diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,38019 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.0,
+  "eval_steps": 500,
+  "global_step": 5427,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.2905577863476845,
+      "learning_rate": 1.226993865030675e-06,
+      "loss": 1.9981,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 8.60152717425747,
+      "learning_rate": 2.45398773006135e-06,
+      "loss": 2.02,
+      "step": 2
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 9.53940432007905,
+      "learning_rate": 3.680981595092025e-06,
+      "loss": 2.047,
+      "step": 3
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 8.015064596950538,
+      "learning_rate": 4.9079754601227e-06,
+      "loss": 2.1075,
+      "step": 4
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 7.459744620478235,
+      "learning_rate": 6.134969325153374e-06,
+      "loss": 1.9171,
+      "step": 5
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.660756140227556,
+      "learning_rate": 7.36196319018405e-06,
+      "loss": 2.0556,
+      "step": 6
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.885433777665061,
+      "learning_rate": 8.588957055214725e-06,
+      "loss": 1.7545,
+      "step": 7
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.6521949301853467,
+      "learning_rate": 9.8159509202454e-06,
+      "loss": 1.4498,
+      "step": 8
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.6828932604373903,
+      "learning_rate": 1.1042944785276074e-05,
+      "loss": 1.7211,
+      "step": 9
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.3343721677155607,
+      "learning_rate": 1.2269938650306748e-05,
+      "loss": 1.5748,
+      "step": 10
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.3371187169880896,
+      "learning_rate": 1.3496932515337424e-05,
+      "loss": 1.4964,
+      "step": 11
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.3687029856862007,
+      "learning_rate": 1.47239263803681e-05,
+      "loss": 1.588,
+      "step": 12
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.3106658156351136,
+      "learning_rate": 1.5950920245398772e-05,
+      "loss": 1.5338,
+      "step": 13
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.079167769083421,
+      "learning_rate": 1.717791411042945e-05,
+      "loss": 1.745,
+      "step": 14
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.3939238984580076,
+      "learning_rate": 1.8404907975460123e-05,
+      "loss": 1.6062,
+      "step": 15
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.0342757484682983,
+      "learning_rate": 1.96319018404908e-05,
+      "loss": 1.4304,
+      "step": 16
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.9492372789977062,
+      "learning_rate": 2.085889570552147e-05,
+      "loss": 1.7929,
+      "step": 17
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.9630350542305112,
+      "learning_rate": 2.208588957055215e-05,
+      "loss": 1.4138,
+      "step": 18
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.9208174071784341,
+      "learning_rate": 2.3312883435582822e-05,
+      "loss": 1.7109,
+      "step": 19
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.8195331033752657,
+      "learning_rate": 2.4539877300613496e-05,
+      "loss": 1.7576,
+      "step": 20
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.7696374705502098,
+      "learning_rate": 2.5766871165644174e-05,
+      "loss": 1.6369,
+      "step": 21
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.8593457344227575,
+      "learning_rate": 2.6993865030674848e-05,
+      "loss": 1.6439,
+      "step": 22
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.7649248332016724,
+      "learning_rate": 2.822085889570552e-05,
+      "loss": 1.5197,
+      "step": 23
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.8833605685032768,
+      "learning_rate": 2.94478527607362e-05,
+      "loss": 1.4393,
+      "step": 24
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.6767579677558123,
+      "learning_rate": 3.067484662576687e-05,
+      "loss": 1.2923,
+      "step": 25
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.6780229820916917,
+      "learning_rate": 3.1901840490797544e-05,
+      "loss": 1.4217,
+      "step": 26
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.6850333132156673,
+      "learning_rate": 3.312883435582822e-05,
+      "loss": 1.5007,
+      "step": 27
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.646393816692974,
+      "learning_rate": 3.43558282208589e-05,
+      "loss": 1.4623,
+      "step": 28
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6523162820479583,
+      "learning_rate": 3.558282208588957e-05,
+      "loss": 1.4672,
+      "step": 29
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6983745920782584,
+      "learning_rate": 3.6809815950920246e-05,
+      "loss": 1.4794,
+      "step": 30
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6770407411908272,
+      "learning_rate": 3.8036809815950924e-05,
+      "loss": 1.5355,
+      "step": 31
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5965073130776971,
+      "learning_rate": 3.92638036809816e-05,
+      "loss": 1.4108,
+      "step": 32
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5809380808063968,
+      "learning_rate": 4.049079754601227e-05,
+      "loss": 1.4227,
+      "step": 33
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6058422039784557,
+      "learning_rate": 4.171779141104294e-05,
+      "loss": 1.3593,
+      "step": 34
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6785877265555837,
+      "learning_rate": 4.2944785276073626e-05,
+      "loss": 1.3981,
+      "step": 35
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5816654932887596,
+      "learning_rate": 4.41717791411043e-05,
+      "loss": 1.4666,
+      "step": 36
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6677557567384471,
+      "learning_rate": 4.539877300613497e-05,
+      "loss": 1.3718,
+      "step": 37
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5838366348484775,
+      "learning_rate": 4.6625766871165645e-05,
+      "loss": 1.5353,
+      "step": 38
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5878895707371306,
+      "learning_rate": 4.785276073619632e-05,
+      "loss": 1.3469,
+      "step": 39
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7367251372928045,
+      "learning_rate": 4.907975460122699e-05,
+      "loss": 1.4122,
+      "step": 40
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6375262107780414,
+      "learning_rate": 5.030674846625767e-05,
+      "loss": 1.3299,
+      "step": 41
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5695248304795745,
+      "learning_rate": 5.153374233128835e-05,
+      "loss": 1.5366,
+      "step": 42
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5202998414513565,
+      "learning_rate": 5.276073619631902e-05,
+      "loss": 1.4326,
+      "step": 43
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6368872077764169,
+      "learning_rate": 5.3987730061349695e-05,
+      "loss": 1.4383,
+      "step": 44
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5679733996528148,
+      "learning_rate": 5.521472392638037e-05,
+      "loss": 1.624,
+      "step": 45
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.591722290374685,
+      "learning_rate": 5.644171779141104e-05,
+      "loss": 1.2775,
+      "step": 46
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5683465483571614,
+      "learning_rate": 5.766871165644172e-05,
+      "loss": 1.3342,
+      "step": 47
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6299744869445382,
+      "learning_rate": 5.88957055214724e-05,
+      "loss": 1.4763,
+      "step": 48
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5689888757085202,
+      "learning_rate": 6.012269938650307e-05,
+      "loss": 1.4454,
+      "step": 49
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.571564732715942,
+      "learning_rate": 6.134969325153375e-05,
+      "loss": 1.3637,
+      "step": 50
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.4949230703867982,
+      "learning_rate": 6.257668711656443e-05,
+      "loss": 1.5494,
+      "step": 51
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.49391515040180656,
+      "learning_rate": 6.380368098159509e-05,
+      "loss": 1.4161,
+      "step": 52
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5229707681697312,
+      "learning_rate": 6.503067484662577e-05,
+      "loss": 1.2066,
+      "step": 53
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5555103812363782,
+      "learning_rate": 6.625766871165644e-05,
+      "loss": 1.3581,
+      "step": 54
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5075151100417339,
+      "learning_rate": 6.748466257668711e-05,
+      "loss": 1.1996,
+      "step": 55
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6838211123291117,
+      "learning_rate": 6.87116564417178e-05,
+      "loss": 1.5435,
+      "step": 56
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.47346306261065885,
+      "learning_rate": 6.993865030674847e-05,
+      "loss": 1.3149,
+      "step": 57
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5615333835343251,
+      "learning_rate": 7.116564417177914e-05,
+      "loss": 1.3024,
+      "step": 58
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5392929876499301,
+      "learning_rate": 7.239263803680982e-05,
+      "loss": 1.4969,
+      "step": 59
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5464960497292252,
+      "learning_rate": 7.361963190184049e-05,
+      "loss": 1.4429,
+      "step": 60
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5325472423867312,
+      "learning_rate": 7.484662576687118e-05,
+      "loss": 1.4705,
+      "step": 61
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5428688515525038,
+      "learning_rate": 7.607361963190185e-05,
+      "loss": 1.1789,
+      "step": 62
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5088510241986943,
+      "learning_rate": 7.730061349693252e-05,
+      "loss": 1.2011,
+      "step": 63
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5586576976510322,
+      "learning_rate": 7.85276073619632e-05,
+      "loss": 1.3758,
+      "step": 64
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5119718192594475,
+      "learning_rate": 7.975460122699386e-05,
+      "loss": 1.4065,
+      "step": 65
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.537125566778921,
+      "learning_rate": 8.098159509202454e-05,
+      "loss": 1.3617,
+      "step": 66
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5755112582830035,
+      "learning_rate": 8.220858895705523e-05,
+      "loss": 1.2627,
+      "step": 67
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5551604748951967,
+      "learning_rate": 8.343558282208588e-05,
+      "loss": 1.4715,
+      "step": 68
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5704764578935172,
+      "learning_rate": 8.466257668711657e-05,
+      "loss": 1.3007,
+      "step": 69
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5126838462346063,
+      "learning_rate": 8.588957055214725e-05,
+      "loss": 1.4847,
+      "step": 70
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5784206864125514,
+      "learning_rate": 8.711656441717791e-05,
+      "loss": 1.2715,
+      "step": 71
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6285045924851832,
+      "learning_rate": 8.83435582822086e-05,
+      "loss": 1.3654,
+      "step": 72
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.46624772514910845,
+      "learning_rate": 8.957055214723928e-05,
+      "loss": 1.2995,
+      "step": 73
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5204635765058973,
+      "learning_rate": 9.079754601226993e-05,
+      "loss": 1.2347,
+      "step": 74
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6029673465756028,
+      "learning_rate": 9.202453987730062e-05,
+      "loss": 1.2532,
+      "step": 75
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5476094724463323,
+      "learning_rate": 9.325153374233129e-05,
+      "loss": 1.5092,
+      "step": 76
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9311311203940346,
+      "learning_rate": 9.447852760736196e-05,
+      "loss": 1.1665,
+      "step": 77
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5531547219834427,
+      "learning_rate": 9.570552147239264e-05,
+      "loss": 1.3519,
+      "step": 78
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5836032937635454,
+      "learning_rate": 9.693251533742331e-05,
+      "loss": 1.44,
+      "step": 79
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5823518010609838,
+      "learning_rate": 9.815950920245399e-05,
+      "loss": 1.1173,
+      "step": 80
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5454905247646865,
+      "learning_rate": 9.938650306748467e-05,
+      "loss": 1.3969,
+      "step": 81
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6036403589297964,
+      "learning_rate": 0.00010061349693251534,
+      "loss": 1.0596,
+      "step": 82
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5222396979663078,
+      "learning_rate": 0.00010184049079754601,
+      "loss": 1.1901,
+      "step": 83
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5356971386185464,
+      "learning_rate": 0.0001030674846625767,
+      "loss": 1.2952,
+      "step": 84
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5217824409597704,
+      "learning_rate": 0.00010429447852760737,
+      "loss": 1.2648,
+      "step": 85
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5082680609192889,
+      "learning_rate": 0.00010552147239263804,
+      "loss": 1.5866,
+      "step": 86
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6522041943713067,
+      "learning_rate": 0.00010674846625766872,
+      "loss": 1.4486,
+      "step": 87
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5571917929264755,
+      "learning_rate": 0.00010797546012269939,
+      "loss": 1.4628,
+      "step": 88
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.4837268330793875,
+      "learning_rate": 0.00010920245398773006,
+      "loss": 1.2189,
+      "step": 89
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6165940875434999,
+      "learning_rate": 0.00011042944785276075,
+      "loss": 1.2151,
+      "step": 90
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.4981148796155304,
+      "learning_rate": 0.00011165644171779142,
+      "loss": 1.3596,
+      "step": 91
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6332638620763956,
+      "learning_rate": 0.00011288343558282209,
+      "loss": 1.4594,
+      "step": 92
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6796054477882439,
+      "learning_rate": 0.00011411042944785277,
+      "loss": 1.2697,
+      "step": 93
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6067443106291492,
+      "learning_rate": 0.00011533742331288344,
+      "loss": 1.4769,
+      "step": 94
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5925856064517426,
+      "learning_rate": 0.00011656441717791411,
+      "loss": 1.5025,
+      "step": 95
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5310497894667259,
+      "learning_rate": 0.0001177914110429448,
+      "loss": 1.488,
+      "step": 96
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8136324289347706,
+      "learning_rate": 0.00011901840490797547,
+      "loss": 1.3852,
+      "step": 97
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5440365896109094,
+      "learning_rate": 0.00012024539877300614,
+      "loss": 1.5093,
+      "step": 98
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5012647831126298,
+      "learning_rate": 0.00012147239263803682,
+      "loss": 1.2701,
+      "step": 99
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.617958377362165,
+      "learning_rate": 0.0001226993865030675,
+      "loss": 1.4864,
+      "step": 100
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.49448354948330303,
+      "learning_rate": 0.00012392638036809816,
+      "loss": 1.3339,
+      "step": 101
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.4800442874745163,
+      "learning_rate": 0.00012515337423312886,
+      "loss": 1.2988,
+      "step": 102
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5590632538347917,
+      "learning_rate": 0.00012638036809815953,
+      "loss": 1.2703,
+      "step": 103
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5006572543712501,
+      "learning_rate": 0.00012760736196319017,
+      "loss": 1.2579,
+      "step": 104
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5086917749645221,
+      "learning_rate": 0.00012883435582822084,
+      "loss": 1.3668,
+      "step": 105
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.48696461030475025,
+      "learning_rate": 0.00013006134969325154,
+      "loss": 1.5487,
+      "step": 106
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5555982168272745,
+      "learning_rate": 0.0001312883435582822,
+      "loss": 1.2003,
+      "step": 107
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.4640323344809559,
+      "learning_rate": 0.00013251533742331288,
+      "loss": 1.3318,
+      "step": 108
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5367038653755865,
+      "learning_rate": 0.00013374233128834358,
+      "loss": 1.2525,
+      "step": 109
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5771325930722303,
+      "learning_rate": 0.00013496932515337422,
+      "loss": 1.4065,
+      "step": 110
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.49127497811841814,
+      "learning_rate": 0.0001361963190184049,
+      "loss": 1.221,
+      "step": 111
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.49326792629066846,
+      "learning_rate": 0.0001374233128834356,
+      "loss": 1.3332,
+      "step": 112
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.467745363834521,
+      "learning_rate": 0.00013865030674846626,
+      "loss": 1.3849,
+      "step": 113
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.4622903974041499,
+      "learning_rate": 0.00013987730061349693,
+      "loss": 1.3516,
+      "step": 114
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5382678447841127,
+      "learning_rate": 0.00014110429447852763,
+      "loss": 1.1172,
+      "step": 115
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.46631778692663756,
+      "learning_rate": 0.00014233128834355828,
+      "loss": 1.5167,
+      "step": 116
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.4542913890506678,
+      "learning_rate": 0.00014355828220858895,
+      "loss": 1.361,
+      "step": 117
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5236763540676828,
+      "learning_rate": 0.00014478527607361964,
+      "loss": 1.3149,
+      "step": 118
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5142605074524307,
+      "learning_rate": 0.00014601226993865031,
+      "loss": 1.3184,
+      "step": 119
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5045038182823308,
+      "learning_rate": 0.00014723926380368098,
+      "loss": 1.3892,
+      "step": 120
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5023215888839109,
+      "learning_rate": 0.00014846625766871168,
+      "loss": 1.3857,
+      "step": 121
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.47848492541905024,
+      "learning_rate": 0.00014969325153374235,
+      "loss": 1.236,
+      "step": 122
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.44385910407098317,
+      "learning_rate": 0.000150920245398773,
+      "loss": 1.3179,
+      "step": 123
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5118913531619579,
+      "learning_rate": 0.0001521472392638037,
+      "loss": 1.298,
+      "step": 124
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5166797195868819,
+      "learning_rate": 0.00015337423312883436,
+      "loss": 1.3227,
+      "step": 125
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6559738815916184,
+      "learning_rate": 0.00015460122699386504,
+      "loss": 1.0786,
+      "step": 126
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.47765331929258653,
+      "learning_rate": 0.00015582822085889573,
+      "loss": 1.3473,
+      "step": 127
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.4623432779072936,
+      "learning_rate": 0.0001570552147239264,
+      "loss": 1.3558,
+      "step": 128
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5050475584197706,
+      "learning_rate": 0.00015828220858895705,
+      "loss": 1.3509,
+      "step": 129
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5534861156335383,
+      "learning_rate": 0.00015950920245398772,
+      "loss": 1.1708,
+      "step": 130
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5208167369090876,
+      "learning_rate": 0.00016073619631901842,
+      "loss": 1.2094,
+      "step": 131
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5067359989027647,
+      "learning_rate": 0.00016196319018404909,
+      "loss": 1.3363,
+      "step": 132
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.46599527625483567,
+      "learning_rate": 0.00016319018404907976,
+      "loss": 1.4461,
+      "step": 133
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5034696068421803,
+      "learning_rate": 0.00016441717791411045,
+      "loss": 1.261,
+      "step": 134
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5252934959715996,
+      "learning_rate": 0.0001656441717791411,
+      "loss": 1.1857,
+      "step": 135
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.4032564610519412,
+      "learning_rate": 0.00016687116564417177,
+      "loss": 1.3335,
+      "step": 136
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5065968775793448,
+      "learning_rate": 0.00016809815950920247,
+      "loss": 1.148,
+      "step": 137
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5442320041786156,
+      "learning_rate": 0.00016932515337423314,
+      "loss": 1.3375,
+      "step": 138
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.47819303174740263,
+      "learning_rate": 0.0001705521472392638,
+      "loss": 1.3024,
+      "step": 139
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.489951707859478,
+      "learning_rate": 0.0001717791411042945,
+      "loss": 1.5563,
+      "step": 140
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.45338307642420794,
+      "learning_rate": 0.00017300613496932518,
+      "loss": 1.3691,
+      "step": 141
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.4527007402077501,
+      "learning_rate": 0.00017423312883435582,
+      "loss": 1.1568,
+      "step": 142
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.4419066128321064,
+      "learning_rate": 0.00017546012269938652,
+      "loss": 1.3463,
+      "step": 143
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.4487113454882587,
+      "learning_rate": 0.0001766871165644172,
+      "loss": 1.246,
+      "step": 144
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.46097765309256383,
+      "learning_rate": 0.00017791411042944786,
+      "loss": 1.2458,
+      "step": 145
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.4891405742621151,
+      "learning_rate": 0.00017914110429447856,
+      "loss": 1.3367,
+      "step": 146
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.536169078136946,
+      "learning_rate": 0.00018036809815950923,
+      "loss": 1.3382,
+      "step": 147
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.4167340201924026,
+      "learning_rate": 0.00018159509202453987,
+      "loss": 1.2018,
+      "step": 148
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.49235032866531364,
+      "learning_rate": 0.00018282208588957057,
+      "loss": 1.2444,
+      "step": 149
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.4579211030151233,
+      "learning_rate": 0.00018404907975460124,
+      "loss": 1.2537,
+      "step": 150
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.4282460190126454,
+      "learning_rate": 0.0001852760736196319,
+      "loss": 1.2146,
+      "step": 151
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.4501191678752776,
+      "learning_rate": 0.00018650306748466258,
+      "loss": 1.2572,
+      "step": 152
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.4443323852869221,
+      "learning_rate": 0.00018773006134969328,
+      "loss": 1.3652,
+      "step": 153
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.4398880300682786,
+      "learning_rate": 0.00018895705521472392,
+      "loss": 1.3028,
+      "step": 154
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.4148867637970492,
+      "learning_rate": 0.0001901840490797546,
+      "loss": 1.278,
+      "step": 155
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.45293944822255516,
+      "learning_rate": 0.0001914110429447853,
+      "loss": 1.1631,
+      "step": 156
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.40155271870466475,
+      "learning_rate": 0.00019263803680981596,
+      "loss": 1.2031,
+      "step": 157
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.45678525356470495,
+      "learning_rate": 0.00019386503067484663,
+      "loss": 1.1569,
+      "step": 158
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.4564158122183654,
+      "learning_rate": 0.00019509202453987733,
+      "loss": 1.329,
+      "step": 159
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.4454709377667363,
+      "learning_rate": 0.00019631901840490797,
+      "loss": 1.4036,
+      "step": 160
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.4224771446995192,
+      "learning_rate": 0.00019754601226993864,
+      "loss": 1.3458,
+      "step": 161
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.42439293288690555,
+      "learning_rate": 0.00019877300613496934,
+      "loss": 1.1814,
+      "step": 162
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5104224261690504,
+      "learning_rate": 0.0002,
+      "loss": 1.3558,
+      "step": 163
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.4322164130766111,
+      "learning_rate": 0.0001999999821910639,
+      "loss": 1.3121,
+      "step": 164
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.4280053308236455,
+      "learning_rate": 0.00019999992876426194,
+      "loss": 1.2828,
+      "step": 165
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.40007995107032573,
+      "learning_rate": 0.00019999983971961317,
+      "loss": 1.4211,
+      "step": 166
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.46426294332858437,
+      "learning_rate": 0.00019999971505714928,
+      "loss": 1.279,
+      "step": 167
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.4329181894051252,
+      "learning_rate": 0.00019999955477691467,
+      "loss": 1.2577,
+      "step": 168
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.3741012690850515,
+      "learning_rate": 0.00019999935887896647,
+      "loss": 1.1394,
+      "step": 169
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.41574765508081574,
+      "learning_rate": 0.00019999912736337437,
+      "loss": 1.3026,
+      "step": 170
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.4333993364286849,
+      "learning_rate": 0.0001999988602302209,
+      "loss": 1.199,
+      "step": 171
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.4017817838162958,
+      "learning_rate": 0.00019999855747960122,
+      "loss": 1.2698,
+      "step": 172
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.39290834974138966,
+      "learning_rate": 0.00019999821911162316,
+      "loss": 1.2309,
+      "step": 173
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.38675291328972583,
+      "learning_rate": 0.00019999784512640716,
+      "loss": 1.3367,
+      "step": 174
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.4624159175389514,
+      "learning_rate": 0.00019999743552408648,
+      "loss": 1.3531,
+      "step": 175
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.43902564254391735,
+      "learning_rate": 0.00019999699030480704,
+      "loss": 1.6063,
+      "step": 176
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.499141860756029,
+      "learning_rate": 0.00019999650946872738,
+      "loss": 1.1721,
+      "step": 177
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.4387721782364717,
+      "learning_rate": 0.0001999959930160188,
+      "loss": 1.1274,
+      "step": 178
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.4706097219247533,
+      "learning_rate": 0.0001999954409468652,
+      "loss": 1.4383,
+      "step": 179
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.4666441880401181,
+      "learning_rate": 0.00019999485326146324,
+      "loss": 1.2922,
+      "step": 180
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.4311775291706574,
+      "learning_rate": 0.0001999942299600222,
+      "loss": 1.3576,
+      "step": 181
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.43224137247605715,
+      "learning_rate": 0.0001999935710427642,
+      "loss": 1.3341,
+      "step": 182
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.41141335860021566,
+      "learning_rate": 0.0001999928765099238,
+      "loss": 1.3477,
+      "step": 183
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.42334839736227714,
+      "learning_rate": 0.00019999214636174845,
+      "loss": 1.5055,
+      "step": 184
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5092909903445749,
+      "learning_rate": 0.00019999138059849822,
+      "loss": 1.1318,
+      "step": 185
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5010592694199599,
+      "learning_rate": 0.00019999057922044582,
+      "loss": 1.3128,
+      "step": 186
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.4307379748709144,
+      "learning_rate": 0.0001999897422278767,
+      "loss": 1.3879,
+      "step": 187
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.49995923673026516,
+      "learning_rate": 0.000199988869621089,
+      "loss": 1.3447,
+      "step": 188
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.48098670186413583,
+      "learning_rate": 0.00019998796140039348,
+      "loss": 1.3979,
+      "step": 189
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.4168714575588646,
+      "learning_rate": 0.00019998701756611368,
+      "loss": 1.2405,
+      "step": 190
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.4709161736096142,
+      "learning_rate": 0.00019998603811858571,
+      "loss": 1.368,
+      "step": 191
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.45243315815256646,
+      "learning_rate": 0.00019998502305815848,
+      "loss": 1.3088,
+      "step": 192
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5475894875067813,
+      "learning_rate": 0.00019998397238519354,
+      "loss": 1.239,
+      "step": 193
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.4824224956180037,
+      "learning_rate": 0.0001999828861000651,
+      "loss": 1.2091,
+      "step": 194
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.550085310172357,
+      "learning_rate": 0.00019998176420316002,
+      "loss": 1.1041,
+      "step": 195
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.3971359935756011,
+      "learning_rate": 0.000199980606694878,
+      "loss": 1.2141,
+      "step": 196
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.43860727906320846,
+      "learning_rate": 0.0001999794135756312,
+      "loss": 1.1889,
+      "step": 197
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.40120174433391165,
+      "learning_rate": 0.0001999781848458447,
+      "loss": 1.2919,
+      "step": 198
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5195588664189194,
+      "learning_rate": 0.00019997692050595606,
+      "loss": 1.4188,
+      "step": 199
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.41326897688409786,
+      "learning_rate": 0.00019997562055641568,
+      "loss": 1.4653,
+      "step": 200
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.4455198891818025,
+      "learning_rate": 0.0001999742849976865,
+      "loss": 1.2463,
+      "step": 201
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.4019477807183807,
+      "learning_rate": 0.00019997291383024428,
+      "loss": 1.3109,
+      "step": 202
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.48582433819912624,
+      "learning_rate": 0.0001999715070545774,
+      "loss": 1.4129,
+      "step": 203
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.3891071871654745,
+      "learning_rate": 0.00019997006467118686,
+      "loss": 1.3835,
+      "step": 204
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.49545819171062727,
+      "learning_rate": 0.00019996858668058646,
+      "loss": 1.0792,
+      "step": 205
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.39379341317736466,
+      "learning_rate": 0.00019996707308330263,
+      "loss": 1.3094,
+      "step": 206
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5423505776891887,
+      "learning_rate": 0.00019996552387987446,
+      "loss": 1.2837,
+      "step": 207
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.42668068294357364,
+      "learning_rate": 0.00019996393907085376,
+      "loss": 1.2822,
+      "step": 208
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.4212161145139675,
+      "learning_rate": 0.00019996231865680497,
+      "loss": 1.598,
+      "step": 209
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.41854089138400774,
+      "learning_rate": 0.00019996066263830531,
+      "loss": 1.0828,
+      "step": 210
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.4743355509288604,
+      "learning_rate": 0.00019995897101594454,
+      "loss": 1.2673,
+      "step": 211
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.4514777615516725,
+      "learning_rate": 0.00019995724379032526,
+      "loss": 1.2002,
+      "step": 212
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.41467371018966476,
+      "learning_rate": 0.0001999554809620626,
+      "loss": 1.3004,
+      "step": 213
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.43166212155463085,
+      "learning_rate": 0.0001999536825317845,
+      "loss": 1.0462,
+      "step": 214
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.624813583349147,
+      "learning_rate": 0.00019995184850013148,
+      "loss": 1.4028,
+      "step": 215
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.4835850237447664,
+      "learning_rate": 0.0001999499788677568,
+      "loss": 1.4125,
+      "step": 216
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5109770565839427,
+      "learning_rate": 0.00019994807363532638,
+      "loss": 1.3367,
+      "step": 217
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6760106189475126,
+      "learning_rate": 0.00019994613280351886,
+      "loss": 1.5947,
+      "step": 218
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.3825279997473377,
+      "learning_rate": 0.00019994415637302547,
+      "loss": 1.2147,
+      "step": 219
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.4554989137173263,
+      "learning_rate": 0.0001999421443445502,
+      "loss": 1.6,
+      "step": 220
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.4033050102925463,
+      "learning_rate": 0.00019994009671880968,
+      "loss": 1.2616,
+      "step": 221
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.42217463508252195,
+      "learning_rate": 0.00019993801349653322,
+      "loss": 1.2934,
+      "step": 222
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.4208816584767837,
+      "learning_rate": 0.00019993589467846284,
+      "loss": 1.2877,
+      "step": 223
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5158862137134327,
+      "learning_rate": 0.0001999337402653532,
+      "loss": 1.2171,
+      "step": 224
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.3880574427852324,
+      "learning_rate": 0.00019993155025797168,
+      "loss": 1.199,
+      "step": 225
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.4667510219033266,
+      "learning_rate": 0.0001999293246570983,
+      "loss": 1.2786,
+      "step": 226
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.452282818453412,
+      "learning_rate": 0.00019992706346352577,
+      "loss": 1.0956,
+      "step": 227
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.4949462391191208,
+      "learning_rate": 0.00019992476667805953,
+      "loss": 1.2919,
+      "step": 228
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.4403457937166642,
+      "learning_rate": 0.00019992243430151755,
+      "loss": 1.2689,
+      "step": 229
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.4145660701059602,
+      "learning_rate": 0.00019992006633473062,
+      "loss": 1.3334,
+      "step": 230
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.41743453135555486,
+      "learning_rate": 0.00019991766277854218,
+      "loss": 1.3746,
+      "step": 231
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5577925694311494,
+      "learning_rate": 0.0001999152236338083,
+      "loss": 1.3083,
+      "step": 232
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.4809163014439084,
+      "learning_rate": 0.00019991274890139774,
+      "loss": 1.3223,
+      "step": 233
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.4023596243918276,
+      "learning_rate": 0.000199910238582192,
+      "loss": 1.3831,
+      "step": 234
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.43458042193159485,
+      "learning_rate": 0.00019990769267708516,
+      "loss": 1.2791,
+      "step": 235
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.39848746491432346,
+      "learning_rate": 0.00019990511118698402,
+      "loss": 1.4319,
+      "step": 236
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.39622688996413136,
+      "learning_rate": 0.00019990249411280803,
+      "loss": 1.2857,
+      "step": 237
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.4872639714446551,
+      "learning_rate": 0.0001998998414554894,
+      "loss": 1.3528,
+      "step": 238
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.3647839055849367,
+      "learning_rate": 0.00019989715321597288,
+      "loss": 1.2505,
+      "step": 239
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.48773751245383523,
+      "learning_rate": 0.00019989442939521602,
+      "loss": 1.3403,
+      "step": 240
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.4313672229683762,
+      "learning_rate": 0.00019989166999418898,
+      "loss": 1.2955,
+      "step": 241
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.44300188028933335,
+      "learning_rate": 0.00019988887501387454,
+      "loss": 1.2923,
+      "step": 242
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.443623554672069,
+      "learning_rate": 0.00019988604445526827,
+      "loss": 1.311,
+      "step": 243
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5162076079749719,
+      "learning_rate": 0.00019988317831937837,
+      "loss": 1.323,
+      "step": 244
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.46730346801409617,
+      "learning_rate": 0.00019988027660722567,
+      "loss": 1.446,
+      "step": 245
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.38496929911492284,
+      "learning_rate": 0.00019987733931984365,
+      "loss": 1.4787,
+      "step": 246
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.40799992907048516,
+      "learning_rate": 0.0001998743664582786,
+      "loss": 1.3133,
+      "step": 247
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.43859106495060335,
+      "learning_rate": 0.00019987135802358933,
+      "loss": 1.2934,
+      "step": 248
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.39777000193728757,
+      "learning_rate": 0.00019986831401684735,
+      "loss": 1.319,
+      "step": 249
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.41188612238123284,
+      "learning_rate": 0.00019986523443913697,
+      "loss": 1.1331,
+      "step": 250
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.42695667271764204,
+      "learning_rate": 0.000199862119291555,
+      "loss": 1.405,
+      "step": 251
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4402997476064092,
+      "learning_rate": 0.000199858968575211,
+      "loss": 1.2695,
+      "step": 252
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.42026668090554953,
+      "learning_rate": 0.00019985578229122722,
+      "loss": 1.2844,
+      "step": 253
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.39085096391158103,
+      "learning_rate": 0.00019985256044073848,
+      "loss": 1.2824,
+      "step": 254
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4848127411034025,
+      "learning_rate": 0.0001998493030248924,
+      "loss": 1.3113,
+      "step": 255
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.425201293823512,
+      "learning_rate": 0.00019984601004484915,
+      "loss": 1.271,
+      "step": 256
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4145611751874817,
+      "learning_rate": 0.00019984268150178167,
+      "loss": 1.1976,
+      "step": 257
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.41728413213433224,
+      "learning_rate": 0.00019983931739687546,
+      "loss": 1.4319,
+      "step": 258
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.5010021063149246,
+      "learning_rate": 0.00019983591773132882,
+      "loss": 1.4714,
+      "step": 259
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.40468163939495266,
+      "learning_rate": 0.00019983248250635257,
+      "loss": 1.3303,
+      "step": 260
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4270898088597985,
+      "learning_rate": 0.00019982901172317031,
+      "loss": 1.2806,
+      "step": 261
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4416002842765301,
+      "learning_rate": 0.0001998255053830182,
+      "loss": 1.3395,
+      "step": 262
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.3796859934585211,
+      "learning_rate": 0.00019982196348714517,
+      "loss": 1.1086,
+      "step": 263
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.3778607490210204,
+      "learning_rate": 0.00019981838603681278,
+      "loss": 1.2492,
+      "step": 264
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.48458902205601406,
+      "learning_rate": 0.00019981477303329518,
+      "loss": 1.4338,
+      "step": 265
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.45761016932712884,
+      "learning_rate": 0.00019981112447787933,
+      "loss": 1.2149,
+      "step": 266
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.40188304801357017,
+      "learning_rate": 0.00019980744037186469,
+      "loss": 1.2207,
+      "step": 267
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4609417229488766,
+      "learning_rate": 0.00019980372071656352,
+      "loss": 1.3676,
+      "step": 268
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.43092635924661804,
+      "learning_rate": 0.00019979996551330065,
+      "loss": 1.486,
+      "step": 269
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.3685399492855458,
+      "learning_rate": 0.0001997961747634136,
+      "loss": 1.445,
+      "step": 270
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.39281457728704966,
+      "learning_rate": 0.00019979234846825253,
+      "loss": 1.0444,
+      "step": 271
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.5054774945059699,
+      "learning_rate": 0.00019978848662918037,
+      "loss": 1.3284,
+      "step": 272
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.42576139547756287,
+      "learning_rate": 0.00019978458924757255,
+      "loss": 1.3244,
+      "step": 273
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.3539018454210635,
+      "learning_rate": 0.00019978065632481726,
+      "loss": 1.2749,
+      "step": 274
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.5000837008824371,
+      "learning_rate": 0.00019977668786231534,
+      "loss": 1.1922,
+      "step": 275
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.40530430792517935,
+      "learning_rate": 0.00019977268386148022,
+      "loss": 1.2206,
+      "step": 276
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4037489715006482,
+      "learning_rate": 0.0001997686443237381,
+      "loss": 1.3932,
+      "step": 277
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.411613820535667,
+      "learning_rate": 0.00019976456925052772,
+      "loss": 1.1409,
+      "step": 278
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4713668448446888,
+      "learning_rate": 0.00019976045864330058,
+      "loss": 1.3348,
+      "step": 279
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.5405049089476842,
+      "learning_rate": 0.00019975631250352077,
+      "loss": 1.2207,
+      "step": 280
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.47127293866342873,
+      "learning_rate": 0.00019975213083266508,
+      "loss": 1.1647,
+      "step": 281
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4576617019966713,
+      "learning_rate": 0.0001997479136322229,
+      "loss": 1.282,
+      "step": 282
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.43271526775406427,
+      "learning_rate": 0.0001997436609036963,
+      "loss": 1.4266,
+      "step": 283
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4754879494759945,
+      "learning_rate": 0.00019973937264860004,
+      "loss": 1.1739,
+      "step": 284
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.3910440607112026,
+      "learning_rate": 0.0001997350488684615,
+      "loss": 1.2194,
+      "step": 285
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4486653000701272,
+      "learning_rate": 0.00019973068956482073,
+      "loss": 1.3289,
+      "step": 286
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.39059485273186034,
+      "learning_rate": 0.00019972629473923042,
+      "loss": 1.4724,
+      "step": 287
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.437099288178825,
+      "learning_rate": 0.00019972186439325587,
+      "loss": 1.3718,
+      "step": 288
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4905573216637573,
+      "learning_rate": 0.00019971739852847514,
+      "loss": 1.3356,
+      "step": 289
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4129213811358112,
+      "learning_rate": 0.0001997128971464788,
+      "loss": 1.1718,
+      "step": 290
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.38591588973462154,
+      "learning_rate": 0.0001997083602488702,
+      "loss": 1.2725,
+      "step": 291
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.40711835136613,
+      "learning_rate": 0.00019970378783726528,
+      "loss": 1.4138,
+      "step": 292
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.39244117794110367,
+      "learning_rate": 0.00019969917991329267,
+      "loss": 1.411,
+      "step": 293
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.5996723895281991,
+      "learning_rate": 0.0001996945364785935,
+      "loss": 1.4607,
+      "step": 294
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4310519927301072,
+      "learning_rate": 0.0001996898575348218,
+      "loss": 1.1772,
+      "step": 295
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4760328520249703,
+      "learning_rate": 0.00019968514308364398,
+      "loss": 1.1061,
+      "step": 296
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.45200258990675685,
+      "learning_rate": 0.00019968039312673935,
+      "loss": 1.3319,
+      "step": 297
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.48847892196061715,
+      "learning_rate": 0.0001996756076657997,
+      "loss": 1.3186,
+      "step": 298
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.4478934642938914,
+      "learning_rate": 0.00019967078670252945,
+      "loss": 1.2525,
+      "step": 299
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.3878264670612438,
+      "learning_rate": 0.0001996659302386458,
+      "loss": 1.3672,
+      "step": 300
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.42352043573256887,
+      "learning_rate": 0.00019966103827587846,
+      "loss": 1.1523,
+      "step": 301
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.3987245341641939,
+      "learning_rate": 0.00019965611081596994,
+      "loss": 1.299,
+      "step": 302
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.4057616971474283,
+      "learning_rate": 0.00019965114786067516,
+      "loss": 1.394,
+      "step": 303
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.39680483952055867,
+      "learning_rate": 0.00019964614941176195,
+      "loss": 1.3941,
+      "step": 304
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.3864398555591148,
+      "learning_rate": 0.00019964111547101054,
+      "loss": 1.2488,
+      "step": 305
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.38016312094607124,
+      "learning_rate": 0.00019963604604021398,
+      "loss": 1.3414,
+      "step": 306
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.41151178853467224,
+      "learning_rate": 0.00019963094112117785,
+      "loss": 1.3687,
+      "step": 307
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.38340504434827116,
+      "learning_rate": 0.0001996258007157205,
+      "loss": 1.2708,
+      "step": 308
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.4023404400888361,
+      "learning_rate": 0.00019962062482567274,
+      "loss": 1.2975,
+      "step": 309
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.37390720399138083,
+      "learning_rate": 0.00019961541345287815,
+      "loss": 1.3882,
+      "step": 310
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.459300499222985,
+      "learning_rate": 0.0001996101665991929,
+      "loss": 1.1879,
+      "step": 311
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.35759624094666215,
+      "learning_rate": 0.0001996048842664858,
+      "loss": 1.2554,
+      "step": 312
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.4262219435553382,
+      "learning_rate": 0.00019959956645663836,
+      "loss": 1.3034,
+      "step": 313
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.41914057627769075,
+      "learning_rate": 0.0001995942131715446,
+      "loss": 1.2825,
+      "step": 314
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.4173880705157873,
+      "learning_rate": 0.00019958882441311126,
+      "loss": 1.4262,
+      "step": 315
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.41185885374211495,
+      "learning_rate": 0.00019958340018325776,
+      "loss": 1.274,
+      "step": 316
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.3769626194810209,
+      "learning_rate": 0.00019957794048391602,
+      "loss": 1.229,
+      "step": 317
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.46660204111447307,
+      "learning_rate": 0.0001995724453170307,
+      "loss": 1.2449,
+      "step": 318
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.45460416493085676,
+      "learning_rate": 0.00019956691468455906,
+      "loss": 1.4253,
+      "step": 319
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.39012320521167565,
+      "learning_rate": 0.000199561348588471,
+      "loss": 1.4216,
+      "step": 320
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.3987054566679674,
+      "learning_rate": 0.00019955574703074905,
+      "loss": 1.2119,
+      "step": 321
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.4181406759749968,
+      "learning_rate": 0.00019955011001338832,
+      "loss": 1.2326,
+      "step": 322
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.42364408247966956,
+      "learning_rate": 0.00019954443753839667,
+      "loss": 1.3577,
+      "step": 323
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.4576008288766158,
+      "learning_rate": 0.00019953872960779445,
+      "loss": 1.1653,
+      "step": 324
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.38762824889991254,
+      "learning_rate": 0.00019953298622361475,
+      "loss": 1.2639,
+      "step": 325
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5130475600941465,
+      "learning_rate": 0.0001995272073879032,
+      "loss": 1.0774,
+      "step": 326
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.3775246549295587,
+      "learning_rate": 0.00019952139310271813,
+      "loss": 1.3901,
+      "step": 327
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.3923237550863354,
+      "learning_rate": 0.00019951554337013048,
+      "loss": 1.223,
+      "step": 328
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.40809557417222864,
+      "learning_rate": 0.00019950965819222374,
+      "loss": 1.4164,
+      "step": 329
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.39947283576763015,
+      "learning_rate": 0.00019950373757109412,
+      "loss": 1.2952,
+      "step": 330
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.3621073513112223,
+      "learning_rate": 0.00019949778150885042,
+      "loss": 1.2448,
+      "step": 331
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.49781848486219293,
+      "learning_rate": 0.00019949179000761406,
+      "loss": 1.342,
+      "step": 332
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.3941717256808793,
+      "learning_rate": 0.0001994857630695191,
+      "loss": 1.1789,
+      "step": 333
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.43312804793746634,
+      "learning_rate": 0.00019947970069671218,
+      "loss": 1.4112,
+      "step": 334
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.44434392769716274,
+      "learning_rate": 0.0001994736028913526,
+      "loss": 1.297,
+      "step": 335
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.4176749075269586,
+      "learning_rate": 0.0001994674696556123,
+      "loss": 1.2345,
+      "step": 336
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.433748255606534,
+      "learning_rate": 0.00019946130099167574,
+      "loss": 1.1572,
+      "step": 337
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.4626519184210942,
+      "learning_rate": 0.00019945509690174012,
+      "loss": 1.4357,
+      "step": 338
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.44636967278799555,
+      "learning_rate": 0.00019944885738801518,
+      "loss": 1.3439,
+      "step": 339
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.40395301198179306,
+      "learning_rate": 0.00019944258245272332,
+      "loss": 1.1183,
+      "step": 340
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.4062128682893621,
+      "learning_rate": 0.00019943627209809952,
+      "loss": 1.245,
+      "step": 341
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5134752590103712,
+      "learning_rate": 0.00019942992632639142,
+      "loss": 1.4064,
+      "step": 342
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.43400219919825034,
+      "learning_rate": 0.00019942354513985923,
+      "loss": 1.2997,
+      "step": 343
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.4143907989541104,
+      "learning_rate": 0.0001994171285407758,
+      "loss": 1.1601,
+      "step": 344
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.4579856215271459,
+      "learning_rate": 0.00019941067653142657,
+      "loss": 1.3283,
+      "step": 345
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.3790895699586321,
+      "learning_rate": 0.00019940418911410963,
+      "loss": 1.3086,
+      "step": 346
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.39984294647520296,
+      "learning_rate": 0.00019939766629113566,
+      "loss": 1.2082,
+      "step": 347
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.4299688533679744,
+      "learning_rate": 0.0001993911080648279,
+      "loss": 1.234,
+      "step": 348
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.39412609149259104,
+      "learning_rate": 0.00019938451443752233,
+      "loss": 1.4289,
+      "step": 349
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5003885628861825,
+      "learning_rate": 0.0001993778854115674,
+      "loss": 1.4617,
+      "step": 350
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.4455425530564577,
+      "learning_rate": 0.00019937122098932428,
+      "loss": 1.2931,
+      "step": 351
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.4505326451082181,
+      "learning_rate": 0.00019936452117316663,
+      "loss": 1.1935,
+      "step": 352
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.4427668444897642,
+      "learning_rate": 0.00019935778596548085,
+      "loss": 1.2742,
+      "step": 353
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.4687793414722574,
+      "learning_rate": 0.00019935101536866582,
+      "loss": 1.2746,
+      "step": 354
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.39470619376180743,
+      "learning_rate": 0.00019934420938513313,
+      "loss": 1.2822,
+      "step": 355
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5179011784171917,
+      "learning_rate": 0.00019933736801730687,
+      "loss": 1.3582,
+      "step": 356
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.47769547655512185,
+      "learning_rate": 0.00019933049126762383,
+      "loss": 1.1582,
+      "step": 357
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.4083304518058515,
+      "learning_rate": 0.00019932357913853336,
+      "loss": 1.2177,
+      "step": 358
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.41382195417742385,
+      "learning_rate": 0.00019931663163249742,
+      "loss": 1.3119,
+      "step": 359
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.48328070407952906,
+      "learning_rate": 0.00019930964875199055,
+      "loss": 1.2981,
+      "step": 360
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.4439969849748576,
+      "learning_rate": 0.0001993026304994999,
+      "loss": 1.1273,
+      "step": 361
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5482357283750525,
+      "learning_rate": 0.00019929557687752524,
+      "loss": 1.1688,
+      "step": 362
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.45531684213541895,
+      "learning_rate": 0.00019928848788857887,
+      "loss": 1.3176,
+      "step": 363
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5194642940389806,
+      "learning_rate": 0.00019928136353518583,
+      "loss": 1.3189,
+      "step": 364
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.44494068947722815,
+      "learning_rate": 0.00019927420381988357,
+      "loss": 1.2025,
+      "step": 365
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.419390599948951,
+      "learning_rate": 0.00019926700874522228,
+      "loss": 1.4973,
+      "step": 366
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.48596064758596624,
+      "learning_rate": 0.00019925977831376467,
+      "loss": 1.114,
+      "step": 367
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.4288276752724884,
+      "learning_rate": 0.0001992525125280861,
+      "loss": 1.1437,
+      "step": 368
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6013989972968181,
+      "learning_rate": 0.00019924521139077443,
+      "loss": 1.0607,
+      "step": 369
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.44735410804297593,
+      "learning_rate": 0.0001992378749044302,
+      "loss": 1.2261,
+      "step": 370
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.41286615187810166,
+      "learning_rate": 0.00019923050307166655,
+      "loss": 1.2723,
+      "step": 371
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.4595969340601655,
+      "learning_rate": 0.00019922309589510913,
+      "loss": 1.3163,
+      "step": 372
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.42145488675436654,
+      "learning_rate": 0.0001992156533773962,
+      "loss": 1.3855,
+      "step": 373
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.4202109913928829,
+      "learning_rate": 0.00019920817552117863,
+      "loss": 1.23,
+      "step": 374
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.4395913897340459,
+      "learning_rate": 0.00019920066232911992,
+      "loss": 1.1588,
+      "step": 375
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.41303209412140374,
+      "learning_rate": 0.0001991931138038961,
+      "loss": 1.3053,
+      "step": 376
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.384050734615511,
+      "learning_rate": 0.00019918552994819573,
+      "loss": 1.1812,
+      "step": 377
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5423772830865649,
+      "learning_rate": 0.00019917791076472008,
+      "loss": 1.1722,
+      "step": 378
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.4130953386148949,
+      "learning_rate": 0.00019917025625618292,
+      "loss": 1.211,
+      "step": 379
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5403943154080612,
+      "learning_rate": 0.00019916256642531064,
+      "loss": 1.3432,
+      "step": 380
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.38978081833567074,
+      "learning_rate": 0.00019915484127484216,
+      "loss": 1.3196,
+      "step": 381
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.45964389975590847,
+      "learning_rate": 0.00019914708080752907,
+      "loss": 1.3186,
+      "step": 382
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.4697838869040222,
+      "learning_rate": 0.0001991392850261354,
+      "loss": 1.1767,
+      "step": 383
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.46670523731360375,
+      "learning_rate": 0.00019913145393343795,
+      "loss": 1.321,
+      "step": 384
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.4597247824481378,
+      "learning_rate": 0.00019912358753222588,
+      "loss": 1.1992,
+      "step": 385
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.4156863121587114,
+      "learning_rate": 0.00019911568582530113,
+      "loss": 1.2999,
+      "step": 386
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.4237436359775387,
+      "learning_rate": 0.000199107748815478,
+      "loss": 1.246,
+      "step": 387
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.4673090350718922,
+      "learning_rate": 0.00019909977650558363,
+      "loss": 1.2546,
+      "step": 388
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.4560429691610961,
+      "learning_rate": 0.0001990917688984575,
+      "loss": 1.1377,
+      "step": 389
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.41401360382995794,
+      "learning_rate": 0.0001990837259969517,
+      "loss": 1.1978,
+      "step": 390
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.41315405311991776,
+      "learning_rate": 0.00019907564780393108,
+      "loss": 1.3222,
+      "step": 391
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.43596377100816625,
+      "learning_rate": 0.0001990675343222728,
+      "loss": 1.0681,
+      "step": 392
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.46930727566306735,
+      "learning_rate": 0.00019905938555486675,
+      "loss": 1.1909,
+      "step": 393
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.4560027153411607,
+      "learning_rate": 0.0001990512015046154,
+      "loss": 1.2308,
+      "step": 394
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.4174181911516993,
+      "learning_rate": 0.00019904298217443366,
+      "loss": 1.3159,
+      "step": 395
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.45805223661555666,
+      "learning_rate": 0.0001990347275672491,
+      "loss": 1.1805,
+      "step": 396
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5135589841783814,
+      "learning_rate": 0.00019902643768600185,
+      "loss": 1.3458,
+      "step": 397
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.38802306407065235,
+      "learning_rate": 0.00019901811253364456,
+      "loss": 1.2656,
+      "step": 398
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.3931441014934635,
+      "learning_rate": 0.0001990097521131425,
+      "loss": 1.3905,
+      "step": 399
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5537126945729326,
+      "learning_rate": 0.00019900135642747347,
+      "loss": 1.275,
+      "step": 400
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.4198378794464167,
+      "learning_rate": 0.00019899292547962788,
+      "loss": 1.1914,
+      "step": 401
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.4738145923388428,
+      "learning_rate": 0.00019898445927260853,
+      "loss": 1.1948,
+      "step": 402
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.46468597182218924,
+      "learning_rate": 0.00019897595780943102,
+      "loss": 1.3085,
+      "step": 403
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.4431252042938539,
+      "learning_rate": 0.00019896742109312335,
+      "loss": 1.4599,
+      "step": 404
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.43511198324729067,
+      "learning_rate": 0.0001989588491267261,
+      "loss": 1.2398,
+      "step": 405
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.4533082522938461,
+      "learning_rate": 0.00019895024191329247,
+      "loss": 1.059,
+      "step": 406
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.41727705984813174,
+      "learning_rate": 0.00019894159945588812,
+      "loss": 1.3021,
+      "step": 407
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.42200045330505925,
+      "learning_rate": 0.00019893292175759131,
+      "loss": 1.2471,
+      "step": 408
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.512967597808161,
+      "learning_rate": 0.00019892420882149287,
+      "loss": 1.1519,
+      "step": 409
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.39107426006974644,
+      "learning_rate": 0.00019891546065069614,
+      "loss": 1.3613,
+      "step": 410
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.4159066749295831,
+      "learning_rate": 0.00019890667724831707,
+      "loss": 1.1447,
+      "step": 411
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.4097089611430613,
+      "learning_rate": 0.00019889785861748412,
+      "loss": 1.1963,
+      "step": 412
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.4469394344548388,
+      "learning_rate": 0.00019888900476133824,
+      "loss": 1.3208,
+      "step": 413
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.3878552418884981,
+      "learning_rate": 0.00019888011568303305,
+      "loss": 1.1388,
+      "step": 414
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.43618705086286097,
+      "learning_rate": 0.0001988711913857346,
+      "loss": 1.2625,
+      "step": 415
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.4286425081300646,
+      "learning_rate": 0.00019886223187262162,
+      "loss": 1.3675,
+      "step": 416
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.4095546089049234,
+      "learning_rate": 0.00019885323714688517,
+      "loss": 1.3694,
+      "step": 417
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.4579870133661384,
+      "learning_rate": 0.0001988442072117291,
+      "loss": 1.2024,
+      "step": 418
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.47977451974290247,
+      "learning_rate": 0.00019883514207036956,
+      "loss": 1.3851,
+      "step": 419
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.40703894353387565,
+      "learning_rate": 0.0001988260417260355,
+      "loss": 1.3355,
+      "step": 420
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.4254897500588801,
+      "learning_rate": 0.00019881690618196816,
+      "loss": 1.2873,
+      "step": 421
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.39816967316263235,
+      "learning_rate": 0.00019880773544142148,
+      "loss": 1.3391,
+      "step": 422
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.36951827572852375,
+      "learning_rate": 0.00019879852950766187,
+      "loss": 1.4625,
+      "step": 423
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.45238753712682817,
+      "learning_rate": 0.0001987892883839683,
+      "loss": 1.3148,
+      "step": 424
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.45111481019324096,
+      "learning_rate": 0.0001987800120736322,
+      "loss": 1.2221,
+      "step": 425
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.38698913242552085,
+      "learning_rate": 0.00019877070057995768,
+      "loss": 1.1148,
+      "step": 426
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.43386120017545104,
+      "learning_rate": 0.00019876135390626122,
+      "loss": 1.3677,
+      "step": 427
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.46475770864164795,
+      "learning_rate": 0.00019875197205587196,
+      "loss": 1.1911,
+      "step": 428
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5217116489564152,
+      "learning_rate": 0.00019874255503213152,
+      "loss": 1.055,
+      "step": 429
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.39617255459125394,
+      "learning_rate": 0.00019873310283839398,
+      "loss": 1.3404,
+      "step": 430
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.4198237899007023,
+      "learning_rate": 0.0001987236154780261,
+      "loss": 1.3627,
+      "step": 431
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.3640818422891405,
+      "learning_rate": 0.000198714092954407,
+      "loss": 1.2703,
+      "step": 432
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.42526246521972527,
+      "learning_rate": 0.00019870453527092845,
+      "loss": 1.3643,
+      "step": 433
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.3836573845698337,
+      "learning_rate": 0.00019869494243099465,
+      "loss": 1.2764,
+      "step": 434
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.39701918427397137,
+      "learning_rate": 0.0001986853144380224,
+      "loss": 1.3797,
+      "step": 435
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.4320717201024364,
+      "learning_rate": 0.00019867565129544096,
+      "loss": 1.1091,
+      "step": 436
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.3777132401765734,
+      "learning_rate": 0.00019866595300669217,
+      "loss": 1.3244,
+      "step": 437
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.40520664155066977,
+      "learning_rate": 0.00019865621957523033,
+      "loss": 1.1648,
+      "step": 438
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.4026206266882511,
+      "learning_rate": 0.00019864645100452227,
+      "loss": 1.3075,
+      "step": 439
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.40998168000842794,
+      "learning_rate": 0.00019863664729804738,
+      "loss": 1.0857,
+      "step": 440
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.3985320802137327,
+      "learning_rate": 0.00019862680845929748,
+      "loss": 1.2886,
+      "step": 441
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.4076609608769866,
+      "learning_rate": 0.00019861693449177702,
+      "loss": 1.4206,
+      "step": 442
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.43744547765719716,
+      "learning_rate": 0.00019860702539900287,
+      "loss": 1.1017,
+      "step": 443
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.45724009583247033,
+      "learning_rate": 0.00019859708118450441,
+      "loss": 1.27,
+      "step": 444
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.3788293395205037,
+      "learning_rate": 0.0001985871018518236,
+      "loss": 1.4226,
+      "step": 445
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.45550199702741473,
+      "learning_rate": 0.00019857708740451483,
+      "loss": 1.5713,
+      "step": 446
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.4652757890123194,
+      "learning_rate": 0.00019856703784614505,
+      "loss": 1.2274,
+      "step": 447
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.44836846004708625,
+      "learning_rate": 0.00019855695318029372,
+      "loss": 1.3177,
+      "step": 448
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.408863823703532,
+      "learning_rate": 0.00019854683341055275,
+      "loss": 1.1963,
+      "step": 449
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.41681343255112235,
+      "learning_rate": 0.00019853667854052663,
+      "loss": 1.3254,
+      "step": 450
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.42409531744580015,
+      "learning_rate": 0.00019852648857383222,
+      "loss": 1.2525,
+      "step": 451
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.4346323455641248,
+      "learning_rate": 0.00019851626351409906,
+      "loss": 1.0736,
+      "step": 452
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.46086761153755956,
+      "learning_rate": 0.00019850600336496904,
+      "loss": 1.3518,
+      "step": 453
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.37828990642136034,
+      "learning_rate": 0.0001984957081300967,
+      "loss": 1.2081,
+      "step": 454
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5442039497138418,
+      "learning_rate": 0.00019848537781314883,
+      "loss": 1.3255,
+      "step": 455
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.3709112734464793,
+      "learning_rate": 0.00019847501241780502,
+      "loss": 1.211,
+      "step": 456
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.3930900235837572,
+      "learning_rate": 0.00019846461194775712,
+      "loss": 1.2339,
+      "step": 457
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.4660476169524567,
+      "learning_rate": 0.00019845417640670956,
+      "loss": 1.2136,
+      "step": 458
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.44935367776035695,
+      "learning_rate": 0.00019844370579837927,
+      "loss": 1.0977,
+      "step": 459
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.43021181987469026,
+      "learning_rate": 0.0001984332001264957,
+      "loss": 1.4831,
+      "step": 460
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5525628236683462,
+      "learning_rate": 0.00019842265939480064,
+      "loss": 1.3803,
+      "step": 461
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.4364298975560644,
+      "learning_rate": 0.00019841208360704858,
+      "loss": 1.3363,
+      "step": 462
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.45574051388048553,
+      "learning_rate": 0.00019840147276700636,
+      "loss": 1.329,
+      "step": 463
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.428945528967339,
+      "learning_rate": 0.00019839082687845335,
+      "loss": 1.4156,
+      "step": 464
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.595294663530642,
+      "learning_rate": 0.00019838014594518132,
+      "loss": 1.1539,
+      "step": 465
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.4292809143281606,
+      "learning_rate": 0.00019836942997099465,
+      "loss": 1.2247,
+      "step": 466
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.4853949829975137,
+      "learning_rate": 0.00019835867895971014,
+      "loss": 1.3192,
+      "step": 467
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.4012278452327183,
+      "learning_rate": 0.00019834789291515705,
+      "loss": 1.2638,
+      "step": 468
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.4017569976996731,
+      "learning_rate": 0.00019833707184117717,
+      "loss": 1.4905,
+      "step": 469
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.44100779511707683,
+      "learning_rate": 0.00019832621574162468,
+      "loss": 1.1975,
+      "step": 470
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.4132694792325637,
+      "learning_rate": 0.00019831532462036636,
+      "loss": 1.2333,
+      "step": 471
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.44010243387333164,
+      "learning_rate": 0.00019830439848128134,
+      "loss": 1.1324,
+      "step": 472
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.4057061711206227,
+      "learning_rate": 0.0001982934373282613,
+      "loss": 1.2249,
+      "step": 473
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.40358922278665255,
+      "learning_rate": 0.0001982824411652104,
+      "loss": 1.3448,
+      "step": 474
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.407152728831155,
+      "learning_rate": 0.0001982714099960452,
+      "loss": 1.2343,
+      "step": 475
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.43192617430112745,
+      "learning_rate": 0.00019826034382469478,
+      "loss": 1.3451,
+      "step": 476
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.4232689059708419,
+      "learning_rate": 0.00019824924265510066,
+      "loss": 1.2113,
+      "step": 477
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.4450278076214006,
+      "learning_rate": 0.00019823810649121688,
+      "loss": 1.3193,
+      "step": 478
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.4417347561971035,
+      "learning_rate": 0.00019822693533700988,
+      "loss": 1.3812,
+      "step": 479
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.41226313310784923,
+      "learning_rate": 0.0001982157291964586,
+      "loss": 1.2289,
+      "step": 480
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.3970392029188392,
+      "learning_rate": 0.00019820448807355437,
+      "loss": 1.188,
+      "step": 481
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.4342352507240826,
+      "learning_rate": 0.00019819321197230109,
+      "loss": 1.2764,
+      "step": 482
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.36891955474487376,
+      "learning_rate": 0.00019818190089671508,
+      "loss": 1.078,
+      "step": 483
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.42294301054402145,
+      "learning_rate": 0.00019817055485082507,
+      "loss": 1.3919,
+      "step": 484
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.4556960975014498,
+      "learning_rate": 0.0001981591738386723,
+      "loss": 1.1733,
+      "step": 485
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.46731819944636077,
+      "learning_rate": 0.00019814775786431045,
+      "loss": 1.157,
+      "step": 486
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.4446621323046658,
+      "learning_rate": 0.00019813630693180562,
+      "loss": 1.2411,
+      "step": 487
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.4252558693705636,
+      "learning_rate": 0.00019812482104523645,
+      "loss": 1.2485,
+      "step": 488
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.4283359205282634,
+      "learning_rate": 0.00019811330020869385,
+      "loss": 1.2658,
+      "step": 489
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.4246356149271799,
+      "learning_rate": 0.00019810174442628142,
+      "loss": 1.1619,
+      "step": 490
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.4039437029985089,
+      "learning_rate": 0.00019809015370211502,
+      "loss": 1.3345,
+      "step": 491
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.38581378984827575,
+      "learning_rate": 0.00019807852804032305,
+      "loss": 1.3997,
+      "step": 492
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6034015465244927,
+      "learning_rate": 0.0001980668674450463,
+      "loss": 1.3132,
+      "step": 493
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.44598723664930945,
+      "learning_rate": 0.00019805517192043802,
+      "loss": 1.2618,
+      "step": 494
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.39981310182589536,
+      "learning_rate": 0.00019804344147066397,
+      "loss": 1.2182,
+      "step": 495
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.3732293643538034,
+      "learning_rate": 0.00019803167609990218,
+      "loss": 1.2908,
+      "step": 496
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.44736327537542614,
+      "learning_rate": 0.0001980198758123433,
+      "loss": 1.3845,
+      "step": 497
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.4442659901705393,
+      "learning_rate": 0.00019800804061219032,
+      "loss": 1.3914,
+      "step": 498
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.4006847142485584,
+      "learning_rate": 0.0001979961705036587,
+      "loss": 1.2415,
+      "step": 499
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.387065742815003,
+      "learning_rate": 0.0001979842654909763,
+      "loss": 1.3684,
+      "step": 500
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.4403865126395757,
+      "learning_rate": 0.00019797232557838343,
+      "loss": 1.2652,
+      "step": 501
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.43716065997066184,
+      "learning_rate": 0.00019796035077013287,
+      "loss": 1.4529,
+      "step": 502
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.4361514694588171,
+      "learning_rate": 0.00019794834107048973,
+      "loss": 1.2235,
+      "step": 503
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.4834669332407145,
+      "learning_rate": 0.0001979362964837317,
+      "loss": 1.1984,
+      "step": 504
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.4588724679467598,
+      "learning_rate": 0.00019792421701414871,
+      "loss": 1.3241,
+      "step": 505
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.49376831685928724,
+      "learning_rate": 0.00019791210266604327,
+      "loss": 1.3246,
+      "step": 506
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.45530870508010546,
+      "learning_rate": 0.00019789995344373024,
+      "loss": 1.3142,
+      "step": 507
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.3873133142035057,
+      "learning_rate": 0.0001978877693515369,
+      "loss": 1.2879,
+      "step": 508
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.4080385603444197,
+      "learning_rate": 0.00019787555039380296,
+      "loss": 1.4696,
+      "step": 509
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.468101488452057,
+      "learning_rate": 0.0001978632965748806,
+      "loss": 1.1737,
+      "step": 510
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.45191371101491323,
+      "learning_rate": 0.00019785100789913428,
+      "loss": 1.2095,
+      "step": 511
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.4263004767510613,
+      "learning_rate": 0.0001978386843709411,
+      "loss": 1.282,
+      "step": 512
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.41594163344130686,
+      "learning_rate": 0.0001978263259946903,
+      "loss": 1.2058,
+      "step": 513
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.3785853113934449,
+      "learning_rate": 0.00019781393277478373,
+      "loss": 1.3849,
+      "step": 514
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.3986767540724619,
+      "learning_rate": 0.00019780150471563558,
+      "loss": 1.2989,
+      "step": 515
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6853083502078515,
+      "learning_rate": 0.00019778904182167254,
+      "loss": 1.2091,
+      "step": 516
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.401334392806089,
+      "learning_rate": 0.0001977765440973335,
+      "loss": 1.2102,
+      "step": 517
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.41139433993491736,
+      "learning_rate": 0.00019776401154706993,
+      "loss": 1.1556,
+      "step": 518
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.4021943790410962,
+      "learning_rate": 0.00019775144417534573,
+      "loss": 1.343,
+      "step": 519
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.44955220218728303,
+      "learning_rate": 0.00019773884198663702,
+      "loss": 1.1802,
+      "step": 520
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.3976155453714967,
+      "learning_rate": 0.00019772620498543247,
+      "loss": 1.2393,
+      "step": 521
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.38559032462713605,
+      "learning_rate": 0.00019771353317623317,
+      "loss": 1.1399,
+      "step": 522
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.4590466204506952,
+      "learning_rate": 0.0001977008265635525,
+      "loss": 1.4512,
+      "step": 523
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.43633786686368775,
+      "learning_rate": 0.0001976880851519163,
+      "loss": 1.2022,
+      "step": 524
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.44145962192482796,
+      "learning_rate": 0.00019767530894586273,
+      "loss": 1.2423,
+      "step": 525
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5401210380198437,
+      "learning_rate": 0.00019766249794994246,
+      "loss": 1.1822,
+      "step": 526
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.4062265987313378,
+      "learning_rate": 0.00019764965216871846,
+      "loss": 1.2925,
+      "step": 527
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.4014370209702807,
+      "learning_rate": 0.0001976367716067662,
+      "loss": 1.3822,
+      "step": 528
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.414085269304349,
+      "learning_rate": 0.00019762385626867336,
+      "loss": 1.106,
+      "step": 529
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.39610982257461713,
+      "learning_rate": 0.00019761090615904022,
+      "loss": 1.1897,
+      "step": 530
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.45246277305074317,
+      "learning_rate": 0.00019759792128247922,
+      "loss": 1.2399,
+      "step": 531
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.41600976527431377,
+      "learning_rate": 0.00019758490164361537,
+      "loss": 1.2397,
+      "step": 532
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.42013061298557997,
+      "learning_rate": 0.00019757184724708595,
+      "loss": 1.2144,
+      "step": 533
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.4953935830046393,
+      "learning_rate": 0.00019755875809754068,
+      "loss": 1.3752,
+      "step": 534
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.37477099789345214,
+      "learning_rate": 0.00019754563419964165,
+      "loss": 1.2747,
+      "step": 535
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.4101745551015663,
+      "learning_rate": 0.0001975324755580633,
+      "loss": 1.0639,
+      "step": 536
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.40036828511174133,
+      "learning_rate": 0.0001975192821774924,
+      "loss": 1.0083,
+      "step": 537
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.4456629709203121,
+      "learning_rate": 0.00019750605406262825,
+      "loss": 1.2834,
+      "step": 538
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.39477917983804195,
+      "learning_rate": 0.00019749279121818235,
+      "loss": 1.3417,
+      "step": 539
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.432530737465089,
+      "learning_rate": 0.00019747949364887872,
+      "loss": 1.2931,
+      "step": 540
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.4763023554240281,
+      "learning_rate": 0.00019746616135945356,
+      "loss": 1.1775,
+      "step": 541
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.4157205531250029,
+      "learning_rate": 0.00019745279435465565,
+      "loss": 1.24,
+      "step": 542
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5364348891656562,
+      "learning_rate": 0.00019743939263924598,
+      "loss": 1.2354,
+      "step": 543
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.4197437747107111,
+      "learning_rate": 0.00019742595621799795,
+      "loss": 1.3026,
+      "step": 544
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.48794995659470985,
+      "learning_rate": 0.00019741248509569736,
+      "loss": 1.2102,
+      "step": 545
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.4291118768792971,
+      "learning_rate": 0.00019739897927714234,
+      "loss": 1.1396,
+      "step": 546
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.4031787067142048,
+      "learning_rate": 0.00019738543876714334,
+      "loss": 1.157,
+      "step": 547
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.4299895943444582,
+      "learning_rate": 0.00019737186357052323,
+      "loss": 1.2745,
+      "step": 548
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.36927659532093915,
+      "learning_rate": 0.0001973582536921172,
+      "loss": 1.331,
+      "step": 549
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.4020811962373291,
+      "learning_rate": 0.00019734460913677277,
+      "loss": 1.1823,
+      "step": 550
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.41115706534723706,
+      "learning_rate": 0.00019733092990934984,
+      "loss": 1.3586,
+      "step": 551
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.3886008799069163,
+      "learning_rate": 0.00019731721601472075,
+      "loss": 1.3249,
+      "step": 552
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.4004999024266242,
+      "learning_rate": 0.00019730346745776997,
+      "loss": 1.1551,
+      "step": 553
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.36930911802588035,
+      "learning_rate": 0.00019728968424339454,
+      "loss": 1.3861,
+      "step": 554
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.49880917997935026,
+      "learning_rate": 0.00019727586637650373,
+      "loss": 1.1828,
+      "step": 555
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.41853112830272793,
+      "learning_rate": 0.0001972620138620191,
+      "loss": 1.1337,
+      "step": 556
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.4495697361070426,
+      "learning_rate": 0.00019724812670487475,
+      "loss": 1.2172,
+      "step": 557
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.41669834167750663,
+      "learning_rate": 0.00019723420491001684,
+      "loss": 1.2796,
+      "step": 558
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.4758455178711545,
+      "learning_rate": 0.00019722024848240413,
+      "loss": 1.3331,
+      "step": 559
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.4537475211418872,
+      "learning_rate": 0.00019720625742700756,
+      "loss": 1.405,
+      "step": 560
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.44116159516896764,
+      "learning_rate": 0.00019719223174881044,
+      "loss": 1.2866,
+      "step": 561
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.42379618903499483,
+      "learning_rate": 0.00019717817145280844,
+      "loss": 1.122,
+      "step": 562
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.3986194563206064,
+      "learning_rate": 0.00019716407654400952,
+      "loss": 1.1975,
+      "step": 563
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.37529180637164167,
+      "learning_rate": 0.00019714994702743396,
+      "loss": 1.3033,
+      "step": 564
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.4092278976928945,
+      "learning_rate": 0.00019713578290811448,
+      "loss": 1.3378,
+      "step": 565
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.40049758410055925,
+      "learning_rate": 0.00019712158419109597,
+      "loss": 1.4041,
+      "step": 566
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.4814350698964384,
+      "learning_rate": 0.0001971073508814357,
+      "loss": 1.0762,
+      "step": 567
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.4042822896759389,
+      "learning_rate": 0.0001970930829842033,
+      "loss": 1.2742,
+      "step": 568
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.4143239314430205,
+      "learning_rate": 0.0001970787805044807,
+      "loss": 1.1833,
+      "step": 569
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.44028371815866024,
+      "learning_rate": 0.00019706444344736213,
+      "loss": 1.1795,
+      "step": 570
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.43290160988403736,
+      "learning_rate": 0.00019705007181795416,
+      "loss": 1.454,
+      "step": 571
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4594603354440774,
+      "learning_rate": 0.00019703566562137562,
+      "loss": 1.1398,
+      "step": 572
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4199068369179442,
+      "learning_rate": 0.0001970212248627577,
+      "loss": 1.1902,
+      "step": 573
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4409727602075107,
+      "learning_rate": 0.00019700674954724388,
+      "loss": 1.2337,
+      "step": 574
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5220202033302318,
+      "learning_rate": 0.00019699223967999005,
+      "loss": 1.2645,
+      "step": 575
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6107709297832473,
+      "learning_rate": 0.0001969776952661642,
+      "loss": 1.1279,
+      "step": 576
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.45239981287058906,
+      "learning_rate": 0.0001969631163109468,
+      "loss": 1.4029,
+      "step": 577
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.3956177868576618,
+      "learning_rate": 0.00019694850281953055,
+      "loss": 1.3069,
+      "step": 578
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4414894731465165,
+      "learning_rate": 0.00019693385479712048,
+      "loss": 1.2014,
+      "step": 579
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4057171711915646,
+      "learning_rate": 0.00019691917224893386,
+      "loss": 1.3332,
+      "step": 580
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.43973106333630413,
+      "learning_rate": 0.00019690445518020036,
+      "loss": 1.2108,
+      "step": 581
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4044979234270597,
+      "learning_rate": 0.00019688970359616183,
+      "loss": 1.2702,
+      "step": 582
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.3884987491277362,
+      "learning_rate": 0.00019687491750207254,
+      "loss": 1.1946,
+      "step": 583
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4004915041195403,
+      "learning_rate": 0.0001968600969031989,
+      "loss": 1.213,
+      "step": 584
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.3977350818658314,
+      "learning_rate": 0.00019684524180481976,
+      "loss": 1.3073,
+      "step": 585
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4647490507242339,
+      "learning_rate": 0.00019683035221222618,
+      "loss": 1.1128,
+      "step": 586
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.38819004171422905,
+      "learning_rate": 0.00019681542813072145,
+      "loss": 1.3426,
+      "step": 587
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.41382011236155386,
+      "learning_rate": 0.0001968004695656213,
+      "loss": 1.1013,
+      "step": 588
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.3956714762063599,
+      "learning_rate": 0.00019678547652225356,
+      "loss": 1.3361,
+      "step": 589
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.44482273467724065,
+      "learning_rate": 0.00019677044900595853,
+      "loss": 1.2735,
+      "step": 590
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4865463913198301,
+      "learning_rate": 0.00019675538702208862,
+      "loss": 1.1606,
+      "step": 591
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.3840315155284186,
+      "learning_rate": 0.00019674029057600864,
+      "loss": 1.174,
+      "step": 592
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.42958295922997264,
+      "learning_rate": 0.00019672515967309557,
+      "loss": 1.2049,
+      "step": 593
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.3923329653525399,
+      "learning_rate": 0.00019670999431873874,
+      "loss": 1.2122,
+      "step": 594
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4600190279326033,
+      "learning_rate": 0.00019669479451833976,
+      "loss": 1.2396,
+      "step": 595
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.41315925003626064,
+      "learning_rate": 0.00019667956027731242,
+      "loss": 1.3401,
+      "step": 596
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4859606212802776,
+      "learning_rate": 0.0001966642916010829,
+      "loss": 1.189,
+      "step": 597
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.445113840687173,
+      "learning_rate": 0.0001966489884950895,
+      "loss": 1.1929,
+      "step": 598
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.498630753683034,
+      "learning_rate": 0.0001966336509647829,
+      "loss": 1.3607,
+      "step": 599
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.41228422027139733,
+      "learning_rate": 0.000196618279015626,
+      "loss": 1.3312,
+      "step": 600
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.42713489861262455,
+      "learning_rate": 0.00019660287265309395,
+      "loss": 1.2674,
+      "step": 601
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4289522500909011,
+      "learning_rate": 0.0001965874318826742,
+      "loss": 1.3383,
+      "step": 602
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5355966301590542,
+      "learning_rate": 0.00019657195670986637,
+      "loss": 1.3787,
+      "step": 603
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.41005612090843274,
+      "learning_rate": 0.00019655644714018243,
+      "loss": 1.1454,
+      "step": 604
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5631929208605212,
+      "learning_rate": 0.00019654090317914656,
+      "loss": 1.2507,
+      "step": 605
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.38138555921368184,
+      "learning_rate": 0.00019652532483229514,
+      "loss": 1.3255,
+      "step": 606
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.41060072342897996,
+      "learning_rate": 0.00019650971210517692,
+      "loss": 1.3044,
+      "step": 607
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.42736433530276485,
+      "learning_rate": 0.00019649406500335276,
+      "loss": 1.0321,
+      "step": 608
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.38193636882831716,
+      "learning_rate": 0.00019647838353239588,
+      "loss": 1.1259,
+      "step": 609
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.44475860902168585,
+      "learning_rate": 0.0001964626676978916,
+      "loss": 1.1198,
+      "step": 610
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4559359508244041,
+      "learning_rate": 0.00019644691750543767,
+      "loss": 1.2189,
+      "step": 611
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.49344062508400055,
+      "learning_rate": 0.0001964311329606439,
+      "loss": 1.0312,
+      "step": 612
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4048489832548174,
+      "learning_rate": 0.00019641531406913243,
+      "loss": 1.3434,
+      "step": 613
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.405617048339034,
+      "learning_rate": 0.00019639946083653762,
+      "loss": 1.2709,
+      "step": 614
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.43618643213448405,
+      "learning_rate": 0.00019638357326850602,
+      "loss": 1.1223,
+      "step": 615
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4104449512961255,
+      "learning_rate": 0.00019636765137069653,
+      "loss": 1.2468,
+      "step": 616
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.393917534122103,
+      "learning_rate": 0.00019635169514878008,
+      "loss": 1.275,
+      "step": 617
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4768270105409465,
+      "learning_rate": 0.00019633570460844002,
+      "loss": 1.2059,
+      "step": 618
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4288263133837961,
+      "learning_rate": 0.0001963196797553718,
+      "loss": 1.1319,
+      "step": 619
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4382776700574148,
+      "learning_rate": 0.00019630362059528316,
+      "loss": 1.1589,
+      "step": 620
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.42950411804539856,
+      "learning_rate": 0.00019628752713389398,
+      "loss": 1.2158,
+      "step": 621
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4406644536361421,
+      "learning_rate": 0.00019627139937693645,
+      "loss": 1.3412,
+      "step": 622
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.39798182741853494,
+      "learning_rate": 0.00019625523733015493,
+      "loss": 1.3375,
+      "step": 623
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.3985426704282314,
+      "learning_rate": 0.00019623904099930602,
+      "loss": 1.295,
+      "step": 624
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.47759374639612484,
+      "learning_rate": 0.00019622281039015843,
+      "loss": 1.0061,
+      "step": 625
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.45504857280785727,
+      "learning_rate": 0.00019620654550849323,
+      "loss": 1.2343,
+      "step": 626
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.38844665518507376,
+      "learning_rate": 0.00019619024636010363,
+      "loss": 1.1998,
+      "step": 627
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.39576893974291794,
+      "learning_rate": 0.000196173912950795,
+      "loss": 1.126,
+      "step": 628
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.41231019643686073,
+      "learning_rate": 0.00019615754528638495,
+      "loss": 1.2523,
+      "step": 629
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5085813630594518,
+      "learning_rate": 0.0001961411433727033,
+      "loss": 1.2404,
+      "step": 630
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.39332059770847233,
+      "learning_rate": 0.00019612470721559207,
+      "loss": 1.2149,
+      "step": 631
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.43661521716486973,
+      "learning_rate": 0.0001961082368209055,
+      "loss": 1.0348,
+      "step": 632
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.44662056261087074,
+      "learning_rate": 0.00019609173219450998,
+      "loss": 1.1176,
+      "step": 633
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.40720564458130787,
+      "learning_rate": 0.00019607519334228406,
+      "loss": 1.1344,
+      "step": 634
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.4430355277956355,
+      "learning_rate": 0.00019605862027011856,
+      "loss": 1.2837,
+      "step": 635
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.4364225596386163,
+      "learning_rate": 0.00019604201298391646,
+      "loss": 1.2013,
+      "step": 636
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5058697248027114,
+      "learning_rate": 0.00019602537148959292,
+      "loss": 1.0656,
+      "step": 637
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.49019550100878967,
+      "learning_rate": 0.0001960086957930753,
+      "loss": 1.1273,
+      "step": 638
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.39080080767480724,
+      "learning_rate": 0.0001959919859003031,
+      "loss": 1.2758,
+      "step": 639
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.42739692885574104,
+      "learning_rate": 0.00019597524181722803,
+      "loss": 1.3143,
+      "step": 640
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.4434190794293887,
+      "learning_rate": 0.000195958463549814,
+      "loss": 1.2326,
+      "step": 641
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.4664915530439378,
+      "learning_rate": 0.00019594165110403707,
+      "loss": 1.2351,
+      "step": 642
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.4051244645895513,
+      "learning_rate": 0.00019592480448588542,
+      "loss": 1.2106,
+      "step": 643
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.3771983311186089,
+      "learning_rate": 0.00019590792370135953,
+      "loss": 1.4315,
+      "step": 644
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.3696733822618877,
+      "learning_rate": 0.00019589100875647196,
+      "loss": 1.2691,
+      "step": 645
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.4123034691012944,
+      "learning_rate": 0.0001958740596572474,
+      "loss": 1.3845,
+      "step": 646
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.40132207081367455,
+      "learning_rate": 0.0001958570764097228,
+      "loss": 1.1765,
+      "step": 647
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.42712230078748786,
+      "learning_rate": 0.0001958400590199473,
+      "loss": 1.2634,
+      "step": 648
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.42491877151399515,
+      "learning_rate": 0.000195823007493982,
+      "loss": 1.3093,
+      "step": 649
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.4024086211596599,
+      "learning_rate": 0.00019580592183790035,
+      "loss": 1.257,
+      "step": 650
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.4175341195255031,
+      "learning_rate": 0.00019578880205778793,
+      "loss": 1.2994,
+      "step": 651
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.4556228980240471,
+      "learning_rate": 0.0001957716481597424,
+      "loss": 1.2614,
+      "step": 652
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.41016514383080904,
+      "learning_rate": 0.00019575446014987363,
+      "loss": 1.0889,
+      "step": 653
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.4327329371445978,
+      "learning_rate": 0.0001957372380343036,
+      "loss": 1.1974,
+      "step": 654
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.4433634119245905,
+      "learning_rate": 0.0001957199818191665,
+      "loss": 1.1529,
+      "step": 655
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.41023110231749255,
+      "learning_rate": 0.0001957026915106086,
+      "loss": 1.3612,
+      "step": 656
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.3719053351797238,
+      "learning_rate": 0.00019568536711478835,
+      "loss": 1.2117,
+      "step": 657
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.39486345777618886,
+      "learning_rate": 0.00019566800863787633,
+      "loss": 1.1445,
+      "step": 658
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.45791282165531794,
+      "learning_rate": 0.00019565061608605526,
+      "loss": 1.1929,
+      "step": 659
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.38505947035292337,
+      "learning_rate": 0.00019563318946551998,
+      "loss": 1.32,
+      "step": 660
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.4019069563584587,
+      "learning_rate": 0.00019561572878247748,
+      "loss": 1.2946,
+      "step": 661
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.4674233690863355,
+      "learning_rate": 0.00019559823404314696,
+      "loss": 1.0958,
+      "step": 662
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.40636251321900174,
+      "learning_rate": 0.00019558070525375952,
+      "loss": 1.1248,
+      "step": 663
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.44575031199540055,
+      "learning_rate": 0.00019556314242055872,
+      "loss": 1.2777,
+      "step": 664
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.4198776066267316,
+      "learning_rate": 0.00019554554554979995,
+      "loss": 1.1588,
+      "step": 665
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.431669468006581,
+      "learning_rate": 0.00019552791464775088,
+      "loss": 1.1326,
+      "step": 666
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5016474775518394,
+      "learning_rate": 0.00019551024972069126,
+      "loss": 1.1827,
+      "step": 667
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5340242013500639,
+      "learning_rate": 0.00019549255077491292,
+      "loss": 1.3799,
+      "step": 668
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.3837464578757683,
+      "learning_rate": 0.00019547481781671992,
+      "loss": 1.1815,
+      "step": 669
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.48432370674209224,
+      "learning_rate": 0.00019545705085242834,
+      "loss": 1.1904,
+      "step": 670
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.4554652369263565,
+      "learning_rate": 0.00019543924988836637,
+      "loss": 1.2891,
+      "step": 671
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.396452957523324,
+      "learning_rate": 0.00019542141493087434,
+      "loss": 1.1592,
+      "step": 672
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.3876630830402,
+      "learning_rate": 0.00019540354598630467,
+      "loss": 1.2317,
+      "step": 673
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5370513687758054,
+      "learning_rate": 0.00019538564306102195,
+      "loss": 1.2259,
+      "step": 674
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.43691023440568044,
+      "learning_rate": 0.00019536770616140276,
+      "loss": 1.164,
+      "step": 675
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.3769921229634343,
+      "learning_rate": 0.0001953497352938359,
+      "loss": 1.2829,
+      "step": 676
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.4213830503620166,
+      "learning_rate": 0.00019533173046472216,
+      "loss": 1.2747,
+      "step": 677
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.42145595202661595,
+      "learning_rate": 0.0001953136916804745,
+      "loss": 1.392,
+      "step": 678
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.40185540124718694,
+      "learning_rate": 0.00019529561894751795,
+      "loss": 1.1238,
+      "step": 679
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.39911345946237486,
+      "learning_rate": 0.00019527751227228963,
+      "loss": 1.2468,
+      "step": 680
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.3991482306934748,
+      "learning_rate": 0.00019525937166123877,
+      "loss": 1.1832,
+      "step": 681
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.4734986073881463,
+      "learning_rate": 0.00019524119712082666,
+      "loss": 1.3147,
+      "step": 682
