diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,8358 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.9995792131285504,
+  "eval_steps": 500,
+  "global_step": 1188,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.001683147485798443,
+      "grad_norm": 0.17560942471027374,
+      "learning_rate": 0.0,
+      "loss": 2.613,
+      "step": 1
+    },
+    {
+      "epoch": 0.003366294971596886,
+      "grad_norm": 0.15861666202545166,
+      "learning_rate": 2.7894294565112984e-06,
+      "loss": 2.6655,
+      "step": 2
+    },
+    {
+      "epoch": 0.005049442457395329,
+      "grad_norm": 0.1817302405834198,
+      "learning_rate": 4.421141086977404e-06,
+      "loss": 2.55,
+      "step": 3
+    },
+    {
+      "epoch": 0.006732589943193772,
+      "grad_norm": 0.17854492366313934,
+      "learning_rate": 5.578858913022597e-06,
+      "loss": 2.7908,
+      "step": 4
+    },
+    {
+      "epoch": 0.008415737428992216,
+      "grad_norm": 0.17169038951396942,
+      "learning_rate": 6.47685462377997e-06,
+      "loss": 2.6868,
+      "step": 5
+    },
+    {
+      "epoch": 0.010098884914790659,
+      "grad_norm": 0.18368647992610931,
+      "learning_rate": 7.210570543488702e-06,
+      "loss": 2.5874,
+      "step": 6
+    },
+    {
+      "epoch": 0.011782032400589101,
+      "grad_norm": 0.19648714363574982,
+      "learning_rate": 7.830918514469461e-06,
+      "loss": 2.6633,
+      "step": 7
+    },
+    {
+      "epoch": 0.013465179886387544,
+      "grad_norm": 0.18358571827411652,
+      "learning_rate": 8.368288369533896e-06,
+      "loss": 2.6355,
+      "step": 8
+    },
+    {
+      "epoch": 0.015148327372185988,
+      "grad_norm": 0.19153611361980438,
+      "learning_rate": 8.842282173954808e-06,
+      "loss": 2.6633,
+      "step": 9
+    },
+    {
+      "epoch": 0.016831474857984433,
+      "grad_norm": 0.20646820962429047,
+      "learning_rate": 9.26628408029127e-06,
+      "loss": 2.7268,
+      "step": 10
+    },
+    {
+      "epoch": 0.018514622343782875,
+      "grad_norm": 0.18688935041427612,
+      "learning_rate": 9.64984045981344e-06,
+      "loss": 2.7832,
+      "step": 11
+    },
+    {
+      "epoch": 0.020197769829581318,
+      "grad_norm": 0.1985747218132019,
+      "learning_rate": 1e-05,
+      "loss": 2.738,
+      "step": 12
+    },
+    {
+      "epoch": 0.02188091731537976,
+      "grad_norm": 0.19321100413799286,
+      "learning_rate": 1e-05,
+      "loss": 2.6206,
+      "step": 13
+    },
+    {
+      "epoch": 0.023564064801178203,
+      "grad_norm": 0.1875382661819458,
+      "learning_rate": 1e-05,
+      "loss": 2.7153,
+      "step": 14
+    },
+    {
+      "epoch": 0.025247212286976645,
+      "grad_norm": 0.18803201615810394,
+      "learning_rate": 1e-05,
+      "loss": 2.5359,
+      "step": 15
+    },
+    {
+      "epoch": 0.026930359772775088,
+      "grad_norm": 0.19693922996520996,
+      "learning_rate": 1e-05,
+      "loss": 2.6082,
+      "step": 16
+    },
+    {
+      "epoch": 0.028613507258573534,
+      "grad_norm": 0.20534300804138184,
+      "learning_rate": 1e-05,
+      "loss": 2.5317,
+      "step": 17
+    },
+    {
+      "epoch": 0.030296654744371977,
+      "grad_norm": 0.22174465656280518,
+      "learning_rate": 1e-05,
+      "loss": 2.6067,
+      "step": 18
+    },
+    {
+      "epoch": 0.03197980223017042,
+      "grad_norm": 0.1947612464427948,
+      "learning_rate": 1e-05,
+      "loss": 2.6824,
+      "step": 19
+    },
+    {
+      "epoch": 0.033662949715968865,
+      "grad_norm": 0.19715926051139832,
+      "learning_rate": 1e-05,
+      "loss": 2.6868,
+      "step": 20
+    },
+    {
+      "epoch": 0.035346097201767304,
+      "grad_norm": 0.19586338102817535,
+      "learning_rate": 1e-05,
+      "loss": 2.6206,
+      "step": 21
+    },
+    {
+      "epoch": 0.03702924468756575,
+      "grad_norm": 0.19280074536800385,
+      "learning_rate": 1e-05,
+      "loss": 2.6023,
+      "step": 22
+    },
+    {
+      "epoch": 0.03871239217336419,
+      "grad_norm": 0.19658198952674866,
+      "learning_rate": 1e-05,
+      "loss": 2.6384,
+      "step": 23
+    },
+    {
+      "epoch": 0.040395539659162635,
+      "grad_norm": 0.17433768510818481,
+      "learning_rate": 1e-05,
+      "loss": 2.5305,
+      "step": 24
+    },
+    {
+      "epoch": 0.042078687144961074,
+      "grad_norm": 0.18013380467891693,
+      "learning_rate": 1e-05,
+      "loss": 2.6519,
+      "step": 25
+    },
+    {
+      "epoch": 0.04376183463075952,
+      "grad_norm": 0.1933555006980896,
+      "learning_rate": 1e-05,
+      "loss": 2.5591,
+      "step": 26
+    },
+    {
+      "epoch": 0.045444982116557966,
+      "grad_norm": 0.18386027216911316,
+      "learning_rate": 1e-05,
+      "loss": 2.6169,
+      "step": 27
+    },
+    {
+      "epoch": 0.047128129602356406,
+      "grad_norm": 0.18173415958881378,
+      "learning_rate": 1e-05,
+      "loss": 2.623,
+      "step": 28
+    },
+    {
+      "epoch": 0.04881127708815485,
+      "grad_norm": 0.19154761731624603,
+      "learning_rate": 1e-05,
+      "loss": 2.5981,
+      "step": 29
+    },
+    {
+      "epoch": 0.05049442457395329,
+      "grad_norm": 0.2001664638519287,
+      "learning_rate": 1e-05,
+      "loss": 2.5066,
+      "step": 30
+    },
+    {
+      "epoch": 0.05217757205975174,
+      "grad_norm": 0.15573543310165405,
+      "learning_rate": 1e-05,
+      "loss": 2.6013,
+      "step": 31
+    },
+    {
+      "epoch": 0.053860719545550176,
+      "grad_norm": 0.16071979701519012,
+      "learning_rate": 1e-05,
+      "loss": 2.4634,
+      "step": 32
+    },
+    {
+      "epoch": 0.05554386703134862,
+      "grad_norm": 0.1769736260175705,
+      "learning_rate": 1e-05,
+      "loss": 2.5491,
+      "step": 33
+    },
+    {
+      "epoch": 0.05722701451714707,
+      "grad_norm": 0.17623937129974365,
+      "learning_rate": 1e-05,
+      "loss": 2.4399,
+      "step": 34
+    },
+    {
+      "epoch": 0.05891016200294551,
+      "grad_norm": 0.17367449402809143,
+      "learning_rate": 1e-05,
+      "loss": 2.5464,
+      "step": 35
+    },
+    {
+      "epoch": 0.06059330948874395,
+      "grad_norm": 0.14842955768108368,
+      "learning_rate": 1e-05,
+      "loss": 2.4174,
+      "step": 36
+    },
+    {
+      "epoch": 0.06227645697454239,
+      "grad_norm": 0.17405100166797638,
+      "learning_rate": 1e-05,
+      "loss": 2.5303,
+      "step": 37
+    },
+    {
+      "epoch": 0.06395960446034084,
+      "grad_norm": 0.145203098654747,
+      "learning_rate": 1e-05,
+      "loss": 2.6428,
+      "step": 38
+    },
+    {
+      "epoch": 0.06564275194613928,
+      "grad_norm": 0.1542726755142212,
+      "learning_rate": 1e-05,
+      "loss": 2.5618,
+      "step": 39
+    },
+    {
+      "epoch": 0.06732589943193773,
+      "grad_norm": 0.14489781856536865,
+      "learning_rate": 1e-05,
+      "loss": 2.6885,
+      "step": 40
+    },
+    {
+      "epoch": 0.06900904691773617,
+      "grad_norm": 0.14798486232757568,
+      "learning_rate": 1e-05,
+      "loss": 2.5322,
+      "step": 41
+    },
+    {
+      "epoch": 0.07069219440353461,
+      "grad_norm": 0.15226829051971436,
+      "learning_rate": 1e-05,
+      "loss": 2.6011,
+      "step": 42
+    },
+    {
+      "epoch": 0.07237534188933305,
+      "grad_norm": 0.14561522006988525,
+      "learning_rate": 1e-05,
+      "loss": 2.5657,
+      "step": 43
+    },
+    {
+      "epoch": 0.0740584893751315,
+      "grad_norm": 0.13787826895713806,
+      "learning_rate": 1e-05,
+      "loss": 2.6011,
+      "step": 44
+    },
+    {
+      "epoch": 0.07574163686092994,
+      "grad_norm": 0.14005698263645172,
+      "learning_rate": 1e-05,
+      "loss": 2.4673,
+      "step": 45
+    },
+    {
+      "epoch": 0.07742478434672838,
+      "grad_norm": 0.13822345435619354,
+      "learning_rate": 1e-05,
+      "loss": 2.512,
+      "step": 46
+    },
+    {
+      "epoch": 0.07910793183252683,
+      "grad_norm": 0.1284177154302597,
+      "learning_rate": 1e-05,
+      "loss": 2.5625,
+      "step": 47
+    },
+    {
+      "epoch": 0.08079107931832527,
+      "grad_norm": 0.1279960423707962,
+      "learning_rate": 1e-05,
+      "loss": 2.46,
+      "step": 48
+    },
+    {
+      "epoch": 0.08247422680412371,
+      "grad_norm": 0.12479826807975769,
+      "learning_rate": 1e-05,
+      "loss": 2.5706,
+      "step": 49
+    },
+    {
+      "epoch": 0.08415737428992215,
+      "grad_norm": 0.12982836365699768,
+      "learning_rate": 1e-05,
+      "loss": 2.5098,
+      "step": 50
+    },
+    {
+      "epoch": 0.0858405217757206,
+      "grad_norm": 0.13269256055355072,
+      "learning_rate": 1e-05,
+      "loss": 2.4688,
+      "step": 51
+    },
+    {
+      "epoch": 0.08752366926151904,
+      "grad_norm": 0.11713477969169617,
+      "learning_rate": 1e-05,
+      "loss": 2.6226,
+      "step": 52
+    },
+    {
+      "epoch": 0.08920681674731748,
+      "grad_norm": 0.11179152131080627,
+      "learning_rate": 1e-05,
+      "loss": 2.4224,
+      "step": 53
+    },
+    {
+      "epoch": 0.09088996423311593,
+      "grad_norm": 0.12146276980638504,
+      "learning_rate": 1e-05,
+      "loss": 2.4639,
+      "step": 54
+    },
+    {
+      "epoch": 0.09257311171891437,
+      "grad_norm": 0.12470445781946182,
+      "learning_rate": 1e-05,
+      "loss": 2.5195,
+      "step": 55
+    },
+    {
+      "epoch": 0.09425625920471281,
+      "grad_norm": 0.11872275173664093,
+      "learning_rate": 1e-05,
+      "loss": 2.5186,
+      "step": 56
+    },
+    {
+      "epoch": 0.09593940669051125,
+      "grad_norm": 0.11616484075784683,
+      "learning_rate": 1e-05,
+      "loss": 2.5581,
+      "step": 57
+    },
+    {
+      "epoch": 0.0976225541763097,
+      "grad_norm": 0.1075875386595726,
+      "learning_rate": 1e-05,
+      "loss": 2.5693,
+      "step": 58
+    },
+    {
+      "epoch": 0.09930570166210814,
+      "grad_norm": 0.10176095366477966,
+      "learning_rate": 1e-05,
+      "loss": 2.521,
+      "step": 59
+    },
+    {
+      "epoch": 0.10098884914790658,
+      "grad_norm": 0.1076890155673027,
+      "learning_rate": 1e-05,
+      "loss": 2.53,
+      "step": 60
+    },
+    {
+      "epoch": 0.10267199663370503,
+      "grad_norm": 0.09105601906776428,
+      "learning_rate": 1e-05,
+      "loss": 2.3733,
+      "step": 61
+    },
+    {
+      "epoch": 0.10435514411950347,
+      "grad_norm": 0.09733142703771591,
+      "learning_rate": 1e-05,
+      "loss": 2.416,
+      "step": 62
+    },
+    {
+      "epoch": 0.10603829160530191,
+      "grad_norm": 0.09099874645471573,
+      "learning_rate": 1e-05,
+      "loss": 2.3774,
+      "step": 63
+    },
+    {
+      "epoch": 0.10772143909110035,
+      "grad_norm": 0.0884426161646843,
+      "learning_rate": 1e-05,
+      "loss": 2.4136,
+      "step": 64
+    },
+    {
+      "epoch": 0.1094045865768988,
+      "grad_norm": 0.08939989656209946,
+      "learning_rate": 1e-05,
+      "loss": 2.4482,
+      "step": 65
+    },
+    {
+      "epoch": 0.11108773406269724,
+      "grad_norm": 0.09078355878591537,
+      "learning_rate": 1e-05,
+      "loss": 2.5256,
+      "step": 66
+    },
+    {
+      "epoch": 0.11277088154849568,
+      "grad_norm": 0.08570227026939392,
+      "learning_rate": 1e-05,
+      "loss": 2.4954,
+      "step": 67
+    },
+    {
+      "epoch": 0.11445402903429414,
+      "grad_norm": 0.0766797736287117,
+      "learning_rate": 1e-05,
+      "loss": 2.3694,
+      "step": 68
+    },
+    {
+      "epoch": 0.11613717652009257,
+      "grad_norm": 0.08015618473291397,
+      "learning_rate": 1e-05,
+      "loss": 2.4724,
+      "step": 69
+    },
+    {
+      "epoch": 0.11782032400589101,
+      "grad_norm": 0.08956343680620193,
+      "learning_rate": 1e-05,
+      "loss": 2.47,
+      "step": 70
+    },
+    {
+      "epoch": 0.11950347149168945,
+      "grad_norm": 0.08134786039590836,
+      "learning_rate": 1e-05,
+      "loss": 2.4482,
+      "step": 71
+    },
+    {
+      "epoch": 0.1211866189774879,
+      "grad_norm": 0.07923366874456406,
+      "learning_rate": 1e-05,
+      "loss": 2.4182,
+      "step": 72
+    },
+    {
+      "epoch": 0.12286976646328635,
+      "grad_norm": 0.07909434288740158,
+      "learning_rate": 1e-05,
+      "loss": 2.3711,
+      "step": 73
+    },
+    {
+      "epoch": 0.12455291394908478,
+      "grad_norm": 0.07540368288755417,
+      "learning_rate": 1e-05,
+      "loss": 2.3962,
+      "step": 74
+    },
+    {
+      "epoch": 0.12623606143488322,
+      "grad_norm": 0.06906846165657043,
+      "learning_rate": 1e-05,
+      "loss": 2.519,
+      "step": 75
+    },
+    {
+      "epoch": 0.12791920892068168,
+      "grad_norm": 0.07301697880029678,
+      "learning_rate": 1e-05,
+      "loss": 2.5537,
+      "step": 76
+    },
+    {
+      "epoch": 0.12960235640648013,
+      "grad_norm": 0.07182423770427704,
+      "learning_rate": 1e-05,
+      "loss": 2.4807,
+      "step": 77
+    },
+    {
+      "epoch": 0.13128550389227855,
+      "grad_norm": 0.06827539950609207,
+      "learning_rate": 1e-05,
+      "loss": 2.5796,
+      "step": 78
+    },
+    {
+      "epoch": 0.132968651378077,
+      "grad_norm": 0.07280007749795914,
+      "learning_rate": 1e-05,
+      "loss": 2.499,
+      "step": 79
+    },
+    {
+      "epoch": 0.13465179886387546,
+      "grad_norm": 0.07410164177417755,
+      "learning_rate": 1e-05,
+      "loss": 2.3418,
+      "step": 80
+    },
+    {
+      "epoch": 0.13633494634967389,
+      "grad_norm": 0.07245635986328125,
+      "learning_rate": 1e-05,
+      "loss": 2.4685,
+      "step": 81
+    },
+    {
+      "epoch": 0.13801809383547234,
+      "grad_norm": 0.06992876529693604,
+      "learning_rate": 1e-05,
+      "loss": 2.4634,
+      "step": 82
+    },
+    {
+      "epoch": 0.13970124132127076,
+      "grad_norm": 0.07322832196950912,
+      "learning_rate": 1e-05,
+      "loss": 2.4949,
+      "step": 83
+    },
+    {
+      "epoch": 0.14138438880706922,
+      "grad_norm": 0.06528163701295853,
+      "learning_rate": 1e-05,
+      "loss": 2.3982,
+      "step": 84
+    },
+    {
+      "epoch": 0.14306753629286767,
+      "grad_norm": 0.06972632557153702,
+      "learning_rate": 1e-05,
+      "loss": 2.4268,
+      "step": 85
+    },
+    {
+      "epoch": 0.1447506837786661,
+      "grad_norm": 0.062493499368429184,
+      "learning_rate": 1e-05,
+      "loss": 2.4309,
+      "step": 86
+    },
+    {
+      "epoch": 0.14643383126446455,
+      "grad_norm": 0.07086165249347687,
+      "learning_rate": 1e-05,
+      "loss": 2.4373,
+      "step": 87
+    },
+    {
+      "epoch": 0.148116978750263,
+      "grad_norm": 0.06631726026535034,
+      "learning_rate": 1e-05,
+      "loss": 2.4141,
+      "step": 88
+    },
+    {
+      "epoch": 0.14980012623606143,
+      "grad_norm": 0.07114582508802414,
+      "learning_rate": 1e-05,
+      "loss": 2.3546,
+      "step": 89
+    },
+    {
+      "epoch": 0.15148327372185988,
+      "grad_norm": 0.06932078301906586,
+      "learning_rate": 1e-05,
+      "loss": 2.4758,
+      "step": 90
+    },
+    {
+      "epoch": 0.15316642120765833,
+      "grad_norm": 0.06153389438986778,
+      "learning_rate": 1e-05,
+      "loss": 2.481,
+      "step": 91
+    },
+    {
+      "epoch": 0.15484956869345676,
+      "grad_norm": 0.06216192990541458,
+      "learning_rate": 1e-05,
+      "loss": 2.4421,
+      "step": 92
+    },
+    {
+      "epoch": 0.1565327161792552,
+      "grad_norm": 0.06554314494132996,
+      "learning_rate": 1e-05,
+      "loss": 2.3008,
+      "step": 93
+    },
+    {
+      "epoch": 0.15821586366505366,
+      "grad_norm": 0.06210967153310776,
+      "learning_rate": 1e-05,
+      "loss": 2.2554,
+      "step": 94
+    },
+    {
+      "epoch": 0.1598990111508521,
+      "grad_norm": 0.06851295381784439,
+      "learning_rate": 1e-05,
+      "loss": 2.5356,
+      "step": 95
+    },
+    {
+      "epoch": 0.16158215863665054,
+      "grad_norm": 0.06121644005179405,
+      "learning_rate": 1e-05,
+      "loss": 2.4299,
+      "step": 96
+    },
+    {
+      "epoch": 0.16326530612244897,
+      "grad_norm": 0.06593657284975052,
+      "learning_rate": 1e-05,
+      "loss": 2.3811,
+      "step": 97
+    },
+    {
+      "epoch": 0.16494845360824742,
+      "grad_norm": 0.06456276774406433,
+      "learning_rate": 1e-05,
+      "loss": 2.3574,
+      "step": 98
+    },
+    {
+      "epoch": 0.16663160109404587,
+      "grad_norm": 0.061866894364356995,
+      "learning_rate": 1e-05,
+      "loss": 2.4758,
+      "step": 99
+    },
+    {
+      "epoch": 0.1683147485798443,
+      "grad_norm": 0.058500371873378754,
+      "learning_rate": 1e-05,
+      "loss": 2.4133,
+      "step": 100
+    },
+    {
+      "epoch": 0.16999789606564275,
+      "grad_norm": 0.06366603821516037,
+      "learning_rate": 1e-05,
+      "loss": 2.3328,
+      "step": 101
+    },
+    {
+      "epoch": 0.1716810435514412,
+      "grad_norm": 0.061924271285533905,
+      "learning_rate": 1e-05,
+      "loss": 2.4047,
+      "step": 102
+    },
+    {
+      "epoch": 0.17336419103723963,
+      "grad_norm": 0.057471342384815216,
+      "learning_rate": 1e-05,
+      "loss": 2.4333,
+      "step": 103
+    },
+    {
+      "epoch": 0.17504733852303808,
+      "grad_norm": 0.05482906475663185,
+      "learning_rate": 1e-05,
+      "loss": 2.3499,
+      "step": 104
+    },
+    {
+      "epoch": 0.17673048600883653,
+      "grad_norm": 0.056116051971912384,
+      "learning_rate": 1e-05,
+      "loss": 2.4653,
+      "step": 105
+    },
+    {
+      "epoch": 0.17841363349463496,
+      "grad_norm": 0.052277661859989166,
+      "learning_rate": 1e-05,
+      "loss": 2.4653,
+      "step": 106
+    },
+    {
+      "epoch": 0.1800967809804334,
+      "grad_norm": 0.06346592307090759,
+      "learning_rate": 1e-05,
+      "loss": 2.3549,
+      "step": 107
+    },
+    {
+      "epoch": 0.18177992846623187,
+      "grad_norm": 0.06070290133357048,
+      "learning_rate": 1e-05,
+      "loss": 2.2886,
+      "step": 108
+    },
+    {
+      "epoch": 0.1834630759520303,
+      "grad_norm": 0.055994004011154175,
+      "learning_rate": 1e-05,
+      "loss": 2.4692,
+      "step": 109
+    },
+    {
+      "epoch": 0.18514622343782874,
+      "grad_norm": 0.05782800912857056,
+      "learning_rate": 1e-05,
+      "loss": 2.3303,
+      "step": 110
+    },
+    {
+      "epoch": 0.18682937092362717,
+      "grad_norm": 0.05491410568356514,
+      "learning_rate": 1e-05,
+      "loss": 2.47,
+      "step": 111
+    },
+    {
+      "epoch": 0.18851251840942562,
+      "grad_norm": 0.060252465307712555,
+      "learning_rate": 1e-05,
+      "loss": 2.5464,
+      "step": 112
+    },
+    {
+      "epoch": 0.19019566589522408,
+      "grad_norm": 0.05614893510937691,
+      "learning_rate": 1e-05,
+      "loss": 2.3457,
+      "step": 113
+    },
+    {
+      "epoch": 0.1918788133810225,
+      "grad_norm": 0.051146939396858215,
+      "learning_rate": 1e-05,
+      "loss": 2.3918,
+      "step": 114
+    },
+    {
+      "epoch": 0.19356196086682095,
+      "grad_norm": 0.05474052205681801,
+      "learning_rate": 1e-05,
+      "loss": 2.3689,
+      "step": 115
+    },
+    {
+      "epoch": 0.1952451083526194,
+      "grad_norm": 0.052064936608076096,
+      "learning_rate": 1e-05,
+      "loss": 2.5073,
+      "step": 116
+    },
+    {
+      "epoch": 0.19692825583841783,
+      "grad_norm": 0.06184034049510956,
+      "learning_rate": 1e-05,
+      "loss": 2.4248,
+      "step": 117
+    },
+    {
+      "epoch": 0.19861140332421628,
+      "grad_norm": 0.05613533779978752,
+      "learning_rate": 1e-05,
+      "loss": 2.5742,
+      "step": 118
+    },
+    {
+      "epoch": 0.20029455081001474,
+      "grad_norm": 0.05547456443309784,
+      "learning_rate": 1e-05,
+      "loss": 2.3884,
+      "step": 119
+    },
+    {
+      "epoch": 0.20197769829581316,
+      "grad_norm": 0.05933033674955368,
+      "learning_rate": 1e-05,
+      "loss": 2.45,
+      "step": 120
+    },
+    {
+      "epoch": 0.20366084578161162,
+      "grad_norm": 0.058600571006536484,
+      "learning_rate": 1e-05,
+      "loss": 2.3875,
+      "step": 121
+    },
+    {
+      "epoch": 0.20534399326741007,
+      "grad_norm": 0.0554657019674778,
+      "learning_rate": 1e-05,
+      "loss": 2.3215,
+      "step": 122
+    },
+    {
+      "epoch": 0.2070271407532085,
+      "grad_norm": 0.05604475364089012,
+      "learning_rate": 1e-05,
+      "loss": 2.3329,
+      "step": 123
+    },
+    {
+      "epoch": 0.20871028823900695,
+      "grad_norm": 0.06094202771782875,
+      "learning_rate": 1e-05,
+      "loss": 2.4177,
+      "step": 124
+    },
+    {
+      "epoch": 0.2103934357248054,
+      "grad_norm": 0.05517999082803726,
+      "learning_rate": 1e-05,
+      "loss": 2.3247,
+      "step": 125
+    },
+    {
+      "epoch": 0.21207658321060383,
+      "grad_norm": 0.05678452178835869,
+      "learning_rate": 1e-05,
+      "loss": 2.3481,
+      "step": 126
+    },
+    {
+      "epoch": 0.21375973069640228,
+      "grad_norm": 0.05295870825648308,
+      "learning_rate": 1e-05,
+      "loss": 2.3694,
+      "step": 127
+    },
+    {
+      "epoch": 0.2154428781822007,
+      "grad_norm": 0.05118125304579735,
+      "learning_rate": 1e-05,
+      "loss": 2.4102,
+      "step": 128
+    },
+    {
+      "epoch": 0.21712602566799916,
+      "grad_norm": 0.05659961327910423,
+      "learning_rate": 1e-05,
+      "loss": 2.3104,
+      "step": 129
+    },
+    {
+      "epoch": 0.2188091731537976,
+      "grad_norm": 0.05049075558781624,
+      "learning_rate": 1e-05,
+      "loss": 2.4949,
+      "step": 130
+    },
+    {
+      "epoch": 0.22049232063959603,
+      "grad_norm": 0.05323097109794617,
+      "learning_rate": 1e-05,
+      "loss": 2.323,
+      "step": 131
+    },
+    {
+      "epoch": 0.2221754681253945,
+      "grad_norm": 0.05309610068798065,
+      "learning_rate": 1e-05,
+      "loss": 2.5203,
+      "step": 132
+    },
+    {
+      "epoch": 0.22385861561119294,
+      "grad_norm": 0.05474167317152023,
+      "learning_rate": 1e-05,
+      "loss": 2.408,
+      "step": 133
+    },
+    {
+      "epoch": 0.22554176309699137,
+      "grad_norm": 0.056433092802762985,
+      "learning_rate": 1e-05,
+      "loss": 2.3779,
+      "step": 134
+    },
+    {
+      "epoch": 0.22722491058278982,
+      "grad_norm": 0.047424182295799255,
+      "learning_rate": 1e-05,
+      "loss": 2.45,
+      "step": 135
+    },
+    {
+      "epoch": 0.22890805806858827,
+      "grad_norm": 0.05422671511769295,
+      "learning_rate": 1e-05,
+      "loss": 2.3397,
+      "step": 136
+    },
+    {
+      "epoch": 0.2305912055543867,
+      "grad_norm": 0.05421329662203789,
+      "learning_rate": 1e-05,
+      "loss": 2.3779,
+      "step": 137
+    },
+    {
+      "epoch": 0.23227435304018515,
+      "grad_norm": 0.057494040578603745,
+      "learning_rate": 1e-05,
+      "loss": 2.4509,
+      "step": 138
+    },
+    {
+      "epoch": 0.2339575005259836,
+      "grad_norm": 0.0516960434615612,
+      "learning_rate": 1e-05,
+      "loss": 2.3647,
+      "step": 139
+    },
+    {
+      "epoch": 0.23564064801178203,
+      "grad_norm": 0.049899645149707794,
+      "learning_rate": 1e-05,
+      "loss": 2.4844,
+      "step": 140
+    },
+    {
+      "epoch": 0.23732379549758048,
+      "grad_norm": 0.05162065476179123,
+      "learning_rate": 1e-05,
+      "loss": 2.3613,
+      "step": 141
+    },
+    {
+      "epoch": 0.2390069429833789,
+      "grad_norm": 0.05812832713127136,
+      "learning_rate": 1e-05,
+      "loss": 2.4548,
+      "step": 142
+    },
+    {
+      "epoch": 0.24069009046917736,
+      "grad_norm": 0.04910556599497795,
+      "learning_rate": 1e-05,
+      "loss": 2.3274,
+      "step": 143
+    },
+    {
+      "epoch": 0.2423732379549758,
+      "grad_norm": 0.05346587672829628,
+      "learning_rate": 1e-05,
+      "loss": 2.325,
+      "step": 144
+    },
+    {
+      "epoch": 0.24405638544077424,
+      "grad_norm": 0.0495002381503582,
+      "learning_rate": 1e-05,
+      "loss": 2.4131,
+      "step": 145
+    },
+    {
+      "epoch": 0.2457395329265727,
+      "grad_norm": 0.05076875165104866,
+      "learning_rate": 1e-05,
+      "loss": 2.3887,
+      "step": 146
+    },
+    {
+      "epoch": 0.24742268041237114,
+      "grad_norm": 0.050955574959516525,
+      "learning_rate": 1e-05,
+      "loss": 2.4517,
