diff --git "a/l2-7b-eu/checkpoint-3300/trainer_state.json" "b/l2-7b-eu/checkpoint-3300/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/l2-7b-eu/checkpoint-3300/trainer_state.json"
@@ -0,0 +1,11591 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.8929779461507239,
+  "eval_steps": 500,
+  "global_step": 3300,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.00027059937762143147,
+      "grad_norm": 4.086390018463135,
+      "learning_rate": 0.0,
+      "loss": 3.2754,
+      "step": 1
+    },
+    {
+      "epoch": 0.0005411987552428629,
+      "grad_norm": 3.758815288543701,
+      "learning_rate": 9.017132551848513e-08,
+      "loss": 3.2863,
+      "step": 2
+    },
+    {
+      "epoch": 0.0010823975104857259,
+      "grad_norm": 3.8250608444213867,
+      "learning_rate": 2.705139765554554e-07,
+      "loss": 3.3425,
+      "step": 4
+    },
+    {
+      "epoch": 0.0016235962657285888,
+      "grad_norm": 3.8092095851898193,
+      "learning_rate": 4.5085662759242564e-07,
+      "loss": 3.3165,
+      "step": 6
+    },
+    {
+      "epoch": 0.0021647950209714517,
+      "grad_norm": 3.7621052265167236,
+      "learning_rate": 6.311992786293959e-07,
+      "loss": 3.3295,
+      "step": 8
+    },
+    {
+      "epoch": 0.002705993776214315,
+      "grad_norm": 3.4136276245117188,
+      "learning_rate": 8.115419296663661e-07,
+      "loss": 3.3073,
+      "step": 10
+    },
+    {
+      "epoch": 0.0032471925314571776,
+      "grad_norm": 2.855100393295288,
+      "learning_rate": 9.918845807033363e-07,
+      "loss": 3.3031,
+      "step": 12
+    },
+    {
+      "epoch": 0.0037883912867000408,
+      "grad_norm": 2.491767406463623,
+      "learning_rate": 1.1722272317403068e-06,
+      "loss": 3.2943,
+      "step": 14
+    },
+    {
+      "epoch": 0.0043295900419429035,
+      "grad_norm": 2.359778642654419,
+      "learning_rate": 1.3525698827772768e-06,
+      "loss": 3.2622,
+      "step": 16
+    },
+    {
+      "epoch": 0.004870788797185766,
+      "grad_norm": 2.037504196166992,
+      "learning_rate": 1.5329125338142473e-06,
+      "loss": 3.239,
+      "step": 18
+    },
+    {
+      "epoch": 0.00541198755242863,
+      "grad_norm": 2.8542497158050537,
+      "learning_rate": 1.7132551848512173e-06,
+      "loss": 3.2031,
+      "step": 20
+    },
+    {
+      "epoch": 0.0059531863076714925,
+      "grad_norm": 2.297046661376953,
+      "learning_rate": 1.8935978358881876e-06,
+      "loss": 3.1721,
+      "step": 22
+    },
+    {
+      "epoch": 0.006494385062914355,
+      "grad_norm": 2.2149112224578857,
+      "learning_rate": 2.0739404869251576e-06,
+      "loss": 3.121,
+      "step": 24
+    },
+    {
+      "epoch": 0.007035583818157218,
+      "grad_norm": 1.8048591613769531,
+      "learning_rate": 2.254283137962128e-06,
+      "loss": 3.0857,
+      "step": 26
+    },
+    {
+      "epoch": 0.0075767825734000815,
+      "grad_norm": 1.7466434240341187,
+      "learning_rate": 2.4346257889990986e-06,
+      "loss": 3.0489,
+      "step": 28
+    },
+    {
+      "epoch": 0.008117981328642944,
+      "grad_norm": 2.1722524166107178,
+      "learning_rate": 2.6149684400360686e-06,
+      "loss": 3.0016,
+      "step": 30
+    },
+    {
+      "epoch": 0.008659180083885807,
+      "grad_norm": 1.364578366279602,
+      "learning_rate": 2.7953110910730386e-06,
+      "loss": 2.9587,
+      "step": 32
+    },
+    {
+      "epoch": 0.00920037883912867,
+      "grad_norm": 1.5823427438735962,
+      "learning_rate": 2.9756537421100095e-06,
+      "loss": 2.931,
+      "step": 34
+    },
+    {
+      "epoch": 0.009741577594371532,
+      "grad_norm": 1.2367908954620361,
+      "learning_rate": 3.1559963931469796e-06,
+      "loss": 2.8953,
+      "step": 36
+    },
+    {
+      "epoch": 0.010282776349614395,
+      "grad_norm": 1.0437366962432861,
+      "learning_rate": 3.3363390441839496e-06,
+      "loss": 2.8412,
+      "step": 38
+    },
+    {
+      "epoch": 0.01082397510485726,
+      "grad_norm": 1.081803798675537,
+      "learning_rate": 3.5166816952209197e-06,
+      "loss": 2.7832,
+      "step": 40
+    },
+    {
+      "epoch": 0.011365173860100122,
+      "grad_norm": 0.9715840220451355,
+      "learning_rate": 3.69702434625789e-06,
+      "loss": 2.7729,
+      "step": 42
+    },
+    {
+      "epoch": 0.011906372615342985,
+      "grad_norm": 0.8603936433792114,
+      "learning_rate": 3.877366997294861e-06,
+      "loss": 2.6904,
+      "step": 44
+    },
+    {
+      "epoch": 0.012447571370585848,
+      "grad_norm": 0.8236231803894043,
+      "learning_rate": 4.057709648331831e-06,
+      "loss": 2.6908,
+      "step": 46
+    },
+    {
+      "epoch": 0.01298877012582871,
+      "grad_norm": 0.7681186199188232,
+      "learning_rate": 4.2380522993688015e-06,
+      "loss": 2.6212,
+      "step": 48
+    },
+    {
+      "epoch": 0.013529968881071573,
+      "grad_norm": 0.8002827167510986,
+      "learning_rate": 4.4183949504057716e-06,
+      "loss": 2.6035,
+      "step": 50
+    },
+    {
+      "epoch": 0.014071167636314436,
+      "grad_norm": 0.6757120490074158,
+      "learning_rate": 4.598737601442742e-06,
+      "loss": 2.595,
+      "step": 52
+    },
+    {
+      "epoch": 0.014612366391557299,
+      "grad_norm": 0.6619369387626648,
+      "learning_rate": 4.779080252479712e-06,
+      "loss": 2.5522,
+      "step": 54
+    },
+    {
+      "epoch": 0.015153565146800163,
+      "grad_norm": 0.6247105598449707,
+      "learning_rate": 4.959422903516682e-06,
+      "loss": 2.5079,
+      "step": 56
+    },
+    {
+      "epoch": 0.015694763902043024,
+      "grad_norm": 0.6559263467788696,
+      "learning_rate": 5.139765554553652e-06,
+      "loss": 2.5009,
+      "step": 58
+    },
+    {
+      "epoch": 0.01623596265728589,
+      "grad_norm": 0.6590877175331116,
+      "learning_rate": 5.320108205590623e-06,
+      "loss": 2.4648,
+      "step": 60
+    },
+    {
+      "epoch": 0.01677716141252875,
+      "grad_norm": 0.6045516133308411,
+      "learning_rate": 5.500450856627593e-06,
+      "loss": 2.421,
+      "step": 62
+    },
+    {
+      "epoch": 0.017318360167771614,
+      "grad_norm": 0.6533932089805603,
+      "learning_rate": 5.680793507664563e-06,
+      "loss": 2.3966,
+      "step": 64
+    },
+    {
+      "epoch": 0.01785955892301448,
+      "grad_norm": 0.6478094458580017,
+      "learning_rate": 5.861136158701533e-06,
+      "loss": 2.3903,
+      "step": 66
+    },
+    {
+      "epoch": 0.01840075767825734,
+      "grad_norm": 0.7349300980567932,
+      "learning_rate": 6.041478809738504e-06,
+      "loss": 2.3552,
+      "step": 68
+    },
+    {
+      "epoch": 0.018941956433500204,
+      "grad_norm": 0.6454821825027466,
+      "learning_rate": 6.221821460775474e-06,
+      "loss": 2.3262,
+      "step": 70
+    },
+    {
+      "epoch": 0.019483155188743065,
+      "grad_norm": 0.7321672439575195,
+      "learning_rate": 6.402164111812444e-06,
+      "loss": 2.3197,
+      "step": 72
+    },
+    {
+      "epoch": 0.02002435394398593,
+      "grad_norm": 0.7664237022399902,
+      "learning_rate": 6.582506762849414e-06,
+      "loss": 2.2992,
+      "step": 74
+    },
+    {
+      "epoch": 0.02056555269922879,
+      "grad_norm": 0.6843811869621277,
+      "learning_rate": 6.762849413886384e-06,
+      "loss": 2.2927,
+      "step": 76
+    },
+    {
+      "epoch": 0.021106751454471655,
+      "grad_norm": 0.7199612259864807,
+      "learning_rate": 6.9431920649233556e-06,
+      "loss": 2.2525,
+      "step": 78
+    },
+    {
+      "epoch": 0.02164795020971452,
+      "grad_norm": 0.778446614742279,
+      "learning_rate": 7.123534715960326e-06,
+      "loss": 2.2267,
+      "step": 80
+    },
+    {
+      "epoch": 0.02218914896495738,
+      "grad_norm": 0.9287930727005005,
+      "learning_rate": 7.303877366997296e-06,
+      "loss": 2.2206,
+      "step": 82
+    },
+    {
+      "epoch": 0.022730347720200245,
+      "grad_norm": 1.033782958984375,
+      "learning_rate": 7.484220018034266e-06,
+      "loss": 2.2063,
+      "step": 84
+    },
+    {
+      "epoch": 0.023271546475443106,
+      "grad_norm": 1.0132615566253662,
+      "learning_rate": 7.664562669071236e-06,
+      "loss": 2.1677,
+      "step": 86
+    },
+    {
+      "epoch": 0.02381274523068597,
+      "grad_norm": 0.9043529033660889,
+      "learning_rate": 7.844905320108207e-06,
+      "loss": 2.1696,
+      "step": 88
+    },
+    {
+      "epoch": 0.02435394398592883,
+      "grad_norm": 0.6718290448188782,
+      "learning_rate": 8.025247971145176e-06,
+      "loss": 2.1492,
+      "step": 90
+    },
+    {
+      "epoch": 0.024895142741171696,
+      "grad_norm": 0.9615944027900696,
+      "learning_rate": 8.205590622182147e-06,
+      "loss": 2.1452,
+      "step": 92
+    },
+    {
+      "epoch": 0.02543634149641456,
+      "grad_norm": 0.9435996413230896,
+      "learning_rate": 8.385933273219116e-06,
+      "loss": 2.1098,
+      "step": 94
+    },
+    {
+      "epoch": 0.02597754025165742,
+      "grad_norm": 0.7614261507987976,
+      "learning_rate": 8.566275924256087e-06,
+      "loss": 2.1286,
+      "step": 96
+    },
+    {
+      "epoch": 0.026518739006900285,
+      "grad_norm": 0.9416339993476868,
+      "learning_rate": 8.746618575293058e-06,
+      "loss": 2.1092,
+      "step": 98
+    },
+    {
+      "epoch": 0.027059937762143146,
+      "grad_norm": 0.9229443073272705,
+      "learning_rate": 8.926961226330027e-06,
+      "loss": 2.0932,
+      "step": 100
+    },
+    {
+      "epoch": 0.02760113651738601,
+      "grad_norm": 0.7135593295097351,
+      "learning_rate": 9.107303877366998e-06,
+      "loss": 2.0699,
+      "step": 102
+    },
+    {
+      "epoch": 0.028142335272628872,
+      "grad_norm": 1.0263723134994507,
+      "learning_rate": 9.287646528403967e-06,
+      "loss": 2.0445,
+      "step": 104
+    },
+    {
+      "epoch": 0.028683534027871736,
+      "grad_norm": 1.0300300121307373,
+      "learning_rate": 9.467989179440938e-06,
+      "loss": 2.0463,
+      "step": 106
+    },
+    {
+      "epoch": 0.029224732783114597,
+      "grad_norm": 0.8331286311149597,
+      "learning_rate": 9.648331830477909e-06,
+      "loss": 2.0381,
+      "step": 108
+    },
+    {
+      "epoch": 0.02976593153835746,
+      "grad_norm": 0.7501435875892639,
+      "learning_rate": 9.828674481514878e-06,
+      "loss": 2.0411,
+      "step": 110
+    },
+    {
+      "epoch": 0.030307130293600326,
+      "grad_norm": 0.6895191073417664,
+      "learning_rate": 1.0009017132551849e-05,
+      "loss": 2.0475,
+      "step": 112
+    },
+    {
+      "epoch": 0.030848329048843187,
+      "grad_norm": 0.95854252576828,
+      "learning_rate": 1.018935978358882e-05,
+      "loss": 2.0071,
+      "step": 114
+    },
+    {
+      "epoch": 0.03138952780408605,
+      "grad_norm": 1.1303929090499878,
+      "learning_rate": 1.036970243462579e-05,
+      "loss": 2.0008,
+      "step": 116
+    },
+    {
+      "epoch": 0.031930726559328916,
+      "grad_norm": 0.7708876729011536,
+      "learning_rate": 1.055004508566276e-05,
+      "loss": 2.0061,
+      "step": 118
+    },
+    {
+      "epoch": 0.03247192531457178,
+      "grad_norm": 0.9773860573768616,
+      "learning_rate": 1.073038773669973e-05,
+      "loss": 2.0096,
+      "step": 120
+    },
+    {
+      "epoch": 0.03301312406981464,
+      "grad_norm": 1.118385910987854,
+      "learning_rate": 1.09107303877367e-05,
+      "loss": 1.9939,
+      "step": 122
+    },
+    {
+      "epoch": 0.0335543228250575,
+      "grad_norm": 0.7215014696121216,
+      "learning_rate": 1.109107303877367e-05,
+      "loss": 1.9515,
+      "step": 124
+    },
+    {
+      "epoch": 0.03409552158030037,
+      "grad_norm": 0.9696834683418274,
+      "learning_rate": 1.1271415689810642e-05,
+      "loss": 1.9639,
+      "step": 126
+    },
+    {
+      "epoch": 0.03463672033554323,
+      "grad_norm": 0.945482611656189,
+      "learning_rate": 1.1451758340847611e-05,
+      "loss": 1.9397,
+      "step": 128
+    },
+    {
+      "epoch": 0.03517791909078609,
+      "grad_norm": 0.7454535365104675,
+      "learning_rate": 1.1632100991884582e-05,
+      "loss": 1.9353,
+      "step": 130
+    },
+    {
+      "epoch": 0.03571911784602896,
+      "grad_norm": 0.7824187278747559,
+      "learning_rate": 1.1812443642921551e-05,
+      "loss": 1.9227,
+      "step": 132
+    },
+    {
+      "epoch": 0.03626031660127182,
+      "grad_norm": 0.7939879894256592,
+      "learning_rate": 1.1992786293958522e-05,
+      "loss": 1.9126,
+      "step": 134
+    },
+    {
+      "epoch": 0.03680151535651468,
+      "grad_norm": 0.7776147723197937,
+      "learning_rate": 1.2173128944995491e-05,
+      "loss": 1.9002,
+      "step": 136
+    },
+    {
+      "epoch": 0.03734271411175754,
+      "grad_norm": 0.6580236554145813,
+      "learning_rate": 1.2353471596032462e-05,
+      "loss": 1.9121,
+      "step": 138
+    },
+    {
+      "epoch": 0.03788391286700041,
+      "grad_norm": 0.7200301289558411,
+      "learning_rate": 1.2533814247069433e-05,
+      "loss": 1.8885,
+      "step": 140
+    },
+    {
+      "epoch": 0.03842511162224327,
+      "grad_norm": 0.7958497405052185,
+      "learning_rate": 1.2714156898106402e-05,
+      "loss": 1.9095,
+      "step": 142
+    },
+    {
+      "epoch": 0.03896631037748613,
+      "grad_norm": 0.9120681881904602,
+      "learning_rate": 1.2894499549143375e-05,
+      "loss": 1.884,
+      "step": 144
+    },
+    {
+      "epoch": 0.039507509132729,
+      "grad_norm": 0.8108247518539429,
+      "learning_rate": 1.3074842200180342e-05,
+      "loss": 1.8656,
+      "step": 146
+    },
+    {
+      "epoch": 0.04004870788797186,
+      "grad_norm": 0.7010449171066284,
+      "learning_rate": 1.3255184851217315e-05,
+      "loss": 1.8635,
+      "step": 148
+    },
+    {
+      "epoch": 0.04058990664321472,
+      "grad_norm": 0.8178524374961853,
+      "learning_rate": 1.3435527502254284e-05,
+      "loss": 1.8933,
+      "step": 150
+    },
+    {
+      "epoch": 0.04113110539845758,
+      "grad_norm": 1.0447405576705933,
+      "learning_rate": 1.3615870153291255e-05,
+      "loss": 1.8523,
+      "step": 152
+    },
+    {
+      "epoch": 0.04167230415370045,
+      "grad_norm": 0.8516271710395813,
+      "learning_rate": 1.3796212804328224e-05,
+      "loss": 1.8528,
+      "step": 154
+    },
+    {
+      "epoch": 0.04221350290894331,
+      "grad_norm": 0.8437328934669495,
+      "learning_rate": 1.3976555455365195e-05,
+      "loss": 1.861,
+      "step": 156
+    },
+    {
+      "epoch": 0.04275470166418617,
+      "grad_norm": 0.851265549659729,
+      "learning_rate": 1.4156898106402164e-05,
+      "loss": 1.8315,
+      "step": 158
+    },
+    {
+      "epoch": 0.04329590041942904,
+      "grad_norm": 0.7337156534194946,
+      "learning_rate": 1.4337240757439135e-05,
+      "loss": 1.8354,
+      "step": 160
+    },
+    {
+      "epoch": 0.0438370991746719,
+      "grad_norm": 0.9754143357276917,
+      "learning_rate": 1.4517583408476104e-05,
+      "loss": 1.8252,
+      "step": 162
+    },
+    {
+      "epoch": 0.04437829792991476,
+      "grad_norm": 0.6172115802764893,
+      "learning_rate": 1.4697926059513075e-05,
+      "loss": 1.8094,
+      "step": 164
+    },
+    {
+      "epoch": 0.04491949668515762,
+      "grad_norm": 0.8304158449172974,
+      "learning_rate": 1.4878268710550044e-05,
+      "loss": 1.8078,
+      "step": 166
+    },
+    {
+      "epoch": 0.04546069544040049,
+      "grad_norm": 0.6388853788375854,
+      "learning_rate": 1.5058611361587017e-05,
+      "loss": 1.8106,
+      "step": 168
+    },
+    {
+      "epoch": 0.04600189419564335,
+      "grad_norm": 0.743231475353241,
+      "learning_rate": 1.5238954012623984e-05,
+      "loss": 1.8144,
+      "step": 170
+    },
+    {
+      "epoch": 0.04654309295088621,
+      "grad_norm": 0.6442289352416992,
+      "learning_rate": 1.5419296663660955e-05,
+      "loss": 1.7831,
+      "step": 172
+    },
+    {
+      "epoch": 0.04708429170612908,
+      "grad_norm": 0.6877187490463257,
+      "learning_rate": 1.559963931469793e-05,
+      "loss": 1.8043,
+      "step": 174
+    },
+    {
+      "epoch": 0.04762549046137194,
+      "grad_norm": 0.9389640688896179,
+      "learning_rate": 1.5779981965734897e-05,
+      "loss": 1.7869,
+      "step": 176
+    },
+    {
+      "epoch": 0.0481666892166148,
+      "grad_norm": 1.0456589460372925,
+      "learning_rate": 1.5960324616771868e-05,
+      "loss": 1.7681,
+      "step": 178
+    },
+    {
+      "epoch": 0.04870788797185766,
+      "grad_norm": 0.9617791175842285,
+      "learning_rate": 1.614066726780884e-05,
+      "loss": 1.7668,
+      "step": 180
+    },
+    {
+      "epoch": 0.04924908672710053,
+      "grad_norm": 0.9334360361099243,
+      "learning_rate": 1.632100991884581e-05,
+      "loss": 1.7893,
+      "step": 182
+    },
+    {
+      "epoch": 0.04979028548234339,
+      "grad_norm": 0.8952531814575195,
+      "learning_rate": 1.6501352569882777e-05,
+      "loss": 1.7758,
+      "step": 184
+    },
+    {
+      "epoch": 0.05033148423758625,
+      "grad_norm": 0.8544924855232239,
+      "learning_rate": 1.6681695220919748e-05,
+      "loss": 1.793,
+      "step": 186
+    },
+    {
+      "epoch": 0.05087268299282912,
+      "grad_norm": 0.7782765030860901,
+      "learning_rate": 1.686203787195672e-05,
+      "loss": 1.768,
+      "step": 188
+    },
+    {
+      "epoch": 0.05141388174807198,
+      "grad_norm": 0.7119695544242859,
+      "learning_rate": 1.704238052299369e-05,
+      "loss": 1.7685,
+      "step": 190
+    },
+    {
+      "epoch": 0.05195508050331484,
+      "grad_norm": 0.9119647145271301,
+      "learning_rate": 1.7222723174030657e-05,
+      "loss": 1.7706,
+      "step": 192
+    },
+    {
+      "epoch": 0.0524962792585577,
+      "grad_norm": 0.6414957642555237,
+      "learning_rate": 1.7403065825067628e-05,
+      "loss": 1.7626,
+      "step": 194
+    },
+    {
+      "epoch": 0.05303747801380057,
+      "grad_norm": 0.8069677352905273,
+      "learning_rate": 1.75834084761046e-05,
+      "loss": 1.7423,
+      "step": 196
+    },
+    {
+      "epoch": 0.05357867676904343,
+      "grad_norm": 0.6549937725067139,
+      "learning_rate": 1.776375112714157e-05,
+      "loss": 1.7428,
+      "step": 198
+    },
+    {
+      "epoch": 0.05411987552428629,
+      "grad_norm": 0.8064024448394775,
+      "learning_rate": 1.7944093778178538e-05,
+      "loss": 1.7448,
+      "step": 200
+    },
+    {
+      "epoch": 0.054661074279529154,
+      "grad_norm": 0.7182701826095581,
+      "learning_rate": 1.8124436429215512e-05,
+      "loss": 1.7248,
+      "step": 202
+    },
+    {
+      "epoch": 0.05520227303477202,
+      "grad_norm": 0.6997919678688049,
+      "learning_rate": 1.830477908025248e-05,
+      "loss": 1.7281,
+      "step": 204
+    },
+    {
+      "epoch": 0.05574347179001488,
+      "grad_norm": 0.7071277499198914,
+      "learning_rate": 1.848512173128945e-05,
+      "loss": 1.714,
+      "step": 206
+    },
+    {
+      "epoch": 0.056284670545257744,
+      "grad_norm": 0.6344273090362549,
+      "learning_rate": 1.866546438232642e-05,
+      "loss": 1.7463,
+      "step": 208
+    },
+    {
+      "epoch": 0.05682586930050061,
+      "grad_norm": 0.7192733883857727,
+      "learning_rate": 1.8845807033363392e-05,
+      "loss": 1.737,
+      "step": 210
+    },
+    {
+      "epoch": 0.05736706805574347,
+      "grad_norm": 0.7418521642684937,
+      "learning_rate": 1.9026149684400363e-05,
+      "loss": 1.7197,
+      "step": 212
+    },
+    {
+      "epoch": 0.057908266810986334,
+      "grad_norm": 0.875845730304718,
+      "learning_rate": 1.920649233543733e-05,
+      "loss": 1.6968,
+      "step": 214
+    },
+    {
+      "epoch": 0.058449465566229195,
+      "grad_norm": 0.7394037842750549,
+      "learning_rate": 1.9386834986474305e-05,
+      "loss": 1.7051,
+      "step": 216
+    },
+    {
+      "epoch": 0.05899066432147206,
+      "grad_norm": 0.6689572930335999,
+      "learning_rate": 1.9567177637511272e-05,
+      "loss": 1.7152,
+      "step": 218
+    },
+    {
+      "epoch": 0.05953186307671492,
+      "grad_norm": 0.7955539226531982,
+      "learning_rate": 1.9747520288548243e-05,
+      "loss": 1.7136,
+      "step": 220
+    },
+    {
+      "epoch": 0.060073061831957784,
+      "grad_norm": 0.7005388140678406,
+      "learning_rate": 1.9927862939585214e-05,
+      "loss": 1.7152,
+      "step": 222
+    },
+    {
+      "epoch": 0.06061426058720065,
+      "grad_norm": 0.6205731630325317,
+      "learning_rate": 2.0108205590622185e-05,
+      "loss": 1.6901,
+      "step": 224
+    },
+    {
+      "epoch": 0.06115545934244351,
+      "grad_norm": 0.7079929709434509,
+      "learning_rate": 2.0288548241659152e-05,
+      "loss": 1.6905,
+      "step": 226
+    },
+    {
+      "epoch": 0.061696658097686374,
+      "grad_norm": 0.6871302723884583,
+      "learning_rate": 2.0468890892696123e-05,
+      "loss": 1.6867,
+      "step": 228
+    },
+    {
+      "epoch": 0.062237856852929235,
+      "grad_norm": 0.7172162532806396,
+      "learning_rate": 2.0649233543733094e-05,
+      "loss": 1.685,
+      "step": 230
+    },
+    {
+      "epoch": 0.0627790556081721,
+      "grad_norm": 0.6729004979133606,
+      "learning_rate": 2.0829576194770065e-05,
+      "loss": 1.6961,
+      "step": 232
+    },
+    {
+      "epoch": 0.06332025436341496,
+      "grad_norm": 0.7335099577903748,
+      "learning_rate": 2.1009918845807033e-05,
+      "loss": 1.6797,
+      "step": 234
+    },
+    {
+      "epoch": 0.06386145311865783,
+      "grad_norm": 0.6398060321807861,
+      "learning_rate": 2.1190261496844003e-05,
+      "loss": 1.7037,
+      "step": 236
+    },
+    {
+      "epoch": 0.0644026518739007,
+      "grad_norm": 0.7026365399360657,
+      "learning_rate": 2.1370604147880974e-05,
+      "loss": 1.6698,
+      "step": 238
+    },
+    {
+      "epoch": 0.06494385062914355,
+      "grad_norm": 0.7972332239151001,
+      "learning_rate": 2.1550946798917945e-05,
+      "loss": 1.6866,
+      "step": 240
+    },
+    {
+      "epoch": 0.06548504938438642,
+      "grad_norm": 0.7363021969795227,
+      "learning_rate": 2.1731289449954913e-05,
+      "loss": 1.6879,
+      "step": 242
+    },
+    {
+      "epoch": 0.06602624813962928,
+      "grad_norm": 0.7071017026901245,
+      "learning_rate": 2.1911632100991887e-05,
+      "loss": 1.6898,
+      "step": 244
+    },
+    {
+      "epoch": 0.06656744689487214,
+      "grad_norm": 0.8030880093574524,
+      "learning_rate": 2.2091974752028858e-05,
+      "loss": 1.6734,
+      "step": 246
+    },
+    {
+      "epoch": 0.067108645650115,
+      "grad_norm": 0.7429569363594055,
+      "learning_rate": 2.2272317403065825e-05,
+      "loss": 1.6722,
+      "step": 248
+    },
+    {
+      "epoch": 0.06764984440535787,
+      "grad_norm": 0.6807804107666016,
+      "learning_rate": 2.2452660054102796e-05,
+      "loss": 1.6697,
+      "step": 250
+    },
+    {
+      "epoch": 0.06819104316060073,
+      "grad_norm": 0.6632562875747681,
+      "learning_rate": 2.2633002705139767e-05,
+      "loss": 1.6453,
+      "step": 252
+    },
+    {
+      "epoch": 0.0687322419158436,
+      "grad_norm": 0.6661680340766907,
+      "learning_rate": 2.2813345356176738e-05,
+      "loss": 1.6701,
+      "step": 254
+    },
+    {
+      "epoch": 0.06927344067108646,
+      "grad_norm": 0.6747105121612549,
+      "learning_rate": 2.2993688007213706e-05,
+      "loss": 1.6729,
+      "step": 256
+    },
+    {
+      "epoch": 0.06981463942632932,
+      "grad_norm": 0.7698473334312439,
+      "learning_rate": 2.317403065825068e-05,
+      "loss": 1.6528,
+      "step": 258
+    },
+    {
+      "epoch": 0.07035583818157218,
+      "grad_norm": 0.6111325621604919,
+      "learning_rate": 2.3354373309287647e-05,
+      "loss": 1.6412,
+      "step": 260
+    },
+    {
+      "epoch": 0.07089703693681504,
+      "grad_norm": 0.7405019998550415,
+      "learning_rate": 2.3534715960324618e-05,
+      "loss": 1.6564,
+      "step": 262
+    },
+    {
+      "epoch": 0.07143823569205791,
+      "grad_norm": 0.6702501773834229,
+      "learning_rate": 2.371505861136159e-05,
+      "loss": 1.654,
+      "step": 264
+    },
+    {
+      "epoch": 0.07197943444730077,
+      "grad_norm": 0.7076373100280762,
+      "learning_rate": 2.389540126239856e-05,
+      "loss": 1.6301,
+      "step": 266
+    },
+    {
+      "epoch": 0.07252063320254364,
+      "grad_norm": 0.7239627242088318,
+      "learning_rate": 2.4075743913435528e-05,
+      "loss": 1.6575,
+      "step": 268
+    },
+    {
+      "epoch": 0.0730618319577865,
+      "grad_norm": 0.753480076789856,
+      "learning_rate": 2.42560865644725e-05,
+      "loss": 1.6603,
+      "step": 270
+    },
+    {
+      "epoch": 0.07360303071302936,
+      "grad_norm": 0.7261641025543213,
+      "learning_rate": 2.443642921550947e-05,
+      "loss": 1.6449,
+      "step": 272
+    },
+    {
+      "epoch": 0.07414422946827222,
+      "grad_norm": 0.6315119862556458,
+      "learning_rate": 2.461677186654644e-05,
+      "loss": 1.6538,
+      "step": 274
+    },
+    {
+      "epoch": 0.07468542822351508,
+      "grad_norm": 0.5698412656784058,
+      "learning_rate": 2.4797114517583408e-05,
+      "loss": 1.6663,
+      "step": 276
+    },
+    {
+      "epoch": 0.07522662697875795,
+      "grad_norm": 0.5968983173370361,
+      "learning_rate": 2.497745716862038e-05,
+      "loss": 1.643,
+      "step": 278
+    },
+    {
+      "epoch": 0.07576782573400082,
+      "grad_norm": 0.561126172542572,
+      "learning_rate": 2.5157799819657353e-05,
+      "loss": 1.6301,
+      "step": 280
+    },
+    {
+      "epoch": 0.07630902448924368,
+      "grad_norm": 0.7290865778923035,
+      "learning_rate": 2.533814247069432e-05,
+      "loss": 1.6412,
+      "step": 282
+    },
+    {
+      "epoch": 0.07685022324448654,
+      "grad_norm": 0.7629122138023376,
+      "learning_rate": 2.5518485121731288e-05,
+      "loss": 1.6335,
+      "step": 284
+    },
+    {
+      "epoch": 0.0773914219997294,
+      "grad_norm": 0.5383496284484863,
+      "learning_rate": 2.5698827772768262e-05,
+      "loss": 1.6226,
+      "step": 286
+    },
+    {
+      "epoch": 0.07793262075497226,
+      "grad_norm": 0.7778373956680298,
+      "learning_rate": 2.5879170423805233e-05,
+      "loss": 1.6333,
+      "step": 288
+    },
+    {
+      "epoch": 0.07847381951021512,
+      "grad_norm": 0.6851366758346558,
+      "learning_rate": 2.60595130748422e-05,
+      "loss": 1.6251,
+      "step": 290
+    },
+    {
+      "epoch": 0.079015018265458,
+      "grad_norm": 0.5947225689888,
+      "learning_rate": 2.623985572587917e-05,
+      "loss": 1.6298,
+      "step": 292
+    },
+    {
+      "epoch": 0.07955621702070086,
+      "grad_norm": 0.9742544889450073,
+      "learning_rate": 2.6420198376916146e-05,
+      "loss": 1.6252,
+      "step": 294
+    },
+    {
+      "epoch": 0.08009741577594372,
+      "grad_norm": 1.2064323425292969,
+      "learning_rate": 2.6600541027953113e-05,
+      "loss": 1.6152,
+      "step": 296
+    },
+    {
+      "epoch": 0.08063861453118658,
+      "grad_norm": 1.0506716966629028,
+      "learning_rate": 2.678088367899008e-05,
+      "loss": 1.6351,
+      "step": 298
+    },
+    {
+      "epoch": 0.08117981328642944,
+      "grad_norm": 1.2992738485336304,
+      "learning_rate": 2.696122633002705e-05,
+      "loss": 1.6193,
+      "step": 300
+    },
+    {
+      "epoch": 0.0817210120416723,
+      "grad_norm": 1.0616599321365356,
+      "learning_rate": 2.7141568981064026e-05,
+      "loss": 1.6135,
+      "step": 302
+    },
+    {
+      "epoch": 0.08226221079691516,
+      "grad_norm": 1.037997841835022,
+      "learning_rate": 2.7321911632100993e-05,
+      "loss": 1.6344,
+      "step": 304
+    },
+    {
+      "epoch": 0.08280340955215804,
+      "grad_norm": 0.8937569856643677,
+      "learning_rate": 2.7502254283137964e-05,
+      "loss": 1.6077,
+      "step": 306
+    },
+    {
+      "epoch": 0.0833446083074009,
+      "grad_norm": 1.1334234476089478,
+      "learning_rate": 2.7682596934174932e-05,
+      "loss": 1.6193,
+      "step": 308
+    },
+    {
+      "epoch": 0.08388580706264376,
+      "grad_norm": 0.8336219191551208,
+      "learning_rate": 2.7862939585211906e-05,
+      "loss": 1.5948,
+      "step": 310
+    },
+    {
+      "epoch": 0.08442700581788662,
+      "grad_norm": 1.1825398206710815,
+      "learning_rate": 2.8043282236248874e-05,
+      "loss": 1.6239,
+      "step": 312
+    },
+    {
+      "epoch": 0.08496820457312948,
+      "grad_norm": 0.7945433259010315,
+      "learning_rate": 2.8223624887285844e-05,
+      "loss": 1.6119,
+      "step": 314
+    },
+    {
+      "epoch": 0.08550940332837234,
+      "grad_norm": 0.6971009969711304,
+      "learning_rate": 2.8403967538322812e-05,
+      "loss": 1.5822,
+      "step": 316
+    },
+    {
+      "epoch": 0.0860506020836152,
+      "grad_norm": 0.6050766706466675,
+      "learning_rate": 2.8584310189359786e-05,
+      "loss": 1.6161,
+      "step": 318
+    },
+    {
+      "epoch": 0.08659180083885808,
+      "grad_norm": 0.6123189330101013,
+      "learning_rate": 2.8764652840396754e-05,
+      "loss": 1.5941,
+      "step": 320
+    },
+    {
+      "epoch": 0.08713299959410094,
+      "grad_norm": 0.5471253395080566,
+      "learning_rate": 2.8944995491433725e-05,
+      "loss": 1.603,
+      "step": 322
+    },
+    {
+      "epoch": 0.0876741983493438,
+      "grad_norm": 0.5793882608413696,
+      "learning_rate": 2.91253381424707e-05,
+      "loss": 1.6076,
+      "step": 324
+    },
+    {
+      "epoch": 0.08821539710458666,
+      "grad_norm": 0.5409413576126099,
+      "learning_rate": 2.9305680793507666e-05,
+      "loss": 1.5825,
+      "step": 326
+    },
+    {
+      "epoch": 0.08875659585982952,
+      "grad_norm": 6.757148265838623,
+      "learning_rate": 2.9486023444544637e-05,
+      "loss": 1.5942,
+      "step": 328
+    },
+    {
+      "epoch": 0.08929779461507238,
+      "grad_norm": 1.3357856273651123,
+      "learning_rate": 2.9666366095581605e-05,
+      "loss": 1.642,
+      "step": 330
+    },
+    {
+      "epoch": 0.08983899337031524,
+      "grad_norm": 0.8245829939842224,
+      "learning_rate": 2.984670874661858e-05,
+      "loss": 1.6062,
+      "step": 332
+    },
+    {
+      "epoch": 0.09038019212555812,
+      "grad_norm": 0.8888993263244629,
+      "learning_rate": 3.0027051397655547e-05,
+      "loss": 1.5952,
+      "step": 334
+    },
+    {
+      "epoch": 0.09092139088080098,
+      "grad_norm": 0.8923915028572083,
+      "learning_rate": 3.0207394048692517e-05,
+      "loss": 1.5977,
+      "step": 336
+    },
+    {
+      "epoch": 0.09146258963604384,
+      "grad_norm": 0.7443459033966064,
+      "learning_rate": 3.0387736699729485e-05,
+      "loss": 1.5738,
+      "step": 338
+    },
+    {
+      "epoch": 0.0920037883912867,
+      "grad_norm": 0.7297430038452148,
+      "learning_rate": 3.056807935076646e-05,
+      "loss": 1.5907,
+      "step": 340
+    },
+    {
+      "epoch": 0.09254498714652956,
+      "grad_norm": 0.6882812976837158,
+      "learning_rate": 3.074842200180343e-05,
+      "loss": 1.5767,
+      "step": 342
+    },
+    {
+      "epoch": 0.09308618590177242,
+      "grad_norm": 0.6150392889976501,
+      "learning_rate": 3.0928764652840394e-05,
+      "loss": 1.5747,
+      "step": 344
+    },
+    {
+      "epoch": 0.09362738465701528,
+      "grad_norm": 0.6230599284172058,
+      "learning_rate": 3.110910730387737e-05,
+      "loss": 1.583,
+      "step": 346
+    },
+    {
+      "epoch": 0.09416858341225816,
+      "grad_norm": 0.6081874966621399,
+      "learning_rate": 3.128944995491434e-05,
+      "loss": 1.5875,
+      "step": 348
+    },
+    {
+      "epoch": 0.09470978216750102,
+      "grad_norm": 0.5467821955680847,
+      "learning_rate": 3.146979260595131e-05,
+      "loss": 1.575,
+      "step": 350
+    },
+    {
+      "epoch": 0.09525098092274388,
+      "grad_norm": 0.5629361271858215,
+      "learning_rate": 3.165013525698828e-05,
+      "loss": 1.5828,
+      "step": 352
+    },
+    {
+      "epoch": 0.09579217967798674,
+      "grad_norm": 0.5995283126831055,
+      "learning_rate": 3.1830477908025245e-05,
+      "loss": 1.5872,
+      "step": 354
+    },
+    {
+      "epoch": 0.0963333784332296,
+      "grad_norm": 0.556450366973877,
+      "learning_rate": 3.201082055906222e-05,
+      "loss": 1.553,
+      "step": 356
+    },
+    {
+      "epoch": 0.09687457718847246,
+      "grad_norm": 0.6498537063598633,
+      "learning_rate": 3.219116321009919e-05,
+      "loss": 1.5667,
+      "step": 358
+    },
+    {
+      "epoch": 0.09741577594371532,
+      "grad_norm": 0.5891172885894775,
+      "learning_rate": 3.237150586113616e-05,
+      "loss": 1.5818,
+      "step": 360
+    },
+    {
+      "epoch": 0.0979569746989582,
+      "grad_norm": 0.6487797498703003,
+      "learning_rate": 3.2551848512173136e-05,
+      "loss": 1.5582,
+      "step": 362
+    },
+    {
+      "epoch": 0.09849817345420106,
+      "grad_norm": 0.5860658884048462,
+      "learning_rate": 3.27321911632101e-05,
+      "loss": 1.5725,
+      "step": 364
+    },
+    {
+      "epoch": 0.09903937220944392,
+      "grad_norm": 0.5619581937789917,
+      "learning_rate": 3.291253381424707e-05,
+      "loss": 1.5779,
+      "step": 366
+    },
+    {
+      "epoch": 0.09958057096468678,
+      "grad_norm": 0.7147429585456848,
+      "learning_rate": 3.309287646528404e-05,
+      "loss": 1.5766,
+      "step": 368
+    },
+    {
+      "epoch": 0.10012176971992964,
+      "grad_norm": 0.5840562582015991,
+      "learning_rate": 3.327321911632101e-05,
+      "loss": 1.5609,
+      "step": 370
+    },
+    {
+      "epoch": 0.1006629684751725,
+      "grad_norm": 0.6277860403060913,
+      "learning_rate": 3.345356176735798e-05,
+      "loss": 1.5645,
+      "step": 372
+    },
+    {
+      "epoch": 0.10120416723041536,
+      "grad_norm": 0.6395567655563354,
+      "learning_rate": 3.3633904418394954e-05,
+      "loss": 1.545,
+      "step": 374
+    },
+    {
+      "epoch": 0.10174536598565824,
+      "grad_norm": 0.6651553511619568,
+      "learning_rate": 3.381424706943192e-05,
+      "loss": 1.5643,
+      "step": 376
+    },
+    {
+      "epoch": 0.1022865647409011,
+      "grad_norm": 0.6691033244132996,
+      "learning_rate": 3.3994589720468896e-05,
+      "loss": 1.5705,
+      "step": 378
+    },
+    {
+      "epoch": 0.10282776349614396,
+      "grad_norm": 0.5426511764526367,
+      "learning_rate": 3.4174932371505863e-05,
+      "loss": 1.536,
+      "step": 380
+    },
+    {
+      "epoch": 0.10336896225138682,
+      "grad_norm": 0.6677694916725159,
+      "learning_rate": 3.435527502254283e-05,
+      "loss": 1.5664,
+      "step": 382
+    },
+    {
+      "epoch": 0.10391016100662968,
+      "grad_norm": 0.5283762216567993,
+      "learning_rate": 3.45356176735798e-05,
+      "loss": 1.5474,
+      "step": 384
+    },
+    {
+      "epoch": 0.10445135976187254,
+      "grad_norm": 0.652812659740448,
+      "learning_rate": 3.471596032461677e-05,
+      "loss": 1.5509,
+      "step": 386
+    },
+    {
+      "epoch": 0.1049925585171154,
+      "grad_norm": 0.8639987111091614,
+      "learning_rate": 3.489630297565375e-05,
+      "loss": 1.5563,
+      "step": 388
+    },
+    {
+      "epoch": 0.10553375727235827,
+      "grad_norm": 0.7726946473121643,
+      "learning_rate": 3.5076645626690715e-05,
+      "loss": 1.5682,
+      "step": 390
+    },
+    {
+      "epoch": 0.10607495602760114,
+      "grad_norm": 0.6511155962944031,
+      "learning_rate": 3.525698827772768e-05,
+      "loss": 1.5571,
+      "step": 392
+    },
+    {
+      "epoch": 0.106616154782844,
+      "grad_norm": 0.6578395962715149,
+      "learning_rate": 3.5437330928764656e-05,
+      "loss": 1.5452,
+      "step": 394
+    },
+    {
+      "epoch": 0.10715735353808686,
+      "grad_norm": 0.642919659614563,
+      "learning_rate": 3.5617673579801624e-05,
+      "loss": 1.5508,
+      "step": 396
+    },
+    {
+      "epoch": 0.10769855229332972,
+      "grad_norm": 0.5190348029136658,
+      "learning_rate": 3.579801623083859e-05,
+      "loss": 1.5432,
+      "step": 398
+    },
+    {
+      "epoch": 0.10823975104857259,
+      "grad_norm": 0.48932549357414246,
+      "learning_rate": 3.5978358881875566e-05,
+      "loss": 1.5544,
+      "step": 400
+    },
+    {
+      "epoch": 0.10878094980381545,
+      "grad_norm": 0.5018340945243835,
+      "learning_rate": 3.615870153291254e-05,
+      "loss": 1.5322,
+      "step": 402
+    },
+    {
+      "epoch": 0.10932214855905831,
+      "grad_norm": 0.5701499581336975,
+      "learning_rate": 3.633904418394951e-05,
+      "loss": 1.5288,
+      "step": 404
+    },
+    {
+      "epoch": 0.10986334731430118,
+      "grad_norm": 0.6049205660820007,
+      "learning_rate": 3.6519386834986475e-05,
+      "loss": 1.5627,
+      "step": 406
+    },
+    {
+      "epoch": 0.11040454606954404,
+      "grad_norm": 0.5781517028808594,
+      "learning_rate": 3.669972948602345e-05,
+      "loss": 1.542,
