diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,8792 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.0,
+  "eval_steps": 500,
+  "global_step": 1250,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "grad_norm": 48.60254041857085,
+      "learning_rate": 5.263157894736843e-07,
+      "loss": 1.3605,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 35.89314672664364,
+      "learning_rate": 1.0526315789473685e-06,
+      "loss": 1.2401,
+      "step": 2
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 60.7990081247422,
+      "learning_rate": 1.5789473684210526e-06,
+      "loss": 1.1649,
+      "step": 3
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 36.51376184380329,
+      "learning_rate": 2.105263157894737e-06,
+      "loss": 1.2448,
+      "step": 4
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 36.98096547145381,
+      "learning_rate": 2.631578947368421e-06,
+      "loss": 1.3259,
+      "step": 5
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 19.840368403961094,
+      "learning_rate": 3.157894736842105e-06,
+      "loss": 1.1235,
+      "step": 6
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 18.991812240245903,
+      "learning_rate": 3.6842105263157896e-06,
+      "loss": 1.0488,
+      "step": 7
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 14.112864562404495,
+      "learning_rate": 4.210526315789474e-06,
+      "loss": 1.0145,
+      "step": 8
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 11.905199352452714,
+      "learning_rate": 4.736842105263158e-06,
+      "loss": 1.0528,
+      "step": 9
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.258681115510464,
+      "learning_rate": 5.263157894736842e-06,
+      "loss": 0.8344,
+      "step": 10
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.468068271446631,
+      "learning_rate": 5.789473684210527e-06,
+      "loss": 0.8258,
+      "step": 11
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.2073891356799695,
+      "learning_rate": 6.31578947368421e-06,
+      "loss": 0.7777,
+      "step": 12
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.42052537085184,
+      "learning_rate": 6.842105263157896e-06,
+      "loss": 0.7737,
+      "step": 13
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.207179625141898,
+      "learning_rate": 7.368421052631579e-06,
+      "loss": 0.7968,
+      "step": 14
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.840526941421514,
+      "learning_rate": 7.894736842105265e-06,
+      "loss": 0.6777,
+      "step": 15
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.6686511595911,
+      "learning_rate": 8.421052631578948e-06,
+      "loss": 0.8208,
+      "step": 16
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.46146169529698,
+      "learning_rate": 8.947368421052632e-06,
+      "loss": 0.7916,
+      "step": 17
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.3633571859308296,
+      "learning_rate": 9.473684210526315e-06,
+      "loss": 0.5994,
+      "step": 18
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.7975096177082355,
+      "learning_rate": 1e-05,
+      "loss": 0.7091,
+      "step": 19
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.848702644768635,
+      "learning_rate": 1.0526315789473684e-05,
+      "loss": 0.6414,
+      "step": 20
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.520282197928678,
+      "learning_rate": 1.105263157894737e-05,
+      "loss": 0.6885,
+      "step": 21
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.912832317546539,
+      "learning_rate": 1.1578947368421053e-05,
+      "loss": 0.5096,
+      "step": 22
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.3336358480278765,
+      "learning_rate": 1.2105263157894737e-05,
+      "loss": 0.4512,
+      "step": 23
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.847456020229534,
+      "learning_rate": 1.263157894736842e-05,
+      "loss": 0.4546,
+      "step": 24
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.912624406249825,
+      "learning_rate": 1.3157894736842108e-05,
+      "loss": 0.5482,
+      "step": 25
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.5759205643404113,
+      "learning_rate": 1.3684210526315791e-05,
+      "loss": 0.4693,
+      "step": 26
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.3883094275895367,
+      "learning_rate": 1.4210526315789475e-05,
+      "loss": 0.51,
+      "step": 27
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.7248048455794103,
+      "learning_rate": 1.4736842105263159e-05,
+      "loss": 0.5592,
+      "step": 28
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.8493552123337693,
+      "learning_rate": 1.5263157894736846e-05,
+      "loss": 0.4027,
+      "step": 29
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.8508740199643166,
+      "learning_rate": 1.578947368421053e-05,
+      "loss": 0.4436,
+      "step": 30
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.45543671036673,
+      "learning_rate": 1.6315789473684213e-05,
+      "loss": 0.464,
+      "step": 31
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.294884635896111,
+      "learning_rate": 1.6842105263157896e-05,
+      "loss": 0.3796,
+      "step": 32
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.623917186038261,
+      "learning_rate": 1.736842105263158e-05,
+      "loss": 0.572,
+      "step": 33
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.0555190621524035,
+      "learning_rate": 1.7894736842105264e-05,
+      "loss": 0.463,
+      "step": 34
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.134418141871718,
+      "learning_rate": 1.8421052631578947e-05,
+      "loss": 0.4648,
+      "step": 35
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.0300812406530055,
+      "learning_rate": 1.894736842105263e-05,
+      "loss": 0.4268,
+      "step": 36
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.9656208592680042,
+      "learning_rate": 1.9473684210526318e-05,
+      "loss": 0.4076,
+      "step": 37
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.297106308343148,
+      "learning_rate": 2e-05,
+      "loss": 0.392,
+      "step": 38
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.278775729812893,
+      "learning_rate": 1.9999966405802828e-05,
+      "loss": 0.4489,
+      "step": 39
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.6821653254673077,
+      "learning_rate": 1.9999865623437014e-05,
+      "loss": 0.4527,
+      "step": 40
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.1520894418798817,
+      "learning_rate": 1.9999697653579705e-05,
+      "loss": 0.3535,
+      "step": 41
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.7413611366019266,
+      "learning_rate": 1.9999462497359468e-05,
+      "loss": 0.4927,
+      "step": 42
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.9573068516195606,
+      "learning_rate": 1.999916015635627e-05,
+      "loss": 0.3845,
+      "step": 43
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.135750361580708,
+      "learning_rate": 1.9998790632601496e-05,
+      "loss": 0.4078,
+      "step": 44
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.042182784623944,
+      "learning_rate": 1.9998353928577917e-05,
+      "loss": 0.3984,
+      "step": 45
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.8618916695510674,
+      "learning_rate": 1.999785004721968e-05,
+      "loss": 0.4136,
+      "step": 46
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.6163481326606473,
+      "learning_rate": 1.999727899191228e-05,
+      "loss": 0.4976,
+      "step": 47
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.466867974348983,
+      "learning_rate": 1.9996640766492542e-05,
+      "loss": 0.3409,
+      "step": 48
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.1921413852007063,
+      "learning_rate": 1.9995935375248608e-05,
+      "loss": 0.4116,
+      "step": 49
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.6860666687778885,
+      "learning_rate": 1.999516282291988e-05,
+      "loss": 0.3526,
+      "step": 50
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.970761788613036,
+      "learning_rate": 1.999432311469702e-05,
+      "loss": 0.39,
+      "step": 51
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.148790374736204,
+      "learning_rate": 1.9993416256221894e-05,
+      "loss": 0.477,
+      "step": 52
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.8495854280786928,
+      "learning_rate": 1.9992442253587533e-05,
+      "loss": 0.4467,
+      "step": 53
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.6437700924653527,
+      "learning_rate": 1.9991401113338103e-05,
+      "loss": 0.3822,
+      "step": 54
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.13624492498233,
+      "learning_rate": 1.9990292842468868e-05,
+      "loss": 0.4819,
+      "step": 55
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.6929943628367687,
+      "learning_rate": 1.998911744842611e-05,
+      "loss": 0.3502,
+      "step": 56
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.7976347652268854,
+      "learning_rate": 1.998787493910712e-05,
+      "loss": 0.4113,
+      "step": 57
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.585758230683453,
+      "learning_rate": 1.9986565322860117e-05,
+      "loss": 0.4017,
+      "step": 58
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.2917334411264303,
+      "learning_rate": 1.99851886084842e-05,
+      "loss": 0.3634,
+      "step": 59
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.6520998947655685,
+      "learning_rate": 1.9983744805229296e-05,
+      "loss": 0.3026,
+      "step": 60
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.468979558179385,
+      "learning_rate": 1.9982233922796087e-05,
+      "loss": 0.3603,
+      "step": 61
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.8251450554893736,
+      "learning_rate": 1.9980655971335944e-05,
+      "loss": 0.2928,
+      "step": 62
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.4576190213905567,
+      "learning_rate": 1.9979010961450876e-05,
+      "loss": 0.3272,
+      "step": 63
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.449635716809827,
+      "learning_rate": 1.9977298904193438e-05,
+      "loss": 0.415,
+      "step": 64
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.2168527165775895,
+      "learning_rate": 1.9975519811066665e-05,
+      "loss": 0.5168,
+      "step": 65
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.010946265742081,
+      "learning_rate": 1.9973673694024002e-05,
+      "loss": 0.3194,
+      "step": 66
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.3855860681847103,
+      "learning_rate": 1.997176056546921e-05,
+      "loss": 0.4533,
+      "step": 67
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.472854386346785,
+      "learning_rate": 1.9969780438256295e-05,
+      "loss": 0.3352,
+      "step": 68
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.2017570680574043,
+      "learning_rate": 1.9967733325689412e-05,
+      "loss": 0.3116,
+      "step": 69
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.801958352314478,
+      "learning_rate": 1.996561924152278e-05,
+      "loss": 0.3785,
+      "step": 70
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.7194369108730396,
+      "learning_rate": 1.99634381999606e-05,
+      "loss": 0.4828,
+      "step": 71
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.394819488279037,
+      "learning_rate": 1.996119021565693e-05,
+      "loss": 0.3358,
+      "step": 72
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.3093567895664115,
+      "learning_rate": 1.9958875303715618e-05,
+      "loss": 0.2748,
+      "step": 73
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.259362170148189,
+      "learning_rate": 1.995649347969019e-05,
+      "loss": 0.3039,
+      "step": 74
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.975563279323876,
+      "learning_rate": 1.995404475958373e-05,
+      "loss": 0.3976,
+      "step": 75
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.419784039357126,
+      "learning_rate": 1.9951529159848805e-05,
+      "loss": 0.3348,
+      "step": 76
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.281341928512009,
+      "learning_rate": 1.9948946697387322e-05,
+      "loss": 0.3401,
+      "step": 77
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.1087881889228988,
+      "learning_rate": 1.9946297389550433e-05,
+      "loss": 0.263,
+      "step": 78
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.639198057741474,
+      "learning_rate": 1.994358125413841e-05,
+      "loss": 0.3027,
+      "step": 79
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.6518187164037905,
+      "learning_rate": 1.9940798309400527e-05,
+      "loss": 0.3789,
+      "step": 80
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.3877370659912067,
+      "learning_rate": 1.993794857403495e-05,
+      "loss": 0.4408,
+      "step": 81
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.995385567063843,
+      "learning_rate": 1.993503206718859e-05,
+      "loss": 0.3925,
+      "step": 82
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.2232618125681203,
+      "learning_rate": 1.993204880845699e-05,
+      "loss": 0.282,
+      "step": 83
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.8970214260593643,
+      "learning_rate": 1.9928998817884185e-05,
+      "loss": 0.3268,
+      "step": 84
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.3977831596079295,
+      "learning_rate": 1.992588211596257e-05,
+      "loss": 0.3398,
+      "step": 85
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.6505205476488403,
+      "learning_rate": 1.992269872363277e-05,
+      "loss": 0.2896,
+      "step": 86
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.673536154000272,
+      "learning_rate": 1.991944866228348e-05,
+      "loss": 0.3762,
+      "step": 87
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.5495620413841635,
+      "learning_rate": 1.9916131953751342e-05,
+      "loss": 0.2878,
+      "step": 88
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.702863582845579,
+      "learning_rate": 1.9912748620320796e-05,
+      "loss": 0.3463,
+      "step": 89
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.8321910963387813,
+      "learning_rate": 1.9909298684723905e-05,
+      "loss": 0.3361,
+      "step": 90
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.884365131994556,
+      "learning_rate": 1.990578217014024e-05,
+      "loss": 0.3617,
+      "step": 91
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.4865597940498807,
+      "learning_rate": 1.9902199100196697e-05,
+      "loss": 0.3244,
+      "step": 92
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.458146087533274,
+      "learning_rate": 1.9898549498967343e-05,
+      "loss": 0.3171,
+      "step": 93
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.9371733606323587,
+      "learning_rate": 1.9894833390973266e-05,
+      "loss": 0.3643,
+      "step": 94
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.8304095813590995,
+      "learning_rate": 1.98910508011824e-05,
+      "loss": 0.3632,
+      "step": 95
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.218718793020854,
+      "learning_rate": 1.9887201755009358e-05,
+      "loss": 0.3553,
+      "step": 96
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.020368297448905,
+      "learning_rate": 1.9883286278315262e-05,
+      "loss": 0.3934,
+      "step": 97
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.39142603903749,
+      "learning_rate": 1.987930439740757e-05,
+      "loss": 0.2706,
+      "step": 98
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.174184873263371,
+      "learning_rate": 1.98752561390399e-05,
+      "loss": 0.2752,
+      "step": 99
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.0120477247756527,
+      "learning_rate": 1.9871141530411854e-05,
+      "loss": 0.3533,
+      "step": 100
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.580695227129875,
+      "learning_rate": 1.9866960599168825e-05,
+      "loss": 0.3744,
+      "step": 101
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.8478732301676497,
+      "learning_rate": 1.986271337340182e-05,
+      "loss": 0.4049,
+      "step": 102
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9142355852738673,
+      "learning_rate": 1.985839988164726e-05,
+      "loss": 0.2745,
+      "step": 103
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.278887974213085,
+      "learning_rate": 1.9854020152886816e-05,
+      "loss": 0.2881,
+      "step": 104
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.7272146416314187,
+      "learning_rate": 1.984957421654717e-05,
+      "loss": 0.3944,
+      "step": 105
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.3803352703736316,
+      "learning_rate": 1.984506210249986e-05,
+      "loss": 0.262,
+      "step": 106
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.035594057707899,
+      "learning_rate": 1.9840483841061058e-05,
+      "loss": 0.3811,
+      "step": 107
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.0457289224654094,
+      "learning_rate": 1.983583946299136e-05,
+      "loss": 0.3255,
+      "step": 108
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.3985562690903723,
+      "learning_rate": 1.9831128999495605e-05,
+      "loss": 0.2482,
+      "step": 109
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.0610268888156695,
+      "learning_rate": 1.982635248222264e-05,
+      "loss": 0.3167,
+      "step": 110
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.4914973894626784,
+      "learning_rate": 1.982150994326511e-05,
+      "loss": 0.3469,
+      "step": 111
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.049519586667049,
+      "learning_rate": 1.9816601415159266e-05,
+      "loss": 0.3055,
+      "step": 112
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.8711780595627188,
+      "learning_rate": 1.981162693088471e-05,
+      "loss": 0.4214,
+      "step": 113
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6864434110364426,
+      "learning_rate": 1.9806586523864212e-05,
+      "loss": 0.2661,
+      "step": 114
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9783615289759517,
+      "learning_rate": 1.9801480227963453e-05,
+      "loss": 0.3394,
+      "step": 115
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.033682841325606,
+      "learning_rate": 1.9796308077490817e-05,
+      "loss": 0.2969,
+      "step": 116
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.54876929709659,
+      "learning_rate": 1.9791070107197155e-05,
+      "loss": 0.367,
+      "step": 117
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.656084912397775,
+      "learning_rate": 1.978576635227554e-05,
+      "loss": 0.4312,
+      "step": 118
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.393147407538132,
+      "learning_rate": 1.978039684836106e-05,
+      "loss": 0.3516,
+      "step": 119
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.2110241591122026,
+      "learning_rate": 1.9774961631530543e-05,
+      "loss": 0.3519,
+      "step": 120
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.5954241037962897,
+      "learning_rate": 1.976946073830234e-05,
+      "loss": 0.4076,
+      "step": 121
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.5229827155020326,
+      "learning_rate": 1.976389420563607e-05,
+      "loss": 0.3995,
+      "step": 122
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.514615433700628,
+      "learning_rate": 1.9758262070932375e-05,
+      "loss": 0.3715,
+      "step": 123
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.06027848990462,
+      "learning_rate": 1.9752564372032655e-05,
+      "loss": 0.3074,
+      "step": 124
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.0739204538312466,
+      "learning_rate": 1.9746801147218844e-05,
+      "loss": 0.3097,
+      "step": 125
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.97836933818748,
+      "learning_rate": 1.9740972435213114e-05,
+      "loss": 0.4227,
+      "step": 126
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9460161454046876,
+      "learning_rate": 1.9735078275177654e-05,
+      "loss": 0.2968,
+      "step": 127
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.1620498463762603,
+      "learning_rate": 1.9729118706714377e-05,
+      "loss": 0.3533,
+      "step": 128
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.1949576922394582,
+      "learning_rate": 1.9723093769864663e-05,
+      "loss": 0.3313,
+      "step": 129
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.3009910262849913,
+      "learning_rate": 1.9717003505109097e-05,
+      "loss": 0.2273,
+      "step": 130
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.405991987612652,
+      "learning_rate": 1.9710847953367193e-05,
+      "loss": 0.2974,
+      "step": 131
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.000338024899964,
+      "learning_rate": 1.970462715599711e-05,
+      "loss": 0.3373,
+      "step": 132
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.8176073475207875,
+      "learning_rate": 1.969834115479539e-05,
+      "loss": 0.3848,
+      "step": 133
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.784763316952781,
+      "learning_rate": 1.9691989991996663e-05,
+      "loss": 0.3022,
+      "step": 134
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.1290677125865116,
+      "learning_rate": 1.9685573710273375e-05,
+      "loss": 0.2818,
+      "step": 135
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.7967830185590423,
+      "learning_rate": 1.967909235273549e-05,
+      "loss": 0.3402,
+      "step": 136
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.5376148858644916,
+      "learning_rate": 1.9672545962930214e-05,
+      "loss": 0.4607,
+      "step": 137
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.4447490847016016,
+      "learning_rate": 1.966593458484168e-05,
+      "loss": 0.3072,
+      "step": 138
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.510203272599188,
+      "learning_rate": 1.9659258262890683e-05,
+      "loss": 0.3246,
+      "step": 139
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.6601770534439706,
+      "learning_rate": 1.9652517041934357e-05,
+      "loss": 0.3447,
+      "step": 140
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.4313564923861795,
+      "learning_rate": 1.9645710967265884e-05,
+      "loss": 0.2969,
+      "step": 141
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.301204187778797,
+      "learning_rate": 1.9638840084614182e-05,
+      "loss": 0.3449,
+      "step": 142
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.4324891519231553,
+      "learning_rate": 1.9631904440143614e-05,
+      "loss": 0.3649,
+      "step": 143
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.375771375825056,
+      "learning_rate": 1.9624904080453656e-05,
+      "loss": 0.2945,
+      "step": 144
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.2051469351019053,
+      "learning_rate": 1.9617839052578605e-05,
+      "loss": 0.2423,
+      "step": 145
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.2775254838339505,
+      "learning_rate": 1.9610709403987248e-05,
+      "loss": 0.3367,
+      "step": 146
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.5865833120563506,
+      "learning_rate": 1.960351518258255e-05,
+      "loss": 0.2663,
+      "step": 147
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.7811326564649548,
+      "learning_rate": 1.9596256436701324e-05,
+      "loss": 0.4021,
+      "step": 148
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9647404834701583,
+      "learning_rate": 1.9588933215113926e-05,
+      "loss": 0.2258,
+      "step": 149
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.6195686997710723,
+      "learning_rate": 1.95815455670239e-05,
+      "loss": 0.3468,
+      "step": 150
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.7408300850319733,
+      "learning_rate": 1.9574093542067673e-05,
+      "loss": 0.2905,
+      "step": 151
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.4846198337244676,
+      "learning_rate": 1.95665771903142e-05,
+      "loss": 0.3065,
+      "step": 152
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.201402370978399,
+      "learning_rate": 1.955899656226464e-05,
+      "loss": 0.2765,
+      "step": 153
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.2048417537082274,
+      "learning_rate": 1.955135170885202e-05,
+      "loss": 0.3261,
+      "step": 154
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8685661741745674,
+      "learning_rate": 1.954364268144088e-05,
+      "loss": 0.2897,
+      "step": 155
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.038261520725671,
+      "learning_rate": 1.9535869531826938e-05,
+      "loss": 0.2876,
+      "step": 156
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.179433620693176,