+      "step": 147
+    },
+    {
+      "epoch": 0.24910582789816957,
+      "grad_norm": 0.05082906410098076,
+      "learning_rate": 1e-05,
+      "loss": 2.3401,
+      "step": 148
+    },
+    {
+      "epoch": 0.250788975383968,
+      "grad_norm": 0.052096717059612274,
+      "learning_rate": 1e-05,
+      "loss": 2.3218,
+      "step": 149
+    },
+    {
+      "epoch": 0.25247212286976645,
+      "grad_norm": 0.052378151565790176,
+      "learning_rate": 1e-05,
+      "loss": 2.4246,
+      "step": 150
+    },
+    {
+      "epoch": 0.2541552703555649,
+      "grad_norm": 0.04881056025624275,
+      "learning_rate": 1e-05,
+      "loss": 2.3435,
+      "step": 151
+    },
+    {
+      "epoch": 0.25583841784136335,
+      "grad_norm": 0.05233067274093628,
+      "learning_rate": 1e-05,
+      "loss": 2.4761,
+      "step": 152
+    },
+    {
+      "epoch": 0.2575215653271618,
+      "grad_norm": 0.05231297388672829,
+      "learning_rate": 1e-05,
+      "loss": 2.4065,
+      "step": 153
+    },
+    {
+      "epoch": 0.25920471281296026,
+      "grad_norm": 0.04649129509925842,
+      "learning_rate": 1e-05,
+      "loss": 2.4175,
+      "step": 154
+    },
+    {
+      "epoch": 0.26088786029875866,
+      "grad_norm": 0.05354660376906395,
+      "learning_rate": 1e-05,
+      "loss": 2.4731,
+      "step": 155
+    },
+    {
+      "epoch": 0.2625710077845571,
+      "grad_norm": 0.05071151629090309,
+      "learning_rate": 1e-05,
+      "loss": 2.4421,
+      "step": 156
+    },
+    {
+      "epoch": 0.26425415527035556,
+      "grad_norm": 0.04953297600150108,
+      "learning_rate": 1e-05,
+      "loss": 2.3134,
+      "step": 157
+    },
+    {
+      "epoch": 0.265937302756154,
+      "grad_norm": 0.051142722368240356,
+      "learning_rate": 1e-05,
+      "loss": 2.3335,
+      "step": 158
+    },
+    {
+      "epoch": 0.26762045024195247,
+      "grad_norm": 0.05187085270881653,
+      "learning_rate": 1e-05,
+      "loss": 2.4387,
+      "step": 159
+    },
+    {
+      "epoch": 0.2693035977277509,
+      "grad_norm": 0.04968629032373428,
+      "learning_rate": 1e-05,
+      "loss": 2.4905,
+      "step": 160
+    },
+    {
+      "epoch": 0.2709867452135493,
+      "grad_norm": 0.053009629249572754,
+      "learning_rate": 1e-05,
+      "loss": 2.4441,
+      "step": 161
+    },
+    {
+      "epoch": 0.27266989269934777,
+      "grad_norm": 0.04917874187231064,
+      "learning_rate": 1e-05,
+      "loss": 2.4763,
+      "step": 162
+    },
+    {
+      "epoch": 0.2743530401851462,
+      "grad_norm": 0.048884451389312744,
+      "learning_rate": 1e-05,
+      "loss": 2.4248,
+      "step": 163
+    },
+    {
+      "epoch": 0.2760361876709447,
+      "grad_norm": 0.049946676939725876,
+      "learning_rate": 1e-05,
+      "loss": 2.5173,
+      "step": 164
+    },
+    {
+      "epoch": 0.27771933515674313,
+      "grad_norm": 0.052534863352775574,
+      "learning_rate": 1e-05,
+      "loss": 2.4558,
+      "step": 165
+    },
+    {
+      "epoch": 0.2794024826425415,
+      "grad_norm": 0.05162844434380531,
+      "learning_rate": 1e-05,
+      "loss": 2.405,
+      "step": 166
+    },
+    {
+      "epoch": 0.28108563012834,
+      "grad_norm": 0.049985259771347046,
+      "learning_rate": 1e-05,
+      "loss": 2.3542,
+      "step": 167
+    },
+    {
+      "epoch": 0.28276877761413843,
+      "grad_norm": 0.05239354074001312,
+      "learning_rate": 1e-05,
+      "loss": 2.3721,
+      "step": 168
+    },
+    {
+      "epoch": 0.2844519250999369,
+      "grad_norm": 0.05592744052410126,
+      "learning_rate": 1e-05,
+      "loss": 2.2701,
+      "step": 169
+    },
+    {
+      "epoch": 0.28613507258573534,
+      "grad_norm": 0.052739113569259644,
+      "learning_rate": 1e-05,
+      "loss": 2.4216,
+      "step": 170
+    },
+    {
+      "epoch": 0.2878182200715338,
+      "grad_norm": 0.04806948080658913,
+      "learning_rate": 1e-05,
+      "loss": 2.3884,
+      "step": 171
+    },
+    {
+      "epoch": 0.2895013675573322,
+      "grad_norm": 0.04990949481725693,
+      "learning_rate": 1e-05,
+      "loss": 2.4419,
+      "step": 172
+    },
+    {
+      "epoch": 0.29118451504313064,
+      "grad_norm": 0.050067439675331116,
+      "learning_rate": 1e-05,
+      "loss": 2.4331,
+      "step": 173
+    },
+    {
+      "epoch": 0.2928676625289291,
+      "grad_norm": 0.0507354810833931,
+      "learning_rate": 1e-05,
+      "loss": 2.406,
+      "step": 174
+    },
+    {
+      "epoch": 0.29455081001472755,
+      "grad_norm": 0.0538686104118824,
+      "learning_rate": 1e-05,
+      "loss": 2.4182,
+      "step": 175
+    },
+    {
+      "epoch": 0.296233957500526,
+      "grad_norm": 0.05205219238996506,
+      "learning_rate": 1e-05,
+      "loss": 2.3401,
+      "step": 176
+    },
+    {
+      "epoch": 0.2979171049863244,
+      "grad_norm": 0.04672086611390114,
+      "learning_rate": 1e-05,
+      "loss": 2.3149,
+      "step": 177
+    },
+    {
+      "epoch": 0.29960025247212285,
+      "grad_norm": 0.051963068544864655,
+      "learning_rate": 1e-05,
+      "loss": 2.2537,
+      "step": 178
+    },
+    {
+      "epoch": 0.3012833999579213,
+      "grad_norm": 0.053639005869627,
+      "learning_rate": 1e-05,
+      "loss": 2.4353,
+      "step": 179
+    },
+    {
+      "epoch": 0.30296654744371976,
+      "grad_norm": 0.05326982960104942,
+      "learning_rate": 1e-05,
+      "loss": 2.334,
+      "step": 180
+    },
+    {
+      "epoch": 0.3046496949295182,
+      "grad_norm": 0.05361334979534149,
+      "learning_rate": 1e-05,
+      "loss": 2.4224,
+      "step": 181
+    },
+    {
+      "epoch": 0.30633284241531666,
+      "grad_norm": 0.05790587514638901,
+      "learning_rate": 1e-05,
+      "loss": 2.334,
+      "step": 182
+    },
+    {
+      "epoch": 0.30801598990111506,
+      "grad_norm": 0.04790763929486275,
+      "learning_rate": 1e-05,
+      "loss": 2.5073,
+      "step": 183
+    },
+    {
+      "epoch": 0.3096991373869135,
+      "grad_norm": 0.054103124886751175,
+      "learning_rate": 1e-05,
+      "loss": 2.3483,
+      "step": 184
+    },
+    {
+      "epoch": 0.31138228487271197,
+      "grad_norm": 0.05902162939310074,
+      "learning_rate": 1e-05,
+      "loss": 2.3301,
+      "step": 185
+    },
+    {
+      "epoch": 0.3130654323585104,
+      "grad_norm": 0.04853544384241104,
+      "learning_rate": 1e-05,
+      "loss": 2.5566,
+      "step": 186
+    },
+    {
+      "epoch": 0.3147485798443089,
+      "grad_norm": 0.055288165807724,
+      "learning_rate": 1e-05,
+      "loss": 2.2903,
+      "step": 187
+    },
+    {
+      "epoch": 0.3164317273301073,
+      "grad_norm": 0.05180734023451805,
+      "learning_rate": 1e-05,
+      "loss": 2.4285,
+      "step": 188
+    },
+    {
+      "epoch": 0.3181148748159057,
+      "grad_norm": 0.04889997839927673,
+      "learning_rate": 1e-05,
+      "loss": 2.2542,
+      "step": 189
+    },
+    {
+      "epoch": 0.3197980223017042,
+      "grad_norm": 0.051011502742767334,
+      "learning_rate": 1e-05,
+      "loss": 2.2893,
+      "step": 190
+    },
+    {
+      "epoch": 0.32148116978750263,
+      "grad_norm": 0.04864371567964554,
+      "learning_rate": 1e-05,
+      "loss": 2.5225,
+      "step": 191
+    },
+    {
+      "epoch": 0.3231643172733011,
+      "grad_norm": 0.05374041944742203,
+      "learning_rate": 1e-05,
+      "loss": 2.4504,
+      "step": 192
+    },
+    {
+      "epoch": 0.32484746475909954,
+      "grad_norm": 0.05158041790127754,
+      "learning_rate": 1e-05,
+      "loss": 2.4683,
+      "step": 193
+    },
+    {
+      "epoch": 0.32653061224489793,
+      "grad_norm": 0.05630083382129669,
+      "learning_rate": 1e-05,
+      "loss": 2.2415,
+      "step": 194
+    },
+    {
+      "epoch": 0.3282137597306964,
+      "grad_norm": 0.05439196154475212,
+      "learning_rate": 1e-05,
+      "loss": 2.3684,
+      "step": 195
+    },
+    {
+      "epoch": 0.32989690721649484,
+      "grad_norm": 0.05023415759205818,
+      "learning_rate": 1e-05,
+      "loss": 2.415,
+      "step": 196
+    },
+    {
+      "epoch": 0.3315800547022933,
+      "grad_norm": 0.05531445890665054,
+      "learning_rate": 1e-05,
+      "loss": 2.4626,
+      "step": 197
+    },
+    {
+      "epoch": 0.33326320218809175,
+      "grad_norm": 0.05087656155228615,
+      "learning_rate": 1e-05,
+      "loss": 2.3936,
+      "step": 198
+    },
+    {
+      "epoch": 0.3349463496738902,
+      "grad_norm": 0.05231088399887085,
+      "learning_rate": 1e-05,
+      "loss": 2.3779,
+      "step": 199
+    },
+    {
+      "epoch": 0.3366294971596886,
+      "grad_norm": 0.0514984093606472,
+      "learning_rate": 1e-05,
+      "loss": 2.3967,
+      "step": 200
+    },
+    {
+      "epoch": 0.33831264464548705,
+      "grad_norm": 0.05334719642996788,
+      "learning_rate": 1e-05,
+      "loss": 2.4604,
+      "step": 201
+    },
+    {
+      "epoch": 0.3399957921312855,
+      "grad_norm": 0.054843124002218246,
+      "learning_rate": 1e-05,
+      "loss": 2.3538,
+      "step": 202
+    },
+    {
+      "epoch": 0.34167893961708395,
+      "grad_norm": 0.04888272285461426,
+      "learning_rate": 1e-05,
+      "loss": 2.4844,
+      "step": 203
+    },
+    {
+      "epoch": 0.3433620871028824,
+      "grad_norm": 0.054122187197208405,
+      "learning_rate": 1e-05,
+      "loss": 2.3291,
+      "step": 204
+    },
+    {
+      "epoch": 0.34504523458868086,
+      "grad_norm": 0.054561201483011246,
+      "learning_rate": 1e-05,
+      "loss": 2.3218,
+      "step": 205
+    },
+    {
+      "epoch": 0.34672838207447926,
+      "grad_norm": 0.04919834062457085,
+      "learning_rate": 1e-05,
+      "loss": 2.4478,
+      "step": 206
+    },
+    {
+      "epoch": 0.3484115295602777,
+      "grad_norm": 0.050551943480968475,
+      "learning_rate": 1e-05,
+      "loss": 2.3755,
+      "step": 207
+    },
+    {
+      "epoch": 0.35009467704607616,
+      "grad_norm": 0.05242514982819557,
+      "learning_rate": 1e-05,
+      "loss": 2.3922,
+      "step": 208
+    },
+    {
+      "epoch": 0.3517778245318746,
+      "grad_norm": 0.06077054515480995,
+      "learning_rate": 1e-05,
+      "loss": 2.3218,
+      "step": 209
+    },
+    {
+      "epoch": 0.35346097201767307,
+      "grad_norm": 0.061367545276880264,
+      "learning_rate": 1e-05,
+      "loss": 2.2957,
+      "step": 210
+    },
+    {
+      "epoch": 0.35514411950347147,
+      "grad_norm": 0.0511772483587265,
+      "learning_rate": 1e-05,
+      "loss": 2.374,
+      "step": 211
+    },
+    {
+      "epoch": 0.3568272669892699,
+      "grad_norm": 0.0496203638613224,
+      "learning_rate": 1e-05,
+      "loss": 2.4182,
+      "step": 212
+    },
+    {
+      "epoch": 0.3585104144750684,
+      "grad_norm": 0.061339233070611954,
+      "learning_rate": 1e-05,
+      "loss": 2.406,
+      "step": 213
+    },
+    {
+      "epoch": 0.3601935619608668,
+      "grad_norm": 0.052460432052612305,
+      "learning_rate": 1e-05,
+      "loss": 2.4309,
+      "step": 214
+    },
+    {
+      "epoch": 0.3618767094466653,
+      "grad_norm": 0.055436089634895325,
+      "learning_rate": 1e-05,
+      "loss": 2.4141,
+      "step": 215
+    },
+    {
+      "epoch": 0.36355985693246373,
+      "grad_norm": 0.05396036058664322,
+      "learning_rate": 1e-05,
+      "loss": 2.2705,
+      "step": 216
+    },
+    {
+      "epoch": 0.36524300441826213,
+      "grad_norm": 0.04853086173534393,
+      "learning_rate": 1e-05,
+      "loss": 2.4473,
+      "step": 217
+    },
+    {
+      "epoch": 0.3669261519040606,
+      "grad_norm": 0.051015399396419525,
+      "learning_rate": 1e-05,
+      "loss": 2.5115,
+      "step": 218
+    },
+    {
+      "epoch": 0.36860929938985904,
+      "grad_norm": 0.05526035279035568,
+      "learning_rate": 1e-05,
+      "loss": 2.3123,
+      "step": 219
+    },
+    {
+      "epoch": 0.3702924468756575,
+      "grad_norm": 0.056169234216213226,
+      "learning_rate": 1e-05,
+      "loss": 2.3447,
+      "step": 220
+    },
+    {
+      "epoch": 0.37197559436145594,
+      "grad_norm": 0.05238133668899536,
+      "learning_rate": 1e-05,
+      "loss": 2.26,
+      "step": 221
+    },
+    {
+      "epoch": 0.37365874184725434,
+      "grad_norm": 0.05587685480713844,
+      "learning_rate": 1e-05,
+      "loss": 2.3083,
+      "step": 222
+    },
+    {
+      "epoch": 0.3753418893330528,
+      "grad_norm": 0.050364553928375244,
+      "learning_rate": 1e-05,
+      "loss": 2.3459,
+      "step": 223
+    },
+    {
+      "epoch": 0.37702503681885124,
+      "grad_norm": 0.0506574809551239,
+      "learning_rate": 1e-05,
+      "loss": 2.4246,
+      "step": 224
+    },
+    {
+      "epoch": 0.3787081843046497,
+      "grad_norm": 0.05842865630984306,
+      "learning_rate": 1e-05,
+      "loss": 2.2617,
+      "step": 225
+    },
+    {
+      "epoch": 0.38039133179044815,
+      "grad_norm": 0.05097496882081032,
+      "learning_rate": 1e-05,
+      "loss": 2.52,
+      "step": 226
+    },
+    {
+      "epoch": 0.3820744792762466,
+      "grad_norm": 0.05665278434753418,
+      "learning_rate": 1e-05,
+      "loss": 2.2715,
+      "step": 227
+    },
+    {
+      "epoch": 0.383757626762045,
+      "grad_norm": 0.053350359201431274,
+      "learning_rate": 1e-05,
+      "loss": 2.3101,
+      "step": 228
+    },
+    {
+      "epoch": 0.38544077424784345,
+      "grad_norm": 0.05481604114174843,
+      "learning_rate": 1e-05,
+      "loss": 2.3347,
+      "step": 229
+    },
+    {
+      "epoch": 0.3871239217336419,
+      "grad_norm": 0.06036606431007385,
+      "learning_rate": 1e-05,
+      "loss": 2.2991,
+      "step": 230
+    },
+    {
+      "epoch": 0.38880706921944036,
+      "grad_norm": 0.0606355145573616,
+      "learning_rate": 1e-05,
+      "loss": 2.4226,
+      "step": 231
+    },
+    {
+      "epoch": 0.3904902167052388,
+      "grad_norm": 0.052770137786865234,
+      "learning_rate": 1e-05,
+      "loss": 2.4539,
+      "step": 232
+    },
+    {
+      "epoch": 0.39217336419103727,
+      "grad_norm": 0.050006203353405,
+      "learning_rate": 1e-05,
+      "loss": 2.3477,
+      "step": 233
+    },
+    {
+      "epoch": 0.39385651167683566,
+      "grad_norm": 0.05640649050474167,
+      "learning_rate": 1e-05,
+      "loss": 2.3123,
+      "step": 234
+    },
+    {
+      "epoch": 0.3955396591626341,
+      "grad_norm": 0.050969429314136505,
+      "learning_rate": 1e-05,
+      "loss": 2.4534,
+      "step": 235
+    },
+    {
+      "epoch": 0.39722280664843257,
+      "grad_norm": 0.05676101893186569,
+      "learning_rate": 1e-05,
+      "loss": 2.3481,
+      "step": 236
+    },
+    {
+      "epoch": 0.398905954134231,
+      "grad_norm": 0.05844707787036896,
+      "learning_rate": 1e-05,
+      "loss": 2.3638,
+      "step": 237
+    },
+    {
+      "epoch": 0.4005891016200295,
+      "grad_norm": 0.053074926137924194,
+      "learning_rate": 1e-05,
+      "loss": 2.3904,
+      "step": 238
+    },
+    {
+      "epoch": 0.4022722491058279,
+      "grad_norm": 0.04979414492845535,
+      "learning_rate": 1e-05,
+      "loss": 2.3855,
+      "step": 239
+    },
+    {
+      "epoch": 0.4039553965916263,
+      "grad_norm": 0.05607665330171585,
+      "learning_rate": 1e-05,
+      "loss": 2.3569,
+      "step": 240
+    },
+    {
+      "epoch": 0.4056385440774248,
+      "grad_norm": 0.05964501202106476,
+      "learning_rate": 1e-05,
+      "loss": 2.3459,
+      "step": 241
+    },
+    {
+      "epoch": 0.40732169156322323,
+      "grad_norm": 0.05849093198776245,
+      "learning_rate": 1e-05,
+      "loss": 2.3213,
+      "step": 242
+    },
+    {
+      "epoch": 0.4090048390490217,
+      "grad_norm": 0.053846072405576706,
+      "learning_rate": 1e-05,
+      "loss": 2.4436,
+      "step": 243
+    },
+    {
+      "epoch": 0.41068798653482014,
+      "grad_norm": 0.054448988288640976,
+      "learning_rate": 1e-05,
+      "loss": 2.3716,
+      "step": 244
+    },
+    {
+      "epoch": 0.41237113402061853,
+      "grad_norm": 0.05229583755135536,
+      "learning_rate": 1e-05,
+      "loss": 2.4099,
+      "step": 245
+    },
+    {
+      "epoch": 0.414054281506417,
+      "grad_norm": 0.05479966476559639,
+      "learning_rate": 1e-05,
+      "loss": 2.4026,
+      "step": 246
+    },
+    {
+      "epoch": 0.41573742899221544,
+      "grad_norm": 0.061799049377441406,
+      "learning_rate": 1e-05,
+      "loss": 2.4072,
+      "step": 247
+    },
+    {
+      "epoch": 0.4174205764780139,
+      "grad_norm": 0.061452727764844894,
+      "learning_rate": 1e-05,
+      "loss": 2.2833,
+      "step": 248
+    },
+    {
+      "epoch": 0.41910372396381235,
+      "grad_norm": 0.05868072435259819,
+      "learning_rate": 1e-05,
+      "loss": 2.3833,
+      "step": 249
+    },
+    {
+      "epoch": 0.4207868714496108,
+      "grad_norm": 0.05926290899515152,
+      "learning_rate": 1e-05,
+      "loss": 2.3645,
+      "step": 250
+    },
+    {
+      "epoch": 0.4224700189354092,
+      "grad_norm": 0.058858342468738556,
+      "learning_rate": 1e-05,
+      "loss": 2.3152,
+      "step": 251
+    },
+    {
+      "epoch": 0.42415316642120765,
+      "grad_norm": 0.058599065989255905,
+      "learning_rate": 1e-05,
+      "loss": 2.2827,
+      "step": 252
+    },
+    {
+      "epoch": 0.4258363139070061,
+      "grad_norm": 0.060381706804037094,
+      "learning_rate": 1e-05,
+      "loss": 2.3024,
+      "step": 253
+    },
+    {
+      "epoch": 0.42751946139280456,
+      "grad_norm": 0.05441940575838089,
+      "learning_rate": 1e-05,
+      "loss": 2.446,
+      "step": 254
+    },
+    {
+      "epoch": 0.429202608878603,
+      "grad_norm": 0.05750846117734909,
+      "learning_rate": 1e-05,
+      "loss": 2.3958,
+      "step": 255
+    },
+    {
+      "epoch": 0.4308857563644014,
+      "grad_norm": 0.060346368700265884,
+      "learning_rate": 1e-05,
+      "loss": 2.2395,
+      "step": 256
+    },
+    {
+      "epoch": 0.43256890385019986,
+      "grad_norm": 0.056383710354566574,
+      "learning_rate": 1e-05,
+      "loss": 2.3518,
+      "step": 257
+    },
+    {
+      "epoch": 0.4342520513359983,
+      "grad_norm": 0.057746805250644684,
+      "learning_rate": 1e-05,
+      "loss": 2.2834,
+      "step": 258
+    },
+    {
+      "epoch": 0.43593519882179677,
+      "grad_norm": 0.051562029868364334,
+      "learning_rate": 1e-05,
+      "loss": 2.3677,
+      "step": 259
+    },
+    {
+      "epoch": 0.4376183463075952,
+      "grad_norm": 0.059988316148519516,
+      "learning_rate": 1e-05,
+      "loss": 2.3372,
+      "step": 260
+    },
+    {
+      "epoch": 0.43930149379339367,
+      "grad_norm": 0.05852155759930611,
+      "learning_rate": 1e-05,
+      "loss": 2.3875,
+      "step": 261
+    },
+    {
+      "epoch": 0.44098464127919207,
+      "grad_norm": 0.06629418581724167,
+      "learning_rate": 1e-05,
+      "loss": 2.4194,
+      "step": 262
+    },
+    {
+      "epoch": 0.4426677887649905,
+      "grad_norm": 0.061044465750455856,
+      "learning_rate": 1e-05,
+      "loss": 2.2466,
+      "step": 263
+    },
+    {
+      "epoch": 0.444350936250789,
+      "grad_norm": 0.056285977363586426,
+      "learning_rate": 1e-05,
+      "loss": 2.3105,
+      "step": 264
+    },
+    {
+      "epoch": 0.44603408373658743,
+      "grad_norm": 0.06135227158665657,
+      "learning_rate": 1e-05,
+      "loss": 2.3853,
+      "step": 265
+    },
+    {
+      "epoch": 0.4477172312223859,
+      "grad_norm": 0.05644640699028969,
+      "learning_rate": 1e-05,
+      "loss": 2.3888,
+      "step": 266
+    },
+    {
+      "epoch": 0.4494003787081843,
+      "grad_norm": 0.06326981633901596,
+      "learning_rate": 1e-05,
+      "loss": 2.3132,
+      "step": 267
+    },
+    {
+      "epoch": 0.45108352619398273,
+      "grad_norm": 0.05710430070757866,
+      "learning_rate": 1e-05,
+      "loss": 2.365,
+      "step": 268
+    },
+    {
+      "epoch": 0.4527666736797812,
+      "grad_norm": 0.05607946217060089,
+      "learning_rate": 1e-05,
+      "loss": 2.4648,
+      "step": 269
+    },
+    {
+      "epoch": 0.45444982116557964,
+      "grad_norm": 0.057825781404972076,
+      "learning_rate": 1e-05,
+      "loss": 2.4189,
+      "step": 270
+    },
+    {
+      "epoch": 0.4561329686513781,
+      "grad_norm": 0.06380680948495865,
+      "learning_rate": 1e-05,
+      "loss": 2.3188,
+      "step": 271
+    },
+    {
+      "epoch": 0.45781611613717654,
+      "grad_norm": 0.06377760320901871,
+      "learning_rate": 1e-05,
+      "loss": 2.2896,
+      "step": 272
+    },
+    {
+      "epoch": 0.45949926362297494,
+      "grad_norm": 0.06210333853960037,
+      "learning_rate": 1e-05,
+      "loss": 2.3663,
+      "step": 273
+    },
+    {
+      "epoch": 0.4611824111087734,
+      "grad_norm": 0.06039275974035263,
+      "learning_rate": 1e-05,
+      "loss": 2.408,
+      "step": 274
+    },
+    {
+      "epoch": 0.46286555859457185,
+      "grad_norm": 0.05442138388752937,
+      "learning_rate": 1e-05,
+      "loss": 2.3843,
+      "step": 275
+    },
+    {
+      "epoch": 0.4645487060803703,
+      "grad_norm": 0.06208937615156174,
+      "learning_rate": 1e-05,
+      "loss": 2.4355,
+      "step": 276
+    },
+    {
+      "epoch": 0.46623185356616875,
+      "grad_norm": 0.0619891993701458,
+      "learning_rate": 1e-05,
+      "loss": 2.3196,
+      "step": 277
+    },
+    {
+      "epoch": 0.4679150010519672,
+      "grad_norm": 0.059192296117544174,
+      "learning_rate": 1e-05,
+      "loss": 2.3237,
+      "step": 278
+    },
+    {
+      "epoch": 0.4695981485377656,
+      "grad_norm": 0.06284468621015549,
+      "learning_rate": 1e-05,
+      "loss": 2.3694,
+      "step": 279
+    },
+    {
+      "epoch": 0.47128129602356406,
+      "grad_norm": 0.06121189519762993,
+      "learning_rate": 1e-05,
+      "loss": 2.3606,
+      "step": 280
+    },
+    {
+      "epoch": 0.4729644435093625,
+      "grad_norm": 0.061919402331113815,
+      "learning_rate": 1e-05,
+      "loss": 2.3381,
+      "step": 281
+    },
+    {
+      "epoch": 0.47464759099516096,
+      "grad_norm": 0.0676443800330162,
+      "learning_rate": 1e-05,
+      "loss": 2.3624,
+      "step": 282
+    },
+    {
+      "epoch": 0.4763307384809594,
+      "grad_norm": 0.060140665620565414,
+      "learning_rate": 1e-05,
+      "loss": 2.4541,
+      "step": 283
+    },
+    {
+      "epoch": 0.4780138859667578,
+      "grad_norm": 0.062285441905260086,
+      "learning_rate": 1e-05,
+      "loss": 2.323,
+      "step": 284
+    },
+    {
+      "epoch": 0.47969703345255627,
+      "grad_norm": 0.06063227355480194,
+      "learning_rate": 1e-05,
+      "loss": 2.3596,
+      "step": 285
+    },
+    {
+      "epoch": 0.4813801809383547,
+      "grad_norm": 0.05906851589679718,
+      "learning_rate": 1e-05,
+      "loss": 2.458,
+      "step": 286
+    },
+    {
+      "epoch": 0.48306332842415317,
+      "grad_norm": 0.05862203240394592,
+      "learning_rate": 1e-05,
+      "loss": 2.291,
+      "step": 287
+    },
+    {
+      "epoch": 0.4847464759099516,
+      "grad_norm": 0.0629325732588768,
+      "learning_rate": 1e-05,
+      "loss": 2.2634,
+      "step": 288
+    },
+    {
+      "epoch": 0.4864296233957501,
+      "grad_norm": 0.06464157998561859,
+      "learning_rate": 1e-05,
+      "loss": 2.2531,
+      "step": 289
+    },
+    {
+      "epoch": 0.4881127708815485,
+      "grad_norm": 0.0547555610537529,
+      "learning_rate": 1e-05,
+      "loss": 2.5339,
+      "step": 290
+    },
+    {
+      "epoch": 0.4897959183673469,
+      "grad_norm": 0.0606168657541275,
+      "learning_rate": 1e-05,
+      "loss": 2.2886,
+      "step": 291
+    },
+    {
+      "epoch": 0.4914790658531454,
+      "grad_norm": 0.058814577758312225,
+      "learning_rate": 1e-05,
+      "loss": 2.3337,
+      "step": 292
+    },
+    {
+      "epoch": 0.49316221333894383,
+      "grad_norm": 0.0691385492682457,
+      "learning_rate": 1e-05,
+      "loss": 2.2904,
+      "step": 293
+    },
+    {
+      "epoch": 0.4948453608247423,
+      "grad_norm": 0.06522157788276672,
+      "learning_rate": 1e-05,
+      "loss": 2.469,
+      "step": 294
+    },
+    {
+      "epoch": 0.4965285083105407,
+      "grad_norm": 0.05957287177443504,
+      "learning_rate": 1e-05,
+      "loss": 2.4095,
+      "step": 295
+    },
+    {
+      "epoch": 0.49821165579633914,
+      "grad_norm": 0.06277060508728027,