+      "step": 408
+    },
+    {
+      "epoch": 0.1109457448247869,
+      "grad_norm": 0.5594660043716431,
+      "learning_rate": 3.688007213706042e-05,
+      "loss": 1.5461,
+      "step": 410
+    },
+    {
+      "epoch": 0.11148694358002977,
+      "grad_norm": 0.5319619178771973,
+      "learning_rate": 3.7060414788097384e-05,
+      "loss": 1.5668,
+      "step": 412
+    },
+    {
+      "epoch": 0.11202814233527263,
+      "grad_norm": 0.5311123728752136,
+      "learning_rate": 3.724075743913435e-05,
+      "loss": 1.528,
+      "step": 414
+    },
+    {
+      "epoch": 0.11256934109051549,
+      "grad_norm": 0.5555101633071899,
+      "learning_rate": 3.7421100090171326e-05,
+      "loss": 1.5392,
+      "step": 416
+    },
+    {
+      "epoch": 0.11311053984575835,
+      "grad_norm": 0.5486223101615906,
+      "learning_rate": 3.76014427412083e-05,
+      "loss": 1.5337,
+      "step": 418
+    },
+    {
+      "epoch": 0.11365173860100122,
+      "grad_norm": 0.5156669020652771,
+      "learning_rate": 3.778178539224527e-05,
+      "loss": 1.5105,
+      "step": 420
+    },
+    {
+      "epoch": 0.11419293735624408,
+      "grad_norm": 0.49596554040908813,
+      "learning_rate": 3.7962128043282235e-05,
+      "loss": 1.515,
+      "step": 422
+    },
+    {
+      "epoch": 0.11473413611148695,
+      "grad_norm": 0.641333281993866,
+      "learning_rate": 3.814247069431921e-05,
+      "loss": 1.5328,
+      "step": 424
+    },
+    {
+      "epoch": 0.1152753348667298,
+      "grad_norm": 0.6106113195419312,
+      "learning_rate": 3.832281334535618e-05,
+      "loss": 1.5189,
+      "step": 426
+    },
+    {
+      "epoch": 0.11581653362197267,
+      "grad_norm": 0.5619134306907654,
+      "learning_rate": 3.8503155996393145e-05,
+      "loss": 1.5295,
+      "step": 428
+    },
+    {
+      "epoch": 0.11635773237721553,
+      "grad_norm": 0.5396978259086609,
+      "learning_rate": 3.868349864743012e-05,
+      "loss": 1.5173,
+      "step": 430
+    },
+    {
+      "epoch": 0.11689893113245839,
+      "grad_norm": 0.5466894507408142,
+      "learning_rate": 3.886384129846709e-05,
+      "loss": 1.5191,
+      "step": 432
+    },
+    {
+      "epoch": 0.11744012988770126,
+      "grad_norm": 0.5601218342781067,
+      "learning_rate": 3.904418394950406e-05,
+      "loss": 1.5285,
+      "step": 434
+    },
+    {
+      "epoch": 0.11798132864294412,
+      "grad_norm": 0.6620492935180664,
+      "learning_rate": 3.922452660054103e-05,
+      "loss": 1.4946,
+      "step": 436
+    },
+    {
+      "epoch": 0.11852252739818699,
+      "grad_norm": 0.49140048027038574,
+      "learning_rate": 3.9404869251578e-05,
+      "loss": 1.512,
+      "step": 438
+    },
+    {
+      "epoch": 0.11906372615342985,
+      "grad_norm": 0.5824118256568909,
+      "learning_rate": 3.958521190261497e-05,
+      "loss": 1.5244,
+      "step": 440
+    },
+    {
+      "epoch": 0.11960492490867271,
+      "grad_norm": 0.4967150092124939,
+      "learning_rate": 3.976555455365194e-05,
+      "loss": 1.5273,
+      "step": 442
+    },
+    {
+      "epoch": 0.12014612366391557,
+      "grad_norm": 0.5089767575263977,
+      "learning_rate": 3.994589720468891e-05,
+      "loss": 1.5119,
+      "step": 444
+    },
+    {
+      "epoch": 0.12068732241915843,
+      "grad_norm": 0.5404312014579773,
+      "learning_rate": 4.0126239855725886e-05,
+      "loss": 1.5072,
+      "step": 446
+    },
+    {
+      "epoch": 0.1212285211744013,
+      "grad_norm": 0.5239550471305847,
+      "learning_rate": 4.0306582506762853e-05,
+      "loss": 1.5336,
+      "step": 448
+    },
+    {
+      "epoch": 0.12176971992964417,
+      "grad_norm": 0.4974781274795532,
+      "learning_rate": 4.048692515779982e-05,
+      "loss": 1.5225,
+      "step": 450
+    },
+    {
+      "epoch": 0.12231091868488703,
+      "grad_norm": 0.5363791584968567,
+      "learning_rate": 4.066726780883679e-05,
+      "loss": 1.5176,
+      "step": 452
+    },
+    {
+      "epoch": 0.12285211744012989,
+      "grad_norm": 0.5095157027244568,
+      "learning_rate": 4.084761045987376e-05,
+      "loss": 1.4936,
+      "step": 454
+    },
+    {
+      "epoch": 0.12339331619537275,
+      "grad_norm": 0.4920356869697571,
+      "learning_rate": 4.102795311091073e-05,
+      "loss": 1.5269,
+      "step": 456
+    },
+    {
+      "epoch": 0.12393451495061561,
+      "grad_norm": 0.4940793514251709,
+      "learning_rate": 4.1208295761947705e-05,
+      "loss": 1.5072,
+      "step": 458
+    },
+    {
+      "epoch": 0.12447571370585847,
+      "grad_norm": 0.4805227220058441,
+      "learning_rate": 4.138863841298467e-05,
+      "loss": 1.4987,
+      "step": 460
+    },
+    {
+      "epoch": 0.12501691246110133,
+      "grad_norm": 0.49683934450149536,
+      "learning_rate": 4.1568981064021646e-05,
+      "loss": 1.5008,
+      "step": 462
+    },
+    {
+      "epoch": 0.1255581112163442,
+      "grad_norm": 0.5283801555633545,
+      "learning_rate": 4.1749323715058614e-05,
+      "loss": 1.5177,
+      "step": 464
+    },
+    {
+      "epoch": 0.12609930997158705,
+      "grad_norm": 0.5395119190216064,
+      "learning_rate": 4.192966636609558e-05,
+      "loss": 1.5106,
+      "step": 466
+    },
+    {
+      "epoch": 0.12664050872682991,
+      "grad_norm": 0.5403693914413452,
+      "learning_rate": 4.211000901713255e-05,
+      "loss": 1.4854,
+      "step": 468
+    },
+    {
+      "epoch": 0.1271817074820728,
+      "grad_norm": 0.4690951406955719,
+      "learning_rate": 4.229035166816952e-05,
+      "loss": 1.5079,
+      "step": 470
+    },
+    {
+      "epoch": 0.12772290623731566,
+      "grad_norm": 0.5077293515205383,
+      "learning_rate": 4.24706943192065e-05,
+      "loss": 1.4953,
+      "step": 472
+    },
+    {
+      "epoch": 0.12826410499255853,
+      "grad_norm": 0.440019816160202,
+      "learning_rate": 4.2651036970243465e-05,
+      "loss": 1.4864,
+      "step": 474
+    },
+    {
+      "epoch": 0.1288053037478014,
+      "grad_norm": 0.48672759532928467,
+      "learning_rate": 4.283137962128044e-05,
+      "loss": 1.5205,
+      "step": 476
+    },
+    {
+      "epoch": 0.12934650250304425,
+      "grad_norm": 0.4732811450958252,
+      "learning_rate": 4.301172227231741e-05,
+      "loss": 1.4998,
+      "step": 478
+    },
+    {
+      "epoch": 0.1298877012582871,
+      "grad_norm": 0.46713048219680786,
+      "learning_rate": 4.3192064923354374e-05,
+      "loss": 1.4893,
+      "step": 480
+    },
+    {
+      "epoch": 0.13042890001352997,
+      "grad_norm": 0.502356231212616,
+      "learning_rate": 4.337240757439134e-05,
+      "loss": 1.5125,
+      "step": 482
+    },
+    {
+      "epoch": 0.13097009876877283,
+      "grad_norm": 0.45067864656448364,
+      "learning_rate": 4.3552750225428316e-05,
+      "loss": 1.4978,
+      "step": 484
+    },
+    {
+      "epoch": 0.1315112975240157,
+      "grad_norm": 0.46964120864868164,
+      "learning_rate": 4.373309287646529e-05,
+      "loss": 1.5006,
+      "step": 486
+    },
+    {
+      "epoch": 0.13205249627925855,
+      "grad_norm": 0.47723180055618286,
+      "learning_rate": 4.391343552750226e-05,
+      "loss": 1.513,
+      "step": 488
+    },
+    {
+      "epoch": 0.1325936950345014,
+      "grad_norm": 0.5100542306900024,
+      "learning_rate": 4.4093778178539225e-05,
+      "loss": 1.5279,
+      "step": 490
+    },
+    {
+      "epoch": 0.13313489378974427,
+      "grad_norm": 0.5344257354736328,
+      "learning_rate": 4.42741208295762e-05,
+      "loss": 1.5193,
+      "step": 492
+    },
+    {
+      "epoch": 0.13367609254498714,
+      "grad_norm": 0.5867893695831299,
+      "learning_rate": 4.445446348061317e-05,
+      "loss": 1.512,
+      "step": 494
+    },
+    {
+      "epoch": 0.13421729130023,
+      "grad_norm": 0.7811394929885864,
+      "learning_rate": 4.4634806131650134e-05,
+      "loss": 1.5038,
+      "step": 496
+    },
+    {
+      "epoch": 0.13475849005547288,
+      "grad_norm": 0.8505339622497559,
+      "learning_rate": 4.48151487826871e-05,
+      "loss": 1.5169,
+      "step": 498
+    },
+    {
+      "epoch": 0.13529968881071575,
+      "grad_norm": 0.6337641477584839,
+      "learning_rate": 4.4995491433724076e-05,
+      "loss": 1.4951,
+      "step": 500
+    },
+    {
+      "epoch": 0.1358408875659586,
+      "grad_norm": 0.7979961633682251,
+      "learning_rate": 4.517583408476105e-05,
+      "loss": 1.5031,
+      "step": 502
+    },
+    {
+      "epoch": 0.13638208632120147,
+      "grad_norm": 0.6946894526481628,
+      "learning_rate": 4.535617673579802e-05,
+      "loss": 1.501,
+      "step": 504
+    },
+    {
+      "epoch": 0.13692328507644433,
+      "grad_norm": 0.6830259561538696,
+      "learning_rate": 4.5536519386834986e-05,
+      "loss": 1.4896,
+      "step": 506
+    },
+    {
+      "epoch": 0.1374644838316872,
+      "grad_norm": 0.5908662676811218,
+      "learning_rate": 4.571686203787196e-05,
+      "loss": 1.4992,
+      "step": 508
+    },
+    {
+      "epoch": 0.13800568258693005,
+      "grad_norm": 0.7655865550041199,
+      "learning_rate": 4.589720468890893e-05,
+      "loss": 1.4911,
+      "step": 510
+    },
+    {
+      "epoch": 0.1385468813421729,
+      "grad_norm": 0.5924785733222961,
+      "learning_rate": 4.6077547339945895e-05,
+      "loss": 1.4719,
+      "step": 512
+    },
+    {
+      "epoch": 0.13908808009741577,
+      "grad_norm": 0.6654263138771057,
+      "learning_rate": 4.625788999098287e-05,
+      "loss": 1.5109,
+      "step": 514
+    },
+    {
+      "epoch": 0.13962927885265863,
+      "grad_norm": 0.5296297073364258,
+      "learning_rate": 4.6438232642019843e-05,
+      "loss": 1.4934,
+      "step": 516
+    },
+    {
+      "epoch": 0.1401704776079015,
+      "grad_norm": 0.5698690414428711,
+      "learning_rate": 4.661857529305681e-05,
+      "loss": 1.4954,
+      "step": 518
+    },
+    {
+      "epoch": 0.14071167636314436,
+      "grad_norm": 0.5790325403213501,
+      "learning_rate": 4.679891794409378e-05,
+      "loss": 1.4673,
+      "step": 520
+    },
+    {
+      "epoch": 0.14125287511838722,
+      "grad_norm": 0.551480770111084,
+      "learning_rate": 4.697926059513075e-05,
+      "loss": 1.476,
+      "step": 522
+    },
+    {
+      "epoch": 0.14179407387363008,
+      "grad_norm": 0.5201780796051025,
+      "learning_rate": 4.715960324616772e-05,
+      "loss": 1.4701,
+      "step": 524
+    },
+    {
+      "epoch": 0.14233527262887297,
+      "grad_norm": 0.46442562341690063,
+      "learning_rate": 4.733994589720469e-05,
+      "loss": 1.4831,
+      "step": 526
+    },
+    {
+      "epoch": 0.14287647138411583,
+      "grad_norm": 0.5558522939682007,
+      "learning_rate": 4.752028854824166e-05,
+      "loss": 1.4729,
+      "step": 528
+    },
+    {
+      "epoch": 0.1434176701393587,
+      "grad_norm": 0.48511791229248047,
+      "learning_rate": 4.7700631199278636e-05,
+      "loss": 1.4742,
+      "step": 530
+    },
+    {
+      "epoch": 0.14395886889460155,
+      "grad_norm": 0.5244829058647156,
+      "learning_rate": 4.7880973850315604e-05,
+      "loss": 1.4928,
+      "step": 532
+    },
+    {
+      "epoch": 0.1445000676498444,
+      "grad_norm": 0.48878946900367737,
+      "learning_rate": 4.806131650135257e-05,
+      "loss": 1.4921,
+      "step": 534
+    },
+    {
+      "epoch": 0.14504126640508727,
+      "grad_norm": 0.5348760485649109,
+      "learning_rate": 4.824165915238954e-05,
+      "loss": 1.4917,
+      "step": 536
+    },
+    {
+      "epoch": 0.14558246516033013,
+      "grad_norm": 0.5444923639297485,
+      "learning_rate": 4.842200180342651e-05,
+      "loss": 1.4546,
+      "step": 538
+    },
+    {
+      "epoch": 0.146123663915573,
+      "grad_norm": 0.494761198759079,
+      "learning_rate": 4.860234445446348e-05,
+      "loss": 1.4751,
+      "step": 540
+    },
+    {
+      "epoch": 0.14666486267081585,
+      "grad_norm": 0.4921441674232483,
+      "learning_rate": 4.8782687105500455e-05,
+      "loss": 1.4767,
+      "step": 542
+    },
+    {
+      "epoch": 0.14720606142605872,
+      "grad_norm": 0.48382577300071716,
+      "learning_rate": 4.896302975653742e-05,
+      "loss": 1.485,
+      "step": 544
+    },
+    {
+      "epoch": 0.14774726018130158,
+      "grad_norm": 0.4616708755493164,
+      "learning_rate": 4.9143372407574397e-05,
+      "loss": 1.4732,
+      "step": 546
+    },
+    {
+      "epoch": 0.14828845893654444,
+      "grad_norm": 0.5030043125152588,
+      "learning_rate": 4.9323715058611364e-05,
+      "loss": 1.4799,
+      "step": 548
+    },
+    {
+      "epoch": 0.1488296576917873,
+      "grad_norm": 0.467230886220932,
+      "learning_rate": 4.950405770964833e-05,
+      "loss": 1.4594,
+      "step": 550
+    },
+    {
+      "epoch": 0.14937085644703016,
+      "grad_norm": 0.42864304780960083,
+      "learning_rate": 4.9684400360685306e-05,
+      "loss": 1.4748,
+      "step": 552
+    },
+    {
+      "epoch": 0.14991205520227305,
+      "grad_norm": 0.43733683228492737,
+      "learning_rate": 4.986474301172227e-05,
+      "loss": 1.462,
+      "step": 554
+    },
+    {
+      "epoch": 0.1504532539575159,
+      "grad_norm": 0.45550286769866943,
+      "learning_rate": 5.004508566275925e-05,
+      "loss": 1.475,
+      "step": 556
+    },
+    {
+      "epoch": 0.15099445271275877,
+      "grad_norm": 0.44999995827674866,
+      "learning_rate": 5.022542831379622e-05,
+      "loss": 1.4794,
+      "step": 558
+    },
+    {
+      "epoch": 0.15153565146800163,
+      "grad_norm": 0.5035279989242554,
+      "learning_rate": 5.040577096483319e-05,
+      "loss": 1.471,
+      "step": 560
+    },
+    {
+      "epoch": 0.1520768502232445,
+      "grad_norm": 0.44605591893196106,
+      "learning_rate": 5.058611361587016e-05,
+      "loss": 1.4461,
+      "step": 562
+    },
+    {
+      "epoch": 0.15261804897848735,
+      "grad_norm": 0.5482723712921143,
+      "learning_rate": 5.0766456266907124e-05,
+      "loss": 1.4597,
+      "step": 564
+    },
+    {
+      "epoch": 0.1531592477337302,
+      "grad_norm": 0.5323627591133118,
+      "learning_rate": 5.094679891794409e-05,
+      "loss": 1.4743,
+      "step": 566
+    },
+    {
+      "epoch": 0.15370044648897307,
+      "grad_norm": 0.5289944410324097,
+      "learning_rate": 5.1127141568981066e-05,
+      "loss": 1.5,
+      "step": 568
+    },
+    {
+      "epoch": 0.15424164524421594,
+      "grad_norm": 0.5446243286132812,
+      "learning_rate": 5.1307484220018034e-05,
+      "loss": 1.4751,
+      "step": 570
+    },
+    {
+      "epoch": 0.1547828439994588,
+      "grad_norm": 0.525830090045929,
+      "learning_rate": 5.1487826871055015e-05,
+      "loss": 1.4639,
+      "step": 572
+    },
+    {
+      "epoch": 0.15532404275470166,
+      "grad_norm": 0.48129191994667053,
+      "learning_rate": 5.166816952209198e-05,
+      "loss": 1.4652,
+      "step": 574
+    },
+    {
+      "epoch": 0.15586524150994452,
+      "grad_norm": 0.47915297746658325,
+      "learning_rate": 5.184851217312895e-05,
+      "loss": 1.4627,
+      "step": 576
+    },
+    {
+      "epoch": 0.15640644026518738,
+      "grad_norm": 0.5229325294494629,
+      "learning_rate": 5.202885482416592e-05,
+      "loss": 1.4525,
+      "step": 578
+    },
+    {
+      "epoch": 0.15694763902043024,
+      "grad_norm": 0.5452600121498108,
+      "learning_rate": 5.2209197475202885e-05,
+      "loss": 1.458,
+      "step": 580
+    },
+    {
+      "epoch": 0.15748883777567313,
+      "grad_norm": 0.427432656288147,
+      "learning_rate": 5.238954012623985e-05,
+      "loss": 1.4773,
+      "step": 582
+    },
+    {
+      "epoch": 0.158030036530916,
+      "grad_norm": 0.450712114572525,
+      "learning_rate": 5.2569882777276827e-05,
+      "loss": 1.469,
+      "step": 584
+    },
+    {
+      "epoch": 0.15857123528615885,
+      "grad_norm": 0.5500516891479492,
+      "learning_rate": 5.27502254283138e-05,
+      "loss": 1.4603,
+      "step": 586
+    },
+    {
+      "epoch": 0.1591124340414017,
+      "grad_norm": 0.457157164812088,
+      "learning_rate": 5.2930568079350775e-05,
+      "loss": 1.4785,
+      "step": 588
+    },
+    {
+      "epoch": 0.15965363279664457,
+      "grad_norm": 0.49750396609306335,
+      "learning_rate": 5.311091073038774e-05,
+      "loss": 1.4603,
+      "step": 590
+    },
+    {
+      "epoch": 0.16019483155188743,
+      "grad_norm": 0.5720525979995728,
+      "learning_rate": 5.329125338142471e-05,
+      "loss": 1.4753,
+      "step": 592
+    },
+    {
+      "epoch": 0.1607360303071303,
+      "grad_norm": 0.4425548315048218,
+      "learning_rate": 5.347159603246168e-05,
+      "loss": 1.462,
+      "step": 594
+    },
+    {
+      "epoch": 0.16127722906237316,
+      "grad_norm": 0.5064132809638977,
+      "learning_rate": 5.3651938683498645e-05,
+      "loss": 1.4596,
+      "step": 596
+    },
+    {
+      "epoch": 0.16181842781761602,
+      "grad_norm": 0.518460750579834,
+      "learning_rate": 5.383228133453562e-05,
+      "loss": 1.4763,
+      "step": 598
+    },
+    {
+      "epoch": 0.16235962657285888,
+      "grad_norm": 0.4613576829433441,
+      "learning_rate": 5.401262398557259e-05,
+      "loss": 1.4487,
+      "step": 600
+    },
+    {
+      "epoch": 0.16290082532810174,
+      "grad_norm": 0.7046213746070862,
+      "learning_rate": 5.419296663660957e-05,
+      "loss": 1.472,
+      "step": 602
+    },
+    {
+      "epoch": 0.1634420240833446,
+      "grad_norm": 0.6164196133613586,
+      "learning_rate": 5.4373309287646535e-05,
+      "loss": 1.4424,
+      "step": 604
+    },
+    {
+      "epoch": 0.16398322283858746,
+      "grad_norm": 0.5106020569801331,
+      "learning_rate": 5.45536519386835e-05,
+      "loss": 1.4567,
+      "step": 606
+    },
+    {
+      "epoch": 0.16452442159383032,
+      "grad_norm": 0.4291236400604248,
+      "learning_rate": 5.473399458972047e-05,
+      "loss": 1.4514,
+      "step": 608
+    },
+    {
+      "epoch": 0.16506562034907318,
+      "grad_norm": 0.46577414870262146,
+      "learning_rate": 5.491433724075744e-05,
+      "loss": 1.4408,
+      "step": 610
+    },
+    {
+      "epoch": 0.16560681910431607,
+      "grad_norm": 0.4729917049407959,
+      "learning_rate": 5.509467989179441e-05,
+      "loss": 1.4493,
+      "step": 612
+    },
+    {
+      "epoch": 0.16614801785955893,
+      "grad_norm": 0.4651925563812256,
+      "learning_rate": 5.527502254283138e-05,
+      "loss": 1.465,
+      "step": 614
+    },
+    {
+      "epoch": 0.1666892166148018,
+      "grad_norm": 0.4756859540939331,
+      "learning_rate": 5.545536519386835e-05,
+      "loss": 1.4641,
+      "step": 616
+    },
+    {
+      "epoch": 0.16723041537004465,
+      "grad_norm": 0.42555975914001465,
+      "learning_rate": 5.563570784490533e-05,
+      "loss": 1.4569,
+      "step": 618
+    },
+    {
+      "epoch": 0.16777161412528752,
+      "grad_norm": 0.5162522196769714,
+      "learning_rate": 5.5816050495942296e-05,
+      "loss": 1.4344,
+      "step": 620
+    },
+    {
+      "epoch": 0.16831281288053038,
+      "grad_norm": 0.5867063999176025,
+      "learning_rate": 5.599639314697926e-05,
+      "loss": 1.4647,
+      "step": 622
+    },
+    {
+      "epoch": 0.16885401163577324,
+      "grad_norm": 0.6629165410995483,
+      "learning_rate": 5.617673579801623e-05,
+      "loss": 1.473,
+      "step": 624
+    },
+    {
+      "epoch": 0.1693952103910161,
+      "grad_norm": 0.5905330777168274,
+      "learning_rate": 5.6357078449053205e-05,
+      "loss": 1.4459,
+      "step": 626
+    },
+    {
+      "epoch": 0.16993640914625896,
+      "grad_norm": 0.7457858324050903,
+      "learning_rate": 5.653742110009017e-05,
+      "loss": 1.4603,
+      "step": 628
+    },
+    {
+      "epoch": 0.17047760790150182,
+      "grad_norm": 0.5977684855461121,
+      "learning_rate": 5.671776375112714e-05,
+      "loss": 1.4621,
+      "step": 630
+    },
+    {
+      "epoch": 0.17101880665674468,
+      "grad_norm": 0.7097992897033691,
+      "learning_rate": 5.689810640216412e-05,
+      "loss": 1.4646,
+      "step": 632
+    },
+    {
+      "epoch": 0.17156000541198754,
+      "grad_norm": 0.5895450711250305,
+      "learning_rate": 5.707844905320109e-05,
+      "loss": 1.4338,
+      "step": 634
+    },
+    {
+      "epoch": 0.1721012041672304,
+      "grad_norm": 0.576877772808075,
+      "learning_rate": 5.7258791704238056e-05,
+      "loss": 1.4666,
+      "step": 636
+    },
+    {
+      "epoch": 0.17264240292247326,
+      "grad_norm": 0.541110098361969,
+      "learning_rate": 5.7439134355275024e-05,
+      "loss": 1.4624,
+      "step": 638
+    },
+    {
+      "epoch": 0.17318360167771615,
+      "grad_norm": 0.5172320604324341,
+      "learning_rate": 5.7619477006312e-05,
+      "loss": 1.473,
+      "step": 640
+    },
+    {
+      "epoch": 0.17372480043295901,
+      "grad_norm": 0.47511357069015503,
+      "learning_rate": 5.7799819657348965e-05,
+      "loss": 1.446,
+      "step": 642
+    },
+    {
+      "epoch": 0.17426599918820188,
+      "grad_norm": 0.48614808917045593,
+      "learning_rate": 5.798016230838593e-05,
+      "loss": 1.4394,
+      "step": 644
+    },
+    {
+      "epoch": 0.17480719794344474,
+      "grad_norm": 0.4435577094554901,
+      "learning_rate": 5.81605049594229e-05,
+      "loss": 1.43,
+      "step": 646
+    },
+    {
+      "epoch": 0.1753483966986876,
+      "grad_norm": 0.4458653926849365,
+      "learning_rate": 5.834084761045988e-05,
+      "loss": 1.46,
+      "step": 648
+    },
+    {
+      "epoch": 0.17588959545393046,
+      "grad_norm": 0.40675726532936096,
+      "learning_rate": 5.852119026149685e-05,
+      "loss": 1.4565,
+      "step": 650
+    },
+    {
+      "epoch": 0.17643079420917332,
+      "grad_norm": 0.4132504165172577,
+      "learning_rate": 5.8701532912533817e-05,
+      "loss": 1.4522,
+      "step": 652
+    },
+    {
+      "epoch": 0.17697199296441618,
+      "grad_norm": 0.40881386399269104,
+      "learning_rate": 5.888187556357079e-05,
+      "loss": 1.4232,
+      "step": 654
+    },
+    {
+      "epoch": 0.17751319171965904,
+      "grad_norm": 0.40527868270874023,
+      "learning_rate": 5.906221821460776e-05,
+      "loss": 1.441,
+      "step": 656
+    },
+    {
+      "epoch": 0.1780543904749019,
+      "grad_norm": 0.40227004885673523,
+      "learning_rate": 5.9242560865644726e-05,
+      "loss": 1.4259,
+      "step": 658
+    },
+    {
+      "epoch": 0.17859558923014476,
+      "grad_norm": 0.4043656289577484,
+      "learning_rate": 5.942290351668169e-05,
+      "loss": 1.4298,
+      "step": 660
+    },
+    {
+      "epoch": 0.17913678798538762,
+      "grad_norm": 0.4288482666015625,
+      "learning_rate": 5.9603246167718674e-05,
+      "loss": 1.4439,
+      "step": 662
+    },
+    {
+      "epoch": 0.17967798674063049,
+      "grad_norm": 0.4385060966014862,
+      "learning_rate": 5.978358881875564e-05,
+      "loss": 1.4237,
+      "step": 664
+    },
+    {
+      "epoch": 0.18021918549587335,
+      "grad_norm": 0.396980345249176,
+      "learning_rate": 5.996393146979261e-05,
+      "loss": 1.4174,
+      "step": 666
+    },
+    {
+      "epoch": 0.18076038425111624,
+      "grad_norm": 0.4060603678226471,
+      "learning_rate": 6.014427412082958e-05,
+      "loss": 1.4479,
+      "step": 668
+    },
+    {
+      "epoch": 0.1813015830063591,
+      "grad_norm": 0.4485025703907013,
+      "learning_rate": 6.032461677186655e-05,
+      "loss": 1.4493,
+      "step": 670
+    },
+    {
+      "epoch": 0.18184278176160196,
+      "grad_norm": 0.44034305214881897,
+      "learning_rate": 6.050495942290352e-05,
+      "loss": 1.4461,
+      "step": 672
+    },
+    {
+      "epoch": 0.18238398051684482,
+      "grad_norm": 0.418074369430542,
+      "learning_rate": 6.0685302073940486e-05,
+      "loss": 1.4287,
+      "step": 674
+    },
+    {
+      "epoch": 0.18292517927208768,
+      "grad_norm": 0.41937318444252014,
+      "learning_rate": 6.0865644724977454e-05,
+      "loss": 1.4338,
+      "step": 676
+    },
+    {
+      "epoch": 0.18346637802733054,
+      "grad_norm": 0.4103530943393707,
+      "learning_rate": 6.104598737601444e-05,
+      "loss": 1.4391,
+      "step": 678
+    },
+    {
+      "epoch": 0.1840075767825734,
+      "grad_norm": 0.4066039025783539,
+      "learning_rate": 6.122633002705141e-05,
+      "loss": 1.4357,
+      "step": 680
+    },
+    {
+      "epoch": 0.18454877553781626,
+      "grad_norm": 0.36903437972068787,
+      "learning_rate": 6.140667267808838e-05,
+      "loss": 1.4111,
+      "step": 682
+    },
+    {
+      "epoch": 0.18508997429305912,
+      "grad_norm": 0.37125757336616516,
+      "learning_rate": 6.158701532912534e-05,
+      "loss": 1.4233,
+      "step": 684
+    },
+    {
+      "epoch": 0.18563117304830198,
+      "grad_norm": 0.44102513790130615,
+      "learning_rate": 6.176735798016231e-05,
+      "loss": 1.4437,
+      "step": 686
+    },
+    {
+      "epoch": 0.18617237180354484,
+      "grad_norm": 0.4337277114391327,
+      "learning_rate": 6.194770063119928e-05,
+      "loss": 1.4425,
+      "step": 688
+    },
+    {
+      "epoch": 0.1867135705587877,
+      "grad_norm": 0.37394315004348755,
+      "learning_rate": 6.212804328223625e-05,
+      "loss": 1.4452,
+      "step": 690
+    },
+    {
+      "epoch": 0.18725476931403057,
+      "grad_norm": 0.41764944791793823,
+      "learning_rate": 6.230838593327321e-05,
+      "loss": 1.4535,
+      "step": 692
+    },
+    {
+      "epoch": 0.18779596806927343,
+      "grad_norm": 0.4214741289615631,
+      "learning_rate": 6.24887285843102e-05,
+      "loss": 1.4391,
+      "step": 694
+    },
+    {
+      "epoch": 0.18833716682451632,
+      "grad_norm": 0.4159027338027954,
+      "learning_rate": 6.266907123534716e-05,
+      "loss": 1.4197,
+      "step": 696
+    },
+    {
+      "epoch": 0.18887836557975918,
+      "grad_norm": 0.38865673542022705,
+      "learning_rate": 6.284941388638413e-05,
+      "loss": 1.4329,
+      "step": 698
+    },
+    {
+      "epoch": 0.18941956433500204,
+      "grad_norm": 0.43646490573883057,
+      "learning_rate": 6.30297565374211e-05,
+      "loss": 1.4147,
+      "step": 700
+    },
+    {
+      "epoch": 0.1899607630902449,
+      "grad_norm": 0.41997334361076355,
+      "learning_rate": 6.321009918845807e-05,
+      "loss": 1.4275,
+      "step": 702
+    },
+    {
+      "epoch": 0.19050196184548776,
+      "grad_norm": 0.38556602597236633,
+      "learning_rate": 6.339044183949505e-05,
+      "loss": 1.4258,
+      "step": 704
+    },
+    {
+      "epoch": 0.19104316060073062,
+      "grad_norm": 0.42955082654953003,
+      "learning_rate": 6.357078449053201e-05,
+      "loss": 1.4201,
+      "step": 706
+    },
+    {
+      "epoch": 0.19158435935597348,
+      "grad_norm": 0.3844427764415741,
+      "learning_rate": 6.3751127141569e-05,
+      "loss": 1.4448,
+      "step": 708
+    },
+    {
+      "epoch": 0.19212555811121634,
+      "grad_norm": 0.4312956929206848,
+      "learning_rate": 6.393146979260596e-05,
+      "loss": 1.4051,
+      "step": 710
+    },
+    {
+      "epoch": 0.1926667568664592,
+      "grad_norm": 0.4556865394115448,
+      "learning_rate": 6.411181244364293e-05,
+      "loss": 1.4305,
+      "step": 712
+    },
+    {
+      "epoch": 0.19320795562170207,
+      "grad_norm": 0.37053731083869934,
+      "learning_rate": 6.42921550946799e-05,
+      "loss": 1.4301,
+      "step": 714
+    },
+    {
+      "epoch": 0.19374915437694493,
+      "grad_norm": 0.3996010720729828,
+      "learning_rate": 6.447249774571686e-05,
+      "loss": 1.4282,
+      "step": 716
+    },
+    {
+      "epoch": 0.1942903531321878,
+      "grad_norm": 0.37610816955566406,
+      "learning_rate": 6.465284039675383e-05,
+      "loss": 1.4277,
+      "step": 718
+    },
+    {
+      "epoch": 0.19483155188743065,
+      "grad_norm": 0.3677166998386383,
+      "learning_rate": 6.48331830477908e-05,
+      "loss": 1.4029,
+      "step": 720
+    },
+    {
+      "epoch": 0.1953727506426735,
+      "grad_norm": 0.3841564357280731,
+      "learning_rate": 6.501352569882777e-05,
+      "loss": 1.4144,
+      "step": 722
+    },
+    {
+      "epoch": 0.1959139493979164,
+      "grad_norm": 0.3687719404697418,
+      "learning_rate": 6.519386834986475e-05,
+      "loss": 1.4079,
+      "step": 724
+    },
+    {
+      "epoch": 0.19645514815315926,
+      "grad_norm": 0.38350847363471985,
+      "learning_rate": 6.537421100090172e-05,
+      "loss": 1.4269,
+      "step": 726
+    },
+    {
+      "epoch": 0.19699634690840212,
+      "grad_norm": 0.39060813188552856,
+      "learning_rate": 6.555455365193868e-05,
+      "loss": 1.4265,
+      "step": 728
+    },
+    {
+      "epoch": 0.19753754566364498,
+      "grad_norm": 0.36068469285964966,
+      "learning_rate": 6.573489630297565e-05,
+      "loss": 1.4325,
+      "step": 730
+    },
+    {
+      "epoch": 0.19807874441888784,
+      "grad_norm": 0.41185086965560913,
+      "learning_rate": 6.591523895401263e-05,
+      "loss": 1.4348,
+      "step": 732
+    },
+    {
+      "epoch": 0.1986199431741307,
+      "grad_norm": 0.4441224932670593,
+      "learning_rate": 6.60955816050496e-05,
+      "loss": 1.4103,
+      "step": 734
+    },
+    {
+      "epoch": 0.19916114192937356,
+      "grad_norm": 0.3727317452430725,
+      "learning_rate": 6.627592425608657e-05,
+      "loss": 1.4188,
+      "step": 736
+    },
+    {
+      "epoch": 0.19970234068461643,
+      "grad_norm": 0.394972562789917,
+      "learning_rate": 6.645626690712355e-05,
+      "loss": 1.4095,
+      "step": 738
+    },
+    {
+      "epoch": 0.20024353943985929,
+      "grad_norm": 0.40716880559921265,
+      "learning_rate": 6.663660955816052e-05,
+      "loss": 1.4127,
+      "step": 740
+    },
+    {
+      "epoch": 0.20078473819510215,
+      "grad_norm": 0.4156644344329834,
+      "learning_rate": 6.681695220919748e-05,
+      "loss": 1.4189,
+      "step": 742
+    },
+    {
+      "epoch": 0.201325936950345,
+      "grad_norm": 0.3787958323955536,
+      "learning_rate": 6.699729486023445e-05,
+      "loss": 1.4221,
+      "step": 744
+    },
+    {
+      "epoch": 0.20186713570558787,
+      "grad_norm": 0.42427608370780945,
+      "learning_rate": 6.717763751127142e-05,
+      "loss": 1.4192,
+      "step": 746
+    },
+    {
+      "epoch": 0.20240833446083073,
+      "grad_norm": 0.4778277277946472,
+      "learning_rate": 6.735798016230839e-05,
+      "loss": 1.4024,
+      "step": 748
+    },
+    {
+      "epoch": 0.2029495332160736,
+      "grad_norm": 0.44801151752471924,
+      "learning_rate": 6.753832281334535e-05,
+      "loss": 1.4222,
+      "step": 750
+    },
+    {
+      "epoch": 0.20349073197131648,
+      "grad_norm": 0.46737611293792725,
+      "learning_rate": 6.771866546438232e-05,
+      "loss": 1.4117,
+      "step": 752
+    },
+    {
+      "epoch": 0.20403193072655934,
+      "grad_norm": 0.4184872806072235,
+      "learning_rate": 6.78990081154193e-05,
+      "loss": 1.4066,
+      "step": 754
+    },
+    {
+      "epoch": 0.2045731294818022,
+      "grad_norm": 0.40458211302757263,
+      "learning_rate": 6.807935076645627e-05,
+      "loss": 1.4274,
+      "step": 756
+    },
+    {
+      "epoch": 0.20511432823704506,
+      "grad_norm": 0.43926185369491577,
+      "learning_rate": 6.825969341749324e-05,
+      "loss": 1.4231,
+      "step": 758
+    },
+    {
+      "epoch": 0.20565552699228792,
+      "grad_norm": 0.4434867203235626,
+      "learning_rate": 6.844003606853022e-05,
+      "loss": 1.4121,
+      "step": 760
+    },
+    {
+      "epoch": 0.20619672574753078,
+      "grad_norm": 0.4500143826007843,
+      "learning_rate": 6.862037871956719e-05,
+      "loss": 1.4179,
+      "step": 762
+    },
+    {
+      "epoch": 0.20673792450277365,
+      "grad_norm": 0.45456650853157043,
+      "learning_rate": 6.880072137060415e-05,
+      "loss": 1.3912,
+      "step": 764
+    },
+    {
+      "epoch": 0.2072791232580165,
+      "grad_norm": 0.4214187264442444,
+      "learning_rate": 6.898106402164112e-05,
+      "loss": 1.3962,
+      "step": 766
+    },
+    {
+      "epoch": 0.20782032201325937,
+      "grad_norm": 0.427682101726532,
+      "learning_rate": 6.916140667267809e-05,
+      "loss": 1.4316,
+      "step": 768
+    },
+    {
+      "epoch": 0.20836152076850223,
+      "grad_norm": 0.44491469860076904,
+      "learning_rate": 6.934174932371507e-05,
+      "loss": 1.4218,
+      "step": 770
+    },
+    {
+      "epoch": 0.2089027195237451,
+      "grad_norm": 0.42736080288887024,
+      "learning_rate": 6.952209197475204e-05,
+      "loss": 1.3931,
+      "step": 772
+    },
+    {
+      "epoch": 0.20944391827898795,
+      "grad_norm": 0.4041571021080017,
+      "learning_rate": 6.9702434625789e-05,
+      "loss": 1.4201,
+      "step": 774
+    },
+    {
+      "epoch": 0.2099851170342308,
+      "grad_norm": 0.4250961244106293,
+      "learning_rate": 6.988277727682597e-05,
+      "loss": 1.4299,
+      "step": 776
+    },
+    {
+      "epoch": 0.21052631578947367,
+      "grad_norm": 0.4335261881351471,
+      "learning_rate": 7.006311992786294e-05,
+      "loss": 1.4125,
+      "step": 778
+    },
+    {
+      "epoch": 0.21106751454471653,
+      "grad_norm": 0.42000851035118103,
+      "learning_rate": 7.02434625788999e-05,
+      "loss": 1.3969,
+      "step": 780
+    },
+    {
+      "epoch": 0.21160871329995942,
+      "grad_norm": 0.38111838698387146,
+      "learning_rate": 7.042380522993687e-05,
+      "loss": 1.3795,
+      "step": 782
+    },
+    {
+      "epoch": 0.21214991205520228,
+      "grad_norm": 0.38366812467575073,
+      "learning_rate": 7.060414788097385e-05,
+      "loss": 1.4041,
+      "step": 784
+    },
+    {
+      "epoch": 0.21269111081044514,
+      "grad_norm": 0.4334602355957031,
+      "learning_rate": 7.078449053201082e-05,
+      "loss": 1.415,
+      "step": 786
+    },
+    {
+      "epoch": 0.213232309565688,
+      "grad_norm": 0.40296411514282227,
+      "learning_rate": 7.096483318304779e-05,
+      "loss": 1.4052,
+      "step": 788
+    },
+    {
+      "epoch": 0.21377350832093087,
+      "grad_norm": 0.4197232723236084,
+      "learning_rate": 7.114517583408477e-05,
+      "loss": 1.4205,
+      "step": 790
+    },
+    {
+      "epoch": 0.21431470707617373,
+      "grad_norm": 0.40287715196609497,
+      "learning_rate": 7.132551848512174e-05,
+      "loss": 1.4047,
+      "step": 792
+    },
+    {
+      "epoch": 0.2148559058314166,
+      "grad_norm": 0.37324196100234985,
+      "learning_rate": 7.15058611361587e-05,
+      "loss": 1.4398,
+      "step": 794
+    },
+    {
+      "epoch": 0.21539710458665945,
+      "grad_norm": 0.4409985840320587,
+      "learning_rate": 7.168620378719567e-05,
+      "loss": 1.3873,
+      "step": 796
+    },
+    {
+      "epoch": 0.2159383033419023,
+      "grad_norm": 0.41441893577575684,
+      "learning_rate": 7.186654643823264e-05,
+      "loss": 1.4174,
+      "step": 798
+    },
+    {
+      "epoch": 0.21647950209714517,
+      "grad_norm": 0.4271719455718994,
+      "learning_rate": 7.204688908926962e-05,
+      "loss": 1.3987,
+      "step": 800
+    },
+    {
+      "epoch": 0.21702070085238803,
+      "grad_norm": 0.4969992935657501,
+      "learning_rate": 7.222723174030659e-05,
+      "loss": 1.4049,
+      "step": 802
+    },
+    {
+      "epoch": 0.2175618996076309,
+      "grad_norm": 0.45711180567741394,
+      "learning_rate": 7.240757439134356e-05,
+      "loss": 1.4061,
+      "step": 804
+    },
+    {
+      "epoch": 0.21810309836287375,
+      "grad_norm": 0.4479979872703552,
+      "learning_rate": 7.258791704238052e-05,
+      "loss": 1.4049,
+      "step": 806
+    },
+    {
+      "epoch": 0.21864429711811662,
+      "grad_norm": 0.4708006978034973,
+      "learning_rate": 7.276825969341749e-05,
+      "loss": 1.3971,
+      "step": 808
+    },
+    {
+      "epoch": 0.2191854958733595,
+      "grad_norm": 0.4387456774711609,
+      "learning_rate": 7.294860234445446e-05,
+      "loss": 1.4272,
+      "step": 810
+    },
+    {
+      "epoch": 0.21972669462860236,
+      "grad_norm": 0.5285756587982178,
+      "learning_rate": 7.312894499549143e-05,
+      "loss": 1.3902,
+      "step": 812
+    },
+    {
+      "epoch": 0.22026789338384523,
+      "grad_norm": 0.5111876726150513,
+      "learning_rate": 7.330928764652841e-05,
+      "loss": 1.4176,
+      "step": 814
+    },
+    {
+      "epoch": 0.2208090921390881,
+      "grad_norm": 0.4643821716308594,
+      "learning_rate": 7.348963029756538e-05,
+      "loss": 1.4216,
+      "step": 816
+    },
+    {
+      "epoch": 0.22135029089433095,
+      "grad_norm": 0.5162214040756226,
+      "learning_rate": 7.366997294860236e-05,
+      "loss": 1.4025,
+      "step": 818
+    },
+    {
+      "epoch": 0.2218914896495738,
+      "grad_norm": 0.4296860992908478,
+      "learning_rate": 7.385031559963932e-05,
+      "loss": 1.3919,
+      "step": 820
+    },
+    {
+      "epoch": 0.22243268840481667,
+      "grad_norm": 0.4449775815010071,
+      "learning_rate": 7.403065825067629e-05,