+      "learning_rate": 1.952803231223674e-05,
+      "loss": 0.2753,
+      "step": 157
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.046890347395123,
+      "learning_rate": 1.95201310753273e-05,
+      "loss": 0.3993,
+      "step": 158
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.8111544068550876,
+      "learning_rate": 1.9512165874185768e-05,
+      "loss": 0.284,
+      "step": 159
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.4995520607151525,
+      "learning_rate": 1.9504136762329046e-05,
+      "loss": 0.3303,
+      "step": 160
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.228827499108435,
+      "learning_rate": 1.949604379370345e-05,
+      "loss": 0.3232,
+      "step": 161
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.31390553334697,
+      "learning_rate": 1.9487887022684336e-05,
+      "loss": 0.3011,
+      "step": 162
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.0420619190076525,
+      "learning_rate": 1.9479666504075737e-05,
+      "loss": 0.3563,
+      "step": 163
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.9863270736505325,
+      "learning_rate": 1.9471382293110004e-05,
+      "loss": 0.3037,
+      "step": 164
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.090603011792817,
+      "learning_rate": 1.946303444544741e-05,
+      "loss": 0.3123,
+      "step": 165
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.1867676870087167,
+      "learning_rate": 1.9454623017175814e-05,
+      "loss": 0.2848,
+      "step": 166
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.2375862840491663,
+      "learning_rate": 1.9446148064810243e-05,
+      "loss": 0.2584,
+      "step": 167
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.073521758948637,
+      "learning_rate": 1.943760964529255e-05,
+      "loss": 0.3982,
+      "step": 168
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.277857419952404,
+      "learning_rate": 1.9429007815990995e-05,
+      "loss": 0.2817,
+      "step": 169
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.860132680882802,
+      "learning_rate": 1.9420342634699893e-05,
+      "loss": 0.3789,
+      "step": 170
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.738288137502009,
+      "learning_rate": 1.9411614159639206e-05,
+      "loss": 0.235,
+      "step": 171
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.0847987605412643,
+      "learning_rate": 1.9402822449454154e-05,
+      "loss": 0.3787,
+      "step": 172
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.678241605511051,
+      "learning_rate": 1.9393967563214833e-05,
+      "loss": 0.3565,
+      "step": 173
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.3794557534236187,
+      "learning_rate": 1.9385049560415794e-05,
+      "loss": 0.3767,
+      "step": 174
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.18026602926335,
+      "learning_rate": 1.937606850097567e-05,
+      "loss": 0.3241,
+      "step": 175
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.839031218674835,
+      "learning_rate": 1.936702444523675e-05,
+      "loss": 0.3692,
+      "step": 176
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.0423888923774403,
+      "learning_rate": 1.9357917453964604e-05,
+      "loss": 0.2939,
+      "step": 177
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.0630953241704497,
+      "learning_rate": 1.9348747588347637e-05,
+      "loss": 0.3033,
+      "step": 178
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.003742056937922,
+      "learning_rate": 1.9339514909996706e-05,
+      "loss": 0.3023,
+      "step": 179
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.183058419141199,
+      "learning_rate": 1.9330219480944693e-05,
+      "loss": 0.3104,
+      "step": 180
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.0962519403436466,
+      "learning_rate": 1.9320861363646094e-05,
+      "loss": 0.3549,
+      "step": 181
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.5773134400966646,
+      "learning_rate": 1.9311440620976597e-05,
+      "loss": 0.4265,
+      "step": 182
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7220486958043284,
+      "learning_rate": 1.9301957316232658e-05,
+      "loss": 0.2717,
+      "step": 183
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.9094558569410807,
+      "learning_rate": 1.929241151313108e-05,
+      "loss": 0.3104,
+      "step": 184
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9500751357137525,
+      "learning_rate": 1.928280327580858e-05,
+      "loss": 0.3202,
+      "step": 185
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.4975322789932215,
+      "learning_rate": 1.9273132668821363e-05,
+      "loss": 0.3126,
+      "step": 186
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.917476755564393,
+      "learning_rate": 1.9263399757144686e-05,
+      "loss": 0.2528,
+      "step": 187
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.143478191577973,
+      "learning_rate": 1.925360460617242e-05,
+      "loss": 0.28,
+      "step": 188
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.5405298702531005,
+      "learning_rate": 1.9243747281716604e-05,
+      "loss": 0.378,
+      "step": 189
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.0155678579375906,
+      "learning_rate": 1.9233827850007028e-05,
+      "loss": 0.2901,
+      "step": 190
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.3060470653500555,
+      "learning_rate": 1.9223846377690754e-05,
+      "loss": 0.3193,
+      "step": 191
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.8900227849807494,
+      "learning_rate": 1.9213802931831697e-05,
+      "loss": 0.3628,
+      "step": 192
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9354932437719492,
+      "learning_rate": 1.9203697579910156e-05,
+      "loss": 0.2377,
+      "step": 193
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.395154833334844,
+      "learning_rate": 1.9193530389822364e-05,
+      "loss": 0.2803,
+      "step": 194
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.1297129261093817,
+      "learning_rate": 1.9183301429880045e-05,
+      "loss": 0.3457,
+      "step": 195
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9248541780760582,
+      "learning_rate": 1.9173010768809934e-05,
+      "loss": 0.283,
+      "step": 196
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.039424995239351,
+      "learning_rate": 1.9162658475753328e-05,
+      "loss": 0.2603,
+      "step": 197
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.4853584640009045,
+      "learning_rate": 1.915224462026563e-05,
+      "loss": 0.394,
+      "step": 198
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.046950886966381,
+      "learning_rate": 1.9141769272315857e-05,
+      "loss": 0.3757,
+      "step": 199
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.036507295142277,
+      "learning_rate": 1.913123250228619e-05,
+      "loss": 0.2404,
+      "step": 200
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.367967992606462,
+      "learning_rate": 1.9120634380971497e-05,
+      "loss": 0.399,
+      "step": 201
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.3778783287603904,
+      "learning_rate": 1.9109974979578852e-05,
+      "loss": 0.2611,
+      "step": 202
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.7971696611984327,
+      "learning_rate": 1.9099254369727062e-05,
+      "loss": 0.415,
+      "step": 203
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.3780933781343143,
+      "learning_rate": 1.9088472623446182e-05,
+      "loss": 0.3033,
+      "step": 204
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.213758758560456,
+      "learning_rate": 1.9077629813177038e-05,
+      "loss": 0.3257,
+      "step": 205
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7096356571332894,
+      "learning_rate": 1.9066726011770725e-05,
+      "loss": 0.2746,
+      "step": 206
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.143024684014943,
+      "learning_rate": 1.9055761292488142e-05,
+      "loss": 0.2936,
+      "step": 207
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.174410493141588,
+      "learning_rate": 1.9044735728999472e-05,
+      "loss": 0.3359,
+      "step": 208
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.8253319924788767,
+      "learning_rate": 1.90336493953837e-05,
+      "loss": 0.3337,
+      "step": 209
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7564814607045327,
+      "learning_rate": 1.9022502366128136e-05,
+      "loss": 0.2585,
+      "step": 210
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8003063075201506,
+      "learning_rate": 1.9011294716127867e-05,
+      "loss": 0.2335,
+      "step": 211
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.0676081597671923,
+      "learning_rate": 1.90000265206853e-05,
+      "loss": 0.3342,
+      "step": 212
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9186653826713473,
+      "learning_rate": 1.898869785550963e-05,
+      "loss": 0.2994,
+      "step": 213
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8847621771847654,
+      "learning_rate": 1.897730879671634e-05,
+      "loss": 0.2403,
+      "step": 214
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6722290718517414,
+      "learning_rate": 1.8965859420826685e-05,
+      "loss": 0.2601,
+      "step": 215
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.7386298568668535,
+      "learning_rate": 1.8954349804767185e-05,
+      "loss": 0.3763,
+      "step": 216
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.5273232386683633,
+      "learning_rate": 1.8942780025869097e-05,
+      "loss": 0.3627,
+      "step": 217
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.0986675503249828,
+      "learning_rate": 1.8931150161867917e-05,
+      "loss": 0.2802,
+      "step": 218
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9864578701687525,
+      "learning_rate": 1.891946029090283e-05,
+      "loss": 0.3077,
+      "step": 219
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.3538318431562364,
+      "learning_rate": 1.8907710491516197e-05,
+      "loss": 0.3164,
+      "step": 220
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.3833491446751607,
+      "learning_rate": 1.889590084265304e-05,
+      "loss": 0.309,
+      "step": 221
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.1589341969698133,
+      "learning_rate": 1.8884031423660492e-05,
+      "loss": 0.2376,
+      "step": 222
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.710090625928746,
+      "learning_rate": 1.887210231428727e-05,
+      "loss": 0.3396,
+      "step": 223
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7501332775825373,
+      "learning_rate": 1.8860113594683148e-05,
+      "loss": 0.2592,
+      "step": 224
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.6999761997206453,
+      "learning_rate": 1.884806534539841e-05,
+      "loss": 0.3225,
+      "step": 225
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.3665867811833197,
+      "learning_rate": 1.8835957647383304e-05,
+      "loss": 0.3459,
+      "step": 226
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.398153025260128,
+      "learning_rate": 1.882379058198751e-05,
+      "loss": 0.2798,
+      "step": 227
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.401401228305016,
+      "learning_rate": 1.8811564230959585e-05,
+      "loss": 0.4071,
+      "step": 228
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.5093960148155054,
+      "learning_rate": 1.8799278676446425e-05,
+      "loss": 0.3185,
+      "step": 229
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.6890640048458447,
+      "learning_rate": 1.878693400099269e-05,
+      "loss": 0.321,
+      "step": 230
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.0892133868732587,
+      "learning_rate": 1.877453028754028e-05,
+      "loss": 0.3648,
+      "step": 231
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.365123613242651,
+      "learning_rate": 1.8762067619427745e-05,
+      "loss": 0.3318,
+      "step": 232
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.2363400437558583,
+      "learning_rate": 1.874954608038976e-05,
+      "loss": 0.3309,
+      "step": 233
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.1067878905016864,
+      "learning_rate": 1.8736965754556527e-05,
+      "loss": 0.3518,
+      "step": 234
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.939495251491112,
+      "learning_rate": 1.8724326726453244e-05,
+      "loss": 0.3071,
+      "step": 235
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.428885090431243,
+      "learning_rate": 1.8711629080999506e-05,
+      "loss": 0.3104,
+      "step": 236
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.711241537545367,
+      "learning_rate": 1.8698872903508755e-05,
+      "loss": 0.3579,
+      "step": 237
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.120518574195615,
+      "learning_rate": 1.86860582796877e-05,
+      "loss": 0.2646,
+      "step": 238
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.228265996988712,
+      "learning_rate": 1.867318529563574e-05,
+      "loss": 0.3185,
+      "step": 239
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.2181290188623852,
+      "learning_rate": 1.866025403784439e-05,
+      "loss": 0.312,
+      "step": 240
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.19939228581308,
+      "learning_rate": 1.8647264593196687e-05,
+      "loss": 0.2936,
+      "step": 241
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.450620349638576,
+      "learning_rate": 1.8634217048966638e-05,
+      "loss": 0.3779,
+      "step": 242
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.2394391810144008,
+      "learning_rate": 1.8621111492818585e-05,
+      "loss": 0.3442,
+      "step": 243
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9643676315139884,
+      "learning_rate": 1.8607948012806664e-05,
+      "loss": 0.2764,
+      "step": 244
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.636425096704688,
+      "learning_rate": 1.8594726697374173e-05,
+      "loss": 0.2618,
+      "step": 245
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9293405761057236,
+      "learning_rate": 1.858144763535302e-05,
+      "loss": 0.2979,
+      "step": 246
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.1636254484392103,
+      "learning_rate": 1.8568110915963082e-05,
+      "loss": 0.3102,
+      "step": 247
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.877917319125591,
+      "learning_rate": 1.855471662881164e-05,
+      "loss": 0.3791,
+      "step": 248
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5670099189407107,
+      "learning_rate": 1.8541264863892755e-05,
+      "loss": 0.1924,
+      "step": 249
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.037307667342044,
+      "learning_rate": 1.852775571158668e-05,
+      "loss": 0.2786,
+      "step": 250
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7947661624098736,
+      "learning_rate": 1.8514189262659235e-05,
+      "loss": 0.2738,
+      "step": 251
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.3960834409580545,
+      "learning_rate": 1.8500565608261215e-05,
+      "loss": 0.2794,
+      "step": 252
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.405818356051298,
+      "learning_rate": 1.8486884839927768e-05,
+      "loss": 0.3228,
+      "step": 253
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.407679300709315,
+      "learning_rate": 1.8473147049577777e-05,
+      "loss": 0.3923,
+      "step": 254
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7377355703462052,
+      "learning_rate": 1.845935232951325e-05,
+      "loss": 0.2699,
+      "step": 255
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.2265681252326313,
+      "learning_rate": 1.8445500772418697e-05,
+      "loss": 0.2771,
+      "step": 256
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.771204006061874,
+      "learning_rate": 1.8431592471360506e-05,
+      "loss": 0.314,
+      "step": 257
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.800034549232218,
+      "learning_rate": 1.8417627519786317e-05,
+      "loss": 0.2969,
+      "step": 258
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.119584706856121,
+      "learning_rate": 1.84036060115244e-05,
+      "loss": 0.3011,
+      "step": 259
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.5245990027253953,
+      "learning_rate": 1.8389528040783014e-05,
+      "loss": 0.2768,
+      "step": 260
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.5058661829859656,
+      "learning_rate": 1.837539370214979e-05,
+      "loss": 0.3769,
+      "step": 261
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.0381135747014842,
+      "learning_rate": 1.836120309059107e-05,
+      "loss": 0.272,
+      "step": 262
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.477627565212375,
+      "learning_rate": 1.8346956301451303e-05,
+      "loss": 0.351,
+      "step": 263
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.2918977462261503,
+      "learning_rate": 1.8332653430452375e-05,
+      "loss": 0.3292,
+      "step": 264
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9162246172960327,
+      "learning_rate": 1.8318294573692987e-05,
+      "loss": 0.3139,
+      "step": 265
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.008395659255134,
+      "learning_rate": 1.8303879827647977e-05,
+      "loss": 0.3573,
+      "step": 266
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4838485512972532,
+      "learning_rate": 1.828940928916772e-05,
+      "loss": 0.2945,
+      "step": 267
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.640319813204331,
+      "learning_rate": 1.8274883055477436e-05,
+      "loss": 0.4545,
+      "step": 268
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.768495647429907,
+      "learning_rate": 1.826030122417656e-05,
+      "loss": 0.3195,
+      "step": 269
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.0356888861889466,
+      "learning_rate": 1.8245663893238075e-05,
+      "loss": 0.3294,
+      "step": 270
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.227539950418251,
+      "learning_rate": 1.8230971161007854e-05,
+      "loss": 0.3633,
+      "step": 271
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8036437828957987,
+      "learning_rate": 1.821622312620401e-05,
+      "loss": 0.2425,
+      "step": 272
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.39188092072465,
+      "learning_rate": 1.8201419887916216e-05,
+      "loss": 0.3828,
+      "step": 273
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.0995389648627083,
+      "learning_rate": 1.8186561545605055e-05,
+      "loss": 0.2721,
+      "step": 274
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8474331750365451,
+      "learning_rate": 1.8171648199101347e-05,
+      "loss": 0.3207,
+      "step": 275
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.981814231675974,
+      "learning_rate": 1.815667994860547e-05,
+      "loss": 0.2448,
+      "step": 276
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9132982718549951,
+      "learning_rate": 1.814165689468669e-05,
+      "loss": 0.29,
+      "step": 277
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.1800413072668343,
+      "learning_rate": 1.8126579138282502e-05,
+      "loss": 0.3054,
+      "step": 278
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.561366521621535,
+      "learning_rate": 1.811144678069793e-05,
+      "loss": 0.3533,
+      "step": 279
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.754016144834769,
+      "learning_rate": 1.809625992360485e-05,
+      "loss": 0.2538,
+      "step": 280
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.162127975656259,
+      "learning_rate": 1.8081018669041324e-05,
+      "loss": 0.293,
+      "step": 281
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.678553745454881,
+      "learning_rate": 1.8065723119410885e-05,
+      "loss": 0.2674,
+      "step": 282
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9784584585424474,
+      "learning_rate": 1.8050373377481878e-05,
+      "loss": 0.2634,
+      "step": 283
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.289698822025373,
+      "learning_rate": 1.803496954638676e-05,
+      "loss": 0.3278,
+      "step": 284
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.0546256448848754,
+      "learning_rate": 1.801951172962139e-05,
+      "loss": 0.3314,
+      "step": 285
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.0189650291686867,
+      "learning_rate": 1.8004000031044363e-05,
+      "loss": 0.3137,
+      "step": 286
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.2793742115240057,
+      "learning_rate": 1.798843455487629e-05,
+      "loss": 0.2916,
+      "step": 287
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.978610211105925,
+      "learning_rate": 1.7972815405699105e-05,
+      "loss": 0.312,
+      "step": 288
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.2311767161123073,
+      "learning_rate": 1.7957142688455362e-05,
+      "loss": 0.3389,
+      "step": 289
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.05360168559885,
+      "learning_rate": 1.7941416508447537e-05,
+      "loss": 0.2494,
+      "step": 290
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.80172085924464,
+      "learning_rate": 1.7925636971337304e-05,
+      "loss": 0.3029,
+      "step": 291
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.327817348612163,
+      "learning_rate": 1.7909804183144837e-05,
+      "loss": 0.2801,
+      "step": 292
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.1848347825813543,
+      "learning_rate": 1.7893918250248106e-05,
+      "loss": 0.2862,
+      "step": 293
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.1760889724855557,
+      "learning_rate": 1.7877979279382135e-05,
+      "loss": 0.2603,
+      "step": 294
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.131768994375875,
+      "learning_rate": 1.7861987377638314e-05,
+      "loss": 0.2858,
+      "step": 295
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.229823163798362,
+      "learning_rate": 1.784594265246366e-05,
+      "loss": 0.3121,
+      "step": 296
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.299352725908318,
+      "learning_rate": 1.782984521166011e-05,
+      "loss": 0.2889,
+      "step": 297
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.5621113668957247,
+      "learning_rate": 1.781369516338378e-05,
+      "loss": 0.4031,
+      "step": 298
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7787192026448813,
+      "learning_rate": 1.7797492616144256e-05,
+      "loss": 0.3196,
+      "step": 299
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.825677662281392,
+      "learning_rate": 1.7781237678803845e-05,
+      "loss": 0.2472,
+      "step": 300
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9138575999411622,
+      "learning_rate": 1.7764930460576867e-05,
+      "loss": 0.3133,
+      "step": 301
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.424454787520677,
+      "learning_rate": 1.77485710710289e-05,
+      "loss": 0.4237,
+      "step": 302
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6596834658332436,
+      "learning_rate": 1.7732159620076053e-05,
+      "loss": 0.2615,
+      "step": 303
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.915662757060755,
+      "learning_rate": 1.7715696217984233e-05,
+      "loss": 0.3257,
+      "step": 304
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8157063327790481,
+      "learning_rate": 1.7699180975368397e-05,
+      "loss": 0.3,
+      "step": 305
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.1772564616176617,
+      "learning_rate": 1.7682614003191807e-05,
+      "loss": 0.3613,
+      "step": 306
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6345766099529573,
+      "learning_rate": 1.7665995412765287e-05,
+      "loss": 0.2321,
+      "step": 307
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6761780023043686,
+      "learning_rate": 1.764932531574648e-05,
+      "loss": 0.2916,
+      "step": 308
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.477586537493892,
+      "learning_rate": 1.7632603824139086e-05,
+      "loss": 0.2836,
+      "step": 309
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.2248085696051487,
+      "learning_rate": 1.761583105029213e-05,
+      "loss": 0.3312,
+      "step": 310
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.873254693799373,
+      "learning_rate": 1.759900710689918e-05,
+      "loss": 0.2996,
+      "step": 311
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.6963403979706424,
+      "learning_rate": 1.7582132106997615e-05,
+      "loss": 0.3688,
+      "step": 312
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9275761025445475,
+      "learning_rate": 1.7565206163967847e-05,
+      "loss": 0.3042,
+      "step": 313