+      "learning_rate": 1e-05,
+      "loss": 2.4697,
+      "step": 296
+    },
+    {
+      "epoch": 0.4998948032821376,
+      "grad_norm": 0.06802426278591156,
+      "learning_rate": 1e-05,
+      "loss": 2.2517,
+      "step": 297
+    },
+    {
+      "epoch": 0.501577950767936,
+      "grad_norm": 0.06365792453289032,
+      "learning_rate": 1e-05,
+      "loss": 2.2942,
+      "step": 298
+    },
+    {
+      "epoch": 0.5032610982537344,
+      "grad_norm": 0.06624794751405716,
+      "learning_rate": 1e-05,
+      "loss": 2.283,
+      "step": 299
+    },
+    {
+      "epoch": 0.5049442457395329,
+      "grad_norm": 0.05979595705866814,
+      "learning_rate": 1e-05,
+      "loss": 2.4387,
+      "step": 300
+    },
+    {
+      "epoch": 0.5066273932253313,
+      "grad_norm": 0.06187634915113449,
+      "learning_rate": 1e-05,
+      "loss": 2.4205,
+      "step": 301
+    },
+    {
+      "epoch": 0.5083105407111298,
+      "grad_norm": 0.06389462947845459,
+      "learning_rate": 1e-05,
+      "loss": 2.2775,
+      "step": 302
+    },
+    {
+      "epoch": 0.5099936881969283,
+      "grad_norm": 0.05831071361899376,
+      "learning_rate": 1e-05,
+      "loss": 2.3892,
+      "step": 303
+    },
+    {
+      "epoch": 0.5116768356827267,
+      "grad_norm": 0.06568494439125061,
+      "learning_rate": 1e-05,
+      "loss": 2.3087,
+      "step": 304
+    },
+    {
+      "epoch": 0.5133599831685252,
+      "grad_norm": 0.062109317630529404,
+      "learning_rate": 1e-05,
+      "loss": 2.3268,
+      "step": 305
+    },
+    {
+      "epoch": 0.5150431306543236,
+      "grad_norm": 0.061168327927589417,
+      "learning_rate": 1e-05,
+      "loss": 2.3093,
+      "step": 306
+    },
+    {
+      "epoch": 0.5167262781401221,
+      "grad_norm": 0.061159648001194,
+      "learning_rate": 1e-05,
+      "loss": 2.3315,
+      "step": 307
+    },
+    {
+      "epoch": 0.5184094256259205,
+      "grad_norm": 0.06269169598817825,
+      "learning_rate": 1e-05,
+      "loss": 2.3442,
+      "step": 308
+    },
+    {
+      "epoch": 0.520092573111719,
+      "grad_norm": 0.06711502373218536,
+      "learning_rate": 1e-05,
+      "loss": 2.2008,
+      "step": 309
+    },
+    {
+      "epoch": 0.5217757205975173,
+      "grad_norm": 0.0663105845451355,
+      "learning_rate": 1e-05,
+      "loss": 2.3502,
+      "step": 310
+    },
+    {
+      "epoch": 0.5234588680833158,
+      "grad_norm": 0.06040646880865097,
+      "learning_rate": 1e-05,
+      "loss": 2.3414,
+      "step": 311
+    },
+    {
+      "epoch": 0.5251420155691142,
+      "grad_norm": 0.06823603063821793,
+      "learning_rate": 1e-05,
+      "loss": 2.3392,
+      "step": 312
+    },
+    {
+      "epoch": 0.5268251630549127,
+      "grad_norm": 0.05944176763296127,
+      "learning_rate": 1e-05,
+      "loss": 2.3193,
+      "step": 313
+    },
+    {
+      "epoch": 0.5285083105407111,
+      "grad_norm": 0.06610157340765,
+      "learning_rate": 1e-05,
+      "loss": 2.2288,
+      "step": 314
+    },
+    {
+      "epoch": 0.5301914580265096,
+      "grad_norm": 0.06880299746990204,
+      "learning_rate": 1e-05,
+      "loss": 2.3529,
+      "step": 315
+    },
+    {
+      "epoch": 0.531874605512308,
+      "grad_norm": 0.06061836704611778,
+      "learning_rate": 1e-05,
+      "loss": 2.3533,
+      "step": 316
+    },
+    {
+      "epoch": 0.5335577529981065,
+      "grad_norm": 0.06552371382713318,
+      "learning_rate": 1e-05,
+      "loss": 2.3579,
+      "step": 317
+    },
+    {
+      "epoch": 0.5352409004839049,
+      "grad_norm": 0.06967922300100327,
+      "learning_rate": 1e-05,
+      "loss": 2.2983,
+      "step": 318
+    },
+    {
+      "epoch": 0.5369240479697034,
+      "grad_norm": 0.06997574120759964,
+      "learning_rate": 1e-05,
+      "loss": 2.355,
+      "step": 319
+    },
+    {
+      "epoch": 0.5386071954555018,
+      "grad_norm": 0.0654403418302536,
+      "learning_rate": 1e-05,
+      "loss": 2.4258,
+      "step": 320
+    },
+    {
+      "epoch": 0.5402903429413002,
+      "grad_norm": 0.06031208485364914,
+      "learning_rate": 1e-05,
+      "loss": 2.4011,
+      "step": 321
+    },
+    {
+      "epoch": 0.5419734904270986,
+      "grad_norm": 0.06496379524469376,
+      "learning_rate": 1e-05,
+      "loss": 2.2429,
+      "step": 322
+    },
+    {
+      "epoch": 0.5436566379128971,
+      "grad_norm": 0.06525281816720963,
+      "learning_rate": 1e-05,
+      "loss": 2.3254,
+      "step": 323
+    },
+    {
+      "epoch": 0.5453397853986955,
+      "grad_norm": 0.07553514093160629,
+      "learning_rate": 1e-05,
+      "loss": 2.2953,
+      "step": 324
+    },
+    {
+      "epoch": 0.547022932884494,
+      "grad_norm": 0.06429509073495865,
+      "learning_rate": 1e-05,
+      "loss": 2.3319,
+      "step": 325
+    },
+    {
+      "epoch": 0.5487060803702924,
+      "grad_norm": 0.0657946914434433,
+      "learning_rate": 1e-05,
+      "loss": 2.3501,
+      "step": 326
+    },
+    {
+      "epoch": 0.5503892278560909,
+      "grad_norm": 0.06548567861318588,
+      "learning_rate": 1e-05,
+      "loss": 2.2781,
+      "step": 327
+    },
+    {
+      "epoch": 0.5520723753418894,
+      "grad_norm": 0.06299672275781631,
+      "learning_rate": 1e-05,
+      "loss": 2.377,
+      "step": 328
+    },
+    {
+      "epoch": 0.5537555228276878,
+      "grad_norm": 0.06381850689649582,
+      "learning_rate": 1e-05,
+      "loss": 2.3945,
+      "step": 329
+    },
+    {
+      "epoch": 0.5554386703134863,
+      "grad_norm": 0.06497140228748322,
+      "learning_rate": 1e-05,
+      "loss": 2.3496,
+      "step": 330
+    },
+    {
+      "epoch": 0.5571218177992847,
+      "grad_norm": 0.06588133424520493,
+      "learning_rate": 1e-05,
+      "loss": 2.3955,
+      "step": 331
+    },
+    {
+      "epoch": 0.558804965285083,
+      "grad_norm": 0.06468643248081207,
+      "learning_rate": 1e-05,
+      "loss": 2.2893,
+      "step": 332
+    },
+    {
+      "epoch": 0.5604881127708815,
+      "grad_norm": 0.07278285920619965,
+      "learning_rate": 1e-05,
+      "loss": 2.3179,
+      "step": 333
+    },
+    {
+      "epoch": 0.56217126025668,
+      "grad_norm": 0.06992325931787491,
+      "learning_rate": 1e-05,
+      "loss": 2.3588,
+      "step": 334
+    },
+    {
+      "epoch": 0.5638544077424784,
+      "grad_norm": 0.06566626578569412,
+      "learning_rate": 1e-05,
+      "loss": 2.4763,
+      "step": 335
+    },
+    {
+      "epoch": 0.5655375552282769,
+      "grad_norm": 0.0633927658200264,
+      "learning_rate": 1e-05,
+      "loss": 2.4685,
+      "step": 336
+    },
+    {
+      "epoch": 0.5672207027140753,
+      "grad_norm": 0.06903122365474701,
+      "learning_rate": 1e-05,
+      "loss": 2.311,
+      "step": 337
+    },
+    {
+      "epoch": 0.5689038501998738,
+      "grad_norm": 0.06421441584825516,
+      "learning_rate": 1e-05,
+      "loss": 2.3589,
+      "step": 338
+    },
+    {
+      "epoch": 0.5705869976856722,
+      "grad_norm": 0.07122648507356644,
+      "learning_rate": 1e-05,
+      "loss": 2.3798,
+      "step": 339
+    },
+    {
+      "epoch": 0.5722701451714707,
+      "grad_norm": 0.06518077105283737,
+      "learning_rate": 1e-05,
+      "loss": 2.4546,
+      "step": 340
+    },
+    {
+      "epoch": 0.5739532926572691,
+      "grad_norm": 0.07509720325469971,
+      "learning_rate": 1e-05,
+      "loss": 2.3341,
+      "step": 341
+    },
+    {
+      "epoch": 0.5756364401430676,
+      "grad_norm": 0.06559302657842636,
+      "learning_rate": 1e-05,
+      "loss": 2.3127,
+      "step": 342
+    },
+    {
+      "epoch": 0.5773195876288659,
+      "grad_norm": 0.06652245670557022,
+      "learning_rate": 1e-05,
+      "loss": 2.3997,
+      "step": 343
+    },
+    {
+      "epoch": 0.5790027351146644,
+      "grad_norm": 0.07472145557403564,
+      "learning_rate": 1e-05,
+      "loss": 2.3237,
+      "step": 344
+    },
+    {
+      "epoch": 0.5806858826004628,
+      "grad_norm": 0.07624109089374542,
+      "learning_rate": 1e-05,
+      "loss": 2.186,
+      "step": 345
+    },
+    {
+      "epoch": 0.5823690300862613,
+      "grad_norm": 0.06387084722518921,
+      "learning_rate": 1e-05,
+      "loss": 2.2717,
+      "step": 346
+    },
+    {
+      "epoch": 0.5840521775720597,
+      "grad_norm": 0.06857839971780777,
+      "learning_rate": 1e-05,
+      "loss": 2.3726,
+      "step": 347
+    },
+    {
+      "epoch": 0.5857353250578582,
+      "grad_norm": 0.06429892778396606,
+      "learning_rate": 1e-05,
+      "loss": 2.4109,
+      "step": 348
+    },
+    {
+      "epoch": 0.5874184725436566,
+      "grad_norm": 0.0720372200012207,
+      "learning_rate": 1e-05,
+      "loss": 2.3291,
+      "step": 349
+    },
+    {
+      "epoch": 0.5891016200294551,
+      "grad_norm": 0.0749678909778595,
+      "learning_rate": 1e-05,
+      "loss": 2.3369,
+      "step": 350
+    },
+    {
+      "epoch": 0.5907847675152536,
+      "grad_norm": 0.0645705908536911,
+      "learning_rate": 1e-05,
+      "loss": 2.3894,
+      "step": 351
+    },
+    {
+      "epoch": 0.592467915001052,
+      "grad_norm": 0.06680341064929962,
+      "learning_rate": 1e-05,
+      "loss": 2.3335,
+      "step": 352
+    },
+    {
+      "epoch": 0.5941510624868505,
+      "grad_norm": 0.07383781671524048,
+      "learning_rate": 1e-05,
+      "loss": 2.2733,
+      "step": 353
+    },
+    {
+      "epoch": 0.5958342099726488,
+      "grad_norm": 0.07338624447584152,
+      "learning_rate": 1e-05,
+      "loss": 2.2236,
+      "step": 354
+    },
+    {
+      "epoch": 0.5975173574584473,
+      "grad_norm": 0.06998410820960999,
+      "learning_rate": 1e-05,
+      "loss": 2.2552,
+      "step": 355
+    },
+    {
+      "epoch": 0.5992005049442457,
+      "grad_norm": 0.06697436422109604,
+      "learning_rate": 1e-05,
+      "loss": 2.4231,
+      "step": 356
+    },
+    {
+      "epoch": 0.6008836524300442,
+      "grad_norm": 0.06693920493125916,
+      "learning_rate": 1e-05,
+      "loss": 2.3296,
+      "step": 357
+    },
+    {
+      "epoch": 0.6025667999158426,
+      "grad_norm": 0.06306028366088867,
+      "learning_rate": 1e-05,
+      "loss": 2.4009,
+      "step": 358
+    },
+    {
+      "epoch": 0.6042499474016411,
+      "grad_norm": 0.0724472776055336,
+      "learning_rate": 1e-05,
+      "loss": 2.2986,
+      "step": 359
+    },
+    {
+      "epoch": 0.6059330948874395,
+      "grad_norm": 0.06711563467979431,
+      "learning_rate": 1e-05,
+      "loss": 2.3755,
+      "step": 360
+    },
+    {
+      "epoch": 0.607616242373238,
+      "grad_norm": 0.07287666201591492,
+      "learning_rate": 1e-05,
+      "loss": 2.325,
+      "step": 361
+    },
+    {
+      "epoch": 0.6092993898590364,
+      "grad_norm": 0.07494334876537323,
+      "learning_rate": 1e-05,
+      "loss": 2.2673,
+      "step": 362
+    },
+    {
+      "epoch": 0.6109825373448349,
+      "grad_norm": 0.07399529218673706,
+      "learning_rate": 1e-05,
+      "loss": 2.3134,
+      "step": 363
+    },
+    {
+      "epoch": 0.6126656848306333,
+      "grad_norm": 0.06705833226442337,
+      "learning_rate": 1e-05,
+      "loss": 2.3772,
+      "step": 364
+    },
+    {
+      "epoch": 0.6143488323164318,
+      "grad_norm": 0.07528689503669739,
+      "learning_rate": 1e-05,
+      "loss": 2.3872,
+      "step": 365
+    },
+    {
+      "epoch": 0.6160319798022301,
+      "grad_norm": 0.06814612448215485,
+      "learning_rate": 1e-05,
+      "loss": 2.2527,
+      "step": 366
+    },
+    {
+      "epoch": 0.6177151272880286,
+      "grad_norm": 0.06929857283830643,
+      "learning_rate": 1e-05,
+      "loss": 2.4138,
+      "step": 367
+    },
+    {
+      "epoch": 0.619398274773827,
+      "grad_norm": 0.07336314767599106,
+      "learning_rate": 1e-05,
+      "loss": 2.4197,
+      "step": 368
+    },
+    {
+      "epoch": 0.6210814222596255,
+      "grad_norm": 0.07009201496839523,
+      "learning_rate": 1e-05,
+      "loss": 2.3943,
+      "step": 369
+    },
+    {
+      "epoch": 0.6227645697454239,
+      "grad_norm": 0.07367721945047379,
+      "learning_rate": 1e-05,
+      "loss": 2.3044,
+      "step": 370
+    },
+    {
+      "epoch": 0.6244477172312224,
+      "grad_norm": 0.07029354572296143,
+      "learning_rate": 1e-05,
+      "loss": 2.3018,
+      "step": 371
+    },
+    {
+      "epoch": 0.6261308647170208,
+      "grad_norm": 0.07852700352668762,
+      "learning_rate": 1e-05,
+      "loss": 2.3727,
+      "step": 372
+    },
+    {
+      "epoch": 0.6278140122028193,
+      "grad_norm": 0.0764508917927742,
+      "learning_rate": 1e-05,
+      "loss": 2.1992,
+      "step": 373
+    },
+    {
+      "epoch": 0.6294971596886177,
+      "grad_norm": 0.0799420177936554,
+      "learning_rate": 1e-05,
+      "loss": 2.2693,
+      "step": 374
+    },
+    {
+      "epoch": 0.6311803071744162,
+      "grad_norm": 0.06878554075956345,
+      "learning_rate": 1e-05,
+      "loss": 2.4749,
+      "step": 375
+    },
+    {
+      "epoch": 0.6328634546602147,
+      "grad_norm": 0.07085944712162018,
+      "learning_rate": 1e-05,
+      "loss": 2.3435,
+      "step": 376
+    },
+    {
+      "epoch": 0.634546602146013,
+      "grad_norm": 0.06489285826683044,
+      "learning_rate": 1e-05,
+      "loss": 2.3257,
+      "step": 377
+    },
+    {
+      "epoch": 0.6362297496318114,
+      "grad_norm": 0.06664973497390747,
+      "learning_rate": 1e-05,
+      "loss": 2.5022,
+      "step": 378
+    },
+    {
+      "epoch": 0.6379128971176099,
+      "grad_norm": 0.07660377770662308,
+      "learning_rate": 1e-05,
+      "loss": 2.3269,
+      "step": 379
+    },
+    {
+      "epoch": 0.6395960446034084,
+      "grad_norm": 0.06934674084186554,
+      "learning_rate": 1e-05,
+      "loss": 2.4021,
+      "step": 380
+    },
+    {
+      "epoch": 0.6412791920892068,
+      "grad_norm": 0.07515530288219452,
+      "learning_rate": 1e-05,
+      "loss": 2.3157,
+      "step": 381
+    },
+    {
+      "epoch": 0.6429623395750053,
+      "grad_norm": 0.07302498072385788,
+      "learning_rate": 1e-05,
+      "loss": 2.3892,
+      "step": 382
+    },
+    {
+      "epoch": 0.6446454870608037,
+      "grad_norm": 0.07303425669670105,
+      "learning_rate": 1e-05,
+      "loss": 2.3765,
+      "step": 383
+    },
+    {
+      "epoch": 0.6463286345466022,
+      "grad_norm": 0.07705460488796234,
+      "learning_rate": 1e-05,
+      "loss": 2.2684,
+      "step": 384
+    },
+    {
+      "epoch": 0.6480117820324006,
+      "grad_norm": 0.07487067580223083,
+      "learning_rate": 1e-05,
+      "loss": 2.3733,
+      "step": 385
+    },
+    {
+      "epoch": 0.6496949295181991,
+      "grad_norm": 0.06538619101047516,
+      "learning_rate": 1e-05,
+      "loss": 2.3789,
+      "step": 386
+    },
+    {
+      "epoch": 0.6513780770039975,
+      "grad_norm": 0.07406684756278992,
+      "learning_rate": 1e-05,
+      "loss": 2.332,
+      "step": 387
+    },
+    {
+      "epoch": 0.6530612244897959,
+      "grad_norm": 0.07246539741754532,
+      "learning_rate": 1e-05,
+      "loss": 2.2302,
+      "step": 388
+    },
+    {
+      "epoch": 0.6547443719755943,
+      "grad_norm": 0.07304323464632034,
+      "learning_rate": 1e-05,
+      "loss": 2.3708,
+      "step": 389
+    },
+    {
+      "epoch": 0.6564275194613928,
+      "grad_norm": 0.07457181811332703,
+      "learning_rate": 1e-05,
+      "loss": 2.2991,
+      "step": 390
+    },
+    {
+      "epoch": 0.6581106669471912,
+      "grad_norm": 0.07300930470228195,
+      "learning_rate": 1e-05,
+      "loss": 2.2423,
+      "step": 391
+    },
+    {
+      "epoch": 0.6597938144329897,
+      "grad_norm": 0.07508236914873123,
+      "learning_rate": 1e-05,
+      "loss": 2.2642,
+      "step": 392
+    },
+    {
+      "epoch": 0.6614769619187881,
+      "grad_norm": 0.07481173425912857,
+      "learning_rate": 1e-05,
+      "loss": 2.3,
+      "step": 393
+    },
+    {
+      "epoch": 0.6631601094045866,
+      "grad_norm": 0.06851742416620255,
+      "learning_rate": 1e-05,
+      "loss": 2.4534,
+      "step": 394
+    },
+    {
+      "epoch": 0.664843256890385,
+      "grad_norm": 0.07536716759204865,
+      "learning_rate": 1e-05,
+      "loss": 2.3264,
+      "step": 395
+    },
+    {
+      "epoch": 0.6665264043761835,
+      "grad_norm": 0.07752048969268799,
+      "learning_rate": 1e-05,
+      "loss": 2.4158,
+      "step": 396
+    },
+    {
+      "epoch": 0.6682095518619819,
+      "grad_norm": 0.06357281655073166,
+      "learning_rate": 1e-05,
+      "loss": 2.4956,
+      "step": 397
+    },
+    {
+      "epoch": 0.6698926993477804,
+      "grad_norm": 0.08333004266023636,
+      "learning_rate": 1e-05,
+      "loss": 2.3921,
+      "step": 398
+    },
+    {
+      "epoch": 0.6715758468335787,
+      "grad_norm": 0.06873282790184021,
+      "learning_rate": 1e-05,
+      "loss": 2.3611,
+      "step": 399
+    },
+    {
+      "epoch": 0.6732589943193772,
+      "grad_norm": 0.07533644139766693,
+      "learning_rate": 1e-05,
+      "loss": 2.3708,
+      "step": 400
+    },
+    {
+      "epoch": 0.6749421418051756,
+      "grad_norm": 0.07756076753139496,
+      "learning_rate": 1e-05,
+      "loss": 2.3003,
+      "step": 401
+    },
+    {
+      "epoch": 0.6766252892909741,
+      "grad_norm": 0.06644177436828613,
+      "learning_rate": 1e-05,
+      "loss": 2.4331,
+      "step": 402
+    },
+    {
+      "epoch": 0.6783084367767725,
+      "grad_norm": 0.07512148469686508,
+      "learning_rate": 1e-05,
+      "loss": 2.2881,
+      "step": 403
+    },
+    {
+      "epoch": 0.679991584262571,
+      "grad_norm": 0.08939874172210693,
+      "learning_rate": 1e-05,
+      "loss": 2.1564,
+      "step": 404
+    },
+    {
+      "epoch": 0.6816747317483695,
+      "grad_norm": 0.07984601706266403,
+      "learning_rate": 1e-05,
+      "loss": 2.3967,
+      "step": 405
+    },
+    {
+      "epoch": 0.6833578792341679,
+      "grad_norm": 0.0724392980337143,
+      "learning_rate": 1e-05,
+      "loss": 2.2859,
+      "step": 406
+    },
+    {
+      "epoch": 0.6850410267199664,
+      "grad_norm": 0.07025589793920517,
+      "learning_rate": 1e-05,
+      "loss": 2.3027,
+      "step": 407
+    },
+    {
+      "epoch": 0.6867241742057648,
+      "grad_norm": 0.07863828539848328,
+      "learning_rate": 1e-05,
+      "loss": 2.3286,
+      "step": 408
+    },
+    {
+      "epoch": 0.6884073216915633,
+      "grad_norm": 0.07466793060302734,
+      "learning_rate": 1e-05,
+      "loss": 2.2849,
+      "step": 409
+    },
+    {
+      "epoch": 0.6900904691773617,
+      "grad_norm": 0.07291209697723389,
+      "learning_rate": 1e-05,
+      "loss": 2.3931,
+      "step": 410
+    },
+    {
+      "epoch": 0.6917736166631601,
+      "grad_norm": 0.072298564016819,
+      "learning_rate": 1e-05,
+      "loss": 2.377,
+      "step": 411
+    },
+    {
+      "epoch": 0.6934567641489585,
+      "grad_norm": 0.06996294856071472,
+      "learning_rate": 1e-05,
+      "loss": 2.3503,
+      "step": 412
+    },
+    {
+      "epoch": 0.695139911634757,
+      "grad_norm": 0.07319701462984085,
+      "learning_rate": 1e-05,
+      "loss": 2.345,
+      "step": 413
+    },
+    {
+      "epoch": 0.6968230591205554,
+      "grad_norm": 0.0768033117055893,
+      "learning_rate": 1e-05,
+      "loss": 2.3679,
+      "step": 414
+    },
+    {
+      "epoch": 0.6985062066063539,
+      "grad_norm": 0.07401002943515778,
+      "learning_rate": 1e-05,
+      "loss": 2.3435,
+      "step": 415
+    },
+    {
+      "epoch": 0.7001893540921523,
+      "grad_norm": 0.07700485736131668,
+      "learning_rate": 1e-05,
+      "loss": 2.3428,
+      "step": 416
+    },
+    {
+      "epoch": 0.7018725015779508,
+      "grad_norm": 0.07446201890707016,
+      "learning_rate": 1e-05,
+      "loss": 2.4133,
+      "step": 417
+    },
+    {
+      "epoch": 0.7035556490637492,
+      "grad_norm": 0.06801878660917282,
+      "learning_rate": 1e-05,
+      "loss": 2.3665,
+      "step": 418
+    },
+    {
+      "epoch": 0.7052387965495477,
+      "grad_norm": 0.07989214360713959,
+      "learning_rate": 1e-05,
+      "loss": 2.3303,
+      "step": 419
+    },
+    {
+      "epoch": 0.7069219440353461,
+      "grad_norm": 0.07385462522506714,
+      "learning_rate": 1e-05,
+      "loss": 2.3608,
+      "step": 420
+    },
+    {
+      "epoch": 0.7086050915211446,
+      "grad_norm": 0.06808451563119888,
+      "learning_rate": 1e-05,
+      "loss": 2.4851,
+      "step": 421
+    },
+    {
+      "epoch": 0.7102882390069429,
+      "grad_norm": 0.07354162633419037,
+      "learning_rate": 1e-05,
+      "loss": 2.3005,
+      "step": 422
+    },
+    {
+      "epoch": 0.7119713864927414,
+      "grad_norm": 0.07730504870414734,
+      "learning_rate": 1e-05,
+      "loss": 2.2815,
+      "step": 423
+    },
+    {
+      "epoch": 0.7136545339785398,
+      "grad_norm": 0.08045239001512527,
+      "learning_rate": 1e-05,
+      "loss": 2.2695,
+      "step": 424
+    },
+    {
+      "epoch": 0.7153376814643383,
+      "grad_norm": 0.07997512817382812,
+      "learning_rate": 1e-05,
+      "loss": 2.3608,
+      "step": 425
+    },
+    {
+      "epoch": 0.7170208289501367,
+      "grad_norm": 0.07076172530651093,
+      "learning_rate": 1e-05,
+      "loss": 2.3411,
+      "step": 426
+    },
+    {
+      "epoch": 0.7187039764359352,
+      "grad_norm": 0.07223929464817047,
+      "learning_rate": 1e-05,
+      "loss": 2.3452,
+      "step": 427
+    },
+    {
+      "epoch": 0.7203871239217337,
+      "grad_norm": 0.07667456567287445,
+      "learning_rate": 1e-05,
+      "loss": 2.333,
+      "step": 428
+    },
+    {
+      "epoch": 0.7220702714075321,
+      "grad_norm": 0.07509643584489822,
+      "learning_rate": 1e-05,
+      "loss": 2.3701,
+      "step": 429
+    },
+    {
+      "epoch": 0.7237534188933306,
+      "grad_norm": 0.08230644464492798,
+      "learning_rate": 1e-05,
+      "loss": 2.3577,
+      "step": 430
+    },
+    {
+      "epoch": 0.725436566379129,
+      "grad_norm": 0.06938886642456055,
+      "learning_rate": 1e-05,
+      "loss": 2.4573,
+      "step": 431
+    },
+    {
+      "epoch": 0.7271197138649275,
+      "grad_norm": 0.07415178418159485,
+      "learning_rate": 1e-05,
+      "loss": 2.2834,
+      "step": 432
+    },
+    {
+      "epoch": 0.7288028613507258,
+      "grad_norm": 0.0821278989315033,
+      "learning_rate": 1e-05,
+      "loss": 2.2744,
+      "step": 433
+    },
+    {
+      "epoch": 0.7304860088365243,
+      "grad_norm": 0.07293502986431122,
+      "learning_rate": 1e-05,
+      "loss": 2.313,
+      "step": 434
+    },
+    {
+      "epoch": 0.7321691563223227,
+      "grad_norm": 0.07829819619655609,
+      "learning_rate": 1e-05,
+      "loss": 2.3849,
+      "step": 435
+    },
+    {
+      "epoch": 0.7338523038081212,
+      "grad_norm": 0.07795297354459763,
+      "learning_rate": 1e-05,
+      "loss": 2.2466,
+      "step": 436
+    },
+    {
+      "epoch": 0.7355354512939196,
+      "grad_norm": 0.06956803798675537,
+      "learning_rate": 1e-05,
+      "loss": 2.4038,
+      "step": 437
+    },
+    {
+      "epoch": 0.7372185987797181,
+      "grad_norm": 0.07948347926139832,
+      "learning_rate": 1e-05,
+      "loss": 2.3042,
+      "step": 438
+    },
+    {
+      "epoch": 0.7389017462655165,
+      "grad_norm": 0.08074218034744263,
+      "learning_rate": 1e-05,
+      "loss": 2.3314,
+      "step": 439
+    },
+    {
+      "epoch": 0.740584893751315,
+      "grad_norm": 0.08029188960790634,
+      "learning_rate": 1e-05,
+      "loss": 2.312,
+      "step": 440
+    },
+    {
+      "epoch": 0.7422680412371134,
+      "grad_norm": 0.0783049538731575,
+      "learning_rate": 1e-05,
+      "loss": 2.307,
+      "step": 441
+    },
+    {
+      "epoch": 0.7439511887229119,
+      "grad_norm": 0.08203115314245224,
+      "learning_rate": 1e-05,
+      "loss": 2.3081,
+      "step": 442
+    },
+    {
+      "epoch": 0.7456343362087103,
+      "grad_norm": 0.08666986972093582,
+      "learning_rate": 1e-05,
+      "loss": 2.3721,
+      "step": 443
+    },
+    {
+      "epoch": 0.7473174836945087,
+      "grad_norm": 0.08097022771835327,
+      "learning_rate": 1e-05,
+      "loss": 2.1912,
+      "step": 444
+    },
+    {
+      "epoch": 0.7490006311803071,
+      "grad_norm": 0.08272138237953186,
+      "learning_rate": 1e-05,
+      "loss": 2.3562,