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.40693335674974535,
+      "learning_rate": 0.0001952229886575266,
+      "loss": 1.201,
+      "step": 683
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.49444086275265503,
+      "learning_rate": 0.00019520474627782424,
+      "loss": 1.1891,
+      "step": 684
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5819745566489066,
+      "learning_rate": 0.00019518646998821696,
+      "loss": 1.3937,
+      "step": 685
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.4400729279389,
+      "learning_rate": 0.00019516815979521445,
+      "loss": 1.1531,
+      "step": 686
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.4131416042725118,
+      "learning_rate": 0.00019514981570533846,
+      "loss": 1.203,
+      "step": 687
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.47957222738299166,
+      "learning_rate": 0.00019513143772512267,
+      "loss": 1.1502,
+      "step": 688
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.377075998404128,
+      "learning_rate": 0.00019511302586111296,
+      "loss": 1.2016,
+      "step": 689
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5168283859080812,
+      "learning_rate": 0.00019509458011986728,
+      "loss": 1.1248,
+      "step": 690
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.4247937557444455,
+      "learning_rate": 0.00019507610050795558,
+      "loss": 1.3033,
+      "step": 691
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.3813411719872442,
+      "learning_rate": 0.00019505758703195988,
+      "loss": 1.3633,
+      "step": 692
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.4177723766046739,
+      "learning_rate": 0.00019503903969847437,
+      "loss": 1.1502,
+      "step": 693
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5988452230176082,
+      "learning_rate": 0.00019502045851410514,
+      "loss": 1.1181,
+      "step": 694
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.38296976586353454,
+      "learning_rate": 0.00019500184348547042,
+      "loss": 1.2019,
+      "step": 695
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.562959449397898,
+      "learning_rate": 0.0001949831946192005,
+      "loss": 1.1786,
+      "step": 696
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.41116493042559715,
+      "learning_rate": 0.0001949645119219377,
+      "loss": 1.2071,
+      "step": 697
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.37619583791374994,
+      "learning_rate": 0.0001949457954003364,
+      "loss": 1.3045,
+      "step": 698
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.45511384066488375,
+      "learning_rate": 0.0001949270450610631,
+      "loss": 1.3149,
+      "step": 699
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.4846039261759903,
+      "learning_rate": 0.00019490826091079615,
+      "loss": 1.2606,
+      "step": 700
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.4560423678923779,
+      "learning_rate": 0.00019488944295622615,
+      "loss": 1.1341,
+      "step": 701
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.393567009582553,
+      "learning_rate": 0.0001948705912040556,
+      "loss": 1.2891,
+      "step": 702
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.3979942185238066,
+      "learning_rate": 0.00019485170566099913,
+      "loss": 1.2249,
+      "step": 703
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.44465141019545557,
+      "learning_rate": 0.00019483278633378336,
+      "loss": 1.1541,
+      "step": 704
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.4189210647797649,
+      "learning_rate": 0.00019481383322914698,
+      "loss": 1.1429,
+      "step": 705
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.39599862079533443,
+      "learning_rate": 0.00019479484635384061,
+      "loss": 1.2789,
+      "step": 706
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.40608084722092425,
+      "learning_rate": 0.00019477582571462705,
+      "loss": 1.1365,
+      "step": 707
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.4409951631005878,
+      "learning_rate": 0.00019475677131828095,
+      "loss": 1.1853,
+      "step": 708
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.4377827789625566,
+      "learning_rate": 0.0001947376831715892,
+      "loss": 1.2048,
+      "step": 709
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6313416281853269,
+      "learning_rate": 0.0001947185612813505,
+      "loss": 1.328,
+      "step": 710
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.4306095652504714,
+      "learning_rate": 0.00019469940565437573,
+      "loss": 1.3011,
+      "step": 711
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.44165416136192026,
+      "learning_rate": 0.00019468021629748768,
+      "loss": 1.2041,
+      "step": 712
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.49072294044982595,
+      "learning_rate": 0.00019466099321752115,
+      "loss": 1.1861,
+      "step": 713
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.4576052744057294,
+      "learning_rate": 0.00019464173642132307,
+      "loss": 1.3764,
+      "step": 714
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.4789217548369063,
+      "learning_rate": 0.00019462244591575222,
+      "loss": 1.2823,
+      "step": 715
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.3715375598264818,
+      "learning_rate": 0.00019460312170767956,
+      "loss": 1.1508,
+      "step": 716
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.4155887235843712,
+      "learning_rate": 0.00019458376380398788,
+      "loss": 1.2187,
+      "step": 717
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.48562923108893774,
+      "learning_rate": 0.0001945643722115721,
+      "loss": 1.2481,
+      "step": 718
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5625183107688074,
+      "learning_rate": 0.0001945449469373391,
+      "loss": 1.308,
+      "step": 719
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.3797496435242234,
+      "learning_rate": 0.00019452548798820768,
+      "loss": 1.1851,
+      "step": 720
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.39738057962251616,
+      "learning_rate": 0.00019450599537110881,
+      "loss": 1.3298,
+      "step": 721
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.41873052214572176,
+      "learning_rate": 0.00019448646909298525,
+      "loss": 1.2748,
+      "step": 722
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.4097637138568657,
+      "learning_rate": 0.0001944669091607919,
+      "loss": 1.2206,
+      "step": 723
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.3921835443525334,
+      "learning_rate": 0.00019444731558149555,
+      "loss": 1.1927,
+      "step": 724
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.41895940503805584,
+      "learning_rate": 0.00019442768836207508,
+      "loss": 1.1407,
+      "step": 725
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.419917506424284,
+      "learning_rate": 0.00019440802750952122,
+      "loss": 1.2347,
+      "step": 726
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.42524564280166066,
+      "learning_rate": 0.00019438833303083678,
+      "loss": 1.3227,
+      "step": 727
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.46651598055969756,
+      "learning_rate": 0.0001943686049330365,
+      "loss": 1.2196,
+      "step": 728
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.42923332079229903,
+      "learning_rate": 0.00019434884322314713,
+      "loss": 1.2668,
+      "step": 729
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.4447494990838177,
+      "learning_rate": 0.00019432904790820735,
+      "loss": 1.2725,
+      "step": 730
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.4082520653763171,
+      "learning_rate": 0.00019430921899526787,
+      "loss": 1.2188,
+      "step": 731
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5574025714820235,
+      "learning_rate": 0.00019428935649139123,
+      "loss": 1.221,
+      "step": 732
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.4110908786568132,
+      "learning_rate": 0.00019426946040365217,
+      "loss": 1.3692,
+      "step": 733
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.4027859442625039,
+      "learning_rate": 0.00019424953073913714,
+      "loss": 1.1448,
+      "step": 734
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.4428510146641056,
+      "learning_rate": 0.00019422956750494473,
+      "loss": 1.0468,
+      "step": 735
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.4019061158683859,
+      "learning_rate": 0.00019420957070818536,
+      "loss": 1.1394,
+      "step": 736
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.42156327825845413,
+      "learning_rate": 0.00019418954035598152,
+      "loss": 1.3722,
+      "step": 737
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.4800200972296228,
+      "learning_rate": 0.00019416947645546756,
+      "loss": 1.3807,
+      "step": 738
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.40455003476817863,
+      "learning_rate": 0.00019414937901378982,
+      "loss": 1.2093,
+      "step": 739
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.4014243694752976,
+      "learning_rate": 0.0001941292480381066,
+      "loss": 1.2256,
+      "step": 740
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.43055014941745556,
+      "learning_rate": 0.0001941090835355881,
+      "loss": 1.2297,
+      "step": 741
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.4712494724541032,
+      "learning_rate": 0.0001940888855134165,
+      "loss": 1.2827,
+      "step": 742
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.41887832476950165,
+      "learning_rate": 0.00019406865397878588,
+      "loss": 1.3553,
+      "step": 743
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.3778631672428122,
+      "learning_rate": 0.00019404838893890235,
+      "loss": 1.248,
+      "step": 744
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.4085165045427522,
+      "learning_rate": 0.00019402809040098382,
+      "loss": 1.2749,
+      "step": 745
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.39837130999573517,
+      "learning_rate": 0.00019400775837226023,
+      "loss": 1.4521,
+      "step": 746
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5254371225693698,
+      "learning_rate": 0.0001939873928599734,
+      "loss": 1.2884,
+      "step": 747
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.43036963741063466,
+      "learning_rate": 0.00019396699387137708,
+      "loss": 1.3164,
+      "step": 748
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.47932884699170764,
+      "learning_rate": 0.00019394656141373702,
+      "loss": 1.1254,
+      "step": 749
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5539949263637862,
+      "learning_rate": 0.0001939260954943307,
+      "loss": 0.9393,
+      "step": 750
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.3872679241643542,
+      "learning_rate": 0.0001939055961204478,
+      "loss": 1.1677,
+      "step": 751
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.42832006251242616,
+      "learning_rate": 0.00019388506329938966,
+      "loss": 1.2631,
+      "step": 752
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.40203978106345056,
+      "learning_rate": 0.00019386449703846968,
+      "loss": 1.1604,
+      "step": 753
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.4971156335529902,
+      "learning_rate": 0.00019384389734501307,
+      "loss": 1.2036,
+      "step": 754
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.42055297714428674,
+      "learning_rate": 0.00019382326422635705,
+      "loss": 1.1498,
+      "step": 755
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.39775320217629573,
+      "learning_rate": 0.00019380259768985068,
+      "loss": 1.2852,
+      "step": 756
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.3962362268563748,
+      "learning_rate": 0.00019378189774285496,
+      "loss": 1.2758,
+      "step": 757
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.413108619875974,
+      "learning_rate": 0.00019376116439274275,
+      "loss": 1.2595,
+      "step": 758
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.41147853493737124,
+      "learning_rate": 0.00019374039764689883,
+      "loss": 1.161,
+      "step": 759
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.3867812072850855,
+      "learning_rate": 0.00019371959751271985,
+      "loss": 1.3595,
+      "step": 760
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.37810542153782045,
+      "learning_rate": 0.00019369876399761443,
+      "loss": 1.2487,
+      "step": 761
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5227092423071035,
+      "learning_rate": 0.000193677897109003,
+      "loss": 1.3207,
+      "step": 762
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.41503950012416124,
+      "learning_rate": 0.0001936569968543179,
+      "loss": 1.2518,
+      "step": 763
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.4103396710300633,
+      "learning_rate": 0.00019363606324100335,
+      "loss": 1.1453,
+      "step": 764
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.47237165641464784,
+      "learning_rate": 0.00019361509627651547,
+      "loss": 1.3052,
+      "step": 765
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.4292880952676403,
+      "learning_rate": 0.00019359409596832224,
+      "loss": 1.2564,
+      "step": 766
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.4946907341710872,
+      "learning_rate": 0.0001935730623239035,
+      "loss": 1.0988,
+      "step": 767
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.39752384445587685,
+      "learning_rate": 0.00019355199535075102,
+      "loss": 1.2671,
+      "step": 768
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.44748776061750445,
+      "learning_rate": 0.00019353089505636838,
+      "loss": 1.3234,
+      "step": 769
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.39731575521819595,
+      "learning_rate": 0.00019350976144827107,
+      "loss": 1.1379,
+      "step": 770
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.3649721849606709,
+      "learning_rate": 0.00019348859453398646,
+      "loss": 1.2149,
+      "step": 771
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.44167028485558696,
+      "learning_rate": 0.00019346739432105374,
+      "loss": 1.0791,
+      "step": 772
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.41505670557688384,
+      "learning_rate": 0.00019344616081702393,
+      "loss": 1.2889,
+      "step": 773
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.44355545890690873,
+      "learning_rate": 0.00019342489402945998,
+      "loss": 1.3148,
+      "step": 774
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.39185195153472324,
+      "learning_rate": 0.00019340359396593672,
+      "loss": 1.4441,
+      "step": 775
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6295664793681127,
+      "learning_rate": 0.0001933822606340407,
+      "loss": 1.1284,
+      "step": 776
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.40138346410237163,
+      "learning_rate": 0.0001933608940413704,
+      "loss": 1.3034,
+      "step": 777
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.398514428650805,
+      "learning_rate": 0.00019333949419553622,
+      "loss": 1.1966,
+      "step": 778
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5253986281800577,
+      "learning_rate": 0.00019331806110416027,
+      "loss": 1.3812,
+      "step": 779
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.4535114242296223,
+      "learning_rate": 0.0001932965947748766,
+      "loss": 1.3184,
+      "step": 780
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.4126691473962807,
+      "learning_rate": 0.000193275095215331,
+      "loss": 1.2769,
+      "step": 781
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.4207700836305723,
+      "learning_rate": 0.0001932535624331812,
+      "loss": 0.9859,
+      "step": 782
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.4007212804689971,
+      "learning_rate": 0.00019323199643609673,
+      "loss": 1.4277,
+      "step": 783
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.4027474290522814,
+      "learning_rate": 0.00019321039723175888,
+      "loss": 1.3389,
+      "step": 784
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.4520849426463411,
+      "learning_rate": 0.00019318876482786094,
+      "loss": 1.3144,
+      "step": 785
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.43067825854133024,
+      "learning_rate": 0.00019316709923210775,
+      "loss": 1.3663,
+      "step": 786
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.4005090103349076,
+      "learning_rate": 0.00019314540045221626,
+      "loss": 1.265,
+      "step": 787
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.4327335398475432,
+      "learning_rate": 0.0001931236684959151,
+      "loss": 1.1107,
+      "step": 788
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.421206281051171,
+      "learning_rate": 0.0001931019033709447,
+      "loss": 1.2339,
+      "step": 789
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.44943326338941475,
+      "learning_rate": 0.00019308010508505734,
+      "loss": 1.1818,
+      "step": 790
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.43762609128583296,
+      "learning_rate": 0.0001930582736460171,
+      "loss": 1.2731,
+      "step": 791
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.47973641632125646,
+      "learning_rate": 0.0001930364090615999,
+      "loss": 1.2683,
+      "step": 792
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.4530270445124136,
+      "learning_rate": 0.0001930145113395934,
+      "loss": 1.2365,
+      "step": 793
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.4091981084523734,
+      "learning_rate": 0.00019299258048779714,
+      "loss": 1.2464,
+      "step": 794
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.42795532582898815,
+      "learning_rate": 0.00019297061651402236,
+      "loss": 1.1381,
+      "step": 795
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.39392474785526604,
+      "learning_rate": 0.00019294861942609228,
+      "loss": 1.1684,
+      "step": 796
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.4768364411501458,
+      "learning_rate": 0.00019292658923184165,
+      "loss": 1.1288,
+      "step": 797
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.45236795936457236,
+      "learning_rate": 0.0001929045259391173,
+      "loss": 1.3481,
+      "step": 798
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.4059090401624415,
+      "learning_rate": 0.00019288242955577757,
+      "loss": 1.2981,
+      "step": 799
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.4587393813709254,
+      "learning_rate": 0.00019286030008969283,
+      "loss": 1.3303,
+      "step": 800
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.41720577267329334,
+      "learning_rate": 0.00019283813754874504,
+      "loss": 1.189,
+      "step": 801
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.4541987066999266,
+      "learning_rate": 0.00019281594194082811,
+      "loss": 1.2247,
+      "step": 802
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.3959200581643399,
+      "learning_rate": 0.0001927937132738476,
+      "loss": 1.3061,
+      "step": 803
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.536639804902753,
+      "learning_rate": 0.00019277145155572085,
+      "loss": 1.2368,
+      "step": 804
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.38624898485700426,
+      "learning_rate": 0.00019274915679437705,
+      "loss": 1.1497,
+      "step": 805
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.41328061736388877,
+      "learning_rate": 0.0001927268289977571,
+      "loss": 1.113,
+      "step": 806
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.4448084131110642,
+      "learning_rate": 0.00019270446817381377,
+      "loss": 1.1078,
+      "step": 807
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.39438696858600586,
+      "learning_rate": 0.0001926820743305114,
+      "loss": 1.2595,
+      "step": 808
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.42201571083086853,
+      "learning_rate": 0.00019265964747582626,
+      "loss": 1.234,
+      "step": 809
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.4317638480660503,
+      "learning_rate": 0.00019263718761774627,
+      "loss": 1.3926,
+      "step": 810
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.46439074910883377,
+      "learning_rate": 0.0001926146947642712,
+      "loss": 1.3878,
+      "step": 811
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.42452790199327345,
+      "learning_rate": 0.00019259216892341253,
+      "loss": 1.3482,
+      "step": 812
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.47004529828425173,
+      "learning_rate": 0.00019256961010319344,
+      "loss": 1.1964,
+      "step": 813
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.3627380422960299,
+      "learning_rate": 0.00019254701831164893,
+      "loss": 1.2332,
+      "step": 814
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.39047840615175716,
+      "learning_rate": 0.00019252439355682573,
+      "loss": 1.2558,
+      "step": 815
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.39070427268786745,
+      "learning_rate": 0.00019250173584678225,
+      "loss": 1.3616,
+      "step": 816
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.45401685055596064,
+      "learning_rate": 0.00019247904518958876,
+      "loss": 1.1443,
+      "step": 817
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.412648541158305,
+      "learning_rate": 0.0001924563215933271,
+      "loss": 1.298,
+      "step": 818
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.4435080552453817,
+      "learning_rate": 0.000192433565066091,
+      "loss": 1.1014,
+      "step": 819
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.38205360339094235,
+      "learning_rate": 0.0001924107756159858,
+      "loss": 1.2597,
+      "step": 820
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.48553196968669543,
+      "learning_rate": 0.0001923879532511287,
+      "loss": 1.1589,
+      "step": 821
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.4983053453300486,
+      "learning_rate": 0.00019236509797964843,
+      "loss": 1.1133,
+      "step": 822
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.4151766446296914,
+      "learning_rate": 0.0001923422098096856,
+      "loss": 1.3541,
+      "step": 823
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.4313136691470291,
+      "learning_rate": 0.00019231928874939255,
+      "loss": 1.2482,
+      "step": 824
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.4466328741324138,
+      "learning_rate": 0.00019229633480693316,
+      "loss": 1.2769,
+      "step": 825
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.41250894624693135,
+      "learning_rate": 0.0001922733479904832,
+      "loss": 1.17,
+      "step": 826
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.4159680458360287,
+      "learning_rate": 0.00019225032830823011,
+      "loss": 1.1952,
+      "step": 827
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.40396774064475643,
+      "learning_rate": 0.00019222727576837295,
+      "loss": 1.1365,
+      "step": 828
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.39095313367026374,
+      "learning_rate": 0.00019220419037912261,
+      "loss": 1.2074,
+      "step": 829
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.44340117052844735,
+      "learning_rate": 0.00019218107214870155,
+      "loss": 1.1797,
+      "step": 830
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.3934251587873682,
+      "learning_rate": 0.000192157921085344,
+      "loss": 1.1984,
+      "step": 831
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5360718272295436,
+      "learning_rate": 0.00019213473719729594,
+      "loss": 1.3352,
+      "step": 832
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.4789321434453119,
+      "learning_rate": 0.0001921115204928149,
+      "loss": 1.2557,
+      "step": 833
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.41493260492546147,
+      "learning_rate": 0.00019208827098017022,
+      "loss": 1.1937,
+      "step": 834
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.45978023207810237,
+      "learning_rate": 0.00019206498866764288,
+      "loss": 1.158,
+      "step": 835
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.4304280013895555,
+      "learning_rate": 0.00019204167356352554,
+      "loss": 1.4127,
+      "step": 836
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.39069278952307407,
+      "learning_rate": 0.00019201832567612253,
+      "loss": 1.2735,
+      "step": 837
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5625374695234071,
+      "learning_rate": 0.0001919949450137499,
+      "loss": 1.1611,
+      "step": 838
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.44618941145853774,
+      "learning_rate": 0.00019197153158473528,
+      "loss": 1.3115,
+      "step": 839
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.380508844257389,
+      "learning_rate": 0.00019194808539741814,
+      "loss": 1.1597,
+      "step": 840
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.3840596414943476,
+      "learning_rate": 0.00019192460646014938,
+      "loss": 1.2551,
+      "step": 841
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.40254577437402933,
+      "learning_rate": 0.0001919010947812918,
+      "loss": 1.073,
+      "step": 842
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.4955256884385489,
+      "learning_rate": 0.00019187755036921978,
+      "loss": 1.384,
+      "step": 843
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.4289011337333196,
+      "learning_rate": 0.00019185397323231925,
+      "loss": 1.202,
+      "step": 844
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.38729595991246396,
+      "learning_rate": 0.0001918303633789879,
+      "loss": 1.1763,
+      "step": 845
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.361380647702892,
+      "learning_rate": 0.0001918067208176351,
+      "loss": 1.2709,
+      "step": 846
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.4152373727434519,
+      "learning_rate": 0.00019178304555668186,
+      "loss": 1.1227,
+      "step": 847
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.39767650518444597,
+      "learning_rate": 0.00019175933760456069,
+      "loss": 1.1899,
+      "step": 848
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.42344060786507304,
+      "learning_rate": 0.00019173559696971594,
+      "loss": 1.2085,
+      "step": 849
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.43165172270108715,
+      "learning_rate": 0.0001917118236606035,
+      "loss": 1.2053,
+      "step": 850
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.4072700269407179,
+      "learning_rate": 0.0001916880176856909,
+      "loss": 1.2407,
+      "step": 851
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.4226710603341286,
+      "learning_rate": 0.00019166417905345736,
+      "loss": 1.1917,
+      "step": 852
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6379341278886934,
+      "learning_rate": 0.00019164030777239367,
+      "loss": 1.0731,
+      "step": 853
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.43994195297288885,
+      "learning_rate": 0.00019161640385100228,
+      "loss": 1.275,
+      "step": 854
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.41921255015220515,
+      "learning_rate": 0.00019159246729779725,
+      "loss": 1.2626,
+      "step": 855
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.3816809551819136,
+      "learning_rate": 0.00019156849812130427,
+      "loss": 1.3405,
+      "step": 856
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.46822789691471434,
+      "learning_rate": 0.00019154449633006064,
+      "loss": 1.148,
+      "step": 857
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.4366616874825709,
+      "learning_rate": 0.0001915204619326153,
+      "loss": 1.2543,
+      "step": 858
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.4328927133876195,
+      "learning_rate": 0.0001914963949375288,
+      "loss": 1.3632,
+      "step": 859
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.44884859878198724,
+      "learning_rate": 0.0001914722953533733,
+      "loss": 1.1548,
+      "step": 860
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.4645861507220917,
+      "learning_rate": 0.00019144816318873253,
+      "loss": 1.403,
+      "step": 861
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.3872709696536166,
+      "learning_rate": 0.00019142399845220188,
+      "loss": 1.2047,
+      "step": 862
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.40370932998399006,
+      "learning_rate": 0.00019139980115238827,
+      "loss": 1.3221,
+      "step": 863
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5599998567514824,
+      "learning_rate": 0.00019137557129791032,
+      "loss": 1.263,
+      "step": 864
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.4254672718118661,
+      "learning_rate": 0.00019135130889739817,
+      "loss": 1.3913,
+      "step": 865
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.42040520244751295,
+      "learning_rate": 0.00019132701395949356,
+      "loss": 1.3305,
+      "step": 866
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.4011576607383039,
+      "learning_rate": 0.0001913026864928498,
+      "loss": 1.2384,
+      "step": 867
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.48013069868937264,
+      "learning_rate": 0.00019127832650613189,
+      "loss": 1.1196,
+      "step": 868
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.4041893509898504,
+      "learning_rate": 0.0001912539340080163,
+      "loss": 1.2537,
+      "step": 869
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.43007343594065317,
+      "learning_rate": 0.0001912295090071911,
+      "loss": 1.1986,
+      "step": 870
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.44499029302511794,
+      "learning_rate": 0.00019120505151235597,
+      "loss": 1.0713,
+      "step": 871
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5150194526436023,
+      "learning_rate": 0.0001911805615322222,
+      "loss": 1.1561,
+      "step": 872
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.39833034607657375,
+      "learning_rate": 0.0001911560390755125,
+      "loss": 1.3401,
+      "step": 873
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.38602186984832265,
+      "learning_rate": 0.0001911314841509613,
+      "loss": 1.1354,
+      "step": 874
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.48830718669904294,
+      "learning_rate": 0.00019110689676731454,
+      "loss": 1.2038,
+      "step": 875
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.4018388860207783,
+      "learning_rate": 0.00019108227693332975,
+      "loss": 1.3433,
+      "step": 876
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.4029344709598993,
+      "learning_rate": 0.00019105762465777598,
+      "loss": 1.0896,
+      "step": 877
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.45661237173399377,
+      "learning_rate": 0.00019103293994943382,
+      "loss": 1.268,
+      "step": 878
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.40438141295966257,
+      "learning_rate": 0.00019100822281709542,
+      "loss": 1.22,
+      "step": 879
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.4257610438665412,
+      "learning_rate": 0.00019098347326956453,
+      "loss": 1.1715,
+      "step": 880
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.43300562940061554,
+      "learning_rate": 0.00019095869131565644,
+      "loss": 1.2071,
+      "step": 881
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.3879878957200197,
+      "learning_rate": 0.00019093387696419788,
+      "loss": 1.2694,
+      "step": 882
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.4238073590458465,
+      "learning_rate": 0.00019090903022402729,
+      "loss": 1.1846,
+      "step": 883
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.48415219379026003,
+      "learning_rate": 0.00019088415110399444,
+      "loss": 1.104,
+      "step": 884
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.44089814568607455,
+      "learning_rate": 0.00019085923961296083,
+      "loss": 1.0628,
+      "step": 885
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.4205670521660812,
+      "learning_rate": 0.00019083429575979935,
+      "loss": 1.1419,
+      "step": 886
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.4000507182835794,
+      "learning_rate": 0.00019080931955339455,
+      "loss": 1.2108,
+      "step": 887
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.3656575093951322,
+      "learning_rate": 0.00019078431100264234,
+      "loss": 1.4073,
+      "step": 888
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.417808756619523,
+      "learning_rate": 0.00019075927011645024,
+      "loss": 1.2921,
+      "step": 889
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.4256288121907213,
+      "learning_rate": 0.00019073419690373731,
+      "loss": 1.2812,
+      "step": 890
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.4188379180794639,
+      "learning_rate": 0.00019070909137343408,
+      "loss": 1.159,
+      "step": 891
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.43118494202925123,
+      "learning_rate": 0.0001906839535344826,
+      "loss": 1.23,
+      "step": 892
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.4223281863588074,
+      "learning_rate": 0.00019065878339583647,
+      "loss": 1.0364,
+      "step": 893
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.4663925484403322,
+      "learning_rate": 0.00019063358096646073,
+      "loss": 1.0427,
+      "step": 894
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.4374061811963192,
+      "learning_rate": 0.00019060834625533196,
+      "loss": 1.266,
+      "step": 895
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.40481395021572636,
+      "learning_rate": 0.0001905830792714382,
+      "loss": 1.1859,
+      "step": 896
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.4130384848039822,
+      "learning_rate": 0.00019055778002377905,
+      "loss": 1.1271,
+      "step": 897
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.39622200548897246,
+      "learning_rate": 0.00019053244852136554,
+      "loss": 1.177,
+      "step": 898
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.49977453441314185,
+      "learning_rate": 0.00019050708477322018,
+      "loss": 1.1857,
+      "step": 899
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.4571695445850005,
+      "learning_rate": 0.00019048168878837708,
+      "loss": 1.2953,
+      "step": 900
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.41811429475757356,
+      "learning_rate": 0.00019045626057588168,
+      "loss": 1.1684,
+      "step": 901
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.4109647055727443,
+      "learning_rate": 0.00019043080014479098,
+      "loss": 1.0853,
+      "step": 902
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.40909296031705294,
+      "learning_rate": 0.00019040530750417348,
+      "loss": 1.3434,
+      "step": 903
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.3843456477902698,
+      "learning_rate": 0.0001903797826631091,
+      "loss": 1.2941,
+      "step": 904
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.3944000573778209,
+      "learning_rate": 0.0001903542256306892,
+      "loss": 1.2048,
+      "step": 905
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.38222965940856135,
+      "learning_rate": 0.0001903286364160167,
+      "loss": 1.2566,
+      "step": 906
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.4341400096930122,
+      "learning_rate": 0.00019030301502820596,
+      "loss": 1.2163,
+      "step": 907
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.41143088017038376,
+      "learning_rate": 0.0001902773614763827,
+      "loss": 1.2485,
+      "step": 908
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.4697355410098832,
+      "learning_rate": 0.00019025167576968422,
+      "loss": 1.2431,
+      "step": 909
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.3695207210021833,
+      "learning_rate": 0.00019022595791725916,
+      "loss": 1.3462,
+      "step": 910
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.3971784373639261,
+      "learning_rate": 0.00019020020792826777,
+      "loss": 1.2226,
+      "step": 911
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.3724694197441415,
+      "learning_rate": 0.0001901744258118816,
+      "loss": 1.3761,
+      "step": 912
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.39806399623199995,
+      "learning_rate": 0.00019014861157728364,
+      "loss": 1.3343,
+      "step": 913
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.42185545031726196,
+      "learning_rate": 0.00019012276523366848,
+      "loss": 1.2069,
+      "step": 914
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.4120235602738129,
+      "learning_rate": 0.0001900968867902419,
+      "loss": 1.3521,
+      "step": 915
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.44557384282152157,
+      "learning_rate": 0.0001900709762562214,
+      "loss": 1.2852,
+      "step": 916
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.4229024065060597,
+      "learning_rate": 0.00019004503364083566,
+      "loss": 1.2152,
+      "step": 917
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.3845162026550113,
+      "learning_rate": 0.00019001905895332493,
+      "loss": 1.344,
+      "step": 918
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.4385329323104083,
+      "learning_rate": 0.0001899930522029408,
+      "loss": 1.1733,
+      "step": 919
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.39973360112576806,
+      "learning_rate": 0.00018996701339894635,
+      "loss": 1.1761,
+      "step": 920
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5098598984531003,
+      "learning_rate": 0.00018994094255061605,
+      "loss": 1.2665,
+      "step": 921
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.41501129332074677,
+      "learning_rate": 0.0001899148396672358,
+      "loss": 1.1719,
+      "step": 922
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.4375418833196446,
+      "learning_rate": 0.00018988870475810282,
+      "loss": 1.1252,
+      "step": 923
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.41840954269661546,
+      "learning_rate": 0.00018986253783252592,
+      "loss": 1.1791,
+      "step": 924
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5223116611468935,
+      "learning_rate": 0.00018983633889982512,
+      "loss": 1.4298,
+      "step": 925
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.4460054856377424,
+      "learning_rate": 0.0001898101079693319,
+      "loss": 1.2528,
+      "step": 926
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.3970605721835403,
+      "learning_rate": 0.00018978384505038924,
+      "loss": 1.2816,
+      "step": 927
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.4513169156455377,
+      "learning_rate": 0.0001897575501523514,
+      "loss": 1.1952,
+      "step": 928
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.41262985664217344,
+      "learning_rate": 0.00018973122328458405,
+      "loss": 1.1255,
+      "step": 929
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.3858334392938787,
+      "learning_rate": 0.00018970486445646424,
+      "loss": 1.2696,
+      "step": 930
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.42048038129213106,
+      "learning_rate": 0.00018967847367738048,
+      "loss": 1.2296,
+      "step": 931
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.40003724599938,
+      "learning_rate": 0.00018965205095673255,
+      "loss": 1.352,
+      "step": 932
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.39214423359911943,
+      "learning_rate": 0.00018962559630393173,
+      "loss": 1.1577,
+      "step": 933
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.44665735406057383,
+      "learning_rate": 0.00018959910972840053,
+      "loss": 1.1946,
+      "step": 934
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.40856917621010586,
+      "learning_rate": 0.00018957259123957295,
+      "loss": 1.3262,
+      "step": 935
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.389339652211477,
+      "learning_rate": 0.00018954604084689427,
+      "loss": 1.3024,
+      "step": 936
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.4844504669643572,
+      "learning_rate": 0.00018951945855982122,
+      "loss": 1.1493,
+      "step": 937
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.517700302764435,
+      "learning_rate": 0.00018949284438782182,
+      "loss": 1.1545,
+      "step": 938
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.4827736628994079,
+      "learning_rate": 0.00018946619834037546,
+      "loss": 1.1458,
+      "step": 939
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.39232933604902814,
+      "learning_rate": 0.00018943952042697296,
+      "loss": 1.2498,
+      "step": 940
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.4334901715588195,
+      "learning_rate": 0.00018941281065711635,
+      "loss": 1.0697,
+      "step": 941
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.4020464214603877,
+      "learning_rate": 0.00018938606904031914,
+      "loss": 1.205,
+      "step": 942
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.398193639903526,
+      "learning_rate": 0.00018935929558610605,
+      "loss": 1.1059,
+      "step": 943
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.41730287186251896,
+      "learning_rate": 0.00018933249030401327,
+      "loss": 1.2625,
+      "step": 944
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.3957894271851279,
+      "learning_rate": 0.00018930565320358827,
+      "loss": 1.307,
+      "step": 945
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.43593377640274567,
+      "learning_rate": 0.00018927878429438985,
+      "loss": 1.2565,
+      "step": 946
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.38150373660998105,
+      "learning_rate": 0.00018925188358598813,
+      "loss": 1.2062,
+      "step": 947
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.3908717081425883,
+      "learning_rate": 0.00018922495108796459,
+      "loss": 1.0937,
+      "step": 948
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.39074937773233526,
+      "learning_rate": 0.000189197986809912,
+      "loss": 1.1237,
+      "step": 949
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.40985420308682935,
+      "learning_rate": 0.00018917099076143443,
+      "loss": 1.3218,
+      "step": 950
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.4087506296367563,
+      "learning_rate": 0.00018914396295214737,
+      "loss": 1.2038,
+      "step": 951
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.38493018619813185,
+      "learning_rate": 0.0001891169033916775,
+      "loss": 1.3003,
+      "step": 952
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.45889224911836823,
+      "learning_rate": 0.00018908981208966288,
+      "loss": 1.2854,
+      "step": 953
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.44867625414719753,
+      "learning_rate": 0.00018906268905575285,
+      "loss": 1.1539,
+      "step": 954
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5381457518540363,
+      "learning_rate": 0.00018903553429960802,
+      "loss": 1.1156,
+      "step": 955
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.4084599965440794,
+      "learning_rate": 0.00018900834783090043,
+      "loss": 1.2484,
+      "step": 956
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.607728963308825,
+      "learning_rate": 0.0001889811296593132,
+      "loss": 1.2307,
+      "step": 957
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.4362201232561436,
+      "learning_rate": 0.000188953879794541,
+      "loss": 1.0896,
+      "step": 958
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.4738519448256955,
+      "learning_rate": 0.00018892659824628952,
+      "loss": 1.2007,
+      "step": 959
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.4260518416715282,
+      "learning_rate": 0.0001888992850242759,
+      "loss": 1.2366,
+      "step": 960
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.43542079150277924,
+      "learning_rate": 0.00018887194013822862,
+      "loss": 1.3779,
+      "step": 961
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.4481122918284739,
+      "learning_rate": 0.00018884456359788724,
+      "loss": 1.2213,
+      "step": 962
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.4042061670221692,
+      "learning_rate": 0.00018881715541300276,
+      "loss": 1.2937,
+      "step": 963
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.4441725070050059,
+      "learning_rate": 0.00018878971559333735,
+      "loss": 1.2069,
+      "step": 964
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.47302700413023774,
+      "learning_rate": 0.00018876224414866454,
+      "loss": 1.1549,
+      "step": 965
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.42081809390928004,
+      "learning_rate": 0.00018873474108876902,
+      "loss": 1.1807,
+      "step": 966
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.4516926250509892,
+      "learning_rate": 0.0001887072064234469,
+      "loss": 1.3504,
+      "step": 967
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.48889640265702733,
+      "learning_rate": 0.00018867964016250528,
+      "loss": 1.1695,
+      "step": 968
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.49552613902020354,
+      "learning_rate": 0.00018865204231576277,
+      "loss": 1.1493,
+      "step": 969
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5907268474491925,
+      "learning_rate": 0.00018862441289304915,
+      "loss": 1.1861,
+      "step": 970
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.43819502062194104,
+      "learning_rate": 0.00018859675190420537,
+      "loss": 1.3287,
+      "step": 971
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.44828153696058587,
+      "learning_rate": 0.0001885690593590838,
+      "loss": 1.2506,
+      "step": 972
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.43902655104810623,
+      "learning_rate": 0.0001885413352675478,
+      "loss": 1.214,
+      "step": 973
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.4455049277068661,
+      "learning_rate": 0.00018851357963947218,
+      "loss": 1.2116,
+      "step": 974
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.40929747474995865,
+      "learning_rate": 0.00018848579248474288,
+      "loss": 1.1832,
+      "step": 975
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.4143703947614673,
+      "learning_rate": 0.00018845797381325708,
+      "loss": 1.2356,
+      "step": 976
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5056004449538829,
+      "learning_rate": 0.00018843012363492325,
+      "loss": 1.119,
+      "step": 977
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.45462371470480845,
+      "learning_rate": 0.000188402241959661,
+      "loss": 1.41,
+      "step": 978
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.39572179657366774,
+      "learning_rate": 0.00018837432879740114,
+      "loss": 1.2947,
+      "step": 979
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.4310739980311259,
+      "learning_rate": 0.00018834638415808584,
+      "loss": 1.2796,
+      "step": 980
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.38336972962117627,
+      "learning_rate": 0.0001883184080516683,
+      "loss": 1.4029,
+      "step": 981
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.41298219809794845,
+      "learning_rate": 0.00018829040048811305,
+      "loss": 1.3686,
+      "step": 982
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.45446325417306926,
+      "learning_rate": 0.0001882623614773958,
+      "loss": 1.2294,
+      "step": 983
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.3883599266353209,
+      "learning_rate": 0.00018823429102950346,
+      "loss": 1.3607,
+      "step": 984
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.4842285435751276,
+      "learning_rate": 0.00018820618915443412,
+      "loss": 1.1718,
+      "step": 985
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.4171740956102097,
+      "learning_rate": 0.000188178055862197,
+      "loss": 1.2121,
+      "step": 986
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.45946933652082567,
+      "learning_rate": 0.0001881498911628127,
+      "loss": 1.1684,
+      "step": 987
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5086386566782248,
+      "learning_rate": 0.00018812169506631277,
+      "loss": 1.1785,
+      "step": 988
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.44371765211130115,
+      "learning_rate": 0.00018809346758274012,
+      "loss": 1.0477,
+      "step": 989
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.3979138340321608,
+      "learning_rate": 0.00018806520872214879,
+      "loss": 1.2098,
+      "step": 990
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.45340103603593596,
+      "learning_rate": 0.00018803691849460393,
+      "loss": 1.1949,
+      "step": 991
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5062178650062923,
+      "learning_rate": 0.000188008596910182,
+      "loss": 1.2677,
+      "step": 992
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.4729167653041863,
+      "learning_rate": 0.00018798024397897046,
+      "loss": 1.2423,
+      "step": 993
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.4084725352578991,
+      "learning_rate": 0.00018795185971106805,
+      "loss": 1.2462,
+      "step": 994
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.4711554873736235,
+      "learning_rate": 0.00018792344411658468,
+      "loss": 1.1638,
+      "step": 995
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.41519453756629127,
+      "learning_rate": 0.00018789499720564136,
+      "loss": 1.2676,
+      "step": 996
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.3996280601561714,
+      "learning_rate": 0.00018786651898837023,
+      "loss": 1.3237,
+      "step": 997
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.39012615461328876,
+      "learning_rate": 0.00018783800947491468,
+      "loss": 1.1895,
+      "step": 998
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5290249154672712,
+      "learning_rate": 0.00018780946867542919,
+      "loss": 1.2825,
+      "step": 999
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.39832368529898793,
+      "learning_rate": 0.00018778089660007936,
+      "loss": 1.2954,
+      "step": 1000
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.430133999909778,
+      "learning_rate": 0.00018775229325904193,
+      "loss": 1.3023,
+      "step": 1001
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.4097485341707277,
+      "learning_rate": 0.00018772365866250485,
+      "loss": 1.3461,
+      "step": 1002
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.4146424251013908,
+      "learning_rate": 0.00018769499282066717,
+      "loss": 1.2595,
+      "step": 1003
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.43891632931182173,
+      "learning_rate": 0.000187666295743739,
+      "loss": 1.0531,
+      "step": 1004
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.4329889574840284,
+      "learning_rate": 0.00018763756744194163,
+      "loss": 1.2302,
+      "step": 1005
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.4231675121399311,
+      "learning_rate": 0.00018760880792550753,
+      "loss": 1.4761,
+      "step": 1006
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.396933859253791,
+      "learning_rate": 0.00018758001720468016,
+      "loss": 1.2613,
+      "step": 1007
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.38779305057942864,
+      "learning_rate": 0.0001875511952897142,
+      "loss": 1.4034,
+      "step": 1008
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.39581726519146576,
+      "learning_rate": 0.00018752234219087538,
+      "loss": 1.2344,
+      "step": 1009
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.480613824827418,
+      "learning_rate": 0.0001874934579184406,
+      "loss": 1.2133,
+      "step": 1010
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5084620111928327,
+      "learning_rate": 0.00018746454248269777,
+      "loss": 1.1794,
+      "step": 1011
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.46289518985979955,
+      "learning_rate": 0.00018743559589394596,
+      "loss": 1.2355,
+      "step": 1012
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.43404074799584486,
+      "learning_rate": 0.00018740661816249538,
+      "loss": 1.2651,
+      "step": 1013
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5219741788371963,
+      "learning_rate": 0.00018737760929866726,
+      "loss": 1.1495,
+      "step": 1014
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.4615237781416015,
+      "learning_rate": 0.0001873485693127939,
+      "loss": 1.2503,
+      "step": 1015
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.4759827980164268,
+      "learning_rate": 0.00018731949821521878,
+      "loss": 1.1466,
+      "step": 1016
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.3759152758522113,
+      "learning_rate": 0.00018729039601629634,
+      "loss": 1.208,
+      "step": 1017
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.46609668384649106,
+      "learning_rate": 0.00018726126272639225,
+      "loss": 1.2127,
+      "step": 1018
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.4259488550357106,
+      "learning_rate": 0.0001872320983558831,
+      "loss": 1.2889,
+      "step": 1019
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.39483853525415574,
+      "learning_rate": 0.00018720290291515665,
+      "loss": 1.2513,
+      "step": 1020
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.40719119890026767,
+      "learning_rate": 0.00018717367641461166,
+      "loss": 1.292,
+      "step": 1021
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.3966655214843322,
+      "learning_rate": 0.00018714441886465803,
+      "loss": 1.187,
+      "step": 1022
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.4653122753948471,
+      "learning_rate": 0.0001871151302757167,
+      "loss": 1.1074,
+      "step": 1023
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.43849692064586415,
+      "learning_rate": 0.00018708581065821954,
+      "loss": 1.1801,
+      "step": 1024
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.38995714676382615,
+      "learning_rate": 0.00018705646002260964,
+      "loss": 1.1504,
+      "step": 1025
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.3844503476315502,
+      "learning_rate": 0.00018702707837934111,
+      "loss": 0.9953,
+      "step": 1026
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.40955758500611017,
+      "learning_rate": 0.000186997665738879,
+      "loss": 1.1173,
+      "step": 1027
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.4517573186130042,
+      "learning_rate": 0.00018696822211169948,
+      "loss": 1.0788,
+      "step": 1028
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.3857214058663221,
+      "learning_rate": 0.00018693874750828977,
+      "loss": 1.3218,
+      "step": 1029
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.484472524626531,
+      "learning_rate": 0.00018690924193914807,
+      "loss": 1.3074,
+      "step": 1030
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.4490568939338803,
+      "learning_rate": 0.00018687970541478364,
+      "loss": 1.2397,
+      "step": 1031
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.4235227375999166,
+      "learning_rate": 0.0001868501379457168,
+      "loss": 1.2557,
+      "step": 1032
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.4268815968986065,
+      "learning_rate": 0.0001868205395424788,
+      "loss": 1.3362,
+      "step": 1033
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.49692955253926413,
+      "learning_rate": 0.000186790910215612,
+      "loss": 1.1285,
+      "step": 1034
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.4396647407064058,
+      "learning_rate": 0.0001867612499756697,
+      "loss": 1.1309,
+      "step": 1035
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.4052879481399543,
+      "learning_rate": 0.00018673155883321626,
+      "loss": 1.2164,
+      "step": 1036
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.47506929433842704,
+      "learning_rate": 0.00018670183679882704,
+      "loss": 1.3218,
+      "step": 1037
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.4199432818084886,
+      "learning_rate": 0.00018667208388308841,
+      "loss": 1.0845,
+      "step": 1038
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.4192145697941194,
+      "learning_rate": 0.00018664230009659768,
+      "loss": 1.3479,
+      "step": 1039
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.45263330217555126,
+      "learning_rate": 0.00018661248544996325,
+      "loss": 1.2566,
+      "step": 1040
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.47541292078899056,
+      "learning_rate": 0.00018658263995380446,
+      "loss": 1.1881,
+      "step": 1041
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.3808988352249516,
+      "learning_rate": 0.00018655276361875163,
+      "loss": 1.3016,
+      "step": 1042
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.373144299272529,
+      "learning_rate": 0.00018652285645544603,
+      "loss": 1.1944,
+      "step": 1043
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.4212394505385918,
+      "learning_rate": 0.00018649291847454,
+      "loss": 1.3134,
+      "step": 1044
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.41026255664647343,
+      "learning_rate": 0.00018646294968669685,
+      "loss": 1.2134,
+      "step": 1045
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.37585261021460126,
+      "learning_rate": 0.00018643295010259075,
+      "loss": 1.2455,
+      "step": 1046
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.40852856562199363,
+      "learning_rate": 0.00018640291973290695,
+      "loss": 1.1969,
+      "step": 1047
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.46555991794078955,
+      "learning_rate": 0.00018637285858834162,
+      "loss": 1.2023,
+      "step": 1048
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.4223974227836315,
+      "learning_rate": 0.0001863427666796019,
+      "loss": 1.1793,
+      "step": 1049
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.4654073859284596,
+      "learning_rate": 0.0001863126440174059,
+      "loss": 1.2436,
+      "step": 1050
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.46929391478231774,
+      "learning_rate": 0.00018628249061248262,
+      "loss": 1.0477,
+      "step": 1051
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.4058750030748448,
+      "learning_rate": 0.00018625230647557217,
+      "loss": 1.2278,
+      "step": 1052
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.4067672726882714,
+      "learning_rate": 0.00018622209161742536,
+      "loss": 1.1263,
+      "step": 1053
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.43337903939971373,
+      "learning_rate": 0.00018619184604880422,
+      "loss": 1.1861,
+      "step": 1054
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.49969091111093505,
+      "learning_rate": 0.00018616156978048146,
+      "loss": 1.095,
+      "step": 1055
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.43030385194079795,
+      "learning_rate": 0.00018613126282324092,
+      "loss": 1.2692,
+      "step": 1056
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6089572366385503,
+      "learning_rate": 0.00018610092518787723,
+      "loss": 1.0702,
+      "step": 1057
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5069822128144771,
+      "learning_rate": 0.00018607055688519606,
+      "loss": 1.0565,
+      "step": 1058
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.3767683355961435,
+      "learning_rate": 0.00018604015792601396,
+      "loss": 1.1438,
+      "step": 1059
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.4393766651757049,
+      "learning_rate": 0.00018600972832115832,
+      "loss": 1.2533,
+      "step": 1060
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.4308911954764944,
+      "learning_rate": 0.0001859792680814676,
+      "loss": 1.1909,
+      "step": 1061
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.3478460097560461,
+      "learning_rate": 0.00018594877721779106,
+      "loss": 1.3159,
+      "step": 1062
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.4280579631014377,
+      "learning_rate": 0.0001859182557409889,
+      "loss": 1.1908,
+      "step": 1063
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.47737005793003645,
+      "learning_rate": 0.0001858877036619322,
+      "loss": 1.166,
+      "step": 1064
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.43720812311630247,
+      "learning_rate": 0.000185857120991503,
+      "loss": 1.0988,
+      "step": 1065
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.45361420259634844,
+      "learning_rate": 0.0001858265077405941,
+      "loss": 1.2902,
+      "step": 1066
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.4481903047721264,
+      "learning_rate": 0.00018579586392010943,
+      "loss": 1.1924,
+      "step": 1067
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.36557709910612896,
+      "learning_rate": 0.00018576518954096358,
+      "loss": 1.1172,
+      "step": 1068
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.4168080259815743,
+      "learning_rate": 0.0001857344846140821,
+      "loss": 1.0633,
+      "step": 1069
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.405904756445125,
+      "learning_rate": 0.0001857037491504015,
+      "loss": 1.1476,
+      "step": 1070
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.4096037413051564,
+      "learning_rate": 0.00018567298316086902,
+      "loss": 1.1981,
+      "step": 1071
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.4205684501591133,
+      "learning_rate": 0.00018564218665644293,
+      "loss": 1.2515,
+      "step": 1072
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.4745463951647767,
+      "learning_rate": 0.00018561135964809223,
+      "loss": 1.331,
+      "step": 1073
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.537507538537619,
+      "learning_rate": 0.00018558050214679685,
+      "loss": 1.277,
+      "step": 1074
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.3545560763005173,
+      "learning_rate": 0.0001855496141635476,
+      "loss": 1.1528,
+      "step": 1075
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5431808471841911,
+      "learning_rate": 0.00018551869570934608,
+      "loss": 1.1293,
+      "step": 1076
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.40448854804219314,
+      "learning_rate": 0.00018548774679520485,
+      "loss": 1.0832,
+      "step": 1077
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.41935817801874525,
+      "learning_rate": 0.0001854567674321472,
+      "loss": 1.2399,
+      "step": 1078
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.47235419770082326,
+      "learning_rate": 0.00018542575763120739,
+      "loss": 1.1893,
+      "step": 1079
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.41787254362126686,
+      "learning_rate": 0.00018539471740343033,
+      "loss": 1.2183,
+      "step": 1080
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.47022535353393097,
+      "learning_rate": 0.00018536364675987202,
+      "loss": 1.1999,
+      "step": 1081
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.4056949468866122,
+      "learning_rate": 0.00018533254571159903,
+      "loss": 1.2016,
+      "step": 1082
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.4488709840574702,
+      "learning_rate": 0.00018530141426968902,
+      "loss": 1.2384,
+      "step": 1083
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.3951053956154525,
+      "learning_rate": 0.0001852702524452303,
+      "loss": 1.03,
+      "step": 1084
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.45412525852702507,
+      "learning_rate": 0.00018523906024932202,
+      "loss": 1.3212,
+      "step": 1085
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.474539779885222,
+      "learning_rate": 0.0001852078376930742,
+      "loss": 1.3351,
+      "step": 1086
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.3976875441825084,
+      "learning_rate": 0.0001851765847876076,
+      "loss": 1.2086,
+      "step": 1087
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.40841805345954285,
+      "learning_rate": 0.00018514530154405394,
+      "loss": 1.1381,
+      "step": 1088
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.41204131694450336,
+      "learning_rate": 0.0001851139879735556,
+      "loss": 1.0874,
+      "step": 1089
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5055655393933399,
+      "learning_rate": 0.00018508264408726577,
+      "loss": 1.2602,
+      "step": 1090
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.423394019684278,
+      "learning_rate": 0.0001850512698963485,
+      "loss": 1.1282,
+      "step": 1091
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5081216453883313,
+      "learning_rate": 0.00018501986541197865,
+      "loss": 1.1523,
+      "step": 1092
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.439321103971142,
+      "learning_rate": 0.00018498843064534176,
+      "loss": 1.1584,
+      "step": 1093
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.4347947406914529,
+      "learning_rate": 0.00018495696560763428,
+      "loss": 1.358,
+      "step": 1094
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.45729741574617605,
+      "learning_rate": 0.00018492547031006336,
+      "loss": 1.1371,
+      "step": 1095
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.4292205398427925,
+      "learning_rate": 0.00018489394476384696,
+      "loss": 1.2805,
+      "step": 1096
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.4637268814939168,
+      "learning_rate": 0.0001848623889802138,
+      "loss": 1.1558,
+      "step": 1097
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.4340308830313106,
+      "learning_rate": 0.0001848308029704034,
+      "loss": 1.2692,
+      "step": 1098
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.37669235242269866,
+      "learning_rate": 0.000184799186745666,
+      "loss": 1.3571,
+      "step": 1099
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.4166263315490858,
+      "learning_rate": 0.00018476754031726265,
+      "loss": 1.3058,
+      "step": 1100
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.4279264801792058,
+      "learning_rate": 0.00018473586369646512,
+      "loss": 1.1472,
+      "step": 1101
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.4047421531380947,
+      "learning_rate": 0.00018470415689455597,
+      "loss": 1.281,
+      "step": 1102
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.3938450928410792,
+      "learning_rate": 0.00018467241992282843,
+      "loss": 1.2185,
+      "step": 1103
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.4551147380551678,
+      "learning_rate": 0.00018464065279258658,
+      "loss": 1.2293,
+      "step": 1104
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.4690054305611647,
+      "learning_rate": 0.00018460885551514522,
+      "loss": 1.2593,
+      "step": 1105
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.3926647817374453,
+      "learning_rate": 0.0001845770281018298,
+      "loss": 1.2719,
+      "step": 1106
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.4256835636921655,
+      "learning_rate": 0.00018454517056397661,
+      "loss": 1.0671,
+      "step": 1107
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.4163956789351297,
+      "learning_rate": 0.00018451328291293264,
+      "loss": 1.0803,
+      "step": 1108
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.4327793451449362,
+      "learning_rate": 0.00018448136516005555,
+      "loss": 1.2482,
+      "step": 1109
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.4308145225744109,
+      "learning_rate": 0.00018444941731671376,
+      "loss": 1.1913,
+      "step": 1110
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.4360466988144919,
+      "learning_rate": 0.00018441743939428648,
+      "loss": 1.1655,
+      "step": 1111
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5071106519916768,
+      "learning_rate": 0.00018438543140416352,
+      "loss": 1.184,
+      "step": 1112
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.45305419099007277,
+      "learning_rate": 0.00018435339335774544,
+      "loss": 1.093,
+      "step": 1113
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.4788799004089455,
+      "learning_rate": 0.0001843213252664435,
+      "loss": 1.2955,
+      "step": 1114
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.4047292868271331,
+      "learning_rate": 0.0001842892271416797,
+      "loss": 1.3995,
+      "step": 1115
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.41973524907329757,
+      "learning_rate": 0.00018425709899488673,
+      "loss": 1.2488,
+      "step": 1116
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.4116525757280027,
+      "learning_rate": 0.0001842249408375079,
+      "loss": 1.2836,
+      "step": 1117
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.40381690409683124,
+      "learning_rate": 0.00018419275268099727,
+      "loss": 1.3076,
+      "step": 1118
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.3990670235924838,
+      "learning_rate": 0.0001841605345368196,
+      "loss": 1.1322,
+      "step": 1119
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.4557919985258954,
+      "learning_rate": 0.00018412828641645027,
+      "loss": 1.1492,
+      "step": 1120
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.41225245407592115,
+      "learning_rate": 0.0001840960083313754,
+      "loss": 1.2144,
+      "step": 1121
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.4063146367180345,
+      "learning_rate": 0.0001840637002930918,
+      "loss": 1.3407,
+      "step": 1122
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.47302251000360224,
+      "learning_rate": 0.00018403136231310684,
+      "loss": 1.2613,
+      "step": 1123
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.4180240178579445,
+      "learning_rate": 0.00018399899440293865,
+      "loss": 1.2414,
+      "step": 1124
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.40242975396276687,
+      "learning_rate": 0.00018396659657411596,
+      "loss": 1.2198,
+      "step": 1125
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.4384328789270641,
+      "learning_rate": 0.00018393416883817824,
+      "loss": 1.3307,
+      "step": 1126
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.4446552338798984,
+      "learning_rate": 0.00018390171120667554,
+      "loss": 1.1275,
+      "step": 1127
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.40315475459611616,
+      "learning_rate": 0.00018386922369116856,
+      "loss": 1.2653,
+      "step": 1128
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.4121559816577027,
+      "learning_rate": 0.00018383670630322865,
+      "loss": 1.0701,
+      "step": 1129
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.4759221235807963,
+      "learning_rate": 0.00018380415905443784,
+      "loss": 1.2562,
+      "step": 1130
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.42209843489184556,
+      "learning_rate": 0.00018377158195638876,
+      "loss": 1.1102,
+      "step": 1131
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.411750681581538,
+      "learning_rate": 0.0001837389750206847,
+      "loss": 1.2987,
+      "step": 1132
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.4242151835412917,
+      "learning_rate": 0.00018370633825893955,
+      "loss": 1.2213,
+      "step": 1133
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.398904531126636,
+      "learning_rate": 0.00018367367168277779,
+      "loss": 1.2161,
+      "step": 1134
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.42104630591741055,
+      "learning_rate": 0.00018364097530383457,
+      "loss": 1.1809,
+      "step": 1135
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.4123556643764555,
+      "learning_rate": 0.00018360824913375568,
+      "loss": 1.2281,
+      "step": 1136
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.4949580199502861,
+      "learning_rate": 0.00018357549318419745,
+      "loss": 1.1946,
+      "step": 1137
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.4591390076461925,
+      "learning_rate": 0.00018354270746682687,
+      "loss": 1.0818,
+      "step": 1138
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.40056269063498434,
+      "learning_rate": 0.00018350989199332154,
+      "loss": 1.2245,
+      "step": 1139
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5100942530665178,
+      "learning_rate": 0.00018347704677536955,
+      "loss": 1.214,
+      "step": 1140
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.41925713443588836,
+      "learning_rate": 0.00018344417182466978,
+      "loss": 1.0876,
+      "step": 1141
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.39377954997922326,
+      "learning_rate": 0.0001834112671529315,
+      "loss": 1.3542,
+      "step": 1142
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5622090630042867,
+      "learning_rate": 0.00018337833277187472,
+      "loss": 1.1644,
+      "step": 1143
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.4500214855875504,
+      "learning_rate": 0.00018334536869322988,
+      "loss": 1.3061,
+      "step": 1144
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.4248913197821054,
+      "learning_rate": 0.00018331237492873818,
+      "loss": 1.305,
+      "step": 1145
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.4982923848699508,
+      "learning_rate": 0.00018327935149015123,
+      "loss": 1.1275,
+      "step": 1146
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.45031144721481847,
+      "learning_rate": 0.00018324629838923132,
+      "loss": 1.0569,
+      "step": 1147
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.4215609294974685,
+      "learning_rate": 0.0001832132156377512,
+      "loss": 1.2469,
+      "step": 1148
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.3742716289678585,
+      "learning_rate": 0.0001831801032474943,
+      "loss": 1.2462,
+      "step": 1149
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.41531503609690806,
+      "learning_rate": 0.00018314696123025454,
+      "loss": 1.1542,
+      "step": 1150
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.4186757903194814,
+      "learning_rate": 0.00018311378959783638,
+      "loss": 1.1584,
+      "step": 1151
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.41465600790494256,
+      "learning_rate": 0.00018308058836205487,
+      "loss": 1.1717,
+      "step": 1152
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.3938088753883019,
+      "learning_rate": 0.00018304735753473555,
+      "loss": 1.3156,
+      "step": 1153
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.49816256566584854,
+      "learning_rate": 0.0001830140971277146,
+      "loss": 1.2332,
+      "step": 1154
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.4095457400363959,
+      "learning_rate": 0.00018298080715283858,
+      "loss": 1.0985,
+      "step": 1155
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.46217798400409227,
+      "learning_rate": 0.00018294748762196473,
+      "loss": 1.2585,
+      "step": 1156
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.4130725885720973,
+      "learning_rate": 0.00018291413854696077,
+      "loss": 1.2092,
+      "step": 1157
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5315586561171906,
+      "learning_rate": 0.0001828807599397049,
+      "loss": 1.177,
+      "step": 1158
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.4420905483291005,
+      "learning_rate": 0.00018284735181208587,
+      "loss": 1.1189,
+      "step": 1159
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.42456738295000523,
+      "learning_rate": 0.00018281391417600295,
+      "loss": 1.237,
+      "step": 1160
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.4042882497342129,
+      "learning_rate": 0.00018278044704336593,
+      "loss": 1.0848,
+      "step": 1161
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.47537010122245993,
+      "learning_rate": 0.00018274695042609507,
+      "loss": 1.1197,
+      "step": 1162
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.3964429766580857,
+      "learning_rate": 0.00018271342433612113,
+      "loss": 1.1493,
+      "step": 1163
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.4256464657525756,
+      "learning_rate": 0.00018267986878538545,
+      "loss": 1.1601,
+      "step": 1164
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.4238648190252956,
+      "learning_rate": 0.00018264628378583976,
+      "loss": 1.254,
+      "step": 1165
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.48948570342084713,
+      "learning_rate": 0.00018261266934944633,
+      "loss": 1.241,
+      "step": 1166
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.4339564630301822,
+      "learning_rate": 0.00018257902548817794,
+      "loss": 1.0665,
+      "step": 1167
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.4732440275999781,
+      "learning_rate": 0.00018254535221401774,
+      "loss": 1.1097,
+      "step": 1168
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.4221240516519539,
+      "learning_rate": 0.0001825116495389595,
+      "loss": 1.0841,
+      "step": 1169
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.46742910281062844,
+      "learning_rate": 0.0001824779174750074,
+      "loss": 1.357,
+      "step": 1170
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5826398920812126,
+      "learning_rate": 0.00018244415603417603,
+      "loss": 1.0945,
+      "step": 1171
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.456156025948686,
+      "learning_rate": 0.00018241036522849055,
+      "loss": 1.1704,
+      "step": 1172
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.43243676067061454,
+      "learning_rate": 0.0001823765450699865,
+      "loss": 1.3262,
+      "step": 1173
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.45964459915248357,
+      "learning_rate": 0.0001823426955707099,
+      "loss": 1.2436,
+      "step": 1174
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.4686291931164835,
+      "learning_rate": 0.00018230881674271722,
+      "loss": 1.2149,
+      "step": 1175
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.4550569260570272,
+      "learning_rate": 0.00018227490859807542,
+      "loss": 1.2923,
+      "step": 1176
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.4421930823641351,
+      "learning_rate": 0.00018224097114886178,
+      "loss": 1.077,
+      "step": 1177
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.4268716215468276,
+      "learning_rate": 0.00018220700440716413,
+      "loss": 1.1334,
+      "step": 1178
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.430246951285977,
+      "learning_rate": 0.00018217300838508073,
+      "loss": 1.1971,
+      "step": 1179
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.45327376123189744,
+      "learning_rate": 0.00018213898309472025,
+      "loss": 1.2132,
+      "step": 1180
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.41095940475376064,
+      "learning_rate": 0.00018210492854820172,
+      "loss": 1.3556,
+      "step": 1181
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.45162325821886706,
+      "learning_rate": 0.00018207084475765467,
+      "loss": 1.1719,
+      "step": 1182
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.4216299758646101,
+      "learning_rate": 0.00018203673173521902,
+      "loss": 1.1182,
+      "step": 1183
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.4161348281065941,
+      "learning_rate": 0.0001820025894930451,
+      "loss": 1.29,
+      "step": 1184
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.40511834472517844,
+      "learning_rate": 0.00018196841804329368,
+      "loss": 1.2903,
+      "step": 1185
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.40053002203522686,
+      "learning_rate": 0.00018193421739813587,
+      "loss": 1.2405,
+      "step": 1186
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5646167304390911,
+      "learning_rate": 0.00018189998756975318,
+      "loss": 1.0068,
+      "step": 1187
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.3625028518057489,
+      "learning_rate": 0.0001818657285703376,
+      "loss": 1.2454,
+      "step": 1188
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.49040040773503507,
+      "learning_rate": 0.00018183144041209145,
+      "loss": 1.2666,
+      "step": 1189
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.4947973599264683,
+      "learning_rate": 0.00018179712310722746,
+      "loss": 1.3164,
+      "step": 1190
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.42978754507702566,
+      "learning_rate": 0.00018176277666796868,
+      "loss": 1.149,
+      "step": 1191
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.42841558640935595,
+      "learning_rate": 0.00018172840110654862,
+      "loss": 1.1945,
+      "step": 1192
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.42992496080457954,
+      "learning_rate": 0.00018169399643521109,
+      "loss": 1.1944,
+      "step": 1193
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.448105855134444,
+      "learning_rate": 0.00018165956266621029,
+      "loss": 1.2494,
+      "step": 1194
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.4728277868878495,
+      "learning_rate": 0.00018162509981181084,
+      "loss": 1.1506,
+      "step": 1195
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.41918719903517565,
+      "learning_rate": 0.00018159060788428767,
+      "loss": 1.2811,
+      "step": 1196
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.39228291969671497,
+      "learning_rate": 0.00018155608689592604,
+      "loss": 1.2829,
+      "step": 1197
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.4142401376356667,
+      "learning_rate": 0.0001815215368590216,
+      "loss": 1.1886,
+      "step": 1198
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.37803247852227023,
+      "learning_rate": 0.00018148695778588033,
+      "loss": 1.3232,
+      "step": 1199
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.4593573808336638,
+      "learning_rate": 0.00018145234968881864,
+      "loss": 1.1287,
+      "step": 1200
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.41914279455811904,
+      "learning_rate": 0.0001814177125801631,
+      "loss": 1.2874,
+      "step": 1201
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.42881561588356065,
+      "learning_rate": 0.00018138304647225074,
+      "loss": 1.2801,
+      "step": 1202
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.37442527040488294,
+      "learning_rate": 0.0001813483513774289,
+      "loss": 1.2932,
+      "step": 1203
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.46633238282251765,
+      "learning_rate": 0.00018131362730805522,
+      "loss": 1.1351,
+      "step": 1204
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.38757351332071277,
+      "learning_rate": 0.0001812788742764977,
+      "loss": 1.3596,
+      "step": 1205
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5991737061535322,
+      "learning_rate": 0.00018124409229513462,
+      "loss": 1.2759,
+      "step": 1206
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.434806548891723,
+      "learning_rate": 0.00018120928137635455,
+      "loss": 1.2969,
+      "step": 1207
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.46248716487193253,
+      "learning_rate": 0.00018117444153255646,
+      "loss": 1.1335,
+      "step": 1208
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.4512988261214917,
+      "learning_rate": 0.00018113957277614948,
+      "loss": 1.215,
+      "step": 1209
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.44625024528226515,
+      "learning_rate": 0.0001811046751195532,
+      "loss": 1.1304,
+      "step": 1210
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.376633059190578,
+      "learning_rate": 0.00018106974857519736,
+      "loss": 1.309,
+      "step": 1211
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.4163438690239004,
+      "learning_rate": 0.0001810347931555221,
+      "loss": 1.1714,
+      "step": 1212
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.3996382338762611,
+      "learning_rate": 0.00018099980887297781,
+      "loss": 1.2697,
+      "step": 1213
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.3985907787300379,
+      "learning_rate": 0.00018096479574002507,
+      "loss": 1.1841,
+      "step": 1214
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.40910341898652425,
+      "learning_rate": 0.0001809297537691349,
+      "loss": 1.1133,
+      "step": 1215
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.4410603927218668,
+      "learning_rate": 0.0001808946829727884,
+      "loss": 1.3338,
+      "step": 1216
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.4011905251588421,
+      "learning_rate": 0.00018085958336347714,
+      "loss": 1.2269,
+      "step": 1217
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.4271772428535343,
+      "learning_rate": 0.0001808244549537028,
+      "loss": 1.1486,
+      "step": 1218
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.42067201560284084,
+      "learning_rate": 0.0001807892977559774,
+      "loss": 1.3273,
+      "step": 1219
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.42417696072732974,
+      "learning_rate": 0.0001807541117828232,
+      "loss": 1.0487,
+      "step": 1220
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.4553360226082746,
+      "learning_rate": 0.0001807188970467726,
+      "loss": 1.2139,
+      "step": 1221
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.4509518827150816,
+      "learning_rate": 0.00018068365356036848,
+      "loss": 1.1455,
+      "step": 1222
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.41037562325844107,
+      "learning_rate": 0.00018064838133616368,
+      "loss": 1.0707,
+      "step": 1223
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.4328301887261335,
+      "learning_rate": 0.00018061308038672154,
+      "loss": 1.2218,
+      "step": 1224
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5327423666842781,
+      "learning_rate": 0.00018057775072461544,
+      "loss": 1.2592,
+      "step": 1225
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.4231582879401159,
+      "learning_rate": 0.00018054239236242907,
+      "loss": 1.1313,
+      "step": 1226
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.41588921347956037,
+      "learning_rate": 0.0001805070053127563,
+      "loss": 1.093,
+      "step": 1227
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.43874494046260254,
+      "learning_rate": 0.0001804715895882013,
+      "loss": 1.2148,
+      "step": 1228
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.4883827870448002,
+      "learning_rate": 0.00018043614520137834,
+      "loss": 1.22,
+      "step": 1229
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.4038827748512788,
+      "learning_rate": 0.00018040067216491197,
+      "loss": 1.3178,
+      "step": 1230
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.39772086476345087,
+      "learning_rate": 0.00018036517049143695,
+      "loss": 1.3108,
+      "step": 1231
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.3945985733483299,
+      "learning_rate": 0.0001803296401935982,
+      "loss": 1.1392,
+      "step": 1232
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.4393133639731285,
+      "learning_rate": 0.0001802940812840509,
+      "loss": 1.2437,
+      "step": 1233
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.4111622662911384,
+      "learning_rate": 0.00018025849377546033,
+      "loss": 1.2885,
+      "step": 1234
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.39692478917651064,
+      "learning_rate": 0.00018022287768050202,
+      "loss": 1.2155,
+      "step": 1235
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.4993165753923811,
+      "learning_rate": 0.00018018723301186165,
+      "loss": 1.1274,
+      "step": 1236
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.39598467074388766,
+      "learning_rate": 0.0001801515597822351,
+      "loss": 1.2387,
+      "step": 1237
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.4753244202766253,
+      "learning_rate": 0.00018011585800432845,
+      "loss": 0.9181,
+      "step": 1238
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.4604877872165509,
+      "learning_rate": 0.00018008012769085786,
+      "loss": 1.2002,
+      "step": 1239
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.4419244388876236,
+      "learning_rate": 0.00018004436885454977,
+      "loss": 1.1422,
+      "step": 1240
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.36249685380467883,
+      "learning_rate": 0.0001800085815081406,
+      "loss": 1.3193,
+      "step": 1241
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.48353467585395604,
+      "learning_rate": 0.00017997276566437717,
+      "loss": 1.1576,
+      "step": 1242
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.48222314430472435,
+      "learning_rate": 0.0001799369213360163,
+      "loss": 1.2074,
+      "step": 1243
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.46201113227446217,
+      "learning_rate": 0.00017990104853582493,
+      "loss": 1.0978,
+      "step": 1244
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.38399906415997925,
+      "learning_rate": 0.00017986514727658017,
+      "loss": 1.1414,
+      "step": 1245
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.42000876882202476,
+      "learning_rate": 0.00017982921757106937,
+      "loss": 1.309,
+      "step": 1246
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5162603909101442,
+      "learning_rate": 0.00017979325943208983,
+      "loss": 1.1046,
+      "step": 1247
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.42766767931389604,
+      "learning_rate": 0.00017975727287244914,
+      "loss": 1.1239,
+      "step": 1248
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.429471035049448,
+      "learning_rate": 0.00017972125790496497,
+      "loss": 1.0864,
+      "step": 1249
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.41894675808843357,
+      "learning_rate": 0.000179685214542465,
+      "loss": 1.252,
+      "step": 1250
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.4547943378871711,
+      "learning_rate": 0.00017964914279778715,
+      "loss": 1.3211,
+      "step": 1251
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.43197826022728447,
+      "learning_rate": 0.0001796130426837794,
+      "loss": 1.192,
+      "step": 1252
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5027174006166188,
+      "learning_rate": 0.00017957691421329987,
+      "loss": 1.1496,
+      "step": 1253
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.4026521262618264,
+      "learning_rate": 0.00017954075739921675,
+      "loss": 1.1899,
+      "step": 1254
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.49102208363638583,
+      "learning_rate": 0.0001795045722544083,
+      "loss": 1.0389,
+      "step": 1255
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.4381067192455368,
+      "learning_rate": 0.00017946835879176294,
+      "loss": 1.142,
+      "step": 1256
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.4017584020768178,
+      "learning_rate": 0.00017943211702417907,
+      "loss": 1.2828,
+      "step": 1257
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.44696509442243215,
+      "learning_rate": 0.00017939584696456527,
+      "loss": 1.3211,
+      "step": 1258
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.4920621549656613,
+      "learning_rate": 0.00017935954862584018,
+      "loss": 1.195,
+      "step": 1259
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5173019275230595,
+      "learning_rate": 0.00017932322202093249,
+      "loss": 1.1377,
+      "step": 1260
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.3965120698576586,
+      "learning_rate": 0.00017928686716278094,
+      "loss": 1.0479,
+      "step": 1261
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.47394505428050265,
+      "learning_rate": 0.00017925048406433437,
+      "loss": 1.2275,
+      "step": 1262
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.4121814819366902,
+      "learning_rate": 0.0001792140727385517,
+      "loss": 1.2047,
+      "step": 1263
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.4412604382282973,
+      "learning_rate": 0.00017917763319840182,
+      "loss": 1.3598,
+      "step": 1264
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.4367664670404755,
+      "learning_rate": 0.00017914116545686372,
+      "loss": 1.1181,
+      "step": 1265
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.474048022963849,
+      "learning_rate": 0.00017910466952692647,
+      "loss": 1.1504,
+      "step": 1266
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.4230406554548749,
+      "learning_rate": 0.0001790681454215891,
+      "loss": 1.3451,
+      "step": 1267
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.4243244396552428,
+      "learning_rate": 0.0001790315931538608,
+      "loss": 1.2807,
+      "step": 1268
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.45812037441187453,
+      "learning_rate": 0.0001789950127367606,
+      "loss": 1.4095,
+      "step": 1269
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.47464216791436775,
+      "learning_rate": 0.00017895840418331773,
+      "loss": 1.3497,
+      "step": 1270
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.45397616196690266,
+      "learning_rate": 0.00017892176750657136,
+      "loss": 1.2593,
+      "step": 1271
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.49518574530554793,
+      "learning_rate": 0.00017888510271957074,
+      "loss": 1.2118,
+      "step": 1272
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.44039293470940555,
+      "learning_rate": 0.000178848409835375,
+      "loss": 1.315,
+      "step": 1273
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.39872310323947713,
+      "learning_rate": 0.00017881168886705343,
+      "loss": 1.1966,
+      "step": 1274
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.43931973722879336,
+      "learning_rate": 0.00017877493982768527,
+      "loss": 1.2787,
+      "step": 1275
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.4650345100118597,
+      "learning_rate": 0.0001787381627303597,
+      "loss": 1.1922,
+      "step": 1276
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.3869719260610495,
+      "learning_rate": 0.00017870135758817592,
+      "loss": 1.4417,
+      "step": 1277
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.4732577506979918,
+      "learning_rate": 0.00017866452441424321,
+      "loss": 1.0814,
+      "step": 1278
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5144845942032602,
+      "learning_rate": 0.00017862766322168075,
+      "loss": 1.1123,
+      "step": 1279
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5371059198671081,
+      "learning_rate": 0.00017859077402361764,
+      "loss": 1.2102,
+      "step": 1280
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.4380168788255874,
+      "learning_rate": 0.0001785538568331931,
+      "loss": 1.2085,
+      "step": 1281
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.48802505533986273,
+      "learning_rate": 0.00017851691166355623,
+      "loss": 1.1819,
+      "step": 1282
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.3985823932154251,
+      "learning_rate": 0.0001784799385278661,
+      "loss": 1.2485,
+      "step": 1283
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.39904750361042457,
+      "learning_rate": 0.00017844293743929176,
+      "loss": 1.1649,
+      "step": 1284
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.3691059581632548,
+      "learning_rate": 0.00017840590841101224,
+      "loss": 1.3318,
+      "step": 1285
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.4238313144139466,
+      "learning_rate": 0.00017836885145621644,
+      "loss": 1.2584,
+      "step": 1286
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.43167565936921465,
+      "learning_rate": 0.0001783317665881033,
+      "loss": 1.2208,
+      "step": 1287
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.38868136252675795,
+      "learning_rate": 0.00017829465381988163,
+      "loss": 1.0485,
+      "step": 1288
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.45023553000967126,
+      "learning_rate": 0.00017825751316477022,
+      "loss": 1.1818,
+      "step": 1289
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.3974940981594699,
+      "learning_rate": 0.00017822034463599778,
+      "loss": 1.2265,
+      "step": 1290
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.4244087850769084,
+      "learning_rate": 0.000178183148246803,
+      "loss": 1.3336,
+      "step": 1291
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.43893067650040507,
+      "learning_rate": 0.00017814592401043435,
+      "loss": 1.2674,
+      "step": 1292
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.4031995220731051,
+      "learning_rate": 0.00017810867194015037,
+      "loss": 1.3047,
+      "step": 1293
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.48071709414537084,
+      "learning_rate": 0.00017807139204921945,
+      "loss": 1.3746,
+      "step": 1294
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.4568520829173446,
+      "learning_rate": 0.0001780340843509199,
+      "loss": 1.1363,
+      "step": 1295
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.4321334225941639,
+      "learning_rate": 0.00017799674885853988,
+      "loss": 1.3084,
+      "step": 1296
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.4550197787816908,
+      "learning_rate": 0.00017795938558537753,
+      "loss": 1.206,
+      "step": 1297
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.4097389064901465,
+      "learning_rate": 0.00017792199454474088,
+      "loss": 1.2204,
+      "step": 1298
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.4403403565118469,
+      "learning_rate": 0.00017788457574994778,
+      "loss": 1.3638,
+      "step": 1299
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.4505973773085649,
+      "learning_rate": 0.00017784712921432603,
+      "loss": 1.1258,
+      "step": 1300
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.46966015979190584,
+      "learning_rate": 0.0001778096549512133,
+      "loss": 1.3737,
+      "step": 1301
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.41450802682782734,
+      "learning_rate": 0.00017777215297395707,
+      "loss": 1.1742,
+      "step": 1302
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.411071402317532,
+      "learning_rate": 0.0001777346232959148,
+      "loss": 1.2982,
+      "step": 1303
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.4353939993343516,
+      "learning_rate": 0.0001776970659304538,
+      "loss": 1.1929,
+      "step": 1304
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.4080042265359988,
+      "learning_rate": 0.0001776594808909511,
+      "loss": 1.335,
+      "step": 1305
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.4231804204889362,
+      "learning_rate": 0.0001776218681907938,
+      "loss": 1.1941,
+      "step": 1306
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.42545318614282507,
+      "learning_rate": 0.00017758422784337863,
+      "loss": 1.3336,
+      "step": 1307
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.3625246564679697,
+      "learning_rate": 0.00017754655986211234,
+      "loss": 1.3342,
+      "step": 1308
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.39835235021926696,
+      "learning_rate": 0.00017750886426041152,
+      "loss": 1.1597,
+      "step": 1309
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.40952465702340723,
+      "learning_rate": 0.00017747114105170245,
+      "loss": 1.167,
+      "step": 1310
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.480453202267986,
+      "learning_rate": 0.00017743339024942135,
+      "loss": 1.2675,
+      "step": 1311
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.49319536071787506,
+      "learning_rate": 0.0001773956118670143,
+      "loss": 1.1149,
+      "step": 1312
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.4508699939400191,
+      "learning_rate": 0.00017735780591793712,
+      "loss": 1.1905,
+      "step": 1313
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.4162008842345468,
+      "learning_rate": 0.00017731997241565549,
+      "loss": 1.1461,
+      "step": 1314
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.4188090746661949,
+      "learning_rate": 0.00017728211137364489,
+      "loss": 1.1749,
+      "step": 1315
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.41250624727220514,
+      "learning_rate": 0.00017724422280539063,
+      "loss": 1.2234,
+      "step": 1316
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.39482290620152793,
+      "learning_rate": 0.00017720630672438782,
+      "loss": 1.2581,
+      "step": 1317
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.46401916311183944,
+      "learning_rate": 0.00017716836314414137,
+      "loss": 1.1181,
+      "step": 1318
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.39762195046650667,
+      "learning_rate": 0.00017713039207816592,
+      "loss": 1.1115,
+      "step": 1319
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.4789061290452024,
+      "learning_rate": 0.000177092393539986,
+      "loss": 1.2181,
+      "step": 1320
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.44639857628897034,
+      "learning_rate": 0.0001770543675431359,
+      "loss": 1.2632,
+      "step": 1321
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.463003242514538,
+      "learning_rate": 0.0001770163141011596,
+      "loss": 1.2769,
+      "step": 1322
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.4325152996180038,
+      "learning_rate": 0.000176978233227611,
+      "loss": 1.1002,
+      "step": 1323
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.40771813550482583,
+      "learning_rate": 0.0001769401249360537,
+      "loss": 1.2375,
+      "step": 1324
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.37503743077506974,
+      "learning_rate": 0.00017690198924006097,
+      "loss": 1.2946,
+      "step": 1325
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.4231465718176906,
+      "learning_rate": 0.00017686382615321602,
+      "loss": 1.4451,
+      "step": 1326
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.44151100325486126,
+      "learning_rate": 0.0001768256356891117,
+      "loss": 1.2087,
+      "step": 1327
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.42033552671983576,
+      "learning_rate": 0.00017678741786135066,
+      "loss": 1.2381,
+      "step": 1328
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.4856774111071269,
+      "learning_rate": 0.00017674917268354522,
+      "loss": 1.3514,
+      "step": 1329
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.42956438092798455,
+      "learning_rate": 0.0001767109001693176,
+      "loss": 1.2389,
+      "step": 1330
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.4270755785641226,
+      "learning_rate": 0.00017667260033229953,
+      "loss": 1.2901,
+      "step": 1331
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.4804671655583362,
+      "learning_rate": 0.0001766342731861327,
+      "loss": 1.01,
+      "step": 1332
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.3699244125572967,
+      "learning_rate": 0.00017659591874446837,
+      "loss": 1.1338,
+      "step": 1333
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.42677060147908324,
+      "learning_rate": 0.0001765575370209676,
+      "loss": 1.2598,
+      "step": 1334
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.3992151745086556,
+      "learning_rate": 0.00017651912802930112,
+      "loss": 1.1779,
+      "step": 1335
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.41974467932074777,
+      "learning_rate": 0.00017648069178314945,
+      "loss": 1.1699,
+      "step": 1336
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.41050542448430405,
+      "learning_rate": 0.0001764422282962027,
+      "loss": 1.3138,
+      "step": 1337
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.42814486170509386,
+      "learning_rate": 0.00017640373758216077,
+      "loss": 1.2355,
+      "step": 1338
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.3969221922153332,
+      "learning_rate": 0.00017636521965473323,
+      "loss": 1.2794,
+      "step": 1339
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5624062673016748,
+      "learning_rate": 0.00017632667452763934,
+      "loss": 0.9789,
+      "step": 1340
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.4568378334062782,
+      "learning_rate": 0.00017628810221460806,
+      "loss": 1.2391,
+      "step": 1341
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.4803968878451528,
+      "learning_rate": 0.00017624950272937802,
+      "loss": 1.1786,
+      "step": 1342
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.41543625145694396,
+      "learning_rate": 0.00017621087608569758,
+      "loss": 1.341,
+      "step": 1343
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.45584697368082777,
+      "learning_rate": 0.0001761722222973247,
+      "loss": 1.1908,
+      "step": 1344
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.4768605878290502,
+      "learning_rate": 0.000176133541378027,
+      "loss": 1.1202,
+      "step": 1345
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.4816297214397447,
+      "learning_rate": 0.00017609483334158181,
+      "loss": 1.1387,
+      "step": 1346
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.4184125766217683,
+      "learning_rate": 0.00017605609820177617,
+      "loss": 1.1647,
+      "step": 1347
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.41902706713355775,
+      "learning_rate": 0.00017601733597240665,
+      "loss": 1.272,
+      "step": 1348
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.46125109529213654,
+      "learning_rate": 0.0001759785466672796,
+      "loss": 1.1571,
+      "step": 1349
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.400321834755756,
+      "learning_rate": 0.00017593973030021086,
+      "loss": 1.118,
+      "step": 1350
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.43435562483445556,
+      "learning_rate": 0.00017590088688502605,
+      "loss": 1.1604,
+      "step": 1351
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.4003974050289482,
+      "learning_rate": 0.00017586201643556031,
+      "loss": 1.0377,
+      "step": 1352
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.41511694277843386,
+      "learning_rate": 0.00017582311896565856,
+      "loss": 1.1772,
+      "step": 1353
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.4581994472826869,
+      "learning_rate": 0.00017578419448917515,
+      "loss": 1.2006,
+      "step": 1354
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.4133992577469832,
+      "learning_rate": 0.00017574524301997423,
+      "loss": 1.1392,
+      "step": 1355
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.4461518844008597,
+      "learning_rate": 0.00017570626457192944,
+      "loss": 1.0983,
+      "step": 1356
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.3847901278287183,
+      "learning_rate": 0.00017566725915892412,
+      "loss": 1.0782,
+      "step": 1357
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.4122252189639809,
+      "learning_rate": 0.00017562822679485112,
+      "loss": 1.2497,
+      "step": 1358
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.4332405750572694,
+      "learning_rate": 0.00017558916749361295,
+      "loss": 1.1334,
+      "step": 1359
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5076728926724542,
+      "learning_rate": 0.00017555008126912168,
+      "loss": 1.1704,
+      "step": 1360
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.4291147466537929,
+      "learning_rate": 0.00017551096813529907,
+      "loss": 1.1499,
+      "step": 1361