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.0767014249772777,
+      "learning_rate": 1.7548229391532572e-05,
+      "loss": 0.3169,
+      "step": 314
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7919429635231559,
+      "learning_rate": 1.7531201903755994e-05,
+      "loss": 0.2714,
+      "step": 315
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.2183254478234824,
+      "learning_rate": 1.7514123815043073e-05,
+      "loss": 0.3313,
+      "step": 316
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.869049065360244,
+      "learning_rate": 1.7496995240138745e-05,
+      "loss": 0.2469,
+      "step": 317
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.1285116739745717,
+      "learning_rate": 1.747981629412715e-05,
+      "loss": 0.2989,
+      "step": 318
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.332735444074386,
+      "learning_rate": 1.7462587092430877e-05,
+      "loss": 0.291,
+      "step": 319
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.007902555326692,
+      "learning_rate": 1.7445307750810153e-05,
+      "loss": 0.2338,
+      "step": 320
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.074668198872314,
+      "learning_rate": 1.742797838536211e-05,
+      "loss": 0.3027,
+      "step": 321
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8245178053269033,
+      "learning_rate": 1.741059911251997e-05,
+      "loss": 0.2782,
+      "step": 322
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8670263646678231,
+      "learning_rate": 1.7393170049052274e-05,
+      "loss": 0.3038,
+      "step": 323
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.4400014121299844,
+      "learning_rate": 1.7375691312062102e-05,
+      "loss": 0.3773,
+      "step": 324
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7893699834417385,
+      "learning_rate": 1.7358163018986282e-05,
+      "loss": 0.236,
+      "step": 325
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.356871958108051,
+      "learning_rate": 1.7340585287594605e-05,
+      "loss": 0.4084,
+      "step": 326
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.119181522421401,
+      "learning_rate": 1.7322958235989015e-05,
+      "loss": 0.3592,
+      "step": 327
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7688080795720216,
+      "learning_rate": 1.730528198260285e-05,
+      "loss": 0.2515,
+      "step": 328
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.0648152043802397,
+      "learning_rate": 1.728755664620002e-05,
+      "loss": 0.3064,
+      "step": 329
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9486429621552654,
+      "learning_rate": 1.7269782345874204e-05,
+      "loss": 0.2793,
+      "step": 330
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.770410943564192,
+      "learning_rate": 1.7251959201048083e-05,
+      "loss": 0.4832,
+      "step": 331
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.416701783014655,
+      "learning_rate": 1.72340873314725e-05,
+      "loss": 0.4427,
+      "step": 332
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.216087293349358,
+      "learning_rate": 1.7216166857225674e-05,
+      "loss": 0.2783,
+      "step": 333
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.3919230833571863,
+      "learning_rate": 1.7198197898712402e-05,
+      "loss": 0.3604,
+      "step": 334
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.854273317784981,
+      "learning_rate": 1.718018057666323e-05,
+      "loss": 0.2478,
+      "step": 335
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9702887673647873,
+      "learning_rate": 1.7162115012133643e-05,
+      "loss": 0.2942,
+      "step": 336
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9030191103194438,
+      "learning_rate": 1.7144001326503276e-05,
+      "loss": 0.2196,
+      "step": 337
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9079753005811861,
+      "learning_rate": 1.7125839641475074e-05,
+      "loss": 0.2952,
+      "step": 338
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.918128245238814,
+      "learning_rate": 1.7107630079074477e-05,
+      "loss": 0.2631,
+      "step": 339
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.5192828679265746,
+      "learning_rate": 1.7089372761648617e-05,
+      "loss": 0.3316,
+      "step": 340
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.522800037570299,
+      "learning_rate": 1.7071067811865477e-05,
+      "loss": 0.3521,
+      "step": 341
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.038572595618781,
+      "learning_rate": 1.7052715352713076e-05,
+      "loss": 0.3539,
+      "step": 342
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.622293826263427,
+      "learning_rate": 1.7034315507498637e-05,
+      "loss": 0.448,
+      "step": 343
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.090457624775084,
+      "learning_rate": 1.7015868399847768e-05,
+      "loss": 0.305,
+      "step": 344
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.5963733633562818,
+      "learning_rate": 1.6997374153703625e-05,
+      "loss": 0.3379,
+      "step": 345
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.917460984311743,
+      "learning_rate": 1.6978832893326074e-05,
+      "loss": 0.2737,
+      "step": 346
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9740793094943232,
+      "learning_rate": 1.6960244743290867e-05,
+      "loss": 0.3296,
+      "step": 347
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.569033540648746,
+      "learning_rate": 1.6941609828488806e-05,
+      "loss": 0.3888,
+      "step": 348
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4877471722417306,
+      "learning_rate": 1.6922928274124887e-05,
+      "loss": 0.2176,
+      "step": 349
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.313163996098816,
+      "learning_rate": 1.690420020571747e-05,
+      "loss": 0.3185,
+      "step": 350
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.1610104071558283,
+      "learning_rate": 1.6885425749097444e-05,
+      "loss": 0.347,
+      "step": 351
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.926355155438669,
+      "learning_rate": 1.686660503040737e-05,
+      "loss": 0.3048,
+      "step": 352
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.550898935585205,
+      "learning_rate": 1.6847738176100632e-05,
+      "loss": 0.2524,
+      "step": 353
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8727473492933593,
+      "learning_rate": 1.6828825312940594e-05,
+      "loss": 0.279,
+      "step": 354
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3169524927684164,
+      "learning_rate": 1.6809866567999752e-05,
+      "loss": 0.2292,
+      "step": 355
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9614958078211489,
+      "learning_rate": 1.6790862068658863e-05,
+      "loss": 0.3098,
+      "step": 356
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.1570857664260474,
+      "learning_rate": 1.677181194260611e-05,
+      "loss": 0.2656,
+      "step": 357
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.2008427558851613,
+      "learning_rate": 1.675271631783623e-05,
+      "loss": 0.3315,
+      "step": 358
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7402168646872844,
+      "learning_rate": 1.673357532264966e-05,
+      "loss": 0.255,
+      "step": 359
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.052681701712784,
+      "learning_rate": 1.671438908565167e-05,
+      "loss": 0.3002,
+      "step": 360
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.0854049905321985,
+      "learning_rate": 1.6695157735751512e-05,
+      "loss": 0.3146,
+      "step": 361
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.556001474745676,
+      "learning_rate": 1.667588140216154e-05,
+      "loss": 0.281,
+      "step": 362
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.328001289077992,
+      "learning_rate": 1.665656021439633e-05,
+      "loss": 0.3679,
+      "step": 363
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.246735349527516,
+      "learning_rate": 1.663719430227186e-05,
+      "loss": 0.3611,
+      "step": 364
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.302536600638992,
+      "learning_rate": 1.6617783795904564e-05,
+      "loss": 0.2772,
+      "step": 365
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7827203381412275,
+      "learning_rate": 1.6598328825710536e-05,
+      "loss": 0.2599,
+      "step": 366
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.0327821167883195,
+      "learning_rate": 1.6578829522404586e-05,
+      "loss": 0.3104,
+      "step": 367
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.3139615273673986,
+      "learning_rate": 1.65592860169994e-05,
+      "loss": 0.3757,
+      "step": 368
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8404784917427333,
+      "learning_rate": 1.653969844080466e-05,
+      "loss": 0.2065,
+      "step": 369
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9029373978368715,
+      "learning_rate": 1.6520066925426146e-05,
+      "loss": 0.2541,
+      "step": 370
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.420515783221278,
+      "learning_rate": 1.650039160276485e-05,
+      "loss": 0.362,
+      "step": 371
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7841307975964125,
+      "learning_rate": 1.648067260501611e-05,
+      "loss": 0.3155,
+      "step": 372
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5202485924429843,
+      "learning_rate": 1.646091006466871e-05,
+      "loss": 0.1921,
+      "step": 373
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7645307356443054,
+      "learning_rate": 1.644110411450398e-05,
+      "loss": 0.2744,
+      "step": 374
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.1778579528461544,
+      "learning_rate": 1.6421254887594918e-05,
+      "loss": 0.3207,
+      "step": 375
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.1144744784839755,
+      "learning_rate": 1.6401362517305296e-05,
+      "loss": 0.3438,
+      "step": 376
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.4036210464297634,
+      "learning_rate": 1.6381427137288756e-05,
+      "loss": 0.2557,
+      "step": 377
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.281318088880689,
+      "learning_rate": 1.6361448881487913e-05,
+      "loss": 0.3522,
+      "step": 378
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.442640792379695,
+      "learning_rate": 1.634142788413346e-05,
+      "loss": 0.3291,
+      "step": 379
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.778962002997131,
+      "learning_rate": 1.6321364279743267e-05,
+      "loss": 0.3112,
+      "step": 380
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.482779982218463,
+      "learning_rate": 1.6301258203121463e-05,
+      "loss": 0.3753,
+      "step": 381
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.255116541310975,
+      "learning_rate": 1.628110978935756e-05,
+      "loss": 0.438,
+      "step": 382
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4043721570913874,
+      "learning_rate": 1.6260919173825507e-05,
+      "loss": 0.2462,
+      "step": 383
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.955749694714099,
+      "learning_rate": 1.6240686492182806e-05,
+      "loss": 0.2731,
+      "step": 384
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5061172100523035,
+      "learning_rate": 1.62204118803696e-05,
+      "loss": 0.2472,
+      "step": 385
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.3645366514423483,
+      "learning_rate": 1.6200095474607753e-05,
+      "loss": 0.4294,
+      "step": 386
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7062872716901005,
+      "learning_rate": 1.6179737411399926e-05,
+      "loss": 0.2888,
+      "step": 387
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7788526606346524,
+      "learning_rate": 1.6159337827528686e-05,
+      "loss": 0.297,
+      "step": 388
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.062084476736332,
+      "learning_rate": 1.6138896860055555e-05,
+      "loss": 0.3007,
+      "step": 389
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4604320059002305,
+      "learning_rate": 1.6118414646320115e-05,
+      "loss": 0.2174,
+      "step": 390
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.346784589374752,
+      "learning_rate": 1.6097891323939063e-05,
+      "loss": 0.2916,
+      "step": 391
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.89550070481857,
+      "learning_rate": 1.6077327030805318e-05,
+      "loss": 0.31,
+      "step": 392
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9449292302402192,
+      "learning_rate": 1.6056721905087056e-05,
+      "loss": 0.2325,
+      "step": 393
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6599175197516962,
+      "learning_rate": 1.6036076085226813e-05,
+      "loss": 0.2587,
+      "step": 394
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.221872605847609,
+      "learning_rate": 1.601538970994054e-05,
+      "loss": 0.2651,
+      "step": 395
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.369743617712892,
+      "learning_rate": 1.599466291821666e-05,
+      "loss": 0.3272,
+      "step": 396
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.152582977278543,
+      "learning_rate": 1.597389584931517e-05,
+      "loss": 0.3633,
+      "step": 397
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.149297985911286,
+      "learning_rate": 1.595308864276666e-05,
+      "loss": 0.3414,
+      "step": 398
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.1819931174923166,
+      "learning_rate": 1.593224143837142e-05,
+      "loss": 0.3172,
+      "step": 399
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.165318987684981,
+      "learning_rate": 1.5911354376198468e-05,
+      "loss": 0.3352,
+      "step": 400
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5544032333205273,
+      "learning_rate": 1.589042759658462e-05,
+      "loss": 0.188,
+      "step": 401
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.1268000986916364,
+      "learning_rate": 1.586946124013354e-05,
+      "loss": 0.234,
+      "step": 402
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6649749080409588,
+      "learning_rate": 1.5848455447714822e-05,
+      "loss": 0.2497,
+      "step": 403
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.5600823534050168,
+      "learning_rate": 1.582741036046301e-05,
+      "loss": 0.3956,
+      "step": 404
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.4435713327056052,
+      "learning_rate": 1.580632611977666e-05,
+      "loss": 0.3828,
+      "step": 405
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6251851298524764,
+      "learning_rate": 1.578520286731741e-05,
+      "loss": 0.289,
+      "step": 406
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7105944882630801,
+      "learning_rate": 1.5764040745008987e-05,
+      "loss": 0.2551,
+      "step": 407
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9471548033131576,
+      "learning_rate": 1.5742839895036305e-05,
+      "loss": 0.3029,
+      "step": 408
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.2009574984159666,
+      "learning_rate": 1.572160045984447e-05,
+      "loss": 0.3057,
+      "step": 409
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4997777963014745,
+      "learning_rate": 1.570032258213783e-05,
+      "loss": 0.279,
+      "step": 410
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.2091983353551163,
+      "learning_rate": 1.5679006404879035e-05,
+      "loss": 0.1804,
+      "step": 411
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.2147374257119607,
+      "learning_rate": 1.565765207128805e-05,
+      "loss": 0.3216,
+      "step": 412
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6847496660853745,
+      "learning_rate": 1.5636259724841224e-05,
+      "loss": 0.2444,
+      "step": 413
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.442032050244461,
+      "learning_rate": 1.561482950927029e-05,
+      "loss": 0.2093,
+      "step": 414
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.7144419107938096,
+      "learning_rate": 1.559336156856143e-05,
+      "loss": 0.2823,
+      "step": 415
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8994572657987778,
+      "learning_rate": 1.5571856046954284e-05,
+      "loss": 0.2811,
+      "step": 416
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.4556233383553328,
+      "learning_rate": 1.5550313088941012e-05,
+      "loss": 0.2636,
+      "step": 417
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9438588716317593,
+      "learning_rate": 1.5528732839265272e-05,
+      "loss": 0.237,
+      "step": 418
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9487422925171727,
+      "learning_rate": 1.550711544292131e-05,
+      "loss": 0.2695,
+      "step": 419
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8474563686885068,
+      "learning_rate": 1.5485461045152937e-05,
+      "loss": 0.332,
+      "step": 420
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.213733875483992,
+      "learning_rate": 1.5463769791452574e-05,
+      "loss": 0.3657,
+      "step": 421
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7730710691561862,
+      "learning_rate": 1.5442041827560274e-05,
+      "loss": 0.2537,
+      "step": 422
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.246874442788659,
+      "learning_rate": 1.5420277299462734e-05,
+      "loss": 0.5124,
+      "step": 423
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.05906376120636,
+      "learning_rate": 1.5398476353392323e-05,
+      "loss": 0.239,
+      "step": 424
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5526941505478917,
+      "learning_rate": 1.537663913582611e-05,
+      "loss": 0.232,
+      "step": 425
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8580732186198845,
+      "learning_rate": 1.5354765793484834e-05,
+      "loss": 0.2396,
+      "step": 426
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5006582448794108,
+      "learning_rate": 1.533285647333198e-05,
+      "loss": 0.2019,
+      "step": 427
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.127615426349164,
+      "learning_rate": 1.531091132257275e-05,
+      "loss": 0.3526,
+      "step": 428
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.041420611859174,
+      "learning_rate": 1.5288930488653094e-05,
+      "loss": 0.2818,
+      "step": 429
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.0158826239887415,
+      "learning_rate": 1.52669141192587e-05,
+      "loss": 0.2433,
+      "step": 430
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.35504555694053,
+      "learning_rate": 1.5244862362314021e-05,
+      "loss": 0.4146,
+      "step": 431
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.321139724637557,
+      "learning_rate": 1.5222775365981272e-05,
+      "loss": 0.3337,
+      "step": 432
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.0652742393182297,
+      "learning_rate": 1.5200653278659431e-05,
+      "loss": 0.2628,
+      "step": 433
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.1040864390034226,
+      "learning_rate": 1.5178496248983254e-05,
+      "loss": 0.3277,
+      "step": 434
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.0478273126880238,
+      "learning_rate": 1.5156304425822269e-05,
+      "loss": 0.3138,
+      "step": 435
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3950221395018922,
+      "learning_rate": 1.5134077958279764e-05,
+      "loss": 0.1842,
+      "step": 436
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.656819521166795,
+      "learning_rate": 1.511181699569181e-05,
+      "loss": 0.2883,
+      "step": 437
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.0042138431638254,
+      "learning_rate": 1.5089521687626243e-05,
+      "loss": 0.2961,
+      "step": 438
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7959393612551642,
+      "learning_rate": 1.5067192183881658e-05,
+      "loss": 0.2915,
+      "step": 439
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5773497666048848,
+      "learning_rate": 1.50448286344864e-05,
+      "loss": 0.2384,
+      "step": 440
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8319515701265676,
+      "learning_rate": 1.5022431189697569e-05,
+      "loss": 0.2124,
+      "step": 441
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.352102926044369,
+      "learning_rate": 1.5000000000000002e-05,
+      "loss": 0.2956,
+      "step": 442
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.6574809214182507,
+      "learning_rate": 1.4977535216105258e-05,
+      "loss": 0.3113,
+      "step": 443
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.022537878921683,
+      "learning_rate": 1.4955036988950617e-05,
+      "loss": 0.3537,
+      "step": 444
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7901743008827358,
+      "learning_rate": 1.4932505469698053e-05,
+      "loss": 0.2454,
+      "step": 445
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.5189391342729044,
+      "learning_rate": 1.4909940809733223e-05,
+      "loss": 0.3997,
+      "step": 446
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7952766424539717,
+      "learning_rate": 1.488734316066446e-05,
+      "loss": 0.3098,
+      "step": 447
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.6459613118945837,
+      "learning_rate": 1.4864712674321733e-05,
+      "loss": 0.4042,
+      "step": 448
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8809594247765489,
+      "learning_rate": 1.484204950275565e-05,
+      "loss": 0.2572,
+      "step": 449
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7345503741548023,
+      "learning_rate": 1.4819353798236427e-05,
+      "loss": 0.2967,
+      "step": 450
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.480786580014742,
+      "learning_rate": 1.4796625713252848e-05,
+      "loss": 0.3428,
+      "step": 451
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9650622625519851,
+      "learning_rate": 1.477386540051127e-05,
+      "loss": 0.3437,
+      "step": 452
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6937032220638994,
+      "learning_rate": 1.4751073012934587e-05,
+      "loss": 0.253,
+      "step": 453
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7558596774994375,
+      "learning_rate": 1.4728248703661183e-05,
+      "loss": 0.2542,
+      "step": 454
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9929358246953341,
+      "learning_rate": 1.4705392626043931e-05,
+      "loss": 0.3633,
+      "step": 455
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.810310474439796,
+      "learning_rate": 1.4682504933649144e-05,
+      "loss": 0.283,
+      "step": 456
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.152036182046435,
+      "learning_rate": 1.4659585780255556e-05,
+      "loss": 0.3102,
+      "step": 457
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.689009597364608,
+      "learning_rate": 1.4636635319853274e-05,
+      "loss": 0.2497,
+      "step": 458
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7511702540468523,
+      "learning_rate": 1.461365370664276e-05,
+      "loss": 0.2242,
+      "step": 459
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8016493801156939,
+      "learning_rate": 1.4590641095033786e-05,
+      "loss": 0.2562,
+      "step": 460
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8330222184027876,
+      "learning_rate": 1.4567597639644387e-05,
+      "loss": 0.348,
+      "step": 461
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8884072768255293,
+      "learning_rate": 1.4544523495299843e-05,
+      "loss": 0.3019,
+      "step": 462
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.125164852414625,
+      "learning_rate": 1.4521418817031627e-05,
+      "loss": 0.27,
+      "step": 463
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.1351697700961187,
+      "learning_rate": 1.4498283760076362e-05,
+      "loss": 0.3036,
+      "step": 464
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7649150773280204,
+      "learning_rate": 1.4475118479874775e-05,
+      "loss": 0.2595,
+      "step": 465
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.561271491793302,
+      "learning_rate": 1.445192313207067e-05,
+      "loss": 0.4268,
+      "step": 466
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.0380333698817577,
+      "learning_rate": 1.4428697872509868e-05,
+      "loss": 0.2958,
+      "step": 467
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8504896113827187,
+      "learning_rate": 1.4405442857239151e-05,
+      "loss": 0.2613,
+      "step": 468
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8094111096940708,
+      "learning_rate": 1.4382158242505236e-05,
+      "loss": 0.2721,
+      "step": 469
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8451009864315955,
+      "learning_rate": 1.4358844184753713e-05,
