diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,36429 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.0,
+  "eval_steps": 500,
+  "global_step": 5198,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.00019238168526356292,
+      "grad_norm": 6.356150899044693,
+      "learning_rate": 0.0,
+      "loss": 1.3232,
+      "step": 1
+    },
+    {
+      "epoch": 0.00038476337052712584,
+      "grad_norm": 7.460174434564686,
+      "learning_rate": 1.282051282051282e-07,
+      "loss": 1.4587,
+      "step": 2
+    },
+    {
+      "epoch": 0.0005771450557906887,
+      "grad_norm": 5.751948289316674,
+      "learning_rate": 2.564102564102564e-07,
+      "loss": 1.4043,
+      "step": 3
+    },
+    {
+      "epoch": 0.0007695267410542517,
+      "grad_norm": 5.880572966101242,
+      "learning_rate": 3.846153846153847e-07,
+      "loss": 1.3223,
+      "step": 4
+    },
+    {
+      "epoch": 0.0009619084263178145,
+      "grad_norm": 5.893318033969335,
+      "learning_rate": 5.128205128205128e-07,
+      "loss": 1.3741,
+      "step": 5
+    },
+    {
+      "epoch": 0.0011542901115813775,
+      "grad_norm": 6.123750392519547,
+      "learning_rate": 6.41025641025641e-07,
+      "loss": 1.37,
+      "step": 6
+    },
+    {
+      "epoch": 0.0013466717968449403,
+      "grad_norm": 5.798504344768712,
+      "learning_rate": 7.692307692307694e-07,
+      "loss": 1.3526,
+      "step": 7
+    },
+    {
+      "epoch": 0.0015390534821085034,
+      "grad_norm": 5.206775620398895,
+      "learning_rate": 8.974358974358975e-07,
+      "loss": 1.3396,
+      "step": 8
+    },
+    {
+      "epoch": 0.0017314351673720662,
+      "grad_norm": 5.627439396081044,
+      "learning_rate": 1.0256410256410257e-06,
+      "loss": 1.3741,
+      "step": 9
+    },
+    {
+      "epoch": 0.001923816852635629,
+      "grad_norm": 5.107427550142596,
+      "learning_rate": 1.153846153846154e-06,
+      "loss": 1.274,
+      "step": 10
+    },
+    {
+      "epoch": 0.002116198537899192,
+      "grad_norm": 4.586843192232132,
+      "learning_rate": 1.282051282051282e-06,
+      "loss": 1.2451,
+      "step": 11
+    },
+    {
+      "epoch": 0.002308580223162755,
+      "grad_norm": 4.695760465398717,
+      "learning_rate": 1.4102564102564104e-06,
+      "loss": 1.3451,
+      "step": 12
+    },
+    {
+      "epoch": 0.002500961908426318,
+      "grad_norm": 3.4767287631957107,
+      "learning_rate": 1.5384615384615387e-06,
+      "loss": 1.31,
+      "step": 13
+    },
+    {
+      "epoch": 0.0026933435936898806,
+      "grad_norm": 2.808292618024251,
+      "learning_rate": 1.6666666666666667e-06,
+      "loss": 1.1952,
+      "step": 14
+    },
+    {
+      "epoch": 0.0028857252789534437,
+      "grad_norm": 2.7243052591621755,
+      "learning_rate": 1.794871794871795e-06,
+      "loss": 1.2295,
+      "step": 15
+    },
+    {
+      "epoch": 0.0030781069642170067,
+      "grad_norm": 2.788705028472469,
+      "learning_rate": 1.9230769230769234e-06,
+      "loss": 1.2607,
+      "step": 16
+    },
+    {
+      "epoch": 0.0032704886494805694,
+      "grad_norm": 0.7135107488790472,
+      "learning_rate": 2.0512820512820513e-06,
+      "loss": 0.2811,
+      "step": 17
+    },
+    {
+      "epoch": 0.0034628703347441324,
+      "grad_norm": 1.8554017739113677,
+      "learning_rate": 2.1794871794871797e-06,
+      "loss": 1.1336,
+      "step": 18
+    },
+    {
+      "epoch": 0.0036552520200076955,
+      "grad_norm": 2.01754428707986,
+      "learning_rate": 2.307692307692308e-06,
+      "loss": 1.1417,
+      "step": 19
+    },
+    {
+      "epoch": 0.003847633705271258,
+      "grad_norm": 2.1410760583713895,
+      "learning_rate": 2.435897435897436e-06,
+      "loss": 1.1683,
+      "step": 20
+    },
+    {
+      "epoch": 0.004040015390534821,
+      "grad_norm": 2.08124187006748,
+      "learning_rate": 2.564102564102564e-06,
+      "loss": 1.1301,
+      "step": 21
+    },
+    {
+      "epoch": 0.004232397075798384,
+      "grad_norm": 2.170425969762234,
+      "learning_rate": 2.6923076923076923e-06,
+      "loss": 1.1828,
+      "step": 22
+    },
+    {
+      "epoch": 0.004424778761061947,
+      "grad_norm": 2.1152227521133025,
+      "learning_rate": 2.8205128205128207e-06,
+      "loss": 1.0477,
+      "step": 23
+    },
+    {
+      "epoch": 0.00461716044632551,
+      "grad_norm": 1.9150160641536391,
+      "learning_rate": 2.948717948717949e-06,
+      "loss": 1.1333,
+      "step": 24
+    },
+    {
+      "epoch": 0.004809542131589073,
+      "grad_norm": 1.7599300452514217,
+      "learning_rate": 3.0769230769230774e-06,
+      "loss": 1.1376,
+      "step": 25
+    },
+    {
+      "epoch": 0.005001923816852636,
+      "grad_norm": 1.7036867822707047,
+      "learning_rate": 3.205128205128206e-06,
+      "loss": 1.0875,
+      "step": 26
+    },
+    {
+      "epoch": 0.005194305502116198,
+      "grad_norm": 1.7674238844125092,
+      "learning_rate": 3.3333333333333333e-06,
+      "loss": 1.1211,
+      "step": 27
+    },
+    {
+      "epoch": 0.005386687187379761,
+      "grad_norm": 1.490880249123068,
+      "learning_rate": 3.4615384615384617e-06,
+      "loss": 1.0775,
+      "step": 28
+    },
+    {
+      "epoch": 0.005579068872643324,
+      "grad_norm": 1.4809150817549896,
+      "learning_rate": 3.58974358974359e-06,
+      "loss": 1.0407,
+      "step": 29
+    },
+    {
+      "epoch": 0.005771450557906887,
+      "grad_norm": 1.5836674755850852,
+      "learning_rate": 3.7179487179487184e-06,
+      "loss": 1.1033,
+      "step": 30
+    },
+    {
+      "epoch": 0.00596383224317045,
+      "grad_norm": 1.5471733710504145,
+      "learning_rate": 3.846153846153847e-06,
+      "loss": 1.0238,
+      "step": 31
+    },
+    {
+      "epoch": 0.0061562139284340135,
+      "grad_norm": 1.6804791145772269,
+      "learning_rate": 3.974358974358974e-06,
+      "loss": 1.0615,
+      "step": 32
+    },
+    {
+      "epoch": 0.006348595613697576,
+      "grad_norm": 1.5218001819450373,
+      "learning_rate": 4.102564102564103e-06,
+      "loss": 1.0391,
+      "step": 33
+    },
+    {
+      "epoch": 0.006540977298961139,
+      "grad_norm": 1.4470931719509301,
+      "learning_rate": 4.230769230769231e-06,
+      "loss": 0.9999,
+      "step": 34
+    },
+    {
+      "epoch": 0.006733358984224702,
+      "grad_norm": 1.319351257398814,
+      "learning_rate": 4.358974358974359e-06,
+      "loss": 0.9936,
+      "step": 35
+    },
+    {
+      "epoch": 0.006925740669488265,
+      "grad_norm": 1.3518468857953607,
+      "learning_rate": 4.487179487179488e-06,
+      "loss": 0.9926,
+      "step": 36
+    },
+    {
+      "epoch": 0.007118122354751828,
+      "grad_norm": 1.380684996792722,
+      "learning_rate": 4.615384615384616e-06,
+      "loss": 1.0446,
+      "step": 37
+    },
+    {
+      "epoch": 0.007310504040015391,
+      "grad_norm": 1.33255859884314,
+      "learning_rate": 4.743589743589744e-06,
+      "loss": 1.0171,
+      "step": 38
+    },
+    {
+      "epoch": 0.007502885725278953,
+      "grad_norm": 1.3194771762897197,
+      "learning_rate": 4.871794871794872e-06,
+      "loss": 1.0117,
+      "step": 39
+    },
+    {
+      "epoch": 0.007695267410542516,
+      "grad_norm": 1.342530930092783,
+      "learning_rate": 5e-06,
+      "loss": 0.9644,
+      "step": 40
+    },
+    {
+      "epoch": 0.007887649095806078,
+      "grad_norm": 1.4082099283802387,
+      "learning_rate": 5.128205128205128e-06,
+      "loss": 1.0024,
+      "step": 41
+    },
+    {
+      "epoch": 0.008080030781069641,
+      "grad_norm": 1.2389155217445196,
+      "learning_rate": 5.256410256410257e-06,
+      "loss": 0.9689,
+      "step": 42
+    },
+    {
+      "epoch": 0.008272412466333205,
+      "grad_norm": 1.2231916402260687,
+      "learning_rate": 5.384615384615385e-06,
+      "loss": 1.0082,
+      "step": 43
+    },
+    {
+      "epoch": 0.008464794151596768,
+      "grad_norm": 1.2181994338720543,
+      "learning_rate": 5.512820512820514e-06,
+      "loss": 0.9581,
+      "step": 44
+    },
+    {
+      "epoch": 0.00865717583686033,
+      "grad_norm": 1.236580683876475,
+      "learning_rate": 5.641025641025641e-06,
+      "loss": 0.9677,
+      "step": 45
+    },
+    {
+      "epoch": 0.008849557522123894,
+      "grad_norm": 1.3784828621217025,
+      "learning_rate": 5.769230769230769e-06,
+      "loss": 1.0024,
+      "step": 46
+    },
+    {
+      "epoch": 0.009041939207387457,
+      "grad_norm": 1.2944371997532096,
+      "learning_rate": 5.897435897435898e-06,
+      "loss": 1.0462,
+      "step": 47
+    },
+    {
+      "epoch": 0.00923432089265102,
+      "grad_norm": 1.2956876805053907,
+      "learning_rate": 6.025641025641026e-06,
+      "loss": 0.9548,
+      "step": 48
+    },
+    {
+      "epoch": 0.009426702577914583,
+      "grad_norm": 1.2840909591180043,
+      "learning_rate": 6.153846153846155e-06,
+      "loss": 1.0241,
+      "step": 49
+    },
+    {
+      "epoch": 0.009619084263178146,
+      "grad_norm": 1.232394676135651,
+      "learning_rate": 6.282051282051282e-06,
+      "loss": 0.999,
+      "step": 50
+    },
+    {
+      "epoch": 0.009811465948441709,
+      "grad_norm": 1.27904397787853,
+      "learning_rate": 6.410256410256412e-06,
+      "loss": 0.9959,
+      "step": 51
+    },
+    {
+      "epoch": 0.010003847633705272,
+      "grad_norm": 1.284461897880254,
+      "learning_rate": 6.538461538461539e-06,
+      "loss": 0.9721,
+      "step": 52
+    },
+    {
+      "epoch": 0.010196229318968833,
+      "grad_norm": 1.2413310630650265,
+      "learning_rate": 6.666666666666667e-06,
+      "loss": 0.964,
+      "step": 53
+    },
+    {
+      "epoch": 0.010388611004232396,
+      "grad_norm": 1.222341840707683,
+      "learning_rate": 6.794871794871796e-06,
+      "loss": 0.9757,
+      "step": 54
+    },
+    {
+      "epoch": 0.01058099268949596,
+      "grad_norm": 1.3637021564967613,
+      "learning_rate": 6.923076923076923e-06,
+      "loss": 1.0509,
+      "step": 55
+    },
+    {
+      "epoch": 0.010773374374759523,
+      "grad_norm": 1.1651712251001558,
+      "learning_rate": 7.051282051282053e-06,
+      "loss": 1.0331,
+      "step": 56
+    },
+    {
+      "epoch": 0.010965756060023086,
+      "grad_norm": 1.1651571052461362,
+      "learning_rate": 7.17948717948718e-06,
+      "loss": 0.9838,
+      "step": 57
+    },
+    {
+      "epoch": 0.011158137745286649,
+      "grad_norm": 1.1718196073687672,
+      "learning_rate": 7.307692307692308e-06,
+      "loss": 0.9933,
+      "step": 58
+    },
+    {
+      "epoch": 0.011350519430550212,
+      "grad_norm": 1.2665570803607102,
+      "learning_rate": 7.435897435897437e-06,
+      "loss": 1.0046,
+      "step": 59
+    },
+    {
+      "epoch": 0.011542901115813775,
+      "grad_norm": 1.2410349716039142,
+      "learning_rate": 7.564102564102564e-06,
+      "loss": 0.9631,
+      "step": 60
+    },
+    {
+      "epoch": 0.011735282801077338,
+      "grad_norm": 1.233196208399184,
+      "learning_rate": 7.692307692307694e-06,
+      "loss": 0.9848,
+      "step": 61
+    },
+    {
+      "epoch": 0.0119276644863409,
+      "grad_norm": 1.191222331975563,
+      "learning_rate": 7.820512820512822e-06,
+      "loss": 1.0303,
+      "step": 62
+    },
+    {
+      "epoch": 0.012120046171604464,
+      "grad_norm": 1.305671820228927,
+      "learning_rate": 7.948717948717949e-06,
+      "loss": 1.0172,
+      "step": 63
+    },
+    {
+      "epoch": 0.012312427856868027,
+      "grad_norm": 1.2916179673830719,
+      "learning_rate": 8.076923076923077e-06,
+      "loss": 0.9476,
+      "step": 64
+    },
+    {
+      "epoch": 0.012504809542131588,
+      "grad_norm": 1.1525627595925938,
+      "learning_rate": 8.205128205128205e-06,
+      "loss": 0.9407,
+      "step": 65
+    },
+    {
+      "epoch": 0.012697191227395151,
+      "grad_norm": 1.3398360470429322,
+      "learning_rate": 8.333333333333334e-06,
+      "loss": 0.9667,
+      "step": 66
+    },
+    {
+      "epoch": 0.012889572912658714,
+      "grad_norm": 1.22154659518903,
+      "learning_rate": 8.461538461538462e-06,
+      "loss": 0.9432,
+      "step": 67
+    },
+    {
+      "epoch": 0.013081954597922277,
+      "grad_norm": 1.1932806740173683,
+      "learning_rate": 8.58974358974359e-06,
+      "loss": 0.9747,
+      "step": 68
+    },
+    {
+      "epoch": 0.01327433628318584,
+      "grad_norm": 1.2617097601976612,
+      "learning_rate": 8.717948717948719e-06,
+      "loss": 1.0119,
+      "step": 69
+    },
+    {
+      "epoch": 0.013466717968449404,
+      "grad_norm": 1.1637460385432554,
+      "learning_rate": 8.846153846153847e-06,
+      "loss": 0.9469,
+      "step": 70
+    },
+    {
+      "epoch": 0.013659099653712967,
+      "grad_norm": 1.308301173461084,
+      "learning_rate": 8.974358974358976e-06,
+      "loss": 0.9364,
+      "step": 71
+    },
+    {
+      "epoch": 0.01385148133897653,
+      "grad_norm": 1.2567033383818984,
+      "learning_rate": 9.102564102564104e-06,
+      "loss": 0.988,
+      "step": 72
+    },
+    {
+      "epoch": 0.014043863024240093,
+      "grad_norm": 1.2342289691783859,
+      "learning_rate": 9.230769230769232e-06,
+      "loss": 0.9568,
+      "step": 73
+    },
+    {
+      "epoch": 0.014236244709503656,
+      "grad_norm": 1.2700383428565953,
+      "learning_rate": 9.358974358974359e-06,
+      "loss": 0.9592,
+      "step": 74
+    },
+    {
+      "epoch": 0.014428626394767219,
+      "grad_norm": 1.3060745110203762,
+      "learning_rate": 9.487179487179487e-06,
+      "loss": 0.9309,
+      "step": 75
+    },
+    {
+      "epoch": 0.014621008080030782,
+      "grad_norm": 1.2798478982575148,
+      "learning_rate": 9.615384615384616e-06,
+      "loss": 0.9655,
+      "step": 76
+    },
+    {
+      "epoch": 0.014813389765294343,
+      "grad_norm": 1.201845804418649,
+      "learning_rate": 9.743589743589744e-06,
+      "loss": 0.9368,
+      "step": 77
+    },
+    {
+      "epoch": 0.015005771450557906,
+      "grad_norm": 1.1367386608783867,
+      "learning_rate": 9.871794871794872e-06,
+      "loss": 0.9925,
+      "step": 78
+    },
+    {
+      "epoch": 0.01519815313582147,
+      "grad_norm": 1.2348844516337532,
+      "learning_rate": 1e-05,
+      "loss": 0.9795,
+      "step": 79
+    },
+    {
+      "epoch": 0.015390534821085032,
+      "grad_norm": 1.3145827943462516,
+      "learning_rate": 1.012820512820513e-05,
+      "loss": 0.914,
+      "step": 80
+    },
+    {
+      "epoch": 0.015582916506348595,
+      "grad_norm": 1.2272418496585038,
+      "learning_rate": 1.0256410256410256e-05,
+      "loss": 0.9686,
+      "step": 81
+    },
+    {
+      "epoch": 0.015775298191612157,
+      "grad_norm": 1.210162911225434,
+      "learning_rate": 1.0384615384615386e-05,
+      "loss": 0.9317,
+      "step": 82
+    },
+    {
+      "epoch": 0.01596767987687572,
+      "grad_norm": 1.196104741347351,
+      "learning_rate": 1.0512820512820514e-05,
+      "loss": 0.9175,
+      "step": 83
+    },
+    {
+      "epoch": 0.016160061562139283,
+      "grad_norm": 1.305742266129155,
+      "learning_rate": 1.0641025641025643e-05,
+      "loss": 0.9573,
+      "step": 84
+    },
+    {
+      "epoch": 0.016352443247402846,
+      "grad_norm": 1.4153109516335347,
+      "learning_rate": 1.076923076923077e-05,
+      "loss": 0.2942,
+      "step": 85
+    },
+    {
+      "epoch": 0.01654482493266641,
+      "grad_norm": 1.4138726641884398,
+      "learning_rate": 1.0897435897435898e-05,
+      "loss": 0.9625,
+      "step": 86
+    },
+    {
+      "epoch": 0.016737206617929972,
+      "grad_norm": 1.2159325640610328,
+      "learning_rate": 1.1025641025641028e-05,
+      "loss": 0.9592,
+      "step": 87
+    },
+    {
+      "epoch": 0.016929588303193535,
+      "grad_norm": 1.2046365101697205,
+      "learning_rate": 1.1153846153846154e-05,
+      "loss": 0.9513,
+      "step": 88
+    },
+    {
+      "epoch": 0.017121969988457098,
+      "grad_norm": 1.240010590070536,
+      "learning_rate": 1.1282051282051283e-05,
+      "loss": 0.9762,
+      "step": 89
+    },
+    {
+      "epoch": 0.01731435167372066,
+      "grad_norm": 1.2256085552227414,
+      "learning_rate": 1.1410256410256411e-05,
+      "loss": 0.9691,
+      "step": 90
+    },
+    {
+      "epoch": 0.017506733358984224,
+      "grad_norm": 1.2215383673245581,
+      "learning_rate": 1.1538461538461538e-05,
+      "loss": 0.9684,
+      "step": 91
+    },
+    {
+      "epoch": 0.017699115044247787,
+      "grad_norm": 1.347213456044837,
+      "learning_rate": 1.1666666666666668e-05,
+      "loss": 0.9749,
+      "step": 92
+    },
+    {
+      "epoch": 0.01789149672951135,
+      "grad_norm": 1.3242297245127812,
+      "learning_rate": 1.1794871794871796e-05,
+      "loss": 0.9556,
+      "step": 93
+    },
+    {
+      "epoch": 0.018083878414774913,
+      "grad_norm": 1.3152328209339867,
+      "learning_rate": 1.1923076923076925e-05,
+      "loss": 0.9261,
+      "step": 94
+    },
+    {
+      "epoch": 0.018276260100038477,
+      "grad_norm": 1.2798858069405359,
+      "learning_rate": 1.2051282051282051e-05,
+      "loss": 0.9409,
+      "step": 95
+    },
+    {
+      "epoch": 0.01846864178530204,
+      "grad_norm": 1.2921697687292342,
+      "learning_rate": 1.217948717948718e-05,
+      "loss": 0.945,
+      "step": 96
+    },
+    {
+      "epoch": 0.018661023470565603,
+      "grad_norm": 1.142260647491554,
+      "learning_rate": 1.230769230769231e-05,
+      "loss": 0.9346,
+      "step": 97
+    },
+    {
+      "epoch": 0.018853405155829166,
+      "grad_norm": 1.2402017470416709,
+      "learning_rate": 1.2435897435897436e-05,
+      "loss": 0.9465,
+      "step": 98
+    },
+    {
+      "epoch": 0.01904578684109273,
+      "grad_norm": 1.126127324418927,
+      "learning_rate": 1.2564102564102565e-05,
+      "loss": 0.9298,
+      "step": 99
+    },
+    {
+      "epoch": 0.019238168526356292,
+      "grad_norm": 1.1494298311668414,
+      "learning_rate": 1.2692307692307693e-05,
+      "loss": 0.9662,
+      "step": 100
+    },
+    {
+      "epoch": 0.019430550211619855,
+      "grad_norm": 1.283194507819127,
+      "learning_rate": 1.2820512820512823e-05,
+      "loss": 1.0265,
+      "step": 101
+    },
+    {
+      "epoch": 0.019622931896883418,
+      "grad_norm": 1.2182110537820618,
+      "learning_rate": 1.294871794871795e-05,
+      "loss": 0.9445,
+      "step": 102
+    },
+    {
+      "epoch": 0.01981531358214698,
+      "grad_norm": 1.2291510459604964,
+      "learning_rate": 1.3076923076923078e-05,
+      "loss": 0.9669,
+      "step": 103
+    },
+    {
+      "epoch": 0.020007695267410544,
+      "grad_norm": 1.1116676357960569,
+      "learning_rate": 1.3205128205128207e-05,
+      "loss": 0.8942,
+      "step": 104
+    },
+    {
+      "epoch": 0.020200076952674107,
+      "grad_norm": 1.1994770595943536,
+      "learning_rate": 1.3333333333333333e-05,
+      "loss": 0.9945,
+      "step": 105
+    },
+    {
+      "epoch": 0.020392458637937667,
+      "grad_norm": 1.226278082946228,
+      "learning_rate": 1.3461538461538463e-05,
+      "loss": 0.9493,
+      "step": 106
+    },
+    {
+      "epoch": 0.02058484032320123,
+      "grad_norm": 1.1872566100518187,
+      "learning_rate": 1.3589743589743592e-05,
+      "loss": 0.9861,
+      "step": 107
+    },
+    {
+      "epoch": 0.020777222008464793,
+      "grad_norm": 1.1794245668786023,
+      "learning_rate": 1.3717948717948718e-05,
+      "loss": 0.8841,
+      "step": 108
+    },
+    {
+      "epoch": 0.020969603693728356,
+      "grad_norm": 1.2372725117873258,
+      "learning_rate": 1.3846153846153847e-05,
+      "loss": 0.9381,
+      "step": 109
+    },
+    {
+      "epoch": 0.02116198537899192,
+      "grad_norm": 1.1607126781804995,
+      "learning_rate": 1.3974358974358975e-05,
+      "loss": 0.9306,
+      "step": 110
+    },
+    {
+      "epoch": 0.021354367064255482,
+      "grad_norm": 1.2431418261793006,
+      "learning_rate": 1.4102564102564105e-05,
+      "loss": 0.9628,
+      "step": 111
+    },
+    {
+      "epoch": 0.021546748749519045,
+      "grad_norm": 1.0785634296726767,
+      "learning_rate": 1.4230769230769232e-05,
+      "loss": 0.9552,
+      "step": 112
+    },
+    {
+      "epoch": 0.021739130434782608,
+      "grad_norm": 1.1595187296804514,
+      "learning_rate": 1.435897435897436e-05,
+      "loss": 0.9013,
+      "step": 113
+    },
+    {
+      "epoch": 0.02193151212004617,
+      "grad_norm": 1.242303190333842,
+      "learning_rate": 1.4487179487179489e-05,
+      "loss": 0.9887,
+      "step": 114
+    },
+    {
+      "epoch": 0.022123893805309734,
+      "grad_norm": 1.2327704153053813,
+      "learning_rate": 1.4615384615384615e-05,
+      "loss": 0.9497,
+      "step": 115
+    },
+    {
+      "epoch": 0.022316275490573297,
+      "grad_norm": 1.2069184184832193,
+      "learning_rate": 1.4743589743589745e-05,
+      "loss": 0.9465,
+      "step": 116
+    },
+    {
+      "epoch": 0.02250865717583686,
+      "grad_norm": 1.2591906722304402,
+      "learning_rate": 1.4871794871794874e-05,
+      "loss": 0.9089,
+      "step": 117
+    },
+    {
+      "epoch": 0.022701038861100423,
+      "grad_norm": 1.2097527048683925,
+      "learning_rate": 1.5000000000000002e-05,
+      "loss": 0.9259,
+      "step": 118
+    },
+    {
+      "epoch": 0.022893420546363986,
+      "grad_norm": 1.1893485069710878,
+      "learning_rate": 1.5128205128205129e-05,
+      "loss": 0.9729,
+      "step": 119
+    },
+    {
+      "epoch": 0.02308580223162755,
+      "grad_norm": 1.1950234333658167,
+      "learning_rate": 1.5256410256410257e-05,
+      "loss": 0.9546,
+      "step": 120
+    },
+    {
+      "epoch": 0.023278183916891113,
+      "grad_norm": 1.2606841359776624,
+      "learning_rate": 1.5384615384615387e-05,
+      "loss": 0.9754,
+      "step": 121
+    },
+    {
+      "epoch": 0.023470565602154676,
+      "grad_norm": 1.2183409132803464,
+      "learning_rate": 1.5512820512820516e-05,
+      "loss": 0.9183,
+      "step": 122
+    },
+    {
+      "epoch": 0.02366294728741824,
+      "grad_norm": 1.3622543226462847,
+      "learning_rate": 1.5641025641025644e-05,
+      "loss": 0.9544,
+      "step": 123
+    },
+    {
+      "epoch": 0.0238553289726818,
+      "grad_norm": 1.1949593434909118,
+      "learning_rate": 1.576923076923077e-05,
+      "loss": 0.982,
+      "step": 124
+    },
+    {
+      "epoch": 0.024047710657945365,
+      "grad_norm": 1.2167742454581438,
+      "learning_rate": 1.5897435897435897e-05,
+      "loss": 0.9378,
+      "step": 125
+    },
+    {
+      "epoch": 0.024240092343208928,
+      "grad_norm": 1.1683673000023957,
+      "learning_rate": 1.602564102564103e-05,
+      "loss": 0.9803,
+      "step": 126
+    },
+    {
+      "epoch": 0.02443247402847249,
+      "grad_norm": 1.2084234881370757,
+      "learning_rate": 1.6153846153846154e-05,
+      "loss": 0.9376,
+      "step": 127
+    },
+    {
+      "epoch": 0.024624855713736054,
+      "grad_norm": 1.1274132368518885,
+      "learning_rate": 1.6282051282051282e-05,
+      "loss": 0.9167,
+      "step": 128
+    },
+    {
+      "epoch": 0.024817237398999614,
+      "grad_norm": 1.221944406241826,
+      "learning_rate": 1.641025641025641e-05,
+      "loss": 0.936,
+      "step": 129
+    },
+    {
+      "epoch": 0.025009619084263177,
+      "grad_norm": 1.2294777809769728,
+      "learning_rate": 1.653846153846154e-05,
+      "loss": 0.9771,
+      "step": 130
+    },
+    {
+      "epoch": 0.02520200076952674,
+      "grad_norm": 1.1360711914699209,
+      "learning_rate": 1.6666666666666667e-05,
+      "loss": 0.897,
+      "step": 131
+    },
+    {
+      "epoch": 0.025394382454790303,
+      "grad_norm": 1.2004224007184938,
+      "learning_rate": 1.6794871794871796e-05,
+      "loss": 0.9791,
+      "step": 132
+    },
+    {
+      "epoch": 0.025586764140053866,
+      "grad_norm": 1.2072951133748322,
+      "learning_rate": 1.6923076923076924e-05,
+      "loss": 0.9543,
+      "step": 133
+    },
+    {
+      "epoch": 0.02577914582531743,
+      "grad_norm": 1.1658321925654267,
+      "learning_rate": 1.7051282051282053e-05,
+      "loss": 0.95,
+      "step": 134
+    },
+    {
+      "epoch": 0.025971527510580992,
+      "grad_norm": 1.3200229628137294,
+      "learning_rate": 1.717948717948718e-05,
+      "loss": 0.9988,
+      "step": 135
+    },
+    {
+      "epoch": 0.026163909195844555,
+      "grad_norm": 1.1537200958162248,
+      "learning_rate": 1.730769230769231e-05,
+      "loss": 0.3321,
+      "step": 136
+    },
+    {
+      "epoch": 0.026356290881108118,
+      "grad_norm": 0.9605622736584558,
+      "learning_rate": 1.7435897435897438e-05,
+      "loss": 0.3231,
+      "step": 137
+    },
+    {
+      "epoch": 0.02654867256637168,
+      "grad_norm": 1.0820013189072422,
+      "learning_rate": 1.7564102564102566e-05,
+      "loss": 0.9784,
+      "step": 138
+    },
+    {
+      "epoch": 0.026741054251635244,
+      "grad_norm": 1.2537291262623462,
+      "learning_rate": 1.7692307692307694e-05,
+      "loss": 0.9814,
+      "step": 139
+    },
+    {
+      "epoch": 0.026933435936898807,
+      "grad_norm": 1.1782240191467104,
+      "learning_rate": 1.7820512820512823e-05,
+      "loss": 0.9813,
+      "step": 140
+    },
+    {
+      "epoch": 0.02712581762216237,
+      "grad_norm": 1.1298548466201883,
+      "learning_rate": 1.794871794871795e-05,
+      "loss": 0.9287,
+      "step": 141
+    },
+    {
+      "epoch": 0.027318199307425933,
+      "grad_norm": 1.1830177873938508,
+      "learning_rate": 1.807692307692308e-05,
+      "loss": 0.9543,
+      "step": 142
+    },
+    {
+      "epoch": 0.027510580992689496,
+      "grad_norm": 1.1654766148549691,
+      "learning_rate": 1.8205128205128208e-05,
+      "loss": 0.8853,
+      "step": 143
+    },
+    {
+      "epoch": 0.02770296267795306,
+      "grad_norm": 1.2115108756958421,
+      "learning_rate": 1.8333333333333333e-05,
+      "loss": 0.9124,
+      "step": 144
+    },
+    {
+      "epoch": 0.027895344363216622,
+      "grad_norm": 1.1986608137359922,
+      "learning_rate": 1.8461538461538465e-05,
+      "loss": 0.8795,
+      "step": 145
+    },
+    {
+      "epoch": 0.028087726048480186,
+      "grad_norm": 1.0995018763036237,
+      "learning_rate": 1.8589743589743593e-05,
+      "loss": 0.8948,
+      "step": 146
+    },
+    {
+      "epoch": 0.02828010773374375,
+      "grad_norm": 1.052728989359722,
+      "learning_rate": 1.8717948717948718e-05,
+      "loss": 0.967,
+      "step": 147
+    },
+    {
+      "epoch": 0.02847248941900731,
+      "grad_norm": 1.2030109021112139,
+      "learning_rate": 1.8846153846153846e-05,
+      "loss": 0.8704,
+      "step": 148
+    },
+    {
+      "epoch": 0.028664871104270875,
+      "grad_norm": 1.1707864754164343,
+      "learning_rate": 1.8974358974358975e-05,
+      "loss": 0.9994,
+      "step": 149
+    },
+    {
+      "epoch": 0.028857252789534438,
+      "grad_norm": 1.3160351372403425,
+      "learning_rate": 1.9102564102564106e-05,
+      "loss": 0.9547,
+      "step": 150
+    },
+    {
+      "epoch": 0.029049634474798,
+      "grad_norm": 1.2234098073772088,
+      "learning_rate": 1.923076923076923e-05,
+      "loss": 0.8837,
+      "step": 151
+    },
+    {
+      "epoch": 0.029242016160061564,
+      "grad_norm": 1.1195457107764701,
+      "learning_rate": 1.935897435897436e-05,
+      "loss": 0.9251,
+      "step": 152
+    },
+    {
+      "epoch": 0.029434397845325123,
+      "grad_norm": 1.1378589157580923,
+      "learning_rate": 1.9487179487179488e-05,
+      "loss": 0.9495,
+      "step": 153
+    },
+    {
+      "epoch": 0.029626779530588686,
+      "grad_norm": 1.1495489969124464,
+      "learning_rate": 1.9615384615384617e-05,
+      "loss": 0.9657,
+      "step": 154
+    },
+    {
+      "epoch": 0.02981916121585225,
+      "grad_norm": 1.1170477632692544,
+      "learning_rate": 1.9743589743589745e-05,
+      "loss": 0.9444,
+      "step": 155
+    },
+    {
+      "epoch": 0.030011542901115813,
+      "grad_norm": 1.2180391857158919,
+      "learning_rate": 1.9871794871794873e-05,
+      "loss": 0.9586,
+      "step": 156
+    },
+    {
+      "epoch": 0.030203924586379376,
+      "grad_norm": 1.1177503576990446,
+      "learning_rate": 2e-05,
+      "loss": 0.9529,
+      "step": 157
+    },
+    {
+      "epoch": 0.03039630627164294,
+      "grad_norm": 1.1956859318196553,
+      "learning_rate": 1.9999998058827844e-05,
+      "loss": 0.9166,
+      "step": 158
+    },
+    {
+      "epoch": 0.030588687956906502,
+      "grad_norm": 1.2318226700740729,
+      "learning_rate": 1.9999992235312136e-05,
+      "loss": 0.8966,
+      "step": 159
+    },
+    {
+      "epoch": 0.030781069642170065,
+      "grad_norm": 1.151064994987078,
+      "learning_rate": 1.9999982529455127e-05,
+      "loss": 0.9745,
+      "step": 160
+    },
+    {
+      "epoch": 0.030973451327433628,
+      "grad_norm": 1.0434302735157488,
+      "learning_rate": 1.9999968941260596e-05,
+      "loss": 0.8945,
+      "step": 161
+    },
+    {
+      "epoch": 0.03116583301269719,
+      "grad_norm": 1.102292982411395,
+      "learning_rate": 1.9999951470733808e-05,
+      "loss": 0.8612,
+      "step": 162
+    },
+    {
+      "epoch": 0.031358214697960754,
+      "grad_norm": 1.198136195106293,
+      "learning_rate": 1.9999930117881548e-05,
+      "loss": 0.9361,
+      "step": 163
+    },
+    {
+      "epoch": 0.031550596383224314,
+      "grad_norm": 1.1441603355112178,
+      "learning_rate": 1.9999904882712115e-05,
+      "loss": 0.9354,
+      "step": 164
+    },
+    {
+      "epoch": 0.03174297806848788,
+      "grad_norm": 1.0959846960664537,
+      "learning_rate": 1.99998757652353e-05,
+      "loss": 0.9662,
+      "step": 165
+    },
+    {
+      "epoch": 0.03193535975375144,
+      "grad_norm": 1.0732157831054754,
+      "learning_rate": 1.9999842765462403e-05,
+      "loss": 0.8986,
+      "step": 166
+    },
+    {
+      "epoch": 0.032127741439015006,
+      "grad_norm": 1.1394459051898742,
+      "learning_rate": 1.999980588340624e-05,
+      "loss": 0.9579,
+      "step": 167
+    },
+    {
+      "epoch": 0.032320123124278566,
+      "grad_norm": 1.140860779385276,
+      "learning_rate": 1.9999765119081132e-05,
+      "loss": 0.964,
+      "step": 168
+    },
+    {
+      "epoch": 0.03251250480954213,
+      "grad_norm": 2.229065026827578,
+      "learning_rate": 1.9999720472502902e-05,
+      "loss": 0.3626,
+      "step": 169
+    },
+    {
+      "epoch": 0.03270488649480569,
+      "grad_norm": 1.1697092725166676,
+      "learning_rate": 1.9999671943688885e-05,
+      "loss": 0.9372,
+      "step": 170
+    },
+    {
+      "epoch": 0.03289726818006926,
+      "grad_norm": 1.1771701523110045,
+      "learning_rate": 1.9999619532657915e-05,
+      "loss": 0.9515,
+      "step": 171
+    },
+    {
+      "epoch": 0.03308964986533282,
+      "grad_norm": 1.170473037220081,
+      "learning_rate": 1.9999563239430352e-05,
+      "loss": 0.9421,
+      "step": 172
+    },
+    {
+      "epoch": 0.033282031550596385,
+      "grad_norm": 1.1266637922773408,
+      "learning_rate": 1.9999503064028043e-05,
+      "loss": 0.9074,
+      "step": 173
+    },
+    {
+      "epoch": 0.033474413235859944,
+      "grad_norm": 1.1919628872689279,
+      "learning_rate": 1.999943900647435e-05,
+      "loss": 0.9927,
+      "step": 174
+    },
+    {
+      "epoch": 0.03366679492112351,
+      "grad_norm": 1.184468124339579,
+      "learning_rate": 1.9999371066794146e-05,
+      "loss": 0.9218,
+      "step": 175
+    },
+    {
+      "epoch": 0.03385917660638707,
+      "grad_norm": 1.1014673950980793,
+      "learning_rate": 1.9999299245013805e-05,
+      "loss": 0.9559,
+      "step": 176
+    },
+    {
+      "epoch": 0.03405155829165064,
+      "grad_norm": 1.112837418627034,
+      "learning_rate": 1.999922354116121e-05,
+      "loss": 0.9183,
+      "step": 177
+    },
+    {
+      "epoch": 0.034243939976914196,
+      "grad_norm": 1.1481209917108948,
+      "learning_rate": 1.999914395526575e-05,
+      "loss": 0.9665,
+      "step": 178
+    },
+    {
+      "epoch": 0.03443632166217776,
+      "grad_norm": 1.1282025080286808,
+      "learning_rate": 1.9999060487358333e-05,
+      "loss": 0.9339,
+      "step": 179
+    },
+    {
+      "epoch": 0.03462870334744132,
+      "grad_norm": 1.0860973400784955,
+      "learning_rate": 1.9998973137471352e-05,
+      "loss": 0.9354,
+      "step": 180
+    },
+    {
+      "epoch": 0.03482108503270489,
+      "grad_norm": 1.0973852222803078,
+      "learning_rate": 1.9998881905638727e-05,
+      "loss": 0.9342,
+      "step": 181
+    },
+    {
+      "epoch": 0.03501346671796845,
+      "grad_norm": 1.1533859531699742,
+      "learning_rate": 1.9998786791895874e-05,
+      "loss": 0.9218,
+      "step": 182
+    },
+    {
+      "epoch": 0.035205848403232015,
+      "grad_norm": 1.170236418328504,
+      "learning_rate": 1.999868779627972e-05,
+      "loss": 0.9582,
+      "step": 183
+    },
+    {
+      "epoch": 0.035398230088495575,
+      "grad_norm": 1.158847373232724,
+      "learning_rate": 1.9998584918828695e-05,
+      "loss": 1.0164,
+      "step": 184
+    },
+    {
+      "epoch": 0.03559061177375914,
+      "grad_norm": 1.0701746756351092,
+      "learning_rate": 1.9998478159582747e-05,
+      "loss": 0.9067,
+      "step": 185
+    },
+    {
+      "epoch": 0.0357829934590227,
+      "grad_norm": 1.1615115834904643,
+      "learning_rate": 1.999836751858332e-05,
+      "loss": 0.9212,
+      "step": 186
+    },
+    {
+      "epoch": 0.03597537514428627,
+      "grad_norm": 1.0773933287912325,
+      "learning_rate": 1.9998252995873367e-05,
+      "loss": 0.9625,
+      "step": 187
+    },
+    {
+      "epoch": 0.03616775682954983,
+      "grad_norm": 1.2512950006535282,
+      "learning_rate": 1.999813459149735e-05,
+      "loss": 0.9277,
+      "step": 188
+    },
+    {
+      "epoch": 0.03636013851481339,
+      "grad_norm": 1.211372414755267,
+      "learning_rate": 1.9998012305501243e-05,
+      "loss": 0.9712,
+      "step": 189
+    },
+    {
+      "epoch": 0.03655252020007695,
+      "grad_norm": 1.1320716042143104,
+      "learning_rate": 1.999788613793251e-05,
+      "loss": 0.9492,
+      "step": 190
+    },
+    {
+      "epoch": 0.03674490188534051,
+      "grad_norm": 1.013681940704365,
+      "learning_rate": 1.999775608884015e-05,
+      "loss": 0.9142,
+      "step": 191
+    },
+    {
+      "epoch": 0.03693728357060408,
+      "grad_norm": 1.1063189173902823,
+      "learning_rate": 1.9997622158274635e-05,
+      "loss": 0.9416,
+      "step": 192
+    },
+    {
+      "epoch": 0.03712966525586764,
+      "grad_norm": 1.122344006188876,
+      "learning_rate": 1.9997484346287973e-05,
+      "loss": 0.9228,
+      "step": 193
+    },
+    {
+      "epoch": 0.037322046941131205,
+      "grad_norm": 1.2013950005361622,
+      "learning_rate": 1.9997342652933668e-05,
+      "loss": 0.9098,
+      "step": 194
+    },
+    {
+      "epoch": 0.037514428626394765,
+      "grad_norm": 1.007770464014413,
+      "learning_rate": 1.9997197078266723e-05,
+      "loss": 0.9503,
+      "step": 195
+    },
+    {
+      "epoch": 0.03770681031165833,
+      "grad_norm": 1.1611237701180686,
+      "learning_rate": 1.999704762234366e-05,
+      "loss": 0.9325,
+      "step": 196
+    },
+    {
+      "epoch": 0.03789919199692189,
+      "grad_norm": 1.1224107288362561,
+      "learning_rate": 1.99968942852225e-05,
+      "loss": 0.8676,
+      "step": 197
+    },
+    {
+      "epoch": 0.03809157368218546,
+      "grad_norm": 1.1336997025743,
+      "learning_rate": 1.9996737066962778e-05,
+      "loss": 0.944,
+      "step": 198
+    },
+    {
+      "epoch": 0.03828395536744902,
+      "grad_norm": 1.1236107498937375,
+      "learning_rate": 1.9996575967625525e-05,
+      "loss": 0.9613,
+      "step": 199
+    },
+    {
+      "epoch": 0.038476337052712584,
+      "grad_norm": 1.1938189399074122,
+      "learning_rate": 1.999641098727329e-05,
+      "loss": 0.8843,
+      "step": 200
+    },
+    {
+      "epoch": 0.03866871873797614,
+      "grad_norm": 1.1338687743775748,
+      "learning_rate": 1.999624212597013e-05,
+      "loss": 0.9374,
+      "step": 201
+    },
+    {
+      "epoch": 0.03886110042323971,
+      "grad_norm": 1.1284916009490493,
+      "learning_rate": 1.9996069383781587e-05,
+      "loss": 0.9556,
+      "step": 202
+    },
+    {
+      "epoch": 0.03905348210850327,
+      "grad_norm": 1.184460604375914,
+      "learning_rate": 1.9995892760774738e-05,
+      "loss": 0.9787,
+      "step": 203
+    },
+    {
+      "epoch": 0.039245863793766836,
+      "grad_norm": 1.1563725745960476,
+      "learning_rate": 1.9995712257018153e-05,
+      "loss": 0.9164,
+      "step": 204
+    },
+    {
+      "epoch": 0.039438245479030395,
+      "grad_norm": 1.1503534589109294,
+      "learning_rate": 1.9995527872581903e-05,
+      "loss": 0.9259,
+      "step": 205
+    },
+    {
+      "epoch": 0.03963062716429396,
+      "grad_norm": 1.1152788081692069,
+      "learning_rate": 1.9995339607537578e-05,
+      "loss": 0.9487,
+      "step": 206
+    },
+    {
+      "epoch": 0.03982300884955752,
+      "grad_norm": 1.5553786355069055,
+      "learning_rate": 1.9995147461958267e-05,
+      "loss": 0.3682,
+      "step": 207
+    },
+    {
+      "epoch": 0.04001539053482109,
+      "grad_norm": 1.144673576564406,
+      "learning_rate": 1.999495143591857e-05,
+      "loss": 0.9157,
+      "step": 208
+    },
+    {
+      "epoch": 0.04020777222008465,
+      "grad_norm": 1.2314464669871907,
+      "learning_rate": 1.999475152949459e-05,
+      "loss": 0.9432,
+      "step": 209
+    },
+    {
+      "epoch": 0.040400153905348214,
+      "grad_norm": 1.1444133640294172,
+      "learning_rate": 1.9994547742763935e-05,
+      "loss": 0.9755,
+      "step": 210
+    },
+    {
+      "epoch": 0.040592535590611774,
+      "grad_norm": 1.1696806781256905,
+      "learning_rate": 1.9994340075805724e-05,
+      "loss": 0.9855,
+      "step": 211
+    },
+    {
+      "epoch": 0.04078491727587533,
+      "grad_norm": 1.0900244239009993,
+      "learning_rate": 1.9994128528700583e-05,
+      "loss": 0.914,
+      "step": 212
+    },
+    {
+      "epoch": 0.0409772989611389,
+      "grad_norm": 1.1358083240305048,
+      "learning_rate": 1.9993913101530635e-05,
+      "loss": 0.8701,
+      "step": 213
+    },
+    {
+      "epoch": 0.04116968064640246,
+      "grad_norm": 1.0604161017044944,
+      "learning_rate": 1.9993693794379525e-05,
+      "loss": 0.965,
+      "step": 214
+    },
+    {
+      "epoch": 0.041362062331666026,
+      "grad_norm": 1.1222632781198847,
+      "learning_rate": 1.9993470607332387e-05,
+      "loss": 0.965,
+      "step": 215
+    },
+    {
+      "epoch": 0.041554444016929586,
+      "grad_norm": 1.0404105820910567,
+      "learning_rate": 1.999324354047588e-05,
+      "loss": 0.9098,
+      "step": 216
+    },
+    {
+      "epoch": 0.04174682570219315,
+      "grad_norm": 1.0826057556572763,
+      "learning_rate": 1.9993012593898146e-05,
+      "loss": 0.9042,
+      "step": 217
+    },
+    {
+      "epoch": 0.04193920738745671,
+      "grad_norm": 1.071143000772332,
+      "learning_rate": 1.9992777767688857e-05,
+      "loss": 0.9839,
+      "step": 218
+    },
+    {
+      "epoch": 0.04213158907272028,
+      "grad_norm": 1.2181406651342852,
+      "learning_rate": 1.9992539061939175e-05,
+      "loss": 0.9481,
+      "step": 219
+    },
+    {
+      "epoch": 0.04232397075798384,
+      "grad_norm": 1.1458046019704866,
+      "learning_rate": 1.999229647674178e-05,
+      "loss": 0.915,
+      "step": 220
+    },
+    {
+      "epoch": 0.042516352443247404,
+      "grad_norm": 0.8999922583862122,
+      "learning_rate": 1.9992050012190845e-05,
+      "loss": 0.3395,
+      "step": 221
+    },
+    {
+      "epoch": 0.042708734128510964,
+      "grad_norm": 1.0908639239769866,
+      "learning_rate": 1.9991799668382058e-05,
+      "loss": 0.931,
+      "step": 222
+    },
+    {
+      "epoch": 0.04290111581377453,
+      "grad_norm": 1.1650916313670048,
+      "learning_rate": 1.9991545445412614e-05,
+      "loss": 1.0285,
+      "step": 223
+    },
+    {
+      "epoch": 0.04309349749903809,
+      "grad_norm": 1.204308518962529,
+      "learning_rate": 1.9991287343381208e-05,
+      "loss": 0.9116,
+      "step": 224
+    },
+    {
+      "epoch": 0.04328587918430166,
+      "grad_norm": 1.1073331994919449,
+      "learning_rate": 1.9991025362388044e-05,
+      "loss": 0.9367,
+      "step": 225
+    },
+    {
+      "epoch": 0.043478260869565216,
+      "grad_norm": 1.063943175725814,
+      "learning_rate": 1.9990759502534835e-05,
+      "loss": 0.9177,
+      "step": 226
+    },
+    {
+      "epoch": 0.04367064255482878,
+      "grad_norm": 1.0904553490740045,
+      "learning_rate": 1.9990489763924796e-05,
+      "loss": 0.932,
+      "step": 227
+    },
+    {
+      "epoch": 0.04386302424009234,
+      "grad_norm": 1.0839496535753372,
+      "learning_rate": 1.9990216146662648e-05,
+      "loss": 0.9741,
+      "step": 228
+    },
+    {
+      "epoch": 0.04405540592535591,
+      "grad_norm": 1.05529302047895,
+      "learning_rate": 1.9989938650854618e-05,
+      "loss": 0.9141,
+      "step": 229
+    },
+    {
+      "epoch": 0.04424778761061947,
+      "grad_norm": 1.2113600161365692,
+      "learning_rate": 1.998965727660844e-05,
+      "loss": 0.9518,
+      "step": 230
+    },
+    {
+      "epoch": 0.044440169295883035,
+      "grad_norm": 1.086797148895117,
+      "learning_rate": 1.9989372024033352e-05,
+      "loss": 0.9545,
+      "step": 231
+    },
+    {
+      "epoch": 0.044632550981146595,
+      "grad_norm": 1.093633257231368,
+      "learning_rate": 1.99890828932401e-05,
+      "loss": 0.8648,
+      "step": 232
+    },
+    {
+      "epoch": 0.04482493266641016,
+      "grad_norm": 1.076979238600669,
+      "learning_rate": 1.9988789884340938e-05,
+      "loss": 0.8875,
+      "step": 233
+    },
+    {
+      "epoch": 0.04501731435167372,
+      "grad_norm": 1.1276103283093384,
+      "learning_rate": 1.9988492997449615e-05,
+      "loss": 0.8716,
+      "step": 234
+    },
+    {
+      "epoch": 0.04520969603693728,
+      "grad_norm": 1.0923930454301125,
+      "learning_rate": 1.9988192232681398e-05,
+      "loss": 0.9635,
+      "step": 235
+    },
+    {
+      "epoch": 0.04540207772220085,
+      "grad_norm": 1.1778361099192753,
+      "learning_rate": 1.9987887590153055e-05,
+      "loss": 0.9164,
+      "step": 236
+    },
+    {
+      "epoch": 0.045594459407464406,
+      "grad_norm": 1.0519843627182615,
+      "learning_rate": 1.9987579069982856e-05,
+      "loss": 0.927,
+      "step": 237
+    },
+    {
+      "epoch": 0.04578684109272797,
+      "grad_norm": 1.052718144633144,
+      "learning_rate": 1.9987266672290577e-05,
+      "loss": 0.9423,
+      "step": 238
+    },
+    {
+      "epoch": 0.04597922277799153,
+      "grad_norm": 1.1358672842611686,
+      "learning_rate": 1.9986950397197503e-05,
+      "loss": 0.9157,
+      "step": 239
+    },
+    {
+      "epoch": 0.0461716044632551,
+      "grad_norm": 1.1683796771677735,
+      "learning_rate": 1.9986630244826425e-05,
+      "loss": 0.9193,
+      "step": 240
+    },
+    {
+      "epoch": 0.04636398614851866,
+      "grad_norm": 1.0953359655052513,
+      "learning_rate": 1.998630621530164e-05,
+      "loss": 0.8945,
+      "step": 241
+    },
+    {
+      "epoch": 0.046556367833782225,
+      "grad_norm": 1.0940005744919787,
+      "learning_rate": 1.998597830874894e-05,
+      "loss": 0.9809,
+      "step": 242
+    },
+    {
+      "epoch": 0.046748749519045785,
+      "grad_norm": 1.045366395267042,
+      "learning_rate": 1.9985646525295634e-05,
+      "loss": 0.9183,
+      "step": 243
+    },
+    {
+      "epoch": 0.04694113120430935,
+      "grad_norm": 1.0654315300047619,
+      "learning_rate": 1.998531086507053e-05,
+      "loss": 0.9669,
+      "step": 244
+    },
+    {
+      "epoch": 0.04713351288957291,
+      "grad_norm": 1.0647197976586886,
+      "learning_rate": 1.9984971328203945e-05,
+      "loss": 0.9498,
+      "step": 245
+    },
+    {
+      "epoch": 0.04732589457483648,
+      "grad_norm": 1.1270610849451108,
+      "learning_rate": 1.9984627914827698e-05,
+      "loss": 0.9312,
+      "step": 246
+    },
+    {
+      "epoch": 0.04751827626010004,
+      "grad_norm": 1.0004154908578013,
+      "learning_rate": 1.9984280625075115e-05,
+      "loss": 0.8594,
+      "step": 247
+    },
+    {
+      "epoch": 0.0477106579453636,
+      "grad_norm": 0.7213984471834568,
+      "learning_rate": 1.9983929459081022e-05,
+      "loss": 0.335,
+      "step": 248
+    },
+    {
+      "epoch": 0.04790303963062716,
+      "grad_norm": 1.0314733732250971,
+      "learning_rate": 1.998357441698176e-05,
+      "loss": 0.9385,
+      "step": 249
+    },
+    {
+      "epoch": 0.04809542131589073,
+      "grad_norm": 1.0309882199372646,
+      "learning_rate": 1.998321549891516e-05,
+      "loss": 0.9669,
+      "step": 250
+    },
+    {
+      "epoch": 0.04828780300115429,
+      "grad_norm": 1.0482743442271598,
+      "learning_rate": 1.9982852705020572e-05,
+      "loss": 0.9485,
+      "step": 251
+    },
+    {
+      "epoch": 0.048480184686417856,
+      "grad_norm": 0.9768211398385649,
+      "learning_rate": 1.9982486035438848e-05,
+      "loss": 0.9009,
+      "step": 252
+    },
+    {
+      "epoch": 0.048672566371681415,
+      "grad_norm": 0.7057516366190139,
+      "learning_rate": 1.9982115490312334e-05,
+      "loss": 0.3256,
+      "step": 253
+    },
+    {
+      "epoch": 0.04886494805694498,
+      "grad_norm": 1.0848510121877522,
+      "learning_rate": 1.9981741069784894e-05,
+      "loss": 0.9653,
+      "step": 254
+    },
+    {
+      "epoch": 0.04905732974220854,
+      "grad_norm": 1.0382554162795765,
+      "learning_rate": 1.9981362774001886e-05,
+      "loss": 0.9627,
+      "step": 255
+    },
+    {
+      "epoch": 0.04924971142747211,
+      "grad_norm": 0.6515838820144549,
+      "learning_rate": 1.9980980603110185e-05,
+      "loss": 0.3348,
+      "step": 256
+    },
+    {
+      "epoch": 0.04944209311273567,
+      "grad_norm": 0.7877883442525636,
+      "learning_rate": 1.9980594557258158e-05,
+      "loss": 0.3388,
+      "step": 257
+    },
+    {
+      "epoch": 0.04963447479799923,
+      "grad_norm": 0.5909940164480085,
+      "learning_rate": 1.9980204636595682e-05,
+      "loss": 0.3047,
+      "step": 258
+    },
+    {
+      "epoch": 0.049826856483262794,
+      "grad_norm": 1.1251013256324067,
+      "learning_rate": 1.9979810841274135e-05,
+      "loss": 0.8974,
+      "step": 259
+    },
+    {
+      "epoch": 0.05001923816852635,
+      "grad_norm": 1.0900967149974456,
+      "learning_rate": 1.9979413171446403e-05,
+      "loss": 0.9456,
+      "step": 260
+    },
+    {
+      "epoch": 0.05021161985378992,
+      "grad_norm": 1.060243558415194,
+      "learning_rate": 1.9979011627266884e-05,
+      "loss": 0.9378,
+      "step": 261
+    },
+    {
+      "epoch": 0.05040400153905348,
+      "grad_norm": 1.0970163261798498,
+      "learning_rate": 1.997860620889146e-05,
+      "loss": 0.9108,
+      "step": 262
+    },
+    {
+      "epoch": 0.050596383224317046,
+      "grad_norm": 1.1592924594109983,
+      "learning_rate": 1.997819691647753e-05,
+      "loss": 0.9048,
+      "step": 263
+    },
+    {
+      "epoch": 0.050788764909580605,
+      "grad_norm": 0.9756007386481391,
+      "learning_rate": 1.9977783750184e-05,
+      "loss": 0.8795,
+      "step": 264
+    },
+    {
+      "epoch": 0.05098114659484417,
+      "grad_norm": 1.0409892679474717,
+      "learning_rate": 1.9977366710171274e-05,
+      "loss": 0.9461,
+      "step": 265
+    },
+    {
+      "epoch": 0.05117352828010773,
+      "grad_norm": 1.07702180398637,
+      "learning_rate": 1.9976945796601258e-05,
+      "loss": 0.8704,
+      "step": 266
+    },
+    {
+      "epoch": 0.0513659099653713,
+      "grad_norm": 1.0721944708413274,
+      "learning_rate": 1.9976521009637366e-05,
+      "loss": 0.955,
+      "step": 267
+    },
+    {
+      "epoch": 0.05155829165063486,
+      "grad_norm": 1.0605206622380619,
+      "learning_rate": 1.997609234944452e-05,
+      "loss": 0.8572,
+      "step": 268
+    },
+    {
+      "epoch": 0.051750673335898424,
+      "grad_norm": 1.2269573826146742,
+      "learning_rate": 1.9975659816189137e-05,
+      "loss": 0.9421,
+      "step": 269
+    },
+    {
+      "epoch": 0.051943055021161984,
+      "grad_norm": 1.0403430962766598,
+      "learning_rate": 1.997522341003914e-05,
+      "loss": 0.9385,
+      "step": 270
+    },
+    {
+      "epoch": 0.05213543670642555,
+      "grad_norm": 0.9491544436146973,
+      "learning_rate": 1.9974783131163957e-05,
+      "loss": 0.8673,
+      "step": 271
+    },
+    {
+      "epoch": 0.05232781839168911,
+      "grad_norm": 1.1468381977377486,
+      "learning_rate": 1.9974338979734523e-05,
+      "loss": 0.8989,
+      "step": 272
+    },
+    {
+      "epoch": 0.052520200076952676,
+      "grad_norm": 1.0175910295646602,
+      "learning_rate": 1.997389095592327e-05,
+      "loss": 0.9674,
+      "step": 273
+    },
+    {
+      "epoch": 0.052712581762216236,
+      "grad_norm": 0.9991414063281017,
+      "learning_rate": 1.9973439059904133e-05,
+      "loss": 0.9389,
+      "step": 274
+    },
+    {
+      "epoch": 0.0529049634474798,
+      "grad_norm": 1.0896473653566254,
+      "learning_rate": 1.9972983291852565e-05,
+      "loss": 0.8955,
+      "step": 275
+    },
+    {
+      "epoch": 0.05309734513274336,
+      "grad_norm": 1.0746382790918951,
+      "learning_rate": 1.9972523651945496e-05,
+      "loss": 0.9574,
+      "step": 276
+    },
+    {
+      "epoch": 0.05328972681800693,
+      "grad_norm": 0.9605657653398932,
+      "learning_rate": 1.9972060140361384e-05,
+      "loss": 0.9037,
+      "step": 277
+    },
+    {
+      "epoch": 0.05348210850327049,
+      "grad_norm": 1.0890241105639293,
+      "learning_rate": 1.997159275728018e-05,
+      "loss": 0.8942,
+      "step": 278
+    },
+    {
+      "epoch": 0.053674490188534055,
+      "grad_norm": 1.1061845088946927,
+      "learning_rate": 1.9971121502883332e-05,
+      "loss": 0.8949,
+      "step": 279
+    },
+    {
+      "epoch": 0.053866871873797614,
+      "grad_norm": 1.077042768767683,
+      "learning_rate": 1.9970646377353802e-05,
+      "loss": 0.9495,
+      "step": 280
+    },
+    {
+      "epoch": 0.054059253559061174,
+      "grad_norm": 0.9655527709189689,
+      "learning_rate": 1.997016738087605e-05,
+      "loss": 0.8614,
+      "step": 281
+    },
+    {
+      "epoch": 0.05425163524432474,
+      "grad_norm": 1.1011491786871404,
+      "learning_rate": 1.9969684513636035e-05,
+      "loss": 0.9364,
+      "step": 282
+    },
+    {
+      "epoch": 0.0544440169295883,
+      "grad_norm": 1.1221808426515643,
+      "learning_rate": 1.9969197775821227e-05,
+      "loss": 0.9373,
+      "step": 283
+    },
+    {
+      "epoch": 0.05463639861485187,
+      "grad_norm": 1.050891622584982,
+      "learning_rate": 1.9968707167620593e-05,
+      "loss": 0.8827,
+      "step": 284
+    },
+    {
+      "epoch": 0.054828780300115426,
+      "grad_norm": 1.011780379005001,
+      "learning_rate": 1.9968212689224603e-05,
+      "loss": 0.8827,
+      "step": 285
+    },
+    {
+      "epoch": 0.05502116198537899,
+      "grad_norm": 1.0577525212404446,
+      "learning_rate": 1.996771434082523e-05,
+      "loss": 0.9243,
+      "step": 286
+    },
+    {
+      "epoch": 0.05521354367064255,
+      "grad_norm": 0.8653442399593965,
+      "learning_rate": 1.9967212122615958e-05,
+      "loss": 0.3137,
+      "step": 287
+    },
+    {
+      "epoch": 0.05540592535590612,
+      "grad_norm": 1.0582591092385902,
+      "learning_rate": 1.9966706034791752e-05,
+      "loss": 0.9507,
+      "step": 288
+    },
+    {
+      "epoch": 0.05559830704116968,
+      "grad_norm": 0.6827323300782722,
+      "learning_rate": 1.9966196077549106e-05,
+      "loss": 0.314,
+      "step": 289
+    },
+    {
+      "epoch": 0.055790688726433245,
+      "grad_norm": 1.0381813024562143,
+      "learning_rate": 1.996568225108599e-05,
+      "loss": 0.9136,
+      "step": 290
+    },
+    {
+      "epoch": 0.055983070411696804,
+      "grad_norm": 1.098442498117459,
+      "learning_rate": 1.99651645556019e-05,
+      "loss": 0.9614,
+      "step": 291
+    },
+    {
+      "epoch": 0.05617545209696037,
+      "grad_norm": 1.0945047607477703,
+      "learning_rate": 1.9964642991297817e-05,
+      "loss": 0.9248,
+      "step": 292
+    },
+    {
+      "epoch": 0.05636783378222393,
+      "grad_norm": 1.0033527380629652,
+      "learning_rate": 1.996411755837623e-05,
+      "loss": 0.9826,
+      "step": 293
+    },
+    {
+      "epoch": 0.0565602154674875,
+      "grad_norm": 1.0855930802033027,
+      "learning_rate": 1.9963588257041137e-05,
+      "loss": 0.9753,
+      "step": 294
+    },
+    {
+      "epoch": 0.05675259715275106,
+      "grad_norm": 0.9738360472422385,
+      "learning_rate": 1.996305508749802e-05,
+      "loss": 0.9174,
+      "step": 295
+    },
+    {
+      "epoch": 0.05694497883801462,
+      "grad_norm": 1.0813803884378976,
+      "learning_rate": 1.9962518049953887e-05,
+      "loss": 0.8765,
+      "step": 296
+    },
+    {
+      "epoch": 0.05713736052327818,
+      "grad_norm": 1.0935256856613744,
+      "learning_rate": 1.9961977144617225e-05,
+      "loss": 0.9202,
+      "step": 297
+    },
+    {
+      "epoch": 0.05732974220854175,
+      "grad_norm": 1.067431980700158,
+      "learning_rate": 1.996143237169803e-05,
+      "loss": 0.8965,
+      "step": 298
+    },
+    {
+      "epoch": 0.05752212389380531,
+      "grad_norm": 0.9825430567755834,
+      "learning_rate": 1.996088373140781e-05,
+      "loss": 0.9427,
+      "step": 299
+    },
+    {
+      "epoch": 0.057714505579068875,
+      "grad_norm": 0.9247162352211907,
+      "learning_rate": 1.9960331223959564e-05,
+      "loss": 0.8735,
+      "step": 300
+    },
+    {
+      "epoch": 0.057906887264332435,
+      "grad_norm": 1.0320980675068654,
+      "learning_rate": 1.995977484956779e-05,
+      "loss": 0.8913,
+      "step": 301
+    },
+    {
+      "epoch": 0.058099268949596,
+      "grad_norm": 1.0900088745981003,
+      "learning_rate": 1.9959214608448495e-05,
+      "loss": 0.92,
+      "step": 302
+    },
+    {
+      "epoch": 0.05829165063485956,
+      "grad_norm": 1.182851385315275,
+      "learning_rate": 1.9958650500819183e-05,
+      "loss": 0.9683,
+      "step": 303
+    },
+    {
+      "epoch": 0.05848403232012313,
+      "grad_norm": 1.0410136438111164,
+      "learning_rate": 1.995808252689886e-05,
+      "loss": 0.9023,
+      "step": 304
+    },
+    {
+      "epoch": 0.05867641400538669,
+      "grad_norm": 1.0477000542503334,
+      "learning_rate": 1.9957510686908034e-05,
+      "loss": 0.9343,
+      "step": 305
+    },
+    {
+      "epoch": 0.05886879569065025,
+      "grad_norm": 1.0539808012654917,
+      "learning_rate": 1.9956934981068713e-05,
+      "loss": 0.9305,
+      "step": 306
+    },
+    {
+      "epoch": 0.05906117737591381,
+      "grad_norm": 1.0904714054934443,
+      "learning_rate": 1.9956355409604402e-05,
+      "loss": 0.9272,
+      "step": 307
+    },
+    {
+      "epoch": 0.05925355906117737,
+      "grad_norm": 0.9405186185873616,
+      "learning_rate": 1.9955771972740118e-05,
+      "loss": 0.9169,
+      "step": 308
+    },
+    {
+      "epoch": 0.05944594074644094,
+      "grad_norm": 1.0588428635846612,
+      "learning_rate": 1.9955184670702363e-05,
+      "loss": 0.9314,
+      "step": 309
+    },
+    {
+      "epoch": 0.0596383224317045,
+      "grad_norm": 1.0606873185273586,
+      "learning_rate": 1.995459350371915e-05,
+      "loss": 1.0004,
+      "step": 310
+    },
+    {
+      "epoch": 0.059830704116968066,
+      "grad_norm": 1.107795445387977,
+      "learning_rate": 1.9953998472019996e-05,
+      "loss": 0.9787,
+      "step": 311
+    },
+    {
+      "epoch": 0.060023085802231625,
+      "grad_norm": 1.0796477424914575,
+      "learning_rate": 1.995339957583591e-05,
+      "loss": 0.3415,
+      "step": 312
+    },
+    {
+      "epoch": 0.06021546748749519,
+      "grad_norm": 0.9577397407239883,
+      "learning_rate": 1.9952796815399403e-05,
+      "loss": 0.8539,
+      "step": 313
+    },
+    {
+      "epoch": 0.06040784917275875,
+      "grad_norm": 1.0856069314016585,
+      "learning_rate": 1.9952190190944484e-05,
+      "loss": 0.9146,
+      "step": 314
+    },
+    {
+      "epoch": 0.06060023085802232,
+      "grad_norm": 1.0340332337997624,
+      "learning_rate": 1.9951579702706668e-05,
+      "loss": 0.9194,
+      "step": 315
+    },
+    {
+      "epoch": 0.06079261254328588,
+      "grad_norm": 1.0662445513221364,
+      "learning_rate": 1.9950965350922975e-05,
+      "loss": 0.9117,
+      "step": 316
+    },
+    {
+      "epoch": 0.060984994228549444,
+      "grad_norm": 1.0533235979302413,
+      "learning_rate": 1.9950347135831907e-05,
+      "loss": 0.7994,
+      "step": 317
+    },
+    {
+      "epoch": 0.061177375913813004,
+      "grad_norm": 0.9543454802036748,
+      "learning_rate": 1.994972505767348e-05,
+      "loss": 0.8884,
+      "step": 318
+    },
+    {
+      "epoch": 0.06136975759907657,
+      "grad_norm": 1.06894990445039,
+      "learning_rate": 1.994909911668921e-05,
+      "loss": 0.9728,
+      "step": 319
+    },
+    {
+      "epoch": 0.06156213928434013,
+      "grad_norm": 1.0911617081873755,
+      "learning_rate": 1.99484693131221e-05,
+      "loss": 0.9712,
+      "step": 320
+    },
+    {
+      "epoch": 0.061754520969603696,
+      "grad_norm": 1.0849706048848975,
+      "learning_rate": 1.994783564721667e-05,
+      "loss": 0.9374,
+      "step": 321
+    },
+    {
+      "epoch": 0.061946902654867256,
+      "grad_norm": 1.0516532844262305,
+      "learning_rate": 1.9947198119218924e-05,
+      "loss": 0.8779,
+      "step": 322
+    },
+    {
+      "epoch": 0.06213928434013082,
+      "grad_norm": 0.9997084404823914,
+      "learning_rate": 1.994655672937638e-05,
+      "loss": 0.9328,
+      "step": 323
+    },
+    {
+      "epoch": 0.06233166602539438,
+      "grad_norm": 0.6942075607182795,
+      "learning_rate": 1.9945911477938044e-05,
+      "loss": 0.3341,
+      "step": 324
+    },
+    {
+      "epoch": 0.06252404771065795,
+      "grad_norm": 0.6455724930344451,
+      "learning_rate": 1.994526236515442e-05,
+      "loss": 0.3396,
+      "step": 325
+    },
+    {
+      "epoch": 0.06271642939592151,
+      "grad_norm": 1.080158917959659,
+      "learning_rate": 1.994460939127753e-05,
+      "loss": 0.8815,
+      "step": 326
+    },
+    {
+      "epoch": 0.06290881108118507,
+      "grad_norm": 0.9991218764561717,
+      "learning_rate": 1.9943952556560863e-05,
+      "loss": 0.8235,
+      "step": 327
+    },
+    {
+      "epoch": 0.06310119276644863,
+      "grad_norm": 1.0316554557292463,
+      "learning_rate": 1.9943291861259433e-05,
+      "loss": 0.9601,
+      "step": 328
+    },
+    {
+      "epoch": 0.0632935744517122,
+      "grad_norm": 1.0649919855376584,
+      "learning_rate": 1.9942627305629747e-05,
+      "loss": 0.9337,
+      "step": 329
+    },
+    {
+      "epoch": 0.06348595613697576,
+      "grad_norm": 1.0742727965076273,
+      "learning_rate": 1.9941958889929808e-05,
+      "loss": 0.9346,
+      "step": 330
+    },
+    {
+      "epoch": 0.06367833782223932,
+      "grad_norm": 1.068706995315869,
+      "learning_rate": 1.9941286614419113e-05,
+      "loss": 0.8656,
+      "step": 331
+    },
+    {
+      "epoch": 0.06387071950750288,
+      "grad_norm": 1.1906792376256072,
+      "learning_rate": 1.994061047935867e-05,
+      "loss": 0.9453,
+      "step": 332
+    },
+    {
+      "epoch": 0.06406310119276645,
+      "grad_norm": 1.1312488426977712,
+      "learning_rate": 1.9939930485010968e-05,
+      "loss": 0.9224,
+      "step": 333
+    },
+    {
+      "epoch": 0.06425548287803001,
+      "grad_norm": 1.0287433063431402,
+      "learning_rate": 1.9939246631640014e-05,
+      "loss": 0.9053,
+      "step": 334
+    },
+    {
+      "epoch": 0.06444786456329357,
+      "grad_norm": 0.965830259627178,
+      "learning_rate": 1.99385589195113e-05,
+      "loss": 0.8631,
+      "step": 335
+    },
+    {
+      "epoch": 0.06464024624855713,
+      "grad_norm": 1.0421959565467722,
+      "learning_rate": 1.9937867348891815e-05,
+      "loss": 0.9298,
+      "step": 336
+    },
+    {
+      "epoch": 0.0648326279338207,
+      "grad_norm": 0.7219135376823084,
+      "learning_rate": 1.9937171920050057e-05,
+      "loss": 0.3106,
+      "step": 337
+    },
+    {
+      "epoch": 0.06502500961908426,
+      "grad_norm": 1.1259655675632791,
+      "learning_rate": 1.9936472633256012e-05,
+      "loss": 0.9034,
+      "step": 338
+    },
+    {
+      "epoch": 0.06521739130434782,
+      "grad_norm": 1.0814298554277322,
+      "learning_rate": 1.9935769488781167e-05,
+      "loss": 0.8858,
+      "step": 339
+    },
+    {
+      "epoch": 0.06540977298961138,
+      "grad_norm": 1.063170827474876,
+      "learning_rate": 1.993506248689851e-05,
+      "loss": 0.9807,
+      "step": 340
+    },
+    {
+      "epoch": 0.06560215467487496,
+      "grad_norm": 1.0536468760411664,
+      "learning_rate": 1.993435162788252e-05,
+      "loss": 0.9701,
+      "step": 341
+    },
+    {
+      "epoch": 0.06579453636013852,
+      "grad_norm": 0.9880696437186115,
+      "learning_rate": 1.993363691200918e-05,
+      "loss": 0.9135,
+      "step": 342
+    },
+    {
+      "epoch": 0.06598691804540208,
+      "grad_norm": 1.0490399507169421,
+      "learning_rate": 1.9932918339555965e-05,
+      "loss": 1.0048,
+      "step": 343
+    },
+    {
+      "epoch": 0.06617929973066564,
+      "grad_norm": 1.0288469362622994,
+      "learning_rate": 1.9932195910801848e-05,
+      "loss": 0.9486,
+      "step": 344
+    },
+    {
+      "epoch": 0.06637168141592921,
+      "grad_norm": 1.0029948255839551,
+      "learning_rate": 1.9931469626027305e-05,
+      "loss": 0.9528,
+      "step": 345
+    },
+    {
+      "epoch": 0.06656406310119277,
+      "grad_norm": 1.026167041609604,
+      "learning_rate": 1.9930739485514304e-05,
+      "loss": 0.9237,
+      "step": 346
+    },
+    {
+      "epoch": 0.06675644478645633,
+      "grad_norm": 1.0323131504765657,
+      "learning_rate": 1.9930005489546308e-05,
+      "loss": 0.9005,
+      "step": 347
+    },
+    {
+      "epoch": 0.06694882647171989,
+      "grad_norm": 0.6266707513449151,
+      "learning_rate": 1.9929267638408277e-05,
+      "loss": 0.3348,
+      "step": 348
+    },
+    {
+      "epoch": 0.06714120815698346,
+      "grad_norm": 0.5811917474013277,
+      "learning_rate": 1.9928525932386678e-05,
+      "loss": 0.3188,
+      "step": 349
+    },
+    {
+      "epoch": 0.06733358984224702,
+      "grad_norm": 1.0945112033958047,
+      "learning_rate": 1.9927780371769463e-05,
+      "loss": 0.893,
+      "step": 350
+    },
+    {
+      "epoch": 0.06752597152751058,
+      "grad_norm": 1.1699279156062343,
+      "learning_rate": 1.9927030956846083e-05,
+      "loss": 0.8793,
+      "step": 351
+    },
+    {
+      "epoch": 0.06771835321277414,
+      "grad_norm": 1.0706549968704706,
+      "learning_rate": 1.992627768790749e-05,
+      "loss": 0.9493,
+      "step": 352
+    },
+    {
+      "epoch": 0.0679107348980377,
+      "grad_norm": 1.1319292680328992,
+      "learning_rate": 1.9925520565246125e-05,
+      "loss": 0.9511,
+      "step": 353
+    },
+    {
+      "epoch": 0.06810311658330127,
+      "grad_norm": 0.9284939685999337,
+      "learning_rate": 1.9924759589155932e-05,
+      "loss": 0.896,
+      "step": 354
+    },
+    {
+      "epoch": 0.06829549826856483,
+      "grad_norm": 1.060519909399317,
+      "learning_rate": 1.9923994759932344e-05,
+      "loss": 0.945,
+      "step": 355
+    },
+    {
+      "epoch": 0.06848787995382839,
+      "grad_norm": 1.0016887287592193,
+      "learning_rate": 1.9923226077872296e-05,
+      "loss": 0.9355,
+      "step": 356
+    },
+    {
+      "epoch": 0.06868026163909195,
+      "grad_norm": 1.12272960650074,
+      "learning_rate": 1.9922453543274223e-05,
+      "loss": 0.9827,
+      "step": 357
+    },
+    {
+      "epoch": 0.06887264332435553,
+      "grad_norm": 1.0284811073231566,
+      "learning_rate": 1.9921677156438044e-05,
+      "loss": 0.9365,
+      "step": 358
+    },
+    {
+      "epoch": 0.06906502500961909,
+      "grad_norm": 1.0234030515917725,
+      "learning_rate": 1.9920896917665178e-05,
+      "loss": 0.9324,
+      "step": 359
+    },
+    {
+      "epoch": 0.06925740669488265,
+      "grad_norm": 1.085209841869588,
+      "learning_rate": 1.992011282725854e-05,
+      "loss": 0.9705,
+      "step": 360
+    },
+    {
+      "epoch": 0.0694497883801462,
+      "grad_norm": 0.971426683575196,
+      "learning_rate": 1.9919324885522548e-05,
+      "loss": 0.8792,
+      "step": 361
+    },
+    {
+      "epoch": 0.06964217006540978,
+      "grad_norm": 0.9991431346297651,
+      "learning_rate": 1.99185330927631e-05,
+      "loss": 0.8733,
+      "step": 362
+    },
+    {
+      "epoch": 0.06983455175067334,
+      "grad_norm": 1.1142049176552107,
+      "learning_rate": 1.99177374492876e-05,
+      "loss": 0.9649,
+      "step": 363
+    },
+    {
+      "epoch": 0.0700269334359369,
+      "grad_norm": 0.6118036845349242,
+      "learning_rate": 1.991693795540494e-05,
+      "loss": 0.3073,
+      "step": 364
+    },
+    {
+      "epoch": 0.07021931512120046,
+      "grad_norm": 1.0901375356046195,
+      "learning_rate": 1.9916134611425522e-05,
+      "loss": 0.8765,
+      "step": 365
+    },
+    {
+      "epoch": 0.07041169680646403,
+      "grad_norm": 1.2544829594799247,
+      "learning_rate": 1.9915327417661226e-05,
+      "loss": 0.871,
+      "step": 366
+    },
+    {
+      "epoch": 0.07060407849172759,
+      "grad_norm": 0.6439298181894014,
+      "learning_rate": 1.991451637442543e-05,
+      "loss": 0.3062,
+      "step": 367
+    },
+    {
+      "epoch": 0.07079646017699115,
+      "grad_norm": 0.9353765247222083,
+      "learning_rate": 1.9913701482033008e-05,
+      "loss": 0.8449,
+      "step": 368
+    },
+    {
+      "epoch": 0.07098884186225471,
+      "grad_norm": 1.0475758702125963,
+      "learning_rate": 1.9912882740800336e-05,
+      "loss": 0.9019,
+      "step": 369
+    },
+    {
+      "epoch": 0.07118122354751828,
+      "grad_norm": 1.0250996088669797,
+      "learning_rate": 1.9912060151045273e-05,
+      "loss": 0.9482,
+      "step": 370
+    },
+    {
+      "epoch": 0.07137360523278184,
+      "grad_norm": 0.9657436269155218,
+      "learning_rate": 1.9911233713087172e-05,
+      "loss": 0.8427,
+      "step": 371
+    },
+    {
+      "epoch": 0.0715659869180454,
+      "grad_norm": 0.6273806476571704,
+      "learning_rate": 1.9910403427246895e-05,
+      "loss": 0.3238,
+      "step": 372
+    },
+    {
+      "epoch": 0.07175836860330896,
+      "grad_norm": 0.9883758046884479,
+      "learning_rate": 1.990956929384678e-05,
+      "loss": 0.9849,
+      "step": 373
+    },
+    {
+      "epoch": 0.07195075028857253,
+      "grad_norm": 0.9532844124529964,
+      "learning_rate": 1.990873131321067e-05,
+      "loss": 0.9233,
+      "step": 374
+    },
+    {
+      "epoch": 0.0721431319738361,
+      "grad_norm": 1.0940821497878852,
+      "learning_rate": 1.9907889485663897e-05,
+      "loss": 0.9493,
+      "step": 375
+    },
+    {
+      "epoch": 0.07233551365909965,
+      "grad_norm": 1.0385026510958237,
+      "learning_rate": 1.9907043811533283e-05,
+      "loss": 0.9682,
+      "step": 376
+    },
+    {
+      "epoch": 0.07252789534436321,
+      "grad_norm": 1.0231666607458285,
+      "learning_rate": 1.9906194291147155e-05,
+      "loss": 0.8859,
+      "step": 377
+    },
+    {
+      "epoch": 0.07272027702962677,
+      "grad_norm": 0.9954814151808695,
+      "learning_rate": 1.9905340924835322e-05,
+      "loss": 0.9075,
+      "step": 378
+    },
+    {
+      "epoch": 0.07291265871489035,
+      "grad_norm": 1.0253586269013863,
+      "learning_rate": 1.9904483712929094e-05,
+      "loss": 0.9688,
+      "step": 379
+    },
+    {
+      "epoch": 0.0731050404001539,
+      "grad_norm": 1.029899790199939,
+      "learning_rate": 1.9903622655761267e-05,
+      "loss": 0.9311,
+      "step": 380
+    },
+    {
+      "epoch": 0.07329742208541747,
+      "grad_norm": 1.128532405891048,
+      "learning_rate": 1.990275775366613e-05,
+      "loss": 0.9463,
+      "step": 381
+    },
+    {
+      "epoch": 0.07348980377068103,
+      "grad_norm": 1.1000912072662656,
+      "learning_rate": 1.9901889006979473e-05,
+      "loss": 0.9234,
+      "step": 382
+    },
+    {
+      "epoch": 0.0736821854559446,
+      "grad_norm": 1.0995801346670553,
+      "learning_rate": 1.990101641603857e-05,
+      "loss": 0.9755,
+      "step": 383
+    },
+    {
+      "epoch": 0.07387456714120816,
+      "grad_norm": 0.6543432250397458,
+      "learning_rate": 1.9900139981182193e-05,
+      "loss": 0.2765,
+      "step": 384
+    },
+    {
+      "epoch": 0.07406694882647172,
+      "grad_norm": 1.0264052461576711,
+      "learning_rate": 1.9899259702750604e-05,
+      "loss": 0.8669,
+      "step": 385
+    },
+    {
+      "epoch": 0.07425933051173528,
+      "grad_norm": 0.9945525983406796,
+      "learning_rate": 1.9898375581085555e-05,
+      "loss": 0.8773,
+      "step": 386
+    },
+    {
+      "epoch": 0.07445171219699885,
+      "grad_norm": 0.9966091372459741,
+      "learning_rate": 1.9897487616530296e-05,
+      "loss": 0.96,
+      "step": 387
+    },
+    {
+      "epoch": 0.07464409388226241,
+      "grad_norm": 0.9830672136659429,
+      "learning_rate": 1.9896595809429565e-05,
+      "loss": 0.9085,
+      "step": 388
+    },
+    {
+      "epoch": 0.07483647556752597,
+      "grad_norm": 0.9923214996306481,
+      "learning_rate": 1.9895700160129593e-05,
+      "loss": 0.9475,
+      "step": 389
+    },
+    {
+      "epoch": 0.07502885725278953,
+      "grad_norm": 1.038503998353339,
+      "learning_rate": 1.9894800668978095e-05,
+      "loss": 0.8524,
+      "step": 390
+    },
+    {
+      "epoch": 0.0752212389380531,
+      "grad_norm": 0.9345795796822466,
+      "learning_rate": 1.9893897336324292e-05,
+      "loss": 0.8763,
+      "step": 391
+    },
+    {
+      "epoch": 0.07541362062331666,
+      "grad_norm": 1.1562581537109837,
+      "learning_rate": 1.9892990162518884e-05,
+      "loss": 0.9078,
+      "step": 392
+    },
+    {
+      "epoch": 0.07560600230858022,
+      "grad_norm": 0.9778957473243307,
+      "learning_rate": 1.9892079147914072e-05,
+      "loss": 0.8958,
+      "step": 393
+    },
+    {
+      "epoch": 0.07579838399384378,
+      "grad_norm": 1.0109125181935874,
+      "learning_rate": 1.9891164292863537e-05,
+      "loss": 0.8989,
+      "step": 394
+    },
+    {
+      "epoch": 0.07599076567910736,
+      "grad_norm": 0.9500548957669385,
+      "learning_rate": 1.9890245597722465e-05,
+      "loss": 0.9362,
+      "step": 395
+    },
+    {
+      "epoch": 0.07618314736437092,