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.44186669586631894,
+      "learning_rate": 0.00017547182810607627,
+      "loss": 1.2446,
+      "step": 1362
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.489954487115797,
+      "learning_rate": 0.00017543266119539422,
+      "loss": 1.1295,
+      "step": 1363
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.4306803716378695,
+      "learning_rate": 0.00017539346741720334,
+      "loss": 1.2107,
+      "step": 1364
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.4017955570241805,
+      "learning_rate": 0.00017535424678546355,
+      "loss": 1.0796,
+      "step": 1365
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.44109586053985445,
+      "learning_rate": 0.00017531499931414444,
+      "loss": 1.2766,
+      "step": 1366
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.45792639977455124,
+      "learning_rate": 0.00017527572501722512,
+      "loss": 1.0823,
+      "step": 1367
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.4420003098431387,
+      "learning_rate": 0.0001752364239086943,
+      "loss": 1.272,
+      "step": 1368
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.4446789632732446,
+      "learning_rate": 0.00017519709600255016,
+      "loss": 1.2099,
+      "step": 1369
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.404870262318969,
+      "learning_rate": 0.00017515774131280046,
+      "loss": 1.1891,
+      "step": 1370
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.44582392834683526,
+      "learning_rate": 0.00017511835985346253,
+      "loss": 1.1872,
+      "step": 1371
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.43043023365784744,
+      "learning_rate": 0.00017507895163856315,
+      "loss": 1.2376,
+      "step": 1372
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.45969399092749785,
+      "learning_rate": 0.00017503951668213874,
+      "loss": 1.1769,
+      "step": 1373
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.46233958528613395,
+      "learning_rate": 0.00017500005499823519,
+      "loss": 0.9474,
+      "step": 1374
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.45996785498628634,
+      "learning_rate": 0.0001749605666009079,
+      "loss": 1.1489,
+      "step": 1375
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.42200111262426815,
+      "learning_rate": 0.0001749210515042218,
+      "loss": 1.0905,
+      "step": 1376
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.4301491224431233,
+      "learning_rate": 0.0001748815097222513,
+      "loss": 1.267,
+      "step": 1377
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.46052245667906017,
+      "learning_rate": 0.00017484194126908038,
+      "loss": 1.2559,
+      "step": 1378
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.4177092463850814,
+      "learning_rate": 0.00017480234615880247,
+      "loss": 1.3717,
+      "step": 1379
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.4330888591953408,
+      "learning_rate": 0.0001747627244055205,
+      "loss": 1.2326,
+      "step": 1380
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.47822140492367105,
+      "learning_rate": 0.00017472307602334693,
+      "loss": 1.3168,
+      "step": 1381
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.4612910088564453,
+      "learning_rate": 0.0001746834010264036,
+      "loss": 1.3653,
+      "step": 1382
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.4033454346780331,
+      "learning_rate": 0.00017464369942882195,
+      "loss": 1.121,
+      "step": 1383
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.4353683920939397,
+      "learning_rate": 0.00017460397124474284,
+      "loss": 1.1793,
+      "step": 1384
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.39394033909027043,
+      "learning_rate": 0.00017456421648831655,
+      "loss": 1.1768,
+      "step": 1385
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.45969084270950467,
+      "learning_rate": 0.000174524435173703,
+      "loss": 1.138,
+      "step": 1386
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.43072496433259433,
+      "learning_rate": 0.0001744846273150713,
+      "loss": 1.197,
+      "step": 1387
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5398931763192616,
+      "learning_rate": 0.0001744447929266003,
+      "loss": 1.105,
+      "step": 1388
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.4259320309321128,
+      "learning_rate": 0.00017440493202247804,
+      "loss": 1.1209,
+      "step": 1389
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.4230551270784258,
+      "learning_rate": 0.00017436504461690222,
+      "loss": 1.2619,
+      "step": 1390
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.4068621741738843,
+      "learning_rate": 0.00017432513072407983,
+      "loss": 1.1703,
+      "step": 1391
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.46960647890417656,
+      "learning_rate": 0.00017428519035822738,
+      "loss": 1.0698,
+      "step": 1392
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.37770220741768,
+      "learning_rate": 0.00017424522353357076,
+      "loss": 1.1268,
+      "step": 1393
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.4126852889543039,
+      "learning_rate": 0.0001742052302643453,
+      "loss": 1.3859,
+      "step": 1394
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.4919598766857543,
+      "learning_rate": 0.00017416521056479577,
+      "loss": 1.1872,
+      "step": 1395
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.38563468114108085,
+      "learning_rate": 0.0001741251644491763,
+      "loss": 1.2439,
+      "step": 1396
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5123894936406493,
+      "learning_rate": 0.00017408509193175053,
+      "loss": 1.3575,
+      "step": 1397
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.43818295767322485,
+      "learning_rate": 0.00017404499302679137,
+      "loss": 1.0269,
+      "step": 1398
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.44484172606886296,
+      "learning_rate": 0.00017400486774858122,
+      "loss": 1.1107,
+      "step": 1399
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.44629815867772044,
+      "learning_rate": 0.00017396471611141187,
+      "loss": 1.2142,
+      "step": 1400
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.4273309166076388,
+      "learning_rate": 0.00017392453812958446,
+      "loss": 1.1813,
+      "step": 1401
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.4574035409129155,
+      "learning_rate": 0.00017388433381740952,
+      "loss": 1.261,
+      "step": 1402
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.4151070194117886,
+      "learning_rate": 0.00017384410318920697,
+      "loss": 1.3402,
+      "step": 1403
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.41827092853043507,
+      "learning_rate": 0.00017380384625930616,
+      "loss": 1.2797,
+      "step": 1404
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.42472019314914217,
+      "learning_rate": 0.0001737635630420457,
+      "loss": 1.3179,
+      "step": 1405
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.3982779339377946,
+      "learning_rate": 0.0001737232535517736,
+      "loss": 1.0811,
+      "step": 1406
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.41011418033778524,
+      "learning_rate": 0.00017368291780284732,
+      "loss": 1.2314,
+      "step": 1407
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.3959871750615605,
+      "learning_rate": 0.0001736425558096335,
+      "loss": 1.29,
+      "step": 1408
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.4058221751795928,
+      "learning_rate": 0.00017360216758650828,
+      "loss": 1.1652,
+      "step": 1409
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.43317389684132396,
+      "learning_rate": 0.00017356175314785707,
+      "loss": 1.2132,
+      "step": 1410
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.40083987628369755,
+      "learning_rate": 0.00017352131250807467,
+      "loss": 1.3991,
+      "step": 1411
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.4324216937575786,
+      "learning_rate": 0.00017348084568156509,
+      "loss": 1.134,
+      "step": 1412
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.4366824318174888,
+      "learning_rate": 0.0001734403526827418,
+      "loss": 1.0184,
+      "step": 1413
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.46925682428689836,
+      "learning_rate": 0.0001733998335260276,
+      "loss": 1.2034,
+      "step": 1414
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.44456072365634625,
+      "learning_rate": 0.00017335928822585447,
+      "loss": 0.9823,
+      "step": 1415
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.4199291381502753,
+      "learning_rate": 0.0001733187167966638,
+      "loss": 1.0838,
+      "step": 1416
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.47608650868026064,
+      "learning_rate": 0.00017327811925290631,
+      "loss": 1.1656,
+      "step": 1417
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.4591654098678873,
+      "learning_rate": 0.00017323749560904195,
+      "loss": 1.1442,
+      "step": 1418
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.580593987172281,
+      "learning_rate": 0.00017319684587954002,
+      "loss": 1.3153,
+      "step": 1419
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.4508260154528983,
+      "learning_rate": 0.00017315617007887903,
+      "loss": 1.2017,
+      "step": 1420
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.40221739953591606,
+      "learning_rate": 0.00017311546822154687,
+      "loss": 1.0833,
+      "step": 1421
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.38774769234817863,
+      "learning_rate": 0.0001730747403220407,
+      "loss": 1.2375,
+      "step": 1422
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5123072672050475,
+      "learning_rate": 0.00017303398639486695,
+      "loss": 1.2682,
+      "step": 1423
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.44713753820209695,
+      "learning_rate": 0.00017299320645454124,
+      "loss": 1.1936,
+      "step": 1424
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.4270057935393677,
+      "learning_rate": 0.00017295240051558853,
+      "loss": 1.2809,
+      "step": 1425
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.3901380441058319,
+      "learning_rate": 0.00017291156859254302,
+      "loss": 1.4071,
+      "step": 1426
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.4379046604463225,
+      "learning_rate": 0.0001728707106999482,
+      "loss": 1.3157,
+      "step": 1427
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.475336899869378,
+      "learning_rate": 0.0001728298268523568,
+      "loss": 1.305,
+      "step": 1428
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.4571307552954528,
+      "learning_rate": 0.00017278891706433072,
+      "loss": 1.1536,
+      "step": 1429
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.46422276152858183,
+      "learning_rate": 0.00017274798135044118,
+      "loss": 1.2209,
+      "step": 1430
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.41622993623754945,
+      "learning_rate": 0.00017270701972526862,
+      "loss": 1.1115,
+      "step": 1431
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.41429807189638534,
+      "learning_rate": 0.0001726660322034027,
+      "loss": 1.1648,
+      "step": 1432
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.4109899505661029,
+      "learning_rate": 0.00017262501879944232,
+      "loss": 1.1914,
+      "step": 1433
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.4803163502506599,
+      "learning_rate": 0.00017258397952799555,
+      "loss": 1.1419,
+      "step": 1434
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.393871786207737,
+      "learning_rate": 0.00017254291440367968,
+      "loss": 1.2145,
+      "step": 1435
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.44362878824021,
+      "learning_rate": 0.0001725018234411213,
+      "loss": 1.0244,
+      "step": 1436
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.3857050191839158,
+      "learning_rate": 0.00017246070665495608,
+      "loss": 1.1649,
+      "step": 1437
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.40006530808159074,
+      "learning_rate": 0.000172419564059829,
+      "loss": 1.2251,
+      "step": 1438
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.41214860935545733,
+      "learning_rate": 0.00017237839567039412,
+      "loss": 1.1944,
+      "step": 1439
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.41193669218480433,
+      "learning_rate": 0.0001723372015013148,
+      "loss": 1.2077,
+      "step": 1440
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.42732367206429334,
+      "learning_rate": 0.00017229598156726347,
+      "loss": 1.2421,
+      "step": 1441
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.4439261335533788,
+      "learning_rate": 0.0001722547358829218,
+      "loss": 1.2002,
+      "step": 1442
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.45726967639166616,
+      "learning_rate": 0.0001722134644629807,
+      "loss": 1.0182,
+      "step": 1443
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.3990603042933494,
+      "learning_rate": 0.0001721721673221401,
+      "loss": 1.2858,
+      "step": 1444
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.4485693350317979,
+      "learning_rate": 0.00017213084447510916,
+      "loss": 1.1839,
+      "step": 1445
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.37143426464099893,
+      "learning_rate": 0.0001720894959366062,
+      "loss": 1.1599,
+      "step": 1446
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.4241777748086523,
+      "learning_rate": 0.00017204812172135874,
+      "loss": 0.9463,
+      "step": 1447
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.4041420400524161,
+      "learning_rate": 0.00017200672184410336,
+      "loss": 1.1281,
+      "step": 1448
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.3805060427896397,
+      "learning_rate": 0.0001719652963195858,
+      "loss": 1.3332,
+      "step": 1449
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.4357080134589215,
+      "learning_rate": 0.00017192384516256102,
+      "loss": 1.119,
+      "step": 1450
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.42822055835443434,
+      "learning_rate": 0.00017188236838779295,
+      "loss": 1.254,
+      "step": 1451
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.4810456664809949,
+      "learning_rate": 0.00017184086601005477,
+      "loss": 1.1589,
+      "step": 1452
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.4106107312743356,
+      "learning_rate": 0.00017179933804412875,
+      "loss": 1.1528,
+      "step": 1453
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.4573773707296973,
+      "learning_rate": 0.00017175778450480625,
+      "loss": 1.1201,
+      "step": 1454
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5472140731281081,
+      "learning_rate": 0.00017171620540688782,
+      "loss": 1.1492,
+      "step": 1455
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.39905562250353765,
+      "learning_rate": 0.00017167460076518299,
+      "loss": 1.1433,
+      "step": 1456
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5081997801804614,
+      "learning_rate": 0.00017163297059451043,
+      "loss": 1.2511,
+      "step": 1457
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.3891125167478373,
+      "learning_rate": 0.00017159131490969797,
+      "loss": 1.1648,
+      "step": 1458
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.44121510189146385,
+      "learning_rate": 0.00017154963372558246,
+      "loss": 1.1972,
+      "step": 1459
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5144140499243485,
+      "learning_rate": 0.00017150792705700982,
+      "loss": 1.1701,
+      "step": 1460
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.4160541943186721,
+      "learning_rate": 0.00017146619491883514,
+      "loss": 1.118,
+      "step": 1461
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.42059503975431084,
+      "learning_rate": 0.00017142443732592248,
+      "loss": 1.1062,
+      "step": 1462
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.4597983505858911,
+      "learning_rate": 0.000171382654293145,
+      "loss": 0.968,
+      "step": 1463
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.4102737404682553,
+      "learning_rate": 0.000171340845835385,
+      "loss": 1.0266,
+      "step": 1464
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5506920191517627,
+      "learning_rate": 0.00017129901196753363,
+      "loss": 1.1324,
+      "step": 1465
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.4288643276075679,
+      "learning_rate": 0.00017125715270449138,
+      "loss": 1.1987,
+      "step": 1466
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.44776126985507103,
+      "learning_rate": 0.00017121526806116748,
+      "loss": 1.2772,
+      "step": 1467
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.39707520959231246,
+      "learning_rate": 0.0001711733580524804,
+      "loss": 1.1232,
+      "step": 1468
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.4151363016483633,
+      "learning_rate": 0.00017113142269335768,
+      "loss": 1.2417,
+      "step": 1469
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.4332434006158144,
+      "learning_rate": 0.00017108946199873568,
+      "loss": 1.2133,
+      "step": 1470
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.41706107554250604,
+      "learning_rate": 0.00017104747598355996,
+      "loss": 1.2052,
+      "step": 1471
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.47242208184671564,
+      "learning_rate": 0.00017100546466278504,
+      "loss": 0.7796,
+      "step": 1472
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5117396986591242,
+      "learning_rate": 0.00017096342805137448,
+      "loss": 1.1468,
+      "step": 1473
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.47178338973981093,
+      "learning_rate": 0.00017092136616430082,
+      "loss": 1.3539,
+      "step": 1474
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.44392048732865486,
+      "learning_rate": 0.00017087927901654557,
+      "loss": 1.1901,
+      "step": 1475
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.4430805829697702,
+      "learning_rate": 0.00017083716662309931,
+      "loss": 1.13,
+      "step": 1476
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.43757019755730836,
+      "learning_rate": 0.00017079502899896158,
+      "loss": 1.2816,
+      "step": 1477
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5089736320146705,
+      "learning_rate": 0.00017075286615914087,
+      "loss": 1.2143,
+      "step": 1478
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.42895531001611953,
+      "learning_rate": 0.00017071067811865476,
+      "loss": 0.9846,
+      "step": 1479
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.47680473593985534,
+      "learning_rate": 0.00017066846489252968,
+      "loss": 1.3386,
+      "step": 1480
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.47661708669014,
+      "learning_rate": 0.00017062622649580104,
+      "loss": 1.1361,
+      "step": 1481
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.44646589660020186,
+      "learning_rate": 0.00017058396294351334,
+      "loss": 1.3713,
+      "step": 1482
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.426948502878878,
+      "learning_rate": 0.00017054167425071995,
+      "loss": 1.348,
+      "step": 1483
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.4345351549749898,
+      "learning_rate": 0.00017049936043248316,
+      "loss": 1.198,
+      "step": 1484
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.4540213091578614,
+      "learning_rate": 0.00017045702150387428,
+      "loss": 1.0838,
+      "step": 1485
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.46871071031503386,
+      "learning_rate": 0.00017041465747997349,
+      "loss": 1.2799,
+      "step": 1486
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.4705369340530799,
+      "learning_rate": 0.00017037226837587002,
+      "loss": 1.3473,
+      "step": 1487
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.4396050303643297,
+      "learning_rate": 0.0001703298542066619,
+      "loss": 1.193,
+      "step": 1488
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.4142132330725613,
+      "learning_rate": 0.00017028741498745621,
+      "loss": 1.1547,
+      "step": 1489
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.4749863227388908,
+      "learning_rate": 0.0001702449507333689,
+      "loss": 1.0823,
+      "step": 1490
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.4461650767294646,
+      "learning_rate": 0.00017020246145952477,
+      "loss": 1.3081,
+      "step": 1491
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.4495258097708363,
+      "learning_rate": 0.00017015994718105768,
+      "loss": 1.0471,
+      "step": 1492
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.4575068376292397,
+      "learning_rate": 0.00017011740791311026,
+      "loss": 0.9931,
+      "step": 1493
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5289268394000792,
+      "learning_rate": 0.00017007484367083408,
+      "loss": 1.1608,
+      "step": 1494
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.3999329337261993,
+      "learning_rate": 0.00017003225446938965,
+      "loss": 1.2342,
+      "step": 1495
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.40050195164032726,
+      "learning_rate": 0.00016998964032394632,
+      "loss": 1.3305,
+      "step": 1496
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.40895600437690033,
+      "learning_rate": 0.00016994700124968234,
+      "loss": 1.3467,
+      "step": 1497
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.43983783968477275,
+      "learning_rate": 0.00016990433726178487,
+      "loss": 1.1656,
+      "step": 1498
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.45326497561569606,
+      "learning_rate": 0.00016986164837544987,
+      "loss": 1.2467,
+      "step": 1499
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.4426123203929683,
+      "learning_rate": 0.00016981893460588223,
+      "loss": 1.1888,
+      "step": 1500
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.4452937953079976,
+      "learning_rate": 0.0001697761959682957,
+      "loss": 1.0405,
+      "step": 1501
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.4468267656489385,
+      "learning_rate": 0.00016973343247791287,
+      "loss": 1.399,
+      "step": 1502
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.3966168813371914,
+      "learning_rate": 0.00016969064414996518,
+      "loss": 1.2748,
+      "step": 1503
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.4690350524558424,
+      "learning_rate": 0.0001696478309996929,
+      "loss": 1.234,
+      "step": 1504
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.4483771268360468,
+      "learning_rate": 0.00016960499304234522,
+      "loss": 1.3082,
+      "step": 1505
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.38624928089086574,
+      "learning_rate": 0.00016956213029318005,
+      "loss": 1.3295,
+      "step": 1506
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5160215064627696,
+      "learning_rate": 0.00016951924276746425,
+      "loss": 1.1713,
+      "step": 1507
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.556398045524901,
+      "learning_rate": 0.00016947633048047337,
+      "loss": 1.0956,
+      "step": 1508
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.43117821737284556,
+      "learning_rate": 0.0001694333934474919,
+      "loss": 1.2348,
+      "step": 1509
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5636472792213614,
+      "learning_rate": 0.00016939043168381307,
+      "loss": 1.0553,
+      "step": 1510
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.4571740862576168,
+      "learning_rate": 0.000169347445204739,
+      "loss": 1.1843,
+      "step": 1511
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.4016829345483073,
+      "learning_rate": 0.00016930443402558048,
+      "loss": 1.1907,
+      "step": 1512
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.3883848489228913,
+      "learning_rate": 0.00016926139816165727,
+      "loss": 1.3277,
+      "step": 1513
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.39577539866373695,
+      "learning_rate": 0.00016921833762829772,
+      "loss": 1.1903,
+      "step": 1514
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.47103298492786333,
+      "learning_rate": 0.00016917525244083918,
+      "loss": 1.1654,
+      "step": 1515
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5041153027964569,
+      "learning_rate": 0.00016913214261462763,
+      "loss": 1.2333,
+      "step": 1516
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.44258322784950666,
+      "learning_rate": 0.00016908900816501788,
+      "loss": 1.3573,
+      "step": 1517
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.4149913988760498,
+      "learning_rate": 0.00016904584910737346,
+      "loss": 1.1688,
+      "step": 1518
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.4420664726212976,
+      "learning_rate": 0.00016900266545706678,
+      "loss": 1.1511,
+      "step": 1519
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.4231512622173876,
+      "learning_rate": 0.00016895945722947888,
+      "loss": 1.2224,
+      "step": 1520
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.42183983956555277,
+      "learning_rate": 0.00016891622443999965,
+      "loss": 1.2239,
+      "step": 1521
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.4467845135875506,
+      "learning_rate": 0.00016887296710402767,
+      "loss": 1.2407,
+      "step": 1522
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.4429697725604104,
+      "learning_rate": 0.00016882968523697028,
+      "loss": 1.1209,
+      "step": 1523
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.3961916093299189,
+      "learning_rate": 0.00016878637885424358,
+      "loss": 1.319,
+      "step": 1524
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.536765456139952,
+      "learning_rate": 0.00016874304797127236,
+      "loss": 1.2277,
+      "step": 1525
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.45446163986691157,
+      "learning_rate": 0.00016869969260349018,
+      "loss": 1.1489,
+      "step": 1526
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.4794191729288296,
+      "learning_rate": 0.0001686563127663393,
+      "loss": 1.1407,
+      "step": 1527
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.3805683071181139,
+      "learning_rate": 0.00016861290847527066,
+      "loss": 1.1719,
+      "step": 1528
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.4155057138101656,
+      "learning_rate": 0.00016856947974574395,
+      "loss": 1.2362,
+      "step": 1529
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.4202047698101505,
+      "learning_rate": 0.00016852602659322757,
+      "loss": 1.2944,
+      "step": 1530
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.4101469629617072,
+      "learning_rate": 0.00016848254903319867,
+      "loss": 1.3144,
+      "step": 1531
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.42555576703272685,
+      "learning_rate": 0.0001684390470811429,
+      "loss": 1.2841,
+      "step": 1532
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.4442319785705862,
+      "learning_rate": 0.0001683955207525549,
+      "loss": 1.2443,
+      "step": 1533
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.4041463519771774,
+      "learning_rate": 0.00016835197006293763,
+      "loss": 1.2346,
+      "step": 1534
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.4844316877507346,
+      "learning_rate": 0.0001683083950278031,
+      "loss": 1.1621,
+      "step": 1535
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.3832326198473463,
+      "learning_rate": 0.00016826479566267168,
+      "loss": 1.2298,
+      "step": 1536
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.4656127773043143,
+      "learning_rate": 0.0001682211719830726,
+      "loss": 1.0707,
+      "step": 1537
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.3937351579684683,
+      "learning_rate": 0.00016817752400454368,
+      "loss": 1.2347,
+      "step": 1538
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.38592858428919347,
+      "learning_rate": 0.00016813385174263137,
+      "loss": 1.3028,
+      "step": 1539
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.3966406210808733,
+      "learning_rate": 0.00016809015521289085,
+      "loss": 1.3357,
+      "step": 1540
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.3728339566300712,
+      "learning_rate": 0.0001680464344308858,
+      "loss": 1.2282,
+      "step": 1541
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.39014131088374476,
+      "learning_rate": 0.00016800268941218876,
+      "loss": 1.2242,
+      "step": 1542
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.4284618233701778,
+      "learning_rate": 0.0001679589201723807,
+      "loss": 1.279,
+      "step": 1543
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.4018651452766923,
+      "learning_rate": 0.0001679151267270513,
+      "loss": 1.2435,
+      "step": 1544
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.45564409147438906,
+      "learning_rate": 0.00016787130909179883,
+      "loss": 1.1608,
+      "step": 1545
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.41934484232434044,
+      "learning_rate": 0.00016782746728223028,
+      "loss": 1.1918,
+      "step": 1546
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.3787926978746803,
+      "learning_rate": 0.0001677836013139611,
+      "loss": 1.2196,
+      "step": 1547
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.43284492369365385,
+      "learning_rate": 0.00016773971120261544,
+      "loss": 1.2155,
+      "step": 1548
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.4503186805425911,
+      "learning_rate": 0.00016769579696382597,
+      "loss": 1.0601,
+      "step": 1549
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.412395744394074,
+      "learning_rate": 0.0001676518586132341,
+      "loss": 1.215,
+      "step": 1550
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.45602630506615777,
+      "learning_rate": 0.00016760789616648966,
+      "loss": 1.2002,
+      "step": 1551
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.44944162706581975,
+      "learning_rate": 0.0001675639096392512,
+      "loss": 1.1641,
+      "step": 1552
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.503970104336148,
+      "learning_rate": 0.00016751989904718572,
+      "loss": 1.1717,
+      "step": 1553
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.44693595873254216,
+      "learning_rate": 0.0001674758644059689,
+      "loss": 1.1313,
+      "step": 1554
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.41663207545821196,
+      "learning_rate": 0.00016743180573128495,
+      "loss": 1.215,
+      "step": 1555
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.42430908225497066,
+      "learning_rate": 0.00016738772303882658,
+      "loss": 1.3302,
+      "step": 1556
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.4796066185093773,
+      "learning_rate": 0.00016734361634429515,
+      "loss": 1.1772,
+      "step": 1557
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.4134367037106033,
+      "learning_rate": 0.0001672994856634005,
+      "loss": 1.2745,
+      "step": 1558
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.4512035858427786,
+      "learning_rate": 0.0001672553310118611,
+      "loss": 1.1916,
+      "step": 1559
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.4532824502536244,
+      "learning_rate": 0.00016721115240540384,
+      "loss": 1.1607,
+      "step": 1560
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5718577302265501,
+      "learning_rate": 0.0001671669498597642,
+      "loss": 1.1216,
+      "step": 1561
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.578176599358481,
+      "learning_rate": 0.0001671227233906862,
+      "loss": 1.2425,
+      "step": 1562
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.45167072322724494,
+      "learning_rate": 0.00016707847301392236,
+      "loss": 1.1454,
+      "step": 1563
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.41360320778692705,
+      "learning_rate": 0.00016703419874523374,
+      "loss": 1.1438,
+      "step": 1564
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.42480513117222113,
+      "learning_rate": 0.00016698990060038985,
+      "loss": 1.1085,
+      "step": 1565
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.42173400191844096,
+      "learning_rate": 0.00016694557859516881,
+      "loss": 1.0151,
+      "step": 1566
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.44810095514713477,
+      "learning_rate": 0.00016690123274535715,
+      "loss": 1.3362,
+      "step": 1567
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.38423380728176837,
+      "learning_rate": 0.00016685686306674988,
+      "loss": 1.2188,
+      "step": 1568
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.3906743498798757,
+      "learning_rate": 0.0001668124695751506,
+      "loss": 1.3361,
+      "step": 1569
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.41596164563512833,
+      "learning_rate": 0.00016676805228637128,
+      "loss": 1.1552,
+      "step": 1570
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.4639880990638579,
+      "learning_rate": 0.00016672361121623238,
+      "loss": 1.1241,
+      "step": 1571
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.4921233257015731,
+      "learning_rate": 0.00016667914638056298,
+      "loss": 1.1629,
+      "step": 1572
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.4331966673864931,
+      "learning_rate": 0.0001666346577952004,
+      "loss": 1.2628,
+      "step": 1573
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.41946938833670405,
+      "learning_rate": 0.0001665901454759906,
+      "loss": 1.0054,
+      "step": 1574
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.4061863097772744,
+      "learning_rate": 0.00016654560943878783,
+      "loss": 1.246,
+      "step": 1575
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.4118858966348525,
+      "learning_rate": 0.00016650104969945497,
+      "loss": 1.1312,
+      "step": 1576
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.42390332090634386,
+      "learning_rate": 0.0001664564662738632,
+      "loss": 1.0709,
+      "step": 1577
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.4968784344120817,
+      "learning_rate": 0.00016641185917789226,
+      "loss": 1.1452,
+      "step": 1578
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.4792539184472965,
+      "learning_rate": 0.00016636722842743013,
+      "loss": 1.1768,
+      "step": 1579
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.3996127995397126,
+      "learning_rate": 0.00016632257403837345,
+      "loss": 1.1753,
+      "step": 1580
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.459878610090864,
+      "learning_rate": 0.00016627789602662708,
+      "loss": 1.1547,
+      "step": 1581
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.439894989425804,
+      "learning_rate": 0.0001662331944081044,
+      "loss": 1.0091,
+      "step": 1582
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.43722723052694956,
+      "learning_rate": 0.00016618846919872718,
+      "loss": 1.2179,
+      "step": 1583
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.48660215271837115,
+      "learning_rate": 0.0001661437204144256,
+      "loss": 1.2213,
+      "step": 1584
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.40059832281380264,
+      "learning_rate": 0.00016609894807113822,
+      "loss": 1.2673,
+      "step": 1585
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.44976726428240726,
+      "learning_rate": 0.00016605415218481199,
+      "loss": 1.1571,
+      "step": 1586
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.43871017413471464,
+      "learning_rate": 0.0001660093327714022,
+      "loss": 1.0785,
+      "step": 1587
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5424585541879717,
+      "learning_rate": 0.00016596448984687267,
+      "loss": 1.1599,
+      "step": 1588
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.4635065954456347,
+      "learning_rate": 0.00016591962342719544,
+      "loss": 1.1443,
+      "step": 1589
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.37740490918888703,
+      "learning_rate": 0.000165874733528351,
+      "loss": 1.3754,
+      "step": 1590
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.46807746854198806,
+      "learning_rate": 0.00016582982016632818,
+      "loss": 1.0321,
+      "step": 1591
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.46803719923474757,
+      "learning_rate": 0.00016578488335712409,
+      "loss": 1.2326,
+      "step": 1592
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.4012697564467126,
+      "learning_rate": 0.00016573992311674437,
+      "loss": 1.2434,
+      "step": 1593
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.393851866916742,
+      "learning_rate": 0.00016569493946120282,
+      "loss": 1.085,
+      "step": 1594
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.37887402778996526,
+      "learning_rate": 0.0001656499324065217,
+      "loss": 1.2283,
+      "step": 1595
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.433412782165206,
+      "learning_rate": 0.00016560490196873155,
+      "loss": 1.0701,
+      "step": 1596
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.3892944292416749,
+      "learning_rate": 0.00016555984816387124,
+      "loss": 1.1717,
+      "step": 1597
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.4114568882124006,
+      "learning_rate": 0.00016551477100798805,
+      "loss": 1.2997,
+      "step": 1598
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.47580861876098945,
+      "learning_rate": 0.00016546967051713743,
+      "loss": 1.1292,
+      "step": 1599
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.3887536471869512,
+      "learning_rate": 0.0001654245467073832,
+      "loss": 1.3312,
+      "step": 1600
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.3983264362403241,
+      "learning_rate": 0.00016537939959479755,
+      "loss": 1.2465,
+      "step": 1601
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5300697680512821,
+      "learning_rate": 0.00016533422919546092,
+      "loss": 1.1712,
+      "step": 1602
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.45948597197280905,
+      "learning_rate": 0.00016528903552546207,
+      "loss": 1.193,
+      "step": 1603
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.4307389095787904,
+      "learning_rate": 0.00016524381860089795,
+      "loss": 1.0148,
+      "step": 1604
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.42687134299304036,
+      "learning_rate": 0.00016519857843787388,
+      "loss": 1.3037,
+      "step": 1605
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.39102586834572334,
+      "learning_rate": 0.00016515331505250348,
+      "loss": 1.241,
+      "step": 1606
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.40940668470072605,
+      "learning_rate": 0.00016510802846090862,
+      "loss": 1.299,
+      "step": 1607
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.4059712672553914,
+      "learning_rate": 0.00016506271867921935,
+      "loss": 1.1823,
+      "step": 1608
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.45459313729434864,
+      "learning_rate": 0.00016501738572357412,
+      "loss": 1.1697,
+      "step": 1609
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.4227555725093705,
+      "learning_rate": 0.00016497202961011952,
+      "loss": 1.186,
+      "step": 1610
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5154105087262955,
+      "learning_rate": 0.00016492665035501046,
+      "loss": 1.1047,
+      "step": 1611
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.40521182254720123,
+      "learning_rate": 0.00016488124797441004,
+      "loss": 1.2132,
+      "step": 1612
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5306228032748829,
+      "learning_rate": 0.00016483582248448964,
+      "loss": 1.1163,
+      "step": 1613
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.4758690008703151,
+      "learning_rate": 0.0001647903739014289,
+      "loss": 1.2368,
+      "step": 1614
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.4791963673125965,
+      "learning_rate": 0.0001647449022414155,
+      "loss": 1.2488,
+      "step": 1615
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.43395780034193737,
+      "learning_rate": 0.00016469940752064562,
+      "loss": 1.1981,
+      "step": 1616
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.4148667121565746,
+      "learning_rate": 0.00016465388975532344,
+      "loss": 1.2394,
+      "step": 1617
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5063239199038609,
+      "learning_rate": 0.0001646083489616614,
+      "loss": 1.2204,
+      "step": 1618
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.40275885731748096,
+      "learning_rate": 0.00016456278515588024,
+      "loss": 1.1557,
+      "step": 1619
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.3854492071455529,
+      "learning_rate": 0.00016451719835420877,
+      "loss": 1.2014,
+      "step": 1620
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.4162172855142556,
+      "learning_rate": 0.00016447158857288403,
+      "loss": 1.2209,
+      "step": 1621
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.3776528341549551,
+      "learning_rate": 0.00016442595582815127,
+      "loss": 1.1734,
+      "step": 1622
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.46075685395669613,
+      "learning_rate": 0.00016438030013626393,
+      "loss": 1.1505,
+      "step": 1623
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.461400836505417,
+      "learning_rate": 0.00016433462151348353,
+      "loss": 1.2368,
+      "step": 1624
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.44282914538517293,
+      "learning_rate": 0.00016428891997607986,
+      "loss": 1.1899,
+      "step": 1625
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.48404740885485226,
+      "learning_rate": 0.00016424319554033084,
+      "loss": 1.0767,
+      "step": 1626
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.4121795031221339,
+      "learning_rate": 0.00016419744822252253,
+      "loss": 1.1866,
+      "step": 1627
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.44920691939017476,
+      "learning_rate": 0.0001641516780389492,
+      "loss": 1.274,
+      "step": 1628
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.4977363952867657,
+      "learning_rate": 0.00016410588500591312,
+      "loss": 1.2278,
+      "step": 1629
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.41981195304585894,
+      "learning_rate": 0.00016406006913972484,
+      "loss": 1.1245,
+      "step": 1630
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5191214386847243,
+      "learning_rate": 0.00016401423045670304,
+      "loss": 1.1904,
+      "step": 1631
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.48839676276548294,
+      "learning_rate": 0.00016396836897317446,
+      "loss": 1.2261,
+      "step": 1632
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.49216745360443964,
+      "learning_rate": 0.00016392248470547394,
+      "loss": 1.1623,
+      "step": 1633
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.43654548378018676,
+      "learning_rate": 0.00016387657766994451,
+      "loss": 1.123,
+      "step": 1634
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.43995594656683934,
+      "learning_rate": 0.0001638306478829373,
+      "loss": 1.1434,
+      "step": 1635
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.45353264157460665,
+      "learning_rate": 0.00016378469536081148,
+      "loss": 1.1295,
+      "step": 1636
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.42112066361205386,
+      "learning_rate": 0.00016373872011993438,
+      "loss": 1.1522,
+      "step": 1637
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.4703627699516045,
+      "learning_rate": 0.00016369272217668144,
+      "loss": 1.1064,
+      "step": 1638
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.39717228996004994,
+      "learning_rate": 0.00016364670154743607,
+      "loss": 1.1965,
+      "step": 1639
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.4265055739427768,
+      "learning_rate": 0.00016360065824858993,
+      "loss": 1.28,
+      "step": 1640
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.45871003102522284,
+      "learning_rate": 0.00016355459229654257,
+      "loss": 1.0099,
+      "step": 1641
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.439830522278743,
+      "learning_rate": 0.00016350850370770178,
+      "loss": 1.2002,
+      "step": 1642
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.46855962612210655,
+      "learning_rate": 0.0001634623924984833,
+      "loss": 1.161,
+      "step": 1643
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.4560701898022395,
+      "learning_rate": 0.00016341625868531096,
+      "loss": 1.1726,
+      "step": 1644
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.4858100539684072,
+      "learning_rate": 0.00016337010228461668,
+      "loss": 1.3182,
+      "step": 1645
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.3914772547943513,
+      "learning_rate": 0.0001633239233128403,
+      "loss": 1.3999,
+      "step": 1646
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.44795574482544215,
+      "learning_rate": 0.00016327772178642986,
+      "loss": 1.1174,
+      "step": 1647
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.4540494306175798,
+      "learning_rate": 0.00016323149772184134,
+      "loss": 1.1113,
+      "step": 1648
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.38353133702902475,
+      "learning_rate": 0.0001631852511355388,
+      "loss": 1.2947,
+      "step": 1649
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.4362260020190405,
+      "learning_rate": 0.00016313898204399423,
+      "loss": 1.3301,
+      "step": 1650
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.4261187573204202,
+      "learning_rate": 0.00016309269046368776,
+      "loss": 1.0933,
+      "step": 1651
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.4900615004251311,
+      "learning_rate": 0.0001630463764111074,
+      "loss": 1.1145,
+      "step": 1652
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.41521729977453636,
+      "learning_rate": 0.00016300003990274927,
+      "loss": 1.2258,
+      "step": 1653
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.502794998108499,
+      "learning_rate": 0.00016295368095511746,
+      "loss": 1.2172,
+      "step": 1654
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.43652823237777455,
+      "learning_rate": 0.000162907299584724,
+      "loss": 1.2746,
+      "step": 1655
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7371546518688992,
+      "learning_rate": 0.000162860895808089,
+      "loss": 1.3165,
+      "step": 1656
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5163216995437213,
+      "learning_rate": 0.00016281446964174046,
+      "loss": 1.1004,
+      "step": 1657
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.3843307283877981,
+      "learning_rate": 0.0001627680211022144,
+      "loss": 1.2276,
+      "step": 1658
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.41415071560156536,
+      "learning_rate": 0.0001627215502060548,
+      "loss": 1.0834,
+      "step": 1659
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.4868824658022943,
+      "learning_rate": 0.0001626750569698136,
+      "loss": 1.1694,
+      "step": 1660
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.42576149729932106,
+      "learning_rate": 0.0001626285414100507,
+      "loss": 1.1107,
+      "step": 1661
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5041380918601903,
+      "learning_rate": 0.00016258200354333396,
+      "loss": 1.1721,
+      "step": 1662
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.4041447494406083,
+      "learning_rate": 0.00016253544338623916,
+      "loss": 1.2929,
+      "step": 1663
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5247320150310668,
+      "learning_rate": 0.0001624888609553501,
+      "loss": 1.3578,
+      "step": 1664
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.46156582104399335,
+      "learning_rate": 0.00016244225626725836,
+      "loss": 1.1977,
+      "step": 1665
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5069701037784781,
+      "learning_rate": 0.00016239562933856365,
+      "loss": 1.0671,
+      "step": 1666
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.4318660752317375,
+      "learning_rate": 0.00016234898018587337,
+      "loss": 1.2188,
+      "step": 1667
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5033047274705422,
+      "learning_rate": 0.00016230230882580303,
+      "loss": 1.2333,
+      "step": 1668
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.43628698903837426,
+      "learning_rate": 0.00016225561527497594,
+      "loss": 1.053,
+      "step": 1669
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.3961762720618994,
+      "learning_rate": 0.00016220889955002337,
+      "loss": 1.1429,
+      "step": 1670
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.45466332186762354,
+      "learning_rate": 0.00016216216166758447,
+      "loss": 1.3935,
+      "step": 1671
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.4398752451002478,
+      "learning_rate": 0.00016211540164430625,
+      "loss": 1.1456,
+      "step": 1672
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.3733465294362133,
+      "learning_rate": 0.00016206861949684366,
+      "loss": 1.194,
+      "step": 1673
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.4058608594305043,
+      "learning_rate": 0.00016202181524185952,
+      "loss": 1.247,
+      "step": 1674
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.4730686974179459,
+      "learning_rate": 0.00016197498889602448,
+      "loss": 1.1531,
+      "step": 1675
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.4216777045306259,
+      "learning_rate": 0.00016192814047601712,
+      "loss": 1.2481,
+      "step": 1676
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.45302074128412045,
+      "learning_rate": 0.00016188126999852381,
+      "loss": 1.1881,
+      "step": 1677
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.39643112014329795,
+      "learning_rate": 0.00016183437748023885,
+      "loss": 1.1416,
+      "step": 1678
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.4073782558781691,
+      "learning_rate": 0.00016178746293786433,
+      "loss": 1.0481,
+      "step": 1679
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.4121756023700516,
+      "learning_rate": 0.00016174052638811024,
+      "loss": 1.3541,
+      "step": 1680
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.40176177412126923,
+      "learning_rate": 0.00016169356784769435,
+      "loss": 1.0346,
+      "step": 1681
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.44860299710613866,
+      "learning_rate": 0.0001616465873333423,
+      "loss": 1.1266,
+      "step": 1682
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.42428353744605735,
+      "learning_rate": 0.0001615995848617876,
+      "loss": 1.4204,
+      "step": 1683
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.4347729719040462,
+      "learning_rate": 0.00016155256044977145,
+      "loss": 1.1499,
+      "step": 1684
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.4690672656305678,
+      "learning_rate": 0.000161505514114043,
+      "loss": 1.0942,
+      "step": 1685
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.4188966157140494,
+      "learning_rate": 0.00016145844587135913,
+      "loss": 1.1403,
+      "step": 1686
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.39438027031658585,
+      "learning_rate": 0.00016141135573848457,
+      "loss": 1.1975,
+      "step": 1687
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.42378137353127016,
+      "learning_rate": 0.0001613642437321918,
+      "loss": 1.2965,
+      "step": 1688
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.4236097204893352,
+      "learning_rate": 0.00016131710986926108,
+      "loss": 1.1889,
+      "step": 1689
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.4082583546667757,
+      "learning_rate": 0.00016126995416648057,
+      "loss": 1.2599,
+      "step": 1690
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.44537770910924573,
+      "learning_rate": 0.0001612227766406461,
+      "loss": 1.2271,
+      "step": 1691
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.4763057955112946,
+      "learning_rate": 0.00016117557730856125,
+      "loss": 1.1816,
+      "step": 1692
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.4425908858027097,
+      "learning_rate": 0.00016112835618703746,
+      "loss": 1.3699,
+      "step": 1693
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.4099762803305613,
+      "learning_rate": 0.0001610811132928939,
+      "loss": 1.1898,
+      "step": 1694
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.4330114967778623,
+      "learning_rate": 0.0001610338486429575,
+      "loss": 1.0895,
+      "step": 1695
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.4367321241180363,
+      "learning_rate": 0.00016098656225406287,
+      "loss": 1.118,
+      "step": 1696
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.39123354327378257,
+      "learning_rate": 0.0001609392541430524,
+      "loss": 1.3701,
+      "step": 1697
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.41166900957595226,
+      "learning_rate": 0.00016089192432677634,
+      "loss": 1.1978,
+      "step": 1698
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.44981772646069046,
+      "learning_rate": 0.00016084457282209243,
+      "loss": 1.1687,
+      "step": 1699
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.46021074302617704,
+      "learning_rate": 0.00016079719964586637,
+      "loss": 1.1758,
+      "step": 1700
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.4100384515737777,
+      "learning_rate": 0.0001607498048149714,
+      "loss": 1.262,
+      "step": 1701
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.4771338635308508,
+      "learning_rate": 0.00016070238834628865,
+      "loss": 1.0152,
+      "step": 1702
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.4626328154987715,
+      "learning_rate": 0.00016065495025670675,
+      "loss": 1.0974,
+      "step": 1703
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.3916275985111847,
+      "learning_rate": 0.00016060749056312222,
+      "loss": 1.2168,
+      "step": 1704
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.41611210493549144,
+      "learning_rate": 0.00016056000928243912,
+      "loss": 1.1878,
+      "step": 1705
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5083971222713042,
+      "learning_rate": 0.0001605125064315693,
+      "loss": 1.1887,
+      "step": 1706
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.48278969074472455,
+      "learning_rate": 0.00016046498202743233,
+      "loss": 1.0193,
+      "step": 1707
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5184137270833924,
+      "learning_rate": 0.00016041743608695533,
+      "loss": 1.1058,
+      "step": 1708
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.47327250466634496,
+      "learning_rate": 0.0001603698686270731,
+      "loss": 1.1303,
+      "step": 1709
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.389822941495763,
+      "learning_rate": 0.00016032227966472827,
+      "loss": 1.1326,
+      "step": 1710
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.397888370473302,
+      "learning_rate": 0.00016027466921687093,
+      "loss": 1.2403,
+      "step": 1711
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.3843820987598725,
+      "learning_rate": 0.00016022703730045895,
+      "loss": 1.146,
+      "step": 1712
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.3856288245500627,
+      "learning_rate": 0.0001601793839324578,
+      "loss": 1.3353,
+      "step": 1713
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.3794759342380828,
+      "learning_rate": 0.00016013170912984058,
+      "loss": 1.2022,
+      "step": 1714
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.374760040752024,
+      "learning_rate": 0.00016008401290958807,
+      "loss": 1.2374,
+      "step": 1715
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.41637784652426957,
+      "learning_rate": 0.00016003629528868859,
+      "loss": 1.1873,
+      "step": 1716
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.40558648104643075,
+      "learning_rate": 0.00015998855628413823,
+      "loss": 1.3967,
+      "step": 1717
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.4425096087435899,
+      "learning_rate": 0.00015994079591294055,
+      "loss": 1.2529,
+      "step": 1718
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.4050461414681135,
+      "learning_rate": 0.00015989301419210675,
+      "loss": 1.0405,
+      "step": 1719
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.42615797728829863,
+      "learning_rate": 0.00015984521113865572,
+      "loss": 1.2471,
+      "step": 1720
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.37056977648598893,
+      "learning_rate": 0.00015979738676961388,
+      "loss": 1.1038,
+      "step": 1721
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.4525796690758856,
+      "learning_rate": 0.00015974954110201526,
+      "loss": 1.2427,
+      "step": 1722
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.39509906073401463,
+      "learning_rate": 0.0001597016741529014,
+      "loss": 1.2498,
+      "step": 1723
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.4215581266847536,
+      "learning_rate": 0.00015965378593932157,
+      "loss": 1.1014,
+      "step": 1724
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.42724808883245896,
+      "learning_rate": 0.00015960587647833249,
+      "loss": 1.2662,
+      "step": 1725
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.487177568180821,
+      "learning_rate": 0.0001595579457869985,
+      "loss": 1.2651,
+      "step": 1726
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.44640046943909156,
+      "learning_rate": 0.0001595099938823915,
+      "loss": 1.1778,
+      "step": 1727
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.4755115779786393,
+      "learning_rate": 0.00015946202078159094,
+      "loss": 1.1531,
+      "step": 1728
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.4511374318508332,
+      "learning_rate": 0.00015941402650168382,
+      "loss": 1.1929,
+      "step": 1729
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.47517656024997607,
+      "learning_rate": 0.00015936601105976465,
+      "loss": 1.2678,
+      "step": 1730
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.49040005347460136,
+      "learning_rate": 0.00015931797447293552,
+      "loss": 1.315,
+      "step": 1731
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.41272814725826973,
+      "learning_rate": 0.0001592699167583061,
+      "loss": 1.261,
+      "step": 1732
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.4100387132519615,
+      "learning_rate": 0.00015922183793299342,
+      "loss": 1.2317,
+      "step": 1733
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.4578750652024175,
+      "learning_rate": 0.00015917373801412223,
+      "loss": 1.1575,
+      "step": 1734
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.4159203104851182,
+      "learning_rate": 0.00015912561701882463,
+      "loss": 1.1398,
+      "step": 1735
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.41105777195617127,
+      "learning_rate": 0.00015907747496424034,
+      "loss": 1.1512,
+      "step": 1736
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.3836186336861135,
+      "learning_rate": 0.00015902931186751652,
+      "loss": 1.1821,
+      "step": 1737
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.4265474069365521,
+      "learning_rate": 0.00015898112774580784,
+      "loss": 1.2885,
+      "step": 1738
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.43475439252339937,
+      "learning_rate": 0.00015893292261627643,
+      "loss": 1.1935,
+      "step": 1739
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.4563940241537527,
+      "learning_rate": 0.00015888469649609198,
+      "loss": 1.269,
+      "step": 1740
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.4487348519758521,
+      "learning_rate": 0.00015883644940243162,
+      "loss": 1.2859,
+      "step": 1741
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.4650891340371591,
+      "learning_rate": 0.00015878818135247987,
+      "loss": 1.0631,
+      "step": 1742
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.4126977836277259,
+      "learning_rate": 0.0001587398923634288,
+      "loss": 1.2923,
+      "step": 1743
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.3914074693789984,
+      "learning_rate": 0.00015869158245247796,
+      "loss": 1.3623,
+      "step": 1744
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5164797630752154,
+      "learning_rate": 0.00015864325163683431,
+      "loss": 1.1013,
+      "step": 1745
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.4167490756961133,
+      "learning_rate": 0.0001585948999337122,
+      "loss": 1.0673,
+      "step": 1746
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.44667590545071534,
+      "learning_rate": 0.00015854652736033354,
+      "loss": 1.0098,
+      "step": 1747
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.41351805972120403,
+      "learning_rate": 0.0001584981339339276,
+      "loss": 1.16,
+      "step": 1748
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.4248604724012829,
+      "learning_rate": 0.00015844971967173102,
+      "loss": 1.1896,
+      "step": 1749
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.45401698292618115,
+      "learning_rate": 0.00015840128459098805,
+      "loss": 1.1223,
+      "step": 1750
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.4327696770196478,
+      "learning_rate": 0.00015835282870895015,
+      "loss": 1.1893,
+      "step": 1751
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.38293566611659635,
+      "learning_rate": 0.00015830435204287628,
+      "loss": 1.185,
+      "step": 1752
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.412085597946236,
+      "learning_rate": 0.00015825585461003283,
+      "loss": 1.2588,
+      "step": 1753
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.4574108494168438,
+      "learning_rate": 0.0001582073364276935,
+      "loss": 1.0687,
+      "step": 1754
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.3981632781005765,
+      "learning_rate": 0.00015815879751313955,
+      "loss": 1.3623,
+      "step": 1755
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5009891093367195,
+      "learning_rate": 0.00015811023788365937,
+      "loss": 1.2079,
+      "step": 1756
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.4134144959880611,
+      "learning_rate": 0.00015806165755654893,
+      "loss": 1.111,
+      "step": 1757
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.47812471516166966,
+      "learning_rate": 0.0001580130565491115,
+      "loss": 1.1363,
+      "step": 1758
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.4652120850167576,
+      "learning_rate": 0.00015796443487865776,
+      "loss": 1.1097,
+      "step": 1759
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.40821604977525694,
+      "learning_rate": 0.00015791579256250565,
+      "loss": 1.1896,
+      "step": 1760
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.4621913611897655,
+      "learning_rate": 0.0001578671296179806,
+      "loss": 1.2208,
+      "step": 1761
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.42415605045189836,
+      "learning_rate": 0.00015781844606241525,
+      "loss": 1.078,
+      "step": 1762
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.4398025993230896,
+      "learning_rate": 0.0001577697419131497,
+      "loss": 1.1491,
+      "step": 1763
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.41357145326378864,
+      "learning_rate": 0.00015772101718753125,
+      "loss": 1.2845,
+      "step": 1764
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5082729457578529,
+      "learning_rate": 0.00015767227190291468,
+      "loss": 1.1158,
+      "step": 1765
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.4241216321941602,
+      "learning_rate": 0.00015762350607666204,
+      "loss": 1.2386,
+      "step": 1766
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.39317966634863677,
+      "learning_rate": 0.00015757471972614265,
+      "loss": 1.3757,
+      "step": 1767
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.46971295574109334,
+      "learning_rate": 0.00015752591286873313,
+      "loss": 1.1275,
+      "step": 1768
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.40768188954825213,
+      "learning_rate": 0.0001574770855218175,
+      "loss": 1.1163,
+      "step": 1769
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.39982358321003786,
+      "learning_rate": 0.00015742823770278702,
+      "loss": 1.0757,
+      "step": 1770
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.42027441410341365,
+      "learning_rate": 0.00015737936942904023,
+      "loss": 1.2549,
+      "step": 1771
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.4638745991804609,
+      "learning_rate": 0.00015733048071798296,
+      "loss": 1.2804,
+      "step": 1772
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.4661800029223617,
+      "learning_rate": 0.00015728157158702833,
+      "loss": 1.05,
+      "step": 1773
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.4028043533420381,
+      "learning_rate": 0.00015723264205359675,
+      "loss": 1.2611,
+      "step": 1774
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.4698659188627507,
+      "learning_rate": 0.00015718369213511585,
+      "loss": 1.1397,
+      "step": 1775
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.4093010266175056,
+      "learning_rate": 0.00015713472184902057,
+      "loss": 1.3195,
+      "step": 1776
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.4141584436928288,
+      "learning_rate": 0.0001570857312127531,
+      "loss": 1.1999,
+      "step": 1777
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.3969581281514913,
+      "learning_rate": 0.00015703672024376282,
+      "loss": 1.2279,
+      "step": 1778
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.4534852289706912,
+      "learning_rate": 0.00015698768895950642,
+      "loss": 1.1913,
+      "step": 1779
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.539405089517775,
+      "learning_rate": 0.0001569386373774478,
+      "loss": 1.2617,
+      "step": 1780
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.43746089572755803,
+      "learning_rate": 0.0001568895655150581,
+      "loss": 1.2087,
+      "step": 1781
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.3932248513334125,
+      "learning_rate": 0.00015684047338981568,
+      "loss": 1.2207,
+      "step": 1782
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.4432081245731642,
+      "learning_rate": 0.00015679136101920605,
+      "loss": 1.1853,
+      "step": 1783
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.4092168415514172,
+      "learning_rate": 0.00015674222842072203,
+      "loss": 1.3283,
+      "step": 1784
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.4071003512176134,
+      "learning_rate": 0.00015669307561186363,
+      "loss": 1.2428,
+      "step": 1785
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.4387285145312934,
+      "learning_rate": 0.00015664390261013798,
+      "loss": 1.0451,
+      "step": 1786
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.4407984351967257,
+      "learning_rate": 0.00015659470943305955,
+      "loss": 1.3593,
+      "step": 1787
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.45520046643627343,
+      "learning_rate": 0.0001565454960981498,
+      "loss": 1.1059,
+      "step": 1788
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.39946385159097847,
+      "learning_rate": 0.0001564962626229375,
+      "loss": 1.3624,
+      "step": 1789
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.46187060796203705,
+      "learning_rate": 0.0001564470090249586,
+      "loss": 1.2443,
+      "step": 1790
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.3914918432079768,
+      "learning_rate": 0.00015639773532175616,
+      "loss": 1.1912,
+      "step": 1791
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.4830162573914962,
+      "learning_rate": 0.00015634844153088043,
+      "loss": 1.148,
+      "step": 1792
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.47686098508991775,
+      "learning_rate": 0.00015629912766988884,
+      "loss": 1.2799,
+      "step": 1793
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.388916074581727,
+      "learning_rate": 0.00015624979375634588,
+      "loss": 1.2811,
+      "step": 1794
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.4823462477273454,
+      "learning_rate": 0.00015620043980782327,
+      "loss": 1.182,
+      "step": 1795
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.38273634699891484,
+      "learning_rate": 0.00015615106584189982,
+      "loss": 1.1567,
+      "step": 1796
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.4585173102532292,
+      "learning_rate": 0.00015610167187616152,
+      "loss": 1.1185,
+      "step": 1797
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.511323465326603,
+      "learning_rate": 0.00015605225792820142,
+      "loss": 1.0439,
+      "step": 1798
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.4730058953125711,
+      "learning_rate": 0.0001560028240156197,
+      "loss": 1.1191,
+      "step": 1799
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.43058326100377375,
+      "learning_rate": 0.00015595337015602377,
+      "loss": 1.2109,
+      "step": 1800
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.45695718938888114,
+      "learning_rate": 0.0001559038963670279,
+      "loss": 1.2519,
+      "step": 1801
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.47507697296622164,
+      "learning_rate": 0.0001558544026662537,
+      "loss": 0.9938,
+      "step": 1802
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5068262055035979,
+      "learning_rate": 0.00015580488907132974,
+      "loss": 1.0632,
+      "step": 1803
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.4246682691804972,
+      "learning_rate": 0.00015575535559989167,
+      "loss": 1.2569,
+      "step": 1804
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.43935477155592706,
+      "learning_rate": 0.00015570580226958232,
+      "loss": 1.2906,
+      "step": 1805
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.45064603614171805,
+      "learning_rate": 0.00015565622909805154,
+      "loss": 1.1077,
+      "step": 1806
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5448754441717751,
+      "learning_rate": 0.00015560663610295617,
+      "loss": 1.0894,
+      "step": 1807
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.4330699839664429,
+      "learning_rate": 0.00015555702330196023,
+      "loss": 1.0632,
+      "step": 1808
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.4171061044952646,
+      "learning_rate": 0.00015550739071273475,
+      "loss": 1.1305,
+      "step": 1809
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.43429496044297056,
+      "learning_rate": 0.00015545773835295775,
+      "loss": 1.1975,
+      "step": 1810
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.4395732644664144,
+      "learning_rate": 0.00015540806624031442,
+      "loss": 1.2937,
+      "step": 1811
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5606108626281344,
+      "learning_rate": 0.00015535837439249682,
+      "loss": 1.1767,
+      "step": 1812
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.39577912172036195,
+      "learning_rate": 0.0001553086628272042,
+      "loss": 1.1883,
+      "step": 1813
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.4686988335852529,
+      "learning_rate": 0.00015525893156214274,
+      "loss": 1.1739,
+      "step": 1814
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.4255067915007002,
+      "learning_rate": 0.00015520918061502569,
+      "loss": 1.2146,
+      "step": 1815
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.43200244770044904,
+      "learning_rate": 0.0001551594100035732,
+      "loss": 1.157,
+      "step": 1816
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.4484435643113129,
+      "learning_rate": 0.00015510961974551257,
+      "loss": 0.9922,
+      "step": 1817
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.41838485790836893,
+      "learning_rate": 0.000155059809858578,
+      "loss": 1.2915,
+      "step": 1818
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.4799732801006926,
+      "learning_rate": 0.00015500998036051074,
+      "loss": 1.1534,
+      "step": 1819
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.41059391865502676,
+      "learning_rate": 0.00015496013126905896,
+      "loss": 1.1972,
+      "step": 1820
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.42342679678639666,
+      "learning_rate": 0.00015491026260197786,
+      "loss": 1.1565,
+      "step": 1821
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.4274385309308075,
+      "learning_rate": 0.0001548603743770296,
+      "loss": 1.1261,
+      "step": 1822
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.42173933587509055,
+      "learning_rate": 0.00015481046661198334,
+      "loss": 1.1431,
+      "step": 1823
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.451534944956152,
+      "learning_rate": 0.00015476053932461512,
+      "loss": 1.4362,
+      "step": 1824
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.39567062039973827,
+      "learning_rate": 0.00015471059253270797,
+      "loss": 1.2291,
+      "step": 1825
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.4210712798493065,
+      "learning_rate": 0.00015466062625405189,
+      "loss": 1.1083,
+      "step": 1826
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.45439134163154704,
+      "learning_rate": 0.0001546106405064438,
+      "loss": 1.1085,
+      "step": 1827
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5064401765554559,
+      "learning_rate": 0.00015456063530768757,
+      "loss": 1.037,
+      "step": 1828
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.39383570021675257,
+      "learning_rate": 0.000154510610675594,
+      "loss": 1.2786,
+      "step": 1829
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.43492849283719764,
+      "learning_rate": 0.0001544605666279808,
+      "loss": 1.1624,
+      "step": 1830
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.41212551719845253,
+      "learning_rate": 0.00015441050318267254,
+      "loss": 1.1386,
+      "step": 1831
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.42593689614485314,
+      "learning_rate": 0.00015436042035750082,
+      "loss": 1.2035,
+      "step": 1832
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.4350905850538593,
+      "learning_rate": 0.00015431031817030402,
+      "loss": 1.0953,
+      "step": 1833
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.4285921287104956,
+      "learning_rate": 0.00015426019663892753,
+      "loss": 1.0943,
+      "step": 1834
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.415373502547973,
+      "learning_rate": 0.00015421005578122356,
+      "loss": 1.0068,
+      "step": 1835
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.41221761163435455,
+      "learning_rate": 0.00015415989561505118,
+      "loss": 1.2899,
+      "step": 1836
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.45016816119278086,
+      "learning_rate": 0.0001541097161582764,
+      "loss": 1.1311,
+      "step": 1837
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.4341753321727164,
+      "learning_rate": 0.0001540595174287721,
+      "loss": 1.1183,
+      "step": 1838
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.4128478734336392,
+      "learning_rate": 0.00015400929944441792,
+      "loss": 1.4366,
+      "step": 1839
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.44651465934536194,
+      "learning_rate": 0.0001539590622231005,
+      "loss": 1.2835,
+      "step": 1840
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.4134748693251177,
+      "learning_rate": 0.00015390880578271328,
+      "loss": 1.3274,
+      "step": 1841
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.4706257500298289,
+      "learning_rate": 0.0001538585301411565,
+      "loss": 1.2298,
+      "step": 1842
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.42958765959404877,
+      "learning_rate": 0.00015380823531633729,
+      "loss": 1.3058,
+      "step": 1843
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.4347876253907526,
+      "learning_rate": 0.00015375792132616957,
+      "loss": 1.1301,
+      "step": 1844
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.43882939817989625,
+      "learning_rate": 0.00015370758818857415,
+      "loss": 1.181,
+      "step": 1845
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.4265575853743481,
+      "learning_rate": 0.0001536572359214786,
+      "loss": 1.1563,
+      "step": 1846
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5029417826810941,
+      "learning_rate": 0.0001536068645428174,
+      "loss": 1.2474,
+      "step": 1847
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5068576605710478,
+      "learning_rate": 0.00015355647407053162,
+      "loss": 1.2244,
+      "step": 1848
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.46380317030910756,
+      "learning_rate": 0.00015350606452256934,
+      "loss": 1.1111,
+      "step": 1849
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.4312982465266956,
+      "learning_rate": 0.0001534556359168854,
+      "loss": 1.1467,
+      "step": 1850
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.45540162608288415,
+      "learning_rate": 0.00015340518827144145,
+      "loss": 1.0173,
+      "step": 1851
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.42104073308179013,
+      "learning_rate": 0.0001533547216042057,
+      "loss": 1.3681,
+      "step": 1852
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5041216793168757,
+      "learning_rate": 0.0001533042359331534,
+      "loss": 1.157,
+      "step": 1853
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.45206899705329506,
+      "learning_rate": 0.00015325373127626652,
+      "loss": 1.0999,
+      "step": 1854
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.3841096768909624,
+      "learning_rate": 0.00015320320765153367,
+      "loss": 1.2063,
+      "step": 1855
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.38944091628228733,
+      "learning_rate": 0.0001531526650769503,
+      "loss": 1.3207,
+      "step": 1856
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.4652529152862946,
+      "learning_rate": 0.00015310210357051863,
+      "loss": 1.2279,
+      "step": 1857
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.47086762662367704,
+      "learning_rate": 0.00015305152315024756,
+      "loss": 1.1386,
+      "step": 1858
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.4359463366693932,
+      "learning_rate": 0.00015300092383415282,
+      "loss": 1.0657,
+      "step": 1859
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.43585900699539765,
+      "learning_rate": 0.0001529503056402567,
+      "loss": 1.2803,
+      "step": 1860
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5139521050210564,
+      "learning_rate": 0.00015289966858658842,
+      "loss": 1.15,
+      "step": 1861
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.43308021683725256,
+      "learning_rate": 0.00015284901269118376,
+      "loss": 1.1816,
+      "step": 1862
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.39135686636576933,
+      "learning_rate": 0.00015279833797208535,
+      "loss": 0.9869,
+      "step": 1863
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5188298262746116,
+      "learning_rate": 0.00015274764444734234,
+      "loss": 1.0499,
+      "step": 1864
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5251560858865996,
+      "learning_rate": 0.0001526969321350108,
+      "loss": 1.3035,
+      "step": 1865
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.4240178660341211,
+      "learning_rate": 0.00015264620105315328,
+      "loss": 1.2018,
+      "step": 1866
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.4294767358881154,
+      "learning_rate": 0.0001525954512198392,
+      "loss": 1.2408,
+      "step": 1867
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.4095094194155571,
+      "learning_rate": 0.00015254468265314447,
+      "loss": 1.1359,
+      "step": 1868
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.3913389681384852,
+      "learning_rate": 0.00015249389537115183,
+      "loss": 1.1643,
+      "step": 1869
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.4430686397489738,
+      "learning_rate": 0.00015244308939195065,
+      "loss": 1.2541,
+      "step": 1870
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.4062887573444519,
+      "learning_rate": 0.00015239226473363687,
+      "loss": 1.1334,
+      "step": 1871
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.44016997634578336,
+      "learning_rate": 0.00015234142141431323,
+      "loss": 1.2477,
+      "step": 1872
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5472151435698095,
+      "learning_rate": 0.000152290559452089,
+      "loss": 1.2459,
+      "step": 1873
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.3781511561416589,
+      "learning_rate": 0.00015223967886508012,
+      "loss": 1.1836,
+      "step": 1874
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.48029839504475597,
+      "learning_rate": 0.0001521887796714092,
+      "loss": 1.0544,
+      "step": 1875
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.4189856038280378,
+      "learning_rate": 0.00015213786188920546,
+      "loss": 1.158,
+      "step": 1876
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.42372926468669886,
+      "learning_rate": 0.00015208692553660467,
+      "loss": 1.2991,
+      "step": 1877
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.4161377632755014,
+      "learning_rate": 0.0001520359706317493,
+      "loss": 1.0662,
+      "step": 1878
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5093081665712184,
+      "learning_rate": 0.0001519849971927884,
+      "loss": 1.1562,
+      "step": 1879
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.41178926759112744,
+      "learning_rate": 0.00015193400523787768,
+      "loss": 1.1868,
+      "step": 1880
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.44061507464088345,
+      "learning_rate": 0.00015188299478517933,
+      "loss": 1.1382,
+      "step": 1881
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.45679496810634496,
+      "learning_rate": 0.00015183196585286218,
+      "loss": 1.2975,
+      "step": 1882
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5902900174358077,
+      "learning_rate": 0.0001517809184591017,
+      "loss": 1.0686,
+      "step": 1883
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.4690175343517927,
+      "learning_rate": 0.00015172985262207987,
+      "loss": 1.233,
+      "step": 1884
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.46177547812349873,
+      "learning_rate": 0.00015167876835998524,
+      "loss": 1.1486,
+      "step": 1885
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.40049337780514066,
+      "learning_rate": 0.0001516276656910129,
+      "loss": 1.3346,
+      "step": 1886
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.403205451583982,
+      "learning_rate": 0.0001515765446333646,
+      "loss": 1.3053,
+      "step": 1887
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.4783087195150144,
+      "learning_rate": 0.0001515254052052486,
+      "loss": 1.0576,
+      "step": 1888
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.4954215526153359,
+      "learning_rate": 0.00015147424742487957,
+      "loss": 1.2103,
+      "step": 1889
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.4536063750540781,
+      "learning_rate": 0.00015142307131047885,
+      "loss": 1.1812,
+      "step": 1890
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.452647384227483,
+      "learning_rate": 0.00015137187688027436,
+      "loss": 1.123,
+      "step": 1891
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.4773042167122206,
+      "learning_rate": 0.00015132066415250042,
+      "loss": 1.2182,
+      "step": 1892
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.38238591411014256,
+      "learning_rate": 0.00015126943314539788,
+      "loss": 1.3002,
+      "step": 1893
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.4864503155265189,
+      "learning_rate": 0.00015121818387721416,
+      "loss": 1.1187,
+      "step": 1894
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.43476472505547376,
+      "learning_rate": 0.00015116691636620318,
+      "loss": 1.0882,
+      "step": 1895
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.4168642493440805,
+      "learning_rate": 0.00015111563063062534,
+      "loss": 1.0001,
+      "step": 1896
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.44706110813600564,
+      "learning_rate": 0.00015106432668874747,
+      "loss": 1.2542,
+      "step": 1897
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.4559353896828977,
+      "learning_rate": 0.00015101300455884297,
+      "loss": 1.1613,
+      "step": 1898
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.47757104998820776,
+      "learning_rate": 0.00015096166425919175,
+      "loss": 1.2109,
+      "step": 1899
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5263530091765437,
+      "learning_rate": 0.0001509103058080801,
+      "loss": 1.2163,
+      "step": 1900
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.4446971893132144,
+      "learning_rate": 0.00015085892922380074,
+      "loss": 1.173,
+      "step": 1901
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.4212442239319362,
+      "learning_rate": 0.00015080753452465296,
+      "loss": 1.1414,
+      "step": 1902
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.4330071521141245,
+      "learning_rate": 0.00015075612172894246,
+      "loss": 1.1926,
+      "step": 1903
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.4566322386407614,
+      "learning_rate": 0.00015070469085498144,
+      "loss": 1.1927,
+      "step": 1904
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.45177939391149763,
+      "learning_rate": 0.00015065324192108838,
+      "loss": 1.1079,
+      "step": 1905
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.432195499612258,
+      "learning_rate": 0.0001506017749455884,
+      "loss": 1.2328,
+      "step": 1906
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.49886142305453796,
+      "learning_rate": 0.00015055028994681284,
+      "loss": 1.1253,
+      "step": 1907
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.42300872084521757,
+      "learning_rate": 0.00015049878694309965,
+      "loss": 1.1615,
+      "step": 1908
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.4281858228518857,
+      "learning_rate": 0.00015044726595279305,
+      "loss": 1.1715,
+      "step": 1909
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.4183052953008053,
+      "learning_rate": 0.00015039572699424375,
+      "loss": 1.2111,
+      "step": 1910
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.3965680338366964,
+      "learning_rate": 0.00015034417008580875,
+      "loss": 1.209,
+      "step": 1911
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.4475619870454584,
+      "learning_rate": 0.0001502925952458517,
+      "loss": 1.3387,
+      "step": 1912
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.4643328345159328,
+      "learning_rate": 0.00015024100249274227,
+      "loss": 1.1267,
+      "step": 1913
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.4290382371183616,
+      "learning_rate": 0.00015018939184485678,
+      "loss": 1.2594,
+      "step": 1914
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.4849038006596817,
+      "learning_rate": 0.00015013776332057786,
+      "loss": 1.2492,
+      "step": 1915
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.4733322993368316,
+      "learning_rate": 0.00015008611693829446,
+      "loss": 1.279,
+      "step": 1916
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.4108683424711202,
+      "learning_rate": 0.00015003445271640197,
+      "loss": 1.2101,
+      "step": 1917
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.41024555100103544,
+      "learning_rate": 0.00014998277067330203,
+      "loss": 1.1817,
+      "step": 1918
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.3983299009673343,
+      "learning_rate": 0.0001499310708274027,
+      "loss": 1.0965,
+      "step": 1919
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6801309889720061,
+      "learning_rate": 0.00014987935319711842,
+      "loss": 1.223,
+      "step": 1920
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.4837569226775585,
+      "learning_rate": 0.00014982761780086987,
+      "loss": 1.2131,
+      "step": 1921
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.4531011040518916,
+      "learning_rate": 0.00014977586465708407,
+      "loss": 1.1758,
+      "step": 1922
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.42977845802752546,
+      "learning_rate": 0.0001497240937841944,
+      "loss": 1.1628,
+      "step": 1923
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.41362111436746535,
+      "learning_rate": 0.00014967230520064056,
+      "loss": 1.2474,
+      "step": 1924
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.4033875139154264,
+      "learning_rate": 0.0001496204989248685,
+      "loss": 1.2673,
+      "step": 1925
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.441859513478225,
+      "learning_rate": 0.00014956867497533058,
+      "loss": 1.1499,
+      "step": 1926
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.45076823072154315,
+      "learning_rate": 0.00014951683337048537,
+      "loss": 1.2494,
+      "step": 1927
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.4051716222595987,
+      "learning_rate": 0.00014946497412879773,
+      "loss": 1.2344,
+      "step": 1928
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.43673770881563884,
+      "learning_rate": 0.00014941309726873879,
+      "loss": 1.1281,
+      "step": 1929
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5287398899958117,
+      "learning_rate": 0.00014936120280878605,
+      "loss": 1.1814,
+      "step": 1930
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5143095286899425,
+      "learning_rate": 0.00014930929076742316,
+      "loss": 1.1345,
+      "step": 1931
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.449801179331484,
+      "learning_rate": 0.0001492573611631401,
+      "loss": 1.1745,
+      "step": 1932
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.3835718793304017,
+      "learning_rate": 0.00014920541401443313,
+      "loss": 1.2627,
+      "step": 1933
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.4430986362251494,
+      "learning_rate": 0.0001491534493398046,
+      "loss": 1.1686,
+      "step": 1934
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.4590057781350295,
+      "learning_rate": 0.00014910146715776338,
+      "loss": 1.2402,
+      "step": 1935
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.42343476018884374,
+      "learning_rate": 0.00014904946748682432,
+      "loss": 1.3542,
+      "step": 1936
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.3911786970672487,
+      "learning_rate": 0.0001489974503455086,
+      "loss": 1.2002,
+      "step": 1937
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.4178600736793033,
+      "learning_rate": 0.00014894541575234365,
+      "loss": 1.3003,
+      "step": 1938
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.41730316862690714,
+      "learning_rate": 0.00014889336372586305,
+      "loss": 1.2897,
+      "step": 1939
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.518035985955232,
+      "learning_rate": 0.00014884129428460665,
+      "loss": 1.1333,
+      "step": 1940
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.46593492460215885,
+      "learning_rate": 0.0001487892074471205,
+      "loss": 1.1883,
+      "step": 1941
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.4195979815546044,
+      "learning_rate": 0.00014873710323195675,
+      "loss": 1.2355,
+      "step": 1942
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5101880438079442,
+      "learning_rate": 0.00014868498165767388,
+      "loss": 1.0926,
+      "step": 1943
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.45538067268868443,
+      "learning_rate": 0.00014863284274283648,
+      "loss": 1.2246,
+      "step": 1944
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5337530859388241,
+      "learning_rate": 0.00014858068650601527,
+      "loss": 1.1541,
+      "step": 1945
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.4603507516121734,
+      "learning_rate": 0.00014852851296578724,
+      "loss": 1.0612,
+      "step": 1946
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.424683262960206,
+      "learning_rate": 0.00014847632214073548,
+      "loss": 1.2838,
+      "step": 1947
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.38651361609692203,
+      "learning_rate": 0.00014842411404944927,
+      "loss": 1.2154,
+      "step": 1948
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.41669934681263837,
+      "learning_rate": 0.000148371888710524,
+      "loss": 1.145,
+      "step": 1949
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.4454383562989557,
+      "learning_rate": 0.0001483196461425612,
+      "loss": 1.1362,
+      "step": 1950
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.4379052642672101,
+      "learning_rate": 0.00014826738636416863,
+      "loss": 1.1242,
+      "step": 1951
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.43239875557072943,
+      "learning_rate": 0.00014821510939396003,
+      "loss": 1.0071,
+      "step": 1952
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.39212590923102636,
+      "learning_rate": 0.00014816281525055543,
+      "loss": 1.1451,
+      "step": 1953
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.4743606089579671,
+      "learning_rate": 0.00014811050395258083,
+      "loss": 1.0627,
+      "step": 1954
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.45428885599760793,
+      "learning_rate": 0.00014805817551866838,
+      "loss": 1.1394,
+      "step": 1955
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.4161681553875841,
+      "learning_rate": 0.00014800582996745643,
+      "loss": 1.0969,
+      "step": 1956
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.4614419694258719,
+      "learning_rate": 0.0001479534673175893,
+      "loss": 1.0918,
+      "step": 1957
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.3963790906356942,
+      "learning_rate": 0.00014790108758771747,
+      "loss": 1.2134,
+      "step": 1958
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.42580798074739523,
+      "learning_rate": 0.0001478486907964975,
+      "loss": 1.034,
+      "step": 1959
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.42260442306614576,
+      "learning_rate": 0.00014779627696259195,
+      "loss": 1.094,
+      "step": 1960
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.42933351207350806,
+      "learning_rate": 0.0001477438461046696,
+      "loss": 1.2703,
+      "step": 1961
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.3687565744048426,
+      "learning_rate": 0.00014769139824140516,
+      "loss": 1.1558,
+      "step": 1962
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.4251222792409212,
+      "learning_rate": 0.0001476389333914794,
+      "loss": 1.0843,
+      "step": 1963
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.497479336084553,
+      "learning_rate": 0.00014758645157357926,
+      "loss": 1.3438,
+      "step": 1964
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.4340108469202065,
+      "learning_rate": 0.00014753395280639762,
+      "loss": 1.3312,
+      "step": 1965
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.41621754650227877,
+      "learning_rate": 0.0001474814371086334,
+      "loss": 1.1633,
+      "step": 1966
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.4024185121505323,
+      "learning_rate": 0.00014742890449899161,
+      "loss": 1.0137,
+      "step": 1967
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6300764151325301,
+      "learning_rate": 0.0001473763549961832,
+      "loss": 1.0168,
+      "step": 1968
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.3879186912200289,
+      "learning_rate": 0.00014732378861892525,
+      "loss": 1.1856,
+      "step": 1969
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.4290987997560587,
+      "learning_rate": 0.00014727120538594074,
+      "loss": 1.1194,
+      "step": 1970
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.44729173921134135,
+      "learning_rate": 0.0001472186053159587,
+      "loss": 1.1183,
+      "step": 1971
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5083750926437817,
+      "learning_rate": 0.00014716598842771413,
+      "loss": 1.2775,
+      "step": 1972
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.4311565817104019,
+      "learning_rate": 0.00014711335473994815,
+      "loss": 1.1038,
+      "step": 1973
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.4245987087819015,
+      "learning_rate": 0.00014706070427140765,
+      "loss": 1.2091,
+      "step": 1974
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.47106528785777924,
+      "learning_rate": 0.00014700803704084564,
+      "loss": 1.3143,
+      "step": 1975
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.41967888938921905,
+      "learning_rate": 0.00014695535306702109,
+      "loss": 1.2708,
+      "step": 1976
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.37331886997936264,
+      "learning_rate": 0.00014690265236869888,
+      "loss": 1.1909,
+      "step": 1977
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.42977188943761996,
+      "learning_rate": 0.0001468499349646499,
+      "loss": 1.0961,
+      "step": 1978
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.4406669021902194,
+      "learning_rate": 0.00014679720087365096,
+      "loss": 1.0575,
+      "step": 1979
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.4840325490072921,
+      "learning_rate": 0.0001467444501144848,
+      "loss": 1.0634,
+      "step": 1980
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5360417613502011,
+      "learning_rate": 0.00014669168270594018,
+      "loss": 1.006,
+      "step": 1981
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.4619218485742844,
+      "learning_rate": 0.00014663889866681164,
+      "loss": 1.1678,
+      "step": 1982
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.38669510285947106,
+      "learning_rate": 0.00014658609801589982,
+      "loss": 1.1842,
+      "step": 1983
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5511439011794443,
+      "learning_rate": 0.00014653328077201115,
+      "loss": 1.2325,
+      "step": 1984
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.502009682564419,
+      "learning_rate": 0.00014648044695395797,
+      "loss": 1.058,
+      "step": 1985
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.46719997165404037,
+      "learning_rate": 0.0001464275965805586,
+      "loss": 1.1323,
+      "step": 1986
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.4819439813234482,
+      "learning_rate": 0.0001463747296706372,
+      "loss": 1.0595,
+      "step": 1987
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.38116268228025996,
+      "learning_rate": 0.0001463218462430239,
+      "loss": 1.3415,
+      "step": 1988
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.3964688073413427,
+      "learning_rate": 0.00014626894631655457,
+      "loss": 1.3577,
+      "step": 1989
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.4272823103509901,
+      "learning_rate": 0.0001462160299100711,
+      "loss": 1.2185,
+      "step": 1990
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.4394895632255822,
+      "learning_rate": 0.00014616309704242113,
+      "loss": 1.1697,
+      "step": 1991
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.3887096929472129,
+      "learning_rate": 0.00014611014773245823,
+      "loss": 1.1859,
+      "step": 1992
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.49258297956577307,
+      "learning_rate": 0.0001460571819990419,
+      "loss": 1.2449,
+      "step": 1993
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.4781892992391851,
+      "learning_rate": 0.00014600419986103734,
+      "loss": 1.3137,
+      "step": 1994
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.3828936101212727,
+      "learning_rate": 0.00014595120133731565,
+      "loss": 1.1736,
+      "step": 1995
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.495633724117446,
+      "learning_rate": 0.00014589818644675378,
+      "loss": 1.1558,
+      "step": 1996
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.4679566653181731,
+      "learning_rate": 0.00014584515520823453,
+      "loss": 1.1162,
+      "step": 1997
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.4322032934941436,
+      "learning_rate": 0.0001457921076406465,
+      "loss": 1.1289,
+      "step": 1998
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.428901899573165,
+      "learning_rate": 0.00014573904376288409,
+      "loss": 1.2856,
+      "step": 1999
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.476533454151738,
+      "learning_rate": 0.0001456859635938475,
+      "loss": 1.1398,
+      "step": 2000
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5714687165811031,
+      "learning_rate": 0.00014563286715244282,
+      "loss": 1.1705,
+      "step": 2001
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.4493881047863828,
+      "learning_rate": 0.00014557975445758181,
+      "loss": 1.0736,
+      "step": 2002
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.42531343203553684,
+      "learning_rate": 0.0001455266255281821,
+      "loss": 1.1834,
+      "step": 2003
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.3895761564960627,
+      "learning_rate": 0.00014547348038316713,
+      "loss": 1.2871,
+      "step": 2004
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.41994938142892846,
+      "learning_rate": 0.00014542031904146602,
+      "loss": 1.2712,
+      "step": 2005
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.4843061857191946,
+      "learning_rate": 0.00014536714152201373,
+      "loss": 1.3542,
+      "step": 2006
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.4133966291530044,
+      "learning_rate": 0.00014531394784375094,
+      "loss": 1.2285,
+      "step": 2007
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.475568554874142,
+      "learning_rate": 0.0001452607380256241,
+      "loss": 1.097,
+      "step": 2008
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.4983732838686681,
+      "learning_rate": 0.00014520751208658546,
+      "loss": 1.2285,
+      "step": 2009
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.4458177092105337,
+      "learning_rate": 0.00014515427004559292,
+      "loss": 1.2223,
+      "step": 2010
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.43373049691385607,
+      "learning_rate": 0.00014510101192161018,
+      "loss": 1.1787,
+      "step": 2011
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.41977017233216923,
+      "learning_rate": 0.00014504773773360666,
+      "loss": 1.1333,
+      "step": 2012
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.42818567386592754,
+      "learning_rate": 0.00014499444750055754,
+      "loss": 1.2326,
+      "step": 2013
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.38791070496218877,
+      "learning_rate": 0.00014494114124144353,
+      "loss": 1.1808,
+      "step": 2014
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.3995556046964628,
+      "learning_rate": 0.00014488781897525131,
+      "loss": 1.1319,
+      "step": 2015
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.47107399508187464,
+      "learning_rate": 0.00014483448072097308,
+      "loss": 1.1345,
+      "step": 2016
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.45211882272172266,
+      "learning_rate": 0.0001447811264976068,
+      "loss": 1.2496,
+      "step": 2017
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.4181621226659709,
+      "learning_rate": 0.00014472775632415615,
+      "loss": 1.2853,
+      "step": 2018
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.40055099303114444,
+      "learning_rate": 0.0001446743702196304,
+      "loss": 1.2099,
+      "step": 2019
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.42334921697888617,
+      "learning_rate": 0.00014462096820304455,
+      "loss": 1.1243,
+      "step": 2020
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.37735792448297806,
+      "learning_rate": 0.00014456755029341924,
+      "loss": 1.321,
+      "step": 2021
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.39797718681908345,
+      "learning_rate": 0.00014451411650978087,
+      "loss": 1.1396,
+      "step": 2022
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.4687804886919832,
+      "learning_rate": 0.00014446066687116136,
+      "loss": 1.0172,
+      "step": 2023
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.4384218888681426,
+      "learning_rate": 0.00014440720139659833,
+      "loss": 1.0314,
+      "step": 2024
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.45378302586998726,
+      "learning_rate": 0.00014435372010513509,
+      "loss": 1.0403,
+      "step": 2025
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.49758897449952477,
+      "learning_rate": 0.0001443002230158205,
+      "loss": 1.1024,
+      "step": 2026
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.41554427267819605,
+      "learning_rate": 0.00014424671014770906,
+      "loss": 1.1206,
+      "step": 2027
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5196056861247762,
+      "learning_rate": 0.00014419318151986097,
+      "loss": 1.4468,
+      "step": 2028
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.42157374613527815,
+      "learning_rate": 0.00014413963715134196,
+      "loss": 1.2139,
+      "step": 2029
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.47132704150992716,
+      "learning_rate": 0.0001440860770612234,
+      "loss": 1.1976,
+      "step": 2030
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.42775283655997665,
+      "learning_rate": 0.00014403250126858224,
+      "loss": 1.247,
+      "step": 2031
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.3947835781965834,
+      "learning_rate": 0.00014397890979250108,
+      "loss": 1.0938,
+      "step": 2032
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.42482339792458174,
+      "learning_rate": 0.000143925302652068,
+      "loss": 1.2048,
+      "step": 2033
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.4467224457702705,
+      "learning_rate": 0.0001438716798663768,
+      "loss": 1.2378,
+      "step": 2034
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5210226790191018,
+      "learning_rate": 0.0001438180414545267,
+      "loss": 1.0891,
+      "step": 2035
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.4821385223625234,
+      "learning_rate": 0.00014376438743562262,
+      "loss": 1.1275,
+      "step": 2036
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.437227875908001,
+      "learning_rate": 0.00014371071782877495,
+      "loss": 1.0614,
+      "step": 2037
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.462537063408154,
+      "learning_rate": 0.00014365703265309968,
+      "loss": 1.2463,
+      "step": 2038
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.4123142755376291,