+      "step": 445
+    },
+    {
+      "epoch": 0.7506837786661056,
+      "grad_norm": 0.08114828914403915,
+      "learning_rate": 1e-05,
+      "loss": 2.3569,
+      "step": 446
+    },
+    {
+      "epoch": 0.752366926151904,
+      "grad_norm": 0.07786712795495987,
+      "learning_rate": 1e-05,
+      "loss": 2.3772,
+      "step": 447
+    },
+    {
+      "epoch": 0.7540500736377025,
+      "grad_norm": 0.07603191584348679,
+      "learning_rate": 1e-05,
+      "loss": 2.2748,
+      "step": 448
+    },
+    {
+      "epoch": 0.7557332211235009,
+      "grad_norm": 0.08364319056272507,
+      "learning_rate": 1e-05,
+      "loss": 2.334,
+      "step": 449
+    },
+    {
+      "epoch": 0.7574163686092994,
+      "grad_norm": 0.07968125492334366,
+      "learning_rate": 1e-05,
+      "loss": 2.3225,
+      "step": 450
+    },
+    {
+      "epoch": 0.7590995160950978,
+      "grad_norm": 0.08204993605613708,
+      "learning_rate": 1e-05,
+      "loss": 2.3107,
+      "step": 451
+    },
+    {
+      "epoch": 0.7607826635808963,
+      "grad_norm": 0.08319111168384552,
+      "learning_rate": 1e-05,
+      "loss": 2.3994,
+      "step": 452
+    },
+    {
+      "epoch": 0.7624658110666948,
+      "grad_norm": 0.07812530547380447,
+      "learning_rate": 1e-05,
+      "loss": 2.2771,
+      "step": 453
+    },
+    {
+      "epoch": 0.7641489585524932,
+      "grad_norm": 0.07962696999311447,
+      "learning_rate": 1e-05,
+      "loss": 2.3094,
+      "step": 454
+    },
+    {
+      "epoch": 0.7658321060382917,
+      "grad_norm": 0.0815802663564682,
+      "learning_rate": 1e-05,
+      "loss": 2.3169,
+      "step": 455
+    },
+    {
+      "epoch": 0.76751525352409,
+      "grad_norm": 0.08460783958435059,
+      "learning_rate": 1e-05,
+      "loss": 2.2443,
+      "step": 456
+    },
+    {
+      "epoch": 0.7691984010098885,
+      "grad_norm": 0.07976390421390533,
+      "learning_rate": 1e-05,
+      "loss": 2.26,
+      "step": 457
+    },
+    {
+      "epoch": 0.7708815484956869,
+      "grad_norm": 0.08143635839223862,
+      "learning_rate": 1e-05,
+      "loss": 2.2517,
+      "step": 458
+    },
+    {
+      "epoch": 0.7725646959814854,
+      "grad_norm": 0.08004558831453323,
+      "learning_rate": 1e-05,
+      "loss": 2.3276,
+      "step": 459
+    },
+    {
+      "epoch": 0.7742478434672838,
+      "grad_norm": 0.0831751599907875,
+      "learning_rate": 1e-05,
+      "loss": 2.2842,
+      "step": 460
+    },
+    {
+      "epoch": 0.7759309909530823,
+      "grad_norm": 0.07613930851221085,
+      "learning_rate": 1e-05,
+      "loss": 2.3958,
+      "step": 461
+    },
+    {
+      "epoch": 0.7776141384388807,
+      "grad_norm": 0.08161590993404388,
+      "learning_rate": 1e-05,
+      "loss": 2.3287,
+      "step": 462
+    },
+    {
+      "epoch": 0.7792972859246792,
+      "grad_norm": 0.08616164326667786,
+      "learning_rate": 1e-05,
+      "loss": 2.3098,
+      "step": 463
+    },
+    {
+      "epoch": 0.7809804334104776,
+      "grad_norm": 0.08720822632312775,
+      "learning_rate": 1e-05,
+      "loss": 2.1388,
+      "step": 464
+    },
+    {
+      "epoch": 0.7826635808962761,
+      "grad_norm": 0.08598899841308594,
+      "learning_rate": 1e-05,
+      "loss": 2.3005,
+      "step": 465
+    },
+    {
+      "epoch": 0.7843467283820745,
+      "grad_norm": 0.07982167601585388,
+      "learning_rate": 1e-05,
+      "loss": 2.3049,
+      "step": 466
+    },
+    {
+      "epoch": 0.7860298758678729,
+      "grad_norm": 0.08733374625444412,
+      "learning_rate": 1e-05,
+      "loss": 2.2747,
+      "step": 467
+    },
+    {
+      "epoch": 0.7877130233536713,
+      "grad_norm": 0.08848235011100769,
+      "learning_rate": 1e-05,
+      "loss": 2.4331,
+      "step": 468
+    },
+    {
+      "epoch": 0.7893961708394698,
+      "grad_norm": 0.08619164675474167,
+      "learning_rate": 1e-05,
+      "loss": 2.2881,
+      "step": 469
+    },
+    {
+      "epoch": 0.7910793183252682,
+      "grad_norm": 0.08046075701713562,
+      "learning_rate": 1e-05,
+      "loss": 2.397,
+      "step": 470
+    },
+    {
+      "epoch": 0.7927624658110667,
+      "grad_norm": 0.08469874411821365,
+      "learning_rate": 1e-05,
+      "loss": 2.3225,
+      "step": 471
+    },
+    {
+      "epoch": 0.7944456132968651,
+      "grad_norm": 0.08878640830516815,
+      "learning_rate": 1e-05,
+      "loss": 2.2832,
+      "step": 472
+    },
+    {
+      "epoch": 0.7961287607826636,
+      "grad_norm": 0.08530005067586899,
+      "learning_rate": 1e-05,
+      "loss": 2.28,
+      "step": 473
+    },
+    {
+      "epoch": 0.797811908268462,
+      "grad_norm": 0.08089161664247513,
+      "learning_rate": 1e-05,
+      "loss": 2.2822,
+      "step": 474
+    },
+    {
+      "epoch": 0.7994950557542605,
+      "grad_norm": 0.0770372822880745,
+      "learning_rate": 1e-05,
+      "loss": 2.4031,
+      "step": 475
+    },
+    {
+      "epoch": 0.801178203240059,
+      "grad_norm": 0.08313820511102676,
+      "learning_rate": 1e-05,
+      "loss": 2.4009,
+      "step": 476
+    },
+    {
+      "epoch": 0.8028613507258574,
+      "grad_norm": 0.08684401214122772,
+      "learning_rate": 1e-05,
+      "loss": 2.4563,
+      "step": 477
+    },
+    {
+      "epoch": 0.8045444982116557,
+      "grad_norm": 0.08352997899055481,
+      "learning_rate": 1e-05,
+      "loss": 2.3242,
+      "step": 478
+    },
+    {
+      "epoch": 0.8062276456974542,
+      "grad_norm": 0.08148252218961716,
+      "learning_rate": 1e-05,
+      "loss": 2.3096,
+      "step": 479
+    },
+    {
+      "epoch": 0.8079107931832527,
+      "grad_norm": 0.08157838881015778,
+      "learning_rate": 1e-05,
+      "loss": 2.3108,
+      "step": 480
+    },
+    {
+      "epoch": 0.8095939406690511,
+      "grad_norm": 0.08561182022094727,
+      "learning_rate": 1e-05,
+      "loss": 2.2327,
+      "step": 481
+    },
+    {
+      "epoch": 0.8112770881548496,
+      "grad_norm": 0.09177689999341965,
+      "learning_rate": 1e-05,
+      "loss": 2.2129,
+      "step": 482
+    },
+    {
+      "epoch": 0.812960235640648,
+      "grad_norm": 0.08262176811695099,
+      "learning_rate": 1e-05,
+      "loss": 2.397,
+      "step": 483
+    },
+    {
+      "epoch": 0.8146433831264465,
+      "grad_norm": 0.08541447669267654,
+      "learning_rate": 1e-05,
+      "loss": 2.2419,
+      "step": 484
+    },
+    {
+      "epoch": 0.8163265306122449,
+      "grad_norm": 0.08732729405164719,
+      "learning_rate": 1e-05,
+      "loss": 2.3328,
+      "step": 485
+    },
+    {
+      "epoch": 0.8180096780980434,
+      "grad_norm": 0.08658833056688309,
+      "learning_rate": 1e-05,
+      "loss": 2.2793,
+      "step": 486
+    },
+    {
+      "epoch": 0.8196928255838418,
+      "grad_norm": 0.0789208933711052,
+      "learning_rate": 1e-05,
+      "loss": 2.4072,
+      "step": 487
+    },
+    {
+      "epoch": 0.8213759730696403,
+      "grad_norm": 0.07870952039957047,
+      "learning_rate": 1e-05,
+      "loss": 2.4082,
+      "step": 488
+    },
+    {
+      "epoch": 0.8230591205554386,
+      "grad_norm": 0.07583601027727127,
+      "learning_rate": 1e-05,
+      "loss": 2.3833,
+      "step": 489
+    },
+    {
+      "epoch": 0.8247422680412371,
+      "grad_norm": 0.08982661366462708,
+      "learning_rate": 1e-05,
+      "loss": 2.2766,
+      "step": 490
+    },
+    {
+      "epoch": 0.8264254155270355,
+      "grad_norm": 0.08841705322265625,
+      "learning_rate": 1e-05,
+      "loss": 2.2581,
+      "step": 491
+    },
+    {
+      "epoch": 0.828108563012834,
+      "grad_norm": 0.08784886449575424,
+      "learning_rate": 1e-05,
+      "loss": 2.2352,
+      "step": 492
+    },
+    {
+      "epoch": 0.8297917104986324,
+      "grad_norm": 0.08765432238578796,
+      "learning_rate": 1e-05,
+      "loss": 2.1957,
+      "step": 493
+    },
+    {
+      "epoch": 0.8314748579844309,
+      "grad_norm": 0.09070983529090881,
+      "learning_rate": 1e-05,
+      "loss": 2.2451,
+      "step": 494
+    },
+    {
+      "epoch": 0.8331580054702293,
+      "grad_norm": 0.08307146281003952,
+      "learning_rate": 1e-05,
+      "loss": 2.3645,
+      "step": 495
+    },
+    {
+      "epoch": 0.8348411529560278,
+      "grad_norm": 0.07774417847394943,
+      "learning_rate": 1e-05,
+      "loss": 2.3921,
+      "step": 496
+    },
+    {
+      "epoch": 0.8365243004418262,
+      "grad_norm": 0.08441779762506485,
+      "learning_rate": 1e-05,
+      "loss": 2.2974,
+      "step": 497
+    },
+    {
+      "epoch": 0.8382074479276247,
+      "grad_norm": 0.08773106336593628,
+      "learning_rate": 1e-05,
+      "loss": 2.3984,
+      "step": 498
+    },
+    {
+      "epoch": 0.8398905954134231,
+      "grad_norm": 0.08157604187726974,
+      "learning_rate": 1e-05,
+      "loss": 2.2946,
+      "step": 499
+    },
+    {
+      "epoch": 0.8415737428992216,
+      "grad_norm": 0.09280236810445786,
+      "learning_rate": 1e-05,
+      "loss": 2.3628,
+      "step": 500
+    },
+    {
+      "epoch": 0.8432568903850199,
+      "grad_norm": 0.08737549185752869,
+      "learning_rate": 1e-05,
+      "loss": 2.2593,
+      "step": 501
+    },
+    {
+      "epoch": 0.8449400378708184,
+      "grad_norm": 0.08917705714702606,
+      "learning_rate": 1e-05,
+      "loss": 2.2435,
+      "step": 502
+    },
+    {
+      "epoch": 0.8466231853566168,
+      "grad_norm": 0.08589258790016174,
+      "learning_rate": 1e-05,
+      "loss": 2.2869,
+      "step": 503
+    },
+    {
+      "epoch": 0.8483063328424153,
+      "grad_norm": 0.08363740891218185,
+      "learning_rate": 1e-05,
+      "loss": 2.1512,
+      "step": 504
+    },
+    {
+      "epoch": 0.8499894803282138,
+      "grad_norm": 0.09710842370986938,
+      "learning_rate": 1e-05,
+      "loss": 2.3042,
+      "step": 505
+    },
+    {
+      "epoch": 0.8516726278140122,
+      "grad_norm": 0.09031599014997482,
+      "learning_rate": 1e-05,
+      "loss": 2.2406,
+      "step": 506
+    },
+    {
+      "epoch": 0.8533557752998107,
+      "grad_norm": 0.08941849321126938,
+      "learning_rate": 1e-05,
+      "loss": 2.2725,
+      "step": 507
+    },
+    {
+      "epoch": 0.8550389227856091,
+      "grad_norm": 0.08926845341920853,
+      "learning_rate": 1e-05,
+      "loss": 2.323,
+      "step": 508
+    },
+    {
+      "epoch": 0.8567220702714076,
+      "grad_norm": 0.08846578001976013,
+      "learning_rate": 1e-05,
+      "loss": 2.3394,
+      "step": 509
+    },
+    {
+      "epoch": 0.858405217757206,
+      "grad_norm": 0.08452317863702774,
+      "learning_rate": 1e-05,
+      "loss": 2.4158,
+      "step": 510
+    },
+    {
+      "epoch": 0.8600883652430045,
+      "grad_norm": 0.08531490713357925,
+      "learning_rate": 1e-05,
+      "loss": 2.3113,
+      "step": 511
+    },
+    {
+      "epoch": 0.8617715127288028,
+      "grad_norm": 0.08221501857042313,
+      "learning_rate": 1e-05,
+      "loss": 2.3826,
+      "step": 512
+    },
+    {
+      "epoch": 0.8634546602146013,
+      "grad_norm": 0.08809410035610199,
+      "learning_rate": 1e-05,
+      "loss": 2.2666,
+      "step": 513
+    },
+    {
+      "epoch": 0.8651378077003997,
+      "grad_norm": 0.0881451964378357,
+      "learning_rate": 1e-05,
+      "loss": 2.4678,
+      "step": 514
+    },
+    {
+      "epoch": 0.8668209551861982,
+      "grad_norm": 0.0958879366517067,
+      "learning_rate": 1e-05,
+      "loss": 2.17,
+      "step": 515
+    },
+    {
+      "epoch": 0.8685041026719966,
+      "grad_norm": 0.08498766273260117,
+      "learning_rate": 1e-05,
+      "loss": 2.4021,
+      "step": 516
+    },
+    {
+      "epoch": 0.8701872501577951,
+      "grad_norm": 0.09182509779930115,
+      "learning_rate": 1e-05,
+      "loss": 2.2476,
+      "step": 517
+    },
+    {
+      "epoch": 0.8718703976435935,
+      "grad_norm": 0.08831535279750824,
+      "learning_rate": 1e-05,
+      "loss": 2.3013,
+      "step": 518
+    },
+    {
+      "epoch": 0.873553545129392,
+      "grad_norm": 0.08792266249656677,
+      "learning_rate": 1e-05,
+      "loss": 2.2463,
+      "step": 519
+    },
+    {
+      "epoch": 0.8752366926151904,
+      "grad_norm": 0.0804978460073471,
+      "learning_rate": 1e-05,
+      "loss": 2.5151,
+      "step": 520
+    },
+    {
+      "epoch": 0.8769198401009889,
+      "grad_norm": 0.09397967159748077,
+      "learning_rate": 1e-05,
+      "loss": 2.2487,
+      "step": 521
+    },
+    {
+      "epoch": 0.8786029875867873,
+      "grad_norm": 0.08882005512714386,
+      "learning_rate": 1e-05,
+      "loss": 2.225,
+      "step": 522
+    },
+    {
+      "epoch": 0.8802861350725857,
+      "grad_norm": 0.08365931361913681,
+      "learning_rate": 1e-05,
+      "loss": 2.4277,
+      "step": 523
+    },
+    {
+      "epoch": 0.8819692825583841,
+      "grad_norm": 0.08842651546001434,
+      "learning_rate": 1e-05,
+      "loss": 2.3884,
+      "step": 524
+    },
+    {
+      "epoch": 0.8836524300441826,
+      "grad_norm": 0.08760154247283936,
+      "learning_rate": 1e-05,
+      "loss": 2.2576,
+      "step": 525
+    },
+    {
+      "epoch": 0.885335577529981,
+      "grad_norm": 0.07843348383903503,
+      "learning_rate": 1e-05,
+      "loss": 2.4143,
+      "step": 526
+    },
+    {
+      "epoch": 0.8870187250157795,
+      "grad_norm": 0.09312726557254791,
+      "learning_rate": 1e-05,
+      "loss": 2.2472,
+      "step": 527
+    },
+    {
+      "epoch": 0.888701872501578,
+      "grad_norm": 0.09460542351007462,
+      "learning_rate": 1e-05,
+      "loss": 2.2043,
+      "step": 528
+    },
+    {
+      "epoch": 0.8903850199873764,
+      "grad_norm": 0.09200920909643173,
+      "learning_rate": 1e-05,
+      "loss": 2.3562,
+      "step": 529
+    },
+    {
+      "epoch": 0.8920681674731749,
+      "grad_norm": 0.08051000535488129,
+      "learning_rate": 1e-05,
+      "loss": 2.4146,
+      "step": 530
+    },
+    {
+      "epoch": 0.8937513149589733,
+      "grad_norm": 0.09969057142734528,
+      "learning_rate": 1e-05,
+      "loss": 2.3342,
+      "step": 531
+    },
+    {
+      "epoch": 0.8954344624447718,
+      "grad_norm": 0.08616895228624344,
+      "learning_rate": 1e-05,
+      "loss": 2.3381,
+      "step": 532
+    },
+    {
+      "epoch": 0.8971176099305702,
+      "grad_norm": 0.09115055203437805,
+      "learning_rate": 1e-05,
+      "loss": 2.2377,
+      "step": 533
+    },
+    {
+      "epoch": 0.8988007574163686,
+      "grad_norm": 0.10309138149023056,
+      "learning_rate": 1e-05,
+      "loss": 2.1418,
+      "step": 534
+    },
+    {
+      "epoch": 0.900483904902167,
+      "grad_norm": 0.09327155351638794,
+      "learning_rate": 1e-05,
+      "loss": 2.312,
+      "step": 535
+    },
+    {
+      "epoch": 0.9021670523879655,
+      "grad_norm": 0.09104789048433304,
+      "learning_rate": 1e-05,
+      "loss": 2.2759,
+      "step": 536
+    },
+    {
+      "epoch": 0.9038501998737639,
+      "grad_norm": 0.08858876675367355,
+      "learning_rate": 1e-05,
+      "loss": 2.4138,
+      "step": 537
+    },
+    {
+      "epoch": 0.9055333473595624,
+      "grad_norm": 0.08850864320993423,
+      "learning_rate": 1e-05,
+      "loss": 2.3915,
+      "step": 538
+    },
+    {
+      "epoch": 0.9072164948453608,
+      "grad_norm": 0.09071122854948044,
+      "learning_rate": 1e-05,
+      "loss": 2.4199,
+      "step": 539
+    },
+    {
+      "epoch": 0.9088996423311593,
+      "grad_norm": 0.08702193200588226,
+      "learning_rate": 1e-05,
+      "loss": 2.3079,
+      "step": 540
+    },
+    {
+      "epoch": 0.9105827898169577,
+      "grad_norm": 0.09564194083213806,
+      "learning_rate": 1e-05,
+      "loss": 2.2996,
+      "step": 541
+    },
+    {
+      "epoch": 0.9122659373027562,
+      "grad_norm": 0.08906988054513931,
+      "learning_rate": 1e-05,
+      "loss": 2.3958,
+      "step": 542
+    },
+    {
+      "epoch": 0.9139490847885546,
+      "grad_norm": 0.08117242157459259,
+      "learning_rate": 1e-05,
+      "loss": 2.5557,
+      "step": 543
+    },
+    {
+      "epoch": 0.9156322322743531,
+      "grad_norm": 0.09870729595422745,
+      "learning_rate": 1e-05,
+      "loss": 2.3542,
+      "step": 544
+    },
+    {
+      "epoch": 0.9173153797601514,
+      "grad_norm": 0.0906287208199501,
+      "learning_rate": 1e-05,
+      "loss": 2.2866,
+      "step": 545
+    },
+    {
+      "epoch": 0.9189985272459499,
+      "grad_norm": 0.08649491518735886,
+      "learning_rate": 1e-05,
+      "loss": 2.3547,
+      "step": 546
+    },
+    {
+      "epoch": 0.9206816747317483,
+      "grad_norm": 0.09572413563728333,
+      "learning_rate": 1e-05,
+      "loss": 2.377,
+      "step": 547
+    },
+    {
+      "epoch": 0.9223648222175468,
+      "grad_norm": 0.08862059563398361,
+      "learning_rate": 1e-05,
+      "loss": 2.3452,
+      "step": 548
+    },
+    {
+      "epoch": 0.9240479697033452,
+      "grad_norm": 0.09061957150697708,
+      "learning_rate": 1e-05,
+      "loss": 2.264,
+      "step": 549
+    },
+    {
+      "epoch": 0.9257311171891437,
+      "grad_norm": 0.10327678918838501,
+      "learning_rate": 1e-05,
+      "loss": 2.3362,
+      "step": 550
+    },
+    {
+      "epoch": 0.9274142646749421,
+      "grad_norm": 0.10101998597383499,
+      "learning_rate": 1e-05,
+      "loss": 2.2091,
+      "step": 551
+    },
+    {
+      "epoch": 0.9290974121607406,
+      "grad_norm": 0.08099676668643951,
+      "learning_rate": 1e-05,
+      "loss": 2.3779,
+      "step": 552
+    },
+    {
+      "epoch": 0.930780559646539,
+      "grad_norm": 0.09572342783212662,
+      "learning_rate": 1e-05,
+      "loss": 2.2186,
+      "step": 553
+    },
+    {
+      "epoch": 0.9324637071323375,
+      "grad_norm": 0.10440348833799362,
+      "learning_rate": 1e-05,
+      "loss": 2.2717,
+      "step": 554
+    },
+    {
+      "epoch": 0.934146854618136,
+      "grad_norm": 0.09859239310026169,
+      "learning_rate": 1e-05,
+      "loss": 2.2964,
+      "step": 555
+    },
+    {
+      "epoch": 0.9358300021039344,
+      "grad_norm": 0.08539914339780807,
+      "learning_rate": 1e-05,
+      "loss": 2.3541,
+      "step": 556
+    },
+    {
+      "epoch": 0.9375131495897328,
+      "grad_norm": 0.09667155891656876,
+      "learning_rate": 1e-05,
+      "loss": 2.2412,
+      "step": 557
+    },
+    {
+      "epoch": 0.9391962970755312,
+      "grad_norm": 0.09381328523159027,
+      "learning_rate": 1e-05,
+      "loss": 2.1632,
+      "step": 558
+    },
+    {
+      "epoch": 0.9408794445613297,
+      "grad_norm": 0.10293637216091156,
+      "learning_rate": 1e-05,
+      "loss": 2.2969,
+      "step": 559
+    },
+    {
+      "epoch": 0.9425625920471281,
+      "grad_norm": 0.08901844918727875,
+      "learning_rate": 1e-05,
+      "loss": 2.2806,
+      "step": 560
+    },
+    {
+      "epoch": 0.9442457395329266,
+      "grad_norm": 0.09931071847677231,
+      "learning_rate": 1e-05,
+      "loss": 2.2671,
+      "step": 561
+    },
+    {
+      "epoch": 0.945928887018725,
+      "grad_norm": 0.08619210124015808,
+      "learning_rate": 1e-05,
+      "loss": 2.428,
+      "step": 562
+    },
+    {
+      "epoch": 0.9476120345045235,
+      "grad_norm": 0.08460855484008789,
+      "learning_rate": 1e-05,
+      "loss": 2.2412,
+      "step": 563
+    },
+    {
+      "epoch": 0.9492951819903219,
+      "grad_norm": 0.09682973474264145,
+      "learning_rate": 1e-05,
+      "loss": 2.3339,
+      "step": 564
+    },
+    {
+      "epoch": 0.9509783294761204,
+      "grad_norm": 0.10189709812402725,
+      "learning_rate": 1e-05,
+      "loss": 2.2268,
+      "step": 565
+    },
+    {
+      "epoch": 0.9526614769619188,
+      "grad_norm": 0.10271991789340973,
+      "learning_rate": 1e-05,
+      "loss": 2.1819,
+      "step": 566
+    },
+    {
+      "epoch": 0.9543446244477173,
+      "grad_norm": 0.0901963859796524,
+      "learning_rate": 1e-05,
+      "loss": 2.3029,
+      "step": 567
+    },
+    {
+      "epoch": 0.9560277719335156,
+      "grad_norm": 0.09148905426263809,
+      "learning_rate": 1e-05,
+      "loss": 2.3362,
+      "step": 568
+    },
+    {
+      "epoch": 0.9577109194193141,
+      "grad_norm": 0.10434332489967346,
+      "learning_rate": 1e-05,
+      "loss": 2.3037,
+      "step": 569
+    },
+    {
+      "epoch": 0.9593940669051125,
+      "grad_norm": 0.0956675261259079,
+      "learning_rate": 1e-05,
+      "loss": 2.3442,
+      "step": 570
+    },
+    {
+      "epoch": 0.961077214390911,
+      "grad_norm": 0.09394146502017975,
+      "learning_rate": 1e-05,
+      "loss": 2.2913,
+      "step": 571
+    },
+    {
+      "epoch": 0.9627603618767094,
+      "grad_norm": 0.09179794043302536,
+      "learning_rate": 1e-05,
+      "loss": 2.21,
+      "step": 572
+    },
+    {
+      "epoch": 0.9644435093625079,
+      "grad_norm": 0.09866604208946228,
+      "learning_rate": 1e-05,
+      "loss": 2.2721,
+      "step": 573
+    },
+    {
+      "epoch": 0.9661266568483063,
+      "grad_norm": 0.10069537162780762,
+      "learning_rate": 1e-05,
+      "loss": 2.1637,
+      "step": 574
+    },
+    {
+      "epoch": 0.9678098043341048,
+      "grad_norm": 0.0923682376742363,
+      "learning_rate": 1e-05,
+      "loss": 2.2343,
+      "step": 575
+    },
+    {
+      "epoch": 0.9694929518199032,
+      "grad_norm": 0.08836492151021957,
+      "learning_rate": 1e-05,
+      "loss": 2.3794,
+      "step": 576
+    },
+    {
+      "epoch": 0.9711760993057017,
+      "grad_norm": 0.0894513726234436,
+      "learning_rate": 1e-05,
+      "loss": 2.2378,
+      "step": 577
+    },
+    {
+      "epoch": 0.9728592467915002,
+      "grad_norm": 0.08647426962852478,
+      "learning_rate": 1e-05,
+      "loss": 2.3589,
+      "step": 578
+    },
+    {
+      "epoch": 0.9745423942772985,
+      "grad_norm": 0.11035202443599701,
+      "learning_rate": 1e-05,
+      "loss": 2.2371,
+      "step": 579
+    },
+    {
+      "epoch": 0.976225541763097,
+      "grad_norm": 0.09551876783370972,
+      "learning_rate": 1e-05,
+      "loss": 2.3353,
+      "step": 580
+    },
+    {
+      "epoch": 0.9779086892488954,
+      "grad_norm": 0.0911082923412323,
+      "learning_rate": 1e-05,
+      "loss": 2.3264,
+      "step": 581
+    },
+    {
+      "epoch": 0.9795918367346939,
+      "grad_norm": 0.10280529409646988,
+      "learning_rate": 1e-05,
+      "loss": 2.2351,
+      "step": 582
+    },
+    {
+      "epoch": 0.9812749842204923,
+      "grad_norm": 0.09424940496683121,
+      "learning_rate": 1e-05,
+      "loss": 2.3464,
+      "step": 583
+    },
+    {
+      "epoch": 0.9829581317062908,
+      "grad_norm": 0.092115618288517,
+      "learning_rate": 1e-05,
+      "loss": 2.2799,
+      "step": 584
+    },
+    {
+      "epoch": 0.9846412791920892,
+      "grad_norm": 0.09771659225225449,
+      "learning_rate": 1e-05,
+      "loss": 2.3777,
+      "step": 585
+    },
+    {
+      "epoch": 0.9863244266778877,
+      "grad_norm": 0.09877105802297592,
+      "learning_rate": 1e-05,
+      "loss": 2.3613,
+      "step": 586
+    },
+    {
+      "epoch": 0.9880075741636861,
+      "grad_norm": 0.09816967695951462,
+      "learning_rate": 1e-05,
+      "loss": 2.2925,
+      "step": 587
+    },
+    {
+      "epoch": 0.9896907216494846,
+      "grad_norm": 0.0874725803732872,
+      "learning_rate": 1e-05,
+      "loss": 2.3154,
+      "step": 588
+    },
+    {
+      "epoch": 0.991373869135283,
+      "grad_norm": 0.09336823225021362,
+      "learning_rate": 1e-05,
+      "loss": 2.3933,
+      "step": 589
+    },
+    {
+      "epoch": 0.9930570166210814,
+      "grad_norm": 0.10439187288284302,
+      "learning_rate": 1e-05,
+      "loss": 2.3655,
+      "step": 590
+    },
+    {
+      "epoch": 0.9947401641068798,
+      "grad_norm": 0.09005751460790634,
+      "learning_rate": 1e-05,
+      "loss": 2.2971,
+      "step": 591
+    },
+    {
+      "epoch": 0.9964233115926783,
+      "grad_norm": 0.10612068325281143,
+      "learning_rate": 1e-05,
+      "loss": 2.3584,
+      "step": 592
+    },
+    {
+      "epoch": 0.9981064590784767,
+      "grad_norm": 0.09101177752017975,
+      "learning_rate": 1e-05,
+      "loss": 2.4402,
+      "step": 593
+    },
+    {
+      "epoch": 0.9997896065642752,
+      "grad_norm": 0.09874800592660904,
+      "learning_rate": 1e-05,
+      "loss": 2.326,
+      "step": 594
+    },
+    {
+      "epoch": 1.0014727540500736,