+      "grad_norm": 1.0395660826324193,
+      "learning_rate": 1.9889323062847516e-05,
+      "loss": 0.9361,
+      "step": 396
+    },
+    {
+      "epoch": 0.07637552904963447,
+      "grad_norm": 1.135126546764037,
+      "learning_rate": 1.988839668859686e-05,
+      "loss": 0.9026,
+      "step": 397
+    },
+    {
+      "epoch": 0.07656791073489803,
+      "grad_norm": 0.9311331847379872,
+      "learning_rate": 1.988746647533014e-05,
+      "loss": 0.9522,
+      "step": 398
+    },
+    {
+      "epoch": 0.0767602924201616,
+      "grad_norm": 1.0046206839811358,
+      "learning_rate": 1.9886532423408495e-05,
+      "loss": 0.9727,
+      "step": 399
+    },
+    {
+      "epoch": 0.07695267410542517,
+      "grad_norm": 1.036868156126891,
+      "learning_rate": 1.9885594533194564e-05,
+      "loss": 0.911,
+      "step": 400
+    },
+    {
+      "epoch": 0.07714505579068873,
+      "grad_norm": 1.0018274649809766,
+      "learning_rate": 1.9884652805052465e-05,
+      "loss": 0.9233,
+      "step": 401
+    },
+    {
+      "epoch": 0.07733743747595229,
+      "grad_norm": 1.0618957763048602,
+      "learning_rate": 1.9883707239347804e-05,
+      "loss": 0.9277,
+      "step": 402
+    },
+    {
+      "epoch": 0.07752981916121585,
+      "grad_norm": 0.9538465425804193,
+      "learning_rate": 1.988275783644769e-05,
+      "loss": 0.8805,
+      "step": 403
+    },
+    {
+      "epoch": 0.07772220084647942,
+      "grad_norm": 1.0349620746453256,
+      "learning_rate": 1.988180459672071e-05,
+      "loss": 0.9654,
+      "step": 404
+    },
+    {
+      "epoch": 0.07791458253174298,
+      "grad_norm": 0.9705138748505167,
+      "learning_rate": 1.988084752053695e-05,
+      "loss": 0.8688,
+      "step": 405
+    },
+    {
+      "epoch": 0.07810696421700654,
+      "grad_norm": 1.0581263772230698,
+      "learning_rate": 1.9879886608267967e-05,
+      "loss": 0.9869,
+      "step": 406
+    },
+    {
+      "epoch": 0.0782993459022701,
+      "grad_norm": 1.0473405407371956,
+      "learning_rate": 1.9878921860286832e-05,
+      "loss": 0.9283,
+      "step": 407
+    },
+    {
+      "epoch": 0.07849172758753367,
+      "grad_norm": 1.0303865431324983,
+      "learning_rate": 1.9877953276968088e-05,
+      "loss": 0.8589,
+      "step": 408
+    },
+    {
+      "epoch": 0.07868410927279723,
+      "grad_norm": 0.9588533164137831,
+      "learning_rate": 1.9876980858687777e-05,
+      "loss": 0.9002,
+      "step": 409
+    },
+    {
+      "epoch": 0.07887649095806079,
+      "grad_norm": 1.1050618227988047,
+      "learning_rate": 1.9876004605823417e-05,
+      "loss": 0.917,
+      "step": 410
+    },
+    {
+      "epoch": 0.07906887264332435,
+      "grad_norm": 1.072669094352535,
+      "learning_rate": 1.987502451875403e-05,
+      "loss": 0.9542,
+      "step": 411
+    },
+    {
+      "epoch": 0.07926125432858792,
+      "grad_norm": 1.003058891028112,
+      "learning_rate": 1.987404059786012e-05,
+      "loss": 0.885,
+      "step": 412
+    },
+    {
+      "epoch": 0.07945363601385148,
+      "grad_norm": 1.0438872867794595,
+      "learning_rate": 1.9873052843523676e-05,
+      "loss": 0.9447,
+      "step": 413
+    },
+    {
+      "epoch": 0.07964601769911504,
+      "grad_norm": 1.1055668519306072,
+      "learning_rate": 1.987206125612818e-05,
+      "loss": 0.9365,
+      "step": 414
+    },
+    {
+      "epoch": 0.0798383993843786,
+      "grad_norm": 1.0556097247391998,
+      "learning_rate": 1.98710658360586e-05,
+      "loss": 0.9037,
+      "step": 415
+    },
+    {
+      "epoch": 0.08003078106964218,
+      "grad_norm": 1.065702058190967,
+      "learning_rate": 1.987006658370139e-05,
+      "loss": 0.9346,
+      "step": 416
+    },
+    {
+      "epoch": 0.08022316275490574,
+      "grad_norm": 1.022491011488092,
+      "learning_rate": 1.9869063499444495e-05,
+      "loss": 0.9036,
+      "step": 417
+    },
+    {
+      "epoch": 0.0804155444401693,
+      "grad_norm": 0.9432999737256113,
+      "learning_rate": 1.9868056583677346e-05,
+      "loss": 0.8832,
+      "step": 418
+    },
+    {
+      "epoch": 0.08060792612543285,
+      "grad_norm": 0.6999679458678303,
+      "learning_rate": 1.9867045836790867e-05,
+      "loss": 0.2994,
+      "step": 419
+    },
+    {
+      "epoch": 0.08080030781069643,
+      "grad_norm": 1.216335757087175,
+      "learning_rate": 1.9866031259177463e-05,
+      "loss": 0.99,
+      "step": 420
+    },
+    {
+      "epoch": 0.08099268949595999,
+      "grad_norm": 1.075952942317633,
+      "learning_rate": 1.9865012851231022e-05,
+      "loss": 0.9117,
+      "step": 421
+    },
+    {
+      "epoch": 0.08118507118122355,
+      "grad_norm": 1.0322406745734685,
+      "learning_rate": 1.9863990613346936e-05,
+      "loss": 0.9391,
+      "step": 422
+    },
+    {
+      "epoch": 0.08137745286648711,
+      "grad_norm": 0.9600484594537925,
+      "learning_rate": 1.986296454592206e-05,
+      "loss": 0.8609,
+      "step": 423
+    },
+    {
+      "epoch": 0.08156983455175067,
+      "grad_norm": 1.0267824020155865,
+      "learning_rate": 1.9861934649354763e-05,
+      "loss": 0.8689,
+      "step": 424
+    },
+    {
+      "epoch": 0.08176221623701424,
+      "grad_norm": 0.6727079104797604,
+      "learning_rate": 1.9860900924044873e-05,
+      "loss": 0.3236,
+      "step": 425
+    },
+    {
+      "epoch": 0.0819545979222778,
+      "grad_norm": 1.0596330983288555,
+      "learning_rate": 1.9859863370393726e-05,
+      "loss": 0.8694,
+      "step": 426
+    },
+    {
+      "epoch": 0.08214697960754136,
+      "grad_norm": 1.0552378240404465,
+      "learning_rate": 1.9858821988804132e-05,
+      "loss": 0.9417,
+      "step": 427
+    },
+    {
+      "epoch": 0.08233936129280492,
+      "grad_norm": 1.084277795510405,
+      "learning_rate": 1.9857776779680393e-05,
+      "loss": 0.9501,
+      "step": 428
+    },
+    {
+      "epoch": 0.08253174297806849,
+      "grad_norm": 0.9188349471656138,
+      "learning_rate": 1.98567277434283e-05,
+      "loss": 0.9162,
+      "step": 429
+    },
+    {
+      "epoch": 0.08272412466333205,
+      "grad_norm": 0.9699920881226191,
+      "learning_rate": 1.9855674880455115e-05,
+      "loss": 0.9101,
+      "step": 430
+    },
+    {
+      "epoch": 0.08291650634859561,
+      "grad_norm": 1.0874560153082604,
+      "learning_rate": 1.98546181911696e-05,
+      "loss": 0.9529,
+      "step": 431
+    },
+    {
+      "epoch": 0.08310888803385917,
+      "grad_norm": 1.04004275807754,
+      "learning_rate": 1.9853557675982e-05,
+      "loss": 0.952,
+      "step": 432
+    },
+    {
+      "epoch": 0.08330126971912274,
+      "grad_norm": 1.013743225281868,
+      "learning_rate": 1.985249333530404e-05,
+      "loss": 0.9395,
+      "step": 433
+    },
+    {
+      "epoch": 0.0834936514043863,
+      "grad_norm": 0.9694672053174421,
+      "learning_rate": 1.9851425169548938e-05,
+      "loss": 0.88,
+      "step": 434
+    },
+    {
+      "epoch": 0.08368603308964986,
+      "grad_norm": 0.9957961100875826,
+      "learning_rate": 1.9850353179131392e-05,
+      "loss": 0.9581,
+      "step": 435
+    },
+    {
+      "epoch": 0.08387841477491342,
+      "grad_norm": 0.6715432336970756,
+      "learning_rate": 1.9849277364467585e-05,
+      "loss": 0.3297,
+      "step": 436
+    },
+    {
+      "epoch": 0.084070796460177,
+      "grad_norm": 1.0539061126342106,
+      "learning_rate": 1.984819772597518e-05,
+      "loss": 0.8682,
+      "step": 437
+    },
+    {
+      "epoch": 0.08426317814544056,
+      "grad_norm": 0.9898493964937313,
+      "learning_rate": 1.9847114264073336e-05,
+      "loss": 0.9277,
+      "step": 438
+    },
+    {
+      "epoch": 0.08445555983070412,
+      "grad_norm": 1.0421406609479855,
+      "learning_rate": 1.984602697918269e-05,
+      "loss": 0.9056,
+      "step": 439
+    },
+    {
+      "epoch": 0.08464794151596768,
+      "grad_norm": 0.9922773668459134,
+      "learning_rate": 1.9844935871725363e-05,
+      "loss": 0.9417,
+      "step": 440
+    },
+    {
+      "epoch": 0.08484032320123125,
+      "grad_norm": 0.9625979435036303,
+      "learning_rate": 1.9843840942124956e-05,
+      "loss": 0.885,
+      "step": 441
+    },
+    {
+      "epoch": 0.08503270488649481,
+      "grad_norm": 0.9799071013097026,
+      "learning_rate": 1.9842742190806566e-05,
+      "loss": 0.96,
+      "step": 442
+    },
+    {
+      "epoch": 0.08522508657175837,
+      "grad_norm": 0.9535229778812249,
+      "learning_rate": 1.984163961819676e-05,
+      "loss": 0.9099,
+      "step": 443
+    },
+    {
+      "epoch": 0.08541746825702193,
+      "grad_norm": 1.019248253699148,
+      "learning_rate": 1.9840533224723595e-05,
+      "loss": 0.9168,
+      "step": 444
+    },
+    {
+      "epoch": 0.08560984994228549,
+      "grad_norm": 1.02579807808078,
+      "learning_rate": 1.9839423010816616e-05,
+      "loss": 0.8622,
+      "step": 445
+    },
+    {
+      "epoch": 0.08580223162754906,
+      "grad_norm": 1.04105837698965,
+      "learning_rate": 1.983830897690684e-05,
+      "loss": 0.9695,
+      "step": 446
+    },
+    {
+      "epoch": 0.08599461331281262,
+      "grad_norm": 1.0379393020744019,
+      "learning_rate": 1.9837191123426777e-05,
+      "loss": 0.9513,
+      "step": 447
+    },
+    {
+      "epoch": 0.08618699499807618,
+      "grad_norm": 0.9811051621676831,
+      "learning_rate": 1.983606945081042e-05,
+      "loss": 0.9315,
+      "step": 448
+    },
+    {
+      "epoch": 0.08637937668333974,
+      "grad_norm": 0.9998357964075436,
+      "learning_rate": 1.983494395949323e-05,
+      "loss": 0.9627,
+      "step": 449
+    },
+    {
+      "epoch": 0.08657175836860331,
+      "grad_norm": 0.9569033846194281,
+      "learning_rate": 1.983381464991217e-05,
+      "loss": 0.8955,
+      "step": 450
+    },
+    {
+      "epoch": 0.08676414005386687,
+      "grad_norm": 0.9562590933922375,
+      "learning_rate": 1.9832681522505676e-05,
+      "loss": 0.861,
+      "step": 451
+    },
+    {
+      "epoch": 0.08695652173913043,
+      "grad_norm": 0.9177384664710877,
+      "learning_rate": 1.9831544577713663e-05,
+      "loss": 0.9095,
+      "step": 452
+    },
+    {
+      "epoch": 0.08714890342439399,
+      "grad_norm": 1.0722951958115343,
+      "learning_rate": 1.983040381597754e-05,
+      "loss": 0.9505,
+      "step": 453
+    },
+    {
+      "epoch": 0.08734128510965757,
+      "grad_norm": 0.7452446844953681,
+      "learning_rate": 1.982925923774018e-05,
+      "loss": 0.3107,
+      "step": 454
+    },
+    {
+      "epoch": 0.08753366679492113,
+      "grad_norm": 0.6639023210170698,
+      "learning_rate": 1.9828110843445954e-05,
+      "loss": 0.3533,
+      "step": 455
+    },
+    {
+      "epoch": 0.08772604848018468,
+      "grad_norm": 1.0516642291636649,
+      "learning_rate": 1.982695863354071e-05,
+      "loss": 0.9074,
+      "step": 456
+    },
+    {
+      "epoch": 0.08791843016544824,
+      "grad_norm": 0.9929126834155242,
+      "learning_rate": 1.9825802608471767e-05,
+      "loss": 0.8592,
+      "step": 457
+    },
+    {
+      "epoch": 0.08811081185071182,
+      "grad_norm": 1.1210682578006375,
+      "learning_rate": 1.982464276868794e-05,
+      "loss": 0.9285,
+      "step": 458
+    },
+    {
+      "epoch": 0.08830319353597538,
+      "grad_norm": 1.061834090069009,
+      "learning_rate": 1.982347911463952e-05,
+      "loss": 0.9951,
+      "step": 459
+    },
+    {
+      "epoch": 0.08849557522123894,
+      "grad_norm": 0.9313842164987207,
+      "learning_rate": 1.9822311646778277e-05,
+      "loss": 0.9405,
+      "step": 460
+    },
+    {
+      "epoch": 0.0886879569065025,
+      "grad_norm": 1.1452959818514865,
+      "learning_rate": 1.982114036555746e-05,
+      "loss": 0.9256,
+      "step": 461
+    },
+    {
+      "epoch": 0.08888033859176607,
+      "grad_norm": 1.0292761249211673,
+      "learning_rate": 1.9819965271431797e-05,
+      "loss": 0.937,
+      "step": 462
+    },
+    {
+      "epoch": 0.08907272027702963,
+      "grad_norm": 1.0317484174952016,
+      "learning_rate": 1.9818786364857506e-05,
+      "loss": 0.8912,
+      "step": 463
+    },
+    {
+      "epoch": 0.08926510196229319,
+      "grad_norm": 0.9236217298635157,
+      "learning_rate": 1.9817603646292278e-05,
+      "loss": 0.8941,
+      "step": 464
+    },
+    {
+      "epoch": 0.08945748364755675,
+      "grad_norm": 0.9742505606827352,
+      "learning_rate": 1.9816417116195287e-05,
+      "loss": 0.897,
+      "step": 465
+    },
+    {
+      "epoch": 0.08964986533282032,
+      "grad_norm": 0.834652119096309,
+      "learning_rate": 1.9815226775027182e-05,
+      "loss": 0.3262,
+      "step": 466
+    },
+    {
+      "epoch": 0.08984224701808388,
+      "grad_norm": 1.0300150034840316,
+      "learning_rate": 1.9814032623250093e-05,
+      "loss": 0.9242,
+      "step": 467
+    },
+    {
+      "epoch": 0.09003462870334744,
+      "grad_norm": 1.0249092145886567,
+      "learning_rate": 1.9812834661327632e-05,
+      "loss": 0.898,
+      "step": 468
+    },
+    {
+      "epoch": 0.090227010388611,
+      "grad_norm": 0.9841774003700253,
+      "learning_rate": 1.9811632889724888e-05,
+      "loss": 0.88,
+      "step": 469
+    },
+    {
+      "epoch": 0.09041939207387456,
+      "grad_norm": 0.9742493204558247,
+      "learning_rate": 1.9810427308908437e-05,
+      "loss": 0.9263,
+      "step": 470
+    },
+    {
+      "epoch": 0.09061177375913813,
+      "grad_norm": 1.0961562100486952,
+      "learning_rate": 1.9809217919346318e-05,
+      "loss": 0.9149,
+      "step": 471
+    },
+    {
+      "epoch": 0.0908041554444017,
+      "grad_norm": 1.0562371696450448,
+      "learning_rate": 1.980800472150806e-05,
+      "loss": 0.9333,
+      "step": 472
+    },
+    {
+      "epoch": 0.09099653712966525,
+      "grad_norm": 0.9897231870436988,
+      "learning_rate": 1.9806787715864674e-05,
+      "loss": 0.8986,
+      "step": 473
+    },
+    {
+      "epoch": 0.09118891881492881,
+      "grad_norm": 0.9872034201237571,
+      "learning_rate": 1.9805566902888637e-05,
+      "loss": 0.9493,
+      "step": 474
+    },
+    {
+      "epoch": 0.09138130050019239,
+      "grad_norm": 0.9182753933629554,
+      "learning_rate": 1.9804342283053916e-05,
+      "loss": 0.8365,
+      "step": 475
+    },
+    {
+      "epoch": 0.09157368218545595,
+      "grad_norm": 0.9359304409197167,
+      "learning_rate": 1.980311385683594e-05,
+      "loss": 0.9309,
+      "step": 476
+    },
+    {
+      "epoch": 0.0917660638707195,
+      "grad_norm": 1.0253981953613478,
+      "learning_rate": 1.980188162471164e-05,
+      "loss": 0.8646,
+      "step": 477
+    },
+    {
+      "epoch": 0.09195844555598306,
+      "grad_norm": 1.058302924879132,
+      "learning_rate": 1.98006455871594e-05,
+      "loss": 0.9281,
+      "step": 478
+    },
+    {
+      "epoch": 0.09215082724124664,
+      "grad_norm": 1.001973809930885,
+      "learning_rate": 1.97994057446591e-05,
+      "loss": 0.8995,
+      "step": 479
+    },
+    {
+      "epoch": 0.0923432089265102,
+      "grad_norm": 1.1431414999899652,
+      "learning_rate": 1.979816209769209e-05,
+      "loss": 0.9606,
+      "step": 480
+    },
+    {
+      "epoch": 0.09253559061177376,
+      "grad_norm": 1.0557988633314797,
+      "learning_rate": 1.9796914646741187e-05,
+      "loss": 0.9344,
+      "step": 481
+    },
+    {
+      "epoch": 0.09272797229703732,
+      "grad_norm": 0.9266429897398724,
+      "learning_rate": 1.9795663392290702e-05,
+      "loss": 0.3352,
+      "step": 482
+    },
+    {
+      "epoch": 0.09292035398230089,
+      "grad_norm": 1.0709694113209445,
+      "learning_rate": 1.9794408334826415e-05,
+      "loss": 0.9387,
+      "step": 483
+    },
+    {
+      "epoch": 0.09311273566756445,
+      "grad_norm": 0.9259525239189322,
+      "learning_rate": 1.979314947483558e-05,
+      "loss": 0.9093,
+      "step": 484
+    },
+    {
+      "epoch": 0.09330511735282801,
+      "grad_norm": 1.0594650973765662,
+      "learning_rate": 1.9791886812806932e-05,
+      "loss": 0.8706,
+      "step": 485
+    },
+    {
+      "epoch": 0.09349749903809157,
+      "grad_norm": 0.6741027343664249,
+      "learning_rate": 1.9790620349230676e-05,
+      "loss": 0.317,
+      "step": 486
+    },
+    {
+      "epoch": 0.09368988072335514,
+      "grad_norm": 1.0331925285088455,
+      "learning_rate": 1.9789350084598504e-05,
+      "loss": 0.9288,
+      "step": 487
+    },
+    {
+      "epoch": 0.0938822624086187,
+      "grad_norm": 1.0161984531677009,
+      "learning_rate": 1.9788076019403565e-05,
+      "loss": 0.8934,
+      "step": 488
+    },
+    {
+      "epoch": 0.09407464409388226,
+      "grad_norm": 0.8993751479632649,
+      "learning_rate": 1.9786798154140507e-05,
+      "loss": 0.8846,
+      "step": 489
+    },
+    {
+      "epoch": 0.09426702577914582,
+      "grad_norm": 1.1104904915599816,
+      "learning_rate": 1.9785516489305437e-05,
+      "loss": 0.9341,
+      "step": 490
+    },
+    {
+      "epoch": 0.0944594074644094,
+      "grad_norm": 0.9223993708575459,
+      "learning_rate": 1.9784231025395936e-05,
+      "loss": 0.9567,
+      "step": 491
+    },
+    {
+      "epoch": 0.09465178914967295,
+      "grad_norm": 0.9702233165038455,
+      "learning_rate": 1.9782941762911075e-05,
+      "loss": 0.9106,
+      "step": 492
+    },
+    {
+      "epoch": 0.09484417083493651,
+      "grad_norm": 0.9486718821028429,
+      "learning_rate": 1.9781648702351383e-05,
+      "loss": 0.9664,
+      "step": 493
+    },
+    {
+      "epoch": 0.09503655252020007,
+      "grad_norm": 1.0141797445631344,
+      "learning_rate": 1.9780351844218874e-05,
+      "loss": 0.9196,
+      "step": 494
+    },
+    {
+      "epoch": 0.09522893420546363,
+      "grad_norm": 0.8773297470256214,
+      "learning_rate": 1.977905118901703e-05,
+      "loss": 0.8911,
+      "step": 495
+    },
+    {
+      "epoch": 0.0954213158907272,
+      "grad_norm": 1.0267147257766212,
+      "learning_rate": 1.977774673725081e-05,
+      "loss": 0.9331,
+      "step": 496
+    },
+    {
+      "epoch": 0.09561369757599077,
+      "grad_norm": 1.0040796029036745,
+      "learning_rate": 1.977643848942665e-05,
+      "loss": 0.8933,
+      "step": 497
+    },
+    {
+      "epoch": 0.09580607926125433,
+      "grad_norm": 0.851700486454822,
+      "learning_rate": 1.977512644605246e-05,
+      "loss": 0.3356,
+      "step": 498
+    },
+    {
+      "epoch": 0.09599846094651789,
+      "grad_norm": 0.9595864969052292,
+      "learning_rate": 1.9773810607637612e-05,
+      "loss": 0.9111,
+      "step": 499
+    },
+    {
+      "epoch": 0.09619084263178146,
+      "grad_norm": 0.9774839283505296,
+      "learning_rate": 1.9772490974692962e-05,
+      "loss": 0.9072,
+      "step": 500
+    },
+    {
+      "epoch": 0.09638322431704502,
+      "grad_norm": 0.9501948845687158,
+      "learning_rate": 1.9771167547730844e-05,
+      "loss": 0.8944,
+      "step": 501
+    },
+    {
+      "epoch": 0.09657560600230858,
+      "grad_norm": 0.994593032780812,
+      "learning_rate": 1.976984032726505e-05,
+      "loss": 0.9085,
+      "step": 502
+    },
+    {
+      "epoch": 0.09676798768757214,
+      "grad_norm": 0.9753050112142359,
+      "learning_rate": 1.976850931381086e-05,
+      "loss": 0.9037,
+      "step": 503
+    },
+    {
+      "epoch": 0.09696036937283571,
+      "grad_norm": 0.9841635275551259,
+      "learning_rate": 1.976717450788501e-05,
+      "loss": 0.9587,
+      "step": 504
+    },
+    {
+      "epoch": 0.09715275105809927,
+      "grad_norm": 0.9929362225265581,
+      "learning_rate": 1.9765835910005726e-05,
+      "loss": 0.8984,
+      "step": 505
+    },
+    {
+      "epoch": 0.09734513274336283,
+      "grad_norm": 1.0905836004839264,
+      "learning_rate": 1.9764493520692685e-05,
+      "loss": 0.9391,
+      "step": 506
+    },
+    {
+      "epoch": 0.09753751442862639,
+      "grad_norm": 0.945657920406541,
+      "learning_rate": 1.9763147340467067e-05,
+      "loss": 0.8931,
+      "step": 507
+    },
+    {
+      "epoch": 0.09772989611388996,
+      "grad_norm": 0.9534697841091134,
+      "learning_rate": 1.9761797369851498e-05,
+      "loss": 0.874,
+      "step": 508
+    },
+    {
+      "epoch": 0.09792227779915352,
+      "grad_norm": 0.990267177185285,
+      "learning_rate": 1.9760443609370074e-05,
+      "loss": 0.3437,
+      "step": 509
+    },
+    {
+      "epoch": 0.09811465948441708,
+      "grad_norm": 1.0487588013059468,
+      "learning_rate": 1.975908605954838e-05,
+      "loss": 0.9531,
+      "step": 510
+    },
+    {
+      "epoch": 0.09830704116968064,
+      "grad_norm": 1.0260253321726265,
+      "learning_rate": 1.9757724720913466e-05,
+      "loss": 0.8997,
+      "step": 511
+    },
+    {
+      "epoch": 0.09849942285494422,
+      "grad_norm": 0.9274626307003517,
+      "learning_rate": 1.9756359593993845e-05,
+      "loss": 0.9163,
+      "step": 512
+    },
+    {
+      "epoch": 0.09869180454020778,
+      "grad_norm": 0.9581453260150369,
+      "learning_rate": 1.975499067931951e-05,
+      "loss": 0.8701,
+      "step": 513
+    },
+    {
+      "epoch": 0.09888418622547133,
+      "grad_norm": 1.1681904254503332,
+      "learning_rate": 1.975361797742192e-05,
+      "loss": 0.8803,
+      "step": 514
+    },
+    {
+      "epoch": 0.0990765679107349,
+      "grad_norm": 1.0310311967811059,
+      "learning_rate": 1.9752241488834002e-05,
+      "loss": 0.9535,
+      "step": 515
+    },
+    {
+      "epoch": 0.09926894959599845,
+      "grad_norm": 0.9639710035760624,
+      "learning_rate": 1.975086121409016e-05,
+      "loss": 0.9051,
+      "step": 516
+    },
+    {
+      "epoch": 0.09946133128126203,
+      "grad_norm": 1.0883855291368902,
+      "learning_rate": 1.974947715372626e-05,
+      "loss": 0.9226,
+      "step": 517
+    },
+    {
+      "epoch": 0.09965371296652559,
+      "grad_norm": 1.008207772273884,
+      "learning_rate": 1.974808930827965e-05,
+      "loss": 0.9086,
+      "step": 518
+    },
+    {
+      "epoch": 0.09984609465178915,
+      "grad_norm": 1.0784335413452422,
+      "learning_rate": 1.9746697678289128e-05,
+      "loss": 0.9895,
+      "step": 519
+    },
+    {
+      "epoch": 0.1000384763370527,
+      "grad_norm": 1.1935990982315445,
+      "learning_rate": 1.9745302264294982e-05,
+      "loss": 0.9265,
+      "step": 520
+    },
+    {
+      "epoch": 0.10023085802231628,
+      "grad_norm": 0.9828814778483683,
+      "learning_rate": 1.9743903066838954e-05,
+      "loss": 0.8917,
+      "step": 521
+    },
+    {
+      "epoch": 0.10042323970757984,
+      "grad_norm": 0.9818218946775493,
+      "learning_rate": 1.9742500086464266e-05,
+      "loss": 0.8353,
+      "step": 522
+    },
+    {
+      "epoch": 0.1006156213928434,
+      "grad_norm": 1.029828927788772,
+      "learning_rate": 1.9741093323715597e-05,
+      "loss": 0.9166,
+      "step": 523
+    },
+    {
+      "epoch": 0.10080800307810696,
+      "grad_norm": 1.072782231989912,
+      "learning_rate": 1.9739682779139107e-05,
+      "loss": 0.8784,
+      "step": 524
+    },
+    {
+      "epoch": 0.10100038476337053,
+      "grad_norm": 0.9664182075044111,
+      "learning_rate": 1.9738268453282414e-05,
+      "loss": 0.8792,
+      "step": 525
+    },
+    {
+      "epoch": 0.10119276644863409,
+      "grad_norm": 1.024310148511455,
+      "learning_rate": 1.9736850346694608e-05,
+      "loss": 0.9847,
+      "step": 526
+    },
+    {
+      "epoch": 0.10138514813389765,
+      "grad_norm": 1.0439015361270212,
+      "learning_rate": 1.973542845992625e-05,
+      "loss": 0.9243,
+      "step": 527
+    },
+    {
+      "epoch": 0.10157752981916121,
+      "grad_norm": 0.8854085941696027,
+      "learning_rate": 1.9734002793529362e-05,
+      "loss": 0.8988,
+      "step": 528
+    },
+    {
+      "epoch": 0.10176991150442478,
+      "grad_norm": 0.9834327124684266,
+      "learning_rate": 1.9732573348057437e-05,
+      "loss": 0.8534,
+      "step": 529
+    },
+    {
+      "epoch": 0.10196229318968834,
+      "grad_norm": 0.9841940612772435,
+      "learning_rate": 1.973114012406544e-05,
+      "loss": 0.9289,
+      "step": 530
+    },
+    {
+      "epoch": 0.1021546748749519,
+      "grad_norm": 0.9542413534527358,
+      "learning_rate": 1.9729703122109788e-05,
+      "loss": 0.9128,
+      "step": 531
+    },
+    {
+      "epoch": 0.10234705656021546,
+      "grad_norm": 0.9271786000250268,
+      "learning_rate": 1.9728262342748384e-05,
+      "loss": 0.9197,
+      "step": 532
+    },
+    {
+      "epoch": 0.10253943824547904,
+      "grad_norm": 1.0316604892646375,
+      "learning_rate": 1.9726817786540584e-05,
+      "loss": 0.8697,
+      "step": 533
+    },
+    {
+      "epoch": 0.1027318199307426,
+      "grad_norm": 1.0545561042031846,
+      "learning_rate": 1.9725369454047215e-05,
+      "loss": 0.896,
+      "step": 534
+    },
+    {
+      "epoch": 0.10292420161600616,
+      "grad_norm": 1.0890906104137266,
+      "learning_rate": 1.9723917345830568e-05,
+      "loss": 0.9,
+      "step": 535
+    },
+    {
+      "epoch": 0.10311658330126972,
+      "grad_norm": 0.975943844138534,
+      "learning_rate": 1.9722461462454405e-05,
+      "loss": 0.9008,
+      "step": 536
+    },
+    {
+      "epoch": 0.10330896498653329,
+      "grad_norm": 0.9388711289917734,
+      "learning_rate": 1.9721001804483947e-05,
+      "loss": 0.9504,
+      "step": 537
+    },
+    {
+      "epoch": 0.10350134667179685,
+      "grad_norm": 0.9430373218327933,
+      "learning_rate": 1.9719538372485887e-05,
+      "loss": 0.8779,
+      "step": 538
+    },
+    {
+      "epoch": 0.10369372835706041,
+      "grad_norm": 1.0614259611173809,
+      "learning_rate": 1.9718071167028376e-05,
+      "loss": 0.8808,
+      "step": 539
+    },
+    {
+      "epoch": 0.10388611004232397,
+      "grad_norm": 1.0161675237109216,
+      "learning_rate": 1.9716600188681038e-05,
+      "loss": 0.8799,
+      "step": 540
+    },
+    {
+      "epoch": 0.10407849172758753,
+      "grad_norm": 0.9290407835372939,
+      "learning_rate": 1.971512543801495e-05,
+      "loss": 0.861,
+      "step": 541
+    },
+    {
+      "epoch": 0.1042708734128511,
+      "grad_norm": 0.9918810880586391,
+      "learning_rate": 1.9713646915602663e-05,
+      "loss": 0.9401,
+      "step": 542
+    },
+    {
+      "epoch": 0.10446325509811466,
+      "grad_norm": 0.952177949718571,
+      "learning_rate": 1.9712164622018197e-05,
+      "loss": 0.9696,
+      "step": 543
+    },
+    {
+      "epoch": 0.10465563678337822,
+      "grad_norm": 1.0272692899522706,
+      "learning_rate": 1.9710678557837024e-05,
+      "loss": 0.9257,
+      "step": 544
+    },
+    {
+      "epoch": 0.10484801846864178,
+      "grad_norm": 1.0081563559560947,
+      "learning_rate": 1.9709188723636088e-05,
+      "loss": 0.9303,
+      "step": 545
+    },
+    {
+      "epoch": 0.10504040015390535,
+      "grad_norm": 1.029051597494217,
+      "learning_rate": 1.970769511999379e-05,
+      "loss": 0.9078,
+      "step": 546
+    },
+    {
+      "epoch": 0.10523278183916891,
+      "grad_norm": 0.9818301984972814,
+      "learning_rate": 1.9706197747490004e-05,
+      "loss": 0.923,
+      "step": 547
+    },
+    {
+      "epoch": 0.10542516352443247,
+      "grad_norm": 0.9247093534074767,
+      "learning_rate": 1.9704696606706055e-05,
+      "loss": 0.831,
+      "step": 548
+    },
+    {
+      "epoch": 0.10561754520969603,
+      "grad_norm": 1.047253401046982,
+      "learning_rate": 1.9703191698224742e-05,
+      "loss": 0.9038,
+      "step": 549
+    },
+    {
+      "epoch": 0.1058099268949596,
+      "grad_norm": 0.9427852708004177,
+      "learning_rate": 1.9701683022630323e-05,
+      "loss": 0.9281,
+      "step": 550
+    },
+    {
+      "epoch": 0.10600230858022316,
+      "grad_norm": 0.9180129909610147,
+      "learning_rate": 1.9700170580508514e-05,
+      "loss": 0.9026,
+      "step": 551
+    },
+    {
+      "epoch": 0.10619469026548672,
+      "grad_norm": 0.9876251086276293,
+      "learning_rate": 1.9698654372446495e-05,
+      "loss": 0.9648,
+      "step": 552
+    },
+    {
+      "epoch": 0.10638707195075028,
+      "grad_norm": 1.0259054550776607,
+      "learning_rate": 1.969713439903292e-05,
+      "loss": 0.9333,
+      "step": 553
+    },
+    {
+      "epoch": 0.10657945363601386,
+      "grad_norm": 1.0565101383896163,
+      "learning_rate": 1.9695610660857886e-05,
+      "loss": 0.9498,
+      "step": 554
+    },
+    {
+      "epoch": 0.10677183532127742,
+      "grad_norm": 1.0022940291048168,
+      "learning_rate": 1.9694083158512965e-05,
+      "loss": 0.9545,
+      "step": 555
+    },
+    {
+      "epoch": 0.10696421700654098,
+      "grad_norm": 0.9089889096254266,
+      "learning_rate": 1.9692551892591185e-05,
+      "loss": 0.8933,
+      "step": 556
+    },
+    {
+      "epoch": 0.10715659869180454,
+      "grad_norm": 1.082765611821337,
+      "learning_rate": 1.9691016863687037e-05,
+      "loss": 0.9057,
+      "step": 557
+    },
+    {
+      "epoch": 0.10734898037706811,
+      "grad_norm": 1.0071311525258733,
+      "learning_rate": 1.968947807239647e-05,
+      "loss": 0.9624,
+      "step": 558
+    },
+    {
+      "epoch": 0.10754136206233167,
+      "grad_norm": 1.121370431933782,
+      "learning_rate": 1.9687935519316897e-05,
+      "loss": 0.3283,
+      "step": 559
+    },
+    {
+      "epoch": 0.10773374374759523,
+      "grad_norm": 0.9384747314000655,
+      "learning_rate": 1.9686389205047186e-05,
+      "loss": 0.9734,
+      "step": 560
+    },
+    {
+      "epoch": 0.10792612543285879,
+      "grad_norm": 1.0265392833332765,
+      "learning_rate": 1.9684839130187678e-05,
+      "loss": 0.9088,
+      "step": 561
+    },
+    {
+      "epoch": 0.10811850711812235,
+      "grad_norm": 1.13720066905895,
+      "learning_rate": 1.968328529534016e-05,
+      "loss": 0.8851,
+      "step": 562
+    },
+    {
+      "epoch": 0.10831088880338592,
+      "grad_norm": 1.0881357613531495,
+      "learning_rate": 1.9681727701107885e-05,
+      "loss": 0.9256,
+      "step": 563
+    },
+    {
+      "epoch": 0.10850327048864948,
+      "grad_norm": 1.0122254836689981,
+      "learning_rate": 1.9680166348095568e-05,
+      "loss": 0.9183,
+      "step": 564
+    },
+    {
+      "epoch": 0.10869565217391304,
+      "grad_norm": 0.9578885519718263,
+      "learning_rate": 1.967860123690937e-05,
+      "loss": 0.938,
+      "step": 565
+    },
+    {
+      "epoch": 0.1088880338591766,
+      "grad_norm": 0.9160376315170603,
+      "learning_rate": 1.9677032368156934e-05,
+      "loss": 0.8856,
+      "step": 566
+    },
+    {
+      "epoch": 0.10908041554444017,
+      "grad_norm": 0.9759082517508688,
+      "learning_rate": 1.967545974244734e-05,
+      "loss": 0.9106,
+      "step": 567
+    },
+    {
+      "epoch": 0.10927279722970373,
+      "grad_norm": 0.9811205148484734,
+      "learning_rate": 1.9673883360391138e-05,
+      "loss": 0.8102,
+      "step": 568
+    },
+    {
+      "epoch": 0.10946517891496729,
+      "grad_norm": 0.9432387605747896,
+      "learning_rate": 1.9672303222600333e-05,
+      "loss": 0.9572,
+      "step": 569
+    },
+    {
+      "epoch": 0.10965756060023085,
+      "grad_norm": 0.8103348128696671,
+      "learning_rate": 1.967071932968839e-05,
+      "loss": 0.322,
+      "step": 570
+    },
+    {
+      "epoch": 0.10984994228549443,
+      "grad_norm": 0.9875100305429135,
+      "learning_rate": 1.9669131682270232e-05,
+      "loss": 0.9376,
+      "step": 571
+    },
+    {
+      "epoch": 0.11004232397075799,
+      "grad_norm": 0.9055508800037539,
+      "learning_rate": 1.9667540280962235e-05,
+      "loss": 0.8232,
+      "step": 572
+    },
+    {
+      "epoch": 0.11023470565602154,
+      "grad_norm": 0.9801605081675344,
+      "learning_rate": 1.966594512638224e-05,
+      "loss": 0.8913,
+      "step": 573
+    },
+    {
+      "epoch": 0.1104270873412851,
+      "grad_norm": 0.984925486383107,
+      "learning_rate": 1.9664346219149538e-05,
+      "loss": 0.915,
+      "step": 574
+    },
+    {
+      "epoch": 0.11061946902654868,
+      "grad_norm": 0.9067054673115367,
+      "learning_rate": 1.966274355988488e-05,
+      "loss": 0.8863,
+      "step": 575
+    },
+    {
+      "epoch": 0.11081185071181224,
+      "grad_norm": 1.0371048901387794,
+      "learning_rate": 1.9661137149210473e-05,
+      "loss": 0.8932,
+      "step": 576
+    },
+    {
+      "epoch": 0.1110042323970758,
+      "grad_norm": 0.9432108342931145,
+      "learning_rate": 1.9659526987749987e-05,
+      "loss": 0.9292,
+      "step": 577
+    },
+    {
+      "epoch": 0.11119661408233936,
+      "grad_norm": 1.0272702017701911,
+      "learning_rate": 1.9657913076128532e-05,
+      "loss": 0.9107,
+      "step": 578
+    },
+    {
+      "epoch": 0.11138899576760293,
+      "grad_norm": 1.029985314915226,
+      "learning_rate": 1.965629541497269e-05,
+      "loss": 0.8797,
+      "step": 579
+    },
+    {
+      "epoch": 0.11158137745286649,
+      "grad_norm": 0.9458740203551704,
+      "learning_rate": 1.9654674004910493e-05,
+      "loss": 0.9167,
+      "step": 580
+    },
+    {
+      "epoch": 0.11177375913813005,
+      "grad_norm": 0.9719537371632296,
+      "learning_rate": 1.9653048846571427e-05,
+      "loss": 0.8915,
+      "step": 581
+    },
+    {
+      "epoch": 0.11196614082339361,
+      "grad_norm": 0.9545348688764776,
+      "learning_rate": 1.9651419940586437e-05,
+      "loss": 0.972,
+      "step": 582
+    },
+    {
+      "epoch": 0.11215852250865718,
+      "grad_norm": 1.0261322355243618,
+      "learning_rate": 1.964978728758791e-05,
+      "loss": 0.8858,
+      "step": 583
+    },
+    {
+      "epoch": 0.11235090419392074,
+      "grad_norm": 0.9660389977242732,
+      "learning_rate": 1.9648150888209715e-05,
+      "loss": 0.9524,
+      "step": 584
+    },
+    {
+      "epoch": 0.1125432858791843,
+      "grad_norm": 0.9534041623809403,
+      "learning_rate": 1.9646510743087144e-05,
+      "loss": 0.9459,
+      "step": 585
+    },
+    {
+      "epoch": 0.11273566756444786,
+      "grad_norm": 1.028597114502671,
+      "learning_rate": 1.964486685285697e-05,
+      "loss": 0.8589,
+      "step": 586
+    },
+    {
+      "epoch": 0.11292804924971142,
+      "grad_norm": 0.8949374186737706,
+      "learning_rate": 1.9643219218157395e-05,
+      "loss": 0.8988,
+      "step": 587
+    },
+    {
+      "epoch": 0.113120430934975,
+      "grad_norm": 0.9358015489797656,
+      "learning_rate": 1.9641567839628092e-05,
+      "loss": 0.8348,
+      "step": 588
+    },
+    {
+      "epoch": 0.11331281262023855,
+      "grad_norm": 0.9698626093114524,
+      "learning_rate": 1.963991271791019e-05,
+      "loss": 0.8501,
+      "step": 589
+    },
+    {
+      "epoch": 0.11350519430550211,
+      "grad_norm": 0.9732337522834856,
+      "learning_rate": 1.9638253853646255e-05,
+      "loss": 0.8748,
+      "step": 590
+    },
+    {
+      "epoch": 0.11369757599076567,
+      "grad_norm": 1.10822042752238,
+      "learning_rate": 1.9636591247480323e-05,
+      "loss": 1.0013,
+      "step": 591
+    },
+    {
+      "epoch": 0.11388995767602925,
+      "grad_norm": 0.9630632492670476,
+      "learning_rate": 1.9634924900057867e-05,
+      "loss": 0.8621,
+      "step": 592
+    },
+    {
+      "epoch": 0.1140823393612928,
+      "grad_norm": 0.9473759925133423,
+      "learning_rate": 1.963325481202583e-05,
+      "loss": 0.8569,
+      "step": 593
+    },
+    {
+      "epoch": 0.11427472104655637,
+      "grad_norm": 1.007997756960106,
+      "learning_rate": 1.963158098403259e-05,
+      "loss": 0.8898,
+      "step": 594
+    },
+    {
+      "epoch": 0.11446710273181993,
+      "grad_norm": 0.6691724400260535,
+      "learning_rate": 1.9629903416727987e-05,
+      "loss": 0.3282,
+      "step": 595
+    },
+    {
+      "epoch": 0.1146594844170835,
+      "grad_norm": 1.0010028325482334,
+      "learning_rate": 1.962822211076331e-05,
+      "loss": 0.8864,
+      "step": 596
+    },
+    {
+      "epoch": 0.11485186610234706,
+      "grad_norm": 1.0060235687144319,
+      "learning_rate": 1.96265370667913e-05,
+      "loss": 0.8619,
+      "step": 597
+    },
+    {
+      "epoch": 0.11504424778761062,
+      "grad_norm": 0.9348811605276396,
+      "learning_rate": 1.9624848285466146e-05,
+      "loss": 0.8959,
+      "step": 598
+    },
+    {
+      "epoch": 0.11523662947287418,
+      "grad_norm": 0.9709767235457031,
+      "learning_rate": 1.9623155767443498e-05,
+      "loss": 0.9151,
+      "step": 599
+    },
+    {
+      "epoch": 0.11542901115813775,
+      "grad_norm": 0.9375073386947848,
+      "learning_rate": 1.9621459513380445e-05,
+      "loss": 0.9297,
+      "step": 600
+    },
+    {
+      "epoch": 0.11562139284340131,
+      "grad_norm": 1.1451297677463959,
+      "learning_rate": 1.9619759523935532e-05,
+      "loss": 0.8438,
+      "step": 601
+    },
+    {
+      "epoch": 0.11581377452866487,
+      "grad_norm": 0.9358074813740274,
+      "learning_rate": 1.9618055799768757e-05,
+      "loss": 0.919,
+      "step": 602
+    },
+    {
+      "epoch": 0.11600615621392843,
+      "grad_norm": 0.6419971340924502,
+      "learning_rate": 1.961634834154156e-05,
+      "loss": 0.3612,
+      "step": 603
+    },
+    {
+      "epoch": 0.116198537899192,
+      "grad_norm": 0.9690076244973352,
+      "learning_rate": 1.9614637149916834e-05,
+      "loss": 0.9557,
+      "step": 604
+    },
+    {
+      "epoch": 0.11639091958445556,
+      "grad_norm": 1.072745956254057,
+      "learning_rate": 1.9612922225558924e-05,
+      "loss": 0.8713,
+      "step": 605
+    },
+    {
+      "epoch": 0.11658330126971912,
+      "grad_norm": 1.037813178126627,
+      "learning_rate": 1.961120356913363e-05,
+      "loss": 0.9658,
+      "step": 606
+    },
+    {
+      "epoch": 0.11677568295498268,
+      "grad_norm": 1.0197975195389488,
+      "learning_rate": 1.960948118130818e-05,
+      "loss": 0.8992,
+      "step": 607
+    },
+    {
+      "epoch": 0.11696806464024626,
+      "grad_norm": 0.9674907372499592,
+      "learning_rate": 1.9607755062751273e-05,
+      "loss": 0.9188,
+      "step": 608
+    },
+    {
+      "epoch": 0.11716044632550981,
+      "grad_norm": 0.9170285289309613,
+      "learning_rate": 1.9606025214133046e-05,
+      "loss": 0.9116,
+      "step": 609
+    },
+    {
+      "epoch": 0.11735282801077337,
+      "grad_norm": 1.0040521594647078,
+      "learning_rate": 1.9604291636125084e-05,
+      "loss": 0.9287,
+      "step": 610
+    },
+    {
+      "epoch": 0.11754520969603693,
+      "grad_norm": 1.0213408343555783,
+      "learning_rate": 1.960255432940043e-05,
+      "loss": 0.8919,
+      "step": 611
+    },
+    {
+      "epoch": 0.1177375913813005,
+      "grad_norm": 0.6541338073962687,
+      "learning_rate": 1.9600813294633552e-05,
+      "loss": 0.3092,
+      "step": 612
+    },
+    {
+      "epoch": 0.11792997306656407,
+      "grad_norm": 0.9721725168123452,
+      "learning_rate": 1.9599068532500394e-05,
+      "loss": 0.8844,
+      "step": 613
+    },
+    {
+      "epoch": 0.11812235475182763,
+      "grad_norm": 1.045022737955378,
+      "learning_rate": 1.9597320043678322e-05,
+      "loss": 0.8925,
+      "step": 614
+    },
+    {
+      "epoch": 0.11831473643709119,
+      "grad_norm": 0.9682480163629889,
+      "learning_rate": 1.9595567828846166e-05,
+      "loss": 0.8533,
+      "step": 615
+    },
+    {
+      "epoch": 0.11850711812235475,
+      "grad_norm": 0.9554230852166382,
+      "learning_rate": 1.9593811888684192e-05,
+      "loss": 0.9236,
+      "step": 616
+    },
+    {
+      "epoch": 0.11869949980761832,
+      "grad_norm": 0.9574494357408831,
+      "learning_rate": 1.9592052223874115e-05,
+      "loss": 0.9155,
+      "step": 617
+    },
+    {
+      "epoch": 0.11889188149288188,
+      "grad_norm": 0.9174632523955739,
+      "learning_rate": 1.959028883509911e-05,
+      "loss": 0.9245,
+      "step": 618
+    },
+    {
+      "epoch": 0.11908426317814544,
+      "grad_norm": 0.9853375234699602,
+      "learning_rate": 1.9588521723043764e-05,
+      "loss": 0.876,
+      "step": 619
+    },
+    {
+      "epoch": 0.119276644863409,
+      "grad_norm": 0.9731958333017595,
+      "learning_rate": 1.958675088839415e-05,
+      "loss": 0.8598,
+      "step": 620
+    },
+    {
+      "epoch": 0.11946902654867257,
+      "grad_norm": 1.027388043715151,
+      "learning_rate": 1.9584976331837758e-05,
+      "loss": 0.8881,
+      "step": 621
+    },
+    {
+      "epoch": 0.11966140823393613,
+      "grad_norm": 1.00762689613545,
+      "learning_rate": 1.9583198054063535e-05,
+      "loss": 0.8792,
+      "step": 622
+    },
+    {
+      "epoch": 0.11985378991919969,
+      "grad_norm": 0.717658632736469,
+      "learning_rate": 1.9581416055761865e-05,
+      "loss": 0.3175,
+      "step": 623
+    },
+    {
+      "epoch": 0.12004617160446325,
+      "grad_norm": 0.9631605755647259,
+      "learning_rate": 1.9579630337624585e-05,
+      "loss": 0.9517,
+      "step": 624
+    },
+    {
+      "epoch": 0.12023855328972682,
+      "grad_norm": 0.9681214850408949,
+      "learning_rate": 1.9577840900344974e-05,
+      "loss": 0.9156,
+      "step": 625
+    },
+    {
+      "epoch": 0.12043093497499038,
+      "grad_norm": 0.6035878812918369,
+      "learning_rate": 1.9576047744617752e-05,
+      "loss": 0.3309,
+      "step": 626
+    },
+    {
+      "epoch": 0.12062331666025394,
+      "grad_norm": 1.0428663753721736,
+      "learning_rate": 1.957425087113908e-05,
+      "loss": 0.9554,
+      "step": 627
+    },
+    {
+      "epoch": 0.1208156983455175,
+      "grad_norm": 0.9230724932297097,
+      "learning_rate": 1.9572450280606568e-05,
+      "loss": 0.9249,
+      "step": 628
+    },
+    {
+      "epoch": 0.12100808003078108,
+      "grad_norm": 1.0249162955242153,
+      "learning_rate": 1.9570645973719273e-05,
+      "loss": 0.9543,
+      "step": 629
+    },
+    {
+      "epoch": 0.12120046171604464,
+      "grad_norm": 0.9737368855915428,
+      "learning_rate": 1.9568837951177677e-05,
+      "loss": 0.9503,
+      "step": 630
+    },
+    {
+      "epoch": 0.1213928434013082,
+      "grad_norm": 0.8326168524779124,
+      "learning_rate": 1.9567026213683728e-05,
+      "loss": 0.3528,
+      "step": 631
+    },
+    {
+      "epoch": 0.12158522508657175,
+      "grad_norm": 0.9281344064243303,
+      "learning_rate": 1.9565210761940798e-05,
+      "loss": 0.8867,
+      "step": 632
+    },
+    {
+      "epoch": 0.12177760677183531,
+      "grad_norm": 0.915111776144517,
+      "learning_rate": 1.956339159665371e-05,
+      "loss": 0.8751,
+      "step": 633
+    },
+    {
+      "epoch": 0.12196998845709889,
+      "grad_norm": 0.6241940227325187,
+      "learning_rate": 1.956156871852873e-05,
+      "loss": 0.3191,
+      "step": 634
+    },
+    {
+      "epoch": 0.12216237014236245,
+      "grad_norm": 0.8622688120842585,
+      "learning_rate": 1.9559742128273558e-05,
+      "loss": 0.9047,
+      "step": 635
+    },
+    {
+      "epoch": 0.12235475182762601,
+      "grad_norm": 1.0067374056559124,
+      "learning_rate": 1.9557911826597337e-05,
+      "loss": 0.9001,
+      "step": 636
+    },
+    {
+      "epoch": 0.12254713351288957,
+      "grad_norm": 1.002311869499229,
+      "learning_rate": 1.9556077814210662e-05,
+      "loss": 0.9673,
+      "step": 637
+    },
+    {
+      "epoch": 0.12273951519815314,
+      "grad_norm": 1.020740897780026,
+      "learning_rate": 1.955424009182555e-05,
+      "loss": 0.9119,
+      "step": 638
+    },
+    {
+      "epoch": 0.1229318968834167,
+      "grad_norm": 1.0118108301441238,
+      "learning_rate": 1.955239866015547e-05,
+      "loss": 0.9474,
+      "step": 639
+    },
+    {
+      "epoch": 0.12312427856868026,
+      "grad_norm": 0.9753816396900973,
+      "learning_rate": 1.9550553519915335e-05,
+      "loss": 0.8748,
+      "step": 640
+    },
+    {
+      "epoch": 0.12331666025394382,
+      "grad_norm": 0.9065063965927997,
+      "learning_rate": 1.954870467182149e-05,
+      "loss": 0.9518,
+      "step": 641
+    },
+    {
+      "epoch": 0.12350904193920739,
+      "grad_norm": 0.9632028759528475,
+      "learning_rate": 1.954685211659172e-05,
+      "loss": 0.8505,
+      "step": 642
+    },
+    {
+      "epoch": 0.12370142362447095,
+      "grad_norm": 0.9910673714538408,
+      "learning_rate": 1.9544995854945248e-05,
+      "loss": 0.8564,
+      "step": 643
+    },
+    {
+      "epoch": 0.12389380530973451,
+      "grad_norm": 0.9732031307990795,
+      "learning_rate": 1.954313588760274e-05,
+      "loss": 0.9095,
+      "step": 644
+    },
+    {
+      "epoch": 0.12408618699499807,
+      "grad_norm": 0.9349587123015122,
+      "learning_rate": 1.9541272215286304e-05,
+      "loss": 0.8761,
+      "step": 645
+    },
+    {
+      "epoch": 0.12427856868026164,
+      "grad_norm": 0.9105944631455443,
+      "learning_rate": 1.9539404838719477e-05,
+      "loss": 0.9011,
+      "step": 646
+    },
+    {
+      "epoch": 0.1244709503655252,
+      "grad_norm": 1.0563287596989304,
+      "learning_rate": 1.9537533758627242e-05,
+      "loss": 0.8528,
+      "step": 647
+    },
+    {
+      "epoch": 0.12466333205078876,
+      "grad_norm": 0.9783377377858462,
+      "learning_rate": 1.953565897573601e-05,
+      "loss": 0.8855,
+      "step": 648
+    },
+    {
+      "epoch": 0.12485571373605232,
+      "grad_norm": 1.0471470781673504,
+      "learning_rate": 1.9533780490773645e-05,
+      "loss": 0.8663,
+      "step": 649
+    },
+    {
+      "epoch": 0.1250480954213159,
+      "grad_norm": 0.8393262041578345,
+      "learning_rate": 1.9531898304469435e-05,
+      "loss": 0.3057,
+      "step": 650
+    },
+    {
+      "epoch": 0.12524047710657946,
+      "grad_norm": 0.9747984420391828,
+      "learning_rate": 1.953001241755411e-05,
+      "loss": 0.8267,
+      "step": 651
+    },
+    {
+      "epoch": 0.12543285879184302,
+      "grad_norm": 0.9172322269189965,
+      "learning_rate": 1.952812283075984e-05,
+      "loss": 0.9049,
+      "step": 652
+    },
+    {
+      "epoch": 0.12562524047710658,
+      "grad_norm": 0.8801191070000007,
+      "learning_rate": 1.952622954482022e-05,
+      "loss": 0.8833,
+      "step": 653
+    },
+    {
+      "epoch": 0.12581762216237014,
+      "grad_norm": 1.028916763638007,
+      "learning_rate": 1.9524332560470293e-05,
+      "loss": 0.8875,
+      "step": 654
+    },
+    {
+      "epoch": 0.1260100038476337,
+      "grad_norm": 1.049693086157148,
+      "learning_rate": 1.9522431878446536e-05,
+      "loss": 0.8789,
+      "step": 655
+    },
+    {
+      "epoch": 0.12620238553289725,
+      "grad_norm": 1.000630245241351,
+      "learning_rate": 1.9520527499486856e-05,
+      "loss": 0.8954,
+      "step": 656
+    },
+    {
+      "epoch": 0.12639476721816084,
+      "grad_norm": 0.9627931891495659,
+      "learning_rate": 1.95186194243306e-05,
+      "loss": 0.9275,
+      "step": 657
+    },
+    {
+      "epoch": 0.1265871489034244,
+      "grad_norm": 0.9790674534093812,
+      "learning_rate": 1.9516707653718546e-05,
+      "loss": 0.8935,
+      "step": 658
+    },
+    {
+      "epoch": 0.12677953058868796,
+      "grad_norm": 0.8771775266521364,
+      "learning_rate": 1.9514792188392914e-05,
+      "loss": 0.8271,
+      "step": 659
+    },
+    {
+      "epoch": 0.12697191227395152,
+      "grad_norm": 0.8909375238931709,
+      "learning_rate": 1.9512873029097347e-05,
+      "loss": 0.9036,
+      "step": 660
+    },
+    {
+      "epoch": 0.12716429395921508,
+      "grad_norm": 0.9224191761032633,
+      "learning_rate": 1.9510950176576933e-05,
+      "loss": 0.8949,
+      "step": 661
+    },
+    {
+      "epoch": 0.12735667564447864,
+      "grad_norm": 1.10754412476127,
+      "learning_rate": 1.950902363157819e-05,
+      "loss": 0.8799,
+      "step": 662
+    },
+    {
+      "epoch": 0.1275490573297422,
+      "grad_norm": 0.930027265289809,
+      "learning_rate": 1.950709339484907e-05,
+      "loss": 0.9058,
+      "step": 663
+    },
+    {
+      "epoch": 0.12774143901500576,
+      "grad_norm": 1.0544104048265828,
+      "learning_rate": 1.9505159467138954e-05,
+      "loss": 0.3276,
+      "step": 664
+    },
+    {
+      "epoch": 0.12793382070026935,
+      "grad_norm": 1.0409203926357495,
+      "learning_rate": 1.9503221849198655e-05,
+      "loss": 0.9046,
+      "step": 665
+    },
+    {
+      "epoch": 0.1281262023855329,
+      "grad_norm": 1.054848498734012,
+      "learning_rate": 1.9501280541780435e-05,
+      "loss": 0.9669,
+      "step": 666
+    },
+    {
+      "epoch": 0.12831858407079647,
+      "grad_norm": 0.6437399492856508,
+      "learning_rate": 1.9499335545637968e-05,
+      "loss": 0.3373,
+      "step": 667
+    },
+    {
+      "epoch": 0.12851096575606002,
+      "grad_norm": 1.1125825831189053,
+      "learning_rate": 1.949738686152637e-05,
+      "loss": 0.9603,
+      "step": 668
+    },
+    {
+      "epoch": 0.12870334744132358,
+      "grad_norm": 0.9042722637800664,
+      "learning_rate": 1.9495434490202188e-05,
+      "loss": 0.8438,
+      "step": 669
+    },
+    {
+      "epoch": 0.12889572912658714,
+      "grad_norm": 0.9354352788067852,
+      "learning_rate": 1.94934784324234e-05,
+      "loss": 0.9258,
+      "step": 670
+    },
+    {
+      "epoch": 0.1290881108118507,
+      "grad_norm": 0.9679463584196953,
+      "learning_rate": 1.9491518688949417e-05,
+      "loss": 0.9102,
+      "step": 671
+    },
+    {
+      "epoch": 0.12928049249711426,
+      "grad_norm": 1.030011308516237,
+      "learning_rate": 1.9489555260541074e-05,
+      "loss": 0.9274,
+      "step": 672
+    },
+    {
+      "epoch": 0.12947287418237785,
+      "grad_norm": 0.8844688120788922,
+      "learning_rate": 1.948758814796064e-05,
+      "loss": 0.8337,
+      "step": 673
+    },
+    {
+      "epoch": 0.1296652558676414,
+      "grad_norm": 0.9707626378335048,
+      "learning_rate": 1.9485617351971827e-05,
+      "loss": 0.8965,
+      "step": 674
+    },
+    {
+      "epoch": 0.12985763755290497,
+      "grad_norm": 1.0797134325884754,
+      "learning_rate": 1.9483642873339753e-05,
+      "loss": 0.9462,
+      "step": 675
+    },
+    {
+      "epoch": 0.13005001923816853,
+      "grad_norm": 1.0127636508716023,
+      "learning_rate": 1.9481664712830987e-05,
+      "loss": 0.8825,
+      "step": 676
+    },
+    {
+      "epoch": 0.1302424009234321,
+      "grad_norm": 0.9472046438353805,
+      "learning_rate": 1.9479682871213515e-05,
+      "loss": 0.9177,
+      "step": 677
+    },
+    {
+      "epoch": 0.13043478260869565,
+      "grad_norm": 0.9960976109018944,
+      "learning_rate": 1.9477697349256756e-05,
+      "loss": 0.9199,
+      "step": 678
+    },
+    {
+      "epoch": 0.1306271642939592,
+      "grad_norm": 0.9135331153574087,
+      "learning_rate": 1.947570814773156e-05,
+      "loss": 0.8895,
+      "step": 679
+    },
+    {
+      "epoch": 0.13081954597922277,
+      "grad_norm": 0.9210490878541144,
+      "learning_rate": 1.9473715267410206e-05,
+      "loss": 0.9466,
+      "step": 680
+    },
+    {
+      "epoch": 0.13101192766448633,
+      "grad_norm": 1.1218877660011508,
+      "learning_rate": 1.9471718709066392e-05,
+      "loss": 0.9361,
+      "step": 681
+    },
+    {
+      "epoch": 0.13120430934974991,
+      "grad_norm": 0.9671023137777117,
+      "learning_rate": 1.9469718473475256e-05,
+      "loss": 0.9331,
+      "step": 682
+    },
+    {
+      "epoch": 0.13139669103501347,
+      "grad_norm": 0.9071137487633956,
+      "learning_rate": 1.9467714561413358e-05,
+      "loss": 0.8341,
+      "step": 683
+    },
+    {
+      "epoch": 0.13158907272027703,
+      "grad_norm": 1.0249635132184443,
+      "learning_rate": 1.9465706973658683e-05,
+      "loss": 0.3506,
+      "step": 684
+    },
+    {
+      "epoch": 0.1317814544055406,
+      "grad_norm": 0.9098785074084939,
+      "learning_rate": 1.9463695710990648e-05,
+      "loss": 0.8369,
+      "step": 685
+    },
+    {
+      "epoch": 0.13197383609080415,
+      "grad_norm": 1.0615321550838632,
+      "learning_rate": 1.946168077419009e-05,
+      "loss": 0.9327,
+      "step": 686
+    },
+    {
+      "epoch": 0.1321662177760677,
+      "grad_norm": 0.8681018414328638,
+      "learning_rate": 1.9459662164039283e-05,
+      "loss": 0.8772,
+      "step": 687
+    },
+    {
+      "epoch": 0.13235859946133127,
+      "grad_norm": 1.0033009340143335,
+      "learning_rate": 1.9457639881321917e-05,
+      "loss": 0.8529,
+      "step": 688
+    },
+    {
+      "epoch": 0.13255098114659483,
+      "grad_norm": 0.9149389825522484,
+      "learning_rate": 1.9455613926823115e-05,
+      "loss": 0.9054,
+      "step": 689
+    },
+    {
+      "epoch": 0.13274336283185842,
+      "grad_norm": 1.1325897146835397,
+      "learning_rate": 1.945358430132942e-05,
+      "loss": 1.0107,
+      "step": 690
+    },
+    {
+      "epoch": 0.13293574451712198,
+      "grad_norm": 0.9448855716559719,
+      "learning_rate": 1.9451551005628803e-05,
+      "loss": 0.8777,
+      "step": 691
+    },
+    {
+      "epoch": 0.13312812620238554,
+      "grad_norm": 0.9896630831431741,
+      "learning_rate": 1.9449514040510654e-05,
+      "loss": 0.8693,
+      "step": 692
+    },
+    {
+      "epoch": 0.1333205078876491,
+      "grad_norm": 0.8595450485233581,
+      "learning_rate": 1.9447473406765803e-05,
+      "loss": 0.8433,
+      "step": 693
+    },
+    {
+      "epoch": 0.13351288957291266,
+      "grad_norm": 0.9875151414393375,
+      "learning_rate": 1.9445429105186487e-05,
+      "loss": 0.9445,
+      "step": 694
+    },
+    {
+      "epoch": 0.13370527125817622,
+      "grad_norm": 0.9254642962692159,
+      "learning_rate": 1.9443381136566382e-05,
+      "loss": 0.885,
+      "step": 695
+    },
+    {
+      "epoch": 0.13389765294343978,
+      "grad_norm": 1.0319633517618163,
+      "learning_rate": 1.9441329501700568e-05,
+      "loss": 0.9461,
+      "step": 696
+    },
+    {
+      "epoch": 0.13409003462870334,
+      "grad_norm": 0.931528497660767,
+      "learning_rate": 1.943927420138557e-05,
+      "loss": 0.822,
+      "step": 697
+    },
+    {
+      "epoch": 0.13428241631396692,
+      "grad_norm": 0.906245920285603,
+      "learning_rate": 1.9437215236419322e-05,
+      "loss": 0.9009,
+      "step": 698
+    },
+    {
+      "epoch": 0.13447479799923048,
+      "grad_norm": 0.940370895420759,
+      "learning_rate": 1.9435152607601187e-05,
+      "loss": 0.8459,
+      "step": 699
+    },
+    {
+      "epoch": 0.13466717968449404,
+      "grad_norm": 0.8966574166154715,
+      "learning_rate": 1.943308631573195e-05,
+      "loss": 0.8272,
+      "step": 700
+    },
+    {
+      "epoch": 0.1348595613697576,
+      "grad_norm": 0.9344977614607324,
+      "learning_rate": 1.9431016361613816e-05,
+      "loss": 0.8667,
+      "step": 701
+    },
+    {
+      "epoch": 0.13505194305502116,
+      "grad_norm": 0.9916374079585701,
+      "learning_rate": 1.9428942746050406e-05,
+      "loss": 0.9064,
+      "step": 702
+    },
+    {
+      "epoch": 0.13524432474028472,
+      "grad_norm": 0.9778810702653192,
+      "learning_rate": 1.9426865469846773e-05,
+      "loss": 0.8794,
+      "step": 703
+    },
+    {
+      "epoch": 0.13543670642554828,
+      "grad_norm": 0.9221277465158829,
+      "learning_rate": 1.9424784533809393e-05,
+      "loss": 0.87,
+      "step": 704
+    },
+    {
+      "epoch": 0.13562908811081184,
+      "grad_norm": 0.9717432541907935,
+      "learning_rate": 1.942269993874615e-05,
+      "loss": 0.8091,
+      "step": 705
+    },
+    {
+      "epoch": 0.1358214697960754,
+      "grad_norm": 0.9791760987609283,
+      "learning_rate": 1.9420611685466358e-05,
+      "loss": 0.906,
+      "step": 706
+    },
+    {
+      "epoch": 0.136013851481339,
+      "grad_norm": 1.0357425661473745,
+      "learning_rate": 1.9418519774780748e-05,
+      "loss": 0.9063,
+      "step": 707
+    },
+    {
+      "epoch": 0.13620623316660255,
+      "grad_norm": 0.8600139479706462,
+      "learning_rate": 1.9416424207501474e-05,
+      "loss": 0.8775,
+      "step": 708
+    },
+    {
+      "epoch": 0.1363986148518661,
+      "grad_norm": 1.1133290885589908,
+      "learning_rate": 1.9414324984442102e-05,
+      "loss": 0.9158,
+      "step": 709
+    },
+    {
+      "epoch": 0.13659099653712967,
+      "grad_norm": 0.8886043668644177,
+      "learning_rate": 1.9412222106417632e-05,
+      "loss": 0.8591,
+      "step": 710
+    },
+    {
+      "epoch": 0.13678337822239323,
+      "grad_norm": 0.950913334371217,
+      "learning_rate": 1.9410115574244462e-05,
+      "loss": 0.922,
+      "step": 711
+    },
+    {
+      "epoch": 0.13697575990765679,
+      "grad_norm": 0.9652425696572761,
+      "learning_rate": 1.9408005388740433e-05,
+      "loss": 0.865,
+      "step": 712
+    },
+    {
+      "epoch": 0.13716814159292035,
+      "grad_norm": 0.9917784900560727,
+      "learning_rate": 1.9405891550724778e-05,
+      "loss": 0.863,
+      "step": 713
+    },
+    {
+      "epoch": 0.1373605232781839,
+      "grad_norm": 1.060069029550011,
+      "learning_rate": 1.940377406101817e-05,
+      "loss": 0.9253,
+      "step": 714
+    },
+    {
+      "epoch": 0.1375529049634475,
+      "grad_norm": 1.0330901292095096,
+      "learning_rate": 1.9401652920442694e-05,
+      "loss": 0.9397,
+      "step": 715
+    },
+    {
+      "epoch": 0.13774528664871105,
+      "grad_norm": 0.9391788584827754,
+      "learning_rate": 1.9399528129821842e-05,
+      "loss": 0.8606,
+      "step": 716
+    },
+    {
+      "epoch": 0.1379376683339746,
+      "grad_norm": 0.8936889674568561,
+      "learning_rate": 1.939739968998054e-05,
+      "loss": 0.9082,
+      "step": 717
+    },
+    {
+      "epoch": 0.13813005001923817,
+      "grad_norm": 0.9973321497490899,
+      "learning_rate": 1.939526760174511e-05,
+      "loss": 0.9235,
+      "step": 718
+    },
+    {
+      "epoch": 0.13832243170450173,
+      "grad_norm": 0.9579960015539128,
+      "learning_rate": 1.939313186594331e-05,
+      "loss": 0.9246,
+      "step": 719
+    },
+    {
+      "epoch": 0.1385148133897653,
+      "grad_norm": 0.9811050603499806,
+      "learning_rate": 1.9390992483404308e-05,
+      "loss": 0.8752,
+      "step": 720
+    },
+    {
+      "epoch": 0.13870719507502885,
+      "grad_norm": 0.9620242582489998,
+      "learning_rate": 1.938884945495868e-05,
+      "loss": 0.9032,
+      "step": 721
+    },
+    {
+      "epoch": 0.1388995767602924,
+      "grad_norm": 0.9169567905058011,
+      "learning_rate": 1.9386702781438425e-05,
+      "loss": 0.9318,
+      "step": 722
+    },
+    {
+      "epoch": 0.139091958445556,
+      "grad_norm": 0.9517347558803031,
+      "learning_rate": 1.938455246367696e-05,
+      "loss": 0.9316,
+      "step": 723
+    },
+    {
+      "epoch": 0.13928434013081956,
+      "grad_norm": 0.934801192444199,
+      "learning_rate": 1.9382398502509107e-05,
+      "loss": 0.8647,
+      "step": 724
+    },
+    {
+      "epoch": 0.13947672181608312,
+      "grad_norm": 0.9866249227210778,
+      "learning_rate": 1.938024089877111e-05,
+      "loss": 0.9202,
+      "step": 725
+    },
+    {
+      "epoch": 0.13966910350134668,
+      "grad_norm": 0.94797108978451,
+      "learning_rate": 1.9378079653300624e-05,
+      "loss": 0.3516,
+      "step": 726
+    },
+    {
+      "epoch": 0.13986148518661023,
+      "grad_norm": 1.1227077124931748,
+      "learning_rate": 1.9375914766936723e-05,
+      "loss": 0.8777,
+      "step": 727
+    },
+    {
+      "epoch": 0.1400538668718738,
+      "grad_norm": 0.9014411585821797,
+      "learning_rate": 1.9373746240519884e-05,
+      "loss": 0.8728,
+      "step": 728
+    },
+    {
+      "epoch": 0.14024624855713735,
+      "grad_norm": 1.00664797372286,
+      "learning_rate": 1.937157407489201e-05,
+      "loss": 0.875,
+      "step": 729
+    },
+    {
+      "epoch": 0.1404386302424009,
+      "grad_norm": 0.9628517851295814,
+      "learning_rate": 1.9369398270896403e-05,
+      "loss": 0.95,
+      "step": 730
+    },
+    {
+      "epoch": 0.14063101192766447,
+      "grad_norm": 0.910828906595872,
+      "learning_rate": 1.936721882937779e-05,
+      "loss": 0.8722,
+      "step": 731
+    },
+    {
+      "epoch": 0.14082339361292806,
+      "grad_norm": 0.9629417197380367,
+      "learning_rate": 1.9365035751182307e-05,
+      "loss": 0.8839,
+      "step": 732
+    },
+    {
+      "epoch": 0.14101577529819162,
+      "grad_norm": 0.9634989868305142,
+      "learning_rate": 1.93628490371575e-05,
+      "loss": 0.92,
+      "step": 733
+    },
+    {
+      "epoch": 0.14120815698345518,
+      "grad_norm": 0.9391120257877041,
+      "learning_rate": 1.9360658688152322e-05,
+      "loss": 0.8322,
+      "step": 734
+    },
+    {
+      "epoch": 0.14140053866871874,
+      "grad_norm": 0.9192421570881326,
+      "learning_rate": 1.9358464705017143e-05,
+      "loss": 0.8228,
+      "step": 735
+    },
+    {
+      "epoch": 0.1415929203539823,
+      "grad_norm": 0.9065922953968022,
+      "learning_rate": 1.9356267088603745e-05,
+      "loss": 0.9601,
+      "step": 736
+    },
+    {
+      "epoch": 0.14178530203924586,
+      "grad_norm": 1.0402042157162363,
+      "learning_rate": 1.9354065839765316e-05,
+      "loss": 0.9229,
+      "step": 737
+    },
+    {
+      "epoch": 0.14197768372450942,
+      "grad_norm": 0.9226211159542477,
+      "learning_rate": 1.9351860959356462e-05,
+      "loss": 0.8759,
+      "step": 738
+    },
+    {
+      "epoch": 0.14217006540977298,
+      "grad_norm": 0.9103191423978265,
+      "learning_rate": 1.9349652448233187e-05,
+      "loss": 0.8894,
+      "step": 739
+    },
+    {
+      "epoch": 0.14236244709503657,
+      "grad_norm": 1.0393756286470484,
+      "learning_rate": 1.934744030725291e-05,
+      "loss": 0.871,
+      "step": 740
+    },
+    {
+      "epoch": 0.14255482878030012,
+      "grad_norm": 1.0951738041099146,
+      "learning_rate": 1.934522453727447e-05,
+      "loss": 0.8472,
+      "step": 741
+    },
+    {
+      "epoch": 0.14274721046556368,
+      "grad_norm": 1.0054527692388375,
+      "learning_rate": 1.93430051391581e-05,
+      "loss": 0.9344,
+      "step": 742
+    },
+    {
+      "epoch": 0.14293959215082724,
+      "grad_norm": 0.9537151820422862,
+      "learning_rate": 1.934078211376544e-05,
+      "loss": 0.8933,
+      "step": 743
+    },
+    {
+      "epoch": 0.1431319738360908,
+      "grad_norm": 0.9584600998882173,
+      "learning_rate": 1.9338555461959554e-05,
+      "loss": 0.8525,
+      "step": 744
+    },
+    {
+      "epoch": 0.14332435552135436,
+      "grad_norm": 0.978501390151673,
+      "learning_rate": 1.93363251846049e-05,
+      "loss": 0.8883,
+      "step": 745
+    },
+    {
+      "epoch": 0.14351673720661792,
+      "grad_norm": 1.0711335026664233,
+      "learning_rate": 1.9334091282567352e-05,
+      "loss": 0.9351,
+      "step": 746
+    },
+    {
+      "epoch": 0.14370911889188148,
+      "grad_norm": 0.9712766571940528,
+      "learning_rate": 1.9331853756714185e-05,
+      "loss": 0.9264,
+      "step": 747
+    },
+    {
+      "epoch": 0.14390150057714507,
+      "grad_norm": 1.0646329595900894,
+      "learning_rate": 1.9329612607914088e-05,
+      "loss": 0.8896,
+      "step": 748
+    },
+    {
+      "epoch": 0.14409388226240863,
+      "grad_norm": 0.9790581643941968,
+      "learning_rate": 1.9327367837037142e-05,
+      "loss": 0.8784,
+      "step": 749
+    },
+    {
+      "epoch": 0.1442862639476722,
+      "grad_norm": 0.816414644382769,
+      "learning_rate": 1.9325119444954855e-05,
+      "loss": 0.3475,
+      "step": 750
+    },
+    {
+      "epoch": 0.14447864563293575,
+      "grad_norm": 0.9681040984661303,
+      "learning_rate": 1.9322867432540126e-05,
+      "loss": 0.906,
+      "step": 751
+    },
+    {
+      "epoch": 0.1446710273181993,
+      "grad_norm": 0.9782215178020337,
+      "learning_rate": 1.9320611800667268e-05,
+      "loss": 0.8644,
+      "step": 752
+    },
+    {
+      "epoch": 0.14486340900346287,
+      "grad_norm": 0.9618186516156919,
+      "learning_rate": 1.9318352550211986e-05,
+      "loss": 0.8728,
+      "step": 753
+    },
+    {
+      "epoch": 0.14505579068872643,
+      "grad_norm": 0.9831372312081971,
+      "learning_rate": 1.9316089682051403e-05,
+      "loss": 0.8227,
+      "step": 754
+    },
+    {
+      "epoch": 0.14524817237399,
+      "grad_norm": 0.9301595723698287,
+      "learning_rate": 1.9313823197064042e-05,
+      "loss": 0.9273,
+      "step": 755
+    },
+    {
+      "epoch": 0.14544055405925355,
+      "grad_norm": 0.9687453941310135,
+      "learning_rate": 1.9311553096129835e-05,
+      "loss": 0.8923,
+      "step": 756
+    },
+    {
+      "epoch": 0.14563293574451713,
+      "grad_norm": 0.9886517438877535,
+      "learning_rate": 1.9309279380130112e-05,
+      "loss": 0.9101,
+      "step": 757
+    },
+    {
+      "epoch": 0.1458253174297807,
+      "grad_norm": 0.9818264193762843,
+      "learning_rate": 1.93070020499476e-05,
+      "loss": 0.8303,
+      "step": 758
+    },
+    {
+      "epoch": 0.14601769911504425,
+      "grad_norm": 0.6216298966350329,
+      "learning_rate": 1.930472110646645e-05,
+      "loss": 0.299,
+      "step": 759
+    },
+    {
+      "epoch": 0.1462100808003078,
+      "grad_norm": 0.9391200221739503,
+      "learning_rate": 1.9302436550572187e-05,
+      "loss": 0.8485,
+      "step": 760
+    },
+    {
+      "epoch": 0.14640246248557137,
+      "grad_norm": 0.9561502769624679,
+      "learning_rate": 1.930014838315177e-05,
+      "loss": 0.8856,
+      "step": 761
+    },
+    {
+      "epoch": 0.14659484417083493,
+      "grad_norm": 0.9733622677640728,
+      "learning_rate": 1.9297856605093534e-05,
+      "loss": 0.8707,
+      "step": 762
+    },
+    {
+      "epoch": 0.1467872258560985,
+      "grad_norm": 0.9974540710702196,
+      "learning_rate": 1.9295561217287226e-05,
+      "loss": 0.8973,
+      "step": 763
+    },
+    {
+      "epoch": 0.14697960754136205,
+      "grad_norm": 0.9158304912853799,
+      "learning_rate": 1.9293262220624002e-05,
+      "loss": 0.9239,
+      "step": 764
+    },
+    {
+      "epoch": 0.14717198922662564,
+      "grad_norm": 0.9475155536277865,
+      "learning_rate": 1.9290959615996407e-05,
+      "loss": 0.933,
+      "step": 765
+    },
+    {
+      "epoch": 0.1473643709118892,
+      "grad_norm": 1.0059766474750447,
+      "learning_rate": 1.9288653404298392e-05,
+      "loss": 0.8568,
+      "step": 766
+    },
+    {
+      "epoch": 0.14755675259715276,
+      "grad_norm": 1.064673546091075,
+      "learning_rate": 1.9286343586425307e-05,
+      "loss": 0.9027,
+      "step": 767
+    },
+    {
+      "epoch": 0.14774913428241632,
+      "grad_norm": 0.9423152186344447,
+      "learning_rate": 1.9284030163273907e-05,
+      "loss": 0.8759,
+      "step": 768
+    },
+    {
+      "epoch": 0.14794151596767988,
+      "grad_norm": 1.0004832833890815,
+      "learning_rate": 1.9281713135742333e-05,
+      "loss": 0.8984,
+      "step": 769
+    },
+    {
+      "epoch": 0.14813389765294344,
+      "grad_norm": 0.95347687452844,
+      "learning_rate": 1.9279392504730147e-05,
+      "loss": 0.8467,
+      "step": 770
+    },
+    {
+      "epoch": 0.148326279338207,
+      "grad_norm": 0.9310408814797889,
+      "learning_rate": 1.9277068271138287e-05,
+      "loss": 0.9207,
+      "step": 771
+    },
+    {
+      "epoch": 0.14851866102347056,
+      "grad_norm": 0.9292626864463759,
+      "learning_rate": 1.9274740435869107e-05,
+      "loss": 0.8573,
+      "step": 772
+    },
+    {
+      "epoch": 0.14871104270873411,
+      "grad_norm": 0.9043224016580271,
+      "learning_rate": 1.927240899982635e-05,
+      "loss": 0.9044,
+      "step": 773
+    },
+    {
+      "epoch": 0.1489034243939977,
+      "grad_norm": 0.9794688537790924,
+      "learning_rate": 1.9270073963915162e-05,
+      "loss": 0.9252,
+      "step": 774
+    },
+    {
+      "epoch": 0.14909580607926126,
+      "grad_norm": 0.9578064968420044,
+      "learning_rate": 1.9267735329042086e-05,
+      "loss": 0.8721,
+      "step": 775
+    },
+    {
+      "epoch": 0.14928818776452482,
+      "grad_norm": 0.9848679941409691,
+      "learning_rate": 1.9265393096115056e-05,
+      "loss": 0.8599,
+      "step": 776
+    },
+    {
+      "epoch": 0.14948056944978838,
+      "grad_norm": 0.9534821393023923,
+      "learning_rate": 1.926304726604341e-05,
+      "loss": 0.9141,
+      "step": 777
+    },
+    {
+      "epoch": 0.14967295113505194,
+      "grad_norm": 1.00816105875212,
+      "learning_rate": 1.9260697839737875e-05,
+      "loss": 0.8099,
+      "step": 778
+    },
+    {
+      "epoch": 0.1498653328203155,
+      "grad_norm": 0.9058642831242009,
+      "learning_rate": 1.925834481811059e-05,
+      "loss": 0.8924,
+      "step": 779
+    },
+    {
+      "epoch": 0.15005771450557906,
+      "grad_norm": 0.9972750976955485,
+      "learning_rate": 1.9255988202075065e-05,
+      "loss": 0.9084,
+      "step": 780
+    },
+    {
+      "epoch": 0.15025009619084262,
+      "grad_norm": 0.9705258745060447,
+      "learning_rate": 1.925362799254623e-05,
+      "loss": 0.9231,
+      "step": 781
+    },
+    {
+      "epoch": 0.1504424778761062,
+      "grad_norm": 0.9880740640191156,
+      "learning_rate": 1.9251264190440398e-05,
+      "loss": 0.8877,
+      "step": 782
+    },
+    {
+      "epoch": 0.15063485956136977,
+      "grad_norm": 0.9027493197165121,
+      "learning_rate": 1.9248896796675277e-05,
+      "loss": 0.8919,
+      "step": 783
+    },
+    {
+      "epoch": 0.15082724124663333,
+      "grad_norm": 0.9411001466917366,
+      "learning_rate": 1.924652581216997e-05,
+      "loss": 0.9279,
+      "step": 784
+    },
+    {
+      "epoch": 0.15101962293189689,
+      "grad_norm": 0.9814145074959807,
+      "learning_rate": 1.9244151237844975e-05,
+      "loss": 0.8936,
+      "step": 785
+    },
+    {
+      "epoch": 0.15121200461716044,
+      "grad_norm": 0.996749342919976,
+      "learning_rate": 1.9241773074622182e-05,
+      "loss": 0.8587,
+      "step": 786
+    },
+    {
+      "epoch": 0.151404386302424,
+      "grad_norm": 1.0324504675189428,
+      "learning_rate": 1.923939132342488e-05,
+      "loss": 0.9326,
+      "step": 787
+    },
+    {
+      "epoch": 0.15159676798768756,
+      "grad_norm": 0.9757720363056218,
+      "learning_rate": 1.923700598517775e-05,
+      "loss": 0.8994,
+      "step": 788
+    },
+    {
+      "epoch": 0.15178914967295112,
+      "grad_norm": 0.9264252372281142,
+      "learning_rate": 1.923461706080685e-05,
+      "loss": 0.9053,
+      "step": 789
+    },
+    {
+      "epoch": 0.1519815313582147,
+      "grad_norm": 0.8340288030235871,
+      "learning_rate": 1.923222455123965e-05,
+      "loss": 0.8513,
+      "step": 790
+    },
+    {
+      "epoch": 0.15217391304347827,
+      "grad_norm": 0.9426432536572661,
+      "learning_rate": 1.9229828457405005e-05,
+      "loss": 0.9253,
+      "step": 791
+    },
+    {
+      "epoch": 0.15236629472874183,
+      "grad_norm": 0.9676350363944066,
+      "learning_rate": 1.9227428780233162e-05,
