0.1602,
+      "step": 1814
+    },
+    {
+      "epoch": 0.9448204060385216,
+      "grad_norm": 0.2457034759853004,
+      "learning_rate": 4.0089598689550126e-05,
+      "loss": 0.149,
+      "step": 1815
+    },
+    {
+      "epoch": 0.9453409682457053,
+      "grad_norm": 0.2466235816689623,
+      "learning_rate": 4.0078394204702895e-05,
+      "loss": 0.151,
+      "step": 1816
+    },
+    {
+      "epoch": 0.9458615304528891,
+      "grad_norm": 0.261112482926145,
+      "learning_rate": 4.0067184957397096e-05,
+      "loss": 0.1536,
+      "step": 1817
+    },
+    {
+      "epoch": 0.9463820926600729,
+      "grad_norm": 0.2758752170090889,
+      "learning_rate": 4.0055970951173116e-05,
+      "loss": 0.1633,
+      "step": 1818
+    },
+    {
+      "epoch": 0.9469026548672567,
+      "grad_norm": 0.266272488103662,
+      "learning_rate": 4.004475218957287e-05,
+      "loss": 0.1568,
+      "step": 1819
+    },
+    {
+      "epoch": 0.9474232170744404,
+      "grad_norm": 0.2733192385804763,
+      "learning_rate": 4.003352867613975e-05,
+      "loss": 0.1526,
+      "step": 1820
+    },
+    {
+      "epoch": 0.9479437792816241,
+      "grad_norm": 0.2502565692501297,
+      "learning_rate": 4.002230041441868e-05,
+      "loss": 0.1498,
+      "step": 1821
+    },
+    {
+      "epoch": 0.9484643414888079,
+      "grad_norm": 0.2981229168833489,
+      "learning_rate": 4.001106740795607e-05,
+      "loss": 0.1535,
+      "step": 1822
+    },
+    {
+      "epoch": 0.9489849036959916,
+      "grad_norm": 0.25792533892269764,
+      "learning_rate": 3.9999829660299806e-05,
+      "loss": 0.1546,
+      "step": 1823
+    },
+    {
+      "epoch": 0.9495054659031754,
+      "grad_norm": 0.2754778565483347,
+      "learning_rate": 3.998858717499931e-05,
+      "loss": 0.154,
+      "step": 1824
+    },
+    {
+      "epoch": 0.9500260281103592,
+      "grad_norm": 0.25581628643187587,
+      "learning_rate": 3.997733995560547e-05,
+      "loss": 0.1588,
+      "step": 1825
+    },
+    {
+      "epoch": 0.950546590317543,
+      "grad_norm": 0.29974295677568763,
+      "learning_rate": 3.9966088005670686e-05,
+      "loss": 0.1581,
+      "step": 1826
+    },
+    {
+      "epoch": 0.9510671525247267,
+      "grad_norm": 0.2566154047129761,
+      "learning_rate": 3.995483132874885e-05,
+      "loss": 0.1521,
+      "step": 1827
+    },
+    {
+      "epoch": 0.9515877147319105,
+      "grad_norm": 0.26876875526607946,
+      "learning_rate": 3.994356992839535e-05,
+      "loss": 0.1608,
+      "step": 1828
+    },
+    {
+      "epoch": 0.9521082769390942,
+      "grad_norm": 0.2774171688241344,
+      "learning_rate": 3.993230380816705e-05,
+      "loss": 0.1515,
+      "step": 1829
+    },
+    {
+      "epoch": 0.9526288391462779,
+      "grad_norm": 0.24564551896719408,
+      "learning_rate": 3.9921032971622306e-05,
+      "loss": 0.1561,
+      "step": 1830
+    },
+    {
+      "epoch": 0.9531494013534617,
+      "grad_norm": 0.2463503110268932,
+      "learning_rate": 3.9909757422321e-05,
+      "loss": 0.1537,
+      "step": 1831
+    },
+    {
+      "epoch": 0.9536699635606455,
+      "grad_norm": 0.267020735900316,
+      "learning_rate": 3.9898477163824454e-05,
+      "loss": 0.1522,
+      "step": 1832
+    },
+    {
+      "epoch": 0.9541905257678293,
+      "grad_norm": 0.2602448863644486,
+      "learning_rate": 3.98871921996955e-05,
+      "loss": 0.1571,
+      "step": 1833
+    },
+    {
+      "epoch": 0.954711087975013,
+      "grad_norm": 0.24776161325403143,
+      "learning_rate": 3.9875902533498465e-05,
+      "loss": 0.1578,
+      "step": 1834
+    },
+    {
+      "epoch": 0.9552316501821968,
+      "grad_norm": 0.25779665449960754,
+      "learning_rate": 3.986460816879913e-05,
+      "loss": 0.1496,
+      "step": 1835
+    },
+    {
+      "epoch": 0.9557522123893806,
+      "grad_norm": 0.2701760173076487,
+      "learning_rate": 3.985330910916482e-05,
+      "loss": 0.1569,
+      "step": 1836
+    },
+    {
+      "epoch": 0.9562727745965642,
+      "grad_norm": 0.25325580552113314,
+      "learning_rate": 3.984200535816427e-05,
+      "loss": 0.1566,
+      "step": 1837
+    },
+    {
+      "epoch": 0.956793336803748,
+      "grad_norm": 0.29664763959854834,
+      "learning_rate": 3.983069691936773e-05,
+      "loss": 0.1534,
+      "step": 1838
+    },
+    {
+      "epoch": 0.9573138990109318,
+      "grad_norm": 0.27443913965950784,
+      "learning_rate": 3.981938379634696e-05,
+      "loss": 0.1587,
+      "step": 1839
+    },
+    {
+      "epoch": 0.9578344612181156,
+      "grad_norm": 0.3036453223080745,
+      "learning_rate": 3.980806599267514e-05,
+      "loss": 0.1622,
+      "step": 1840
+    },
+    {
+      "epoch": 0.9583550234252993,
+      "grad_norm": 0.2703683477848925,
+      "learning_rate": 3.979674351192697e-05,
+      "loss": 0.1512,
+      "step": 1841
+    },
+    {
+      "epoch": 0.9588755856324831,
+      "grad_norm": 0.25118682959451855,
+      "learning_rate": 3.978541635767862e-05,
+      "loss": 0.1479,
+      "step": 1842
+    },
+    {
+      "epoch": 0.9593961478396669,
+      "grad_norm": 0.27007643798115183,
+      "learning_rate": 3.977408453350773e-05,
+      "loss": 0.1478,
+      "step": 1843
+    },
+    {
+      "epoch": 0.9599167100468506,
+      "grad_norm": 0.24905501625360876,
+      "learning_rate": 3.976274804299342e-05,
+      "loss": 0.1492,
+      "step": 1844
+    },
+    {
+      "epoch": 0.9604372722540343,
+      "grad_norm": 0.2721594566362824,
+      "learning_rate": 3.975140688971628e-05,
+      "loss": 0.1577,
+      "step": 1845
+    },
+    {
+      "epoch": 0.9609578344612181,
+      "grad_norm": 0.26822647420622714,
+      "learning_rate": 3.974006107725837e-05,
+      "loss": 0.1542,
+      "step": 1846
+    },
+    {
+      "epoch": 0.9614783966684018,
+      "grad_norm": 0.2610073474569343,
+      "learning_rate": 3.972871060920323e-05,
+      "loss": 0.1573,
+      "step": 1847
+    },
+    {
+      "epoch": 0.9619989588755856,
+      "grad_norm": 0.24741924184000724,
+      "learning_rate": 3.971735548913586e-05,
+      "loss": 0.1572,
+      "step": 1848
+    },
+    {
+      "epoch": 0.9625195210827694,
+      "grad_norm": 0.2497560572681831,
+      "learning_rate": 3.970599572064275e-05,
+      "loss": 0.1627,
+      "step": 1849
+    },
+    {
+      "epoch": 0.9630400832899532,
+      "grad_norm": 0.261255891586638,
+      "learning_rate": 3.969463130731183e-05,
+      "loss": 0.146,
+      "step": 1850
+    },
+    {
+      "epoch": 0.9635606454971369,
+      "grad_norm": 0.2661862951079788,
+      "learning_rate": 3.968326225273251e-05,
+      "loss": 0.1549,
+      "step": 1851
+    },
+    {
+      "epoch": 0.9640812077043207,
+      "grad_norm": 0.26696523143923034,
+      "learning_rate": 3.9671888560495676e-05,
+      "loss": 0.1584,
+      "step": 1852
+    },
+    {
+      "epoch": 0.9646017699115044,
+      "grad_norm": 0.28425969486144664,
+      "learning_rate": 3.966051023419366e-05,
+      "loss": 0.1585,
+      "step": 1853
+    },
+    {
+      "epoch": 0.9651223321186881,
+      "grad_norm": 0.2252880385971139,
+      "learning_rate": 3.964912727742027e-05,
+      "loss": 0.1468,
+      "step": 1854
+    },
+    {
+      "epoch": 0.9656428943258719,
+      "grad_norm": 0.2746297685611525,
+      "learning_rate": 3.963773969377077e-05,
+      "loss": 0.1526,
+      "step": 1855
+    },
+    {
+      "epoch": 0.9661634565330557,
+      "grad_norm": 0.2562090900129971,
+      "learning_rate": 3.9626347486841896e-05,
+      "loss": 0.1574,
+      "step": 1856
+    },
+    {
+      "epoch": 0.9666840187402395,
+      "grad_norm": 0.2641947002589958,
+      "learning_rate": 3.961495066023184e-05,
+      "loss": 0.1579,
+      "step": 1857
+    },
+    {
+      "epoch": 0.9672045809474232,
+      "grad_norm": 0.2576610491884646,
+      "learning_rate": 3.9603549217540235e-05,
+      "loss": 0.1566,
+      "step": 1858
+    },
+    {
+      "epoch": 0.967725143154607,
+      "grad_norm": 0.22546307618447003,
+      "learning_rate": 3.959214316236821e-05,
+      "loss": 0.1451,
+      "step": 1859
+    },
+    {
+      "epoch": 0.9682457053617908,
+      "grad_norm": 0.27228400782836965,
+      "learning_rate": 3.95807324983183e-05,
+      "loss": 0.1544,
+      "step": 1860
+    },
+    {
+      "epoch": 0.9687662675689744,
+      "grad_norm": 0.2404970285210459,
+      "learning_rate": 3.956931722899454e-05,
+      "loss": 0.1586,
+      "step": 1861
+    },
+    {
+      "epoch": 0.9692868297761582,
+      "grad_norm": 0.2483984351066557,
+      "learning_rate": 3.955789735800241e-05,
+      "loss": 0.1519,
+      "step": 1862
+    },
+    {
+      "epoch": 0.969807391983342,
+      "grad_norm": 0.26080827724904837,
+      "learning_rate": 3.954647288894883e-05,
+      "loss": 0.1577,
+      "step": 1863
+    },
+    {
+      "epoch": 0.9703279541905258,
+      "grad_norm": 0.24279634879644127,
+      "learning_rate": 3.953504382544216e-05,
+      "loss": 0.1529,
+      "step": 1864
+    },
+    {
+      "epoch": 0.9708485163977095,
+      "grad_norm": 0.24834103365291627,
+      "learning_rate": 3.952361017109226e-05,
+      "loss": 0.1494,
+      "step": 1865
+    },
+    {
+      "epoch": 0.9713690786048933,
+      "grad_norm": 0.24460951509898635,
+      "learning_rate": 3.95121719295104e-05,
+      "loss": 0.1571,
+      "step": 1866
+    },
+    {
+      "epoch": 0.9718896408120771,
+      "grad_norm": 0.26683915335760894,
+      "learning_rate": 3.95007291043093e-05,
+      "loss": 0.1531,
+      "step": 1867
+    },
+    {
+      "epoch": 0.9724102030192608,
+      "grad_norm": 0.24054568963656242,
+      "learning_rate": 3.9489281699103145e-05,
+      "loss": 0.1516,
+      "step": 1868
+    },
+    {
+      "epoch": 0.9729307652264445,
+      "grad_norm": 0.24375115372025324,
+      "learning_rate": 3.947782971750755e-05,
+      "loss": 0.1488,
+      "step": 1869
+    },
+    {
+      "epoch": 0.9734513274336283,
+      "grad_norm": 0.25412067934433286,
+      "learning_rate": 3.94663731631396e-05,
+      "loss": 0.1607,
+      "step": 1870
+    },
+    {
+      "epoch": 0.973971889640812,
+      "grad_norm": 0.2397702543825332,
+      "learning_rate": 3.945491203961779e-05,
+      "loss": 0.1511,
+      "step": 1871
+    },
+    {
+      "epoch": 0.9744924518479958,
+      "grad_norm": 0.2806540915038973,
+      "learning_rate": 3.94434463505621e-05,
+      "loss": 0.154,
+      "step": 1872
+    },
+    {
+      "epoch": 0.9750130140551796,
+      "grad_norm": 0.2546854416214566,
+      "learning_rate": 3.9431976099593896e-05,
+      "loss": 0.1605,
+      "step": 1873
+    },
+    {
+      "epoch": 0.9755335762623634,
+      "grad_norm": 0.2575288721304389,
+      "learning_rate": 3.942050129033603e-05,
+      "loss": 0.1485,
+      "step": 1874
+    },
+    {
+      "epoch": 0.9760541384695471,
+      "grad_norm": 0.26908803144635207,
+      "learning_rate": 3.9409021926412795e-05,
+      "loss": 0.1549,
+      "step": 1875
+    },
+    {
+      "epoch": 0.9765747006767309,
+      "grad_norm": 0.29289715788111687,
+      "learning_rate": 3.9397538011449894e-05,
+      "loss": 0.1563,
+      "step": 1876
+    },
+    {
+      "epoch": 0.9770952628839146,
+      "grad_norm": 0.23831369536611632,
+      "learning_rate": 3.938604954907449e-05,
+      "loss": 0.1541,
+      "step": 1877
+    },
+    {
+      "epoch": 0.9776158250910983,
+      "grad_norm": 0.2589658450603122,
+      "learning_rate": 3.9374556542915167e-05,
+      "loss": 0.1519,
+      "step": 1878
+    },
+    {
+      "epoch": 0.9781363872982821,
+      "grad_norm": 0.3008301402515451,
+      "learning_rate": 3.936305899660195e-05,
+      "loss": 0.1554,
+      "step": 1879
+    },
+    {
+      "epoch": 0.9786569495054659,
+      "grad_norm": 0.2579134695099253,
+      "learning_rate": 3.935155691376631e-05,
+      "loss": 0.16,
+      "step": 1880
+    },
+    {
+      "epoch": 0.9791775117126497,
+      "grad_norm": 0.2755461345867494,
+      "learning_rate": 3.934005029804112e-05,
+      "loss": 0.1495,
+      "step": 1881
+    },
+    {
+      "epoch": 0.9796980739198334,
+      "grad_norm": 0.25850824899183367,
+      "learning_rate": 3.9328539153060725e-05,
+      "loss": 0.1589,
+      "step": 1882
+    },
+    {
+      "epoch": 0.9802186361270172,
+      "grad_norm": 0.2823202784003672,
+      "learning_rate": 3.931702348246087e-05,
+      "loss": 0.1588,
+      "step": 1883
+    },
+    {
+      "epoch": 0.980739198334201,
+      "grad_norm": 0.25907134737796256,
+      "learning_rate": 3.930550328987875e-05,
+      "loss": 0.1516,
+      "step": 1884
+    },
+    {
+      "epoch": 0.9812597605413846,
+      "grad_norm": 0.26429422471247077,
+      "learning_rate": 3.929397857895297e-05,
+      "loss": 0.1571,
+      "step": 1885
+    },
+    {
+      "epoch": 0.9817803227485684,
+      "grad_norm": 0.25776033042758834,
+      "learning_rate": 3.928244935332356e-05,
+      "loss": 0.1565,
+      "step": 1886
+    },
+    {
+      "epoch": 0.9823008849557522,
+      "grad_norm": 0.23117386280694466,
+      "learning_rate": 3.9270915616632e-05,
+      "loss": 0.1532,
+      "step": 1887
+    },
+    {
+      "epoch": 0.982821447162936,
+      "grad_norm": 0.2589601378005726,
+      "learning_rate": 3.9259377372521176e-05,
+      "loss": 0.1603,
+      "step": 1888
+    },
+    {
+      "epoch": 0.9833420093701197,
+      "grad_norm": 0.24736053021683718,
+      "learning_rate": 3.924783462463541e-05,
+      "loss": 0.1482,
+      "step": 1889
+    },
+    {
+      "epoch": 0.9838625715773035,
+      "grad_norm": 0.24050172557296737,
+      "learning_rate": 3.923628737662043e-05,
+      "loss": 0.1558,
+      "step": 1890
+    },
+    {
+      "epoch": 0.9843831337844873,
+      "grad_norm": 0.25204285474497545,
+      "learning_rate": 3.9224735632123395e-05,
+      "loss": 0.1588,
+      "step": 1891
+    },
+    {
+      "epoch": 0.984903695991671,
+      "grad_norm": 0.24569292644864882,
+      "learning_rate": 3.921317939479289e-05,
+      "loss": 0.149,
+      "step": 1892
+    },
+    {
+      "epoch": 0.9854242581988547,
+      "grad_norm": 0.2591242502253903,
+      "learning_rate": 3.920161866827889e-05,
+      "loss": 0.1515,
+      "step": 1893
+    },
+    {
+      "epoch": 0.9859448204060385,
+      "grad_norm": 0.25314114488184825,
+      "learning_rate": 3.919005345623285e-05,
+      "loss": 0.1583,
+      "step": 1894
+    },
+    {
+      "epoch": 0.9864653826132223,
+      "grad_norm": 0.2707662697095083,
+      "learning_rate": 3.917848376230757e-05,
+      "loss": 0.1474,
+      "step": 1895
+    },
+    {
+      "epoch": 0.986985944820406,
+      "grad_norm": 0.2489741320109898,
+      "learning_rate": 3.916690959015731e-05,
+      "loss": 0.1604,
+      "step": 1896
+    },
+    {
+      "epoch": 0.9875065070275898,
+      "grad_norm": 0.26650693500099853,
+      "learning_rate": 3.915533094343773e-05,
+      "loss": 0.1484,
+      "step": 1897
+    },
+    {
+      "epoch": 0.9880270692347736,
+      "grad_norm": 0.24576821619143538,
+      "learning_rate": 3.914374782580591e-05,
+      "loss": 0.1533,
+      "step": 1898
+    },
+    {
+      "epoch": 0.9885476314419573,
+      "grad_norm": 0.2242445667075516,
+      "learning_rate": 3.913216024092032e-05,
+      "loss": 0.1453,
+      "step": 1899
+    },
+    {
+      "epoch": 0.9890681936491411,
+      "grad_norm": 0.24170044956082523,
+      "learning_rate": 3.912056819244089e-05,
+      "loss": 0.15,
+      "step": 1900
+    },
+    {
+      "epoch": 0.9895887558563248,
+      "grad_norm": 0.23999183798149767,
+      "learning_rate": 3.910897168402889e-05,
+      "loss": 0.1445,
+      "step": 1901
+    },
+    {
+      "epoch": 0.9901093180635085,
+      "grad_norm": 0.24978308599133406,
+      "learning_rate": 3.909737071934707e-05,
+      "loss": 0.1516,
+      "step": 1902
+    },
+    {
+      "epoch": 0.9906298802706923,
+      "grad_norm": 0.2477655018309477,
+      "learning_rate": 3.9085765302059554e-05,
+      "loss": 0.1617,
+      "step": 1903
+    },
+    {
+      "epoch": 0.9911504424778761,
+      "grad_norm": 0.24166671338937565,
+      "learning_rate": 3.907415543583184e-05,
+      "loss": 0.1537,
+      "step": 1904
+    },
+    {
+      "epoch": 0.9916710046850599,
+      "grad_norm": 0.24364319933406497,
+      "learning_rate": 3.9062541124330884e-05,
+      "loss": 0.1529,
+      "step": 1905
+    },
+    {
+      "epoch": 0.9921915668922436,
+      "grad_norm": 0.2646257000445511,
+      "learning_rate": 3.905092237122504e-05,
+      "loss": 0.1554,
+      "step": 1906
+    },
+    {
+      "epoch": 0.9927121290994274,
+      "grad_norm": 0.2496525283069466,
+      "learning_rate": 3.903929918018403e-05,
+      "loss": 0.1511,
+      "step": 1907
+    },
+    {
+      "epoch": 0.9932326913066112,
+      "grad_norm": 0.2309200833828225,
+      "learning_rate": 3.902767155487901e-05,
+      "loss": 0.154,
+      "step": 1908
+    },
+    {
+      "epoch": 0.9937532535137948,
+      "grad_norm": 0.2451270375417471,
+      "learning_rate": 3.9016039498982515e-05,
+      "loss": 0.1453,
+      "step": 1909
+    },
+    {
+      "epoch": 0.9942738157209786,
+      "grad_norm": 0.2699658100912516,
+      "learning_rate": 3.90044030161685e-05,
+      "loss": 0.1545,
+      "step": 1910
+    },
+    {
+      "epoch": 0.9947943779281624,
+      "grad_norm": 0.26692292482327223,
+      "learning_rate": 3.8992762110112304e-05,
+      "loss": 0.1561,
+      "step": 1911
+    },
+    {
+      "epoch": 0.9953149401353462,
+      "grad_norm": 0.24585278390934867,
+      "learning_rate": 3.8981116784490666e-05,
+      "loss": 0.147,
+      "step": 1912
+    },
+    {
+      "epoch": 0.9958355023425299,
+      "grad_norm": 0.25147680505901215,
+      "learning_rate": 3.896946704298172e-05,
+      "loss": 0.1521,
+      "step": 1913
+    },
+    {
+      "epoch": 0.9963560645497137,
+      "grad_norm": 0.25230893137578037,
+      "learning_rate": 3.8957812889265e-05,
+      "loss": 0.1541,
+      "step": 1914
+    },
+    {
+      "epoch": 0.9968766267568975,
+      "grad_norm": 0.24038428447771637,
+      "learning_rate": 3.8946154327021434e-05,
+      "loss": 0.1498,
+      "step": 1915
+    },
+    {
+      "epoch": 0.9973971889640812,
+      "grad_norm": 0.2466022950313137,
+      "learning_rate": 3.893449135993333e-05,
+      "loss": 0.1632,
+      "step": 1916
+    },
+    {
+      "epoch": 0.9979177511712649,
+      "grad_norm": 0.25366577440341587,
+      "learning_rate": 3.89228239916844e-05,
+      "loss": 0.146,
+      "step": 1917
+    },
+    {
+      "epoch": 0.9984383133784487,
+      "grad_norm": 0.24171289918790415,
+      "learning_rate": 3.8911152225959743e-05,
+      "loss": 0.1561,
+      "step": 1918
+    },
+    {
+      "epoch": 0.9989588755856325,
+      "grad_norm": 0.2581505141982399,
+      "learning_rate": 3.889947606644584e-05,
+      "loss": 0.1595,
+      "step": 1919
+    },
+    {
+      "epoch": 0.9994794377928162,
+      "grad_norm": 0.24747053780836764,
+      "learning_rate": 3.888779551683057e-05,
+      "loss": 0.1519,
+      "step": 1920
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.237124505232837,
+      "learning_rate": 3.8876110580803186e-05,
+      "loss": 0.1481,
+      "step": 1921
+    },
+    {
+      "epoch": 1.0005205622071838,
+      "grad_norm": 0.2661627018308619,
+      "learning_rate": 3.886442126205435e-05,
+      "loss": 0.1194,
+      "step": 1922
+    },
+    {
+      "epoch": 1.0010411244143675,
+      "grad_norm": 0.2520650370090721,
+      "learning_rate": 3.8852727564276086e-05,
+      "loss": 0.1134,
+      "step": 1923
+    },
+    {
+      "epoch": 1.0015616866215513,
+      "grad_norm": 0.22725768987546488,
+      "learning_rate": 3.884102949116181e-05,
+      "loss": 0.1149,
+      "step": 1924
+    },
+    {
+      "epoch": 1.002082248828735,
+      "grad_norm": 0.25305995229234013,
+      "learning_rate": 3.8829327046406304e-05,
+      "loss": 0.1111,
+      "step": 1925
+    },
+    {
+      "epoch": 1.0026028110359189,
+      "grad_norm": 0.27939931613711055,
+      "learning_rate": 3.881762023370576e-05,
+      "loss": 0.1104,
+      "step": 1926
+    },
+    {
+      "epoch": 1.0031233732431026,
+      "grad_norm": 0.3043520089700027,
+      "learning_rate": 3.880590905675773e-05,
+      "loss": 0.1232,
+      "step": 1927
+    },
+    {
+      "epoch": 1.0036439354502864,
+      "grad_norm": 0.261721428329474,
+      "learning_rate": 3.879419351926115e-05,
+      "loss": 0.1076,
+      "step": 1928
+    },
+    {
+      "epoch": 1.0041644976574702,
+      "grad_norm": 0.2700146216167829,
+      "learning_rate": 3.878247362491633e-05,
+      "loss": 0.1143,
+      "step": 1929
+    },
+    {
+      "epoch": 1.0046850598646537,
+      "grad_norm": 0.2512971409772084,
+      "learning_rate": 3.877074937742495e-05,
+      "loss": 0.1063,
+      "step": 1930
+    },
+    {
+      "epoch": 1.0052056220718375,
+      "grad_norm": 0.24157054726586163,
+      "learning_rate": 3.8759020780490094e-05,
+      "loss": 0.1139,
+      "step": 1931
+    },
+    {
+      "epoch": 1.0057261842790213,
+      "grad_norm": 0.25231694869276755,
+      "learning_rate": 3.8747287837816184e-05,
+      "loss": 0.1167,
+      "step": 1932
+    },
+    {
+      "epoch": 1.006246746486205,
+      "grad_norm": 0.28056071702883195,
+      "learning_rate": 3.8735550553109024e-05,
+      "loss": 0.1148,
+      "step": 1933
+    },
+    {
+      "epoch": 1.0067673086933888,
+      "grad_norm": 0.23802118673016223,
+      "learning_rate": 3.87238089300758e-05,
+      "loss": 0.1061,
+      "step": 1934
+    },
+    {
+      "epoch": 1.0072878709005726,
+      "grad_norm": 0.2518313655139843,
+      "learning_rate": 3.8712062972425077e-05,
+      "loss": 0.1144,
+      "step": 1935
+    },
+    {
+      "epoch": 1.0078084331077564,
+      "grad_norm": 0.2687038083950964,
+      "learning_rate": 3.870031268386676e-05,
+      "loss": 0.1153,
+      "step": 1936
+    },
+    {
+      "epoch": 1.0083289953149401,
+      "grad_norm": 0.2618181075208367,
+      "learning_rate": 3.868855806811212e-05,
+      "loss": 0.1177,
+      "step": 1937
+    },
+    {
+      "epoch": 1.008849557522124,
+      "grad_norm": 0.2685854416587716,
+      "learning_rate": 3.867679912887385e-05,
+      "loss": 0.1186,
+      "step": 1938
+    },
+    {
+      "epoch": 1.0093701197293077,
+      "grad_norm": 0.26559471123613715,
+      "learning_rate": 3.866503586986595e-05,
+      "loss": 0.1161,
+      "step": 1939
+    },
+    {
+      "epoch": 1.0098906819364915,
+      "grad_norm": 0.24409238460263524,
+      "learning_rate": 3.865326829480381e-05,
+      "loss": 0.1102,
+      "step": 1940
+    },
+    {
+      "epoch": 1.0104112441436752,
+      "grad_norm": 0.2937743298818115,
+      "learning_rate": 3.864149640740417e-05,
+      "loss": 0.1145,
+      "step": 1941
+    },
+    {
+      "epoch": 1.010931806350859,
+      "grad_norm": 0.25001092468218145,
+      "learning_rate": 3.862972021138514e-05,
+      "loss": 0.1117,
+      "step": 1942
+    },
+    {
+      "epoch": 1.0114523685580428,
+      "grad_norm": 0.24032698365614474,
+      "learning_rate": 3.86179397104662e-05,
+      "loss": 0.1083,
+      "step": 1943
+    },
+    {
+      "epoch": 1.0119729307652265,
+      "grad_norm": 0.26454925482485414,
+      "learning_rate": 3.860615490836817e-05,
+      "loss": 0.1122,
+      "step": 1944
+    },
+    {
+      "epoch": 1.01249349297241,
+      "grad_norm": 0.25931975068403923,
+      "learning_rate": 3.859436580881325e-05,
+      "loss": 0.113,
+      "step": 1945
+    },
+    {
+      "epoch": 1.0130140551795939,
+      "grad_norm": 0.28362163389354017,
+      "learning_rate": 3.858257241552498e-05,
+      "loss": 0.1158,
+      "step": 1946
+    },
+    {
+      "epoch": 1.0135346173867776,
+      "grad_norm": 0.24978572696444806,
+      "learning_rate": 3.857077473222825e-05,
+      "loss": 0.1084,
+      "step": 1947
+    },
+    {
+      "epoch": 1.0140551795939614,
+      "grad_norm": 0.24629450541328152,
+      "learning_rate": 3.855897276264934e-05,
+      "loss": 0.1108,
+      "step": 1948
+    },
+    {
+      "epoch": 1.0145757418011452,
+      "grad_norm": 0.24448076133842503,
+      "learning_rate": 3.8547166510515854e-05,
+      "loss": 0.1142,
+      "step": 1949
+    },
+    {
+      "epoch": 1.015096304008329,
+      "grad_norm": 0.25912362617154444,
+      "learning_rate": 3.8535355979556755e-05,
+      "loss": 0.1185,
+      "step": 1950
+    },
+    {
+      "epoch": 1.0156168662155127,
+      "grad_norm": 0.26406004163474434,
+      "learning_rate": 3.852354117350235e-05,
+      "loss": 0.1137,
+      "step": 1951
+    },
+    {
+      "epoch": 1.0161374284226965,
+      "grad_norm": 0.23169769727224848,
+      "learning_rate": 3.8511722096084313e-05,
+      "loss": 0.1058,
+      "step": 1952
+    },
+    {
+      "epoch": 1.0166579906298803,
+      "grad_norm": 0.2420173219027427,
+      "learning_rate": 3.8499898751035656e-05,
+      "loss": 0.1091,
+      "step": 1953
+    },
+    {
+      "epoch": 1.017178552837064,
+      "grad_norm": 0.24237055917616832,
+      "learning_rate": 3.848807114209074e-05,
+      "loss": 0.1121,
+      "step": 1954
+    },
+    {
+      "epoch": 1.0176991150442478,
+      "grad_norm": 0.2417178678114401,
+      "learning_rate": 3.8476239272985284e-05,
+      "loss": 0.1138,
+      "step": 1955
+    },
+    {
+      "epoch": 1.0182196772514316,
+      "grad_norm": 0.24659252456869502,
+      "learning_rate": 3.846440314745633e-05,
+      "loss": 0.1173,
+      "step": 1956
+    },
+    {
+      "epoch": 1.0187402394586154,
+      "grad_norm": 0.234545477823226,
+      "learning_rate": 3.8452562769242276e-05,
+      "loss": 0.1136,
+      "step": 1957
+    },
+    {
+      "epoch": 1.0192608016657991,
+      "grad_norm": 0.23798416850355458,
+      "learning_rate": 3.844071814208288e-05,
+      "loss": 0.1121,
+      "step": 1958
+    },
+    {
+      "epoch": 1.019781363872983,
+      "grad_norm": 0.26896997384709637,
+      "learning_rate": 3.842886926971922e-05,
+      "loss": 0.1119,
+      "step": 1959
+    },
+    {
+      "epoch": 1.0203019260801667,
+      "grad_norm": 0.2461574213164607,
+      "learning_rate": 3.8417016155893716e-05,
+      "loss": 0.1127,
+      "step": 1960
+    },
+    {
+      "epoch": 1.0208224882873504,
+      "grad_norm": 0.23767808403209464,
+      "learning_rate": 3.840515880435013e-05,
+      "loss": 0.1085,
+      "step": 1961
+    },
+    {
+      "epoch": 1.021343050494534,
+      "grad_norm": 0.2713143335946204,
+      "learning_rate": 3.839329721883358e-05,
+      "loss": 0.1111,
+      "step": 1962
+    },
+    {
+      "epoch": 1.0218636127017178,
+      "grad_norm": 0.2599294251623557,
+      "learning_rate": 3.8381431403090494e-05,
+      "loss": 0.1126,
+      "step": 1963
+    },
+    {
+      "epoch": 1.0223841749089015,
+      "grad_norm": 0.24545293936346022,
+      "learning_rate": 3.8369561360868656e-05,
+      "loss": 0.1105,
+      "step": 1964
+    },
+    {
+      "epoch": 1.0229047371160853,
+      "grad_norm": 0.24512412624134056,
+      "learning_rate": 3.835768709591717e-05,
+      "loss": 0.1097,
+      "step": 1965
+    },
+    {
+      "epoch": 1.023425299323269,
+      "grad_norm": 0.2458438658532382,
+      "learning_rate": 3.8345808611986485e-05,
+      "loss": 0.1068,
+      "step": 1966
+    },
+    {
+      "epoch": 1.0239458615304529,
+      "grad_norm": 0.2426897803892748,
+      "learning_rate": 3.8333925912828384e-05,
+      "loss": 0.1108,
+      "step": 1967
+    },
+    {
+      "epoch": 1.0244664237376366,
+      "grad_norm": 0.24168253147617336,
+      "learning_rate": 3.832203900219597e-05,
+      "loss": 0.1103,
+      "step": 1968
+    },
+    {
+      "epoch": 1.0249869859448204,
+      "grad_norm": 0.25236496660379437,
+      "learning_rate": 3.8310147883843684e-05,
+      "loss": 0.109,
+      "step": 1969
+    },
+    {
+      "epoch": 1.0255075481520042,
+      "grad_norm": 0.25308298024794296,
+      "learning_rate": 3.829825256152729e-05,
+      "loss": 0.1119,
+      "step": 1970
+    },
+    {
+      "epoch": 1.026028110359188,
+      "grad_norm": 0.256204183314334,
+      "learning_rate": 3.82863530390039e-05,
+      "loss": 0.1108,
+      "step": 1971
+    },
+    {
+      "epoch": 1.0265486725663717,
+      "grad_norm": 0.24156358649582654,
+      "learning_rate": 3.8274449320031926e-05,
+      "loss": 0.1084,
+      "step": 1972
+    },
+    {
+      "epoch": 1.0270692347735555,
+      "grad_norm": 0.24950618643926795,
+      "learning_rate": 3.826254140837111e-05,
+      "loss": 0.1087,
+      "step": 1973
+    },
+    {
+      "epoch": 1.0275897969807393,
+      "grad_norm": 0.2704998223408513,
+      "learning_rate": 3.8250629307782535e-05,
+      "loss": 0.1133,
+      "step": 1974
+    },
+    {
+      "epoch": 1.028110359187923,
+      "grad_norm": 0.25017914226933313,
+      "learning_rate": 3.8238713022028595e-05,
+      "loss": 0.1105,
+      "step": 1975
+    },
+    {
+      "epoch": 1.0286309213951068,
+      "grad_norm": 0.2327093355495663,
+      "learning_rate": 3.8226792554873004e-05,
+      "loss": 0.113,
+      "step": 1976
+    },
+    {
+      "epoch": 1.0291514836022904,
+      "grad_norm": 0.2719152461496494,
+      "learning_rate": 3.821486791008081e-05,
+      "loss": 0.1239,
+      "step": 1977
+    },
+    {
+      "epoch": 1.0296720458094741,
+      "grad_norm": 0.2397787036646458,
+      "learning_rate": 3.820293909141835e-05,
+      "loss": 0.1153,
+      "step": 1978
+    },
+    {
+      "epoch": 1.030192608016658,
+      "grad_norm": 0.24038853770587257,
+      "learning_rate": 3.819100610265332e-05,
+      "loss": 0.1184,
+      "step": 1979
+    },
+    {
+      "epoch": 1.0307131702238417,
+      "grad_norm": 0.24961313789239628,
+      "learning_rate": 3.8179068947554705e-05,
+      "loss": 0.1124,
+      "step": 1980
+    },
+    {
+      "epoch": 1.0312337324310255,
+      "grad_norm": 0.2382377547825293,
+      "learning_rate": 3.8167127629892815e-05,
+      "loss": 0.11,
+      "step": 1981
+    },
+    {
+      "epoch": 1.0317542946382092,
+      "grad_norm": 0.2532706658437514,
+      "learning_rate": 3.815518215343928e-05,
+      "loss": 0.1179,
+      "step": 1982
+    },
+    {
+      "epoch": 1.032274856845393,
+      "grad_norm": 0.2656733745200885,
+      "learning_rate": 3.8143232521967023e-05,
+      "loss": 0.1149,
+      "step": 1983
+    },
+    {
+      "epoch": 1.0327954190525768,
+      "grad_norm": 0.23840852166865217,
+      "learning_rate": 3.813127873925031e-05,
+      "loss": 0.1068,
+      "step": 1984
+    },
+    {
+      "epoch": 1.0333159812597605,
+      "grad_norm": 0.25521240664720873,
+      "learning_rate": 3.811932080906468e-05,
+      "loss": 0.1085,
+      "step": 1985
+    },
+    {
+      "epoch": 1.0338365434669443,
+      "grad_norm": 0.24469566160163492,
+      "learning_rate": 3.8107358735187036e-05,
+      "loss": 0.1128,
+      "step": 1986
+    },
+    {
+      "epoch": 1.034357105674128,
+      "grad_norm": 0.25281133665260414,
+      "learning_rate": 3.809539252139553e-05,
+      "loss": 0.1128,
+      "step": 1987
+    },
+    {
+      "epoch": 1.0348776678813119,
+      "grad_norm": 0.236637048209141,
+      "learning_rate": 3.8083422171469666e-05,
+      "loss": 0.104,
+      "step": 1988
+    },
+    {
+      "epoch": 1.0353982300884956,
+      "grad_norm": 0.2729698170542485,
+      "learning_rate": 3.807144768919022e-05,
+      "loss": 0.1153,
+      "step": 1989
+    },
+    {
+      "epoch": 1.0359187922956794,
+      "grad_norm": 0.26111532376784635,
+      "learning_rate": 3.8059469078339305e-05,
+      "loss": 0.1096,
+      "step": 1990
+    },
+    {
+      "epoch": 1.0364393545028632,
+      "grad_norm": 0.2528115929930001,
+      "learning_rate": 3.8047486342700314e-05,
+      "loss": 0.1117,
+      "step": 1991
+    },
+    {
+      "epoch": 1.036959916710047,
+      "grad_norm": 0.2614038022166877,
+      "learning_rate": 3.803549948605797e-05,
+      "loss": 0.1145,
+      "step": 1992
+    },
+    {
+      "epoch": 1.0374804789172307,
+      "grad_norm": 0.2584809095985116,
+      "learning_rate": 3.8023508512198256e-05,
+      "loss": 0.1125,
+      "step": 1993
+    },
+    {
+      "epoch": 1.0380010411244143,
+      "grad_norm": 0.2510595198139831,
+      "learning_rate": 3.8011513424908504e-05,
+      "loss": 0.1182,
+      "step": 1994
+    },
+    {
+      "epoch": 1.038521603331598,
+      "grad_norm": 0.2490798615545912,
+      "learning_rate": 3.7999514227977304e-05,
+      "loss": 0.1069,
+      "step": 1995
+    },
+    {
+      "epoch": 1.0390421655387818,
+      "grad_norm": 0.23506443265432583,
+      "learning_rate": 3.798751092519456e-05,
+      "loss": 0.1091,
+      "step": 1996
+    },
+    {
+      "epoch": 1.0395627277459656,
+      "grad_norm": 0.2542184882450879,
+      "learning_rate": 3.7975503520351487e-05,
+      "loss": 0.1168,
+      "step": 1997
+    },
+    {
+      "epoch": 1.0400832899531494,
+      "grad_norm": 0.2622709219446434,
+      "learning_rate": 3.796349201724058e-05,
+      "loss": 0.1121,
+      "step": 1998
+    },
+    {
+      "epoch": 1.0406038521603331,
+      "grad_norm": 0.24618731175514902,
+      "learning_rate": 3.795147641965561e-05,
+      "loss": 0.1143,
+      "step": 1999
+    },
+    {
+      "epoch": 1.041124414367517,
+      "grad_norm": 0.26095944678719163,
+      "learning_rate": 3.7939456731391684e-05,
+      "loss": 0.1157,
+      "step": 2000
+    },
+    {
+      "epoch": 1.0416449765747007,
+      "grad_norm": 0.24783987249021047,
+      "learning_rate": 3.792743295624517e-05,
+      "loss": 0.1168,
+      "step": 2001
+    },
+    {
+      "epoch": 1.0421655387818844,
+      "grad_norm": 0.2528026533974678,
+      "learning_rate": 3.791540509801373e-05,
+      "loss": 0.1114,
+      "step": 2002
+    },
+    {
+      "epoch": 1.0426861009890682,
+      "grad_norm": 0.23900993330244627,
+      "learning_rate": 3.7903373160496345e-05,
+      "loss": 0.1089,
+      "step": 2003
+    },
+    {
+      "epoch": 1.043206663196252,
+      "grad_norm": 0.2464392822573349,
+      "learning_rate": 3.789133714749323e-05,
+      "loss": 0.1115,
+      "step": 2004
+    },
+    {
+      "epoch": 1.0437272254034358,
+      "grad_norm": 0.26170131605706815,
+      "learning_rate": 3.787929706280594e-05,
+      "loss": 0.1181,
+      "step": 2005
+    },
+    {
+      "epoch": 1.0442477876106195,
+      "grad_norm": 0.2366142963015739,
+      "learning_rate": 3.786725291023728e-05,
+      "loss": 0.1084,
+      "step": 2006
+    },
+    {
+      "epoch": 1.0447683498178033,
+      "grad_norm": 0.266920747864476,
+      "learning_rate": 3.785520469359138e-05,
+      "loss": 0.1193,
+      "step": 2007
+    },
+    {
+      "epoch": 1.045288912024987,
+      "grad_norm": 0.27917270091917423,
+      "learning_rate": 3.784315241667359e-05,
+      "loss": 0.1216,
+      "step": 2008
+    },
+    {
+      "epoch": 1.0458094742321706,
+      "grad_norm": 0.2524069132942275,
+      "learning_rate": 3.7831096083290606e-05,
+      "loss": 0.1145,
+      "step": 2009
+    },
+    {
+      "epoch": 1.0463300364393544,
+      "grad_norm": 0.27357050117506915,
+      "learning_rate": 3.781903569725036e-05,
+      "loss": 0.1171,
+      "step": 2010
+    },
+    {
+      "epoch": 1.0468505986465382,
+      "grad_norm": 0.23696843787160707,
+      "learning_rate": 3.780697126236211e-05,
+      "loss": 0.1143,
+      "step": 2011
+    },
+    {
+      "epoch": 1.047371160853722,
+      "grad_norm": 0.23491005269928766,
+      "learning_rate": 3.779490278243634e-05,
+      "loss": 0.1082,
+      "step": 2012
+    },
+    {
+      "epoch": 1.0478917230609057,
+      "grad_norm": 0.24529859449463776,
+      "learning_rate": 3.778283026128485e-05,
+      "loss": 0.1138,
+      "step": 2013
+    },
+    {
+      "epoch": 1.0484122852680895,
+      "grad_norm": 0.24704977515279603,
+      "learning_rate": 3.7770753702720704e-05,
+      "loss": 0.1124,
+      "step": 2014
+    },
+    {
+      "epoch": 1.0489328474752733,
+      "grad_norm": 0.26691179513701707,
+      "learning_rate": 3.775867311055823e-05,
+      "loss": 0.1162,
+      "step": 2015
+    },
+    {
+      "epoch": 1.049453409682457,
+      "grad_norm": 0.2427635803168682,
+      "learning_rate": 3.7746588488613066e-05,
+      "loss": 0.1085,
+      "step": 2016
+    },
+    {
+      "epoch": 1.0499739718896408,
+      "grad_norm": 0.2551238836654876,
+      "learning_rate": 3.773449984070207e-05,
+      "loss": 0.11,
+      "step": 2017
+    },
+    {
+      "epoch": 1.0504945340968246,
+      "grad_norm": 0.24382085261536607,
+      "learning_rate": 3.7722407170643416e-05,
+      "loss": 0.1103,
+      "step": 2018
+    },
+    {
+      "epoch": 1.0510150963040084,
+      "grad_norm": 0.2349225844442972,
+      "learning_rate": 3.7710310482256526e-05,
+      "loss": 0.115,
+      "step": 2019
+    },
+    {
+      "epoch": 1.0515356585111921,
+      "grad_norm": 0.24160862206632094,
+      "learning_rate": 3.7698209779362105e-05,
+      "loss": 0.1111,
+      "step": 2020
+    },
+    {
+      "epoch": 1.052056220718376,
+      "grad_norm": 0.24412984723875725,
+      "learning_rate": 3.768610506578211e-05,
+      "loss": 0.1079,
+      "step": 2021
+    },
+    {
+      "epoch": 1.0525767829255597,
+      "grad_norm": 0.24330022038065835,
+      "learning_rate": 3.7673996345339765e-05,
+      "loss": 0.1101,
+      "step": 2022
+    },
+    {
+      "epoch": 1.0530973451327434,
+      "grad_norm": 0.26374011548874077,
+      "learning_rate": 3.7661883621859585e-05,
+      "loss": 0.1063,
+      "step": 2023
+    },
+    {
+      "epoch": 1.0536179073399272,
+      "grad_norm": 0.24698833556767213,
+      "learning_rate": 3.764976689916732e-05,
+      "loss": 0.1128,
+      "step": 2024
+    },
+    {
+      "epoch": 1.054138469547111,
+      "grad_norm": 0.2534386235100205,
+      "learning_rate": 3.7637646181090006e-05,
+      "loss": 0.1116,
+      "step": 2025
+    },
+    {
+      "epoch": 1.0546590317542945,
+      "grad_norm": 0.2698669420224393,
+      "learning_rate": 3.7625521471455914e-05,
+      "loss": 0.1146,
+      "step": 2026
+    },
+    {
+      "epoch": 1.0551795939614783,
+      "grad_norm": 0.2399199660805788,
+      "learning_rate": 3.76133927740946e-05,
+      "loss": 0.1147,
+      "step": 2027
+    },
+    {
+      "epoch": 1.055700156168662,
+      "grad_norm": 0.24762045246035963,
+      "learning_rate": 3.760126009283688e-05,
+      "loss": 0.1114,
+      "step": 2028
+    },
+    {
+      "epoch": 1.0562207183758459,
+      "grad_norm": 0.24458494730365296,
+      "learning_rate": 3.758912343151481e-05,
+      "loss": 0.1138,
+      "step": 2029
+    },
+    {
+      "epoch": 1.0567412805830296,
+      "grad_norm": 0.23831510708786413,
+      "learning_rate": 3.75769827939617e-05,
+      "loss": 0.1093,
+      "step": 2030
+    },
+    {
+      "epoch": 1.0572618427902134,
+      "grad_norm": 0.24160472062392452,
+      "learning_rate": 3.7564838184012155e-05,
+      "loss": 0.1113,
+      "step": 2031
+    },
+    {
+      "epoch": 1.0577824049973972,
+      "grad_norm": 0.25777500815863746,
+      "learning_rate": 3.755268960550199e-05,
+      "loss": 0.1126,
+      "step": 2032
+    },
+    {
+      "epoch": 1.058302967204581,
+      "grad_norm": 0.2469273261838509,
+      "learning_rate": 3.754053706226829e-05,
+      "loss": 0.1093,
+      "step": 2033
+    },
+    {
+      "epoch": 1.0588235294117647,
+      "grad_norm": 0.2403499464753065,
+      "learning_rate": 3.75283805581494e-05,
+      "loss": 0.1093,
+      "step": 2034
+    },
+    {
+      "epoch": 1.0593440916189485,
+      "grad_norm": 0.28349200510154626,
+      "learning_rate": 3.751622009698492e-05,
+      "loss": 0.1128,
+      "step": 2035
+    },
+    {
+      "epoch": 1.0598646538261323,
+      "grad_norm": 0.25717282657195967,
+      "learning_rate": 3.7504055682615676e-05,
+      "loss": 0.1128,
+      "step": 2036
+    },
+    {
+      "epoch": 1.060385216033316,
+      "grad_norm": 0.2604631820265362,
+      "learning_rate": 3.749188731888375e-05,
+      "loss": 0.1094,
+      "step": 2037
+    },
+    {
+      "epoch": 1.0609057782404998,
+      "grad_norm": 0.26157189061505526,
+      "learning_rate": 3.7479715009632486e-05,
+      "loss": 0.1152,
+      "step": 2038
+    },
+    {
+      "epoch": 1.0614263404476836,
+      "grad_norm": 0.2558378121103914,
+      "learning_rate": 3.7467538758706476e-05,
+      "loss": 0.1084,
+      "step": 2039
+    },
+    {
+      "epoch": 1.0619469026548674,
+      "grad_norm": 0.24072034047832436,
+      "learning_rate": 3.7455358569951535e-05,
+      "loss": 0.109,
+      "step": 2040
+    },
+    {
+      "epoch": 1.062467464862051,
+      "grad_norm": 0.2560891463466215,
+      "learning_rate": 3.744317444721473e-05,
+      "loss": 0.1097,
+      "step": 2041
+    },
+    {
+      "epoch": 1.0629880270692347,
+      "grad_norm": 0.2441562331368522,
+      "learning_rate": 3.743098639434438e-05,
+      "loss": 0.1126,
+      "step": 2042
+    },
+    {
+      "epoch": 1.0635085892764184,
+      "grad_norm": 0.25909422394913734,
+      "learning_rate": 3.7418794415190037e-05,
+      "loss": 0.1126,
+      "step": 2043
+    },
+    {
+      "epoch": 1.0640291514836022,
+      "grad_norm": 0.25192159362102,
+      "learning_rate": 3.74065985136025e-05,
+      "loss": 0.113,
+      "step": 2044
+    },
+    {
+      "epoch": 1.064549713690786,
+      "grad_norm": 0.24579868026393167,
+      "learning_rate": 3.73943986934338e-05,
+      "loss": 0.1136,
+      "step": 2045
+    },
+    {
+      "epoch": 1.0650702758979698,
+      "grad_norm": 0.25827104066791506,
+      "learning_rate": 3.738219495853721e-05,
+      "loss": 0.1161,
+      "step": 2046
+    },
+    {
+      "epoch": 1.0655908381051535,
+      "grad_norm": 0.23400282041641796,
+      "learning_rate": 3.736998731276722e-05,
+      "loss": 0.1115,
+      "step": 2047
+    },
+    {
+      "epoch": 1.0661114003123373,
+      "grad_norm": 0.26815101610132586,
+      "learning_rate": 3.7357775759979605e-05,
+      "loss": 0.1113,
+      "step": 2048
+    },
+    {
+      "epoch": 1.066631962519521,
+      "grad_norm": 0.24275941822819658,
+      "learning_rate": 3.734556030403131e-05,
+      "loss": 0.1115,
+      "step": 2049
+    },
+    {
+      "epoch": 1.0671525247267049,
+      "grad_norm": 0.2718950911696723,
+      "learning_rate": 3.733334094878057e-05,
+      "loss": 0.1155,
+      "step": 2050
+    },
+    {
+      "epoch": 1.0676730869338886,
+      "grad_norm": 0.2598394861210938,
+      "learning_rate": 3.73211176980868e-05,
+      "loss": 0.112,
+      "step": 2051
+    },
+    {
+      "epoch": 1.0681936491410724,
+      "grad_norm": 0.2694385440911815,
+      "learning_rate": 3.73088905558107e-05,
+      "loss": 0.1155,
+      "step": 2052
+    },
+    {
+      "epoch": 1.0687142113482562,
+      "grad_norm": 0.24106350307174138,
+      "learning_rate": 3.7296659525814146e-05,
+      "loss": 0.1106,
+      "step": 2053
+    },
+    {
+      "epoch": 1.06923477355544,
+      "grad_norm": 0.262198101864761,
+      "learning_rate": 3.7284424611960275e-05,
+      "loss": 0.1111,
+      "step": 2054
+    },
+    {
+      "epoch": 1.0697553357626237,
+      "grad_norm": 0.2546146585988391,
+      "learning_rate": 3.727218581811346e-05,
+      "loss": 0.1142,
+      "step": 2055
+    },
+    {
+      "epoch": 1.0702758979698075,
+      "grad_norm": 0.24247156568617306,
+      "learning_rate": 3.725994314813925e-05,
+      "loss": 0.1093,
+      "step": 2056
+    },
+    {
+      "epoch": 1.0707964601769913,
+      "grad_norm": 0.23958179241477287,
+      "learning_rate": 3.724769660590447e-05,
+      "loss": 0.1062,
+      "step": 2057
+    },
+    {
+      "epoch": 1.0713170223841748,
+      "grad_norm": 0.2396103519223926,
+      "learning_rate": 3.723544619527714e-05,
+      "loss": 0.1158,
+      "step": 2058
+    },
+    {
+      "epoch": 1.0718375845913586,
+      "grad_norm": 0.24141752361795885,
+      "learning_rate": 3.722319192012652e-05,
+      "loss": 0.1163,
+      "step": 2059
+    },
+    {
+      "epoch": 1.0723581467985424,
+      "grad_norm": 0.2530546244951208,
+      "learning_rate": 3.721093378432306e-05,
+      "loss": 0.1103,
+      "step": 2060
+    },
+    {
+      "epoch": 1.0728787090057261,
+      "grad_norm": 0.2304892834331245,
+      "learning_rate": 3.7198671791738475e-05,
+      "loss": 0.1138,
+      "step": 2061
+    },
+    {
+      "epoch": 1.07339927121291,
+      "grad_norm": 0.24116172581054715,
+      "learning_rate": 3.718640594624566e-05,
+      "loss": 0.1131,
+      "step": 2062
+    },
+    {
+      "epoch": 1.0739198334200937,
+      "grad_norm": 0.2514432482339406,
+      "learning_rate": 3.7174136251718736e-05,
+      "loss": 0.1147,
+      "step": 2063
+    },
+    {
+      "epoch": 1.0744403956272774,
+      "grad_norm": 0.24575627264010017,
+      "learning_rate": 3.716186271203305e-05,
+      "loss": 0.1112,
+      "step": 2064
+    },
+    {
+      "epoch": 1.0749609578344612,
+      "grad_norm": 0.2408037168496382,
+      "learning_rate": 3.714958533106515e-05,
+      "loss": 0.1094,
+      "step": 2065
+    },
+    {
+      "epoch": 1.075481520041645,
+      "grad_norm": 0.248061764942323,
+      "learning_rate": 3.713730411269282e-05,
+      "loss": 0.1165,
+      "step": 2066
+    },
+    {
+      "epoch": 1.0760020822488288,
+      "grad_norm": 0.2372533367139487,
+      "learning_rate": 3.7125019060795024e-05,
+      "loss": 0.1062,
+      "step": 2067
+    },
+    {
+      "epoch": 1.0765226444560125,
+      "grad_norm": 0.2383605045988699,
+      "learning_rate": 3.711273017925196e-05,
+      "loss": 0.1029,
+      "step": 2068
+    },
+    {
+      "epoch": 1.0770432066631963,
+      "grad_norm": 0.2609070795749889,
+      "learning_rate": 3.710043747194503e-05,
+      "loss": 0.1099,
+      "step": 2069
+    },
+    {
+      "epoch": 1.07756376887038,
+      "grad_norm": 0.25094133107875816,
+      "learning_rate": 3.708814094275683e-05,
+      "loss": 0.1104,
+      "step": 2070
+    },
+    {
+      "epoch": 1.0780843310775639,
+      "grad_norm": 0.2555320702860627,
+      "learning_rate": 3.70758405955712e-05,
+      "loss": 0.1118,
+      "step": 2071
+    },
+    {
+      "epoch": 1.0786048932847476,
+      "grad_norm": 0.2557360841701921,
+      "learning_rate": 3.7063536434273135e-05,
+      "loss": 0.1078,
+      "step": 2072
+    },
+    {
+      "epoch": 1.0791254554919312,
+      "grad_norm": 0.30210930503756817,
+      "learning_rate": 3.705122846274889e-05,
+      "loss": 0.1116,
+      "step": 2073
+    },
+    {
+      "epoch": 1.079646017699115,
+      "grad_norm": 0.283413324681584,
+      "learning_rate": 3.703891668488587e-05,
+      "loss": 0.1168,
+      "step": 2074
+    },
+    {
+      "epoch": 1.0801665799062987,
+      "grad_norm": 0.22741963249503896,
+      "learning_rate": 3.702660110457272e-05,
+      "loss": 0.1086,
+      "step": 2075
+    },
+    {
+      "epoch": 1.0806871421134825,
+      "grad_norm": 0.24697100297311528,
+      "learning_rate": 3.7014281725699276e-05,
+      "loss": 0.1138,
+      "step": 2076
+    },
+    {
+      "epoch": 1.0812077043206663,
+      "grad_norm": 0.2555266097287812,
+      "learning_rate": 3.700195855215656e-05,
+      "loss": 0.1094,
+      "step": 2077
+    },
+    {
+      "epoch": 1.08172826652785,
+      "grad_norm": 0.25611897463687516,
+      "learning_rate": 3.6989631587836814e-05,
+      "loss": 0.1131,
+      "step": 2078
+    },
+    {
+      "epoch": 1.0822488287350338,
+      "grad_norm": 0.24097390848814207,
+      "learning_rate": 3.6977300836633466e-05,
+      "loss": 0.1082,
+      "step": 2079
+    },
+    {
+      "epoch": 1.0827693909422176,
+      "grad_norm": 0.27090232208959586,
+      "learning_rate": 3.696496630244114e-05,
+      "loss": 0.113,
+      "step": 2080
+    },
+    {
+      "epoch": 1.0832899531494014,
+      "grad_norm": 0.2435755711000546,
+      "learning_rate": 3.695262798915564e-05,
+      "loss": 0.1082,
+      "step": 2081
+    },
+    {
+      "epoch": 1.0838105153565851,
+      "grad_norm": 0.2440321008013806,
+      "learning_rate": 3.694028590067401e-05,
+      "loss": 0.1088,
+      "step": 2082
+    },
+    {
+      "epoch": 1.084331077563769,
+      "grad_norm": 0.2523999632061997,
+      "learning_rate": 3.6927940040894424e-05,
+      "loss": 0.1123,
+      "step": 2083
+    },
+    {
+      "epoch": 1.0848516397709527,
+      "grad_norm": 0.25809490414472963,
+      "learning_rate": 3.691559041371631e-05,
+      "loss": 0.1156,
+      "step": 2084
+    },
+    {
+      "epoch": 1.0853722019781364,
+      "grad_norm": 0.2764366832810147,
+      "learning_rate": 3.6903237023040235e-05,
+      "loss": 0.1098,
+      "step": 2085
+    },
+    {
+      "epoch": 1.0858927641853202,
+      "grad_norm": 0.24467790732933922,
+      "learning_rate": 3.689087987276797e-05,
+      "loss": 0.1083,
+      "step": 2086
+    },
+    {
+      "epoch": 1.086413326392504,
+      "grad_norm": 0.2590589962094532,
+      "learning_rate": 3.68785189668025e-05,
+      "loss": 0.1142,
+      "step": 2087
+    },
+    {
+      "epoch": 1.0869338885996878,
+      "grad_norm": 0.23869572486032326,
+      "learning_rate": 3.686615430904795e-05,
+      "loss": 0.1141,
+      "step": 2088
+    },
+    {
+      "epoch": 1.0874544508068715,
+      "grad_norm": 0.24558579990983817,
+      "learning_rate": 3.685378590340968e-05,
+      "loss": 0.1086,
+      "step": 2089
+    },
+    {
+      "epoch": 1.087975013014055,
+      "grad_norm": 0.2451590372072089,
+      "learning_rate": 3.684141375379418e-05,
+      "loss": 0.1132,
+      "step": 2090
+    },
+    {
+      "epoch": 1.0884955752212389,
+      "grad_norm": 0.24564234860463044,
+      "learning_rate": 3.6829037864109176e-05,
+      "loss": 0.1109,
+      "step": 2091
+    },
+    {
+      "epoch": 1.0890161374284226,
+      "grad_norm": 0.26529150512656974,
+      "learning_rate": 3.6816658238263525e-05,
+      "loss": 0.1199,
+      "step": 2092
+    },
+    {
+      "epoch": 1.0895366996356064,
+      "grad_norm": 0.2585931700922823,
+      "learning_rate": 3.680427488016731e-05,
+      "loss": 0.1194,
+      "step": 2093
+    },
+    {
+      "epoch": 1.0900572618427902,
+      "grad_norm": 0.2464657000667077,
+      "learning_rate": 3.679188779373177e-05,
+      "loss": 0.1147,
+      "step": 2094
+    },
+    {
+      "epoch": 1.090577824049974,
+      "grad_norm": 0.2452855437235244,
+      "learning_rate": 3.677949698286931e-05,
+      "loss": 0.1149,
+      "step": 2095
+    },
+    {
+      "epoch": 1.0910983862571577,
+      "grad_norm": 0.2534553426066007,
+      "learning_rate": 3.676710245149353e-05,
+      "loss": 0.1171,
+      "step": 2096
+    },
+    {
+      "epoch": 1.0916189484643415,
+      "grad_norm": 0.23973694981857727,
+      "learning_rate": 3.675470420351921e-05,
+      "loss": 0.1158,
+      "step": 2097
+    },
+    {
+      "epoch": 1.0921395106715253,
+      "grad_norm": 0.24051783041794145,
+      "learning_rate": 3.6742302242862284e-05,
+      "loss": 0.1105,
+      "step": 2098
+    },
+    {
+      "epoch": 1.092660072878709,
+      "grad_norm": 0.24443405018149944,
+      "learning_rate": 3.6729896573439867e-05,
+      "loss": 0.1116,
+      "step": 2099
+    },
+    {
+      "epoch": 1.0931806350858928,
+      "grad_norm": 0.2597312632422425,
+      "learning_rate": 3.671748719917025e-05,
+      "loss": 0.1124,
+      "step": 2100
+    },
+    {
+      "epoch": 1.0937011972930766,
+      "grad_norm": 0.24822503893302403,
+      "learning_rate": 3.6705074123972885e-05,
+      "loss": 0.1096,
+      "step": 2101
+    },
+    {
+      "epoch": 1.0942217595002603,
+      "grad_norm": 0.24740697095379377,
+      "learning_rate": 3.669265735176842e-05,
+      "loss": 0.1083,
+      "step": 2102
+    },
+    {
+      "epoch": 1.0947423217074441,
+      "grad_norm": 0.2645450510720223,
+      "learning_rate": 3.668023688647863e-05,
+      "loss": 0.1126,
+      "step": 2103
+    },
+    {
+      "epoch": 1.095262883914628,
+      "grad_norm": 0.2535454890824587,
+      "learning_rate": 3.666781273202646e-05,
+      "loss": 0.1208,
+      "step": 2104
+    },
+    {
+      "epoch": 1.0957834461218114,
+      "grad_norm": 0.2575428275855921,
+      "learning_rate": 3.6655384892336075e-05,
+      "loss": 0.1168,
+      "step": 2105
+    },
+    {
+      "epoch": 1.0963040083289952,
+      "grad_norm": 0.2309286179235019,
+      "learning_rate": 3.664295337133274e-05,
+      "loss": 0.1124,
+      "step": 2106
+    },
+    {
+      "epoch": 1.096824570536179,
+      "grad_norm": 0.2512339973023405,
+      "learning_rate": 3.6630518172942915e-05,
+      "loss": 0.1151,
+      "step": 2107
+    },
+    {
+      "epoch": 1.0973451327433628,
+      "grad_norm": 0.2386785747665995,
+      "learning_rate": 3.6618079301094216e-05,
+      "loss": 0.1172,
+      "step": 2108
+    },
+    {
+      "epoch": 1.0978656949505465,
+      "grad_norm": 0.23724836929193713,
+      "learning_rate": 3.660563675971541e-05,
+      "loss": 0.1115,
+      "step": 2109
+    },
+    {
+      "epoch": 1.0983862571577303,
+      "grad_norm": 0.23600725505996434,
+      "learning_rate": 3.659319055273644e-05,
+      "loss": 0.1144,
+      "step": 2110
+    },
+    {
+      "epoch": 1.098906819364914,
+      "grad_norm": 0.23772868705388486,
+      "learning_rate": 3.6580740684088396e-05,
+      "loss": 0.113,
+      "step": 2111
+    },
+    {
+      "epoch": 1.0994273815720979,
+      "grad_norm": 0.237076578253062,
+      "learning_rate": 3.656828715770352e-05,
+      "loss": 0.1093,
+      "step": 2112
+    },
+    {
+      "epoch": 1.0999479437792816,
+      "grad_norm": 0.24382321669254056,
+      "learning_rate": 3.655582997751521e-05,
+      "loss": 0.1106,
+      "step": 2113
+    },
+    {
+      "epoch": 1.1004685059864654,
+      "grad_norm": 0.2377191561622143,
+      "learning_rate": 3.654336914745804e-05,
+      "loss": 0.1149,
+      "step": 2114
+    },
+    {
+      "epoch": 1.1009890681936492,
+      "grad_norm": 0.2490830706313632,
+      "learning_rate": 3.65309046714677e-05,
+      "loss": 0.1216,
+      "step": 2115
+    },
+    {
+      "epoch": 1.101509630400833,
+      "grad_norm": 0.23768207413624293,
+      "learning_rate": 3.651843655348107e-05,
+      "loss": 0.1116,
+      "step": 2116
+    },
+    {
+      "epoch": 1.1020301926080167,
+      "grad_norm": 0.23880298996606503,
+      "learning_rate": 3.650596479743616e-05,
+      "loss": 0.1111,
+      "step": 2117
+    },
+    {
+      "epoch": 1.1025507548152005,
+      "grad_norm": 0.241400343263037,
+      "learning_rate": 3.649348940727212e-05,
+      "loss": 0.1106,
+      "step": 2118
+    },
+    {
+      "epoch": 1.1030713170223843,
+      "grad_norm": 0.22973487630981618,
+      "learning_rate": 3.6481010386929264e-05,
+      "loss": 0.1107,
+      "step": 2119
+    },
+    {
+      "epoch": 1.103591879229568,
+      "grad_norm": 0.24451820553198225,
+      "learning_rate": 3.6468527740349045e-05,
+      "loss": 0.115,
+      "step": 2120
+    },
+    {
+      "epoch": 1.1041124414367518,
+      "grad_norm": 0.27311985417564316,
+      "learning_rate": 3.645604147147408e-05,
+      "loss": 0.1123,
+      "step": 2121
+    },
+    {
+      "epoch": 1.1046330036439354,
+      "grad_norm": 0.2438128553800211,
+      "learning_rate": 3.644355158424808e-05,
+      "loss": 0.1138,
+      "step": 2122
+    },
+    {
+      "epoch": 1.1051535658511191,
+      "grad_norm": 0.25561227871892656,
+      "learning_rate": 3.6431058082615964e-05,
+      "loss": 0.1236,
+      "step": 2123
+    },
+    {
+      "epoch": 1.105674128058303,
+      "grad_norm": 0.24292979052263167,
+      "learning_rate": 3.6418560970523745e-05,
+      "loss": 0.1126,
+      "step": 2124
+    },
+    {
+      "epoch": 1.1061946902654867,
+      "grad_norm": 0.23356756437347412,
+      "learning_rate": 3.6406060251918604e-05,
+      "loss": 0.1092,
+      "step": 2125
+    },
+    {
+      "epoch": 1.1067152524726704,
+      "grad_norm": 0.2510189996756236,
+      "learning_rate": 3.6393555930748846e-05,
+      "loss": 0.1134,
+      "step": 2126
+    },
+    {
+      "epoch": 1.1072358146798542,
+      "grad_norm": 0.2389541755449117,
+      "learning_rate": 3.63810480109639e-05,
+      "loss": 0.1101,
+      "step": 2127
+    },
+    {
+      "epoch": 1.107756376887038,
+      "grad_norm": 0.24947876309951245,
+      "learning_rate": 3.636853649651438e-05,
+      "loss": 0.1134,
+      "step": 2128
+    },
+    {
+      "epoch": 1.1082769390942218,
+      "grad_norm": 0.24327455652806065,
+      "learning_rate": 3.6356021391351976e-05,
+      "loss": 0.1128,
+      "step": 2129
+    },
+    {
+      "epoch": 1.1087975013014055,
+      "grad_norm": 0.25621361610147303,
+      "learning_rate": 3.634350269942956e-05,
+      "loss": 0.115,
+      "step": 2130
+    },
+    {
+      "epoch": 1.1093180635085893,
+      "grad_norm": 0.24171392174728737,
+      "learning_rate": 3.633098042470111e-05,
+      "loss": 0.1145,
+      "step": 2131
+    },
+    {
+      "epoch": 1.109838625715773,
+      "grad_norm": 0.24255117880153898,
+      "learning_rate": 3.631845457112174e-05,
+      "loss": 0.1118,
+      "step": 2132
+    },
+    {
+      "epoch": 1.1103591879229568,
+      "grad_norm": 0.2589821697017109,
+      "learning_rate": 3.63059251426477e-05,
+      "loss": 0.1118,
+      "step": 2133
+    },
+    {
+      "epoch": 1.1108797501301406,
+      "grad_norm": 0.2546318202945326,
+      "learning_rate": 3.6293392143236374e-05,
+      "loss": 0.1149,
+      "step": 2134
+    },
+    {
+      "epoch": 1.1114003123373244,
+      "grad_norm": 0.24797440602614088,
+      "learning_rate": 3.628085557684625e-05,
+      "loss": 0.1151,
+      "step": 2135
+    },
+    {
+      "epoch": 1.1119208745445082,
+      "grad_norm": 0.25596352809519574,
+      "learning_rate": 3.6268315447436976e-05,
+      "loss": 0.1104,
+      "step": 2136
+    },
+    {
+      "epoch": 1.1124414367516917,
+      "grad_norm": 0.24903661132125504,
+      "learning_rate": 3.6255771758969303e-05,
+      "loss": 0.1149,
+      "step": 2137
+    },
+    {
+      "epoch": 1.1129619989588755,
+      "grad_norm": 0.24389645955042044,
+      "learning_rate": 3.62432245154051e-05,
+      "loss": 0.1155,
+      "step": 2138
+    },
+    {
+      "epoch": 1.1134825611660593,
+      "grad_norm": 0.235781861888092,
+      "learning_rate": 3.6230673720707393e-05,
+      "loss": 0.112,
+      "step": 2139
+    },
+    {
+      "epoch": 1.114003123373243,
+      "grad_norm": 0.2308900894585995,
+      "learning_rate": 3.621811937884029e-05,
+      "loss": 0.1077,
+      "step": 2140
+    },
+    {
+      "epoch": 1.1145236855804268,
+      "grad_norm": 0.2753295219850416,
+      "learning_rate": 3.620556149376904e-05,
+      "loss": 0.1156,
+      "step": 2141
+    },
+    {
+      "epoch": 1.1150442477876106,
+      "grad_norm": 0.2470693560397974,
+      "learning_rate": 3.6193000069460006e-05,
+      "loss": 0.1096,
+      "step": 2142
+    },
+    {
+      "epoch": 1.1155648099947943,
+      "grad_norm": 0.24067635208687654,
+      "learning_rate": 3.618043510988068e-05,
+      "loss": 0.1102,
+      "step": 2143
+    },
+    {
+      "epoch": 1.1160853722019781,
+      "grad_norm": 0.2553351905414518,
+      "learning_rate": 3.616786661899965e-05,
+      "loss": 0.1135,
+      "step": 2144
+    },
+    {
+      "epoch": 1.116605934409162,
+      "grad_norm": 0.26177624105631986,
+      "learning_rate": 3.6155294600786625e-05,
+      "loss": 0.1082,
+      "step": 2145
+    },
+    {
+      "epoch": 1.1171264966163457,
+      "grad_norm": 0.2657931741696369,
+      "learning_rate": 3.6142719059212456e-05,
+      "loss": 0.1168,
+      "step": 2146
+    },
+    {
+      "epoch": 1.1176470588235294,
+      "grad_norm": 0.25488377144439966,
+      "learning_rate": 3.613013999824906e-05,
+      "loss": 0.113,
+      "step": 2147
+    },
+    {
+      "epoch": 1.1181676210307132,
+      "grad_norm": 0.2755475530717958,
+      "learning_rate": 3.6117557421869506e-05,
+      "loss": 0.1098,
+      "step": 2148
+    },
+    {
+      "epoch": 1.118688183237897,
+      "grad_norm": 0.25201657034289093,
+      "learning_rate": 3.6104971334047956e-05,
+      "loss": 0.1086,
+      "step": 2149
+    },
+    {
+      "epoch": 1.1192087454450808,
+      "grad_norm": 0.2357938145901859,
+      "learning_rate": 3.609238173875966e-05,
+      "loss": 0.1087,
+      "step": 2150
+    },
+    {
+      "epoch": 1.1197293076522645,
+      "grad_norm": 0.2540699466524578,
+      "learning_rate": 3.6079788639981036e-05,
+      "loss": 0.1143,
+      "step": 2151
+    },
+    {
+      "epoch": 1.1202498698594483,
+      "grad_norm": 0.24546722256380496,
+      "learning_rate": 3.606719204168954e-05,
+      "loss": 0.1117,
+      "step": 2152
+    },
+    {
+      "epoch": 1.120770432066632,
+      "grad_norm": 0.23557984732078704,
+      "learning_rate": 3.6054591947863784e-05,
+      "loss": 0.1108,
+      "step": 2153
+    },
+    {
+      "epoch": 1.1212909942738156,
+      "grad_norm": 0.24342719518637673,
+      "learning_rate": 3.604198836248344e-05,
+      "loss": 0.1106,
+      "step": 2154
+    },
+    {
+      "epoch": 1.1218115564809994,
+      "grad_norm": 0.24766981664804752,
+      "learning_rate": 3.602938128952933e-05,
+      "loss": 0.1145,
+      "step": 2155
+    },
+    {
+      "epoch": 1.1223321186881832,
+      "grad_norm": 0.25337709609875236,
+      "learning_rate": 3.6016770732983344e-05,
+      "loss": 0.1107,
+      "step": 2156
+    },
+    {
+      "epoch": 1.122852680895367,
+      "grad_norm": 0.2777055344080702,
+      "learning_rate": 3.600415669682849e-05,
+      "loss": 0.1171,
+      "step": 2157
+    },
+    {
+      "epoch": 1.1233732431025507,
+      "grad_norm": 0.24802200183807002,
+      "learning_rate": 3.599153918504886e-05,
+      "loss": 0.1148,
+      "step": 2158
+    },
+    {
+      "epoch": 1.1238938053097345,
+      "grad_norm": 0.249326912022627,
+      "learning_rate": 3.597891820162964e-05,
+      "loss": 0.114,
+      "step": 2159
+    },
+    {
+      "epoch": 1.1244143675169183,
+      "grad_norm": 0.24976539124029015,
+      "learning_rate": 3.596629375055716e-05,
+      "loss": 0.1108,
+      "step": 2160
+    },
+    {
+      "epoch": 1.124934929724102,
+      "grad_norm": 0.2549908652545141,
+      "learning_rate": 3.5953665835818765e-05,
+      "loss": 0.1161,
+      "step": 2161
+    },
+    {
+      "epoch": 1.1254554919312858,
+      "grad_norm": 0.2589331285801467,
+      "learning_rate": 3.594103446140297e-05,
+      "loss": 0.1144,
+      "step": 2162
+    },
+    {
+      "epoch": 1.1259760541384696,
+      "grad_norm": 0.24712526867777315,
+      "learning_rate": 3.592839963129934e-05,
+      "loss": 0.1202,
+      "step": 2163
+    },
+    {
+      "epoch": 1.1264966163456533,
+      "grad_norm": 0.252358949933663,
+      "learning_rate": 3.591576134949854e-05,
+      "loss": 0.1127,
+      "step": 2164
+    },
+    {
+      "epoch": 1.1270171785528371,
+      "grad_norm": 0.2357834325734446,
+      "learning_rate": 3.590311961999233e-05,
+      "loss": 0.1059,
+      "step": 2165
+    },
+    {
+      "epoch": 1.127537740760021,
+      "grad_norm": 0.25635650100122037,
+      "learning_rate": 3.589047444677355e-05,
+      "loss": 0.1154,
+      "step": 2166
+    },
+    {
+      "epoch": 1.1280583029672047,
+      "grad_norm": 0.28412878079518017,
+      "learning_rate": 3.587782583383615e-05,
+      "loss": 0.1173,
+      "step": 2167
+    },
+    {
+      "epoch": 1.1285788651743884,
+      "grad_norm": 0.26029129448235633,
+      "learning_rate": 3.586517378517514e-05,
+      "loss": 0.1131,
+      "step": 2168
+    },
+    {
+      "epoch": 1.129099427381572,
+      "grad_norm": 0.2592810117661404,
+      "learning_rate": 3.585251830478663e-05,
+      "loss": 0.1118,
+      "step": 2169
+    },
+    {
+      "epoch": 1.1296199895887558,
+      "grad_norm": 0.25743920274899534,
+      "learning_rate": 3.58398593966678e-05,
+      "loss": 0.1113,
+      "step": 2170
+    },
+    {
+      "epoch": 1.1301405517959395,
+      "grad_norm": 0.26024239886218314,
+      "learning_rate": 3.582719706481693e-05,
+      "loss": 0.1187,
+      "step": 2171
+    },
+    {
+      "epoch": 1.1306611140031233,
+      "grad_norm": 0.2534072347951432,
+      "learning_rate": 3.581453131323337e-05,
+      "loss": 0.1157,
+      "step": 2172
+    },
+    {
+      "epoch": 1.131181676210307,
+      "grad_norm": 0.2537355619987752,
+      "learning_rate": 3.580186214591756e-05,
+      "loss": 0.1095,
+      "step": 2173
+    },
+    {
+      "epoch": 1.1317022384174908,
+      "grad_norm": 0.23632427238631776,
+      "learning_rate": 3.578918956687101e-05,
+      "loss": 0.112,
+      "step": 2174
+    },
+    {
+      "epoch": 1.1322228006246746,
+      "grad_norm": 0.2412909737617186,
+      "learning_rate": 3.5776513580096315e-05,
+      "loss": 0.111,
+      "step": 2175
+    },
+    {
+      "epoch": 1.1327433628318584,
+      "grad_norm": 0.2483748166841977,
+      "learning_rate": 3.576383418959713e-05,
+      "loss": 0.1088,
+      "step": 2176
+    },
+    {
+      "epoch": 1.1332639250390422,
+      "grad_norm": 0.25025876140950926,
+      "learning_rate": 3.57511513993782e-05,
+      "loss": 0.1154,
+      "step": 2177
+    },
+    {
+      "epoch": 1.133784487246226,
+      "grad_norm": 0.24527483686430052,
+      "learning_rate": 3.5738465213445345e-05,
+      "loss": 0.1133,
+      "step": 2178
+    },
+    {
+      "epoch": 1.1343050494534097,
+      "grad_norm": 0.2400418074455827,
+      "learning_rate": 3.572577563580545e-05,
+      "loss": 0.1087,
+      "step": 2179
+    },
+    {
+      "epoch": 1.1348256116605935,
+      "grad_norm": 0.2473622001205245,
+      "learning_rate": 3.571308267046647e-05,
+      "loss": 0.1094,
+      "step": 2180
+    },
+    {
+      "epoch": 1.1353461738677773,
+      "grad_norm": 0.22926206403165528,
+      "learning_rate": 3.5700386321437446e-05,
+      "loss": 0.103,
+      "step": 2181
+    },
+    {
+      "epoch": 1.135866736074961,
+      "grad_norm": 0.2530508376636812,
+      "learning_rate": 3.5687686592728465e-05,
+      "loss": 0.1164,
+      "step": 2182
+    },
+    {
+      "epoch": 1.1363872982821448,
+      "grad_norm": 0.2547621831001377,
+      "learning_rate": 3.567498348835069e-05,
+      "loss": 0.1165,
+      "step": 2183
+    },
+    {
+      "epoch": 1.1369078604893286,
+      "grad_norm": 0.24555495677708386,
+      "learning_rate": 3.566227701231637e-05,
+      "loss": 0.1146,
+      "step": 2184
+    },
+    {
+      "epoch": 1.1374284226965123,
+      "grad_norm": 0.24522163832061175,
+      "learning_rate": 3.5649567168638786e-05,
+      "loss": 0.1094,
+      "step": 2185
+    },
+    {
+      "epoch": 1.1379489849036961,
+      "grad_norm": 0.24316629123363587,
+      "learning_rate": 3.56368539613323e-05,
+      "loss": 0.111,
+      "step": 2186
+    },
+    {
+      "epoch": 1.1384695471108797,
+      "grad_norm": 0.24441005322777853,
+      "learning_rate": 3.562413739441234e-05,
+      "loss": 0.1114,
+      "step": 2187
+    },
+    {
+      "epoch": 1.1389901093180634,
+      "grad_norm": 0.25037434739181624,
+      "learning_rate": 3.561141747189538e-05,
+      "loss": 0.1154,
+      "step": 2188
+    },
+    {
+      "epoch": 1.1395106715252472,
+      "grad_norm": 0.2561173459299334,
+      "learning_rate": 3.559869419779897e-05,
+      "loss": 0.1142,
+      "step": 2189
+    },
+    {
+      "epoch": 1.140031233732431,
+      "grad_norm": 0.23707593564275278,
+      "learning_rate": 3.558596757614172e-05,
+      "loss": 0.1099,
+      "step": 2190
+    },
+    {
+      "epoch": 1.1405517959396148,
+      "grad_norm": 0.2500778362026866,
+      "learning_rate": 3.5573237610943264e-05,
+      "loss": 0.1161,
+      "step": 2191
+    },
+    {
+      "epoch": 1.1410723581467985,
+      "grad_norm": 0.27492670789651996,
+      "learning_rate": 3.556050430622435e-05,
+      "loss": 0.1192,
+      "step": 2192
+    },
+    {
+      "epoch": 1.1415929203539823,
+      "grad_norm": 0.2605598668612606,
+      "learning_rate": 3.5547767666006735e-05,
+      "loss": 0.1132,
+      "step": 2193
+    },
+    {
+      "epoch": 1.142113482561166,
+      "grad_norm": 0.2605622258673461,
+      "learning_rate": 3.553502769431323e-05,
+      "loss": 0.114,
+      "step": 2194
+    },
+    {
+      "epoch": 1.1426340447683498,
+      "grad_norm": 0.2577447414552421,
+      "learning_rate": 3.5522284395167724e-05,
+      "loss": 0.1149,
+      "step": 2195
+    },
+    {
+      "epoch": 1.1431546069755336,
+      "grad_norm": 0.2728095113979058,
+      "learning_rate": 3.550953777259515e-05,
+      "loss": 0.1194,
+      "step": 2196
+    },
+    {
+      "epoch": 1.1436751691827174,
+      "grad_norm": 0.22575779206167243,
+      "learning_rate": 3.549678783062147e-05,
+      "loss": 0.1104,
+      "step": 2197
+    },
+    {
+      "epoch": 1.1441957313899012,
+      "grad_norm": 0.2551987577448281,
+      "learning_rate": 3.54840345732737e-05,
+      "loss": 0.1158,
+      "step": 2198
+    },
+    {
+      "epoch": 1.144716293597085,
+      "grad_norm": 0.24948023939005337,
+      "learning_rate": 3.547127800457994e-05,
+      "loss": 0.1139,
+      "step": 2199
+    },
+    {
+      "epoch": 1.1452368558042687,
+      "grad_norm": 0.260804602778483,
+      "learning_rate": 3.54585181285693e-05,
+      "loss": 0.1142,
+      "step": 2200
+    },
+    {
+      "epoch": 1.1457574180114523,
+      "grad_norm": 0.24626439137019285,
+      "learning_rate": 3.5445754949271924e-05,
+      "loss": 0.114,
+      "step": 2201
+    },
+    {
+      "epoch": 1.146277980218636,
+      "grad_norm": 0.23624464258694697,
+      "learning_rate": 3.543298847071904e-05,
+      "loss": 0.1101,
+      "step": 2202
+    },
+    {
+      "epoch": 1.1467985424258198,
+      "grad_norm": 0.2557734293730534,
+      "learning_rate": 3.542021869694289e-05,
+      "loss": 0.1176,
+      "step": 2203
+    },
+    {
+      "epoch": 1.1473191046330036,
+      "grad_norm": 0.25015139608483516,
+      "learning_rate": 3.5407445631976756e-05,
+      "loss": 0.1131,
+      "step": 2204
+    },
+    {
+      "epoch": 1.1478396668401873,
+      "grad_norm": 0.23610501538977566,
+      "learning_rate": 3.5394669279854966e-05,
+      "loss": 0.1153,
+      "step": 2205
+    },
+    {
+      "epoch": 1.1483602290473711,
+      "grad_norm": 0.22976718602322604,
+      "learning_rate": 3.53818896446129e-05,
+      "loss": 0.1184,
+      "step": 2206
+    },
+    {
+      "epoch": 1.148880791254555,
+      "grad_norm": 0.240542256545283,
+      "learning_rate": 3.536910673028695e-05,
+      "loss": 0.1121,
+      "step": 2207
+    },
+    {
+      "epoch": 1.1494013534617387,
+      "grad_norm": 0.23366916566434953,
+      "learning_rate": 3.5356320540914556e-05,
+      "loss": 0.1139,
+      "step": 2208
+    },
+    {
+      "epoch": 1.1499219156689224,
+      "grad_norm": 0.2277321141291514,
+      "learning_rate": 3.534353108053419e-05,
+      "loss": 0.1108,
+      "step": 2209
+    },
+    {
+      "epoch": 1.1504424778761062,
+      "grad_norm": 0.2245190916941551,
+      "learning_rate": 3.5330738353185364e-05,
+      "loss": 0.1106,
+      "step": 2210
+    },
+    {
+      "epoch": 1.15096304008329,
+      "grad_norm": 0.2441026791856793,
+      "learning_rate": 3.531794236290862e-05,
+      "loss": 0.1111,
+      "step": 2211
+    },
+    {
+      "epoch": 1.1514836022904738,
+      "grad_norm": 0.2693550306160406,
+      "learning_rate": 3.530514311374552e-05,
+      "loss": 0.1185,
+      "step": 2212
+    },
+    {
+      "epoch": 1.1520041644976575,
+      "grad_norm": 0.2583486674649021,
+      "learning_rate": 3.529234060973867e-05,
+      "loss": 0.1053,
+      "step": 2213
+    },
+    {
+      "epoch": 1.1525247267048413,
+      "grad_norm": 0.24969080269353847,
+      "learning_rate": 3.527953485493168e-05,
+      "loss": 0.1149,
+      "step": 2214
+    },
+    {
+      "epoch": 1.153045288912025,
+      "grad_norm": 0.29109726318240187,
+      "learning_rate": 3.526672585336923e-05,
+      "loss": 0.113,
+      "step": 2215
+    },
+    {
+      "epoch": 1.1535658511192088,
+      "grad_norm": 0.2342749727367158,
+      "learning_rate": 3.525391360909697e-05,
+      "loss": 0.109,
+      "step": 2216
+    },
+    {
+      "epoch": 1.1540864133263926,
+      "grad_norm": 0.2500797999992455,
+      "learning_rate": 3.524109812616161e-05,
+      "loss": 0.1144,
+      "step": 2217
+    },
+    {
+      "epoch": 1.1546069755335764,
+      "grad_norm": 0.2500104491421713,
+      "learning_rate": 3.5228279408610895e-05,
+      "loss": 0.1139,
+      "step": 2218
+    },
+    {
+      "epoch": 1.15512753774076,
+      "grad_norm": 0.23246173483256596,
+      "learning_rate": 3.521545746049356e-05,
+      "loss": 0.1107,
+      "step": 2219
+    },
+    {
+      "epoch": 1.1556480999479437,
+      "grad_norm": 0.25759769841196517,
+      "learning_rate": 3.5202632285859364e-05,
+      "loss": 0.115,
+      "step": 2220
+    },
+    {
+      "epoch": 1.1561686621551275,
+      "grad_norm": 0.2518042175741102,
+      "learning_rate": 3.518980388875911e-05,
+      "loss": 0.1126,
+      "step": 2221
+    },
+    {
+      "epoch": 1.1566892243623113,
+      "grad_norm": 0.24279674908930854,
+      "learning_rate": 3.517697227324459e-05,
+      "loss": 0.1125,
+      "step": 2222
+    },
+    {
+      "epoch": 1.157209786569495,
+      "grad_norm": 0.24501321952749994,
+      "learning_rate": 3.516413744336863e-05,
+      "loss": 0.1096,
+      "step": 2223
+    },
+    {
+      "epoch": 1.1577303487766788,
+      "grad_norm": 0.24613219825749424,
+      "learning_rate": 3.5151299403185075e-05,
+      "loss": 0.1119,
+      "step": 2224
+    },
+    {
+      "epoch": 1.1582509109838626,
+      "grad_norm": 0.2614423597655891,
+      "learning_rate": 3.513845815674877e-05,
+      "loss": 0.1139,
+      "step": 2225
+    },
+    {
+      "epoch": 1.1587714731910463,
+      "grad_norm": 0.24042769917249954,
+      "learning_rate": 3.512561370811556e-05,
+      "loss": 0.1088,
+      "step": 2226
+    },
+    {
+      "epoch": 1.1592920353982301,
+      "grad_norm": 0.25698514658916083,
+      "learning_rate": 3.5112766061342344e-05,
+      "loss": 0.112,
+      "step": 2227
+    },
+    {
+      "epoch": 1.1598125976054139,
+      "grad_norm": 0.27566732529397137,
+      "learning_rate": 3.5099915220487e-05,
+      "loss": 0.1123,
+      "step": 2228
+    },
+    {
+      "epoch": 1.1603331598125977,
+      "grad_norm": 0.2612613930290791,
+      "learning_rate": 3.5087061189608425e-05,
+      "loss": 0.1138,
+      "step": 2229
+    },
+    {
+      "epoch": 1.1608537220197814,
+      "grad_norm": 0.2609146607471247,
+      "learning_rate": 3.50742039727665e-05,
+      "loss": 0.1131,
+      "step": 2230
+    },
+    {
+      "epoch": 1.1613742842269652,
+      "grad_norm": 0.23452789382351913,
+      "learning_rate": 3.506134357402216e-05,
+      "loss": 0.1155,
+      "step": 2231
+    },
+    {
+      "epoch": 1.161894846434149,
+      "grad_norm": 0.25381787284216584,
+      "learning_rate": 3.504847999743729e-05,
+      "loss": 0.1169,
+      "step": 2232
+    },
+    {
+      "epoch": 1.1624154086413325,
+      "grad_norm": 0.2467153144538376,
+      "learning_rate": 3.503561324707484e-05,
+      "loss": 0.1113,
+      "step": 2233
+    },
+    {
+      "epoch": 1.1629359708485163,
+      "grad_norm": 0.2365562667950988,
+      "learning_rate": 3.50227433269987e-05,
+      "loss": 0.1152,
+      "step": 2234
+    },
+    {
+      "epoch": 1.1634565330557,
+      "grad_norm": 0.23867811869220612,
+      "learning_rate": 3.50098702412738e-05,
+      "loss": 0.1157,
+      "step": 2235
+    },
+    {
+      "epoch": 1.1639770952628838,
+      "grad_norm": 0.26099442663987993,
+      "learning_rate": 3.4996993993966057e-05,
+      "loss": 0.117,
+      "step": 2236
+    },
+    {
+      "epoch": 1.1644976574700676,
+      "grad_norm": 0.23935359957958538,
+      "learning_rate": 3.498411458914238e-05,
+      "loss": 0.1111,
+      "step": 2237
+    },
+    {
+      "epoch": 1.1650182196772514,
+      "grad_norm": 0.25105828547638254,
+      "learning_rate": 3.497123203087072e-05,
+      "loss": 0.1158,
+      "step": 2238
+    },
+    {
+      "epoch": 1.1655387818844352,
+      "grad_norm": 0.23331248980187663,
+      "learning_rate": 3.4958346323219946e-05,
+      "loss": 0.1037,
+      "step": 2239
+    },
+    {
+      "epoch": 1.166059344091619,
+      "grad_norm": 0.23447637588391998,
+      "learning_rate": 3.494545747025999e-05,
+      "loss": 0.1111,
+      "step": 2240
+    },
+    {
+      "epoch": 1.1665799062988027,
+      "grad_norm": 0.2526432724715221,
+      "learning_rate": 3.493256547606174e-05,
+      "loss": 0.1169,
+      "step": 2241
+    },
+    {
+      "epoch": 1.1671004685059865,
+      "grad_norm": 0.23593617679344556,
+      "learning_rate": 3.4919670344697085e-05,
+      "loss": 0.1119,
+      "step": 2242
+    },
+    {
+      "epoch": 1.1676210307131702,
+      "grad_norm": 0.24062003884271277,
+      "learning_rate": 3.490677208023892e-05,
+      "loss": 0.11,
+      "step": 2243
+    },
+    {
+      "epoch": 1.168141592920354,
+      "grad_norm": 0.2606487027308462,
+      "learning_rate": 3.489387068676111e-05,
+      "loss": 0.1188,
+      "step": 2244
+    },
+    {
+      "epoch": 1.1686621551275378,
+      "grad_norm": 0.23693423265149932,
+      "learning_rate": 3.4880966168338515e-05,
+      "loss": 0.1099,
+      "step": 2245
+    },
+    {
+      "epoch": 1.1691827173347216,
+      "grad_norm": 0.24304853176359928,
+      "learning_rate": 3.4868058529046985e-05,
+      "loss": 0.1194,
+      "step": 2246
+    },
+    {
+      "epoch": 1.1697032795419053,
+      "grad_norm": 0.25302863952962923,
+      "learning_rate": 3.485514777296335e-05,
+      "loss": 0.1136,
+      "step": 2247
+    },
+    {
+      "epoch": 1.170223841749089,
+      "grad_norm": 0.26167932489486967,
+      "learning_rate": 3.484223390416543e-05,
+      "loss": 0.113,
+      "step": 2248
+    },
+    {
+      "epoch": 1.1707444039562729,
+      "grad_norm": 0.2571613695259598,
+      "learning_rate": 3.4829316926732025e-05,
+      "loss": 0.1168,
+      "step": 2249
+    },
+    {
+      "epoch": 1.1712649661634567,
+      "grad_norm": 0.24293615449947745,
+      "learning_rate": 3.481639684474291e-05,
+      "loss": 0.1082,
+      "step": 2250
+    },
+    {
+      "epoch": 1.1717855283706402,
+      "grad_norm": 0.2473405511928646,
+      "learning_rate": 3.4803473662278865e-05,
+      "loss": 0.1122,
+      "step": 2251
+    },
+    {
+      "epoch": 1.172306090577824,
+      "grad_norm": 0.25163795289435603,
+      "learning_rate": 3.479054738342162e-05,
+      "loss": 0.1146,
+      "step": 2252
+    },
+    {
+      "epoch": 1.1728266527850078,
+      "grad_norm": 0.25041465835608206,
+      "learning_rate": 3.47776180122539e-05,
+      "loss": 0.1166,
+      "step": 2253
+    },
+    {
+      "epoch": 1.1733472149921915,
+      "grad_norm": 0.24547313540848606,
+      "learning_rate": 3.47646855528594e-05,
+      "loss": 0.1148,
+      "step": 2254
+    },
+    {
+      "epoch": 1.1738677771993753,
+      "grad_norm": 0.23742977846106733,
+      "learning_rate": 3.4751750009322795e-05,
+      "loss": 0.1063,
+      "step": 2255
+    },
+    {
+      "epoch": 1.174388339406559,
+      "grad_norm": 0.23708783235578984,
+      "learning_rate": 3.473881138572973e-05,
+      "loss": 0.1155,
+      "step": 2256
+    },
+    {
+      "epoch": 1.1749089016137428,
+      "grad_norm": 0.23348367213721333,
+      "learning_rate": 3.472586968616682e-05,
+      "loss": 0.1067,
+      "step": 2257
+    },
+    {
+      "epoch": 1.1754294638209266,
+      "grad_norm": 0.2502161259943564,
+      "learning_rate": 3.4712924914721664e-05,
+      "loss": 0.111,
+      "step": 2258
+    },
+    {
+      "epoch": 1.1759500260281104,
+      "grad_norm": 0.26009313181484484,
+      "learning_rate": 3.469997707548281e-05,
+      "loss": 0.117,
+      "step": 2259
+    },
+    {
+      "epoch": 1.1764705882352942,
+      "grad_norm": 0.23101072131923964,
+      "learning_rate": 3.468702617253981e-05,
+      "loss": 0.1092,
+      "step": 2260
+    },
+    {
+      "epoch": 1.176991150442478,
+      "grad_norm": 0.2546409766040149,
+      "learning_rate": 3.4674072209983144e-05,
+      "loss": 0.1142,
+      "step": 2261
+    },
+    {
+      "epoch": 1.1775117126496617,
+      "grad_norm": 0.23922221142024416,
+      "learning_rate": 3.466111519190428e-05,
+      "loss": 0.1107,
+      "step": 2262
+    },
+    {
+      "epoch": 1.1780322748568455,
+      "grad_norm": 0.244686444864269,
+      "learning_rate": 3.464815512239565e-05,
+      "loss": 0.1113,
+      "step": 2263
+    },
+    {
+      "epoch": 1.1785528370640292,
+      "grad_norm": 0.23856819223838555,
+      "learning_rate": 3.463519200555064e-05,
+      "loss": 0.1113,
+      "step": 2264
+    },
+    {
+      "epoch": 1.1790733992712128,
+      "grad_norm": 0.26366348156801095,
+      "learning_rate": 3.462222584546363e-05,
+      "loss": 0.1079,
+      "step": 2265
+    },
+    {
+      "epoch": 1.1795939614783966,
+      "grad_norm": 0.24664767010017477,
+      "learning_rate": 3.46092566462299e-05,
+      "loss": 0.1131,
+      "step": 2266
+    },
+    {
+      "epoch": 1.1801145236855803,
+      "grad_norm": 0.24616840057630468,
+      "learning_rate": 3.4596284411945755e-05,
+      "loss": 0.1173,
+      "step": 2267
+    },
+    {
+      "epoch": 1.1806350858927641,
+      "grad_norm": 0.25367146760865916,
+      "learning_rate": 3.4583309146708406e-05,
+      "loss": 0.1154,
+      "step": 2268
+    },
+    {
+      "epoch": 1.1811556480999479,
+      "grad_norm": 0.23507264194144847,
+      "learning_rate": 3.457033085461607e-05,
+      "loss": 0.1037,
+      "step": 2269
+    },
+    {
+      "epoch": 1.1816762103071317,
+      "grad_norm": 0.23989782007015442,
+      "learning_rate": 3.455734953976789e-05,
+      "loss": 0.1128,
+      "step": 2270
+    },
+    {
+      "epoch": 1.1821967725143154,
+      "grad_norm": 0.2405068672443556,
+      "learning_rate": 3.454436520626396e-05,
+      "loss": 0.1128,
+      "step": 2271
+    },
+    {
+      "epoch": 1.1827173347214992,
+      "grad_norm": 0.2301126062735549,
+      "learning_rate": 3.453137785820534e-05,
+      "loss": 0.1091,
+      "step": 2272
+    },
+    {
+      "epoch": 1.183237896928683,
+      "grad_norm": 0.25627883239774635,
+      "learning_rate": 3.451838749969404e-05,
+      "loss": 0.1088,
+      "step": 2273
+    },
+    {
+      "epoch": 1.1837584591358667,
+      "grad_norm": 0.23620278190068159,
+      "learning_rate": 3.450539413483302e-05,
+      "loss": 0.1144,
+      "step": 2274
+    },
+    {
+      "epoch": 1.1842790213430505,
+      "grad_norm": 0.25644395644621587,
+      "learning_rate": 3.4492397767726195e-05,
+      "loss": 0.1152,
+      "step": 2275
+    },
+    {
+      "epoch": 1.1847995835502343,
+      "grad_norm": 0.22348749190590206,
+      "learning_rate": 3.4479398402478406e-05,
+      "loss": 0.108,
+      "step": 2276
+    },
+    {
+      "epoch": 1.185320145757418,
+      "grad_norm": 0.2526453426350537,
+      "learning_rate": 3.4466396043195484e-05,
+      "loss": 0.1153,
+      "step": 2277
+    },
+    {
+      "epoch": 1.1858407079646018,
+      "grad_norm": 0.2562747166960448,
+      "learning_rate": 3.445339069398415e-05,
+      "loss": 0.1194,
+      "step": 2278
+    },
+    {
+      "epoch": 1.1863612701717856,
+      "grad_norm": 0.23680974190421952,
+      "learning_rate": 3.444038235895212e-05,
+      "loss": 0.11,
+      "step": 2279
+    },
+    {
+      "epoch": 1.1868818323789694,
+      "grad_norm": 0.2545225578398651,
+      "learning_rate": 3.442737104220801e-05,
+      "loss": 0.1118,
+      "step": 2280
+    },
+    {
+      "epoch": 1.1874023945861532,
+      "grad_norm": 0.2461192137870807,
+      "learning_rate": 3.441435674786143e-05,
+      "loss": 0.1104,
+      "step": 2281
+    },
+    {
+      "epoch": 1.187922956793337,
+      "grad_norm": 0.2538161714997669,
+      "learning_rate": 3.4401339480022874e-05,
+      "loss": 0.1118,
+      "step": 2282
+    },
+    {
+      "epoch": 1.1884435190005205,
+      "grad_norm": 0.25314168955138006,
+      "learning_rate": 3.4388319242803806e-05,
+      "loss": 0.1091,
+      "step": 2283
+    },
+    {
+      "epoch": 1.1889640812077042,
+      "grad_norm": 0.2587245929932049,
+      "learning_rate": 3.437529604031663e-05,
+      "loss": 0.1125,
+      "step": 2284
+    },
+    {
+      "epoch": 1.189484643414888,
+      "grad_norm": 0.24732781819326619,
+      "learning_rate": 3.436226987667467e-05,
+      "loss": 0.1095,
+      "step": 2285
+    },
+    {
+      "epoch": 1.1900052056220718,
+      "grad_norm": 0.26399681078473874,
+      "learning_rate": 3.4349240755992216e-05,
+      "loss": 0.1144,
+      "step": 2286
+    },
+    {
+      "epoch": 1.1905257678292556,
+      "grad_norm": 0.254509691032438,
+      "learning_rate": 3.433620868238444e-05,
+      "loss": 0.1195,
+      "step": 2287
+    },
+    {
+      "epoch": 1.1910463300364393,
+      "grad_norm": 0.27960613887789765,
+      "learning_rate": 3.4323173659967506e-05,
+      "loss": 0.1147,
+      "step": 2288
+    },
+    {
+      "epoch": 1.191566892243623,
+      "grad_norm": 0.23122617317358943,
+      "learning_rate": 3.431013569285846e-05,
+      "loss": 0.1082,
+      "step": 2289
+    },
+    {
+      "epoch": 1.1920874544508069,
+      "grad_norm": 0.2452337995607666,
+      "learning_rate": 3.429709478517531e-05,
+      "loss": 0.1121,
+      "step": 2290
+    },
+    {
+      "epoch": 1.1926080166579907,
+      "grad_norm": 0.24517649301196232,
+      "learning_rate": 3.428405094103696e-05,
+      "loss": 0.1145,
+      "step": 2291
+    },
+    {
+      "epoch": 1.1931285788651744,
+      "grad_norm": 0.24289572986099975,
+      "learning_rate": 3.42710041645633e-05,
+      "loss": 0.1138,
+      "step": 2292
+    },
+    {
+      "epoch": 1.1936491410723582,
+      "grad_norm": 0.26963497608876424,
+      "learning_rate": 3.425795445987508e-05,
+      "loss": 0.1186,
+      "step": 2293
+    },
+    {
+      "epoch": 1.194169703279542,
+      "grad_norm": 0.2525547241342057,
+      "learning_rate": 3.4244901831094014e-05,
+      "loss": 0.1145,
+      "step": 2294
+    },
+    {
+      "epoch": 1.1946902654867257,
+      "grad_norm": 0.25154080236686166,
+      "learning_rate": 3.4231846282342725e-05,
+      "loss": 0.1158,
+      "step": 2295
+    },
+    {
+      "epoch": 1.1952108276939095,
+      "grad_norm": 0.2607865126803791,
+      "learning_rate": 3.4218787817744773e-05,
+      "loss": 0.1145,
+      "step": 2296
+    },
+    {
+      "epoch": 1.195731389901093,
+      "grad_norm": 0.28235116430206414,
+      "learning_rate": 3.420572644142463e-05,
+      "loss": 0.1157,
+      "step": 2297
+    },
+    {
+      "epoch": 1.1962519521082768,
+      "grad_norm": 0.23089667791953067,
+      "learning_rate": 3.419266215750767e-05,
+      "loss": 0.1131,
+      "step": 2298
+    },
+    {
+      "epoch": 1.1967725143154606,
+      "grad_norm": 0.2293790099648737,
+      "learning_rate": 3.4179594970120215e-05,
+      "loss": 0.1151,
+      "step": 2299
+    },
+    {
+      "epoch": 1.1972930765226444,
+      "grad_norm": 0.24575677024213127,
+      "learning_rate": 3.4166524883389476e-05,
+      "loss": 0.1128,
+      "step": 2300
+    },
+    {
+      "epoch": 1.1978136387298282,
+      "grad_norm": 0.25216046527261354,
+      "learning_rate": 3.415345190144362e-05,
+      "loss": 0.1137,
+      "step": 2301
+    },
+    {
+      "epoch": 1.198334200937012,
+      "grad_norm": 0.24305569802868002,
+      "learning_rate": 3.414037602841168e-05,
+      "loss": 0.1147,
+      "step": 2302
+    },
+    {
+      "epoch": 1.1988547631441957,
+      "grad_norm": 0.27127800645249683,
+      "learning_rate": 3.4127297268423636e-05,
+      "loss": 0.1181,
+      "step": 2303
+    },
+    {
+      "epoch": 1.1993753253513795,
+      "grad_norm": 0.23242248281346625,
+      "learning_rate": 3.411421562561037e-05,
+      "loss": 0.1128,
+      "step": 2304
+    },
+    {
+      "epoch": 1.1998958875585632,
+      "grad_norm": 0.2745166695623737,
+      "learning_rate": 3.410113110410366e-05,
+      "loss": 0.1117,
+      "step": 2305
+    },
+    {
+      "epoch": 1.200416449765747,
+      "grad_norm": 0.23978776170340857,
+      "learning_rate": 3.408804370803623e-05,
+      "loss": 0.1108,
+      "step": 2306
+    },
+    {
+      "epoch": 1.2009370119729308,
+      "grad_norm": 0.24928731609898175,
+      "learning_rate": 3.407495344154167e-05,
+      "loss": 0.1161,
+      "step": 2307
+    },
+    {
+      "epoch": 1.2014575741801146,
+      "grad_norm": 0.23242055883453808,
+      "learning_rate": 3.40618603087545e-05,
+      "loss": 0.1114,
+      "step": 2308
+    },
+    {