+      "grad_norm": 0.1025647521018982,
+      "learning_rate": 1e-05,
+      "loss": 2.4041,
+      "step": 595
+    },
+    {
+      "epoch": 1.003155901535872,
+      "grad_norm": 0.11109832674264908,
+      "learning_rate": 1e-05,
+      "loss": 2.2881,
+      "step": 596
+    },
+    {
+      "epoch": 1.0048390490216705,
+      "grad_norm": 0.09670565277338028,
+      "learning_rate": 1e-05,
+      "loss": 2.2003,
+      "step": 597
+    },
+    {
+      "epoch": 1.0065221965074689,
+      "grad_norm": 0.09513822942972183,
+      "learning_rate": 1e-05,
+      "loss": 2.3225,
+      "step": 598
+    },
+    {
+      "epoch": 1.0082053439932674,
+      "grad_norm": 0.11121483892202377,
+      "learning_rate": 1e-05,
+      "loss": 2.4143,
+      "step": 599
+    },
+    {
+      "epoch": 1.0098884914790658,
+      "grad_norm": 0.09941378980875015,
+      "learning_rate": 1e-05,
+      "loss": 2.333,
+      "step": 600
+    },
+    {
+      "epoch": 1.0115716389648644,
+      "grad_norm": 0.09730757772922516,
+      "learning_rate": 1e-05,
+      "loss": 2.3638,
+      "step": 601
+    },
+    {
+      "epoch": 1.0132547864506627,
+      "grad_norm": 0.10626422613859177,
+      "learning_rate": 1e-05,
+      "loss": 2.2303,
+      "step": 602
+    },
+    {
+      "epoch": 1.0149379339364613,
+      "grad_norm": 0.0958971306681633,
+      "learning_rate": 1e-05,
+      "loss": 2.3906,
+      "step": 603
+    },
+    {
+      "epoch": 1.0166210814222596,
+      "grad_norm": 0.10065159201622009,
+      "learning_rate": 1e-05,
+      "loss": 2.3425,
+      "step": 604
+    },
+    {
+      "epoch": 1.0183042289080582,
+      "grad_norm": 0.08671624213457108,
+      "learning_rate": 1e-05,
+      "loss": 2.2742,
+      "step": 605
+    },
+    {
+      "epoch": 1.0199873763938565,
+      "grad_norm": 0.09528376907110214,
+      "learning_rate": 1e-05,
+      "loss": 2.3765,
+      "step": 606
+    },
+    {
+      "epoch": 1.0216705238796548,
+      "grad_norm": 0.09153752028942108,
+      "learning_rate": 1e-05,
+      "loss": 2.2983,
+      "step": 607
+    },
+    {
+      "epoch": 1.0233536713654534,
+      "grad_norm": 0.10145740956068039,
+      "learning_rate": 1e-05,
+      "loss": 2.1774,
+      "step": 608
+    },
+    {
+      "epoch": 1.0250368188512518,
+      "grad_norm": 0.09908965229988098,
+      "learning_rate": 1e-05,
+      "loss": 2.3479,
+      "step": 609
+    },
+    {
+      "epoch": 1.0267199663370503,
+      "grad_norm": 0.09253786504268646,
+      "learning_rate": 1e-05,
+      "loss": 2.3228,
+      "step": 610
+    },
+    {
+      "epoch": 1.0284031138228487,
+      "grad_norm": 0.094690702855587,
+      "learning_rate": 1e-05,
+      "loss": 2.2864,
+      "step": 611
+    },
+    {
+      "epoch": 1.0300862613086472,
+      "grad_norm": 0.09160283952951431,
+      "learning_rate": 1e-05,
+      "loss": 2.4285,
+      "step": 612
+    },
+    {
+      "epoch": 1.0317694087944456,
+      "grad_norm": 0.10157333314418793,
+      "learning_rate": 1e-05,
+      "loss": 2.1316,
+      "step": 613
+    },
+    {
+      "epoch": 1.0334525562802441,
+      "grad_norm": 0.10498999804258347,
+      "learning_rate": 1e-05,
+      "loss": 2.373,
+      "step": 614
+    },
+    {
+      "epoch": 1.0351357037660425,
+      "grad_norm": 0.09599211066961288,
+      "learning_rate": 1e-05,
+      "loss": 2.3511,
+      "step": 615
+    },
+    {
+      "epoch": 1.036818851251841,
+      "grad_norm": 0.1121436059474945,
+      "learning_rate": 1e-05,
+      "loss": 2.127,
+      "step": 616
+    },
+    {
+      "epoch": 1.0385019987376394,
+      "grad_norm": 0.10269173234701157,
+      "learning_rate": 1e-05,
+      "loss": 2.2659,
+      "step": 617
+    },
+    {
+      "epoch": 1.040185146223438,
+      "grad_norm": 0.0945139229297638,
+      "learning_rate": 1e-05,
+      "loss": 2.3281,
+      "step": 618
+    },
+    {
+      "epoch": 1.0418682937092363,
+      "grad_norm": 0.09318878501653671,
+      "learning_rate": 1e-05,
+      "loss": 2.3247,
+      "step": 619
+    },
+    {
+      "epoch": 1.0435514411950346,
+      "grad_norm": 0.10471779108047485,
+      "learning_rate": 1e-05,
+      "loss": 2.3098,
+      "step": 620
+    },
+    {
+      "epoch": 1.0452345886808332,
+      "grad_norm": 0.10514305531978607,
+      "learning_rate": 1e-05,
+      "loss": 2.3647,
+      "step": 621
+    },
+    {
+      "epoch": 1.0469177361666315,
+      "grad_norm": 0.09875541925430298,
+      "learning_rate": 1e-05,
+      "loss": 2.4204,
+      "step": 622
+    },
+    {
+      "epoch": 1.04860088365243,
+      "grad_norm": 0.10112539678812027,
+      "learning_rate": 1e-05,
+      "loss": 2.3269,
+      "step": 623
+    },
+    {
+      "epoch": 1.0502840311382284,
+      "grad_norm": 0.09719318896532059,
+      "learning_rate": 1e-05,
+      "loss": 2.3223,
+      "step": 624
+    },
+    {
+      "epoch": 1.051967178624027,
+      "grad_norm": 0.09615301340818405,
+      "learning_rate": 1e-05,
+      "loss": 2.2798,
+      "step": 625
+    },
+    {
+      "epoch": 1.0536503261098253,
+      "grad_norm": 0.09600812941789627,
+      "learning_rate": 1e-05,
+      "loss": 2.2738,
+      "step": 626
+    },
+    {
+      "epoch": 1.055333473595624,
+      "grad_norm": 0.09326303005218506,
+      "learning_rate": 1e-05,
+      "loss": 2.23,
+      "step": 627
+    },
+    {
+      "epoch": 1.0570166210814222,
+      "grad_norm": 0.09689430892467499,
+      "learning_rate": 1e-05,
+      "loss": 2.2582,
+      "step": 628
+    },
+    {
+      "epoch": 1.0586997685672208,
+      "grad_norm": 0.10389314591884613,
+      "learning_rate": 1e-05,
+      "loss": 2.3733,
+      "step": 629
+    },
+    {
+      "epoch": 1.0603829160530192,
+      "grad_norm": 0.09320785105228424,
+      "learning_rate": 1e-05,
+      "loss": 2.3315,
+      "step": 630
+    },
+    {
+      "epoch": 1.0620660635388175,
+      "grad_norm": 0.10638166218996048,
+      "learning_rate": 1e-05,
+      "loss": 2.4058,
+      "step": 631
+    },
+    {
+      "epoch": 1.063749211024616,
+      "grad_norm": 0.09525519609451294,
+      "learning_rate": 1e-05,
+      "loss": 2.2803,
+      "step": 632
+    },
+    {
+      "epoch": 1.0654323585104144,
+      "grad_norm": 0.09904535114765167,
+      "learning_rate": 1e-05,
+      "loss": 2.3613,
+      "step": 633
+    },
+    {
+      "epoch": 1.067115505996213,
+      "grad_norm": 0.10914106667041779,
+      "learning_rate": 1e-05,
+      "loss": 2.3955,
+      "step": 634
+    },
+    {
+      "epoch": 1.0687986534820113,
+      "grad_norm": 0.10424593091011047,
+      "learning_rate": 1e-05,
+      "loss": 2.2332,
+      "step": 635
+    },
+    {
+      "epoch": 1.0704818009678099,
+      "grad_norm": 0.10360780358314514,
+      "learning_rate": 1e-05,
+      "loss": 2.3127,
+      "step": 636
+    },
+    {
+      "epoch": 1.0721649484536082,
+      "grad_norm": 0.11223631352186203,
+      "learning_rate": 1e-05,
+      "loss": 2.201,
+      "step": 637
+    },
+    {
+      "epoch": 1.0738480959394068,
+      "grad_norm": 0.09491337090730667,
+      "learning_rate": 1e-05,
+      "loss": 2.3129,
+      "step": 638
+    },
+    {
+      "epoch": 1.0755312434252051,
+      "grad_norm": 0.09244826436042786,
+      "learning_rate": 1e-05,
+      "loss": 2.3728,
+      "step": 639
+    },
+    {
+      "epoch": 1.0772143909110037,
+      "grad_norm": 0.0922231450676918,
+      "learning_rate": 1e-05,
+      "loss": 2.3225,
+      "step": 640
+    },
+    {
+      "epoch": 1.078897538396802,
+      "grad_norm": 0.10818596929311752,
+      "learning_rate": 1e-05,
+      "loss": 2.3104,
+      "step": 641
+    },
+    {
+      "epoch": 1.0805806858826004,
+      "grad_norm": 0.09497258812189102,
+      "learning_rate": 1e-05,
+      "loss": 2.3176,
+      "step": 642
+    },
+    {
+      "epoch": 1.082263833368399,
+      "grad_norm": 0.10034379363059998,
+      "learning_rate": 1e-05,
+      "loss": 2.3943,
+      "step": 643
+    },
+    {
+      "epoch": 1.0839469808541973,
+      "grad_norm": 0.10024038702249527,
+      "learning_rate": 1e-05,
+      "loss": 2.3127,
+      "step": 644
+    },
+    {
+      "epoch": 1.0856301283399958,
+      "grad_norm": 0.10074039548635483,
+      "learning_rate": 1e-05,
+      "loss": 2.2351,
+      "step": 645
+    },
+    {
+      "epoch": 1.0873132758257942,
+      "grad_norm": 0.09631813317537308,
+      "learning_rate": 1e-05,
+      "loss": 2.3101,
+      "step": 646
+    },
+    {
+      "epoch": 1.0889964233115927,
+      "grad_norm": 0.10632781684398651,
+      "learning_rate": 1e-05,
+      "loss": 2.3669,
+      "step": 647
+    },
+    {
+      "epoch": 1.090679570797391,
+      "grad_norm": 0.10795175284147263,
+      "learning_rate": 1e-05,
+      "loss": 2.3064,
+      "step": 648
+    },
+    {
+      "epoch": 1.0923627182831896,
+      "grad_norm": 0.11120691895484924,
+      "learning_rate": 1e-05,
+      "loss": 2.2911,
+      "step": 649
+    },
+    {
+      "epoch": 1.094045865768988,
+      "grad_norm": 0.10034749656915665,
+      "learning_rate": 1e-05,
+      "loss": 2.3696,
+      "step": 650
+    },
+    {
+      "epoch": 1.0957290132547866,
+      "grad_norm": 0.10955310612916946,
+      "learning_rate": 1e-05,
+      "loss": 2.3464,
+      "step": 651
+    },
+    {
+      "epoch": 1.097412160740585,
+      "grad_norm": 0.09739572554826736,
+      "learning_rate": 1e-05,
+      "loss": 2.325,
+      "step": 652
+    },
+    {
+      "epoch": 1.0990953082263832,
+      "grad_norm": 0.10152111947536469,
+      "learning_rate": 1e-05,
+      "loss": 2.3745,
+      "step": 653
+    },
+    {
+      "epoch": 1.1007784557121818,
+      "grad_norm": 0.10103686153888702,
+      "learning_rate": 1e-05,
+      "loss": 2.3303,
+      "step": 654
+    },
+    {
+      "epoch": 1.1024616031979801,
+      "grad_norm": 0.1003558412194252,
+      "learning_rate": 1e-05,
+      "loss": 2.312,
+      "step": 655
+    },
+    {
+      "epoch": 1.1041447506837787,
+      "grad_norm": 0.10518987476825714,
+      "learning_rate": 1e-05,
+      "loss": 2.3444,
+      "step": 656
+    },
+    {
+      "epoch": 1.105827898169577,
+      "grad_norm": 0.09896016865968704,
+      "learning_rate": 1e-05,
+      "loss": 2.2532,
+      "step": 657
+    },
+    {
+      "epoch": 1.1075110456553756,
+      "grad_norm": 0.09725090116262436,
+      "learning_rate": 1e-05,
+      "loss": 2.3625,
+      "step": 658
+    },
+    {
+      "epoch": 1.109194193141174,
+      "grad_norm": 0.09022284299135208,
+      "learning_rate": 1e-05,
+      "loss": 2.3743,
+      "step": 659
+    },
+    {
+      "epoch": 1.1108773406269725,
+      "grad_norm": 0.10471490770578384,
+      "learning_rate": 1e-05,
+      "loss": 2.3416,
+      "step": 660
+    },
+    {
+      "epoch": 1.1125604881127709,
+      "grad_norm": 0.10991263389587402,
+      "learning_rate": 1e-05,
+      "loss": 2.3214,
+      "step": 661
+    },
+    {
+      "epoch": 1.1142436355985694,
+      "grad_norm": 0.10231148451566696,
+      "learning_rate": 1e-05,
+      "loss": 2.2832,
+      "step": 662
+    },
+    {
+      "epoch": 1.1159267830843678,
+      "grad_norm": 0.09433937072753906,
+      "learning_rate": 1e-05,
+      "loss": 2.2645,
+      "step": 663
+    },
+    {
+      "epoch": 1.117609930570166,
+      "grad_norm": 0.13238666951656342,
+      "learning_rate": 1e-05,
+      "loss": 2.2483,
+      "step": 664
+    },
+    {
+      "epoch": 1.1192930780559647,
+      "grad_norm": 0.10956214368343353,
+      "learning_rate": 1e-05,
+      "loss": 2.2321,
+      "step": 665
+    },
+    {
+      "epoch": 1.120976225541763,
+      "grad_norm": 0.11065597832202911,
+      "learning_rate": 1e-05,
+      "loss": 2.1869,
+      "step": 666
+    },
+    {
+      "epoch": 1.1226593730275616,
+      "grad_norm": 0.10971678793430328,
+      "learning_rate": 1e-05,
+      "loss": 2.1855,
+      "step": 667
+    },
+    {
+      "epoch": 1.12434252051336,
+      "grad_norm": 0.11080143600702286,
+      "learning_rate": 1e-05,
+      "loss": 2.198,
+      "step": 668
+    },
+    {
+      "epoch": 1.1260256679991585,
+      "grad_norm": 0.10381001979112625,
+      "learning_rate": 1e-05,
+      "loss": 2.3384,
+      "step": 669
+    },
+    {
+      "epoch": 1.1277088154849568,
+      "grad_norm": 0.1026921421289444,
+      "learning_rate": 1e-05,
+      "loss": 2.2458,
+      "step": 670
+    },
+    {
+      "epoch": 1.1293919629707554,
+      "grad_norm": 0.10585295408964157,
+      "learning_rate": 1e-05,
+      "loss": 2.1859,
+      "step": 671
+    },
+    {
+      "epoch": 1.1310751104565537,
+      "grad_norm": 0.10650487244129181,
+      "learning_rate": 1e-05,
+      "loss": 2.2662,
+      "step": 672
+    },
+    {
+      "epoch": 1.1327582579423523,
+      "grad_norm": 0.10717649012804031,
+      "learning_rate": 1e-05,
+      "loss": 2.3088,
+      "step": 673
+    },
+    {
+      "epoch": 1.1344414054281506,
+      "grad_norm": 0.10479724407196045,
+      "learning_rate": 1e-05,
+      "loss": 2.3042,
+      "step": 674
+    },
+    {
+      "epoch": 1.136124552913949,
+      "grad_norm": 0.10629065334796906,
+      "learning_rate": 1e-05,
+      "loss": 2.3481,
+      "step": 675
+    },
+    {
+      "epoch": 1.1378077003997475,
+      "grad_norm": 0.10375174880027771,
+      "learning_rate": 1e-05,
+      "loss": 2.3845,
+      "step": 676
+    },
+    {
+      "epoch": 1.1394908478855459,
+      "grad_norm": 0.10122872143983841,
+      "learning_rate": 1e-05,
+      "loss": 2.335,
+      "step": 677
+    },
+    {
+      "epoch": 1.1411739953713445,
+      "grad_norm": 0.09846247732639313,
+      "learning_rate": 1e-05,
+      "loss": 2.4028,
+      "step": 678
+    },
+    {
+      "epoch": 1.1428571428571428,
+      "grad_norm": 0.11501342058181763,
+      "learning_rate": 1e-05,
+      "loss": 2.2419,
+      "step": 679
+    },
+    {
+      "epoch": 1.1445402903429414,
+      "grad_norm": 0.11248493194580078,
+      "learning_rate": 1e-05,
+      "loss": 2.1294,
+      "step": 680
+    },
+    {
+      "epoch": 1.1462234378287397,
+      "grad_norm": 0.1141652762889862,
+      "learning_rate": 1e-05,
+      "loss": 2.2842,
+      "step": 681
+    },
+    {
+      "epoch": 1.1479065853145383,
+      "grad_norm": 0.10232444107532501,
+      "learning_rate": 1e-05,
+      "loss": 2.1798,
+      "step": 682
+    },
+    {
+      "epoch": 1.1495897328003366,
+      "grad_norm": 0.10624698549509048,
+      "learning_rate": 1e-05,
+      "loss": 2.2474,
+      "step": 683
+    },
+    {
+      "epoch": 1.1512728802861352,
+      "grad_norm": 0.10583934187889099,
+      "learning_rate": 1e-05,
+      "loss": 2.2917,
+      "step": 684
+    },
+    {
+      "epoch": 1.1529560277719335,
+      "grad_norm": 0.10667344182729721,
+      "learning_rate": 1e-05,
+      "loss": 2.2581,
+      "step": 685
+    },
+    {
+      "epoch": 1.1546391752577319,
+      "grad_norm": 0.10415381193161011,
+      "learning_rate": 1e-05,
+      "loss": 2.3325,
+      "step": 686
+    },
+    {
+      "epoch": 1.1563223227435304,
+      "grad_norm": 0.109574094414711,
+      "learning_rate": 1e-05,
+      "loss": 2.3306,
+      "step": 687
+    },
+    {
+      "epoch": 1.1580054702293288,
+      "grad_norm": 0.10537154227495193,
+      "learning_rate": 1e-05,
+      "loss": 2.3396,
+      "step": 688
+    },
+    {
+      "epoch": 1.1596886177151273,
+      "grad_norm": 0.10670781880617142,
+      "learning_rate": 1e-05,
+      "loss": 2.2518,
+      "step": 689
+    },
+    {
+      "epoch": 1.1613717652009257,
+      "grad_norm": 0.10296822339296341,
+      "learning_rate": 1e-05,
+      "loss": 2.3911,
+      "step": 690
+    },
+    {
+      "epoch": 1.1630549126867242,
+      "grad_norm": 0.10323610156774521,
+      "learning_rate": 1e-05,
+      "loss": 2.415,
+      "step": 691
+    },
+    {
+      "epoch": 1.1647380601725226,
+      "grad_norm": 0.09952528029680252,
+      "learning_rate": 1e-05,
+      "loss": 2.3674,
+      "step": 692
+    },
+    {
+      "epoch": 1.1664212076583211,
+      "grad_norm": 0.10683920234441757,
+      "learning_rate": 1e-05,
+      "loss": 2.1606,
+      "step": 693
+    },
+    {
+      "epoch": 1.1681043551441195,
+      "grad_norm": 0.10594907402992249,
+      "learning_rate": 1e-05,
+      "loss": 2.3633,
+      "step": 694
+    },
+    {
+      "epoch": 1.169787502629918,
+      "grad_norm": 0.1164483055472374,
+      "learning_rate": 1e-05,
+      "loss": 2.272,
+      "step": 695
+    },
+    {
+      "epoch": 1.1714706501157164,
+      "grad_norm": 0.1053275316953659,
+      "learning_rate": 1e-05,
+      "loss": 2.3361,
+      "step": 696
+    },
+    {
+      "epoch": 1.1731537976015147,
+      "grad_norm": 0.11722961068153381,
+      "learning_rate": 1e-05,
+      "loss": 2.1008,
+      "step": 697
+    },
+    {
+      "epoch": 1.1748369450873133,
+      "grad_norm": 0.11388476192951202,
+      "learning_rate": 1e-05,
+      "loss": 2.3129,
+      "step": 698
+    },
+    {
+      "epoch": 1.1765200925731116,
+      "grad_norm": 0.1149948239326477,
+      "learning_rate": 1e-05,
+      "loss": 2.3503,
+      "step": 699
+    },
+    {
+      "epoch": 1.1782032400589102,
+      "grad_norm": 0.09305736422538757,
+      "learning_rate": 1e-05,
+      "loss": 2.3811,
+      "step": 700
+    },
+    {
+      "epoch": 1.1798863875447085,
+      "grad_norm": 0.1027708575129509,
+      "learning_rate": 1e-05,
+      "loss": 2.3262,
+      "step": 701
+    },
+    {
+      "epoch": 1.181569535030507,
+      "grad_norm": 0.1058826595544815,
+      "learning_rate": 1e-05,
+      "loss": 2.2576,
+      "step": 702
+    },
+    {
+      "epoch": 1.1832526825163054,
+      "grad_norm": 0.1003696396946907,
+      "learning_rate": 1e-05,
+      "loss": 2.2759,
+      "step": 703
+    },
+    {
+      "epoch": 1.184935830002104,
+      "grad_norm": 0.11113473027944565,
+      "learning_rate": 1e-05,
+      "loss": 2.4163,
+      "step": 704
+    },
+    {
+      "epoch": 1.1866189774879023,
+      "grad_norm": 0.10945228487253189,
+      "learning_rate": 1e-05,
+      "loss": 2.2725,
+      "step": 705
+    },
+    {
+      "epoch": 1.188302124973701,
+      "grad_norm": 0.1079326868057251,
+      "learning_rate": 1e-05,
+      "loss": 2.3048,
+      "step": 706
+    },
+    {
+      "epoch": 1.1899852724594993,
+      "grad_norm": 0.10752802342176437,
+      "learning_rate": 1e-05,
+      "loss": 2.2145,
+      "step": 707
+    },
+    {
+      "epoch": 1.1916684199452976,
+      "grad_norm": 0.10588284581899643,
+      "learning_rate": 1e-05,
+      "loss": 2.3025,
+      "step": 708
+    },
+    {
+      "epoch": 1.1933515674310962,
+      "grad_norm": 0.1051083654165268,
+      "learning_rate": 1e-05,
+      "loss": 2.3198,
+      "step": 709
+    },
+    {
+      "epoch": 1.1950347149168945,
+      "grad_norm": 0.11915988475084305,
+      "learning_rate": 1e-05,
+      "loss": 2.2456,
+      "step": 710
+    },
+    {
+      "epoch": 1.196717862402693,
+      "grad_norm": 0.10947719216346741,
+      "learning_rate": 1e-05,
+      "loss": 2.3479,
+      "step": 711
+    },
+    {
+      "epoch": 1.1984010098884914,
+      "grad_norm": 0.11522776633501053,
+      "learning_rate": 1e-05,
+      "loss": 2.2898,
+      "step": 712
+    },
+    {
+      "epoch": 1.20008415737429,
+      "grad_norm": 0.10741020739078522,
+      "learning_rate": 1e-05,
+      "loss": 2.3198,
+      "step": 713
+    },
+    {
+      "epoch": 1.2017673048600883,
+      "grad_norm": 0.10589215159416199,
+      "learning_rate": 1e-05,
+      "loss": 2.2812,
+      "step": 714
+    },
+    {
+      "epoch": 1.2034504523458869,
+      "grad_norm": 0.10151232033967972,
+      "learning_rate": 1e-05,
+      "loss": 2.429,
+      "step": 715
+    },
+    {
+      "epoch": 1.2051335998316852,
+      "grad_norm": 0.11951622366905212,
+      "learning_rate": 1e-05,
+      "loss": 2.1932,
+      "step": 716
+    },
+    {
+      "epoch": 1.2068167473174838,
+      "grad_norm": 0.11722715198993683,
+      "learning_rate": 1e-05,
+      "loss": 2.2356,
+      "step": 717
+    },
+    {
+      "epoch": 1.2084998948032821,
+      "grad_norm": 0.11441315710544586,
+      "learning_rate": 1e-05,
+      "loss": 2.2891,
+      "step": 718
+    },
+    {
+      "epoch": 1.2101830422890805,
+      "grad_norm": 0.10936987400054932,
+      "learning_rate": 1e-05,
+      "loss": 2.2843,
+      "step": 719
+    },
+    {
+      "epoch": 1.211866189774879,
+      "grad_norm": 0.12374020367860794,
+      "learning_rate": 1e-05,
+      "loss": 2.2944,
+      "step": 720
+    },
+    {
+      "epoch": 1.2135493372606774,
+      "grad_norm": 0.11024117469787598,
+      "learning_rate": 1e-05,
+      "loss": 2.2595,
+      "step": 721
+    },
+    {
+      "epoch": 1.215232484746476,
+      "grad_norm": 0.09707245975732803,
+      "learning_rate": 1e-05,
+      "loss": 2.3867,
+      "step": 722
+    },
+    {
+      "epoch": 1.2169156322322743,
+      "grad_norm": 0.11022404581308365,
+      "learning_rate": 1e-05,
+      "loss": 2.375,
+      "step": 723
+    },
+    {
+      "epoch": 1.2185987797180728,
+      "grad_norm": 0.10732002556324005,
+      "learning_rate": 1e-05,
+      "loss": 2.3674,
+      "step": 724
+    },
+    {
+      "epoch": 1.2202819272038712,
+      "grad_norm": 0.11548677086830139,
+      "learning_rate": 1e-05,
+      "loss": 2.3284,
+      "step": 725
+    },
+    {
+      "epoch": 1.2219650746896698,
+      "grad_norm": 0.10313412547111511,
+      "learning_rate": 1e-05,
+      "loss": 2.4128,
+      "step": 726
+    },
+    {
+      "epoch": 1.223648222175468,
+      "grad_norm": 0.12717945873737335,
+      "learning_rate": 1e-05,
+      "loss": 2.2847,
+      "step": 727
+    },
+    {
+      "epoch": 1.2253313696612667,
+      "grad_norm": 0.11565182358026505,
+      "learning_rate": 1e-05,
+      "loss": 2.2695,
+      "step": 728
+    },
+    {
+      "epoch": 1.227014517147065,
+      "grad_norm": 0.10489466041326523,
+      "learning_rate": 1e-05,
+      "loss": 2.3394,
+      "step": 729
+    },
+    {
+      "epoch": 1.2286976646328633,
+      "grad_norm": 0.11056289076805115,
+      "learning_rate": 1e-05,
+      "loss": 2.4165,
+      "step": 730
+    },
+    {
+      "epoch": 1.230380812118662,
+      "grad_norm": 0.12048956751823425,
+      "learning_rate": 1e-05,
+      "loss": 2.2289,
+      "step": 731
+    },
+    {
+      "epoch": 1.2320639596044602,
+      "grad_norm": 0.10263136774301529,
+      "learning_rate": 1e-05,
+      "loss": 2.3306,
+      "step": 732
+    },
+    {
+      "epoch": 1.2337471070902588,
+      "grad_norm": 0.11179950088262558,
+      "learning_rate": 1e-05,
+      "loss": 2.3481,
+      "step": 733
+    },
+    {
+      "epoch": 1.2354302545760572,
+      "grad_norm": 0.10484311729669571,
+      "learning_rate": 1e-05,
+      "loss": 2.2703,
+      "step": 734
+    },
+    {
+      "epoch": 1.2371134020618557,
+      "grad_norm": 0.1182483434677124,
+      "learning_rate": 1e-05,
+      "loss": 2.2328,
+      "step": 735
+    },
+    {
+      "epoch": 1.238796549547654,
+      "grad_norm": 0.11377429217100143,
+      "learning_rate": 1e-05,
+      "loss": 2.3657,
+      "step": 736
+    },
+    {
+      "epoch": 1.2404796970334526,
+      "grad_norm": 0.11151503771543503,
+      "learning_rate": 1e-05,
+      "loss": 2.3542,
+      "step": 737
+    },
+    {
+      "epoch": 1.242162844519251,
+      "grad_norm": 0.12628555297851562,
+      "learning_rate": 1e-05,
+      "loss": 2.2634,
+      "step": 738
+    },
+    {
+      "epoch": 1.2438459920050495,
+      "grad_norm": 0.10311713814735413,
+      "learning_rate": 1e-05,
+      "loss": 2.2717,
+      "step": 739
+    },
+    {
+      "epoch": 1.2455291394908479,
+      "grad_norm": 0.12768767774105072,
+      "learning_rate": 1e-05,
+      "loss": 2.1725,
+      "step": 740
+    },
+    {
+      "epoch": 1.2472122869766462,
+      "grad_norm": 0.12390502542257309,
+      "learning_rate": 1e-05,
+      "loss": 2.1708,
+      "step": 741
+    },
+    {
+      "epoch": 1.2488954344624448,
+      "grad_norm": 0.10566207021474838,
+      "learning_rate": 1e-05,
+      "loss": 2.3469,
+      "step": 742
+    },
+    {
+      "epoch": 1.2505785819482433,
+      "grad_norm": 0.10176009684801102,
+      "learning_rate": 1e-05,
+      "loss": 2.3159,
+      "step": 743
+    },
+    {
+      "epoch": 1.2522617294340417,
+      "grad_norm": 0.10881732404232025,
+      "learning_rate": 1e-05,