+      "loss": 0.2508,
+      "step": 470
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4184369060807616,
+      "learning_rate": 1.4335500840627988e-05,
+      "loss": 0.208,
+      "step": 471
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.1158834118020655,
+      "learning_rate": 1.4312128366968244e-05,
+      "loss": 0.2629,
+      "step": 472
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.3693919275045663,
+      "learning_rate": 1.4288726920810381e-05,
+      "loss": 0.3818,
+      "step": 473
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5636740891713368,
+      "learning_rate": 1.4265296659384956e-05,
+      "loss": 0.2369,
+      "step": 474
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.200704759747656,
+      "learning_rate": 1.4241837740116133e-05,
+      "loss": 0.3506,
+      "step": 475
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.1643259001356636,
+      "learning_rate": 1.4218350320620625e-05,
+      "loss": 0.3591,
+      "step": 476
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.1092608570864133,
+      "learning_rate": 1.4194834558706632e-05,
+      "loss": 0.3503,
+      "step": 477
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.06478314792136,
+      "learning_rate": 1.4171290612372781e-05,
+      "loss": 0.2355,
+      "step": 478
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6958980700470296,
+      "learning_rate": 1.4147718639807071e-05,
+      "loss": 0.2471,
+      "step": 479
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8849495582079205,
+      "learning_rate": 1.4124118799385797e-05,
+      "loss": 0.2712,
+      "step": 480
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8906644949931855,
+      "learning_rate": 1.4100491249672499e-05,
+      "loss": 0.293,
+      "step": 481
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.931395556520391,
+      "learning_rate": 1.4076836149416889e-05,
+      "loss": 0.2185,
+      "step": 482
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7527868454175564,
+      "learning_rate": 1.405315365755379e-05,
+      "loss": 0.3091,
+      "step": 483
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.900785095851623,
+      "learning_rate": 1.4029443933202059e-05,
+      "loss": 0.2492,
+      "step": 484
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.233763201074291,
+      "learning_rate": 1.4005707135663529e-05,
+      "loss": 0.292,
+      "step": 485
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9402674198597119,
+      "learning_rate": 1.3981943424421932e-05,
+      "loss": 0.2757,
+      "step": 486
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.05490281674057,
+      "learning_rate": 1.3958152959141824e-05,
+      "loss": 0.3103,
+      "step": 487
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.689652030822919,
+      "learning_rate": 1.3934335899667526e-05,
+      "loss": 0.28,
+      "step": 488
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5225074972151993,
+      "learning_rate": 1.3910492406022033e-05,
+      "loss": 0.2385,
+      "step": 489
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8009132690551426,
+      "learning_rate": 1.3886622638405953e-05,
+      "loss": 0.2212,
+      "step": 490
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7608649807194425,
+      "learning_rate": 1.386272675719642e-05,
+      "loss": 0.252,
+      "step": 491
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5938821438110429,
+      "learning_rate": 1.3838804922946027e-05,
+      "loss": 0.2367,
+      "step": 492
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.095727220463024,
+      "learning_rate": 1.381485729638173e-05,
+      "loss": 0.313,
+      "step": 493
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7733127567846445,
+      "learning_rate": 1.3790884038403796e-05,
+      "loss": 0.2192,
+      "step": 494
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8184052409785323,
+      "learning_rate": 1.3766885310084687e-05,
+      "loss": 0.2479,
+      "step": 495
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.2549269220342394,
+      "learning_rate": 1.374286127266801e-05,
+      "loss": 0.3441,
+      "step": 496
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7087572677999134,
+      "learning_rate": 1.3718812087567414e-05,
+      "loss": 0.2949,
+      "step": 497
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.614404012946222,
+      "learning_rate": 1.3694737916365517e-05,
+      "loss": 0.2777,
+      "step": 498
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.0160740135079345,
+      "learning_rate": 1.36706389208128e-05,
+      "loss": 0.295,
+      "step": 499
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7024108871165236,
+      "learning_rate": 1.3646515262826551e-05,
+      "loss": 0.2313,
+      "step": 500
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8676689841332443,
+      "learning_rate": 1.3622367104489757e-05,
+      "loss": 0.3114,
+      "step": 501
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5476973040713458,
+      "learning_rate": 1.3598194608050011e-05,
+      "loss": 0.2199,
+      "step": 502
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9531175723466303,
+      "learning_rate": 1.357399793591844e-05,
+      "loss": 0.2401,
+      "step": 503
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.836145460893975,
+      "learning_rate": 1.354977725066859e-05,
+      "loss": 0.2551,
+      "step": 504
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.5413490229034283,
+      "learning_rate": 1.3525532715035365e-05,
+      "loss": 0.3665,
+      "step": 505
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7210272719851107,
+      "learning_rate": 1.3501264491913909e-05,
+      "loss": 0.2342,
+      "step": 506
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.152971499183195,
+      "learning_rate": 1.3476972744358507e-05,
+      "loss": 0.277,
+      "step": 507
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.963619315388926,
+      "learning_rate": 1.3452657635581521e-05,
+      "loss": 0.2778,
+      "step": 508
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.261926107950733,
+      "learning_rate": 1.3428319328952254e-05,
+      "loss": 0.3748,
+      "step": 509
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.002171823722819,
+      "learning_rate": 1.3403957987995884e-05,
+      "loss": 0.2425,
+      "step": 510
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.273686726573573,
+      "learning_rate": 1.337957377639235e-05,
+      "loss": 0.3396,
+      "step": 511
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.3312186957241137,
+      "learning_rate": 1.335516685797525e-05,
+      "loss": 0.3483,
+      "step": 512
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.9943390495589555,
+      "learning_rate": 1.333073739673076e-05,
+      "loss": 0.3134,
+      "step": 513
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.9690670721926964,
+      "learning_rate": 1.3306285556796494e-05,
+      "loss": 0.2985,
+      "step": 514
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.00802440137394,
+      "learning_rate": 1.3281811502460448e-05,
+      "loss": 0.2536,
+      "step": 515
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.301585415154793,
+      "learning_rate": 1.3257315398159865e-05,
+      "loss": 0.2793,
+      "step": 516
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7858563085952386,
+      "learning_rate": 1.3232797408480128e-05,
+      "loss": 0.2701,
+      "step": 517
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.017873008101678,
+      "learning_rate": 1.3208257698153677e-05,
+      "loss": 0.3334,
+      "step": 518
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8364335369046931,
+      "learning_rate": 1.3183696432058889e-05,
+      "loss": 0.2536,
+      "step": 519
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7056348936779424,
+      "learning_rate": 1.3159113775218963e-05,
+      "loss": 0.3034,
+      "step": 520
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.227118298645537,
+      "learning_rate": 1.3134509892800821e-05,
+      "loss": 0.3322,
+      "step": 521
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.117256632364151,
+      "learning_rate": 1.3109884950114007e-05,
+      "loss": 0.2889,
+      "step": 522
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4273413888554127,
+      "learning_rate": 1.3085239112609549e-05,
+      "loss": 0.2379,
+      "step": 523
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9575603339371808,
+      "learning_rate": 1.3060572545878875e-05,
+      "loss": 0.2725,
+      "step": 524
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8170930579170255,
+      "learning_rate": 1.3035885415652685e-05,
+      "loss": 0.3094,
+      "step": 525
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8284584543080296,
+      "learning_rate": 1.3011177887799846e-05,
+      "loss": 0.2835,
+      "step": 526
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.569945283347339,
+      "learning_rate": 1.2986450128326267e-05,
+      "loss": 0.2072,
+      "step": 527
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8533615577812252,
+      "learning_rate": 1.2961702303373795e-05,
+      "loss": 0.2079,
+      "step": 528
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9399595281432296,
+      "learning_rate": 1.2936934579219094e-05,
+      "loss": 0.2467,
+      "step": 529
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.213415657483411,
+      "learning_rate": 1.2912147122272523e-05,
+      "loss": 0.246,
+      "step": 530
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.4888356873561297,
+      "learning_rate": 1.2887340099077024e-05,
+      "loss": 0.3111,
+      "step": 531
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9573742795078264,
+      "learning_rate": 1.2862513676307009e-05,
+      "loss": 0.2122,
+      "step": 532
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7014691865915594,
+      "learning_rate": 1.283766802076722e-05,
+      "loss": 0.2022,
+      "step": 533
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.190373861234172,
+      "learning_rate": 1.2812803299391629e-05,
+      "loss": 0.2752,
+      "step": 534
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.6302470512377956,
+      "learning_rate": 1.2787919679242307e-05,
+      "loss": 0.3832,
+      "step": 535
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.6985835144579573,
+      "learning_rate": 1.2763017327508304e-05,
+      "loss": 0.3706,
+      "step": 536
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.1919668554044596,
+      "learning_rate": 1.2738096411504521e-05,
+      "loss": 0.3172,
+      "step": 537
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.965514372155199,
+      "learning_rate": 1.271315709867059e-05,
+      "loss": 0.2695,
+      "step": 538
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9788197014312077,
+      "learning_rate": 1.2688199556569753e-05,
+      "loss": 0.3215,
+      "step": 539
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8117898000767236,
+      "learning_rate": 1.2663223952887724e-05,
+      "loss": 0.2785,
+      "step": 540
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.758141248372573,
+      "learning_rate": 1.263823045543158e-05,
+      "loss": 0.4488,
+      "step": 541
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.444676019103691,
+      "learning_rate": 1.2613219232128608e-05,
+      "loss": 0.2588,
+      "step": 542
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7579524367776718,
+      "learning_rate": 1.2588190451025209e-05,
+      "loss": 0.2807,
+      "step": 543
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.0539384138121735,
+      "learning_rate": 1.2563144280285742e-05,
+      "loss": 0.3425,
+      "step": 544
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.123962428661618,
+      "learning_rate": 1.2538080888191408e-05,
+      "loss": 0.3437,
+      "step": 545
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.848843665190907,
+      "learning_rate": 1.2513000443139112e-05,
+      "loss": 0.2328,
+      "step": 546
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9068583881882215,
+      "learning_rate": 1.2487903113640338e-05,
+      "loss": 0.255,
+      "step": 547
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5499884922847627,
+      "learning_rate": 1.2462789068320016e-05,
+      "loss": 0.2282,
+      "step": 548
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7518125236459046,
+      "learning_rate": 1.2437658475915378e-05,
+      "loss": 0.2218,
+      "step": 549
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8114924338777028,
+      "learning_rate": 1.2412511505274845e-05,
+      "loss": 0.2805,
+      "step": 550
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9244797085738425,
+      "learning_rate": 1.2387348325356873e-05,
+      "loss": 0.3397,
+      "step": 551
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5720493713091632,
+      "learning_rate": 1.2362169105228828e-05,
+      "loss": 0.2528,
+      "step": 552
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8428242120952283,
+      "learning_rate": 1.2336974014065844e-05,
+      "loss": 0.2768,
+      "step": 553
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.835184040116667,
+      "learning_rate": 1.23117632211497e-05,
+      "loss": 0.2833,
+      "step": 554
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.214221991382723,
+      "learning_rate": 1.2286536895867656e-05,
+      "loss": 0.3365,
+      "step": 555
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.235743345209664,
+      "learning_rate": 1.2261295207711347e-05,
+      "loss": 0.3073,
+      "step": 556
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.628463064921636,
+      "learning_rate": 1.2236038326275628e-05,
+      "loss": 0.2442,
+      "step": 557
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6633400921663062,
+      "learning_rate": 1.221076642125742e-05,
+      "loss": 0.2829,
+      "step": 558
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.9337054901311905,
+      "learning_rate": 1.2185479662454596e-05,
+      "loss": 0.2395,
+      "step": 559
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.906099833447381,
+      "learning_rate": 1.2160178219764838e-05,
+      "loss": 0.2846,
+      "step": 560
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.9714262464040193,
+      "learning_rate": 1.2134862263184467e-05,
+      "loss": 0.2152,
+      "step": 561
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.057145897252437,
+      "learning_rate": 1.2109531962807333e-05,
+      "loss": 0.2386,
+      "step": 562
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.9480146225635264,
+      "learning_rate": 1.2084187488823657e-05,
+      "loss": 0.3043,
+      "step": 563
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.9995365117357957,
+      "learning_rate": 1.2058829011518896e-05,
+      "loss": 0.2283,
+      "step": 564
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.9136198948337104,
+      "learning_rate": 1.2033456701272577e-05,
+      "loss": 0.3082,
+      "step": 565
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.307133524485423,
+      "learning_rate": 1.2008070728557186e-05,
+      "loss": 0.2434,
+      "step": 566
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.1483944519592906,
+      "learning_rate": 1.1982671263936996e-05,
+      "loss": 0.3163,
+      "step": 567
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.0153764114524515,
+      "learning_rate": 1.195725847806693e-05,
+      "loss": 0.2928,
+      "step": 568
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7916502641369683,
+      "learning_rate": 1.193183254169142e-05,
+      "loss": 0.2365,
+      "step": 569
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.91929055961395,
+      "learning_rate": 1.1906393625643244e-05,
+      "loss": 0.267,
+      "step": 570
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.725549936564891,
+      "learning_rate": 1.1880941900842398e-05,
+      "loss": 0.294,
+      "step": 571
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.867199788127604,
+      "learning_rate": 1.1855477538294934e-05,
+      "loss": 0.2054,
+      "step": 572
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.1796669209124824,
+      "learning_rate": 1.1830000709091816e-05,
+      "loss": 0.3281,
+      "step": 573
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6222853120409504,
+      "learning_rate": 1.1804511584407763e-05,
+      "loss": 0.2956,
+      "step": 574
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.04945883602689,
+      "learning_rate": 1.177901033550012e-05,
+      "loss": 0.3231,
+      "step": 575
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.826195057643915,
+      "learning_rate": 1.1753497133707678e-05,
+      "loss": 0.3016,
+      "step": 576
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 8.971990719339775,
+      "learning_rate": 1.1727972150449545e-05,
+      "loss": 0.2818,
+      "step": 577
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6422662404105821,
+      "learning_rate": 1.1702435557223988e-05,
+      "loss": 0.2689,
+      "step": 578
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.762833226202305,
+      "learning_rate": 1.1676887525607272e-05,
+      "loss": 0.2159,
+      "step": 579
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2910649198888533,
+      "learning_rate": 1.1651328227252516e-05,
+      "loss": 0.2244,
+      "step": 580
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.094888043379137,
+      "learning_rate": 1.1625757833888552e-05,
+      "loss": 0.3594,
+      "step": 581
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.2689593600580564,
+      "learning_rate": 1.1600176517318742e-05,
+      "loss": 0.3697,
+      "step": 582
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.036837172227918,
+      "learning_rate": 1.1574584449419841e-05,
+      "loss": 0.3415,
+      "step": 583
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8312182661707044,
+      "learning_rate": 1.1548981802140849e-05,
+      "loss": 0.2814,
+      "step": 584
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7233203263190564,
+      "learning_rate": 1.1523368747501839e-05,
+      "loss": 0.2371,
+      "step": 585
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.354957898292099,
+      "learning_rate": 1.1497745457592817e-05,
+      "loss": 0.2179,
+      "step": 586
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.852344400666275,
+      "learning_rate": 1.1472112104572547e-05,
+      "loss": 0.2883,
+      "step": 587
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.2117818179444004,
+      "learning_rate": 1.1446468860667422e-05,
+      "loss": 0.3362,
+      "step": 588
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6374093138437715,
+      "learning_rate": 1.142081589817027e-05,
+      "loss": 0.229,
+      "step": 589
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.44747695223584,
+      "learning_rate": 1.1395153389439232e-05,
+      "loss": 0.3219,
+      "step": 590
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7629632446504988,
+      "learning_rate": 1.1369481506896582e-05,
+      "loss": 0.2567,
+      "step": 591
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.1288861116559445,
+      "learning_rate": 1.1343800423027583e-05,
+      "loss": 0.3142,
+      "step": 592
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4937719998915224,
+      "learning_rate": 1.1318110310379303e-05,
+      "loss": 0.2684,
+      "step": 593
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.9641163799417516,
+      "learning_rate": 1.129241134155949e-05,
+      "loss": 0.3032,
+      "step": 594
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.320726435802104,
+      "learning_rate": 1.1266703689235395e-05,
+      "loss": 0.3314,
+      "step": 595
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.600756878312125,
+      "learning_rate": 1.1240987526132595e-05,
+      "loss": 0.2048,
+      "step": 596
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5669187735152597,
+      "learning_rate": 1.1215263025033869e-05,
+      "loss": 0.2504,
+      "step": 597
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.673604639988355,
+      "learning_rate": 1.1189530358778005e-05,
+      "loss": 0.267,
+      "step": 598
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6903134828244797,
+      "learning_rate": 1.1163789700258656e-05,
+      "loss": 0.2873,
+      "step": 599
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8345305254478188,
+      "learning_rate": 1.1138041222423177e-05,
+      "loss": 0.2501,
+      "step": 600
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7035373535636231,
+      "learning_rate": 1.111228509827145e-05,
+      "loss": 0.2597,
+      "step": 601
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.0008621254077523,
+      "learning_rate": 1.1086521500854746e-05,
+      "loss": 0.299,
+      "step": 602
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.098817320177265,
+      "learning_rate": 1.1060750603274535e-05,
+      "loss": 0.3879,
+      "step": 603
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.0169678737020864,
+      "learning_rate": 1.1034972578681338e-05,
+      "loss": 0.2962,
+      "step": 604
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7427358350656112,
+      "learning_rate": 1.1009187600273565e-05,
+      "loss": 0.233,
+      "step": 605
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7268783633290732,
+      "learning_rate": 1.0983395841296349e-05,
+      "loss": 0.3154,
+      "step": 606
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.1796515190823067,
+      "learning_rate": 1.0957597475040373e-05,
+      "loss": 0.3123,
+      "step": 607
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.0775099568321305,
+      "learning_rate": 1.0931792674840718e-05,
+      "loss": 0.3513,
+      "step": 608
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9943573933644843,
+      "learning_rate": 1.0905981614075693e-05,
+      "loss": 0.2742,
+      "step": 609
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9559678557187534,
+      "learning_rate": 1.0880164466165675e-05,
+      "loss": 0.261,
+      "step": 610
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5524280949298477,
+      "learning_rate": 1.0854341404571929e-05,
+      "loss": 0.2293,
+      "step": 611
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8032833160238992,
+      "learning_rate": 1.0828512602795462e-05,
+      "loss": 0.2201,
+      "step": 612
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.991664339586316,
+      "learning_rate": 1.0802678234375852e-05,
+      "loss": 0.2977,
+      "step": 613
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.2614790312717767,
+      "learning_rate": 1.0776838472890065e-05,
+      "loss": 0.3178,
+      "step": 614
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.6688609853378997,
+      "learning_rate": 1.075099349195131e-05,
+      "loss": 0.3176,
+      "step": 615
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9523644646192964,
+      "learning_rate": 1.0725143465207868e-05,
+      "loss": 0.3127,
+      "step": 616
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.0367494023975357,
+      "learning_rate": 1.0699288566341914e-05,
+      "loss": 0.2484,
+      "step": 617
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.70860182909142,
+      "learning_rate": 1.0673428969068365e-05,
+      "loss": 0.2455,
+      "step": 618
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.0921836159299128,
+      "learning_rate": 1.06475648471337e-05,
+      "loss": 0.2931,
+      "step": 619
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7894757861594108,
+      "learning_rate": 1.0621696374314807e-05,
+      "loss": 0.2389,
+      "step": 620
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.9138153204173056,
+      "learning_rate": 1.0595823724417795e-05,
+      "loss": 0.2499,
+      "step": 621
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.732159740978979,
+      "learning_rate": 1.0569947071276847e-05,
+      "loss": 0.2528,
+      "step": 622
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.604322234601798,
+      "learning_rate": 1.0544066588753044e-05,
+      "loss": 0.3118,
+      "step": 623
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.857528267039127,
+      "learning_rate": 1.0518182450733185e-05,
+      "loss": 0.3384,
+      "step": 624
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8319294463625793,
+      "learning_rate": 1.0492294831128641e-05,
+      "loss": 0.2126,
+      "step": 625
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7949572184557832,
+      "learning_rate": 1.0466403903874176e-05,
+      "loss": 0.2624,
+      "step": 626
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.3350124712513427,