+      "loss": 0.8769,
+      "step": 792
+    },
+    {
+      "epoch": 0.1525586764140054,
+      "grad_norm": 0.9844075160262186,
+      "learning_rate": 1.922502552065576e-05,
+      "loss": 0.8645,
+      "step": 793
+    },
+    {
+      "epoch": 0.15275105809926895,
+      "grad_norm": 0.9934459099331502,
+      "learning_rate": 1.922261867960582e-05,
+      "loss": 0.8754,
+      "step": 794
+    },
+    {
+      "epoch": 0.1529434397845325,
+      "grad_norm": 1.0024273419368994,
+      "learning_rate": 1.9220208258017763e-05,
+      "loss": 0.9032,
+      "step": 795
+    },
+    {
+      "epoch": 0.15313582146979607,
+      "grad_norm": 0.9051321407034704,
+      "learning_rate": 1.92177942568274e-05,
+      "loss": 0.8886,
+      "step": 796
+    },
+    {
+      "epoch": 0.15332820315505963,
+      "grad_norm": 0.9087738571521314,
+      "learning_rate": 1.921537667697193e-05,
+      "loss": 0.9072,
+      "step": 797
+    },
+    {
+      "epoch": 0.1535205848403232,
+      "grad_norm": 0.9760458793784145,
+      "learning_rate": 1.9212955519389938e-05,
+      "loss": 0.9023,
+      "step": 798
+    },
+    {
+      "epoch": 0.15371296652558677,
+      "grad_norm": 0.9246169603551622,
+      "learning_rate": 1.9210530785021405e-05,
+      "loss": 0.9168,
+      "step": 799
+    },
+    {
+      "epoch": 0.15390534821085033,
+      "grad_norm": 0.9738486535853831,
+      "learning_rate": 1.9208102474807692e-05,
+      "loss": 0.8737,
+      "step": 800
+    },
+    {
+      "epoch": 0.1540977298961139,
+      "grad_norm": 0.9978166455631136,
+      "learning_rate": 1.920567058969155e-05,
+      "loss": 0.9361,
+      "step": 801
+    },
+    {
+      "epoch": 0.15429011158137745,
+      "grad_norm": 0.9388857842010891,
+      "learning_rate": 1.920323513061713e-05,
+      "loss": 0.8788,
+      "step": 802
+    },
+    {
+      "epoch": 0.154482493266641,
+      "grad_norm": 0.9689678543729954,
+      "learning_rate": 1.9200796098529956e-05,
+      "loss": 0.9172,
+      "step": 803
+    },
+    {
+      "epoch": 0.15467487495190457,
+      "grad_norm": 0.9293098924723674,
+      "learning_rate": 1.919835349437694e-05,
+      "loss": 0.8226,
+      "step": 804
+    },
+    {
+      "epoch": 0.15486725663716813,
+      "grad_norm": 1.0500698754646924,
+      "learning_rate": 1.9195907319106394e-05,
+      "loss": 0.9058,
+      "step": 805
+    },
+    {
+      "epoch": 0.1550596383224317,
+      "grad_norm": 0.99707074535541,
+      "learning_rate": 1.9193457573667996e-05,
+      "loss": 0.9162,
+      "step": 806
+    },
+    {
+      "epoch": 0.15525202000769528,
+      "grad_norm": 0.9867373292429993,
+      "learning_rate": 1.919100425901283e-05,
+      "loss": 0.868,
+      "step": 807
+    },
+    {
+      "epoch": 0.15544440169295884,
+      "grad_norm": 0.9553696132170069,
+      "learning_rate": 1.9188547376093355e-05,
+      "loss": 0.8687,
+      "step": 808
+    },
+    {
+      "epoch": 0.1556367833782224,
+      "grad_norm": 0.9833855394316614,
+      "learning_rate": 1.918608692586342e-05,
+      "loss": 0.876,
+      "step": 809
+    },
+    {
+      "epoch": 0.15582916506348596,
+      "grad_norm": 1.0070995345362173,
+      "learning_rate": 1.918362290927825e-05,
+      "loss": 0.9152,
+      "step": 810
+    },
+    {
+      "epoch": 0.15602154674874952,
+      "grad_norm": 0.8897837954069994,
+      "learning_rate": 1.9181155327294468e-05,
+      "loss": 0.8816,
+      "step": 811
+    },
+    {
+      "epoch": 0.15621392843401308,
+      "grad_norm": 0.8975684147583171,
+      "learning_rate": 1.9178684180870072e-05,
+      "loss": 0.8746,
+      "step": 812
+    },
+    {
+      "epoch": 0.15640631011927664,
+      "grad_norm": 1.029564217004512,
+      "learning_rate": 1.9176209470964446e-05,
+      "loss": 0.8553,
+      "step": 813
+    },
+    {
+      "epoch": 0.1565986918045402,
+      "grad_norm": 0.9179508195168602,
+      "learning_rate": 1.9173731198538354e-05,
+      "loss": 0.8881,
+      "step": 814
+    },
+    {
+      "epoch": 0.15679107348980378,
+      "grad_norm": 1.006793915112619,
+      "learning_rate": 1.9171249364553956e-05,
+      "loss": 0.8812,
+      "step": 815
+    },
+    {
+      "epoch": 0.15698345517506734,
+      "grad_norm": 0.8847805533705115,
+      "learning_rate": 1.9168763969974773e-05,
+      "loss": 0.9265,
+      "step": 816
+    },
+    {
+      "epoch": 0.1571758368603309,
+      "grad_norm": 0.9190068045725875,
+      "learning_rate": 1.916627501576573e-05,
+      "loss": 0.8157,
+      "step": 817
+    },
+    {
+      "epoch": 0.15736821854559446,
+      "grad_norm": 0.9766125630312996,
+      "learning_rate": 1.916378250289312e-05,
+      "loss": 0.9133,
+      "step": 818
+    },
+    {
+      "epoch": 0.15756060023085802,
+      "grad_norm": 0.973150317241632,
+      "learning_rate": 1.9161286432324628e-05,
+      "loss": 0.828,
+      "step": 819
+    },
+    {
+      "epoch": 0.15775298191612158,
+      "grad_norm": 1.0319166842908112,
+      "learning_rate": 1.9158786805029307e-05,
+      "loss": 0.9291,
+      "step": 820
+    },
+    {
+      "epoch": 0.15794536360138514,
+      "grad_norm": 1.0226527887338348,
+      "learning_rate": 1.9156283621977603e-05,
+      "loss": 0.9202,
+      "step": 821
+    },
+    {
+      "epoch": 0.1581377452866487,
+      "grad_norm": 1.075533682087793,
+      "learning_rate": 1.9153776884141336e-05,
+      "loss": 0.948,
+      "step": 822
+    },
+    {
+      "epoch": 0.15833012697191226,
+      "grad_norm": 0.9438585320511588,
+      "learning_rate": 1.915126659249371e-05,
+      "loss": 0.8549,
+      "step": 823
+    },
+    {
+      "epoch": 0.15852250865717585,
+      "grad_norm": 0.9137083953870705,
+      "learning_rate": 1.9148752748009304e-05,
+      "loss": 0.796,
+      "step": 824
+    },
+    {
+      "epoch": 0.1587148903424394,
+      "grad_norm": 0.7008797727642785,
+      "learning_rate": 1.914623535166408e-05,
+      "loss": 0.3541,
+      "step": 825
+    },
+    {
+      "epoch": 0.15890727202770297,
+      "grad_norm": 1.0286313296624383,
+      "learning_rate": 1.9143714404435382e-05,
+      "loss": 0.9045,
+      "step": 826
+    },
+    {
+      "epoch": 0.15909965371296653,
+      "grad_norm": 1.0199261671506532,
+      "learning_rate": 1.9141189907301922e-05,
+      "loss": 0.8077,
+      "step": 827
+    },
+    {
+      "epoch": 0.1592920353982301,
+      "grad_norm": 0.8669376800704948,
+      "learning_rate": 1.9138661861243802e-05,
+      "loss": 0.8605,
+      "step": 828
+    },
+    {
+      "epoch": 0.15948441708349365,
+      "grad_norm": 1.0196636891145734,
+      "learning_rate": 1.913613026724249e-05,
+      "loss": 0.8708,
+      "step": 829
+    },
+    {
+      "epoch": 0.1596767987687572,
+      "grad_norm": 0.8845976882087305,
+      "learning_rate": 1.9133595126280848e-05,
+      "loss": 0.8921,
+      "step": 830
+    },
+    {
+      "epoch": 0.15986918045402077,
+      "grad_norm": 0.8499375067942851,
+      "learning_rate": 1.9131056439343095e-05,
+      "loss": 0.8493,
+      "step": 831
+    },
+    {
+      "epoch": 0.16006156213928435,
+      "grad_norm": 1.0070230833802218,
+      "learning_rate": 1.9128514207414838e-05,
+      "loss": 0.8972,
+      "step": 832
+    },
+    {
+      "epoch": 0.1602539438245479,
+      "grad_norm": 1.0007819671895526,
+      "learning_rate": 1.9125968431483068e-05,
+      "loss": 0.8411,
+      "step": 833
+    },
+    {
+      "epoch": 0.16044632550981147,
+      "grad_norm": 0.9279188791253913,
+      "learning_rate": 1.9123419112536132e-05,
+      "loss": 0.9526,
+      "step": 834
+    },
+    {
+      "epoch": 0.16063870719507503,
+      "grad_norm": 1.0415583268961968,
+      "learning_rate": 1.912086625156377e-05,
+      "loss": 0.9381,
+      "step": 835
+    },
+    {
+      "epoch": 0.1608310888803386,
+      "grad_norm": 0.9724582977636075,
+      "learning_rate": 1.911830984955709e-05,
+      "loss": 0.8482,
+      "step": 836
+    },
+    {
+      "epoch": 0.16102347056560215,
+      "grad_norm": 0.9516177406629052,
+      "learning_rate": 1.911574990750857e-05,
+      "loss": 0.8378,
+      "step": 837
+    },
+    {
+      "epoch": 0.1612158522508657,
+      "grad_norm": 0.9825864646501627,
+      "learning_rate": 1.9113186426412073e-05,
+      "loss": 0.8747,
+      "step": 838
+    },
+    {
+      "epoch": 0.16140823393612927,
+      "grad_norm": 0.978593866727499,
+      "learning_rate": 1.9110619407262828e-05,
+      "loss": 0.903,
+      "step": 839
+    },
+    {
+      "epoch": 0.16160061562139286,
+      "grad_norm": 0.9451847452181323,
+      "learning_rate": 1.9108048851057447e-05,
+      "loss": 0.9138,
+      "step": 840
+    },
+    {
+      "epoch": 0.16179299730665642,
+      "grad_norm": 0.9351975625912665,
+      "learning_rate": 1.9105474758793897e-05,
+      "loss": 0.9076,
+      "step": 841
+    },
+    {
+      "epoch": 0.16198537899191998,
+      "grad_norm": 0.9791438006533443,
+      "learning_rate": 1.9102897131471536e-05,
+      "loss": 0.8559,
+      "step": 842
+    },
+    {
+      "epoch": 0.16217776067718354,
+      "grad_norm": 0.9123126519006082,
+      "learning_rate": 1.9100315970091088e-05,
+      "loss": 0.8922,
+      "step": 843
+    },
+    {
+      "epoch": 0.1623701423624471,
+      "grad_norm": 0.9758349047012158,
+      "learning_rate": 1.9097731275654645e-05,
+      "loss": 0.8912,
+      "step": 844
+    },
+    {
+      "epoch": 0.16256252404771065,
+      "grad_norm": 0.9513951821423611,
+      "learning_rate": 1.909514304916568e-05,
+      "loss": 0.9305,
+      "step": 845
+    },
+    {
+      "epoch": 0.16275490573297421,
+      "grad_norm": 1.0173780876657186,
+      "learning_rate": 1.9092551291629026e-05,
+      "loss": 0.8947,
+      "step": 846
+    },
+    {
+      "epoch": 0.16294728741823777,
+      "grad_norm": 0.9040832669540396,
+      "learning_rate": 1.9089956004050893e-05,
+      "loss": 0.8971,
+      "step": 847
+    },
+    {
+      "epoch": 0.16313966910350133,
+      "grad_norm": 0.8647419554702771,
+      "learning_rate": 1.908735718743887e-05,
+      "loss": 0.9051,
+      "step": 848
+    },
+    {
+      "epoch": 0.16333205078876492,
+      "grad_norm": 0.9686874986622255,
+      "learning_rate": 1.908475484280189e-05,
+      "loss": 0.8605,
+      "step": 849
+    },
+    {
+      "epoch": 0.16352443247402848,
+      "grad_norm": 0.9712193859028443,
+      "learning_rate": 1.908214897115029e-05,
+      "loss": 0.9319,
+      "step": 850
+    },
+    {
+      "epoch": 0.16371681415929204,
+      "grad_norm": 0.8817866524878152,
+      "learning_rate": 1.907953957349575e-05,
+      "loss": 0.8337,
+      "step": 851
+    },
+    {
+      "epoch": 0.1639091958445556,
+      "grad_norm": 0.9679733722163556,
+      "learning_rate": 1.907692665085133e-05,
+      "loss": 0.8641,
+      "step": 852
+    },
+    {
+      "epoch": 0.16410157752981916,
+      "grad_norm": 0.9020438021023617,
+      "learning_rate": 1.9074310204231457e-05,
+      "loss": 0.8481,
+      "step": 853
+    },
+    {
+      "epoch": 0.16429395921508272,
+      "grad_norm": 0.8748258174166414,
+      "learning_rate": 1.9071690234651923e-05,
+      "loss": 0.9637,
+      "step": 854
+    },
+    {
+      "epoch": 0.16448634090034628,
+      "grad_norm": 1.0857010492196817,
+      "learning_rate": 1.9069066743129893e-05,
+      "loss": 0.9187,
+      "step": 855
+    },
+    {
+      "epoch": 0.16467872258560984,
+      "grad_norm": 0.9596752740236532,
+      "learning_rate": 1.90664397306839e-05,
+      "loss": 0.8755,
+      "step": 856
+    },
+    {
+      "epoch": 0.16487110427087343,
+      "grad_norm": 0.9522895442254394,
+      "learning_rate": 1.9063809198333832e-05,
+      "loss": 0.8592,
+      "step": 857
+    },
+    {
+      "epoch": 0.16506348595613698,
+      "grad_norm": 0.688349710087415,
+      "learning_rate": 1.9061175147100957e-05,
+      "loss": 0.3584,
+      "step": 858
+    },
+    {
+      "epoch": 0.16525586764140054,
+      "grad_norm": 1.0141089506450243,
+      "learning_rate": 1.905853757800791e-05,
+      "loss": 0.9231,
+      "step": 859
+    },
+    {
+      "epoch": 0.1654482493266641,
+      "grad_norm": 0.9246851795677415,
+      "learning_rate": 1.9055896492078675e-05,
+      "loss": 0.8617,
+      "step": 860
+    },
+    {
+      "epoch": 0.16564063101192766,
+      "grad_norm": 0.9544132782176149,
+      "learning_rate": 1.905325189033862e-05,
+      "loss": 0.8882,
+      "step": 861
+    },
+    {
+      "epoch": 0.16583301269719122,
+      "grad_norm": 0.9297754653461484,
+      "learning_rate": 1.905060377381447e-05,
+      "loss": 0.9039,
+      "step": 862
+    },
+    {
+      "epoch": 0.16602539438245478,
+      "grad_norm": 1.0418010797892088,
+      "learning_rate": 1.904795214353431e-05,
+      "loss": 0.9273,
+      "step": 863
+    },
+    {
+      "epoch": 0.16621777606771834,
+      "grad_norm": 0.9746324750705642,
+      "learning_rate": 1.90452970005276e-05,
+      "loss": 0.917,
+      "step": 864
+    },
+    {
+      "epoch": 0.16641015775298193,
+      "grad_norm": 0.9736378846188537,
+      "learning_rate": 1.9042638345825155e-05,
+      "loss": 0.9657,
+      "step": 865
+    },
+    {
+      "epoch": 0.1666025394382455,
+      "grad_norm": 0.9900873898136588,
+      "learning_rate": 1.9039976180459158e-05,
+      "loss": 0.8897,
+      "step": 866
+    },
+    {
+      "epoch": 0.16679492112350905,
+      "grad_norm": 0.898435876222092,
+      "learning_rate": 1.9037310505463153e-05,
+      "loss": 0.9353,
+      "step": 867
+    },
+    {
+      "epoch": 0.1669873028087726,
+      "grad_norm": 0.8618715199666401,
+      "learning_rate": 1.9034641321872043e-05,
+      "loss": 0.8366,
+      "step": 868
+    },
+    {
+      "epoch": 0.16717968449403617,
+      "grad_norm": 0.9637040665874266,
+      "learning_rate": 1.9031968630722104e-05,
+      "loss": 0.8748,
+      "step": 869
+    },
+    {
+      "epoch": 0.16737206617929973,
+      "grad_norm": 0.8872773641837096,
+      "learning_rate": 1.902929243305096e-05,
+      "loss": 0.8085,
+      "step": 870
+    },
+    {
+      "epoch": 0.1675644478645633,
+      "grad_norm": 1.0212699941719188,
+      "learning_rate": 1.902661272989761e-05,
+      "loss": 0.89,
+      "step": 871
+    },
+    {
+      "epoch": 0.16775682954982685,
+      "grad_norm": 0.9856382349583251,
+      "learning_rate": 1.9023929522302394e-05,
+      "loss": 0.9346,
+      "step": 872
+    },
+    {
+      "epoch": 0.1679492112350904,
+      "grad_norm": 1.0218127841891251,
+      "learning_rate": 1.9021242811307044e-05,
+      "loss": 0.8466,
+      "step": 873
+    },
+    {
+      "epoch": 0.168141592920354,
+      "grad_norm": 0.986117752043892,
+      "learning_rate": 1.901855259795462e-05,
+      "loss": 0.8344,
+      "step": 874
+    },
+    {
+      "epoch": 0.16833397460561755,
+      "grad_norm": 2.480639697074005,
+      "learning_rate": 1.9015858883289556e-05,
+      "loss": 0.8795,
+      "step": 875
+    },
+    {
+      "epoch": 0.1685263562908811,
+      "grad_norm": 0.9579744543830574,
+      "learning_rate": 1.9013161668357655e-05,
+      "loss": 0.9481,
+      "step": 876
+    },
+    {
+      "epoch": 0.16871873797614467,
+      "grad_norm": 0.9928084484352208,
+      "learning_rate": 1.901046095420606e-05,
+      "loss": 0.8405,
+      "step": 877
+    },
+    {
+      "epoch": 0.16891111966140823,
+      "grad_norm": 1.049198574651676,
+      "learning_rate": 1.9007756741883284e-05,
+      "loss": 0.8876,
+      "step": 878
+    },
+    {
+      "epoch": 0.1691035013466718,
+      "grad_norm": 0.9691617466704329,
+      "learning_rate": 1.9005049032439193e-05,
+      "loss": 0.909,
+      "step": 879
+    },
+    {
+      "epoch": 0.16929588303193535,
+      "grad_norm": 0.8566619005511822,
+      "learning_rate": 1.9002337826925012e-05,
+      "loss": 0.3401,
+      "step": 880
+    },
+    {
+      "epoch": 0.1694882647171989,
+      "grad_norm": 0.8815540587280513,
+      "learning_rate": 1.899962312639333e-05,
+      "loss": 0.8303,
+      "step": 881
+    },
+    {
+      "epoch": 0.1696806464024625,
+      "grad_norm": 0.9522733754458971,
+      "learning_rate": 1.8996904931898085e-05,
+      "loss": 0.9311,
+      "step": 882
+    },
+    {
+      "epoch": 0.16987302808772606,
+      "grad_norm": 1.0110834157160402,
+      "learning_rate": 1.899418324449457e-05,
+      "loss": 0.9571,
+      "step": 883
+    },
+    {
+      "epoch": 0.17006540977298962,
+      "grad_norm": 0.9657805650693476,
+      "learning_rate": 1.8991458065239444e-05,
+      "loss": 0.8597,
+      "step": 884
+    },
+    {
+      "epoch": 0.17025779145825318,
+      "grad_norm": 0.985517048708571,
+      "learning_rate": 1.8988729395190712e-05,
+      "loss": 0.899,
+      "step": 885
+    },
+    {
+      "epoch": 0.17045017314351674,
+      "grad_norm": 1.047450348265624,
+      "learning_rate": 1.8985997235407735e-05,
+      "loss": 0.8805,
+      "step": 886
+    },
+    {
+      "epoch": 0.1706425548287803,
+      "grad_norm": 1.013218925298836,
+      "learning_rate": 1.898326158695124e-05,
+      "loss": 0.872,
+      "step": 887
+    },
+    {
+      "epoch": 0.17083493651404386,
+      "grad_norm": 0.9318947801481622,
+      "learning_rate": 1.8980522450883287e-05,
+      "loss": 0.9303,
+      "step": 888
+    },
+    {
+      "epoch": 0.17102731819930742,
+      "grad_norm": 0.8708634081450302,
+      "learning_rate": 1.8977779828267314e-05,
+      "loss": 0.8934,
+      "step": 889
+    },
+    {
+      "epoch": 0.17121969988457098,
+      "grad_norm": 0.9766844583801871,
+      "learning_rate": 1.8975033720168094e-05,
+      "loss": 0.8903,
+      "step": 890
+    },
+    {
+      "epoch": 0.17141208156983456,
+      "grad_norm": 0.751130360029687,
+      "learning_rate": 1.897228412765177e-05,
+      "loss": 0.3694,
+      "step": 891
+    },
+    {
+      "epoch": 0.17160446325509812,
+      "grad_norm": 0.9462063809434174,
+      "learning_rate": 1.896953105178582e-05,
+      "loss": 0.8981,
+      "step": 892
+    },
+    {
+      "epoch": 0.17179684494036168,
+      "grad_norm": 0.8038659496858896,
+      "learning_rate": 1.8966774493639084e-05,
+      "loss": 0.3452,
+      "step": 893
+    },
+    {
+      "epoch": 0.17198922662562524,
+      "grad_norm": 0.9610223380460479,
+      "learning_rate": 1.896401445428176e-05,
+      "loss": 0.9188,
+      "step": 894
+    },
+    {
+      "epoch": 0.1721816083108888,
+      "grad_norm": 0.6674266061634574,
+      "learning_rate": 1.896125093478538e-05,
+      "loss": 0.3353,
+      "step": 895
+    },
+    {
+      "epoch": 0.17237398999615236,
+      "grad_norm": 0.9180276182644383,
+      "learning_rate": 1.895848393622284e-05,
+      "loss": 0.8749,
+      "step": 896
+    },
+    {
+      "epoch": 0.17256637168141592,
+      "grad_norm": 1.0204053835542115,
+      "learning_rate": 1.895571345966839e-05,
+      "loss": 0.9323,
+      "step": 897
+    },
+    {
+      "epoch": 0.17275875336667948,
+      "grad_norm": 0.8841668675260473,
+      "learning_rate": 1.8952939506197622e-05,
+      "loss": 0.8816,
+      "step": 898
+    },
+    {
+      "epoch": 0.17295113505194307,
+      "grad_norm": 1.0166999448519238,
+      "learning_rate": 1.8950162076887477e-05,
+      "loss": 0.8631,
+      "step": 899
+    },
+    {
+      "epoch": 0.17314351673720663,
+      "grad_norm": 0.9826707554219699,
+      "learning_rate": 1.894738117281625e-05,
+      "loss": 0.9002,
+      "step": 900
+    },
+    {
+      "epoch": 0.17333589842247019,
+      "grad_norm": 0.9194574854661118,
+      "learning_rate": 1.8944596795063584e-05,
+      "loss": 0.8675,
+      "step": 901
+    },
+    {
+      "epoch": 0.17352828010773375,
+      "grad_norm": 0.9721848488944356,
+      "learning_rate": 1.894180894471047e-05,
+      "loss": 0.8207,
+      "step": 902
+    },
+    {
+      "epoch": 0.1737206617929973,
+      "grad_norm": 1.1036658513881612,
+      "learning_rate": 1.8939017622839253e-05,
+      "loss": 0.3372,
+      "step": 903
+    },
+    {
+      "epoch": 0.17391304347826086,
+      "grad_norm": 0.9502939554848907,
+      "learning_rate": 1.8936222830533613e-05,
+      "loss": 0.836,
+      "step": 904
+    },
+    {
+      "epoch": 0.17410542516352442,
+      "grad_norm": 1.0010844485247676,
+      "learning_rate": 1.8933424568878586e-05,
+      "loss": 0.9182,
+      "step": 905
+    },
+    {
+      "epoch": 0.17429780684878798,
+      "grad_norm": 0.971191823537388,
+      "learning_rate": 1.8930622838960555e-05,
+      "loss": 0.9156,
+      "step": 906
+    },
+    {
+      "epoch": 0.17449018853405157,
+      "grad_norm": 0.9163061644978593,
+      "learning_rate": 1.8927817641867244e-05,
+      "loss": 0.8614,
+      "step": 907
+    },
+    {
+      "epoch": 0.17468257021931513,
+      "grad_norm": 0.9777425966608534,
+      "learning_rate": 1.8925008978687737e-05,
+      "loss": 0.8763,
+      "step": 908
+    },
+    {
+      "epoch": 0.1748749519045787,
+      "grad_norm": 0.921417407567931,
+      "learning_rate": 1.8922196850512446e-05,
+      "loss": 0.875,
+      "step": 909
+    },
+    {
+      "epoch": 0.17506733358984225,
+      "grad_norm": 0.9329351732252843,
+      "learning_rate": 1.8919381258433135e-05,
+      "loss": 0.8995,
+      "step": 910
+    },
+    {
+      "epoch": 0.1752597152751058,
+      "grad_norm": 0.8861517023420692,
+      "learning_rate": 1.8916562203542916e-05,
+      "loss": 0.8299,
+      "step": 911
+    },
+    {
+      "epoch": 0.17545209696036937,
+      "grad_norm": 1.11264997207102,
+      "learning_rate": 1.8913739686936244e-05,
+      "loss": 0.8072,
+      "step": 912
+    },
+    {
+      "epoch": 0.17564447864563293,
+      "grad_norm": 0.9497145987004891,
+      "learning_rate": 1.8910913709708918e-05,
+      "loss": 0.8743,
+      "step": 913
+    },
+    {
+      "epoch": 0.1758368603308965,
+      "grad_norm": 0.9364406588257441,
+      "learning_rate": 1.8908084272958077e-05,
+      "loss": 0.8526,
+      "step": 914
+    },
+    {
+      "epoch": 0.17602924201616005,
+      "grad_norm": 0.8103508198866236,
+      "learning_rate": 1.8905251377782206e-05,
+      "loss": 0.3683,
+      "step": 915
+    },
+    {
+      "epoch": 0.17622162370142364,
+      "grad_norm": 1.0179011225012622,
+      "learning_rate": 1.8902415025281136e-05,
+      "loss": 0.9414,
+      "step": 916
+    },
+    {
+      "epoch": 0.1764140053866872,
+      "grad_norm": 1.0220135536048063,
+      "learning_rate": 1.889957521655603e-05,
+      "loss": 0.9551,
+      "step": 917
+    },
+    {
+      "epoch": 0.17660638707195075,
+      "grad_norm": 1.028098332386793,
+      "learning_rate": 1.8896731952709408e-05,
+      "loss": 0.9077,
+      "step": 918
+    },
+    {
+      "epoch": 0.17679876875721431,
+      "grad_norm": 0.9794958410084269,
+      "learning_rate": 1.8893885234845117e-05,
+      "loss": 0.865,
+      "step": 919
+    },
+    {
+      "epoch": 0.17699115044247787,
+      "grad_norm": 0.9181070358142106,
+      "learning_rate": 1.8891035064068354e-05,
+      "loss": 0.8799,
+      "step": 920
+    },
+    {
+      "epoch": 0.17718353212774143,
+      "grad_norm": 0.9907442962368943,
+      "learning_rate": 1.888818144148565e-05,
+      "loss": 0.8728,
+      "step": 921
+    },
+    {
+      "epoch": 0.177375913813005,
+      "grad_norm": 0.9005215910515729,
+      "learning_rate": 1.888532436820488e-05,
+      "loss": 0.8839,
+      "step": 922
+    },
+    {
+      "epoch": 0.17756829549826855,
+      "grad_norm": 0.9229007720542642,
+      "learning_rate": 1.8882463845335263e-05,
+      "loss": 0.8593,
+      "step": 923
+    },
+    {
+      "epoch": 0.17776067718353214,
+      "grad_norm": 0.9802358590404685,
+      "learning_rate": 1.8879599873987343e-05,
+      "loss": 0.8761,
+      "step": 924
+    },
+    {
+      "epoch": 0.1779530588687957,
+      "grad_norm": 0.9623005402874502,
+      "learning_rate": 1.8876732455273022e-05,
+      "loss": 0.8986,
+      "step": 925
+    },
+    {
+      "epoch": 0.17814544055405926,
+      "grad_norm": 0.9046187176292793,
+      "learning_rate": 1.8873861590305527e-05,
+      "loss": 0.8677,
+      "step": 926
+    },
+    {
+      "epoch": 0.17833782223932282,
+      "grad_norm": 0.9836831440984217,
+      "learning_rate": 1.8870987280199428e-05,
+      "loss": 0.8455,
+      "step": 927
+    },
+    {
+      "epoch": 0.17853020392458638,
+      "grad_norm": 0.8468029068837767,
+      "learning_rate": 1.886810952607063e-05,
+      "loss": 0.864,
+      "step": 928
+    },
+    {
+      "epoch": 0.17872258560984994,
+      "grad_norm": 0.9537655028469293,
+      "learning_rate": 1.8865228329036372e-05,
+      "loss": 0.8564,
+      "step": 929
+    },
+    {
+      "epoch": 0.1789149672951135,
+      "grad_norm": 0.6865427828385117,
+      "learning_rate": 1.886234369021524e-05,
+      "loss": 0.3492,
+      "step": 930
+    },
+    {
+      "epoch": 0.17910734898037706,
+      "grad_norm": 0.8840816886797802,
+      "learning_rate": 1.885945561072715e-05,
+      "loss": 0.9002,
+      "step": 931
+    },
+    {
+      "epoch": 0.17929973066564064,
+      "grad_norm": 0.8942274598351084,
+      "learning_rate": 1.885656409169335e-05,
+      "loss": 0.9106,
+      "step": 932
+    },
+    {
+      "epoch": 0.1794921123509042,
+      "grad_norm": 0.9078498945818361,
+      "learning_rate": 1.885366913423643e-05,
+      "loss": 0.8565,
+      "step": 933
+    },
+    {
+      "epoch": 0.17968449403616776,
+      "grad_norm": 0.8773456115955539,
+      "learning_rate": 1.8850770739480312e-05,
+      "loss": 0.8751,
+      "step": 934
+    },
+    {
+      "epoch": 0.17987687572143132,
+      "grad_norm": 0.9718730796426609,
+      "learning_rate": 1.8847868908550252e-05,
+      "loss": 0.933,
+      "step": 935
+    },
+    {
+      "epoch": 0.18006925740669488,
+      "grad_norm": 0.9474654602828411,
+      "learning_rate": 1.8844963642572837e-05,
+      "loss": 0.8757,
+      "step": 936
+    },
+    {
+      "epoch": 0.18026163909195844,
+      "grad_norm": 0.9600093171989512,
+      "learning_rate": 1.8842054942676e-05,
+      "loss": 0.8292,
+      "step": 937
+    },
+    {
+      "epoch": 0.180454020777222,
+      "grad_norm": 0.9718892445837143,
+      "learning_rate": 1.8839142809988987e-05,
+      "loss": 0.9263,
+      "step": 938
+    },
+    {
+      "epoch": 0.18064640246248556,
+      "grad_norm": 0.9182286931111926,
+      "learning_rate": 1.88362272456424e-05,
+      "loss": 0.8586,
+      "step": 939
+    },
+    {
+      "epoch": 0.18083878414774912,
+      "grad_norm": 0.7006120794339297,
+      "learning_rate": 1.8833308250768153e-05,
+      "loss": 0.3301,
+      "step": 940
+    },
+    {
+      "epoch": 0.1810311658330127,
+      "grad_norm": 0.9046668387327943,
+      "learning_rate": 1.8830385826499507e-05,
+      "loss": 0.8487,
+      "step": 941
+    },
+    {
+      "epoch": 0.18122354751827627,
+      "grad_norm": 1.1142942929607107,
+      "learning_rate": 1.882745997397104e-05,
+      "loss": 0.8062,
+      "step": 942
+    },
+    {
+      "epoch": 0.18141592920353983,
+      "grad_norm": 0.9941845680471323,
+      "learning_rate": 1.8824530694318675e-05,
+      "loss": 0.8775,
+      "step": 943
+    },
+    {
+      "epoch": 0.1816083108888034,
+      "grad_norm": 0.8851676905920371,
+      "learning_rate": 1.882159798867966e-05,
+      "loss": 0.885,
+      "step": 944
+    },
+    {
+      "epoch": 0.18180069257406695,
+      "grad_norm": 0.9114266572546215,
+      "learning_rate": 1.8818661858192562e-05,
+      "loss": 0.8441,
+      "step": 945
+    },
+    {
+      "epoch": 0.1819930742593305,
+      "grad_norm": 0.9687196355625464,
+      "learning_rate": 1.88157223039973e-05,
+      "loss": 0.8794,
+      "step": 946
+    },
+    {
+      "epoch": 0.18218545594459407,
+      "grad_norm": 1.0785432986660761,
+      "learning_rate": 1.8812779327235106e-05,
+      "loss": 0.8542,
+      "step": 947
+    },
+    {
+      "epoch": 0.18237783762985763,
+      "grad_norm": 0.9120976122315204,
+      "learning_rate": 1.880983292904854e-05,
+      "loss": 0.9422,
+      "step": 948
+    },
+    {
+      "epoch": 0.1825702193151212,
+      "grad_norm": 1.0251686530565194,
+      "learning_rate": 1.88068831105815e-05,
+      "loss": 0.878,
+      "step": 949
+    },
+    {
+      "epoch": 0.18276260100038477,
+      "grad_norm": 0.8359849029497917,
+      "learning_rate": 1.8803929872979214e-05,
+      "loss": 0.8516,
+      "step": 950
+    },
+    {
+      "epoch": 0.18295498268564833,
+      "grad_norm": 0.9104923528360851,
+      "learning_rate": 1.8800973217388215e-05,
+      "loss": 0.896,
+      "step": 951
+    },
+    {
+      "epoch": 0.1831473643709119,
+      "grad_norm": 0.9083837498465834,
+      "learning_rate": 1.879801314495639e-05,
+      "loss": 0.9079,
+      "step": 952
+    },
+    {
+      "epoch": 0.18333974605617545,
+      "grad_norm": 0.9603341101839917,
+      "learning_rate": 1.879504965683294e-05,
+      "loss": 0.9522,
+      "step": 953
+    },
+    {
+      "epoch": 0.183532127741439,
+      "grad_norm": 0.9615415983466601,
+      "learning_rate": 1.8792082754168385e-05,
+      "loss": 0.8639,
+      "step": 954
+    },
+    {
+      "epoch": 0.18372450942670257,
+      "grad_norm": 0.9380127415810599,
+      "learning_rate": 1.878911243811459e-05,
+      "loss": 0.8359,
+      "step": 955
+    },
+    {
+      "epoch": 0.18391689111196613,
+      "grad_norm": 0.934114359959676,
+      "learning_rate": 1.8786138709824726e-05,
+      "loss": 0.8185,
+      "step": 956
+    },
+    {
+      "epoch": 0.18410927279722972,
+      "grad_norm": 0.6513436264023759,
+      "learning_rate": 1.8783161570453295e-05,
+      "loss": 0.3611,
+      "step": 957
+    },
+    {
+      "epoch": 0.18430165448249328,
+      "grad_norm": 0.9203133754143358,
+      "learning_rate": 1.878018102115614e-05,
+      "loss": 0.8773,
+      "step": 958
+    },
+    {
+      "epoch": 0.18449403616775684,
+      "grad_norm": 0.9681451208822649,
+      "learning_rate": 1.8777197063090394e-05,
+      "loss": 0.8688,
+      "step": 959
+    },
+    {
+      "epoch": 0.1846864178530204,
+      "grad_norm": 1.045785744221588,
+      "learning_rate": 1.877420969741454e-05,
+      "loss": 0.8936,
+      "step": 960
+    },
+    {
+      "epoch": 0.18487879953828396,
+      "grad_norm": 1.0481450643220274,
+      "learning_rate": 1.877121892528838e-05,
+      "loss": 0.8627,
+      "step": 961
+    },
+    {
+      "epoch": 0.18507118122354752,
+      "grad_norm": 0.9207027209188703,
+      "learning_rate": 1.876822474787303e-05,
+      "loss": 0.9344,
+      "step": 962
+    },
+    {
+      "epoch": 0.18526356290881107,
+      "grad_norm": 0.9971793738871656,
+      "learning_rate": 1.8765227166330933e-05,
+      "loss": 0.9319,
+      "step": 963
+    },
+    {
+      "epoch": 0.18545594459407463,
+      "grad_norm": 0.9285441450176842,
+      "learning_rate": 1.8762226181825857e-05,
+      "loss": 0.9059,
+      "step": 964
+    },
+    {
+      "epoch": 0.1856483262793382,
+      "grad_norm": 0.9114561732909712,
+      "learning_rate": 1.875922179552288e-05,
+      "loss": 0.8251,
+      "step": 965
+    },
+    {
+      "epoch": 0.18584070796460178,
+      "grad_norm": 0.9070566812097544,
+      "learning_rate": 1.875621400858842e-05,
+      "loss": 0.9178,
+      "step": 966
+    },
+    {
+      "epoch": 0.18603308964986534,
+      "grad_norm": 0.878917842971885,
+      "learning_rate": 1.875320282219019e-05,
+      "loss": 0.8786,
+      "step": 967
+    },
+    {
+      "epoch": 0.1862254713351289,
+      "grad_norm": 0.930485304434635,
+      "learning_rate": 1.8750188237497247e-05,
+      "loss": 0.9173,
+      "step": 968
+    },
+    {
+      "epoch": 0.18641785302039246,
+      "grad_norm": 0.9533471956968331,
+      "learning_rate": 1.874717025567995e-05,
+      "loss": 0.9176,
+      "step": 969
+    },
+    {
+      "epoch": 0.18661023470565602,
+      "grad_norm": 0.928882623869484,
+      "learning_rate": 1.874414887790999e-05,
+      "loss": 0.9184,
+      "step": 970
+    },
+    {
+      "epoch": 0.18680261639091958,
+      "grad_norm": 0.8948657653912855,
+      "learning_rate": 1.8741124105360363e-05,
+      "loss": 0.9167,
+      "step": 971
+    },
+    {
+      "epoch": 0.18699499807618314,
+      "grad_norm": 1.006299446158606,
+      "learning_rate": 1.873809593920539e-05,
+      "loss": 0.8657,
+      "step": 972
+    },
+    {
+      "epoch": 0.1871873797614467,
+      "grad_norm": 0.9462035244109573,
+      "learning_rate": 1.8735064380620717e-05,
+      "loss": 0.8997,
+      "step": 973
+    },
+    {
+      "epoch": 0.18737976144671029,
+      "grad_norm": 0.8322720541834384,
+      "learning_rate": 1.873202943078329e-05,
+      "loss": 0.8461,
+      "step": 974
+    },
+    {
+      "epoch": 0.18757214313197385,
+      "grad_norm": 1.0610603277185615,
+      "learning_rate": 1.8728991090871387e-05,
+      "loss": 0.9112,
+      "step": 975
+    },
+    {
+      "epoch": 0.1877645248172374,
+      "grad_norm": 0.9722765836182882,
+      "learning_rate": 1.8725949362064596e-05,
+      "loss": 0.9078,
+      "step": 976
+    },
+    {
+      "epoch": 0.18795690650250096,
+      "grad_norm": 1.0101410760979044,
+      "learning_rate": 1.8722904245543817e-05,
+      "loss": 0.8172,
+      "step": 977
+    },
+    {
+      "epoch": 0.18814928818776452,
+      "grad_norm": 1.0052335814174413,
+      "learning_rate": 1.871985574249127e-05,
+      "loss": 0.8614,
+      "step": 978
+    },
+    {
+      "epoch": 0.18834166987302808,
+      "grad_norm": 0.979761512596374,
+      "learning_rate": 1.8716803854090495e-05,
+      "loss": 0.9046,
+      "step": 979
+    },
+    {
+      "epoch": 0.18853405155829164,
+      "grad_norm": 0.9466922675791113,
+      "learning_rate": 1.8713748581526334e-05,
+      "loss": 0.8958,
+      "step": 980
+    },
+    {
+      "epoch": 0.1887264332435552,
+      "grad_norm": 0.9092836233011384,
+      "learning_rate": 1.871068992598495e-05,
+      "loss": 0.8753,
+      "step": 981
+    },
+    {
+      "epoch": 0.1889188149288188,
+      "grad_norm": 0.9002711702081604,
+      "learning_rate": 1.8707627888653816e-05,
+      "loss": 0.8478,
+      "step": 982
+    },
+    {
+      "epoch": 0.18911119661408235,
+      "grad_norm": 0.8684815371746799,
+      "learning_rate": 1.8704562470721728e-05,
+      "loss": 0.8538,
+      "step": 983
+    },
+    {
+      "epoch": 0.1893035782993459,
+      "grad_norm": 0.7191979240961676,
+      "learning_rate": 1.870149367337878e-05,
+      "loss": 0.3681,
+      "step": 984
+    },
+    {
+      "epoch": 0.18949595998460947,
+      "grad_norm": 1.0409942689377714,
+      "learning_rate": 1.8698421497816386e-05,
+      "loss": 0.8566,
+      "step": 985
+    },
+    {
+      "epoch": 0.18968834166987303,
+      "grad_norm": 0.8649016300461364,
+      "learning_rate": 1.869534594522727e-05,
+      "loss": 0.8513,
+      "step": 986
+    },
+    {
+      "epoch": 0.1898807233551366,
+      "grad_norm": 0.8938117061147584,
+      "learning_rate": 1.8692267016805473e-05,
+      "loss": 0.8446,
+      "step": 987
+    },
+    {
+      "epoch": 0.19007310504040015,
+      "grad_norm": 0.8603410677754029,
+      "learning_rate": 1.8689184713746333e-05,
+      "loss": 0.7869,
+      "step": 988
+    },
+    {
+      "epoch": 0.1902654867256637,
+      "grad_norm": 0.9076131149459004,
+      "learning_rate": 1.868609903724651e-05,
+      "loss": 0.8574,
+      "step": 989
+    },
+    {
+      "epoch": 0.19045786841092727,
+      "grad_norm": 0.6113316529431607,
+      "learning_rate": 1.8683009988503972e-05,
+      "loss": 0.3674,
+      "step": 990
+    },
+    {
+      "epoch": 0.19065025009619085,
+      "grad_norm": 0.932789988911279,
+      "learning_rate": 1.867991756871799e-05,
+      "loss": 0.8504,
+      "step": 991
+    },
+    {
+      "epoch": 0.1908426317814544,
+      "grad_norm": 0.952254289718065,
+      "learning_rate": 1.867682177908915e-05,
+      "loss": 0.9072,
+      "step": 992
+    },
+    {
+      "epoch": 0.19103501346671797,
+      "grad_norm": 0.9037924709325529,
+      "learning_rate": 1.867372262081934e-05,
+      "loss": 0.9011,
+      "step": 993
+    },
+    {
+      "epoch": 0.19122739515198153,
+      "grad_norm": 0.9539826584737561,
+      "learning_rate": 1.8670620095111766e-05,
+      "loss": 0.8952,
+      "step": 994
+    },
+    {
+      "epoch": 0.1914197768372451,
+      "grad_norm": 0.9123968758910869,
+      "learning_rate": 1.8667514203170934e-05,
+      "loss": 0.9539,
+      "step": 995
+    },
+    {
+      "epoch": 0.19161215852250865,
+      "grad_norm": 0.9620025718844469,
+      "learning_rate": 1.8664404946202658e-05,
+      "loss": 0.848,
+      "step": 996
+    },
+    {
+      "epoch": 0.1918045402077722,
+      "grad_norm": 0.9097351697192403,
+      "learning_rate": 1.8661292325414058e-05,
+      "loss": 0.8746,
+      "step": 997
+    },
+    {
+      "epoch": 0.19199692189303577,
+      "grad_norm": 0.9264817405336419,
+      "learning_rate": 1.865817634201356e-05,
+      "loss": 0.8765,
+      "step": 998
+    },
+    {
+      "epoch": 0.19218930357829936,
+      "grad_norm": 0.9632504891047245,
+      "learning_rate": 1.8655056997210893e-05,
+      "loss": 0.8942,
+      "step": 999
+    },
+    {
+      "epoch": 0.19238168526356292,
+      "grad_norm": 1.0095247912950702,
+      "learning_rate": 1.8651934292217097e-05,
+      "loss": 0.8857,
+      "step": 1000
+    },
+    {
+      "epoch": 0.19257406694882648,
+      "grad_norm": 0.8580365850577722,
+      "learning_rate": 1.864880822824452e-05,
+      "loss": 0.8711,
+      "step": 1001
+    },
+    {
+      "epoch": 0.19276644863409004,
+      "grad_norm": 0.9811943447965918,
+      "learning_rate": 1.8645678806506795e-05,
+      "loss": 0.8854,
+      "step": 1002
+    },
+    {
+      "epoch": 0.1929588303193536,
+      "grad_norm": 0.8186579723074203,
+      "learning_rate": 1.864254602821888e-05,
+      "loss": 0.8594,
+      "step": 1003
+    },
+    {
+      "epoch": 0.19315121200461716,
+      "grad_norm": 0.8888607883812975,
+      "learning_rate": 1.8639409894597026e-05,
+      "loss": 0.8946,
+      "step": 1004
+    },
+    {
+      "epoch": 0.19334359368988072,
+      "grad_norm": 0.8626505105797818,
+      "learning_rate": 1.8636270406858786e-05,
+      "loss": 0.8751,
+      "step": 1005
+    },
+    {
+      "epoch": 0.19353597537514428,
+      "grad_norm": 0.9021832149103206,
+      "learning_rate": 1.8633127566223023e-05,
+      "loss": 0.8335,
+      "step": 1006
+    },
+    {
+      "epoch": 0.19372835706040784,
+      "grad_norm": 0.8935511974434246,
+      "learning_rate": 1.862998137390989e-05,
+      "loss": 0.8346,
+      "step": 1007
+    },
+    {
+      "epoch": 0.19392073874567142,
+      "grad_norm": 0.9997215746438762,
+      "learning_rate": 1.8626831831140845e-05,
+      "loss": 0.886,
+      "step": 1008
+    },
+    {
+      "epoch": 0.19411312043093498,
+      "grad_norm": 0.9615871202913168,
+      "learning_rate": 1.8623678939138652e-05,
+      "loss": 0.8863,
+      "step": 1009
+    },
+    {
+      "epoch": 0.19430550211619854,
+      "grad_norm": 0.8654971847797834,
+      "learning_rate": 1.8620522699127374e-05,
+      "loss": 0.8718,
+      "step": 1010
+    },
+    {
+      "epoch": 0.1944978838014621,
+      "grad_norm": 1.13766203306829,
+      "learning_rate": 1.8617363112332376e-05,
+      "loss": 0.8654,
+      "step": 1011
+    },
+    {
+      "epoch": 0.19469026548672566,
+      "grad_norm": 0.9101847399971971,
+      "learning_rate": 1.8614200179980307e-05,
+      "loss": 0.8398,
+      "step": 1012
+    },
+    {
+      "epoch": 0.19488264717198922,
+      "grad_norm": 0.8871755410472815,
+      "learning_rate": 1.8611033903299136e-05,
+      "loss": 0.889,
+      "step": 1013
+    },
+    {
+      "epoch": 0.19507502885725278,
+      "grad_norm": 0.9370532750853839,
+      "learning_rate": 1.8607864283518116e-05,
+      "loss": 0.8442,
+      "step": 1014
+    },
+    {
+      "epoch": 0.19526741054251634,
+      "grad_norm": 0.8701806719937798,
+      "learning_rate": 1.8604691321867804e-05,
+      "loss": 0.8728,
+      "step": 1015
+    },
+    {
+      "epoch": 0.19545979222777993,
+      "grad_norm": 0.957712883757812,
+      "learning_rate": 1.8601515019580053e-05,
+      "loss": 0.9187,
+      "step": 1016
+    },
+    {
+      "epoch": 0.1956521739130435,
+      "grad_norm": 0.9296201408932494,
+      "learning_rate": 1.8598335377888012e-05,
+      "loss": 0.8676,
+      "step": 1017
+    },
+    {
+      "epoch": 0.19584455559830705,
+      "grad_norm": 0.6843377397284007,
+      "learning_rate": 1.8595152398026128e-05,
+      "loss": 0.3178,
+      "step": 1018
+    },
+    {
+      "epoch": 0.1960369372835706,
+      "grad_norm": 0.6688007671939152,
+      "learning_rate": 1.8591966081230142e-05,
+      "loss": 0.3305,
+      "step": 1019
+    },
+    {
+      "epoch": 0.19622931896883417,
+      "grad_norm": 0.9030043334184993,
+      "learning_rate": 1.8588776428737095e-05,
+      "loss": 0.8671,
+      "step": 1020
+    },
+    {
+      "epoch": 0.19642170065409773,
+      "grad_norm": 0.937763769513175,
+      "learning_rate": 1.858558344178532e-05,
+      "loss": 0.8862,
+      "step": 1021
+    },
+    {
+      "epoch": 0.19661408233936128,
+      "grad_norm": 1.0056197953036874,
+      "learning_rate": 1.8582387121614437e-05,
+      "loss": 0.9016,
+      "step": 1022
+    },
+    {
+      "epoch": 0.19680646402462484,
+      "grad_norm": 0.992365223740527,
+      "learning_rate": 1.857918746946538e-05,
+      "loss": 0.9638,
+      "step": 1023
+    },
+    {
+      "epoch": 0.19699884570988843,
+      "grad_norm": 1.0180141740119626,
+      "learning_rate": 1.8575984486580353e-05,
+      "loss": 0.9048,
+      "step": 1024
+    },
+    {
+      "epoch": 0.197191227395152,
+      "grad_norm": 0.9489527458719221,
+      "learning_rate": 1.857277817420287e-05,
+      "loss": 0.8934,
+      "step": 1025
+    },
+    {
+      "epoch": 0.19738360908041555,
+      "grad_norm": 0.9634250827667012,
+      "learning_rate": 1.8569568533577727e-05,
+      "loss": 0.9565,
+      "step": 1026
+    },
+    {
+      "epoch": 0.1975759907656791,
+      "grad_norm": 0.9886799734819981,
+      "learning_rate": 1.8566355565951023e-05,
+      "loss": 0.8883,
+      "step": 1027
+    },
+    {
+      "epoch": 0.19776837245094267,
+      "grad_norm": 0.9743923440991857,
+      "learning_rate": 1.8563139272570142e-05,
+      "loss": 0.899,
+      "step": 1028
+    },
+    {
+      "epoch": 0.19796075413620623,
+      "grad_norm": 1.0037934641198316,
+      "learning_rate": 1.8559919654683756e-05,
+      "loss": 0.9059,
+      "step": 1029
+    },
+    {
+      "epoch": 0.1981531358214698,
+      "grad_norm": 0.9394858110257254,
+      "learning_rate": 1.8556696713541833e-05,
+      "loss": 0.9073,
+      "step": 1030
+    },
+    {
+      "epoch": 0.19834551750673335,
+      "grad_norm": 0.8693980135027408,
+      "learning_rate": 1.855347045039563e-05,
+      "loss": 0.8685,
+      "step": 1031
+    },
+    {
+      "epoch": 0.1985378991919969,
+      "grad_norm": 0.9083204547092582,
+      "learning_rate": 1.8550240866497697e-05,
+      "loss": 0.8808,
+      "step": 1032
+    },
+    {
+      "epoch": 0.1987302808772605,
+      "grad_norm": 0.9703007424179821,
+      "learning_rate": 1.854700796310186e-05,
+      "loss": 0.878,
+      "step": 1033
+    },
+    {
+      "epoch": 0.19892266256252406,
+      "grad_norm": 0.9919177983723136,
+      "learning_rate": 1.8543771741463254e-05,
+      "loss": 0.8976,
+      "step": 1034
+    },
+    {
+      "epoch": 0.19911504424778761,
+      "grad_norm": 0.8715635086198168,
+      "learning_rate": 1.8540532202838286e-05,
+      "loss": 0.8677,
+      "step": 1035
+    },
+    {
+      "epoch": 0.19930742593305117,
+      "grad_norm": 0.852949962033007,
+      "learning_rate": 1.8537289348484658e-05,
+      "loss": 0.8195,
+      "step": 1036
+    },
+    {
+      "epoch": 0.19949980761831473,
+      "grad_norm": 1.444125779166957,
+      "learning_rate": 1.8534043179661357e-05,
+      "loss": 0.354,
+      "step": 1037
+    },
+    {
+      "epoch": 0.1996921893035783,
+      "grad_norm": 1.027024172617474,
+      "learning_rate": 1.8530793697628658e-05,
+      "loss": 0.9655,
+      "step": 1038
+    },
+    {
+      "epoch": 0.19988457098884185,
+      "grad_norm": 1.0068960362033594,
+      "learning_rate": 1.8527540903648122e-05,
+      "loss": 0.9145,
+      "step": 1039
+    },
+    {
+      "epoch": 0.2000769526741054,
+      "grad_norm": 1.039362509344859,
+      "learning_rate": 1.8524284798982595e-05,
+      "loss": 0.8776,
+      "step": 1040
+    },
+    {
+      "epoch": 0.200269334359369,
+      "grad_norm": 0.9214291124280953,
+      "learning_rate": 1.852102538489621e-05,
+      "loss": 0.8559,
+      "step": 1041
+    },
+    {
+      "epoch": 0.20046171604463256,
+      "grad_norm": 0.9254709024506141,
+      "learning_rate": 1.8517762662654383e-05,
+      "loss": 0.363,
+      "step": 1042
+    },
+    {
+      "epoch": 0.20065409772989612,
+      "grad_norm": 0.9559166760116719,
+      "learning_rate": 1.851449663352381e-05,
+      "loss": 0.842,
+      "step": 1043
+    },
+    {
+      "epoch": 0.20084647941515968,
+      "grad_norm": 0.9305739124094535,
+      "learning_rate": 1.851122729877249e-05,
+      "loss": 0.8336,
+      "step": 1044
+    },
+    {
+      "epoch": 0.20103886110042324,
+      "grad_norm": 0.9934074922857435,
+      "learning_rate": 1.8507954659669677e-05,
+      "loss": 0.8679,
+      "step": 1045
+    },
+    {
+      "epoch": 0.2012312427856868,
+      "grad_norm": 0.9060957061316686,
+      "learning_rate": 1.850467871748593e-05,
+      "loss": 0.8383,
+      "step": 1046
+    },
+    {
+      "epoch": 0.20142362447095036,
+      "grad_norm": 0.8995403945377359,
+      "learning_rate": 1.850139947349308e-05,
+      "loss": 0.9124,
+      "step": 1047
+    },
+    {
+      "epoch": 0.20161600615621392,
+      "grad_norm": 0.9909529702974132,
+      "learning_rate": 1.8498116928964244e-05,
+      "loss": 0.8749,
+      "step": 1048
+    },
+    {
+      "epoch": 0.2018083878414775,
+      "grad_norm": 0.8710685382093771,
+      "learning_rate": 1.849483108517381e-05,
+      "loss": 0.8562,
+      "step": 1049
+    },
+    {
+      "epoch": 0.20200076952674106,
+      "grad_norm": 0.9364324281170491,
+      "learning_rate": 1.849154194339747e-05,
+      "loss": 0.8702,
+      "step": 1050
+    },
+    {
+      "epoch": 0.20219315121200462,
+      "grad_norm": 0.8823702718521891,
+      "learning_rate": 1.8488249504912173e-05,
+      "loss": 0.8689,
+      "step": 1051
+    },
+    {
+      "epoch": 0.20238553289726818,
+      "grad_norm": 0.8787330053604641,
+      "learning_rate": 1.8484953770996163e-05,
+      "loss": 0.8099,
+      "step": 1052
+    },
+    {
+      "epoch": 0.20257791458253174,
+      "grad_norm": 0.9935684390437574,
+      "learning_rate": 1.848165474292895e-05,
+      "loss": 0.858,
+      "step": 1053
+    },
+    {
+      "epoch": 0.2027702962677953,
+      "grad_norm": 0.9683370071005181,
+      "learning_rate": 1.8478352421991334e-05,
+      "loss": 0.9293,
+      "step": 1054
+    },
+    {
+      "epoch": 0.20296267795305886,
+      "grad_norm": 0.9280101241857818,
+      "learning_rate": 1.847504680946539e-05,
+      "loss": 0.8556,
+      "step": 1055
+    },
+    {
+      "epoch": 0.20315505963832242,
+      "grad_norm": 0.8887200251742223,
+      "learning_rate": 1.847173790663447e-05,
+      "loss": 0.8448,
+      "step": 1056
+    },
+    {
+      "epoch": 0.20334744132358598,
+      "grad_norm": 0.8537507548603561,
+      "learning_rate": 1.8468425714783206e-05,
+      "loss": 0.9182,
+      "step": 1057
+    },
+    {
+      "epoch": 0.20353982300884957,
+      "grad_norm": 0.9428747655912598,
+      "learning_rate": 1.84651102351975e-05,
+      "loss": 0.8822,
+      "step": 1058
+    },
+    {
+      "epoch": 0.20373220469411313,
+      "grad_norm": 0.916802165983943,
+      "learning_rate": 1.846179146916454e-05,
+      "loss": 0.8605,
+      "step": 1059
+    },
+    {
+      "epoch": 0.2039245863793767,
+      "grad_norm": 0.8627838117966022,
+      "learning_rate": 1.8458469417972783e-05,
+      "loss": 0.3417,
+      "step": 1060
+    },
+    {
+      "epoch": 0.20411696806464025,
+      "grad_norm": 0.8596894653629039,
+      "learning_rate": 1.8455144082911965e-05,
+      "loss": 0.865,
+      "step": 1061
+    },
+    {
+      "epoch": 0.2043093497499038,
+      "grad_norm": 0.917372166669498,
+      "learning_rate": 1.8451815465273097e-05,
+      "loss": 0.8702,
+      "step": 1062
+    },
+    {
+      "epoch": 0.20450173143516737,
+      "grad_norm": 0.8837235154553809,
+      "learning_rate": 1.8448483566348456e-05,
+      "loss": 0.8743,
+      "step": 1063
+    },
+    {
+      "epoch": 0.20469411312043093,
+      "grad_norm": 0.9165097729244109,
+      "learning_rate": 1.8445148387431605e-05,
+      "loss": 0.8798,
+      "step": 1064
+    },
+    {
+      "epoch": 0.20488649480569449,
+      "grad_norm": 0.8468923840347553,
+      "learning_rate": 1.8441809929817382e-05,
+      "loss": 0.9004,
+      "step": 1065
+    },
+    {
+      "epoch": 0.20507887649095807,
+      "grad_norm": 0.9237260513052958,
+      "learning_rate": 1.8438468194801876e-05,
+      "loss": 0.8426,
+      "step": 1066
+    },
+    {
+      "epoch": 0.20527125817622163,
+      "grad_norm": 0.979395288879235,
+      "learning_rate": 1.8435123183682475e-05,
+      "loss": 0.8225,
+      "step": 1067
+    },
+    {
+      "epoch": 0.2054636398614852,
+      "grad_norm": 0.8969698745847375,
+      "learning_rate": 1.8431774897757824e-05,
+      "loss": 0.8757,
+      "step": 1068
+    },
+    {
+      "epoch": 0.20565602154674875,
+      "grad_norm": 0.9248933899585544,
+      "learning_rate": 1.8428423338327847e-05,
+      "loss": 0.9565,
+      "step": 1069
+    },
+    {
+      "epoch": 0.2058484032320123,
+      "grad_norm": 0.8858750250220245,
+      "learning_rate": 1.8425068506693727e-05,
+      "loss": 0.8511,
+      "step": 1070
+    },
+    {
+      "epoch": 0.20604078491727587,
+      "grad_norm": 0.9612062781802696,
+      "learning_rate": 1.842171040415793e-05,
+      "loss": 0.8598,
+      "step": 1071
+    },
+    {
+      "epoch": 0.20623316660253943,
+      "grad_norm": 0.9106780437651628,
+      "learning_rate": 1.8418349032024185e-05,
+      "loss": 0.9109,
+      "step": 1072
+    },
+    {
+      "epoch": 0.206425548287803,
+      "grad_norm": 0.8717916939636734,
+      "learning_rate": 1.8414984391597492e-05,
+      "loss": 0.8694,
+      "step": 1073
+    },
+    {
+      "epoch": 0.20661792997306658,
+      "grad_norm": 0.925408294885155,
+      "learning_rate": 1.8411616484184126e-05,
+      "loss": 0.847,
+      "step": 1074
+    },
+    {
+      "epoch": 0.20681031165833014,
+      "grad_norm": 0.917594675666447,
+      "learning_rate": 1.8408245311091618e-05,
+      "loss": 0.8357,
+      "step": 1075
+    },
+    {
+      "epoch": 0.2070026933435937,
+      "grad_norm": 0.948454645627743,
+      "learning_rate": 1.8404870873628774e-05,
+      "loss": 0.9067,
+      "step": 1076
+    },
+    {
+      "epoch": 0.20719507502885726,
+      "grad_norm": 0.8184151140741305,
+      "learning_rate": 1.8401493173105675e-05,
+      "loss": 0.8379,
+      "step": 1077
+    },
+    {
+      "epoch": 0.20738745671412082,
+      "grad_norm": 0.8753913307186019,
+      "learning_rate": 1.8398112210833648e-05,
+      "loss": 0.8571,
+      "step": 1078
+    },
+    {
+      "epoch": 0.20757983839938438,
+      "grad_norm": 0.8866615797385177,
+      "learning_rate": 1.8394727988125308e-05,
+      "loss": 0.8819,
+      "step": 1079
+    },
+    {
+      "epoch": 0.20777222008464794,
+      "grad_norm": 0.8673527826477565,
+      "learning_rate": 1.8391340506294524e-05,
+      "loss": 0.8223,
+      "step": 1080
+    },
+    {
+      "epoch": 0.2079646017699115,
+      "grad_norm": 0.9433935956507706,
+      "learning_rate": 1.8387949766656434e-05,
+      "loss": 0.8794,
+      "step": 1081
+    },
+    {
+      "epoch": 0.20815698345517505,
+      "grad_norm": 1.0271673111613704,
+      "learning_rate": 1.8384555770527438e-05,
+      "loss": 0.8697,
+      "step": 1082
+    },
+    {
+      "epoch": 0.20834936514043864,
+      "grad_norm": 0.8931491463607244,
+      "learning_rate": 1.8381158519225204e-05,
+      "loss": 0.8807,
+      "step": 1083
+    },
+    {
+      "epoch": 0.2085417468257022,
+      "grad_norm": 0.9120864971489071,
+      "learning_rate": 1.8377758014068662e-05,
+      "loss": 0.9008,
+      "step": 1084
+    },
+    {
+      "epoch": 0.20873412851096576,
+      "grad_norm": 0.7913186851831782,
+      "learning_rate": 1.8374354256378e-05,
+      "loss": 0.3721,
+      "step": 1085
+    },
+    {
+      "epoch": 0.20892651019622932,
+      "grad_norm": 0.9368516045577856,
+      "learning_rate": 1.837094724747468e-05,
+      "loss": 0.8734,
+      "step": 1086
+    },
+    {
+      "epoch": 0.20911889188149288,
+      "grad_norm": 1.0348029365388791,
+      "learning_rate": 1.8367536988681422e-05,
+      "loss": 0.92,
+      "step": 1087
+    },
+    {
+      "epoch": 0.20931127356675644,
+      "grad_norm": 0.8755107531757421,
+      "learning_rate": 1.83641234813222e-05,
+      "loss": 0.8624,
+      "step": 1088
+    },
+    {
+      "epoch": 0.20950365525202,
+      "grad_norm": 0.9627011133285763,
+      "learning_rate": 1.8360706726722253e-05,
+      "loss": 0.7853,
+      "step": 1089
+    },
+    {
+      "epoch": 0.20969603693728356,
+      "grad_norm": 1.01247604543826,
+      "learning_rate": 1.835728672620809e-05,
+      "loss": 0.9209,
+      "step": 1090
+    },
+    {
+      "epoch": 0.20988841862254715,
+      "grad_norm": 0.9228638580723751,
+      "learning_rate": 1.8353863481107473e-05,
+      "loss": 0.8406,
+      "step": 1091
+    },
+    {
+      "epoch": 0.2100808003078107,
+      "grad_norm": 1.0492495959621204,
+      "learning_rate": 1.835043699274942e-05,
+      "loss": 0.8739,
+      "step": 1092
+    },
+    {
+      "epoch": 0.21027318199307427,
+      "grad_norm": 1.033667396060039,
+      "learning_rate": 1.8347007262464206e-05,
+      "loss": 0.8145,
+      "step": 1093
+    },
+    {
+      "epoch": 0.21046556367833782,
+      "grad_norm": 1.0067567826243082,
+      "learning_rate": 1.8343574291583385e-05,
+      "loss": 0.8197,
+      "step": 1094
+    },
+    {
+      "epoch": 0.21065794536360138,
+      "grad_norm": 0.9025391683329841,
+      "learning_rate": 1.8340138081439743e-05,
+      "loss": 0.8677,
+      "step": 1095
+    },
+    {
+      "epoch": 0.21085032704886494,
+      "grad_norm": 0.8839997713291201,
+      "learning_rate": 1.833669863336734e-05,
+      "loss": 0.8332,
+      "step": 1096
+    },
+    {
+      "epoch": 0.2110427087341285,
+      "grad_norm": 0.9714104061195746,
+      "learning_rate": 1.833325594870148e-05,
+      "loss": 0.9051,
+      "step": 1097
+    },
+    {
+      "epoch": 0.21123509041939206,
+      "grad_norm": 0.9656223190328947,
+      "learning_rate": 1.8329810028778747e-05,
+      "loss": 0.8906,
+      "step": 1098
+    },
+    {
+      "epoch": 0.21142747210465565,
+      "grad_norm": 0.9821324963499407,
+      "learning_rate": 1.8326360874936952e-05,
+      "loss": 0.8823,
+      "step": 1099
+    },
+    {
+      "epoch": 0.2116198537899192,
+      "grad_norm": 0.9862525161080206,
+      "learning_rate": 1.8322908488515182e-05,
+      "loss": 0.9067,
+      "step": 1100
+    },
+    {
+      "epoch": 0.21181223547518277,
+      "grad_norm": 0.9324437281870467,
+      "learning_rate": 1.8319452870853772e-05,
+      "loss": 0.9114,
+      "step": 1101
+    },
+    {
+      "epoch": 0.21200461716044633,
+      "grad_norm": 0.8884282982859285,
+      "learning_rate": 1.8315994023294306e-05,
+      "loss": 0.8729,
+      "step": 1102
+    },
+    {
+      "epoch": 0.2121969988457099,
+      "grad_norm": 0.8896823739263302,
+      "learning_rate": 1.8312531947179634e-05,
+      "loss": 0.9164,
+      "step": 1103
+    },
+    {
+      "epoch": 0.21238938053097345,
+      "grad_norm": 0.9356604664448148,
+      "learning_rate": 1.8309066643853854e-05,
+      "loss": 0.8862,
+      "step": 1104
+    },
+    {
+      "epoch": 0.212581762216237,
+      "grad_norm": 0.9509687847358337,
+      "learning_rate": 1.8305598114662312e-05,
+      "loss": 0.9235,
+      "step": 1105
+    },
+    {
+      "epoch": 0.21277414390150057,
+      "grad_norm": 0.9658866439498481,
+      "learning_rate": 1.830212636095161e-05,
+      "loss": 0.8699,
+      "step": 1106
+    },
+    {
+      "epoch": 0.21296652558676413,
+      "grad_norm": 0.7126070184631219,
+      "learning_rate": 1.8298651384069605e-05,
+      "loss": 0.3378,
+      "step": 1107
+    },
+    {
+      "epoch": 0.21315890727202771,
+      "grad_norm": 0.936781247769052,
+      "learning_rate": 1.8295173185365405e-05,
+      "loss": 0.855,
+      "step": 1108
+    },
+    {
+      "epoch": 0.21335128895729127,
+      "grad_norm": 0.9618469251642381,
+      "learning_rate": 1.829169176618936e-05,
+      "loss": 0.8669,
+      "step": 1109
+    },
+    {
+      "epoch": 0.21354367064255483,
+      "grad_norm": 0.9176704322272974,
+      "learning_rate": 1.828820712789308e-05,
+      "loss": 0.8561,
+      "step": 1110
+    },
+    {
+      "epoch": 0.2137360523278184,
+      "grad_norm": 0.9831773504433754,
+      "learning_rate": 1.828471927182942e-05,
+      "loss": 0.879,
+      "step": 1111
+    },
+    {
+      "epoch": 0.21392843401308195,
+      "grad_norm": 0.9550991481213037,
+      "learning_rate": 1.828122819935249e-05,
+      "loss": 0.8389,
+      "step": 1112
+    },
+    {
+      "epoch": 0.2141208156983455,
+      "grad_norm": 1.0155252309795229,
+      "learning_rate": 1.8277733911817642e-05,
+      "loss": 0.8271,
+      "step": 1113
+    },
+    {
+      "epoch": 0.21431319738360907,
+      "grad_norm": 0.9750757365307605,
+      "learning_rate": 1.8274236410581478e-05,
+      "loss": 0.8981,
+      "step": 1114
+    },
+    {
+      "epoch": 0.21450557906887263,
+      "grad_norm": 1.0805363121583078,
+      "learning_rate": 1.827073569700185e-05,
+      "loss": 0.8875,
+      "step": 1115
+    },
+    {
+      "epoch": 0.21469796075413622,
+      "grad_norm": 1.0037273433697196,
+      "learning_rate": 1.8267231772437854e-05,
+      "loss": 0.8978,
+      "step": 1116
+    },
+    {
+      "epoch": 0.21489034243939978,
+      "grad_norm": 0.8712283896862058,
+      "learning_rate": 1.8263724638249834e-05,
+      "loss": 0.8501,
+      "step": 1117
+    },
+    {
+      "epoch": 0.21508272412466334,
+      "grad_norm": 0.9300461042353921,
+      "learning_rate": 1.8260214295799382e-05,
+      "loss": 0.8711,
+      "step": 1118
+    },
+    {
+      "epoch": 0.2152751058099269,
+      "grad_norm": 0.698264244019554,
+      "learning_rate": 1.825670074644933e-05,
+      "loss": 0.3451,
+      "step": 1119
+    },
+    {
+      "epoch": 0.21546748749519046,
+      "grad_norm": 0.9464591220192733,
+      "learning_rate": 1.8253183991563768e-05,
+      "loss": 0.8813,
+      "step": 1120
+    },
+    {
+      "epoch": 0.21565986918045402,
+      "grad_norm": 0.9876271095153261,
+      "learning_rate": 1.824966403250801e-05,
+      "loss": 0.9003,
+      "step": 1121
+    },
+    {
+      "epoch": 0.21585225086571758,
+      "grad_norm": 0.6287230371297933,
+      "learning_rate": 1.8246140870648633e-05,
+      "loss": 0.3562,
+      "step": 1122
+    },
+    {
+      "epoch": 0.21604463255098114,
+      "grad_norm": 0.9247983477399722,
+      "learning_rate": 1.8242614507353446e-05,
+      "loss": 0.8879,
+      "step": 1123
+    },
+    {
+      "epoch": 0.2162370142362447,
+      "grad_norm": 0.9372779933466898,
+      "learning_rate": 1.8239084943991507e-05,
+      "loss": 0.8403,
+      "step": 1124
+    },
+    {
+      "epoch": 0.21642939592150828,
+      "grad_norm": 1.0017352181951724,
+      "learning_rate": 1.823555218193311e-05,
+      "loss": 0.8888,
+      "step": 1125
+    },
+    {
+      "epoch": 0.21662177760677184,
+      "grad_norm": 0.6974369470227099,
+      "learning_rate": 1.8232016222549797e-05,
+      "loss": 0.3873,
+      "step": 1126
+    },
+    {
+      "epoch": 0.2168141592920354,
+      "grad_norm": 0.9318212094803401,
+      "learning_rate": 1.8228477067214352e-05,
+      "loss": 0.8869,
+      "step": 1127
+    },
+    {
+      "epoch": 0.21700654097729896,
+      "grad_norm": 1.0453861919095235,
+      "learning_rate": 1.8224934717300794e-05,
+      "loss": 0.8955,
+      "step": 1128
+    },
+    {
+      "epoch": 0.21719892266256252,
+      "grad_norm": 0.9373241376013453,
+      "learning_rate": 1.8221389174184385e-05,
+      "loss": 0.8062,
+      "step": 1129
+    },
+    {
+      "epoch": 0.21739130434782608,
+      "grad_norm": 0.8541642743965372,
+      "learning_rate": 1.8217840439241633e-05,
+      "loss": 0.9079,
+      "step": 1130
+    },
+    {
+      "epoch": 0.21758368603308964,
+      "grad_norm": 1.0056621588974641,
+      "learning_rate": 1.8214288513850267e-05,
+      "loss": 0.8449,
+      "step": 1131
+    },
+    {
+      "epoch": 0.2177760677183532,
+      "grad_norm": 0.7847781494200494,
+      "learning_rate": 1.8210733399389277e-05,
+      "loss": 0.8241,
+      "step": 1132
+    },
+    {
+      "epoch": 0.2179684494036168,
+      "grad_norm": 0.9203924434877772,
+      "learning_rate": 1.820717509723888e-05,
+      "loss": 0.8704,
+      "step": 1133
+    },
+    {
+      "epoch": 0.21816083108888035,
+      "grad_norm": 0.9887238895363091,
+      "learning_rate": 1.8203613608780525e-05,
+      "loss": 0.9296,
+      "step": 1134
+    },
+    {
+      "epoch": 0.2183532127741439,
+      "grad_norm": 0.9546208257204792,
+      "learning_rate": 1.8200048935396908e-05,
+      "loss": 0.8309,
+      "step": 1135
+    },
+    {
+      "epoch": 0.21854559445940747,
+      "grad_norm": 0.8924717140220111,
+      "learning_rate": 1.819648107847196e-05,
+      "loss": 0.8922,
+      "step": 1136
+    },
+    {
+      "epoch": 0.21873797614467103,
+      "grad_norm": 0.92325557480455,
+      "learning_rate": 1.8192910039390844e-05,
+      "loss": 0.9022,
+      "step": 1137
+    },
+    {
+      "epoch": 0.21893035782993459,
+      "grad_norm": 0.7787857601462344,
+      "learning_rate": 1.8189335819539963e-05,
+      "loss": 0.8282,
+      "step": 1138
+    },
+    {
+      "epoch": 0.21912273951519814,
+      "grad_norm": 0.9419978978487937,
+      "learning_rate": 1.8185758420306947e-05,
+      "loss": 0.9229,
+      "step": 1139
+    },
+    {
+      "epoch": 0.2193151212004617,
+      "grad_norm": 0.6384651357991911,
+      "learning_rate": 1.818217784308067e-05,
+      "loss": 0.3471,
+      "step": 1140
+    },
+    {
+      "epoch": 0.2195075028857253,
+      "grad_norm": 1.0013875800941123,
+      "learning_rate": 1.817859408925123e-05,
+      "loss": 0.8444,
+      "step": 1141
+    },
+    {
+      "epoch": 0.21969988457098885,
+      "grad_norm": 0.8772900330196058,
+      "learning_rate": 1.817500716020997e-05,
+      "loss": 0.7967,
+      "step": 1142
+    },
+    {
+      "epoch": 0.2198922662562524,
+      "grad_norm": 0.9866699861092602,
+      "learning_rate": 1.8171417057349457e-05,
+      "loss": 0.8772,
+      "step": 1143
+    },
+    {
+      "epoch": 0.22008464794151597,
+      "grad_norm": 0.9237088348831368,
+      "learning_rate": 1.816782378206349e-05,
+      "loss": 0.9496,
+      "step": 1144
+    },
+    {
+      "epoch": 0.22027702962677953,
+      "grad_norm": 1.0054172600076743,
+      "learning_rate": 1.8164227335747108e-05,
+      "loss": 0.901,
+      "step": 1145
+    },
+    {
+      "epoch": 0.2204694113120431,
+      "grad_norm": 0.9396604272665069,
+      "learning_rate": 1.8160627719796568e-05,
+      "loss": 0.9125,
+      "step": 1146
+    },
+    {
+      "epoch": 0.22066179299730665,
+      "grad_norm": 0.8671689305200011,
+      "learning_rate": 1.815702493560937e-05,
+      "loss": 0.893,
+      "step": 1147
+    },
+    {
+      "epoch": 0.2208541746825702,
+      "grad_norm": 0.9516756447758423,
+      "learning_rate": 1.8153418984584238e-05,
+      "loss": 0.8927,
+      "step": 1148
+    },
+    {
+      "epoch": 0.22104655636783377,
+      "grad_norm": 0.9534757788491468,
+      "learning_rate": 1.8149809868121125e-05,
+      "loss": 0.8602,
+      "step": 1149
+    },
+    {
+      "epoch": 0.22123893805309736,
+      "grad_norm": 0.9712358622610803,
+      "learning_rate": 1.8146197587621217e-05,
+      "loss": 0.9066,
+      "step": 1150
+    },
+    {
+      "epoch": 0.22143131973836092,
+      "grad_norm": 0.867896496762974,
+      "learning_rate": 1.814258214448692e-05,
+      "loss": 0.8905,
+      "step": 1151
+    },
+    {
+      "epoch": 0.22162370142362448,
+      "grad_norm": 0.8682058065772847,
+      "learning_rate": 1.8138963540121878e-05,
+      "loss": 0.837,
+      "step": 1152
+    },
+    {
+      "epoch": 0.22181608310888803,
+      "grad_norm": 0.8816243364823496,
+      "learning_rate": 1.813534177593096e-05,
+      "loss": 0.8808,
+      "step": 1153
+    },
+    {
+      "epoch": 0.2220084647941516,
+      "grad_norm": 0.9517314118235902,
+      "learning_rate": 1.8131716853320254e-05,
+      "loss": 0.8527,
+      "step": 1154
+    },
+    {
+      "epoch": 0.22220084647941515,
+      "grad_norm": 0.9687557268310836,
+      "learning_rate": 1.8128088773697086e-05,
+      "loss": 0.8831,
+      "step": 1155
+    },
+    {
+      "epoch": 0.2223932281646787,
+      "grad_norm": 0.912758822856774,
+      "learning_rate": 1.8124457538469996e-05,
+      "loss": 0.8142,
+      "step": 1156
+    },
+    {
+      "epoch": 0.22258560984994227,
+      "grad_norm": 0.975173757334746,
+      "learning_rate": 1.8120823149048753e-05,
+      "loss": 0.8869,
+      "step": 1157
+    },
+    {
+      "epoch": 0.22277799153520586,
+      "grad_norm": 0.6622712227318327,
+      "learning_rate": 1.811718560684436e-05,
+      "loss": 0.3617,
+      "step": 1158
+    },
+    {
+      "epoch": 0.22297037322046942,
+      "grad_norm": 0.8861184978764818,
+      "learning_rate": 1.8113544913269025e-05,
+      "loss": 0.9046,
+      "step": 1159
+    },
+    {
+      "epoch": 0.22316275490573298,
+      "grad_norm": 0.9744004285877005,
+      "learning_rate": 1.8109901069736202e-05,
+      "loss": 0.8632,
+      "step": 1160
+    },
+    {
+      "epoch": 0.22335513659099654,
+      "grad_norm": 0.9542925048251781,
+      "learning_rate": 1.8106254077660552e-05,
+      "loss": 0.8521,
+      "step": 1161
+    },
+    {
+      "epoch": 0.2235475182762601,
+      "grad_norm": 0.9309158872093203,
+      "learning_rate": 1.810260393845796e-05,
+      "loss": 0.8926,
+      "step": 1162
+    },
+    {
+      "epoch": 0.22373989996152366,
+      "grad_norm": 0.9097764201698093,
+      "learning_rate": 1.809895065354554e-05,
+      "loss": 0.8701,
+      "step": 1163
+    },
+    {
+      "epoch": 0.22393228164678722,
+      "grad_norm": 0.936742691052022,
+      "learning_rate": 1.8095294224341622e-05,
+      "loss": 0.8711,
+      "step": 1164
+    },
+    {
+      "epoch": 0.22412466333205078,
+      "grad_norm": 0.9293441642805539,
+      "learning_rate": 1.8091634652265755e-05,
+      "loss": 0.8377,
+      "step": 1165
+    },
+    {
+      "epoch": 0.22431704501731436,
+      "grad_norm": 0.8862550157989745,
+      "learning_rate": 1.8087971938738715e-05,
+      "loss": 0.8503,
+      "step": 1166
+    },
+    {
+      "epoch": 0.22450942670257792,
+      "grad_norm": 0.9633523520171046,
+      "learning_rate": 1.808430608518249e-05,
+      "loss": 0.8679,
+      "step": 1167
+    },
+    {
+      "epoch": 0.22470180838784148,
+      "grad_norm": 0.8878750292303486,
+      "learning_rate": 1.808063709302029e-05,
+      "loss": 0.8702,
+      "step": 1168
+    },
+    {
+      "epoch": 0.22489419007310504,
+      "grad_norm": 0.8306485690283155,
+      "learning_rate": 1.807696496367655e-05,
+      "loss": 0.8911,
+      "step": 1169
+    },
+    {
+      "epoch": 0.2250865717583686,
+      "grad_norm": 0.9316234610790993,
+      "learning_rate": 1.8073289698576913e-05,
+      "loss": 0.8568,
+      "step": 1170
+    },
+    {
+      "epoch": 0.22527895344363216,
+      "grad_norm": 0.8677664004185213,
+      "learning_rate": 1.8069611299148236e-05,
+      "loss": 0.8511,
+      "step": 1171
+    },
+    {
+      "epoch": 0.22547133512889572,
+      "grad_norm": 0.6453990754958097,
+      "learning_rate": 1.8065929766818617e-05,
+      "loss": 0.3174,
+      "step": 1172
+    },
+    {
+      "epoch": 0.22566371681415928,
+      "grad_norm": 0.9096454370404886,
+      "learning_rate": 1.806224510301734e-05,
+      "loss": 0.8705,
+      "step": 1173
+    },
+    {
+      "epoch": 0.22585609849942284,
+      "grad_norm": 0.9290015344549541,
+      "learning_rate": 1.8058557309174926e-05,
+      "loss": 0.8355,
+      "step": 1174
+    },
+    {
+      "epoch": 0.22604848018468643,
+      "grad_norm": 0.9396036515859738,
+      "learning_rate": 1.8054866386723096e-05,
+      "loss": 0.8813,
+      "step": 1175
+    },
+    {
+      "epoch": 0.22624086186995,
+      "grad_norm": 0.979336401210774,
+      "learning_rate": 1.80511723370948e-05,
+      "loss": 0.926,
+      "step": 1176
+    },
+    {
+      "epoch": 0.22643324355521355,
+      "grad_norm": 0.8256136962485165,
+      "learning_rate": 1.804747516172419e-05,
+      "loss": 0.8301,
+      "step": 1177
+    },
+    {
+      "epoch": 0.2266256252404771,
+      "grad_norm": 0.8787481373572503,
+      "learning_rate": 1.8043774862046644e-05,
+      "loss": 0.8883,
+      "step": 1178
+    },
+    {
+      "epoch": 0.22681800692574067,
+      "grad_norm": 0.9966401832513306,
+      "learning_rate": 1.804007143949874e-05,
+      "loss": 0.7872,
+      "step": 1179
+    },
+    {
+      "epoch": 0.22701038861100423,
+      "grad_norm": 0.9442973240461491,
+      "learning_rate": 1.8036364895518272e-05,
+      "loss": 0.8155,
+      "step": 1180
+    },
+    {
+      "epoch": 0.2272027702962678,
+      "grad_norm": 1.1145260917382283,
+      "learning_rate": 1.8032655231544253e-05,
+      "loss": 0.8804,
+      "step": 1181
+    },
+    {
+      "epoch": 0.22739515198153135,
+      "grad_norm": 0.9863465451512602,
+      "learning_rate": 1.8028942449016903e-05,
+      "loss": 0.8512,
+      "step": 1182
+    },
+    {
+      "epoch": 0.22758753366679493,
+      "grad_norm": 0.973402576948008,
+      "learning_rate": 1.8025226549377647e-05,
+      "loss": 0.8717,
+      "step": 1183
+    },
+    {
+      "epoch": 0.2277799153520585,
+      "grad_norm": 0.8974140297852222,
+      "learning_rate": 1.8021507534069133e-05,
+      "loss": 0.8062,
+      "step": 1184
+    },
+    {
+      "epoch": 0.22797229703732205,
+      "grad_norm": 1.0072989181566243,
+      "learning_rate": 1.8017785404535198e-05,
+      "loss": 0.8957,
+      "step": 1185
+    },
+    {
+      "epoch": 0.2281646787225856,
+      "grad_norm": 0.9451069703131727,
+      "learning_rate": 1.8014060162220916e-05,
+      "loss": 0.8589,
+      "step": 1186
+    },
+    {
+      "epoch": 0.22835706040784917,
+      "grad_norm": 0.8649594576584796,
+      "learning_rate": 1.801033180857254e-05,
+      "loss": 0.8947,
+      "step": 1187
+    },
+    {
+      "epoch": 0.22854944209311273,