+      "epoch": 1.2019781363872983,
+      "grad_norm": 0.2418508216693535,
+      "learning_rate": 3.404876431381014e-05,
+      "loss": 0.1097,
+      "step": 2309
+    },
+    {
+      "epoch": 1.202498698594482,
+      "grad_norm": 0.2644872421572009,
+      "learning_rate": 3.403566546084493e-05,
+      "loss": 0.1198,
+      "step": 2310
+    },
+    {
+      "epoch": 1.2030192608016659,
+      "grad_norm": 0.2432882408100227,
+      "learning_rate": 3.4022563753996075e-05,
+      "loss": 0.1161,
+      "step": 2311
+    },
+    {
+      "epoch": 1.2035398230088497,
+      "grad_norm": 0.25959575629843024,
+      "learning_rate": 3.400945919740171e-05,
+      "loss": 0.1136,
+      "step": 2312
+    },
+    {
+      "epoch": 1.2040603852160334,
+      "grad_norm": 0.2331391415291296,
+      "learning_rate": 3.3996351795200865e-05,
+      "loss": 0.1096,
+      "step": 2313
+    },
+    {
+      "epoch": 1.2045809474232172,
+      "grad_norm": 0.24943855765809428,
+      "learning_rate": 3.3983241551533465e-05,
+      "loss": 0.1091,
+      "step": 2314
+    },
+    {
+      "epoch": 1.2051015096304007,
+      "grad_norm": 0.24791662409738544,
+      "learning_rate": 3.397012847054035e-05,
+      "loss": 0.1148,
+      "step": 2315
+    },
+    {
+      "epoch": 1.2056220718375845,
+      "grad_norm": 0.2618458677286621,
+      "learning_rate": 3.3957012556363224e-05,
+      "loss": 0.1119,
+      "step": 2316
+    },
+    {
+      "epoch": 1.2061426340447683,
+      "grad_norm": 0.2635221612106524,
+      "learning_rate": 3.394389381314471e-05,
+      "loss": 0.112,
+      "step": 2317
+    },
+    {
+      "epoch": 1.206663196251952,
+      "grad_norm": 0.24840413289756066,
+      "learning_rate": 3.393077224502832e-05,
+      "loss": 0.1104,
+      "step": 2318
+    },
+    {
+      "epoch": 1.2071837584591358,
+      "grad_norm": 0.2504601503439982,
+      "learning_rate": 3.391764785615845e-05,
+      "loss": 0.1137,
+      "step": 2319
+    },
+    {
+      "epoch": 1.2077043206663196,
+      "grad_norm": 0.22826339447866847,
+      "learning_rate": 3.3904520650680405e-05,
+      "loss": 0.1084,
+      "step": 2320
+    },
+    {
+      "epoch": 1.2082248828735034,
+      "grad_norm": 0.23110036065955847,
+      "learning_rate": 3.3891390632740345e-05,
+      "loss": 0.1063,
+      "step": 2321
+    },
+    {
+      "epoch": 1.2087454450806872,
+      "grad_norm": 0.25396241886265036,
+      "learning_rate": 3.387825780648536e-05,
+      "loss": 0.1209,
+      "step": 2322
+    },
+    {
+      "epoch": 1.209266007287871,
+      "grad_norm": 0.2437164195108459,
+      "learning_rate": 3.386512217606339e-05,
+      "loss": 0.1202,
+      "step": 2323
+    },
+    {
+      "epoch": 1.2097865694950547,
+      "grad_norm": 0.23771865004331263,
+      "learning_rate": 3.38519837456233e-05,
+      "loss": 0.1127,
+      "step": 2324
+    },
+    {
+      "epoch": 1.2103071317022385,
+      "grad_norm": 0.22893673951370747,
+      "learning_rate": 3.383884251931481e-05,
+      "loss": 0.1098,
+      "step": 2325
+    },
+    {
+      "epoch": 1.2108276939094222,
+      "grad_norm": 0.24464255415023003,
+      "learning_rate": 3.3825698501288516e-05,
+      "loss": 0.1114,
+      "step": 2326
+    },
+    {
+      "epoch": 1.211348256116606,
+      "grad_norm": 0.24641719397311446,
+      "learning_rate": 3.381255169569594e-05,
+      "loss": 0.1163,
+      "step": 2327
+    },
+    {
+      "epoch": 1.2118688183237898,
+      "grad_norm": 0.24202099555940623,
+      "learning_rate": 3.3799402106689416e-05,
+      "loss": 0.1122,
+      "step": 2328
+    },
+    {
+      "epoch": 1.2123893805309733,
+      "grad_norm": 0.23976840972172184,
+      "learning_rate": 3.3786249738422235e-05,
+      "loss": 0.1138,
+      "step": 2329
+    },
+    {
+      "epoch": 1.212909942738157,
+      "grad_norm": 0.26118454788338863,
+      "learning_rate": 3.3773094595048506e-05,
+      "loss": 0.1105,
+      "step": 2330
+    },
+    {
+      "epoch": 1.2134305049453409,
+      "grad_norm": 0.2504712191955974,
+      "learning_rate": 3.375993668072324e-05,
+      "loss": 0.1124,
+      "step": 2331
+    },
+    {
+      "epoch": 1.2139510671525247,
+      "grad_norm": 0.2325507565735545,
+      "learning_rate": 3.374677599960231e-05,
+      "loss": 0.1098,
+      "step": 2332
+    },
+    {
+      "epoch": 1.2144716293597084,
+      "grad_norm": 0.2590513905450211,
+      "learning_rate": 3.3733612555842486e-05,
+      "loss": 0.1124,
+      "step": 2333
+    },
+    {
+      "epoch": 1.2149921915668922,
+      "grad_norm": 0.24603883688487388,
+      "learning_rate": 3.3720446353601394e-05,
+      "loss": 0.1122,
+      "step": 2334
+    },
+    {
+      "epoch": 1.215512753774076,
+      "grad_norm": 0.2377595048373133,
+      "learning_rate": 3.370727739703752e-05,
+      "loss": 0.1124,
+      "step": 2335
+    },
+    {
+      "epoch": 1.2160333159812597,
+      "grad_norm": 0.23641590886231106,
+      "learning_rate": 3.369410569031024e-05,
+      "loss": 0.1114,
+      "step": 2336
+    },
+    {
+      "epoch": 1.2165538781884435,
+      "grad_norm": 0.2498625675970159,
+      "learning_rate": 3.36809312375798e-05,
+      "loss": 0.1145,
+      "step": 2337
+    },
+    {
+      "epoch": 1.2170744403956273,
+      "grad_norm": 0.22842565347219712,
+      "learning_rate": 3.36677540430073e-05,
+      "loss": 0.1129,
+      "step": 2338
+    },
+    {
+      "epoch": 1.217595002602811,
+      "grad_norm": 0.22898607545193125,
+      "learning_rate": 3.365457411075471e-05,
+      "loss": 0.1127,
+      "step": 2339
+    },
+    {
+      "epoch": 1.2181155648099948,
+      "grad_norm": 0.23953214969485537,
+      "learning_rate": 3.3641391444984864e-05,
+      "loss": 0.1112,
+      "step": 2340
+    },
+    {
+      "epoch": 1.2186361270171786,
+      "grad_norm": 0.24104626520668057,
+      "learning_rate": 3.362820604986147e-05,
+      "loss": 0.1125,
+      "step": 2341
+    },
+    {
+      "epoch": 1.2191566892243624,
+      "grad_norm": 0.23748691484838422,
+      "learning_rate": 3.361501792954908e-05,
+      "loss": 0.1095,
+      "step": 2342
+    },
+    {
+      "epoch": 1.2196772514315461,
+      "grad_norm": 0.2483740020552502,
+      "learning_rate": 3.360182708821312e-05,
+      "loss": 0.1158,
+      "step": 2343
+    },
+    {
+      "epoch": 1.22019781363873,
+      "grad_norm": 0.25425874287469785,
+      "learning_rate": 3.358863353001987e-05,
+      "loss": 0.1225,
+      "step": 2344
+    },
+    {
+      "epoch": 1.2207183758459137,
+      "grad_norm": 0.2586255317538337,
+      "learning_rate": 3.3575437259136474e-05,
+      "loss": 0.1164,
+      "step": 2345
+    },
+    {
+      "epoch": 1.2212389380530975,
+      "grad_norm": 0.23402320390581927,
+      "learning_rate": 3.3562238279730916e-05,
+      "loss": 0.1147,
+      "step": 2346
+    },
+    {
+      "epoch": 1.221759500260281,
+      "grad_norm": 0.2477059723049737,
+      "learning_rate": 3.354903659597207e-05,
+      "loss": 0.1115,
+      "step": 2347
+    },
+    {
+      "epoch": 1.2222800624674648,
+      "grad_norm": 0.23650878937783718,
+      "learning_rate": 3.353583221202962e-05,
+      "loss": 0.1067,
+      "step": 2348
+    },
+    {
+      "epoch": 1.2228006246746486,
+      "grad_norm": 0.24360478631058402,
+      "learning_rate": 3.352262513207413e-05,
+      "loss": 0.1113,
+      "step": 2349
+    },
+    {
+      "epoch": 1.2233211868818323,
+      "grad_norm": 0.239918587840492,
+      "learning_rate": 3.350941536027702e-05,
+      "loss": 0.1103,
+      "step": 2350
+    },
+    {
+      "epoch": 1.223841749089016,
+      "grad_norm": 0.23335823537471334,
+      "learning_rate": 3.349620290081055e-05,
+      "loss": 0.107,
+      "step": 2351
+    },
+    {
+      "epoch": 1.2243623112961999,
+      "grad_norm": 0.23023317202230778,
+      "learning_rate": 3.348298775784782e-05,
+      "loss": 0.1139,
+      "step": 2352
+    },
+    {
+      "epoch": 1.2248828735033837,
+      "grad_norm": 0.24137499470962637,
+      "learning_rate": 3.3469769935562796e-05,
+      "loss": 0.1111,
+      "step": 2353
+    },
+    {
+      "epoch": 1.2254034357105674,
+      "grad_norm": 0.23984175856275775,
+      "learning_rate": 3.345654943813027e-05,
+      "loss": 0.1126,
+      "step": 2354
+    },
+    {
+      "epoch": 1.2259239979177512,
+      "grad_norm": 0.2511343023919521,
+      "learning_rate": 3.34433262697259e-05,
+      "loss": 0.1235,
+      "step": 2355
+    },
+    {
+      "epoch": 1.226444560124935,
+      "grad_norm": 0.2248493142834007,
+      "learning_rate": 3.343010043452618e-05,
+      "loss": 0.11,
+      "step": 2356
+    },
+    {
+      "epoch": 1.2269651223321187,
+      "grad_norm": 0.23619956004612885,
+      "learning_rate": 3.3416871936708436e-05,
+      "loss": 0.1122,
+      "step": 2357
+    },
+    {
+      "epoch": 1.2274856845393025,
+      "grad_norm": 0.24154490816188087,
+      "learning_rate": 3.340364078045085e-05,
+      "loss": 0.118,
+      "step": 2358
+    },
+    {
+      "epoch": 1.2280062467464863,
+      "grad_norm": 0.23906919393439444,
+      "learning_rate": 3.339040696993243e-05,
+      "loss": 0.112,
+      "step": 2359
+    },
+    {
+      "epoch": 1.22852680895367,
+      "grad_norm": 0.23515780174889195,
+      "learning_rate": 3.3377170509333035e-05,
+      "loss": 0.1065,
+      "step": 2360
+    },
+    {
+      "epoch": 1.2290473711608536,
+      "grad_norm": 0.2448541840615606,
+      "learning_rate": 3.336393140283335e-05,
+      "loss": 0.1104,
+      "step": 2361
+    },
+    {
+      "epoch": 1.2295679333680374,
+      "grad_norm": 0.23502697559009228,
+      "learning_rate": 3.33506896546149e-05,
+      "loss": 0.1168,
+      "step": 2362
+    },
+    {
+      "epoch": 1.2300884955752212,
+      "grad_norm": 0.24547086115985342,
+      "learning_rate": 3.333744526886006e-05,
+      "loss": 0.1178,
+      "step": 2363
+    },
+    {
+      "epoch": 1.230609057782405,
+      "grad_norm": 0.2238019166845714,
+      "learning_rate": 3.3324198249752004e-05,
+      "loss": 0.1089,
+      "step": 2364
+    },
+    {
+      "epoch": 1.2311296199895887,
+      "grad_norm": 0.25163269197743565,
+      "learning_rate": 3.331094860147477e-05,
+      "loss": 0.1149,
+      "step": 2365
+    },
+    {
+      "epoch": 1.2316501821967725,
+      "grad_norm": 0.24302067777188527,
+      "learning_rate": 3.329769632821321e-05,
+      "loss": 0.1126,
+      "step": 2366
+    },
+    {
+      "epoch": 1.2321707444039562,
+      "grad_norm": 0.24052489202021649,
+      "learning_rate": 3.328444143415301e-05,
+      "loss": 0.1126,
+      "step": 2367
+    },
+    {
+      "epoch": 1.23269130661114,
+      "grad_norm": 0.23963819746590062,
+      "learning_rate": 3.327118392348068e-05,
+      "loss": 0.1111,
+      "step": 2368
+    },
+    {
+      "epoch": 1.2332118688183238,
+      "grad_norm": 0.2554674901122427,
+      "learning_rate": 3.325792380038356e-05,
+      "loss": 0.122,
+      "step": 2369
+    },
+    {
+      "epoch": 1.2337324310255076,
+      "grad_norm": 0.24101776937698122,
+      "learning_rate": 3.324466106904981e-05,
+      "loss": 0.1108,
+      "step": 2370
+    },
+    {
+      "epoch": 1.2342529932326913,
+      "grad_norm": 0.23475211548593766,
+      "learning_rate": 3.323139573366842e-05,
+      "loss": 0.1143,
+      "step": 2371
+    },
+    {
+      "epoch": 1.234773555439875,
+      "grad_norm": 0.23437107317802344,
+      "learning_rate": 3.32181277984292e-05,
+      "loss": 0.109,
+      "step": 2372
+    },
+    {
+      "epoch": 1.2352941176470589,
+      "grad_norm": 0.23380076784273857,
+      "learning_rate": 3.320485726752278e-05,
+      "loss": 0.1119,
+      "step": 2373
+    },
+    {
+      "epoch": 1.2358146798542426,
+      "grad_norm": 0.2540138119726505,
+      "learning_rate": 3.3191584145140626e-05,
+      "loss": 0.1171,
+      "step": 2374
+    },
+    {
+      "epoch": 1.2363352420614264,
+      "grad_norm": 0.2411702497970775,
+      "learning_rate": 3.317830843547499e-05,
+      "loss": 0.1161,
+      "step": 2375
+    },
+    {
+      "epoch": 1.2368558042686102,
+      "grad_norm": 0.233210046923344,
+      "learning_rate": 3.316503014271896e-05,
+      "loss": 0.1103,
+      "step": 2376
+    },
+    {
+      "epoch": 1.237376366475794,
+      "grad_norm": 0.24776647975273905,
+      "learning_rate": 3.3151749271066444e-05,
+      "loss": 0.1155,
+      "step": 2377
+    },
+    {
+      "epoch": 1.2378969286829777,
+      "grad_norm": 0.2273242590925185,
+      "learning_rate": 3.3138465824712164e-05,
+      "loss": 0.1124,
+      "step": 2378
+    },
+    {
+      "epoch": 1.2384174908901613,
+      "grad_norm": 0.2394314835262025,
+      "learning_rate": 3.312517980785164e-05,
+      "loss": 0.1128,
+      "step": 2379
+    },
+    {
+      "epoch": 1.238938053097345,
+      "grad_norm": 0.2448659548311951,
+      "learning_rate": 3.311189122468122e-05,
+      "loss": 0.1196,
+      "step": 2380
+    },
+    {
+      "epoch": 1.2394586153045288,
+      "grad_norm": 0.23448498331985973,
+      "learning_rate": 3.309860007939806e-05,
+      "loss": 0.1116,
+      "step": 2381
+    },
+    {
+      "epoch": 1.2399791775117126,
+      "grad_norm": 0.22911095022022188,
+      "learning_rate": 3.308530637620011e-05,
+      "loss": 0.1121,
+      "step": 2382
+    },
+    {
+      "epoch": 1.2404997397188964,
+      "grad_norm": 0.24025893459427192,
+      "learning_rate": 3.307201011928616e-05,
+      "loss": 0.1144,
+      "step": 2383
+    },
+    {
+      "epoch": 1.2410203019260801,
+      "grad_norm": 0.23259241617711043,
+      "learning_rate": 3.305871131285577e-05,
+      "loss": 0.1094,
+      "step": 2384
+    },
+    {
+      "epoch": 1.241540864133264,
+      "grad_norm": 0.2361852244544814,
+      "learning_rate": 3.3045409961109324e-05,
+      "loss": 0.1071,
+      "step": 2385
+    },
+    {
+      "epoch": 1.2420614263404477,
+      "grad_norm": 0.2561842268409719,
+      "learning_rate": 3.3032106068248014e-05,
+      "loss": 0.1198,
+      "step": 2386
+    },
+    {
+      "epoch": 1.2425819885476315,
+      "grad_norm": 0.2511394825375184,
+      "learning_rate": 3.301879963847383e-05,
+      "loss": 0.1106,
+      "step": 2387
+    },
+    {
+      "epoch": 1.2431025507548152,
+      "grad_norm": 0.22935577553554087,
+      "learning_rate": 3.3005490675989545e-05,
+      "loss": 0.1103,
+      "step": 2388
+    },
+    {
+      "epoch": 1.243623112961999,
+      "grad_norm": 0.24634598430303592,
+      "learning_rate": 3.2992179184998774e-05,
+      "loss": 0.1118,
+      "step": 2389
+    },
+    {
+      "epoch": 1.2441436751691828,
+      "grad_norm": 0.24037850632563773,
+      "learning_rate": 3.2978865169705885e-05,
+      "loss": 0.115,
+      "step": 2390
+    },
+    {
+      "epoch": 1.2446642373763666,
+      "grad_norm": 0.24026168911843124,
+      "learning_rate": 3.296554863431607e-05,
+      "loss": 0.1093,
+      "step": 2391
+    },
+    {
+      "epoch": 1.2451847995835503,
+      "grad_norm": 0.24022065182744629,
+      "learning_rate": 3.295222958303532e-05,
+      "loss": 0.1115,
+      "step": 2392
+    },
+    {
+      "epoch": 1.2457053617907339,
+      "grad_norm": 0.23185501118099175,
+      "learning_rate": 3.2938908020070404e-05,
+      "loss": 0.1101,
+      "step": 2393
+    },
+    {
+      "epoch": 1.2462259239979177,
+      "grad_norm": 0.23099404002234822,
+      "learning_rate": 3.292558394962888e-05,
+      "loss": 0.1132,
+      "step": 2394
+    },
+    {
+      "epoch": 1.2467464862051014,
+      "grad_norm": 0.24176376810983632,
+      "learning_rate": 3.2912257375919126e-05,
+      "loss": 0.1148,
+      "step": 2395
+    },
+    {
+      "epoch": 1.2472670484122852,
+      "grad_norm": 0.24222391889430964,
+      "learning_rate": 3.289892830315028e-05,
+      "loss": 0.1098,
+      "step": 2396
+    },
+    {
+      "epoch": 1.247787610619469,
+      "grad_norm": 0.23277468724181424,
+      "learning_rate": 3.28855967355323e-05,
+      "loss": 0.1081,
+      "step": 2397
+    },
+    {
+      "epoch": 1.2483081728266527,
+      "grad_norm": 0.2319208224863446,
+      "learning_rate": 3.2872262677275906e-05,
+      "loss": 0.1127,
+      "step": 2398
+    },
+    {
+      "epoch": 1.2488287350338365,
+      "grad_norm": 0.22600332372977186,
+      "learning_rate": 3.285892613259261e-05,
+      "loss": 0.1082,
+      "step": 2399
+    },
+    {
+      "epoch": 1.2493492972410203,
+      "grad_norm": 0.2380744039289219,
+      "learning_rate": 3.2845587105694716e-05,
+      "loss": 0.1157,
+      "step": 2400
+    },
+    {
+      "epoch": 1.249869859448204,
+      "grad_norm": 0.24199537993154474,
+      "learning_rate": 3.283224560079532e-05,
+      "loss": 0.1085,
+      "step": 2401
+    },
+    {
+      "epoch": 1.2503904216553878,
+      "grad_norm": 0.25841375982911335,
+      "learning_rate": 3.281890162210829e-05,
+      "loss": 0.1129,
+      "step": 2402
+    },
+    {
+      "epoch": 1.2509109838625716,
+      "grad_norm": 0.23379214223234784,
+      "learning_rate": 3.2805555173848254e-05,
+      "loss": 0.11,
+      "step": 2403
+    },
+    {
+      "epoch": 1.2514315460697554,
+      "grad_norm": 0.24808893501537926,
+      "learning_rate": 3.279220626023065e-05,
+      "loss": 0.1103,
+      "step": 2404
+    },
+    {
+      "epoch": 1.2519521082769391,
+      "grad_norm": 0.23963781789500202,
+      "learning_rate": 3.27788548854717e-05,
+      "loss": 0.1126,
+      "step": 2405
+    },
+    {
+      "epoch": 1.252472670484123,
+      "grad_norm": 0.2536793830039519,
+      "learning_rate": 3.276550105378838e-05,
+      "loss": 0.1188,
+      "step": 2406
+    },
+    {
+      "epoch": 1.2529932326913067,
+      "grad_norm": 0.22592288227783439,
+      "learning_rate": 3.275214476939845e-05,
+      "loss": 0.1111,
+      "step": 2407
+    },
+    {
+      "epoch": 1.2535137948984905,
+      "grad_norm": 0.2184518861484896,
+      "learning_rate": 3.273878603652045e-05,
+      "loss": 0.1101,
+      "step": 2408
+    },
+    {
+      "epoch": 1.2540343571056742,
+      "grad_norm": 0.22611064467072162,
+      "learning_rate": 3.272542485937369e-05,
+      "loss": 0.1078,
+      "step": 2409
+    },
+    {
+      "epoch": 1.254554919312858,
+      "grad_norm": 0.24252402472546458,
+      "learning_rate": 3.271206124217825e-05,
+      "loss": 0.1089,
+      "step": 2410
+    },
+    {
+      "epoch": 1.2550754815200418,
+      "grad_norm": 0.24827439034038407,
+      "learning_rate": 3.269869518915497e-05,
+      "loss": 0.1153,
+      "step": 2411
+    },
+    {
+      "epoch": 1.2555960437272253,
+      "grad_norm": 0.24447695733165611,
+      "learning_rate": 3.268532670452549e-05,
+      "loss": 0.111,
+      "step": 2412
+    },
+    {
+      "epoch": 1.256116605934409,
+      "grad_norm": 0.24933714928678632,
+      "learning_rate": 3.2671955792512186e-05,
+      "loss": 0.1113,
+      "step": 2413
+    },
+    {
+      "epoch": 1.2566371681415929,
+      "grad_norm": 0.2378454669519632,
+      "learning_rate": 3.265858245733824e-05,
+      "loss": 0.1152,
+      "step": 2414
+    },
+    {
+      "epoch": 1.2571577303487766,
+      "grad_norm": 0.2512478037323881,
+      "learning_rate": 3.2645206703227536e-05,
+      "loss": 0.1206,
+      "step": 2415
+    },
+    {
+      "epoch": 1.2576782925559604,
+      "grad_norm": 0.250096683928609,
+      "learning_rate": 3.263182853440479e-05,
+      "loss": 0.1135,
+      "step": 2416
+    },
+    {
+      "epoch": 1.2581988547631442,
+      "grad_norm": 0.23599764911190643,
+      "learning_rate": 3.261844795509542e-05,
+      "loss": 0.116,
+      "step": 2417
+    },
+    {
+      "epoch": 1.258719416970328,
+      "grad_norm": 0.2311757103757818,
+      "learning_rate": 3.260506496952567e-05,
+      "loss": 0.1154,
+      "step": 2418
+    },
+    {
+      "epoch": 1.2592399791775117,
+      "grad_norm": 0.23201479378425932,
+      "learning_rate": 3.259167958192249e-05,
+      "loss": 0.114,
+      "step": 2419
+    },
+    {
+      "epoch": 1.2597605413846955,
+      "grad_norm": 0.2365740831566155,
+      "learning_rate": 3.257829179651361e-05,
+      "loss": 0.1095,
+      "step": 2420
+    },
+    {
+      "epoch": 1.2602811035918793,
+      "grad_norm": 0.22738663769119088,
+      "learning_rate": 3.2564901617527513e-05,
+      "loss": 0.1069,
+      "step": 2421
+    },
+    {
+      "epoch": 1.260801665799063,
+      "grad_norm": 0.24513932299659597,
+      "learning_rate": 3.2551509049193444e-05,
+      "loss": 0.1126,
+      "step": 2422
+    },
+    {
+      "epoch": 1.2613222280062468,
+      "grad_norm": 0.2295326150055126,
+      "learning_rate": 3.253811409574141e-05,
+      "loss": 0.1135,
+      "step": 2423
+    },
+    {
+      "epoch": 1.2618427902134304,
+      "grad_norm": 0.2392260915445472,
+      "learning_rate": 3.252471676140215e-05,
+      "loss": 0.1133,
+      "step": 2424
+    },
+    {
+      "epoch": 1.2623633524206141,
+      "grad_norm": 0.2281032715398985,
+      "learning_rate": 3.251131705040716e-05,
+      "loss": 0.1086,
+      "step": 2425
+    },
+    {
+      "epoch": 1.262883914627798,
+      "grad_norm": 0.240395636614695,
+      "learning_rate": 3.24979149669887e-05,
+      "loss": 0.1145,
+      "step": 2426
+    },
+    {
+      "epoch": 1.2634044768349817,
+      "grad_norm": 0.2406505726843499,
+      "learning_rate": 3.2484510515379776e-05,
+      "loss": 0.1122,
+      "step": 2427
+    },
+    {
+      "epoch": 1.2639250390421655,
+      "grad_norm": 0.2645085343890866,
+      "learning_rate": 3.247110369981413e-05,
+      "loss": 0.1151,
+      "step": 2428
+    },
+    {
+      "epoch": 1.2644456012493492,
+      "grad_norm": 0.22557528612610186,
+      "learning_rate": 3.245769452452626e-05,
+      "loss": 0.1092,
+      "step": 2429
+    },
+    {
+      "epoch": 1.264966163456533,
+      "grad_norm": 0.23590364280733392,
+      "learning_rate": 3.244428299375141e-05,
+      "loss": 0.1157,
+      "step": 2430
+    },
+    {
+      "epoch": 1.2654867256637168,
+      "grad_norm": 0.23355810219083442,
+      "learning_rate": 3.243086911172555e-05,
+      "loss": 0.1112,
+      "step": 2431
+    },
+    {
+      "epoch": 1.2660072878709006,
+      "grad_norm": 0.22768610677426723,
+      "learning_rate": 3.241745288268544e-05,
+      "loss": 0.1112,
+      "step": 2432
+    },
+    {
+      "epoch": 1.2665278500780843,
+      "grad_norm": 0.2213720698521258,
+      "learning_rate": 3.240403431086853e-05,
+      "loss": 0.1115,
+      "step": 2433
+    },
+    {
+      "epoch": 1.267048412285268,
+      "grad_norm": 0.2511275509634126,
+      "learning_rate": 3.239061340051302e-05,
+      "loss": 0.1144,
+      "step": 2434
+    },
+    {
+      "epoch": 1.2675689744924519,
+      "grad_norm": 0.22474601806469852,
+      "learning_rate": 3.237719015585787e-05,
+      "loss": 0.1084,
+      "step": 2435
+    },
+    {
+      "epoch": 1.2680895366996356,
+      "grad_norm": 0.22967485481677846,
+      "learning_rate": 3.236376458114276e-05,
+      "loss": 0.1053,
+      "step": 2436
+    },
+    {
+      "epoch": 1.2686100989068194,
+      "grad_norm": 0.24508106364086055,
+      "learning_rate": 3.235033668060813e-05,
+      "loss": 0.1141,
+      "step": 2437
+    },
+    {
+      "epoch": 1.2691306611140032,
+      "grad_norm": 0.23936751039871776,
+      "learning_rate": 3.233690645849511e-05,
+      "loss": 0.1126,
+      "step": 2438
+    },
+    {
+      "epoch": 1.269651223321187,
+      "grad_norm": 0.2423970044613412,
+      "learning_rate": 3.23234739190456e-05,
+      "loss": 0.11,
+      "step": 2439
+    },
+    {
+      "epoch": 1.2701717855283707,
+      "grad_norm": 0.2408077607865761,
+      "learning_rate": 3.2310039066502224e-05,
+      "loss": 0.1186,
+      "step": 2440
+    },
+    {
+      "epoch": 1.2706923477355545,
+      "grad_norm": 0.24264263902809302,
+      "learning_rate": 3.229660190510833e-05,
+      "loss": 0.1168,
+      "step": 2441
+    },
+    {
+      "epoch": 1.2712129099427383,
+      "grad_norm": 0.23451047905283273,
+      "learning_rate": 3.2283162439108004e-05,
+      "loss": 0.1114,
+      "step": 2442
+    },
+    {
+      "epoch": 1.271733472149922,
+      "grad_norm": 0.24187390316258348,
+      "learning_rate": 3.226972067274605e-05,
+      "loss": 0.1111,
+      "step": 2443
+    },
+    {
+      "epoch": 1.2722540343571056,
+      "grad_norm": 0.23365628213616502,
+      "learning_rate": 3.2256276610268e-05,
+      "loss": 0.1177,
+      "step": 2444
+    },
+    {
+      "epoch": 1.2727745965642894,
+      "grad_norm": 0.2427390698922539,
+      "learning_rate": 3.224283025592011e-05,
+      "loss": 0.1111,
+      "step": 2445
+    },
+    {
+      "epoch": 1.2732951587714731,
+      "grad_norm": 0.2512975614866838,
+      "learning_rate": 3.22293816139494e-05,
+      "loss": 0.1182,
+      "step": 2446
+    },
+    {
+      "epoch": 1.273815720978657,
+      "grad_norm": 0.24295218403998542,
+      "learning_rate": 3.2215930688603525e-05,
+      "loss": 0.1165,
+      "step": 2447
+    },
+    {
+      "epoch": 1.2743362831858407,
+      "grad_norm": 0.23644527244752322,
+      "learning_rate": 3.220247748413094e-05,
+      "loss": 0.113,
+      "step": 2448
+    },
+    {
+      "epoch": 1.2748568453930245,
+      "grad_norm": 0.23785821712712837,
+      "learning_rate": 3.21890220047808e-05,
+      "loss": 0.114,
+      "step": 2449
+    },
+    {
+      "epoch": 1.2753774076002082,
+      "grad_norm": 0.23632989350892925,
+      "learning_rate": 3.217556425480296e-05,
+      "loss": 0.1111,
+      "step": 2450
+    },
+    {
+      "epoch": 1.275897969807392,
+      "grad_norm": 0.23263875426854946,
+      "learning_rate": 3.216210423844801e-05,
+      "loss": 0.1102,
+      "step": 2451
+    },
+    {
+      "epoch": 1.2764185320145758,
+      "grad_norm": 0.23188957689089493,
+      "learning_rate": 3.214864195996723e-05,
+      "loss": 0.1075,
+      "step": 2452
+    },
+    {
+      "epoch": 1.2769390942217596,
+      "grad_norm": 0.23375003589359838,
+      "learning_rate": 3.213517742361267e-05,
+      "loss": 0.1082,
+      "step": 2453
+    },
+    {
+      "epoch": 1.2774596564289433,
+      "grad_norm": 0.24021397633189484,
+      "learning_rate": 3.212171063363702e-05,
+      "loss": 0.1166,
+      "step": 2454
+    },
+    {
+      "epoch": 1.277980218636127,
+      "grad_norm": 0.2322622124803816,
+      "learning_rate": 3.2108241594293756e-05,
+      "loss": 0.1139,
+      "step": 2455
+    },
+    {
+      "epoch": 1.2785007808433106,
+      "grad_norm": 0.23120316428655097,
+      "learning_rate": 3.209477030983698e-05,
+      "loss": 0.112,
+      "step": 2456
+    },
+    {
+      "epoch": 1.2790213430504944,
+      "grad_norm": 0.23819406522964934,
+      "learning_rate": 3.20812967845216e-05,
+      "loss": 0.1065,
+      "step": 2457
+    },
+    {
+      "epoch": 1.2795419052576782,
+      "grad_norm": 0.23305129874267824,
+      "learning_rate": 3.206782102260316e-05,
+      "loss": 0.1111,
+      "step": 2458
+    },
+    {
+      "epoch": 1.280062467464862,
+      "grad_norm": 0.26056719528908606,
+      "learning_rate": 3.205434302833792e-05,
+      "loss": 0.1104,
+      "step": 2459
+    },
+    {
+      "epoch": 1.2805830296720457,
+      "grad_norm": 0.23577408055653468,
+      "learning_rate": 3.204086280598289e-05,
+      "loss": 0.1192,
+      "step": 2460
+    },
+    {
+      "epoch": 1.2811035918792295,
+      "grad_norm": 0.2659768658535487,
+      "learning_rate": 3.202738035979571e-05,
+      "loss": 0.1197,
+      "step": 2461
+    },
+    {
+      "epoch": 1.2816241540864133,
+      "grad_norm": 0.2439223726100321,
+      "learning_rate": 3.2013895694034804e-05,
+      "loss": 0.107,
+      "step": 2462
+    },
+    {
+      "epoch": 1.282144716293597,
+      "grad_norm": 0.23562944836320895,
+      "learning_rate": 3.200040881295922e-05,
+      "loss": 0.1104,
+      "step": 2463
+    },
+    {
+      "epoch": 1.2826652785007808,
+      "grad_norm": 0.25461295273652873,
+      "learning_rate": 3.198691972082878e-05,
+      "loss": 0.1142,
+      "step": 2464
+    },
+    {
+      "epoch": 1.2831858407079646,
+      "grad_norm": 0.23055086268252875,
+      "learning_rate": 3.197342842190394e-05,
+      "loss": 0.1072,
+      "step": 2465
+    },
+    {
+      "epoch": 1.2837064029151484,
+      "grad_norm": 0.2770919732305317,
+      "learning_rate": 3.1959934920445894e-05,
+      "loss": 0.1161,
+      "step": 2466
+    },
+    {
+      "epoch": 1.2842269651223321,
+      "grad_norm": 0.24915210522689957,
+      "learning_rate": 3.194643922071651e-05,
+      "loss": 0.1139,
+      "step": 2467
+    },
+    {
+      "epoch": 1.284747527329516,
+      "grad_norm": 0.23485697563408167,
+      "learning_rate": 3.193294132697835e-05,
+      "loss": 0.1111,
+      "step": 2468
+    },
+    {
+      "epoch": 1.2852680895366997,
+      "grad_norm": 0.2376200199679929,
+      "learning_rate": 3.191944124349471e-05,
+      "loss": 0.1075,
+      "step": 2469
+    },
+    {
+      "epoch": 1.2857886517438835,
+      "grad_norm": 0.23788594569376167,
+      "learning_rate": 3.190593897452951e-05,
+      "loss": 0.1086,
+      "step": 2470
+    },
+    {
+      "epoch": 1.2863092139510672,
+      "grad_norm": 0.2338576028196406,
+      "learning_rate": 3.189243452434741e-05,
+      "loss": 0.1147,
+      "step": 2471
+    },
+    {
+      "epoch": 1.286829776158251,
+      "grad_norm": 0.23732556832773136,
+      "learning_rate": 3.187892789721373e-05,
+      "loss": 0.1112,
+      "step": 2472
+    },
+    {
+      "epoch": 1.2873503383654348,
+      "grad_norm": 0.24378666037810956,
+      "learning_rate": 3.186541909739452e-05,
+      "loss": 0.1124,
+      "step": 2473
+    },
+    {
+      "epoch": 1.2878709005726185,
+      "grad_norm": 0.23279147802128494,
+      "learning_rate": 3.185190812915646e-05,
+      "loss": 0.1099,
+      "step": 2474
+    },
+    {
+      "epoch": 1.2883914627798023,
+      "grad_norm": 0.2237628651766288,
+      "learning_rate": 3.1838394996766946e-05,
+      "loss": 0.1116,
+      "step": 2475
+    },
+    {
+      "epoch": 1.2889120249869859,
+      "grad_norm": 0.22815623654435227,
+      "learning_rate": 3.182487970449407e-05,
+      "loss": 0.1094,
+      "step": 2476
+    },
+    {
+      "epoch": 1.2894325871941696,
+      "grad_norm": 0.2324375751464136,
+      "learning_rate": 3.181136225660657e-05,
+      "loss": 0.1078,
+      "step": 2477
+    },
+    {
+      "epoch": 1.2899531494013534,
+      "grad_norm": 0.24381943224649627,
+      "learning_rate": 3.179784265737392e-05,
+      "loss": 0.1186,
+      "step": 2478
+    },
+    {
+      "epoch": 1.2904737116085372,
+      "grad_norm": 0.22358232434006062,
+      "learning_rate": 3.178432091106619e-05,
+      "loss": 0.1086,
+      "step": 2479
+    },
+    {
+      "epoch": 1.290994273815721,
+      "grad_norm": 0.22505649527448798,
+      "learning_rate": 3.1770797021954216e-05,
+      "loss": 0.1149,
+      "step": 2480
+    },
+    {
+      "epoch": 1.2915148360229047,
+      "grad_norm": 0.2271327003271489,
+      "learning_rate": 3.1757270994309445e-05,
+      "loss": 0.1151,
+      "step": 2481
+    },
+    {
+      "epoch": 1.2920353982300885,
+      "grad_norm": 0.22354543223054793,
+      "learning_rate": 3.174374283240405e-05,
+      "loss": 0.1157,
+      "step": 2482
+    },
+    {
+      "epoch": 1.2925559604372723,
+      "grad_norm": 0.23601648934888142,
+      "learning_rate": 3.1730212540510835e-05,
+      "loss": 0.1134,
+      "step": 2483
+    },
+    {
+      "epoch": 1.293076522644456,
+      "grad_norm": 0.23493593163696477,
+      "learning_rate": 3.1716680122903294e-05,
+      "loss": 0.1146,
+      "step": 2484
+    },
+    {
+      "epoch": 1.2935970848516398,
+      "grad_norm": 0.24903210748416696,
+      "learning_rate": 3.170314558385562e-05,
+      "loss": 0.112,
+      "step": 2485
+    },
+    {
+      "epoch": 1.2941176470588236,
+      "grad_norm": 0.24092859338470038,
+      "learning_rate": 3.1689608927642624e-05,
+      "loss": 0.114,
+      "step": 2486
+    },
+    {
+      "epoch": 1.2946382092660074,
+      "grad_norm": 0.23580201734250328,
+      "learning_rate": 3.1676070158539825e-05,
+      "loss": 0.1132,
+      "step": 2487
+    },
+    {
+      "epoch": 1.295158771473191,
+      "grad_norm": 0.2356380954462469,
+      "learning_rate": 3.166252928082339e-05,
+      "loss": 0.1087,
+      "step": 2488
+    },
+    {
+      "epoch": 1.2956793336803747,
+      "grad_norm": 0.23755644465406894,
+      "learning_rate": 3.164898629877016e-05,
+      "loss": 0.1109,
+      "step": 2489
+    },
+    {
+      "epoch": 1.2961998958875585,
+      "grad_norm": 0.2319470903145725,
+      "learning_rate": 3.1635441216657636e-05,
+      "loss": 0.1085,
+      "step": 2490
+    },
+    {
+      "epoch": 1.2967204580947422,
+      "grad_norm": 0.23994882725651862,
+      "learning_rate": 3.1621894038763995e-05,
+      "loss": 0.1142,
+      "step": 2491
+    },
+    {
+      "epoch": 1.297241020301926,
+      "grad_norm": 0.23110760341343395,
+      "learning_rate": 3.1608344769368056e-05,
+      "loss": 0.105,
+      "step": 2492
+    },
+    {
+      "epoch": 1.2977615825091098,
+      "grad_norm": 0.2493451968765941,
+      "learning_rate": 3.1594793412749315e-05,
+      "loss": 0.1144,
+      "step": 2493
+    },
+    {
+      "epoch": 1.2982821447162936,
+      "grad_norm": 0.23459448167768657,
+      "learning_rate": 3.158123997318792e-05,
+      "loss": 0.1117,
+      "step": 2494
+    },
+    {
+      "epoch": 1.2988027069234773,
+      "grad_norm": 0.23267743984705816,
+      "learning_rate": 3.1567684454964675e-05,
+      "loss": 0.1092,
+      "step": 2495
+    },
+    {
+      "epoch": 1.299323269130661,
+      "grad_norm": 0.23495505350422138,
+      "learning_rate": 3.155412686236105e-05,
+      "loss": 0.1099,
+      "step": 2496
+    },
+    {
+      "epoch": 1.2998438313378449,
+      "grad_norm": 0.2271281943706107,
+      "learning_rate": 3.1540567199659154e-05,
+      "loss": 0.1104,
+      "step": 2497
+    },
+    {
+      "epoch": 1.3003643935450286,
+      "grad_norm": 0.25317690129572223,
+      "learning_rate": 3.152700547114177e-05,
+      "loss": 0.1118,
+      "step": 2498
+    },
+    {
+      "epoch": 1.3008849557522124,
+      "grad_norm": 0.24086272384333318,
+      "learning_rate": 3.15134416810923e-05,
+      "loss": 0.1132,
+      "step": 2499
+    },
+    {
+      "epoch": 1.3014055179593962,
+      "grad_norm": 0.25095110856441216,
+      "learning_rate": 3.149987583379486e-05,
+      "loss": 0.1184,
+      "step": 2500
+    },
+    {
+      "epoch": 1.30192608016658,
+      "grad_norm": 0.23992994556687544,
+      "learning_rate": 3.1486307933534143e-05,
+      "loss": 0.1095,
+      "step": 2501
+    },
+    {
+      "epoch": 1.3024466423737637,
+      "grad_norm": 0.2363209984551393,
+      "learning_rate": 3.147273798459553e-05,
+      "loss": 0.1114,
+      "step": 2502
+    },
+    {
+      "epoch": 1.3029672045809475,
+      "grad_norm": 0.2524480432523037,
+      "learning_rate": 3.145916599126506e-05,
+      "loss": 0.1124,
+      "step": 2503
+    },
+    {
+      "epoch": 1.3034877667881313,
+      "grad_norm": 0.22828236428464427,
+      "learning_rate": 3.1445591957829374e-05,
+      "loss": 0.105,
+      "step": 2504
+    },
+    {
+      "epoch": 1.304008328995315,
+      "grad_norm": 0.25760549471351335,
+      "learning_rate": 3.14320158885758e-05,
+      "loss": 0.1099,
+      "step": 2505
+    },
+    {
+      "epoch": 1.3045288912024988,
+      "grad_norm": 0.24230787911345597,
+      "learning_rate": 3.141843778779229e-05,
+      "loss": 0.1096,
+      "step": 2506
+    },
+    {
+      "epoch": 1.3050494534096826,
+      "grad_norm": 0.24593151994037304,
+      "learning_rate": 3.140485765976743e-05,
+      "loss": 0.1068,
+      "step": 2507
+    },
+    {
+      "epoch": 1.3055700156168661,
+      "grad_norm": 0.2367624744628246,
+      "learning_rate": 3.1391275508790476e-05,
+      "loss": 0.1089,
+      "step": 2508
+    },
+    {
+      "epoch": 1.30609057782405,
+      "grad_norm": 0.25001842938229535,
+      "learning_rate": 3.1377691339151285e-05,
+      "loss": 0.1154,
+      "step": 2509
+    },
+    {
+      "epoch": 1.3066111400312337,
+      "grad_norm": 0.23444841996962343,
+      "learning_rate": 3.136410515514038e-05,
+      "loss": 0.1111,
+      "step": 2510
+    },
+    {
+      "epoch": 1.3071317022384175,
+      "grad_norm": 0.25857566739815296,
+      "learning_rate": 3.13505169610489e-05,
+      "loss": 0.1184,
+      "step": 2511
+    },
+    {
+      "epoch": 1.3076522644456012,
+      "grad_norm": 0.23910560119186738,
+      "learning_rate": 3.133692676116865e-05,
+      "loss": 0.1072,
+      "step": 2512
+    },
+    {
+      "epoch": 1.308172826652785,
+      "grad_norm": 0.23446735180490522,
+      "learning_rate": 3.132333455979202e-05,
+      "loss": 0.1109,
+      "step": 2513
+    },
+    {
+      "epoch": 1.3086933888599688,
+      "grad_norm": 0.24121685895466385,
+      "learning_rate": 3.130974036121208e-05,
+      "loss": 0.1137,
+      "step": 2514
+    },
+    {
+      "epoch": 1.3092139510671525,
+      "grad_norm": 0.2471492387424443,
+      "learning_rate": 3.12961441697225e-05,
+      "loss": 0.1115,
+      "step": 2515
+    },
+    {
+      "epoch": 1.3097345132743363,
+      "grad_norm": 0.23961316944400285,
+      "learning_rate": 3.1282545989617595e-05,
+      "loss": 0.1174,
+      "step": 2516
+    },
+    {
+      "epoch": 1.31025507548152,
+      "grad_norm": 0.22528913743073048,
+      "learning_rate": 3.126894582519231e-05,
+      "loss": 0.1122,
+      "step": 2517
+    },
+    {
+      "epoch": 1.3107756376887039,
+      "grad_norm": 0.2222728768218916,
+      "learning_rate": 3.1255343680742195e-05,
+      "loss": 0.1088,
+      "step": 2518
+    },
+    {
+      "epoch": 1.3112961998958876,
+      "grad_norm": 0.2364525553318129,
+      "learning_rate": 3.1241739560563446e-05,
+      "loss": 0.1092,
+      "step": 2519
+    },
+    {
+      "epoch": 1.3118167621030712,
+      "grad_norm": 0.23549419626533213,
+      "learning_rate": 3.122813346895288e-05,
+      "loss": 0.1126,
+      "step": 2520
+    },
+    {
+      "epoch": 1.312337324310255,
+      "grad_norm": 0.22438892761571402,
+      "learning_rate": 3.121452541020793e-05,
+      "loss": 0.1057,
+      "step": 2521
+    },
+    {
+      "epoch": 1.3128578865174387,
+      "grad_norm": 0.2331411062533715,
+      "learning_rate": 3.1200915388626654e-05,
+      "loss": 0.1132,
+      "step": 2522
+    },
+    {
+      "epoch": 1.3133784487246225,
+      "grad_norm": 0.23459665267739663,
+      "learning_rate": 3.118730340850774e-05,
+      "loss": 0.1075,
+      "step": 2523
+    },
+    {
+      "epoch": 1.3138990109318063,
+      "grad_norm": 0.2347453526216641,
+      "learning_rate": 3.1173689474150476e-05,
+      "loss": 0.1098,
+      "step": 2524
+    },
+    {
+      "epoch": 1.31441957313899,
+      "grad_norm": 0.23323798829157819,
+      "learning_rate": 3.116007358985477e-05,
+      "loss": 0.1091,
+      "step": 2525
+    },
+    {
+      "epoch": 1.3149401353461738,
+      "grad_norm": 0.23608030886582382,
+      "learning_rate": 3.1146455759921166e-05,
+      "loss": 0.1116,
+      "step": 2526
+    },
+    {
+      "epoch": 1.3154606975533576,
+      "grad_norm": 0.24356320449137384,
+      "learning_rate": 3.11328359886508e-05,
+      "loss": 0.1072,
+      "step": 2527
+    },
+    {
+      "epoch": 1.3159812597605414,
+      "grad_norm": 0.23069187776813016,
+      "learning_rate": 3.111921428034544e-05,
+      "loss": 0.1072,
+      "step": 2528
+    },
+    {
+      "epoch": 1.3165018219677251,
+      "grad_norm": 0.24927589485206192,
+      "learning_rate": 3.110559063930743e-05,
+      "loss": 0.1165,
+      "step": 2529
+    },
+    {
+      "epoch": 1.317022384174909,
+      "grad_norm": 0.22698192027918843,
+      "learning_rate": 3.109196506983978e-05,
+      "loss": 0.1098,
+      "step": 2530
+    },
+    {
+      "epoch": 1.3175429463820927,
+      "grad_norm": 0.2503079852778464,
+      "learning_rate": 3.107833757624605e-05,
+      "loss": 0.1106,
+      "step": 2531
+    },
+    {
+      "epoch": 1.3180635085892765,
+      "grad_norm": 0.2459913765268515,
+      "learning_rate": 3.1064708162830466e-05,
+      "loss": 0.1176,
+      "step": 2532
+    },
+    {
+      "epoch": 1.3185840707964602,
+      "grad_norm": 0.24673270505003705,
+      "learning_rate": 3.105107683389781e-05,
+      "loss": 0.1162,
+      "step": 2533
+    },
+    {
+      "epoch": 1.319104633003644,
+      "grad_norm": 0.22092262416559946,
+      "learning_rate": 3.10374435937535e-05,
+      "loss": 0.1067,
+      "step": 2534
+    },
+    {
+      "epoch": 1.3196251952108278,
+      "grad_norm": 0.23778340367233575,
+      "learning_rate": 3.102380844670355e-05,
+      "loss": 0.1156,
+      "step": 2535
+    },
+    {
+      "epoch": 1.3201457574180115,
+      "grad_norm": 0.24483607513262326,
+      "learning_rate": 3.101017139705455e-05,
+      "loss": 0.1145,
+      "step": 2536
+    },
+    {
+      "epoch": 1.3206663196251953,
+      "grad_norm": 0.22769610941345692,
+      "learning_rate": 3.099653244911375e-05,
+      "loss": 0.1144,
+      "step": 2537
+    },
+    {
+      "epoch": 1.321186881832379,
+      "grad_norm": 0.24187095569563516,
+      "learning_rate": 3.098289160718895e-05,
+      "loss": 0.1146,
+      "step": 2538
+    },
+    {
+      "epoch": 1.3217074440395629,
+      "grad_norm": 0.22264068349148974,
+      "learning_rate": 3.096924887558855e-05,
+      "loss": 0.1111,
+      "step": 2539
+    },
+    {
+      "epoch": 1.3222280062467464,
+      "grad_norm": 0.24298069839995018,
+      "learning_rate": 3.095560425862157e-05,
+      "loss": 0.1138,
+      "step": 2540
+    },
+    {
+      "epoch": 1.3227485684539302,
+      "grad_norm": 0.23672363156144616,
+      "learning_rate": 3.094195776059763e-05,
+      "loss": 0.1086,
+      "step": 2541
+    },
+    {
+      "epoch": 1.323269130661114,
+      "grad_norm": 0.23121118320526707,
+      "learning_rate": 3.09283093858269e-05,
+      "loss": 0.1157,
+      "step": 2542
+    },
+    {
+      "epoch": 1.3237896928682977,
+      "grad_norm": 0.22330970680800596,
+      "learning_rate": 3.0914659138620186e-05,
+      "loss": 0.1062,
+      "step": 2543
+    },
+    {
+      "epoch": 1.3243102550754815,
+      "grad_norm": 0.24667833728809804,
+      "learning_rate": 3.090100702328888e-05,
+      "loss": 0.1203,
+      "step": 2544
+    },
+    {
+      "epoch": 1.3248308172826653,
+      "grad_norm": 0.24428638534778402,
+      "learning_rate": 3.088735304414494e-05,
+      "loss": 0.1183,
+      "step": 2545
+    },
+    {
+      "epoch": 1.325351379489849,
+      "grad_norm": 0.2357143337600258,
+      "learning_rate": 3.087369720550094e-05,
+      "loss": 0.1158,
+      "step": 2546
+    },
+    {
+      "epoch": 1.3258719416970328,
+      "grad_norm": 0.21965988630691494,
+      "learning_rate": 3.0860039511670024e-05,
+      "loss": 0.1103,
+      "step": 2547
+    },
+    {
+      "epoch": 1.3263925039042166,
+      "grad_norm": 0.22044710681349058,
+      "learning_rate": 3.084637996696592e-05,
+      "loss": 0.1148,
+      "step": 2548
+    },
+    {
+      "epoch": 1.3269130661114004,
+      "grad_norm": 0.2305177255025451,
+      "learning_rate": 3.083271857570297e-05,
+      "loss": 0.1093,
+      "step": 2549
+    },
+    {
+      "epoch": 1.3274336283185841,
+      "grad_norm": 0.23136166042256415,
+      "learning_rate": 3.0819055342196054e-05,
+      "loss": 0.1078,
+      "step": 2550
+    },
+    {
+      "epoch": 1.327954190525768,
+      "grad_norm": 0.2421297560595075,
+      "learning_rate": 3.080539027076066e-05,
+      "loss": 0.1125,
+      "step": 2551
+    },
+    {
+      "epoch": 1.3284747527329515,
+      "grad_norm": 0.22632609868379913,
+      "learning_rate": 3.0791723365712867e-05,
+      "loss": 0.1098,
+      "step": 2552
+    },
+    {
+      "epoch": 1.3289953149401352,
+      "grad_norm": 0.23994272711042058,
+      "learning_rate": 3.077805463136931e-05,
+      "loss": 0.111,
+      "step": 2553
+    },
+    {
+      "epoch": 1.329515877147319,
+      "grad_norm": 0.24457717811072027,
+      "learning_rate": 3.07643840720472e-05,
+      "loss": 0.1168,
+      "step": 2554
+    },
+    {
+      "epoch": 1.3300364393545028,
+      "grad_norm": 0.2375550486078291,
+      "learning_rate": 3.075071169206437e-05,
+      "loss": 0.1164,
+      "step": 2555
+    },
+    {
+      "epoch": 1.3305570015616865,
+      "grad_norm": 0.23885122754311797,
+      "learning_rate": 3.073703749573916e-05,
+      "loss": 0.1135,
+      "step": 2556
+    },
+    {
+      "epoch": 1.3310775637688703,
+      "grad_norm": 0.24172905569894784,
+      "learning_rate": 3.072336148739053e-05,
+      "loss": 0.1105,
+      "step": 2557
+    },
+    {
+      "epoch": 1.331598125976054,
+      "grad_norm": 0.24554361390660961,
+      "learning_rate": 3.0709683671338e-05,
+      "loss": 0.1169,
+      "step": 2558
+    },
+    {
+      "epoch": 1.3321186881832379,
+      "grad_norm": 0.22416722525242608,
+      "learning_rate": 3.069600405190167e-05,
+      "loss": 0.1099,
+      "step": 2559
+    },
+    {
+      "epoch": 1.3326392503904216,
+      "grad_norm": 0.23312900420360225,
+      "learning_rate": 3.068232263340218e-05,
+      "loss": 0.1126,
+      "step": 2560
+    },
+    {
+      "epoch": 1.3331598125976054,
+      "grad_norm": 0.23606607210162725,
+      "learning_rate": 3.066863942016077e-05,
+      "loss": 0.1089,
+      "step": 2561
+    },
+    {
+      "epoch": 1.3336803748047892,
+      "grad_norm": 0.2373211884453036,
+      "learning_rate": 3.0654954416499244e-05,
+      "loss": 0.109,
+      "step": 2562
+    },
+    {
+      "epoch": 1.334200937011973,
+      "grad_norm": 0.2343989018463782,
+      "learning_rate": 3.064126762673994e-05,
+      "loss": 0.1125,
+      "step": 2563
+    },
+    {
+      "epoch": 1.3347214992191567,
+      "grad_norm": 0.26796891784177396,
+      "learning_rate": 3.062757905520582e-05,
+      "loss": 0.122,
+      "step": 2564
+    },
+    {
+      "epoch": 1.3352420614263405,
+      "grad_norm": 0.24223466770689397,
+      "learning_rate": 3.0613888706220336e-05,
+      "loss": 0.1134,
+      "step": 2565
+    },
+    {
+      "epoch": 1.3357626236335243,
+      "grad_norm": 0.2549438338386172,
+      "learning_rate": 3.060019658410755e-05,
+      "loss": 0.117,
+      "step": 2566
+    },
+    {
+      "epoch": 1.336283185840708,
+      "grad_norm": 0.25893651098604287,
+      "learning_rate": 3.0586502693192074e-05,
+      "loss": 0.116,
+      "step": 2567
+    },
+    {
+      "epoch": 1.3368037480478918,
+      "grad_norm": 0.22061547809765672,
+      "learning_rate": 3.0572807037799075e-05,
+      "loss": 0.1069,
+      "step": 2568
+    },
+    {
+      "epoch": 1.3373243102550756,
+      "grad_norm": 0.2707795969101939,
+      "learning_rate": 3.055910962225428e-05,
+      "loss": 0.1084,
+      "step": 2569
+    },
+    {
+      "epoch": 1.3378448724622594,
+      "grad_norm": 0.23308154556319333,
+      "learning_rate": 3.054541045088396e-05,
+      "loss": 0.115,
+      "step": 2570
+    },
+    {
+      "epoch": 1.3383654346694431,
+      "grad_norm": 0.2567417526177613,
+      "learning_rate": 3.053170952801496e-05,
+      "loss": 0.115,
+      "step": 2571
+    },
+    {
+      "epoch": 1.3388859968766267,
+      "grad_norm": 0.2362561720344537,
+      "learning_rate": 3.0518006857974666e-05,
+      "loss": 0.1142,
+      "step": 2572
+    },
+    {
+      "epoch": 1.3394065590838105,
+      "grad_norm": 0.24896491322044217,
+      "learning_rate": 3.0504302445091027e-05,
+      "loss": 0.1142,
+      "step": 2573
+    },
+    {
+      "epoch": 1.3399271212909942,
+      "grad_norm": 0.26160908249539727,
+      "learning_rate": 3.0490596293692525e-05,
+      "loss": 0.1179,
+      "step": 2574
+    },
+    {
+      "epoch": 1.340447683498178,
+      "grad_norm": 0.23521571929263785,
+      "learning_rate": 3.0476888408108202e-05,
+      "loss": 0.1162,
+      "step": 2575
+    },
+    {
+      "epoch": 1.3409682457053618,
+      "grad_norm": 0.24915979845519137,
+      "learning_rate": 3.0463178792667645e-05,
+      "loss": 0.1127,
+      "step": 2576
+    },
+    {
+      "epoch": 1.3414888079125455,
+      "grad_norm": 0.23236370763142386,
+      "learning_rate": 3.0449467451700997e-05,
+      "loss": 0.1052,
+      "step": 2577
+    },
+    {
+      "epoch": 1.3420093701197293,
+      "grad_norm": 0.2204117728067703,
+      "learning_rate": 3.0435754389538928e-05,
+      "loss": 0.1048,
+      "step": 2578
+    },
+    {
+      "epoch": 1.342529932326913,
+      "grad_norm": 0.23813101057229322,
+      "learning_rate": 3.0422039610512666e-05,
+      "loss": 0.109,
+      "step": 2579
+    },
+    {
+      "epoch": 1.3430504945340969,
+      "grad_norm": 0.24713378801317953,
+      "learning_rate": 3.0408323118953968e-05,
+      "loss": 0.1121,
+      "step": 2580
+    },
+    {
+      "epoch": 1.3435710567412806,
+      "grad_norm": 0.25584268149164247,
+      "learning_rate": 3.0394604919195156e-05,
+      "loss": 0.1118,
+      "step": 2581
+    },
+    {
+      "epoch": 1.3440916189484644,
+      "grad_norm": 0.24392930456555278,
+      "learning_rate": 3.0380885015569067e-05,
+      "loss": 0.1083,
+      "step": 2582
+    },
+    {
+      "epoch": 1.3446121811556482,
+      "grad_norm": 0.2352412179715951,
+      "learning_rate": 3.036716341240908e-05,
+      "loss": 0.1099,
+      "step": 2583
+    },
+    {
+      "epoch": 1.3451327433628317,
+      "grad_norm": 0.25301035157344787,
+      "learning_rate": 3.0353440114049126e-05,
+      "loss": 0.1077,
+      "step": 2584
+    },
+    {
+      "epoch": 1.3456533055700155,
+      "grad_norm": 0.27102376417563373,
+      "learning_rate": 3.0339715124823652e-05,
+      "loss": 0.1133,
+      "step": 2585
+    },
+    {
+      "epoch": 1.3461738677771993,
+      "grad_norm": 0.245471066517425,
+      "learning_rate": 3.0325988449067654e-05,
+      "loss": 0.1063,
+      "step": 2586
+    },
+    {
+      "epoch": 1.346694429984383,
+      "grad_norm": 0.25444136673567697,
+      "learning_rate": 3.031226009111665e-05,
+      "loss": 0.1114,
+      "step": 2587
+    },
+    {
+      "epoch": 1.3472149921915668,
+      "grad_norm": 0.24276271471445565,
+      "learning_rate": 3.0298530055306708e-05,
+      "loss": 0.1131,
+      "step": 2588
+    },
+    {
+      "epoch": 1.3477355543987506,
+      "grad_norm": 0.26370328884810446,
+      "learning_rate": 3.028479834597439e-05,
+      "loss": 0.1118,
+      "step": 2589
+    },
+    {
+      "epoch": 1.3482561166059344,
+      "grad_norm": 0.252507222324151,
+      "learning_rate": 3.027106496745683e-05,
+      "loss": 0.1088,
+      "step": 2590
+    },
+    {
+      "epoch": 1.3487766788131181,
+      "grad_norm": 0.22861875840700396,
+      "learning_rate": 3.025732992409166e-05,
+      "loss": 0.1117,
+      "step": 2591
+    },
+    {
+      "epoch": 1.349297241020302,
+      "grad_norm": 0.24956628943142115,
+      "learning_rate": 3.0243593220217044e-05,
+      "loss": 0.1116,
+      "step": 2592
+    },
+    {
+      "epoch": 1.3498178032274857,
+      "grad_norm": 0.2390835790110749,
+      "learning_rate": 3.0229854860171662e-05,
+      "loss": 0.1131,
+      "step": 2593
+    },
+    {
+      "epoch": 1.3503383654346695,
+      "grad_norm": 0.24074530354960633,
+      "learning_rate": 3.021611484829475e-05,
+      "loss": 0.1111,
+      "step": 2594
+    },
+    {
+      "epoch": 1.3508589276418532,
+      "grad_norm": 0.2444101290555521,
+      "learning_rate": 3.0202373188926037e-05,
+      "loss": 0.1084,
+      "step": 2595
+    },
+    {
+      "epoch": 1.351379489849037,
+      "grad_norm": 0.24713525990174354,
+      "learning_rate": 3.0188629886405763e-05,
+      "loss": 0.1054,
+      "step": 2596
+    },
+    {
+      "epoch": 1.3519000520562208,
+      "grad_norm": 0.25164204870542045,
+      "learning_rate": 3.017488494507471e-05,
+      "loss": 0.1121,
+      "step": 2597
+    },
+    {
+      "epoch": 1.3524206142634045,
+      "grad_norm": 0.24569308639230078,
+      "learning_rate": 3.0161138369274177e-05,
+      "loss": 0.1121,
+      "step": 2598
+    },
+    {
+      "epoch": 1.3529411764705883,
+      "grad_norm": 0.24172753415281717,
+      "learning_rate": 3.0147390163345972e-05,
+      "loss": 0.1104,
+      "step": 2599
+    },
+    {
+      "epoch": 1.353461738677772,
+      "grad_norm": 0.2713692191610928,
+      "learning_rate": 3.013364033163241e-05,
+      "loss": 0.1175,
+      "step": 2600
+    },
+    {
+      "epoch": 1.3539823008849559,
+      "grad_norm": 0.2610319613907833,
+      "learning_rate": 3.0119888878476338e-05,
+      "loss": 0.1154,
+      "step": 2601
+    },
+    {
+      "epoch": 1.3545028630921396,
+      "grad_norm": 0.22606906047311914,
+      "learning_rate": 3.0106135808221093e-05,
+      "loss": 0.1057,
+      "step": 2602
+    },
+    {
+      "epoch": 1.3550234252993234,
+      "grad_norm": 0.2416710905492764,
+      "learning_rate": 3.009238112521054e-05,
+      "loss": 0.1111,
+      "step": 2603
+    },
+    {
+      "epoch": 1.355543987506507,
+      "grad_norm": 0.23725466138349213,
+      "learning_rate": 3.007862483378906e-05,
+      "loss": 0.114,
+      "step": 2604
+    },
+    {
+      "epoch": 1.3560645497136907,
+      "grad_norm": 0.22201764810825417,
+      "learning_rate": 3.0064866938301507e-05,
+      "loss": 0.1091,
+      "step": 2605
+    },
+    {
+      "epoch": 1.3565851119208745,
+      "grad_norm": 0.26553203846707873,
+      "learning_rate": 3.005110744309328e-05,
+      "loss": 0.1095,
+      "step": 2606
+    },
+    {
+      "epoch": 1.3571056741280583,
+      "grad_norm": 0.22897934347302173,
+      "learning_rate": 3.003734635251026e-05,
+      "loss": 0.1042,
+      "step": 2607
+    },
+    {
+      "epoch": 1.357626236335242,
+      "grad_norm": 0.22965225339232517,
+      "learning_rate": 3.0023583670898848e-05,
+      "loss": 0.1123,
+      "step": 2608
+    },
+    {
+      "epoch": 1.3581467985424258,
+      "grad_norm": 0.2374884978816423,
+      "learning_rate": 3.0009819402605938e-05,
+      "loss": 0.1065,
+      "step": 2609
+    },
+    {
+      "epoch": 1.3586673607496096,
+      "grad_norm": 0.23724855307575515,
+      "learning_rate": 2.999605355197892e-05,
+      "loss": 0.1138,
+      "step": 2610
+    },
+    {
+      "epoch": 1.3591879229567934,
+      "grad_norm": 0.2522681898496753,
+      "learning_rate": 2.9982286123365694e-05,
+      "loss": 0.1127,
+      "step": 2611
+    },
+    {
+      "epoch": 1.3597084851639771,
+      "grad_norm": 0.2390455975944688,
+      "learning_rate": 2.9968517121114652e-05,
+      "loss": 0.1103,
+      "step": 2612
+    },
+    {
+      "epoch": 1.360229047371161,
+      "grad_norm": 0.22922231459543907,
+      "learning_rate": 2.9954746549574697e-05,
+      "loss": 0.1017,
+      "step": 2613
+    },
+    {
+      "epoch": 1.3607496095783447,
+      "grad_norm": 0.23580922604859134,
+      "learning_rate": 2.9940974413095203e-05,
+      "loss": 0.1099,
+      "step": 2614
+    },
+    {
+      "epoch": 1.3612701717855284,
+      "grad_norm": 0.22858569901780046,
+      "learning_rate": 2.9927200716026055e-05,
+      "loss": 0.111,
+      "step": 2615
+    },
+    {
+      "epoch": 1.361790733992712,
+      "grad_norm": 0.2532805937856553,
+      "learning_rate": 2.9913425462717625e-05,
+      "loss": 0.113,
+      "step": 2616
+    },
+    {
+      "epoch": 1.3623112961998958,
+      "grad_norm": 0.23253715435079703,
+      "learning_rate": 2.989964865752079e-05,
+      "loss": 0.11,
+      "step": 2617
+    },
+    {
+      "epoch": 1.3628318584070795,
+      "grad_norm": 0.23333001277583243,
+      "learning_rate": 2.9885870304786896e-05,
+      "loss": 0.1116,
+      "step": 2618
+    },
+    {
+      "epoch": 1.3633524206142633,
+      "grad_norm": 0.21615247744152524,
+      "learning_rate": 2.9872090408867785e-05,
+      "loss": 0.1058,
+      "step": 2619
+    },
+    {
+      "epoch": 1.363872982821447,
+      "grad_norm": 0.2367438794292196,
+      "learning_rate": 2.9858308974115808e-05,
+      "loss": 0.112,
+      "step": 2620
+    },
+    {
+      "epoch": 1.3643935450286309,
+      "grad_norm": 0.23367571489326353,
+      "learning_rate": 2.9844526004883755e-05,
+      "loss": 0.1045,
+      "step": 2621
+    },
+    {
+      "epoch": 1.3649141072358146,
+      "grad_norm": 0.24556639627942822,
+      "learning_rate": 2.9830741505524958e-05,
+      "loss": 0.1217,
+      "step": 2622
+    },
+    {
+      "epoch": 1.3654346694429984,
+      "grad_norm": 0.23637425075744797,
+      "learning_rate": 2.9816955480393187e-05,
+      "loss": 0.111,
+      "step": 2623
+    },
+    {
+      "epoch": 1.3659552316501822,
+      "grad_norm": 0.25593497993008707,
+      "learning_rate": 2.9803167933842714e-05,
+      "loss": 0.11,
+      "step": 2624
+    },
+    {
+      "epoch": 1.366475793857366,
+      "grad_norm": 0.23985020724360726,
+      "learning_rate": 2.9789378870228283e-05,
+      "loss": 0.1109,
+      "step": 2625
+    },
+    {
+      "epoch": 1.3669963560645497,
+      "grad_norm": 0.23470299239423129,
+      "learning_rate": 2.9775588293905132e-05,
+      "loss": 0.1123,
+      "step": 2626
+    },
+    {
+      "epoch": 1.3675169182717335,
+      "grad_norm": 0.2331930805433966,
+      "learning_rate": 2.976179620922896e-05,
+      "loss": 0.1086,
+      "step": 2627
+    },
+    {
+      "epoch": 1.3680374804789173,
+      "grad_norm": 0.23535519041672734,
+      "learning_rate": 2.9748002620555944e-05,
+      "loss": 0.1084,
+      "step": 2628
+    },
+    {
+      "epoch": 1.368558042686101,
+      "grad_norm": 0.2506834392739139,
+      "learning_rate": 2.9734207532242754e-05,
+      "loss": 0.1134,
+      "step": 2629
+    },
+    {
+      "epoch": 1.3690786048932848,
+      "grad_norm": 0.22986664916695576,
+      "learning_rate": 2.9720410948646504e-05,
+      "loss": 0.1081,
+      "step": 2630
+    },
+    {
+      "epoch": 1.3695991671004686,
+      "grad_norm": 0.24834546542727068,
+      "learning_rate": 2.970661287412482e-05,
+      "loss": 0.1162,
+      "step": 2631
+    },
+    {
+      "epoch": 1.3701197293076524,
+      "grad_norm": 0.2308044732796431,
+      "learning_rate": 2.969281331303576e-05,
+      "loss": 0.1112,
+      "step": 2632
+    },
+    {
+      "epoch": 1.3706402915148361,
+      "grad_norm": 0.24372786534936366,
+      "learning_rate": 2.967901226973787e-05,
+      "loss": 0.1123,
+      "step": 2633
+    },
+    {
+      "epoch": 1.37116085372202,
+      "grad_norm": 0.23353876211260266,
+      "learning_rate": 2.966520974859016e-05,
+      "loss": 0.1063,
+      "step": 2634
+    },
+    {
+      "epoch": 1.3716814159292037,
+      "grad_norm": 0.23840470512997244,
+      "learning_rate": 2.965140575395211e-05,
+      "loss": 0.1123,
+      "step": 2635
+    },
+    {
+      "epoch": 1.3722019781363872,
+      "grad_norm": 0.2308432390077942,
+      "learning_rate": 2.9637600290183675e-05,
+      "loss": 0.1126,
+      "step": 2636
+    },
+    {
+      "epoch": 1.372722540343571,
+      "grad_norm": 0.23561930060975914,
+      "learning_rate": 2.9623793361645247e-05,
+      "loss": 0.1106,
+      "step": 2637
+    },
+    {
+      "epoch": 1.3732431025507548,
+      "grad_norm": 0.21718824756279217,
+      "learning_rate": 2.96099849726977e-05,
+      "loss": 0.1078,
+      "step": 2638
+    },
+    {
+      "epoch": 1.3737636647579385,
+      "grad_norm": 0.23549809364298932,
+      "learning_rate": 2.9596175127702368e-05,
+      "loss": 0.1155,
+      "step": 2639
+    },
+    {
+      "epoch": 1.3742842269651223,
+      "grad_norm": 0.2351146169717648,
+      "learning_rate": 2.958236383102105e-05,
+      "loss": 0.1089,
+      "step": 2640
+    },
+    {
+      "epoch": 1.374804789172306,
+      "grad_norm": 0.230720734914438,
+      "learning_rate": 2.956855108701599e-05,
+      "loss": 0.1125,
+      "step": 2641
+    },
+    {
+      "epoch": 1.3753253513794899,
+      "grad_norm": 0.23350953234570837,
+      "learning_rate": 2.9554736900049883e-05,
+      "loss": 0.1123,
+      "step": 2642
+    },
+    {
+      "epoch": 1.3758459135866736,
+      "grad_norm": 0.2272348761594922,
+      "learning_rate": 2.954092127448591e-05,
+      "loss": 0.1126,
+      "step": 2643
+    },
+    {
+      "epoch": 1.3763664757938574,
+      "grad_norm": 0.26121567593064826,
+      "learning_rate": 2.9527104214687685e-05,
+      "loss": 0.1171,
+      "step": 2644
+    },
+    {
+      "epoch": 1.3768870380010412,
+      "grad_norm": 0.24347665782131095,
+      "learning_rate": 2.951328572501928e-05,
+      "loss": 0.114,
+      "step": 2645
+    },
+    {
+      "epoch": 1.377407600208225,
+      "grad_norm": 0.22666019429301196,
+      "learning_rate": 2.94994658098452e-05,
+      "loss": 0.109,
+      "step": 2646
+    },
+    {
+      "epoch": 1.3779281624154087,
+      "grad_norm": 0.24888395219943815,
+      "learning_rate": 2.9485644473530437e-05,
+      "loss": 0.114,
+      "step": 2647
+    },
+    {
+      "epoch": 1.3784487246225923,
+      "grad_norm": 0.2339315941555209,
+      "learning_rate": 2.9471821720440406e-05,
+      "loss": 0.1054,
+      "step": 2648
+    },
+    {
+      "epoch": 1.378969286829776,
+      "grad_norm": 0.22715324176204416,
+      "learning_rate": 2.9457997554940974e-05,
+      "loss": 0.1071,
+      "step": 2649
+    },
+    {
+      "epoch": 1.3794898490369598,
+      "grad_norm": 0.23984980511126433,
+      "learning_rate": 2.944417198139846e-05,
+      "loss": 0.1104,
+      "step": 2650
+    },
+    {
+      "epoch": 1.3800104112441436,
+      "grad_norm": 0.2388805658491324,
+      "learning_rate": 2.9430345004179614e-05,
+      "loss": 0.1112,
+      "step": 2651
+    },
+    {
+      "epoch": 1.3805309734513274,
+      "grad_norm": 0.2609240796303999,
+      "learning_rate": 2.9416516627651647e-05,
+      "loss": 0.1201,
+      "step": 2652
+    },
+    {
+      "epoch": 1.3810515356585111,
+      "grad_norm": 0.22009015612627597,
+      "learning_rate": 2.9402686856182205e-05,
+      "loss": 0.1059,
+      "step": 2653
+    },
+    {
+      "epoch": 1.381572097865695,
+      "grad_norm": 0.25819973582961897,
+      "learning_rate": 2.9388855694139373e-05,
+      "loss": 0.1136,
+      "step": 2654
+    },
+    {
+      "epoch": 1.3820926600728787,
+      "grad_norm": 0.2626233123096237,
+      "learning_rate": 2.9375023145891666e-05,
+      "loss": 0.1182,
+      "step": 2655
+    },
+    {
+      "epoch": 1.3826132222800624,
+      "grad_norm": 0.24547878318931401,
+      "learning_rate": 2.936118921580806e-05,
+      "loss": 0.112,
+      "step": 2656
+    },
+    {
+      "epoch": 1.3831337844872462,
+      "grad_norm": 0.22938479950092555,
+      "learning_rate": 2.9347353908257936e-05,
+      "loss": 0.1112,
+      "step": 2657
+    },
+    {
+      "epoch": 1.38365434669443,
+      "grad_norm": 0.23335117709653277,
+      "learning_rate": 2.9333517227611152e-05,
+      "loss": 0.1116,
+      "step": 2658
+    },
+    {
+      "epoch": 1.3841749089016138,
+      "grad_norm": 0.2514180524501941,
+      "learning_rate": 2.9319679178237957e-05,
+      "loss": 0.1115,
+      "step": 2659
+    },
+    {
+      "epoch": 1.3846954711087975,
+      "grad_norm": 0.22641541429241346,
+      "learning_rate": 2.9305839764509058e-05,
+      "loss": 0.1071,
+      "step": 2660
+    },
+    {
+      "epoch": 1.3852160333159813,
+      "grad_norm": 0.23200240748290513,
+      "learning_rate": 2.929199899079558e-05,
+      "loss": 0.1067,
+      "step": 2661
+    },
+    {
+      "epoch": 1.385736595523165,
+      "grad_norm": 0.2390937747494344,
+      "learning_rate": 2.9278156861469096e-05,
+      "loss": 0.1081,
+      "step": 2662
+    },
+    {
+      "epoch": 1.3862571577303489,
+      "grad_norm": 0.23468619522538356,
+      "learning_rate": 2.9264313380901588e-05,
+      "loss": 0.1097,
+      "step": 2663
+    },
+    {
+      "epoch": 1.3867777199375326,
+      "grad_norm": 0.24345058195506755,
+      "learning_rate": 2.9250468553465466e-05,
+      "loss": 0.1089,
+      "step": 2664
+    },
+    {
+      "epoch": 1.3872982821447164,
+      "grad_norm": 0.25475277322641765,
+      "learning_rate": 2.9236622383533575e-05,
+      "loss": 0.1113,
+      "step": 2665
+    },
+    {
+      "epoch": 1.3878188443519002,
+      "grad_norm": 0.2532468565384966,
+      "learning_rate": 2.9222774875479176e-05,
+      "loss": 0.1079,
+      "step": 2666
+    },
+    {
+      "epoch": 1.388339406559084,
+      "grad_norm": 0.2512503551401563,
+      "learning_rate": 2.920892603367596e-05,
+      "loss": 0.1107,
+      "step": 2667
+    },
+    {
+      "epoch": 1.3888599687662675,
+      "grad_norm": 0.24339764746526638,
+      "learning_rate": 2.919507586249805e-05,
+      "loss": 0.1161,
+      "step": 2668
+    },
+    {
+      "epoch": 1.3893805309734513,
+      "grad_norm": 0.2523651063330094,
+      "learning_rate": 2.9181224366319947e-05,
+      "loss": 0.1139,
+      "step": 2669
+    },
+    {
+      "epoch": 1.389901093180635,
+      "grad_norm": 0.22988428164397576,
+      "learning_rate": 2.916737154951662e-05,
+      "loss": 0.1095,
+      "step": 2670
+    },
+    {
+      "epoch": 1.3904216553878188,
+      "grad_norm": 0.24090438564750494,
+      "learning_rate": 2.9153517416463418e-05,
+      "loss": 0.1097,
+      "step": 2671
+    },
+    {
+      "epoch": 1.3909422175950026,
+      "grad_norm": 0.24074539123393496,
+      "learning_rate": 2.913966197153613e-05,
+      "loss": 0.1068,
+      "step": 2672
+    },
+    {
+      "epoch": 1.3914627798021864,
+      "grad_norm": 0.2374411651100617,
+      "learning_rate": 2.9125805219110952e-05,
+      "loss": 0.1135,
+      "step": 2673
+    },
+    {
+      "epoch": 1.3919833420093701,
+      "grad_norm": 0.24364817979891987,
+      "learning_rate": 2.9111947163564478e-05,
+      "loss": 0.1077,
+      "step": 2674
+    },
+    {
+      "epoch": 1.392503904216554,
+      "grad_norm": 0.239001833403062,
+      "learning_rate": 2.9098087809273743e-05,
+      "loss": 0.1089,
+      "step": 2675
+    },
+    {
+      "epoch": 1.3930244664237377,
+      "grad_norm": 0.2415390586855587,
+      "learning_rate": 2.908422716061617e-05,
+      "loss": 0.108,
+      "step": 2676
+    },
+    {
+      "epoch": 1.3935450286309214,
+      "grad_norm": 0.26779259818902096,
+      "learning_rate": 2.9070365221969598e-05,
+      "loss": 0.1099,
+      "step": 2677
+    },
+    {
+      "epoch": 1.3940655908381052,
+      "grad_norm": 0.23464593639934386,
+      "learning_rate": 2.9056501997712267e-05,
+      "loss": 0.1111,
+      "step": 2678
+    },
+    {
+      "epoch": 1.394586153045289,
+      "grad_norm": 0.2343125934015767,
+      "learning_rate": 2.904263749222283e-05,
+      "loss": 0.1055,
+      "step": 2679
+    },
+    {
+      "epoch": 1.3951067152524725,
+      "grad_norm": 0.2366101209417697,
+      "learning_rate": 2.9028771709880342e-05,
+      "loss": 0.1104,
+      "step": 2680
+    },
+    {
+      "epoch": 1.3956272774596563,
+      "grad_norm": 0.24371168991559053,
+      "learning_rate": 2.9014904655064273e-05,
+      "loss": 0.1132,
+      "step": 2681
+    },
+    {
+      "epoch": 1.39614783966684,
+      "grad_norm": 0.23630952126438518,
+      "learning_rate": 2.9001036332154474e-05,
+      "loss": 0.1088,
+      "step": 2682
+    },
+    {
+      "epoch": 1.3966684018740239,
+      "grad_norm": 0.24538418864834158,
+      "learning_rate": 2.8987166745531207e-05,
+      "loss": 0.1126,
+      "step": 2683
+    },
+    {
+      "epoch": 1.3971889640812076,
+      "grad_norm": 0.23134351061219888,
+      "learning_rate": 2.897329589957514e-05,
+      "loss": 0.1144,
+      "step": 2684
+    },
+    {
+      "epoch": 1.3977095262883914,
+      "grad_norm": 0.23678786376287386,
+      "learning_rate": 2.8959423798667317e-05,
+      "loss": 0.1088,
+      "step": 2685
+    },
+    {
+      "epoch": 1.3982300884955752,
+      "grad_norm": 0.23975398835505177,
+      "learning_rate": 2.894555044718921e-05,
+      "loss": 0.1089,
+      "step": 2686
+    },
+    {
+      "epoch": 1.398750650702759,
+      "grad_norm": 0.23608185526452935,
+      "learning_rate": 2.893167584952266e-05,
+      "loss": 0.11,
+      "step": 2687
+    },
+    {
+      "epoch": 1.3992712129099427,
+      "grad_norm": 0.24040631010421307,
+      "learning_rate": 2.8917800010049917e-05,
+      "loss": 0.1169,
+      "step": 2688
+    },
+    {
+      "epoch": 1.3997917751171265,
+      "grad_norm": 0.2414692039364617,
+      "learning_rate": 2.8903922933153606e-05,
+      "loss": 0.1097,
+      "step": 2689
+    },
+    {
+      "epoch": 1.4003123373243103,
+      "grad_norm": 0.24897062492291208,
+      "learning_rate": 2.8890044623216763e-05,
+      "loss": 0.1149,
+      "step": 2690
+    },
+    {
+      "epoch": 1.400832899531494,
+      "grad_norm": 0.23371010023798855,
+      "learning_rate": 2.8876165084622797e-05,
+      "loss": 0.1115,
+      "step": 2691
+    },
+    {
+      "epoch": 1.4013534617386778,
+      "grad_norm": 0.21963840974379567,
+      "learning_rate": 2.8862284321755517e-05,
+      "loss": 0.1007,
+      "step": 2692
+    },
+    {
+      "epoch": 1.4018740239458616,
+      "grad_norm": 0.23159324050922378,
+      "learning_rate": 2.8848402338999115e-05,
+      "loss": 0.1115,
+      "step": 2693
+    },
+    {
+      "epoch": 1.4023945861530454,
+      "grad_norm": 0.23496545202276656,
+      "learning_rate": 2.8834519140738158e-05,
+      "loss": 0.1143,
+      "step": 2694
+    },
+    {
+      "epoch": 1.4029151483602291,
+      "grad_norm": 0.23229098782752947,
+      "learning_rate": 2.882063473135763e-05,
+      "loss": 0.1135,
+      "step": 2695
+    },
+    {
+      "epoch": 1.403435710567413,
+      "grad_norm": 0.23844892538632909,
+      "learning_rate": 2.880674911524284e-05,
+      "loss": 0.1125,
+      "step": 2696
+    },
+    {
+      "epoch": 1.4039562727745967,
+      "grad_norm": 0.22375205668172357,
+      "learning_rate": 2.8792862296779538e-05,
+      "loss": 0.1086,
+      "step": 2697
+    },
+    {
+      "epoch": 1.4044768349817804,
+      "grad_norm": 0.243507183609586,
+      "learning_rate": 2.8778974280353817e-05,
+      "loss": 0.1111,
+      "step": 2698
+    },
+    {
+      "epoch": 1.4049973971889642,
+      "grad_norm": 0.23876948702376113,
+      "learning_rate": 2.8765085070352153e-05,
+      "loss": 0.1116,
+      "step": 2699
+    },
+    {
+      "epoch": 1.4055179593961478,
+      "grad_norm": 0.23325769806586572,
+      "learning_rate": 2.8751194671161423e-05,
+      "loss": 0.1098,
+      "step": 2700
+    },
+    {
+      "epoch": 1.4060385216033315,
+      "grad_norm": 0.2508356243169636,
+      "learning_rate": 2.8737303087168837e-05,
+      "loss": 0.1071,
+      "step": 2701
+    },
+    {
+      "epoch": 1.4065590838105153,
+      "grad_norm": 0.24922685226760763,
+      "learning_rate": 2.8723410322762027e-05,
+      "loss": 0.1152,
+      "step": 2702
+    },
+    {
+      "epoch": 1.407079646017699,
+      "grad_norm": 0.2487978849373613,
+      "learning_rate": 2.8709516382328962e-05,
+      "loss": 0.113,
+      "step": 2703
+    },
+    {
+      "epoch": 1.4076002082248829,
+      "grad_norm": 0.24928009890621933,
+      "learning_rate": 2.8695621270258e-05,
+      "loss": 0.1174,
+      "step": 2704
+    },
+    {
+      "epoch": 1.4081207704320666,
+      "grad_norm": 0.23579163389801594,
+      "learning_rate": 2.8681724990937857e-05,
+      "loss": 0.1105,
+      "step": 2705
+    },
+    {
+      "epoch": 1.4086413326392504,
+      "grad_norm": 0.23415300436456926,
+      "learning_rate": 2.8667827548757624e-05,
+      "loss": 0.1102,
+      "step": 2706
+    },
+    {
+      "epoch": 1.4091618948464342,
+      "grad_norm": 0.22972294165859275,
+      "learning_rate": 2.865392894810678e-05,
+      "loss": 0.1061,
+      "step": 2707
+    },
+    {
+      "epoch": 1.409682457053618,
+      "grad_norm": 0.2345573651214216,
+      "learning_rate": 2.8640029193375128e-05,
+      "loss": 0.1046,
+      "step": 2708
+    },
+    {
+      "epoch": 1.4102030192608017,
+      "grad_norm": 0.23651806913382753,
+      "learning_rate": 2.8626128288952862e-05,
+      "loss": 0.1069,
+      "step": 2709
+    },
+    {
+      "epoch": 1.4107235814679855,
+      "grad_norm": 0.24904333761243203,
+      "learning_rate": 2.8612226239230532e-05,
+      "loss": 0.1129,
+      "step": 2710
+    },
+    {
+      "epoch": 1.4112441436751693,
+      "grad_norm": 0.23917081649980002,
+      "learning_rate": 2.8598323048599067e-05,
+      "loss": 0.1081,
+      "step": 2711
+    },
+    {
+      "epoch": 1.4117647058823528,
+      "grad_norm": 0.2456854467567518,
+      "learning_rate": 2.8584418721449724e-05,
+      "loss": 0.1103,
+      "step": 2712
+    },
+    {
+      "epoch": 1.4122852680895366,
+      "grad_norm": 0.2452287893186469,
+      "learning_rate": 2.8570513262174152e-05,
+      "loss": 0.1174,
+      "step": 2713
+    },
+    {
+      "epoch": 1.4128058302967204,
+      "grad_norm": 0.23238334523577805,
+      "learning_rate": 2.855660667516433e-05,
+      "loss": 0.1099,
+      "step": 2714
+    },
+    {
+      "epoch": 1.4133263925039041,
+      "grad_norm": 0.23389506698918613,
+      "learning_rate": 2.854269896481261e-05,
+      "loss": 0.1084,
+      "step": 2715
+    },
+    {
+      "epoch": 1.413846954711088,
+      "grad_norm": 0.23215202724524295,
+      "learning_rate": 2.8528790135511695e-05,
+      "loss": 0.1136,
+      "step": 2716
+    },
+    {
+      "epoch": 1.4143675169182717,
+      "grad_norm": 0.2394218131173443,
+      "learning_rate": 2.851488019165464e-05,
+      "loss": 0.1073,
+      "step": 2717
+    },
+    {
+      "epoch": 1.4148880791254554,
+      "grad_norm": 0.2260015168454527,
+      "learning_rate": 2.8500969137634853e-05,
+      "loss": 0.1111,
+      "step": 2718
+    },
+    {
+      "epoch": 1.4154086413326392,
+      "grad_norm": 0.2196531969922503,
+      "learning_rate": 2.8487056977846083e-05,
+      "loss": 0.1122,
+      "step": 2719
+    },
+    {
+      "epoch": 1.415929203539823,
+      "grad_norm": 0.23260699506685206,
+      "learning_rate": 2.8473143716682455e-05,
+      "loss": 0.1157,
+      "step": 2720
+    },
+    {
+      "epoch": 1.4164497657470068,
+      "grad_norm": 0.23668138862358012,
+      "learning_rate": 2.8459229358538407e-05,
+      "loss": 0.1115,
+      "step": 2721
+    },
+    {
+      "epoch": 1.4169703279541905,
+      "grad_norm": 0.2246223176402912,
+      "learning_rate": 2.8445313907808756e-05,
+      "loss": 0.1084,
+      "step": 2722
+    },
+    {
+      "epoch": 1.4174908901613743,
+      "grad_norm": 0.2301496115534156,
+      "learning_rate": 2.843139736888864e-05,
+      "loss": 0.1076,
+      "step": 2723
+    },
+    {
+      "epoch": 1.418011452368558,
+      "grad_norm": 0.22756253474564722,
+      "learning_rate": 2.841747974617355e-05,
+      "loss": 0.1086,
+      "step": 2724
+    },
+    {
+      "epoch": 1.4185320145757419,
+      "grad_norm": 0.22250208266300767,
+      "learning_rate": 2.8403561044059324e-05,
+      "loss": 0.1086,
+      "step": 2725
+    },
+    {
+      "epoch": 1.4190525767829256,
+      "grad_norm": 0.2254005836949918,
+      "learning_rate": 2.8389641266942124e-05,
+      "loss": 0.104,
+      "step": 2726
+    },
+    {
+      "epoch": 1.4195731389901094,
+      "grad_norm": 0.23920028284992914,
+      "learning_rate": 2.8375720419218488e-05,
+      "loss": 0.1161,
+      "step": 2727
+    },
+    {
+      "epoch": 1.4200937011972932,
+      "grad_norm": 0.23743660758278437,
+      "learning_rate": 2.836179850528523e-05,
+      "loss": 0.1085,
+      "step": 2728
+    },
+    {
+      "epoch": 1.420614263404477,
+      "grad_norm": 0.22280403669098997,
+      "learning_rate": 2.8347875529539576e-05,
+      "loss": 0.1074,
+      "step": 2729
+    },
+    {
+      "epoch": 1.4211348256116607,
+      "grad_norm": 0.2304981362316628,
+      "learning_rate": 2.8333951496379023e-05,
+      "loss": 0.1071,
+      "step": 2730
+    },
+    {
+      "epoch": 1.4216553878188445,
+      "grad_norm": 0.23280393574418745,
+      "learning_rate": 2.8320026410201445e-05,
+      "loss": 0.1102,
+      "step": 2731
+    },
+    {
+      "epoch": 1.422175950026028,
+      "grad_norm": 0.23611313379761717,
+      "learning_rate": 2.8306100275405024e-05,
+      "loss": 0.1101,
+      "step": 2732
+    },
+    {
+      "epoch": 1.4226965122332118,
+      "grad_norm": 0.2241908878458328,
+      "learning_rate": 2.829217309638828e-05,
+      "loss": 0.1102,
+      "step": 2733
+    },
+    {
+      "epoch": 1.4232170744403956,
+      "grad_norm": 0.2447870254970879,
+      "learning_rate": 2.827824487755007e-05,
+      "loss": 0.1119,
+      "step": 2734
+    },
+    {
+      "epoch": 1.4237376366475794,
+      "grad_norm": 0.23587811793675378,
+      "learning_rate": 2.8264315623289568e-05,
+      "loss": 0.115,
+      "step": 2735
+    },
+    {
+      "epoch": 1.4242581988547631,
+      "grad_norm": 0.23445583906080528,
+      "learning_rate": 2.8250385338006297e-05,
+      "loss": 0.1127,
+      "step": 2736
+    },
+    {
+      "epoch": 1.424778761061947,
+      "grad_norm": 0.22467550766807312,
+      "learning_rate": 2.823645402610006e-05,
+      "loss": 0.1073,
+      "step": 2737
+    },
+    {
+      "epoch": 1.4252993232691307,
+      "grad_norm": 0.22942028138580609,
+      "learning_rate": 2.8222521691971037e-05,
+      "loss": 0.1081,
+      "step": 2738
+    },
+    {
+      "epoch": 1.4258198854763144,
+      "grad_norm": 0.2258282159083254,
+      "learning_rate": 2.8208588340019703e-05,
+      "loss": 0.1137,
+      "step": 2739
+    },
+    {
+      "epoch": 1.4263404476834982,
+      "grad_norm": 0.2385778073200138,
+      "learning_rate": 2.8194653974646858e-05,
+      "loss": 0.1173,
+      "step": 2740
+    },
+    {
+      "epoch": 1.426861009890682,
+      "grad_norm": 0.2279308225257302,
+      "learning_rate": 2.8180718600253613e-05,
+      "loss": 0.1118,
+      "step": 2741
+    },
+    {
+      "epoch": 1.4273815720978658,
+      "grad_norm": 0.23064677854327556,
+      "learning_rate": 2.8166782221241418e-05,
+      "loss": 0.112,
+      "step": 2742
+    },
+    {
+      "epoch": 1.4279021343050495,
+      "grad_norm": 0.23251063489281165,
+      "learning_rate": 2.8152844842012034e-05,
+      "loss": 0.1141,
+      "step": 2743
+    },
+    {
+      "epoch": 1.428422696512233,
+      "grad_norm": 0.23020617762610876,
+      "learning_rate": 2.8138906466967518e-05,
+      "loss": 0.1081,
+      "step": 2744
+    },
+    {
+      "epoch": 1.4289432587194169,
+      "grad_norm": 0.2222890423486755,
+      "learning_rate": 2.812496710051028e-05,
+      "loss": 0.1065,
+      "step": 2745
+    },
+    {
+      "epoch": 1.4294638209266006,
+      "grad_norm": 0.2274018551434016,
+      "learning_rate": 2.8111026747043002e-05,
+      "loss": 0.1057,
+      "step": 2746
+    },
+    {
+      "epoch": 1.4299843831337844,
+      "grad_norm": 0.236085240555113,
+      "learning_rate": 2.80970854109687e-05,
+      "loss": 0.1056,
+      "step": 2747
+    },
+    {
+      "epoch": 1.4305049453409682,
+      "grad_norm": 0.2224073810854335,
+      "learning_rate": 2.80831430966907e-05,
+      "loss": 0.1094,
+      "step": 2748
+    },
+    {
+      "epoch": 1.431025507548152,
+      "grad_norm": 0.22383924274617603,
+      "learning_rate": 2.806919980861264e-05,
+      "loss": 0.1106,
+      "step": 2749
+    },
+    {
+      "epoch": 1.4315460697553357,
+      "grad_norm": 0.23203392604509374,
+      "learning_rate": 2.805525555113845e-05,
+      "loss": 0.1113,
+      "step": 2750
+    },
+    {
+      "epoch": 1.4320666319625195,
+      "grad_norm": 0.22742895594575752,
+      "learning_rate": 2.804131032867237e-05,
+      "loss": 0.1094,
+      "step": 2751
+    },
+    {
+      "epoch": 1.4325871941697033,
+      "grad_norm": 0.22277598495191936,
+      "learning_rate": 2.8027364145618967e-05,
+      "loss": 0.1107,
+      "step": 2752
+    },
+    {
+      "epoch": 1.433107756376887,
+      "grad_norm": 0.23109249376380572,
+      "learning_rate": 2.8013417006383076e-05,
+      "loss": 0.1089,
+      "step": 2753
+    },
+    {
+      "epoch": 1.4336283185840708,
+      "grad_norm": 0.23662969693774688,
+      "learning_rate": 2.799946891536987e-05,
+      "loss": 0.1129,
+      "step": 2754
+    },
+    {
+      "epoch": 1.4341488807912546,
+      "grad_norm": 0.2260021154026626,
+      "learning_rate": 2.7985519876984795e-05,
+      "loss": 0.1082,
+      "step": 2755
+    },
+    {
+      "epoch": 1.4346694429984383,
+      "grad_norm": 0.23726551030073853,
+      "learning_rate": 2.7971569895633604e-05,
+      "loss": 0.114,
+      "step": 2756
+    },
+    {
+      "epoch": 1.4351900052056221,
+      "grad_norm": 0.23082581298133342,
+      "learning_rate": 2.7957618975722362e-05,
+      "loss": 0.1095,
+      "step": 2757
+    },
+    {
+      "epoch": 1.435710567412806,
+      "grad_norm": 0.24003960853374687,
+      "learning_rate": 2.7943667121657412e-05,
+      "loss": 0.1118,
+      "step": 2758
+    },
+    {
+      "epoch": 1.4362311296199897,
+      "grad_norm": 0.234089806557142,
+      "learning_rate": 2.7929714337845396e-05,
+      "loss": 0.1101,
+      "step": 2759
+    },
+    {
+      "epoch": 1.4367516918271734,
+      "grad_norm": 0.23113107032487773,
+      "learning_rate": 2.7915760628693256e-05,
+      "loss": 0.1101,
+      "step": 2760
+    },
+    {
+      "epoch": 1.4372722540343572,
+      "grad_norm": 0.2219497904207552,
+      "learning_rate": 2.7901805998608217e-05,
+      "loss": 0.1059,
+      "step": 2761
+    },
+    {
+      "epoch": 1.437792816241541,
+      "grad_norm": 0.22605082077104915,
+      "learning_rate": 2.7887850451997805e-05,
+      "loss": 0.1068,
+      "step": 2762
+    },
+    {
+      "epoch": 1.4383133784487248,
+      "grad_norm": 0.22649858856913815,
+      "learning_rate": 2.787389399326984e-05,
+      "loss": 0.1095,
+      "step": 2763
+    },
+    {
+      "epoch": 1.4388339406559083,
+      "grad_norm": 0.2368963632813257,
+      "learning_rate": 2.7859936626832407e-05,
+      "loss": 0.1098,
+      "step": 2764
+    },
+    {
+      "epoch": 1.439354502863092,
+      "grad_norm": 0.22188698169925974,
+      "learning_rate": 2.7845978357093895e-05,
+      "loss": 0.1114,
+      "step": 2765
+    },
+    {
+      "epoch": 1.4398750650702759,
+      "grad_norm": 0.22991662693273585,
+      "learning_rate": 2.7832019188462977e-05,
+      "loss": 0.1128,
+      "step": 2766
+    },
+    {
+      "epoch": 1.4403956272774596,
+      "grad_norm": 0.25211405079579136,
+      "learning_rate": 2.7818059125348616e-05,
+      "loss": 0.1215,
+      "step": 2767
+    },
+    {
+      "epoch": 1.4409161894846434,
+      "grad_norm": 0.24479114751525108,
+      "learning_rate": 2.7804098172160038e-05,
+      "loss": 0.1172,
+      "step": 2768
+    },
+    {
+      "epoch": 1.4414367516918272,
+      "grad_norm": 0.22826017987069053,
+      "learning_rate": 2.779013633330676e-05,
+      "loss": 0.1108,
+      "step": 2769
+    },
+    {
+      "epoch": 1.441957313899011,
+      "grad_norm": 0.2346902016087554,
+      "learning_rate": 2.7776173613198592e-05,
+      "loss": 0.107,
+      "step": 2770
+    },
+    {
+      "epoch": 1.4424778761061947,
+      "grad_norm": 0.2353334199015863,
+      "learning_rate": 2.7762210016245605e-05,
+      "loss": 0.1132,
+      "step": 2771
+    },
+    {
+      "epoch": 1.4429984383133785,
+      "grad_norm": 0.2153683891298237,
+      "learning_rate": 2.7748245546858155e-05,
+      "loss": 0.1088,
+      "step": 2772
+    },
+    {
+      "epoch": 1.4435190005205623,
+      "grad_norm": 0.23715391047211573,
+      "learning_rate": 2.7734280209446865e-05,
+      "loss": 0.1097,
+      "step": 2773
+    },
+    {
+      "epoch": 1.444039562727746,
+      "grad_norm": 0.23698576922062276,
+      "learning_rate": 2.7720314008422636e-05,
+      "loss": 0.1105,
+      "step": 2774
+    },
+    {
+      "epoch": 1.4445601249349298,
+      "grad_norm": 0.24630914098283596,
+      "learning_rate": 2.770634694819666e-05,
+      "loss": 0.1119,
+      "step": 2775
+    },
+    {
+      "epoch": 1.4450806871421134,
+      "grad_norm": 0.24175525799827727,
+      "learning_rate": 2.7692379033180376e-05,
+      "loss": 0.111,
+      "step": 2776
+    },
+    {
+      "epoch": 1.4456012493492971,
+      "grad_norm": 0.22721483418002544,
+      "learning_rate": 2.7678410267785492e-05,
+      "loss": 0.1094,
+      "step": 2777
+    },
+    {
+      "epoch": 1.446121811556481,
+      "grad_norm": 0.24330824085525268,
+      "learning_rate": 2.7664440656424014e-05,
+      "loss": 0.112,
+      "step": 2778
+    },
+    {
+      "epoch": 1.4466423737636647,
+      "grad_norm": 0.22575100884114555,
+      "learning_rate": 2.7650470203508177e-05,
+      "loss": 0.1048,
+      "step": 2779
+    },
+    {
+      "epoch": 1.4471629359708484,
+      "grad_norm": 0.2302645170045773,
+      "learning_rate": 2.7636498913450508e-05,
+      "loss": 0.1088,
+      "step": 2780
+    },
+    {
+      "epoch": 1.4476834981780322,
+      "grad_norm": 0.24802303184573393,
+      "learning_rate": 2.7622526790663795e-05,
+      "loss": 0.1163,
+      "step": 2781
+    },
+    {
+      "epoch": 1.448204060385216,
+      "grad_norm": 0.23957252091431663,
+      "learning_rate": 2.760855383956108e-05,
+      "loss": 0.1104,
+      "step": 2782
+    },
+    {
+      "epoch": 1.4487246225923998,
+      "grad_norm": 0.2226274711661623,
+      "learning_rate": 2.7594580064555664e-05,
+      "loss": 0.1073,
+      "step": 2783
+    },
+    {
+      "epoch": 1.4492451847995835,
+      "grad_norm": 0.22154143748354907,
+      "learning_rate": 2.7580605470061126e-05,
+      "loss": 0.1063,
+      "step": 2784
+    },
+    {
+      "epoch": 1.4497657470067673,
+      "grad_norm": 0.2213471609280991,
+      "learning_rate": 2.7566630060491288e-05,
+      "loss": 0.1068,
+      "step": 2785
+    },
+    {
+      "epoch": 1.450286309213951,
+      "grad_norm": 0.22697992240357193,
+      "learning_rate": 2.7552653840260234e-05,
+      "loss": 0.1122,
+      "step": 2786
+    },
+    {
+      "epoch": 1.4508068714211348,
+      "grad_norm": 0.23909687959750933,
+      "learning_rate": 2.7538676813782315e-05,
+      "loss": 0.1109,
+      "step": 2787
+    },
+    {
+      "epoch": 1.4513274336283186,
+      "grad_norm": 0.22865206395405765,
+      "learning_rate": 2.752469898547211e-05,
+      "loss": 0.111,
+      "step": 2788
+    },
+    {
+      "epoch": 1.4518479958355024,
+      "grad_norm": 0.230864102308563,
+      "learning_rate": 2.751072035974448e-05,
+      "loss": 0.1101,
+      "step": 2789
+    },
+    {
+      "epoch": 1.4523685580426862,
+      "grad_norm": 0.22847986253277383,
+      "learning_rate": 2.749674094101452e-05,
+      "loss": 0.1059,
+      "step": 2790
+    },
+    {
+      "epoch": 1.45288912024987,
+      "grad_norm": 0.22746929861816742,
+      "learning_rate": 2.748276073369759e-05,
+      "loss": 0.1077,
+      "step": 2791
+    },
+    {
+      "epoch": 1.4534096824570537,
+      "grad_norm": 0.2490634493102354,
+      "learning_rate": 2.7468779742209272e-05,
+      "loss": 0.1149,
+      "step": 2792
+    },
+    {
+      "epoch": 1.4539302446642375,
+      "grad_norm": 0.23550416923749126,
+      "learning_rate": 2.745479797096543e-05,
+      "loss": 0.1078,
+      "step": 2793
+    },
+    {
+      "epoch": 1.4544508068714213,
+      "grad_norm": 0.23727643035605134,
+      "learning_rate": 2.744081542438215e-05,
+      "loss": 0.1102,
+      "step": 2794
+    },
+    {
+      "epoch": 1.454971369078605,
+      "grad_norm": 0.23639067924333218,
+      "learning_rate": 2.7426832106875772e-05,
+      "loss": 0.1123,
+      "step": 2795
+    },
+    {
+      "epoch": 1.4554919312857886,
+      "grad_norm": 0.22586950416876225,
+      "learning_rate": 2.741284802286288e-05,
+      "loss": 0.1079,
+      "step": 2796
+    },
+    {
+      "epoch": 1.4560124934929723,
+      "grad_norm": 0.23428485678259794,
+      "learning_rate": 2.7398863176760297e-05,
+      "loss": 0.1103,
+      "step": 2797
+    },
+    {
+      "epoch": 1.4565330557001561,
+      "grad_norm": 0.23671222076156578,
+      "learning_rate": 2.7384877572985096e-05,
+      "loss": 0.1131,
+      "step": 2798
+    },
+    {
+      "epoch": 1.45705361790734,
+      "grad_norm": 0.24175271615548818,
+      "learning_rate": 2.7370891215954568e-05,
+      "loss": 0.1079,
+      "step": 2799
+    },
+    {
+      "epoch": 1.4575741801145237,
+      "grad_norm": 0.25201562246791964,
+      "learning_rate": 2.7356904110086267e-05,
+      "loss": 0.1139,
+      "step": 2800
+    },
+    {
+      "epoch": 1.4580947423217074,
+      "grad_norm": 0.23688810004618258,
+      "learning_rate": 2.7342916259797964e-05,
+      "loss": 0.114,
+      "step": 2801
+    },
+    {
+      "epoch": 1.4586153045288912,
+      "grad_norm": 0.25496678426814795,