+      "learning_rate": 1.0440509842926768e-05,
+      "loss": 0.2573,
+      "step": 627
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8899114484260906,
+      "learning_rate": 1.0414612822264457e-05,
+      "loss": 0.2336,
+      "step": 628
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.08203660477542,
+      "learning_rate": 1.0388713015885161e-05,
+      "loss": 0.355,
+      "step": 629
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5453599564780907,
+      "learning_rate": 1.0362810597805526e-05,
+      "loss": 0.2005,
+      "step": 630
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4880436797334258,
+      "learning_rate": 1.0336905742059742e-05,
+      "loss": 0.2083,
+      "step": 631
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.2129712910857706,
+      "learning_rate": 1.031099862269837e-05,
+      "loss": 0.2938,
+      "step": 632
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.007015377715633,
+      "learning_rate": 1.028508941378719e-05,
+      "loss": 0.326,
+      "step": 633
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.430415446584424,
+      "learning_rate": 1.0259178289406011e-05,
+      "loss": 0.3133,
+      "step": 634
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7430708368390706,
+      "learning_rate": 1.0233265423647523e-05,
+      "loss": 0.2283,
+      "step": 635
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.475276647543626,
+      "learning_rate": 1.0207350990616107e-05,
+      "loss": 0.2085,
+      "step": 636
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5906030539698848,
+      "learning_rate": 1.0181435164426676e-05,
+      "loss": 0.2401,
+      "step": 637
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.260933647724671,
+      "learning_rate": 1.0155518119203511e-05,
+      "loss": 0.2861,
+      "step": 638
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8853987068919458,
+      "learning_rate": 1.0129600029079072e-05,
+      "loss": 0.2366,
+      "step": 639
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.0976061677630002,
+      "learning_rate": 1.0103681068192845e-05,
+      "loss": 0.2928,
+      "step": 640
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.2519798986265807,
+      "learning_rate": 1.0077761410690172e-05,
+      "loss": 0.3527,
+      "step": 641
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.9980492224694486,
+      "learning_rate": 1.0051841230721065e-05,
+      "loss": 0.3277,
+      "step": 642
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.1108410240606466,
+      "learning_rate": 1.0025920702439051e-05,
+      "loss": 0.2718,
+      "step": 643
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.499661079196125,
+      "learning_rate": 1e-05,
+      "loss": 0.3087,
+      "step": 644
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.795724082595488,
+      "learning_rate": 9.97407929756095e-06,
+      "loss": 0.3023,
+      "step": 645
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.467066192741996,
+      "learning_rate": 9.948158769278939e-06,
+      "loss": 0.2076,
+      "step": 646
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.050471519058259,
+      "learning_rate": 9.92223858930983e-06,
+      "loss": 0.3034,
+      "step": 647
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.220856023054313,
+      "learning_rate": 9.896318931807155e-06,
+      "loss": 0.3391,
+      "step": 648
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.863571935358034,
+      "learning_rate": 9.870399970920932e-06,
+      "loss": 0.2584,
+      "step": 649
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8386913522932375,
+      "learning_rate": 9.844481880796492e-06,
+      "loss": 0.2296,
+      "step": 650
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.159763851238493,
+      "learning_rate": 9.818564835573324e-06,
+      "loss": 0.3015,
+      "step": 651
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.1871325388107845,
+      "learning_rate": 9.7926490093839e-06,
+      "loss": 0.3556,
+      "step": 652
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.452578997980729,
+      "learning_rate": 9.766734576352478e-06,
+      "loss": 0.3296,
+      "step": 653
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3715404202620423,
+      "learning_rate": 9.740821710593989e-06,
+      "loss": 0.1949,
+      "step": 654
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6850708561729166,
+      "learning_rate": 9.714910586212815e-06,
+      "loss": 0.2519,
+      "step": 655
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.9975915546839027,
+      "learning_rate": 9.689001377301634e-06,
+      "loss": 0.3526,
+      "step": 656
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6032989326955913,
+      "learning_rate": 9.663094257940258e-06,
+      "loss": 0.258,
+      "step": 657
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5020075466048406,
+      "learning_rate": 9.637189402194477e-06,
+      "loss": 0.2157,
+      "step": 658
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5360060738402725,
+      "learning_rate": 9.61128698411484e-06,
+      "loss": 0.2529,
+      "step": 659
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.1523288036462875,
+      "learning_rate": 9.585387177735548e-06,
+      "loss": 0.3479,
+      "step": 660
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.526681384178872,
+      "learning_rate": 9.559490157073236e-06,
+      "loss": 0.2498,
+      "step": 661
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9706114035878841,
+      "learning_rate": 9.533596096125826e-06,
+      "loss": 0.2578,
+      "step": 662
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.0875827900456034,
+      "learning_rate": 9.507705168871359e-06,
+      "loss": 0.2958,
+      "step": 663
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3150336652807166,
+      "learning_rate": 9.481817549266817e-06,
+      "loss": 0.2239,
+      "step": 664
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.755272176537075,
+      "learning_rate": 9.45593341124696e-06,
+      "loss": 0.2453,
+      "step": 665
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2853026912980152,
+      "learning_rate": 9.430052928723153e-06,
+      "loss": 0.1516,
+      "step": 666
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.0816660750021545,
+      "learning_rate": 9.404176275582208e-06,
+      "loss": 0.3357,
+      "step": 667
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9012164667801466,
+      "learning_rate": 9.378303625685196e-06,
+      "loss": 0.2896,
+      "step": 668
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.038751603631483,
+      "learning_rate": 9.352435152866299e-06,
+      "loss": 0.2958,
+      "step": 669
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7628021191667054,
+      "learning_rate": 9.326571030931636e-06,
+      "loss": 0.254,
+      "step": 670
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.140386509095083,
+      "learning_rate": 9.300711433658088e-06,
+      "loss": 0.3249,
+      "step": 671
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.0219557296847683,
+      "learning_rate": 9.274856534792138e-06,
+      "loss": 0.3125,
+      "step": 672
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7783897851664183,
+      "learning_rate": 9.249006508048695e-06,
+      "loss": 0.3127,
+      "step": 673
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.3950371883220853,
+      "learning_rate": 9.223161527109938e-06,
+      "loss": 0.3485,
+      "step": 674
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.685073081651275,
+      "learning_rate": 9.197321765624153e-06,
+      "loss": 0.2779,
+      "step": 675
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.846355813733675,
+      "learning_rate": 9.17148739720454e-06,
+      "loss": 0.292,
+      "step": 676
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.1325807316923835,
+      "learning_rate": 9.145658595428075e-06,
+      "loss": 0.3394,
+      "step": 677
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.1928723467900304,
+      "learning_rate": 9.119835533834332e-06,
+      "loss": 0.3057,
+      "step": 678
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8305417609648569,
+      "learning_rate": 9.09401838592431e-06,
+      "loss": 0.2657,
+      "step": 679
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6794809573883291,
+      "learning_rate": 9.068207325159285e-06,
+      "loss": 0.2233,
+      "step": 680
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9638987984684106,
+      "learning_rate": 9.042402524959632e-06,
+      "loss": 0.284,
+      "step": 681
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.767069934592504,
+      "learning_rate": 9.016604158703654e-06,
+      "loss": 0.2898,
+      "step": 682
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9395833869891466,
+      "learning_rate": 8.990812399726435e-06,
+      "loss": 0.2359,
+      "step": 683
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6200623938126053,
+      "learning_rate": 8.965027421318666e-06,
+      "loss": 0.2537,
+      "step": 684
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9061001116103824,
+      "learning_rate": 8.939249396725468e-06,
+      "loss": 0.249,
+      "step": 685
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7573100057844808,
+      "learning_rate": 8.913478499145255e-06,
+      "loss": 0.2884,
+      "step": 686
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9056792765093784,
+      "learning_rate": 8.887714901728551e-06,
+      "loss": 0.2455,
+      "step": 687
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9591625204073388,
+      "learning_rate": 8.861958777576826e-06,
+      "loss": 0.2489,
+      "step": 688
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.389773070992383,
+      "learning_rate": 8.836210299741346e-06,
+      "loss": 0.1992,
+      "step": 689
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9368736859469793,
+      "learning_rate": 8.810469641222001e-06,
+      "loss": 0.2221,
+      "step": 690
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.0589002820916567,
+      "learning_rate": 8.784736974966135e-06,
+      "loss": 0.2342,
+      "step": 691
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4905403992972885,
+      "learning_rate": 8.759012473867407e-06,
+      "loss": 0.2684,
+      "step": 692
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7473535053835778,
+      "learning_rate": 8.73329631076461e-06,
+      "loss": 0.243,
+      "step": 693
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6813683752097253,
+      "learning_rate": 8.707588658440511e-06,
+      "loss": 0.2492,
+      "step": 694
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.408970858642131,
+      "learning_rate": 8.681889689620699e-06,
+      "loss": 0.2003,
+      "step": 695
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.716823703272264,
+      "learning_rate": 8.656199576972424e-06,
+      "loss": 0.2784,
+      "step": 696
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.411386721721133,
+      "learning_rate": 8.630518493103421e-06,
+      "loss": 0.3147,
+      "step": 697
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.770920978175344,
+      "learning_rate": 8.604846610560771e-06,
+      "loss": 0.279,
+      "step": 698
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.3271735591024725,
+      "learning_rate": 8.579184101829734e-06,
+      "loss": 0.3377,
+      "step": 699
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.3419344175534484,
+      "learning_rate": 8.553531139332583e-06,
+      "loss": 0.3811,
+      "step": 700
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.5626317676383743,
+      "learning_rate": 8.527887895427454e-06,
+      "loss": 0.2962,
+      "step": 701
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9362226850688755,
+      "learning_rate": 8.502254542407186e-06,
+      "loss": 0.3329,
+      "step": 702
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7470223843486916,
+      "learning_rate": 8.476631252498163e-06,
+      "loss": 0.252,
+      "step": 703
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8326252457768055,
+      "learning_rate": 8.451018197859153e-06,
+      "loss": 0.305,
+      "step": 704
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.914344829480951,
+      "learning_rate": 8.425415550580162e-06,
+      "loss": 0.2827,
+      "step": 705
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7401454344432838,
+      "learning_rate": 8.399823482681263e-06,
+      "loss": 0.2237,
+      "step": 706
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.915407039290035,
+      "learning_rate": 8.374242166111448e-06,
+      "loss": 0.2602,
+      "step": 707
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7662030359060474,
+      "learning_rate": 8.348671772747488e-06,
+      "loss": 0.2613,
+      "step": 708
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8378807564270505,
+      "learning_rate": 8.323112474392731e-06,
+      "loss": 0.2295,
+      "step": 709
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.1373704073838935,
+      "learning_rate": 8.297564442776014e-06,
+      "loss": 0.2583,
+      "step": 710
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9134412975123845,
+      "learning_rate": 8.272027849550457e-06,
+      "loss": 0.251,
+      "step": 711
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7321493305498081,
+      "learning_rate": 8.246502866292324e-06,
+      "loss": 0.2462,
+      "step": 712
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7422502864431673,
+      "learning_rate": 8.22098966449988e-06,
+      "loss": 0.3001,
+      "step": 713
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8695974387785772,
+      "learning_rate": 8.195488415592238e-06,
+      "loss": 0.2629,
+      "step": 714
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.0689409685964963,
+      "learning_rate": 8.169999290908189e-06,
+      "loss": 0.2823,
+      "step": 715
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.1939503731878625,
+      "learning_rate": 8.144522461705067e-06,
+      "loss": 0.2832,
+      "step": 716
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7859543663180524,
+      "learning_rate": 8.119058099157605e-06,
+      "loss": 0.2787,
+      "step": 717
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.2440954170050214,
+      "learning_rate": 8.09360637435676e-06,
+      "loss": 0.2857,
+      "step": 718
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8539424786835583,
+      "learning_rate": 8.068167458308582e-06,
+      "loss": 0.2208,
+      "step": 719
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9465753497269007,
+      "learning_rate": 8.042741521933071e-06,
+      "loss": 0.2725,
+      "step": 720
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.0697271722885526,
+      "learning_rate": 8.017328736063005e-06,
+      "loss": 0.2857,
+      "step": 721
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8427390960314034,
+      "learning_rate": 7.991929271442817e-06,
+      "loss": 0.2416,
+      "step": 722
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3575832964819068,
+      "learning_rate": 7.966543298727426e-06,
+      "loss": 0.2328,
+      "step": 723
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.23600118477056,
+      "learning_rate": 7.941170988481108e-06,
+      "loss": 0.3954,
+      "step": 724
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4238788399103068,
+      "learning_rate": 7.915812511176348e-06,
+      "loss": 0.2308,
+      "step": 725
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5635993113710267,
+      "learning_rate": 7.89046803719267e-06,
+      "loss": 0.2139,
+      "step": 726
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8370804480852818,
+      "learning_rate": 7.865137736815536e-06,
+      "loss": 0.2773,
+      "step": 727
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.1761902290910675,
+      "learning_rate": 7.839821780235168e-06,
+      "loss": 0.2539,
+      "step": 728
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.740237960651226,
+      "learning_rate": 7.814520337545405e-06,
+      "loss": 0.2879,
+      "step": 729
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.564863716974201,
+      "learning_rate": 7.789233578742583e-06,
+      "loss": 0.2321,
+      "step": 730
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8466822510532337,
+      "learning_rate": 7.763961673724379e-06,
+      "loss": 0.3398,
+      "step": 731
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9795137872908464,
+      "learning_rate": 7.738704792288654e-06,
+      "loss": 0.2737,
+      "step": 732
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6011860774130708,
+      "learning_rate": 7.713463104132345e-06,
+      "loss": 0.2026,
+      "step": 733
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6552476365616033,
+      "learning_rate": 7.688236778850307e-06,
+      "loss": 0.2569,
+      "step": 734
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.0217820875771344,
+      "learning_rate": 7.663025985934158e-06,
+      "loss": 0.304,
+      "step": 735
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5150619815753843,
+      "learning_rate": 7.637830894771176e-06,
+      "loss": 0.2189,
+      "step": 736
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.186718062247446,
+      "learning_rate": 7.61265167464313e-06,
+      "loss": 0.3316,
+      "step": 737
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.107306318442609,
+      "learning_rate": 7.587488494725157e-06,
+      "loss": 0.2399,
+      "step": 738
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4123410080473793,
+      "learning_rate": 7.5623415240846235e-06,
+      "loss": 0.2493,
+      "step": 739
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.1775732705159556,
+      "learning_rate": 7.537210931679988e-06,
+      "loss": 0.3429,
+      "step": 740
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.0867680464812977,
+      "learning_rate": 7.512096886359663e-06,
+      "loss": 0.2466,
+      "step": 741
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.369684623629669,
+      "learning_rate": 7.48699955686089e-06,
+      "loss": 0.21,
+      "step": 742
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.061299527975879,
+      "learning_rate": 7.4619191118085955e-06,
+      "loss": 0.3019,
+      "step": 743
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.2068568998917577,
+      "learning_rate": 7.4368557197142596e-06,
+      "loss": 0.3616,
+      "step": 744
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.211709484730348,
+      "learning_rate": 7.411809548974792e-06,
+      "loss": 0.2979,
+      "step": 745
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6484394682313261,
+      "learning_rate": 7.3867807678713965e-06,
+      "loss": 0.2686,
+      "step": 746
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.819593888775896,
+      "learning_rate": 7.361769544568424e-06,
+      "loss": 0.2769,
+      "step": 747
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.0270942833528975,
+      "learning_rate": 7.336776047112277e-06,
+      "loss": 0.3073,
+      "step": 748
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8410653613922618,
+      "learning_rate": 7.311800443430251e-06,
+      "loss": 0.2065,
+      "step": 749
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.11937231287671,
+      "learning_rate": 7.286842901329413e-06,
+      "loss": 0.2401,
+      "step": 750
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6752790498990626,
+      "learning_rate": 7.26190358849548e-06,
+      "loss": 0.2336,
+      "step": 751
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.585871964147853,
+      "learning_rate": 7.236982672491699e-06,
+      "loss": 0.2328,
+      "step": 752
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4850708268042807,
+      "learning_rate": 7.212080320757695e-06,
+      "loss": 0.2361,
+      "step": 753
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.810180128194031,
+      "learning_rate": 7.187196700608373e-06,
+      "loss": 0.2118,
+      "step": 754
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9604044922011872,
+      "learning_rate": 7.162331979232784e-06,
+      "loss": 0.2935,
+      "step": 755
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.3782851976916053,
+      "learning_rate": 7.137486323692994e-06,
+      "loss": 0.3103,
+      "step": 756
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.1641128770713647,
+      "learning_rate": 7.1126599009229766e-06,
+      "loss": 0.3441,
+      "step": 757
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8820104796769928,
+      "learning_rate": 7.0878528777274814e-06,
+      "loss": 0.2598,
+      "step": 758
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8562917314614056,
+      "learning_rate": 7.063065420780909e-06,
+      "loss": 0.2472,
+      "step": 759
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.121603337381749,
+      "learning_rate": 7.0382976966262065e-06,
+      "loss": 0.3149,
+      "step": 760
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.3963555305399518,
+      "learning_rate": 7.013549871673736e-06,
+      "loss": 0.2907,
+      "step": 761
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.916425141616387,
+      "learning_rate": 6.988822112200157e-06,
+      "loss": 0.245,
+      "step": 762
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.0267284952286784,
+      "learning_rate": 6.964114584347316e-06,
+      "loss": 0.2846,
+      "step": 763
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.0835299646218517,
+      "learning_rate": 6.939427454121128e-06,
+      "loss": 0.3285,
+      "step": 764
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.110661975800217,
+      "learning_rate": 6.914760887390453e-06,
+      "loss": 0.2585,
+      "step": 765
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8797700943372209,
+      "learning_rate": 6.890115049885995e-06,
+      "loss": 0.2546,
+      "step": 766
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.12025172688243,
+      "learning_rate": 6.865490107199182e-06,
+      "loss": 0.3522,
+      "step": 767
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.4608708843959843,
+      "learning_rate": 6.840886224781039e-06,
+      "loss": 0.4119,
+      "step": 768
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8579288038132629,
+      "learning_rate": 6.816303567941111e-06,
+      "loss": 0.2418,
+      "step": 769
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.7794135410103595,
+      "learning_rate": 6.791742301846325e-06,
+      "loss": 0.3241,
+      "step": 770
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.937994466079078,
+      "learning_rate": 6.767202591519876e-06,
+      "loss": 0.2847,
+      "step": 771
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6099233206079304,
+      "learning_rate": 6.742684601840142e-06,
+      "loss": 0.2654,
+      "step": 772
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7824149383848715,
+      "learning_rate": 6.718188497539554e-06,
+      "loss": 0.2328,
+      "step": 773
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.1762391597184894,
+      "learning_rate": 6.693714443203507e-06,
+      "loss": 0.3638,
+      "step": 774
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7104692317924,
+      "learning_rate": 6.669262603269246e-06,
+      "loss": 0.2263,
+      "step": 775
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7275338948194552,
+      "learning_rate": 6.644833142024752e-06,
+      "loss": 0.2567,
+      "step": 776
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.166136590966985,
+      "learning_rate": 6.620426223607655e-06,
+      "loss": 0.3237,
+      "step": 777
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9817672311971697,
+      "learning_rate": 6.59604201200412e-06,
+      "loss": 0.1942,
+      "step": 778
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.039979494344524,
+      "learning_rate": 6.571680671047749e-06,
+      "loss": 0.244,
+      "step": 779
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9571588104884596,
+      "learning_rate": 6.547342364418482e-06,
+      "loss": 0.2502,
+      "step": 780
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.680652340274426,
+      "learning_rate": 6.523027255641494e-06,
+      "loss": 0.2289,
+      "step": 781
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8417236190828588,
+      "learning_rate": 6.498735508086094e-06,
+      "loss": 0.2579,
+      "step": 782
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7507950484803283,
+      "learning_rate": 6.474467284964634e-06,
+      "loss": 0.2502,
+      "step": 783
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8084082798639443,
+      "learning_rate": 6.450222749331414e-06,