+      "grad_norm": 0.9681611259841009,
+      "learning_rate": 1.8006600345037558e-05,
+      "loss": 0.8434,
+      "step": 1188
+    },
+    {
+      "epoch": 0.2287418237783763,
+      "grad_norm": 0.9496249989295624,
+      "learning_rate": 1.8002865773064644e-05,
+      "loss": 0.8761,
+      "step": 1189
+    },
+    {
+      "epoch": 0.22893420546363985,
+      "grad_norm": 0.9117321563656284,
+      "learning_rate": 1.799912809410369e-05,
+      "loss": 0.8767,
+      "step": 1190
+    },
+    {
+      "epoch": 0.22912658714890344,
+      "grad_norm": 0.9548433966132102,
+      "learning_rate": 1.799538730960579e-05,
+      "loss": 0.8839,
+      "step": 1191
+    },
+    {
+      "epoch": 0.229318968834167,
+      "grad_norm": 0.8897204417705885,
+      "learning_rate": 1.799164342102325e-05,
+      "loss": 0.8618,
+      "step": 1192
+    },
+    {
+      "epoch": 0.22951135051943056,
+      "grad_norm": 0.7752193039646431,
+      "learning_rate": 1.7987896429809573e-05,
+      "loss": 0.3451,
+      "step": 1193
+    },
+    {
+      "epoch": 0.22970373220469412,
+      "grad_norm": 0.970206967755609,
+      "learning_rate": 1.798414633741947e-05,
+      "loss": 0.7992,
+      "step": 1194
+    },
+    {
+      "epoch": 0.22989611388995768,
+      "grad_norm": 0.8885030799369101,
+      "learning_rate": 1.7980393145308857e-05,
+      "loss": 0.9078,
+      "step": 1195
+    },
+    {
+      "epoch": 0.23008849557522124,
+      "grad_norm": 1.0336392684350968,
+      "learning_rate": 1.797663685493485e-05,
+      "loss": 0.8775,
+      "step": 1196
+    },
+    {
+      "epoch": 0.2302808772604848,
+      "grad_norm": 0.9001446424314463,
+      "learning_rate": 1.7972877467755777e-05,
+      "loss": 0.8532,
+      "step": 1197
+    },
+    {
+      "epoch": 0.23047325894574835,
+      "grad_norm": 0.9447888237125295,
+      "learning_rate": 1.7969114985231152e-05,
+      "loss": 0.9284,
+      "step": 1198
+    },
+    {
+      "epoch": 0.23066564063101191,
+      "grad_norm": 0.8428375356696568,
+      "learning_rate": 1.796534940882171e-05,
+      "loss": 0.7916,
+      "step": 1199
+    },
+    {
+      "epoch": 0.2308580223162755,
+      "grad_norm": 0.890495239255743,
+      "learning_rate": 1.7961580739989365e-05,
+      "loss": 0.9123,
+      "step": 1200
+    },
+    {
+      "epoch": 0.23105040400153906,
+      "grad_norm": 0.8773104071960834,
+      "learning_rate": 1.795780898019726e-05,
+      "loss": 0.9096,
+      "step": 1201
+    },
+    {
+      "epoch": 0.23124278568680262,
+      "grad_norm": 0.8747609688025001,
+      "learning_rate": 1.795403413090971e-05,
+      "loss": 0.8212,
+      "step": 1202
+    },
+    {
+      "epoch": 0.23143516737206618,
+      "grad_norm": 0.9219425062831278,
+      "learning_rate": 1.7950256193592243e-05,
+      "loss": 0.8641,
+      "step": 1203
+    },
+    {
+      "epoch": 0.23162754905732974,
+      "grad_norm": 0.9167485313881659,
+      "learning_rate": 1.794647516971159e-05,
+      "loss": 0.831,
+      "step": 1204
+    },
+    {
+      "epoch": 0.2318199307425933,
+      "grad_norm": 0.9331405630348729,
+      "learning_rate": 1.7942691060735666e-05,
+      "loss": 0.8596,
+      "step": 1205
+    },
+    {
+      "epoch": 0.23201231242785686,
+      "grad_norm": 1.0090975281884678,
+      "learning_rate": 1.79389038681336e-05,
+      "loss": 0.8035,
+      "step": 1206
+    },
+    {
+      "epoch": 0.23220469411312042,
+      "grad_norm": 0.9416091736137084,
+      "learning_rate": 1.7935113593375707e-05,
+      "loss": 0.8549,
+      "step": 1207
+    },
+    {
+      "epoch": 0.232397075798384,
+      "grad_norm": 0.8693340806429757,
+      "learning_rate": 1.7931320237933503e-05,
+      "loss": 0.9212,
+      "step": 1208
+    },
+    {
+      "epoch": 0.23258945748364757,
+      "grad_norm": 0.9375691761749666,
+      "learning_rate": 1.79275238032797e-05,
+      "loss": 0.8468,
+      "step": 1209
+    },
+    {
+      "epoch": 0.23278183916891113,
+      "grad_norm": 0.957037808677727,
+      "learning_rate": 1.7923724290888205e-05,
+      "loss": 0.877,
+      "step": 1210
+    },
+    {
+      "epoch": 0.23297422085417469,
+      "grad_norm": 0.9132090274223479,
+      "learning_rate": 1.791992170223412e-05,
+      "loss": 0.8369,
+      "step": 1211
+    },
+    {
+      "epoch": 0.23316660253943824,
+      "grad_norm": 0.9019120943944012,
+      "learning_rate": 1.791611603879374e-05,
+      "loss": 0.882,
+      "step": 1212
+    },
+    {
+      "epoch": 0.2333589842247018,
+      "grad_norm": 0.9196672315725283,
+      "learning_rate": 1.791230730204455e-05,
+      "loss": 0.8647,
+      "step": 1213
+    },
+    {
+      "epoch": 0.23355136590996536,
+      "grad_norm": 0.8901168718313226,
+      "learning_rate": 1.7908495493465236e-05,
+      "loss": 0.9327,
+      "step": 1214
+    },
+    {
+      "epoch": 0.23374374759522892,
+      "grad_norm": 0.8692230375592941,
+      "learning_rate": 1.7904680614535675e-05,
+      "loss": 0.8271,
+      "step": 1215
+    },
+    {
+      "epoch": 0.2339361292804925,
+      "grad_norm": 0.9057690162670959,
+      "learning_rate": 1.7900862666736935e-05,
+      "loss": 0.8451,
+      "step": 1216
+    },
+    {
+      "epoch": 0.23412851096575607,
+      "grad_norm": 0.9367364006771105,
+      "learning_rate": 1.789704165155127e-05,
+      "loss": 0.853,
+      "step": 1217
+    },
+    {
+      "epoch": 0.23432089265101963,
+      "grad_norm": 0.8772963453337094,
+      "learning_rate": 1.7893217570462134e-05,
+      "loss": 0.853,
+      "step": 1218
+    },
+    {
+      "epoch": 0.2345132743362832,
+      "grad_norm": 0.8824019895280809,
+      "learning_rate": 1.7889390424954168e-05,
+      "loss": 0.8658,
+      "step": 1219
+    },
+    {
+      "epoch": 0.23470565602154675,
+      "grad_norm": 1.1104382654554112,
+      "learning_rate": 1.78855602165132e-05,
+      "loss": 0.8777,
+      "step": 1220
+    },
+    {
+      "epoch": 0.2348980377068103,
+      "grad_norm": 1.0364734787400376,
+      "learning_rate": 1.7881726946626244e-05,
+      "loss": 0.8526,
+      "step": 1221
+    },
+    {
+      "epoch": 0.23509041939207387,
+      "grad_norm": 0.8923434177432201,
+      "learning_rate": 1.787789061678151e-05,
+      "loss": 0.8831,
+      "step": 1222
+    },
+    {
+      "epoch": 0.23528280107733743,
+      "grad_norm": 0.9356705548533087,
+      "learning_rate": 1.78740512284684e-05,
+      "loss": 0.8723,
+      "step": 1223
+    },
+    {
+      "epoch": 0.235475182762601,
+      "grad_norm": 0.8690731911676538,
+      "learning_rate": 1.787020878317749e-05,
+      "loss": 0.8406,
+      "step": 1224
+    },
+    {
+      "epoch": 0.23566756444786457,
+      "grad_norm": 0.932466729747037,
+      "learning_rate": 1.7866363282400555e-05,
+      "loss": 0.8706,
+      "step": 1225
+    },
+    {
+      "epoch": 0.23585994613312813,
+      "grad_norm": 0.9195417014577383,
+      "learning_rate": 1.7862514727630543e-05,
+      "loss": 0.9081,
+      "step": 1226
+    },
+    {
+      "epoch": 0.2360523278183917,
+      "grad_norm": 0.8375616015564664,
+      "learning_rate": 1.7858663120361597e-05,
+      "loss": 0.897,
+      "step": 1227
+    },
+    {
+      "epoch": 0.23624470950365525,
+      "grad_norm": 0.8549908998872318,
+      "learning_rate": 1.785480846208905e-05,
+      "loss": 0.8368,
+      "step": 1228
+    },
+    {
+      "epoch": 0.2364370911889188,
+      "grad_norm": 0.8909381174151334,
+      "learning_rate": 1.7850950754309405e-05,
+      "loss": 0.8433,
+      "step": 1229
+    },
+    {
+      "epoch": 0.23662947287418237,
+      "grad_norm": 0.924858393881658,
+      "learning_rate": 1.7847089998520365e-05,
+      "loss": 0.918,
+      "step": 1230
+    },
+    {
+      "epoch": 0.23682185455944593,
+      "grad_norm": 0.7947720129329198,
+      "learning_rate": 1.7843226196220803e-05,
+      "loss": 0.8239,
+      "step": 1231
+    },
+    {
+      "epoch": 0.2370142362447095,
+      "grad_norm": 0.86267864652118,
+      "learning_rate": 1.783935934891078e-05,
+      "loss": 0.9048,
+      "step": 1232
+    },
+    {
+      "epoch": 0.23720661792997308,
+      "grad_norm": 0.9053363603949303,
+      "learning_rate": 1.7835489458091544e-05,
+      "loss": 0.8501,
+      "step": 1233
+    },
+    {
+      "epoch": 0.23739899961523664,
+      "grad_norm": 0.8672745656018541,
+      "learning_rate": 1.7831616525265515e-05,
+      "loss": 0.9332,
+      "step": 1234
+    },
+    {
+      "epoch": 0.2375913813005002,
+      "grad_norm": 0.8671654915564192,
+      "learning_rate": 1.7827740551936296e-05,
+      "loss": 0.9107,
+      "step": 1235
+    },
+    {
+      "epoch": 0.23778376298576376,
+      "grad_norm": 0.8849636870422262,
+      "learning_rate": 1.7823861539608686e-05,
+      "loss": 0.9357,
+      "step": 1236
+    },
+    {
+      "epoch": 0.23797614467102732,
+      "grad_norm": 0.8827977374518553,
+      "learning_rate": 1.7819979489788638e-05,
+      "loss": 0.8223,
+      "step": 1237
+    },
+    {
+      "epoch": 0.23816852635629088,
+      "grad_norm": 0.9118384989762928,
+      "learning_rate": 1.7816094403983298e-05,
+      "loss": 0.8819,
+      "step": 1238
+    },
+    {
+      "epoch": 0.23836090804155444,
+      "grad_norm": 0.8988422958924583,
+      "learning_rate": 1.7812206283701002e-05,
+      "loss": 0.8734,
+      "step": 1239
+    },
+    {
+      "epoch": 0.238553289726818,
+      "grad_norm": 0.9632493826684722,
+      "learning_rate": 1.7808315130451244e-05,
+      "loss": 0.8183,
+      "step": 1240
+    },
+    {
+      "epoch": 0.23874567141208156,
+      "grad_norm": 0.9056701790152845,
+      "learning_rate": 1.78044209457447e-05,
+      "loss": 0.928,
+      "step": 1241
+    },
+    {
+      "epoch": 0.23893805309734514,
+      "grad_norm": 0.9717857748135311,
+      "learning_rate": 1.7800523731093232e-05,
+      "loss": 0.8582,
+      "step": 1242
+    },
+    {
+      "epoch": 0.2391304347826087,
+      "grad_norm": 0.8731032019092507,
+      "learning_rate": 1.7796623488009875e-05,
+      "loss": 0.8791,
+      "step": 1243
+    },
+    {
+      "epoch": 0.23932281646787226,
+      "grad_norm": 0.8539735501301605,
+      "learning_rate": 1.7792720218008826e-05,
+      "loss": 0.8574,
+      "step": 1244
+    },
+    {
+      "epoch": 0.23951519815313582,
+      "grad_norm": 0.8477449714035654,
+      "learning_rate": 1.7788813922605488e-05,
+      "loss": 0.8519,
+      "step": 1245
+    },
+    {
+      "epoch": 0.23970757983839938,
+      "grad_norm": 0.915782122789951,
+      "learning_rate": 1.7784904603316402e-05,
+      "loss": 0.8508,
+      "step": 1246
+    },
+    {
+      "epoch": 0.23989996152366294,
+      "grad_norm": 0.8725962819554685,
+      "learning_rate": 1.7780992261659305e-05,
+      "loss": 0.8669,
+      "step": 1247
+    },
+    {
+      "epoch": 0.2400923432089265,
+      "grad_norm": 0.9085298007049676,
+      "learning_rate": 1.777707689915311e-05,
+      "loss": 0.8234,
+      "step": 1248
+    },
+    {
+      "epoch": 0.24028472489419006,
+      "grad_norm": 0.8409395634150729,
+      "learning_rate": 1.777315851731789e-05,
+      "loss": 0.8583,
+      "step": 1249
+    },
+    {
+      "epoch": 0.24047710657945365,
+      "grad_norm": 0.8597646285325263,
+      "learning_rate": 1.7769237117674893e-05,
+      "loss": 0.8731,
+      "step": 1250
+    },
+    {
+      "epoch": 0.2406694882647172,
+      "grad_norm": 0.9676320810808983,
+      "learning_rate": 1.7765312701746543e-05,
+      "loss": 0.8888,
+      "step": 1251
+    },
+    {
+      "epoch": 0.24086186994998077,
+      "grad_norm": 0.9168860093045579,
+      "learning_rate": 1.7761385271056436e-05,
+      "loss": 0.9006,
+      "step": 1252
+    },
+    {
+      "epoch": 0.24105425163524433,
+      "grad_norm": 0.8655267662471231,
+      "learning_rate": 1.7757454827129338e-05,
+      "loss": 0.3449,
+      "step": 1253
+    },
+    {
+      "epoch": 0.24124663332050789,
+      "grad_norm": 0.9156429645839868,
+      "learning_rate": 1.7753521371491174e-05,
+      "loss": 0.7859,
+      "step": 1254
+    },
+    {
+      "epoch": 0.24143901500577145,
+      "grad_norm": 0.9922494878361331,
+      "learning_rate": 1.7749584905669057e-05,
+      "loss": 0.8821,
+      "step": 1255
+    },
+    {
+      "epoch": 0.241631396691035,
+      "grad_norm": 1.0111016145679224,
+      "learning_rate": 1.774564543119125e-05,
+      "loss": 0.8943,
+      "step": 1256
+    },
+    {
+      "epoch": 0.24182377837629856,
+      "grad_norm": 0.976300061655159,
+      "learning_rate": 1.7741702949587196e-05,
+      "loss": 0.9044,
+      "step": 1257
+    },
+    {
+      "epoch": 0.24201616006156215,
+      "grad_norm": 0.8977142827764442,
+      "learning_rate": 1.7737757462387507e-05,
+      "loss": 0.9194,
+      "step": 1258
+    },
+    {
+      "epoch": 0.2422085417468257,
+      "grad_norm": 0.9259621919280702,
+      "learning_rate": 1.7733808971123946e-05,
+      "loss": 0.8476,
+      "step": 1259
+    },
+    {
+      "epoch": 0.24240092343208927,
+      "grad_norm": 1.0773955121046375,
+      "learning_rate": 1.7729857477329463e-05,
+      "loss": 0.8552,
+      "step": 1260
+    },
+    {
+      "epoch": 0.24259330511735283,
+      "grad_norm": 0.8301915577301328,
+      "learning_rate": 1.7725902982538162e-05,
+      "loss": 0.846,
+      "step": 1261
+    },
+    {
+      "epoch": 0.2427856868026164,
+      "grad_norm": 0.9366327850575039,
+      "learning_rate": 1.772194548828531e-05,
+      "loss": 0.828,
+      "step": 1262
+    },
+    {
+      "epoch": 0.24297806848787995,
+      "grad_norm": 0.7720750073271803,
+      "learning_rate": 1.7717984996107346e-05,
+      "loss": 0.3592,
+      "step": 1263
+    },
+    {
+      "epoch": 0.2431704501731435,
+      "grad_norm": 0.87872988212691,
+      "learning_rate": 1.771402150754187e-05,
+      "loss": 0.8382,
+      "step": 1264
+    },
+    {
+      "epoch": 0.24336283185840707,
+      "grad_norm": 0.9209047757528778,
+      "learning_rate": 1.7710055024127637e-05,
+      "loss": 0.8727,
+      "step": 1265
+    },
+    {
+      "epoch": 0.24355521354367063,
+      "grad_norm": 0.8326639825189668,
+      "learning_rate": 1.7706085547404582e-05,
+      "loss": 0.8913,
+      "step": 1266
+    },
+    {
+      "epoch": 0.24374759522893422,
+      "grad_norm": 0.8338221800537239,
+      "learning_rate": 1.770211307891379e-05,
+      "loss": 0.8244,
+      "step": 1267
+    },
+    {
+      "epoch": 0.24393997691419778,
+      "grad_norm": 0.8613756960878451,
+      "learning_rate": 1.769813762019751e-05,
+      "loss": 0.8247,
+      "step": 1268
+    },
+    {
+      "epoch": 0.24413235859946134,
+      "grad_norm": 0.9139437172654927,
+      "learning_rate": 1.769415917279915e-05,
+      "loss": 0.8794,
+      "step": 1269
+    },
+    {
+      "epoch": 0.2443247402847249,
+      "grad_norm": 0.957739240621011,
+      "learning_rate": 1.7690177738263284e-05,
+      "loss": 0.9135,
+      "step": 1270
+    },
+    {
+      "epoch": 0.24451712196998845,
+      "grad_norm": 0.8251273621899513,
+      "learning_rate": 1.7686193318135635e-05,
+      "loss": 0.8755,
+      "step": 1271
+    },
+    {
+      "epoch": 0.24470950365525201,
+      "grad_norm": 0.929470830557934,
+      "learning_rate": 1.76822059139631e-05,
+      "loss": 0.8677,
+      "step": 1272
+    },
+    {
+      "epoch": 0.24490188534051557,
+      "grad_norm": 0.9467224892480495,
+      "learning_rate": 1.7678215527293724e-05,
+      "loss": 0.9007,
+      "step": 1273
+    },
+    {
+      "epoch": 0.24509426702577913,
+      "grad_norm": 0.8892496328586916,
+      "learning_rate": 1.767422215967671e-05,
+      "loss": 0.9201,
+      "step": 1274
+    },
+    {
+      "epoch": 0.24528664871104272,
+      "grad_norm": 0.8884943757105267,
+      "learning_rate": 1.767022581266242e-05,
+      "loss": 0.9355,
+      "step": 1275
+    },
+    {
+      "epoch": 0.24547903039630628,
+      "grad_norm": 0.9732675101052116,
+      "learning_rate": 1.766622648780238e-05,
+      "loss": 0.8115,
+      "step": 1276
+    },
+    {
+      "epoch": 0.24567141208156984,
+      "grad_norm": 0.8912674724694419,
+      "learning_rate": 1.766222418664926e-05,
+      "loss": 0.9024,
+      "step": 1277
+    },
+    {
+      "epoch": 0.2458637937668334,
+      "grad_norm": 0.8577441269075269,
+      "learning_rate": 1.765821891075689e-05,
+      "loss": 0.8653,
+      "step": 1278
+    },
+    {
+      "epoch": 0.24605617545209696,
+      "grad_norm": 0.8666040135437711,
+      "learning_rate": 1.7654210661680263e-05,
+      "loss": 0.915,
+      "step": 1279
+    },
+    {
+      "epoch": 0.24624855713736052,
+      "grad_norm": 0.6877053309873518,
+      "learning_rate": 1.765019944097551e-05,
+      "loss": 0.3393,
+      "step": 1280
+    },
+    {
+      "epoch": 0.24644093882262408,
+      "grad_norm": 0.6926288629270706,
+      "learning_rate": 1.7646185250199936e-05,
+      "loss": 0.3224,
+      "step": 1281
+    },
+    {
+      "epoch": 0.24663332050788764,
+      "grad_norm": 1.0046626028539298,
+      "learning_rate": 1.7642168090911976e-05,
+      "loss": 0.9223,
+      "step": 1282
+    },
+    {
+      "epoch": 0.24682570219315123,
+      "grad_norm": 0.9350207824540647,
+      "learning_rate": 1.763814796467124e-05,
+      "loss": 0.8578,
+      "step": 1283
+    },
+    {
+      "epoch": 0.24701808387841478,
+      "grad_norm": 0.9640903829194173,
+      "learning_rate": 1.763412487303847e-05,
+      "loss": 0.9175,
+      "step": 1284
+    },
+    {
+      "epoch": 0.24721046556367834,
+      "grad_norm": 1.0076829598119386,
+      "learning_rate": 1.7630098817575578e-05,
+      "loss": 0.8823,
+      "step": 1285
+    },
+    {
+      "epoch": 0.2474028472489419,
+      "grad_norm": 0.9467976122896563,
+      "learning_rate": 1.762606979984561e-05,
+      "loss": 0.8814,
+      "step": 1286
+    },
+    {
+      "epoch": 0.24759522893420546,
+      "grad_norm": 0.9828290820272959,
+      "learning_rate": 1.7622037821412775e-05,
+      "loss": 0.8062,
+      "step": 1287
+    },
+    {
+      "epoch": 0.24778761061946902,
+      "grad_norm": 0.9340072203880309,
+      "learning_rate": 1.7618002883842426e-05,
+      "loss": 0.8698,
+      "step": 1288
+    },
+    {
+      "epoch": 0.24797999230473258,
+      "grad_norm": 0.8823585373646313,
+      "learning_rate": 1.7613964988701057e-05,
+      "loss": 0.9219,
+      "step": 1289
+    },
+    {
+      "epoch": 0.24817237398999614,
+      "grad_norm": 0.9711784447345366,
+      "learning_rate": 1.7609924137556326e-05,
+      "loss": 0.7958,
+      "step": 1290
+    },
+    {
+      "epoch": 0.2483647556752597,
+      "grad_norm": 0.8837330969400775,
+      "learning_rate": 1.7605880331977022e-05,
+      "loss": 0.901,
+      "step": 1291
+    },
+    {
+      "epoch": 0.2485571373605233,
+      "grad_norm": 0.902968448130955,
+      "learning_rate": 1.76018335735331e-05,
+      "loss": 0.9159,
+      "step": 1292
+    },
+    {
+      "epoch": 0.24874951904578685,
+      "grad_norm": 0.9369031967812395,
+      "learning_rate": 1.7597783863795644e-05,
+      "loss": 0.8398,
+      "step": 1293
+    },
+    {
+      "epoch": 0.2489419007310504,
+      "grad_norm": 0.8887714152053197,
+      "learning_rate": 1.7593731204336895e-05,
+      "loss": 0.9025,
+      "step": 1294
+    },
+    {
+      "epoch": 0.24913428241631397,
+      "grad_norm": 0.8591570024664369,
+      "learning_rate": 1.7589675596730233e-05,
+      "loss": 0.9177,
+      "step": 1295
+    },
+    {
+      "epoch": 0.24932666410157753,
+      "grad_norm": 0.8466617757459987,
+      "learning_rate": 1.758561704255018e-05,
+      "loss": 0.8949,
+      "step": 1296
+    },
+    {
+      "epoch": 0.2495190457868411,
+      "grad_norm": 0.8994257608993943,
+      "learning_rate": 1.7581555543372413e-05,
+      "loss": 0.8593,
+      "step": 1297
+    },
+    {
+      "epoch": 0.24971142747210465,
+      "grad_norm": 0.8806188595187883,
+      "learning_rate": 1.7577491100773744e-05,
+      "loss": 0.832,
+      "step": 1298
+    },
+    {
+      "epoch": 0.2499038091573682,
+      "grad_norm": 0.900781000979358,
+      "learning_rate": 1.7573423716332128e-05,
+      "loss": 0.8547,
+      "step": 1299
+    },
+    {
+      "epoch": 0.2500961908426318,
+      "grad_norm": 0.892256369565147,
+      "learning_rate": 1.7569353391626665e-05,
+      "loss": 0.8807,
+      "step": 1300
+    },
+    {
+      "epoch": 0.2502885725278953,
+      "grad_norm": 0.9378148483861315,
+      "learning_rate": 1.7565280128237595e-05,
+      "loss": 0.3517,
+      "step": 1301
+    },
+    {
+      "epoch": 0.2504809542131589,
+      "grad_norm": 0.9825142968191728,
+      "learning_rate": 1.75612039277463e-05,
+      "loss": 0.8891,
+      "step": 1302
+    },
+    {
+      "epoch": 0.25067333589842244,
+      "grad_norm": 0.8505916863415818,
+      "learning_rate": 1.75571247917353e-05,
+      "loss": 0.8512,
+      "step": 1303
+    },
+    {
+      "epoch": 0.25086571758368603,
+      "grad_norm": 0.6369916651409336,
+      "learning_rate": 1.7553042721788255e-05,
+      "loss": 0.3757,
+      "step": 1304
+    },
+    {
+      "epoch": 0.2510580992689496,
+      "grad_norm": 0.9405393341412137,
+      "learning_rate": 1.754895771948997e-05,
+      "loss": 0.9074,
+      "step": 1305
+    },
+    {
+      "epoch": 0.25125048095421315,
+      "grad_norm": 0.9404529202053448,
+      "learning_rate": 1.754486978642637e-05,
+      "loss": 0.8658,
+      "step": 1306
+    },
+    {
+      "epoch": 0.25144286263947674,
+      "grad_norm": 0.9574994625214353,
+      "learning_rate": 1.7540778924184553e-05,
+      "loss": 0.8972,
+      "step": 1307
+    },
+    {
+      "epoch": 0.25163524432474027,
+      "grad_norm": 0.9248329241044715,
+      "learning_rate": 1.7536685134352717e-05,
+      "loss": 0.9175,
+      "step": 1308
+    },
+    {
+      "epoch": 0.25182762601000386,
+      "grad_norm": 1.0153152356111874,
+      "learning_rate": 1.7532588418520215e-05,
+      "loss": 0.9347,
+      "step": 1309
+    },
+    {
+      "epoch": 0.2520200076952674,
+      "grad_norm": 0.9714910929426758,
+      "learning_rate": 1.7528488778277535e-05,
+      "loss": 0.8212,
+      "step": 1310
+    },
+    {
+      "epoch": 0.252212389380531,
+      "grad_norm": 0.8642941929833182,
+      "learning_rate": 1.75243862152163e-05,
+      "loss": 0.8092,
+      "step": 1311
+    },
+    {
+      "epoch": 0.2524047710657945,
+      "grad_norm": 0.9582309556160071,
+      "learning_rate": 1.752028073092926e-05,
+      "loss": 0.8996,
+      "step": 1312
+    },
+    {
+      "epoch": 0.2525971527510581,
+      "grad_norm": 0.9015509649049903,
+      "learning_rate": 1.7516172327010314e-05,
+      "loss": 0.8981,
+      "step": 1313
+    },
+    {
+      "epoch": 0.2527895344363217,
+      "grad_norm": 0.9318386991190069,
+      "learning_rate": 1.751206100505448e-05,
+      "loss": 0.9308,
+      "step": 1314
+    },
+    {
+      "epoch": 0.2529819161215852,
+      "grad_norm": 0.9699871575402621,
+      "learning_rate": 1.7507946766657914e-05,
+      "loss": 0.8612,
+      "step": 1315
+    },
+    {
+      "epoch": 0.2531742978068488,
+      "grad_norm": 0.8710450149540049,
+      "learning_rate": 1.7503829613417905e-05,
+      "loss": 0.8338,
+      "step": 1316
+    },
+    {
+      "epoch": 0.25336667949211233,
+      "grad_norm": 0.8819924367040325,
+      "learning_rate": 1.749970954693288e-05,
+      "loss": 0.8712,
+      "step": 1317
+    },
+    {
+      "epoch": 0.2535590611773759,
+      "grad_norm": 1.0193697459720803,
+      "learning_rate": 1.7495586568802384e-05,
+      "loss": 0.8766,
+      "step": 1318
+    },
+    {
+      "epoch": 0.25375144286263945,
+      "grad_norm": 0.9795257000971911,
+      "learning_rate": 1.7491460680627105e-05,
+      "loss": 0.87,
+      "step": 1319
+    },
+    {
+      "epoch": 0.25394382454790304,
+      "grad_norm": 0.9548403683997206,
+      "learning_rate": 1.7487331884008845e-05,
+      "loss": 0.9389,
+      "step": 1320
+    },
+    {
+      "epoch": 0.25413620623316663,
+      "grad_norm": 0.8955557574157,
+      "learning_rate": 1.7483200180550554e-05,
+      "loss": 0.8364,
+      "step": 1321
+    },
+    {
+      "epoch": 0.25432858791843016,
+      "grad_norm": 0.9387907758656351,
+      "learning_rate": 1.74790655718563e-05,
+      "loss": 0.3401,
+      "step": 1322
+    },
+    {
+      "epoch": 0.25452096960369375,
+      "grad_norm": 0.8956101795915207,
+      "learning_rate": 1.747492805953128e-05,
+      "loss": 0.8352,
+      "step": 1323
+    },
+    {
+      "epoch": 0.2547133512889573,
+      "grad_norm": 0.9685160034291265,
+      "learning_rate": 1.7470787645181818e-05,
+      "loss": 0.8745,
+      "step": 1324
+    },
+    {
+      "epoch": 0.25490573297422087,
+      "grad_norm": 0.8910428913434121,
+      "learning_rate": 1.7466644330415362e-05,
+      "loss": 0.8362,
+      "step": 1325
+    },
+    {
+      "epoch": 0.2550981146594844,
+      "grad_norm": 0.9839891827859032,
+      "learning_rate": 1.7462498116840496e-05,
+      "loss": 0.9152,
+      "step": 1326
+    },
+    {
+      "epoch": 0.255290496344748,
+      "grad_norm": 1.0168258125352563,
+      "learning_rate": 1.745834900606692e-05,
+      "loss": 0.8163,
+      "step": 1327
+    },
+    {
+      "epoch": 0.2554828780300115,
+      "grad_norm": 0.963119829907931,
+      "learning_rate": 1.7454196999705458e-05,
+      "loss": 0.9118,
+      "step": 1328
+    },
+    {
+      "epoch": 0.2556752597152751,
+      "grad_norm": 0.9208753011922463,
+      "learning_rate": 1.7450042099368066e-05,
+      "loss": 0.8584,
+      "step": 1329
+    },
+    {
+      "epoch": 0.2558676414005387,
+      "grad_norm": 0.9036705079335566,
+      "learning_rate": 1.7445884306667823e-05,
+      "loss": 0.8296,
+      "step": 1330
+    },
+    {
+      "epoch": 0.2560600230858022,
+      "grad_norm": 0.8608303771833808,
+      "learning_rate": 1.7441723623218917e-05,
+      "loss": 0.8352,
+      "step": 1331
+    },
+    {
+      "epoch": 0.2562524047710658,
+      "grad_norm": 0.9205886001835745,
+      "learning_rate": 1.7437560050636678e-05,
+      "loss": 0.8151,
+      "step": 1332
+    },
+    {
+      "epoch": 0.25644478645632934,
+      "grad_norm": 0.9648840049711503,
+      "learning_rate": 1.7433393590537543e-05,
+      "loss": 0.8913,
+      "step": 1333
+    },
+    {
+      "epoch": 0.25663716814159293,
+      "grad_norm": 0.9319685068317297,
+      "learning_rate": 1.7429224244539077e-05,
+      "loss": 0.8909,
+      "step": 1334
+    },
+    {
+      "epoch": 0.25682954982685646,
+      "grad_norm": 0.9152277827085439,
+      "learning_rate": 1.7425052014259965e-05,
+      "loss": 0.9031,
+      "step": 1335
+    },
+    {
+      "epoch": 0.25702193151212005,
+      "grad_norm": 1.0127261480965872,
+      "learning_rate": 1.7420876901320006e-05,
+      "loss": 0.8961,
+      "step": 1336
+    },
+    {
+      "epoch": 0.2572143131973836,
+      "grad_norm": 0.8767968148871966,
+      "learning_rate": 1.7416698907340128e-05,
+      "loss": 0.909,
+      "step": 1337
+    },
+    {
+      "epoch": 0.25740669488264717,
+      "grad_norm": 0.9935235715629275,
+      "learning_rate": 1.741251803394237e-05,
+      "loss": 0.8525,
+      "step": 1338
+    },
+    {
+      "epoch": 0.25759907656791076,
+      "grad_norm": 0.823635253496099,
+      "learning_rate": 1.740833428274989e-05,
+      "loss": 0.8682,
+      "step": 1339
+    },
+    {
+      "epoch": 0.2577914582531743,
+      "grad_norm": 0.8644667032376702,
+      "learning_rate": 1.7404147655386966e-05,
+      "loss": 0.9032,
+      "step": 1340
+    },
+    {
+      "epoch": 0.2579838399384379,
+      "grad_norm": 0.9646118132551428,
+      "learning_rate": 1.739995815347899e-05,
+      "loss": 0.9156,
+      "step": 1341
+    },
+    {
+      "epoch": 0.2581762216237014,
+      "grad_norm": 0.8880588269562915,
+      "learning_rate": 1.739576577865247e-05,
+      "loss": 0.8696,
+      "step": 1342
+    },
+    {
+      "epoch": 0.258368603308965,
+      "grad_norm": 1.0128286006194054,
+      "learning_rate": 1.739157053253503e-05,
+      "loss": 0.8826,
+      "step": 1343
+    },
+    {
+      "epoch": 0.2585609849942285,
+      "grad_norm": 0.8711132520661519,
+      "learning_rate": 1.738737241675541e-05,
+      "loss": 0.9023,
+      "step": 1344
+    },
+    {
+      "epoch": 0.2587533666794921,
+      "grad_norm": 0.9163970747714321,
+      "learning_rate": 1.7383171432943466e-05,
+      "loss": 0.8696,
+      "step": 1345
+    },
+    {
+      "epoch": 0.2589457483647557,
+      "grad_norm": 0.9321833519358579,
+      "learning_rate": 1.737896758273016e-05,
+      "loss": 0.9324,
+      "step": 1346
+    },
+    {
+      "epoch": 0.25913813005001923,
+      "grad_norm": 0.9749182274671203,
+      "learning_rate": 1.7374760867747574e-05,
+      "loss": 0.8394,
+      "step": 1347
+    },
+    {
+      "epoch": 0.2593305117352828,
+      "grad_norm": 0.9520952983536605,
+      "learning_rate": 1.7370551289628895e-05,
+      "loss": 0.8423,
+      "step": 1348
+    },
+    {
+      "epoch": 0.25952289342054635,
+      "grad_norm": 0.9118200633625917,
+      "learning_rate": 1.7366338850008432e-05,
+      "loss": 0.819,
+      "step": 1349
+    },
+    {
+      "epoch": 0.25971527510580994,
+      "grad_norm": 0.9921548239610416,
+      "learning_rate": 1.73621235505216e-05,
+      "loss": 0.9499,
+      "step": 1350
+    },
+    {
+      "epoch": 0.25990765679107347,
+      "grad_norm": 0.8644849457388681,
+      "learning_rate": 1.7357905392804918e-05,
+      "loss": 0.8855,
+      "step": 1351
+    },
+    {
+      "epoch": 0.26010003847633706,
+      "grad_norm": 0.8090162022871645,
+      "learning_rate": 1.735368437849602e-05,
+      "loss": 0.3483,
+      "step": 1352
+    },
+    {
+      "epoch": 0.2602924201616006,
+      "grad_norm": 0.8494514988743438,
+      "learning_rate": 1.7349460509233654e-05,
+      "loss": 0.883,
+      "step": 1353
+    },
+    {
+      "epoch": 0.2604848018468642,
+      "grad_norm": 0.9350986728648616,
+      "learning_rate": 1.734523378665767e-05,
+      "loss": 0.9316,
+      "step": 1354
+    },
+    {
+      "epoch": 0.26067718353212777,
+      "grad_norm": 0.8748673596664809,
+      "learning_rate": 1.7341004212409026e-05,
+      "loss": 0.8235,
+      "step": 1355
+    },
+    {
+      "epoch": 0.2608695652173913,
+      "grad_norm": 0.8652520156018371,
+      "learning_rate": 1.7336771788129785e-05,
+      "loss": 0.9072,
+      "step": 1356
+    },
+    {
+      "epoch": 0.2610619469026549,
+      "grad_norm": 0.9386003175919059,
+      "learning_rate": 1.7332536515463126e-05,
+      "loss": 0.8741,
+      "step": 1357
+    },
+    {
+      "epoch": 0.2612543285879184,
+      "grad_norm": 0.9513103473157519,
+      "learning_rate": 1.7328298396053324e-05,
+      "loss": 0.9084,
+      "step": 1358
+    },
+    {
+      "epoch": 0.261446710273182,
+      "grad_norm": 0.9334978785792942,
+      "learning_rate": 1.7324057431545768e-05,
+      "loss": 0.8795,
+      "step": 1359
+    },
+    {
+      "epoch": 0.26163909195844554,
+      "grad_norm": 0.9196532559682081,
+      "learning_rate": 1.7319813623586935e-05,
+      "loss": 0.9268,
+      "step": 1360
+    },
+    {
+      "epoch": 0.2618314736437091,
+      "grad_norm": 0.9404006047209355,
+      "learning_rate": 1.7315566973824433e-05,
+      "loss": 0.9178,
+      "step": 1361
+    },
+    {
+      "epoch": 0.26202385532897265,
+      "grad_norm": 0.8068420849287683,
+      "learning_rate": 1.7311317483906946e-05,
+      "loss": 0.8974,
+      "step": 1362
+    },
+    {
+      "epoch": 0.26221623701423624,
+      "grad_norm": 0.9275059924227689,
+      "learning_rate": 1.730706515548427e-05,
+      "loss": 0.8882,
+      "step": 1363
+    },
+    {
+      "epoch": 0.26240861869949983,
+      "grad_norm": 0.8388930737439524,
+      "learning_rate": 1.730280999020732e-05,
+      "loss": 0.8699,
+      "step": 1364
+    },
+    {
+      "epoch": 0.26260100038476336,
+      "grad_norm": 0.9009539357425607,
+      "learning_rate": 1.729855198972808e-05,
+      "loss": 0.8408,
+      "step": 1365
+    },
+    {
+      "epoch": 0.26279338207002695,
+      "grad_norm": 0.8047871052540713,
+      "learning_rate": 1.729429115569967e-05,
+      "loss": 0.8542,
+      "step": 1366
+    },
+    {
+      "epoch": 0.2629857637552905,
+      "grad_norm": 0.8494977708938227,
+      "learning_rate": 1.729002748977628e-05,
+      "loss": 0.8808,
+      "step": 1367
+    },
+    {
+      "epoch": 0.26317814544055407,
+      "grad_norm": 0.8787143741312471,
+      "learning_rate": 1.7285760993613215e-05,
+      "loss": 0.8649,
+      "step": 1368
+    },
+    {
+      "epoch": 0.2633705271258176,
+      "grad_norm": 0.9158946339894979,
+      "learning_rate": 1.7281491668866874e-05,
+      "loss": 0.8615,
+      "step": 1369
+    },
+    {
+      "epoch": 0.2635629088110812,
+      "grad_norm": 0.9038240806682061,
+      "learning_rate": 1.727721951719476e-05,
+      "loss": 0.8776,
+      "step": 1370
+    },
+    {
+      "epoch": 0.2637552904963448,
+      "grad_norm": 0.9539518044629862,
+      "learning_rate": 1.7272944540255468e-05,
+      "loss": 0.9063,
+      "step": 1371
+    },
+    {
+      "epoch": 0.2639476721816083,
+      "grad_norm": 0.8823376208887337,
+      "learning_rate": 1.726866673970869e-05,
+      "loss": 0.8764,
+      "step": 1372
+    },
+    {
+      "epoch": 0.2641400538668719,
+      "grad_norm": 0.9720871255146147,
+      "learning_rate": 1.7264386117215216e-05,
+      "loss": 0.8204,
+      "step": 1373
+    },
+    {
+      "epoch": 0.2643324355521354,
+      "grad_norm": 0.9392952554710255,
+      "learning_rate": 1.7260102674436933e-05,
+      "loss": 0.8239,
+      "step": 1374
+    },
+    {
+      "epoch": 0.264524817237399,
+      "grad_norm": 0.7166126216245662,
+      "learning_rate": 1.7255816413036818e-05,
+      "loss": 0.3479,
+      "step": 1375
+    },
+    {
+      "epoch": 0.26471719892266254,
+      "grad_norm": 0.9185406300942623,
+      "learning_rate": 1.7251527334678946e-05,
+      "loss": 0.9122,
+      "step": 1376
+    },
+    {
+      "epoch": 0.26490958060792613,
+      "grad_norm": 0.9254462756843875,
+      "learning_rate": 1.7247235441028486e-05,
+      "loss": 0.8564,
+      "step": 1377
+    },
+    {
+      "epoch": 0.26510196229318966,
+      "grad_norm": 0.9602928575643633,
+      "learning_rate": 1.7242940733751696e-05,
+      "loss": 0.8862,
+      "step": 1378
+    },
+    {
+      "epoch": 0.26529434397845325,
+      "grad_norm": 0.8990679093717878,
+      "learning_rate": 1.7238643214515934e-05,
+      "loss": 0.8275,
+      "step": 1379
+    },
+    {
+      "epoch": 0.26548672566371684,
+      "grad_norm": 0.9279529057204758,
+      "learning_rate": 1.7234342884989642e-05,
+      "loss": 0.8932,
+      "step": 1380
+    },
+    {
+      "epoch": 0.26567910734898037,
+      "grad_norm": 0.9609303241433416,
+      "learning_rate": 1.7230039746842352e-05,
+      "loss": 0.8508,
+      "step": 1381
+    },
+    {
+      "epoch": 0.26587148903424396,
+      "grad_norm": 0.9388362259068539,
+      "learning_rate": 1.7225733801744698e-05,
+      "loss": 0.9111,
+      "step": 1382
+    },
+    {
+      "epoch": 0.2660638707195075,
+      "grad_norm": 0.9121538495783347,
+      "learning_rate": 1.7221425051368394e-05,
+      "loss": 0.8446,
+      "step": 1383
+    },
+    {
+      "epoch": 0.2662562524047711,
+      "grad_norm": 0.8650662807052015,
+      "learning_rate": 1.7217113497386245e-05,
+      "loss": 0.8811,
+      "step": 1384
+    },
+    {
+      "epoch": 0.2664486340900346,
+      "grad_norm": 0.9374707114407996,
+      "learning_rate": 1.721279914147214e-05,
+      "loss": 0.8762,
+      "step": 1385
+    },
+    {
+      "epoch": 0.2666410157752982,
+      "grad_norm": 0.8648148618132696,
+      "learning_rate": 1.7208481985301065e-05,
+      "loss": 0.9042,
+      "step": 1386
+    },
+    {
+      "epoch": 0.2668333974605617,
+      "grad_norm": 0.8088039324365659,
+      "learning_rate": 1.7204162030549093e-05,
+      "loss": 0.847,
+      "step": 1387
+    },
+    {
+      "epoch": 0.2670257791458253,
+      "grad_norm": 0.9291419204027112,
+      "learning_rate": 1.7199839278893368e-05,
+      "loss": 0.8146,
+      "step": 1388
+    },
+    {
+      "epoch": 0.2672181608310889,
+      "grad_norm": 0.9317857642374777,
+      "learning_rate": 1.719551373201214e-05,
+      "loss": 0.8596,
+      "step": 1389
+    },
+    {
+      "epoch": 0.26741054251635243,
+      "grad_norm": 0.9333073071348196,
+      "learning_rate": 1.7191185391584736e-05,
+      "loss": 0.9406,
+      "step": 1390
+    },
+    {
+      "epoch": 0.267602924201616,
+      "grad_norm": 0.8703394114319886,
+      "learning_rate": 1.7186854259291558e-05,
+      "loss": 0.8557,
+      "step": 1391
+    },
+    {
+      "epoch": 0.26779530588687955,
+      "grad_norm": 0.8024041563604332,
+      "learning_rate": 1.7182520336814105e-05,
+      "loss": 0.8382,
+      "step": 1392
+    },
+    {
+      "epoch": 0.26798768757214314,
+      "grad_norm": 0.9400822696666822,
+      "learning_rate": 1.717818362583496e-05,
+      "loss": 0.8655,
+      "step": 1393
+    },
+    {
+      "epoch": 0.2681800692574067,
+      "grad_norm": 0.9281958000369064,
+      "learning_rate": 1.7173844128037777e-05,
+      "loss": 0.8753,
+      "step": 1394
+    },
+    {
+      "epoch": 0.26837245094267026,
+      "grad_norm": 0.806146181190936,
+      "learning_rate": 1.71695018451073e-05,
+      "loss": 0.8347,
+      "step": 1395
+    },
+    {
+      "epoch": 0.26856483262793385,
+      "grad_norm": 0.8302787526504728,
+      "learning_rate": 1.7165156778729355e-05,
+      "loss": 0.9125,
+      "step": 1396
+    },
+    {
+      "epoch": 0.2687572143131974,
+      "grad_norm": 0.9090116840315128,
+      "learning_rate": 1.7160808930590845e-05,
+      "loss": 0.8987,
+      "step": 1397
+    },
+    {
+      "epoch": 0.26894959599846097,
+      "grad_norm": 0.8815228848664296,
+      "learning_rate": 1.7156458302379753e-05,
+      "loss": 0.8092,
+      "step": 1398
+    },
+    {
+      "epoch": 0.2691419776837245,
+      "grad_norm": 0.9419833341808777,
+      "learning_rate": 1.7152104895785147e-05,
+      "loss": 0.8624,
+      "step": 1399
+    },
+    {
+      "epoch": 0.2693343593689881,
+      "grad_norm": 0.852283595733743,
+      "learning_rate": 1.7147748712497162e-05,
+      "loss": 0.9201,
+      "step": 1400
+    },
+    {
+      "epoch": 0.2695267410542516,
+      "grad_norm": 0.9038075410503199,
+      "learning_rate": 1.7143389754207026e-05,
+      "loss": 0.8971,
+      "step": 1401
+    },
+    {
+      "epoch": 0.2697191227395152,
+      "grad_norm": 1.020099155510674,
+      "learning_rate": 1.713902802260703e-05,
+      "loss": 0.8996,
+      "step": 1402
+    },
+    {
+      "epoch": 0.26991150442477874,
+      "grad_norm": 0.8560001828892442,
+      "learning_rate": 1.7134663519390557e-05,
+      "loss": 0.8708,
+      "step": 1403
+    },
+    {
+      "epoch": 0.2701038861100423,
+      "grad_norm": 0.7988922756194091,
+      "learning_rate": 1.7130296246252048e-05,
+      "loss": 0.8681,
+      "step": 1404
+    },
+    {
+      "epoch": 0.2702962677953059,
+      "grad_norm": 0.9830929030485014,
+      "learning_rate": 1.7125926204887034e-05,
+      "loss": 0.8501,
+      "step": 1405
+    },
+    {
+      "epoch": 0.27048864948056944,
+      "grad_norm": 0.956528198461936,
+      "learning_rate": 1.712155339699211e-05,
+      "loss": 0.881,
+      "step": 1406
+    },
+    {
+      "epoch": 0.27068103116583303,
+      "grad_norm": 0.7323608959307211,
+      "learning_rate": 1.7117177824264962e-05,
+      "loss": 0.3524,
+      "step": 1407
+    },
+    {
+      "epoch": 0.27087341285109656,
+      "grad_norm": 0.8585680343673602,
+      "learning_rate": 1.7112799488404327e-05,
+      "loss": 0.8333,
+      "step": 1408
+    },
+    {
+      "epoch": 0.27106579453636015,
+      "grad_norm": 0.9307805225607007,
+      "learning_rate": 1.7108418391110033e-05,
+      "loss": 0.9008,
+      "step": 1409
+    },
+    {
+      "epoch": 0.2712581762216237,
+      "grad_norm": 0.8624994500183429,
+      "learning_rate": 1.7104034534082968e-05,
+      "loss": 0.8281,
+      "step": 1410
+    },
+    {
+      "epoch": 0.27145055790688727,
+      "grad_norm": 0.8425316883816042,
+      "learning_rate": 1.7099647919025096e-05,
+      "loss": 0.8558,
+      "step": 1411
+    },
+    {
+      "epoch": 0.2716429395921508,
+      "grad_norm": 0.9866326475860748,
+      "learning_rate": 1.7095258547639456e-05,
+      "loss": 0.8452,
+      "step": 1412
+    },
+    {
+      "epoch": 0.2718353212774144,
+      "grad_norm": 0.879191537165796,
+      "learning_rate": 1.709086642163015e-05,
+      "loss": 0.8703,
+      "step": 1413
+    },
+    {
+      "epoch": 0.272027702962678,
+      "grad_norm": 0.82145275616546,
+      "learning_rate": 1.7086471542702355e-05,
+      "loss": 0.8662,
+      "step": 1414
+    },
+    {
+      "epoch": 0.2722200846479415,
+      "grad_norm": 0.9126366008638321,
+      "learning_rate": 1.708207391256231e-05,
+      "loss": 0.8148,
+      "step": 1415
+    },
+    {
+      "epoch": 0.2724124663332051,
+      "grad_norm": 1.0293159989974139,
+      "learning_rate": 1.707767353291733e-05,
+      "loss": 0.8682,
+      "step": 1416
+    },
+    {
+      "epoch": 0.2726048480184686,
+      "grad_norm": 0.9849437018358961,
+      "learning_rate": 1.7073270405475796e-05,
+      "loss": 0.8425,
+      "step": 1417
+    },
+    {
+      "epoch": 0.2727972297037322,
+      "grad_norm": 0.9264894002740893,
+      "learning_rate": 1.7068864531947147e-05,
+      "loss": 0.8829,
+      "step": 1418
+    },
+    {
+      "epoch": 0.27298961138899575,
+      "grad_norm": 0.8700377276635641,
+      "learning_rate": 1.70644559140419e-05,
+      "loss": 0.8464,
+      "step": 1419
+    },
+    {
+      "epoch": 0.27318199307425933,
+      "grad_norm": 0.8714331865936815,
+      "learning_rate": 1.706004455347163e-05,
+      "loss": 0.8702,
+      "step": 1420
+    },
+    {
+      "epoch": 0.2733743747595229,
+      "grad_norm": 1.0072112110307212,
+      "learning_rate": 1.705563045194898e-05,
+      "loss": 0.8678,
+      "step": 1421
+    },
+    {
+      "epoch": 0.27356675644478645,
+      "grad_norm": 0.859165900169817,
+      "learning_rate": 1.7051213611187657e-05,
+      "loss": 0.8274,
+      "step": 1422
+    },
+    {
+      "epoch": 0.27375913813005004,
+      "grad_norm": 0.8235266246064139,
+      "learning_rate": 1.704679403290243e-05,
+      "loss": 0.8707,
+      "step": 1423
+    },
+    {
+      "epoch": 0.27395151981531357,
+      "grad_norm": 0.866404883313361,
+      "learning_rate": 1.7042371718809132e-05,
+      "loss": 0.9124,
+      "step": 1424
+    },
+    {
+      "epoch": 0.27414390150057716,
+      "grad_norm": 0.9662383119843795,
+      "learning_rate": 1.7037946670624652e-05,
+      "loss": 0.8534,
+      "step": 1425
+    },
+    {
+      "epoch": 0.2743362831858407,
+      "grad_norm": 0.8933412004998451,
+      "learning_rate": 1.7033518890066956e-05,
+      "loss": 0.8486,
+      "step": 1426
+    },
+    {
+      "epoch": 0.2745286648711043,
+      "grad_norm": 0.9288198789007696,
+      "learning_rate": 1.7029088378855055e-05,
+      "loss": 0.8731,
+      "step": 1427
+    },
+    {
+      "epoch": 0.2747210465563678,
+      "grad_norm": 0.952165722606425,
+      "learning_rate": 1.7024655138709025e-05,
+      "loss": 0.8421,
+      "step": 1428
+    },
+    {
+      "epoch": 0.2749134282416314,
+      "grad_norm": 0.7587851255547033,
+      "learning_rate": 1.7020219171350004e-05,
+      "loss": 0.3565,
+      "step": 1429
+    },
+    {
+      "epoch": 0.275105809926895,
+      "grad_norm": 0.9943526547419366,
+      "learning_rate": 1.7015780478500187e-05,
+      "loss": 0.8414,
+      "step": 1430
+    },
+    {
+      "epoch": 0.2752981916121585,
+      "grad_norm": 0.9194475063125657,
+      "learning_rate": 1.701133906188283e-05,
+      "loss": 0.8373,
+      "step": 1431
+    },
+    {
+      "epoch": 0.2754905732974221,
+      "grad_norm": 0.9441090960616387,
+      "learning_rate": 1.700689492322224e-05,
+      "loss": 0.9042,
+      "step": 1432
+    },
+    {
+      "epoch": 0.27568295498268564,
+      "grad_norm": 0.9277541430743157,
+      "learning_rate": 1.700244806424379e-05,
+      "loss": 0.8366,
+      "step": 1433
+    },
+    {
+      "epoch": 0.2758753366679492,
+      "grad_norm": 0.8821051913730245,
+      "learning_rate": 1.6997998486673893e-05,
+      "loss": 0.8893,
+      "step": 1434
+    },
+    {
+      "epoch": 0.27606771835321275,
+      "grad_norm": 0.9953270876484174,
+      "learning_rate": 1.699354619224004e-05,
+      "loss": 0.8467,
+      "step": 1435
+    },
+    {
+      "epoch": 0.27626010003847634,
+      "grad_norm": 0.8495841599321242,
+      "learning_rate": 1.698909118267076e-05,
+      "loss": 0.8922,
+      "step": 1436
+    },
+    {
+      "epoch": 0.2764524817237399,
+      "grad_norm": 0.8336220121495342,
+      "learning_rate": 1.6984633459695646e-05,
+      "loss": 0.8023,
+      "step": 1437
+    },
+    {
+      "epoch": 0.27664486340900346,
+      "grad_norm": 0.9085787607413458,
+      "learning_rate": 1.6980173025045328e-05,
+      "loss": 0.9043,
+      "step": 1438
+    },
+    {
+      "epoch": 0.27683724509426705,
+      "grad_norm": 0.9827452363668173,
+      "learning_rate": 1.697570988045151e-05,
+      "loss": 0.9165,
+      "step": 1439
+    },
+    {
+      "epoch": 0.2770296267795306,
+      "grad_norm": 0.9776700749321026,
+      "learning_rate": 1.6971244027646937e-05,
+      "loss": 0.8255,
+      "step": 1440
+    },
+    {
+      "epoch": 0.27722200846479417,
+      "grad_norm": 0.9656226224419884,
+      "learning_rate": 1.69667754683654e-05,
+      "loss": 0.8354,
+      "step": 1441
+    },
+    {
+      "epoch": 0.2774143901500577,
+      "grad_norm": 0.9318363965703169,
+      "learning_rate": 1.6962304204341758e-05,
+      "loss": 0.8169,
+      "step": 1442
+    },
+    {
+      "epoch": 0.2776067718353213,
+      "grad_norm": 0.7970618036229608,
+      "learning_rate": 1.6957830237311904e-05,
+      "loss": 0.8759,
+      "step": 1443
+    },
+    {
+      "epoch": 0.2777991535205848,
+      "grad_norm": 0.9362620802672995,
+      "learning_rate": 1.6953353569012784e-05,
+      "loss": 0.8255,
+      "step": 1444
+    },
+    {
+      "epoch": 0.2779915352058484,
+      "grad_norm": 0.9240864190865675,
+      "learning_rate": 1.6948874201182402e-05,
+      "loss": 0.8785,
+      "step": 1445
+    },
+    {
+      "epoch": 0.278183916891112,
+      "grad_norm": 0.9327245547022847,
+      "learning_rate": 1.6944392135559798e-05,
+      "loss": 0.8402,
+      "step": 1446
+    },
+    {
+      "epoch": 0.2783762985763755,
+      "grad_norm": 0.8989171571786677,
+      "learning_rate": 1.6939907373885062e-05,
+      "loss": 0.8372,
+      "step": 1447
+    },
+    {
+      "epoch": 0.2785686802616391,
+      "grad_norm": 0.9640465060824644,
+      "learning_rate": 1.6935419917899335e-05,
+      "loss": 0.8662,
+      "step": 1448
+    },
+    {
+      "epoch": 0.27876106194690264,
+      "grad_norm": 0.866303649150593,
+      "learning_rate": 1.6930929769344807e-05,
+      "loss": 0.872,
+      "step": 1449
+    },
+    {
+      "epoch": 0.27895344363216623,
+      "grad_norm": 0.9401383236948329,
+      "learning_rate": 1.69264369299647e-05,
+      "loss": 0.8647,
+      "step": 1450
+    },
+    {
+      "epoch": 0.27914582531742976,
+      "grad_norm": 0.8526741140403978,
+      "learning_rate": 1.692194140150329e-05,
+      "loss": 0.8788,
+      "step": 1451
+    },
+    {
+      "epoch": 0.27933820700269335,
+      "grad_norm": 0.7556526939178891,
+      "learning_rate": 1.69174431857059e-05,
+      "loss": 0.3684,
+      "step": 1452
+    },
+    {
+      "epoch": 0.2795305886879569,
+      "grad_norm": 0.8997875693402243,
+      "learning_rate": 1.6912942284318898e-05,
+      "loss": 0.8438,
+      "step": 1453
+    },
+    {
+      "epoch": 0.27972297037322047,
+      "grad_norm": 0.9206279194710927,
+      "learning_rate": 1.6908438699089674e-05,
+      "loss": 0.8389,
+      "step": 1454
+    },
+    {
+      "epoch": 0.27991535205848406,
+      "grad_norm": 0.8921241092755989,
+      "learning_rate": 1.690393243176668e-05,
+      "loss": 0.7846,
+      "step": 1455
+    },
+    {
+      "epoch": 0.2801077337437476,
+      "grad_norm": 0.9371914640389231,
+      "learning_rate": 1.6899423484099413e-05,
+      "loss": 0.8413,
+      "step": 1456
+    },
+    {
+      "epoch": 0.2803001154290112,
+      "grad_norm": 0.9209350052109672,
+      "learning_rate": 1.6894911857838394e-05,
+      "loss": 0.8947,
+      "step": 1457
+    },
+    {
+      "epoch": 0.2804924971142747,
+      "grad_norm": 0.9289564996454789,
+      "learning_rate": 1.689039755473519e-05,
+      "loss": 0.8119,
+      "step": 1458
+    },
+    {
+      "epoch": 0.2806848787995383,
+      "grad_norm": 0.8980421807826351,
+      "learning_rate": 1.6885880576542417e-05,
+      "loss": 0.8859,
+      "step": 1459
+    },
+    {
+      "epoch": 0.2808772604848018,
+      "grad_norm": 0.854730822984798,
+      "learning_rate": 1.6881360925013712e-05,
+      "loss": 0.8816,
+      "step": 1460
+    },
+    {
+      "epoch": 0.2810696421700654,
+      "grad_norm": 0.9190983184315106,
+      "learning_rate": 1.6876838601903765e-05,
+      "loss": 0.8758,
+      "step": 1461
+    },
+    {
+      "epoch": 0.28126202385532895,
+      "grad_norm": 0.9562041631059883,
+      "learning_rate": 1.6872313608968296e-05,
+      "loss": 0.8815,
+      "step": 1462
+    },
+    {
+      "epoch": 0.28145440554059253,
+      "grad_norm": 0.9616612823775406,
+      "learning_rate": 1.6867785947964065e-05,
+      "loss": 0.8648,
+      "step": 1463
+    },
+    {
+      "epoch": 0.2816467872258561,
+      "grad_norm": 0.8434838049389829,
+      "learning_rate": 1.6863255620648866e-05,
+      "loss": 0.8151,
+      "step": 1464
+    },
+    {
+      "epoch": 0.28183916891111965,
+      "grad_norm": 0.8482616324577655,
+      "learning_rate": 1.685872262878152e-05,
+      "loss": 0.8907,
+      "step": 1465
+    },
+    {
+      "epoch": 0.28203155059638324,
+      "grad_norm": 0.7973820143206627,
+      "learning_rate": 1.6854186974121903e-05,
+      "loss": 0.8725,
+      "step": 1466
+    },
+    {
+      "epoch": 0.28222393228164677,
+      "grad_norm": 0.913975506400572,
+      "learning_rate": 1.68496486584309e-05,
+      "loss": 0.8752,
+      "step": 1467
+    },
+    {
+      "epoch": 0.28241631396691036,
+      "grad_norm": 0.8690958460167033,
+      "learning_rate": 1.6845107683470453e-05,
+      "loss": 0.8416,
+      "step": 1468
+    },
+    {
+      "epoch": 0.2826086956521739,
+      "grad_norm": 0.889599290738194,
+      "learning_rate": 1.6840564051003517e-05,
+      "loss": 0.867,
+      "step": 1469
+    },
+    {
+      "epoch": 0.2828010773374375,
+      "grad_norm": 0.9409108926868642,
+      "learning_rate": 1.6836017762794087e-05,
+      "loss": 0.8568,
+      "step": 1470
+    },
+    {
+      "epoch": 0.28299345902270107,
+      "grad_norm": 0.8871612524974856,
+      "learning_rate": 1.6831468820607192e-05,
+      "loss": 0.9269,
+      "step": 1471
+    },
+    {
+      "epoch": 0.2831858407079646,
+      "grad_norm": 0.9141228731528063,
+      "learning_rate": 1.6826917226208886e-05,
+      "loss": 0.812,
+      "step": 1472
+    },
+    {
+      "epoch": 0.2833782223932282,
+      "grad_norm": 0.9321069411148923,
+      "learning_rate": 1.6822362981366257e-05,
+      "loss": 0.8467,
+      "step": 1473
+    },
+    {
+      "epoch": 0.2835706040784917,
+      "grad_norm": 0.9455514102330176,
+      "learning_rate": 1.6817806087847417e-05,
+      "loss": 0.7971,
+      "step": 1474
+    },
+    {
+      "epoch": 0.2837629857637553,
+      "grad_norm": 0.8852795700797353,
+      "learning_rate": 1.681324654742151e-05,
+      "loss": 0.8206,
+      "step": 1475
+    },
+    {
+      "epoch": 0.28395536744901884,
+      "grad_norm": 0.9296463623557975,
+      "learning_rate": 1.6808684361858706e-05,
+      "loss": 0.836,
+      "step": 1476
+    },
+    {
+      "epoch": 0.2841477491342824,
+      "grad_norm": 0.9644173291131044,
+      "learning_rate": 1.6804119532930202e-05,
+      "loss": 0.8383,
+      "step": 1477
+    },
+    {
+      "epoch": 0.28434013081954596,
+      "grad_norm": 0.975766867800839,
+      "learning_rate": 1.6799552062408225e-05,
+      "loss": 0.8182,
+      "step": 1478
+    },
+    {
+      "epoch": 0.28453251250480954,
+      "grad_norm": 0.7986601445777287,
+      "learning_rate": 1.6794981952066018e-05,
+      "loss": 0.373,
+      "step": 1479
+    },
+    {
+      "epoch": 0.28472489419007313,
+      "grad_norm": 0.7866342647808597,
+      "learning_rate": 1.6790409203677863e-05,
+      "loss": 0.8553,
+      "step": 1480
+    },
+    {
+      "epoch": 0.28491727587533666,
+      "grad_norm": 0.8618591183943156,
+      "learning_rate": 1.6785833819019052e-05,
+      "loss": 0.864,
+      "step": 1481
+    },
+    {
+      "epoch": 0.28510965756060025,
+      "grad_norm": 0.9157266107191265,
+      "learning_rate": 1.678125579986591e-05,
+      "loss": 0.8981,
+      "step": 1482
+    },
+    {
+      "epoch": 0.2853020392458638,
+      "grad_norm": 0.8588935419107674,
+      "learning_rate": 1.677667514799578e-05,
+      "loss": 0.8067,
+      "step": 1483
+    },
+    {
+      "epoch": 0.28549442093112737,
+      "grad_norm": 0.9099511637427563,
+      "learning_rate": 1.6772091865187032e-05,
+      "loss": 0.8712,
+      "step": 1484
+    },
+    {
+      "epoch": 0.2856868026163909,
+      "grad_norm": 0.8814654239328586,
+      "learning_rate": 1.676750595321905e-05,
+      "loss": 0.8147,
+      "step": 1485
+    },
+    {
+      "epoch": 0.2858791843016545,
+      "grad_norm": 0.9222692176945962,
+      "learning_rate": 1.6762917413872246e-05,
+      "loss": 0.853,
+      "step": 1486
+    },
+    {
+      "epoch": 0.286071565986918,
+      "grad_norm": 0.9145659276912992,
+      "learning_rate": 1.675832624892805e-05,
+      "loss": 0.7603,
+      "step": 1487
+    },
+    {
+      "epoch": 0.2862639476721816,
+      "grad_norm": 0.9643651353107544,
+      "learning_rate": 1.6753732460168907e-05,
+      "loss": 0.8626,
+      "step": 1488
+    },
+    {
+      "epoch": 0.2864563293574452,
+      "grad_norm": 0.907441096965074,
+      "learning_rate": 1.674913604937828e-05,
+      "loss": 0.861,
+      "step": 1489
+    },
+    {
+      "epoch": 0.2866487110427087,
+      "grad_norm": 0.9262424380178506,
+      "learning_rate": 1.6744537018340662e-05,
+      "loss": 0.8639,
+      "step": 1490
+    },
+    {
+      "epoch": 0.2868410927279723,
+      "grad_norm": 0.8809614823206697,
+      "learning_rate": 1.6739935368841555e-05,
+      "loss": 0.8514,
+      "step": 1491
+    },
+    {
+      "epoch": 0.28703347441323585,
+      "grad_norm": 0.9956349542814713,
+      "learning_rate": 1.6735331102667475e-05,
+      "loss": 0.8467,
+      "step": 1492
+    },
+    {
+      "epoch": 0.28722585609849943,
+      "grad_norm": 0.9384976918473367,
+      "learning_rate": 1.6730724221605955e-05,
+      "loss": 0.8329,
+      "step": 1493
+    },
+    {
+      "epoch": 0.28741823778376296,
+      "grad_norm": 0.9703811534013784,
+      "learning_rate": 1.6726114727445547e-05,
+      "loss": 0.8737,
+      "step": 1494
+    },
+    {
+      "epoch": 0.28761061946902655,
+      "grad_norm": 0.9186413462277588,
+      "learning_rate": 1.6721502621975813e-05,
+      "loss": 0.8885,
+      "step": 1495
+    },
+    {
+      "epoch": 0.28780300115429014,
+      "grad_norm": 0.8987349699471533,
+      "learning_rate": 1.6716887906987332e-05,
+      "loss": 0.8894,
+      "step": 1496
+    },
+    {
+      "epoch": 0.28799538283955367,
+      "grad_norm": 0.93594376398792,
+      "learning_rate": 1.6712270584271703e-05,
+      "loss": 0.898,
+      "step": 1497
+    },
+    {
+      "epoch": 0.28818776452481726,
+      "grad_norm": 0.9209177335509535,
+      "learning_rate": 1.670765065562152e-05,
+      "loss": 0.8353,
+      "step": 1498
+    },
+    {
+      "epoch": 0.2883801462100808,
+      "grad_norm": 0.9405909867275929,
+      "learning_rate": 1.67030281228304e-05,
+      "loss": 0.8578,
+      "step": 1499
+    },
+    {
+      "epoch": 0.2885725278953444,
+      "grad_norm": 0.9236855152282274,
+      "learning_rate": 1.6698402987692968e-05,
+      "loss": 0.8904,
+      "step": 1500
+    },
+    {
+      "epoch": 0.2887649095806079,
+      "grad_norm": 0.8296995565529333,
+      "learning_rate": 1.6693775252004866e-05,
+      "loss": 0.8674,
+      "step": 1501
+    },
+    {
+      "epoch": 0.2889572912658715,
+      "grad_norm": 0.8615563964106138,
+      "learning_rate": 1.668914491756274e-05,
+      "loss": 0.8037,
+      "step": 1502
+    },
+    {
+      "epoch": 0.28914967295113503,
+      "grad_norm": 0.9169717040655676,
+      "learning_rate": 1.668451198616424e-05,
+      "loss": 0.8652,
+      "step": 1503
+    },
+    {
+      "epoch": 0.2893420546363986,
+      "grad_norm": 0.7605180677976029,
+      "learning_rate": 1.6679876459608033e-05,
+      "loss": 0.3288,
+      "step": 1504
+    },
+    {
+      "epoch": 0.2895344363216622,
+      "grad_norm": 0.9572698597833463,
+      "learning_rate": 1.667523833969379e-05,
+      "loss": 0.8535,
+      "step": 1505
+    },
+    {
+      "epoch": 0.28972681800692573,
+      "grad_norm": 0.8748678917705399,
+      "learning_rate": 1.667059762822219e-05,
+      "loss": 0.89,
+      "step": 1506
+    },
+    {
+      "epoch": 0.2899191996921893,
+      "grad_norm": 0.8295877445371332,
+      "learning_rate": 1.666595432699491e-05,
+      "loss": 0.8427,
+      "step": 1507
+    },
+    {
+      "epoch": 0.29011158137745285,
+      "grad_norm": 0.8622781044369935,
+      "learning_rate": 1.6661308437814652e-05,
+      "loss": 0.8432,
+      "step": 1508
+    },
+    {
+      "epoch": 0.29030396306271644,
+      "grad_norm": 0.8917058858791409,
+      "learning_rate": 1.6656659962485097e-05,
+      "loss": 0.8857,
+      "step": 1509
+    },
+    {
+      "epoch": 0.29049634474798,
+      "grad_norm": 0.8627643394024034,
+      "learning_rate": 1.6652008902810952e-05,
+      "loss": 0.8012,
+      "step": 1510
+    },
+    {
+      "epoch": 0.29068872643324356,
+      "grad_norm": 0.8789276314486144,
+      "learning_rate": 1.6647355260597915e-05,
+      "loss": 0.8561,
+      "step": 1511
+    },
+    {
+      "epoch": 0.2908811081185071,
+      "grad_norm": 0.9984338137083408,
+      "learning_rate": 1.664269903765269e-05,
+      "loss": 0.9502,
+      "step": 1512
+    },
+    {
+      "epoch": 0.2910734898037707,
+      "grad_norm": 0.8796481558015427,
+      "learning_rate": 1.6638040235782983e-05,
+      "loss": 0.9024,
+      "step": 1513
+    },
+    {
+      "epoch": 0.29126587148903427,
+      "grad_norm": 0.8314372933323695,
+      "learning_rate": 1.6633378856797505e-05,
+      "loss": 0.9024,
+      "step": 1514
+    },
+    {
+      "epoch": 0.2914582531742978,
+      "grad_norm": 0.9277322959728481,
+      "learning_rate": 1.662871490250596e-05,
+      "loss": 0.8929,
+      "step": 1515
+    },
+    {
+      "epoch": 0.2916506348595614,
+      "grad_norm": 0.8435115666459351,
+      "learning_rate": 1.662404837471905e-05,
+      "loss": 0.888,
+      "step": 1516
+    },
+    {
+      "epoch": 0.2918430165448249,
+      "grad_norm": 0.8744007691977231,
+      "learning_rate": 1.66193792752485e-05,
+      "loss": 0.8606,
+      "step": 1517
+    },
+    {
+      "epoch": 0.2920353982300885,
+      "grad_norm": 0.9183207309956976,
+      "learning_rate": 1.6614707605906995e-05,
+      "loss": 0.881,
+      "step": 1518
+    },
+    {
+      "epoch": 0.29222777991535204,
+      "grad_norm": 0.9090152060911204,
+      "learning_rate": 1.661003336850825e-05,
+      "loss": 0.9008,
+      "step": 1519
+    },
+    {
+      "epoch": 0.2924201616006156,
+      "grad_norm": 0.9376217065459664,
+      "learning_rate": 1.660535656486696e-05,
+      "loss": 0.8546,
+      "step": 1520
+    },
+    {
+      "epoch": 0.2926125432858792,
+      "grad_norm": 0.9401968202987555,
+      "learning_rate": 1.660067719679882e-05,
+      "loss": 0.9427,
+      "step": 1521
+    },
+    {
+      "epoch": 0.29280492497114274,
+      "grad_norm": 0.9181573119624051,
+      "learning_rate": 1.6595995266120528e-05,
+      "loss": 0.8594,
+      "step": 1522
+    },
+    {
+      "epoch": 0.29299730665640633,
+      "grad_norm": 0.9030857533842448,
+      "learning_rate": 1.6591310774649766e-05,
+      "loss": 0.8353,
+      "step": 1523
+    },
+    {
+      "epoch": 0.29318968834166986,
+      "grad_norm": 0.922800118871425,
+      "learning_rate": 1.6586623724205216e-05,
+      "loss": 0.8844,
+      "step": 1524
+    },
+    {
+      "epoch": 0.29338207002693345,
+      "grad_norm": 0.9576258709943511,
+      "learning_rate": 1.6581934116606554e-05,
+      "loss": 0.8781,
+      "step": 1525
+    },
+    {
+      "epoch": 0.293574451712197,
+      "grad_norm": 0.8643184202466663,
+      "learning_rate": 1.657724195367444e-05,
+      "loss": 0.8244,
+      "step": 1526
+    },
+    {
+      "epoch": 0.29376683339746057,
+      "grad_norm": 0.8500666146123014,
+      "learning_rate": 1.657254723723054e-05,
+      "loss": 0.8458,
+      "step": 1527
+    },
+    {
+      "epoch": 0.2939592150827241,
+      "grad_norm": 0.8799729692504109,
+      "learning_rate": 1.6567849969097505e-05,
+      "loss": 0.8587,
+      "step": 1528
+    },
+    {
+      "epoch": 0.2941515967679877,
+      "grad_norm": 0.8444817723167254,
+      "learning_rate": 1.6563150151098973e-05,
+      "loss": 0.3248,
+      "step": 1529
+    },
+    {
+      "epoch": 0.2943439784532513,
+      "grad_norm": 0.9677522613370997,
+      "learning_rate": 1.6558447785059577e-05,
+      "loss": 0.9234,
+      "step": 1530
+    },
+    {
+      "epoch": 0.2945363601385148,
+      "grad_norm": 0.9458967136577793,
+      "learning_rate": 1.655374287280494e-05,
+      "loss": 0.8986,
+      "step": 1531
+    },
+    {
+      "epoch": 0.2947287418237784,
+      "grad_norm": 0.8592845366707627,
+      "learning_rate": 1.6549035416161662e-05,
+      "loss": 0.8424,
+      "step": 1532
+    },
+    {
+      "epoch": 0.2949211235090419,
+      "grad_norm": 0.8474679989433095,
+      "learning_rate": 1.654432541695735e-05,
+      "loss": 0.8576,
+      "step": 1533
+    },
+    {
+      "epoch": 0.2951135051943055,
+      "grad_norm": 0.8836553526929412,
+      "learning_rate": 1.653961287702058e-05,
+      "loss": 0.838,
+      "step": 1534
+    },
+    {
+      "epoch": 0.29530588687956905,
+      "grad_norm": 0.8511860559926913,
+      "learning_rate": 1.653489779818093e-05,
+      "loss": 0.8004,
+      "step": 1535
+    },
+    {
+      "epoch": 0.29549826856483263,
+      "grad_norm": 0.7917006298889878,
+      "learning_rate": 1.6530180182268946e-05,
+      "loss": 0.8271,
+      "step": 1536
+    },
+    {
+      "epoch": 0.29569065025009617,
+      "grad_norm": 0.8906568826520794,
+      "learning_rate": 1.652546003111618e-05,
+      "loss": 0.8367,
+      "step": 1537
+    },
+    {
+      "epoch": 0.29588303193535975,
+      "grad_norm": 0.8377418693460059,
+      "learning_rate": 1.652073734655515e-05,
+      "loss": 0.8358,
+      "step": 1538
+    },
+    {
+      "epoch": 0.29607541362062334,
+      "grad_norm": 0.8810050849180725,
+      "learning_rate": 1.6516012130419366e-05,
+      "loss": 0.8172,
+      "step": 1539
+    },
+    {
+      "epoch": 0.29626779530588687,
+      "grad_norm": 0.9325398138601191,
+      "learning_rate": 1.6511284384543317e-05,
+      "loss": 0.8941,
+      "step": 1540
+    },
+    {
+      "epoch": 0.29646017699115046,
+      "grad_norm": 1.0266151206068497,
+      "learning_rate": 1.6506554110762483e-05,
+      "loss": 0.8794,
+      "step": 1541
+    },
+    {
+      "epoch": 0.296652558676414,
+      "grad_norm": 0.900803679261653,
+      "learning_rate": 1.650182131091332e-05,
+      "loss": 0.8517,
+      "step": 1542
+    },
+    {
+      "epoch": 0.2968449403616776,
+      "grad_norm": 0.9246961101586452,
+      "learning_rate": 1.6497085986833252e-05,
+      "loss": 0.8447,
+      "step": 1543
+    },
+    {
+      "epoch": 0.2970373220469411,
+      "grad_norm": 0.9349087566733381,
+      "learning_rate": 1.6492348140360704e-05,
+      "loss": 0.8376,
+      "step": 1544
+    },
+    {
+      "epoch": 0.2972297037322047,
+      "grad_norm": 0.8825990842501937,
+      "learning_rate": 1.6487607773335074e-05,
+      "loss": 0.8008,
+      "step": 1545
+    },
+    {
+      "epoch": 0.29742208541746823,
+      "grad_norm": 0.8290047062696579,
+      "learning_rate": 1.648286488759673e-05,
+      "loss": 0.8718,
+      "step": 1546
+    },
+    {
+      "epoch": 0.2976144671027318,
+      "grad_norm": 0.8717857506055448,
+      "learning_rate": 1.6478119484987026e-05,
+      "loss": 0.8345,
+      "step": 1547
+    },
+    {
+      "epoch": 0.2978068487879954,
+      "grad_norm": 0.7282154573200073,
+      "learning_rate": 1.6473371567348287e-05,
+      "loss": 0.3311,
+      "step": 1548
+    },
+    {
+      "epoch": 0.29799923047325894,
+      "grad_norm": 0.8742679161454444,
+      "learning_rate": 1.6468621136523823e-05,
+      "loss": 0.8341,
+      "step": 1549
+    },
+    {
+      "epoch": 0.2981916121585225,
+      "grad_norm": 0.906076936299084,
+      "learning_rate": 1.646386819435791e-05,
+      "loss": 0.8507,
+      "step": 1550
+    },
+    {
+      "epoch": 0.29838399384378606,
+      "grad_norm": 1.026421824365286,
+      "learning_rate": 1.6459112742695807e-05,
+      "loss": 0.8045,
+      "step": 1551
+    },
+    {
+      "epoch": 0.29857637552904964,
+      "grad_norm": 0.9607611896405842,
+      "learning_rate": 1.6454354783383748e-05,
+      "loss": 0.9229,
+      "step": 1552
+    },
+    {
+      "epoch": 0.2987687572143132,
+      "grad_norm": 0.8247145372458107,
+      "learning_rate": 1.644959431826893e-05,
+      "loss": 0.8271,
+      "step": 1553
+    },
+    {
+      "epoch": 0.29896113889957676,
+      "grad_norm": 0.8566315857866287,
+      "learning_rate": 1.6444831349199528e-05,
+      "loss": 0.826,
+      "step": 1554
+    },
+    {
+      "epoch": 0.29915352058484035,
+      "grad_norm": 0.8573592133138535,
+      "learning_rate": 1.6440065878024697e-05,
+      "loss": 0.891,
+      "step": 1555
+    },
+    {
+      "epoch": 0.2993459022701039,
+      "grad_norm": 0.915938357767272,
+      "learning_rate": 1.6435297906594553e-05,
+      "loss": 0.779,
+      "step": 1556
+    },
+    {
+      "epoch": 0.29953828395536747,
+      "grad_norm": 0.9627678594208273,
+      "learning_rate": 1.643052743676019e-05,
+      "loss": 0.8512,
+      "step": 1557
+    },
+    {
+      "epoch": 0.299730665640631,
+      "grad_norm": 0.6816395263590602,
+      "learning_rate": 1.6425754470373667e-05,
+      "loss": 0.3801,
+      "step": 1558
+    },
+    {
+      "epoch": 0.2999230473258946,
+      "grad_norm": 0.9605790886130494,
+      "learning_rate": 1.642097900928801e-05,
+      "loss": 0.8275,
+      "step": 1559
+    },
+    {
+      "epoch": 0.3001154290111581,
+      "grad_norm": 0.886579924389122,
+      "learning_rate": 1.6416201055357225e-05,
+      "loss": 0.7802,
+      "step": 1560
+    },
+    {
+      "epoch": 0.3003078106964217,
+      "grad_norm": 0.8438600299904057,
+      "learning_rate": 1.641142061043627e-05,
+      "loss": 0.9189,
+      "step": 1561
+    },
+    {
+      "epoch": 0.30050019238168524,
+      "grad_norm": 0.9540771675970636,
+      "learning_rate": 1.640663767638108e-05,
+      "loss": 0.8601,
+      "step": 1562
+    },
+    {
+      "epoch": 0.3006925740669488,
+      "grad_norm": 0.893789107797163,
+      "learning_rate": 1.6401852255048564e-05,
+      "loss": 0.853,
+      "step": 1563
+    },
+    {
+      "epoch": 0.3008849557522124,
+      "grad_norm": 0.6373127183821419,
+      "learning_rate": 1.6397064348296578e-05,
+      "loss": 0.3691,
+      "step": 1564
+    },
+    {
+      "epoch": 0.30107733743747594,
+      "grad_norm": 1.0175272499455714,
+      "learning_rate": 1.6392273957983955e-05,
+      "loss": 0.8831,
+      "step": 1565
+    },
+    {
+      "epoch": 0.30126971912273953,
+      "grad_norm": 0.9219740469402519,
+      "learning_rate": 1.638748108597049e-05,
+      "loss": 0.8742,
+      "step": 1566
+    },
+    {
+      "epoch": 0.30146210080800306,
+      "grad_norm": 0.9119738464282401,
+      "learning_rate": 1.6382685734116934e-05,
+      "loss": 0.855,
+      "step": 1567
+    },
+    {
+      "epoch": 0.30165448249326665,
+      "grad_norm": 0.8571311410694311,
+      "learning_rate": 1.6377887904285018e-05,
+      "loss": 0.8204,
+      "step": 1568
+    },
+    {
+      "epoch": 0.3018468641785302,
+      "grad_norm": 0.9562320183869502,
+      "learning_rate": 1.637308759833742e-05,
+      "loss": 0.8907,
+      "step": 1569
+    },
+    {
+      "epoch": 0.30203924586379377,
+      "grad_norm": 0.8466390224683671,
+      "learning_rate": 1.6368284818137787e-05,
+      "loss": 0.8695,
+      "step": 1570
+    },
+    {
+      "epoch": 0.3022316275490573,
+      "grad_norm": 0.9069716684412936,
+      "learning_rate": 1.636347956555072e-05,
+      "loss": 0.8524,
+      "step": 1571
+    },
+    {
+      "epoch": 0.3024240092343209,
+      "grad_norm": 0.8263728755783778,
+      "learning_rate": 1.635867184244178e-05,
+      "loss": 0.8271,
+      "step": 1572
+    },
+    {
+      "epoch": 0.3026163909195845,
+      "grad_norm": 0.9837589087457346,
+      "learning_rate": 1.63538616506775e-05,
+      "loss": 0.8987,
+      "step": 1573
+    },
+    {
+      "epoch": 0.302808772604848,
+      "grad_norm": 0.9085024745355368,
+      "learning_rate": 1.6349048992125358e-05,
+      "loss": 0.8547,
+      "step": 1574
+    },
+    {
+      "epoch": 0.3030011542901116,
+      "grad_norm": 0.8290184094515922,
+      "learning_rate": 1.634423386865379e-05,
+      "loss": 0.8749,
+      "step": 1575
+    },
+    {
+      "epoch": 0.30319353597537513,
+      "grad_norm": 0.9370856587416664,
+      "learning_rate": 1.6339416282132196e-05,
+      "loss": 0.9286,
+      "step": 1576
+    },
+    {
+      "epoch": 0.3033859176606387,
+      "grad_norm": 0.8560200104119186,
+      "learning_rate": 1.633459623443093e-05,
+      "loss": 0.8682,
+      "step": 1577
+    },
+    {
+      "epoch": 0.30357829934590225,
+      "grad_norm": 0.8108849620453558,
+      "learning_rate": 1.6329773727421297e-05,
+      "loss": 0.8574,
+      "step": 1578
+    },
+    {
+      "epoch": 0.30377068103116583,
+      "grad_norm": 0.9214154917422018,
+      "learning_rate": 1.6324948762975567e-05,
+      "loss": 0.8443,
+      "step": 1579
+    },
+    {
+      "epoch": 0.3039630627164294,
+      "grad_norm": 0.8999718359567525,
+      "learning_rate": 1.632012134296695e-05,
+      "loss": 0.8381,
+      "step": 1580
+    },
+    {
+      "epoch": 0.30415544440169295,
+      "grad_norm": 0.7069770376394836,
+      "learning_rate": 1.6315291469269617e-05,
+      "loss": 0.3836,
+      "step": 1581
+    },
+    {
+      "epoch": 0.30434782608695654,
+      "grad_norm": 0.6223353689657546,
+      "learning_rate": 1.63104591437587e-05,
+      "loss": 0.3402,
+      "step": 1582
+    },
+    {