+      "learning_rate": 0.0001436033319277183,
+      "loss": 1.2594,
+      "step": 2039
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.4108573431112248,
+      "learning_rate": 0.00014354961567175786,
+      "loss": 1.1303,
+      "step": 2040
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.4047684096186533,
+      "learning_rate": 0.000143495883904351,
+      "loss": 1.2302,
+      "step": 2041
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.4733576267408301,
+      "learning_rate": 0.00014344213664463576,
+      "loss": 1.1735,
+      "step": 2042
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.4504449193012932,
+      "learning_rate": 0.00014338837391175582,
+      "loss": 1.1814,
+      "step": 2043
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.4516836616690366,
+      "learning_rate": 0.00014333459572486032,
+      "loss": 1.1426,
+      "step": 2044
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.39648388833908815,
+      "learning_rate": 0.00014328080210310388,
+      "loss": 1.3325,
+      "step": 2045
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.428898663585617,
+      "learning_rate": 0.00014322699306564668,
+      "loss": 1.1872,
+      "step": 2046
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.48299668438756477,
+      "learning_rate": 0.00014317316863165434,
+      "loss": 1.2536,
+      "step": 2047
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.4776227787719985,
+      "learning_rate": 0.00014311932882029796,
+      "loss": 1.1067,
+      "step": 2048
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.42723633751258233,
+      "learning_rate": 0.00014306547365075414,
+      "loss": 1.1396,
+      "step": 2049
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.40096278509477656,
+      "learning_rate": 0.00014301160314220498,
+      "loss": 1.0754,
+      "step": 2050
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.40777657401575723,
+      "learning_rate": 0.00014295771731383797,
+      "loss": 1.1538,
+      "step": 2051
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.43987083772822044,
+      "learning_rate": 0.0001429038161848461,
+      "loss": 1.2251,
+      "step": 2052
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.4430136962510553,
+      "learning_rate": 0.0001428498997744278,
+      "loss": 1.4366,
+      "step": 2053
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.40056338716457773,
+      "learning_rate": 0.000142795968101787,
+      "loss": 1.245,
+      "step": 2054
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.472545945490941,
+      "learning_rate": 0.00014274202118613294,
+      "loss": 0.9885,
+      "step": 2055
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.43553021457771834,
+      "learning_rate": 0.0001426880590466804,
+      "loss": 1.0913,
+      "step": 2056
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5214309886997988,
+      "learning_rate": 0.00014263408170264955,
+      "loss": 1.0481,
+      "step": 2057
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5142805798312672,
+      "learning_rate": 0.00014258008917326597,
+      "loss": 1.1004,
+      "step": 2058
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.45526096849635894,
+      "learning_rate": 0.00014252608147776065,
+      "loss": 1.2382,
+      "step": 2059
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.4574717547726295,
+      "learning_rate": 0.00014247205863536996,
+      "loss": 1.3829,
+      "step": 2060
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.44843848073072684,
+      "learning_rate": 0.0001424180206653357,
+      "loss": 1.1301,
+      "step": 2061
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.42295695832628283,
+      "learning_rate": 0.0001423639675869051,
+      "loss": 1.1851,
+      "step": 2062
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.43622814813648286,
+      "learning_rate": 0.00014230989941933065,
+      "loss": 1.107,
+      "step": 2063
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.4260940742424721,
+      "learning_rate": 0.00014225581618187026,
+      "loss": 1.3325,
+      "step": 2064
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.4544376304115898,
+      "learning_rate": 0.00014220171789378728,
+      "loss": 1.0749,
+      "step": 2065
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.42340907849646053,
+      "learning_rate": 0.00014214760457435037,
+      "loss": 1.058,
+      "step": 2066
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.49711688005432764,
+      "learning_rate": 0.0001420934762428335,
+      "loss": 1.0468,
+      "step": 2067
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5668559241720175,
+      "learning_rate": 0.00014203933291851608,
+      "loss": 1.1393,
+      "step": 2068
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.4317449415917551,
+      "learning_rate": 0.00014198517462068279,
+      "loss": 1.2884,
+      "step": 2069
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.4166986225207518,
+      "learning_rate": 0.00014193100136862365,
+      "loss": 1.1361,
+      "step": 2070
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.42267452963106167,
+      "learning_rate": 0.00014187681318163404,
+      "loss": 1.1966,
+      "step": 2071
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.3978012914688716,
+      "learning_rate": 0.00014182261007901462,
+      "loss": 1.1867,
+      "step": 2072
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.4560100122185351,
+      "learning_rate": 0.0001417683920800714,
+      "loss": 1.1344,
+      "step": 2073
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.46045857760979747,
+      "learning_rate": 0.00014171415920411566,
+      "loss": 1.0833,
+      "step": 2074
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.4499621537026508,
+      "learning_rate": 0.00014165991147046403,
+      "loss": 1.0548,
+      "step": 2075
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5211818610069717,
+      "learning_rate": 0.00014160564889843832,
+      "loss": 1.2278,
+      "step": 2076
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.4050493875413598,
+      "learning_rate": 0.0001415513715073658,
+      "loss": 1.258,
+      "step": 2077
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.3953590583029378,
+      "learning_rate": 0.00014149707931657886,
+      "loss": 1.3745,
+      "step": 2078
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.42233096372056284,
+      "learning_rate": 0.0001414427723454153,
+      "loss": 1.1051,
+      "step": 2079
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.4283417972853378,
+      "learning_rate": 0.00014138845061321796,
+      "loss": 1.1798,
+      "step": 2080
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.41729442045363474,
+      "learning_rate": 0.00014133411413933523,
+      "loss": 1.2146,
+      "step": 2081
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.40304489388571396,
+      "learning_rate": 0.00014127976294312053,
+      "loss": 1.1939,
+      "step": 2082
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.4376013679911511,
+      "learning_rate": 0.00014122539704393265,
+      "loss": 1.0692,
+      "step": 2083
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.46886531524670644,
+      "learning_rate": 0.0001411710164611355,
+      "loss": 1.1271,
+      "step": 2084
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.4216546991554233,
+      "learning_rate": 0.00014111662121409833,
+      "loss": 1.2536,
+      "step": 2085
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.520025504662005,
+      "learning_rate": 0.0001410622113221956,
+      "loss": 1.1404,
+      "step": 2086
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.44693803678612276,
+      "learning_rate": 0.0001410077868048069,
+      "loss": 1.2274,
+      "step": 2087
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6168488957323836,
+      "learning_rate": 0.0001409533476813171,
+      "loss": 1.2941,
+      "step": 2088
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.4456959520311065,
+      "learning_rate": 0.00014089889397111624,
+      "loss": 1.1673,
+      "step": 2089
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.3975198877688334,
+      "learning_rate": 0.00014084442569359964,
+      "loss": 1.0625,
+      "step": 2090
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.4274429610099362,
+      "learning_rate": 0.00014078994286816768,
+      "loss": 1.2436,
+      "step": 2091
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.39277319825876456,
+      "learning_rate": 0.00014073544551422597,
+      "loss": 1.3199,
+      "step": 2092
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.4746273476309984,
+      "learning_rate": 0.00014068093365118535,
+      "loss": 1.14,
+      "step": 2093
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.48109057335569855,
+      "learning_rate": 0.00014062640729846178,
+      "loss": 1.2096,
+      "step": 2094
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.3990852941910486,
+      "learning_rate": 0.0001405718664754764,
+      "loss": 1.1907,
+      "step": 2095
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.45887023426952517,
+      "learning_rate": 0.00014051731120165542,
+      "loss": 1.2339,
+      "step": 2096
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5232148902283226,
+      "learning_rate": 0.00014046274149643036,
+      "loss": 1.4206,
+      "step": 2097
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.3990277176905189,
+      "learning_rate": 0.0001404081573792377,
+      "loss": 1.3099,
+      "step": 2098
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5543733350471501,
+      "learning_rate": 0.00014035355886951923,
+      "loss": 1.2723,
+      "step": 2099
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.4840418728767712,
+      "learning_rate": 0.00014029894598672173,
+      "loss": 1.2613,
+      "step": 2100
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5012446836617263,
+      "learning_rate": 0.00014024431875029713,
+      "loss": 1.1287,
+      "step": 2101
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.43923761586595156,
+      "learning_rate": 0.00014018967717970254,
+      "loss": 1.2647,
+      "step": 2102
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.48632448419690627,
+      "learning_rate": 0.00014013502129440005,
+      "loss": 1.0771,
+      "step": 2103
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.46370738892520813,
+      "learning_rate": 0.000140080351113857,
+      "loss": 1.2307,
+      "step": 2104
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.3833431715038961,
+      "learning_rate": 0.0001400256666575457,
+      "loss": 1.1346,
+      "step": 2105
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.42721112198517747,
+      "learning_rate": 0.0001399709679449436,
+      "loss": 1.2276,
+      "step": 2106
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.41537102332741443,
+      "learning_rate": 0.00013991625499553325,
+      "loss": 1.1035,
+      "step": 2107
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.3951765864247545,
+      "learning_rate": 0.0001398615278288022,
+      "loss": 1.0971,
+      "step": 2108
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.41027020931282177,
+      "learning_rate": 0.0001398067864642431,
+      "loss": 1.1831,
+      "step": 2109
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.4245298045728807,
+      "learning_rate": 0.00013975203092135365,
+      "loss": 1.0441,
+      "step": 2110
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.4378410435098881,
+      "learning_rate": 0.00013969726121963667,
+      "loss": 1.2161,
+      "step": 2111
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.40356897567387556,
+      "learning_rate": 0.0001396424773785999,
+      "loss": 1.2449,
+      "step": 2112
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.45894042767557475,
+      "learning_rate": 0.00013958767941775618,
+      "loss": 0.9754,
+      "step": 2113
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.44838734880223,
+      "learning_rate": 0.00013953286735662342,
+      "loss": 1.3496,
+      "step": 2114
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.44342987668258077,
+      "learning_rate": 0.0001394780412147245,
+      "loss": 1.3417,
+      "step": 2115
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.4508764790355972,
+      "learning_rate": 0.00013942320101158732,
+      "loss": 1.0407,
+      "step": 2116
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.46591267557404337,
+      "learning_rate": 0.00013936834676674478,
+      "loss": 1.2324,
+      "step": 2117
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.4553712938830467,
+      "learning_rate": 0.00013931347849973475,
+      "loss": 1.0759,
+      "step": 2118
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.4505340470808664,
+      "learning_rate": 0.00013925859623010024,
+      "loss": 1.184,
+      "step": 2119
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.41381966197255715,
+      "learning_rate": 0.00013920369997738908,
+      "loss": 1.0514,
+      "step": 2120
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.4090819758992069,
+      "learning_rate": 0.0001391487897611542,
+      "loss": 1.2324,
+      "step": 2121
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.483118978002014,
+      "learning_rate": 0.0001390938656009534,
+      "loss": 1.2908,
+      "step": 2122
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.45433734385977953,
+      "learning_rate": 0.00013903892751634947,
+      "loss": 1.2154,
+      "step": 2123
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.44835116244014017,
+      "learning_rate": 0.0001389839755269103,
+      "loss": 1.2523,
+      "step": 2124
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.4923764833745098,
+      "learning_rate": 0.0001389290096522085,
+      "loss": 1.256,
+      "step": 2125
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.4303835622529688,
+      "learning_rate": 0.00013887402991182178,
+      "loss": 1.0648,
+      "step": 2126
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.42305389941859034,
+      "learning_rate": 0.0001388190363253328,
+      "loss": 1.257,
+      "step": 2127
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.42488071561950214,
+      "learning_rate": 0.00013876402891232906,
+      "loss": 1.3437,
+      "step": 2128
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.44124078102823583,
+      "learning_rate": 0.00013870900769240305,
+      "loss": 1.0686,
+      "step": 2129
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.4126028331152009,
+      "learning_rate": 0.00013865397268515215,
+      "loss": 1.1828,
+      "step": 2130
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.40126016141083426,
+      "learning_rate": 0.00013859892391017865,
+      "loss": 1.1717,
+      "step": 2131
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.43236804674821105,
+      "learning_rate": 0.00013854386138708978,
+      "loss": 1.0394,
+      "step": 2132
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.4613299485652511,
+      "learning_rate": 0.0001384887851354976,
+      "loss": 1.182,
+      "step": 2133
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.39305967011139853,
+      "learning_rate": 0.00013843369517501912,
+      "loss": 1.0749,
+      "step": 2134
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.4313555303458508,
+      "learning_rate": 0.00013837859152527623,
+      "loss": 0.9969,
+      "step": 2135
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.4252232760747602,
+      "learning_rate": 0.00013832347420589564,
+      "loss": 1.1047,
+      "step": 2136
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.44741169593442914,
+      "learning_rate": 0.000138268343236509,
+      "loss": 1.2943,
+      "step": 2137
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.42736282203157966,
+      "learning_rate": 0.00013821319863675276,
+      "loss": 1.0845,
+      "step": 2138
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.4871619491542356,
+      "learning_rate": 0.00013815804042626828,
+      "loss": 1.1146,
+      "step": 2139
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.44025546282516215,
+      "learning_rate": 0.0001381028686247017,
+      "loss": 1.2546,
+      "step": 2140
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.3810244379420192,
+      "learning_rate": 0.00013804768325170408,
+      "loss": 1.2996,
+      "step": 2141
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.4073055924155801,
+      "learning_rate": 0.00013799248432693127,
+      "loss": 1.1,
+      "step": 2142
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.4476049243795729,
+      "learning_rate": 0.00013793727187004395,
+      "loss": 0.9695,
+      "step": 2143
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.3892626426918247,
+      "learning_rate": 0.00013788204590070764,
+      "loss": 1.1059,
+      "step": 2144
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.43835649643647534,
+      "learning_rate": 0.0001378268064385926,
+      "loss": 1.2806,
+      "step": 2145
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.4684932191264051,
+      "learning_rate": 0.000137771553503374,
+      "loss": 1.235,
+      "step": 2146
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.4720338468744819,
+      "learning_rate": 0.00013771628711473172,
+      "loss": 0.9987,
+      "step": 2147
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.4875175289577886,
+      "learning_rate": 0.00013766100729235053,
+      "loss": 1.2227,
+      "step": 2148
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.4424363656436083,
+      "learning_rate": 0.0001376057140559199,
+      "loss": 0.9542,
+      "step": 2149
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.4338194926721232,
+      "learning_rate": 0.00013755040742513406,
+      "loss": 1.2487,
+      "step": 2150
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.4400498200126344,
+      "learning_rate": 0.00013749508741969213,
+      "loss": 1.2561,
+      "step": 2151
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.4543030721182541,
+      "learning_rate": 0.00013743975405929786,
+      "loss": 1.0126,
+      "step": 2152
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.3984818779169799,
+      "learning_rate": 0.00013738440736365984,
+      "loss": 1.2968,
+      "step": 2153
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.4358834340144631,
+      "learning_rate": 0.0001373290473524914,
+      "loss": 1.041,
+      "step": 2154
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5669485816020295,
+      "learning_rate": 0.00013727367404551055,
+      "loss": 1.0079,
+      "step": 2155
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.4924697478625036,
+      "learning_rate": 0.00013721828746244015,
+      "loss": 1.0878,
+      "step": 2156
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.42340822286675583,
+      "learning_rate": 0.00013716288762300765,
+      "loss": 1.2347,
+      "step": 2157
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.4407788767395031,
+      "learning_rate": 0.00013710747454694536,
+      "loss": 1.0388,
+      "step": 2158
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.446150022230011,
+      "learning_rate": 0.0001370520482539902,
+      "loss": 1.2069,
+      "step": 2159
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.49628395520590723,
+      "learning_rate": 0.00013699660876388383,
+      "loss": 1.1295,
+      "step": 2160
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.3747666383554271,
+      "learning_rate": 0.00013694115609637263,
+      "loss": 1.0942,
+      "step": 2161
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.4302086533713984,
+      "learning_rate": 0.00013688569027120768,
+      "loss": 1.1553,
+      "step": 2162
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5114022597206201,
+      "learning_rate": 0.0001368302113081447,
+      "loss": 1.2269,
+      "step": 2163
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.4159446031543904,
+      "learning_rate": 0.00013677471922694418,
+      "loss": 1.0447,
+      "step": 2164
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.4405149638496812,
+      "learning_rate": 0.0001367192140473711,
+      "loss": 0.8909,
+      "step": 2165
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.42334459311655487,
+      "learning_rate": 0.00013666369578919528,
+      "loss": 1.1039,
+      "step": 2166
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.4408961373432183,
+      "learning_rate": 0.0001366081644721912,
+      "loss": 1.2294,
+      "step": 2167
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.3911927811201082,
+      "learning_rate": 0.00013655262011613785,
+      "loss": 1.2088,
+      "step": 2168
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.42454750707896394,
+      "learning_rate": 0.000136497062740819,
+      "loss": 1.2016,
+      "step": 2169
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.45948460397354923,
+      "learning_rate": 0.00013644149236602298,
+      "loss": 1.1812,
+      "step": 2170
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.44170128460849506,
+      "learning_rate": 0.00013638590901154276,
+      "loss": 1.2851,
+      "step": 2171
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.4110473737727612,
+      "learning_rate": 0.000136330312697176,
+      "loss": 1.3798,
+      "step": 2172
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.4282282358670717,
+      "learning_rate": 0.0001362747034427249,
+      "loss": 1.0161,
+      "step": 2173
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.471068882771124,
+      "learning_rate": 0.00013621908126799622,
+      "loss": 1.0745,
+      "step": 2174
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.4246009257003082,
+      "learning_rate": 0.00013616344619280156,
+      "loss": 1.1961,
+      "step": 2175
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.4126798103670225,
+      "learning_rate": 0.0001361077982369568,
+      "loss": 1.124,
+      "step": 2176
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.4240922718189132,
+      "learning_rate": 0.0001360521374202826,
+      "loss": 1.1544,
+      "step": 2177
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.42868852274134095,
+      "learning_rate": 0.00013599646376260418,
+      "loss": 1.2661,
+      "step": 2178
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.45371690264723064,
+      "learning_rate": 0.00013594077728375128,
+      "loss": 1.3639,
+      "step": 2179
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.4082040949394695,
+      "learning_rate": 0.0001358850780035583,
+      "loss": 1.0738,
+      "step": 2180
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.40105434449767324,
+      "learning_rate": 0.0001358293659418641,
+      "loss": 1.1001,
+      "step": 2181
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.4936757867442796,
+      "learning_rate": 0.00013577364111851207,
+      "loss": 1.0647,
+      "step": 2182
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.47980574697502515,
+      "learning_rate": 0.0001357179035533503,
+      "loss": 1.1647,
+      "step": 2183
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.4279663136608911,
+      "learning_rate": 0.0001356621532662313,
+      "loss": 1.1321,
+      "step": 2184
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.39743283874506746,
+      "learning_rate": 0.00013560639027701215,
+      "loss": 1.1504,
+      "step": 2185
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.4091172266551295,
+      "learning_rate": 0.00013555061460555437,
+      "loss": 1.1512,
+      "step": 2186
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.4299901964129109,
+      "learning_rate": 0.0001354948262717241,
+      "loss": 1.179,
+      "step": 2187
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.38719266665135493,
+      "learning_rate": 0.00013543902529539206,
+      "loss": 1.2491,
+      "step": 2188
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.4414864229817937,
+      "learning_rate": 0.0001353832116964332,
+      "loss": 1.1698,
+      "step": 2189
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.39058085745860616,
+      "learning_rate": 0.00013532738549472723,
+      "loss": 1.193,
+      "step": 2190
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.44778516495745796,
+      "learning_rate": 0.00013527154671015825,
+      "loss": 1.0856,
+      "step": 2191
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.4343564652716607,
+      "learning_rate": 0.00013521569536261487,
+      "loss": 1.026,
+      "step": 2192
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.46495695667799214,
+      "learning_rate": 0.00013515983147199007,
+      "loss": 1.1047,
+      "step": 2193
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.49104145314693926,
+      "learning_rate": 0.00013510395505818146,
+      "loss": 1.1852,
+      "step": 2194
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.4407697468157063,
+      "learning_rate": 0.00013504806614109098,
+      "loss": 1.2198,
+      "step": 2195
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.38980846050788326,
+      "learning_rate": 0.00013499216474062512,
+      "loss": 1.1739,
+      "step": 2196
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.43786652452210956,
+      "learning_rate": 0.00013493625087669474,
+      "loss": 1.1637,
+      "step": 2197
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.504511914259361,
+      "learning_rate": 0.00013488032456921518,
+      "loss": 1.2451,
+      "step": 2198
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.4538979043170104,
+      "learning_rate": 0.00013482438583810614,
+      "loss": 1.2679,
+      "step": 2199
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.4000153155472987,
+      "learning_rate": 0.0001347684347032919,
+      "loss": 1.117,
+      "step": 2200
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.434505461978965,
+      "learning_rate": 0.00013471247118470103,
+      "loss": 1.1453,
+      "step": 2201
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.4273373020578937,
+      "learning_rate": 0.00013465649530226652,
+      "loss": 1.2433,
+      "step": 2202
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.39350080325109094,
+      "learning_rate": 0.0001346005070759258,
+      "loss": 1.1321,
+      "step": 2203
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.42749465165193545,
+      "learning_rate": 0.00013454450652562071,
+      "loss": 1.3551,
+      "step": 2204
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.4272587271399394,
+      "learning_rate": 0.0001344884936712974,
+      "loss": 1.1208,
+      "step": 2205
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.4119751957709363,
+      "learning_rate": 0.00013443246853290654,
+      "loss": 1.1337,
+      "step": 2206
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.3979413232915234,
+      "learning_rate": 0.00013437643113040301,
+      "loss": 1.2163,
+      "step": 2207
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.40052939179397384,
+      "learning_rate": 0.00013432038148374618,
+      "loss": 1.2217,
+      "step": 2208
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.3837169433529659,
+      "learning_rate": 0.00013426431961289975,
+      "loss": 1.424,
+      "step": 2209
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.4050696218132727,
+      "learning_rate": 0.0001342082455378317,
+      "loss": 1.2858,
+      "step": 2210
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.45419216086367026,
+      "learning_rate": 0.0001341521592785145,
+      "loss": 1.0937,
+      "step": 2211
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.40340993520000196,
+      "learning_rate": 0.00013409606085492486,
+      "loss": 1.2994,
+      "step": 2212
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.4198373144160479,
+      "learning_rate": 0.00013403995028704385,
+      "loss": 1.2329,
+      "step": 2213
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.4569021769736206,
+      "learning_rate": 0.00013398382759485683,
+      "loss": 1.2788,
+      "step": 2214
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5139114794208176,
+      "learning_rate": 0.00013392769279835354,
+      "loss": 1.2214,
+      "step": 2215
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.42112396172879196,
+      "learning_rate": 0.00013387154591752799,
+      "loss": 1.3423,
+      "step": 2216
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.3910227543831881,
+      "learning_rate": 0.00013381538697237853,
+      "loss": 1.0998,
+      "step": 2217
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.46136647181800594,
+      "learning_rate": 0.0001337592159829077,
+      "loss": 1.2152,
+      "step": 2218
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.4430057743521499,
+      "learning_rate": 0.00013370303296912249,
+      "loss": 1.324,
+      "step": 2219
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.395731712897511,
+      "learning_rate": 0.00013364683795103404,
+      "loss": 1.3274,
+      "step": 2220
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5607367655154614,
+      "learning_rate": 0.0001335906309486579,
+      "loss": 1.1801,
+      "step": 2221
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.42679307809515343,
+      "learning_rate": 0.00013353441198201373,
+      "loss": 1.0995,
+      "step": 2222
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.42281716825508947,
+      "learning_rate": 0.00013347818107112554,
+      "loss": 1.1302,
+      "step": 2223
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.4578000868786269,
+      "learning_rate": 0.00013342193823602164,
+      "loss": 1.1257,
+      "step": 2224
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.4444054424166417,
+      "learning_rate": 0.00013336568349673448,
+      "loss": 1.1038,
+      "step": 2225
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.4198103980672787,
+      "learning_rate": 0.00013330941687330082,
+      "loss": 1.2596,
+      "step": 2226
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.399116733363804,
+      "learning_rate": 0.0001332531383857616,
+      "loss": 1.4114,
+      "step": 2227
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.4205710891157664,
+      "learning_rate": 0.00013319684805416208,
+      "loss": 1.0222,
+      "step": 2228
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.45996484472994825,
+      "learning_rate": 0.00013314054589855164,
+      "loss": 1.2088,
+      "step": 2229
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.4146333134792787,
+      "learning_rate": 0.00013308423193898392,
+      "loss": 1.2586,
+      "step": 2230
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.42153357859216317,
+      "learning_rate": 0.00013302790619551674,
+      "loss": 1.1426,
+      "step": 2231
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.47042745913614814,
+      "learning_rate": 0.00013297156868821212,
+      "loss": 1.2129,
+      "step": 2232
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.4321864960520822,
+      "learning_rate": 0.00013291521943713632,
+      "loss": 1.2639,
+      "step": 2233
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.4137947362519655,
+      "learning_rate": 0.00013285885846235973,
+      "loss": 1.1449,
+      "step": 2234
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.4661887723705618,
+      "learning_rate": 0.0001328024857839569,
+      "loss": 1.1059,
+      "step": 2235
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.40000799474328996,
+      "learning_rate": 0.00013274610142200664,
+      "loss": 1.0224,
+      "step": 2236
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.43684530371371105,
+      "learning_rate": 0.00013268970539659182,
+      "loss": 1.372,
+      "step": 2237
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.42963450100022454,
+      "learning_rate": 0.00013263329772779953,
+      "loss": 1.2254,
+      "step": 2238
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.4943509966128191,
+      "learning_rate": 0.00013257687843572092,
+      "loss": 1.1192,
+      "step": 2239
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.45156034011781215,
+      "learning_rate": 0.00013252044754045138,
+      "loss": 1.1666,
+      "step": 2240
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.46353969448217114,
+      "learning_rate": 0.00013246400506209045,
+      "loss": 1.0629,
+      "step": 2241
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.4652797933112969,
+      "learning_rate": 0.00013240755102074162,
+      "loss": 1.095,
+      "step": 2242
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.3959121961576266,
+      "learning_rate": 0.00013235108543651272,
+      "loss": 1.1713,
+      "step": 2243
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.40909966774694156,
+      "learning_rate": 0.00013229460832951554,
+      "loss": 1.2641,
+      "step": 2244
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.4299838136214135,
+      "learning_rate": 0.00013223811971986606,
+      "loss": 1.234,
+      "step": 2245
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.40645225935697243,
+      "learning_rate": 0.00013218161962768427,
+      "loss": 1.2076,
+      "step": 2246
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.46685745157274405,
+      "learning_rate": 0.00013212510807309434,
+      "loss": 1.1159,
+      "step": 2247
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5005192313347224,
+      "learning_rate": 0.00013206858507622447,
+      "loss": 1.2256,
+      "step": 2248
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.3951969795453201,
+      "learning_rate": 0.00013201205065720698,
+      "loss": 1.1911,
+      "step": 2249
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.383434982588342,
+      "learning_rate": 0.00013195550483617818,
+      "loss": 1.2029,
+      "step": 2250
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.3842244055721399,
+      "learning_rate": 0.0001318989476332785,
+      "loss": 1.2317,
+      "step": 2251
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.4565166646013192,
+      "learning_rate": 0.00013184237906865243,
+      "loss": 0.9395,
+      "step": 2252
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.42546537067868495,
+      "learning_rate": 0.00013178579916244847,
+      "loss": 1.0081,
+      "step": 2253
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.40555189305787287,
+      "learning_rate": 0.00013172920793481918,
+      "loss": 1.2552,
+      "step": 2254
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.40305269956684453,
+      "learning_rate": 0.00013167260540592114,
+      "loss": 1.3225,
+      "step": 2255
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.4463177251593675,
+      "learning_rate": 0.00013161599159591502,
+      "loss": 1.3793,
+      "step": 2256
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.4142228212045942,
+      "learning_rate": 0.00013155936652496544,
+      "loss": 1.119,
+      "step": 2257
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.43496407389181846,
+      "learning_rate": 0.00013150273021324095,
+      "loss": 1.1783,
+      "step": 2258
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.41729612636481045,
+      "learning_rate": 0.00013144608268091435,
+      "loss": 1.2714,
+      "step": 2259
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.434422688584761,
+      "learning_rate": 0.00013138942394816215,
+      "loss": 1.3038,
+      "step": 2260
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.40818833235561863,
+      "learning_rate": 0.00013133275403516507,
+      "loss": 1.2495,
+      "step": 2261
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.41646676468485355,
+      "learning_rate": 0.00013127607296210773,
+      "loss": 0.9155,
+      "step": 2262
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.44414845689970806,
+      "learning_rate": 0.00013121938074917865,
+      "loss": 1.1769,
+      "step": 2263
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.44487776709900473,
+      "learning_rate": 0.00013116267741657045,
+      "loss": 1.2519,
+      "step": 2264
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.45281254803571,
+      "learning_rate": 0.00013110596298447963,
+      "loss": 1.154,
+      "step": 2265
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.44194416836515366,
+      "learning_rate": 0.00013104923747310666,
+      "loss": 1.0987,
+      "step": 2266
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.4139179293110785,
+      "learning_rate": 0.000130992500902656,
+      "loss": 1.2494,
+      "step": 2267
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.4711748098735488,
+      "learning_rate": 0.0001309357532933359,
+      "loss": 1.1805,
+      "step": 2268
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.42340193375347457,
+      "learning_rate": 0.0001308789946653588,
+      "loss": 1.2706,
+      "step": 2269
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.4865655582305204,
+      "learning_rate": 0.00013082222503894085,
+      "loss": 1.0943,
+      "step": 2270
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.4291507148014026,
+      "learning_rate": 0.00013076544443430213,
+      "loss": 1.1929,
+      "step": 2271
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.43247630787843105,
+      "learning_rate": 0.00013070865287166677,
+      "loss": 1.1084,
+      "step": 2272
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.430106647964287,
+      "learning_rate": 0.00013065185037126263,
+      "loss": 1.0211,
+      "step": 2273
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.4528768631636721,
+      "learning_rate": 0.00013059503695332165,
+      "loss": 1.1558,
+      "step": 2274
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.4351062698801986,
+      "learning_rate": 0.00013053821263807946,
+      "loss": 1.2176,
+      "step": 2275
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.4045636521053758,
+      "learning_rate": 0.00013048137744577572,
+      "loss": 1.2259,
+      "step": 2276
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.4931012304736558,
+      "learning_rate": 0.00013042453139665397,
+      "loss": 1.2631,
+      "step": 2277
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.4384155302479605,
+      "learning_rate": 0.00013036767451096148,
+      "loss": 1.2881,
+      "step": 2278
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.41112731104146666,
+      "learning_rate": 0.0001303108068089495,
+      "loss": 1.1884,
+      "step": 2279
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.41690482784853206,
+      "learning_rate": 0.00013025392831087307,
+      "loss": 1.1138,
+      "step": 2280
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.4091170281066683,
+      "learning_rate": 0.00013019703903699114,
+      "loss": 1.1813,
+      "step": 2281
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.4648363070281301,
+      "learning_rate": 0.00013014013900756644,
+      "loss": 1.1517,
+      "step": 2282
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.4740477563102422,
+      "learning_rate": 0.00013008322824286555,
+      "loss": 1.1122,
+      "step": 2283
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.39633984986017107,
+      "learning_rate": 0.00013002630676315884,
+      "loss": 1.2972,
+      "step": 2284
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.4197783786303029,
+      "learning_rate": 0.00012996937458872057,
+      "loss": 1.1585,
+      "step": 2285
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.48771681000826056,
+      "learning_rate": 0.00012991243173982877,
+      "loss": 1.0958,
+      "step": 2286
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.39814248856005746,
+      "learning_rate": 0.00012985547823676525,
+      "loss": 1.1541,
+      "step": 2287
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.4689802448084067,
+      "learning_rate": 0.00012979851409981565,
+      "loss": 1.0182,
+      "step": 2288
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.4303089911667587,
+      "learning_rate": 0.00012974153934926938,
+      "loss": 1.1465,
+      "step": 2289
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.43133924079886476,
+      "learning_rate": 0.00012968455400541966,
+      "loss": 1.1107,
+      "step": 2290
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.42599742113918615,
+      "learning_rate": 0.00012962755808856342,
+      "loss": 1.1836,
+      "step": 2291
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.44108064154216436,
+      "learning_rate": 0.00012957055161900137,
+      "loss": 1.1912,
+      "step": 2292
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.42796090015054794,
+      "learning_rate": 0.00012951353461703808,
+      "loss": 1.2361,
+      "step": 2293
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.4470698617949849,
+      "learning_rate": 0.00012945650710298175,
+      "loss": 1.2515,
+      "step": 2294
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.3990890433751725,
+      "learning_rate": 0.00012939946909714433,
+      "loss": 1.2719,
+      "step": 2295
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.46089746227992545,
+      "learning_rate": 0.0001293424206198416,
+      "loss": 1.0696,
+      "step": 2296
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.4100773278805265,
+      "learning_rate": 0.000129285361691393,
+      "loss": 1.2692,
+      "step": 2297
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.42311773762078525,
+      "learning_rate": 0.00012922829233212165,
+      "loss": 1.2931,
+      "step": 2298
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.4576393225548987,
+      "learning_rate": 0.00012917121256235455,
+      "loss": 1.1847,
+      "step": 2299
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.46253415998536745,
+      "learning_rate": 0.0001291141224024222,
+      "loss": 1.2542,
+      "step": 2300
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.4552726503931759,
+      "learning_rate": 0.00012905702187265895,
+      "loss": 0.9907,
+      "step": 2301
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.48172962405891184,
+      "learning_rate": 0.00012899991099340284,
+      "loss": 1.1968,
+      "step": 2302
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.3812946911273757,
+      "learning_rate": 0.00012894278978499543,
+      "loss": 1.2274,
+      "step": 2303
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.41499425576353594,
+      "learning_rate": 0.00012888565826778216,
+      "loss": 1.2373,
+      "step": 2304
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.40249885907564603,
+      "learning_rate": 0.00012882851646211206,
+      "loss": 1.2895,
+      "step": 2305
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.3975429310557063,
+      "learning_rate": 0.0001287713643883378,
+      "loss": 1.1388,
+      "step": 2306
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.4154869349101822,
+      "learning_rate": 0.00012871420206681571,
+      "loss": 1.2251,
+      "step": 2307
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.4270975609694272,
+      "learning_rate": 0.00012865702951790587,
+      "loss": 0.9923,
+      "step": 2308
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.40844125212422083,
+      "learning_rate": 0.00012859984676197187,
+      "loss": 1.2024,
+      "step": 2309
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.42080160293553265,
+      "learning_rate": 0.000128542653819381,
+      "loss": 1.2319,
+      "step": 2310
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.40441268581434553,
+      "learning_rate": 0.00012848545071050414,
+      "loss": 1.1179,
+      "step": 2311
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.48048767258198793,
+      "learning_rate": 0.00012842823745571588,
+      "loss": 1.0093,
+      "step": 2312
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.4883893903812722,
+      "learning_rate": 0.00012837101407539434,
+      "loss": 1.1385,
+      "step": 2313
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.46467626671074613,
+      "learning_rate": 0.00012831378058992128,
+      "loss": 0.9319,
+      "step": 2314
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.4831149726883403,
+      "learning_rate": 0.000128256537019682,
+      "loss": 1.1914,
+      "step": 2315
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.40815838055736375,
+      "learning_rate": 0.00012819928338506548,
+      "loss": 1.2447,
+      "step": 2316
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.43543915130423294,
+      "learning_rate": 0.00012814201970646425,
+      "loss": 1.1409,
+      "step": 2317
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.461088899381986,
+      "learning_rate": 0.0001280847460042744,
+      "loss": 1.2025,
+      "step": 2318
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.38453052337533755,
+      "learning_rate": 0.00012802746229889563,
+      "loss": 1.0726,
+      "step": 2319
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.395962650443037,
+      "learning_rate": 0.00012797016861073113,
+      "loss": 1.3503,
+      "step": 2320
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.4222003354777775,
+      "learning_rate": 0.00012791286496018772,
+      "loss": 1.1398,
+      "step": 2321
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.462433837857333,
+      "learning_rate": 0.00012785555136767576,
+      "loss": 1.1148,
+      "step": 2322
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.415910623264656,
+      "learning_rate": 0.00012779822785360912,
+      "loss": 1.3319,
+      "step": 2323
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.42264240413437426,
+      "learning_rate": 0.0001277408944384052,
+      "loss": 1.1844,
+      "step": 2324
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.392575466777858,
+      "learning_rate": 0.00012768355114248494,
+      "loss": 1.1632,
+      "step": 2325
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.4041114746621208,
+      "learning_rate": 0.00012762619798627282,
+      "loss": 1.1651,
+      "step": 2326
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.4393333116831791,
+      "learning_rate": 0.00012756883499019683,
+      "loss": 1.2162,
+      "step": 2327
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.4153889408364148,
+      "learning_rate": 0.0001275114621746884,
+      "loss": 1.274,
+      "step": 2328
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5046389668203737,
+      "learning_rate": 0.00012745407956018254,
+      "loss": 1.2429,
+      "step": 2329
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.4063442019747821,
+      "learning_rate": 0.00012739668716711772,
+      "loss": 1.1204,
+      "step": 2330
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.4033656869390118,
+      "learning_rate": 0.00012733928501593587,
+      "loss": 1.2087,
+      "step": 2331
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.4140069965870761,
+      "learning_rate": 0.00012728187312708238,
+      "loss": 1.1255,
+      "step": 2332
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.48876925103032065,
+      "learning_rate": 0.00012722445152100624,
+      "loss": 1.0867,
+      "step": 2333
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.43092999108625796,
+      "learning_rate": 0.00012716702021815977,
+      "loss": 1.1761,
+      "step": 2334
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.4379486619316274,
+      "learning_rate": 0.0001271095792389987,
+      "loss": 1.069,
+      "step": 2335
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.4438426484817456,
+      "learning_rate": 0.0001270521286039824,
+      "loss": 1.1989,
+      "step": 2336
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.4365801414094866,
+      "learning_rate": 0.00012699466833357342,
+      "loss": 1.2436,
+      "step": 2337
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.46498248279585114,
+      "learning_rate": 0.00012693719844823805,
+      "loss": 1.1201,
+      "step": 2338
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.41117603513049755,
+      "learning_rate": 0.00012687971896844575,
+      "loss": 1.178,
+      "step": 2339
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.40225382102935525,
+      "learning_rate": 0.00012682222991466948,
+      "loss": 1.1873,
+      "step": 2340
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.4070213473442209,
+      "learning_rate": 0.00012676473130738566,
+      "loss": 1.4923,
+      "step": 2341
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.416403743005072,
+      "learning_rate": 0.00012670722316707403,
+      "loss": 1.178,
+      "step": 2342
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.46041431357426116,
+      "learning_rate": 0.00012664970551421777,
+      "loss": 1.0268,
+      "step": 2343
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.42139202273671217,
+      "learning_rate": 0.0001265921783693035,
+      "loss": 1.1345,
+      "step": 2344
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.45545745853885533,
+      "learning_rate": 0.00012653464175282108,
+      "loss": 1.1446,
+      "step": 2345
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.4386126571263804,
+      "learning_rate": 0.0001264770956852639,
+      "loss": 1.239,
+      "step": 2346
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.40293794854029197,
+      "learning_rate": 0.00012641954018712863,
+      "loss": 1.1398,
+      "step": 2347
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.39995167209665317,
+      "learning_rate": 0.00012636197527891526,
+      "loss": 1.2102,
+      "step": 2348
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.40077312370755924,
+      "learning_rate": 0.00012630440098112725,
+      "loss": 1.1475,
+      "step": 2349
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.42904390930909264,
+      "learning_rate": 0.0001262468173142713,
+      "loss": 1.1951,
+      "step": 2350
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.39349412391012534,
+      "learning_rate": 0.00012618922429885753,
+      "loss": 1.232,
+      "step": 2351
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.43539444015735224,
+      "learning_rate": 0.0001261316219553993,
+      "loss": 1.1262,
+      "step": 2352
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5186144381748614,
+      "learning_rate": 0.00012607401030441336,
+      "loss": 1.1747,
+      "step": 2353
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.38136521760966724,
+      "learning_rate": 0.00012601638936641976,
+      "loss": 1.3548,
+      "step": 2354
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.4187861302738498,
+      "learning_rate": 0.00012595875916194188,
+      "loss": 1.2279,
+      "step": 2355
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.4173623728549111,
+      "learning_rate": 0.00012590111971150626,
+      "loss": 1.0507,
+      "step": 2356
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.428595271250096,
+      "learning_rate": 0.00012584347103564297,
+      "loss": 1.0981,
+      "step": 2357
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.4018483231221196,
+      "learning_rate": 0.00012578581315488514,
+      "loss": 1.2823,
+      "step": 2358
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5043659948107583,
+      "learning_rate": 0.0001257281460897694,
+      "loss": 1.1749,
+      "step": 2359
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.4271218180542123,
+      "learning_rate": 0.00012567046986083544,
+      "loss": 1.2285,
+      "step": 2360
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.4297985715745637,
+      "learning_rate": 0.00012561278448862634,
+      "loss": 1.1336,
+      "step": 2361
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.4259014970835036,
+      "learning_rate": 0.00012555508999368838,
+      "loss": 1.1284,
+      "step": 2362
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.42654040207191724,
+      "learning_rate": 0.00012549738639657115,
+      "loss": 1.1376,
+      "step": 2363
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.3969477522584373,
+      "learning_rate": 0.00012543967371782738,
+      "loss": 1.2839,
+      "step": 2364
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.44223705764356636,
+      "learning_rate": 0.00012538195197801317,
+      "loss": 1.337,
+      "step": 2365
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.4517277824241131,
+      "learning_rate": 0.00012532422119768775,
+      "loss": 1.2127,
+      "step": 2366
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.45556831736960424,
+      "learning_rate": 0.0001252664813974136,
+      "loss": 1.1621,
+      "step": 2367
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.4002106699831191,
+      "learning_rate": 0.00012520873259775636,
+      "loss": 1.2579,
+      "step": 2368
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.3861869399038417,
+      "learning_rate": 0.00012515097481928495,
+      "loss": 1.2099,
+      "step": 2369
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.37675097436930455,
+      "learning_rate": 0.0001250932080825715,
+      "loss": 1.242,
+      "step": 2370
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.4047724744875969,
+      "learning_rate": 0.00012503543240819127,
+      "loss": 1.1571,
+      "step": 2371
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.4714940734932751,
+      "learning_rate": 0.0001249776478167227,
+      "loss": 1.1383,
+      "step": 2372
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.39635026722850725,
+      "learning_rate": 0.00012491985432874745,
+      "loss": 1.21,
+      "step": 2373
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.4414964601813941,
+      "learning_rate": 0.00012486205196485034,
+      "loss": 1.3132,
+      "step": 2374
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.4695350279052181,
+      "learning_rate": 0.00012480424074561933,
+      "loss": 0.9809,
+      "step": 2375
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.4667724451728448,
+      "learning_rate": 0.0001247464206916456,
+      "loss": 1.1343,
+      "step": 2376
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.41049760955221054,
+      "learning_rate": 0.00012468859182352332,
+      "loss": 1.3624,
+      "step": 2377
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.45789583566063413,
+      "learning_rate": 0.00012463075416184994,
+      "loss": 1.2117,
+      "step": 2378
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.4507560825505181,
+      "learning_rate": 0.00012457290772722608,
+      "loss": 1.3348,
+      "step": 2379
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.42960393321210344,
+      "learning_rate": 0.0001245150525402553,
+      "loss": 1.1296,
+      "step": 2380
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.3975348010520652,
+      "learning_rate": 0.00012445718862154445,
+      "loss": 1.1837,
+      "step": 2381
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.462980924273746,
+      "learning_rate": 0.00012439931599170342,
+      "loss": 1.1087,
+      "step": 2382
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.40742877569931185,
+      "learning_rate": 0.00012434143467134523,
+      "loss": 1.0699,
+      "step": 2383
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.4141762306382595,
+      "learning_rate": 0.0001242835446810859,
+      "loss": 1.168,
+      "step": 2384
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.41204672364560374,
+      "learning_rate": 0.00012422564604154467,
+      "loss": 1.2634,
+      "step": 2385
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.43099417651041766,
+      "learning_rate": 0.0001241677387733438,
+      "loss": 1.2037,
+      "step": 2386
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.4254128195103118,
+      "learning_rate": 0.00012410982289710865,
+      "loss": 1.1714,
+      "step": 2387
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.4481516108268839,
+      "learning_rate": 0.00012405189843346756,
+      "loss": 1.1247,
+      "step": 2388
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.42951123296368976,
+      "learning_rate": 0.00012399396540305205,
+      "loss": 1.216,
+      "step": 2389
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.3738215547345975,
+      "learning_rate": 0.00012393602382649655,
+      "loss": 1.146,
+      "step": 2390
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.39263204843048244,
+      "learning_rate": 0.00012387807372443873,
+      "loss": 1.2335,
+      "step": 2391
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.37819155971474827,
+      "learning_rate": 0.00012382011511751913,
+      "loss": 1.1908,
+      "step": 2392
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.43370274779315854,
+      "learning_rate": 0.0001237621480263813,
+      "loss": 1.0863,
+      "step": 2393
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.385486719704166,
+      "learning_rate": 0.00012370417247167202,
+      "loss": 1.1801,
+      "step": 2394
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.40346064598723186,
+      "learning_rate": 0.0001236461884740409,
+      "loss": 1.2326,
+      "step": 2395
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.4417016162332854,
+      "learning_rate": 0.0001235881960541405,
+      "loss": 1.135,
+      "step": 2396
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.44659077484348825,
+      "learning_rate": 0.00012353019523262665,
+      "loss": 1.2354,
+      "step": 2397
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.46008561792918923,
+      "learning_rate": 0.00012347218603015793,
+      "loss": 0.9907,
+      "step": 2398
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.44621326444908765,
+      "learning_rate": 0.00012341416846739595,
+      "loss": 1.3026,
+      "step": 2399
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.40544293670776027,
+      "learning_rate": 0.0001233561425650054,
+      "loss": 1.0586,
+      "step": 2400
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.44094103568960874,
+      "learning_rate": 0.00012329810834365383,
+      "loss": 1.1356,
+      "step": 2401
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.4143933226824256,
+      "learning_rate": 0.0001232400658240118,
+      "loss": 1.102,
+      "step": 2402
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.4132902296209223,
+      "learning_rate": 0.00012318201502675285,
+      "loss": 1.2507,
+      "step": 2403
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5238026401757537,
+      "learning_rate": 0.0001231239559725534,
+      "loss": 1.1226,
+      "step": 2404
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.4301900918624624,
+      "learning_rate": 0.00012306588868209284,
+      "loss": 1.2483,
+      "step": 2405
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.46906438959963004,
+      "learning_rate": 0.00012300781317605355,
+      "loss": 1.1253,
+      "step": 2406
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.445577127481049,
+      "learning_rate": 0.0001229497294751208,
+      "loss": 1.1767,
+      "step": 2407
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.40798508079845813,
+      "learning_rate": 0.00012289163759998269,
+      "loss": 1.0818,
+      "step": 2408
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.4301566799228255,
+      "learning_rate": 0.00012283353757133037,
+      "loss": 1.2676,
+      "step": 2409
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.4966065581632544,
+      "learning_rate": 0.00012277542940985779,
+      "loss": 1.1876,
+      "step": 2410
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.4488992215529306,
+      "learning_rate": 0.0001227173131362619,
+      "loss": 1.1473,
+      "step": 2411
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.3954036014220161,
+      "learning_rate": 0.00012265918877124246,
+      "loss": 1.2067,
+      "step": 2412
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.4114195114307291,
+      "learning_rate": 0.0001226010563355021,
+      "loss": 1.2652,
+      "step": 2413
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.3863437852725348,
+      "learning_rate": 0.00012254291584974636,
+      "loss": 1.313,
+      "step": 2414
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.4376856057136768,
+      "learning_rate": 0.00012248476733468368,
+      "loss": 1.1477,
+      "step": 2415
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.423957687415224,
+      "learning_rate": 0.00012242661081102534,
+      "loss": 1.1297,
+      "step": 2416
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.4196151504412718,
+      "learning_rate": 0.00012236844629948538,
+      "loss": 1.2301,
+      "step": 2417
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.4209255681654184,
+      "learning_rate": 0.00012231027382078081,
+      "loss": 1.1512,
+      "step": 2418
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.49597831434483164,
+      "learning_rate": 0.00012225209339563145,
+      "loss": 1.1427,
+      "step": 2419
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.3969132798415712,
+      "learning_rate": 0.00012219390504475988,
+      "loss": 1.0197,
+      "step": 2420
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.4236732141835515,
+      "learning_rate": 0.0001221357087888916,
+      "loss": 1.1154,
+      "step": 2421
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.42476371925417616,
+      "learning_rate": 0.00012207750464875484,
+      "loss": 1.1408,
+      "step": 2422
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.38356341433455526,
+      "learning_rate": 0.00012201929264508065,
+      "loss": 1.1796,
+      "step": 2423
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.4852567101072414,
+      "learning_rate": 0.00012196107279860301,
+      "loss": 1.0783,
+      "step": 2424
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.43874227664967486,
+      "learning_rate": 0.00012190284513005849,
+      "loss": 1.1268,
+      "step": 2425
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.4737868923855923,
+      "learning_rate": 0.00012184460966018659,
+      "loss": 1.2342,
+      "step": 2426
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.49420396198370586,
+      "learning_rate": 0.00012178636640972953,
+      "loss": 1.2192,
+      "step": 2427
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.4563883703028032,
+      "learning_rate": 0.00012172811539943234,
+      "loss": 1.1518,
+      "step": 2428
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.4535307542510057,
+      "learning_rate": 0.00012166985665004278,
+      "loss": 1.0657,
+      "step": 2429
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.39362549986112333,
+      "learning_rate": 0.00012161159018231136,
+      "loss": 1.2082,
+      "step": 2430
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.36836715148192606,
+      "learning_rate": 0.00012155331601699136,
+      "loss": 1.0084,
+      "step": 2431
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.3803251852158082,
+      "learning_rate": 0.00012149503417483885,
+      "loss": 1.3516,
+      "step": 2432
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.4846813704873836,
+      "learning_rate": 0.00012143674467661248,
+      "loss": 1.2233,
+      "step": 2433
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.417112110404014,
+      "learning_rate": 0.00012137844754307381,
+      "loss": 0.9569,
+      "step": 2434
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.4429846759175786,
+      "learning_rate": 0.00012132014279498703,
+      "loss": 1.0672,
+      "step": 2435
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.3903979236496923,
+      "learning_rate": 0.00012126183045311904,
+      "loss": 1.2872,
+      "step": 2436
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.43323045990009684,
+      "learning_rate": 0.00012120351053823942,
+      "loss": 1.1264,
+      "step": 2437
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.4955960099939004,
+      "learning_rate": 0.00012114518307112053,
+      "loss": 1.2207,
+      "step": 2438
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.46028536722912544,
+      "learning_rate": 0.00012108684807253734,
+      "loss": 1.1115,
+      "step": 2439
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5070196704631956,
+      "learning_rate": 0.00012102850556326761,
+      "loss": 1.2365,
+      "step": 2440
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.4545866378560307,
+      "learning_rate": 0.00012097015556409161,
+      "loss": 1.306,
+      "step": 2441
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.4194816902485913,
+      "learning_rate": 0.00012091179809579238,
+      "loss": 1.1201,
+      "step": 2442
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.4546396561626711,
+      "learning_rate": 0.00012085343317915565,
+      "loss": 1.125,
+      "step": 2443
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.4255938898678441,
+      "learning_rate": 0.00012079506083496974,
+      "loss": 1.2134,
+      "step": 2444
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.4373711809629479,
+      "learning_rate": 0.00012073668108402565,
+      "loss": 1.0882,
+      "step": 2445
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.38119705119450503,
+      "learning_rate": 0.00012067829394711696,
+      "loss": 1.2909,
+      "step": 2446
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.46029833326942904,
+      "learning_rate": 0.00012061989944503998,
+      "loss": 1.0661,
+      "step": 2447
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.4398760478158,
+      "learning_rate": 0.00012056149759859357,
+      "loss": 1.2126,
+      "step": 2448
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.3879305597885016,
+      "learning_rate": 0.0001205030884285792,
+      "loss": 1.1494,
+      "step": 2449
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.4123772544955544,
+      "learning_rate": 0.00012044467195580102,
+      "loss": 1.1365,
+      "step": 2450
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.42144110281641584,
+      "learning_rate": 0.00012038624820106572,
+      "loss": 1.2231,
+      "step": 2451
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.4286929675274135,
+      "learning_rate": 0.00012032781718518257,
+      "loss": 1.1835,
+      "step": 2452
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.41932241507006335,
+      "learning_rate": 0.00012026937892896351,
+      "loss": 1.0736,
+      "step": 2453
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.420614129992187,
+      "learning_rate": 0.00012021093345322292,
+      "loss": 1.3127,
+      "step": 2454
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.42996318036330944,
+      "learning_rate": 0.0001201524807787779,
+      "loss": 1.119,
+      "step": 2455
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.45640038288030976,
+      "learning_rate": 0.00012009402092644804,
+      "loss": 1.2982,
+      "step": 2456
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.4792506130980692,
+      "learning_rate": 0.00012003555391705545,
+      "loss": 1.1413,
+      "step": 2457
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.46498189676082136,
+      "learning_rate": 0.00011997707977142492,
+      "loss": 1.1492,
+      "step": 2458
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.42683703484086594,
+      "learning_rate": 0.0001199185985103836,
+      "loss": 1.1554,
+      "step": 2459
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.4105955223683733,
+      "learning_rate": 0.00011986011015476138,
+      "loss": 1.1179,
+      "step": 2460
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.4050332005776959,
+      "learning_rate": 0.00011980161472539046,
+      "loss": 1.1334,
+      "step": 2461
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.38635958116910624,
+      "learning_rate": 0.00011974311224310573,
+      "loss": 1.0987,
+      "step": 2462
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.49999266834669,
+      "learning_rate": 0.0001196846027287445,
+      "loss": 1.1577,
+      "step": 2463
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.42587147499819633,
+      "learning_rate": 0.00011962608620314664,
+      "loss": 1.2013,
+      "step": 2464
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.4138637612833237,
+      "learning_rate": 0.00011956756268715448,
+      "loss": 1.188,
+      "step": 2465
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.44578023546173373,
+      "learning_rate": 0.00011950903220161285,
+      "loss": 1.0645,
+      "step": 2466
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.39625818788176465,
+      "learning_rate": 0.00011945049476736905,
+      "loss": 1.105,
+      "step": 2467
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.4366901955079131,
+      "learning_rate": 0.0001193919504052729,
+      "loss": 1.1473,
+      "step": 2468
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.39748833116292653,
+      "learning_rate": 0.0001193333991361766,
+      "loss": 1.1107,
+      "step": 2469
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.46380819064008594,
+      "learning_rate": 0.00011927484098093493,
+      "loss": 1.24,
+      "step": 2470
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.4605438921369201,
+      "learning_rate": 0.00011921627596040503,
+      "loss": 1.0682,
+      "step": 2471
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.42970258923041793,
+      "learning_rate": 0.00011915770409544654,
+      "loss": 1.2191,
+      "step": 2472
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.4318329538570119,
+      "learning_rate": 0.00011909912540692148,
+      "loss": 1.1499,
+      "step": 2473
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.4426845474889169,
+      "learning_rate": 0.00011904053991569433,
+      "loss": 1.1502,
+      "step": 2474
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.38413712438808795,
+      "learning_rate": 0.00011898194764263197,
+      "loss": 1.0402,
+      "step": 2475
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.4186118268328292,
+      "learning_rate": 0.00011892334860860382,
+      "loss": 1.1291,
+      "step": 2476
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.4564114023622362,
+      "learning_rate": 0.00011886474283448153,
+      "loss": 1.2008,
+      "step": 2477
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.3814674766694374,
+      "learning_rate": 0.00011880613034113921,
+      "loss": 1.2126,
+      "step": 2478
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.39491698336213243,
+      "learning_rate": 0.00011874751114945346,
+      "loss": 1.1585,
+      "step": 2479
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.4411487082462044,
+      "learning_rate": 0.00011868888528030312,
+      "loss": 1.1623,
+      "step": 2480
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.45187403058070946,
+      "learning_rate": 0.00011863025275456953,
+      "loss": 1.145,
+      "step": 2481
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.43152173344988926,
+      "learning_rate": 0.00011857161359313633,
+      "loss": 1.1409,
+      "step": 2482
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.47329783803269126,
+      "learning_rate": 0.00011851296781688952,
+      "loss": 1.1398,
+      "step": 2483
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5382952029990064,
+      "learning_rate": 0.00011845431544671748,
+      "loss": 0.9918,
+      "step": 2484
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.49852131725271354,
+      "learning_rate": 0.00011839565650351097,
+      "loss": 0.9512,
+      "step": 2485
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.4980127370473026,
+      "learning_rate": 0.00011833699100816302,
+      "loss": 1.2417,
+      "step": 2486
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.3824378838705919,
+      "learning_rate": 0.00011827831898156905,
+      "loss": 1.1329,
+      "step": 2487
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.44770861767996784,
+      "learning_rate": 0.00011821964044462679,
+      "loss": 1.1814,
+      "step": 2488
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.47122963687066816,
+      "learning_rate": 0.00011816095541823629,
+      "loss": 1.1034,
+      "step": 2489
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.49325150107822235,
+      "learning_rate": 0.00011810226392329988,
+      "loss": 1.1817,
+      "step": 2490
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.40565128103679327,
+      "learning_rate": 0.00011804356598072223,
+      "loss": 1.2067,
+      "step": 2491
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5286003026817951,
+      "learning_rate": 0.00011798486161141034,
+      "loss": 0.8964,
+      "step": 2492
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.4374114598907731,
+      "learning_rate": 0.00011792615083627345,
+      "loss": 1.0722,
+      "step": 2493
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.4730973460284043,
+      "learning_rate": 0.00011786743367622301,
+      "loss": 1.1543,
+      "step": 2494
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.44051380448311633,
+      "learning_rate": 0.0001178087101521729,
+      "loss": 1.181,
+      "step": 2495
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.458407230465703,
+      "learning_rate": 0.00011774998028503916,
+      "loss": 1.1167,
+      "step": 2496
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.4751940215624145,
+      "learning_rate": 0.00011769124409574014,
+      "loss": 1.1277,
+      "step": 2497
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.43873280889113414,
+      "learning_rate": 0.00011763250160519639,
+      "loss": 1.0174,
+      "step": 2498
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.42393699361946696,
+      "learning_rate": 0.00011757375283433076,
+      "loss": 1.123,
+      "step": 2499
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.4777626803182419,
+      "learning_rate": 0.00011751499780406829,
+      "loss": 1.0314,
+      "step": 2500
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.4416887389999516,
+      "learning_rate": 0.00011745623653533633,
+      "loss": 1.1848,
+      "step": 2501
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.43009544258011156,
+      "learning_rate": 0.0001173974690490643,
+      "loss": 1.0603,
+      "step": 2502
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5230697112944794,
+      "learning_rate": 0.000117338695366184,
+      "loss": 1.1473,
+      "step": 2503
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.43073281314258915,
+      "learning_rate": 0.00011727991550762936,
+      "loss": 0.9644,
+      "step": 2504
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.39746084068099385,
+      "learning_rate": 0.0001172211294943365,
+      "loss": 1.2164,
+      "step": 2505
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.4455777011694188,
+      "learning_rate": 0.00011716233734724371,
+      "loss": 1.1143,
+      "step": 2506
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.4749947167008822,
+      "learning_rate": 0.00011710353908729156,
+      "loss": 1.3033,
+      "step": 2507
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.41844155867220645,
+      "learning_rate": 0.0001170447347354227,
+      "loss": 1.2661,
+      "step": 2508
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.435223466425302,
+      "learning_rate": 0.00011698592431258205,
+      "loss": 1.2514,
+      "step": 2509
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.4169699581679024,
+      "learning_rate": 0.00011692710783971654,
+      "loss": 1.3185,
+      "step": 2510
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.4106925622943266,
+      "learning_rate": 0.00011686828533777539,
+      "loss": 1.1336,
+      "step": 2511
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.463992603658453,
+      "learning_rate": 0.00011680945682770994,
+      "loss": 1.3057,
+      "step": 2512
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.4241523433742991,
+      "learning_rate": 0.00011675062233047364,
+      "loss": 1.1342,
+      "step": 2513
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5447976473936714,
+      "learning_rate": 0.00011669178186702212,
+      "loss": 1.0111,
+      "step": 2514
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.43170427169979997,
+      "learning_rate": 0.00011663293545831302,
+      "loss": 0.9598,
+      "step": 2515
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.40779754180961664,
+      "learning_rate": 0.00011657408312530624,
+      "loss": 1.2707,
+      "step": 2516
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5409698373295824,
+      "learning_rate": 0.00011651522488896371,
+      "loss": 1.1297,
+      "step": 2517
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.3961863756075187,
+      "learning_rate": 0.00011645636077024949,
+      "loss": 1.3697,
+      "step": 2518
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.4450239792932599,
+      "learning_rate": 0.00011639749079012972,
+      "loss": 1.3369,
+      "step": 2519
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.4498397757018136,
+      "learning_rate": 0.00011633861496957263,
+      "loss": 1.0874,
+      "step": 2520
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.4209241251161517,
+      "learning_rate": 0.00011627973332954855,
+      "loss": 1.0289,
+      "step": 2521
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.4464520644474054,
+      "learning_rate": 0.00011622084589102986,
+      "loss": 1.1943,
+      "step": 2522
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.41819558414341185,
+      "learning_rate": 0.00011616195267499102,
+      "loss": 1.1489,
+      "step": 2523
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.4738759922311017,
+      "learning_rate": 0.00011610305370240851,
+      "loss": 1.0456,
+      "step": 2524
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.49864295823644655,
+      "learning_rate": 0.00011604414899426093,
+      "loss": 1.1081,
+      "step": 2525
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.4297441460332717,
+      "learning_rate": 0.00011598523857152887,
+      "loss": 1.1635,
+      "step": 2526
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.4234665514285073,
+      "learning_rate": 0.00011592632245519495,
+      "loss": 1.2128,
+      "step": 2527
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.40063507431262957,
+      "learning_rate": 0.00011586740066624384,
+      "loss": 1.0245,
+      "step": 2528
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.4833419809676442,
+      "learning_rate": 0.00011580847322566224,
+      "loss": 1.0734,
+      "step": 2529
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5410806864473747,
+      "learning_rate": 0.00011574954015443888,
+      "loss": 1.1249,
+      "step": 2530
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.4770174079022598,
+      "learning_rate": 0.00011569060147356441,
+      "loss": 1.1199,
+      "step": 2531
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.3834345713140124,
+      "learning_rate": 0.00011563165720403157,
+      "loss": 1.0307,
+      "step": 2532
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.4869971164635193,
+      "learning_rate": 0.00011557270736683505,
+      "loss": 1.0397,
+      "step": 2533
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.4260060548175767,
+      "learning_rate": 0.00011551375198297151,
+      "loss": 1.2122,
+      "step": 2534
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5396141466824338,
+      "learning_rate": 0.00011545479107343963,
+      "loss": 1.1926,
+      "step": 2535
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.39815322646100554,
+      "learning_rate": 0.00011539582465924001,
+      "loss": 1.1129,
+      "step": 2536
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.42671462725165366,
+      "learning_rate": 0.00011533685276137522,
+      "loss": 1.1408,
+      "step": 2537
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.42208725721545143,
+      "learning_rate": 0.00011527787540084985,
+      "loss": 1.2828,
+      "step": 2538
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.4191896493204638,
+      "learning_rate": 0.00011521889259867032,
+      "loss": 1.099,
+      "step": 2539
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.4231159106972044,
+      "learning_rate": 0.00011515990437584507,
+      "loss": 1.1653,
+      "step": 2540
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.45276491675260794,
+      "learning_rate": 0.00011510091075338444,
+      "loss": 0.837,
+      "step": 2541
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.501142884370324,
+      "learning_rate": 0.00011504191175230077,
+      "loss": 1.2676,
+      "step": 2542
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.4241236070814755,
+      "learning_rate": 0.00011498290739360815,
+      "loss": 1.2145,
+      "step": 2543
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.40750666037366007,
+      "learning_rate": 0.00011492389769832275,
+      "loss": 0.9854,
+      "step": 2544
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.42751445925783393,
+      "learning_rate": 0.00011486488268746254,
+      "loss": 1.0857,
+      "step": 2545
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.4562648512976861,
+      "learning_rate": 0.0001148058623820474,
+      "loss": 1.1301,
+      "step": 2546
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.447773568411726,
+      "learning_rate": 0.00011474683680309912,
+      "loss": 1.2755,
+      "step": 2547
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.49910760304525176,
+      "learning_rate": 0.00011468780597164135,
+      "loss": 1.1179,
+      "step": 2548
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.4062481711380424,
+      "learning_rate": 0.00011462876990869962,
+      "loss": 0.9316,
+      "step": 2549
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.39343938933602346,
+      "learning_rate": 0.00011456972863530133,
+      "loss": 1.0817,
+      "step": 2550
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.4030498443948385,
+      "learning_rate": 0.0001145106821724757,
+      "loss": 1.1889,
+      "step": 2551
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.3990992760711147,
+      "learning_rate": 0.00011445163054125385,
+      "loss": 1.244,
+      "step": 2552
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.39900404819812796,
+      "learning_rate": 0.00011439257376266868,
+      "loss": 1.2054,
+      "step": 2553
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.4879229710312798,
+      "learning_rate": 0.00011433351185775502,
+      "loss": 1.322,
+      "step": 2554
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.49312066432938356,
+      "learning_rate": 0.0001142744448475494,
+      "loss": 1.2046,
+      "step": 2555
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.450877459001166,
+      "learning_rate": 0.00011421537275309027,
+      "loss": 0.928,
+      "step": 2556
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.4742746889184881,
+      "learning_rate": 0.00011415629559541782,
+      "loss": 1.1723,
+      "step": 2557
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.43482852356203466,
+      "learning_rate": 0.00011409721339557414,
+      "loss": 1.1477,
+      "step": 2558
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.4442835544889002,
+      "learning_rate": 0.00011403812617460296,
+      "loss": 1.1319,
+      "step": 2559
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.4213177730231936,
+      "learning_rate": 0.00011397903395354996,
+      "loss": 1.2628,
+      "step": 2560
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.4798522974469049,
+      "learning_rate": 0.00011391993675346251,
+      "loss": 1.06,
+      "step": 2561
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.4506389873721285,
+      "learning_rate": 0.00011386083459538978,
+      "loss": 1.1637,
+      "step": 2562
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.4508660755619113,
+      "learning_rate": 0.00011380172750038269,
+      "loss": 1.2012,
+      "step": 2563
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.47991787393270036,
+      "learning_rate": 0.00011374261548949395,
+      "loss": 1.1215,
+      "step": 2564
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.4571983558574211,
+      "learning_rate": 0.00011368349858377796,
+      "loss": 1.0768,
+      "step": 2565
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.39801524603072264,
+      "learning_rate": 0.00011362437680429097,
+      "loss": 1.1684,
+      "step": 2566
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.4543121007996487,
+      "learning_rate": 0.00011356525017209085,
+      "loss": 1.1112,
+      "step": 2567
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.4169317919358828,
+      "learning_rate": 0.00011350611870823726,
+      "loss": 1.3288,
+      "step": 2568
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.42126550005623575,
+      "learning_rate": 0.00011344698243379154,
+      "loss": 1.2781,
+      "step": 2569
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.38733823301543774,
+      "learning_rate": 0.00011338784136981687,
+      "loss": 1.3004,
+      "step": 2570
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.4378696719919442,
+      "learning_rate": 0.0001133286955373779,
+      "loss": 1.0738,
+      "step": 2571
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.41655139187820633,
+      "learning_rate": 0.00011326954495754121,
+      "loss": 1.2585,
+      "step": 2572
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.4124617112902373,
+      "learning_rate": 0.00011321038965137496,
+      "loss": 1.3417,
+      "step": 2573
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.4306703947822757,
+      "learning_rate": 0.00011315122963994903,
+      "loss": 1.1388,
+      "step": 2574
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.4749153570032497,
+      "learning_rate": 0.00011309206494433488,
+      "loss": 1.2793,
+      "step": 2575
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.4362912377985656,
+      "learning_rate": 0.0001130328955856058,
+      "loss": 1.1446,
+      "step": 2576
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5709439539954194,
+      "learning_rate": 0.0001129737215848366,
+      "loss": 1.2298,
+      "step": 2577
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.41252707159642205,
+      "learning_rate": 0.00011291454296310389,
+      "loss": 1.2364,
+      "step": 2578
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.39041422103314694,
+      "learning_rate": 0.00011285535974148576,
+      "loss": 1.1974,
+      "step": 2579
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.4322106028535349,
+      "learning_rate": 0.00011279617194106198,
+      "loss": 1.1613,
+      "step": 2580
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.4653774783537383,
+      "learning_rate": 0.00011273697958291405,
+      "loss": 1.0733,
+      "step": 2581
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.40833272590457875,
+      "learning_rate": 0.00011267778268812504,
+      "loss": 1.3063,
+      "step": 2582
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.4492221245480568,
+      "learning_rate": 0.00011261858127777956,
+      "loss": 1.124,
+      "step": 2583
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.4459366332263464,
+      "learning_rate": 0.00011255937537296393,
+      "loss": 1.0721,
+      "step": 2584
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.4501209725322147,
+      "learning_rate": 0.00011250016499476602,
+      "loss": 1.0599,
+      "step": 2585
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.42840609322226647,
+      "learning_rate": 0.00011244095016427534,
+      "loss": 1.2195,
+      "step": 2586
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.39173481810504107,
+      "learning_rate": 0.00011238173090258293,
+      "loss": 1.1841,
+      "step": 2587
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.4109681995644072,
+      "learning_rate": 0.00011232250723078141,
+      "loss": 1.3021,
+      "step": 2588
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.4000915370862346,
+      "learning_rate": 0.00011226327916996504,
+      "loss": 1.2244,
+      "step": 2589
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5161524811623353,
+      "learning_rate": 0.00011220404674122955,
+      "loss": 1.0963,
+      "step": 2590
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.4168838123478484,
+      "learning_rate": 0.00011214480996567232,
+      "loss": 1.1153,
+      "step": 2591
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.458867457926017,
+      "learning_rate": 0.00011208556886439216,
+      "loss": 1.0559,
+      "step": 2592
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.48477413567899835,
+      "learning_rate": 0.00011202632345848954,
+      "loss": 1.1572,
+      "step": 2593
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.3826180125373306,
+      "learning_rate": 0.00011196707376906641,
+      "loss": 1.2015,
+      "step": 2594
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.41595732424507137,
+      "learning_rate": 0.00011190781981722623,
+      "loss": 1.0798,
+      "step": 2595
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.41940540389604797,
+      "learning_rate": 0.00011184856162407398,
+      "loss": 0.9946,
+      "step": 2596
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.4464129046678245,
+      "learning_rate": 0.00011178929921071622,
+      "loss": 1.1987,
+      "step": 2597
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.4959532596516926,
+      "learning_rate": 0.00011173003259826094,
+      "loss": 1.1694,
+      "step": 2598
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.4319004378552327,
+      "learning_rate": 0.00011167076180781764,
+      "loss": 1.1035,
+      "step": 2599
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.43382220840008145,
+      "learning_rate": 0.0001116114868604973,
+      "loss": 0.9915,
+      "step": 2600
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.4648502824050836,
+      "learning_rate": 0.0001115522077774124,
+      "loss": 1.1123,
+      "step": 2601
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.3971841860702026,
+      "learning_rate": 0.00011149292457967691,
+      "loss": 1.1946,
+      "step": 2602
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.48000927867731547,
+      "learning_rate": 0.00011143363728840625,
+      "loss": 1.2215,
+      "step": 2603
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5038654425465321,
+      "learning_rate": 0.00011137434592471725,
+      "loss": 1.1019,
+      "step": 2604
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.4132375560077463,
+      "learning_rate": 0.00011131505050972826,
+      "loss": 1.1529,
+      "step": 2605
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.4184444823622112,
+      "learning_rate": 0.00011125575106455906,
+      "loss": 1.1872,
+      "step": 2606
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5632339226482768,
+      "learning_rate": 0.00011119644761033078,
+      "loss": 1.0018,
+      "step": 2607
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.47441806472208325,
+      "learning_rate": 0.00011113714016816614,
+      "loss": 1.1996,
+      "step": 2608
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.43235590656439454,
+      "learning_rate": 0.00011107782875918913,
+      "loss": 1.0887,
+      "step": 2609
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.47699431557111804,
+      "learning_rate": 0.00011101851340452527,
+      "loss": 1.1557,
+      "step": 2610
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.425196865273347,
+      "learning_rate": 0.00011095919412530136,
+      "loss": 1.2179,
+      "step": 2611
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.4303558977155775,
+      "learning_rate": 0.00011089987094264569,
+      "loss": 1.2146,
+      "step": 2612
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.50629642834796,
+      "learning_rate": 0.0001108405438776879,
+      "loss": 1.0304,
+      "step": 2613
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.40214713101139804,
+      "learning_rate": 0.00011078121295155905,
+      "loss": 1.2,
+      "step": 2614
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.4480201861597104,
+      "learning_rate": 0.00011072187818539159,
+      "loss": 1.1543,
+      "step": 2615
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.45319249384069454,
+      "learning_rate": 0.00011066253960031923,
+      "loss": 1.0912,
+      "step": 2616
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.39034406705052244,
+      "learning_rate": 0.00011060319721747714,
+      "loss": 1.089,
+      "step": 2617
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.4830192190839973,
+      "learning_rate": 0.0001105438510580018,
+      "loss": 1.0917,
+      "step": 2618
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.4856198296686446,
+      "learning_rate": 0.0001104845011430311,
+      "loss": 1.2152,
+      "step": 2619
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.3797594860249326,
+      "learning_rate": 0.00011042514749370418,
+      "loss": 1.1409,
+      "step": 2620
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.4138750773292461,
+      "learning_rate": 0.00011036579013116155,
+      "loss": 1.1236,
+      "step": 2621
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.41362286202067755,
+      "learning_rate": 0.00011030642907654503,
+      "loss": 1.21,
+      "step": 2622
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.43483649914696554,
+      "learning_rate": 0.00011024706435099776,
+      "loss": 1.0771,
+      "step": 2623
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.39181241710764825,
+      "learning_rate": 0.0001101876959756642,
+      "loss": 1.2517,
+      "step": 2624
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.39948246143317523,
+      "learning_rate": 0.00011012832397169013,
+      "loss": 1.2977,
+      "step": 2625
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.41951276022218464,
+      "learning_rate": 0.00011006894836022256,
+      "loss": 1.2694,
+      "step": 2626
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.41046265533109205,
+      "learning_rate": 0.00011000956916240985,
+      "loss": 1.0345,
+      "step": 2627
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.4304574588872903,
+      "learning_rate": 0.00010995018639940156,
+      "loss": 1.0566,
+      "step": 2628
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.4219599199383658,
+      "learning_rate": 0.0001098908000923486,
+      "loss": 1.131,
+      "step": 2629
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.49121204065379576,
+      "learning_rate": 0.00010983141026240311,
+      "loss": 1.3337,
+      "step": 2630
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.43203939883065257,
+      "learning_rate": 0.00010977201693071852,
+      "loss": 1.2803,
+      "step": 2631
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.500485433442954,
+      "learning_rate": 0.00010971262011844938,
+      "loss": 1.1458,
+      "step": 2632
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.4355906095399076,
+      "learning_rate": 0.0001096532198467516,
+      "loss": 1.2466,
+      "step": 2633
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.45512192143870606,
+      "learning_rate": 0.00010959381613678235,
+      "loss": 1.1117,
+      "step": 2634
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.37234878162656326,
+      "learning_rate": 0.00010953440900969994,
+      "loss": 1.0608,
+      "step": 2635
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.432246465989668,
+      "learning_rate": 0.00010947499848666388,
+      "loss": 1.1414,
+      "step": 2636
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.3975649819342618,
+      "learning_rate": 0.00010941558458883496,
+      "loss": 1.2332,
+      "step": 2637
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.40004851567585686,
+      "learning_rate": 0.00010935616733737517,
+      "loss": 1.0218,
+      "step": 2638
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.40106225705342174,
+      "learning_rate": 0.00010929674675344767,
+      "loss": 1.0722,
+      "step": 2639
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.4323100673104924,
+      "learning_rate": 0.00010923732285821678,
+      "loss": 0.9918,
+      "step": 2640
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.42166113381215325,
+      "learning_rate": 0.00010917789567284804,
+      "loss": 1.105,
+      "step": 2641
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5395783999624649,
+      "learning_rate": 0.00010911846521850817,
+      "loss": 1.1263,
+      "step": 2642
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.4105136897861753,
+      "learning_rate": 0.00010905903151636501,
+      "loss": 1.2841,
+      "step": 2643
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.4692367744458125,
+      "learning_rate": 0.00010899959458758754,
+      "loss": 1.0933,
+      "step": 2644
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.47822760142911264,
+      "learning_rate": 0.000108940154453346,
+      "loss": 1.1257,
+      "step": 2645
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.4729182019876271,
+      "learning_rate": 0.00010888071113481164,
+      "loss": 1.1408,
+      "step": 2646
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.4415355800118505,
+      "learning_rate": 0.00010882126465315698,
+      "loss": 1.2957,
+      "step": 2647
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.4835246947860025,
+      "learning_rate": 0.00010876181502955553,
+      "loss": 1.0545,
+      "step": 2648
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.460384795836417,
+      "learning_rate": 0.000108702362285182,
+      "loss": 1.2743,
+      "step": 2649
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.4464020508913504,
+      "learning_rate": 0.0001086429064412122,
+      "loss": 1.4477,
+      "step": 2650
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.467719053187985,
+      "learning_rate": 0.00010858344751882304,
+      "loss": 1.3019,
+      "step": 2651
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.3965584591918505,
+      "learning_rate": 0.00010852398553919252,
+      "loss": 1.076,
+      "step": 2652
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.37557098190541566,
+      "learning_rate": 0.0001084645205234997,
+      "loss": 1.1977,
+      "step": 2653
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.4724580016260082,
+      "learning_rate": 0.00010840505249292476,
+      "loss": 1.2739,
+      "step": 2654
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.4037529097848332,
+      "learning_rate": 0.000108345581468649,
+      "loss": 1.1009,
+      "step": 2655
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.3854633020818454,
+      "learning_rate": 0.00010828610747185468,
+      "loss": 1.137,
+      "step": 2656
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.42382208542610617,
+      "learning_rate": 0.0001082266305237252,
+      "loss": 1.0167,
+      "step": 2657
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.460981252703356,
+      "learning_rate": 0.00010816715064544495,
+      "loss": 1.1393,
+      "step": 2658
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.4448923189299561,
+      "learning_rate": 0.00010810766785819946,
+      "loss": 1.3626,
+      "step": 2659
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.4236373215004859,
+      "learning_rate": 0.00010804818218317517,
+      "loss": 1.1756,
+      "step": 2660
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5020436411012712,
+      "learning_rate": 0.00010798869364155963,
+      "loss": 1.145,
+      "step": 2661
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.48618399762504677,
+      "learning_rate": 0.00010792920225454139,
+      "loss": 1.0732,
+      "step": 2662
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.47211554287592633,
+      "learning_rate": 0.00010786970804331007,
+      "loss": 1.2364,
+      "step": 2663
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.4031991740218947,
+      "learning_rate": 0.00010781021102905617,
+      "loss": 1.297,
+      "step": 2664
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.3855839749943919,
+      "learning_rate": 0.00010775071123297128,
+      "loss": 1.134,
+      "step": 2665
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.48660127000203274,
+      "learning_rate": 0.00010769120867624798,
+      "loss": 1.2175,
+      "step": 2666
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.42115878200932366,
+      "learning_rate": 0.00010763170338007978,
+      "loss": 1.2698,
+      "step": 2667
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.44251736445993295,
+      "learning_rate": 0.00010757219536566125,
+      "loss": 1.0359,
+      "step": 2668
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.4137290731344686,
+      "learning_rate": 0.00010751268465418785,
+      "loss": 1.1613,
+      "step": 2669
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.39564888575423907,
+      "learning_rate": 0.000107453171266856,
+      "loss": 1.2569,
+      "step": 2670
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.44237080493960057,
+      "learning_rate": 0.00010739365522486318,
+      "loss": 1.1443,
+      "step": 2671
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.4522518151559356,
+      "learning_rate": 0.00010733413654940767,
+      "loss": 1.2017,
+      "step": 2672
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.4101637494423119,
+      "learning_rate": 0.00010727461526168878,
+      "loss": 1.1769,
+      "step": 2673
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5428054964238067,
+      "learning_rate": 0.0001072150913829067,
+      "loss": 1.1372,
+      "step": 2674
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.4093562131612201,
+      "learning_rate": 0.00010715556493426262,
+      "loss": 1.0168,
+      "step": 2675
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.4229345608564538,
+      "learning_rate": 0.00010709603593695856,
+      "loss": 1.2979,
+      "step": 2676
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.4003925660436943,
+      "learning_rate": 0.00010703650441219749,
+      "loss": 1.16,
+      "step": 2677
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.4342649218027764,
+      "learning_rate": 0.00010697697038118326,
+      "loss": 1.1316,
+      "step": 2678
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.4654055946164656,
+      "learning_rate": 0.00010691743386512064,
+      "loss": 1.1983,
+      "step": 2679
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.4957813110795794,
+      "learning_rate": 0.00010685789488521527,
+      "loss": 1.1735,
+      "step": 2680
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5160985831257657,
+      "learning_rate": 0.00010679835346267366,
+      "loss": 1.1965,
+      "step": 2681
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.40196820335798333,
+      "learning_rate": 0.0001067388096187032,
+      "loss": 1.2106,
+      "step": 2682
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.4127256199466812,
+      "learning_rate": 0.00010667926337451217,
+      "loss": 1.11,
+      "step": 2683
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.41921656056024864,
+      "learning_rate": 0.00010661971475130963,
+      "loss": 1.058,
+      "step": 2684
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.40637411681709346,
+      "learning_rate": 0.00010656016377030555,
+      "loss": 1.1356,
+      "step": 2685
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.4496769478817081,
+      "learning_rate": 0.00010650061045271072,
+      "loss": 1.0894,
+      "step": 2686
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.4232775582747194,
+      "learning_rate": 0.00010644105481973674,
+      "loss": 1.1978,
+      "step": 2687
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.4561170608875535,
+      "learning_rate": 0.00010638149689259613,
+      "loss": 1.1272,
+      "step": 2688
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.42688542832846554,
+      "learning_rate": 0.00010632193669250207,
+      "loss": 1.1229,
+      "step": 2689
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.4058057102494925,
+      "learning_rate": 0.00010626237424066872,
+      "loss": 1.1246,
+      "step": 2690
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.41920047560251367,
+      "learning_rate": 0.00010620280955831087,
+      "loss": 1.0658,
+      "step": 2691
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.49553531797201605,
+      "learning_rate": 0.00010614324266664428,
+      "loss": 1.0838,
+      "step": 2692
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.46144827656547704,
+      "learning_rate": 0.00010608367358688536,
+      "loss": 1.2265,
+      "step": 2693
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.3815484771742994,
+      "learning_rate": 0.00010602410234025136,
+      "loss": 1.2157,
+      "step": 2694
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.42107710036133694,
+      "learning_rate": 0.00010596452894796028,
+      "loss": 1.1809,
+      "step": 2695
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.44654951391928605,
+      "learning_rate": 0.0001059049534312309,
+      "loss": 1.22,
+      "step": 2696
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.4106796295011577,
+      "learning_rate": 0.00010584537581128276,
+      "loss": 1.1589,
+      "step": 2697
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.4017479053216145,
+      "learning_rate": 0.00010578579610933612,
+      "loss": 1.068,
+      "step": 2698
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.4447425784408912,
+      "learning_rate": 0.000105726214346612,
+      "loss": 1.0645,
+      "step": 2699
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.47312287817594273,
+      "learning_rate": 0.00010566663054433223,
+      "loss": 1.069,
+      "step": 2700
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.3871640780872376,
+      "learning_rate": 0.00010560704472371919,
+      "loss": 1.2331,
+      "step": 2701
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.40236065185311604,
+      "learning_rate": 0.00010554745690599612,
+      "loss": 1.1431,
+      "step": 2702
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.38894482992547674,
+      "learning_rate": 0.00010548786711238697,
+      "loss": 1.3086,
+      "step": 2703
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.45217799220009103,
+      "learning_rate": 0.00010542827536411632,
+      "loss": 1.2247,
+      "step": 2704
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.47644895813021815,
+      "learning_rate": 0.00010536868168240949,
+      "loss": 1.0274,
+      "step": 2705
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.44281698113121776,
+      "learning_rate": 0.00010530908608849246,
+      "loss": 1.0189,
+      "step": 2706
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5174485061238785,
+      "learning_rate": 0.00010524948860359193,
+      "loss": 1.1299,
+      "step": 2707
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.45230409959067575,
+      "learning_rate": 0.0001051898892489353,
+      "loss": 1.2498,
+      "step": 2708
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.4592908516831924,
+      "learning_rate": 0.00010513028804575049,
+      "loss": 1.1223,
+      "step": 2709
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.4348776410874984,
+      "learning_rate": 0.00010507068501526627,
+      "loss": 1.0335,
+      "step": 2710
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.41415134726055136,
+      "learning_rate": 0.00010501108017871192,
+      "loss": 1.2078,
+      "step": 2711
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.40446711543680147,
+      "learning_rate": 0.0001049514735573175,
+      "loss": 1.118,
+      "step": 2712
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.41919158480833024,
+      "learning_rate": 0.00010489186517231348,
+      "loss": 1.1772,
+      "step": 2713
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.4366838605812132,
+      "learning_rate": 0.00010483225504493122,
+      "loss": 1.1653,
+      "step": 2714
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.4492345023222059,
+      "learning_rate": 0.00010477264319640252,
+      "loss": 1.0772,