+      "loss": 2.2966,
+      "step": 744
+    },
+    {
+      "epoch": 1.25394487691984,
+      "grad_norm": 0.11917608976364136,
+      "learning_rate": 1e-05,
+      "loss": 2.395,
+      "step": 745
+    },
+    {
+      "epoch": 1.2556280244056386,
+      "grad_norm": 0.09600858390331268,
+      "learning_rate": 1e-05,
+      "loss": 2.3479,
+      "step": 746
+    },
+    {
+      "epoch": 1.257311171891437,
+      "grad_norm": 0.11550504714250565,
+      "learning_rate": 1e-05,
+      "loss": 2.301,
+      "step": 747
+    },
+    {
+      "epoch": 1.2589943193772355,
+      "grad_norm": 0.10588584840297699,
+      "learning_rate": 1e-05,
+      "loss": 2.4163,
+      "step": 748
+    },
+    {
+      "epoch": 1.2606774668630338,
+      "grad_norm": 0.10998673737049103,
+      "learning_rate": 1e-05,
+      "loss": 2.3379,
+      "step": 749
+    },
+    {
+      "epoch": 1.2623606143488324,
+      "grad_norm": 0.10513128340244293,
+      "learning_rate": 1e-05,
+      "loss": 2.3795,
+      "step": 750
+    },
+    {
+      "epoch": 1.2640437618346307,
+      "grad_norm": 0.11185754835605621,
+      "learning_rate": 1e-05,
+      "loss": 2.2583,
+      "step": 751
+    },
+    {
+      "epoch": 1.265726909320429,
+      "grad_norm": 0.10794227570295334,
+      "learning_rate": 1e-05,
+      "loss": 2.285,
+      "step": 752
+    },
+    {
+      "epoch": 1.2674100568062276,
+      "grad_norm": 0.12522459030151367,
+      "learning_rate": 1e-05,
+      "loss": 2.2292,
+      "step": 753
+    },
+    {
+      "epoch": 1.2690932042920262,
+      "grad_norm": 0.11628364026546478,
+      "learning_rate": 1e-05,
+      "loss": 2.3342,
+      "step": 754
+    },
+    {
+      "epoch": 1.2707763517778246,
+      "grad_norm": 0.12842795252799988,
+      "learning_rate": 1e-05,
+      "loss": 2.1455,
+      "step": 755
+    },
+    {
+      "epoch": 1.272459499263623,
+      "grad_norm": 0.11268262565135956,
+      "learning_rate": 1e-05,
+      "loss": 2.2241,
+      "step": 756
+    },
+    {
+      "epoch": 1.2741426467494215,
+      "grad_norm": 0.11674508452415466,
+      "learning_rate": 1e-05,
+      "loss": 2.2677,
+      "step": 757
+    },
+    {
+      "epoch": 1.2758257942352198,
+      "grad_norm": 0.11475373059511185,
+      "learning_rate": 1e-05,
+      "loss": 2.4075,
+      "step": 758
+    },
+    {
+      "epoch": 1.2775089417210184,
+      "grad_norm": 0.11378497630357742,
+      "learning_rate": 1e-05,
+      "loss": 2.3032,
+      "step": 759
+    },
+    {
+      "epoch": 1.2791920892068167,
+      "grad_norm": 0.10426255315542221,
+      "learning_rate": 1e-05,
+      "loss": 2.2488,
+      "step": 760
+    },
+    {
+      "epoch": 1.2808752366926153,
+      "grad_norm": 0.11820263415575027,
+      "learning_rate": 1e-05,
+      "loss": 2.197,
+      "step": 761
+    },
+    {
+      "epoch": 1.2825583841784136,
+      "grad_norm": 0.10741489380598068,
+      "learning_rate": 1e-05,
+      "loss": 2.2811,
+      "step": 762
+    },
+    {
+      "epoch": 1.284241531664212,
+      "grad_norm": 0.115534208714962,
+      "learning_rate": 1e-05,
+      "loss": 2.3105,
+      "step": 763
+    },
+    {
+      "epoch": 1.2859246791500105,
+      "grad_norm": 0.1159248948097229,
+      "learning_rate": 1e-05,
+      "loss": 2.2963,
+      "step": 764
+    },
+    {
+      "epoch": 1.287607826635809,
+      "grad_norm": 0.11940732598304749,
+      "learning_rate": 1e-05,
+      "loss": 2.3274,
+      "step": 765
+    },
+    {
+      "epoch": 1.2892909741216074,
+      "grad_norm": 0.11882008612155914,
+      "learning_rate": 1e-05,
+      "loss": 2.2405,
+      "step": 766
+    },
+    {
+      "epoch": 1.2909741216074058,
+      "grad_norm": 0.10939499735832214,
+      "learning_rate": 1e-05,
+      "loss": 2.3008,
+      "step": 767
+    },
+    {
+      "epoch": 1.2926572690932043,
+      "grad_norm": 0.11414020508527756,
+      "learning_rate": 1e-05,
+      "loss": 2.3164,
+      "step": 768
+    },
+    {
+      "epoch": 1.2943404165790027,
+      "grad_norm": 0.11446741968393326,
+      "learning_rate": 1e-05,
+      "loss": 2.2524,
+      "step": 769
+    },
+    {
+      "epoch": 1.2960235640648012,
+      "grad_norm": 0.12233757227659225,
+      "learning_rate": 1e-05,
+      "loss": 2.3997,
+      "step": 770
+    },
+    {
+      "epoch": 1.2977067115505996,
+      "grad_norm": 0.11746780574321747,
+      "learning_rate": 1e-05,
+      "loss": 2.2241,
+      "step": 771
+    },
+    {
+      "epoch": 1.2993898590363981,
+      "grad_norm": 0.12653754651546478,
+      "learning_rate": 1e-05,
+      "loss": 2.2181,
+      "step": 772
+    },
+    {
+      "epoch": 1.3010730065221965,
+      "grad_norm": 0.11092430353164673,
+      "learning_rate": 1e-05,
+      "loss": 2.194,
+      "step": 773
+    },
+    {
+      "epoch": 1.3027561540079948,
+      "grad_norm": 0.11273445188999176,
+      "learning_rate": 1e-05,
+      "loss": 2.2821,
+      "step": 774
+    },
+    {
+      "epoch": 1.3044393014937934,
+      "grad_norm": 0.10755831003189087,
+      "learning_rate": 1e-05,
+      "loss": 2.3381,
+      "step": 775
+    },
+    {
+      "epoch": 1.306122448979592,
+      "grad_norm": 0.10324183851480484,
+      "learning_rate": 1e-05,
+      "loss": 2.4531,
+      "step": 776
+    },
+    {
+      "epoch": 1.3078055964653903,
+      "grad_norm": 0.1238187626004219,
+      "learning_rate": 1e-05,
+      "loss": 2.2378,
+      "step": 777
+    },
+    {
+      "epoch": 1.3094887439511886,
+      "grad_norm": 0.10919329524040222,
+      "learning_rate": 1e-05,
+      "loss": 2.3157,
+      "step": 778
+    },
+    {
+      "epoch": 1.3111718914369872,
+      "grad_norm": 0.11661651730537415,
+      "learning_rate": 1e-05,
+      "loss": 2.3889,
+      "step": 779
+    },
+    {
+      "epoch": 1.3128550389227855,
+      "grad_norm": 0.11324804276227951,
+      "learning_rate": 1e-05,
+      "loss": 2.366,
+      "step": 780
+    },
+    {
+      "epoch": 1.314538186408584,
+      "grad_norm": 0.11539211124181747,
+      "learning_rate": 1e-05,
+      "loss": 2.2661,
+      "step": 781
+    },
+    {
+      "epoch": 1.3162213338943825,
+      "grad_norm": 0.12013803422451019,
+      "learning_rate": 1e-05,
+      "loss": 2.2388,
+      "step": 782
+    },
+    {
+      "epoch": 1.317904481380181,
+      "grad_norm": 0.1297876238822937,
+      "learning_rate": 1e-05,
+      "loss": 2.338,
+      "step": 783
+    },
+    {
+      "epoch": 1.3195876288659794,
+      "grad_norm": 0.11792443692684174,
+      "learning_rate": 1e-05,
+      "loss": 2.3162,
+      "step": 784
+    },
+    {
+      "epoch": 1.3212707763517777,
+      "grad_norm": 0.11543410271406174,
+      "learning_rate": 1e-05,
+      "loss": 2.325,
+      "step": 785
+    },
+    {
+      "epoch": 1.3229539238375763,
+      "grad_norm": 0.11507069319486618,
+      "learning_rate": 1e-05,
+      "loss": 2.3389,
+      "step": 786
+    },
+    {
+      "epoch": 1.3246370713233748,
+      "grad_norm": 0.11883421987295151,
+      "learning_rate": 1e-05,
+      "loss": 2.3784,
+      "step": 787
+    },
+    {
+      "epoch": 1.3263202188091732,
+      "grad_norm": 0.11997753381729126,
+      "learning_rate": 1e-05,
+      "loss": 2.2183,
+      "step": 788
+    },
+    {
+      "epoch": 1.3280033662949715,
+      "grad_norm": 0.12312667816877365,
+      "learning_rate": 1e-05,
+      "loss": 2.2661,
+      "step": 789
+    },
+    {
+      "epoch": 1.32968651378077,
+      "grad_norm": 0.1280994415283203,
+      "learning_rate": 1e-05,
+      "loss": 2.235,
+      "step": 790
+    },
+    {
+      "epoch": 1.3313696612665684,
+      "grad_norm": 0.12460897862911224,
+      "learning_rate": 1e-05,
+      "loss": 2.2775,
+      "step": 791
+    },
+    {
+      "epoch": 1.333052808752367,
+      "grad_norm": 0.11441405862569809,
+      "learning_rate": 1e-05,
+      "loss": 2.2642,
+      "step": 792
+    },
+    {
+      "epoch": 1.3347359562381653,
+      "grad_norm": 0.1078685000538826,
+      "learning_rate": 1e-05,
+      "loss": 2.3174,
+      "step": 793
+    },
+    {
+      "epoch": 1.3364191037239639,
+      "grad_norm": 0.11945922672748566,
+      "learning_rate": 1e-05,
+      "loss": 2.3101,
+      "step": 794
+    },
+    {
+      "epoch": 1.3381022512097622,
+      "grad_norm": 0.11506087332963943,
+      "learning_rate": 1e-05,
+      "loss": 2.3167,
+      "step": 795
+    },
+    {
+      "epoch": 1.3397853986955606,
+      "grad_norm": 0.12365138530731201,
+      "learning_rate": 1e-05,
+      "loss": 2.3044,
+      "step": 796
+    },
+    {
+      "epoch": 1.3414685461813591,
+      "grad_norm": 0.12331211566925049,
+      "learning_rate": 1e-05,
+      "loss": 2.2058,
+      "step": 797
+    },
+    {
+      "epoch": 1.3431516936671577,
+      "grad_norm": 0.12298640608787537,
+      "learning_rate": 1e-05,
+      "loss": 2.21,
+      "step": 798
+    },
+    {
+      "epoch": 1.344834841152956,
+      "grad_norm": 0.12047012150287628,
+      "learning_rate": 1e-05,
+      "loss": 2.2781,
+      "step": 799
+    },
+    {
+      "epoch": 1.3465179886387544,
+      "grad_norm": 0.12428031861782074,
+      "learning_rate": 1e-05,
+      "loss": 2.3032,
+      "step": 800
+    },
+    {
+      "epoch": 1.348201136124553,
+      "grad_norm": 0.1128249540925026,
+      "learning_rate": 1e-05,
+      "loss": 2.3135,
+      "step": 801
+    },
+    {
+      "epoch": 1.3498842836103513,
+      "grad_norm": 0.12616464495658875,
+      "learning_rate": 1e-05,
+      "loss": 2.1487,
+      "step": 802
+    },
+    {
+      "epoch": 1.3515674310961499,
+      "grad_norm": 0.11388704925775528,
+      "learning_rate": 1e-05,
+      "loss": 2.2346,
+      "step": 803
+    },
+    {
+      "epoch": 1.3532505785819482,
+      "grad_norm": 0.10213828831911087,
+      "learning_rate": 1e-05,
+      "loss": 2.2859,
+      "step": 804
+    },
+    {
+      "epoch": 1.3549337260677468,
+      "grad_norm": 0.1226121038198471,
+      "learning_rate": 1e-05,
+      "loss": 2.2183,
+      "step": 805
+    },
+    {
+      "epoch": 1.356616873553545,
+      "grad_norm": 0.11445735394954681,
+      "learning_rate": 1e-05,
+      "loss": 2.3784,
+      "step": 806
+    },
+    {
+      "epoch": 1.3583000210393434,
+      "grad_norm": 0.11648505181074142,
+      "learning_rate": 1e-05,
+      "loss": 2.3442,
+      "step": 807
+    },
+    {
+      "epoch": 1.359983168525142,
+      "grad_norm": 0.1296563744544983,
+      "learning_rate": 1e-05,
+      "loss": 2.2469,
+      "step": 808
+    },
+    {
+      "epoch": 1.3616663160109406,
+      "grad_norm": 0.12322400510311127,
+      "learning_rate": 1e-05,
+      "loss": 2.2915,
+      "step": 809
+    },
+    {
+      "epoch": 1.363349463496739,
+      "grad_norm": 0.11419309675693512,
+      "learning_rate": 1e-05,
+      "loss": 2.3024,
+      "step": 810
+    },
+    {
+      "epoch": 1.3650326109825373,
+      "grad_norm": 0.12253374606370926,
+      "learning_rate": 1e-05,
+      "loss": 2.2969,
+      "step": 811
+    },
+    {
+      "epoch": 1.3667157584683358,
+      "grad_norm": 0.1254422962665558,
+      "learning_rate": 1e-05,
+      "loss": 2.364,
+      "step": 812
+    },
+    {
+      "epoch": 1.3683989059541342,
+      "grad_norm": 0.12984994053840637,
+      "learning_rate": 1e-05,
+      "loss": 2.2936,
+      "step": 813
+    },
+    {
+      "epoch": 1.3700820534399327,
+      "grad_norm": 0.1182006224989891,
+      "learning_rate": 1e-05,
+      "loss": 2.2673,
+      "step": 814
+    },
+    {
+      "epoch": 1.371765200925731,
+      "grad_norm": 0.12920832633972168,
+      "learning_rate": 1e-05,
+      "loss": 2.1582,
+      "step": 815
+    },
+    {
+      "epoch": 1.3734483484115296,
+      "grad_norm": 0.1216689869761467,
+      "learning_rate": 1e-05,
+      "loss": 2.3479,
+      "step": 816
+    },
+    {
+      "epoch": 1.375131495897328,
+      "grad_norm": 0.12459319084882736,
+      "learning_rate": 1e-05,
+      "loss": 2.1868,
+      "step": 817
+    },
+    {
+      "epoch": 1.3768146433831263,
+      "grad_norm": 0.11144936084747314,
+      "learning_rate": 1e-05,
+      "loss": 2.3663,
+      "step": 818
+    },
+    {
+      "epoch": 1.3784977908689249,
+      "grad_norm": 0.1110294982790947,
+      "learning_rate": 1e-05,
+      "loss": 2.3164,
+      "step": 819
+    },
+    {
+      "epoch": 1.3801809383547234,
+      "grad_norm": 0.11903022974729538,
+      "learning_rate": 1e-05,
+      "loss": 2.2589,
+      "step": 820
+    },
+    {
+      "epoch": 1.3818640858405218,
+      "grad_norm": 0.10610275715589523,
+      "learning_rate": 1e-05,
+      "loss": 2.4153,
+      "step": 821
+    },
+    {
+      "epoch": 1.3835472333263201,
+      "grad_norm": 0.11972808837890625,
+      "learning_rate": 1e-05,
+      "loss": 2.3901,
+      "step": 822
+    },
+    {
+      "epoch": 1.3852303808121187,
+      "grad_norm": 0.10772975534200668,
+      "learning_rate": 1e-05,
+      "loss": 2.3555,
+      "step": 823
+    },
+    {
+      "epoch": 1.386913528297917,
+      "grad_norm": 0.11757270246744156,
+      "learning_rate": 1e-05,
+      "loss": 2.2677,
+      "step": 824
+    },
+    {
+      "epoch": 1.3885966757837156,
+      "grad_norm": 0.1217508539557457,
+      "learning_rate": 1e-05,
+      "loss": 2.2267,
+      "step": 825
+    },
+    {
+      "epoch": 1.390279823269514,
+      "grad_norm": 0.10996967554092407,
+      "learning_rate": 1e-05,
+      "loss": 2.3965,
+      "step": 826
+    },
+    {
+      "epoch": 1.3919629707553125,
+      "grad_norm": 0.13068005442619324,
+      "learning_rate": 1e-05,
+      "loss": 2.1991,
+      "step": 827
+    },
+    {
+      "epoch": 1.3936461182411108,
+      "grad_norm": 0.12149260193109512,
+      "learning_rate": 1e-05,
+      "loss": 2.2775,
+      "step": 828
+    },
+    {
+      "epoch": 1.3953292657269092,
+      "grad_norm": 0.1100870743393898,
+      "learning_rate": 1e-05,
+      "loss": 2.2571,
+      "step": 829
+    },
+    {
+      "epoch": 1.3970124132127077,
+      "grad_norm": 0.10005280375480652,
+      "learning_rate": 1e-05,
+      "loss": 2.2808,
+      "step": 830
+    },
+    {
+      "epoch": 1.3986955606985063,
+      "grad_norm": 0.11633820086717606,
+      "learning_rate": 1e-05,
+      "loss": 2.3215,
+      "step": 831
+    },
+    {
+      "epoch": 1.4003787081843047,
+      "grad_norm": 0.11901983618736267,
+      "learning_rate": 1e-05,
+      "loss": 2.4236,
+      "step": 832
+    },
+    {
+      "epoch": 1.402061855670103,
+      "grad_norm": 0.11173246055841446,
+      "learning_rate": 1e-05,
+      "loss": 2.3457,
+      "step": 833
+    },
+    {
+      "epoch": 1.4037450031559016,
+      "grad_norm": 0.10333243012428284,
+      "learning_rate": 1e-05,
+      "loss": 2.2659,
+      "step": 834
+    },
+    {
+      "epoch": 1.4054281506417,
+      "grad_norm": 0.13903972506523132,
+      "learning_rate": 1e-05,
+      "loss": 2.1946,
+      "step": 835
+    },
+    {
+      "epoch": 1.4071112981274985,
+      "grad_norm": 0.11832322925329208,
+      "learning_rate": 1e-05,
+      "loss": 2.3223,
+      "step": 836
+    },
+    {
+      "epoch": 1.4087944456132968,
+      "grad_norm": 0.10906493663787842,
+      "learning_rate": 1e-05,
+      "loss": 2.4316,
+      "step": 837
+    },
+    {
+      "epoch": 1.4104775930990954,
+      "grad_norm": 0.10980133712291718,
+      "learning_rate": 1e-05,
+      "loss": 2.3525,
+      "step": 838
+    },
+    {
+      "epoch": 1.4121607405848937,
+      "grad_norm": 0.12958386540412903,
+      "learning_rate": 1e-05,
+      "loss": 2.3081,
+      "step": 839
+    },
+    {
+      "epoch": 1.413843888070692,
+      "grad_norm": 0.1342059075832367,
+      "learning_rate": 1e-05,
+      "loss": 2.3564,
+      "step": 840
+    },
+    {
+      "epoch": 1.4155270355564906,
+      "grad_norm": 0.1362716406583786,
+      "learning_rate": 1e-05,
+      "loss": 2.2435,
+      "step": 841
+    },
+    {
+      "epoch": 1.4172101830422892,
+      "grad_norm": 0.10814797878265381,
+      "learning_rate": 1e-05,
+      "loss": 2.3373,
+      "step": 842
+    },
+    {
+      "epoch": 1.4188933305280875,
+      "grad_norm": 0.111182801425457,
+      "learning_rate": 1e-05,
+      "loss": 2.2921,
+      "step": 843
+    },
+    {
+      "epoch": 1.4205764780138859,
+      "grad_norm": 0.11161399632692337,
+      "learning_rate": 1e-05,
+      "loss": 2.3816,
+      "step": 844
+    },
+    {
+      "epoch": 1.4222596254996844,
+      "grad_norm": 0.1261526495218277,
+      "learning_rate": 1e-05,
+      "loss": 2.4082,
+      "step": 845
+    },
+    {
+      "epoch": 1.4239427729854828,
+      "grad_norm": 0.10805182158946991,
+      "learning_rate": 1e-05,
+      "loss": 2.3622,
+      "step": 846
+    },
+    {
+      "epoch": 1.4256259204712813,
+      "grad_norm": 0.12294517457485199,
+      "learning_rate": 1e-05,
+      "loss": 2.3638,
+      "step": 847
+    },
+    {
+      "epoch": 1.4273090679570797,
+      "grad_norm": 0.10903607308864594,
+      "learning_rate": 1e-05,
+      "loss": 2.3484,
+      "step": 848
+    },
+    {
+      "epoch": 1.4289922154428782,
+      "grad_norm": 0.12460491806268692,
+      "learning_rate": 1e-05,
+      "loss": 2.2046,
+      "step": 849
+    },
+    {
+      "epoch": 1.4306753629286766,
+      "grad_norm": 0.13793089985847473,
+      "learning_rate": 1e-05,
+      "loss": 2.2437,
+      "step": 850
+    },
+    {
+      "epoch": 1.4323585104144752,
+      "grad_norm": 0.11700379103422165,
+      "learning_rate": 1e-05,
+      "loss": 2.2288,
+      "step": 851
+    },
+    {
+      "epoch": 1.4340416579002735,
+      "grad_norm": 0.11343109607696533,
+      "learning_rate": 1e-05,
+      "loss": 2.2501,
+      "step": 852
+    },
+    {
+      "epoch": 1.435724805386072,
+      "grad_norm": 0.10918331891298294,
+      "learning_rate": 1e-05,
+      "loss": 2.47,
+      "step": 853
+    },
+    {
+      "epoch": 1.4374079528718704,
+      "grad_norm": 0.12782573699951172,
+      "learning_rate": 1e-05,
+      "loss": 2.2281,
+      "step": 854
+    },
+    {
+      "epoch": 1.4390911003576687,
+      "grad_norm": 0.12039442360401154,
+      "learning_rate": 1e-05,
+      "loss": 2.2766,
+      "step": 855
+    },
+    {
+      "epoch": 1.4407742478434673,
+      "grad_norm": 0.13949096202850342,
+      "learning_rate": 1e-05,
+      "loss": 2.198,
+      "step": 856
+    },
+    {
+      "epoch": 1.4424573953292656,
+      "grad_norm": 0.13327306509017944,
+      "learning_rate": 1e-05,
+      "loss": 2.2253,
+      "step": 857
+    },
+    {
+      "epoch": 1.4441405428150642,
+      "grad_norm": 0.1229238212108612,
+      "learning_rate": 1e-05,
+      "loss": 2.3147,
+      "step": 858
+    },
+    {
+      "epoch": 1.4458236903008626,
+      "grad_norm": 0.13407859206199646,
+      "learning_rate": 1e-05,
+      "loss": 2.2532,
+      "step": 859
+    },
+    {
+      "epoch": 1.4475068377866611,
+      "grad_norm": 0.1280384659767151,
+      "learning_rate": 1e-05,
+      "loss": 2.3174,
+      "step": 860
+    },
+    {
+      "epoch": 1.4491899852724595,
+      "grad_norm": 0.1532362997531891,
+      "learning_rate": 1e-05,
+      "loss": 2.1671,
+      "step": 861
+    },
+    {
+      "epoch": 1.450873132758258,
+      "grad_norm": 0.1134854182600975,
+      "learning_rate": 1e-05,
+      "loss": 2.3607,
+      "step": 862
+    },
+    {
+      "epoch": 1.4525562802440564,
+      "grad_norm": 0.11682198196649551,
+      "learning_rate": 1e-05,
+      "loss": 2.4041,
+      "step": 863
+    },
+    {
+      "epoch": 1.454239427729855,
+      "grad_norm": 0.11356412619352341,
+      "learning_rate": 1e-05,
+      "loss": 2.2756,
+      "step": 864
+    },
+    {
+      "epoch": 1.4559225752156533,
+      "grad_norm": 0.11278104037046432,
+      "learning_rate": 1e-05,
+      "loss": 2.2983,
+      "step": 865
+    },
+    {
+      "epoch": 1.4576057227014516,
+      "grad_norm": 0.13442599773406982,
+      "learning_rate": 1e-05,
+      "loss": 2.2593,
+      "step": 866
+    },
+    {
+      "epoch": 1.4592888701872502,
+      "grad_norm": 0.1254800707101822,
+      "learning_rate": 1e-05,
+      "loss": 2.3213,
+      "step": 867
+    },
+    {
+      "epoch": 1.4609720176730487,
+      "grad_norm": 0.12374315410852432,
+      "learning_rate": 1e-05,
+      "loss": 2.4221,
+      "step": 868
+    },
+    {
+      "epoch": 1.462655165158847,
+      "grad_norm": 0.13577024638652802,
+      "learning_rate": 1e-05,
+      "loss": 2.2473,
+      "step": 869
+    },
+    {
+      "epoch": 1.4643383126446454,
+      "grad_norm": 0.12822799384593964,
+      "learning_rate": 1e-05,
+      "loss": 2.3057,
+      "step": 870
+    },
+    {
+      "epoch": 1.466021460130444,
+      "grad_norm": 0.1283286213874817,
+      "learning_rate": 1e-05,
+      "loss": 2.374,
+      "step": 871
+    },
+    {
+      "epoch": 1.4677046076162423,
+      "grad_norm": 0.12054271996021271,
+      "learning_rate": 1e-05,
+      "loss": 2.3369,
+      "step": 872
+    },
+    {
+      "epoch": 1.469387755102041,
+      "grad_norm": 0.127189502120018,
+      "learning_rate": 1e-05,
+      "loss": 2.3167,
+      "step": 873
+    },
+    {
+      "epoch": 1.4710709025878392,
+      "grad_norm": 0.12767814099788666,
+      "learning_rate": 1e-05,
+      "loss": 2.2695,
+      "step": 874
+    },
+    {
+      "epoch": 1.4727540500736378,
+      "grad_norm": 0.12026406079530716,
+      "learning_rate": 1e-05,
+      "loss": 2.3313,
+      "step": 875
+    },
+    {
+      "epoch": 1.4744371975594361,
+      "grad_norm": 0.13317981362342834,
+      "learning_rate": 1e-05,
+      "loss": 2.209,
+      "step": 876
+    },
+    {
+      "epoch": 1.4761203450452345,
+      "grad_norm": 0.12904947996139526,
+      "learning_rate": 1e-05,
+      "loss": 2.2344,
+      "step": 877
+    },
+    {
+      "epoch": 1.477803492531033,
+      "grad_norm": 0.13126946985721588,
+      "learning_rate": 1e-05,
+      "loss": 2.2888,
+      "step": 878
+    },
+    {
+      "epoch": 1.4794866400168316,
+      "grad_norm": 0.128869891166687,
+      "learning_rate": 1e-05,
+      "loss": 2.1996,
+      "step": 879
+    },
+    {
+      "epoch": 1.48116978750263,
+      "grad_norm": 0.1279861181974411,
+      "learning_rate": 1e-05,
+      "loss": 2.1873,
+      "step": 880
+    },
+    {
+      "epoch": 1.4828529349884283,
+      "grad_norm": 0.11732237040996552,
+      "learning_rate": 1e-05,
+      "loss": 2.3259,
+      "step": 881
+    },
+    {
+      "epoch": 1.4845360824742269,
+      "grad_norm": 0.1279248595237732,
+      "learning_rate": 1e-05,
+      "loss": 2.386,
+      "step": 882
+    },
+    {
+      "epoch": 1.4862192299600252,
+      "grad_norm": 0.13578535616397858,
+      "learning_rate": 1e-05,
+      "loss": 2.2937,
+      "step": 883
+    },
+    {
+      "epoch": 1.4879023774458238,
+      "grad_norm": 0.13534606993198395,
+      "learning_rate": 1e-05,
+      "loss": 2.239,
+      "step": 884
+    },
+    {
+      "epoch": 1.489585524931622,
+      "grad_norm": 0.12359879165887833,
+      "learning_rate": 1e-05,
+      "loss": 2.3572,
+      "step": 885
+    },
+    {
+      "epoch": 1.4912686724174207,
+      "grad_norm": 0.1236250028014183,
+      "learning_rate": 1e-05,
+      "loss": 2.188,
+      "step": 886
+    },
+    {
+      "epoch": 1.492951819903219,
+      "grad_norm": 0.12695659697055817,
+      "learning_rate": 1e-05,
+      "loss": 2.2637,
+      "step": 887
+    },
+    {
+      "epoch": 1.4946349673890174,
+      "grad_norm": 0.1281343400478363,
+      "learning_rate": 1e-05,
+      "loss": 2.2961,
+      "step": 888
+    },
+    {
+      "epoch": 1.496318114874816,
+      "grad_norm": 0.12446150928735733,
+      "learning_rate": 1e-05,
+      "loss": 2.3362,
+      "step": 889
+    },
+    {
+      "epoch": 1.4980012623606145,
+      "grad_norm": 0.12564988434314728,
+      "learning_rate": 1e-05,
+      "loss": 2.288,
+      "step": 890
+    },
+    {
+      "epoch": 1.4996844098464128,
+      "grad_norm": 0.14049400389194489,
+      "learning_rate": 1e-05,
+      "loss": 2.2867,
+      "step": 891
+    },
+    {
+      "epoch": 1.5013675573322112,
+      "grad_norm": 0.12252961844205856,
+      "learning_rate": 1e-05,
+      "loss": 2.3511,
+      "step": 892
+    },
+    {
+      "epoch": 1.5030507048180097,
+      "grad_norm": 0.15993735194206238,
+      "learning_rate": 1e-05,
+      "loss": 2.0931,
+      "step": 893
+    },
+    {
+      "epoch": 1.504733852303808,