+      "epoch": 0.3045402077722201,
+      "grad_norm": 0.8642903387751513,
+      "learning_rate": 1.6305624368310265e-05,
+      "loss": 0.8538,
+      "step": 1583
+    },
+    {
+      "epoch": 0.30473258945748366,
+      "grad_norm": 0.9193075539422011,
+      "learning_rate": 1.630078714480134e-05,
+      "loss": 0.8416,
+      "step": 1584
+    },
+    {
+      "epoch": 0.3049249711427472,
+      "grad_norm": 0.9855141215035249,
+      "learning_rate": 1.6295947475109904e-05,
+      "loss": 0.8624,
+      "step": 1585
+    },
+    {
+      "epoch": 0.3051173528280108,
+      "grad_norm": 0.9088348257554549,
+      "learning_rate": 1.629110536111488e-05,
+      "loss": 0.8734,
+      "step": 1586
+    },
+    {
+      "epoch": 0.3053097345132743,
+      "grad_norm": 0.8750688867140418,
+      "learning_rate": 1.628626080469615e-05,
+      "loss": 0.819,
+      "step": 1587
+    },
+    {
+      "epoch": 0.3055021161985379,
+      "grad_norm": 0.9894177519848726,
+      "learning_rate": 1.628141380773453e-05,
+      "loss": 0.9148,
+      "step": 1588
+    },
+    {
+      "epoch": 0.3056944978838015,
+      "grad_norm": 0.8469160605691728,
+      "learning_rate": 1.6276564372111797e-05,
+      "loss": 0.8257,
+      "step": 1589
+    },
+    {
+      "epoch": 0.305886879569065,
+      "grad_norm": 0.9120624137998166,
+      "learning_rate": 1.6271712499710663e-05,
+      "loss": 0.8949,
+      "step": 1590
+    },
+    {
+      "epoch": 0.3060792612543286,
+      "grad_norm": 0.7721870948423383,
+      "learning_rate": 1.62668581924148e-05,
+      "loss": 0.3437,
+      "step": 1591
+    },
+    {
+      "epoch": 0.30627164293959214,
+      "grad_norm": 0.8587771829034572,
+      "learning_rate": 1.6262001452108807e-05,
+      "loss": 0.8706,
+      "step": 1592
+    },
+    {
+      "epoch": 0.3064640246248557,
+      "grad_norm": 0.9554459866413392,
+      "learning_rate": 1.6257142280678247e-05,
+      "loss": 0.8969,
+      "step": 1593
+    },
+    {
+      "epoch": 0.30665640631011926,
+      "grad_norm": 0.8896780993742694,
+      "learning_rate": 1.6252280680009613e-05,
+      "loss": 0.8489,
+      "step": 1594
+    },
+    {
+      "epoch": 0.30684878799538284,
+      "grad_norm": 0.8457536700615383,
+      "learning_rate": 1.6247416651990343e-05,
+      "loss": 0.7876,
+      "step": 1595
+    },
+    {
+      "epoch": 0.3070411696806464,
+      "grad_norm": 0.8325884747592281,
+      "learning_rate": 1.624255019850883e-05,
+      "loss": 0.9012,
+      "step": 1596
+    },
+    {
+      "epoch": 0.30723355136590996,
+      "grad_norm": 0.8829774827193305,
+      "learning_rate": 1.6237681321454387e-05,
+      "loss": 0.8825,
+      "step": 1597
+    },
+    {
+      "epoch": 0.30742593305117355,
+      "grad_norm": 0.8305115364818458,
+      "learning_rate": 1.623281002271729e-05,
+      "loss": 0.8085,
+      "step": 1598
+    },
+    {
+      "epoch": 0.3076183147364371,
+      "grad_norm": 0.8497681288367605,
+      "learning_rate": 1.6227936304188738e-05,
+      "loss": 0.8274,
+      "step": 1599
+    },
+    {
+      "epoch": 0.30781069642170067,
+      "grad_norm": 0.9688503017722131,
+      "learning_rate": 1.622306016776088e-05,
+      "loss": 0.8812,
+      "step": 1600
+    },
+    {
+      "epoch": 0.3080030781069642,
+      "grad_norm": 0.9407778882932615,
+      "learning_rate": 1.6218181615326795e-05,
+      "loss": 0.9102,
+      "step": 1601
+    },
+    {
+      "epoch": 0.3081954597922278,
+      "grad_norm": 0.8626128694040522,
+      "learning_rate": 1.6213300648780515e-05,
+      "loss": 0.7894,
+      "step": 1602
+    },
+    {
+      "epoch": 0.3083878414774913,
+      "grad_norm": 0.8826254389459387,
+      "learning_rate": 1.620841727001699e-05,
+      "loss": 0.8153,
+      "step": 1603
+    },
+    {
+      "epoch": 0.3085802231627549,
+      "grad_norm": 0.9601227941384879,
+      "learning_rate": 1.6203531480932114e-05,
+      "loss": 0.864,
+      "step": 1604
+    },
+    {
+      "epoch": 0.3087726048480185,
+      "grad_norm": 0.8355745733367002,
+      "learning_rate": 1.619864328342273e-05,
+      "loss": 0.8597,
+      "step": 1605
+    },
+    {
+      "epoch": 0.308964986533282,
+      "grad_norm": 0.90851058906388,
+      "learning_rate": 1.6193752679386593e-05,
+      "loss": 0.8229,
+      "step": 1606
+    },
+    {
+      "epoch": 0.3091573682185456,
+      "grad_norm": 0.8532495704007541,
+      "learning_rate": 1.6188859670722414e-05,
+      "loss": 0.8516,
+      "step": 1607
+    },
+    {
+      "epoch": 0.30934974990380915,
+      "grad_norm": 0.8903836611062547,
+      "learning_rate": 1.6183964259329817e-05,
+      "loss": 0.8255,
+      "step": 1608
+    },
+    {
+      "epoch": 0.30954213158907273,
+      "grad_norm": 0.9292970036163868,
+      "learning_rate": 1.6179066447109376e-05,
+      "loss": 0.8363,
+      "step": 1609
+    },
+    {
+      "epoch": 0.30973451327433627,
+      "grad_norm": 0.9269739642780833,
+      "learning_rate": 1.6174166235962588e-05,
+      "loss": 0.8889,
+      "step": 1610
+    },
+    {
+      "epoch": 0.30992689495959985,
+      "grad_norm": 0.8863002141508217,
+      "learning_rate": 1.6169263627791886e-05,
+      "loss": 0.8614,
+      "step": 1611
+    },
+    {
+      "epoch": 0.3101192766448634,
+      "grad_norm": 0.824713069483705,
+      "learning_rate": 1.616435862450063e-05,
+      "loss": 0.3688,
+      "step": 1612
+    },
+    {
+      "epoch": 0.31031165833012697,
+      "grad_norm": 0.6895408892311093,
+      "learning_rate": 1.615945122799311e-05,
+      "loss": 0.3498,
+      "step": 1613
+    },
+    {
+      "epoch": 0.31050404001539056,
+      "grad_norm": 0.9128119609439963,
+      "learning_rate": 1.6154541440174547e-05,
+      "loss": 0.8057,
+      "step": 1614
+    },
+    {
+      "epoch": 0.3106964217006541,
+      "grad_norm": 1.0428609977288927,
+      "learning_rate": 1.614962926295109e-05,
+      "loss": 0.7899,
+      "step": 1615
+    },
+    {
+      "epoch": 0.3108888033859177,
+      "grad_norm": 0.9413933397777404,
+      "learning_rate": 1.6144714698229814e-05,
+      "loss": 0.8901,
+      "step": 1616
+    },
+    {
+      "epoch": 0.3110811850711812,
+      "grad_norm": 0.8818682855757238,
+      "learning_rate": 1.6139797747918725e-05,
+      "loss": 0.8639,
+      "step": 1617
+    },
+    {
+      "epoch": 0.3112735667564448,
+      "grad_norm": 0.8917260489978234,
+      "learning_rate": 1.613487841392675e-05,
+      "loss": 0.8443,
+      "step": 1618
+    },
+    {
+      "epoch": 0.31146594844170833,
+      "grad_norm": 0.8937641786846655,
+      "learning_rate": 1.612995669816375e-05,
+      "loss": 0.8828,
+      "step": 1619
+    },
+    {
+      "epoch": 0.3116583301269719,
+      "grad_norm": 0.8557170770225866,
+      "learning_rate": 1.6125032602540492e-05,
+      "loss": 0.9066,
+      "step": 1620
+    },
+    {
+      "epoch": 0.31185071181223545,
+      "grad_norm": 0.8966394391582696,
+      "learning_rate": 1.6120106128968686e-05,
+      "loss": 0.8751,
+      "step": 1621
+    },
+    {
+      "epoch": 0.31204309349749904,
+      "grad_norm": 0.9280395205864794,
+      "learning_rate": 1.6115177279360965e-05,
+      "loss": 0.809,
+      "step": 1622
+    },
+    {
+      "epoch": 0.3122354751827626,
+      "grad_norm": 0.9422618883122279,
+      "learning_rate": 1.611024605563087e-05,
+      "loss": 0.8649,
+      "step": 1623
+    },
+    {
+      "epoch": 0.31242785686802615,
+      "grad_norm": 0.8835415581038802,
+      "learning_rate": 1.610531245969287e-05,
+      "loss": 0.9078,
+      "step": 1624
+    },
+    {
+      "epoch": 0.31262023855328974,
+      "grad_norm": 0.8120080332117713,
+      "learning_rate": 1.6100376493462368e-05,
+      "loss": 0.8208,
+      "step": 1625
+    },
+    {
+      "epoch": 0.3128126202385533,
+      "grad_norm": 0.9058155707531742,
+      "learning_rate": 1.6095438158855668e-05,
+      "loss": 0.3201,
+      "step": 1626
+    },
+    {
+      "epoch": 0.31300500192381686,
+      "grad_norm": 0.9315915155362287,
+      "learning_rate": 1.609049745779e-05,
+      "loss": 0.874,
+      "step": 1627
+    },
+    {
+      "epoch": 0.3131973836090804,
+      "grad_norm": 0.8673389880725348,
+      "learning_rate": 1.6085554392183517e-05,
+      "loss": 0.8336,
+      "step": 1628
+    },
+    {
+      "epoch": 0.313389765294344,
+      "grad_norm": 0.9008608806041691,
+      "learning_rate": 1.608060896395529e-05,
+      "loss": 0.8772,
+      "step": 1629
+    },
+    {
+      "epoch": 0.31358214697960757,
+      "grad_norm": 0.9201628085467389,
+      "learning_rate": 1.60756611750253e-05,
+      "loss": 0.8105,
+      "step": 1630
+    },
+    {
+      "epoch": 0.3137745286648711,
+      "grad_norm": 0.9695393304446149,
+      "learning_rate": 1.6070711027314446e-05,
+      "loss": 0.9472,
+      "step": 1631
+    },
+    {
+      "epoch": 0.3139669103501347,
+      "grad_norm": 0.7172459652664743,
+      "learning_rate": 1.606575852274456e-05,
+      "loss": 0.357,
+      "step": 1632
+    },
+    {
+      "epoch": 0.3141592920353982,
+      "grad_norm": 0.8618675375097724,
+      "learning_rate": 1.6060803663238357e-05,
+      "loss": 0.8731,
+      "step": 1633
+    },
+    {
+      "epoch": 0.3143516737206618,
+      "grad_norm": 0.9377316955868895,
+      "learning_rate": 1.6055846450719498e-05,
+      "loss": 0.872,
+      "step": 1634
+    },
+    {
+      "epoch": 0.31454405540592534,
+      "grad_norm": 0.8144862373695538,
+      "learning_rate": 1.6050886887112535e-05,
+      "loss": 0.8848,
+      "step": 1635
+    },
+    {
+      "epoch": 0.3147364370911889,
+      "grad_norm": 0.8998299321758159,
+      "learning_rate": 1.6045924974342945e-05,
+      "loss": 0.7975,
+      "step": 1636
+    },
+    {
+      "epoch": 0.31492881877645246,
+      "grad_norm": 0.8241921518317038,
+      "learning_rate": 1.604096071433711e-05,
+      "loss": 0.8804,
+      "step": 1637
+    },
+    {
+      "epoch": 0.31512120046171604,
+      "grad_norm": 0.9217471921434475,
+      "learning_rate": 1.6035994109022333e-05,
+      "loss": 0.8689,
+      "step": 1638
+    },
+    {
+      "epoch": 0.31531358214697963,
+      "grad_norm": 0.9261089114219587,
+      "learning_rate": 1.6031025160326814e-05,
+      "loss": 0.8067,
+      "step": 1639
+    },
+    {
+      "epoch": 0.31550596383224316,
+      "grad_norm": 0.8825077960885497,
+      "learning_rate": 1.6026053870179678e-05,
+      "loss": 0.8427,
+      "step": 1640
+    },
+    {
+      "epoch": 0.31569834551750675,
+      "grad_norm": 0.9575254470658041,
+      "learning_rate": 1.6021080240510943e-05,
+      "loss": 0.8782,
+      "step": 1641
+    },
+    {
+      "epoch": 0.3158907272027703,
+      "grad_norm": 0.8328568208547074,
+      "learning_rate": 1.601610427325155e-05,
+      "loss": 0.874,
+      "step": 1642
+    },
+    {
+      "epoch": 0.31608310888803387,
+      "grad_norm": 0.7902125234717571,
+      "learning_rate": 1.6011125970333333e-05,
+      "loss": 0.8507,
+      "step": 1643
+    },
+    {
+      "epoch": 0.3162754905732974,
+      "grad_norm": 0.9300249146829,
+      "learning_rate": 1.600614533368905e-05,
+      "loss": 0.8314,
+      "step": 1644
+    },
+    {
+      "epoch": 0.316467872258561,
+      "grad_norm": 0.8739410984824524,
+      "learning_rate": 1.6001162365252348e-05,
+      "loss": 0.8642,
+      "step": 1645
+    },
+    {
+      "epoch": 0.3166602539438245,
+      "grad_norm": 1.001243749491582,
+      "learning_rate": 1.5996177066957787e-05,
+      "loss": 0.8246,
+      "step": 1646
+    },
+    {
+      "epoch": 0.3168526356290881,
+      "grad_norm": 0.8871491929612809,
+      "learning_rate": 1.5991189440740838e-05,
+      "loss": 0.8984,
+      "step": 1647
+    },
+    {
+      "epoch": 0.3170450173143517,
+      "grad_norm": 0.8831409759732674,
+      "learning_rate": 1.5986199488537867e-05,
+      "loss": 0.8718,
+      "step": 1648
+    },
+    {
+      "epoch": 0.3172373989996152,
+      "grad_norm": 0.8783959229798706,
+      "learning_rate": 1.598120721228614e-05,
+      "loss": 0.8676,
+      "step": 1649
+    },
+    {
+      "epoch": 0.3174297806848788,
+      "grad_norm": 0.9229355652976761,
+      "learning_rate": 1.5976212613923836e-05,
+      "loss": 0.8621,
+      "step": 1650
+    },
+    {
+      "epoch": 0.31762216237014235,
+      "grad_norm": 0.8524320236061053,
+      "learning_rate": 1.5971215695390026e-05,
+      "loss": 0.8369,
+      "step": 1651
+    },
+    {
+      "epoch": 0.31781454405540593,
+      "grad_norm": 0.8942636569315422,
+      "learning_rate": 1.5966216458624692e-05,
+      "loss": 0.8462,
+      "step": 1652
+    },
+    {
+      "epoch": 0.31800692574066947,
+      "grad_norm": 0.949062986125999,
+      "learning_rate": 1.5961214905568705e-05,
+      "loss": 0.908,
+      "step": 1653
+    },
+    {
+      "epoch": 0.31819930742593305,
+      "grad_norm": 0.9363328185584795,
+      "learning_rate": 1.595621103816384e-05,
+      "loss": 0.8782,
+      "step": 1654
+    },
+    {
+      "epoch": 0.31839168911119664,
+      "grad_norm": 0.819700955510055,
+      "learning_rate": 1.5951204858352772e-05,
+      "loss": 0.8169,
+      "step": 1655
+    },
+    {
+      "epoch": 0.3185840707964602,
+      "grad_norm": 0.8836009221047447,
+      "learning_rate": 1.594619636807907e-05,
+      "loss": 0.8982,
+      "step": 1656
+    },
+    {
+      "epoch": 0.31877645248172376,
+      "grad_norm": 0.9463202956028015,
+      "learning_rate": 1.5941185569287206e-05,
+      "loss": 0.8442,
+      "step": 1657
+    },
+    {
+      "epoch": 0.3189688341669873,
+      "grad_norm": 0.964744660341475,
+      "learning_rate": 1.5936172463922542e-05,
+      "loss": 0.8237,
+      "step": 1658
+    },
+    {
+      "epoch": 0.3191612158522509,
+      "grad_norm": 0.875513162279359,
+      "learning_rate": 1.593115705393134e-05,
+      "loss": 0.8772,
+      "step": 1659
+    },
+    {
+      "epoch": 0.3193535975375144,
+      "grad_norm": 0.877512623113615,
+      "learning_rate": 1.5926139341260755e-05,
+      "loss": 0.8897,
+      "step": 1660
+    },
+    {
+      "epoch": 0.319545979222778,
+      "grad_norm": 0.8787368640117373,
+      "learning_rate": 1.5921119327858835e-05,
+      "loss": 0.8777,
+      "step": 1661
+    },
+    {
+      "epoch": 0.31973836090804153,
+      "grad_norm": 0.9597172699515409,
+      "learning_rate": 1.5916097015674518e-05,
+      "loss": 0.8808,
+      "step": 1662
+    },
+    {
+      "epoch": 0.3199307425933051,
+      "grad_norm": 0.8075046342239901,
+      "learning_rate": 1.5911072406657646e-05,
+      "loss": 0.8372,
+      "step": 1663
+    },
+    {
+      "epoch": 0.3201231242785687,
+      "grad_norm": 0.9152409916697921,
+      "learning_rate": 1.5906045502758943e-05,
+      "loss": 0.8078,
+      "step": 1664
+    },
+    {
+      "epoch": 0.32031550596383224,
+      "grad_norm": 0.8574698053182739,
+      "learning_rate": 1.590101630593002e-05,
+      "loss": 0.9421,
+      "step": 1665
+    },
+    {
+      "epoch": 0.3205078876490958,
+      "grad_norm": 0.865729206914995,
+      "learning_rate": 1.5895984818123392e-05,
+      "loss": 0.8621,
+      "step": 1666
+    },
+    {
+      "epoch": 0.32070026933435936,
+      "grad_norm": 0.9065657717112973,
+      "learning_rate": 1.5890951041292453e-05,
+      "loss": 0.8461,
+      "step": 1667
+    },
+    {
+      "epoch": 0.32089265101962294,
+      "grad_norm": 0.8792788462039044,
+      "learning_rate": 1.588591497739149e-05,
+      "loss": 0.813,
+      "step": 1668
+    },
+    {
+      "epoch": 0.3210850327048865,
+      "grad_norm": 0.8295974847556616,
+      "learning_rate": 1.5880876628375668e-05,
+      "loss": 0.8259,
+      "step": 1669
+    },
+    {
+      "epoch": 0.32127741439015006,
+      "grad_norm": 0.8401646591174956,
+      "learning_rate": 1.587583599620106e-05,
+      "loss": 0.8085,
+      "step": 1670
+    },
+    {
+      "epoch": 0.3214697960754136,
+      "grad_norm": 0.9369706894186589,
+      "learning_rate": 1.5870793082824604e-05,
+      "loss": 0.855,
+      "step": 1671
+    },
+    {
+      "epoch": 0.3216621777606772,
+      "grad_norm": 0.8804328733541004,
+      "learning_rate": 1.5865747890204138e-05,
+      "loss": 0.3666,
+      "step": 1672
+    },
+    {
+      "epoch": 0.32185455944594077,
+      "grad_norm": 0.8374424741669184,
+      "learning_rate": 1.5860700420298377e-05,
+      "loss": 0.8778,
+      "step": 1673
+    },
+    {
+      "epoch": 0.3220469411312043,
+      "grad_norm": 0.9092838178062409,
+      "learning_rate": 1.5855650675066924e-05,
+      "loss": 0.8515,
+      "step": 1674
+    },
+    {
+      "epoch": 0.3222393228164679,
+      "grad_norm": 0.896689753263929,
+      "learning_rate": 1.5850598656470265e-05,
+      "loss": 0.835,
+      "step": 1675
+    },
+    {
+      "epoch": 0.3224317045017314,
+      "grad_norm": 0.9560378722855755,
+      "learning_rate": 1.584554436646976e-05,
+      "loss": 0.8735,
+      "step": 1676
+    },
+    {
+      "epoch": 0.322624086186995,
+      "grad_norm": 0.9331066455513892,
+      "learning_rate": 1.5840487807027665e-05,
+      "loss": 0.8398,
+      "step": 1677
+    },
+    {
+      "epoch": 0.32281646787225854,
+      "grad_norm": 0.8409364748501198,
+      "learning_rate": 1.5835428980107113e-05,
+      "loss": 0.8798,
+      "step": 1678
+    },
+    {
+      "epoch": 0.3230088495575221,
+      "grad_norm": 1.0349121023928163,
+      "learning_rate": 1.583036788767211e-05,
+      "loss": 0.8465,
+      "step": 1679
+    },
+    {
+      "epoch": 0.3232012312427857,
+      "grad_norm": 0.8849167468723924,
+      "learning_rate": 1.5825304531687548e-05,
+      "loss": 0.8318,
+      "step": 1680
+    },
+    {
+      "epoch": 0.32339361292804925,
+      "grad_norm": 0.9424612778860867,
+      "learning_rate": 1.5820238914119195e-05,
+      "loss": 0.7765,
+      "step": 1681
+    },
+    {
+      "epoch": 0.32358599461331283,
+      "grad_norm": 0.8838262854766268,
+      "learning_rate": 1.5815171036933697e-05,
+      "loss": 0.8989,
+      "step": 1682
+    },
+    {
+      "epoch": 0.32377837629857636,
+      "grad_norm": 0.9225735745982718,
+      "learning_rate": 1.5810100902098582e-05,
+      "loss": 0.8873,
+      "step": 1683
+    },
+    {
+      "epoch": 0.32397075798383995,
+      "grad_norm": 0.8589047548627686,
+      "learning_rate": 1.580502851158225e-05,
+      "loss": 0.863,
+      "step": 1684
+    },
+    {
+      "epoch": 0.3241631396691035,
+      "grad_norm": 0.9235520751868377,
+      "learning_rate": 1.5799953867353975e-05,
+      "loss": 0.8804,
+      "step": 1685
+    },
+    {
+      "epoch": 0.32435552135436707,
+      "grad_norm": 0.9329601296325627,
+      "learning_rate": 1.579487697138391e-05,
+      "loss": 0.8021,
+      "step": 1686
+    },
+    {
+      "epoch": 0.3245479030396306,
+      "grad_norm": 0.8737320717034389,
+      "learning_rate": 1.5789797825643086e-05,
+      "loss": 0.8814,
+      "step": 1687
+    },
+    {
+      "epoch": 0.3247402847248942,
+      "grad_norm": 0.915046276452217,
+      "learning_rate": 1.5784716432103394e-05,
+      "loss": 0.9324,
+      "step": 1688
+    },
+    {
+      "epoch": 0.3249326664101578,
+      "grad_norm": 0.9268093464545314,
+      "learning_rate": 1.5779632792737608e-05,
+      "loss": 0.8596,
+      "step": 1689
+    },
+    {
+      "epoch": 0.3251250480954213,
+      "grad_norm": 0.8269781271101063,
+      "learning_rate": 1.5774546909519376e-05,
+      "loss": 0.8274,
+      "step": 1690
+    },
+    {
+      "epoch": 0.3253174297806849,
+      "grad_norm": 1.0075038719181226,
+      "learning_rate": 1.5769458784423206e-05,
+      "loss": 0.8698,
+      "step": 1691
+    },
+    {
+      "epoch": 0.32550981146594843,
+      "grad_norm": 0.856168568307091,
+      "learning_rate": 1.5764368419424488e-05,
+      "loss": 0.8774,
+      "step": 1692
+    },
+    {
+      "epoch": 0.325702193151212,
+      "grad_norm": 0.8941385879777548,
+      "learning_rate": 1.575927581649948e-05,
+      "loss": 0.8979,
+      "step": 1693
+    },
+    {
+      "epoch": 0.32589457483647555,
+      "grad_norm": 0.9305455249771228,
+      "learning_rate": 1.5754180977625303e-05,
+      "loss": 0.8281,
+      "step": 1694
+    },
+    {
+      "epoch": 0.32608695652173914,
+      "grad_norm": 0.8646406536803535,
+      "learning_rate": 1.574908390477995e-05,
+      "loss": 0.8694,
+      "step": 1695
+    },
+    {
+      "epoch": 0.32627933820700267,
+      "grad_norm": 0.866238484426705,
+      "learning_rate": 1.5743984599942273e-05,
+      "loss": 0.8504,
+      "step": 1696
+    },
+    {
+      "epoch": 0.32647171989226625,
+      "grad_norm": 0.873434164448445,
+      "learning_rate": 1.5738883065092005e-05,
+      "loss": 0.8145,
+      "step": 1697
+    },
+    {
+      "epoch": 0.32666410157752984,
+      "grad_norm": 0.9041768597080385,
+      "learning_rate": 1.5733779302209735e-05,
+      "loss": 0.8299,
+      "step": 1698
+    },
+    {
+      "epoch": 0.3268564832627934,
+      "grad_norm": 0.9295722880545448,
+      "learning_rate": 1.572867331327692e-05,
+      "loss": 0.8704,
+      "step": 1699
+    },
+    {
+      "epoch": 0.32704886494805696,
+      "grad_norm": 0.9380987971137182,
+      "learning_rate": 1.5723565100275884e-05,
+      "loss": 0.8731,
+      "step": 1700
+    },
+    {
+      "epoch": 0.3272412466333205,
+      "grad_norm": 0.9093126889736369,
+      "learning_rate": 1.5718454665189806e-05,
+      "loss": 0.8168,
+      "step": 1701
+    },
+    {
+      "epoch": 0.3274336283185841,
+      "grad_norm": 0.9053278209217086,
+      "learning_rate": 1.5713342010002733e-05,
+      "loss": 0.8144,
+      "step": 1702
+    },
+    {
+      "epoch": 0.3276260100038476,
+      "grad_norm": 0.8955610082142926,
+      "learning_rate": 1.5708227136699578e-05,
+      "loss": 0.7963,
+      "step": 1703
+    },
+    {
+      "epoch": 0.3278183916891112,
+      "grad_norm": 0.920873103818156,
+      "learning_rate": 1.5703110047266105e-05,
+      "loss": 0.8315,
+      "step": 1704
+    },
+    {
+      "epoch": 0.3280107733743748,
+      "grad_norm": 0.8686962108762084,
+      "learning_rate": 1.569799074368895e-05,
+      "loss": 0.8171,
+      "step": 1705
+    },
+    {
+      "epoch": 0.3282031550596383,
+      "grad_norm": 0.973411526351427,
+      "learning_rate": 1.5692869227955603e-05,
+      "loss": 0.7726,
+      "step": 1706
+    },
+    {
+      "epoch": 0.3283955367449019,
+      "grad_norm": 0.9498386123472706,
+      "learning_rate": 1.5687745502054407e-05,
+      "loss": 0.8618,
+      "step": 1707
+    },
+    {
+      "epoch": 0.32858791843016544,
+      "grad_norm": 0.9338635624768461,
+      "learning_rate": 1.5682619567974575e-05,
+      "loss": 0.9103,
+      "step": 1708
+    },
+    {
+      "epoch": 0.328780300115429,
+      "grad_norm": 0.8741443625106621,
+      "learning_rate": 1.567749142770617e-05,
+      "loss": 0.846,
+      "step": 1709
+    },
+    {
+      "epoch": 0.32897268180069256,
+      "grad_norm": 0.9585101994877012,
+      "learning_rate": 1.5672361083240106e-05,
+      "loss": 0.8865,
+      "step": 1710
+    },
+    {
+      "epoch": 0.32916506348595614,
+      "grad_norm": 0.8740628946027289,
+      "learning_rate": 1.5667228536568167e-05,
+      "loss": 0.3635,
+      "step": 1711
+    },
+    {
+      "epoch": 0.3293574451712197,
+      "grad_norm": 0.8830909188081387,
+      "learning_rate": 1.566209378968298e-05,
+      "loss": 0.7963,
+      "step": 1712
+    },
+    {
+      "epoch": 0.32954982685648326,
+      "grad_norm": 0.9023212421662575,
+      "learning_rate": 1.565695684457803e-05,
+      "loss": 0.8973,
+      "step": 1713
+    },
+    {
+      "epoch": 0.32974220854174685,
+      "grad_norm": 0.9192547981795229,
+      "learning_rate": 1.5651817703247666e-05,
+      "loss": 0.8638,
+      "step": 1714
+    },
+    {
+      "epoch": 0.3299345902270104,
+      "grad_norm": 0.9018799108713677,
+      "learning_rate": 1.5646676367687067e-05,
+      "loss": 0.8137,
+      "step": 1715
+    },
+    {
+      "epoch": 0.33012697191227397,
+      "grad_norm": 0.6901453116360025,
+      "learning_rate": 1.564153283989228e-05,
+      "loss": 0.3539,
+      "step": 1716
+    },
+    {
+      "epoch": 0.3303193535975375,
+      "grad_norm": 0.9158331402627286,
+      "learning_rate": 1.5636387121860207e-05,
+      "loss": 0.7641,
+      "step": 1717
+    },
+    {
+      "epoch": 0.3305117352828011,
+      "grad_norm": 0.9609275923494512,
+      "learning_rate": 1.5631239215588578e-05,
+      "loss": 0.8139,
+      "step": 1718
+    },
+    {
+      "epoch": 0.3307041169680646,
+      "grad_norm": 0.9394087053980968,
+      "learning_rate": 1.5626089123076004e-05,
+      "loss": 0.8895,
+      "step": 1719
+    },
+    {
+      "epoch": 0.3308964986533282,
+      "grad_norm": 0.831347534746285,
+      "learning_rate": 1.5620936846321917e-05,
+      "loss": 0.8287,
+      "step": 1720
+    },
+    {
+      "epoch": 0.33108888033859174,
+      "grad_norm": 0.9161878968295166,
+      "learning_rate": 1.561578238732661e-05,
+      "loss": 0.8281,
+      "step": 1721
+    },
+    {
+      "epoch": 0.3312812620238553,
+      "grad_norm": 0.9139106082223993,
+      "learning_rate": 1.561062574809123e-05,
+      "loss": 0.8792,
+      "step": 1722
+    },
+    {
+      "epoch": 0.3314736437091189,
+      "grad_norm": 0.8501671169500225,
+      "learning_rate": 1.5605466930617747e-05,
+      "loss": 0.8536,
+      "step": 1723
+    },
+    {
+      "epoch": 0.33166602539438245,
+      "grad_norm": 0.8340679855342941,
+      "learning_rate": 1.5600305936909005e-05,
+      "loss": 0.9132,
+      "step": 1724
+    },
+    {
+      "epoch": 0.33185840707964603,
+      "grad_norm": 0.9729830900681893,
+      "learning_rate": 1.559514276896867e-05,
+      "loss": 0.9079,
+      "step": 1725
+    },
+    {
+      "epoch": 0.33205078876490957,
+      "grad_norm": 0.8645480615107756,
+      "learning_rate": 1.558997742880127e-05,
+      "loss": 0.8466,
+      "step": 1726
+    },
+    {
+      "epoch": 0.33224317045017315,
+      "grad_norm": 0.8178548254105622,
+      "learning_rate": 1.5584809918412158e-05,
+      "loss": 0.8008,
+      "step": 1727
+    },
+    {
+      "epoch": 0.3324355521354367,
+      "grad_norm": 0.8804136245621014,
+      "learning_rate": 1.557964023980755e-05,
+      "loss": 0.3548,
+      "step": 1728
+    },
+    {
+      "epoch": 0.3326279338207003,
+      "grad_norm": 0.8676715956796711,
+      "learning_rate": 1.5574468394994486e-05,
+      "loss": 0.8556,
+      "step": 1729
+    },
+    {
+      "epoch": 0.33282031550596386,
+      "grad_norm": 0.8632095660499571,
+      "learning_rate": 1.5569294385980856e-05,
+      "loss": 0.7983,
+      "step": 1730
+    },
+    {
+      "epoch": 0.3330126971912274,
+      "grad_norm": 0.9051007502317857,
+      "learning_rate": 1.556411821477539e-05,
+      "loss": 0.8673,
+      "step": 1731
+    },
+    {
+      "epoch": 0.333205078876491,
+      "grad_norm": 0.9359132161707349,
+      "learning_rate": 1.5558939883387657e-05,
+      "loss": 0.7683,
+      "step": 1732
+    },
+    {
+      "epoch": 0.3333974605617545,
+      "grad_norm": 0.8779044794846749,
+      "learning_rate": 1.5553759393828058e-05,
+      "loss": 0.8988,
+      "step": 1733
+    },
+    {
+      "epoch": 0.3335898422470181,
+      "grad_norm": 0.9083788784219076,
+      "learning_rate": 1.554857674810784e-05,
+      "loss": 0.8661,
+      "step": 1734
+    },
+    {
+      "epoch": 0.33378222393228163,
+      "grad_norm": 0.8932650873155474,
+      "learning_rate": 1.554339194823909e-05,
+      "loss": 0.7461,
+      "step": 1735
+    },
+    {
+      "epoch": 0.3339746056175452,
+      "grad_norm": 0.874439842110843,
+      "learning_rate": 1.553820499623472e-05,
+      "loss": 0.7842,
+      "step": 1736
+    },
+    {
+      "epoch": 0.33416698730280875,
+      "grad_norm": 0.8691549979863386,
+      "learning_rate": 1.553301589410848e-05,
+      "loss": 0.8444,
+      "step": 1737
+    },
+    {
+      "epoch": 0.33435936898807234,
+      "grad_norm": 0.9382418775151821,
+      "learning_rate": 1.5527824643874968e-05,
+      "loss": 0.8032,
+      "step": 1738
+    },
+    {
+      "epoch": 0.3345517506733359,
+      "grad_norm": 0.8756506819762452,
+      "learning_rate": 1.5522631247549598e-05,
+      "loss": 0.8645,
+      "step": 1739
+    },
+    {
+      "epoch": 0.33474413235859946,
+      "grad_norm": 0.8930051283214528,
+      "learning_rate": 1.5517435707148628e-05,
+      "loss": 0.9081,
+      "step": 1740
+    },
+    {
+      "epoch": 0.33493651404386304,
+      "grad_norm": 0.8963293956250221,
+      "learning_rate": 1.5512238024689144e-05,
+      "loss": 0.8525,
+      "step": 1741
+    },
+    {
+      "epoch": 0.3351288957291266,
+      "grad_norm": 0.9092678349030451,
+      "learning_rate": 1.550703820218907e-05,
+      "loss": 0.8669,
+      "step": 1742
+    },
+    {
+      "epoch": 0.33532127741439016,
+      "grad_norm": 0.9950797529390312,
+      "learning_rate": 1.550183624166715e-05,
+      "loss": 0.8955,
+      "step": 1743
+    },
+    {
+      "epoch": 0.3355136590996537,
+      "grad_norm": 0.907892443531239,
+      "learning_rate": 1.549663214514297e-05,
+      "loss": 0.791,
+      "step": 1744
+    },
+    {
+      "epoch": 0.3357060407849173,
+      "grad_norm": 0.8180788059320417,
+      "learning_rate": 1.5491425914636934e-05,
+      "loss": 0.8708,
+      "step": 1745
+    },
+    {
+      "epoch": 0.3358984224701808,
+      "grad_norm": 0.8905920159587426,
+      "learning_rate": 1.5486217552170283e-05,
+      "loss": 0.8577,
+      "step": 1746
+    },
+    {
+      "epoch": 0.3360908041554444,
+      "grad_norm": 0.8979892120457673,
+      "learning_rate": 1.548100705976508e-05,
+      "loss": 0.9044,
+      "step": 1747
+    },
+    {
+      "epoch": 0.336283185840708,
+      "grad_norm": 0.8420577417379433,
+      "learning_rate": 1.5475794439444226e-05,
+      "loss": 0.8117,
+      "step": 1748
+    },
+    {
+      "epoch": 0.3364755675259715,
+      "grad_norm": 0.9086739723127221,
+      "learning_rate": 1.5470579693231432e-05,
+      "loss": 0.8469,
+      "step": 1749
+    },
+    {
+      "epoch": 0.3366679492112351,
+      "grad_norm": 0.8802985998293118,
+      "learning_rate": 1.5465362823151245e-05,
+      "loss": 0.8794,
+      "step": 1750
+    },
+    {
+      "epoch": 0.33686033089649864,
+      "grad_norm": 1.0052215261022055,
+      "learning_rate": 1.5460143831229026e-05,
+      "loss": 0.9013,
+      "step": 1751
+    },
+    {
+      "epoch": 0.3370527125817622,
+      "grad_norm": 0.866544485503468,
+      "learning_rate": 1.545492271949098e-05,
+      "loss": 0.8661,
+      "step": 1752
+    },
+    {
+      "epoch": 0.33724509426702576,
+      "grad_norm": 0.797262818008557,
+      "learning_rate": 1.544969948996411e-05,
+      "loss": 0.787,
+      "step": 1753
+    },
+    {
+      "epoch": 0.33743747595228935,
+      "grad_norm": 0.9521750695511414,
+      "learning_rate": 1.544447414467626e-05,
+      "loss": 0.8501,
+      "step": 1754
+    },
+    {
+      "epoch": 0.33762985763755293,
+      "grad_norm": 0.879706940214121,
+      "learning_rate": 1.5439246685656093e-05,
+      "loss": 0.8276,
+      "step": 1755
+    },
+    {
+      "epoch": 0.33782223932281646,
+      "grad_norm": 0.9151184543366531,
+      "learning_rate": 1.5434017114933082e-05,
+      "loss": 0.888,
+      "step": 1756
+    },
+    {
+      "epoch": 0.33801462100808005,
+      "grad_norm": 0.8646817801379176,
+      "learning_rate": 1.5428785434537527e-05,
+      "loss": 0.8544,
+      "step": 1757
+    },
+    {
+      "epoch": 0.3382070026933436,
+      "grad_norm": 0.8845261598699534,
+      "learning_rate": 1.542355164650055e-05,
+      "loss": 0.8171,
+      "step": 1758
+    },
+    {
+      "epoch": 0.33839938437860717,
+      "grad_norm": 0.9105281750010585,
+      "learning_rate": 1.541831575285408e-05,
+      "loss": 0.8802,
+      "step": 1759
+    },
+    {
+      "epoch": 0.3385917660638707,
+      "grad_norm": 0.876535841959437,
+      "learning_rate": 1.541307775563088e-05,
+      "loss": 0.8429,
+      "step": 1760
+    },
+    {
+      "epoch": 0.3387841477491343,
+      "grad_norm": 0.9365036817973641,
+      "learning_rate": 1.540783765686452e-05,
+      "loss": 0.8952,
+      "step": 1761
+    },
+    {
+      "epoch": 0.3389765294343978,
+      "grad_norm": 0.9745934008517372,
+      "learning_rate": 1.540259545858938e-05,
+      "loss": 0.8915,
+      "step": 1762
+    },
+    {
+      "epoch": 0.3391689111196614,
+      "grad_norm": 0.9914802125226702,
+      "learning_rate": 1.539735116284067e-05,
+      "loss": 0.7948,
+      "step": 1763
+    },
+    {
+      "epoch": 0.339361292804925,
+      "grad_norm": 0.8794990505987774,
+      "learning_rate": 1.53921047716544e-05,
+      "loss": 0.7922,
+      "step": 1764
+    },
+    {
+      "epoch": 0.33955367449018853,
+      "grad_norm": 0.925068329867901,
+      "learning_rate": 1.53868562870674e-05,
+      "loss": 0.8697,
+      "step": 1765
+    },
+    {
+      "epoch": 0.3397460561754521,
+      "grad_norm": 0.7229753998586073,
+      "learning_rate": 1.5381605711117318e-05,
+      "loss": 0.3578,
+      "step": 1766
+    },
+    {
+      "epoch": 0.33993843786071565,
+      "grad_norm": 0.9614745264928451,
+      "learning_rate": 1.5376353045842604e-05,
+      "loss": 0.8991,
+      "step": 1767
+    },
+    {
+      "epoch": 0.34013081954597923,
+      "grad_norm": 0.870826136679754,
+      "learning_rate": 1.5371098293282526e-05,
+      "loss": 0.8815,
+      "step": 1768
+    },
+    {
+      "epoch": 0.34032320123124277,
+      "grad_norm": 0.9211051180356321,
+      "learning_rate": 1.5365841455477158e-05,
+      "loss": 0.8398,
+      "step": 1769
+    },
+    {
+      "epoch": 0.34051558291650635,
+      "grad_norm": 0.9854125374648295,
+      "learning_rate": 1.5360582534467382e-05,
+      "loss": 0.8279,
+      "step": 1770
+    },
+    {
+      "epoch": 0.3407079646017699,
+      "grad_norm": 0.5942428120971539,
+      "learning_rate": 1.5355321532294897e-05,
+      "loss": 0.3577,
+      "step": 1771
+    },
+    {
+      "epoch": 0.3409003462870335,
+      "grad_norm": 0.9355958993417136,
+      "learning_rate": 1.5350058451002204e-05,
+      "loss": 0.797,
+      "step": 1772
+    },
+    {
+      "epoch": 0.34109272797229706,
+      "grad_norm": 0.8831949770698221,
+      "learning_rate": 1.5344793292632614e-05,
+      "loss": 0.7836,
+      "step": 1773
+    },
+    {
+      "epoch": 0.3412851096575606,
+      "grad_norm": 0.8944467183387402,
+      "learning_rate": 1.533952605923024e-05,
+      "loss": 0.8255,
+      "step": 1774
+    },
+    {
+      "epoch": 0.3414774913428242,
+      "grad_norm": 0.9063309068399807,
+      "learning_rate": 1.5334256752840007e-05,
+      "loss": 0.857,
+      "step": 1775
+    },
+    {
+      "epoch": 0.3416698730280877,
+      "grad_norm": 0.8822000746789396,
+      "learning_rate": 1.532898537550764e-05,
+      "loss": 0.8305,
+      "step": 1776
+    },
+    {
+      "epoch": 0.3418622547133513,
+      "grad_norm": 0.850174047989685,
+      "learning_rate": 1.532371192927966e-05,
+      "loss": 0.7579,
+      "step": 1777
+    },
+    {
+      "epoch": 0.34205463639861483,
+      "grad_norm": 0.8171310939771944,
+      "learning_rate": 1.5318436416203412e-05,
+      "loss": 0.8603,
+      "step": 1778
+    },
+    {
+      "epoch": 0.3422470180838784,
+      "grad_norm": 0.8402765400742352,
+      "learning_rate": 1.531315883832703e-05,
+      "loss": 0.7791,
+      "step": 1779
+    },
+    {
+      "epoch": 0.34243939976914195,
+      "grad_norm": 0.7228024133046372,
+      "learning_rate": 1.530787919769945e-05,
+      "loss": 0.3768,
+      "step": 1780
+    },
+    {
+      "epoch": 0.34263178145440554,
+      "grad_norm": 0.8718702895992549,
+      "learning_rate": 1.5302597496370408e-05,
+      "loss": 0.8009,
+      "step": 1781
+    },
+    {
+      "epoch": 0.3428241631396691,
+      "grad_norm": 0.8746291376719535,
+      "learning_rate": 1.5297313736390447e-05,
+      "loss": 0.8652,
+      "step": 1782
+    },
+    {
+      "epoch": 0.34301654482493266,
+      "grad_norm": 0.8831948071278526,
+      "learning_rate": 1.5292027919810898e-05,
+      "loss": 0.8116,
+      "step": 1783
+    },
+    {
+      "epoch": 0.34320892651019624,
+      "grad_norm": 0.8384772128398085,
+      "learning_rate": 1.52867400486839e-05,
+      "loss": 0.8238,
+      "step": 1784
+    },
+    {
+      "epoch": 0.3434013081954598,
+      "grad_norm": 0.893212792010188,
+      "learning_rate": 1.528145012506239e-05,
+      "loss": 0.8641,
+      "step": 1785
+    },
+    {
+      "epoch": 0.34359368988072336,
+      "grad_norm": 0.635076602915607,
+      "learning_rate": 1.5276158151000096e-05,
+      "loss": 0.3456,
+      "step": 1786
+    },
+    {
+      "epoch": 0.3437860715659869,
+      "grad_norm": 0.8769970631496764,
+      "learning_rate": 1.5270864128551542e-05,
+      "loss": 0.8284,
+      "step": 1787
+    },
+    {
+      "epoch": 0.3439784532512505,
+      "grad_norm": 0.8488021935400117,
+      "learning_rate": 1.5265568059772053e-05,
+      "loss": 0.8572,
+      "step": 1788
+    },
+    {
+      "epoch": 0.34417083493651407,
+      "grad_norm": 0.9056645392911827,
+      "learning_rate": 1.5260269946717746e-05,
+      "loss": 0.8622,
+      "step": 1789
+    },
+    {
+      "epoch": 0.3443632166217776,
+      "grad_norm": 0.830858024854033,
+      "learning_rate": 1.5254969791445526e-05,
+      "loss": 0.7894,
+      "step": 1790
+    },
+    {
+      "epoch": 0.3445555983070412,
+      "grad_norm": 0.9260235616614961,
+      "learning_rate": 1.5249667596013102e-05,
+      "loss": 0.827,
+      "step": 1791
+    },
+    {
+      "epoch": 0.3447479799923047,
+      "grad_norm": 0.8337089717299104,
+      "learning_rate": 1.5244363362478967e-05,
+      "loss": 0.8542,
+      "step": 1792
+    },
+    {
+      "epoch": 0.3449403616775683,
+      "grad_norm": 0.8400887325285339,
+      "learning_rate": 1.5239057092902404e-05,
+      "loss": 0.8386,
+      "step": 1793
+    },
+    {
+      "epoch": 0.34513274336283184,
+      "grad_norm": 0.9088056872074686,
+      "learning_rate": 1.523374878934349e-05,
+      "loss": 0.8622,
+      "step": 1794
+    },
+    {
+      "epoch": 0.3453251250480954,
+      "grad_norm": 0.8369547253430362,
+      "learning_rate": 1.5228438453863095e-05,
+      "loss": 0.8311,
+      "step": 1795
+    },
+    {
+      "epoch": 0.34551750673335896,
+      "grad_norm": 0.913388440945517,
+      "learning_rate": 1.522312608852287e-05,
+      "loss": 0.8975,
+      "step": 1796
+    },
+    {
+      "epoch": 0.34570988841862255,
+      "grad_norm": 0.83503103967194,
+      "learning_rate": 1.5217811695385263e-05,
+      "loss": 0.8815,
+      "step": 1797
+    },
+    {
+      "epoch": 0.34590227010388613,
+      "grad_norm": 0.6634096560008125,
+      "learning_rate": 1.52124952765135e-05,
+      "loss": 0.3829,
+      "step": 1798
+    },
+    {
+      "epoch": 0.34609465178914967,
+      "grad_norm": 0.6285687381682104,
+      "learning_rate": 1.5207176833971598e-05,
+      "loss": 0.349,
+      "step": 1799
+    },
+    {
+      "epoch": 0.34628703347441325,
+      "grad_norm": 0.8620997314230443,
+      "learning_rate": 1.520185636982436e-05,
+      "loss": 0.8682,
+      "step": 1800
+    },
+    {
+      "epoch": 0.3464794151596768,
+      "grad_norm": 0.8892423962265654,
+      "learning_rate": 1.5196533886137376e-05,
+      "loss": 0.8264,
+      "step": 1801
+    },
+    {
+      "epoch": 0.34667179684494037,
+      "grad_norm": 0.8671122477816628,
+      "learning_rate": 1.5191209384977014e-05,
+      "loss": 0.8363,
+      "step": 1802
+    },
+    {
+      "epoch": 0.3468641785302039,
+      "grad_norm": 0.6698092854103417,
+      "learning_rate": 1.5185882868410431e-05,
+      "loss": 0.3496,
+      "step": 1803
+    },
+    {
+      "epoch": 0.3470565602154675,
+      "grad_norm": 0.9243918569192682,
+      "learning_rate": 1.5180554338505564e-05,
+      "loss": 0.8034,
+      "step": 1804
+    },
+    {
+      "epoch": 0.347248941900731,
+      "grad_norm": 0.8264204102693415,
+      "learning_rate": 1.517522379733113e-05,
+      "loss": 0.8277,
+      "step": 1805
+    },
+    {
+      "epoch": 0.3474413235859946,
+      "grad_norm": 0.890324077151211,
+      "learning_rate": 1.5169891246956629e-05,
+      "loss": 0.8146,
+      "step": 1806
+    },
+    {
+      "epoch": 0.3476337052712582,
+      "grad_norm": 0.8998994446966276,
+      "learning_rate": 1.5164556689452346e-05,
+      "loss": 0.7863,
+      "step": 1807
+    },
+    {
+      "epoch": 0.34782608695652173,
+      "grad_norm": 0.9338529179573043,
+      "learning_rate": 1.5159220126889329e-05,
+      "loss": 0.8638,
+      "step": 1808
+    },
+    {
+      "epoch": 0.3480184686417853,
+      "grad_norm": 0.9195539744743201,
+      "learning_rate": 1.5153881561339426e-05,
+      "loss": 0.8312,
+      "step": 1809
+    },
+    {
+      "epoch": 0.34821085032704885,
+      "grad_norm": 1.2803802390659802,
+      "learning_rate": 1.5148540994875242e-05,
+      "loss": 0.8777,
+      "step": 1810
+    },
+    {
+      "epoch": 0.34840323201231244,
+      "grad_norm": 0.6780858581890885,
+      "learning_rate": 1.5143198429570181e-05,
+      "loss": 0.3535,
+      "step": 1811
+    },
+    {
+      "epoch": 0.34859561369757597,
+      "grad_norm": 0.8866069800220362,
+      "learning_rate": 1.5137853867498403e-05,
+      "loss": 0.8744,
+      "step": 1812
+    },
+    {
+      "epoch": 0.34878799538283956,
+      "grad_norm": 0.8786556164013535,
+      "learning_rate": 1.5132507310734847e-05,
+      "loss": 0.9069,
+      "step": 1813
+    },
+    {
+      "epoch": 0.34898037706810314,
+      "grad_norm": 0.8573495040906238,
+      "learning_rate": 1.5127158761355241e-05,
+      "loss": 0.8229,
+      "step": 1814
+    },
+    {
+      "epoch": 0.3491727587533667,
+      "grad_norm": 0.8797845192266451,
+      "learning_rate": 1.512180822143607e-05,
+      "loss": 0.8507,
+      "step": 1815
+    },
+    {
+      "epoch": 0.34936514043863026,
+      "grad_norm": 0.9461631271590555,
+      "learning_rate": 1.5116455693054594e-05,
+      "loss": 0.8323,
+      "step": 1816
+    },
+    {
+      "epoch": 0.3495575221238938,
+      "grad_norm": 0.8784039124919184,
+      "learning_rate": 1.5111101178288858e-05,
+      "loss": 0.8418,
+      "step": 1817
+    },
+    {
+      "epoch": 0.3497499038091574,
+      "grad_norm": 0.8917214525207184,
+      "learning_rate": 1.510574467921766e-05,
+      "loss": 0.7563,
+      "step": 1818
+    },
+    {
+      "epoch": 0.3499422854944209,
+      "grad_norm": 0.8802231387611068,
+      "learning_rate": 1.5100386197920585e-05,
+      "loss": 0.8802,
+      "step": 1819
+    },
+    {
+      "epoch": 0.3501346671796845,
+      "grad_norm": 0.8419845221628783,
+      "learning_rate": 1.5095025736477977e-05,
+      "loss": 0.8262,
+      "step": 1820
+    },
+    {
+      "epoch": 0.35032704886494803,
+      "grad_norm": 0.9171206700787049,
+      "learning_rate": 1.5089663296970952e-05,
+      "loss": 0.8018,
+      "step": 1821
+    },
+    {
+      "epoch": 0.3505194305502116,
+      "grad_norm": 0.946660901636356,
+      "learning_rate": 1.5084298881481388e-05,
+      "loss": 0.8481,
+      "step": 1822
+    },
+    {
+      "epoch": 0.3507118122354752,
+      "grad_norm": 0.8739199327508707,
+      "learning_rate": 1.5078932492091942e-05,
+      "loss": 0.8777,
+      "step": 1823
+    },
+    {
+      "epoch": 0.35090419392073874,
+      "grad_norm": 0.9728229021393214,
+      "learning_rate": 1.5073564130886032e-05,
+      "loss": 0.7598,
+      "step": 1824
+    },
+    {
+      "epoch": 0.3510965756060023,
+      "grad_norm": 0.819712046560312,
+      "learning_rate": 1.506819379994784e-05,
+      "loss": 0.8241,
+      "step": 1825
+    },
+    {
+      "epoch": 0.35128895729126586,
+      "grad_norm": 0.8870567481965237,
+      "learning_rate": 1.5062821501362308e-05,
+      "loss": 0.8776,
+      "step": 1826
+    },
+    {
+      "epoch": 0.35148133897652944,
+      "grad_norm": 0.8772563038662087,
+      "learning_rate": 1.5057447237215152e-05,
+      "loss": 0.8068,
+      "step": 1827
+    },
+    {
+      "epoch": 0.351673720661793,
+      "grad_norm": 0.9621306582370549,
+      "learning_rate": 1.5052071009592846e-05,
+      "loss": 0.8582,
+      "step": 1828
+    },
+    {
+      "epoch": 0.35186610234705656,
+      "grad_norm": 0.8327555867390762,
+      "learning_rate": 1.5046692820582625e-05,
+      "loss": 0.8474,
+      "step": 1829
+    },
+    {
+      "epoch": 0.3520584840323201,
+      "grad_norm": 0.9035443787468379,
+      "learning_rate": 1.504131267227249e-05,
+      "loss": 0.8966,
+      "step": 1830
+    },
+    {
+      "epoch": 0.3522508657175837,
+      "grad_norm": 0.8770866241756334,
+      "learning_rate": 1.5035930566751198e-05,
+      "loss": 0.9014,
+      "step": 1831
+    },
+    {
+      "epoch": 0.35244324740284727,
+      "grad_norm": 0.8501039431811862,
+      "learning_rate": 1.5030546506108268e-05,
+      "loss": 0.8286,
+      "step": 1832
+    },
+    {
+      "epoch": 0.3526356290881108,
+      "grad_norm": 0.8680795986207568,
+      "learning_rate": 1.5025160492433976e-05,
+      "loss": 0.8485,
+      "step": 1833
+    },
+    {
+      "epoch": 0.3528280107733744,
+      "grad_norm": 0.8971870188986942,
+      "learning_rate": 1.501977252781936e-05,
+      "loss": 0.8751,
+      "step": 1834
+    },
+    {
+      "epoch": 0.3530203924586379,
+      "grad_norm": 0.8940301012548304,
+      "learning_rate": 1.5014382614356213e-05,
+      "loss": 0.7404,
+      "step": 1835
+    },
+    {
+      "epoch": 0.3532127741439015,
+      "grad_norm": 0.9197805911341448,
+      "learning_rate": 1.5008990754137088e-05,
+      "loss": 0.8041,
+      "step": 1836
+    },
+    {
+      "epoch": 0.35340515582916504,
+      "grad_norm": 0.8768534093889365,
+      "learning_rate": 1.5003596949255284e-05,
+      "loss": 0.8481,
+      "step": 1837
+    },
+    {
+      "epoch": 0.35359753751442863,
+      "grad_norm": 0.874607763990165,
+      "learning_rate": 1.4998201201804867e-05,
+      "loss": 0.8114,
+      "step": 1838
+    },
+    {
+      "epoch": 0.3537899191996922,
+      "grad_norm": 0.8891158223299255,
+      "learning_rate": 1.499280351388065e-05,
+      "loss": 0.8391,
+      "step": 1839
+    },
+    {
+      "epoch": 0.35398230088495575,
+      "grad_norm": 0.8817541832773547,
+      "learning_rate": 1.49874038875782e-05,
+      "loss": 0.833,
+      "step": 1840
+    },
+    {
+      "epoch": 0.35417468257021933,
+      "grad_norm": 1.0173790782745433,
+      "learning_rate": 1.498200232499384e-05,
+      "loss": 0.8727,
+      "step": 1841
+    },
+    {
+      "epoch": 0.35436706425548287,
+      "grad_norm": 0.8624807390394732,
+      "learning_rate": 1.4976598828224643e-05,
+      "loss": 0.7712,
+      "step": 1842
+    },
+    {
+      "epoch": 0.35455944594074645,
+      "grad_norm": 0.8927020021896733,
+      "learning_rate": 1.497119339936843e-05,
+      "loss": 0.8505,
+      "step": 1843
+    },
+    {
+      "epoch": 0.35475182762601,
+      "grad_norm": 0.8926059520333656,
+      "learning_rate": 1.4965786040523779e-05,
+      "loss": 0.8553,
+      "step": 1844
+    },
+    {
+      "epoch": 0.3549442093112736,
+      "grad_norm": 0.7449257470667174,
+      "learning_rate": 1.496037675379001e-05,
+      "loss": 0.3249,
+      "step": 1845
+    },
+    {
+      "epoch": 0.3551365909965371,
+      "grad_norm": 0.9392424037355539,
+      "learning_rate": 1.4954965541267192e-05,
+      "loss": 0.7756,
+      "step": 1846
+    },
+    {
+      "epoch": 0.3553289726818007,
+      "grad_norm": 1.020610387728308,
+      "learning_rate": 1.494955240505615e-05,
+      "loss": 0.8395,
+      "step": 1847
+    },
+    {
+      "epoch": 0.3555213543670643,
+      "grad_norm": 0.878117121161596,
+      "learning_rate": 1.494413734725844e-05,
+      "loss": 0.8127,
+      "step": 1848
+    },
+    {
+      "epoch": 0.3557137360523278,
+      "grad_norm": 0.9133483829466856,
+      "learning_rate": 1.4938720369976385e-05,
+      "loss": 0.8738,
+      "step": 1849
+    },
+    {
+      "epoch": 0.3559061177375914,
+      "grad_norm": 0.9239168936472874,
+      "learning_rate": 1.4933301475313036e-05,
+      "loss": 0.9064,
+      "step": 1850
+    },
+    {
+      "epoch": 0.35609849942285493,
+      "grad_norm": 0.8744989450589917,
+      "learning_rate": 1.4927880665372197e-05,
+      "loss": 0.8275,
+      "step": 1851
+    },
+    {
+      "epoch": 0.3562908811081185,
+      "grad_norm": 0.9340257819931126,
+      "learning_rate": 1.4922457942258411e-05,
+      "loss": 0.8787,
+      "step": 1852
+    },
+    {
+      "epoch": 0.35648326279338205,
+      "grad_norm": 0.9659559099064629,
+      "learning_rate": 1.4917033308076967e-05,
+      "loss": 0.8897,
+      "step": 1853
+    },
+    {
+      "epoch": 0.35667564447864564,
+      "grad_norm": 0.8226445878988482,
+      "learning_rate": 1.4911606764933892e-05,
+      "loss": 0.8615,
+      "step": 1854
+    },
+    {
+      "epoch": 0.35686802616390917,
+      "grad_norm": 0.7711464712616716,
+      "learning_rate": 1.490617831493596e-05,
+      "loss": 0.8194,
+      "step": 1855
+    },
+    {
+      "epoch": 0.35706040784917276,
+      "grad_norm": 0.8506478815620172,
+      "learning_rate": 1.4900747960190682e-05,
+      "loss": 0.8706,
+      "step": 1856
+    },
+    {
+      "epoch": 0.35725278953443634,
+      "grad_norm": 0.870552880234674,
+      "learning_rate": 1.489531570280631e-05,
+      "loss": 0.8029,
+      "step": 1857
+    },
+    {
+      "epoch": 0.3574451712196999,
+      "grad_norm": 0.8700247434461938,
+      "learning_rate": 1.488988154489183e-05,
+      "loss": 0.8251,
+      "step": 1858
+    },
+    {
+      "epoch": 0.35763755290496346,
+      "grad_norm": 0.8682686640934995,
+      "learning_rate": 1.4884445488556972e-05,
+      "loss": 0.8556,
+      "step": 1859
+    },
+    {
+      "epoch": 0.357829934590227,
+      "grad_norm": 0.7566999856199353,
+      "learning_rate": 1.4879007535912198e-05,
+      "loss": 0.3498,
+      "step": 1860
+    },
+    {
+      "epoch": 0.3580223162754906,
+      "grad_norm": 0.8204042110993384,
+      "learning_rate": 1.4873567689068708e-05,
+      "loss": 0.8661,
+      "step": 1861
+    },
+    {
+      "epoch": 0.3582146979607541,
+      "grad_norm": 0.8289072164665804,
+      "learning_rate": 1.4868125950138442e-05,
+      "loss": 0.8209,
+      "step": 1862
+    },
+    {
+      "epoch": 0.3584070796460177,
+      "grad_norm": 0.8186402504632571,
+      "learning_rate": 1.4862682321234064e-05,
+      "loss": 0.8829,
+      "step": 1863
+    },
+    {
+      "epoch": 0.3585994613312813,
+      "grad_norm": 0.8521469497870516,
+      "learning_rate": 1.4857236804468983e-05,
+      "loss": 0.817,
+      "step": 1864
+    },
+    {
+      "epoch": 0.3587918430165448,
+      "grad_norm": 0.9115937193946495,
+      "learning_rate": 1.4851789401957338e-05,
+      "loss": 0.8005,
+      "step": 1865
+    },
+    {
+      "epoch": 0.3589842247018084,
+      "grad_norm": 0.888985143861451,
+      "learning_rate": 1.4846340115813993e-05,
+      "loss": 0.8621,
+      "step": 1866
+    },
+    {
+      "epoch": 0.35917660638707194,
+      "grad_norm": 0.9155596835962704,
+      "learning_rate": 1.484088894815455e-05,
+      "loss": 0.9188,
+      "step": 1867
+    },
+    {
+      "epoch": 0.3593689880723355,
+      "grad_norm": 0.6750639219769413,
+      "learning_rate": 1.4835435901095341e-05,
+      "loss": 0.3589,
+      "step": 1868
+    },
+    {
+      "epoch": 0.35956136975759906,
+      "grad_norm": 0.7808531484594875,
+      "learning_rate": 1.4829980976753426e-05,
+      "loss": 0.8433,
+      "step": 1869
+    },
+    {
+      "epoch": 0.35975375144286265,
+      "grad_norm": 0.8672835487406537,
+      "learning_rate": 1.4824524177246597e-05,
+      "loss": 0.7794,
+      "step": 1870
+    },
+    {
+      "epoch": 0.3599461331281262,
+      "grad_norm": 0.9205870530052864,
+      "learning_rate": 1.4819065504693365e-05,
+      "loss": 0.9028,
+      "step": 1871
+    },
+    {
+      "epoch": 0.36013851481338977,
+      "grad_norm": 0.8784860563989858,
+      "learning_rate": 1.4813604961212984e-05,
+      "loss": 0.8956,
+      "step": 1872
+    },
+    {
+      "epoch": 0.36033089649865335,
+      "grad_norm": 0.903782363946679,
+      "learning_rate": 1.4808142548925417e-05,
+      "loss": 0.8449,
+      "step": 1873
+    },
+    {
+      "epoch": 0.3605232781839169,
+      "grad_norm": 0.9522485345511679,
+      "learning_rate": 1.4802678269951365e-05,
+      "loss": 0.8465,
+      "step": 1874
+    },
+    {
+      "epoch": 0.36071565986918047,
+      "grad_norm": 0.9162100237546593,
+      "learning_rate": 1.4797212126412243e-05,
+      "loss": 0.7909,
+      "step": 1875
+    },
+    {
+      "epoch": 0.360908041554444,
+      "grad_norm": 0.8749137598391301,
+      "learning_rate": 1.4791744120430202e-05,
+      "loss": 0.8789,
+      "step": 1876
+    },
+    {
+      "epoch": 0.3611004232397076,
+      "grad_norm": 0.8809329167436161,
+      "learning_rate": 1.4786274254128112e-05,
+      "loss": 0.8466,
+      "step": 1877
+    },
+    {
+      "epoch": 0.3612928049249711,
+      "grad_norm": 0.9311955859389907,
+      "learning_rate": 1.4780802529629559e-05,
+      "loss": 0.8498,
+      "step": 1878
+    },
+    {
+      "epoch": 0.3614851866102347,
+      "grad_norm": 0.9403101706164905,
+      "learning_rate": 1.4775328949058856e-05,
+      "loss": 0.919,
+      "step": 1879
+    },
+    {
+      "epoch": 0.36167756829549824,
+      "grad_norm": 0.8852256047087369,
+      "learning_rate": 1.4769853514541037e-05,
+      "loss": 0.7865,
+      "step": 1880
+    },
+    {
+      "epoch": 0.36186994998076183,
+      "grad_norm": 0.8912712321455594,
+      "learning_rate": 1.4764376228201848e-05,
+      "loss": 0.8057,
+      "step": 1881
+    },
+    {
+      "epoch": 0.3620623316660254,
+      "grad_norm": 0.8373980989587336,
+      "learning_rate": 1.475889709216777e-05,
+      "loss": 0.8885,
+      "step": 1882
+    },
+    {
+      "epoch": 0.36225471335128895,
+      "grad_norm": 0.8933390080484326,
+      "learning_rate": 1.4753416108565985e-05,
+      "loss": 0.8429,
+      "step": 1883
+    },
+    {
+      "epoch": 0.36244709503655254,
+      "grad_norm": 0.9291355591282663,
+      "learning_rate": 1.47479332795244e-05,
+      "loss": 0.8848,
+      "step": 1884
+    },
+    {
+      "epoch": 0.36263947672181607,
+      "grad_norm": 0.8529214328721796,
+      "learning_rate": 1.4742448607171644e-05,
+      "loss": 0.3805,
+      "step": 1885
+    },
+    {
+      "epoch": 0.36283185840707965,
+      "grad_norm": 0.878957238325444,
+      "learning_rate": 1.473696209363705e-05,
+      "loss": 0.829,
+      "step": 1886
+    },
+    {
+      "epoch": 0.3630242400923432,
+      "grad_norm": 0.9267820119700405,
+      "learning_rate": 1.4731473741050673e-05,
+      "loss": 0.8217,
+      "step": 1887
+    },
+    {
+      "epoch": 0.3632166217776068,
+      "grad_norm": 0.8980772145505213,
+      "learning_rate": 1.4725983551543279e-05,
+      "loss": 0.8284,
+      "step": 1888
+    },
+    {
+      "epoch": 0.36340900346287036,
+      "grad_norm": 0.8772100396678629,
+      "learning_rate": 1.472049152724635e-05,
+      "loss": 0.8954,
+      "step": 1889
+    },
+    {
+      "epoch": 0.3636013851481339,
+      "grad_norm": 0.9163295303167824,
+      "learning_rate": 1.471499767029208e-05,
+      "loss": 0.8723,
+      "step": 1890
+    },
+    {
+      "epoch": 0.3637937668333975,
+      "grad_norm": 0.9377771949535597,
+      "learning_rate": 1.470950198281337e-05,
+      "loss": 0.8119,
+      "step": 1891
+    },
+    {
+      "epoch": 0.363986148518661,
+      "grad_norm": 0.9703215892749681,
+      "learning_rate": 1.470400446694384e-05,
+      "loss": 0.8311,
+      "step": 1892
+    },
+    {
+      "epoch": 0.3641785302039246,
+      "grad_norm": 0.8965181639490445,
+      "learning_rate": 1.4698505124817811e-05,
+      "loss": 0.8199,
+      "step": 1893
+    },
+    {
+      "epoch": 0.36437091188918813,
+      "grad_norm": 0.6711432576973532,
+      "learning_rate": 1.4693003958570318e-05,
+      "loss": 0.3352,
+      "step": 1894
+    },
+    {
+      "epoch": 0.3645632935744517,
+      "grad_norm": 0.9172927528533508,
+      "learning_rate": 1.4687500970337103e-05,
+      "loss": 0.8749,
+      "step": 1895
+    },
+    {
+      "epoch": 0.36475567525971525,
+      "grad_norm": 0.8768205291191323,
+      "learning_rate": 1.4681996162254618e-05,
+      "loss": 0.9067,
+      "step": 1896
+    },
+    {
+      "epoch": 0.36494805694497884,
+      "grad_norm": 0.9105368630279895,
+      "learning_rate": 1.4676489536460015e-05,
+      "loss": 0.9111,
+      "step": 1897
+    },
+    {
+      "epoch": 0.3651404386302424,
+      "grad_norm": 0.8380492978395624,
+      "learning_rate": 1.467098109509116e-05,
+      "loss": 0.8246,
+      "step": 1898
+    },
+    {
+      "epoch": 0.36533282031550596,
+      "grad_norm": 0.9392837728795966,
+      "learning_rate": 1.4665470840286614e-05,
+      "loss": 0.87,
+      "step": 1899
+    },
+    {
+      "epoch": 0.36552520200076954,
+      "grad_norm": 0.8062415816921075,
+      "learning_rate": 1.4659958774185654e-05,
+      "loss": 0.8589,
+      "step": 1900
+    },
+    {
+      "epoch": 0.3657175836860331,
+      "grad_norm": 0.8327130569234407,
+      "learning_rate": 1.4654444898928249e-05,
+      "loss": 0.8868,
+      "step": 1901
+    },
+    {
+      "epoch": 0.36590996537129666,
+      "grad_norm": 0.8502150966403594,
+      "learning_rate": 1.4648929216655077e-05,
+      "loss": 0.878,
+      "step": 1902
+    },
+    {
+      "epoch": 0.3661023470565602,
+      "grad_norm": 0.8541158655886063,
+      "learning_rate": 1.4643411729507517e-05,
+      "loss": 0.8479,
+      "step": 1903
+    },
+    {
+      "epoch": 0.3662947287418238,
+      "grad_norm": 0.9271691881005671,
+      "learning_rate": 1.4637892439627644e-05,
+      "loss": 0.9083,
+      "step": 1904
+    },
+    {
+      "epoch": 0.3664871104270873,
+      "grad_norm": 0.8964898187765169,
+      "learning_rate": 1.4632371349158241e-05,
+      "loss": 0.8769,
+      "step": 1905
+    },
+    {
+      "epoch": 0.3666794921123509,
+      "grad_norm": 0.8298717747980526,
+      "learning_rate": 1.4626848460242782e-05,
+      "loss": 0.825,
+      "step": 1906
+    },
+    {
+      "epoch": 0.3668718737976145,
+      "grad_norm": 0.9093478029865297,
+      "learning_rate": 1.4621323775025444e-05,
+      "loss": 0.842,
+      "step": 1907
+    },
+    {
+      "epoch": 0.367064255482878,
+      "grad_norm": 0.914303349567162,
+      "learning_rate": 1.4615797295651099e-05,
+      "loss": 0.8904,
+      "step": 1908
+    },
+    {
+      "epoch": 0.3672566371681416,
+      "grad_norm": 0.8693767432630188,
+      "learning_rate": 1.4610269024265317e-05,
+      "loss": 0.8236,
+      "step": 1909
+    },
+    {
+      "epoch": 0.36744901885340514,
+      "grad_norm": 0.6893164249031472,
+      "learning_rate": 1.4604738963014365e-05,
+      "loss": 0.387,
+      "step": 1910
+    },
+    {
+      "epoch": 0.36764140053866873,
+      "grad_norm": 0.8449430104283409,
+      "learning_rate": 1.4599207114045202e-05,
+      "loss": 0.8666,
+      "step": 1911
+    },
+    {
+      "epoch": 0.36783378222393226,
+      "grad_norm": 0.7879634011444212,
+      "learning_rate": 1.4593673479505482e-05,
+      "loss": 0.8146,
+      "step": 1912
+    },
+    {
+      "epoch": 0.36802616390919585,
+      "grad_norm": 0.8946566706907879,
+      "learning_rate": 1.4588138061543551e-05,
+      "loss": 0.9095,
+      "step": 1913
+    },
+    {
+      "epoch": 0.36821854559445943,
+      "grad_norm": 0.5768283821850404,
+      "learning_rate": 1.458260086230845e-05,
+      "loss": 0.3365,
+      "step": 1914
+    },
+    {
+      "epoch": 0.36841092727972297,
+      "grad_norm": 0.8711497072855326,
+      "learning_rate": 1.4577061883949912e-05,
+      "loss": 0.8122,
+      "step": 1915
+    },
+    {
+      "epoch": 0.36860330896498655,
+      "grad_norm": 0.891745410061075,
+      "learning_rate": 1.4571521128618358e-05,
+      "loss": 0.8559,
+      "step": 1916
+    },
+    {
+      "epoch": 0.3687956906502501,
+      "grad_norm": 0.8972371436962714,
+      "learning_rate": 1.4565978598464895e-05,
+      "loss": 0.8374,
+      "step": 1917
+    },
+    {
+      "epoch": 0.3689880723355137,
+      "grad_norm": 0.8775778597994708,
+      "learning_rate": 1.4560434295641338e-05,
+      "loss": 0.7813,
+      "step": 1918
+    },
+    {
+      "epoch": 0.3691804540207772,
+      "grad_norm": 0.8106778374934542,
+      "learning_rate": 1.455488822230016e-05,
+      "loss": 0.8099,
+      "step": 1919
+    },
+    {
+      "epoch": 0.3693728357060408,
+      "grad_norm": 0.8399026580771787,
+      "learning_rate": 1.4549340380594545e-05,
+      "loss": 0.8346,
+      "step": 1920
+    },
+    {
+      "epoch": 0.3695652173913043,
+      "grad_norm": 0.866350482265557,
+      "learning_rate": 1.454379077267836e-05,
+      "loss": 0.8339,
+      "step": 1921
+    },
+    {
+      "epoch": 0.3697575990765679,
+      "grad_norm": 0.9358327163318275,
+      "learning_rate": 1.4538239400706147e-05,
+      "loss": 0.8556,
+      "step": 1922
+    },
+    {
+      "epoch": 0.3699499807618315,
+      "grad_norm": 0.7122322874184328,
+      "learning_rate": 1.4532686266833143e-05,
+      "loss": 0.3532,
+      "step": 1923
+    },
+    {
+      "epoch": 0.37014236244709503,
+      "grad_norm": 0.9541786252543494,
+      "learning_rate": 1.4527131373215265e-05,
+      "loss": 0.8473,
+      "step": 1924
+    },
+    {
+      "epoch": 0.3703347441323586,
+      "grad_norm": 0.602449188975326,
+      "learning_rate": 1.4521574722009115e-05,
+      "loss": 0.3279,
+      "step": 1925
+    },
+    {
+      "epoch": 0.37052712581762215,
+      "grad_norm": 0.9069437684198839,
+      "learning_rate": 1.4516016315371974e-05,
+      "loss": 0.8836,
+      "step": 1926
+    },
+    {
+      "epoch": 0.37071950750288574,
+      "grad_norm": 0.8722347932871511,
+      "learning_rate": 1.4510456155461807e-05,
+      "loss": 0.8284,
+      "step": 1927
+    },
+    {
+      "epoch": 0.37091188918814927,
+      "grad_norm": 0.9490029684722797,
+      "learning_rate": 1.4504894244437264e-05,
+      "loss": 0.8716,
+      "step": 1928
+    },
+    {
+      "epoch": 0.37110427087341286,
+      "grad_norm": 0.8876906899695144,
+      "learning_rate": 1.4499330584457667e-05,
+      "loss": 0.8471,
+      "step": 1929
+    },
+    {
+      "epoch": 0.3712966525586764,
+      "grad_norm": 0.8503691944468037,
+      "learning_rate": 1.4493765177683017e-05,
+      "loss": 0.8618,
+      "step": 1930
+    },
+    {
+      "epoch": 0.37148903424394,
+      "grad_norm": 0.8776122126946996,
+      "learning_rate": 1.4488198026274007e-05,
+      "loss": 0.8669,
+      "step": 1931
+    },
+    {
+      "epoch": 0.37168141592920356,
+      "grad_norm": 0.9009073115445365,
+      "learning_rate": 1.4482629132391985e-05,
+      "loss": 0.8544,
+      "step": 1932
+    },
+    {
+      "epoch": 0.3718737976144671,
+      "grad_norm": 1.0289032226332604,
+      "learning_rate": 1.4477058498198993e-05,
+      "loss": 0.3628,
+      "step": 1933
+    },
+    {
+      "epoch": 0.3720661792997307,
+      "grad_norm": 0.8553324712337323,
+      "learning_rate": 1.4471486125857743e-05,
+      "loss": 0.8003,
+      "step": 1934
+    },
+    {
+      "epoch": 0.3722585609849942,
+      "grad_norm": 0.8557628868815889,
+      "learning_rate": 1.446591201753162e-05,
+      "loss": 0.8706,
+      "step": 1935
+    },
+    {
+      "epoch": 0.3724509426702578,
+      "grad_norm": 0.9226810919668857,
+      "learning_rate": 1.4460336175384688e-05,
+      "loss": 0.8511,
+      "step": 1936
+    },
+    {
+      "epoch": 0.37264332435552133,
+      "grad_norm": 0.8590411344246365,
+      "learning_rate": 1.4454758601581675e-05,
+      "loss": 0.8749,
+      "step": 1937
+    },
+    {
+      "epoch": 0.3728357060407849,
+      "grad_norm": 0.8535156558840834,
+      "learning_rate": 1.4449179298287999e-05,
+      "loss": 0.7925,
+      "step": 1938
+    },
+    {
+      "epoch": 0.3730280877260485,
+      "grad_norm": 0.9108345040044094,
+      "learning_rate": 1.4443598267669723e-05,
+      "loss": 0.7891,
+      "step": 1939
+    },
+    {
+      "epoch": 0.37322046941131204,
+      "grad_norm": 0.9251288566123396,
+      "learning_rate": 1.4438015511893602e-05,
+      "loss": 0.9057,
+      "step": 1940
+    },
+    {
+      "epoch": 0.3734128510965756,
+      "grad_norm": 0.9052038595330075,
+      "learning_rate": 1.4432431033127056e-05,
+      "loss": 0.8455,
+      "step": 1941
+    },
+    {
+      "epoch": 0.37360523278183916,
+      "grad_norm": 0.7872435572612206,
+      "learning_rate": 1.442684483353817e-05,
+      "loss": 0.7534,
+      "step": 1942
+    },
+    {
+      "epoch": 0.37379761446710275,
+      "grad_norm": 0.84545315778054,
+      "learning_rate": 1.4421256915295697e-05,
+      "loss": 0.786,
+      "step": 1943
+    },
+    {
+      "epoch": 0.3739899961523663,
+      "grad_norm": 0.8451720129338766,
+      "learning_rate": 1.4415667280569064e-05,
+      "loss": 0.871,
+      "step": 1944
+    },
+    {
+      "epoch": 0.37418237783762986,
+      "grad_norm": 0.8922649652077709,
+      "learning_rate": 1.4410075931528356e-05,
+      "loss": 0.8312,
+      "step": 1945
+    },
+    {
+      "epoch": 0.3743747595228934,
+      "grad_norm": 0.8801524726829687,
+      "learning_rate": 1.4404482870344322e-05,
+      "loss": 0.8936,
+      "step": 1946
+    },
+    {
+      "epoch": 0.374567141208157,
+      "grad_norm": 0.8684067108035717,
+      "learning_rate": 1.4398888099188396e-05,
+      "loss": 0.8311,
+      "step": 1947
+    },
+    {
+      "epoch": 0.37475952289342057,
+      "grad_norm": 0.8912977953891604,
+      "learning_rate": 1.4393291620232646e-05,
+      "loss": 0.8589,
+      "step": 1948
+    },
+    {
+      "epoch": 0.3749519045786841,
+      "grad_norm": 0.9003636451775235,
+      "learning_rate": 1.4387693435649826e-05,
+      "loss": 0.8772,
+      "step": 1949
+    },
+    {
+      "epoch": 0.3751442862639477,
+      "grad_norm": 0.9678372284743665,
+      "learning_rate": 1.4382093547613338e-05,
+      "loss": 0.787,
+      "step": 1950
+    },
+    {
+      "epoch": 0.3753366679492112,
+      "grad_norm": 0.8987076057057346,
+      "learning_rate": 1.4376491958297263e-05,
+      "loss": 0.896,
+      "step": 1951
+    },
+    {
+      "epoch": 0.3755290496344748,
+      "grad_norm": 0.881387255355625,
+      "learning_rate": 1.4370888669876317e-05,
+      "loss": 0.8587,
+      "step": 1952
+    },
+    {
+      "epoch": 0.37572143131973834,
+      "grad_norm": 0.8692587662187784,
+      "learning_rate": 1.4365283684525895e-05,
+      "loss": 0.7932,
+      "step": 1953
+    },
+    {
+      "epoch": 0.37591381300500193,
+      "grad_norm": 0.928113969501491,
+      "learning_rate": 1.4359677004422045e-05,
+      "loss": 0.8637,
+      "step": 1954
+    },
+    {
+      "epoch": 0.37610619469026546,
+      "grad_norm": 0.9109132812077769,
+      "learning_rate": 1.4354068631741476e-05,
+      "loss": 0.9242,
+      "step": 1955
+    },
+    {
+      "epoch": 0.37629857637552905,
+      "grad_norm": 0.8785525662512085,
+      "learning_rate": 1.4348458568661548e-05,
+      "loss": 0.8745,
+      "step": 1956
+    },
+    {
+      "epoch": 0.37649095806079264,
+      "grad_norm": 0.8664903813755966,
+      "learning_rate": 1.434284681736028e-05,
+      "loss": 0.8727,
+      "step": 1957
+    },
+    {
+      "epoch": 0.37668333974605617,
+      "grad_norm": 0.8534621634160187,
+      "learning_rate": 1.4337233380016354e-05,
+      "loss": 0.843,
+      "step": 1958
+    },
+    {
+      "epoch": 0.37687572143131975,
+      "grad_norm": 0.9623019194851549,
+      "learning_rate": 1.433161825880909e-05,
+      "loss": 0.8449,
+      "step": 1959
+    },
+    {
+      "epoch": 0.3770681031165833,
+      "grad_norm": 0.7753878644538447,
+      "learning_rate": 1.432600145591848e-05,
+      "loss": 0.3678,
+      "step": 1960
+    },
+    {
+      "epoch": 0.3772604848018469,
+      "grad_norm": 0.8593811857114981,
+      "learning_rate": 1.4320382973525151e-05,
+      "loss": 0.8484,
+      "step": 1961
+    },
+    {
+      "epoch": 0.3774528664871104,
+      "grad_norm": 0.9179689168548779,
+      "learning_rate": 1.43147628138104e-05,
+      "loss": 0.891,
+      "step": 1962
+    },
+    {
+      "epoch": 0.377645248172374,
+      "grad_norm": 0.8661430974380462,
+      "learning_rate": 1.4309140978956161e-05,
+      "loss": 0.8646,
+      "step": 1963
+    },
+    {
+      "epoch": 0.3778376298576376,
+      "grad_norm": 0.8608771582535418,
+      "learning_rate": 1.430351747114503e-05,
+      "loss": 0.7821,
+      "step": 1964
+    },
+    {
+      "epoch": 0.3780300115429011,
+      "grad_norm": 0.6240839984267407,
+      "learning_rate": 1.429789229256024e-05,
+      "loss": 0.2782,
+      "step": 1965
+    },
+    {
+      "epoch": 0.3782223932281647,
+      "grad_norm": 0.6319658378755023,
+      "learning_rate": 1.429226544538568e-05,
+      "loss": 0.3028,
+      "step": 1966
+    },
+    {
+      "epoch": 0.37841477491342823,
+      "grad_norm": 0.8452875984824434,
+      "learning_rate": 1.4286636931805887e-05,
+      "loss": 0.8389,
+      "step": 1967
+    },
+    {
+      "epoch": 0.3786071565986918,
+      "grad_norm": 0.841703101259149,
+      "learning_rate": 1.4281006754006045e-05,
+      "loss": 0.8643,
+      "step": 1968
+    },
+    {
+      "epoch": 0.37879953828395535,
+      "grad_norm": 0.8420877306309484,
+      "learning_rate": 1.427537491417198e-05,
+      "loss": 0.7727,
+      "step": 1969
+    },
+    {
+      "epoch": 0.37899191996921894,
+      "grad_norm": 0.9135751757399391,
+      "learning_rate": 1.426974141449017e-05,
+      "loss": 0.8585,
+      "step": 1970
+    },
+    {
+      "epoch": 0.37918430165448247,
+      "grad_norm": 0.9360631574821197,
+      "learning_rate": 1.4264106257147732e-05,
+      "loss": 0.8486,
+      "step": 1971
+    },
+    {
+      "epoch": 0.37937668333974606,
+      "grad_norm": 0.856223958968654,
+      "learning_rate": 1.4258469444332423e-05,
+      "loss": 0.9035,
+      "step": 1972
+    },
+    {
+      "epoch": 0.37956906502500964,
+      "grad_norm": 0.8528901885137723,
+      "learning_rate": 1.4252830978232658e-05,
+      "loss": 0.8277,
+      "step": 1973
+    },
+    {
+      "epoch": 0.3797614467102732,
+      "grad_norm": 0.9268172490678079,
+      "learning_rate": 1.4247190861037474e-05,
+      "loss": 0.8762,
+      "step": 1974
+    },
+    {
+      "epoch": 0.37995382839553676,
+      "grad_norm": 0.8242922014818741,
+      "learning_rate": 1.4241549094936567e-05,
+      "loss": 0.8239,
+      "step": 1975
+    },
+    {
+      "epoch": 0.3801462100808003,
+      "grad_norm": 0.8830147201070933,
+      "learning_rate": 1.4235905682120255e-05,
+      "loss": 0.8276,
+      "step": 1976
+    },
+    {
+      "epoch": 0.3803385917660639,
+      "grad_norm": 0.8704148141158102,
+      "learning_rate": 1.4230260624779512e-05,
+      "loss": 0.7979,
+      "step": 1977