+      "step": 2715
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.424506352576007,
+      "learning_rate": 0.00010471302964795988,
+      "loss": 0.9895,
+      "step": 2716
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.43625941088684594,
+      "learning_rate": 0.00010465341442083635,
+      "loss": 1.1526,
+      "step": 2717
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.47414434228862284,
+      "learning_rate": 0.00010459379753626562,
+      "loss": 0.9755,
+      "step": 2718
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.37846895965015764,
+      "learning_rate": 0.00010453417901548195,
+      "loss": 1.106,
+      "step": 2719
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.41243576204984117,
+      "learning_rate": 0.00010447455887972022,
+      "loss": 1.3277,
+      "step": 2720
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.41630904457901735,
+      "learning_rate": 0.0001044149371502158,
+      "loss": 1.1844,
+      "step": 2721
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5188735253416226,
+      "learning_rate": 0.0001043553138482047,
+      "loss": 1.2889,
+      "step": 2722
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.46248175532044217,
+      "learning_rate": 0.00010429568899492348,
+      "loss": 1.0932,
+      "step": 2723
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5219547851880618,
+      "learning_rate": 0.00010423606261160928,
+      "loss": 1.1942,
+      "step": 2724
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.39726703773381644,
+      "learning_rate": 0.00010417643471949967,
+      "loss": 1.1581,
+      "step": 2725
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5232840082943007,
+      "learning_rate": 0.00010411680533983288,
+      "loss": 1.1408,
+      "step": 2726
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.4392629768223627,
+      "learning_rate": 0.00010405717449384766,
+      "loss": 1.0467,
+      "step": 2727
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.4285018924632825,
+      "learning_rate": 0.00010399754220278319,
+      "loss": 1.2264,
+      "step": 2728
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.43417508935824817,
+      "learning_rate": 0.00010393790848787928,
+      "loss": 1.2036,
+      "step": 2729
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.483463099430095,
+      "learning_rate": 0.00010387827337037612,
+      "loss": 1.1478,
+      "step": 2730
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.44742471886510415,
+      "learning_rate": 0.0001038186368715145,
+      "loss": 1.1978,
+      "step": 2731
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.4057314194513662,
+      "learning_rate": 0.00010375899901253572,
+      "loss": 1.0653,
+      "step": 2732
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.41921036544606444,
+      "learning_rate": 0.00010369935981468144,
+      "loss": 1.2221,
+      "step": 2733
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.4361770342665071,
+      "learning_rate": 0.00010363971929919391,
+      "loss": 1.2607,
+      "step": 2734
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.4673433237738596,
+      "learning_rate": 0.00010358007748731582,
+      "loss": 1.2209,
+      "step": 2735
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.4096032661358655,
+      "learning_rate": 0.00010352043440029033,
+      "loss": 1.1715,
+      "step": 2736
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.47880188215629094,
+      "learning_rate": 0.00010346079005936102,
+      "loss": 1.1969,
+      "step": 2737
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.405753981772802,
+      "learning_rate": 0.00010340114448577189,
+      "loss": 1.135,
+      "step": 2738
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.4734420165627424,
+      "learning_rate": 0.00010334149770076747,
+      "loss": 1.2067,
+      "step": 2739
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.48454119068255236,
+      "learning_rate": 0.00010328184972559267,
+      "loss": 1.0709,
+      "step": 2740
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.3747251659053613,
+      "learning_rate": 0.00010322220058149286,
+      "loss": 1.0514,
+      "step": 2741
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.46478429261344634,
+      "learning_rate": 0.00010316255028971374,
+      "loss": 1.0268,
+      "step": 2742
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.45942009706035886,
+      "learning_rate": 0.00010310289887150149,
+      "loss": 1.1458,
+      "step": 2743
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.40759785467008136,
+      "learning_rate": 0.00010304324634810268,
+      "loss": 1.0646,
+      "step": 2744
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.4704729827011044,
+      "learning_rate": 0.0001029835927407643,
+      "loss": 1.0307,
+      "step": 2745
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.46644565466318,
+      "learning_rate": 0.00010292393807073366,
+      "loss": 1.043,
+      "step": 2746
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.47452499673291126,
+      "learning_rate": 0.00010286428235925849,
+      "loss": 1.1318,
+      "step": 2747
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.4346788273274708,
+      "learning_rate": 0.0001028046256275869,
+      "loss": 1.1002,
+      "step": 2748
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.46011405222555246,
+      "learning_rate": 0.00010274496789696734,
+      "loss": 1.0046,
+      "step": 2749
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.44491787209997075,
+      "learning_rate": 0.00010268530918864863,
+      "loss": 1.2025,
+      "step": 2750
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.4518058587376427,
+      "learning_rate": 0.0001026256495238799,
+      "loss": 1.0832,
+      "step": 2751
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.41558160260569593,
+      "learning_rate": 0.00010256598892391069,
+      "loss": 1.045,
+      "step": 2752
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.44853584450236866,
+      "learning_rate": 0.00010250632740999084,
+      "loss": 0.9875,
+      "step": 2753
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.44105489560126554,
+      "learning_rate": 0.00010244666500337046,
+      "loss": 1.0918,
+      "step": 2754
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.4517892928376344,
+      "learning_rate": 0.00010238700172530009,
+      "loss": 1.1529,
+      "step": 2755
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.528768575895962,
+      "learning_rate": 0.00010232733759703047,
+      "loss": 1.0672,
+      "step": 2756
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.4275553691865887,
+      "learning_rate": 0.00010226767263981276,
+      "loss": 1.3339,
+      "step": 2757
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.4453525321127512,
+      "learning_rate": 0.00010220800687489831,
+      "loss": 1.1971,
+      "step": 2758
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.43031119318676864,
+      "learning_rate": 0.00010214834032353877,
+      "loss": 1.1022,
+      "step": 2759
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.4458986655979861,
+      "learning_rate": 0.00010208867300698612,
+      "loss": 1.2299,
+      "step": 2760
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5048639215531708,
+      "learning_rate": 0.00010202900494649259,
+      "loss": 1.167,
+      "step": 2761
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.40631990051326894,
+      "learning_rate": 0.00010196933616331067,
+      "loss": 1.1846,
+      "step": 2762
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.40005463432986965,
+      "learning_rate": 0.0001019096666786931,
+      "loss": 1.2331,
+      "step": 2763
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.4157935090698685,
+      "learning_rate": 0.00010184999651389291,
+      "loss": 1.3045,
+      "step": 2764
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.41725717450483596,
+      "learning_rate": 0.00010179032569016334,
+      "loss": 1.2198,
+      "step": 2765
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5174436615526928,
+      "learning_rate": 0.00010173065422875783,
+      "loss": 1.0005,
+      "step": 2766
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.4765914001048855,
+      "learning_rate": 0.00010167098215093009,
+      "loss": 1.3229,
+      "step": 2767
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.4819535346346199,
+      "learning_rate": 0.0001016113094779341,
+      "loss": 1.212,
+      "step": 2768
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.4537477880576474,
+      "learning_rate": 0.00010155163623102398,
+      "loss": 1.1277,
+      "step": 2769
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.533171890189172,
+      "learning_rate": 0.000101491962431454,
+      "loss": 1.2448,
+      "step": 2770
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.4021396362260267,
+      "learning_rate": 0.00010143228810047875,
+      "loss": 1.2286,
+      "step": 2771
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.44395944067034426,
+      "learning_rate": 0.00010137261325935297,
+      "loss": 1.1437,
+      "step": 2772
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.40538603058285044,
+      "learning_rate": 0.00010131293792933157,
+      "loss": 1.2658,
+      "step": 2773
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.3876845761735228,
+      "learning_rate": 0.00010125326213166959,
+      "loss": 1.3583,
+      "step": 2774
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.47139997318405413,
+      "learning_rate": 0.00010119358588762232,
+      "loss": 1.2594,
+      "step": 2775
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.4570201942341838,
+      "learning_rate": 0.00010113390921844512,
+      "loss": 1.1459,
+      "step": 2776
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.40787692179362744,
+      "learning_rate": 0.00010107423214539365,
+      "loss": 1.4056,
+      "step": 2777
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.4055586678101137,
+      "learning_rate": 0.0001010145546897235,
+      "loss": 1.0924,
+      "step": 2778
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.4539940037123861,
+      "learning_rate": 0.00010095487687269054,
+      "loss": 1.1778,
+      "step": 2779
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.3995188073354822,
+      "learning_rate": 0.0001008951987155508,
+      "loss": 1.3043,
+      "step": 2780
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.43096165433449607,
+      "learning_rate": 0.00010083552023956031,
+      "loss": 1.1031,
+      "step": 2781
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.40496237660777773,
+      "learning_rate": 0.00010077584146597529,
+      "loss": 1.3248,
+      "step": 2782
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.43382322406166146,
+      "learning_rate": 0.00010071616241605205,
+      "loss": 1.1185,
+      "step": 2783
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.3986663301696873,
+      "learning_rate": 0.00010065648311104698,
+      "loss": 1.1841,
+      "step": 2784
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.45962574261419475,
+      "learning_rate": 0.00010059680357221662,
+      "loss": 1.1571,
+      "step": 2785
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.3990467375807131,
+      "learning_rate": 0.00010053712382081751,
+      "loss": 1.2767,
+      "step": 2786
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.4730589882541846,
+      "learning_rate": 0.00010047744387810632,
+      "loss": 1.2214,
+      "step": 2787
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.4244954443022012,
+      "learning_rate": 0.00010041776376533978,
+      "loss": 1.1431,
+      "step": 2788
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.40379294680492867,
+      "learning_rate": 0.00010035808350377469,
+      "loss": 1.323,
+      "step": 2789
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.4041435622308113,
+      "learning_rate": 0.00010029840311466787,
+      "loss": 1.1606,
+      "step": 2790
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.41216214834612724,
+      "learning_rate": 0.00010023872261927621,
+      "loss": 1.0433,
+      "step": 2791
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.4208390770777852,
+      "learning_rate": 0.00010017904203885661,
+      "loss": 1.1527,
+      "step": 2792
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5021913745282548,
+      "learning_rate": 0.00010011936139466607,
+      "loss": 1.0203,
+      "step": 2793
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.434561196336741,
+      "learning_rate": 0.00010005968070796154,
+      "loss": 0.9439,
+      "step": 2794
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.42167839739207424,
+      "learning_rate": 0.0001,
+      "loss": 1.2442,
+      "step": 2795
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.4454267193908089,
+      "learning_rate": 9.99403192920385e-05,
+      "loss": 1.101,
+      "step": 2796
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.4110656150153553,
+      "learning_rate": 9.988063860533395e-05,
+      "loss": 1.238,
+      "step": 2797
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.3900149679056662,
+      "learning_rate": 9.982095796114342e-05,
+      "loss": 1.0498,
+      "step": 2798
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6159947205953532,
+      "learning_rate": 9.976127738072381e-05,
+      "loss": 1.3573,
+      "step": 2799
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.4207708321782306,
+      "learning_rate": 9.970159688533216e-05,
+      "loss": 1.1689,
+      "step": 2800
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.4455388130496562,
+      "learning_rate": 9.964191649622532e-05,
+      "loss": 1.0934,
+      "step": 2801
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.44192735278442247,
+      "learning_rate": 9.958223623466025e-05,
+      "loss": 0.9961,
+      "step": 2802
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.3879457841426218,
+      "learning_rate": 9.952255612189368e-05,
+      "loss": 1.2022,
+      "step": 2803
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.43266496948895095,
+      "learning_rate": 9.946287617918252e-05,
+      "loss": 1.1288,
+      "step": 2804
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.39206513220027966,
+      "learning_rate": 9.94031964277834e-05,
+      "loss": 1.0857,
+      "step": 2805
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.43192573097135767,
+      "learning_rate": 9.934351688895305e-05,
+      "loss": 1.2024,
+      "step": 2806
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.45624920646223943,
+      "learning_rate": 9.9283837583948e-05,
+      "loss": 1.1692,
+      "step": 2807
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.43939494684286406,
+      "learning_rate": 9.922415853402472e-05,
+      "loss": 1.3143,
+      "step": 2808
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.43119639359355943,
+      "learning_rate": 9.916447976043972e-05,
+      "loss": 1.1335,
+      "step": 2809
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.468661982120142,
+      "learning_rate": 9.910480128444922e-05,
+      "loss": 1.0492,
+      "step": 2810
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.4903105484311023,
+      "learning_rate": 9.904512312730948e-05,
+      "loss": 1.3073,
+      "step": 2811
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.4883664650438566,
+      "learning_rate": 9.898544531027652e-05,
+      "loss": 1.1593,
+      "step": 2812
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.3902386621277154,
+      "learning_rate": 9.892576785460639e-05,
+      "loss": 1.3279,
+      "step": 2813
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.4741646392399749,
+      "learning_rate": 9.886609078155488e-05,
+      "loss": 1.1832,
+      "step": 2814
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.4614523531525487,
+      "learning_rate": 9.880641411237772e-05,
+      "loss": 1.2063,
+      "step": 2815
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.43212722700315864,
+      "learning_rate": 9.874673786833041e-05,
+      "loss": 1.256,
+      "step": 2816
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.40473740336555836,
+      "learning_rate": 9.868706207066846e-05,
+      "loss": 1.1874,
+      "step": 2817
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.48013982019082296,
+      "learning_rate": 9.862738674064705e-05,
+      "loss": 1.0709,
+      "step": 2818
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.4644392524313026,
+      "learning_rate": 9.856771189952126e-05,
+      "loss": 1.1533,
+      "step": 2819
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.3996110970683142,
+      "learning_rate": 9.850803756854604e-05,
+      "loss": 1.2934,
+      "step": 2820
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.494835027060612,
+      "learning_rate": 9.844836376897605e-05,
+      "loss": 1.1754,
+      "step": 2821
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.46729069285436686,
+      "learning_rate": 9.838869052206594e-05,
+      "loss": 1.2671,
+      "step": 2822
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.47917885416632133,
+      "learning_rate": 9.83290178490699e-05,
+      "loss": 1.0478,
+      "step": 2823
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.4168688846734565,
+      "learning_rate": 9.826934577124221e-05,
+      "loss": 1.2276,
+      "step": 2824
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.4073379555254438,
+      "learning_rate": 9.820967430983668e-05,
+      "loss": 1.1578,
+      "step": 2825
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.47057094217764145,
+      "learning_rate": 9.815000348610713e-05,
+      "loss": 1.1956,
+      "step": 2826
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.42884450324338524,
+      "learning_rate": 9.809033332130693e-05,
+      "loss": 0.9868,
+      "step": 2827
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.4131221368423854,
+      "learning_rate": 9.803066383668935e-05,
+      "loss": 1.3047,
+      "step": 2828
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.4969449447100712,
+      "learning_rate": 9.797099505350744e-05,
+      "loss": 1.161,
+      "step": 2829
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.44482720035464607,
+      "learning_rate": 9.79113269930139e-05,
+      "loss": 1.1765,
+      "step": 2830
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.42926959585619534,
+      "learning_rate": 9.785165967646127e-05,
+      "loss": 1.1202,
+      "step": 2831
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.4091269768798844,
+      "learning_rate": 9.779199312510171e-05,
+      "loss": 1.0856,
+      "step": 2832
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.47050916974229184,
+      "learning_rate": 9.773232736018725e-05,
+      "loss": 1.1282,
+      "step": 2833
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5085910236954856,
+      "learning_rate": 9.767266240296952e-05,
+      "loss": 0.9049,
+      "step": 2834
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5705377887452143,
+      "learning_rate": 9.761299827469992e-05,
+      "loss": 1.139,
+      "step": 2835
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.4409788338900152,
+      "learning_rate": 9.755333499662954e-05,
+      "loss": 1.0392,
+      "step": 2836
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.4009941194415464,
+      "learning_rate": 9.749367259000918e-05,
+      "loss": 1.2377,
+      "step": 2837
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.4014167143638734,
+      "learning_rate": 9.743401107608933e-05,
+      "loss": 1.2802,
+      "step": 2838
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5250475133392236,
+      "learning_rate": 9.737435047612012e-05,
+      "loss": 1.154,
+      "step": 2839
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.4207002153307308,
+      "learning_rate": 9.731469081135141e-05,
+      "loss": 1.1325,
+      "step": 2840
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.42508056315649995,
+      "learning_rate": 9.725503210303266e-05,
+      "loss": 1.1867,
+      "step": 2841
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.4301000677223001,
+      "learning_rate": 9.719537437241312e-05,
+      "loss": 1.0679,
+      "step": 2842
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.4317789236877776,
+      "learning_rate": 9.713571764074152e-05,
+      "loss": 1.2433,
+      "step": 2843
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.39984136197607356,
+      "learning_rate": 9.707606192926637e-05,
+      "loss": 1.1537,
+      "step": 2844
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.47195381642076484,
+      "learning_rate": 9.70164072592357e-05,
+      "loss": 1.2611,
+      "step": 2845
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.453322995720399,
+      "learning_rate": 9.695675365189734e-05,
+      "loss": 1.1574,
+      "step": 2846
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.4943300327853518,
+      "learning_rate": 9.689710112849852e-05,
+      "loss": 1.1915,
+      "step": 2847
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.44764995035044736,
+      "learning_rate": 9.68374497102863e-05,
+      "loss": 1.1037,
+      "step": 2848
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5091917595968889,
+      "learning_rate": 9.677779941850719e-05,
+      "loss": 0.8863,
+      "step": 2849
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.43329347618908143,
+      "learning_rate": 9.671815027440734e-05,
+      "loss": 1.1786,
+      "step": 2850
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.39398479845255474,
+      "learning_rate": 9.665850229923258e-05,
+      "loss": 1.1533,
+      "step": 2851
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.3945132352498431,
+      "learning_rate": 9.659885551422814e-05,
+      "loss": 1.3225,
+      "step": 2852
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.415843772261947,
+      "learning_rate": 9.653920994063903e-05,
+      "loss": 1.1585,
+      "step": 2853
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.437345857994555,
+      "learning_rate": 9.647956559970968e-05,
+      "loss": 1.0546,
+      "step": 2854
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.511573355501523,
+      "learning_rate": 9.641992251268419e-05,
+      "loss": 1.1402,
+      "step": 2855
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.46645809911085456,
+      "learning_rate": 9.636028070080608e-05,
+      "loss": 1.1056,
+      "step": 2856
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.47649023423283,
+      "learning_rate": 9.630064018531858e-05,
+      "loss": 0.9923,
+      "step": 2857
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.37694998371831984,
+      "learning_rate": 9.62410009874643e-05,
+      "loss": 0.9109,
+      "step": 2858
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.412526699411181,
+      "learning_rate": 9.618136312848551e-05,
+      "loss": 1.1988,
+      "step": 2859
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.41258550729251864,
+      "learning_rate": 9.612172662962393e-05,
+      "loss": 1.0675,
+      "step": 2860
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.43372532855243373,
+      "learning_rate": 9.606209151212076e-05,
+      "loss": 1.115,
+      "step": 2861
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.48043389128516517,
+      "learning_rate": 9.600245779721682e-05,
+      "loss": 1.151,
+      "step": 2862
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.3925389951497158,
+      "learning_rate": 9.594282550615235e-05,
+      "loss": 1.3027,
+      "step": 2863
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.4113861470140319,
+      "learning_rate": 9.588319466016713e-05,
+      "loss": 1.2866,
+      "step": 2864
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.4266282275828395,
+      "learning_rate": 9.582356528050034e-05,
+      "loss": 1.1674,
+      "step": 2865
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.4085983503142901,
+      "learning_rate": 9.576393738839076e-05,
+      "loss": 1.1778,
+      "step": 2866
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.43313372763715974,
+      "learning_rate": 9.570431100507651e-05,
+      "loss": 1.2478,
+      "step": 2867
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.4664179873947757,
+      "learning_rate": 9.564468615179533e-05,
+      "loss": 1.1441,
+      "step": 2868
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.47844361167104615,
+      "learning_rate": 9.558506284978422e-05,
+      "loss": 1.3093,
+      "step": 2869
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.4280843712560286,
+      "learning_rate": 9.552544112027981e-05,
+      "loss": 1.3255,
+      "step": 2870
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.4253225896340581,
+      "learning_rate": 9.546582098451809e-05,
+      "loss": 1.2011,
+      "step": 2871
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5097794920022186,
+      "learning_rate": 9.540620246373439e-05,
+      "loss": 1.2654,
+      "step": 2872
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.4485663210016047,
+      "learning_rate": 9.534658557916368e-05,
+      "loss": 1.1996,
+      "step": 2873
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.4419049511581746,
+      "learning_rate": 9.528697035204016e-05,
+      "loss": 1.1557,
+      "step": 2874
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5491404747716918,
+      "learning_rate": 9.52273568035975e-05,
+      "loss": 1.163,
+      "step": 2875
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.4333800443911744,
+      "learning_rate": 9.516774495506879e-05,
+      "loss": 1.1696,
+      "step": 2876
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.40320272299711407,
+      "learning_rate": 9.510813482768653e-05,
+      "loss": 1.198,
+      "step": 2877
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.441432594539895,
+      "learning_rate": 9.504852644268254e-05,
+      "loss": 1.1799,
+      "step": 2878
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.38219285691528915,
+      "learning_rate": 9.498891982128809e-05,
+      "loss": 1.2372,
+      "step": 2879
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.44879220568783157,
+      "learning_rate": 9.492931498473377e-05,
+      "loss": 1.0877,
+      "step": 2880
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.4782522870413934,
+      "learning_rate": 9.486971195424952e-05,
+      "loss": 1.0785,
+      "step": 2881
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5786031306978457,
+      "learning_rate": 9.481011075106475e-05,
+      "loss": 1.0453,
+      "step": 2882
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.4442432456611923,
+      "learning_rate": 9.475051139640809e-05,
+      "loss": 1.2706,
+      "step": 2883
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.4401712767465556,
+      "learning_rate": 9.469091391150758e-05,
+      "loss": 1.023,
+      "step": 2884
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.4875662159191855,
+      "learning_rate": 9.463131831759053e-05,
+      "loss": 1.1674,
+      "step": 2885
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.42164283346003173,
+      "learning_rate": 9.457172463588371e-05,
+      "loss": 1.1162,
+      "step": 2886
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.41166062513743296,
+      "learning_rate": 9.451213288761304e-05,
+      "loss": 1.1449,
+      "step": 2887
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.3733501373775528,
+      "learning_rate": 9.445254309400389e-05,
+      "loss": 1.172,
+      "step": 2888
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.45897204974801653,
+      "learning_rate": 9.439295527628081e-05,
+      "loss": 1.2177,
+      "step": 2889
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.4472614764010617,
+      "learning_rate": 9.43333694556678e-05,
+      "loss": 1.1427,
+      "step": 2890
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.40168660035293724,
+      "learning_rate": 9.4273785653388e-05,
+      "loss": 1.296,
+      "step": 2891
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.4094121249829959,
+      "learning_rate": 9.42142038906639e-05,
+      "loss": 1.1053,
+      "step": 2892
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.4660744129494079,
+      "learning_rate": 9.41546241887173e-05,
+      "loss": 1.1591,
+      "step": 2893
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.392790527481713,
+      "learning_rate": 9.40950465687691e-05,
+      "loss": 1.121,
+      "step": 2894
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.44271031116041987,
+      "learning_rate": 9.403547105203974e-05,
+      "loss": 1.1607,
+      "step": 2895
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.4090659635970277,
+      "learning_rate": 9.397589765974865e-05,
+      "loss": 1.2128,
+      "step": 2896
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.45668503182331194,
+      "learning_rate": 9.391632641311466e-05,
+      "loss": 1.2887,
+      "step": 2897
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.4176296622552325,
+      "learning_rate": 9.385675733335573e-05,
+      "loss": 1.1407,
+      "step": 2898
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.40349856817294,
+      "learning_rate": 9.379719044168915e-05,
+      "loss": 1.2357,
+      "step": 2899
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.49417883214218905,
+      "learning_rate": 9.373762575933129e-05,
+      "loss": 1.0339,
+      "step": 2900
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.39306534607395666,
+      "learning_rate": 9.367806330749794e-05,
+      "loss": 1.1354,
+      "step": 2901
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.42732202182527945,
+      "learning_rate": 9.36185031074039e-05,
+      "loss": 1.2712,
+      "step": 2902
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.46661121998622235,
+      "learning_rate": 9.355894518026328e-05,
+      "loss": 1.2153,
+      "step": 2903
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.43796240729962566,
+      "learning_rate": 9.349938954728934e-05,
+      "loss": 1.0093,
+      "step": 2904
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.43691968382479807,
+      "learning_rate": 9.343983622969448e-05,
+      "loss": 1.1696,
+      "step": 2905
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.4822997895343087,
+      "learning_rate": 9.338028524869041e-05,
+      "loss": 1.0963,
+      "step": 2906
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.39716654595745926,
+      "learning_rate": 9.332073662548784e-05,
+      "loss": 1.258,
+      "step": 2907
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.4152577523057511,
+      "learning_rate": 9.326119038129681e-05,
+      "loss": 1.1301,
+      "step": 2908
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.42969238756648837,
+      "learning_rate": 9.320164653732634e-05,
+      "loss": 1.067,
+      "step": 2909
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.42943859402078133,
+      "learning_rate": 9.314210511478474e-05,
+      "loss": 1.2055,
+      "step": 2910
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.4648186177633878,
+      "learning_rate": 9.308256613487937e-05,
+      "loss": 1.1671,
+      "step": 2911
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.48363614975096064,
+      "learning_rate": 9.302302961881676e-05,
+      "loss": 1.188,
+      "step": 2912
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.4294128126998854,
+      "learning_rate": 9.296349558780255e-05,
+      "loss": 1.2839,
+      "step": 2913
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.3832149261243446,
+      "learning_rate": 9.290396406304145e-05,
+      "loss": 1.0208,
+      "step": 2914
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.4127785241731782,
+      "learning_rate": 9.28444350657374e-05,
+      "loss": 1.2039,
+      "step": 2915
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.4488197308271869,
+      "learning_rate": 9.27849086170933e-05,
+      "loss": 1.2428,
+      "step": 2916
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.3663411913311612,
+      "learning_rate": 9.272538473831126e-05,
+      "loss": 1.2884,
+      "step": 2917
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.47447933033850537,
+      "learning_rate": 9.266586345059234e-05,
+      "loss": 1.2022,
+      "step": 2918
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.471391569639413,
+      "learning_rate": 9.260634477513683e-05,
+      "loss": 1.2088,
+      "step": 2919
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.4086769882561515,
+      "learning_rate": 9.254682873314399e-05,
+      "loss": 1.213,
+      "step": 2920
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5305300863291568,
+      "learning_rate": 9.248731534581218e-05,
+      "loss": 1.1275,
+      "step": 2921
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5244894451646166,
+      "learning_rate": 9.242780463433875e-05,
+      "loss": 1.0435,
+      "step": 2922
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5068723505448164,
+      "learning_rate": 9.236829661992023e-05,
+      "loss": 1.238,
+      "step": 2923
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.39315850197752633,
+      "learning_rate": 9.230879132375207e-05,
+      "loss": 1.1614,
+      "step": 2924
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.442494193205375,
+      "learning_rate": 9.224928876702874e-05,
+      "loss": 1.2134,
+      "step": 2925
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.41889570608380433,
+      "learning_rate": 9.218978897094387e-05,
+      "loss": 1.0873,
+      "step": 2926
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.4618729669659348,
+      "learning_rate": 9.213029195668996e-05,
+      "loss": 1.1144,
+      "step": 2927
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.41600278958732134,
+      "learning_rate": 9.207079774545863e-05,
+      "loss": 1.1533,
+      "step": 2928
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.3972869688690637,
+      "learning_rate": 9.201130635844038e-05,
+      "loss": 1.2132,
+      "step": 2929
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.4536769351415445,
+      "learning_rate": 9.195181781682485e-05,
+      "loss": 1.1718,
+      "step": 2930
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.4439101612600338,
+      "learning_rate": 9.189233214180056e-05,
+      "loss": 1.1851,
+      "step": 2931
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.4337322166462216,
+      "learning_rate": 9.183284935455507e-05,
+      "loss": 1.2126,
+      "step": 2932
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.39177616054971504,
+      "learning_rate": 9.177336947627486e-05,
+      "loss": 1.3495,
+      "step": 2933
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.45266946152298515,
+      "learning_rate": 9.171389252814534e-05,
+      "loss": 1.093,
+      "step": 2934
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.42117084324969045,
+      "learning_rate": 9.165441853135104e-05,
+      "loss": 1.1621,
+      "step": 2935
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.46718780587037456,
+      "learning_rate": 9.159494750707526e-05,
+      "loss": 1.1542,
+      "step": 2936
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.41476183434661684,
+      "learning_rate": 9.153547947650036e-05,
+      "loss": 1.0224,
+      "step": 2937
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.4375255050116256,
+      "learning_rate": 9.147601446080751e-05,
+      "loss": 0.8486,
+      "step": 2938
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.41259041604849556,
+      "learning_rate": 9.141655248117698e-05,
+      "loss": 1.1603,
+      "step": 2939
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.45225581551770455,
+      "learning_rate": 9.135709355878779e-05,
+      "loss": 1.0106,
+      "step": 2940
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5093319033517698,
+      "learning_rate": 9.129763771481802e-05,
+      "loss": 1.1944,
+      "step": 2941
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.41390875925928566,
+      "learning_rate": 9.123818497044446e-05,
+      "loss": 1.1755,
+      "step": 2942
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.47843341825371105,
+      "learning_rate": 9.117873534684303e-05,
+      "loss": 1.1014,
+      "step": 2943
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.42983975152924647,
+      "learning_rate": 9.111928886518837e-05,
+      "loss": 1.1211,
+      "step": 2944
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.4163560080155607,
+      "learning_rate": 9.105984554665404e-05,
+      "loss": 1.0947,
+      "step": 2945
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.42673824045150915,
+      "learning_rate": 9.10004054124125e-05,
+      "loss": 1.2458,
+      "step": 2946
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.43749277649869267,
+      "learning_rate": 9.094096848363502e-05,
+      "loss": 1.1588,
+      "step": 2947
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.3994001162199933,
+      "learning_rate": 9.088153478149189e-05,
+      "loss": 1.0862,
+      "step": 2948
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.42254871720906245,
+      "learning_rate": 9.082210432715197e-05,
+      "loss": 0.9887,
+      "step": 2949
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.39309753900554073,
+      "learning_rate": 9.076267714178323e-05,
+      "loss": 1.1844,
+      "step": 2950
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.4430634842349722,
+      "learning_rate": 9.070325324655234e-05,
+      "loss": 1.152,
+      "step": 2951
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.38283869238840196,
+      "learning_rate": 9.064383266262485e-05,
+      "loss": 1.3643,
+      "step": 2952
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.40555451448755725,
+      "learning_rate": 9.058441541116503e-05,
+      "loss": 1.1764,
+      "step": 2953
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.413653939848638,
+      "learning_rate": 9.052500151333615e-05,
+      "loss": 1.0381,
+      "step": 2954
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.506251829273982,
+      "learning_rate": 9.046559099030012e-05,
+      "loss": 1.0759,
+      "step": 2955
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.43014588450490887,
+      "learning_rate": 9.040618386321768e-05,
+      "loss": 1.1837,
+      "step": 2956
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.43383907790831855,
+      "learning_rate": 9.034678015324842e-05,
+      "loss": 1.2558,
+      "step": 2957
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.44173218213867915,
+      "learning_rate": 9.028737988155065e-05,
+      "loss": 1.003,
+      "step": 2958
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5119329121445159,
+      "learning_rate": 9.022798306928153e-05,
+      "loss": 0.9561,
+      "step": 2959
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.4246329125505336,
+      "learning_rate": 9.016858973759689e-05,
+      "loss": 1.103,
+      "step": 2960
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.40173572930553203,
+      "learning_rate": 9.010919990765141e-05,
+      "loss": 1.1793,
+      "step": 2961
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.4573699355509967,
+      "learning_rate": 9.004981360059844e-05,
+      "loss": 1.0019,
+      "step": 2962
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5049129804692625,
+      "learning_rate": 8.999043083759017e-05,
+      "loss": 1.1745,
+      "step": 2963
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.4247591478779381,
+      "learning_rate": 8.993105163977745e-05,
+      "loss": 1.1547,
+      "step": 2964
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.52417826437679,
+      "learning_rate": 8.987167602830989e-05,
+      "loss": 1.0554,
+      "step": 2965
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.45867767982991164,
+      "learning_rate": 8.981230402433583e-05,
+      "loss": 1.2462,
+      "step": 2966
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.43040735687933795,
+      "learning_rate": 8.975293564900226e-05,
+      "loss": 1.1156,
+      "step": 2967
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.45110804269171995,
+      "learning_rate": 8.9693570923455e-05,
+      "loss": 1.1611,
+      "step": 2968
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.45780731351595794,
+      "learning_rate": 8.963420986883848e-05,
+      "loss": 0.9267,
+      "step": 2969
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.44147114672026344,
+      "learning_rate": 8.957485250629584e-05,
+      "loss": 1.018,
+      "step": 2970
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.3835157018586977,
+      "learning_rate": 8.951549885696889e-05,
+      "loss": 1.3744,
+      "step": 2971
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5144339676294876,
+      "learning_rate": 8.945614894199821e-05,
+      "loss": 1.0163,
+      "step": 2972
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.42201930038381136,
+      "learning_rate": 8.939680278252287e-05,
+      "loss": 1.0796,
+      "step": 2973
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.491747659253328,
+      "learning_rate": 8.93374603996808e-05,
+      "loss": 1.0883,
+      "step": 2974
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.4215443917538474,
+      "learning_rate": 8.927812181460843e-05,
+      "loss": 1.171,
+      "step": 2975
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5781465698792495,
+      "learning_rate": 8.921878704844096e-05,
+      "loss": 0.9161,
+      "step": 2976
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.4814802234922726,
+      "learning_rate": 8.915945612231214e-05,
+      "loss": 1.2181,
+      "step": 2977
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.39980181227760686,
+      "learning_rate": 8.910012905735435e-05,
+      "loss": 1.1674,
+      "step": 2978
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.416731291033609,
+      "learning_rate": 8.904080587469868e-05,
+      "loss": 1.2549,
+      "step": 2979
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.4924258671477863,
+      "learning_rate": 8.898148659547476e-05,
+      "loss": 1.0733,
+      "step": 2980
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.4051179642397773,
+      "learning_rate": 8.892217124081088e-05,
+      "loss": 1.1734,
+      "step": 2981
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.4275855825531411,
+      "learning_rate": 8.886285983183386e-05,
+      "loss": 1.222,
+      "step": 2982
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.4659269099216692,
+      "learning_rate": 8.880355238966923e-05,
+      "loss": 1.2001,
+      "step": 2983
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.41439177232891705,
+      "learning_rate": 8.874424893544098e-05,
+      "loss": 1.178,
+      "step": 2984
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.4088738547545015,
+      "learning_rate": 8.868494949027176e-05,
+      "loss": 1.1675,
+      "step": 2985
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.46497153152537574,
+      "learning_rate": 8.862565407528275e-05,
+      "loss": 1.2357,
+      "step": 2986
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.41460993907627963,
+      "learning_rate": 8.856636271159377e-05,
+      "loss": 1.1835,
+      "step": 2987
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.40510756516161006,
+      "learning_rate": 8.85070754203231e-05,
+      "loss": 1.1741,
+      "step": 2988
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.4493737436509115,
+      "learning_rate": 8.844779222258761e-05,
+      "loss": 1.212,
+      "step": 2989
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.4719973565738947,
+      "learning_rate": 8.838851313950275e-05,
+      "loss": 1.0926,
+      "step": 2990
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.4919350464975654,
+      "learning_rate": 8.832923819218238e-05,
+      "loss": 0.9486,
+      "step": 2991
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.43777242452575843,
+      "learning_rate": 8.826996740173908e-05,
+      "loss": 1.2446,
+      "step": 2992
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.4811041114001849,
+      "learning_rate": 8.821070078928379e-05,
+      "loss": 1.0689,
+      "step": 2993
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.43131987771054014,
+      "learning_rate": 8.815143837592604e-05,
+      "loss": 1.2538,
+      "step": 2994
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.4145052326021495,
+      "learning_rate": 8.809218018277378e-05,
+      "loss": 1.1628,
+      "step": 2995
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.4170302988546796,
+      "learning_rate": 8.803292623093363e-05,
+      "loss": 1.2011,
+      "step": 2996
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5010220400511376,
+      "learning_rate": 8.797367654151052e-05,
+      "loss": 1.0228,
+      "step": 2997
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.42689984109129336,
+      "learning_rate": 8.791443113560787e-05,
+      "loss": 1.1479,
+      "step": 2998
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.4491946889983466,
+      "learning_rate": 8.785519003432774e-05,
+      "loss": 0.9919,
+      "step": 2999
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.4311247396888756,
+      "learning_rate": 8.779595325877047e-05,
+      "loss": 1.0344,
+      "step": 3000
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.43185679304334335,
+      "learning_rate": 8.7736720830035e-05,
+      "loss": 1.32,
+      "step": 3001
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.4170743075956058,
+      "learning_rate": 8.767749276921859e-05,
+      "loss": 1.2649,
+      "step": 3002
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.4419807857313315,
+      "learning_rate": 8.761826909741709e-05,
+      "loss": 1.2436,
+      "step": 3003
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.4199968035314524,
+      "learning_rate": 8.755904983572466e-05,
+      "loss": 1.1027,
+      "step": 3004
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.3917890039399922,
+      "learning_rate": 8.7499835005234e-05,
+      "loss": 1.3308,
+      "step": 3005
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.4751888271718363,
+      "learning_rate": 8.744062462703608e-05,
+      "loss": 1.0812,
+      "step": 3006
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.44501667195619815,
+      "learning_rate": 8.738141872222048e-05,
+      "loss": 1.048,
+      "step": 3007
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.4102378887016195,
+      "learning_rate": 8.732221731187501e-05,
+      "loss": 1.0829,
+      "step": 3008
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.44877347473465967,
+      "learning_rate": 8.726302041708597e-05,
+      "loss": 0.9086,
+      "step": 3009
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.4259870023996932,
+      "learning_rate": 8.720382805893805e-05,
+      "loss": 1.0418,
+      "step": 3010
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5483058372379935,
+      "learning_rate": 8.714464025851427e-05,
+      "loss": 1.1264,
+      "step": 3011
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.4258529373212258,
+      "learning_rate": 8.708545703689613e-05,
+      "loss": 1.1725,
+      "step": 3012
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.442329532826965,
+      "learning_rate": 8.702627841516338e-05,
+      "loss": 1.1436,
+      "step": 3013
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.4913934525366949,
+      "learning_rate": 8.696710441439421e-05,
+      "loss": 1.1211,
+      "step": 3014
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.42128591746418076,
+      "learning_rate": 8.690793505566511e-05,
+      "loss": 1.3057,
+      "step": 3015
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.4366508964377608,
+      "learning_rate": 8.684877036005101e-05,
+      "loss": 1.1532,
+      "step": 3016
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.45719412737318066,
+      "learning_rate": 8.678961034862505e-05,
+      "loss": 1.1033,
+      "step": 3017
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.3978665368976384,
+      "learning_rate": 8.673045504245881e-05,
+      "loss": 1.1314,
+      "step": 3018
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.4636276756498325,
+      "learning_rate": 8.667130446262214e-05,
+      "loss": 1.2183,
+      "step": 3019
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.4162175602215256,
+      "learning_rate": 8.661215863018317e-05,
+      "loss": 1.1943,
+      "step": 3020
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.46110808267625564,
+      "learning_rate": 8.655301756620847e-05,
+      "loss": 1.1147,
+      "step": 3021
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.45705204286176254,
+      "learning_rate": 8.649388129176277e-05,
+      "loss": 1.156,
+      "step": 3022
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.41629209506272374,
+      "learning_rate": 8.643474982790919e-05,
+      "loss": 1.3202,
+      "step": 3023
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.45939231044612605,
+      "learning_rate": 8.637562319570905e-05,
+      "loss": 1.1312,
+      "step": 3024
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.464422686098712,
+      "learning_rate": 8.631650141622207e-05,
+      "loss": 1.3201,
+      "step": 3025
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.44410684362189656,
+      "learning_rate": 8.625738451050606e-05,
+      "loss": 1.2828,
+      "step": 3026
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.4375323847587673,
+      "learning_rate": 8.619827249961733e-05,
+      "loss": 1.1074,
+      "step": 3027
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.4163013030738542,
+      "learning_rate": 8.613916540461024e-05,
+      "loss": 1.2823,
+      "step": 3028
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.4391822847324534,
+      "learning_rate": 8.608006324653753e-05,
+      "loss": 1.1687,
+      "step": 3029
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.40001046367823695,
+      "learning_rate": 8.602096604645009e-05,
+      "loss": 1.2482,
+      "step": 3030
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.42479592657762233,
+      "learning_rate": 8.596187382539705e-05,
+      "loss": 1.1487,
+      "step": 3031
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.44886294348626354,
+      "learning_rate": 8.590278660442591e-05,
+      "loss": 1.1735,
+      "step": 3032
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.4059128912215593,
+      "learning_rate": 8.584370440458219e-05,
+      "loss": 1.1675,
+      "step": 3033
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.4168977085562774,
+      "learning_rate": 8.578462724690977e-05,
+      "loss": 1.1953,
+      "step": 3034
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.48230359051733535,
+      "learning_rate": 8.57255551524506e-05,
+      "loss": 1.0941,
+      "step": 3035
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.49127851982403953,
+      "learning_rate": 8.5666488142245e-05,
+      "loss": 1.0664,
+      "step": 3036
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.43770340621828496,
+      "learning_rate": 8.560742623733131e-05,
+      "loss": 1.1182,
+      "step": 3037
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.4364011308203448,
+      "learning_rate": 8.554836945874618e-05,
+      "loss": 0.9813,
+      "step": 3038
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.4833382200079468,
+      "learning_rate": 8.548931782752429e-05,
+      "loss": 1.1078,
+      "step": 3039
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6255598069742176,
+      "learning_rate": 8.543027136469868e-05,
+      "loss": 1.1452,
+      "step": 3040
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.40255908804425944,
+      "learning_rate": 8.53712300913004e-05,
+      "loss": 1.0913,
+      "step": 3041
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.458131533852643,
+      "learning_rate": 8.531219402835867e-05,
+      "loss": 1.1195,
+      "step": 3042
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.43866264017389683,
+      "learning_rate": 8.525316319690092e-05,
+      "loss": 1.0544,
+      "step": 3043
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.42167140488632904,
+      "learning_rate": 8.519413761795261e-05,
+      "loss": 1.1324,
+      "step": 3044
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.4836718751058472,
+      "learning_rate": 8.51351173125375e-05,
+      "loss": 0.9125,
+      "step": 3045
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.3942242461478137,
+      "learning_rate": 8.507610230167726e-05,
+      "loss": 1.1853,
+      "step": 3046
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.39299774532645065,
+      "learning_rate": 8.501709260639186e-05,
+      "loss": 1.3243,
+      "step": 3047
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.4505848993565066,
+      "learning_rate": 8.495808824769924e-05,
+      "loss": 1.0877,
+      "step": 3048
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.47800264997969405,
+      "learning_rate": 8.489908924661557e-05,
+      "loss": 1.2289,
+      "step": 3049
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.4287691666411929,
+      "learning_rate": 8.484009562415496e-05,
+      "loss": 1.1075,
+      "step": 3050
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.4040944389268758,
+      "learning_rate": 8.47811074013297e-05,
+      "loss": 1.0858,
+      "step": 3051
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.39117045433986547,
+      "learning_rate": 8.472212459915019e-05,
+      "loss": 1.3418,
+      "step": 3052
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5524530668819022,
+      "learning_rate": 8.46631472386248e-05,
+      "loss": 0.994,
+      "step": 3053
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.40428544868845423,
+      "learning_rate": 8.460417534076004e-05,
+      "loss": 1.2577,
+      "step": 3054
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.45097728390303954,
+      "learning_rate": 8.454520892656038e-05,
+      "loss": 1.1683,
+      "step": 3055
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5030007362075821,
+      "learning_rate": 8.448624801702851e-05,
+      "loss": 1.0178,
+      "step": 3056
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.4287687665819392,
+      "learning_rate": 8.442729263316497e-05,
+      "loss": 1.3716,
+      "step": 3057
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.37959397452641824,
+      "learning_rate": 8.436834279596845e-05,
+      "loss": 1.1801,
+      "step": 3058
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.46776549665981465,
+      "learning_rate": 8.430939852643558e-05,
+      "loss": 1.1294,
+      "step": 3059
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.41561837144972436,
+      "learning_rate": 8.425045984556114e-05,
+      "loss": 1.1556,
+      "step": 3060
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.4743444906416586,
+      "learning_rate": 8.419152677433778e-05,
+      "loss": 1.1844,
+      "step": 3061
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.45705621886249675,
+      "learning_rate": 8.413259933375618e-05,
+      "loss": 1.1521,
+      "step": 3062
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.41281576686171,
+      "learning_rate": 8.407367754480511e-05,
+      "loss": 1.132,
+      "step": 3063
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.48881534584001485,
+      "learning_rate": 8.401476142847115e-05,
+      "loss": 1.1202,
+      "step": 3064
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5326151136357101,
+      "learning_rate": 8.39558510057391e-05,
+      "loss": 1.0395,
+      "step": 3065
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.4210957515435548,
+      "learning_rate": 8.38969462975915e-05,
+      "loss": 1.0757,
+      "step": 3066
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.4658588520065389,
+      "learning_rate": 8.383804732500902e-05,
+      "loss": 1.2785,
+      "step": 3067
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.40991618683414893,
+      "learning_rate": 8.377915410897013e-05,
+      "loss": 1.1913,
+      "step": 3068
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.40017277524407746,
+      "learning_rate": 8.372026667045146e-05,
+      "loss": 0.983,
+      "step": 3069
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.40425372657539044,
+      "learning_rate": 8.366138503042738e-05,
+      "loss": 1.0406,
+      "step": 3070
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.46800467585163175,
+      "learning_rate": 8.36025092098703e-05,
+      "loss": 1.2549,
+      "step": 3071
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.40969567594886147,
+      "learning_rate": 8.354363922975055e-05,
+      "loss": 1.072,
+      "step": 3072
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.4668973954204942,
+      "learning_rate": 8.34847751110363e-05,
+      "loss": 1.1162,
+      "step": 3073
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.4502213200519017,
+      "learning_rate": 8.342591687469381e-05,
+      "loss": 1.1654,
+      "step": 3074
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.4075305457344678,
+      "learning_rate": 8.336706454168701e-05,
+      "loss": 1.0819,
+      "step": 3075
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.4541955831404862,
+      "learning_rate": 8.330821813297793e-05,
+      "loss": 1.1135,
+      "step": 3076
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5197689269665257,
+      "learning_rate": 8.324937766952638e-05,
+      "loss": 1.1472,
+      "step": 3077
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.4783497029320874,
+      "learning_rate": 8.319054317229009e-05,
+      "loss": 1.1555,
+      "step": 3078
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.4154369255507592,
+      "learning_rate": 8.313171466222461e-05,
+      "loss": 1.1777,
+      "step": 3079
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.4223544452662213,
+      "learning_rate": 8.307289216028348e-05,
+      "loss": 1.2923,
+      "step": 3080
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.490003657034207,
+      "learning_rate": 8.301407568741799e-05,
+      "loss": 0.9738,
+      "step": 3081
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.4120135019090449,
+      "learning_rate": 8.295526526457732e-05,
+      "loss": 1.1591,
+      "step": 3082
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.3956613094982569,
+      "learning_rate": 8.289646091270849e-05,
+      "loss": 1.2062,
+      "step": 3083
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.4214294184903558,
+      "learning_rate": 8.28376626527563e-05,
+      "loss": 1.1823,
+      "step": 3084
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.4672455906316201,
+      "learning_rate": 8.277887050566354e-05,
+      "loss": 1.2448,
+      "step": 3085
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.4949524902150619,
+      "learning_rate": 8.272008449237065e-05,
+      "loss": 1.1491,
+      "step": 3086
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.48229964064584185,
+      "learning_rate": 8.266130463381602e-05,
+      "loss": 1.0536,
+      "step": 3087
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.49372114266091127,
+      "learning_rate": 8.260253095093569e-05,
+      "loss": 1.0538,
+      "step": 3088
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.4282609804504164,
+      "learning_rate": 8.254376346466371e-05,
+      "loss": 1.2214,
+      "step": 3089
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.47457432331257066,
+      "learning_rate": 8.24850021959317e-05,
+      "loss": 1.2794,
+      "step": 3090
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.44570500955730535,
+      "learning_rate": 8.242624716566927e-05,
+      "loss": 0.8829,
+      "step": 3091
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.44337664472473937,
+      "learning_rate": 8.236749839480361e-05,
+      "loss": 1.2194,
+      "step": 3092
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5038037188541759,
+      "learning_rate": 8.230875590425988e-05,
+      "loss": 0.9455,
+      "step": 3093
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.42608772722921956,
+      "learning_rate": 8.225001971496087e-05,
+      "loss": 1.1142,
+      "step": 3094
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.41907486050889714,
+      "learning_rate": 8.219128984782712e-05,
+      "loss": 1.2865,
+      "step": 3095
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.44824381327671703,
+      "learning_rate": 8.213256632377702e-05,
+      "loss": 1.1297,
+      "step": 3096
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.4894621823999029,
+      "learning_rate": 8.207384916372658e-05,
+      "loss": 1.0936,
+      "step": 3097
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.3883493479050727,
+      "learning_rate": 8.20151383885897e-05,
+      "loss": 1.1576,
+      "step": 3098
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.4438771524140683,
+      "learning_rate": 8.195643401927777e-05,
+      "loss": 1.1778,
+      "step": 3099
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.3990819741264743,
+      "learning_rate": 8.189773607670015e-05,
+      "loss": 1.1381,
+      "step": 3100
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.3735364600967238,
+      "learning_rate": 8.183904458176374e-05,
+      "loss": 1.1819,
+      "step": 3101
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5197315254226554,
+      "learning_rate": 8.178035955537325e-05,
+      "loss": 0.9597,
+      "step": 3102
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.4473695617231478,
+      "learning_rate": 8.172168101843099e-05,
+      "loss": 1.1518,
+      "step": 3103
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.4421978741381111,
+      "learning_rate": 8.1663008991837e-05,
+      "loss": 1.1075,
+      "step": 3104
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.4397514012290237,
+      "learning_rate": 8.160434349648907e-05,
+      "loss": 1.1495,
+      "step": 3105
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.4553087052512707,
+      "learning_rate": 8.154568455328254e-05,
+      "loss": 1.0801,
+      "step": 3106
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.45706112738534754,
+      "learning_rate": 8.148703218311053e-05,
+      "loss": 1.0796,
+      "step": 3107
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.46032762448308123,
+      "learning_rate": 8.14283864068637e-05,
+      "loss": 1.1565,
+      "step": 3108
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.4258234406475577,
+      "learning_rate": 8.136974724543049e-05,
+      "loss": 1.3147,
+      "step": 3109
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.4072734485653768,
+      "learning_rate": 8.131111471969688e-05,
+      "loss": 1.1905,
+      "step": 3110
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.4541169611087308,
+      "learning_rate": 8.125248885054656e-05,
+      "loss": 1.2486,
+      "step": 3111
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.4188185535768401,
+      "learning_rate": 8.119386965886077e-05,
+      "loss": 1.2225,
+      "step": 3112
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.4615955511160636,
+      "learning_rate": 8.113525716551851e-05,
+      "loss": 1.1692,
+      "step": 3113
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.42482862164417673,
+      "learning_rate": 8.107665139139622e-05,
+      "loss": 1.1031,
+      "step": 3114
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.4426072120222796,
+      "learning_rate": 8.101805235736804e-05,
+      "loss": 1.1313,
+      "step": 3115
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.44284118631154973,
+      "learning_rate": 8.095946008430573e-05,
+      "loss": 1.1586,
+      "step": 3116
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.4604004102086753,
+      "learning_rate": 8.090087459307854e-05,
+      "loss": 1.2554,
+      "step": 3117
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.445729774359789,
+      "learning_rate": 8.084229590455349e-05,
+      "loss": 1.0058,
+      "step": 3118
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.3920763706871372,
+      "learning_rate": 8.078372403959498e-05,
+      "loss": 1.2964,
+      "step": 3119
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.37792662217020007,
+      "learning_rate": 8.072515901906508e-05,
+      "loss": 1.2609,
+      "step": 3120
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.4095627569211982,
+      "learning_rate": 8.06666008638234e-05,
+      "loss": 1.1431,
+      "step": 3121
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.43349377782571535,
+      "learning_rate": 8.060804959472713e-05,
+      "loss": 1.2048,
+      "step": 3122
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.38973428837263535,
+      "learning_rate": 8.054950523263096e-05,
+      "loss": 1.2005,
+      "step": 3123
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.44666867197815446,
+      "learning_rate": 8.049096779838719e-05,
+      "loss": 1.2262,
+      "step": 3124
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.4344059507425389,
+      "learning_rate": 8.043243731284557e-05,
+      "loss": 1.1181,
+      "step": 3125
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.4576106074441134,
+      "learning_rate": 8.03739137968534e-05,
+      "loss": 1.1445,
+      "step": 3126
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.4622648433424955,
+      "learning_rate": 8.031539727125554e-05,
+      "loss": 1.1506,
+      "step": 3127
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.4101797170378805,
+      "learning_rate": 8.025688775689428e-05,
+      "loss": 1.1995,
+      "step": 3128
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.4333303969023405,
+      "learning_rate": 8.019838527460957e-05,
+      "loss": 1.1276,
+      "step": 3129
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.4244568335676038,
+      "learning_rate": 8.013988984523866e-05,
+      "loss": 1.0998,
+      "step": 3130
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.42668895376288746,
+      "learning_rate": 8.008140148961641e-05,
+      "loss": 1.195,
+      "step": 3131
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.45446818860646704,
+      "learning_rate": 8.002292022857509e-05,
+      "loss": 1.2136,
+      "step": 3132
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.43772466517138253,
+      "learning_rate": 7.996444608294456e-05,
+      "loss": 1.1886,
+      "step": 3133
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.4206109813719567,
+      "learning_rate": 7.990597907355198e-05,
+      "loss": 1.19,
+      "step": 3134
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.417074446842182,
+      "learning_rate": 7.984751922122214e-05,
+      "loss": 1.095,
+      "step": 3135
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.40849518738544377,
+      "learning_rate": 7.978906654677713e-05,
+      "loss": 1.2175,
+      "step": 3136
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.4129004437285376,
+      "learning_rate": 7.973062107103653e-05,
+      "loss": 1.0839,
+      "step": 3137
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.42153557926603713,
+      "learning_rate": 7.967218281481745e-05,
+      "loss": 1.1093,
+      "step": 3138
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.4524152144127746,
+      "learning_rate": 7.96137517989343e-05,
+      "loss": 1.1085,
+      "step": 3139
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.4206632380757849,
+      "learning_rate": 7.955532804419899e-05,
+      "loss": 1.2703,
+      "step": 3140
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.4702276832348514,
+      "learning_rate": 7.949691157142079e-05,
+      "loss": 1.0518,
+      "step": 3141
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.43221963336203273,
+      "learning_rate": 7.943850240140644e-05,
+      "loss": 0.9401,
+      "step": 3142
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.41611199999619064,
+      "learning_rate": 7.938010055496002e-05,
+      "loss": 1.271,
+      "step": 3143
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.41471124244422974,
+      "learning_rate": 7.932170605288307e-05,
+      "loss": 0.9872,
+      "step": 3144
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.39211867151866625,
+      "learning_rate": 7.926331891597436e-05,
+      "loss": 1.1711,
+      "step": 3145
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.4507937485909854,
+      "learning_rate": 7.920493916503027e-05,
+      "loss": 1.2034,
+      "step": 3146
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.4198913317096475,
+      "learning_rate": 7.914656682084437e-05,
+      "loss": 1.0854,
+      "step": 3147
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.41000595151966535,
+      "learning_rate": 7.908820190420763e-05,
+      "loss": 1.0837,
+      "step": 3148
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.4066329654494708,
+      "learning_rate": 7.902984443590844e-05,
+      "loss": 1.2814,
+      "step": 3149
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.3776093472088061,
+      "learning_rate": 7.89714944367324e-05,
+      "loss": 1.3413,
+      "step": 3150
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.46508469198142965,
+      "learning_rate": 7.891315192746267e-05,
+      "loss": 1.1374,
+      "step": 3151
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.4503817965157996,
+      "learning_rate": 7.885481692887947e-05,
+      "loss": 1.2232,
+      "step": 3152
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.45313487076268516,
+      "learning_rate": 7.87964894617606e-05,
+      "loss": 1.0442,
+      "step": 3153
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.4040497360845687,
+      "learning_rate": 7.873816954688099e-05,
+      "loss": 1.0048,
+      "step": 3154
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.4881120589487679,
+      "learning_rate": 7.867985720501301e-05,
+      "loss": 0.9358,
+      "step": 3155
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.4178084301932167,
+      "learning_rate": 7.862155245692623e-05,
+      "loss": 1.1614,
+      "step": 3156
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.39935354197027634,
+      "learning_rate": 7.856325532338753e-05,
+      "loss": 1.0926,
+      "step": 3157
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.42916147172279917,
+      "learning_rate": 7.85049658251612e-05,
+      "loss": 1.219,
+      "step": 3158
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.44465439100582543,
+      "learning_rate": 7.844668398300865e-05,
+      "loss": 1.1627,
+      "step": 3159
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.4049394283216283,
+      "learning_rate": 7.838840981768868e-05,
+      "loss": 1.2066,
+      "step": 3160
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.441868612110153,
+      "learning_rate": 7.833014334995723e-05,
+      "loss": 0.9876,
+      "step": 3161
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.508580146084791,
+      "learning_rate": 7.827188460056768e-05,
+      "loss": 1.1455,
+      "step": 3162
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.4742565124123365,
+      "learning_rate": 7.821363359027048e-05,
+      "loss": 1.216,
+      "step": 3163
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.43213998240915225,
+      "learning_rate": 7.815539033981344e-05,
+      "loss": 1.0937,
+      "step": 3164
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.39831126529554933,
+      "learning_rate": 7.809715486994152e-05,
+      "loss": 1.2574,
+      "step": 3165
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.43565408421876767,
+      "learning_rate": 7.803892720139701e-05,
+      "loss": 1.1399,
+      "step": 3166
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.4134378714687648,
+      "learning_rate": 7.798070735491936e-05,
+      "loss": 1.136,
+      "step": 3167
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.44080855566295823,
+      "learning_rate": 7.79224953512452e-05,
+      "loss": 1.0191,
+      "step": 3168
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.4360299922097329,
+      "learning_rate": 7.786429121110845e-05,
+      "loss": 1.1127,
+      "step": 3169
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.49271673227454565,
+      "learning_rate": 7.780609495524012e-05,
+      "loss": 1.2967,
+      "step": 3170
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.4855778705312797,
+      "learning_rate": 7.774790660436858e-05,
+      "loss": 1.0552,
+      "step": 3171
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.42632852579432223,
+      "learning_rate": 7.76897261792192e-05,
+      "loss": 1.0742,
+      "step": 3172
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.41623647899241945,
+      "learning_rate": 7.763155370051465e-05,
+      "loss": 1.1817,
+      "step": 3173
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.4788602402140263,
+      "learning_rate": 7.757338918897468e-05,
+      "loss": 1.0066,
+      "step": 3174
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5068754718222824,
+      "learning_rate": 7.751523266531634e-05,
+      "loss": 1.2071,
+      "step": 3175
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.40563388778852383,
+      "learning_rate": 7.745708415025363e-05,
+      "loss": 1.0826,
+      "step": 3176
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.49054373981454386,
+      "learning_rate": 7.739894366449794e-05,
+      "loss": 1.0396,
+      "step": 3177
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.4903579970846335,
+      "learning_rate": 7.73408112287576e-05,
+      "loss": 0.9969,
+      "step": 3178
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.38836204705178706,
+      "learning_rate": 7.728268686373814e-05,
+      "loss": 1.0801,
+      "step": 3179
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.46162079618470736,
+      "learning_rate": 7.722457059014225e-05,
+      "loss": 1.0354,
+      "step": 3180
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.400431519197561,
+      "learning_rate": 7.716646242866967e-05,
+      "loss": 1.1895,
+      "step": 3181
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.43440425774692343,
+      "learning_rate": 7.710836240001737e-05,
+      "loss": 1.2074,
+      "step": 3182
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.3890990275395917,
+      "learning_rate": 7.705027052487925e-05,
+      "loss": 1.2954,
+      "step": 3183
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.38925786522065126,
+      "learning_rate": 7.699218682394649e-05,
+      "loss": 1.1601,
+      "step": 3184
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.4512241787737713,
+      "learning_rate": 7.693411131790717e-05,
+      "loss": 1.0791,
+      "step": 3185
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.38396135147792587,
+      "learning_rate": 7.687604402744663e-05,
+      "loss": 1.2965,
+      "step": 3186
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.4474826804953935,
+      "learning_rate": 7.681798497324716e-05,
+      "loss": 1.2089,
+      "step": 3187
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.4216842084548054,
+      "learning_rate": 7.675993417598821e-05,
+      "loss": 1.1098,
+      "step": 3188
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5058929032745334,
+      "learning_rate": 7.670189165634621e-05,
+      "loss": 1.0686,
+      "step": 3189
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.44001633056499345,
+      "learning_rate": 7.664385743499462e-05,
+      "loss": 1.0889,
+      "step": 3190
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.469562824226135,
+      "learning_rate": 7.658583153260407e-05,
+      "loss": 1.1704,
+      "step": 3191
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.4153639189163479,
+      "learning_rate": 7.65278139698421e-05,
+      "loss": 1.0845,
+      "step": 3192
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.43442475848576023,
+      "learning_rate": 7.646980476737338e-05,
+      "loss": 1.1369,
+      "step": 3193
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.4296402481630513,
+      "learning_rate": 7.641180394585947e-05,
+      "loss": 1.2755,
+      "step": 3194
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.4152459956050618,
+      "learning_rate": 7.635381152595915e-05,
+      "loss": 1.122,
+      "step": 3195
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.443453457969796,
+      "learning_rate": 7.629582752832798e-05,
+      "loss": 1.1733,
+      "step": 3196
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.45580598256138766,
+      "learning_rate": 7.62378519736187e-05,
+      "loss": 1.104,
+      "step": 3197
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5045041506084248,
+      "learning_rate": 7.617988488248089e-05,
+      "loss": 1.2028,
+      "step": 3198
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.490701166359041,
+      "learning_rate": 7.612192627556128e-05,
+      "loss": 1.2193,
+      "step": 3199
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.4427179124353798,
+      "learning_rate": 7.606397617350349e-05,
+      "loss": 1.2176,
+      "step": 3200
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.41285393445970753,
+      "learning_rate": 7.600603459694799e-05,
+      "loss": 1.1389,
+      "step": 3201
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.48543381491852566,
+      "learning_rate": 7.594810156653247e-05,
+      "loss": 1.1976,
+      "step": 3202
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.4191376487115718,
+      "learning_rate": 7.589017710289139e-05,
+      "loss": 1.1788,
+      "step": 3203
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.44781647413653447,
+      "learning_rate": 7.583226122665622e-05,
+      "loss": 1.0756,
+      "step": 3204
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5015990638280492,
+      "learning_rate": 7.577435395845533e-05,
+      "loss": 1.0689,
+      "step": 3205
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.41984592638646506,
+      "learning_rate": 7.571645531891411e-05,
+      "loss": 1.0924,
+      "step": 3206
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.43051463629226716,
+      "learning_rate": 7.56585653286548e-05,
+      "loss": 1.0669,
+      "step": 3207
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.4382844916651777,
+      "learning_rate": 7.56006840082966e-05,
+      "loss": 1.0993,
+      "step": 3208
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.4061105498762294,
+      "learning_rate": 7.554281137845558e-05,
+      "loss": 1.1954,
+      "step": 3209
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.42705699622226795,
+      "learning_rate": 7.54849474597447e-05,
+      "loss": 1.1169,
+      "step": 3210
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.41951023768980894,
+      "learning_rate": 7.542709227277396e-05,
+      "loss": 1.1656,
+      "step": 3211
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.46715904640102723,
+      "learning_rate": 7.536924583815007e-05,
+      "loss": 1.2018,
+      "step": 3212
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.44065781927058645,
+      "learning_rate": 7.531140817647673e-05,
+      "loss": 1.0982,
+      "step": 3213
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.4248749201072954,
+      "learning_rate": 7.525357930835444e-05,
+      "loss": 1.0857,
+      "step": 3214
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.45974589122263837,
+      "learning_rate": 7.519575925438067e-05,
+      "loss": 1.0636,
+      "step": 3215
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.4215704052136378,
+      "learning_rate": 7.513794803514966e-05,
+      "loss": 1.3749,
+      "step": 3216
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.4213111910291883,
+      "learning_rate": 7.508014567125256e-05,
+      "loss": 1.0908,
+      "step": 3217
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.371716467140509,
+      "learning_rate": 7.502235218327731e-05,
+      "loss": 1.0206,
+      "step": 3218
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5057963869381996,
+      "learning_rate": 7.496456759180875e-05,
+      "loss": 1.2204,
+      "step": 3219
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.4361315136531467,
+      "learning_rate": 7.490679191742853e-05,
+      "loss": 1.194,
+      "step": 3220
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.3837099514962716,
+      "learning_rate": 7.484902518071506e-05,
+      "loss": 1.0669,
+      "step": 3221
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.4371417612913872,
+      "learning_rate": 7.47912674022437e-05,
+      "loss": 1.1009,
+      "step": 3222
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.4465988266017963,
+      "learning_rate": 7.473351860258643e-05,
+      "loss": 1.0536,
+      "step": 3223
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.4112518065672341,
+      "learning_rate": 7.467577880231227e-05,
+      "loss": 1.0027,
+      "step": 3224
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.4654265480446376,
+      "learning_rate": 7.461804802198681e-05,
+      "loss": 1.1545,
+      "step": 3225
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.40907518627662653,
+      "learning_rate": 7.456032628217262e-05,
+      "loss": 1.088,
+      "step": 3226
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5403570274460063,
+      "learning_rate": 7.450261360342888e-05,
+      "loss": 1.237,
+      "step": 3227
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5073818203439244,
+      "learning_rate": 7.444491000631165e-05,
+      "loss": 1.0857,
+      "step": 3228
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.45721905342699165,
+      "learning_rate": 7.438721551137367e-05,
+      "loss": 1.0039,
+      "step": 3229
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.45670303937503753,
+      "learning_rate": 7.432953013916457e-05,
+      "loss": 1.0376,
+      "step": 3230
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.4567149618979133,
+      "learning_rate": 7.427185391023063e-05,
+      "loss": 1.0716,
+      "step": 3231
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.4651467548510654,
+      "learning_rate": 7.421418684511486e-05,
+      "loss": 1.2402,
+      "step": 3232
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.4147711823656278,
+      "learning_rate": 7.41565289643571e-05,
+      "loss": 0.9797,
+      "step": 3233
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.43329586725632574,
+      "learning_rate": 7.409888028849377e-05,
+      "loss": 1.3144,
+      "step": 3234
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.3854928180840778,
+      "learning_rate": 7.404124083805819e-05,
+      "loss": 1.1012,
+      "step": 3235
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.4427725668736682,
+      "learning_rate": 7.398361063358025e-05,
+      "loss": 1.218,
+      "step": 3236
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.40432880119993,
+      "learning_rate": 7.392598969558667e-05,
+      "loss": 1.1513,
+      "step": 3237
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.4195163926511318,
+      "learning_rate": 7.38683780446007e-05,
+      "loss": 1.1899,
+      "step": 3238
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.4562827715181376,
+      "learning_rate": 7.381077570114249e-05,
+      "loss": 0.9675,
+      "step": 3239
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.41455239734156263,
+      "learning_rate": 7.375318268572871e-05,
+      "loss": 1.0745,
+      "step": 3240
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.44682995694629013,
+      "learning_rate": 7.369559901887277e-05,
+      "loss": 1.1512,
+      "step": 3241
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.39458985383545037,
+      "learning_rate": 7.363802472108477e-05,
+      "loss": 1.0688,
+      "step": 3242
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5024801493798046,
+      "learning_rate": 7.358045981287141e-05,
+      "loss": 1.2409,
+      "step": 3243
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.4158450236422402,
+      "learning_rate": 7.352290431473612e-05,
+      "loss": 1.1986,
+      "step": 3244
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.4381635089150221,
+      "learning_rate": 7.346535824717893e-05,
+      "loss": 1.103,
+      "step": 3245
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.44404526461010824,
+      "learning_rate": 7.340782163069655e-05,
+      "loss": 1.0508,
+      "step": 3246
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.38252017404158883,
+      "learning_rate": 7.335029448578222e-05,
+      "loss": 1.3408,
+      "step": 3247
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5340654531662435,
+      "learning_rate": 7.329277683292599e-05,
+      "loss": 1.0066,
+      "step": 3248
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.4324339832724528,
+      "learning_rate": 7.323526869261437e-05,
+      "loss": 0.8834,
+      "step": 3249
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.4390163479092811,
+      "learning_rate": 7.317777008533054e-05,
+      "loss": 1.169,
+      "step": 3250
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.4549063467093044,
+      "learning_rate": 7.312028103155426e-05,
+      "loss": 0.9285,
+      "step": 3251
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.463408351919376,
+      "learning_rate": 7.306280155176198e-05,
+      "loss": 1.201,
+      "step": 3252
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.38740958952050014,
+      "learning_rate": 7.30053316664266e-05,
+      "loss": 1.1085,
+      "step": 3253
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.4061988379584434,
+      "learning_rate": 7.294787139601764e-05,
+      "loss": 1.1366,
+      "step": 3254
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.44776070793082046,
+      "learning_rate": 7.289042076100132e-05,
+      "loss": 1.2494,
+      "step": 3255
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.43857388641368406,
+      "learning_rate": 7.283297978184027e-05,
+      "loss": 1.2047,
+      "step": 3256
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.41619526441679433,
+      "learning_rate": 7.277554847899378e-05,
+      "loss": 1.1463,
+      "step": 3257
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.533936888160872,
+      "learning_rate": 7.271812687291759e-05,
+      "loss": 1.1467,
+      "step": 3258
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.45116009588664346,
+      "learning_rate": 7.266071498406417e-05,
+      "loss": 1.0544,
+      "step": 3259
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.47887730018154284,
+      "learning_rate": 7.26033128328823e-05,
+      "loss": 1.1726,
+      "step": 3260
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5124983929621847,
+      "learning_rate": 7.254592043981748e-05,
+      "loss": 1.1162,
+      "step": 3261
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.44399344207044045,
+      "learning_rate": 7.248853782531164e-05,
+      "loss": 1.2041,
+      "step": 3262
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.437423738717388,
+      "learning_rate": 7.243116500980318e-05,
+      "loss": 1.0294,
+      "step": 3263
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.47013912558299925,
+      "learning_rate": 7.237380201372719e-05,
+      "loss": 1.0639,
+      "step": 3264
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.4063748554512925,
+      "learning_rate": 7.231644885751507e-05,
+      "loss": 1.0771,
+      "step": 3265
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.43009143378653625,
+      "learning_rate": 7.225910556159483e-05,
+      "loss": 1.2035,
+      "step": 3266
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.3898877605359894,
+      "learning_rate": 7.220177214639088e-05,
+      "loss": 1.1434,
+      "step": 3267
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.450121740139184,
+      "learning_rate": 7.214444863232425e-05,
+      "loss": 1.1783,
+      "step": 3268
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5003169044725179,
+      "learning_rate": 7.208713503981228e-05,
+      "loss": 1.012,
+      "step": 3269
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.41558454679823637,
+      "learning_rate": 7.202983138926888e-05,
+      "loss": 1.1129,
+      "step": 3270
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.3954570571960559,
+      "learning_rate": 7.197253770110438e-05,
+      "loss": 1.0888,
+      "step": 3271
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.4007210613614983,
+      "learning_rate": 7.191525399572561e-05,
+      "loss": 1.1295,
+      "step": 3272
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.41291735684271824,
+      "learning_rate": 7.185798029353579e-05,
+      "loss": 1.2503,
+      "step": 3273
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.36624467388578585,
+      "learning_rate": 7.180071661493454e-05,
+      "loss": 1.0462,
+      "step": 3274
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.4764314721063882,
+      "learning_rate": 7.174346298031804e-05,
+      "loss": 0.9972,
+      "step": 3275
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.4397037343937754,
+      "learning_rate": 7.168621941007873e-05,
+      "loss": 1.2401,
+      "step": 3276
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.4429419235321856,
+      "learning_rate": 7.16289859246057e-05,
+      "loss": 1.2513,
+      "step": 3277
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.44274700234202513,
+      "learning_rate": 7.157176254428411e-05,
+      "loss": 1.1359,
+      "step": 3278
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.4228756174089166,
+      "learning_rate": 7.151454928949585e-05,
+      "loss": 1.2806,
+      "step": 3279
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.43546735414726234,
+      "learning_rate": 7.145734618061902e-05,
+      "loss": 1.2346,
+      "step": 3280
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.4078530975309651,
+      "learning_rate": 7.140015323802816e-05,
+      "loss": 1.2187,
+      "step": 3281
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.39318162024042136,
+      "learning_rate": 7.134297048209412e-05,
+      "loss": 1.1155,
+      "step": 3282
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.42999438408964313,
+      "learning_rate": 7.128579793318428e-05,
+      "loss": 1.0723,
+      "step": 3283
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.4221165212832297,
+      "learning_rate": 7.122863561166224e-05,
+      "loss": 1.1938,
+      "step": 3284
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.44879708568807075,
+      "learning_rate": 7.117148353788797e-05,
+      "loss": 1.0367,
+      "step": 3285
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.4182901091669516,
+      "learning_rate": 7.111434173221786e-05,
+      "loss": 1.1556,
+      "step": 3286
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.3839123435320263,
+      "learning_rate": 7.105721021500458e-05,
+      "loss": 1.0647,
+      "step": 3287
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.47864212047946,
+      "learning_rate": 7.100008900659721e-05,
+      "loss": 0.9319,
+      "step": 3288
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.44560986801175784,
+      "learning_rate": 7.094297812734103e-05,
+      "loss": 1.0967,
+      "step": 3289
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.4836445582910997,
+      "learning_rate": 7.088587759757782e-05,
+      "loss": 1.1395,
+      "step": 3290
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.3817391947341694,
+      "learning_rate": 7.082878743764545e-05,
+      "loss": 1.0844,
+      "step": 3291
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.3983229371640626,
+      "learning_rate": 7.077170766787835e-05,
+      "loss": 1.3514,
+      "step": 3292
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.40906559152437977,
+      "learning_rate": 7.071463830860703e-05,
+      "loss": 1.349,
+      "step": 3293
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.41599787511285385,
+      "learning_rate": 7.065757938015843e-05,
+      "loss": 1.1433,
+      "step": 3294
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.4673180958707515,
+      "learning_rate": 7.060053090285572e-05,
+      "loss": 1.194,
+      "step": 3295
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.43472845162273677,
+      "learning_rate": 7.054349289701828e-05,
+      "loss": 0.8294,
+      "step": 3296
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.44659768077223094,
+      "learning_rate": 7.048646538296195e-05,
+      "loss": 1.2001,
+      "step": 3297
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.43135016142316823,
+      "learning_rate": 7.042944838099863e-05,
+      "loss": 1.166,
+      "step": 3298
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.4467110546144613,
+      "learning_rate": 7.037244191143661e-05,
+      "loss": 1.1389,
+      "step": 3299
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5360355730250492,
+      "learning_rate": 7.031544599458036e-05,
+      "loss": 1.1779,
+      "step": 3300
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.4963274231487852,
+      "learning_rate": 7.025846065073064e-05,
+      "loss": 1.1324,
+      "step": 3301
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.4741417043728474,
+      "learning_rate": 7.020148590018435e-05,
+      "loss": 1.1334,
+      "step": 3302
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.4542105603866223,
+      "learning_rate": 7.014452176323477e-05,
+      "loss": 1.2795,
+      "step": 3303
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.4351698860243642,
+      "learning_rate": 7.008756826017124e-05,
+      "loss": 1.1168,
+      "step": 3304
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.4037806018735641,
+      "learning_rate": 7.003062541127946e-05,
+      "loss": 1.0922,
+      "step": 3305
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.42288932204271346,
+      "learning_rate": 6.997369323684123e-05,
+      "loss": 1.1823,
+      "step": 3306
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.43031188132155396,
+      "learning_rate": 6.991677175713449e-05,
+      "loss": 1.1499,
+      "step": 3307
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.426063369865606,
+      "learning_rate": 6.98598609924336e-05,
+      "loss": 1.1427,
+      "step": 3308
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.4718154775468632,
+      "learning_rate": 6.98029609630089e-05,
+      "loss": 0.9094,
+      "step": 3309
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.39940698115888296,
+      "learning_rate": 6.974607168912696e-05,
+      "loss": 1.0086,
+      "step": 3310
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.4144499375326889,
+      "learning_rate": 6.968919319105052e-05,
+      "loss": 1.0218,
+      "step": 3311
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.43501280832708916,
+      "learning_rate": 6.963232548903853e-05,
+      "loss": 1.1157,
+      "step": 3312
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.45425647830291904,
+      "learning_rate": 6.957546860334604e-05,
+      "loss": 1.1757,
+      "step": 3313
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.4766094441865091,
+      "learning_rate": 6.951862255422427e-05,
+      "loss": 1.2065,
+      "step": 3314
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5405723919595059,
+      "learning_rate": 6.946178736192053e-05,
+      "loss": 1.0686,
+      "step": 3315
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5096127575864485,
+      "learning_rate": 6.940496304667839e-05,
+      "loss": 1.1048,
+      "step": 3316
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.44761485271087187,
+      "learning_rate": 6.934814962873739e-05,
+      "loss": 1.1121,
+      "step": 3317
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.45237169044802533,
+      "learning_rate": 6.929134712833328e-05,
+      "loss": 1.101,
+      "step": 3318
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.41756261797224364,
+      "learning_rate": 6.92345555656979e-05,
+      "loss": 1.3052,
+      "step": 3319
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.490830378338134,
+      "learning_rate": 6.917777496105917e-05,
+      "loss": 1.1706,
+      "step": 3320
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.44819047936599843,
+      "learning_rate": 6.912100533464121e-05,
+      "loss": 1.0624,
+      "step": 3321
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.4890011907417924,
+      "learning_rate": 6.906424670666409e-05,
+      "loss": 1.0703,
+      "step": 3322
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.445860156422466,
+      "learning_rate": 6.900749909734406e-05,
+      "loss": 1.1251,
+      "step": 3323
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.43546028649959756,
+      "learning_rate": 6.895076252689335e-05,
+      "loss": 1.1281,
+      "step": 3324
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5361230378793229,
+      "learning_rate": 6.889403701552039e-05,
+      "loss": 1.0079,
+      "step": 3325
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.3913868950622121,
+      "learning_rate": 6.88373225834296e-05,
+      "loss": 1.2033,
+      "step": 3326
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.40718579381218395,
+      "learning_rate": 6.878061925082137e-05,
+      "loss": 1.1119,
+      "step": 3327
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.4452429742119691,
+      "learning_rate": 6.872392703789232e-05,
+      "loss": 1.0952,
+      "step": 3328
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.4194703311016238,
+      "learning_rate": 6.866724596483495e-05,
+      "loss": 1.0387,
+      "step": 3329
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.4570701420261807,
+      "learning_rate": 6.861057605183787e-05,
+      "loss": 1.167,
+      "step": 3330
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.42309853164134614,
+      "learning_rate": 6.855391731908567e-05,
+      "loss": 1.2158,
+      "step": 3331
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.4516507247124152,
+      "learning_rate": 6.849726978675906e-05,
+      "loss": 1.1359,
+      "step": 3332
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5131475149970341,
+      "learning_rate": 6.84406334750346e-05,
+      "loss": 1.0326,
+      "step": 3333
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.418458498170454,
+      "learning_rate": 6.838400840408499e-05,
+      "loss": 1.1815,
+      "step": 3334
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.44576693763713754,
+      "learning_rate": 6.832739459407885e-05,
+      "loss": 1.0021,
+      "step": 3335
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.44489938040417787,
+      "learning_rate": 6.827079206518083e-05,
+      "loss": 1.1128,
+      "step": 3336
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.4119660761172966,
+      "learning_rate": 6.821420083755157e-05,
+      "loss": 1.228,
+      "step": 3337
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.41937618080955746,
+      "learning_rate": 6.81576209313476e-05,
+      "loss": 1.0813,
+      "step": 3338
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.4254318846497289,
+      "learning_rate": 6.810105236672155e-05,
+      "loss": 1.204,
+      "step": 3339
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.43503637046249,
+      "learning_rate": 6.804449516382183e-05,
+      "loss": 1.1215,
+      "step": 3340
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.4344950002484197,
+      "learning_rate": 6.798794934279304e-05,
+      "loss": 1.1251,
+      "step": 3341
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5193278972090825,
+      "learning_rate": 6.793141492377552e-05,
+      "loss": 1.0708,
+      "step": 3342
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.41940767776962357,
+      "learning_rate": 6.787489192690569e-05,
+      "loss": 1.0053,
+      "step": 3343
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.4512437132192107,
+      "learning_rate": 6.781838037231571e-05,
+      "loss": 1.1538,
+      "step": 3344
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.399024819085027,
+      "learning_rate": 6.776188028013396e-05,
+      "loss": 1.0587,
+      "step": 3345
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.46794955128972754,
+      "learning_rate": 6.770539167048446e-05,
+      "loss": 1.1084,
+      "step": 3346
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5414415904859735,
+      "learning_rate": 6.764891456348729e-05,
+      "loss": 1.0991,
+      "step": 3347
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.37380456252439026,
+      "learning_rate": 6.759244897925841e-05,
+      "loss": 1.1689,
+      "step": 3348
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.38506436469207306,
+      "learning_rate": 6.753599493790959e-05,
+      "loss": 1.1016,
+      "step": 3349
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.4603559469136704,
+      "learning_rate": 6.747955245954862e-05,
+      "loss": 1.2295,
+      "step": 3350
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.41529394127720826,
+      "learning_rate": 6.74231215642791e-05,
+      "loss": 1.1312,
+      "step": 3351
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.43195483907520194,
+      "learning_rate": 6.736670227220052e-05,
+      "loss": 0.9997,
+      "step": 3352
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.38774791739301684,
+      "learning_rate": 6.73102946034082e-05,
+      "loss": 1.1127,
+      "step": 3353
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.43418673298003496,
+      "learning_rate": 6.725389857799337e-05,
+      "loss": 1.0462,
+      "step": 3354
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.44992168222185214,
+      "learning_rate": 6.719751421604309e-05,
+      "loss": 1.1794,
+      "step": 3355
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.39771566267514985,
+      "learning_rate": 6.714114153764029e-05,
+      "loss": 1.2313,
+      "step": 3356
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.4301930977770521,
+      "learning_rate": 6.708478056286369e-05,
+      "loss": 1.2294,
+      "step": 3357
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.42207721482914085,
+      "learning_rate": 6.70284313117879e-05,
+      "loss": 1.1648,
+      "step": 3358
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.3897726151721619,
+      "learning_rate": 6.697209380448333e-05,
+      "loss": 1.1227,
+      "step": 3359
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5001031681676907,
+      "learning_rate": 6.691576806101612e-05,
+      "loss": 1.056,
+      "step": 3360
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.4240743463530688,
+      "learning_rate": 6.68594541014484e-05,
+      "loss": 1.1936,
+      "step": 3361
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.4622982133499847,
+      "learning_rate": 6.680315194583794e-05,
+      "loss": 1.0644,
+      "step": 3362
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5528811224563405,
+      "learning_rate": 6.674686161423843e-05,
+      "loss": 1.2276,
+      "step": 3363
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.4250270070009707,
+      "learning_rate": 6.669058312669919e-05,
+      "loss": 1.0938,
+      "step": 3364
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.42181666769437753,
+      "learning_rate": 6.663431650326553e-05,
+      "loss": 1.0405,
+      "step": 3365
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.4299133331245127,
+      "learning_rate": 6.657806176397836e-05,
+      "loss": 0.9962,
+      "step": 3366
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.41084155699175057,
+      "learning_rate": 6.652181892887446e-05,
+      "loss": 1.1518,
+      "step": 3367
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.4911332705175731,
+      "learning_rate": 6.646558801798627e-05,
+      "loss": 1.138,
+      "step": 3368
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.3763384953727361,
+      "learning_rate": 6.640936905134212e-05,
+      "loss": 1.1628,
+      "step": 3369
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.43490895955705056,
+      "learning_rate": 6.635316204896598e-05,
+      "loss": 1.1989,
+      "step": 3370
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.43922090864553026,
+      "learning_rate": 6.629696703087754e-05,
+      "loss": 1.0147,
+      "step": 3371
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.44139166551306414,
+      "learning_rate": 6.624078401709235e-05,
+      "loss": 1.3404,
+      "step": 3372
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.48157901064635317,
+      "learning_rate": 6.61846130276215e-05,
+      "loss": 1.1381,
+      "step": 3373
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.486655302335105,
+      "learning_rate": 6.612845408247204e-05,
+      "loss": 1.0744,
+      "step": 3374
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.3777145993127069,
+      "learning_rate": 6.607230720164647e-05,
+      "loss": 1.1322,
+      "step": 3375
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.4468838804869244,
+      "learning_rate": 6.60161724051432e-05,
+      "loss": 1.3023,
+      "step": 3376
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.4340650930378557,
+      "learning_rate": 6.596004971295615e-05,
+      "loss": 1.0482,
+      "step": 3377
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.46164907701091507,
+      "learning_rate": 6.590393914507517e-05,
+      "loss": 1.0681,
+      "step": 3378
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5070420594183342,
+      "learning_rate": 6.584784072148555e-05,
+      "loss": 1.1512,
+      "step": 3379
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.41216935724741255,
+      "learning_rate": 6.579175446216833e-05,
+      "loss": 1.3629,
+      "step": 3380
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.48488181738142627,
+      "learning_rate": 6.573568038710032e-05,
+      "loss": 1.1712,
+      "step": 3381
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.4429903203710049,
+      "learning_rate": 6.567961851625386e-05,
+      "loss": 1.2102,
+      "step": 3382
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.41123570175918095,
+      "learning_rate": 6.562356886959704e-05,
+      "loss": 1.1173,
+      "step": 3383
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.3952353243635426,
+      "learning_rate": 6.556753146709348e-05,
+      "loss": 1.3018,
+      "step": 3384
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.4148420032954845,
+      "learning_rate": 6.55115063287026e-05,
+      "loss": 1.2076,
+      "step": 3385
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.43280877112501176,
+      "learning_rate": 6.545549347437931e-05,
+      "loss": 1.2292,
+      "step": 3386
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.4505808221085309,
+      "learning_rate": 6.539949292407421e-05,
+      "loss": 1.256,
+      "step": 3387
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.38031010625362915,
+      "learning_rate": 6.534350469773348e-05,
+      "loss": 1.1534,
+      "step": 3388
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.44719806952625063,
+      "learning_rate": 6.528752881529898e-05,
+      "loss": 1.1052,
+      "step": 3389
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.4593030015889371,
+      "learning_rate": 6.523156529670811e-05,
+      "loss": 0.9861,
+      "step": 3390
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.49051658381382807,
+      "learning_rate": 6.517561416189387e-05,
+      "loss": 1.0712,
+      "step": 3391
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.4575998308032589,
+      "learning_rate": 6.511967543078488e-05,
+      "loss": 1.0249,
+      "step": 3392
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.4758694454242847,
+      "learning_rate": 6.506374912330527e-05,
+      "loss": 1.0725,
+      "step": 3393
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.4646700742299722,
+      "learning_rate": 6.50078352593749e-05,
+      "loss": 1.1528,
+      "step": 3394
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.45850592372059357,
+      "learning_rate": 6.495193385890901e-05,
+      "loss": 1.1448,
+      "step": 3395