+      "grad_norm": 0.13673749566078186,
+      "learning_rate": 1e-05,
+      "loss": 2.2998,
+      "step": 894
+    },
+    {
+      "epoch": 1.5064169997896064,
+      "grad_norm": 0.11770147830247879,
+      "learning_rate": 1e-05,
+      "loss": 2.2883,
+      "step": 895
+    },
+    {
+      "epoch": 1.508100147275405,
+      "grad_norm": 0.11792504787445068,
+      "learning_rate": 1e-05,
+      "loss": 2.1893,
+      "step": 896
+    },
+    {
+      "epoch": 1.5097832947612035,
+      "grad_norm": 0.1405222862958908,
+      "learning_rate": 1e-05,
+      "loss": 2.2645,
+      "step": 897
+    },
+    {
+      "epoch": 1.5114664422470019,
+      "grad_norm": 0.1401311457157135,
+      "learning_rate": 1e-05,
+      "loss": 2.2085,
+      "step": 898
+    },
+    {
+      "epoch": 1.5131495897328002,
+      "grad_norm": 0.14068666100502014,
+      "learning_rate": 1e-05,
+      "loss": 2.2711,
+      "step": 899
+    },
+    {
+      "epoch": 1.5148327372185988,
+      "grad_norm": 0.12995976209640503,
+      "learning_rate": 1e-05,
+      "loss": 2.2883,
+      "step": 900
+    },
+    {
+      "epoch": 1.5165158847043974,
+      "grad_norm": 0.12454178184270859,
+      "learning_rate": 1e-05,
+      "loss": 2.2515,
+      "step": 901
+    },
+    {
+      "epoch": 1.5181990321901957,
+      "grad_norm": 0.12165191769599915,
+      "learning_rate": 1e-05,
+      "loss": 2.3621,
+      "step": 902
+    },
+    {
+      "epoch": 1.519882179675994,
+      "grad_norm": 0.1413601189851761,
+      "learning_rate": 1e-05,
+      "loss": 2.27,
+      "step": 903
+    },
+    {
+      "epoch": 1.5215653271617926,
+      "grad_norm": 0.13545894622802734,
+      "learning_rate": 1e-05,
+      "loss": 2.3008,
+      "step": 904
+    },
+    {
+      "epoch": 1.523248474647591,
+      "grad_norm": 0.12211872637271881,
+      "learning_rate": 1e-05,
+      "loss": 2.3921,
+      "step": 905
+    },
+    {
+      "epoch": 1.5249316221333893,
+      "grad_norm": 0.13053253293037415,
+      "learning_rate": 1e-05,
+      "loss": 2.2434,
+      "step": 906
+    },
+    {
+      "epoch": 1.5266147696191879,
+      "grad_norm": 0.12977124750614166,
+      "learning_rate": 1e-05,
+      "loss": 2.2366,
+      "step": 907
+    },
+    {
+      "epoch": 1.5282979171049864,
+      "grad_norm": 0.13451719284057617,
+      "learning_rate": 1e-05,
+      "loss": 2.3154,
+      "step": 908
+    },
+    {
+      "epoch": 1.5299810645907848,
+      "grad_norm": 0.11067184805870056,
+      "learning_rate": 1e-05,
+      "loss": 2.3296,
+      "step": 909
+    },
+    {
+      "epoch": 1.531664212076583,
+      "grad_norm": 0.12281223386526108,
+      "learning_rate": 1e-05,
+      "loss": 2.2479,
+      "step": 910
+    },
+    {
+      "epoch": 1.5333473595623817,
+      "grad_norm": 0.12240397185087204,
+      "learning_rate": 1e-05,
+      "loss": 2.3416,
+      "step": 911
+    },
+    {
+      "epoch": 1.5350305070481802,
+      "grad_norm": 0.14465166628360748,
+      "learning_rate": 1e-05,
+      "loss": 2.1801,
+      "step": 912
+    },
+    {
+      "epoch": 1.5367136545339786,
+      "grad_norm": 0.1263197958469391,
+      "learning_rate": 1e-05,
+      "loss": 2.2583,
+      "step": 913
+    },
+    {
+      "epoch": 1.538396802019777,
+      "grad_norm": 0.14653970301151276,
+      "learning_rate": 1e-05,
+      "loss": 2.2939,
+      "step": 914
+    },
+    {
+      "epoch": 1.5400799495055755,
+      "grad_norm": 0.1311267763376236,
+      "learning_rate": 1e-05,
+      "loss": 2.2517,
+      "step": 915
+    },
+    {
+      "epoch": 1.5417630969913738,
+      "grad_norm": 0.13173674046993256,
+      "learning_rate": 1e-05,
+      "loss": 2.309,
+      "step": 916
+    },
+    {
+      "epoch": 1.5434462444771722,
+      "grad_norm": 0.13140322268009186,
+      "learning_rate": 1e-05,
+      "loss": 2.1447,
+      "step": 917
+    },
+    {
+      "epoch": 1.5451293919629707,
+      "grad_norm": 0.12431302666664124,
+      "learning_rate": 1e-05,
+      "loss": 2.3315,
+      "step": 918
+    },
+    {
+      "epoch": 1.5468125394487693,
+      "grad_norm": 0.14358630776405334,
+      "learning_rate": 1e-05,
+      "loss": 2.2634,
+      "step": 919
+    },
+    {
+      "epoch": 1.5484956869345676,
+      "grad_norm": 0.1297353357076645,
+      "learning_rate": 1e-05,
+      "loss": 2.2489,
+      "step": 920
+    },
+    {
+      "epoch": 1.550178834420366,
+      "grad_norm": 0.12963449954986572,
+      "learning_rate": 1e-05,
+      "loss": 2.1533,
+      "step": 921
+    },
+    {
+      "epoch": 1.5518619819061645,
+      "grad_norm": 0.11558603495359421,
+      "learning_rate": 1e-05,
+      "loss": 2.2688,
+      "step": 922
+    },
+    {
+      "epoch": 1.553545129391963,
+      "grad_norm": 0.14222054183483124,
+      "learning_rate": 1e-05,
+      "loss": 2.2385,
+      "step": 923
+    },
+    {
+      "epoch": 1.5552282768777614,
+      "grad_norm": 0.1376868486404419,
+      "learning_rate": 1e-05,
+      "loss": 2.2051,
+      "step": 924
+    },
+    {
+      "epoch": 1.5569114243635598,
+      "grad_norm": 0.12993879616260529,
+      "learning_rate": 1e-05,
+      "loss": 2.3445,
+      "step": 925
+    },
+    {
+      "epoch": 1.5585945718493583,
+      "grad_norm": 0.14503213763237,
+      "learning_rate": 1e-05,
+      "loss": 2.215,
+      "step": 926
+    },
+    {
+      "epoch": 1.560277719335157,
+      "grad_norm": 0.1302722692489624,
+      "learning_rate": 1e-05,
+      "loss": 2.1945,
+      "step": 927
+    },
+    {
+      "epoch": 1.561960866820955,
+      "grad_norm": 0.13545845448970795,
+      "learning_rate": 1e-05,
+      "loss": 2.3059,
+      "step": 928
+    },
+    {
+      "epoch": 1.5636440143067536,
+      "grad_norm": 0.12279404699802399,
+      "learning_rate": 1e-05,
+      "loss": 2.3511,
+      "step": 929
+    },
+    {
+      "epoch": 1.5653271617925522,
+      "grad_norm": 0.13220550119876862,
+      "learning_rate": 1e-05,
+      "loss": 2.2837,
+      "step": 930
+    },
+    {
+      "epoch": 1.5670103092783505,
+      "grad_norm": 0.1407599151134491,
+      "learning_rate": 1e-05,
+      "loss": 2.2905,
+      "step": 931
+    },
+    {
+      "epoch": 1.5686934567641488,
+      "grad_norm": 0.12597431242465973,
+      "learning_rate": 1e-05,
+      "loss": 2.366,
+      "step": 932
+    },
+    {
+      "epoch": 1.5703766042499474,
+      "grad_norm": 0.12998835742473602,
+      "learning_rate": 1e-05,
+      "loss": 2.1067,
+      "step": 933
+    },
+    {
+      "epoch": 1.572059751735746,
+      "grad_norm": 0.14708921313285828,
+      "learning_rate": 1e-05,
+      "loss": 2.2687,
+      "step": 934
+    },
+    {
+      "epoch": 1.5737428992215443,
+      "grad_norm": 0.13333402574062347,
+      "learning_rate": 1e-05,
+      "loss": 2.3381,
+      "step": 935
+    },
+    {
+      "epoch": 1.5754260467073427,
+      "grad_norm": 0.14774633944034576,
+      "learning_rate": 1e-05,
+      "loss": 2.163,
+      "step": 936
+    },
+    {
+      "epoch": 1.5771091941931412,
+      "grad_norm": 0.1283462792634964,
+      "learning_rate": 1e-05,
+      "loss": 2.3892,
+      "step": 937
+    },
+    {
+      "epoch": 1.5787923416789398,
+      "grad_norm": 0.12011823058128357,
+      "learning_rate": 1e-05,
+      "loss": 2.2758,
+      "step": 938
+    },
+    {
+      "epoch": 1.580475489164738,
+      "grad_norm": 0.11618427187204361,
+      "learning_rate": 1e-05,
+      "loss": 2.2545,
+      "step": 939
+    },
+    {
+      "epoch": 1.5821586366505365,
+      "grad_norm": 0.12683863937854767,
+      "learning_rate": 1e-05,
+      "loss": 2.291,
+      "step": 940
+    },
+    {
+      "epoch": 1.583841784136335,
+      "grad_norm": 0.13158243894577026,
+      "learning_rate": 1e-05,
+      "loss": 2.3066,
+      "step": 941
+    },
+    {
+      "epoch": 1.5855249316221334,
+      "grad_norm": 0.13269281387329102,
+      "learning_rate": 1e-05,
+      "loss": 2.3442,
+      "step": 942
+    },
+    {
+      "epoch": 1.5872080791079317,
+      "grad_norm": 0.14047692716121674,
+      "learning_rate": 1e-05,
+      "loss": 2.3092,
+      "step": 943
+    },
+    {
+      "epoch": 1.5888912265937303,
+      "grad_norm": 0.1387140154838562,
+      "learning_rate": 1e-05,
+      "loss": 2.1482,
+      "step": 944
+    },
+    {
+      "epoch": 1.5905743740795288,
+      "grad_norm": 0.13907848298549652,
+      "learning_rate": 1e-05,
+      "loss": 2.3484,
+      "step": 945
+    },
+    {
+      "epoch": 1.5922575215653272,
+      "grad_norm": 0.13114407658576965,
+      "learning_rate": 1e-05,
+      "loss": 2.2195,
+      "step": 946
+    },
+    {
+      "epoch": 1.5939406690511255,
+      "grad_norm": 0.1368924379348755,
+      "learning_rate": 1e-05,
+      "loss": 2.322,
+      "step": 947
+    },
+    {
+      "epoch": 1.595623816536924,
+      "grad_norm": 0.141913041472435,
+      "learning_rate": 1e-05,
+      "loss": 2.2336,
+      "step": 948
+    },
+    {
+      "epoch": 1.5973069640227227,
+      "grad_norm": 0.13295848667621613,
+      "learning_rate": 1e-05,
+      "loss": 2.3081,
+      "step": 949
+    },
+    {
+      "epoch": 1.5989901115085208,
+      "grad_norm": 0.12306110560894012,
+      "learning_rate": 1e-05,
+      "loss": 2.3354,
+      "step": 950
+    },
+    {
+      "epoch": 1.6006732589943193,
+      "grad_norm": 0.12122649699449539,
+      "learning_rate": 1e-05,
+      "loss": 2.2839,
+      "step": 951
+    },
+    {
+      "epoch": 1.602356406480118,
+      "grad_norm": 0.13046576082706451,
+      "learning_rate": 1e-05,
+      "loss": 2.385,
+      "step": 952
+    },
+    {
+      "epoch": 1.6040395539659162,
+      "grad_norm": 0.1272476315498352,
+      "learning_rate": 1e-05,
+      "loss": 2.4153,
+      "step": 953
+    },
+    {
+      "epoch": 1.6057227014517146,
+      "grad_norm": 0.13073799014091492,
+      "learning_rate": 1e-05,
+      "loss": 2.2854,
+      "step": 954
+    },
+    {
+      "epoch": 1.6074058489375131,
+      "grad_norm": 0.12583526968955994,
+      "learning_rate": 1e-05,
+      "loss": 2.3318,
+      "step": 955
+    },
+    {
+      "epoch": 1.6090889964233117,
+      "grad_norm": 0.1474972665309906,
+      "learning_rate": 1e-05,
+      "loss": 2.2542,
+      "step": 956
+    },
+    {
+      "epoch": 1.61077214390911,
+      "grad_norm": 0.13445797562599182,
+      "learning_rate": 1e-05,
+      "loss": 2.3645,
+      "step": 957
+    },
+    {
+      "epoch": 1.6124552913949084,
+      "grad_norm": 0.13466110825538635,
+      "learning_rate": 1e-05,
+      "loss": 2.3394,
+      "step": 958
+    },
+    {
+      "epoch": 1.614138438880707,
+      "grad_norm": 0.13525816798210144,
+      "learning_rate": 1e-05,
+      "loss": 2.2471,
+      "step": 959
+    },
+    {
+      "epoch": 1.6158215863665055,
+      "grad_norm": 0.1377459019422531,
+      "learning_rate": 1e-05,
+      "loss": 2.2478,
+      "step": 960
+    },
+    {
+      "epoch": 1.6175047338523036,
+      "grad_norm": 0.1405583918094635,
+      "learning_rate": 1e-05,
+      "loss": 2.2146,
+      "step": 961
+    },
+    {
+      "epoch": 1.6191878813381022,
+      "grad_norm": 0.11743167042732239,
+      "learning_rate": 1e-05,
+      "loss": 2.3555,
+      "step": 962
+    },
+    {
+      "epoch": 1.6208710288239008,
+      "grad_norm": 0.13644517958164215,
+      "learning_rate": 1e-05,
+      "loss": 2.2155,
+      "step": 963
+    },
+    {
+      "epoch": 1.6225541763096991,
+      "grad_norm": 0.12609997391700745,
+      "learning_rate": 1e-05,
+      "loss": 2.2593,
+      "step": 964
+    },
+    {
+      "epoch": 1.6242373237954975,
+      "grad_norm": 0.13276560604572296,
+      "learning_rate": 1e-05,
+      "loss": 2.1737,
+      "step": 965
+    },
+    {
+      "epoch": 1.625920471281296,
+      "grad_norm": 0.13567714393138885,
+      "learning_rate": 1e-05,
+      "loss": 2.3336,
+      "step": 966
+    },
+    {
+      "epoch": 1.6276036187670946,
+      "grad_norm": 0.12559200823307037,
+      "learning_rate": 1e-05,
+      "loss": 2.3494,
+      "step": 967
+    },
+    {
+      "epoch": 1.629286766252893,
+      "grad_norm": 0.13090649247169495,
+      "learning_rate": 1e-05,
+      "loss": 2.1851,
+      "step": 968
+    },
+    {
+      "epoch": 1.6309699137386913,
+      "grad_norm": 0.15777987241744995,
+      "learning_rate": 1e-05,
+      "loss": 2.2205,
+      "step": 969
+    },
+    {
+      "epoch": 1.6326530612244898,
+      "grad_norm": 0.1433715522289276,
+      "learning_rate": 1e-05,
+      "loss": 2.2295,
+      "step": 970
+    },
+    {
+      "epoch": 1.6343362087102884,
+      "grad_norm": 0.1218508929014206,
+      "learning_rate": 1e-05,
+      "loss": 2.3762,
+      "step": 971
+    },
+    {
+      "epoch": 1.6360193561960865,
+      "grad_norm": 0.14540942013263702,
+      "learning_rate": 1e-05,
+      "loss": 2.2139,
+      "step": 972
+    },
+    {
+      "epoch": 1.637702503681885,
+      "grad_norm": 0.14829136431217194,
+      "learning_rate": 1e-05,
+      "loss": 2.2871,
+      "step": 973
+    },
+    {
+      "epoch": 1.6393856511676836,
+      "grad_norm": 0.12728969752788544,
+      "learning_rate": 1e-05,
+      "loss": 2.2917,
+      "step": 974
+    },
+    {
+      "epoch": 1.641068798653482,
+      "grad_norm": 0.1471221148967743,
+      "learning_rate": 1e-05,
+      "loss": 2.2012,
+      "step": 975
+    },
+    {
+      "epoch": 1.6427519461392803,
+      "grad_norm": 0.13320200145244598,
+      "learning_rate": 1e-05,
+      "loss": 2.2771,
+      "step": 976
+    },
+    {
+      "epoch": 1.644435093625079,
+      "grad_norm": 0.1363966464996338,
+      "learning_rate": 1e-05,
+      "loss": 2.3086,
+      "step": 977
+    },
+    {
+      "epoch": 1.6461182411108775,
+      "grad_norm": 0.13870568573474884,
+      "learning_rate": 1e-05,
+      "loss": 2.2898,
+      "step": 978
+    },
+    {
+      "epoch": 1.6478013885966758,
+      "grad_norm": 0.15152350068092346,
+      "learning_rate": 1e-05,
+      "loss": 2.2994,
+      "step": 979
+    },
+    {
+      "epoch": 1.6494845360824741,
+      "grad_norm": 0.13830937445163727,
+      "learning_rate": 1e-05,
+      "loss": 2.2108,
+      "step": 980
+    },
+    {
+      "epoch": 1.6511676835682727,
+      "grad_norm": 0.15544220805168152,
+      "learning_rate": 1e-05,
+      "loss": 2.4043,
+      "step": 981
+    },
+    {
+      "epoch": 1.6528508310540713,
+      "grad_norm": 0.13135483860969543,
+      "learning_rate": 1e-05,
+      "loss": 2.2373,
+      "step": 982
+    },
+    {
+      "epoch": 1.6545339785398696,
+      "grad_norm": 0.12355194985866547,
+      "learning_rate": 1e-05,
+      "loss": 2.4163,
+      "step": 983
+    },
+    {
+      "epoch": 1.656217126025668,
+      "grad_norm": 0.14110660552978516,
+      "learning_rate": 1e-05,
+      "loss": 2.2031,
+      "step": 984
+    },
+    {
+      "epoch": 1.6579002735114665,
+      "grad_norm": 0.13077346980571747,
+      "learning_rate": 1e-05,
+      "loss": 2.3601,
+      "step": 985
+    },
+    {
+      "epoch": 1.6595834209972649,
+      "grad_norm": 0.14212660491466522,
+      "learning_rate": 1e-05,
+      "loss": 2.197,
+      "step": 986
+    },
+    {
+      "epoch": 1.6612665684830632,
+      "grad_norm": 0.12336140871047974,
+      "learning_rate": 1e-05,
+      "loss": 2.4146,
+      "step": 987
+    },
+    {
+      "epoch": 1.6629497159688618,
+      "grad_norm": 0.15291054546833038,
+      "learning_rate": 1e-05,
+      "loss": 2.2764,
+      "step": 988
+    },
+    {
+      "epoch": 1.6646328634546603,
+      "grad_norm": 0.1272605061531067,
+      "learning_rate": 1e-05,
+      "loss": 2.2703,
+      "step": 989
+    },
+    {
+      "epoch": 1.6663160109404587,
+      "grad_norm": 0.13462689518928528,
+      "learning_rate": 1e-05,
+      "loss": 2.3188,
+      "step": 990
+    },
+    {
+      "epoch": 1.667999158426257,
+      "grad_norm": 0.13290910422801971,
+      "learning_rate": 1e-05,
+      "loss": 2.2172,
+      "step": 991
+    },
+    {
+      "epoch": 1.6696823059120556,
+      "grad_norm": 0.15105758607387543,
+      "learning_rate": 1e-05,
+      "loss": 2.2156,
+      "step": 992
+    },
+    {
+      "epoch": 1.6713654533978541,
+      "grad_norm": 0.13150456547737122,
+      "learning_rate": 1e-05,
+      "loss": 2.3362,
+      "step": 993
+    },
+    {
+      "epoch": 1.6730486008836525,
+      "grad_norm": 0.13139204680919647,
+      "learning_rate": 1e-05,
+      "loss": 2.3833,
+      "step": 994
+    },
+    {
+      "epoch": 1.6747317483694508,
+      "grad_norm": 0.14886420965194702,
+      "learning_rate": 1e-05,
+      "loss": 2.1893,
+      "step": 995
+    },
+    {
+      "epoch": 1.6764148958552494,
+      "grad_norm": 0.13227102160453796,
+      "learning_rate": 1e-05,
+      "loss": 2.4055,
+      "step": 996
+    },
+    {
+      "epoch": 1.6780980433410477,
+      "grad_norm": 0.12545333802700043,
+      "learning_rate": 1e-05,
+      "loss": 2.3311,
+      "step": 997
+    },
+    {
+      "epoch": 1.679781190826846,
+      "grad_norm": 0.13391169905662537,
+      "learning_rate": 1e-05,
+      "loss": 2.3022,
+      "step": 998
+    },
+    {
+      "epoch": 1.6814643383126446,
+      "grad_norm": 0.13013269007205963,
+      "learning_rate": 1e-05,
+      "loss": 2.2318,
+      "step": 999
+    },
+    {
+      "epoch": 1.6831474857984432,
+      "grad_norm": 0.1331031173467636,
+      "learning_rate": 1e-05,
+      "loss": 2.3022,
+      "step": 1000
+    },
+    {
+      "epoch": 1.6848306332842415,
+      "grad_norm": 0.14438873529434204,
+      "learning_rate": 1e-05,
+      "loss": 2.2388,
+      "step": 1001
+    },
+    {
+      "epoch": 1.6865137807700399,
+      "grad_norm": 0.1422380954027176,
+      "learning_rate": 1e-05,
+      "loss": 2.3145,
+      "step": 1002
+    },
+    {
+      "epoch": 1.6881969282558384,
+      "grad_norm": 0.13909044861793518,
+      "learning_rate": 1e-05,
+      "loss": 2.2249,
+      "step": 1003
+    },
+    {
+      "epoch": 1.689880075741637,
+      "grad_norm": 0.14147858321666718,
+      "learning_rate": 1e-05,
+      "loss": 2.3179,
+      "step": 1004
+    },
+    {
+      "epoch": 1.6915632232274354,
+      "grad_norm": 0.13203288614749908,
+      "learning_rate": 1e-05,
+      "loss": 2.1912,
+      "step": 1005
+    },
+    {
+      "epoch": 1.6932463707132337,
+      "grad_norm": 0.14461839199066162,
+      "learning_rate": 1e-05,
+      "loss": 2.1982,
+      "step": 1006
+    },
+    {
+      "epoch": 1.6949295181990323,
+      "grad_norm": 0.14539021253585815,
+      "learning_rate": 1e-05,
+      "loss": 2.2917,
+      "step": 1007
+    },
+    {
+      "epoch": 1.6966126656848306,
+      "grad_norm": 0.14774973690509796,
+      "learning_rate": 1e-05,
+      "loss": 2.2639,
+      "step": 1008
+    },
+    {
+      "epoch": 1.698295813170629,
+      "grad_norm": 0.14927157759666443,
+      "learning_rate": 1e-05,
+      "loss": 2.1956,
+      "step": 1009
+    },
+    {
+      "epoch": 1.6999789606564275,
+      "grad_norm": 0.1286613643169403,
+      "learning_rate": 1e-05,
+      "loss": 2.292,
+      "step": 1010
+    },
+    {
+      "epoch": 1.701662108142226,
+      "grad_norm": 0.12883049249649048,
+      "learning_rate": 1e-05,
+      "loss": 2.2573,
+      "step": 1011
+    },
+    {
+      "epoch": 1.7033452556280244,
+      "grad_norm": 0.14129754900932312,
+      "learning_rate": 1e-05,
+      "loss": 2.334,
+      "step": 1012
+    },
+    {
+      "epoch": 1.7050284031138228,
+      "grad_norm": 0.13216479122638702,
+      "learning_rate": 1e-05,
+      "loss": 2.2664,
+      "step": 1013
+    },
+    {
+      "epoch": 1.7067115505996213,
+      "grad_norm": 0.12611788511276245,
+      "learning_rate": 1e-05,
+      "loss": 2.3159,
+      "step": 1014
+    },
+    {
+      "epoch": 1.7083946980854199,
+      "grad_norm": 0.14012207090854645,
+      "learning_rate": 1e-05,
+      "loss": 2.4026,
+      "step": 1015
+    },
+    {
+      "epoch": 1.7100778455712182,
+      "grad_norm": 0.14449255168437958,
+      "learning_rate": 1e-05,
+      "loss": 2.3313,
+      "step": 1016
+    },
+    {
+      "epoch": 1.7117609930570166,
+      "grad_norm": 0.15093393623828888,
+      "learning_rate": 1e-05,
+      "loss": 2.2075,
+      "step": 1017
+    },
+    {
+      "epoch": 1.7134441405428151,
+      "grad_norm": 0.15169350802898407,
+      "learning_rate": 1e-05,
+      "loss": 2.1926,
+      "step": 1018
+    },
+    {
+      "epoch": 1.7151272880286135,
+      "grad_norm": 0.13613849878311157,
+      "learning_rate": 1e-05,
+      "loss": 2.3394,
+      "step": 1019
+    },
+    {
+      "epoch": 1.7168104355144118,
+      "grad_norm": 0.13525283336639404,
+      "learning_rate": 1e-05,
+      "loss": 2.2234,
+      "step": 1020
+    },
+    {
+      "epoch": 1.7184935830002104,
+      "grad_norm": 0.1529736965894699,
+      "learning_rate": 1e-05,
+      "loss": 2.1866,
+      "step": 1021
+    },
+    {
+      "epoch": 1.720176730486009,
+      "grad_norm": 0.13723863661289215,
+      "learning_rate": 1e-05,
+      "loss": 2.3027,
+      "step": 1022
+    },
+    {
+      "epoch": 1.7218598779718073,
+      "grad_norm": 0.16251115500926971,
+      "learning_rate": 1e-05,
+      "loss": 2.3428,
+      "step": 1023
+    },
+    {
+      "epoch": 1.7235430254576056,
+      "grad_norm": 0.1440790742635727,
+      "learning_rate": 1e-05,
+      "loss": 2.3298,
+      "step": 1024
+    },
+    {
+      "epoch": 1.7252261729434042,
+      "grad_norm": 0.13486018776893616,
+      "learning_rate": 1e-05,
+      "loss": 2.3826,
+      "step": 1025
+    },
+    {
+      "epoch": 1.7269093204292028,
+      "grad_norm": 0.15616028010845184,
+      "learning_rate": 1e-05,
+      "loss": 2.0817,
+      "step": 1026
+    },
+    {
+      "epoch": 1.728592467915001,
+      "grad_norm": 0.15306299924850464,
+      "learning_rate": 1e-05,
+      "loss": 2.2601,
+      "step": 1027
+    },
+    {
+      "epoch": 1.7302756154007994,
+      "grad_norm": 0.14421014487743378,
+      "learning_rate": 1e-05,
+      "loss": 2.1998,
+      "step": 1028
+    },
+    {
+      "epoch": 1.731958762886598,
+      "grad_norm": 0.14438478648662567,
+      "learning_rate": 1e-05,
+      "loss": 2.262,
+      "step": 1029
+    },
+    {
+      "epoch": 1.7336419103723963,
+      "grad_norm": 0.13325351476669312,
+      "learning_rate": 1e-05,
+      "loss": 2.2852,
+      "step": 1030
+    },
+    {
+      "epoch": 1.7353250578581947,
+      "grad_norm": 0.14232920110225677,
+      "learning_rate": 1e-05,
+      "loss": 2.3147,
+      "step": 1031
+    },
+    {
+      "epoch": 1.7370082053439933,
+      "grad_norm": 0.1394515186548233,
+      "learning_rate": 1e-05,
+      "loss": 2.2781,
+      "step": 1032
+    },
+    {
+      "epoch": 1.7386913528297918,
+      "grad_norm": 0.12838682532310486,
+      "learning_rate": 1e-05,
+      "loss": 2.2827,
+      "step": 1033
+    },
+    {
+      "epoch": 1.7403745003155902,
+      "grad_norm": 0.15612417459487915,
+      "learning_rate": 1e-05,
+      "loss": 2.3108,
+      "step": 1034
+    },
+    {
+      "epoch": 1.7420576478013885,
+      "grad_norm": 0.14740139245986938,
+      "learning_rate": 1e-05,
+      "loss": 2.2412,
+      "step": 1035
+    },
+    {
+      "epoch": 1.743740795287187,
+      "grad_norm": 0.1541980355978012,
+      "learning_rate": 1e-05,
+      "loss": 2.3156,
+      "step": 1036
+    },
+    {
+      "epoch": 1.7454239427729856,
+      "grad_norm": 0.14056488871574402,
+      "learning_rate": 1e-05,
+      "loss": 2.1829,
+      "step": 1037
+    },
+    {
+      "epoch": 1.747107090258784,
+      "grad_norm": 0.143393874168396,
+      "learning_rate": 1e-05,
+      "loss": 2.2717,
+      "step": 1038
+    },
+    {
+      "epoch": 1.7487902377445823,
+      "grad_norm": 0.14296631515026093,
+      "learning_rate": 1e-05,
+      "loss": 2.342,
+      "step": 1039
+    },
+    {
+      "epoch": 1.7504733852303809,
+      "grad_norm": 0.13753627240657806,
+      "learning_rate": 1e-05,
+      "loss": 2.324,
+      "step": 1040
+    },
+    {
+      "epoch": 1.7521565327161792,
+      "grad_norm": 0.13361461460590363,
+      "learning_rate": 1e-05,
+      "loss": 2.3549,
+      "step": 1041
+    },
+    {
+      "epoch": 1.7538396802019776,
+      "grad_norm": 0.16176526248455048,
+      "learning_rate": 1e-05,
+      "loss": 2.0996,
+      "step": 1042
+    },
+    {
+      "epoch": 1.7555228276877761,
+      "grad_norm": 0.14512574672698975,
+      "learning_rate": 1e-05,