+    },
+    {
+      "epoch": 0.3805309734513274,
+      "grad_norm": 0.8371888103804064,
+      "learning_rate": 1.4224613925105947e-05,
+      "loss": 0.8375,
+      "step": 1978
+    },
+    {
+      "epoch": 0.380723355136591,
+      "grad_norm": 0.8698203839904782,
+      "learning_rate": 1.4218965585291792e-05,
+      "loss": 0.8354,
+      "step": 1979
+    },
+    {
+      "epoch": 0.38091573682185453,
+      "grad_norm": 0.9121304228468888,
+      "learning_rate": 1.4213315607529939e-05,
+      "loss": 0.8064,
+      "step": 1980
+    },
+    {
+      "epoch": 0.3811081185071181,
+      "grad_norm": 0.8775981333260872,
+      "learning_rate": 1.4207663994013896e-05,
+      "loss": 0.8643,
+      "step": 1981
+    },
+    {
+      "epoch": 0.3813005001923817,
+      "grad_norm": 0.9005900471838773,
+      "learning_rate": 1.4202010746937815e-05,
+      "loss": 0.8765,
+      "step": 1982
+    },
+    {
+      "epoch": 0.38149288187764524,
+      "grad_norm": 0.8764223259247378,
+      "learning_rate": 1.4196355868496485e-05,
+      "loss": 0.8617,
+      "step": 1983
+    },
+    {
+      "epoch": 0.3816852635629088,
+      "grad_norm": 0.876077856102243,
+      "learning_rate": 1.4190699360885323e-05,
+      "loss": 0.782,
+      "step": 1984
+    },
+    {
+      "epoch": 0.38187764524817236,
+      "grad_norm": 0.8194640975670845,
+      "learning_rate": 1.4185041226300376e-05,
+      "loss": 0.809,
+      "step": 1985
+    },
+    {
+      "epoch": 0.38207002693343595,
+      "grad_norm": 0.928181281643078,
+      "learning_rate": 1.4179381466938332e-05,
+      "loss": 0.8049,
+      "step": 1986
+    },
+    {
+      "epoch": 0.3822624086186995,
+      "grad_norm": 0.9413175880122328,
+      "learning_rate": 1.4173720084996501e-05,
+      "loss": 0.9138,
+      "step": 1987
+    },
+    {
+      "epoch": 0.38245479030396307,
+      "grad_norm": 0.8548461891869018,
+      "learning_rate": 1.4168057082672828e-05,
+      "loss": 0.8633,
+      "step": 1988
+    },
+    {
+      "epoch": 0.38264717198922665,
+      "grad_norm": 0.8626031334437284,
+      "learning_rate": 1.4162392462165884e-05,
+      "loss": 0.8084,
+      "step": 1989
+    },
+    {
+      "epoch": 0.3828395536744902,
+      "grad_norm": 0.8765529711885668,
+      "learning_rate": 1.4156726225674874e-05,
+      "loss": 0.8473,
+      "step": 1990
+    },
+    {
+      "epoch": 0.3830319353597538,
+      "grad_norm": 0.8811087902554386,
+      "learning_rate": 1.415105837539962e-05,
+      "loss": 0.8426,
+      "step": 1991
+    },
+    {
+      "epoch": 0.3832243170450173,
+      "grad_norm": 0.8289160407206098,
+      "learning_rate": 1.414538891354058e-05,
+      "loss": 0.8489,
+      "step": 1992
+    },
+    {
+      "epoch": 0.3834166987302809,
+      "grad_norm": 0.8491674723295126,
+      "learning_rate": 1.4139717842298835e-05,
+      "loss": 0.872,
+      "step": 1993
+    },
+    {
+      "epoch": 0.3836090804155444,
+      "grad_norm": 0.8866256147331425,
+      "learning_rate": 1.4134045163876086e-05,
+      "loss": 0.8261,
+      "step": 1994
+    },
+    {
+      "epoch": 0.383801462100808,
+      "grad_norm": 0.8208503322995722,
+      "learning_rate": 1.4128370880474667e-05,
+      "loss": 0.8373,
+      "step": 1995
+    },
+    {
+      "epoch": 0.38399384378607154,
+      "grad_norm": 0.9156512169778923,
+      "learning_rate": 1.412269499429753e-05,
+      "loss": 0.8632,
+      "step": 1996
+    },
+    {
+      "epoch": 0.38418622547133513,
+      "grad_norm": 0.9647889548942796,
+      "learning_rate": 1.4117017507548244e-05,
+      "loss": 0.8624,
+      "step": 1997
+    },
+    {
+      "epoch": 0.3843786071565987,
+      "grad_norm": 0.9519676447514565,
+      "learning_rate": 1.4111338422431013e-05,
+      "loss": 0.774,
+      "step": 1998
+    },
+    {
+      "epoch": 0.38457098884186225,
+      "grad_norm": 0.8776626146563772,
+      "learning_rate": 1.4105657741150648e-05,
+      "loss": 0.872,
+      "step": 1999
+    },
+    {
+      "epoch": 0.38476337052712584,
+      "grad_norm": 0.8204695552616907,
+      "learning_rate": 1.4099975465912584e-05,
+      "loss": 0.8401,
+      "step": 2000
+    },
+    {
+      "epoch": 0.38495575221238937,
+      "grad_norm": 0.8969465589619798,
+      "learning_rate": 1.4094291598922877e-05,
+      "loss": 0.832,
+      "step": 2001
+    },
+    {
+      "epoch": 0.38514813389765296,
+      "grad_norm": 0.8697422076825257,
+      "learning_rate": 1.40886061423882e-05,
+      "loss": 0.8284,
+      "step": 2002
+    },
+    {
+      "epoch": 0.3853405155829165,
+      "grad_norm": 0.8867825859044887,
+      "learning_rate": 1.4082919098515846e-05,
+      "loss": 0.7853,
+      "step": 2003
+    },
+    {
+      "epoch": 0.3855328972681801,
+      "grad_norm": 0.8730529316582181,
+      "learning_rate": 1.407723046951372e-05,
+      "loss": 0.8324,
+      "step": 2004
+    },
+    {
+      "epoch": 0.3857252789534436,
+      "grad_norm": 0.8796849160123348,
+      "learning_rate": 1.4071540257590341e-05,
+      "loss": 0.8801,
+      "step": 2005
+    },
+    {
+      "epoch": 0.3859176606387072,
+      "grad_norm": 0.8391232337983586,
+      "learning_rate": 1.4065848464954848e-05,
+      "loss": 0.7991,
+      "step": 2006
+    },
+    {
+      "epoch": 0.3861100423239708,
+      "grad_norm": 1.1629769847795888,
+      "learning_rate": 1.4060155093816988e-05,
+      "loss": 0.3431,
+      "step": 2007
+    },
+    {
+      "epoch": 0.3863024240092343,
+      "grad_norm": 0.8693977279607927,
+      "learning_rate": 1.4054460146387124e-05,
+      "loss": 0.7797,
+      "step": 2008
+    },
+    {
+      "epoch": 0.3864948056944979,
+      "grad_norm": 0.9441430368136199,
+      "learning_rate": 1.4048763624876233e-05,
+      "loss": 0.8068,
+      "step": 2009
+    },
+    {
+      "epoch": 0.38668718737976143,
+      "grad_norm": 0.9340363491423056,
+      "learning_rate": 1.4043065531495904e-05,
+      "loss": 0.8056,
+      "step": 2010
+    },
+    {
+      "epoch": 0.386879569065025,
+      "grad_norm": 0.7721618839191898,
+      "learning_rate": 1.4037365868458325e-05,
+      "loss": 0.8157,
+      "step": 2011
+    },
+    {
+      "epoch": 0.38707195075028855,
+      "grad_norm": 0.63726826514089,
+      "learning_rate": 1.4031664637976305e-05,
+      "loss": 0.3584,
+      "step": 2012
+    },
+    {
+      "epoch": 0.38726433243555214,
+      "grad_norm": 0.859936728943249,
+      "learning_rate": 1.402596184226326e-05,
+      "loss": 0.8492,
+      "step": 2013
+    },
+    {
+      "epoch": 0.38745671412081567,
+      "grad_norm": 0.9190191614530161,
+      "learning_rate": 1.4020257483533208e-05,
+      "loss": 0.8401,
+      "step": 2014
+    },
+    {
+      "epoch": 0.38764909580607926,
+      "grad_norm": 0.8192937348285032,
+      "learning_rate": 1.401455156400078e-05,
+      "loss": 0.8502,
+      "step": 2015
+    },
+    {
+      "epoch": 0.38784147749134285,
+      "grad_norm": 0.8714721188008335,
+      "learning_rate": 1.400884408588121e-05,
+      "loss": 0.8719,
+      "step": 2016
+    },
+    {
+      "epoch": 0.3880338591766064,
+      "grad_norm": 0.7935440059147492,
+      "learning_rate": 1.400313505139034e-05,
+      "loss": 0.8479,
+      "step": 2017
+    },
+    {
+      "epoch": 0.38822624086186996,
+      "grad_norm": 0.8833316790034615,
+      "learning_rate": 1.3997424462744607e-05,
+      "loss": 0.8613,
+      "step": 2018
+    },
+    {
+      "epoch": 0.3884186225471335,
+      "grad_norm": 0.9308584589763687,
+      "learning_rate": 1.3991712322161065e-05,
+      "loss": 0.848,
+      "step": 2019
+    },
+    {
+      "epoch": 0.3886110042323971,
+      "grad_norm": 0.8894291665728187,
+      "learning_rate": 1.3985998631857359e-05,
+      "loss": 0.8617,
+      "step": 2020
+    },
+    {
+      "epoch": 0.3888033859176606,
+      "grad_norm": 0.7618895888170787,
+      "learning_rate": 1.398028339405174e-05,
+      "loss": 0.3645,
+      "step": 2021
+    },
+    {
+      "epoch": 0.3889957676029242,
+      "grad_norm": 1.0241267423703553,
+      "learning_rate": 1.3974566610963068e-05,
+      "loss": 0.8272,
+      "step": 2022
+    },
+    {
+      "epoch": 0.3891881492881878,
+      "grad_norm": 0.9583158848450247,
+      "learning_rate": 1.3968848284810785e-05,
+      "loss": 0.8153,
+      "step": 2023
+    },
+    {
+      "epoch": 0.3893805309734513,
+      "grad_norm": 0.9693989169287925,
+      "learning_rate": 1.3963128417814951e-05,
+      "loss": 0.9124,
+      "step": 2024
+    },
+    {
+      "epoch": 0.3895729126587149,
+      "grad_norm": 0.9117355521427527,
+      "learning_rate": 1.3957407012196204e-05,
+      "loss": 0.8866,
+      "step": 2025
+    },
+    {
+      "epoch": 0.38976529434397844,
+      "grad_norm": 0.9518072393936184,
+      "learning_rate": 1.3951684070175802e-05,
+      "loss": 0.8752,
+      "step": 2026
+    },
+    {
+      "epoch": 0.38995767602924203,
+      "grad_norm": 0.8320825089813021,
+      "learning_rate": 1.3945959593975582e-05,
+      "loss": 0.8234,
+      "step": 2027
+    },
+    {
+      "epoch": 0.39015005771450556,
+      "grad_norm": 0.8634524971398813,
+      "learning_rate": 1.3940233585817984e-05,
+      "loss": 0.8035,
+      "step": 2028
+    },
+    {
+      "epoch": 0.39034243939976915,
+      "grad_norm": 0.8864788470407442,
+      "learning_rate": 1.3934506047926042e-05,
+      "loss": 0.8566,
+      "step": 2029
+    },
+    {
+      "epoch": 0.3905348210850327,
+      "grad_norm": 0.9171359186289367,
+      "learning_rate": 1.3928776982523384e-05,
+      "loss": 0.8533,
+      "step": 2030
+    },
+    {
+      "epoch": 0.39072720277029627,
+      "grad_norm": 0.8359677043422674,
+      "learning_rate": 1.3923046391834229e-05,
+      "loss": 0.7653,
+      "step": 2031
+    },
+    {
+      "epoch": 0.39091958445555985,
+      "grad_norm": 0.6363199244537943,
+      "learning_rate": 1.3917314278083391e-05,
+      "loss": 0.3657,
+      "step": 2032
+    },
+    {
+      "epoch": 0.3911119661408234,
+      "grad_norm": 0.8403225575161393,
+      "learning_rate": 1.3911580643496272e-05,
+      "loss": 0.7785,
+      "step": 2033
+    },
+    {
+      "epoch": 0.391304347826087,
+      "grad_norm": 0.8878481100484779,
+      "learning_rate": 1.3905845490298867e-05,
+      "loss": 0.8713,
+      "step": 2034
+    },
+    {
+      "epoch": 0.3914967295113505,
+      "grad_norm": 0.642004333482726,
+      "learning_rate": 1.390010882071776e-05,
+      "loss": 0.3634,
+      "step": 2035
+    },
+    {
+      "epoch": 0.3916891111966141,
+      "grad_norm": 0.8559947820833292,
+      "learning_rate": 1.3894370636980128e-05,
+      "loss": 0.8294,
+      "step": 2036
+    },
+    {
+      "epoch": 0.3918814928818776,
+      "grad_norm": 0.9657344778070545,
+      "learning_rate": 1.3888630941313728e-05,
+      "loss": 0.9065,
+      "step": 2037
+    },
+    {
+      "epoch": 0.3920738745671412,
+      "grad_norm": 0.6637408249877516,
+      "learning_rate": 1.3882889735946901e-05,
+      "loss": 0.3308,
+      "step": 2038
+    },
+    {
+      "epoch": 0.39226625625240474,
+      "grad_norm": 0.8818284172503242,
+      "learning_rate": 1.3877147023108592e-05,
+      "loss": 0.8923,
+      "step": 2039
+    },
+    {
+      "epoch": 0.39245863793766833,
+      "grad_norm": 0.6340518633170447,
+      "learning_rate": 1.3871402805028314e-05,
+      "loss": 0.3571,
+      "step": 2040
+    },
+    {
+      "epoch": 0.3926510196229319,
+      "grad_norm": 0.9262696270624857,
+      "learning_rate": 1.3865657083936167e-05,
+      "loss": 0.8617,
+      "step": 2041
+    },
+    {
+      "epoch": 0.39284340130819545,
+      "grad_norm": 0.8813287896579024,
+      "learning_rate": 1.3859909862062844e-05,
+      "loss": 0.8831,
+      "step": 2042
+    },
+    {
+      "epoch": 0.39303578299345904,
+      "grad_norm": 0.9682390134662141,
+      "learning_rate": 1.385416114163961e-05,
+      "loss": 0.8015,
+      "step": 2043
+    },
+    {
+      "epoch": 0.39322816467872257,
+      "grad_norm": 0.9144438787111883,
+      "learning_rate": 1.3848410924898321e-05,
+      "loss": 0.8246,
+      "step": 2044
+    },
+    {
+      "epoch": 0.39342054636398616,
+      "grad_norm": 0.8369662978666328,
+      "learning_rate": 1.3842659214071406e-05,
+      "loss": 0.8877,
+      "step": 2045
+    },
+    {
+      "epoch": 0.3936129280492497,
+      "grad_norm": 0.8432509058187084,
+      "learning_rate": 1.3836906011391878e-05,
+      "loss": 0.8539,
+      "step": 2046
+    },
+    {
+      "epoch": 0.3938053097345133,
+      "grad_norm": 0.8499079510151644,
+      "learning_rate": 1.3831151319093323e-05,
+      "loss": 0.7929,
+      "step": 2047
+    },
+    {
+      "epoch": 0.39399769141977686,
+      "grad_norm": 0.9608755519246626,
+      "learning_rate": 1.382539513940992e-05,
+      "loss": 0.8772,
+      "step": 2048
+    },
+    {
+      "epoch": 0.3941900731050404,
+      "grad_norm": 0.8460679391602755,
+      "learning_rate": 1.3819637474576411e-05,
+      "loss": 0.8118,
+      "step": 2049
+    },
+    {
+      "epoch": 0.394382454790304,
+      "grad_norm": 0.8239895639233569,
+      "learning_rate": 1.381387832682812e-05,
+      "loss": 0.8211,
+      "step": 2050
+    },
+    {
+      "epoch": 0.3945748364755675,
+      "grad_norm": 0.8961953231849655,
+      "learning_rate": 1.380811769840095e-05,
+      "loss": 0.8532,
+      "step": 2051
+    },
+    {
+      "epoch": 0.3947672181608311,
+      "grad_norm": 0.9115617971691796,
+      "learning_rate": 1.3802355591531366e-05,
+      "loss": 0.8078,
+      "step": 2052
+    },
+    {
+      "epoch": 0.39495959984609463,
+      "grad_norm": 0.690802097091316,
+      "learning_rate": 1.3796592008456427e-05,
+      "loss": 0.3552,
+      "step": 2053
+    },
+    {
+      "epoch": 0.3951519815313582,
+      "grad_norm": 0.8969544745609039,
+      "learning_rate": 1.3790826951413747e-05,
+      "loss": 0.8011,
+      "step": 2054
+    },
+    {
+      "epoch": 0.39534436321662175,
+      "grad_norm": 0.8727382882969533,
+      "learning_rate": 1.3785060422641526e-05,
+      "loss": 0.8171,
+      "step": 2055
+    },
+    {
+      "epoch": 0.39553674490188534,
+      "grad_norm": 1.0383098306733627,
+      "learning_rate": 1.3779292424378521e-05,
+      "loss": 0.7815,
+      "step": 2056
+    },
+    {
+      "epoch": 0.3957291265871489,
+      "grad_norm": 0.8707701463583911,
+      "learning_rate": 1.3773522958864076e-05,
+      "loss": 0.8371,
+      "step": 2057
+    },
+    {
+      "epoch": 0.39592150827241246,
+      "grad_norm": 0.9150109396720484,
+      "learning_rate": 1.3767752028338091e-05,
+      "loss": 0.8883,
+      "step": 2058
+    },
+    {
+      "epoch": 0.39611388995767605,
+      "grad_norm": 0.8720440885236931,
+      "learning_rate": 1.376197963504104e-05,
+      "loss": 0.8132,
+      "step": 2059
+    },
+    {
+      "epoch": 0.3963062716429396,
+      "grad_norm": 0.9985885600948802,
+      "learning_rate": 1.3756205781213965e-05,
+      "loss": 0.8083,
+      "step": 2060
+    },
+    {
+      "epoch": 0.39649865332820317,
+      "grad_norm": 0.8977690984334928,
+      "learning_rate": 1.375043046909848e-05,
+      "loss": 0.8382,
+      "step": 2061
+    },
+    {
+      "epoch": 0.3966910350134667,
+      "grad_norm": 0.9207542950776698,
+      "learning_rate": 1.3744653700936752e-05,
+      "loss": 0.8008,
+      "step": 2062
+    },
+    {
+      "epoch": 0.3968834166987303,
+      "grad_norm": 0.877006218582839,
+      "learning_rate": 1.3738875478971526e-05,
+      "loss": 0.9234,
+      "step": 2063
+    },
+    {
+      "epoch": 0.3970757983839938,
+      "grad_norm": 0.9008404186683074,
+      "learning_rate": 1.3733095805446107e-05,
+      "loss": 0.8135,
+      "step": 2064
+    },
+    {
+      "epoch": 0.3972681800692574,
+      "grad_norm": 0.925303293971945,
+      "learning_rate": 1.372731468260436e-05,
+      "loss": 0.8501,
+      "step": 2065
+    },
+    {
+      "epoch": 0.397460561754521,
+      "grad_norm": 0.8477045381917556,
+      "learning_rate": 1.372153211269072e-05,
+      "loss": 0.8373,
+      "step": 2066
+    },
+    {
+      "epoch": 0.3976529434397845,
+      "grad_norm": 0.9120381576575621,
+      "learning_rate": 1.3715748097950176e-05,
+      "loss": 0.8629,
+      "step": 2067
+    },
+    {
+      "epoch": 0.3978453251250481,
+      "grad_norm": 0.9152486135389536,
+      "learning_rate": 1.3709962640628284e-05,
+      "loss": 0.8661,
+      "step": 2068
+    },
+    {
+      "epoch": 0.39803770681031164,
+      "grad_norm": 0.9104074257460003,
+      "learning_rate": 1.3704175742971158e-05,
+      "loss": 0.8184,
+      "step": 2069
+    },
+    {
+      "epoch": 0.39823008849557523,
+      "grad_norm": 0.8190578131661178,
+      "learning_rate": 1.369838740722547e-05,
+      "loss": 0.7793,
+      "step": 2070
+    },
+    {
+      "epoch": 0.39842247018083876,
+      "grad_norm": 0.9214110127817693,
+      "learning_rate": 1.3692597635638452e-05,
+      "loss": 0.7988,
+      "step": 2071
+    },
+    {
+      "epoch": 0.39861485186610235,
+      "grad_norm": 0.8558599199402902,
+      "learning_rate": 1.368680643045789e-05,
+      "loss": 0.8086,
+      "step": 2072
+    },
+    {
+      "epoch": 0.39880723355136594,
+      "grad_norm": 0.9648160011334016,
+      "learning_rate": 1.3681013793932132e-05,
+      "loss": 0.806,
+      "step": 2073
+    },
+    {
+      "epoch": 0.39899961523662947,
+      "grad_norm": 0.8748968602072144,
+      "learning_rate": 1.3675219728310076e-05,
+      "loss": 0.8296,
+      "step": 2074
+    },
+    {
+      "epoch": 0.39919199692189306,
+      "grad_norm": 0.8652082580429211,
+      "learning_rate": 1.3669424235841185e-05,
+      "loss": 0.8435,
+      "step": 2075
+    },
+    {
+      "epoch": 0.3993843786071566,
+      "grad_norm": 0.8784269657894864,
+      "learning_rate": 1.3663627318775459e-05,
+      "loss": 0.8462,
+      "step": 2076
+    },
+    {
+      "epoch": 0.3995767602924202,
+      "grad_norm": 0.8856728930053237,
+      "learning_rate": 1.3657828979363468e-05,
+      "loss": 0.8601,
+      "step": 2077
+    },
+    {
+      "epoch": 0.3997691419776837,
+      "grad_norm": 0.6902820781941593,
+      "learning_rate": 1.3652029219856324e-05,
+      "loss": 0.3825,
+      "step": 2078
+    },
+    {
+      "epoch": 0.3999615236629473,
+      "grad_norm": 0.8643155615511857,
+      "learning_rate": 1.3646228042505694e-05,
+      "loss": 0.804,
+      "step": 2079
+    },
+    {
+      "epoch": 0.4001539053482108,
+      "grad_norm": 0.8877063002161711,
+      "learning_rate": 1.3640425449563793e-05,
+      "loss": 0.8411,
+      "step": 2080
+    },
+    {
+      "epoch": 0.4003462870334744,
+      "grad_norm": 0.9084041872257295,
+      "learning_rate": 1.3634621443283389e-05,
+      "loss": 0.8465,
+      "step": 2081
+    },
+    {
+      "epoch": 0.400538668718738,
+      "grad_norm": 0.9353632352668791,
+      "learning_rate": 1.36288160259178e-05,
+      "loss": 0.8467,
+      "step": 2082
+    },
+    {
+      "epoch": 0.40073105040400153,
+      "grad_norm": 0.8819909865974206,
+      "learning_rate": 1.3623009199720882e-05,
+      "loss": 0.8848,
+      "step": 2083
+    },
+    {
+      "epoch": 0.4009234320892651,
+      "grad_norm": 0.7993723575451664,
+      "learning_rate": 1.3617200966947053e-05,
+      "loss": 0.867,
+      "step": 2084
+    },
+    {
+      "epoch": 0.40111581377452865,
+      "grad_norm": 0.8952372861043845,
+      "learning_rate": 1.3611391329851262e-05,
+      "loss": 0.8413,
+      "step": 2085
+    },
+    {
+      "epoch": 0.40130819545979224,
+      "grad_norm": 0.8851223614755085,
+      "learning_rate": 1.3605580290689013e-05,
+      "loss": 0.8829,
+      "step": 2086
+    },
+    {
+      "epoch": 0.40150057714505577,
+      "grad_norm": 0.8896604905901719,
+      "learning_rate": 1.3599767851716353e-05,
+      "loss": 0.8544,
+      "step": 2087
+    },
+    {
+      "epoch": 0.40169295883031936,
+      "grad_norm": 0.8463095544079874,
+      "learning_rate": 1.3593954015189867e-05,
+      "loss": 0.8463,
+      "step": 2088
+    },
+    {
+      "epoch": 0.4018853405155829,
+      "grad_norm": 0.9904538852634925,
+      "learning_rate": 1.3588138783366692e-05,
+      "loss": 0.816,
+      "step": 2089
+    },
+    {
+      "epoch": 0.4020777222008465,
+      "grad_norm": 0.8694041783928232,
+      "learning_rate": 1.3582322158504495e-05,
+      "loss": 0.8449,
+      "step": 2090
+    },
+    {
+      "epoch": 0.40227010388611006,
+      "grad_norm": 0.922450056007762,
+      "learning_rate": 1.3576504142861496e-05,
+      "loss": 0.8906,
+      "step": 2091
+    },
+    {
+      "epoch": 0.4024624855713736,
+      "grad_norm": 0.844642541705955,
+      "learning_rate": 1.3570684738696444e-05,
+      "loss": 0.7816,
+      "step": 2092
+    },
+    {
+      "epoch": 0.4026548672566372,
+      "grad_norm": 0.926498293065094,
+      "learning_rate": 1.3564863948268631e-05,
+      "loss": 0.9049,
+      "step": 2093
+    },
+    {
+      "epoch": 0.4028472489419007,
+      "grad_norm": 0.8406276301645524,
+      "learning_rate": 1.3559041773837898e-05,
+      "loss": 0.8216,
+      "step": 2094
+    },
+    {
+      "epoch": 0.4030396306271643,
+      "grad_norm": 0.8742754430607708,
+      "learning_rate": 1.3553218217664603e-05,
+      "loss": 0.8575,
+      "step": 2095
+    },
+    {
+      "epoch": 0.40323201231242783,
+      "grad_norm": 0.8446102507889349,
+      "learning_rate": 1.3547393282009656e-05,
+      "loss": 0.7689,
+      "step": 2096
+    },
+    {
+      "epoch": 0.4034243939976914,
+      "grad_norm": 0.8647901379544353,
+      "learning_rate": 1.3541566969134496e-05,
+      "loss": 0.8606,
+      "step": 2097
+    },
+    {
+      "epoch": 0.403616775682955,
+      "grad_norm": 0.8524244894564602,
+      "learning_rate": 1.3535739281301102e-05,
+      "loss": 0.801,
+      "step": 2098
+    },
+    {
+      "epoch": 0.40380915736821854,
+      "grad_norm": 0.8393473714079781,
+      "learning_rate": 1.3529910220771975e-05,
+      "loss": 0.7858,
+      "step": 2099
+    },
+    {
+      "epoch": 0.40400153905348213,
+      "grad_norm": 0.9214568053681031,
+      "learning_rate": 1.3524079789810163e-05,
+      "loss": 0.806,
+      "step": 2100
+    },
+    {
+      "epoch": 0.40419392073874566,
+      "grad_norm": 0.8435760883386615,
+      "learning_rate": 1.3518247990679241e-05,
+      "loss": 0.8633,
+      "step": 2101
+    },
+    {
+      "epoch": 0.40438630242400925,
+      "grad_norm": 0.9910553286353078,
+      "learning_rate": 1.3512414825643312e-05,
+      "loss": 0.8152,
+      "step": 2102
+    },
+    {
+      "epoch": 0.4045786841092728,
+      "grad_norm": 0.8118902443757195,
+      "learning_rate": 1.3506580296967011e-05,
+      "loss": 0.7759,
+      "step": 2103
+    },
+    {
+      "epoch": 0.40477106579453637,
+      "grad_norm": 0.8665955592543542,
+      "learning_rate": 1.3500744406915505e-05,
+      "loss": 0.86,
+      "step": 2104
+    },
+    {
+      "epoch": 0.4049634474797999,
+      "grad_norm": 0.9725921762432256,
+      "learning_rate": 1.3494907157754485e-05,
+      "loss": 0.8162,
+      "step": 2105
+    },
+    {
+      "epoch": 0.4051558291650635,
+      "grad_norm": 0.8296536279275598,
+      "learning_rate": 1.348906855175017e-05,
+      "loss": 0.8383,
+      "step": 2106
+    },
+    {
+      "epoch": 0.4053482108503271,
+      "grad_norm": 0.8014040739377851,
+      "learning_rate": 1.3483228591169315e-05,
+      "loss": 0.7548,
+      "step": 2107
+    },
+    {
+      "epoch": 0.4055405925355906,
+      "grad_norm": 0.8360125980990794,
+      "learning_rate": 1.347738727827919e-05,
+      "loss": 0.755,
+      "step": 2108
+    },
+    {
+      "epoch": 0.4057329742208542,
+      "grad_norm": 0.8608833581473412,
+      "learning_rate": 1.3471544615347591e-05,
+      "loss": 0.8307,
+      "step": 2109
+    },
+    {
+      "epoch": 0.4059253559061177,
+      "grad_norm": 0.9165179075951666,
+      "learning_rate": 1.3465700604642847e-05,
+      "loss": 0.8387,
+      "step": 2110
+    },
+    {
+      "epoch": 0.4061177375913813,
+      "grad_norm": 0.6703785830907473,
+      "learning_rate": 1.34598552484338e-05,
+      "loss": 0.3349,
+      "step": 2111
+    },
+    {
+      "epoch": 0.40631011927664484,
+      "grad_norm": 0.9203809684157206,
+      "learning_rate": 1.3454008548989816e-05,
+      "loss": 0.8054,
+      "step": 2112
+    },
+    {
+      "epoch": 0.40650250096190843,
+      "grad_norm": 0.9079436430440454,
+      "learning_rate": 1.3448160508580789e-05,
+      "loss": 0.8964,
+      "step": 2113
+    },
+    {
+      "epoch": 0.40669488264717196,
+      "grad_norm": 0.6815899276330024,
+      "learning_rate": 1.3442311129477133e-05,
+      "loss": 0.3321,
+      "step": 2114
+    },
+    {
+      "epoch": 0.40688726433243555,
+      "grad_norm": 0.9464282397625672,
+      "learning_rate": 1.343646041394977e-05,
+      "loss": 0.8468,
+      "step": 2115
+    },
+    {
+      "epoch": 0.40707964601769914,
+      "grad_norm": 0.8605682495106847,
+      "learning_rate": 1.3430608364270156e-05,
+      "loss": 0.8764,
+      "step": 2116
+    },
+    {
+      "epoch": 0.40727202770296267,
+      "grad_norm": 0.8202739936855856,
+      "learning_rate": 1.3424754982710256e-05,
+      "loss": 0.8236,
+      "step": 2117
+    },
+    {
+      "epoch": 0.40746440938822626,
+      "grad_norm": 0.8506930169285526,
+      "learning_rate": 1.3418900271542552e-05,
+      "loss": 0.8605,
+      "step": 2118
+    },
+    {
+      "epoch": 0.4076567910734898,
+      "grad_norm": 0.9489747010997296,
+      "learning_rate": 1.3413044233040045e-05,
+      "loss": 0.8489,
+      "step": 2119
+    },
+    {
+      "epoch": 0.4078491727587534,
+      "grad_norm": 0.79975733011484,
+      "learning_rate": 1.3407186869476253e-05,
+      "loss": 0.7983,
+      "step": 2120
+    },
+    {
+      "epoch": 0.4080415544440169,
+      "grad_norm": 0.8062251257527177,
+      "learning_rate": 1.3401328183125208e-05,
+      "loss": 0.8408,
+      "step": 2121
+    },
+    {
+      "epoch": 0.4082339361292805,
+      "grad_norm": 0.9151186032628874,
+      "learning_rate": 1.339546817626145e-05,
+      "loss": 0.8469,
+      "step": 2122
+    },
+    {
+      "epoch": 0.4084263178145441,
+      "grad_norm": 0.8127814792478547,
+      "learning_rate": 1.3389606851160037e-05,
+      "loss": 0.8402,
+      "step": 2123
+    },
+    {
+      "epoch": 0.4086186994998076,
+      "grad_norm": 0.8504476040227046,
+      "learning_rate": 1.3383744210096537e-05,
+      "loss": 0.8498,
+      "step": 2124
+    },
+    {
+      "epoch": 0.4088110811850712,
+      "grad_norm": 0.8380055422200285,
+      "learning_rate": 1.3377880255347026e-05,
+      "loss": 0.866,
+      "step": 2125
+    },
+    {
+      "epoch": 0.40900346287033473,
+      "grad_norm": 0.9083394858444152,
+      "learning_rate": 1.3372014989188098e-05,
+      "loss": 0.7963,
+      "step": 2126
+    },
+    {
+      "epoch": 0.4091958445555983,
+      "grad_norm": 0.8660402348060943,
+      "learning_rate": 1.3366148413896851e-05,
+      "loss": 0.8216,
+      "step": 2127
+    },
+    {
+      "epoch": 0.40938822624086185,
+      "grad_norm": 0.9292172692338028,
+      "learning_rate": 1.3360280531750886e-05,
+      "loss": 0.9575,
+      "step": 2128
+    },
+    {
+      "epoch": 0.40958060792612544,
+      "grad_norm": 0.8185270021430485,
+      "learning_rate": 1.3354411345028324e-05,
+      "loss": 0.8955,
+      "step": 2129
+    },
+    {
+      "epoch": 0.40977298961138897,
+      "grad_norm": 0.8056711943261903,
+      "learning_rate": 1.3348540856007782e-05,
+      "loss": 0.8244,
+      "step": 2130
+    },
+    {
+      "epoch": 0.40996537129665256,
+      "grad_norm": 0.8507695931234929,
+      "learning_rate": 1.3342669066968385e-05,
+      "loss": 0.8377,
+      "step": 2131
+    },
+    {
+      "epoch": 0.41015775298191615,
+      "grad_norm": 0.8316127429533285,
+      "learning_rate": 1.3336795980189763e-05,
+      "loss": 0.8466,
+      "step": 2132
+    },
+    {
+      "epoch": 0.4103501346671797,
+      "grad_norm": 0.8548338002738683,
+      "learning_rate": 1.3330921597952056e-05,
+      "loss": 0.8148,
+      "step": 2133
+    },
+    {
+      "epoch": 0.41054251635244327,
+      "grad_norm": 0.7958709150664741,
+      "learning_rate": 1.3325045922535896e-05,
+      "loss": 0.8594,
+      "step": 2134
+    },
+    {
+      "epoch": 0.4107348980377068,
+      "grad_norm": 0.9272042844594004,
+      "learning_rate": 1.3319168956222423e-05,
+      "loss": 0.8191,
+      "step": 2135
+    },
+    {
+      "epoch": 0.4109272797229704,
+      "grad_norm": 0.8792275241676584,
+      "learning_rate": 1.331329070129328e-05,
+      "loss": 0.8292,
+      "step": 2136
+    },
+    {
+      "epoch": 0.4111196614082339,
+      "grad_norm": 0.8585224524031134,
+      "learning_rate": 1.3307411160030608e-05,
+      "loss": 0.8559,
+      "step": 2137
+    },
+    {
+      "epoch": 0.4113120430934975,
+      "grad_norm": 0.8492770086512145,
+      "learning_rate": 1.3301530334717046e-05,
+      "loss": 0.8896,
+      "step": 2138
+    },
+    {
+      "epoch": 0.41150442477876104,
+      "grad_norm": 0.822310539047366,
+      "learning_rate": 1.3295648227635729e-05,
+      "loss": 0.8312,
+      "step": 2139
+    },
+    {
+      "epoch": 0.4116968064640246,
+      "grad_norm": 0.8519713381451982,
+      "learning_rate": 1.32897648410703e-05,
+      "loss": 0.8574,
+      "step": 2140
+    },
+    {
+      "epoch": 0.4118891881492882,
+      "grad_norm": 0.8745660050249422,
+      "learning_rate": 1.328388017730489e-05,
+      "loss": 0.8273,
+      "step": 2141
+    },
+    {
+      "epoch": 0.41208156983455174,
+      "grad_norm": 0.6871288933833851,
+      "learning_rate": 1.327799423862413e-05,
+      "loss": 0.3388,
+      "step": 2142
+    },
+    {
+      "epoch": 0.41227395151981533,
+      "grad_norm": 0.8327562999551475,
+      "learning_rate": 1.3272107027313142e-05,
+      "loss": 0.84,
+      "step": 2143
+    },
+    {
+      "epoch": 0.41246633320507886,
+      "grad_norm": 0.8631871633207269,
+      "learning_rate": 1.3266218545657541e-05,
+      "loss": 0.8598,
+      "step": 2144
+    },
+    {
+      "epoch": 0.41265871489034245,
+      "grad_norm": 0.8671806052719575,
+      "learning_rate": 1.326032879594344e-05,
+      "loss": 0.8027,
+      "step": 2145
+    },
+    {
+      "epoch": 0.412851096575606,
+      "grad_norm": 0.8882227954849092,
+      "learning_rate": 1.3254437780457448e-05,
+      "loss": 0.8388,
+      "step": 2146
+    },
+    {
+      "epoch": 0.41304347826086957,
+      "grad_norm": 0.87130207568727,
+      "learning_rate": 1.3248545501486654e-05,
+      "loss": 0.8451,
+      "step": 2147
+    },
+    {
+      "epoch": 0.41323585994613315,
+      "grad_norm": 0.8619761852303519,
+      "learning_rate": 1.3242651961318646e-05,
+      "loss": 0.8058,
+      "step": 2148
+    },
+    {
+      "epoch": 0.4134282416313967,
+      "grad_norm": 0.8865884007857314,
+      "learning_rate": 1.32367571622415e-05,
+      "loss": 0.8892,
+      "step": 2149
+    },
+    {
+      "epoch": 0.4136206233166603,
+      "grad_norm": 0.9023223458659257,
+      "learning_rate": 1.3230861106543777e-05,
+      "loss": 0.8774,
+      "step": 2150
+    },
+    {
+      "epoch": 0.4138130050019238,
+      "grad_norm": 0.9073384704122378,
+      "learning_rate": 1.3224963796514532e-05,
+      "loss": 0.8441,
+      "step": 2151
+    },
+    {
+      "epoch": 0.4140053866871874,
+      "grad_norm": 0.8256726575408279,
+      "learning_rate": 1.32190652344433e-05,
+      "loss": 0.8073,
+      "step": 2152
+    },
+    {
+      "epoch": 0.4141977683724509,
+      "grad_norm": 0.8900900148454839,
+      "learning_rate": 1.3213165422620111e-05,
+      "loss": 0.824,
+      "step": 2153
+    },
+    {
+      "epoch": 0.4143901500577145,
+      "grad_norm": 0.642048262530444,
+      "learning_rate": 1.3207264363335472e-05,
+      "loss": 0.3343,
+      "step": 2154
+    },
+    {
+      "epoch": 0.41458253174297804,
+      "grad_norm": 0.9215675725312158,
+      "learning_rate": 1.3201362058880375e-05,
+      "loss": 0.8167,
+      "step": 2155
+    },
+    {
+      "epoch": 0.41477491342824163,
+      "grad_norm": 0.8182481135675947,
+      "learning_rate": 1.3195458511546307e-05,
+      "loss": 0.8234,
+      "step": 2156
+    },
+    {
+      "epoch": 0.4149672951135052,
+      "grad_norm": 0.9093920991091725,
+      "learning_rate": 1.3189553723625217e-05,
+      "loss": 0.8911,
+      "step": 2157
+    },
+    {
+      "epoch": 0.41515967679876875,
+      "grad_norm": 0.7863934703154372,
+      "learning_rate": 1.318364769740955e-05,
+      "loss": 0.819,
+      "step": 2158
+    },
+    {
+      "epoch": 0.41535205848403234,
+      "grad_norm": 0.8736339657120485,
+      "learning_rate": 1.3177740435192235e-05,
+      "loss": 0.8437,
+      "step": 2159
+    },
+    {
+      "epoch": 0.41554444016929587,
+      "grad_norm": 0.9676674752638393,
+      "learning_rate": 1.3171831939266668e-05,
+      "loss": 0.8889,
+      "step": 2160
+    },
+    {
+      "epoch": 0.41573682185455946,
+      "grad_norm": 0.8089419131722437,
+      "learning_rate": 1.3165922211926734e-05,
+      "loss": 0.8025,
+      "step": 2161
+    },
+    {
+      "epoch": 0.415929203539823,
+      "grad_norm": 0.9201245427670722,
+      "learning_rate": 1.3160011255466791e-05,
+      "loss": 0.8293,
+      "step": 2162
+    },
+    {
+      "epoch": 0.4161215852250866,
+      "grad_norm": 0.9217506405692699,
+      "learning_rate": 1.3154099072181677e-05,
+      "loss": 0.8511,
+      "step": 2163
+    },
+    {
+      "epoch": 0.4163139669103501,
+      "grad_norm": 0.8280179190869759,
+      "learning_rate": 1.3148185664366704e-05,
+      "loss": 0.8117,
+      "step": 2164
+    },
+    {
+      "epoch": 0.4165063485956137,
+      "grad_norm": 0.6909424560599243,
+      "learning_rate": 1.314227103431766e-05,
+      "loss": 0.3216,
+      "step": 2165
+    },
+    {
+      "epoch": 0.4166987302808773,
+      "grad_norm": 0.9478086641583104,
+      "learning_rate": 1.3136355184330809e-05,
+      "loss": 0.828,
+      "step": 2166
+    },
+    {
+      "epoch": 0.4168911119661408,
+      "grad_norm": 0.8889519594156869,
+      "learning_rate": 1.3130438116702888e-05,
+      "loss": 0.8389,
+      "step": 2167
+    },
+    {
+      "epoch": 0.4170834936514044,
+      "grad_norm": 0.8773975800858947,
+      "learning_rate": 1.3124519833731106e-05,
+      "loss": 0.8029,
+      "step": 2168
+    },
+    {
+      "epoch": 0.41727587533666793,
+      "grad_norm": 0.870162774886634,
+      "learning_rate": 1.3118600337713146e-05,
+      "loss": 0.8611,
+      "step": 2169
+    },
+    {
+      "epoch": 0.4174682570219315,
+      "grad_norm": 0.8809534097453019,
+      "learning_rate": 1.3112679630947156e-05,
+      "loss": 0.8858,
+      "step": 2170
+    },
+    {
+      "epoch": 0.41766063870719505,
+      "grad_norm": 0.9114460943762933,
+      "learning_rate": 1.310675771573176e-05,
+      "loss": 0.8271,
+      "step": 2171
+    },
+    {
+      "epoch": 0.41785302039245864,
+      "grad_norm": 0.8475586534081836,
+      "learning_rate": 1.310083459436605e-05,
+      "loss": 0.8097,
+      "step": 2172
+    },
+    {
+      "epoch": 0.41804540207772223,
+      "grad_norm": 0.8453208138640264,
+      "learning_rate": 1.3094910269149587e-05,
+      "loss": 0.8359,
+      "step": 2173
+    },
+    {
+      "epoch": 0.41823778376298576,
+      "grad_norm": 0.8518721239645016,
+      "learning_rate": 1.3088984742382395e-05,
+      "loss": 0.8341,
+      "step": 2174
+    },
+    {
+      "epoch": 0.41843016544824935,
+      "grad_norm": 0.8924094147926884,
+      "learning_rate": 1.3083058016364972e-05,
+      "loss": 0.8726,
+      "step": 2175
+    },
+    {
+      "epoch": 0.4186225471335129,
+      "grad_norm": 0.9527946868650066,
+      "learning_rate": 1.3077130093398274e-05,
+      "loss": 0.7995,
+      "step": 2176
+    },
+    {
+      "epoch": 0.41881492881877647,
+      "grad_norm": 0.8733966884535977,
+      "learning_rate": 1.3071200975783725e-05,
+      "loss": 0.8936,
+      "step": 2177
+    },
+    {
+      "epoch": 0.41900731050404,
+      "grad_norm": 0.8688979444358064,
+      "learning_rate": 1.3065270665823206e-05,
+      "loss": 0.8286,
+      "step": 2178
+    },
+    {
+      "epoch": 0.4191996921893036,
+      "grad_norm": 0.9156848442586754,
+      "learning_rate": 1.3059339165819082e-05,
+      "loss": 0.8517,
+      "step": 2179
+    },
+    {
+      "epoch": 0.4193920738745671,
+      "grad_norm": 0.8617316183884849,
+      "learning_rate": 1.3053406478074155e-05,
+      "loss": 0.8741,
+      "step": 2180
+    },
+    {
+      "epoch": 0.4195844555598307,
+      "grad_norm": 0.8731971484451726,
+      "learning_rate": 1.3047472604891701e-05,
+      "loss": 0.7671,
+      "step": 2181
+    },
+    {
+      "epoch": 0.4197768372450943,
+      "grad_norm": 0.827308957257778,
+      "learning_rate": 1.3041537548575455e-05,
+      "loss": 0.8528,
+      "step": 2182
+    },
+    {
+      "epoch": 0.4199692189303578,
+      "grad_norm": 0.873602385717881,
+      "learning_rate": 1.303560131142961e-05,
+      "loss": 0.8547,
+      "step": 2183
+    },
+    {
+      "epoch": 0.4201616006156214,
+      "grad_norm": 0.8953621930906135,
+      "learning_rate": 1.3029663895758814e-05,
+      "loss": 0.7747,
+      "step": 2184
+    },
+    {
+      "epoch": 0.42035398230088494,
+      "grad_norm": 0.8685514290005595,
+      "learning_rate": 1.3023725303868183e-05,
+      "loss": 0.8271,
+      "step": 2185
+    },
+    {
+      "epoch": 0.42054636398614853,
+      "grad_norm": 0.9187398496640021,
+      "learning_rate": 1.3017785538063277e-05,
+      "loss": 0.8712,
+      "step": 2186
+    },
+    {
+      "epoch": 0.42073874567141206,
+      "grad_norm": 0.8887954599782006,
+      "learning_rate": 1.3011844600650121e-05,
+      "loss": 0.8433,
+      "step": 2187
+    },
+    {
+      "epoch": 0.42093112735667565,
+      "grad_norm": 0.8391747994177812,
+      "learning_rate": 1.300590249393519e-05,
+      "loss": 0.8786,
+      "step": 2188
+    },
+    {
+      "epoch": 0.4211235090419392,
+      "grad_norm": 0.8751520766981662,
+      "learning_rate": 1.2999959220225416e-05,
+      "loss": 0.8701,
+      "step": 2189
+    },
+    {
+      "epoch": 0.42131589072720277,
+      "grad_norm": 0.6396119697634094,
+      "learning_rate": 1.299401478182818e-05,
+      "loss": 0.3401,
+      "step": 2190
+    },
+    {
+      "epoch": 0.42150827241246636,
+      "grad_norm": 0.8101184132452338,
+      "learning_rate": 1.2988069181051314e-05,
+      "loss": 0.8472,
+      "step": 2191
+    },
+    {
+      "epoch": 0.4217006540977299,
+      "grad_norm": 0.8878249576823884,
+      "learning_rate": 1.2982122420203114e-05,
+      "loss": 0.8528,
+      "step": 2192
+    },
+    {
+      "epoch": 0.4218930357829935,
+      "grad_norm": 0.8110348738012501,
+      "learning_rate": 1.2976174501592313e-05,
+      "loss": 0.8248,
+      "step": 2193
+    },
+    {
+      "epoch": 0.422085417468257,
+      "grad_norm": 0.8039893085986051,
+      "learning_rate": 1.2970225427528098e-05,
+      "loss": 0.8581,
+      "step": 2194
+    },
+    {
+      "epoch": 0.4222777991535206,
+      "grad_norm": 0.7869981579694302,
+      "learning_rate": 1.2964275200320104e-05,
+      "loss": 0.8456,
+      "step": 2195
+    },
+    {
+      "epoch": 0.4224701808387841,
+      "grad_norm": 0.8267848613216165,
+      "learning_rate": 1.2958323822278413e-05,
+      "loss": 0.8569,
+      "step": 2196
+    },
+    {
+      "epoch": 0.4226625625240477,
+      "grad_norm": 0.8875043430970213,
+      "learning_rate": 1.2952371295713558e-05,
+      "loss": 0.7905,
+      "step": 2197
+    },
+    {
+      "epoch": 0.4228549442093113,
+      "grad_norm": 0.8575870378840547,
+      "learning_rate": 1.2946417622936512e-05,
+      "loss": 0.8424,
+      "step": 2198
+    },
+    {
+      "epoch": 0.42304732589457483,
+      "grad_norm": 0.8353514575738099,
+      "learning_rate": 1.2940462806258696e-05,
+      "loss": 0.8299,
+      "step": 2199
+    },
+    {
+      "epoch": 0.4232397075798384,
+      "grad_norm": 0.8494221758800369,
+      "learning_rate": 1.2934506847991976e-05,
+      "loss": 0.788,
+      "step": 2200
+    },
+    {
+      "epoch": 0.42343208926510195,
+      "grad_norm": 0.849798467120386,
+      "learning_rate": 1.2928549750448661e-05,
+      "loss": 0.8405,
+      "step": 2201
+    },
+    {
+      "epoch": 0.42362447095036554,
+      "grad_norm": 0.9040356625218666,
+      "learning_rate": 1.2922591515941498e-05,
+      "loss": 0.806,
+      "step": 2202
+    },
+    {
+      "epoch": 0.42381685263562907,
+      "grad_norm": 0.8973255859170027,
+      "learning_rate": 1.2916632146783683e-05,
+      "loss": 0.8913,
+      "step": 2203
+    },
+    {
+      "epoch": 0.42400923432089266,
+      "grad_norm": 0.8722121885876217,
+      "learning_rate": 1.2910671645288841e-05,
+      "loss": 0.7749,
+      "step": 2204
+    },
+    {
+      "epoch": 0.4242016160061562,
+      "grad_norm": 0.9680882422820586,
+      "learning_rate": 1.2904710013771054e-05,
+      "loss": 0.8797,
+      "step": 2205
+    },
+    {
+      "epoch": 0.4243939976914198,
+      "grad_norm": 0.9157317106974844,
+      "learning_rate": 1.2898747254544826e-05,
+      "loss": 0.8064,
+      "step": 2206
+    },
+    {
+      "epoch": 0.42458637937668336,
+      "grad_norm": 0.8725030345809746,
+      "learning_rate": 1.2892783369925105e-05,
+      "loss": 0.8376,
+      "step": 2207
+    },
+    {
+      "epoch": 0.4247787610619469,
+      "grad_norm": 0.8882490044381467,
+      "learning_rate": 1.2886818362227283e-05,
+      "loss": 0.8496,
+      "step": 2208
+    },
+    {
+      "epoch": 0.4249711427472105,
+      "grad_norm": 0.9251347056106201,
+      "learning_rate": 1.2880852233767174e-05,
+      "loss": 0.8638,
+      "step": 2209
+    },
+    {
+      "epoch": 0.425163524432474,
+      "grad_norm": 0.988873582397908,
+      "learning_rate": 1.2874884986861038e-05,
+      "loss": 0.854,
+      "step": 2210
+    },
+    {
+      "epoch": 0.4253559061177376,
+      "grad_norm": 0.9239345198796975,
+      "learning_rate": 1.2868916623825561e-05,
+      "loss": 0.8312,
+      "step": 2211
+    },
+    {
+      "epoch": 0.42554828780300114,
+      "grad_norm": 0.8658460397030684,
+      "learning_rate": 1.2862947146977876e-05,
+      "loss": 0.8507,
+      "step": 2212
+    },
+    {
+      "epoch": 0.4257406694882647,
+      "grad_norm": 0.8692060198791819,
+      "learning_rate": 1.2856976558635532e-05,
+      "loss": 0.8815,
+      "step": 2213
+    },
+    {
+      "epoch": 0.42593305117352825,
+      "grad_norm": 0.9077713598517076,
+      "learning_rate": 1.2851004861116519e-05,
+      "loss": 0.8566,
+      "step": 2214
+    },
+    {
+      "epoch": 0.42612543285879184,
+      "grad_norm": 0.8490150939978919,
+      "learning_rate": 1.2845032056739257e-05,
+      "loss": 0.8152,
+      "step": 2215
+    },
+    {
+      "epoch": 0.42631781454405543,
+      "grad_norm": 0.8365062233979498,
+      "learning_rate": 1.2839058147822595e-05,
+      "loss": 0.8461,
+      "step": 2216
+    },
+    {
+      "epoch": 0.42651019622931896,
+      "grad_norm": 0.8138915180809184,
+      "learning_rate": 1.2833083136685803e-05,
+      "loss": 0.8416,
+      "step": 2217
+    },
+    {
+      "epoch": 0.42670257791458255,
+      "grad_norm": 0.8943605001149315,
+      "learning_rate": 1.2827107025648595e-05,
+      "loss": 0.8492,
+      "step": 2218
+    },
+    {
+      "epoch": 0.4268949595998461,
+      "grad_norm": 0.9689534503154031,
+      "learning_rate": 1.2821129817031099e-05,
+      "loss": 0.8425,
+      "step": 2219
+    },
+    {
+      "epoch": 0.42708734128510967,
+      "grad_norm": 0.8879943311004034,
+      "learning_rate": 1.2815151513153874e-05,
+      "loss": 0.8677,
+      "step": 2220
+    },
+    {
+      "epoch": 0.4272797229703732,
+      "grad_norm": 0.8618633084909669,
+      "learning_rate": 1.2809172116337903e-05,
+      "loss": 0.8352,
+      "step": 2221
+    },
+    {
+      "epoch": 0.4274721046556368,
+      "grad_norm": 0.8248810596123108,
+      "learning_rate": 1.2803191628904594e-05,
+      "loss": 0.7933,
+      "step": 2222
+    },
+    {
+      "epoch": 0.4276644863409004,
+      "grad_norm": 0.8476136571552648,
+      "learning_rate": 1.2797210053175779e-05,
+      "loss": 0.8287,
+      "step": 2223
+    },
+    {
+      "epoch": 0.4278568680261639,
+      "grad_norm": 0.8736225843434168,
+      "learning_rate": 1.2791227391473706e-05,
+      "loss": 0.8142,
+      "step": 2224
+    },
+    {
+      "epoch": 0.4280492497114275,
+      "grad_norm": 0.9515671662749485,
+      "learning_rate": 1.2785243646121059e-05,
+      "loss": 0.8609,
+      "step": 2225
+    },
+    {
+      "epoch": 0.428241631396691,
+      "grad_norm": 0.8963456642222687,
+      "learning_rate": 1.277925881944093e-05,
+      "loss": 0.8701,
+      "step": 2226
+    },
+    {
+      "epoch": 0.4284340130819546,
+      "grad_norm": 0.9106952599365449,
+      "learning_rate": 1.2773272913756833e-05,
+      "loss": 0.9058,
+      "step": 2227
+    },
+    {
+      "epoch": 0.42862639476721814,
+      "grad_norm": 0.881066838040749,
+      "learning_rate": 1.2767285931392705e-05,
+      "loss": 0.8341,
+      "step": 2228
+    },
+    {
+      "epoch": 0.42881877645248173,
+      "grad_norm": 0.9334384586722079,
+      "learning_rate": 1.27612978746729e-05,
+      "loss": 0.886,
+      "step": 2229
+    },
+    {
+      "epoch": 0.42901115813774526,
+      "grad_norm": 0.8490693143684024,
+      "learning_rate": 1.2755308745922182e-05,
+      "loss": 0.842,
+      "step": 2230
+    },
+    {
+      "epoch": 0.42920353982300885,
+      "grad_norm": 0.9334007577216312,
+      "learning_rate": 1.2749318547465742e-05,
+      "loss": 0.7983,
+      "step": 2231
+    },
+    {
+      "epoch": 0.42939592150827244,
+      "grad_norm": 0.9357826663934721,
+      "learning_rate": 1.2743327281629181e-05,
+      "loss": 0.8381,
+      "step": 2232
+    },
+    {
+      "epoch": 0.42958830319353597,
+      "grad_norm": 0.6413378549594018,
+      "learning_rate": 1.2737334950738512e-05,
+      "loss": 0.3458,
+      "step": 2233
+    },
+    {
+      "epoch": 0.42978068487879956,
+      "grad_norm": 0.654714265205191,
+      "learning_rate": 1.273134155712017e-05,
+      "loss": 0.3687,
+      "step": 2234
+    },
+    {
+      "epoch": 0.4299730665640631,
+      "grad_norm": 0.8194828463779399,
+      "learning_rate": 1.272534710310099e-05,
+      "loss": 0.8673,
+      "step": 2235
+    },
+    {
+      "epoch": 0.4301654482493267,
+      "grad_norm": 0.825310537311817,
+      "learning_rate": 1.2719351591008228e-05,
+      "loss": 0.7346,
+      "step": 2236
+    },
+    {
+      "epoch": 0.4303578299345902,
+      "grad_norm": 0.9211172426412261,
+      "learning_rate": 1.2713355023169547e-05,
+      "loss": 0.8694,
+      "step": 2237
+    },
+    {
+      "epoch": 0.4305502116198538,
+      "grad_norm": 0.7852572841863174,
+      "learning_rate": 1.2707357401913022e-05,
+      "loss": 0.8207,
+      "step": 2238
+    },
+    {
+      "epoch": 0.4307425933051173,
+      "grad_norm": 0.8282511701404058,
+      "learning_rate": 1.270135872956714e-05,
+      "loss": 0.8453,
+      "step": 2239
+    },
+    {
+      "epoch": 0.4309349749903809,
+      "grad_norm": 0.9182401895869606,
+      "learning_rate": 1.2695359008460785e-05,
+      "loss": 0.7315,
+      "step": 2240
+    },
+    {
+      "epoch": 0.4311273566756445,
+      "grad_norm": 0.8879249784782809,
+      "learning_rate": 1.2689358240923264e-05,
+      "loss": 0.7965,
+      "step": 2241
+    },
+    {
+      "epoch": 0.43131973836090803,
+      "grad_norm": 0.8210203143246622,
+      "learning_rate": 1.2683356429284273e-05,
+      "loss": 0.8124,
+      "step": 2242
+    },
+    {
+      "epoch": 0.4315121200461716,
+      "grad_norm": 0.8398402680425243,
+      "learning_rate": 1.2677353575873926e-05,
+      "loss": 0.8431,
+      "step": 2243
+    },
+    {
+      "epoch": 0.43170450173143515,
+      "grad_norm": 0.6768808767846733,
+      "learning_rate": 1.2671349683022736e-05,
+      "loss": 0.3675,
+      "step": 2244
+    },
+    {
+      "epoch": 0.43189688341669874,
+      "grad_norm": 0.658058231011825,
+      "learning_rate": 1.2665344753061622e-05,
+      "loss": 0.3674,
+      "step": 2245
+    },
+    {
+      "epoch": 0.43208926510196227,
+      "grad_norm": 0.8981699556134266,
+      "learning_rate": 1.2659338788321904e-05,
+      "loss": 0.8248,
+      "step": 2246
+    },
+    {
+      "epoch": 0.43228164678722586,
+      "grad_norm": 0.9523497848191392,
+      "learning_rate": 1.2653331791135308e-05,
+      "loss": 0.8569,
+      "step": 2247
+    },
+    {
+      "epoch": 0.4324740284724894,
+      "grad_norm": 0.9399920191152544,
+      "learning_rate": 1.2647323763833952e-05,
+      "loss": 0.7965,
+      "step": 2248
+    },
+    {
+      "epoch": 0.432666410157753,
+      "grad_norm": 0.9473346283801329,
+      "learning_rate": 1.264131470875036e-05,
+      "loss": 0.8646,
+      "step": 2249
+    },
+    {
+      "epoch": 0.43285879184301657,
+      "grad_norm": 0.9549585322513965,
+      "learning_rate": 1.2635304628217452e-05,
+      "loss": 0.9248,
+      "step": 2250
+    },
+    {
+      "epoch": 0.4330511735282801,
+      "grad_norm": 0.6885649282452616,
+      "learning_rate": 1.2629293524568555e-05,
+      "loss": 0.3453,
+      "step": 2251
+    },
+    {
+      "epoch": 0.4332435552135437,
+      "grad_norm": 0.8415557389523821,
+      "learning_rate": 1.2623281400137383e-05,
+      "loss": 0.7847,
+      "step": 2252
+    },
+    {
+      "epoch": 0.4334359368988072,
+      "grad_norm": 0.9821809520186289,
+      "learning_rate": 1.2617268257258051e-05,
+      "loss": 0.8675,
+      "step": 2253
+    },
+    {
+      "epoch": 0.4336283185840708,
+      "grad_norm": 0.8570795919307674,
+      "learning_rate": 1.2611254098265063e-05,
+      "loss": 0.8521,
+      "step": 2254
+    },
+    {
+      "epoch": 0.43382070026933434,
+      "grad_norm": 0.8723955451828692,
+      "learning_rate": 1.2605238925493326e-05,
+      "loss": 0.8991,
+      "step": 2255
+    },
+    {
+      "epoch": 0.4340130819545979,
+      "grad_norm": 0.9020317826308721,
+      "learning_rate": 1.2599222741278136e-05,
+      "loss": 0.7946,
+      "step": 2256
+    },
+    {
+      "epoch": 0.4342054636398615,
+      "grad_norm": 0.8736964816451105,
+      "learning_rate": 1.2593205547955185e-05,
+      "loss": 0.8759,
+      "step": 2257
+    },
+    {
+      "epoch": 0.43439784532512504,
+      "grad_norm": 0.8797053393813774,
+      "learning_rate": 1.2587187347860554e-05,
+      "loss": 0.7939,
+      "step": 2258
+    },
+    {
+      "epoch": 0.43459022701038863,
+      "grad_norm": 0.8605835312649487,
+      "learning_rate": 1.2581168143330716e-05,
+      "loss": 0.8265,
+      "step": 2259
+    },
+    {
+      "epoch": 0.43478260869565216,
+      "grad_norm": 0.8993240265906781,
+      "learning_rate": 1.2575147936702531e-05,
+      "loss": 0.8301,
+      "step": 2260
+    },
+    {
+      "epoch": 0.43497499038091575,
+      "grad_norm": 0.9548724857122258,
+      "learning_rate": 1.2569126730313255e-05,
+      "loss": 0.8631,
+      "step": 2261
+    },
+    {
+      "epoch": 0.4351673720661793,
+      "grad_norm": 0.6404311941697903,
+      "learning_rate": 1.2563104526500523e-05,
+      "loss": 0.3633,
+      "step": 2262
+    },
+    {
+      "epoch": 0.43535975375144287,
+      "grad_norm": 0.9690144651174785,
+      "learning_rate": 1.2557081327602361e-05,
+      "loss": 0.7994,
+      "step": 2263
+    },
+    {
+      "epoch": 0.4355521354367064,
+      "grad_norm": 0.979341240837771,
+      "learning_rate": 1.2551057135957187e-05,
+      "loss": 0.8122,
+      "step": 2264
+    },
+    {
+      "epoch": 0.43574451712197,
+      "grad_norm": 0.8405954115766662,
+      "learning_rate": 1.2545031953903796e-05,
+      "loss": 0.8286,
+      "step": 2265
+    },
+    {
+      "epoch": 0.4359368988072336,
+      "grad_norm": 0.8787830934398129,
+      "learning_rate": 1.2539005783781374e-05,
+      "loss": 0.8824,
+      "step": 2266
+    },
+    {
+      "epoch": 0.4361292804924971,
+      "grad_norm": 0.8673129568173078,
+      "learning_rate": 1.2532978627929486e-05,
+      "loss": 0.8648,
+      "step": 2267
+    },
+    {
+      "epoch": 0.4363216621777607,
+      "grad_norm": 1.0300622806925044,
+      "learning_rate": 1.2526950488688083e-05,
+      "loss": 0.8462,
+      "step": 2268
+    },
+    {
+      "epoch": 0.4365140438630242,
+      "grad_norm": 0.8877205972468946,
+      "learning_rate": 1.2520921368397492e-05,
+      "loss": 0.8211,
+      "step": 2269
+    },
+    {
+      "epoch": 0.4367064255482878,
+      "grad_norm": 0.8647502343229451,
+      "learning_rate": 1.2514891269398429e-05,
+      "loss": 0.8086,
+      "step": 2270
+    },
+    {
+      "epoch": 0.43689880723355135,
+      "grad_norm": 0.6721300572403824,
+      "learning_rate": 1.2508860194031986e-05,
+      "loss": 0.3567,
+      "step": 2271
+    },
+    {
+      "epoch": 0.43709118891881493,
+      "grad_norm": 0.8029781445510202,
+      "learning_rate": 1.2502828144639629e-05,
+      "loss": 0.8271,
+      "step": 2272
+    },
+    {
+      "epoch": 0.43728357060407846,
+      "grad_norm": 0.9581039117267324,
+      "learning_rate": 1.2496795123563218e-05,
+      "loss": 0.8525,
+      "step": 2273
+    },
+    {
+      "epoch": 0.43747595228934205,
+      "grad_norm": 0.8522866125820424,
+      "learning_rate": 1.249076113314497e-05,
+      "loss": 0.7477,
+      "step": 2274
+    },
+    {
+      "epoch": 0.43766833397460564,
+      "grad_norm": 0.8782095454303463,
+      "learning_rate": 1.248472617572749e-05,
+      "loss": 0.8609,
+      "step": 2275
+    },
+    {
+      "epoch": 0.43786071565986917,
+      "grad_norm": 0.8784850687377274,
+      "learning_rate": 1.2478690253653756e-05,
+      "loss": 0.8245,
+      "step": 2276
+    },
+    {
+      "epoch": 0.43805309734513276,
+      "grad_norm": 0.9543109010558956,
+      "learning_rate": 1.2472653369267122e-05,
+      "loss": 0.7552,
+      "step": 2277
+    },
+    {
+      "epoch": 0.4382454790303963,
+      "grad_norm": 0.8749080491364793,
+      "learning_rate": 1.2466615524911316e-05,
+      "loss": 0.8218,
+      "step": 2278
+    },
+    {
+      "epoch": 0.4384378607156599,
+      "grad_norm": 0.9155654924046184,
+      "learning_rate": 1.2460576722930432e-05,
+      "loss": 0.8404,
+      "step": 2279
+    },
+    {
+      "epoch": 0.4386302424009234,
+      "grad_norm": 0.9386701697028416,
+      "learning_rate": 1.2454536965668949e-05,
+      "loss": 0.9065,
+      "step": 2280
+    },
+    {
+      "epoch": 0.438822624086187,
+      "grad_norm": 0.8395113561196779,
+      "learning_rate": 1.24484962554717e-05,
+      "loss": 0.77,
+      "step": 2281
+    },
+    {
+      "epoch": 0.4390150057714506,
+      "grad_norm": 0.8667175303804131,
+      "learning_rate": 1.24424545946839e-05,
+      "loss": 0.8921,
+      "step": 2282
+    },
+    {
+      "epoch": 0.4392073874567141,
+      "grad_norm": 0.8083154357711307,
+      "learning_rate": 1.2436411985651131e-05,
+      "loss": 0.813,
+      "step": 2283
+    },
+    {
+      "epoch": 0.4393997691419777,
+      "grad_norm": 0.9849313230559015,
+      "learning_rate": 1.2430368430719342e-05,
+      "loss": 0.8769,
+      "step": 2284
+    },
+    {
+      "epoch": 0.43959215082724123,
+      "grad_norm": 0.9594210739746961,
+      "learning_rate": 1.242432393223485e-05,
+      "loss": 0.8059,
+      "step": 2285
+    },
+    {
+      "epoch": 0.4397845325125048,
+      "grad_norm": 0.9348610992949489,
+      "learning_rate": 1.2418278492544328e-05,
+      "loss": 0.8175,
+      "step": 2286
+    },
+    {
+      "epoch": 0.43997691419776835,
+      "grad_norm": 0.8418270216349524,
+      "learning_rate": 1.2412232113994841e-05,
+      "loss": 0.8393,
+      "step": 2287
+    },
+    {
+      "epoch": 0.44016929588303194,
+      "grad_norm": 0.7837091974158702,
+      "learning_rate": 1.2406184798933786e-05,
+      "loss": 0.843,
+      "step": 2288
+    },
+    {
+      "epoch": 0.4403616775682955,
+      "grad_norm": 0.7890722212914054,
+      "learning_rate": 1.2400136549708945e-05,
+      "loss": 0.8504,
+      "step": 2289
+    },
+    {
+      "epoch": 0.44055405925355906,
+      "grad_norm": 0.8746819688566863,
+      "learning_rate": 1.239408736866846e-05,
+      "loss": 0.7914,
+      "step": 2290
+    },
+    {
+      "epoch": 0.44074644093882265,
+      "grad_norm": 0.8533177406174416,
+      "learning_rate": 1.2388037258160823e-05,
+      "loss": 0.8087,
+      "step": 2291
+    },
+    {
+      "epoch": 0.4409388226240862,
+      "grad_norm": 0.8569923631961269,
+      "learning_rate": 1.23819862205349e-05,
+      "loss": 0.8334,
+      "step": 2292
+    },
+    {
+      "epoch": 0.44113120430934977,
+      "grad_norm": 0.8219958285201809,
+      "learning_rate": 1.2375934258139917e-05,
+      "loss": 0.8551,
+      "step": 2293
+    },
+    {
+      "epoch": 0.4413235859946133,
+      "grad_norm": 0.8822379117605617,
+      "learning_rate": 1.2369881373325448e-05,
+      "loss": 0.8958,
+      "step": 2294
+    },
+    {
+      "epoch": 0.4415159676798769,
+      "grad_norm": 0.6779474466824668,
+      "learning_rate": 1.236382756844143e-05,
+      "loss": 0.3504,
+      "step": 2295
+    },
+    {
+      "epoch": 0.4417083493651404,
+      "grad_norm": 0.8278814836902233,
+      "learning_rate": 1.2357772845838159e-05,
+      "loss": 0.7961,
+      "step": 2296
+    },
+    {
+      "epoch": 0.441900731050404,
+      "grad_norm": 0.8812975503747402,
+      "learning_rate": 1.2351717207866292e-05,
+      "loss": 0.8102,
+      "step": 2297
+    },
+    {
+      "epoch": 0.44209311273566754,
+      "grad_norm": 0.8885034940493376,
+      "learning_rate": 1.2345660656876832e-05,
+      "loss": 0.8721,
+      "step": 2298
+    },
+    {
+      "epoch": 0.4422854944209311,
+      "grad_norm": 0.9144045388318797,
+      "learning_rate": 1.233960319522114e-05,
+      "loss": 0.8722,
+      "step": 2299
+    },
+    {
+      "epoch": 0.4424778761061947,
+      "grad_norm": 0.8715720545544644,
+      "learning_rate": 1.2333544825250938e-05,
+      "loss": 0.8061,
+      "step": 2300
+    },
+    {
+      "epoch": 0.44267025779145824,
+      "grad_norm": 0.8627272682096624,
+      "learning_rate": 1.2327485549318285e-05,
+      "loss": 0.8555,
+      "step": 2301
+    },
+    {
+      "epoch": 0.44286263947672183,
+      "grad_norm": 0.8499651762528888,
+      "learning_rate": 1.2321425369775601e-05,
+      "loss": 0.8231,
+      "step": 2302
+    },
+    {
+      "epoch": 0.44305502116198536,
+      "grad_norm": 0.8575208307868059,
+      "learning_rate": 1.2315364288975665e-05,
+      "loss": 0.8116,
+      "step": 2303
+    },
+    {
+      "epoch": 0.44324740284724895,
+      "grad_norm": 0.8509648935062077,
+      "learning_rate": 1.2309302309271587e-05,
+      "loss": 0.8462,
+      "step": 2304
+    },
+    {
+      "epoch": 0.4434397845325125,
+      "grad_norm": 0.8829272117798993,
+      "learning_rate": 1.2303239433016842e-05,
+      "loss": 0.8559,
+      "step": 2305
+    },
+    {
+      "epoch": 0.44363216621777607,
+      "grad_norm": 0.6599184630513326,
+      "learning_rate": 1.2297175662565248e-05,
+      "loss": 0.3442,
+      "step": 2306
+    },
+    {
+      "epoch": 0.44382454790303966,
+      "grad_norm": 0.9550684063841587,
+      "learning_rate": 1.229111100027097e-05,
+      "loss": 0.8598,
+      "step": 2307
+    },
+    {
+      "epoch": 0.4440169295883032,
+      "grad_norm": 0.6298707495256025,
+      "learning_rate": 1.228504544848851e-05,
+      "loss": 0.3514,
+      "step": 2308
+    },
+    {
+      "epoch": 0.4442093112735668,
+      "grad_norm": 0.8475026345010844,
+      "learning_rate": 1.2278979009572736e-05,
+      "loss": 0.7592,
+      "step": 2309
+    },
+    {
+      "epoch": 0.4444016929588303,
+      "grad_norm": 0.8583876529295867,
+      "learning_rate": 1.2272911685878841e-05,
+      "loss": 0.832,
+      "step": 2310
+    },
+    {
+      "epoch": 0.4445940746440939,
+      "grad_norm": 0.8692092806875851,
+      "learning_rate": 1.2266843479762372e-05,
+      "loss": 0.8002,
+      "step": 2311
+    },
+    {
+      "epoch": 0.4447864563293574,
+      "grad_norm": 0.8190856252627713,
+      "learning_rate": 1.2260774393579209e-05,
+      "loss": 0.8495,
+      "step": 2312
+    },
+    {
+      "epoch": 0.444978838014621,
+      "grad_norm": 0.9253167710556744,
+      "learning_rate": 1.2254704429685593e-05,
+      "loss": 0.8585,
+      "step": 2313
+    },
+    {
+      "epoch": 0.44517121969988455,
+      "grad_norm": 0.7984204359738976,
+      "learning_rate": 1.2248633590438084e-05,
+      "loss": 0.7992,
+      "step": 2314
+    },
+    {
+      "epoch": 0.44536360138514813,
+      "grad_norm": 0.9356268611317827,
+      "learning_rate": 1.2242561878193589e-05,
+      "loss": 0.8602,
+      "step": 2315
+    },
+    {
+      "epoch": 0.4455559830704117,
+      "grad_norm": 0.6589914177823059,
+      "learning_rate": 1.2236489295309362e-05,
+      "loss": 0.3117,
+      "step": 2316
+    },
+    {
+      "epoch": 0.44574836475567525,
+      "grad_norm": 0.9891784530486593,
+      "learning_rate": 1.2230415844142984e-05,
+      "loss": 0.8676,
+      "step": 2317
+    },
+    {
+      "epoch": 0.44594074644093884,
+      "grad_norm": 0.7882500800965437,
+      "learning_rate": 1.2224341527052378e-05,
+      "loss": 0.8244,
+      "step": 2318
+    },
+    {
+      "epoch": 0.44613312812620237,
+      "grad_norm": 0.9925607101851653,
+      "learning_rate": 1.2218266346395811e-05,
+      "loss": 0.8658,
+      "step": 2319
+    },
+    {
+      "epoch": 0.44632550981146596,
+      "grad_norm": 0.8871356612271751,
+      "learning_rate": 1.221219030453187e-05,
+      "loss": 0.8478,
+      "step": 2320
+    },
+    {
+      "epoch": 0.4465178914967295,
+      "grad_norm": 0.768644413689389,
+      "learning_rate": 1.220611340381948e-05,
+      "loss": 0.8325,
+      "step": 2321
+    },
+    {
+      "epoch": 0.4467102731819931,
+      "grad_norm": 0.8264822913104579,
+      "learning_rate": 1.2200035646617912e-05,
+      "loss": 0.778,
+      "step": 2322
+    },
+    {
+      "epoch": 0.4469026548672566,
+      "grad_norm": 0.9006044490923129,
+      "learning_rate": 1.2193957035286757e-05,
+      "loss": 0.8273,
+      "step": 2323
+    },
+    {
+      "epoch": 0.4470950365525202,
+      "grad_norm": 0.9397047731628465,
+      "learning_rate": 1.2187877572185937e-05,
+      "loss": 0.8838,
+      "step": 2324
+    },
+    {
+      "epoch": 0.4472874182377838,
+      "grad_norm": 0.881532892810587,
+      "learning_rate": 1.2181797259675713e-05,
+      "loss": 0.8756,
+      "step": 2325
+    },
+    {
+      "epoch": 0.4474797999230473,
+      "grad_norm": 0.8745822159129341,
+      "learning_rate": 1.2175716100116677e-05,
+      "loss": 0.8283,
+      "step": 2326
+    },
+    {
+      "epoch": 0.4476721816083109,
+      "grad_norm": 0.9301603161622939,
+      "learning_rate": 1.2169634095869736e-05,
+      "loss": 0.8932,
+      "step": 2327
+    },
+    {
+      "epoch": 0.44786456329357444,
+      "grad_norm": 0.8767820210237329,
+      "learning_rate": 1.2163551249296132e-05,
+      "loss": 0.8176,
+      "step": 2328
+    },
+    {
+      "epoch": 0.448056944978838,
+      "grad_norm": 0.8764014410465126,
+      "learning_rate": 1.2157467562757443e-05,
+      "loss": 0.8141,
+      "step": 2329
+    },
+    {
+      "epoch": 0.44824932666410156,
+      "grad_norm": 0.8879819798839603,
+      "learning_rate": 1.2151383038615563e-05,
+      "loss": 0.8143,
+      "step": 2330
+    },
+    {
+      "epoch": 0.44844170834936514,
+      "grad_norm": 0.8759476105731123,
+      "learning_rate": 1.214529767923271e-05,
+      "loss": 0.8239,
+      "step": 2331
+    },
+    {
+      "epoch": 0.44863409003462873,
+      "grad_norm": 0.8637613797576239,
+      "learning_rate": 1.2139211486971436e-05,
+      "loss": 0.8397,
+      "step": 2332
+    },
+    {
+      "epoch": 0.44882647171989226,
+      "grad_norm": 0.8704715966740368,
+      "learning_rate": 1.213312446419461e-05,
+      "loss": 0.8609,
+      "step": 2333
+    },
+    {
+      "epoch": 0.44901885340515585,
+      "grad_norm": 0.8464466923506119,
+      "learning_rate": 1.2127036613265418e-05,
+      "loss": 0.801,
+      "step": 2334
+    },
+    {
+      "epoch": 0.4492112350904194,
+      "grad_norm": 0.7110371969732149,
+      "learning_rate": 1.2120947936547375e-05,
+      "loss": 0.315,
+      "step": 2335
+    },
+    {
+      "epoch": 0.44940361677568297,
+      "grad_norm": 0.6948506900195015,
+      "learning_rate": 1.2114858436404322e-05,
+      "loss": 0.3635,
+      "step": 2336
+    },
+    {
+      "epoch": 0.4495959984609465,
+      "grad_norm": 0.9615952743470197,
+      "learning_rate": 1.2108768115200405e-05,
+      "loss": 0.864,
+      "step": 2337
+    },
+    {
+      "epoch": 0.4497883801462101,
+      "grad_norm": 0.9271754933696533,
+      "learning_rate": 1.2102676975300095e-05,
+      "loss": 0.827,
+      "step": 2338
+    },
+    {
+      "epoch": 0.4499807618314736,
+      "grad_norm": 0.8537188344856211,
+      "learning_rate": 1.209658501906819e-05,
+      "loss": 0.7657,
+      "step": 2339
+    },
+    {
+      "epoch": 0.4501731435167372,
+      "grad_norm": 0.8825399725168994,
+      "learning_rate": 1.2090492248869795e-05,
+      "loss": 0.8667,
+      "step": 2340
+    },
+    {
+      "epoch": 0.4503655252020008,
+      "grad_norm": 0.8100573289536702,
+      "learning_rate": 1.2084398667070325e-05,
+      "loss": 0.8333,
+      "step": 2341
+    },
+    {
+      "epoch": 0.4505579068872643,
+      "grad_norm": 0.876887308584304,
+      "learning_rate": 1.2078304276035527e-05,
+      "loss": 0.8432,
+      "step": 2342
+    },
+    {
+      "epoch": 0.4507502885725279,
+      "grad_norm": 0.9059098163751174,
+      "learning_rate": 1.2072209078131451e-05,
+      "loss": 0.8365,
+      "step": 2343
+    },
+    {
+      "epoch": 0.45094267025779144,
+      "grad_norm": 0.8879131822950563,
+      "learning_rate": 1.2066113075724461e-05,
+      "loss": 0.8586,
+      "step": 2344
+    },
+    {
+      "epoch": 0.45113505194305503,
+      "grad_norm": 0.8011037750228858,
+      "learning_rate": 1.206001627118124e-05,
+      "loss": 0.786,
+      "step": 2345
+    },
+    {
+      "epoch": 0.45132743362831856,
+      "grad_norm": 0.8220170723266673,
+      "learning_rate": 1.2053918666868776e-05,
+      "loss": 0.7813,
+      "step": 2346
+    },
+    {
+      "epoch": 0.45151981531358215,
+      "grad_norm": 0.8995644963734111,
+      "learning_rate": 1.2047820265154362e-05,
+      "loss": 0.842,
+      "step": 2347
+    },
+    {
+      "epoch": 0.4517121969988457,
+      "grad_norm": 0.8560093708839547,
+      "learning_rate": 1.2041721068405614e-05,
+      "loss": 0.7949,
+      "step": 2348
+    },
+    {
+      "epoch": 0.45190457868410927,
+      "grad_norm": 0.8725689160342955,
+      "learning_rate": 1.203562107899045e-05,
+      "loss": 0.8188,
+      "step": 2349
+    },
+    {
+      "epoch": 0.45209696036937286,
+      "grad_norm": 0.9273485870912288,
+      "learning_rate": 1.2029520299277095e-05,
+      "loss": 0.8914,
+      "step": 2350
+    },
+    {
+      "epoch": 0.4522893420546364,
+      "grad_norm": 0.9085769525612106,
+      "learning_rate": 1.2023418731634078e-05,
+      "loss": 0.83,
+      "step": 2351
+    },
+    {
+      "epoch": 0.4524817237399,
+      "grad_norm": 0.9740824493730424,
+      "learning_rate": 1.2017316378430244e-05,
+      "loss": 0.8277,
+      "step": 2352
+    },
+    {
+      "epoch": 0.4526741054251635,
+      "grad_norm": 0.8452749519729551,
+      "learning_rate": 1.2011213242034733e-05,
+      "loss": 0.7869,
+      "step": 2353
+    },
+    {
+      "epoch": 0.4528664871104271,
+      "grad_norm": 0.9527406243156945,
+      "learning_rate": 1.2005109324816992e-05,
+      "loss": 0.9221,
+      "step": 2354
+    },
+    {
+      "epoch": 0.45305886879569063,
+      "grad_norm": 0.8666453528755266,
+      "learning_rate": 1.1999004629146775e-05,
+      "loss": 0.8562,
+      "step": 2355
+    },
+    {
+      "epoch": 0.4532512504809542,
+      "grad_norm": 0.8354221475217387,
+      "learning_rate": 1.1992899157394133e-05,
+      "loss": 0.7974,
+      "step": 2356
+    },
+    {
+      "epoch": 0.4534436321662178,
+      "grad_norm": 0.8645206730831727,
+      "learning_rate": 1.1986792911929418e-05,
+      "loss": 0.7996,
+      "step": 2357
+    },
+    {
+      "epoch": 0.45363601385148133,
+      "grad_norm": 0.922279797749516,
+      "learning_rate": 1.198068589512329e-05,
+      "loss": 0.7933,
+      "step": 2358
+    },
+    {
+      "epoch": 0.4538283955367449,
+      "grad_norm": 0.8632149103794537,
+      "learning_rate": 1.1974578109346702e-05,
+      "loss": 0.322,
+      "step": 2359
+    },
+    {
+      "epoch": 0.45402077722200845,
+      "grad_norm": 1.0022011897799352,
+      "learning_rate": 1.1968469556970905e-05,
+      "loss": 0.9096,
+      "step": 2360
+    },
+    {
+      "epoch": 0.45421315890727204,
+      "grad_norm": 0.8798335997004467,
+      "learning_rate": 1.1962360240367445e-05,
+      "loss": 0.8506,
+      "step": 2361
+    },
+    {
+      "epoch": 0.4544055405925356,
+      "grad_norm": 0.9678196648813475,
+      "learning_rate": 1.1956250161908179e-05,
+      "loss": 0.7641,
+      "step": 2362
+    },
+    {
+      "epoch": 0.45459792227779916,
+      "grad_norm": 1.0085352223934576,
+      "learning_rate": 1.195013932396524e-05,
+      "loss": 0.832,
+      "step": 2363
+    },
+    {
+      "epoch": 0.4547903039630627,
+      "grad_norm": 0.8804918789724577,
+      "learning_rate": 1.1944027728911072e-05,
+      "loss": 0.8492,
+      "step": 2364
+    },
+    {
+      "epoch": 0.4549826856483263,
+      "grad_norm": 0.9777434154524499,
+      "learning_rate": 1.1937915379118406e-05,
+      "loss": 0.9003,
+      "step": 2365
+    },
+    {
+      "epoch": 0.45517506733358987,
+      "grad_norm": 0.8825668262691795,
+      "learning_rate": 1.1931802276960265e-05,
+      "loss": 0.8531,
+      "step": 2366
+    },
+    {
+      "epoch": 0.4553674490188534,
+      "grad_norm": 0.8946607522528334,
+      "learning_rate": 1.1925688424809965e-05,
+      "loss": 0.8711,
+      "step": 2367
+    },
+    {
+      "epoch": 0.455559830704117,
+      "grad_norm": 0.8645685850176394,
+      "learning_rate": 1.1919573825041115e-05,
+      "loss": 0.8395,
+      "step": 2368
+    },
+    {
+      "epoch": 0.4557522123893805,
+      "grad_norm": 0.8026106056501281,
+      "learning_rate": 1.1913458480027614e-05,
+      "loss": 0.8984,
+      "step": 2369
+    },
+    {
+      "epoch": 0.4559445940746441,
+      "grad_norm": 0.892077316779484,
+      "learning_rate": 1.1907342392143646e-05,
+      "loss": 0.8067,
+      "step": 2370
+    },
+    {
+      "epoch": 0.45613697575990764,
+      "grad_norm": 0.8768349946490408,
+      "learning_rate": 1.1901225563763694e-05,
+      "loss": 0.769,
+      "step": 2371
+    },
+    {
+      "epoch": 0.4563293574451712,
+      "grad_norm": 0.7792500151384875,