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5010476246988357,
+      "learning_rate": 6.489604494181857e-05,
+      "loss": 1.0646,
+      "step": 3396
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.4643885487039272,
+      "learning_rate": 6.484016852800993e-05,
+      "loss": 1.0668,
+      "step": 3397
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.4632120572271518,
+      "learning_rate": 6.478430463738516e-05,
+      "loss": 0.9748,
+      "step": 3398
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.4300950888736018,
+      "learning_rate": 6.472845328984175e-05,
+      "loss": 1.1748,
+      "step": 3399
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.46152794312802176,
+      "learning_rate": 6.467261450527279e-05,
+      "loss": 1.0539,
+      "step": 3400
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.4285279908399653,
+      "learning_rate": 6.461678830356684e-05,
+      "loss": 1.083,
+      "step": 3401
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.4207046045893617,
+      "learning_rate": 6.456097470460798e-05,
+      "loss": 1.1505,
+      "step": 3402
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.40945713402262635,
+      "learning_rate": 6.450517372827591e-05,
+      "loss": 1.1324,
+      "step": 3403
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.44811354413895316,
+      "learning_rate": 6.444938539444565e-05,
+      "loss": 1.042,
+      "step": 3404
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.4331460164998924,
+      "learning_rate": 6.43936097229879e-05,
+      "loss": 1.1556,
+      "step": 3405
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.4969857874644648,
+      "learning_rate": 6.43378467337687e-05,
+      "loss": 1.1102,
+      "step": 3406
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.4048673115021053,
+      "learning_rate": 6.42820964466497e-05,
+      "loss": 1.2286,
+      "step": 3407
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.40841882764245035,
+      "learning_rate": 6.422635888148791e-05,
+      "loss": 1.1823,
+      "step": 3408
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.4778699205077795,
+      "learning_rate": 6.417063405813594e-05,
+      "loss": 1.0986,
+      "step": 3409
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.46103948017765733,
+      "learning_rate": 6.411492199644171e-05,
+      "loss": 1.2039,
+      "step": 3410
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.44639199543485303,
+      "learning_rate": 6.405922271624874e-05,
+      "loss": 1.1395,
+      "step": 3411
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.43926501234454085,
+      "learning_rate": 6.400353623739587e-05,
+      "loss": 1.0544,
+      "step": 3412
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.4173209739101363,
+      "learning_rate": 6.394786257971743e-05,
+      "loss": 1.2208,
+      "step": 3413
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.4747052424976873,
+      "learning_rate": 6.389220176304325e-05,
+      "loss": 0.9904,
+      "step": 3414
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.4406770102558871,
+      "learning_rate": 6.383655380719848e-05,
+      "loss": 1.1278,
+      "step": 3415
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.44062479687799433,
+      "learning_rate": 6.378091873200379e-05,
+      "loss": 1.1838,
+      "step": 3416
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.45237033864639176,
+      "learning_rate": 6.372529655727513e-05,
+      "loss": 1.0071,
+      "step": 3417
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.4520222166303622,
+      "learning_rate": 6.366968730282403e-05,
+      "loss": 1.051,
+      "step": 3418
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.47808974632950374,
+      "learning_rate": 6.361409098845725e-05,
+      "loss": 1.0737,
+      "step": 3419
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.4198184560260746,
+      "learning_rate": 6.355850763397707e-05,
+      "loss": 1.1255,
+      "step": 3420
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.41571728231490396,
+      "learning_rate": 6.350293725918102e-05,
+      "loss": 1.1368,
+      "step": 3421
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.4913019948187437,
+      "learning_rate": 6.344737988386215e-05,
+      "loss": 1.1281,
+      "step": 3422
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.45048102092137754,
+      "learning_rate": 6.339183552780884e-05,
+      "loss": 1.1162,
+      "step": 3423
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.4163649991487393,
+      "learning_rate": 6.333630421080472e-05,
+      "loss": 1.1698,
+      "step": 3424
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.40543507676678375,
+      "learning_rate": 6.328078595262894e-05,
+      "loss": 1.1496,
+      "step": 3425
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5100992852112777,
+      "learning_rate": 6.322528077305585e-05,
+      "loss": 1.0517,
+      "step": 3426
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.4776348890398004,
+      "learning_rate": 6.316978869185532e-05,
+      "loss": 1.069,
+      "step": 3427
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.4413954862225715,
+      "learning_rate": 6.311430972879232e-05,
+      "loss": 1.0824,
+      "step": 3428
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.39929238500834907,
+      "learning_rate": 6.305884390362738e-05,
+      "loss": 1.2414,
+      "step": 3429
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.43251932381815883,
+      "learning_rate": 6.300339123611618e-05,
+      "loss": 1.0931,
+      "step": 3430
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.434910975771727,
+      "learning_rate": 6.294795174600985e-05,
+      "loss": 0.9684,
+      "step": 3431
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.4977569665111876,
+      "learning_rate": 6.28925254530547e-05,
+      "loss": 1.0765,
+      "step": 3432
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5084067055209829,
+      "learning_rate": 6.283711237699236e-05,
+      "loss": 0.9711,
+      "step": 3433
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5107025842308474,
+      "learning_rate": 6.27817125375599e-05,
+      "loss": 0.9981,
+      "step": 3434
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.45763792802682146,
+      "learning_rate": 6.272632595448947e-05,
+      "loss": 1.233,
+      "step": 3435
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.4490912231218561,
+      "learning_rate": 6.267095264750866e-05,
+      "loss": 0.9698,
+      "step": 3436
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.43416977368359566,
+      "learning_rate": 6.261559263634017e-05,
+      "loss": 0.9837,
+      "step": 3437
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.4163085980870379,
+      "learning_rate": 6.256024594070217e-05,
+      "loss": 1.1667,
+      "step": 3438
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.4316132238220486,
+      "learning_rate": 6.250491258030791e-05,
+      "loss": 1.1066,
+      "step": 3439
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.43914185160855956,
+      "learning_rate": 6.244959257486596e-05,
+      "loss": 1.1359,
+      "step": 3440
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.404036752771784,
+      "learning_rate": 6.239428594408013e-05,
+      "loss": 1.0694,
+      "step": 3441
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.39571358246339433,
+      "learning_rate": 6.23389927076495e-05,
+      "loss": 1.1073,
+      "step": 3442
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.4607056140479705,
+      "learning_rate": 6.22837128852683e-05,
+      "loss": 1.0817,
+      "step": 3443
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5504297995528895,
+      "learning_rate": 6.222844649662604e-05,
+      "loss": 1.0891,
+      "step": 3444
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.43430299714168236,
+      "learning_rate": 6.217319356140745e-05,
+      "loss": 1.1011,
+      "step": 3445
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.4684119551135448,
+      "learning_rate": 6.211795409929238e-05,
+      "loss": 1.0507,
+      "step": 3446
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.44774109964623393,
+      "learning_rate": 6.206272812995607e-05,
+      "loss": 1.0428,
+      "step": 3447
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.44846235095687786,
+      "learning_rate": 6.200751567306874e-05,
+      "loss": 1.1025,
+      "step": 3448
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.37979287317290256,
+      "learning_rate": 6.195231674829593e-05,
+      "loss": 1.161,
+      "step": 3449
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5816773320618406,
+      "learning_rate": 6.18971313752983e-05,
+      "loss": 1.161,
+      "step": 3450
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.42404922990443356,
+      "learning_rate": 6.184195957373176e-05,
+      "loss": 1.0676,
+      "step": 3451
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.49973223296610353,
+      "learning_rate": 6.178680136324727e-05,
+      "loss": 1.0873,
+      "step": 3452
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.38430704823474876,
+      "learning_rate": 6.173165676349103e-05,
+      "loss": 1.1776,
+      "step": 3453
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.40508546638699366,
+      "learning_rate": 6.167652579410439e-05,
+      "loss": 1.1512,
+      "step": 3454
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.44787766925449846,
+      "learning_rate": 6.162140847472381e-05,
+      "loss": 1.153,
+      "step": 3455
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.42389034408509,
+      "learning_rate": 6.156630482498092e-05,
+      "loss": 0.9871,
+      "step": 3456
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.4227205619861611,
+      "learning_rate": 6.151121486450243e-05,
+      "loss": 1.2214,
+      "step": 3457
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.4055720921817765,
+      "learning_rate": 6.145613861291026e-05,
+      "loss": 1.2067,
+      "step": 3458
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.44682914621779257,
+      "learning_rate": 6.140107608982136e-05,
+      "loss": 1.0443,
+      "step": 3459
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.4692831419342991,
+      "learning_rate": 6.134602731484788e-05,
+      "loss": 1.1397,
+      "step": 3460
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.4618130617264273,
+      "learning_rate": 6.129099230759695e-05,
+      "loss": 1.1033,
+      "step": 3461
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.4477695148611071,
+      "learning_rate": 6.123597108767095e-05,
+      "loss": 1.1463,
+      "step": 3462
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.46907110680540937,
+      "learning_rate": 6.118096367466721e-05,
+      "loss": 1.3251,
+      "step": 3463
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.401080756502236,
+      "learning_rate": 6.112597008817823e-05,
+      "loss": 1.0701,
+      "step": 3464
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.4449199981443137,
+      "learning_rate": 6.107099034779155e-05,
+      "loss": 1.1868,
+      "step": 3465
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5367221983429203,
+      "learning_rate": 6.1016024473089726e-05,
+      "loss": 1.0036,
+      "step": 3466
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.4085150170285111,
+      "learning_rate": 6.0961072483650526e-05,
+      "loss": 0.9542,
+      "step": 3467
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.46983781548129666,
+      "learning_rate": 6.090613439904662e-05,
+      "loss": 0.9831,
+      "step": 3468
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.43123202780123593,
+      "learning_rate": 6.0851210238845834e-05,
+      "loss": 1.0618,
+      "step": 3469
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.39777470237276585,
+      "learning_rate": 6.07963000226109e-05,
+      "loss": 1.1082,
+      "step": 3470
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.4767071739711476,
+      "learning_rate": 6.0741403769899765e-05,
+      "loss": 1.2108,
+      "step": 3471
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.4749789825849407,
+      "learning_rate": 6.0686521500265234e-05,
+      "loss": 1.0804,
+      "step": 3472
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.4494084072252133,
+      "learning_rate": 6.063165323325526e-05,
+      "loss": 1.0137,
+      "step": 3473
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.4314942967670764,
+      "learning_rate": 6.057679898841269e-05,
+      "loss": 0.9274,
+      "step": 3474
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.39247843193533233,
+      "learning_rate": 6.05219587852755e-05,
+      "loss": 1.1254,
+      "step": 3475
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.44230895298601514,
+      "learning_rate": 6.046713264337658e-05,
+      "loss": 1.1481,
+      "step": 3476
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.41044017835910235,
+      "learning_rate": 6.041232058224382e-05,
+      "loss": 1.1632,
+      "step": 3477
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.4392966528498281,
+      "learning_rate": 6.0357522621400134e-05,
+      "loss": 1.1906,
+      "step": 3478
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5073260361282818,
+      "learning_rate": 6.030273878036334e-05,
+      "loss": 1.2095,
+      "step": 3479
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.46309804540908844,
+      "learning_rate": 6.0247969078646384e-05,
+      "loss": 1.1171,
+      "step": 3480
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.4169002762205245,
+      "learning_rate": 6.019321353575692e-05,
+      "loss": 1.1533,
+      "step": 3481
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.4122469197100042,
+      "learning_rate": 6.013847217119784e-05,
+      "loss": 1.0207,
+      "step": 3482
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.4401970068890709,
+      "learning_rate": 6.008374500446676e-05,
+      "loss": 0.9976,
+      "step": 3483
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5879449550384784,
+      "learning_rate": 6.002903205505641e-05,
+      "loss": 1.01,
+      "step": 3484
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.3997072886523605,
+      "learning_rate": 5.9974333342454326e-05,
+      "loss": 1.2786,
+      "step": 3485
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.49230511979251845,
+      "learning_rate": 5.991964888614301e-05,
+      "loss": 1.0902,
+      "step": 3486
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.48908826203290456,
+      "learning_rate": 5.986497870559997e-05,
+      "loss": 1.1168,
+      "step": 3487
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.43135514068483033,
+      "learning_rate": 5.98103228202975e-05,
+      "loss": 1.1989,
+      "step": 3488
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.43155611843279773,
+      "learning_rate": 5.9755681249702896e-05,
+      "loss": 1.0949,
+      "step": 3489
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.4215140837650032,
+      "learning_rate": 5.970105401327828e-05,
+      "loss": 1.0223,
+      "step": 3490
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.45393632330921474,
+      "learning_rate": 5.964644113048079e-05,
+      "loss": 1.1642,
+      "step": 3491
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.4798948942701216,
+      "learning_rate": 5.9591842620762294e-05,
+      "loss": 1.0397,
+      "step": 3492
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.4361046415814585,
+      "learning_rate": 5.953725850356968e-05,
+      "loss": 1.0856,
+      "step": 3493
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.44266805438503126,
+      "learning_rate": 5.948268879834458e-05,
+      "loss": 1.095,
+      "step": 3494
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.4108395403174933,
+      "learning_rate": 5.9428133524523646e-05,
+      "loss": 1.1931,
+      "step": 3495
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.4536965263719057,
+      "learning_rate": 5.937359270153825e-05,
+      "loss": 1.0225,
+      "step": 3496
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.4665883541323011,
+      "learning_rate": 5.931906634881468e-05,
+      "loss": 0.9334,
+      "step": 3497
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.43383692179300226,
+      "learning_rate": 5.926455448577407e-05,
+      "loss": 1.1756,
+      "step": 3498
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.46724960950984434,
+      "learning_rate": 5.921005713183235e-05,
+      "loss": 1.1803,
+      "step": 3499
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5104819035268147,
+      "learning_rate": 5.9155574306400395e-05,
+      "loss": 1.0296,
+      "step": 3500
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.44244094668832523,
+      "learning_rate": 5.910110602888377e-05,
+      "loss": 1.2021,
+      "step": 3501
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.44795378098676675,
+      "learning_rate": 5.904665231868294e-05,
+      "loss": 1.1795,
+      "step": 3502
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.40198579663073325,
+      "learning_rate": 5.89922131951931e-05,
+      "loss": 0.8737,
+      "step": 3503
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5356473279910984,
+      "learning_rate": 5.893778867780444e-05,
+      "loss": 1.2116,
+      "step": 3504
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.4145989390344316,
+      "learning_rate": 5.8883378785901644e-05,
+      "loss": 1.208,
+      "step": 3505
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.4952555026107272,
+      "learning_rate": 5.88289835388645e-05,
+      "loss": 1.0721,
+      "step": 3506
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.44139633889094027,
+      "learning_rate": 5.877460295606738e-05,
+      "loss": 1.231,
+      "step": 3507
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.48097628917595014,
+      "learning_rate": 5.872023705687948e-05,
+      "loss": 1.0341,
+      "step": 3508
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.4958327718905778,
+      "learning_rate": 5.866588586066481e-05,
+      "loss": 0.9687,
+      "step": 3509
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.406400633864402,
+      "learning_rate": 5.861154938678204e-05,
+      "loss": 1.1543,
+      "step": 3510
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.4455547888562037,
+      "learning_rate": 5.8557227654584754e-05,
+      "loss": 1.0255,
+      "step": 3511
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.44417966136594483,
+      "learning_rate": 5.8502920683421134e-05,
+      "loss": 1.2001,
+      "step": 3512
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.46091623997742937,
+      "learning_rate": 5.8448628492634225e-05,
+      "loss": 1.1785,
+      "step": 3513
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.48591297794136756,
+      "learning_rate": 5.839435110156167e-05,
+      "loss": 1.0297,
+      "step": 3514
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.44072027591562324,
+      "learning_rate": 5.834008852953603e-05,
+      "loss": 1.1838,
+      "step": 3515
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.42941507368452425,
+      "learning_rate": 5.828584079588435e-05,
+      "loss": 1.2063,
+      "step": 3516
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.40923221572349694,
+      "learning_rate": 5.8231607919928636e-05,
+      "loss": 1.0173,
+      "step": 3517
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.39408732381065487,
+      "learning_rate": 5.817738992098543e-05,
+      "loss": 0.9619,
+      "step": 3518
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.43462972896259605,
+      "learning_rate": 5.812318681836597e-05,
+      "loss": 1.1832,
+      "step": 3519
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.4172208931737817,
+      "learning_rate": 5.8068998631376384e-05,
+      "loss": 1.1679,
+      "step": 3520
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.43824683059799263,
+      "learning_rate": 5.801482537931722e-05,
+      "loss": 1.1341,
+      "step": 3521
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.4218424800282871,
+      "learning_rate": 5.796066708148393e-05,
+      "loss": 1.1566,
+      "step": 3522
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.46185547778768915,
+      "learning_rate": 5.790652375716652e-05,
+      "loss": 1.2227,
+      "step": 3523
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.4475100002067,
+      "learning_rate": 5.785239542564968e-05,
+      "loss": 1.1997,
+      "step": 3524
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.41803771677102664,
+      "learning_rate": 5.779828210621273e-05,
+      "loss": 1.0665,
+      "step": 3525
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.46254242241697,
+      "learning_rate": 5.774418381812977e-05,
+      "loss": 0.949,
+      "step": 3526
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.41424165203221724,
+      "learning_rate": 5.769010058066937e-05,
+      "loss": 1.1374,
+      "step": 3527
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.4224916930564516,
+      "learning_rate": 5.763603241309492e-05,
+      "loss": 1.1411,
+      "step": 3528
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.435714893698284,
+      "learning_rate": 5.758197933466429e-05,
+      "loss": 1.155,
+      "step": 3529
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.44807267234809905,
+      "learning_rate": 5.7527941364630066e-05,
+      "loss": 0.9806,
+      "step": 3530
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.4804393982013838,
+      "learning_rate": 5.74739185222394e-05,
+      "loss": 1.1982,
+      "step": 3531
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.4161475415142702,
+      "learning_rate": 5.7419910826734056e-05,
+      "loss": 1.2637,
+      "step": 3532
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.4171757563303151,
+      "learning_rate": 5.7365918297350484e-05,
+      "loss": 1.1711,
+      "step": 3533
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.4286355806703411,
+      "learning_rate": 5.731194095331961e-05,
+      "loss": 1.1939,
+      "step": 3534
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.47623930104286955,
+      "learning_rate": 5.7257978813867094e-05,
+      "loss": 1.1832,
+      "step": 3535
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.39095134454625263,
+      "learning_rate": 5.720403189821302e-05,
+      "loss": 1.3388,
+      "step": 3536
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.42436115399722196,
+      "learning_rate": 5.71501002255722e-05,
+      "loss": 1.0297,
+      "step": 3537
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.45048754398847873,
+      "learning_rate": 5.709618381515394e-05,
+      "loss": 1.0831,
+      "step": 3538
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.4213692356539341,
+      "learning_rate": 5.704228268616208e-05,
+      "loss": 1.1018,
+      "step": 3539
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.42207875510406145,
+      "learning_rate": 5.698839685779507e-05,
+      "loss": 1.0387,
+      "step": 3540
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.4973637209722719,
+      "learning_rate": 5.693452634924587e-05,
+      "loss": 1.1159,
+      "step": 3541
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.4098597643579094,
+      "learning_rate": 5.688067117970207e-05,
+      "loss": 1.1926,
+      "step": 3542
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.40406660922461407,
+      "learning_rate": 5.682683136834567e-05,
+      "loss": 1.1491,
+      "step": 3543
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.39502279794277906,
+      "learning_rate": 5.6773006934353326e-05,
+      "loss": 1.2474,
+      "step": 3544
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.4741981777824592,
+      "learning_rate": 5.6719197896896106e-05,
+      "loss": 1.3641,
+      "step": 3545
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.39104219108966276,
+      "learning_rate": 5.666540427513969e-05,
+      "loss": 1.1519,
+      "step": 3546
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.4571282877139187,
+      "learning_rate": 5.6611626088244194e-05,
+      "loss": 0.9874,
+      "step": 3547
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.4933945791846753,
+      "learning_rate": 5.6557863355364286e-05,
+      "loss": 1.0208,
+      "step": 3548
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.4076983908491341,
+      "learning_rate": 5.650411609564907e-05,
+      "loss": 1.2204,
+      "step": 3549
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.484400680451012,
+      "learning_rate": 5.645038432824217e-05,
+      "loss": 1.121,
+      "step": 3550
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.3888537663872469,
+      "learning_rate": 5.639666807228175e-05,
+      "loss": 1.1367,
+      "step": 3551
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.4254946912281158,
+      "learning_rate": 5.634296734690035e-05,
+      "loss": 1.0479,
+      "step": 3552
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.40282683130547925,
+      "learning_rate": 5.628928217122507e-05,
+      "loss": 1.1609,
+      "step": 3553
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.4251233664471547,
+      "learning_rate": 5.623561256437736e-05,
+      "loss": 1.1255,
+      "step": 3554
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.4119110709042794,
+      "learning_rate": 5.6181958545473325e-05,
+      "loss": 1.1891,
+      "step": 3555
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.43547744729071486,
+      "learning_rate": 5.612832013362321e-05,
+      "loss": 1.196,
+      "step": 3556
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.4847744575358152,
+      "learning_rate": 5.6074697347932e-05,
+      "loss": 1.1981,
+      "step": 3557
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.4099880794668844,
+      "learning_rate": 5.6021090207498926e-05,
+      "loss": 1.1561,
+      "step": 3558
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.48482412013013493,
+      "learning_rate": 5.5967498731417765e-05,
+      "loss": 1.2037,
+      "step": 3559
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.48511785821170444,
+      "learning_rate": 5.5913922938776645e-05,
+      "loss": 1.0736,
+      "step": 3560
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.45835995315070055,
+      "learning_rate": 5.586036284865805e-05,
+      "loss": 1.1269,
+      "step": 3561
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.4423852537116012,
+      "learning_rate": 5.580681848013906e-05,
+      "loss": 1.1282,
+      "step": 3562
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.362673432735491,
+      "learning_rate": 5.575328985229098e-05,
+      "loss": 1.2869,
+      "step": 3563
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.43472762202770027,
+      "learning_rate": 5.569977698417956e-05,
+      "loss": 1.0317,
+      "step": 3564
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5047062938410539,
+      "learning_rate": 5.564627989486494e-05,
+      "loss": 1.1417,
+      "step": 3565
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.4271651524169534,
+      "learning_rate": 5.5592798603401696e-05,
+      "loss": 1.2868,
+      "step": 3566
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.43065975363504366,
+      "learning_rate": 5.5539333128838653e-05,
+      "loss": 1.0555,
+      "step": 3567
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.372790997537263,
+      "learning_rate": 5.548588349021915e-05,
+      "loss": 1.1138,
+      "step": 3568
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.4454199087446004,
+      "learning_rate": 5.543244970658074e-05,
+      "loss": 0.9429,
+      "step": 3569
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.38527491758433696,
+      "learning_rate": 5.537903179695547e-05,
+      "loss": 1.101,
+      "step": 3570
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.4629055240909443,
+      "learning_rate": 5.5325629780369635e-05,
+      "loss": 0.9808,
+      "step": 3571
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.4306748168043986,
+      "learning_rate": 5.5272243675843896e-05,
+      "loss": 1.2024,
+      "step": 3572
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.3963620407634554,
+      "learning_rate": 5.521887350239323e-05,
+      "loss": 1.0824,
+      "step": 3573
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.47377041621793414,
+      "learning_rate": 5.5165519279026935e-05,
+      "loss": 1.0934,
+      "step": 3574
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6258780106525346,
+      "learning_rate": 5.511218102474872e-05,
+      "loss": 1.2682,
+      "step": 3575
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.42041787155991045,
+      "learning_rate": 5.505885875855647e-05,
+      "loss": 1.1591,
+      "step": 3576
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.4231877425076663,
+      "learning_rate": 5.5005552499442504e-05,
+      "loss": 1.2222,
+      "step": 3577
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.40932147890678106,
+      "learning_rate": 5.495226226639331e-05,
+      "loss": 1.0847,
+      "step": 3578
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.40839704690424866,
+      "learning_rate": 5.48989880783898e-05,
+      "loss": 1.2634,
+      "step": 3579
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.40541728475040656,
+      "learning_rate": 5.484572995440709e-05,
+      "loss": 1.0906,
+      "step": 3580
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.4118709523532234,
+      "learning_rate": 5.479248791341458e-05,
+      "loss": 1.2165,
+      "step": 3581
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.4388142637692349,
+      "learning_rate": 5.473926197437594e-05,
+      "loss": 1.022,
+      "step": 3582
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.4245960348186348,
+      "learning_rate": 5.468605215624909e-05,
+      "loss": 0.9876,
+      "step": 3583
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.47914335565948885,
+      "learning_rate": 5.463285847798632e-05,
+      "loss": 1.1027,
+      "step": 3584
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.4260095292580753,
+      "learning_rate": 5.4579680958533986e-05,
+      "loss": 1.1128,
+      "step": 3585
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.40117809479135463,
+      "learning_rate": 5.4526519616832885e-05,
+      "loss": 1.1479,
+      "step": 3586
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.3978319577754821,
+      "learning_rate": 5.4473374471817906e-05,
+      "loss": 1.0893,
+      "step": 3587
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.48477724183251325,
+      "learning_rate": 5.442024554241822e-05,
+      "loss": 0.9069,
+      "step": 3588
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.4108024600789549,
+      "learning_rate": 5.436713284755721e-05,
+      "loss": 1.1785,
+      "step": 3589
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.4061158437155526,
+      "learning_rate": 5.431403640615252e-05,
+      "loss": 1.138,
+      "step": 3590
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.43934520853170494,
+      "learning_rate": 5.4260956237115965e-05,
+      "loss": 1.0149,
+      "step": 3591
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.4169918720882398,
+      "learning_rate": 5.420789235935353e-05,
+      "loss": 1.0679,
+      "step": 3592
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.4303459515052516,
+      "learning_rate": 5.415484479176549e-05,
+      "loss": 1.0794,
+      "step": 3593
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.4137077725850348,
+      "learning_rate": 5.410181355324622e-05,
+      "loss": 1.1862,
+      "step": 3594
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.4694788499058781,
+      "learning_rate": 5.4048798662684376e-05,
+      "loss": 1.1739,
+      "step": 3595
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.41853706306574356,
+      "learning_rate": 5.39958001389627e-05,
+      "loss": 1.0748,
+      "step": 3596
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.4343849169582247,
+      "learning_rate": 5.394281800095813e-05,
+      "loss": 1.1231,
+      "step": 3597
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.3935493524830143,
+      "learning_rate": 5.388985226754175e-05,
+      "loss": 1.0735,
+      "step": 3598
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.44290519810603624,
+      "learning_rate": 5.3836902957578904e-05,
+      "loss": 1.1916,
+      "step": 3599
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.44668352004533757,
+      "learning_rate": 5.378397008992893e-05,
+      "loss": 1.1816,
+      "step": 3600
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.4325421548048769,
+      "learning_rate": 5.373105368344544e-05,
+      "loss": 1.1871,
+      "step": 3601
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.3979860456918098,
+      "learning_rate": 5.3678153756976135e-05,
+      "loss": 1.2037,
+      "step": 3602
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.4217592904356969,
+      "learning_rate": 5.362527032936277e-05,
+      "loss": 1.2599,
+      "step": 3603
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.41255573647682464,
+      "learning_rate": 5.357240341944145e-05,
+      "loss": 1.1994,
+      "step": 3604
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.42758069830519374,
+      "learning_rate": 5.351955304604207e-05,
+      "loss": 0.9812,
+      "step": 3605
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.43974718825680537,
+      "learning_rate": 5.346671922798891e-05,
+      "loss": 1.0692,
+      "step": 3606
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.4186775098297169,
+      "learning_rate": 5.341390198410019e-05,
+      "loss": 1.1826,
+      "step": 3607
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.4466338617712497,
+      "learning_rate": 5.336110133318837e-05,
+      "loss": 1.0258,
+      "step": 3608
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.47153337626397307,
+      "learning_rate": 5.3308317294059826e-05,
+      "loss": 1.0103,
+      "step": 3609
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.3971138240009304,
+      "learning_rate": 5.32555498855152e-05,
+      "loss": 1.3204,
+      "step": 3610
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.4195281784848684,
+      "learning_rate": 5.320279912634907e-05,
+      "loss": 1.2383,
+      "step": 3611
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.40757834855138964,
+      "learning_rate": 5.3150065035350136e-05,
+      "loss": 1.2524,
+      "step": 3612
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.4125982453387227,
+      "learning_rate": 5.309734763130116e-05,
+      "loss": 1.1963,
+      "step": 3613
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.45227860231446865,
+      "learning_rate": 5.304464693297895e-05,
+      "loss": 1.0615,
+      "step": 3614
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.4013642799262556,
+      "learning_rate": 5.299196295915441e-05,
+      "loss": 0.9275,
+      "step": 3615
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.39128001505714544,
+      "learning_rate": 5.2939295728592374e-05,
+      "loss": 0.9478,
+      "step": 3616
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5371408039737074,
+      "learning_rate": 5.2886645260051894e-05,
+      "loss": 1.1296,
+      "step": 3617
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.4068553629314962,
+      "learning_rate": 5.283401157228584e-05,
+      "loss": 1.0666,
+      "step": 3618
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.4150511558111666,
+      "learning_rate": 5.278139468404133e-05,
+      "loss": 1.2872,
+      "step": 3619
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.382186649704804,
+      "learning_rate": 5.2728794614059306e-05,
+      "loss": 1.2579,
+      "step": 3620
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.44317502244385004,
+      "learning_rate": 5.26762113810748e-05,
+      "loss": 1.2359,
+      "step": 3621
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.44078198076808417,
+      "learning_rate": 5.2623645003816804e-05,
+      "loss": 1.0787,
+      "step": 3622
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.4224265732746059,
+      "learning_rate": 5.257109550100843e-05,
+      "loss": 1.2983,
+      "step": 3623
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.419078259276571,
+      "learning_rate": 5.251856289136664e-05,
+      "loss": 1.1055,
+      "step": 3624
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5245916803808178,
+      "learning_rate": 5.2466047193602395e-05,
+      "loss": 1.0998,
+      "step": 3625
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.47124051282459584,
+      "learning_rate": 5.241354842642077e-05,
+      "loss": 1.1777,
+      "step": 3626
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.45329017399984883,
+      "learning_rate": 5.236106660852058e-05,
+      "loss": 1.1497,
+      "step": 3627
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.38269036477529533,
+      "learning_rate": 5.2308601758594866e-05,
+      "loss": 1.1421,
+      "step": 3628
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.45261662643954564,
+      "learning_rate": 5.225615389533043e-05,
+      "loss": 0.9862,
+      "step": 3629
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.4222555165265691,
+      "learning_rate": 5.220372303740807e-05,
+      "loss": 0.9857,
+      "step": 3630
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.4180043113645979,
+      "learning_rate": 5.215130920350252e-05,
+      "loss": 1.113,
+      "step": 3631
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.41554337262016205,
+      "learning_rate": 5.209891241228254e-05,
+      "loss": 1.1317,
+      "step": 3632
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.4795830198376781,
+      "learning_rate": 5.204653268241069e-05,
+      "loss": 1.124,
+      "step": 3633
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.4818420611022101,
+      "learning_rate": 5.199417003254358e-05,
+      "loss": 1.2096,
+      "step": 3634
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.4874540330085074,
+      "learning_rate": 5.1941824481331626e-05,
+      "loss": 1.0258,
+      "step": 3635
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.4372316743921917,
+      "learning_rate": 5.1889496047419216e-05,
+      "loss": 1.1231,
+      "step": 3636
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.3860992920738133,
+      "learning_rate": 5.183718474944464e-05,
+      "loss": 1.1771,
+      "step": 3637
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.4488038223562485,
+      "learning_rate": 5.1784890606039984e-05,
+      "loss": 1.1052,
+      "step": 3638
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.46510271265825254,
+      "learning_rate": 5.173261363583142e-05,
+      "loss": 1.0496,
+      "step": 3639
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.4036669440978354,
+      "learning_rate": 5.168035385743879e-05,
+      "loss": 1.1253,
+      "step": 3640
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.4547846720980273,
+      "learning_rate": 5.162811128947602e-05,
+      "loss": 1.129,
+      "step": 3641
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.43577179952675027,
+      "learning_rate": 5.157588595055073e-05,
+      "loss": 1.1632,
+      "step": 3642
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.43698382198241953,
+      "learning_rate": 5.1523677859264516e-05,
+      "loss": 1.169,
+      "step": 3643
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.43457818329541503,
+      "learning_rate": 5.147148703421277e-05,
+      "loss": 1.1352,
+      "step": 3644
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.4652738122779284,
+      "learning_rate": 5.141931349398477e-05,
+      "loss": 1.2072,
+      "step": 3645
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.45818811767649337,
+      "learning_rate": 5.136715725716358e-05,
+      "loss": 1.1096,
+      "step": 3646
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5023773263982487,
+      "learning_rate": 5.1315018342326124e-05,
+      "loss": 1.1494,
+      "step": 3647
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5239876217435108,
+      "learning_rate": 5.126289676804328e-05,
+      "loss": 1.0977,
+      "step": 3648
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.519024719824936,
+      "learning_rate": 5.1210792552879525e-05,
+      "loss": 1.2413,
+      "step": 3649
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.46632085129514894,
+      "learning_rate": 5.1158705715393364e-05,
+      "loss": 0.9642,
+      "step": 3650
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.39696800921612535,
+      "learning_rate": 5.110663627413694e-05,
+      "loss": 1.1079,
+      "step": 3651
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.43235708965779185,
+      "learning_rate": 5.105458424765637e-05,
+      "loss": 0.9855,
+      "step": 3652
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.46037567722736567,
+      "learning_rate": 5.100254965449143e-05,
+      "loss": 1.2553,
+      "step": 3653
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.44834489218239587,
+      "learning_rate": 5.095053251317572e-05,
+      "loss": 1.1887,
+      "step": 3654
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.39086627788032535,
+      "learning_rate": 5.089853284223667e-05,
+      "loss": 1.1812,
+      "step": 3655
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.4259969741717848,
+      "learning_rate": 5.0846550660195394e-05,
+      "loss": 1.0709,
+      "step": 3656
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.4506154529975262,
+      "learning_rate": 5.079458598556692e-05,
+      "loss": 1.0577,
+      "step": 3657
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.46536073799111655,
+      "learning_rate": 5.07426388368599e-05,
+      "loss": 1.1236,
+      "step": 3658
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.42506035890691923,
+      "learning_rate": 5.069070923257685e-05,
+      "loss": 1.0547,
+      "step": 3659
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.44430645071328473,
+      "learning_rate": 5.063879719121398e-05,
+      "loss": 1.1332,
+      "step": 3660
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.4521874762438369,
+      "learning_rate": 5.058690273126123e-05,
+      "loss": 1.0079,
+      "step": 3661
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.42228687755537314,
+      "learning_rate": 5.053502587120229e-05,
+      "loss": 1.1981,
+      "step": 3662
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.4826211685429359,
+      "learning_rate": 5.0483166629514654e-05,
+      "loss": 1.0632,
+      "step": 3663
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.4706043811866479,
+      "learning_rate": 5.043132502466941e-05,
+      "loss": 1.043,
+      "step": 3664
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.37218307339182644,
+      "learning_rate": 5.0379501075131494e-05,
+      "loss": 0.9418,
+      "step": 3665
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.466872575212506,
+      "learning_rate": 5.032769479935948e-05,
+      "loss": 1.0869,
+      "step": 3666
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.48485890342789795,
+      "learning_rate": 5.0275906215805625e-05,
+      "loss": 1.0138,
+      "step": 3667
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.44325747694575385,
+      "learning_rate": 5.022413534291597e-05,
+      "loss": 1.0493,
+      "step": 3668
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.48980255548964174,
+      "learning_rate": 5.0172382199130167e-05,
+      "loss": 1.1082,
+      "step": 3669
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.39550998090350126,
+      "learning_rate": 5.01206468028816e-05,
+      "loss": 1.1323,
+      "step": 3670
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.38383559713047716,
+      "learning_rate": 5.006892917259728e-05,
+      "loss": 1.0785,
+      "step": 3671
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.41325539756087915,
+      "learning_rate": 5.001722932669799e-05,
+      "loss": 1.0915,
+      "step": 3672
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.4174095730952444,
+      "learning_rate": 4.996554728359804e-05,
+      "loss": 1.211,
+      "step": 3673
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.4335098073474942,
+      "learning_rate": 4.9913883061705534e-05,
+      "loss": 1.0975,
+      "step": 3674
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.4138404091392316,
+      "learning_rate": 4.986223667942214e-05,
+      "loss": 1.2342,
+      "step": 3675
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.45952221095122253,
+      "learning_rate": 4.981060815514322e-05,
+      "loss": 1.0556,
+      "step": 3676
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.43792194285945885,
+      "learning_rate": 4.9758997507257763e-05,
+      "loss": 0.9733,
+      "step": 3677
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.4580539497712538,
+      "learning_rate": 4.970740475414837e-05,
+      "loss": 1.3141,
+      "step": 3678
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.41096037224289356,
+      "learning_rate": 4.965582991419127e-05,
+      "loss": 1.2973,
+      "step": 3679
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.40527222170105015,
+      "learning_rate": 4.960427300575629e-05,
+      "loss": 1.1399,
+      "step": 3680
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.4550237480818617,
+      "learning_rate": 4.9552734047206986e-05,
+      "loss": 1.0358,
+      "step": 3681
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.4359552677644107,
+      "learning_rate": 4.9501213056900355e-05,
+      "loss": 1.1356,
+      "step": 3682
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.42238506165813033,
+      "learning_rate": 4.944971005318716e-05,
+      "loss": 1.1568,
+      "step": 3683
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.4057334783527927,
+      "learning_rate": 4.939822505441163e-05,
+      "loss": 1.2536,
+      "step": 3684
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.4638440554517315,
+      "learning_rate": 4.9346758078911634e-05,
+      "loss": 0.9619,
+      "step": 3685
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.40141709794768865,
+      "learning_rate": 4.929530914501857e-05,
+      "loss": 1.0056,
+      "step": 3686
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.42785837260801896,
+      "learning_rate": 4.924387827105754e-05,
+      "loss": 1.3152,
+      "step": 3687
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.4267350593127641,
+      "learning_rate": 4.919246547534708e-05,
+      "loss": 1.1125,
+      "step": 3688
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.3956881129449576,
+      "learning_rate": 4.9141070776199296e-05,
+      "loss": 1.2692,
+      "step": 3689
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.49693351838460226,
+      "learning_rate": 4.908969419191997e-05,
+      "loss": 1.2512,
+      "step": 3690
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.4591878328861195,
+      "learning_rate": 4.903833574080825e-05,
+      "loss": 1.0777,
+      "step": 3691
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.3800492888299361,
+      "learning_rate": 4.8986995441157016e-05,
+      "loss": 1.0587,
+      "step": 3692
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.47441284254405214,
+      "learning_rate": 4.893567331125256e-05,
+      "loss": 1.119,
+      "step": 3693
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.42393942094710446,
+      "learning_rate": 4.8884369369374715e-05,
+      "loss": 1.2412,
+      "step": 3694
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5490533686707227,
+      "learning_rate": 4.8833083633796826e-05,
+      "loss": 1.0171,
+      "step": 3695
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.4429129391287522,
+      "learning_rate": 4.878181612278586e-05,
+      "loss": 1.0314,
+      "step": 3696
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5065727276343879,
+      "learning_rate": 4.873056685460213e-05,
+      "loss": 1.1263,
+      "step": 3697
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5269886608364802,
+      "learning_rate": 4.867933584749961e-05,
+      "loss": 0.9517,
+      "step": 3698
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.4114417385227601,
+      "learning_rate": 4.862812311972567e-05,
+      "loss": 1.0846,
+      "step": 3699
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.3888103950680494,
+      "learning_rate": 4.857692868952113e-05,
+      "loss": 1.2313,
+      "step": 3700
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.4408921170620224,
+      "learning_rate": 4.8525752575120454e-05,
+      "loss": 1.2413,
+      "step": 3701
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.48869035335456673,
+      "learning_rate": 4.8474594794751447e-05,
+      "loss": 1.0047,
+      "step": 3702
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.42359883073996396,
+      "learning_rate": 4.8423455366635404e-05,
+      "loss": 1.0815,
+      "step": 3703
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5206995915520345,
+      "learning_rate": 4.8372334308987086e-05,
+      "loss": 1.1075,
+      "step": 3704
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.45190472256120046,
+      "learning_rate": 4.8321231640014794e-05,
+      "loss": 1.1298,
+      "step": 3705
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.47657444021136625,
+      "learning_rate": 4.827014737792013e-05,
+      "loss": 0.9629,
+      "step": 3706
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.4917226151792126,
+      "learning_rate": 4.8219081540898295e-05,
+      "loss": 1.1779,
+      "step": 3707
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.3896723858365976,
+      "learning_rate": 4.816803414713782e-05,
+      "loss": 1.2323,
+      "step": 3708
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.41587587402719634,
+      "learning_rate": 4.811700521482071e-05,
+      "loss": 1.1144,
+      "step": 3709
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5727305186302124,
+      "learning_rate": 4.806599476212237e-05,
+      "loss": 1.0941,
+      "step": 3710
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.4388706226906959,
+      "learning_rate": 4.80150028072116e-05,
+      "loss": 1.0931,
+      "step": 3711
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.3932330515740056,
+      "learning_rate": 4.796402936825074e-05,
+      "loss": 1.2931,
+      "step": 3712
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.3981550375983292,
+      "learning_rate": 4.791307446339536e-05,
+      "loss": 1.1075,
+      "step": 3713
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5020767315528754,
+      "learning_rate": 4.786213811079459e-05,
+      "loss": 1.2411,
+      "step": 3714
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.49297408965491296,
+      "learning_rate": 4.781122032859079e-05,
+      "loss": 1.1461,
+      "step": 3715
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.41798260590896225,
+      "learning_rate": 4.7760321134919875e-05,
+      "loss": 1.0727,
+      "step": 3716
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.4213774583812704,
+      "learning_rate": 4.770944054791101e-05,
+      "loss": 1.1205,
+      "step": 3717
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.4290042590601909,
+      "learning_rate": 4.7658578585686784e-05,
+      "loss": 1.0762,
+      "step": 3718
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.4648230995743925,
+      "learning_rate": 4.760773526636315e-05,
+      "loss": 1.0107,
+      "step": 3719
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.39457862137906863,
+      "learning_rate": 4.755691060804938e-05,
+      "loss": 1.2435,
+      "step": 3720
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5059984728979937,
+      "learning_rate": 4.7506104628848204e-05,
+      "loss": 1.0361,
+      "step": 3721
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.4014175754353781,
+      "learning_rate": 4.7455317346855546e-05,
+      "loss": 1.1107,
+      "step": 3722
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.44997874880930994,
+      "learning_rate": 4.740454878016084e-05,
+      "loss": 1.1904,
+      "step": 3723
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.4136286589079954,
+      "learning_rate": 4.735379894684671e-05,
+      "loss": 1.2244,
+      "step": 3724
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5065271364033525,
+      "learning_rate": 4.730306786498921e-05,
+      "loss": 1.0088,
+      "step": 3725
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5648025955392068,
+      "learning_rate": 4.7252355552657657e-05,
+      "loss": 1.0491,
+      "step": 3726
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.38275752810469643,
+      "learning_rate": 4.72016620279147e-05,
+      "loss": 1.0237,
+      "step": 3727
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.480966414671511,
+      "learning_rate": 4.715098730881623e-05,
+      "loss": 1.279,
+      "step": 3728
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.42017190227087087,
+      "learning_rate": 4.71003314134116e-05,
+      "loss": 1.056,
+      "step": 3729
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.4636710608068418,
+      "learning_rate": 4.704969435974334e-05,
+      "loss": 1.3196,
+      "step": 3730
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.4029187350861382,
+      "learning_rate": 4.699907616584721e-05,
+      "loss": 1.0967,
+      "step": 3731
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.49283772226229067,
+      "learning_rate": 4.694847684975245e-05,
+      "loss": 1.155,
+      "step": 3732
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.48731832411345144,
+      "learning_rate": 4.68978964294814e-05,
+      "loss": 1.1677,
+      "step": 3733
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.38532417654553464,
+      "learning_rate": 4.684733492304974e-05,
+      "loss": 1.1925,
+      "step": 3734
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.41639118754582205,
+      "learning_rate": 4.6796792348466356e-05,
+      "loss": 1.0835,
+      "step": 3735
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.43689342470428927,
+      "learning_rate": 4.674626872373351e-05,
+      "loss": 1.1423,
+      "step": 3736
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.43162018854656053,
+      "learning_rate": 4.669576406684658e-05,
+      "loss": 1.1599,
+      "step": 3737
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.4578030000074734,
+      "learning_rate": 4.6645278395794324e-05,
+      "loss": 1.033,
+      "step": 3738
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.42466308560080257,
+      "learning_rate": 4.6594811728558584e-05,
+      "loss": 1.1847,
+      "step": 3739
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.4863221652994277,
+      "learning_rate": 4.6544364083114575e-05,
+      "loss": 1.1498,
+      "step": 3740
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.4148547211656974,
+      "learning_rate": 4.6493935477430664e-05,
+      "loss": 1.1443,
+      "step": 3741
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.40078001061737495,
+      "learning_rate": 4.644352592946842e-05,
+      "loss": 1.1754,
+      "step": 3742
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.45036783467134295,
+      "learning_rate": 4.6393135457182676e-05,
+      "loss": 0.9838,
+      "step": 3743
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.40191427475246994,
+      "learning_rate": 4.634276407852139e-05,
+      "loss": 1.295,
+      "step": 3744
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.4465591307798951,
+      "learning_rate": 4.629241181142587e-05,
+      "loss": 1.0528,
+      "step": 3745
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.42405855251794433,
+      "learning_rate": 4.6242078673830426e-05,
+      "loss": 1.1184,
+      "step": 3746
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.48941251800742747,
+      "learning_rate": 4.6191764683662744e-05,
+      "loss": 1.0112,
+      "step": 3747
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5145259126509486,
+      "learning_rate": 4.61414698588435e-05,
+      "loss": 1.2344,
+      "step": 3748
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.4709220608707974,
+      "learning_rate": 4.609119421728674e-05,
+      "loss": 1.2875,
+      "step": 3749
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.4829537114290681,
+      "learning_rate": 4.6040937776899517e-05,
+      "loss": 1.1491,
+      "step": 3750
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.4759417802270143,
+      "learning_rate": 4.599070055558211e-05,
+      "loss": 1.2096,
+      "step": 3751
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.4149014322921493,
+      "learning_rate": 4.594048257122797e-05,
+      "loss": 1.0866,
+      "step": 3752
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.4633474624218409,
+      "learning_rate": 4.589028384172361e-05,
+      "loss": 1.0985,
+      "step": 3753
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.43088537235787977,
+      "learning_rate": 4.584010438494885e-05,
+      "loss": 1.056,
+      "step": 3754
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.4069230021263997,
+      "learning_rate": 4.578994421877645e-05,
+      "loss": 1.0948,
+      "step": 3755
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.4151874927055056,
+      "learning_rate": 4.573980336107247e-05,
+      "loss": 1.109,
+      "step": 3756
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.4294830510690449,
+      "learning_rate": 4.568968182969595e-05,
+      "loss": 0.9809,
+      "step": 3757
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.4451397094037277,
+      "learning_rate": 4.563957964249923e-05,
+      "loss": 1.1425,
+      "step": 3758
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.46423338643775147,
+      "learning_rate": 4.558949681732747e-05,
+      "loss": 0.9788,
+      "step": 3759
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.39501273387213337,
+      "learning_rate": 4.5539433372019246e-05,
+      "loss": 1.25,
+      "step": 3760
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.38679948544368103,
+      "learning_rate": 4.548938932440604e-05,
+      "loss": 1.1717,
+      "step": 3761
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.4049560178496815,
+      "learning_rate": 4.543936469231244e-05,
+      "loss": 1.1137,
+      "step": 3762
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.44164110790696115,
+      "learning_rate": 4.538935949355623e-05,
+      "loss": 1.1394,
+      "step": 3763
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.606049575558055,
+      "learning_rate": 4.533937374594813e-05,
+      "loss": 0.9469,
+      "step": 3764
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.40759663961816245,
+      "learning_rate": 4.5289407467292064e-05,
+      "loss": 1.1715,
+      "step": 3765
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.44863159566754607,
+      "learning_rate": 4.5239460675384936e-05,
+      "loss": 1.0969,
+      "step": 3766
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.4343637922433307,
+      "learning_rate": 4.5189533388016715e-05,
+      "loss": 1.0384,
+      "step": 3767
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.4141486207941082,
+      "learning_rate": 4.5139625622970404e-05,
+      "loss": 1.128,
+      "step": 3768
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.4438253134283439,
+      "learning_rate": 4.508973739802217e-05,
+      "loss": 1.1344,
+      "step": 3769
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.4339864860899427,
+      "learning_rate": 4.5039868730941046e-05,
+      "loss": 1.1066,
+      "step": 3770
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.4154202059525997,
+      "learning_rate": 4.499001963948929e-05,
+      "loss": 1.1426,
+      "step": 3771
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.4305791113691321,
+      "learning_rate": 4.494019014142202e-05,
+      "loss": 1.0515,
+      "step": 3772
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5082676060308273,
+      "learning_rate": 4.489038025448743e-05,
+      "loss": 1.0448,
+      "step": 3773
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.41261739189816204,
+      "learning_rate": 4.484058999642681e-05,
+      "loss": 1.2012,
+      "step": 3774
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.4569221106532274,
+      "learning_rate": 4.479081938497435e-05,
+      "loss": 1.1315,
+      "step": 3775
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.42680480861197506,
+      "learning_rate": 4.474106843785728e-05,
+      "loss": 1.1035,
+      "step": 3776
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.4284279715944373,
+      "learning_rate": 4.46913371727958e-05,
+      "loss": 1.1648,
+      "step": 3777
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.43233419041195975,
+      "learning_rate": 4.464162560750319e-05,
+      "loss": 1.1335,
+      "step": 3778
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.40206992791119217,
+      "learning_rate": 4.45919337596856e-05,
+      "loss": 1.0769,
+      "step": 3779
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.45414882808074625,
+      "learning_rate": 4.4542261647042263e-05,
+      "loss": 1.2079,
+      "step": 3780
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.4656550593020722,
+      "learning_rate": 4.4492609287265254e-05,
+      "loss": 0.9491,
+      "step": 3781
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.460268571796791,
+      "learning_rate": 4.444297669803981e-05,
+      "loss": 1.2498,
+      "step": 3782
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.400432386968432,
+      "learning_rate": 4.439336389704388e-05,
+      "loss": 1.2044,
+      "step": 3783
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5154603987891748,
+      "learning_rate": 4.43437709019485e-05,
+      "loss": 1.0686,
+      "step": 3784
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.4628189830211725,
+      "learning_rate": 4.4294197730417705e-05,
+      "loss": 1.071,
+      "step": 3785
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.409491108700332,
+      "learning_rate": 4.424464440010834e-05,
+      "loss": 1.0619,
+      "step": 3786
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.440260100856978,
+      "learning_rate": 4.41951109286703e-05,
+      "loss": 1.2243,
+      "step": 3787
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.4316156074868064,
+      "learning_rate": 4.4145597333746314e-05,
+      "loss": 1.2278,
+      "step": 3788
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.3981169083559841,
+      "learning_rate": 4.409610363297211e-05,
+      "loss": 1.1243,
+      "step": 3789
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.42751572300188384,
+      "learning_rate": 4.4046629843976275e-05,
+      "loss": 1.1246,
+      "step": 3790
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.4571660539219093,
+      "learning_rate": 4.39971759843803e-05,
+      "loss": 1.0851,
+      "step": 3791
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.43180281003212223,
+      "learning_rate": 4.394774207179859e-05,
+      "loss": 1.0903,
+      "step": 3792
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.4543917030709054,
+      "learning_rate": 4.38983281238385e-05,
+      "loss": 0.9914,
+      "step": 3793
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.42744435667123404,
+      "learning_rate": 4.384893415810021e-05,
+      "loss": 1.1304,
+      "step": 3794
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.47554216130988614,
+      "learning_rate": 4.379956019217675e-05,
+      "loss": 0.9919,
+      "step": 3795
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.4370604783178502,
+      "learning_rate": 4.3750206243654146e-05,
+      "loss": 1.1588,
+      "step": 3796
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.4493074060518425,
+      "learning_rate": 4.3700872330111166e-05,
+      "loss": 1.1371,
+      "step": 3797
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.4288245905191339,
+      "learning_rate": 4.365155846911957e-05,
+      "loss": 1.1673,
+      "step": 3798
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.4315111569067553,
+      "learning_rate": 4.3602264678243854e-05,
+      "loss": 1.2161,
+      "step": 3799
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.48747426453748577,
+      "learning_rate": 4.355299097504143e-05,
+      "loss": 1.0065,
+      "step": 3800
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.44141758452460755,
+      "learning_rate": 4.350373737706251e-05,
+      "loss": 1.204,
+      "step": 3801
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.49785321988344705,
+      "learning_rate": 4.345450390185024e-05,
+      "loss": 1.11,
+      "step": 3802
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.4100588847079784,
+      "learning_rate": 4.340529056694047e-05,
+      "loss": 1.2691,
+      "step": 3803
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.44143791908631963,
+      "learning_rate": 4.335609738986201e-05,
+      "loss": 1.2553,
+      "step": 3804
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.4444388212752621,
+      "learning_rate": 4.3306924388136405e-05,
+      "loss": 1.1547,
+      "step": 3805
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.4534902382457652,
+      "learning_rate": 4.3257771579277964e-05,
+      "loss": 1.1347,
+      "step": 3806
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.4585156050264357,
+      "learning_rate": 4.3208638980794e-05,
+      "loss": 1.2106,
+      "step": 3807
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.4055654845855362,
+      "learning_rate": 4.3159526610184365e-05,
+      "loss": 0.972,
+      "step": 3808
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5333822686255315,
+      "learning_rate": 4.3110434484941934e-05,
+      "loss": 1.2104,
+      "step": 3809
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.43057633503049847,
+      "learning_rate": 4.306136262255219e-05,
+      "loss": 1.0753,
+      "step": 3810
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.4835896399068684,
+      "learning_rate": 4.3012311040493594e-05,
+      "loss": 1.1436,
+      "step": 3811
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.41575951453758214,
+      "learning_rate": 4.296327975623716e-05,
+      "loss": 1.227,
+      "step": 3812
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.39930282674131495,
+      "learning_rate": 4.291426878724691e-05,
+      "loss": 1.1761,
+      "step": 3813
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.41830323942881537,
+      "learning_rate": 4.286527815097944e-05,
+      "loss": 1.1147,
+      "step": 3814
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.4928190493849124,
+      "learning_rate": 4.281630786488418e-05,
+      "loss": 1.0909,
+      "step": 3815
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.40582552076921535,
+      "learning_rate": 4.27673579464033e-05,
+      "loss": 1.3019,
+      "step": 3816
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5939354056786278,
+      "learning_rate": 4.271842841297169e-05,
+      "loss": 1.0662,
+      "step": 3817
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.3999393723351442,
+      "learning_rate": 4.266951928201708e-05,
+      "loss": 1.0296,
+      "step": 3818
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.4915430629730649,
+      "learning_rate": 4.2620630570959775e-05,
+      "loss": 1.0568,
+      "step": 3819
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.475465188458468,
+      "learning_rate": 4.257176229721299e-05,
+      "loss": 1.1467,
+      "step": 3820
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.41152732443736,
+      "learning_rate": 4.252291447818247e-05,
+      "loss": 1.0931,
+      "step": 3821
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.4280694717395832,
+      "learning_rate": 4.247408713126687e-05,
+      "loss": 1.2991,
+      "step": 3822
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.46238345894972527,
+      "learning_rate": 4.242528027385738e-05,
+      "loss": 1.0237,
+      "step": 3823
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.47910530171067883,
+      "learning_rate": 4.237649392333799e-05,
+      "loss": 1.0772,
+      "step": 3824
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.45665859441476664,
+      "learning_rate": 4.232772809708535e-05,
+      "loss": 1.2533,
+      "step": 3825
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.46227958486391907,
+      "learning_rate": 4.227898281246877e-05,
+      "loss": 1.1868,
+      "step": 3826
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.4322994784337255,
+      "learning_rate": 4.2230258086850374e-05,
+      "loss": 1.1359,
+      "step": 3827
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.4182705071394501,
+      "learning_rate": 4.2181553937584775e-05,
+      "loss": 1.1104,
+      "step": 3828
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.46180477469820796,
+      "learning_rate": 4.213287038201943e-05,
+      "loss": 1.0302,
+      "step": 3829
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.4379138758816305,
+      "learning_rate": 4.208420743749434e-05,
+      "loss": 1.156,
+      "step": 3830
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.4099911356365074,
+      "learning_rate": 4.2035565121342246e-05,
+      "loss": 1.142,
+      "step": 3831
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.4797296331736634,
+      "learning_rate": 4.1986943450888506e-05,
+      "loss": 1.1095,
+      "step": 3832
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.4714021974116638,
+      "learning_rate": 4.1938342443451095e-05,
+      "loss": 1.1794,
+      "step": 3833
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.42359707569216354,
+      "learning_rate": 4.1889762116340635e-05,
+      "loss": 1.3779,
+      "step": 3834
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.39893126982446586,
+      "learning_rate": 4.184120248686048e-05,
+      "loss": 1.1009,
+      "step": 3835
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.44391464913915774,
+      "learning_rate": 4.179266357230649e-05,
+      "loss": 1.0324,
+      "step": 3836
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5822465827967821,
+      "learning_rate": 4.174414538996717e-05,
+      "loss": 1.1458,
+      "step": 3837
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.40152434839204165,
+      "learning_rate": 4.1695647957123727e-05,
+      "loss": 1.0083,
+      "step": 3838
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.41932766986951847,
+      "learning_rate": 4.164717129104988e-05,
+      "loss": 1.0827,
+      "step": 3839
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.4241421966472261,
+      "learning_rate": 4.159871540901199e-05,
+      "loss": 1.3244,
+      "step": 3840
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.3843750028970654,
+      "learning_rate": 4.155028032826897e-05,
+      "loss": 1.1385,
+      "step": 3841
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.41134978324238686,
+      "learning_rate": 4.1501866066072436e-05,
+      "loss": 1.0625,
+      "step": 3842
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.4149606650767285,
+      "learning_rate": 4.1453472639666457e-05,
+      "loss": 1.0853,
+      "step": 3843
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.4639268336331141,
+      "learning_rate": 4.1405100066287803e-05,
+      "loss": 0.9677,
+      "step": 3844
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.45859645762203416,
+      "learning_rate": 4.135674836316569e-05,
+      "loss": 1.0528,
+      "step": 3845
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.4961007647692676,
+      "learning_rate": 4.130841754752203e-05,
+      "loss": 1.012,
+      "step": 3846
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.43071847101493943,
+      "learning_rate": 4.12601076365712e-05,
+      "loss": 0.9908,
+      "step": 3847
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.46320654610746126,
+      "learning_rate": 4.1211818647520164e-05,
+      "loss": 1.0395,
+      "step": 3848
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.43491705960573895,
+      "learning_rate": 4.116355059756843e-05,
+      "loss": 1.1302,
+      "step": 3849
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.41303832558333986,
+      "learning_rate": 4.111530350390801e-05,
+      "loss": 1.0664,
+      "step": 3850
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.463894360654933,
+      "learning_rate": 4.106707738372357e-05,
+      "loss": 1.1125,
+      "step": 3851
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.4170360924703485,
+      "learning_rate": 4.1018872254192166e-05,
+      "loss": 1.1095,
+      "step": 3852
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.44688672533878576,
+      "learning_rate": 4.097068813248349e-05,
+      "loss": 1.3112,
+      "step": 3853
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5769573802571853,
+      "learning_rate": 4.092252503575965e-05,
+      "loss": 1.1189,
+      "step": 3854
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.3979109151946672,
+      "learning_rate": 4.087438298117536e-05,
+      "loss": 1.1149,
+      "step": 3855
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.4403863500249152,
+      "learning_rate": 4.082626198587779e-05,
+      "loss": 0.9581,
+      "step": 3856
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.45974745182791216,
+      "learning_rate": 4.07781620670066e-05,
+      "loss": 1.1993,
+      "step": 3857
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.4246207001147996,
+      "learning_rate": 4.073008324169396e-05,
+      "loss": 1.0067,
+      "step": 3858
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.49892975285773083,
+      "learning_rate": 4.0682025527064486e-05,
+      "loss": 1.0217,
+      "step": 3859
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.4192805557909134,
+      "learning_rate": 4.063398894023538e-05,
+      "loss": 1.0268,
+      "step": 3860
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.39761168253743684,
+      "learning_rate": 4.05859734983162e-05,
+      "loss": 1.1121,
+      "step": 3861
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.4100671769720258,
+      "learning_rate": 4.053797921840907e-05,
+      "loss": 1.0795,
+      "step": 3862
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.39815171246068876,
+      "learning_rate": 4.049000611760852e-05,
+      "loss": 1.2996,
+      "step": 3863
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.4373161770801694,
+      "learning_rate": 4.044205421300154e-05,
+      "loss": 1.2018,
+      "step": 3864
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.4589155783972582,
+      "learning_rate": 4.0394123521667524e-05,
+      "loss": 1.0251,
+      "step": 3865
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.4599901026985476,
+      "learning_rate": 4.034621406067847e-05,
+      "loss": 1.0888,
+      "step": 3866
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.40872521919238464,
+      "learning_rate": 4.029832584709864e-05,
+      "loss": 1.1071,
+      "step": 3867
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.4998660707395915,
+      "learning_rate": 4.025045889798478e-05,
+      "loss": 1.0916,
+      "step": 3868
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.42585714202517333,
+      "learning_rate": 4.020261323038614e-05,
+      "loss": 1.141,
+      "step": 3869
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.4582064357081592,
+      "learning_rate": 4.015478886134427e-05,
+      "loss": 1.1612,
+      "step": 3870
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.4221439570881287,
+      "learning_rate": 4.010698580789325e-05,
+      "loss": 1.0529,
+      "step": 3871
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.419840278337427,
+      "learning_rate": 4.0059204087059485e-05,
+      "loss": 1.107,
+      "step": 3872
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.36973162391065845,
+      "learning_rate": 4.001144371586181e-05,
+      "loss": 1.0454,
+      "step": 3873
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5084709913659953,
+      "learning_rate": 3.9963704711311404e-05,
+      "loss": 0.8629,
+      "step": 3874
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.43494779503902287,
+      "learning_rate": 3.991598709041196e-05,
+      "loss": 1.1942,
+      "step": 3875
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.4746542736606154,
+      "learning_rate": 3.9868290870159405e-05,
+      "loss": 1.1117,
+      "step": 3876
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.3909611101415937,
+      "learning_rate": 3.9820616067542204e-05,
+      "loss": 1.0476,
+      "step": 3877
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.4689084523937681,
+      "learning_rate": 3.9772962699541075e-05,
+      "loss": 1.2145,
+      "step": 3878
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5502562175064195,
+      "learning_rate": 3.972533078312907e-05,
+      "loss": 0.9705,
+      "step": 3879
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.4211653768843274,
+      "learning_rate": 3.967772033527175e-05,
+      "loss": 1.1129,
+      "step": 3880
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.42531266918406824,
+      "learning_rate": 3.963013137292691e-05,
+      "loss": 1.1355,
+      "step": 3881
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.4676820749392726,
+      "learning_rate": 3.958256391304472e-05,
+      "loss": 1.0636,
+      "step": 3882
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5333282467016559,
+      "learning_rate": 3.9535017972567675e-05,
+      "loss": 0.9908,
+      "step": 3883
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.3988956767315207,
+      "learning_rate": 3.948749356843069e-05,
+      "loss": 1.2826,
+      "step": 3884
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.4399836136358537,
+      "learning_rate": 3.943999071756088e-05,
+      "loss": 1.1316,
+      "step": 3885
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.4999649430507637,
+      "learning_rate": 3.93925094368778e-05,
+      "loss": 0.9852,
+      "step": 3886
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.4965596892572626,
+      "learning_rate": 3.934504974329326e-05,
+      "loss": 1.0564,
+      "step": 3887
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.4792962759562115,
+      "learning_rate": 3.9297611653711383e-05,
+      "loss": 1.0809,
+      "step": 3888
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.4384680866184174,
+      "learning_rate": 3.925019518502861e-05,
+      "loss": 0.9735,
+      "step": 3889
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.43400629486165193,
+      "learning_rate": 3.920280035413365e-05,
+      "loss": 1.1448,
+      "step": 3890
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.451332592866665,
+      "learning_rate": 3.915542717790759e-05,
+      "loss": 1.1935,
+      "step": 3891
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5251714463887087,
+      "learning_rate": 3.910807567322369e-05,
+      "loss": 1.065,
+      "step": 3892
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5032327169714736,
+      "learning_rate": 3.9060745856947603e-05,
+      "loss": 1.1536,
+      "step": 3893
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.41155628238882264,
+      "learning_rate": 3.901343774593714e-05,
+      "loss": 1.1946,
+      "step": 3894
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.4475659926011557,
+      "learning_rate": 3.896615135704251e-05,
+      "loss": 1.1208,
+      "step": 3895
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.4210526847332113,
+      "learning_rate": 3.8918886707106106e-05,
+      "loss": 1.1221,
+      "step": 3896
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.44587277653410173,
+      "learning_rate": 3.887164381296256e-05,
+      "loss": 1.1463,
+      "step": 3897
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.511046432625359,
+      "learning_rate": 3.882442269143877e-05,
+      "loss": 1.0522,
+      "step": 3898
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.40154971296967745,
+      "learning_rate": 3.877722335935394e-05,
+      "loss": 1.2366,
+      "step": 3899
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.3956748008098747,
+      "learning_rate": 3.8730045833519465e-05,
+      "loss": 1.0665,
+      "step": 3900
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.43574849766430984,
+      "learning_rate": 3.868289013073893e-05,
+      "loss": 0.983,
+      "step": 3901
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.4028488756540075,
+      "learning_rate": 3.863575626780826e-05,
+      "loss": 0.9322,
+      "step": 3902
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.4456623824661753,
+      "learning_rate": 3.858864426151545e-05,
+      "loss": 1.0739,
+      "step": 3903
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.44299316872043454,
+      "learning_rate": 3.854155412864089e-05,
+      "loss": 1.1049,
+      "step": 3904
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.4712273677321107,
+      "learning_rate": 3.849448588595703e-05,
+      "loss": 1.0506,
+      "step": 3905
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.47926254076483576,
+      "learning_rate": 3.844743955022859e-05,
+      "loss": 1.0085,
+      "step": 3906
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.439165881024421,
+      "learning_rate": 3.840041513821243e-05,
+      "loss": 1.1616,
+      "step": 3907
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.46666725274830906,
+      "learning_rate": 3.8353412666657705e-05,
+      "loss": 1.1096,
+      "step": 3908
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.47879291217571324,
+      "learning_rate": 3.830643215230565e-05,
+      "loss": 1.1125,
+      "step": 3909
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.3742964144538944,
+      "learning_rate": 3.825947361188977e-05,
+      "loss": 1.229,
+      "step": 3910
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.39171549974628883,
+      "learning_rate": 3.821253706213569e-05,
+      "loss": 1.1636,
+      "step": 3911
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.38583447822618716,
+      "learning_rate": 3.816562251976119e-05,
+      "loss": 1.1027,
+      "step": 3912
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.45467881676117167,
+      "learning_rate": 3.811873000147622e-05,
+      "loss": 1.0791,
+      "step": 3913
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.4195640300458811,
+      "learning_rate": 3.80718595239829e-05,
+      "loss": 1.1397,
+      "step": 3914
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.44468718591337175,
+      "learning_rate": 3.802501110397553e-05,
+      "loss": 1.2805,
+      "step": 3915
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.37895133646042006,
+      "learning_rate": 3.7978184758140476e-05,
+      "loss": 1.0806,
+      "step": 3916
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.42039512494823866,
+      "learning_rate": 3.793138050315635e-05,
+      "loss": 1.1841,
+      "step": 3917
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.39999792641629817,
+      "learning_rate": 3.788459835569373e-05,
+      "loss": 1.0189,
+      "step": 3918
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.47639334463345456,
+      "learning_rate": 3.783783833241554e-05,
+      "loss": 1.1967,
+      "step": 3919
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5061563920672297,
+      "learning_rate": 3.779110044997665e-05,
+      "loss": 1.0678,
+      "step": 3920
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.48420041158672533,
+      "learning_rate": 3.774438472502409e-05,
+      "loss": 0.9801,
+      "step": 3921
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.45760465608280915,
+      "learning_rate": 3.769769117419703e-05,
+      "loss": 1.1529,
+      "step": 3922
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.46064220633829894,
+      "learning_rate": 3.7651019814126654e-05,
+      "loss": 1.0793,
+      "step": 3923
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.4597457888157605,
+      "learning_rate": 3.760437066143641e-05,
+      "loss": 1.0045,
+      "step": 3924
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.40774912071948377,
+      "learning_rate": 3.7557743732741625e-05,
+      "loss": 1.047,
+      "step": 3925
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.3936270328563426,
+      "learning_rate": 3.7511139044649926e-05,
+      "loss": 1.144,
+      "step": 3926
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.4175548114787764,
+      "learning_rate": 3.7464556613760826e-05,
+      "loss": 1.0292,
+      "step": 3927
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.4321586270189877,
+      "learning_rate": 3.741799645666606e-05,
+      "loss": 1.1126,
+      "step": 3928
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.39945307119909906,
+      "learning_rate": 3.737145858994934e-05,
+      "loss": 1.1933,
+      "step": 3929
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.4467919327393091,
+      "learning_rate": 3.7324943030186455e-05,
+      "loss": 1.1291,
+      "step": 3930
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.4192323944747912,
+      "learning_rate": 3.727844979394526e-05,
+      "loss": 1.0031,
+      "step": 3931
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.4199114135728555,
+      "learning_rate": 3.7231978897785635e-05,
+      "loss": 1.2142,
+      "step": 3932
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.447041103966067,
+      "learning_rate": 3.7185530358259565e-05,
+      "loss": 1.1323,
+      "step": 3933
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5386306093052604,
+      "learning_rate": 3.7139104191911004e-05,
+      "loss": 1.167,
+      "step": 3934
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.4138622421757206,
+      "learning_rate": 3.709270041527599e-05,
+      "loss": 1.0149,
+      "step": 3935
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.4767587433019716,
+      "learning_rate": 3.704631904488256e-05,
+      "loss": 1.0817,
+      "step": 3936
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5041666304782385,
+      "learning_rate": 3.6999960097250755e-05,
+      "loss": 1.1265,
+      "step": 3937
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.406211201194204,
+      "learning_rate": 3.6953623588892603e-05,
+      "loss": 1.2379,
+      "step": 3938
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5453603178024163,
+      "learning_rate": 3.6907309536312276e-05,
+      "loss": 1.1118,
+      "step": 3939
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.3920513835225689,
+      "learning_rate": 3.686101795600576e-05,
+      "loss": 1.2614,
+      "step": 3940
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.3919771389633014,
+      "learning_rate": 3.6814748864461216e-05,
+      "loss": 1.2383,
+      "step": 3941
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.4234541292102585,
+      "learning_rate": 3.6768502278158665e-05,
+      "loss": 1.1605,
+      "step": 3942
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.44680543252135324,
+      "learning_rate": 3.672227821357014e-05,
+      "loss": 1.0461,
+      "step": 3943
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.4644360395198038,
+      "learning_rate": 3.667607668715972e-05,
+      "loss": 1.086,
+      "step": 3944
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.39567798467940657,
+      "learning_rate": 3.662989771538338e-05,
+      "loss": 1.06,
+      "step": 3945
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.49773215482649685,
+      "learning_rate": 3.6583741314689076e-05,
+      "loss": 1.0779,
+      "step": 3946
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.4277230182539193,
+      "learning_rate": 3.6537607501516715e-05,
+      "loss": 1.2986,
+      "step": 3947
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.4017672916430747,
+      "learning_rate": 3.6491496292298246e-05,
+      "loss": 1.1233,
+      "step": 3948
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.42861888506397783,
+      "learning_rate": 3.644540770345742e-05,
+      "loss": 1.0715,
+      "step": 3949
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5509668073751158,
+      "learning_rate": 3.63993417514101e-05,
+      "loss": 1.1093,
+      "step": 3950
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.41949447323607253,
+      "learning_rate": 3.635329845256391e-05,
+      "loss": 1.1468,
+      "step": 3951
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.3780741425645238,
+      "learning_rate": 3.6307277823318574e-05,
+      "loss": 1.2427,
+      "step": 3952
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.40676705727680984,
+      "learning_rate": 3.626127988006563e-05,
+      "loss": 0.9555,
+      "step": 3953
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.4311786696817371,
+      "learning_rate": 3.6215304639188555e-05,
+      "loss": 1.1799,
+      "step": 3954
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.44054723667261375,
+      "learning_rate": 3.616935211706275e-05,
+      "loss": 1.1371,
+      "step": 3955
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.42120363565063434,
+      "learning_rate": 3.61234223300555e-05,
+      "loss": 0.9771,
+      "step": 3956
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.3999761111266724,
+      "learning_rate": 3.607751529452609e-05,
+      "loss": 1.1748,
+      "step": 3957
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.4534175146079946,
+      "learning_rate": 3.6031631026825554e-05,
+      "loss": 1.2695,
+      "step": 3958
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.4556232211931274,
+      "learning_rate": 3.598576954329695e-05,
+      "loss": 1.0329,
+      "step": 3959
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5492095102834316,
+      "learning_rate": 3.5939930860275115e-05,
+      "loss": 1.1123,
+      "step": 3960
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.47677896298342004,
+      "learning_rate": 3.589411499408691e-05,
+      "loss": 1.1029,
+      "step": 3961
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.4401535758495748,
+      "learning_rate": 3.5848321961050823e-05,
+      "loss": 0.9605,
+      "step": 3962
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.44880337611907384,
+      "learning_rate": 3.5802551777477476e-05,
+      "loss": 1.1295,
+      "step": 3963
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.4416026279567653,
+      "learning_rate": 3.575680445966918e-05,
+      "loss": 1.1539,
+      "step": 3964
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.4125093373442285,
+      "learning_rate": 3.571108002392014e-05,
+      "loss": 1.1922,
+      "step": 3965
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.43222690588984275,
+      "learning_rate": 3.56653784865165e-05,
+      "loss": 1.1901,
+      "step": 3966
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.4335325340803002,
+      "learning_rate": 3.5619699863736076e-05,
+      "loss": 1.2395,
+      "step": 3967
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.4290670933528832,
+      "learning_rate": 3.557404417184873e-05,
+      "loss": 1.0734,
+      "step": 3968
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.40500185750313034,
+      "learning_rate": 3.5528411427115984e-05,
+      "loss": 1.1471,
+      "step": 3969
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.3963204278653148,
+      "learning_rate": 3.548280164579126e-05,
+      "loss": 1.1184,
+      "step": 3970
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.4486893972198056,
+      "learning_rate": 3.543721484411976e-05,
+      "loss": 1.0997,
+      "step": 3971
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.4735488271706566,
+      "learning_rate": 3.53916510383386e-05,
+      "loss": 1.0197,
+      "step": 3972
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.4621997057166897,
+      "learning_rate": 3.5346110244676564e-05,
+      "loss": 1.0059,
+      "step": 3973
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.454763766081732,
+      "learning_rate": 3.53005924793544e-05,
+      "loss": 1.1833,
+      "step": 3974
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.4013498613962083,
+      "learning_rate": 3.525509775858451e-05,
+      "loss": 1.2262,
+      "step": 3975
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.42449659739581275,
+      "learning_rate": 3.520962609857112e-05,
+      "loss": 1.1867,
+      "step": 3976
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.46663239131399387,
+      "learning_rate": 3.516417751551035e-05,
+      "loss": 1.2504,
+      "step": 3977
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.4813487757524383,
+      "learning_rate": 3.5118752025589965e-05,
+      "loss": 0.9043,
+      "step": 3978
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.444405994384315,
+      "learning_rate": 3.5073349644989564e-05,
+      "loss": 1.2528,
+      "step": 3979
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.38414914866388367,
+      "learning_rate": 3.5027970389880486e-05,
+      "loss": 1.2175,
+      "step": 3980
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.48487743498881897,
+      "learning_rate": 3.498261427642589e-05,
+      "loss": 1.1019,
+      "step": 3981
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.4508576938322537,
+      "learning_rate": 3.493728132078064e-05,
+      "loss": 1.1651,
+      "step": 3982
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.44472517679280565,
+      "learning_rate": 3.489197153909139e-05,
+      "loss": 1.1967,
+      "step": 3983
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.42400192099220185,
+      "learning_rate": 3.484668494749653e-05,
+      "loss": 1.1146,
+      "step": 3984
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5469076044843727,
+      "learning_rate": 3.4801421562126144e-05,
+      "loss": 1.1511,
+      "step": 3985
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.434459991697995,
+      "learning_rate": 3.47561813991021e-05,
+      "loss": 1.0748,
+      "step": 3986
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.38226881065347024,
+      "learning_rate": 3.4710964474537966e-05,
+      "loss": 1.1926,
+      "step": 3987
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.45497796678058466,
+      "learning_rate": 3.466577080453909e-05,
+      "loss": 1.083,
+      "step": 3988
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5614475369968267,
+      "learning_rate": 3.462060040520244e-05,
+      "loss": 1.0471,
+      "step": 3989
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.43502275729235157,
+      "learning_rate": 3.457545329261681e-05,
+      "loss": 1.0794,
+      "step": 3990
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.3813121451220138,
+      "learning_rate": 3.4530329482862587e-05,
+      "loss": 1.246,
+      "step": 3991
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.46860697350215597,
+      "learning_rate": 3.448522899201198e-05,
+      "loss": 1.1295,
+      "step": 3992
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.3930484172822048,
+      "learning_rate": 3.444015183612876e-05,
+      "loss": 1.2341,
+      "step": 3993
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.434318551987863,
+      "learning_rate": 3.439509803126848e-05,
+      "loss": 1.1863,
+      "step": 3994
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.39928617630282415,
+      "learning_rate": 3.4350067593478356e-05,
+      "loss": 1.2098,
+      "step": 3995
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5123365062787516,
+      "learning_rate": 3.4305060538797206e-05,
+      "loss": 1.171,
+      "step": 3996
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.40718224082345744,
+      "learning_rate": 3.426007688325566e-05,
+      "loss": 1.1334,
+      "step": 3997
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.4332160954007031,
+      "learning_rate": 3.42151166428759e-05,
+      "loss": 1.3806,
+      "step": 3998
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.4296262868180955,
+      "learning_rate": 3.4170179833671846e-05,
+      "loss": 1.204,
+      "step": 3999
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.38488750955643825,
+      "learning_rate": 3.4125266471648976e-05,
+      "loss": 1.1676,
+      "step": 4000
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.4592153945598863,
+      "learning_rate": 3.408037657280454e-05,
+      "loss": 1.1305,
+      "step": 4001
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.3903529030851228,
+      "learning_rate": 3.403551015312734e-05,
+      "loss": 1.1043,
+      "step": 4002
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.4661440986211775,
+      "learning_rate": 3.399066722859782e-05,
+      "loss": 1.2312,
+      "step": 4003
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.3855576160634624,
+      "learning_rate": 3.394584781518805e-05,
+      "loss": 1.0411,
+      "step": 4004
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.4202083942604936,
+      "learning_rate": 3.390105192886182e-05,
+      "loss": 1.099,
+      "step": 4005
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.4661471646734709,
+      "learning_rate": 3.385627958557445e-05,
+      "loss": 1.0689,
+      "step": 4006
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.47187805680397504,
+      "learning_rate": 3.381153080127284e-05,
+      "loss": 1.1155,
+      "step": 4007
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.48199760850175044,
+      "learning_rate": 3.3766805591895636e-05,
+      "loss": 1.0336,
+      "step": 4008
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.4259182259194616,
+      "learning_rate": 3.372210397337294e-05,
+      "loss": 1.1099,
+      "step": 4009
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.3802043349533506,
+      "learning_rate": 3.367742596162661e-05,
+      "loss": 1.1378,
+      "step": 4010
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.419614130484787,
+      "learning_rate": 3.363277157256988e-05,
+      "loss": 1.141,
+      "step": 4011
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.38220266707880685,
+      "learning_rate": 3.3588140822107784e-05,
+      "loss": 1.1853,
+      "step": 4012
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.4184810419907848,
+      "learning_rate": 3.354353372613678e-05,
+      "loss": 1.1645,
+      "step": 4013
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.3986196302751116,
+      "learning_rate": 3.349895030054504e-05,
+      "loss": 1.0706,
+      "step": 4014
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.4230781324005517,
+      "learning_rate": 3.345439056121216e-05,
+      "loss": 1.0499,
+      "step": 4015
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.42496924447353496,
+      "learning_rate": 3.340985452400943e-05,
+      "loss": 1.1819,
+      "step": 4016
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.42238053186976854,
+      "learning_rate": 3.336534220479961e-05,
+      "loss": 1.1563,
+      "step": 4017
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.4511848787764692,
+      "learning_rate": 3.332085361943705e-05,
+      "loss": 1.0901,
+      "step": 4018
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.4893972674756453,
+      "learning_rate": 3.327638878376764e-05,
+      "loss": 1.1113,
+      "step": 4019
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.4367115275048279,
+      "learning_rate": 3.323194771362875e-05,
+      "loss": 1.1253,
+      "step": 4020
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.36522580798540805,
+      "learning_rate": 3.318753042484944e-05,
+      "loss": 1.143,
+      "step": 4021
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.47646082431813547,
+      "learning_rate": 3.3143136933250117e-05,
+      "loss": 1.0595,
+      "step": 4022
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.39901865870293135,
+      "learning_rate": 3.3098767254642883e-05,
+      "loss": 1.0896,
+      "step": 4023
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6393175815072265,
+      "learning_rate": 3.305442140483118e-05,
+      "loss": 1.039,
+      "step": 4024
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.45226893875201846,
+      "learning_rate": 3.301009939961014e-05,
+      "loss": 1.0873,
+      "step": 4025
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.398493051982934,
+      "learning_rate": 3.296580125476629e-05,
+      "loss": 1.1211,
+      "step": 4026
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.4308974007250614,
+      "learning_rate": 3.292152698607768e-05,
+      "loss": 1.1667,
+      "step": 4027
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.4452552985675732,
+      "learning_rate": 3.287727660931385e-05,
+      "loss": 1.0662,
+      "step": 4028
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.3913060560840543,
+      "learning_rate": 3.2833050140235824e-05,
+      "loss": 1.1891,
+      "step": 4029
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.44178897027947944,
+      "learning_rate": 3.278884759459619e-05,
+      "loss": 1.0736,
+      "step": 4030
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.38939773876669836,
+      "learning_rate": 3.27446689881389e-05,
+      "loss": 1.1786,
+      "step": 4031
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.444456044279783,
+      "learning_rate": 3.2700514336599483e-05,
+      "loss": 1.1572,
+      "step": 4032
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.4345300076441374,
+      "learning_rate": 3.2656383655704835e-05,
+      "loss": 1.2017,
+      "step": 4033
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.4255713973460075,
+      "learning_rate": 3.2612276961173416e-05,
+      "loss": 1.1013,
+      "step": 4034
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.48563093811952845,
+      "learning_rate": 3.256819426871507e-05,
+      "loss": 1.0729,
+      "step": 4035
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.438488466552937,
+      "learning_rate": 3.2524135594031116e-05,
+      "loss": 1.1114,
+      "step": 4036
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.4619285948916418,
+      "learning_rate": 3.2480100952814305e-05,
+      "loss": 1.0502,
+      "step": 4037
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.44050502477024706,
+      "learning_rate": 3.243609036074882e-05,
+      "loss": 1.1631,
+      "step": 4038
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.46020012389741977,
+      "learning_rate": 3.239210383351034e-05,
+      "loss": 1.1585,
+      "step": 4039
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.421806940031608,
+      "learning_rate": 3.23481413867659e-05,
+      "loss": 1.0923,
+      "step": 4040
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.39604516506771587,
+      "learning_rate": 3.2304203036174016e-05,
+      "loss": 1.1003,
+      "step": 4041
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.44589004458717635,
+      "learning_rate": 3.22602887973846e-05,
+      "loss": 1.2161,
+      "step": 4042
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.41863005951494453,
+      "learning_rate": 3.2216398686038926e-05,
+      "loss": 1.1872,
+      "step": 4043
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.4274357891952626,
+      "learning_rate": 3.2172532717769734e-05,
+      "loss": 1.0828,
+      "step": 4044
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.4263844671169171,
+      "learning_rate": 3.212869090820116e-05,
+      "loss": 1.0467,
+      "step": 4045
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.40347263889658086,
+      "learning_rate": 3.2084873272948705e-05,
+      "loss": 0.9242,
+      "step": 4046
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.3986337338681856,
+      "learning_rate": 3.2041079827619314e-05,
+      "loss": 1.1473,
+      "step": 4047
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.49907116406457913,
+      "learning_rate": 3.199731058781127e-05,
+      "loss": 1.1253,
+      "step": 4048
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.43375344614135525,
+      "learning_rate": 3.195356556911418e-05,
+      "loss": 1.2115,
+      "step": 4049
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.433722545165402,
+      "learning_rate": 3.1909844787109186e-05,
+      "loss": 1.2091,
+      "step": 4050
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.465547110449235,
+      "learning_rate": 3.1866148257368665e-05,
+      "loss": 1.1272,
+      "step": 4051
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.4150696808892286,
+      "learning_rate": 3.182247599545637e-05,
+      "loss": 1.1268,
+      "step": 4052
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.4153695534880818,
+      "learning_rate": 3.177882801692742e-05,
+      "loss": 1.0742,
+      "step": 4053
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.40261056238977666,
+      "learning_rate": 3.173520433732834e-05,
+      "loss": 1.2431,
+      "step": 4054
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5104225474160486,
+      "learning_rate": 3.169160497219692e-05,
+      "loss": 1.0966,
+      "step": 4055
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.4143871078259418,
+      "learning_rate": 3.164802993706236e-05,
+      "loss": 1.1033,
+      "step": 4056
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.41715536011044607,
+      "learning_rate": 3.160447924744512e-05,
+      "loss": 0.9822,
+      "step": 4057
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.46176010570015324,
+      "learning_rate": 3.156095291885708e-05,
+      "loss": 1.1843,
+      "step": 4058
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.4319593986887865,
+      "learning_rate": 3.15174509668014e-05,
+      "loss": 0.94,
+      "step": 4059
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.4429267833568218,
+      "learning_rate": 3.147397340677244e-05,
+      "loss": 1.2408,
+      "step": 4060
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.503748312394986,
+      "learning_rate": 3.14305202542561e-05,
+      "loss": 1.2605,
+      "step": 4061
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.43197525980747337,
+      "learning_rate": 3.138709152472938e-05,
+      "loss": 0.9677,
+      "step": 4062
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.3901097182032211,
+      "learning_rate": 3.1343687233660754e-05,
+      "loss": 1.1884,
+      "step": 4063
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.4448608449533428,
+      "learning_rate": 3.130030739650983e-05,
+      "loss": 1.1309,
+      "step": 4064
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.4177774849347701,
+      "learning_rate": 3.125695202872765e-05,
+      "loss": 1.0396,
+      "step": 4065
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.40274746703248404,
+      "learning_rate": 3.1213621145756447e-05,
+      "loss": 1.1204,
+      "step": 4066
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.40214318396942145,
+      "learning_rate": 3.117031476302975e-05,
+      "loss": 1.3253,
+      "step": 4067
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.46377190375435523,
+      "learning_rate": 3.112703289597234e-05,
+      "loss": 1.0915,
+      "step": 4068
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.4069771851988625,
+      "learning_rate": 3.108377556000037e-05,
+      "loss": 1.2634,
+      "step": 4069
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.43821198679690837,
+      "learning_rate": 3.1040542770521156e-05,
+      "loss": 0.9574,
+      "step": 4070
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.4797341096794143,
+      "learning_rate": 3.099733454293324e-05,
+      "loss": 1.1594,
+      "step": 4071
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.4919071819748191,
+      "learning_rate": 3.095415089262657e-05,
+      "loss": 1.0977,
+      "step": 4072
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.47477198820702643,
+      "learning_rate": 3.091099183498214e-05,
+      "loss": 0.9845,
+      "step": 4073
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.42554364047657134,
+      "learning_rate": 3.086785738537239e-05,
+      "loss": 1.286,
+      "step": 4074