+      "loss": 0.32,
+      "step": 784
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.2526702250171216,
+      "learning_rate": 6.426002064081565e-06,
+      "loss": 0.3481,
+      "step": 785
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.909785646107146,
+      "learning_rate": 6.40180539194999e-06,
+      "loss": 0.2803,
+      "step": 786
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8134151111142809,
+      "learning_rate": 6.377632895510248e-06,
+      "loss": 0.2465,
+      "step": 787
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.7155913087330097,
+      "learning_rate": 6.35348473717345e-06,
+      "loss": 0.4239,
+      "step": 788
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.577301053913633,
+      "learning_rate": 6.329361079187199e-06,
+      "loss": 0.2506,
+      "step": 789
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7753632554801293,
+      "learning_rate": 6.305262083634488e-06,
+      "loss": 0.2695,
+      "step": 790
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.19246127069483,
+      "learning_rate": 6.281187912432587e-06,
+      "loss": 0.3557,
+      "step": 791
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5659295166802625,
+      "learning_rate": 6.2571387273319905e-06,
+      "loss": 0.2695,
+      "step": 792
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8487247045088673,
+      "learning_rate": 6.233114689915316e-06,
+      "loss": 0.2652,
+      "step": 793
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5346195654255266,
+      "learning_rate": 6.209115961596208e-06,
+      "loss": 0.2475,
+      "step": 794
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8728887617878336,
+      "learning_rate": 6.1851427036182696e-06,
+      "loss": 0.2652,
+      "step": 795
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9044082710275902,
+      "learning_rate": 6.1611950770539766e-06,
+      "loss": 0.2661,
+      "step": 796
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.6455901933838546,
+      "learning_rate": 6.137273242803581e-06,
+      "loss": 0.3343,
+      "step": 797
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.0561627754494536,
+      "learning_rate": 6.113377361594048e-06,
+      "loss": 0.3318,
+      "step": 798
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.710547264376754,
+      "learning_rate": 6.0895075939779705e-06,
+      "loss": 0.2586,
+      "step": 799
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.5825166136418742,
+      "learning_rate": 6.065664100332478e-06,
+      "loss": 0.3625,
+      "step": 800
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.590240850136802,
+      "learning_rate": 6.041847040858177e-06,
+      "loss": 0.1974,
+      "step": 801
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4426197798860692,
+      "learning_rate": 6.018056575578075e-06,
+      "loss": 0.2023,
+      "step": 802
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.0249441359081226,
+      "learning_rate": 5.994292864336473e-06,
+      "loss": 0.3006,
+      "step": 803
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9763595272017949,
+      "learning_rate": 5.970556066797941e-06,
+      "loss": 0.2527,
+      "step": 804
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.1274698904290483,
+      "learning_rate": 5.9468463424462146e-06,
+      "loss": 0.2836,
+      "step": 805
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6396568935664884,
+      "learning_rate": 5.923163850583114e-06,
+      "loss": 0.2601,
+      "step": 806
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.4905209493046985,
+      "learning_rate": 5.899508750327502e-06,
+      "loss": 0.3347,
+      "step": 807
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.86505404652595,
+      "learning_rate": 5.875881200614208e-06,
+      "loss": 0.2689,
+      "step": 808
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8960019373545756,
+      "learning_rate": 5.852281360192933e-06,
+      "loss": 0.2615,
+      "step": 809
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.36025005092698,
+      "learning_rate": 5.828709387627219e-06,
+      "loss": 0.2768,
+      "step": 810
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3701135734425034,
+      "learning_rate": 5.80516544129337e-06,
+      "loss": 0.1988,
+      "step": 811
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7516812972892462,
+      "learning_rate": 5.781649679379379e-06,
+      "loss": 0.2604,
+      "step": 812
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4459045171656397,
+      "learning_rate": 5.758162259883867e-06,
+      "loss": 0.1918,
+      "step": 813
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7513968129817934,
+      "learning_rate": 5.7347033406150494e-06,
+      "loss": 0.2546,
+      "step": 814
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7415168967269428,
+      "learning_rate": 5.711273079189621e-06,
+      "loss": 0.2248,
+      "step": 815
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5123709515298638,
+      "learning_rate": 5.687871633031754e-06,
+      "loss": 0.2031,
+      "step": 816
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.598182895373857,
+      "learning_rate": 5.664499159372017e-06,
+      "loss": 0.3691,
+      "step": 817
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4193538794405987,
+      "learning_rate": 5.64115581524629e-06,
+      "loss": 0.2177,
+      "step": 818
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.905161913821604,
+      "learning_rate": 5.617841757494762e-06,
+      "loss": 0.2835,
+      "step": 819
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.102083052982321,
+      "learning_rate": 5.594557142760853e-06,
+      "loss": 0.3016,
+      "step": 820
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.2761940983796167,
+      "learning_rate": 5.571302127490133e-06,
+      "loss": 0.3509,
+      "step": 821
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.277744764697184,
+      "learning_rate": 5.548076867929331e-06,
+      "loss": 0.2858,
+      "step": 822
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7070140403796947,
+      "learning_rate": 5.524881520125229e-06,
+      "loss": 0.2232,
+      "step": 823
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.915971628592982,
+      "learning_rate": 5.501716239923642e-06,
+      "loss": 0.2862,
+      "step": 824
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9129074042445322,
+      "learning_rate": 5.4785811829683764e-06,
+      "loss": 0.2869,
+      "step": 825
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6475461446423851,
+      "learning_rate": 5.455476504700161e-06,
+      "loss": 0.2089,
+      "step": 826
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.1519296552582547,
+      "learning_rate": 5.432402360355616e-06,
+      "loss": 0.2914,
+      "step": 827
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.219671875855405,
+      "learning_rate": 5.4093589049662175e-06,
+      "loss": 0.3412,
+      "step": 828
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.714197759692349,
+      "learning_rate": 5.386346293357242e-06,
+      "loss": 0.2313,
+      "step": 829
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.5027628517881846,
+      "learning_rate": 5.3633646801467255e-06,
+      "loss": 0.3819,
+      "step": 830
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.813344312078274,
+      "learning_rate": 5.340414219744451e-06,
+      "loss": 0.2589,
+      "step": 831
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.2453999187971676,
+      "learning_rate": 5.31749506635086e-06,
+      "loss": 0.2981,
+      "step": 832
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8015076932483307,
+      "learning_rate": 5.294607373956071e-06,
+      "loss": 0.2588,
+      "step": 833
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.1720611042347624,
+      "learning_rate": 5.271751296338823e-06,
+      "loss": 0.2953,
+      "step": 834
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.822196770831624,
+      "learning_rate": 5.248926987065417e-06,
+      "loss": 0.2931,
+      "step": 835
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.0288907542049603,
+      "learning_rate": 5.226134599488728e-06,
+      "loss": 0.3015,
+      "step": 836
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.606122975712571,
+      "learning_rate": 5.2033742867471586e-06,
+      "loss": 0.2496,
+      "step": 837
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.926511830402839,
+      "learning_rate": 5.1806462017635775e-06,
+      "loss": 0.2775,
+      "step": 838
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7046572764822447,
+      "learning_rate": 5.15795049724435e-06,
+      "loss": 0.2097,
+      "step": 839
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.139068743994829,
+      "learning_rate": 5.135287325678271e-06,
+      "loss": 0.2945,
+      "step": 840
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.3382499483296915,
+      "learning_rate": 5.112656839335544e-06,
+      "loss": 0.2717,
+      "step": 841
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.5489341490977715,
+      "learning_rate": 5.090059190266779e-06,
+      "loss": 0.2297,
+      "step": 842
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.929739578300716,
+      "learning_rate": 5.067494530301953e-06,
+      "loss": 0.2937,
+      "step": 843
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.688606273121372,
+      "learning_rate": 5.044963011049384e-06,
+      "loss": 0.2313,
+      "step": 844
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.345953555234023,
+      "learning_rate": 5.022464783894743e-06,
+      "loss": 0.372,
+      "step": 845
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8505091067979924,
+      "learning_rate": 5.000000000000003e-06,
+      "loss": 0.281,
+      "step": 846
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.506976721199778,
+      "learning_rate": 4.977568810302432e-06,
+      "loss": 0.232,
+      "step": 847
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6407950451029856,
+      "learning_rate": 4.955171365513603e-06,
+      "loss": 0.2228,
+      "step": 848
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.599770395848796,
+      "learning_rate": 4.932807816118347e-06,
+      "loss": 0.2348,
+      "step": 849
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.1228077377531895,
+      "learning_rate": 4.910478312373757e-06,
+      "loss": 0.2572,
+      "step": 850
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7090102227337063,
+      "learning_rate": 4.88818300430819e-06,
+      "loss": 0.2325,
+      "step": 851
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8950875251067498,
+      "learning_rate": 4.865922041720239e-06,
+      "loss": 0.2548,
+      "step": 852
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5009240917340398,
+      "learning_rate": 4.843695574177737e-06,
+      "loss": 0.1798,
+      "step": 853
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5171623126881457,
+      "learning_rate": 4.821503751016746e-06,
+      "loss": 0.2368,
+      "step": 854
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.2217784851424653,
+      "learning_rate": 4.799346721340571e-06,
+      "loss": 0.2635,
+      "step": 855
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6726727263291783,
+      "learning_rate": 4.777224634018732e-06,
+      "loss": 0.2305,
+      "step": 856
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5970766816679292,
+      "learning_rate": 4.7551376376859794e-06,
+      "loss": 0.2486,
+      "step": 857
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8282756770609114,
+      "learning_rate": 4.733085880741301e-06,
+      "loss": 0.2258,
+      "step": 858
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8021282161394163,
+      "learning_rate": 4.711069511346909e-06,
+      "loss": 0.2533,
+      "step": 859
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8622337232706356,
+      "learning_rate": 4.689088677427249e-06,
+      "loss": 0.2628,
+      "step": 860
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.5580543061415226,
+      "learning_rate": 4.667143526668022e-06,
+      "loss": 0.3121,
+      "step": 861
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6231857953982278,
+      "learning_rate": 4.645234206515171e-06,
+      "loss": 0.2643,
+      "step": 862
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5262446570330437,
+      "learning_rate": 4.623360864173893e-06,
+      "loss": 0.2512,
+      "step": 863
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.522437154126145,
+      "learning_rate": 4.601523646607675e-06,
+      "loss": 0.2477,
+      "step": 864
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5879991031824943,
+      "learning_rate": 4.579722700537268e-06,
+      "loss": 0.2674,
+      "step": 865
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.013364625188082,
+      "learning_rate": 4.557958172439726e-06,
+      "loss": 0.4444,
+      "step": 866
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.799358920343576,
+      "learning_rate": 4.536230208547425e-06,
+      "loss": 0.2368,
+      "step": 867
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.956106849110676,
+      "learning_rate": 4.5145389548470645e-06,
+      "loss": 0.2445,
+      "step": 868
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5453317819490386,
+      "learning_rate": 4.492884557078688e-06,
+      "loss": 0.2464,
+      "step": 869
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.477291903477527,
+      "learning_rate": 4.471267160734731e-06,
+      "loss": 0.2515,
+      "step": 870
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.810092610902386,
+      "learning_rate": 4.449686911058992e-06,
+      "loss": 0.2465,
+      "step": 871
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.3340906296515653,
+      "learning_rate": 4.4281439530457174e-06,
+      "loss": 0.2296,
+      "step": 872
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9130306552063339,
+      "learning_rate": 4.4066384314385755e-06,
+      "loss": 0.2599,
+      "step": 873
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.776612076322738,
+      "learning_rate": 4.385170490729712e-06,
+      "loss": 0.3193,
+      "step": 874
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9804752262674616,
+      "learning_rate": 4.36374027515878e-06,
+      "loss": 0.2713,
+      "step": 875
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6897331235488673,
+      "learning_rate": 4.342347928711953e-06,
+      "loss": 0.2245,
+      "step": 876
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.029522379259432,
+      "learning_rate": 4.320993595120969e-06,
+      "loss": 0.2631,
+      "step": 877
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.1769099014874875,
+      "learning_rate": 4.299677417862174e-06,
+      "loss": 0.3171,
+      "step": 878
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.690743115404439,
+      "learning_rate": 4.278399540155536e-06,
+      "loss": 0.2477,
+      "step": 879
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.488705706366445,
+      "learning_rate": 4.257160104963695e-06,
+      "loss": 0.3136,
+      "step": 880
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6774752784282572,
+      "learning_rate": 4.2359592549910145e-06,
+      "loss": 0.1692,
+      "step": 881
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9676187400989076,
+      "learning_rate": 4.214797132682597e-06,
+      "loss": 0.2493,
+      "step": 882
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6946929404614504,
+      "learning_rate": 4.193673880223339e-06,
+      "loss": 0.2747,
+      "step": 883
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.935665546193616,
+      "learning_rate": 4.172589639536992e-06,
+      "loss": 0.2747,
+      "step": 884
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.938611898607862,
+      "learning_rate": 4.151544552285178e-06,
+      "loss": 0.3066,
+      "step": 885
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9430796788762474,
+      "learning_rate": 4.130538759866457e-06,
+      "loss": 0.2913,
+      "step": 886
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6893016033444104,
+      "learning_rate": 4.109572403415386e-06,
+      "loss": 0.23,
+      "step": 887
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6669863700924643,
+      "learning_rate": 4.088645623801534e-06,
+      "loss": 0.219,
+      "step": 888
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.177237248640568,
+      "learning_rate": 4.067758561628577e-06,
+      "loss": 0.2593,
+      "step": 889
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.445017111944672,
+      "learning_rate": 4.046911357233343e-06,
+      "loss": 0.3743,
+      "step": 890
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.0263857774688745,
+      "learning_rate": 4.026104150684835e-06,
+      "loss": 0.2894,
+      "step": 891
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.214076834154425,
+      "learning_rate": 4.00533708178334e-06,
+      "loss": 0.3055,
+      "step": 892
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8491534643694523,
+      "learning_rate": 3.984610290059467e-06,
+      "loss": 0.2483,
+      "step": 893
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.0864996537735236,
+      "learning_rate": 3.9639239147731865e-06,
+      "loss": 0.2977,
+      "step": 894
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.518977429747083,
+      "learning_rate": 3.943278094912946e-06,
+      "loss": 0.3909,
+      "step": 895
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8430207575686766,
+      "learning_rate": 3.9226729691946865e-06,
+      "loss": 0.2309,
+      "step": 896
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5685678811930743,
+      "learning_rate": 3.902108676060937e-06,
+      "loss": 0.2341,
+      "step": 897
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.1181407558758147,
+      "learning_rate": 3.881585353679891e-06,
+      "loss": 0.2572,
+      "step": 898
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9186450278547378,
+      "learning_rate": 3.861103139944448e-06,
+      "loss": 0.265,
+      "step": 899
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.646266791160965,
+      "learning_rate": 3.840662172471315e-06,
+      "loss": 0.2546,
+      "step": 900
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6907626028232738,
+      "learning_rate": 3.8202625886000745e-06,
+      "loss": 0.2462,
+      "step": 901
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.079071161563802,
+      "learning_rate": 3.799904525392251e-06,
+      "loss": 0.231,
+      "step": 902
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6050203728815282,
+      "learning_rate": 3.7795881196303996e-06,
+      "loss": 0.2116,
+      "step": 903
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.675470240404623,
+      "learning_rate": 3.759313507817196e-06,
+      "loss": 0.2679,
+      "step": 904
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.850984938367741,
+      "learning_rate": 3.739080826174498e-06,
+      "loss": 0.3196,
+      "step": 905
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9981294449239055,
+      "learning_rate": 3.718890210642442e-06,
+      "loss": 0.2924,
+      "step": 906
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9295081629858126,
+      "learning_rate": 3.6987417968785365e-06,
+      "loss": 0.2104,
+      "step": 907
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8068361121579488,
+      "learning_rate": 3.6786357202567367e-06,
+      "loss": 0.3137,
+      "step": 908
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7334238116847467,
+      "learning_rate": 3.658572115866541e-06,
+      "loss": 0.2643,
+      "step": 909
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5092334537286647,
+      "learning_rate": 3.638551118512089e-06,
+      "loss": 0.2166,
+      "step": 910
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8963530539025175,
+      "learning_rate": 3.618572862711247e-06,
+      "loss": 0.2888,
+      "step": 911
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2997655874538818,
+      "learning_rate": 3.5986374826947067e-06,
+      "loss": 0.2343,
+      "step": 912
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3597778204686617,
+      "learning_rate": 3.5787451124050832e-06,
+      "loss": 0.199,
+      "step": 913
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.685966678875132,
+      "learning_rate": 3.558895885496023e-06,
+      "loss": 0.2205,
+      "step": 914
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.403178326462928,
+      "learning_rate": 3.5390899353312934e-06,
+      "loss": 0.2058,
+      "step": 915
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.649215250464992,
+      "learning_rate": 3.519327394983888e-06,
+      "loss": 0.254,
+      "step": 916
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.784007943129409,
+      "learning_rate": 3.4996083972351514e-06,
+      "loss": 0.2338,
+      "step": 917
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5425709234643024,
+      "learning_rate": 3.479933074573858e-06,
+      "loss": 0.2127,
+      "step": 918
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.072846212502085,
+      "learning_rate": 3.4603015591953393e-06,
+      "loss": 0.2506,
+      "step": 919
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9617017545169801,
+      "learning_rate": 3.440713983000601e-06,
+      "loss": 0.2623,
+      "step": 920
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.1949608945418095,
+      "learning_rate": 3.421170477595419e-06,
+      "loss": 0.287,
+      "step": 921
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.0223352734966333,
+      "learning_rate": 3.401671174289469e-06,
+      "loss": 0.2437,
+      "step": 922
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.726939680121309,
+      "learning_rate": 3.3822162040954355e-06,
+      "loss": 0.2669,
+      "step": 923
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3958468201182181,
+      "learning_rate": 3.3628056977281456e-06,
+      "loss": 0.1795,
+      "step": 924
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5349920574626108,
+      "learning_rate": 3.3434397856036705e-06,
+      "loss": 0.2277,
+      "step": 925
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6265746387819953,
+      "learning_rate": 3.3241185978384636e-06,
+      "loss": 0.212,
+      "step": 926
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.990986006877363,
+      "learning_rate": 3.304842264248489e-06,
+      "loss": 0.3214,
+      "step": 927
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9233533246133991,
+      "learning_rate": 3.2856109143483316e-06,
+      "loss": 0.275,
+      "step": 928
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5915126926393386,
+      "learning_rate": 3.266424677350346e-06,
+      "loss": 0.2297,
+      "step": 929
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6176077751978817,
+      "learning_rate": 3.2472836821637744e-06,
+      "loss": 0.2224,
+      "step": 930
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5393085991541886,
+      "learning_rate": 3.228188057393895e-06,
+      "loss": 0.2248,
+      "step": 931
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.060362274625342,
+      "learning_rate": 3.209137931341143e-06,
+      "loss": 0.2691,
+      "step": 932
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8804183328512907,
+      "learning_rate": 3.190133432000252e-06,
+      "loss": 0.2816,
+      "step": 933
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7918052113764629,
+      "learning_rate": 3.1711746870594083e-06,
+      "loss": 0.2553,
+      "step": 934
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.0199052401985345,
+      "learning_rate": 3.1522618238993728e-06,
+      "loss": 0.2894,
+      "step": 935
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.262932683368157,
+      "learning_rate": 3.1333949695926323e-06,
+      "loss": 0.3589,
+      "step": 936
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8111099043452628,
+      "learning_rate": 3.114574250902558e-06,
+      "loss": 0.2961,
+      "step": 937
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.180703372509301,
+      "learning_rate": 3.0957997942825337e-06,
+      "loss": 0.2751,
+      "step": 938
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.1348008623457293,
+      "learning_rate": 3.077071725875116e-06,
+      "loss": 0.3002,
+      "step": 939
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5856791522786344,
+      "learning_rate": 3.0583901715111965e-06,
+      "loss": 0.2568,
+      "step": 940
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6940017773069984,
+      "learning_rate": 3.039755256709134e-06,
+      "loss": 0.2467,
+      "step": 941