+      "learning_rate": 2.7328927669507675e-05,
+      "loss": 0.1123,
+      "step": 2802
+    },
+    {
+      "epoch": 1.459135866736075,
+      "grad_norm": 0.24119391804523413,
+      "learning_rate": 2.7314938343633656e-05,
+      "loss": 0.1117,
+      "step": 2803
+    },
+    {
+      "epoch": 1.4596564289432588,
+      "grad_norm": 0.23981545441457577,
+      "learning_rate": 2.7300948286594373e-05,
+      "loss": 0.1138,
+      "step": 2804
+    },
+    {
+      "epoch": 1.4601769911504425,
+      "grad_norm": 0.238290521313305,
+      "learning_rate": 2.7286957502808546e-05,
+      "loss": 0.1088,
+      "step": 2805
+    },
+    {
+      "epoch": 1.4606975533576263,
+      "grad_norm": 0.23496530630575055,
+      "learning_rate": 2.7272965996695116e-05,
+      "loss": 0.1091,
+      "step": 2806
+    },
+    {
+      "epoch": 1.46121811556481,
+      "grad_norm": 0.2219092228421417,
+      "learning_rate": 2.7258973772673247e-05,
+      "loss": 0.1069,
+      "step": 2807
+    },
+    {
+      "epoch": 1.4617386777719936,
+      "grad_norm": 0.23788495609882543,
+      "learning_rate": 2.7244980835162342e-05,
+      "loss": 0.1059,
+      "step": 2808
+    },
+    {
+      "epoch": 1.4622592399791774,
+      "grad_norm": 0.22578800709661323,
+      "learning_rate": 2.7230987188582008e-05,
+      "loss": 0.1111,
+      "step": 2809
+    },
+    {
+      "epoch": 1.4627798021863612,
+      "grad_norm": 0.2642738917819999,
+      "learning_rate": 2.7216992837352108e-05,
+      "loss": 0.111,
+      "step": 2810
+    },
+    {
+      "epoch": 1.463300364393545,
+      "grad_norm": 0.23319413750842335,
+      "learning_rate": 2.7202997785892688e-05,
+      "loss": 0.1102,
+      "step": 2811
+    },
+    {
+      "epoch": 1.4638209266007287,
+      "grad_norm": 0.25360100989894624,
+      "learning_rate": 2.718900203862406e-05,
+      "loss": 0.1128,
+      "step": 2812
+    },
+    {
+      "epoch": 1.4643414888079125,
+      "grad_norm": 0.2639966129821883,
+      "learning_rate": 2.7175005599966718e-05,
+      "loss": 0.1165,
+      "step": 2813
+    },
+    {
+      "epoch": 1.4648620510150963,
+      "grad_norm": 0.22862001367725718,
+      "learning_rate": 2.7161008474341393e-05,
+      "loss": 0.1059,
+      "step": 2814
+    },
+    {
+      "epoch": 1.46538261322228,
+      "grad_norm": 0.24334822660492816,
+      "learning_rate": 2.714701066616902e-05,
+      "loss": 0.1124,
+      "step": 2815
+    },
+    {
+      "epoch": 1.4659031754294638,
+      "grad_norm": 0.23329003664914877,
+      "learning_rate": 2.713301217987077e-05,
+      "loss": 0.1096,
+      "step": 2816
+    },
+    {
+      "epoch": 1.4664237376366476,
+      "grad_norm": 0.24018365846353792,
+      "learning_rate": 2.7119013019868013e-05,
+      "loss": 0.1137,
+      "step": 2817
+    },
+    {
+      "epoch": 1.4669442998438313,
+      "grad_norm": 0.22712008229642378,
+      "learning_rate": 2.710501319058233e-05,
+      "loss": 0.1098,
+      "step": 2818
+    },
+    {
+      "epoch": 1.4674648620510151,
+      "grad_norm": 0.22145732141152524,
+      "learning_rate": 2.7091012696435525e-05,
+      "loss": 0.1097,
+      "step": 2819
+    },
+    {
+      "epoch": 1.467985424258199,
+      "grad_norm": 0.22548306513700106,
+      "learning_rate": 2.70770115418496e-05,
+      "loss": 0.1084,
+      "step": 2820
+    },
+    {
+      "epoch": 1.4685059864653827,
+      "grad_norm": 0.2316717588456251,
+      "learning_rate": 2.706300973124678e-05,
+      "loss": 0.1119,
+      "step": 2821
+    },
+    {
+      "epoch": 1.4690265486725664,
+      "grad_norm": 0.24444595497147328,
+      "learning_rate": 2.7049007269049483e-05,
+      "loss": 0.112,
+      "step": 2822
+    },
+    {
+      "epoch": 1.4695471108797502,
+      "grad_norm": 0.223605647603151,
+      "learning_rate": 2.7035004159680332e-05,
+      "loss": 0.1069,
+      "step": 2823
+    },
+    {
+      "epoch": 1.470067673086934,
+      "grad_norm": 0.22488128372014973,
+      "learning_rate": 2.702100040756217e-05,
+      "loss": 0.1057,
+      "step": 2824
+    },
+    {
+      "epoch": 1.4705882352941178,
+      "grad_norm": 0.24269771382490615,
+      "learning_rate": 2.700699601711803e-05,
+      "loss": 0.111,
+      "step": 2825
+    },
+    {
+      "epoch": 1.4711087975013015,
+      "grad_norm": 0.23271593883892053,
+      "learning_rate": 2.699299099277115e-05,
+      "loss": 0.1034,
+      "step": 2826
+    },
+    {
+      "epoch": 1.4716293597084853,
+      "grad_norm": 0.2238190756200398,
+      "learning_rate": 2.6978985338944966e-05,
+      "loss": 0.1096,
+      "step": 2827
+    },
+    {
+      "epoch": 1.4721499219156688,
+      "grad_norm": 0.2466339353589823,
+      "learning_rate": 2.6964979060063123e-05,
+      "loss": 0.1116,
+      "step": 2828
+    },
+    {
+      "epoch": 1.4726704841228526,
+      "grad_norm": 0.2540666543177738,
+      "learning_rate": 2.6950972160549444e-05,
+      "loss": 0.1133,
+      "step": 2829
+    },
+    {
+      "epoch": 1.4731910463300364,
+      "grad_norm": 0.2250856415814824,
+      "learning_rate": 2.6936964644827973e-05,
+      "loss": 0.1078,
+      "step": 2830
+    },
+    {
+      "epoch": 1.4737116085372202,
+      "grad_norm": 0.23250731739398878,
+      "learning_rate": 2.692295651732293e-05,
+      "loss": 0.1102,
+      "step": 2831
+    },
+    {
+      "epoch": 1.474232170744404,
+      "grad_norm": 0.24620466821834674,
+      "learning_rate": 2.6908947782458728e-05,
+      "loss": 0.1111,
+      "step": 2832
+    },
+    {
+      "epoch": 1.4747527329515877,
+      "grad_norm": 0.2245002310950658,
+      "learning_rate": 2.6894938444659974e-05,
+      "loss": 0.1032,
+      "step": 2833
+    },
+    {
+      "epoch": 1.4752732951587715,
+      "grad_norm": 0.2327464778529271,
+      "learning_rate": 2.6880928508351484e-05,
+      "loss": 0.1098,
+      "step": 2834
+    },
+    {
+      "epoch": 1.4757938573659553,
+      "grad_norm": 0.2376686699449246,
+      "learning_rate": 2.6866917977958246e-05,
+      "loss": 0.1101,
+      "step": 2835
+    },
+    {
+      "epoch": 1.476314419573139,
+      "grad_norm": 0.2413687823738175,
+      "learning_rate": 2.685290685790542e-05,
+      "loss": 0.1125,
+      "step": 2836
+    },
+    {
+      "epoch": 1.4768349817803228,
+      "grad_norm": 0.23618732799015563,
+      "learning_rate": 2.6838895152618387e-05,
+      "loss": 0.112,
+      "step": 2837
+    },
+    {
+      "epoch": 1.4773555439875066,
+      "grad_norm": 0.22955043165979686,
+      "learning_rate": 2.682488286652269e-05,
+      "loss": 0.1059,
+      "step": 2838
+    },
+    {
+      "epoch": 1.4778761061946903,
+      "grad_norm": 0.27218809301190455,
+      "learning_rate": 2.6810870004044063e-05,
+      "loss": 0.1124,
+      "step": 2839
+    },
+    {
+      "epoch": 1.478396668401874,
+      "grad_norm": 0.2370646380825962,
+      "learning_rate": 2.6796856569608414e-05,
+      "loss": 0.1144,
+      "step": 2840
+    },
+    {
+      "epoch": 1.4789172306090577,
+      "grad_norm": 0.2379047779082474,
+      "learning_rate": 2.6782842567641842e-05,
+      "loss": 0.107,
+      "step": 2841
+    },
+    {
+      "epoch": 1.4794377928162414,
+      "grad_norm": 0.24483160304496035,
+      "learning_rate": 2.6768828002570623e-05,
+      "loss": 0.1099,
+      "step": 2842
+    },
+    {
+      "epoch": 1.4799583550234252,
+      "grad_norm": 0.2323379686773366,
+      "learning_rate": 2.6754812878821206e-05,
+      "loss": 0.1085,
+      "step": 2843
+    },
+    {
+      "epoch": 1.480478917230609,
+      "grad_norm": 0.23946063633545,
+      "learning_rate": 2.6740797200820228e-05,
+      "loss": 0.1096,
+      "step": 2844
+    },
+    {
+      "epoch": 1.4809994794377928,
+      "grad_norm": 0.24336186169601903,
+      "learning_rate": 2.672678097299447e-05,
+      "loss": 0.1111,
+      "step": 2845
+    },
+    {
+      "epoch": 1.4815200416449765,
+      "grad_norm": 0.23709636885936694,
+      "learning_rate": 2.6712764199770936e-05,
+      "loss": 0.1129,
+      "step": 2846
+    },
+    {
+      "epoch": 1.4820406038521603,
+      "grad_norm": 0.23972559290010753,
+      "learning_rate": 2.6698746885576746e-05,
+      "loss": 0.1137,
+      "step": 2847
+    },
+    {
+      "epoch": 1.482561166059344,
+      "grad_norm": 0.23976136195183834,
+      "learning_rate": 2.668472903483925e-05,
+      "loss": 0.1121,
+      "step": 2848
+    },
+    {
+      "epoch": 1.4830817282665278,
+      "grad_norm": 0.24400625472890314,
+      "learning_rate": 2.6670710651985924e-05,
+      "loss": 0.1135,
+      "step": 2849
+    },
+    {
+      "epoch": 1.4836022904737116,
+      "grad_norm": 0.2367697438563428,
+      "learning_rate": 2.6656691741444423e-05,
+      "loss": 0.107,
+      "step": 2850
+    },
+    {
+      "epoch": 1.4841228526808954,
+      "grad_norm": 0.23848746621745118,
+      "learning_rate": 2.6642672307642575e-05,
+      "loss": 0.1079,
+      "step": 2851
+    },
+    {
+      "epoch": 1.4846434148880792,
+      "grad_norm": 0.23917458954785298,
+      "learning_rate": 2.6628652355008364e-05,
+      "loss": 0.1135,
+      "step": 2852
+    },
+    {
+      "epoch": 1.485163977095263,
+      "grad_norm": 0.23079110138661027,
+      "learning_rate": 2.661463188796996e-05,
+      "loss": 0.113,
+      "step": 2853
+    },
+    {
+      "epoch": 1.4856845393024467,
+      "grad_norm": 0.22660723854160528,
+      "learning_rate": 2.6600610910955652e-05,
+      "loss": 0.1084,
+      "step": 2854
+    },
+    {
+      "epoch": 1.4862051015096305,
+      "grad_norm": 0.24763834613940575,
+      "learning_rate": 2.6586589428393944e-05,
+      "loss": 0.1139,
+      "step": 2855
+    },
+    {
+      "epoch": 1.4867256637168142,
+      "grad_norm": 0.22651685710665284,
+      "learning_rate": 2.6572567444713453e-05,
+      "loss": 0.1088,
+      "step": 2856
+    },
+    {
+      "epoch": 1.487246225923998,
+      "grad_norm": 0.23203385277212307,
+      "learning_rate": 2.655854496434299e-05,
+      "loss": 0.1082,
+      "step": 2857
+    },
+    {
+      "epoch": 1.4877667881311818,
+      "grad_norm": 0.23006346404589428,
+      "learning_rate": 2.6544521991711498e-05,
+      "loss": 0.109,
+      "step": 2858
+    },
+    {
+      "epoch": 1.4882873503383656,
+      "grad_norm": 0.24467746804171445,
+      "learning_rate": 2.6530498531248078e-05,
+      "loss": 0.1078,
+      "step": 2859
+    },
+    {
+      "epoch": 1.4888079125455491,
+      "grad_norm": 0.2376855115614867,
+      "learning_rate": 2.6516474587382002e-05,
+      "loss": 0.1038,
+      "step": 2860
+    },
+    {
+      "epoch": 1.489328474752733,
+      "grad_norm": 0.2278413610792057,
+      "learning_rate": 2.650245016454268e-05,
+      "loss": 0.1084,
+      "step": 2861
+    },
+    {
+      "epoch": 1.4898490369599167,
+      "grad_norm": 0.24707716899895538,
+      "learning_rate": 2.6488425267159688e-05,
+      "loss": 0.1044,
+      "step": 2862
+    },
+    {
+      "epoch": 1.4903695991671004,
+      "grad_norm": 0.2587262411800931,
+      "learning_rate": 2.6474399899662715e-05,
+      "loss": 0.1116,
+      "step": 2863
+    },
+    {
+      "epoch": 1.4908901613742842,
+      "grad_norm": 0.2435423020073306,
+      "learning_rate": 2.646037406648165e-05,
+      "loss": 0.1115,
+      "step": 2864
+    },
+    {
+      "epoch": 1.491410723581468,
+      "grad_norm": 0.23050479938532595,
+      "learning_rate": 2.6446347772046492e-05,
+      "loss": 0.1084,
+      "step": 2865
+    },
+    {
+      "epoch": 1.4919312857886518,
+      "grad_norm": 0.2374569801662734,
+      "learning_rate": 2.6432321020787403e-05,
+      "loss": 0.1088,
+      "step": 2866
+    },
+    {
+      "epoch": 1.4924518479958355,
+      "grad_norm": 0.2312604690786547,
+      "learning_rate": 2.641829381713468e-05,
+      "loss": 0.1109,
+      "step": 2867
+    },
+    {
+      "epoch": 1.4929724102030193,
+      "grad_norm": 0.23090596721393158,
+      "learning_rate": 2.6404266165518767e-05,
+      "loss": 0.114,
+      "step": 2868
+    },
+    {
+      "epoch": 1.493492972410203,
+      "grad_norm": 0.23229013774491258,
+      "learning_rate": 2.6390238070370255e-05,
+      "loss": 0.108,
+      "step": 2869
+    },
+    {
+      "epoch": 1.4940135346173868,
+      "grad_norm": 0.22185984755752802,
+      "learning_rate": 2.6376209536119856e-05,
+      "loss": 0.1074,
+      "step": 2870
+    },
+    {
+      "epoch": 1.4945340968245706,
+      "grad_norm": 0.22282503223676492,
+      "learning_rate": 2.6362180567198447e-05,
+      "loss": 0.1084,
+      "step": 2871
+    },
+    {
+      "epoch": 1.4950546590317542,
+      "grad_norm": 0.22439529261749314,
+      "learning_rate": 2.6348151168037028e-05,
+      "loss": 0.106,
+      "step": 2872
+    },
+    {
+      "epoch": 1.495575221238938,
+      "grad_norm": 0.23223806149855644,
+      "learning_rate": 2.633412134306672e-05,
+      "loss": 0.1107,
+      "step": 2873
+    },
+    {
+      "epoch": 1.4960957834461217,
+      "grad_norm": 0.23202694735599594,
+      "learning_rate": 2.6320091096718817e-05,
+      "loss": 0.1108,
+      "step": 2874
+    },
+    {
+      "epoch": 1.4966163456533055,
+      "grad_norm": 0.22150351815304223,
+      "learning_rate": 2.6306060433424712e-05,
+      "loss": 0.1089,
+      "step": 2875
+    },
+    {
+      "epoch": 1.4971369078604893,
+      "grad_norm": 0.2356343347877159,
+      "learning_rate": 2.629202935761595e-05,
+      "loss": 0.1066,
+      "step": 2876
+    },
+    {
+      "epoch": 1.497657470067673,
+      "grad_norm": 0.22946145700792592,
+      "learning_rate": 2.6277997873724182e-05,
+      "loss": 0.1092,
+      "step": 2877
+    },
+    {
+      "epoch": 1.4981780322748568,
+      "grad_norm": 0.24032297573317024,
+      "learning_rate": 2.6263965986181215e-05,
+      "loss": 0.1172,
+      "step": 2878
+    },
+    {
+      "epoch": 1.4986985944820406,
+      "grad_norm": 0.23337639493576984,
+      "learning_rate": 2.6249933699418965e-05,
+      "loss": 0.1092,
+      "step": 2879
+    },
+    {
+      "epoch": 1.4992191566892243,
+      "grad_norm": 0.23600113599122988,
+      "learning_rate": 2.6235901017869495e-05,
+      "loss": 0.1109,
+      "step": 2880
+    },
+    {
+      "epoch": 1.4997397188964081,
+      "grad_norm": 0.2239919599001527,
+      "learning_rate": 2.6221867945964966e-05,
+      "loss": 0.107,
+      "step": 2881
+    },
+    {
+      "epoch": 1.5002602811035919,
+      "grad_norm": 0.2438916295121852,
+      "learning_rate": 2.6207834488137677e-05,
+      "loss": 0.1071,
+      "step": 2882
+    },
+    {
+      "epoch": 1.5007808433107757,
+      "grad_norm": 0.24034447143613874,
+      "learning_rate": 2.6193800648820055e-05,
+      "loss": 0.1119,
+      "step": 2883
+    },
+    {
+      "epoch": 1.5013014055179594,
+      "grad_norm": 0.21632247281172787,
+      "learning_rate": 2.6179766432444624e-05,
+      "loss": 0.105,
+      "step": 2884
+    },
+    {
+      "epoch": 1.5018219677251432,
+      "grad_norm": 0.23446966101143912,
+      "learning_rate": 2.6165731843444063e-05,
+      "loss": 0.1053,
+      "step": 2885
+    },
+    {
+      "epoch": 1.502342529932327,
+      "grad_norm": 0.23530370722928862,
+      "learning_rate": 2.6151696886251126e-05,
+      "loss": 0.1079,
+      "step": 2886
+    },
+    {
+      "epoch": 1.5028630921395107,
+      "grad_norm": 0.2489383930332058,
+      "learning_rate": 2.6137661565298726e-05,
+      "loss": 0.1121,
+      "step": 2887
+    },
+    {
+      "epoch": 1.5033836543466945,
+      "grad_norm": 0.24643956447343202,
+      "learning_rate": 2.6123625885019854e-05,
+      "loss": 0.1091,
+      "step": 2888
+    },
+    {
+      "epoch": 1.5039042165538783,
+      "grad_norm": 0.23991702848744126,
+      "learning_rate": 2.6109589849847643e-05,
+      "loss": 0.1164,
+      "step": 2889
+    },
+    {
+      "epoch": 1.504424778761062,
+      "grad_norm": 0.2381158559586777,
+      "learning_rate": 2.6095553464215322e-05,
+      "loss": 0.1087,
+      "step": 2890
+    },
+    {
+      "epoch": 1.5049453409682458,
+      "grad_norm": 0.2264207570228489,
+      "learning_rate": 2.6081516732556226e-05,
+      "loss": 0.1077,
+      "step": 2891
+    },
+    {
+      "epoch": 1.5054659031754296,
+      "grad_norm": 0.23184141890912552,
+      "learning_rate": 2.606747965930383e-05,
+      "loss": 0.1093,
+      "step": 2892
+    },
+    {
+      "epoch": 1.5059864653826134,
+      "grad_norm": 0.2300329290920324,
+      "learning_rate": 2.605344224889167e-05,
+      "loss": 0.1114,
+      "step": 2893
+    },
+    {
+      "epoch": 1.506507027589797,
+      "grad_norm": 0.23186346307565206,
+      "learning_rate": 2.6039404505753433e-05,
+      "loss": 0.1052,
+      "step": 2894
+    },
+    {
+      "epoch": 1.5070275897969807,
+      "grad_norm": 0.22392258697887885,
+      "learning_rate": 2.602536643432288e-05,
+      "loss": 0.1041,
+      "step": 2895
+    },
+    {
+      "epoch": 1.5075481520041645,
+      "grad_norm": 0.24950950930317453,
+      "learning_rate": 2.601132803903389e-05,
+      "loss": 0.1097,
+      "step": 2896
+    },
+    {
+      "epoch": 1.5080687142113482,
+      "grad_norm": 0.22767870898021633,
+      "learning_rate": 2.599728932432044e-05,
+      "loss": 0.1049,
+      "step": 2897
+    },
+    {
+      "epoch": 1.508589276418532,
+      "grad_norm": 0.23533974509856762,
+      "learning_rate": 2.5983250294616618e-05,
+      "loss": 0.108,
+      "step": 2898
+    },
+    {
+      "epoch": 1.5091098386257158,
+      "grad_norm": 0.24447947730787556,
+      "learning_rate": 2.596921095435659e-05,
+      "loss": 0.1109,
+      "step": 2899
+    },
+    {
+      "epoch": 1.5096304008328996,
+      "grad_norm": 0.24152156647479783,
+      "learning_rate": 2.595517130797464e-05,
+      "loss": 0.1164,
+      "step": 2900
+    },
+    {
+      "epoch": 1.5101509630400833,
+      "grad_norm": 0.23201234882111582,
+      "learning_rate": 2.5941131359905146e-05,
+      "loss": 0.1057,
+      "step": 2901
+    },
+    {
+      "epoch": 1.5106715252472669,
+      "grad_norm": 0.22589534043240497,
+      "learning_rate": 2.5927091114582565e-05,
+      "loss": 0.1099,
+      "step": 2902
+    },
+    {
+      "epoch": 1.5111920874544507,
+      "grad_norm": 0.22101671654491578,
+      "learning_rate": 2.5913050576441477e-05,
+      "loss": 0.109,
+      "step": 2903
+    },
+    {
+      "epoch": 1.5117126496616344,
+      "grad_norm": 0.2186466588558665,
+      "learning_rate": 2.589900974991652e-05,
+      "loss": 0.1088,
+      "step": 2904
+    },
+    {
+      "epoch": 1.5122332118688182,
+      "grad_norm": 0.22498932105359454,
+      "learning_rate": 2.588496863944244e-05,
+      "loss": 0.1034,
+      "step": 2905
+    },
+    {
+      "epoch": 1.512753774076002,
+      "grad_norm": 0.21166235287356577,
+      "learning_rate": 2.5870927249454097e-05,
+      "loss": 0.1037,
+      "step": 2906
+    },
+    {
+      "epoch": 1.5132743362831858,
+      "grad_norm": 0.22559742909440408,
+      "learning_rate": 2.5856885584386393e-05,
+      "loss": 0.1045,
+      "step": 2907
+    },
+    {
+      "epoch": 1.5137948984903695,
+      "grad_norm": 0.22659709971519004,
+      "learning_rate": 2.584284364867435e-05,
+      "loss": 0.106,
+      "step": 2908
+    },
+    {
+      "epoch": 1.5143154606975533,
+      "grad_norm": 0.24273341457463493,
+      "learning_rate": 2.582880144675305e-05,
+      "loss": 0.1155,
+      "step": 2909
+    },
+    {
+      "epoch": 1.514836022904737,
+      "grad_norm": 0.23888844864219327,
+      "learning_rate": 2.5814758983057684e-05,
+      "loss": 0.1058,
+      "step": 2910
+    },
+    {
+      "epoch": 1.5153565851119208,
+      "grad_norm": 0.24063817042473679,
+      "learning_rate": 2.5800716262023515e-05,
+      "loss": 0.1078,
+      "step": 2911
+    },
+    {
+      "epoch": 1.5158771473191046,
+      "grad_norm": 0.2403037315106119,
+      "learning_rate": 2.5786673288085898e-05,
+      "loss": 0.1127,
+      "step": 2912
+    },
+    {
+      "epoch": 1.5163977095262884,
+      "grad_norm": 0.2273826223314166,
+      "learning_rate": 2.577263006568025e-05,
+      "loss": 0.1064,
+      "step": 2913
+    },
+    {
+      "epoch": 1.5169182717334722,
+      "grad_norm": 0.21909822724724917,
+      "learning_rate": 2.5758586599242057e-05,
+      "loss": 0.1036,
+      "step": 2914
+    },
+    {
+      "epoch": 1.517438833940656,
+      "grad_norm": 0.23969847647923984,
+      "learning_rate": 2.5744542893206924e-05,
+      "loss": 0.1106,
+      "step": 2915
+    },
+    {
+      "epoch": 1.5179593961478397,
+      "grad_norm": 0.24016614079483722,
+      "learning_rate": 2.5730498952010502e-05,
+      "loss": 0.1074,
+      "step": 2916
+    },
+    {
+      "epoch": 1.5184799583550235,
+      "grad_norm": 0.2308633624750452,
+      "learning_rate": 2.5716454780088512e-05,
+      "loss": 0.1115,
+      "step": 2917
+    },
+    {
+      "epoch": 1.5190005205622072,
+      "grad_norm": 0.23198908006842925,
+      "learning_rate": 2.570241038187675e-05,
+      "loss": 0.1128,
+      "step": 2918
+    },
+    {
+      "epoch": 1.519521082769391,
+      "grad_norm": 0.23442644263464943,
+      "learning_rate": 2.5688365761811116e-05,
+      "loss": 0.1049,
+      "step": 2919
+    },
+    {
+      "epoch": 1.5200416449765748,
+      "grad_norm": 0.22883414451848424,
+      "learning_rate": 2.5674320924327533e-05,
+      "loss": 0.1097,
+      "step": 2920
+    },
+    {
+      "epoch": 1.5205622071837586,
+      "grad_norm": 0.2361125411690794,
+      "learning_rate": 2.566027587386203e-05,
+      "loss": 0.1072,
+      "step": 2921
+    },
+    {
+      "epoch": 1.5210827693909423,
+      "grad_norm": 0.23120541766705457,
+      "learning_rate": 2.5646230614850673e-05,
+      "loss": 0.1082,
+      "step": 2922
+    },
+    {
+      "epoch": 1.521603331598126,
+      "grad_norm": 0.22388494741212342,
+      "learning_rate": 2.5632185151729616e-05,
+      "loss": 0.1091,
+      "step": 2923
+    },
+    {
+      "epoch": 1.5221238938053099,
+      "grad_norm": 0.2260163271545998,
+      "learning_rate": 2.5618139488935072e-05,
+      "loss": 0.1066,
+      "step": 2924
+    },
+    {
+      "epoch": 1.5226444560124937,
+      "grad_norm": 0.2365959253566713,
+      "learning_rate": 2.5604093630903307e-05,
+      "loss": 0.1147,
+      "step": 2925
+    },
+    {
+      "epoch": 1.5231650182196772,
+      "grad_norm": 0.2195917844320001,
+      "learning_rate": 2.559004758207067e-05,
+      "loss": 0.1016,
+      "step": 2926
+    },
+    {
+      "epoch": 1.523685580426861,
+      "grad_norm": 0.22196040442631998,
+      "learning_rate": 2.557600134687354e-05,
+      "loss": 0.1106,
+      "step": 2927
+    },
+    {
+      "epoch": 1.5242061426340447,
+      "grad_norm": 0.2249212499365945,
+      "learning_rate": 2.5561954929748382e-05,
+      "loss": 0.1091,
+      "step": 2928
+    },
+    {
+      "epoch": 1.5247267048412285,
+      "grad_norm": 0.21665296060051842,
+      "learning_rate": 2.5547908335131704e-05,
+      "loss": 0.1066,
+      "step": 2929
+    },
+    {
+      "epoch": 1.5252472670484123,
+      "grad_norm": 0.23396144154716453,
+      "learning_rate": 2.5533861567460077e-05,
+      "loss": 0.1122,
+      "step": 2930
+    },
+    {
+      "epoch": 1.525767829255596,
+      "grad_norm": 0.23254614834471804,
+      "learning_rate": 2.5519814631170125e-05,
+      "loss": 0.1099,
+      "step": 2931
+    },
+    {
+      "epoch": 1.5262883914627798,
+      "grad_norm": 0.2313903436634836,
+      "learning_rate": 2.550576753069852e-05,
+      "loss": 0.111,
+      "step": 2932
+    },
+    {
+      "epoch": 1.5268089536699636,
+      "grad_norm": 0.22589258005003615,
+      "learning_rate": 2.5491720270481994e-05,
+      "loss": 0.1086,
+      "step": 2933
+    },
+    {
+      "epoch": 1.5273295158771472,
+      "grad_norm": 0.24350507728473914,
+      "learning_rate": 2.5477672854957325e-05,
+      "loss": 0.1149,
+      "step": 2934
+    },
+    {
+      "epoch": 1.527850078084331,
+      "grad_norm": 0.2296799862656636,
+      "learning_rate": 2.5463625288561343e-05,
+      "loss": 0.1095,
+      "step": 2935
+    },
+    {
+      "epoch": 1.5283706402915147,
+      "grad_norm": 0.21068213738843436,
+      "learning_rate": 2.5449577575730908e-05,
+      "loss": 0.1037,
+      "step": 2936
+    },
+    {
+      "epoch": 1.5288912024986985,
+      "grad_norm": 0.2282417696222234,
+      "learning_rate": 2.5435529720902955e-05,
+      "loss": 0.1078,
+      "step": 2937
+    },
+    {
+      "epoch": 1.5294117647058822,
+      "grad_norm": 0.2287731855036949,
+      "learning_rate": 2.5421481728514456e-05,
+      "loss": 0.1077,
+      "step": 2938
+    },
+    {
+      "epoch": 1.529932326913066,
+      "grad_norm": 0.24416895094839683,
+      "learning_rate": 2.5407433603002417e-05,
+      "loss": 0.116,
+      "step": 2939
+    },
+    {
+      "epoch": 1.5304528891202498,
+      "grad_norm": 0.2254694810188856,
+      "learning_rate": 2.5393385348803877e-05,
+      "loss": 0.1101,
+      "step": 2940
+    },
+    {
+      "epoch": 1.5309734513274336,
+      "grad_norm": 0.22885825764719317,
+      "learning_rate": 2.5379336970355938e-05,
+      "loss": 0.1105,
+      "step": 2941
+    },
+    {
+      "epoch": 1.5314940135346173,
+      "grad_norm": 0.2240956766046063,
+      "learning_rate": 2.5365288472095734e-05,
+      "loss": 0.1105,
+      "step": 2942
+    },
+    {
+      "epoch": 1.532014575741801,
+      "grad_norm": 0.22399611393029464,
+      "learning_rate": 2.5351239858460423e-05,
+      "loss": 0.1116,
+      "step": 2943
+    },
+    {
+      "epoch": 1.5325351379489849,
+      "grad_norm": 0.2291513822745778,
+      "learning_rate": 2.5337191133887232e-05,
+      "loss": 0.1106,
+      "step": 2944
+    },
+    {
+      "epoch": 1.5330557001561687,
+      "grad_norm": 0.2278099394652461,
+      "learning_rate": 2.5323142302813384e-05,
+      "loss": 0.1108,
+      "step": 2945
+    },
+    {
+      "epoch": 1.5335762623633524,
+      "grad_norm": 0.23495759529423726,
+      "learning_rate": 2.5309093369676158e-05,
+      "loss": 0.1122,
+      "step": 2946
+    },
+    {
+      "epoch": 1.5340968245705362,
+      "grad_norm": 0.22433879417057873,
+      "learning_rate": 2.529504433891286e-05,
+      "loss": 0.1094,
+      "step": 2947
+    },
+    {
+      "epoch": 1.53461738677772,
+      "grad_norm": 0.2312010873910697,
+      "learning_rate": 2.5280995214960835e-05,
+      "loss": 0.1108,
+      "step": 2948
+    },
+    {
+      "epoch": 1.5351379489849037,
+      "grad_norm": 0.22683977448776502,
+      "learning_rate": 2.5266946002257447e-05,
+      "loss": 0.1078,
+      "step": 2949
+    },
+    {
+      "epoch": 1.5356585111920875,
+      "grad_norm": 0.24008434305692997,
+      "learning_rate": 2.525289670524008e-05,
+      "loss": 0.1058,
+      "step": 2950
+    },
+    {
+      "epoch": 1.5361790733992713,
+      "grad_norm": 0.2276628830953262,
+      "learning_rate": 2.523884732834617e-05,
+      "loss": 0.1156,
+      "step": 2951
+    },
+    {
+      "epoch": 1.536699635606455,
+      "grad_norm": 0.23018808698410154,
+      "learning_rate": 2.522479787601315e-05,
+      "loss": 0.1107,
+      "step": 2952
+    },
+    {
+      "epoch": 1.5372201978136388,
+      "grad_norm": 0.22488086495321882,
+      "learning_rate": 2.521074835267851e-05,
+      "loss": 0.1084,
+      "step": 2953
+    },
+    {
+      "epoch": 1.5377407600208226,
+      "grad_norm": 0.23467533242982103,
+      "learning_rate": 2.519669876277973e-05,
+      "loss": 0.11,
+      "step": 2954
+    },
+    {
+      "epoch": 1.5382613222280064,
+      "grad_norm": 0.24014862239225046,
+      "learning_rate": 2.5182649110754324e-05,
+      "loss": 0.117,
+      "step": 2955
+    },
+    {
+      "epoch": 1.5387818844351902,
+      "grad_norm": 0.2313151546820494,
+      "learning_rate": 2.5168599401039833e-05,
+      "loss": 0.1118,
+      "step": 2956
+    },
+    {
+      "epoch": 1.539302446642374,
+      "grad_norm": 0.24807712232283863,
+      "learning_rate": 2.515454963807381e-05,
+      "loss": 0.1088,
+      "step": 2957
+    },
+    {
+      "epoch": 1.5398230088495575,
+      "grad_norm": 0.22300185095946995,
+      "learning_rate": 2.514049982629381e-05,
+      "loss": 0.107,
+      "step": 2958
+    },
+    {
+      "epoch": 1.5403435710567412,
+      "grad_norm": 0.24240071957864304,
+      "learning_rate": 2.5126449970137427e-05,
+      "loss": 0.1118,
+      "step": 2959
+    },
+    {
+      "epoch": 1.540864133263925,
+      "grad_norm": 0.22785169569105423,
+      "learning_rate": 2.5112400074042264e-05,
+      "loss": 0.1097,
+      "step": 2960
+    },
+    {
+      "epoch": 1.5413846954711088,
+      "grad_norm": 0.23607456456416684,
+      "learning_rate": 2.509835014244592e-05,
+      "loss": 0.1047,
+      "step": 2961
+    },
+    {
+      "epoch": 1.5419052576782926,
+      "grad_norm": 0.24205109765791188,
+      "learning_rate": 2.5084300179786036e-05,
+      "loss": 0.1086,
+      "step": 2962
+    },
+    {
+      "epoch": 1.5424258198854763,
+      "grad_norm": 0.22178597028098188,
+      "learning_rate": 2.5070250190500223e-05,
+      "loss": 0.1079,
+      "step": 2963
+    },
+    {
+      "epoch": 1.54294638209266,
+      "grad_norm": 0.24576503225959828,
+      "learning_rate": 2.5056200179026128e-05,
+      "loss": 0.1124,
+      "step": 2964
+    },
+    {
+      "epoch": 1.5434669442998439,
+      "grad_norm": 0.219334903872518,
+      "learning_rate": 2.5042150149801408e-05,
+      "loss": 0.1045,
+      "step": 2965
+    },
+    {
+      "epoch": 1.5439875065070274,
+      "grad_norm": 0.2253492148046559,
+      "learning_rate": 2.5028100107263714e-05,
+      "loss": 0.1078,
+      "step": 2966
+    },
+    {
+      "epoch": 1.5445080687142112,
+      "grad_norm": 0.21980719949962024,
+      "learning_rate": 2.501405005585069e-05,
+      "loss": 0.1101,
+      "step": 2967
+    },
+    {
+      "epoch": 1.545028630921395,
+      "grad_norm": 0.22443739038898264,
+      "learning_rate": 2.5e-05,
+      "loss": 0.1075,
+      "step": 2968
+    },
+    {
+      "epoch": 1.5455491931285787,
+      "grad_norm": 0.22056246246510713,
+      "learning_rate": 2.4985949944149315e-05,
+      "loss": 0.1076,
+      "step": 2969
+    },
+    {
+      "epoch": 1.5460697553357625,
+      "grad_norm": 0.22545188726465185,
+      "learning_rate": 2.4971899892736295e-05,
+      "loss": 0.1095,
+      "step": 2970
+    },
+    {
+      "epoch": 1.5465903175429463,
+      "grad_norm": 0.2127466187917586,
+      "learning_rate": 2.4957849850198588e-05,
+      "loss": 0.1023,
+      "step": 2971
+    },
+    {
+      "epoch": 1.54711087975013,
+      "grad_norm": 0.24537811434768622,
+      "learning_rate": 2.494379982097387e-05,
+      "loss": 0.112,
+      "step": 2972
+    },
+    {
+      "epoch": 1.5476314419573138,
+      "grad_norm": 0.23557533614284934,
+      "learning_rate": 2.4929749809499786e-05,
+      "loss": 0.1109,
+      "step": 2973
+    },
+    {
+      "epoch": 1.5481520041644976,
+      "grad_norm": 0.23075351228444654,
+      "learning_rate": 2.4915699820213973e-05,
+      "loss": 0.1092,
+      "step": 2974
+    },
+    {
+      "epoch": 1.5486725663716814,
+      "grad_norm": 0.23468413572684516,
+      "learning_rate": 2.4901649857554082e-05,
+      "loss": 0.1106,
+      "step": 2975
+    },
+    {
+      "epoch": 1.5491931285788652,
+      "grad_norm": 0.2282770092156467,
+      "learning_rate": 2.488759992595774e-05,
+      "loss": 0.1055,
+      "step": 2976
+    },
+    {
+      "epoch": 1.549713690786049,
+      "grad_norm": 0.2344915741816126,
+      "learning_rate": 2.487355002986258e-05,
+      "loss": 0.1105,
+      "step": 2977
+    },
+    {
+      "epoch": 1.5502342529932327,
+      "grad_norm": 0.23092806515292114,
+      "learning_rate": 2.4859500173706195e-05,
+      "loss": 0.1101,
+      "step": 2978
+    },
+    {
+      "epoch": 1.5507548152004165,
+      "grad_norm": 0.25062475394919426,
+      "learning_rate": 2.48454503619262e-05,
+      "loss": 0.1087,
+      "step": 2979
+    },
+    {
+      "epoch": 1.5512753774076002,
+      "grad_norm": 0.22139929458785265,
+      "learning_rate": 2.4831400598960162e-05,
+      "loss": 0.1072,
+      "step": 2980
+    },
+    {
+      "epoch": 1.551795939614784,
+      "grad_norm": 0.24617079179057558,
+      "learning_rate": 2.4817350889245675e-05,
+      "loss": 0.1092,
+      "step": 2981
+    },
+    {
+      "epoch": 1.5523165018219678,
+      "grad_norm": 0.23491552403906255,
+      "learning_rate": 2.4803301237220277e-05,
+      "loss": 0.1052,
+      "step": 2982
+    },
+    {
+      "epoch": 1.5528370640291516,
+      "grad_norm": 0.22975191075816628,
+      "learning_rate": 2.4789251647321497e-05,
+      "loss": 0.1085,
+      "step": 2983
+    },
+    {
+      "epoch": 1.5533576262363353,
+      "grad_norm": 0.23209505246845444,
+      "learning_rate": 2.4775202123986855e-05,
+      "loss": 0.1108,
+      "step": 2984
+    },
+    {
+      "epoch": 1.553878188443519,
+      "grad_norm": 0.22191390499400956,
+      "learning_rate": 2.4761152671653835e-05,
+      "loss": 0.1046,
+      "step": 2985
+    },
+    {
+      "epoch": 1.5543987506507029,
+      "grad_norm": 0.22032710409929804,
+      "learning_rate": 2.4747103294759928e-05,
+      "loss": 0.1037,
+      "step": 2986
+    },
+    {
+      "epoch": 1.5549193128578866,
+      "grad_norm": 0.22728163589619113,
+      "learning_rate": 2.4733053997742562e-05,
+      "loss": 0.1102,
+      "step": 2987
+    },
+    {
+      "epoch": 1.5554398750650704,
+      "grad_norm": 0.22954424667321993,
+      "learning_rate": 2.4719004785039168e-05,
+      "loss": 0.1086,
+      "step": 2988
+    },
+    {
+      "epoch": 1.5559604372722542,
+      "grad_norm": 0.22332893329361106,
+      "learning_rate": 2.4704955661087137e-05,
+      "loss": 0.1095,
+      "step": 2989
+    },
+    {
+      "epoch": 1.5564809994794377,
+      "grad_norm": 0.2228519433822219,
+      "learning_rate": 2.4690906630323844e-05,
+      "loss": 0.1113,
+      "step": 2990
+    },
+    {
+      "epoch": 1.5570015616866215,
+      "grad_norm": 0.22383437614333498,
+      "learning_rate": 2.4676857697186625e-05,
+      "loss": 0.1107,
+      "step": 2991
+    },
+    {
+      "epoch": 1.5575221238938053,
+      "grad_norm": 0.2230728740150953,
+      "learning_rate": 2.4662808866112773e-05,
+      "loss": 0.1112,
+      "step": 2992
+    },
+    {
+      "epoch": 1.558042686100989,
+      "grad_norm": 0.21685734122134906,
+      "learning_rate": 2.464876014153958e-05,
+      "loss": 0.1088,
+      "step": 2993
+    },
+    {
+      "epoch": 1.5585632483081728,
+      "grad_norm": 0.2203672898784778,
+      "learning_rate": 2.4634711527904272e-05,
+      "loss": 0.1081,
+      "step": 2994
+    },
+    {
+      "epoch": 1.5590838105153566,
+      "grad_norm": 0.2197549868768869,
+      "learning_rate": 2.4620663029644068e-05,
+      "loss": 0.1042,
+      "step": 2995
+    },
+    {
+      "epoch": 1.5596043727225404,
+      "grad_norm": 0.22043617116963546,
+      "learning_rate": 2.460661465119613e-05,
+      "loss": 0.1069,
+      "step": 2996
+    },
+    {
+      "epoch": 1.5601249349297241,
+      "grad_norm": 0.23842553602217398,
+      "learning_rate": 2.459256639699759e-05,
+      "loss": 0.1137,
+      "step": 2997
+    },
+    {
+      "epoch": 1.5606454971369077,
+      "grad_norm": 0.21377025661406016,
+      "learning_rate": 2.457851827148554e-05,
+      "loss": 0.104,
+      "step": 2998
+    },
+    {
+      "epoch": 1.5611660593440915,
+      "grad_norm": 0.2274138298267601,
+      "learning_rate": 2.456447027909704e-05,
+      "loss": 0.1097,
+      "step": 2999
+    },
+    {
+      "epoch": 1.5616866215512752,
+      "grad_norm": 0.22165119897394892,
+      "learning_rate": 2.45504224242691e-05,
+      "loss": 0.1094,
+      "step": 3000
+    },
+    {
+      "epoch": 1.562207183758459,
+      "grad_norm": 0.21490549783538904,
+      "learning_rate": 2.453637471143867e-05,
+      "loss": 0.1125,
+      "step": 3001
+    },
+    {
+      "epoch": 1.5627277459656428,
+      "grad_norm": 0.21227852422993695,
+      "learning_rate": 2.4522327145042684e-05,
+      "loss": 0.1069,
+      "step": 3002
+    },
+    {
+      "epoch": 1.5632483081728266,
+      "grad_norm": 0.21598368023419817,
+      "learning_rate": 2.450827972951801e-05,
+      "loss": 0.1067,
+      "step": 3003
+    },
+    {
+      "epoch": 1.5637688703800103,
+      "grad_norm": 0.21825295793980276,
+      "learning_rate": 2.4494232469301485e-05,
+      "loss": 0.1071,
+      "step": 3004
+    },
+    {
+      "epoch": 1.564289432587194,
+      "grad_norm": 0.22548740973175538,
+      "learning_rate": 2.4480185368829877e-05,
+      "loss": 0.1129,
+      "step": 3005
+    },
+    {
+      "epoch": 1.5648099947943779,
+      "grad_norm": 0.22232346176457024,
+      "learning_rate": 2.446613843253993e-05,
+      "loss": 0.1067,
+      "step": 3006
+    },
+    {
+      "epoch": 1.5653305570015617,
+      "grad_norm": 0.24009122736060118,
+      "learning_rate": 2.4452091664868298e-05,
+      "loss": 0.111,
+      "step": 3007
+    },
+    {
+      "epoch": 1.5658511192087454,
+      "grad_norm": 0.2243097388008218,
+      "learning_rate": 2.4438045070251624e-05,
+      "loss": 0.1036,
+      "step": 3008
+    },
+    {
+      "epoch": 1.5663716814159292,
+      "grad_norm": 0.23190645268050183,
+      "learning_rate": 2.4423998653126472e-05,
+      "loss": 0.1124,
+      "step": 3009
+    },
+    {
+      "epoch": 1.566892243623113,
+      "grad_norm": 0.3486307228352415,
+      "learning_rate": 2.440995241792934e-05,
+      "loss": 0.1043,
+      "step": 3010
+    },
+    {
+      "epoch": 1.5674128058302967,
+      "grad_norm": 0.23470724899809928,
+      "learning_rate": 2.43959063690967e-05,
+      "loss": 0.1075,
+      "step": 3011
+    },
+    {
+      "epoch": 1.5679333680374805,
+      "grad_norm": 0.23329367665913123,
+      "learning_rate": 2.4381860511064933e-05,
+      "loss": 0.1139,
+      "step": 3012
+    },
+    {
+      "epoch": 1.5684539302446643,
+      "grad_norm": 0.23185338653990387,
+      "learning_rate": 2.4367814848270387e-05,
+      "loss": 0.1082,
+      "step": 3013
+    },
+    {
+      "epoch": 1.568974492451848,
+      "grad_norm": 0.2259910671539203,
+      "learning_rate": 2.435376938514933e-05,
+      "loss": 0.1114,
+      "step": 3014
+    },
+    {
+      "epoch": 1.5694950546590318,
+      "grad_norm": 0.2427562483119206,
+      "learning_rate": 2.4339724126137974e-05,
+      "loss": 0.1119,
+      "step": 3015
+    },
+    {
+      "epoch": 1.5700156168662156,
+      "grad_norm": 0.25899173179638624,
+      "learning_rate": 2.4325679075672462e-05,
+      "loss": 0.116,
+      "step": 3016
+    },
+    {
+      "epoch": 1.5705361790733994,
+      "grad_norm": 0.22878997419672623,
+      "learning_rate": 2.431163423818889e-05,
+      "loss": 0.1085,
+      "step": 3017
+    },
+    {
+      "epoch": 1.5710567412805831,
+      "grad_norm": 0.2384972240792906,
+      "learning_rate": 2.4297589618123258e-05,
+      "loss": 0.1147,
+      "step": 3018
+    },
+    {
+      "epoch": 1.571577303487767,
+      "grad_norm": 0.22510819520011185,
+      "learning_rate": 2.4283545219911503e-05,
+      "loss": 0.1069,
+      "step": 3019
+    },
+    {
+      "epoch": 1.5720978656949507,
+      "grad_norm": 0.2230992006833652,
+      "learning_rate": 2.4269501047989514e-05,
+      "loss": 0.1089,
+      "step": 3020
+    },
+    {
+      "epoch": 1.5726184279021345,
+      "grad_norm": 0.22853011702632323,
+      "learning_rate": 2.425545710679308e-05,
+      "loss": 0.1091,
+      "step": 3021
+    },
+    {
+      "epoch": 1.573138990109318,
+      "grad_norm": 0.2242485692166422,
+      "learning_rate": 2.424141340075795e-05,
+      "loss": 0.1049,
+      "step": 3022
+    },
+    {
+      "epoch": 1.5736595523165018,
+      "grad_norm": 0.2322211744480236,
+      "learning_rate": 2.422736993431976e-05,
+      "loss": 0.1125,
+      "step": 3023
+    },
+    {
+      "epoch": 1.5741801145236856,
+      "grad_norm": 0.2076629175930266,
+      "learning_rate": 2.4213326711914108e-05,
+      "loss": 0.102,
+      "step": 3024
+    },
+    {
+      "epoch": 1.5747006767308693,
+      "grad_norm": 0.2217478061639669,
+      "learning_rate": 2.419928373797648e-05,
+      "loss": 0.1059,
+      "step": 3025
+    },
+    {
+      "epoch": 1.575221238938053,
+      "grad_norm": 0.22413761407508365,
+      "learning_rate": 2.418524101694232e-05,
+      "loss": 0.1088,
+      "step": 3026
+    },
+    {
+      "epoch": 1.5757418011452369,
+      "grad_norm": 0.2261030701823896,
+      "learning_rate": 2.4171198553246967e-05,
+      "loss": 0.113,
+      "step": 3027
+    },
+    {
+      "epoch": 1.5762623633524206,
+      "grad_norm": 0.24473343847561574,
+      "learning_rate": 2.4157156351325668e-05,
+      "loss": 0.1062,
+      "step": 3028
+    },
+    {
+      "epoch": 1.5767829255596044,
+      "grad_norm": 0.22190551778893397,
+      "learning_rate": 2.414311441561362e-05,
+      "loss": 0.1086,
+      "step": 3029
+    },
+    {
+      "epoch": 1.577303487766788,
+      "grad_norm": 0.23472397117142715,
+      "learning_rate": 2.4129072750545912e-05,
+      "loss": 0.1137,
+      "step": 3030
+    },
+    {
+      "epoch": 1.5778240499739717,
+      "grad_norm": 0.22344372474806895,
+      "learning_rate": 2.4115031360557562e-05,
+      "loss": 0.1087,
+      "step": 3031
+    },
+    {
+      "epoch": 1.5783446121811555,
+      "grad_norm": 0.25965872299574905,
+      "learning_rate": 2.4100990250083487e-05,
+      "loss": 0.1139,
+      "step": 3032
+    },
+    {
+      "epoch": 1.5788651743883393,
+      "grad_norm": 0.22568577090311404,
+      "learning_rate": 2.4086949423558526e-05,
+      "loss": 0.1098,
+      "step": 3033
+    },
+    {
+      "epoch": 1.579385736595523,
+      "grad_norm": 0.21814595333115117,
+      "learning_rate": 2.4072908885417438e-05,
+      "loss": 0.1084,
+      "step": 3034
+    },
+    {
+      "epoch": 1.5799062988027068,
+      "grad_norm": 0.22292260467366531,
+      "learning_rate": 2.4058868640094857e-05,
+      "loss": 0.1087,
+      "step": 3035
+    },
+    {
+      "epoch": 1.5804268610098906,
+      "grad_norm": 0.22464387093431798,
+      "learning_rate": 2.404482869202537e-05,
+      "loss": 0.1043,
+      "step": 3036
+    },
+    {
+      "epoch": 1.5809474232170744,
+      "grad_norm": 0.22574493698081854,
+      "learning_rate": 2.4030789045643418e-05,
+      "loss": 0.1046,
+      "step": 3037
+    },
+    {
+      "epoch": 1.5814679854242581,
+      "grad_norm": 0.23794985547646166,
+      "learning_rate": 2.401674970538339e-05,
+      "loss": 0.1092,
+      "step": 3038
+    },
+    {
+      "epoch": 1.581988547631442,
+      "grad_norm": 0.23061947403120528,
+      "learning_rate": 2.4002710675679565e-05,
+      "loss": 0.1083,
+      "step": 3039
+    },
+    {
+      "epoch": 1.5825091098386257,
+      "grad_norm": 0.3114813728434248,
+      "learning_rate": 2.3988671960966113e-05,
+      "loss": 0.1135,
+      "step": 3040
+    },
+    {
+      "epoch": 1.5830296720458095,
+      "grad_norm": 0.24350087448310825,
+      "learning_rate": 2.3974633565677126e-05,
+      "loss": 0.1126,
+      "step": 3041
+    },
+    {
+      "epoch": 1.5835502342529932,
+      "grad_norm": 0.23325201805390236,
+      "learning_rate": 2.3960595494246573e-05,
+      "loss": 0.1088,
+      "step": 3042
+    },
+    {
+      "epoch": 1.584070796460177,
+      "grad_norm": 0.22534337357637105,
+      "learning_rate": 2.394655775110833e-05,
+      "loss": 0.1109,
+      "step": 3043
+    },
+    {
+      "epoch": 1.5845913586673608,
+      "grad_norm": 0.2186403250472862,
+      "learning_rate": 2.393252034069617e-05,
+      "loss": 0.1078,
+      "step": 3044
+    },
+    {
+      "epoch": 1.5851119208745446,
+      "grad_norm": 0.23893857707748875,
+      "learning_rate": 2.3918483267443777e-05,
+      "loss": 0.1096,
+      "step": 3045
+    },
+    {
+      "epoch": 1.5856324830817283,
+      "grad_norm": 0.2296564204097977,
+      "learning_rate": 2.3904446535784687e-05,
+      "loss": 0.1117,
+      "step": 3046
+    },
+    {
+      "epoch": 1.586153045288912,
+      "grad_norm": 0.2256094183111909,
+      "learning_rate": 2.3890410150152363e-05,
+      "loss": 0.11,
+      "step": 3047
+    },
+    {
+      "epoch": 1.5866736074960959,
+      "grad_norm": 0.2197512896748002,
+      "learning_rate": 2.387637411498015e-05,
+      "loss": 0.1085,
+      "step": 3048
+    },
+    {
+      "epoch": 1.5871941697032796,
+      "grad_norm": 0.22684246333884714,
+      "learning_rate": 2.386233843470128e-05,
+      "loss": 0.107,
+      "step": 3049
+    },
+    {
+      "epoch": 1.5877147319104634,
+      "grad_norm": 0.22474155235201673,
+      "learning_rate": 2.384830311374888e-05,
+      "loss": 0.1067,
+      "step": 3050
+    },
+    {
+      "epoch": 1.5882352941176472,
+      "grad_norm": 0.22213948016741225,
+      "learning_rate": 2.3834268156555943e-05,
+      "loss": 0.1066,
+      "step": 3051
+    },
+    {
+      "epoch": 1.588755856324831,
+      "grad_norm": 0.234760326851895,
+      "learning_rate": 2.3820233567555378e-05,
+      "loss": 0.1059,
+      "step": 3052
+    },
+    {
+      "epoch": 1.5892764185320147,
+      "grad_norm": 0.22463049283556413,
+      "learning_rate": 2.3806199351179948e-05,
+      "loss": 0.1119,
+      "step": 3053
+    },
+    {
+      "epoch": 1.5897969807391983,
+      "grad_norm": 0.2233971474985051,
+      "learning_rate": 2.379216551186233e-05,
+      "loss": 0.1039,
+      "step": 3054
+    },
+    {
+      "epoch": 1.590317542946382,
+      "grad_norm": 0.23716623599975512,
+      "learning_rate": 2.3778132054035043e-05,
+      "loss": 0.1077,
+      "step": 3055
+    },
+    {
+      "epoch": 1.5908381051535658,
+      "grad_norm": 0.22256838771775073,
+      "learning_rate": 2.3764098982130507e-05,
+      "loss": 0.1058,
+      "step": 3056
+    },
+    {
+      "epoch": 1.5913586673607496,
+      "grad_norm": 0.22537466848570817,
+      "learning_rate": 2.3750066300581037e-05,
+      "loss": 0.1039,
+      "step": 3057
+    },
+    {
+      "epoch": 1.5918792295679334,
+      "grad_norm": 0.2796215036036369,
+      "learning_rate": 2.3736034013818788e-05,
+      "loss": 0.1144,
+      "step": 3058
+    },
+    {
+      "epoch": 1.5923997917751171,
+      "grad_norm": 0.23128267449784587,
+      "learning_rate": 2.3722002126275824e-05,
+      "loss": 0.1105,
+      "step": 3059
+    },
+    {
+      "epoch": 1.592920353982301,
+      "grad_norm": 0.22108759830456698,
+      "learning_rate": 2.3707970642384056e-05,
+      "loss": 0.1059,
+      "step": 3060
+    },
+    {
+      "epoch": 1.5934409161894847,
+      "grad_norm": 0.23163491105689502,
+      "learning_rate": 2.3693939566575287e-05,
+      "loss": 0.1085,
+      "step": 3061
+    },
+    {
+      "epoch": 1.5939614783966682,
+      "grad_norm": 0.226474109078299,
+      "learning_rate": 2.3679908903281182e-05,
+      "loss": 0.1108,
+      "step": 3062
+    },
+    {
+      "epoch": 1.594482040603852,
+      "grad_norm": 0.2360898550334602,
+      "learning_rate": 2.366587865693328e-05,
+      "loss": 0.1046,
+      "step": 3063
+    },
+    {
+      "epoch": 1.5950026028110358,
+      "grad_norm": 0.21596317159749592,
+      "learning_rate": 2.3651848831962985e-05,
+      "loss": 0.1065,
+      "step": 3064
+    },
+    {
+      "epoch": 1.5955231650182196,
+      "grad_norm": 0.2340159211967676,
+      "learning_rate": 2.363781943280156e-05,
+      "loss": 0.1064,
+      "step": 3065
+    },
+    {
+      "epoch": 1.5960437272254033,
+      "grad_norm": 0.23811590741012165,
+      "learning_rate": 2.3623790463880153e-05,
+      "loss": 0.1101,
+      "step": 3066
+    },
+    {
+      "epoch": 1.596564289432587,
+      "grad_norm": 0.24057517611837248,
+      "learning_rate": 2.3609761929629755e-05,
+      "loss": 0.1038,
+      "step": 3067
+    },
+    {
+      "epoch": 1.5970848516397709,
+      "grad_norm": 0.22888857683722608,
+      "learning_rate": 2.3595733834481236e-05,
+      "loss": 0.1093,
+      "step": 3068
+    },
+    {
+      "epoch": 1.5976054138469546,
+      "grad_norm": 0.23000911098983517,
+      "learning_rate": 2.3581706182865324e-05,
+      "loss": 0.1133,
+      "step": 3069
+    },
+    {
+      "epoch": 1.5981259760541384,
+      "grad_norm": 0.23001119374170786,
+      "learning_rate": 2.3567678979212602e-05,
+      "loss": 0.1064,
+      "step": 3070
+    },
+    {
+      "epoch": 1.5986465382613222,
+      "grad_norm": 0.22412995503191058,
+      "learning_rate": 2.3553652227953507e-05,
+      "loss": 0.1138,
+      "step": 3071
+    },
+    {
+      "epoch": 1.599167100468506,
+      "grad_norm": 0.23039758715555397,
+      "learning_rate": 2.3539625933518354e-05,
+      "loss": 0.1095,
+      "step": 3072
+    },
+    {
+      "epoch": 1.5996876626756897,
+      "grad_norm": 0.2356318998221753,
+      "learning_rate": 2.3525600100337294e-05,
+      "loss": 0.1139,
+      "step": 3073
+    },
+    {
+      "epoch": 1.6002082248828735,
+      "grad_norm": 0.21699411021910314,
+      "learning_rate": 2.3511574732840325e-05,
+      "loss": 0.1024,
+      "step": 3074
+    },
+    {
+      "epoch": 1.6007287870900573,
+      "grad_norm": 0.2185662528539939,
+      "learning_rate": 2.3497549835457328e-05,
+      "loss": 0.1085,
+      "step": 3075
+    },
+    {
+      "epoch": 1.601249349297241,
+      "grad_norm": 0.22922474819533187,
+      "learning_rate": 2.3483525412618e-05,
+      "loss": 0.1097,
+      "step": 3076
+    },
+    {
+      "epoch": 1.6017699115044248,
+      "grad_norm": 0.21577899387549374,
+      "learning_rate": 2.3469501468751928e-05,
+      "loss": 0.1074,
+      "step": 3077
+    },
+    {
+      "epoch": 1.6022904737116086,
+      "grad_norm": 0.22051359137252985,
+      "learning_rate": 2.3455478008288508e-05,
+      "loss": 0.1039,
+      "step": 3078
+    },
+    {
+      "epoch": 1.6028110359187924,
+      "grad_norm": 0.23827856861336888,
+      "learning_rate": 2.3441455035657013e-05,
+      "loss": 0.1068,
+      "step": 3079
+    },
+    {
+      "epoch": 1.6033315981259761,
+      "grad_norm": 0.2338622329393293,
+      "learning_rate": 2.3427432555286543e-05,
+      "loss": 0.1072,
+      "step": 3080
+    },
+    {
+      "epoch": 1.60385216033316,
+      "grad_norm": 0.23740858644971805,
+      "learning_rate": 2.341341057160606e-05,
+      "loss": 0.1114,
+      "step": 3081
+    },
+    {
+      "epoch": 1.6043727225403437,
+      "grad_norm": 0.23132959270597508,
+      "learning_rate": 2.3399389089044354e-05,
+      "loss": 0.1058,
+      "step": 3082
+    },
+    {
+      "epoch": 1.6048932847475275,
+      "grad_norm": 0.253632972193079,
+      "learning_rate": 2.3385368112030052e-05,
+      "loss": 0.1089,
+      "step": 3083
+    },
+    {
+      "epoch": 1.6054138469547112,
+      "grad_norm": 0.2249408980909488,
+      "learning_rate": 2.337134764499164e-05,
+      "loss": 0.1059,
+      "step": 3084
+    },
+    {
+      "epoch": 1.605934409161895,
+      "grad_norm": 0.2220495865666503,
+      "learning_rate": 2.335732769235743e-05,
+      "loss": 0.105,
+      "step": 3085
+    },
+    {
+      "epoch": 1.6064549713690786,
+      "grad_norm": 0.2442265679104099,
+      "learning_rate": 2.3343308258555587e-05,
+      "loss": 0.1086,
+      "step": 3086
+    },
+    {
+      "epoch": 1.6069755335762623,
+      "grad_norm": 0.2336403991069815,
+      "learning_rate": 2.332928934801408e-05,
+      "loss": 0.1075,
+      "step": 3087
+    },
+    {
+      "epoch": 1.607496095783446,
+      "grad_norm": 0.22276349197581477,
+      "learning_rate": 2.3315270965160753e-05,
+      "loss": 0.1055,
+      "step": 3088
+    },
+    {
+      "epoch": 1.6080166579906299,
+      "grad_norm": 0.2280395730745695,
+      "learning_rate": 2.330125311442325e-05,
+      "loss": 0.1044,
+      "step": 3089
+    },
+    {
+      "epoch": 1.6085372201978136,
+      "grad_norm": 0.23389298695008165,
+      "learning_rate": 2.3287235800229073e-05,
+      "loss": 0.1047,
+      "step": 3090
+    },
+    {
+      "epoch": 1.6090577824049974,
+      "grad_norm": 0.22779890717062137,
+      "learning_rate": 2.327321902700554e-05,
+      "loss": 0.1092,
+      "step": 3091
+    },
+    {
+      "epoch": 1.6095783446121812,
+      "grad_norm": 0.21949265647763663,
+      "learning_rate": 2.3259202799179785e-05,
+      "loss": 0.1036,
+      "step": 3092
+    },
+    {
+      "epoch": 1.610098906819365,
+      "grad_norm": 0.22277867696874984,
+      "learning_rate": 2.3245187121178804e-05,
+      "loss": 0.1106,
+      "step": 3093
+    },
+    {
+      "epoch": 1.6106194690265485,
+      "grad_norm": 0.23819505127712554,
+      "learning_rate": 2.323117199742938e-05,
+      "loss": 0.1102,
+      "step": 3094
+    },
+    {
+      "epoch": 1.6111400312337323,
+      "grad_norm": 0.22647009120490577,
+      "learning_rate": 2.3217157432358164e-05,
+      "loss": 0.1048,
+      "step": 3095
+    },
+    {
+      "epoch": 1.611660593440916,
+      "grad_norm": 0.23823079455303328,
+      "learning_rate": 2.320314343039159e-05,
+      "loss": 0.1067,
+      "step": 3096
+    },
+    {
+      "epoch": 1.6121811556480998,
+      "grad_norm": 0.22520992501315207,
+      "learning_rate": 2.3189129995955943e-05,
+      "loss": 0.1053,
+      "step": 3097
+    },
+    {
+      "epoch": 1.6127017178552836,
+      "grad_norm": 0.2258137397120401,
+      "learning_rate": 2.3175117133477313e-05,
+      "loss": 0.1053,
+      "step": 3098
+    },
+    {
+      "epoch": 1.6132222800624674,
+      "grad_norm": 0.2271783410129132,
+      "learning_rate": 2.3161104847381608e-05,
+      "loss": 0.1102,
+      "step": 3099
+    },
+    {
+      "epoch": 1.6137428422696511,
+      "grad_norm": 0.2271396794206383,
+      "learning_rate": 2.314709314209459e-05,
+      "loss": 0.1025,
+      "step": 3100
+    },
+    {
+      "epoch": 1.614263404476835,
+      "grad_norm": 0.2512629230037695,
+      "learning_rate": 2.3133082022041767e-05,
+      "loss": 0.1088,
+      "step": 3101
+    },
+    {
+      "epoch": 1.6147839666840187,
+      "grad_norm": 0.23119313550176498,
+      "learning_rate": 2.3119071491648525e-05,
+      "loss": 0.108,
+      "step": 3102
+    },
+    {
+      "epoch": 1.6153045288912025,
+      "grad_norm": 0.21317276454406853,
+      "learning_rate": 2.310506155534003e-05,
+      "loss": 0.1034,
+      "step": 3103
+    },
+    {
+      "epoch": 1.6158250910983862,
+      "grad_norm": 0.24282928727082653,
+      "learning_rate": 2.3091052217541278e-05,
+      "loss": 0.11,
+      "step": 3104
+    },
+    {
+      "epoch": 1.61634565330557,
+      "grad_norm": 0.2408607507520557,
+      "learning_rate": 2.307704348267708e-05,
+      "loss": 0.1087,
+      "step": 3105
+    },
+    {
+      "epoch": 1.6168662155127538,
+      "grad_norm": 0.23585883743968147,
+      "learning_rate": 2.3063035355172026e-05,
+      "loss": 0.1117,
+      "step": 3106
+    },
+    {
+      "epoch": 1.6173867777199376,
+      "grad_norm": 0.23123042557678272,
+      "learning_rate": 2.304902783945056e-05,
+      "loss": 0.1056,
+      "step": 3107
+    },
+    {
+      "epoch": 1.6179073399271213,
+      "grad_norm": 0.23648141825819088,
+      "learning_rate": 2.3035020939936876e-05,
+      "loss": 0.1037,
+      "step": 3108
+    },
+    {
+      "epoch": 1.618427902134305,
+      "grad_norm": 0.22436906961143638,
+      "learning_rate": 2.302101466105504e-05,
+      "loss": 0.1017,
+      "step": 3109
+    },
+    {
+      "epoch": 1.6189484643414889,
+      "grad_norm": 0.21301990592667625,
+      "learning_rate": 2.3007009007228857e-05,
+      "loss": 0.1059,
+      "step": 3110
+    },
+    {
+      "epoch": 1.6194690265486726,
+      "grad_norm": 0.23396887540615707,
+      "learning_rate": 2.2993003982881975e-05,
+      "loss": 0.1062,
+      "step": 3111
+    },
+    {
+      "epoch": 1.6199895887558564,
+      "grad_norm": 0.2433106836264054,
+      "learning_rate": 2.2978999592437837e-05,
+      "loss": 0.1058,
+      "step": 3112
+    },
+    {
+      "epoch": 1.6205101509630402,
+      "grad_norm": 0.2216670932634246,
+      "learning_rate": 2.296499584031967e-05,
+      "loss": 0.1049,
+      "step": 3113
+    },
+    {
+      "epoch": 1.621030713170224,
+      "grad_norm": 0.23403580449109418,
+      "learning_rate": 2.2950992730950523e-05,
+      "loss": 0.1047,
+      "step": 3114
+    },
+    {
+      "epoch": 1.6215512753774077,
+      "grad_norm": 0.23966172688921225,
+      "learning_rate": 2.293699026875322e-05,
+      "loss": 0.1048,
+      "step": 3115
+    },
+    {
+      "epoch": 1.6220718375845915,
+      "grad_norm": 0.22806661382306423,
+      "learning_rate": 2.29229884581504e-05,
+      "loss": 0.1082,
+      "step": 3116
+    },
+    {
+      "epoch": 1.6225923997917753,
+      "grad_norm": 0.23681871295661272,
+      "learning_rate": 2.2908987303564474e-05,
+      "loss": 0.1044,
+      "step": 3117
+    },
+    {
+      "epoch": 1.6231129619989588,
+      "grad_norm": 0.22656571389726227,
+      "learning_rate": 2.2894986809417676e-05,
+      "loss": 0.1077,
+      "step": 3118
+    },
+    {
+      "epoch": 1.6236335242061426,
+      "grad_norm": 0.23062976184034878,
+      "learning_rate": 2.2880986980131996e-05,
+      "loss": 0.1031,
+      "step": 3119
+    },
+    {
+      "epoch": 1.6241540864133264,
+      "grad_norm": 0.22553061329604537,
+      "learning_rate": 2.2866987820129234e-05,
+      "loss": 0.1053,
+      "step": 3120
+    },
+    {
+      "epoch": 1.6246746486205101,
+      "grad_norm": 0.22883909177019793,
+      "learning_rate": 2.2852989333830988e-05,
+      "loss": 0.1059,
+      "step": 3121
+    },
+    {
+      "epoch": 1.625195210827694,
+      "grad_norm": 0.2267238669608568,
+      "learning_rate": 2.2838991525658616e-05,
+      "loss": 0.1092,
+      "step": 3122
+    },
+    {
+      "epoch": 1.6257157730348777,
+      "grad_norm": 0.2221538641240812,
+      "learning_rate": 2.282499440003329e-05,
+      "loss": 0.1019,
+      "step": 3123
+    },
+    {
+      "epoch": 1.6262363352420615,
+      "grad_norm": 0.23065218500300072,
+      "learning_rate": 2.281099796137594e-05,
+      "loss": 0.1073,
+      "step": 3124
+    },
+    {
+      "epoch": 1.6267568974492452,
+      "grad_norm": 0.21528998054447926,
+      "learning_rate": 2.279700221410731e-05,
+      "loss": 0.1038,
+      "step": 3125
+    },
+    {
+      "epoch": 1.6272774596564288,
+      "grad_norm": 0.22406127559380892,
+      "learning_rate": 2.2783007162647894e-05,
+      "loss": 0.1092,
+      "step": 3126
+    },
+    {
+      "epoch": 1.6277980218636126,
+      "grad_norm": 0.22732856950130653,
+      "learning_rate": 2.2769012811417998e-05,
+      "loss": 0.1043,
+      "step": 3127
+    },
+    {
+      "epoch": 1.6283185840707963,
+      "grad_norm": 0.24391940916747334,
+      "learning_rate": 2.275501916483767e-05,
+      "loss": 0.1084,
+      "step": 3128
+    },
+    {
+      "epoch": 1.62883914627798,
+      "grad_norm": 0.212950903066082,
+      "learning_rate": 2.274102622732676e-05,
+      "loss": 0.1021,
+      "step": 3129
+    },
+    {
+      "epoch": 1.6293597084851639,
+      "grad_norm": 0.22582328484751868,
+      "learning_rate": 2.2727034003304893e-05,
+      "loss": 0.1078,
+      "step": 3130
+    },
+    {
+      "epoch": 1.6298802706923476,
+      "grad_norm": 0.23178125241301198,
+      "learning_rate": 2.2713042497191456e-05,
+      "loss": 0.1052,
+      "step": 3131
+    },
+    {
+      "epoch": 1.6304008328995314,
+      "grad_norm": 0.22817672478922343,
+      "learning_rate": 2.2699051713405633e-05,
+      "loss": 0.113,
+      "step": 3132
+    },
+    {
+      "epoch": 1.6309213951067152,
+      "grad_norm": 0.21735110824642498,
+      "learning_rate": 2.2685061656366347e-05,
+      "loss": 0.1048,
+      "step": 3133
+    },
+    {
+      "epoch": 1.631441957313899,
+      "grad_norm": 0.219691390085283,
+      "learning_rate": 2.2671072330492328e-05,
+      "loss": 0.1063,
+      "step": 3134
+    },
+    {
+      "epoch": 1.6319625195210827,
+      "grad_norm": 0.23114194424793824,
+      "learning_rate": 2.2657083740202035e-05,
+      "loss": 0.1096,
+      "step": 3135
+    },
+    {
+      "epoch": 1.6324830817282665,
+      "grad_norm": 0.2194312392722462,
+      "learning_rate": 2.264309588991374e-05,
+      "loss": 0.1066,
+      "step": 3136
+    },
+    {
+      "epoch": 1.6330036439354503,
+      "grad_norm": 0.21507036864686763,
+      "learning_rate": 2.2629108784045438e-05,
+      "loss": 0.1077,
+      "step": 3137
+    },
+    {
+      "epoch": 1.633524206142634,
+      "grad_norm": 0.22777523976635303,
+      "learning_rate": 2.2615122427014913e-05,
+      "loss": 0.1086,
+      "step": 3138
+    },
+    {
+      "epoch": 1.6340447683498178,
+      "grad_norm": 0.22055091654855918,
+      "learning_rate": 2.260113682323971e-05,
+      "loss": 0.1035,
+      "step": 3139
+    },
+    {
+      "epoch": 1.6345653305570016,
+      "grad_norm": 0.2238767964565613,
+      "learning_rate": 2.2587151977137122e-05,
+      "loss": 0.1077,
+      "step": 3140
+    },
+    {
+      "epoch": 1.6350858927641854,
+      "grad_norm": 0.22760165259373993,
+      "learning_rate": 2.2573167893124237e-05,
+      "loss": 0.1128,
+      "step": 3141
+    },
+    {
+      "epoch": 1.6356064549713691,
+      "grad_norm": 0.22042911403521775,
+      "learning_rate": 2.2559184575617857e-05,
+      "loss": 0.1073,
+      "step": 3142
+    },
+    {
+      "epoch": 1.636127017178553,
+      "grad_norm": 0.2278962819089735,
+      "learning_rate": 2.254520202903458e-05,
+      "loss": 0.107,
+      "step": 3143
+    },
+    {
+      "epoch": 1.6366475793857367,
+      "grad_norm": 0.22694087159169066,
+      "learning_rate": 2.253122025779073e-05,
+      "loss": 0.1092,
+      "step": 3144
+    },
+    {
+      "epoch": 1.6371681415929205,
+      "grad_norm": 0.22199970972024982,
+      "learning_rate": 2.2517239266302424e-05,
+      "loss": 0.1088,
+      "step": 3145
+    },
+    {
+      "epoch": 1.6376887038001042,
+      "grad_norm": 0.22587462774553319,
+      "learning_rate": 2.2503259058985487e-05,
+      "loss": 0.1039,
+      "step": 3146
+    },
+    {
+      "epoch": 1.638209266007288,
+      "grad_norm": 0.22588919883230146,
+      "learning_rate": 2.2489279640255526e-05,
+      "loss": 0.1018,
+      "step": 3147
+    },
+    {
+      "epoch": 1.6387298282144718,
+      "grad_norm": 0.2361781102822983,
+      "learning_rate": 2.2475301014527897e-05,
+      "loss": 0.1078,
+      "step": 3148
+    },
+    {
+      "epoch": 1.6392503904216555,
+      "grad_norm": 0.22178776083606402,
+      "learning_rate": 2.246132318621769e-05,
+      "loss": 0.1065,
+      "step": 3149
+    },
+    {
+      "epoch": 1.639770952628839,
+      "grad_norm": 0.2301808192860147,
+      "learning_rate": 2.2447346159739772e-05,
+      "loss": 0.1069,
+      "step": 3150
+    },
+    {
+      "epoch": 1.6402915148360229,
+      "grad_norm": 0.23564349495869355,
+      "learning_rate": 2.2433369939508718e-05,
+      "loss": 0.1132,
+      "step": 3151
+    },
+    {
+      "epoch": 1.6408120770432066,
+      "grad_norm": 0.2142534971466308,
+      "learning_rate": 2.241939452993888e-05,
+      "loss": 0.1035,
+      "step": 3152
+    },
+    {
+      "epoch": 1.6413326392503904,
+      "grad_norm": 0.2269015837271862,
+      "learning_rate": 2.2405419935444338e-05,
+      "loss": 0.1146,
+      "step": 3153
+    },
+    {
+      "epoch": 1.6418532014575742,
+      "grad_norm": 0.23097131531972812,
+      "learning_rate": 2.2391446160438933e-05,
+      "loss": 0.1092,
+      "step": 3154
+    },
+    {
+      "epoch": 1.642373763664758,
+      "grad_norm": 0.22445851872935704,
+      "learning_rate": 2.2377473209336214e-05,
+      "loss": 0.1059,
+      "step": 3155
+    },
+    {
+      "epoch": 1.6428943258719417,
+      "grad_norm": 0.21849931770987602,
+      "learning_rate": 2.2363501086549498e-05,
+      "loss": 0.1067,
+      "step": 3156
+    },
+    {
+      "epoch": 1.6434148880791255,
+      "grad_norm": 0.22288897541665642,
+      "learning_rate": 2.234952979649183e-05,
+      "loss": 0.1072,
+      "step": 3157
+    },
+    {
+      "epoch": 1.643935450286309,
+      "grad_norm": 0.23784579556458776,
+      "learning_rate": 2.233555934357599e-05,
+      "loss": 0.1146,
+      "step": 3158
+    },
+    {
+      "epoch": 1.6444560124934928,
+      "grad_norm": 0.22729963389585045,
+      "learning_rate": 2.232158973221451e-05,
+      "loss": 0.1048,
+      "step": 3159
+    },
+    {
+      "epoch": 1.6449765747006766,
+      "grad_norm": 0.22444048797078522,
+      "learning_rate": 2.230762096681963e-05,
+      "loss": 0.1032,
+      "step": 3160
+    },
+    {
+      "epoch": 1.6454971369078604,
+      "grad_norm": 0.2274706569746683,
+      "learning_rate": 2.2293653051803344e-05,
+      "loss": 0.1125,
+      "step": 3161
+    },
+    {
+      "epoch": 1.6460176991150441,
+      "grad_norm": 0.2215985773170227,
+      "learning_rate": 2.2279685991577363e-05,
+      "loss": 0.1106,
+      "step": 3162
+    },
+    {
+      "epoch": 1.646538261322228,
+      "grad_norm": 0.2248236336469974,
+      "learning_rate": 2.2265719790553147e-05,
+      "loss": 0.1056,
+      "step": 3163
+    },
+    {
+      "epoch": 1.6470588235294117,
+      "grad_norm": 0.21576271667103167,
+      "learning_rate": 2.225175445314186e-05,
+      "loss": 0.1046,
+      "step": 3164
+    },
+    {
+      "epoch": 1.6475793857365955,
+      "grad_norm": 0.23144058603703913,
+      "learning_rate": 2.2237789983754405e-05,
+      "loss": 0.1113,
+      "step": 3165
+    },
+    {
+      "epoch": 1.6480999479437792,
+      "grad_norm": 0.2143838630761341,
+      "learning_rate": 2.2223826386801417e-05,
+      "loss": 0.1041,
+      "step": 3166
+    },
+    {
+      "epoch": 1.648620510150963,
+      "grad_norm": 0.2199889181336657,
+      "learning_rate": 2.2209863666693244e-05,
+      "loss": 0.1041,
+      "step": 3167
+    },
+    {
+      "epoch": 1.6491410723581468,
+      "grad_norm": 0.23070227108778804,
+      "learning_rate": 2.2195901827839965e-05,
+      "loss": 0.1092,
+      "step": 3168
+    },
+    {
+      "epoch": 1.6496616345653305,
+      "grad_norm": 0.21963888599506054,
+      "learning_rate": 2.2181940874651393e-05,
+      "loss": 0.1067,
+      "step": 3169
+    },
+    {
+      "epoch": 1.6501821967725143,
+      "grad_norm": 0.22331117871742942,
+      "learning_rate": 2.216798081153702e-05,
+      "loss": 0.107,
+      "step": 3170
+    },
+    {
+      "epoch": 1.650702758979698,
+      "grad_norm": 0.23747544142009083,
+      "learning_rate": 2.2154021642906107e-05,
+      "loss": 0.1125,
+      "step": 3171
+    },
+    {
+      "epoch": 1.6512233211868819,
+      "grad_norm": 0.2304050364200822,
+      "learning_rate": 2.2140063373167606e-05,
+      "loss": 0.1068,
+      "step": 3172
+    },
+    {
+      "epoch": 1.6517438833940656,
+      "grad_norm": 0.23358412449420016,
+      "learning_rate": 2.212610600673017e-05,
+      "loss": 0.1059,
+      "step": 3173
+    },
+    {
+      "epoch": 1.6522644456012494,
+      "grad_norm": 0.21636641417173472,
+      "learning_rate": 2.21121495480022e-05,
+      "loss": 0.1055,
+      "step": 3174
+    },
+    {
+      "epoch": 1.6527850078084332,
+      "grad_norm": 0.22691961656100734,
+      "learning_rate": 2.2098194001391785e-05,
+      "loss": 0.1039,
+      "step": 3175
+    },
+    {
+      "epoch": 1.653305570015617,
+      "grad_norm": 0.23081013769275266,
+      "learning_rate": 2.2084239371306753e-05,
+      "loss": 0.1108,
+      "step": 3176
+    },
+    {
+      "epoch": 1.6538261322228007,
+      "grad_norm": 0.24275473044090903,
+      "learning_rate": 2.2070285662154607e-05,
+      "loss": 0.1108,
+      "step": 3177
+    },
+    {
+      "epoch": 1.6543466944299845,
+      "grad_norm": 0.2161343954274846,
+      "learning_rate": 2.2056332878342594e-05,
+      "loss": 0.106,
+      "step": 3178
+    },
+    {
+      "epoch": 1.6548672566371683,
+      "grad_norm": 0.22720407589672256,
+      "learning_rate": 2.2042381024277637e-05,
+      "loss": 0.1044,
+      "step": 3179
+    },
+    {
+      "epoch": 1.655387818844352,
+      "grad_norm": 0.23757875519411803,
+      "learning_rate": 2.202843010436639e-05,
+      "loss": 0.1122,
+      "step": 3180
+    },
+    {
+      "epoch": 1.6559083810515358,
+      "grad_norm": 0.209322137149078,
+      "learning_rate": 2.2014480123015214e-05,
+      "loss": 0.1052,
+      "step": 3181
+    },
+    {
+      "epoch": 1.6564289432587194,
+      "grad_norm": 0.2238782072859011,
+      "learning_rate": 2.2000531084630137e-05,
+      "loss": 0.1092,
+      "step": 3182
+    },
+    {
+      "epoch": 1.6569495054659031,
+      "grad_norm": 0.21411441676663537,
+      "learning_rate": 2.1986582993616926e-05,
+      "loss": 0.0983,
+      "step": 3183
+    },
+    {
+      "epoch": 1.657470067673087,
+      "grad_norm": 0.23607167726054,
+      "learning_rate": 2.1972635854381042e-05,
+      "loss": 0.1062,
+      "step": 3184
+    },
+    {
+      "epoch": 1.6579906298802707,
+      "grad_norm": 0.23634245749199465,
+      "learning_rate": 2.1958689671327635e-05,
+      "loss": 0.1077,
+      "step": 3185
+    },
+    {
+      "epoch": 1.6585111920874545,
+      "grad_norm": 0.26476311947101455,
+      "learning_rate": 2.1944744448861557e-05,
+      "loss": 0.1029,
+      "step": 3186
+    },
+    {
+      "epoch": 1.6590317542946382,
+      "grad_norm": 0.2225889940109205,
+      "learning_rate": 2.1930800191387366e-05,
+      "loss": 0.1019,
+      "step": 3187
+    },
+    {
+      "epoch": 1.659552316501822,
+      "grad_norm": 0.2123029031111446,
+      "learning_rate": 2.1916856903309298e-05,
+      "loss": 0.1013,
+      "step": 3188
+    },
+    {
+      "epoch": 1.6600728787090058,
+      "grad_norm": 0.23091095057929104,
+      "learning_rate": 2.19029145890313e-05,
+      "loss": 0.1065,
+      "step": 3189
+    },
+    {
+      "epoch": 1.6605934409161893,
+      "grad_norm": 0.24015311654835608,
+      "learning_rate": 2.188897325295701e-05,
+      "loss": 0.1039,
+      "step": 3190
+    },
+    {
+      "epoch": 1.661114003123373,
+      "grad_norm": 0.23370275976758934,
+      "learning_rate": 2.187503289948973e-05,
+      "loss": 0.1104,
+      "step": 3191
+    },
+    {
+      "epoch": 1.6616345653305569,
+      "grad_norm": 0.22827631936172066,
+      "learning_rate": 2.1861093533032488e-05,
+      "loss": 0.1033,
+      "step": 3192
+    },
+    {
+      "epoch": 1.6621551275377406,
+      "grad_norm": 0.2208651631543092,
+      "learning_rate": 2.1847155157987972e-05,
+      "loss": 0.1046,
+      "step": 3193
+    },
+    {
+      "epoch": 1.6626756897449244,
+      "grad_norm": 0.22601221991866668,
+      "learning_rate": 2.1833217778758584e-05,
+      "loss": 0.1077,
+      "step": 3194
+    },
+    {
+      "epoch": 1.6631962519521082,
+      "grad_norm": 0.22968672480262192,
+      "learning_rate": 2.1819281399746392e-05,
+      "loss": 0.104,
+      "step": 3195
+    },
+    {
+      "epoch": 1.663716814159292,
+      "grad_norm": 0.23140457155651345,
+      "learning_rate": 2.180534602535315e-05,
+      "loss": 0.1068,
+      "step": 3196
+    },
+    {
+      "epoch": 1.6642373763664757,
+      "grad_norm": 0.2153723166687647,
+      "learning_rate": 2.17914116599803e-05,
+      "loss": 0.1063,
+      "step": 3197
+    },
+    {
+      "epoch": 1.6647579385736595,
+      "grad_norm": 0.23114435518218374,
+      "learning_rate": 2.1777478308028965e-05,
+      "loss": 0.1089,
+      "step": 3198
+    },
+    {
+      "epoch": 1.6652785007808433,
+      "grad_norm": 0.22821342024755512,
+      "learning_rate": 2.176354597389995e-05,
+      "loss": 0.1054,
+      "step": 3199
+    },
+    {
+      "epoch": 1.665799062988027,
+      "grad_norm": 0.2339992932085678,
+      "learning_rate": 2.1749614661993715e-05,
+      "loss": 0.1113,
+      "step": 3200
+    },
+    {
+      "epoch": 1.6663196251952108,
+      "grad_norm": 0.22259182696219096,
+      "learning_rate": 2.1735684376710435e-05,
+      "loss": 0.1044,
+      "step": 3201
+    },
+    {
+      "epoch": 1.6668401874023946,
+      "grad_norm": 0.2219044472166739,
+      "learning_rate": 2.1721755122449932e-05,
+      "loss": 0.1077,
+      "step": 3202
+    },
+    {
+      "epoch": 1.6673607496095784,
+      "grad_norm": 0.21727491507702684,
+      "learning_rate": 2.1707826903611726e-05,
+      "loss": 0.1028,
+      "step": 3203
+    },
+    {
+      "epoch": 1.6678813118167621,
+      "grad_norm": 0.2299535668577538,
+      "learning_rate": 2.169389972459498e-05,
+      "loss": 0.1098,
+      "step": 3204
+    },
+    {
+      "epoch": 1.668401874023946,
+      "grad_norm": 0.21563012365258863,
+      "learning_rate": 2.1679973589798564e-05,
+      "loss": 0.103,
+      "step": 3205
+    },
+    {
+      "epoch": 1.6689224362311297,
+      "grad_norm": 0.22726073168216962,
+      "learning_rate": 2.166604850362098e-05,
+      "loss": 0.1082,
+      "step": 3206
+    },
+    {
+      "epoch": 1.6694429984383135,
+      "grad_norm": 0.23344147957321737,
+      "learning_rate": 2.165212447046043e-05,
+      "loss": 0.11,
+      "step": 3207
+    },
+    {
+      "epoch": 1.6699635606454972,
+      "grad_norm": 0.21101470102115977,
+      "learning_rate": 2.163820149471478e-05,
+      "loss": 0.1067,
+      "step": 3208
+    },
+    {
+      "epoch": 1.670484122852681,
+      "grad_norm": 0.21852389821603915,
+      "learning_rate": 2.1624279580781525e-05,
+      "loss": 0.1094,
+      "step": 3209
+    },
+    {
+      "epoch": 1.6710046850598648,
+      "grad_norm": 0.21706627954133206,
+      "learning_rate": 2.1610358733057882e-05,
+      "loss": 0.1092,
+      "step": 3210
+    },
+    {
+      "epoch": 1.6715252472670485,
+      "grad_norm": 0.2260538276253177,
+      "learning_rate": 2.1596438955940682e-05,
+      "loss": 0.1079,
+      "step": 3211
+    },
+    {
+      "epoch": 1.6720458094742323,
+      "grad_norm": 0.23132753540493692,
+      "learning_rate": 2.1582520253826454e-05,
+      "loss": 0.1061,
+      "step": 3212
+    },
+    {
+      "epoch": 1.672566371681416,
+      "grad_norm": 0.22764404827151757,
+      "learning_rate": 2.156860263111136e-05,
+      "loss": 0.1087,
+      "step": 3213
+    },
+    {
+      "epoch": 1.6730869338885996,
+      "grad_norm": 0.23998945954385206,
+      "learning_rate": 2.155468609219125e-05,
+      "loss": 0.1068,
+      "step": 3214
+    },
+    {
+      "epoch": 1.6736074960957834,
+      "grad_norm": 0.22270579464295384,
+      "learning_rate": 2.154077064146159e-05,
+      "loss": 0.1031,
+      "step": 3215
+    },
+    {
+      "epoch": 1.6741280583029672,
+      "grad_norm": 0.23422116557049794,
+      "learning_rate": 2.152685628331755e-05,
+      "loss": 0.1083,
+      "step": 3216
+    },
+    {
+      "epoch": 1.674648620510151,
+      "grad_norm": 0.23808475492769957,
+      "learning_rate": 2.1512943022153926e-05,
+      "loss": 0.1108,
+      "step": 3217
+    },
+    {
+      "epoch": 1.6751691827173347,
+      "grad_norm": 0.2336184489419489,
+      "learning_rate": 2.149903086236516e-05,
+      "loss": 0.1072,
+      "step": 3218
+    },
+    {
+      "epoch": 1.6756897449245185,
+      "grad_norm": 0.21905361177253538,
+      "learning_rate": 2.1485119808345372e-05,
+      "loss": 0.1037,
+      "step": 3219
+    },
+    {
+      "epoch": 1.6762103071317023,
+      "grad_norm": 0.20981047900721037,
+      "learning_rate": 2.147120986448831e-05,
+      "loss": 0.0976,
+      "step": 3220
+    },
+    {
+      "epoch": 1.676730869338886,
+      "grad_norm": 0.2545642241580596,
+      "learning_rate": 2.1457301035187397e-05,
+      "loss": 0.1093,
+      "step": 3221
+    },
+    {
+      "epoch": 1.6772514315460696,
+      "grad_norm": 0.2218754138460833,
+      "learning_rate": 2.1443393324835675e-05,
+      "loss": 0.1047,
+      "step": 3222
+    },
+    {
+      "epoch": 1.6777719937532534,
+      "grad_norm": 0.23088102794993018,
+      "learning_rate": 2.1429486737825854e-05,
+      "loss": 0.1104,
+      "step": 3223
+    },
+    {
+      "epoch": 1.6782925559604371,
+      "grad_norm": 0.22005937755272137,
+      "learning_rate": 2.1415581278550275e-05,
+      "loss": 0.1073,
+      "step": 3224
+    },
+    {
+      "epoch": 1.678813118167621,
+      "grad_norm": 0.23400677719772744,
+      "learning_rate": 2.140167695140094e-05,
+      "loss": 0.1103,
+      "step": 3225
+    },
+    {
+      "epoch": 1.6793336803748047,
+      "grad_norm": 0.2365964192307508,
+      "learning_rate": 2.1387773760769474e-05,
+      "loss": 0.1085,
+      "step": 3226
+    },
+    {
+      "epoch": 1.6798542425819885,
+      "grad_norm": 0.21637550071988382,
+      "learning_rate": 2.137387171104715e-05,
+      "loss": 0.107,
+      "step": 3227
+    },
+    {
+      "epoch": 1.6803748047891722,
+      "grad_norm": 0.22333862046046274,
+      "learning_rate": 2.1359970806624885e-05,
+      "loss": 0.1084,
+      "step": 3228
+    },
+    {
+      "epoch": 1.680895366996356,
+      "grad_norm": 0.23203824461964603,
+      "learning_rate": 2.134607105189323e-05,
+      "loss": 0.1074,
+      "step": 3229
+    },
+    {
+      "epoch": 1.6814159292035398,
+      "grad_norm": 0.22765546349056484,
+      "learning_rate": 2.1332172451242378e-05,
+      "loss": 0.1069,
+      "step": 3230
+    },
+    {
+      "epoch": 1.6819364914107235,
+      "grad_norm": 0.23666560244713625,
+      "learning_rate": 2.131827500906215e-05,
+      "loss": 0.1034,
+      "step": 3231
+    },
+    {
+      "epoch": 1.6824570536179073,
+      "grad_norm": 0.2374292849318095,
+      "learning_rate": 2.1304378729742007e-05,
+      "loss": 0.1068,
+      "step": 3232
+    },
+    {
+      "epoch": 1.682977615825091,
+      "grad_norm": 0.2324037821047654,
+      "learning_rate": 2.129048361767104e-05,
+      "loss": 0.1038,
+      "step": 3233
+    },
+    {
+      "epoch": 1.6834981780322749,
+      "grad_norm": 0.23887705711017124,
+      "learning_rate": 2.127658967723797e-05,
+      "loss": 0.1116,
+      "step": 3234
+    },
+    {
+      "epoch": 1.6840187402394586,
+      "grad_norm": 0.24478758959999747,
+      "learning_rate": 2.126269691283117e-05,
+      "loss": 0.1078,
+      "step": 3235
+    },
+    {
+      "epoch": 1.6845393024466424,
+      "grad_norm": 0.24871328706272613,
+      "learning_rate": 2.124880532883859e-05,
+      "loss": 0.1081,
+      "step": 3236
+    },
+    {
+      "epoch": 1.6850598646538262,
+      "grad_norm": 0.2516650860777226,
+      "learning_rate": 2.123491492964785e-05,
+      "loss": 0.1085,
+      "step": 3237
+    },
+    {
+      "epoch": 1.68558042686101,
+      "grad_norm": 0.2253323571868253,
+      "learning_rate": 2.1221025719646193e-05,
+      "loss": 0.1019,
+      "step": 3238
+    },
+    {
+      "epoch": 1.6861009890681937,
+      "grad_norm": 0.22356011592203995,
+      "learning_rate": 2.1207137703220465e-05,
+      "loss": 0.1069,
+      "step": 3239
+    },
+    {
+      "epoch": 1.6866215512753775,
+      "grad_norm": 0.22129126827686282,
+      "learning_rate": 2.119325088475716e-05,
+      "loss": 0.1021,
+      "step": 3240
+    },
+    {
+      "epoch": 1.6871421134825613,
+      "grad_norm": 0.24363724324274239,
+      "learning_rate": 2.1179365268642374e-05,
+      "loss": 0.1073,
+      "step": 3241
+    },
+    {
+      "epoch": 1.687662675689745,
+      "grad_norm": 0.22571502175977035,
+      "learning_rate": 2.1165480859261838e-05,
+      "loss": 0.1086,
+      "step": 3242
+    },
+    {
+      "epoch": 1.6881832378969288,
+      "grad_norm": 0.2333938281028589,
+      "learning_rate": 2.1151597661000884e-05,
+      "loss": 0.1116,
+      "step": 3243
+    },
+    {
+      "epoch": 1.6887038001041126,
+      "grad_norm": 0.2418535802191957,
+      "learning_rate": 2.1137715678244492e-05,
+      "loss": 0.1122,
+      "step": 3244
+    },
+    {
+      "epoch": 1.6892243623112964,
+      "grad_norm": 0.2226092711205871,
+      "learning_rate": 2.1123834915377212e-05,
+      "loss": 0.1077,
+      "step": 3245
+    },
+    {
+      "epoch": 1.68974492451848,
+      "grad_norm": 0.2218422154239549,
+      "learning_rate": 2.1109955376783247e-05,
+      "loss": 0.1063,
+      "step": 3246
+    },
+    {
+      "epoch": 1.6902654867256637,
+      "grad_norm": 0.22432155018614183,
+      "learning_rate": 2.1096077066846404e-05,
+      "loss": 0.1042,
+      "step": 3247
+    },
+    {
+      "epoch": 1.6907860489328475,
+      "grad_norm": 0.23580900748913308,
+      "learning_rate": 2.1082199989950093e-05,
+      "loss": 0.1044,
+      "step": 3248
+    },
+    {
+      "epoch": 1.6913066111400312,
+      "grad_norm": 0.22358704155742054,
+      "learning_rate": 2.1068324150477346e-05,
+      "loss": 0.1069,
+      "step": 3249
+    },
+    {
+      "epoch": 1.691827173347215,
+      "grad_norm": 0.22829186908041899,
+      "learning_rate": 2.105444955281079e-05,
+      "loss": 0.1048,
+      "step": 3250
+    },
+    {
+      "epoch": 1.6923477355543988,
+      "grad_norm": 0.22170478945882835,
+      "learning_rate": 2.1040576201332685e-05,
+      "loss": 0.1044,
+      "step": 3251
+    },
+    {
+      "epoch": 1.6928682977615825,
+      "grad_norm": 0.23340110469012113,
+      "learning_rate": 2.1026704100424864e-05,
+      "loss": 0.1078,
+      "step": 3252
+    },
+    {
+      "epoch": 1.6933888599687663,
+      "grad_norm": 0.22247959374329498,
+      "learning_rate": 2.10128332544688e-05,
+      "loss": 0.1072,
+      "step": 3253
+    },
+    {
+      "epoch": 1.6939094221759499,
+      "grad_norm": 0.24563634319942448,
+      "learning_rate": 2.0998963667845535e-05,
+      "loss": 0.1069,
+      "step": 3254
+    },
+    {
+      "epoch": 1.6944299843831336,
+      "grad_norm": 0.2222993782818447,
+      "learning_rate": 2.0985095344935733e-05,
+      "loss": 0.1089,
+      "step": 3255
+    },
+    {
+      "epoch": 1.6949505465903174,
+      "grad_norm": 0.23032803072350344,
+      "learning_rate": 2.0971228290119664e-05,
+      "loss": 0.1065,
+      "step": 3256
+    },
+    {
+      "epoch": 1.6954711087975012,
+      "grad_norm": 0.22538817632991662,
+      "learning_rate": 2.0957362507777176e-05,
+      "loss": 0.1109,
+      "step": 3257
+    },
+    {
+      "epoch": 1.695991671004685,
+      "grad_norm": 0.22220876229186903,
+      "learning_rate": 2.0943498002287743e-05,
+      "loss": 0.1089,
+      "step": 3258
+    },
+    {
+      "epoch": 1.6965122332118687,
+      "grad_norm": 0.22756130766303528,
+      "learning_rate": 2.0929634778030408e-05,
+      "loss": 0.1082,
+      "step": 3259
+    },
+    {
+      "epoch": 1.6970327954190525,
+      "grad_norm": 0.2324022767031974,
+      "learning_rate": 2.0915772839383834e-05,
+      "loss": 0.1048,
+      "step": 3260
+    },
+    {
+      "epoch": 1.6975533576262363,
+      "grad_norm": 0.2224932908070099,
+      "learning_rate": 2.0901912190726256e-05,
+      "loss": 0.1052,
+      "step": 3261
+    },
+    {
+      "epoch": 1.69807391983342,
+      "grad_norm": 0.22064157879488797,
+      "learning_rate": 2.0888052836435524e-05,
+      "loss": 0.1034,
+      "step": 3262
+    },
+    {
+      "epoch": 1.6985944820406038,
+      "grad_norm": 0.22010208727090902,
+      "learning_rate": 2.087419478088906e-05,
+      "loss": 0.1048,
+      "step": 3263
+    },
+    {
+      "epoch": 1.6991150442477876,
+      "grad_norm": 0.2454058370043552,
+      "learning_rate": 2.0860338028463876e-05,
+      "loss": 0.1032,
+      "step": 3264
+    },
+    {
+      "epoch": 1.6996356064549714,
+      "grad_norm": 0.22591759232291647,
+      "learning_rate": 2.084648258353659e-05,
+      "loss": 0.1041,
+      "step": 3265
+    },
+    {
+      "epoch": 1.7001561686621551,
+      "grad_norm": 0.2269876270874768,
+      "learning_rate": 2.0832628450483388e-05,
+      "loss": 0.1054,
+      "step": 3266
+    },
+    {
+      "epoch": 1.700676730869339,
+      "grad_norm": 0.2323776942869489,
+      "learning_rate": 2.081877563368006e-05,
+      "loss": 0.107,
+      "step": 3267
+    },
+    {
+      "epoch": 1.7011972930765227,
+      "grad_norm": 0.23268241409871376,
+      "learning_rate": 2.0804924137501955e-05,
+      "loss": 0.1056,
+      "step": 3268
+    },
+    {
+      "epoch": 1.7017178552837064,
+      "grad_norm": 0.2356301736735351,
+      "learning_rate": 2.0791073966324037e-05,
+      "loss": 0.1035,
+      "step": 3269
+    },
+    {
+      "epoch": 1.7022384174908902,
+      "grad_norm": 0.22499277174575544,
+      "learning_rate": 2.0777225124520823e-05,
+      "loss": 0.1054,
+      "step": 3270
+    },
+    {
+      "epoch": 1.702758979698074,
+      "grad_norm": 0.2304220231846641,
+      "learning_rate": 2.0763377616466427e-05,
+      "loss": 0.108,
+      "step": 3271
+    },
+    {
+      "epoch": 1.7032795419052578,
+      "grad_norm": 0.21825438924376672,
+      "learning_rate": 2.0749531446534546e-05,
+      "loss": 0.1046,
+      "step": 3272
+    },
+    {
+      "epoch": 1.7038001041124415,
+      "grad_norm": 0.2237918265772953,
+      "learning_rate": 2.073568661909842e-05,
+      "loss": 0.1069,
+      "step": 3273
+    },
+    {
+      "epoch": 1.7043206663196253,
+      "grad_norm": 0.22614928472918566,
+      "learning_rate": 2.072184313853091e-05,
+      "loss": 0.1063,
+      "step": 3274
+    },
+    {
+      "epoch": 1.704841228526809,
+      "grad_norm": 0.21936797781205583,
+      "learning_rate": 2.070800100920442e-05,
+      "loss": 0.105,
+      "step": 3275
+    },
+    {
+      "epoch": 1.7053617907339929,
+      "grad_norm": 0.21812208623696525,
+      "learning_rate": 2.069416023549095e-05,
+      "loss": 0.1045,
+      "step": 3276
+    },
+    {
+      "epoch": 1.7058823529411766,
+      "grad_norm": 0.21349218293678732,
+      "learning_rate": 2.068032082176205e-05,
+      "loss": 0.1017,
+      "step": 3277
+    },
+    {
+      "epoch": 1.7064029151483602,
+      "grad_norm": 0.23388102739671973,
+      "learning_rate": 2.0666482772388853e-05,
+      "loss": 0.1065,
+      "step": 3278
+    },
+    {
+      "epoch": 1.706923477355544,
+      "grad_norm": 0.2114916830371581,
+      "learning_rate": 2.0652646091742063e-05,
+      "loss": 0.1001,
+      "step": 3279
+    },
+    {
+      "epoch": 1.7074440395627277,
+      "grad_norm": 0.22727043325186014,
+      "learning_rate": 2.0638810784191946e-05,
+      "loss": 0.1113,
+      "step": 3280
+    },
+    {
+      "epoch": 1.7079646017699115,
+      "grad_norm": 0.2509541792776318,
+      "learning_rate": 2.0624976854108347e-05,
+      "loss": 0.1133,
+      "step": 3281
+    },
+    {
+      "epoch": 1.7084851639770953,
+      "grad_norm": 0.22983253253459388,
+      "learning_rate": 2.061114430586064e-05,
+      "loss": 0.1056,
+      "step": 3282
+    },
+    {
+      "epoch": 1.709005726184279,
+      "grad_norm": 0.22546592705998422,
+      "learning_rate": 2.0597313143817804e-05,
+      "loss": 0.1074,
+      "step": 3283
+    },
+    {
+      "epoch": 1.7095262883914628,
+      "grad_norm": 0.22447733642224385,
+      "learning_rate": 2.0583483372348356e-05,
+      "loss": 0.1051,
+      "step": 3284
+    },
+    {
+      "epoch": 1.7100468505986466,
+      "grad_norm": 0.22044225269872403,
+      "learning_rate": 2.056965499582039e-05,
+      "loss": 0.107,
+      "step": 3285
+    },
+    {
+      "epoch": 1.7105674128058301,
+      "grad_norm": 0.23172108119736853,
+      "learning_rate": 2.055582801860155e-05,
+      "loss": 0.1101,
+      "step": 3286
+    },
+    {
+      "epoch": 1.711087975013014,
+      "grad_norm": 0.22418510588075263,
+      "learning_rate": 2.0542002445059032e-05,
+      "loss": 0.1031,
+      "step": 3287
+    },
+    {
+      "epoch": 1.7116085372201977,
+      "grad_norm": 0.21742221769182057,
+      "learning_rate": 2.0528178279559596e-05,
+      "loss": 0.1112,
+      "step": 3288
+    },
+    {
+      "epoch": 1.7121290994273815,
+      "grad_norm": 0.2285310956948703,
+      "learning_rate": 2.0514355526469566e-05,
+      "loss": 0.1082,
+      "step": 3289
+    },
+    {
+      "epoch": 1.7126496616345652,
+      "grad_norm": 0.23733520699349175,
+      "learning_rate": 2.0500534190154808e-05,
+      "loss": 0.1108,
+      "step": 3290
+    },
+    {
+      "epoch": 1.713170223841749,
+      "grad_norm": 0.224702532100698,
+      "learning_rate": 2.0486714274980732e-05,
+      "loss": 0.1054,
+      "step": 3291
+    },
+    {
+      "epoch": 1.7136907860489328,
+      "grad_norm": 0.23380286724902663,
+      "learning_rate": 2.0472895785312324e-05,
+      "loss": 0.1042,
+      "step": 3292
+    },
+    {
+      "epoch": 1.7142113482561165,
+      "grad_norm": 0.22363815675761445,
+      "learning_rate": 2.0459078725514092e-05,
+      "loss": 0.0998,
+      "step": 3293
+    },
+    {
+      "epoch": 1.7147319104633003,
+      "grad_norm": 0.23504429753004497,
+      "learning_rate": 2.0445263099950123e-05,
+      "loss": 0.1084,
+      "step": 3294
+    },
+    {
+      "epoch": 1.715252472670484,
+      "grad_norm": 0.22724060712973546,