+      "loss": 1.4002,
+      "step": 822
+    },
+    {
+      "epoch": 0.22297388716005953,
+      "grad_norm": 0.39713212847709656,
+      "learning_rate": 7.421100090171326e-05,
+      "loss": 1.4012,
+      "step": 824
+    },
+    {
+      "epoch": 0.2235150859153024,
+      "grad_norm": 0.41655346751213074,
+      "learning_rate": 7.439134355275023e-05,
+      "loss": 1.4155,
+      "step": 826
+    },
+    {
+      "epoch": 0.22405628467054525,
+      "grad_norm": 0.3751365542411804,
+      "learning_rate": 7.45716862037872e-05,
+      "loss": 1.4021,
+      "step": 828
+    },
+    {
+      "epoch": 0.2245974834257881,
+      "grad_norm": 0.41483408212661743,
+      "learning_rate": 7.475202885482417e-05,
+      "loss": 1.4207,
+      "step": 830
+    },
+    {
+      "epoch": 0.22513868218103097,
+      "grad_norm": 0.397360235452652,
+      "learning_rate": 7.493237150586114e-05,
+      "loss": 1.392,
+      "step": 832
+    },
+    {
+      "epoch": 0.22567988093627384,
+      "grad_norm": 0.3874877691268921,
+      "learning_rate": 7.511271415689811e-05,
+      "loss": 1.4143,
+      "step": 834
+    },
+    {
+      "epoch": 0.2262210796915167,
+      "grad_norm": 0.4382254481315613,
+      "learning_rate": 7.529305680793508e-05,
+      "loss": 1.4109,
+      "step": 836
+    },
+    {
+      "epoch": 0.22676227844675959,
+      "grad_norm": 0.3728530704975128,
+      "learning_rate": 7.547339945897204e-05,
+      "loss": 1.4215,
+      "step": 838
+    },
+    {
+      "epoch": 0.22730347720200245,
+      "grad_norm": 0.41155338287353516,
+      "learning_rate": 7.565374211000901e-05,
+      "loss": 1.3963,
+      "step": 840
+    },
+    {
+      "epoch": 0.2278446759572453,
+      "grad_norm": 0.3550320267677307,
+      "learning_rate": 7.5834084761046e-05,
+      "loss": 1.3998,
+      "step": 842
+    },
+    {
+      "epoch": 0.22838587471248817,
+      "grad_norm": 0.3858035206794739,
+      "learning_rate": 7.601442741208296e-05,
+      "loss": 1.387,
+      "step": 844
+    },
+    {
+      "epoch": 0.22892707346773103,
+      "grad_norm": 0.38636457920074463,
+      "learning_rate": 7.619477006311994e-05,
+      "loss": 1.387,
+      "step": 846
+    },
+    {
+      "epoch": 0.2294682722229739,
+      "grad_norm": 0.41915518045425415,
+      "learning_rate": 7.637511271415691e-05,
+      "loss": 1.3917,
+      "step": 848
+    },
+    {
+      "epoch": 0.23000947097821675,
+      "grad_norm": 0.35796865820884705,
+      "learning_rate": 7.655545536519388e-05,
+      "loss": 1.406,
+      "step": 850
+    },
+    {
+      "epoch": 0.2305506697334596,
+      "grad_norm": 0.35221853852272034,
+      "learning_rate": 7.673579801623084e-05,
+      "loss": 1.3892,
+      "step": 852
+    },
+    {
+      "epoch": 0.23109186848870247,
+      "grad_norm": 0.3815077245235443,
+      "learning_rate": 7.691614066726781e-05,
+      "loss": 1.3845,
+      "step": 854
+    },
+    {
+      "epoch": 0.23163306724394533,
+      "grad_norm": 0.3554491400718689,
+      "learning_rate": 7.709648331830478e-05,
+      "loss": 1.3644,
+      "step": 856
+    },
+    {
+      "epoch": 0.2321742659991882,
+      "grad_norm": 0.3762814998626709,
+      "learning_rate": 7.727682596934175e-05,
+      "loss": 1.3976,
+      "step": 858
+    },
+    {
+      "epoch": 0.23271546475443106,
+      "grad_norm": 0.34575173258781433,
+      "learning_rate": 7.745716862037873e-05,
+      "loss": 1.3925,
+      "step": 860
+    },
+    {
+      "epoch": 0.23325666350967392,
+      "grad_norm": 0.37864556908607483,
+      "learning_rate": 7.76375112714157e-05,
+      "loss": 1.3993,
+      "step": 862
+    },
+    {
+      "epoch": 0.23379786226491678,
+      "grad_norm": 0.34448474645614624,
+      "learning_rate": 7.781785392245266e-05,
+      "loss": 1.3855,
+      "step": 864
+    },
+    {
+      "epoch": 0.23433906102015967,
+      "grad_norm": 0.40932390093803406,
+      "learning_rate": 7.799819657348963e-05,
+      "loss": 1.395,
+      "step": 866
+    },
+    {
+      "epoch": 0.23488025977540253,
+      "grad_norm": 0.3737650513648987,
+      "learning_rate": 7.81785392245266e-05,
+      "loss": 1.3918,
+      "step": 868
+    },
+    {
+      "epoch": 0.2354214585306454,
+      "grad_norm": 0.42988118529319763,
+      "learning_rate": 7.835888187556357e-05,
+      "loss": 1.3837,
+      "step": 870
+    },
+    {
+      "epoch": 0.23596265728588825,
+      "grad_norm": 0.3865496814250946,
+      "learning_rate": 7.853922452660055e-05,
+      "loss": 1.3976,
+      "step": 872
+    },
+    {
+      "epoch": 0.2365038560411311,
+      "grad_norm": 0.3682670295238495,
+      "learning_rate": 7.871956717763751e-05,
+      "loss": 1.3792,
+      "step": 874
+    },
+    {
+      "epoch": 0.23704505479637397,
+      "grad_norm": 0.4236462712287903,
+      "learning_rate": 7.88999098286745e-05,
+      "loss": 1.4032,
+      "step": 876
+    },
+    {
+      "epoch": 0.23758625355161683,
+      "grad_norm": 0.3742213249206543,
+      "learning_rate": 7.908025247971146e-05,
+      "loss": 1.3709,
+      "step": 878
+    },
+    {
+      "epoch": 0.2381274523068597,
+      "grad_norm": 0.38234424591064453,
+      "learning_rate": 7.926059513074843e-05,
+      "loss": 1.3862,
+      "step": 880
+    },
+    {
+      "epoch": 0.23866865106210255,
+      "grad_norm": 0.37414151430130005,
+      "learning_rate": 7.94409377817854e-05,
+      "loss": 1.3751,
+      "step": 882
+    },
+    {
+      "epoch": 0.23920984981734542,
+      "grad_norm": 0.3838132619857788,
+      "learning_rate": 7.962128043282237e-05,
+      "loss": 1.3805,
+      "step": 884
+    },
+    {
+      "epoch": 0.23975104857258828,
+      "grad_norm": 0.3818622827529907,
+      "learning_rate": 7.980162308385933e-05,
+      "loss": 1.3735,
+      "step": 886
+    },
+    {
+      "epoch": 0.24029224732783114,
+      "grad_norm": 0.38791927695274353,
+      "learning_rate": 7.99819657348963e-05,
+      "loss": 1.3958,
+      "step": 888
+    },
+    {
+      "epoch": 0.240833446083074,
+      "grad_norm": 0.4164978861808777,
+      "learning_rate": 8.016230838593328e-05,
+      "loss": 1.421,
+      "step": 890
+    },
+    {
+      "epoch": 0.24137464483831686,
+      "grad_norm": 0.3721414804458618,
+      "learning_rate": 8.034265103697025e-05,
+      "loss": 1.3977,
+      "step": 892
+    },
+    {
+      "epoch": 0.24191584359355975,
+      "grad_norm": 0.37698984146118164,
+      "learning_rate": 8.052299368800722e-05,
+      "loss": 1.3854,
+      "step": 894
+    },
+    {
+      "epoch": 0.2424570423488026,
+      "grad_norm": 0.3553116023540497,
+      "learning_rate": 8.070333633904418e-05,
+      "loss": 1.3925,
+      "step": 896
+    },
+    {
+      "epoch": 0.24299824110404547,
+      "grad_norm": 0.37809059023857117,
+      "learning_rate": 8.088367899008115e-05,
+      "loss": 1.368,
+      "step": 898
+    },
+    {
+      "epoch": 0.24353943985928833,
+      "grad_norm": 0.3835943043231964,
+      "learning_rate": 8.106402164111813e-05,
+      "loss": 1.3992,
+      "step": 900
+    },
+    {
+      "epoch": 0.2440806386145312,
+      "grad_norm": 0.4013379216194153,
+      "learning_rate": 8.12443642921551e-05,
+      "loss": 1.3912,
+      "step": 902
+    },
+    {
+      "epoch": 0.24462183736977405,
+      "grad_norm": 0.37845560908317566,
+      "learning_rate": 8.142470694319207e-05,
+      "loss": 1.3934,
+      "step": 904
+    },
+    {
+      "epoch": 0.24516303612501691,
+      "grad_norm": 0.39762255549430847,
+      "learning_rate": 8.160504959422905e-05,
+      "loss": 1.3782,
+      "step": 906
+    },
+    {
+      "epoch": 0.24570423488025978,
+      "grad_norm": 0.36652496457099915,
+      "learning_rate": 8.178539224526602e-05,
+      "loss": 1.3787,
+      "step": 908
+    },
+    {
+      "epoch": 0.24624543363550264,
+      "grad_norm": 0.39953047037124634,
+      "learning_rate": 8.196573489630298e-05,
+      "loss": 1.3752,
+      "step": 910
+    },
+    {
+      "epoch": 0.2467866323907455,
+      "grad_norm": 0.35875022411346436,
+      "learning_rate": 8.214607754733995e-05,
+      "loss": 1.3768,
+      "step": 912
+    },
+    {
+      "epoch": 0.24732783114598836,
+      "grad_norm": 0.3617067337036133,
+      "learning_rate": 8.232642019837692e-05,
+      "loss": 1.3859,
+      "step": 914
+    },
+    {
+      "epoch": 0.24786902990123122,
+      "grad_norm": 0.38250839710235596,
+      "learning_rate": 8.250676284941389e-05,
+      "loss": 1.3897,
+      "step": 916
+    },
+    {
+      "epoch": 0.24841022865647408,
+      "grad_norm": 0.3404116928577423,
+      "learning_rate": 8.268710550045085e-05,
+      "loss": 1.3933,
+      "step": 918
+    },
+    {
+      "epoch": 0.24895142741171694,
+      "grad_norm": 0.3547706604003906,
+      "learning_rate": 8.286744815148782e-05,
+      "loss": 1.3787,
+      "step": 920
+    },
+    {
+      "epoch": 0.2494926261669598,
+      "grad_norm": 0.32752275466918945,
+      "learning_rate": 8.30477908025248e-05,
+      "loss": 1.3905,
+      "step": 922
+    },
+    {
+      "epoch": 0.25003382492220266,
+      "grad_norm": 0.3413980007171631,
+      "learning_rate": 8.322813345356177e-05,
+      "loss": 1.385,
+      "step": 924
+    },
+    {
+      "epoch": 0.25057502367744555,
+      "grad_norm": 0.5574982762336731,
+      "learning_rate": 8.340847610459874e-05,
+      "loss": 1.3869,
+      "step": 926
+    },
+    {
+      "epoch": 0.2511162224326884,
+      "grad_norm": 0.41128844022750854,
+      "learning_rate": 8.358881875563572e-05,
+      "loss": 1.3583,
+      "step": 928
+    },
+    {
+      "epoch": 0.2516574211879313,
+      "grad_norm": 0.3476073145866394,
+      "learning_rate": 8.376916140667269e-05,
+      "loss": 1.3832,
+      "step": 930
+    },
+    {
+      "epoch": 0.2521986199431741,
+      "grad_norm": 0.34838998317718506,
+      "learning_rate": 8.394950405770965e-05,
+      "loss": 1.3748,
+      "step": 932
+    },
+    {
+      "epoch": 0.252739818698417,
+      "grad_norm": 0.3552824556827545,
+      "learning_rate": 8.412984670874662e-05,
+      "loss": 1.3936,
+      "step": 934
+    },
+    {
+      "epoch": 0.25328101745365983,
+      "grad_norm": 0.34918278455734253,
+      "learning_rate": 8.43101893597836e-05,
+      "loss": 1.3733,
+      "step": 936
+    },
+    {
+      "epoch": 0.2538222162089027,
+      "grad_norm": 0.431455135345459,
+      "learning_rate": 8.449053201082057e-05,
+      "loss": 1.3924,
+      "step": 938
+    },
+    {
+      "epoch": 0.2543634149641456,
+      "grad_norm": 0.37811046838760376,
+      "learning_rate": 8.467087466185754e-05,
+      "loss": 1.3861,
+      "step": 940
+    },
+    {
+      "epoch": 0.25490461371938844,
+      "grad_norm": 0.35659778118133545,
+      "learning_rate": 8.48512173128945e-05,
+      "loss": 1.3736,
+      "step": 942
+    },
+    {
+      "epoch": 0.25544581247463133,
+      "grad_norm": 0.4327319264411926,
+      "learning_rate": 8.503155996393147e-05,
+      "loss": 1.3883,
+      "step": 944
+    },
+    {
+      "epoch": 0.25598701122987416,
+      "grad_norm": 0.39134231209754944,
+      "learning_rate": 8.521190261496844e-05,
+      "loss": 1.3704,
+      "step": 946
+    },
+    {
+      "epoch": 0.25652820998511705,
+      "grad_norm": 0.39573270082473755,
+      "learning_rate": 8.53922452660054e-05,
+      "loss": 1.4047,
+      "step": 948
+    },
+    {
+      "epoch": 0.2570694087403599,
+      "grad_norm": 0.3299993872642517,
+      "learning_rate": 8.557258791704237e-05,
+      "loss": 1.3778,
+      "step": 950
+    },
+    {
+      "epoch": 0.2576106074956028,
+      "grad_norm": 0.3559456765651703,
+      "learning_rate": 8.575293056807936e-05,
+      "loss": 1.3794,
+      "step": 952
+    },
+    {
+      "epoch": 0.2581518062508456,
+      "grad_norm": 0.36347028613090515,
+      "learning_rate": 8.593327321911632e-05,
+      "loss": 1.3817,
+      "step": 954
+    },
+    {
+      "epoch": 0.2586930050060885,
+      "grad_norm": 0.39882585406303406,
+      "learning_rate": 8.611361587015329e-05,
+      "loss": 1.3565,
+      "step": 956
+    },
+    {
+      "epoch": 0.2592342037613313,
+      "grad_norm": 0.3932117223739624,
+      "learning_rate": 8.629395852119027e-05,
+      "loss": 1.396,
+      "step": 958
+    },
+    {
+      "epoch": 0.2597754025165742,
+      "grad_norm": 0.3526294231414795,
+      "learning_rate": 8.647430117222724e-05,
+      "loss": 1.3624,
+      "step": 960
+    },
+    {
+      "epoch": 0.26031660127181705,
+      "grad_norm": 0.3804738223552704,
+      "learning_rate": 8.66546438232642e-05,
+      "loss": 1.3616,
+      "step": 962
+    },
+    {
+      "epoch": 0.26085780002705994,
+      "grad_norm": 0.36557725071907043,
+      "learning_rate": 8.683498647430117e-05,
+      "loss": 1.3997,
+      "step": 964
+    },
+    {
+      "epoch": 0.2613989987823028,
+      "grad_norm": 0.3574380874633789,
+      "learning_rate": 8.701532912533815e-05,
+      "loss": 1.3901,
+      "step": 966
+    },
+    {
+      "epoch": 0.26194019753754566,
+      "grad_norm": 0.4025056064128876,
+      "learning_rate": 8.719567177637512e-05,
+      "loss": 1.3707,
+      "step": 968
+    },
+    {
+      "epoch": 0.26248139629278855,
+      "grad_norm": 0.3687063157558441,
+      "learning_rate": 8.737601442741209e-05,
+      "loss": 1.3679,
+      "step": 970
+    },
+    {
+      "epoch": 0.2630225950480314,
+      "grad_norm": 0.3697878420352936,
+      "learning_rate": 8.755635707844906e-05,
+      "loss": 1.3981,
+      "step": 972
+    },
+    {
+      "epoch": 0.26356379380327427,
+      "grad_norm": 0.34241798520088196,
+      "learning_rate": 8.773669972948602e-05,
+      "loss": 1.3728,
+      "step": 974
+    },
+    {
+      "epoch": 0.2641049925585171,
+      "grad_norm": 0.40002745389938354,
+      "learning_rate": 8.791704238052299e-05,
+      "loss": 1.3732,
+      "step": 976
+    },
+    {
+      "epoch": 0.26464619131376,
+      "grad_norm": 0.42943906784057617,
+      "learning_rate": 8.809738503155996e-05,
+      "loss": 1.3731,
+      "step": 978
+    },
+    {
+      "epoch": 0.2651873900690028,
+      "grad_norm": 0.37437063455581665,
+      "learning_rate": 8.827772768259693e-05,
+      "loss": 1.372,
+      "step": 980
+    },
+    {
+      "epoch": 0.2657285888242457,
+      "grad_norm": 0.3378891944885254,
+      "learning_rate": 8.845807033363391e-05,
+      "loss": 1.3777,
+      "step": 982
+    },
+    {
+      "epoch": 0.26626978757948855,
+      "grad_norm": 0.32884734869003296,
+      "learning_rate": 8.863841298467088e-05,
+      "loss": 1.3639,
+      "step": 984
+    },
+    {
+      "epoch": 0.26681098633473144,
+      "grad_norm": 0.3945903480052948,
+      "learning_rate": 8.881875563570786e-05,
+      "loss": 1.3722,
+      "step": 986
+    },
+    {
+      "epoch": 0.26735218508997427,
+      "grad_norm": 0.39569205045700073,
+      "learning_rate": 8.899909828674482e-05,
+      "loss": 1.376,
+      "step": 988
+    },
+    {
+      "epoch": 0.26789338384521716,
+      "grad_norm": 0.31659135222435,
+      "learning_rate": 8.917944093778179e-05,
+      "loss": 1.3807,
+      "step": 990
+    },
+    {
+      "epoch": 0.26843458260046,
+      "grad_norm": 0.44032666087150574,
+      "learning_rate": 8.935978358881876e-05,
+      "loss": 1.3986,
+      "step": 992
+    },
+    {
+      "epoch": 0.2689757813557029,
+      "grad_norm": 0.3445993661880493,
+      "learning_rate": 8.954012623985573e-05,
+      "loss": 1.3589,
+      "step": 994
+    },
+    {
+      "epoch": 0.26951698011094577,
+      "grad_norm": 0.3693557679653168,
+      "learning_rate": 8.97204688908927e-05,
+      "loss": 1.3593,
+      "step": 996
+    },
+    {
+      "epoch": 0.2700581788661886,
+      "grad_norm": 0.3965442478656769,
+      "learning_rate": 8.990081154192968e-05,
+      "loss": 1.3909,
+      "step": 998
+    },
+    {
+      "epoch": 0.2705993776214315,
+      "grad_norm": 0.4038390815258026,
+      "learning_rate": 9.008115419296664e-05,
+      "loss": 1.3629,
+      "step": 1000
+    },
+    {
+      "epoch": 0.2711405763766743,
+      "grad_norm": 0.36394256353378296,
+      "learning_rate": 9.026149684400361e-05,
+      "loss": 1.3812,
+      "step": 1002
+    },
+    {
+      "epoch": 0.2716817751319172,
+      "grad_norm": 0.4527181386947632,
+      "learning_rate": 9.044183949504058e-05,
+      "loss": 1.3692,
+      "step": 1004
+    },
+    {
+      "epoch": 0.27222297388716005,
+      "grad_norm": 0.37700143456459045,
+      "learning_rate": 9.062218214607755e-05,
+      "loss": 1.3652,
+      "step": 1006
+    },
+    {
+      "epoch": 0.27276417264240294,
+      "grad_norm": 0.45016244053840637,
+      "learning_rate": 9.080252479711451e-05,
+      "loss": 1.3657,
+      "step": 1008
+    },
+    {
+      "epoch": 0.27330537139764577,
+      "grad_norm": 0.42159709334373474,
+      "learning_rate": 9.09828674481515e-05,
+      "loss": 1.3702,
+      "step": 1010
+    },
+    {
+      "epoch": 0.27384657015288866,
+      "grad_norm": 0.3884572982788086,
+      "learning_rate": 9.116321009918846e-05,
+      "loss": 1.3535,
+      "step": 1012
+    },
+    {
+      "epoch": 0.2743877689081315,
+      "grad_norm": 0.37507420778274536,
+      "learning_rate": 9.134355275022544e-05,
+      "loss": 1.3659,
+      "step": 1014
+    },
+    {
+      "epoch": 0.2749289676633744,
+      "grad_norm": 0.35269656777381897,
+      "learning_rate": 9.152389540126241e-05,
+      "loss": 1.3623,
+      "step": 1016
+    },
+    {
+      "epoch": 0.2754701664186172,
+      "grad_norm": 0.3543412387371063,
+      "learning_rate": 9.170423805229938e-05,
+      "loss": 1.3695,
+      "step": 1018
+    },
+    {
+      "epoch": 0.2760113651738601,
+      "grad_norm": 0.3173674941062927,
+      "learning_rate": 9.188458070333635e-05,
+      "loss": 1.3572,
+      "step": 1020
+    },
+    {
+      "epoch": 0.276552563929103,
+      "grad_norm": 0.3729746341705322,
+      "learning_rate": 9.206492335437331e-05,
+      "loss": 1.3888,
+      "step": 1022
+    },
+    {
+      "epoch": 0.2770937626843458,
+      "grad_norm": 0.33210429549217224,
+      "learning_rate": 9.224526600541028e-05,
+      "loss": 1.3395,
+      "step": 1024
+    },
+    {
+      "epoch": 0.2776349614395887,
+      "grad_norm": 0.338366836309433,
+      "learning_rate": 9.242560865644725e-05,
+      "loss": 1.3498,
+      "step": 1026
+    },
+    {
+      "epoch": 0.27817616019483155,
+      "grad_norm": 0.3367864191532135,
+      "learning_rate": 9.260595130748423e-05,
+      "loss": 1.3548,
+      "step": 1028
+    },
+    {
+      "epoch": 0.27871735895007443,
+      "grad_norm": 0.40313002467155457,
+      "learning_rate": 9.27862939585212e-05,
+      "loss": 1.4059,
+      "step": 1030
+    },
+    {
+      "epoch": 0.27925855770531727,
+      "grad_norm": 0.3434394299983978,
+      "learning_rate": 9.296663660955816e-05,
+      "loss": 1.3522,
+      "step": 1032
+    },
+    {
+      "epoch": 0.27979975646056016,
+      "grad_norm": 0.35454580187797546,
+      "learning_rate": 9.314697926059513e-05,
+      "loss": 1.3838,
+      "step": 1034
+    },
+    {
+      "epoch": 0.280340955215803,
+      "grad_norm": 0.3280038833618164,
+      "learning_rate": 9.33273219116321e-05,
+      "loss": 1.3753,
+      "step": 1036
+    },
+    {
+      "epoch": 0.2808821539710459,
+      "grad_norm": 0.4306875169277191,
+      "learning_rate": 9.350766456266907e-05,
+      "loss": 1.3807,
+      "step": 1038
+    },
+    {
+      "epoch": 0.2814233527262887,
+      "grad_norm": 0.3500923812389374,
+      "learning_rate": 9.368800721370605e-05,
+      "loss": 1.36,
+      "step": 1040
+    },
+    {
+      "epoch": 0.2819645514815316,
+      "grad_norm": 0.3702130913734436,
+      "learning_rate": 9.386834986474301e-05,
+      "loss": 1.3919,
+      "step": 1042
+    },
+    {
+      "epoch": 0.28250575023677443,
+      "grad_norm": 0.3651416599750519,
+      "learning_rate": 9.404869251578e-05,
+      "loss": 1.3805,
+      "step": 1044
+    },
+    {
+      "epoch": 0.2830469489920173,
+      "grad_norm": 0.35927796363830566,
+      "learning_rate": 9.422903516681696e-05,
+      "loss": 1.3507,
+      "step": 1046
+    },
+    {
+      "epoch": 0.28358814774726016,
+      "grad_norm": 0.36750975251197815,
+      "learning_rate": 9.440937781785393e-05,
+      "loss": 1.3475,
+      "step": 1048
+    },
+    {
+      "epoch": 0.28412934650250304,
+      "grad_norm": 0.31946998834609985,
+      "learning_rate": 9.45897204688909e-05,
+      "loss": 1.3708,
+      "step": 1050
+    },
+    {
+      "epoch": 0.28467054525774593,
+      "grad_norm": 0.3447932302951813,
+      "learning_rate": 9.477006311992787e-05,
+      "loss": 1.3519,
+      "step": 1052
+    },
+    {
+      "epoch": 0.28521174401298877,
+      "grad_norm": 0.31405511498451233,
+      "learning_rate": 9.495040577096483e-05,
+      "loss": 1.3806,
+      "step": 1054
+    },
+    {
+      "epoch": 0.28575294276823165,
+      "grad_norm": 0.3198442757129669,
+      "learning_rate": 9.51307484220018e-05,
+      "loss": 1.368,
+      "step": 1056
+    },
+    {
+      "epoch": 0.2862941415234745,
+      "grad_norm": 0.33328956365585327,
+      "learning_rate": 9.531109107303878e-05,
+      "loss": 1.3429,
+      "step": 1058
+    },
+    {
+      "epoch": 0.2868353402787174,
+      "grad_norm": 0.29432907700538635,
+      "learning_rate": 9.549143372407575e-05,
+      "loss": 1.3698,
+      "step": 1060
+    },
+    {
+      "epoch": 0.2873765390339602,
+      "grad_norm": 0.3468937575817108,
+      "learning_rate": 9.567177637511272e-05,
+      "loss": 1.356,
+      "step": 1062
+    },
+    {
+      "epoch": 0.2879177377892031,
+      "grad_norm": 0.3619658350944519,
+      "learning_rate": 9.585211902614968e-05,
+      "loss": 1.3596,
+      "step": 1064
+    },
+    {
+      "epoch": 0.28845893654444593,
+      "grad_norm": 0.3384917378425598,
+      "learning_rate": 9.603246167718665e-05,
+      "loss": 1.3693,
+      "step": 1066
+    },
+    {
+      "epoch": 0.2890001352996888,
+      "grad_norm": 0.3724029064178467,
+      "learning_rate": 9.621280432822363e-05,
+      "loss": 1.3639,
+      "step": 1068
+    },
+    {
+      "epoch": 0.28954133405493165,
+      "grad_norm": 0.7029115557670593,
+      "learning_rate": 9.63931469792606e-05,
+      "loss": 1.3557,
+      "step": 1070
+    },
+    {
+      "epoch": 0.29008253281017454,
+      "grad_norm": 0.5529230833053589,
+      "learning_rate": 9.657348963029757e-05,
+      "loss": 1.3657,
+      "step": 1072
+    },
+    {
+      "epoch": 0.2906237315654174,
+      "grad_norm": 0.4254820644855499,
+      "learning_rate": 9.675383228133455e-05,
+      "loss": 1.3633,
+      "step": 1074
+    },
+    {
+      "epoch": 0.29116493032066026,
+      "grad_norm": 0.4930615723133087,
+      "learning_rate": 9.693417493237152e-05,
+      "loss": 1.3714,
+      "step": 1076
+    },
+    {
+      "epoch": 0.2917061290759031,
+      "grad_norm": 0.4455857574939728,
+      "learning_rate": 9.711451758340848e-05,
+      "loss": 1.3615,
+      "step": 1078
+    },
+    {
+      "epoch": 0.292247327831146,
+      "grad_norm": 0.4171796441078186,
+      "learning_rate": 9.729486023444545e-05,
+      "loss": 1.3673,
+      "step": 1080
+    },
+    {
+      "epoch": 0.2927885265863889,
+      "grad_norm": 0.37810683250427246,
+      "learning_rate": 9.747520288548242e-05,
+      "loss": 1.3683,
+      "step": 1082
+    },
+    {
+      "epoch": 0.2933297253416317,
+      "grad_norm": 0.4057900905609131,
+      "learning_rate": 9.765554553651939e-05,
+      "loss": 1.3674,
+      "step": 1084
+    },
+    {
+      "epoch": 0.2938709240968746,
+      "grad_norm": 0.40583640336990356,
+      "learning_rate": 9.783588818755635e-05,
+      "loss": 1.3566,
+      "step": 1086
+    },
+    {
+      "epoch": 0.29441212285211743,
+      "grad_norm": 0.39454150199890137,
+      "learning_rate": 9.801623083859334e-05,
+      "loss": 1.3611,
+      "step": 1088
+    },
+    {
+      "epoch": 0.2949533216073603,
+      "grad_norm": 0.42229679226875305,
+      "learning_rate": 9.81965734896303e-05,
+      "loss": 1.3726,
+      "step": 1090
+    },
+    {
+      "epoch": 0.29549452036260315,
+      "grad_norm": 0.3274170160293579,
+      "learning_rate": 9.837691614066727e-05,
+      "loss": 1.3375,
+      "step": 1092
+    },
+    {
+      "epoch": 0.29603571911784604,
+      "grad_norm": 0.40999388694763184,
+      "learning_rate": 9.855725879170424e-05,
+      "loss": 1.3548,
+      "step": 1094
+    },
+    {
+      "epoch": 0.2965769178730889,
+      "grad_norm": 0.33515796065330505,
+      "learning_rate": 9.873760144274122e-05,
+      "loss": 1.3903,
+      "step": 1096
+    },
+    {
+      "epoch": 0.29711811662833176,
+      "grad_norm": 0.3834095597267151,
+      "learning_rate": 9.891794409377819e-05,
+      "loss": 1.3653,
+      "step": 1098
+    },
+    {
+      "epoch": 0.2976593153835746,
+      "grad_norm": 0.34850651025772095,
+      "learning_rate": 9.909828674481515e-05,
+      "loss": 1.3573,
+      "step": 1100
+    },
+    {
+      "epoch": 0.2982005141388175,
+      "grad_norm": 0.3811749815940857,
+      "learning_rate": 9.927862939585212e-05,
+      "loss": 1.3843,
+      "step": 1102
+    },
+    {
+      "epoch": 0.2987417128940603,
+      "grad_norm": 0.3308597803115845,
+      "learning_rate": 9.94589720468891e-05,
+      "loss": 1.3492,
+      "step": 1104
+    },
+    {
+      "epoch": 0.2992829116493032,
+      "grad_norm": 0.31952470541000366,
+      "learning_rate": 9.963931469792607e-05,
+      "loss": 1.3586,
+      "step": 1106
+    },
+    {
+      "epoch": 0.2998241104045461,
+      "grad_norm": 0.3433592915534973,
+      "learning_rate": 9.981965734896304e-05,
+      "loss": 1.3524,
+      "step": 1108
+    },
+    {
+      "epoch": 0.30036530915978893,
+      "grad_norm": 0.4547680914402008,
+      "learning_rate": 0.0001,
+      "loss": 1.3562,
+      "step": 1110
+    },
+    {
+      "epoch": 0.3009065079150318,
+      "grad_norm": 0.4963592290878296,
+      "learning_rate": 9.999999008881264e-05,
+      "loss": 1.3452,
+      "step": 1112
+    },
+    {
+      "epoch": 0.30144770667027465,
+      "grad_norm": 1.1111193895339966,
+      "learning_rate": 9.999996035525452e-05,
+      "loss": 1.3732,
+      "step": 1114
+    },
+    {
+      "epoch": 0.30198890542551754,
+      "grad_norm": 0.6860964298248291,
+      "learning_rate": 9.999991079933739e-05,
+      "loss": 1.3689,
+      "step": 1116
+    },
+    {
+      "epoch": 0.3025301041807604,
+      "grad_norm": 0.7344204783439636,
+      "learning_rate": 9.999984142108093e-05,
+      "loss": 1.3575,
+      "step": 1118
+    },
+    {
+      "epoch": 0.30307130293600326,
+      "grad_norm": 0.6534725427627563,
+      "learning_rate": 9.999975222051263e-05,
+      "loss": 1.376,
+      "step": 1120
+    },
+    {
+      "epoch": 0.3036125016912461,
+      "grad_norm": 0.5108229517936707,
+      "learning_rate": 9.999964319766785e-05,
+      "loss": 1.3741,
+      "step": 1122
+    },
+    {
+      "epoch": 0.304153700446489,
+      "grad_norm": 0.4888688325881958,
+      "learning_rate": 9.99995143525898e-05,
+      "loss": 1.3555,
+      "step": 1124
+    },
+    {
+      "epoch": 0.3046948992017318,
+      "grad_norm": 0.42808806896209717,
+      "learning_rate": 9.999936568532962e-05,
+      "loss": 1.3548,
+      "step": 1126
+    },
+    {
+      "epoch": 0.3052360979569747,
+      "grad_norm": 0.3921727240085602,
+      "learning_rate": 9.999919719594617e-05,
+      "loss": 1.3559,
+      "step": 1128
+    },
+    {
+      "epoch": 0.30577729671221754,
+      "grad_norm": 0.3473529517650604,
+      "learning_rate": 9.999900888450628e-05,
+      "loss": 1.3603,
+      "step": 1130
+    },
+    {
+      "epoch": 0.3063184954674604,
+      "grad_norm": 0.3337381184101105,
+      "learning_rate": 9.999880075108464e-05,
+      "loss": 1.3642,
+      "step": 1132
+    },
+    {
+      "epoch": 0.30685969422270326,
+      "grad_norm": 0.3363231122493744,
+      "learning_rate": 9.99985727957637e-05,
+      "loss": 1.3606,
+      "step": 1134
+    },
+    {
+      "epoch": 0.30740089297794615,
+      "grad_norm": 0.32726484537124634,
+      "learning_rate": 9.999832501863386e-05,
+      "loss": 1.3493,
+      "step": 1136
+    },
+    {
+      "epoch": 0.30794209173318904,
+      "grad_norm": 0.3190646767616272,
+      "learning_rate": 9.999805741979338e-05,
+      "loss": 1.3518,
+      "step": 1138
+    },
+    {
+      "epoch": 0.30848329048843187,
+      "grad_norm": 0.31244540214538574,
+      "learning_rate": 9.999776999934831e-05,
+      "loss": 1.3495,
+      "step": 1140
+    },
+    {
+      "epoch": 0.30902448924367476,
+      "grad_norm": 0.3286384344100952,
+      "learning_rate": 9.999746275741261e-05,
+      "loss": 1.3517,
+      "step": 1142
+    },
+    {
+      "epoch": 0.3095656879989176,
+      "grad_norm": 0.3630046546459198,
+      "learning_rate": 9.99971356941081e-05,
+      "loss": 1.3641,
+      "step": 1144
+    },
+    {
+      "epoch": 0.3101068867541605,
+      "grad_norm": 0.30771151185035706,
+      "learning_rate": 9.999678880956443e-05,
+      "loss": 1.3571,
+      "step": 1146
+    },
+    {
+      "epoch": 0.3106480855094033,
+      "grad_norm": 0.30026301741600037,
+      "learning_rate": 9.99964221039191e-05,
+      "loss": 1.3541,
+      "step": 1148
+    },
+    {
+      "epoch": 0.3111892842646462,
+      "grad_norm": 0.3128298223018646,
+      "learning_rate": 9.999603557731754e-05,
+      "loss": 1.3556,
+      "step": 1150
+    },
+    {
+      "epoch": 0.31173048301988904,
+      "grad_norm": 0.30185452103614807,
+      "learning_rate": 9.999562922991293e-05,
+      "loss": 1.3484,
+      "step": 1152
+    },
+    {
+      "epoch": 0.3122716817751319,
+      "grad_norm": 0.3274635076522827,
+      "learning_rate": 9.99952030618664e-05,
+      "loss": 1.3729,
+      "step": 1154
+    },
+    {
+      "epoch": 0.31281288053037476,
+      "grad_norm": 0.30549076199531555,
+      "learning_rate": 9.999475707334692e-05,
+      "loss": 1.3642,
+      "step": 1156
+    },
+    {
+      "epoch": 0.31335407928561765,
+      "grad_norm": 0.3147718906402588,
+      "learning_rate": 9.999429126453126e-05,
+      "loss": 1.3493,
+      "step": 1158
+    },
+    {
+      "epoch": 0.3138952780408605,
+      "grad_norm": 0.6205586791038513,
+      "learning_rate": 9.99938056356041e-05,
+      "loss": 1.3623,
+      "step": 1160
+    },
+    {
+      "epoch": 0.31443647679610337,
+      "grad_norm": 0.3471706211566925,
+      "learning_rate": 9.999330018675798e-05,
+      "loss": 1.3533,
+      "step": 1162
+    },
+    {
+      "epoch": 0.31497767555134626,
+      "grad_norm": 1.3515815734863281,
+      "learning_rate": 9.999277491819328e-05,
+      "loss": 1.3565,
+      "step": 1164
+    },
+    {
+      "epoch": 0.3155188743065891,
+      "grad_norm": 733.9155883789062,
+      "learning_rate": 9.999222983011824e-05,
+      "loss": 5.2143,
+      "step": 1166
+    },
+    {
+      "epoch": 0.316060073061832,
+      "grad_norm": 2.9439170360565186,
+      "learning_rate": 9.999166492274894e-05,
+      "loss": 1.4438,
+      "step": 1168
+    },
+    {
+      "epoch": 0.3166012718170748,
+      "grad_norm": 1.5871142148971558,
+      "learning_rate": 9.999108019630938e-05,
+      "loss": 1.4426,
+      "step": 1170
+    },
+    {
+      "epoch": 0.3171424705723177,
+      "grad_norm": 711.9217529296875,
+      "learning_rate": 9.999047565103132e-05,
+      "loss": 3.6935,
+      "step": 1172
+    },
+    {
+      "epoch": 0.31768366932756054,
+      "grad_norm": 100.76264953613281,
+      "learning_rate": 9.998985128715448e-05,
+      "loss": 4.2396,
+      "step": 1174
+    },
+    {
+      "epoch": 0.3182248680828034,
+      "grad_norm": 108.88189697265625,
+      "learning_rate": 9.998920710492634e-05,
+      "loss": 4.9929,
+      "step": 1176
+    },
+    {
+      "epoch": 0.31876606683804626,
+      "grad_norm": 72.18595123291016,
+      "learning_rate": 9.998854310460233e-05,
+      "loss": 6.0375,
+      "step": 1178
+    },
+    {
+      "epoch": 0.31930726559328915,
+      "grad_norm": 59.48538589477539,
+      "learning_rate": 9.998785928644567e-05,
+      "loss": 5.8932,
+      "step": 1180
+    },
+    {
+      "epoch": 0.319848464348532,
+      "grad_norm": 36.32703399658203,
+      "learning_rate": 9.998715565072744e-05,
+      "loss": 6.5369,
+      "step": 1182
+    },
+    {
+      "epoch": 0.32038966310377487,
+      "grad_norm": 18.565351486206055,
+      "learning_rate": 9.998643219772664e-05,
+      "loss": 6.1671,
+      "step": 1184
+    },
+    {
+      "epoch": 0.3209308618590177,
+      "grad_norm": 45.84898376464844,
+      "learning_rate": 9.998568892773003e-05,
+      "loss": 5.9379,
+      "step": 1186
+    },
+    {
+      "epoch": 0.3214720606142606,
+      "grad_norm": 66.2480239868164,
+      "learning_rate": 9.998492584103232e-05,
+      "loss": 5.7071,
+      "step": 1188
+    },
+    {
+      "epoch": 0.3220132593695034,
+      "grad_norm": 41.693092346191406,
+      "learning_rate": 9.998414293793599e-05,
+      "loss": 6.3198,
+      "step": 1190
+    },
+    {
+      "epoch": 0.3225544581247463,
+      "grad_norm": 19.323413848876953,
+      "learning_rate": 9.998334021875147e-05,
+      "loss": 5.377,
+      "step": 1192
+    },
+    {
+      "epoch": 0.3230956568799892,
+      "grad_norm": 15.907301902770996,
+      "learning_rate": 9.998251768379696e-05,
+      "loss": 4.5293,
+      "step": 1194
+    },
+    {
+      "epoch": 0.32363685563523203,
+      "grad_norm": 80.1374740600586,
+      "learning_rate": 9.998167533339857e-05,
+      "loss": 4.3471,
+      "step": 1196
+    },
+    {
+      "epoch": 0.3241780543904749,
+      "grad_norm": 23.298336029052734,
+      "learning_rate": 9.998081316789024e-05,
+      "loss": 3.7461,
+      "step": 1198
+    },
+    {
+      "epoch": 0.32471925314571776,
+      "grad_norm": 82.48027801513672,
+      "learning_rate": 9.997993118761378e-05,
+      "loss": 4.1647,
+      "step": 1200
+    },
+    {
+      "epoch": 0.32526045190096065,
+      "grad_norm": 27.916913986206055,
+      "learning_rate": 9.997902939291883e-05,
+      "loss": 3.9092,
+      "step": 1202
+    },
+    {
+      "epoch": 0.3258016506562035,
+      "grad_norm": 15.70148754119873,
+      "learning_rate": 9.997810778416293e-05,
+      "loss": 3.1628,
+      "step": 1204
+    },
+    {
+      "epoch": 0.32634284941144637,
+      "grad_norm": 18.33330535888672,
+      "learning_rate": 9.997716636171142e-05,
+      "loss": 2.8777,
+      "step": 1206
+    },
+    {
+      "epoch": 0.3268840481666892,
+      "grad_norm": 10.6620512008667,
+      "learning_rate": 9.997620512593755e-05,
+      "loss": 2.3009,
+      "step": 1208
+    },
+    {
+      "epoch": 0.3274252469219321,
+      "grad_norm": 32.01799011230469,
+      "learning_rate": 9.99752240772224e-05,
+      "loss": 1.9617,
+      "step": 1210
+    },
+    {
+      "epoch": 0.3279664456771749,
+      "grad_norm": 5.677090644836426,
+      "learning_rate": 9.997422321595488e-05,
+      "loss": 1.8401,
+      "step": 1212
+    },
+    {
+      "epoch": 0.3285076444324178,
+      "grad_norm": 8.914667129516602,
+      "learning_rate": 9.997320254253179e-05,
+      "loss": 1.6707,
+      "step": 1214
+    },
+    {
+      "epoch": 0.32904884318766064,
+      "grad_norm": 2.3725008964538574,
+      "learning_rate": 9.997216205735779e-05,
+      "loss": 1.5757,
+      "step": 1216
+    },
+    {
+      "epoch": 0.32959004194290353,
+      "grad_norm": 2.418389320373535,
+      "learning_rate": 9.997110176084538e-05,
+      "loss": 1.5154,
+      "step": 1218
+    },
+    {
+      "epoch": 0.33013124069814637,
+      "grad_norm": 2.802185297012329,
+      "learning_rate": 9.997002165341487e-05,
+      "loss": 1.4883,
+      "step": 1220
+    },
+    {
+      "epoch": 0.33067243945338926,
+      "grad_norm": 2.1769211292266846,
+      "learning_rate": 9.996892173549452e-05,
+      "loss": 1.445,
+      "step": 1222
+    },
+    {
+      "epoch": 0.33121363820863214,
+      "grad_norm": 1.799670934677124,
+      "learning_rate": 9.996780200752035e-05,
+      "loss": 1.4276,
+      "step": 1224
+    },
+    {
+      "epoch": 0.331754836963875,
+      "grad_norm": 3.2545313835144043,
+      "learning_rate": 9.996666246993627e-05,
+      "loss": 1.4394,
+      "step": 1226
+    },
+    {
+      "epoch": 0.33229603571911787,
+      "grad_norm": 1.1922351121902466,
+      "learning_rate": 9.996550312319408e-05,
+      "loss": 1.4359,
+      "step": 1228
+    },
+    {
+      "epoch": 0.3328372344743607,
+      "grad_norm": 2.6813228130340576,
+      "learning_rate": 9.996432396775339e-05,
+      "loss": 1.4229,
+      "step": 1230
+    },
+    {
+      "epoch": 0.3333784332296036,
+      "grad_norm": 1.6968843936920166,
+      "learning_rate": 9.996312500408165e-05,
+      "loss": 1.4281,
+      "step": 1232
+    },
+    {
+      "epoch": 0.3339196319848464,
+      "grad_norm": 1.3502254486083984,
+      "learning_rate": 9.996190623265421e-05,
+      "loss": 1.408,
+      "step": 1234
+    },
+    {
+      "epoch": 0.3344608307400893,
+      "grad_norm": 1.2809518575668335,
+      "learning_rate": 9.996066765395424e-05,