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.974696925499119,
+      "learning_rate": 3.021167106673928e-06,
+      "loss": 0.2767,
+      "step": 942
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.503892274988601,
+      "learning_rate": 3.0026258462963787e-06,
+      "loss": 0.1722,
+      "step": 943
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.153554440122579,
+      "learning_rate": 2.9841316001522345e-06,
+      "loss": 0.3149,
+      "step": 944
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.75095568108896,
+      "learning_rate": 2.9656844925013638e-06,
+      "loss": 0.2601,
+      "step": 945
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8607532467725405,
+      "learning_rate": 2.94728464728693e-06,
+      "loss": 0.3038,
+      "step": 946
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.626934063795533,
+      "learning_rate": 2.9289321881345257e-06,
+      "loss": 0.1833,
+      "step": 947
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9274778528150571,
+      "learning_rate": 2.910627238351383e-06,
+      "loss": 0.2845,
+      "step": 948
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.3559754751169395,
+      "learning_rate": 2.8923699209255285e-06,
+      "loss": 0.2273,
+      "step": 949
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6582819252008927,
+      "learning_rate": 2.8741603585249312e-06,
+      "loss": 0.2841,
+      "step": 950
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.654622795384472,
+      "learning_rate": 2.855998673496728e-06,
+      "loss": 0.2547,
+      "step": 951
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.168483542872676,
+      "learning_rate": 2.837884987866363e-06,
+      "loss": 0.2858,
+      "step": 952
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6177763948492443,
+      "learning_rate": 2.8198194233367747e-06,
+      "loss": 0.2487,
+      "step": 953
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8494390147186466,
+      "learning_rate": 2.8018021012875994e-06,
+      "loss": 0.271,
+      "step": 954
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6944744681663106,
+      "learning_rate": 2.783833142774328e-06,
+      "loss": 0.2145,
+      "step": 955
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.356672636203177,
+      "learning_rate": 2.7659126685275028e-06,
+      "loss": 0.242,
+      "step": 956
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7071797812952967,
+      "learning_rate": 2.74804079895192e-06,
+      "loss": 0.1947,
+      "step": 957
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.680836923707193,
+      "learning_rate": 2.7302176541257984e-06,
+      "loss": 0.2331,
+      "step": 958
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.668963787322095,
+      "learning_rate": 2.7124433537999838e-06,
+      "loss": 0.2216,
+      "step": 959
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4706495302727334,
+      "learning_rate": 2.694718017397151e-06,
+      "loss": 0.2001,
+      "step": 960
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6545100367424903,
+      "learning_rate": 2.677041764010988e-06,
+      "loss": 0.2329,
+      "step": 961
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.950623456067504,
+      "learning_rate": 2.6594147124053983e-06,
+      "loss": 0.2655,
+      "step": 962
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.3036306131593465,
+      "learning_rate": 2.641836981013719e-06,
+      "loss": 0.3089,
+      "step": 963
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.4138904957865215,
+      "learning_rate": 2.6243086879379e-06,
+      "loss": 0.4055,
+      "step": 964
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.946390349319703,
+      "learning_rate": 2.6068299509477267e-06,
+      "loss": 0.3118,
+      "step": 965
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7931647327520022,
+      "learning_rate": 2.5894008874800323e-06,
+      "loss": 0.246,
+      "step": 966
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8355050420939845,
+      "learning_rate": 2.572021614637892e-06,
+      "loss": 0.2431,
+      "step": 967
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5108304857763393,
+      "learning_rate": 2.5546922491898497e-06,
+      "loss": 0.1944,
+      "step": 968
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.346782967322365,
+      "learning_rate": 2.537412907569127e-06,
+      "loss": 0.2079,
+      "step": 969
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4907687498692637,
+      "learning_rate": 2.5201837058728506e-06,
+      "loss": 0.2222,
+      "step": 970
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.006852991655727,
+      "learning_rate": 2.5030047598612585e-06,
+      "loss": 0.2125,
+      "step": 971
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.8979712050124378,
+      "learning_rate": 2.485876184956928e-06,
+      "loss": 0.4267,
+      "step": 972
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6213096691389526,
+      "learning_rate": 2.468798096244007e-06,
+      "loss": 0.2109,
+      "step": 973
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.9873316318148522,
+      "learning_rate": 2.451770608467432e-06,
+      "loss": 0.2406,
+      "step": 974
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.186626149431908,
+      "learning_rate": 2.4347938360321564e-06,
+      "loss": 0.3419,
+      "step": 975
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7968422881940356,
+      "learning_rate": 2.417867893002387e-06,
+      "loss": 0.2259,
+      "step": 976
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8541653929442405,
+      "learning_rate": 2.400992893100822e-06,
+      "loss": 0.2574,
+      "step": 977
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.0761176343433854,
+      "learning_rate": 2.3841689497078746e-06,
+      "loss": 0.3144,
+      "step": 978
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.1785093636632977,
+      "learning_rate": 2.3673961758609156e-06,
+      "loss": 0.2764,
+      "step": 979
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.9317088428685394,
+      "learning_rate": 2.3506746842535244e-06,
+      "loss": 0.5288,
+      "step": 980
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7937271040727196,
+      "learning_rate": 2.3340045872347173e-06,
+      "loss": 0.2357,
+      "step": 981
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.838531651124889,
+      "learning_rate": 2.317385996808195e-06,
+      "loss": 0.256,
+      "step": 982
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.283906968927328,
+      "learning_rate": 2.3008190246316033e-06,
+      "loss": 0.2639,
+      "step": 983
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.158590692135789,
+      "learning_rate": 2.2843037820157678e-06,
+      "loss": 0.3043,
+      "step": 984
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6092953541482968,
+      "learning_rate": 2.26784037992395e-06,
+      "loss": 0.2322,
+      "step": 985
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.97563915984023,
+      "learning_rate": 2.251428928971102e-06,
+      "loss": 0.2471,
+      "step": 986
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.629050153917207,
+      "learning_rate": 2.2350695394231346e-06,
+      "loss": 0.2425,
+      "step": 987
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6882685576225636,
+      "learning_rate": 2.218762321196156e-06,
+      "loss": 0.2276,
+      "step": 988
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6805409089233097,
+      "learning_rate": 2.2025073838557454e-06,
+      "loss": 0.2604,
+      "step": 989
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8115589091967659,
+      "learning_rate": 2.186304836616221e-06,
+      "loss": 0.2522,
+      "step": 990
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.9373751118106743,
+      "learning_rate": 2.170154788339892e-06,
+      "loss": 0.2744,
+      "step": 991
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7419361886427593,
+      "learning_rate": 2.1540573475363402e-06,
+      "loss": 0.2447,
+      "step": 992
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.9284641554213888,
+      "learning_rate": 2.1380126223616894e-06,
+      "loss": 0.2173,
+      "step": 993
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.1195771363082847,
+      "learning_rate": 2.122020720617869e-06,
+      "loss": 0.3005,
+      "step": 994
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.770085685562092,
+      "learning_rate": 2.106081749751897e-06,
+      "loss": 0.2074,
+      "step": 995
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.499879939600023,
+      "learning_rate": 2.090195816855164e-06,
+      "loss": 0.167,
+      "step": 996
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.307606423009887,
+      "learning_rate": 2.0743630286627005e-06,
+      "loss": 0.3208,
+      "step": 997
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6977856293372309,
+      "learning_rate": 2.058583491552465e-06,
+      "loss": 0.2204,
+      "step": 998
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5998489674345802,
+      "learning_rate": 2.0428573115446394e-06,
+      "loss": 0.2267,
+      "step": 999
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.1114488918832883,
+      "learning_rate": 2.0271845943008984e-06,
+      "loss": 0.3243,
+      "step": 1000
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.1802349543638213,
+      "learning_rate": 2.011565445123711e-06,
+      "loss": 0.2929,
+      "step": 1001
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.2640749367528925,
+      "learning_rate": 1.9959999689556407e-06,
+      "loss": 0.3503,
+      "step": 1002
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.905369014558016,
+      "learning_rate": 1.9804882703786122e-06,
+      "loss": 0.273,
+      "step": 1003
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.9385398442468922,
+      "learning_rate": 1.9650304536132426e-06,
+      "loss": 0.2166,
+      "step": 1004
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4130939284561759,
+      "learning_rate": 1.9496266225181247e-06,
+      "loss": 0.2331,
+      "step": 1005
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.554232419563932,
+      "learning_rate": 1.9342768805891176e-06,
+      "loss": 0.1955,
+      "step": 1006
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7286117804200054,
+      "learning_rate": 1.918981330958678e-06,
+      "loss": 0.2455,
+      "step": 1007
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.344247788752057,
+      "learning_rate": 1.9037400763951508e-06,
+      "loss": 0.2516,
+      "step": 1008
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.116552663192002,
+      "learning_rate": 1.8885532193020706e-06,
+      "loss": 0.2739,
+      "step": 1009
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5784074402197446,
+      "learning_rate": 1.8734208617174986e-06,
+      "loss": 0.2572,
+      "step": 1010
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6104349407900382,
+      "learning_rate": 1.8583431053133127e-06,
+      "loss": 0.2411,
+      "step": 1011
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6986377833707487,
+      "learning_rate": 1.8433200513945338e-06,
+      "loss": 0.2578,
+      "step": 1012
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5459871249460182,
+      "learning_rate": 1.8283518008986566e-06,
+      "loss": 0.1919,
+      "step": 1013
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4656179228758885,
+      "learning_rate": 1.813438454394948e-06,
+      "loss": 0.2173,
+      "step": 1014
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.2051276084515803,
+      "learning_rate": 1.7985801120837865e-06,
+      "loss": 0.2693,
+      "step": 1015
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.9929337623116723,
+      "learning_rate": 1.7837768737959937e-06,
+      "loss": 0.2439,
+      "step": 1016
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.3310422723848814,
+      "learning_rate": 1.7690288389921495e-06,
+      "loss": 0.3406,
+      "step": 1017
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5929192652600608,
+      "learning_rate": 1.7543361067619269e-06,
+      "loss": 0.1961,
+      "step": 1018
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.253150783294225,
+      "learning_rate": 1.7396987758234418e-06,
+      "loss": 0.2971,
+      "step": 1019
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.9551151379061291,
+      "learning_rate": 1.7251169445225658e-06,
+      "loss": 0.2702,
+      "step": 1020
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8443350566644119,
+      "learning_rate": 1.7105907108322816e-06,
+      "loss": 0.2168,
+      "step": 1021
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8953709979345807,
+      "learning_rate": 1.6961201723520248e-06,
+      "loss": 0.2209,
+      "step": 1022
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6933804163691488,
+      "learning_rate": 1.6817054263070176e-06,
+      "loss": 0.2651,
+      "step": 1023
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.9741371681122017,
+      "learning_rate": 1.6673465695476233e-06,
+      "loss": 0.2624,
+      "step": 1024
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7937690474707826,
+      "learning_rate": 1.6530436985486997e-06,
+      "loss": 0.227,
+      "step": 1025
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.0398781094509846,
+      "learning_rate": 1.6387969094089318e-06,
+      "loss": 0.2504,
+      "step": 1026
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.0034581357121066,
+      "learning_rate": 1.6246062978502165e-06,
+      "loss": 0.2697,
+      "step": 1027
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3490783836586475,
+      "learning_rate": 1.6104719592169905e-06,
+      "loss": 0.1582,
+      "step": 1028
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.553829042342749,
+      "learning_rate": 1.5963939884756042e-06,
+      "loss": 0.1961,
+      "step": 1029
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5629315635602483,
+      "learning_rate": 1.5823724802136863e-06,
+      "loss": 0.214,
+      "step": 1030
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5501021255130174,
+      "learning_rate": 1.5684075286394983e-06,
+      "loss": 0.1857,
+      "step": 1031
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6851671227244354,
+      "learning_rate": 1.5544992275813053e-06,
+      "loss": 0.1912,
+      "step": 1032
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.0503928029573855,
+      "learning_rate": 1.5406476704867524e-06,
+      "loss": 0.3187,
+      "step": 1033
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.611805408069118,
+      "learning_rate": 1.5268529504222262e-06,
+      "loss": 0.2549,
+      "step": 1034
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.0144593850719943,
+      "learning_rate": 1.5131151600722338e-06,
+      "loss": 0.301,
+      "step": 1035
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7665966680162037,
+      "learning_rate": 1.4994343917387854e-06,
+      "loss": 0.2406,
+      "step": 1036
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5618392868502702,
+      "learning_rate": 1.485810737340767e-06,
+      "loss": 0.2677,
+      "step": 1037
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.9288584503490473,
+      "learning_rate": 1.4722442884133214e-06,
+      "loss": 0.2304,
+      "step": 1038
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5617186763180901,
+      "learning_rate": 1.4587351361072455e-06,
+      "loss": 0.2567,
+      "step": 1039
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.770241801620364,
+      "learning_rate": 1.4452833711883629e-06,
+      "loss": 0.2577,
+      "step": 1040
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.9382896906192402,
+      "learning_rate": 1.4318890840369181e-06,
+      "loss": 0.2542,
+      "step": 1041
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6049626081328412,
+      "learning_rate": 1.4185523646469822e-06,
+      "loss": 0.2083,
+      "step": 1042
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6914009615957128,
+      "learning_rate": 1.405273302625828e-06,
+      "loss": 0.2374,
+      "step": 1043
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8576240619220135,
+      "learning_rate": 1.3920519871933425e-06,
+      "loss": 0.235,
+      "step": 1044
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.9143500706855923,
+      "learning_rate": 1.3788885071814173e-06,
+      "loss": 0.3512,
+      "step": 1045
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.885115331951092,
+      "learning_rate": 1.3657829510333653e-06,
+      "loss": 0.2132,
+      "step": 1046
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.698087441819351,
+      "learning_rate": 1.3527354068033139e-06,
+      "loss": 0.2516,
+      "step": 1047
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6111456973292821,
+      "learning_rate": 1.339745962155613e-06,
+      "loss": 0.2265,
+      "step": 1048
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4189744222385203,
+      "learning_rate": 1.326814704364262e-06,
+      "loss": 0.2044,
+      "step": 1049
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8257357511262042,
+      "learning_rate": 1.313941720312303e-06,
+      "loss": 0.2578,
+      "step": 1050
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5300273413158851,
+      "learning_rate": 1.3011270964912458e-06,
+      "loss": 0.2133,
+      "step": 1051
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.9035816080597627,
+      "learning_rate": 1.2883709190004956e-06,
+      "loss": 0.2663,
+      "step": 1052
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4919891884156362,
+      "learning_rate": 1.2756732735467581e-06,
+      "loss": 0.2153,
+      "step": 1053
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.489188386807827,
+      "learning_rate": 1.263034245443473e-06,
+      "loss": 0.2072,
+      "step": 1054
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.0790616040249184,
+      "learning_rate": 1.2504539196102438e-06,
+      "loss": 0.3164,
+      "step": 1055
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7799690128110763,
+      "learning_rate": 1.2379323805722575e-06,
+      "loss": 0.267,
+      "step": 1056
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.1522746470134906,
+      "learning_rate": 1.2254697124597237e-06,
+      "loss": 0.2844,
+      "step": 1057
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4489560689229093,
+      "learning_rate": 1.2130659990073146e-06,
+      "loss": 0.1939,
+      "step": 1058
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.139743701541494,
+      "learning_rate": 1.2007213235535785e-06,
+      "loss": 0.4079,
+      "step": 1059
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9988394205499171,
+      "learning_rate": 1.1884357690404157e-06,
+      "loss": 0.2531,
+      "step": 1060
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4279144078503483,
+      "learning_rate": 1.176209418012495e-06,
+      "loss": 0.162,
+      "step": 1061
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7124786659484386,
+      "learning_rate": 1.1640423526166987e-06,
+      "loss": 0.225,
+      "step": 1062
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7979914687264473,
+      "learning_rate": 1.1519346546015908e-06,
+      "loss": 0.249,
+      "step": 1063
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7353155680409826,
+      "learning_rate": 1.1398864053168534e-06,
+      "loss": 0.2624,
+      "step": 1064
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8080169204151373,
+      "learning_rate": 1.127897685712731e-06,
+      "loss": 0.2759,
+      "step": 1065
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.0284757482671933,
+      "learning_rate": 1.1159685763395113e-06,
+      "loss": 0.2559,
+      "step": 1066
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5144844908790105,
+      "learning_rate": 1.1040991573469629e-06,
+      "loss": 0.2071,
+      "step": 1067
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.2401085581760296,
+      "learning_rate": 1.0922895084838036e-06,
+      "loss": 0.2199,
+      "step": 1068
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.629115436288061,
+      "learning_rate": 1.0805397090971738e-06,
+      "loss": 0.2697,
+      "step": 1069
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.076570153773264,
+      "learning_rate": 1.0688498381320855e-06,
+      "loss": 0.2594,
+      "step": 1070
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8580443460941836,
+      "learning_rate": 1.057219974130903e-06,
+      "loss": 0.2128,
+      "step": 1071
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.1981221023851834,
+      "learning_rate": 1.0456501952328191e-06,
+      "loss": 0.1426,
+      "step": 1072
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.607708557965466,
+      "learning_rate": 1.0341405791733183e-06,
+      "loss": 0.2416,
+      "step": 1073
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.981390470792631,
+      "learning_rate": 1.022691203283661e-06,
+      "loss": 0.2385,
+      "step": 1074
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7331067592776395,
+      "learning_rate": 1.0113021444903725e-06,
+      "loss": 0.2727,
+      "step": 1075
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6434505773641759,
+      "learning_rate": 9.999734793146998e-07,
+      "loss": 0.246,
+      "step": 1076
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6784574564006565,
+      "learning_rate": 9.887052838721322e-07,
+      "loss": 0.2591,
+      "step": 1077
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8787331378064278,
+      "learning_rate": 9.77497633871868e-07,
+      "loss": 0.2378,
+      "step": 1078
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6362732770702062,
+      "learning_rate": 9.663506046162986e-07,
+      "loss": 0.2236,
+      "step": 1079
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4273077402383598,
+      "learning_rate": 9.5526427100053e-07,
+      "loss": 0.1615,
+      "step": 1080
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7635521575059474,
+      "learning_rate": 9.44238707511862e-07,
+      "loss": 0.2497,
+      "step": 1081
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.4497383712551444,
+      "learning_rate": 9.332739882292752e-07,
+      "loss": 0.3468,
+      "step": 1082
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8689717300129733,
+      "learning_rate": 9.22370186822965e-07,
+      "loss": 0.2115,
+      "step": 1083
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.1692904506738553,
+      "learning_rate": 9.115273765538202e-07,
+      "loss": 0.2912,
+      "step": 1084
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8477047973625857,
+      "learning_rate": 9.0074563027294e-07,
+      "loss": 0.254,
+      "step": 1085
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6084097443123828,
+      "learning_rate": 8.900250204211513e-07,
+      "loss": 0.2346,
+      "step": 1086
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9758442942520758,
+      "learning_rate": 8.793656190285071e-07,
+      "loss": 0.2914,
+      "step": 1087
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5800324714803446,
+      "learning_rate": 8.687674977138116e-07,
+      "loss": 0.2053,
+      "step": 1088
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.903023910592132,
+      "learning_rate": 8.582307276841461e-07,
+      "loss": 0.2486,
+      "step": 1089
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8721484797697183,
+      "learning_rate": 8.477553797343729e-07,
+      "loss": 0.2525,
+      "step": 1090
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.210408987867441,
+      "learning_rate": 8.373415242466721e-07,
+      "loss": 0.3476,
+      "step": 1091
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8145618381264705,
+      "learning_rate": 8.269892311900696e-07,
+      "loss": 0.2253,
+      "step": 1092
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.503548842568088,
+      "learning_rate": 8.166985701199581e-07,
+      "loss": 0.2078,
+      "step": 1093
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.970993383274634,
+      "learning_rate": 8.06469610177636e-07,
+      "loss": 0.2661,
+      "step": 1094
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.2541842404634007,
+      "learning_rate": 7.963024200898462e-07,
+      "loss": 0.2943,
+      "step": 1095
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.0793321004842507,
+      "learning_rate": 7.861970681683051e-07,
+      "loss": 0.248,
+      "step": 1096
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4323451359398107,
+      "learning_rate": 7.761536223092459e-07,
+      "loss": 0.219,
+      "step": 1097