+      "learning_rate": 2.043144891298402e-05,
+      "loss": 0.1048,
+      "step": 3295
+    },
+    {
+      "epoch": 1.7157730348776679,
+      "grad_norm": 0.23183122000489004,
+      "learning_rate": 2.0417636168978954e-05,
+      "loss": 0.0994,
+      "step": 3296
+    },
+    {
+      "epoch": 1.7162935970848516,
+      "grad_norm": 0.2417008008534072,
+      "learning_rate": 2.040382487229763e-05,
+      "loss": 0.108,
+      "step": 3297
+    },
+    {
+      "epoch": 1.7168141592920354,
+      "grad_norm": 0.21706572332557506,
+      "learning_rate": 2.03900150273023e-05,
+      "loss": 0.1044,
+      "step": 3298
+    },
+    {
+      "epoch": 1.7173347214992192,
+      "grad_norm": 0.2250832016912324,
+      "learning_rate": 2.0376206638354766e-05,
+      "loss": 0.1072,
+      "step": 3299
+    },
+    {
+      "epoch": 1.717855283706403,
+      "grad_norm": 0.21945238165897196,
+      "learning_rate": 2.036239970981633e-05,
+      "loss": 0.1056,
+      "step": 3300
+    },
+    {
+      "epoch": 1.7183758459135867,
+      "grad_norm": 0.2217903336181355,
+      "learning_rate": 2.0348594246047893e-05,
+      "loss": 0.1022,
+      "step": 3301
+    },
+    {
+      "epoch": 1.7188964081207705,
+      "grad_norm": 0.21983648892682253,
+      "learning_rate": 2.0334790251409845e-05,
+      "loss": 0.1006,
+      "step": 3302
+    },
+    {
+      "epoch": 1.7194169703279543,
+      "grad_norm": 0.2271718283385726,
+      "learning_rate": 2.0320987730262132e-05,
+      "loss": 0.1071,
+      "step": 3303
+    },
+    {
+      "epoch": 1.719937532535138,
+      "grad_norm": 0.2199145587427751,
+      "learning_rate": 2.0307186686964245e-05,
+      "loss": 0.1073,
+      "step": 3304
+    },
+    {
+      "epoch": 1.7204580947423218,
+      "grad_norm": 0.2231189412324603,
+      "learning_rate": 2.029338712587518e-05,
+      "loss": 0.1042,
+      "step": 3305
+    },
+    {
+      "epoch": 1.7209786569495056,
+      "grad_norm": 0.22383118284350628,
+      "learning_rate": 2.027958905135349e-05,
+      "loss": 0.1073,
+      "step": 3306
+    },
+    {
+      "epoch": 1.7214992191566894,
+      "grad_norm": 0.2179631138427114,
+      "learning_rate": 2.0265792467757248e-05,
+      "loss": 0.107,
+      "step": 3307
+    },
+    {
+      "epoch": 1.7220197813638731,
+      "grad_norm": 0.2367574980395077,
+      "learning_rate": 2.0251997379444062e-05,
+      "loss": 0.1114,
+      "step": 3308
+    },
+    {
+      "epoch": 1.722540343571057,
+      "grad_norm": 0.2103000613187075,
+      "learning_rate": 2.0238203790771054e-05,
+      "loss": 0.1028,
+      "step": 3309
+    },
+    {
+      "epoch": 1.7230609057782404,
+      "grad_norm": 0.21798620187439052,
+      "learning_rate": 2.0224411706094877e-05,
+      "loss": 0.1044,
+      "step": 3310
+    },
+    {
+      "epoch": 1.7235814679854242,
+      "grad_norm": 0.20804477935141746,
+      "learning_rate": 2.0210621129771722e-05,
+      "loss": 0.101,
+      "step": 3311
+    },
+    {
+      "epoch": 1.724102030192608,
+      "grad_norm": 0.22253756149900092,
+      "learning_rate": 2.019683206615729e-05,
+      "loss": 0.1045,
+      "step": 3312
+    },
+    {
+      "epoch": 1.7246225923997918,
+      "grad_norm": 0.21956300028974318,
+      "learning_rate": 2.018304451960682e-05,
+      "loss": 0.1073,
+      "step": 3313
+    },
+    {
+      "epoch": 1.7251431546069755,
+      "grad_norm": 0.21104867004615668,
+      "learning_rate": 2.016925849447504e-05,
+      "loss": 0.1049,
+      "step": 3314
+    },
+    {
+      "epoch": 1.7256637168141593,
+      "grad_norm": 0.21007160419878684,
+      "learning_rate": 2.015547399511624e-05,
+      "loss": 0.1002,
+      "step": 3315
+    },
+    {
+      "epoch": 1.726184279021343,
+      "grad_norm": 0.21210442829270762,
+      "learning_rate": 2.0141691025884195e-05,
+      "loss": 0.1039,
+      "step": 3316
+    },
+    {
+      "epoch": 1.7267048412285269,
+      "grad_norm": 0.20410554097153225,
+      "learning_rate": 2.0127909591132217e-05,
+      "loss": 0.1004,
+      "step": 3317
+    },
+    {
+      "epoch": 1.7272254034357104,
+      "grad_norm": 0.23935102740693862,
+      "learning_rate": 2.0114129695213114e-05,
+      "loss": 0.1056,
+      "step": 3318
+    },
+    {
+      "epoch": 1.7277459656428942,
+      "grad_norm": 0.23313707905631195,
+      "learning_rate": 2.0100351342479216e-05,
+      "loss": 0.1102,
+      "step": 3319
+    },
+    {
+      "epoch": 1.728266527850078,
+      "grad_norm": 0.23139683180145568,
+      "learning_rate": 2.008657453728238e-05,
+      "loss": 0.1099,
+      "step": 3320
+    },
+    {
+      "epoch": 1.7287870900572617,
+      "grad_norm": 0.24872510544166382,
+      "learning_rate": 2.007279928397395e-05,
+      "loss": 0.1066,
+      "step": 3321
+    },
+    {
+      "epoch": 1.7293076522644455,
+      "grad_norm": 0.23200706308434033,
+      "learning_rate": 2.0059025586904807e-05,
+      "loss": 0.1054,
+      "step": 3322
+    },
+    {
+      "epoch": 1.7298282144716293,
+      "grad_norm": 0.21942541308934396,
+      "learning_rate": 2.004525345042531e-05,
+      "loss": 0.1028,
+      "step": 3323
+    },
+    {
+      "epoch": 1.730348776678813,
+      "grad_norm": 0.2609746790506161,
+      "learning_rate": 2.003148287888535e-05,
+      "loss": 0.1088,
+      "step": 3324
+    },
+    {
+      "epoch": 1.7308693388859968,
+      "grad_norm": 0.2366478927625931,
+      "learning_rate": 2.0017713876634305e-05,
+      "loss": 0.1059,
+      "step": 3325
+    },
+    {
+      "epoch": 1.7313899010931806,
+      "grad_norm": 0.22258749938323874,
+      "learning_rate": 2.000394644802109e-05,
+      "loss": 0.105,
+      "step": 3326
+    },
+    {
+      "epoch": 1.7319104633003644,
+      "grad_norm": 0.22200613011043596,
+      "learning_rate": 1.9990180597394075e-05,
+      "loss": 0.1024,
+      "step": 3327
+    },
+    {
+      "epoch": 1.7324310255075481,
+      "grad_norm": 0.22635255297590737,
+      "learning_rate": 1.9976416329101154e-05,
+      "loss": 0.1112,
+      "step": 3328
+    },
+    {
+      "epoch": 1.732951587714732,
+      "grad_norm": 0.21816833986696646,
+      "learning_rate": 1.9962653647489745e-05,
+      "loss": 0.1083,
+      "step": 3329
+    },
+    {
+      "epoch": 1.7334721499219157,
+      "grad_norm": 0.2320214874149184,
+      "learning_rate": 1.9948892556906727e-05,
+      "loss": 0.1063,
+      "step": 3330
+    },
+    {
+      "epoch": 1.7339927121290994,
+      "grad_norm": 0.21304345539614472,
+      "learning_rate": 1.99351330616985e-05,
+      "loss": 0.1058,
+      "step": 3331
+    },
+    {
+      "epoch": 1.7345132743362832,
+      "grad_norm": 0.21513285495170326,
+      "learning_rate": 1.9921375166210948e-05,
+      "loss": 0.1045,
+      "step": 3332
+    },
+    {
+      "epoch": 1.735033836543467,
+      "grad_norm": 0.23454273842647397,
+      "learning_rate": 1.990761887478946e-05,
+      "loss": 0.1115,
+      "step": 3333
+    },
+    {
+      "epoch": 1.7355543987506508,
+      "grad_norm": 0.22338156044764015,
+      "learning_rate": 1.989386419177891e-05,
+      "loss": 0.1082,
+      "step": 3334
+    },
+    {
+      "epoch": 1.7360749609578345,
+      "grad_norm": 0.20836963730690877,
+      "learning_rate": 1.988011112152367e-05,
+      "loss": 0.1041,
+      "step": 3335
+    },
+    {
+      "epoch": 1.7365955231650183,
+      "grad_norm": 0.2236097112876602,
+      "learning_rate": 1.9866359668367594e-05,
+      "loss": 0.1086,
+      "step": 3336
+    },
+    {
+      "epoch": 1.737116085372202,
+      "grad_norm": 0.21123759334297273,
+      "learning_rate": 1.9852609836654034e-05,
+      "loss": 0.1062,
+      "step": 3337
+    },
+    {
+      "epoch": 1.7376366475793859,
+      "grad_norm": 0.22237040640657335,
+      "learning_rate": 1.9838861630725826e-05,
+      "loss": 0.108,
+      "step": 3338
+    },
+    {
+      "epoch": 1.7381572097865696,
+      "grad_norm": 0.2495840089981045,
+      "learning_rate": 1.982511505492529e-05,
+      "loss": 0.1076,
+      "step": 3339
+    },
+    {
+      "epoch": 1.7386777719937534,
+      "grad_norm": 0.22958506111561988,
+      "learning_rate": 1.9811370113594246e-05,
+      "loss": 0.1047,
+      "step": 3340
+    },
+    {
+      "epoch": 1.7391983342009372,
+      "grad_norm": 0.216172102598489,
+      "learning_rate": 1.9797626811073972e-05,
+      "loss": 0.1027,
+      "step": 3341
+    },
+    {
+      "epoch": 1.7397188964081207,
+      "grad_norm": 0.2221979457825088,
+      "learning_rate": 1.9783885151705252e-05,
+      "loss": 0.1056,
+      "step": 3342
+    },
+    {
+      "epoch": 1.7402394586153045,
+      "grad_norm": 0.22693748401563166,
+      "learning_rate": 1.9770145139828333e-05,
+      "loss": 0.1086,
+      "step": 3343
+    },
+    {
+      "epoch": 1.7407600208224883,
+      "grad_norm": 0.2273052352822081,
+      "learning_rate": 1.975640677978297e-05,
+      "loss": 0.1023,
+      "step": 3344
+    },
+    {
+      "epoch": 1.741280583029672,
+      "grad_norm": 0.23018043315752323,
+      "learning_rate": 1.974267007590835e-05,
+      "loss": 0.1073,
+      "step": 3345
+    },
+    {
+      "epoch": 1.7418011452368558,
+      "grad_norm": 0.21977354781533529,
+      "learning_rate": 1.9728935032543174e-05,
+      "loss": 0.1025,
+      "step": 3346
+    },
+    {
+      "epoch": 1.7423217074440396,
+      "grad_norm": 0.22170578071619307,
+      "learning_rate": 1.9715201654025614e-05,
+      "loss": 0.1053,
+      "step": 3347
+    },
+    {
+      "epoch": 1.7428422696512234,
+      "grad_norm": 0.22697352459866138,
+      "learning_rate": 1.9701469944693298e-05,
+      "loss": 0.1045,
+      "step": 3348
+    },
+    {
+      "epoch": 1.7433628318584071,
+      "grad_norm": 0.21983474061455138,
+      "learning_rate": 1.9687739908883352e-05,
+      "loss": 0.1015,
+      "step": 3349
+    },
+    {
+      "epoch": 1.7438833940655907,
+      "grad_norm": 0.22230299934979136,
+      "learning_rate": 1.967401155093235e-05,
+      "loss": 0.1075,
+      "step": 3350
+    },
+    {
+      "epoch": 1.7444039562727744,
+      "grad_norm": 0.2278927921689901,
+      "learning_rate": 1.9660284875176354e-05,
+      "loss": 0.104,
+      "step": 3351
+    },
+    {
+      "epoch": 1.7449245184799582,
+      "grad_norm": 0.2258066335818804,
+      "learning_rate": 1.9646559885950876e-05,
+      "loss": 0.1072,
+      "step": 3352
+    },
+    {
+      "epoch": 1.745445080687142,
+      "grad_norm": 0.21809762204524716,
+      "learning_rate": 1.9632836587590928e-05,
+      "loss": 0.1057,
+      "step": 3353
+    },
+    {
+      "epoch": 1.7459656428943258,
+      "grad_norm": 0.2332642653387811,
+      "learning_rate": 1.9619114984430946e-05,
+      "loss": 0.1101,
+      "step": 3354
+    },
+    {
+      "epoch": 1.7464862051015095,
+      "grad_norm": 0.22377591593309243,
+      "learning_rate": 1.960539508080485e-05,
+      "loss": 0.1034,
+      "step": 3355
+    },
+    {
+      "epoch": 1.7470067673086933,
+      "grad_norm": 0.2214624251282975,
+      "learning_rate": 1.9591676881046038e-05,
+      "loss": 0.1051,
+      "step": 3356
+    },
+    {
+      "epoch": 1.747527329515877,
+      "grad_norm": 0.22969375504708825,
+      "learning_rate": 1.957796038948734e-05,
+      "loss": 0.1042,
+      "step": 3357
+    },
+    {
+      "epoch": 1.7480478917230609,
+      "grad_norm": 0.2212002516504855,
+      "learning_rate": 1.9564245610461078e-05,
+      "loss": 0.1056,
+      "step": 3358
+    },
+    {
+      "epoch": 1.7485684539302446,
+      "grad_norm": 0.226595022233284,
+      "learning_rate": 1.955053254829901e-05,
+      "loss": 0.1022,
+      "step": 3359
+    },
+    {
+      "epoch": 1.7490890161374284,
+      "grad_norm": 0.21300220127544178,
+      "learning_rate": 1.9536821207332357e-05,
+      "loss": 0.1032,
+      "step": 3360
+    },
+    {
+      "epoch": 1.7496095783446122,
+      "grad_norm": 0.2237621475683394,
+      "learning_rate": 1.95231115918918e-05,
+      "loss": 0.1029,
+      "step": 3361
+    },
+    {
+      "epoch": 1.750130140551796,
+      "grad_norm": 0.22048593748404666,
+      "learning_rate": 1.9509403706307484e-05,
+      "loss": 0.1053,
+      "step": 3362
+    },
+    {
+      "epoch": 1.7506507027589797,
+      "grad_norm": 0.22433810240135404,
+      "learning_rate": 1.9495697554908986e-05,
+      "loss": 0.102,
+      "step": 3363
+    },
+    {
+      "epoch": 1.7511712649661635,
+      "grad_norm": 0.22380034181909983,
+      "learning_rate": 1.948199314202534e-05,
+      "loss": 0.0986,
+      "step": 3364
+    },
+    {
+      "epoch": 1.7516918271733473,
+      "grad_norm": 0.22338005341310205,
+      "learning_rate": 1.946829047198505e-05,
+      "loss": 0.1035,
+      "step": 3365
+    },
+    {
+      "epoch": 1.752212389380531,
+      "grad_norm": 0.230442140004699,
+      "learning_rate": 1.945458954911605e-05,
+      "loss": 0.108,
+      "step": 3366
+    },
+    {
+      "epoch": 1.7527329515877148,
+      "grad_norm": 0.2304987431815853,
+      "learning_rate": 1.944089037774573e-05,
+      "loss": 0.1082,
+      "step": 3367
+    },
+    {
+      "epoch": 1.7532535137948986,
+      "grad_norm": 0.2214657123371355,
+      "learning_rate": 1.942719296220093e-05,
+      "loss": 0.1027,
+      "step": 3368
+    },
+    {
+      "epoch": 1.7537740760020823,
+      "grad_norm": 0.20755785482599343,
+      "learning_rate": 1.9413497306807925e-05,
+      "loss": 0.1005,
+      "step": 3369
+    },
+    {
+      "epoch": 1.7542946382092661,
+      "grad_norm": 0.2109976471942146,
+      "learning_rate": 1.9399803415892454e-05,
+      "loss": 0.1001,
+      "step": 3370
+    },
+    {
+      "epoch": 1.75481520041645,
+      "grad_norm": 0.22412574394179605,
+      "learning_rate": 1.9386111293779673e-05,
+      "loss": 0.1011,
+      "step": 3371
+    },
+    {
+      "epoch": 1.7553357626236337,
+      "grad_norm": 0.2276297126469934,
+      "learning_rate": 1.937242094479419e-05,
+      "loss": 0.1016,
+      "step": 3372
+    },
+    {
+      "epoch": 1.7558563248308174,
+      "grad_norm": 0.224838207969345,
+      "learning_rate": 1.9358732373260056e-05,
+      "loss": 0.1049,
+      "step": 3373
+    },
+    {
+      "epoch": 1.756376887038001,
+      "grad_norm": 0.2211883964455484,
+      "learning_rate": 1.934504558350076e-05,
+      "loss": 0.1054,
+      "step": 3374
+    },
+    {
+      "epoch": 1.7568974492451848,
+      "grad_norm": 0.22067590673603887,
+      "learning_rate": 1.933136057983923e-05,
+      "loss": 0.1013,
+      "step": 3375
+    },
+    {
+      "epoch": 1.7574180114523685,
+      "grad_norm": 0.22690702349673778,
+      "learning_rate": 1.931767736659782e-05,
+      "loss": 0.1032,
+      "step": 3376
+    },
+    {
+      "epoch": 1.7579385736595523,
+      "grad_norm": 0.22520906999437268,
+      "learning_rate": 1.930399594809834e-05,
+      "loss": 0.1069,
+      "step": 3377
+    },
+    {
+      "epoch": 1.758459135866736,
+      "grad_norm": 0.22649242565754607,
+      "learning_rate": 1.9290316328662e-05,
+      "loss": 0.1049,
+      "step": 3378
+    },
+    {
+      "epoch": 1.7589796980739199,
+      "grad_norm": 0.22332071105437934,
+      "learning_rate": 1.9276638512609474e-05,
+      "loss": 0.1059,
+      "step": 3379
+    },
+    {
+      "epoch": 1.7595002602811036,
+      "grad_norm": 0.21894476859270845,
+      "learning_rate": 1.926296250426085e-05,
+      "loss": 0.1038,
+      "step": 3380
+    },
+    {
+      "epoch": 1.7600208224882874,
+      "grad_norm": 0.2303137164013817,
+      "learning_rate": 1.9249288307935642e-05,
+      "loss": 0.1055,
+      "step": 3381
+    },
+    {
+      "epoch": 1.760541384695471,
+      "grad_norm": 0.22126996251729855,
+      "learning_rate": 1.9235615927952804e-05,
+      "loss": 0.107,
+      "step": 3382
+    },
+    {
+      "epoch": 1.7610619469026547,
+      "grad_norm": 0.21780067256343733,
+      "learning_rate": 1.92219453686307e-05,
+      "loss": 0.1032,
+      "step": 3383
+    },
+    {
+      "epoch": 1.7615825091098385,
+      "grad_norm": 0.21443493771388783,
+      "learning_rate": 1.9208276634287143e-05,
+      "loss": 0.1047,
+      "step": 3384
+    },
+    {
+      "epoch": 1.7621030713170223,
+      "grad_norm": 0.22561009131363186,
+      "learning_rate": 1.9194609729239344e-05,
+      "loss": 0.1056,
+      "step": 3385
+    },
+    {
+      "epoch": 1.762623633524206,
+      "grad_norm": 0.22455128134416372,
+      "learning_rate": 1.9180944657803956e-05,
+      "loss": 0.1071,
+      "step": 3386
+    },
+    {
+      "epoch": 1.7631441957313898,
+      "grad_norm": 0.2264693768229407,
+      "learning_rate": 1.9167281424297035e-05,
+      "loss": 0.1109,
+      "step": 3387
+    },
+    {
+      "epoch": 1.7636647579385736,
+      "grad_norm": 0.2107756363167944,
+      "learning_rate": 1.9153620033034075e-05,
+      "loss": 0.1059,
+      "step": 3388
+    },
+    {
+      "epoch": 1.7641853201457574,
+      "grad_norm": 0.23268370370945063,
+      "learning_rate": 1.9139960488329985e-05,
+      "loss": 0.1125,
+      "step": 3389
+    },
+    {
+      "epoch": 1.7647058823529411,
+      "grad_norm": 0.2373633791421892,
+      "learning_rate": 1.912630279449906e-05,
+      "loss": 0.1025,
+      "step": 3390
+    },
+    {
+      "epoch": 1.765226444560125,
+      "grad_norm": 0.2310037923502031,
+      "learning_rate": 1.911264695585506e-05,
+      "loss": 0.1089,
+      "step": 3391
+    },
+    {
+      "epoch": 1.7657470067673087,
+      "grad_norm": 0.2123561360419581,
+      "learning_rate": 1.9098992976711123e-05,
+      "loss": 0.1023,
+      "step": 3392
+    },
+    {
+      "epoch": 1.7662675689744924,
+      "grad_norm": 0.2306983798259102,
+      "learning_rate": 1.9085340861379813e-05,
+      "loss": 0.1126,
+      "step": 3393
+    },
+    {
+      "epoch": 1.7667881311816762,
+      "grad_norm": 0.23277860448314053,
+      "learning_rate": 1.9071690614173102e-05,
+      "loss": 0.1029,
+      "step": 3394
+    },
+    {
+      "epoch": 1.76730869338886,
+      "grad_norm": 0.2087088586057,
+      "learning_rate": 1.9058042239402378e-05,
+      "loss": 0.0985,
+      "step": 3395
+    },
+    {
+      "epoch": 1.7678292555960438,
+      "grad_norm": 0.21479438193487035,
+      "learning_rate": 1.9044395741378425e-05,
+      "loss": 0.1002,
+      "step": 3396
+    },
+    {
+      "epoch": 1.7683498178032275,
+      "grad_norm": 0.22921709100541876,
+      "learning_rate": 1.903075112441145e-05,
+      "loss": 0.1058,
+      "step": 3397
+    },
+    {
+      "epoch": 1.7688703800104113,
+      "grad_norm": 0.22434060260788052,
+      "learning_rate": 1.9017108392811065e-05,
+      "loss": 0.1044,
+      "step": 3398
+    },
+    {
+      "epoch": 1.769390942217595,
+      "grad_norm": 0.23525431934160432,
+      "learning_rate": 1.9003467550886253e-05,
+      "loss": 0.1125,
+      "step": 3399
+    },
+    {
+      "epoch": 1.7699115044247788,
+      "grad_norm": 0.22009803271545916,
+      "learning_rate": 1.8989828602945454e-05,
+      "loss": 0.1047,
+      "step": 3400
+    },
+    {
+      "epoch": 1.7704320666319626,
+      "grad_norm": 0.22537635925080748,
+      "learning_rate": 1.897619155329646e-05,
+      "loss": 0.1043,
+      "step": 3401
+    },
+    {
+      "epoch": 1.7709526288391464,
+      "grad_norm": 0.2367752202655961,
+      "learning_rate": 1.8962556406246505e-05,
+      "loss": 0.1035,
+      "step": 3402
+    },
+    {
+      "epoch": 1.7714731910463302,
+      "grad_norm": 0.20474450022021,
+      "learning_rate": 1.8948923166102192e-05,
+      "loss": 0.0994,
+      "step": 3403
+    },
+    {
+      "epoch": 1.771993753253514,
+      "grad_norm": 0.21437274278815563,
+      "learning_rate": 1.893529183716954e-05,
+      "loss": 0.0997,
+      "step": 3404
+    },
+    {
+      "epoch": 1.7725143154606977,
+      "grad_norm": 0.22322450199964058,
+      "learning_rate": 1.892166242375395e-05,
+      "loss": 0.0996,
+      "step": 3405
+    },
+    {
+      "epoch": 1.7730348776678813,
+      "grad_norm": 0.22023539278452614,
+      "learning_rate": 1.8908034930160228e-05,
+      "loss": 0.1039,
+      "step": 3406
+    },
+    {
+      "epoch": 1.773555439875065,
+      "grad_norm": 0.2114044276184025,
+      "learning_rate": 1.889440936069258e-05,
+      "loss": 0.1026,
+      "step": 3407
+    },
+    {
+      "epoch": 1.7740760020822488,
+      "grad_norm": 0.21695286591737722,
+      "learning_rate": 1.8880785719654577e-05,
+      "loss": 0.1037,
+      "step": 3408
+    },
+    {
+      "epoch": 1.7745965642894326,
+      "grad_norm": 0.2363632816801158,
+      "learning_rate": 1.8867164011349208e-05,
+      "loss": 0.1071,
+      "step": 3409
+    },
+    {
+      "epoch": 1.7751171264966163,
+      "grad_norm": 0.22771210911879075,
+      "learning_rate": 1.885354424007884e-05,
+      "loss": 0.1083,
+      "step": 3410
+    },
+    {
+      "epoch": 1.7756376887038001,
+      "grad_norm": 0.2274868597566939,
+      "learning_rate": 1.8839926410145235e-05,
+      "loss": 0.1077,
+      "step": 3411
+    },
+    {
+      "epoch": 1.776158250910984,
+      "grad_norm": 0.21891521186548685,
+      "learning_rate": 1.882631052584953e-05,
+      "loss": 0.1022,
+      "step": 3412
+    },
+    {
+      "epoch": 1.7766788131181677,
+      "grad_norm": 0.22044753666985667,
+      "learning_rate": 1.8812696591492265e-05,
+      "loss": 0.1027,
+      "step": 3413
+    },
+    {
+      "epoch": 1.7771993753253512,
+      "grad_norm": 0.23004226671888997,
+      "learning_rate": 1.8799084611373345e-05,
+      "loss": 0.1085,
+      "step": 3414
+    },
+    {
+      "epoch": 1.777719937532535,
+      "grad_norm": 0.21243628528006495,
+      "learning_rate": 1.8785474589792074e-05,
+      "loss": 0.1044,
+      "step": 3415
+    },
+    {
+      "epoch": 1.7782404997397188,
+      "grad_norm": 0.220110579204587,
+      "learning_rate": 1.8771866531047133e-05,
+      "loss": 0.1023,
+      "step": 3416
+    },
+    {
+      "epoch": 1.7787610619469025,
+      "grad_norm": 0.22305394623742394,
+      "learning_rate": 1.8758260439436563e-05,
+      "loss": 0.1081,
+      "step": 3417
+    },
+    {
+      "epoch": 1.7792816241540863,
+      "grad_norm": 0.2276585098729061,
+      "learning_rate": 1.8744656319257817e-05,
+      "loss": 0.1073,
+      "step": 3418
+    },
+    {
+      "epoch": 1.77980218636127,
+      "grad_norm": 0.22252174513686673,
+      "learning_rate": 1.87310541748077e-05,
+      "loss": 0.1079,
+      "step": 3419
+    },
+    {
+      "epoch": 1.7803227485684539,
+      "grad_norm": 0.21570267838003582,
+      "learning_rate": 1.8717454010382407e-05,
+      "loss": 0.1049,
+      "step": 3420
+    },
+    {
+      "epoch": 1.7808433107756376,
+      "grad_norm": 0.21778548124027283,
+      "learning_rate": 1.87038558302775e-05,
+      "loss": 0.1014,
+      "step": 3421
+    },
+    {
+      "epoch": 1.7813638729828214,
+      "grad_norm": 0.21552687109805088,
+      "learning_rate": 1.8690259638787926e-05,
+      "loss": 0.0991,
+      "step": 3422
+    },
+    {
+      "epoch": 1.7818844351900052,
+      "grad_norm": 0.2345671900413482,
+      "learning_rate": 1.867666544020798e-05,
+      "loss": 0.107,
+      "step": 3423
+    },
+    {
+      "epoch": 1.782404997397189,
+      "grad_norm": 0.23258757335297145,
+      "learning_rate": 1.866307323883136e-05,
+      "loss": 0.1105,
+      "step": 3424
+    },
+    {
+      "epoch": 1.7829255596043727,
+      "grad_norm": 0.21889647232741902,
+      "learning_rate": 1.8649483038951107e-05,
+      "loss": 0.107,
+      "step": 3425
+    },
+    {
+      "epoch": 1.7834461218115565,
+      "grad_norm": 0.22215635618720045,
+      "learning_rate": 1.863589484485963e-05,
+      "loss": 0.1066,
+      "step": 3426
+    },
+    {
+      "epoch": 1.7839666840187403,
+      "grad_norm": 0.21415158989479102,
+      "learning_rate": 1.8622308660848724e-05,
+      "loss": 0.1029,
+      "step": 3427
+    },
+    {
+      "epoch": 1.784487246225924,
+      "grad_norm": 0.22781030552769713,
+      "learning_rate": 1.860872449120953e-05,
+      "loss": 0.1087,
+      "step": 3428
+    },
+    {
+      "epoch": 1.7850078084331078,
+      "grad_norm": 0.22290517689214315,
+      "learning_rate": 1.8595142340232575e-05,
+      "loss": 0.1019,
+      "step": 3429
+    },
+    {
+      "epoch": 1.7855283706402916,
+      "grad_norm": 0.21633222759985674,
+      "learning_rate": 1.8581562212207714e-05,
+      "loss": 0.1021,
+      "step": 3430
+    },
+    {
+      "epoch": 1.7860489328474753,
+      "grad_norm": 0.230030532368222,
+      "learning_rate": 1.8567984111424205e-05,
+      "loss": 0.1109,
+      "step": 3431
+    },
+    {
+      "epoch": 1.7865694950546591,
+      "grad_norm": 0.22735979206413187,
+      "learning_rate": 1.8554408042170628e-05,
+      "loss": 0.1089,
+      "step": 3432
+    },
+    {
+      "epoch": 1.787090057261843,
+      "grad_norm": 0.2253631691551036,
+      "learning_rate": 1.8540834008734943e-05,
+      "loss": 0.1023,
+      "step": 3433
+    },
+    {
+      "epoch": 1.7876106194690267,
+      "grad_norm": 0.21969328560623388,
+      "learning_rate": 1.8527262015404477e-05,
+      "loss": 0.1046,
+      "step": 3434
+    },
+    {
+      "epoch": 1.7881311816762104,
+      "grad_norm": 0.22505681618644893,
+      "learning_rate": 1.8513692066465866e-05,
+      "loss": 0.1061,
+      "step": 3435
+    },
+    {
+      "epoch": 1.7886517438833942,
+      "grad_norm": 0.22392160124207264,
+      "learning_rate": 1.8500124166205152e-05,
+      "loss": 0.1055,
+      "step": 3436
+    },
+    {
+      "epoch": 1.789172306090578,
+      "grad_norm": 0.22703973700198465,
+      "learning_rate": 1.84865583189077e-05,
+      "loss": 0.1053,
+      "step": 3437
+    },
+    {
+      "epoch": 1.7896928682977615,
+      "grad_norm": 0.2235806887467908,
+      "learning_rate": 1.847299452885824e-05,
+      "loss": 0.1059,
+      "step": 3438
+    },
+    {
+      "epoch": 1.7902134305049453,
+      "grad_norm": 0.22095971299700543,
+      "learning_rate": 1.8459432800340855e-05,
+      "loss": 0.1015,
+      "step": 3439
+    },
+    {
+      "epoch": 1.790733992712129,
+      "grad_norm": 0.23339506867569496,
+      "learning_rate": 1.8445873137638953e-05,
+      "loss": 0.1044,
+      "step": 3440
+    },
+    {
+      "epoch": 1.7912545549193128,
+      "grad_norm": 0.21709782713883458,
+      "learning_rate": 1.8432315545035328e-05,
+      "loss": 0.1034,
+      "step": 3441
+    },
+    {
+      "epoch": 1.7917751171264966,
+      "grad_norm": 0.2165477854014385,
+      "learning_rate": 1.841876002681208e-05,
+      "loss": 0.1041,
+      "step": 3442
+    },
+    {
+      "epoch": 1.7922956793336804,
+      "grad_norm": 0.23599068704496834,
+      "learning_rate": 1.840520658725069e-05,
+      "loss": 0.1082,
+      "step": 3443
+    },
+    {
+      "epoch": 1.7928162415408642,
+      "grad_norm": 0.22660187330424075,
+      "learning_rate": 1.8391655230631953e-05,
+      "loss": 0.1083,
+      "step": 3444
+    },
+    {
+      "epoch": 1.793336803748048,
+      "grad_norm": 0.21818235771020988,
+      "learning_rate": 1.837810596123601e-05,
+      "loss": 0.1041,
+      "step": 3445
+    },
+    {
+      "epoch": 1.7938573659552315,
+      "grad_norm": 0.21184422424865806,
+      "learning_rate": 1.836455878334237e-05,
+      "loss": 0.1013,
+      "step": 3446
+    },
+    {
+      "epoch": 1.7943779281624153,
+      "grad_norm": 0.21802505383850326,
+      "learning_rate": 1.8351013701229846e-05,
+      "loss": 0.1054,
+      "step": 3447
+    },
+    {
+      "epoch": 1.794898490369599,
+      "grad_norm": 0.2155910468353952,
+      "learning_rate": 1.833747071917662e-05,
+      "loss": 0.1011,
+      "step": 3448
+    },
+    {
+      "epoch": 1.7954190525767828,
+      "grad_norm": 0.21140794526123036,
+      "learning_rate": 1.832392984146018e-05,
+      "loss": 0.1022,
+      "step": 3449
+    },
+    {
+      "epoch": 1.7959396147839666,
+      "grad_norm": 0.20930278134313027,
+      "learning_rate": 1.8310391072357382e-05,
+      "loss": 0.0955,
+      "step": 3450
+    },
+    {
+      "epoch": 1.7964601769911503,
+      "grad_norm": 0.2316473180383841,
+      "learning_rate": 1.829685441614438e-05,
+      "loss": 0.1055,
+      "step": 3451
+    },
+    {
+      "epoch": 1.7969807391983341,
+      "grad_norm": 0.23166286020742097,
+      "learning_rate": 1.8283319877096705e-05,
+      "loss": 0.1025,
+      "step": 3452
+    },
+    {
+      "epoch": 1.797501301405518,
+      "grad_norm": 0.22381973015644815,
+      "learning_rate": 1.8269787459489174e-05,
+      "loss": 0.1048,
+      "step": 3453
+    },
+    {
+      "epoch": 1.7980218636127017,
+      "grad_norm": 0.22774693899244658,
+      "learning_rate": 1.8256257167595957e-05,
+      "loss": 0.1028,
+      "step": 3454
+    },
+    {
+      "epoch": 1.7985424258198854,
+      "grad_norm": 0.22509542841432917,
+      "learning_rate": 1.8242729005690557e-05,
+      "loss": 0.1012,
+      "step": 3455
+    },
+    {
+      "epoch": 1.7990629880270692,
+      "grad_norm": 0.22202303234657486,
+      "learning_rate": 1.822920297804579e-05,
+      "loss": 0.1025,
+      "step": 3456
+    },
+    {
+      "epoch": 1.799583550234253,
+      "grad_norm": 0.2268089623653908,
+      "learning_rate": 1.8215679088933813e-05,
+      "loss": 0.1046,
+      "step": 3457
+    },
+    {
+      "epoch": 1.8001041124414368,
+      "grad_norm": 0.22470714993267865,
+      "learning_rate": 1.8202157342626087e-05,
+      "loss": 0.1064,
+      "step": 3458
+    },
+    {
+      "epoch": 1.8006246746486205,
+      "grad_norm": 0.2153205950636323,
+      "learning_rate": 1.8188637743393423e-05,
+      "loss": 0.1003,
+      "step": 3459
+    },
+    {
+      "epoch": 1.8011452368558043,
+      "grad_norm": 0.20954593624683357,
+      "learning_rate": 1.8175120295505925e-05,
+      "loss": 0.1031,
+      "step": 3460
+    },
+    {
+      "epoch": 1.801665799062988,
+      "grad_norm": 0.21853659329976693,
+      "learning_rate": 1.8161605003233056e-05,
+      "loss": 0.1043,
+      "step": 3461
+    },
+    {
+      "epoch": 1.8021863612701718,
+      "grad_norm": 0.21719611628693516,
+      "learning_rate": 1.8148091870843554e-05,
+      "loss": 0.1029,
+      "step": 3462
+    },
+    {
+      "epoch": 1.8027069234773556,
+      "grad_norm": 0.21032988240795134,
+      "learning_rate": 1.813458090260549e-05,
+      "loss": 0.1012,
+      "step": 3463
+    },
+    {
+      "epoch": 1.8032274856845394,
+      "grad_norm": 0.22087081939591202,
+      "learning_rate": 1.8121072102786274e-05,
+      "loss": 0.1082,
+      "step": 3464
+    },
+    {
+      "epoch": 1.8037480478917232,
+      "grad_norm": 0.2190982660554764,
+      "learning_rate": 1.81075654756526e-05,
+      "loss": 0.1058,
+      "step": 3465
+    },
+    {
+      "epoch": 1.804268610098907,
+      "grad_norm": 0.22040569387096418,
+      "learning_rate": 1.8094061025470498e-05,
+      "loss": 0.1001,
+      "step": 3466
+    },
+    {
+      "epoch": 1.8047891723060907,
+      "grad_norm": 0.21603368220817645,
+      "learning_rate": 1.8080558756505294e-05,
+      "loss": 0.0998,
+      "step": 3467
+    },
+    {
+      "epoch": 1.8053097345132745,
+      "grad_norm": 0.20574278416396982,
+      "learning_rate": 1.8067058673021646e-05,
+      "loss": 0.099,
+      "step": 3468
+    },
+    {
+      "epoch": 1.8058302967204583,
+      "grad_norm": 0.2277361813536661,
+      "learning_rate": 1.8053560779283495e-05,
+      "loss": 0.1053,
+      "step": 3469
+    },
+    {
+      "epoch": 1.8063508589276418,
+      "grad_norm": 0.21219090161169762,
+      "learning_rate": 1.804006507955411e-05,
+      "loss": 0.1013,
+      "step": 3470
+    },
+    {
+      "epoch": 1.8068714211348256,
+      "grad_norm": 0.22700160243005704,
+      "learning_rate": 1.8026571578096068e-05,
+      "loss": 0.1038,
+      "step": 3471
+    },
+    {
+      "epoch": 1.8073919833420093,
+      "grad_norm": 0.2369122624279523,
+      "learning_rate": 1.8013080279171228e-05,
+      "loss": 0.1123,
+      "step": 3472
+    },
+    {
+      "epoch": 1.8079125455491931,
+      "grad_norm": 0.22360440845976917,
+      "learning_rate": 1.799959118704078e-05,
+      "loss": 0.1048,
+      "step": 3473
+    },
+    {
+      "epoch": 1.808433107756377,
+      "grad_norm": 0.2104592319944269,
+      "learning_rate": 1.7986104305965205e-05,
+      "loss": 0.0973,
+      "step": 3474
+    },
+    {
+      "epoch": 1.8089536699635607,
+      "grad_norm": 0.218593446956845,
+      "learning_rate": 1.7972619640204296e-05,
+      "loss": 0.1043,
+      "step": 3475
+    },
+    {
+      "epoch": 1.8094742321707444,
+      "grad_norm": 0.22148524142264853,
+      "learning_rate": 1.795913719401712e-05,
+      "loss": 0.1004,
+      "step": 3476
+    },
+    {
+      "epoch": 1.8099947943779282,
+      "grad_norm": 0.2155675656645299,
+      "learning_rate": 1.7945656971662085e-05,
+      "loss": 0.1029,
+      "step": 3477
+    },
+    {
+      "epoch": 1.8105153565851118,
+      "grad_norm": 0.23004321978824924,
+      "learning_rate": 1.7932178977396848e-05,
+      "loss": 0.1071,
+      "step": 3478
+    },
+    {
+      "epoch": 1.8110359187922955,
+      "grad_norm": 0.2323235989966905,
+      "learning_rate": 1.79187032154784e-05,
+      "loss": 0.1088,
+      "step": 3479
+    },
+    {
+      "epoch": 1.8115564809994793,
+      "grad_norm": 0.2174771770164572,
+      "learning_rate": 1.7905229690163023e-05,
+      "loss": 0.1032,
+      "step": 3480
+    },
+    {
+      "epoch": 1.812077043206663,
+      "grad_norm": 0.21461562545103127,
+      "learning_rate": 1.789175840570626e-05,
+      "loss": 0.1015,
+      "step": 3481
+    },
+    {
+      "epoch": 1.8125976054138468,
+      "grad_norm": 0.21080577930269348,
+      "learning_rate": 1.7878289366362984e-05,
+      "loss": 0.1057,
+      "step": 3482
+    },
+    {
+      "epoch": 1.8131181676210306,
+      "grad_norm": 0.2166558884121671,
+      "learning_rate": 1.7864822576387342e-05,
+      "loss": 0.1021,
+      "step": 3483
+    },
+    {
+      "epoch": 1.8136387298282144,
+      "grad_norm": 0.22936772903160424,
+      "learning_rate": 1.7851358040032774e-05,
+      "loss": 0.1058,
+      "step": 3484
+    },
+    {
+      "epoch": 1.8141592920353982,
+      "grad_norm": 0.21693822231883433,
+      "learning_rate": 1.7837895761552002e-05,
+      "loss": 0.0993,
+      "step": 3485
+    },
+    {
+      "epoch": 1.814679854242582,
+      "grad_norm": 0.21424260587284594,
+      "learning_rate": 1.782443574519705e-05,
+      "loss": 0.1045,
+      "step": 3486
+    },
+    {
+      "epoch": 1.8152004164497657,
+      "grad_norm": 0.21784231944571822,
+      "learning_rate": 1.7810977995219203e-05,
+      "loss": 0.101,
+      "step": 3487
+    },
+    {
+      "epoch": 1.8157209786569495,
+      "grad_norm": 0.2337885632751051,
+      "learning_rate": 1.7797522515869062e-05,
+      "loss": 0.1076,
+      "step": 3488
+    },
+    {
+      "epoch": 1.8162415408641333,
+      "grad_norm": 0.2281419186886461,
+      "learning_rate": 1.778406931139649e-05,
+      "loss": 0.1054,
+      "step": 3489
+    },
+    {
+      "epoch": 1.816762103071317,
+      "grad_norm": 0.21997153529814273,
+      "learning_rate": 1.777061838605062e-05,
+      "loss": 0.1057,
+      "step": 3490
+    },
+    {
+      "epoch": 1.8172826652785008,
+      "grad_norm": 0.21614117988760198,
+      "learning_rate": 1.7757169744079893e-05,
+      "loss": 0.1016,
+      "step": 3491
+    },
+    {
+      "epoch": 1.8178032274856846,
+      "grad_norm": 0.214012852621221,
+      "learning_rate": 1.774372338973201e-05,
+      "loss": 0.1044,
+      "step": 3492
+    },
+    {
+      "epoch": 1.8183237896928683,
+      "grad_norm": 0.21662067485067615,
+      "learning_rate": 1.7730279327253962e-05,
+      "loss": 0.1017,
+      "step": 3493
+    },
+    {
+      "epoch": 1.8188443519000521,
+      "grad_norm": 0.22564115079401081,
+      "learning_rate": 1.7716837560892e-05,
+      "loss": 0.1052,
+      "step": 3494
+    },
+    {
+      "epoch": 1.8193649141072359,
+      "grad_norm": 0.21323173454794608,
+      "learning_rate": 1.7703398094891673e-05,
+      "loss": 0.0973,
+      "step": 3495
+    },
+    {
+      "epoch": 1.8198854763144197,
+      "grad_norm": 0.21838902892220904,
+      "learning_rate": 1.7689960933497778e-05,
+      "loss": 0.1028,
+      "step": 3496
+    },
+    {
+      "epoch": 1.8204060385216034,
+      "grad_norm": 0.2271905848355706,
+      "learning_rate": 1.76765260809544e-05,
+      "loss": 0.1036,
+      "step": 3497
+    },
+    {
+      "epoch": 1.8209266007287872,
+      "grad_norm": 0.22011061951043087,
+      "learning_rate": 1.7663093541504905e-05,
+      "loss": 0.1019,
+      "step": 3498
+    },
+    {
+      "epoch": 1.821447162935971,
+      "grad_norm": 0.22881874170767702,
+      "learning_rate": 1.7649663319391883e-05,
+      "loss": 0.1073,
+      "step": 3499
+    },
+    {
+      "epoch": 1.8219677251431547,
+      "grad_norm": 0.22620885690592923,
+      "learning_rate": 1.7636235418857245e-05,
+      "loss": 0.1025,
+      "step": 3500
+    },
+    {
+      "epoch": 1.8224882873503385,
+      "grad_norm": 0.23410418806456498,
+      "learning_rate": 1.7622809844142137e-05,
+      "loss": 0.102,
+      "step": 3501
+    },
+    {
+      "epoch": 1.823008849557522,
+      "grad_norm": 0.22101701713784802,
+      "learning_rate": 1.760938659948699e-05,
+      "loss": 0.102,
+      "step": 3502
+    },
+    {
+      "epoch": 1.8235294117647058,
+      "grad_norm": 0.22714340766149158,
+      "learning_rate": 1.7595965689131484e-05,
+      "loss": 0.1001,
+      "step": 3503
+    },
+    {
+      "epoch": 1.8240499739718896,
+      "grad_norm": 0.22057919964249992,
+      "learning_rate": 1.7582547117314563e-05,
+      "loss": 0.102,
+      "step": 3504
+    },
+    {
+      "epoch": 1.8245705361790734,
+      "grad_norm": 0.21511543280496126,
+      "learning_rate": 1.7569130888274446e-05,
+      "loss": 0.1043,
+      "step": 3505
+    },
+    {
+      "epoch": 1.8250910983862572,
+      "grad_norm": 0.21440452554779588,
+      "learning_rate": 1.7555717006248594e-05,
+      "loss": 0.0959,
+      "step": 3506
+    },
+    {
+      "epoch": 1.825611660593441,
+      "grad_norm": 0.22709181091908154,
+      "learning_rate": 1.7542305475473746e-05,
+      "loss": 0.1043,
+      "step": 3507
+    },
+    {
+      "epoch": 1.8261322228006247,
+      "grad_norm": 0.2268531896037747,
+      "learning_rate": 1.7528896300185878e-05,
+      "loss": 0.103,
+      "step": 3508
+    },
+    {
+      "epoch": 1.8266527850078085,
+      "grad_norm": 0.22430552658597233,
+      "learning_rate": 1.751548948462023e-05,
+      "loss": 0.1041,
+      "step": 3509
+    },
+    {
+      "epoch": 1.827173347214992,
+      "grad_norm": 0.22793206081910328,
+      "learning_rate": 1.7502085033011302e-05,
+      "loss": 0.1019,
+      "step": 3510
+    },
+    {
+      "epoch": 1.8276939094221758,
+      "grad_norm": 0.23464781588284941,
+      "learning_rate": 1.748868294959284e-05,
+      "loss": 0.1054,
+      "step": 3511
+    },
+    {
+      "epoch": 1.8282144716293596,
+      "grad_norm": 0.22172768017872468,
+      "learning_rate": 1.7475283238597857e-05,
+      "loss": 0.1017,
+      "step": 3512
+    },
+    {
+      "epoch": 1.8287350338365433,
+      "grad_norm": 0.2518798662916246,
+      "learning_rate": 1.746188590425859e-05,
+      "loss": 0.1055,
+      "step": 3513
+    },
+    {
+      "epoch": 1.8292555960437271,
+      "grad_norm": 0.21708297533262325,
+      "learning_rate": 1.7448490950806552e-05,
+      "loss": 0.1041,
+      "step": 3514
+    },
+    {
+      "epoch": 1.829776158250911,
+      "grad_norm": 0.23188297562996182,
+      "learning_rate": 1.7435098382472486e-05,
+      "loss": 0.1094,
+      "step": 3515
+    },
+    {
+      "epoch": 1.8302967204580947,
+      "grad_norm": 0.21477736177121873,
+      "learning_rate": 1.74217082034864e-05,
+      "loss": 0.1037,
+      "step": 3516
+    },
+    {
+      "epoch": 1.8308172826652784,
+      "grad_norm": 0.2174218899457457,
+      "learning_rate": 1.740832041807752e-05,
+      "loss": 0.1041,
+      "step": 3517
+    },
+    {
+      "epoch": 1.8313378448724622,
+      "grad_norm": 0.22509449830210138,
+      "learning_rate": 1.7394935030474335e-05,
+      "loss": 0.0994,
+      "step": 3518
+    },
+    {
+      "epoch": 1.831858407079646,
+      "grad_norm": 0.2141160924638327,
+      "learning_rate": 1.738155204490458e-05,
+      "loss": 0.0995,
+      "step": 3519
+    },
+    {
+      "epoch": 1.8323789692868298,
+      "grad_norm": 0.23187384577530246,
+      "learning_rate": 1.7368171465595222e-05,
+      "loss": 0.1086,
+      "step": 3520
+    },
+    {
+      "epoch": 1.8328995314940135,
+      "grad_norm": 0.2130230900152308,
+      "learning_rate": 1.735479329677247e-05,
+      "loss": 0.1006,
+      "step": 3521
+    },
+    {
+      "epoch": 1.8334200937011973,
+      "grad_norm": 0.2235989907682875,
+      "learning_rate": 1.7341417542661767e-05,
+      "loss": 0.1055,
+      "step": 3522
+    },
+    {
+      "epoch": 1.833940655908381,
+      "grad_norm": 0.2304050686195293,
+      "learning_rate": 1.732804420748781e-05,
+      "loss": 0.1042,
+      "step": 3523
+    },
+    {
+      "epoch": 1.8344612181155648,
+      "grad_norm": 0.21354769016325825,
+      "learning_rate": 1.731467329547451e-05,
+      "loss": 0.0985,
+      "step": 3524
+    },
+    {
+      "epoch": 1.8349817803227486,
+      "grad_norm": 0.22881459978388433,
+      "learning_rate": 1.7301304810845037e-05,
+      "loss": 0.1009,
+      "step": 3525
+    },
+    {
+      "epoch": 1.8355023425299324,
+      "grad_norm": 0.21108826324764815,
+      "learning_rate": 1.7287938757821765e-05,
+      "loss": 0.1042,
+      "step": 3526
+    },
+    {
+      "epoch": 1.8360229047371162,
+      "grad_norm": 0.23139125405119665,
+      "learning_rate": 1.7274575140626318e-05,
+      "loss": 0.1075,
+      "step": 3527
+    },
+    {
+      "epoch": 1.8365434669443,
+      "grad_norm": 0.2174665102211108,
+      "learning_rate": 1.7261213963479556e-05,
+      "loss": 0.1041,
+      "step": 3528
+    },
+    {
+      "epoch": 1.8370640291514837,
+      "grad_norm": 0.22610247555368754,
+      "learning_rate": 1.724785523060155e-05,
+      "loss": 0.1009,
+      "step": 3529
+    },
+    {
+      "epoch": 1.8375845913586675,
+      "grad_norm": 0.22191495291139376,
+      "learning_rate": 1.7234498946211625e-05,
+      "loss": 0.1066,
+      "step": 3530
+    },
+    {
+      "epoch": 1.8381051535658512,
+      "grad_norm": 0.22121309383263021,
+      "learning_rate": 1.7221145114528297e-05,
+      "loss": 0.1068,
+      "step": 3531
+    },
+    {
+      "epoch": 1.838625715773035,
+      "grad_norm": 0.2133266726862183,
+      "learning_rate": 1.7207793739769352e-05,
+      "loss": 0.1021,
+      "step": 3532
+    },
+    {
+      "epoch": 1.8391462779802188,
+      "grad_norm": 0.218435643614945,
+      "learning_rate": 1.7194444826151752e-05,
+      "loss": 0.1027,
+      "step": 3533
+    },
+    {
+      "epoch": 1.8396668401874023,
+      "grad_norm": 0.21306246655698843,
+      "learning_rate": 1.7181098377891723e-05,
+      "loss": 0.1028,
+      "step": 3534
+    },
+    {
+      "epoch": 1.8401874023945861,
+      "grad_norm": 0.2316381773531802,
+      "learning_rate": 1.7167754399204683e-05,
+      "loss": 0.1087,
+      "step": 3535
+    },
+    {
+      "epoch": 1.8407079646017699,
+      "grad_norm": 0.2324977841120054,
+      "learning_rate": 1.7154412894305283e-05,
+      "loss": 0.1083,
+      "step": 3536
+    },
+    {
+      "epoch": 1.8412285268089537,
+      "grad_norm": 0.21543732870890425,
+      "learning_rate": 1.7141073867407397e-05,
+      "loss": 0.1004,
+      "step": 3537
+    },
+    {
+      "epoch": 1.8417490890161374,
+      "grad_norm": 0.21111111952706546,
+      "learning_rate": 1.71277373227241e-05,
+      "loss": 0.1035,
+      "step": 3538
+    },
+    {
+      "epoch": 1.8422696512233212,
+      "grad_norm": 0.20655633250382696,
+      "learning_rate": 1.7114403264467703e-05,
+      "loss": 0.098,
+      "step": 3539
+    },
+    {
+      "epoch": 1.842790213430505,
+      "grad_norm": 0.22438903873520755,
+      "learning_rate": 1.710107169684972e-05,
+      "loss": 0.1067,
+      "step": 3540
+    },
+    {
+      "epoch": 1.8433107756376887,
+      "grad_norm": 0.2199314935079129,
+      "learning_rate": 1.7087742624080883e-05,
+      "loss": 0.1021,
+      "step": 3541
+    },
+    {
+      "epoch": 1.8438313378448723,
+      "grad_norm": 0.2188004654542698,
+      "learning_rate": 1.7074416050371122e-05,
+      "loss": 0.1063,
+      "step": 3542
+    },
+    {
+      "epoch": 1.844351900052056,
+      "grad_norm": 0.22136277736501705,
+      "learning_rate": 1.7061091979929612e-05,
+      "loss": 0.1041,
+      "step": 3543
+    },
+    {
+      "epoch": 1.8448724622592398,
+      "grad_norm": 0.22383412015373572,
+      "learning_rate": 1.7047770416964688e-05,
+      "loss": 0.1037,
+      "step": 3544
+    },
+    {
+      "epoch": 1.8453930244664236,
+      "grad_norm": 0.22366283880625565,
+      "learning_rate": 1.7034451365683927e-05,
+      "loss": 0.1057,
+      "step": 3545
+    },
+    {
+      "epoch": 1.8459135866736074,
+      "grad_norm": 0.22075942408527013,
+      "learning_rate": 1.702113483029412e-05,
+      "loss": 0.1017,
+      "step": 3546
+    },
+    {
+      "epoch": 1.8464341488807912,
+      "grad_norm": 0.2313958085020113,
+      "learning_rate": 1.700782081500123e-05,
+      "loss": 0.1046,
+      "step": 3547
+    },
+    {
+      "epoch": 1.846954711087975,
+      "grad_norm": 0.22215124468148365,
+      "learning_rate": 1.6994509324010457e-05,
+      "loss": 0.1038,
+      "step": 3548
+    },
+    {
+      "epoch": 1.8474752732951587,
+      "grad_norm": 0.2140104139101645,
+      "learning_rate": 1.6981200361526177e-05,
+      "loss": 0.1016,
+      "step": 3549
+    },
+    {
+      "epoch": 1.8479958355023425,
+      "grad_norm": 0.2199758204376425,
+      "learning_rate": 1.6967893931751988e-05,
+      "loss": 0.0985,
+      "step": 3550
+    },
+    {
+      "epoch": 1.8485163977095262,
+      "grad_norm": 0.21980628404627928,
+      "learning_rate": 1.695459003889068e-05,
+      "loss": 0.1027,
+      "step": 3551
+    },
+    {
+      "epoch": 1.84903695991671,
+      "grad_norm": 0.22964814499456035,
+      "learning_rate": 1.694128868714424e-05,
+      "loss": 0.1008,
+      "step": 3552
+    },
+    {
+      "epoch": 1.8495575221238938,
+      "grad_norm": 0.2194752903061473,
+      "learning_rate": 1.692798988071385e-05,
+      "loss": 0.1043,
+      "step": 3553
+    },
+    {
+      "epoch": 1.8500780843310776,
+      "grad_norm": 0.22695773488044693,
+      "learning_rate": 1.6914693623799894e-05,
+      "loss": 0.1063,
+      "step": 3554
+    },
+    {
+      "epoch": 1.8505986465382613,
+      "grad_norm": 0.22960187036725627,
+      "learning_rate": 1.690139992060195e-05,
+      "loss": 0.1032,
+      "step": 3555
+    },
+    {
+      "epoch": 1.851119208745445,
+      "grad_norm": 0.2174160418036627,
+      "learning_rate": 1.6888108775318785e-05,
+      "loss": 0.1009,
+      "step": 3556
+    },
+    {
+      "epoch": 1.8516397709526289,
+      "grad_norm": 0.2234493817495722,
+      "learning_rate": 1.6874820192148365e-05,
+      "loss": 0.104,
+      "step": 3557
+    },
+    {
+      "epoch": 1.8521603331598127,
+      "grad_norm": 0.220450115790806,
+      "learning_rate": 1.686153417528784e-05,
+      "loss": 0.1016,
+      "step": 3558
+    },
+    {
+      "epoch": 1.8526808953669964,
+      "grad_norm": 0.2274125870596465,
+      "learning_rate": 1.684825072893356e-05,
+      "loss": 0.1098,
+      "step": 3559
+    },
+    {
+      "epoch": 1.8532014575741802,
+      "grad_norm": 0.22368136276174644,
+      "learning_rate": 1.6834969857281042e-05,
+      "loss": 0.1048,
+      "step": 3560
+    },
+    {
+      "epoch": 1.853722019781364,
+      "grad_norm": 0.22278441077025057,
+      "learning_rate": 1.682169156452502e-05,
+      "loss": 0.1021,
+      "step": 3561
+    },
+    {
+      "epoch": 1.8542425819885477,
+      "grad_norm": 0.2144032315295473,
+      "learning_rate": 1.6808415854859384e-05,
+      "loss": 0.1009,
+      "step": 3562
+    },
+    {
+      "epoch": 1.8547631441957315,
+      "grad_norm": 0.21883351924688985,
+      "learning_rate": 1.679514273247722e-05,
+      "loss": 0.1046,
+      "step": 3563
+    },
+    {
+      "epoch": 1.8552837064029153,
+      "grad_norm": 0.21804994979018316,
+      "learning_rate": 1.678187220157081e-05,
+      "loss": 0.105,
+      "step": 3564
+    },
+    {
+      "epoch": 1.855804268610099,
+      "grad_norm": 0.23179533965357424,
+      "learning_rate": 1.6768604266331585e-05,
+      "loss": 0.1049,
+      "step": 3565
+    },
+    {
+      "epoch": 1.8563248308172826,
+      "grad_norm": 0.21771973013714693,
+      "learning_rate": 1.675533893095019e-05,
+      "loss": 0.1042,
+      "step": 3566
+    },
+    {
+      "epoch": 1.8568453930244664,
+      "grad_norm": 0.21510845344937907,
+      "learning_rate": 1.6742076199616448e-05,
+      "loss": 0.1035,
+      "step": 3567
+    },
+    {
+      "epoch": 1.8573659552316502,
+      "grad_norm": 0.22503185695899153,
+      "learning_rate": 1.6728816076519322e-05,
+      "loss": 0.1045,
+      "step": 3568
+    },
+    {
+      "epoch": 1.857886517438834,
+      "grad_norm": 0.21278464572186798,
+      "learning_rate": 1.671555856584699e-05,
+      "loss": 0.1043,
+      "step": 3569
+    },
+    {
+      "epoch": 1.8584070796460177,
+      "grad_norm": 0.21442678992556535,
+      "learning_rate": 1.6702303671786797e-05,
+      "loss": 0.1019,
+      "step": 3570
+    },
+    {
+      "epoch": 1.8589276418532015,
+      "grad_norm": 0.21995013323000376,
+      "learning_rate": 1.6689051398525233e-05,
+      "loss": 0.0996,
+      "step": 3571
+    },
+    {
+      "epoch": 1.8594482040603852,
+      "grad_norm": 0.2166463003463736,
+      "learning_rate": 1.6675801750247998e-05,
+      "loss": 0.1001,
+      "step": 3572
+    },
+    {
+      "epoch": 1.859968766267569,
+      "grad_norm": 0.21168851210455933,
+      "learning_rate": 1.6662554731139944e-05,
+      "loss": 0.0986,
+      "step": 3573
+    },
+    {
+      "epoch": 1.8604893284747526,
+      "grad_norm": 0.2187117583742439,
+      "learning_rate": 1.66493103453851e-05,
+      "loss": 0.099,
+      "step": 3574
+    },
+    {
+      "epoch": 1.8610098906819363,
+      "grad_norm": 0.2323426842133383,
+      "learning_rate": 1.6636068597166655e-05,
+      "loss": 0.1083,
+      "step": 3575
+    },
+    {
+      "epoch": 1.8615304528891201,
+      "grad_norm": 0.226528190668166,
+      "learning_rate": 1.6622829490666974e-05,
+      "loss": 0.1015,
+      "step": 3576
+    },
+    {
+      "epoch": 1.8620510150963039,
+      "grad_norm": 0.2099990904532976,
+      "learning_rate": 1.6609593030067574e-05,
+      "loss": 0.102,
+      "step": 3577
+    },
+    {
+      "epoch": 1.8625715773034877,
+      "grad_norm": 0.21653151598199424,
+      "learning_rate": 1.6596359219549158e-05,
+      "loss": 0.102,
+      "step": 3578
+    },
+    {
+      "epoch": 1.8630921395106714,
+      "grad_norm": 0.23703398831513672,
+      "learning_rate": 1.6583128063291576e-05,
+      "loss": 0.0998,
+      "step": 3579
+    },
+    {
+      "epoch": 1.8636127017178552,
+      "grad_norm": 0.2204375202877476,
+      "learning_rate": 1.6569899565473828e-05,
+      "loss": 0.1021,
+      "step": 3580
+    },
+    {
+      "epoch": 1.864133263925039,
+      "grad_norm": 0.21805115310094073,
+      "learning_rate": 1.6556673730274107e-05,
+      "loss": 0.1049,
+      "step": 3581
+    },
+    {
+      "epoch": 1.8646538261322227,
+      "grad_norm": 0.22105359053989074,
+      "learning_rate": 1.6543450561869732e-05,
+      "loss": 0.1069,
+      "step": 3582
+    },
+    {
+      "epoch": 1.8651743883394065,
+      "grad_norm": 0.226961101870007,
+      "learning_rate": 1.6530230064437213e-05,
+      "loss": 0.1039,
+      "step": 3583
+    },
+    {
+      "epoch": 1.8656949505465903,
+      "grad_norm": 0.21715026468515009,
+      "learning_rate": 1.651701224215218e-05,
+      "loss": 0.1058,
+      "step": 3584
+    },
+    {
+      "epoch": 1.866215512753774,
+      "grad_norm": 0.2185791885590469,
+      "learning_rate": 1.6503797099189453e-05,
+      "loss": 0.1033,
+      "step": 3585
+    },
+    {
+      "epoch": 1.8667360749609578,
+      "grad_norm": 0.21828087490161857,
+      "learning_rate": 1.6490584639722976e-05,
+      "loss": 0.1044,
+      "step": 3586
+    },
+    {
+      "epoch": 1.8672566371681416,
+      "grad_norm": 0.2258399116146915,
+      "learning_rate": 1.6477374867925867e-05,
+      "loss": 0.1046,
+      "step": 3587
+    },
+    {
+      "epoch": 1.8677771993753254,
+      "grad_norm": 0.21880046178089685,
+      "learning_rate": 1.646416778797039e-05,
+      "loss": 0.1029,
+      "step": 3588
+    },
+    {
+      "epoch": 1.8682977615825092,
+      "grad_norm": 0.22303778060525425,
+      "learning_rate": 1.645096340402794e-05,
+      "loss": 0.1014,
+      "step": 3589
+    },
+    {
+      "epoch": 1.868818323789693,
+      "grad_norm": 0.22130295830879798,
+      "learning_rate": 1.6437761720269087e-05,
+      "loss": 0.1021,
+      "step": 3590
+    },
+    {
+      "epoch": 1.8693388859968767,
+      "grad_norm": 0.2164153839615916,
+      "learning_rate": 1.642456274086353e-05,
+      "loss": 0.1033,
+      "step": 3591
+    },
+    {
+      "epoch": 1.8698594482040605,
+      "grad_norm": 0.22284825275166165,
+      "learning_rate": 1.6411366469980134e-05,
+      "loss": 0.1024,
+      "step": 3592
+    },
+    {
+      "epoch": 1.8703800104112442,
+      "grad_norm": 0.2110508578950555,
+      "learning_rate": 1.6398172911786883e-05,
+      "loss": 0.1024,
+      "step": 3593
+    },
+    {
+      "epoch": 1.870900572618428,
+      "grad_norm": 0.20736136520524462,
+      "learning_rate": 1.6384982070450922e-05,
+      "loss": 0.0949,
+      "step": 3594
+    },
+    {
+      "epoch": 1.8714211348256118,
+      "grad_norm": 0.21550440293843662,
+      "learning_rate": 1.637179395013853e-05,
+      "loss": 0.1,
+      "step": 3595
+    },
+    {
+      "epoch": 1.8719416970327956,
+      "grad_norm": 0.22901002637779894,
+      "learning_rate": 1.6358608555015135e-05,
+      "loss": 0.1085,
+      "step": 3596
+    },
+    {
+      "epoch": 1.8724622592399793,
+      "grad_norm": 0.21708597501510324,
+      "learning_rate": 1.6345425889245298e-05,
+      "loss": 0.0956,
+      "step": 3597
+    },
+    {
+      "epoch": 1.8729828214471629,
+      "grad_norm": 0.21299348538165125,
+      "learning_rate": 1.6332245956992703e-05,
+      "loss": 0.0975,
+      "step": 3598
+    },
+    {
+      "epoch": 1.8735033836543467,
+      "grad_norm": 0.21235644848347995,
+      "learning_rate": 1.6319068762420204e-05,
+      "loss": 0.1009,
+      "step": 3599
+    },
+    {
+      "epoch": 1.8740239458615304,
+      "grad_norm": 0.21460664666278179,
+      "learning_rate": 1.6305894309689763e-05,
+      "loss": 0.1024,
+      "step": 3600
+    },
+    {
+      "epoch": 1.8745445080687142,
+      "grad_norm": 0.22828274571427395,
+      "learning_rate": 1.629272260296249e-05,
+      "loss": 0.1071,
+      "step": 3601
+    },
+    {
+      "epoch": 1.875065070275898,
+      "grad_norm": 0.22326950946792548,
+      "learning_rate": 1.6279553646398615e-05,
+      "loss": 0.1061,
+      "step": 3602
+    },
+    {
+      "epoch": 1.8755856324830817,
+      "grad_norm": 0.21887665058394934,
+      "learning_rate": 1.626638744415752e-05,
+      "loss": 0.1052,
+      "step": 3603
+    },
+    {
+      "epoch": 1.8761061946902655,
+      "grad_norm": 0.22556670911635793,
+      "learning_rate": 1.625322400039769e-05,
+      "loss": 0.1034,
+      "step": 3604
+    },
+    {
+      "epoch": 1.8766267568974493,
+      "grad_norm": 0.22477039664046272,
+      "learning_rate": 1.6240063319276767e-05,
+      "loss": 0.1066,
+      "step": 3605
+    },
+    {
+      "epoch": 1.8771473191046328,
+      "grad_norm": 0.208448037377274,
+      "learning_rate": 1.6226905404951503e-05,
+      "loss": 0.0964,
+      "step": 3606
+    },
+    {
+      "epoch": 1.8776678813118166,
+      "grad_norm": 0.20964154673856936,
+      "learning_rate": 1.621375026157777e-05,
+      "loss": 0.0974,
+      "step": 3607
+    },
+    {
+      "epoch": 1.8781884435190004,
+      "grad_norm": 0.21982252986036593,
+      "learning_rate": 1.6200597893310586e-05,
+      "loss": 0.1041,
+      "step": 3608
+    },
+    {
+      "epoch": 1.8787090057261842,
+      "grad_norm": 0.21630070655518385,
+      "learning_rate": 1.618744830430407e-05,
+      "loss": 0.1028,
+      "step": 3609
+    },
+    {
+      "epoch": 1.879229567933368,
+      "grad_norm": 0.22686308982838688,
+      "learning_rate": 1.6174301498711486e-05,
+      "loss": 0.1037,
+      "step": 3610
+    },
+    {
+      "epoch": 1.8797501301405517,
+      "grad_norm": 0.22043831684254314,
+      "learning_rate": 1.6161157480685197e-05,
+      "loss": 0.103,
+      "step": 3611
+    },
+    {
+      "epoch": 1.8802706923477355,
+      "grad_norm": 0.2178337556402931,
+      "learning_rate": 1.6148016254376702e-05,
+      "loss": 0.0978,
+      "step": 3612
+    },
+    {
+      "epoch": 1.8807912545549192,
+      "grad_norm": 0.22627246524746905,
+      "learning_rate": 1.613487782393661e-05,
+      "loss": 0.1046,
+      "step": 3613
+    },
+    {
+      "epoch": 1.881311816762103,
+      "grad_norm": 0.22358601640561931,
+      "learning_rate": 1.6121742193514648e-05,
+      "loss": 0.0986,
+      "step": 3614
+    },
+    {
+      "epoch": 1.8818323789692868,
+      "grad_norm": 0.22423924786442362,
+      "learning_rate": 1.610860936725967e-05,
+      "loss": 0.1001,
+      "step": 3615
+    },
+    {
+      "epoch": 1.8823529411764706,
+      "grad_norm": 0.2187539346723351,
+      "learning_rate": 1.6095479349319607e-05,
+      "loss": 0.106,
+      "step": 3616
+    },
+    {
+      "epoch": 1.8828735033836543,
+      "grad_norm": 0.22405774990004992,
+      "learning_rate": 1.6082352143841555e-05,
+      "loss": 0.0999,
+      "step": 3617
+    },
+    {
+      "epoch": 1.883394065590838,
+      "grad_norm": 0.2268274693952615,
+      "learning_rate": 1.6069227754971683e-05,
+      "loss": 0.1007,
+      "step": 3618
+    },
+    {
+      "epoch": 1.8839146277980219,
+      "grad_norm": 0.2387625325422801,
+      "learning_rate": 1.6056106186855292e-05,
+      "loss": 0.1072,
+      "step": 3619
+    },
+    {
+      "epoch": 1.8844351900052057,
+      "grad_norm": 0.21912658060501727,
+      "learning_rate": 1.6042987443636775e-05,
+      "loss": 0.1043,
+      "step": 3620
+    },
+    {
+      "epoch": 1.8849557522123894,
+      "grad_norm": 0.22608005928160363,
+      "learning_rate": 1.6029871529459656e-05,
+      "loss": 0.1034,
+      "step": 3621
+    },
+    {
+      "epoch": 1.8854763144195732,
+      "grad_norm": 0.21470577776047628,
+      "learning_rate": 1.601675844846653e-05,
+      "loss": 0.0991,
+      "step": 3622
+    },
+    {
+      "epoch": 1.885996876626757,
+      "grad_norm": 0.21496261618823018,
+      "learning_rate": 1.600364820479914e-05,
+      "loss": 0.0995,
+      "step": 3623
+    },
+    {
+      "epoch": 1.8865174388339407,
+      "grad_norm": 0.20783054254928857,
+      "learning_rate": 1.5990540802598302e-05,
+      "loss": 0.0999,
+      "step": 3624
+    },
+    {
+      "epoch": 1.8870380010411245,
+      "grad_norm": 0.20378345368619807,
+      "learning_rate": 1.5977436246003937e-05,
+      "loss": 0.0936,
+      "step": 3625
+    },
+    {
+      "epoch": 1.8875585632483083,
+      "grad_norm": 0.2186409812177155,
+      "learning_rate": 1.5964334539155084e-05,
+      "loss": 0.1048,
+      "step": 3626
+    },
+    {
+      "epoch": 1.888079125455492,
+      "grad_norm": 0.22599765981965825,
+      "learning_rate": 1.5951235686189857e-05,
+      "loss": 0.1043,
+      "step": 3627
+    },
+    {
+      "epoch": 1.8885996876626758,
+      "grad_norm": 0.22083477562213608,
+      "learning_rate": 1.5938139691245505e-05,
+      "loss": 0.1018,
+      "step": 3628
+    },
+    {
+      "epoch": 1.8891202498698596,
+      "grad_norm": 0.2170688955855864,
+      "learning_rate": 1.5925046558458333e-05,
+      "loss": 0.1021,
+      "step": 3629
+    },
+    {
+      "epoch": 1.8896408120770432,
+      "grad_norm": 0.2148159717859146,
+      "learning_rate": 1.5911956291963775e-05,
+      "loss": 0.0983,
+      "step": 3630
+    },
+    {
+      "epoch": 1.890161374284227,
+      "grad_norm": 0.20873764640196554,
+      "learning_rate": 1.5898868895896334e-05,
+      "loss": 0.0974,
+      "step": 3631
+    },
+    {
+      "epoch": 1.8906819364914107,
+      "grad_norm": 0.3148651473791158,
+      "learning_rate": 1.5885784374389632e-05,
+      "loss": 0.1065,
+      "step": 3632
+    },
+    {
+      "epoch": 1.8912024986985945,
+      "grad_norm": 0.22482430722857094,
+      "learning_rate": 1.5872702731576373e-05,
+      "loss": 0.1042,
+      "step": 3633
+    },
+    {
+      "epoch": 1.8917230609057782,
+      "grad_norm": 0.22637467340246845,
+      "learning_rate": 1.585962397158833e-05,
+      "loss": 0.1087,
+      "step": 3634
+    },
+    {
+      "epoch": 1.892243623112962,
+      "grad_norm": 0.22354140941775033,
+      "learning_rate": 1.584654809855639e-05,
+      "loss": 0.0974,
+      "step": 3635
+    },
+    {
+      "epoch": 1.8927641853201458,
+      "grad_norm": 0.21626315203073032,
+      "learning_rate": 1.5833475116610523e-05,
+      "loss": 0.1005,
+      "step": 3636
+    },
+    {
+      "epoch": 1.8932847475273296,
+      "grad_norm": 0.2113336278345908,
+      "learning_rate": 1.582040502987979e-05,
+      "loss": 0.1041,
+      "step": 3637
+    },
+    {
+      "epoch": 1.893805309734513,
+      "grad_norm": 0.22087048570487774,
+      "learning_rate": 1.5807337842492337e-05,
+      "loss": 0.1036,
+      "step": 3638
+    },
+    {
+      "epoch": 1.8943258719416969,
+      "grad_norm": 0.21572578777284565,
+      "learning_rate": 1.5794273558575374e-05,
+      "loss": 0.0984,
+      "step": 3639
+    },
+    {
+      "epoch": 1.8948464341488807,
+      "grad_norm": 0.22769892060924338,
+      "learning_rate": 1.5781212182255226e-05,
+      "loss": 0.1122,
+      "step": 3640
+    },
+    {
+      "epoch": 1.8953669963560644,
+      "grad_norm": 0.213430358027192,
+      "learning_rate": 1.5768153717657268e-05,
+      "loss": 0.0988,
+      "step": 3641
+    },
+    {
+      "epoch": 1.8958875585632482,
+      "grad_norm": 0.2622995076619564,
+      "learning_rate": 1.5755098168905992e-05,
+      "loss": 0.098,
+      "step": 3642
+    },
+    {
+      "epoch": 1.896408120770432,
+      "grad_norm": 0.2307094720665387,
+      "learning_rate": 1.574204554012493e-05,
+      "loss": 0.0983,
+      "step": 3643
+    },
+    {
+      "epoch": 1.8969286829776157,
+      "grad_norm": 0.22053805471841584,
+      "learning_rate": 1.572899583543671e-05,
+      "loss": 0.1017,
+      "step": 3644
+    },
+    {
+      "epoch": 1.8974492451847995,
+      "grad_norm": 0.21911020706637074,
+      "learning_rate": 1.571594905896304e-05,
+      "loss": 0.1048,
+      "step": 3645
+    },
+    {
+      "epoch": 1.8979698073919833,
+      "grad_norm": 0.21973155556064952,
+      "learning_rate": 1.5702905214824705e-05,
+      "loss": 0.1047,
+      "step": 3646
+    },
+    {
+      "epoch": 1.898490369599167,
+      "grad_norm": 0.2446184787139059,
+      "learning_rate": 1.5689864307141548e-05,
+      "loss": 0.1003,
+      "step": 3647
+    },
+    {
+      "epoch": 1.8990109318063508,
+      "grad_norm": 0.21721103382874216,
+      "learning_rate": 1.56768263400325e-05,
+      "loss": 0.0934,
+      "step": 3648
+    },
+    {
+      "epoch": 1.8995314940135346,
+      "grad_norm": 0.22688257718189353,
+      "learning_rate": 1.566379131761556e-05,
+      "loss": 0.104,
+      "step": 3649
+    },
+    {
+      "epoch": 1.9000520562207184,
+      "grad_norm": 0.21482237576043078,
+      "learning_rate": 1.5650759244007783e-05,
+      "loss": 0.0964,
+      "step": 3650
+    },
+    {
+      "epoch": 1.9005726184279021,
+      "grad_norm": 0.22722159279382412,
+      "learning_rate": 1.5637730123325327e-05,
+      "loss": 0.104,
+      "step": 3651
+    },
+    {
+      "epoch": 1.901093180635086,
+      "grad_norm": 0.22648407542435803,
+      "learning_rate": 1.562470395968338e-05,
+      "loss": 0.1043,
+      "step": 3652
+    },
+    {
+      "epoch": 1.9016137428422697,
+      "grad_norm": 0.21123512550603266,
+      "learning_rate": 1.56116807571962e-05,
+      "loss": 0.0991,
+      "step": 3653
+    },
+    {
+      "epoch": 1.9021343050494535,
+      "grad_norm": 0.22587891194139853,
+      "learning_rate": 1.559866051997714e-05,
+      "loss": 0.1019,
+      "step": 3654
+    },
+    {
+      "epoch": 1.9026548672566372,
+      "grad_norm": 0.22485896129735808,
+      "learning_rate": 1.5585643252138577e-05,
+      "loss": 0.0991,
+      "step": 3655
+    },
+    {
+      "epoch": 1.903175429463821,
+      "grad_norm": 0.21654611342402638,
+      "learning_rate": 1.557262895779199e-05,
+      "loss": 0.1018,
+      "step": 3656
+    },
+    {
+      "epoch": 1.9036959916710048,
+      "grad_norm": 0.22992554106238536,
+      "learning_rate": 1.5559617641047886e-05,
+      "loss": 0.1052,
+      "step": 3657
+    },
+    {
+      "epoch": 1.9042165538781886,
+      "grad_norm": 0.2261900928255639,
+      "learning_rate": 1.5546609306015856e-05,
+      "loss": 0.1038,
+      "step": 3658
+    },
+    {
+      "epoch": 1.9047371160853723,
+      "grad_norm": 0.22144384444707318,
+      "learning_rate": 1.5533603956804522e-05,
+      "loss": 0.1015,
+      "step": 3659
+    },
+    {
+      "epoch": 1.905257678292556,
+      "grad_norm": 0.21449867047073035,
+      "learning_rate": 1.5520601597521596e-05,
+      "loss": 0.0986,
+      "step": 3660
+    },
+    {
+      "epoch": 1.9057782404997399,
+      "grad_norm": 0.22083165166943855,
+      "learning_rate": 1.5507602232273814e-05,
+      "loss": 0.0978,
+      "step": 3661
+    },
+    {
+      "epoch": 1.9062988027069234,
+      "grad_norm": 0.21908132571042455,
+      "learning_rate": 1.5494605865166983e-05,
+      "loss": 0.1009,
+      "step": 3662
+    },
+    {
+      "epoch": 1.9068193649141072,
+      "grad_norm": 0.2256794889232402,
+      "learning_rate": 1.5481612500305964e-05,
+      "loss": 0.1064,
+      "step": 3663
+    },
+    {
+      "epoch": 1.907339927121291,
+      "grad_norm": 0.20932873751015288,
+      "learning_rate": 1.5468622141794664e-05,
+      "loss": 0.1019,
+      "step": 3664
+    },
+    {
+      "epoch": 1.9078604893284747,
+      "grad_norm": 0.22557308680633772,
+      "learning_rate": 1.5455634793736046e-05,
+      "loss": 0.1051,
+      "step": 3665
+    },
+    {
+      "epoch": 1.9083810515356585,
+      "grad_norm": 0.21717795714503,
+      "learning_rate": 1.544265046023211e-05,
+      "loss": 0.1043,
+      "step": 3666
+    },
+    {
+      "epoch": 1.9089016137428423,
+      "grad_norm": 0.21382166007000752,
+      "learning_rate": 1.542966914538393e-05,
+      "loss": 0.1003,
+      "step": 3667
+    },
+    {
+      "epoch": 1.909422175950026,
+      "grad_norm": 0.21047577575173457,
+      "learning_rate": 1.541669085329159e-05,
+      "loss": 0.1021,
+      "step": 3668
+    },
+    {
+      "epoch": 1.9099427381572098,
+      "grad_norm": 0.21885604669207556,
+      "learning_rate": 1.540371558805425e-05,
+      "loss": 0.1045,
+      "step": 3669
+    },
+    {
+      "epoch": 1.9104633003643934,
+      "grad_norm": 0.2263376773671937,
+      "learning_rate": 1.539074335377011e-05,
+      "loss": 0.1079,
+      "step": 3670
+    },
+    {
+      "epoch": 1.9109838625715772,
+      "grad_norm": 0.21167752773460963,
+      "learning_rate": 1.537777415453638e-05,
+      "loss": 0.0996,
+      "step": 3671
+    },
+    {
+      "epoch": 1.911504424778761,
+      "grad_norm": 0.21780236850803966,
+      "learning_rate": 1.536480799444936e-05,
+      "loss": 0.104,
+      "step": 3672
+    },
+    {
+      "epoch": 1.9120249869859447,
+      "grad_norm": 0.2202957087255931,
+      "learning_rate": 1.5351844877604353e-05,
+      "loss": 0.101,
+      "step": 3673
+    },
+    {
+      "epoch": 1.9125455491931285,
+      "grad_norm": 0.2215233885356788,
+      "learning_rate": 1.5338884808095726e-05,
+      "loss": 0.1031,
+      "step": 3674
+    },
+    {
+      "epoch": 1.9130661114003122,
+      "grad_norm": 0.221036149093325,
+      "learning_rate": 1.5325927790016858e-05,
+      "loss": 0.1009,
+      "step": 3675
+    },
+    {
+      "epoch": 1.913586673607496,
+      "grad_norm": 0.21072634910142585,
+      "learning_rate": 1.5312973827460194e-05,
+      "loss": 0.1022,
+      "step": 3676
+    },
+    {
+      "epoch": 1.9141072358146798,
+      "grad_norm": 0.21579345962748916,
+      "learning_rate": 1.5300022924517186e-05,
+      "loss": 0.1014,
+      "step": 3677
+    },
+    {
+      "epoch": 1.9146277980218636,
+      "grad_norm": 0.22972340517503928,
+      "learning_rate": 1.528707508527834e-05,
+      "loss": 0.1066,
+      "step": 3678
+    },
+    {
+      "epoch": 1.9151483602290473,
+      "grad_norm": 0.22399399495289485,
+      "learning_rate": 1.527413031383319e-05,
+      "loss": 0.1062,
+      "step": 3679
+    },
+    {
+      "epoch": 1.915668922436231,
+      "grad_norm": 0.22928343502766305,
+      "learning_rate": 1.5261188614270278e-05,
+      "loss": 0.1044,
+      "step": 3680
+    },
+    {
+      "epoch": 1.9161894846434149,
+      "grad_norm": 0.20874236255449022,
+      "learning_rate": 1.5248249990677212e-05,
+      "loss": 0.1011,
+      "step": 3681
+    },
+    {
+      "epoch": 1.9167100468505986,
+      "grad_norm": 0.2238709567201973,
+      "learning_rate": 1.5235314447140603e-05,
+      "loss": 0.1049,
+      "step": 3682
+    },
+    {
+      "epoch": 1.9172306090577824,
+      "grad_norm": 0.22429081990026067,
+      "learning_rate": 1.5222381987746104e-05,
+      "loss": 0.1058,
+      "step": 3683
+    },
+    {
+      "epoch": 1.9177511712649662,
+      "grad_norm": 0.21422872549018027,
+      "learning_rate": 1.5209452616578379e-05,
+      "loss": 0.0997,
+      "step": 3684
+    },
+    {
+      "epoch": 1.91827173347215,
+      "grad_norm": 0.2133382379880559,
+      "learning_rate": 1.5196526337721137e-05,
+      "loss": 0.1031,
+      "step": 3685
+    },
+    {
+      "epoch": 1.9187922956793337,
+      "grad_norm": 0.2191123247317634,
+      "learning_rate": 1.5183603155257087e-05,
+      "loss": 0.1003,
+      "step": 3686
+    },
+    {
+      "epoch": 1.9193128578865175,
+      "grad_norm": 0.2203372017928507,
+      "learning_rate": 1.517068307326798e-05,
+      "loss": 0.1008,
+      "step": 3687
+    },
+    {
+      "epoch": 1.9198334200937013,
+      "grad_norm": 0.22009662025394322,
+      "learning_rate": 1.5157766095834581e-05,
+      "loss": 0.1015,
+      "step": 3688
+    },
+    {
+      "epoch": 1.920353982300885,
+      "grad_norm": 0.21359112700683444,
+      "learning_rate": 1.5144852227036658e-05,
+      "loss": 0.1021,
+      "step": 3689
+    },
+    {
+      "epoch": 1.9208745445080688,
+      "grad_norm": 0.2333651788145784,
+      "learning_rate": 1.5131941470953026e-05,
+      "loss": 0.1082,
+      "step": 3690
+    },
+    {
+      "epoch": 1.9213951067152526,
+      "grad_norm": 0.2199424676535258,
+      "learning_rate": 1.5119033831661489e-05,
+      "loss": 0.0969,
+      "step": 3691
+    },
+    {
+      "epoch": 1.9219156689224364,
+      "grad_norm": 0.21869088466324416,
+      "learning_rate": 1.5106129313238898e-05,
+      "loss": 0.1047,
+      "step": 3692
+    },
+    {
+      "epoch": 1.9224362311296201,
+      "grad_norm": 0.2257137266532471,
+      "learning_rate": 1.5093227919761082e-05,
+      "loss": 0.1059,
+      "step": 3693
+    },
+    {
+      "epoch": 1.9229567933368037,
+      "grad_norm": 0.2042823698266939,
+      "learning_rate": 1.5080329655302916e-05,
+      "loss": 0.0964,
+      "step": 3694
+    },
+    {
+      "epoch": 1.9234773555439875,
+      "grad_norm": 0.21289243041563377,
+      "learning_rate": 1.5067434523938263e-05,
+      "loss": 0.0969,
+      "step": 3695
+    },
+    {
+      "epoch": 1.9239979177511712,
+      "grad_norm": 0.2157354773694145,
+      "learning_rate": 1.5054542529740009e-05,
+      "loss": 0.1054,
+      "step": 3696
+    },
+    {
+      "epoch": 1.924518479958355,
+      "grad_norm": 0.20594700902414997,
+      "learning_rate": 1.504165367678006e-05,
+      "loss": 0.0991,
+      "step": 3697
+    },
+    {
+      "epoch": 1.9250390421655388,
+      "grad_norm": 0.2251525996090509,
+      "learning_rate": 1.5028767969129288e-05,
+      "loss": 0.0977,
+      "step": 3698
+    },
+    {
+      "epoch": 1.9255596043727226,
+      "grad_norm": 0.21960241687587326,
+      "learning_rate": 1.5015885410857616e-05,
+      "loss": 0.0994,
+      "step": 3699
+    },
+    {
+      "epoch": 1.9260801665799063,
+      "grad_norm": 0.20915888582744993,
+      "learning_rate": 1.5003006006033948e-05,
+      "loss": 0.0996,
+      "step": 3700
+    },
+    {
+      "epoch": 1.92660072878709,
+      "grad_norm": 0.2225064521826276,
+      "learning_rate": 1.4990129758726203e-05,
+      "loss": 0.1028,
+      "step": 3701
+    },
+    {
+      "epoch": 1.9271212909942737,
+      "grad_norm": 0.22458058269108447,
+      "learning_rate": 1.4977256673001305e-05,
+      "loss": 0.1027,
+      "step": 3702
+    },
+    {
+      "epoch": 1.9276418532014574,
+      "grad_norm": 0.21114252037998082,
+      "learning_rate": 1.4964386752925163e-05,
+      "loss": 0.0958,
+      "step": 3703
+    },
+    {
+      "epoch": 1.9281624154086412,
+      "grad_norm": 0.21678013606532165,
+      "learning_rate": 1.4951520002562705e-05,
+      "loss": 0.0992,
+      "step": 3704
+    },
+    {
+      "epoch": 1.928682977615825,
+      "grad_norm": 0.20924637264966345,
+      "learning_rate": 1.4938656425977842e-05,
+      "loss": 0.0971,
+      "step": 3705
+    },
+    {
+      "epoch": 1.9292035398230087,
+      "grad_norm": 0.21121389074080546,
+      "learning_rate": 1.4925796027233505e-05,
+      "loss": 0.1028,
+      "step": 3706
+    },
+    {
+      "epoch": 1.9297241020301925,
+      "grad_norm": 0.2188455100447437,
+      "learning_rate": 1.4912938810391591e-05,
+      "loss": 0.1019,
+      "step": 3707
+    },
+    {
+      "epoch": 1.9302446642373763,
+      "grad_norm": 0.22810063283033222,
+      "learning_rate": 1.4900084779513004e-05,
+      "loss": 0.1034,
+      "step": 3708
+    },
+    {
+      "epoch": 1.93076522644456,
+      "grad_norm": 0.2144538872962715,
+      "learning_rate": 1.488723393865766e-05,
+      "loss": 0.1016,
+      "step": 3709
+    },
+    {
+      "epoch": 1.9312857886517438,
+      "grad_norm": 0.20918446229699433,
+      "learning_rate": 1.487438629188444e-05,
+      "loss": 0.0993,
+      "step": 3710
+    },
+    {
+      "epoch": 1.9318063508589276,
+      "grad_norm": 0.23023524817864033,
+      "learning_rate": 1.4861541843251242e-05,
+      "loss": 0.1019,
+      "step": 3711
+    },
+    {
+      "epoch": 1.9323269130661114,
+      "grad_norm": 0.22076188050168008,
+      "learning_rate": 1.4848700596814926e-05,
+      "loss": 0.101,
+      "step": 3712
+    },
+    {
+      "epoch": 1.9328474752732951,
+      "grad_norm": 0.21201820939269886,
+      "learning_rate": 1.483586255663137e-05,
+      "loss": 0.1034,
+      "step": 3713
+    },
+    {
+      "epoch": 1.933368037480479,
+      "grad_norm": 0.2154132891161572,
+      "learning_rate": 1.482302772675541e-05,
+      "loss": 0.0982,
+      "step": 3714
+    },
+    {
+      "epoch": 1.9338885996876627,
+      "grad_norm": 0.22979358381751835,
+      "learning_rate": 1.4810196111240898e-05,
+      "loss": 0.105,
+      "step": 3715
+    },
+    {
+      "epoch": 1.9344091618948465,
+      "grad_norm": 0.2170365978314031,
+      "learning_rate": 1.4797367714140642e-05,
+      "loss": 0.1036,
+      "step": 3716
+    },
+    {
+      "epoch": 1.9349297241020302,
+      "grad_norm": 0.21104199550912994,
+      "learning_rate": 1.4784542539506447e-05,
+      "loss": 0.0989,
+      "step": 3717
+    },
+    {
+      "epoch": 1.935450286309214,
+      "grad_norm": 0.21704690446239305,
+      "learning_rate": 1.477172059138911e-05,
+      "loss": 0.0988,
+      "step": 3718
+    },
+    {
+      "epoch": 1.9359708485163978,
+      "grad_norm": 0.2216423717123099,
+      "learning_rate": 1.4758901873838387e-05,
+      "loss": 0.1008,
+      "step": 3719
+    },
+    {
+      "epoch": 1.9364914107235816,
+      "grad_norm": 0.2224915756445421,
+      "learning_rate": 1.4746086390903041e-05,
+      "loss": 0.1028,
+      "step": 3720
+    },
+    {
+      "epoch": 1.9370119729307653,
+      "grad_norm": 0.22165523853861174,
+      "learning_rate": 1.4733274146630782e-05,
+      "loss": 0.1027,
+      "step": 3721
+    },
+    {
+      "epoch": 1.937532535137949,
+      "grad_norm": 0.2104242110992599,
+      "learning_rate": 1.472046514506832e-05,
+      "loss": 0.0999,
+      "step": 3722
+    },
+    {
+      "epoch": 1.9380530973451329,
+      "grad_norm": 0.2183583161427088,
+      "learning_rate": 1.4707659390261336e-05,
+      "loss": 0.1008,
+      "step": 3723
+    },
+    {
+      "epoch": 1.9385736595523166,
+      "grad_norm": 0.2205753671797551,
+      "learning_rate": 1.4694856886254484e-05,
+      "loss": 0.0993,
+      "step": 3724
+    },
+    {
+      "epoch": 1.9390942217595004,
+      "grad_norm": 0.21265574296368048,
+      "learning_rate": 1.4682057637091386e-05,
+      "loss": 0.099,
+      "step": 3725
+    },
+    {
+      "epoch": 1.939614783966684,
+      "grad_norm": 0.21653176063173926,
+      "learning_rate": 1.4669261646814637e-05,
+      "loss": 0.0994,
+      "step": 3726
+    },
+    {
+      "epoch": 1.9401353461738677,
+      "grad_norm": 0.23412308864317125,
+      "learning_rate": 1.4656468919465816e-05,
+      "loss": 0.1081,
+      "step": 3727
+    },
+    {
+      "epoch": 1.9406559083810515,
+      "grad_norm": 0.22672128076546094,
+      "learning_rate": 1.4643679459085451e-05,
+      "loss": 0.0958,
+      "step": 3728
+    },
+    {
+      "epoch": 1.9411764705882353,
+      "grad_norm": 0.21722673840560922,
+      "learning_rate": 1.4630893269713058e-05,
+      "loss": 0.101,
+      "step": 3729
+    },
+    {
+      "epoch": 1.941697032795419,
+      "grad_norm": 0.21147974417849058,
+      "learning_rate": 1.4618110355387105e-05,
+      "loss": 0.1029,
+      "step": 3730
+    },
+    {
+      "epoch": 1.9422175950026028,
+      "grad_norm": 0.21231022676958888,
+      "learning_rate": 1.4605330720145036e-05,
+      "loss": 0.1004,
+      "step": 3731
+    },
+    {
+      "epoch": 1.9427381572097866,
+      "grad_norm": 0.2202858395550634,
+      "learning_rate": 1.4592554368023248e-05,
+      "loss": 0.1061,
+      "step": 3732
+    },
+    {
+      "epoch": 1.9432587194169704,
+      "grad_norm": 0.2079520954395182,
+      "learning_rate": 1.4579781303057122e-05,
+      "loss": 0.0993,
+      "step": 3733
+    },
+    {
+      "epoch": 1.943779281624154,
+      "grad_norm": 0.2232459002547435,
+      "learning_rate": 1.4567011529280963e-05,
+      "loss": 0.1016,
+      "step": 3734
+    },
+    {
+      "epoch": 1.9442998438313377,
+      "grad_norm": 0.2131387702891851,
+      "learning_rate": 1.4554245050728085e-05,
+      "loss": 0.0976,
+      "step": 3735
+    },
+    {
+      "epoch": 1.9448204060385215,
+      "grad_norm": 0.21742449638032005,
+      "learning_rate": 1.4541481871430712e-05,
+      "loss": 0.0997,
+      "step": 3736
+    },
+    {
+      "epoch": 1.9453409682457052,
+      "grad_norm": 0.21341755263048834,
+      "learning_rate": 1.4528721995420065e-05,
+      "loss": 0.1032,
+      "step": 3737
+    },
+    {
+      "epoch": 1.945861530452889,
+      "grad_norm": 0.22009074356689978,
+      "learning_rate": 1.4515965426726297e-05,
+      "loss": 0.1059,
+      "step": 3738
+    },
+    {
+      "epoch": 1.9463820926600728,
+      "grad_norm": 0.21912610193463983,
+      "learning_rate": 1.4503212169378549e-05,
+      "loss": 0.1001,
+      "step": 3739
+    },
+    {
+      "epoch": 1.9469026548672566,
+      "grad_norm": 0.2170439457377092,
+      "learning_rate": 1.449046222740486e-05,
+      "loss": 0.1057,
+      "step": 3740
+    },
+    {
+      "epoch": 1.9474232170744403,
+      "grad_norm": 0.21785895921005488,
+      "learning_rate": 1.4477715604832277e-05,
+      "loss": 0.1031,
+      "step": 3741
+    },
+    {
+      "epoch": 1.947943779281624,
+      "grad_norm": 0.21495265954061052,
+      "learning_rate": 1.4464972305686777e-05,
+      "loss": 0.0986,
+      "step": 3742
+    },
+    {
+      "epoch": 1.9484643414888079,
+      "grad_norm": 0.20444844893689895,
+      "learning_rate": 1.4452232333993271e-05,
+      "loss": 0.0953,
+      "step": 3743
+    },
+    {
+      "epoch": 1.9489849036959916,
+      "grad_norm": 0.22666704840339114,
+      "learning_rate": 1.4439495693775657e-05,
+      "loss": 0.1055,
+      "step": 3744
+    },
+    {
+      "epoch": 1.9495054659031754,
+      "grad_norm": 0.21327639711763433,
+      "learning_rate": 1.4426762389056735e-05,
+      "loss": 0.1034,
+      "step": 3745
+    },
+    {
+      "epoch": 1.9500260281103592,
+      "grad_norm": 0.2184419717310296,
+      "learning_rate": 1.4414032423858287e-05,
+      "loss": 0.1047,
+      "step": 3746
+    },
+    {
+      "epoch": 1.950546590317543,
+      "grad_norm": 0.21193552764734155,
+      "learning_rate": 1.4401305802201029e-05,
+      "loss": 0.1026,
+      "step": 3747
+    },
+    {
+      "epoch": 1.9510671525247267,
+      "grad_norm": 0.2209102393312676,
+      "learning_rate": 1.4388582528104628e-05,
+      "loss": 0.1011,
+      "step": 3748
+    },
+    {
+      "epoch": 1.9515877147319105,
+      "grad_norm": 0.21657115471489805,
+      "learning_rate": 1.4375862605587668e-05,
+      "loss": 0.0986,
+      "step": 3749
+    },
+    {
+      "epoch": 1.9521082769390943,
+      "grad_norm": 0.21434326778156354,
+      "learning_rate": 1.4363146038667702e-05,
+      "loss": 0.1012,
+      "step": 3750
+    },
+    {
+      "epoch": 1.952628839146278,
+      "grad_norm": 0.21320540069825253,
+      "learning_rate": 1.4350432831361221e-05,
+      "loss": 0.0988,
+      "step": 3751
+    },
+    {
+      "epoch": 1.9531494013534618,
+      "grad_norm": 0.22627494817868693,
+      "learning_rate": 1.4337722987683632e-05,
+      "loss": 0.101,
+      "step": 3752
+    },
+    {
+      "epoch": 1.9536699635606456,
+      "grad_norm": 0.2194395485328437,
+      "learning_rate": 1.4325016511649315e-05,
+      "loss": 0.0982,
+      "step": 3753
+    },
+    {
+      "epoch": 1.9541905257678294,
+      "grad_norm": 0.21191340497198383,
+      "learning_rate": 1.4312313407271543e-05,
+      "loss": 0.099,
+      "step": 3754
+    },
+    {
+      "epoch": 1.9547110879750131,
+      "grad_norm": 0.21476576576191445,
+      "learning_rate": 1.429961367856256e-05,
+      "loss": 0.0995,
+      "step": 3755
+    },
+    {
+      "epoch": 1.955231650182197,
+      "grad_norm": 0.21770044547304337,
+      "learning_rate": 1.4286917329533528e-05,
+      "loss": 0.1015,
+      "step": 3756
+    },
+    {
+      "epoch": 1.9557522123893807,
+      "grad_norm": 0.21158808266880086,
+      "learning_rate": 1.427422436419456e-05,
+      "loss": 0.098,
+      "step": 3757
+    },
+    {
+      "epoch": 1.9562727745965642,
+      "grad_norm": 0.21361489248889914,
+      "learning_rate": 1.4261534786554661e-05,
+      "loss": 0.1001,
+      "step": 3758
+    },
+    {
+      "epoch": 1.956793336803748,
+      "grad_norm": 0.21171491539393147,
+      "learning_rate": 1.4248848600621801e-05,
+      "loss": 0.1007,
+      "step": 3759
+    },
+    {
+      "epoch": 1.9573138990109318,
+      "grad_norm": 0.21237977936588603,
+      "learning_rate": 1.4236165810402879e-05,
+      "loss": 0.0974,
+      "step": 3760
+    },
+    {
+      "epoch": 1.9578344612181156,
+      "grad_norm": 0.22772433653264357,
+      "learning_rate": 1.4223486419903692e-05,
+      "loss": 0.105,
+      "step": 3761
+    },
+    {
+      "epoch": 1.9583550234252993,
+      "grad_norm": 0.21822253874910252,
+      "learning_rate": 1.4210810433128997e-05,
+      "loss": 0.1027,
+      "step": 3762
+    },
+    {
+      "epoch": 1.958875585632483,
+      "grad_norm": 0.21968663113465622,
+      "learning_rate": 1.4198137854082443e-05,
+      "loss": 0.0979,
+      "step": 3763
+    },
+    {
+      "epoch": 1.9593961478396669,
+      "grad_norm": 0.21509071607759678,
+      "learning_rate": 1.4185468686766628e-05,
+      "loss": 0.1002,
+      "step": 3764
+    },
+    {
+      "epoch": 1.9599167100468506,
+      "grad_norm": 0.21833030367787082,
+      "learning_rate": 1.4172802935183071e-05,
+      "loss": 0.104,
+      "step": 3765
+    },
+    {
+      "epoch": 1.9604372722540342,
+      "grad_norm": 0.21150315924836155,
+      "learning_rate": 1.41601406033322e-05,
+      "loss": 0.0976,
+      "step": 3766
+    },
+    {
+      "epoch": 1.960957834461218,
+      "grad_norm": 0.22119462315092003,
+      "learning_rate": 1.4147481695213377e-05,
+      "loss": 0.101,
+      "step": 3767
+    },
+    {
+      "epoch": 1.9614783966684017,
+      "grad_norm": 0.22377655608418107,
+      "learning_rate": 1.4134826214824859e-05,
+      "loss": 0.1087,
+      "step": 3768
+    },
+    {
+      "epoch": 1.9619989588755855,
+      "grad_norm": 0.214726683314058,
+      "learning_rate": 1.4122174166163853e-05,
+      "loss": 0.1011,
+      "step": 3769
+    },
+    {
+      "epoch": 1.9625195210827693,
+      "grad_norm": 0.22428658200555723,
+      "learning_rate": 1.410952555322645e-05,
+      "loss": 0.1002,
+      "step": 3770
+    },
+    {
+      "epoch": 1.963040083289953,
+      "grad_norm": 0.22440945050923508,
+      "learning_rate": 1.4096880380007673e-05,
+      "loss": 0.1079,
+      "step": 3771
+    },
+    {
+      "epoch": 1.9635606454971368,
+      "grad_norm": 0.22006424334394925,
+      "learning_rate": 1.4084238650501471e-05,
+      "loss": 0.103,
+      "step": 3772
+    },
+    {
+      "epoch": 1.9640812077043206,
+      "grad_norm": 0.2078275175323549,
+      "learning_rate": 1.4071600368700668e-05,
+      "loss": 0.0965,
+      "step": 3773
+    },
+    {
+      "epoch": 1.9646017699115044,
+      "grad_norm": 0.20908961325681313,
+      "learning_rate": 1.4058965538597033e-05,
+      "loss": 0.0958,
+      "step": 3774
+    },
+    {
+      "epoch": 1.9651223321186881,
+      "grad_norm": 0.23113004080134392,
+      "learning_rate": 1.4046334164181232e-05,
+      "loss": 0.1046,
+      "step": 3775
+    },
+    {
+      "epoch": 1.965642894325872,
+      "grad_norm": 0.21341710865458288,
+      "learning_rate": 1.4033706249442852e-05,
+      "loss": 0.0992,
+      "step": 3776
+    },
+    {
+      "epoch": 1.9661634565330557,
+      "grad_norm": 0.2255057605756392,
+      "learning_rate": 1.4021081798370356e-05,
+      "loss": 0.1076,
+      "step": 3777
+    },
+    {
+      "epoch": 1.9666840187402395,
+      "grad_norm": 0.22209298875300015,
+      "learning_rate": 1.4008460814951151e-05,
+      "loss": 0.1057,
+      "step": 3778
+    },
+    {
+      "epoch": 1.9672045809474232,
+      "grad_norm": 0.22187334155214908,
+      "learning_rate": 1.3995843303171517e-05,
+      "loss": 0.0982,
+      "step": 3779
+    },
+    {
+      "epoch": 1.967725143154607,
+      "grad_norm": 0.2103303538717618,
+      "learning_rate": 1.3983229267016653e-05,
+      "loss": 0.1011,
+      "step": 3780
+    },
+    {
+      "epoch": 1.9682457053617908,
+      "grad_norm": 0.2040935740244857,
+      "learning_rate": 1.3970618710470676e-05,
+      "loss": 0.0998,
+      "step": 3781
+    },
+    {
+      "epoch": 1.9687662675689745,
+      "grad_norm": 0.22402864584457863,
+      "learning_rate": 1.395801163751656e-05,
+      "loss": 0.1026,
+      "step": 3782
+    },
+    {
+      "epoch": 1.9692868297761583,
+      "grad_norm": 0.20692376854563577,
+      "learning_rate": 1.3945408052136222e-05,
+      "loss": 0.0981,
+      "step": 3783
+    },
+    {
+      "epoch": 1.969807391983342,
+      "grad_norm": 0.22980259328124997,
+      "learning_rate": 1.3932807958310456e-05,
+      "loss": 0.1076,
+      "step": 3784
+    },
+    {
+      "epoch": 1.9703279541905259,
+      "grad_norm": 0.20367490497979784,
+      "learning_rate": 1.392021136001897e-05,
+      "loss": 0.0968,
+      "step": 3785
+    },
+    {
+      "epoch": 1.9708485163977096,
+      "grad_norm": 0.20898246158228093,
+      "learning_rate": 1.3907618261240334e-05,
+      "loss": 0.1033,
+      "step": 3786
+    },
+    {
+      "epoch": 1.9713690786048934,
+      "grad_norm": 0.2113639999243892,
+      "learning_rate": 1.3895028665952058e-05,
+      "loss": 0.1014,
+      "step": 3787
+    },
+    {
+      "epoch": 1.9718896408120772,