+      "loss": 1.4176,
+      "step": 1236
+    },
+    {
+      "epoch": 0.33500202949533214,
+      "grad_norm": 1.0455057621002197,
+      "learning_rate": 9.995940926847279e-05,
+      "loss": 1.4056,
+      "step": 1238
+    },
+    {
+      "epoch": 0.33554322825057503,
+      "grad_norm": 1.3292824029922485,
+      "learning_rate": 9.99581310767087e-05,
+      "loss": 1.4033,
+      "step": 1240
+    },
+    {
+      "epoch": 0.33608442700581787,
+      "grad_norm": 1.5960067510604858,
+      "learning_rate": 9.995683307916875e-05,
+      "loss": 1.379,
+      "step": 1242
+    },
+    {
+      "epoch": 0.33662562576106075,
+      "grad_norm": 1.0471105575561523,
+      "learning_rate": 9.99555152763675e-05,
+      "loss": 1.3823,
+      "step": 1244
+    },
+    {
+      "epoch": 0.3371668245163036,
+      "grad_norm": 2.339273452758789,
+      "learning_rate": 9.99541776688274e-05,
+      "loss": 1.3698,
+      "step": 1246
+    },
+    {
+      "epoch": 0.3377080232715465,
+      "grad_norm": 0.81674724817276,
+      "learning_rate": 9.995282025707875e-05,
+      "loss": 1.4154,
+      "step": 1248
+    },
+    {
+      "epoch": 0.33824922202678936,
+      "grad_norm": 0.6240290999412537,
+      "learning_rate": 9.995144304165968e-05,
+      "loss": 1.4035,
+      "step": 1250
+    },
+    {
+      "epoch": 0.3387904207820322,
+      "grad_norm": 2.281787872314453,
+      "learning_rate": 9.995004602311619e-05,
+      "loss": 1.3906,
+      "step": 1252
+    },
+    {
+      "epoch": 0.3393316195372751,
+      "grad_norm": 0.6818395853042603,
+      "learning_rate": 9.99486292020021e-05,
+      "loss": 1.3853,
+      "step": 1254
+    },
+    {
+      "epoch": 0.3398728182925179,
+      "grad_norm": 6.299881935119629,
+      "learning_rate": 9.994719257887915e-05,
+      "loss": 1.3856,
+      "step": 1256
+    },
+    {
+      "epoch": 0.3404140170477608,
+      "grad_norm": 0.8173750638961792,
+      "learning_rate": 9.994573615431686e-05,
+      "loss": 1.3871,
+      "step": 1258
+    },
+    {
+      "epoch": 0.34095521580300364,
+      "grad_norm": 2.155395746231079,
+      "learning_rate": 9.994425992889262e-05,
+      "loss": 1.3382,
+      "step": 1260
+    },
+    {
+      "epoch": 0.34149641455824653,
+      "grad_norm": 0.5846114754676819,
+      "learning_rate": 9.99427639031917e-05,
+      "loss": 1.3978,
+      "step": 1262
+    },
+    {
+      "epoch": 0.34203761331348936,
+      "grad_norm": 0.6624069213867188,
+      "learning_rate": 9.994124807780717e-05,
+      "loss": 1.3792,
+      "step": 1264
+    },
+    {
+      "epoch": 0.34257881206873225,
+      "grad_norm": 0.5708588361740112,
+      "learning_rate": 9.993971245333998e-05,
+      "loss": 1.3677,
+      "step": 1266
+    },
+    {
+      "epoch": 0.3431200108239751,
+      "grad_norm": 0.5245474576950073,
+      "learning_rate": 9.993815703039894e-05,
+      "loss": 1.3672,
+      "step": 1268
+    },
+    {
+      "epoch": 0.343661209579218,
+      "grad_norm": 0.501871645450592,
+      "learning_rate": 9.993658180960069e-05,
+      "loss": 1.3674,
+      "step": 1270
+    },
+    {
+      "epoch": 0.3442024083344608,
+      "grad_norm": 0.5990382432937622,
+      "learning_rate": 9.993498679156969e-05,
+      "loss": 1.3804,
+      "step": 1272
+    },
+    {
+      "epoch": 0.3447436070897037,
+      "grad_norm": 0.42392146587371826,
+      "learning_rate": 9.993337197693833e-05,
+      "loss": 1.3628,
+      "step": 1274
+    },
+    {
+      "epoch": 0.34528480584494653,
+      "grad_norm": 0.46936917304992676,
+      "learning_rate": 9.993173736634676e-05,
+      "loss": 1.3696,
+      "step": 1276
+    },
+    {
+      "epoch": 0.3458260046001894,
+      "grad_norm": 0.52222740650177,
+      "learning_rate": 9.993008296044304e-05,
+      "loss": 1.3697,
+      "step": 1278
+    },
+    {
+      "epoch": 0.3463672033554323,
+      "grad_norm": 0.3582518398761749,
+      "learning_rate": 9.992840875988305e-05,
+      "loss": 1.3825,
+      "step": 1280
+    },
+    {
+      "epoch": 0.34690840211067514,
+      "grad_norm": 0.3533988296985626,
+      "learning_rate": 9.99267147653305e-05,
+      "loss": 1.361,
+      "step": 1282
+    },
+    {
+      "epoch": 0.34744960086591803,
+      "grad_norm": 0.35905274748802185,
+      "learning_rate": 9.992500097745702e-05,
+      "loss": 1.3721,
+      "step": 1284
+    },
+    {
+      "epoch": 0.34799079962116086,
+      "grad_norm": 0.3057416081428528,
+      "learning_rate": 9.9923267396942e-05,
+      "loss": 1.369,
+      "step": 1286
+    },
+    {
+      "epoch": 0.34853199837640375,
+      "grad_norm": 0.3299311101436615,
+      "learning_rate": 9.992151402447272e-05,
+      "loss": 1.358,
+      "step": 1288
+    },
+    {
+      "epoch": 0.3490731971316466,
+      "grad_norm": 0.3086453080177307,
+      "learning_rate": 9.99197408607443e-05,
+      "loss": 1.3534,
+      "step": 1290
+    },
+    {
+      "epoch": 0.3496143958868895,
+      "grad_norm": 0.3111782968044281,
+      "learning_rate": 9.991794790645969e-05,
+      "loss": 1.3605,
+      "step": 1292
+    },
+    {
+      "epoch": 0.3501555946421323,
+      "grad_norm": 0.3231568932533264,
+      "learning_rate": 9.991613516232974e-05,
+      "loss": 1.3543,
+      "step": 1294
+    },
+    {
+      "epoch": 0.3506967933973752,
+      "grad_norm": 0.3288814425468445,
+      "learning_rate": 9.991430262907309e-05,
+      "loss": 1.3521,
+      "step": 1296
+    },
+    {
+      "epoch": 0.35123799215261803,
+      "grad_norm": 0.3239436745643616,
+      "learning_rate": 9.991245030741622e-05,
+      "loss": 1.3335,
+      "step": 1298
+    },
+    {
+      "epoch": 0.3517791909078609,
+      "grad_norm": 0.3560773730278015,
+      "learning_rate": 9.991057819809353e-05,
+      "loss": 1.3487,
+      "step": 1300
+    },
+    {
+      "epoch": 0.35232038966310375,
+      "grad_norm": 0.4387347400188446,
+      "learning_rate": 9.990868630184716e-05,
+      "loss": 1.3548,
+      "step": 1302
+    },
+    {
+      "epoch": 0.35286158841834664,
+      "grad_norm": 0.32067278027534485,
+      "learning_rate": 9.990677461942717e-05,
+      "loss": 1.3471,
+      "step": 1304
+    },
+    {
+      "epoch": 0.3534027871735895,
+      "grad_norm": 0.4399580955505371,
+      "learning_rate": 9.990484315159146e-05,
+      "loss": 1.3588,
+      "step": 1306
+    },
+    {
+      "epoch": 0.35394398592883236,
+      "grad_norm": 0.9175602793693542,
+      "learning_rate": 9.990289189910571e-05,
+      "loss": 1.3432,
+      "step": 1308
+    },
+    {
+      "epoch": 0.35448518468407525,
+      "grad_norm": 0.45273318886756897,
+      "learning_rate": 9.990092086274352e-05,
+      "loss": 1.3434,
+      "step": 1310
+    },
+    {
+      "epoch": 0.3550263834393181,
+      "grad_norm": 0.3346487879753113,
+      "learning_rate": 9.989893004328632e-05,
+      "loss": 1.3339,
+      "step": 1312
+    },
+    {
+      "epoch": 0.35556758219456097,
+      "grad_norm": 0.4779951870441437,
+      "learning_rate": 9.989691944152333e-05,
+      "loss": 1.3561,
+      "step": 1314
+    },
+    {
+      "epoch": 0.3561087809498038,
+      "grad_norm": 0.6359366774559021,
+      "learning_rate": 9.989488905825166e-05,
+      "loss": 1.3499,
+      "step": 1316
+    },
+    {
+      "epoch": 0.3566499797050467,
+      "grad_norm": 0.5867050290107727,
+      "learning_rate": 9.989283889427625e-05,
+      "loss": 1.3791,
+      "step": 1318
+    },
+    {
+      "epoch": 0.3571911784602895,
+      "grad_norm": 1.869691014289856,
+      "learning_rate": 9.989076895040989e-05,
+      "loss": 1.3663,
+      "step": 1320
+    },
+    {
+      "epoch": 0.3577323772155324,
+      "grad_norm": 2.7147843837738037,
+      "learning_rate": 9.98886792274732e-05,
+      "loss": 1.358,
+      "step": 1322
+    },
+    {
+      "epoch": 0.35827357597077525,
+      "grad_norm": 0.8717885613441467,
+      "learning_rate": 9.988656972629465e-05,
+      "loss": 1.34,
+      "step": 1324
+    },
+    {
+      "epoch": 0.35881477472601814,
+      "grad_norm": 0.7126337885856628,
+      "learning_rate": 9.988444044771054e-05,
+      "loss": 1.3281,
+      "step": 1326
+    },
+    {
+      "epoch": 0.35935597348126097,
+      "grad_norm": 0.7409217357635498,
+      "learning_rate": 9.988229139256502e-05,
+      "loss": 1.3571,
+      "step": 1328
+    },
+    {
+      "epoch": 0.35989717223650386,
+      "grad_norm": 0.5892549157142639,
+      "learning_rate": 9.988012256171006e-05,
+      "loss": 1.3269,
+      "step": 1330
+    },
+    {
+      "epoch": 0.3604383709917467,
+      "grad_norm": 0.4858717620372772,
+      "learning_rate": 9.98779339560055e-05,
+      "loss": 1.3506,
+      "step": 1332
+    },
+    {
+      "epoch": 0.3609795697469896,
+      "grad_norm": 0.37409740686416626,
+      "learning_rate": 9.987572557631903e-05,
+      "loss": 1.3339,
+      "step": 1334
+    },
+    {
+      "epoch": 0.36152076850223247,
+      "grad_norm": 0.38315168023109436,
+      "learning_rate": 9.987349742352611e-05,
+      "loss": 1.3404,
+      "step": 1336
+    },
+    {
+      "epoch": 0.3620619672574753,
+      "grad_norm": 0.32702726125717163,
+      "learning_rate": 9.987124949851014e-05,
+      "loss": 1.3595,
+      "step": 1338
+    },
+    {
+      "epoch": 0.3626031660127182,
+      "grad_norm": 0.3133656680583954,
+      "learning_rate": 9.986898180216226e-05,
+      "loss": 1.3428,
+      "step": 1340
+    },
+    {
+      "epoch": 0.363144364767961,
+      "grad_norm": 0.2916230857372284,
+      "learning_rate": 9.986669433538152e-05,
+      "loss": 1.3381,
+      "step": 1342
+    },
+    {
+      "epoch": 0.3636855635232039,
+      "grad_norm": 0.28036215901374817,
+      "learning_rate": 9.986438709907476e-05,
+      "loss": 1.3447,
+      "step": 1344
+    },
+    {
+      "epoch": 0.36422676227844675,
+      "grad_norm": 0.30352699756622314,
+      "learning_rate": 9.98620600941567e-05,
+      "loss": 1.3427,
+      "step": 1346
+    },
+    {
+      "epoch": 0.36476796103368964,
+      "grad_norm": 0.3100769519805908,
+      "learning_rate": 9.985971332154984e-05,
+      "loss": 1.3603,
+      "step": 1348
+    },
+    {
+      "epoch": 0.36530915978893247,
+      "grad_norm": 0.2933647930622101,
+      "learning_rate": 9.98573467821846e-05,
+      "loss": 1.3646,
+      "step": 1350
+    },
+    {
+      "epoch": 0.36585035854417536,
+      "grad_norm": 0.2938663959503174,
+      "learning_rate": 9.985496047699916e-05,
+      "loss": 1.3763,
+      "step": 1352
+    },
+    {
+      "epoch": 0.3663915572994182,
+      "grad_norm": 0.2916519343852997,
+      "learning_rate": 9.985255440693955e-05,
+      "loss": 1.3431,
+      "step": 1354
+    },
+    {
+      "epoch": 0.3669327560546611,
+      "grad_norm": 0.2954147756099701,
+      "learning_rate": 9.985012857295968e-05,
+      "loss": 1.338,
+      "step": 1356
+    },
+    {
+      "epoch": 0.3674739548099039,
+      "grad_norm": 0.2839341163635254,
+      "learning_rate": 9.984768297602125e-05,
+      "loss": 1.3653,
+      "step": 1358
+    },
+    {
+      "epoch": 0.3680151535651468,
+      "grad_norm": 0.2878473699092865,
+      "learning_rate": 9.984521761709382e-05,
+      "loss": 1.3302,
+      "step": 1360
+    },
+    {
+      "epoch": 0.3685563523203897,
+      "grad_norm": 0.2859325408935547,
+      "learning_rate": 9.984273249715478e-05,
+      "loss": 1.3273,
+      "step": 1362
+    },
+    {
+      "epoch": 0.3690975510756325,
+      "grad_norm": 0.28399959206581116,
+      "learning_rate": 9.984022761718933e-05,
+      "loss": 1.3516,
+      "step": 1364
+    },
+    {
+      "epoch": 0.3696387498308754,
+      "grad_norm": 0.29740169644355774,
+      "learning_rate": 9.983770297819052e-05,
+      "loss": 1.3389,
+      "step": 1366
+    },
+    {
+      "epoch": 0.37017994858611825,
+      "grad_norm": 0.3143361806869507,
+      "learning_rate": 9.983515858115928e-05,
+      "loss": 1.3557,
+      "step": 1368
+    },
+    {
+      "epoch": 0.37072114734136113,
+      "grad_norm": 0.30783936381340027,
+      "learning_rate": 9.983259442710429e-05,
+      "loss": 1.3498,
+      "step": 1370
+    },
+    {
+      "epoch": 0.37126234609660397,
+      "grad_norm": 0.297091543674469,
+      "learning_rate": 9.983001051704211e-05,
+      "loss": 1.3308,
+      "step": 1372
+    },
+    {
+      "epoch": 0.37180354485184686,
+      "grad_norm": 0.3118893504142761,
+      "learning_rate": 9.982740685199712e-05,
+      "loss": 1.3372,
+      "step": 1374
+    },
+    {
+      "epoch": 0.3723447436070897,
+      "grad_norm": 0.2826865017414093,
+      "learning_rate": 9.982478343300155e-05,
+      "loss": 1.3488,
+      "step": 1376
+    },
+    {
+      "epoch": 0.3728859423623326,
+      "grad_norm": 0.2829175889492035,
+      "learning_rate": 9.982214026109544e-05,
+      "loss": 1.3693,
+      "step": 1378
+    },
+    {
+      "epoch": 0.3734271411175754,
+      "grad_norm": 0.3026389479637146,
+      "learning_rate": 9.981947733732668e-05,
+      "loss": 1.3276,
+      "step": 1380
+    },
+    {
+      "epoch": 0.3739683398728183,
+      "grad_norm": 0.30112889409065247,
+      "learning_rate": 9.981679466275096e-05,
+      "loss": 1.3441,
+      "step": 1382
+    },
+    {
+      "epoch": 0.37450953862806113,
+      "grad_norm": 0.27241262793540955,
+      "learning_rate": 9.981409223843183e-05,
+      "loss": 1.3373,
+      "step": 1384
+    },
+    {
+      "epoch": 0.375050737383304,
+      "grad_norm": 0.2804114520549774,
+      "learning_rate": 9.981137006544066e-05,
+      "loss": 1.344,
+      "step": 1386
+    },
+    {
+      "epoch": 0.37559193613854686,
+      "grad_norm": 0.27698764204978943,
+      "learning_rate": 9.980862814485665e-05,
+      "loss": 1.3543,
+      "step": 1388
+    },
+    {
+      "epoch": 0.37613313489378974,
+      "grad_norm": 0.29283177852630615,
+      "learning_rate": 9.980586647776681e-05,
+      "loss": 1.3332,
+      "step": 1390
+    },
+    {
+      "epoch": 0.37667433364903263,
+      "grad_norm": 0.2896028459072113,
+      "learning_rate": 9.980308506526604e-05,
+      "loss": 1.3392,
+      "step": 1392
+    },
+    {
+      "epoch": 0.37721553240427547,
+      "grad_norm": 0.27882838249206543,
+      "learning_rate": 9.980028390845697e-05,
+      "loss": 1.336,
+      "step": 1394
+    },
+    {
+      "epoch": 0.37775673115951836,
+      "grad_norm": 0.2886262834072113,
+      "learning_rate": 9.979746300845015e-05,
+      "loss": 1.3331,
+      "step": 1396
+    },
+    {
+      "epoch": 0.3782979299147612,
+      "grad_norm": 0.3085189163684845,
+      "learning_rate": 9.97946223663639e-05,
+      "loss": 1.3296,
+      "step": 1398
+    },
+    {
+      "epoch": 0.3788391286700041,
+      "grad_norm": 0.3342386484146118,
+      "learning_rate": 9.97917619833244e-05,
+      "loss": 1.351,
+      "step": 1400
+    },
+    {
+      "epoch": 0.3793803274252469,
+      "grad_norm": 0.3263756036758423,
+      "learning_rate": 9.978888186046562e-05,
+      "loss": 1.3526,
+      "step": 1402
+    },
+    {
+      "epoch": 0.3799215261804898,
+      "grad_norm": 0.292346715927124,
+      "learning_rate": 9.97859819989294e-05,
+      "loss": 1.3498,
+      "step": 1404
+    },
+    {
+      "epoch": 0.38046272493573263,
+      "grad_norm": 0.29072263836860657,
+      "learning_rate": 9.978306239986536e-05,
+      "loss": 1.3423,
+      "step": 1406
+    },
+    {
+      "epoch": 0.3810039236909755,
+      "grad_norm": 0.3350834548473358,
+      "learning_rate": 9.978012306443101e-05,
+      "loss": 1.3559,
+      "step": 1408
+    },
+    {
+      "epoch": 0.38154512244621835,
+      "grad_norm": 0.28721559047698975,
+      "learning_rate": 9.977716399379157e-05,
+      "loss": 1.3294,
+      "step": 1410
+    },
+    {
+      "epoch": 0.38208632120146124,
+      "grad_norm": 0.3062276244163513,
+      "learning_rate": 9.977418518912023e-05,
+      "loss": 1.3457,
+      "step": 1412
+    },
+    {
+      "epoch": 0.3826275199567041,
+      "grad_norm": 0.30255332589149475,
+      "learning_rate": 9.977118665159791e-05,
+      "loss": 1.3371,
+      "step": 1414
+    },
+    {
+      "epoch": 0.38316871871194697,
+      "grad_norm": 0.2800199091434479,
+      "learning_rate": 9.976816838241334e-05,
+      "loss": 1.3439,
+      "step": 1416
+    },
+    {
+      "epoch": 0.3837099174671898,
+      "grad_norm": 0.2754746675491333,
+      "learning_rate": 9.976513038276312e-05,
+      "loss": 1.3303,
+      "step": 1418
+    },
+    {
+      "epoch": 0.3842511162224327,
+      "grad_norm": 0.29933616518974304,
+      "learning_rate": 9.976207265385168e-05,
+      "loss": 1.3365,
+      "step": 1420
+    },
+    {
+      "epoch": 0.3847923149776756,
+      "grad_norm": 0.3023386001586914,
+      "learning_rate": 9.975899519689122e-05,
+      "loss": 1.3164,
+      "step": 1422
+    },
+    {
+      "epoch": 0.3853335137329184,
+      "grad_norm": 0.2901383936405182,
+      "learning_rate": 9.975589801310181e-05,
+      "loss": 1.3209,
+      "step": 1424
+    },
+    {
+      "epoch": 0.3858747124881613,
+      "grad_norm": 0.28566035628318787,
+      "learning_rate": 9.975278110371131e-05,
+      "loss": 1.3301,
+      "step": 1426
+    },
+    {
+      "epoch": 0.38641591124340413,
+      "grad_norm": 0.3010505735874176,
+      "learning_rate": 9.974964446995543e-05,
+      "loss": 1.319,
+      "step": 1428
+    },
+    {
+      "epoch": 0.386957109998647,
+      "grad_norm": 0.2977135479450226,
+      "learning_rate": 9.974648811307766e-05,
+      "loss": 1.3311,
+      "step": 1430
+    },
+    {
+      "epoch": 0.38749830875388985,
+      "grad_norm": 0.28914034366607666,
+      "learning_rate": 9.974331203432932e-05,
+      "loss": 1.343,
+      "step": 1432
+    },
+    {
+      "epoch": 0.38803950750913274,
+      "grad_norm": 0.2842980623245239,
+      "learning_rate": 9.974011623496958e-05,
+      "loss": 1.3162,
+      "step": 1434
+    },
+    {
+      "epoch": 0.3885807062643756,
+      "grad_norm": 0.3048929274082184,
+      "learning_rate": 9.97369007162654e-05,
+      "loss": 1.3166,
+      "step": 1436
+    },
+    {
+      "epoch": 0.38912190501961846,
+      "grad_norm": 0.3024531304836273,
+      "learning_rate": 9.973366547949157e-05,
+      "loss": 1.3156,
+      "step": 1438
+    },
+    {
+      "epoch": 0.3896631037748613,
+      "grad_norm": 0.2911103367805481,
+      "learning_rate": 9.973041052593068e-05,
+      "loss": 1.3314,
+      "step": 1440
+    },
+    {
+      "epoch": 0.3902043025301042,
+      "grad_norm": 0.30932334065437317,
+      "learning_rate": 9.972713585687317e-05,
+      "loss": 1.3144,
+      "step": 1442
+    },
+    {
+      "epoch": 0.390745501285347,
+      "grad_norm": 0.302971750497818,
+      "learning_rate": 9.972384147361725e-05,
+      "loss": 1.3431,
+      "step": 1444
+    },
+    {
+      "epoch": 0.3912867000405899,
+      "grad_norm": 0.32412296533584595,
+      "learning_rate": 9.972052737746898e-05,
+      "loss": 1.3167,
+      "step": 1446
+    },
+    {
+      "epoch": 0.3918278987958328,
+      "grad_norm": 0.4637945890426636,
+      "learning_rate": 9.97171935697422e-05,
+      "loss": 1.3433,
+      "step": 1448
+    },
+    {
+      "epoch": 0.39236909755107563,
+      "grad_norm": 0.32690081000328064,
+      "learning_rate": 9.971384005175864e-05,
+      "loss": 1.3327,
+      "step": 1450
+    },
+    {
+      "epoch": 0.3929102963063185,
+      "grad_norm": 0.3049994111061096,
+      "learning_rate": 9.971046682484776e-05,
+      "loss": 1.3401,
+      "step": 1452
+    },
+    {
+      "epoch": 0.39345149506156135,
+      "grad_norm": 0.306095689535141,
+      "learning_rate": 9.970707389034688e-05,
+      "loss": 1.3205,
+      "step": 1454
+    },
+    {
+      "epoch": 0.39399269381680424,
+      "grad_norm": 0.3375592529773712,
+      "learning_rate": 9.970366124960111e-05,
+      "loss": 1.3243,
+      "step": 1456
+    },
+    {
+      "epoch": 0.3945338925720471,
+      "grad_norm": 0.30508387088775635,
+      "learning_rate": 9.970022890396338e-05,
+      "loss": 1.3342,
+      "step": 1458
+    },
+    {
+      "epoch": 0.39507509132728996,
+      "grad_norm": 0.2996918261051178,
+      "learning_rate": 9.969677685479444e-05,
+      "loss": 1.3457,
+      "step": 1460
+    },
+    {
+      "epoch": 0.3956162900825328,
+      "grad_norm": 0.29500269889831543,
+      "learning_rate": 9.969330510346286e-05,
+      "loss": 1.3306,
+      "step": 1462
+    },
+    {
+      "epoch": 0.3961574888377757,
+      "grad_norm": 0.28392598032951355,
+      "learning_rate": 9.9689813651345e-05,
+      "loss": 1.3347,
+      "step": 1464
+    },
+    {
+      "epoch": 0.3966986875930185,
+      "grad_norm": 0.2859434485435486,
+      "learning_rate": 9.968630249982503e-05,
+      "loss": 1.3342,
+      "step": 1466
+    },
+    {
+      "epoch": 0.3972398863482614,
+      "grad_norm": 0.3038876950740814,
+      "learning_rate": 9.968277165029494e-05,
+      "loss": 1.3248,
+      "step": 1468
+    },
+    {
+      "epoch": 0.39778108510350424,
+      "grad_norm": 0.3060581088066101,
+      "learning_rate": 9.967922110415454e-05,
+      "loss": 1.3403,
+      "step": 1470
+    },
+    {
+      "epoch": 0.39832228385874713,
+      "grad_norm": 0.30475133657455444,
+      "learning_rate": 9.96756508628114e-05,
+      "loss": 1.3338,
+      "step": 1472
+    },
+    {
+      "epoch": 0.39886348261398996,
+      "grad_norm": 0.33263343572616577,
+      "learning_rate": 9.967206092768095e-05,
+      "loss": 1.3209,
+      "step": 1474
+    },
+    {
+      "epoch": 0.39940468136923285,
+      "grad_norm": 0.2895435094833374,
+      "learning_rate": 9.966845130018645e-05,
+      "loss": 1.3352,
+      "step": 1476
+    },
+    {
+      "epoch": 0.39994588012447574,
+      "grad_norm": 0.27237775921821594,
+      "learning_rate": 9.966482198175886e-05,
+      "loss": 1.3239,
+      "step": 1478
+    },
+    {
+      "epoch": 0.40048707887971857,
+      "grad_norm": 0.2740168571472168,
+      "learning_rate": 9.966117297383707e-05,
+      "loss": 1.3371,
+      "step": 1480
+    },
+    {
+      "epoch": 0.40102827763496146,
+      "grad_norm": 0.30601269006729126,
+      "learning_rate": 9.965750427786768e-05,
+      "loss": 1.343,
+      "step": 1482
+    },
+    {
+      "epoch": 0.4015694763902043,
+      "grad_norm": 0.28768840432167053,
+      "learning_rate": 9.965381589530518e-05,
+      "loss": 1.3442,
+      "step": 1484
+    },
+    {
+      "epoch": 0.4021106751454472,
+      "grad_norm": 0.28244882822036743,
+      "learning_rate": 9.965010782761177e-05,
+      "loss": 1.3336,
+      "step": 1486
+    },
+    {
+      "epoch": 0.40265187390069,
+      "grad_norm": 0.2694818079471588,
+      "learning_rate": 9.964638007625754e-05,
+      "loss": 1.3448,
+      "step": 1488
+    },
+    {
+      "epoch": 0.4031930726559329,
+      "grad_norm": 0.29507288336753845,
+      "learning_rate": 9.964263264272033e-05,
+      "loss": 1.327,
+      "step": 1490
+    },
+    {
+      "epoch": 0.40373427141117574,
+      "grad_norm": 0.3036315143108368,
+      "learning_rate": 9.963886552848581e-05,
+      "loss": 1.3289,
+      "step": 1492
+    },
+    {
+      "epoch": 0.4042754701664186,
+      "grad_norm": 0.2737107574939728,
+      "learning_rate": 9.963507873504744e-05,
+      "loss": 1.3281,
+      "step": 1494
+    },
+    {
+      "epoch": 0.40481666892166146,
+      "grad_norm": 0.29833105206489563,
+      "learning_rate": 9.963127226390647e-05,
+      "loss": 1.3378,
+      "step": 1496
+    },
+    {
+      "epoch": 0.40535786767690435,
+      "grad_norm": 0.32203689217567444,
+      "learning_rate": 9.9627446116572e-05,
+      "loss": 1.3158,
+      "step": 1498
+    },
+    {
+      "epoch": 0.4058990664321472,
+      "grad_norm": 0.27837038040161133,
+      "learning_rate": 9.962360029456086e-05,
+      "loss": 1.3051,
+      "step": 1500
+    },
+    {
+      "epoch": 0.40644026518739007,
+      "grad_norm": 0.2688932418823242,
+      "learning_rate": 9.961973479939774e-05,
+      "loss": 1.339,
+      "step": 1502
+    },
+    {
+      "epoch": 0.40698146394263296,
+      "grad_norm": 0.2779388725757599,
+      "learning_rate": 9.96158496326151e-05,
+      "loss": 1.3264,
+      "step": 1504
+    },
+    {
+      "epoch": 0.4075226626978758,
+      "grad_norm": 0.27401190996170044,
+      "learning_rate": 9.961194479575321e-05,
+      "loss": 1.3139,
+      "step": 1506
+    },
+    {
+      "epoch": 0.4080638614531187,
+      "grad_norm": 0.270448237657547,
+      "learning_rate": 9.960802029036012e-05,
+      "loss": 1.3253,
+      "step": 1508
+    },
+    {
+      "epoch": 0.4086050602083615,
+      "grad_norm": 0.29150158166885376,
+      "learning_rate": 9.96040761179917e-05,
+      "loss": 1.3324,
+      "step": 1510
+    },
+    {
+      "epoch": 0.4091462589636044,
+      "grad_norm": 0.2666511833667755,
+      "learning_rate": 9.960011228021159e-05,
+      "loss": 1.325,
+      "step": 1512
+    },
+    {
+      "epoch": 0.40968745771884724,
+      "grad_norm": 0.2782241106033325,
+      "learning_rate": 9.959612877859125e-05,
+      "loss": 1.3162,
+      "step": 1514
+    },
+    {
+      "epoch": 0.4102286564740901,
+      "grad_norm": 0.2845720946788788,
+      "learning_rate": 9.959212561470996e-05,
+      "loss": 1.3316,
+      "step": 1516
+    },
+    {
+      "epoch": 0.41076985522933296,
+      "grad_norm": 0.27991780638694763,
+      "learning_rate": 9.958810279015473e-05,
+      "loss": 1.3121,
+      "step": 1518
+    },
+    {
+      "epoch": 0.41131105398457585,
+      "grad_norm": 0.2804965674877167,
+      "learning_rate": 9.958406030652043e-05,
+      "loss": 1.3246,
+      "step": 1520
+    },
+    {
+      "epoch": 0.4118522527398187,
+      "grad_norm": 0.2732795178890228,
+      "learning_rate": 9.957999816540965e-05,
+      "loss": 1.3217,
+      "step": 1522
+    },
+    {
+      "epoch": 0.41239345149506157,
+      "grad_norm": 0.28181079030036926,
+      "learning_rate": 9.957591636843284e-05,
+      "loss": 1.3374,
+      "step": 1524
+    },
+    {
+      "epoch": 0.4129346502503044,
+      "grad_norm": 0.3096240162849426,
+      "learning_rate": 9.957181491720822e-05,
+      "loss": 1.3324,
+      "step": 1526
+    },
+    {
+      "epoch": 0.4134758490055473,
+      "grad_norm": 0.2709742486476898,
+      "learning_rate": 9.95676938133618e-05,
+      "loss": 1.3055,
+      "step": 1528
+    },
+    {
+      "epoch": 0.4140170477607901,
+      "grad_norm": 0.27309080958366394,
+      "learning_rate": 9.956355305852736e-05,
+      "loss": 1.313,
+      "step": 1530
+    },
+    {
+      "epoch": 0.414558246516033,
+      "grad_norm": 0.29801151156425476,
+      "learning_rate": 9.955939265434652e-05,
+      "loss": 1.3185,
+      "step": 1532
+    },
+    {
+      "epoch": 0.4150994452712759,
+      "grad_norm": 0.28698021173477173,
+      "learning_rate": 9.955521260246865e-05,
+      "loss": 1.3214,
+      "step": 1534
+    },
+    {
+      "epoch": 0.41564064402651874,
+      "grad_norm": 0.2641914188861847,
+      "learning_rate": 9.955101290455093e-05,
+      "loss": 1.317,
+      "step": 1536
+    },
+    {
+      "epoch": 0.4161818427817616,
+      "grad_norm": 0.26065558195114136,
+      "learning_rate": 9.954679356225832e-05,
+      "loss": 1.3253,
+      "step": 1538
+    },
+    {
+      "epoch": 0.41672304153700446,
+      "grad_norm": 0.27157294750213623,
+      "learning_rate": 9.954255457726354e-05,
+      "loss": 1.3218,
+      "step": 1540
+    },
+    {
+      "epoch": 0.41726424029224735,
+      "grad_norm": 0.2833496630191803,
+      "learning_rate": 9.953829595124715e-05,
+      "loss": 1.32,
+      "step": 1542
+    },
+    {
+      "epoch": 0.4178054390474902,
+      "grad_norm": 0.2757824659347534,
+      "learning_rate": 9.953401768589745e-05,
+      "loss": 1.3165,
+      "step": 1544
+    },
+    {
+      "epoch": 0.41834663780273307,
+      "grad_norm": 0.2609362304210663,
+      "learning_rate": 9.952971978291059e-05,
+      "loss": 1.3229,
+      "step": 1546
+    },
+    {
+      "epoch": 0.4188878365579759,
+      "grad_norm": 0.2863214313983917,
+      "learning_rate": 9.952540224399043e-05,
+      "loss": 1.3217,
+      "step": 1548
+    },
+    {
+      "epoch": 0.4194290353132188,
+      "grad_norm": 0.27573657035827637,
+      "learning_rate": 9.952106507084864e-05,
+      "loss": 1.3151,
+      "step": 1550
+    },
+    {
+      "epoch": 0.4199702340684616,
+      "grad_norm": 0.26843398809432983,
+      "learning_rate": 9.95167082652047e-05,
+      "loss": 1.3185,
+      "step": 1552
+    },
+    {
+      "epoch": 0.4205114328237045,
+      "grad_norm": 0.25903749465942383,
+      "learning_rate": 9.951233182878585e-05,
+      "loss": 1.3142,
+      "step": 1554
+    },
+    {
+      "epoch": 0.42105263157894735,
+      "grad_norm": 0.27221450209617615,
+      "learning_rate": 9.950793576332713e-05,
+      "loss": 1.3119,
+      "step": 1556
+    },
+    {
+      "epoch": 0.42159383033419023,
+      "grad_norm": 0.2897038161754608,
+      "learning_rate": 9.950352007057134e-05,
+      "loss": 1.3217,
+      "step": 1558
+    },
+    {
+      "epoch": 0.42213502908943307,
+      "grad_norm": 0.2515231668949127,
+      "learning_rate": 9.949908475226905e-05,
+      "loss": 1.3263,
+      "step": 1560
+    },
+    {
+      "epoch": 0.42267622784467596,
+      "grad_norm": 0.26686710119247437,
+      "learning_rate": 9.949462981017865e-05,
+      "loss": 1.3269,
+      "step": 1562
+    },
+    {
+      "epoch": 0.42321742659991884,
+      "grad_norm": 0.2747204899787903,
+      "learning_rate": 9.949015524606629e-05,
+      "loss": 1.3208,
+      "step": 1564
+    },
+    {
+      "epoch": 0.4237586253551617,
+      "grad_norm": 0.25866395235061646,
+      "learning_rate": 9.948566106170589e-05,
+      "loss": 1.3273,
+      "step": 1566
+    },
+    {
+      "epoch": 0.42429982411040457,
+      "grad_norm": 0.2659189999103546,
+      "learning_rate": 9.948114725887918e-05,
+      "loss": 1.2955,
+      "step": 1568
+    },
+    {
+      "epoch": 0.4248410228656474,
+      "grad_norm": 0.25262853503227234,
+      "learning_rate": 9.947661383937563e-05,
+      "loss": 1.284,
+      "step": 1570
+    },
+    {
+      "epoch": 0.4253822216208903,
+      "grad_norm": 0.24780422449111938,
+      "learning_rate": 9.94720608049925e-05,
+      "loss": 1.3168,
+      "step": 1572
+    },
+    {
+      "epoch": 0.4259234203761331,
+      "grad_norm": 0.2663845121860504,
+      "learning_rate": 9.946748815753484e-05,
+      "loss": 1.313,
+      "step": 1574
+    },
+    {
+      "epoch": 0.426464619131376,
+      "grad_norm": 0.2906511425971985,
+      "learning_rate": 9.946289589881545e-05,
+      "loss": 1.3197,
+      "step": 1576
+    },
+    {
+      "epoch": 0.42700581788661884,
+      "grad_norm": 0.28401264548301697,
+      "learning_rate": 9.945828403065493e-05,
+      "loss": 1.3254,
+      "step": 1578
+    },
+    {
+      "epoch": 0.42754701664186173,
+      "grad_norm": 0.27820122241973877,
+      "learning_rate": 9.945365255488164e-05,
+      "loss": 1.3153,
+      "step": 1580
+    },
+    {
+      "epoch": 0.42808821539710457,
+      "grad_norm": 0.2573559880256653,
+      "learning_rate": 9.944900147333173e-05,
+      "loss": 1.3144,
+      "step": 1582
+    },
+    {
+      "epoch": 0.42862941415234745,
+      "grad_norm": 0.2536357343196869,
+      "learning_rate": 9.944433078784909e-05,
+      "loss": 1.3172,
+      "step": 1584
+    },
+    {
+      "epoch": 0.4291706129075903,
+      "grad_norm": 0.2745160758495331,
+      "learning_rate": 9.94396405002854e-05,
+      "loss": 1.3023,
+      "step": 1586
+    },
+    {
+      "epoch": 0.4297118116628332,
+      "grad_norm": 0.290393203496933,
+      "learning_rate": 9.943493061250013e-05,
+      "loss": 1.3095,
+      "step": 1588
+    },
+    {
+      "epoch": 0.43025301041807607,
+      "grad_norm": 0.29357218742370605,
+      "learning_rate": 9.94302011263605e-05,
+      "loss": 1.3232,
+      "step": 1590
+    },
+    {
+      "epoch": 0.4307942091733189,
+      "grad_norm": 0.2756180167198181,
+      "learning_rate": 9.94254520437415e-05,
+      "loss": 1.3179,
+      "step": 1592
+    },
+    {
+      "epoch": 0.4313354079285618,
+      "grad_norm": 0.30225417017936707,
+      "learning_rate": 9.942068336652589e-05,
+      "loss": 1.3353,
+      "step": 1594
+    },
+    {
+      "epoch": 0.4318766066838046,
+      "grad_norm": 0.26694637537002563,
+      "learning_rate": 9.94158950966042e-05,
+      "loss": 1.318,
+      "step": 1596
+    },
+    {
+      "epoch": 0.4324178054390475,
+      "grad_norm": 0.2528863549232483,
+      "learning_rate": 9.941108723587471e-05,
+      "loss": 1.3282,
+      "step": 1598
+    },
+    {
+      "epoch": 0.43295900419429034,
+      "grad_norm": 0.25261232256889343,
+      "learning_rate": 9.940625978624353e-05,
+      "loss": 1.3178,
+      "step": 1600
+    },
+    {
+      "epoch": 0.43350020294953323,
+      "grad_norm": 0.2624775767326355,
+      "learning_rate": 9.940141274962444e-05,
+      "loss": 1.31,
+      "step": 1602
+    },
+    {
+      "epoch": 0.43404140170477606,
+      "grad_norm": 0.260810524225235,
+      "learning_rate": 9.939654612793908e-05,
+      "loss": 1.3162,
+      "step": 1604
+    },
+    {
+      "epoch": 0.43458260046001895,
+      "grad_norm": 0.2815745174884796,
+      "learning_rate": 9.939165992311676e-05,
+      "loss": 1.3112,
+      "step": 1606
+    },
+    {
+      "epoch": 0.4351237992152618,
+      "grad_norm": 0.2773973345756531,
+      "learning_rate": 9.938675413709466e-05,
+      "loss": 1.3,
+      "step": 1608
+    },
+    {
+      "epoch": 0.4356649979705047,
+      "grad_norm": 0.26486915349960327,
+      "learning_rate": 9.938182877181763e-05,
+      "loss": 1.3193,
+      "step": 1610
+    },
+    {
+      "epoch": 0.4362061967257475,
+      "grad_norm": 0.26103830337524414,
+      "learning_rate": 9.937688382923832e-05,
+      "loss": 1.3244,
+      "step": 1612
+    },
+    {
+      "epoch": 0.4367473954809904,
+      "grad_norm": 0.2556493878364563,
+      "learning_rate": 9.937191931131716e-05,
+      "loss": 1.3087,
+      "step": 1614
+    },
+    {
+      "epoch": 0.43728859423623323,
+      "grad_norm": 0.2739090919494629,
+      "learning_rate": 9.93669352200223e-05,
+      "loss": 1.3009,
+      "step": 1616
+    },
+    {
+      "epoch": 0.4378297929914761,
+      "grad_norm": 0.26297444105148315,
+      "learning_rate": 9.936193155732967e-05,
+      "loss": 1.2971,
+      "step": 1618
+    },
+    {
+      "epoch": 0.438370991746719,
+      "grad_norm": 0.2587411403656006,
+      "learning_rate": 9.935690832522297e-05,
+      "loss": 1.3259,
+      "step": 1620
+    },
+    {
+      "epoch": 0.43891219050196184,
+      "grad_norm": 0.2419731616973877,
+      "learning_rate": 9.935186552569366e-05,
+      "loss": 1.3123,
+      "step": 1622
+    },
+    {
+      "epoch": 0.43945338925720473,
+      "grad_norm": 0.27424389123916626,
+      "learning_rate": 9.934680316074092e-05,
+      "loss": 1.3196,
+      "step": 1624
+    },
+    {
+      "epoch": 0.43999458801244756,
+      "grad_norm": 0.258242666721344,
+      "learning_rate": 9.934172123237173e-05,
+      "loss": 1.3044,
+      "step": 1626
+    },
+    {
+      "epoch": 0.44053578676769045,
+      "grad_norm": 0.2621035575866699,
+      "learning_rate": 9.933661974260078e-05,
+      "loss": 1.3111,
+      "step": 1628
+    },
+    {
+      "epoch": 0.4410769855229333,
+      "grad_norm": 0.25349390506744385,
+      "learning_rate": 9.93314986934506e-05,
+      "loss": 1.3025,
+      "step": 1630
+    },
+    {
+      "epoch": 0.4416181842781762,
+      "grad_norm": 0.2615620195865631,
+      "learning_rate": 9.932635808695136e-05,
+      "loss": 1.3291,
+      "step": 1632
+    },
+    {
+      "epoch": 0.442159383033419,
+      "grad_norm": 0.2933880686759949,
+      "learning_rate": 9.932119792514105e-05,
+      "loss": 1.3327,
+      "step": 1634
+    },
+    {
+      "epoch": 0.4427005817886619,
+      "grad_norm": 0.2584700286388397,
+      "learning_rate": 9.931601821006544e-05,
+      "loss": 1.3031,
+      "step": 1636
+    },
+    {
+      "epoch": 0.44324178054390473,
+      "grad_norm": 0.2718084156513214,
+      "learning_rate": 9.931081894377797e-05,
+      "loss": 1.3053,
+      "step": 1638
+    },
+    {
+      "epoch": 0.4437829792991476,
+      "grad_norm": 0.27105703949928284,
+      "learning_rate": 9.93056001283399e-05,
+      "loss": 1.3012,
+      "step": 1640
+    },
+    {
+      "epoch": 0.44432417805439045,
+      "grad_norm": 0.27265292406082153,
+      "learning_rate": 9.930036176582021e-05,
+      "loss": 1.2957,
+      "step": 1642
+    },
+    {
+      "epoch": 0.44486537680963334,
+      "grad_norm": 0.26121169328689575,
+      "learning_rate": 9.929510385829564e-05,
+      "loss": 1.3062,
+      "step": 1644
+    },
+    {
+      "epoch": 0.44540657556487623,
+      "grad_norm": 0.26841971278190613,
+      "learning_rate": 9.928982640785067e-05,
+      "loss": 1.3192,
+      "step": 1646
+    },
+    {
+      "epoch": 0.44594777432011906,
+      "grad_norm": 0.27634862065315247,
+      "learning_rate": 9.928452941657755e-05,
+      "loss": 1.3005,
+      "step": 1648
+    },
+    {