+      "loss": 2.3289,
+      "step": 1043
+    },
+    {
+      "epoch": 1.7572059751735747,
+      "grad_norm": 0.14329467713832855,
+      "learning_rate": 1e-05,
+      "loss": 2.2429,
+      "step": 1044
+    },
+    {
+      "epoch": 1.758889122659373,
+      "grad_norm": 0.1415308713912964,
+      "learning_rate": 1e-05,
+      "loss": 2.2976,
+      "step": 1045
+    },
+    {
+      "epoch": 1.7605722701451714,
+      "grad_norm": 0.13017630577087402,
+      "learning_rate": 1e-05,
+      "loss": 2.3142,
+      "step": 1046
+    },
+    {
+      "epoch": 1.76225541763097,
+      "grad_norm": 0.14865103363990784,
+      "learning_rate": 1e-05,
+      "loss": 2.2659,
+      "step": 1047
+    },
+    {
+      "epoch": 1.7639385651167685,
+      "grad_norm": 0.13973674178123474,
+      "learning_rate": 1e-05,
+      "loss": 2.1975,
+      "step": 1048
+    },
+    {
+      "epoch": 1.7656217126025668,
+      "grad_norm": 0.12378077954053879,
+      "learning_rate": 1e-05,
+      "loss": 2.4469,
+      "step": 1049
+    },
+    {
+      "epoch": 1.7673048600883652,
+      "grad_norm": 0.13462629914283752,
+      "learning_rate": 1e-05,
+      "loss": 2.332,
+      "step": 1050
+    },
+    {
+      "epoch": 1.7689880075741637,
+      "grad_norm": 0.14375431835651398,
+      "learning_rate": 1e-05,
+      "loss": 2.2834,
+      "step": 1051
+    },
+    {
+      "epoch": 1.770671155059962,
+      "grad_norm": 0.1413864940404892,
+      "learning_rate": 1e-05,
+      "loss": 2.2769,
+      "step": 1052
+    },
+    {
+      "epoch": 1.7723543025457604,
+      "grad_norm": 0.15052342414855957,
+      "learning_rate": 1e-05,
+      "loss": 2.2522,
+      "step": 1053
+    },
+    {
+      "epoch": 1.774037450031559,
+      "grad_norm": 0.15616975724697113,
+      "learning_rate": 1e-05,
+      "loss": 2.1501,
+      "step": 1054
+    },
+    {
+      "epoch": 1.7757205975173576,
+      "grad_norm": 0.16257071495056152,
+      "learning_rate": 1e-05,
+      "loss": 2.1545,
+      "step": 1055
+    },
+    {
+      "epoch": 1.777403745003156,
+      "grad_norm": 0.13512100279331207,
+      "learning_rate": 1e-05,
+      "loss": 2.2218,
+      "step": 1056
+    },
+    {
+      "epoch": 1.7790868924889542,
+      "grad_norm": 0.1581428200006485,
+      "learning_rate": 1e-05,
+      "loss": 2.1865,
+      "step": 1057
+    },
+    {
+      "epoch": 1.7807700399747528,
+      "grad_norm": 0.13829343020915985,
+      "learning_rate": 1e-05,
+      "loss": 2.3337,
+      "step": 1058
+    },
+    {
+      "epoch": 1.7824531874605514,
+      "grad_norm": 0.16639141738414764,
+      "learning_rate": 1e-05,
+      "loss": 2.2325,
+      "step": 1059
+    },
+    {
+      "epoch": 1.7841363349463497,
+      "grad_norm": 0.1412006914615631,
+      "learning_rate": 1e-05,
+      "loss": 2.3389,
+      "step": 1060
+    },
+    {
+      "epoch": 1.785819482432148,
+      "grad_norm": 0.13130658864974976,
+      "learning_rate": 1e-05,
+      "loss": 2.3376,
+      "step": 1061
+    },
+    {
+      "epoch": 1.7875026299179466,
+      "grad_norm": 0.1495353728532791,
+      "learning_rate": 1e-05,
+      "loss": 2.2666,
+      "step": 1062
+    },
+    {
+      "epoch": 1.789185777403745,
+      "grad_norm": 0.15077506005764008,
+      "learning_rate": 1e-05,
+      "loss": 2.228,
+      "step": 1063
+    },
+    {
+      "epoch": 1.7908689248895433,
+      "grad_norm": 0.1426386535167694,
+      "learning_rate": 1e-05,
+      "loss": 2.2727,
+      "step": 1064
+    },
+    {
+      "epoch": 1.7925520723753419,
+      "grad_norm": 0.14268244802951813,
+      "learning_rate": 1e-05,
+      "loss": 2.3643,
+      "step": 1065
+    },
+    {
+      "epoch": 1.7942352198611404,
+      "grad_norm": 0.14923584461212158,
+      "learning_rate": 1e-05,
+      "loss": 2.333,
+      "step": 1066
+    },
+    {
+      "epoch": 1.7959183673469388,
+      "grad_norm": 0.15571311116218567,
+      "learning_rate": 1e-05,
+      "loss": 2.3171,
+      "step": 1067
+    },
+    {
+      "epoch": 1.7976015148327371,
+      "grad_norm": 0.13931907713413239,
+      "learning_rate": 1e-05,
+      "loss": 2.2164,
+      "step": 1068
+    },
+    {
+      "epoch": 1.7992846623185357,
+      "grad_norm": 0.1513443887233734,
+      "learning_rate": 1e-05,
+      "loss": 2.2885,
+      "step": 1069
+    },
+    {
+      "epoch": 1.8009678098043342,
+      "grad_norm": 0.14123128354549408,
+      "learning_rate": 1e-05,
+      "loss": 2.3517,
+      "step": 1070
+    },
+    {
+      "epoch": 1.8026509572901326,
+      "grad_norm": 0.16668306291103363,
+      "learning_rate": 1e-05,
+      "loss": 2.1907,
+      "step": 1071
+    },
+    {
+      "epoch": 1.804334104775931,
+      "grad_norm": 0.14049063622951508,
+      "learning_rate": 1e-05,
+      "loss": 2.4216,
+      "step": 1072
+    },
+    {
+      "epoch": 1.8060172522617295,
+      "grad_norm": 0.13806495070457458,
+      "learning_rate": 1e-05,
+      "loss": 2.3367,
+      "step": 1073
+    },
+    {
+      "epoch": 1.8077003997475278,
+      "grad_norm": 0.14562048017978668,
+      "learning_rate": 1e-05,
+      "loss": 2.2303,
+      "step": 1074
+    },
+    {
+      "epoch": 1.8093835472333262,
+      "grad_norm": 0.16803675889968872,
+      "learning_rate": 1e-05,
+      "loss": 2.2404,
+      "step": 1075
+    },
+    {
+      "epoch": 1.8110666947191247,
+      "grad_norm": 0.14971864223480225,
+      "learning_rate": 1e-05,
+      "loss": 2.1941,
+      "step": 1076
+    },
+    {
+      "epoch": 1.8127498422049233,
+      "grad_norm": 0.162116140127182,
+      "learning_rate": 1e-05,
+      "loss": 2.2034,
+      "step": 1077
+    },
+    {
+      "epoch": 1.8144329896907216,
+      "grad_norm": 0.1417408138513565,
+      "learning_rate": 1e-05,
+      "loss": 2.2991,
+      "step": 1078
+    },
+    {
+      "epoch": 1.81611613717652,
+      "grad_norm": 0.14334024488925934,
+      "learning_rate": 1e-05,
+      "loss": 2.3796,
+      "step": 1079
+    },
+    {
+      "epoch": 1.8177992846623185,
+      "grad_norm": 0.13600003719329834,
+      "learning_rate": 1e-05,
+      "loss": 2.2322,
+      "step": 1080
+    },
+    {
+      "epoch": 1.8194824321481171,
+      "grad_norm": 0.1557435244321823,
+      "learning_rate": 1e-05,
+      "loss": 2.2151,
+      "step": 1081
+    },
+    {
+      "epoch": 1.8211655796339155,
+      "grad_norm": 0.14444471895694733,
+      "learning_rate": 1e-05,
+      "loss": 2.2778,
+      "step": 1082
+    },
+    {
+      "epoch": 1.8228487271197138,
+      "grad_norm": 0.15237338840961456,
+      "learning_rate": 1e-05,
+      "loss": 2.1863,
+      "step": 1083
+    },
+    {
+      "epoch": 1.8245318746055124,
+      "grad_norm": 0.1488647758960724,
+      "learning_rate": 1e-05,
+      "loss": 2.1194,
+      "step": 1084
+    },
+    {
+      "epoch": 1.8262150220913107,
+      "grad_norm": 0.14532509446144104,
+      "learning_rate": 1e-05,
+      "loss": 2.3018,
+      "step": 1085
+    },
+    {
+      "epoch": 1.827898169577109,
+      "grad_norm": 0.1438300609588623,
+      "learning_rate": 1e-05,
+      "loss": 2.3542,
+      "step": 1086
+    },
+    {
+      "epoch": 1.8295813170629076,
+      "grad_norm": 0.13162897527217865,
+      "learning_rate": 1e-05,
+      "loss": 2.3762,
+      "step": 1087
+    },
+    {
+      "epoch": 1.8312644645487062,
+      "grad_norm": 0.14388734102249146,
+      "learning_rate": 1e-05,
+      "loss": 2.3097,
+      "step": 1088
+    },
+    {
+      "epoch": 1.8329476120345045,
+      "grad_norm": 0.1633898764848709,
+      "learning_rate": 1e-05,
+      "loss": 2.1975,
+      "step": 1089
+    },
+    {
+      "epoch": 1.8346307595203029,
+      "grad_norm": 0.14513400197029114,
+      "learning_rate": 1e-05,
+      "loss": 2.3562,
+      "step": 1090
+    },
+    {
+      "epoch": 1.8363139070061014,
+      "grad_norm": 0.1562061607837677,
+      "learning_rate": 1e-05,
+      "loss": 2.2384,
+      "step": 1091
+    },
+    {
+      "epoch": 1.8379970544919,
+      "grad_norm": 0.14833082258701324,
+      "learning_rate": 1e-05,
+      "loss": 2.199,
+      "step": 1092
+    },
+    {
+      "epoch": 1.8396802019776983,
+      "grad_norm": 0.14182843267917633,
+      "learning_rate": 1e-05,
+      "loss": 2.2632,
+      "step": 1093
+    },
+    {
+      "epoch": 1.8413633494634967,
+      "grad_norm": 0.16517210006713867,
+      "learning_rate": 1e-05,
+      "loss": 2.2719,
+      "step": 1094
+    },
+    {
+      "epoch": 1.8430464969492952,
+      "grad_norm": 0.1563366949558258,
+      "learning_rate": 1e-05,
+      "loss": 2.2285,
+      "step": 1095
+    },
+    {
+      "epoch": 1.8447296444350936,
+      "grad_norm": 0.1349581480026245,
+      "learning_rate": 1e-05,
+      "loss": 2.2998,
+      "step": 1096
+    },
+    {
+      "epoch": 1.846412791920892,
+      "grad_norm": 0.14647842943668365,
+      "learning_rate": 1e-05,
+      "loss": 2.2588,
+      "step": 1097
+    },
+    {
+      "epoch": 1.8480959394066905,
+      "grad_norm": 0.1527308076620102,
+      "learning_rate": 1e-05,
+      "loss": 2.1945,
+      "step": 1098
+    },
+    {
+      "epoch": 1.849779086892489,
+      "grad_norm": 0.16208425164222717,
+      "learning_rate": 1e-05,
+      "loss": 2.1692,
+      "step": 1099
+    },
+    {
+      "epoch": 1.8514622343782874,
+      "grad_norm": 0.15897248685359955,
+      "learning_rate": 1e-05,
+      "loss": 2.3582,
+      "step": 1100
+    },
+    {
+      "epoch": 1.8531453818640857,
+      "grad_norm": 0.14687612652778625,
+      "learning_rate": 1e-05,
+      "loss": 2.3057,
+      "step": 1101
+    },
+    {
+      "epoch": 1.8548285293498843,
+      "grad_norm": 0.1631488800048828,
+      "learning_rate": 1e-05,
+      "loss": 2.2521,
+      "step": 1102
+    },
+    {
+      "epoch": 1.8565116768356829,
+      "grad_norm": 0.14686156809329987,
+      "learning_rate": 1e-05,
+      "loss": 2.313,
+      "step": 1103
+    },
+    {
+      "epoch": 1.8581948243214812,
+      "grad_norm": 0.162966787815094,
+      "learning_rate": 1e-05,
+      "loss": 2.1968,
+      "step": 1104
+    },
+    {
+      "epoch": 1.8598779718072795,
+      "grad_norm": 0.15387648344039917,
+      "learning_rate": 1e-05,
+      "loss": 2.3059,
+      "step": 1105
+    },
+    {
+      "epoch": 1.861561119293078,
+      "grad_norm": 0.1489906907081604,
+      "learning_rate": 1e-05,
+      "loss": 2.2195,
+      "step": 1106
+    },
+    {
+      "epoch": 1.8632442667788764,
+      "grad_norm": 0.14351260662078857,
+      "learning_rate": 1e-05,
+      "loss": 2.2656,
+      "step": 1107
+    },
+    {
+      "epoch": 1.8649274142646748,
+      "grad_norm": 0.16010256111621857,
+      "learning_rate": 1e-05,
+      "loss": 2.3252,
+      "step": 1108
+    },
+    {
+      "epoch": 1.8666105617504734,
+      "grad_norm": 0.14475148916244507,
+      "learning_rate": 1e-05,
+      "loss": 2.2878,
+      "step": 1109
+    },
+    {
+      "epoch": 1.868293709236272,
+      "grad_norm": 0.14097367227077484,
+      "learning_rate": 1e-05,
+      "loss": 2.3716,
+      "step": 1110
+    },
+    {
+      "epoch": 1.8699768567220703,
+      "grad_norm": 0.15699978172779083,
+      "learning_rate": 1e-05,
+      "loss": 2.1678,
+      "step": 1111
+    },
+    {
+      "epoch": 1.8716600042078686,
+      "grad_norm": 0.1370065063238144,
+      "learning_rate": 1e-05,
+      "loss": 2.3315,
+      "step": 1112
+    },
+    {
+      "epoch": 1.8733431516936672,
+      "grad_norm": 0.1498231291770935,
+      "learning_rate": 1e-05,
+      "loss": 2.2949,
+      "step": 1113
+    },
+    {
+      "epoch": 1.8750262991794657,
+      "grad_norm": 0.13267523050308228,
+      "learning_rate": 1e-05,
+      "loss": 2.3535,
+      "step": 1114
+    },
+    {
+      "epoch": 1.876709446665264,
+      "grad_norm": 0.1453379988670349,
+      "learning_rate": 1e-05,
+      "loss": 2.2791,
+      "step": 1115
+    },
+    {
+      "epoch": 1.8783925941510624,
+      "grad_norm": 0.15499484539031982,
+      "learning_rate": 1e-05,
+      "loss": 2.2085,
+      "step": 1116
+    },
+    {
+      "epoch": 1.880075741636861,
+      "grad_norm": 0.14418251812458038,
+      "learning_rate": 1e-05,
+      "loss": 2.2793,
+      "step": 1117
+    },
+    {
+      "epoch": 1.8817588891226595,
+      "grad_norm": 0.13686548173427582,
+      "learning_rate": 1e-05,
+      "loss": 2.4175,
+      "step": 1118
+    },
+    {
+      "epoch": 1.8834420366084577,
+      "grad_norm": 0.17202888429164886,
+      "learning_rate": 1e-05,
+      "loss": 2.2196,
+      "step": 1119
+    },
+    {
+      "epoch": 1.8851251840942562,
+      "grad_norm": 0.1437048763036728,
+      "learning_rate": 1e-05,
+      "loss": 2.2688,
+      "step": 1120
+    },
+    {
+      "epoch": 1.8868083315800548,
+      "grad_norm": 0.13868288695812225,
+      "learning_rate": 1e-05,
+      "loss": 2.2971,
+      "step": 1121
+    },
+    {
+      "epoch": 1.8884914790658531,
+      "grad_norm": 0.133874773979187,
+      "learning_rate": 1e-05,
+      "loss": 2.3228,
+      "step": 1122
+    },
+    {
+      "epoch": 1.8901746265516515,
+      "grad_norm": 0.15967018902301788,
+      "learning_rate": 1e-05,
+      "loss": 2.2346,
+      "step": 1123
+    },
+    {
+      "epoch": 1.89185777403745,
+      "grad_norm": 0.15074019134044647,
+      "learning_rate": 1e-05,
+      "loss": 2.3577,
+      "step": 1124
+    },
+    {
+      "epoch": 1.8935409215232486,
+      "grad_norm": 0.13931475579738617,
+      "learning_rate": 1e-05,
+      "loss": 2.3789,
+      "step": 1125
+    },
+    {
+      "epoch": 1.895224069009047,
+      "grad_norm": 0.15354882180690765,
+      "learning_rate": 1e-05,
+      "loss": 2.184,
+      "step": 1126
+    },
+    {
+      "epoch": 1.8969072164948453,
+      "grad_norm": 0.15907764434814453,
+      "learning_rate": 1e-05,
+      "loss": 2.3638,
+      "step": 1127
+    },
+    {
+      "epoch": 1.8985903639806438,
+      "grad_norm": 0.13138049840927124,
+      "learning_rate": 1e-05,
+      "loss": 2.4543,
+      "step": 1128
+    },
+    {
+      "epoch": 1.9002735114664424,
+      "grad_norm": 0.14568856358528137,
+      "learning_rate": 1e-05,
+      "loss": 2.3064,
+      "step": 1129
+    },
+    {
+      "epoch": 1.9019566589522405,
+      "grad_norm": 0.1426182985305786,
+      "learning_rate": 1e-05,
+      "loss": 2.3223,
+      "step": 1130
+    },
+    {
+      "epoch": 1.903639806438039,
+      "grad_norm": 0.13313454389572144,
+      "learning_rate": 1e-05,
+      "loss": 2.3953,
+      "step": 1131
+    },
+    {
+      "epoch": 1.9053229539238377,
+      "grad_norm": 0.16987952589988708,
+      "learning_rate": 1e-05,
+      "loss": 2.1274,
+      "step": 1132
+    },
+    {
+      "epoch": 1.907006101409636,
+      "grad_norm": 0.1408863216638565,
+      "learning_rate": 1e-05,
+      "loss": 2.3242,
+      "step": 1133
+    },
+    {
+      "epoch": 1.9086892488954343,
+      "grad_norm": 0.14704225957393646,
+      "learning_rate": 1e-05,
+      "loss": 2.3687,
+      "step": 1134
+    },
+    {
+      "epoch": 1.910372396381233,
+      "grad_norm": 0.18410103023052216,
+      "learning_rate": 1e-05,
+      "loss": 2.1222,
+      "step": 1135
+    },
+    {
+      "epoch": 1.9120555438670315,
+      "grad_norm": 0.13889069855213165,
+      "learning_rate": 1e-05,
+      "loss": 2.3165,
+      "step": 1136
+    },
+    {
+      "epoch": 1.9137386913528298,
+      "grad_norm": 0.1532329022884369,
+      "learning_rate": 1e-05,
+      "loss": 2.2913,
+      "step": 1137
+    },
+    {
+      "epoch": 1.9154218388386282,
+      "grad_norm": 0.14806988835334778,
+      "learning_rate": 1e-05,
+      "loss": 2.2239,
+      "step": 1138
+    },
+    {
+      "epoch": 1.9171049863244267,
+      "grad_norm": 0.14964371919631958,
+      "learning_rate": 1e-05,
+      "loss": 2.2639,
+      "step": 1139
+    },
+    {
+      "epoch": 1.9187881338102253,
+      "grad_norm": 0.15137715637683868,
+      "learning_rate": 1e-05,
+      "loss": 2.3096,
+      "step": 1140
+    },
+    {
+      "epoch": 1.9204712812960234,
+      "grad_norm": 0.15892736613750458,
+      "learning_rate": 1e-05,
+      "loss": 2.3163,
+      "step": 1141
+    },
+    {
+      "epoch": 1.922154428781822,
+      "grad_norm": 0.15544387698173523,
+      "learning_rate": 1e-05,
+      "loss": 2.1825,
+      "step": 1142
+    },
+    {
+      "epoch": 1.9238375762676205,
+      "grad_norm": 0.14712852239608765,
+      "learning_rate": 1e-05,
+      "loss": 2.2659,
+      "step": 1143
+    },
+    {
+      "epoch": 1.9255207237534189,
+      "grad_norm": 0.1436305195093155,
+      "learning_rate": 1e-05,
+      "loss": 2.3101,
+      "step": 1144
+    },
+    {
+      "epoch": 1.9272038712392172,
+      "grad_norm": 0.16642406582832336,
+      "learning_rate": 1e-05,
+      "loss": 2.2156,
+      "step": 1145
+    },
+    {
+      "epoch": 1.9288870187250158,
+      "grad_norm": 0.16517338156700134,
+      "learning_rate": 1e-05,
+      "loss": 2.2561,
+      "step": 1146
+    },
+    {
+      "epoch": 1.9305701662108143,
+      "grad_norm": 0.1337500959634781,
+      "learning_rate": 1e-05,
+      "loss": 2.3818,
+      "step": 1147
+    },
+    {
+      "epoch": 1.9322533136966127,
+      "grad_norm": 0.15977586805820465,
+      "learning_rate": 1e-05,
+      "loss": 2.2377,
+      "step": 1148
+    },
+    {
+      "epoch": 1.933936461182411,
+      "grad_norm": 0.14951424300670624,
+      "learning_rate": 1e-05,
+      "loss": 2.2269,
+      "step": 1149
+    },
+    {
+      "epoch": 1.9356196086682096,
+      "grad_norm": 0.13450993597507477,
+      "learning_rate": 1e-05,
+      "loss": 2.3442,
+      "step": 1150
+    },
+    {
+      "epoch": 1.9373027561540082,
+      "grad_norm": 0.16469308733940125,
+      "learning_rate": 1e-05,
+      "loss": 2.3123,
+      "step": 1151
+    },
+    {
+      "epoch": 1.9389859036398063,
+      "grad_norm": 0.14135532081127167,
+      "learning_rate": 1e-05,
+      "loss": 2.387,
+      "step": 1152
+    },
+    {
+      "epoch": 1.9406690511256048,
+      "grad_norm": 0.13864876329898834,
+      "learning_rate": 1e-05,
+      "loss": 2.2661,
+      "step": 1153
+    },
+    {
+      "epoch": 1.9423521986114034,
+      "grad_norm": 0.16291983425617218,
+      "learning_rate": 1e-05,
+      "loss": 2.2617,
+      "step": 1154
+    },
+    {
+      "epoch": 1.9440353460972017,
+      "grad_norm": 0.13341820240020752,
+      "learning_rate": 1e-05,
+      "loss": 2.4299,
+      "step": 1155
+    },
+    {
+      "epoch": 1.945718493583,
+      "grad_norm": 0.15701517462730408,
+      "learning_rate": 1e-05,
+      "loss": 2.2211,
+      "step": 1156
+    },
+    {
+      "epoch": 1.9474016410687987,
+      "grad_norm": 0.16075365245342255,
+      "learning_rate": 1e-05,
+      "loss": 2.1801,
+      "step": 1157
+    },
+    {
+      "epoch": 1.9490847885545972,
+      "grad_norm": 0.15631234645843506,
+      "learning_rate": 1e-05,
+      "loss": 2.2152,
+      "step": 1158
+    },
+    {
+      "epoch": 1.9507679360403956,
+      "grad_norm": 0.16927126049995422,
+      "learning_rate": 1e-05,
+      "loss": 2.1776,
+      "step": 1159
+    },
+    {
+      "epoch": 1.952451083526194,
+      "grad_norm": 0.15192179381847382,
+      "learning_rate": 1e-05,
+      "loss": 2.2812,
+      "step": 1160
+    },
+    {
+      "epoch": 1.9541342310119925,
+      "grad_norm": 0.145833820104599,
+      "learning_rate": 1e-05,
+      "loss": 2.3124,
+      "step": 1161
+    },
+    {
+      "epoch": 1.955817378497791,
+      "grad_norm": 0.16952313482761383,
+      "learning_rate": 1e-05,
+      "loss": 2.1085,
+      "step": 1162
+    },
+    {
+      "epoch": 1.9575005259835891,
+      "grad_norm": 0.1629469394683838,
+      "learning_rate": 1e-05,
+      "loss": 2.2267,
+      "step": 1163
+    },
+    {
+      "epoch": 1.9591836734693877,
+      "grad_norm": 0.16672489047050476,
+      "learning_rate": 1e-05,
+      "loss": 2.3783,
+      "step": 1164
+    },
+    {
+      "epoch": 1.9608668209551863,
+      "grad_norm": 0.14810308814048767,
+      "learning_rate": 1e-05,
+      "loss": 2.3723,
+      "step": 1165
+    },
+    {
+      "epoch": 1.9625499684409846,
+      "grad_norm": 0.1435479074716568,
+      "learning_rate": 1e-05,
+      "loss": 2.2615,
+      "step": 1166
+    },
+    {
+      "epoch": 1.964233115926783,
+      "grad_norm": 0.149140864610672,
+      "learning_rate": 1e-05,
+      "loss": 2.2134,
+      "step": 1167
+    },
+    {
+      "epoch": 1.9659162634125815,
+      "grad_norm": 0.17785809934139252,
+      "learning_rate": 1e-05,
+      "loss": 2.1993,
+      "step": 1168
+    },
+    {
+      "epoch": 1.96759941089838,
+      "grad_norm": 0.15931861102581024,
+      "learning_rate": 1e-05,
+      "loss": 2.1807,
+      "step": 1169
+    },
+    {
+      "epoch": 1.9692825583841784,
+      "grad_norm": 0.16015268862247467,
+      "learning_rate": 1e-05,
+      "loss": 2.2737,
+      "step": 1170
+    },
+    {
+      "epoch": 1.9709657058699768,
+      "grad_norm": 0.14189362525939941,
+      "learning_rate": 1e-05,
+      "loss": 2.3416,
+      "step": 1171
+    },
+    {
+      "epoch": 1.9726488533557753,
+      "grad_norm": 0.1655077338218689,
+      "learning_rate": 1e-05,
+      "loss": 2.184,
+      "step": 1172
+    },
+    {
+      "epoch": 1.974332000841574,
+      "grad_norm": 0.17838408052921295,
+      "learning_rate": 1e-05,
+      "loss": 2.2466,
+      "step": 1173
+    },
+    {
+      "epoch": 1.9760151483273722,
+      "grad_norm": 0.16605247557163239,
+      "learning_rate": 1e-05,
+      "loss": 2.2019,
+      "step": 1174
+    },
+    {
+      "epoch": 1.9776982958131706,
+      "grad_norm": 0.15444627404212952,
+      "learning_rate": 1e-05,
+      "loss": 2.2382,
+      "step": 1175
+    },
+    {
+      "epoch": 1.9793814432989691,
+      "grad_norm": 0.15730591118335724,
+      "learning_rate": 1e-05,
+      "loss": 2.3335,
+      "step": 1176
+    },
+    {
+      "epoch": 1.9810645907847675,
+      "grad_norm": 0.17332051694393158,
+      "learning_rate": 1e-05,
+      "loss": 2.17,
+      "step": 1177
+    },
+    {
+      "epoch": 1.9827477382705658,
+      "grad_norm": 0.15129022300243378,
+      "learning_rate": 1e-05,
+      "loss": 2.2584,
+      "step": 1178
+    },
+    {
+      "epoch": 1.9844308857563644,
+      "grad_norm": 0.16302135586738586,
+      "learning_rate": 1e-05,
+      "loss": 2.1904,
+      "step": 1179
+    },
+    {
+      "epoch": 1.986114033242163,
+      "grad_norm": 0.14117322862148285,
+      "learning_rate": 1e-05,
+      "loss": 2.3611,
+      "step": 1180
+    },
+    {
+      "epoch": 1.9877971807279613,
+      "grad_norm": 0.14415599405765533,
+      "learning_rate": 1e-05,
+      "loss": 2.3503,
+      "step": 1181
+    },
+    {
+      "epoch": 1.9894803282137596,
+      "grad_norm": 0.15894141793251038,
+      "learning_rate": 1e-05,
+      "loss": 2.2253,
+      "step": 1182
+    },
+    {
+      "epoch": 1.9911634756995582,
+      "grad_norm": 0.15063215792179108,
+      "learning_rate": 1e-05,
+      "loss": 2.303,
+      "step": 1183
+    },
+    {
+      "epoch": 1.9928466231853568,
+      "grad_norm": 0.15843670070171356,
+      "learning_rate": 1e-05,
+      "loss": 2.2959,
+      "step": 1184
+    },
+    {
+      "epoch": 1.9945297706711551,
+      "grad_norm": 0.1457902193069458,
+      "learning_rate": 1e-05,
+      "loss": 2.3396,
+      "step": 1185
+    },
+    {
+      "epoch": 1.9962129181569535,
+      "grad_norm": 0.1694038361310959,
+      "learning_rate": 1e-05,
+      "loss": 2.3169,
+      "step": 1186
+    },
+    {
+      "epoch": 1.997896065642752,
+      "grad_norm": 0.16121593117713928,
+      "learning_rate": 1e-05,
+      "loss": 2.2754,
+      "step": 1187
+    },
+    {
+      "epoch": 1.9995792131285504,
+      "grad_norm": 0.16226674616336823,
+      "learning_rate": 1e-05,
+      "loss": 2.2498,
+      "step": 1188
+    },
+    {
+      "epoch": 1.9995792131285504,
+      "step": 1188,
+      "total_flos": 2.494777289795961e+18,
+      "train_loss": 2.3332799018834174,
+      "train_runtime": 81586.2049,
+      "train_samples_per_second": 0.932,
+      "train_steps_per_second": 0.015
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 1188,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 2,
+  "save_steps": 1000,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 2.494777289795961e+18,
+  "train_batch_size": 8,
+  "trial_name": null,
+  "trial_params": null
+}