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.45877906827911696,
+      "learning_rate": 3.0824747559160836e-05,
+      "loss": 1.0395,
+      "step": 4075
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.404695853534555,
+      "learning_rate": 3.078166237170229e-05,
+      "loss": 1.0598,
+      "step": 4076
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5226875308676264,
+      "learning_rate": 3.073860183834276e-05,
+      "loss": 1.1434,
+      "step": 4077
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.4374918516755399,
+      "learning_rate": 3.069556597441953e-05,
+      "loss": 1.2178,
+      "step": 4078
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.4791242340241139,
+      "learning_rate": 3.065255479526101e-05,
+      "loss": 1.0109,
+      "step": 4079
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.4003042807933513,
+      "learning_rate": 3.0609568316186935e-05,
+      "loss": 0.969,
+      "step": 4080
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.43658395756595925,
+      "learning_rate": 3.056660655250813e-05,
+      "loss": 1.1847,
+      "step": 4081
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.44034184965792117,
+      "learning_rate": 3.052366951952663e-05,
+      "loss": 0.9366,
+      "step": 4082
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.42069594347239375,
+      "learning_rate": 3.0480757232535772e-05,
+      "loss": 1.065,
+      "step": 4083
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.4375310103348434,
+      "learning_rate": 3.0437869706819955e-05,
+      "loss": 1.1593,
+      "step": 4084
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.4269386260773027,
+      "learning_rate": 3.0395006957654802e-05,
+      "loss": 1.0414,
+      "step": 4085
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.4166296584124347,
+      "learning_rate": 3.0352169000307086e-05,
+      "loss": 1.0693,
+      "step": 4086
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.38260884630725,
+      "learning_rate": 3.0309355850034837e-05,
+      "loss": 1.2478,
+      "step": 4087
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.41553237645165136,
+      "learning_rate": 3.026656752208712e-05,
+      "loss": 1.2717,
+      "step": 4088
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.4743063430568139,
+      "learning_rate": 3.0223804031704295e-05,
+      "loss": 1.1471,
+      "step": 4089
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.4595108938689279,
+      "learning_rate": 3.018106539411778e-05,
+      "loss": 1.185,
+      "step": 4090
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.3970594372191495,
+      "learning_rate": 3.0138351624550164e-05,
+      "loss": 1.0338,
+      "step": 4091
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.44425114378410846,
+      "learning_rate": 3.009566273821518e-05,
+      "loss": 1.0877,
+      "step": 4092
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.38609119127111075,
+      "learning_rate": 3.0052998750317675e-05,
+      "loss": 1.004,
+      "step": 4093
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.44785475759382043,
+      "learning_rate": 3.0010359676053713e-05,
+      "loss": 1.0686,
+      "step": 4094
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.38211271353341403,
+      "learning_rate": 2.9967745530610357e-05,
+      "loss": 1.2403,
+      "step": 4095
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.4269559602627749,
+      "learning_rate": 2.9925156329165937e-05,
+      "loss": 1.1653,
+      "step": 4096
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.41632595292503627,
+      "learning_rate": 2.9882592086889736e-05,
+      "loss": 1.1402,
+      "step": 4097
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.4022425932050193,
+      "learning_rate": 2.984005281894232e-05,
+      "loss": 1.0723,
+      "step": 4098
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.40316607030965385,
+      "learning_rate": 2.979753854047522e-05,
+      "loss": 1.2718,
+      "step": 4099
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.4494848833629546,
+      "learning_rate": 2.975504926663113e-05,
+      "loss": 1.2359,
+      "step": 4100
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.40391588500402575,
+      "learning_rate": 2.9712585012543816e-05,
+      "loss": 1.112,
+      "step": 4101
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.42685060119358104,
+      "learning_rate": 2.967014579333811e-05,
+      "loss": 1.1175,
+      "step": 4102
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.48179165610973784,
+      "learning_rate": 2.9627731624130016e-05,
+      "loss": 1.0591,
+      "step": 4103
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5137821631883573,
+      "learning_rate": 2.9585342520026514e-05,
+      "loss": 0.9376,
+      "step": 4104
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.4163579801469035,
+      "learning_rate": 2.9542978496125762e-05,
+      "loss": 1.1393,
+      "step": 4105
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.45795710164479797,
+      "learning_rate": 2.9500639567516842e-05,
+      "loss": 1.2249,
+      "step": 4106
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.41287038498247414,
+      "learning_rate": 2.9458325749280057e-05,
+      "loss": 1.1309,
+      "step": 4107
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.4974910419520055,
+      "learning_rate": 2.9416037056486658e-05,
+      "loss": 1.0425,
+      "step": 4108
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.38918073632302896,
+      "learning_rate": 2.9373773504198963e-05,
+      "loss": 1.2437,
+      "step": 4109
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5123830329531147,
+      "learning_rate": 2.9331535107470342e-05,
+      "loss": 1.1997,
+      "step": 4110
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.3996005810290748,
+      "learning_rate": 2.9289321881345254e-05,
+      "loss": 1.191,
+      "step": 4111
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.4656189527956771,
+      "learning_rate": 2.924713384085913e-05,
+      "loss": 1.0284,
+      "step": 4112
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.4540328626677855,
+      "learning_rate": 2.9204971001038428e-05,
+      "loss": 0.8967,
+      "step": 4113
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.4111794987855104,
+      "learning_rate": 2.91628333769007e-05,
+      "loss": 1.0129,
+      "step": 4114
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.41583247877262913,
+      "learning_rate": 2.9120720983454463e-05,
+      "loss": 1.0257,
+      "step": 4115
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.38776338329844073,
+      "learning_rate": 2.9078633835699233e-05,
+      "loss": 1.016,
+      "step": 4116
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.4922548262875228,
+      "learning_rate": 2.9036571948625537e-05,
+      "loss": 1.107,
+      "step": 4117
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.4066772986646165,
+      "learning_rate": 2.8994535337214978e-05,
+      "loss": 1.182,
+      "step": 4118
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.4903880626980828,
+      "learning_rate": 2.8952524016440052e-05,
+      "loss": 1.0146,
+      "step": 4119
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5725484824790122,
+      "learning_rate": 2.8910538001264354e-05,
+      "loss": 1.1675,
+      "step": 4120
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.41091165871276414,
+      "learning_rate": 2.8868577306642343e-05,
+      "loss": 1.1219,
+      "step": 4121
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.4628282904612589,
+      "learning_rate": 2.8826641947519595e-05,
+      "loss": 1.1286,
+      "step": 4122
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5835612525715682,
+      "learning_rate": 2.8784731938832556e-05,
+      "loss": 1.1139,
+      "step": 4123
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.4175595921608017,
+      "learning_rate": 2.8742847295508678e-05,
+      "loss": 1.1487,
+      "step": 4124
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.4541692176169396,
+      "learning_rate": 2.8700988032466392e-05,
+      "loss": 1.1972,
+      "step": 4125
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.41644597458442084,
+      "learning_rate": 2.865915416461503e-05,
+      "loss": 1.1602,
+      "step": 4126
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.44007305673107383,
+      "learning_rate": 2.8617345706855003e-05,
+      "loss": 1.0928,
+      "step": 4127
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.4952564302371592,
+      "learning_rate": 2.8575562674077517e-05,
+      "loss": 1.1224,
+      "step": 4128
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.4346947936712108,
+      "learning_rate": 2.8533805081164867e-05,
+      "loss": 1.188,
+      "step": 4129
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5666101222277315,
+      "learning_rate": 2.8492072942990157e-05,
+      "loss": 1.2752,
+      "step": 4130
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.466650550087827,
+      "learning_rate": 2.845036627441755e-05,
+      "loss": 1.1477,
+      "step": 4131
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.43683340346721045,
+      "learning_rate": 2.8408685090302044e-05,
+      "loss": 1.1493,
+      "step": 4132
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.46343898188151134,
+      "learning_rate": 2.8367029405489597e-05,
+      "loss": 0.9962,
+      "step": 4133
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.44542218787965804,
+      "learning_rate": 2.8325399234817064e-05,
+      "loss": 1.0075,
+      "step": 4134
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.409532916581491,
+      "learning_rate": 2.828379459311219e-05,
+      "loss": 1.0741,
+      "step": 4135
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.4033069144059048,
+      "learning_rate": 2.8242215495193746e-05,
+      "loss": 1.0766,
+      "step": 4136
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.4437405433078432,
+      "learning_rate": 2.8200661955871242e-05,
+      "loss": 1.095,
+      "step": 4137
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.4817770538064987,
+      "learning_rate": 2.815913398994524e-05,
+      "loss": 1.1045,
+      "step": 4138
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.38838363579020396,
+      "learning_rate": 2.8117631612207084e-05,
+      "loss": 1.2624,
+      "step": 4139
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.43033117899530215,
+      "learning_rate": 2.8076154837439027e-05,
+      "loss": 1.1371,
+      "step": 4140
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.38474556186595615,
+      "learning_rate": 2.8034703680414186e-05,
+      "loss": 1.3902,
+      "step": 4141
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.49887412691258776,
+      "learning_rate": 2.7993278155896653e-05,
+      "loss": 1.1449,
+      "step": 4142
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.38462634560319975,
+      "learning_rate": 2.7951878278641285e-05,
+      "loss": 1.0618,
+      "step": 4143
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.4698306973541561,
+      "learning_rate": 2.7910504063393793e-05,
+      "loss": 1.1069,
+      "step": 4144
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5061521402295021,
+      "learning_rate": 2.786915552489088e-05,
+      "loss": 1.0458,
+      "step": 4145
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.47443251616469634,
+      "learning_rate": 2.782783267785992e-05,
+      "loss": 1.1068,
+      "step": 4146
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.4311719840513317,
+      "learning_rate": 2.778653553701932e-05,
+      "loss": 1.2088,
+      "step": 4147
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.4605409709013759,
+      "learning_rate": 2.7745264117078207e-05,
+      "loss": 1.1087,
+      "step": 4148
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.4653403140523103,
+      "learning_rate": 2.7704018432736566e-05,
+      "loss": 0.9944,
+      "step": 4149
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.44829956310126934,
+      "learning_rate": 2.7662798498685227e-05,
+      "loss": 0.8808,
+      "step": 4150
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.42429272860354844,
+      "learning_rate": 2.7621604329605898e-05,
+      "loss": 1.1233,
+      "step": 4151
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.4269359743150881,
+      "learning_rate": 2.7580435940171e-05,
+      "loss": 1.0414,
+      "step": 4152
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.4697371816378722,
+      "learning_rate": 2.7539293345043926e-05,
+      "loss": 1.2241,
+      "step": 4153
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.4123834667690057,
+      "learning_rate": 2.749817655887873e-05,
+      "loss": 1.1244,
+      "step": 4154
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.440884080608528,
+      "learning_rate": 2.745708559632032e-05,
+      "loss": 1.0931,
+      "step": 4155
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.45395243280984937,
+      "learning_rate": 2.741602047200448e-05,
+      "loss": 1.0106,
+      "step": 4156
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.4920952361363041,
+      "learning_rate": 2.7374981200557703e-05,
+      "loss": 1.0846,
+      "step": 4157
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.47232285314485867,
+      "learning_rate": 2.7333967796597315e-05,
+      "loss": 1.142,
+      "step": 4158
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.3969024976247107,
+      "learning_rate": 2.7292980274731382e-05,
+      "loss": 1.2364,
+      "step": 4159
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.4435943862769823,
+      "learning_rate": 2.7252018649558842e-05,
+      "loss": 1.0867,
+      "step": 4160
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.4574928574402072,
+      "learning_rate": 2.7211082935669285e-05,
+      "loss": 1.0855,
+      "step": 4161
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.4121306109288966,
+      "learning_rate": 2.717017314764323e-05,
+      "loss": 0.9828,
+      "step": 4162
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.45360628764081634,
+      "learning_rate": 2.7129289300051787e-05,
+      "loss": 0.9753,
+      "step": 4163
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.41036120820903255,
+      "learning_rate": 2.7088431407457015e-05,
+      "loss": 1.1167,
+      "step": 4164
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5235066452469641,
+      "learning_rate": 2.7047599484411535e-05,
+      "loss": 1.1896,
+      "step": 4165
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.4264882816880962,
+      "learning_rate": 2.7006793545458807e-05,
+      "loss": 1.2828,
+      "step": 4166
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.45521985530630954,
+      "learning_rate": 2.6966013605133088e-05,
+      "loss": 1.235,
+      "step": 4167
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.4255400931165722,
+      "learning_rate": 2.692525967795929e-05,
+      "loss": 1.175,
+      "step": 4168
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.4353162467550098,
+      "learning_rate": 2.6884531778453126e-05,
+      "loss": 1.1837,
+      "step": 4169
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.45894293298499705,
+      "learning_rate": 2.684382992112098e-05,
+      "loss": 1.1659,
+      "step": 4170
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.43719269544031136,
+      "learning_rate": 2.6803154120460007e-05,
+      "loss": 1.136,
+      "step": 4171
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.44103330768107535,
+      "learning_rate": 2.6762504390958066e-05,
+      "loss": 1.1762,
+      "step": 4172
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5027002748947255,
+      "learning_rate": 2.672188074709371e-05,
+      "loss": 1.2577,
+      "step": 4173
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.43947095667272423,
+      "learning_rate": 2.6681283203336193e-05,
+      "loss": 1.1587,
+      "step": 4174
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.45264083617129885,
+      "learning_rate": 2.6640711774145543e-05,
+      "loss": 1.0928,
+      "step": 4175
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.467192107718259,
+      "learning_rate": 2.660016647397243e-05,
+      "loss": 1.0328,
+      "step": 4176
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.48702408449188755,
+      "learning_rate": 2.6559647317258183e-05,
+      "loss": 0.9693,
+      "step": 4177
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.4479903438720699,
+      "learning_rate": 2.6519154318434934e-05,
+      "loss": 1.0245,
+      "step": 4178
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.39124966089723445,
+      "learning_rate": 2.6478687491925357e-05,
+      "loss": 1.1034,
+      "step": 4179
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.37214681497092084,
+      "learning_rate": 2.6438246852142934e-05,
+      "loss": 1.1964,
+      "step": 4180
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.4981201304794601,
+      "learning_rate": 2.639783241349174e-05,
+      "loss": 1.083,
+      "step": 4181
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.4499960482028722,
+      "learning_rate": 2.635744419036653e-05,
+      "loss": 1.0187,
+      "step": 4182
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.44991671606806477,
+      "learning_rate": 2.63170821971527e-05,
+      "loss": 0.9964,
+      "step": 4183
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.4610173767336911,
+      "learning_rate": 2.6276746448226398e-05,
+      "loss": 0.989,
+      "step": 4184
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.42288552687368647,
+      "learning_rate": 2.6236436957954304e-05,
+      "loss": 1.155,
+      "step": 4185
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.42552634256890026,
+      "learning_rate": 2.6196153740693842e-05,
+      "loss": 1.2119,
+      "step": 4186
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.43434586433169053,
+      "learning_rate": 2.6155896810793036e-05,
+      "loss": 0.8908,
+      "step": 4187
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.4941413411481656,
+      "learning_rate": 2.6115666182590513e-05,
+      "loss": 1.075,
+      "step": 4188
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.43405636559119476,
+      "learning_rate": 2.6075461870415595e-05,
+      "loss": 1.0051,
+      "step": 4189
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.45997045660667063,
+      "learning_rate": 2.6035283888588158e-05,
+      "loss": 0.9418,
+      "step": 4190
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.4340240029897283,
+      "learning_rate": 2.599513225141881e-05,
+      "loss": 1.0469,
+      "step": 4191
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.49699367022186225,
+      "learning_rate": 2.595500697320865e-05,
+      "loss": 1.1015,
+      "step": 4192
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.42772544372510374,
+      "learning_rate": 2.59149080682495e-05,
+      "loss": 1.2164,
+      "step": 4193
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.4398044003442902,
+      "learning_rate": 2.5874835550823683e-05,
+      "loss": 1.1889,
+      "step": 4194
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.4168345063951611,
+      "learning_rate": 2.5834789435204243e-05,
+      "loss": 1.1741,
+      "step": 4195
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.501955511372995,
+      "learning_rate": 2.5794769735654723e-05,
+      "loss": 1.2018,
+      "step": 4196
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.44038689723497,
+      "learning_rate": 2.5754776466429274e-05,
+      "loss": 1.1866,
+      "step": 4197
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.4027751097048484,
+      "learning_rate": 2.5714809641772664e-05,
+      "loss": 1.0541,
+      "step": 4198
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.4072228088645122,
+      "learning_rate": 2.567486927592018e-05,
+      "loss": 1.1435,
+      "step": 4199
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.4319057826839221,
+      "learning_rate": 2.5634955383097804e-05,
+      "loss": 0.8963,
+      "step": 4200
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.3952894126891799,
+      "learning_rate": 2.559506797752196e-05,
+      "loss": 1.1687,
+      "step": 4201
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.38560874154895575,
+      "learning_rate": 2.5555207073399733e-05,
+      "loss": 1.1172,
+      "step": 4202
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.38331311303140525,
+      "learning_rate": 2.5515372684928683e-05,
+      "loss": 0.9779,
+      "step": 4203
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.42153494734774316,
+      "learning_rate": 2.5475564826297027e-05,
+      "loss": 1.2196,
+      "step": 4204
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.4279840276963783,
+      "learning_rate": 2.5435783511683443e-05,
+      "loss": 1.1089,
+      "step": 4205
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.37292366077500433,
+      "learning_rate": 2.5396028755257207e-05,
+      "loss": 1.2457,
+      "step": 4206
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.4261090855628609,
+      "learning_rate": 2.5356300571178094e-05,
+      "loss": 0.9995,
+      "step": 4207
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.42986667083399205,
+      "learning_rate": 2.531659897359643e-05,
+      "loss": 1.0471,
+      "step": 4208
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.4442384679109188,
+      "learning_rate": 2.527692397665311e-05,
+      "loss": 1.1692,
+      "step": 4209
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.4871043932828831,
+      "learning_rate": 2.523727559447949e-05,
+      "loss": 0.9819,
+      "step": 4210
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.40670930497600194,
+      "learning_rate": 2.5197653841197543e-05,
+      "loss": 1.1715,
+      "step": 4211
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.436800673921742,
+      "learning_rate": 2.5158058730919608e-05,
+      "loss": 0.953,
+      "step": 4212
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5250300504272307,
+      "learning_rate": 2.511849027774874e-05,
+      "loss": 1.1602,
+      "step": 4213
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.42252235037840363,
+      "learning_rate": 2.5078948495778233e-05,
+      "loss": 1.2262,
+      "step": 4214
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.38896002065853946,
+      "learning_rate": 2.503943339909214e-05,
+      "loss": 1.0928,
+      "step": 4215
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.4328975411631521,
+      "learning_rate": 2.4999945001764824e-05,
+      "loss": 1.1022,
+      "step": 4216
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.4463279488532637,
+      "learning_rate": 2.4960483317861273e-05,
+      "loss": 1.1198,
+      "step": 4217
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.4129417366557657,
+      "learning_rate": 2.4921048361436882e-05,
+      "loss": 1.2078,
+      "step": 4218
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.41938261816366285,
+      "learning_rate": 2.4881640146537498e-05,
+      "loss": 1.0048,
+      "step": 4219
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.48814695862279345,
+      "learning_rate": 2.484225868719956e-05,
+      "loss": 1.259,
+      "step": 4220
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.4566573449664895,
+      "learning_rate": 2.4802903997449865e-05,
+      "loss": 1.1833,
+      "step": 4221
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.4792573766211381,
+      "learning_rate": 2.476357609130573e-05,
+      "loss": 1.1282,
+      "step": 4222
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.41766786435351466,
+      "learning_rate": 2.4724274982774865e-05,
+      "loss": 1.2459,
+      "step": 4223
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.41554526905822914,
+      "learning_rate": 2.4685000685855576e-05,
+      "loss": 0.9731,
+      "step": 4224
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.4879878735068277,
+      "learning_rate": 2.464575321453646e-05,
+      "loss": 1.1924,
+      "step": 4225
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.4455455194429363,
+      "learning_rate": 2.4606532582796693e-05,
+      "loss": 1.1684,
+      "step": 4226
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.4423944735441804,
+      "learning_rate": 2.4567338804605756e-05,
+      "loss": 1.0115,
+      "step": 4227
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.4662744749565376,
+      "learning_rate": 2.4528171893923723e-05,
+      "loss": 0.8985,
+      "step": 4228
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.46796252612067,
+      "learning_rate": 2.448903186470096e-05,
+      "loss": 1.004,
+      "step": 4229
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.516920978692492,
+      "learning_rate": 2.444991873087833e-05,
+      "loss": 1.1054,
+      "step": 4230
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.4938510133230365,
+      "learning_rate": 2.4410832506387095e-05,
+      "loss": 1.0906,
+      "step": 4231
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.523650455281826,
+      "learning_rate": 2.4371773205148895e-05,
+      "loss": 1.0988,
+      "step": 4232
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.4324285114311776,
+      "learning_rate": 2.4332740841075908e-05,
+      "loss": 1.2014,
+      "step": 4233
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.4376389403949988,
+      "learning_rate": 2.429373542807054e-05,
+      "loss": 1.0859,
+      "step": 4234
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.4475150219713845,
+      "learning_rate": 2.4254756980025773e-05,
+      "loss": 1.1747,
+      "step": 4235
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.4665760398034787,
+      "learning_rate": 2.4215805510824828e-05,
+      "loss": 0.9775,
+      "step": 4236
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.42185527189415306,
+      "learning_rate": 2.417688103434146e-05,
+      "loss": 1.0438,
+      "step": 4237
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.42988172341682424,
+      "learning_rate": 2.413798356443969e-05,
+      "loss": 1.047,
+      "step": 4238
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.4198521885622046,
+      "learning_rate": 2.4099113114973992e-05,
+      "loss": 1.1373,
+      "step": 4239
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.3951724306070046,
+      "learning_rate": 2.4060269699789184e-05,
+      "loss": 1.0334,
+      "step": 4240
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.4508044572513338,
+      "learning_rate": 2.4021453332720422e-05,
+      "loss": 1.1249,
+      "step": 4241
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5062207544155691,
+      "learning_rate": 2.398266402759335e-05,
+      "loss": 1.1379,
+      "step": 4242
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.469625073627503,
+      "learning_rate": 2.394390179822382e-05,
+      "loss": 1.0368,
+      "step": 4243
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.48214656726447147,
+      "learning_rate": 2.3905166658418188e-05,
+      "loss": 1.1447,
+      "step": 4244
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.45681263462951205,
+      "learning_rate": 2.386645862197303e-05,
+      "loss": 1.3039,
+      "step": 4245
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.47797153891286287,
+      "learning_rate": 2.3827777702675348e-05,
+      "loss": 1.1601,
+      "step": 4246
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.4033464573607413,
+      "learning_rate": 2.3789123914302436e-05,
+      "loss": 1.1517,
+      "step": 4247
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.4016509486329039,
+      "learning_rate": 2.3750497270621986e-05,
+      "loss": 1.2941,
+      "step": 4248
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.41227603595347606,
+      "learning_rate": 2.371189778539197e-05,
+      "loss": 1.2821,
+      "step": 4249
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.487683425473845,
+      "learning_rate": 2.3673325472360687e-05,
+      "loss": 1.1512,
+      "step": 4250
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.4514628888332395,
+      "learning_rate": 2.3634780345266806e-05,
+      "loss": 1.0439,
+      "step": 4251
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.44694269768529477,
+      "learning_rate": 2.3596262417839255e-05,
+      "loss": 1.0111,
+      "step": 4252
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.4215247449751002,
+      "learning_rate": 2.3557771703797325e-05,
+      "loss": 1.1387,
+      "step": 4253
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.399996069256357,
+      "learning_rate": 2.3519308216850577e-05,
+      "loss": 1.197,
+      "step": 4254
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.4423835705209643,
+      "learning_rate": 2.348087197069889e-05,
+      "loss": 1.1078,
+      "step": 4255
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.4587315662178466,
+      "learning_rate": 2.3442462979032408e-05,
+      "loss": 1.238,
+      "step": 4256
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.4870740293450337,
+      "learning_rate": 2.3404081255531652e-05,
+      "loss": 1.0212,
+      "step": 4257
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.4892664611974982,
+      "learning_rate": 2.3365726813867304e-05,
+      "loss": 1.1093,
+      "step": 4258
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.4644350235771718,
+      "learning_rate": 2.3327399667700477e-05,
+      "loss": 1.0312,
+      "step": 4259
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.48488933519732924,
+      "learning_rate": 2.3289099830682448e-05,
+      "loss": 1.2143,
+      "step": 4260
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5388182702530805,
+      "learning_rate": 2.325082731645477e-05,
+      "loss": 1.1207,
+      "step": 4261
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.46385428530955664,
+      "learning_rate": 2.3212582138649396e-05,
+      "loss": 0.9946,
+      "step": 4262
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.48894678657520074,
+      "learning_rate": 2.317436431088832e-05,
+      "loss": 1.1161,
+      "step": 4263
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.4938695496254484,
+      "learning_rate": 2.3136173846784014e-05,
+      "loss": 1.1306,
+      "step": 4264
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.4702701025031784,
+      "learning_rate": 2.3098010759939036e-05,
+      "loss": 1.1038,
+      "step": 4265
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.458394308513747,
+      "learning_rate": 2.3059875063946347e-05,
+      "loss": 1.2774,
+      "step": 4266
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.4298628561300465,
+      "learning_rate": 2.3021766772388986e-05,
+      "loss": 1.2271,
+      "step": 4267
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5788808609256099,
+      "learning_rate": 2.298368589884039e-05,
+      "loss": 0.9901,
+      "step": 4268
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5063264053914709,
+      "learning_rate": 2.294563245686413e-05,
+      "loss": 1.0679,
+      "step": 4269
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.45078651717885077,
+      "learning_rate": 2.290760646001402e-05,
+      "loss": 1.1766,
+      "step": 4270
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.42101033391623904,
+      "learning_rate": 2.2869607921834113e-05,
+      "loss": 1.0465,
+      "step": 4271
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.4401398894604397,
+      "learning_rate": 2.2831636855858663e-05,
+      "loss": 1.1741,
+      "step": 4272
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.45379401090478333,
+      "learning_rate": 2.2793693275612204e-05,
+      "loss": 1.1313,
+      "step": 4273
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5175052939855869,
+      "learning_rate": 2.2755777194609374e-05,
+      "loss": 0.99,
+      "step": 4274
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.4234573280667234,
+      "learning_rate": 2.2717888626355134e-05,
+      "loss": 1.1642,
+      "step": 4275
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.4346797212595524,
+      "learning_rate": 2.268002758434452e-05,
+      "loss": 1.1894,
+      "step": 4276
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.3898789755060201,
+      "learning_rate": 2.2642194082062895e-05,
+      "loss": 1.1483,
+      "step": 4277
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.420057128490195,
+      "learning_rate": 2.260438813298572e-05,
+      "loss": 0.9899,
+      "step": 4278
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.4238404174943811,
+      "learning_rate": 2.2566609750578673e-05,
+      "loss": 1.1213,
+      "step": 4279
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.4402309762518351,
+      "learning_rate": 2.252885894829757e-05,
+      "loss": 1.0625,
+      "step": 4280
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.4013874464731777,
+      "learning_rate": 2.249113573958852e-05,
+      "loss": 1.1953,
+      "step": 4281
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.38741131689506264,
+      "learning_rate": 2.245344013788767e-05,
+      "loss": 1.0272,
+      "step": 4282
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5454501648212262,
+      "learning_rate": 2.2415772156621382e-05,
+      "loss": 1.0644,
+      "step": 4283
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.4229399132269001,
+      "learning_rate": 2.2378131809206248e-05,
+      "loss": 1.158,
+      "step": 4284
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.4553490773122571,
+      "learning_rate": 2.2340519109048898e-05,
+      "loss": 1.2712,
+      "step": 4285
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.43378841019986636,
+      "learning_rate": 2.2302934069546222e-05,
+      "loss": 1.1184,
+      "step": 4286
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.39482974494368706,
+      "learning_rate": 2.2265376704085193e-05,
+      "loss": 1.0933,
+      "step": 4287
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.458593966585598,
+      "learning_rate": 2.2227847026042946e-05,
+      "loss": 1.0413,
+      "step": 4288
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.40826032054544326,
+      "learning_rate": 2.2190345048786733e-05,
+      "loss": 1.1899,
+      "step": 4289
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.49456722609554166,
+      "learning_rate": 2.2152870785673996e-05,
+      "loss": 1.0385,
+      "step": 4290
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.45743846111140773,
+      "learning_rate": 2.211542425005223e-05,
+      "loss": 1.1416,
+      "step": 4291
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.4486485086014614,
+      "learning_rate": 2.2078005455259143e-05,
+      "loss": 0.9809,
+      "step": 4292
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.49383115756032037,
+      "learning_rate": 2.2040614414622484e-05,
+      "loss": 1.1076,
+      "step": 4293
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.39841677689050875,
+      "learning_rate": 2.2003251141460156e-05,
+      "loss": 1.1805,
+      "step": 4294
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.42716280899956927,
+      "learning_rate": 2.196591564908016e-05,
+      "loss": 1.1438,
+      "step": 4295
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5272570866132619,
+      "learning_rate": 2.1928607950780566e-05,
+      "loss": 1.0509,
+      "step": 4296
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.4348868262010981,
+      "learning_rate": 2.189132805984965e-05,
+      "loss": 0.9762,
+      "step": 4297
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.46573552203049423,
+      "learning_rate": 2.1854075989565658e-05,
+      "loss": 1.179,
+      "step": 4298
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.44561711064043313,
+      "learning_rate": 2.181685175319702e-05,
+      "loss": 0.9832,
+      "step": 4299
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.4809469671329899,
+      "learning_rate": 2.1779655364002195e-05,
+      "loss": 1.0,
+      "step": 4300
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.4444044128938168,
+      "learning_rate": 2.174248683522978e-05,
+      "loss": 1.0562,
+      "step": 4301
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.4453681719868396,
+      "learning_rate": 2.170534618011839e-05,
+      "loss": 1.1337,
+      "step": 4302
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.4636185967808964,
+      "learning_rate": 2.1668233411896734e-05,
+      "loss": 1.2835,
+      "step": 4303
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.4112744726127679,
+      "learning_rate": 2.1631148543783597e-05,
+      "loss": 0.9991,
+      "step": 4304
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.45872823140947816,
+      "learning_rate": 2.1594091588987774e-05,
+      "loss": 1.1548,
+      "step": 4305
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.39288213616129647,
+      "learning_rate": 2.1557062560708252e-05,
+      "loss": 1.174,
+      "step": 4306
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.42013761360676505,
+      "learning_rate": 2.1520061472133902e-05,
+      "loss": 1.1226,
+      "step": 4307
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.46952822665616145,
+      "learning_rate": 2.148308833644378e-05,
+      "loss": 1.1402,
+      "step": 4308
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.40555674672370606,
+      "learning_rate": 2.14461431668069e-05,
+      "loss": 1.1029,
+      "step": 4309
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.534543834333514,
+      "learning_rate": 2.1409225976382363e-05,
+      "loss": 1.0385,
+      "step": 4310
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.4448342600148856,
+      "learning_rate": 2.137233677831929e-05,
+      "loss": 1.2221,
+      "step": 4311
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.4379566999453861,
+      "learning_rate": 2.1335475585756803e-05,
+      "loss": 1.1124,
+      "step": 4312
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.4040624313938186,
+      "learning_rate": 2.12986424118241e-05,
+      "loss": 1.1244,
+      "step": 4313
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.4328029803717822,
+      "learning_rate": 2.126183726964034e-05,
+      "loss": 1.1743,
+      "step": 4314
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.48491479267312987,
+      "learning_rate": 2.122506017231477e-05,
+      "loss": 1.0188,
+      "step": 4315
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.44474263713278656,
+      "learning_rate": 2.118831113294657e-05,
+      "loss": 1.0215,
+      "step": 4316
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.4401761709628928,
+      "learning_rate": 2.1151590164625012e-05,
+      "loss": 1.0949,
+      "step": 4317
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5029665623166237,
+      "learning_rate": 2.1114897280429304e-05,
+      "loss": 1.0139,
+      "step": 4318
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.40758327176390746,
+      "learning_rate": 2.1078232493428664e-05,
+      "loss": 1.2527,
+      "step": 4319
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.39522198144422344,
+      "learning_rate": 2.104159581668229e-05,
+      "loss": 1.1632,
+      "step": 4320
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.44259262148970235,
+      "learning_rate": 2.100498726323943e-05,
+      "loss": 1.094,
+      "step": 4321
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.4883282517611793,
+      "learning_rate": 2.0968406846139233e-05,
+      "loss": 1.3193,
+      "step": 4322
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.48155743763361375,
+      "learning_rate": 2.0931854578410905e-05,
+      "loss": 0.9413,
+      "step": 4323
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.509608966538246,
+      "learning_rate": 2.089533047307356e-05,
+      "loss": 1.1709,
+      "step": 4324
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.44108729284603027,
+      "learning_rate": 2.0858834543136286e-05,
+      "loss": 1.0798,
+      "step": 4325
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.49400961998536913,
+      "learning_rate": 2.0822366801598204e-05,
+      "loss": 1.1505,
+      "step": 4326
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.4400932280902056,
+      "learning_rate": 2.0785927261448335e-05,
+      "loss": 1.1791,
+      "step": 4327
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.4464955068877079,
+      "learning_rate": 2.0749515935665654e-05,
+      "loss": 1.1581,
+      "step": 4328
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.4673015639724096,
+      "learning_rate": 2.0713132837219075e-05,
+      "loss": 1.0145,
+      "step": 4329
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.4069635396696683,
+      "learning_rate": 2.067677797906754e-05,
+      "loss": 1.0769,
+      "step": 4330
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.4173783871724525,
+      "learning_rate": 2.064045137415982e-05,
+      "loss": 1.2368,
+      "step": 4331
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.40912670666324025,
+      "learning_rate": 2.0604153035434735e-05,
+      "loss": 1.1811,
+      "step": 4332
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.489668229840435,
+      "learning_rate": 2.056788297582093e-05,
+      "loss": 1.0743,
+      "step": 4333
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.4554298700904282,
+      "learning_rate": 2.0531641208237084e-05,
+      "loss": 1.1053,
+      "step": 4334
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.4941995688389929,
+      "learning_rate": 2.04954277455917e-05,
+      "loss": 1.2817,
+      "step": 4335
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.4614307002489454,
+      "learning_rate": 2.045924260078327e-05,
+      "loss": 0.9252,
+      "step": 4336
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5022035836425526,
+      "learning_rate": 2.0423085786700135e-05,
+      "loss": 0.9983,
+      "step": 4337
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.4291271969700389,
+      "learning_rate": 2.03869573162206e-05,
+      "loss": 1.1306,
+      "step": 4338
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.42578153921222955,
+      "learning_rate": 2.035085720221288e-05,
+      "loss": 1.0154,
+      "step": 4339
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.4572791739828534,
+      "learning_rate": 2.0314785457535014e-05,
+      "loss": 1.1362,
+      "step": 4340
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.3804092578573056,
+      "learning_rate": 2.0278742095035065e-05,
+      "loss": 1.1703,
+      "step": 4341
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.4330412319485793,
+      "learning_rate": 2.0242727127550865e-05,
+      "loss": 1.0099,
+      "step": 4342
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.38732753334621317,
+      "learning_rate": 2.0206740567910187e-05,
+      "loss": 1.0299,
+      "step": 4343
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.4002619831576191,
+      "learning_rate": 2.0170782428930658e-05,
+      "loss": 1.0951,
+      "step": 4344
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.45297468873864766,
+      "learning_rate": 2.013485272341984e-05,
+      "loss": 1.0422,
+      "step": 4345
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5290422009690384,
+      "learning_rate": 2.009895146417512e-05,
+      "loss": 1.0405,
+      "step": 4346
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.4422846102094539,
+      "learning_rate": 2.0063078663983714e-05,
+      "loss": 1.0384,
+      "step": 4347
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.41937869613425666,
+      "learning_rate": 2.0027234335622826e-05,
+      "loss": 0.9499,
+      "step": 4348
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5856233588031869,
+      "learning_rate": 1.999141849185938e-05,
+      "loss": 1.148,
+      "step": 4349
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.45871385148157356,
+      "learning_rate": 1.9955631145450272e-05,
+      "loss": 1.2497,
+      "step": 4350
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.4196891146266073,
+      "learning_rate": 1.9919872309142152e-05,
+      "loss": 1.0351,
+      "step": 4351
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.3973608315761954,
+      "learning_rate": 1.988414199567158e-05,
+      "loss": 1.1224,
+      "step": 4352
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.4777025733133256,
+      "learning_rate": 1.9848440217764897e-05,
+      "loss": 1.015,
+      "step": 4353
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.42923837436340107,
+      "learning_rate": 1.9812766988138375e-05,
+      "loss": 1.0748,
+      "step": 4354
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.4217197868673152,
+      "learning_rate": 1.9777122319497986e-05,
+      "loss": 1.0658,
+      "step": 4355
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.44585180166942,
+      "learning_rate": 1.9741506224539687e-05,
+      "loss": 1.0147,
+      "step": 4356
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.45491016516580046,
+      "learning_rate": 1.9705918715949123e-05,
+      "loss": 1.084,
+      "step": 4357
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.4783432022693347,
+      "learning_rate": 1.9670359806401785e-05,
+      "loss": 1.1155,
+      "step": 4358
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.4747081617357751,
+      "learning_rate": 1.9634829508563057e-05,
+      "loss": 0.9557,
+      "step": 4359
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.39627153847384133,
+      "learning_rate": 1.9599327835088056e-05,
+      "loss": 1.247,
+      "step": 4360
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.48098556044771706,
+      "learning_rate": 1.9563854798621706e-05,
+      "loss": 1.0691,
+      "step": 4361
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.45572575788961134,
+      "learning_rate": 1.9528410411798727e-05,
+      "loss": 1.1636,
+      "step": 4362
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.4319885425867567,
+      "learning_rate": 1.9492994687243714e-05,
+      "loss": 1.0655,
+      "step": 4363
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.47754006434822355,
+      "learning_rate": 1.945760763757094e-05,
+      "loss": 1.1042,
+      "step": 4364
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.4672190014738716,
+      "learning_rate": 1.9422249275384573e-05,
+      "loss": 1.1274,
+      "step": 4365
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.530364360149989,
+      "learning_rate": 1.9386919613278475e-05,
+      "loss": 1.1081,
+      "step": 4366
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.440588889968773,
+      "learning_rate": 1.9351618663836325e-05,
+      "loss": 1.295,
+      "step": 4367
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.44567201142471274,
+      "learning_rate": 1.9316346439631573e-05,
+      "loss": 1.088,
+      "step": 4368
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5001830490554149,
+      "learning_rate": 1.9281102953227416e-05,
+      "loss": 1.0109,
+      "step": 4369
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.41231379523977907,
+      "learning_rate": 1.9245888217176854e-05,
+      "loss": 1.0767,
+      "step": 4370
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.4205600623334083,
+      "learning_rate": 1.9210702244022617e-05,
+      "loss": 1.0827,
+      "step": 4371
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.46010106238166243,
+      "learning_rate": 1.9175545046297218e-05,
+      "loss": 1.0931,
+      "step": 4372
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.45923517888630944,
+      "learning_rate": 1.9140416636522872e-05,
+      "loss": 1.1949,
+      "step": 4373
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.4447254788442033,
+      "learning_rate": 1.9105317027211612e-05,
+      "loss": 1.1788,
+      "step": 4374
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.420182693522501,
+      "learning_rate": 1.907024623086515e-05,
+      "loss": 1.1068,
+      "step": 4375
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.414869656378557,
+      "learning_rate": 1.903520425997496e-05,
+      "loss": 1.019,
+      "step": 4376
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.44134168818471614,
+      "learning_rate": 1.9000191127022237e-05,
+      "loss": 1.0874,
+      "step": 4377
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.402372945766694,
+      "learning_rate": 1.896520684447789e-05,
+      "loss": 1.0381,
+      "step": 4378
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.45197734234775316,
+      "learning_rate": 1.893025142480265e-05,
+      "loss": 0.9239,
+      "step": 4379
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.47707403992392866,
+      "learning_rate": 1.889532488044681e-05,
+      "loss": 1.0165,
+      "step": 4380
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5311866250324059,
+      "learning_rate": 1.886042722385053e-05,
+      "loss": 1.1145,
+      "step": 4381
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.3821566301946028,
+      "learning_rate": 1.882555846744356e-05,
+      "loss": 1.0408,
+      "step": 4382
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.45722288300073255,
+      "learning_rate": 1.8790718623645455e-05,
+      "loss": 1.0834,
+      "step": 4383
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.397209058130468,
+      "learning_rate": 1.8755907704865416e-05,
+      "loss": 1.1067,
+      "step": 4384
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.4456468570024379,
+      "learning_rate": 1.872112572350233e-05,
+      "loss": 1.0539,
+      "step": 4385
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.4902100303831539,
+      "learning_rate": 1.8686372691944787e-05,
+      "loss": 0.9646,
+      "step": 4386
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.4273835149937985,
+      "learning_rate": 1.8651648622571128e-05,
+      "loss": 0.8861,
+      "step": 4387
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.4257578954010938,
+      "learning_rate": 1.861695352774929e-05,
+      "loss": 1.242,
+      "step": 4388
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.49010054045145346,
+      "learning_rate": 1.858228741983692e-05,
+      "loss": 1.0763,
+      "step": 4389
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.46058730002367354,
+      "learning_rate": 1.8547650311181387e-05,
+      "loss": 1.1218,
+      "step": 4390
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.4135389373023044,
+      "learning_rate": 1.851304221411967e-05,
+      "loss": 1.2448,
+      "step": 4391
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.4241113242781263,
+      "learning_rate": 1.847846314097843e-05,
+      "loss": 1.0094,
+      "step": 4392
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.41707824038774466,
+      "learning_rate": 1.8443913104073983e-05,
+      "loss": 1.1239,
+      "step": 4393
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.3915924705929446,
+      "learning_rate": 1.8409392115712366e-05,
+      "loss": 1.1647,
+      "step": 4394
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.4232524524660463,
+      "learning_rate": 1.837490018818917e-05,
+      "loss": 1.1207,
+      "step": 4395
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.44109465979682666,
+      "learning_rate": 1.8340437333789727e-05,
+      "loss": 1.1448,
+      "step": 4396
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.47167167377139274,
+      "learning_rate": 1.830600356478893e-05,
+      "loss": 1.1481,
+      "step": 4397
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.42755313484328145,
+      "learning_rate": 1.8271598893451404e-05,
+      "loss": 1.2627,
+      "step": 4398
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.4724973174853592,
+      "learning_rate": 1.8237223332031327e-05,
+      "loss": 1.1829,
+      "step": 4399
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.3919078585792821,
+      "learning_rate": 1.820287689277256e-05,
+      "loss": 1.1644,
+      "step": 4400
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.4406555043919689,
+      "learning_rate": 1.8168559587908552e-05,
+      "loss": 1.0431,
+      "step": 4401
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.40546733235343757,
+      "learning_rate": 1.8134271429662396e-05,
+      "loss": 1.1897,
+      "step": 4402
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.4388642387993802,
+      "learning_rate": 1.8100012430246837e-05,
+      "loss": 0.999,
+      "step": 4403
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.4939717380722525,
+      "learning_rate": 1.8065782601864156e-05,
+      "loss": 1.2027,
+      "step": 4404
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.4118791208023366,
+      "learning_rate": 1.8031581956706346e-05,
+      "loss": 1.1941,
+      "step": 4405
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.425896090462011,
+      "learning_rate": 1.7997410506954893e-05,
+      "loss": 1.1971,
+      "step": 4406
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.3709664103606557,
+      "learning_rate": 1.7963268264780985e-05,
+      "loss": 1.1381,
+      "step": 4407
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.4301343519005976,
+      "learning_rate": 1.792915524234534e-05,
+      "loss": 1.0781,
+      "step": 4408
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.4112591866763652,
+      "learning_rate": 1.7895071451798306e-05,
+      "loss": 1.1255,
+      "step": 4409
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.38359806346627223,
+      "learning_rate": 1.7861016905279782e-05,
+      "loss": 1.0195,
+      "step": 4410
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.46194181978785837,
+      "learning_rate": 1.7826991614919265e-05,
+      "loss": 1.072,
+      "step": 4411
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.4312227225646645,
+      "learning_rate": 1.779299559283588e-05,
+      "loss": 1.1275,
+      "step": 4412
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.43570051264914644,
+      "learning_rate": 1.775902885113824e-05,
+      "loss": 1.0941,
+      "step": 4413
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.4041089682177866,
+      "learning_rate": 1.772509140192462e-05,
+      "loss": 1.0652,
+      "step": 4414
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.4421388554405724,
+      "learning_rate": 1.7691183257282772e-05,
+      "loss": 1.1821,
+      "step": 4415
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.45468743703516096,
+      "learning_rate": 1.7657304429290134e-05,
+      "loss": 0.9995,
+      "step": 4416
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.4588058104559715,
+      "learning_rate": 1.7623454930013515e-05,
+      "loss": 1.1425,
+      "step": 4417
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.49078429636550885,
+      "learning_rate": 1.7589634771509465e-05,
+      "loss": 0.9523,
+      "step": 4418
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.44980698026350885,
+      "learning_rate": 1.7555843965823992e-05,
+      "loss": 1.1677,
+      "step": 4419
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.43004286039366213,
+      "learning_rate": 1.752208252499261e-05,
+      "loss": 0.9442,
+      "step": 4420
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.40930645587782294,
+      "learning_rate": 1.7488350461040504e-05,
+      "loss": 1.2147,
+      "step": 4421
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.47326298763330604,
+      "learning_rate": 1.7454647785982257e-05,
+      "loss": 1.0091,
+      "step": 4422
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.48468915898935866,
+      "learning_rate": 1.7420974511822085e-05,
+      "loss": 0.9856,
+      "step": 4423
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.4224728629808044,
+      "learning_rate": 1.7387330650553678e-05,
+      "loss": 1.1436,
+      "step": 4424
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.4298839202822309,
+      "learning_rate": 1.7353716214160266e-05,
+      "loss": 1.1471,
+      "step": 4425
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.44659644729957115,
+      "learning_rate": 1.7320131214614566e-05,
+      "loss": 1.082,
+      "step": 4426
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.38716231356006026,
+      "learning_rate": 1.7286575663878877e-05,
+      "loss": 1.1461,
+      "step": 4427
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.3992054397632307,
+      "learning_rate": 1.7253049573904944e-05,
+      "loss": 1.1112,
+      "step": 4428
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.42483211483431205,
+      "learning_rate": 1.721955295663409e-05,
+      "loss": 1.1145,
+      "step": 4429
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.43056662954496394,
+      "learning_rate": 1.7186085823997067e-05,
+      "loss": 1.1508,
+      "step": 4430
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.40681131244190977,
+      "learning_rate": 1.7152648187914134e-05,
+      "loss": 1.147,
+      "step": 4431
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.42420388007947923,
+      "learning_rate": 1.7119240060295105e-05,
+      "loss": 1.1354,
+      "step": 4432
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5923029841212558,
+      "learning_rate": 1.7085861453039243e-05,
+      "loss": 1.085,
+      "step": 4433
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6319792667620752,
+      "learning_rate": 1.705251237803528e-05,
+      "loss": 1.1223,
+      "step": 4434
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.40779023690644856,
+      "learning_rate": 1.7019192847161425e-05,
+      "loss": 1.0005,
+      "step": 4435
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.408859671061738,
+      "learning_rate": 1.698590287228544e-05,
+      "loss": 1.1548,
+      "step": 4436
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.4145888556637396,
+      "learning_rate": 1.6952642465264447e-05,
+      "loss": 1.1025,
+      "step": 4437
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.4995400536282099,
+      "learning_rate": 1.6919411637945147e-05,
+      "loss": 0.9652,
+      "step": 4438
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.45669261050887866,
+      "learning_rate": 1.6886210402163615e-05,
+      "loss": 0.9919,
+      "step": 4439
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.42870982081840736,
+      "learning_rate": 1.6853038769745467e-05,
+      "loss": 1.0342,
+      "step": 4440
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5090110743084709,
+      "learning_rate": 1.6819896752505738e-05,
+      "loss": 0.9524,
+      "step": 4441
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.46613637359236343,
+      "learning_rate": 1.6786784362248818e-05,
+      "loss": 1.057,
+      "step": 4442
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.46806834139551173,
+      "learning_rate": 1.6753701610768724e-05,
+      "loss": 1.0415,
+      "step": 4443
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.44110319620898425,
+      "learning_rate": 1.672064850984878e-05,
+      "loss": 1.1298,
+      "step": 4444
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.4033505232261622,
+      "learning_rate": 1.668762507126185e-05,
+      "loss": 1.0812,
+      "step": 4445
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.3877477212916835,
+      "learning_rate": 1.665463130677011e-05,
+      "loss": 1.0345,
+      "step": 4446
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5485999479001011,
+      "learning_rate": 1.6621667228125302e-05,
+      "loss": 0.9939,
+      "step": 4447
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.4697018598848335,
+      "learning_rate": 1.6588732847068513e-05,
+      "loss": 1.1203,
+      "step": 4448
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.482439656971217,
+      "learning_rate": 1.6555828175330245e-05,
+      "loss": 1.0458,
+      "step": 4449
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.42760467569117916,
+      "learning_rate": 1.6522953224630443e-05,
+      "loss": 0.9077,
+      "step": 4450
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5454431434580298,
+      "learning_rate": 1.6490108006678494e-05,
+      "loss": 1.0671,
+      "step": 4451
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.43159975181619864,
+      "learning_rate": 1.6457292533173153e-05,
+      "loss": 0.9849,
+      "step": 4452
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.47388903410838923,
+      "learning_rate": 1.6424506815802566e-05,
+      "loss": 1.1144,
+      "step": 4453
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.3947187723576146,
+      "learning_rate": 1.639175086624435e-05,
+      "loss": 1.1894,
+      "step": 4454
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.49078821094688974,
+      "learning_rate": 1.635902469616544e-05,
+      "loss": 0.8781,
+      "step": 4455
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.46317861882149275,
+      "learning_rate": 1.6326328317222238e-05,
+      "loss": 1.092,
+      "step": 4456
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.40410394171386727,
+      "learning_rate": 1.6293661741060485e-05,
+      "loss": 1.1324,
+      "step": 4457
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.3969515281922041,
+      "learning_rate": 1.626102497931532e-05,
+      "loss": 1.1994,
+      "step": 4458
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.4774279691194705,
+      "learning_rate": 1.6228418043611227e-05,
+      "loss": 1.0581,
+      "step": 4459
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.39413443477261584,
+      "learning_rate": 1.619584094556217e-05,
+      "loss": 0.9829,
+      "step": 4460
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.4053949077760275,
+      "learning_rate": 1.616329369677134e-05,
+      "loss": 1.2489,
+      "step": 4461
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.43450230262933104,
+      "learning_rate": 1.613077630883145e-05,
+      "loss": 1.0541,
+      "step": 4462
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.4105456126894247,
+      "learning_rate": 1.6098288793324477e-05,
+      "loss": 1.226,
+      "step": 4463
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.3969917093580335,
+      "learning_rate": 1.6065831161821744e-05,
+      "loss": 1.178,
+      "step": 4464
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.453450947854435,
+      "learning_rate": 1.6033403425884052e-05,
+      "loss": 1.1255,
+      "step": 4465
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.48704414743783947,
+      "learning_rate": 1.600100559706137e-05,
+      "loss": 1.142,
+      "step": 4466
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.4270100112831191,
+      "learning_rate": 1.5968637686893186e-05,
+      "loss": 1.168,
+      "step": 4467
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.44755775585875374,
+      "learning_rate": 1.5936299706908208e-05,
+      "loss": 0.9623,
+      "step": 4468
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.49169733412974337,
+      "learning_rate": 1.5903991668624597e-05,
+      "loss": 0.9602,
+      "step": 4469
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.463502371005581,
+      "learning_rate": 1.587171358354973e-05,
+      "loss": 1.0996,
+      "step": 4470
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.4617856753865499,
+      "learning_rate": 1.5839465463180425e-05,
+      "loss": 0.993,
+      "step": 4471
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.40789249092999125,
+      "learning_rate": 1.580724731900276e-05,
+      "loss": 1.183,
+      "step": 4472
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.40477902493011686,
+      "learning_rate": 1.5775059162492146e-05,
+      "loss": 1.1387,
+      "step": 4473
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.4227818517160228,
+      "learning_rate": 1.574290100511332e-05,
+      "loss": 1.1191,
+      "step": 4474
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.41127731408384416,
+      "learning_rate": 1.57107728583203e-05,
+      "loss": 1.1421,
+      "step": 4475
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.4834868707299869,
+      "learning_rate": 1.5678674733556507e-05,
+      "loss": 1.086,
+      "step": 4476
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.4842817501654564,
+      "learning_rate": 1.564660664225457e-05,
+      "loss": 1.1922,
+      "step": 4477
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.44922296879575657,
+      "learning_rate": 1.5614568595836498e-05,
+      "loss": 1.0433,
+      "step": 4478
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.39149533332909325,
+      "learning_rate": 1.5582560605713504e-05,
+      "loss": 1.1673,
+      "step": 4479
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.4512802368885381,
+      "learning_rate": 1.555058268328622e-05,
+      "loss": 0.8782,
+      "step": 4480
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.45229102098446483,
+      "learning_rate": 1.5518634839944468e-05,
+      "loss": 1.0925,
+      "step": 4481
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.47458554771119765,
+      "learning_rate": 1.5486717087067393e-05,
+      "loss": 0.9257,
+      "step": 4482
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.43554774153875997,
+      "learning_rate": 1.545482943602341e-05,
+      "loss": 1.1139,
+      "step": 4483
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.4909116110654624,
+      "learning_rate": 1.5422971898170212e-05,
+      "loss": 1.2745,
+      "step": 4484
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.39540082000038923,
+      "learning_rate": 1.5391144484854815e-05,
+      "loss": 0.9764,
+      "step": 4485
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.43581769449183544,
+      "learning_rate": 1.535934720741342e-05,
+      "loss": 1.0937,
+      "step": 4486
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.42603761518587824,
+      "learning_rate": 1.5327580077171587e-05,
+      "loss": 1.2892,
+      "step": 4487
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.41833823292692035,
+      "learning_rate": 1.5295843105444053e-05,
+      "loss": 1.0162,
+      "step": 4488
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.47523041405763683,
+      "learning_rate": 1.5264136303534894e-05,
+      "loss": 0.9873,
+      "step": 4489
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.4038375584461333,
+      "learning_rate": 1.5232459682737366e-05,
+      "loss": 1.0381,
+      "step": 4490
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.41596054815216144,
+      "learning_rate": 1.5200813254334012e-05,
+      "loss": 1.0594,
+      "step": 4491
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.39919621804368094,
+      "learning_rate": 1.5169197029596604e-05,
+      "loss": 1.138,
+      "step": 4492
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.4084745463460917,
+      "learning_rate": 1.5137611019786213e-05,
+      "loss": 1.2162,
+      "step": 4493
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.43141461509851153,
+      "learning_rate": 1.5106055236153061e-05,
+      "loss": 1.1444,
+      "step": 4494
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.4284958882640726,
+      "learning_rate": 1.5074529689936645e-05,
+      "loss": 1.0995,
+      "step": 4495
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.4294212599040114,
+      "learning_rate": 1.5043034392365729e-05,
+      "loss": 0.9851,
+      "step": 4496
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.4363511605672097,
+      "learning_rate": 1.501156935465825e-05,
+      "loss": 1.1731,
+      "step": 4497
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.41630222207843975,
+      "learning_rate": 1.4980134588021377e-05,
+      "loss": 0.9671,
+      "step": 4498
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.4225394934460105,
+      "learning_rate": 1.4948730103651498e-05,
+      "loss": 0.9958,
+      "step": 4499
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.42511279271503316,
+      "learning_rate": 1.4917355912734243e-05,
+      "loss": 1.2282,
+      "step": 4500
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.4449272385704028,
+      "learning_rate": 1.4886012026444406e-05,
+      "loss": 1.0551,
+      "step": 4501
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.46225876796596904,
+      "learning_rate": 1.4854698455946058e-05,
+      "loss": 1.2058,
+      "step": 4502
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.43802752651098237,
+      "learning_rate": 1.4823415212392377e-05,
+      "loss": 1.1059,
+      "step": 4503
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5082100139636094,
+      "learning_rate": 1.4792162306925827e-05,
+      "loss": 1.0164,
+      "step": 4504
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.39612479753544094,
+      "learning_rate": 1.4760939750678004e-05,
+      "loss": 1.1943,
+      "step": 4505
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.46532884454350004,
+      "learning_rate": 1.4729747554769735e-05,
+      "loss": 1.0953,
+      "step": 4506
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.4544919620108365,
+      "learning_rate": 1.4698585730310998e-05,
+      "loss": 1.1622,
+      "step": 4507
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.39305281322047353,
+      "learning_rate": 1.4667454288400962e-05,
+      "loss": 1.1808,
+      "step": 4508
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.4168099820829916,
+      "learning_rate": 1.463635324012802e-05,
+      "loss": 1.0697,
+      "step": 4509
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.3898403083452531,
+      "learning_rate": 1.460528259656967e-05,
+      "loss": 1.2911,
+      "step": 4510
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.46478061267233245,
+      "learning_rate": 1.4574242368792657e-05,
+      "loss": 1.2245,
+      "step": 4511
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.3848646806315029,
+      "learning_rate": 1.4543232567852783e-05,
+      "loss": 1.105,
+      "step": 4512
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.43649878489354843,
+      "learning_rate": 1.4512253204795156e-05,
+      "loss": 1.16,
+      "step": 4513
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.46762043393467795,
+      "learning_rate": 1.4481304290653919e-05,
+      "loss": 1.0482,
+      "step": 4514
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.43412387053416035,
+      "learning_rate": 1.4450385836452429e-05,
+      "loss": 1.3572,
+      "step": 4515
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.47004962914346093,
+      "learning_rate": 1.4419497853203189e-05,
+      "loss": 0.8688,
+      "step": 4516
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.3945533330231974,
+      "learning_rate": 1.4388640351907801e-05,
+      "loss": 1.1725,
+      "step": 4517
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.4544487425978656,
+      "learning_rate": 1.4357813343557103e-05,
+      "loss": 1.1975,
+      "step": 4518
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.40552085708314056,
+      "learning_rate": 1.4327016839130969e-05,
+      "loss": 1.111,
+      "step": 4519
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.41805854497253375,
+      "learning_rate": 1.4296250849598514e-05,
+      "loss": 1.0007,
+      "step": 4520
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.4244528407161119,
+      "learning_rate": 1.4265515385917904e-05,
+      "loss": 1.0956,
+      "step": 4521
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.43082206172428694,
+      "learning_rate": 1.4234810459036452e-05,
+      "loss": 1.2174,
+      "step": 4522
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.45753483612501505,
+      "learning_rate": 1.4204136079890584e-05,
+      "loss": 1.1265,
+      "step": 4523
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.4428632880355992,
+      "learning_rate": 1.4173492259405896e-05,
+      "loss": 1.1225,
+      "step": 4524
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.4409686206294535,
+      "learning_rate": 1.4142879008497056e-05,
+      "loss": 0.9833,
+      "step": 4525
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.4014436503943478,
+      "learning_rate": 1.411229633806781e-05,
+      "loss": 0.9845,
+      "step": 4526
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.4391146746464223,
+      "learning_rate": 1.408174425901112e-05,
+      "loss": 1.1688,
+      "step": 4527
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.4340734000287139,
+      "learning_rate": 1.4051222782208939e-05,
+      "loss": 0.9628,
+      "step": 4528
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.49776977923246196,
+      "learning_rate": 1.4020731918532393e-05,
+      "loss": 0.9882,
+      "step": 4529
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.4441565707635681,
+      "learning_rate": 1.3990271678841682e-05,
+      "loss": 1.0594,
+      "step": 4530
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.45784833624860066,
+      "learning_rate": 1.3959842073986085e-05,
+      "loss": 1.0632,
+      "step": 4531
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.4431861156405426,
+      "learning_rate": 1.3929443114803942e-05,
+      "loss": 1.1575,
+      "step": 4532
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.4032331333154148,
+      "learning_rate": 1.389907481212278e-05,
+      "loss": 1.0444,
+      "step": 4533
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.46473864783728414,
+      "learning_rate": 1.3868737176759106e-05,
+      "loss": 1.3351,
+      "step": 4534
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.3986057285401156,
+      "learning_rate": 1.383843021951855e-05,
+      "loss": 1.3197,
+      "step": 4535
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.4102181839974426,
+      "learning_rate": 1.3808153951195812e-05,
+      "loss": 0.99,
+      "step": 4536
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.39759623000695804,
+      "learning_rate": 1.3777908382574622e-05,
+      "loss": 0.9633,
+      "step": 4537
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.4433095034287833,
+      "learning_rate": 1.3747693524427862e-05,
+      "loss": 1.0752,
+      "step": 4538
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.4366657306277558,
+      "learning_rate": 1.371750938751739e-05,
+      "loss": 1.2395,
+      "step": 4539
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.4242769450566901,
+      "learning_rate": 1.368735598259414e-05,
+      "loss": 1.2228,
+      "step": 4540
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.3945614343842941,
+      "learning_rate": 1.3657233320398122e-05,
+      "loss": 1.3059,
+      "step": 4541
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.4341793109221358,
+      "learning_rate": 1.3627141411658406e-05,
+      "loss": 1.118,
+      "step": 4542
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.44085727233537075,
+      "learning_rate": 1.3597080267093066e-05,
+      "loss": 1.0703,
+      "step": 4543
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.41485777959450787,
+      "learning_rate": 1.3567049897409268e-05,
+      "loss": 1.1119,
+      "step": 4544
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.48798432095337746,
+      "learning_rate": 1.3537050313303179e-05,
+      "loss": 1.0456,
+      "step": 4545
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.47493794811815254,
+      "learning_rate": 1.3507081525459997e-05,
+      "loss": 1.0442,
+      "step": 4546
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.4335389881256516,
+      "learning_rate": 1.3477143544553995e-05,
+      "loss": 1.1514,
+      "step": 4547
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.4372524219683798,
+      "learning_rate": 1.3447236381248395e-05,
+      "loss": 0.9997,
+      "step": 4548
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.4564923685039287,
+      "learning_rate": 1.341736004619556e-05,
+      "loss": 0.8942,
+      "step": 4549
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.46086234842487567,
+      "learning_rate": 1.338751455003674e-05,
+      "loss": 1.0973,
+      "step": 4550
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.45510379042784294,
+      "learning_rate": 1.335769990340232e-05,
+      "loss": 0.9895,
+      "step": 4551
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.43121131980601635,
+      "learning_rate": 1.33279161169116e-05,
+      "loss": 1.1927,
+      "step": 4552
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.41360503488844114,
+      "learning_rate": 1.3298163201172964e-05,
+      "loss": 1.143,
+      "step": 4553
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.4137454207085425,
+      "learning_rate": 1.3268441166783752e-05,
+      "loss": 1.1637,
+      "step": 4554
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.3973169196756405,
+      "learning_rate": 1.3238750024330338e-05,
+      "loss": 1.0255,
+      "step": 4555
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.4801582518259874,
+      "learning_rate": 1.3209089784388019e-05,
+      "loss": 1.1829,
+      "step": 4556
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.4839283451176308,
+      "learning_rate": 1.317946045752122e-05,
+      "loss": 0.968,
+      "step": 4557
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.4749859948693398,
+      "learning_rate": 1.3149862054283225e-05,
+      "loss": 1.1386,
+      "step": 4558
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.4407883275603721,
+      "learning_rate": 1.3120294585216353e-05,
+      "loss": 0.971,
+      "step": 4559
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5809130954612504,
+      "learning_rate": 1.3090758060851938e-05,
+      "loss": 0.9645,
+      "step": 4560
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.40633627945241724,
+      "learning_rate": 1.3061252491710229e-05,
+      "loss": 1.2353,
+      "step": 4561
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.4538356291831223,
+      "learning_rate": 1.3031777888300523e-05,
+      "loss": 0.9873,
+      "step": 4562
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.4053244003539211,
+      "learning_rate": 1.300233426112103e-05,
+      "loss": 1.1116,
+      "step": 4563
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5020894434352963,
+      "learning_rate": 1.2972921620658928e-05,
+      "loss": 0.983,
+      "step": 4564
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.40490705443647834,
+      "learning_rate": 1.2943539977390361e-05,
+      "loss": 1.0688,
+      "step": 4565
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.43124128347125873,
+      "learning_rate": 1.291418934178048e-05,
+      "loss": 1.1137,
+      "step": 4566
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.4307103835890677,
+      "learning_rate": 1.2884869724283332e-05,
+      "loss": 0.9925,
+      "step": 4567
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.39201335001730014,
+      "learning_rate": 1.2855581135341976e-05,
+      "loss": 1.17,
+      "step": 4568
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.4839796049760975,
+      "learning_rate": 1.2826323585388355e-05,
+      "loss": 1.1828,
+      "step": 4569
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.4120725022439862,
+      "learning_rate": 1.279709708484339e-05,
+      "loss": 1.1585,
+      "step": 4570
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.41515771214024516,
+      "learning_rate": 1.2767901644116941e-05,
+      "loss": 1.2845,
+      "step": 4571
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.38968190054957447,
+      "learning_rate": 1.273873727360777e-05,
+      "loss": 1.1731,
+      "step": 4572
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.408234590159623,
+      "learning_rate": 1.2709603983703678e-05,
+      "loss": 0.9135,
+      "step": 4573
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.4081903350763686,
+      "learning_rate": 1.2680501784781241e-05,
+      "loss": 1.0257,
+      "step": 4574
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.4166891013095214,
+      "learning_rate": 1.2651430687206112e-05,
+      "loss": 1.1414,
+      "step": 4575
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.45036905594193877,
+      "learning_rate": 1.2622390701332754e-05,
+      "loss": 0.9955,
+      "step": 4576
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.4556276614358003,
+      "learning_rate": 1.2593381837504625e-05,
+      "loss": 1.0797,
+      "step": 4577
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.45343715453599137,
+      "learning_rate": 1.256440410605404e-05,
+      "loss": 1.1743,
+      "step": 4578
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.40135196308983867,
+      "learning_rate": 1.2535457517302263e-05,
+      "loss": 1.21,
+      "step": 4579
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.39197009940174454,
+      "learning_rate": 1.2506542081559447e-05,
+      "loss": 1.2852,
+      "step": 4580
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.4565264662684635,
+      "learning_rate": 1.2477657809124631e-05,
+      "loss": 1.0356,
+      "step": 4581
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.42966138443177254,
+      "learning_rate": 1.2448804710285833e-05,
+      "loss": 1.0919,
+      "step": 4582
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.4856291965763213,
+      "learning_rate": 1.2419982795319852e-05,
+      "loss": 1.0371,
+      "step": 4583
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5089386494978089,
+      "learning_rate": 1.2391192074492496e-05,
+      "loss": 1.0752,
+      "step": 4584
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.43132993815761816,
+      "learning_rate": 1.2362432558058356e-05,
+      "loss": 1.2128,
+      "step": 4585
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.44971691599995856,
+      "learning_rate": 1.2333704256261014e-05,
+      "loss": 1.1433,
+      "step": 4586
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.46033253641424055,
+      "learning_rate": 1.230500717933285e-05,
+      "loss": 1.1296,
+      "step": 4587
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.411896153000042,
+      "learning_rate": 1.2276341337495157e-05,
+      "loss": 1.0697,
+      "step": 4588
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.44391067076522867,
+      "learning_rate": 1.2247706740958087e-05,
+      "loss": 1.254,
+      "step": 4589
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.43209656635926835,
+      "learning_rate": 1.2219103399920673e-05,
+      "loss": 1.0519,
+      "step": 4590
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.40761638985656085,
+      "learning_rate": 1.2190531324570843e-05,
+      "loss": 0.9074,
+      "step": 4591
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.39487730372563284,
+      "learning_rate": 1.216199052508532e-05,
+      "loss": 1.0967,
+      "step": 4592
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.43121875703403495,
+      "learning_rate": 1.2133481011629777e-05,
+      "loss": 1.1311,
+      "step": 4593
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.41583979703451,
+      "learning_rate": 1.2105002794358666e-05,
+      "loss": 1.2131,
+      "step": 4594
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.4168501340809143,
+      "learning_rate": 1.207655588341534e-05,
+      "loss": 1.0869,
+      "step": 4595
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.43822546806165846,
+      "learning_rate": 1.2048140288931942e-05,
+      "loss": 1.0023,
+      "step": 4596
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.40585257438989536,
+      "learning_rate": 1.201975602102956e-05,
+      "loss": 0.9778,
+      "step": 4597
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.45760797728472047,
+      "learning_rate": 1.199140308981802e-05,
+      "loss": 1.1317,
+      "step": 4598
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.4270577513434468,
+      "learning_rate": 1.1963081505396068e-05,
+      "loss": 1.0735,
+      "step": 4599
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.4825495143416405,
+      "learning_rate": 1.1934791277851242e-05,
+      "loss": 1.1235,
+      "step": 4600
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.536241406780845,
+      "learning_rate": 1.1906532417259886e-05,
+      "loss": 1.1481,
+      "step": 4601
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.42675923991502956,
+      "learning_rate": 1.1878304933687246e-05,
+      "loss": 1.0617,
+      "step": 4602
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.4396101907201179,
+      "learning_rate": 1.1850108837187335e-05,
+      "loss": 0.9878,
+      "step": 4603
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.4907961239193105,
+      "learning_rate": 1.1821944137803009e-05,
+      "loss": 0.9653,
+      "step": 4604
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.454947455187568,
+      "learning_rate": 1.1793810845565901e-05,
+      "loss": 1.1526,
+      "step": 4605
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.44879207293019313,
+      "learning_rate": 1.1765708970496548e-05,
+      "loss": 1.0899,
+      "step": 4606
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.4442667433232714,
+      "learning_rate": 1.1737638522604177e-05,
+      "loss": 1.0248,
+      "step": 4607
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.4187433093297504,
+      "learning_rate": 1.1709599511886948e-05,
+      "loss": 1.2504,
+      "step": 4608
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.44061855144321044,
+      "learning_rate": 1.1681591948331705e-05,
+      "loss": 1.0495,
+      "step": 4609
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.47983457010574304,
+      "learning_rate": 1.1653615841914189e-05,
+      "loss": 0.9758,
+      "step": 4610
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.4365393313514957,
+      "learning_rate": 1.1625671202598875e-05,
+      "loss": 1.1322,
+      "step": 4611
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.473736193250286,
+      "learning_rate": 1.1597758040339046e-05,
+      "loss": 1.1006,
+      "step": 4612
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.43996193303426245,
+      "learning_rate": 1.1569876365076793e-05,
+      "loss": 1.1509,
+      "step": 4613
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.4015820407287281,
+      "learning_rate": 1.1542026186742926e-05,
+      "loss": 1.093,
+      "step": 4614
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.4416063445976301,
+      "learning_rate": 1.1514207515257147e-05,
+      "loss": 1.1541,
+      "step": 4615
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.41087030050456796,
+      "learning_rate": 1.1486420360527838e-05,
+      "loss": 1.0142,
+      "step": 4616
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.43042292385552366,
+      "learning_rate": 1.1458664732452218e-05,
+      "loss": 1.1372,
+      "step": 4617
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5416373208064854,
+      "learning_rate": 1.1430940640916221e-05,
+      "loss": 0.9327,
+      "step": 4618
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.4493405326075682,
+      "learning_rate": 1.1403248095794628e-05,
+      "loss": 1.1993,
+      "step": 4619
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.4310349730558723,
+      "learning_rate": 1.1375587106950858e-05,
+      "loss": 1.1432,
+      "step": 4620
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.49800459424311105,
+      "learning_rate": 1.1347957684237243e-05,
+      "loss": 0.973,
+      "step": 4621
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.39576235294054923,
+      "learning_rate": 1.1320359837494754e-05,
+      "loss": 1.1953,
+      "step": 4622
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.4819860993288274,
+      "learning_rate": 1.129279357655315e-05,
+      "loss": 1.0834,
+      "step": 4623
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.44969305795125397,
+      "learning_rate": 1.1265258911230981e-05,
+      "loss": 1.0073,
+      "step": 4624
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.4320214215432609,
+      "learning_rate": 1.1237755851335463e-05,
+      "loss": 1.0025,
+      "step": 4625
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.4592054377181996,
+      "learning_rate": 1.1210284406662652e-05,
+      "loss": 1.1152,
+      "step": 4626
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5276893261690359,
+      "learning_rate": 1.1182844586997266e-05,
+      "loss": 1.0184,
+      "step": 4627
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.39323290276715733,
+      "learning_rate": 1.1155436402112785e-05,
+      "loss": 1.1767,
+      "step": 4628
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.4467193539786629,
+      "learning_rate": 1.11280598617714e-05,
+      "loss": 1.116,
+      "step": 4629
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.4123000498674916,
+      "learning_rate": 1.1100714975724103e-05,
+      "loss": 1.1156,
+      "step": 4630
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5297832887198088,
+      "learning_rate": 1.107340175371051e-05,
+      "loss": 1.0627,
+      "step": 4631
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.4528078035937782,
+      "learning_rate": 1.1046120205459043e-05,
+      "loss": 1.0549,
+      "step": 4632
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.37278620075168456,
+      "learning_rate": 1.1018870340686804e-05,
+      "loss": 1.1042,
+      "step": 4633
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.4024160983619849,
+      "learning_rate": 1.0991652169099598e-05,
+      "loss": 1.0952,
+      "step": 4634
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5125394137607182,
+      "learning_rate": 1.096446570039198e-05,
+      "loss": 0.9475,
+      "step": 4635
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5610223332843564,
+      "learning_rate": 1.0937310944247181e-05,
+      "loss": 1.2202,
+      "step": 4636
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.40378605955852176,
+      "learning_rate": 1.0910187910337144e-05,
+      "loss": 1.0406,
+      "step": 4637
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.41177889012996344,
+      "learning_rate": 1.0883096608322507e-05,
+      "loss": 1.3368,
+      "step": 4638
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.3915738942810706,
+      "learning_rate": 1.085603704785264e-05,
+      "loss": 1.0721,
+      "step": 4639
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.39101822863020114,
+      "learning_rate": 1.0829009238565558e-05,
+      "loss": 1.0549,
+      "step": 4640
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.4750802116184324,
+      "learning_rate": 1.0802013190088024e-05,
+      "loss": 1.0872,
+      "step": 4641
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.4571551033551232,
+      "learning_rate": 1.0775048912035435e-05,
+      "loss": 1.1196,
+      "step": 4642
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.4689375185604434,
+      "learning_rate": 1.0748116414011888e-05,
+      "loss": 1.2187,
+      "step": 4643
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.42787578694325973,
+      "learning_rate": 1.0721215705610178e-05,
+      "loss": 1.0868,
+      "step": 4644
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.41892883596736524,
+      "learning_rate": 1.0694346796411747e-05,
+      "loss": 1.092,
+      "step": 4645
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.4500740003932142,
+      "learning_rate": 1.0667509695986754e-05,
+      "loss": 1.139,
+      "step": 4646
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.40019306513769887,
+      "learning_rate": 1.0640704413893965e-05,
+      "loss": 0.9798,
+      "step": 4647
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.42221050495332807,
+      "learning_rate": 1.0613930959680906e-05,
+      "loss": 1.1087,
+      "step": 4648
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.43441768099382483,
+      "learning_rate": 1.0587189342883652e-05,
+      "loss": 1.1291,
+      "step": 4649
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.4410397250804397,
+      "learning_rate": 1.056047957302705e-05,
+      "loss": 1.0671,
+      "step": 4650
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.39645518778844546,
+      "learning_rate": 1.0533801659624531e-05,
+      "loss": 1.1665,
+      "step": 4651
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.4215515240473389,
+      "learning_rate": 1.0507155612178199e-05,
+      "loss": 1.0073,
+      "step": 4652
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.455377702052344,
+      "learning_rate": 1.0480541440178816e-05,
+      "loss": 1.0434,
+      "step": 4653
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.4145620445155117,
+      "learning_rate": 1.0453959153105752e-05,
+      "loss": 1.212,
+      "step": 4654
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.522395472122109,
+      "learning_rate": 1.0427408760427093e-05,
+      "loss": 1.1612,
+      "step": 4655
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.4592985867157704,
+      "learning_rate": 1.0400890271599483e-05,
+      "loss": 1.0581,
+      "step": 4656
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.3819172118021537,
+      "learning_rate": 1.0374403696068291e-05,
+      "loss": 1.3468,
+      "step": 4657
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.4709446095388339,
+      "learning_rate": 1.0347949043267436e-05,
+      "loss": 1.1222,
+      "step": 4658
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.39810924699279854,
+      "learning_rate": 1.0321526322619534e-05,
+      "loss": 1.1053,
+      "step": 4659
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.4383310406061623,
+      "learning_rate": 1.0295135543535772e-05,
+      "loss": 1.1383,
+      "step": 4660
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.4288431180832216,
+      "learning_rate": 1.0268776715415984e-05,
+      "loss": 1.0481,
+      "step": 4661
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.42516074508596197,
+      "learning_rate": 1.0242449847648616e-05,
+      "loss": 1.1342,
+      "step": 4662
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5335049414525177,
+      "learning_rate": 1.0216154949610768e-05,
+      "loss": 1.1213,
+      "step": 4663
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.3883128003339493,
+      "learning_rate": 1.0189892030668114e-05,
+      "loss": 1.0891,
+      "step": 4664
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.41594166235784064,
+      "learning_rate": 1.0163661100174904e-05,
+      "loss": 1.1023,
+      "step": 4665
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.4187765215340906,
+      "learning_rate": 1.013746216747411e-05,
+      "loss": 1.1386,
+      "step": 4666
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.38757808099602553,
+      "learning_rate": 1.0111295241897157e-05,
+      "loss": 1.1588,
+      "step": 4667
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.4274834354021409,
+      "learning_rate": 1.0085160332764242e-05,
+      "loss": 1.1244,
+      "step": 4668
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5475871747495635,
+      "learning_rate": 1.0059057449383958e-05,
+      "loss": 0.9028,
+      "step": 4669
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.4221218875877746,
+      "learning_rate": 1.0032986601053673e-05,
+      "loss": 1.0368,
+      "step": 4670
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.4411113714185451,
+      "learning_rate": 1.0006947797059219e-05,
+      "loss": 1.1314,
+      "step": 4671
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.47113335836171566,
+      "learning_rate": 9.980941046675107e-06,
+      "loss": 1.0675,
+      "step": 4672
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.4543914732164602,
+      "learning_rate": 9.954966359164352e-06,
+      "loss": 1.0554,
+      "step": 4673
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.44106970735691675,
+      "learning_rate": 9.92902374377862e-06,
+      "loss": 1.0236,
+      "step": 4674
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5152577201087392,
+      "learning_rate": 9.903113209758096e-06,
+      "loss": 0.9626,
+      "step": 4675
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.4117144536841753,
+      "learning_rate": 9.877234766331567e-06,
+      "loss": 0.9876,
+      "step": 4676
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.43219754298602464,
+      "learning_rate": 9.851388422716379e-06,
+      "loss": 1.1121,
+      "step": 4677
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.3983147109231423,
+      "learning_rate": 9.825574188118437e-06,
+      "loss": 1.126,
+      "step": 4678
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.43175413835278914,
+      "learning_rate": 9.79979207173225e-06,
+      "loss": 1.1109,
+      "step": 4679
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.4274532184838412,
+      "learning_rate": 9.774042082740831e-06,
+      "loss": 1.2029,
+      "step": 4680
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.43761366690876935,
+      "learning_rate": 9.748324230315809e-06,
+      "loss": 1.0237,
+      "step": 4681
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.38727228558876353,
+      "learning_rate": 9.72263852361731e-06,
+      "loss": 1.1825,
+      "step": 4682
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.47099958615627463,
+      "learning_rate": 9.696984971794065e-06,
+      "loss": 1.1103,
+      "step": 4683
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.4215763717564122,
+      "learning_rate": 9.671363583983295e-06,
+      "loss": 1.1824,
+      "step": 4684
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.4841717255803452,
+      "learning_rate": 9.64577436931081e-06,
+      "loss": 1.2807,
+      "step": 4685
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.40163285265993004,
+      "learning_rate": 9.620217336890936e-06,
+      "loss": 1.2712,
+      "step": 4686
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.4837576788945594,
+      "learning_rate": 9.594692495826529e-06,
+      "loss": 1.1247,
+      "step": 4687
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.43501736121299905,
+      "learning_rate": 9.569199855209033e-06,
+      "loss": 0.9851,
+      "step": 4688
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.4131296589147409,
+      "learning_rate": 9.543739424118336e-06,
+      "loss": 1.1045,
+      "step": 4689
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.40199952240547304,
+      "learning_rate": 9.518311211622944e-06,
+      "loss": 1.1574,
+      "step": 4690
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5193409002759954,
+      "learning_rate": 9.492915226779808e-06,
+      "loss": 1.0002,
+      "step": 4691
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.4365572777851364,
+      "learning_rate": 9.467551478634485e-06,
+      "loss": 1.0818,
+      "step": 4692
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.4282901129560482,
+      "learning_rate": 9.442219976220979e-06,
+      "loss": 1.1012,
+      "step": 4693
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.3893760810089681,
+      "learning_rate": 9.416920728561818e-06,
+      "loss": 1.0479,
+      "step": 4694
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.44866559431420466,
+      "learning_rate": 9.391653744668072e-06,
+      "loss": 1.1597,
+      "step": 4695
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.48679429594749996,
+      "learning_rate": 9.366419033539275e-06,
+      "loss": 1.2782,
+      "step": 4696
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.47504939910166105,
+      "learning_rate": 9.34121660416354e-06,
+      "loss": 1.0098,
+      "step": 4697
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.4348867502249225,
+      "learning_rate": 9.316046465517381e-06,
+      "loss": 1.1635,
+      "step": 4698
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.44495080713558804,
+      "learning_rate": 9.29090862656593e-06,
+      "loss": 1.0128,
+      "step": 4699
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.42041606439537943,
+      "learning_rate": 9.265803096262715e-06,
+      "loss": 0.9852,
+      "step": 4700
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.4111926700223509,
+      "learning_rate": 9.240729883549792e-06,
+      "loss": 1.3196,
+      "step": 4701
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.4834944774887555,
+      "learning_rate": 9.215688997357686e-06,
+      "loss": 1.0195,
+      "step": 4702
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.47531794822813167,
+      "learning_rate": 9.19068044660547e-06,
+      "loss": 1.2518,
+      "step": 4703
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.4208852326927555,
+      "learning_rate": 9.165704240200623e-06,
+      "loss": 1.0351,
+      "step": 4704
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.42480690455143644,
+      "learning_rate": 9.14076038703917e-06,
+      "loss": 1.1028,
+      "step": 4705
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.42494269184479655,
+      "learning_rate": 9.115848896005575e-06,
+      "loss": 0.886,
+      "step": 4706
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.4012558411331378,
+      "learning_rate": 9.090969775972736e-06,
+      "loss": 1.1236,
+      "step": 4707
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.48118844938590943,
+      "learning_rate": 9.066123035802122e-06,
+      "loss": 1.0826,
+      "step": 4708
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.44778448669879584,
+      "learning_rate": 9.041308684343596e-06,
+      "loss": 1.1893,
+      "step": 4709
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.4683259988703086,
+      "learning_rate": 9.016526730435493e-06,
+      "loss": 1.0804,
+      "step": 4710
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.41833159017332844,
+      "learning_rate": 8.991777182904605e-06,
+      "loss": 1.3299,
+      "step": 4711
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.44403267972790017,
+      "learning_rate": 8.967060050566224e-06,
+      "loss": 1.1875,
+      "step": 4712
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.41400708662942687,
+      "learning_rate": 8.942375342224041e-06,
+      "loss": 0.9784,
+      "step": 4713
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.3980601644369456,
+      "learning_rate": 8.917723066670247e-06,
+      "loss": 1.2328,
+      "step": 4714
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.43644190283647993,
+      "learning_rate": 8.89310323268544e-06,
+      "loss": 1.0747,
+      "step": 4715
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.44170867099509475,
+      "learning_rate": 8.868515849038705e-06,
+      "loss": 1.1587,
+      "step": 4716
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.46277696643658567,
+      "learning_rate": 8.843960924487538e-06,
+      "loss": 1.0648,
+      "step": 4717
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.43548676029462985,
+      "learning_rate": 8.819438467777841e-06,
+      "loss": 1.0974,
+      "step": 4718
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5077665029397419,
+      "learning_rate": 8.79494848764404e-06,
+      "loss": 1.194,
+      "step": 4719
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.49963841424188815,
+      "learning_rate": 8.770490992808911e-06,
+      "loss": 1.0814,
+      "step": 4720
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5828347592116789,
+      "learning_rate": 8.74606599198372e-06,
+      "loss": 1.0633,
+      "step": 4721
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.48075421824373105,
+      "learning_rate": 8.72167349386811e-06,
+      "loss": 1.1516,
+      "step": 4722
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.3981869776019179,
+      "learning_rate": 8.697313507150184e-06,
+      "loss": 1.2055,
+      "step": 4723
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.49170536992688274,
+      "learning_rate": 8.672986040506459e-06,
+      "loss": 1.074,
+      "step": 4724
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.430342445487443,
+      "learning_rate": 8.648691102601858e-06,
+      "loss": 1.1457,
+      "step": 4725
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.42893623155438826,
+      "learning_rate": 8.624428702089682e-06,
+      "loss": 0.968,
+      "step": 4726
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.4655241170797992,
+      "learning_rate": 8.600198847611729e-06,
+      "loss": 1.1081,
+      "step": 4727
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.3997431788911307,
+      "learning_rate": 8.576001547798152e-06,
+      "loss": 1.208,
+      "step": 4728
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5146069326087775,
+      "learning_rate": 8.551836811267478e-06,
+      "loss": 1.0837,
+      "step": 4729
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5105664360822739,
+      "learning_rate": 8.527704646626711e-06,
+      "loss": 0.9748,
+      "step": 4730
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.49893106112481145,
+      "learning_rate": 8.503605062471187e-06,
+      "loss": 1.0651,
+      "step": 4731
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5377216268091646,
+      "learning_rate": 8.479538067384706e-06,
+      "loss": 1.0222,
+      "step": 4732
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.45758490298750004,
+      "learning_rate": 8.45550366993938e-06,
+      "loss": 1.0685,
+      "step": 4733
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.4649857298731256,
+      "learning_rate": 8.431501878695769e-06,
+      "loss": 1.0985,
+      "step": 4734
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.4729358657701154,
+      "learning_rate": 8.40753270220277e-06,
+      "loss": 1.1958,
+      "step": 4735
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.40751935691562474,
+      "learning_rate": 8.383596148997742e-06,
+      "loss": 1.0259,
+      "step": 4736
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.4531823961029114,
+      "learning_rate": 8.35969222760633e-06,
+      "loss": 1.081,
+      "step": 4737
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6211896908066965,
+      "learning_rate": 8.33582094654265e-06,
+      "loss": 1.0523,
+      "step": 4738
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.4142303536604132,
+      "learning_rate": 8.311982314309109e-06,
+      "loss": 1.1033,
+      "step": 4739
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.4716113250312623,
+      "learning_rate": 8.288176339396514e-06,
+      "loss": 1.239,
+      "step": 4740
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.3887875966496256,
+      "learning_rate": 8.264403030284074e-06,
+      "loss": 1.0732,
+      "step": 4741
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5637063299021418,
+      "learning_rate": 8.240662395439325e-06,
+      "loss": 1.1024,
+      "step": 4742
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.4767788975337744,
+      "learning_rate": 8.216954443318181e-06,
+      "loss": 1.0464,
+      "step": 4743
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.4574007271226469,
+      "learning_rate": 8.193279182364888e-06,
+      "loss": 1.0851,
+      "step": 4744
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.4049768742144153,
+      "learning_rate": 8.1696366210121e-06,
+      "loss": 1.1666,
+      "step": 4745
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.45397876803032466,
+      "learning_rate": 8.146026767680758e-06,
+      "loss": 0.9577,
+      "step": 4746
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.43922311854162766,
+      "learning_rate": 8.122449630780238e-06,
+      "loss": 1.033,
+      "step": 4747
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.43461060666759915,
+      "learning_rate": 8.098905218708186e-06,
+      "loss": 1.0701,
+      "step": 4748
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.41720033369345,
+      "learning_rate": 8.075393539850628e-06,
+      "loss": 1.2248,
+      "step": 4749
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.4409291992189723,
+      "learning_rate": 8.051914602581912e-06,
+      "loss": 1.016,
+      "step": 4750
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.40790646513730866,
+      "learning_rate": 8.028468415264723e-06,
+      "loss": 1.0008,
+      "step": 4751
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.429208839241914,