+      "grad_norm": 0.2175367277530161,
+      "learning_rate": 1.38824425781305e-05,
+      "loss": 0.1008,
+      "step": 3788
+    },
+    {
+      "epoch": 1.972410203019261,
+      "grad_norm": 0.21182935807043565,
+      "learning_rate": 1.3869860001750942e-05,
+      "loss": 0.1013,
+      "step": 3789
+    },
+    {
+      "epoch": 1.9729307652264445,
+      "grad_norm": 0.22349166273303656,
+      "learning_rate": 1.3857280940787559e-05,
+      "loss": 0.0988,
+      "step": 3790
+    },
+    {
+      "epoch": 1.9734513274336283,
+      "grad_norm": 0.21190197883778825,
+      "learning_rate": 1.3844705399213379e-05,
+      "loss": 0.0995,
+      "step": 3791
+    },
+    {
+      "epoch": 1.973971889640812,
+      "grad_norm": 0.2193754184132712,
+      "learning_rate": 1.3832133381000359e-05,
+      "loss": 0.105,
+      "step": 3792
+    },
+    {
+      "epoch": 1.9744924518479958,
+      "grad_norm": 0.22565894195227415,
+      "learning_rate": 1.3819564890119325e-05,
+      "loss": 0.0992,
+      "step": 3793
+    },
+    {
+      "epoch": 1.9750130140551796,
+      "grad_norm": 0.2223063511163848,
+      "learning_rate": 1.3806999930539998e-05,
+      "loss": 0.1082,
+      "step": 3794
+    },
+    {
+      "epoch": 1.9755335762623634,
+      "grad_norm": 0.2182868394326183,
+      "learning_rate": 1.3794438506230967e-05,
+      "loss": 0.0969,
+      "step": 3795
+    },
+    {
+      "epoch": 1.9760541384695471,
+      "grad_norm": 0.21985719559638187,
+      "learning_rate": 1.378188062115972e-05,
+      "loss": 0.1031,
+      "step": 3796
+    },
+    {
+      "epoch": 1.976574700676731,
+      "grad_norm": 0.21709916965896411,
+      "learning_rate": 1.3769326279292616e-05,
+      "loss": 0.1,
+      "step": 3797
+    },
+    {
+      "epoch": 1.9770952628839145,
+      "grad_norm": 0.21263835779355764,
+      "learning_rate": 1.3756775484594896e-05,
+      "loss": 0.1004,
+      "step": 3798
+    },
+    {
+      "epoch": 1.9776158250910982,
+      "grad_norm": 0.21782936591583624,
+      "learning_rate": 1.374422824103071e-05,
+      "loss": 0.102,
+      "step": 3799
+    },
+    {
+      "epoch": 1.978136387298282,
+      "grad_norm": 0.21792766988581672,
+      "learning_rate": 1.373168455256303e-05,
+      "loss": 0.1007,
+      "step": 3800
+    },
+    {
+      "epoch": 1.9786569495054658,
+      "grad_norm": 0.20963925747095527,
+      "learning_rate": 1.3719144423153751e-05,
+      "loss": 0.0944,
+      "step": 3801
+    },
+    {
+      "epoch": 1.9791775117126496,
+      "grad_norm": 0.2122594378833588,
+      "learning_rate": 1.370660785676363e-05,
+      "loss": 0.0998,
+      "step": 3802
+    },
+    {
+      "epoch": 1.9796980739198333,
+      "grad_norm": 0.2213288489164458,
+      "learning_rate": 1.3694074857352308e-05,
+      "loss": 0.1002,
+      "step": 3803
+    },
+    {
+      "epoch": 1.980218636127017,
+      "grad_norm": 0.21750190142433626,
+      "learning_rate": 1.3681545428878268e-05,
+      "loss": 0.1015,
+      "step": 3804
+    },
+    {
+      "epoch": 1.9807391983342009,
+      "grad_norm": 0.21026897260525107,
+      "learning_rate": 1.3669019575298902e-05,
+      "loss": 0.097,
+      "step": 3805
+    },
+    {
+      "epoch": 1.9812597605413846,
+      "grad_norm": 0.22344596462757446,
+      "learning_rate": 1.3656497300570448e-05,
+      "loss": 0.1017,
+      "step": 3806
+    },
+    {
+      "epoch": 1.9817803227485684,
+      "grad_norm": 0.23421057868440046,
+      "learning_rate": 1.3643978608648028e-05,
+      "loss": 0.1042,
+      "step": 3807
+    },
+    {
+      "epoch": 1.9823008849557522,
+      "grad_norm": 0.2200839534373584,
+      "learning_rate": 1.3631463503485634e-05,
+      "loss": 0.0985,
+      "step": 3808
+    },
+    {
+      "epoch": 1.982821447162936,
+      "grad_norm": 0.21843886850429678,
+      "learning_rate": 1.3618951989036102e-05,
+      "loss": 0.1027,
+      "step": 3809
+    },
+    {
+      "epoch": 1.9833420093701197,
+      "grad_norm": 0.21096923350471475,
+      "learning_rate": 1.3606444069251162e-05,
+      "loss": 0.0993,
+      "step": 3810
+    },
+    {
+      "epoch": 1.9838625715773035,
+      "grad_norm": 0.21343331941361038,
+      "learning_rate": 1.3593939748081393e-05,
+      "loss": 0.0993,
+      "step": 3811
+    },
+    {
+      "epoch": 1.9843831337844873,
+      "grad_norm": 0.22152103395034903,
+      "learning_rate": 1.3581439029476255e-05,
+      "loss": 0.1011,
+      "step": 3812
+    },
+    {
+      "epoch": 1.984903695991671,
+      "grad_norm": 0.20750708820454272,
+      "learning_rate": 1.3568941917384036e-05,
+      "loss": 0.1011,
+      "step": 3813
+    },
+    {
+      "epoch": 1.9854242581988548,
+      "grad_norm": 0.22282972962697625,
+      "learning_rate": 1.3556448415751927e-05,
+      "loss": 0.1,
+      "step": 3814
+    },
+    {
+      "epoch": 1.9859448204060386,
+      "grad_norm": 0.21973378480721914,
+      "learning_rate": 1.3543958528525934e-05,
+      "loss": 0.0998,
+      "step": 3815
+    },
+    {
+      "epoch": 1.9864653826132224,
+      "grad_norm": 0.21975469487000557,
+      "learning_rate": 1.3531472259650956e-05,
+      "loss": 0.1027,
+      "step": 3816
+    },
+    {
+      "epoch": 1.9869859448204061,
+      "grad_norm": 0.20704264881854378,
+      "learning_rate": 1.3518989613070745e-05,
+      "loss": 0.0992,
+      "step": 3817
+    },
+    {
+      "epoch": 1.98750650702759,
+      "grad_norm": 0.2276001326716296,
+      "learning_rate": 1.3506510592727889e-05,
+      "loss": 0.103,
+      "step": 3818
+    },
+    {
+      "epoch": 1.9880270692347737,
+      "grad_norm": 0.21862242414923339,
+      "learning_rate": 1.3494035202563842e-05,
+      "loss": 0.102,
+      "step": 3819
+    },
+    {
+      "epoch": 1.9885476314419575,
+      "grad_norm": 0.20514072992531857,
+      "learning_rate": 1.3481563446518924e-05,
+      "loss": 0.0979,
+      "step": 3820
+    },
+    {
+      "epoch": 1.9890681936491412,
+      "grad_norm": 0.219390795219004,
+      "learning_rate": 1.3469095328532305e-05,
+      "loss": 0.1015,
+      "step": 3821
+    },
+    {
+      "epoch": 1.9895887558563248,
+      "grad_norm": 0.2234331591905043,
+      "learning_rate": 1.3456630852541968e-05,
+      "loss": 0.0986,
+      "step": 3822
+    },
+    {
+      "epoch": 1.9901093180635085,
+      "grad_norm": 0.22043672801905884,
+      "learning_rate": 1.3444170022484803e-05,
+      "loss": 0.1,
+      "step": 3823
+    },
+    {
+      "epoch": 1.9906298802706923,
+      "grad_norm": 0.2121077118944676,
+      "learning_rate": 1.3431712842296495e-05,
+      "loss": 0.0992,
+      "step": 3824
+    },
+    {
+      "epoch": 1.991150442477876,
+      "grad_norm": 0.2098280218821776,
+      "learning_rate": 1.3419259315911612e-05,
+      "loss": 0.0974,
+      "step": 3825
+    },
+    {
+      "epoch": 1.9916710046850599,
+      "grad_norm": 0.22151509852945767,
+      "learning_rate": 1.3406809447263569e-05,
+      "loss": 0.1058,
+      "step": 3826
+    },
+    {
+      "epoch": 1.9921915668922436,
+      "grad_norm": 0.2188216176247912,
+      "learning_rate": 1.3394363240284596e-05,
+      "loss": 0.1001,
+      "step": 3827
+    },
+    {
+      "epoch": 1.9927121290994274,
+      "grad_norm": 0.2175007496547397,
+      "learning_rate": 1.3381920698905787e-05,
+      "loss": 0.0997,
+      "step": 3828
+    },
+    {
+      "epoch": 1.9932326913066112,
+      "grad_norm": 0.22386791931729272,
+      "learning_rate": 1.3369481827057084e-05,
+      "loss": 0.1046,
+      "step": 3829
+    },
+    {
+      "epoch": 1.9937532535137947,
+      "grad_norm": 0.22189589045670546,
+      "learning_rate": 1.3357046628667266e-05,
+      "loss": 0.102,
+      "step": 3830
+    },
+    {
+      "epoch": 1.9942738157209785,
+      "grad_norm": 0.22222680495180996,
+      "learning_rate": 1.3344615107663929e-05,
+      "loss": 0.1015,
+      "step": 3831
+    },
+    {
+      "epoch": 1.9947943779281623,
+      "grad_norm": 0.21165370943689257,
+      "learning_rate": 1.3332187267973545e-05,
+      "loss": 0.1,
+      "step": 3832
+    },
+    {
+      "epoch": 1.995314940135346,
+      "grad_norm": 0.21493189063483512,
+      "learning_rate": 1.3319763113521388e-05,
+      "loss": 0.0988,
+      "step": 3833
+    },
+    {
+      "epoch": 1.9958355023425298,
+      "grad_norm": 0.2152720610885134,
+      "learning_rate": 1.3307342648231587e-05,
+      "loss": 0.1046,
+      "step": 3834
+    },
+    {
+      "epoch": 1.9963560645497136,
+      "grad_norm": 0.22504773486380045,
+      "learning_rate": 1.3294925876027112e-05,
+      "loss": 0.1018,
+      "step": 3835
+    },
+    {
+      "epoch": 1.9968766267568974,
+      "grad_norm": 0.21719340129386458,
+      "learning_rate": 1.3282512800829761e-05,
+      "loss": 0.0994,
+      "step": 3836
+    },
+    {
+      "epoch": 1.9973971889640811,
+      "grad_norm": 0.2351408224727131,
+      "learning_rate": 1.3270103426560138e-05,
+      "loss": 0.1031,
+      "step": 3837
+    },
+    {
+      "epoch": 1.997917751171265,
+      "grad_norm": 0.22730448971582903,
+      "learning_rate": 1.3257697757137722e-05,
+      "loss": 0.1018,
+      "step": 3838
+    },
+    {
+      "epoch": 1.9984383133784487,
+      "grad_norm": 0.2240262905997672,
+      "learning_rate": 1.3245295796480789e-05,
+      "loss": 0.0991,
+      "step": 3839
+    },
+    {
+      "epoch": 1.9989588755856325,
+      "grad_norm": 0.20159272727142402,
+      "learning_rate": 1.3232897548506473e-05,
+      "loss": 0.0929,
+      "step": 3840
+    },
+    {
+      "epoch": 1.9994794377928162,
+      "grad_norm": 0.2185282556675355,
+      "learning_rate": 1.3220503017130702e-05,
+      "loss": 0.1029,
+      "step": 3841
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.21010022623920993,
+      "learning_rate": 1.3208112206268241e-05,
+      "loss": 0.0951,
+      "step": 3842
+    },
+    {
+      "epoch": 2.0005205622071838,
+      "grad_norm": 0.23888745487077462,
+      "learning_rate": 1.3195725119832692e-05,
+      "loss": 0.0589,
+      "step": 3843
+    },
+    {
+      "epoch": 2.0010411244143675,
+      "grad_norm": 0.22113727902582608,
+      "learning_rate": 1.3183341761736474e-05,
+      "loss": 0.059,
+      "step": 3844
+    },
+    {
+      "epoch": 2.0015616866215513,
+      "grad_norm": 0.2100996955115601,
+      "learning_rate": 1.3170962135890838e-05,
+      "loss": 0.057,
+      "step": 3845
+    },
+    {
+      "epoch": 2.002082248828735,
+      "grad_norm": 0.21218819629958469,
+      "learning_rate": 1.3158586246205823e-05,
+      "loss": 0.0553,
+      "step": 3846
+    },
+    {
+      "epoch": 2.002602811035919,
+      "grad_norm": 0.20870010973163983,
+      "learning_rate": 1.3146214096590326e-05,
+      "loss": 0.0527,
+      "step": 3847
+    },
+    {
+      "epoch": 2.0031233732431026,
+      "grad_norm": 0.2324010648172641,
+      "learning_rate": 1.3133845690952045e-05,
+      "loss": 0.0533,
+      "step": 3848
+    },
+    {
+      "epoch": 2.0036439354502864,
+      "grad_norm": 0.27402223811767706,
+      "learning_rate": 1.3121481033197508e-05,
+      "loss": 0.0548,
+      "step": 3849
+    },
+    {
+      "epoch": 2.00416449765747,
+      "grad_norm": 0.3136132661537596,
+      "learning_rate": 1.3109120127232039e-05,
+      "loss": 0.0561,
+      "step": 3850
+    },
+    {
+      "epoch": 2.004685059864654,
+      "grad_norm": 0.28713280826038895,
+      "learning_rate": 1.3096762976959776e-05,
+      "loss": 0.0554,
+      "step": 3851
+    },
+    {
+      "epoch": 2.0052056220718377,
+      "grad_norm": 0.2568407408897936,
+      "learning_rate": 1.3084409586283696e-05,
+      "loss": 0.0537,
+      "step": 3852
+    },
+    {
+      "epoch": 2.0057261842790215,
+      "grad_norm": 0.2422352941267545,
+      "learning_rate": 1.307205995910557e-05,
+      "loss": 0.0562,
+      "step": 3853
+    },
+    {
+      "epoch": 2.0062467464862053,
+      "grad_norm": 0.2393420163757943,
+      "learning_rate": 1.3059714099326e-05,
+      "loss": 0.055,
+      "step": 3854
+    },
+    {
+      "epoch": 2.006767308693389,
+      "grad_norm": 0.22062809727347588,
+      "learning_rate": 1.3047372010844361e-05,
+      "loss": 0.0532,
+      "step": 3855
+    },
+    {
+      "epoch": 2.007287870900573,
+      "grad_norm": 0.22318736134320938,
+      "learning_rate": 1.3035033697558868e-05,
+      "loss": 0.052,
+      "step": 3856
+    },
+    {
+      "epoch": 2.0078084331077566,
+      "grad_norm": 0.2175095710848304,
+      "learning_rate": 1.302269916336653e-05,
+      "loss": 0.0532,
+      "step": 3857
+    },
+    {
+      "epoch": 2.0083289953149404,
+      "grad_norm": 0.2043884688220176,
+      "learning_rate": 1.3010368412163187e-05,
+      "loss": 0.0522,
+      "step": 3858
+    },
+    {
+      "epoch": 2.0088495575221237,
+      "grad_norm": 0.21515354889304936,
+      "learning_rate": 1.2998041447843448e-05,
+      "loss": 0.055,
+      "step": 3859
+    },
+    {
+      "epoch": 2.0093701197293075,
+      "grad_norm": 0.22420587505565287,
+      "learning_rate": 1.2985718274300731e-05,
+      "loss": 0.0516,
+      "step": 3860
+    },
+    {
+      "epoch": 2.0098906819364912,
+      "grad_norm": 0.2049175307303528,
+      "learning_rate": 1.2973398895427283e-05,
+      "loss": 0.0519,
+      "step": 3861
+    },
+    {
+      "epoch": 2.010411244143675,
+      "grad_norm": 0.21591901082933884,
+      "learning_rate": 1.2961083315114131e-05,
+      "loss": 0.0527,
+      "step": 3862
+    },
+    {
+      "epoch": 2.0109318063508588,
+      "grad_norm": 0.22484304767098973,
+      "learning_rate": 1.2948771537251119e-05,
+      "loss": 0.053,
+      "step": 3863
+    },
+    {
+      "epoch": 2.0114523685580425,
+      "grad_norm": 0.22695563724281367,
+      "learning_rate": 1.2936463565726864e-05,
+      "loss": 0.0522,
+      "step": 3864
+    },
+    {
+      "epoch": 2.0119729307652263,
+      "grad_norm": 0.22138427669788502,
+      "learning_rate": 1.2924159404428803e-05,
+      "loss": 0.052,
+      "step": 3865
+    },
+    {
+      "epoch": 2.01249349297241,
+      "grad_norm": 0.2260270817057609,
+      "learning_rate": 1.2911859057243165e-05,
+      "loss": 0.052,
+      "step": 3866
+    },
+    {
+      "epoch": 2.013014055179594,
+      "grad_norm": 0.22818335275933718,
+      "learning_rate": 1.2899562528054981e-05,
+      "loss": 0.0514,
+      "step": 3867
+    },
+    {
+      "epoch": 2.0135346173867776,
+      "grad_norm": 0.2291490389139092,
+      "learning_rate": 1.2887269820748044e-05,
+      "loss": 0.0522,
+      "step": 3868
+    },
+    {
+      "epoch": 2.0140551795939614,
+      "grad_norm": 0.2253847436578246,
+      "learning_rate": 1.2874980939204984e-05,
+      "loss": 0.0514,
+      "step": 3869
+    },
+    {
+      "epoch": 2.014575741801145,
+      "grad_norm": 0.2376467323645594,
+      "learning_rate": 1.2862695887307186e-05,
+      "loss": 0.0531,
+      "step": 3870
+    },
+    {
+      "epoch": 2.015096304008329,
+      "grad_norm": 0.22645666631710104,
+      "learning_rate": 1.285041466893485e-05,
+      "loss": 0.052,
+      "step": 3871
+    },
+    {
+      "epoch": 2.0156168662155127,
+      "grad_norm": 0.2422877418801926,
+      "learning_rate": 1.2838137287966961e-05,
+      "loss": 0.0507,
+      "step": 3872
+    },
+    {
+      "epoch": 2.0161374284226965,
+      "grad_norm": 0.2506319561703479,
+      "learning_rate": 1.282586374828127e-05,
+      "loss": 0.0503,
+      "step": 3873
+    },
+    {
+      "epoch": 2.0166579906298803,
+      "grad_norm": 0.23453280208493882,
+      "learning_rate": 1.2813594053754346e-05,
+      "loss": 0.0543,
+      "step": 3874
+    },
+    {
+      "epoch": 2.017178552837064,
+      "grad_norm": 0.22699601063395236,
+      "learning_rate": 1.2801328208261526e-05,
+      "loss": 0.0524,
+      "step": 3875
+    },
+    {
+      "epoch": 2.017699115044248,
+      "grad_norm": 0.24112589242280094,
+      "learning_rate": 1.2789066215676943e-05,
+      "loss": 0.0522,
+      "step": 3876
+    },
+    {
+      "epoch": 2.0182196772514316,
+      "grad_norm": 0.23871504911812172,
+      "learning_rate": 1.2776808079873487e-05,
+      "loss": 0.0535,
+      "step": 3877
+    },
+    {
+      "epoch": 2.0187402394586154,
+      "grad_norm": 0.23353444655198632,
+      "learning_rate": 1.2764553804722867e-05,
+      "loss": 0.052,
+      "step": 3878
+    },
+    {
+      "epoch": 2.019260801665799,
+      "grad_norm": 0.21500168671332048,
+      "learning_rate": 1.2752303394095538e-05,
+      "loss": 0.0499,
+      "step": 3879
+    },
+    {
+      "epoch": 2.019781363872983,
+      "grad_norm": 0.22650014917043784,
+      "learning_rate": 1.2740056851860754e-05,
+      "loss": 0.0556,
+      "step": 3880
+    },
+    {
+      "epoch": 2.0203019260801667,
+      "grad_norm": 0.23063435161871682,
+      "learning_rate": 1.2727814181886555e-05,
+      "loss": 0.0529,
+      "step": 3881
+    },
+    {
+      "epoch": 2.0208224882873504,
+      "grad_norm": 0.22502317978846956,
+      "learning_rate": 1.2715575388039724e-05,
+      "loss": 0.0531,
+      "step": 3882
+    },
+    {
+      "epoch": 2.021343050494534,
+      "grad_norm": 0.23365639149073505,
+      "learning_rate": 1.2703340474185854e-05,
+      "loss": 0.0542,
+      "step": 3883
+    },
+    {
+      "epoch": 2.021863612701718,
+      "grad_norm": 0.22024899646428664,
+      "learning_rate": 1.2691109444189303e-05,
+      "loss": 0.0503,
+      "step": 3884
+    },
+    {
+      "epoch": 2.0223841749089018,
+      "grad_norm": 0.23270362452207452,
+      "learning_rate": 1.2678882301913202e-05,
+      "loss": 0.0551,
+      "step": 3885
+    },
+    {
+      "epoch": 2.0229047371160855,
+      "grad_norm": 0.23138418992597184,
+      "learning_rate": 1.2666659051219437e-05,
+      "loss": 0.053,
+      "step": 3886
+    },
+    {
+      "epoch": 2.0234252993232693,
+      "grad_norm": 0.2239621991704845,
+      "learning_rate": 1.2654439695968696e-05,
+      "loss": 0.0508,
+      "step": 3887
+    },
+    {
+      "epoch": 2.023945861530453,
+      "grad_norm": 0.21409030138660093,
+      "learning_rate": 1.2642224240020404e-05,
+      "loss": 0.0521,
+      "step": 3888
+    },
+    {
+      "epoch": 2.024466423737637,
+      "grad_norm": 0.2342885910165884,
+      "learning_rate": 1.263001268723278e-05,
+      "loss": 0.0511,
+      "step": 3889
+    },
+    {
+      "epoch": 2.02498698594482,
+      "grad_norm": 0.230010672562304,
+      "learning_rate": 1.2617805041462805e-05,
+      "loss": 0.0532,
+      "step": 3890
+    },
+    {
+      "epoch": 2.025507548152004,
+      "grad_norm": 0.22439174131590303,
+      "learning_rate": 1.2605601306566205e-05,
+      "loss": 0.0506,
+      "step": 3891
+    },
+    {
+      "epoch": 2.0260281103591877,
+      "grad_norm": 0.22388621675313647,
+      "learning_rate": 1.2593401486397499e-05,
+      "loss": 0.0517,
+      "step": 3892
+    },
+    {
+      "epoch": 2.0265486725663715,
+      "grad_norm": 0.22601607131458107,
+      "learning_rate": 1.258120558480996e-05,
+      "loss": 0.0526,
+      "step": 3893
+    },
+    {
+      "epoch": 2.0270692347735553,
+      "grad_norm": 0.22319982470499325,
+      "learning_rate": 1.2569013605655627e-05,
+      "loss": 0.0503,
+      "step": 3894
+    },
+    {
+      "epoch": 2.027589796980739,
+      "grad_norm": 0.23487166926356898,
+      "learning_rate": 1.2556825552785273e-05,
+      "loss": 0.0527,
+      "step": 3895
+    },
+    {
+      "epoch": 2.028110359187923,
+      "grad_norm": 0.23072170266065373,
+      "learning_rate": 1.2544641430048479e-05,
+      "loss": 0.052,
+      "step": 3896
+    },
+    {
+      "epoch": 2.0286309213951066,
+      "grad_norm": 0.22670085314206623,
+      "learning_rate": 1.2532461241293531e-05,
+      "loss": 0.0505,
+      "step": 3897
+    },
+    {
+      "epoch": 2.0291514836022904,
+      "grad_norm": 0.23480717165950493,
+      "learning_rate": 1.2520284990367514e-05,
+      "loss": 0.052,
+      "step": 3898
+    },
+    {
+      "epoch": 2.029672045809474,
+      "grad_norm": 0.22717862102204162,
+      "learning_rate": 1.2508112681116263e-05,
+      "loss": 0.0513,
+      "step": 3899
+    },
+    {
+      "epoch": 2.030192608016658,
+      "grad_norm": 0.2259905619825204,
+      "learning_rate": 1.2495944317384337e-05,
+      "loss": 0.0513,
+      "step": 3900
+    },
+    {
+      "epoch": 2.0307131702238417,
+      "grad_norm": 0.23979551126797663,
+      "learning_rate": 1.2483779903015086e-05,
+      "loss": 0.0498,
+      "step": 3901
+    },
+    {
+      "epoch": 2.0312337324310255,
+      "grad_norm": 0.22847333964606212,
+      "learning_rate": 1.2471619441850596e-05,
+      "loss": 0.0509,
+      "step": 3902
+    },
+    {
+      "epoch": 2.0317542946382092,
+      "grad_norm": 0.23177006094278615,
+      "learning_rate": 1.2459462937731708e-05,
+      "loss": 0.0506,
+      "step": 3903
+    },
+    {
+      "epoch": 2.032274856845393,
+      "grad_norm": 0.24262261994408266,
+      "learning_rate": 1.2447310394498019e-05,
+      "loss": 0.053,
+      "step": 3904
+    },
+    {
+      "epoch": 2.0327954190525768,
+      "grad_norm": 0.23652782752700235,
+      "learning_rate": 1.2435161815987859e-05,
+      "loss": 0.0528,
+      "step": 3905
+    },
+    {
+      "epoch": 2.0333159812597605,
+      "grad_norm": 0.23162625176396265,
+      "learning_rate": 1.2423017206038307e-05,
+      "loss": 0.0537,
+      "step": 3906
+    },
+    {
+      "epoch": 2.0338365434669443,
+      "grad_norm": 0.22628246516711711,
+      "learning_rate": 1.2410876568485203e-05,
+      "loss": 0.0523,
+      "step": 3907
+    },
+    {
+      "epoch": 2.034357105674128,
+      "grad_norm": 0.23814151638305578,
+      "learning_rate": 1.2398739907163124e-05,
+      "loss": 0.0518,
+      "step": 3908
+    },
+    {
+      "epoch": 2.034877667881312,
+      "grad_norm": 0.24212468958455002,
+      "learning_rate": 1.2386607225905405e-05,
+      "loss": 0.0522,
+      "step": 3909
+    },
+    {
+      "epoch": 2.0353982300884956,
+      "grad_norm": 0.22458499825758288,
+      "learning_rate": 1.2374478528544092e-05,
+      "loss": 0.0503,
+      "step": 3910
+    },
+    {
+      "epoch": 2.0359187922956794,
+      "grad_norm": 0.2154706609637643,
+      "learning_rate": 1.2362353818910002e-05,
+      "loss": 0.0505,
+      "step": 3911
+    },
+    {
+      "epoch": 2.036439354502863,
+      "grad_norm": 0.23207911313268095,
+      "learning_rate": 1.2350233100832678e-05,
+      "loss": 0.0523,
+      "step": 3912
+    },
+    {
+      "epoch": 2.036959916710047,
+      "grad_norm": 0.2275097951477535,
+      "learning_rate": 1.2338116378140424e-05,
+      "loss": 0.0508,
+      "step": 3913
+    },
+    {
+      "epoch": 2.0374804789172307,
+      "grad_norm": 0.24046310726808087,
+      "learning_rate": 1.2326003654660249e-05,
+      "loss": 0.0486,
+      "step": 3914
+    },
+    {
+      "epoch": 2.0380010411244145,
+      "grad_norm": 0.25638052879623396,
+      "learning_rate": 1.2313894934217907e-05,
+      "loss": 0.0506,
+      "step": 3915
+    },
+    {
+      "epoch": 2.0385216033315983,
+      "grad_norm": 0.2241861365775457,
+      "learning_rate": 1.2301790220637904e-05,
+      "loss": 0.0504,
+      "step": 3916
+    },
+    {
+      "epoch": 2.039042165538782,
+      "grad_norm": 0.21754058305545665,
+      "learning_rate": 1.2289689517743475e-05,
+      "loss": 0.0478,
+      "step": 3917
+    },
+    {
+      "epoch": 2.039562727745966,
+      "grad_norm": 0.2471236978327747,
+      "learning_rate": 1.2277592829356593e-05,
+      "loss": 0.0539,
+      "step": 3918
+    },
+    {
+      "epoch": 2.0400832899531496,
+      "grad_norm": 0.23244298598394517,
+      "learning_rate": 1.2265500159297935e-05,
+      "loss": 0.0514,
+      "step": 3919
+    },
+    {
+      "epoch": 2.0406038521603334,
+      "grad_norm": 0.2434195925375697,
+      "learning_rate": 1.2253411511386938e-05,
+      "loss": 0.0529,
+      "step": 3920
+    },
+    {
+      "epoch": 2.041124414367517,
+      "grad_norm": 0.23277356231448715,
+      "learning_rate": 1.2241326889441763e-05,
+      "loss": 0.0509,
+      "step": 3921
+    },
+    {
+      "epoch": 2.041644976574701,
+      "grad_norm": 0.2326730216527539,
+      "learning_rate": 1.2229246297279302e-05,
+      "loss": 0.0499,
+      "step": 3922
+    },
+    {
+      "epoch": 2.0421655387818842,
+      "grad_norm": 0.22920084540866484,
+      "learning_rate": 1.2217169738715162e-05,
+      "loss": 0.052,
+      "step": 3923
+    },
+    {
+      "epoch": 2.042686100989068,
+      "grad_norm": 0.2283017342208316,
+      "learning_rate": 1.2205097217563668e-05,
+      "loss": 0.0503,
+      "step": 3924
+    },
+    {
+      "epoch": 2.0432066631962518,
+      "grad_norm": 0.2316172935733046,
+      "learning_rate": 1.2193028737637897e-05,
+      "loss": 0.0519,
+      "step": 3925
+    },
+    {
+      "epoch": 2.0437272254034355,
+      "grad_norm": 0.2165080484448344,
+      "learning_rate": 1.2180964302749637e-05,
+      "loss": 0.0483,
+      "step": 3926
+    },
+    {
+      "epoch": 2.0442477876106193,
+      "grad_norm": 0.21978990050785774,
+      "learning_rate": 1.2168903916709404e-05,
+      "loss": 0.0499,
+      "step": 3927
+    },
+    {
+      "epoch": 2.044768349817803,
+      "grad_norm": 0.24452756991961144,
+      "learning_rate": 1.2156847583326414e-05,
+      "loss": 0.0532,
+      "step": 3928
+    },
+    {
+      "epoch": 2.045288912024987,
+      "grad_norm": 0.21604477471728578,
+      "learning_rate": 1.2144795306408626e-05,
+      "loss": 0.0484,
+      "step": 3929
+    },
+    {
+      "epoch": 2.0458094742321706,
+      "grad_norm": 0.22490194735030902,
+      "learning_rate": 1.213274708976271e-05,
+      "loss": 0.0507,
+      "step": 3930
+    },
+    {
+      "epoch": 2.0463300364393544,
+      "grad_norm": 0.23117597236920664,
+      "learning_rate": 1.2120702937194061e-05,
+      "loss": 0.0499,
+      "step": 3931
+    },
+    {
+      "epoch": 2.046850598646538,
+      "grad_norm": 0.2302153815021224,
+      "learning_rate": 1.2108662852506778e-05,
+      "loss": 0.0514,
+      "step": 3932
+    },
+    {
+      "epoch": 2.047371160853722,
+      "grad_norm": 0.22991731459055237,
+      "learning_rate": 1.2096626839503666e-05,
+      "loss": 0.0504,
+      "step": 3933
+    },
+    {
+      "epoch": 2.0478917230609057,
+      "grad_norm": 0.2336931030300491,
+      "learning_rate": 1.2084594901986271e-05,
+      "loss": 0.051,
+      "step": 3934
+    },
+    {
+      "epoch": 2.0484122852680895,
+      "grad_norm": 0.23611354689827643,
+      "learning_rate": 1.2072567043754837e-05,
+      "loss": 0.0505,
+      "step": 3935
+    },
+    {
+      "epoch": 2.0489328474752733,
+      "grad_norm": 0.24039532523121837,
+      "learning_rate": 1.2060543268608329e-05,
+      "loss": 0.0533,
+      "step": 3936
+    },
+    {
+      "epoch": 2.049453409682457,
+      "grad_norm": 0.23726727147966595,
+      "learning_rate": 1.2048523580344398e-05,
+      "loss": 0.0507,
+      "step": 3937
+    },
+    {
+      "epoch": 2.049973971889641,
+      "grad_norm": 0.2385742302487926,
+      "learning_rate": 1.2036507982759431e-05,
+      "loss": 0.0526,
+      "step": 3938
+    },
+    {
+      "epoch": 2.0504945340968246,
+      "grad_norm": 0.22713140069481194,
+      "learning_rate": 1.2024496479648514e-05,
+      "loss": 0.0507,
+      "step": 3939
+    },
+    {
+      "epoch": 2.0510150963040084,
+      "grad_norm": 0.22422405740318607,
+      "learning_rate": 1.2012489074805444e-05,
+      "loss": 0.049,
+      "step": 3940
+    },
+    {
+      "epoch": 2.051535658511192,
+      "grad_norm": 0.23105843716868302,
+      "learning_rate": 1.200048577202271e-05,
+      "loss": 0.0529,
+      "step": 3941
+    },
+    {
+      "epoch": 2.052056220718376,
+      "grad_norm": 0.23528988007664547,
+      "learning_rate": 1.1988486575091507e-05,
+      "loss": 0.0533,
+      "step": 3942
+    },
+    {
+      "epoch": 2.0525767829255597,
+      "grad_norm": 0.23894532699169194,
+      "learning_rate": 1.1976491487801748e-05,
+      "loss": 0.0524,
+      "step": 3943
+    },
+    {
+      "epoch": 2.0530973451327434,
+      "grad_norm": 0.23433351722426457,
+      "learning_rate": 1.1964500513942034e-05,
+      "loss": 0.0503,
+      "step": 3944
+    },
+    {
+      "epoch": 2.053617907339927,
+      "grad_norm": 0.2429361501735258,
+      "learning_rate": 1.1952513657299691e-05,
+      "loss": 0.0525,
+      "step": 3945
+    },
+    {
+      "epoch": 2.054138469547111,
+      "grad_norm": 0.2357366347477472,
+      "learning_rate": 1.1940530921660703e-05,
+      "loss": 0.0508,
+      "step": 3946
+    },
+    {
+      "epoch": 2.0546590317542948,
+      "grad_norm": 0.22479006903552398,
+      "learning_rate": 1.1928552310809785e-05,
+      "loss": 0.0499,
+      "step": 3947
+    },
+    {
+      "epoch": 2.0551795939614785,
+      "grad_norm": 0.22971193296557763,
+      "learning_rate": 1.191657782853034e-05,
+      "loss": 0.0505,
+      "step": 3948
+    },
+    {
+      "epoch": 2.0557001561686623,
+      "grad_norm": 0.2251516109143194,
+      "learning_rate": 1.1904607478604476e-05,
+      "loss": 0.0508,
+      "step": 3949
+    },
+    {
+      "epoch": 2.056220718375846,
+      "grad_norm": 0.21994611285526833,
+      "learning_rate": 1.1892641264812978e-05,
+      "loss": 0.0491,
+      "step": 3950
+    },
+    {
+      "epoch": 2.05674128058303,
+      "grad_norm": 0.23498529308957097,
+      "learning_rate": 1.1880679190935323e-05,
+      "loss": 0.0505,
+      "step": 3951
+    },
+    {
+      "epoch": 2.0572618427902136,
+      "grad_norm": 0.24181869335547906,
+      "learning_rate": 1.1868721260749699e-05,
+      "loss": 0.0506,
+      "step": 3952
+    },
+    {
+      "epoch": 2.0577824049973974,
+      "grad_norm": 0.251727766194006,
+      "learning_rate": 1.1856767478032979e-05,
+      "loss": 0.0524,
+      "step": 3953
+    },
+    {
+      "epoch": 2.0583029672045807,
+      "grad_norm": 0.22110211479371478,
+      "learning_rate": 1.1844817846560732e-05,
+      "loss": 0.0489,
+      "step": 3954
+    },
+    {
+      "epoch": 2.0588235294117645,
+      "grad_norm": 0.25357982403761625,
+      "learning_rate": 1.183287237010719e-05,
+      "loss": 0.0524,
+      "step": 3955
+    },
+    {
+      "epoch": 2.0593440916189483,
+      "grad_norm": 0.2435984952058527,
+      "learning_rate": 1.1820931052445297e-05,
+      "loss": 0.053,
+      "step": 3956
+    },
+    {
+      "epoch": 2.059864653826132,
+      "grad_norm": 0.23320265572221055,
+      "learning_rate": 1.180899389734668e-05,
+      "loss": 0.0494,
+      "step": 3957
+    },
+    {
+      "epoch": 2.060385216033316,
+      "grad_norm": 0.23298375809694774,
+      "learning_rate": 1.1797060908581656e-05,
+      "loss": 0.052,
+      "step": 3958
+    },
+    {
+      "epoch": 2.0609057782404996,
+      "grad_norm": 0.24071234192220964,
+      "learning_rate": 1.1785132089919208e-05,
+      "loss": 0.0513,
+      "step": 3959
+    },
+    {
+      "epoch": 2.0614263404476834,
+      "grad_norm": 0.21300744847374067,
+      "learning_rate": 1.1773207445127005e-05,
+      "loss": 0.0488,
+      "step": 3960
+    },
+    {
+      "epoch": 2.061946902654867,
+      "grad_norm": 0.22968959003299735,
+      "learning_rate": 1.1761286977971412e-05,
+      "loss": 0.0515,
+      "step": 3961
+    },
+    {
+      "epoch": 2.062467464862051,
+      "grad_norm": 0.2293890367687117,
+      "learning_rate": 1.1749370692217465e-05,
+      "loss": 0.0525,
+      "step": 3962
+    },
+    {
+      "epoch": 2.0629880270692347,
+      "grad_norm": 0.22697976848422804,
+      "learning_rate": 1.1737458591628897e-05,
+      "loss": 0.053,
+      "step": 3963
+    },
+    {
+      "epoch": 2.0635085892764184,
+      "grad_norm": 0.23070259973942472,
+      "learning_rate": 1.1725550679968084e-05,
+      "loss": 0.0515,
+      "step": 3964
+    },
+    {
+      "epoch": 2.064029151483602,
+      "grad_norm": 0.21962051644788558,
+      "learning_rate": 1.1713646960996102e-05,
+      "loss": 0.0509,
+      "step": 3965
+    },
+    {
+      "epoch": 2.064549713690786,
+      "grad_norm": 0.23542235445056928,
+      "learning_rate": 1.1701747438472704e-05,
+      "loss": 0.0513,
+      "step": 3966
+    },
+    {
+      "epoch": 2.0650702758979698,
+      "grad_norm": 0.2249101062878514,
+      "learning_rate": 1.1689852116156313e-05,
+      "loss": 0.051,
+      "step": 3967
+    },
+    {
+      "epoch": 2.0655908381051535,
+      "grad_norm": 0.22365269003898253,
+      "learning_rate": 1.1677960997804047e-05,
+      "loss": 0.0519,
+      "step": 3968
+    },
+    {
+      "epoch": 2.0661114003123373,
+      "grad_norm": 0.23393573206151985,
+      "learning_rate": 1.1666074087171627e-05,
+      "loss": 0.0505,
+      "step": 3969
+    },
+    {
+      "epoch": 2.066631962519521,
+      "grad_norm": 0.22998763268787198,
+      "learning_rate": 1.1654191388013521e-05,
+      "loss": 0.0484,
+      "step": 3970
+    },
+    {
+      "epoch": 2.067152524726705,
+      "grad_norm": 0.2439217995498246,
+      "learning_rate": 1.1642312904082835e-05,
+      "loss": 0.0517,
+      "step": 3971
+    },
+    {
+      "epoch": 2.0676730869338886,
+      "grad_norm": 0.24085294418610648,
+      "learning_rate": 1.163043863913135e-05,
+      "loss": 0.0538,
+      "step": 3972
+    },
+    {
+      "epoch": 2.0681936491410724,
+      "grad_norm": 0.22810652141332463,
+      "learning_rate": 1.1618568596909512e-05,
+      "loss": 0.0488,
+      "step": 3973
+    },
+    {
+      "epoch": 2.068714211348256,
+      "grad_norm": 0.23657856963046003,
+      "learning_rate": 1.1606702781166423e-05,
+      "loss": 0.0501,
+      "step": 3974
+    },
+    {
+      "epoch": 2.06923477355544,
+      "grad_norm": 0.23158656572125352,
+      "learning_rate": 1.1594841195649866e-05,
+      "loss": 0.0516,
+      "step": 3975
+    },
+    {
+      "epoch": 2.0697553357626237,
+      "grad_norm": 0.23564568907963002,
+      "learning_rate": 1.1582983844106282e-05,
+      "loss": 0.0523,
+      "step": 3976
+    },
+    {
+      "epoch": 2.0702758979698075,
+      "grad_norm": 0.21940900341049915,
+      "learning_rate": 1.1571130730280786e-05,
+      "loss": 0.0505,
+      "step": 3977
+    },
+    {
+      "epoch": 2.0707964601769913,
+      "grad_norm": 0.23826470500495703,
+      "learning_rate": 1.1559281857917125e-05,
+      "loss": 0.0502,
+      "step": 3978
+    },
+    {
+      "epoch": 2.071317022384175,
+      "grad_norm": 0.22049292585636585,
+      "learning_rate": 1.1547437230757726e-05,
+      "loss": 0.0504,
+      "step": 3979
+    },
+    {
+      "epoch": 2.071837584591359,
+      "grad_norm": 0.23925268301002867,
+      "learning_rate": 1.1535596852543675e-05,
+      "loss": 0.0528,
+      "step": 3980
+    },
+    {
+      "epoch": 2.0723581467985426,
+      "grad_norm": 0.22626393109224524,
+      "learning_rate": 1.1523760727014721e-05,
+      "loss": 0.0521,
+      "step": 3981
+    },
+    {
+      "epoch": 2.0728787090057263,
+      "grad_norm": 0.23358296619905644,
+      "learning_rate": 1.1511928857909266e-05,
+      "loss": 0.0521,
+      "step": 3982
+    },
+    {
+      "epoch": 2.07339927121291,
+      "grad_norm": 0.22578239853084178,
+      "learning_rate": 1.1500101248964348e-05,
+      "loss": 0.049,
+      "step": 3983
+    },
+    {
+      "epoch": 2.073919833420094,
+      "grad_norm": 0.22302089050267151,
+      "learning_rate": 1.1488277903915689e-05,
+      "loss": 0.0496,
+      "step": 3984
+    },
+    {
+      "epoch": 2.0744403956272777,
+      "grad_norm": 0.22645035221882673,
+      "learning_rate": 1.147645882649765e-05,
+      "loss": 0.0511,
+      "step": 3985
+    },
+    {
+      "epoch": 2.0749609578344614,
+      "grad_norm": 0.23257285368266278,
+      "learning_rate": 1.1464644020443253e-05,
+      "loss": 0.0498,
+      "step": 3986
+    },
+    {
+      "epoch": 2.0754815200416448,
+      "grad_norm": 0.23394643638643065,
+      "learning_rate": 1.1452833489484155e-05,
+      "loss": 0.0508,
+      "step": 3987
+    },
+    {
+      "epoch": 2.0760020822488285,
+      "grad_norm": 0.2302396225330739,
+      "learning_rate": 1.1441027237350663e-05,
+      "loss": 0.0501,
+      "step": 3988
+    },
+    {
+      "epoch": 2.0765226444560123,
+      "grad_norm": 0.2364153136455663,
+      "learning_rate": 1.142922526777175e-05,
+      "loss": 0.0532,
+      "step": 3989
+    },
+    {
+      "epoch": 2.077043206663196,
+      "grad_norm": 0.24046454093029992,
+      "learning_rate": 1.1417427584475027e-05,
+      "loss": 0.0526,
+      "step": 3990
+    },
+    {
+      "epoch": 2.07756376887038,
+      "grad_norm": 0.23257494130614306,
+      "learning_rate": 1.1405634191186759e-05,
+      "loss": 0.0515,
+      "step": 3991
+    },
+    {
+      "epoch": 2.0780843310775636,
+      "grad_norm": 0.2327035472109291,
+      "learning_rate": 1.1393845091631833e-05,
+      "loss": 0.0497,
+      "step": 3992
+    },
+    {
+      "epoch": 2.0786048932847474,
+      "grad_norm": 0.23438697636486627,
+      "learning_rate": 1.1382060289533804e-05,
+      "loss": 0.0507,
+      "step": 3993
+    },
+    {
+      "epoch": 2.079125455491931,
+      "grad_norm": 0.23128247990194226,
+      "learning_rate": 1.1370279788614856e-05,
+      "loss": 0.0508,
+      "step": 3994
+    },
+    {
+      "epoch": 2.079646017699115,
+      "grad_norm": 0.23693622205792916,
+      "learning_rate": 1.1358503592595837e-05,
+      "loss": 0.0508,
+      "step": 3995
+    },
+    {
+      "epoch": 2.0801665799062987,
+      "grad_norm": 0.23110049467220117,
+      "learning_rate": 1.1346731705196204e-05,
+      "loss": 0.0501,
+      "step": 3996
+    },
+    {
+      "epoch": 2.0806871421134825,
+      "grad_norm": 0.24306539756134846,
+      "learning_rate": 1.1334964130134055e-05,
+      "loss": 0.0538,
+      "step": 3997
+    },
+    {
+      "epoch": 2.0812077043206663,
+      "grad_norm": 0.219892553239681,
+      "learning_rate": 1.132320087112615e-05,
+      "loss": 0.0478,
+      "step": 3998
+    },
+    {
+      "epoch": 2.08172826652785,
+      "grad_norm": 0.24126403657509093,
+      "learning_rate": 1.1311441931887873e-05,
+      "loss": 0.0519,
+      "step": 3999
+    },
+    {
+      "epoch": 2.082248828735034,
+      "grad_norm": 0.2248359123191256,
+      "learning_rate": 1.1299687316133256e-05,
+      "loss": 0.0497,
+      "step": 4000
+    },
+    {
+      "epoch": 2.0827693909422176,
+      "grad_norm": 0.24495968302747279,
+      "learning_rate": 1.1287937027574933e-05,
+      "loss": 0.0536,
+      "step": 4001
+    },
+    {
+      "epoch": 2.0832899531494014,
+      "grad_norm": 0.2280996896351997,
+      "learning_rate": 1.1276191069924197e-05,
+      "loss": 0.0516,
+      "step": 4002
+    },
+    {
+      "epoch": 2.083810515356585,
+      "grad_norm": 0.22983408181704557,
+      "learning_rate": 1.1264449446890975e-05,
+      "loss": 0.0497,
+      "step": 4003
+    },
+    {
+      "epoch": 2.084331077563769,
+      "grad_norm": 0.22345330078372413,
+      "learning_rate": 1.1252712162183825e-05,
+      "loss": 0.0481,
+      "step": 4004
+    },
+    {
+      "epoch": 2.0848516397709527,
+      "grad_norm": 0.22929949981841313,
+      "learning_rate": 1.1240979219509917e-05,
+      "loss": 0.0508,
+      "step": 4005
+    },
+    {
+      "epoch": 2.0853722019781364,
+      "grad_norm": 0.22837492034567583,
+      "learning_rate": 1.1229250622575052e-05,
+      "loss": 0.0494,
+      "step": 4006
+    },
+    {
+      "epoch": 2.08589276418532,
+      "grad_norm": 0.2343978470902539,
+      "learning_rate": 1.1217526375083675e-05,
+      "loss": 0.0499,
+      "step": 4007
+    },
+    {
+      "epoch": 2.086413326392504,
+      "grad_norm": 0.2336189431913341,
+      "learning_rate": 1.120580648073885e-05,
+      "loss": 0.0516,
+      "step": 4008
+    },
+    {
+      "epoch": 2.0869338885996878,
+      "grad_norm": 0.23282836656117412,
+      "learning_rate": 1.1194090943242278e-05,
+      "loss": 0.0503,
+      "step": 4009
+    },
+    {
+      "epoch": 2.0874544508068715,
+      "grad_norm": 0.24746886009644017,
+      "learning_rate": 1.1182379766294244e-05,
+      "loss": 0.0522,
+      "step": 4010
+    },
+    {
+      "epoch": 2.0879750130140553,
+      "grad_norm": 0.23120603188037314,
+      "learning_rate": 1.1170672953593696e-05,
+      "loss": 0.0501,
+      "step": 4011
+    },
+    {
+      "epoch": 2.088495575221239,
+      "grad_norm": 0.23047710533447974,
+      "learning_rate": 1.1158970508838193e-05,
+      "loss": 0.0482,
+      "step": 4012
+    },
+    {
+      "epoch": 2.089016137428423,
+      "grad_norm": 0.2524446742959235,
+      "learning_rate": 1.1147272435723918e-05,
+      "loss": 0.05,
+      "step": 4013
+    },
+    {
+      "epoch": 2.0895366996356066,
+      "grad_norm": 0.25459941682903553,
+      "learning_rate": 1.113557873794566e-05,
+      "loss": 0.0521,
+      "step": 4014
+    },
+    {
+      "epoch": 2.0900572618427904,
+      "grad_norm": 0.22652476025578092,
+      "learning_rate": 1.1123889419196821e-05,
+      "loss": 0.0514,
+      "step": 4015
+    },
+    {
+      "epoch": 2.090577824049974,
+      "grad_norm": 0.2288103168103935,
+      "learning_rate": 1.1112204483169439e-05,
+      "loss": 0.0506,
+      "step": 4016
+    },
+    {
+      "epoch": 2.091098386257158,
+      "grad_norm": 0.24150820992396882,
+      "learning_rate": 1.1100523933554166e-05,
+      "loss": 0.05,
+      "step": 4017
+    },
+    {
+      "epoch": 2.0916189484643413,
+      "grad_norm": 0.2322103337709988,
+      "learning_rate": 1.108884777404027e-05,
+      "loss": 0.0507,
+      "step": 4018
+    },
+    {
+      "epoch": 2.092139510671525,
+      "grad_norm": 0.22947198807908026,
+      "learning_rate": 1.1077176008315606e-05,
+      "loss": 0.0509,
+      "step": 4019
+    },
+    {
+      "epoch": 2.092660072878709,
+      "grad_norm": 0.23943146290614956,
+      "learning_rate": 1.1065508640066672e-05,
+      "loss": 0.0505,
+      "step": 4020
+    },
+    {
+      "epoch": 2.0931806350858926,
+      "grad_norm": 0.2487064925589514,
+      "learning_rate": 1.1053845672978567e-05,
+      "loss": 0.0535,
+      "step": 4021
+    },
+    {
+      "epoch": 2.0937011972930764,
+      "grad_norm": 0.23593594385346,
+      "learning_rate": 1.1042187110735e-05,
+      "loss": 0.0503,
+      "step": 4022
+    },
+    {
+      "epoch": 2.09422175950026,
+      "grad_norm": 0.24229015214508823,
+      "learning_rate": 1.1030532957018288e-05,
+      "loss": 0.0512,
+      "step": 4023
+    },
+    {
+      "epoch": 2.094742321707444,
+      "grad_norm": 0.2257929833420632,
+      "learning_rate": 1.1018883215509343e-05,
+      "loss": 0.0485,
+      "step": 4024
+    },
+    {
+      "epoch": 2.0952628839146277,
+      "grad_norm": 0.23317922327013857,
+      "learning_rate": 1.10072378898877e-05,
+      "loss": 0.0496,
+      "step": 4025
+    },
+    {
+      "epoch": 2.0957834461218114,
+      "grad_norm": 0.2375777735308176,
+      "learning_rate": 1.0995596983831502e-05,
+      "loss": 0.0512,
+      "step": 4026
+    },
+    {
+      "epoch": 2.096304008328995,
+      "grad_norm": 0.23856279889569232,
+      "learning_rate": 1.0983960501017492e-05,
+      "loss": 0.0495,
+      "step": 4027
+    },
+    {
+      "epoch": 2.096824570536179,
+      "grad_norm": 0.23551216040588757,
+      "learning_rate": 1.0972328445121e-05,
+      "loss": 0.0495,
+      "step": 4028
+    },
+    {
+      "epoch": 2.0973451327433628,
+      "grad_norm": 0.24501438351869415,
+      "learning_rate": 1.0960700819815973e-05,
+      "loss": 0.0491,
+      "step": 4029
+    },
+    {
+      "epoch": 2.0978656949505465,
+      "grad_norm": 0.23782057609972138,
+      "learning_rate": 1.0949077628774961e-05,
+      "loss": 0.0488,
+      "step": 4030
+    },
+    {
+      "epoch": 2.0983862571577303,
+      "grad_norm": 0.23261361096280053,
+      "learning_rate": 1.0937458875669119e-05,
+      "loss": 0.0513,
+      "step": 4031
+    },
+    {
+      "epoch": 2.098906819364914,
+      "grad_norm": 0.24475133620833012,
+      "learning_rate": 1.0925844564168175e-05,
+      "loss": 0.05,
+      "step": 4032
+    },
+    {
+      "epoch": 2.099427381572098,
+      "grad_norm": 0.28758926057418177,
+      "learning_rate": 1.0914234697940465e-05,
+      "loss": 0.0496,
+      "step": 4033
+    },
+    {
+      "epoch": 2.0999479437792816,
+      "grad_norm": 0.22436454655048274,
+      "learning_rate": 1.0902629280652931e-05,
+      "loss": 0.0506,
+      "step": 4034
+    },
+    {
+      "epoch": 2.1004685059864654,
+      "grad_norm": 0.234024748448541,
+      "learning_rate": 1.0891028315971105e-05,
+      "loss": 0.0485,
+      "step": 4035
+    },
+    {
+      "epoch": 2.100989068193649,
+      "grad_norm": 0.22684434792800956,
+      "learning_rate": 1.0879431807559116e-05,
+      "loss": 0.0463,
+      "step": 4036
+    },
+    {
+      "epoch": 2.101509630400833,
+      "grad_norm": 0.23052518036453548,
+      "learning_rate": 1.0867839759079682e-05,
+      "loss": 0.0489,
+      "step": 4037
+    },
+    {
+      "epoch": 2.1020301926080167,
+      "grad_norm": 0.23554676365556773,
+      "learning_rate": 1.0856252174194096e-05,
+      "loss": 0.0501,
+      "step": 4038
+    },
+    {
+      "epoch": 2.1025507548152005,
+      "grad_norm": 0.2364179349009896,
+      "learning_rate": 1.084466905656227e-05,
+      "loss": 0.0486,
+      "step": 4039
+    },
+    {
+      "epoch": 2.1030713170223843,
+      "grad_norm": 0.2315928309233159,
+      "learning_rate": 1.0833090409842694e-05,
+      "loss": 0.0485,
+      "step": 4040
+    },
+    {
+      "epoch": 2.103591879229568,
+      "grad_norm": 0.24327580350563419,
+      "learning_rate": 1.0821516237692434e-05,
+      "loss": 0.0524,
+      "step": 4041
+    },
+    {
+      "epoch": 2.104112441436752,
+      "grad_norm": 0.22953852222832158,
+      "learning_rate": 1.080994654376716e-05,
+      "loss": 0.0475,
+      "step": 4042
+    },
+    {
+      "epoch": 2.1046330036439356,
+      "grad_norm": 0.23269426088814224,
+      "learning_rate": 1.0798381331721109e-05,
+      "loss": 0.0491,
+      "step": 4043
+    },
+    {
+      "epoch": 2.1051535658511193,
+      "grad_norm": 0.2393902668332969,
+      "learning_rate": 1.0786820605207117e-05,
+      "loss": 0.0496,
+      "step": 4044
+    },
+    {
+      "epoch": 2.105674128058303,
+      "grad_norm": 0.23157080328354862,
+      "learning_rate": 1.0775264367876605e-05,
+      "loss": 0.0509,
+      "step": 4045
+    },
+    {
+      "epoch": 2.106194690265487,
+      "grad_norm": 0.24639805362000278,
+      "learning_rate": 1.0763712623379577e-05,
+      "loss": 0.0517,
+      "step": 4046
+    },
+    {
+      "epoch": 2.1067152524726707,
+      "grad_norm": 0.25130598484708905,
+      "learning_rate": 1.0752165375364593e-05,
+      "loss": 0.0528,
+      "step": 4047
+    },
+    {
+      "epoch": 2.1072358146798544,
+      "grad_norm": 0.22906779523998783,
+      "learning_rate": 1.0740622627478821e-05,
+      "loss": 0.0505,
+      "step": 4048
+    },
+    {
+      "epoch": 2.107756376887038,
+      "grad_norm": 0.22056035942244043,
+      "learning_rate": 1.0729084383368005e-05,
+      "loss": 0.0489,
+      "step": 4049
+    },
+    {
+      "epoch": 2.108276939094222,
+      "grad_norm": 0.24068696987121577,
+      "learning_rate": 1.0717550646676443e-05,
+      "loss": 0.0508,
+      "step": 4050
+    },
+    {
+      "epoch": 2.1087975013014053,
+      "grad_norm": 0.22714859742711338,
+      "learning_rate": 1.0706021421047047e-05,
+      "loss": 0.0506,
+      "step": 4051
+    },
+    {
+      "epoch": 2.109318063508589,
+      "grad_norm": 0.22747067082854852,
+      "learning_rate": 1.0694496710121257e-05,
+      "loss": 0.0491,
+      "step": 4052
+    },
+    {
+      "epoch": 2.109838625715773,
+      "grad_norm": 0.23375027504810572,
+      "learning_rate": 1.0682976517539128e-05,
+      "loss": 0.05,
+      "step": 4053
+    },
+    {
+      "epoch": 2.1103591879229566,
+      "grad_norm": 0.22132944162068763,
+      "learning_rate": 1.0671460846939274e-05,
+      "loss": 0.0482,
+      "step": 4054
+    },
+    {
+      "epoch": 2.1108797501301404,
+      "grad_norm": 0.22882347387684723,
+      "learning_rate": 1.0659949701958885e-05,
+      "loss": 0.0499,
+      "step": 4055
+    },
+    {
+      "epoch": 2.111400312337324,
+      "grad_norm": 0.2312263646586655,
+      "learning_rate": 1.0648443086233697e-05,
+      "loss": 0.0511,
+      "step": 4056
+    },
+    {
+      "epoch": 2.111920874544508,
+      "grad_norm": 0.24220776541349695,
+      "learning_rate": 1.063694100339805e-05,
+      "loss": 0.0516,
+      "step": 4057
+    },
+    {
+      "epoch": 2.1124414367516917,
+      "grad_norm": 0.22625059430843475,
+      "learning_rate": 1.0625443457084841e-05,
+      "loss": 0.0494,
+      "step": 4058
+    },
+    {
+      "epoch": 2.1129619989588755,
+      "grad_norm": 0.23299358120448602,
+      "learning_rate": 1.0613950450925513e-05,
+      "loss": 0.0495,
+      "step": 4059
+    },
+    {
+      "epoch": 2.1134825611660593,
+      "grad_norm": 0.23342314525886373,
+      "learning_rate": 1.060246198855011e-05,
+      "loss": 0.0494,
+      "step": 4060
+    },
+    {
+      "epoch": 2.114003123373243,
+      "grad_norm": 0.2421280557624441,
+      "learning_rate": 1.059097807358721e-05,
+      "loss": 0.0503,
+      "step": 4061
+    },
+    {
+      "epoch": 2.114523685580427,
+      "grad_norm": 0.22689249003133796,
+      "learning_rate": 1.0579498709663968e-05,
+      "loss": 0.0479,
+      "step": 4062
+    },
+    {
+      "epoch": 2.1150442477876106,
+      "grad_norm": 0.2392848913272859,
+      "learning_rate": 1.0568023900406107e-05,
+      "loss": 0.052,
+      "step": 4063
+    },
+    {
+      "epoch": 2.1155648099947943,
+      "grad_norm": 0.25646230307725143,
+      "learning_rate": 1.0556553649437914e-05,
+      "loss": 0.0513,
+      "step": 4064
+    },
+    {
+      "epoch": 2.116085372201978,
+      "grad_norm": 0.2416279360872608,
+      "learning_rate": 1.0545087960382211e-05,
+      "loss": 0.0515,
+      "step": 4065
+    },
+    {
+      "epoch": 2.116605934409162,
+      "grad_norm": 0.2231969043709799,
+      "learning_rate": 1.05336268368604e-05,
+      "loss": 0.0491,
+      "step": 4066
+    },
+    {
+      "epoch": 2.1171264966163457,
+      "grad_norm": 0.2425556380715717,
+      "learning_rate": 1.0522170282492444e-05,
+      "loss": 0.0495,
+      "step": 4067
+    },
+    {
+      "epoch": 2.1176470588235294,
+      "grad_norm": 0.2383579552897591,
+      "learning_rate": 1.051071830089686e-05,
+      "loss": 0.0517,
+      "step": 4068
+    },
+    {
+      "epoch": 2.118167621030713,
+      "grad_norm": 0.24083746604101072,
+      "learning_rate": 1.049927089569071e-05,
+      "loss": 0.0514,
+      "step": 4069
+    },
+    {
+      "epoch": 2.118688183237897,
+      "grad_norm": 0.23603705033415373,
+      "learning_rate": 1.048782807048961e-05,
+      "loss": 0.0509,
+      "step": 4070
+    },
+    {
+      "epoch": 2.1192087454450808,
+      "grad_norm": 0.22699841700025328,
+      "learning_rate": 1.0476389828907743e-05,
+      "loss": 0.0507,
+      "step": 4071
+    },
+    {
+      "epoch": 2.1197293076522645,
+      "grad_norm": 0.2384365332964614,
+      "learning_rate": 1.046495617455784e-05,
+      "loss": 0.0506,
+      "step": 4072
+    },
+    {
+      "epoch": 2.1202498698594483,
+      "grad_norm": 0.23686492769560208,
+      "learning_rate": 1.0453527111051184e-05,
+      "loss": 0.0499,
+      "step": 4073
+    },
+    {
+      "epoch": 2.120770432066632,
+      "grad_norm": 0.23537377480311067,
+      "learning_rate": 1.0442102641997594e-05,
+      "loss": 0.0495,
+      "step": 4074
+    },
+    {
+      "epoch": 2.121290994273816,
+      "grad_norm": 0.23207116343425835,
+      "learning_rate": 1.0430682771005456e-05,
+      "loss": 0.0525,
+      "step": 4075
+    },
+    {
+      "epoch": 2.1218115564809996,
+      "grad_norm": 0.23332099551170052,
+      "learning_rate": 1.0419267501681699e-05,
+      "loss": 0.0498,
+      "step": 4076
+    },
+    {
+      "epoch": 2.1223321186881834,
+      "grad_norm": 0.23684848540512882,
+      "learning_rate": 1.0407856837631798e-05,
+      "loss": 0.0521,
+      "step": 4077
+    },
+    {
+      "epoch": 2.122852680895367,
+      "grad_norm": 0.22933556276740186,
+      "learning_rate": 1.0396450782459771e-05,
+      "loss": 0.0497,
+      "step": 4078
+    },
+    {
+      "epoch": 2.123373243102551,
+      "grad_norm": 0.2288593333175558,
+      "learning_rate": 1.0385049339768168e-05,
+      "loss": 0.0506,
+      "step": 4079
+    },
+    {
+      "epoch": 2.1238938053097347,
+      "grad_norm": 0.23192903613988014,
+      "learning_rate": 1.0373652513158105e-05,
+      "loss": 0.0513,
+      "step": 4080
+    },
+    {
+      "epoch": 2.1244143675169185,
+      "grad_norm": 0.23148424092687478,
+      "learning_rate": 1.0362260306229229e-05,
+      "loss": 0.0517,
+      "step": 4081
+    },
+    {
+      "epoch": 2.124934929724102,
+      "grad_norm": 0.23760449025559408,
+      "learning_rate": 1.0350872722579741e-05,
+      "loss": 0.0506,
+      "step": 4082
+    },
+    {
+      "epoch": 2.1254554919312856,
+      "grad_norm": 0.23072320281257291,
+      "learning_rate": 1.0339489765806345e-05,
+      "loss": 0.0519,
+      "step": 4083
+    },
+    {
+      "epoch": 2.1259760541384694,
+      "grad_norm": 0.2273989358106142,
+      "learning_rate": 1.032811143950433e-05,
+      "loss": 0.0492,
+      "step": 4084
+    },
+    {
+      "epoch": 2.126496616345653,
+      "grad_norm": 0.2286725964642402,
+      "learning_rate": 1.0316737747267486e-05,
+      "loss": 0.0508,
+      "step": 4085
+    },
+    {
+      "epoch": 2.127017178552837,
+      "grad_norm": 0.2403403105915755,
+      "learning_rate": 1.0305368692688174e-05,
+      "loss": 0.0498,
+      "step": 4086
+    },
+    {
+      "epoch": 2.1275377407600207,
+      "grad_norm": 0.22835020945257542,
+      "learning_rate": 1.0294004279357259e-05,
+      "loss": 0.0494,
+      "step": 4087
+    },
+    {
+      "epoch": 2.1280583029672044,
+      "grad_norm": 0.23886208707779596,
+      "learning_rate": 1.028264451086414e-05,
+      "loss": 0.05,
+      "step": 4088
+    },
+    {
+      "epoch": 2.128578865174388,
+      "grad_norm": 0.23621522502410017,
+      "learning_rate": 1.0271289390796771e-05,
+      "loss": 0.0495,
+      "step": 4089
+    },
+    {
+      "epoch": 2.129099427381572,
+      "grad_norm": 0.23986277451442098,
+      "learning_rate": 1.0259938922741627e-05,
+      "loss": 0.0505,
+      "step": 4090
+    },
+    {
+      "epoch": 2.1296199895887558,
+      "grad_norm": 0.2306221065006331,
+      "learning_rate": 1.0248593110283725e-05,
+      "loss": 0.0503,
+      "step": 4091
+    },
+    {
+      "epoch": 2.1301405517959395,
+      "grad_norm": 0.23069214444668543,
+      "learning_rate": 1.023725195700658e-05,
+      "loss": 0.0497,
+      "step": 4092
+    },
+    {
+      "epoch": 2.1306611140031233,
+      "grad_norm": 0.22731226780333583,
+      "learning_rate": 1.0225915466492267e-05,
+      "loss": 0.0492,
+      "step": 4093
+    },
+    {
+      "epoch": 2.131181676210307,
+      "grad_norm": 0.23439529582276133,
+      "learning_rate": 1.0214583642321376e-05,
+      "loss": 0.0499,
+      "step": 4094
+    },
+    {
+      "epoch": 2.131702238417491,
+      "grad_norm": 0.2390100567918747,
+      "learning_rate": 1.0203256488073034e-05,
+      "loss": 0.0498,
+      "step": 4095
+    },
+    {
+      "epoch": 2.1322228006246746,
+      "grad_norm": 0.22979409308662924,
+      "learning_rate": 1.0191934007324874e-05,
+      "loss": 0.0503,
+      "step": 4096
+    },
+    {
+      "epoch": 2.1327433628318584,
+      "grad_norm": 0.23224468806672807,
+      "learning_rate": 1.0180616203653054e-05,
+      "loss": 0.0498,
+      "step": 4097
+    },
+    {
+      "epoch": 2.133263925039042,
+      "grad_norm": 0.2310744812755954,
+      "learning_rate": 1.0169303080632272e-05,
+      "loss": 0.05,
+      "step": 4098
+    },
+    {
+      "epoch": 2.133784487246226,
+      "grad_norm": 0.2266485347614057,
+      "learning_rate": 1.0157994641835736e-05,
+      "loss": 0.05,
+      "step": 4099
+    },
+    {
+      "epoch": 2.1343050494534097,
+      "grad_norm": 0.231117064885941,
+      "learning_rate": 1.014669089083518e-05,
+      "loss": 0.0484,
+      "step": 4100
+    },
+    {
+      "epoch": 2.1348256116605935,
+      "grad_norm": 0.23246610511746577,
+      "learning_rate": 1.0135391831200866e-05,
+      "loss": 0.0509,
+      "step": 4101
+    },
+    {
+      "epoch": 2.1353461738677773,
+      "grad_norm": 0.24212771834516583,
+      "learning_rate": 1.012409746650154e-05,
+      "loss": 0.053,
+      "step": 4102
+    },
+    {
+      "epoch": 2.135866736074961,
+      "grad_norm": 0.23361341416282064,
+      "learning_rate": 1.01128078003045e-05,
+      "loss": 0.0482,
+      "step": 4103
+    },
+    {
+      "epoch": 2.136387298282145,
+      "grad_norm": 0.23836572088407046,
+      "learning_rate": 1.0101522836175555e-05,
+      "loss": 0.0504,
+      "step": 4104
+    },
+    {
+      "epoch": 2.1369078604893286,
+      "grad_norm": 0.2259664441526069,
+      "learning_rate": 1.0090242577679005e-05,
+      "loss": 0.0495,
+      "step": 4105
+    },
+    {
+      "epoch": 2.1374284226965123,
+      "grad_norm": 0.22082689150726414,
+      "learning_rate": 1.0078967028377697e-05,
+      "loss": 0.049,
+      "step": 4106
+    },
+    {
+      "epoch": 2.137948984903696,
+      "grad_norm": 0.2342587477443236,
+      "learning_rate": 1.006769619183296e-05,
+      "loss": 0.0502,
+      "step": 4107
+    },
+    {
+      "epoch": 2.13846954711088,
+      "grad_norm": 0.22943741247932514,
+      "learning_rate": 1.0056430071604653e-05,
+      "loss": 0.0487,
+      "step": 4108
+    },
+    {
+      "epoch": 2.1389901093180637,
+      "grad_norm": 0.22043167610871012,
+      "learning_rate": 1.0045168671251143e-05,
+      "loss": 0.049,
+      "step": 4109
+    },
+    {
+      "epoch": 2.1395106715252474,
+      "grad_norm": 0.2405489283322427,
+      "learning_rate": 1.0033911994329314e-05,
+      "loss": 0.051,
+      "step": 4110
+    },
+    {
+      "epoch": 2.140031233732431,
+      "grad_norm": 0.234193224915324,
+      "learning_rate": 1.0022660044394534e-05,
+      "loss": 0.0483,
+      "step": 4111
+    },
+    {
+      "epoch": 2.140551795939615,
+      "grad_norm": 0.22465165622300198,
+      "learning_rate": 1.0011412825000694e-05,
+      "loss": 0.0478,
+      "step": 4112
+    },
+    {
+      "epoch": 2.1410723581467987,
+      "grad_norm": 0.23278809432046188,
+      "learning_rate": 1.00001703397002e-05,
+      "loss": 0.0479,
+      "step": 4113
+    },
+    {
+      "epoch": 2.1415929203539825,
+      "grad_norm": 0.23156952891318974,
+      "learning_rate": 9.988932592043937e-06,
+      "loss": 0.048,
+      "step": 4114
+    },
+    {
+      "epoch": 2.1421134825611663,
+      "grad_norm": 0.24878058256395674,
+      "learning_rate": 9.977699585581324e-06,
+      "loss": 0.0529,
+      "step": 4115
+    },
+    {
+      "epoch": 2.1426340447683496,
+      "grad_norm": 0.2401759125829604,
+      "learning_rate": 9.966471323860251e-06,
+      "loss": 0.0507,
+      "step": 4116
+    },
+    {
+      "epoch": 2.1431546069755334,
+      "grad_norm": 0.2318433491416454,
+      "learning_rate": 9.955247810427137e-06,
+      "loss": 0.0506,
+      "step": 4117
+    },
+    {
+      "epoch": 2.143675169182717,
+      "grad_norm": 0.23626591123934843,
+      "learning_rate": 9.944029048826887e-06,
+      "loss": 0.0498,
+      "step": 4118
+    },
+    {
+      "epoch": 2.144195731389901,
+      "grad_norm": 0.23359239052440017,
+      "learning_rate": 9.932815042602913e-06,
+      "loss": 0.053,
+      "step": 4119
+    },
+    {
+      "epoch": 2.1447162935970847,
+      "grad_norm": 0.24007159621963173,
+      "learning_rate": 9.921605795297109e-06,
+      "loss": 0.0495,
+      "step": 4120
+    },
+    {
+      "epoch": 2.1452368558042685,
+      "grad_norm": 0.23718809404737304,
+      "learning_rate": 9.910401310449883e-06,
+      "loss": 0.0497,
+      "step": 4121
+    },
+    {
+      "epoch": 2.1457574180114523,
+      "grad_norm": 0.23382317156311533,
+      "learning_rate": 9.899201591600138e-06,
+      "loss": 0.0487,
+      "step": 4122
+    },
+    {
+      "epoch": 2.146277980218636,
+      "grad_norm": 0.23007554978616346,
+      "learning_rate": 9.888006642285255e-06,
+      "loss": 0.0514,
+      "step": 4123
+    },
+    {
+      "epoch": 2.14679854242582,
+      "grad_norm": 0.2360311167478857,
+      "learning_rate": 9.876816466041133e-06,
+      "loss": 0.0514,
+      "step": 4124
+    },
+    {
+      "epoch": 2.1473191046330036,
+      "grad_norm": 0.23498144817531677,
+      "learning_rate": 9.865631066402137e-06,
+      "loss": 0.0528,
+      "step": 4125
+    },
+    {
+      "epoch": 2.1478396668401873,
+      "grad_norm": 0.22215864699197535,
+      "learning_rate": 9.854450446901143e-06,
+      "loss": 0.0463,
+      "step": 4126
+    },
+    {
+      "epoch": 2.148360229047371,
+      "grad_norm": 0.22531836871998293,
+      "learning_rate": 9.843274611069509e-06,
+      "loss": 0.0472,
+      "step": 4127
+    },
+    {
+      "epoch": 2.148880791254555,
+      "grad_norm": 0.24034863053588068,
+      "learning_rate": 9.832103562437096e-06,
+      "loss": 0.0514,
+      "step": 4128
+    },
+    {
+      "epoch": 2.1494013534617387,
+      "grad_norm": 0.2526527336235642,
+      "learning_rate": 9.820937304532221e-06,
+      "loss": 0.0511,
+      "step": 4129
+    },
+    {
+      "epoch": 2.1499219156689224,
+      "grad_norm": 0.23965378370422113,
+      "learning_rate": 9.809775840881718e-06,
+      "loss": 0.0493,
+      "step": 4130
+    },
+    {
+      "epoch": 2.150442477876106,
+      "grad_norm": 0.23284542516117132,
+      "learning_rate": 9.798619175010907e-06,
+      "loss": 0.0486,
+      "step": 4131
+    },
+    {
+      "epoch": 2.15096304008329,
+      "grad_norm": 0.2517441778747228,
+      "learning_rate": 9.787467310443562e-06,
+      "loss": 0.0518,
+      "step": 4132
+    },
+    {
+      "epoch": 2.1514836022904738,
+      "grad_norm": 0.24113620179268663,
+      "learning_rate": 9.77632025070198e-06,
+      "loss": 0.0487,
+      "step": 4133
+    },
+    {
+      "epoch": 2.1520041644976575,
+      "grad_norm": 0.23483751096204794,
+      "learning_rate": 9.765177999306904e-06,
+      "loss": 0.05,
+      "step": 4134
+    },
+    {
+      "epoch": 2.1525247267048413,
+      "grad_norm": 0.2372927103941721,
+      "learning_rate": 9.754040559777583e-06,
+      "loss": 0.0498,
+      "step": 4135
+    },
+    {
+      "epoch": 2.153045288912025,
+      "grad_norm": 0.23825778506276546,
+      "learning_rate": 9.742907935631737e-06,
+      "loss": 0.049,
+      "step": 4136
+    },
+    {
+      "epoch": 2.153565851119209,
+      "grad_norm": 0.24176693832701435,
+      "learning_rate": 9.731780130385578e-06,
+      "loss": 0.0497,
+      "step": 4137
+    },
+    {
+      "epoch": 2.1540864133263926,
+      "grad_norm": 0.2335246650469084,
+      "learning_rate": 9.720657147553769e-06,
+      "loss": 0.05,
+      "step": 4138
+    },
+    {
+      "epoch": 2.1546069755335764,
+      "grad_norm": 0.23861089006529668,
+      "learning_rate": 9.709538990649472e-06,
+      "loss": 0.0503,
+      "step": 4139
+    },
+    {
+      "epoch": 2.15512753774076,
+      "grad_norm": 0.2291209928615605,
+      "learning_rate": 9.698425663184324e-06,
+      "loss": 0.0485,
+      "step": 4140
+    },
+    {
+      "epoch": 2.155648099947944,
+      "grad_norm": 0.2342784317879674,
+      "learning_rate": 9.68731716866842e-06,
+      "loss": 0.049,
+      "step": 4141
+    },
+    {
+      "epoch": 2.1561686621551277,
+      "grad_norm": 0.2320214212518064,
+      "learning_rate": 9.676213510610352e-06,
+      "loss": 0.0485,
+      "step": 4142
+    },
+    {
+      "epoch": 2.1566892243623115,
+      "grad_norm": 0.2377587377352587,
+      "learning_rate": 9.665114692517158e-06,
+      "loss": 0.0523,
+      "step": 4143
+    },
+    {
+      "epoch": 2.1572097865694952,
+      "grad_norm": 0.2348296445693297,
+      "learning_rate": 9.654020717894366e-06,
+      "loss": 0.0502,
+      "step": 4144
+    },
+    {
+      "epoch": 2.157730348776679,
+      "grad_norm": 0.21794445970232668,
+      "learning_rate": 9.642931590245973e-06,
+      "loss": 0.0474,
+      "step": 4145
+    },
+    {
+      "epoch": 2.1582509109838623,
+      "grad_norm": 0.2350475368286949,
+      "learning_rate": 9.63184731307445e-06,
+      "loss": 0.0502,
+      "step": 4146
+    },
+    {
+      "epoch": 2.158771473191046,
+      "grad_norm": 0.23095612217802725,
+      "learning_rate": 9.620767889880708e-06,
+      "loss": 0.0483,
+      "step": 4147
+    },
+    {
+      "epoch": 2.15929203539823,
+      "grad_norm": 0.21882801219665537,
+      "learning_rate": 9.609693324164154e-06,
+      "loss": 0.0476,
+      "step": 4148
+    },
+    {
+      "epoch": 2.1598125976054137,
+      "grad_norm": 0.23959480333572203,
+      "learning_rate": 9.59862361942266e-06,
+      "loss": 0.0499,
+      "step": 4149
+    },
+    {
+      "epoch": 2.1603331598125974,
+      "grad_norm": 0.2382618434451453,
+      "learning_rate": 9.587558779152536e-06,
+      "loss": 0.0483,
+      "step": 4150
+    },
+    {
+      "epoch": 2.160853722019781,
+      "grad_norm": 0.2481562495202432,
+      "learning_rate": 9.576498806848591e-06,
+      "loss": 0.0515,
+      "step": 4151
+    },
+    {
+      "epoch": 2.161374284226965,
+      "grad_norm": 0.23415717741022837,
+      "learning_rate": 9.565443706004065e-06,
+      "loss": 0.0506,
+      "step": 4152
+    },
+    {
+      "epoch": 2.1618948464341488,
+      "grad_norm": 0.2310272602868718,
+      "learning_rate": 9.554393480110677e-06,
+      "loss": 0.0501,
+      "step": 4153
+    },
+    {
+      "epoch": 2.1624154086413325,
+      "grad_norm": 0.21717437554416877,
+      "learning_rate": 9.543348132658602e-06,
+      "loss": 0.0471,
+      "step": 4154
+    },
+    {
+      "epoch": 2.1629359708485163,
+      "grad_norm": 0.24030597109135618,
+      "learning_rate": 9.532307667136493e-06,
+      "loss": 0.0503,
+      "step": 4155
+    },
+    {
+      "epoch": 2.1634565330557,
+      "grad_norm": 0.2271124234150207,
+      "learning_rate": 9.521272087031414e-06,
+      "loss": 0.0474,
+      "step": 4156
+    },
+    {
+      "epoch": 2.163977095262884,
+      "grad_norm": 0.23172566333461286,
+      "learning_rate": 9.510241395828926e-06,
+      "loss": 0.0509,
+      "step": 4157
+    },
+    {
+      "epoch": 2.1644976574700676,
+      "grad_norm": 0.23463565756110716,
+      "learning_rate": 9.499215597013048e-06,
+      "loss": 0.0519,
+      "step": 4158
+    },
+    {
+      "epoch": 2.1650182196772514,
+      "grad_norm": 0.2269853064996004,
+      "learning_rate": 9.488194694066219e-06,
+      "loss": 0.0503,
+      "step": 4159
+    },
+    {
+      "epoch": 2.165538781884435,
+      "grad_norm": 0.22448361065078423,
+      "learning_rate": 9.47717869046937e-06,
+      "loss": 0.0488,
+      "step": 4160
+    },
+    {
+      "epoch": 2.166059344091619,
+      "grad_norm": 0.2298514916947985,
+      "learning_rate": 9.466167589701855e-06,
+      "loss": 0.0492,
+      "step": 4161
+    },
+    {
+      "epoch": 2.1665799062988027,
+      "grad_norm": 0.2411385315001518,
+      "learning_rate": 9.455161395241496e-06,
+      "loss": 0.0514,
+      "step": 4162
+    },
+    {
+      "epoch": 2.1671004685059865,
+      "grad_norm": 0.23344249520016908,
+      "learning_rate": 9.444160110564562e-06,
+      "loss": 0.0507,
+      "step": 4163
+    },
+    {
+      "epoch": 2.1676210307131702,
+      "grad_norm": 0.22970482297994116,
+      "learning_rate": 9.433163739145773e-06,
+      "loss": 0.049,
+      "step": 4164
+    },
+    {
+      "epoch": 2.168141592920354,
+      "grad_norm": 0.24137131622878738,
+      "learning_rate": 9.422172284458303e-06,
+      "loss": 0.0501,
+      "step": 4165
+    },
+    {
+      "epoch": 2.168662155127538,
+      "grad_norm": 0.2391941048873966,
+      "learning_rate": 9.411185749973744e-06,
+      "loss": 0.0491,
+      "step": 4166
+    },
+    {
+      "epoch": 2.1691827173347216,
+      "grad_norm": 0.22597120615731292,
+      "learning_rate": 9.400204139162178e-06,
+      "loss": 0.0478,
+      "step": 4167
+    },
+    {
+      "epoch": 2.1697032795419053,
+      "grad_norm": 0.22407522799206378,
+      "learning_rate": 9.389227455492083e-06,
+      "loss": 0.0488,
+      "step": 4168
+    },
+    {
+      "epoch": 2.170223841749089,
+      "grad_norm": 0.23341996865358303,
+      "learning_rate": 9.378255702430425e-06,
+      "loss": 0.0495,
+      "step": 4169
+    },
+    {
+      "epoch": 2.170744403956273,
+      "grad_norm": 0.22756065098090528,
+      "learning_rate": 9.367288883442596e-06,
+      "loss": 0.0465,
+      "step": 4170
+    },
+    {
+      "epoch": 2.1712649661634567,
+      "grad_norm": 0.24032557476752722,
+      "learning_rate": 9.356327001992412e-06,
+      "loss": 0.0514,
+      "step": 4171
+    },
+    {
+      "epoch": 2.1717855283706404,
+      "grad_norm": 0.23279773656960784,
+      "learning_rate": 9.345370061542158e-06,
+      "loss": 0.049,
+      "step": 4172
+    },
+    {
+      "epoch": 2.172306090577824,
+      "grad_norm": 0.227184511533179,
+      "learning_rate": 9.334418065552538e-06,
+      "loss": 0.048,
+      "step": 4173
+    },
+    {
+      "epoch": 2.172826652785008,
+      "grad_norm": 0.2388318118994056,
+      "learning_rate": 9.323471017482718e-06,
+      "loss": 0.0492,
+      "step": 4174
+    },
+    {
+      "epoch": 2.1733472149921917,
+      "grad_norm": 0.23862458637474987,
+      "learning_rate": 9.312528920790265e-06,
+      "loss": 0.0507,
+      "step": 4175
+    },
+    {
+      "epoch": 2.1738677771993755,
+      "grad_norm": 0.2300694111237474,
+      "learning_rate": 9.301591778931218e-06,
+      "loss": 0.0496,
+      "step": 4176
+    },
+    {
+      "epoch": 2.1743883394065593,
+      "grad_norm": 0.22487725179561208,
+      "learning_rate": 9.290659595360018e-06,
+      "loss": 0.0474,
+      "step": 4177
+    },
+    {
+      "epoch": 2.174908901613743,
+      "grad_norm": 0.24227495997845988,
+      "learning_rate": 9.27973237352957e-06,
+      "loss": 0.0484,
+      "step": 4178
+    },
+    {
+      "epoch": 2.175429463820927,
+      "grad_norm": 0.24971312513542707,
+      "learning_rate": 9.268810116891205e-06,
+      "loss": 0.0522,
+      "step": 4179
+    },
+    {
+      "epoch": 2.17595002602811,
+      "grad_norm": 0.23240074321999316,
+      "learning_rate": 9.257892828894663e-06,
+      "loss": 0.0509,
+      "step": 4180
+    },
+    {
+      "epoch": 2.176470588235294,
+      "grad_norm": 0.23411180744875743,
+      "learning_rate": 9.24698051298814e-06,
+      "loss": 0.0507,
+      "step": 4181
+    },
+    {
+      "epoch": 2.1769911504424777,
+      "grad_norm": 0.2283915870645207,
+      "learning_rate": 9.236073172618254e-06,
+      "loss": 0.0496,
+      "step": 4182
+    },
+    {
+      "epoch": 2.1775117126496615,
+      "grad_norm": 0.21862188186366638,
+      "learning_rate": 9.225170811230058e-06,
+      "loss": 0.0488,
+      "step": 4183
+    },
+    {
+      "epoch": 2.1780322748568453,
+      "grad_norm": 0.25870559139681704,
+      "learning_rate": 9.214273432267009e-06,
+      "loss": 0.0491,
+      "step": 4184
+    },
+    {
+      "epoch": 2.178552837064029,
+      "grad_norm": 0.2338660696928279,
+      "learning_rate": 9.203381039171022e-06,
+      "loss": 0.0488,
+      "step": 4185
+    },
+    {
+      "epoch": 2.179073399271213,
+      "grad_norm": 0.23653181541138485,
+      "learning_rate": 9.192493635382407e-06,
+      "loss": 0.0482,
+      "step": 4186
+    },
+    {
+      "epoch": 2.1795939614783966,
+      "grad_norm": 0.2475444397957008,
+      "learning_rate": 9.181611224339917e-06,
+      "loss": 0.051,
+      "step": 4187
+    },
+    {
+      "epoch": 2.1801145236855803,
+      "grad_norm": 0.23144206027749797,
+      "learning_rate": 9.170733809480738e-06,
+      "loss": 0.0478,
+      "step": 4188
+    },
+    {
+      "epoch": 2.180635085892764,
+      "grad_norm": 0.23938498072235687,
+      "learning_rate": 9.159861394240444e-06,
+      "loss": 0.0494,
+      "step": 4189
+    },
+    {
+      "epoch": 2.181155648099948,
+      "grad_norm": 0.24401500298605283,
+      "learning_rate": 9.148993982053058e-06,
+      "loss": 0.05,
+      "step": 4190
+    },
+    {
+      "epoch": 2.1816762103071317,
+      "grad_norm": 0.23262393333357667,
+      "learning_rate": 9.138131576351014e-06,
+      "loss": 0.0487,
+      "step": 4191
+    },
+    {
+      "epoch": 2.1821967725143154,
+      "grad_norm": 0.23923195938864855,
+      "learning_rate": 9.127274180565177e-06,
+      "loss": 0.0497,
+      "step": 4192
+    },
+    {
+      "epoch": 2.182717334721499,
+      "grad_norm": 0.2385711033015507,
+      "learning_rate": 9.116421798124794e-06,
+      "loss": 0.0508,
+      "step": 4193
+    },
+    {
+      "epoch": 2.183237896928683,
+      "grad_norm": 0.2305410960253734,
+      "learning_rate": 9.105574432457576e-06,
+      "loss": 0.0504,
+      "step": 4194
+    },
+    {
+      "epoch": 2.1837584591358667,
+      "grad_norm": 0.2315059122487158,
+      "learning_rate": 9.094732086989608e-06,
+      "loss": 0.0481,
+      "step": 4195
+    },
+    {
+      "epoch": 2.1842790213430505,
+      "grad_norm": 0.2388138458267806,
+      "learning_rate": 9.083894765145412e-06,
+      "loss": 0.0495,
+      "step": 4196
+    },
+    {
+      "epoch": 2.1847995835502343,
+      "grad_norm": 0.24338846982564916,
+      "learning_rate": 9.073062470347928e-06,
+      "loss": 0.0497,
+      "step": 4197
+    },
+    {
+      "epoch": 2.185320145757418,
+      "grad_norm": 0.24700761283541492,
+      "learning_rate": 9.062235206018488e-06,
+      "loss": 0.0512,
+      "step": 4198
+    },
+    {
+      "epoch": 2.185840707964602,
+      "grad_norm": 0.22202347104845527,
+      "learning_rate": 9.051412975576849e-06,
+      "loss": 0.0481,
+      "step": 4199
+    },
+    {
+      "epoch": 2.1863612701717856,
+      "grad_norm": 0.23160644856981238,
+      "learning_rate": 9.040595782441172e-06,
+      "loss": 0.0495,
+      "step": 4200
+    },
+    {
+      "epoch": 2.1868818323789694,
+      "grad_norm": 0.2192099194389428,
+      "learning_rate": 9.029783630028044e-06,
+      "loss": 0.0495,
+      "step": 4201
+    },
+    {
+      "epoch": 2.187402394586153,
+      "grad_norm": 0.23563639581238982,
+      "learning_rate": 9.018976521752426e-06,
+      "loss": 0.0519,
+      "step": 4202
+    },
+    {
+      "epoch": 2.187922956793337,
+      "grad_norm": 0.2229417962071288,
+      "learning_rate": 9.008174461027724e-06,
+      "loss": 0.0482,
+      "step": 4203
+    },
+    {
+      "epoch": 2.1884435190005207,
+      "grad_norm": 0.23972449679471294,
+      "learning_rate": 8.997377451265715e-06,
+      "loss": 0.0513,
+      "step": 4204
+    },
+    {
+      "epoch": 2.1889640812077045,
+      "grad_norm": 0.226247431389423,
+      "learning_rate": 8.986585495876605e-06,
+      "loss": 0.0493,
+      "step": 4205
+    },
+    {
+      "epoch": 2.1894846434148882,
+      "grad_norm": 0.23325199021404394,
+      "learning_rate": 8.975798598269002e-06,
+      "loss": 0.0506,
+      "step": 4206
+    },
+    {
+      "epoch": 2.190005205622072,
+      "grad_norm": 0.2245553471409382,
+      "learning_rate": 8.965016761849898e-06,
+      "loss": 0.0487,
+      "step": 4207
+    },
+    {
+      "epoch": 2.190525767829256,
+      "grad_norm": 0.23199873840527224,
+      "learning_rate": 8.954239990024704e-06,
+      "loss": 0.0485,
+      "step": 4208
+    },
+    {
+      "epoch": 2.1910463300364396,
+      "grad_norm": 0.22307074327270973,
+      "learning_rate": 8.943468286197224e-06,
+      "loss": 0.0477,
+      "step": 4209
+    },
+    {
+      "epoch": 2.191566892243623,
+      "grad_norm": 0.23538404056571152,
+      "learning_rate": 8.932701653769676e-06,
+      "loss": 0.0492,
+      "step": 4210
+    },
+    {
+      "epoch": 2.1920874544508067,
+      "grad_norm": 0.23621294854326094,
+      "learning_rate": 8.921940096142645e-06,
+      "loss": 0.0495,
+      "step": 4211
+    },
+    {
+      "epoch": 2.1926080166579904,
+      "grad_norm": 0.2286930249072395,
+      "learning_rate": 8.911183616715148e-06,
+      "loss": 0.0481,
+      "step": 4212
+    },
+    {
+      "epoch": 2.193128578865174,
+      "grad_norm": 0.2466558181780694,
+      "learning_rate": 8.900432218884567e-06,
+      "loss": 0.0509,
+      "step": 4213
+    },
+    {
+      "epoch": 2.193649141072358,
+      "grad_norm": 0.2372365759392663,
+      "learning_rate": 8.8896859060467e-06,
+      "loss": 0.0497,
+      "step": 4214
+    },
+    {
+      "epoch": 2.1941697032795417,
+      "grad_norm": 0.2319389953511392,
+      "learning_rate": 8.878944681595742e-06,
+      "loss": 0.0496,
+      "step": 4215
+    },
+    {
+      "epoch": 2.1946902654867255,
+      "grad_norm": 0.23364662530422042,
+      "learning_rate": 8.868208548924253e-06,
+      "loss": 0.0491,
+      "step": 4216
+    },
+    {
+      "epoch": 2.1952108276939093,
+      "grad_norm": 0.24276463476612706,
+      "learning_rate": 8.857477511423215e-06,
+      "loss": 0.05,
+      "step": 4217
+    },
+    {
+      "epoch": 2.195731389901093,
+      "grad_norm": 0.23011307596727132,
+      "learning_rate": 8.846751572481984e-06,
+      "loss": 0.0477,
+      "step": 4218
+    },
+    {
+      "epoch": 2.196251952108277,
+      "grad_norm": 0.2337315315374658,
+      "learning_rate": 8.836030735488327e-06,
+      "loss": 0.0489,
+      "step": 4219
+    },
+    {
+      "epoch": 2.1967725143154606,
+      "grad_norm": 0.2325310579261457,
+      "learning_rate": 8.825315003828358e-06,
+      "loss": 0.0465,
+      "step": 4220
+    },
+    {
+      "epoch": 2.1972930765226444,
+      "grad_norm": 0.22830561027317142,
+      "learning_rate": 8.814604380886623e-06,
+      "loss": 0.0476,
+      "step": 4221
+    },
+    {
+      "epoch": 2.197813638729828,
+      "grad_norm": 0.24012212633480134,
+      "learning_rate": 8.803898870046023e-06,
+      "loss": 0.0488,
+      "step": 4222
+    },
+    {
+      "epoch": 2.198334200937012,
+      "grad_norm": 0.23919713359665457,
+      "learning_rate": 8.79319847468786e-06,
+      "loss": 0.0489,
+      "step": 4223
+    },
+    {
+      "epoch": 2.1988547631441957,
+      "grad_norm": 0.2506742677947288,
+      "learning_rate": 8.782503198191828e-06,
+      "loss": 0.0514,
+      "step": 4224
+    },
+    {
+      "epoch": 2.1993753253513795,
+      "grad_norm": 0.23617636186386712,
+      "learning_rate": 8.771813043935972e-06,
+      "loss": 0.0494,
+      "step": 4225
+    },
+    {
+      "epoch": 2.1998958875585632,
+      "grad_norm": 0.23474992131532096,
+      "learning_rate": 8.761128015296754e-06,
+      "loss": 0.0494,
+      "step": 4226
+    },
+    {
+      "epoch": 2.200416449765747,
+      "grad_norm": 0.24128412011120356,
+      "learning_rate": 8.750448115649001e-06,
+      "loss": 0.0506,
+      "step": 4227
+    },
+    {
+      "epoch": 2.200937011972931,
+      "grad_norm": 0.23586375626059694,
+      "learning_rate": 8.739773348365928e-06,
+      "loss": 0.05,
+      "step": 4228
+    },
+    {
+      "epoch": 2.2014575741801146,
+      "grad_norm": 0.22292550967559244,
+      "learning_rate": 8.729103716819112e-06,
+      "loss": 0.0481,
+      "step": 4229
+    },
+    {
+      "epoch": 2.2019781363872983,
+      "grad_norm": 0.23125396362101913,
+      "learning_rate": 8.71843922437853e-06,
+      "loss": 0.0473,
+      "step": 4230
+    },
+    {
+      "epoch": 2.202498698594482,
+      "grad_norm": 0.22704085751147418,
+      "learning_rate": 8.707779874412514e-06,
+      "loss": 0.047,
+      "step": 4231
+    },
+    {
+      "epoch": 2.203019260801666,
+      "grad_norm": 0.23357200148046692,
+      "learning_rate": 8.697125670287787e-06,
+      "loss": 0.0486,
+      "step": 4232
+    },
+    {
+      "epoch": 2.2035398230088497,
+      "grad_norm": 0.24120365615695358,
+      "learning_rate": 8.686476615369451e-06,
+      "loss": 0.0507,
+      "step": 4233
+    },
+    {
+      "epoch": 2.2040603852160334,
+      "grad_norm": 0.2440416527287281,
+      "learning_rate": 8.67583271302096e-06,
+      "loss": 0.0498,
+      "step": 4234
+    },
+    {
+      "epoch": 2.204580947423217,
+      "grad_norm": 0.22550408972403665,
+      "learning_rate": 8.665193966604157e-06,
+      "loss": 0.0461,
+      "step": 4235
+    },
+    {
+      "epoch": 2.205101509630401,
+      "grad_norm": 0.24657257357860465,
+      "learning_rate": 8.654560379479257e-06,
+      "loss": 0.0508,
+      "step": 4236
+    },
+    {
+      "epoch": 2.2056220718375847,
+      "grad_norm": 0.2378711148974515,
+      "learning_rate": 8.643931955004839e-06,
+      "loss": 0.0504,
+      "step": 4237
+    },
+    {
+      "epoch": 2.2061426340447685,
+      "grad_norm": 0.23414764136210095,
+      "learning_rate": 8.633308696537865e-06,
+      "loss": 0.0486,
+      "step": 4238
+    },
+    {
+      "epoch": 2.2066631962519523,
+      "grad_norm": 0.24236213181260452,
+      "learning_rate": 8.622690607433644e-06,
+      "loss": 0.0506,
+      "step": 4239
+    },
+    {
+      "epoch": 2.207183758459136,
+      "grad_norm": 0.22663158777101486,
+      "learning_rate": 8.612077691045856e-06,
+      "loss": 0.0481,
+      "step": 4240
+    },
+    {
+      "epoch": 2.20770432066632,
+      "grad_norm": 0.23127324675787353,
+      "learning_rate": 8.601469950726562e-06,
+      "loss": 0.0494,
+      "step": 4241
+    },
+    {
+      "epoch": 2.2082248828735036,
+      "grad_norm": 0.23219971597920153,
+      "learning_rate": 8.59086738982618e-06,
+      "loss": 0.0482,
+      "step": 4242
+    },
+    {
+      "epoch": 2.2087454450806874,
+      "grad_norm": 0.22566023224763826,
+      "learning_rate": 8.580270011693498e-06,
+      "loss": 0.0501,
+      "step": 4243
+    },
+    {
+      "epoch": 2.2092660072878707,
+      "grad_norm": 0.22931201717326824,
+      "learning_rate": 8.569677819675646e-06,
+      "loss": 0.0469,
+      "step": 4244
+    },
+    {
+      "epoch": 2.2097865694950545,
+      "grad_norm": 0.22932887506999267,
+      "learning_rate": 8.55909081711814e-06,
+      "loss": 0.0474,
+      "step": 4245
+    },
+    {
+      "epoch": 2.2103071317022382,
+      "grad_norm": 0.23290519077904787,
+      "learning_rate": 8.548509007364849e-06,
+      "loss": 0.0484,
+      "step": 4246
+    },
+    {
+      "epoch": 2.210827693909422,
+      "grad_norm": 0.23050032981873889,
+      "learning_rate": 8.537932393758008e-06,
+      "loss": 0.0495,
+      "step": 4247
+    },
+    {
+      "epoch": 2.211348256116606,
+      "grad_norm": 0.23320169879502017,
+      "learning_rate": 8.527360979638196e-06,
+      "loss": 0.0487,
+      "step": 4248
+    },
+    {
+      "epoch": 2.2118688183237896,
+      "grad_norm": 0.22937417737271626,
+      "learning_rate": 8.51679476834435e-06,
+      "loss": 0.0492,
+      "step": 4249
+    },
+    {
+      "epoch": 2.2123893805309733,
+      "grad_norm": 0.23866247343146085,
+      "learning_rate": 8.506233763213776e-06,
+      "loss": 0.0481,
+      "step": 4250
+    },
+    {
+      "epoch": 2.212909942738157,
+      "grad_norm": 0.2544913382799314,
+      "learning_rate": 8.495677967582135e-06,
+      "loss": 0.0493,
+      "step": 4251
+    },
+    {
+      "epoch": 2.213430504945341,
+      "grad_norm": 0.2335235599829452,
+      "learning_rate": 8.485127384783446e-06,
+      "loss": 0.0486,
+      "step": 4252
+    },
+    {
+      "epoch": 2.2139510671525247,
+      "grad_norm": 0.24283971222288098,
+      "learning_rate": 8.474582018150054e-06,
+      "loss": 0.0484,
+      "step": 4253
+    },
+    {
+      "epoch": 2.2144716293597084,
+      "grad_norm": 0.23164997555170744,
+      "learning_rate": 8.464041871012687e-06,
+      "loss": 0.0497,
+      "step": 4254
+    },
+    {
+      "epoch": 2.214992191566892,
+      "grad_norm": 0.2413814098062649,
+      "learning_rate": 8.453506946700418e-06,
+      "loss": 0.0493,
+      "step": 4255
+    },
+    {
+      "epoch": 2.215512753774076,
+      "grad_norm": 0.2454607147482343,
+      "learning_rate": 8.442977248540667e-06,
+      "loss": 0.0479,
+      "step": 4256
+    },
+    {
+      "epoch": 2.2160333159812597,
+      "grad_norm": 0.23779862580138247,
+      "learning_rate": 8.4324527798592e-06,
+      "loss": 0.0497,
+      "step": 4257
+    },
+    {
+      "epoch": 2.2165538781884435,
+      "grad_norm": 0.23762422063366767,
+      "learning_rate": 8.421933543980126e-06,
+      "loss": 0.0505,
+      "step": 4258
+    },
+    {
+      "epoch": 2.2170744403956273,
+      "grad_norm": 0.23130632658477607,
+      "learning_rate": 8.411419544225913e-06,
+      "loss": 0.0487,
+      "step": 4259
+    },
+    {
+      "epoch": 2.217595002602811,
+      "grad_norm": 0.22356019790315676,
+      "learning_rate": 8.400910783917377e-06,
+      "loss": 0.0472,
+      "step": 4260
+    },
+    {
+      "epoch": 2.218115564809995,
+      "grad_norm": 0.22887925032024237,
+      "learning_rate": 8.390407266373674e-06,
+      "loss": 0.0476,
+      "step": 4261
+    },
+    {
+      "epoch": 2.2186361270171786,
+      "grad_norm": 0.23463897003025824,
+      "learning_rate": 8.379908994912294e-06,
+      "loss": 0.049,
+      "step": 4262
+    },
+    {
+      "epoch": 2.2191566892243624,
+      "grad_norm": 0.23240197161886428,
+      "learning_rate": 8.369415972849088e-06,
+      "loss": 0.0495,
+      "step": 4263
+    },
+    {
+      "epoch": 2.219677251431546,
+      "grad_norm": 0.22919016167185535,
+      "learning_rate": 8.358928203498236e-06,
+      "loss": 0.0502,
+      "step": 4264
+    },
+    {
+      "epoch": 2.22019781363873,
+      "grad_norm": 0.2202440420250215,
+      "learning_rate": 8.348445690172274e-06,
+      "loss": 0.0476,
+      "step": 4265
+    },
+    {
+      "epoch": 2.2207183758459137,
+      "grad_norm": 0.23110580036654815,
+      "learning_rate": 8.337968436182054e-06,
+      "loss": 0.0503,
+      "step": 4266
+    },
+    {
+      "epoch": 2.2212389380530975,
+      "grad_norm": 0.23758638402574997,
+      "learning_rate": 8.327496444836793e-06,
+      "loss": 0.0509,
+      "step": 4267
+    },
+    {
+      "epoch": 2.2217595002602812,
+      "grad_norm": 0.22791558989136954,
+      "learning_rate": 8.317029719444016e-06,
+      "loss": 0.0469,
+      "step": 4268
+    },
+    {
+      "epoch": 2.222280062467465,
+      "grad_norm": 0.23245148595947804,
+      "learning_rate": 8.306568263309616e-06,
+      "loss": 0.049,
+      "step": 4269
+    },
+    {
+      "epoch": 2.222800624674649,
+      "grad_norm": 0.22057289323649362,
+      "learning_rate": 8.296112079737808e-06,
+      "loss": 0.047,
+      "step": 4270
+    },
+    {
+      "epoch": 2.2233211868818326,
+      "grad_norm": 0.2291814857700997,
+      "learning_rate": 8.28566117203113e-06,
+      "loss": 0.0493,
+      "step": 4271
+    },
+    {
+      "epoch": 2.2238417490890163,
+      "grad_norm": 0.23917673247453083,
+      "learning_rate": 8.275215543490475e-06,
+      "loss": 0.0492,
+      "step": 4272
+    },
+    {
+      "epoch": 2.2243623112962,
+      "grad_norm": 0.23646886044374246,
+      "learning_rate": 8.264775197415053e-06,
+      "loss": 0.0495,
+      "step": 4273
+    },
+    {
+      "epoch": 2.2248828735033834,
+      "grad_norm": 0.23541029580737305,
+      "learning_rate": 8.254340137102426e-06,
+      "loss": 0.0485,
+      "step": 4274
+    },
+    {
+      "epoch": 2.225403435710567,
+      "grad_norm": 0.25343510785437356,
+      "learning_rate": 8.243910365848448e-06,
+      "loss": 0.051,
+      "step": 4275
+    },
+    {
+      "epoch": 2.225923997917751,
+      "grad_norm": 0.2298835859276625,
+      "learning_rate": 8.233485886947346e-06,
+      "loss": 0.0489,
+      "step": 4276
+    },
+    {
+      "epoch": 2.2264445601249347,
+      "grad_norm": 0.22912963408681325,
+      "learning_rate": 8.22306670369164e-06,
+      "loss": 0.0487,
+      "step": 4277
+    },
+    {
+      "epoch": 2.2269651223321185,
+      "grad_norm": 0.23556170673164437,
+      "learning_rate": 8.2126528193722e-06,
+      "loss": 0.0485,
+      "step": 4278
+    },
+    {
+      "epoch": 2.2274856845393023,
+      "grad_norm": 0.22736010372358842,
+      "learning_rate": 8.202244237278223e-06,
+      "loss": 0.0474,
+      "step": 4279
+    },
+    {
+      "epoch": 2.228006246746486,
+      "grad_norm": 0.2427914749744837,
+      "learning_rate": 8.19184096069721e-06,
+      "loss": 0.0501,
+      "step": 4280
+    },
+    {
+      "epoch": 2.22852680895367,
+      "grad_norm": 0.2351054578306983,
+      "learning_rate": 8.181442992915e-06,
+      "loss": 0.0486,
+      "step": 4281
+    },
+    {
+      "epoch": 2.2290473711608536,