+      "learning_rate": 1.1895107997262516e-05,
+      "loss": 0.795,
+      "step": 2372
+    },
+    {
+      "epoch": 0.45652173913043476,
+      "grad_norm": 0.878425344430481,
+      "learning_rate": 1.1888989695015166e-05,
+      "loss": 0.8174,
+      "step": 2373
+    },
+    {
+      "epoch": 0.45671412081569834,
+      "grad_norm": 0.8772462331222917,
+      "learning_rate": 1.1882870659396968e-05,
+      "loss": 0.8338,
+      "step": 2374
+    },
+    {
+      "epoch": 0.45690650250096193,
+      "grad_norm": 0.915909989496588,
+      "learning_rate": 1.1876750892783558e-05,
+      "loss": 0.8754,
+      "step": 2375
+    },
+    {
+      "epoch": 0.45709888418622546,
+      "grad_norm": 0.9226459216996034,
+      "learning_rate": 1.1870630397550831e-05,
+      "loss": 0.872,
+      "step": 2376
+    },
+    {
+      "epoch": 0.45729126587148905,
+      "grad_norm": 0.9151933313322156,
+      "learning_rate": 1.1864509176074974e-05,
+      "loss": 0.8725,
+      "step": 2377
+    },
+    {
+      "epoch": 0.4574836475567526,
+      "grad_norm": 0.8320610351875081,
+      "learning_rate": 1.185838723073246e-05,
+      "loss": 0.8881,
+      "step": 2378
+    },
+    {
+      "epoch": 0.45767602924201617,
+      "grad_norm": 0.8614350852018031,
+      "learning_rate": 1.1852264563900038e-05,
+      "loss": 0.8328,
+      "step": 2379
+    },
+    {
+      "epoch": 0.4578684109272797,
+      "grad_norm": 0.8676431401424447,
+      "learning_rate": 1.1846141177954733e-05,
+      "loss": 0.8435,
+      "step": 2380
+    },
+    {
+      "epoch": 0.4580607926125433,
+      "grad_norm": 0.8645078051156099,
+      "learning_rate": 1.1840017075273861e-05,
+      "loss": 0.8609,
+      "step": 2381
+    },
+    {
+      "epoch": 0.4582531742978069,
+      "grad_norm": 0.8639766976015363,
+      "learning_rate": 1.1833892258235008e-05,
+      "loss": 0.8086,
+      "step": 2382
+    },
+    {
+      "epoch": 0.4584455559830704,
+      "grad_norm": 0.9024492230869008,
+      "learning_rate": 1.1827766729216035e-05,
+      "loss": 0.869,
+      "step": 2383
+    },
+    {
+      "epoch": 0.458637937668334,
+      "grad_norm": 0.8868192431574006,
+      "learning_rate": 1.1821640490595086e-05,
+      "loss": 0.8039,
+      "step": 2384
+    },
+    {
+      "epoch": 0.4588303193535975,
+      "grad_norm": 0.8381719899676017,
+      "learning_rate": 1.181551354475058e-05,
+      "loss": 0.7891,
+      "step": 2385
+    },
+    {
+      "epoch": 0.4590227010388611,
+      "grad_norm": 0.9617360878459909,
+      "learning_rate": 1.1809385894061206e-05,
+      "loss": 0.8799,
+      "step": 2386
+    },
+    {
+      "epoch": 0.45921508272412465,
+      "grad_norm": 0.9564324165760316,
+      "learning_rate": 1.1803257540905926e-05,
+      "loss": 0.8797,
+      "step": 2387
+    },
+    {
+      "epoch": 0.45940746440938823,
+      "grad_norm": 0.8715317694551533,
+      "learning_rate": 1.1797128487663982e-05,
+      "loss": 0.8133,
+      "step": 2388
+    },
+    {
+      "epoch": 0.45959984609465176,
+      "grad_norm": 0.8583777984877526,
+      "learning_rate": 1.1790998736714882e-05,
+      "loss": 0.8092,
+      "step": 2389
+    },
+    {
+      "epoch": 0.45979222777991535,
+      "grad_norm": 0.914145431394924,
+      "learning_rate": 1.1784868290438404e-05,
+      "loss": 0.8181,
+      "step": 2390
+    },
+    {
+      "epoch": 0.45998460946517894,
+      "grad_norm": 0.8723136672943416,
+      "learning_rate": 1.1778737151214606e-05,
+      "loss": 0.7925,
+      "step": 2391
+    },
+    {
+      "epoch": 0.46017699115044247,
+      "grad_norm": 0.8720889975310768,
+      "learning_rate": 1.17726053214238e-05,
+      "loss": 0.8074,
+      "step": 2392
+    },
+    {
+      "epoch": 0.46036937283570606,
+      "grad_norm": 0.8229551560010447,
+      "learning_rate": 1.1766472803446577e-05,
+      "loss": 0.837,
+      "step": 2393
+    },
+    {
+      "epoch": 0.4605617545209696,
+      "grad_norm": 0.9180783282080991,
+      "learning_rate": 1.1760339599663788e-05,
+      "loss": 0.824,
+      "step": 2394
+    },
+    {
+      "epoch": 0.4607541362062332,
+      "grad_norm": 0.8853328175072497,
+      "learning_rate": 1.1754205712456556e-05,
+      "loss": 0.8396,
+      "step": 2395
+    },
+    {
+      "epoch": 0.4609465178914967,
+      "grad_norm": 1.021207748892328,
+      "learning_rate": 1.1748071144206266e-05,
+      "loss": 0.8047,
+      "step": 2396
+    },
+    {
+      "epoch": 0.4611388995767603,
+      "grad_norm": 0.715988701295286,
+      "learning_rate": 1.1741935897294572e-05,
+      "loss": 0.339,
+      "step": 2397
+    },
+    {
+      "epoch": 0.46133128126202383,
+      "grad_norm": 0.9079546332968439,
+      "learning_rate": 1.1735799974103388e-05,
+      "loss": 0.8595,
+      "step": 2398
+    },
+    {
+      "epoch": 0.4615236629472874,
+      "grad_norm": 0.8273487194248393,
+      "learning_rate": 1.1729663377014888e-05,
+      "loss": 0.7646,
+      "step": 2399
+    },
+    {
+      "epoch": 0.461716044632551,
+      "grad_norm": 0.8743500437099636,
+      "learning_rate": 1.172352610841151e-05,
+      "loss": 0.8374,
+      "step": 2400
+    },
+    {
+      "epoch": 0.46190842631781454,
+      "grad_norm": 0.878530944600156,
+      "learning_rate": 1.1717388170675954e-05,
+      "loss": 0.8267,
+      "step": 2401
+    },
+    {
+      "epoch": 0.4621008080030781,
+      "grad_norm": 0.8204453036997391,
+      "learning_rate": 1.1711249566191179e-05,
+      "loss": 0.8335,
+      "step": 2402
+    },
+    {
+      "epoch": 0.46229318968834165,
+      "grad_norm": 0.8879059609333461,
+      "learning_rate": 1.17051102973404e-05,
+      "loss": 0.7967,
+      "step": 2403
+    },
+    {
+      "epoch": 0.46248557137360524,
+      "grad_norm": 0.8222885940016378,
+      "learning_rate": 1.1698970366507096e-05,
+      "loss": 0.8539,
+      "step": 2404
+    },
+    {
+      "epoch": 0.4626779530588688,
+      "grad_norm": 0.8607417069230987,
+      "learning_rate": 1.1692829776074999e-05,
+      "loss": 0.8167,
+      "step": 2405
+    },
+    {
+      "epoch": 0.46287033474413236,
+      "grad_norm": 0.8388598023774019,
+      "learning_rate": 1.1686688528428099e-05,
+      "loss": 0.8382,
+      "step": 2406
+    },
+    {
+      "epoch": 0.46306271642939595,
+      "grad_norm": 0.8827227217260947,
+      "learning_rate": 1.1680546625950635e-05,
+      "loss": 0.8873,
+      "step": 2407
+    },
+    {
+      "epoch": 0.4632550981146595,
+      "grad_norm": 0.8191701753312177,
+      "learning_rate": 1.167440407102711e-05,
+      "loss": 0.8084,
+      "step": 2408
+    },
+    {
+      "epoch": 0.46344747979992307,
+      "grad_norm": 0.8303832058142013,
+      "learning_rate": 1.1668260866042271e-05,
+      "loss": 0.7601,
+      "step": 2409
+    },
+    {
+      "epoch": 0.4636398614851866,
+      "grad_norm": 0.8889841362779751,
+      "learning_rate": 1.1662117013381126e-05,
+      "loss": 0.8484,
+      "step": 2410
+    },
+    {
+      "epoch": 0.4638322431704502,
+      "grad_norm": 0.9249811832028727,
+      "learning_rate": 1.1655972515428928e-05,
+      "loss": 0.8712,
+      "step": 2411
+    },
+    {
+      "epoch": 0.4640246248557137,
+      "grad_norm": 0.9570612581354815,
+      "learning_rate": 1.1649827374571182e-05,
+      "loss": 0.8101,
+      "step": 2412
+    },
+    {
+      "epoch": 0.4642170065409773,
+      "grad_norm": 0.8991588333425649,
+      "learning_rate": 1.1643681593193642e-05,
+      "loss": 0.8195,
+      "step": 2413
+    },
+    {
+      "epoch": 0.46440938822624084,
+      "grad_norm": 0.8733454017730965,
+      "learning_rate": 1.1637535173682318e-05,
+      "loss": 0.8605,
+      "step": 2414
+    },
+    {
+      "epoch": 0.4646017699115044,
+      "grad_norm": 0.6487536492493559,
+      "learning_rate": 1.1631388118423457e-05,
+      "loss": 0.3374,
+      "step": 2415
+    },
+    {
+      "epoch": 0.464794151596768,
+      "grad_norm": 0.6277933310165678,
+      "learning_rate": 1.1625240429803553e-05,
+      "loss": 0.3711,
+      "step": 2416
+    },
+    {
+      "epoch": 0.46498653328203154,
+      "grad_norm": 0.8908409918205057,
+      "learning_rate": 1.1619092110209361e-05,
+      "loss": 0.8728,
+      "step": 2417
+    },
+    {
+      "epoch": 0.46517891496729513,
+      "grad_norm": 0.9784965378548123,
+      "learning_rate": 1.1612943162027863e-05,
+      "loss": 0.7934,
+      "step": 2418
+    },
+    {
+      "epoch": 0.46537129665255866,
+      "grad_norm": 0.9720099588861187,
+      "learning_rate": 1.1606793587646295e-05,
+      "loss": 0.8567,
+      "step": 2419
+    },
+    {
+      "epoch": 0.46556367833782225,
+      "grad_norm": 0.9430431070604413,
+      "learning_rate": 1.160064338945213e-05,
+      "loss": 0.876,
+      "step": 2420
+    },
+    {
+      "epoch": 0.4657560600230858,
+      "grad_norm": 0.5987691504509225,
+      "learning_rate": 1.1594492569833093e-05,
+      "loss": 0.3553,
+      "step": 2421
+    },
+    {
+      "epoch": 0.46594844170834937,
+      "grad_norm": 0.9040072617378982,
+      "learning_rate": 1.1588341131177137e-05,
+      "loss": 0.8607,
+      "step": 2422
+    },
+    {
+      "epoch": 0.4661408233936129,
+      "grad_norm": 0.8794574690932538,
+      "learning_rate": 1.1582189075872467e-05,
+      "loss": 0.8238,
+      "step": 2423
+    },
+    {
+      "epoch": 0.4663332050788765,
+      "grad_norm": 0.8508166705142206,
+      "learning_rate": 1.1576036406307523e-05,
+      "loss": 0.8336,
+      "step": 2424
+    },
+    {
+      "epoch": 0.4665255867641401,
+      "grad_norm": 0.8518934637112546,
+      "learning_rate": 1.156988312487098e-05,
+      "loss": 0.88,
+      "step": 2425
+    },
+    {
+      "epoch": 0.4667179684494036,
+      "grad_norm": 0.9003411950175114,
+      "learning_rate": 1.1563729233951757e-05,
+      "loss": 0.8326,
+      "step": 2426
+    },
+    {
+      "epoch": 0.4669103501346672,
+      "grad_norm": 0.842029467126434,
+      "learning_rate": 1.1557574735939003e-05,
+      "loss": 0.8146,
+      "step": 2427
+    },
+    {
+      "epoch": 0.4671027318199307,
+      "grad_norm": 0.8498595171204176,
+      "learning_rate": 1.1551419633222107e-05,
+      "loss": 0.8309,
+      "step": 2428
+    },
+    {
+      "epoch": 0.4672951135051943,
+      "grad_norm": 0.8666191228077967,
+      "learning_rate": 1.1545263928190692e-05,
+      "loss": 0.8495,
+      "step": 2429
+    },
+    {
+      "epoch": 0.46748749519045785,
+      "grad_norm": 0.8815432510113842,
+      "learning_rate": 1.1539107623234618e-05,
+      "loss": 0.8568,
+      "step": 2430
+    },
+    {
+      "epoch": 0.46767987687572143,
+      "grad_norm": 0.8625987851957607,
+      "learning_rate": 1.153295072074397e-05,
+      "loss": 0.8287,
+      "step": 2431
+    },
+    {
+      "epoch": 0.467872258560985,
+      "grad_norm": 0.9452725168191192,
+      "learning_rate": 1.1526793223109072e-05,
+      "loss": 0.8001,
+      "step": 2432
+    },
+    {
+      "epoch": 0.46806464024624855,
+      "grad_norm": 0.8960641483971498,
+      "learning_rate": 1.1520635132720475e-05,
+      "loss": 0.8504,
+      "step": 2433
+    },
+    {
+      "epoch": 0.46825702193151214,
+      "grad_norm": 0.8516763689078983,
+      "learning_rate": 1.1514476451968961e-05,
+      "loss": 0.7959,
+      "step": 2434
+    },
+    {
+      "epoch": 0.4684494036167757,
+      "grad_norm": 0.9040130045302802,
+      "learning_rate": 1.1508317183245545e-05,
+      "loss": 0.8089,
+      "step": 2435
+    },
+    {
+      "epoch": 0.46864178530203926,
+      "grad_norm": 0.8988876554009692,
+      "learning_rate": 1.1502157328941466e-05,
+      "loss": 0.8662,
+      "step": 2436
+    },
+    {
+      "epoch": 0.4688341669873028,
+      "grad_norm": 0.8735117995579051,
+      "learning_rate": 1.149599689144819e-05,
+      "loss": 0.8835,
+      "step": 2437
+    },
+    {
+      "epoch": 0.4690265486725664,
+      "grad_norm": 0.8218346102137168,
+      "learning_rate": 1.1489835873157414e-05,
+      "loss": 0.8289,
+      "step": 2438
+    },
+    {
+      "epoch": 0.4692189303578299,
+      "grad_norm": 0.9022400199195932,
+      "learning_rate": 1.1483674276461053e-05,
+      "loss": 0.8586,
+      "step": 2439
+    },
+    {
+      "epoch": 0.4694113120430935,
+      "grad_norm": 0.9065444602969533,
+      "learning_rate": 1.1477512103751254e-05,
+      "loss": 0.8394,
+      "step": 2440
+    },
+    {
+      "epoch": 0.4696036937283571,
+      "grad_norm": 1.0091843107136513,
+      "learning_rate": 1.1471349357420384e-05,
+      "loss": 0.8137,
+      "step": 2441
+    },
+    {
+      "epoch": 0.4697960754136206,
+      "grad_norm": 0.8637085883782307,
+      "learning_rate": 1.1465186039861033e-05,
+      "loss": 0.8559,
+      "step": 2442
+    },
+    {
+      "epoch": 0.4699884570988842,
+      "grad_norm": 1.023074671713285,
+      "learning_rate": 1.1459022153466016e-05,
+      "loss": 0.8502,
+      "step": 2443
+    },
+    {
+      "epoch": 0.47018083878414774,
+      "grad_norm": 0.897380914257127,
+      "learning_rate": 1.1452857700628362e-05,
+      "loss": 0.7983,
+      "step": 2444
+    },
+    {
+      "epoch": 0.4703732204694113,
+      "grad_norm": 0.9108833952962292,
+      "learning_rate": 1.1446692683741326e-05,
+      "loss": 0.8561,
+      "step": 2445
+    },
+    {
+      "epoch": 0.47056560215467486,
+      "grad_norm": 0.8475633874538261,
+      "learning_rate": 1.1440527105198377e-05,
+      "loss": 0.7863,
+      "step": 2446
+    },
+    {
+      "epoch": 0.47075798383993844,
+      "grad_norm": 0.9086677561445438,
+      "learning_rate": 1.143436096739321e-05,
+      "loss": 0.8428,
+      "step": 2447
+    },
+    {
+      "epoch": 0.470950365525202,
+      "grad_norm": 0.9285538038071398,
+      "learning_rate": 1.1428194272719729e-05,
+      "loss": 0.8744,
+      "step": 2448
+    },
+    {
+      "epoch": 0.47114274721046556,
+      "grad_norm": 0.8431666044948113,
+      "learning_rate": 1.1422027023572052e-05,
+      "loss": 0.8354,
+      "step": 2449
+    },
+    {
+      "epoch": 0.47133512889572915,
+      "grad_norm": 0.8704442860239846,
+      "learning_rate": 1.1415859222344525e-05,
+      "loss": 0.8424,
+      "step": 2450
+    },
+    {
+      "epoch": 0.4715275105809927,
+      "grad_norm": 0.8585200858830194,
+      "learning_rate": 1.14096908714317e-05,
+      "loss": 0.8127,
+      "step": 2451
+    },
+    {
+      "epoch": 0.47171989226625627,
+      "grad_norm": 0.8732790878622554,
+      "learning_rate": 1.1403521973228342e-05,
+      "loss": 0.8707,
+      "step": 2452
+    },
+    {
+      "epoch": 0.4719122739515198,
+      "grad_norm": 0.916403568860353,
+      "learning_rate": 1.1397352530129428e-05,
+      "loss": 0.8654,
+      "step": 2453
+    },
+    {
+      "epoch": 0.4721046556367834,
+      "grad_norm": 0.8258991961142171,
+      "learning_rate": 1.139118254453015e-05,
+      "loss": 0.8019,
+      "step": 2454
+    },
+    {
+      "epoch": 0.4722970373220469,
+      "grad_norm": 0.7069985551080531,
+      "learning_rate": 1.1385012018825907e-05,
+      "loss": 0.342,
+      "step": 2455
+    },
+    {
+      "epoch": 0.4724894190073105,
+      "grad_norm": 0.8467753459691201,
+      "learning_rate": 1.1378840955412313e-05,
+      "loss": 0.8668,
+      "step": 2456
+    },
+    {
+      "epoch": 0.4726818006925741,
+      "grad_norm": 0.9557216413849312,
+      "learning_rate": 1.1372669356685185e-05,
+      "loss": 0.8261,
+      "step": 2457
+    },
+    {
+      "epoch": 0.4728741823778376,
+      "grad_norm": 0.9637655731307316,
+      "learning_rate": 1.1366497225040549e-05,
+      "loss": 0.8199,
+      "step": 2458
+    },
+    {
+      "epoch": 0.4730665640631012,
+      "grad_norm": 0.9166460305995202,
+      "learning_rate": 1.1360324562874643e-05,
+      "loss": 0.8566,
+      "step": 2459
+    },
+    {
+      "epoch": 0.47325894574836475,
+      "grad_norm": 0.8415759393776855,
+      "learning_rate": 1.1354151372583901e-05,
+      "loss": 0.8299,
+      "step": 2460
+    },
+    {
+      "epoch": 0.47345132743362833,
+      "grad_norm": 0.8862305477390658,
+      "learning_rate": 1.1347977656564974e-05,
+      "loss": 0.8185,
+      "step": 2461
+    },
+    {
+      "epoch": 0.47364370911889186,
+      "grad_norm": 0.9285310187458734,
+      "learning_rate": 1.1341803417214705e-05,
+      "loss": 0.8291,
+      "step": 2462
+    },
+    {
+      "epoch": 0.47383609080415545,
+      "grad_norm": 0.9773621702381619,
+      "learning_rate": 1.1335628656930153e-05,
+      "loss": 0.7989,
+      "step": 2463
+    },
+    {
+      "epoch": 0.474028472489419,
+      "grad_norm": 0.8415705876281806,
+      "learning_rate": 1.132945337810857e-05,
+      "loss": 0.81,
+      "step": 2464
+    },
+    {
+      "epoch": 0.47422085417468257,
+      "grad_norm": 0.8977327613146072,
+      "learning_rate": 1.132327758314741e-05,
+      "loss": 0.8791,
+      "step": 2465
+    },
+    {
+      "epoch": 0.47441323585994616,
+      "grad_norm": 0.8647495468935528,
+      "learning_rate": 1.131710127444433e-05,
+      "loss": 0.8189,
+      "step": 2466
+    },
+    {
+      "epoch": 0.4746056175452097,
+      "grad_norm": 0.8500352129227557,
+      "learning_rate": 1.1310924454397187e-05,
+      "loss": 0.825,
+      "step": 2467
+    },
+    {
+      "epoch": 0.4747979992304733,
+      "grad_norm": 0.941791221378926,
+      "learning_rate": 1.1304747125404031e-05,
+      "loss": 0.8791,
+      "step": 2468
+    },
+    {
+      "epoch": 0.4749903809157368,
+      "grad_norm": 0.8322862588051256,
+      "learning_rate": 1.129856928986312e-05,
+      "loss": 0.7703,
+      "step": 2469
+    },
+    {
+      "epoch": 0.4751827626010004,
+      "grad_norm": 0.9816228954066327,
+      "learning_rate": 1.12923909501729e-05,
+      "loss": 0.8185,
+      "step": 2470
+    },
+    {
+      "epoch": 0.47537514428626393,
+      "grad_norm": 0.8099130416868081,
+      "learning_rate": 1.1286212108732015e-05,
+      "loss": 0.7765,
+      "step": 2471
+    },
+    {
+      "epoch": 0.4755675259715275,
+      "grad_norm": 0.9025963239423589,
+      "learning_rate": 1.1280032767939302e-05,
+      "loss": 0.7989,
+      "step": 2472
+    },
+    {
+      "epoch": 0.47575990765679105,
+      "grad_norm": 0.9441917747807949,
+      "learning_rate": 1.1273852930193798e-05,
+      "loss": 0.8788,
+      "step": 2473
+    },
+    {
+      "epoch": 0.47595228934205464,
+      "grad_norm": 0.842499401131679,
+      "learning_rate": 1.1267672597894725e-05,
+      "loss": 0.6911,
+      "step": 2474
+    },
+    {
+      "epoch": 0.4761446710273182,
+      "grad_norm": 0.7921111816313595,
+      "learning_rate": 1.12614917734415e-05,
+      "loss": 0.8104,
+      "step": 2475
+    },
+    {
+      "epoch": 0.47633705271258175,
+      "grad_norm": 0.896221079926793,
+      "learning_rate": 1.1255310459233737e-05,
+      "loss": 0.9049,
+      "step": 2476
+    },
+    {
+      "epoch": 0.47652943439784534,
+      "grad_norm": 0.8103071595440834,
+      "learning_rate": 1.1249128657671233e-05,
+      "loss": 0.838,
+      "step": 2477
+    },
+    {
+      "epoch": 0.4767218160831089,
+      "grad_norm": 0.9486808063181051,
+      "learning_rate": 1.1242946371153974e-05,
+      "loss": 0.8597,
+      "step": 2478
+    },
+    {
+      "epoch": 0.47691419776837246,
+      "grad_norm": 0.9085500991561115,
+      "learning_rate": 1.1236763602082136e-05,
+      "loss": 0.8302,
+      "step": 2479
+    },
+    {
+      "epoch": 0.477106579453636,
+      "grad_norm": 0.782892018755486,
+      "learning_rate": 1.1230580352856088e-05,
+      "loss": 0.849,
+      "step": 2480
+    },
+    {
+      "epoch": 0.4772989611388996,
+      "grad_norm": 0.8515233436639673,
+      "learning_rate": 1.1224396625876375e-05,
+      "loss": 0.798,
+      "step": 2481
+    },
+    {
+      "epoch": 0.4774913428241631,
+      "grad_norm": 0.923542676929046,
+      "learning_rate": 1.1218212423543734e-05,
+      "loss": 0.8976,
+      "step": 2482
+    },
+    {
+      "epoch": 0.4776837245094267,
+      "grad_norm": 0.857618905917189,
+      "learning_rate": 1.1212027748259086e-05,
+      "loss": 0.8009,
+      "step": 2483
+    },
+    {
+      "epoch": 0.4778761061946903,
+      "grad_norm": 0.8565609941034293,
+      "learning_rate": 1.1205842602423537e-05,
+      "loss": 0.7975,
+      "step": 2484
+    },
+    {
+      "epoch": 0.4780684878799538,
+      "grad_norm": 0.8911706181518042,
+      "learning_rate": 1.1199656988438373e-05,
+      "loss": 0.8764,
+      "step": 2485
+    },
+    {
+      "epoch": 0.4782608695652174,
+      "grad_norm": 0.918519168867406,
+      "learning_rate": 1.1193470908705055e-05,
+      "loss": 0.8289,
+      "step": 2486
+    },
+    {
+      "epoch": 0.47845325125048094,
+      "grad_norm": 0.9276523385668535,
+      "learning_rate": 1.1187284365625241e-05,
+      "loss": 0.8145,
+      "step": 2487
+    },
+    {
+      "epoch": 0.4786456329357445,
+      "grad_norm": 0.8116626110057639,
+      "learning_rate": 1.1181097361600754e-05,
+      "loss": 0.7307,
+      "step": 2488
+    },
+    {
+      "epoch": 0.47883801462100806,
+      "grad_norm": 0.9746078698596525,
+      "learning_rate": 1.1174909899033608e-05,
+      "loss": 0.8729,
+      "step": 2489
+    },
+    {
+      "epoch": 0.47903039630627164,
+      "grad_norm": 0.8261612503487294,
+      "learning_rate": 1.1168721980325987e-05,
+      "loss": 0.8498,
+      "step": 2490
+    },
+    {
+      "epoch": 0.47922277799153523,
+      "grad_norm": 0.8463165265578103,
+      "learning_rate": 1.1162533607880251e-05,
+      "loss": 0.8579,
+      "step": 2491
+    },
+    {
+      "epoch": 0.47941515967679876,
+      "grad_norm": 0.9043129067305639,
+      "learning_rate": 1.1156344784098942e-05,
+      "loss": 0.8121,
+      "step": 2492
+    },
+    {
+      "epoch": 0.47960754136206235,
+      "grad_norm": 0.8400317178403267,
+      "learning_rate": 1.1150155511384772e-05,
+      "loss": 0.8645,
+      "step": 2493
+    },
+    {
+      "epoch": 0.4797999230473259,
+      "grad_norm": 0.9130437658038021,
+      "learning_rate": 1.1143965792140631e-05,
+      "loss": 0.8868,
+      "step": 2494
+    },
+    {
+      "epoch": 0.47999230473258947,
+      "grad_norm": 0.8856362242806035,
+      "learning_rate": 1.1137775628769584e-05,
+      "loss": 0.8828,
+      "step": 2495
+    },
+    {
+      "epoch": 0.480184686417853,
+      "grad_norm": 0.8823270623090583,
+      "learning_rate": 1.1131585023674863e-05,
+      "loss": 0.7892,
+      "step": 2496
+    },
+    {
+      "epoch": 0.4803770681031166,
+      "grad_norm": 0.8348978783546626,
+      "learning_rate": 1.1125393979259874e-05,
+      "loss": 0.8228,
+      "step": 2497
+    },
+    {
+      "epoch": 0.4805694497883801,
+      "grad_norm": 0.881266784546975,
+      "learning_rate": 1.1119202497928192e-05,
+      "loss": 0.8727,
+      "step": 2498
+    },
+    {
+      "epoch": 0.4807618314736437,
+      "grad_norm": 0.8501801394763433,
+      "learning_rate": 1.1113010582083568e-05,
+      "loss": 0.8444,
+      "step": 2499
+    },
+    {
+      "epoch": 0.4809542131589073,
+      "grad_norm": 0.8449973464032419,
+      "learning_rate": 1.1106818234129913e-05,
+      "loss": 0.842,
+      "step": 2500
+    },
+    {
+      "epoch": 0.4811465948441708,
+      "grad_norm": 0.9414666025020358,
+      "learning_rate": 1.1100625456471307e-05,
+      "loss": 0.8086,
+      "step": 2501
+    },
+    {
+      "epoch": 0.4813389765294344,
+      "grad_norm": 0.8619454340691937,
+      "learning_rate": 1.1094432251512006e-05,
+      "loss": 0.8142,
+      "step": 2502
+    },
+    {
+      "epoch": 0.48153135821469795,
+      "grad_norm": 0.9315388250142821,
+      "learning_rate": 1.1088238621656422e-05,
+      "loss": 0.8778,
+      "step": 2503
+    },
+    {
+      "epoch": 0.48172373989996153,
+      "grad_norm": 0.8788498333521515,
+      "learning_rate": 1.1082044569309138e-05,
+      "loss": 0.8622,
+      "step": 2504
+    },
+    {
+      "epoch": 0.48191612158522507,
+      "grad_norm": 0.7264272459910095,
+      "learning_rate": 1.1075850096874894e-05,
+      "loss": 0.3421,
+      "step": 2505
+    },
+    {
+      "epoch": 0.48210850327048865,
+      "grad_norm": 0.8253368207888183,
+      "learning_rate": 1.1069655206758603e-05,
+      "loss": 0.8695,
+      "step": 2506
+    },
+    {
+      "epoch": 0.4823008849557522,
+      "grad_norm": 0.9365241167014862,
+      "learning_rate": 1.1063459901365325e-05,
+      "loss": 0.8243,
+      "step": 2507
+    },
+    {
+      "epoch": 0.48249326664101577,
+      "grad_norm": 0.8235812479718633,
+      "learning_rate": 1.1057264183100303e-05,
+      "loss": 0.8039,
+      "step": 2508
+    },
+    {
+      "epoch": 0.48268564832627936,
+      "grad_norm": 0.7936155041230321,
+      "learning_rate": 1.1051068054368921e-05,
+      "loss": 0.819,
+      "step": 2509
+    },
+    {
+      "epoch": 0.4828780300115429,
+      "grad_norm": 0.912117041297376,
+      "learning_rate": 1.104487151757673e-05,
+      "loss": 0.7748,
+      "step": 2510
+    },
+    {
+      "epoch": 0.4830704116968065,
+      "grad_norm": 0.973296652801658,
+      "learning_rate": 1.1038674575129442e-05,
+      "loss": 0.84,
+      "step": 2511
+    },
+    {
+      "epoch": 0.48326279338207,
+      "grad_norm": 0.6334273159513958,
+      "learning_rate": 1.1032477229432921e-05,
+      "loss": 0.3573,
+      "step": 2512
+    },
+    {
+      "epoch": 0.4834551750673336,
+      "grad_norm": 0.8857931492042463,
+      "learning_rate": 1.1026279482893187e-05,
+      "loss": 0.8436,
+      "step": 2513
+    },
+    {
+      "epoch": 0.48364755675259713,
+      "grad_norm": 0.6373634122808391,
+      "learning_rate": 1.1020081337916425e-05,
+      "loss": 0.365,
+      "step": 2514
+    },
+    {
+      "epoch": 0.4838399384378607,
+      "grad_norm": 0.8837182537651718,
+      "learning_rate": 1.1013882796908963e-05,
+      "loss": 0.7934,
+      "step": 2515
+    },
+    {
+      "epoch": 0.4840323201231243,
+      "grad_norm": 0.8411378025379436,
+      "learning_rate": 1.1007683862277292e-05,
+      "loss": 0.807,
+      "step": 2516
+    },
+    {
+      "epoch": 0.48422470180838784,
+      "grad_norm": 0.8882713163404408,
+      "learning_rate": 1.1001484536428052e-05,
+      "loss": 0.8488,
+      "step": 2517
+    },
+    {
+      "epoch": 0.4844170834936514,
+      "grad_norm": 0.8621348920805133,
+      "learning_rate": 1.0995284821768029e-05,
+      "loss": 0.8424,
+      "step": 2518
+    },
+    {
+      "epoch": 0.48460946517891496,
+      "grad_norm": 0.92105741069832,
+      "learning_rate": 1.098908472070417e-05,
+      "loss": 0.8603,
+      "step": 2519
+    },
+    {
+      "epoch": 0.48480184686417854,
+      "grad_norm": 0.8980164635076009,
+      "learning_rate": 1.0982884235643567e-05,
+      "loss": 0.8616,
+      "step": 2520
+    },
+    {
+      "epoch": 0.4849942285494421,
+      "grad_norm": 0.9015836918751803,
+      "learning_rate": 1.0976683368993464e-05,
+      "loss": 0.9239,
+      "step": 2521
+    },
+    {
+      "epoch": 0.48518661023470566,
+      "grad_norm": 0.8521799111984827,
+      "learning_rate": 1.0970482123161249e-05,
+      "loss": 0.8502,
+      "step": 2522
+    },
+    {
+      "epoch": 0.4853789919199692,
+      "grad_norm": 0.9541217151431299,
+      "learning_rate": 1.0964280500554459e-05,
+      "loss": 0.8802,
+      "step": 2523
+    },
+    {
+      "epoch": 0.4855713736052328,
+      "grad_norm": 0.8437885532699209,
+      "learning_rate": 1.0958078503580776e-05,
+      "loss": 0.8219,
+      "step": 2524
+    },
+    {
+      "epoch": 0.48576375529049637,
+      "grad_norm": 0.9206257561171342,
+      "learning_rate": 1.0951876134648032e-05,
+      "loss": 0.8272,
+      "step": 2525
+    },
+    {
+      "epoch": 0.4859561369757599,
+      "grad_norm": 0.9218558106830181,
+      "learning_rate": 1.0945673396164198e-05,
+      "loss": 0.8548,
+      "step": 2526
+    },
+    {
+      "epoch": 0.4861485186610235,
+      "grad_norm": 0.8506413776230586,
+      "learning_rate": 1.0939470290537389e-05,
+      "loss": 0.8003,
+      "step": 2527
+    },
+    {
+      "epoch": 0.486340900346287,
+      "grad_norm": 0.6247825297149427,
+      "learning_rate": 1.0933266820175868e-05,
+      "loss": 0.3451,
+      "step": 2528
+    },
+    {
+      "epoch": 0.4865332820315506,
+      "grad_norm": 0.8181012943088564,
+      "learning_rate": 1.0927062987488035e-05,
+      "loss": 0.8619,
+      "step": 2529
+    },
+    {
+      "epoch": 0.48672566371681414,
+      "grad_norm": 0.8285413914544218,
+      "learning_rate": 1.0920858794882429e-05,
+      "loss": 0.8645,
+      "step": 2530
+    },
+    {
+      "epoch": 0.4869180454020777,
+      "grad_norm": 0.6246529210484331,
+      "learning_rate": 1.0914654244767736e-05,
+      "loss": 0.3181,
+      "step": 2531
+    },
+    {
+      "epoch": 0.48711042708734126,
+      "grad_norm": 0.8943080443997915,
+      "learning_rate": 1.0908449339552769e-05,
+      "loss": 0.7875,
+      "step": 2532
+    },
+    {
+      "epoch": 0.48730280877260485,
+      "grad_norm": 0.8999061599117342,
+      "learning_rate": 1.0902244081646489e-05,
+      "loss": 0.7745,
+      "step": 2533
+    },
+    {
+      "epoch": 0.48749519045786843,
+      "grad_norm": 0.823228084989382,
+      "learning_rate": 1.0896038473457993e-05,
+      "loss": 0.8336,
+      "step": 2534
+    },
+    {
+      "epoch": 0.48768757214313196,
+      "grad_norm": 0.8635435571008842,
+      "learning_rate": 1.0889832517396511e-05,
+      "loss": 0.8619,
+      "step": 2535
+    },
+    {
+      "epoch": 0.48787995382839555,
+      "grad_norm": 0.8727937889859324,
+      "learning_rate": 1.0883626215871408e-05,
+      "loss": 0.8489,
+      "step": 2536
+    },
+    {
+      "epoch": 0.4880723355136591,
+      "grad_norm": 0.9452613028231284,
+      "learning_rate": 1.0877419571292183e-05,
+      "loss": 0.8751,
+      "step": 2537
+    },
+    {
+      "epoch": 0.48826471719892267,
+      "grad_norm": 0.8870610261738784,
+      "learning_rate": 1.0871212586068469e-05,
+      "loss": 0.8218,
+      "step": 2538
+    },
+    {
+      "epoch": 0.4884570988841862,
+      "grad_norm": 0.7685402637319125,
+      "learning_rate": 1.0865005262610033e-05,
+      "loss": 0.8214,
+      "step": 2539
+    },
+    {
+      "epoch": 0.4886494805694498,
+      "grad_norm": 0.8964178687360336,
+      "learning_rate": 1.085879760332677e-05,
+      "loss": 0.8165,
+      "step": 2540
+    },
+    {
+      "epoch": 0.4888418622547134,
+      "grad_norm": 0.8259490094258694,
+      "learning_rate": 1.085258961062871e-05,
+      "loss": 0.802,
+      "step": 2541
+    },
+    {
+      "epoch": 0.4890342439399769,
+      "grad_norm": 0.8861720616344434,
+      "learning_rate": 1.0846381286926007e-05,
+      "loss": 0.8269,
+      "step": 2542
+    },
+    {
+      "epoch": 0.4892266256252405,
+      "grad_norm": 0.8912931380192163,
+      "learning_rate": 1.0840172634628948e-05,
+      "loss": 0.8031,
+      "step": 2543
+    },
+    {
+      "epoch": 0.48941900731050403,
+      "grad_norm": 0.7173243785634518,
+      "learning_rate": 1.0833963656147944e-05,
+      "loss": 0.342,
+      "step": 2544
+    },
+    {
+      "epoch": 0.4896113889957676,
+      "grad_norm": 0.8711374700508764,
+      "learning_rate": 1.082775435389353e-05,
+      "loss": 0.8212,
+      "step": 2545
+    },
+    {
+      "epoch": 0.48980377068103115,
+      "grad_norm": 0.7740432451972798,
+      "learning_rate": 1.0821544730276379e-05,
+      "loss": 0.7907,
+      "step": 2546
+    },
+    {
+      "epoch": 0.48999615236629473,
+      "grad_norm": 0.8371115024392256,
+      "learning_rate": 1.0815334787707277e-05,
+      "loss": 0.8363,
+      "step": 2547
+    },
+    {
+      "epoch": 0.49018853405155827,
+      "grad_norm": 0.8307158068852631,
+      "learning_rate": 1.0809124528597138e-05,
+      "loss": 0.8351,
+      "step": 2548
+    },
+    {
+      "epoch": 0.49038091573682185,
+      "grad_norm": 0.9387200526142775,
+      "learning_rate": 1.0802913955356998e-05,
+      "loss": 0.7963,
+      "step": 2549
+    },
+    {
+      "epoch": 0.49057329742208544,
+      "grad_norm": 0.8722006361822684,
+      "learning_rate": 1.0796703070398016e-05,
+      "loss": 0.8387,
+      "step": 2550
+    },
+    {
+      "epoch": 0.490765679107349,
+      "grad_norm": 0.969374019262298,
+      "learning_rate": 1.079049187613147e-05,
+      "loss": 0.8222,
+      "step": 2551
+    },
+    {
+      "epoch": 0.49095806079261256,
+      "grad_norm": 0.8477527476318504,
+      "learning_rate": 1.0784280374968761e-05,
+      "loss": 0.8811,
+      "step": 2552
+    },
+    {
+      "epoch": 0.4911504424778761,
+      "grad_norm": 0.9383503724665705,
+      "learning_rate": 1.0778068569321403e-05,
+      "loss": 0.8294,
+      "step": 2553
+    },
+    {
+      "epoch": 0.4913428241631397,
+      "grad_norm": 0.9400924130704694,
+      "learning_rate": 1.077185646160104e-05,
+      "loss": 0.8403,
+      "step": 2554
+    },
+    {
+      "epoch": 0.4915352058484032,
+      "grad_norm": 0.9100393802683202,
+      "learning_rate": 1.0765644054219422e-05,
+      "loss": 0.7947,
+      "step": 2555
+    },
+    {
+      "epoch": 0.4917275875336668,
+      "grad_norm": 0.9483137928115372,
+      "learning_rate": 1.0759431349588421e-05,
+      "loss": 0.8724,
+      "step": 2556
+    },
+    {
+      "epoch": 0.49191996921893033,
+      "grad_norm": 0.843875896472559,
+      "learning_rate": 1.0753218350120023e-05,
+      "loss": 0.8107,
+      "step": 2557
+    },
+    {
+      "epoch": 0.4921123509041939,
+      "grad_norm": 0.8261477782146383,
+      "learning_rate": 1.0747005058226325e-05,
+      "loss": 0.8012,
+      "step": 2558
+    },
+    {
+      "epoch": 0.4923047325894575,
+      "grad_norm": 0.9119986932600129,
+      "learning_rate": 1.0740791476319543e-05,
+      "loss": 0.8294,
+      "step": 2559
+    },
+    {
+      "epoch": 0.49249711427472104,
+      "grad_norm": 0.904979905146405,
+      "learning_rate": 1.0734577606812007e-05,
+      "loss": 0.8089,
+      "step": 2560
+    },
+    {
+      "epoch": 0.4926894959599846,
+      "grad_norm": 0.8632683491927127,
+      "learning_rate": 1.0728363452116149e-05,
+      "loss": 0.8102,
+      "step": 2561
+    },
+    {
+      "epoch": 0.49288187764524816,
+      "grad_norm": 0.9042399906504426,
+      "learning_rate": 1.0722149014644523e-05,
+      "loss": 0.8129,
+      "step": 2562
+    },
+    {
+      "epoch": 0.49307425933051174,
+      "grad_norm": 0.8573248371590367,
+      "learning_rate": 1.0715934296809782e-05,
+      "loss": 0.8361,
+      "step": 2563
+    },
+    {
+      "epoch": 0.4932666410157753,
+      "grad_norm": 0.9961832396507999,
+      "learning_rate": 1.0709719301024698e-05,
+      "loss": 0.8608,
+      "step": 2564
+    },
+    {
+      "epoch": 0.49345902270103886,
+      "grad_norm": 0.9065246785633096,
+      "learning_rate": 1.0703504029702148e-05,
+      "loss": 0.8303,
+      "step": 2565
+    },
+    {
+      "epoch": 0.49365140438630245,
+      "grad_norm": 0.6444587209002096,
+      "learning_rate": 1.0697288485255107e-05,
+      "loss": 0.3432,
+      "step": 2566
+    },
+    {
+      "epoch": 0.493843786071566,
+      "grad_norm": 0.8566723951944532,
+      "learning_rate": 1.0691072670096669e-05,
+      "loss": 0.813,
+      "step": 2567
+    },
+    {
+      "epoch": 0.49403616775682957,
+      "grad_norm": 0.8906978345479568,
+      "learning_rate": 1.0684856586640026e-05,
+      "loss": 0.8222,
+      "step": 2568
+    },
+    {
+      "epoch": 0.4942285494420931,
+      "grad_norm": 0.8391216645156295,
+      "learning_rate": 1.0678640237298476e-05,
+      "loss": 0.858,
+      "step": 2569
+    },
+    {
+      "epoch": 0.4944209311273567,
+      "grad_norm": 0.8667267201354854,
+      "learning_rate": 1.0672423624485423e-05,
+      "loss": 0.811,
+      "step": 2570
+    },
+    {
+      "epoch": 0.4946133128126202,
+      "grad_norm": 0.8273327055440312,
+      "learning_rate": 1.0666206750614363e-05,
+      "loss": 0.8535,
+      "step": 2571
+    },
+    {
+      "epoch": 0.4948056944978838,
+      "grad_norm": 0.8601174112282853,
+      "learning_rate": 1.0659989618098904e-05,
+      "loss": 0.858,
+      "step": 2572
+    },
+    {
+      "epoch": 0.49499807618314734,
+      "grad_norm": 0.8927781409896841,
+      "learning_rate": 1.065377222935275e-05,
+      "loss": 0.8894,
+      "step": 2573
+    },
+    {
+      "epoch": 0.4951904578684109,
+      "grad_norm": 0.8700493576384803,
+      "learning_rate": 1.0647554586789708e-05,
+      "loss": 0.8172,
+      "step": 2574
+    },
+    {
+      "epoch": 0.4953828395536745,
+      "grad_norm": 0.8365779949235772,
+      "learning_rate": 1.064133669282368e-05,
+      "loss": 0.8503,
+      "step": 2575
+    },
+    {
+      "epoch": 0.49557522123893805,
+      "grad_norm": 0.8637181816393922,
+      "learning_rate": 1.0635118549868668e-05,
+      "loss": 0.7898,
+      "step": 2576
+    },
+    {
+      "epoch": 0.49576760292420163,
+      "grad_norm": 0.8767604142919323,
+      "learning_rate": 1.0628900160338764e-05,
+      "loss": 0.8051,
+      "step": 2577
+    },
+    {
+      "epoch": 0.49595998460946517,
+      "grad_norm": 0.8384846488844532,
+      "learning_rate": 1.0622681526648167e-05,
+      "loss": 0.8062,
+      "step": 2578
+    },
+    {
+      "epoch": 0.49615236629472875,
+      "grad_norm": 0.7676074860599659,
+      "learning_rate": 1.0616462651211156e-05,
+      "loss": 0.7313,
+      "step": 2579
+    },
+    {
+      "epoch": 0.4963447479799923,
+      "grad_norm": 0.6356024618572844,
+      "learning_rate": 1.0610243536442125e-05,
+      "loss": 0.3447,
+      "step": 2580
+    },
+    {
+      "epoch": 0.49653712966525587,
+      "grad_norm": 0.9062027292211287,
+      "learning_rate": 1.0604024184755539e-05,
+      "loss": 0.9009,
+      "step": 2581
+    },
+    {
+      "epoch": 0.4967295113505194,
+      "grad_norm": 0.8621316361293955,
+      "learning_rate": 1.0597804598565969e-05,
+      "loss": 0.8387,
+      "step": 2582
+    },
+    {
+      "epoch": 0.496921893035783,
+      "grad_norm": 0.8695394105403499,
+      "learning_rate": 1.0591584780288069e-05,
+      "loss": 0.8326,
+      "step": 2583
+    },
+    {
+      "epoch": 0.4971142747210466,
+      "grad_norm": 0.8478268109736532,
+      "learning_rate": 1.0585364732336587e-05,
+      "loss": 0.807,
+      "step": 2584
+    },
+    {
+      "epoch": 0.4973066564063101,
+      "grad_norm": 0.8518501451240285,
+      "learning_rate": 1.0579144457126365e-05,
+      "loss": 0.8219,
+      "step": 2585
+    },
+    {
+      "epoch": 0.4974990380915737,
+      "grad_norm": 0.8749408311693839,
+      "learning_rate": 1.057292395707232e-05,
+      "loss": 0.8327,
+      "step": 2586
+    },
+    {
+      "epoch": 0.49769141977683723,
+      "grad_norm": 0.6156122944709419,
+      "learning_rate": 1.0566703234589471e-05,
+      "loss": 0.3617,
+      "step": 2587
+    },
+    {
+      "epoch": 0.4978838014621008,
+      "grad_norm": 0.8624096080097357,
+      "learning_rate": 1.0560482292092912e-05,
+      "loss": 0.868,
+      "step": 2588
+    },
+    {
+      "epoch": 0.49807618314736435,
+      "grad_norm": 0.8658537602609531,
+      "learning_rate": 1.0554261131997833e-05,
+      "loss": 0.8339,
+      "step": 2589
+    },
+    {
+      "epoch": 0.49826856483262794,
+      "grad_norm": 0.9205264801640329,
+      "learning_rate": 1.0548039756719497e-05,
+      "loss": 0.8639,
+      "step": 2590
+    },
+    {
+      "epoch": 0.4984609465178915,
+      "grad_norm": 0.905450401963335,
+      "learning_rate": 1.054181816867326e-05,
+      "loss": 0.8798,
+      "step": 2591
+    },
+    {
+      "epoch": 0.49865332820315506,
+      "grad_norm": 0.9502330791540927,
+      "learning_rate": 1.053559637027455e-05,
+      "loss": 0.8777,
+      "step": 2592
+    },
+    {
+      "epoch": 0.49884570988841864,
+      "grad_norm": 0.8970882254877709,
+      "learning_rate": 1.0529374363938888e-05,
+      "loss": 0.7368,
+      "step": 2593
+    },
+    {
+      "epoch": 0.4990380915736822,
+      "grad_norm": 0.9468459884876799,
+      "learning_rate": 1.0523152152081875e-05,
+      "loss": 0.8468,
+      "step": 2594
+    },
+    {
+      "epoch": 0.49923047325894576,
+      "grad_norm": 0.840950531354001,
+      "learning_rate": 1.051692973711918e-05,
+      "loss": 0.8706,
+      "step": 2595
+    },
+    {
+      "epoch": 0.4994228549442093,
+      "grad_norm": 0.635179262406969,
+      "learning_rate": 1.0510707121466568e-05,
+      "loss": 0.319,
+      "step": 2596
+    },
+    {
+      "epoch": 0.4996152366294729,
+      "grad_norm": 0.8484238278329653,
+      "learning_rate": 1.0504484307539864e-05,
+      "loss": 0.8346,
+      "step": 2597
+    },
+    {
+      "epoch": 0.4998076183147364,
+      "grad_norm": 0.9618870575481446,
+      "learning_rate": 1.0498261297754984e-05,
+      "loss": 0.8457,
+      "step": 2598
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8998263986087602,
+      "learning_rate": 1.0492038094527907e-05,
+      "loss": 0.8156,
+      "step": 2599
+    },
+    {
+      "epoch": 0.5001923816852636,
+      "grad_norm": 0.8653020180255447,
+      "learning_rate": 1.0485814700274706e-05,
+      "loss": 0.7675,
+      "step": 2600
+    },
+    {
+      "epoch": 0.5003847633705272,
+      "grad_norm": 0.8778240213807781,
+      "learning_rate": 1.047959111741151e-05,
+      "loss": 0.8865,
+      "step": 2601
+    },
+    {
+      "epoch": 0.5005771450557907,
+      "grad_norm": 0.9157418801827073,
+      "learning_rate": 1.0473367348354529e-05,
+      "loss": 0.8135,
+      "step": 2602
+    },
+    {
+      "epoch": 0.5007695267410542,
+      "grad_norm": 0.8829739446780719,
+      "learning_rate": 1.0467143395520044e-05,
+      "loss": 0.7916,
+      "step": 2603
+    },
+    {
+      "epoch": 0.5009619084263178,
+      "grad_norm": 0.9847372470328253,
+      "learning_rate": 1.046091926132441e-05,
+      "loss": 0.8137,
+      "step": 2604
+    },
+    {
+      "epoch": 0.5011542901115814,
+      "grad_norm": 0.8592529896191317,
+      "learning_rate": 1.0454694948184045e-05,
+      "loss": 0.8377,
+      "step": 2605
+    },
+    {
+      "epoch": 0.5013466717968449,
+      "grad_norm": 0.8467558612685046,
+      "learning_rate": 1.044847045851545e-05,
+      "loss": 0.859,
+      "step": 2606
+    },
+    {
+      "epoch": 0.5015390534821085,
+      "grad_norm": 0.9243880503388,
+      "learning_rate": 1.044224579473518e-05,
+      "loss": 0.7994,
+      "step": 2607
+    },
+    {
+      "epoch": 0.5017314351673721,
+      "grad_norm": 0.8612807779237887,
+      "learning_rate": 1.0436020959259862e-05,
+      "loss": 0.8016,
+      "step": 2608
+    },
+    {
+      "epoch": 0.5019238168526357,
+      "grad_norm": 0.8310658399422,
+      "learning_rate": 1.0429795954506203e-05,
+      "loss": 0.8324,
+      "step": 2609
+    },
+    {
+      "epoch": 0.5021161985378992,
+      "grad_norm": 0.8709458118260301,
+      "learning_rate": 1.0423570782890951e-05,
+      "loss": 0.7667,
+      "step": 2610
+    },
+    {
+      "epoch": 0.5023085802231627,
+      "grad_norm": 0.8533836565734517,
+      "learning_rate": 1.0417345446830938e-05,
+      "loss": 0.7866,
+      "step": 2611
+    },
+    {
+      "epoch": 0.5025009619084263,
+      "grad_norm": 0.8565003994810553,
+      "learning_rate": 1.0411119948743052e-05,
+      "loss": 0.8205,
+      "step": 2612
+    },
+    {
+      "epoch": 0.5026933435936899,
+      "grad_norm": 0.8485522635901058,
+      "learning_rate": 1.0404894291044247e-05,
+      "loss": 0.7989,
+      "step": 2613
+    },
+    {
+      "epoch": 0.5028857252789535,
+      "grad_norm": 0.8428318579732937,
+      "learning_rate": 1.0398668476151538e-05,
+      "loss": 0.9005,
+      "step": 2614
+    },
+    {
+      "epoch": 0.503078106964217,
+      "grad_norm": 0.8513541337625051,
+      "learning_rate": 1.0392442506482e-05,
+      "loss": 0.8197,
+      "step": 2615
+    },
+    {
+      "epoch": 0.5032704886494805,
+      "grad_norm": 0.9090729136934544,
+      "learning_rate": 1.038621638445277e-05,
+      "loss": 0.8749,
+      "step": 2616
+    },
+    {
+      "epoch": 0.5034628703347441,
+      "grad_norm": 0.8629495343664416,
+      "learning_rate": 1.037999011248104e-05,
+      "loss": 0.8507,
+      "step": 2617
+    },
+    {
+      "epoch": 0.5036552520200077,
+      "grad_norm": 0.862338816474559,
+      "learning_rate": 1.0373763692984062e-05,
+      "loss": 0.8378,
+      "step": 2618
+    },
+    {
+      "epoch": 0.5038476337052713,
+      "grad_norm": 0.8751756177875394,
+      "learning_rate": 1.0367537128379154e-05,
+      "loss": 0.793,
+      "step": 2619
+    },
+    {
+      "epoch": 0.5040400153905348,
+      "grad_norm": 0.9690439284795973,
+      "learning_rate": 1.0361310421083677e-05,
+      "loss": 0.8075,
+      "step": 2620
+    },
+    {
+      "epoch": 0.5042323970757984,
+      "grad_norm": 0.8475599354037624,
+      "learning_rate": 1.0355083573515052e-05,
+      "loss": 0.8049,
+      "step": 2621
+    },
+    {
+      "epoch": 0.504424778761062,
+      "grad_norm": 0.8489514603324821,
+      "learning_rate": 1.0348856588090764e-05,
+      "loss": 0.8623,
+      "step": 2622
+    },
+    {
+      "epoch": 0.5046171604463255,
+      "grad_norm": 0.926778960641885,
+      "learning_rate": 1.0342629467228331e-05,
+      "loss": 0.7607,
+      "step": 2623
+    },
+    {
+      "epoch": 0.504809542131589,
+      "grad_norm": 0.8813089272652068,
+      "learning_rate": 1.0336402213345345e-05,
+      "loss": 0.8311,
+      "step": 2624
+    },
+    {
+      "epoch": 0.5050019238168526,
+      "grad_norm": 0.8626709131904453,
+      "learning_rate": 1.0330174828859434e-05,
+      "loss": 0.8336,
+      "step": 2625
+    },
+    {
+      "epoch": 0.5051943055021162,
+      "grad_norm": 0.8025365794327385,
+      "learning_rate": 1.0323947316188288e-05,
+      "loss": 0.8072,
+      "step": 2626
+    },
+    {
+      "epoch": 0.5053866871873798,
+      "grad_norm": 0.8660285274939085,
+      "learning_rate": 1.031771967774964e-05,
+      "loss": 0.8555,
+      "step": 2627
+    },
+    {
+      "epoch": 0.5055790688726434,
+      "grad_norm": 0.8883978878514266,
+      "learning_rate": 1.0311491915961271e-05,
+      "loss": 0.871,
+      "step": 2628
+    },
+    {
+      "epoch": 0.5057714505579068,
+      "grad_norm": 0.8092254782509501,
+      "learning_rate": 1.030526403324102e-05,
+      "loss": 0.8194,
+      "step": 2629
+    },
+    {
+      "epoch": 0.5059638322431704,
+      "grad_norm": 0.9050663846378942,
+      "learning_rate": 1.0299036032006759e-05,
+      "loss": 0.8078,
+      "step": 2630
+    },
+    {
+      "epoch": 0.506156213928434,
+      "grad_norm": 0.8973027571004367,
+      "learning_rate": 1.0292807914676412e-05,
+      "loss": 0.8149,
+      "step": 2631
+    },
+    {
+      "epoch": 0.5063485956136976,
+      "grad_norm": 0.9838319933136038,
+      "learning_rate": 1.0286579683667952e-05,
+      "loss": 0.8273,
+      "step": 2632
+    },
+    {
+      "epoch": 0.5065409772989612,
+      "grad_norm": 0.8078839719389556,
+      "learning_rate": 1.0280351341399392e-05,
+      "loss": 0.8244,
+      "step": 2633
+    },
+    {
+      "epoch": 0.5067333589842247,
+      "grad_norm": 0.8667296962237256,
+      "learning_rate": 1.027412289028879e-05,
+      "loss": 0.8665,
+      "step": 2634
+    },
+    {
+      "epoch": 0.5069257406694883,
+      "grad_norm": 0.6281613099627591,
+      "learning_rate": 1.0267894332754243e-05,
+      "loss": 0.3582,
+      "step": 2635
+    },
+    {
+      "epoch": 0.5071181223547518,
+      "grad_norm": 0.8587189820535744,
+      "learning_rate": 1.0261665671213891e-05,
+      "loss": 0.7886,
+      "step": 2636
+    },
+    {
+      "epoch": 0.5073105040400154,
+      "grad_norm": 0.8735785042397007,
+      "learning_rate": 1.0255436908085919e-05,
+      "loss": 0.8487,
+      "step": 2637
+    },
+    {
+      "epoch": 0.5075028857252789,
+      "grad_norm": 0.6317365701145404,
+      "learning_rate": 1.024920804578854e-05,
+      "loss": 0.3574,
+      "step": 2638
+    },
+    {
+      "epoch": 0.5076952674105425,
+      "grad_norm": 0.8044598966797523,
+      "learning_rate": 1.0242979086740019e-05,
+      "loss": 0.8164,
+      "step": 2639
+    },
+    {
+      "epoch": 0.5078876490958061,
+      "grad_norm": 0.9212712659003272,
+      "learning_rate": 1.023675003335865e-05,
+      "loss": 0.7923,
+      "step": 2640
+    },
+    {
+      "epoch": 0.5080800307810697,
+      "grad_norm": 0.6268895753108752,
+      "learning_rate": 1.0230520888062765e-05,
+      "loss": 0.3443,
+      "step": 2641
+    },
+    {
+      "epoch": 0.5082724124663333,
+      "grad_norm": 0.9431591790201737,
+      "learning_rate": 1.0224291653270739e-05,
+      "loss": 0.863,
+      "step": 2642
+    },
+    {
+      "epoch": 0.5084647941515967,
+      "grad_norm": 0.9105984817721815,
+      "learning_rate": 1.0218062331400969e-05,
+      "loss": 0.8115,
+      "step": 2643
+    },
+    {
+      "epoch": 0.5086571758368603,
+      "grad_norm": 0.8141939961829593,
+      "learning_rate": 1.0211832924871889e-05,
+      "loss": 0.8259,
+      "step": 2644
+    },
+    {
+      "epoch": 0.5088495575221239,
+      "grad_norm": 0.9127591101249732,
+      "learning_rate": 1.0205603436101978e-05,
+      "loss": 0.8801,
+      "step": 2645
+    },
+    {
+      "epoch": 0.5090419392073875,
+      "grad_norm": 0.9015510785559816,
+      "learning_rate": 1.0199373867509734e-05,
+      "loss": 0.8307,
+      "step": 2646
+    },
+    {
+      "epoch": 0.509234320892651,
+      "grad_norm": 0.8671499489088926,
+      "learning_rate": 1.019314422151369e-05,
+      "loss": 0.8442,
+      "step": 2647
+    },
+    {
+      "epoch": 0.5094267025779146,
+      "grad_norm": 0.9101741502658528,
+      "learning_rate": 1.0186914500532408e-05,
+      "loss": 0.8256,
+      "step": 2648
+    },
+    {
+      "epoch": 0.5096190842631781,
+      "grad_norm": 0.9038150706630497,
+      "learning_rate": 1.0180684706984483e-05,
+      "loss": 0.8247,
+      "step": 2649
+    },
+    {
+      "epoch": 0.5098114659484417,
+      "grad_norm": 0.8924870035793543,
+      "learning_rate": 1.0174454843288533e-05,
+      "loss": 0.8625,
+      "step": 2650
+    },
+    {
+      "epoch": 0.5100038476337053,
+      "grad_norm": 0.933965382139782,
+      "learning_rate": 1.0168224911863205e-05,
+      "loss": 0.8413,
+      "step": 2651
+    },
+    {
+      "epoch": 0.5101962293189688,
+      "grad_norm": 0.919458892228402,
+      "learning_rate": 1.0161994915127173e-05,
+      "loss": 0.7775,
+      "step": 2652
+    },
+    {
+      "epoch": 0.5103886110042324,
+      "grad_norm": 0.9665252126139605,
+      "learning_rate": 1.015576485549914e-05,
+      "loss": 0.841,
+      "step": 2653
+    },
+    {
+      "epoch": 0.510580992689496,
+      "grad_norm": 0.8409739313897645,
+      "learning_rate": 1.0149534735397823e-05,
+      "loss": 0.8181,
+      "step": 2654
+    },
+    {
+      "epoch": 0.5107733743747596,
+      "grad_norm": 0.9380048075963482,
+      "learning_rate": 1.0143304557241979e-05,
+      "loss": 0.8413,
+      "step": 2655
+    },
+    {
+      "epoch": 0.510965756060023,
+      "grad_norm": 0.990020899601608,
+      "learning_rate": 1.0137074323450372e-05,
+      "loss": 0.8379,
+      "step": 2656
+    },
+    {
+      "epoch": 0.5111581377452866,
+      "grad_norm": 0.8847530322789068,
+      "learning_rate": 1.0130844036441787e-05,
+      "loss": 0.7762,
+      "step": 2657
+    },
+    {
+      "epoch": 0.5113505194305502,
+      "grad_norm": 0.89203182118148,
+      "learning_rate": 1.0124613698635043e-05,
+      "loss": 0.8301,
+      "step": 2658
+    },
+    {
+      "epoch": 0.5115429011158138,
+      "grad_norm": 0.8332480313910846,
+      "learning_rate": 1.0118383312448973e-05,
+      "loss": 0.8754,
+      "step": 2659
+    },
+    {
+      "epoch": 0.5117352828010774,
+      "grad_norm": 0.6836825955262472,
+      "learning_rate": 1.0112152880302426e-05,
+      "loss": 0.3432,
+      "step": 2660
+    },
+    {
+      "epoch": 0.5119276644863409,
+      "grad_norm": 0.8392077387832436,
+      "learning_rate": 1.0105922404614265e-05,
+      "loss": 0.8598,
+      "step": 2661
+    },
+    {
+      "epoch": 0.5121200461716044,
+      "grad_norm": 0.8591052067401225,
+      "learning_rate": 1.0099691887803385e-05,
+      "loss": 0.8708,
+      "step": 2662
+    },
+    {
+      "epoch": 0.512312427856868,
+      "grad_norm": 0.8648589413575638,
+      "learning_rate": 1.0093461332288678e-05,
+      "loss": 0.8722,
+      "step": 2663
+    },
+    {
+      "epoch": 0.5125048095421316,
+      "grad_norm": 0.6505952101105859,
+      "learning_rate": 1.0087230740489065e-05,
+      "loss": 0.3457,
+      "step": 2664
+    },
+    {
+      "epoch": 0.5126971912273951,
+      "grad_norm": 0.8051188094944637,
+      "learning_rate": 1.0081000114823473e-05,
+      "loss": 0.8528,
+      "step": 2665
+    },
+    {
+      "epoch": 0.5128895729126587,
+      "grad_norm": 0.9153856081300382,
+      "learning_rate": 1.007476945771085e-05,
+      "loss": 0.8415,
+      "step": 2666
+    },
+    {
+      "epoch": 0.5130819545979223,
+      "grad_norm": 0.9347162707091003,
+      "learning_rate": 1.006853877157015e-05,
+      "loss": 0.836,
+      "step": 2667
+    },
+    {
+      "epoch": 0.5132743362831859,
+      "grad_norm": 0.9134143011125259,
+      "learning_rate": 1.0062308058820337e-05,
+      "loss": 0.8535,
+      "step": 2668
+    },
+    {
+      "epoch": 0.5134667179684494,
+      "grad_norm": 0.8702437567276248,
+      "learning_rate": 1.0056077321880393e-05,
+      "loss": 0.8391,
+      "step": 2669
+    },
+    {
+      "epoch": 0.5136590996537129,
+      "grad_norm": 0.8278675907563775,
+      "learning_rate": 1.0049846563169297e-05,
+      "loss": 0.8242,
+      "step": 2670
+    },
+    {
+      "epoch": 0.5138514813389765,
+      "grad_norm": 0.8516679000849016,
+      "learning_rate": 1.0043615785106051e-05,
+      "loss": 0.8392,
+      "step": 2671
+    },
+    {
+      "epoch": 0.5140438630242401,
+      "grad_norm": 0.8425956349200947,
+      "learning_rate": 1.0037384990109658e-05,
+      "loss": 0.8036,
+      "step": 2672
+    },
+    {
+      "epoch": 0.5142362447095037,
+      "grad_norm": 0.8464259405389564,
+      "learning_rate": 1.0031154180599123e-05,
+      "loss": 0.827,
+      "step": 2673
+    },
+    {
+      "epoch": 0.5144286263947672,
+      "grad_norm": 0.8772043280406407,
+      "learning_rate": 1.0024923358993458e-05,
+      "loss": 0.8561,
+      "step": 2674
+    },
+    {
+      "epoch": 0.5146210080800308,
+      "grad_norm": 0.8524096084173326,
+      "learning_rate": 1.0018692527711695e-05,
+      "loss": 0.8216,
+      "step": 2675
+    },
+    {
+      "epoch": 0.5148133897652943,
+      "grad_norm": 0.8202683108327804,
+      "learning_rate": 1.0012461689172846e-05,
+      "loss": 0.7972,
+      "step": 2676
+    },
+    {
+      "epoch": 0.5150057714505579,
+      "grad_norm": 0.850277609644827,
+      "learning_rate": 1.0006230845795937e-05,
+      "loss": 0.8575,
+      "step": 2677
+    },
+    {
+      "epoch": 0.5151981531358215,
+      "grad_norm": 0.8731486673106419,
+      "learning_rate": 1e-05,
+      "loss": 0.7433,
+      "step": 2678
+    },
+    {
+      "epoch": 0.515390534821085,
+      "grad_norm": 0.9131553839443313,
+      "learning_rate": 9.993769154204063e-06,
+      "loss": 0.8518,
+      "step": 2679
+    },
+    {
+      "epoch": 0.5155829165063486,
+      "grad_norm": 0.8692860023179773,
+      "learning_rate": 9.987538310827159e-06,
+      "loss": 0.7932,
+      "step": 2680
+    },
+    {
+      "epoch": 0.5157752981916122,
+      "grad_norm": 0.879050250555243,
+      "learning_rate": 9.981307472288308e-06,
+      "loss": 0.8517,
+      "step": 2681
+    },
+    {
+      "epoch": 0.5159676798768758,
+      "grad_norm": 0.8539122554253337,
+      "learning_rate": 9.975076641006542e-06,
+      "loss": 0.8167,
+      "step": 2682
+    },
+    {
+      "epoch": 0.5161600615621393,
+      "grad_norm": 0.8854505971951345,
+      "learning_rate": 9.968845819400883e-06,
+      "loss": 0.8327,
+      "step": 2683
+    },
+    {
+      "epoch": 0.5163524432474028,
+      "grad_norm": 0.9158359961125537,
+      "learning_rate": 9.962615009890346e-06,
+      "loss": 0.8696,
+      "step": 2684
+    },
+    {
+      "epoch": 0.5165448249326664,
+      "grad_norm": 0.9050947854434997,
+      "learning_rate": 9.956384214893949e-06,
+      "loss": 0.8401,
+      "step": 2685
+    },
+    {
+      "epoch": 0.51673720661793,
+      "grad_norm": 0.9047529639633572,
+      "learning_rate": 9.950153436830707e-06,
+      "loss": 0.905,
+      "step": 2686
+    },
+    {
+      "epoch": 0.5169295883031936,
+      "grad_norm": 0.8621312456598139,
+      "learning_rate": 9.94392267811961e-06,
+      "loss": 0.8433,
+      "step": 2687
+    },
+    {
+      "epoch": 0.517121969988457,
+      "grad_norm": 1.032991490767895,
+      "learning_rate": 9.937691941179665e-06,
+      "loss": 0.8237,
+      "step": 2688
+    },
+    {
+      "epoch": 0.5173143516737206,
+      "grad_norm": 0.856564084418986,
+      "learning_rate": 9.931461228429856e-06,
+      "loss": 0.8401,
+      "step": 2689
+    },
+    {
+      "epoch": 0.5175067333589842,
+      "grad_norm": 0.9581032117694721,
+      "learning_rate": 9.925230542289151e-06,
+      "loss": 0.8202,
+      "step": 2690
+    },
+    {
+      "epoch": 0.5176991150442478,
+      "grad_norm": 0.6532873735198744,
+      "learning_rate": 9.91899988517653e-06,
+      "loss": 0.3493,
+      "step": 2691
+    },
+    {
+      "epoch": 0.5178914967295114,
+      "grad_norm": 0.8929239706137039,
+      "learning_rate": 9.912769259510938e-06,
+      "loss": 0.848,
+      "step": 2692
+    },
+    {
+      "epoch": 0.5180838784147749,
+      "grad_norm": 0.8605512735436497,
+      "learning_rate": 9.906538667711324e-06,
+      "loss": 0.8216,
+      "step": 2693
+    },
+    {
+      "epoch": 0.5182762601000385,
+      "grad_norm": 0.9083203197231146,
+      "learning_rate": 9.90030811219662e-06,
+      "loss": 0.8332,
+      "step": 2694
+    },
+    {
+      "epoch": 0.518468641785302,
+      "grad_norm": 0.9005723593841868,
+      "learning_rate": 9.894077595385736e-06,
+      "loss": 0.8005,
+      "step": 2695
+    },
+    {
+      "epoch": 0.5186610234705656,
+      "grad_norm": 0.8385915668552527,
+      "learning_rate": 9.887847119697577e-06,
+      "loss": 0.8976,
+      "step": 2696
+    },
+    {
+      "epoch": 0.5188534051558291,
+      "grad_norm": 0.8794215942150431,
+      "learning_rate": 9.881616687551032e-06,
+      "loss": 0.8114,
+      "step": 2697
+    },
+    {
+      "epoch": 0.5190457868410927,
+      "grad_norm": 0.8518847628288407,
+      "learning_rate": 9.875386301364958e-06,
+      "loss": 0.7656,
+      "step": 2698
+    },
+    {
+      "epoch": 0.5192381685263563,
+      "grad_norm": 0.798299079966406,
+      "learning_rate": 9.869155963558215e-06,
+      "loss": 0.8094,
+      "step": 2699
+    },
+    {
+      "epoch": 0.5194305502116199,
+      "grad_norm": 0.9315667601854666,
+      "learning_rate": 9.862925676549635e-06,
+      "loss": 0.8294,
+      "step": 2700
+    },
+    {
+      "epoch": 0.5196229318968835,
+      "grad_norm": 0.8548006051041143,
+      "learning_rate": 9.856695442758023e-06,
+      "loss": 0.8242,
+      "step": 2701
+    },
+    {
+      "epoch": 0.5198153135821469,
+      "grad_norm": 0.8339054021614364,
+      "learning_rate": 9.850465264602175e-06,
+      "loss": 0.8033,
+      "step": 2702
+    },
+    {
+      "epoch": 0.5200076952674105,
+      "grad_norm": 0.835540755574849,
+      "learning_rate": 9.844235144500865e-06,
+      "loss": 0.7804,
+      "step": 2703
+    },
+    {
+      "epoch": 0.5202000769526741,
+      "grad_norm": 0.8421579063200216,
+      "learning_rate": 9.83800508487283e-06,
+      "loss": 0.802,
+      "step": 2704
+    },
+    {
+      "epoch": 0.5203924586379377,
+      "grad_norm": 0.8672669582711163,
+      "learning_rate": 9.831775088136797e-06,
+      "loss": 0.7861,
+      "step": 2705
+    },
+    {
+      "epoch": 0.5205848403232012,
+      "grad_norm": 0.850654723514553,
+      "learning_rate": 9.82554515671147e-06,
+      "loss": 0.8218,
+      "step": 2706
+    },
+    {
+      "epoch": 0.5207772220084648,
+      "grad_norm": 0.8573921236238331,
+      "learning_rate": 9.819315293015519e-06,
+      "loss": 0.7946,
+      "step": 2707
+    },
+    {
+      "epoch": 0.5209696036937284,
+      "grad_norm": 0.8324380509489536,
+      "learning_rate": 9.813085499467594e-06,
+      "loss": 0.8509,
+      "step": 2708
+    },
+    {
+      "epoch": 0.5211619853789919,
+      "grad_norm": 0.8814191160443499,
+      "learning_rate": 9.806855778486314e-06,
+      "loss": 0.8448,
+      "step": 2709
+    },
+    {
+      "epoch": 0.5213543670642555,
+      "grad_norm": 0.8064334923928348,
+      "learning_rate": 9.800626132490268e-06,
+      "loss": 0.7865,
+      "step": 2710
+    },
+    {
+      "epoch": 0.521546748749519,
+      "grad_norm": 0.825541029258176,
+      "learning_rate": 9.794396563898022e-06,
+      "loss": 0.7609,
+      "step": 2711
+    },
+    {
+      "epoch": 0.5217391304347826,
+      "grad_norm": 0.820862532377526,
+      "learning_rate": 9.788167075128113e-06,
+      "loss": 0.7908,
+      "step": 2712
+    },
+    {
+      "epoch": 0.5219315121200462,
+      "grad_norm": 0.8898738873939938,
+      "learning_rate": 9.781937668599035e-06,
+      "loss": 0.7934,
+      "step": 2713
+    },
+    {
+      "epoch": 0.5221238938053098,
+      "grad_norm": 0.8251242078946859,
+      "learning_rate": 9.775708346729263e-06,
+      "loss": 0.815,
+      "step": 2714
+    },
+    {
+      "epoch": 0.5223162754905732,
+      "grad_norm": 0.6803374866387527,
+      "learning_rate": 9.769479111937238e-06,
+      "loss": 0.3363,
+      "step": 2715
+    },
+    {
+      "epoch": 0.5225086571758368,
+      "grad_norm": 0.8310519945471133,
+      "learning_rate": 9.763249966641352e-06,
+      "loss": 0.8682,
+      "step": 2716
+    },
+    {
+      "epoch": 0.5227010388611004,
+      "grad_norm": 0.8910398361160671,
+      "learning_rate": 9.757020913259986e-06,
+      "loss": 0.8578,
+      "step": 2717
+    },
+    {
+      "epoch": 0.522893420546364,
+      "grad_norm": 0.8919326145822306,
+      "learning_rate": 9.750791954211464e-06,
+      "loss": 0.7984,
+      "step": 2718
+    },
+    {
+      "epoch": 0.5230858022316276,
+      "grad_norm": 0.844638072130958,
+      "learning_rate": 9.744563091914085e-06,
+      "loss": 0.8447,
+      "step": 2719
+    },
+    {
+      "epoch": 0.5232781839168911,
+      "grad_norm": 0.7873011420887455,
+      "learning_rate": 9.738334328786114e-06,
+      "loss": 0.8751,
+      "step": 2720
+    },
+    {
+      "epoch": 0.5234705656021547,
+      "grad_norm": 0.8431066891888871,
+      "learning_rate": 9.732105667245759e-06,
+      "loss": 0.884,
+      "step": 2721
+    },
+    {
+      "epoch": 0.5236629472874182,
+      "grad_norm": 0.8730110837041755,
+      "learning_rate": 9.725877109711212e-06,
+      "loss": 0.8401,
+      "step": 2722
+    },
+    {
+      "epoch": 0.5238553289726818,
+      "grad_norm": 0.894959095764068,
+      "learning_rate": 9.719648658600611e-06,
+      "loss": 0.8155,
+      "step": 2723
+    },
+    {
+      "epoch": 0.5240477106579453,
+      "grad_norm": 0.9028141275497908,
+      "learning_rate": 9.71342031633205e-06,
+      "loss": 0.8622,
+      "step": 2724
+    },
+    {
+      "epoch": 0.5242400923432089,
+      "grad_norm": 0.794503319843355,
+      "learning_rate": 9.70719208532359e-06,
+      "loss": 0.8046,
+      "step": 2725
+    },
+    {
+      "epoch": 0.5244324740284725,
+      "grad_norm": 0.9126709509059278,
+      "learning_rate": 9.700963967993246e-06,
+      "loss": 0.8626,
+      "step": 2726
+    },
+    {
+      "epoch": 0.5246248557137361,
+      "grad_norm": 0.8532984448066787,
+      "learning_rate": 9.694735966758982e-06,
+      "loss": 0.7763,
+      "step": 2727
+    },
+    {
+      "epoch": 0.5248172373989997,
+      "grad_norm": 0.8913465114434047,
+      "learning_rate": 9.688508084038729e-06,
+      "loss": 0.7679,
+      "step": 2728
+    },
+    {
+      "epoch": 0.5250096190842631,
+      "grad_norm": 0.8249971595164921,
+      "learning_rate": 9.682280322250365e-06,
+      "loss": 0.8375,
+      "step": 2729
+    },
+    {
+      "epoch": 0.5252020007695267,
+      "grad_norm": 0.87189743461422,
+      "learning_rate": 9.676052683811715e-06,
+      "loss": 0.8437,
+      "step": 2730
+    },
+    {
+      "epoch": 0.5253943824547903,
+      "grad_norm": 0.7953817555264027,
+      "learning_rate": 9.669825171140568e-06,
+      "loss": 0.7924,
+      "step": 2731
+    },
+    {
+      "epoch": 0.5255867641400539,
+      "grad_norm": 0.83372106193568,
+      "learning_rate": 9.66359778665466e-06,
+      "loss": 0.8099,
+      "step": 2732
+    },
+    {
+      "epoch": 0.5257791458253175,
+      "grad_norm": 0.8505267509713246,
+      "learning_rate": 9.657370532771672e-06,
+      "loss": 0.8318,
+      "step": 2733
+    },
+    {
+      "epoch": 0.525971527510581,
+      "grad_norm": 0.8935267197982699,
+      "learning_rate": 9.651143411909241e-06,
+      "loss": 0.8767,
+      "step": 2734
+    },
+    {
+      "epoch": 0.5261639091958445,
+      "grad_norm": 0.8262033693584799,
+      "learning_rate": 9.64491642648495e-06,
+      "loss": 0.8686,
+      "step": 2735
+    },
+    {
+      "epoch": 0.5263562908811081,
+      "grad_norm": 0.8748814596785551,
+      "learning_rate": 9.638689578916326e-06,
+      "loss": 0.8685,
+      "step": 2736
+    },
+    {
+      "epoch": 0.5265486725663717,
+      "grad_norm": 0.9169096567736947,
+      "learning_rate": 9.632462871620847e-06,
+      "loss": 0.8231,
+      "step": 2737
+    },
+    {
+      "epoch": 0.5267410542516352,
+      "grad_norm": 0.6507801481316838,
+      "learning_rate": 9.62623630701594e-06,
+      "loss": 0.3311,
+      "step": 2738
+    },
+    {
+      "epoch": 0.5269334359368988,
+      "grad_norm": 0.8134936556035963,
+      "learning_rate": 9.620009887518963e-06,
+      "loss": 0.8041,
+      "step": 2739
+    },
+    {
+      "epoch": 0.5271258176221624,
+      "grad_norm": 0.8638241893478145,
+      "learning_rate": 9.613783615547233e-06,
+      "loss": 0.7965,
+      "step": 2740
+    },
+    {
+      "epoch": 0.527318199307426,
+      "grad_norm": 0.8543839197851906,
+      "learning_rate": 9.607557493518006e-06,
+      "loss": 0.7897,
+      "step": 2741
+    },
+    {
+      "epoch": 0.5275105809926895,
+      "grad_norm": 0.835294559579335,
+      "learning_rate": 9.601331523848464e-06,
+      "loss": 0.7775,
+      "step": 2742
+    },
+    {
+      "epoch": 0.527702962677953,
+      "grad_norm": 0.8707701894017678,
+      "learning_rate": 9.595105708955758e-06,
+      "loss": 0.8754,
+      "step": 2743
+    },
+    {
+      "epoch": 0.5278953443632166,
+      "grad_norm": 0.8759767731443833,
+      "learning_rate": 9.588880051256951e-06,
+      "loss": 0.804,
+      "step": 2744
+    },
+    {
+      "epoch": 0.5280877260484802,
+      "grad_norm": 0.8679675082298095,
+      "learning_rate": 9.582654553169064e-06,
+      "loss": 0.7998,
+      "step": 2745
+    },
+    {
+      "epoch": 0.5282801077337438,
+      "grad_norm": 0.6312852145931005,
+      "learning_rate": 9.576429217109054e-06,
+      "loss": 0.3292,
+      "step": 2746
+    },
+    {
+      "epoch": 0.5284724894190073,
+      "grad_norm": 0.884534012321794,
+      "learning_rate": 9.5702040454938e-06,
+      "loss": 0.8027,
+      "step": 2747
+    },
+    {
+      "epoch": 0.5286648711042709,
+      "grad_norm": 0.7019751933745078,
+      "learning_rate": 9.563979040740138e-06,
+      "loss": 0.3415,
+      "step": 2748
+    },
+    {
+      "epoch": 0.5288572527895344,
+      "grad_norm": 0.8944545098253368,
+      "learning_rate": 9.557754205264826e-06,
+      "loss": 0.8636,
+      "step": 2749
+    },
+    {
+      "epoch": 0.529049634474798,
+      "grad_norm": 0.9896861500469162,
+      "learning_rate": 9.551529541484554e-06,
+      "loss": 0.8517,
+      "step": 2750
+    },
+    {
+      "epoch": 0.5292420161600616,
+      "grad_norm": 0.9310669052060959,
+      "learning_rate": 9.545305051815957e-06,
+      "loss": 0.8915,
+      "step": 2751
+    },
+    {
+      "epoch": 0.5294343978453251,
+      "grad_norm": 0.8863277690499666,
+      "learning_rate": 9.539080738675597e-06,
+      "loss": 0.8411,
+      "step": 2752
+    },
+    {
+      "epoch": 0.5296267795305887,
+      "grad_norm": 0.888894629305449,
+      "learning_rate": 9.53285660447996e-06,
+      "loss": 0.7973,
+      "step": 2753
+    },
+    {
+      "epoch": 0.5298191612158523,
+      "grad_norm": 0.861521682296519,
+      "learning_rate": 9.526632651645476e-06,
+      "loss": 0.8479,
+      "step": 2754
+    },
+    {
+      "epoch": 0.5300115429011159,
+      "grad_norm": 0.9434589701948916,
+      "learning_rate": 9.520408882588497e-06,
+      "loss": 0.8501,
+      "step": 2755
+    },
+    {
+      "epoch": 0.5302039245863793,
+      "grad_norm": 0.7855409622776376,
+      "learning_rate": 9.514185299725299e-06,
+      "loss": 0.8018,
+      "step": 2756
+    },
+    {
+      "epoch": 0.5303963062716429,
+      "grad_norm": 0.8296352561345101,
+      "learning_rate": 9.507961905472093e-06,
+      "loss": 0.8198,
+      "step": 2757
+    },
+    {
+      "epoch": 0.5305886879569065,
+      "grad_norm": 0.8361147395224573,
+      "learning_rate": 9.501738702245023e-06,
+      "loss": 0.8046,
+      "step": 2758
+    },
+    {
+      "epoch": 0.5307810696421701,
+      "grad_norm": 0.8423805487275045,
+      "learning_rate": 9.495515692460138e-06,
+      "loss": 0.7753,
+      "step": 2759
+    },
+    {
+      "epoch": 0.5309734513274337,
+      "grad_norm": 0.8450236397168238,
+      "learning_rate": 9.489292878533436e-06,
+      "loss": 0.9156,
+      "step": 2760
+    },
+    {
+      "epoch": 0.5311658330126972,
+      "grad_norm": 0.9257510579824286,
+      "learning_rate": 9.483070262880823e-06,
+      "loss": 0.8375,
+      "step": 2761
+    },
+    {
+      "epoch": 0.5313582146979607,
+      "grad_norm": 0.8923605594460242,
+      "learning_rate": 9.476847847918126e-06,
+      "loss": 0.8866,
+      "step": 2762
+    },
+    {
+      "epoch": 0.5315505963832243,
+      "grad_norm": 0.82584822417588,
+      "learning_rate": 9.47062563606111e-06,
+      "loss": 0.8046,
+      "step": 2763
+    },
+    {
+      "epoch": 0.5317429780684879,
+      "grad_norm": 0.8767226976081574,
+      "learning_rate": 9.464403629725454e-06,
+      "loss": 0.8158,
+      "step": 2764
+    },
+    {
+      "epoch": 0.5319353597537514,
+      "grad_norm": 0.90938372720237,
+      "learning_rate": 9.458181831326744e-06,
+      "loss": 0.848,
+      "step": 2765
+    },
+    {
+      "epoch": 0.532127741439015,
+      "grad_norm": 0.9560359654021291,
+      "learning_rate": 9.451960243280506e-06,
+      "loss": 0.8648,
+      "step": 2766
+    },
+    {
+      "epoch": 0.5323201231242786,
+      "grad_norm": 0.8848198639727689,
+      "learning_rate": 9.44573886800217e-06,
+      "loss": 0.8264,
+      "step": 2767
+    },
+    {
+      "epoch": 0.5325125048095422,
+      "grad_norm": 0.9718027430455471,