+      "epoch": 0.44648897307536195,
+      "grad_norm": 0.25527122616767883,
+      "learning_rate": 9.927921288657623e-05,
+      "loss": 1.3121,
+      "step": 1650
+    },
+    {
+      "epoch": 0.4470301718306048,
+      "grad_norm": 0.2733294665813446,
+      "learning_rate": 9.927387681995443e-05,
+      "loss": 1.3051,
+      "step": 1652
+    },
+    {
+      "epoch": 0.44757137058584767,
+      "grad_norm": 0.2783257067203522,
+      "learning_rate": 9.926852121882766e-05,
+      "loss": 1.2947,
+      "step": 1654
+    },
+    {
+      "epoch": 0.4481125693410905,
+      "grad_norm": 0.2672583758831024,
+      "learning_rate": 9.926314608531911e-05,
+      "loss": 1.3272,
+      "step": 1656
+    },
+    {
+      "epoch": 0.4486537680963334,
+      "grad_norm": 0.2568219304084778,
+      "learning_rate": 9.925775142155974e-05,
+      "loss": 1.3025,
+      "step": 1658
+    },
+    {
+      "epoch": 0.4491949668515762,
+      "grad_norm": 0.2576539218425751,
+      "learning_rate": 9.925233722968826e-05,
+      "loss": 1.2715,
+      "step": 1660
+    },
+    {
+      "epoch": 0.4497361656068191,
+      "grad_norm": 0.25898897647857666,
+      "learning_rate": 9.924690351185109e-05,
+      "loss": 1.3039,
+      "step": 1662
+    },
+    {
+      "epoch": 0.45027736436206195,
+      "grad_norm": 0.25795668363571167,
+      "learning_rate": 9.924145027020242e-05,
+      "loss": 1.3115,
+      "step": 1664
+    },
+    {
+      "epoch": 0.45081856311730484,
+      "grad_norm": 0.2781166136264801,
+      "learning_rate": 9.92359775069042e-05,
+      "loss": 1.3017,
+      "step": 1666
+    },
+    {
+      "epoch": 0.45135976187254767,
+      "grad_norm": 0.2871512770652771,
+      "learning_rate": 9.923048522412608e-05,
+      "loss": 1.3206,
+      "step": 1668
+    },
+    {
+      "epoch": 0.45190096062779056,
+      "grad_norm": 0.27760595083236694,
+      "learning_rate": 9.922497342404544e-05,
+      "loss": 1.3214,
+      "step": 1670
+    },
+    {
+      "epoch": 0.4524421593830334,
+      "grad_norm": 0.26959067583084106,
+      "learning_rate": 9.921944210884746e-05,
+      "loss": 1.3144,
+      "step": 1672
+    },
+    {
+      "epoch": 0.4529833581382763,
+      "grad_norm": 0.2662011384963989,
+      "learning_rate": 9.921389128072498e-05,
+      "loss": 1.3022,
+      "step": 1674
+    },
+    {
+      "epoch": 0.45352455689351917,
+      "grad_norm": 0.28014811873435974,
+      "learning_rate": 9.920832094187861e-05,
+      "loss": 1.3104,
+      "step": 1676
+    },
+    {
+      "epoch": 0.454065755648762,
+      "grad_norm": 0.2560974955558777,
+      "learning_rate": 9.920273109451673e-05,
+      "loss": 1.3113,
+      "step": 1678
+    },
+    {
+      "epoch": 0.4546069544040049,
+      "grad_norm": 0.285339891910553,
+      "learning_rate": 9.91971217408554e-05,
+      "loss": 1.3126,
+      "step": 1680
+    },
+    {
+      "epoch": 0.4551481531592477,
+      "grad_norm": 0.29105204343795776,
+      "learning_rate": 9.919149288311843e-05,
+      "loss": 1.3248,
+      "step": 1682
+    },
+    {
+      "epoch": 0.4556893519144906,
+      "grad_norm": 0.2868146002292633,
+      "learning_rate": 9.918584452353739e-05,
+      "loss": 1.3217,
+      "step": 1684
+    },
+    {
+      "epoch": 0.45623055066973345,
+      "grad_norm": 0.26717278361320496,
+      "learning_rate": 9.918017666435152e-05,
+      "loss": 1.2991,
+      "step": 1686
+    },
+    {
+      "epoch": 0.45677174942497634,
+      "grad_norm": 0.2560403048992157,
+      "learning_rate": 9.917448930780786e-05,
+      "loss": 1.3091,
+      "step": 1688
+    },
+    {
+      "epoch": 0.45731294818021917,
+      "grad_norm": 0.2610042989253998,
+      "learning_rate": 9.916878245616114e-05,
+      "loss": 1.2948,
+      "step": 1690
+    },
+    {
+      "epoch": 0.45785414693546206,
+      "grad_norm": 0.27322304248809814,
+      "learning_rate": 9.916305611167382e-05,
+      "loss": 1.3121,
+      "step": 1692
+    },
+    {
+      "epoch": 0.4583953456907049,
+      "grad_norm": 0.26559844613075256,
+      "learning_rate": 9.91573102766161e-05,
+      "loss": 1.307,
+      "step": 1694
+    },
+    {
+      "epoch": 0.4589365444459478,
+      "grad_norm": 0.2677384316921234,
+      "learning_rate": 9.91515449532659e-05,
+      "loss": 1.2925,
+      "step": 1696
+    },
+    {
+      "epoch": 0.4594777432011906,
+      "grad_norm": 0.2670448422431946,
+      "learning_rate": 9.914576014390888e-05,
+      "loss": 1.3051,
+      "step": 1698
+    },
+    {
+      "epoch": 0.4600189419564335,
+      "grad_norm": 0.2537919878959656,
+      "learning_rate": 9.91399558508384e-05,
+      "loss": 1.3047,
+      "step": 1700
+    },
+    {
+      "epoch": 0.46056014071167634,
+      "grad_norm": 0.2712916433811188,
+      "learning_rate": 9.913413207635555e-05,
+      "loss": 1.2949,
+      "step": 1702
+    },
+    {
+      "epoch": 0.4611013394669192,
+      "grad_norm": 0.27910125255584717,
+      "learning_rate": 9.912828882276917e-05,
+      "loss": 1.336,
+      "step": 1704
+    },
+    {
+      "epoch": 0.4616425382221621,
+      "grad_norm": 0.25917065143585205,
+      "learning_rate": 9.91224260923958e-05,
+      "loss": 1.2938,
+      "step": 1706
+    },
+    {
+      "epoch": 0.46218373697740495,
+      "grad_norm": 0.265024334192276,
+      "learning_rate": 9.91165438875597e-05,
+      "loss": 1.2876,
+      "step": 1708
+    },
+    {
+      "epoch": 0.46272493573264784,
+      "grad_norm": 0.2637651860713959,
+      "learning_rate": 9.911064221059286e-05,
+      "loss": 1.3128,
+      "step": 1710
+    },
+    {
+      "epoch": 0.46326613448789067,
+      "grad_norm": 0.25448864698410034,
+      "learning_rate": 9.910472106383495e-05,
+      "loss": 1.3289,
+      "step": 1712
+    },
+    {
+      "epoch": 0.46380733324313356,
+      "grad_norm": 0.24903124570846558,
+      "learning_rate": 9.909878044963346e-05,
+      "loss": 1.3013,
+      "step": 1714
+    },
+    {
+      "epoch": 0.4643485319983764,
+      "grad_norm": 0.258848637342453,
+      "learning_rate": 9.909282037034347e-05,
+      "loss": 1.3052,
+      "step": 1716
+    },
+    {
+      "epoch": 0.4648897307536193,
+      "grad_norm": 0.25806304812431335,
+      "learning_rate": 9.908684082832787e-05,
+      "loss": 1.286,
+      "step": 1718
+    },
+    {
+      "epoch": 0.4654309295088621,
+      "grad_norm": 0.26794132590293884,
+      "learning_rate": 9.908084182595723e-05,
+      "loss": 1.3069,
+      "step": 1720
+    },
+    {
+      "epoch": 0.465972128264105,
+      "grad_norm": 0.26079118251800537,
+      "learning_rate": 9.907482336560983e-05,
+      "loss": 1.3145,
+      "step": 1722
+    },
+    {
+      "epoch": 0.46651332701934783,
+      "grad_norm": 0.25958481431007385,
+      "learning_rate": 9.906878544967169e-05,
+      "loss": 1.3098,
+      "step": 1724
+    },
+    {
+      "epoch": 0.4670545257745907,
+      "grad_norm": 0.2390812784433365,
+      "learning_rate": 9.906272808053652e-05,
+      "loss": 1.3085,
+      "step": 1726
+    },
+    {
+      "epoch": 0.46759572452983356,
+      "grad_norm": 0.263637900352478,
+      "learning_rate": 9.905665126060574e-05,
+      "loss": 1.2933,
+      "step": 1728
+    },
+    {
+      "epoch": 0.46813692328507645,
+      "grad_norm": 0.2462746798992157,
+      "learning_rate": 9.90505549922885e-05,
+      "loss": 1.2877,
+      "step": 1730
+    },
+    {
+      "epoch": 0.46867812204031933,
+      "grad_norm": 0.244845911860466,
+      "learning_rate": 9.904443927800164e-05,
+      "loss": 1.325,
+      "step": 1732
+    },
+    {
+      "epoch": 0.46921932079556217,
+      "grad_norm": 0.28249332308769226,
+      "learning_rate": 9.903830412016974e-05,
+      "loss": 1.313,
+      "step": 1734
+    },
+    {
+      "epoch": 0.46976051955080506,
+      "grad_norm": 0.29556336998939514,
+      "learning_rate": 9.903214952122504e-05,
+      "loss": 1.3142,
+      "step": 1736
+    },
+    {
+      "epoch": 0.4703017183060479,
+      "grad_norm": 0.2746431827545166,
+      "learning_rate": 9.902597548360754e-05,
+      "loss": 1.3096,
+      "step": 1738
+    },
+    {
+      "epoch": 0.4708429170612908,
+      "grad_norm": 0.2979538142681122,
+      "learning_rate": 9.901978200976492e-05,
+      "loss": 1.2849,
+      "step": 1740
+    },
+    {
+      "epoch": 0.4713841158165336,
+      "grad_norm": 0.2766527235507965,
+      "learning_rate": 9.901356910215255e-05,
+      "loss": 1.3089,
+      "step": 1742
+    },
+    {
+      "epoch": 0.4719253145717765,
+      "grad_norm": 0.25000783801078796,
+      "learning_rate": 9.900733676323353e-05,
+      "loss": 1.308,
+      "step": 1744
+    },
+    {
+      "epoch": 0.47246651332701933,
+      "grad_norm": 0.26226234436035156,
+      "learning_rate": 9.900108499547864e-05,
+      "loss": 1.3041,
+      "step": 1746
+    },
+    {
+      "epoch": 0.4730077120822622,
+      "grad_norm": 0.2794544994831085,
+      "learning_rate": 9.899481380136642e-05,
+      "loss": 1.3312,
+      "step": 1748
+    },
+    {
+      "epoch": 0.47354891083750505,
+      "grad_norm": 0.24771127104759216,
+      "learning_rate": 9.898852318338303e-05,
+      "loss": 1.2853,
+      "step": 1750
+    },
+    {
+      "epoch": 0.47409010959274794,
+      "grad_norm": 0.2811632752418518,
+      "learning_rate": 9.898221314402238e-05,
+      "loss": 1.3019,
+      "step": 1752
+    },
+    {
+      "epoch": 0.4746313083479908,
+      "grad_norm": 0.2812533378601074,
+      "learning_rate": 9.897588368578608e-05,
+      "loss": 1.3298,
+      "step": 1754
+    },
+    {
+      "epoch": 0.47517250710323367,
+      "grad_norm": 0.25955653190612793,
+      "learning_rate": 9.896953481118341e-05,
+      "loss": 1.3093,
+      "step": 1756
+    },
+    {
+      "epoch": 0.4757137058584765,
+      "grad_norm": 0.2653108537197113,
+      "learning_rate": 9.896316652273136e-05,
+      "loss": 1.2898,
+      "step": 1758
+    },
+    {
+      "epoch": 0.4762549046137194,
+      "grad_norm": 0.27985796332359314,
+      "learning_rate": 9.895677882295466e-05,
+      "loss": 1.2928,
+      "step": 1760
+    },
+    {
+      "epoch": 0.4767961033689623,
+      "grad_norm": 0.2889133393764496,
+      "learning_rate": 9.895037171438568e-05,
+      "loss": 1.3088,
+      "step": 1762
+    },
+    {
+      "epoch": 0.4773373021242051,
+      "grad_norm": 0.2615009546279907,
+      "learning_rate": 9.894394519956448e-05,
+      "loss": 1.3212,
+      "step": 1764
+    },
+    {
+      "epoch": 0.477878500879448,
+      "grad_norm": 0.24938960373401642,
+      "learning_rate": 9.893749928103885e-05,
+      "loss": 1.2982,
+      "step": 1766
+    },
+    {
+      "epoch": 0.47841969963469083,
+      "grad_norm": 0.27132853865623474,
+      "learning_rate": 9.893103396136427e-05,
+      "loss": 1.294,
+      "step": 1768
+    },
+    {
+      "epoch": 0.4789608983899337,
+      "grad_norm": 0.2632822096347809,
+      "learning_rate": 9.89245492431039e-05,
+      "loss": 1.2852,
+      "step": 1770
+    },
+    {
+      "epoch": 0.47950209714517655,
+      "grad_norm": 0.27269670367240906,
+      "learning_rate": 9.891804512882856e-05,
+      "loss": 1.2934,
+      "step": 1772
+    },
+    {
+      "epoch": 0.48004329590041944,
+      "grad_norm": 0.2572595179080963,
+      "learning_rate": 9.891152162111683e-05,
+      "loss": 1.2719,
+      "step": 1774
+    },
+    {
+      "epoch": 0.4805844946556623,
+      "grad_norm": 0.2708267867565155,
+      "learning_rate": 9.890497872255489e-05,
+      "loss": 1.2907,
+      "step": 1776
+    },
+    {
+      "epoch": 0.48112569341090516,
+      "grad_norm": 0.28407028317451477,
+      "learning_rate": 9.889841643573671e-05,
+      "loss": 1.2977,
+      "step": 1778
+    },
+    {
+      "epoch": 0.481666892166148,
+      "grad_norm": 0.26248103380203247,
+      "learning_rate": 9.889183476326386e-05,
+      "loss": 1.2993,
+      "step": 1780
+    },
+    {
+      "epoch": 0.4822080909213909,
+      "grad_norm": 0.26148512959480286,
+      "learning_rate": 9.888523370774563e-05,
+      "loss": 1.2893,
+      "step": 1782
+    },
+    {
+      "epoch": 0.4827492896766337,
+      "grad_norm": 0.2815425395965576,
+      "learning_rate": 9.8878613271799e-05,
+      "loss": 1.3015,
+      "step": 1784
+    },
+    {
+      "epoch": 0.4832904884318766,
+      "grad_norm": 0.26061713695526123,
+      "learning_rate": 9.887197345804862e-05,
+      "loss": 1.2781,
+      "step": 1786
+    },
+    {
+      "epoch": 0.4838316871871195,
+      "grad_norm": 0.2641533613204956,
+      "learning_rate": 9.886531426912683e-05,
+      "loss": 1.2993,
+      "step": 1788
+    },
+    {
+      "epoch": 0.48437288594236233,
+      "grad_norm": 0.25920137763023376,
+      "learning_rate": 9.885863570767364e-05,
+      "loss": 1.2955,
+      "step": 1790
+    },
+    {
+      "epoch": 0.4849140846976052,
+      "grad_norm": 0.24002158641815186,
+      "learning_rate": 9.885193777633676e-05,
+      "loss": 1.2932,
+      "step": 1792
+    },
+    {
+      "epoch": 0.48545528345284805,
+      "grad_norm": 0.2643393576145172,
+      "learning_rate": 9.884522047777157e-05,
+      "loss": 1.2963,
+      "step": 1794
+    },
+    {
+      "epoch": 0.48599648220809094,
+      "grad_norm": 0.2522197663784027,
+      "learning_rate": 9.883848381464112e-05,
+      "loss": 1.2947,
+      "step": 1796
+    },
+    {
+      "epoch": 0.4865376809633338,
+      "grad_norm": 0.2431286871433258,
+      "learning_rate": 9.883172778961613e-05,
+      "loss": 1.3112,
+      "step": 1798
+    },
+    {
+      "epoch": 0.48707887971857666,
+      "grad_norm": 0.26892608404159546,
+      "learning_rate": 9.882495240537505e-05,
+      "loss": 1.2904,
+      "step": 1800
+    },
+    {
+      "epoch": 0.4876200784738195,
+      "grad_norm": 0.2528528571128845,
+      "learning_rate": 9.881815766460392e-05,
+      "loss": 1.2949,
+      "step": 1802
+    },
+    {
+      "epoch": 0.4881612772290624,
+      "grad_norm": 0.2614927291870117,
+      "learning_rate": 9.881134356999652e-05,
+      "loss": 1.288,
+      "step": 1804
+    },
+    {
+      "epoch": 0.4887024759843052,
+      "grad_norm": 0.2523605227470398,
+      "learning_rate": 9.880451012425426e-05,
+      "loss": 1.3029,
+      "step": 1806
+    },
+    {
+      "epoch": 0.4892436747395481,
+      "grad_norm": 0.24303248524665833,
+      "learning_rate": 9.879765733008627e-05,
+      "loss": 1.3107,
+      "step": 1808
+    },
+    {
+      "epoch": 0.48978487349479094,
+      "grad_norm": 0.2470557987689972,
+      "learning_rate": 9.879078519020933e-05,
+      "loss": 1.2856,
+      "step": 1810
+    },
+    {
+      "epoch": 0.49032607225003383,
+      "grad_norm": 0.2526317536830902,
+      "learning_rate": 9.878389370734784e-05,
+      "loss": 1.2965,
+      "step": 1812
+    },
+    {
+      "epoch": 0.49086727100527666,
+      "grad_norm": 0.2483314871788025,
+      "learning_rate": 9.877698288423394e-05,
+      "loss": 1.3016,
+      "step": 1814
+    },
+    {
+      "epoch": 0.49140846976051955,
+      "grad_norm": 0.24746839702129364,
+      "learning_rate": 9.877005272360741e-05,
+      "loss": 1.2944,
+      "step": 1816
+    },
+    {
+      "epoch": 0.49194966851576244,
+      "grad_norm": 0.24739988148212433,
+      "learning_rate": 9.876310322821568e-05,
+      "loss": 1.3037,
+      "step": 1818
+    },
+    {
+      "epoch": 0.4924908672710053,
+      "grad_norm": 0.2740204632282257,
+      "learning_rate": 9.875613440081387e-05,
+      "loss": 1.3116,
+      "step": 1820
+    },
+    {
+      "epoch": 0.49303206602624816,
+      "grad_norm": 0.27116379141807556,
+      "learning_rate": 9.874914624416475e-05,
+      "loss": 1.288,
+      "step": 1822
+    },
+    {
+      "epoch": 0.493573264781491,
+      "grad_norm": 0.24231554567813873,
+      "learning_rate": 9.874213876103878e-05,
+      "loss": 1.2975,
+      "step": 1824
+    },
+    {
+      "epoch": 0.4941144635367339,
+      "grad_norm": 0.2590995728969574,
+      "learning_rate": 9.873511195421402e-05,
+      "loss": 1.2678,
+      "step": 1826
+    },
+    {
+      "epoch": 0.4946556622919767,
+      "grad_norm": 0.25694531202316284,
+      "learning_rate": 9.872806582647625e-05,
+      "loss": 1.28,
+      "step": 1828
+    },
+    {
+      "epoch": 0.4951968610472196,
+      "grad_norm": 0.25455620884895325,
+      "learning_rate": 9.87210003806189e-05,
+      "loss": 1.2942,
+      "step": 1830
+    },
+    {
+      "epoch": 0.49573805980246244,
+      "grad_norm": 0.2639889121055603,
+      "learning_rate": 9.871391561944302e-05,
+      "loss": 1.3161,
+      "step": 1832
+    },
+    {
+      "epoch": 0.4962792585577053,
+      "grad_norm": 0.271282821893692,
+      "learning_rate": 9.870681154575737e-05,
+      "loss": 1.3071,
+      "step": 1834
+    },
+    {
+      "epoch": 0.49682045731294816,
+      "grad_norm": 0.26479372382164,
+      "learning_rate": 9.869968816237833e-05,
+      "loss": 1.2841,
+      "step": 1836
+    },
+    {
+      "epoch": 0.49736165606819105,
+      "grad_norm": 0.26040130853652954,
+      "learning_rate": 9.869254547212997e-05,
+      "loss": 1.2989,
+      "step": 1838
+    },
+    {
+      "epoch": 0.4979028548234339,
+      "grad_norm": 0.26563623547554016,
+      "learning_rate": 9.868538347784396e-05,
+      "loss": 1.2965,
+      "step": 1840
+    },
+    {
+      "epoch": 0.49844405357867677,
+      "grad_norm": 0.26089224219322205,
+      "learning_rate": 9.867820218235969e-05,
+      "loss": 1.3071,
+      "step": 1842
+    },
+    {
+      "epoch": 0.4989852523339196,
+      "grad_norm": 0.27151811122894287,
+      "learning_rate": 9.867100158852412e-05,
+      "loss": 1.287,
+      "step": 1844
+    },
+    {
+      "epoch": 0.4995264510891625,
+      "grad_norm": 0.2477792203426361,
+      "learning_rate": 9.866378169919192e-05,
+      "loss": 1.2894,
+      "step": 1846
+    },
+    {
+      "epoch": 0.5000676498444053,
+      "grad_norm": 0.24871942400932312,
+      "learning_rate": 9.865654251722545e-05,
+      "loss": 1.3024,
+      "step": 1848
+    },
+    {
+      "epoch": 0.5006088485996483,
+      "grad_norm": 0.26377877593040466,
+      "learning_rate": 9.86492840454946e-05,
+      "loss": 1.2939,
+      "step": 1850
+    },
+    {
+      "epoch": 0.5011500473548911,
+      "grad_norm": 0.258228063583374,
+      "learning_rate": 9.8642006286877e-05,
+      "loss": 1.291,
+      "step": 1852
+    },
+    {
+      "epoch": 0.5016912461101339,
+      "grad_norm": 0.26982301473617554,
+      "learning_rate": 9.86347092442579e-05,
+      "loss": 1.2845,
+      "step": 1854
+    },
+    {
+      "epoch": 0.5022324448653768,
+      "grad_norm": 0.24094600975513458,
+      "learning_rate": 9.862739292053021e-05,
+      "loss": 1.2744,
+      "step": 1856
+    },
+    {
+      "epoch": 0.5027736436206197,
+      "grad_norm": 0.25840380787849426,
+      "learning_rate": 9.862005731859442e-05,
+      "loss": 1.2966,
+      "step": 1858
+    },
+    {
+      "epoch": 0.5033148423758625,
+      "grad_norm": 0.26734429597854614,
+      "learning_rate": 9.861270244135877e-05,
+      "loss": 1.2856,
+      "step": 1860
+    },
+    {
+      "epoch": 0.5038560411311054,
+      "grad_norm": 0.24431397020816803,
+      "learning_rate": 9.860532829173903e-05,
+      "loss": 1.2871,
+      "step": 1862
+    },
+    {
+      "epoch": 0.5043972398863482,
+      "grad_norm": 0.25425857305526733,
+      "learning_rate": 9.859793487265869e-05,
+      "loss": 1.2822,
+      "step": 1864
+    },
+    {
+      "epoch": 0.5049384386415912,
+      "grad_norm": 0.25332111120224,
+      "learning_rate": 9.859052218704885e-05,
+      "loss": 1.2723,
+      "step": 1866
+    },
+    {
+      "epoch": 0.505479637396834,
+      "grad_norm": 0.24775418639183044,
+      "learning_rate": 9.858309023784826e-05,
+      "loss": 1.2934,
+      "step": 1868
+    },
+    {
+      "epoch": 0.5060208361520768,
+      "grad_norm": 0.24880458414554596,
+      "learning_rate": 9.857563902800328e-05,
+      "loss": 1.3041,
+      "step": 1870
+    },
+    {
+      "epoch": 0.5065620349073197,
+      "grad_norm": 0.2574135959148407,
+      "learning_rate": 9.856816856046793e-05,
+      "loss": 1.2855,
+      "step": 1872
+    },
+    {
+      "epoch": 0.5071032336625626,
+      "grad_norm": 0.26873350143432617,
+      "learning_rate": 9.856067883820386e-05,
+      "loss": 1.3055,
+      "step": 1874
+    },
+    {
+      "epoch": 0.5076444324178054,
+      "grad_norm": 0.23742420971393585,
+      "learning_rate": 9.855316986418036e-05,
+      "loss": 1.3029,
+      "step": 1876
+    },
+    {
+      "epoch": 0.5081856311730483,
+      "grad_norm": 0.2398921549320221,
+      "learning_rate": 9.854564164137432e-05,
+      "loss": 1.2849,
+      "step": 1878
+    },
+    {
+      "epoch": 0.5087268299282912,
+      "grad_norm": 0.25182288885116577,
+      "learning_rate": 9.85380941727703e-05,
+      "loss": 1.2981,
+      "step": 1880
+    },
+    {
+      "epoch": 0.509268028683534,
+      "grad_norm": 0.23373378813266754,
+      "learning_rate": 9.853052746136048e-05,
+      "loss": 1.2772,
+      "step": 1882
+    },
+    {
+      "epoch": 0.5098092274387769,
+      "grad_norm": 0.2581213712692261,
+      "learning_rate": 9.852294151014466e-05,
+      "loss": 1.3147,
+      "step": 1884
+    },
+    {
+      "epoch": 0.5103504261940197,
+      "grad_norm": 0.26642751693725586,
+      "learning_rate": 9.851533632213028e-05,
+      "loss": 1.2885,
+      "step": 1886
+    },
+    {
+      "epoch": 0.5108916249492627,
+      "grad_norm": 0.24029181897640228,
+      "learning_rate": 9.850771190033237e-05,
+      "loss": 1.297,
+      "step": 1888
+    },
+    {
+      "epoch": 0.5114328237045055,
+      "grad_norm": 0.2555221915245056,
+      "learning_rate": 9.850006824777364e-05,
+      "loss": 1.284,
+      "step": 1890
+    },
+    {
+      "epoch": 0.5119740224597483,
+      "grad_norm": 0.2723660171031952,
+      "learning_rate": 9.849240536748439e-05,
+      "loss": 1.2821,
+      "step": 1892
+    },
+    {
+      "epoch": 0.5125152212149912,
+      "grad_norm": 0.24772705137729645,
+      "learning_rate": 9.848472326250253e-05,
+      "loss": 1.2743,
+      "step": 1894
+    },
+    {
+      "epoch": 0.5130564199702341,
+      "grad_norm": 0.2344834804534912,
+      "learning_rate": 9.847702193587365e-05,
+      "loss": 1.286,
+      "step": 1896
+    },
+    {
+      "epoch": 0.5135976187254769,
+      "grad_norm": 0.23948362469673157,
+      "learning_rate": 9.846930139065088e-05,
+      "loss": 1.2673,
+      "step": 1898
+    },
+    {
+      "epoch": 0.5141388174807198,
+      "grad_norm": 0.27207908034324646,
+      "learning_rate": 9.846156162989503e-05,
+      "loss": 1.3041,
+      "step": 1900
+    },
+    {
+      "epoch": 0.5146800162359627,
+      "grad_norm": 0.2407965511083603,
+      "learning_rate": 9.845380265667454e-05,
+      "loss": 1.2875,
+      "step": 1902
+    },
+    {
+      "epoch": 0.5152212149912055,
+      "grad_norm": 0.2517203688621521,
+      "learning_rate": 9.844602447406538e-05,
+      "loss": 1.2855,
+      "step": 1904
+    },
+    {
+      "epoch": 0.5157624137464484,
+      "grad_norm": 0.24267178773880005,
+      "learning_rate": 9.843822708515123e-05,
+      "loss": 1.2711,
+      "step": 1906
+    },
+    {
+      "epoch": 0.5163036125016912,
+      "grad_norm": 0.23933006823062897,
+      "learning_rate": 9.843041049302331e-05,
+      "loss": 1.3094,
+      "step": 1908
+    },
+    {
+      "epoch": 0.5168448112569342,
+      "grad_norm": 0.21948301792144775,
+      "learning_rate": 9.842257470078054e-05,
+      "loss": 1.2686,
+      "step": 1910
+    },
+    {
+      "epoch": 0.517386010012177,
+      "grad_norm": 0.239594966173172,
+      "learning_rate": 9.841471971152933e-05,
+      "loss": 1.2959,
+      "step": 1912
+    },
+    {
+      "epoch": 0.5179272087674198,
+      "grad_norm": 0.26850634813308716,
+      "learning_rate": 9.840684552838385e-05,
+      "loss": 1.2969,
+      "step": 1914
+    },
+    {
+      "epoch": 0.5184684075226627,
+      "grad_norm": 0.26066869497299194,
+      "learning_rate": 9.839895215446573e-05,
+      "loss": 1.2935,
+      "step": 1916
+    },
+    {
+      "epoch": 0.5190096062779056,
+      "grad_norm": 0.25288596749305725,
+      "learning_rate": 9.839103959290433e-05,
+      "loss": 1.2922,
+      "step": 1918
+    },
+    {
+      "epoch": 0.5195508050331484,
+      "grad_norm": 0.24453966319561005,
+      "learning_rate": 9.838310784683655e-05,
+      "loss": 1.3058,
+      "step": 1920
+    },
+    {
+      "epoch": 0.5200920037883913,
+      "grad_norm": 0.25353509187698364,
+      "learning_rate": 9.837515691940689e-05,
+      "loss": 1.3161,
+      "step": 1922
+    },
+    {
+      "epoch": 0.5206332025436341,
+      "grad_norm": 0.24898375570774078,
+      "learning_rate": 9.836718681376749e-05,
+      "loss": 1.2925,
+      "step": 1924
+    },
+    {
+      "epoch": 0.521174401298877,
+      "grad_norm": 0.2576977014541626,
+      "learning_rate": 9.835919753307807e-05,
+      "loss": 1.2916,
+      "step": 1926
+    },
+    {
+      "epoch": 0.5217156000541199,
+      "grad_norm": 0.25432518124580383,
+      "learning_rate": 9.8351189080506e-05,
+      "loss": 1.2866,
+      "step": 1928
+    },
+    {
+      "epoch": 0.5222567988093627,
+      "grad_norm": 0.2504200339317322,
+      "learning_rate": 9.834316145922615e-05,
+      "loss": 1.2728,
+      "step": 1930
+    },
+    {
+      "epoch": 0.5227979975646057,
+      "grad_norm": 0.2627692222595215,
+      "learning_rate": 9.83351146724211e-05,
+      "loss": 1.2853,
+      "step": 1932
+    },
+    {
+      "epoch": 0.5233391963198485,
+      "grad_norm": 0.2776716351509094,
+      "learning_rate": 9.832704872328094e-05,
+      "loss": 1.2881,
+      "step": 1934
+    },
+    {
+      "epoch": 0.5238803950750913,
+      "grad_norm": 0.24669450521469116,
+      "learning_rate": 9.831896361500344e-05,
+      "loss": 1.2681,
+      "step": 1936
+    },
+    {
+      "epoch": 0.5244215938303342,
+      "grad_norm": 0.24949464201927185,
+      "learning_rate": 9.831085935079387e-05,
+      "loss": 1.2851,
+      "step": 1938
+    },
+    {
+      "epoch": 0.5249627925855771,
+      "grad_norm": 0.2585392892360687,
+      "learning_rate": 9.830273593386518e-05,
+      "loss": 1.2796,
+      "step": 1940
+    },
+    {
+      "epoch": 0.5255039913408199,
+      "grad_norm": 0.26086801290512085,
+      "learning_rate": 9.829459336743787e-05,
+      "loss": 1.293,
+      "step": 1942
+    },
+    {
+      "epoch": 0.5260451900960628,
+      "grad_norm": 0.25490057468414307,
+      "learning_rate": 9.828643165474006e-05,
+      "loss": 1.2824,
+      "step": 1944
+    },
+    {
+      "epoch": 0.5265863888513056,
+      "grad_norm": 0.24865177273750305,
+      "learning_rate": 9.827825079900739e-05,
+      "loss": 1.2835,
+      "step": 1946
+    },
+    {
+      "epoch": 0.5271275876065485,
+      "grad_norm": 0.25498902797698975,
+      "learning_rate": 9.827005080348317e-05,
+      "loss": 1.2931,
+      "step": 1948
+    },
+    {
+      "epoch": 0.5276687863617914,
+      "grad_norm": 0.2585375905036926,
+      "learning_rate": 9.826183167141828e-05,
+      "loss": 1.2659,
+      "step": 1950
+    },
+    {
+      "epoch": 0.5282099851170342,
+      "grad_norm": 0.2300305813550949,
+      "learning_rate": 9.825359340607116e-05,
+      "loss": 1.3019,
+      "step": 1952
+    },
+    {
+      "epoch": 0.528751183872277,
+      "grad_norm": 0.24674038589000702,
+      "learning_rate": 9.824533601070784e-05,
+      "loss": 1.2784,
+      "step": 1954
+    },
+    {
+      "epoch": 0.52929238262752,
+      "grad_norm": 0.23458759486675262,
+      "learning_rate": 9.823705948860195e-05,
+      "loss": 1.2779,
+      "step": 1956
+    },
+    {
+      "epoch": 0.5298335813827628,
+      "grad_norm": 0.24736309051513672,
+      "learning_rate": 9.822876384303472e-05,
+      "loss": 1.3083,
+      "step": 1958
+    },
+    {
+      "epoch": 0.5303747801380057,
+      "grad_norm": 0.25108450651168823,
+      "learning_rate": 9.82204490772949e-05,
+      "loss": 1.3044,
+      "step": 1960
+    },
+    {
+      "epoch": 0.5309159788932486,
+      "grad_norm": 0.23308375477790833,
+      "learning_rate": 9.82121151946789e-05,
+      "loss": 1.2694,
+      "step": 1962
+    },
+    {
+      "epoch": 0.5314571776484914,
+      "grad_norm": 0.2283206284046173,
+      "learning_rate": 9.820376219849064e-05,
+      "loss": 1.2735,
+      "step": 1964
+    },
+    {
+      "epoch": 0.5319983764037343,
+      "grad_norm": 0.24121573567390442,
+      "learning_rate": 9.819539009204164e-05,
+      "loss": 1.2799,
+      "step": 1966
+    },
+    {
+      "epoch": 0.5325395751589771,
+      "grad_norm": 0.24135661125183105,
+      "learning_rate": 9.8186998878651e-05,
+      "loss": 1.295,
+      "step": 1968
+    },
+    {
+      "epoch": 0.53308077391422,
+      "grad_norm": 0.24390241503715515,
+      "learning_rate": 9.817858856164542e-05,
+      "loss": 1.2812,
+      "step": 1970
+    },
+    {
+      "epoch": 0.5336219726694629,
+      "grad_norm": 0.24739502370357513,
+      "learning_rate": 9.817015914435913e-05,
+      "loss": 1.2872,
+      "step": 1972
+    },
+    {
+      "epoch": 0.5341631714247057,
+      "grad_norm": 0.25517916679382324,
+      "learning_rate": 9.816171063013395e-05,
+      "loss": 1.2718,
+      "step": 1974
+    },
+    {
+      "epoch": 0.5347043701799485,
+      "grad_norm": 0.25479528307914734,
+      "learning_rate": 9.815324302231928e-05,
+      "loss": 1.2952,
+      "step": 1976
+    },
+    {
+      "epoch": 0.5352455689351915,
+      "grad_norm": 0.24998174607753754,
+      "learning_rate": 9.814475632427206e-05,
+      "loss": 1.2914,
+      "step": 1978
+    },
+    {
+      "epoch": 0.5357867676904343,
+      "grad_norm": 0.2341603934764862,
+      "learning_rate": 9.813625053935686e-05,
+      "loss": 1.2793,
+      "step": 1980
+    },
+    {
+      "epoch": 0.5363279664456772,
+      "grad_norm": 0.23716285824775696,
+      "learning_rate": 9.812772567094574e-05,
+      "loss": 1.2872,
+      "step": 1982
+    },
+    {
+      "epoch": 0.53686916520092,
+      "grad_norm": 0.2324230819940567,
+      "learning_rate": 9.81191817224184e-05,
+      "loss": 1.2604,
+      "step": 1984
+    },
+    {
+      "epoch": 0.5374103639561629,
+      "grad_norm": 0.24399405717849731,
+      "learning_rate": 9.811061869716205e-05,
+      "loss": 1.2972,
+      "step": 1986
+    },
+    {
+      "epoch": 0.5379515627114058,
+      "grad_norm": 0.24572497606277466,
+      "learning_rate": 9.810203659857145e-05,
+      "loss": 1.2784,
+      "step": 1988
+    },
+    {
+      "epoch": 0.5384927614666486,
+      "grad_norm": 0.22993844747543335,
+      "learning_rate": 9.8093435430049e-05,
+      "loss": 1.2886,
+      "step": 1990
+    },
+    {
+      "epoch": 0.5390339602218915,
+      "grad_norm": 0.24518661201000214,
+      "learning_rate": 9.808481519500458e-05,
+      "loss": 1.2622,
+      "step": 1992
+    },
+    {
+      "epoch": 0.5395751589771344,
+      "grad_norm": 0.2601888179779053,
+      "learning_rate": 9.807617589685568e-05,
+      "loss": 1.2739,
+      "step": 1994
+    },
+    {
+      "epoch": 0.5401163577323772,
+      "grad_norm": 0.24736261367797852,
+      "learning_rate": 9.80675175390273e-05,
+      "loss": 1.2748,
+      "step": 1996
+    },
+    {
+      "epoch": 0.54065755648762,
+      "grad_norm": 0.2332574725151062,
+      "learning_rate": 9.805884012495203e-05,
+      "loss": 1.2639,
+      "step": 1998
+    },
+    {
+      "epoch": 0.541198755242863,
+      "grad_norm": 0.2662294805049896,
+      "learning_rate": 9.805014365807004e-05,
+      "loss": 1.2914,
+      "step": 2000
+    },
+    {
+      "epoch": 0.5417399539981058,
+      "grad_norm": 0.28600943088531494,
+      "learning_rate": 9.804142814182902e-05,
+      "loss": 1.2657,
+      "step": 2002
+    },
+    {
+      "epoch": 0.5422811527533486,
+      "grad_norm": 0.2814892530441284,
+      "learning_rate": 9.803269357968416e-05,
+      "loss": 1.2839,
+      "step": 2004
+    },
+    {
+      "epoch": 0.5428223515085915,
+      "grad_norm": 0.24939605593681335,
+      "learning_rate": 9.802393997509833e-05,
+      "loss": 1.2692,
+      "step": 2006
+    },
+    {
+      "epoch": 0.5433635502638344,
+      "grad_norm": 0.2562806308269501,
+      "learning_rate": 9.801516733154181e-05,
+      "loss": 1.291,
+      "step": 2008
+    },
+    {
+      "epoch": 0.5439047490190773,
+      "grad_norm": 0.2617442011833191,
+      "learning_rate": 9.800637565249255e-05,
+      "loss": 1.2808,
+      "step": 2010
+    },
+    {
+      "epoch": 0.5444459477743201,
+      "grad_norm": 0.2421412616968155,
+      "learning_rate": 9.799756494143593e-05,
+      "loss": 1.2733,
+      "step": 2012
+    },
+    {
+      "epoch": 0.5449871465295629,
+      "grad_norm": 0.25231024622917175,
+      "learning_rate": 9.798873520186497e-05,
+      "loss": 1.2695,
+      "step": 2014
+    },
+    {
+      "epoch": 0.5455283452848059,
+      "grad_norm": 0.25108659267425537,
+      "learning_rate": 9.79798864372802e-05,
+      "loss": 1.298,
+      "step": 2016
+    },
+    {
+      "epoch": 0.5460695440400487,
+      "grad_norm": 0.24615678191184998,
+      "learning_rate": 9.79710186511897e-05,
+      "loss": 1.3127,
+      "step": 2018
+    },
+    {
+      "epoch": 0.5466107427952915,
+      "grad_norm": 0.23436503112316132,
+      "learning_rate": 9.796213184710904e-05,
+      "loss": 1.2896,
+      "step": 2020
+    },
+    {
+      "epoch": 0.5471519415505345,
+      "grad_norm": 0.23453901708126068,
+      "learning_rate": 9.79532260285614e-05,
+      "loss": 1.2761,
+      "step": 2022
+    },
+    {
+      "epoch": 0.5476931403057773,
+      "grad_norm": 0.2413233071565628,
+      "learning_rate": 9.794430119907748e-05,
+      "loss": 1.2744,
+      "step": 2024
+    },
+    {
+      "epoch": 0.5482343390610201,
+      "grad_norm": 0.2426893562078476,
+      "learning_rate": 9.793535736219546e-05,
+      "loss": 1.2615,
+      "step": 2026
+    },
+    {
+      "epoch": 0.548775537816263,
+      "grad_norm": 0.23853014409542084,
+      "learning_rate": 9.792639452146115e-05,
+      "loss": 1.2897,
+      "step": 2028
+    },
+    {
+      "epoch": 0.5493167365715059,
+      "grad_norm": 0.24866445362567902,
+      "learning_rate": 9.791741268042784e-05,
+      "loss": 1.2957,
+      "step": 2030
+    },
+    {
+      "epoch": 0.5498579353267488,
+      "grad_norm": 0.24467822909355164,
+      "learning_rate": 9.790841184265633e-05,
+      "loss": 1.2867,
+      "step": 2032
+    },
+    {
+      "epoch": 0.5503991340819916,
+      "grad_norm": 0.2393324077129364,
+      "learning_rate": 9.7899392011715e-05,
+      "loss": 1.3061,
+      "step": 2034
+    },
+    {
+      "epoch": 0.5509403328372344,
+      "grad_norm": 0.23834531009197235,
+      "learning_rate": 9.789035319117974e-05,
+      "loss": 1.2957,
+      "step": 2036
+    },
+    {
+      "epoch": 0.5514815315924774,
+      "grad_norm": 0.2603852450847626,
+      "learning_rate": 9.788129538463397e-05,
+      "loss": 1.2897,
+      "step": 2038
+    },
+    {
+      "epoch": 0.5520227303477202,
+      "grad_norm": 0.26540425419807434,
+      "learning_rate": 9.787221859566861e-05,
+      "loss": 1.2829,
+      "step": 2040
+    },
+    {
+      "epoch": 0.552563929102963,
+      "grad_norm": 0.25125250220298767,
+      "learning_rate": 9.786312282788216e-05,
+      "loss": 1.2708,
+      "step": 2042
+    },
+    {
+      "epoch": 0.553105127858206,
+      "grad_norm": 0.23911471664905548,
+      "learning_rate": 9.785400808488061e-05,
+      "loss": 1.2949,
+      "step": 2044
+    },
+    {
+      "epoch": 0.5536463266134488,
+      "grad_norm": 0.23871150612831116,
+      "learning_rate": 9.784487437027746e-05,
+      "loss": 1.2863,
+      "step": 2046
+    },
+    {
+      "epoch": 0.5541875253686916,
+      "grad_norm": 0.25253376364707947,
+      "learning_rate": 9.783572168769376e-05,
+      "loss": 1.2797,
+      "step": 2048
+    },
+    {
+      "epoch": 0.5547287241239345,
+      "grad_norm": 0.25140559673309326,
+      "learning_rate": 9.782655004075807e-05,
+      "loss": 1.2666,
+      "step": 2050
+    },
+    {
+      "epoch": 0.5552699228791774,
+      "grad_norm": 0.25297242403030396,
+      "learning_rate": 9.781735943310646e-05,
+      "loss": 1.2935,
+      "step": 2052
+    },
+    {
+      "epoch": 0.5558111216344203,
+      "grad_norm": 0.28536322712898254,
+      "learning_rate": 9.780814986838252e-05,
+      "loss": 1.2891,
+      "step": 2054
+    },
+    {
+      "epoch": 0.5563523203896631,
+      "grad_norm": 0.28267911076545715,
+      "learning_rate": 9.779892135023738e-05,
+      "loss": 1.2846,
+      "step": 2056
+    },
+    {
+      "epoch": 0.5568935191449059,
+      "grad_norm": 0.24850498139858246,
+      "learning_rate": 9.778967388232964e-05,
+      "loss": 1.2823,
+      "step": 2058
+    },
+    {
+      "epoch": 0.5574347179001489,
+      "grad_norm": 0.4929364025592804,
+      "learning_rate": 9.778040746832544e-05,
+      "loss": 1.2681,
+      "step": 2060
+    },
+    {
+      "epoch": 0.5579759166553917,
+      "grad_norm": 0.25423306226730347,