+      "grad_norm": 0.22778548939677618,
+      "learning_rate": 8.171050337215767e-06,
+      "loss": 0.0483,
+      "step": 4282
+    },
+    {
+      "epoch": 2.2295679333680374,
+      "grad_norm": 0.22980336566734544,
+      "learning_rate": 8.160662996881996e-06,
+      "loss": 0.0468,
+      "step": 4283
+    },
+    {
+      "epoch": 2.230088495575221,
+      "grad_norm": 0.23245660876536978,
+      "learning_rate": 8.150280975194478e-06,
+      "loss": 0.0477,
+      "step": 4284
+    },
+    {
+      "epoch": 2.230609057782405,
+      "grad_norm": 0.22749204175239007,
+      "learning_rate": 8.139904275432354e-06,
+      "loss": 0.0489,
+      "step": 4285
+    },
+    {
+      "epoch": 2.2311296199895887,
+      "grad_norm": 0.23848791969754046,
+      "learning_rate": 8.129532900873051e-06,
+      "loss": 0.0495,
+      "step": 4286
+    },
+    {
+      "epoch": 2.2316501821967725,
+      "grad_norm": 0.24418712222939992,
+      "learning_rate": 8.119166854792345e-06,
+      "loss": 0.048,
+      "step": 4287
+    },
+    {
+      "epoch": 2.2321707444039562,
+      "grad_norm": 0.23832224809407973,
+      "learning_rate": 8.10880614046432e-06,
+      "loss": 0.0502,
+      "step": 4288
+    },
+    {
+      "epoch": 2.23269130661114,
+      "grad_norm": 0.2349337209019424,
+      "learning_rate": 8.098450761161356e-06,
+      "loss": 0.0488,
+      "step": 4289
+    },
+    {
+      "epoch": 2.233211868818324,
+      "grad_norm": 0.236101048702289,
+      "learning_rate": 8.08810072015417e-06,
+      "loss": 0.0486,
+      "step": 4290
+    },
+    {
+      "epoch": 2.2337324310255076,
+      "grad_norm": 0.2436594526550554,
+      "learning_rate": 8.07775602071179e-06,
+      "loss": 0.0502,
+      "step": 4291
+    },
+    {
+      "epoch": 2.2342529932326913,
+      "grad_norm": 0.24215638153778066,
+      "learning_rate": 8.067416666101562e-06,
+      "loss": 0.0501,
+      "step": 4292
+    },
+    {
+      "epoch": 2.234773555439875,
+      "grad_norm": 0.22910339256480491,
+      "learning_rate": 8.057082659589115e-06,
+      "loss": 0.0485,
+      "step": 4293
+    },
+    {
+      "epoch": 2.235294117647059,
+      "grad_norm": 0.23759220672459413,
+      "learning_rate": 8.046754004438429e-06,
+      "loss": 0.0501,
+      "step": 4294
+    },
+    {
+      "epoch": 2.2358146798542426,
+      "grad_norm": 0.2396620539258785,
+      "learning_rate": 8.036430703911754e-06,
+      "loss": 0.0505,
+      "step": 4295
+    },
+    {
+      "epoch": 2.2363352420614264,
+      "grad_norm": 0.21656808633265354,
+      "learning_rate": 8.026112761269683e-06,
+      "loss": 0.0467,
+      "step": 4296
+    },
+    {
+      "epoch": 2.23685580426861,
+      "grad_norm": 0.23694210045408912,
+      "learning_rate": 8.015800179771105e-06,
+      "loss": 0.0492,
+      "step": 4297
+    },
+    {
+      "epoch": 2.237376366475794,
+      "grad_norm": 0.23066475593043279,
+      "learning_rate": 8.005492962673197e-06,
+      "loss": 0.0485,
+      "step": 4298
+    },
+    {
+      "epoch": 2.2378969286829777,
+      "grad_norm": 0.23780414518123882,
+      "learning_rate": 7.99519111323147e-06,
+      "loss": 0.0502,
+      "step": 4299
+    },
+    {
+      "epoch": 2.2384174908901615,
+      "grad_norm": 0.22874416790471078,
+      "learning_rate": 7.984894634699725e-06,
+      "loss": 0.0477,
+      "step": 4300
+    },
+    {
+      "epoch": 2.2389380530973453,
+      "grad_norm": 0.225516589384533,
+      "learning_rate": 7.974603530330069e-06,
+      "loss": 0.0478,
+      "step": 4301
+    },
+    {
+      "epoch": 2.239458615304529,
+      "grad_norm": 0.24222066580737353,
+      "learning_rate": 7.964317803372918e-06,
+      "loss": 0.0509,
+      "step": 4302
+    },
+    {
+      "epoch": 2.239979177511713,
+      "grad_norm": 0.24287142818992802,
+      "learning_rate": 7.95403745707698e-06,
+      "loss": 0.0497,
+      "step": 4303
+    },
+    {
+      "epoch": 2.2404997397188966,
+      "grad_norm": 0.24632751357990357,
+      "learning_rate": 7.943762494689252e-06,
+      "loss": 0.0503,
+      "step": 4304
+    },
+    {
+      "epoch": 2.2410203019260804,
+      "grad_norm": 0.2450475168217388,
+      "learning_rate": 7.93349291945506e-06,
+      "loss": 0.0484,
+      "step": 4305
+    },
+    {
+      "epoch": 2.241540864133264,
+      "grad_norm": 0.24379993992170595,
+      "learning_rate": 7.92322873461801e-06,
+      "loss": 0.0478,
+      "step": 4306
+    },
+    {
+      "epoch": 2.242061426340448,
+      "grad_norm": 0.24148406455105237,
+      "learning_rate": 7.912969943420018e-06,
+      "loss": 0.049,
+      "step": 4307
+    },
+    {
+      "epoch": 2.2425819885476312,
+      "grad_norm": 0.25211178202149287,
+      "learning_rate": 7.902716549101272e-06,
+      "loss": 0.0519,
+      "step": 4308
+    },
+    {
+      "epoch": 2.243102550754815,
+      "grad_norm": 0.22380206855764131,
+      "learning_rate": 7.892468554900278e-06,
+      "loss": 0.0457,
+      "step": 4309
+    },
+    {
+      "epoch": 2.243623112961999,
+      "grad_norm": 0.23150832565863866,
+      "learning_rate": 7.88222596405383e-06,
+      "loss": 0.048,
+      "step": 4310
+    },
+    {
+      "epoch": 2.2441436751691826,
+      "grad_norm": 0.24560507817753072,
+      "learning_rate": 7.871988779797024e-06,
+      "loss": 0.0509,
+      "step": 4311
+    },
+    {
+      "epoch": 2.2446642373763663,
+      "grad_norm": 0.22181091893953384,
+      "learning_rate": 7.861757005363232e-06,
+      "loss": 0.0475,
+      "step": 4312
+    },
+    {
+      "epoch": 2.24518479958355,
+      "grad_norm": 0.23679248354187557,
+      "learning_rate": 7.851530643984111e-06,
+      "loss": 0.0484,
+      "step": 4313
+    },
+    {
+      "epoch": 2.245705361790734,
+      "grad_norm": 0.23595668958031096,
+      "learning_rate": 7.841309698889638e-06,
+      "loss": 0.0501,
+      "step": 4314
+    },
+    {
+      "epoch": 2.2462259239979177,
+      "grad_norm": 0.22600608264316838,
+      "learning_rate": 7.831094173308056e-06,
+      "loss": 0.0483,
+      "step": 4315
+    },
+    {
+      "epoch": 2.2467464862051014,
+      "grad_norm": 0.23927132977588947,
+      "learning_rate": 7.820884070465914e-06,
+      "loss": 0.0497,
+      "step": 4316
+    },
+    {
+      "epoch": 2.247267048412285,
+      "grad_norm": 0.2291797616221194,
+      "learning_rate": 7.810679393588025e-06,
+      "loss": 0.049,
+      "step": 4317
+    },
+    {
+      "epoch": 2.247787610619469,
+      "grad_norm": 0.2252726239114742,
+      "learning_rate": 7.800480145897501e-06,
+      "loss": 0.0479,
+      "step": 4318
+    },
+    {
+      "epoch": 2.2483081728266527,
+      "grad_norm": 0.2209306053422971,
+      "learning_rate": 7.790286330615749e-06,
+      "loss": 0.0466,
+      "step": 4319
+    },
+    {
+      "epoch": 2.2488287350338365,
+      "grad_norm": 0.21256290410992668,
+      "learning_rate": 7.780097950962447e-06,
+      "loss": 0.0459,
+      "step": 4320
+    },
+    {
+      "epoch": 2.2493492972410203,
+      "grad_norm": 0.23304751234583473,
+      "learning_rate": 7.76991501015556e-06,
+      "loss": 0.0494,
+      "step": 4321
+    },
+    {
+      "epoch": 2.249869859448204,
+      "grad_norm": 0.24109946012783512,
+      "learning_rate": 7.759737511411325e-06,
+      "loss": 0.05,
+      "step": 4322
+    },
+    {
+      "epoch": 2.250390421655388,
+      "grad_norm": 0.23100844503880485,
+      "learning_rate": 7.749565457944274e-06,
+      "loss": 0.0477,
+      "step": 4323
+    },
+    {
+      "epoch": 2.2509109838625716,
+      "grad_norm": 0.22756640638272935,
+      "learning_rate": 7.73939885296722e-06,
+      "loss": 0.0488,
+      "step": 4324
+    },
+    {
+      "epoch": 2.2514315460697554,
+      "grad_norm": 0.2411358442421993,
+      "learning_rate": 7.729237699691254e-06,
+      "loss": 0.0514,
+      "step": 4325
+    },
+    {
+      "epoch": 2.251952108276939,
+      "grad_norm": 0.2385049139121903,
+      "learning_rate": 7.719082001325728e-06,
+      "loss": 0.0489,
+      "step": 4326
+    },
+    {
+      "epoch": 2.252472670484123,
+      "grad_norm": 0.24245891150143284,
+      "learning_rate": 7.70893176107829e-06,
+      "loss": 0.0492,
+      "step": 4327
+    },
+    {
+      "epoch": 2.2529932326913067,
+      "grad_norm": 0.23962567363070575,
+      "learning_rate": 7.698786982154857e-06,
+      "loss": 0.0486,
+      "step": 4328
+    },
+    {
+      "epoch": 2.2535137948984905,
+      "grad_norm": 0.23067229481030735,
+      "learning_rate": 7.688647667759633e-06,
+      "loss": 0.0482,
+      "step": 4329
+    },
+    {
+      "epoch": 2.2540343571056742,
+      "grad_norm": 0.23572306878031657,
+      "learning_rate": 7.678513821095076e-06,
+      "loss": 0.0488,
+      "step": 4330
+    },
+    {
+      "epoch": 2.254554919312858,
+      "grad_norm": 0.22767821391268697,
+      "learning_rate": 7.668385445361923e-06,
+      "loss": 0.0478,
+      "step": 4331
+    },
+    {
+      "epoch": 2.255075481520042,
+      "grad_norm": 0.22836659984706298,
+      "learning_rate": 7.658262543759184e-06,
+      "loss": 0.0465,
+      "step": 4332
+    },
+    {
+      "epoch": 2.2555960437272256,
+      "grad_norm": 0.2393729125564174,
+      "learning_rate": 7.648145119484152e-06,
+      "loss": 0.0482,
+      "step": 4333
+    },
+    {
+      "epoch": 2.2561166059344093,
+      "grad_norm": 0.23094639008000092,
+      "learning_rate": 7.638033175732385e-06,
+      "loss": 0.0483,
+      "step": 4334
+    },
+    {
+      "epoch": 2.256637168141593,
+      "grad_norm": 0.22558491929936555,
+      "learning_rate": 7.627926715697689e-06,
+      "loss": 0.0473,
+      "step": 4335
+    },
+    {
+      "epoch": 2.257157730348777,
+      "grad_norm": 0.23284991884566408,
+      "learning_rate": 7.617825742572163e-06,
+      "loss": 0.0488,
+      "step": 4336
+    },
+    {
+      "epoch": 2.25767829255596,
+      "grad_norm": 0.23896169808363904,
+      "learning_rate": 7.607730259546164e-06,
+      "loss": 0.0491,
+      "step": 4337
+    },
+    {
+      "epoch": 2.258198854763144,
+      "grad_norm": 0.229443652276642,
+      "learning_rate": 7.597640269808323e-06,
+      "loss": 0.0483,
+      "step": 4338
+    },
+    {
+      "epoch": 2.2587194169703277,
+      "grad_norm": 0.22477269530377628,
+      "learning_rate": 7.5875557765455245e-06,
+      "loss": 0.0485,
+      "step": 4339
+    },
+    {
+      "epoch": 2.2592399791775115,
+      "grad_norm": 0.2248750508583482,
+      "learning_rate": 7.577476782942905e-06,
+      "loss": 0.046,
+      "step": 4340
+    },
+    {
+      "epoch": 2.2597605413846953,
+      "grad_norm": 0.24282761232837097,
+      "learning_rate": 7.567403292183892e-06,
+      "loss": 0.0488,
+      "step": 4341
+    },
+    {
+      "epoch": 2.260281103591879,
+      "grad_norm": 0.24559688750387063,
+      "learning_rate": 7.557335307450164e-06,
+      "loss": 0.0487,
+      "step": 4342
+    },
+    {
+      "epoch": 2.260801665799063,
+      "grad_norm": 0.2279229009237183,
+      "learning_rate": 7.547272831921665e-06,
+      "loss": 0.0473,
+      "step": 4343
+    },
+    {
+      "epoch": 2.2613222280062466,
+      "grad_norm": 0.24163113811010245,
+      "learning_rate": 7.5372158687765784e-06,
+      "loss": 0.0488,
+      "step": 4344
+    },
+    {
+      "epoch": 2.2618427902134304,
+      "grad_norm": 0.22542343579424412,
+      "learning_rate": 7.527164421191369e-06,
+      "loss": 0.0473,
+      "step": 4345
+    },
+    {
+      "epoch": 2.262363352420614,
+      "grad_norm": 0.24354092974905714,
+      "learning_rate": 7.51711849234075e-06,
+      "loss": 0.0493,
+      "step": 4346
+    },
+    {
+      "epoch": 2.262883914627798,
+      "grad_norm": 0.2315919364564454,
+      "learning_rate": 7.507078085397701e-06,
+      "loss": 0.0481,
+      "step": 4347
+    },
+    {
+      "epoch": 2.2634044768349817,
+      "grad_norm": 0.23710368286301195,
+      "learning_rate": 7.497043203533444e-06,
+      "loss": 0.0481,
+      "step": 4348
+    },
+    {
+      "epoch": 2.2639250390421655,
+      "grad_norm": 0.23553140969761996,
+      "learning_rate": 7.487013849917454e-06,
+      "loss": 0.0479,
+      "step": 4349
+    },
+    {
+      "epoch": 2.2644456012493492,
+      "grad_norm": 0.24202468474373762,
+      "learning_rate": 7.476990027717473e-06,
+      "loss": 0.0502,
+      "step": 4350
+    },
+    {
+      "epoch": 2.264966163456533,
+      "grad_norm": 0.24252991253463546,
+      "learning_rate": 7.46697174009949e-06,
+      "loss": 0.0506,
+      "step": 4351
+    },
+    {
+      "epoch": 2.265486725663717,
+      "grad_norm": 0.24450739997748402,
+      "learning_rate": 7.456958990227761e-06,
+      "loss": 0.0493,
+      "step": 4352
+    },
+    {
+      "epoch": 2.2660072878709006,
+      "grad_norm": 0.23512299197952063,
+      "learning_rate": 7.446951781264755e-06,
+      "loss": 0.05,
+      "step": 4353
+    },
+    {
+      "epoch": 2.2665278500780843,
+      "grad_norm": 0.2349636567214315,
+      "learning_rate": 7.436950116371225e-06,
+      "loss": 0.0478,
+      "step": 4354
+    },
+    {
+      "epoch": 2.267048412285268,
+      "grad_norm": 0.22313490108520748,
+      "learning_rate": 7.4269539987061625e-06,
+      "loss": 0.0489,
+      "step": 4355
+    },
+    {
+      "epoch": 2.267568974492452,
+      "grad_norm": 0.24266098411924822,
+      "learning_rate": 7.416963431426815e-06,
+      "loss": 0.0508,
+      "step": 4356
+    },
+    {
+      "epoch": 2.2680895366996356,
+      "grad_norm": 0.22709180519413907,
+      "learning_rate": 7.406978417688659e-06,
+      "loss": 0.0476,
+      "step": 4357
+    },
+    {
+      "epoch": 2.2686100989068194,
+      "grad_norm": 0.2252843789851944,
+      "learning_rate": 7.396998960645418e-06,
+      "loss": 0.0475,
+      "step": 4358
+    },
+    {
+      "epoch": 2.269130661114003,
+      "grad_norm": 0.2280964931147517,
+      "learning_rate": 7.387025063449082e-06,
+      "loss": 0.0479,
+      "step": 4359
+    },
+    {
+      "epoch": 2.269651223321187,
+      "grad_norm": 0.22442020507845567,
+      "learning_rate": 7.377056729249865e-06,
+      "loss": 0.0474,
+      "step": 4360
+    },
+    {
+      "epoch": 2.2701717855283707,
+      "grad_norm": 0.23536807823855144,
+      "learning_rate": 7.3670939611962446e-06,
+      "loss": 0.0471,
+      "step": 4361
+    },
+    {
+      "epoch": 2.2706923477355545,
+      "grad_norm": 0.22341853555386254,
+      "learning_rate": 7.357136762434908e-06,
+      "loss": 0.0467,
+      "step": 4362
+    },
+    {
+      "epoch": 2.2712129099427383,
+      "grad_norm": 0.2314457941140156,
+      "learning_rate": 7.347185136110807e-06,
+      "loss": 0.0474,
+      "step": 4363
+    },
+    {
+      "epoch": 2.271733472149922,
+      "grad_norm": 0.23444389464917653,
+      "learning_rate": 7.337239085367134e-06,
+      "loss": 0.0473,
+      "step": 4364
+    },
+    {
+      "epoch": 2.272254034357106,
+      "grad_norm": 0.22895085279319216,
+      "learning_rate": 7.32729861334531e-06,
+      "loss": 0.0482,
+      "step": 4365
+    },
+    {
+      "epoch": 2.2727745965642896,
+      "grad_norm": 0.2301198469063954,
+      "learning_rate": 7.317363723185017e-06,
+      "loss": 0.0466,
+      "step": 4366
+    },
+    {
+      "epoch": 2.2732951587714734,
+      "grad_norm": 0.2375738571090302,
+      "learning_rate": 7.3074344180241225e-06,
+      "loss": 0.0493,
+      "step": 4367
+    },
+    {
+      "epoch": 2.273815720978657,
+      "grad_norm": 0.23547330811705733,
+      "learning_rate": 7.297510700998783e-06,
+      "loss": 0.0467,
+      "step": 4368
+    },
+    {
+      "epoch": 2.274336283185841,
+      "grad_norm": 0.23704718711434258,
+      "learning_rate": 7.2875925752433655e-06,
+      "loss": 0.0484,
+      "step": 4369
+    },
+    {
+      "epoch": 2.2748568453930247,
+      "grad_norm": 0.23007390803788796,
+      "learning_rate": 7.277680043890475e-06,
+      "loss": 0.0462,
+      "step": 4370
+    },
+    {
+      "epoch": 2.2753774076002085,
+      "grad_norm": 0.2382599584192981,
+      "learning_rate": 7.267773110070964e-06,
+      "loss": 0.0489,
+      "step": 4371
+    },
+    {
+      "epoch": 2.2758979698073922,
+      "grad_norm": 0.23792105081665424,
+      "learning_rate": 7.25787177691388e-06,
+      "loss": 0.0485,
+      "step": 4372
+    },
+    {
+      "epoch": 2.2764185320145756,
+      "grad_norm": 0.22893605244305398,
+      "learning_rate": 7.2479760475465395e-06,
+      "loss": 0.0457,
+      "step": 4373
+    },
+    {
+      "epoch": 2.2769390942217593,
+      "grad_norm": 0.2372593732772901,
+      "learning_rate": 7.238085925094468e-06,
+      "loss": 0.0486,
+      "step": 4374
+    },
+    {
+      "epoch": 2.277459656428943,
+      "grad_norm": 0.23222430381532452,
+      "learning_rate": 7.22820141268144e-06,
+      "loss": 0.0473,
+      "step": 4375
+    },
+    {
+      "epoch": 2.277980218636127,
+      "grad_norm": 0.23165532275317738,
+      "learning_rate": 7.2183225134294345e-06,
+      "loss": 0.0489,
+      "step": 4376
+    },
+    {
+      "epoch": 2.2785007808433106,
+      "grad_norm": 0.22857397337810045,
+      "learning_rate": 7.2084492304586586e-06,
+      "loss": 0.0485,
+      "step": 4377
+    },
+    {
+      "epoch": 2.2790213430504944,
+      "grad_norm": 0.2308547953684119,
+      "learning_rate": 7.19858156688756e-06,
+      "loss": 0.0459,
+      "step": 4378
+    },
+    {
+      "epoch": 2.279541905257678,
+      "grad_norm": 0.22575144523819185,
+      "learning_rate": 7.188719525832813e-06,
+      "loss": 0.0476,
+      "step": 4379
+    },
+    {
+      "epoch": 2.280062467464862,
+      "grad_norm": 0.24027575061928344,
+      "learning_rate": 7.1788631104093145e-06,
+      "loss": 0.0496,
+      "step": 4380
+    },
+    {
+      "epoch": 2.2805830296720457,
+      "grad_norm": 0.23540572654705236,
+      "learning_rate": 7.1690123237301596e-06,
+      "loss": 0.0501,
+      "step": 4381
+    },
+    {
+      "epoch": 2.2811035918792295,
+      "grad_norm": 0.24772965477869333,
+      "learning_rate": 7.159167168906694e-06,
+      "loss": 0.049,
+      "step": 4382
+    },
+    {
+      "epoch": 2.2816241540864133,
+      "grad_norm": 0.2320276288812207,
+      "learning_rate": 7.149327649048482e-06,
+      "loss": 0.047,
+      "step": 4383
+    },
+    {
+      "epoch": 2.282144716293597,
+      "grad_norm": 0.23303073974073554,
+      "learning_rate": 7.1394937672633e-06,
+      "loss": 0.0481,
+      "step": 4384
+    },
+    {
+      "epoch": 2.282665278500781,
+      "grad_norm": 0.2353449984385499,
+      "learning_rate": 7.129665526657145e-06,
+      "loss": 0.0479,
+      "step": 4385
+    },
+    {
+      "epoch": 2.2831858407079646,
+      "grad_norm": 0.23270277910500065,
+      "learning_rate": 7.119842930334222e-06,
+      "loss": 0.0479,
+      "step": 4386
+    },
+    {
+      "epoch": 2.2837064029151484,
+      "grad_norm": 0.24467249136218208,
+      "learning_rate": 7.110025981396975e-06,
+      "loss": 0.0503,
+      "step": 4387
+    },
+    {
+      "epoch": 2.284226965122332,
+      "grad_norm": 0.23634108219594238,
+      "learning_rate": 7.100214682946049e-06,
+      "loss": 0.0483,
+      "step": 4388
+    },
+    {
+      "epoch": 2.284747527329516,
+      "grad_norm": 0.2243524743535289,
+      "learning_rate": 7.090409038080317e-06,
+      "loss": 0.0477,
+      "step": 4389
+    },
+    {
+      "epoch": 2.2852680895366997,
+      "grad_norm": 0.2272063584018309,
+      "learning_rate": 7.080609049896844e-06,
+      "loss": 0.0486,
+      "step": 4390
+    },
+    {
+      "epoch": 2.2857886517438835,
+      "grad_norm": 0.22104862431300737,
+      "learning_rate": 7.0708147214909315e-06,
+      "loss": 0.0479,
+      "step": 4391
+    },
+    {
+      "epoch": 2.2863092139510672,
+      "grad_norm": 0.2324349902565072,
+      "learning_rate": 7.06102605595608e-06,
+      "loss": 0.0489,
+      "step": 4392
+    },
+    {
+      "epoch": 2.286829776158251,
+      "grad_norm": 0.23945786205814168,
+      "learning_rate": 7.051243056384016e-06,
+      "loss": 0.0473,
+      "step": 4393
+    },
+    {
+      "epoch": 2.2873503383654348,
+      "grad_norm": 0.2290382839925783,
+      "learning_rate": 7.04146572586466e-06,
+      "loss": 0.0476,
+      "step": 4394
+    },
+    {
+      "epoch": 2.2878709005726185,
+      "grad_norm": 0.22924448362356376,
+      "learning_rate": 7.031694067486136e-06,
+      "loss": 0.0474,
+      "step": 4395
+    },
+    {
+      "epoch": 2.2883914627798023,
+      "grad_norm": 0.23714424830350492,
+      "learning_rate": 7.0219280843348e-06,
+      "loss": 0.05,
+      "step": 4396
+    },
+    {
+      "epoch": 2.288912024986986,
+      "grad_norm": 0.2261718669538399,
+      "learning_rate": 7.012167779495201e-06,
+      "loss": 0.0471,
+      "step": 4397
+    },
+    {
+      "epoch": 2.28943258719417,
+      "grad_norm": 0.22758504532465043,
+      "learning_rate": 7.002413156050108e-06,
+      "loss": 0.0467,
+      "step": 4398
+    },
+    {
+      "epoch": 2.2899531494013536,
+      "grad_norm": 0.23789065938410056,
+      "learning_rate": 6.9926642170804665e-06,
+      "loss": 0.0485,
+      "step": 4399
+    },
+    {
+      "epoch": 2.2904737116085374,
+      "grad_norm": 0.24091702145465044,
+      "learning_rate": 6.982920965665457e-06,
+      "loss": 0.0455,
+      "step": 4400
+    },
+    {
+      "epoch": 2.2909942738157207,
+      "grad_norm": 0.2359330189743294,
+      "learning_rate": 6.9731834048824465e-06,
+      "loss": 0.0486,
+      "step": 4401
+    },
+    {
+      "epoch": 2.2915148360229045,
+      "grad_norm": 0.23437207513741382,
+      "learning_rate": 6.963451537807023e-06,
+      "loss": 0.0473,
+      "step": 4402
+    },
+    {
+      "epoch": 2.2920353982300883,
+      "grad_norm": 0.23735518742716438,
+      "learning_rate": 6.953725367512951e-06,
+      "loss": 0.0476,
+      "step": 4403
+    },
+    {
+      "epoch": 2.292555960437272,
+      "grad_norm": 0.2462383376150093,
+      "learning_rate": 6.944004897072201e-06,
+      "loss": 0.0487,
+      "step": 4404
+    },
+    {
+      "epoch": 2.293076522644456,
+      "grad_norm": 0.2411982549992137,
+      "learning_rate": 6.934290129554957e-06,
+      "loss": 0.0477,
+      "step": 4405
+    },
+    {
+      "epoch": 2.2935970848516396,
+      "grad_norm": 0.218847924533852,
+      "learning_rate": 6.924581068029598e-06,
+      "loss": 0.0456,
+      "step": 4406
+    },
+    {
+      "epoch": 2.2941176470588234,
+      "grad_norm": 0.22811627743109275,
+      "learning_rate": 6.914877715562704e-06,
+      "loss": 0.0464,
+      "step": 4407
+    },
+    {
+      "epoch": 2.294638209266007,
+      "grad_norm": 0.2463222783668148,
+      "learning_rate": 6.905180075219025e-06,
+      "loss": 0.0476,
+      "step": 4408
+    },
+    {
+      "epoch": 2.295158771473191,
+      "grad_norm": 0.2494378835645998,
+      "learning_rate": 6.895488150061541e-06,
+      "loss": 0.0481,
+      "step": 4409
+    },
+    {
+      "epoch": 2.2956793336803747,
+      "grad_norm": 0.22371944092940846,
+      "learning_rate": 6.88580194315141e-06,
+      "loss": 0.0456,
+      "step": 4410
+    },
+    {
+      "epoch": 2.2961998958875585,
+      "grad_norm": 0.23726791642543021,
+      "learning_rate": 6.876121457547996e-06,
+      "loss": 0.0477,
+      "step": 4411
+    },
+    {
+      "epoch": 2.2967204580947422,
+      "grad_norm": 0.22025059869257912,
+      "learning_rate": 6.866446696308837e-06,
+      "loss": 0.0444,
+      "step": 4412
+    },
+    {
+      "epoch": 2.297241020301926,
+      "grad_norm": 0.2373689563171938,
+      "learning_rate": 6.856777662489669e-06,
+      "loss": 0.0473,
+      "step": 4413
+    },
+    {
+      "epoch": 2.29776158250911,
+      "grad_norm": 0.22953651254536686,
+      "learning_rate": 6.847114359144427e-06,
+      "loss": 0.0476,
+      "step": 4414
+    },
+    {
+      "epoch": 2.2982821447162936,
+      "grad_norm": 0.2314904782948286,
+      "learning_rate": 6.83745678932523e-06,
+      "loss": 0.0475,
+      "step": 4415
+    },
+    {
+      "epoch": 2.2988027069234773,
+      "grad_norm": 0.23068456271896137,
+      "learning_rate": 6.8278049560824035e-06,
+      "loss": 0.0497,
+      "step": 4416
+    },
+    {
+      "epoch": 2.299323269130661,
+      "grad_norm": 0.22509485065033707,
+      "learning_rate": 6.818158862464422e-06,
+      "loss": 0.0473,
+      "step": 4417
+    },
+    {
+      "epoch": 2.299843831337845,
+      "grad_norm": 0.23754030254959319,
+      "learning_rate": 6.8085185115179836e-06,
+      "loss": 0.0484,
+      "step": 4418
+    },
+    {
+      "epoch": 2.3003643935450286,
+      "grad_norm": 0.23154423326948806,
+      "learning_rate": 6.798883906287956e-06,
+      "loss": 0.0476,
+      "step": 4419
+    },
+    {
+      "epoch": 2.3008849557522124,
+      "grad_norm": 0.2276877249333266,
+      "learning_rate": 6.789255049817406e-06,
+      "loss": 0.0474,
+      "step": 4420
+    },
+    {
+      "epoch": 2.301405517959396,
+      "grad_norm": 0.2455145987692723,
+      "learning_rate": 6.779631945147566e-06,
+      "loss": 0.0484,
+      "step": 4421
+    },
+    {
+      "epoch": 2.30192608016658,
+      "grad_norm": 0.23746214997113513,
+      "learning_rate": 6.770014595317853e-06,
+      "loss": 0.0464,
+      "step": 4422
+    },
+    {
+      "epoch": 2.3024466423737637,
+      "grad_norm": 0.22566035826376565,
+      "learning_rate": 6.760403003365884e-06,
+      "loss": 0.0471,
+      "step": 4423
+    },
+    {
+      "epoch": 2.3029672045809475,
+      "grad_norm": 0.23320331066661007,
+      "learning_rate": 6.750797172327442e-06,
+      "loss": 0.0492,
+      "step": 4424
+    },
+    {
+      "epoch": 2.3034877667881313,
+      "grad_norm": 0.22163307688943074,
+      "learning_rate": 6.741197105236505e-06,
+      "loss": 0.0456,
+      "step": 4425
+    },
+    {
+      "epoch": 2.304008328995315,
+      "grad_norm": 0.24047914367787057,
+      "learning_rate": 6.731602805125206e-06,
+      "loss": 0.0492,
+      "step": 4426
+    },
+    {
+      "epoch": 2.304528891202499,
+      "grad_norm": 0.22844822202025908,
+      "learning_rate": 6.72201427502388e-06,
+      "loss": 0.0477,
+      "step": 4427
+    },
+    {
+      "epoch": 2.3050494534096826,
+      "grad_norm": 0.22813096140516748,
+      "learning_rate": 6.712431517961029e-06,
+      "loss": 0.0472,
+      "step": 4428
+    },
+    {
+      "epoch": 2.3055700156168664,
+      "grad_norm": 0.22811760187365832,
+      "learning_rate": 6.702854536963343e-06,
+      "loss": 0.0472,
+      "step": 4429
+    },
+    {
+      "epoch": 2.30609057782405,
+      "grad_norm": 0.23994023415059773,
+      "learning_rate": 6.69328333505567e-06,
+      "loss": 0.0484,
+      "step": 4430
+    },
+    {
+      "epoch": 2.306611140031234,
+      "grad_norm": 0.23874833178382263,
+      "learning_rate": 6.683717915261034e-06,
+      "loss": 0.0498,
+      "step": 4431
+    },
+    {
+      "epoch": 2.3071317022384177,
+      "grad_norm": 0.2331323610830941,
+      "learning_rate": 6.674158280600645e-06,
+      "loss": 0.0479,
+      "step": 4432
+    },
+    {
+      "epoch": 2.3076522644456015,
+      "grad_norm": 0.23190741443914148,
+      "learning_rate": 6.6646044340938854e-06,
+      "loss": 0.0474,
+      "step": 4433
+    },
+    {
+      "epoch": 2.3081728266527852,
+      "grad_norm": 0.23628413869646864,
+      "learning_rate": 6.655056378758298e-06,
+      "loss": 0.0498,
+      "step": 4434
+    },
+    {
+      "epoch": 2.308693388859969,
+      "grad_norm": 0.2313885805739463,
+      "learning_rate": 6.645514117609616e-06,
+      "loss": 0.0475,
+      "step": 4435
+    },
+    {
+      "epoch": 2.3092139510671528,
+      "grad_norm": 0.2233459799496242,
+      "learning_rate": 6.6359776536617096e-06,
+      "loss": 0.0467,
+      "step": 4436
+    },
+    {
+      "epoch": 2.309734513274336,
+      "grad_norm": 0.2363879717090792,
+      "learning_rate": 6.626446989926652e-06,
+      "loss": 0.0484,
+      "step": 4437
+    },
+    {
+      "epoch": 2.31025507548152,
+      "grad_norm": 0.23027837691324807,
+      "learning_rate": 6.616922129414671e-06,
+      "loss": 0.0462,
+      "step": 4438
+    },
+    {
+      "epoch": 2.3107756376887036,
+      "grad_norm": 0.22061521966437242,
+      "learning_rate": 6.6074030751341496e-06,
+      "loss": 0.046,
+      "step": 4439
+    },
+    {
+      "epoch": 2.3112961998958874,
+      "grad_norm": 0.236489044723267,
+      "learning_rate": 6.597889830091664e-06,
+      "loss": 0.0491,
+      "step": 4440
+    },
+    {
+      "epoch": 2.311816762103071,
+      "grad_norm": 0.2256938496890428,
+      "learning_rate": 6.5883823972919205e-06,
+      "loss": 0.0463,
+      "step": 4441
+    },
+    {
+      "epoch": 2.312337324310255,
+      "grad_norm": 0.23195296992115286,
+      "learning_rate": 6.5788807797378196e-06,
+      "loss": 0.0486,
+      "step": 4442
+    },
+    {
+      "epoch": 2.3128578865174387,
+      "grad_norm": 0.2274872199904384,
+      "learning_rate": 6.569384980430415e-06,
+      "loss": 0.0466,
+      "step": 4443
+    },
+    {
+      "epoch": 2.3133784487246225,
+      "grad_norm": 0.2438659322406287,
+      "learning_rate": 6.559895002368927e-06,
+      "loss": 0.0502,
+      "step": 4444
+    },
+    {
+      "epoch": 2.3138990109318063,
+      "grad_norm": 0.24195755478933637,
+      "learning_rate": 6.5504108485507175e-06,
+      "loss": 0.0493,
+      "step": 4445
+    },
+    {
+      "epoch": 2.31441957313899,
+      "grad_norm": 0.2312187971962166,
+      "learning_rate": 6.5409325219713325e-06,
+      "loss": 0.0468,
+      "step": 4446
+    },
+    {
+      "epoch": 2.314940135346174,
+      "grad_norm": 0.23686834836229043,
+      "learning_rate": 6.531460025624475e-06,
+      "loss": 0.0473,
+      "step": 4447
+    },
+    {
+      "epoch": 2.3154606975533576,
+      "grad_norm": 0.22867574310554584,
+      "learning_rate": 6.521993362501988e-06,
+      "loss": 0.0458,
+      "step": 4448
+    },
+    {
+      "epoch": 2.3159812597605414,
+      "grad_norm": 0.23825031277590675,
+      "learning_rate": 6.512532535593896e-06,
+      "loss": 0.0474,
+      "step": 4449
+    },
+    {
+      "epoch": 2.316501821967725,
+      "grad_norm": 0.23330630762399085,
+      "learning_rate": 6.503077547888353e-06,
+      "loss": 0.047,
+      "step": 4450
+    },
+    {
+      "epoch": 2.317022384174909,
+      "grad_norm": 0.22866429601766594,
+      "learning_rate": 6.493628402371693e-06,
+      "loss": 0.047,
+      "step": 4451
+    },
+    {
+      "epoch": 2.3175429463820927,
+      "grad_norm": 0.23326929579367334,
+      "learning_rate": 6.484185102028398e-06,
+      "loss": 0.0467,
+      "step": 4452
+    },
+    {
+      "epoch": 2.3180635085892765,
+      "grad_norm": 0.21423278359694856,
+      "learning_rate": 6.474747649841103e-06,
+      "loss": 0.0458,
+      "step": 4453
+    },
+    {
+      "epoch": 2.3185840707964602,
+      "grad_norm": 0.22292986350540375,
+      "learning_rate": 6.465316048790587e-06,
+      "loss": 0.0473,
+      "step": 4454
+    },
+    {
+      "epoch": 2.319104633003644,
+      "grad_norm": 0.24046086305383696,
+      "learning_rate": 6.4558903018557936e-06,
+      "loss": 0.0506,
+      "step": 4455
+    },
+    {
+      "epoch": 2.3196251952108278,
+      "grad_norm": 0.23153182365151787,
+      "learning_rate": 6.446470412013817e-06,
+      "loss": 0.0471,
+      "step": 4456
+    },
+    {
+      "epoch": 2.3201457574180115,
+      "grad_norm": 0.2256223599408448,
+      "learning_rate": 6.437056382239884e-06,
+      "loss": 0.0475,
+      "step": 4457
+    },
+    {
+      "epoch": 2.3206663196251953,
+      "grad_norm": 0.22640346868399505,
+      "learning_rate": 6.427648215507398e-06,
+      "loss": 0.0497,
+      "step": 4458
+    },
+    {
+      "epoch": 2.321186881832379,
+      "grad_norm": 0.22663691130270067,
+      "learning_rate": 6.418245914787882e-06,
+      "loss": 0.0451,
+      "step": 4459
+    },
+    {
+      "epoch": 2.321707444039563,
+      "grad_norm": 0.22821336766562575,
+      "learning_rate": 6.408849483051024e-06,
+      "loss": 0.0472,
+      "step": 4460
+    },
+    {
+      "epoch": 2.3222280062467466,
+      "grad_norm": 0.23385635491282356,
+      "learning_rate": 6.399458923264659e-06,
+      "loss": 0.0482,
+      "step": 4461
+    },
+    {
+      "epoch": 2.3227485684539304,
+      "grad_norm": 0.2320806105076779,
+      "learning_rate": 6.3900742383947664e-06,
+      "loss": 0.0459,
+      "step": 4462
+    },
+    {
+      "epoch": 2.323269130661114,
+      "grad_norm": 0.2389305017153277,
+      "learning_rate": 6.380695431405453e-06,
+      "loss": 0.0496,
+      "step": 4463
+    },
+    {
+      "epoch": 2.323789692868298,
+      "grad_norm": 0.22813071098401094,
+      "learning_rate": 6.371322505258992e-06,
+      "loss": 0.0457,
+      "step": 4464
+    },
+    {
+      "epoch": 2.3243102550754813,
+      "grad_norm": 0.24014119920977195,
+      "learning_rate": 6.361955462915795e-06,
+      "loss": 0.0474,
+      "step": 4465
+    },
+    {
+      "epoch": 2.324830817282665,
+      "grad_norm": 0.2416415435940098,
+      "learning_rate": 6.352594307334395e-06,
+      "loss": 0.0473,
+      "step": 4466
+    },
+    {
+      "epoch": 2.325351379489849,
+      "grad_norm": 0.2656651285253611,
+      "learning_rate": 6.343239041471497e-06,
+      "loss": 0.0479,
+      "step": 4467
+    },
+    {
+      "epoch": 2.3258719416970326,
+      "grad_norm": 0.23550173668414345,
+      "learning_rate": 6.333889668281912e-06,
+      "loss": 0.0473,
+      "step": 4468
+    },
+    {
+      "epoch": 2.3263925039042164,
+      "grad_norm": 0.23828823064619004,
+      "learning_rate": 6.324546190718614e-06,
+      "loss": 0.0457,
+      "step": 4469
+    },
+    {
+      "epoch": 2.3269130661114,
+      "grad_norm": 0.24471560830756434,
+      "learning_rate": 6.3152086117327116e-06,
+      "loss": 0.0488,
+      "step": 4470
+    },
+    {
+      "epoch": 2.327433628318584,
+      "grad_norm": 0.22933514872804697,
+      "learning_rate": 6.305876934273452e-06,
+      "loss": 0.0468,
+      "step": 4471
+    },
+    {
+      "epoch": 2.3279541905257677,
+      "grad_norm": 0.23426764882810178,
+      "learning_rate": 6.296551161288197e-06,
+      "loss": 0.0485,
+      "step": 4472
+    },
+    {
+      "epoch": 2.3284747527329515,
+      "grad_norm": 0.22856682242539877,
+      "learning_rate": 6.28723129572247e-06,
+      "loss": 0.0462,
+      "step": 4473
+    },
+    {
+      "epoch": 2.3289953149401352,
+      "grad_norm": 0.23846002941423738,
+      "learning_rate": 6.277917340519918e-06,
+      "loss": 0.0482,
+      "step": 4474
+    },
+    {
+      "epoch": 2.329515877147319,
+      "grad_norm": 0.234480561210467,
+      "learning_rate": 6.268609298622327e-06,
+      "loss": 0.0482,
+      "step": 4475
+    },
+    {
+      "epoch": 2.3300364393545028,
+      "grad_norm": 0.24047174883310446,
+      "learning_rate": 6.259307172969606e-06,
+      "loss": 0.0477,
+      "step": 4476
+    },
+    {
+      "epoch": 2.3305570015616865,
+      "grad_norm": 0.2344548406347454,
+      "learning_rate": 6.250010966499786e-06,
+      "loss": 0.0484,
+      "step": 4477
+    },
+    {
+      "epoch": 2.3310775637688703,
+      "grad_norm": 0.22642383076214023,
+      "learning_rate": 6.240720682149054e-06,
+      "loss": 0.0471,
+      "step": 4478
+    },
+    {
+      "epoch": 2.331598125976054,
+      "grad_norm": 0.22569085877899356,
+      "learning_rate": 6.231436322851711e-06,
+      "loss": 0.0465,
+      "step": 4479
+    },
+    {
+      "epoch": 2.332118688183238,
+      "grad_norm": 0.21706380279472942,
+      "learning_rate": 6.222157891540198e-06,
+      "loss": 0.0465,
+      "step": 4480
+    },
+    {
+      "epoch": 2.3326392503904216,
+      "grad_norm": 0.22733120940204257,
+      "learning_rate": 6.21288539114506e-06,
+      "loss": 0.0475,
+      "step": 4481
+    },
+    {
+      "epoch": 2.3331598125976054,
+      "grad_norm": 0.22939434669257294,
+      "learning_rate": 6.203618824594995e-06,
+      "loss": 0.0463,
+      "step": 4482
+    },
+    {
+      "epoch": 2.333680374804789,
+      "grad_norm": 0.23734452631886388,
+      "learning_rate": 6.194358194816813e-06,
+      "loss": 0.0477,
+      "step": 4483
+    },
+    {
+      "epoch": 2.334200937011973,
+      "grad_norm": 0.23411419581489817,
+      "learning_rate": 6.1851035047354595e-06,
+      "loss": 0.0477,
+      "step": 4484
+    },
+    {
+      "epoch": 2.3347214992191567,
+      "grad_norm": 0.24002134917837561,
+      "learning_rate": 6.175854757273989e-06,
+      "loss": 0.0481,
+      "step": 4485
+    },
+    {
+      "epoch": 2.3352420614263405,
+      "grad_norm": 0.2370166595867462,
+      "learning_rate": 6.166611955353577e-06,
+      "loss": 0.0506,
+      "step": 4486
+    },
+    {
+      "epoch": 2.3357626236335243,
+      "grad_norm": 0.23745847434442083,
+      "learning_rate": 6.157375101893543e-06,
+      "loss": 0.047,
+      "step": 4487
+    },
+    {
+      "epoch": 2.336283185840708,
+      "grad_norm": 0.24957266727201047,
+      "learning_rate": 6.148144199811309e-06,
+      "loss": 0.0479,
+      "step": 4488
+    },
+    {
+      "epoch": 2.336803748047892,
+      "grad_norm": 0.2351547804630439,
+      "learning_rate": 6.138919252022435e-06,
+      "loss": 0.0474,
+      "step": 4489
+    },
+    {
+      "epoch": 2.3373243102550756,
+      "grad_norm": 0.24163133481886173,
+      "learning_rate": 6.129700261440574e-06,
+      "loss": 0.0475,
+      "step": 4490
+    },
+    {
+      "epoch": 2.3378448724622594,
+      "grad_norm": 0.24059572984473415,
+      "learning_rate": 6.120487230977517e-06,
+      "loss": 0.0502,
+      "step": 4491
+    },
+    {
+      "epoch": 2.338365434669443,
+      "grad_norm": 0.2355645081535075,
+      "learning_rate": 6.1112801635431704e-06,
+      "loss": 0.0497,
+      "step": 4492
+    },
+    {
+      "epoch": 2.338885996876627,
+      "grad_norm": 0.23939602704815524,
+      "learning_rate": 6.102079062045559e-06,
+      "loss": 0.0479,
+      "step": 4493
+    },
+    {
+      "epoch": 2.3394065590838107,
+      "grad_norm": 0.22977889226796178,
+      "learning_rate": 6.092883929390816e-06,
+      "loss": 0.0456,
+      "step": 4494
+    },
+    {
+      "epoch": 2.3399271212909944,
+      "grad_norm": 0.23478062109634587,
+      "learning_rate": 6.083694768483181e-06,
+      "loss": 0.0481,
+      "step": 4495
+    },
+    {
+      "epoch": 2.340447683498178,
+      "grad_norm": 0.22765522951948453,
+      "learning_rate": 6.074511582225029e-06,
+      "loss": 0.0484,
+      "step": 4496
+    },
+    {
+      "epoch": 2.340968245705362,
+      "grad_norm": 0.21743859076143826,
+      "learning_rate": 6.065334373516834e-06,
+      "loss": 0.0471,
+      "step": 4497
+    },
+    {
+      "epoch": 2.3414888079125458,
+      "grad_norm": 0.2314187819934621,
+      "learning_rate": 6.056163145257187e-06,
+      "loss": 0.0484,
+      "step": 4498
+    },
+    {
+      "epoch": 2.3420093701197295,
+      "grad_norm": 0.23842056215195048,
+      "learning_rate": 6.046997900342796e-06,
+      "loss": 0.048,
+      "step": 4499
+    },
+    {
+      "epoch": 2.3425299323269133,
+      "grad_norm": 0.22448513602756287,
+      "learning_rate": 6.037838641668459e-06,
+      "loss": 0.0459,
+      "step": 4500
+    },
+    {
+      "epoch": 2.3430504945340966,
+      "grad_norm": 0.2236326297612647,
+      "learning_rate": 6.028685372127099e-06,
+      "loss": 0.0469,
+      "step": 4501
+    },
+    {
+      "epoch": 2.3435710567412804,
+      "grad_norm": 0.22213192913695828,
+      "learning_rate": 6.019538094609759e-06,
+      "loss": 0.0466,
+      "step": 4502
+    },
+    {
+      "epoch": 2.344091618948464,
+      "grad_norm": 0.22847968405344984,
+      "learning_rate": 6.010396812005553e-06,
+      "loss": 0.0468,
+      "step": 4503
+    },
+    {
+      "epoch": 2.344612181155648,
+      "grad_norm": 0.23492086518084526,
+      "learning_rate": 6.00126152720174e-06,
+      "loss": 0.047,
+      "step": 4504
+    },
+    {
+      "epoch": 2.3451327433628317,
+      "grad_norm": 0.2394105213441764,
+      "learning_rate": 5.992132243083656e-06,
+      "loss": 0.0496,
+      "step": 4505
+    },
+    {
+      "epoch": 2.3456533055700155,
+      "grad_norm": 0.24111441487478028,
+      "learning_rate": 5.983008962534761e-06,
+      "loss": 0.0461,
+      "step": 4506
+    },
+    {
+      "epoch": 2.3461738677771993,
+      "grad_norm": 0.23107118946181654,
+      "learning_rate": 5.973891688436608e-06,
+      "loss": 0.0476,
+      "step": 4507
+    },
+    {
+      "epoch": 2.346694429984383,
+      "grad_norm": 0.23065663806986544,
+      "learning_rate": 5.964780423668867e-06,
+      "loss": 0.0481,
+      "step": 4508
+    },
+    {
+      "epoch": 2.347214992191567,
+      "grad_norm": 0.23820356615082777,
+      "learning_rate": 5.9556751711092844e-06,
+      "loss": 0.0472,
+      "step": 4509
+    },
+    {
+      "epoch": 2.3477355543987506,
+      "grad_norm": 0.2365896072704537,
+      "learning_rate": 5.94657593363373e-06,
+      "loss": 0.0473,
+      "step": 4510
+    },
+    {
+      "epoch": 2.3482561166059344,
+      "grad_norm": 0.23479232602004582,
+      "learning_rate": 5.9374827141161715e-06,
+      "loss": 0.0489,
+      "step": 4511
+    },
+    {
+      "epoch": 2.348776678813118,
+      "grad_norm": 0.23324065251358814,
+      "learning_rate": 5.928395515428656e-06,
+      "loss": 0.0469,
+      "step": 4512
+    },
+    {
+      "epoch": 2.349297241020302,
+      "grad_norm": 0.2271112729337206,
+      "learning_rate": 5.919314340441362e-06,
+      "loss": 0.0462,
+      "step": 4513
+    },
+    {
+      "epoch": 2.3498178032274857,
+      "grad_norm": 0.23075689196359045,
+      "learning_rate": 5.9102391920225315e-06,
+      "loss": 0.0484,
+      "step": 4514
+    },
+    {
+      "epoch": 2.3503383654346695,
+      "grad_norm": 0.22572112690840151,
+      "learning_rate": 5.901170073038523e-06,
+      "loss": 0.0466,
+      "step": 4515
+    },
+    {
+      "epoch": 2.3508589276418532,
+      "grad_norm": 0.23256450262742248,
+      "learning_rate": 5.89210698635379e-06,
+      "loss": 0.0461,
+      "step": 4516
+    },
+    {
+      "epoch": 2.351379489849037,
+      "grad_norm": 0.220207578677318,
+      "learning_rate": 5.883049934830884e-06,
+      "loss": 0.0465,
+      "step": 4517
+    },
+    {
+      "epoch": 2.3519000520562208,
+      "grad_norm": 0.22721249545677813,
+      "learning_rate": 5.873998921330426e-06,
+      "loss": 0.0455,
+      "step": 4518
+    },
+    {
+      "epoch": 2.3524206142634045,
+      "grad_norm": 0.2632945933248785,
+      "learning_rate": 5.864953948711155e-06,
+      "loss": 0.0481,
+      "step": 4519
+    },
+    {
+      "epoch": 2.3529411764705883,
+      "grad_norm": 0.23515126483913215,
+      "learning_rate": 5.855915019829902e-06,
+      "loss": 0.0469,
+      "step": 4520
+    },
+    {
+      "epoch": 2.353461738677772,
+      "grad_norm": 0.22931919368137274,
+      "learning_rate": 5.846882137541571e-06,
+      "loss": 0.0474,
+      "step": 4521
+    },
+    {
+      "epoch": 2.353982300884956,
+      "grad_norm": 0.24606325790404943,
+      "learning_rate": 5.837855304699175e-06,
+      "loss": 0.0484,
+      "step": 4522
+    },
+    {
+      "epoch": 2.3545028630921396,
+      "grad_norm": 0.23172001395098654,
+      "learning_rate": 5.828834524153795e-06,
+      "loss": 0.0475,
+      "step": 4523
+    },
+    {
+      "epoch": 2.3550234252993234,
+      "grad_norm": 0.21936760204415123,
+      "learning_rate": 5.819819798754625e-06,
+      "loss": 0.0458,
+      "step": 4524
+    },
+    {
+      "epoch": 2.355543987506507,
+      "grad_norm": 0.2193218125572941,
+      "learning_rate": 5.810811131348929e-06,
+      "loss": 0.0459,
+      "step": 4525
+    },
+    {
+      "epoch": 2.356064549713691,
+      "grad_norm": 0.2345064969588428,
+      "learning_rate": 5.801808524782071e-06,
+      "loss": 0.0468,
+      "step": 4526
+    },
+    {
+      "epoch": 2.3565851119208747,
+      "grad_norm": 0.24851347312329655,
+      "learning_rate": 5.792811981897484e-06,
+      "loss": 0.0492,
+      "step": 4527
+    },
+    {
+      "epoch": 2.3571056741280585,
+      "grad_norm": 0.23050774006520552,
+      "learning_rate": 5.783821505536696e-06,
+      "loss": 0.0467,
+      "step": 4528
+    },
+    {
+      "epoch": 2.357626236335242,
+      "grad_norm": 0.23084948710195863,
+      "learning_rate": 5.7748370985393295e-06,
+      "loss": 0.0454,
+      "step": 4529
+    },
+    {
+      "epoch": 2.3581467985424256,
+      "grad_norm": 0.24791040020859287,
+      "learning_rate": 5.765858763743062e-06,
+      "loss": 0.0476,
+      "step": 4530
+    },
+    {
+      "epoch": 2.3586673607496094,
+      "grad_norm": 0.23318412091442703,
+      "learning_rate": 5.756886503983683e-06,
+      "loss": 0.0475,
+      "step": 4531
+    },
+    {
+      "epoch": 2.359187922956793,
+      "grad_norm": 0.23289524027494776,
+      "learning_rate": 5.747920322095035e-06,
+      "loss": 0.0464,
+      "step": 4532
+    },
+    {
+      "epoch": 2.359708485163977,
+      "grad_norm": 0.23663155315479364,
+      "learning_rate": 5.738960220909068e-06,
+      "loss": 0.0465,
+      "step": 4533
+    },
+    {
+      "epoch": 2.3602290473711607,
+      "grad_norm": 0.2330747611757856,
+      "learning_rate": 5.730006203255792e-06,
+      "loss": 0.0466,
+      "step": 4534
+    },
+    {
+      "epoch": 2.3607496095783445,
+      "grad_norm": 0.23854565465214606,
+      "learning_rate": 5.721058271963311e-06,
+      "loss": 0.0466,
+      "step": 4535
+    },
+    {
+      "epoch": 2.3612701717855282,
+      "grad_norm": 0.23694164646216895,
+      "learning_rate": 5.712116429857789e-06,
+      "loss": 0.0473,
+      "step": 4536
+    },
+    {
+      "epoch": 2.361790733992712,
+      "grad_norm": 0.23150554800278614,
+      "learning_rate": 5.7031806797634755e-06,
+      "loss": 0.0466,
+      "step": 4537
+    },
+    {
+      "epoch": 2.3623112961998958,
+      "grad_norm": 0.23630006270170592,
+      "learning_rate": 5.694251024502709e-06,
+      "loss": 0.0484,
+      "step": 4538
+    },
+    {
+      "epoch": 2.3628318584070795,
+      "grad_norm": 0.2338320873775774,
+      "learning_rate": 5.685327466895874e-06,
+      "loss": 0.0471,
+      "step": 4539
+    },
+    {
+      "epoch": 2.3633524206142633,
+      "grad_norm": 0.2377908975210259,
+      "learning_rate": 5.6764100097614595e-06,
+      "loss": 0.0488,
+      "step": 4540
+    },
+    {
+      "epoch": 2.363872982821447,
+      "grad_norm": 0.23304039641551305,
+      "learning_rate": 5.667498655916001e-06,
+      "loss": 0.0462,
+      "step": 4541
+    },
+    {
+      "epoch": 2.364393545028631,
+      "grad_norm": 0.23372925814357076,
+      "learning_rate": 5.6585934081741205e-06,
+      "loss": 0.0471,
+      "step": 4542
+    },
+    {
+      "epoch": 2.3649141072358146,
+      "grad_norm": 0.23054147188963464,
+      "learning_rate": 5.649694269348516e-06,
+      "loss": 0.0487,
+      "step": 4543
+    },
+    {
+      "epoch": 2.3654346694429984,
+      "grad_norm": 0.2295991682759455,
+      "learning_rate": 5.640801242249952e-06,
+      "loss": 0.0466,
+      "step": 4544
+    },
+    {
+      "epoch": 2.365955231650182,
+      "grad_norm": 0.23056386741672127,
+      "learning_rate": 5.631914329687249e-06,
+      "loss": 0.0478,
+      "step": 4545
+    },
+    {
+      "epoch": 2.366475793857366,
+      "grad_norm": 0.22736615156094703,
+      "learning_rate": 5.623033534467315e-06,
+      "loss": 0.0459,
+      "step": 4546
+    },
+    {
+      "epoch": 2.3669963560645497,
+      "grad_norm": 0.23702181937720163,
+      "learning_rate": 5.614158859395122e-06,
+      "loss": 0.0481,
+      "step": 4547
+    },
+    {
+      "epoch": 2.3675169182717335,
+      "grad_norm": 0.2266535316694567,
+      "learning_rate": 5.605290307273694e-06,
+      "loss": 0.046,
+      "step": 4548
+    },
+    {
+      "epoch": 2.3680374804789173,
+      "grad_norm": 0.23582655875491848,
+      "learning_rate": 5.596427880904148e-06,
+      "loss": 0.0472,
+      "step": 4549
+    },
+    {
+      "epoch": 2.368558042686101,
+      "grad_norm": 0.24143542099238016,
+      "learning_rate": 5.587571583085632e-06,
+      "loss": 0.0467,
+      "step": 4550
+    },
+    {
+      "epoch": 2.369078604893285,
+      "grad_norm": 0.2307564717560343,
+      "learning_rate": 5.5787214166153875e-06,
+      "loss": 0.0467,
+      "step": 4551
+    },
+    {
+      "epoch": 2.3695991671004686,
+      "grad_norm": 0.24506784373841559,
+      "learning_rate": 5.569877384288708e-06,
+      "loss": 0.0474,
+      "step": 4552
+    },
+    {
+      "epoch": 2.3701197293076524,
+      "grad_norm": 0.32435382511813377,
+      "learning_rate": 5.5610394888989585e-06,
+      "loss": 0.0476,
+      "step": 4553
+    },
+    {
+      "epoch": 2.370640291514836,
+      "grad_norm": 0.2342486669021339,
+      "learning_rate": 5.5522077332375436e-06,
+      "loss": 0.0474,
+      "step": 4554
+    },
+    {
+      "epoch": 2.37116085372202,
+      "grad_norm": 0.21410206999675588,
+      "learning_rate": 5.543382120093946e-06,
+      "loss": 0.0442,
+      "step": 4555
+    },
+    {
+      "epoch": 2.3716814159292037,
+      "grad_norm": 0.2338596233871188,
+      "learning_rate": 5.5345626522557175e-06,
+      "loss": 0.0455,
+      "step": 4556
+    },
+    {
+      "epoch": 2.3722019781363874,
+      "grad_norm": 0.23916410275167593,
+      "learning_rate": 5.525749332508437e-06,
+      "loss": 0.0469,
+      "step": 4557
+    },
+    {
+      "epoch": 2.372722540343571,
+      "grad_norm": 0.23457736048555594,
+      "learning_rate": 5.51694216363578e-06,
+      "loss": 0.0456,
+      "step": 4558
+    },
+    {
+      "epoch": 2.373243102550755,
+      "grad_norm": 0.25322127032434727,
+      "learning_rate": 5.5081411484194435e-06,
+      "loss": 0.048,
+      "step": 4559
+    },
+    {
+      "epoch": 2.3737636647579388,
+      "grad_norm": 0.24019282093163075,
+      "learning_rate": 5.499346289639206e-06,
+      "loss": 0.0466,
+      "step": 4560
+    },
+    {
+      "epoch": 2.3742842269651225,
+      "grad_norm": 0.2264918241309705,
+      "learning_rate": 5.490557590072892e-06,
+      "loss": 0.0454,
+      "step": 4561
+    },
+    {
+      "epoch": 2.3748047891723063,
+      "grad_norm": 0.2357276481863,
+      "learning_rate": 5.4817750524963904e-06,
+      "loss": 0.047,
+      "step": 4562
+    },
+    {
+      "epoch": 2.37532535137949,
+      "grad_norm": 0.23230594226469076,
+      "learning_rate": 5.472998679683619e-06,
+      "loss": 0.0465,
+      "step": 4563
+    },
+    {
+      "epoch": 2.375845913586674,
+      "grad_norm": 0.22533858615611319,
+      "learning_rate": 5.4642284744065715e-06,
+      "loss": 0.0468,
+      "step": 4564
+    },
+    {
+      "epoch": 2.376366475793857,
+      "grad_norm": 0.22724858506067902,
+      "learning_rate": 5.455464439435299e-06,
+      "loss": 0.0481,
+      "step": 4565
+    },
+    {
+      "epoch": 2.376887038001041,
+      "grad_norm": 0.23606749438431365,
+      "learning_rate": 5.446706577537869e-06,
+      "loss": 0.0466,
+      "step": 4566
+    },
+    {
+      "epoch": 2.3774076002082247,
+      "grad_norm": 0.2370377994578295,
+      "learning_rate": 5.437954891480443e-06,
+      "loss": 0.0494,
+      "step": 4567
+    },
+    {
+      "epoch": 2.3779281624154085,
+      "grad_norm": 0.2240890350805136,
+      "learning_rate": 5.4292093840271955e-06,
+      "loss": 0.0458,
+      "step": 4568
+    },
+    {
+      "epoch": 2.3784487246225923,
+      "grad_norm": 0.2289474671965338,
+      "learning_rate": 5.420470057940372e-06,
+      "loss": 0.0467,
+      "step": 4569
+    },
+    {
+      "epoch": 2.378969286829776,
+      "grad_norm": 0.2298982841879016,
+      "learning_rate": 5.411736915980253e-06,
+      "loss": 0.049,
+      "step": 4570
+    },
+    {
+      "epoch": 2.37948984903696,
+      "grad_norm": 0.2268863902053528,
+      "learning_rate": 5.403009960905178e-06,
+      "loss": 0.0461,
+      "step": 4571
+    },
+    {
+      "epoch": 2.3800104112441436,
+      "grad_norm": 0.2436459860162038,
+      "learning_rate": 5.394289195471527e-06,
+      "loss": 0.0487,
+      "step": 4572
+    },
+    {
+      "epoch": 2.3805309734513274,
+      "grad_norm": 0.2632454708511737,
+      "learning_rate": 5.385574622433714e-06,
+      "loss": 0.0469,
+      "step": 4573
+    },
+    {
+      "epoch": 2.381051535658511,
+      "grad_norm": 0.2301709888777297,
+      "learning_rate": 5.3768662445442204e-06,
+      "loss": 0.0475,
+      "step": 4574
+    },
+    {
+      "epoch": 2.381572097865695,
+      "grad_norm": 0.22767022475445184,
+      "learning_rate": 5.368164064553541e-06,
+      "loss": 0.0473,
+      "step": 4575
+    },
+    {
+      "epoch": 2.3820926600728787,
+      "grad_norm": 0.22628560424820487,
+      "learning_rate": 5.359468085210237e-06,
+      "loss": 0.045,
+      "step": 4576
+    },
+    {
+      "epoch": 2.3826132222800624,
+      "grad_norm": 0.23342340422854826,
+      "learning_rate": 5.3507783092609095e-06,
+      "loss": 0.0486,
+      "step": 4577
+    },
+    {
+      "epoch": 2.383133784487246,
+      "grad_norm": 0.2245786257913053,
+      "learning_rate": 5.342094739450179e-06,
+      "loss": 0.0464,
+      "step": 4578
+    },
+    {
+      "epoch": 2.38365434669443,
+      "grad_norm": 0.2302258019263189,
+      "learning_rate": 5.333417378520733e-06,
+      "loss": 0.0474,
+      "step": 4579
+    },
+    {
+      "epoch": 2.3841749089016138,
+      "grad_norm": 0.2268167472789496,
+      "learning_rate": 5.324746229213282e-06,
+      "loss": 0.0455,
+      "step": 4580
+    },
+    {
+      "epoch": 2.3846954711087975,
+      "grad_norm": 0.235065648886395,
+      "learning_rate": 5.316081294266587e-06,
+      "loss": 0.0476,
+      "step": 4581
+    },
+    {
+      "epoch": 2.3852160333159813,
+      "grad_norm": 0.23538123207470538,
+      "learning_rate": 5.30742257641742e-06,
+      "loss": 0.0481,
+      "step": 4582
+    },
+    {
+      "epoch": 2.385736595523165,
+      "grad_norm": 0.23818619605234898,
+      "learning_rate": 5.298770078400628e-06,
+      "loss": 0.0468,
+      "step": 4583
+    },
+    {
+      "epoch": 2.386257157730349,
+      "grad_norm": 0.2320901545098414,
+      "learning_rate": 5.290123802949051e-06,
+      "loss": 0.0451,
+      "step": 4584
+    },
+    {
+      "epoch": 2.3867777199375326,
+      "grad_norm": 0.23469735771174544,
+      "learning_rate": 5.2814837527935975e-06,
+      "loss": 0.047,
+      "step": 4585
+    },
+    {
+      "epoch": 2.3872982821447164,
+      "grad_norm": 0.23767900812756343,
+      "learning_rate": 5.272849930663204e-06,
+      "loss": 0.0489,
+      "step": 4586
+    },
+    {
+      "epoch": 2.3878188443519,
+      "grad_norm": 0.23370065886236566,
+      "learning_rate": 5.264222339284816e-06,
+      "loss": 0.0461,
+      "step": 4587
+    },
+    {
+      "epoch": 2.388339406559084,
+      "grad_norm": 0.23170361003181508,
+      "learning_rate": 5.255600981383438e-06,
+      "loss": 0.0476,
+      "step": 4588
+    },
+    {
+      "epoch": 2.3888599687662677,
+      "grad_norm": 0.23695214156506306,
+      "learning_rate": 5.246985859682094e-06,
+      "loss": 0.0481,
+      "step": 4589
+    },
+    {
+      "epoch": 2.3893805309734515,
+      "grad_norm": 0.22780483346006977,
+      "learning_rate": 5.238376976901849e-06,
+      "loss": 0.0471,
+      "step": 4590
+    },
+    {
+      "epoch": 2.3899010931806353,
+      "grad_norm": 0.23083988277274833,
+      "learning_rate": 5.229774335761775e-06,
+      "loss": 0.0459,
+      "step": 4591
+    },
+    {
+      "epoch": 2.390421655387819,
+      "grad_norm": 0.22986426177716607,
+      "learning_rate": 5.221177938978999e-06,
+      "loss": 0.0446,
+      "step": 4592
+    },
+    {
+      "epoch": 2.3909422175950024,
+      "grad_norm": 0.23342974438965086,
+      "learning_rate": 5.2125877892686496e-06,
+      "loss": 0.0472,
+      "step": 4593
+    },
+    {
+      "epoch": 2.391462779802186,
+      "grad_norm": 0.23195959114003814,
+      "learning_rate": 5.204003889343906e-06,
+      "loss": 0.0461,
+      "step": 4594
+    },
+    {
+      "epoch": 2.39198334200937,
+      "grad_norm": 0.2216035633570572,
+      "learning_rate": 5.195426241915963e-06,
+      "loss": 0.045,
+      "step": 4595
+    },
+    {
+      "epoch": 2.3925039042165537,
+      "grad_norm": 0.2132065666076437,
+      "learning_rate": 5.186854849694034e-06,
+      "loss": 0.0445,
+      "step": 4596
+    },
+    {
+      "epoch": 2.3930244664237375,
+      "grad_norm": 0.23129878723213484,
+      "learning_rate": 5.178289715385368e-06,
+      "loss": 0.0469,
+      "step": 4597
+    },
+    {
+      "epoch": 2.3935450286309212,
+      "grad_norm": 0.2388488336500406,
+      "learning_rate": 5.169730841695233e-06,
+      "loss": 0.047,
+      "step": 4598
+    },
+    {
+      "epoch": 2.394065590838105,
+      "grad_norm": 0.22989393580863302,
+      "learning_rate": 5.161178231326927e-06,
+      "loss": 0.046,
+      "step": 4599
+    },
+    {
+      "epoch": 2.3945861530452888,
+      "grad_norm": 0.24387299564354334,
+      "learning_rate": 5.152631886981746e-06,
+      "loss": 0.0478,
+      "step": 4600
+    },
+    {
+      "epoch": 2.3951067152524725,
+      "grad_norm": 0.2184126453644583,
+      "learning_rate": 5.144091811359039e-06,
+      "loss": 0.0437,
+      "step": 4601
+    },
+    {
+      "epoch": 2.3956272774596563,
+      "grad_norm": 0.23254797708689787,
+      "learning_rate": 5.1355580071561465e-06,
+      "loss": 0.0455,
+      "step": 4602
+    },
+    {
+      "epoch": 2.39614783966684,
+      "grad_norm": 0.24362953858535674,
+      "learning_rate": 5.127030477068445e-06,
+      "loss": 0.0462,
+      "step": 4603
+    },
+    {
+      "epoch": 2.396668401874024,
+      "grad_norm": 0.23599914959344417,
+      "learning_rate": 5.118509223789336e-06,
+      "loss": 0.0467,
+      "step": 4604
+    },
+    {
+      "epoch": 2.3971889640812076,
+      "grad_norm": 0.22043950196000547,
+      "learning_rate": 5.109994250010211e-06,
+      "loss": 0.0439,
+      "step": 4605
+    },
+    {
+      "epoch": 2.3977095262883914,
+      "grad_norm": 0.23070921769768574,
+      "learning_rate": 5.101485558420505e-06,
+      "loss": 0.0467,
+      "step": 4606
+    },
+    {
+      "epoch": 2.398230088495575,
+      "grad_norm": 0.24197283057250074,
+      "learning_rate": 5.092983151707656e-06,
+      "loss": 0.0481,
+      "step": 4607
+    },
+    {
+      "epoch": 2.398750650702759,
+      "grad_norm": 0.22154942089827273,
+      "learning_rate": 5.0844870325571255e-06,
+      "loss": 0.0448,
+      "step": 4608
+    },
+    {
+      "epoch": 2.3992712129099427,
+      "grad_norm": 0.24014632166672964,
+      "learning_rate": 5.0759972036523715e-06,
+      "loss": 0.0479,
+      "step": 4609
+    },
+    {
+      "epoch": 2.3997917751171265,
+      "grad_norm": 0.22536943970116005,
+      "learning_rate": 5.067513667674892e-06,
+      "loss": 0.0457,
+      "step": 4610
+    },
+    {
+      "epoch": 2.4003123373243103,
+      "grad_norm": 0.2250611130271378,
+      "learning_rate": 5.059036427304167e-06,
+      "loss": 0.0447,
+      "step": 4611
+    },
+    {
+      "epoch": 2.400832899531494,
+      "grad_norm": 0.2408162294102492,
+      "learning_rate": 5.050565485217712e-06,
+      "loss": 0.0477,
+      "step": 4612
+    },
+    {
+      "epoch": 2.401353461738678,
+      "grad_norm": 0.24625623591280466,
+      "learning_rate": 5.04210084409105e-06,
+      "loss": 0.0484,
+      "step": 4613
+    },
+    {
+      "epoch": 2.4018740239458616,
+      "grad_norm": 0.2481755981581284,
+      "learning_rate": 5.033642506597694e-06,
+      "loss": 0.048,
+      "step": 4614
+    },
+    {
+      "epoch": 2.4023945861530454,
+      "grad_norm": 0.23987593424996412,
+      "learning_rate": 5.025190475409189e-06,
+      "loss": 0.0468,
+      "step": 4615
+    },
+    {
+      "epoch": 2.402915148360229,
+      "grad_norm": 0.23074712648149856,
+      "learning_rate": 5.01674475319508e-06,
+      "loss": 0.0473,
+      "step": 4616
+    },
+    {
+      "epoch": 2.403435710567413,
+      "grad_norm": 0.22920234691129646,
+      "learning_rate": 5.008305342622923e-06,
+      "loss": 0.0463,
+      "step": 4617
+    },
+    {
+      "epoch": 2.4039562727745967,
+      "grad_norm": 0.24431092684106212,
+      "learning_rate": 4.99987224635827e-06,
+      "loss": 0.0486,
+      "step": 4618
+    },
+    {
+      "epoch": 2.4044768349817804,
+      "grad_norm": 0.24071342837263257,
+      "learning_rate": 4.99144546706469e-06,
+      "loss": 0.0482,
+      "step": 4619
+    },
+    {
+      "epoch": 2.404997397188964,
+      "grad_norm": 0.22487442588291262,
+      "learning_rate": 4.9830250074037435e-06,
+      "loss": 0.0463,
+      "step": 4620
+    },
+    {
+      "epoch": 2.405517959396148,
+      "grad_norm": 0.23677496930387812,
+      "learning_rate": 4.97461087003501e-06,
+      "loss": 0.0472,
+      "step": 4621
+    },
+    {
+      "epoch": 2.4060385216033318,
+      "grad_norm": 0.2339609503507166,
+      "learning_rate": 4.966203057616073e-06,
+      "loss": 0.0463,
+      "step": 4622
+    },
+    {
+      "epoch": 2.4065590838105155,
+      "grad_norm": 0.22648734693696182,
+      "learning_rate": 4.9578015728024955e-06,
+      "loss": 0.0452,
+      "step": 4623
+    },
+    {
+      "epoch": 2.4070796460176993,
+      "grad_norm": 0.23140249756070344,
+      "learning_rate": 4.949406418247865e-06,
+      "loss": 0.046,
+      "step": 4624
+    },
+    {
+      "epoch": 2.407600208224883,
+      "grad_norm": 0.22814922822292663,
+      "learning_rate": 4.941017596603761e-06,
+      "loss": 0.0468,
+      "step": 4625
+    },
+    {
+      "epoch": 2.408120770432067,
+      "grad_norm": 0.24189988700195564,
+      "learning_rate": 4.9326351105197704e-06,
+      "loss": 0.0486,
+      "step": 4626
+    },
+    {
+      "epoch": 2.4086413326392506,
+      "grad_norm": 0.2363782391459408,
+      "learning_rate": 4.92425896264346e-06,
+      "loss": 0.0477,
+      "step": 4627
+    },
+    {
+      "epoch": 2.4091618948464344,
+      "grad_norm": 0.23153910984513215,
+      "learning_rate": 4.915889155620423e-06,
+      "loss": 0.0483,
+      "step": 4628
+    },
+    {
+      "epoch": 2.4096824570536177,
+      "grad_norm": 0.23130021099353434,
+      "learning_rate": 4.907525692094217e-06,
+      "loss": 0.0474,
+      "step": 4629
+    },
+    {
+      "epoch": 2.4102030192608015,
+      "grad_norm": 0.25831436035426264,
+      "learning_rate": 4.89916857470642e-06,
+      "loss": 0.0506,
+      "step": 4630
+    },
+    {
+      "epoch": 2.4107235814679853,
+      "grad_norm": 0.2321273528575011,
+      "learning_rate": 4.890817806096606e-06,
+      "loss": 0.049,
+      "step": 4631
+    },
+    {
+      "epoch": 2.411244143675169,
+      "grad_norm": 0.23553678312517642,
+      "learning_rate": 4.882473388902323e-06,
+      "loss": 0.0473,
+      "step": 4632
+    },
+    {
+      "epoch": 2.411764705882353,
+      "grad_norm": 0.23546720700809365,
+      "learning_rate": 4.874135325759133e-06,
+      "loss": 0.0479,
+      "step": 4633
+    },
+    {
+      "epoch": 2.4122852680895366,
+      "grad_norm": 0.24113411652212086,
+      "learning_rate": 4.8658036193005855e-06,
+      "loss": 0.047,
+      "step": 4634
+    },
+    {
+      "epoch": 2.4128058302967204,
+      "grad_norm": 0.2324247954219874,
+      "learning_rate": 4.857478272158217e-06,
+      "loss": 0.0481,
+      "step": 4635
+    },
+    {
+      "epoch": 2.413326392503904,
+      "grad_norm": 0.23429448270749925,
+      "learning_rate": 4.849159286961571e-06,
+      "loss": 0.0474,
+      "step": 4636
+    },
+    {
+      "epoch": 2.413846954711088,
+      "grad_norm": 0.23663034771539931,
+      "learning_rate": 4.840846666338161e-06,
+      "loss": 0.047,
+      "step": 4637
+    },
+    {
+      "epoch": 2.4143675169182717,
+      "grad_norm": 0.2270730528314231,
+      "learning_rate": 4.8325404129134915e-06,
+      "loss": 0.0461,
+      "step": 4638
+    },
+    {
+      "epoch": 2.4148880791254554,
+      "grad_norm": 0.22378060847821732,
+      "learning_rate": 4.824240529311075e-06,
+      "loss": 0.0466,
+      "step": 4639
+    },
+    {
+      "epoch": 2.415408641332639,
+      "grad_norm": 0.225573866033014,
+      "learning_rate": 4.815947018152397e-06,
+      "loss": 0.0464,
+      "step": 4640
+    },
+    {
+      "epoch": 2.415929203539823,
+      "grad_norm": 0.23586252079175646,
+      "learning_rate": 4.807659882056945e-06,
+      "loss": 0.0466,
+      "step": 4641
+    },
+    {
+      "epoch": 2.4164497657470068,
+      "grad_norm": 0.22986325100510877,
+      "learning_rate": 4.799379123642162e-06,
+      "loss": 0.0446,
+      "step": 4642
+    },
+    {
+      "epoch": 2.4169703279541905,
+      "grad_norm": 0.23705062124271192,
+      "learning_rate": 4.791104745523509e-06,
+      "loss": 0.0479,
+      "step": 4643
+    },
+    {
+      "epoch": 2.4174908901613743,
+      "grad_norm": 0.23732585219476587,
+      "learning_rate": 4.78283675031442e-06,
+      "loss": 0.047,
+      "step": 4644
+    },
+    {
+      "epoch": 2.418011452368558,
+      "grad_norm": 0.23331514934441308,
+      "learning_rate": 4.7745751406263165e-06,
+      "loss": 0.0459,
+      "step": 4645
+    },
+    {
+      "epoch": 2.418532014575742,
+      "grad_norm": 0.23781982825270778,
+      "learning_rate": 4.766319919068593e-06,
+      "loss": 0.0476,
+      "step": 4646
+    },
+    {
+      "epoch": 2.4190525767829256,
+      "grad_norm": 0.24131605817905036,
+      "learning_rate": 4.758071088248628e-06,
+      "loss": 0.0483,
+      "step": 4647
+    },
+    {
+      "epoch": 2.4195731389901094,
+      "grad_norm": 0.2351381427413349,
+      "learning_rate": 4.7498286507717895e-06,
+      "loss": 0.047,
+      "step": 4648
+    },
+    {
+      "epoch": 2.420093701197293,
+      "grad_norm": 0.23702310918812222,
+      "learning_rate": 4.741592609241427e-06,
+      "loss": 0.0471,
+      "step": 4649
+    },
+    {
+      "epoch": 2.420614263404477,
+      "grad_norm": 0.2214982247537357,
+      "learning_rate": 4.733362966258869e-06,
+      "loss": 0.0439,
+      "step": 4650
+    },
+    {
+      "epoch": 2.4211348256116607,
+      "grad_norm": 0.23411796590355238,
+      "learning_rate": 4.725139724423411e-06,
+      "loss": 0.0461,
+      "step": 4651
+    },
+    {
+      "epoch": 2.4216553878188445,
+      "grad_norm": 0.23041508871761626,
+      "learning_rate": 4.716922886332334e-06,
+      "loss": 0.0449,
+      "step": 4652
+    },
+    {
+      "epoch": 2.4221759500260283,
+      "grad_norm": 0.2402349549586395,
+      "learning_rate": 4.7087124545809045e-06,
+      "loss": 0.046,
+      "step": 4653
+    },
+    {
+      "epoch": 2.422696512233212,
+      "grad_norm": 0.23216773676851069,
+      "learning_rate": 4.700508431762365e-06,
+      "loss": 0.0465,
+      "step": 4654
+    },
+    {
+      "epoch": 2.423217074440396,
+      "grad_norm": 0.2348648584242627,
+      "learning_rate": 4.692310820467919e-06,
+      "loss": 0.0478,
+      "step": 4655
+    },
+    {
+      "epoch": 2.4237376366475796,
+      "grad_norm": 0.22464964421666625,
+      "learning_rate": 4.684119623286748e-06,
+      "loss": 0.0445,
+      "step": 4656
+    },
+    {
+      "epoch": 2.424258198854763,
+      "grad_norm": 0.2246231882200223,
+      "learning_rate": 4.675934842806018e-06,
+      "loss": 0.0469,
+      "step": 4657
+    },
+    {
+      "epoch": 2.4247787610619467,
+      "grad_norm": 0.2288877796143045,
+      "learning_rate": 4.667756481610866e-06,
+      "loss": 0.0465,
+      "step": 4658
+    },
+    {
+      "epoch": 2.4252993232691304,
+      "grad_norm": 0.22672281731855962,
+      "learning_rate": 4.6595845422844035e-06,
+      "loss": 0.0461,
+      "step": 4659
+    },
+    {
+      "epoch": 2.425819885476314,
+      "grad_norm": 0.23471740148435588,
+      "learning_rate": 4.6514190274076996e-06,
+      "loss": 0.0478,
+      "step": 4660
+    },
+    {
+      "epoch": 2.426340447683498,
+      "grad_norm": 0.22362959400916294,
+      "learning_rate": 4.643259939559807e-06,
+      "loss": 0.0458,
+      "step": 4661
+    },
+    {
+      "epoch": 2.4268610098906818,
+      "grad_norm": 0.2227801341848505,
+      "learning_rate": 4.6351072813177495e-06,
+      "loss": 0.046,
+      "step": 4662
+    },
+    {
+      "epoch": 2.4273815720978655,
+      "grad_norm": 0.22031118417632997,
+      "learning_rate": 4.626961055256515e-06,
+      "loss": 0.0439,
+      "step": 4663
+    },
+    {
+      "epoch": 2.4279021343050493,
+      "grad_norm": 0.22982141836032516,
+      "learning_rate": 4.618821263949061e-06,
+      "loss": 0.0462,
+      "step": 4664
+    },
+    {
+      "epoch": 2.428422696512233,
+      "grad_norm": 0.23150711698760643,
+      "learning_rate": 4.610687909966304e-06,
+      "loss": 0.0456,
+      "step": 4665
+    },
+    {
+      "epoch": 2.428943258719417,
+      "grad_norm": 0.23045867200414713,
+      "learning_rate": 4.602560995877142e-06,
+      "loss": 0.0457,
+      "step": 4666
+    },
+    {
+      "epoch": 2.4294638209266006,
+      "grad_norm": 0.23591693108732284,
+      "learning_rate": 4.594440524248431e-06,
+      "loss": 0.0442,
+      "step": 4667
+    },
+    {
+      "epoch": 2.4299843831337844,
+      "grad_norm": 0.22810723603892324,
+      "learning_rate": 4.586326497645002e-06,
+      "loss": 0.046,
+      "step": 4668
+    },
+    {
+      "epoch": 2.430504945340968,
+      "grad_norm": 0.2399613227524779,
+      "learning_rate": 4.578218918629632e-06,
+      "loss": 0.0476,
+      "step": 4669
+    },
+    {
+      "epoch": 2.431025507548152,
+      "grad_norm": 0.2464507638072641,
+      "learning_rate": 4.570117789763073e-06,
+      "loss": 0.0473,
+      "step": 4670
+    },
+    {
+      "epoch": 2.4315460697553357,
+      "grad_norm": 0.23144494070866378,
+      "learning_rate": 4.562023113604041e-06,
+      "loss": 0.0461,
+      "step": 4671
+    },
+    {
+      "epoch": 2.4320666319625195,
+      "grad_norm": 0.2325135941641269,
+      "learning_rate": 4.553934892709216e-06,
+      "loss": 0.0477,
+      "step": 4672
+    },
+    {
+      "epoch": 2.4325871941697033,
+      "grad_norm": 0.22934582369828932,
+      "learning_rate": 4.545853129633226e-06,
+      "loss": 0.0468,
+      "step": 4673
+    },
+    {
+      "epoch": 2.433107756376887,
+      "grad_norm": 0.23396484474390528,
+      "learning_rate": 4.5377778269286766e-06,
+      "loss": 0.0468,
+      "step": 4674
+    },
+    {
+      "epoch": 2.433628318584071,
+      "grad_norm": 0.23530318910470144,
+      "learning_rate": 4.529708987146114e-06,
+      "loss": 0.0462,
+      "step": 4675
+    },
+    {
+      "epoch": 2.4341488807912546,
+      "grad_norm": 0.23344486790079444,
+      "learning_rate": 4.521646612834057e-06,
+      "loss": 0.0464,
+      "step": 4676
+    },
+    {
+      "epoch": 2.4346694429984383,
+      "grad_norm": 0.23330292282907472,
+      "learning_rate": 4.513590706538989e-06,
+      "loss": 0.0463,
+      "step": 4677
+    },
+    {
+      "epoch": 2.435190005205622,
+      "grad_norm": 0.22437040772272743,
+      "learning_rate": 4.5055412708053245e-06,
+      "loss": 0.0477,
+      "step": 4678
+    },
+    {
+      "epoch": 2.435710567412806,
+      "grad_norm": 0.2265909891553176,
+      "learning_rate": 4.497498308175454e-06,
+      "loss": 0.0461,
+      "step": 4679
+    },
+    {
+      "epoch": 2.4362311296199897,
+      "grad_norm": 0.23121513283352574,
+      "learning_rate": 4.489461821189725e-06,
+      "loss": 0.0464,
+      "step": 4680
+    },
+    {
+      "epoch": 2.4367516918271734,
+      "grad_norm": 0.2528579823174406,
+      "learning_rate": 4.481431812386436e-06,
+      "loss": 0.0495,
+      "step": 4681
+    },
+    {
+      "epoch": 2.437272254034357,
+      "grad_norm": 0.23289575073759872,
+      "learning_rate": 4.473408284301825e-06,
+      "loss": 0.0459,
+      "step": 4682
+    },
+    {
+      "epoch": 2.437792816241541,
+      "grad_norm": 0.23015881999073623,
+      "learning_rate": 4.465391239470112e-06,
+      "loss": 0.0453,
+      "step": 4683
+    },
+    {
+      "epoch": 2.4383133784487248,
+      "grad_norm": 0.22624808663286283,
+      "learning_rate": 4.457380680423434e-06,
+      "loss": 0.0453,
+      "step": 4684
+    },
+    {
+      "epoch": 2.4388339406559085,
+      "grad_norm": 0.2296916685174956,
+      "learning_rate": 4.4493766096919136e-06,
+      "loss": 0.0464,
+      "step": 4685
+    },
+    {
+      "epoch": 2.4393545028630923,
+      "grad_norm": 0.21738846706883652,
+      "learning_rate": 4.441379029803605e-06,
+      "loss": 0.0425,
+      "step": 4686
+    },
+    {
+      "epoch": 2.439875065070276,
+      "grad_norm": 0.2423083039591658,
+      "learning_rate": 4.433387943284511e-06,
+      "loss": 0.0468,
+      "step": 4687
+    },
+    {
+      "epoch": 2.44039562727746,
+      "grad_norm": 0.2341956963383124,
+      "learning_rate": 4.425403352658591e-06,
+      "loss": 0.0447,
+      "step": 4688
+    },
+    {
+      "epoch": 2.4409161894846436,
+      "grad_norm": 0.24101151048749272,
+      "learning_rate": 4.417425260447753e-06,
+      "loss": 0.0475,
+      "step": 4689
+    },
+    {
+      "epoch": 2.4414367516918274,
+      "grad_norm": 0.24063994225163093,
+      "learning_rate": 4.4094536691718505e-06,
+      "loss": 0.0465,
+      "step": 4690
+    },
+    {
+      "epoch": 2.441957313899011,
+      "grad_norm": 0.23071169180768644,
+      "learning_rate": 4.401488581348679e-06,
+      "loss": 0.0471,
+      "step": 4691
+    },
+    {
+      "epoch": 2.442477876106195,
+      "grad_norm": 0.2355448253070943,
+      "learning_rate": 4.393529999493989e-06,
+      "loss": 0.0471,
+      "step": 4692
+    },
+    {
+      "epoch": 2.4429984383133783,
+      "grad_norm": 0.23632196722116713,
+      "learning_rate": 4.385577926121464e-06,
+      "loss": 0.0474,
+      "step": 4693
+    },
+    {
+      "epoch": 2.443519000520562,
+      "grad_norm": 0.23534929674479932,
+      "learning_rate": 4.3776323637427395e-06,
+      "loss": 0.0452,
+      "step": 4694
+    },
+    {
+      "epoch": 2.444039562727746,
+      "grad_norm": 0.24304563929485506,
+      "learning_rate": 4.369693314867407e-06,
+      "loss": 0.047,
+      "step": 4695
+    },
+    {
+      "epoch": 2.4445601249349296,
+      "grad_norm": 0.23188182601997595,
+      "learning_rate": 4.3617607820029686e-06,
+      "loss": 0.045,
+      "step": 4696
+    },
+    {
+      "epoch": 2.4450806871421134,
+      "grad_norm": 0.23588489710907637,
+      "learning_rate": 4.353834767654896e-06,
+      "loss": 0.046,
+      "step": 4697
+    },
+    {
+      "epoch": 2.445601249349297,
+      "grad_norm": 0.2423879338466431,
+      "learning_rate": 4.345915274326595e-06,
+      "loss": 0.0486,
+      "step": 4698
+    },
+    {
+      "epoch": 2.446121811556481,
+      "grad_norm": 0.23190042633660438,
+      "learning_rate": 4.338002304519406e-06,
+      "loss": 0.0476,
+      "step": 4699
+    },
+    {
+      "epoch": 2.4466423737636647,
+      "grad_norm": 0.2371416200568294,
+      "learning_rate": 4.330095860732625e-06,
+      "loss": 0.0475,
+      "step": 4700
+    },
+    {
+      "epoch": 2.4471629359708484,
+      "grad_norm": 0.233119464425362,
+      "learning_rate": 4.322195945463464e-06,
+      "loss": 0.0453,
+      "step": 4701
+    },
+    {
+      "epoch": 2.447683498178032,
+      "grad_norm": 0.23660527240553358,
+      "learning_rate": 4.314302561207079e-06,
+      "loss": 0.047,
+      "step": 4702
+    },
+    {
+      "epoch": 2.448204060385216,
+      "grad_norm": 0.2225359791140698,
+      "learning_rate": 4.306415710456577e-06,
+      "loss": 0.0451,
+      "step": 4703
+    },
+    {
+      "epoch": 2.4487246225923998,
+      "grad_norm": 0.22352359233272487,
+      "learning_rate": 4.2985353957029876e-06,
+      "loss": 0.0438,
+      "step": 4704
+    },
+    {
+      "epoch": 2.4492451847995835,
+      "grad_norm": 0.22659412226171605,
+      "learning_rate": 4.29066161943529e-06,
+      "loss": 0.0453,
+      "step": 4705
+    },
+    {
+      "epoch": 2.4497657470067673,
+      "grad_norm": 0.23202678432385068,
+      "learning_rate": 4.282794384140379e-06,
+      "loss": 0.0456,
+      "step": 4706
+    },
+    {
+      "epoch": 2.450286309213951,
+      "grad_norm": 0.23436332924353653,
+      "learning_rate": 4.274933692303093e-06,
+      "loss": 0.0467,
+      "step": 4707
+    },
+    {
+      "epoch": 2.450806871421135,
+      "grad_norm": 0.23568145553682646,
+      "learning_rate": 4.267079546406211e-06,
+      "loss": 0.0457,
+      "step": 4708
+    },
+    {
+      "epoch": 2.4513274336283186,
+      "grad_norm": 0.23688538546245036,
+      "learning_rate": 4.259231948930442e-06,
+      "loss": 0.0465,
+      "step": 4709
+    },
+    {
+      "epoch": 2.4518479958355024,
+      "grad_norm": 0.22584612440650176,
+      "learning_rate": 4.251390902354413e-06,
+      "loss": 0.045,
+      "step": 4710
+    },
+    {
+      "epoch": 2.452368558042686,
+      "grad_norm": 0.22787648763996438,
+      "learning_rate": 4.243556409154692e-06,
+      "loss": 0.0447,
+      "step": 4711
+    },
+    {
+      "epoch": 2.45288912024987,
+      "grad_norm": 0.2308201824461422,
+      "learning_rate": 4.235728471805775e-06,
+      "loss": 0.0452,
+      "step": 4712
+    },
+    {
+      "epoch": 2.4534096824570537,
+      "grad_norm": 0.23117909367110293,
+      "learning_rate": 4.227907092780095e-06,
+      "loss": 0.0455,
+      "step": 4713
+    },
+    {
+      "epoch": 2.4539302446642375,
+      "grad_norm": 0.22631631172759079,
+      "learning_rate": 4.22009227454801e-06,
+      "loss": 0.0463,
+      "step": 4714
+    },
+    {
+      "epoch": 2.4544508068714213,
+      "grad_norm": 0.2344996227202645,
+      "learning_rate": 4.212284019577792e-06,
+      "loss": 0.0469,
+      "step": 4715
+    },
+    {
+      "epoch": 2.454971369078605,
+      "grad_norm": 0.22826293638212433,
+      "learning_rate": 4.204482330335657e-06,
+      "loss": 0.046,
+      "step": 4716
+    },
+    {
+      "epoch": 2.455491931285789,
+      "grad_norm": 0.24057079364699233,
+      "learning_rate": 4.196687209285744e-06,
+      "loss": 0.0475,
+      "step": 4717
+    },
+    {
+      "epoch": 2.4560124934929726,
+      "grad_norm": 0.23451233031179078,
+      "learning_rate": 4.188898658890117e-06,
+      "loss": 0.0462,
+      "step": 4718
+    },
+    {
+      "epoch": 2.4565330557001563,
+      "grad_norm": 0.23186747086362713,
+      "learning_rate": 4.1811166816087595e-06,
+      "loss": 0.0464,
+      "step": 4719
+    },
+    {
+      "epoch": 2.45705361790734,
+      "grad_norm": 0.22706734857716063,
+      "learning_rate": 4.173341279899576e-06,
+      "loss": 0.0442,
+      "step": 4720
+    },
+    {
+      "epoch": 2.4575741801145234,
+      "grad_norm": 0.23846594551848097,
+      "learning_rate": 4.165572456218405e-06,
+      "loss": 0.0492,
+      "step": 4721
+    },
+    {
+      "epoch": 2.458094742321707,
+      "grad_norm": 0.22761321129821427,
+      "learning_rate": 4.157810213019003e-06,
+      "loss": 0.0467,
+      "step": 4722
+    },
+    {
+      "epoch": 2.458615304528891,
+      "grad_norm": 0.21913183574661646,
+      "learning_rate": 4.150054552753055e-06,
+      "loss": 0.0449,
+      "step": 4723
+    },
+    {
+      "epoch": 2.4591358667360748,
+      "grad_norm": 0.2258958824847465,
+      "learning_rate": 4.1423054778701455e-06,
+      "loss": 0.0463,
+      "step": 4724
+    },
+    {
+      "epoch": 2.4596564289432585,
+      "grad_norm": 0.2535114295876158,
+      "learning_rate": 4.1345629908178e-06,
+      "loss": 0.0446,
+      "step": 4725
+    },
+    {
+      "epoch": 2.4601769911504423,
+      "grad_norm": 0.2188190545994277,
+      "learning_rate": 4.126827094041455e-06,
+      "loss": 0.045,
+      "step": 4726
+    },
+    {
+      "epoch": 2.460697553357626,
+      "grad_norm": 0.22451949242641625,
+      "learning_rate": 4.119097789984472e-06,
+      "loss": 0.0474,
+      "step": 4727
+    },
+    {
+      "epoch": 2.46121811556481,
+      "grad_norm": 0.23179831019249533,
+      "learning_rate": 4.111375081088123e-06,
+      "loss": 0.0456,
+      "step": 4728
+    },
+    {
+      "epoch": 2.4617386777719936,
+      "grad_norm": 0.23401390682247894,
+      "learning_rate": 4.103658969791588e-06,
+      "loss": 0.047,
+      "step": 4729
+    },
+    {
+      "epoch": 2.4622592399791774,
+      "grad_norm": 0.29525830006055254,
+      "learning_rate": 4.095949458531984e-06,
+      "loss": 0.045,
+      "step": 4730
+    },
+    {
+      "epoch": 2.462779802186361,
+      "grad_norm": 0.2259193500356729,
+      "learning_rate": 4.088246549744331e-06,
+      "loss": 0.0438,
+      "step": 4731
+    },
+    {
+      "epoch": 2.463300364393545,
+      "grad_norm": 0.22553481026955677,
+      "learning_rate": 4.0805502458615725e-06,
+      "loss": 0.0441,
+      "step": 4732
+    },
+    {
+      "epoch": 2.4638209266007287,
+      "grad_norm": 0.2533883156131254,
+      "learning_rate": 4.07286054931455e-06,
+      "loss": 0.0473,
+      "step": 4733
+    },
+    {
+      "epoch": 2.4643414888079125,
+      "grad_norm": 0.23645871037063934,
+      "learning_rate": 4.065177462532027e-06,
+      "loss": 0.0453,
+      "step": 4734
+    },
+    {
+      "epoch": 2.4648620510150963,
+      "grad_norm": 0.2390469842852221,
+      "learning_rate": 4.057500987940688e-06,
+      "loss": 0.0467,
+      "step": 4735
+    },
+    {
+      "epoch": 2.46538261322228,
+      "grad_norm": 0.24040218439850058,
+      "learning_rate": 4.04983112796512e-06,
+      "loss": 0.0464,
+      "step": 4736
+    },
+    {
+      "epoch": 2.465903175429464,
+      "grad_norm": 0.2323783034515998,
+      "learning_rate": 4.04216788502782e-06,
+      "loss": 0.0466,
+      "step": 4737
+    },
+    {
+      "epoch": 2.4664237376366476,
+      "grad_norm": 0.2421242614235067,
+      "learning_rate": 4.03451126154919e-06,
+      "loss": 0.0457,
+      "step": 4738
+    },
+    {
+      "epoch": 2.4669442998438313,
+      "grad_norm": 0.2241947463119316,
+      "learning_rate": 4.0268612599475534e-06,
+      "loss": 0.0442,
+      "step": 4739
+    },
+    {
+      "epoch": 2.467464862051015,
+      "grad_norm": 0.23390631221316124,
+      "learning_rate": 4.019217882639137e-06,
+      "loss": 0.0459,
+      "step": 4740
+    },
+    {
+      "epoch": 2.467985424258199,
+      "grad_norm": 0.23604571830215357,
+      "learning_rate": 4.011581132038078e-06,
+      "loss": 0.046,
+      "step": 4741
+    },
+    {
+      "epoch": 2.4685059864653827,
+      "grad_norm": 0.23679414599656148,
+      "learning_rate": 4.003951010556412e-06,
+      "loss": 0.0477,
+      "step": 4742
+    },
+    {
+      "epoch": 2.4690265486725664,
+      "grad_norm": 0.22466780147244902,
+      "learning_rate": 3.996327520604087e-06,
+      "loss": 0.0453,
+      "step": 4743
+    },
+    {
+      "epoch": 2.46954711087975,
+      "grad_norm": 0.23183844848666738,
+      "learning_rate": 3.9887106645889574e-06,
+      "loss": 0.0458,
+      "step": 4744
+    },
+    {
+      "epoch": 2.470067673086934,
+      "grad_norm": 0.2258352174710472,
+      "learning_rate": 3.981100444916788e-06,
+      "loss": 0.0446,
+      "step": 4745
+    },
+    {
+      "epoch": 2.4705882352941178,
+      "grad_norm": 0.23149767246195083,
+      "learning_rate": 3.973496863991233e-06,
+      "loss": 0.0447,
+      "step": 4746
+    },
+    {
+      "epoch": 2.4711087975013015,
+      "grad_norm": 0.23290847900975084,
+      "learning_rate": 3.965899924213851e-06,
+      "loss": 0.0446,
+      "step": 4747
+    },
+    {
+      "epoch": 2.4716293597084853,
+      "grad_norm": 0.2249778446324839,
+      "learning_rate": 3.958309627984116e-06,
+      "loss": 0.045,
+      "step": 4748
+    },
+    {
+      "epoch": 2.472149921915669,
+      "grad_norm": 0.21861880310195234,
+      "learning_rate": 3.950725977699396e-06,
+      "loss": 0.0443,
+      "step": 4749
+    },
+    {
+      "epoch": 2.472670484122853,
+      "grad_norm": 0.23380723096798692,
+      "learning_rate": 3.943148975754968e-06,
+      "loss": 0.0456,
+      "step": 4750
+    },
+    {
+      "epoch": 2.4731910463300366,
+      "grad_norm": 0.25812439912963125,
+      "learning_rate": 3.9355786245439896e-06,
+      "loss": 0.0492,
+      "step": 4751
+    },
+    {
+      "epoch": 2.4737116085372204,
+      "grad_norm": 0.2346925416518511,
+      "learning_rate": 3.928014926457532e-06,
+      "loss": 0.0469,
+      "step": 4752
+    },
+    {
+      "epoch": 2.474232170744404,
+      "grad_norm": 0.2324299060972178,
+      "learning_rate": 3.920457883884571e-06,
+      "loss": 0.0469,
+      "step": 4753
+    },
+    {
+      "epoch": 2.474752732951588,
+      "grad_norm": 0.22340130582451653,
+      "learning_rate": 3.9129074992119705e-06,
+      "loss": 0.0452,
+      "step": 4754
+    },
+    {
+      "epoch": 2.4752732951587717,
+      "grad_norm": 0.23185760003678735,
+      "learning_rate": 3.905363774824492e-06,
+      "loss": 0.0461,
+      "step": 4755
+    },
+    {
+      "epoch": 2.4757938573659555,
+      "grad_norm": 0.22369373177962135,
+      "learning_rate": 3.897826713104786e-06,
+      "loss": 0.0429,
+      "step": 4756
+    },
+    {
+      "epoch": 2.476314419573139,
+      "grad_norm": 0.22152244736340082,
+      "learning_rate": 3.8902963164334145e-06,
+      "loss": 0.0454,
+      "step": 4757
+    },
+    {
+      "epoch": 2.4768349817803226,
+      "grad_norm": 0.2288794662126069,
+      "learning_rate": 3.882772587188827e-06,
+      "loss": 0.044,
+      "step": 4758
+    },
+    {
+      "epoch": 2.4773555439875063,
+      "grad_norm": 0.22786850218271076,
+      "learning_rate": 3.875255527747376e-06,
+      "loss": 0.0455,
+      "step": 4759
+    },
+    {
+      "epoch": 2.47787610619469,
+      "grad_norm": 0.24127063199154952,
+      "learning_rate": 3.867745140483281e-06,
+      "loss": 0.0494,
+      "step": 4760
+    },
+    {
+      "epoch": 2.478396668401874,
+      "grad_norm": 0.22651816469440886,
+      "learning_rate": 3.860241427768682e-06,
+      "loss": 0.0474,
+      "step": 4761
+    },
+    {
+      "epoch": 2.4789172306090577,
+      "grad_norm": 0.23292917754764078,
+      "learning_rate": 3.852744391973601e-06,
+      "loss": 0.0453,
+      "step": 4762
+    },
+    {
+      "epoch": 2.4794377928162414,
+      "grad_norm": 0.23481848280563425,
+      "learning_rate": 3.845254035465951e-06,
+      "loss": 0.0462,
+      "step": 4763
+    },
+    {
+      "epoch": 2.479958355023425,
+      "grad_norm": 0.23409115715668288,
+      "learning_rate": 3.8377703606115425e-06,
+      "loss": 0.0456,
+      "step": 4764
+    },
+    {
+      "epoch": 2.480478917230609,
+      "grad_norm": 0.22688970649832368,
+      "learning_rate": 3.830293369774049e-06,
+      "loss": 0.046,
+      "step": 4765
+    },
+    {
+      "epoch": 2.4809994794377928,
+      "grad_norm": 0.24911131431251246,
+      "learning_rate": 3.822823065315062e-06,
+      "loss": 0.0477,
+      "step": 4766
+    },
+    {
+      "epoch": 2.4815200416449765,
+      "grad_norm": 0.23613842700910898,
+      "learning_rate": 3.815359449594053e-06,
+      "loss": 0.0464,
+      "step": 4767
+    },
+    {
+      "epoch": 2.4820406038521603,
+      "grad_norm": 0.22630425395246193,
+      "learning_rate": 3.8079025249683766e-06,
+      "loss": 0.0454,
+      "step": 4768
+    },
+    {
+      "epoch": 2.482561166059344,
+      "grad_norm": 0.23834407534931806,
+      "learning_rate": 3.800452293793283e-06,
+      "loss": 0.0474,
+      "step": 4769
+    },
+    {
+      "epoch": 2.483081728266528,
+      "grad_norm": 0.22299016955003192,
+      "learning_rate": 3.7930087584218924e-06,
+      "loss": 0.0447,
+      "step": 4770
+    },
+    {
+      "epoch": 2.4836022904737116,
+      "grad_norm": 0.22180738154042506,
+      "learning_rate": 3.785571921205225e-06,
+      "loss": 0.0444,
+      "step": 4771
+    },
+    {
+      "epoch": 2.4841228526808954,
+      "grad_norm": 0.23606413057780049,
+      "learning_rate": 3.7781417844921785e-06,
+      "loss": 0.0462,
+      "step": 4772
+    },
+    {
+      "epoch": 2.484643414888079,
+      "grad_norm": 0.23226564075490944,
+      "learning_rate": 3.770718350629543e-06,
+      "loss": 0.0471,
+      "step": 4773
+    },
+    {
+      "epoch": 2.485163977095263,
+      "grad_norm": 0.222173152299585,
+      "learning_rate": 3.7633016219619786e-06,
+      "loss": 0.044,
+      "step": 4774
+    },
+    {
+      "epoch": 2.4856845393024467,
+      "grad_norm": 0.22670387438422338,
+      "learning_rate": 3.755891600832026e-06,
+      "loss": 0.0451,
+      "step": 4775
+    },
+    {
+      "epoch": 2.4862051015096305,
+      "grad_norm": 0.24457041056832837,
+      "learning_rate": 3.748488289580124e-06,
+      "loss": 0.0471,
+      "step": 4776
+    },
+    {
+      "epoch": 2.4867256637168142,