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.223009593916491,
+      "learning_rate": 7.661721499929753e-07,
+      "loss": 0.3207,
+      "step": 1098
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.15631076631837,
+      "learning_rate": 7.562527182833978e-07,
+      "loss": 0.2821,
+      "step": 1099
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.83217141842868,
+      "learning_rate": 7.463953938275859e-07,
+      "loss": 0.2463,
+      "step": 1100
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7945591900376303,
+      "learning_rate": 7.366002428553154e-07,
+      "loss": 0.2689,
+      "step": 1101
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6401381233328671,
+      "learning_rate": 7.268673311786378e-07,
+      "loss": 0.2225,
+      "step": 1102
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.345183288068179,
+      "learning_rate": 7.171967241914224e-07,
+      "loss": 0.2579,
+      "step": 1103
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.154992194024568,
+      "learning_rate": 7.07588486868922e-07,
+      "loss": 0.3283,
+      "step": 1104
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2305934639872602,
+      "learning_rate": 6.980426837673437e-07,
+      "loss": 0.1764,
+      "step": 1105
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.150422986343797,
+      "learning_rate": 6.885593790234057e-07,
+      "loss": 0.3109,
+      "step": 1106
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.9487027632840348,
+      "learning_rate": 6.791386363539065e-07,
+      "loss": 0.2511,
+      "step": 1107
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3624494939574974,
+      "learning_rate": 6.697805190553086e-07,
+      "loss": 0.2182,
+      "step": 1108
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.9629653911358407,
+      "learning_rate": 6.604850900032956e-07,
+      "loss": 0.2803,
+      "step": 1109
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6686070386269276,
+      "learning_rate": 6.512524116523633e-07,
+      "loss": 0.1942,
+      "step": 1110
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8184381048891454,
+      "learning_rate": 6.420825460353975e-07,
+      "loss": 0.234,
+      "step": 1111
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8292676950113214,
+      "learning_rate": 6.329755547632499e-07,
+      "loss": 0.2588,
+      "step": 1112
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.14785054063878,
+      "learning_rate": 6.239314990243339e-07,
+      "loss": 0.2758,
+      "step": 1113
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.445041782478947,
+      "learning_rate": 6.149504395842087e-07,
+      "loss": 0.1945,
+      "step": 1114
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.458479886365711,
+      "learning_rate": 6.0603243678517e-07,
+      "loss": 0.3026,
+      "step": 1115
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.11544638899036,
+      "learning_rate": 5.971775505458444e-07,
+      "loss": 0.2975,
+      "step": 1116
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5675593571621884,
+      "learning_rate": 5.883858403607967e-07,
+      "loss": 0.2438,
+      "step": 1117
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.2184826624622627,
+      "learning_rate": 5.796573653001091e-07,
+      "loss": 0.2709,
+      "step": 1118
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8643405635039205,
+      "learning_rate": 5.709921840090072e-07,
+      "loss": 0.2948,
+      "step": 1119
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5694826821845747,
+      "learning_rate": 5.62390354707455e-07,
+      "loss": 0.2275,
+      "step": 1120
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5156052671076892,
+      "learning_rate": 5.538519351897575e-07,
+      "loss": 0.2244,
+      "step": 1121
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.0819892972861216,
+      "learning_rate": 5.453769828241872e-07,
+      "loss": 0.3315,
+      "step": 1122
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.5892607193569708,
+      "learning_rate": 5.369655545525909e-07,
+      "loss": 0.3366,
+      "step": 1123
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.0297523191929456,
+      "learning_rate": 5.286177068899989e-07,
+      "loss": 0.2691,
+      "step": 1124
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5852040501856501,
+      "learning_rate": 5.203334959242634e-07,
+      "loss": 0.2311,
+      "step": 1125
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.1572968911595205,
+      "learning_rate": 5.121129773156663e-07,
+      "loss": 0.2527,
+      "step": 1126
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.0453850393520736,
+      "learning_rate": 5.039562062965508e-07,
+      "loss": 0.2716,
+      "step": 1127
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8774391499144594,
+      "learning_rate": 4.95863237670956e-07,
+      "loss": 0.2746,
+      "step": 1128
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.0000300303285536,
+      "learning_rate": 4.878341258142349e-07,
+      "loss": 0.2602,
+      "step": 1129
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8976121097618184,
+      "learning_rate": 4.798689246727006e-07,
+      "loss": 0.2403,
+      "step": 1130
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.436462843251761,
+      "learning_rate": 4.7196768776326397e-07,
+      "loss": 0.2295,
+      "step": 1131
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.724066891804642,
+      "learning_rate": 4.6413046817306404e-07,
+      "loss": 0.1926,
+      "step": 1132
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.9793371188350892,
+      "learning_rate": 4.563573185591219e-07,
+      "loss": 0.2746,
+      "step": 1133
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5767007072650288,
+      "learning_rate": 4.4864829114798394e-07,
+      "loss": 0.2291,
+      "step": 1134
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.664299581980807,
+      "learning_rate": 4.4100343773536226e-07,
+      "loss": 0.2361,
+      "step": 1135
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.0924593265935583,
+      "learning_rate": 4.3342280968580287e-07,
+      "loss": 0.2756,
+      "step": 1136
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6395170627049216,
+      "learning_rate": 4.259064579323302e-07,
+      "loss": 0.2264,
+      "step": 1137
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5970799593500762,
+      "learning_rate": 4.184544329761009e-07,
+      "loss": 0.229,
+      "step": 1138
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.182004710086821,
+      "learning_rate": 4.11066784886075e-07,
+      "loss": 0.3343,
+      "step": 1139
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7698497542021756,
+      "learning_rate": 4.037435632986786e-07,
+      "loss": 0.2768,
+      "step": 1140
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.819895206197229,
+      "learning_rate": 3.96484817417454e-07,
+      "loss": 0.2596,
+      "step": 1141
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.0223853119407713,
+      "learning_rate": 3.8929059601275463e-07,
+      "loss": 0.2485,
+      "step": 1142
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.022319275762578,
+      "learning_rate": 3.8216094742139833e-07,
+      "loss": 0.275,
+      "step": 1143
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.0916811134300453,
+      "learning_rate": 3.750959195463466e-07,
+      "loss": 0.2983,
+      "step": 1144
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.751594212716712,
+      "learning_rate": 3.6809555985639065e-07,
+      "loss": 0.208,
+      "step": 1145
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.385537278039897,
+      "learning_rate": 3.611599153858214e-07,
+      "loss": 0.2972,
+      "step": 1146
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.2409653452191423,
+      "learning_rate": 3.5428903273411865e-07,
+      "loss": 0.3121,
+      "step": 1147
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.0139033846091645,
+      "learning_rate": 3.474829580656436e-07,
+      "loss": 0.3174,
+      "step": 1148
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.91488746376917,
+      "learning_rate": 3.4074173710931804e-07,
+      "loss": 0.2736,
+      "step": 1149
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.673409190170197,
+      "learning_rate": 3.3406541515832e-07,
+      "loss": 0.2287,
+      "step": 1150
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7600273523357843,
+      "learning_rate": 3.2745403706978876e-07,
+      "loss": 0.2166,
+      "step": 1151
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5479515229778715,
+      "learning_rate": 3.209076472645112e-07,
+      "loss": 0.2092,
+      "step": 1152
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8353681175432979,
+      "learning_rate": 3.1442628972662703e-07,
+      "loss": 0.3097,
+      "step": 1153
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7186461066312657,
+      "learning_rate": 3.080100080033388e-07,
+      "loss": 0.2302,
+      "step": 1154
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.490131470295163,
+      "learning_rate": 3.016588452046132e-07,
+      "loss": 0.2988,
+      "step": 1155
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8254395188861843,
+      "learning_rate": 2.9537284400289354e-07,
+      "loss": 0.2099,
+      "step": 1156
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.9317831257279952,
+      "learning_rate": 2.8915204663281014e-07,
+      "loss": 0.2755,
+      "step": 1157
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.9044530009379896,
+      "learning_rate": 2.829964948909048e-07,
+      "loss": 0.2599,
+      "step": 1158
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.028039260610861,
+      "learning_rate": 2.769062301353398e-07,
+      "loss": 0.2548,
+      "step": 1159
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6524032088131855,
+      "learning_rate": 2.708812932856253e-07,
+      "loss": 0.1937,
+      "step": 1160
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.573824843820764,
+      "learning_rate": 2.649217248223468e-07,
+      "loss": 0.2193,
+      "step": 1161
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.091481605696416,
+      "learning_rate": 2.5902756478688674e-07,
+      "loss": 0.3083,
+      "step": 1162
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.489498923190435,
+      "learning_rate": 2.5319885278115907e-07,
+      "loss": 0.1469,
+      "step": 1163
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.9782590951388326,
+      "learning_rate": 2.474356279673462e-07,
+      "loss": 0.2463,
+      "step": 1164
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4647787821345215,
+      "learning_rate": 2.4173792906762806e-07,
+      "loss": 0.2227,
+      "step": 1165
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.034144523410484,
+      "learning_rate": 2.3610579436392999e-07,
+      "loss": 0.2374,
+      "step": 1166
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7225315166453739,
+      "learning_rate": 2.3053926169765984e-07,
+      "loss": 0.218,
+      "step": 1167
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3404018893048908,
+      "learning_rate": 2.2503836846945792e-07,
+      "loss": 0.198,
+      "step": 1168
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.3954420631646602,
+      "learning_rate": 2.1960315163894075e-07,
+      "loss": 0.2733,
+      "step": 1169
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.934892509654841,
+      "learning_rate": 2.1423364772445886e-07,
+      "loss": 0.2712,
+      "step": 1170
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.690523390032749,
+      "learning_rate": 2.0892989280284825e-07,
+      "loss": 0.2585,
+      "step": 1171
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.2024579481976647,
+      "learning_rate": 2.036919225091827e-07,
+      "loss": 0.2626,
+      "step": 1172
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8870857206700493,
+      "learning_rate": 1.9851977203654839e-07,
+      "loss": 0.1944,
+      "step": 1173
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.781749146894803,
+      "learning_rate": 1.9341347613579086e-07,
+      "loss": 0.2071,
+      "step": 1174
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.948549718871209,
+      "learning_rate": 1.8837306911529185e-07,
+      "loss": 0.295,
+      "step": 1175
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6870970988575948,
+      "learning_rate": 1.8339858484073935e-07,
+      "loss": 0.226,
+      "step": 1176
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9419992404111,
+      "learning_rate": 1.784900567348913e-07,
+      "loss": 0.2746,
+      "step": 1177
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.1023113751631537,
+      "learning_rate": 1.7364751777736334e-07,
+      "loss": 0.3058,
+      "step": 1178
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.476859237886749,
+      "learning_rate": 1.6887100050439587e-07,
+      "loss": 0.2365,
+      "step": 1179
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9830005254933283,
+      "learning_rate": 1.6416053700863965e-07,
+      "loss": 0.2764,
+      "step": 1180
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7476064815606922,
+      "learning_rate": 1.595161589389449e-07,
+      "loss": 0.2578,
+      "step": 1181
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2919070393253038,
+      "learning_rate": 1.5493789750014032e-07,
+      "loss": 0.1889,
+      "step": 1182
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.248108868009186,
+      "learning_rate": 1.504257834528311e-07,
+      "loss": 0.2576,
+      "step": 1183
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.770475360150387,
+      "learning_rate": 1.459798471131868e-07,
+      "loss": 0.2445,
+      "step": 1184
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.93257359669872,
+      "learning_rate": 1.4160011835273936e-07,
+      "loss": 0.2886,
+      "step": 1185
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.0547275163131533,
+      "learning_rate": 1.3728662659818205e-07,
+      "loss": 0.3056,
+      "step": 1186
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.05653058843975,
+      "learning_rate": 1.3303940083117527e-07,
+      "loss": 0.2307,
+      "step": 1187
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8564552698989694,
+      "learning_rate": 1.2885846958814673e-07,
+      "loss": 0.2447,
+      "step": 1188
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4862371640366514,
+      "learning_rate": 1.2474386096010037e-07,
+      "loss": 0.1962,
+      "step": 1189
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3335426432602813,
+      "learning_rate": 1.206956025924333e-07,
+      "loss": 0.2149,
+      "step": 1190
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9383374912058284,
+      "learning_rate": 1.1671372168474137e-07,
+      "loss": 0.2511,
+      "step": 1191
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.0159139844573124,
+      "learning_rate": 1.1279824499064396e-07,
+      "loss": 0.2569,
+      "step": 1192
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8027140386415839,
+      "learning_rate": 1.0894919881760168e-07,
+      "loss": 0.2752,
+      "step": 1193
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7497164453237868,
+      "learning_rate": 1.0516660902673448e-07,
+      "loss": 0.2237,
+      "step": 1194
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7524906394461068,
+      "learning_rate": 1.014505010326583e-07,
+      "loss": 0.2426,
+      "step": 1195
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.9015157578913895,
+      "learning_rate": 9.780089980330643e-08,
+      "loss": 0.2367,
+      "step": 1196
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8091956868397134,
+      "learning_rate": 9.42178298597607e-08,
+      "loss": 0.2108,
+      "step": 1197
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7079179710567602,
+      "learning_rate": 9.070131527609604e-08,
+      "loss": 0.2549,
+      "step": 1198
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.0012786450489655,
+      "learning_rate": 8.725137967920739e-08,
+      "loss": 0.2362,
+      "step": 1199
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6799446307225845,
+      "learning_rate": 8.386804624865851e-08,
+      "loss": 0.2287,
+      "step": 1200
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2707413035131545,
+      "learning_rate": 8.055133771652346e-08,
+      "loss": 0.1707,
+      "step": 1201
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.240844331643583,
+      "learning_rate": 7.730127636723539e-08,
+      "loss": 0.2938,
+      "step": 1202
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8077151139509218,
+      "learning_rate": 7.411788403743236e-08,
+      "loss": 0.2064,
+      "step": 1203
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.463428596541932,
+      "learning_rate": 7.100118211581852e-08,
+      "loss": 0.216,
+      "step": 1204
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8579701300807716,
+      "learning_rate": 6.795119154301199e-08,
+      "loss": 0.2622,
+      "step": 1205
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.889983109137829,
+      "learning_rate": 6.496793281141056e-08,
+      "loss": 0.2207,
+      "step": 1206
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9249117499497341,
+      "learning_rate": 6.205142596505177e-08,
+      "loss": 0.2789,
+      "step": 1207
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.0731734610871966,
+      "learning_rate": 5.920169059947412e-08,
+      "loss": 0.2718,
+      "step": 1208
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.476765494755492,
+      "learning_rate": 5.6418745861593905e-08,
+      "loss": 0.4248,
+      "step": 1209
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.1868528878565128,
+      "learning_rate": 5.37026104495697e-08,
+      "loss": 0.3682,
+      "step": 1210
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.60588326536286,
+      "learning_rate": 5.105330261267916e-08,
+      "loss": 0.2397,
+      "step": 1211
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.988214848749438,
+      "learning_rate": 4.8470840151195745e-08,
+      "loss": 0.2641,
+      "step": 1212
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.427518099981952,
+      "learning_rate": 4.595524041627109e-08,
+      "loss": 0.2897,
+      "step": 1213
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4469407635891465,
+      "learning_rate": 4.350652030981395e-08,
+      "loss": 0.2708,
+      "step": 1214
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.61869636519982,
+      "learning_rate": 4.1124696284383644e-08,
+      "loss": 0.1713,
+      "step": 1215
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7304087932509222,
+      "learning_rate": 3.8809784343072364e-08,
+      "loss": 0.2796,
+      "step": 1216
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.3509790467777107,
+      "learning_rate": 3.6561800039403016e-08,
+      "loss": 0.2884,
+      "step": 1217
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6569368307990178,
+      "learning_rate": 3.438075847721933e-08,
+      "loss": 0.2357,
+      "step": 1218
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.377016569549657,
+      "learning_rate": 3.2266674310589276e-08,
+      "loss": 0.214,
+      "step": 1219
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8801587671172033,
+      "learning_rate": 3.0219561743707326e-08,
+      "loss": 0.206,
+      "step": 1220
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.222157784273952,
+      "learning_rate": 2.8239434530792364e-08,
+      "loss": 0.2952,
+      "step": 1221
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8976048919400084,
+      "learning_rate": 2.6326305976001054e-08,
+      "loss": 0.3109,
+      "step": 1222
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7465668776876373,
+      "learning_rate": 2.4480188933336812e-08,
+      "loss": 0.2183,
+      "step": 1223
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8655547145953026,
+      "learning_rate": 2.2701095806565432e-08,
+      "loss": 0.2624,
+      "step": 1224
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.9202134401828403,
+      "learning_rate": 2.0989038549125152e-08,
+      "loss": 0.3012,
+      "step": 1225
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8141355211746353,
+      "learning_rate": 1.9344028664056715e-08,
+      "loss": 0.271,
+      "step": 1226
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.149284486477867,
+      "learning_rate": 1.7766077203915655e-08,
+      "loss": 0.3302,
+      "step": 1227
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.3524206969808823,
+      "learning_rate": 1.6255194770704586e-08,
+      "loss": 0.3171,
+      "step": 1228
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.012305463045916,
+      "learning_rate": 1.4811391515799911e-08,
+      "loss": 0.2089,
+      "step": 1229
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.017582806511981,
+      "learning_rate": 1.3434677139885222e-08,
+      "loss": 0.2611,
+      "step": 1230
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.613515590334642,
+      "learning_rate": 1.2125060892881345e-08,
+      "loss": 0.239,
+      "step": 1231
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.454180215027866,
+      "learning_rate": 1.0882551573891953e-08,
+      "loss": 0.173,
+      "step": 1232
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.606757287726453,
+      "learning_rate": 9.707157531134714e-09,
+      "loss": 0.2215,
+      "step": 1233
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.538723066462182,
+      "learning_rate": 8.59888666189579e-09,
+      "loss": 0.2393,
+      "step": 1234
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8384817094233377,
+      "learning_rate": 7.557746412468758e-09,
+      "loss": 0.227,
+      "step": 1235
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.966983790686683,
+      "learning_rate": 6.583743778106888e-09,
+      "loss": 0.2717,
+      "step": 1236
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.0175004257834193,
+      "learning_rate": 5.676885302978719e-09,
+      "loss": 0.2976,
+      "step": 1237
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.105840626114623,
+      "learning_rate": 4.837177080119215e-09,
+      "loss": 0.3113,
+      "step": 1238
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.2993857734684404,
+      "learning_rate": 4.064624751394242e-09,
+      "loss": 0.2948,
+      "step": 1239
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6503345660550335,
+      "learning_rate": 3.3592335074594805e-09,
+      "loss": 0.2222,
+      "step": 1240
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2733823560068471,
+      "learning_rate": 2.7210080877237978e-09,
+      "loss": 0.1937,
+      "step": 1241
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5863288882920181,
+      "learning_rate": 2.149952780321485e-09,
+      "loss": 0.1769,
+      "step": 1242
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6579688873141334,
+      "learning_rate": 1.6460714220833952e-09,
+      "loss": 0.2163,
+      "step": 1243
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.805966529584832,
+      "learning_rate": 1.209367398504746e-09,
+      "loss": 0.2301,
+      "step": 1244
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.142574731152733,
+      "learning_rate": 8.39843643731797e-10,
+      "loss": 0.366,
+      "step": 1245
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.00762337546855,
+      "learning_rate": 5.375026405352035e-10,
+      "loss": 0.3333,
+      "step": 1246
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.0681991372071358,
+      "learning_rate": 3.023464202944748e-10,
+      "loss": 0.2887,
+      "step": 1247
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.3712925913849663,
+      "learning_rate": 1.3437656298687096e-10,
+      "loss": 0.2392,
+      "step": 1248
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.8748392210135587,
+      "learning_rate": 3.3594197175190743e-11,
+      "loss": 0.2562,
+      "step": 1249
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.8181806222269268,
+      "learning_rate": 0.0,
+      "loss": 0.301,
+      "step": 1250
+    },
+    {
+      "epoch": 1.0,
+      "step": 1250,
+      "total_flos": 0.0,
+      "train_loss": 0.2979340986609459,
+      "train_runtime": 15445.7821,
+      "train_samples_per_second": 5.179,
+      "train_steps_per_second": 0.081
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 1250,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 300,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": false,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 0.0,
+  "train_batch_size": 4,
+  "trial_name": null,
+  "trial_params": null
+}