+      "learning_rate": 9.777112211189843e-05,
+      "loss": 1.2765,
+      "step": 2062
+    },
+    {
+      "epoch": 0.5585171154106345,
+      "grad_norm": 0.23608753085136414,
+      "learning_rate": 9.776181781672977e-05,
+      "loss": 1.2756,
+      "step": 2064
+    },
+    {
+      "epoch": 0.5590583141658774,
+      "grad_norm": 0.3117451071739197,
+      "learning_rate": 9.775249458650812e-05,
+      "loss": 1.2731,
+      "step": 2066
+    },
+    {
+      "epoch": 0.5595995129211203,
+      "grad_norm": 0.2454603612422943,
+      "learning_rate": 9.774315242492965e-05,
+      "loss": 1.2821,
+      "step": 2068
+    },
+    {
+      "epoch": 0.5601407116763631,
+      "grad_norm": 0.3214171528816223,
+      "learning_rate": 9.773379133569804e-05,
+      "loss": 1.2964,
+      "step": 2070
+    },
+    {
+      "epoch": 0.560681910431606,
+      "grad_norm": 0.23589906096458435,
+      "learning_rate": 9.772441132252448e-05,
+      "loss": 1.2794,
+      "step": 2072
+    },
+    {
+      "epoch": 0.5612231091868489,
+      "grad_norm": 0.23020370304584503,
+      "learning_rate": 9.771501238912763e-05,
+      "loss": 1.2753,
+      "step": 2074
+    },
+    {
+      "epoch": 0.5617643079420918,
+      "grad_norm": 0.2368050515651703,
+      "learning_rate": 9.77055945392337e-05,
+      "loss": 1.3048,
+      "step": 2076
+    },
+    {
+      "epoch": 0.5623055066973346,
+      "grad_norm": 0.2581866383552551,
+      "learning_rate": 9.769615777657633e-05,
+      "loss": 1.2765,
+      "step": 2078
+    },
+    {
+      "epoch": 0.5628467054525774,
+      "grad_norm": 0.2481439858675003,
+      "learning_rate": 9.768670210489675e-05,
+      "loss": 1.2957,
+      "step": 2080
+    },
+    {
+      "epoch": 0.5633879042078204,
+      "grad_norm": 0.2861919701099396,
+      "learning_rate": 9.767722752794361e-05,
+      "loss": 1.2647,
+      "step": 2082
+    },
+    {
+      "epoch": 0.5639291029630632,
+      "grad_norm": 0.2552880346775055,
+      "learning_rate": 9.766773404947309e-05,
+      "loss": 1.2675,
+      "step": 2084
+    },
+    {
+      "epoch": 0.564470301718306,
+      "grad_norm": 0.251891165971756,
+      "learning_rate": 9.765822167324885e-05,
+      "loss": 1.2799,
+      "step": 2086
+    },
+    {
+      "epoch": 0.5650115004735489,
+      "grad_norm": 0.25395113229751587,
+      "learning_rate": 9.764869040304205e-05,
+      "loss": 1.2916,
+      "step": 2088
+    },
+    {
+      "epoch": 0.5655526992287918,
+      "grad_norm": 0.2496347427368164,
+      "learning_rate": 9.763914024263136e-05,
+      "loss": 1.2722,
+      "step": 2090
+    },
+    {
+      "epoch": 0.5660938979840346,
+      "grad_norm": 0.24722573161125183,
+      "learning_rate": 9.762957119580287e-05,
+      "loss": 1.2722,
+      "step": 2092
+    },
+    {
+      "epoch": 0.5666350967392775,
+      "grad_norm": 0.23567502200603485,
+      "learning_rate": 9.761998326635026e-05,
+      "loss": 1.2681,
+      "step": 2094
+    },
+    {
+      "epoch": 0.5671762954945203,
+      "grad_norm": 0.2396802455186844,
+      "learning_rate": 9.76103764580746e-05,
+      "loss": 1.2509,
+      "step": 2096
+    },
+    {
+      "epoch": 0.5677174942497633,
+      "grad_norm": 0.24394263327121735,
+      "learning_rate": 9.76007507747845e-05,
+      "loss": 1.2863,
+      "step": 2098
+    },
+    {
+      "epoch": 0.5682586930050061,
+      "grad_norm": 0.23184406757354736,
+      "learning_rate": 9.759110622029604e-05,
+      "loss": 1.2827,
+      "step": 2100
+    },
+    {
+      "epoch": 0.5687998917602489,
+      "grad_norm": 0.24522008001804352,
+      "learning_rate": 9.758144279843275e-05,
+      "loss": 1.2716,
+      "step": 2102
+    },
+    {
+      "epoch": 0.5693410905154919,
+      "grad_norm": 0.2323451191186905,
+      "learning_rate": 9.757176051302573e-05,
+      "loss": 1.2747,
+      "step": 2104
+    },
+    {
+      "epoch": 0.5698822892707347,
+      "grad_norm": 0.26049986481666565,
+      "learning_rate": 9.756205936791344e-05,
+      "loss": 1.2976,
+      "step": 2106
+    },
+    {
+      "epoch": 0.5704234880259775,
+      "grad_norm": 0.24207298457622528,
+      "learning_rate": 9.75523393669419e-05,
+      "loss": 1.2637,
+      "step": 2108
+    },
+    {
+      "epoch": 0.5709646867812204,
+      "grad_norm": 0.23590189218521118,
+      "learning_rate": 9.754260051396459e-05,
+      "loss": 1.2725,
+      "step": 2110
+    },
+    {
+      "epoch": 0.5715058855364633,
+      "grad_norm": 0.25714561343193054,
+      "learning_rate": 9.753284281284243e-05,
+      "loss": 1.2751,
+      "step": 2112
+    },
+    {
+      "epoch": 0.5720470842917061,
+      "grad_norm": 0.2563743591308594,
+      "learning_rate": 9.752306626744385e-05,
+      "loss": 1.2552,
+      "step": 2114
+    },
+    {
+      "epoch": 0.572588283046949,
+      "grad_norm": 0.23137059807777405,
+      "learning_rate": 9.751327088164474e-05,
+      "loss": 1.2826,
+      "step": 2116
+    },
+    {
+      "epoch": 0.5731294818021918,
+      "grad_norm": 0.23491452634334564,
+      "learning_rate": 9.750345665932844e-05,
+      "loss": 1.2909,
+      "step": 2118
+    },
+    {
+      "epoch": 0.5736706805574348,
+      "grad_norm": 0.23278982937335968,
+      "learning_rate": 9.749362360438579e-05,
+      "loss": 1.2904,
+      "step": 2120
+    },
+    {
+      "epoch": 0.5742118793126776,
+      "grad_norm": 0.22500935196876526,
+      "learning_rate": 9.748377172071508e-05,
+      "loss": 1.2822,
+      "step": 2122
+    },
+    {
+      "epoch": 0.5747530780679204,
+      "grad_norm": 0.23745082318782806,
+      "learning_rate": 9.747390101222205e-05,
+      "loss": 1.279,
+      "step": 2124
+    },
+    {
+      "epoch": 0.5752942768231633,
+      "grad_norm": 0.24000118672847748,
+      "learning_rate": 9.746401148281993e-05,
+      "loss": 1.2806,
+      "step": 2126
+    },
+    {
+      "epoch": 0.5758354755784062,
+      "grad_norm": 0.24468575417995453,
+      "learning_rate": 9.74541031364294e-05,
+      "loss": 1.2867,
+      "step": 2128
+    },
+    {
+      "epoch": 0.576376674333649,
+      "grad_norm": 0.23120936751365662,
+      "learning_rate": 9.744417597697859e-05,
+      "loss": 1.2666,
+      "step": 2130
+    },
+    {
+      "epoch": 0.5769178730888919,
+      "grad_norm": 0.25929006934165955,
+      "learning_rate": 9.743423000840309e-05,
+      "loss": 1.2672,
+      "step": 2132
+    },
+    {
+      "epoch": 0.5774590718441348,
+      "grad_norm": 0.2530214786529541,
+      "learning_rate": 9.742426523464598e-05,
+      "loss": 1.2769,
+      "step": 2134
+    },
+    {
+      "epoch": 0.5780002705993776,
+      "grad_norm": 0.2752387225627899,
+      "learning_rate": 9.741428165965775e-05,
+      "loss": 1.2562,
+      "step": 2136
+    },
+    {
+      "epoch": 0.5785414693546205,
+      "grad_norm": 0.25394052267074585,
+      "learning_rate": 9.740427928739638e-05,
+      "loss": 1.2758,
+      "step": 2138
+    },
+    {
+      "epoch": 0.5790826681098633,
+      "grad_norm": 0.25311702489852905,
+      "learning_rate": 9.739425812182728e-05,
+      "loss": 1.2603,
+      "step": 2140
+    },
+    {
+      "epoch": 0.5796238668651063,
+      "grad_norm": 0.26108497381210327,
+      "learning_rate": 9.738421816692329e-05,
+      "loss": 1.2627,
+      "step": 2142
+    },
+    {
+      "epoch": 0.5801650656203491,
+      "grad_norm": 0.2541772425174713,
+      "learning_rate": 9.737415942666476e-05,
+      "loss": 1.2752,
+      "step": 2144
+    },
+    {
+      "epoch": 0.5807062643755919,
+      "grad_norm": 0.24984823167324066,
+      "learning_rate": 9.736408190503943e-05,
+      "loss": 1.2673,
+      "step": 2146
+    },
+    {
+      "epoch": 0.5812474631308348,
+      "grad_norm": 0.2763904333114624,
+      "learning_rate": 9.735398560604251e-05,
+      "loss": 1.2936,
+      "step": 2148
+    },
+    {
+      "epoch": 0.5817886618860777,
+      "grad_norm": 0.26247066259384155,
+      "learning_rate": 9.734387053367669e-05,
+      "loss": 1.2636,
+      "step": 2150
+    },
+    {
+      "epoch": 0.5823298606413205,
+      "grad_norm": 0.27749454975128174,
+      "learning_rate": 9.7333736691952e-05,
+      "loss": 1.2857,
+      "step": 2152
+    },
+    {
+      "epoch": 0.5828710593965634,
+      "grad_norm": 0.39380860328674316,
+      "learning_rate": 9.732358408488602e-05,
+      "loss": 1.2916,
+      "step": 2154
+    },
+    {
+      "epoch": 0.5834122581518062,
+      "grad_norm": 0.2505074441432953,
+      "learning_rate": 9.731341271650372e-05,
+      "loss": 1.2548,
+      "step": 2156
+    },
+    {
+      "epoch": 0.5839534569070491,
+      "grad_norm": 0.2549828588962555,
+      "learning_rate": 9.730322259083751e-05,
+      "loss": 1.2884,
+      "step": 2158
+    },
+    {
+      "epoch": 0.584494655662292,
+      "grad_norm": 0.24714533984661102,
+      "learning_rate": 9.729301371192724e-05,
+      "loss": 1.2823,
+      "step": 2160
+    },
+    {
+      "epoch": 0.5850358544175348,
+      "grad_norm": 0.24945247173309326,
+      "learning_rate": 9.728278608382018e-05,
+      "loss": 1.2976,
+      "step": 2162
+    },
+    {
+      "epoch": 0.5855770531727778,
+      "grad_norm": 0.2512315511703491,
+      "learning_rate": 9.727253971057109e-05,
+      "loss": 1.2883,
+      "step": 2164
+    },
+    {
+      "epoch": 0.5861182519280206,
+      "grad_norm": 0.2401745468378067,
+      "learning_rate": 9.726227459624207e-05,
+      "loss": 1.2637,
+      "step": 2166
+    },
+    {
+      "epoch": 0.5866594506832634,
+      "grad_norm": 0.260251522064209,
+      "learning_rate": 9.725199074490271e-05,
+      "loss": 1.2618,
+      "step": 2168
+    },
+    {
+      "epoch": 0.5872006494385063,
+      "grad_norm": 0.2533782124519348,
+      "learning_rate": 9.724168816063004e-05,
+      "loss": 1.2825,
+      "step": 2170
+    },
+    {
+      "epoch": 0.5877418481937492,
+      "grad_norm": 0.2545458972454071,
+      "learning_rate": 9.723136684750847e-05,
+      "loss": 1.2784,
+      "step": 2172
+    },
+    {
+      "epoch": 0.588283046948992,
+      "grad_norm": 0.24370916187763214,
+      "learning_rate": 9.722102680962988e-05,
+      "loss": 1.2601,
+      "step": 2174
+    },
+    {
+      "epoch": 0.5888242457042349,
+      "grad_norm": 0.23707440495491028,
+      "learning_rate": 9.721066805109353e-05,
+      "loss": 1.2818,
+      "step": 2176
+    },
+    {
+      "epoch": 0.5893654444594777,
+      "grad_norm": 0.22903890907764435,
+      "learning_rate": 9.720029057600615e-05,
+      "loss": 1.2686,
+      "step": 2178
+    },
+    {
+      "epoch": 0.5899066432147206,
+      "grad_norm": 0.22820548713207245,
+      "learning_rate": 9.718989438848182e-05,
+      "loss": 1.2749,
+      "step": 2180
+    },
+    {
+      "epoch": 0.5904478419699635,
+      "grad_norm": 0.2249859720468521,
+      "learning_rate": 9.717947949264214e-05,
+      "loss": 1.2649,
+      "step": 2182
+    },
+    {
+      "epoch": 0.5909890407252063,
+      "grad_norm": 0.23568090796470642,
+      "learning_rate": 9.716904589261602e-05,
+      "loss": 1.2764,
+      "step": 2184
+    },
+    {
+      "epoch": 0.5915302394804492,
+      "grad_norm": 0.24089080095291138,
+      "learning_rate": 9.715859359253987e-05,
+      "loss": 1.2801,
+      "step": 2186
+    },
+    {
+      "epoch": 0.5920714382356921,
+      "grad_norm": 0.2259254902601242,
+      "learning_rate": 9.714812259655746e-05,
+      "loss": 1.2805,
+      "step": 2188
+    },
+    {
+      "epoch": 0.5926126369909349,
+      "grad_norm": 0.23276519775390625,
+      "learning_rate": 9.713763290881999e-05,
+      "loss": 1.2635,
+      "step": 2190
+    },
+    {
+      "epoch": 0.5931538357461777,
+      "grad_norm": 0.24884091317653656,
+      "learning_rate": 9.712712453348607e-05,
+      "loss": 1.2984,
+      "step": 2192
+    },
+    {
+      "epoch": 0.5936950345014207,
+      "grad_norm": 0.23471422493457794,
+      "learning_rate": 9.711659747472171e-05,
+      "loss": 1.2742,
+      "step": 2194
+    },
+    {
+      "epoch": 0.5942362332566635,
+      "grad_norm": 0.25790145993232727,
+      "learning_rate": 9.710605173670037e-05,
+      "loss": 1.2865,
+      "step": 2196
+    },
+    {
+      "epoch": 0.5947774320119064,
+      "grad_norm": 0.24584504961967468,
+      "learning_rate": 9.709548732360285e-05,
+      "loss": 1.2826,
+      "step": 2198
+    },
+    {
+      "epoch": 0.5953186307671492,
+      "grad_norm": 0.23682548105716705,
+      "learning_rate": 9.708490423961741e-05,
+      "loss": 1.2499,
+      "step": 2200
+    },
+    {
+      "epoch": 0.5958598295223921,
+      "grad_norm": 0.24267072975635529,
+      "learning_rate": 9.707430248893964e-05,
+      "loss": 1.2514,
+      "step": 2202
+    },
+    {
+      "epoch": 0.596401028277635,
+      "grad_norm": 0.2546815276145935,
+      "learning_rate": 9.706368207577264e-05,
+      "loss": 1.2755,
+      "step": 2204
+    },
+    {
+      "epoch": 0.5969422270328778,
+      "grad_norm": 0.24322691559791565,
+      "learning_rate": 9.70530430043268e-05,
+      "loss": 1.2817,
+      "step": 2206
+    },
+    {
+      "epoch": 0.5974834257881206,
+      "grad_norm": 0.22995691001415253,
+      "learning_rate": 9.704238527882e-05,
+      "loss": 1.2487,
+      "step": 2208
+    },
+    {
+      "epoch": 0.5980246245433636,
+      "grad_norm": 0.25768396258354187,
+      "learning_rate": 9.70317089034774e-05,
+      "loss": 1.2956,
+      "step": 2210
+    },
+    {
+      "epoch": 0.5985658232986064,
+      "grad_norm": 0.2691928744316101,
+      "learning_rate": 9.702101388253167e-05,
+      "loss": 1.2704,
+      "step": 2212
+    },
+    {
+      "epoch": 0.5991070220538492,
+      "grad_norm": 0.2356652021408081,
+      "learning_rate": 9.701030022022282e-05,
+      "loss": 1.2548,
+      "step": 2214
+    },
+    {
+      "epoch": 0.5996482208090922,
+      "grad_norm": 0.24094751477241516,
+      "learning_rate": 9.699956792079825e-05,
+      "loss": 1.2616,
+      "step": 2216
+    },
+    {
+      "epoch": 0.600189419564335,
+      "grad_norm": 0.2429209202528,
+      "learning_rate": 9.698881698851274e-05,
+      "loss": 1.2603,
+      "step": 2218
+    },
+    {
+      "epoch": 0.6007306183195779,
+      "grad_norm": 0.24263691902160645,
+      "learning_rate": 9.69780474276285e-05,
+      "loss": 1.2788,
+      "step": 2220
+    },
+    {
+      "epoch": 0.6012718170748207,
+      "grad_norm": 0.25438740849494934,
+      "learning_rate": 9.696725924241506e-05,
+      "loss": 1.2823,
+      "step": 2222
+    },
+    {
+      "epoch": 0.6018130158300636,
+      "grad_norm": 0.258472204208374,
+      "learning_rate": 9.695645243714939e-05,
+      "loss": 1.2673,
+      "step": 2224
+    },
+    {
+      "epoch": 0.6023542145853065,
+      "grad_norm": 0.2571878433227539,
+      "learning_rate": 9.694562701611583e-05,
+      "loss": 1.295,
+      "step": 2226
+    },
+    {
+      "epoch": 0.6028954133405493,
+      "grad_norm": 0.2430989295244217,
+      "learning_rate": 9.693478298360607e-05,
+      "loss": 1.2595,
+      "step": 2228
+    },
+    {
+      "epoch": 0.6034366120957921,
+      "grad_norm": 0.23489908874034882,
+      "learning_rate": 9.692392034391922e-05,
+      "loss": 1.2773,
+      "step": 2230
+    },
+    {
+      "epoch": 0.6039778108510351,
+      "grad_norm": 0.2507382035255432,
+      "learning_rate": 9.691303910136171e-05,
+      "loss": 1.2782,
+      "step": 2232
+    },
+    {
+      "epoch": 0.6045190096062779,
+      "grad_norm": 0.23355726897716522,
+      "learning_rate": 9.690213926024743e-05,
+      "loss": 1.263,
+      "step": 2234
+    },
+    {
+      "epoch": 0.6050602083615207,
+      "grad_norm": 0.2275291532278061,
+      "learning_rate": 9.689122082489754e-05,
+      "loss": 1.2677,
+      "step": 2236
+    },
+    {
+      "epoch": 0.6056014071167636,
+      "grad_norm": 0.2314850389957428,
+      "learning_rate": 9.688028379964068e-05,
+      "loss": 1.2646,
+      "step": 2238
+    },
+    {
+      "epoch": 0.6061426058720065,
+      "grad_norm": 0.24879969656467438,
+      "learning_rate": 9.686932818881278e-05,
+      "loss": 1.2704,
+      "step": 2240
+    },
+    {
+      "epoch": 0.6066838046272494,
+      "grad_norm": 0.23156161606311798,
+      "learning_rate": 9.685835399675717e-05,
+      "loss": 1.2795,
+      "step": 2242
+    },
+    {
+      "epoch": 0.6072250033824922,
+      "grad_norm": 0.22952421009540558,
+      "learning_rate": 9.684736122782454e-05,
+      "loss": 1.2597,
+      "step": 2244
+    },
+    {
+      "epoch": 0.6077662021377351,
+      "grad_norm": 1.0009911060333252,
+      "learning_rate": 9.683634988637293e-05,
+      "loss": 1.2504,
+      "step": 2246
+    },
+    {
+      "epoch": 0.608307400892978,
+      "grad_norm": 0.2649003267288208,
+      "learning_rate": 9.682531997676777e-05,
+      "loss": 1.2376,
+      "step": 2248
+    },
+    {
+      "epoch": 0.6088485996482208,
+      "grad_norm": 0.31321823596954346,
+      "learning_rate": 9.681427150338187e-05,
+      "loss": 1.2607,
+      "step": 2250
+    },
+    {
+      "epoch": 0.6093897984034636,
+      "grad_norm": 0.3142634332180023,
+      "learning_rate": 9.680320447059532e-05,
+      "loss": 1.261,
+      "step": 2252
+    },
+    {
+      "epoch": 0.6099309971587066,
+      "grad_norm": 0.31247085332870483,
+      "learning_rate": 9.679211888279565e-05,
+      "loss": 1.2685,
+      "step": 2254
+    },
+    {
+      "epoch": 0.6104721959139494,
+      "grad_norm": 0.25763556361198425,
+      "learning_rate": 9.67810147443777e-05,
+      "loss": 1.2542,
+      "step": 2256
+    },
+    {
+      "epoch": 0.6110133946691922,
+      "grad_norm": 0.2788141667842865,
+      "learning_rate": 9.676989205974367e-05,
+      "loss": 1.2747,
+      "step": 2258
+    },
+    {
+      "epoch": 0.6115545934244351,
+      "grad_norm": 0.26279813051223755,
+      "learning_rate": 9.675875083330315e-05,
+      "loss": 1.261,
+      "step": 2260
+    },
+    {
+      "epoch": 0.612095792179678,
+      "grad_norm": 0.24764376878738403,
+      "learning_rate": 9.674759106947302e-05,
+      "loss": 1.2632,
+      "step": 2262
+    },
+    {
+      "epoch": 0.6126369909349209,
+      "grad_norm": 0.2378121018409729,
+      "learning_rate": 9.673641277267756e-05,
+      "loss": 1.2569,
+      "step": 2264
+    },
+    {
+      "epoch": 0.6131781896901637,
+      "grad_norm": 0.25457054376602173,
+      "learning_rate": 9.672521594734838e-05,
+      "loss": 1.2667,
+      "step": 2266
+    },
+    {
+      "epoch": 0.6137193884454065,
+      "grad_norm": 0.2589806616306305,
+      "learning_rate": 9.67140005979244e-05,
+      "loss": 1.2515,
+      "step": 2268
+    },
+    {
+      "epoch": 0.6142605872006495,
+      "grad_norm": 0.23375307023525238,
+      "learning_rate": 9.670276672885195e-05,
+      "loss": 1.2608,
+      "step": 2270
+    },
+    {
+      "epoch": 0.6148017859558923,
+      "grad_norm": 0.22978229820728302,
+      "learning_rate": 9.669151434458468e-05,
+      "loss": 1.2516,
+      "step": 2272
+    },
+    {
+      "epoch": 0.6153429847111351,
+      "grad_norm": 0.22958585619926453,
+      "learning_rate": 9.668024344958353e-05,
+      "loss": 1.2617,
+      "step": 2274
+    },
+    {
+      "epoch": 0.6158841834663781,
+      "grad_norm": 0.22783328592777252,
+      "learning_rate": 9.666895404831685e-05,
+      "loss": 1.2732,
+      "step": 2276
+    },
+    {
+      "epoch": 0.6164253822216209,
+      "grad_norm": 0.2413301318883896,
+      "learning_rate": 9.665764614526027e-05,
+      "loss": 1.2501,
+      "step": 2278
+    },
+    {
+      "epoch": 0.6169665809768637,
+      "grad_norm": 0.2591399550437927,
+      "learning_rate": 9.66463197448968e-05,
+      "loss": 1.2657,
+      "step": 2280
+    },
+    {
+      "epoch": 0.6175077797321066,
+      "grad_norm": 0.23001301288604736,
+      "learning_rate": 9.663497485171675e-05,
+      "loss": 1.2698,
+      "step": 2282
+    },
+    {
+      "epoch": 0.6180489784873495,
+      "grad_norm": 0.21373826265335083,
+      "learning_rate": 9.662361147021779e-05,
+      "loss": 1.2651,
+      "step": 2284
+    },
+    {
+      "epoch": 0.6185901772425924,
+      "grad_norm": 0.2302403599023819,
+      "learning_rate": 9.66122296049049e-05,
+      "loss": 1.2786,
+      "step": 2286
+    },
+    {
+      "epoch": 0.6191313759978352,
+      "grad_norm": 0.24121953547000885,
+      "learning_rate": 9.660082926029038e-05,
+      "loss": 1.2639,
+      "step": 2288
+    },
+    {
+      "epoch": 0.619672574753078,
+      "grad_norm": 0.22190925478935242,
+      "learning_rate": 9.658941044089387e-05,
+      "loss": 1.2472,
+      "step": 2290
+    },
+    {
+      "epoch": 0.620213773508321,
+      "grad_norm": 0.22907890379428864,
+      "learning_rate": 9.657797315124234e-05,
+      "loss": 1.2475,
+      "step": 2292
+    },
+    {
+      "epoch": 0.6207549722635638,
+      "grad_norm": 0.23821642994880676,
+      "learning_rate": 9.656651739587008e-05,
+      "loss": 1.2689,
+      "step": 2294
+    },
+    {
+      "epoch": 0.6212961710188066,
+      "grad_norm": 0.25953301787376404,
+      "learning_rate": 9.655504317931867e-05,
+      "loss": 1.2587,
+      "step": 2296
+    },
+    {
+      "epoch": 0.6218373697740495,
+      "grad_norm": 0.24054677784442902,
+      "learning_rate": 9.654355050613705e-05,
+      "loss": 1.2541,
+      "step": 2298
+    },
+    {
+      "epoch": 0.6223785685292924,
+      "grad_norm": 0.23474815487861633,
+      "learning_rate": 9.65320393808815e-05,
+      "loss": 1.2605,
+      "step": 2300
+    },
+    {
+      "epoch": 0.6229197672845352,
+      "grad_norm": 0.23981337249279022,
+      "learning_rate": 9.652050980811551e-05,
+      "loss": 1.267,
+      "step": 2302
+    },
+    {
+      "epoch": 0.6234609660397781,
+      "grad_norm": 0.21820946037769318,
+      "learning_rate": 9.650896179240997e-05,
+      "loss": 1.2555,
+      "step": 2304
+    },
+    {
+      "epoch": 0.624002164795021,
+      "grad_norm": 0.2165161520242691,
+      "learning_rate": 9.64973953383431e-05,
+      "loss": 1.271,
+      "step": 2306
+    },
+    {
+      "epoch": 0.6245433635502639,
+      "grad_norm": 0.22105515003204346,
+      "learning_rate": 9.648581045050035e-05,
+      "loss": 1.2663,
+      "step": 2308
+    },
+    {
+      "epoch": 0.6250845623055067,
+      "grad_norm": 0.22626088559627533,
+      "learning_rate": 9.647420713347454e-05,
+      "loss": 1.2645,
+      "step": 2310
+    },
+    {
+      "epoch": 0.6256257610607495,
+      "grad_norm": 0.2347354292869568,
+      "learning_rate": 9.646258539186577e-05,
+      "loss": 1.2372,
+      "step": 2312
+    },
+    {
+      "epoch": 0.6261669598159925,
+      "grad_norm": 0.2388308346271515,
+      "learning_rate": 9.645094523028144e-05,
+      "loss": 1.2652,
+      "step": 2314
+    },
+    {
+      "epoch": 0.6267081585712353,
+      "grad_norm": 0.2252940982580185,
+      "learning_rate": 9.643928665333628e-05,
+      "loss": 1.2595,
+      "step": 2316
+    },
+    {
+      "epoch": 0.6272493573264781,
+      "grad_norm": 0.24020199477672577,
+      "learning_rate": 9.64276096656523e-05,
+      "loss": 1.3079,
+      "step": 2318
+    },
+    {
+      "epoch": 0.627790556081721,
+      "grad_norm": 0.23432402312755585,
+      "learning_rate": 9.64159142718588e-05,
+      "loss": 1.2718,
+      "step": 2320
+    },
+    {
+      "epoch": 0.6283317548369639,
+      "grad_norm": 0.22962002456188202,
+      "learning_rate": 9.640420047659239e-05,
+      "loss": 1.2606,
+      "step": 2322
+    },
+    {
+      "epoch": 0.6288729535922067,
+      "grad_norm": 0.21251855790615082,
+      "learning_rate": 9.6392468284497e-05,
+      "loss": 1.2568,
+      "step": 2324
+    },
+    {
+      "epoch": 0.6294141523474496,
+      "grad_norm": 0.2140374481678009,
+      "learning_rate": 9.63807177002238e-05,
+      "loss": 1.276,
+      "step": 2326
+    },
+    {
+      "epoch": 0.6299553511026925,
+      "grad_norm": 0.21366523206233978,
+      "learning_rate": 9.636894872843132e-05,
+      "loss": 1.2521,
+      "step": 2328
+    },
+    {
+      "epoch": 0.6304965498579354,
+      "grad_norm": 0.22407646477222443,
+      "learning_rate": 9.635716137378528e-05,
+      "loss": 1.2692,
+      "step": 2330
+    },
+    {
+      "epoch": 0.6310377486131782,
+      "grad_norm": 0.24414391815662384,
+      "learning_rate": 9.63453556409588e-05,
+      "loss": 1.2554,
+      "step": 2332
+    },
+    {
+      "epoch": 0.631578947368421,
+      "grad_norm": 0.23787756264209747,
+      "learning_rate": 9.633353153463219e-05,
+      "loss": 1.2471,
+      "step": 2334
+    },
+    {
+      "epoch": 0.632120146123664,
+      "grad_norm": 0.2248927801847458,
+      "learning_rate": 9.63216890594931e-05,
+      "loss": 1.2586,
+      "step": 2336
+    },
+    {
+      "epoch": 0.6326613448789068,
+      "grad_norm": 0.2224208265542984,
+      "learning_rate": 9.630982822023648e-05,
+      "loss": 1.293,
+      "step": 2338
+    },
+    {
+      "epoch": 0.6332025436341496,
+      "grad_norm": 0.2312854379415512,
+      "learning_rate": 9.62979490215645e-05,
+      "loss": 1.2604,
+      "step": 2340
+    },
+    {
+      "epoch": 0.6337437423893925,
+      "grad_norm": 0.22154025733470917,
+      "learning_rate": 9.628605146818665e-05,
+      "loss": 1.2645,
+      "step": 2342
+    },
+    {
+      "epoch": 0.6342849411446354,
+      "grad_norm": 0.2136823982000351,
+      "learning_rate": 9.627413556481968e-05,
+      "loss": 1.2375,
+      "step": 2344
+    },
+    {
+      "epoch": 0.6348261398998782,
+      "grad_norm": 0.21541404724121094,
+      "learning_rate": 9.626220131618763e-05,
+      "loss": 1.2771,
+      "step": 2346
+    },
+    {
+      "epoch": 0.6353673386551211,
+      "grad_norm": 0.22025029361248016,
+      "learning_rate": 9.625024872702178e-05,
+      "loss": 1.261,
+      "step": 2348
+    },
+    {
+      "epoch": 0.6359085374103639,
+      "grad_norm": 0.2375534474849701,
+      "learning_rate": 9.623827780206073e-05,
+      "loss": 1.2808,
+      "step": 2350
+    },
+    {
+      "epoch": 0.6364497361656068,
+      "grad_norm": 0.23297767341136932,
+      "learning_rate": 9.62262885460503e-05,
+      "loss": 1.2697,
+      "step": 2352
+    },
+    {
+      "epoch": 0.6369909349208497,
+      "grad_norm": 0.24082797765731812,
+      "learning_rate": 9.621428096374363e-05,
+      "loss": 1.2347,
+      "step": 2354
+    },
+    {
+      "epoch": 0.6375321336760925,
+      "grad_norm": 0.22009813785552979,
+      "learning_rate": 9.620225505990105e-05,
+      "loss": 1.2631,
+      "step": 2356
+    },
+    {
+      "epoch": 0.6380733324313355,
+      "grad_norm": 0.22501374781131744,
+      "learning_rate": 9.619021083929025e-05,
+      "loss": 1.2563,
+      "step": 2358
+    },
+    {
+      "epoch": 0.6386145311865783,
+      "grad_norm": 0.22494594752788544,
+      "learning_rate": 9.61781483066861e-05,
+      "loss": 1.2532,
+      "step": 2360
+    },
+    {
+      "epoch": 0.6391557299418211,
+      "grad_norm": 0.3569008409976959,
+      "learning_rate": 9.616606746687078e-05,
+      "loss": 1.2684,
+      "step": 2362
+    },
+    {
+      "epoch": 0.639696928697064,
+      "grad_norm": 207.0965576171875,
+      "learning_rate": 9.61539683246337e-05,
+      "loss": 1.3637,
+      "step": 2364
+    },
+    {
+      "epoch": 0.6402381274523069,
+      "grad_norm": 0.4599202573299408,
+      "learning_rate": 9.614185088477152e-05,
+      "loss": 1.292,
+      "step": 2366
+    },
+    {
+      "epoch": 0.6407793262075497,
+      "grad_norm": 0.3244802951812744,
+      "learning_rate": 9.61297151520882e-05,
+      "loss": 1.2585,
+      "step": 2368
+    },
+    {
+      "epoch": 0.6413205249627926,
+      "grad_norm": 0.30332016944885254,
+      "learning_rate": 9.611756113139488e-05,
+      "loss": 1.2619,
+      "step": 2370
+    },
+    {
+      "epoch": 0.6418617237180354,
+      "grad_norm": 0.2982909083366394,
+      "learning_rate": 9.610538882751001e-05,
+      "loss": 1.2637,
+      "step": 2372
+    },
+    {
+      "epoch": 0.6424029224732783,
+      "grad_norm": 5.417288303375244,
+      "learning_rate": 9.609319824525928e-05,
+      "loss": 1.2713,
+      "step": 2374
+    },
+    {
+      "epoch": 0.6429441212285212,
+      "grad_norm": 0.4198252260684967,
+      "learning_rate": 9.608098938947562e-05,
+      "loss": 1.2541,
+      "step": 2376
+    },
+    {
+      "epoch": 0.643485319983764,
+      "grad_norm": 0.8178582191467285,
+      "learning_rate": 9.606876226499918e-05,
+      "loss": 1.2884,
+      "step": 2378
+    },
+    {
+      "epoch": 0.6440265187390068,
+      "grad_norm": 0.33514025807380676,
+      "learning_rate": 9.60565168766774e-05,
+      "loss": 1.2719,
+      "step": 2380
+    },
+    {
+      "epoch": 0.6445677174942498,
+      "grad_norm": 0.2973354756832123,
+      "learning_rate": 9.60442532293649e-05,
+      "loss": 1.2515,
+      "step": 2382
+    },
+    {
+      "epoch": 0.6451089162494926,
+      "grad_norm": 0.4670213758945465,
+      "learning_rate": 9.603197132792359e-05,
+      "loss": 1.2665,
+      "step": 2384
+    },
+    {
+      "epoch": 0.6456501150047355,
+      "grad_norm": 0.3197322189807892,
+      "learning_rate": 9.60196711772226e-05,
+      "loss": 1.2574,
+      "step": 2386
+    },
+    {
+      "epoch": 0.6461913137599784,
+      "grad_norm": 1.1344069242477417,
+      "learning_rate": 9.600735278213828e-05,
+      "loss": 1.2689,
+      "step": 2388
+    },
+    {
+      "epoch": 0.6467325125152212,
+      "grad_norm": 0.5379347801208496,
+      "learning_rate": 9.599501614755425e-05,
+      "loss": 1.249,
+      "step": 2390
+    },
+    {
+      "epoch": 0.6472737112704641,
+      "grad_norm": 0.33201339840888977,
+      "learning_rate": 9.598266127836131e-05,
+      "loss": 1.2729,
+      "step": 2392
+    },
+    {
+      "epoch": 0.6478149100257069,
+      "grad_norm": 8.969808578491211,
+      "learning_rate": 9.597028817945753e-05,
+      "loss": 1.2768,
+      "step": 2394
+    },
+    {
+      "epoch": 0.6483561087809498,
+      "grad_norm": 0.3650411069393158,
+      "learning_rate": 9.595789685574821e-05,
+      "loss": 1.2511,
+      "step": 2396
+    },
+    {
+      "epoch": 0.6488973075361927,
+      "grad_norm": 0.8414996862411499,
+      "learning_rate": 9.594548731214583e-05,
+      "loss": 1.2707,
+      "step": 2398
+    },
+    {
+      "epoch": 0.6494385062914355,
+      "grad_norm": 0.5362874269485474,
+      "learning_rate": 9.593305955357016e-05,
+      "loss": 1.2453,
+      "step": 2400
+    },
+    {
+      "epoch": 0.6499797050466783,
+      "grad_norm": 0.40546804666519165,
+      "learning_rate": 9.592061358494813e-05,
+      "loss": 1.2665,
+      "step": 2402
+    },
+    {
+      "epoch": 0.6505209038019213,
+      "grad_norm": 0.29758453369140625,
+      "learning_rate": 9.590814941121389e-05,
+      "loss": 1.2538,
+      "step": 2404
+    },
+    {
+      "epoch": 0.6510621025571641,
+      "grad_norm": 0.2636415660381317,
+      "learning_rate": 9.589566703730888e-05,
+      "loss": 1.2457,
+      "step": 2406
+    },
+    {
+      "epoch": 0.651603301312407,
+      "grad_norm": 0.2844487130641937,
+      "learning_rate": 9.588316646818168e-05,
+      "loss": 1.257,
+      "step": 2408
+    },
+    {
+      "epoch": 0.6521445000676498,
+      "grad_norm": 0.2777060568332672,
+      "learning_rate": 9.587064770878808e-05,
+      "loss": 1.2506,
+      "step": 2410
+    },
+    {
+      "epoch": 0.6526856988228927,
+      "grad_norm": 0.2585492730140686,
+      "learning_rate": 9.585811076409117e-05,
+      "loss": 1.2472,
+      "step": 2412
+    },
+    {
+      "epoch": 0.6532268975781356,
+      "grad_norm": 0.24312525987625122,
+      "learning_rate": 9.584555563906116e-05,
+      "loss": 1.2703,
+      "step": 2414
+    },
+    {
+      "epoch": 0.6537680963333784,
+      "grad_norm": 0.2286798357963562,
+      "learning_rate": 9.583298233867549e-05,
+      "loss": 1.2582,
+      "step": 2416
+    },
+    {
+      "epoch": 0.6543092950886213,
+      "grad_norm": 0.22804994881153107,
+      "learning_rate": 9.582039086791883e-05,
+      "loss": 1.2538,
+      "step": 2418
+    },
+    {
+      "epoch": 0.6548504938438642,
+      "grad_norm": 0.2244635969400406,
+      "learning_rate": 9.580778123178303e-05,
+      "loss": 1.2481,
+      "step": 2420
+    },
+    {
+      "epoch": 0.655391692599107,
+      "grad_norm": 0.22303158044815063,
+      "learning_rate": 9.579515343526714e-05,
+      "loss": 1.2574,
+      "step": 2422
+    },
+    {
+      "epoch": 0.6559328913543498,
+      "grad_norm": 0.2208811491727829,
+      "learning_rate": 9.578250748337742e-05,
+      "loss": 1.2579,
+      "step": 2424
+    },
+    {
+      "epoch": 0.6564740901095928,
+      "grad_norm": 0.20853403210639954,
+      "learning_rate": 9.576984338112736e-05,
+      "loss": 1.2619,
+      "step": 2426
+    },
+    {
+      "epoch": 0.6570152888648356,
+      "grad_norm": 0.20974035561084747,
+      "learning_rate": 9.575716113353757e-05,
+      "loss": 1.2605,
+      "step": 2428
+    },
+    {
+      "epoch": 0.6575564876200785,
+      "grad_norm": 0.22891463339328766,
+      "learning_rate": 9.57444607456359e-05,
+      "loss": 1.2586,
+      "step": 2430
+    },
+    {
+      "epoch": 0.6580976863753213,
+      "grad_norm": 0.21693287789821625,
+      "learning_rate": 9.57317422224574e-05,
+      "loss": 1.2505,
+      "step": 2432
+    },
+    {
+      "epoch": 0.6586388851305642,
+      "grad_norm": 0.21806494891643524,
+      "learning_rate": 9.57190055690443e-05,
+      "loss": 1.261,
+      "step": 2434
+    },
+    {
+      "epoch": 0.6591800838858071,
+      "grad_norm": 0.24015147984027863,
+      "learning_rate": 9.570625079044601e-05,
+      "loss": 1.2564,
+      "step": 2436
+    },
+    {
+      "epoch": 0.6597212826410499,
+      "grad_norm": 0.26577669382095337,
+      "learning_rate": 9.569347789171912e-05,
+      "loss": 1.2716,
+      "step": 2438
+    },
+    {
+      "epoch": 0.6602624813962927,
+      "grad_norm": 0.2382255643606186,
+      "learning_rate": 9.568068687792741e-05,
+      "loss": 1.2465,
+      "step": 2440
+    },
+    {
+      "epoch": 0.6608036801515357,
+      "grad_norm": 0.22770415246486664,
+      "learning_rate": 9.566787775414188e-05,
+      "loss": 1.2229,
+      "step": 2442
+    },
+    {
+      "epoch": 0.6613448789067785,
+      "grad_norm": 0.23449081182479858,
+      "learning_rate": 9.565505052544065e-05,
+      "loss": 1.2582,
+      "step": 2444
+    },
+    {
+      "epoch": 0.6618860776620213,
+      "grad_norm": 0.22105945646762848,
+      "learning_rate": 9.564220519690903e-05,
+      "loss": 1.2505,
+      "step": 2446
+    },
+    {
+      "epoch": 0.6624272764172643,
+      "grad_norm": 0.22349369525909424,
+      "learning_rate": 9.562934177363953e-05,
+      "loss": 1.2578,
+      "step": 2448
+    },
+    {
+      "epoch": 0.6629684751725071,
+      "grad_norm": 0.23770608007907867,
+      "learning_rate": 9.561646026073184e-05,
+      "loss": 1.2399,
+      "step": 2450
+    },
+    {
+      "epoch": 0.66350967392775,
+      "grad_norm": 0.2204604148864746,
+      "learning_rate": 9.56035606632928e-05,
+      "loss": 1.2512,
+      "step": 2452
+    },
+    {
+      "epoch": 0.6640508726829928,
+      "grad_norm": 0.2204030454158783,
+      "learning_rate": 9.559064298643638e-05,
+      "loss": 1.2821,
+      "step": 2454
+    },
+    {
+      "epoch": 0.6645920714382357,
+      "grad_norm": 0.2169465720653534,
+      "learning_rate": 9.55777072352838e-05,
+      "loss": 1.2529,
+      "step": 2456
+    },
+    {
+      "epoch": 0.6651332701934786,
+      "grad_norm": 0.2273695021867752,
+      "learning_rate": 9.55647534149634e-05,
+      "loss": 1.2497,
+      "step": 2458
+    },
+    {
+      "epoch": 0.6656744689487214,
+      "grad_norm": 0.22077496349811554,
+      "learning_rate": 9.555178153061069e-05,
+      "loss": 1.2433,
+      "step": 2460
+    },
+    {
+      "epoch": 0.6662156677039642,
+      "grad_norm": 0.2203417718410492,
+      "learning_rate": 9.553879158736833e-05,
+      "loss": 1.2464,
+      "step": 2462
+    },
+    {
+      "epoch": 0.6667568664592072,
+      "grad_norm": 0.22205059230327606,
+      "learning_rate": 9.552578359038617e-05,
+      "loss": 1.2611,
+      "step": 2464
+    },
+    {
+      "epoch": 0.66729806521445,
+      "grad_norm": 0.2206515222787857,
+      "learning_rate": 9.551275754482119e-05,
+      "loss": 1.2624,
+      "step": 2466
+    },
+    {
+      "epoch": 0.6678392639696928,
+      "grad_norm": 0.21758343279361725,
+      "learning_rate": 9.549971345583753e-05,
+      "loss": 1.2406,
+      "step": 2468
+    },
+    {
+      "epoch": 0.6683804627249358,
+      "grad_norm": 0.21517138183116913,
+      "learning_rate": 9.548665132860647e-05,
+      "loss": 1.2538,
+      "step": 2470
+    },
+    {
+      "epoch": 0.6689216614801786,
+      "grad_norm": 0.21490350365638733,
+      "learning_rate": 9.547357116830648e-05,
+      "loss": 1.2534,
+      "step": 2472
+    },
+    {
+      "epoch": 0.6694628602354215,
+      "grad_norm": 0.2156359702348709,
+      "learning_rate": 9.546047298012315e-05,
+      "loss": 1.2459,
+      "step": 2474
+    },
+    {
+      "epoch": 0.6700040589906643,