+      "grad_norm": 0.2292671621830906,
+      "learning_rate": 3.7410916905445763e-06,
+      "loss": 0.0461,
+      "step": 4777
+    },
+    {
+      "epoch": 2.487246225923998,
+      "grad_norm": 0.2319982917968305,
+      "learning_rate": 3.7337018060615847e-06,
+      "loss": 0.0455,
+      "step": 4778
+    },
+    {
+      "epoch": 2.487766788131182,
+      "grad_norm": 0.22435841846525945,
+      "learning_rate": 3.7263186384652064e-06,
+      "loss": 0.043,
+      "step": 4779
+    },
+    {
+      "epoch": 2.4882873503383656,
+      "grad_norm": 0.22949123049534018,
+      "learning_rate": 3.7189421900873905e-06,
+      "loss": 0.045,
+      "step": 4780
+    },
+    {
+      "epoch": 2.4888079125455493,
+      "grad_norm": 0.23210355715087153,
+      "learning_rate": 3.71157246325797e-06,
+      "loss": 0.0441,
+      "step": 4781
+    },
+    {
+      "epoch": 2.489328474752733,
+      "grad_norm": 0.2260444385170093,
+      "learning_rate": 3.7042094603046473e-06,
+      "loss": 0.0436,
+      "step": 4782
+    },
+    {
+      "epoch": 2.489849036959917,
+      "grad_norm": 0.2293597694356391,
+      "learning_rate": 3.696853183552998e-06,
+      "loss": 0.0453,
+      "step": 4783
+    },
+    {
+      "epoch": 2.4903695991671007,
+      "grad_norm": 0.22728917524183118,
+      "learning_rate": 3.6895036353264716e-06,
+      "loss": 0.0473,
+      "step": 4784
+    },
+    {
+      "epoch": 2.490890161374284,
+      "grad_norm": 0.2258597144539425,
+      "learning_rate": 3.6821608179464006e-06,
+      "loss": 0.0447,
+      "step": 4785
+    },
+    {
+      "epoch": 2.4914107235814678,
+      "grad_norm": 0.23529432877833512,
+      "learning_rate": 3.674824733731991e-06,
+      "loss": 0.0454,
+      "step": 4786
+    },
+    {
+      "epoch": 2.4919312857886515,
+      "grad_norm": 0.2276587426156354,
+      "learning_rate": 3.6674953850003245e-06,
+      "loss": 0.0457,
+      "step": 4787
+    },
+    {
+      "epoch": 2.4924518479958353,
+      "grad_norm": 0.23595370854468917,
+      "learning_rate": 3.6601727740663395e-06,
+      "loss": 0.0466,
+      "step": 4788
+    },
+    {
+      "epoch": 2.492972410203019,
+      "grad_norm": 0.23081012893994207,
+      "learning_rate": 3.652856903242863e-06,
+      "loss": 0.0437,
+      "step": 4789
+    },
+    {
+      "epoch": 2.493492972410203,
+      "grad_norm": 0.23017085621336414,
+      "learning_rate": 3.6455477748405853e-06,
+      "loss": 0.0445,
+      "step": 4790
+    },
+    {
+      "epoch": 2.4940135346173866,
+      "grad_norm": 0.23217068455139228,
+      "learning_rate": 3.638245391168077e-06,
+      "loss": 0.0476,
+      "step": 4791
+    },
+    {
+      "epoch": 2.4945340968245704,
+      "grad_norm": 0.2224420779312252,
+      "learning_rate": 3.630949754531765e-06,
+      "loss": 0.0455,
+      "step": 4792
+    },
+    {
+      "epoch": 2.495054659031754,
+      "grad_norm": 0.2332204491629984,
+      "learning_rate": 3.623660867235945e-06,
+      "loss": 0.0455,
+      "step": 4793
+    },
+    {
+      "epoch": 2.495575221238938,
+      "grad_norm": 0.2302258850640075,
+      "learning_rate": 3.6163787315827894e-06,
+      "loss": 0.0461,
+      "step": 4794
+    },
+    {
+      "epoch": 2.4960957834461217,
+      "grad_norm": 0.23170180876721022,
+      "learning_rate": 3.609103349872342e-06,
+      "loss": 0.0462,
+      "step": 4795
+    },
+    {
+      "epoch": 2.4966163456533055,
+      "grad_norm": 0.22548808881382684,
+      "learning_rate": 3.6018347244025085e-06,
+      "loss": 0.0456,
+      "step": 4796
+    },
+    {
+      "epoch": 2.4971369078604893,
+      "grad_norm": 0.22318432696419793,
+      "learning_rate": 3.5945728574690474e-06,
+      "loss": 0.0451,
+      "step": 4797
+    },
+    {
+      "epoch": 2.497657470067673,
+      "grad_norm": 0.2328804121095348,
+      "learning_rate": 3.5873177513655985e-06,
+      "loss": 0.0465,
+      "step": 4798
+    },
+    {
+      "epoch": 2.498178032274857,
+      "grad_norm": 0.2362488676580557,
+      "learning_rate": 3.5800694083836685e-06,
+      "loss": 0.0458,
+      "step": 4799
+    },
+    {
+      "epoch": 2.4986985944820406,
+      "grad_norm": 0.22505683785956507,
+      "learning_rate": 3.5728278308126195e-06,
+      "loss": 0.0445,
+      "step": 4800
+    },
+    {
+      "epoch": 2.4992191566892243,
+      "grad_norm": 0.23832222888676316,
+      "learning_rate": 3.5655930209396783e-06,
+      "loss": 0.0451,
+      "step": 4801
+    },
+    {
+      "epoch": 2.499739718896408,
+      "grad_norm": 0.23018716395069688,
+      "learning_rate": 3.5583649810499246e-06,
+      "loss": 0.0442,
+      "step": 4802
+    },
+    {
+      "epoch": 2.500260281103592,
+      "grad_norm": 0.22556770408037974,
+      "learning_rate": 3.551143713426319e-06,
+      "loss": 0.0438,
+      "step": 4803
+    },
+    {
+      "epoch": 2.5007808433107757,
+      "grad_norm": 0.22278746289847356,
+      "learning_rate": 3.543929220349673e-06,
+      "loss": 0.0452,
+      "step": 4804
+    },
+    {
+      "epoch": 2.5013014055179594,
+      "grad_norm": 0.22592179865196593,
+      "learning_rate": 3.536721504098664e-06,
+      "loss": 0.0454,
+      "step": 4805
+    },
+    {
+      "epoch": 2.501821967725143,
+      "grad_norm": 0.2191833418264602,
+      "learning_rate": 3.529520566949812e-06,
+      "loss": 0.0429,
+      "step": 4806
+    },
+    {
+      "epoch": 2.502342529932327,
+      "grad_norm": 0.23567814923196181,
+      "learning_rate": 3.522326411177515e-06,
+      "loss": 0.0455,
+      "step": 4807
+    },
+    {
+      "epoch": 2.5028630921395107,
+      "grad_norm": 0.2329494968156347,
+      "learning_rate": 3.5151390390540245e-06,
+      "loss": 0.0456,
+      "step": 4808
+    },
+    {
+      "epoch": 2.5033836543466945,
+      "grad_norm": 0.23201928458314705,
+      "learning_rate": 3.5079584528494497e-06,
+      "loss": 0.0453,
+      "step": 4809
+    },
+    {
+      "epoch": 2.5039042165538783,
+      "grad_norm": 0.22257219879809637,
+      "learning_rate": 3.5007846548317487e-06,
+      "loss": 0.0432,
+      "step": 4810
+    },
+    {
+      "epoch": 2.504424778761062,
+      "grad_norm": 0.2357881279026739,
+      "learning_rate": 3.4936176472667337e-06,
+      "loss": 0.0455,
+      "step": 4811
+    },
+    {
+      "epoch": 2.504945340968246,
+      "grad_norm": 0.22156754335142384,
+      "learning_rate": 3.486457432418089e-06,
+      "loss": 0.0439,
+      "step": 4812
+    },
+    {
+      "epoch": 2.5054659031754296,
+      "grad_norm": 0.222223788188508,
+      "learning_rate": 3.479304012547338e-06,
+      "loss": 0.0437,
+      "step": 4813
+    },
+    {
+      "epoch": 2.5059864653826134,
+      "grad_norm": 0.23233184259143652,
+      "learning_rate": 3.4721573899138743e-06,
+      "loss": 0.045,
+      "step": 4814
+    },
+    {
+      "epoch": 2.506507027589797,
+      "grad_norm": 0.23772916739161976,
+      "learning_rate": 3.4650175667749223e-06,
+      "loss": 0.0466,
+      "step": 4815
+    },
+    {
+      "epoch": 2.507027589796981,
+      "grad_norm": 0.2380129235798541,
+      "learning_rate": 3.457884545385573e-06,
+      "loss": 0.0463,
+      "step": 4816
+    },
+    {
+      "epoch": 2.5075481520041647,
+      "grad_norm": 0.2428567210468212,
+      "learning_rate": 3.450758327998768e-06,
+      "loss": 0.0455,
+      "step": 4817
+    },
+    {
+      "epoch": 2.5080687142113485,
+      "grad_norm": 0.24116904429043992,
+      "learning_rate": 3.4436389168653023e-06,
+      "loss": 0.0441,
+      "step": 4818
+    },
+    {
+      "epoch": 2.5085892764185322,
+      "grad_norm": 0.2311244048249909,
+      "learning_rate": 3.436526314233815e-06,
+      "loss": 0.0459,
+      "step": 4819
+    },
+    {
+      "epoch": 2.509109838625716,
+      "grad_norm": 0.22619706111589447,
+      "learning_rate": 3.42942052235079e-06,
+      "loss": 0.0452,
+      "step": 4820
+    },
+    {
+      "epoch": 2.5096304008329,
+      "grad_norm": 0.22462384968951704,
+      "learning_rate": 3.4223215434605714e-06,
+      "loss": 0.0437,
+      "step": 4821
+    },
+    {
+      "epoch": 2.5101509630400836,
+      "grad_norm": 0.23196960288458288,
+      "learning_rate": 3.4152293798053486e-06,
+      "loss": 0.046,
+      "step": 4822
+    },
+    {
+      "epoch": 2.510671525247267,
+      "grad_norm": 0.22816810301577573,
+      "learning_rate": 3.408144033625163e-06,
+      "loss": 0.0449,
+      "step": 4823
+    },
+    {
+      "epoch": 2.5111920874544507,
+      "grad_norm": 0.22363210681026655,
+      "learning_rate": 3.401065507157883e-06,
+      "loss": 0.0426,
+      "step": 4824
+    },
+    {
+      "epoch": 2.5117126496616344,
+      "grad_norm": 0.22254268395086219,
+      "learning_rate": 3.393993802639245e-06,
+      "loss": 0.0445,
+      "step": 4825
+    },
+    {
+      "epoch": 2.512233211868818,
+      "grad_norm": 0.222987753608597,
+      "learning_rate": 3.3869289223028204e-06,
+      "loss": 0.0444,
+      "step": 4826
+    },
+    {
+      "epoch": 2.512753774076002,
+      "grad_norm": 0.24197883530585623,
+      "learning_rate": 3.3798708683800305e-06,
+      "loss": 0.0446,
+      "step": 4827
+    },
+    {
+      "epoch": 2.5132743362831858,
+      "grad_norm": 0.236727669006638,
+      "learning_rate": 3.372819643100139e-06,
+      "loss": 0.0467,
+      "step": 4828
+    },
+    {
+      "epoch": 2.5137948984903695,
+      "grad_norm": 0.24252487797178882,
+      "learning_rate": 3.3657752486902396e-06,
+      "loss": 0.046,
+      "step": 4829
+    },
+    {
+      "epoch": 2.5143154606975533,
+      "grad_norm": 0.23031235971733746,
+      "learning_rate": 3.3587376873752853e-06,
+      "loss": 0.0456,
+      "step": 4830
+    },
+    {
+      "epoch": 2.514836022904737,
+      "grad_norm": 0.22614796596777614,
+      "learning_rate": 3.351706961378068e-06,
+      "loss": 0.0439,
+      "step": 4831
+    },
+    {
+      "epoch": 2.515356585111921,
+      "grad_norm": 0.22456575523764083,
+      "learning_rate": 3.344683072919216e-06,
+      "loss": 0.0451,
+      "step": 4832
+    },
+    {
+      "epoch": 2.5158771473191046,
+      "grad_norm": 0.22971764854819687,
+      "learning_rate": 3.337666024217209e-06,
+      "loss": 0.0457,
+      "step": 4833
+    },
+    {
+      "epoch": 2.5163977095262884,
+      "grad_norm": 0.23225244680033916,
+      "learning_rate": 3.3306558174883427e-06,
+      "loss": 0.0469,
+      "step": 4834
+    },
+    {
+      "epoch": 2.516918271733472,
+      "grad_norm": 0.22408742635883183,
+      "learning_rate": 3.323652454946774e-06,
+      "loss": 0.0453,
+      "step": 4835
+    },
+    {
+      "epoch": 2.517438833940656,
+      "grad_norm": 0.22183055756851636,
+      "learning_rate": 3.3166559388044945e-06,
+      "loss": 0.0434,
+      "step": 4836
+    },
+    {
+      "epoch": 2.5179593961478397,
+      "grad_norm": 0.2260016674787736,
+      "learning_rate": 3.3096662712713224e-06,
+      "loss": 0.0448,
+      "step": 4837
+    },
+    {
+      "epoch": 2.5184799583550235,
+      "grad_norm": 0.2301369168176432,
+      "learning_rate": 3.3026834545549252e-06,
+      "loss": 0.0448,
+      "step": 4838
+    },
+    {
+      "epoch": 2.5190005205622072,
+      "grad_norm": 0.21883186300554672,
+      "learning_rate": 3.295707490860797e-06,
+      "loss": 0.0436,
+      "step": 4839
+    },
+    {
+      "epoch": 2.519521082769391,
+      "grad_norm": 0.22202859390787064,
+      "learning_rate": 3.288738382392273e-06,
+      "loss": 0.0449,
+      "step": 4840
+    },
+    {
+      "epoch": 2.520041644976575,
+      "grad_norm": 0.22761610071343236,
+      "learning_rate": 3.2817761313505226e-06,
+      "loss": 0.0468,
+      "step": 4841
+    },
+    {
+      "epoch": 2.5205622071837586,
+      "grad_norm": 0.22523059843548607,
+      "learning_rate": 3.2748207399345534e-06,
+      "loss": 0.0431,
+      "step": 4842
+    },
+    {
+      "epoch": 2.5210827693909423,
+      "grad_norm": 0.22572775303764772,
+      "learning_rate": 3.267872210341194e-06,
+      "loss": 0.0445,
+      "step": 4843
+    },
+    {
+      "epoch": 2.521603331598126,
+      "grad_norm": 0.23141678410475952,
+      "learning_rate": 3.2609305447651145e-06,
+      "loss": 0.0473,
+      "step": 4844
+    },
+    {
+      "epoch": 2.52212389380531,
+      "grad_norm": 0.4886518611240305,
+      "learning_rate": 3.2539957453988244e-06,
+      "loss": 0.0473,
+      "step": 4845
+    },
+    {
+      "epoch": 2.5226444560124937,
+      "grad_norm": 0.23911782956874156,
+      "learning_rate": 3.2470678144326442e-06,
+      "loss": 0.0469,
+      "step": 4846
+    },
+    {
+      "epoch": 2.523165018219677,
+      "grad_norm": 0.22949597789881793,
+      "learning_rate": 3.2401467540547485e-06,
+      "loss": 0.044,
+      "step": 4847
+    },
+    {
+      "epoch": 2.5236855804268608,
+      "grad_norm": 0.22907687990668468,
+      "learning_rate": 3.233232566451119e-06,
+      "loss": 0.0448,
+      "step": 4848
+    },
+    {
+      "epoch": 2.5242061426340445,
+      "grad_norm": 0.23226555702427493,
+      "learning_rate": 3.2263252538055816e-06,
+      "loss": 0.0457,
+      "step": 4849
+    },
+    {
+      "epoch": 2.5247267048412283,
+      "grad_norm": 0.22719820874022031,
+      "learning_rate": 3.2194248182997904e-06,
+      "loss": 0.0446,
+      "step": 4850
+    },
+    {
+      "epoch": 2.525247267048412,
+      "grad_norm": 0.2207201168091615,
+      "learning_rate": 3.2125312621132274e-06,
+      "loss": 0.0455,
+      "step": 4851
+    },
+    {
+      "epoch": 2.525767829255596,
+      "grad_norm": 0.21985764504983096,
+      "learning_rate": 3.2056445874231873e-06,
+      "loss": 0.0449,
+      "step": 4852
+    },
+    {
+      "epoch": 2.5262883914627796,
+      "grad_norm": 0.22836848944195356,
+      "learning_rate": 3.198764796404807e-06,
+      "loss": 0.0449,
+      "step": 4853
+    },
+    {
+      "epoch": 2.5268089536699634,
+      "grad_norm": 0.24033077551127982,
+      "learning_rate": 3.191891891231055e-06,
+      "loss": 0.0457,
+      "step": 4854
+    },
+    {
+      "epoch": 2.527329515877147,
+      "grad_norm": 0.2412658133769983,
+      "learning_rate": 3.1850258740726975e-06,
+      "loss": 0.0458,
+      "step": 4855
+    },
+    {
+      "epoch": 2.527850078084331,
+      "grad_norm": 0.22859774083041914,
+      "learning_rate": 3.178166747098357e-06,
+      "loss": 0.0436,
+      "step": 4856
+    },
+    {
+      "epoch": 2.5283706402915147,
+      "grad_norm": 0.23096275204448605,
+      "learning_rate": 3.171314512474452e-06,
+      "loss": 0.0458,
+      "step": 4857
+    },
+    {
+      "epoch": 2.5288912024986985,
+      "grad_norm": 0.22706448675029148,
+      "learning_rate": 3.1644691723652448e-06,
+      "loss": 0.0437,
+      "step": 4858
+    },
+    {
+      "epoch": 2.5294117647058822,
+      "grad_norm": 0.2378141750228416,
+      "learning_rate": 3.1576307289328117e-06,
+      "loss": 0.0483,
+      "step": 4859
+    },
+    {
+      "epoch": 2.529932326913066,
+      "grad_norm": 0.2289339944308606,
+      "learning_rate": 3.1507991843370526e-06,
+      "loss": 0.0437,
+      "step": 4860
+    },
+    {
+      "epoch": 2.53045288912025,
+      "grad_norm": 0.23778111816368155,
+      "learning_rate": 3.1439745407356835e-06,
+      "loss": 0.0471,
+      "step": 4861
+    },
+    {
+      "epoch": 2.5309734513274336,
+      "grad_norm": 0.2176967688986415,
+      "learning_rate": 3.1371568002842437e-06,
+      "loss": 0.0434,
+      "step": 4862
+    },
+    {
+      "epoch": 2.5314940135346173,
+      "grad_norm": 0.22833640651862924,
+      "learning_rate": 3.1303459651361026e-06,
+      "loss": 0.0459,
+      "step": 4863
+    },
+    {
+      "epoch": 2.532014575741801,
+      "grad_norm": 0.23296274792690516,
+      "learning_rate": 3.123542037442426e-06,
+      "loss": 0.0448,
+      "step": 4864
+    },
+    {
+      "epoch": 2.532535137948985,
+      "grad_norm": 0.23036570209219906,
+      "learning_rate": 3.1167450193522214e-06,
+      "loss": 0.0456,
+      "step": 4865
+    },
+    {
+      "epoch": 2.5330557001561687,
+      "grad_norm": 0.22673874717998874,
+      "learning_rate": 3.1099549130122944e-06,
+      "loss": 0.0444,
+      "step": 4866
+    },
+    {
+      "epoch": 2.5335762623633524,
+      "grad_norm": 0.23336621523932544,
+      "learning_rate": 3.10317172056728e-06,
+      "loss": 0.0459,
+      "step": 4867
+    },
+    {
+      "epoch": 2.534096824570536,
+      "grad_norm": 0.22381256494910945,
+      "learning_rate": 3.0963954441596277e-06,
+      "loss": 0.0426,
+      "step": 4868
+    },
+    {
+      "epoch": 2.53461738677772,
+      "grad_norm": 0.22193048669411228,
+      "learning_rate": 3.0896260859296035e-06,
+      "loss": 0.0428,
+      "step": 4869
+    },
+    {
+      "epoch": 2.5351379489849037,
+      "grad_norm": 0.233380501372848,
+      "learning_rate": 3.082863648015277e-06,
+      "loss": 0.0448,
+      "step": 4870
+    },
+    {
+      "epoch": 2.5356585111920875,
+      "grad_norm": 0.24772658711823495,
+      "learning_rate": 3.076108132552549e-06,
+      "loss": 0.0464,
+      "step": 4871
+    },
+    {
+      "epoch": 2.5361790733992713,
+      "grad_norm": 0.2333756921366926,
+      "learning_rate": 3.0693595416751207e-06,
+      "loss": 0.0449,
+      "step": 4872
+    },
+    {
+      "epoch": 2.536699635606455,
+      "grad_norm": 0.22180824102144125,
+      "learning_rate": 3.0626178775145175e-06,
+      "loss": 0.0445,
+      "step": 4873
+    },
+    {
+      "epoch": 2.537220197813639,
+      "grad_norm": 0.23966984007720837,
+      "learning_rate": 3.0558831422000695e-06,
+      "loss": 0.0465,
+      "step": 4874
+    },
+    {
+      "epoch": 2.5377407600208226,
+      "grad_norm": 0.23454749585264367,
+      "learning_rate": 3.0491553378589084e-06,
+      "loss": 0.0446,
+      "step": 4875
+    },
+    {
+      "epoch": 2.5382613222280064,
+      "grad_norm": 0.24426571149502913,
+      "learning_rate": 3.042434466615998e-06,
+      "loss": 0.0478,
+      "step": 4876
+    },
+    {
+      "epoch": 2.53878188443519,
+      "grad_norm": 0.2350973499571893,
+      "learning_rate": 3.0357205305940993e-06,
+      "loss": 0.0436,
+      "step": 4877
+    },
+    {
+      "epoch": 2.539302446642374,
+      "grad_norm": 0.2326275497622743,
+      "learning_rate": 3.0290135319137908e-06,
+      "loss": 0.0465,
+      "step": 4878
+    },
+    {
+      "epoch": 2.5398230088495577,
+      "grad_norm": 0.2281455888156863,
+      "learning_rate": 3.0223134726934472e-06,
+      "loss": 0.043,
+      "step": 4879
+    },
+    {
+      "epoch": 2.5403435710567415,
+      "grad_norm": 0.24217000636895653,
+      "learning_rate": 3.015620355049262e-06,
+      "loss": 0.0473,
+      "step": 4880
+    },
+    {
+      "epoch": 2.5408641332639252,
+      "grad_norm": 0.23048820939944398,
+      "learning_rate": 3.0089341810952327e-06,
+      "loss": 0.0443,
+      "step": 4881
+    },
+    {
+      "epoch": 2.541384695471109,
+      "grad_norm": 0.22378119144167463,
+      "learning_rate": 3.0022549529431704e-06,
+      "loss": 0.0438,
+      "step": 4882
+    },
+    {
+      "epoch": 2.541905257678293,
+      "grad_norm": 0.2296752533850872,
+      "learning_rate": 2.995582672702679e-06,
+      "loss": 0.0452,
+      "step": 4883
+    },
+    {
+      "epoch": 2.5424258198854766,
+      "grad_norm": 0.2243414666365788,
+      "learning_rate": 2.9889173424811735e-06,
+      "loss": 0.0442,
+      "step": 4884
+    },
+    {
+      "epoch": 2.5429463820926603,
+      "grad_norm": 0.2304455679727754,
+      "learning_rate": 2.98225896438388e-06,
+      "loss": 0.0443,
+      "step": 4885
+    },
+    {
+      "epoch": 2.543466944299844,
+      "grad_norm": 0.23018490529524177,
+      "learning_rate": 2.9756075405138222e-06,
+      "loss": 0.0454,
+      "step": 4886
+    },
+    {
+      "epoch": 2.5439875065070274,
+      "grad_norm": 0.24131348157714574,
+      "learning_rate": 2.9689630729718337e-06,
+      "loss": 0.0454,
+      "step": 4887
+    },
+    {
+      "epoch": 2.544508068714211,
+      "grad_norm": 0.21657168953862416,
+      "learning_rate": 2.9623255638565384e-06,
+      "loss": 0.0437,
+      "step": 4888
+    },
+    {
+      "epoch": 2.545028630921395,
+      "grad_norm": 0.23781682691026923,
+      "learning_rate": 2.9556950152643757e-06,
+      "loss": 0.0456,
+      "step": 4889
+    },
+    {
+      "epoch": 2.5455491931285787,
+      "grad_norm": 0.23395997912589891,
+      "learning_rate": 2.9490714292895822e-06,
+      "loss": 0.0451,
+      "step": 4890
+    },
+    {
+      "epoch": 2.5460697553357625,
+      "grad_norm": 0.2439448653848847,
+      "learning_rate": 2.9424548080241978e-06,
+      "loss": 0.0458,
+      "step": 4891
+    },
+    {
+      "epoch": 2.5465903175429463,
+      "grad_norm": 0.2184786715073653,
+      "learning_rate": 2.9358451535580534e-06,
+      "loss": 0.043,
+      "step": 4892
+    },
+    {
+      "epoch": 2.54711087975013,
+      "grad_norm": 0.22699072515393065,
+      "learning_rate": 2.9292424679787824e-06,
+      "loss": 0.0438,
+      "step": 4893
+    },
+    {
+      "epoch": 2.547631441957314,
+      "grad_norm": 0.23484480726234366,
+      "learning_rate": 2.9226467533718244e-06,
+      "loss": 0.0445,
+      "step": 4894
+    },
+    {
+      "epoch": 2.5481520041644976,
+      "grad_norm": 0.23093997831111027,
+      "learning_rate": 2.9160580118204104e-06,
+      "loss": 0.0452,
+      "step": 4895
+    },
+    {
+      "epoch": 2.5486725663716814,
+      "grad_norm": 0.23405936670463198,
+      "learning_rate": 2.9094762454055847e-06,
+      "loss": 0.0452,
+      "step": 4896
+    },
+    {
+      "epoch": 2.549193128578865,
+      "grad_norm": 0.23354436325740416,
+      "learning_rate": 2.902901456206156e-06,
+      "loss": 0.0452,
+      "step": 4897
+    },
+    {
+      "epoch": 2.549713690786049,
+      "grad_norm": 0.25244543484114845,
+      "learning_rate": 2.89633364629876e-06,
+      "loss": 0.0492,
+      "step": 4898
+    },
+    {
+      "epoch": 2.5502342529932327,
+      "grad_norm": 0.22617895203553662,
+      "learning_rate": 2.889772817757813e-06,
+      "loss": 0.0448,
+      "step": 4899
+    },
+    {
+      "epoch": 2.5507548152004165,
+      "grad_norm": 0.23586461110043583,
+      "learning_rate": 2.8832189726555383e-06,
+      "loss": 0.0461,
+      "step": 4900
+    },
+    {
+      "epoch": 2.5512753774076002,
+      "grad_norm": 0.22350522590122324,
+      "learning_rate": 2.8766721130619315e-06,
+      "loss": 0.0464,
+      "step": 4901
+    },
+    {
+      "epoch": 2.551795939614784,
+      "grad_norm": 0.23067382345830922,
+      "learning_rate": 2.8701322410448095e-06,
+      "loss": 0.0446,
+      "step": 4902
+    },
+    {
+      "epoch": 2.552316501821968,
+      "grad_norm": 0.23541409906186989,
+      "learning_rate": 2.8635993586697553e-06,
+      "loss": 0.0455,
+      "step": 4903
+    },
+    {
+      "epoch": 2.5528370640291516,
+      "grad_norm": 0.2196421849316751,
+      "learning_rate": 2.8570734680001627e-06,
+      "loss": 0.0417,
+      "step": 4904
+    },
+    {
+      "epoch": 2.5533576262363353,
+      "grad_norm": 0.21999629966823,
+      "learning_rate": 2.850554571097211e-06,
+      "loss": 0.0435,
+      "step": 4905
+    },
+    {
+      "epoch": 2.553878188443519,
+      "grad_norm": 0.22111791678119153,
+      "learning_rate": 2.844042670019878e-06,
+      "loss": 0.0446,
+      "step": 4906
+    },
+    {
+      "epoch": 2.554398750650703,
+      "grad_norm": 0.23774001294637379,
+      "learning_rate": 2.837537766824913e-06,
+      "loss": 0.0455,
+      "step": 4907
+    },
+    {
+      "epoch": 2.5549193128578866,
+      "grad_norm": 0.22855614328754512,
+      "learning_rate": 2.8310398635668755e-06,
+      "loss": 0.0449,
+      "step": 4908
+    },
+    {
+      "epoch": 2.5554398750650704,
+      "grad_norm": 0.22786759432794976,
+      "learning_rate": 2.824548962298107e-06,
+      "loss": 0.0436,
+      "step": 4909
+    },
+    {
+      "epoch": 2.555960437272254,
+      "grad_norm": 0.2217342917078937,
+      "learning_rate": 2.8180650650687287e-06,
+      "loss": 0.0432,
+      "step": 4910
+    },
+    {
+      "epoch": 2.5564809994794375,
+      "grad_norm": 0.2337668784988823,
+      "learning_rate": 2.811588173926666e-06,
+      "loss": 0.0463,
+      "step": 4911
+    },
+    {
+      "epoch": 2.5570015616866213,
+      "grad_norm": 0.23907810436310875,
+      "learning_rate": 2.8051182909176133e-06,
+      "loss": 0.0467,
+      "step": 4912
+    },
+    {
+      "epoch": 2.557522123893805,
+      "grad_norm": 0.23815342311596505,
+      "learning_rate": 2.7986554180850665e-06,
+      "loss": 0.0464,
+      "step": 4913
+    },
+    {
+      "epoch": 2.558042686100989,
+      "grad_norm": 0.2310077217504325,
+      "learning_rate": 2.7921995574702986e-06,
+      "loss": 0.0449,
+      "step": 4914
+    },
+    {
+      "epoch": 2.5585632483081726,
+      "grad_norm": 0.22853140276683548,
+      "learning_rate": 2.7857507111123755e-06,
+      "loss": 0.0456,
+      "step": 4915
+    },
+    {
+      "epoch": 2.5590838105153564,
+      "grad_norm": 0.24059012312722752,
+      "learning_rate": 2.779308881048137e-06,
+      "loss": 0.0431,
+      "step": 4916
+    },
+    {
+      "epoch": 2.55960437272254,
+      "grad_norm": 0.24488720395307423,
+      "learning_rate": 2.7728740693122147e-06,
+      "loss": 0.0471,
+      "step": 4917
+    },
+    {
+      "epoch": 2.560124934929724,
+      "grad_norm": 0.22806652160138638,
+      "learning_rate": 2.7664462779370293e-06,
+      "loss": 0.0449,
+      "step": 4918
+    },
+    {
+      "epoch": 2.5606454971369077,
+      "grad_norm": 0.23288075306841058,
+      "learning_rate": 2.7600255089527626e-06,
+      "loss": 0.0447,
+      "step": 4919
+    },
+    {
+      "epoch": 2.5611660593440915,
+      "grad_norm": 0.23782264448640814,
+      "learning_rate": 2.7536117643874067e-06,
+      "loss": 0.0473,
+      "step": 4920
+    },
+    {
+      "epoch": 2.5616866215512752,
+      "grad_norm": 0.23345197126927117,
+      "learning_rate": 2.747205046266707e-06,
+      "loss": 0.0452,
+      "step": 4921
+    },
+    {
+      "epoch": 2.562207183758459,
+      "grad_norm": 0.22295930818431847,
+      "learning_rate": 2.7408053566142124e-06,
+      "loss": 0.0453,
+      "step": 4922
+    },
+    {
+      "epoch": 2.562727745965643,
+      "grad_norm": 0.2400982174167174,
+      "learning_rate": 2.734412697451236e-06,
+      "loss": 0.0448,
+      "step": 4923
+    },
+    {
+      "epoch": 2.5632483081728266,
+      "grad_norm": 0.2226346510144299,
+      "learning_rate": 2.7280270707968874e-06,
+      "loss": 0.0413,
+      "step": 4924
+    },
+    {
+      "epoch": 2.5637688703800103,
+      "grad_norm": 0.23435826768260656,
+      "learning_rate": 2.721648478668032e-06,
+      "loss": 0.044,
+      "step": 4925
+    },
+    {
+      "epoch": 2.564289432587194,
+      "grad_norm": 0.22124720780327345,
+      "learning_rate": 2.715276923079335e-06,
+      "loss": 0.0441,
+      "step": 4926
+    },
+    {
+      "epoch": 2.564809994794378,
+      "grad_norm": 0.23111451493845334,
+      "learning_rate": 2.708912406043229e-06,
+      "loss": 0.0443,
+      "step": 4927
+    },
+    {
+      "epoch": 2.5653305570015617,
+      "grad_norm": 0.22812981752484784,
+      "learning_rate": 2.70255492956992e-06,
+      "loss": 0.0437,
+      "step": 4928
+    },
+    {
+      "epoch": 2.5658511192087454,
+      "grad_norm": 0.2242696338157624,
+      "learning_rate": 2.6962044956674035e-06,
+      "loss": 0.044,
+      "step": 4929
+    },
+    {
+      "epoch": 2.566371681415929,
+      "grad_norm": 0.2526958000740198,
+      "learning_rate": 2.689861106341432e-06,
+      "loss": 0.0468,
+      "step": 4930
+    },
+    {
+      "epoch": 2.566892243623113,
+      "grad_norm": 0.2275462053374598,
+      "learning_rate": 2.6835247635955463e-06,
+      "loss": 0.0447,
+      "step": 4931
+    },
+    {
+      "epoch": 2.5674128058302967,
+      "grad_norm": 0.2237994819214662,
+      "learning_rate": 2.6771954694310597e-06,
+      "loss": 0.0438,
+      "step": 4932
+    },
+    {
+      "epoch": 2.5679333680374805,
+      "grad_norm": 0.22406574776804847,
+      "learning_rate": 2.670873225847062e-06,
+      "loss": 0.0435,
+      "step": 4933
+    },
+    {
+      "epoch": 2.5684539302446643,
+      "grad_norm": 0.2270917489269596,
+      "learning_rate": 2.664558034840403e-06,
+      "loss": 0.0447,
+      "step": 4934
+    },
+    {
+      "epoch": 2.568974492451848,
+      "grad_norm": 0.23089342107757108,
+      "learning_rate": 2.658249898405718e-06,
+      "loss": 0.0456,
+      "step": 4935
+    },
+    {
+      "epoch": 2.569495054659032,
+      "grad_norm": 0.2313393425399634,
+      "learning_rate": 2.6519488185354157e-06,
+      "loss": 0.0436,
+      "step": 4936
+    },
+    {
+      "epoch": 2.5700156168662156,
+      "grad_norm": 0.2221957268696646,
+      "learning_rate": 2.6456547972196625e-06,
+      "loss": 0.0443,
+      "step": 4937
+    },
+    {
+      "epoch": 2.5705361790733994,
+      "grad_norm": 0.23170888298200445,
+      "learning_rate": 2.6393678364464074e-06,
+      "loss": 0.0452,
+      "step": 4938
+    },
+    {
+      "epoch": 2.571056741280583,
+      "grad_norm": 0.2308739427020387,
+      "learning_rate": 2.6330879382013617e-06,
+      "loss": 0.0466,
+      "step": 4939
+    },
+    {
+      "epoch": 2.571577303487767,
+      "grad_norm": 0.23012613337882634,
+      "learning_rate": 2.6268151044680113e-06,
+      "loss": 0.0443,
+      "step": 4940
+    },
+    {
+      "epoch": 2.5720978656949507,
+      "grad_norm": 0.23042253316392428,
+      "learning_rate": 2.620549337227607e-06,
+      "loss": 0.0443,
+      "step": 4941
+    },
+    {
+      "epoch": 2.5726184279021345,
+      "grad_norm": 0.22649967085410116,
+      "learning_rate": 2.6142906384591798e-06,
+      "loss": 0.0443,
+      "step": 4942
+    },
+    {
+      "epoch": 2.5731389901093182,
+      "grad_norm": 0.22980513686165474,
+      "learning_rate": 2.6080390101395043e-06,
+      "loss": 0.0443,
+      "step": 4943
+    },
+    {
+      "epoch": 2.573659552316502,
+      "grad_norm": 0.22269196931932628,
+      "learning_rate": 2.6017944542431393e-06,
+      "loss": 0.043,
+      "step": 4944
+    },
+    {
+      "epoch": 2.574180114523686,
+      "grad_norm": 0.2220316442992769,
+      "learning_rate": 2.5955569727424163e-06,
+      "loss": 0.043,
+      "step": 4945
+    },
+    {
+      "epoch": 2.5747006767308696,
+      "grad_norm": 0.2382111826632995,
+      "learning_rate": 2.589326567607406e-06,
+      "loss": 0.045,
+      "step": 4946
+    },
+    {
+      "epoch": 2.5752212389380533,
+      "grad_norm": 0.2300585724875798,
+      "learning_rate": 2.5831032408059758e-06,
+      "loss": 0.0431,
+      "step": 4947
+    },
+    {
+      "epoch": 2.575741801145237,
+      "grad_norm": 0.2279139534194501,
+      "learning_rate": 2.576886994303729e-06,
+      "loss": 0.0461,
+      "step": 4948
+    },
+    {
+      "epoch": 2.576262363352421,
+      "grad_norm": 0.23023216030690563,
+      "learning_rate": 2.5706778300640527e-06,
+      "loss": 0.0454,
+      "step": 4949
+    },
+    {
+      "epoch": 2.5767829255596046,
+      "grad_norm": 0.22354558169408728,
+      "learning_rate": 2.564475750048087e-06,
+      "loss": 0.0452,
+      "step": 4950
+    },
+    {
+      "epoch": 2.577303487766788,
+      "grad_norm": 0.23125790378504177,
+      "learning_rate": 2.5582807562147455e-06,
+      "loss": 0.0448,
+      "step": 4951
+    },
+    {
+      "epoch": 2.5778240499739717,
+      "grad_norm": 0.22611942246092737,
+      "learning_rate": 2.552092850520682e-06,
+      "loss": 0.0445,
+      "step": 4952
+    },
+    {
+      "epoch": 2.5783446121811555,
+      "grad_norm": 0.2241937478852757,
+      "learning_rate": 2.545912034920331e-06,
+      "loss": 0.0434,
+      "step": 4953
+    },
+    {
+      "epoch": 2.5788651743883393,
+      "grad_norm": 0.228033785778967,
+      "learning_rate": 2.5397383113658883e-06,
+      "loss": 0.0461,
+      "step": 4954
+    },
+    {
+      "epoch": 2.579385736595523,
+      "grad_norm": 0.2248949405146001,
+      "learning_rate": 2.533571681807295e-06,
+      "loss": 0.0429,
+      "step": 4955
+    },
+    {
+      "epoch": 2.579906298802707,
+      "grad_norm": 0.231228519351519,
+      "learning_rate": 2.527412148192265e-06,
+      "loss": 0.0463,
+      "step": 4956
+    },
+    {
+      "epoch": 2.5804268610098906,
+      "grad_norm": 0.2171525044347602,
+      "learning_rate": 2.5212597124662564e-06,
+      "loss": 0.0439,
+      "step": 4957
+    },
+    {
+      "epoch": 2.5809474232170744,
+      "grad_norm": 0.23091260307400774,
+      "learning_rate": 2.5151143765725027e-06,
+      "loss": 0.046,
+      "step": 4958
+    },
+    {
+      "epoch": 2.581467985424258,
+      "grad_norm": 0.21624018139865478,
+      "learning_rate": 2.5089761424519853e-06,
+      "loss": 0.0422,
+      "step": 4959
+    },
+    {
+      "epoch": 2.581988547631442,
+      "grad_norm": 0.224287761302505,
+      "learning_rate": 2.5028450120434505e-06,
+      "loss": 0.0438,
+      "step": 4960
+    },
+    {
+      "epoch": 2.5825091098386257,
+      "grad_norm": 0.2339375258097004,
+      "learning_rate": 2.4967209872833822e-06,
+      "loss": 0.0449,
+      "step": 4961
+    },
+    {
+      "epoch": 2.5830296720458095,
+      "grad_norm": 0.2317454115663667,
+      "learning_rate": 2.4906040701060367e-06,
+      "loss": 0.0453,
+      "step": 4962
+    },
+    {
+      "epoch": 2.5835502342529932,
+      "grad_norm": 0.224028657841797,
+      "learning_rate": 2.484494262443429e-06,
+      "loss": 0.0427,
+      "step": 4963
+    },
+    {
+      "epoch": 2.584070796460177,
+      "grad_norm": 0.22363155592376474,
+      "learning_rate": 2.4783915662253103e-06,
+      "loss": 0.0432,
+      "step": 4964
+    },
+    {
+      "epoch": 2.584591358667361,
+      "grad_norm": 0.24028921492138963,
+      "learning_rate": 2.472295983379205e-06,
+      "loss": 0.0463,
+      "step": 4965
+    },
+    {
+      "epoch": 2.5851119208745446,
+      "grad_norm": 0.21789704784809844,
+      "learning_rate": 2.466207515830374e-06,
+      "loss": 0.043,
+      "step": 4966
+    },
+    {
+      "epoch": 2.5856324830817283,
+      "grad_norm": 0.23331471651872346,
+      "learning_rate": 2.4601261655018385e-06,
+      "loss": 0.045,
+      "step": 4967
+    },
+    {
+      "epoch": 2.586153045288912,
+      "grad_norm": 0.23970513589566356,
+      "learning_rate": 2.4540519343143774e-06,
+      "loss": 0.0456,
+      "step": 4968
+    },
+    {
+      "epoch": 2.586673607496096,
+      "grad_norm": 0.22417695780216113,
+      "learning_rate": 2.447984824186514e-06,
+      "loss": 0.043,
+      "step": 4969
+    },
+    {
+      "epoch": 2.5871941697032796,
+      "grad_norm": 0.22276549554035466,
+      "learning_rate": 2.441924837034529e-06,
+      "loss": 0.0423,
+      "step": 4970
+    },
+    {
+      "epoch": 2.5877147319104634,
+      "grad_norm": 0.22293938392333162,
+      "learning_rate": 2.4358719747724356e-06,
+      "loss": 0.0426,
+      "step": 4971
+    },
+    {
+      "epoch": 2.588235294117647,
+      "grad_norm": 0.23230080511957873,
+      "learning_rate": 2.429826239312022e-06,
+      "loss": 0.0465,
+      "step": 4972
+    },
+    {
+      "epoch": 2.588755856324831,
+      "grad_norm": 0.24276681968930813,
+      "learning_rate": 2.4237876325628017e-06,
+      "loss": 0.0445,
+      "step": 4973
+    },
+    {
+      "epoch": 2.5892764185320147,
+      "grad_norm": 0.22442049970967431,
+      "learning_rate": 2.417756156432055e-06,
+      "loss": 0.0432,
+      "step": 4974
+    },
+    {
+      "epoch": 2.589796980739198,
+      "grad_norm": 0.23419001522856595,
+      "learning_rate": 2.411731812824808e-06,
+      "loss": 0.0456,
+      "step": 4975
+    },
+    {
+      "epoch": 2.590317542946382,
+      "grad_norm": 0.22466863339121676,
+      "learning_rate": 2.4057146036438135e-06,
+      "loss": 0.0439,
+      "step": 4976
+    },
+    {
+      "epoch": 2.5908381051535656,
+      "grad_norm": 0.22312298115872314,
+      "learning_rate": 2.3997045307895954e-06,
+      "loss": 0.0432,
+      "step": 4977
+    },
+    {
+      "epoch": 2.5913586673607494,
+      "grad_norm": 0.23722999672753922,
+      "learning_rate": 2.3937015961604165e-06,
+      "loss": 0.0461,
+      "step": 4978
+    },
+    {
+      "epoch": 2.591879229567933,
+      "grad_norm": 0.22912082388275234,
+      "learning_rate": 2.3877058016522816e-06,
+      "loss": 0.0447,
+      "step": 4979
+    },
+    {
+      "epoch": 2.592399791775117,
+      "grad_norm": 0.2355711003414976,
+      "learning_rate": 2.381717149158935e-06,
+      "loss": 0.0458,
+      "step": 4980
+    },
+    {
+      "epoch": 2.5929203539823007,
+      "grad_norm": 0.2281375391556885,
+      "learning_rate": 2.3757356405718854e-06,
+      "loss": 0.0441,
+      "step": 4981
+    },
+    {
+      "epoch": 2.5934409161894845,
+      "grad_norm": 0.23163425813558036,
+      "learning_rate": 2.369761277780355e-06,
+      "loss": 0.0445,
+      "step": 4982
+    },
+    {
+      "epoch": 2.5939614783966682,
+      "grad_norm": 0.24549021694669057,
+      "learning_rate": 2.3637940626713346e-06,
+      "loss": 0.0471,
+      "step": 4983
+    },
+    {
+      "epoch": 2.594482040603852,
+      "grad_norm": 0.2406789780817914,
+      "learning_rate": 2.357833997129552e-06,
+      "loss": 0.0463,
+      "step": 4984
+    },
+    {
+      "epoch": 2.595002602811036,
+      "grad_norm": 0.23803338315683714,
+      "learning_rate": 2.3518810830374634e-06,
+      "loss": 0.0448,
+      "step": 4985
+    },
+    {
+      "epoch": 2.5955231650182196,
+      "grad_norm": 0.2315546330265174,
+      "learning_rate": 2.3459353222752835e-06,
+      "loss": 0.0471,
+      "step": 4986
+    },
+    {
+      "epoch": 2.5960437272254033,
+      "grad_norm": 0.21931566525318558,
+      "learning_rate": 2.3399967167209576e-06,
+      "loss": 0.0425,
+      "step": 4987
+    },
+    {
+      "epoch": 2.596564289432587,
+      "grad_norm": 0.22188942682547205,
+      "learning_rate": 2.3340652682501767e-06,
+      "loss": 0.0432,
+      "step": 4988
+    },
+    {
+      "epoch": 2.597084851639771,
+      "grad_norm": 0.2292900755538919,
+      "learning_rate": 2.328140978736365e-06,
+      "loss": 0.0432,
+      "step": 4989
+    },
+    {
+      "epoch": 2.5976054138469546,
+      "grad_norm": 0.23064657071914493,
+      "learning_rate": 2.3222238500506923e-06,
+      "loss": 0.0449,
+      "step": 4990
+    },
+    {
+      "epoch": 2.5981259760541384,
+      "grad_norm": 0.22546541475421025,
+      "learning_rate": 2.3163138840620546e-06,
+      "loss": 0.0439,
+      "step": 4991
+    },
+    {
+      "epoch": 2.598646538261322,
+      "grad_norm": 0.23350588400038483,
+      "learning_rate": 2.310411082637101e-06,
+      "loss": 0.0465,
+      "step": 4992
+    },
+    {
+      "epoch": 2.599167100468506,
+      "grad_norm": 0.2314847500853126,
+      "learning_rate": 2.3045154476402154e-06,
+      "loss": 0.0453,
+      "step": 4993
+    },
+    {
+      "epoch": 2.5996876626756897,
+      "grad_norm": 0.23217992293328432,
+      "learning_rate": 2.2986269809335022e-06,
+      "loss": 0.0446,
+      "step": 4994
+    },
+    {
+      "epoch": 2.6002082248828735,
+      "grad_norm": 0.21965356479125087,
+      "learning_rate": 2.2927456843768206e-06,
+      "loss": 0.0426,
+      "step": 4995
+    },
+    {
+      "epoch": 2.6007287870900573,
+      "grad_norm": 0.2350569114494813,
+      "learning_rate": 2.286871559827758e-06,
+      "loss": 0.0451,
+      "step": 4996
+    },
+    {
+      "epoch": 2.601249349297241,
+      "grad_norm": 0.24155775370205634,
+      "learning_rate": 2.2810046091416374e-06,
+      "loss": 0.0479,
+      "step": 4997
+    },
+    {
+      "epoch": 2.601769911504425,
+      "grad_norm": 0.21928794475027916,
+      "learning_rate": 2.2751448341715083e-06,
+      "loss": 0.0453,
+      "step": 4998
+    },
+    {
+      "epoch": 2.6022904737116086,
+      "grad_norm": 0.2478180725379701,
+      "learning_rate": 2.269292236768167e-06,
+      "loss": 0.0443,
+      "step": 4999
+    },
+    {
+      "epoch": 2.6028110359187924,
+      "grad_norm": 0.22140285555828385,
+      "learning_rate": 2.2634468187801296e-06,
+      "loss": 0.0439,
+      "step": 5000
+    },
+    {
+      "epoch": 2.603331598125976,
+      "grad_norm": 0.22451869562094484,
+      "learning_rate": 2.257608582053655e-06,
+      "loss": 0.0446,
+      "step": 5001
+    },
+    {
+      "epoch": 2.60385216033316,
+      "grad_norm": 0.23546031458995,
+      "learning_rate": 2.251777528432736e-06,
+      "loss": 0.0451,
+      "step": 5002
+    },
+    {
+      "epoch": 2.6043727225403437,
+      "grad_norm": 0.2324359389521522,
+      "learning_rate": 2.2459536597590786e-06,
+      "loss": 0.0452,
+      "step": 5003
+    },
+    {
+      "epoch": 2.6048932847475275,
+      "grad_norm": 0.2206269507203779,
+      "learning_rate": 2.240136977872137e-06,
+      "loss": 0.0432,
+      "step": 5004
+    },
+    {
+      "epoch": 2.6054138469547112,
+      "grad_norm": 0.22420136480404548,
+      "learning_rate": 2.2343274846090918e-06,
+      "loss": 0.0438,
+      "step": 5005
+    },
+    {
+      "epoch": 2.605934409161895,
+      "grad_norm": 0.22114323880099848,
+      "learning_rate": 2.228525181804855e-06,
+      "loss": 0.0447,
+      "step": 5006
+    },
+    {
+      "epoch": 2.6064549713690788,
+      "grad_norm": 0.22862454975827587,
+      "learning_rate": 2.2227300712920534e-06,
+      "loss": 0.0433,
+      "step": 5007
+    },
+    {
+      "epoch": 2.6069755335762625,
+      "grad_norm": 0.22833262874322383,
+      "learning_rate": 2.2169421549010636e-06,
+      "loss": 0.0444,
+      "step": 5008
+    },
+    {
+      "epoch": 2.6074960957834463,
+      "grad_norm": 0.23904644665049268,
+      "learning_rate": 2.2111614344599683e-06,
+      "loss": 0.0459,
+      "step": 5009
+    },
+    {
+      "epoch": 2.60801665799063,
+      "grad_norm": 0.2305455365922837,
+      "learning_rate": 2.205387911794593e-06,
+      "loss": 0.0453,
+      "step": 5010
+    },
+    {
+      "epoch": 2.608537220197814,
+      "grad_norm": 0.2328552560660912,
+      "learning_rate": 2.19962158872849e-06,
+      "loss": 0.0452,
+      "step": 5011
+    },
+    {
+      "epoch": 2.6090577824049976,
+      "grad_norm": 0.22388814260008266,
+      "learning_rate": 2.1938624670829218e-06,
+      "loss": 0.045,
+      "step": 5012
+    },
+    {
+      "epoch": 2.6095783446121814,
+      "grad_norm": 0.22661115206828292,
+      "learning_rate": 2.1881105486768945e-06,
+      "loss": 0.0447,
+      "step": 5013
+    },
+    {
+      "epoch": 2.610098906819365,
+      "grad_norm": 0.21667312091492352,
+      "learning_rate": 2.18236583532713e-06,
+      "loss": 0.0415,
+      "step": 5014
+    },
+    {
+      "epoch": 2.6106194690265485,
+      "grad_norm": 0.23591261253651902,
+      "learning_rate": 2.1766283288480793e-06,
+      "loss": 0.0442,
+      "step": 5015
+    },
+    {
+      "epoch": 2.6111400312337323,
+      "grad_norm": 0.2345912526719469,
+      "learning_rate": 2.1708980310519045e-06,
+      "loss": 0.0437,
+      "step": 5016
+    },
+    {
+      "epoch": 2.611660593440916,
+      "grad_norm": 0.2268571102942271,
+      "learning_rate": 2.165174943748513e-06,
+      "loss": 0.044,
+      "step": 5017
+    },
+    {
+      "epoch": 2.6121811556481,
+      "grad_norm": 0.22718616461448765,
+      "learning_rate": 2.159459068745512e-06,
+      "loss": 0.0449,
+      "step": 5018
+    },
+    {
+      "epoch": 2.6127017178552836,
+      "grad_norm": 0.22275521228810305,
+      "learning_rate": 2.1537504078482427e-06,
+      "loss": 0.0413,
+      "step": 5019
+    },
+    {
+      "epoch": 2.6132222800624674,
+      "grad_norm": 0.23325969759410764,
+      "learning_rate": 2.1480489628597723e-06,
+      "loss": 0.0457,
+      "step": 5020
+    },
+    {
+      "epoch": 2.613742842269651,
+      "grad_norm": 0.23677448755403346,
+      "learning_rate": 2.142354735580873e-06,
+      "loss": 0.0458,
+      "step": 5021
+    },
+    {
+      "epoch": 2.614263404476835,
+      "grad_norm": 0.2231689376626884,
+      "learning_rate": 2.1366677278100487e-06,
+      "loss": 0.0445,
+      "step": 5022
+    },
+    {
+      "epoch": 2.6147839666840187,
+      "grad_norm": 0.22687556800748557,
+      "learning_rate": 2.1309879413435292e-06,
+      "loss": 0.0447,
+      "step": 5023
+    },
+    {
+      "epoch": 2.6153045288912025,
+      "grad_norm": 0.23231196286407194,
+      "learning_rate": 2.125315377975251e-06,
+      "loss": 0.0447,
+      "step": 5024
+    },
+    {
+      "epoch": 2.6158250910983862,
+      "grad_norm": 0.23049253549431217,
+      "learning_rate": 2.1196500394968678e-06,
+      "loss": 0.0455,
+      "step": 5025
+    },
+    {
+      "epoch": 2.61634565330557,
+      "grad_norm": 0.22425972462104576,
+      "learning_rate": 2.113991927697767e-06,
+      "loss": 0.0449,
+      "step": 5026
+    },
+    {
+      "epoch": 2.616866215512754,
+      "grad_norm": 0.22468948485303378,
+      "learning_rate": 2.1083410443650365e-06,
+      "loss": 0.0444,
+      "step": 5027
+    },
+    {
+      "epoch": 2.6173867777199376,
+      "grad_norm": 0.2384716964753025,
+      "learning_rate": 2.102697391283487e-06,
+      "loss": 0.0461,
+      "step": 5028
+    },
+    {
+      "epoch": 2.6179073399271213,
+      "grad_norm": 0.2316329860082873,
+      "learning_rate": 2.0970609702356563e-06,
+      "loss": 0.045,
+      "step": 5029
+    },
+    {
+      "epoch": 2.618427902134305,
+      "grad_norm": 0.22706235586772133,
+      "learning_rate": 2.0914317830017764e-06,
+      "loss": 0.0442,
+      "step": 5030
+    },
+    {
+      "epoch": 2.618948464341489,
+      "grad_norm": 0.2283217603622717,
+      "learning_rate": 2.085809831359814e-06,
+      "loss": 0.0434,
+      "step": 5031
+    },
+    {
+      "epoch": 2.6194690265486726,
+      "grad_norm": 0.23412543514217077,
+      "learning_rate": 2.08019511708544e-06,
+      "loss": 0.044,
+      "step": 5032
+    },
+    {
+      "epoch": 2.6199895887558564,
+      "grad_norm": 0.23322446403477373,
+      "learning_rate": 2.0745876419520442e-06,
+      "loss": 0.0454,
+      "step": 5033
+    },
+    {
+      "epoch": 2.62051015096304,
+      "grad_norm": 0.2225537808805055,
+      "learning_rate": 2.0689874077307325e-06,
+      "loss": 0.0441,
+      "step": 5034
+    },
+    {
+      "epoch": 2.621030713170224,
+      "grad_norm": 0.22326275306638776,
+      "learning_rate": 2.0633944161903144e-06,
+      "loss": 0.0427,
+      "step": 5035
+    },
+    {
+      "epoch": 2.6215512753774077,
+      "grad_norm": 0.2242939602375942,
+      "learning_rate": 2.0578086690973135e-06,
+      "loss": 0.045,
+      "step": 5036
+    },
+    {
+      "epoch": 2.6220718375845915,
+      "grad_norm": 0.22728481687598673,
+      "learning_rate": 2.052230168215971e-06,
+      "loss": 0.0456,
+      "step": 5037
+    },
+    {
+      "epoch": 2.6225923997917753,
+      "grad_norm": 0.22262615082530046,
+      "learning_rate": 2.0466589153082416e-06,
+      "loss": 0.0433,
+      "step": 5038
+    },
+    {
+      "epoch": 2.6231129619989586,
+      "grad_norm": 0.22560826041711385,
+      "learning_rate": 2.041094912133784e-06,
+      "loss": 0.0446,
+      "step": 5039
+    },
+    {
+      "epoch": 2.6236335242061424,
+      "grad_norm": 0.23114517964831685,
+      "learning_rate": 2.035538160449968e-06,
+      "loss": 0.0466,
+      "step": 5040
+    },
+    {
+      "epoch": 2.624154086413326,
+      "grad_norm": 0.2420652950928089,
+      "learning_rate": 2.0299886620118722e-06,
+      "loss": 0.0458,
+      "step": 5041
+    },
+    {
+      "epoch": 2.62467464862051,
+      "grad_norm": 0.22783597835324096,
+      "learning_rate": 2.02444641857229e-06,
+      "loss": 0.0446,
+      "step": 5042
+    },
+    {
+      "epoch": 2.6251952108276937,
+      "grad_norm": 0.22160692042897737,
+      "learning_rate": 2.018911431881723e-06,
+      "loss": 0.0446,
+      "step": 5043
+    },
+    {
+      "epoch": 2.6257157730348775,
+      "grad_norm": 0.2322363544425751,
+      "learning_rate": 2.0133837036883734e-06,
+      "loss": 0.0451,
+      "step": 5044
+    },
+    {
+      "epoch": 2.6262363352420612,
+      "grad_norm": 0.22850250345514178,
+      "learning_rate": 2.007863235738153e-06,
+      "loss": 0.0452,
+      "step": 5045
+    },
+    {
+      "epoch": 2.626756897449245,
+      "grad_norm": 0.22005963804167653,
+      "learning_rate": 2.0023500297746828e-06,
+      "loss": 0.0439,
+      "step": 5046
+    },
+    {
+      "epoch": 2.627277459656429,
+      "grad_norm": 0.2215482537319163,
+      "learning_rate": 1.9968440875392902e-06,
+      "loss": 0.0429,
+      "step": 5047
+    },
+    {
+      "epoch": 2.6277980218636126,
+      "grad_norm": 0.22063378797132666,
+      "learning_rate": 1.9913454107710173e-06,
+      "loss": 0.0425,
+      "step": 5048
+    },
+    {
+      "epoch": 2.6283185840707963,
+      "grad_norm": 0.2300782059424934,
+      "learning_rate": 1.9858540012065886e-06,
+      "loss": 0.0443,
+      "step": 5049
+    },
+    {
+      "epoch": 2.62883914627798,
+      "grad_norm": 0.21421042163884607,
+      "learning_rate": 1.9803698605804497e-06,
+      "loss": 0.0419,
+      "step": 5050
+    },
+    {
+      "epoch": 2.629359708485164,
+      "grad_norm": 0.2273905924198931,
+      "learning_rate": 1.974892990624752e-06,
+      "loss": 0.0448,
+      "step": 5051
+    },
+    {
+      "epoch": 2.6298802706923476,
+      "grad_norm": 0.23000101131311979,
+      "learning_rate": 1.96942339306935e-06,
+      "loss": 0.0449,
+      "step": 5052
+    },
+    {
+      "epoch": 2.6304008328995314,
+      "grad_norm": 0.2306219483746907,
+      "learning_rate": 1.96396106964179e-06,
+      "loss": 0.0434,
+      "step": 5053
+    },
+    {
+      "epoch": 2.630921395106715,
+      "grad_norm": 0.2254745000498489,
+      "learning_rate": 1.9585060220673247e-06,
+      "loss": 0.0441,
+      "step": 5054
+    },
+    {
+      "epoch": 2.631441957313899,
+      "grad_norm": 0.23129142941993588,
+      "learning_rate": 1.953058252068915e-06,
+      "loss": 0.0448,
+      "step": 5055
+    },
+    {
+      "epoch": 2.6319625195210827,
+      "grad_norm": 0.22233175432556077,
+      "learning_rate": 1.9476177613672237e-06,
+      "loss": 0.0438,
+      "step": 5056
+    },
+    {
+      "epoch": 2.6324830817282665,
+      "grad_norm": 0.22444580299782396,
+      "learning_rate": 1.942184551680612e-06,
+      "loss": 0.0451,
+      "step": 5057
+    },
+    {
+      "epoch": 2.6330036439354503,
+      "grad_norm": 0.22412724223445896,
+      "learning_rate": 1.9367586247251323e-06,
+      "loss": 0.044,
+      "step": 5058
+    },
+    {
+      "epoch": 2.633524206142634,
+      "grad_norm": 0.24134427074612705,
+      "learning_rate": 1.93133998221455e-06,
+      "loss": 0.0462,
+      "step": 5059
+    },
+    {
+      "epoch": 2.634044768349818,
+      "grad_norm": 0.23306486452552985,
+      "learning_rate": 1.9259286258603263e-06,
+      "loss": 0.0448,
+      "step": 5060
+    },
+    {
+      "epoch": 2.6345653305570016,
+      "grad_norm": 0.2357629035963088,
+      "learning_rate": 1.9205245573716197e-06,
+      "loss": 0.0464,
+      "step": 5061
+    },
+    {
+      "epoch": 2.6350858927641854,
+      "grad_norm": 0.22558172778330046,
+      "learning_rate": 1.9151277784552864e-06,
+      "loss": 0.0442,
+      "step": 5062
+    },
+    {
+      "epoch": 2.635606454971369,
+      "grad_norm": 0.23769331834375482,
+      "learning_rate": 1.9097382908158713e-06,
+      "loss": 0.0448,
+      "step": 5063
+    },
+    {
+      "epoch": 2.636127017178553,
+      "grad_norm": 0.22352280580432798,
+      "learning_rate": 1.9043560961556323e-06,
+      "loss": 0.0448,
+      "step": 5064
+    },
+    {
+      "epoch": 2.6366475793857367,
+      "grad_norm": 0.24155333261345938,
+      "learning_rate": 1.898981196174518e-06,
+      "loss": 0.0456,
+      "step": 5065
+    },
+    {
+      "epoch": 2.6371681415929205,
+      "grad_norm": 0.2153187541406319,
+      "learning_rate": 1.8936135925701732e-06,
+      "loss": 0.0428,
+      "step": 5066
+    },
+    {
+      "epoch": 2.6376887038001042,
+      "grad_norm": 0.23310151041795749,
+      "learning_rate": 1.8882532870379331e-06,
+      "loss": 0.0447,
+      "step": 5067
+    },
+    {
+      "epoch": 2.638209266007288,
+      "grad_norm": 0.23186210725957584,
+      "learning_rate": 1.8829002812708302e-06,
+      "loss": 0.0456,
+      "step": 5068
+    },
+    {
+      "epoch": 2.6387298282144718,
+      "grad_norm": 0.24386297462518527,
+      "learning_rate": 1.8775545769595975e-06,
+      "loss": 0.0454,
+      "step": 5069
+    },
+    {
+      "epoch": 2.6392503904216555,
+      "grad_norm": 0.21918598958276325,
+      "learning_rate": 1.8722161757926597e-06,
+      "loss": 0.0433,
+      "step": 5070
+    },
+    {
+      "epoch": 2.6397709526288393,
+      "grad_norm": 0.22662770705374358,
+      "learning_rate": 1.866885079456121e-06,
+      "loss": 0.0403,
+      "step": 5071
+    },
+    {
+      "epoch": 2.640291514836023,
+      "grad_norm": 0.22853002659541766,
+      "learning_rate": 1.8615612896338036e-06,
+      "loss": 0.0438,
+      "step": 5072
+    },
+    {
+      "epoch": 2.640812077043207,
+      "grad_norm": 0.22941524162811514,
+      "learning_rate": 1.856244808007196e-06,
+      "loss": 0.044,
+      "step": 5073
+    },
+    {
+      "epoch": 2.6413326392503906,
+      "grad_norm": 0.22276184054208367,
+      "learning_rate": 1.8509356362554963e-06,
+      "loss": 0.0451,
+      "step": 5074
+    },
+    {
+      "epoch": 2.6418532014575744,
+      "grad_norm": 0.22604762542260493,
+      "learning_rate": 1.8456337760555915e-06,
+      "loss": 0.0447,
+      "step": 5075
+    },
+    {
+      "epoch": 2.642373763664758,
+      "grad_norm": 0.23677583990143997,
+      "learning_rate": 1.840339229082047e-06,
+      "loss": 0.0455,
+      "step": 5076
+    },
+    {
+      "epoch": 2.642894325871942,
+      "grad_norm": 0.2286482680320122,
+      "learning_rate": 1.8350519970071312e-06,
+      "loss": 0.0442,
+      "step": 5077
+    },
+    {
+      "epoch": 2.6434148880791257,
+      "grad_norm": 0.22736556778997194,
+      "learning_rate": 1.8297720815007969e-06,
+      "loss": 0.0453,
+      "step": 5078
+    },
+    {
+      "epoch": 2.643935450286309,
+      "grad_norm": 0.22259072643024455,
+      "learning_rate": 1.824499484230696e-06,
+      "loss": 0.0435,
+      "step": 5079
+    },
+    {
+      "epoch": 2.644456012493493,
+      "grad_norm": 0.2199119162247651,
+      "learning_rate": 1.819234206862147e-06,
+      "loss": 0.0417,
+      "step": 5080
+    },
+    {
+      "epoch": 2.6449765747006766,
+      "grad_norm": 0.24083225482663506,
+      "learning_rate": 1.8139762510581804e-06,
+      "loss": 0.0451,
+      "step": 5081
+    },
+    {
+      "epoch": 2.6454971369078604,
+      "grad_norm": 0.23292508889547486,
+      "learning_rate": 1.8087256184794953e-06,
+      "loss": 0.0463,
+      "step": 5082
+    },
+    {
+      "epoch": 2.646017699115044,
+      "grad_norm": 0.2300747118454459,
+      "learning_rate": 1.8034823107844878e-06,
+      "loss": 0.0445,
+      "step": 5083
+    },
+    {
+      "epoch": 2.646538261322228,
+      "grad_norm": 0.2363535678951791,
+      "learning_rate": 1.798246329629244e-06,
+      "loss": 0.0431,
+      "step": 5084
+    },
+    {
+      "epoch": 2.6470588235294117,
+      "grad_norm": 0.22966034657954063,
+      "learning_rate": 1.793017676667519e-06,
+      "loss": 0.0437,
+      "step": 5085
+    },
+    {
+      "epoch": 2.6475793857365955,
+      "grad_norm": 0.22141333009963216,
+      "learning_rate": 1.7877963535507748e-06,
+      "loss": 0.044,
+      "step": 5086
+    },
+    {
+      "epoch": 2.6480999479437792,
+      "grad_norm": 0.22871432021572957,
+      "learning_rate": 1.7825823619281452e-06,
+      "loss": 0.0427,
+      "step": 5087
+    },
+    {
+      "epoch": 2.648620510150963,
+      "grad_norm": 0.22928264134273937,
+      "learning_rate": 1.7773757034464545e-06,
+      "loss": 0.0458,
+      "step": 5088
+    },
+    {
+      "epoch": 2.6491410723581468,
+      "grad_norm": 0.22874179067788555,
+      "learning_rate": 1.7721763797501984e-06,
+      "loss": 0.0448,
+      "step": 5089
+    },
+    {
+      "epoch": 2.6496616345653305,
+      "grad_norm": 0.23127911334565235,
+      "learning_rate": 1.766984392481577e-06,
+      "loss": 0.0439,
+      "step": 5090
+    },
+    {
+      "epoch": 2.6501821967725143,
+      "grad_norm": 0.22687967610681098,
+      "learning_rate": 1.761799743280454e-06,
+      "loss": 0.0447,
+      "step": 5091
+    },
+    {
+      "epoch": 2.650702758979698,
+      "grad_norm": 0.21699331621599646,
+      "learning_rate": 1.75662243378438e-06,
+      "loss": 0.0419,
+      "step": 5092
+    },
+    {
+      "epoch": 2.651223321186882,
+      "grad_norm": 0.23305173793665826,
+      "learning_rate": 1.751452465628603e-06,
+      "loss": 0.0435,
+      "step": 5093
+    },
+    {
+      "epoch": 2.6517438833940656,
+      "grad_norm": 0.230107160982093,
+      "learning_rate": 1.7462898404460275e-06,
+      "loss": 0.041,
+      "step": 5094
+    },
+    {
+      "epoch": 2.6522644456012494,
+      "grad_norm": 0.2298327980073945,
+      "learning_rate": 1.7411345598672523e-06,
+      "loss": 0.044,
+      "step": 5095
+    },
+    {
+      "epoch": 2.652785007808433,
+      "grad_norm": 0.2309553511497464,
+      "learning_rate": 1.7359866255205609e-06,
+      "loss": 0.0447,
+      "step": 5096
+    },
+    {
+      "epoch": 2.653305570015617,
+      "grad_norm": 0.23228524615867757,
+      "learning_rate": 1.7308460390319025e-06,
+      "loss": 0.0443,
+      "step": 5097
+    },
+    {
+      "epoch": 2.6538261322228007,
+      "grad_norm": 0.2323659311301738,
+      "learning_rate": 1.7257128020249258e-06,
+      "loss": 0.0464,
+      "step": 5098
+    },
+    {
+      "epoch": 2.6543466944299845,
+      "grad_norm": 0.23831555902873636,
+      "learning_rate": 1.7205869161209365e-06,
+      "loss": 0.0465,
+      "step": 5099
+    },
+    {
+      "epoch": 2.6548672566371683,
+      "grad_norm": 0.22428319592649987,
+      "learning_rate": 1.7154683829389284e-06,
+      "loss": 0.0421,
+      "step": 5100
+    },
+    {
+      "epoch": 2.655387818844352,
+      "grad_norm": 0.23529402262221027,
+      "learning_rate": 1.7103572040955696e-06,
+      "loss": 0.0442,
+      "step": 5101
+    },
+    {
+      "epoch": 2.655908381051536,
+      "grad_norm": 0.23011207860973742,
+      "learning_rate": 1.7052533812052157e-06,
+      "loss": 0.0448,
+      "step": 5102
+    },
+    {
+      "epoch": 2.656428943258719,
+      "grad_norm": 0.22964408679947323,
+      "learning_rate": 1.7001569158798914e-06,
+      "loss": 0.0439,
+      "step": 5103
+    },
+    {
+      "epoch": 2.656949505465903,
+      "grad_norm": 0.22620381387212382,
+      "learning_rate": 1.6950678097292893e-06,
+      "loss": 0.0432,
+      "step": 5104
+    },
+    {
+      "epoch": 2.6574700676730867,
+      "grad_norm": 0.23347058886316926,
+      "learning_rate": 1.6899860643607907e-06,
+      "loss": 0.0458,
+      "step": 5105
+    },
+    {
+      "epoch": 2.6579906298802705,
+      "grad_norm": 0.2229703409152849,
+      "learning_rate": 1.6849116813794503e-06,
+      "loss": 0.0436,
+      "step": 5106
+    },
+    {
+      "epoch": 2.6585111920874542,
+      "grad_norm": 0.22996346900113476,
+      "learning_rate": 1.6798446623879915e-06,
+      "loss": 0.0448,
+      "step": 5107
+    },
+    {
+      "epoch": 2.659031754294638,
+      "grad_norm": 0.22933075899722521,
+      "learning_rate": 1.6747850089868178e-06,
+      "loss": 0.0454,
+      "step": 5108
+    },
+    {
+      "epoch": 2.659552316501822,
+      "grad_norm": 0.22374222520186948,
+      "learning_rate": 1.669732722773995e-06,
+      "loss": 0.045,
+      "step": 5109
+    },
+    {
+      "epoch": 2.6600728787090056,
+      "grad_norm": 0.23066068965417216,
+      "learning_rate": 1.6646878053452776e-06,
+      "loss": 0.0431,
+      "step": 5110
+    },
+    {
+      "epoch": 2.6605934409161893,
+      "grad_norm": 0.224015399002532,
+      "learning_rate": 1.659650258294082e-06,
+      "loss": 0.0422,
+      "step": 5111
+    },
+    {
+      "epoch": 2.661114003123373,
+      "grad_norm": 0.23582456384176945,
+      "learning_rate": 1.6546200832115028e-06,
+      "loss": 0.0447,
+      "step": 5112
+    },
+    {
+      "epoch": 2.661634565330557,
+      "grad_norm": 0.2261905982992855,
+      "learning_rate": 1.649597281686302e-06,
+      "loss": 0.0414,
+      "step": 5113
+    },
+    {
+      "epoch": 2.6621551275377406,
+      "grad_norm": 0.23148944353074696,
+      "learning_rate": 1.644581855304911e-06,
+      "loss": 0.0451,
+      "step": 5114
+    },
+    {
+      "epoch": 2.6626756897449244,
+      "grad_norm": 0.22331597314340787,
+      "learning_rate": 1.639573805651437e-06,
+      "loss": 0.0419,
+      "step": 5115
+    },
+    {
+      "epoch": 2.663196251952108,
+      "grad_norm": 0.22397006275035072,
+      "learning_rate": 1.6345731343076626e-06,
+      "loss": 0.0437,
+      "step": 5116
+    },
+    {
+      "epoch": 2.663716814159292,
+      "grad_norm": 0.2324005116144098,
+      "learning_rate": 1.629579842853024e-06,
+      "loss": 0.0457,
+      "step": 5117
+    },
+    {
+      "epoch": 2.6642373763664757,
+      "grad_norm": 0.22967664976854812,
+      "learning_rate": 1.624593932864632e-06,
+      "loss": 0.0445,
+      "step": 5118
+    },
+    {
+      "epoch": 2.6647579385736595,
+      "grad_norm": 0.23099217882260017,
+      "learning_rate": 1.6196154059172742e-06,
+      "loss": 0.0437,
+      "step": 5119
+    },
+    {
+      "epoch": 2.6652785007808433,
+      "grad_norm": 0.23529689019600636,
+      "learning_rate": 1.6146442635834008e-06,
+      "loss": 0.0452,
+      "step": 5120
+    },
+    {
+      "epoch": 2.665799062988027,
+      "grad_norm": 0.24041486589271255,
+      "learning_rate": 1.6096805074331338e-06,
+      "loss": 0.0459,
+      "step": 5121
+    },
+    {
+      "epoch": 2.666319625195211,
+      "grad_norm": 0.2268769288085849,
+      "learning_rate": 1.6047241390342498e-06,
+      "loss": 0.0449,
+      "step": 5122
+    },
+    {
+      "epoch": 2.6668401874023946,
+      "grad_norm": 0.22397701381379848,
+      "learning_rate": 1.599775159952205e-06,
+      "loss": 0.0448,
+      "step": 5123
+    },
+    {
+      "epoch": 2.6673607496095784,
+      "grad_norm": 0.2243649170475879,
+      "learning_rate": 1.5948335717501179e-06,
+      "loss": 0.0422,
+      "step": 5124
+    },
+    {
+      "epoch": 2.667881311816762,
+      "grad_norm": 0.23886065101491008,
+      "learning_rate": 1.5898993759887765e-06,
+      "loss": 0.0452,
+      "step": 5125
+    },
+    {
+      "epoch": 2.668401874023946,
+      "grad_norm": 0.22837358130534788,
+      "learning_rate": 1.5849725742266231e-06,
+      "loss": 0.043,
+      "step": 5126
+    },
+    {
+      "epoch": 2.6689224362311297,
+      "grad_norm": 0.22398559556865122,
+      "learning_rate": 1.5800531680197683e-06,
+      "loss": 0.0443,
+      "step": 5127
+    },
+    {
+      "epoch": 2.6694429984383135,
+      "grad_norm": 0.22326877672343834,
+      "learning_rate": 1.5751411589219945e-06,
+      "loss": 0.0415,
+      "step": 5128
+    },
+    {
+      "epoch": 2.6699635606454972,
+      "grad_norm": 0.24686809704433404,
+      "learning_rate": 1.570236548484741e-06,
+      "loss": 0.0448,
+      "step": 5129
+    },
+    {
+      "epoch": 2.670484122852681,
+      "grad_norm": 0.22496284749102272,
+      "learning_rate": 1.5653393382571158e-06,
+      "loss": 0.0436,
+      "step": 5130
+    },
+    {
+      "epoch": 2.6710046850598648,
+      "grad_norm": 0.2216536409582885,
+      "learning_rate": 1.560449529785879e-06,
+      "loss": 0.0427,
+      "step": 5131
+    },
+    {
+      "epoch": 2.6715252472670485,
+      "grad_norm": 0.23039373528584386,
+      "learning_rate": 1.5555671246154647e-06,
+      "loss": 0.0456,
+      "step": 5132
+    },
+    {
+      "epoch": 2.6720458094742323,
+      "grad_norm": 0.2267257512361847,
+      "learning_rate": 1.5506921242879612e-06,
+      "loss": 0.0438,
+      "step": 5133
+    },
+    {
+      "epoch": 2.672566371681416,
+      "grad_norm": 0.22909660719838928,
+      "learning_rate": 1.5458245303431262e-06,
+      "loss": 0.0442,
+      "step": 5134
+    },
+    {
+      "epoch": 2.6730869338886,
+      "grad_norm": 0.23207465894660437,
+      "learning_rate": 1.5409643443183658e-06,
+      "loss": 0.0452,
+      "step": 5135
+    },
+    {
+      "epoch": 2.6736074960957836,
+      "grad_norm": 0.23464400344667644,
+      "learning_rate": 1.5361115677487548e-06,
+      "loss": 0.0452,
+      "step": 5136
+    },
+    {
+      "epoch": 2.6741280583029674,
+      "grad_norm": 0.23955054324523992,
+      "learning_rate": 1.5312662021670227e-06,
+      "loss": 0.0451,
+      "step": 5137
+    },
+    {
+      "epoch": 2.674648620510151,
+      "grad_norm": 0.22781012646604124,
+      "learning_rate": 1.5264282491035676e-06,
+      "loss": 0.0427,
+      "step": 5138
+    },
+    {
+      "epoch": 2.675169182717335,
+      "grad_norm": 0.22466051885015176,
+      "learning_rate": 1.5215977100864392e-06,
+      "loss": 0.0442,
+      "step": 5139
+    },
+    {
+      "epoch": 2.6756897449245187,
+      "grad_norm": 0.223780756891236,
+      "learning_rate": 1.5167745866413424e-06,
+      "loss": 0.0442,
+      "step": 5140
+    },
+    {
+      "epoch": 2.6762103071317025,
+      "grad_norm": 0.21773883268442826,
+      "learning_rate": 1.5119588802916445e-06,
+      "loss": 0.0426,
+      "step": 5141
+    },
+    {
+      "epoch": 2.6767308693388863,
+      "grad_norm": 0.23066656195993457,
+      "learning_rate": 1.5071505925583735e-06,
+      "loss": 0.044,
+      "step": 5142
+    },
+    {
+      "epoch": 2.6772514315460696,
+      "grad_norm": 0.2226523056001122,
+      "learning_rate": 1.5023497249602086e-06,
+      "loss": 0.0438,
+      "step": 5143
+    },
+    {
+      "epoch": 2.6777719937532534,
+      "grad_norm": 0.2188974751683572,
+      "learning_rate": 1.49755627901349e-06,
+      "loss": 0.0435,
+      "step": 5144
+    },
+    {
+      "epoch": 2.678292555960437,
+      "grad_norm": 0.2229966329377455,
+      "learning_rate": 1.4927702562322037e-06,
+      "loss": 0.0428,
+      "step": 5145
+    },
+    {
+      "epoch": 2.678813118167621,
+      "grad_norm": 0.24359513195681573,
+      "learning_rate": 1.4879916581280045e-06,
+      "loss": 0.0463,
+      "step": 5146
+    },
+    {
+      "epoch": 2.6793336803748047,
+      "grad_norm": 0.23261019586712137,
+      "learning_rate": 1.4832204862101906e-06,
+      "loss": 0.0429,
+      "step": 5147
+    },
+    {
+      "epoch": 2.6798542425819885,
+      "grad_norm": 0.2157186684270539,
+      "learning_rate": 1.4784567419857314e-06,
+      "loss": 0.0425,
+      "step": 5148
+    },
+    {
+      "epoch": 2.6803748047891722,
+      "grad_norm": 0.23508661356548638,
+      "learning_rate": 1.4737004269592236e-06,
+      "loss": 0.045,
+      "step": 5149
+    },
+    {
+      "epoch": 2.680895366996356,
+      "grad_norm": 0.22272454558258475,
+      "learning_rate": 1.468951542632943e-06,
+      "loss": 0.0437,
+      "step": 5150
+    },
+    {
+      "epoch": 2.6814159292035398,
+      "grad_norm": 0.22708962558308424,
+      "learning_rate": 1.4642100905068068e-06,
+      "loss": 0.0451,
+      "step": 5151
+    },
+    {
+      "epoch": 2.6819364914107235,
+      "grad_norm": 0.22570865365739282,
+      "learning_rate": 1.4594760720783863e-06,
+      "loss": 0.0426,
+      "step": 5152
+    },
+    {
+      "epoch": 2.6824570536179073,
+      "grad_norm": 0.22957967759748524,
+      "learning_rate": 1.4547494888429074e-06,
+      "loss": 0.0432,
+      "step": 5153
+    },
+    {
+      "epoch": 2.682977615825091,
+      "grad_norm": 0.22887509990760044,
+      "learning_rate": 1.4500303422932348e-06,
+      "loss": 0.0434,
+      "step": 5154
+    },
+    {
+      "epoch": 2.683498178032275,
+      "grad_norm": 0.23098550242907323,
+      "learning_rate": 1.4453186339199037e-06,
+      "loss": 0.0452,
+      "step": 5155
+    },
+    {
+      "epoch": 2.6840187402394586,
+      "grad_norm": 0.22547319635189877,
+      "learning_rate": 1.4406143652110875e-06,
+      "loss": 0.0423,
+      "step": 5156
+    },
+    {
+      "epoch": 2.6845393024466424,
+      "grad_norm": 0.236534240012221,
+      "learning_rate": 1.4359175376526174e-06,
+      "loss": 0.0469,
+      "step": 5157
+    },
+    {
+      "epoch": 2.685059864653826,
+      "grad_norm": 0.22335001771494034,
+      "learning_rate": 1.431228152727962e-06,
+      "loss": 0.0441,
+      "step": 5158
+    },
+    {
+      "epoch": 2.68558042686101,
+      "grad_norm": 0.22432122506376032,
+      "learning_rate": 1.4265462119182532e-06,
+      "loss": 0.0421,
+      "step": 5159
+    },
+    {
+      "epoch": 2.6861009890681937,
+      "grad_norm": 0.22882569581384093,
+      "learning_rate": 1.4218717167022638e-06,
+      "loss": 0.0432,
+      "step": 5160
+    },
+    {
+      "epoch": 2.6866215512753775,
+      "grad_norm": 0.23103308439566217,
+      "learning_rate": 1.4172046685564212e-06,
+      "loss": 0.0447,
+      "step": 5161
+    },
+    {
+      "epoch": 2.6871421134825613,
+      "grad_norm": 0.22503412132865003,
+      "learning_rate": 1.412545068954796e-06,
+      "loss": 0.0435,
+      "step": 5162
+    },
+    {
+      "epoch": 2.687662675689745,
+      "grad_norm": 0.2228402807168184,
+      "learning_rate": 1.4078929193690998e-06,
+      "loss": 0.0426,
+      "step": 5163
+    },
+    {
+      "epoch": 2.688183237896929,
+      "grad_norm": 0.2287986065956918,
+      "learning_rate": 1.4032482212686993e-06,
+      "loss": 0.0434,
+      "step": 5164
+    },
+    {
+      "epoch": 2.6887038001041126,
+      "grad_norm": 0.2320753787979263,
+      "learning_rate": 1.3986109761206095e-06,
+      "loss": 0.0441,
+      "step": 5165
+    },
+    {
+      "epoch": 2.6892243623112964,
+      "grad_norm": 0.23129496956085294,
+      "learning_rate": 1.3939811853894896e-06,
+      "loss": 0.0456,
+      "step": 5166
+    },
+    {
+      "epoch": 2.6897449245184797,
+      "grad_norm": 0.2345935034216518,
+      "learning_rate": 1.389358850537642e-06,
+      "loss": 0.0448,
+      "step": 5167
+    },
+    {
+      "epoch": 2.6902654867256635,
+      "grad_norm": 0.23234893177636318,
+      "learning_rate": 1.38474397302501e-06,
+      "loss": 0.0442,
+      "step": 5168
+    },
+    {
+      "epoch": 2.6907860489328472,
+      "grad_norm": 0.23634107034074991,
+      "learning_rate": 1.3801365543091916e-06,
+      "loss": 0.0458,
+      "step": 5169
+    },
+    {
+      "epoch": 2.691306611140031,
+      "grad_norm": 0.23696395645252785,
+      "learning_rate": 1.3755365958454254e-06,
+      "loss": 0.0451,
+      "step": 5170
+    },
+    {
+      "epoch": 2.6918271733472148,
+      "grad_norm": 0.22469488655336936,
+      "learning_rate": 1.3709440990865908e-06,
+      "loss": 0.0447,
+      "step": 5171
+    },
+    {
+      "epoch": 2.6923477355543985,
+      "grad_norm": 0.23085933725904473,
+      "learning_rate": 1.366359065483211e-06,
+      "loss": 0.0446,
+      "step": 5172
+    },
+    {
+      "epoch": 2.6928682977615823,
+      "grad_norm": 0.23659193340498502,
+      "learning_rate": 1.3617814964834523e-06,
+      "loss": 0.0455,
+      "step": 5173
+    },
+    {
+      "epoch": 2.693388859968766,
+      "grad_norm": 0.2189403460488524,
+      "learning_rate": 1.3572113935331226e-06,
+      "loss": 0.044,
+      "step": 5174
+    },
+    {
+      "epoch": 2.69390942217595,
+      "grad_norm": 0.2247914833766047,
+      "learning_rate": 1.3526487580756752e-06,
+      "loss": 0.0441,
+      "step": 5175
+    },
+    {
+      "epoch": 2.6944299843831336,
+      "grad_norm": 0.21893245384497728,
+      "learning_rate": 1.3480935915522075e-06,
+      "loss": 0.043,
+      "step": 5176
+    },
+    {
+      "epoch": 2.6949505465903174,
+      "grad_norm": 0.23444838887341896,
+      "learning_rate": 1.3435458954014463e-06,
+      "loss": 0.0453,
+      "step": 5177
+    },
+    {
+      "epoch": 2.695471108797501,
+      "grad_norm": 0.23107549495251808,
+      "learning_rate": 1.3390056710597649e-06,
+      "loss": 0.0442,
+      "step": 5178
+    },
+    {
+      "epoch": 2.695991671004685,
+      "grad_norm": 0.22611578165673227,
+      "learning_rate": 1.3344729199611827e-06,
+      "loss": 0.0458,
+      "step": 5179
+    },
+    {
+      "epoch": 2.6965122332118687,
+      "grad_norm": 0.22290496338439902,
+      "learning_rate": 1.3299476435373548e-06,
+      "loss": 0.0429,
+      "step": 5180
+    },
+    {
+      "epoch": 2.6970327954190525,
+      "grad_norm": 0.21442995257650893,
+      "learning_rate": 1.3254298432175682e-06,
+      "loss": 0.0424,
+      "step": 5181
+    },
+    {
+      "epoch": 2.6975533576262363,
+      "grad_norm": 0.21876017392018013,
+      "learning_rate": 1.320919520428754e-06,
+      "loss": 0.0432,
+      "step": 5182
+    },
+    {
+      "epoch": 2.69807391983342,
+      "grad_norm": 0.22822179337752255,
+      "learning_rate": 1.3164166765954861e-06,
+      "loss": 0.0454,
+      "step": 5183
+    },
+    {
+      "epoch": 2.698594482040604,
+      "grad_norm": 0.23403568093314686,
+      "learning_rate": 1.3119213131399688e-06,
+      "loss": 0.0468,
+      "step": 5184
+    },
+    {
+      "epoch": 2.6991150442477876,
+      "grad_norm": 0.22524839908609418,
+      "learning_rate": 1.3074334314820551e-06,
+      "loss": 0.0447,
+      "step": 5185
+    },
+    {
+      "epoch": 2.6996356064549714,
+      "grad_norm": 0.22270469667567147,
+      "learning_rate": 1.3029530330392143e-06,
+      "loss": 0.0442,
+      "step": 5186
+    },
+    {
+      "epoch": 2.700156168662155,
+      "grad_norm": 0.23012017240768629,
+      "learning_rate": 1.2984801192265749e-06,
+      "loss": 0.0439,
+      "step": 5187
+    },
+    {
+      "epoch": 2.700676730869339,
+      "grad_norm": 0.21642941461801396,
+      "learning_rate": 1.2940146914568852e-06,
+      "loss": 0.0438,
+      "step": 5188
+    },
+    {
+      "epoch": 2.7011972930765227,
+      "grad_norm": 0.22328673189570775,
+      "learning_rate": 1.2895567511405414e-06,
+      "loss": 0.0439,
+      "step": 5189
+    },
+    {
+      "epoch": 2.7017178552837064,
+      "grad_norm": 0.23823980764075436,
+      "learning_rate": 1.285106299685565e-06,
+      "loss": 0.0457,
+      "step": 5190
+    },
+    {
+      "epoch": 2.70223841749089,
+      "grad_norm": 0.22391174202738537,
+      "learning_rate": 1.280663338497609e-06,
+      "loss": 0.0448,
+      "step": 5191
+    },
+    {
+      "epoch": 2.702758979698074,
+      "grad_norm": 0.22413624516419742,
+      "learning_rate": 1.276227868979976e-06,
+      "loss": 0.0422,
+      "step": 5192
+    },
+    {
+      "epoch": 2.7032795419052578,
+      "grad_norm": 0.219580469635659,
+      "learning_rate": 1.2717998925335927e-06,
+      "loss": 0.0425,
+      "step": 5193
+    },
+    {
+      "epoch": 2.7038001041124415,
+      "grad_norm": 0.23393120071768486,
+      "learning_rate": 1.267379410557018e-06,
+      "loss": 0.0463,
+      "step": 5194
+    },
+    {
+      "epoch": 2.7043206663196253,
+      "grad_norm": 0.22651080633930457,
+      "learning_rate": 1.2629664244464463e-06,
+      "loss": 0.0446,
+      "step": 5195
+    },
+    {
+      "epoch": 2.704841228526809,
+      "grad_norm": 0.2345516458739699,
+      "learning_rate": 1.258560935595704e-06,
+      "loss": 0.0457,
+      "step": 5196
+    },
+    {
+      "epoch": 2.705361790733993,
+      "grad_norm": 0.21826699278019435,
+      "learning_rate": 1.2541629453962479e-06,
+      "loss": 0.0427,
+      "step": 5197
+    },
+    {
+      "epoch": 2.7058823529411766,
+      "grad_norm": 0.21948973250772175,
+      "learning_rate": 1.2497724552371747e-06,
+      "loss": 0.0428,
+      "step": 5198
+    },
+    {
+      "epoch": 2.7064029151483604,
+      "grad_norm": 0.23291910481811462,
+      "learning_rate": 1.2453894665052008e-06,
+      "loss": 0.0445,
+      "step": 5199
+    },
+    {
+      "epoch": 2.706923477355544,
+      "grad_norm": 0.21755848524659785,
+      "learning_rate": 1.2410139805846738e-06,
+      "loss": 0.0423,
+      "step": 5200
+    },
+    {
+      "epoch": 2.707444039562728,
+      "grad_norm": 0.2319847511448208,
+      "learning_rate": 1.2366459988575774e-06,
+      "loss": 0.0438,
+      "step": 5201
+    },
+    {
+      "epoch": 2.7079646017699117,
+      "grad_norm": 0.2303536354287684,
+      "learning_rate": 1.2322855227035301e-06,
+      "loss": 0.0439,
+      "step": 5202
+    },
+    {
+      "epoch": 2.7084851639770955,
+      "grad_norm": 0.23065828693762852,
+      "learning_rate": 1.2279325534997693e-06,
+      "loss": 0.0446,
+      "step": 5203
+    },
+    {
+      "epoch": 2.7090057261842793,
+      "grad_norm": 0.24141647025860344,
+      "learning_rate": 1.2235870926211619e-06,
+      "loss": 0.0457,
+      "step": 5204
+    },
+    {
+      "epoch": 2.709526288391463,
+      "grad_norm": 0.2350329888138472,
+      "learning_rate": 1.2192491414402096e-06,
+      "loss": 0.044,
+      "step": 5205
+    },
+    {
+      "epoch": 2.710046850598647,
+      "grad_norm": 0.231757509837116,
+      "learning_rate": 1.2149187013270392e-06,
+      "loss": 0.045,
+      "step": 5206
+    },
+    {
+      "epoch": 2.71056741280583,
+      "grad_norm": 0.2317717587228352,
+      "learning_rate": 1.2105957736494089e-06,
+      "loss": 0.0435,
+      "step": 5207
+    },
+    {
+      "epoch": 2.711087975013014,
+      "grad_norm": 0.22055288340775916,
+      "learning_rate": 1.2062803597726963e-06,
+      "loss": 0.0411,
+      "step": 5208
+    },
+    {
+      "epoch": 2.7116085372201977,
+      "grad_norm": 0.2268087132163455,
+      "learning_rate": 1.2019724610599081e-06,
+      "loss": 0.0436,
+      "step": 5209
+    },
+    {
+      "epoch": 2.7121290994273815,
+      "grad_norm": 0.2303811370184185,
+      "learning_rate": 1.197672078871681e-06,
+      "loss": 0.044,
+      "step": 5210
+    },
+    {
+      "epoch": 2.7126496616345652,
+      "grad_norm": 0.23071049199969476,
+      "learning_rate": 1.193379214566276e-06,
+      "loss": 0.0446,
+      "step": 5211
+    },
+    {
+      "epoch": 2.713170223841749,
+      "grad_norm": 0.22811091103571507,
+      "learning_rate": 1.1890938694995829e-06,
+      "loss": 0.0432,
+      "step": 5212
+    },
+    {
+      "epoch": 2.7136907860489328,
+      "grad_norm": 0.2379584189281919,
+      "learning_rate": 1.1848160450251083e-06,
+      "loss": 0.0463,
+      "step": 5213
+    },
+    {
+      "epoch": 2.7142113482561165,
+      "grad_norm": 0.22764918185318186,
+      "learning_rate": 1.180545742493988e-06,
+      "loss": 0.0442,
+      "step": 5214
+    },
+    {
+      "epoch": 2.7147319104633003,
+      "grad_norm": 0.2254826094616572,
+      "learning_rate": 1.1762829632549849e-06,
+      "loss": 0.044,
+      "step": 5215
+    },
+    {
+      "epoch": 2.715252472670484,
+      "grad_norm": 0.22681141939496943,
+      "learning_rate": 1.1720277086544857e-06,
+      "loss": 0.0433,
+      "step": 5216
+    },
+    {
+      "epoch": 2.715773034877668,
+      "grad_norm": 0.21876872488915144,
+      "learning_rate": 1.1677799800364958e-06,
+      "loss": 0.0429,
+      "step": 5217
+    },
+    {
+      "epoch": 2.7162935970848516,
+      "grad_norm": 0.22378659997771677,
+      "learning_rate": 1.1635397787426366e-06,
+      "loss": 0.0438,
+      "step": 5218
+    },
+    {
+      "epoch": 2.7168141592920354,
+      "grad_norm": 0.2265493115245886,
+      "learning_rate": 1.159307106112173e-06,
+      "loss": 0.0442,
+      "step": 5219
+    },
+    {
+      "epoch": 2.717334721499219,
+      "grad_norm": 0.2298762644808635,
+      "learning_rate": 1.1550819634819743e-06,
+      "loss": 0.0443,
+      "step": 5220
+    },
+    {
+      "epoch": 2.717855283706403,
+      "grad_norm": 0.22228468158301753,
+      "learning_rate": 1.1508643521865397e-06,
+      "loss": 0.0429,
+      "step": 5221
+    },
+    {
+      "epoch": 2.7183758459135867,
+      "grad_norm": 0.23151020928005936,
+      "learning_rate": 1.1466542735579844e-06,
+      "loss": 0.0427,
+      "step": 5222
+    },
+    {
+      "epoch": 2.7188964081207705,
+      "grad_norm": 0.2214334447059201,
+      "learning_rate": 1.14245172892605e-06,
+      "loss": 0.0433,
+      "step": 5223
+    },
+    {
+      "epoch": 2.7194169703279543,
+      "grad_norm": 0.2283368759330757,
+      "learning_rate": 1.1382567196180916e-06,
+      "loss": 0.0437,
+      "step": 5224
+    },
+    {
+      "epoch": 2.719937532535138,
+      "grad_norm": 0.21320513712505007,
+      "learning_rate": 1.1340692469590964e-06,
+      "loss": 0.0422,
+      "step": 5225
+    },
+    {
+      "epoch": 2.720458094742322,
+      "grad_norm": 0.21709414283045186,
+      "learning_rate": 1.1298893122716563e-06,
+      "loss": 0.0438,
+      "step": 5226
+    },
+    {
+      "epoch": 2.7209786569495056,
+      "grad_norm": 0.2274072043634905,
+      "learning_rate": 1.125716916875988e-06,
+      "loss": 0.0433,
+      "step": 5227
+    },
+    {
+      "epoch": 2.7214992191566894,
+      "grad_norm": 0.2275711914621425,
+      "learning_rate": 1.1215520620899311e-06,
+      "loss": 0.0439,
+      "step": 5228
+    },
+    {
+      "epoch": 2.722019781363873,
+      "grad_norm": 0.22221898121836894,
+      "learning_rate": 1.1173947492289395e-06,
+      "loss": 0.0428,
+      "step": 5229
+    },
+    {
+      "epoch": 2.722540343571057,
+      "grad_norm": 0.21297199594248953,
+      "learning_rate": 1.1132449796060872e-06,
+      "loss": 0.0416,
+      "step": 5230
+    },
+    {
+      "epoch": 2.7230609057782402,
+      "grad_norm": 0.2255176774021255,
+      "learning_rate": 1.1091027545320654e-06,
+      "loss": 0.0426,
+      "step": 5231
+    },
+    {
+      "epoch": 2.723581467985424,
+      "grad_norm": 0.22910733482663895,
+      "learning_rate": 1.1049680753151798e-06,
+      "loss": 0.0449,
+      "step": 5232
+    },
+    {
+      "epoch": 2.7241020301926078,
+      "grad_norm": 0.22540358774719207,
+      "learning_rate": 1.1008409432613525e-06,
+      "loss": 0.0423,
+      "step": 5233
+    },
+    {
+      "epoch": 2.7246225923997915,
+      "grad_norm": 0.22668282262981723,
+      "learning_rate": 1.0967213596741327e-06,
+      "loss": 0.0417,
+      "step": 5234
+    },
+    {
+      "epoch": 2.7251431546069753,
+      "grad_norm": 0.22941299415233102,
+      "learning_rate": 1.0926093258546655e-06,
+      "loss": 0.0445,
+      "step": 5235
+    },
+    {
+      "epoch": 2.725663716814159,
+      "grad_norm": 0.2460091226274877,
+      "learning_rate": 1.0885048431017313e-06,
+      "loss": 0.0474,
+      "step": 5236
+    },
+    {
+      "epoch": 2.726184279021343,
+      "grad_norm": 0.22696800191590433,
+      "learning_rate": 1.0844079127117074e-06,
+      "loss": 0.0441,
+      "step": 5237
+    },
+    {
+      "epoch": 2.7267048412285266,
+      "grad_norm": 0.2279206422911676,
+      "learning_rate": 1.0803185359786028e-06,
+      "loss": 0.0435,
+      "step": 5238
+    },
+    {
+      "epoch": 2.7272254034357104,
+      "grad_norm": 0.2206185588040236,
+      "learning_rate": 1.0762367141940287e-06,
+      "loss": 0.0439,
+      "step": 5239
+    },
+    {
+      "epoch": 2.727745965642894,
+      "grad_norm": 0.22844650094437752,
+      "learning_rate": 1.0721624486472209e-06,
+      "loss": 0.0436,
+      "step": 5240
+    },
+    {
+      "epoch": 2.728266527850078,
+      "grad_norm": 0.22100974494576425,
+      "learning_rate": 1.0680957406250135e-06,
+      "loss": 0.0436,
+      "step": 5241
+    },
+    {
+      "epoch": 2.7287870900572617,
+      "grad_norm": 0.22075747241403315,
+      "learning_rate": 1.0640365914118682e-06,
+      "loss": 0.0424,
+      "step": 5242
+    },
+    {
+      "epoch": 2.7293076522644455,
+      "grad_norm": 0.230695539655403,
+      "learning_rate": 1.0599850022898539e-06,
+      "loss": 0.043,
+      "step": 5243
+    },
+    {
+      "epoch": 2.7298282144716293,
+      "grad_norm": 0.23168258489285376,
+      "learning_rate": 1.055940974538641e-06,
+      "loss": 0.0431,
+      "step": 5244
+    },
+    {
+      "epoch": 2.730348776678813,
+      "grad_norm": 0.22697802722231222,
+      "learning_rate": 1.0519045094355363e-06,
+      "loss": 0.0432,
+      "step": 5245
+    },
+    {
+      "epoch": 2.730869338885997,
+      "grad_norm": 0.2308978573693367,
+      "learning_rate": 1.0478756082554304e-06,
+      "loss": 0.0441,
+      "step": 5246
+    },
+    {
+      "epoch": 2.7313899010931806,
+      "grad_norm": 0.2304130228518842,
+      "learning_rate": 1.0438542722708445e-06,
+      "loss": 0.0427,
+      "step": 5247
+    },
+    {
+      "epoch": 2.7319104633003644,
+      "grad_norm": 0.2282468794565119,
+      "learning_rate": 1.0398405027519016e-06,
+      "loss": 0.0424,
+      "step": 5248
+    },
+    {
+      "epoch": 2.732431025507548,
+      "grad_norm": 0.21865000074849475,
+      "learning_rate": 1.0358343009663428e-06,
+      "loss": 0.0422,
+      "step": 5249
+    },
+    {
+      "epoch": 2.732951587714732,
+      "grad_norm": 0.21986811718537888,
+      "learning_rate": 1.0318356681795039e-06,
+      "loss": 0.0409,
+      "step": 5250
+    },
+    {
+      "epoch": 2.7334721499219157,
+      "grad_norm": 0.22870095194546886,
+      "learning_rate": 1.0278446056543406e-06,
+      "loss": 0.0434,
+      "step": 5251
+    },
+    {
+      "epoch": 2.7339927121290994,
+      "grad_norm": 0.22636557592510134,
+      "learning_rate": 1.0238611146514253e-06,
+      "loss": 0.0429,
+      "step": 5252
+    },
+    {
+      "epoch": 2.734513274336283,
+      "grad_norm": 0.21850314176206617,
+      "learning_rate": 1.0198851964289185e-06,
+      "loss": 0.0431,
+      "step": 5253
+    },
+    {
+      "epoch": 2.735033836543467,
+      "grad_norm": 0.2186586118194875,
+      "learning_rate": 1.01591685224261e-06,
+      "loss": 0.0418,
+      "step": 5254
+    },
+    {
+      "epoch": 2.7355543987506508,
+      "grad_norm": 0.24147015358727544,
+      "learning_rate": 1.0119560833458775e-06,
+      "loss": 0.0444,
+      "step": 5255
+    },
+    {
+      "epoch": 2.7360749609578345,
+      "grad_norm": 0.2315952637276368,
+      "learning_rate": 1.0080028909897233e-06,
+      "loss": 0.0463,
+      "step": 5256
+    },
+    {
+      "epoch": 2.7365955231650183,
+      "grad_norm": 0.2200895690176689,
+      "learning_rate": 1.0040572764227458e-06,
+      "loss": 0.0432,
+      "step": 5257
+    },
+    {
+      "epoch": 2.737116085372202,
+      "grad_norm": 0.22812841777700557,
+      "learning_rate": 1.0001192408911592e-06,
+      "loss": 0.0442,
+      "step": 5258
+    },
+    {
+      "epoch": 2.737636647579386,
+      "grad_norm": 0.2224052677158258,
+      "learning_rate": 9.961887856387714e-07,
+      "loss": 0.0426,
+      "step": 5259
+    },
+    {
+      "epoch": 2.7381572097865696,
+      "grad_norm": 0.22285780659280513,
+      "learning_rate": 9.92265911907006e-07,
+      "loss": 0.0429,
+      "step": 5260
+    },
+    {
+      "epoch": 2.7386777719937534,
+      "grad_norm": 0.22292964049968192,
+      "learning_rate": 9.883506209348914e-07,
+      "loss": 0.0415,
+      "step": 5261
+    },
+    {
+      "epoch": 2.739198334200937,
+      "grad_norm": 0.2253363805591684,
+      "learning_rate": 9.84442913959055e-07,
+      "loss": 0.0445,
+      "step": 5262
+    },
+    {
+      "epoch": 2.739718896408121,
+      "grad_norm": 0.22914905603731295,
+      "learning_rate": 9.805427922137373e-07,
+      "loss": 0.0436,
+      "step": 5263
+    },
+    {
+      "epoch": 2.7402394586153047,
+      "grad_norm": 0.22423612517145042,
+      "learning_rate": 9.766502569307722e-07,
+      "loss": 0.0443,
+      "step": 5264
+    },
+    {
+      "epoch": 2.7407600208224885,
+      "grad_norm": 0.22723159502834736,
+      "learning_rate": 9.727653093396044e-07,
+      "loss": 0.0447,
+      "step": 5265
+    },
+    {
+      "epoch": 2.7412805830296723,
+      "grad_norm": 0.22822872201192684,
+      "learning_rate": 9.688879506672854e-07,
+      "loss": 0.043,
+      "step": 5266
+    },
+    {
+      "epoch": 2.741801145236856,
+      "grad_norm": 0.21393631920738038,
+      "learning_rate": 9.650181821384636e-07,
+      "loss": 0.0412,
+      "step": 5267
+    },
+    {
+      "epoch": 2.74232170744404,
+      "grad_norm": 0.22094569188225746,
+      "learning_rate": 9.611560049753915e-07,
+      "loss": 0.0434,
+      "step": 5268
+    },
+    {
+      "epoch": 2.7428422696512236,
+      "grad_norm": 0.22305606790036275,
+      "learning_rate": 9.573014203979242e-07,
+      "loss": 0.0425,
+      "step": 5269
+    },
+    {
+      "epoch": 2.7433628318584073,
+      "grad_norm": 0.22902027508759243,
+      "learning_rate": 9.534544296235181e-07,