+      "learning_rate": 8.005054986250137e-06,
+      "loss": 1.0943,
+      "step": 4752
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.3928903582607072,
+      "learning_rate": 7.981674323877487e-06,
+      "loss": 1.1631,
+      "step": 4753
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.4339755900307398,
+      "learning_rate": 7.958326436474473e-06,
+      "loss": 1.1672,
+      "step": 4754
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.44912542514076936,
+      "learning_rate": 7.935011332357112e-06,
+      "loss": 1.0155,
+      "step": 4755
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.44570371960754646,
+      "learning_rate": 7.911729019829783e-06,
+      "loss": 1.0247,
+      "step": 4756
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.4952030598949717,
+      "learning_rate": 7.888479507185109e-06,
+      "loss": 1.0973,
+      "step": 4757
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.4471241550951928,
+      "learning_rate": 7.865262802704088e-06,
+      "loss": 1.0367,
+      "step": 4758
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.45119933311602195,
+      "learning_rate": 7.842078914656015e-06,
+      "loss": 1.1575,
+      "step": 4759
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.49751220546900354,
+      "learning_rate": 7.818927851298474e-06,
+      "loss": 1.0242,
+      "step": 4760
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.42722038398404805,
+      "learning_rate": 7.795809620877426e-06,
+      "loss": 1.3202,
+      "step": 4761
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.4758489901143063,
+      "learning_rate": 7.772724231627048e-06,
+      "loss": 1.0096,
+      "step": 4762
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.4644691586212372,
+      "learning_rate": 7.749671691769911e-06,
+      "loss": 1.1973,
+      "step": 4763
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.49680391417744024,
+      "learning_rate": 7.726652009516788e-06,
+      "loss": 1.0161,
+      "step": 4764
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5943678496385155,
+      "learning_rate": 7.703665193066845e-06,
+      "loss": 1.0669,
+      "step": 4765
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.38485323382504644,
+      "learning_rate": 7.680711250607487e-06,
+      "loss": 1.0657,
+      "step": 4766
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.41620740143323587,
+      "learning_rate": 7.657790190314406e-06,
+      "loss": 1.0186,
+      "step": 4767
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.45618980573035905,
+      "learning_rate": 7.634902020351575e-06,
+      "loss": 1.0113,
+      "step": 4768
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.4081964030867149,
+      "learning_rate": 7.612046748871327e-06,
+      "loss": 1.2166,
+      "step": 4769
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.40193734015378335,
+      "learning_rate": 7.589224384014193e-06,
+      "loss": 1.0989,
+      "step": 4770
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.469074879126985,
+      "learning_rate": 7.566434933909006e-06,
+      "loss": 1.1914,
+      "step": 4771
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.441375690422459,
+      "learning_rate": 7.543678406672905e-06,
+      "loss": 1.1185,
+      "step": 4772
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.3855789966050383,
+      "learning_rate": 7.520954810411262e-06,
+      "loss": 1.0893,
+      "step": 4773
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.42681723075074196,
+      "learning_rate": 7.49826415321776e-06,
+      "loss": 0.9199,
+      "step": 4774
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.49708115777441675,
+      "learning_rate": 7.475606443174288e-06,
+      "loss": 1.2344,
+      "step": 4775
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.44314365741626804,
+      "learning_rate": 7.452981688351079e-06,
+      "loss": 1.1909,
+      "step": 4776
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.47265439996415215,
+      "learning_rate": 7.430389896806567e-06,
+      "loss": 1.0319,
+      "step": 4777
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.4557880653882937,
+      "learning_rate": 7.40783107658749e-06,
+      "loss": 1.1269,
+      "step": 4778
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5046336897241174,
+      "learning_rate": 7.385305235728801e-06,
+      "loss": 1.13,
+      "step": 4779
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.4806291241893041,
+      "learning_rate": 7.362812382253737e-06,
+      "loss": 1.0265,
+      "step": 4780
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.4287265116801635,
+      "learning_rate": 7.340352524173777e-06,
+      "loss": 0.9494,
+      "step": 4781
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.41754206789511517,
+      "learning_rate": 7.317925669488623e-06,
+      "loss": 0.9892,
+      "step": 4782
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.4932344719825304,
+      "learning_rate": 7.295531826186264e-06,
+      "loss": 1.1528,
+      "step": 4783
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.4281614379895274,
+      "learning_rate": 7.273171002242885e-06,
+      "loss": 1.221,
+      "step": 4784
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.4697624557375974,
+      "learning_rate": 7.25084320562297e-06,
+      "loss": 1.1727,
+      "step": 4785
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.4666565739931536,
+      "learning_rate": 7.2285484442791705e-06,
+      "loss": 1.1501,
+      "step": 4786
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.45755970128822115,
+      "learning_rate": 7.206286726152434e-06,
+      "loss": 1.2521,
+      "step": 4787
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.46485113623362373,
+      "learning_rate": 7.184058059171894e-06,
+      "loss": 1.1197,
+      "step": 4788
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.4107871716545619,
+      "learning_rate": 7.161862451254941e-06,
+      "loss": 1.0793,
+      "step": 4789
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.4405256405537926,
+      "learning_rate": 7.139699910307185e-06,
+      "loss": 0.9429,
+      "step": 4790
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6106460621497151,
+      "learning_rate": 7.117570444222432e-06,
+      "loss": 1.0873,
+      "step": 4791
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.46226422854274096,
+      "learning_rate": 7.095474060882734e-06,
+      "loss": 1.0232,
+      "step": 4792
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.4526894542001507,
+      "learning_rate": 7.073410768158339e-06,
+      "loss": 0.9434,
+      "step": 4793
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.44234595796309994,
+      "learning_rate": 7.051380573907751e-06,
+      "loss": 1.1919,
+      "step": 4794
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.44917844775928734,
+      "learning_rate": 7.029383485977625e-06,
+      "loss": 1.1709,
+      "step": 4795
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.4580724058886441,
+      "learning_rate": 7.007419512202884e-06,
+      "loss": 1.1312,
+      "step": 4796
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5157594815353771,
+      "learning_rate": 6.985488660406625e-06,
+      "loss": 1.1345,
+      "step": 4797
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.41931366926527813,
+      "learning_rate": 6.963590938400133e-06,
+      "loss": 1.0288,
+      "step": 4798
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.4301897443605237,
+      "learning_rate": 6.941726353982913e-06,
+      "loss": 1.0013,
+      "step": 4799
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.45078363069982064,
+      "learning_rate": 6.919894914942682e-06,
+      "loss": 1.0811,
+      "step": 4800
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.4114643998158335,
+      "learning_rate": 6.898096629055318e-06,
+      "loss": 1.2615,
+      "step": 4801
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.41743019296418055,
+      "learning_rate": 6.876331504084899e-06,
+      "loss": 1.177,
+      "step": 4802
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.4064501577651484,
+      "learning_rate": 6.854599547783736e-06,
+      "loss": 1.1,
+      "step": 4803
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.4330657777800185,
+      "learning_rate": 6.832900767892237e-06,
+      "loss": 1.0243,
+      "step": 4804
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.42938549065515763,
+      "learning_rate": 6.811235172139097e-06,
+      "loss": 1.2016,
+      "step": 4805
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.4650699492176313,
+      "learning_rate": 6.7896027682411215e-06,
+      "loss": 0.9877,
+      "step": 4806
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.4282096488160844,
+      "learning_rate": 6.768003563903303e-06,
+      "loss": 1.1214,
+      "step": 4807
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.43492181985207323,
+      "learning_rate": 6.7464375668188085e-06,
+      "loss": 1.0854,
+      "step": 4808
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.45646656173666295,
+      "learning_rate": 6.724904784669028e-06,
+      "loss": 1.2157,
+      "step": 4809
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.45300240349663523,
+      "learning_rate": 6.703405225123427e-06,
+      "loss": 1.1994,
+      "step": 4810
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5136941319260311,
+      "learning_rate": 6.681938895839746e-06,
+      "loss": 1.2303,
+      "step": 4811
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.403763977869309,
+      "learning_rate": 6.660505804463802e-06,
+      "loss": 1.2011,
+      "step": 4812
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.4628122696283248,
+      "learning_rate": 6.639105958629599e-06,
+      "loss": 0.9048,
+      "step": 4813
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.4856530352152948,
+      "learning_rate": 6.61773936595933e-06,
+      "loss": 1.1289,
+      "step": 4814
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.4273029969065295,
+      "learning_rate": 6.596406034063318e-06,
+      "loss": 1.2644,
+      "step": 4815
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.40554638076037086,
+      "learning_rate": 6.5751059705400295e-06,
+      "loss": 1.0343,
+      "step": 4816
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.44052877543358154,
+      "learning_rate": 6.553839182976085e-06,
+      "loss": 0.9797,
+      "step": 4817
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.4870685901942123,
+      "learning_rate": 6.532605678946291e-06,
+      "loss": 1.1379,
+      "step": 4818
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.42319405737306437,
+      "learning_rate": 6.5114054660135315e-06,
+      "loss": 1.1455,
+      "step": 4819
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.4205569213848013,
+      "learning_rate": 6.490238551728922e-06,
+      "loss": 1.114,
+      "step": 4820
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.4160613979403996,
+      "learning_rate": 6.469104943631621e-06,
+      "loss": 1.0202,
+      "step": 4821
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5287346814231406,
+      "learning_rate": 6.448004649249018e-06,
+      "loss": 1.0794,
+      "step": 4822
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.4220631275280108,
+      "learning_rate": 6.426937676096534e-06,
+      "loss": 1.1255,
+      "step": 4823
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.44307745742199056,
+      "learning_rate": 6.405904031677801e-06,
+      "loss": 0.9924,
+      "step": 4824
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.4428133472773169,
+      "learning_rate": 6.3849037234845584e-06,
+      "loss": 1.1636,
+      "step": 4825
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.47057114460242344,
+      "learning_rate": 6.363936758996658e-06,
+      "loss": 1.1781,
+      "step": 4826
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.4656203225659522,
+      "learning_rate": 6.343003145682114e-06,
+      "loss": 0.9152,
+      "step": 4827
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.4320181734863634,
+      "learning_rate": 6.322102890996995e-06,
+      "loss": 1.1372,
+      "step": 4828
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.43248463021428796,
+      "learning_rate": 6.301236002385568e-06,
+      "loss": 1.1475,
+      "step": 4829
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.435177738384679,
+      "learning_rate": 6.280402487280157e-06,
+      "loss": 1.1684,
+      "step": 4830
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.4832591314969914,
+      "learning_rate": 6.2596023531012015e-06,
+      "loss": 0.9407,
+      "step": 4831
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.442501182191794,
+      "learning_rate": 6.2388356072572765e-06,
+      "loss": 0.8869,
+      "step": 4832
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.4298347216469419,
+      "learning_rate": 6.218102257145064e-06,
+      "loss": 0.969,
+      "step": 4833
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.3831040506917484,
+      "learning_rate": 6.1974023101493364e-06,
+      "loss": 1.2754,
+      "step": 4834
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.4106389404042277,
+      "learning_rate": 6.176735773642961e-06,
+      "loss": 1.1396,
+      "step": 4835
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.40978400880899746,
+      "learning_rate": 6.156102654986951e-06,
+      "loss": 1.0242,
+      "step": 4836
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.45467038552408534,
+      "learning_rate": 6.1355029615303375e-06,
+      "loss": 1.2141,
+      "step": 4837
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.39536756568145764,
+      "learning_rate": 6.114936700610341e-06,
+      "loss": 0.9818,
+      "step": 4838
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.427844971496436,
+      "learning_rate": 6.094403879552213e-06,
+      "loss": 1.062,
+      "step": 4839
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.4358216892050629,
+      "learning_rate": 6.0739045056692925e-06,
+      "loss": 1.154,
+      "step": 4840
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.48191303968613824,
+      "learning_rate": 6.053438586263005e-06,
+      "loss": 1.0637,
+      "step": 4841
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.43631407018182183,
+      "learning_rate": 6.03300612862292e-06,
+      "loss": 1.2279,
+      "step": 4842
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.4518919851010567,
+      "learning_rate": 6.012607140026605e-06,
+      "loss": 0.8924,
+      "step": 4843
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.44657537978486767,
+      "learning_rate": 5.992241627739781e-06,
+      "loss": 1.1933,
+      "step": 4844
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.40347508274951677,
+      "learning_rate": 5.97190959901619e-06,
+      "loss": 1.2609,
+      "step": 4845
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.4721563453477822,
+      "learning_rate": 5.951611061097672e-06,
+      "loss": 1.02,
+      "step": 4846
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.49414145972293533,
+      "learning_rate": 5.931346021214123e-06,
+      "loss": 1.2979,
+      "step": 4847
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.4990287494387647,
+      "learning_rate": 5.911114486583525e-06,
+      "loss": 0.9999,
+      "step": 4848
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.4370232459844368,
+      "learning_rate": 5.890916464411933e-06,
+      "loss": 1.2486,
+      "step": 4849
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.44285408221252437,
+      "learning_rate": 5.870751961893417e-06,
+      "loss": 0.9406,
+      "step": 4850
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.44089469073053167,
+      "learning_rate": 5.850620986210198e-06,
+      "loss": 0.9293,
+      "step": 4851
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.4507970454873354,
+      "learning_rate": 5.830523544532451e-06,
+      "loss": 1.0678,
+      "step": 4852
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.4409994998078844,
+      "learning_rate": 5.810459644018506e-06,
+      "loss": 0.9863,
+      "step": 4853
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.4655386779304326,
+      "learning_rate": 5.790429291814659e-06,
+      "loss": 1.2128,
+      "step": 4854
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.38952181869203656,
+      "learning_rate": 5.770432495055311e-06,
+      "loss": 0.9989,
+      "step": 4855
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.4154748846324804,
+      "learning_rate": 5.7504692608628785e-06,
+      "loss": 1.0677,
+      "step": 4856
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.40256471743850014,
+      "learning_rate": 5.730539596347851e-06,
+      "loss": 1.2238,
+      "step": 4857
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.4100197567114653,
+      "learning_rate": 5.710643508608771e-06,
+      "loss": 1.1419,
+      "step": 4858
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.46023601470427516,
+      "learning_rate": 5.69078100473216e-06,
+      "loss": 1.1544,
+      "step": 4859
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.4594340652660134,
+      "learning_rate": 5.670952091792658e-06,
+      "loss": 1.0226,
+      "step": 4860
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.4327581770382769,
+      "learning_rate": 5.6511567768528835e-06,
+      "loss": 1.088,
+      "step": 4861
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.4308898442452065,
+      "learning_rate": 5.631395066963507e-06,
+      "loss": 1.1813,
+      "step": 4862
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.43643145743835315,
+      "learning_rate": 5.611666969163243e-06,
+      "loss": 1.2432,
+      "step": 4863
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.43248535554386136,
+      "learning_rate": 5.591972490478814e-06,
+      "loss": 1.0171,
+      "step": 4864
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.4013949647904614,
+      "learning_rate": 5.572311637924965e-06,
+      "loss": 1.2348,
+      "step": 4865
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.41334602811497023,
+      "learning_rate": 5.552684418504462e-06,
+      "loss": 0.9695,
+      "step": 4866
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.4574784262801688,
+      "learning_rate": 5.533090839208133e-06,
+      "loss": 1.1072,
+      "step": 4867
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.40973269715476207,
+      "learning_rate": 5.513530907014763e-06,
+      "loss": 1.038,
+      "step": 4868
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.41973118516373575,
+      "learning_rate": 5.494004628891214e-06,
+      "loss": 1.2203,
+      "step": 4869
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.43525341959999275,
+      "learning_rate": 5.47451201179231e-06,
+      "loss": 1.1377,
+      "step": 4870
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.4573821234999226,
+      "learning_rate": 5.455053062660931e-06,
+      "loss": 1.0477,
+      "step": 4871
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.46288954343060523,
+      "learning_rate": 5.4356277884278996e-06,
+      "loss": 0.991,
+      "step": 4872
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.44577206641975997,
+      "learning_rate": 5.416236196012126e-06,
+      "loss": 1.2135,
+      "step": 4873
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.4172490682966123,
+      "learning_rate": 5.396878292320451e-06,
+      "loss": 1.163,
+      "step": 4874
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.41201194324856066,
+      "learning_rate": 5.377554084247771e-06,
+      "loss": 1.2682,
+      "step": 4875
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.43715715917046727,
+      "learning_rate": 5.358263578676959e-06,
+      "loss": 1.0769,
+      "step": 4876
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.46575707904038144,
+      "learning_rate": 5.33900678247885e-06,
+      "loss": 1.1033,
+      "step": 4877
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.42797379035639527,
+      "learning_rate": 5.319783702512349e-06,
+      "loss": 1.179,
+      "step": 4878
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.4253852048618836,
+      "learning_rate": 5.300594345624288e-06,
+      "loss": 1.0927,
+      "step": 4879
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.602672408436662,
+      "learning_rate": 5.281438718649501e-06,
+      "loss": 0.9253,
+      "step": 4880
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.4498955206981569,
+      "learning_rate": 5.262316828410807e-06,
+      "loss": 1.1247,
+      "step": 4881
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.3908132232587374,
+      "learning_rate": 5.243228681719048e-06,
+      "loss": 1.0442,
+      "step": 4882
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.4616605283160026,
+      "learning_rate": 5.224174285372974e-06,
+      "loss": 1.1229,
+      "step": 4883
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.4313545203037363,
+      "learning_rate": 5.2051536461594e-06,
+      "loss": 1.0245,
+      "step": 4884
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.3933090599307443,
+      "learning_rate": 5.18616677085304e-06,
+      "loss": 1.2045,
+      "step": 4885
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.43913665980564487,
+      "learning_rate": 5.16721366621663e-06,
+      "loss": 1.0725,
+      "step": 4886
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.40874404578493745,
+      "learning_rate": 5.14829433900087e-06,
+      "loss": 1.2214,
+      "step": 4887
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.4396891196886409,
+      "learning_rate": 5.129408795944412e-06,
+      "loss": 1.2025,
+      "step": 4888
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.448830779475878,
+      "learning_rate": 5.110557043773878e-06,
+      "loss": 1.0613,
+      "step": 4889
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.42164915403444714,
+      "learning_rate": 5.091739089203862e-06,
+      "loss": 1.1267,
+      "step": 4890
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.46258581574987623,
+      "learning_rate": 5.0729549389369245e-06,
+      "loss": 0.979,
+      "step": 4891
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5489373285536467,
+      "learning_rate": 5.054204599663581e-06,
+      "loss": 1.1671,
+      "step": 4892
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.3839394161969445,
+      "learning_rate": 5.035488078062301e-06,
+      "loss": 1.1734,
+      "step": 4893
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.45018372591847616,
+      "learning_rate": 5.0168053807995055e-06,
+      "loss": 1.0516,
+      "step": 4894
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.4819407166915906,
+      "learning_rate": 4.998156514529595e-06,
+      "loss": 1.101,
+      "step": 4895
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.4448001110176714,
+      "learning_rate": 4.979541485894889e-06,
+      "loss": 1.0814,
+      "step": 4896
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.4436623264876953,
+      "learning_rate": 4.960960301525652e-06,
+      "loss": 1.1122,
+      "step": 4897
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.45487435591311864,
+      "learning_rate": 4.942412968040111e-06,
+      "loss": 1.1465,
+      "step": 4898
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.4714921426961155,
+      "learning_rate": 4.923899492044437e-06,
+      "loss": 1.0895,
+      "step": 4899
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.43620464452053076,
+      "learning_rate": 4.905419880132733e-06,
+      "loss": 1.115,
+      "step": 4900
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.43615573135341096,
+      "learning_rate": 4.886974138887035e-06,
+      "loss": 1.0969,
+      "step": 4901
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.4464478382186725,
+      "learning_rate": 4.868562274877353e-06,
+      "loss": 1.1304,
+      "step": 4902
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.512824980314021,
+      "learning_rate": 4.850184294661575e-06,
+      "loss": 0.9848,
+      "step": 4903
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.43862775623083156,
+      "learning_rate": 4.831840204785553e-06,
+      "loss": 1.1333,
+      "step": 4904
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5065317482628473,
+      "learning_rate": 4.813530011783052e-06,
+      "loss": 1.1492,
+      "step": 4905
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.397993078295894,
+      "learning_rate": 4.795253722175786e-06,
+      "loss": 1.1582,
+      "step": 4906
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.42317134283857005,
+      "learning_rate": 4.777011342473392e-06,
+      "loss": 0.9564,
+      "step": 4907
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.4154315184449695,
+      "learning_rate": 4.758802879173374e-06,
+      "loss": 1.0655,
+      "step": 4908
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5014633064944694,
+      "learning_rate": 4.740628338761255e-06,
+      "loss": 1.2347,
+      "step": 4909
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.3782813275462487,
+      "learning_rate": 4.722487727710368e-06,
+      "loss": 0.9809,
+      "step": 4910
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.4336870798165031,
+      "learning_rate": 4.704381052482055e-06,
+      "loss": 1.1635,
+      "step": 4911
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.4167956227398333,
+      "learning_rate": 4.686308319525512e-06,
+      "loss": 1.0579,
+      "step": 4912
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.4469277371533868,
+      "learning_rate": 4.668269535277858e-06,
+      "loss": 1.2141,
+      "step": 4913
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.39734472919322894,
+      "learning_rate": 4.650264706164109e-06,
+      "loss": 1.2492,
+      "step": 4914
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.3829139624022374,
+      "learning_rate": 4.632293838597246e-06,
+      "loss": 1.2458,
+      "step": 4915
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.40386331965609806,
+      "learning_rate": 4.614356938978059e-06,
+      "loss": 1.161,
+      "step": 4916
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.4692531858473411,
+      "learning_rate": 4.596454013695328e-06,
+      "loss": 1.0536,
+      "step": 4917
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.4234290661948093,
+      "learning_rate": 4.578585069125685e-06,
+      "loss": 1.3209,
+      "step": 4918
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.41006561182800083,
+      "learning_rate": 4.560750111633638e-06,
+      "loss": 1.1253,
+      "step": 4919
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.425999745575952,
+      "learning_rate": 4.542949147571684e-06,
+      "loss": 1.1451,
+      "step": 4920
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.4219365712896067,
+      "learning_rate": 4.5251821832800746e-06,
+      "loss": 1.1406,
+      "step": 4921
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.46545081965337115,
+      "learning_rate": 4.50744922508708e-06,
+      "loss": 1.0251,
+      "step": 4922
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5133509101642485,
+      "learning_rate": 4.489750279308757e-06,
+      "loss": 1.0822,
+      "step": 4923
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.39599999877429637,
+      "learning_rate": 4.472085352249134e-06,
+      "loss": 1.1187,
+      "step": 4924
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.4093168867104127,
+      "learning_rate": 4.454454450200051e-06,
+      "loss": 1.0445,
+      "step": 4925
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.4199963409957844,
+      "learning_rate": 4.436857579441289e-06,
+      "loss": 1.0459,
+      "step": 4926
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.4302560400898706,
+      "learning_rate": 4.4192947462404654e-06,
+      "loss": 1.3078,
+      "step": 4927
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.4508875498004485,
+      "learning_rate": 4.401765956853088e-06,
+      "loss": 1.0197,
+      "step": 4928
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.3817944541932066,
+      "learning_rate": 4.3842712175225285e-06,
+      "loss": 1.1626,
+      "step": 4929
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.4892408295426448,
+      "learning_rate": 4.366810534480037e-06,
+      "loss": 1.1357,
+      "step": 4930
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.48438464309792867,
+      "learning_rate": 4.349383913944771e-06,
+      "loss": 0.9207,
+      "step": 4931
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.4604276261855715,
+      "learning_rate": 4.331991362123689e-06,
+      "loss": 1.0702,
+      "step": 4932
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.4620340354655377,
+      "learning_rate": 4.314632885211667e-06,
+      "loss": 1.0135,
+      "step": 4933
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.3960260093813619,
+      "learning_rate": 4.297308489391405e-06,
+      "loss": 1.0992,
+      "step": 4934
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5541117543511827,
+      "learning_rate": 4.280018180833501e-06,
+      "loss": 1.0972,
+      "step": 4935
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.41673436638672595,
+      "learning_rate": 4.262761965696404e-06,
+      "loss": 1.2126,
+      "step": 4936
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.45193561502756163,
+      "learning_rate": 4.24553985012639e-06,
+      "loss": 1.0219,
+      "step": 4937
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.48465534308392316,
+      "learning_rate": 4.228351840257605e-06,
+      "loss": 1.0617,
+      "step": 4938
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.48305679207161173,
+      "learning_rate": 4.211197942212086e-06,
+      "loss": 1.1471,
+      "step": 4939
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.44399297884713035,
+      "learning_rate": 4.194078162099657e-06,
+      "loss": 1.041,
+      "step": 4940
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.4465920709796698,
+      "learning_rate": 4.176992506018018e-06,
+      "loss": 1.1673,
+      "step": 4941
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.41467256442353,
+      "learning_rate": 4.159940980052734e-06,
+      "loss": 0.9824,
+      "step": 4942
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.46139386057016946,
+      "learning_rate": 4.142923590277181e-06,
+      "loss": 1.1472,
+      "step": 4943
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.40796958110668335,
+      "learning_rate": 4.12594034275261e-06,
+      "loss": 1.2072,
+      "step": 4944
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.46598363449303176,
+      "learning_rate": 4.1089912435280705e-06,
+      "loss": 0.9943,
+      "step": 4945
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.40681977779082845,
+      "learning_rate": 4.09207629864049e-06,
+      "loss": 1.1054,
+      "step": 4946
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.4818403087382363,
+      "learning_rate": 4.075195514114593e-06,
+      "loss": 1.1034,
+      "step": 4947
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.4322773877212745,
+      "learning_rate": 4.058348895962971e-06,
+      "loss": 1.0352,
+      "step": 4948
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.4579510099957429,
+      "learning_rate": 4.041536450186012e-06,
+      "loss": 0.9358,
+      "step": 4949
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.39635466942378605,
+      "learning_rate": 4.024758182771982e-06,
+      "loss": 1.0826,
+      "step": 4950
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.4196954266871242,
+      "learning_rate": 4.008014099696922e-06,
+      "loss": 1.1203,
+      "step": 4951
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.45592026595807866,
+      "learning_rate": 3.991304206924728e-06,
+      "loss": 1.2028,
+      "step": 4952
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.4820608904093585,
+      "learning_rate": 3.974628510407097e-06,
+      "loss": 1.1314,
+      "step": 4953
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.41518674577606507,
+      "learning_rate": 3.957987016083553e-06,
+      "loss": 1.014,
+      "step": 4954
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.4851493542591833,
+      "learning_rate": 3.941379729881456e-06,
+      "loss": 0.9127,
+      "step": 4955
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.43973272726920304,
+      "learning_rate": 3.924806657715951e-06,
+      "loss": 1.0274,
+      "step": 4956
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.4569722934717319,
+      "learning_rate": 3.908267805490051e-06,
+      "loss": 1.0983,
+      "step": 4957
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.4947740049729834,
+      "learning_rate": 3.8917631790944985e-06,
+      "loss": 1.0927,
+      "step": 4958
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.4654855555234969,
+      "learning_rate": 3.875292784407924e-06,
+      "loss": 1.1577,
+      "step": 4959
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5541901688027825,
+      "learning_rate": 3.858856627296714e-06,
+      "loss": 1.0992,
+      "step": 4960
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.4056412460670293,
+      "learning_rate": 3.842454713615084e-06,
+      "loss": 1.1618,
+      "step": 4961
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.38950561583540394,
+      "learning_rate": 3.826087049205041e-06,
+      "loss": 1.0646,
+      "step": 4962
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.4893076382588843,
+      "learning_rate": 3.8097536398963963e-06,
+      "loss": 1.0425,
+      "step": 4963
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.3953728147655422,
+      "learning_rate": 3.793454491506776e-06,
+      "loss": 1.1073,
+      "step": 4964
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.46623209242974184,
+      "learning_rate": 3.7771896098415695e-06,
+      "loss": 1.1094,
+      "step": 4965
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.4068747995802555,
+      "learning_rate": 3.760959000694009e-06,
+      "loss": 1.1601,
+      "step": 4966
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.43346984775215086,
+      "learning_rate": 3.74476266984507e-06,
+      "loss": 0.9712,
+      "step": 4967
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5086543094269163,
+      "learning_rate": 3.7286006230635495e-06,
+      "loss": 1.0728,
+      "step": 4968
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.42840788370538035,
+      "learning_rate": 3.7124728661060317e-06,
+      "loss": 1.0808,
+      "step": 4969
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.41167317863645997,
+      "learning_rate": 3.6963794047168763e-06,
+      "loss": 1.1715,
+      "step": 4970
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.42047532641883933,
+      "learning_rate": 3.6803202446282214e-06,
+      "loss": 0.9767,
+      "step": 4971
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.4808955679384015,
+      "learning_rate": 3.664295391559991e-06,
+      "loss": 1.0839,
+      "step": 4972
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.42986796311958425,
+      "learning_rate": 3.648304851219919e-06,
+      "loss": 1.1364,
+      "step": 4973
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.4841473784052731,
+      "learning_rate": 3.6323486293034835e-06,
+      "loss": 1.147,
+      "step": 4974
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.45555291167576106,
+      "learning_rate": 3.6164267314939713e-06,
+      "loss": 1.0683,
+      "step": 4975
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.4607546133414873,
+      "learning_rate": 3.600539163462402e-06,
+      "loss": 1.1092,
+      "step": 4976
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.41694601837394224,
+      "learning_rate": 3.5846859308675927e-06,
+      "loss": 1.0518,
+      "step": 4977
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.43904309636910827,
+      "learning_rate": 3.568867039356116e-06,
+      "loss": 1.0524,
+      "step": 4978
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.4860304238284254,
+      "learning_rate": 3.5530824945623542e-06,
+      "loss": 1.0253,
+      "step": 4979
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.4830360147909272,
+      "learning_rate": 3.5373323021083873e-06,
+      "loss": 1.1108,
+      "step": 4980
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.4670931051875092,
+      "learning_rate": 3.5216164676041386e-06,
+      "loss": 1.0473,
+      "step": 4981
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.4467627044265851,
+      "learning_rate": 3.505934996647242e-06,
+      "loss": 0.9911,
+      "step": 4982
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.4128719395657158,
+      "learning_rate": 3.4902878948230854e-06,
+      "loss": 1.1186,
+      "step": 4983
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.4228754943316035,
+      "learning_rate": 3.474675167704855e-06,
+      "loss": 1.1794,
+      "step": 4984
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5736097305732496,
+      "learning_rate": 3.459096820853469e-06,
+      "loss": 1.0842,
+      "step": 4985
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.3820495015059072,
+      "learning_rate": 3.4435528598175894e-06,
+      "loss": 1.1468,
+      "step": 4986
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.4065671534619588,
+      "learning_rate": 3.4280432901336425e-06,
+      "loss": 1.0471,
+      "step": 4987
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.418032158397448,
+      "learning_rate": 3.4125681173258316e-06,
+      "loss": 1.1641,
+      "step": 4988
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.4569004309103617,
+      "learning_rate": 3.3971273469060592e-06,
+      "loss": 0.9901,
+      "step": 4989
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.40975014843054636,
+      "learning_rate": 3.381720984374015e-06,
+      "loss": 1.0095,
+      "step": 4990
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.4579796505691732,
+      "learning_rate": 3.366349035217109e-06,
+      "loss": 1.1898,
+      "step": 4991
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.3975887588265909,
+      "learning_rate": 3.351011504910517e-06,
+      "loss": 1.0946,
+      "step": 4992
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5038332596250411,
+      "learning_rate": 3.3357083989171256e-06,
+      "loss": 1.1832,
+      "step": 4993
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.4159956299257592,
+      "learning_rate": 3.320439722687574e-06,
+      "loss": 1.1153,
+      "step": 4994
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.412157883027807,
+      "learning_rate": 3.3052054816602452e-06,
+      "loss": 1.1376,
+      "step": 4995
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.41447691989367025,
+      "learning_rate": 3.2900056812612436e-06,
+      "loss": 1.0868,
+      "step": 4996
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.4070541132029471,
+      "learning_rate": 3.274840326904438e-06,
+      "loss": 0.9454,
+      "step": 4997
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5067648627632406,
+      "learning_rate": 3.259709423991375e-06,
+      "loss": 1.082,
+      "step": 4998
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.4227266274302929,
+      "learning_rate": 3.2446129779113877e-06,
+      "loss": 0.9999,
+      "step": 4999
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.4379090129754789,
+      "learning_rate": 3.229550994041497e-06,
+      "loss": 1.0227,
+      "step": 5000
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.4755601553908275,
+      "learning_rate": 3.214523477746456e-06,
+      "loss": 1.2137,
+      "step": 5001
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.4480945834956784,
+      "learning_rate": 3.199530434378739e-06,
+      "loss": 0.9184,
+      "step": 5002
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.42068318612788147,
+      "learning_rate": 3.184571869278574e-06,
+      "loss": 1.1197,
+      "step": 5003
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.39093539498007135,
+      "learning_rate": 3.169647787773866e-06,
+      "loss": 1.2473,
+      "step": 5004
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.43473488986640496,
+      "learning_rate": 3.154758195180252e-06,
+      "loss": 1.0577,
+      "step": 5005
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.40257409696844243,
+      "learning_rate": 3.1399030968011e-06,
+      "loss": 1.1783,
+      "step": 5006
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.4326882910403301,
+      "learning_rate": 3.1250824979274675e-06,
+      "loss": 1.1376,
+      "step": 5007
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.437600477011577,
+      "learning_rate": 3.1102964038381644e-06,
+      "loss": 1.096,
+      "step": 5008
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.4280372654597047,
+      "learning_rate": 3.095544819799656e-06,
+      "loss": 1.1099,
+      "step": 5009
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.4509656763316527,
+      "learning_rate": 3.080827751066151e-06,
+      "loss": 1.0457,
+      "step": 5010
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.41203442259805934,
+      "learning_rate": 3.0661452028795336e-06,
+      "loss": 1.1279,
+      "step": 5011
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.4309402491198156,
+      "learning_rate": 3.0514971804694535e-06,
+      "loss": 0.9996,
+      "step": 5012
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.4524865613357975,
+      "learning_rate": 3.036883689053205e-06,
+      "loss": 1.1688,
+      "step": 5013
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.4879085032094116,
+      "learning_rate": 3.0223047338358013e-06,
+      "loss": 1.0826,
+      "step": 5014
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.4050956059448305,
+      "learning_rate": 3.007760320009967e-06,
+      "loss": 1.07,
+      "step": 5015
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.4411560867678746,
+      "learning_rate": 2.9932504527560913e-06,
+      "loss": 1.3005,
+      "step": 5016
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.4326682030256378,
+      "learning_rate": 2.9787751372423066e-06,
+      "loss": 1.2108,
+      "step": 5017
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.45341711106393995,
+      "learning_rate": 2.9643343786244006e-06,
+      "loss": 1.1188,
+      "step": 5018
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.4594167021509271,
+      "learning_rate": 2.9499281820458692e-06,
+      "loss": 1.0802,
+      "step": 5019
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.44938797719720636,
+      "learning_rate": 2.9355565526378635e-06,
+      "loss": 0.8508,
+      "step": 5020
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.43885672526277897,
+      "learning_rate": 2.921219495519301e-06,
+      "loss": 1.1829,
+      "step": 5021
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.4789569488871049,
+      "learning_rate": 2.906917015796695e-06,
+      "loss": 1.01,
+      "step": 5022
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.44455574875633563,
+      "learning_rate": 2.8926491185643058e-06,
+      "loss": 1.1082,
+      "step": 5023
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.45993943301158546,
+      "learning_rate": 2.8784158089040556e-06,
+      "loss": 1.2604,
+      "step": 5024
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.46928597320822574,
+      "learning_rate": 2.8642170918855347e-06,
+      "loss": 1.1079,
+      "step": 5025
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.4029760534394289,
+      "learning_rate": 2.8500529725660417e-06,
+      "loss": 1.1609,
+      "step": 5026
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.41640227881182984,
+      "learning_rate": 2.835923455990508e-06,
+      "loss": 1.1817,
+      "step": 5027
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.4268514645515142,
+      "learning_rate": 2.821828547191585e-06,
+      "loss": 1.2074,
+      "step": 5028
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.42542280107467023,
+      "learning_rate": 2.8077682511895687e-06,
+      "loss": 1.0657,
+      "step": 5029
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.4242009332542491,
+      "learning_rate": 2.7937425729924527e-06,
+      "loss": 1.0456,
+      "step": 5030
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.4105197676074723,
+      "learning_rate": 2.7797515175958744e-06,
+      "loss": 1.1837,
+      "step": 5031
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.4435857077439962,
+      "learning_rate": 2.7657950899831585e-06,
+      "loss": 1.1205,
+      "step": 5032
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.4523989563790952,
+      "learning_rate": 2.7518732951252847e-06,
+      "loss": 1.1711,
+      "step": 5033
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.3840058665272871,
+      "learning_rate": 2.7379861379808856e-06,
+      "loss": 1.0965,
+      "step": 5034
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.44631487644182943,
+      "learning_rate": 2.7241336234962944e-06,
+      "loss": 1.0225,
+      "step": 5035
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5152647450340428,
+      "learning_rate": 2.7103157566054525e-06,
+      "loss": 1.0325,
+      "step": 5036
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.4276644520485558,
+      "learning_rate": 2.6965325422300237e-06,
+      "loss": 1.102,
+      "step": 5037
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5053765146151775,
+      "learning_rate": 2.682783985279269e-06,
+      "loss": 1.0337,
+      "step": 5038
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.49327375371335785,
+      "learning_rate": 2.6690700906501496e-06,
+      "loss": 0.9589,
+      "step": 5039
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.4395169620275134,
+      "learning_rate": 2.655390863227247e-06,
+      "loss": 0.9991,
+      "step": 5040
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.44531833490676626,
+      "learning_rate": 2.6417463078828307e-06,
+      "loss": 1.0118,
+      "step": 5041
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.42472895347078693,
+      "learning_rate": 2.628136429476791e-06,
+      "loss": 1.0156,
+      "step": 5042
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.4871886005576437,
+      "learning_rate": 2.6145612328566717e-06,
+      "loss": 0.9081,
+      "step": 5043
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.45055876430046865,
+      "learning_rate": 2.6010207228576723e-06,
+      "loss": 1.1141,
+      "step": 5044
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.45510873789798617,
+      "learning_rate": 2.5875149043026348e-06,
+      "loss": 1.0934,
+      "step": 5045
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.4306977631098254,
+      "learning_rate": 2.574043782002056e-06,
+      "loss": 1.1969,
+      "step": 5046
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.46227162431499264,
+      "learning_rate": 2.5606073607540305e-06,
+      "loss": 0.9175,
+      "step": 5047
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.41487554940457627,
+      "learning_rate": 2.5472056453443638e-06,
+      "loss": 1.0821,
+      "step": 5048
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.46665758421865416,
+      "learning_rate": 2.5338386405464377e-06,
+      "loss": 1.075,
+      "step": 5049
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.38593200653202564,
+      "learning_rate": 2.52050635112131e-06,
+      "loss": 1.0063,
+      "step": 5050
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.42440487834873286,
+      "learning_rate": 2.5072087818176382e-06,
+      "loss": 1.1531,
+      "step": 5051
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.4469533075326513,
+      "learning_rate": 2.4939459373717556e-06,
+      "loss": 1.3444,
+      "step": 5052
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.47184247958067355,
+      "learning_rate": 2.4807178225075945e-06,
+      "loss": 1.1043,
+      "step": 5053
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.4606718508593389,
+      "learning_rate": 2.4675244419367305e-06,
+      "loss": 1.1342,
+      "step": 5054
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.390989396445886,
+      "learning_rate": 2.4543658003583604e-06,
+      "loss": 1.0052,
+      "step": 5055
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.4230015125333437,
+      "learning_rate": 2.4412419024593127e-06,
+      "loss": 1.1195,
+      "step": 5056
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.44641826461343154,
+      "learning_rate": 2.428152752914059e-06,
+      "loss": 1.0802,
+      "step": 5057
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.4130940654149731,
+      "learning_rate": 2.4150983563846486e-06,
+      "loss": 1.1516,
+      "step": 5058
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5256053570210785,
+      "learning_rate": 2.402078717520795e-06,
+      "loss": 1.0633,
+      "step": 5059
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.42241822670316487,
+      "learning_rate": 2.3890938409597995e-06,
+      "loss": 1.1811,
+      "step": 5060
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.4347829934173333,
+      "learning_rate": 2.37614373132663e-06,
+      "loss": 0.9853,
+      "step": 5061
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.49577465150870487,
+      "learning_rate": 2.3632283932338074e-06,
+      "loss": 1.1448,
+      "step": 5062
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.48385262733210843,
+      "learning_rate": 2.3503478312815298e-06,
+      "loss": 0.8948,
+      "step": 5063
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.44711648846761476,
+      "learning_rate": 2.3375020500575606e-06,
+      "loss": 1.1381,
+      "step": 5064
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.3783224386367112,
+      "learning_rate": 2.324691054137296e-06,
+      "loss": 1.3505,
+      "step": 5065
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.4358721199661109,
+      "learning_rate": 2.3119148480837406e-06,
+      "loss": 1.3179,
+      "step": 5066
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.4082598663820162,
+      "learning_rate": 2.2991734364475215e-06,
+      "loss": 1.1439,
+      "step": 5067
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.45972328855164746,
+      "learning_rate": 2.286466823766831e-06,
+      "loss": 1.0525,
+      "step": 5068
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.47208325457076267,
+      "learning_rate": 2.2737950145675147e-06,
+      "loss": 1.037,
+      "step": 5069
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.45248859056938756,
+      "learning_rate": 2.261158013362996e-06,
+      "loss": 1.2847,
+      "step": 5070
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.433766324786765,
+      "learning_rate": 2.248555824654297e-06,
+      "loss": 1.1313,
+      "step": 5071
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.47315456765918634,
+      "learning_rate": 2.2359884529300603e-06,
+      "loss": 1.1969,
+      "step": 5072
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.43627319663511227,
+      "learning_rate": 2.2234559026665157e-06,
+      "loss": 1.1975,
+      "step": 5073
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.3961115077330131,
+      "learning_rate": 2.2109581783275047e-06,
+      "loss": 1.0223,
+      "step": 5074
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.4888318457427915,
+      "learning_rate": 2.19849528436441e-06,
+      "loss": 1.2438,
+      "step": 5075
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.4907958162383656,
+      "learning_rate": 2.186067225216293e-06,
+      "loss": 1.1315,
+      "step": 5076
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.36478339595762976,
+      "learning_rate": 2.1736740053097338e-06,
+      "loss": 1.1157,
+      "step": 5077
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.47096829340528396,
+      "learning_rate": 2.1613156290589465e-06,
+      "loss": 1.1616,
+      "step": 5078
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.48102763812713445,
+      "learning_rate": 2.148992100865721e-06,
+      "loss": 0.9868,
+      "step": 5079
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.44583768966773873,
+      "learning_rate": 2.136703425119424e-06,
+      "loss": 1.0164,
+      "step": 5080
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.45022687602530354,
+      "learning_rate": 2.1244496061970433e-06,
+      "loss": 1.0675,
+      "step": 5081
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.3941256253920535,
+      "learning_rate": 2.1122306484631204e-06,
+      "loss": 1.0618,
+      "step": 5082
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.42568469323802394,
+      "learning_rate": 2.1000465562697856e-06,
+      "loss": 1.0373,
+      "step": 5083
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.4670713192046501,
+      "learning_rate": 2.087897333956734e-06,
+      "loss": 1.0899,
+      "step": 5084
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.4316308541618242,
+      "learning_rate": 2.075782985851293e-06,
+      "loss": 0.9661,
+      "step": 5085
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.3868104755565469,
+      "learning_rate": 2.0637035162683115e-06,
+      "loss": 0.9623,
+      "step": 5086
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.4177619373396268,
+      "learning_rate": 2.0516589295102583e-06,
+      "loss": 1.228,
+      "step": 5087
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.412675065368626,
+      "learning_rate": 2.0396492298671467e-06,
+      "loss": 1.0331,
+      "step": 5088
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.43407825460746263,
+      "learning_rate": 2.0276744216165654e-06,
+      "loss": 1.1046,
+      "step": 5089
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.48757607730227653,
+      "learning_rate": 2.0157345090237145e-06,
+      "loss": 1.0994,
+      "step": 5090
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.4550499795414422,
+      "learning_rate": 2.003829496341325e-06,
+      "loss": 1.08,
+      "step": 5091
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.4545894022486251,
+      "learning_rate": 1.9919593878096944e-06,
+      "loss": 1.1042,
+      "step": 5092
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.4080253017036463,
+      "learning_rate": 1.980124187656718e-06,
+      "loss": 1.1942,
+      "step": 5093
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.44947721021988657,
+      "learning_rate": 1.968323900097835e-06,
+      "loss": 1.1067,
+      "step": 5094
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.43433052659189586,
+      "learning_rate": 1.956558529336061e-06,
+      "loss": 1.0728,
+      "step": 5095
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.4093459114174658,
+      "learning_rate": 1.944828079561978e-06,
+      "loss": 1.2275,
+      "step": 5096
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.4483838713724983,
+      "learning_rate": 1.9331325549536984e-06,
+      "loss": 1.1238,
+      "step": 5097
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.4049969575197429,
+      "learning_rate": 1.921471959676957e-06,
+      "loss": 1.2421,
+      "step": 5098
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.37615625791440505,
+      "learning_rate": 1.9098462978849873e-06,
+      "loss": 1.2228,
+      "step": 5099
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.39805150483527313,
+      "learning_rate": 1.898255573718588e-06,
+      "loss": 1.2306,
+      "step": 5100
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.46098876295857105,
+      "learning_rate": 1.8866997913061457e-06,
+      "loss": 1.0211,
+      "step": 5101
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7040765064940239,
+      "learning_rate": 1.87517895476359e-06,
+      "loss": 1.0672,
+      "step": 5102
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5072350919723623,
+      "learning_rate": 1.8636930681943831e-06,
+      "loss": 1.2773,
+      "step": 5103
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.4248922957739785,
+      "learning_rate": 1.8522421356895637e-06,
+      "loss": 1.0848,
+      "step": 5104
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.47797499462693765,
+      "learning_rate": 1.8408261613277022e-06,
+      "loss": 1.0665,
+      "step": 5105
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.491011346514947,
+      "learning_rate": 1.8294451491749464e-06,
+      "loss": 1.0841,
+      "step": 5106
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.49636315688314053,
+      "learning_rate": 1.8180991032849426e-06,
+      "loss": 1.0327,
+      "step": 5107
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.38669948537007426,
+      "learning_rate": 1.8067880276989134e-06,
+      "loss": 1.088,
+      "step": 5108
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.39244342211763944,
+      "learning_rate": 1.7955119264456476e-06,
+      "loss": 1.2192,
+      "step": 5109
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.4301103424130786,
+      "learning_rate": 1.7842708035414434e-06,
+      "loss": 1.1532,
+      "step": 5110
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.42616145043840126,
+      "learning_rate": 1.7730646629901315e-06,
+      "loss": 1.1633,
+      "step": 5111
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.3920320054903487,
+      "learning_rate": 1.761893508783119e-06,
+      "loss": 1.3622,
+      "step": 5112
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.412151474122597,
+      "learning_rate": 1.7507573448993231e-06,
+      "loss": 1.1202,
+      "step": 5113
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.44513063770406053,
+      "learning_rate": 1.7396561753052154e-06,
+      "loss": 1.1669,
+      "step": 5114
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.4066581551570856,
+      "learning_rate": 1.7285900039547998e-06,
+      "loss": 1.2122,
+      "step": 5115
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.47152819098107057,
+      "learning_rate": 1.717558834789601e-06,
+      "loss": 1.0376,
+      "step": 5116
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.41400286435079203,
+      "learning_rate": 1.7065626717386874e-06,
+      "loss": 1.199,
+      "step": 5117
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.415947824538069,
+      "learning_rate": 1.695601518718659e-06,
+      "loss": 1.2497,
+      "step": 5118
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.41334107821710603,
+      "learning_rate": 1.684675379633649e-06,
+      "loss": 1.0593,
+      "step": 5119
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.46094229610107845,
+      "learning_rate": 1.6737842583753216e-06,
+      "loss": 1.1138,
+      "step": 5120
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.49919263242592654,
+      "learning_rate": 1.662928158822863e-06,
+      "loss": 1.0554,
+      "step": 5121
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.45541288187118417,
+      "learning_rate": 1.6521070848429576e-06,
+      "loss": 1.1653,
+      "step": 5122
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.4821774639343692,
+      "learning_rate": 1.6413210402898893e-06,
+      "loss": 0.9867,
+      "step": 5123
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.4248383672137894,
+      "learning_rate": 1.6305700290053627e-06,
+      "loss": 1.0833,
+      "step": 5124
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.42392011361700627,
+      "learning_rate": 1.6198540548187036e-06,
+      "loss": 1.0964,
+      "step": 5125
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.45145348726637646,
+      "learning_rate": 1.609173121546681e-06,
+      "loss": 1.0271,
+      "step": 5126
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.4086645460130038,
+      "learning_rate": 1.5985272329936407e-06,
+      "loss": 1.1078,
+      "step": 5127
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.48325146635131994,
+      "learning_rate": 1.5879163929514052e-06,
+      "loss": 1.106,
+      "step": 5128
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.394900396599991,
+      "learning_rate": 1.577340605199351e-06,
+      "loss": 0.8408,
+      "step": 5129
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5089399741107743,
+      "learning_rate": 1.5667998735043321e-06,
+      "loss": 1.0517,
+      "step": 5130
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5104014814630932,
+      "learning_rate": 1.5562942016207338e-06,
+      "loss": 1.1218,
+      "step": 5131
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.42905300244827527,
+      "learning_rate": 1.545823593290463e-06,
+      "loss": 1.0174,
+      "step": 5132
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.4245980976548461,
+      "learning_rate": 1.5353880522429032e-06,
+      "loss": 1.2119,
+      "step": 5133
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.4457914442887692,
+      "learning_rate": 1.5249875821950034e-06,
+      "loss": 1.1526,
+      "step": 5134
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.44176479097945254,
+      "learning_rate": 1.5146221868511668e-06,
+      "loss": 1.1314,
+      "step": 5135
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.47426863535177183,
+      "learning_rate": 1.5042918699033403e-06,
+      "loss": 0.9391,
+      "step": 5136
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.3840440780305627,
+      "learning_rate": 1.4939966350309476e-06,
+      "loss": 1.1078,
+      "step": 5137
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.3989677480625158,
+      "learning_rate": 1.483736485900944e-06,
+      "loss": 1.2131,
+      "step": 5138
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.45108265342756637,
+      "learning_rate": 1.473511426167784e-06,
+      "loss": 0.9391,
+      "step": 5139
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.4482963868918586,
+      "learning_rate": 1.4633214594734102e-06,
+      "loss": 1.1114,
+      "step": 5140
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.4460240874362524,
+      "learning_rate": 1.4531665894472634e-06,
+      "loss": 0.9274,
+      "step": 5141
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.43113511451593817,
+      "learning_rate": 1.4430468197062841e-06,
+      "loss": 1.1607,
+      "step": 5142
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.4456433042257504,
+      "learning_rate": 1.4329621538549444e-06,
+      "loss": 1.0279,
+      "step": 5143
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.4342116906750268,
+      "learning_rate": 1.4229125954851707e-06,
+      "loss": 1.2443,
+      "step": 5144
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.405957867477203,
+      "learning_rate": 1.4128981481764115e-06,
+      "loss": 1.1242,
+      "step": 5145
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.4700785203886421,
+      "learning_rate": 1.4029188154955908e-06,
+      "loss": 1.0344,
+      "step": 5146
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.4047070005876291,
+      "learning_rate": 1.3929746009971433e-06,
+      "loss": 1.1608,
+      "step": 5147
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5198232567349258,
+      "learning_rate": 1.3830655082229803e-06,
+      "loss": 1.009,
+      "step": 5148
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.45827901969255724,
+      "learning_rate": 1.373191540702523e-06,
+      "loss": 1.0511,
+      "step": 5149
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.48604434156864174,
+      "learning_rate": 1.363352701952636e-06,
+      "loss": 1.0071,
+      "step": 5150
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.4485964002714696,
+      "learning_rate": 1.353548995477738e-06,
+      "loss": 1.1073,
+      "step": 5151
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.42804997913336973,
+      "learning_rate": 1.3437804247696917e-06,
+      "loss": 1.1047,
+      "step": 5152
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5623674336972254,
+      "learning_rate": 1.3340469933078359e-06,
+      "loss": 1.2201,
+      "step": 5153
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.43722718250174875,
+      "learning_rate": 1.324348704559042e-06,
+      "loss": 0.9352,
+      "step": 5154
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.44083784210938226,
+      "learning_rate": 1.3146855619776134e-06,
+      "loss": 1.0327,
+      "step": 5155
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.42502410970031734,
+      "learning_rate": 1.3050575690053636e-06,
+      "loss": 1.3026,
+      "step": 5156
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.4002092614085613,
+      "learning_rate": 1.2954647290715605e-06,
+      "loss": 1.2742,
+      "step": 5157
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.46276625596043547,
+      "learning_rate": 1.2859070455930045e-06,
+      "loss": 1.1238,
+      "step": 5158
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.4187272820550083,
+      "learning_rate": 1.2763845219739056e-06,
+      "loss": 1.0286,
+      "step": 5159
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5268384632252388,
+      "learning_rate": 1.2668971616060066e-06,
+      "loss": 1.102,
+      "step": 5160
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.46517267328529316,
+      "learning_rate": 1.257444967868482e-06,
+      "loss": 1.1018,
+      "step": 5161
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.42445445430881473,
+      "learning_rate": 1.2480279441280274e-06,
+      "loss": 1.0109,
+      "step": 5162
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.4540673019879382,
+      "learning_rate": 1.2386460937387822e-06,
+      "loss": 1.1851,
+      "step": 5163
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5302462969465027,
+      "learning_rate": 1.2292994200423403e-06,
+      "loss": 0.909,
+      "step": 5164
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.4285175626667748,
+      "learning_rate": 1.219987926367816e-06,
+      "loss": 1.0531,
+      "step": 5165
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.43620174145107177,
+      "learning_rate": 1.2107116160317344e-06,
+      "loss": 0.9958,
+      "step": 5166
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.41955853464284243,
+      "learning_rate": 1.2014704923381414e-06,
+      "loss": 1.2081,
+      "step": 5167
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.4581231157863012,
+      "learning_rate": 1.1922645585785264e-06,
+      "loss": 1.1386,
+      "step": 5168
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.4173200713435814,
+      "learning_rate": 1.183093818031844e-06,
+      "loss": 1.178,
+      "step": 5169
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.4047581490202959,
+      "learning_rate": 1.1739582739645038e-06,
+      "loss": 1.0593,
+      "step": 5170
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.4603796070935751,
+      "learning_rate": 1.1648579296304253e-06,
+      "loss": 1.1125,
+      "step": 5171
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.4413034720823877,
+      "learning_rate": 1.1557927882709264e-06,
+      "loss": 1.0543,
+      "step": 5172
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.42183780302993823,
+      "learning_rate": 1.1467628531148357e-06,
+      "loss": 1.022,
+      "step": 5173
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.4333567960237378,
+      "learning_rate": 1.1377681273784135e-06,
+      "loss": 1.2949,
+      "step": 5174
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.4185840021138616,
+      "learning_rate": 1.1288086142653864e-06,
+      "loss": 1.0062,
+      "step": 5175
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.4352511811126047,
+      "learning_rate": 1.119884316966957e-06,
+      "loss": 1.0524,
+      "step": 5176
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.4487426171290615,
+      "learning_rate": 1.1109952386617605e-06,
+      "loss": 1.1211,
+      "step": 5177
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.4670689543658417,
+      "learning_rate": 1.1021413825158977e-06,
+      "loss": 1.1876,
+      "step": 5178
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.4412160317816144,
+      "learning_rate": 1.0933227516829347e-06,
+      "loss": 0.9257,
+      "step": 5179
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5148651416192587,
+      "learning_rate": 1.0845393493038592e-06,
+      "loss": 1.1645,
+      "step": 5180
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.48202387133019586,
+      "learning_rate": 1.0757911785071463e-06,
+      "loss": 1.0892,
+      "step": 5181
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.4572295036891286,
+      "learning_rate": 1.0670782424087033e-06,
+      "loss": 1.1365,
+      "step": 5182
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.426966651462916,
+      "learning_rate": 1.0584005441119038e-06,
+      "loss": 1.1376,
+      "step": 5183
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.41015373550583367,
+      "learning_rate": 1.0497580867075418e-06,
+      "loss": 1.0781,
+      "step": 5184
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.46443720977473385,
+      "learning_rate": 1.0411508732738884e-06,
+      "loss": 1.0804,
+      "step": 5185
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.4749337524291308,
+      "learning_rate": 1.0325789068766467e-06,
+      "loss": 1.2445,
+      "step": 5186
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.46617271622612577,
+      "learning_rate": 1.0240421905689745e-06,
+      "loss": 0.9067,
+      "step": 5187
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.41379910810477283,
+      "learning_rate": 1.0155407273914619e-06,
+      "loss": 1.187,
+      "step": 5188
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.487696134959113,
+      "learning_rate": 1.0070745203721532e-06,
+      "loss": 1.032,
+      "step": 5189
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.435531164356504,
+      "learning_rate": 9.986435725265141e-07,
+      "loss": 1.1973,
+      "step": 5190
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.37262111466260806,
+      "learning_rate": 9.902478868574982e-07,
+      "loss": 1.0955,
+      "step": 5191
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5311408232944393,
+      "learning_rate": 9.818874663554357e-07,
+      "loss": 0.9972,
+      "step": 5192
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.40932305555821097,
+      "learning_rate": 9.735623139981664e-07,
+      "loss": 1.121,
+      "step": 5193
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.45402214389780116,
+      "learning_rate": 9.65272432750919e-07,
+      "loss": 1.2712,
+      "step": 5194
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.4525172389324443,
+      "learning_rate": 9.570178255663532e-07,
+      "loss": 0.9672,
+      "step": 5195
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.4625049975210951,
+      "learning_rate": 9.487984953846174e-07,
+      "loss": 1.1967,
+      "step": 5196
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5019399898335791,
+      "learning_rate": 9.406144451332366e-07,
+      "loss": 1.1814,
+      "step": 5197
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.40499378843379086,
+      "learning_rate": 9.324656777272123e-07,
+      "loss": 1.105,
+      "step": 5198
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.43950234783946346,
+      "learning_rate": 9.24352196068945e-07,
+      "loss": 1.161,
+      "step": 5199
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.47580443172405296,
+      "learning_rate": 9.162740030482897e-07,
+      "loss": 0.9647,
+      "step": 5200
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.40639144423558543,
+      "learning_rate": 9.082311015425338e-07,
+      "loss": 1.0658,
+      "step": 5201
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.4402699806162486,
+      "learning_rate": 9.002234944163967e-07,
+      "loss": 1.119,
+      "step": 5202
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.42285942169600194,
+      "learning_rate": 8.922511845219971e-07,
+      "loss": 1.224,
+      "step": 5203
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.46513936573895964,
+      "learning_rate": 8.843141746989081e-07,
+      "loss": 1.2303,
+      "step": 5204
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.4757798229248753,
+      "learning_rate": 8.764124677741236e-07,
+      "loss": 0.9901,
+      "step": 5205
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.4309871266857302,
+      "learning_rate": 8.685460665620704e-07,
+      "loss": 1.0852,
+      "step": 5206
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5457610238129948,
+      "learning_rate": 8.607149738645848e-07,
+      "loss": 1.01,
+      "step": 5207
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.41088351975324944,
+      "learning_rate": 8.529191924709468e-07,
+      "loss": 1.0457,
+      "step": 5208
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.4551568433389947,
+      "learning_rate": 8.451587251578352e-07,
+      "loss": 1.1036,
+      "step": 5209
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.43417713873092983,
+      "learning_rate": 8.374335746893613e-07,
+      "loss": 1.1561,
+      "step": 5210
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.4563258077199253,
+      "learning_rate": 8.297437438170797e-07,
+      "loss": 1.2428,
+      "step": 5211
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.46358957110476795,
+      "learning_rate": 8.220892352799215e-07,
+      "loss": 1.2026,
+      "step": 5212
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.41800455455285906,
+      "learning_rate": 8.144700518042836e-07,
+      "loss": 1.1048,
+      "step": 5213
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.4006138463998516,
+      "learning_rate": 8.068861961039287e-07,
+      "loss": 1.2509,
+      "step": 5214
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.44461117500520075,
+      "learning_rate": 7.993376708800848e-07,
+      "loss": 1.2645,
+      "step": 5215
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.42808278978975384,
+      "learning_rate": 7.91824478821368e-07,
+      "loss": 1.2138,
+      "step": 5216
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.431291868020283,
+      "learning_rate": 7.843466226038155e-07,
+      "loss": 1.1257,
+      "step": 5217
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.4285867259111711,
+      "learning_rate": 7.769041048908965e-07,
+      "loss": 1.161,
+      "step": 5218
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.4731097400100977,
+      "learning_rate": 7.694969283334575e-07,
+      "loss": 0.9706,
+      "step": 5219
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.4385309632021775,
+      "learning_rate": 7.621250955697878e-07,
+      "loss": 1.0245,
+      "step": 5220
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5063875638656269,
+      "learning_rate": 7.547886092255762e-07,
+      "loss": 1.0596,
+      "step": 5221
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.4290409656402819,
+      "learning_rate": 7.474874719139214e-07,
+      "loss": 1.0952,
+      "step": 5222
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.38878869993657544,
+      "learning_rate": 7.402216862353317e-07,
+      "loss": 1.2055,
+      "step": 5223
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.4296447364755756,
+      "learning_rate": 7.329912547777262e-07,
+      "loss": 0.9812,
+      "step": 5224
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.4262444663141015,
+      "learning_rate": 7.257961801164337e-07,
+      "loss": 1.1275,
+      "step": 5225
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.4580682614771181,
+      "learning_rate": 7.186364648141819e-07,
+      "loss": 1.094,
+      "step": 5226
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.4710751043506765,
+      "learning_rate": 7.115121114211199e-07,
+      "loss": 1.1989,
+      "step": 5227
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.41030774897449107,
+      "learning_rate": 7.044231224747844e-07,
+      "loss": 1.2384,
+      "step": 5228
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.4242820618371173,
+      "learning_rate": 6.973695005001113e-07,
+      "loss": 1.1265,
+      "step": 5229
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.38598388219540214,
+      "learning_rate": 6.903512480094576e-07,
+      "loss": 1.0589,
+      "step": 5230
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.4156966611959161,
+      "learning_rate": 6.833683675025904e-07,
+      "loss": 1.2364,
+      "step": 5231
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.43597598071390165,
+      "learning_rate": 6.764208614666423e-07,
+      "loss": 1.0064,
+      "step": 5232
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.4504220170770441,
+      "learning_rate": 6.695087323761784e-07,
+      "loss": 1.0911,
+      "step": 5233
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.4643280845462525,
+      "learning_rate": 6.626319826931404e-07,
+      "loss": 0.9819,
+      "step": 5234
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.45509939115341264,
+      "learning_rate": 6.557906148669024e-07,
+      "loss": 1.1744,
+      "step": 5235
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.4264149545821727,
+      "learning_rate": 6.48984631334204e-07,
+      "loss": 1.0999,
+      "step": 5236
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.4609227940479632,
+      "learning_rate": 6.422140345191729e-07,
+      "loss": 1.045,
+      "step": 5237
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.40963076051354186,
+      "learning_rate": 6.3547882683338e-07,
+      "loss": 1.1955,
+      "step": 5238
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.43319515550379645,
+      "learning_rate": 6.287790106757396e-07,
+      "loss": 1.0715,
+      "step": 5239
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.4349601472236186,
+      "learning_rate": 6.221145884326096e-07,
+      "loss": 1.1576,
+      "step": 5240
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.47985837672661236,
+      "learning_rate": 6.154855624776801e-07,
+      "loss": 1.1973,
+      "step": 5241
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.43686863419303873,
+      "learning_rate": 6.088919351721067e-07,
+      "loss": 1.0746,
+      "step": 5242
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.44605165371507444,
+      "learning_rate": 6.023337088643665e-07,
+      "loss": 1.0485,
+      "step": 5243
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.415856434322536,
+      "learning_rate": 5.958108858903799e-07,
+      "loss": 1.0261,
+      "step": 5244
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.4488449201083173,
+      "learning_rate": 5.893234685734439e-07,
+      "loss": 1.113,
+      "step": 5245
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.4528701283949548,
+      "learning_rate": 5.828714592242212e-07,
+      "loss": 1.0783,
+      "step": 5246
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.4364521125962032,
+      "learning_rate": 5.764548601407849e-07,
+      "loss": 1.1175,
+      "step": 5247
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.425603759690955,
+      "learning_rate": 5.700736736085843e-07,
+      "loss": 1.1242,
+      "step": 5248
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.3927028653777134,
+      "learning_rate": 5.637279019004793e-07,
+      "loss": 1.1448,
+      "step": 5249
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.43318700744946675,
+      "learning_rate": 5.57417547276684e-07,
+      "loss": 1.2523,
+      "step": 5250
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.4685909473829075,
+      "learning_rate": 5.51142611984834e-07,
+      "loss": 1.042,
+      "step": 5251
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.4531938881887009,
+      "learning_rate": 5.449030982598968e-07,
+      "loss": 1.1209,
+      "step": 5252
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.3926137207600696,
+      "learning_rate": 5.386990083242837e-07,
+      "loss": 1.1176,
+      "step": 5253
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.4272095939922754,
+      "learning_rate": 5.325303443877272e-07,
+      "loss": 1.098,
+      "step": 5254
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.412775496660864,
+      "learning_rate": 5.26397108647414e-07,
+      "loss": 1.0466,
+      "step": 5255
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.4484366051135617,
+      "learning_rate": 5.202993032878301e-07,
+      "loss": 1.1236,
+      "step": 5256
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.45151004284023705,
+      "learning_rate": 5.142369304809158e-07,
+      "loss": 1.0649,
+      "step": 5257
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.39263548111340996,
+      "learning_rate": 5.082099923859551e-07,
+      "loss": 1.1237,
+      "step": 5258
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.4167637895070872,
+      "learning_rate": 5.022184911495864e-07,
+      "loss": 1.0797,
+      "step": 5259
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.4273914781617841,
+      "learning_rate": 4.962624289059026e-07,
+      "loss": 1.0393,
+      "step": 5260
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.4421970739248616,
+      "learning_rate": 4.903418077762844e-07,
+      "loss": 1.298,
+      "step": 5261
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.487879905285035,
+      "learning_rate": 4.844566298695563e-07,
+      "loss": 1.0787,
+      "step": 5262
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.47507020491961266,
+      "learning_rate": 4.786068972818747e-07,
+      "loss": 1.1218,
+      "step": 5263
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5445639023394767,
+      "learning_rate": 4.7279261209680627e-07,
+      "loss": 1.0413,
+      "step": 5264
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.4169967479955348,
+      "learning_rate": 4.670137763852611e-07,
+      "loss": 1.0188,
+      "step": 5265
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.44416578274174434,
+      "learning_rate": 4.6127039220555943e-07,
+      "loss": 0.9715,
+      "step": 5266
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.3844640792319549,
+      "learning_rate": 4.555624616033427e-07,
+      "loss": 1.1726,
+      "step": 5267
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.4148035279939807,
+      "learning_rate": 4.498899866116735e-07,
+      "loss": 1.1747,
+      "step": 5268
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5276233370775006,
+      "learning_rate": 4.442529692509689e-07,
+      "loss": 1.1068,
+      "step": 5269
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.43982004514662026,
+      "learning_rate": 4.386514115290119e-07,
+      "loss": 1.1094,
+      "step": 5270
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.4615738286809717,
+      "learning_rate": 4.3308531544095087e-07,
+      "loss": 1.0856,
+      "step": 5271
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.41820568215914605,
+      "learning_rate": 4.2755468296931113e-07,
+      "loss": 1.1092,
+      "step": 5272
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.4229129598781426,
+      "learning_rate": 4.220595160840057e-07,
+      "loss": 1.0485,
+      "step": 5273
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.376920674749232,
+      "learning_rate": 4.16599816742258e-07,
+      "loss": 1.1028,
+      "step": 5274
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.46738528143956026,
+      "learning_rate": 4.111755868887346e-07,
+      "loss": 1.1168,
+      "step": 5275
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.4685572626809478,
+      "learning_rate": 4.0578682845541226e-07,
+      "loss": 0.9788,
+      "step": 5276
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.4096145487660686,
+      "learning_rate": 4.0043354336165573e-07,
+      "loss": 1.0237,
+      "step": 5277
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.4876679217781047,
+      "learning_rate": 3.951157335141953e-07,
+      "loss": 1.1121,
+      "step": 5278
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.440255735522709,
+      "learning_rate": 3.8983340080710476e-07,
+      "loss": 0.9407,
+      "step": 5279
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.4786457997681977,
+      "learning_rate": 3.845865471218679e-07,
+      "loss": 1.0947,
+      "step": 5280
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.43410438107890903,
+      "learning_rate": 3.7937517432726777e-07,
+      "loss": 1.1017,
+      "step": 5281
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5152218732128618,
+      "learning_rate": 3.7419928427951944e-07,
+      "loss": 1.0445,
+      "step": 5282
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.41305092663504267,
+      "learning_rate": 3.6905887882213717e-07,
+      "loss": 1.262,
+      "step": 5283
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.4320073428647209,
+      "learning_rate": 3.639539597860342e-07,
+      "loss": 1.103,
+      "step": 5284
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.3995572102391015,
+      "learning_rate": 3.588845289894782e-07,
+      "loss": 1.0447,
+      "step": 5285
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5109382889851929,
+      "learning_rate": 3.5385058823809156e-07,
+      "loss": 0.9416,
+      "step": 5286
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.41117487162158184,
+      "learning_rate": 3.488521393248401e-07,
+      "loss": 1.1408,
+      "step": 5287
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.4509389206475228,
+      "learning_rate": 3.438891840300995e-07,
+      "loss": 1.1345,
+      "step": 5288
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.40649761438557636,
+      "learning_rate": 3.389617241215337e-07,
+      "loss": 1.0509,
+      "step": 5289
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.42022640823063845,
+      "learning_rate": 3.3406976135421653e-07,
+      "loss": 1.004,
+      "step": 5290
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5147113101991124,
+      "learning_rate": 3.292132974705653e-07,
+      "loss": 1.1901,
+      "step": 5291
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.4371808371978573,
+      "learning_rate": 3.2439233420032965e-07,
+      "loss": 1.0394,
+      "step": 5292
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.43928472539860347,
+      "learning_rate": 3.196068732606583e-07,
+      "loss": 1.232,
+      "step": 5293
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.8025538435843999,
+      "learning_rate": 3.1485691635601e-07,
+      "loss": 1.0073,
+      "step": 5294
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.4626697929426316,
+      "learning_rate": 3.1014246517823145e-07,
+      "loss": 0.9824,
+      "step": 5295
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.4709657961351348,
+      "learning_rate": 3.054635214065016e-07,
+      "loss": 1.1802,
+      "step": 5296
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.4431623253079953,
+      "learning_rate": 3.0082008670736517e-07,
+      "loss": 1.0385,
+      "step": 5297
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.490784240494799,
+      "learning_rate": 2.962121627347214e-07,
+      "loss": 1.0391,
+      "step": 5298
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.43732548353922707,
+      "learning_rate": 2.916397511298019e-07,
+      "loss": 1.2252,
+      "step": 5299
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5579181053263739,
+      "learning_rate": 2.8710285352120393e-07,
+      "loss": 1.0628,
+      "step": 5300
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.4203516449047966,
+      "learning_rate": 2.826014715248904e-07,
+      "loss": 1.1268,
+      "step": 5301
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.4673224888983139,
+      "learning_rate": 2.781356067441343e-07,
+      "loss": 0.9598,
+      "step": 5302
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.4348728459549277,
+      "learning_rate": 2.7370526076959666e-07,
+      "loss": 1.2364,
+      "step": 5303
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.4368649508864878,
+      "learning_rate": 2.693104351792708e-07,
+      "loss": 0.9943,
+      "step": 5304
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.4127056344367273,
+      "learning_rate": 2.649511315384934e-07,
+      "loss": 1.0136,
+      "step": 5305
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5646858322298262,
+      "learning_rate": 2.6062735139996684e-07,
+      "loss": 1.0969,
+      "step": 5306
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.43693714567966224,
+      "learning_rate": 2.5633909630371487e-07,
+      "loss": 1.106,
+      "step": 5307
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.4302855471635986,
+      "learning_rate": 2.520863677771268e-07,
+      "loss": 0.9025,
+      "step": 5308
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.44314671917649356,
+      "learning_rate": 2.478691673349465e-07,
+      "loss": 1.2062,
+      "step": 5309
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.4289928157459252,
+      "learning_rate": 2.4368749647925013e-07,
+      "loss": 1.2944,
+      "step": 5310
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.44942521043989075,
+      "learning_rate": 2.395413566994353e-07,
+      "loss": 1.1275,
+      "step": 5311
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.4745988223026685,
+      "learning_rate": 2.3543074947229848e-07,
+      "loss": 1.0007,
+      "step": 5312
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.4145767794309972,
+      "learning_rate": 2.313556762619351e-07,
+      "loss": 1.0019,
+      "step": 5313
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.44946913216343864,
+      "learning_rate": 2.273161385197953e-07,
+      "loss": 1.2284,
+      "step": 5314
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.49449592965402983,
+      "learning_rate": 2.2331213768468363e-07,
+      "loss": 1.097,
+      "step": 5315
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.4562931638146519,
+      "learning_rate": 2.1934367518274823e-07,
+      "loss": 1.0867,
+      "step": 5316
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.4300054348263976,
+      "learning_rate": 2.1541075242745846e-07,
+      "loss": 1.2531,
+      "step": 5317
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.39599513186123714,
+      "learning_rate": 2.1151337081963817e-07,
+      "loss": 1.0649,
+      "step": 5318
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.4313726557030586,
+      "learning_rate": 2.0765153174745478e-07,
+      "loss": 1.2467,
+      "step": 5319
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.41122944004287104,
+      "learning_rate": 2.0382523658641906e-07,
+      "loss": 0.9059,
+      "step": 5320
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.4641874793880597,
+      "learning_rate": 2.000344866993631e-07,
+      "loss": 1.0311,
+      "step": 5321
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.44866959293590764,
+      "learning_rate": 1.962792834364846e-07,
+      "loss": 1.0515,
+      "step": 5322
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.41260102230820145,
+      "learning_rate": 1.925596281353026e-07,
+      "loss": 1.1152,
+      "step": 5323
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.48373001628779594,
+      "learning_rate": 1.888755221206795e-07,
+      "loss": 0.9544,
+      "step": 5324
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.41988252326796605,
+      "learning_rate": 1.852269667048101e-07,
+      "loss": 1.0832,
+      "step": 5325
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.45621749688526464,
+      "learning_rate": 1.8161396318723266e-07,
+      "loss": 0.9646,
+      "step": 5326
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.4038937699690394,
+      "learning_rate": 1.7803651285482893e-07,
+      "loss": 1.1271,
+      "step": 5327
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.42461217912211136,
+      "learning_rate": 1.744946169818018e-07,
+      "loss": 1.0966,
+      "step": 5328
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.42567374499143507,
+      "learning_rate": 1.7098827682970886e-07,
+      "loss": 1.0827,
+      "step": 5329
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.45489167210124254,
+      "learning_rate": 1.6751749364743995e-07,
+      "loss": 1.2095,
+      "step": 5330
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.4580275522716424,
+      "learning_rate": 1.6408226867118403e-07,
+      "loss": 0.9822,
+      "step": 5331
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.41723036411299635,
+      "learning_rate": 1.6068260312452898e-07,
+      "loss": 1.0681,
+      "step": 5332
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.40145823329317215,
+      "learning_rate": 1.5731849821833954e-07,
+      "loss": 1.324,
+      "step": 5333
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.42594911897690374,
+      "learning_rate": 1.5398995515085724e-07,
+      "loss": 0.99,
+      "step": 5334
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.44325883247154074,
+      "learning_rate": 1.506969751076226e-07,
+      "loss": 1.2428,
+      "step": 5335
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.44003096690088817,
+      "learning_rate": 1.4743955926154185e-07,
+      "loss": 1.0695,
+      "step": 5336
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.4372525373305373,
+      "learning_rate": 1.4421770877280915e-07,
+      "loss": 0.9638,
+      "step": 5337
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.44627504062133705,
+      "learning_rate": 1.4103142478900655e-07,
+      "loss": 1.1023,
+      "step": 5338
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.42794834218492284,
+      "learning_rate": 1.378807084450151e-07,
+      "loss": 1.0842,
+      "step": 5339
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5492770038433381,
+      "learning_rate": 1.3476556086303715e-07,
+      "loss": 0.9912,
+      "step": 5340
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.47710532067826605,
+      "learning_rate": 1.3168598315264068e-07,
+      "loss": 1.1084,
+      "step": 5341
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.4345268843630331,
+      "learning_rate": 1.2864197641070387e-07,
+      "loss": 1.0325,
+      "step": 5342
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.4647657506795041,
+      "learning_rate": 1.2563354172142606e-07,
+      "loss": 1.0211,
+      "step": 5343
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.4452031993662504,
+      "learning_rate": 1.2266068015635013e-07,
+      "loss": 1.1371,
+      "step": 5344
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.450171458628293,
+      "learning_rate": 1.1972339277435129e-07,
+      "loss": 1.0768,
+      "step": 5345
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.39183448073655547,
+      "learning_rate": 1.1682168062163713e-07,
+      "loss": 1.3453,
+      "step": 5346
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.4590245463620408,
+      "learning_rate": 1.1395554473171422e-07,
+      "loss": 1.0682,
+      "step": 5347
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5305429751690773,
+      "learning_rate": 1.1112498612545486e-07,
+      "loss": 1.0935,
+      "step": 5348
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.41720912570316177,
+      "learning_rate": 1.0833000581103036e-07,
+      "loss": 1.0731,
+      "step": 5349
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.49660424179006807,
+      "learning_rate": 1.0557060478396663e-07,
+      "loss": 1.1924,
+      "step": 5350
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.4547983600124261,
+      "learning_rate": 1.0284678402711079e-07,
+      "loss": 0.9739,
+      "step": 5351
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.4436289388303694,
+      "learning_rate": 1.0015854451060901e-07,
+      "loss": 1.0605,
+      "step": 5352
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.44582289863907093,
+      "learning_rate": 9.750588719196207e-08,
+      "loss": 1.1555,
+      "step": 5353
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.40500550443137295,
+      "learning_rate": 9.488881301599195e-08,
+      "loss": 1.1748,
+      "step": 5354
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.4539293592773089,
+      "learning_rate": 9.230732291485301e-08,
+      "loss": 1.0811,
+      "step": 5355
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.41022961145201875,
+      "learning_rate": 8.976141780799862e-08,
+      "loss": 1.0469,
+      "step": 5356
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.40519877387594827,
+      "learning_rate": 8.725109860224789e-08,
+      "loss": 0.9955,
+      "step": 5357
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.44246430300796513,
+      "learning_rate": 8.477636619170781e-08,
+      "loss": 0.9857,
+      "step": 5358
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.4527316894040448,
+      "learning_rate": 8.233722145782885e-08,
+      "loss": 1.0358,
+      "step": 5359
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.4528664253797944,
+      "learning_rate": 7.993366526938273e-08,
+      "loss": 1.1067,
+      "step": 5360
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.44406508384842475,
+      "learning_rate": 7.756569848246242e-08,
+      "loss": 1.1701,
+      "step": 5361
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.44078764075541754,
+      "learning_rate": 7.523332194049326e-08,
+      "loss": 1.2453,
+      "step": 5362
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.498000866151354,
+      "learning_rate": 7.293653647421073e-08,
+      "loss": 0.9948,
+      "step": 5363
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.4169632457837521,
+      "learning_rate": 7.067534290169376e-08,
+      "loss": 1.3115,
+      "step": 5364
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.49068676488060964,
+      "learning_rate": 6.844974202832033e-08,
+      "loss": 1.2426,
+      "step": 5365
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.43672974404397313,
+      "learning_rate": 6.625973464680079e-08,
+      "loss": 1.0557,
+      "step": 5366
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.46361595905547925,
+      "learning_rate": 6.410532153717786e-08,
+      "loss": 0.981,
+      "step": 5367
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5046356801955793,
+      "learning_rate": 6.198650346680435e-08,
+      "loss": 1.0946,
+      "step": 5368
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.4149865698440182,
+      "learning_rate": 5.990328119034328e-08,
+      "loss": 1.0717,
+      "step": 5369
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.4412780368017275,
+      "learning_rate": 5.7855655449823297e-08,
+      "loss": 1.1406,
+      "step": 5370
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.42424195204032206,
+      "learning_rate": 5.584362697453882e-08,
+      "loss": 1.2821,
+      "step": 5371
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.44108582688405445,
+      "learning_rate": 5.3867196481149904e-08,
+      "loss": 0.9717,
+      "step": 5372
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.43466149808076077,
+      "learning_rate": 5.192636467360457e-08,
+      "loss": 1.0288,
+      "step": 5373
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.43030773824280266,
+      "learning_rate": 5.002113224319427e-08,
+      "loss": 1.043,
+      "step": 5374
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.4128224804882282,
+      "learning_rate": 4.8151499868520634e-08,
+      "loss": 1.0711,
+      "step": 5375
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.4651568330243433,
+      "learning_rate": 4.631746821550653e-08,
+      "loss": 1.1794,
+      "step": 5376
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.40174716820068956,
+      "learning_rate": 4.451903793740719e-08,
+      "loss": 1.3519,
+      "step": 5377
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.4727389578301178,
+      "learning_rate": 4.275620967475469e-08,
+      "loss": 1.1018,
+      "step": 5378
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.3883419231859818,
+      "learning_rate": 4.102898405545785e-08,
+      "loss": 1.0821,
+      "step": 5379
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.4145533555188952,
+      "learning_rate": 3.933736169471347e-08,
+      "loss": 1.1165,
+      "step": 5380
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.44664981495686296,
+      "learning_rate": 3.7681343195028474e-08,
+      "loss": 1.0851,
+      "step": 5381
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.4792018509921765,
+      "learning_rate": 3.606092914625325e-08,
+      "loss": 1.1706,
+      "step": 5382
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.4240608713135061,
+      "learning_rate": 3.447612012554835e-08,
+      "loss": 1.1303,
+      "step": 5383
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.3963063323677778,
+      "learning_rate": 3.2926916697373357e-08,
+      "loss": 1.0857,
+      "step": 5384
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.42355101085103036,
+      "learning_rate": 3.141331941354242e-08,
+      "loss": 1.1946,
+      "step": 5385
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.4284812594313166,
+      "learning_rate": 2.9935328813146536e-08,
+      "loss": 1.2294,
+      "step": 5386
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.4280457180570557,
+      "learning_rate": 2.8492945422620155e-08,
+      "loss": 1.038,
+      "step": 5387
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.4218279548517095,
+      "learning_rate": 2.7086169755718983e-08,
+      "loss": 1.3062,
+      "step": 5388
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.41525691088086497,
+      "learning_rate": 2.5715002313497772e-08,
+      "loss": 1.1882,
+      "step": 5389
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5235492714092416,
+      "learning_rate": 2.437944358433253e-08,
+      "loss": 1.125,
+      "step": 5390
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.393635735279244,
+      "learning_rate": 2.307949404394272e-08,
+      "loss": 1.1767,
+      "step": 5391
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.4088415648959348,
+      "learning_rate": 2.1815154155313543e-08,
+      "loss": 1.2466,
+      "step": 5392
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.3767327299785733,
+      "learning_rate": 2.0586424368795875e-08,
+      "loss": 1.2521,
+      "step": 5393
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.44953019221183016,
+      "learning_rate": 1.9393305122028528e-08,
+      "loss": 1.0692,
+      "step": 5394
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.4312806657864153,
+      "learning_rate": 1.8235796839982665e-08,
+      "loss": 1.0587,
+      "step": 5395
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.46702199851155135,
+      "learning_rate": 1.7113899934939616e-08,
+      "loss": 1.0977,
+      "step": 5396
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.4190286648430261,
+      "learning_rate": 1.6027614806479742e-08,
+      "loss": 1.1197,
+      "step": 5397
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.44933392698152386,
+      "learning_rate": 1.497694184151577e-08,
+      "loss": 1.1278,
+      "step": 5398
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.4548012205664979,
+      "learning_rate": 1.3961881414292778e-08,
+      "loss": 1.0017,
+      "step": 5399
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.3974342972609528,
+      "learning_rate": 1.2982433886354894e-08,
+      "loss": 1.0574,
+      "step": 5400
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.42100890358701687,
+      "learning_rate": 1.2038599606534195e-08,
+      "loss": 1.1475,
+      "step": 5401
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.4250644294481628,
+      "learning_rate": 1.1130378911028416e-08,
+      "loss": 1.1643,
+      "step": 5402
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.4231704986650131,
+      "learning_rate": 1.0257772123312137e-08,
+      "loss": 1.1959,
+      "step": 5403
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.43532311661636336,
+      "learning_rate": 9.420779554192294e-09,
+      "loss": 1.1894,
+      "step": 5404
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.4597865424007111,
+      "learning_rate": 8.619401501797075e-09,
+      "loss": 1.0053,
+      "step": 5405
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.40840220535584676,
+      "learning_rate": 7.853638251553718e-09,
+      "loss": 0.9585,
+      "step": 5406
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.441705852217399,
+      "learning_rate": 7.123490076210715e-09,
+      "loss": 1.2582,
+      "step": 5407
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.37659001411163956,
+      "learning_rate": 6.428957235826705e-09,
+      "loss": 0.8886,
+      "step": 5408
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.39371875251846067,
+      "learning_rate": 5.770039977792685e-09,
+      "loss": 1.1085,
+      "step": 5409
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.40812845140424403,
+      "learning_rate": 5.146738536787599e-09,
+      "loss": 1.0697,
+      "step": 5410
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5400584982729886,
+      "learning_rate": 4.559053134822744e-09,
+      "loss": 0.9769,
+      "step": 5411
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.4524738660148234,
+      "learning_rate": 4.0069839812195696e-09,
+      "loss": 1.0725,
+      "step": 5412
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.46722928309337536,
+      "learning_rate": 3.4905312726207783e-09,
+      "loss": 1.0711,
+      "step": 5413
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.4274325745892134,
+      "learning_rate": 3.0096951929570184e-09,
+      "loss": 1.1261,
+      "step": 5414
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.44751625273800705,
+      "learning_rate": 2.5644759135134976e-09,
+      "loss": 1.0141,
+      "step": 5415
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.4409318228566828,
+      "learning_rate": 2.154873592852269e-09,
+      "loss": 1.0978,
+      "step": 5416
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.4268051531358153,
+      "learning_rate": 1.7808883768677398e-09,
+      "loss": 1.1808,
+      "step": 5417
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.4331543991885026,
+      "learning_rate": 1.4425203987755709e-09,
+      "loss": 1.0686,
+      "step": 5418
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.45463115429778705,
+      "learning_rate": 1.1397697790793694e-09,
+      "loss": 1.0249,
+      "step": 5419
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.46430706765881313,
+      "learning_rate": 8.726366256261998e-10,
+      "loss": 1.1969,
+      "step": 5420
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.4138709520121256,
+      "learning_rate": 6.411210335621753e-10,
+      "loss": 1.0431,
+      "step": 5421
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.460398286392873,
+      "learning_rate": 4.4522308533245793e-10,
+      "loss": 1.1499,
+      "step": 5422
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.4577333773457859,
+      "learning_rate": 2.849428507367691e-10,
+      "loss": 1.0655,
+      "step": 5423
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.44222594111251246,
+      "learning_rate": 1.6028038684057222e-10,
+      "loss": 0.8774,
+      "step": 5424
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.46145042129885244,
+      "learning_rate": 7.123573805278838e-11,
+      "loss": 1.0719,
+      "step": 5425
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.49717710028697093,
+      "learning_rate": 1.7808936103591577e-11,
+      "loss": 0.9506,
+      "step": 5426
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.42189228327446016,
+      "learning_rate": 0.0,
+      "loss": 1.1727,
+      "step": 5427
+    },
+    {
+      "epoch": 1.0,
+      "step": 5427,
+      "total_flos": 4307202935947264.0,
+      "train_loss": 1.1681652979141708,
+      "train_runtime": 16865.9608,
+      "train_samples_per_second": 41.184,
+      "train_steps_per_second": 0.322
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 5427,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 500,
+  "total_flos": 4307202935947264.0,
+  "train_batch_size": 8,
+  "trial_name": null,
+  "trial_params": null
+}