+      "grad_norm": 0.2196791172027588,
+      "learning_rate": 9.544735676924923e-05,
+      "loss": 1.2534,
+      "step": 2476
+    },
+    {
+      "epoch": 0.6705452577459072,
+      "grad_norm": 0.22666549682617188,
+      "learning_rate": 9.54342225408846e-05,
+      "loss": 1.252,
+      "step": 2478
+    },
+    {
+      "epoch": 0.6710864565011501,
+      "grad_norm": 0.2314993143081665,
+      "learning_rate": 9.54210703002363e-05,
+      "loss": 1.2478,
+      "step": 2480
+    },
+    {
+      "epoch": 0.6716276552563929,
+      "grad_norm": 0.2225077599287033,
+      "learning_rate": 9.54079000525185e-05,
+      "loss": 1.2465,
+      "step": 2482
+    },
+    {
+      "epoch": 0.6721688540116357,
+      "grad_norm": 0.22268906235694885,
+      "learning_rate": 9.539471180295249e-05,
+      "loss": 1.2453,
+      "step": 2484
+    },
+    {
+      "epoch": 0.6727100527668787,
+      "grad_norm": 0.30744513869285583,
+      "learning_rate": 9.538150555676677e-05,
+      "loss": 1.2874,
+      "step": 2486
+    },
+    {
+      "epoch": 0.6732512515221215,
+      "grad_norm": 0.27435171604156494,
+      "learning_rate": 9.536828131919686e-05,
+      "loss": 1.2533,
+      "step": 2488
+    },
+    {
+      "epoch": 0.6737924502773643,
+      "grad_norm": 0.5657795667648315,
+      "learning_rate": 9.535503909548553e-05,
+      "loss": 1.2567,
+      "step": 2490
+    },
+    {
+      "epoch": 0.6743336490326072,
+      "grad_norm": 0.4795803129673004,
+      "learning_rate": 9.53417788908826e-05,
+      "loss": 1.2563,
+      "step": 2492
+    },
+    {
+      "epoch": 0.6748748477878501,
+      "grad_norm": 0.3125123977661133,
+      "learning_rate": 9.532850071064503e-05,
+      "loss": 1.251,
+      "step": 2494
+    },
+    {
+      "epoch": 0.675416046543093,
+      "grad_norm": 0.2949443459510803,
+      "learning_rate": 9.531520456003696e-05,
+      "loss": 1.2491,
+      "step": 2496
+    },
+    {
+      "epoch": 0.6759572452983358,
+      "grad_norm": 0.289389967918396,
+      "learning_rate": 9.530189044432959e-05,
+      "loss": 1.2571,
+      "step": 2498
+    },
+    {
+      "epoch": 0.6764984440535787,
+      "grad_norm": 0.24411126971244812,
+      "learning_rate": 9.528855836880127e-05,
+      "loss": 1.2528,
+      "step": 2500
+    },
+    {
+      "epoch": 0.6770396428088216,
+      "grad_norm": 0.38176965713500977,
+      "learning_rate": 9.527520833873748e-05,
+      "loss": 1.2462,
+      "step": 2502
+    },
+    {
+      "epoch": 0.6775808415640644,
+      "grad_norm": 0.25295090675354004,
+      "learning_rate": 9.52618403594308e-05,
+      "loss": 1.2601,
+      "step": 2504
+    },
+    {
+      "epoch": 0.6781220403193072,
+      "grad_norm": 0.24630951881408691,
+      "learning_rate": 9.524845443618091e-05,
+      "loss": 1.2398,
+      "step": 2506
+    },
+    {
+      "epoch": 0.6786632390745502,
+      "grad_norm": 0.25156068801879883,
+      "learning_rate": 9.523505057429466e-05,
+      "loss": 1.2429,
+      "step": 2508
+    },
+    {
+      "epoch": 0.679204437829793,
+      "grad_norm": 0.23003700375556946,
+      "learning_rate": 9.522162877908596e-05,
+      "loss": 1.2569,
+      "step": 2510
+    },
+    {
+      "epoch": 0.6797456365850358,
+      "grad_norm": 0.2248392552137375,
+      "learning_rate": 9.520818905587585e-05,
+      "loss": 1.2506,
+      "step": 2512
+    },
+    {
+      "epoch": 0.6802868353402787,
+      "grad_norm": 0.22383219003677368,
+      "learning_rate": 9.519473140999246e-05,
+      "loss": 1.2294,
+      "step": 2514
+    },
+    {
+      "epoch": 0.6808280340955216,
+      "grad_norm": 0.22723117470741272,
+      "learning_rate": 9.518125584677106e-05,
+      "loss": 1.2658,
+      "step": 2516
+    },
+    {
+      "epoch": 0.6813692328507645,
+      "grad_norm": 0.24425800144672394,
+      "learning_rate": 9.516776237155402e-05,
+      "loss": 1.233,
+      "step": 2518
+    },
+    {
+      "epoch": 0.6819104316060073,
+      "grad_norm": 0.22345170378684998,
+      "learning_rate": 9.515425098969075e-05,
+      "loss": 1.248,
+      "step": 2520
+    },
+    {
+      "epoch": 0.6824516303612501,
+      "grad_norm": 0.21297229826450348,
+      "learning_rate": 9.514072170653782e-05,
+      "loss": 1.2453,
+      "step": 2522
+    },
+    {
+      "epoch": 0.6829928291164931,
+      "grad_norm": 0.21216444671154022,
+      "learning_rate": 9.51271745274589e-05,
+      "loss": 1.2473,
+      "step": 2524
+    },
+    {
+      "epoch": 0.6835340278717359,
+      "grad_norm": 0.2091735154390335,
+      "learning_rate": 9.511360945782472e-05,
+      "loss": 1.2451,
+      "step": 2526
+    },
+    {
+      "epoch": 0.6840752266269787,
+      "grad_norm": 0.21291106939315796,
+      "learning_rate": 9.510002650301313e-05,
+      "loss": 1.2772,
+      "step": 2528
+    },
+    {
+      "epoch": 0.6846164253822217,
+      "grad_norm": 0.21953986585140228,
+      "learning_rate": 9.508642566840901e-05,
+      "loss": 1.2533,
+      "step": 2530
+    },
+    {
+      "epoch": 0.6851576241374645,
+      "grad_norm": 0.21948380768299103,
+      "learning_rate": 9.507280695940446e-05,
+      "loss": 1.2797,
+      "step": 2532
+    },
+    {
+      "epoch": 0.6856988228927073,
+      "grad_norm": 0.21971148252487183,
+      "learning_rate": 9.505917038139851e-05,
+      "loss": 1.2609,
+      "step": 2534
+    },
+    {
+      "epoch": 0.6862400216479502,
+      "grad_norm": 0.21478046476840973,
+      "learning_rate": 9.504551593979738e-05,
+      "loss": 1.2625,
+      "step": 2536
+    },
+    {
+      "epoch": 0.6867812204031931,
+      "grad_norm": 0.21927322447299957,
+      "learning_rate": 9.503184364001431e-05,
+      "loss": 1.2415,
+      "step": 2538
+    },
+    {
+      "epoch": 0.687322419158436,
+      "grad_norm": 0.2084941267967224,
+      "learning_rate": 9.501815348746971e-05,
+      "loss": 1.2455,
+      "step": 2540
+    },
+    {
+      "epoch": 0.6878636179136788,
+      "grad_norm": 0.20336540043354034,
+      "learning_rate": 9.500444548759095e-05,
+      "loss": 1.2505,
+      "step": 2542
+    },
+    {
+      "epoch": 0.6884048166689216,
+      "grad_norm": 0.21661430597305298,
+      "learning_rate": 9.499071964581256e-05,
+      "loss": 1.235,
+      "step": 2544
+    },
+    {
+      "epoch": 0.6889460154241646,
+      "grad_norm": 0.2240605354309082,
+      "learning_rate": 9.497697596757609e-05,
+      "loss": 1.2546,
+      "step": 2546
+    },
+    {
+      "epoch": 0.6894872141794074,
+      "grad_norm": 0.2289547622203827,
+      "learning_rate": 9.496321445833022e-05,
+      "loss": 1.2387,
+      "step": 2548
+    },
+    {
+      "epoch": 0.6900284129346502,
+      "grad_norm": 0.22886811196804047,
+      "learning_rate": 9.494943512353063e-05,
+      "loss": 1.2531,
+      "step": 2550
+    },
+    {
+      "epoch": 0.6905696116898931,
+      "grad_norm": 0.2151922732591629,
+      "learning_rate": 9.493563796864014e-05,
+      "loss": 1.2447,
+      "step": 2552
+    },
+    {
+      "epoch": 0.691110810445136,
+      "grad_norm": 0.2263440489768982,
+      "learning_rate": 9.492182299912857e-05,
+      "loss": 1.245,
+      "step": 2554
+    },
+    {
+      "epoch": 0.6916520092003788,
+      "grad_norm": 0.23101641237735748,
+      "learning_rate": 9.490799022047286e-05,
+      "loss": 1.2253,
+      "step": 2556
+    },
+    {
+      "epoch": 0.6921932079556217,
+      "grad_norm": 0.2258201241493225,
+      "learning_rate": 9.489413963815694e-05,
+      "loss": 1.2477,
+      "step": 2558
+    },
+    {
+      "epoch": 0.6927344067108646,
+      "grad_norm": 0.2227460741996765,
+      "learning_rate": 9.488027125767187e-05,
+      "loss": 1.2215,
+      "step": 2560
+    },
+    {
+      "epoch": 0.6932756054661074,
+      "grad_norm": 0.2213139533996582,
+      "learning_rate": 9.48663850845157e-05,
+      "loss": 1.2308,
+      "step": 2562
+    },
+    {
+      "epoch": 0.6938168042213503,
+      "grad_norm": 0.22192241251468658,
+      "learning_rate": 9.485248112419363e-05,
+      "loss": 1.2487,
+      "step": 2564
+    },
+    {
+      "epoch": 0.6943580029765931,
+      "grad_norm": 0.21532469987869263,
+      "learning_rate": 9.483855938221777e-05,
+      "loss": 1.2498,
+      "step": 2566
+    },
+    {
+      "epoch": 0.6948992017318361,
+      "grad_norm": 0.21143551170825958,
+      "learning_rate": 9.482461986410743e-05,
+      "loss": 1.2453,
+      "step": 2568
+    },
+    {
+      "epoch": 0.6954404004870789,
+      "grad_norm": 0.21282954514026642,
+      "learning_rate": 9.481066257538886e-05,
+      "loss": 1.2499,
+      "step": 2570
+    },
+    {
+      "epoch": 0.6959815992423217,
+      "grad_norm": 0.219988152384758,
+      "learning_rate": 9.47966875215954e-05,
+      "loss": 1.2478,
+      "step": 2572
+    },
+    {
+      "epoch": 0.6965227979975646,
+      "grad_norm": 0.21327020227909088,
+      "learning_rate": 9.478269470826744e-05,
+      "loss": 1.2364,
+      "step": 2574
+    },
+    {
+      "epoch": 0.6970639967528075,
+      "grad_norm": 0.2091750204563141,
+      "learning_rate": 9.476868414095237e-05,
+      "loss": 1.2494,
+      "step": 2576
+    },
+    {
+      "epoch": 0.6976051955080503,
+      "grad_norm": 0.2145649939775467,
+      "learning_rate": 9.475465582520466e-05,
+      "loss": 1.254,
+      "step": 2578
+    },
+    {
+      "epoch": 0.6981463942632932,
+      "grad_norm": 0.21477670967578888,
+      "learning_rate": 9.474060976658578e-05,
+      "loss": 1.2678,
+      "step": 2580
+    },
+    {
+      "epoch": 0.6986875930185361,
+      "grad_norm": 0.21862445771694183,
+      "learning_rate": 9.472654597066431e-05,
+      "loss": 1.2512,
+      "step": 2582
+    },
+    {
+      "epoch": 0.699228791773779,
+      "grad_norm": 0.21111270785331726,
+      "learning_rate": 9.471246444301574e-05,
+      "loss": 1.2587,
+      "step": 2584
+    },
+    {
+      "epoch": 0.6997699905290218,
+      "grad_norm": 0.21332062780857086,
+      "learning_rate": 9.469836518922269e-05,
+      "loss": 1.2569,
+      "step": 2586
+    },
+    {
+      "epoch": 0.7003111892842646,
+      "grad_norm": 0.21386279165744781,
+      "learning_rate": 9.468424821487476e-05,
+      "loss": 1.2308,
+      "step": 2588
+    },
+    {
+      "epoch": 0.7008523880395076,
+      "grad_norm": 0.20638014376163483,
+      "learning_rate": 9.46701135255686e-05,
+      "loss": 1.2453,
+      "step": 2590
+    },
+    {
+      "epoch": 0.7013935867947504,
+      "grad_norm": 0.2437312752008438,
+      "learning_rate": 9.465596112690787e-05,
+      "loss": 1.2523,
+      "step": 2592
+    },
+    {
+      "epoch": 0.7019347855499932,
+      "grad_norm": 0.22395059466362,
+      "learning_rate": 9.464179102450325e-05,
+      "loss": 1.2535,
+      "step": 2594
+    },
+    {
+      "epoch": 0.7024759843052361,
+      "grad_norm": 0.22118812799453735,
+      "learning_rate": 9.462760322397246e-05,
+      "loss": 1.2488,
+      "step": 2596
+    },
+    {
+      "epoch": 0.703017183060479,
+      "grad_norm": 0.22880488634109497,
+      "learning_rate": 9.461339773094021e-05,
+      "loss": 1.2407,
+      "step": 2598
+    },
+    {
+      "epoch": 0.7035583818157218,
+      "grad_norm": 0.21199798583984375,
+      "learning_rate": 9.45991745510382e-05,
+      "loss": 1.2476,
+      "step": 2600
+    },
+    {
+      "epoch": 0.7040995805709647,
+      "grad_norm": 0.20646455883979797,
+      "learning_rate": 9.458493368990519e-05,
+      "loss": 1.2556,
+      "step": 2602
+    },
+    {
+      "epoch": 0.7046407793262075,
+      "grad_norm": 0.2136593908071518,
+      "learning_rate": 9.457067515318698e-05,
+      "loss": 1.2567,
+      "step": 2604
+    },
+    {
+      "epoch": 0.7051819780814504,
+      "grad_norm": 0.214664489030838,
+      "learning_rate": 9.455639894653627e-05,
+      "loss": 1.266,
+      "step": 2606
+    },
+    {
+      "epoch": 0.7057231768366933,
+      "grad_norm": 0.2101629078388214,
+      "learning_rate": 9.454210507561285e-05,
+      "loss": 1.2499,
+      "step": 2608
+    },
+    {
+      "epoch": 0.7062643755919361,
+      "grad_norm": 0.2157791256904602,
+      "learning_rate": 9.452779354608348e-05,
+      "loss": 1.2421,
+      "step": 2610
+    },
+    {
+      "epoch": 0.706805574347179,
+      "grad_norm": 0.20827960968017578,
+      "learning_rate": 9.451346436362196e-05,
+      "loss": 1.2566,
+      "step": 2612
+    },
+    {
+      "epoch": 0.7073467731024219,
+      "grad_norm": 0.21283753216266632,
+      "learning_rate": 9.449911753390901e-05,
+      "loss": 1.2561,
+      "step": 2614
+    },
+    {
+      "epoch": 0.7078879718576647,
+      "grad_norm": 0.22358572483062744,
+      "learning_rate": 9.448475306263245e-05,
+      "loss": 1.2418,
+      "step": 2616
+    },
+    {
+      "epoch": 0.7084291706129076,
+      "grad_norm": 0.21198727190494537,
+      "learning_rate": 9.4470370955487e-05,
+      "loss": 1.2511,
+      "step": 2618
+    },
+    {
+      "epoch": 0.7089703693681505,
+      "grad_norm": 0.21495653688907623,
+      "learning_rate": 9.445597121817442e-05,
+      "loss": 1.2294,
+      "step": 2620
+    },
+    {
+      "epoch": 0.7095115681233933,
+      "grad_norm": 0.21378777921199799,
+      "learning_rate": 9.444155385640345e-05,
+      "loss": 1.2375,
+      "step": 2622
+    },
+    {
+      "epoch": 0.7100527668786362,
+      "grad_norm": 0.21197205781936646,
+      "learning_rate": 9.442711887588981e-05,
+      "loss": 1.251,
+      "step": 2624
+    },
+    {
+      "epoch": 0.710593965633879,
+      "grad_norm": 0.21979504823684692,
+      "learning_rate": 9.441266628235624e-05,
+      "loss": 1.2467,
+      "step": 2626
+    },
+    {
+      "epoch": 0.7111351643891219,
+      "grad_norm": 0.21565599739551544,
+      "learning_rate": 9.43981960815324e-05,
+      "loss": 1.22,
+      "step": 2628
+    },
+    {
+      "epoch": 0.7116763631443648,
+      "grad_norm": 0.19891119003295898,
+      "learning_rate": 9.438370827915499e-05,
+      "loss": 1.215,
+      "step": 2630
+    },
+    {
+      "epoch": 0.7122175618996076,
+      "grad_norm": 0.21079830825328827,
+      "learning_rate": 9.436920288096764e-05,
+      "loss": 1.2407,
+      "step": 2632
+    },
+    {
+      "epoch": 0.7127587606548504,
+      "grad_norm": 0.21531549096107483,
+      "learning_rate": 9.435467989272099e-05,
+      "loss": 1.2348,
+      "step": 2634
+    },
+    {
+      "epoch": 0.7132999594100934,
+      "grad_norm": 0.22583681344985962,
+      "learning_rate": 9.434013932017265e-05,
+      "loss": 1.2567,
+      "step": 2636
+    },
+    {
+      "epoch": 0.7138411581653362,
+      "grad_norm": 0.24707137048244476,
+      "learning_rate": 9.432558116908718e-05,
+      "loss": 1.244,
+      "step": 2638
+    },
+    {
+      "epoch": 0.714382356920579,
+      "grad_norm": 0.23890820145606995,
+      "learning_rate": 9.431100544523614e-05,
+      "loss": 1.2361,
+      "step": 2640
+    },
+    {
+      "epoch": 0.714923555675822,
+      "grad_norm": 0.2275097668170929,
+      "learning_rate": 9.429641215439802e-05,
+      "loss": 1.2337,
+      "step": 2642
+    },
+    {
+      "epoch": 0.7154647544310648,
+      "grad_norm": 0.22068314254283905,
+      "learning_rate": 9.42818013023583e-05,
+      "loss": 1.246,
+      "step": 2644
+    },
+    {
+      "epoch": 0.7160059531863077,
+      "grad_norm": 0.22214053571224213,
+      "learning_rate": 9.426717289490943e-05,
+      "loss": 1.2507,
+      "step": 2646
+    },
+    {
+      "epoch": 0.7165471519415505,
+      "grad_norm": 0.21483547985553741,
+      "learning_rate": 9.425252693785078e-05,
+      "loss": 1.2223,
+      "step": 2648
+    },
+    {
+      "epoch": 0.7170883506967934,
+      "grad_norm": 0.21457841992378235,
+      "learning_rate": 9.423786343698872e-05,
+      "loss": 1.2494,
+      "step": 2650
+    },
+    {
+      "epoch": 0.7176295494520363,
+      "grad_norm": 0.20471327006816864,
+      "learning_rate": 9.422318239813656e-05,
+      "loss": 1.2426,
+      "step": 2652
+    },
+    {
+      "epoch": 0.7181707482072791,
+      "grad_norm": 0.20799721777439117,
+      "learning_rate": 9.420848382711455e-05,
+      "loss": 1.2409,
+      "step": 2654
+    },
+    {
+      "epoch": 0.7187119469625219,
+      "grad_norm": 0.2095753401517868,
+      "learning_rate": 9.41937677297499e-05,
+      "loss": 1.2349,
+      "step": 2656
+    },
+    {
+      "epoch": 0.7192531457177649,
+      "grad_norm": 0.2103864848613739,
+      "learning_rate": 9.417903411187678e-05,
+      "loss": 1.2432,
+      "step": 2658
+    },
+    {
+      "epoch": 0.7197943444730077,
+      "grad_norm": 0.20874999463558197,
+      "learning_rate": 9.416428297933631e-05,
+      "loss": 1.24,
+      "step": 2660
+    },
+    {
+      "epoch": 0.7203355432282506,
+      "grad_norm": 0.21667924523353577,
+      "learning_rate": 9.41495143379765e-05,
+      "loss": 1.254,
+      "step": 2662
+    },
+    {
+      "epoch": 0.7208767419834934,
+      "grad_norm": 0.20849965512752533,
+      "learning_rate": 9.413472819365237e-05,
+      "loss": 1.2494,
+      "step": 2664
+    },
+    {
+      "epoch": 0.7214179407387363,
+      "grad_norm": 0.2131972759962082,
+      "learning_rate": 9.411992455222585e-05,
+      "loss": 1.2233,
+      "step": 2666
+    },
+    {
+      "epoch": 0.7219591394939792,
+      "grad_norm": 0.21590593457221985,
+      "learning_rate": 9.410510341956579e-05,
+      "loss": 1.2428,
+      "step": 2668
+    },
+    {
+      "epoch": 0.722500338249222,
+      "grad_norm": 0.21747298538684845,
+      "learning_rate": 9.409026480154801e-05,
+      "loss": 1.2495,
+      "step": 2670
+    },
+    {
+      "epoch": 0.7230415370044649,
+      "grad_norm": 0.21579551696777344,
+      "learning_rate": 9.407540870405523e-05,
+      "loss": 1.2513,
+      "step": 2672
+    },
+    {
+      "epoch": 0.7235827357597078,
+      "grad_norm": 0.20697540044784546,
+      "learning_rate": 9.40605351329771e-05,
+      "loss": 1.2364,
+      "step": 2674
+    },
+    {
+      "epoch": 0.7241239345149506,
+      "grad_norm": 0.215818852186203,
+      "learning_rate": 9.404564409421024e-05,
+      "loss": 1.2242,
+      "step": 2676
+    },
+    {
+      "epoch": 0.7246651332701934,
+      "grad_norm": 0.21552613377571106,
+      "learning_rate": 9.403073559365816e-05,
+      "loss": 1.2378,
+      "step": 2678
+    },
+    {
+      "epoch": 0.7252063320254364,
+      "grad_norm": 0.20463980734348297,
+      "learning_rate": 9.401580963723127e-05,
+      "loss": 1.2144,
+      "step": 2680
+    },
+    {
+      "epoch": 0.7257475307806792,
+      "grad_norm": 0.20748072862625122,
+      "learning_rate": 9.400086623084696e-05,
+      "loss": 1.2422,
+      "step": 2682
+    },
+    {
+      "epoch": 0.726288729535922,
+      "grad_norm": 0.21622253954410553,
+      "learning_rate": 9.398590538042948e-05,
+      "loss": 1.2466,
+      "step": 2684
+    },
+    {
+      "epoch": 0.7268299282911649,
+      "grad_norm": 0.21229557693004608,
+      "learning_rate": 9.397092709191005e-05,
+      "loss": 1.2533,
+      "step": 2686
+    },
+    {
+      "epoch": 0.7273711270464078,
+      "grad_norm": 0.2206655591726303,
+      "learning_rate": 9.395593137122676e-05,
+      "loss": 1.2368,
+      "step": 2688
+    },
+    {
+      "epoch": 0.7279123258016507,
+      "grad_norm": 0.22106198966503143,
+      "learning_rate": 9.39409182243246e-05,
+      "loss": 1.2523,
+      "step": 2690
+    },
+    {
+      "epoch": 0.7284535245568935,
+      "grad_norm": 0.21155452728271484,
+      "learning_rate": 9.392588765715554e-05,
+      "loss": 1.2558,
+      "step": 2692
+    },
+    {
+      "epoch": 0.7289947233121363,
+      "grad_norm": 0.2205546647310257,
+      "learning_rate": 9.39108396756784e-05,
+      "loss": 1.2409,
+      "step": 2694
+    },
+    {
+      "epoch": 0.7295359220673793,
+      "grad_norm": 0.2159835547208786,
+      "learning_rate": 9.389577428585888e-05,
+      "loss": 1.248,
+      "step": 2696
+    },
+    {
+      "epoch": 0.7300771208226221,
+      "grad_norm": 0.20885945856571198,
+      "learning_rate": 9.388069149366966e-05,
+      "loss": 1.2388,
+      "step": 2698
+    },
+    {
+      "epoch": 0.7306183195778649,
+      "grad_norm": 0.2038174420595169,
+      "learning_rate": 9.386559130509026e-05,
+      "loss": 1.2213,
+      "step": 2700
+    },
+    {
+      "epoch": 0.7311595183331079,
+      "grad_norm": 0.21526674926280975,
+      "learning_rate": 9.385047372610709e-05,
+      "loss": 1.2369,
+      "step": 2702
+    },
+    {
+      "epoch": 0.7317007170883507,
+      "grad_norm": 0.21164047718048096,
+      "learning_rate": 9.383533876271349e-05,
+      "loss": 1.2414,
+      "step": 2704
+    },
+    {
+      "epoch": 0.7322419158435935,
+      "grad_norm": 0.20897522568702698,
+      "learning_rate": 9.38201864209097e-05,
+      "loss": 1.2396,
+      "step": 2706
+    },
+    {
+      "epoch": 0.7327831145988364,
+      "grad_norm": 0.20399637520313263,
+      "learning_rate": 9.38050167067028e-05,
+      "loss": 1.2254,
+      "step": 2708
+    },
+    {
+      "epoch": 0.7333243133540793,
+      "grad_norm": 0.21298326551914215,
+      "learning_rate": 9.37898296261068e-05,
+      "loss": 1.2465,
+      "step": 2710
+    },
+    {
+      "epoch": 0.7338655121093222,
+      "grad_norm": 0.2132457196712494,
+      "learning_rate": 9.377462518514257e-05,
+      "loss": 1.2309,
+      "step": 2712
+    },
+    {
+      "epoch": 0.734406710864565,
+      "grad_norm": 0.22349213063716888,
+      "learning_rate": 9.375940338983789e-05,
+      "loss": 1.2446,
+      "step": 2714
+    },
+    {
+      "epoch": 0.7349479096198078,
+      "grad_norm": 0.2120126485824585,
+      "learning_rate": 9.374416424622738e-05,
+      "loss": 1.2468,
+      "step": 2716
+    },
+    {
+      "epoch": 0.7354891083750508,
+      "grad_norm": 0.21380288898944855,
+      "learning_rate": 9.372890776035259e-05,
+      "loss": 1.2379,
+      "step": 2718
+    },
+    {
+      "epoch": 0.7360303071302936,
+      "grad_norm": 0.20808126032352448,
+      "learning_rate": 9.371363393826187e-05,
+      "loss": 1.2507,
+      "step": 2720
+    },
+    {
+      "epoch": 0.7365715058855364,
+      "grad_norm": 0.21934252977371216,
+      "learning_rate": 9.369834278601052e-05,
+      "loss": 1.2572,
+      "step": 2722
+    },
+    {
+      "epoch": 0.7371127046407794,
+      "grad_norm": 0.20850279927253723,
+      "learning_rate": 9.36830343096607e-05,
+      "loss": 1.2538,
+      "step": 2724
+    },
+    {
+      "epoch": 0.7376539033960222,
+      "grad_norm": 0.216210275888443,
+      "learning_rate": 9.366770851528137e-05,
+      "loss": 1.2307,
+      "step": 2726
+    },
+    {
+      "epoch": 0.738195102151265,
+      "grad_norm": 0.20865590870380402,
+      "learning_rate": 9.365236540894842e-05,
+      "loss": 1.2293,
+      "step": 2728
+    },
+    {
+      "epoch": 0.7387363009065079,
+      "grad_norm": 0.20254139602184296,
+      "learning_rate": 9.363700499674462e-05,
+      "loss": 1.2543,
+      "step": 2730
+    },
+    {
+      "epoch": 0.7392774996617508,
+      "grad_norm": 0.21307919919490814,
+      "learning_rate": 9.36216272847595e-05,
+      "loss": 1.2353,
+      "step": 2732
+    },
+    {
+      "epoch": 0.7398186984169937,
+      "grad_norm": 0.21729370951652527,
+      "learning_rate": 9.360623227908957e-05,
+      "loss": 1.2384,
+      "step": 2734
+    },
+    {
+      "epoch": 0.7403598971722365,
+      "grad_norm": 0.2232731133699417,
+      "learning_rate": 9.359081998583812e-05,
+      "loss": 1.2237,
+      "step": 2736
+    },
+    {
+      "epoch": 0.7409010959274793,
+      "grad_norm": 0.2216210663318634,
+      "learning_rate": 9.357539041111531e-05,
+      "loss": 1.2338,
+      "step": 2738
+    },
+    {
+      "epoch": 0.7414422946827223,
+      "grad_norm": 0.22623343765735626,
+      "learning_rate": 9.355994356103818e-05,
+      "loss": 1.2307,
+      "step": 2740
+    },
+    {
+      "epoch": 0.7419834934379651,
+      "grad_norm": 0.23438121378421783,
+      "learning_rate": 9.354447944173059e-05,
+      "loss": 1.2363,
+      "step": 2742
+    },
+    {
+      "epoch": 0.7425246921932079,
+      "grad_norm": 0.22417870163917542,
+      "learning_rate": 9.352899805932322e-05,
+      "loss": 1.2658,
+      "step": 2744
+    },
+    {
+      "epoch": 0.7430658909484508,
+      "grad_norm": 0.23071123659610748,
+      "learning_rate": 9.351349941995366e-05,
+      "loss": 1.2414,
+      "step": 2746
+    },
+    {
+      "epoch": 0.7436070897036937,
+      "grad_norm": 0.2217751145362854,
+      "learning_rate": 9.349798352976629e-05,
+      "loss": 1.2392,
+      "step": 2748
+    },
+    {
+      "epoch": 0.7441482884589365,
+      "grad_norm": 0.23684772849082947,
+      "learning_rate": 9.348245039491235e-05,
+      "loss": 1.2503,
+      "step": 2750
+    },
+    {
+      "epoch": 0.7446894872141794,
+      "grad_norm": 0.22171282768249512,
+      "learning_rate": 9.34669000215499e-05,
+      "loss": 1.2287,
+      "step": 2752
+    },
+    {
+      "epoch": 0.7452306859694223,
+      "grad_norm": 0.22197501361370087,
+      "learning_rate": 9.345133241584387e-05,
+      "loss": 1.1991,
+      "step": 2754
+    },
+    {
+      "epoch": 0.7457718847246652,
+      "grad_norm": 0.2621997594833374,
+      "learning_rate": 9.343574758396598e-05,
+      "loss": 1.2346,
+      "step": 2756
+    },
+    {
+      "epoch": 0.746313083479908,
+      "grad_norm": 0.21338815987110138,
+      "learning_rate": 9.342014553209482e-05,
+      "loss": 1.2437,
+      "step": 2758
+    },
+    {
+      "epoch": 0.7468542822351508,
+      "grad_norm": 0.21545028686523438,
+      "learning_rate": 9.340452626641574e-05,
+      "loss": 1.2558,
+      "step": 2760
+    },
+    {
+      "epoch": 0.7473954809903938,
+      "grad_norm": 0.20994403958320618,
+      "learning_rate": 9.338888979312101e-05,
+      "loss": 1.2272,
+      "step": 2762
+    },
+    {
+      "epoch": 0.7479366797456366,
+      "grad_norm": 0.21458999812602997,
+      "learning_rate": 9.337323611840964e-05,
+      "loss": 1.2522,
+      "step": 2764
+    },
+    {
+      "epoch": 0.7484778785008794,
+      "grad_norm": 0.21569861471652985,
+      "learning_rate": 9.335756524848751e-05,
+      "loss": 1.2348,
+      "step": 2766
+    },
+    {
+      "epoch": 0.7490190772561223,
+      "grad_norm": 0.21545200049877167,
+      "learning_rate": 9.334187718956727e-05,
+      "loss": 1.2351,
+      "step": 2768
+    },
+    {
+      "epoch": 0.7495602760113652,
+      "grad_norm": 0.20550045371055603,
+      "learning_rate": 9.332617194786844e-05,
+      "loss": 1.2333,
+      "step": 2770
+    },
+    {
+      "epoch": 0.750101474766608,
+      "grad_norm": 0.21121762692928314,
+      "learning_rate": 9.331044952961729e-05,
+      "loss": 1.2347,
+      "step": 2772
+    },
+    {
+      "epoch": 0.7506426735218509,
+      "grad_norm": 0.2130371332168579,
+      "learning_rate": 9.329470994104697e-05,
+      "loss": 1.2384,
+      "step": 2774
+    },
+    {
+      "epoch": 0.7511838722770937,
+      "grad_norm": 0.2100599855184555,
+      "learning_rate": 9.327895318839739e-05,
+      "loss": 1.2572,
+      "step": 2776
+    },
+    {
+      "epoch": 0.7517250710323367,
+      "grad_norm": 0.2141609936952591,
+      "learning_rate": 9.326317927791526e-05,
+      "loss": 1.2493,
+      "step": 2778
+    },
+    {
+      "epoch": 0.7522662697875795,
+      "grad_norm": 0.20215147733688354,
+      "learning_rate": 9.32473882158541e-05,
+      "loss": 1.2328,
+      "step": 2780
+    },
+    {
+      "epoch": 0.7528074685428223,
+      "grad_norm": 0.20483511686325073,
+      "learning_rate": 9.323158000847428e-05,
+      "loss": 1.2467,
+      "step": 2782
+    },
+    {
+      "epoch": 0.7533486672980653,
+      "grad_norm": 0.21057730913162231,
+      "learning_rate": 9.32157546620429e-05,
+      "loss": 1.226,
+      "step": 2784
+    },
+    {
+      "epoch": 0.7538898660533081,
+      "grad_norm": 0.21694819629192352,
+      "learning_rate": 9.319991218283385e-05,
+      "loss": 1.2269,
+      "step": 2786
+    },
+    {
+      "epoch": 0.7544310648085509,
+      "grad_norm": 0.22040502727031708,
+      "learning_rate": 9.318405257712788e-05,
+      "loss": 1.2336,
+      "step": 2788
+    },
+    {
+      "epoch": 0.7549722635637938,
+      "grad_norm": 0.21631376445293427,
+      "learning_rate": 9.31681758512125e-05,
+      "loss": 1.2426,
+      "step": 2790
+    },
+    {
+      "epoch": 0.7555134623190367,
+      "grad_norm": 0.2051629275083542,
+      "learning_rate": 9.315228201138194e-05,
+      "loss": 1.2528,
+      "step": 2792
+    },
+    {
+      "epoch": 0.7560546610742795,
+      "grad_norm": 0.2192334532737732,
+      "learning_rate": 9.313637106393733e-05,
+      "loss": 1.2262,
+      "step": 2794
+    },
+    {
+      "epoch": 0.7565958598295224,
+      "grad_norm": 0.21075467765331268,
+      "learning_rate": 9.31204430151865e-05,
+      "loss": 1.2464,
+      "step": 2796
+    },
+    {
+      "epoch": 0.7571370585847652,
+      "grad_norm": 0.20599377155303955,
+      "learning_rate": 9.31044978714441e-05,
+      "loss": 1.2284,
+      "step": 2798
+    },
+    {
+      "epoch": 0.7576782573400082,
+      "grad_norm": 0.20556782186031342,
+      "learning_rate": 9.308853563903153e-05,
+      "loss": 1.2337,
+      "step": 2800
+    },
+    {
+      "epoch": 0.758219456095251,
+      "grad_norm": 0.2129114270210266,
+      "learning_rate": 9.307255632427698e-05,
+      "loss": 1.2351,
+      "step": 2802
+    },
+    {
+      "epoch": 0.7587606548504938,
+      "grad_norm": 0.22170618176460266,
+      "learning_rate": 9.305655993351539e-05,
+      "loss": 1.2509,
+      "step": 2804
+    },
+    {
+      "epoch": 0.7593018536057367,
+      "grad_norm": 0.2149934470653534,
+      "learning_rate": 9.304054647308853e-05,
+      "loss": 1.2506,
+      "step": 2806
+    },
+    {
+      "epoch": 0.7598430523609796,
+      "grad_norm": 0.21681110560894012,
+      "learning_rate": 9.302451594934488e-05,
+      "loss": 1.2446,
+      "step": 2808
+    },
+    {
+      "epoch": 0.7603842511162224,
+      "grad_norm": 0.2138003557920456,
+      "learning_rate": 9.300846836863966e-05,
+      "loss": 1.2315,
+      "step": 2810
+    },
+    {
+      "epoch": 0.7609254498714653,
+      "grad_norm": 0.21714389324188232,
+      "learning_rate": 9.299240373733495e-05,
+      "loss": 1.2237,
+      "step": 2812
+    },
+    {
+      "epoch": 0.7614666486267082,
+      "grad_norm": 0.20781435072422028,
+      "learning_rate": 9.297632206179951e-05,
+      "loss": 1.235,
+      "step": 2814
+    },
+    {
+      "epoch": 0.762007847381951,
+      "grad_norm": 0.21106529235839844,
+      "learning_rate": 9.296022334840889e-05,
+      "loss": 1.241,
+      "step": 2816
+    },
+    {
+      "epoch": 0.7625490461371939,
+      "grad_norm": 0.20268838107585907,
+      "learning_rate": 9.294410760354537e-05,
+      "loss": 1.2582,
+      "step": 2818
+    },
+    {
+      "epoch": 0.7630902448924367,
+      "grad_norm": 0.20030425488948822,
+      "learning_rate": 9.292797483359801e-05,
+      "loss": 1.2428,
+      "step": 2820
+    },
+    {
+      "epoch": 0.7636314436476797,
+      "grad_norm": 0.2100449800491333,
+      "learning_rate": 9.291182504496258e-05,
+      "loss": 1.2367,
+      "step": 2822
+    },
+    {
+      "epoch": 0.7641726424029225,
+      "grad_norm": 0.21454234421253204,
+      "learning_rate": 9.289565824404165e-05,
+      "loss": 1.2261,
+      "step": 2824
+    },
+    {
+      "epoch": 0.7647138411581653,
+      "grad_norm": 0.21005463600158691,
+      "learning_rate": 9.28794744372445e-05,
+      "loss": 1.237,
+      "step": 2826
+    },
+    {
+      "epoch": 0.7652550399134082,
+      "grad_norm": 0.2123933732509613,
+      "learning_rate": 9.286327363098717e-05,
+      "loss": 1.2115,
+      "step": 2828
+    },
+    {
+      "epoch": 0.7657962386686511,
+      "grad_norm": 0.2171681821346283,
+      "learning_rate": 9.284705583169239e-05,
+      "loss": 1.2415,
+      "step": 2830
+    },
+    {
+      "epoch": 0.7663374374238939,
+      "grad_norm": 0.21714134514331818,
+      "learning_rate": 9.283082104578972e-05,
+      "loss": 1.237,
+      "step": 2832
+    },
+    {
+      "epoch": 0.7668786361791368,
+      "grad_norm": 0.20111635327339172,
+      "learning_rate": 9.281456927971536e-05,
+      "loss": 1.2237,
+      "step": 2834
+    },
+    {
+      "epoch": 0.7674198349343796,
+      "grad_norm": 0.20329216122627258,
+      "learning_rate": 9.279830053991232e-05,
+      "loss": 1.2338,
+      "step": 2836
+    },
+    {
+      "epoch": 0.7679610336896225,
+      "grad_norm": 0.21309934556484222,
+      "learning_rate": 9.278201483283026e-05,
+      "loss": 1.2314,
+      "step": 2838
+    },
+    {
+      "epoch": 0.7685022324448654,
+      "grad_norm": 0.2031441330909729,
+      "learning_rate": 9.276571216492562e-05,
+      "loss": 1.2016,
+      "step": 2840
+    },
+    {
+      "epoch": 0.7690434312001082,
+      "grad_norm": 0.20808055996894836,
+      "learning_rate": 9.274939254266157e-05,
+      "loss": 1.2367,
+      "step": 2842
+    },
+    {
+      "epoch": 0.7695846299553512,
+      "grad_norm": 0.2009919136762619,
+      "learning_rate": 9.273305597250797e-05,
+      "loss": 1.2488,
+      "step": 2844
+    },
+    {
+      "epoch": 0.770125828710594,
+      "grad_norm": 0.2107824683189392,
+      "learning_rate": 9.27167024609414e-05,
+      "loss": 1.2268,
+      "step": 2846
+    },
+    {
+      "epoch": 0.7706670274658368,
+      "grad_norm": 0.21597540378570557,
+      "learning_rate": 9.270033201444517e-05,
+      "loss": 1.251,
+      "step": 2848
+    },
+    {
+      "epoch": 0.7712082262210797,
+      "grad_norm": 0.21570149064064026,
+      "learning_rate": 9.268394463950934e-05,
+      "loss": 1.2291,
+      "step": 2850
+    },
+    {
+      "epoch": 0.7717494249763226,
+      "grad_norm": 0.2073807716369629,
+      "learning_rate": 9.266754034263061e-05,
+      "loss": 1.2229,
+      "step": 2852
+    },
+    {
+      "epoch": 0.7722906237315654,
+      "grad_norm": 0.21396131813526154,
+      "learning_rate": 9.265111913031243e-05,
+      "loss": 1.2354,
+      "step": 2854
+    },
+    {
+      "epoch": 0.7728318224868083,
+      "grad_norm": 0.22582246363162994,
+      "learning_rate": 9.263468100906494e-05,
+      "loss": 1.2304,
+      "step": 2856
+    },
+    {
+      "epoch": 0.7733730212420511,
+      "grad_norm": 0.20621982216835022,
+      "learning_rate": 9.2618225985405e-05,
+      "loss": 1.2178,
+      "step": 2858
+    },
+    {
+      "epoch": 0.773914219997294,
+      "grad_norm": 0.20848476886749268,
+      "learning_rate": 9.260175406585619e-05,
+      "loss": 1.2256,
+      "step": 2860
+    },
+    {
+      "epoch": 0.7744554187525369,
+      "grad_norm": 0.20401284098625183,
+      "learning_rate": 9.258526525694871e-05,
+      "loss": 1.2201,
+      "step": 2862
+    },
+    {
+      "epoch": 0.7749966175077797,
+      "grad_norm": 0.20163412392139435,
+      "learning_rate": 9.256875956521953e-05,
+      "loss": 1.2537,
+      "step": 2864
+    },
+    {
+      "epoch": 0.7755378162630227,
+      "grad_norm": 0.21364827454090118,
+      "learning_rate": 9.255223699721229e-05,
+      "loss": 1.2473,
+      "step": 2866
+    },
+    {
+      "epoch": 0.7760790150182655,
+      "grad_norm": 0.20431366562843323,
+      "learning_rate": 9.253569755947732e-05,
+      "loss": 1.2261,
+      "step": 2868
+    },
+    {
+      "epoch": 0.7766202137735083,
+      "grad_norm": 0.20286224782466888,
+      "learning_rate": 9.251914125857167e-05,
+      "loss": 1.2227,
+      "step": 2870
+    },
+    {
+      "epoch": 0.7771614125287511,
+      "grad_norm": 0.20155613124370575,
+      "learning_rate": 9.2502568101059e-05,
+      "loss": 1.223,
+      "step": 2872
+    },
+    {
+      "epoch": 0.7777026112839941,
+      "grad_norm": 0.20083202421665192,
+      "learning_rate": 9.24859780935097e-05,
+      "loss": 1.2361,
+      "step": 2874
+    },
+    {
+      "epoch": 0.7782438100392369,
+      "grad_norm": 0.20565393567085266,
+      "learning_rate": 9.246937124250086e-05,
+      "loss": 1.2454,
+      "step": 2876
+    },
+    {
+      "epoch": 0.7787850087944798,
+      "grad_norm": 0.20953059196472168,
+      "learning_rate": 9.24527475546162e-05,
+      "loss": 1.2402,
+      "step": 2878
+    },
+    {
+      "epoch": 0.7793262075497226,
+      "grad_norm": 0.20281900465488434,
+      "learning_rate": 9.243610703644616e-05,
+      "loss": 1.243,
+      "step": 2880
+    },
+    {
+      "epoch": 0.7798674063049655,
+      "grad_norm": 0.2107364684343338,
+      "learning_rate": 9.241944969458784e-05,
+      "loss": 1.2531,
+      "step": 2882
+    },
+    {
+      "epoch": 0.7804086050602084,
+      "grad_norm": 0.2223702222108841,
+      "learning_rate": 9.240277553564495e-05,
+      "loss": 1.2358,
+      "step": 2884
+    },
+    {
+      "epoch": 0.7809498038154512,
+      "grad_norm": 0.21632793545722961,
+      "learning_rate": 9.2386084566228e-05,
+      "loss": 1.2228,
+      "step": 2886
+    },
+    {
+      "epoch": 0.781491002570694,
+      "grad_norm": 0.20548765361309052,
+      "learning_rate": 9.2369376792954e-05,