+      "loss": 0.0428,
+      "step": 5270
+    },
+    {
+      "epoch": 2.7438833940655907,
+      "grad_norm": 0.23242670385926198,
+      "learning_rate": 9.49615033867235e-07,
+      "loss": 0.0446,
+      "step": 5271
+    },
+    {
+      "epoch": 2.7444039562727744,
+      "grad_norm": 0.2347141380798654,
+      "learning_rate": 9.45783234341735e-07,
+      "loss": 0.0425,
+      "step": 5272
+    },
+    {
+      "epoch": 2.744924518479958,
+      "grad_norm": 0.22332579684782305,
+      "learning_rate": 9.419590322572725e-07,
+      "loss": 0.0443,
+      "step": 5273
+    },
+    {
+      "epoch": 2.745445080687142,
+      "grad_norm": 0.22189445785006218,
+      "learning_rate": 9.381424288217117e-07,
+      "loss": 0.0423,
+      "step": 5274
+    },
+    {
+      "epoch": 2.7459656428943258,
+      "grad_norm": 0.2261260353066565,
+      "learning_rate": 9.343334252405133e-07,
+      "loss": 0.0434,
+      "step": 5275
+    },
+    {
+      "epoch": 2.7464862051015095,
+      "grad_norm": 0.22674984875383666,
+      "learning_rate": 9.30532022716743e-07,
+      "loss": 0.0434,
+      "step": 5276
+    },
+    {
+      "epoch": 2.7470067673086933,
+      "grad_norm": 0.22247669966218622,
+      "learning_rate": 9.26738222451054e-07,
+      "loss": 0.0431,
+      "step": 5277
+    },
+    {
+      "epoch": 2.747527329515877,
+      "grad_norm": 0.22557924637748528,
+      "learning_rate": 9.229520256417073e-07,
+      "loss": 0.0456,
+      "step": 5278
+    },
+    {
+      "epoch": 2.748047891723061,
+      "grad_norm": 0.22992144412583332,
+      "learning_rate": 9.191734334845603e-07,
+      "loss": 0.0452,
+      "step": 5279
+    },
+    {
+      "epoch": 2.7485684539302446,
+      "grad_norm": 0.22300684919494487,
+      "learning_rate": 9.154024471730721e-07,
+      "loss": 0.0438,
+      "step": 5280
+    },
+    {
+      "epoch": 2.7490890161374284,
+      "grad_norm": 0.22245433446016094,
+      "learning_rate": 9.1163906789829e-07,
+      "loss": 0.042,
+      "step": 5281
+    },
+    {
+      "epoch": 2.749609578344612,
+      "grad_norm": 0.22206770415357324,
+      "learning_rate": 9.078832968488632e-07,
+      "loss": 0.0434,
+      "step": 5282
+    },
+    {
+      "epoch": 2.750130140551796,
+      "grad_norm": 0.23376405716262028,
+      "learning_rate": 9.041351352110427e-07,
+      "loss": 0.0431,
+      "step": 5283
+    },
+    {
+      "epoch": 2.7506507027589797,
+      "grad_norm": 0.22966704496277124,
+      "learning_rate": 9.003945841686707e-07,
+      "loss": 0.0434,
+      "step": 5284
+    },
+    {
+      "epoch": 2.7511712649661635,
+      "grad_norm": 0.22838593078199046,
+      "learning_rate": 8.966616449031906e-07,
+      "loss": 0.0422,
+      "step": 5285
+    },
+    {
+      "epoch": 2.7516918271733473,
+      "grad_norm": 0.22176204273902209,
+      "learning_rate": 8.929363185936346e-07,
+      "loss": 0.0443,
+      "step": 5286
+    },
+    {
+      "epoch": 2.752212389380531,
+      "grad_norm": 0.22581365238705586,
+      "learning_rate": 8.892186064166335e-07,
+      "loss": 0.0441,
+      "step": 5287
+    },
+    {
+      "epoch": 2.752732951587715,
+      "grad_norm": 0.2244870202646647,
+      "learning_rate": 8.855085095464149e-07,
+      "loss": 0.0429,
+      "step": 5288
+    },
+    {
+      "epoch": 2.7532535137948986,
+      "grad_norm": 0.22803770453270514,
+      "learning_rate": 8.818060291548053e-07,
+      "loss": 0.0436,
+      "step": 5289
+    },
+    {
+      "epoch": 2.7537740760020823,
+      "grad_norm": 0.2250742592276933,
+      "learning_rate": 8.781111664112162e-07,
+      "loss": 0.042,
+      "step": 5290
+    },
+    {
+      "epoch": 2.754294638209266,
+      "grad_norm": 0.2253420328172488,
+      "learning_rate": 8.744239224826534e-07,
+      "loss": 0.0419,
+      "step": 5291
+    },
+    {
+      "epoch": 2.75481520041645,
+      "grad_norm": 0.2264536838481127,
+      "learning_rate": 8.707442985337239e-07,
+      "loss": 0.0441,
+      "step": 5292
+    },
+    {
+      "epoch": 2.7553357626236337,
+      "grad_norm": 0.2202483024468351,
+      "learning_rate": 8.670722957266231e-07,
+      "loss": 0.0443,
+      "step": 5293
+    },
+    {
+      "epoch": 2.7558563248308174,
+      "grad_norm": 0.2303490443562307,
+      "learning_rate": 8.634079152211427e-07,
+      "loss": 0.0429,
+      "step": 5294
+    },
+    {
+      "epoch": 2.7563768870380008,
+      "grad_norm": 0.22221701362424096,
+      "learning_rate": 8.597511581746626e-07,
+      "loss": 0.0433,
+      "step": 5295
+    },
+    {
+      "epoch": 2.7568974492451845,
+      "grad_norm": 0.23951455762675466,
+      "learning_rate": 8.56102025742156e-07,
+      "loss": 0.0447,
+      "step": 5296
+    },
+    {
+      "epoch": 2.7574180114523683,
+      "grad_norm": 0.2374125127215411,
+      "learning_rate": 8.524605190761897e-07,
+      "loss": 0.0451,
+      "step": 5297
+    },
+    {
+      "epoch": 2.757938573659552,
+      "grad_norm": 0.23192693160940323,
+      "learning_rate": 8.488266393269245e-07,
+      "loss": 0.0432,
+      "step": 5298
+    },
+    {
+      "epoch": 2.758459135866736,
+      "grad_norm": 0.2342718094829394,
+      "learning_rate": 8.452003876421033e-07,
+      "loss": 0.0437,
+      "step": 5299
+    },
+    {
+      "epoch": 2.7589796980739196,
+      "grad_norm": 0.2250652941295221,
+      "learning_rate": 8.415817651670654e-07,
+      "loss": 0.0437,
+      "step": 5300
+    },
+    {
+      "epoch": 2.7595002602811034,
+      "grad_norm": 0.22979466231086934,
+      "learning_rate": 8.379707730447439e-07,
+      "loss": 0.0437,
+      "step": 5301
+    },
+    {
+      "epoch": 2.760020822488287,
+      "grad_norm": 0.23361883586709467,
+      "learning_rate": 8.343674124156542e-07,
+      "loss": 0.0443,
+      "step": 5302
+    },
+    {
+      "epoch": 2.760541384695471,
+      "grad_norm": 0.22631184004097923,
+      "learning_rate": 8.307716844179081e-07,
+      "loss": 0.0416,
+      "step": 5303
+    },
+    {
+      "epoch": 2.7610619469026547,
+      "grad_norm": 0.23184232037276983,
+      "learning_rate": 8.271835901872055e-07,
+      "loss": 0.0439,
+      "step": 5304
+    },
+    {
+      "epoch": 2.7615825091098385,
+      "grad_norm": 0.23202421193452336,
+      "learning_rate": 8.236031308568287e-07,
+      "loss": 0.0431,
+      "step": 5305
+    },
+    {
+      "epoch": 2.7621030713170223,
+      "grad_norm": 0.22241507265189012,
+      "learning_rate": 8.200303075576565e-07,
+      "loss": 0.0425,
+      "step": 5306
+    },
+    {
+      "epoch": 2.762623633524206,
+      "grad_norm": 0.22697877013798187,
+      "learning_rate": 8.164651214181556e-07,
+      "loss": 0.044,
+      "step": 5307
+    },
+    {
+      "epoch": 2.76314419573139,
+      "grad_norm": 0.2247465473110427,
+      "learning_rate": 8.129075735643698e-07,
+      "loss": 0.0429,
+      "step": 5308
+    },
+    {
+      "epoch": 2.7636647579385736,
+      "grad_norm": 0.22274117285283673,
+      "learning_rate": 8.093576651199447e-07,
+      "loss": 0.0425,
+      "step": 5309
+    },
+    {
+      "epoch": 2.7641853201457574,
+      "grad_norm": 0.22736433771242023,
+      "learning_rate": 8.058153972061027e-07,
+      "loss": 0.0415,
+      "step": 5310
+    },
+    {
+      "epoch": 2.764705882352941,
+      "grad_norm": 0.22964935642008533,
+      "learning_rate": 8.022807709416575e-07,
+      "loss": 0.0433,
+      "step": 5311
+    },
+    {
+      "epoch": 2.765226444560125,
+      "grad_norm": 0.23653451017732308,
+      "learning_rate": 7.987537874430101e-07,
+      "loss": 0.0447,
+      "step": 5312
+    },
+    {
+      "epoch": 2.7657470067673087,
+      "grad_norm": 0.22232202569716866,
+      "learning_rate": 7.952344478241503e-07,
+      "loss": 0.042,
+      "step": 5313
+    },
+    {
+      "epoch": 2.7662675689744924,
+      "grad_norm": 0.22865924001277593,
+      "learning_rate": 7.917227531966387e-07,
+      "loss": 0.0443,
+      "step": 5314
+    },
+    {
+      "epoch": 2.766788131181676,
+      "grad_norm": 0.2199762097546008,
+      "learning_rate": 7.88218704669641e-07,
+      "loss": 0.0416,
+      "step": 5315
+    },
+    {
+      "epoch": 2.76730869338886,
+      "grad_norm": 0.22907960763658383,
+      "learning_rate": 7.847223033498968e-07,
+      "loss": 0.0443,
+      "step": 5316
+    },
+    {
+      "epoch": 2.7678292555960438,
+      "grad_norm": 0.2224849881577553,
+      "learning_rate": 7.812335503417284e-07,
+      "loss": 0.0434,
+      "step": 5317
+    },
+    {
+      "epoch": 2.7683498178032275,
+      "grad_norm": 0.2229443345455532,
+      "learning_rate": 7.777524467470515e-07,
+      "loss": 0.0423,
+      "step": 5318
+    },
+    {
+      "epoch": 2.7688703800104113,
+      "grad_norm": 0.22816646258582196,
+      "learning_rate": 7.742789936653561e-07,
+      "loss": 0.0429,
+      "step": 5319
+    },
+    {
+      "epoch": 2.769390942217595,
+      "grad_norm": 0.22839763781886158,
+      "learning_rate": 7.708131921937229e-07,
+      "loss": 0.0441,
+      "step": 5320
+    },
+    {
+      "epoch": 2.769911504424779,
+      "grad_norm": 0.22727070637222285,
+      "learning_rate": 7.673550434268123e-07,
+      "loss": 0.045,
+      "step": 5321
+    },
+    {
+      "epoch": 2.7704320666319626,
+      "grad_norm": 0.22633561873190844,
+      "learning_rate": 7.639045484568702e-07,
+      "loss": 0.045,
+      "step": 5322
+    },
+    {
+      "epoch": 2.7709526288391464,
+      "grad_norm": 0.22765311723057183,
+      "learning_rate": 7.60461708373722e-07,
+      "loss": 0.044,
+      "step": 5323
+    },
+    {
+      "epoch": 2.77147319104633,
+      "grad_norm": 0.2272959845769333,
+      "learning_rate": 7.570265242647784e-07,
+      "loss": 0.0432,
+      "step": 5324
+    },
+    {
+      "epoch": 2.771993753253514,
+      "grad_norm": 0.2181349991575582,
+      "learning_rate": 7.535989972150298e-07,
+      "loss": 0.0426,
+      "step": 5325
+    },
+    {
+      "epoch": 2.7725143154606977,
+      "grad_norm": 0.21996140541009873,
+      "learning_rate": 7.501791283070436e-07,
+      "loss": 0.0431,
+      "step": 5326
+    },
+    {
+      "epoch": 2.7730348776678815,
+      "grad_norm": 0.22240120810935823,
+      "learning_rate": 7.467669186209836e-07,
+      "loss": 0.0426,
+      "step": 5327
+    },
+    {
+      "epoch": 2.7735554398750653,
+      "grad_norm": 0.23526222986344317,
+      "learning_rate": 7.433623692345765e-07,
+      "loss": 0.0428,
+      "step": 5328
+    },
+    {
+      "epoch": 2.774076002082249,
+      "grad_norm": 0.23391764152544686,
+      "learning_rate": 7.399654812231399e-07,
+      "loss": 0.0441,
+      "step": 5329
+    },
+    {
+      "epoch": 2.774596564289433,
+      "grad_norm": 0.23704308426871976,
+      "learning_rate": 7.365762556595685e-07,
+      "loss": 0.0442,
+      "step": 5330
+    },
+    {
+      "epoch": 2.7751171264966166,
+      "grad_norm": 0.22376362085484436,
+      "learning_rate": 7.331946936143392e-07,
+      "loss": 0.0456,
+      "step": 5331
+    },
+    {
+      "epoch": 2.7756376887038003,
+      "grad_norm": 0.2265634568261286,
+      "learning_rate": 7.298207961555031e-07,
+      "loss": 0.0443,
+      "step": 5332
+    },
+    {
+      "epoch": 2.776158250910984,
+      "grad_norm": 0.23042926713303066,
+      "learning_rate": 7.264545643486997e-07,
+      "loss": 0.0429,
+      "step": 5333
+    },
+    {
+      "epoch": 2.776678813118168,
+      "grad_norm": 0.2213224298794227,
+      "learning_rate": 7.230959992571368e-07,
+      "loss": 0.0421,
+      "step": 5334
+    },
+    {
+      "epoch": 2.777199375325351,
+      "grad_norm": 0.2210558855748437,
+      "learning_rate": 7.197451019416073e-07,
+      "loss": 0.0425,
+      "step": 5335
+    },
+    {
+      "epoch": 2.777719937532535,
+      "grad_norm": 0.2258634646023793,
+      "learning_rate": 7.164018734604816e-07,
+      "loss": 0.0437,
+      "step": 5336
+    },
+    {
+      "epoch": 2.7782404997397188,
+      "grad_norm": 0.23229199016420934,
+      "learning_rate": 7.130663148697037e-07,
+      "loss": 0.0464,
+      "step": 5337
+    },
+    {
+      "epoch": 2.7787610619469025,
+      "grad_norm": 0.22326405991841328,
+      "learning_rate": 7.097384272228003e-07,
+      "loss": 0.043,
+      "step": 5338
+    },
+    {
+      "epoch": 2.7792816241540863,
+      "grad_norm": 0.23391749933423464,
+      "learning_rate": 7.064182115708723e-07,
+      "loss": 0.0445,
+      "step": 5339
+    },
+    {
+      "epoch": 2.77980218636127,
+      "grad_norm": 0.23115860660431092,
+      "learning_rate": 7.031056689626031e-07,
+      "loss": 0.0437,
+      "step": 5340
+    },
+    {
+      "epoch": 2.780322748568454,
+      "grad_norm": 0.22704336146442725,
+      "learning_rate": 6.998008004442391e-07,
+      "loss": 0.0457,
+      "step": 5341
+    },
+    {
+      "epoch": 2.7808433107756376,
+      "grad_norm": 0.2226141103882042,
+      "learning_rate": 6.965036070596175e-07,
+      "loss": 0.0437,
+      "step": 5342
+    },
+    {
+      "epoch": 2.7813638729828214,
+      "grad_norm": 0.22972774937475848,
+      "learning_rate": 6.932140898501471e-07,
+      "loss": 0.0437,
+      "step": 5343
+    },
+    {
+      "epoch": 2.781884435190005,
+      "grad_norm": 0.22763949291942495,
+      "learning_rate": 6.899322498548022e-07,
+      "loss": 0.0444,
+      "step": 5344
+    },
+    {
+      "epoch": 2.782404997397189,
+      "grad_norm": 0.2269364277590439,
+      "learning_rate": 6.866580881101508e-07,
+      "loss": 0.0455,
+      "step": 5345
+    },
+    {
+      "epoch": 2.7829255596043727,
+      "grad_norm": 0.22415876770946747,
+      "learning_rate": 6.833916056503187e-07,
+      "loss": 0.0413,
+      "step": 5346
+    },
+    {
+      "epoch": 2.7834461218115565,
+      "grad_norm": 0.2169182386149355,
+      "learning_rate": 6.801328035070137e-07,
+      "loss": 0.0411,
+      "step": 5347
+    },
+    {
+      "epoch": 2.7839666840187403,
+      "grad_norm": 0.22859639804739612,
+      "learning_rate": 6.768816827095182e-07,
+      "loss": 0.0439,
+      "step": 5348
+    },
+    {
+      "epoch": 2.784487246225924,
+      "grad_norm": 0.2252301895314262,
+      "learning_rate": 6.736382442846911e-07,
+      "loss": 0.0435,
+      "step": 5349
+    },
+    {
+      "epoch": 2.785007808433108,
+      "grad_norm": 0.23199340249555175,
+      "learning_rate": 6.70402489256955e-07,
+      "loss": 0.0433,
+      "step": 5350
+    },
+    {
+      "epoch": 2.7855283706402916,
+      "grad_norm": 0.2217393408623345,
+      "learning_rate": 6.671744186483143e-07,
+      "loss": 0.0427,
+      "step": 5351
+    },
+    {
+      "epoch": 2.7860489328474753,
+      "grad_norm": 0.22705878796369014,
+      "learning_rate": 6.639540334783478e-07,
+      "loss": 0.0436,
+      "step": 5352
+    },
+    {
+      "epoch": 2.786569495054659,
+      "grad_norm": 0.22046258969292104,
+      "learning_rate": 6.60741334764195e-07,
+      "loss": 0.0408,
+      "step": 5353
+    },
+    {
+      "epoch": 2.787090057261843,
+      "grad_norm": 0.22592912598051434,
+      "learning_rate": 6.575363235205856e-07,
+      "loss": 0.0434,
+      "step": 5354
+    },
+    {
+      "epoch": 2.7876106194690267,
+      "grad_norm": 0.22293933544483777,
+      "learning_rate": 6.543390007598016e-07,
+      "loss": 0.0433,
+      "step": 5355
+    },
+    {
+      "epoch": 2.7881311816762104,
+      "grad_norm": 0.2200500939054284,
+      "learning_rate": 6.511493674917102e-07,
+      "loss": 0.0418,
+      "step": 5356
+    },
+    {
+      "epoch": 2.788651743883394,
+      "grad_norm": 0.22913113334418797,
+      "learning_rate": 6.479674247237472e-07,
+      "loss": 0.0443,
+      "step": 5357
+    },
+    {
+      "epoch": 2.789172306090578,
+      "grad_norm": 0.23960510693606646,
+      "learning_rate": 6.447931734609197e-07,
+      "loss": 0.0434,
+      "step": 5358
+    },
+    {
+      "epoch": 2.7896928682977613,
+      "grad_norm": 0.23198268720084506,
+      "learning_rate": 6.416266147058009e-07,
+      "loss": 0.0436,
+      "step": 5359
+    },
+    {
+      "epoch": 2.790213430504945,
+      "grad_norm": 0.23155582454440707,
+      "learning_rate": 6.38467749458535e-07,
+      "loss": 0.0445,
+      "step": 5360
+    },
+    {
+      "epoch": 2.790733992712129,
+      "grad_norm": 0.2298256419187941,
+      "learning_rate": 6.353165787168464e-07,
+      "loss": 0.0437,
+      "step": 5361
+    },
+    {
+      "epoch": 2.7912545549193126,
+      "grad_norm": 0.2254102181052845,
+      "learning_rate": 6.321731034760164e-07,
+      "loss": 0.0433,
+      "step": 5362
+    },
+    {
+      "epoch": 2.7917751171264964,
+      "grad_norm": 0.23436390964542603,
+      "learning_rate": 6.290373247289011e-07,
+      "loss": 0.0458,
+      "step": 5363
+    },
+    {
+      "epoch": 2.79229567933368,
+      "grad_norm": 0.21766623446970718,
+      "learning_rate": 6.259092434659247e-07,
+      "loss": 0.0416,
+      "step": 5364
+    },
+    {
+      "epoch": 2.792816241540864,
+      "grad_norm": 0.21981768528318218,
+      "learning_rate": 6.227888606750803e-07,
+      "loss": 0.042,
+      "step": 5365
+    },
+    {
+      "epoch": 2.7933368037480477,
+      "grad_norm": 0.23061777624720375,
+      "learning_rate": 6.196761773419324e-07,
+      "loss": 0.0445,
+      "step": 5366
+    },
+    {
+      "epoch": 2.7938573659552315,
+      "grad_norm": 0.21991675403724478,
+      "learning_rate": 6.165711944496083e-07,
+      "loss": 0.0409,
+      "step": 5367
+    },
+    {
+      "epoch": 2.7943779281624153,
+      "grad_norm": 0.23345443481213357,
+      "learning_rate": 6.134739129788125e-07,
+      "loss": 0.0458,
+      "step": 5368
+    },
+    {
+      "epoch": 2.794898490369599,
+      "grad_norm": 0.23028111148067149,
+      "learning_rate": 6.103843339078014e-07,
+      "loss": 0.0436,
+      "step": 5369
+    },
+    {
+      "epoch": 2.795419052576783,
+      "grad_norm": 0.2302827035468124,
+      "learning_rate": 6.073024582124165e-07,
+      "loss": 0.0435,
+      "step": 5370
+    },
+    {
+      "epoch": 2.7959396147839666,
+      "grad_norm": 0.23249897658678184,
+      "learning_rate": 6.042282868660515e-07,
+      "loss": 0.0458,
+      "step": 5371
+    },
+    {
+      "epoch": 2.7964601769911503,
+      "grad_norm": 0.22510823189817172,
+      "learning_rate": 6.011618208396768e-07,
+      "loss": 0.0426,
+      "step": 5372
+    },
+    {
+      "epoch": 2.796980739198334,
+      "grad_norm": 0.22638808421418086,
+      "learning_rate": 5.981030611018234e-07,
+      "loss": 0.042,
+      "step": 5373
+    },
+    {
+      "epoch": 2.797501301405518,
+      "grad_norm": 0.22413378380890733,
+      "learning_rate": 5.950520086185878e-07,
+      "loss": 0.0424,
+      "step": 5374
+    },
+    {
+      "epoch": 2.7980218636127017,
+      "grad_norm": 0.2430762884956928,
+      "learning_rate": 5.920086643536354e-07,
+      "loss": 0.047,
+      "step": 5375
+    },
+    {
+      "epoch": 2.7985424258198854,
+      "grad_norm": 0.232168583757956,
+      "learning_rate": 5.889730292681972e-07,
+      "loss": 0.0438,
+      "step": 5376
+    },
+    {
+      "epoch": 2.799062988027069,
+      "grad_norm": 0.24007634677429804,
+      "learning_rate": 5.859451043210701e-07,
+      "loss": 0.0442,
+      "step": 5377
+    },
+    {
+      "epoch": 2.799583550234253,
+      "grad_norm": 0.22672307210185036,
+      "learning_rate": 5.829248904686085e-07,
+      "loss": 0.0433,
+      "step": 5378
+    },
+    {
+      "epoch": 2.8001041124414368,
+      "grad_norm": 0.2346606956607048,
+      "learning_rate": 5.799123886647439e-07,
+      "loss": 0.0451,
+      "step": 5379
+    },
+    {
+      "epoch": 2.8006246746486205,
+      "grad_norm": 0.22725576849779167,
+      "learning_rate": 5.769075998609569e-07,
+      "loss": 0.0429,
+      "step": 5380
+    },
+    {
+      "epoch": 2.8011452368558043,
+      "grad_norm": 0.23830547232482183,
+      "learning_rate": 5.73910525006302e-07,
+      "loss": 0.0442,
+      "step": 5381
+    },
+    {
+      "epoch": 2.801665799062988,
+      "grad_norm": 0.22960806546029328,
+      "learning_rate": 5.709211650473972e-07,
+      "loss": 0.0425,
+      "step": 5382
+    },
+    {
+      "epoch": 2.802186361270172,
+      "grad_norm": 0.2201304972527062,
+      "learning_rate": 5.679395209284178e-07,
+      "loss": 0.0426,
+      "step": 5383
+    },
+    {
+      "epoch": 2.8027069234773556,
+      "grad_norm": 0.22778314103337913,
+      "learning_rate": 5.649655935911075e-07,
+      "loss": 0.0435,
+      "step": 5384
+    },
+    {
+      "epoch": 2.8032274856845394,
+      "grad_norm": 0.22560465254179524,
+      "learning_rate": 5.619993839747733e-07,
+      "loss": 0.0432,
+      "step": 5385
+    },
+    {
+      "epoch": 2.803748047891723,
+      "grad_norm": 0.23421440146211206,
+      "learning_rate": 5.590408930162799e-07,
+      "loss": 0.0435,
+      "step": 5386
+    },
+    {
+      "epoch": 2.804268610098907,
+      "grad_norm": 0.21559388239552502,
+      "learning_rate": 5.560901216500575e-07,
+      "loss": 0.0411,
+      "step": 5387
+    },
+    {
+      "epoch": 2.8047891723060907,
+      "grad_norm": 0.22569321419302593,
+      "learning_rate": 5.531470708080965e-07,
+      "loss": 0.0428,
+      "step": 5388
+    },
+    {
+      "epoch": 2.8053097345132745,
+      "grad_norm": 0.2256230654004025,
+      "learning_rate": 5.502117414199481e-07,
+      "loss": 0.0432,
+      "step": 5389
+    },
+    {
+      "epoch": 2.8058302967204583,
+      "grad_norm": 0.23309729929515796,
+      "learning_rate": 5.472841344127261e-07,
+      "loss": 0.043,
+      "step": 5390
+    },
+    {
+      "epoch": 2.806350858927642,
+      "grad_norm": 0.22323967104644515,
+      "learning_rate": 5.443642507111074e-07,
+      "loss": 0.0429,
+      "step": 5391
+    },
+    {
+      "epoch": 2.806871421134826,
+      "grad_norm": 0.2200278159127524,
+      "learning_rate": 5.414520912373239e-07,
+      "loss": 0.0415,
+      "step": 5392
+    },
+    {
+      "epoch": 2.8073919833420096,
+      "grad_norm": 0.22357673987141546,
+      "learning_rate": 5.38547656911173e-07,
+      "loss": 0.0434,
+      "step": 5393
+    },
+    {
+      "epoch": 2.8079125455491933,
+      "grad_norm": 0.23721653321770933,
+      "learning_rate": 5.3565094865001e-07,
+      "loss": 0.0436,
+      "step": 5394
+    },
+    {
+      "epoch": 2.808433107756377,
+      "grad_norm": 0.23374671104516043,
+      "learning_rate": 5.327619673687528e-07,
+      "loss": 0.0442,
+      "step": 5395
+    },
+    {
+      "epoch": 2.808953669963561,
+      "grad_norm": 0.2243628561143931,
+      "learning_rate": 5.298807139798689e-07,
+      "loss": 0.0434,
+      "step": 5396
+    },
+    {
+      "epoch": 2.8094742321707447,
+      "grad_norm": 0.21338440786570576,
+      "learning_rate": 5.270071893934026e-07,
+      "loss": 0.0422,
+      "step": 5397
+    },
+    {
+      "epoch": 2.8099947943779284,
+      "grad_norm": 0.21726078260576132,
+      "learning_rate": 5.24141394516936e-07,
+      "loss": 0.0425,
+      "step": 5398
+    },
+    {
+      "epoch": 2.8105153565851118,
+      "grad_norm": 0.2357056904995546,
+      "learning_rate": 5.212833302556258e-07,
+      "loss": 0.0441,
+      "step": 5399
+    },
+    {
+      "epoch": 2.8110359187922955,
+      "grad_norm": 0.2218114034723653,
+      "learning_rate": 5.184329975121832e-07,
+      "loss": 0.0414,
+      "step": 5400
+    },
+    {
+      "epoch": 2.8115564809994793,
+      "grad_norm": 0.21777410706096698,
+      "learning_rate": 5.155903971868742e-07,
+      "loss": 0.0425,
+      "step": 5401
+    },
+    {
+      "epoch": 2.812077043206663,
+      "grad_norm": 0.21863538909227456,
+      "learning_rate": 5.127555301775223e-07,
+      "loss": 0.0411,
+      "step": 5402
+    },
+    {
+      "epoch": 2.812597605413847,
+      "grad_norm": 0.2272082792608195,
+      "learning_rate": 5.099283973795111e-07,
+      "loss": 0.0438,
+      "step": 5403
+    },
+    {
+      "epoch": 2.8131181676210306,
+      "grad_norm": 0.21938578677958842,
+      "learning_rate": 5.071089996857848e-07,
+      "loss": 0.0422,
+      "step": 5404
+    },
+    {
+      "epoch": 2.8136387298282144,
+      "grad_norm": 0.22905873049766085,
+      "learning_rate": 5.042973379868365e-07,
+      "loss": 0.0422,
+      "step": 5405
+    },
+    {
+      "epoch": 2.814159292035398,
+      "grad_norm": 0.22649704219116712,
+      "learning_rate": 5.014934131707199e-07,
+      "loss": 0.0435,
+      "step": 5406
+    },
+    {
+      "epoch": 2.814679854242582,
+      "grad_norm": 0.22467215809657948,
+      "learning_rate": 4.98697226123046e-07,
+      "loss": 0.0418,
+      "step": 5407
+    },
+    {
+      "epoch": 2.8152004164497657,
+      "grad_norm": 0.22783791469459005,
+      "learning_rate": 4.959087777269805e-07,
+      "loss": 0.0446,
+      "step": 5408
+    },
+    {
+      "epoch": 2.8157209786569495,
+      "grad_norm": 0.2214050794833035,
+      "learning_rate": 4.931280688632467e-07,
+      "loss": 0.0408,
+      "step": 5409
+    },
+    {
+      "epoch": 2.8162415408641333,
+      "grad_norm": 0.22088031305505856,
+      "learning_rate": 4.90355100410117e-07,
+      "loss": 0.0416,
+      "step": 5410
+    },
+    {
+      "epoch": 2.816762103071317,
+      "grad_norm": 0.22653460185601557,
+      "learning_rate": 4.875898732434298e-07,
+      "loss": 0.0427,
+      "step": 5411
+    },
+    {
+      "epoch": 2.817282665278501,
+      "grad_norm": 0.21289858185605157,
+      "learning_rate": 4.848323882365668e-07,
+      "loss": 0.0413,
+      "step": 5412
+    },
+    {
+      "epoch": 2.8178032274856846,
+      "grad_norm": 0.2309828690362825,
+      "learning_rate": 4.820826462604788e-07,
+      "loss": 0.0434,
+      "step": 5413
+    },
+    {
+      "epoch": 2.8183237896928683,
+      "grad_norm": 0.2374990881444587,
+      "learning_rate": 4.793406481836515e-07,
+      "loss": 0.0427,
+      "step": 5414
+    },
+    {
+      "epoch": 2.818844351900052,
+      "grad_norm": 0.23637721721606647,
+      "learning_rate": 4.7660639487214496e-07,
+      "loss": 0.0464,
+      "step": 5415
+    },
+    {
+      "epoch": 2.819364914107236,
+      "grad_norm": 0.22388241364899847,
+      "learning_rate": 4.738798871895572e-07,
+      "loss": 0.0434,
+      "step": 5416
+    },
+    {
+      "epoch": 2.8198854763144197,
+      "grad_norm": 0.2210864475178364,
+      "learning_rate": 4.7116112599704666e-07,
+      "loss": 0.0435,
+      "step": 5417
+    },
+    {
+      "epoch": 2.8204060385216034,
+      "grad_norm": 0.2259240497819733,
+      "learning_rate": 4.6845011215332914e-07,
+      "loss": 0.0427,
+      "step": 5418
+    },
+    {
+      "epoch": 2.820926600728787,
+      "grad_norm": 0.219204603117465,
+      "learning_rate": 4.6574684651466415e-07,
+      "loss": 0.0411,
+      "step": 5419
+    },
+    {
+      "epoch": 2.821447162935971,
+      "grad_norm": 0.21818043914111,
+      "learning_rate": 4.6305132993487155e-07,
+      "loss": 0.0413,
+      "step": 5420
+    },
+    {
+      "epoch": 2.8219677251431547,
+      "grad_norm": 0.21659493110106742,
+      "learning_rate": 4.6036356326532024e-07,
+      "loss": 0.0402,
+      "step": 5421
+    },
+    {
+      "epoch": 2.8224882873503385,
+      "grad_norm": 0.22715592904564377,
+      "learning_rate": 4.57683547354934e-07,
+      "loss": 0.0437,
+      "step": 5422
+    },
+    {
+      "epoch": 2.823008849557522,
+      "grad_norm": 0.23159567590291835,
+      "learning_rate": 4.5501128305018013e-07,
+      "loss": 0.0432,
+      "step": 5423
+    },
+    {
+      "epoch": 2.8235294117647056,
+      "grad_norm": 0.21733811243688106,
+      "learning_rate": 4.523467711950946e-07,
+      "loss": 0.0413,
+      "step": 5424
+    },
+    {
+      "epoch": 2.8240499739718894,
+      "grad_norm": 0.23788228121357813,
+      "learning_rate": 4.496900126312431e-07,
+      "loss": 0.0433,
+      "step": 5425
+    },
+    {
+      "epoch": 2.824570536179073,
+      "grad_norm": 0.22471884600489062,
+      "learning_rate": 4.4704100819776e-07,
+      "loss": 0.0426,
+      "step": 5426
+    },
+    {
+      "epoch": 2.825091098386257,
+      "grad_norm": 0.22556479297052626,
+      "learning_rate": 4.443997587313231e-07,
+      "loss": 0.0436,
+      "step": 5427
+    },
+    {
+      "epoch": 2.8256116605934407,
+      "grad_norm": 0.2204685863010288,
+      "learning_rate": 4.4176626506616245e-07,
+      "loss": 0.0426,
+      "step": 5428
+    },
+    {
+      "epoch": 2.8261322228006245,
+      "grad_norm": 0.2222480259682644,
+      "learning_rate": 4.391405280340544e-07,
+      "loss": 0.0425,
+      "step": 5429
+    },
+    {
+      "epoch": 2.8266527850078083,
+      "grad_norm": 0.223467726694392,
+      "learning_rate": 4.365225484643326e-07,
+      "loss": 0.0435,
+      "step": 5430
+    },
+    {
+      "epoch": 2.827173347214992,
+      "grad_norm": 0.22428205507604543,
+      "learning_rate": 4.339123271838746e-07,
+      "loss": 0.0419,
+      "step": 5431
+    },
+    {
+      "epoch": 2.827693909422176,
+      "grad_norm": 0.22403734441364156,
+      "learning_rate": 4.3130986501711547e-07,
+      "loss": 0.0418,
+      "step": 5432
+    },
+    {
+      "epoch": 2.8282144716293596,
+      "grad_norm": 0.2274253646113262,
+      "learning_rate": 4.2871516278602806e-07,
+      "loss": 0.0439,
+      "step": 5433
+    },
+    {
+      "epoch": 2.8287350338365433,
+      "grad_norm": 0.22023423609927847,
+      "learning_rate": 4.2612822131013754e-07,
+      "loss": 0.0431,
+      "step": 5434
+    },
+    {
+      "epoch": 2.829255596043727,
+      "grad_norm": 0.23148682456065023,
+      "learning_rate": 4.235490414065263e-07,
+      "loss": 0.0442,
+      "step": 5435
+    },
+    {
+      "epoch": 2.829776158250911,
+      "grad_norm": 0.21570556980256503,
+      "learning_rate": 4.2097762388981775e-07,
+      "loss": 0.0404,
+      "step": 5436
+    },
+    {
+      "epoch": 2.8302967204580947,
+      "grad_norm": 0.2261675056659339,
+      "learning_rate": 4.1841396957218446e-07,
+      "loss": 0.0434,
+      "step": 5437
+    },
+    {
+      "epoch": 2.8308172826652784,
+      "grad_norm": 0.23651027031881064,
+      "learning_rate": 4.158580792633482e-07,
+      "loss": 0.0459,
+      "step": 5438
+    },
+    {
+      "epoch": 2.831337844872462,
+      "grad_norm": 0.22013855746990973,
+      "learning_rate": 4.1330995377057703e-07,
+      "loss": 0.0426,
+      "step": 5439
+    },
+    {
+      "epoch": 2.831858407079646,
+      "grad_norm": 0.21970369162169295,
+      "learning_rate": 4.107695938986883e-07,
+      "loss": 0.0412,
+      "step": 5440
+    },
+    {
+      "epoch": 2.8323789692868298,
+      "grad_norm": 0.23453520708614153,
+      "learning_rate": 4.0823700045004854e-07,
+      "loss": 0.0449,
+      "step": 5441
+    },
+    {
+      "epoch": 2.8328995314940135,
+      "grad_norm": 0.22928185978871454,
+      "learning_rate": 4.057121742245651e-07,
+      "loss": 0.0428,
+      "step": 5442
+    },
+    {
+      "epoch": 2.8334200937011973,
+      "grad_norm": 0.2378862033367092,
+      "learning_rate": 4.031951160196945e-07,
+      "loss": 0.0438,
+      "step": 5443
+    },
+    {
+      "epoch": 2.833940655908381,
+      "grad_norm": 0.22721678410714027,
+      "learning_rate": 4.0068582663044527e-07,
+      "loss": 0.0452,
+      "step": 5444
+    },
+    {
+      "epoch": 2.834461218115565,
+      "grad_norm": 0.22411659172958168,
+      "learning_rate": 3.98184306849364e-07,
+      "loss": 0.0434,
+      "step": 5445
+    },
+    {
+      "epoch": 2.8349817803227486,
+      "grad_norm": 0.22502599203444407,
+      "learning_rate": 3.9569055746654927e-07,
+      "loss": 0.0441,
+      "step": 5446
+    },
+    {
+      "epoch": 2.8355023425299324,
+      "grad_norm": 0.22506720047885997,
+      "learning_rate": 3.9320457926964313e-07,
+      "loss": 0.0434,
+      "step": 5447
+    },
+    {
+      "epoch": 2.836022904737116,
+      "grad_norm": 0.2329833809778821,
+      "learning_rate": 3.9072637304383155e-07,
+      "loss": 0.0429,
+      "step": 5448
+    },
+    {
+      "epoch": 2.8365434669443,
+      "grad_norm": 0.22427732647548107,
+      "learning_rate": 3.882559395718466e-07,
+      "loss": 0.0416,
+      "step": 5449
+    },
+    {
+      "epoch": 2.8370640291514837,
+      "grad_norm": 0.22799511119215582,
+      "learning_rate": 3.857932796339697e-07,
+      "loss": 0.0434,
+      "step": 5450
+    },
+    {
+      "epoch": 2.8375845913586675,
+      "grad_norm": 0.22666653298135034,
+      "learning_rate": 3.833383940080232e-07,
+      "loss": 0.0431,
+      "step": 5451
+    },
+    {
+      "epoch": 2.8381051535658512,
+      "grad_norm": 0.22778792498675723,
+      "learning_rate": 3.808912834693701e-07,
+      "loss": 0.0424,
+      "step": 5452
+    },
+    {
+      "epoch": 2.838625715773035,
+      "grad_norm": 0.22353151517718023,
+      "learning_rate": 3.784519487909255e-07,
+      "loss": 0.0421,
+      "step": 5453
+    },
+    {
+      "epoch": 2.839146277980219,
+      "grad_norm": 0.22576570564471596,
+      "learning_rate": 3.7602039074314254e-07,
+      "loss": 0.0432,
+      "step": 5454
+    },
+    {
+      "epoch": 2.8396668401874026,
+      "grad_norm": 0.21923761704451078,
+      "learning_rate": 3.7359661009402356e-07,
+      "loss": 0.0408,
+      "step": 5455
+    },
+    {
+      "epoch": 2.8401874023945863,
+      "grad_norm": 0.22437305305796867,
+      "learning_rate": 3.71180607609109e-07,
+      "loss": 0.0436,
+      "step": 5456
+    },
+    {
+      "epoch": 2.84070796460177,
+      "grad_norm": 0.22396548451251694,
+      "learning_rate": 3.687723840514828e-07,
+      "loss": 0.043,
+      "step": 5457
+    },
+    {
+      "epoch": 2.841228526808954,
+      "grad_norm": 0.22949152227259195,
+      "learning_rate": 3.6637194018177556e-07,
+      "loss": 0.0431,
+      "step": 5458
+    },
+    {
+      "epoch": 2.8417490890161377,
+      "grad_norm": 0.23054445515120947,
+      "learning_rate": 3.63979276758164e-07,
+      "loss": 0.0438,
+      "step": 5459
+    },
+    {
+      "epoch": 2.8422696512233214,
+      "grad_norm": 0.23505182113223505,
+      "learning_rate": 3.6159439453635757e-07,
+      "loss": 0.0446,
+      "step": 5460
+    },
+    {
+      "epoch": 2.842790213430505,
+      "grad_norm": 0.2304779734911683,
+      "learning_rate": 3.5921729426961206e-07,
+      "loss": 0.044,
+      "step": 5461
+    },
+    {
+      "epoch": 2.843310775637689,
+      "grad_norm": 0.22385791465509375,
+      "learning_rate": 3.568479767087296e-07,
+      "loss": 0.0431,
+      "step": 5462
+    },
+    {
+      "epoch": 2.8438313378448723,
+      "grad_norm": 0.2264655835605014,
+      "learning_rate": 3.544864426020478e-07,
+      "loss": 0.0442,
+      "step": 5463
+    },
+    {
+      "epoch": 2.844351900052056,
+      "grad_norm": 0.22304778819677265,
+      "learning_rate": 3.521326926954532e-07,
+      "loss": 0.044,
+      "step": 5464
+    },
+    {
+      "epoch": 2.84487246225924,
+      "grad_norm": 0.21800081748182232,
+      "learning_rate": 3.497867277323652e-07,
+      "loss": 0.0424,
+      "step": 5465
+    },
+    {
+      "epoch": 2.8453930244664236,
+      "grad_norm": 0.23473879191053773,
+      "learning_rate": 3.474485484537521e-07,
+      "loss": 0.045,
+      "step": 5466
+    },
+    {
+      "epoch": 2.8459135866736074,
+      "grad_norm": 0.23264235355245735,
+      "learning_rate": 3.451181555981148e-07,
+      "loss": 0.043,
+      "step": 5467
+    },
+    {
+      "epoch": 2.846434148880791,
+      "grad_norm": 0.22835863639242554,
+      "learning_rate": 3.42795549901509e-07,
+      "loss": 0.0433,
+      "step": 5468
+    },
+    {
+      "epoch": 2.846954711087975,
+      "grad_norm": 0.2296814079782182,
+      "learning_rate": 3.4048073209751175e-07,
+      "loss": 0.044,
+      "step": 5469
+    },
+    {
+      "epoch": 2.8474752732951587,
+      "grad_norm": 0.22896223448226471,
+      "learning_rate": 3.381737029172577e-07,
+      "loss": 0.043,
+      "step": 5470
+    },
+    {
+      "epoch": 2.8479958355023425,
+      "grad_norm": 0.2229011394279904,
+      "learning_rate": 3.358744630894084e-07,
+      "loss": 0.0433,
+      "step": 5471
+    },
+    {
+      "epoch": 2.8485163977095262,
+      "grad_norm": 0.22816466058853094,
+      "learning_rate": 3.335830133401746e-07,
+      "loss": 0.0434,
+      "step": 5472
+    },
+    {
+      "epoch": 2.84903695991671,
+      "grad_norm": 0.23157385199630254,
+      "learning_rate": 3.3129935439329963e-07,
+      "loss": 0.0439,
+      "step": 5473
+    },
+    {
+      "epoch": 2.849557522123894,
+      "grad_norm": 0.22032768627149715,
+      "learning_rate": 3.290234869700731e-07,
+      "loss": 0.0424,
+      "step": 5474
+    },
+    {
+      "epoch": 2.8500780843310776,
+      "grad_norm": 0.23190042597170488,
+      "learning_rate": 3.2675541178931456e-07,
+      "loss": 0.044,
+      "step": 5475
+    },
+    {
+      "epoch": 2.8505986465382613,
+      "grad_norm": 0.23176287027892795,
+      "learning_rate": 3.244951295673926e-07,
+      "loss": 0.0429,
+      "step": 5476
+    },
+    {
+      "epoch": 2.851119208745445,
+      "grad_norm": 0.2242543810542042,
+      "learning_rate": 3.222426410182111e-07,
+      "loss": 0.0421,
+      "step": 5477
+    },
+    {
+      "epoch": 2.851639770952629,
+      "grad_norm": 0.2206514598162686,
+      "learning_rate": 3.199979468532038e-07,
+      "loss": 0.0426,
+      "step": 5478
+    },
+    {
+      "epoch": 2.8521603331598127,
+      "grad_norm": 0.23274276038647804,
+      "learning_rate": 3.1776104778135364e-07,
+      "loss": 0.0435,
+      "step": 5479
+    },
+    {
+      "epoch": 2.8526808953669964,
+      "grad_norm": 0.22635998598505638,
+      "learning_rate": 3.155319445091787e-07,
+      "loss": 0.043,
+      "step": 5480
+    },
+    {
+      "epoch": 2.85320145757418,
+      "grad_norm": 0.2324227117826993,
+      "learning_rate": 3.1331063774072965e-07,
+      "loss": 0.044,
+      "step": 5481
+    },
+    {
+      "epoch": 2.853722019781364,
+      "grad_norm": 0.22289276085299753,
+      "learning_rate": 3.1109712817760374e-07,
+      "loss": 0.0431,
+      "step": 5482
+    },
+    {
+      "epoch": 2.8542425819885477,
+      "grad_norm": 0.2306524725426029,
+      "learning_rate": 3.0889141651892495e-07,
+      "loss": 0.0439,
+      "step": 5483
+    },
+    {
+      "epoch": 2.8547631441957315,
+      "grad_norm": 0.23473790460862667,
+      "learning_rate": 3.0669350346136106e-07,
+      "loss": 0.0446,
+      "step": 5484
+    },
+    {
+      "epoch": 2.8552837064029153,
+      "grad_norm": 0.2322283298990947,
+      "learning_rate": 3.045033896991178e-07,
+      "loss": 0.044,
+      "step": 5485
+    },
+    {
+      "epoch": 2.855804268610099,
+      "grad_norm": 0.23255630154350312,
+      "learning_rate": 3.0232107592393364e-07,
+      "loss": 0.0436,
+      "step": 5486
+    },
+    {
+      "epoch": 2.8563248308172824,
+      "grad_norm": 0.22101058262042608,
+      "learning_rate": 3.001465628250849e-07,
+      "loss": 0.0414,
+      "step": 5487
+    },
+    {
+      "epoch": 2.856845393024466,
+      "grad_norm": 0.22455476284506215,
+      "learning_rate": 2.979798510893833e-07,
+      "loss": 0.0418,
+      "step": 5488
+    },
+    {
+      "epoch": 2.85736595523165,
+      "grad_norm": 0.2183840949547814,
+      "learning_rate": 2.958209414011759e-07,
+      "loss": 0.042,
+      "step": 5489
+    },
+    {
+      "epoch": 2.8578865174388337,
+      "grad_norm": 0.2277491006600283,
+      "learning_rate": 2.936698344423505e-07,
+      "loss": 0.0439,
+      "step": 5490
+    },
+    {
+      "epoch": 2.8584070796460175,
+      "grad_norm": 0.22275283037062169,
+      "learning_rate": 2.9152653089232764e-07,
+      "loss": 0.0412,
+      "step": 5491
+    },
+    {
+      "epoch": 2.8589276418532013,
+      "grad_norm": 0.23345487136636467,
+      "learning_rate": 2.8939103142805457e-07,
+      "loss": 0.0423,
+      "step": 5492
+    },
+    {
+      "epoch": 2.859448204060385,
+      "grad_norm": 0.2319513079584263,
+      "learning_rate": 2.8726333672402796e-07,
+      "loss": 0.0446,
+      "step": 5493
+    },
+    {
+      "epoch": 2.859968766267569,
+      "grad_norm": 0.22778088078659994,
+      "learning_rate": 2.8514344745227126e-07,
+      "loss": 0.0428,
+      "step": 5494
+    },
+    {
+      "epoch": 2.8604893284747526,
+      "grad_norm": 0.2332741422674495,
+      "learning_rate": 2.8303136428234624e-07,
+      "loss": 0.0434,
+      "step": 5495
+    },
+    {
+      "epoch": 2.8610098906819363,
+      "grad_norm": 0.22627378712005314,
+      "learning_rate": 2.809270878813441e-07,
+      "loss": 0.0442,
+      "step": 5496
+    },
+    {
+      "epoch": 2.86153045288912,
+      "grad_norm": 0.22239061683349978,
+      "learning_rate": 2.788306189138945e-07,
+      "loss": 0.0428,
+      "step": 5497
+    },
+    {
+      "epoch": 2.862051015096304,
+      "grad_norm": 0.2248198343709385,
+      "learning_rate": 2.767419580421593e-07,
+      "loss": 0.0429,
+      "step": 5498
+    },
+    {
+      "epoch": 2.8625715773034877,
+      "grad_norm": 0.2304909297159804,
+      "learning_rate": 2.7466110592583585e-07,
+      "loss": 0.0413,
+      "step": 5499
+    },
+    {
+      "epoch": 2.8630921395106714,
+      "grad_norm": 0.23465634470891614,
+      "learning_rate": 2.725880632221511e-07,
+      "loss": 0.0442,
+      "step": 5500
+    },
+    {
+      "epoch": 2.863612701717855,
+      "grad_norm": 0.22387193157763116,
+      "learning_rate": 2.705228305858731e-07,
+      "loss": 0.0418,
+      "step": 5501
+    },
+    {
+      "epoch": 2.864133263925039,
+      "grad_norm": 0.23336695279278744,
+      "learning_rate": 2.684654086692939e-07,
+      "loss": 0.0445,
+      "step": 5502
+    },
+    {
+      "epoch": 2.8646538261322227,
+      "grad_norm": 0.21957744263463164,
+      "learning_rate": 2.664157981222437e-07,
+      "loss": 0.0435,
+      "step": 5503
+    },
+    {
+      "epoch": 2.8651743883394065,
+      "grad_norm": 0.2170786887345366,
+      "learning_rate": 2.6437399959208797e-07,
+      "loss": 0.0421,
+      "step": 5504
+    },
+    {
+      "epoch": 2.8656949505465903,
+      "grad_norm": 0.22627011415059628,
+      "learning_rate": 2.6234001372372194e-07,
+      "loss": 0.0428,
+      "step": 5505
+    },
+    {
+      "epoch": 2.866215512753774,
+      "grad_norm": 0.2196527927457996,
+      "learning_rate": 2.603138411595707e-07,
+      "loss": 0.0422,
+      "step": 5506
+    },
+    {
+      "epoch": 2.866736074960958,
+      "grad_norm": 0.2273020912046598,
+      "learning_rate": 2.5829548253959445e-07,
+      "loss": 0.043,
+      "step": 5507
+    },
+    {
+      "epoch": 2.8672566371681416,
+      "grad_norm": 0.22215073225966223,
+      "learning_rate": 2.5628493850128334e-07,
+      "loss": 0.0425,
+      "step": 5508
+    },
+    {
+      "epoch": 2.8677771993753254,
+      "grad_norm": 0.21964426631376047,
+      "learning_rate": 2.5428220967965986e-07,
+      "loss": 0.0419,
+      "step": 5509
+    },
+    {
+      "epoch": 2.868297761582509,
+      "grad_norm": 0.23320229447875457,
+      "learning_rate": 2.5228729670728477e-07,
+      "loss": 0.0438,
+      "step": 5510
+    },
+    {
+      "epoch": 2.868818323789693,
+      "grad_norm": 0.22196795916795412,
+      "learning_rate": 2.503002002142374e-07,
+      "loss": 0.0422,
+      "step": 5511
+    },
+    {
+      "epoch": 2.8693388859968767,
+      "grad_norm": 0.22835206156602514,
+      "learning_rate": 2.483209208281406e-07,
+      "loss": 0.0431,
+      "step": 5512
+    },
+    {
+      "epoch": 2.8698594482040605,
+      "grad_norm": 0.22386662881380173,
+      "learning_rate": 2.4634945917414164e-07,
+      "loss": 0.0421,
+      "step": 5513
+    },
+    {
+      "epoch": 2.8703800104112442,
+      "grad_norm": 0.2352898314338022,
+      "learning_rate": 2.4438581587491737e-07,
+      "loss": 0.0445,
+      "step": 5514
+    },
+    {
+      "epoch": 2.870900572618428,
+      "grad_norm": 0.23787340624984157,
+      "learning_rate": 2.424299915506828e-07,
+      "loss": 0.0447,
+      "step": 5515
+    },
+    {
+      "epoch": 2.871421134825612,
+      "grad_norm": 0.2205758396912121,
+      "learning_rate": 2.4048198681917154e-07,
+      "loss": 0.0432,
+      "step": 5516
+    },
+    {
+      "epoch": 2.8719416970327956,
+      "grad_norm": 0.21752952294602812,
+      "learning_rate": 2.3854180229565816e-07,
+      "loss": 0.0416,
+      "step": 5517
+    },
+    {
+      "epoch": 2.8724622592399793,
+      "grad_norm": 0.22801262969470634,
+      "learning_rate": 2.3660943859294127e-07,
+      "loss": 0.0438,
+      "step": 5518
+    },
+    {
+      "epoch": 2.872982821447163,
+      "grad_norm": 0.23350404962787336,
+      "learning_rate": 2.3468489632135772e-07,
+      "loss": 0.044,
+      "step": 5519
+    },
+    {
+      "epoch": 2.873503383654347,
+      "grad_norm": 0.23598923493996454,
+      "learning_rate": 2.3276817608875734e-07,
+      "loss": 0.0442,
+      "step": 5520
+    },
+    {
+      "epoch": 2.8740239458615306,
+      "grad_norm": 0.2246003814846989,
+      "learning_rate": 2.3085927850053646e-07,
+      "loss": 0.0423,
+      "step": 5521
+    },
+    {
+      "epoch": 2.8745445080687144,
+      "grad_norm": 0.22646231927768828,
+      "learning_rate": 2.2895820415961e-07,
+      "loss": 0.0444,
+      "step": 5522
+    },
+    {
+      "epoch": 2.875065070275898,
+      "grad_norm": 0.22076704915255055,
+      "learning_rate": 2.2706495366643098e-07,
+      "loss": 0.0422,
+      "step": 5523
+    },
+    {
+      "epoch": 2.875585632483082,
+      "grad_norm": 0.22233921973664836,
+      "learning_rate": 2.2517952761897387e-07,
+      "loss": 0.0406,
+      "step": 5524
+    },
+    {
+      "epoch": 2.8761061946902657,
+      "grad_norm": 0.22865295080762318,
+      "learning_rate": 2.233019266127373e-07,
+      "loss": 0.0421,
+      "step": 5525
+    },
+    {
+      "epoch": 2.8766267568974495,
+      "grad_norm": 0.22596505815503942,
+      "learning_rate": 2.2143215124076354e-07,
+      "loss": 0.0429,
+      "step": 5526
+    },
+    {
+      "epoch": 2.877147319104633,
+      "grad_norm": 0.22657611067678246,
+      "learning_rate": 2.195702020936108e-07,
+      "loss": 0.0425,
+      "step": 5527
+    },
+    {
+      "epoch": 2.8776678813118166,
+      "grad_norm": 0.22817410965615426,
+      "learning_rate": 2.1771607975937246e-07,
+      "loss": 0.0432,
+      "step": 5528
+    },
+    {
+      "epoch": 2.8781884435190004,
+      "grad_norm": 0.22318595175463884,
+      "learning_rate": 2.158697848236607e-07,
+      "loss": 0.0425,
+      "step": 5529
+    },
+    {
+      "epoch": 2.878709005726184,
+      "grad_norm": 0.23144548762355127,
+      "learning_rate": 2.1403131786962848e-07,
+      "loss": 0.043,
+      "step": 5530
+    },
+    {
+      "epoch": 2.879229567933368,
+      "grad_norm": 0.22008731677973759,
+      "learning_rate": 2.1220067947794187e-07,
+      "loss": 0.0423,
+      "step": 5531
+    },
+    {
+      "epoch": 2.8797501301405517,
+      "grad_norm": 0.22081632722175687,
+      "learning_rate": 2.1037787022681055e-07,
+      "loss": 0.0411,
+      "step": 5532
+    },
+    {
+      "epoch": 2.8802706923477355,
+      "grad_norm": 0.2261541595421086,
+      "learning_rate": 2.0856289069195457e-07,
+      "loss": 0.0435,
+      "step": 5533
+    },
+    {
+      "epoch": 2.8807912545549192,
+      "grad_norm": 0.2325478049396348,
+      "learning_rate": 2.0675574144663202e-07,
+      "loss": 0.0455,
+      "step": 5534
+    },
+    {
+      "epoch": 2.881311816762103,
+      "grad_norm": 0.22779678676819293,
+      "learning_rate": 2.0495642306162244e-07,
+      "loss": 0.0423,
+      "step": 5535
+    },
+    {
+      "epoch": 2.881832378969287,
+      "grad_norm": 0.2708994243214654,
+      "learning_rate": 2.031649361052379e-07,
+      "loss": 0.0435,
+      "step": 5536
+    },
+    {
+      "epoch": 2.8823529411764706,
+      "grad_norm": 0.225376654268015,
+      "learning_rate": 2.0138128114331467e-07,
+      "loss": 0.0425,
+      "step": 5537
+    },
+    {
+      "epoch": 2.8828735033836543,
+      "grad_norm": 0.21849928271688246,
+      "learning_rate": 1.9960545873920765e-07,
+      "loss": 0.0422,
+      "step": 5538
+    },
+    {
+      "epoch": 2.883394065590838,
+      "grad_norm": 0.2240071596809921,
+      "learning_rate": 1.978374694538071e-07,
+      "loss": 0.0446,
+      "step": 5539
+    },
+    {
+      "epoch": 2.883914627798022,
+      "grad_norm": 0.21762364229612807,
+      "learning_rate": 1.960773138455302e-07,
+      "loss": 0.0424,
+      "step": 5540
+    },
+    {
+      "epoch": 2.8844351900052057,
+      "grad_norm": 0.22194985729963923,
+      "learning_rate": 1.9432499247031e-07,
+      "loss": 0.0419,
+      "step": 5541
+    },
+    {
+      "epoch": 2.8849557522123894,
+      "grad_norm": 0.2192694312980397,
+      "learning_rate": 1.925805058816177e-07,
+      "loss": 0.0414,
+      "step": 5542
+    },
+    {
+      "epoch": 2.885476314419573,
+      "grad_norm": 0.2458743689393578,
+      "learning_rate": 1.9084385463043475e-07,
+      "loss": 0.0451,
+      "step": 5543
+    },
+    {
+      "epoch": 2.885996876626757,
+      "grad_norm": 0.22425721644334612,
+      "learning_rate": 1.891150392652835e-07,
+      "loss": 0.0417,
+      "step": 5544
+    },
+    {
+      "epoch": 2.8865174388339407,
+      "grad_norm": 0.2275200585924636,
+      "learning_rate": 1.873940603322022e-07,
+      "loss": 0.0437,
+      "step": 5545
+    },
+    {
+      "epoch": 2.8870380010411245,
+      "grad_norm": 0.22808079161183273,
+      "learning_rate": 1.85680918374756e-07,
+      "loss": 0.0427,
+      "step": 5546
+    },
+    {
+      "epoch": 2.8875585632483083,
+      "grad_norm": 0.2232387790940575,
+      "learning_rate": 1.8397561393403427e-07,
+      "loss": 0.0443,
+      "step": 5547
+    },
+    {
+      "epoch": 2.888079125455492,
+      "grad_norm": 0.22740682941731324,
+      "learning_rate": 1.8227814754865068e-07,
+      "loss": 0.0444,
+      "step": 5548
+    },
+    {
+      "epoch": 2.888599687662676,
+      "grad_norm": 0.2304311796701836,
+      "learning_rate": 1.8058851975474577e-07,
+      "loss": 0.0446,
+      "step": 5549
+    },
+    {
+      "epoch": 2.8891202498698596,
+      "grad_norm": 0.2347568933658674,
+      "learning_rate": 1.7890673108598433e-07,
+      "loss": 0.0424,
+      "step": 5550
+    },
+    {
+      "epoch": 2.889640812077043,
+      "grad_norm": 0.22430118829279044,
+      "learning_rate": 1.7723278207354976e-07,
+      "loss": 0.0433,
+      "step": 5551
+    },
+    {
+      "epoch": 2.8901613742842267,
+      "grad_norm": 0.22066156804762518,
+      "learning_rate": 1.7556667324615527e-07,
+      "loss": 0.0417,
+      "step": 5552
+    },
+    {
+      "epoch": 2.8906819364914105,
+      "grad_norm": 0.22032601369478175,
+      "learning_rate": 1.7390840513003548e-07,
+      "loss": 0.0415,
+      "step": 5553
+    },
+    {
+      "epoch": 2.8912024986985942,
+      "grad_norm": 0.23669903664743355,
+      "learning_rate": 1.722579782489464e-07,
+      "loss": 0.0443,
+      "step": 5554
+    },
+    {
+      "epoch": 2.891723060905778,
+      "grad_norm": 0.23261439457929947,
+      "learning_rate": 1.7061539312417108e-07,
+      "loss": 0.0434,
+      "step": 5555
+    },
+    {
+      "epoch": 2.892243623112962,
+      "grad_norm": 0.22614962993197088,
+      "learning_rate": 1.689806502745167e-07,
+      "loss": 0.042,
+      "step": 5556
+    },
+    {
+      "epoch": 2.8927641853201456,
+      "grad_norm": 0.22566726168640677,
+      "learning_rate": 1.6735375021630916e-07,
+      "loss": 0.0437,
+      "step": 5557
+    },
+    {
+      "epoch": 2.8932847475273293,
+      "grad_norm": 0.23008423849366555,
+      "learning_rate": 1.6573469346339576e-07,
+      "loss": 0.0432,
+      "step": 5558
+    },
+    {
+      "epoch": 2.893805309734513,
+      "grad_norm": 0.22692870640049476,
+      "learning_rate": 1.6412348052715632e-07,
+      "loss": 0.0448,
+      "step": 5559
+    },
+    {
+      "epoch": 2.894325871941697,
+      "grad_norm": 0.22253713385008417,
+      "learning_rate": 1.6252011191648653e-07,
+      "loss": 0.0418,
+      "step": 5560
+    },
+    {
+      "epoch": 2.8948464341488807,
+      "grad_norm": 0.24561203289796596,
+      "learning_rate": 1.6092458813779797e-07,
+      "loss": 0.043,
+      "step": 5561
+    },
+    {
+      "epoch": 2.8953669963560644,
+      "grad_norm": 0.2272264787299964,
+      "learning_rate": 1.593369096950348e-07,
+      "loss": 0.0422,
+      "step": 5562
+    },
+    {
+      "epoch": 2.895887558563248,
+      "grad_norm": 0.21728188819033073,
+      "learning_rate": 1.5775707708966248e-07,
+      "loss": 0.0418,
+      "step": 5563
+    },
+    {
+      "epoch": 2.896408120770432,
+      "grad_norm": 0.23004082744114976,
+      "learning_rate": 1.5618509082066246e-07,
+      "loss": 0.0437,
+      "step": 5564
+    },
+    {
+      "epoch": 2.8969286829776157,
+      "grad_norm": 0.24070902532286687,
+      "learning_rate": 1.5462095138454314e-07,
+      "loss": 0.043,
+      "step": 5565
+    },
+    {
+      "epoch": 2.8974492451847995,
+      "grad_norm": 0.2340610425537053,
+      "learning_rate": 1.5306465927533154e-07,
+      "loss": 0.0439,
+      "step": 5566
+    },
+    {
+      "epoch": 2.8979698073919833,
+      "grad_norm": 0.2266359975592615,
+      "learning_rate": 1.515162149845789e-07,
+      "loss": 0.0427,
+      "step": 5567
+    },
+    {
+      "epoch": 2.898490369599167,
+      "grad_norm": 0.22807262882625087,
+      "learning_rate": 1.4997561900135238e-07,
+      "loss": 0.0428,
+      "step": 5568
+    },
+    {
+      "epoch": 2.899010931806351,
+      "grad_norm": 0.22716325142678023,
+      "learning_rate": 1.4844287181224603e-07,
+      "loss": 0.0439,
+      "step": 5569
+    },
+    {
+      "epoch": 2.8995314940135346,
+      "grad_norm": 0.22620114345008535,
+      "learning_rate": 1.469179739013754e-07,
+      "loss": 0.0443,
+      "step": 5570
+    },
+    {
+      "epoch": 2.9000520562207184,
+      "grad_norm": 0.22824850058053361,
+      "learning_rate": 1.4540092575036636e-07,
+      "loss": 0.043,
+      "step": 5571
+    },
+    {
+      "epoch": 2.900572618427902,
+      "grad_norm": 0.23480974832174725,
+      "learning_rate": 1.4389172783838277e-07,
+      "loss": 0.0467,
+      "step": 5572
+    },
+    {
+      "epoch": 2.901093180635086,
+      "grad_norm": 0.22686928685216148,
+      "learning_rate": 1.4239038064209343e-07,
+      "loss": 0.0426,
+      "step": 5573
+    },
+    {
+      "epoch": 2.9016137428422697,
+      "grad_norm": 0.2185130750861107,
+      "learning_rate": 1.4089688463569394e-07,
+      "loss": 0.0412,
+      "step": 5574
+    },
+    {
+      "epoch": 2.9021343050494535,
+      "grad_norm": 0.22340460528624712,
+      "learning_rate": 1.3941124029090425e-07,
+      "loss": 0.0427,
+      "step": 5575
+    },
+    {
+      "epoch": 2.9026548672566372,
+      "grad_norm": 0.24387629963245097,
+      "learning_rate": 1.379334480769545e-07,
+      "loss": 0.046,
+      "step": 5576
+    },
+    {
+      "epoch": 2.903175429463821,
+      "grad_norm": 0.22899996540170645,
+      "learning_rate": 1.3646350846060197e-07,
+      "loss": 0.0453,
+      "step": 5577
+    },
+    {
+      "epoch": 2.903695991671005,
+      "grad_norm": 0.22580749048473495,
+      "learning_rate": 1.3500142190612797e-07,
+      "loss": 0.043,
+      "step": 5578
+    },
+    {
+      "epoch": 2.9042165538781886,
+      "grad_norm": 0.22259129283476642,
+      "learning_rate": 1.335471888753187e-07,
+      "loss": 0.0431,
+      "step": 5579
+    },
+    {
+      "epoch": 2.9047371160853723,
+      "grad_norm": 0.2222893681291658,
+      "learning_rate": 1.3210080982749284e-07,
+      "loss": 0.0424,
+      "step": 5580
+    },
+    {
+      "epoch": 2.905257678292556,
+      "grad_norm": 0.21864221805966788,
+      "learning_rate": 1.306622852194822e-07,
+      "loss": 0.0413,
+      "step": 5581
+    },
+    {
+      "epoch": 2.90577824049974,
+      "grad_norm": 0.2293590339534558,
+      "learning_rate": 1.2923161550564278e-07,
+      "loss": 0.0436,
+      "step": 5582
+    },
+    {
+      "epoch": 2.9062988027069236,
+      "grad_norm": 0.228497915949144,
+      "learning_rate": 1.2780880113784365e-07,
+      "loss": 0.0449,
+      "step": 5583
+    },
+    {
+      "epoch": 2.9068193649141074,
+      "grad_norm": 0.21641699537000803,
+      "learning_rate": 1.2639384256547816e-07,
+      "loss": 0.0413,
+      "step": 5584
+    },
+    {
+      "epoch": 2.907339927121291,
+      "grad_norm": 0.21723573180469424,
+      "learning_rate": 1.2498674023545543e-07,
+      "loss": 0.042,
+      "step": 5585
+    },
+    {
+      "epoch": 2.907860489328475,
+      "grad_norm": 0.2299450832808966,
+      "learning_rate": 1.2358749459220332e-07,
+      "loss": 0.0431,
+      "step": 5586
+    },
+    {
+      "epoch": 2.9083810515356587,
+      "grad_norm": 0.23308971409684184,
+      "learning_rate": 1.221961060776683e-07,
+      "loss": 0.0439,
+      "step": 5587
+    },
+    {
+      "epoch": 2.9089016137428425,
+      "grad_norm": 0.23005044711744224,
+      "learning_rate": 1.2081257513131828e-07,
+      "loss": 0.0447,
+      "step": 5588
+    },
+    {
+      "epoch": 2.9094221759500263,
+      "grad_norm": 0.22157757520601382,
+      "learning_rate": 1.1943690219013148e-07,
+      "loss": 0.0424,
+      "step": 5589
+    },
+    {
+      "epoch": 2.90994273815721,
+      "grad_norm": 0.22673099799671645,
+      "learning_rate": 1.1806908768861314e-07,
+      "loss": 0.0427,
+      "step": 5590
+    },
+    {
+      "epoch": 2.9104633003643934,
+      "grad_norm": 0.22789349189024055,
+      "learning_rate": 1.167091320587843e-07,
+      "loss": 0.0433,
+      "step": 5591
+    },
+    {
+      "epoch": 2.910983862571577,
+      "grad_norm": 0.221483966989294,
+      "learning_rate": 1.1535703573017919e-07,
+      "loss": 0.0421,
+      "step": 5592
+    },
+    {
+      "epoch": 2.911504424778761,
+      "grad_norm": 0.22112758106997762,
+      "learning_rate": 1.1401279912985341e-07,
+      "loss": 0.0426,
+      "step": 5593
+    },
+    {
+      "epoch": 2.9120249869859447,
+      "grad_norm": 0.22109349725590008,
+      "learning_rate": 1.1267642268238121e-07,
+      "loss": 0.0426,
+      "step": 5594
+    },
+    {
+      "epoch": 2.9125455491931285,
+      "grad_norm": 0.22080664506069955,
+      "learning_rate": 1.1134790680984996e-07,
+      "loss": 0.0418,
+      "step": 5595
+    },
+    {
+      "epoch": 2.9130661114003122,
+      "grad_norm": 0.22213280394837107,
+      "learning_rate": 1.1002725193186847e-07,
+      "loss": 0.0414,
+      "step": 5596
+    },
+    {
+      "epoch": 2.913586673607496,
+      "grad_norm": 0.230005852903613,
+      "learning_rate": 1.0871445846555861e-07,
+      "loss": 0.0451,
+      "step": 5597
+    },
+    {
+      "epoch": 2.91410723581468,
+      "grad_norm": 0.22684379339260194,
+      "learning_rate": 1.0740952682556371e-07,
+      "loss": 0.0439,
+      "step": 5598
+    },
+    {
+      "epoch": 2.9146277980218636,
+      "grad_norm": 0.2309550050238035,
+      "learning_rate": 1.0611245742404297e-07,
+      "loss": 0.0443,
+      "step": 5599
+    },
+    {
+      "epoch": 2.9151483602290473,
+      "grad_norm": 0.22694425419899925,
+      "learning_rate": 1.0482325067066868e-07,
+      "loss": 0.0429,
+      "step": 5600
+    },
+    {
+      "epoch": 2.915668922436231,
+      "grad_norm": 0.2355942815015955,
+      "learning_rate": 1.0354190697263454e-07,
+      "loss": 0.0416,
+      "step": 5601
+    },
+    {
+      "epoch": 2.916189484643415,
+      "grad_norm": 0.22915067675452885,
+      "learning_rate": 1.0226842673464742e-07,
+      "loss": 0.0431,
+      "step": 5602
+    },
+    {
+      "epoch": 2.9167100468505986,
+      "grad_norm": 0.22859882687372848,
+      "learning_rate": 1.0100281035893277e-07,
+      "loss": 0.0421,
+      "step": 5603
+    },
+    {
+      "epoch": 2.9172306090577824,
+      "grad_norm": 0.22612284310863184,
+      "learning_rate": 9.97450582452264e-08,
+      "loss": 0.0432,
+      "step": 5604
+    },
+    {
+      "epoch": 2.917751171264966,
+      "grad_norm": 0.2371748286947038,
+      "learning_rate": 9.849517079079384e-08,
+      "loss": 0.0442,
+      "step": 5605
+    },
+    {
+      "epoch": 2.91827173347215,
+      "grad_norm": 0.23429013057909479,
+      "learning_rate": 9.725314839039989e-08,
+      "loss": 0.0456,
+      "step": 5606
+    },
+    {
+      "epoch": 2.9187922956793337,
+      "grad_norm": 0.2252893473582409,
+      "learning_rate": 9.60189914363363e-08,
+      "loss": 0.0435,
+      "step": 5607
+    },
+    {
+      "epoch": 2.9193128578865175,
+      "grad_norm": 0.22061594822107208,
+      "learning_rate": 9.479270031840514e-08,
+      "loss": 0.0426,
+      "step": 5608
+    },
+    {
+      "epoch": 2.9198334200937013,
+      "grad_norm": 0.22394052877244155,
+      "learning_rate": 9.357427542392716e-08,
+      "loss": 0.0443,
+      "step": 5609
+    },
+    {
+      "epoch": 2.920353982300885,
+      "grad_norm": 0.22924733216319026,
+      "learning_rate": 9.236371713774172e-08,
+      "loss": 0.0432,
+      "step": 5610
+    },
+    {
+      "epoch": 2.920874544508069,
+      "grad_norm": 0.22008904881055802,
+      "learning_rate": 9.116102584219299e-08,
+      "loss": 0.0411,
+      "step": 5611
+    },
+    {
+      "epoch": 2.9213951067152526,
+      "grad_norm": 0.23123570570029328,
+      "learning_rate": 8.996620191714933e-08,
+      "loss": 0.0448,
+      "step": 5612
+    },
+    {
+      "epoch": 2.9219156689224364,
+      "grad_norm": 0.22502509485305863,
+      "learning_rate": 8.877924573999496e-08,
+      "loss": 0.0435,
+      "step": 5613
+    },
+    {
+      "epoch": 2.92243623112962,
+      "grad_norm": 0.22614275559557945,
+      "learning_rate": 8.760015768562169e-08,
+      "loss": 0.0431,
+      "step": 5614
+    },
+    {
+      "epoch": 2.9229567933368035,
+      "grad_norm": 0.23309402951892166,
+      "learning_rate": 8.642893812644271e-08,
+      "loss": 0.043,
+      "step": 5615
+    },
+    {
+      "epoch": 2.9234773555439872,
+      "grad_norm": 0.23455703296115385,
+      "learning_rate": 8.526558743238156e-08,
+      "loss": 0.0459,
+      "step": 5616
+    },
+    {
+      "epoch": 2.923997917751171,
+      "grad_norm": 0.22213718683676725,
+      "learning_rate": 8.411010597088043e-08,
+      "loss": 0.0411,
+      "step": 5617
+    },
+    {
+      "epoch": 2.924518479958355,
+      "grad_norm": 0.2309034800189144,
+      "learning_rate": 8.296249410689461e-08,
+      "loss": 0.0446,
+      "step": 5618
+    },
+    {
+      "epoch": 2.9250390421655386,
+      "grad_norm": 0.22884832459468085,
+      "learning_rate": 8.182275220288971e-08,
+      "loss": 0.0437,
+      "step": 5619
+    },
+    {
+      "epoch": 2.9255596043727223,
+      "grad_norm": 0.22329303957735033,
+      "learning_rate": 8.069088061885277e-08,
+      "loss": 0.0421,
+      "step": 5620
+    },
+    {
+      "epoch": 2.926080166579906,
+      "grad_norm": 0.22435831062356618,
+      "learning_rate": 7.956687971228116e-08,
+      "loss": 0.0445,
+      "step": 5621
+    },
+    {
+      "epoch": 2.92660072878709,
+      "grad_norm": 0.22004471253385186,
+      "learning_rate": 7.845074983818534e-08,
+      "loss": 0.0423,
+      "step": 5622
+    },
+    {
+      "epoch": 2.9271212909942737,
+      "grad_norm": 0.2281975415710203,
+      "learning_rate": 7.734249134909166e-08,
+      "loss": 0.0443,
+      "step": 5623
+    },
+    {
+      "epoch": 2.9276418532014574,
+      "grad_norm": 0.2347633299429142,
+      "learning_rate": 7.624210459504233e-08,
+      "loss": 0.0468,
+      "step": 5624
+    },
+    {
+      "epoch": 2.928162415408641,
+      "grad_norm": 0.2244782906812353,
+      "learning_rate": 7.514958992358434e-08,
+      "loss": 0.0429,
+      "step": 5625
+    },
+    {
+      "epoch": 2.928682977615825,
+      "grad_norm": 0.22082459228471335,
+      "learning_rate": 7.406494767979167e-08,
+      "loss": 0.0404,
+      "step": 5626
+    },
+    {
+      "epoch": 2.9292035398230087,
+      "grad_norm": 0.23891153241426735,
+      "learning_rate": 7.298817820623749e-08,
+      "loss": 0.0452,
+      "step": 5627
+    },
+    {
+      "epoch": 2.9297241020301925,
+      "grad_norm": 0.2285958330718903,
+      "learning_rate": 7.191928184302477e-08,
+      "loss": 0.0424,
+      "step": 5628
+    },
+    {
+      "epoch": 2.9302446642373763,
+      "grad_norm": 0.21899788706540083,
+      "learning_rate": 7.085825892775288e-08,
+      "loss": 0.0439,
+      "step": 5629
+    },
+    {
+      "epoch": 2.93076522644456,
+      "grad_norm": 0.22669475507459996,
+      "learning_rate": 6.980510979554545e-08,
+      "loss": 0.0419,
+      "step": 5630
+    },
+    {
+      "epoch": 2.931285788651744,
+      "grad_norm": 0.23827004869207266,
+      "learning_rate": 6.875983477903635e-08,
+      "loss": 0.0448,
+      "step": 5631
+    },
+    {
+      "epoch": 2.9318063508589276,
+      "grad_norm": 0.23319352811948646,
+      "learning_rate": 6.772243420836988e-08,
+      "loss": 0.0425,
+      "step": 5632
+    },
+    {
+      "epoch": 2.9323269130661114,
+      "grad_norm": 0.21762653350885441,
+      "learning_rate": 6.66929084112089e-08,
+      "loss": 0.0409,
+      "step": 5633
+    },
+    {
+      "epoch": 2.932847475273295,
+      "grad_norm": 0.2387237574668359,
+      "learning_rate": 6.567125771272387e-08,
+      "loss": 0.0452,
+      "step": 5634
+    },
+    {
+      "epoch": 2.933368037480479,
+      "grad_norm": 0.23525862361486116,
+      "learning_rate": 6.465748243559556e-08,
+      "loss": 0.0429,
+      "step": 5635
+    },
+    {
+      "epoch": 2.9338885996876627,
+      "grad_norm": 0.23174264458375252,
+      "learning_rate": 6.365158290002893e-08,
+      "loss": 0.0442,
+      "step": 5636
+    },
+    {
+      "epoch": 2.9344091618948465,
+      "grad_norm": 0.2253382202261745,
+      "learning_rate": 6.26535594237282e-08,
+      "loss": 0.0426,
+      "step": 5637
+    },
+    {
+      "epoch": 2.9349297241020302,
+      "grad_norm": 0.23139331028310633,
+      "learning_rate": 6.166341232191896e-08,
+      "loss": 0.0437,
+      "step": 5638
+    },
+    {
+      "epoch": 2.935450286309214,
+      "grad_norm": 0.22693372243517235,
+      "learning_rate": 6.068114190733442e-08,
+      "loss": 0.0429,
+      "step": 5639
+    },
+    {
+      "epoch": 2.935970848516398,
+      "grad_norm": 0.2286453647937163,
+      "learning_rate": 5.970674849022084e-08,
+      "loss": 0.0427,
+      "step": 5640
+    },
+    {
+      "epoch": 2.9364914107235816,
+      "grad_norm": 0.21710776971071516,
+      "learning_rate": 5.874023237833759e-08,
+      "loss": 0.0422,
+      "step": 5641
+    },
+    {
+      "epoch": 2.9370119729307653,
+      "grad_norm": 0.22418298046187546,
+      "learning_rate": 5.7781593876954366e-08,
+      "loss": 0.0432,
+      "step": 5642
+    },
+    {
+      "epoch": 2.937532535137949,
+      "grad_norm": 0.2227493313624406,
+      "learning_rate": 5.683083328885397e-08,
+      "loss": 0.0413,
+      "step": 5643
+    },
+    {
+      "epoch": 2.938053097345133,
+      "grad_norm": 0.22433420679943167,
+      "learning_rate": 5.58879509143323e-08,
+      "loss": 0.0422,
+      "step": 5644
+    },
+    {
+      "epoch": 2.9385736595523166,
+      "grad_norm": 0.2201777114475942,
+      "learning_rate": 5.495294705119558e-08,
+      "loss": 0.0427,
+      "step": 5645
+    },
+    {
+      "epoch": 2.9390942217595004,
+      "grad_norm": 0.2272931098330437,
+      "learning_rate": 5.402582199476036e-08,
+      "loss": 0.0427,
+      "step": 5646
+    },
+    {
+      "epoch": 2.939614783966684,
+      "grad_norm": 0.21987207104156062,
+      "learning_rate": 5.3106576037856295e-08,
+      "loss": 0.0416,
+      "step": 5647
+    },
+    {
+      "epoch": 2.940135346173868,
+      "grad_norm": 0.23629433671926608,
+      "learning_rate": 5.2195209470823345e-08,
+      "loss": 0.0434,
+      "step": 5648
+    },
+    {
+      "epoch": 2.9406559083810517,
+      "grad_norm": 0.226357025922961,
+      "learning_rate": 5.1291722581511826e-08,
+      "loss": 0.0437,
+      "step": 5649
+    },
+    {
+      "epoch": 2.9411764705882355,
+      "grad_norm": 0.22887056354520308,
+      "learning_rate": 5.039611565529068e-08,
+      "loss": 0.0424,
+      "step": 5650
+    },
+    {
+      "epoch": 2.9416970327954193,
+      "grad_norm": 0.22765259780951763,
+      "learning_rate": 4.950838897503085e-08,
+      "loss": 0.0434,
+      "step": 5651
+    },
+    {
+      "epoch": 2.942217595002603,
+      "grad_norm": 0.22353601786311908,
+      "learning_rate": 4.86285428211164e-08,
+      "loss": 0.0419,
+      "step": 5652
+    },
+    {
+      "epoch": 2.942738157209787,
+      "grad_norm": 0.22628985062359203,
+      "learning_rate": 4.7756577471444466e-08,
+      "loss": 0.0427,
+      "step": 5653
+    },
+    {
+      "epoch": 2.9432587194169706,
+      "grad_norm": 0.2334402923764293,
+      "learning_rate": 4.6892493201422525e-08,
+      "loss": 0.0443,
+      "step": 5654
+    },
+    {
+      "epoch": 2.943779281624154,
+      "grad_norm": 0.2213843214467801,
+      "learning_rate": 4.603629028397116e-08,
+      "loss": 0.0411,
+      "step": 5655
+    },
+    {
+      "epoch": 2.9442998438313377,
+      "grad_norm": 0.22400947744966332,
+      "learning_rate": 4.518796898951572e-08,
+      "loss": 0.0424,
+      "step": 5656
+    },
+    {
+      "epoch": 2.9448204060385215,
+      "grad_norm": 0.2205563956277841,
+      "learning_rate": 4.434752958599464e-08,
+      "loss": 0.0419,
+      "step": 5657
+    },
+    {
+      "epoch": 2.9453409682457052,
+      "grad_norm": 0.2218373906484088,
+      "learning_rate": 4.351497233886226e-08,
+      "loss": 0.0423,
+      "step": 5658
+    },
+    {
+      "epoch": 2.945861530452889,
+      "grad_norm": 0.2257691605384441,
+      "learning_rate": 4.26902975110749e-08,
+      "loss": 0.0434,
+      "step": 5659
+    },
+    {
+      "epoch": 2.946382092660073,
+      "grad_norm": 0.21762331110270125,
+      "learning_rate": 4.1873505363104746e-08,
+      "loss": 0.0414,
+      "step": 5660
+    },
+    {
+      "epoch": 2.9469026548672566,
+      "grad_norm": 0.2270394287344173,
+      "learning_rate": 4.106459615293434e-08,
+      "loss": 0.0434,
+      "step": 5661
+    },
+    {
+      "epoch": 2.9474232170744403,
+      "grad_norm": 0.22235806351067547,
+      "learning_rate": 4.026357013605098e-08,
+      "loss": 0.0432,
+      "step": 5662
+    },
+    {
+      "epoch": 2.947943779281624,
+      "grad_norm": 0.20548692652481407,
+      "learning_rate": 3.9470427565460596e-08,
+      "loss": 0.0413,
+      "step": 5663
+    },
+    {
+      "epoch": 2.948464341488808,
+      "grad_norm": 0.2279626929922583,
+      "learning_rate": 3.8685168691671156e-08,
+      "loss": 0.042,
+      "step": 5664
+    },
+    {
+      "epoch": 2.9489849036959916,
+      "grad_norm": 0.2257806223440966,
+      "learning_rate": 3.7907793762703705e-08,
+      "loss": 0.0418,
+      "step": 5665
+    },
+    {
+      "epoch": 2.9495054659031754,
+      "grad_norm": 0.2270748445679294,
+      "learning_rate": 3.713830302409238e-08,
+      "loss": 0.0433,
+      "step": 5666
+    },
+    {
+      "epoch": 2.950026028110359,
+      "grad_norm": 0.22232084572422514,
+      "learning_rate": 3.637669671887611e-08,
+      "loss": 0.0418,
+      "step": 5667
+    },
+    {
+      "epoch": 2.950546590317543,
+      "grad_norm": 0.22207686160720008,
+      "learning_rate": 3.562297508760415e-08,
+      "loss": 0.042,
+      "step": 5668
+    },
+    {
+      "epoch": 2.9510671525247267,
+      "grad_norm": 0.2335518580519873,
+      "learning_rate": 3.4877138368341614e-08,
+      "loss": 0.0447,
+      "step": 5669
+    },
+    {
+      "epoch": 2.9515877147319105,
+      "grad_norm": 0.23000775212512042,
+      "learning_rate": 3.413918679665284e-08,
+      "loss": 0.0434,
+      "step": 5670
+    },
+    {
+      "epoch": 2.9521082769390943,
+      "grad_norm": 0.2200961120523701,
+      "learning_rate": 3.3409120605623624e-08,
+      "loss": 0.0416,
+      "step": 5671
+    },
+    {
+      "epoch": 2.952628839146278,
+      "grad_norm": 0.2211023507432535,
+      "learning_rate": 3.268694002583617e-08,
+      "loss": 0.0421,
+      "step": 5672
+    },
+    {
+      "epoch": 2.953149401353462,
+      "grad_norm": 0.23087187119179148,
+      "learning_rate": 3.197264528539135e-08,
+      "loss": 0.0447,
+      "step": 5673
+    },
+    {
+      "epoch": 2.9536699635606456,
+      "grad_norm": 0.22974934193368118,
+      "learning_rate": 3.1266236609900356e-08,
+      "loss": 0.0436,
+      "step": 5674
+    },
+    {
+      "epoch": 2.9541905257678294,
+      "grad_norm": 0.23183069263185715,
+      "learning_rate": 3.056771422247362e-08,
+      "loss": 0.0428,
+      "step": 5675
+    },
+    {
+      "epoch": 2.954711087975013,
+      "grad_norm": 0.22537547711234884,
+      "learning_rate": 2.9877078343740206e-08,
+      "loss": 0.0435,
+      "step": 5676
+    },
+    {
+      "epoch": 2.955231650182197,
+      "grad_norm": 0.22654670232931923,
+      "learning_rate": 2.919432919183396e-08,
+      "loss": 0.042,
+      "step": 5677
+    },
+    {
+      "epoch": 2.9557522123893807,
+      "grad_norm": 0.22542875112228397,
+      "learning_rate": 2.851946698240182e-08,
+      "loss": 0.0431,
+      "step": 5678
+    },
+    {
+      "epoch": 2.956272774596564,
+      "grad_norm": 0.22216150114544114,
+      "learning_rate": 2.7852491928595514e-08,
+      "loss": 0.0407,
+      "step": 5679
+    },
+    {
+      "epoch": 2.956793336803748,
+      "grad_norm": 0.22284731710279135,
+      "learning_rate": 2.7193404241074306e-08,
+      "loss": 0.0428,
+      "step": 5680
+    },
+    {
+      "epoch": 2.9573138990109316,
+      "grad_norm": 0.22243490757474335,
+      "learning_rate": 2.654220412801056e-08,
+      "loss": 0.0416,
+      "step": 5681
+    },
+    {
+      "epoch": 2.9578344612181153,
+      "grad_norm": 0.22588973540324708,
+      "learning_rate": 2.5898891795084202e-08,
+      "loss": 0.0429,
+      "step": 5682
+    },
+    {
+      "epoch": 2.958355023425299,
+      "grad_norm": 0.2378710459634585,
+      "learning_rate": 2.5263467445479915e-08,
+      "loss": 0.0448,
+      "step": 5683
+    },
+    {
+      "epoch": 2.958875585632483,
+      "grad_norm": 0.23245001982292812,
+      "learning_rate": 2.4635931279898273e-08,
+      "loss": 0.043,
+      "step": 5684
+    },
+    {
+      "epoch": 2.9593961478396666,
+      "grad_norm": 0.2277675922701488,
+      "learning_rate": 2.4016283496544613e-08,
+      "loss": 0.043,
+      "step": 5685
+    },
+    {
+      "epoch": 2.9599167100468504,
+      "grad_norm": 0.21835077756145146,
+      "learning_rate": 2.340452429112905e-08,
+      "loss": 0.0416,
+      "step": 5686
+    },
+    {
+      "epoch": 2.960437272254034,
+      "grad_norm": 0.20768373110222885,
+      "learning_rate": 2.2800653856874797e-08,
+      "loss": 0.0403,
+      "step": 5687
+    },
+    {
+      "epoch": 2.960957834461218,
+      "grad_norm": 0.22352062317713317,
+      "learning_rate": 2.2204672384512625e-08,
+      "loss": 0.0415,
+      "step": 5688
+    },
+    {
+      "epoch": 2.9614783966684017,
+      "grad_norm": 0.22503022396452138,
+      "learning_rate": 2.161658006228362e-08,
+      "loss": 0.0422,
+      "step": 5689
+    },
+    {
+      "epoch": 2.9619989588755855,
+      "grad_norm": 0.22166037598844507,
+      "learning_rate": 2.1036377075930867e-08,
+      "loss": 0.0415,
+      "step": 5690
+    },
+    {
+      "epoch": 2.9625195210827693,
+      "grad_norm": 0.22650211899906575,
+      "learning_rate": 2.046406360871056e-08,
+      "loss": 0.0425,
+      "step": 5691
+    },
+    {
+      "epoch": 2.963040083289953,
+      "grad_norm": 0.22401814975398185,
+      "learning_rate": 1.989963984138643e-08,
+      "loss": 0.0444,
+      "step": 5692
+    },
+    {
+      "epoch": 2.963560645497137,
+      "grad_norm": 0.22589420895588494,
+      "learning_rate": 1.9343105952229768e-08,
+      "loss": 0.0425,
+      "step": 5693
+    },
+    {
+      "epoch": 2.9640812077043206,
+      "grad_norm": 0.2303703387090916,
+      "learning_rate": 1.8794462117022182e-08,
+      "loss": 0.0437,
+      "step": 5694
+    },
+    {
+      "epoch": 2.9646017699115044,
+      "grad_norm": 0.22853971241469895,
+      "learning_rate": 1.8253708509047285e-08,
+      "loss": 0.0427,
+      "step": 5695
+    },
+    {
+      "epoch": 2.965122332118688,
+      "grad_norm": 0.22039445211667255,
+      "learning_rate": 1.7720845299101783e-08,
+      "loss": 0.0427,
+      "step": 5696
+    },
+    {
+      "epoch": 2.965642894325872,
+      "grad_norm": 0.2267933732428841,
+      "learning_rate": 1.7195872655487166e-08,
+      "loss": 0.0439,
+      "step": 5697
+    },
+    {
+      "epoch": 2.9661634565330557,
+      "grad_norm": 0.22063737426257915,
+      "learning_rate": 1.6678790744015238e-08,
+      "loss": 0.0427,
+      "step": 5698
+    },
+    {
+      "epoch": 2.9666840187402395,
+      "grad_norm": 0.22278000386951152,
+      "learning_rate": 1.616959972800536e-08,
+      "loss": 0.0429,
+      "step": 5699
+    },
+    {
+      "epoch": 2.9672045809474232,
+      "grad_norm": 0.2207802651418027,
+      "learning_rate": 1.5668299768284434e-08,
+      "loss": 0.0415,
+      "step": 5700
+    },
+    {
+      "epoch": 2.967725143154607,
+      "grad_norm": 0.22385079884702969,
+      "learning_rate": 1.5174891023184146e-08,
+      "loss": 0.044,
+      "step": 5701
+    },
+    {
+      "epoch": 2.9682457053617908,
+      "grad_norm": 0.2247189896311398,
+      "learning_rate": 1.4689373648549277e-08,
+      "loss": 0.0434,
+      "step": 5702
+    },
+    {
+      "epoch": 2.9687662675689745,
+      "grad_norm": 0.2352896279762599,
+      "learning_rate": 1.421174779772383e-08,
+      "loss": 0.0438,
+      "step": 5703
+    },
+    {
+      "epoch": 2.9692868297761583,
+      "grad_norm": 0.21791843470524855,
+      "learning_rate": 1.3742013621564909e-08,
+      "loss": 0.0411,
+      "step": 5704
+    },
+    {
+      "epoch": 2.969807391983342,
+      "grad_norm": 0.21979737042742506,
+      "learning_rate": 1.3280171268442721e-08,
+      "loss": 0.0421,
+      "step": 5705
+    },
+    {
+      "epoch": 2.970327954190526,
+      "grad_norm": 0.22711419377584297,
+      "learning_rate": 1.282622088422114e-08,
+      "loss": 0.0422,
+      "step": 5706
+    },
+    {
+      "epoch": 2.9708485163977096,
+      "grad_norm": 0.2214357911141881,
+      "learning_rate": 1.2380162612282697e-08,
+      "loss": 0.0423,
+      "step": 5707
+    },
+    {
+      "epoch": 2.9713690786048934,
+      "grad_norm": 0.23079498217049907,
+      "learning_rate": 1.1941996593514693e-08,
+      "loss": 0.044,
+      "step": 5708
+    },
+    {
+      "epoch": 2.971889640812077,
+      "grad_norm": 0.2328227144923505,
+      "learning_rate": 1.1511722966306426e-08,
+      "loss": 0.0442,
+      "step": 5709
+    },
+    {
+      "epoch": 2.972410203019261,
+      "grad_norm": 0.23439895371290795,
+      "learning_rate": 1.10893418665603e-08,
+      "loss": 0.043,
+      "step": 5710
+    },
+    {
+      "epoch": 2.9729307652264447,
+      "grad_norm": 0.22428047997601785,
+      "learning_rate": 1.0674853427683484e-08,
+      "loss": 0.0444,
+      "step": 5711
+    },
+    {
+      "epoch": 2.9734513274336285,
+      "grad_norm": 0.22842869772809338,
+      "learning_rate": 1.0268257780590707e-08,
+      "loss": 0.0417,
+      "step": 5712
+    },
+    {
+      "epoch": 2.9739718896408123,
+      "grad_norm": 0.21987057018297476,
+      "learning_rate": 9.869555053704239e-09,
+      "loss": 0.0413,
+      "step": 5713
+    },
+    {
+      "epoch": 2.974492451847996,
+      "grad_norm": 0.2298608691054948,
+      "learning_rate": 9.4787453729539e-09,
+      "loss": 0.0437,
+      "step": 5714
+    },
+    {
+      "epoch": 2.97501301405518,
+      "grad_norm": 0.2253370055551647,
+      "learning_rate": 9.095828861771516e-09,
+      "loss": 0.0432,
+      "step": 5715
+    },
+    {
+      "epoch": 2.9755335762623636,
+      "grad_norm": 0.2278149249962827,
+      "learning_rate": 8.720805641104779e-09,
+      "loss": 0.0434,
+      "step": 5716
+    },
+    {
+      "epoch": 2.9760541384695474,
+      "grad_norm": 0.2314267428074255,
+      "learning_rate": 8.353675829403385e-09,
+      "loss": 0.0443,
+      "step": 5717
+    },
+    {
+      "epoch": 2.976574700676731,
+      "grad_norm": 0.22400674704669293,
+      "learning_rate": 7.994439542619025e-09,
+      "loss": 0.0425,
+      "step": 5718
+    },
+    {
+      "epoch": 2.9770952628839145,
+      "grad_norm": 0.22281537791883546,
+      "learning_rate": 7.643096894222046e-09,
+      "loss": 0.0427,
+      "step": 5719
+    },
+    {
+      "epoch": 2.9776158250910982,
+      "grad_norm": 0.22059637224805248,
+      "learning_rate": 7.299647995176462e-09,
+      "loss": 0.0427,
+      "step": 5720
+    },
+    {
+      "epoch": 2.978136387298282,
+      "grad_norm": 0.22754394710380124,
+      "learning_rate": 6.964092953962165e-09,
+      "loss": 0.0436,
+      "step": 5721
+    },
+    {
+      "epoch": 2.978656949505466,
+      "grad_norm": 0.23398045942704954,
+      "learning_rate": 6.63643187656382e-09,
+      "loss": 0.0447,
+      "step": 5722
+    },
+    {
+      "epoch": 2.9791775117126496,
+      "grad_norm": 0.22256790802648144,
+      "learning_rate": 6.316664866470867e-09,
+      "loss": 0.0434,
+      "step": 5723
+    },
+    {
+      "epoch": 2.9796980739198333,
+      "grad_norm": 0.22504707512425381,
+      "learning_rate": 6.004792024680295e-09,
+      "loss": 0.0433,
+      "step": 5724
+    },
+    {
+      "epoch": 2.980218636127017,
+      "grad_norm": 0.2172897922400504,
+      "learning_rate": 5.700813449699416e-09,
+      "loss": 0.043,
+      "step": 5725
+    },
+    {
+      "epoch": 2.980739198334201,
+      "grad_norm": 0.23460775231227962,
+      "learning_rate": 5.404729237531991e-09,
+      "loss": 0.045,
+      "step": 5726
+    },
+    {
+      "epoch": 2.9812597605413846,
+      "grad_norm": 0.2093848412957586,
+      "learning_rate": 5.116539481703208e-09,
+      "loss": 0.0402,
+      "step": 5727
+    },
+    {
+      "epoch": 2.9817803227485684,
+      "grad_norm": 0.22035954463083757,
+      "learning_rate": 4.836244273231927e-09,
+      "loss": 0.0423,
+      "step": 5728
+    },
+    {
+      "epoch": 2.982300884955752,
+      "grad_norm": 0.24018779005516844,
+      "learning_rate": 4.56384370064733e-09,
+      "loss": 0.0433,
+      "step": 5729
+    },
+    {
+      "epoch": 2.982821447162936,
+      "grad_norm": 0.22737460452274955,
+      "learning_rate": 4.299337849991703e-09,
+      "loss": 0.0441,
+      "step": 5730
+    },
+    {
+      "epoch": 2.9833420093701197,
+      "grad_norm": 0.219433427420989,
+      "learning_rate": 4.042726804801e-09,
+      "loss": 0.0412,
+      "step": 5731
+    },
+    {
+      "epoch": 2.9838625715773035,
+      "grad_norm": 0.22548984426888996,
+      "learning_rate": 3.794010646132606e-09,
+      "loss": 0.0428,
+      "step": 5732
+    },
+    {
+      "epoch": 2.9843831337844873,
+      "grad_norm": 0.2231171978367076,
+      "learning_rate": 3.553189452537575e-09,
+      "loss": 0.0435,
+      "step": 5733
+    },
+    {
+      "epoch": 2.984903695991671,
+      "grad_norm": 0.2261689918270949,
+      "learning_rate": 3.3202633000772865e-09,
+      "loss": 0.0423,
+      "step": 5734
+    },
+    {
+      "epoch": 2.985424258198855,
+      "grad_norm": 0.24047982640447405,
+      "learning_rate": 3.0952322623262197e-09,
+      "loss": 0.0442,
+      "step": 5735
+    },
+    {
+      "epoch": 2.9859448204060386,
+      "grad_norm": 0.2198384025823191,
+      "learning_rate": 2.878096410355302e-09,
+      "loss": 0.0421,
+      "step": 5736
+    },
+    {
+      "epoch": 2.9864653826132224,
+      "grad_norm": 0.2205868075205334,
+      "learning_rate": 2.6688558127485607e-09,
+      "loss": 0.0414,
+      "step": 5737
+    },
+    {
+      "epoch": 2.986985944820406,
+      "grad_norm": 0.23578679799030974,
+      "learning_rate": 2.4675105355920213e-09,
+      "loss": 0.0445,
+      "step": 5738
+    },
+    {
+      "epoch": 2.98750650702759,
+      "grad_norm": 0.22520079740469365,
+      "learning_rate": 2.2740606424792587e-09,
+      "loss": 0.0416,
+      "step": 5739
+    },
+    {
+      "epoch": 2.9880270692347737,
+      "grad_norm": 0.23337576964018655,
+      "learning_rate": 2.088506194514173e-09,
+      "loss": 0.043,
+      "step": 5740
+    },
+    {
+      "epoch": 2.9885476314419575,
+      "grad_norm": 0.23186028702019476,
+      "learning_rate": 1.9108472502998855e-09,
+      "loss": 0.0432,
+      "step": 5741
+    },
+    {
+      "epoch": 2.9890681936491412,
+      "grad_norm": 0.2184071173157055,
+      "learning_rate": 1.7410838659498442e-09,
+      "loss": 0.0418,
+      "step": 5742
+    },
+    {
+      "epoch": 2.9895887558563246,
+      "grad_norm": 0.2199140316024418,
+      "learning_rate": 1.579216095087821e-09,
+      "loss": 0.0428,
+      "step": 5743
+    },
+    {
+      "epoch": 2.9901093180635083,
+      "grad_norm": 0.22353723184238114,
+      "learning_rate": 1.4252439888340353e-09,
+      "loss": 0.0435,
+      "step": 5744
+    },
+    {
+      "epoch": 2.990629880270692,
+      "grad_norm": 0.22673756486431215,
+      "learning_rate": 1.2791675958218064e-09,
+      "loss": 0.0439,
+      "step": 5745
+    },
+    {
+      "epoch": 2.991150442477876,
+      "grad_norm": 0.22545208764365354,
+      "learning_rate": 1.140986962186452e-09,
+      "loss": 0.0438,
+      "step": 5746
+    },
+    {
+      "epoch": 2.9916710046850596,
+      "grad_norm": 0.23688504795341422,
+      "learning_rate": 1.010702131576391e-09,
+      "loss": 0.0442,
+      "step": 5747
+    },
+    {
+      "epoch": 2.9921915668922434,
+      "grad_norm": 0.23253053657185024,
+      "learning_rate": 8.883131451392635e-10,
+      "loss": 0.044,
+      "step": 5748
+    },
+    {
+      "epoch": 2.992712129099427,
+      "grad_norm": 0.2379163810816669,
+      "learning_rate": 7.738200415302599e-10,
+      "loss": 0.0452,
+      "step": 5749
+    },
+    {
+      "epoch": 2.993232691306611,
+      "grad_norm": 0.21509722288903138,
+      "learning_rate": 6.672228569148953e-10,
+      "loss": 0.0425,
+      "step": 5750
+    },
+    {
+      "epoch": 2.9937532535137947,
+      "grad_norm": 0.22479935136282975,
+      "learning_rate": 5.685216249579073e-10,
+      "loss": 0.0421,
+      "step": 5751
+    },
+    {
+      "epoch": 2.9942738157209785,
+      "grad_norm": 0.2329527890243757,
+      "learning_rate": 4.777163768343585e-10,
+      "loss": 0.0431,
+      "step": 5752
+    },
+    {
+      "epoch": 2.9947943779281623,
+      "grad_norm": 0.21818439935194653,
+      "learning_rate": 3.94807141224085e-10,
+      "loss": 0.0418,
+      "step": 5753
+    },
+    {
+      "epoch": 2.995314940135346,
+      "grad_norm": 0.23421104271035925,
+      "learning_rate": 3.197939443172482e-10,
+      "loss": 0.0437,
+      "step": 5754
+    },
+    {
+      "epoch": 2.99583550234253,
+      "grad_norm": 0.21714292300858615,
+      "learning_rate": 2.526768098060073e-10,
+      "loss": 0.0426,
+      "step": 5755
+    },
+    {
+      "epoch": 2.9963560645497136,
+      "grad_norm": 0.23045424040900458,
+      "learning_rate": 1.9345575888451983e-10,
+      "loss": 0.0436,
+      "step": 5756
+    },
+    {
+      "epoch": 2.9968766267568974,
+      "grad_norm": 0.22483821530138692,
+      "learning_rate": 1.421308102628194e-10,
+      "loss": 0.0433,
+      "step": 5757
+    },
+    {
+      "epoch": 2.997397188964081,
+      "grad_norm": 0.22259646441449232,
+      "learning_rate": 9.870198014738652e-11,
+      "loss": 0.0449,
+      "step": 5758
+    },
+    {
+      "epoch": 2.997917751171265,
+      "grad_norm": 0.22244406662721328,
+      "learning_rate": 6.316928225780228e-11,
+      "loss": 0.0437,
+      "step": 5759
+    },
+    {
+      "epoch": 2.9984383133784487,
+      "grad_norm": 0.22721923240497144,
+      "learning_rate": 3.553272781842143e-11,
+      "loss": 0.0434,
+      "step": 5760
+    },
+    {
+      "epoch": 2.9989588755856325,
+      "grad_norm": 0.22864277231584557,
+      "learning_rate": 1.5792325552821398e-11,
+      "loss": 0.0435,
+      "step": 5761
+    },
+    {
+      "epoch": 2.9994794377928162,
+      "grad_norm": 0.22668136835192348,
+      "learning_rate": 3.948081700455575e-12,
+      "loss": 0.0439,
+      "step": 5762
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.21769778535390255,
+      "learning_rate": 0.0,
+      "loss": 0.042,
+      "step": 5763
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 5763,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 240,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 2413100531712000.0,
+  "train_batch_size": 4,
+  "trial_name": null,
+  "trial_params": null
+}