+      "loss": 1.2211,
+      "step": 2888
+    },
+    {
+      "epoch": 0.782032201325937,
+      "grad_norm": 0.2088259905576706,
+      "learning_rate": 9.235265222244676e-05,
+      "loss": 1.2279,
+      "step": 2890
+    },
+    {
+      "epoch": 0.7825734000811798,
+      "grad_norm": 0.20582431554794312,
+      "learning_rate": 9.233591086133666e-05,
+      "loss": 1.2302,
+      "step": 2892
+    },
+    {
+      "epoch": 0.7831145988364226,
+      "grad_norm": 0.20779459178447723,
+      "learning_rate": 9.23191527162608e-05,
+      "loss": 1.2463,
+      "step": 2894
+    },
+    {
+      "epoch": 0.7836557975916656,
+      "grad_norm": 0.20974069833755493,
+      "learning_rate": 9.23023777938629e-05,
+      "loss": 1.2299,
+      "step": 2896
+    },
+    {
+      "epoch": 0.7841969963469084,
+      "grad_norm": 0.21260978281497955,
+      "learning_rate": 9.228558610079331e-05,
+      "loss": 1.2288,
+      "step": 2898
+    },
+    {
+      "epoch": 0.7847381951021513,
+      "grad_norm": 0.21376334130764008,
+      "learning_rate": 9.226877764370908e-05,
+      "loss": 1.2365,
+      "step": 2900
+    },
+    {
+      "epoch": 0.7852793938573941,
+      "grad_norm": 0.2044704556465149,
+      "learning_rate": 9.225195242927387e-05,
+      "loss": 1.2308,
+      "step": 2902
+    },
+    {
+      "epoch": 0.785820592612637,
+      "grad_norm": 0.20474430918693542,
+      "learning_rate": 9.2235110464158e-05,
+      "loss": 1.2247,
+      "step": 2904
+    },
+    {
+      "epoch": 0.7863617913678799,
+      "grad_norm": 0.20992141962051392,
+      "learning_rate": 9.221825175503842e-05,
+      "loss": 1.2319,
+      "step": 2906
+    },
+    {
+      "epoch": 0.7869029901231227,
+      "grad_norm": 0.21773672103881836,
+      "learning_rate": 9.220137630859874e-05,
+      "loss": 1.2248,
+      "step": 2908
+    },
+    {
+      "epoch": 0.7874441888783655,
+      "grad_norm": 0.21675674617290497,
+      "learning_rate": 9.218448413152913e-05,
+      "loss": 1.235,
+      "step": 2910
+    },
+    {
+      "epoch": 0.7879853876336085,
+      "grad_norm": 0.2061057984828949,
+      "learning_rate": 9.216757523052653e-05,
+      "loss": 1.2381,
+      "step": 2912
+    },
+    {
+      "epoch": 0.7885265863888513,
+      "grad_norm": 0.20627647638320923,
+      "learning_rate": 9.215064961229438e-05,
+      "loss": 1.2299,
+      "step": 2914
+    },
+    {
+      "epoch": 0.7890677851440941,
+      "grad_norm": 0.616995096206665,
+      "learning_rate": 9.213370728354283e-05,
+      "loss": 1.2687,
+      "step": 2916
+    },
+    {
+      "epoch": 0.789608983899337,
+      "grad_norm": 0.23795656859874725,
+      "learning_rate": 9.21167482509886e-05,
+      "loss": 1.2376,
+      "step": 2918
+    },
+    {
+      "epoch": 0.7901501826545799,
+      "grad_norm": 0.2422102987766266,
+      "learning_rate": 9.209977252135506e-05,
+      "loss": 1.2429,
+      "step": 2920
+    },
+    {
+      "epoch": 0.7906913814098228,
+      "grad_norm": 0.2670714259147644,
+      "learning_rate": 9.208278010137222e-05,
+      "loss": 1.2101,
+      "step": 2922
+    },
+    {
+      "epoch": 0.7912325801650656,
+      "grad_norm": 0.294065922498703,
+      "learning_rate": 9.206577099777664e-05,
+      "loss": 1.2465,
+      "step": 2924
+    },
+    {
+      "epoch": 0.7917737789203085,
+      "grad_norm": 0.2536512315273285,
+      "learning_rate": 9.204874521731158e-05,
+      "loss": 1.224,
+      "step": 2926
+    },
+    {
+      "epoch": 0.7923149776755514,
+      "grad_norm": 0.23533768951892853,
+      "learning_rate": 9.203170276672681e-05,
+      "loss": 1.2375,
+      "step": 2928
+    },
+    {
+      "epoch": 0.7928561764307942,
+      "grad_norm": 0.29083576798439026,
+      "learning_rate": 9.201464365277883e-05,
+      "loss": 1.2247,
+      "step": 2930
+    },
+    {
+      "epoch": 0.793397375186037,
+      "grad_norm": 0.278577595949173,
+      "learning_rate": 9.199756788223067e-05,
+      "loss": 1.2459,
+      "step": 2932
+    },
+    {
+      "epoch": 0.79393857394128,
+      "grad_norm": 0.28439784049987793,
+      "learning_rate": 9.198047546185193e-05,
+      "loss": 1.224,
+      "step": 2934
+    },
+    {
+      "epoch": 0.7944797726965228,
+      "grad_norm": 0.45653703808784485,
+      "learning_rate": 9.196336639841892e-05,
+      "loss": 1.2389,
+      "step": 2936
+    },
+    {
+      "epoch": 0.7950209714517656,
+      "grad_norm": 0.43514567613601685,
+      "learning_rate": 9.194624069871442e-05,
+      "loss": 1.2365,
+      "step": 2938
+    },
+    {
+      "epoch": 0.7955621702070085,
+      "grad_norm": 0.39155837893486023,
+      "learning_rate": 9.192909836952794e-05,
+      "loss": 1.2364,
+      "step": 2940
+    },
+    {
+      "epoch": 0.7961033689622514,
+      "grad_norm": 0.2874111235141754,
+      "learning_rate": 9.191193941765546e-05,
+      "loss": 1.2255,
+      "step": 2942
+    },
+    {
+      "epoch": 0.7966445677174943,
+      "grad_norm": 0.2747213840484619,
+      "learning_rate": 9.189476384989963e-05,
+      "loss": 1.2283,
+      "step": 2944
+    },
+    {
+      "epoch": 0.7971857664727371,
+      "grad_norm": 0.2161729484796524,
+      "learning_rate": 9.187757167306966e-05,
+      "loss": 1.2346,
+      "step": 2946
+    },
+    {
+      "epoch": 0.7977269652279799,
+      "grad_norm": 0.22049780189990997,
+      "learning_rate": 9.186036289398134e-05,
+      "loss": 1.2422,
+      "step": 2948
+    },
+    {
+      "epoch": 0.7982681639832229,
+      "grad_norm": 0.22543483972549438,
+      "learning_rate": 9.184313751945704e-05,
+      "loss": 1.2366,
+      "step": 2950
+    },
+    {
+      "epoch": 0.7988093627384657,
+      "grad_norm": 0.227324977517128,
+      "learning_rate": 9.182589555632572e-05,
+      "loss": 1.2251,
+      "step": 2952
+    },
+    {
+      "epoch": 0.7993505614937085,
+      "grad_norm": 0.21022869646549225,
+      "learning_rate": 9.180863701142293e-05,
+      "loss": 1.2337,
+      "step": 2954
+    },
+    {
+      "epoch": 0.7998917602489515,
+      "grad_norm": 0.21206127107143402,
+      "learning_rate": 9.179136189159074e-05,
+      "loss": 1.2277,
+      "step": 2956
+    },
+    {
+      "epoch": 0.8004329590041943,
+      "grad_norm": 0.22186043858528137,
+      "learning_rate": 9.177407020367788e-05,
+      "loss": 1.2471,
+      "step": 2958
+    },
+    {
+      "epoch": 0.8009741577594371,
+      "grad_norm": 0.24853624403476715,
+      "learning_rate": 9.175676195453955e-05,
+      "loss": 1.245,
+      "step": 2960
+    },
+    {
+      "epoch": 0.80151535651468,
+      "grad_norm": 0.24665629863739014,
+      "learning_rate": 9.173943715103757e-05,
+      "loss": 1.2357,
+      "step": 2962
+    },
+    {
+      "epoch": 0.8020565552699229,
+      "grad_norm": 0.25784316658973694,
+      "learning_rate": 9.172209580004035e-05,
+      "loss": 1.2382,
+      "step": 2964
+    },
+    {
+      "epoch": 0.8025977540251658,
+      "grad_norm": 0.22399091720581055,
+      "learning_rate": 9.170473790842278e-05,
+      "loss": 1.2208,
+      "step": 2966
+    },
+    {
+      "epoch": 0.8031389527804086,
+      "grad_norm": 0.2071622610092163,
+      "learning_rate": 9.168736348306638e-05,
+      "loss": 1.217,
+      "step": 2968
+    },
+    {
+      "epoch": 0.8036801515356514,
+      "grad_norm": 0.20199859142303467,
+      "learning_rate": 9.166997253085918e-05,
+      "loss": 1.2489,
+      "step": 2970
+    },
+    {
+      "epoch": 0.8042213502908944,
+      "grad_norm": 0.22274377942085266,
+      "learning_rate": 9.165256505869581e-05,
+      "loss": 1.2417,
+      "step": 2972
+    },
+    {
+      "epoch": 0.8047625490461372,
+      "grad_norm": 0.2210993617773056,
+      "learning_rate": 9.163514107347738e-05,
+      "loss": 1.2395,
+      "step": 2974
+    },
+    {
+      "epoch": 0.80530374780138,
+      "grad_norm": 0.22720351815223694,
+      "learning_rate": 9.161770058211161e-05,
+      "loss": 1.2454,
+      "step": 2976
+    },
+    {
+      "epoch": 0.8058449465566229,
+      "grad_norm": 0.22297875583171844,
+      "learning_rate": 9.160024359151274e-05,
+      "loss": 1.2279,
+      "step": 2978
+    },
+    {
+      "epoch": 0.8063861453118658,
+      "grad_norm": 0.2187831699848175,
+      "learning_rate": 9.158277010860153e-05,
+      "loss": 1.2481,
+      "step": 2980
+    },
+    {
+      "epoch": 0.8069273440671086,
+      "grad_norm": 0.2193155139684677,
+      "learning_rate": 9.15652801403053e-05,
+      "loss": 1.2349,
+      "step": 2982
+    },
+    {
+      "epoch": 0.8074685428223515,
+      "grad_norm": 0.21247895061969757,
+      "learning_rate": 9.154777369355793e-05,
+      "loss": 1.2109,
+      "step": 2984
+    },
+    {
+      "epoch": 0.8080097415775944,
+      "grad_norm": 0.2187221348285675,
+      "learning_rate": 9.15302507752998e-05,
+      "loss": 1.2453,
+      "step": 2986
+    },
+    {
+      "epoch": 0.8085509403328373,
+      "grad_norm": 0.2421714961528778,
+      "learning_rate": 9.151271139247782e-05,
+      "loss": 1.2325,
+      "step": 2988
+    },
+    {
+      "epoch": 0.8090921390880801,
+      "grad_norm": 0.25165337324142456,
+      "learning_rate": 9.149515555204542e-05,
+      "loss": 1.2345,
+      "step": 2990
+    },
+    {
+      "epoch": 0.8096333378433229,
+      "grad_norm": 0.6235466003417969,
+      "learning_rate": 9.147758326096259e-05,
+      "loss": 1.2307,
+      "step": 2992
+    },
+    {
+      "epoch": 0.8101745365985659,
+      "grad_norm": 1.8468120098114014,
+      "learning_rate": 9.14599945261958e-05,
+      "loss": 1.2226,
+      "step": 2994
+    },
+    {
+      "epoch": 0.8107157353538087,
+      "grad_norm": 0.6419652104377747,
+      "learning_rate": 9.144238935471809e-05,
+      "loss": 1.237,
+      "step": 2996
+    },
+    {
+      "epoch": 0.8112569341090515,
+      "grad_norm": 9.949187278747559,
+      "learning_rate": 9.142476775350895e-05,
+      "loss": 1.2359,
+      "step": 2998
+    },
+    {
+      "epoch": 0.8117981328642944,
+      "grad_norm": 51.645015716552734,
+      "learning_rate": 9.140712972955445e-05,
+      "loss": 3.9273,
+      "step": 3000
+    },
+    {
+      "epoch": 0.8123393316195373,
+      "grad_norm": 162.77838134765625,
+      "learning_rate": 9.138947528984714e-05,
+      "loss": 7.3207,
+      "step": 3002
+    },
+    {
+      "epoch": 0.8128805303747801,
+      "grad_norm": 143.2057342529297,
+      "learning_rate": 9.137180444138604e-05,
+      "loss": 6.7469,
+      "step": 3004
+    },
+    {
+      "epoch": 0.813421729130023,
+      "grad_norm": 24.596132278442383,
+      "learning_rate": 9.135411719117677e-05,
+      "loss": 6.7275,
+      "step": 3006
+    },
+    {
+      "epoch": 0.8139629278852659,
+      "grad_norm": 113.07401275634766,
+      "learning_rate": 9.133641354623135e-05,
+      "loss": 6.8831,
+      "step": 3008
+    },
+    {
+      "epoch": 0.8145041266405088,
+      "grad_norm": 54.99746322631836,
+      "learning_rate": 9.131869351356836e-05,
+      "loss": 6.957,
+      "step": 3010
+    },
+    {
+      "epoch": 0.8150453253957516,
+      "grad_norm": 41.836971282958984,
+      "learning_rate": 9.130095710021287e-05,
+      "loss": 6.9197,
+      "step": 3012
+    },
+    {
+      "epoch": 0.8155865241509944,
+      "grad_norm": 16.159334182739258,
+      "learning_rate": 9.128320431319643e-05,
+      "loss": 6.6745,
+      "step": 3014
+    },
+    {
+      "epoch": 0.8161277229062374,
+      "grad_norm": 35.35638427734375,
+      "learning_rate": 9.12654351595571e-05,
+      "loss": 6.5526,
+      "step": 3016
+    },
+    {
+      "epoch": 0.8166689216614802,
+      "grad_norm": 13.418272018432617,
+      "learning_rate": 9.124764964633941e-05,
+      "loss": 6.5421,
+      "step": 3018
+    },
+    {
+      "epoch": 0.817210120416723,
+      "grad_norm": 76.91167449951172,
+      "learning_rate": 9.122984778059436e-05,
+      "loss": 6.598,
+      "step": 3020
+    },
+    {
+      "epoch": 0.8177513191719659,
+      "grad_norm": 29.271677017211914,
+      "learning_rate": 9.121202956937949e-05,
+      "loss": 6.746,
+      "step": 3022
+    },
+    {
+      "epoch": 0.8182925179272088,
+      "grad_norm": 6.679831027984619,
+      "learning_rate": 9.119419501975876e-05,
+      "loss": 6.6051,
+      "step": 3024
+    },
+    {
+      "epoch": 0.8188337166824516,
+      "grad_norm": 2.620072364807129,
+      "learning_rate": 9.117634413880264e-05,
+      "loss": 6.4967,
+      "step": 3026
+    },
+    {
+      "epoch": 0.8193749154376945,
+      "grad_norm": 35.06947708129883,
+      "learning_rate": 9.115847693358808e-05,
+      "loss": 6.5821,
+      "step": 3028
+    },
+    {
+      "epoch": 0.8199161141929373,
+      "grad_norm": 7.402040004730225,
+      "learning_rate": 9.114059341119846e-05,
+      "loss": 6.6581,
+      "step": 3030
+    },
+    {
+      "epoch": 0.8204573129481803,
+      "grad_norm": 9.003362655639648,
+      "learning_rate": 9.112269357872367e-05,
+      "loss": 6.5743,
+      "step": 3032
+    },
+    {
+      "epoch": 0.8209985117034231,
+      "grad_norm": 19.55501937866211,
+      "learning_rate": 9.110477744326008e-05,
+      "loss": 6.5286,
+      "step": 3034
+    },
+    {
+      "epoch": 0.8215397104586659,
+      "grad_norm": 16.157516479492188,
+      "learning_rate": 9.108684501191048e-05,
+      "loss": 6.5647,
+      "step": 3036
+    },
+    {
+      "epoch": 0.8220809092139089,
+      "grad_norm": 11.481574058532715,
+      "learning_rate": 9.10688962917841e-05,
+      "loss": 6.5452,
+      "step": 3038
+    },
+    {
+      "epoch": 0.8226221079691517,
+      "grad_norm": 12.504718780517578,
+      "learning_rate": 9.105093128999672e-05,
+      "loss": 6.4514,
+      "step": 3040
+    },
+    {
+      "epoch": 0.8231633067243945,
+      "grad_norm": 12.287551879882812,
+      "learning_rate": 9.103295001367049e-05,
+      "loss": 6.4944,
+      "step": 3042
+    },
+    {
+      "epoch": 0.8237045054796374,
+      "grad_norm": 6.438191890716553,
+      "learning_rate": 9.101495246993405e-05,
+      "loss": 6.5133,
+      "step": 3044
+    },
+    {
+      "epoch": 0.8242457042348803,
+      "grad_norm": 10.490718841552734,
+      "learning_rate": 9.099693866592249e-05,
+      "loss": 6.6264,
+      "step": 3046
+    },
+    {
+      "epoch": 0.8247869029901231,
+      "grad_norm": 3.485543966293335,
+      "learning_rate": 9.097890860877732e-05,
+      "loss": 6.6551,
+      "step": 3048
+    },
+    {
+      "epoch": 0.825328101745366,
+      "grad_norm": 7.037774562835693,
+      "learning_rate": 9.096086230564653e-05,
+      "loss": 6.5162,
+      "step": 3050
+    },
+    {
+      "epoch": 0.8258693005006088,
+      "grad_norm": 7.147155284881592,
+      "learning_rate": 9.094279976368452e-05,
+      "loss": 6.4698,
+      "step": 3052
+    },
+    {
+      "epoch": 0.8264104992558517,
+      "grad_norm": 6.157515048980713,
+      "learning_rate": 9.092472099005212e-05,
+      "loss": 6.452,
+      "step": 3054
+    },
+    {
+      "epoch": 0.8269516980110946,
+      "grad_norm": 24.150165557861328,
+      "learning_rate": 9.090662599191666e-05,
+      "loss": 6.4559,
+      "step": 3056
+    },
+    {
+      "epoch": 0.8274928967663374,
+      "grad_norm": 2.645434617996216,
+      "learning_rate": 9.088851477645181e-05,
+      "loss": 6.4264,
+      "step": 3058
+    },
+    {
+      "epoch": 0.8280340955215802,
+      "grad_norm": 26.285385131835938,
+      "learning_rate": 9.087038735083775e-05,
+      "loss": 6.4145,
+      "step": 3060
+    },
+    {
+      "epoch": 0.8285752942768232,
+      "grad_norm": 37.51917266845703,
+      "learning_rate": 9.085224372226105e-05,
+      "loss": 6.6315,
+      "step": 3062
+    },
+    {
+      "epoch": 0.829116493032066,
+      "grad_norm": 4.452447891235352,
+      "learning_rate": 9.083408389791468e-05,
+      "loss": 6.5029,
+      "step": 3064
+    },
+    {
+      "epoch": 0.8296576917873089,
+      "grad_norm": 2.6523871421813965,
+      "learning_rate": 9.081590788499807e-05,
+      "loss": 6.4376,
+      "step": 3066
+    },
+    {
+      "epoch": 0.8301988905425518,
+      "grad_norm": 8.193984031677246,
+      "learning_rate": 9.079771569071706e-05,
+      "loss": 6.4238,
+      "step": 3068
+    },
+    {
+      "epoch": 0.8307400892977946,
+      "grad_norm": 6.554834365844727,
+      "learning_rate": 9.07795073222839e-05,
+      "loss": 6.4618,
+      "step": 3070
+    },
+    {
+      "epoch": 0.8312812880530375,
+      "grad_norm": 9.348752975463867,
+      "learning_rate": 9.076128278691726e-05,
+      "loss": 6.4341,
+      "step": 3072
+    },
+    {
+      "epoch": 0.8318224868082803,
+      "grad_norm": 5.486053466796875,
+      "learning_rate": 9.07430420918422e-05,
+      "loss": 6.4018,
+      "step": 3074
+    },
+    {
+      "epoch": 0.8323636855635232,
+      "grad_norm": 6.757177352905273,
+      "learning_rate": 9.07247852442902e-05,
+      "loss": 6.4055,
+      "step": 3076
+    },
+    {
+      "epoch": 0.8329048843187661,
+      "grad_norm": 2.9033281803131104,
+      "learning_rate": 9.070651225149913e-05,
+      "loss": 6.3955,
+      "step": 3078
+    },
+    {
+      "epoch": 0.8334460830740089,
+      "grad_norm": 9.029162406921387,
+      "learning_rate": 9.068822312071328e-05,
+      "loss": 6.3819,
+      "step": 3080
+    },
+    {
+      "epoch": 0.8339872818292517,
+      "grad_norm": 1.991795539855957,
+      "learning_rate": 9.066991785918333e-05,
+      "loss": 6.4096,
+      "step": 3082
+    },
+    {
+      "epoch": 0.8345284805844947,
+      "grad_norm": 1.6264301538467407,
+      "learning_rate": 9.065159647416637e-05,
+      "loss": 6.3804,
+      "step": 3084
+    },
+    {
+      "epoch": 0.8350696793397375,
+      "grad_norm": 2.102339506149292,
+      "learning_rate": 9.063325897292587e-05,
+      "loss": 6.3781,
+      "step": 3086
+    },
+    {
+      "epoch": 0.8356108780949804,
+      "grad_norm": 1.4373193979263306,
+      "learning_rate": 9.061490536273164e-05,
+      "loss": 6.3785,
+      "step": 3088
+    },
+    {
+      "epoch": 0.8361520768502232,
+      "grad_norm": 1.3095070123672485,
+      "learning_rate": 9.059653565085997e-05,
+      "loss": 6.3726,
+      "step": 3090
+    },
+    {
+      "epoch": 0.8366932756054661,
+      "grad_norm": 1.7201792001724243,
+      "learning_rate": 9.057814984459347e-05,
+      "loss": 6.3754,
+      "step": 3092
+    },
+    {
+      "epoch": 0.837234474360709,
+      "grad_norm": 2.334815740585327,
+      "learning_rate": 9.055974795122113e-05,
+      "loss": 6.3692,
+      "step": 3094
+    },
+    {
+      "epoch": 0.8377756731159518,
+      "grad_norm": 2.3444583415985107,
+      "learning_rate": 9.054132997803837e-05,
+      "loss": 6.353,
+      "step": 3096
+    },
+    {
+      "epoch": 0.8383168718711947,
+      "grad_norm": 4.413025379180908,
+      "learning_rate": 9.052289593234693e-05,
+      "loss": 6.345,
+      "step": 3098
+    },
+    {
+      "epoch": 0.8388580706264376,
+      "grad_norm": 6.012264251708984,
+      "learning_rate": 9.050444582145495e-05,
+      "loss": 6.3102,
+      "step": 3100
+    },
+    {
+      "epoch": 0.8393992693816804,
+      "grad_norm": 16.267066955566406,
+      "learning_rate": 9.04859796526769e-05,
+      "loss": 6.2907,
+      "step": 3102
+    },
+    {
+      "epoch": 0.8399404681369232,
+      "grad_norm": 22.159154891967773,
+      "learning_rate": 9.046749743333369e-05,
+      "loss": 6.2638,
+      "step": 3104
+    },
+    {
+      "epoch": 0.8404816668921662,
+      "grad_norm": 12.406789779663086,
+      "learning_rate": 9.044899917075251e-05,
+      "loss": 6.2353,
+      "step": 3106
+    },
+    {
+      "epoch": 0.841022865647409,
+      "grad_norm": 13.376330375671387,
+      "learning_rate": 9.043048487226697e-05,
+      "loss": 6.2521,
+      "step": 3108
+    },
+    {
+      "epoch": 0.8415640644026519,
+      "grad_norm": 56.80801773071289,
+      "learning_rate": 9.041195454521702e-05,
+      "loss": 6.525,
+      "step": 3110
+    },
+    {
+      "epoch": 0.8421052631578947,
+      "grad_norm": 25.448945999145508,
+      "learning_rate": 9.039340819694897e-05,
+      "loss": 6.2921,
+      "step": 3112
+    },
+    {
+      "epoch": 0.8426464619131376,
+      "grad_norm": 14.529874801635742,
+      "learning_rate": 9.037484583481544e-05,
+      "loss": 6.3741,
+      "step": 3114
+    },
+    {
+      "epoch": 0.8431876606683805,
+      "grad_norm": 60.734962463378906,
+      "learning_rate": 9.035626746617547e-05,
+      "loss": 6.4179,
+      "step": 3116
+    },
+    {
+      "epoch": 0.8437288594236233,
+      "grad_norm": 65.94615936279297,
+      "learning_rate": 9.033767309839438e-05,
+      "loss": 7.1273,
+      "step": 3118
+    },
+    {
+      "epoch": 0.8442700581788661,
+      "grad_norm": 18.26249122619629,
+      "learning_rate": 9.031906273884388e-05,
+      "loss": 6.5407,
+      "step": 3120
+    },
+    {
+      "epoch": 0.8448112569341091,
+      "grad_norm": 35.836326599121094,
+      "learning_rate": 9.030043639490197e-05,
+      "loss": 6.3452,
+      "step": 3122
+    },
+    {
+      "epoch": 0.8453524556893519,
+      "grad_norm": 31.0974178314209,
+      "learning_rate": 9.028179407395305e-05,
+      "loss": 6.7146,
+      "step": 3124
+    },
+    {
+      "epoch": 0.8458936544445947,
+      "grad_norm": 20.31138801574707,
+      "learning_rate": 9.026313578338782e-05,
+      "loss": 6.3789,
+      "step": 3126
+    },
+    {
+      "epoch": 0.8464348531998377,
+      "grad_norm": 22.042652130126953,
+      "learning_rate": 9.024446153060328e-05,
+      "loss": 6.2103,
+      "step": 3128
+    },
+    {
+      "epoch": 0.8469760519550805,
+      "grad_norm": 24.26772117614746,
+      "learning_rate": 9.022577132300283e-05,
+      "loss": 6.2928,
+      "step": 3130
+    },
+    {
+      "epoch": 0.8475172507103234,
+      "grad_norm": 23.121070861816406,
+      "learning_rate": 9.020706516799615e-05,
+      "loss": 6.3883,
+      "step": 3132
+    },
+    {
+      "epoch": 0.8480584494655662,
+      "grad_norm": 22.5421199798584,
+      "learning_rate": 9.018834307299922e-05,
+      "loss": 6.2152,
+      "step": 3134
+    },
+    {
+      "epoch": 0.8485996482208091,
+      "grad_norm": 15.424832344055176,
+      "learning_rate": 9.016960504543439e-05,
+      "loss": 6.2134,
+      "step": 3136
+    },
+    {
+      "epoch": 0.849140846976052,
+      "grad_norm": 4.891734600067139,
+      "learning_rate": 9.015085109273029e-05,
+      "loss": 6.137,
+      "step": 3138
+    },
+    {
+      "epoch": 0.8496820457312948,
+      "grad_norm": 3.1807522773742676,
+      "learning_rate": 9.01320812223219e-05,
+      "loss": 6.075,
+      "step": 3140
+    },
+    {
+      "epoch": 0.8502232444865376,
+      "grad_norm": 3.0970818996429443,
+      "learning_rate": 9.011329544165047e-05,
+      "loss": 6.0443,
+      "step": 3142
+    },
+    {
+      "epoch": 0.8507644432417806,
+      "grad_norm": 1.8074407577514648,
+      "learning_rate": 9.009449375816358e-05,
+      "loss": 6.0025,
+      "step": 3144
+    },
+    {
+      "epoch": 0.8513056419970234,
+      "grad_norm": 2.570284843444824,
+      "learning_rate": 9.007567617931512e-05,
+      "loss": 5.9863,
+      "step": 3146
+    },
+    {
+      "epoch": 0.8518468407522662,
+      "grad_norm": 4.418685436248779,
+      "learning_rate": 9.005684271256525e-05,
+      "loss": 5.9694,
+      "step": 3148
+    },
+    {
+      "epoch": 0.8523880395075092,
+      "grad_norm": 6.861987113952637,
+      "learning_rate": 9.003799336538046e-05,
+      "loss": 5.9542,
+      "step": 3150
+    },
+    {
+      "epoch": 0.852929238262752,
+      "grad_norm": 4.844227313995361,
+      "learning_rate": 9.001912814523353e-05,
+      "loss": 5.9202,
+      "step": 3152
+    },
+    {
+      "epoch": 0.8534704370179949,
+      "grad_norm": 3.8856005668640137,
+      "learning_rate": 9.000024705960352e-05,
+      "loss": 5.9001,
+      "step": 3154
+    },
+    {
+      "epoch": 0.8540116357732377,
+      "grad_norm": 7.827765464782715,
+      "learning_rate": 8.998135011597583e-05,
+      "loss": 5.9147,
+      "step": 3156
+    },
+    {
+      "epoch": 0.8545528345284806,
+      "grad_norm": 8.146918296813965,
+      "learning_rate": 8.996243732184206e-05,
+      "loss": 5.8791,
+      "step": 3158
+    },
+    {
+      "epoch": 0.8550940332837235,
+      "grad_norm": 2.4434125423431396,
+      "learning_rate": 8.994350868470015e-05,
+      "loss": 5.8594,
+      "step": 3160
+    },
+    {
+      "epoch": 0.8556352320389663,
+      "grad_norm": 13.466486930847168,
+      "learning_rate": 8.99245642120543e-05,
+      "loss": 5.9065,
+      "step": 3162
+    },
+    {
+      "epoch": 0.8561764307942091,
+      "grad_norm": 15.81242561340332,
+      "learning_rate": 8.990560391141503e-05,
+      "loss": 5.8803,
+      "step": 3164
+    },
+    {
+      "epoch": 0.8567176295494521,
+      "grad_norm": 9.728450775146484,
+      "learning_rate": 8.988662779029909e-05,
+      "loss": 5.9393,
+      "step": 3166
+    },
+    {
+      "epoch": 0.8572588283046949,
+      "grad_norm": 15.22099494934082,
+      "learning_rate": 8.98676358562295e-05,
+      "loss": 5.9002,
+      "step": 3168
+    },
+    {
+      "epoch": 0.8578000270599377,
+      "grad_norm": 6.861898422241211,
+      "learning_rate": 8.98486281167356e-05,
+      "loss": 5.8433,
+      "step": 3170
+    },
+    {
+      "epoch": 0.8583412258151806,
+      "grad_norm": 1.44621741771698,
+      "learning_rate": 8.982960457935293e-05,
+      "loss": 5.8085,
+      "step": 3172
+    },
+    {
+      "epoch": 0.8588824245704235,
+      "grad_norm": 6.946808815002441,
+      "learning_rate": 8.981056525162332e-05,
+      "loss": 5.8282,
+      "step": 3174
+    },
+    {
+      "epoch": 0.8594236233256664,
+      "grad_norm": 7.4296555519104,
+      "learning_rate": 8.979151014109488e-05,
+      "loss": 5.8066,
+      "step": 3176
+    },
+    {
+      "epoch": 0.8599648220809092,
+      "grad_norm": 1.931650161743164,
+      "learning_rate": 8.977243925532196e-05,
+      "loss": 5.7569,
+      "step": 3178
+    },
+    {
+      "epoch": 0.8605060208361521,
+      "grad_norm": 1.9481853246688843,
+      "learning_rate": 8.975335260186515e-05,
+      "loss": 5.7733,
+      "step": 3180
+    },
+    {
+      "epoch": 0.861047219591395,
+      "grad_norm": 5.015133380889893,
+      "learning_rate": 8.973425018829134e-05,
+      "loss": 5.7617,
+      "step": 3182
+    },
+    {
+      "epoch": 0.8615884183466378,
+      "grad_norm": 9.222882270812988,
+      "learning_rate": 8.971513202217359e-05,
+      "loss": 5.7758,
+      "step": 3184
+    },
+    {
+      "epoch": 0.8621296171018806,
+      "grad_norm": 9.576149940490723,
+      "learning_rate": 8.969599811109128e-05,
+      "loss": 5.7407,
+      "step": 3186
+    },
+    {
+      "epoch": 0.8626708158571236,
+      "grad_norm": 10.73227310180664,
+      "learning_rate": 8.967684846262997e-05,
+      "loss": 5.746,
+      "step": 3188
+    },
+    {
+      "epoch": 0.8632120146123664,
+      "grad_norm": 3.917820692062378,
+      "learning_rate": 8.965768308438155e-05,
+      "loss": 5.7495,
+      "step": 3190
+    },
+    {
+      "epoch": 0.8637532133676092,
+      "grad_norm": 16.54001808166504,
+      "learning_rate": 8.963850198394402e-05,
+      "loss": 5.8388,
+      "step": 3192
+    },
+    {
+      "epoch": 0.8642944121228521,
+      "grad_norm": 7.986750602722168,
+      "learning_rate": 8.961930516892172e-05,
+      "loss": 5.7186,
+      "step": 3194
+    },
+    {
+      "epoch": 0.864835610878095,
+      "grad_norm": 6.636990547180176,
+      "learning_rate": 8.960009264692518e-05,
+      "loss": 5.7177,
+      "step": 3196
+    },
+    {
+      "epoch": 0.8653768096333379,
+      "grad_norm": 10.408087730407715,
+      "learning_rate": 8.958086442557111e-05,
+      "loss": 5.7057,
+      "step": 3198
+    },
+    {
+      "epoch": 0.8659180083885807,
+      "grad_norm": 7.17811918258667,
+      "learning_rate": 8.956162051248253e-05,
+      "loss": 5.6626,
+      "step": 3200
+    },
+    {
+      "epoch": 0.8664592071438235,
+      "grad_norm": 3.971498489379883,
+      "learning_rate": 8.954236091528865e-05,
+      "loss": 5.6487,
+      "step": 3202
+    },
+    {
+      "epoch": 0.8670004058990665,
+      "grad_norm": 12.803196907043457,
+      "learning_rate": 8.952308564162486e-05,
+      "loss": 5.6778,
+      "step": 3204
+    },
+    {
+      "epoch": 0.8675416046543093,
+      "grad_norm": 6.128145694732666,
+      "learning_rate": 8.950379469913281e-05,
+      "loss": 5.6023,
+      "step": 3206
+    },
+    {
+      "epoch": 0.8680828034095521,
+      "grad_norm": 2.572782278060913,
+      "learning_rate": 8.948448809546033e-05,
+      "loss": 5.5705,
+      "step": 3208
+    },
+    {
+      "epoch": 0.8686240021647951,
+      "grad_norm": 7.732709884643555,
+      "learning_rate": 8.94651658382615e-05,
+      "loss": 5.5648,
+      "step": 3210
+    },
+    {
+      "epoch": 0.8691652009200379,
+      "grad_norm": 5.356006145477295,
+      "learning_rate": 8.944582793519657e-05,
+      "loss": 5.544,
+      "step": 3212
+    },
+    {
+      "epoch": 0.8697063996752807,
+      "grad_norm": 6.853109359741211,
+      "learning_rate": 8.9426474393932e-05,
+      "loss": 5.5661,
+      "step": 3214
+    },
+    {
+      "epoch": 0.8702475984305236,
+      "grad_norm": 6.834400177001953,
+      "learning_rate": 8.940710522214044e-05,
+      "loss": 5.4848,
+      "step": 3216
+    },
+    {
+      "epoch": 0.8707887971857665,
+      "grad_norm": 3.9893531799316406,
+      "learning_rate": 8.938772042750078e-05,
+      "loss": 5.4885,
+      "step": 3218
+    },
+    {
+      "epoch": 0.8713299959410093,
+      "grad_norm": 5.868941307067871,
+      "learning_rate": 8.936832001769805e-05,
+      "loss": 5.4513,
+      "step": 3220
+    },
+    {
+      "epoch": 0.8718711946962522,
+      "grad_norm": 7.860538482666016,
+      "learning_rate": 8.934890400042351e-05,
+      "loss": 5.3947,
+      "step": 3222
+    },
+    {
+      "epoch": 0.872412393451495,
+      "grad_norm": 7.617331027984619,
+      "learning_rate": 8.932947238337456e-05,
+      "loss": 5.3621,
+      "step": 3224
+    },
+    {
+      "epoch": 0.872953592206738,
+      "grad_norm": 13.739005088806152,
+      "learning_rate": 8.931002517425484e-05,
+      "loss": 5.3659,
+      "step": 3226
+    },
+    {
+      "epoch": 0.8734947909619808,
+      "grad_norm": 32.63043212890625,
+      "learning_rate": 8.929056238077416e-05,
+      "loss": 5.4167,
+      "step": 3228
+    },
+    {
+      "epoch": 0.8740359897172236,
+      "grad_norm": 10.652158737182617,
+      "learning_rate": 8.927108401064847e-05,
+      "loss": 5.44,
+      "step": 3230
+    },
+    {
+      "epoch": 0.8745771884724665,
+      "grad_norm": 8.121881484985352,
+      "learning_rate": 8.925159007159994e-05,
+      "loss": 5.321,
+      "step": 3232
+    },
+    {
+      "epoch": 0.8751183872277094,
+      "grad_norm": 14.56945514678955,
+      "learning_rate": 8.923208057135688e-05,
+      "loss": 5.3423,
+      "step": 3234
+    },
+    {
+      "epoch": 0.8756595859829522,
+      "grad_norm": 19.6383113861084,
+      "learning_rate": 8.92125555176538e-05,
+      "loss": 5.3221,
+      "step": 3236
+    },
+    {
+      "epoch": 0.8762007847381951,
+      "grad_norm": 6.179553985595703,
+      "learning_rate": 8.919301491823133e-05,
+      "loss": 5.2839,
+      "step": 3238
+    },
+    {
+      "epoch": 0.876741983493438,
+      "grad_norm": 14.963215827941895,
+      "learning_rate": 8.917345878083631e-05,
+      "loss": 5.2851,
+      "step": 3240
+    },
+    {
+      "epoch": 0.8772831822486808,
+      "grad_norm": 9.411989212036133,
+      "learning_rate": 8.915388711322173e-05,
+      "loss": 5.2005,
+      "step": 3242
+    },
+    {
+      "epoch": 0.8778243810039237,
+      "grad_norm": 4.158975601196289,
+      "learning_rate": 8.91342999231467e-05,
+      "loss": 5.1834,
+      "step": 3244
+    },
+    {
+      "epoch": 0.8783655797591665,
+      "grad_norm": 6.446174621582031,
+      "learning_rate": 8.911469721837655e-05,
+      "loss": 5.1028,
+      "step": 3246
+    },
+    {
+      "epoch": 0.8789067785144095,
+      "grad_norm": 7.448984622955322,
+      "learning_rate": 8.909507900668269e-05,
+      "loss": 5.0743,
+      "step": 3248
+    },
+    {
+      "epoch": 0.8794479772696523,
+      "grad_norm": 13.622466087341309,
+      "learning_rate": 8.907544529584273e-05,
+      "loss": 5.0598,
+      "step": 3250
+    },
+    {
+      "epoch": 0.8799891760248951,
+      "grad_norm": 7.748263835906982,
+      "learning_rate": 8.905579609364041e-05,
+      "loss": 5.009,
+      "step": 3252
+    },
+    {
+      "epoch": 0.880530374780138,
+      "grad_norm": 7.795956611633301,
+      "learning_rate": 8.903613140786558e-05,
+      "loss": 4.9863,
+      "step": 3254
+    },
+    {
+      "epoch": 0.8810715735353809,
+      "grad_norm": 4.537596702575684,
+      "learning_rate": 8.901645124631428e-05,
+      "loss": 4.9221,
+      "step": 3256
+    },
+    {
+      "epoch": 0.8816127722906237,
+      "grad_norm": 9.186697006225586,
+      "learning_rate": 8.899675561678863e-05,
+      "loss": 4.9103,
+      "step": 3258
+    },
+    {
+      "epoch": 0.8821539710458666,
+      "grad_norm": 6.000752925872803,
+      "learning_rate": 8.897704452709697e-05,
+      "loss": 4.8237,
+      "step": 3260
+    },
+    {
+      "epoch": 0.8826951698011094,
+      "grad_norm": 3.594982862472534,
+      "learning_rate": 8.895731798505366e-05,
+      "loss": 4.8564,
+      "step": 3262
+    },
+    {
+      "epoch": 0.8832363685563523,
+      "grad_norm": 6.02200174331665,
+      "learning_rate": 8.893757599847927e-05,
+      "loss": 4.8013,
+      "step": 3264
+    },
+    {
+      "epoch": 0.8837775673115952,
+      "grad_norm": 4.270401477813721,
+      "learning_rate": 8.891781857520044e-05,
+      "loss": 4.7153,
+      "step": 3266
+    },
+    {
+      "epoch": 0.884318766066838,
+      "grad_norm": 3.288161516189575,
+      "learning_rate": 8.889804572304995e-05,
+      "loss": 4.6645,
+      "step": 3268
+    },
+    {
+      "epoch": 0.884859964822081,
+      "grad_norm": 3.725242853164673,
+      "learning_rate": 8.887825744986674e-05,
+      "loss": 4.6046,
+      "step": 3270
+    },
+    {
+      "epoch": 0.8854011635773238,
+      "grad_norm": 5.191972732543945,
+      "learning_rate": 8.885845376349574e-05,
+      "loss": 4.5773,
+      "step": 3272
+    },
+    {
+      "epoch": 0.8859423623325666,
+      "grad_norm": 4.172395706176758,
+      "learning_rate": 8.883863467178814e-05,
+      "loss": 4.5419,
+      "step": 3274
+    },
+    {
+      "epoch": 0.8864835610878095,
+      "grad_norm": 3.393796920776367,
+      "learning_rate": 8.881880018260116e-05,
+      "loss": 4.4957,
+      "step": 3276
+    },
+    {
+      "epoch": 0.8870247598430524,
+      "grad_norm": 4.771553993225098,
+      "learning_rate": 8.87989503037981e-05,
+      "loss": 4.4532,
+      "step": 3278
+    },
+    {
+      "epoch": 0.8875659585982952,
+      "grad_norm": 3.170055389404297,
+      "learning_rate": 8.877908504324843e-05,
+      "loss": 4.3909,
+      "step": 3280
+    },
+    {
+      "epoch": 0.8881071573535381,
+      "grad_norm": 2.902820110321045,
+      "learning_rate": 8.875920440882767e-05,
+      "loss": 4.3511,
+      "step": 3282
+    },
+    {
+      "epoch": 0.8886483561087809,
+      "grad_norm": 3.18314528465271,
+      "learning_rate": 8.873930840841745e-05,
+      "loss": 4.3131,
+      "step": 3284
+    },
+    {
+      "epoch": 0.8891895548640238,
+      "grad_norm": 2.8853447437286377,
+      "learning_rate": 8.871939704990548e-05,
+      "loss": 4.2587,
+      "step": 3286
+    },
+    {
+      "epoch": 0.8897307536192667,
+      "grad_norm": 3.534005880355835,
+      "learning_rate": 8.869947034118557e-05,
+      "loss": 4.2324,
+      "step": 3288
+    },
+    {
+      "epoch": 0.8902719523745095,
+      "grad_norm": 3.527754068374634,
+      "learning_rate": 8.867952829015761e-05,
+      "loss": 4.2085,
+      "step": 3290
+    },
+    {
+      "epoch": 0.8908131511297525,
+      "grad_norm": 3.0101592540740967,
+      "learning_rate": 8.86595709047276e-05,
+      "loss": 4.1537,
+      "step": 3292
+    },
+    {
+      "epoch": 0.8913543498849953,
+      "grad_norm": 2.5968217849731445,
+      "learning_rate": 8.863959819280759e-05,
+      "loss": 4.1401,
+      "step": 3294
+    },
+    {
+      "epoch": 0.8918955486402381,
+      "grad_norm": 2.631981134414673,
+      "learning_rate": 8.861961016231569e-05,
+      "loss": 4.0802,
+      "step": 3296
+    },
+    {
+      "epoch": 0.892436747395481,
+      "grad_norm": 2.6680941581726074,
+      "learning_rate": 8.859960682117612e-05,
+      "loss": 4.085,
+      "step": 3298
+    },
+    {
+      "epoch": 0.8929779461507239,
+      "grad_norm": 3.104212999343872,
+      "learning_rate": 8.857958817731915e-05,
+      "loss": 4.0593,
+      "step": 3300
+    }
+  ],
+  "logging_steps": 2,
+  "max_steps": 11088,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 300,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1.3718037840191816e+20,
+  "train_batch_size": 8,
+  "trial_name": null,
+  "trial_params": null
+}