+      "learning_rate": 9.43951770790709e-06,
+      "loss": 0.8188,
+      "step": 2768
+    },
+    {
+      "epoch": 0.5327048864948057,
+      "grad_norm": 0.6281250769190692,
+      "learning_rate": 9.433296765410534e-06,
+      "loss": 0.3375,
+      "step": 2769
+    },
+    {
+      "epoch": 0.5328972681800692,
+      "grad_norm": 0.9241477070188259,
+      "learning_rate": 9.427076042927683e-06,
+      "loss": 0.826,
+      "step": 2770
+    },
+    {
+      "epoch": 0.5330896498653328,
+      "grad_norm": 0.9142438111342238,
+      "learning_rate": 9.420855542873638e-06,
+      "loss": 0.7969,
+      "step": 2771
+    },
+    {
+      "epoch": 0.5332820315505964,
+      "grad_norm": 0.8108887335980554,
+      "learning_rate": 9.414635267663416e-06,
+      "loss": 0.7746,
+      "step": 2772
+    },
+    {
+      "epoch": 0.53347441323586,
+      "grad_norm": 0.8510574704318876,
+      "learning_rate": 9.408415219711934e-06,
+      "loss": 0.7419,
+      "step": 2773
+    },
+    {
+      "epoch": 0.5336667949211235,
+      "grad_norm": 0.8589902585534405,
+      "learning_rate": 9.402195401434036e-06,
+      "loss": 0.8104,
+      "step": 2774
+    },
+    {
+      "epoch": 0.533859176606387,
+      "grad_norm": 0.8842335777658542,
+      "learning_rate": 9.395975815244468e-06,
+      "loss": 0.8129,
+      "step": 2775
+    },
+    {
+      "epoch": 0.5340515582916506,
+      "grad_norm": 0.8433922825031764,
+      "learning_rate": 9.389756463557878e-06,
+      "loss": 0.8437,
+      "step": 2776
+    },
+    {
+      "epoch": 0.5342439399769142,
+      "grad_norm": 0.7001749629217404,
+      "learning_rate": 9.383537348788844e-06,
+      "loss": 0.3581,
+      "step": 2777
+    },
+    {
+      "epoch": 0.5344363216621778,
+      "grad_norm": 0.6252101628378834,
+      "learning_rate": 9.377318473351838e-06,
+      "loss": 0.3745,
+      "step": 2778
+    },
+    {
+      "epoch": 0.5346287033474413,
+      "grad_norm": 0.8316448704384114,
+      "learning_rate": 9.371099839661238e-06,
+      "loss": 0.7879,
+      "step": 2779
+    },
+    {
+      "epoch": 0.5348210850327049,
+      "grad_norm": 0.8040810357279009,
+      "learning_rate": 9.364881450131335e-06,
+      "loss": 0.8059,
+      "step": 2780
+    },
+    {
+      "epoch": 0.5350134667179685,
+      "grad_norm": 0.8315338328052742,
+      "learning_rate": 9.358663307176323e-06,
+      "loss": 0.833,
+      "step": 2781
+    },
+    {
+      "epoch": 0.535205848403232,
+      "grad_norm": 0.8838523584675149,
+      "learning_rate": 9.352445413210294e-06,
+      "loss": 0.8324,
+      "step": 2782
+    },
+    {
+      "epoch": 0.5353982300884956,
+      "grad_norm": 0.847005941988616,
+      "learning_rate": 9.346227770647251e-06,
+      "loss": 0.7938,
+      "step": 2783
+    },
+    {
+      "epoch": 0.5355906117737591,
+      "grad_norm": 0.8939149319420954,
+      "learning_rate": 9.3400103819011e-06,
+      "loss": 0.8259,
+      "step": 2784
+    },
+    {
+      "epoch": 0.5357829934590227,
+      "grad_norm": 0.929662306963428,
+      "learning_rate": 9.33379324938564e-06,
+      "loss": 0.8385,
+      "step": 2785
+    },
+    {
+      "epoch": 0.5359753751442863,
+      "grad_norm": 0.9188540326220511,
+      "learning_rate": 9.327576375514582e-06,
+      "loss": 0.819,
+      "step": 2786
+    },
+    {
+      "epoch": 0.5361677568295499,
+      "grad_norm": 0.6787275331823793,
+      "learning_rate": 9.321359762701527e-06,
+      "loss": 0.3719,
+      "step": 2787
+    },
+    {
+      "epoch": 0.5363601385148133,
+      "grad_norm": 0.7923822970479446,
+      "learning_rate": 9.315143413359975e-06,
+      "loss": 0.7583,
+      "step": 2788
+    },
+    {
+      "epoch": 0.5365525202000769,
+      "grad_norm": 0.8618370964660496,
+      "learning_rate": 9.308927329903333e-06,
+      "loss": 0.8189,
+      "step": 2789
+    },
+    {
+      "epoch": 0.5367449018853405,
+      "grad_norm": 0.9079471408977736,
+      "learning_rate": 9.302711514744897e-06,
+      "loss": 0.8554,
+      "step": 2790
+    },
+    {
+      "epoch": 0.5369372835706041,
+      "grad_norm": 0.8822182782567434,
+      "learning_rate": 9.296495970297855e-06,
+      "loss": 0.7737,
+      "step": 2791
+    },
+    {
+      "epoch": 0.5371296652558677,
+      "grad_norm": 0.8420812349977617,
+      "learning_rate": 9.290280698975307e-06,
+      "loss": 0.8285,
+      "step": 2792
+    },
+    {
+      "epoch": 0.5373220469411312,
+      "grad_norm": 0.8632047995738846,
+      "learning_rate": 9.284065703190221e-06,
+      "loss": 0.8302,
+      "step": 2793
+    },
+    {
+      "epoch": 0.5375144286263948,
+      "grad_norm": 0.8711365709120678,
+      "learning_rate": 9.27785098535548e-06,
+      "loss": 0.7553,
+      "step": 2794
+    },
+    {
+      "epoch": 0.5377068103116583,
+      "grad_norm": 0.8884476160430452,
+      "learning_rate": 9.271636547883856e-06,
+      "loss": 0.7967,
+      "step": 2795
+    },
+    {
+      "epoch": 0.5378991919969219,
+      "grad_norm": 0.9764610133276349,
+      "learning_rate": 9.265422393187998e-06,
+      "loss": 0.8557,
+      "step": 2796
+    },
+    {
+      "epoch": 0.5380915736821854,
+      "grad_norm": 0.8169109414422511,
+      "learning_rate": 9.259208523680457e-06,
+      "loss": 0.7793,
+      "step": 2797
+    },
+    {
+      "epoch": 0.538283955367449,
+      "grad_norm": 0.846785851774684,
+      "learning_rate": 9.252994941773679e-06,
+      "loss": 0.7932,
+      "step": 2798
+    },
+    {
+      "epoch": 0.5384763370527126,
+      "grad_norm": 0.8165948386777315,
+      "learning_rate": 9.24678164987998e-06,
+      "loss": 0.7708,
+      "step": 2799
+    },
+    {
+      "epoch": 0.5386687187379762,
+      "grad_norm": 0.9025804801936065,
+      "learning_rate": 9.24056865041158e-06,
+      "loss": 0.8214,
+      "step": 2800
+    },
+    {
+      "epoch": 0.5388611004232398,
+      "grad_norm": 1.0456952798320092,
+      "learning_rate": 9.234355945780581e-06,
+      "loss": 0.8502,
+      "step": 2801
+    },
+    {
+      "epoch": 0.5390534821085032,
+      "grad_norm": 0.8580803051739965,
+      "learning_rate": 9.228143538398963e-06,
+      "loss": 0.8428,
+      "step": 2802
+    },
+    {
+      "epoch": 0.5392458637937668,
+      "grad_norm": 0.9358973732833039,
+      "learning_rate": 9.221931430678598e-06,
+      "loss": 0.811,
+      "step": 2803
+    },
+    {
+      "epoch": 0.5394382454790304,
+      "grad_norm": 0.9549101359772725,
+      "learning_rate": 9.215719625031245e-06,
+      "loss": 0.8621,
+      "step": 2804
+    },
+    {
+      "epoch": 0.539630627164294,
+      "grad_norm": 0.80382250255684,
+      "learning_rate": 9.209508123868534e-06,
+      "loss": 0.799,
+      "step": 2805
+    },
+    {
+      "epoch": 0.5398230088495575,
+      "grad_norm": 0.8549292936163121,
+      "learning_rate": 9.203296929601986e-06,
+      "loss": 0.7947,
+      "step": 2806
+    },
+    {
+      "epoch": 0.5400153905348211,
+      "grad_norm": 0.9229374123304265,
+      "learning_rate": 9.197086044643004e-06,
+      "loss": 0.7962,
+      "step": 2807
+    },
+    {
+      "epoch": 0.5402077722200846,
+      "grad_norm": 0.8840410592869056,
+      "learning_rate": 9.190875471402865e-06,
+      "loss": 0.8192,
+      "step": 2808
+    },
+    {
+      "epoch": 0.5404001539053482,
+      "grad_norm": 0.8234296274720394,
+      "learning_rate": 9.184665212292723e-06,
+      "loss": 0.81,
+      "step": 2809
+    },
+    {
+      "epoch": 0.5405925355906118,
+      "grad_norm": 0.6217875329947623,
+      "learning_rate": 9.178455269723623e-06,
+      "loss": 0.3481,
+      "step": 2810
+    },
+    {
+      "epoch": 0.5407849172758753,
+      "grad_norm": 0.8681117155029896,
+      "learning_rate": 9.172245646106471e-06,
+      "loss": 0.812,
+      "step": 2811
+    },
+    {
+      "epoch": 0.5409772989611389,
+      "grad_norm": 0.6350997021277557,
+      "learning_rate": 9.166036343852061e-06,
+      "loss": 0.3657,
+      "step": 2812
+    },
+    {
+      "epoch": 0.5411696806464025,
+      "grad_norm": 0.9372818196066508,
+      "learning_rate": 9.159827365371055e-06,
+      "loss": 0.8752,
+      "step": 2813
+    },
+    {
+      "epoch": 0.5413620623316661,
+      "grad_norm": 0.8390774792351325,
+      "learning_rate": 9.153618713073995e-06,
+      "loss": 0.8008,
+      "step": 2814
+    },
+    {
+      "epoch": 0.5415544440169295,
+      "grad_norm": 0.9615787295651008,
+      "learning_rate": 9.14741038937129e-06,
+      "loss": 0.8374,
+      "step": 2815
+    },
+    {
+      "epoch": 0.5417468257021931,
+      "grad_norm": 0.9968007268176491,
+      "learning_rate": 9.141202396673232e-06,
+      "loss": 0.8285,
+      "step": 2816
+    },
+    {
+      "epoch": 0.5419392073874567,
+      "grad_norm": 0.8638666630912483,
+      "learning_rate": 9.13499473738997e-06,
+      "loss": 0.8568,
+      "step": 2817
+    },
+    {
+      "epoch": 0.5421315890727203,
+      "grad_norm": 0.8489920786798096,
+      "learning_rate": 9.128787413931536e-06,
+      "loss": 0.8284,
+      "step": 2818
+    },
+    {
+      "epoch": 0.5423239707579839,
+      "grad_norm": 0.8607324870124281,
+      "learning_rate": 9.122580428707822e-06,
+      "loss": 0.7978,
+      "step": 2819
+    },
+    {
+      "epoch": 0.5425163524432474,
+      "grad_norm": 0.8673812494204222,
+      "learning_rate": 9.116373784128597e-06,
+      "loss": 0.7806,
+      "step": 2820
+    },
+    {
+      "epoch": 0.542708734128511,
+      "grad_norm": 0.8336582530449876,
+      "learning_rate": 9.110167482603494e-06,
+      "loss": 0.8566,
+      "step": 2821
+    },
+    {
+      "epoch": 0.5429011158137745,
+      "grad_norm": 0.865832732861329,
+      "learning_rate": 9.10396152654201e-06,
+      "loss": 0.8311,
+      "step": 2822
+    },
+    {
+      "epoch": 0.5430934974990381,
+      "grad_norm": 0.9035255812745883,
+      "learning_rate": 9.097755918353513e-06,
+      "loss": 0.839,
+      "step": 2823
+    },
+    {
+      "epoch": 0.5432858791843016,
+      "grad_norm": 0.8288249188884831,
+      "learning_rate": 9.091550660447236e-06,
+      "loss": 0.8419,
+      "step": 2824
+    },
+    {
+      "epoch": 0.5434782608695652,
+      "grad_norm": 0.8734159482400402,
+      "learning_rate": 9.08534575523227e-06,
+      "loss": 0.8055,
+      "step": 2825
+    },
+    {
+      "epoch": 0.5436706425548288,
+      "grad_norm": 0.9577511578689157,
+      "learning_rate": 9.079141205117573e-06,
+      "loss": 0.794,
+      "step": 2826
+    },
+    {
+      "epoch": 0.5438630242400924,
+      "grad_norm": 0.8992845904119022,
+      "learning_rate": 9.072937012511968e-06,
+      "loss": 0.8983,
+      "step": 2827
+    },
+    {
+      "epoch": 0.544055405925356,
+      "grad_norm": 0.8348026767050263,
+      "learning_rate": 9.066733179824134e-06,
+      "loss": 0.7158,
+      "step": 2828
+    },
+    {
+      "epoch": 0.5442477876106194,
+      "grad_norm": 0.8558564791872738,
+      "learning_rate": 9.060529709462613e-06,
+      "loss": 0.8487,
+      "step": 2829
+    },
+    {
+      "epoch": 0.544440169295883,
+      "grad_norm": 0.9101505728120213,
+      "learning_rate": 9.054326603835807e-06,
+      "loss": 0.8557,
+      "step": 2830
+    },
+    {
+      "epoch": 0.5446325509811466,
+      "grad_norm": 0.9762521478442119,
+      "learning_rate": 9.048123865351971e-06,
+      "loss": 0.8534,
+      "step": 2831
+    },
+    {
+      "epoch": 0.5448249326664102,
+      "grad_norm": 0.8766303398170076,
+      "learning_rate": 9.041921496419225e-06,
+      "loss": 0.8165,
+      "step": 2832
+    },
+    {
+      "epoch": 0.5450173143516738,
+      "grad_norm": 0.8859110261949145,
+      "learning_rate": 9.035719499445545e-06,
+      "loss": 0.7784,
+      "step": 2833
+    },
+    {
+      "epoch": 0.5452096960369373,
+      "grad_norm": 0.6937939033386182,
+      "learning_rate": 9.029517876838755e-06,
+      "loss": 0.3361,
+      "step": 2834
+    },
+    {
+      "epoch": 0.5454020777222008,
+      "grad_norm": 0.8493384007874506,
+      "learning_rate": 9.023316631006536e-06,
+      "loss": 0.8319,
+      "step": 2835
+    },
+    {
+      "epoch": 0.5455944594074644,
+      "grad_norm": 0.642261501278718,
+      "learning_rate": 9.017115764356436e-06,
+      "loss": 0.3522,
+      "step": 2836
+    },
+    {
+      "epoch": 0.545786841092728,
+      "grad_norm": 0.8196918248416661,
+      "learning_rate": 9.010915279295833e-06,
+      "loss": 0.8055,
+      "step": 2837
+    },
+    {
+      "epoch": 0.5459792227779915,
+      "grad_norm": 0.6412829905781539,
+      "learning_rate": 9.004715178231975e-06,
+      "loss": 0.3288,
+      "step": 2838
+    },
+    {
+      "epoch": 0.5461716044632551,
+      "grad_norm": 0.8199181793176952,
+      "learning_rate": 8.998515463571953e-06,
+      "loss": 0.7842,
+      "step": 2839
+    },
+    {
+      "epoch": 0.5463639861485187,
+      "grad_norm": 0.879820963519553,
+      "learning_rate": 8.992316137722711e-06,
+      "loss": 0.7861,
+      "step": 2840
+    },
+    {
+      "epoch": 0.5465563678337823,
+      "grad_norm": 0.805344838741774,
+      "learning_rate": 8.986117203091042e-06,
+      "loss": 0.8717,
+      "step": 2841
+    },
+    {
+      "epoch": 0.5467487495190458,
+      "grad_norm": 0.8509414631430218,
+      "learning_rate": 8.97991866208358e-06,
+      "loss": 0.7739,
+      "step": 2842
+    },
+    {
+      "epoch": 0.5469411312043093,
+      "grad_norm": 0.8741406514247216,
+      "learning_rate": 8.973720517106814e-06,
+      "loss": 0.8647,
+      "step": 2843
+    },
+    {
+      "epoch": 0.5471335128895729,
+      "grad_norm": 0.8425875902021079,
+      "learning_rate": 8.967522770567086e-06,
+      "loss": 0.79,
+      "step": 2844
+    },
+    {
+      "epoch": 0.5473258945748365,
+      "grad_norm": 0.9223386157020266,
+      "learning_rate": 8.961325424870561e-06,
+      "loss": 0.8708,
+      "step": 2845
+    },
+    {
+      "epoch": 0.5475182762601001,
+      "grad_norm": 0.8702189950112492,
+      "learning_rate": 8.955128482423271e-06,
+      "loss": 0.8668,
+      "step": 2846
+    },
+    {
+      "epoch": 0.5477106579453636,
+      "grad_norm": 0.8811895540000872,
+      "learning_rate": 8.948931945631082e-06,
+      "loss": 0.8683,
+      "step": 2847
+    },
+    {
+      "epoch": 0.5479030396306271,
+      "grad_norm": 0.874104060062377,
+      "learning_rate": 8.9427358168997e-06,
+      "loss": 0.786,
+      "step": 2848
+    },
+    {
+      "epoch": 0.5480954213158907,
+      "grad_norm": 0.8705185327466095,
+      "learning_rate": 8.936540098634675e-06,
+      "loss": 0.7735,
+      "step": 2849
+    },
+    {
+      "epoch": 0.5482878030011543,
+      "grad_norm": 0.8351265827724964,
+      "learning_rate": 8.930344793241404e-06,
+      "loss": 0.7892,
+      "step": 2850
+    },
+    {
+      "epoch": 0.5484801846864179,
+      "grad_norm": 0.9626729804469332,
+      "learning_rate": 8.924149903125108e-06,
+      "loss": 0.8332,
+      "step": 2851
+    },
+    {
+      "epoch": 0.5486725663716814,
+      "grad_norm": 0.8205091517479032,
+      "learning_rate": 8.917955430690865e-06,
+      "loss": 0.8024,
+      "step": 2852
+    },
+    {
+      "epoch": 0.548864948056945,
+      "grad_norm": 0.7926701692980015,
+      "learning_rate": 8.91176137834358e-06,
+      "loss": 0.7917,
+      "step": 2853
+    },
+    {
+      "epoch": 0.5490573297422086,
+      "grad_norm": 0.9306866963571934,
+      "learning_rate": 8.905567748487997e-06,
+      "loss": 0.8341,
+      "step": 2854
+    },
+    {
+      "epoch": 0.5492497114274721,
+      "grad_norm": 0.6482004314882388,
+      "learning_rate": 8.899374543528695e-06,
+      "loss": 0.3702,
+      "step": 2855
+    },
+    {
+      "epoch": 0.5494420931127356,
+      "grad_norm": 0.8702306188491451,
+      "learning_rate": 8.893181765870094e-06,
+      "loss": 0.7979,
+      "step": 2856
+    },
+    {
+      "epoch": 0.5496344747979992,
+      "grad_norm": 0.8729559442520741,
+      "learning_rate": 8.886989417916435e-06,
+      "loss": 0.7966,
+      "step": 2857
+    },
+    {
+      "epoch": 0.5498268564832628,
+      "grad_norm": 1.0387854569844264,
+      "learning_rate": 8.88079750207181e-06,
+      "loss": 0.8062,
+      "step": 2858
+    },
+    {
+      "epoch": 0.5500192381685264,
+      "grad_norm": 0.8494954516928321,
+      "learning_rate": 8.87460602074013e-06,
+      "loss": 0.7748,
+      "step": 2859
+    },
+    {
+      "epoch": 0.55021161985379,
+      "grad_norm": 0.8341741387129651,
+      "learning_rate": 8.86841497632514e-06,
+      "loss": 0.7479,
+      "step": 2860
+    },
+    {
+      "epoch": 0.5504040015390534,
+      "grad_norm": 0.8727652903763965,
+      "learning_rate": 8.862224371230418e-06,
+      "loss": 0.7771,
+      "step": 2861
+    },
+    {
+      "epoch": 0.550596383224317,
+      "grad_norm": 0.8856058722325291,
+      "learning_rate": 8.85603420785937e-06,
+      "loss": 0.8926,
+      "step": 2862
+    },
+    {
+      "epoch": 0.5507887649095806,
+      "grad_norm": 0.9084756805639895,
+      "learning_rate": 8.84984448861523e-06,
+      "loss": 0.7856,
+      "step": 2863
+    },
+    {
+      "epoch": 0.5509811465948442,
+      "grad_norm": 0.8136898515958757,
+      "learning_rate": 8.84365521590106e-06,
+      "loss": 0.7398,
+      "step": 2864
+    },
+    {
+      "epoch": 0.5511735282801077,
+      "grad_norm": 0.8133245951004003,
+      "learning_rate": 8.837466392119752e-06,
+      "loss": 0.7806,
+      "step": 2865
+    },
+    {
+      "epoch": 0.5513659099653713,
+      "grad_norm": 0.8427982346732021,
+      "learning_rate": 8.831278019674017e-06,
+      "loss": 0.7943,
+      "step": 2866
+    },
+    {
+      "epoch": 0.5515582916506349,
+      "grad_norm": 0.8887551495174795,
+      "learning_rate": 8.825090100966396e-06,
+      "loss": 0.8055,
+      "step": 2867
+    },
+    {
+      "epoch": 0.5517506733358984,
+      "grad_norm": 0.8499383294907575,
+      "learning_rate": 8.818902638399247e-06,
+      "loss": 0.84,
+      "step": 2868
+    },
+    {
+      "epoch": 0.551943055021162,
+      "grad_norm": 0.8781743509794816,
+      "learning_rate": 8.81271563437476e-06,
+      "loss": 0.7928,
+      "step": 2869
+    },
+    {
+      "epoch": 0.5521354367064255,
+      "grad_norm": 0.8435142052666665,
+      "learning_rate": 8.806529091294948e-06,
+      "loss": 0.8491,
+      "step": 2870
+    },
+    {
+      "epoch": 0.5523278183916891,
+      "grad_norm": 0.8823008580180244,
+      "learning_rate": 8.800343011561633e-06,
+      "loss": 0.8179,
+      "step": 2871
+    },
+    {
+      "epoch": 0.5525202000769527,
+      "grad_norm": 0.7186095231812679,
+      "learning_rate": 8.794157397576464e-06,
+      "loss": 0.3466,
+      "step": 2872
+    },
+    {
+      "epoch": 0.5527125817622163,
+      "grad_norm": 0.8417676415566809,
+      "learning_rate": 8.787972251740916e-06,
+      "loss": 0.7741,
+      "step": 2873
+    },
+    {
+      "epoch": 0.5529049634474797,
+      "grad_norm": 0.8999957420500337,
+      "learning_rate": 8.781787576456269e-06,
+      "loss": 0.8208,
+      "step": 2874
+    },
+    {
+      "epoch": 0.5530973451327433,
+      "grad_norm": 0.830040222008236,
+      "learning_rate": 8.775603374123627e-06,
+      "loss": 0.7959,
+      "step": 2875
+    },
+    {
+      "epoch": 0.5532897268180069,
+      "grad_norm": 0.8671830893129526,
+      "learning_rate": 8.769419647143917e-06,
+      "loss": 0.7998,
+      "step": 2876
+    },
+    {
+      "epoch": 0.5534821085032705,
+      "grad_norm": 0.8747213353584236,
+      "learning_rate": 8.763236397917865e-06,
+      "loss": 0.8165,
+      "step": 2877
+    },
+    {
+      "epoch": 0.5536744901885341,
+      "grad_norm": 0.8564810285642023,
+      "learning_rate": 8.757053628846028e-06,
+      "loss": 0.8005,
+      "step": 2878
+    },
+    {
+      "epoch": 0.5538668718737976,
+      "grad_norm": 0.8979865160935562,
+      "learning_rate": 8.75087134232877e-06,
+      "loss": 0.8151,
+      "step": 2879
+    },
+    {
+      "epoch": 0.5540592535590612,
+      "grad_norm": 0.8882685001335168,
+      "learning_rate": 8.744689540766265e-06,
+      "loss": 0.8245,
+      "step": 2880
+    },
+    {
+      "epoch": 0.5542516352443247,
+      "grad_norm": 0.8489060719221071,
+      "learning_rate": 8.738508226558499e-06,
+      "loss": 0.7997,
+      "step": 2881
+    },
+    {
+      "epoch": 0.5544440169295883,
+      "grad_norm": 0.8793772142595864,
+      "learning_rate": 8.73232740210528e-06,
+      "loss": 0.8254,
+      "step": 2882
+    },
+    {
+      "epoch": 0.5546363986148518,
+      "grad_norm": 0.8317902127777473,
+      "learning_rate": 8.726147069806206e-06,
+      "loss": 0.818,
+      "step": 2883
+    },
+    {
+      "epoch": 0.5548287803001154,
+      "grad_norm": 0.8498836310377664,
+      "learning_rate": 8.719967232060698e-06,
+      "loss": 0.7484,
+      "step": 2884
+    },
+    {
+      "epoch": 0.555021161985379,
+      "grad_norm": 0.9183673735416015,
+      "learning_rate": 8.713787891267988e-06,
+      "loss": 0.8143,
+      "step": 2885
+    },
+    {
+      "epoch": 0.5552135436706426,
+      "grad_norm": 0.8767596883280362,
+      "learning_rate": 8.707609049827102e-06,
+      "loss": 0.7967,
+      "step": 2886
+    },
+    {
+      "epoch": 0.5554059253559062,
+      "grad_norm": 0.9219488016971755,
+      "learning_rate": 8.70143071013688e-06,
+      "loss": 0.7732,
+      "step": 2887
+    },
+    {
+      "epoch": 0.5555983070411696,
+      "grad_norm": 0.8868462028051292,
+      "learning_rate": 8.695252874595972e-06,
+      "loss": 0.7797,
+      "step": 2888
+    },
+    {
+      "epoch": 0.5557906887264332,
+      "grad_norm": 0.7692697090271274,
+      "learning_rate": 8.689075545602816e-06,
+      "loss": 0.3718,
+      "step": 2889
+    },
+    {
+      "epoch": 0.5559830704116968,
+      "grad_norm": 0.8723266082852207,
+      "learning_rate": 8.68289872555567e-06,
+      "loss": 0.8294,
+      "step": 2890
+    },
+    {
+      "epoch": 0.5561754520969604,
+      "grad_norm": 0.8665008257428618,
+      "learning_rate": 8.676722416852594e-06,
+      "loss": 0.8036,
+      "step": 2891
+    },
+    {
+      "epoch": 0.556367833782224,
+      "grad_norm": 0.8930957395383884,
+      "learning_rate": 8.670546621891434e-06,
+      "loss": 0.8355,
+      "step": 2892
+    },
+    {
+      "epoch": 0.5565602154674875,
+      "grad_norm": 0.7900605338249951,
+      "learning_rate": 8.66437134306985e-06,
+      "loss": 0.7838,
+      "step": 2893
+    },
+    {
+      "epoch": 0.556752597152751,
+      "grad_norm": 0.8847060853052108,
+      "learning_rate": 8.658196582785297e-06,
+      "loss": 0.8154,
+      "step": 2894
+    },
+    {
+      "epoch": 0.5569449788380146,
+      "grad_norm": 0.944282746449644,
+      "learning_rate": 8.652022343435027e-06,
+      "loss": 0.812,
+      "step": 2895
+    },
+    {
+      "epoch": 0.5571373605232782,
+      "grad_norm": 0.8759416081019398,
+      "learning_rate": 8.645848627416102e-06,
+      "loss": 0.8139,
+      "step": 2896
+    },
+    {
+      "epoch": 0.5573297422085417,
+      "grad_norm": 0.892750746253494,
+      "learning_rate": 8.63967543712536e-06,
+      "loss": 0.8271,
+      "step": 2897
+    },
+    {
+      "epoch": 0.5575221238938053,
+      "grad_norm": 0.9375628511824584,
+      "learning_rate": 8.633502774959453e-06,
+      "loss": 0.8131,
+      "step": 2898
+    },
+    {
+      "epoch": 0.5577145055790689,
+      "grad_norm": 0.8591897605837595,
+      "learning_rate": 8.627330643314818e-06,
+      "loss": 0.792,
+      "step": 2899
+    },
+    {
+      "epoch": 0.5579068872643325,
+      "grad_norm": 0.894152797446417,
+      "learning_rate": 8.62115904458769e-06,
+      "loss": 0.7767,
+      "step": 2900
+    },
+    {
+      "epoch": 0.558099268949596,
+      "grad_norm": 0.9403817961205942,
+      "learning_rate": 8.614987981174093e-06,
+      "loss": 0.812,
+      "step": 2901
+    },
+    {
+      "epoch": 0.5582916506348595,
+      "grad_norm": 0.9083064540321489,
+      "learning_rate": 8.608817455469854e-06,
+      "loss": 0.8913,
+      "step": 2902
+    },
+    {
+      "epoch": 0.5584840323201231,
+      "grad_norm": 0.8504884779082656,
+      "learning_rate": 8.602647469870573e-06,
+      "loss": 0.8694,
+      "step": 2903
+    },
+    {
+      "epoch": 0.5586764140053867,
+      "grad_norm": 0.8924980663995151,
+      "learning_rate": 8.596478026771658e-06,
+      "loss": 0.8346,
+      "step": 2904
+    },
+    {
+      "epoch": 0.5588687956906503,
+      "grad_norm": 0.8659611815392535,
+      "learning_rate": 8.590309128568303e-06,
+      "loss": 0.8548,
+      "step": 2905
+    },
+    {
+      "epoch": 0.5590611773759138,
+      "grad_norm": 0.8140034012303768,
+      "learning_rate": 8.584140777655476e-06,
+      "loss": 0.7996,
+      "step": 2906
+    },
+    {
+      "epoch": 0.5592535590611774,
+      "grad_norm": 0.8503241663101949,
+      "learning_rate": 8.57797297642795e-06,
+      "loss": 0.8145,
+      "step": 2907
+    },
+    {
+      "epoch": 0.5594459407464409,
+      "grad_norm": 0.8065933401671029,
+      "learning_rate": 8.571805727280278e-06,
+      "loss": 0.8491,
+      "step": 2908
+    },
+    {
+      "epoch": 0.5596383224317045,
+      "grad_norm": 0.9272634166110356,
+      "learning_rate": 8.565639032606794e-06,
+      "loss": 0.7879,
+      "step": 2909
+    },
+    {
+      "epoch": 0.5598307041169681,
+      "grad_norm": 0.8974298878580199,
+      "learning_rate": 8.559472894801623e-06,
+      "loss": 0.7824,
+      "step": 2910
+    },
+    {
+      "epoch": 0.5600230858022316,
+      "grad_norm": 0.8507912884973065,
+      "learning_rate": 8.553307316258678e-06,
+      "loss": 0.7776,
+      "step": 2911
+    },
+    {
+      "epoch": 0.5602154674874952,
+      "grad_norm": 0.862134921703722,
+      "learning_rate": 8.547142299371642e-06,
+      "loss": 0.8245,
+      "step": 2912
+    },
+    {
+      "epoch": 0.5604078491727588,
+      "grad_norm": 0.8520155678099587,
+      "learning_rate": 8.540977846533986e-06,
+      "loss": 0.8128,
+      "step": 2913
+    },
+    {
+      "epoch": 0.5606002308580224,
+      "grad_norm": 0.8473403281866664,
+      "learning_rate": 8.534813960138968e-06,
+      "loss": 0.7914,
+      "step": 2914
+    },
+    {
+      "epoch": 0.5607926125432858,
+      "grad_norm": 0.8937992695332417,
+      "learning_rate": 8.528650642579618e-06,
+      "loss": 0.8547,
+      "step": 2915
+    },
+    {
+      "epoch": 0.5609849942285494,
+      "grad_norm": 0.9518876941117348,
+      "learning_rate": 8.52248789624875e-06,
+      "loss": 0.9025,
+      "step": 2916
+    },
+    {
+      "epoch": 0.561177375913813,
+      "grad_norm": 0.9127686087069266,
+      "learning_rate": 8.516325723538949e-06,
+      "loss": 0.7958,
+      "step": 2917
+    },
+    {
+      "epoch": 0.5613697575990766,
+      "grad_norm": 0.8827520287483926,
+      "learning_rate": 8.510164126842591e-06,
+      "loss": 0.8522,
+      "step": 2918
+    },
+    {
+      "epoch": 0.5615621392843402,
+      "grad_norm": 0.8588738156986986,
+      "learning_rate": 8.504003108551814e-06,
+      "loss": 0.8649,
+      "step": 2919
+    },
+    {
+      "epoch": 0.5617545209696037,
+      "grad_norm": 0.9240155872107624,
+      "learning_rate": 8.497842671058539e-06,
+      "loss": 0.7976,
+      "step": 2920
+    },
+    {
+      "epoch": 0.5619469026548672,
+      "grad_norm": 0.8767305315888647,
+      "learning_rate": 8.491682816754456e-06,
+      "loss": 0.8611,
+      "step": 2921
+    },
+    {
+      "epoch": 0.5621392843401308,
+      "grad_norm": 0.8899357119313409,
+      "learning_rate": 8.485523548031044e-06,
+      "loss": 0.8396,
+      "step": 2922
+    },
+    {
+      "epoch": 0.5623316660253944,
+      "grad_norm": 0.8921693590507347,
+      "learning_rate": 8.479364867279529e-06,
+      "loss": 0.8351,
+      "step": 2923
+    },
+    {
+      "epoch": 0.5625240477106579,
+      "grad_norm": 0.8891168607593513,
+      "learning_rate": 8.47320677689093e-06,
+      "loss": 0.8079,
+      "step": 2924
+    },
+    {
+      "epoch": 0.5627164293959215,
+      "grad_norm": 0.902884421474475,
+      "learning_rate": 8.467049279256034e-06,
+      "loss": 0.8779,
+      "step": 2925
+    },
+    {
+      "epoch": 0.5629088110811851,
+      "grad_norm": 0.8339869638520038,
+      "learning_rate": 8.460892376765387e-06,
+      "loss": 0.7761,
+      "step": 2926
+    },
+    {
+      "epoch": 0.5631011927664487,
+      "grad_norm": 0.8966456851415097,
+      "learning_rate": 8.45473607180931e-06,
+      "loss": 0.83,
+      "step": 2927
+    },
+    {
+      "epoch": 0.5632935744517122,
+      "grad_norm": 0.8784961966237712,
+      "learning_rate": 8.448580366777898e-06,
+      "loss": 0.8462,
+      "step": 2928
+    },
+    {
+      "epoch": 0.5634859561369757,
+      "grad_norm": 0.8894734042501298,
+      "learning_rate": 8.442425264061e-06,
+      "loss": 0.7891,
+      "step": 2929
+    },
+    {
+      "epoch": 0.5636783378222393,
+      "grad_norm": 0.9168376138250717,
+      "learning_rate": 8.436270766048245e-06,
+      "loss": 0.8371,
+      "step": 2930
+    },
+    {
+      "epoch": 0.5638707195075029,
+      "grad_norm": 0.8185067180292277,
+      "learning_rate": 8.430116875129023e-06,
+      "loss": 0.8072,
+      "step": 2931
+    },
+    {
+      "epoch": 0.5640631011927665,
+      "grad_norm": 0.927032644864294,
+      "learning_rate": 8.42396359369248e-06,
+      "loss": 0.8629,
+      "step": 2932
+    },
+    {
+      "epoch": 0.56425548287803,
+      "grad_norm": 0.8595119361111665,
+      "learning_rate": 8.417810924127533e-06,
+      "loss": 0.8444,
+      "step": 2933
+    },
+    {
+      "epoch": 0.5644478645632935,
+      "grad_norm": 0.8107893074664902,
+      "learning_rate": 8.411658868822866e-06,
+      "loss": 0.7704,
+      "step": 2934
+    },
+    {
+      "epoch": 0.5646402462485571,
+      "grad_norm": 0.9169441147030931,
+      "learning_rate": 8.40550743016691e-06,
+      "loss": 0.7954,
+      "step": 2935
+    },
+    {
+      "epoch": 0.5648326279338207,
+      "grad_norm": 0.8021812226086812,
+      "learning_rate": 8.39935661054787e-06,
+      "loss": 0.8011,
+      "step": 2936
+    },
+    {
+      "epoch": 0.5650250096190843,
+      "grad_norm": 0.8184078814416893,
+      "learning_rate": 8.393206412353709e-06,
+      "loss": 0.8132,
+      "step": 2937
+    },
+    {
+      "epoch": 0.5652173913043478,
+      "grad_norm": 0.915766413400768,
+      "learning_rate": 8.38705683797214e-06,
+      "loss": 0.7658,
+      "step": 2938
+    },
+    {
+      "epoch": 0.5654097729896114,
+      "grad_norm": 0.9056635099443426,
+      "learning_rate": 8.38090788979064e-06,
+      "loss": 0.7566,
+      "step": 2939
+    },
+    {
+      "epoch": 0.565602154674875,
+      "grad_norm": 0.9187098097361955,
+      "learning_rate": 8.374759570196448e-06,
+      "loss": 0.8081,
+      "step": 2940
+    },
+    {
+      "epoch": 0.5657945363601385,
+      "grad_norm": 0.8634275094982933,
+      "learning_rate": 8.368611881576547e-06,
+      "loss": 0.8011,
+      "step": 2941
+    },
+    {
+      "epoch": 0.5659869180454021,
+      "grad_norm": 0.9065848122220095,
+      "learning_rate": 8.362464826317687e-06,
+      "loss": 0.8219,
+      "step": 2942
+    },
+    {
+      "epoch": 0.5661792997306656,
+      "grad_norm": 0.8237327890713795,
+      "learning_rate": 8.35631840680636e-06,
+      "loss": 0.8179,
+      "step": 2943
+    },
+    {
+      "epoch": 0.5663716814159292,
+      "grad_norm": 0.8998423027626428,
+      "learning_rate": 8.35017262542882e-06,
+      "loss": 0.8352,
+      "step": 2944
+    },
+    {
+      "epoch": 0.5665640631011928,
+      "grad_norm": 0.8623712382754144,
+      "learning_rate": 8.344027484571075e-06,
+      "loss": 0.7689,
+      "step": 2945
+    },
+    {
+      "epoch": 0.5667564447864564,
+      "grad_norm": 0.9140707425639395,
+      "learning_rate": 8.337882986618877e-06,
+      "loss": 0.7916,
+      "step": 2946
+    },
+    {
+      "epoch": 0.5669488264717198,
+      "grad_norm": 0.9393862936153426,
+      "learning_rate": 8.331739133957729e-06,
+      "loss": 0.7912,
+      "step": 2947
+    },
+    {
+      "epoch": 0.5671412081569834,
+      "grad_norm": 0.7973020205340585,
+      "learning_rate": 8.325595928972894e-06,
+      "loss": 0.8134,
+      "step": 2948
+    },
+    {
+      "epoch": 0.567333589842247,
+      "grad_norm": 0.8921511140968901,
+      "learning_rate": 8.319453374049367e-06,
+      "loss": 0.784,
+      "step": 2949
+    },
+    {
+      "epoch": 0.5675259715275106,
+      "grad_norm": 0.7975823512564189,
+      "learning_rate": 8.313311471571903e-06,
+      "loss": 0.7641,
+      "step": 2950
+    },
+    {
+      "epoch": 0.5677183532127742,
+      "grad_norm": 0.8796845714232029,
+      "learning_rate": 8.307170223925003e-06,
+      "loss": 0.8102,
+      "step": 2951
+    },
+    {
+      "epoch": 0.5679107348980377,
+      "grad_norm": 0.8283669973216768,
+      "learning_rate": 8.301029633492907e-06,
+      "loss": 0.8438,
+      "step": 2952
+    },
+    {
+      "epoch": 0.5681031165833013,
+      "grad_norm": 0.8194256359009269,
+      "learning_rate": 8.294889702659602e-06,
+      "loss": 0.8334,
+      "step": 2953
+    },
+    {
+      "epoch": 0.5682954982685648,
+      "grad_norm": 0.8423771799907173,
+      "learning_rate": 8.288750433808828e-06,
+      "loss": 0.8412,
+      "step": 2954
+    },
+    {
+      "epoch": 0.5684878799538284,
+      "grad_norm": 0.8446277169971869,
+      "learning_rate": 8.282611829324049e-06,
+      "loss": 0.8036,
+      "step": 2955
+    },
+    {
+      "epoch": 0.5686802616390919,
+      "grad_norm": 0.9951621326064969,
+      "learning_rate": 8.276473891588492e-06,
+      "loss": 0.8397,
+      "step": 2956
+    },
+    {
+      "epoch": 0.5688726433243555,
+      "grad_norm": 0.8567133496096431,
+      "learning_rate": 8.270336622985116e-06,
+      "loss": 0.8462,
+      "step": 2957
+    },
+    {
+      "epoch": 0.5690650250096191,
+      "grad_norm": 0.8728521513961071,
+      "learning_rate": 8.264200025896616e-06,
+      "loss": 0.8174,
+      "step": 2958
+    },
+    {
+      "epoch": 0.5692574066948827,
+      "grad_norm": 0.8770090657061295,
+      "learning_rate": 8.258064102705428e-06,
+      "loss": 0.8443,
+      "step": 2959
+    },
+    {
+      "epoch": 0.5694497883801463,
+      "grad_norm": 0.8610807764013058,
+      "learning_rate": 8.251928855793736e-06,
+      "loss": 0.7818,
+      "step": 2960
+    },
+    {
+      "epoch": 0.5696421700654097,
+      "grad_norm": 0.8274002327919278,
+      "learning_rate": 8.245794287543447e-06,
+      "loss": 0.8032,
+      "step": 2961
+    },
+    {
+      "epoch": 0.5698345517506733,
+      "grad_norm": 0.9383272110822968,
+      "learning_rate": 8.239660400336213e-06,
+      "loss": 0.8581,
+      "step": 2962
+    },
+    {
+      "epoch": 0.5700269334359369,
+      "grad_norm": 0.8299400305590873,
+      "learning_rate": 8.233527196553428e-06,
+      "loss": 0.792,
+      "step": 2963
+    },
+    {
+      "epoch": 0.5702193151212005,
+      "grad_norm": 0.8307491966516248,
+      "learning_rate": 8.227394678576204e-06,
+      "loss": 0.8106,
+      "step": 2964
+    },
+    {
+      "epoch": 0.570411696806464,
+      "grad_norm": 0.9266222672767408,
+      "learning_rate": 8.221262848785395e-06,
+      "loss": 0.8724,
+      "step": 2965
+    },
+    {
+      "epoch": 0.5706040784917276,
+      "grad_norm": 0.8499956168676399,
+      "learning_rate": 8.215131709561597e-06,
+      "loss": 0.7919,
+      "step": 2966
+    },
+    {
+      "epoch": 0.5707964601769911,
+      "grad_norm": 0.9025414888254327,
+      "learning_rate": 8.20900126328512e-06,
+      "loss": 0.8111,
+      "step": 2967
+    },
+    {
+      "epoch": 0.5709888418622547,
+      "grad_norm": 0.871423515933429,
+      "learning_rate": 8.202871512336023e-06,
+      "loss": 0.8558,
+      "step": 2968
+    },
+    {
+      "epoch": 0.5711812235475183,
+      "grad_norm": 0.694236534714168,
+      "learning_rate": 8.196742459094079e-06,
+      "loss": 0.3506,
+      "step": 2969
+    },
+    {
+      "epoch": 0.5713736052327818,
+      "grad_norm": 0.9157208539728942,
+      "learning_rate": 8.190614105938796e-06,
+      "loss": 0.7936,
+      "step": 2970
+    },
+    {
+      "epoch": 0.5715659869180454,
+      "grad_norm": 0.8491290624061908,
+      "learning_rate": 8.184486455249424e-06,
+      "loss": 0.8275,
+      "step": 2971
+    },
+    {
+      "epoch": 0.571758368603309,
+      "grad_norm": 0.8286881456066588,
+      "learning_rate": 8.178359509404916e-06,
+      "loss": 0.7806,
+      "step": 2972
+    },
+    {
+      "epoch": 0.5719507502885726,
+      "grad_norm": 0.8353722791372507,
+      "learning_rate": 8.172233270783966e-06,
+      "loss": 0.7976,
+      "step": 2973
+    },
+    {
+      "epoch": 0.572143131973836,
+      "grad_norm": 0.9325510405931856,
+      "learning_rate": 8.166107741764997e-06,
+      "loss": 0.8047,
+      "step": 2974
+    },
+    {
+      "epoch": 0.5723355136590996,
+      "grad_norm": 0.9806444797040953,
+      "learning_rate": 8.15998292472614e-06,
+      "loss": 0.8263,
+      "step": 2975
+    },
+    {
+      "epoch": 0.5725278953443632,
+      "grad_norm": 0.9131216352872054,
+      "learning_rate": 8.153858822045267e-06,
+      "loss": 0.7924,
+      "step": 2976
+    },
+    {
+      "epoch": 0.5727202770296268,
+      "grad_norm": 0.8755001643530459,
+      "learning_rate": 8.147735436099967e-06,
+      "loss": 0.8277,
+      "step": 2977
+    },
+    {
+      "epoch": 0.5729126587148904,
+      "grad_norm": 0.8638046216392907,
+      "learning_rate": 8.141612769267543e-06,
+      "loss": 0.7986,
+      "step": 2978
+    },
+    {
+      "epoch": 0.5731050404001539,
+      "grad_norm": 0.9142352429178533,
+      "learning_rate": 8.135490823925027e-06,
+      "loss": 0.8217,
+      "step": 2979
+    },
+    {
+      "epoch": 0.5732974220854175,
+      "grad_norm": 0.820096169152406,
+      "learning_rate": 8.129369602449176e-06,
+      "loss": 0.7857,
+      "step": 2980
+    },
+    {
+      "epoch": 0.573489803770681,
+      "grad_norm": 0.8299766869484221,
+      "learning_rate": 8.123249107216446e-06,
+      "loss": 0.7991,
+      "step": 2981
+    },
+    {
+      "epoch": 0.5736821854559446,
+      "grad_norm": 0.8980635778317129,
+      "learning_rate": 8.117129340603032e-06,
+      "loss": 0.7785,
+      "step": 2982
+    },
+    {
+      "epoch": 0.5738745671412081,
+      "grad_norm": 0.9143863239259215,
+      "learning_rate": 8.111010304984841e-06,
+      "loss": 0.7391,
+      "step": 2983
+    },
+    {
+      "epoch": 0.5740669488264717,
+      "grad_norm": 0.7955942141798994,
+      "learning_rate": 8.104892002737488e-06,
+      "loss": 0.8027,
+      "step": 2984
+    },
+    {
+      "epoch": 0.5742593305117353,
+      "grad_norm": 0.9189142818345554,
+      "learning_rate": 8.098774436236308e-06,
+      "loss": 0.836,
+      "step": 2985
+    },
+    {
+      "epoch": 0.5744517121969989,
+      "grad_norm": 0.8052846917015606,
+      "learning_rate": 8.092657607856356e-06,
+      "loss": 0.3218,
+      "step": 2986
+    },
+    {
+      "epoch": 0.5746440938822625,
+      "grad_norm": 0.8582784860992996,
+      "learning_rate": 8.086541519972388e-06,
+      "loss": 0.8133,
+      "step": 2987
+    },
+    {
+      "epoch": 0.5748364755675259,
+      "grad_norm": 0.8519074747313994,
+      "learning_rate": 8.080426174958886e-06,
+      "loss": 0.8156,
+      "step": 2988
+    },
+    {
+      "epoch": 0.5750288572527895,
+      "grad_norm": 0.8232395679497525,
+      "learning_rate": 8.074311575190039e-06,
+      "loss": 0.7889,
+      "step": 2989
+    },
+    {
+      "epoch": 0.5752212389380531,
+      "grad_norm": 0.657730138571481,
+      "learning_rate": 8.068197723039738e-06,
+      "loss": 0.3593,
+      "step": 2990
+    },
+    {
+      "epoch": 0.5754136206233167,
+      "grad_norm": 0.6455559181757063,
+      "learning_rate": 8.062084620881598e-06,
+      "loss": 0.3302,
+      "step": 2991
+    },
+    {
+      "epoch": 0.5756060023085803,
+      "grad_norm": 0.8560749259395606,
+      "learning_rate": 8.055972271088933e-06,
+      "loss": 0.7595,
+      "step": 2992
+    },
+    {
+      "epoch": 0.5757983839938438,
+      "grad_norm": 0.8990237561867473,
+      "learning_rate": 8.049860676034762e-06,
+      "loss": 0.7735,
+      "step": 2993
+    },
+    {
+      "epoch": 0.5759907656791073,
+      "grad_norm": 0.8444606061887701,
+      "learning_rate": 8.043749838091828e-06,
+      "loss": 0.8111,
+      "step": 2994
+    },
+    {
+      "epoch": 0.5761831473643709,
+      "grad_norm": 0.8824426228954376,
+      "learning_rate": 8.037639759632558e-06,
+      "loss": 0.8477,
+      "step": 2995
+    },
+    {
+      "epoch": 0.5763755290496345,
+      "grad_norm": 0.7878687974964536,
+      "learning_rate": 8.031530443029099e-06,
+      "loss": 0.8088,
+      "step": 2996
+    },
+    {
+      "epoch": 0.576567910734898,
+      "grad_norm": 0.832787185927806,
+      "learning_rate": 8.025421890653303e-06,
+      "loss": 0.7807,
+      "step": 2997
+    },
+    {
+      "epoch": 0.5767602924201616,
+      "grad_norm": 0.8410399745600443,
+      "learning_rate": 8.019314104876712e-06,
+      "loss": 0.8227,
+      "step": 2998
+    },
+    {
+      "epoch": 0.5769526741054252,
+      "grad_norm": 0.6327189774609113,
+      "learning_rate": 8.013207088070582e-06,
+      "loss": 0.3418,
+      "step": 2999
+    },
+    {
+      "epoch": 0.5771450557906888,
+      "grad_norm": 0.925926999299207,
+      "learning_rate": 8.007100842605872e-06,
+      "loss": 0.7657,
+      "step": 3000
+    },
+    {
+      "epoch": 0.5773374374759523,
+      "grad_norm": 0.8420320998023307,
+      "learning_rate": 8.000995370853227e-06,
+      "loss": 0.8102,
+      "step": 3001
+    },
+    {
+      "epoch": 0.5775298191612158,
+      "grad_norm": 0.787423239745502,
+      "learning_rate": 7.994890675183008e-06,
+      "loss": 0.7412,
+      "step": 3002
+    },
+    {
+      "epoch": 0.5777222008464794,
+      "grad_norm": 0.6389813030496694,
+      "learning_rate": 7.98878675796527e-06,
+      "loss": 0.351,
+      "step": 3003
+    },
+    {
+      "epoch": 0.577914582531743,
+      "grad_norm": 0.9762098544843392,
+      "learning_rate": 7.98268362156976e-06,
+      "loss": 0.8724,
+      "step": 3004
+    },
+    {
+      "epoch": 0.5781069642170066,
+      "grad_norm": 0.8444748818826158,
+      "learning_rate": 7.976581268365924e-06,
+      "loss": 0.7975,
+      "step": 3005
+    },
+    {
+      "epoch": 0.5782993459022701,
+      "grad_norm": 0.9134300607670007,
+      "learning_rate": 7.97047970072291e-06,
+      "loss": 0.815,
+      "step": 3006
+    },
+    {
+      "epoch": 0.5784917275875336,
+      "grad_norm": 0.9635213755049868,
+      "learning_rate": 7.964378921009552e-06,
+      "loss": 0.8035,
+      "step": 3007
+    },
+    {
+      "epoch": 0.5786841092727972,
+      "grad_norm": 0.8622853565481334,
+      "learning_rate": 7.958278931594385e-06,
+      "loss": 0.8368,
+      "step": 3008
+    },
+    {
+      "epoch": 0.5788764909580608,
+      "grad_norm": 0.828199614928027,
+      "learning_rate": 7.952179734845642e-06,
+      "loss": 0.8023,
+      "step": 3009
+    },
+    {
+      "epoch": 0.5790688726433244,
+      "grad_norm": 0.8468530642044558,
+      "learning_rate": 7.946081333131227e-06,
+      "loss": 0.819,
+      "step": 3010
+    },
+    {
+      "epoch": 0.5792612543285879,
+      "grad_norm": 0.9122577695556974,
+      "learning_rate": 7.93998372881876e-06,
+      "loss": 0.8303,
+      "step": 3011
+    },
+    {
+      "epoch": 0.5794536360138515,
+      "grad_norm": 0.8046647705254382,
+      "learning_rate": 7.93388692427554e-06,
+      "loss": 0.8211,
+      "step": 3012
+    },
+    {
+      "epoch": 0.5796460176991151,
+      "grad_norm": 0.9480415822338887,
+      "learning_rate": 7.92779092186855e-06,
+      "loss": 0.8144,
+      "step": 3013
+    },
+    {
+      "epoch": 0.5798383993843786,
+      "grad_norm": 0.8170345985676086,
+      "learning_rate": 7.921695723964473e-06,
+      "loss": 0.8202,
+      "step": 3014
+    },
+    {
+      "epoch": 0.5800307810696421,
+      "grad_norm": 0.8583717252511984,
+      "learning_rate": 7.915601332929678e-06,
+      "loss": 0.7894,
+      "step": 3015
+    },
+    {
+      "epoch": 0.5802231627549057,
+      "grad_norm": 0.8344653679148634,
+      "learning_rate": 7.90950775113021e-06,
+      "loss": 0.8153,
+      "step": 3016
+    },
+    {
+      "epoch": 0.5804155444401693,
+      "grad_norm": 0.8726799899973606,
+      "learning_rate": 7.903414980931813e-06,
+      "loss": 0.8469,
+      "step": 3017
+    },
+    {
+      "epoch": 0.5806079261254329,
+      "grad_norm": 0.9065726198947617,
+      "learning_rate": 7.897323024699907e-06,
+      "loss": 0.8422,
+      "step": 3018
+    },
+    {
+      "epoch": 0.5808003078106965,
+      "grad_norm": 0.8573049766766739,
+      "learning_rate": 7.8912318847996e-06,
+      "loss": 0.8816,
+      "step": 3019
+    },
+    {
+      "epoch": 0.58099268949596,
+      "grad_norm": 0.8479373613957353,
+      "learning_rate": 7.885141563595685e-06,
+      "loss": 0.8161,
+      "step": 3020
+    },
+    {
+      "epoch": 0.5811850711812235,
+      "grad_norm": 0.8460170543173143,
+      "learning_rate": 7.879052063452626e-06,
+      "loss": 0.8297,
+      "step": 3021
+    },
+    {
+      "epoch": 0.5813774528664871,
+      "grad_norm": 0.9504491177316865,
+      "learning_rate": 7.872963386734584e-06,
+      "loss": 0.8847,
+      "step": 3022
+    },
+    {
+      "epoch": 0.5815698345517507,
+      "grad_norm": 0.8502446331479117,
+      "learning_rate": 7.866875535805394e-06,
+      "loss": 0.7598,
+      "step": 3023
+    },
+    {
+      "epoch": 0.5817622162370142,
+      "grad_norm": 0.7776298721634414,
+      "learning_rate": 7.860788513028566e-06,
+      "loss": 0.7542,
+      "step": 3024
+    },
+    {
+      "epoch": 0.5819545979222778,
+      "grad_norm": 0.8128723114833408,
+      "learning_rate": 7.85470232076729e-06,
+      "loss": 0.7648,
+      "step": 3025
+    },
+    {
+      "epoch": 0.5821469796075414,
+      "grad_norm": 0.8927973126416995,
+      "learning_rate": 7.848616961384442e-06,
+      "loss": 0.817,
+      "step": 3026
+    },
+    {
+      "epoch": 0.582339361292805,
+      "grad_norm": 0.8341357450547211,
+      "learning_rate": 7.842532437242559e-06,
+      "loss": 0.7843,
+      "step": 3027
+    },
+    {
+      "epoch": 0.5825317429780685,
+      "grad_norm": 0.8447439845466294,
+      "learning_rate": 7.83644875070387e-06,
+      "loss": 0.772,
+      "step": 3028
+    },
+    {
+      "epoch": 0.582724124663332,
+      "grad_norm": 0.8716557840787396,
+      "learning_rate": 7.83036590413027e-06,
+      "loss": 0.8458,
+      "step": 3029
+    },
+    {
+      "epoch": 0.5829165063485956,
+      "grad_norm": 0.961637746573162,
+      "learning_rate": 7.824283899883327e-06,
+      "loss": 0.8066,
+      "step": 3030
+    },
+    {
+      "epoch": 0.5831088880338592,
+      "grad_norm": 0.6045739307082572,
+      "learning_rate": 7.818202740324287e-06,
+      "loss": 0.3346,
+      "step": 3031
+    },
+    {
+      "epoch": 0.5833012697191228,
+      "grad_norm": 0.8764701159355671,
+      "learning_rate": 7.812122427814068e-06,
+      "loss": 0.8019,
+      "step": 3032
+    },
+    {
+      "epoch": 0.5834936514043862,
+      "grad_norm": 0.8245016309442418,
+      "learning_rate": 7.806042964713248e-06,
+      "loss": 0.8164,
+      "step": 3033
+    },
+    {
+      "epoch": 0.5836860330896498,
+      "grad_norm": 0.8900597017321739,
+      "learning_rate": 7.79996435338209e-06,
+      "loss": 0.8428,
+      "step": 3034
+    },
+    {
+      "epoch": 0.5838784147749134,
+      "grad_norm": 0.6069490201473675,
+      "learning_rate": 7.793886596180521e-06,
+      "loss": 0.3584,
+      "step": 3035
+    },
+    {
+      "epoch": 0.584070796460177,
+      "grad_norm": 0.850445167432794,
+      "learning_rate": 7.787809695468134e-06,
+      "loss": 0.8031,
+      "step": 3036
+    },
+    {
+      "epoch": 0.5842631781454406,
+      "grad_norm": 0.8191644332072716,
+      "learning_rate": 7.78173365360419e-06,
+      "loss": 0.8341,
+      "step": 3037
+    },
+    {
+      "epoch": 0.5844555598307041,
+      "grad_norm": 0.9278635751014758,
+      "learning_rate": 7.775658472947623e-06,
+      "loss": 0.8628,
+      "step": 3038
+    },
+    {
+      "epoch": 0.5846479415159677,
+      "grad_norm": 0.8756129811608546,
+      "learning_rate": 7.769584155857019e-06,
+      "loss": 0.7364,
+      "step": 3039
+    },
+    {
+      "epoch": 0.5848403232012312,
+      "grad_norm": 0.8295265019869823,
+      "learning_rate": 7.763510704690645e-06,
+      "loss": 0.8651,
+      "step": 3040
+    },
+    {
+      "epoch": 0.5850327048864948,
+      "grad_norm": 0.9344473167308979,
+      "learning_rate": 7.757438121806414e-06,
+      "loss": 0.8213,
+      "step": 3041
+    },
+    {
+      "epoch": 0.5852250865717584,
+      "grad_norm": 0.8623958405237314,
+      "learning_rate": 7.75136640956192e-06,
+      "loss": 0.8331,
+      "step": 3042
+    },
+    {
+      "epoch": 0.5854174682570219,
+      "grad_norm": 0.8702482180834638,
+      "learning_rate": 7.745295570314412e-06,
+      "loss": 0.7669,
+      "step": 3043
+    },
+    {
+      "epoch": 0.5856098499422855,
+      "grad_norm": 0.8590528652770578,
+      "learning_rate": 7.739225606420793e-06,
+      "loss": 0.8152,
+      "step": 3044
+    },
+    {
+      "epoch": 0.5858022316275491,
+      "grad_norm": 0.9411240483946355,
+      "learning_rate": 7.733156520237633e-06,
+      "loss": 0.8205,
+      "step": 3045
+    },
+    {
+      "epoch": 0.5859946133128127,
+      "grad_norm": 0.8217855614917635,
+      "learning_rate": 7.727088314121165e-06,
+      "loss": 0.8118,
+      "step": 3046
+    },
+    {
+      "epoch": 0.5861869949980761,
+      "grad_norm": 0.8999959342760475,
+      "learning_rate": 7.721020990427268e-06,
+      "loss": 0.7623,
+      "step": 3047
+    },
+    {
+      "epoch": 0.5863793766833397,
+      "grad_norm": 0.8076819201098527,
+      "learning_rate": 7.714954551511489e-06,
+      "loss": 0.8286,
+      "step": 3048
+    },
+    {
+      "epoch": 0.5865717583686033,
+      "grad_norm": 0.9838264078372583,
+      "learning_rate": 7.708888999729036e-06,
+      "loss": 0.8059,
+      "step": 3049
+    },
+    {
+      "epoch": 0.5867641400538669,
+      "grad_norm": 0.838853410660032,
+      "learning_rate": 7.702824337434756e-06,
+      "loss": 0.8026,
+      "step": 3050
+    },
+    {
+      "epoch": 0.5869565217391305,
+      "grad_norm": 0.8905900233665991,
+      "learning_rate": 7.69676056698316e-06,
+      "loss": 0.7756,
+      "step": 3051
+    },
+    {
+      "epoch": 0.587148903424394,
+      "grad_norm": 0.9419407481925166,
+      "learning_rate": 7.690697690728417e-06,
+      "loss": 0.8578,
+      "step": 3052
+    },
+    {
+      "epoch": 0.5873412851096576,
+      "grad_norm": 0.922329873017219,
+      "learning_rate": 7.68463571102434e-06,
+      "loss": 0.8371,
+      "step": 3053
+    },
+    {
+      "epoch": 0.5875336667949211,
+      "grad_norm": 0.8962603288584164,
+      "learning_rate": 7.678574630224399e-06,
+      "loss": 0.7945,
+      "step": 3054
+    },
+    {
+      "epoch": 0.5877260484801847,
+      "grad_norm": 0.9288464944469271,
+      "learning_rate": 7.672514450681721e-06,
+      "loss": 0.8166,
+      "step": 3055
+    },
+    {
+      "epoch": 0.5879184301654482,
+      "grad_norm": 0.8634105386173708,
+      "learning_rate": 7.666455174749066e-06,
+      "loss": 0.8214,
+      "step": 3056
+    },
+    {
+      "epoch": 0.5881108118507118,
+      "grad_norm": 0.6261273498903727,
+      "learning_rate": 7.66039680477886e-06,
+      "loss": 0.3741,
+      "step": 3057
+    },
+    {
+      "epoch": 0.5883031935359754,
+      "grad_norm": 0.9649566729576305,
+      "learning_rate": 7.654339343123173e-06,
+      "loss": 0.8091,
+      "step": 3058
+    },
+    {
+      "epoch": 0.588495575221239,
+      "grad_norm": 0.855065413574077,
+      "learning_rate": 7.648282792133711e-06,
+      "loss": 0.8261,
+      "step": 3059
+    },
+    {
+      "epoch": 0.5886879569065026,
+      "grad_norm": 0.8528849063628179,
+      "learning_rate": 7.642227154161841e-06,
+      "loss": 0.7888,
+      "step": 3060
+    },
+    {
+      "epoch": 0.588880338591766,
+      "grad_norm": 0.8227876373967833,
+      "learning_rate": 7.636172431558575e-06,
+      "loss": 0.7963,
+      "step": 3061
+    },
+    {
+      "epoch": 0.5890727202770296,
+      "grad_norm": 0.7792393397490404,
+      "learning_rate": 7.630118626674557e-06,
+      "loss": 0.7655,
+      "step": 3062
+    },
+    {
+      "epoch": 0.5892651019622932,
+      "grad_norm": 0.7981874050972689,
+      "learning_rate": 7.6240657418600846e-06,
+      "loss": 0.8212,
+      "step": 3063
+    },
+    {
+      "epoch": 0.5894574836475568,
+      "grad_norm": 0.9132338607333731,
+      "learning_rate": 7.618013779465101e-06,
+      "loss": 0.8542,
+      "step": 3064
+    },
+    {
+      "epoch": 0.5896498653328203,
+      "grad_norm": 0.8812682269125363,
+      "learning_rate": 7.611962741839178e-06,
+      "loss": 0.8213,
+      "step": 3065
+    },
+    {
+      "epoch": 0.5898422470180839,
+      "grad_norm": 0.9454804009935125,
+      "learning_rate": 7.6059126313315466e-06,
+      "loss": 0.7802,
+      "step": 3066
+    },
+    {
+      "epoch": 0.5900346287033474,
+      "grad_norm": 0.9162351450178304,
+      "learning_rate": 7.599863450291056e-06,
+      "loss": 0.8518,
+      "step": 3067
+    },
+    {
+      "epoch": 0.590227010388611,
+      "grad_norm": 0.9469107188140875,
+      "learning_rate": 7.593815201066215e-06,
+      "loss": 0.8665,
+      "step": 3068
+    },
+    {
+      "epoch": 0.5904193920738746,
+      "grad_norm": 0.8793636229932824,
+      "learning_rate": 7.587767886005164e-06,
+      "loss": 0.8374,
+      "step": 3069
+    },
+    {
+      "epoch": 0.5906117737591381,
+      "grad_norm": 0.8899672133097257,
+      "learning_rate": 7.581721507455672e-06,
+      "loss": 0.8306,
+      "step": 3070
+    },
+    {
+      "epoch": 0.5908041554444017,
+      "grad_norm": 0.8515749981885254,
+      "learning_rate": 7.575676067765154e-06,
+      "loss": 0.7981,
+      "step": 3071
+    },
+    {
+      "epoch": 0.5909965371296653,
+      "grad_norm": 0.882770628205672,
+      "learning_rate": 7.569631569280662e-06,
+      "loss": 0.7921,
+      "step": 3072
+    },
+    {
+      "epoch": 0.5911889188149289,
+      "grad_norm": 0.6056549360218914,
+      "learning_rate": 7.563588014348871e-06,
+      "loss": 0.3194,
+      "step": 3073
+    },
+    {
+      "epoch": 0.5913813005001923,
+      "grad_norm": 0.8374614030515412,
+      "learning_rate": 7.5575454053161e-06,
+      "loss": 0.7966,
+      "step": 3074
+    },
+    {
+      "epoch": 0.5915736821854559,
+      "grad_norm": 0.7961777627180441,
+      "learning_rate": 7.551503744528304e-06,
+      "loss": 0.8506,
+      "step": 3075
+    },
+    {
+      "epoch": 0.5917660638707195,
+      "grad_norm": 0.9367484561296915,
+      "learning_rate": 7.545463034331054e-06,
+      "loss": 0.8052,
+      "step": 3076
+    },
+    {
+      "epoch": 0.5919584455559831,
+      "grad_norm": 1.0168594708037124,
+      "learning_rate": 7.539423277069568e-06,
+      "loss": 0.761,
+      "step": 3077
+    },
+    {
+      "epoch": 0.5921508272412467,
+      "grad_norm": 0.8960211740060643,
+      "learning_rate": 7.53338447508869e-06,
+      "loss": 0.7819,
+      "step": 3078
+    },
+    {
+      "epoch": 0.5923432089265102,
+      "grad_norm": 0.8657871279325146,
+      "learning_rate": 7.52734663073288e-06,
+      "loss": 0.7829,
+      "step": 3079
+    },
+    {
+      "epoch": 0.5925355906117737,
+      "grad_norm": 0.8963658616046548,
+      "learning_rate": 7.521309746346246e-06,
+      "loss": 0.8283,
+      "step": 3080
+    },
+    {
+      "epoch": 0.5927279722970373,
+      "grad_norm": 0.8595910064340286,
+      "learning_rate": 7.515273824272516e-06,
+      "loss": 0.8074,
+      "step": 3081
+    },
+    {
+      "epoch": 0.5929203539823009,
+      "grad_norm": 0.924953620684594,
+      "learning_rate": 7.509238866855033e-06,
+      "loss": 0.7838,
+      "step": 3082
+    },
+    {
+      "epoch": 0.5931127356675644,
+      "grad_norm": 0.8708295399703321,
+      "learning_rate": 7.503204876436785e-06,
+      "loss": 0.8134,
+      "step": 3083
+    },
+    {
+      "epoch": 0.593305117352828,
+      "grad_norm": 0.8252221099743346,
+      "learning_rate": 7.497171855360372e-06,
+      "loss": 0.8152,
+      "step": 3084
+    },
+    {
+      "epoch": 0.5934974990380916,
+      "grad_norm": 0.8031782290377106,
+      "learning_rate": 7.491139805968018e-06,
+      "loss": 0.8155,
+      "step": 3085
+    },
+    {
+      "epoch": 0.5936898807233552,
+      "grad_norm": 0.9166018777534907,
+      "learning_rate": 7.485108730601571e-06,
+      "loss": 0.7961,
+      "step": 3086
+    },
+    {
+      "epoch": 0.5938822624086187,
+      "grad_norm": 0.9876024876431779,
+      "learning_rate": 7.4790786316025125e-06,
+      "loss": 0.9059,
+      "step": 3087
+    },
+    {
+      "epoch": 0.5940746440938822,
+      "grad_norm": 0.9483398666357434,
+      "learning_rate": 7.473049511311921e-06,
+      "loss": 0.791,
+      "step": 3088
+    },
+    {
+      "epoch": 0.5942670257791458,
+      "grad_norm": 0.9187046039057428,
+      "learning_rate": 7.467021372070515e-06,
+      "loss": 0.8155,
+      "step": 3089
+    },
+    {
+      "epoch": 0.5944594074644094,
+      "grad_norm": 0.8930715128608752,
+      "learning_rate": 7.46099421621863e-06,
+      "loss": 0.8034,
+      "step": 3090
+    },
+    {
+      "epoch": 0.594651789149673,
+      "grad_norm": 0.7839155746549944,
+      "learning_rate": 7.4549680460962044e-06,
+      "loss": 0.757,
+      "step": 3091
+    },
+    {
+      "epoch": 0.5948441708349365,
+      "grad_norm": 0.8452097987425827,
+      "learning_rate": 7.448942864042819e-06,
+      "loss": 0.8267,
+      "step": 3092
+    },
+    {
+      "epoch": 0.5950365525202,
+      "grad_norm": 0.8822549797773905,
+      "learning_rate": 7.4429186723976425e-06,
+      "loss": 0.7997,
+      "step": 3093
+    },
+    {
+      "epoch": 0.5952289342054636,
+      "grad_norm": 0.8620459348263011,
+      "learning_rate": 7.43689547349948e-06,
+      "loss": 0.788,
+      "step": 3094
+    },
+    {
+      "epoch": 0.5954213158907272,
+      "grad_norm": 0.8637604885035979,
+      "learning_rate": 7.43087326968675e-06,
+      "loss": 0.8077,
+      "step": 3095
+    },
+    {
+      "epoch": 0.5956136975759908,
+      "grad_norm": 0.9209303118404927,
+      "learning_rate": 7.42485206329747e-06,
+      "loss": 0.8964,
+      "step": 3096
+    },
+    {
+      "epoch": 0.5958060792612543,
+      "grad_norm": 0.920291080997183,
+      "learning_rate": 7.418831856669286e-06,
+      "loss": 0.7868,
+      "step": 3097
+    },
+    {
+      "epoch": 0.5959984609465179,
+      "grad_norm": 0.8462017114535356,
+      "learning_rate": 7.41281265213945e-06,
+      "loss": 0.7886,
+      "step": 3098
+    },
+    {
+      "epoch": 0.5961908426317815,
+      "grad_norm": 0.6545635555715053,
+      "learning_rate": 7.406794452044816e-06,
+      "loss": 0.3673,
+      "step": 3099
+    },
+    {
+      "epoch": 0.596383224317045,
+      "grad_norm": 0.8961974493063639,
+      "learning_rate": 7.400777258721865e-06,
+      "loss": 0.8022,
+      "step": 3100
+    },
+    {
+      "epoch": 0.5965756060023086,
+      "grad_norm": 0.8712489269901554,
+      "learning_rate": 7.394761074506679e-06,
+      "loss": 0.7957,
+      "step": 3101
+    },
+    {
+      "epoch": 0.5967679876875721,
+      "grad_norm": 0.8343904165057744,
+      "learning_rate": 7.3887459017349405e-06,
+      "loss": 0.7461,
+      "step": 3102
+    },
+    {
+      "epoch": 0.5969603693728357,
+      "grad_norm": 0.8578563055431729,
+      "learning_rate": 7.382731742741953e-06,
+      "loss": 0.8277,
+      "step": 3103
+    },
+    {
+      "epoch": 0.5971527510580993,
+      "grad_norm": 0.8493467374329317,
+      "learning_rate": 7.376718599862621e-06,
+      "loss": 0.7548,
+      "step": 3104
+    },
+    {
+      "epoch": 0.5973451327433629,
+      "grad_norm": 0.8930306676487114,
+      "learning_rate": 7.370706475431446e-06,
+      "loss": 0.7879,
+      "step": 3105
+    },
+    {
+      "epoch": 0.5975375144286263,
+      "grad_norm": 0.900895017168664,
+      "learning_rate": 7.364695371782547e-06,
+      "loss": 0.8246,
+      "step": 3106
+    },
+    {
+      "epoch": 0.5977298961138899,
+      "grad_norm": 0.8607076725171413,
+      "learning_rate": 7.358685291249644e-06,
+      "loss": 0.8096,
+      "step": 3107
+    },
+    {
+      "epoch": 0.5979222777991535,
+      "grad_norm": 0.9130671388020919,
+      "learning_rate": 7.352676236166051e-06,
+      "loss": 0.7905,
+      "step": 3108
+    },
+    {
+      "epoch": 0.5981146594844171,
+      "grad_norm": 0.8904242523360975,
+      "learning_rate": 7.346668208864695e-06,
+      "loss": 0.7964,
+      "step": 3109
+    },
+    {
+      "epoch": 0.5983070411696807,
+      "grad_norm": 0.864439913064672,
+      "learning_rate": 7.3406612116781e-06,
+      "loss": 0.7564,
+      "step": 3110
+    },
+    {
+      "epoch": 0.5984994228549442,
+      "grad_norm": 0.8965777633341443,
+      "learning_rate": 7.33465524693838e-06,
+      "loss": 0.7698,
+      "step": 3111
+    },
+    {
+      "epoch": 0.5986918045402078,
+      "grad_norm": 0.8118566384128423,
+      "learning_rate": 7.328650316977265e-06,
+      "loss": 0.8015,
+      "step": 3112
+    },
+    {
+      "epoch": 0.5988841862254713,
+      "grad_norm": 1.0147897552282832,
+      "learning_rate": 7.322646424126079e-06,
+      "loss": 0.7917,
+      "step": 3113
+    },
+    {
+      "epoch": 0.5990765679107349,
+      "grad_norm": 0.8868070605714417,
+      "learning_rate": 7.316643570715729e-06,
+      "loss": 0.7148,
+      "step": 3114
+    },
+    {
+      "epoch": 0.5992689495959984,
+      "grad_norm": 0.9764323818789789,
+      "learning_rate": 7.310641759076742e-06,
+      "loss": 0.8673,
+      "step": 3115
+    },
+    {
+      "epoch": 0.599461331281262,
+      "grad_norm": 0.8407298446067433,
+      "learning_rate": 7.304640991539216e-06,
+      "loss": 0.7873,
+      "step": 3116
+    },
+    {
+      "epoch": 0.5996537129665256,
+      "grad_norm": 0.7935276363752035,
+      "learning_rate": 7.2986412704328625e-06,
+      "loss": 0.8145,
+      "step": 3117
+    },
+    {
+      "epoch": 0.5998460946517892,
+      "grad_norm": 0.8582624609446616,
+      "learning_rate": 7.292642598086982e-06,
+      "loss": 0.7755,
+      "step": 3118
+    },
+    {
+      "epoch": 0.6000384763370528,
+      "grad_norm": 0.9187197959824424,
+      "learning_rate": 7.286644976830457e-06,
+      "loss": 0.7664,
+      "step": 3119
+    },
+    {
+      "epoch": 0.6002308580223162,
+      "grad_norm": 0.96134733723011,
+      "learning_rate": 7.280648408991775e-06,
+      "loss": 0.9046,
+      "step": 3120
+    },
+    {
+      "epoch": 0.6004232397075798,
+      "grad_norm": 0.861520268007993,
+      "learning_rate": 7.274652896899015e-06,
+      "loss": 0.8574,
+      "step": 3121
+    },
+    {
+      "epoch": 0.6006156213928434,
+      "grad_norm": 0.9374163091623261,
+      "learning_rate": 7.268658442879834e-06,
+      "loss": 0.7657,
+      "step": 3122
+    },
+    {
+      "epoch": 0.600808003078107,
+      "grad_norm": 0.9123727765837264,
+      "learning_rate": 7.262665049261489e-06,
+      "loss": 0.8787,
+      "step": 3123
+    },
+    {
+      "epoch": 0.6010003847633705,
+      "grad_norm": 0.828863783561032,
+      "learning_rate": 7.256672718370824e-06,
+      "loss": 0.7397,
+      "step": 3124
+    },
+    {
+      "epoch": 0.6011927664486341,
+      "grad_norm": 0.8385400168552523,
+      "learning_rate": 7.250681452534261e-06,
+      "loss": 0.7989,
+      "step": 3125
+    },
+    {
+      "epoch": 0.6013851481338977,
+      "grad_norm": 0.8931828855974093,
+      "learning_rate": 7.2446912540778196e-06,
+      "loss": 0.81,
+      "step": 3126
+    },
+    {
+      "epoch": 0.6015775298191612,
+      "grad_norm": 0.8625182574694334,
+      "learning_rate": 7.238702125327106e-06,
+      "loss": 0.8033,
+      "step": 3127
+    },
+    {
+      "epoch": 0.6017699115044248,
+      "grad_norm": 0.7784839473669195,
+      "learning_rate": 7.232714068607296e-06,
+      "loss": 0.775,
+      "step": 3128
+    },
+    {
+      "epoch": 0.6019622931896883,
+      "grad_norm": 0.8835223888983896,
+      "learning_rate": 7.226727086243168e-06,
+      "loss": 0.8349,
+      "step": 3129
+    },
+    {
+      "epoch": 0.6021546748749519,
+      "grad_norm": 0.9950716943378705,
+      "learning_rate": 7.220741180559074e-06,
+      "loss": 0.7826,
+      "step": 3130
+    },
+    {
+      "epoch": 0.6023470565602155,
+      "grad_norm": 0.8586052461876977,
+      "learning_rate": 7.214756353878942e-06,
+      "loss": 0.7997,
+      "step": 3131
+    },
+    {
+      "epoch": 0.6025394382454791,
+      "grad_norm": 0.8987357874420159,
+      "learning_rate": 7.208772608526293e-06,
+      "loss": 0.7358,
+      "step": 3132
+    },
+    {
+      "epoch": 0.6027318199307425,
+      "grad_norm": 0.9083414350183959,
+      "learning_rate": 7.202789946824227e-06,
+      "loss": 0.813,
+      "step": 3133
+    },
+    {
+      "epoch": 0.6029242016160061,
+      "grad_norm": 0.5835214227168214,
+      "learning_rate": 7.1968083710954075e-06,
+      "loss": 0.3663,
+      "step": 3134
+    },
+    {
+      "epoch": 0.6031165833012697,
+      "grad_norm": 0.8513599105573064,
+      "learning_rate": 7.1908278836621e-06,
+      "loss": 0.7844,
+      "step": 3135
+    },
+    {
+      "epoch": 0.6033089649865333,
+      "grad_norm": 0.8696866321047856,
+      "learning_rate": 7.184848486846128e-06,
+      "loss": 0.7648,
+      "step": 3136
+    },
+    {
+      "epoch": 0.6035013466717969,
+      "grad_norm": 0.8955215727728201,
+      "learning_rate": 7.178870182968904e-06,
+      "loss": 0.8203,
+      "step": 3137
+    },
+    {
+      "epoch": 0.6036937283570604,
+      "grad_norm": 0.8236279870026496,
+      "learning_rate": 7.1728929743514065e-06,
+      "loss": 0.8318,
+      "step": 3138
+    },
+    {
+      "epoch": 0.603886110042324,
+      "grad_norm": 0.970005284669574,
+      "learning_rate": 7.166916863314199e-06,
+      "loss": 0.8045,
+      "step": 3139
+    },
+    {
+      "epoch": 0.6040784917275875,
+      "grad_norm": 0.8511805441116412,
+      "learning_rate": 7.1609418521774095e-06,
+      "loss": 0.8808,
+      "step": 3140
+    },
+    {
+      "epoch": 0.6042708734128511,
+      "grad_norm": 0.9203176973182462,
+      "learning_rate": 7.154967943260748e-06,
+      "loss": 0.8506,
+      "step": 3141
+    },
+    {
+      "epoch": 0.6044632550981146,
+      "grad_norm": 0.9371325915207933,
+      "learning_rate": 7.148995138883483e-06,
+      "loss": 0.808,
+      "step": 3142
+    },
+    {
+      "epoch": 0.6046556367833782,
+      "grad_norm": 0.8906548701417926,
+      "learning_rate": 7.143023441364471e-06,
+      "loss": 0.7915,
+      "step": 3143
+    },
+    {
+      "epoch": 0.6048480184686418,
+      "grad_norm": 0.898448879804392,
+      "learning_rate": 7.13705285302213e-06,
+      "loss": 0.808,
+      "step": 3144
+    },
+    {
+      "epoch": 0.6050404001539054,
+      "grad_norm": 0.9149630991088521,
+      "learning_rate": 7.131083376174441e-06,
+      "loss": 0.8278,
+      "step": 3145
+    },
+    {
+      "epoch": 0.605232781839169,
+      "grad_norm": 0.7939699411758951,
+      "learning_rate": 7.125115013138966e-06,
+      "loss": 0.7719,
+      "step": 3146
+    },
+    {
+      "epoch": 0.6054251635244324,
+      "grad_norm": 0.8394153510168796,
+      "learning_rate": 7.119147766232832e-06,
+      "loss": 0.8253,
+      "step": 3147
+    },
+    {
+      "epoch": 0.605617545209696,
+      "grad_norm": 0.7149235119097777,
+      "learning_rate": 7.113181637772721e-06,
+      "loss": 0.7639,
+      "step": 3148
+    },
+    {
+      "epoch": 0.6058099268949596,
+      "grad_norm": 0.6064989865257036,
+      "learning_rate": 7.107216630074895e-06,
+      "loss": 0.3252,
+      "step": 3149
+    },
+    {
+      "epoch": 0.6060023085802232,
+      "grad_norm": 0.8781222337142656,
+      "learning_rate": 7.1012527454551795e-06,
+      "loss": 0.847,
+      "step": 3150
+    },
+    {
+      "epoch": 0.6061946902654868,
+      "grad_norm": 0.8384041076045359,
+      "learning_rate": 7.09528998622895e-06,
+      "loss": 0.7787,
+      "step": 3151
+    },
+    {
+      "epoch": 0.6063870719507503,
+      "grad_norm": 0.8488616507643492,
+      "learning_rate": 7.089328354711159e-06,
+      "loss": 0.8487,
+      "step": 3152
+    },
+    {
+      "epoch": 0.6065794536360138,
+      "grad_norm": 0.9074047118656978,
+      "learning_rate": 7.083367853216323e-06,
+      "loss": 0.8333,
+      "step": 3153
+    },
+    {
+      "epoch": 0.6067718353212774,
+      "grad_norm": 0.8608699170883415,
+      "learning_rate": 7.077408484058505e-06,
+      "loss": 0.7398,
+      "step": 3154
+    },
+    {
+      "epoch": 0.606964217006541,
+      "grad_norm": 0.8648151759791979,
+      "learning_rate": 7.071450249551342e-06,
+      "loss": 0.8532,
+      "step": 3155
+    },
+    {
+      "epoch": 0.6071565986918045,
+      "grad_norm": 0.9508996826462156,
+      "learning_rate": 7.065493152008026e-06,
+      "loss": 0.839,
+      "step": 3156
+    },
+    {
+      "epoch": 0.6073489803770681,
+      "grad_norm": 0.9332209533322023,
+      "learning_rate": 7.059537193741306e-06,
+      "loss": 0.82,
+      "step": 3157
+    },
+    {
+      "epoch": 0.6075413620623317,
+      "grad_norm": 0.855719122449895,
+      "learning_rate": 7.053582377063489e-06,
+      "loss": 0.7955,
+      "step": 3158
+    },
+    {
+      "epoch": 0.6077337437475953,
+      "grad_norm": 0.9646406715980913,
+      "learning_rate": 7.047628704286446e-06,
+      "loss": 0.8858,
+      "step": 3159
+    },
+    {
+      "epoch": 0.6079261254328588,
+      "grad_norm": 0.6460764122220205,
+      "learning_rate": 7.041676177721588e-06,
+      "loss": 0.3584,
+      "step": 3160
+    },
+    {
+      "epoch": 0.6081185071181223,
+      "grad_norm": 0.6644569931388979,
+      "learning_rate": 7.035724799679898e-06,
+      "loss": 0.3626,
+      "step": 3161
+    },
+    {
+      "epoch": 0.6083108888033859,
+      "grad_norm": 0.8224561617918454,
+      "learning_rate": 7.029774572471904e-06,
+      "loss": 0.8109,
+      "step": 3162
+    },
+    {
+      "epoch": 0.6085032704886495,
+      "grad_norm": 0.8590250462043976,
+      "learning_rate": 7.023825498407689e-06,
+      "loss": 0.8117,
+      "step": 3163
+    },
+    {
+      "epoch": 0.6086956521739131,
+      "grad_norm": 0.9650295747203016,
+      "learning_rate": 7.0178775797968855e-06,
+      "loss": 0.8033,
+      "step": 3164
+    },
+    {
+      "epoch": 0.6088880338591766,
+      "grad_norm": 0.9104567371504835,
+      "learning_rate": 7.011930818948688e-06,
+      "loss": 0.8541,
+      "step": 3165
+    },
+    {
+      "epoch": 0.6090804155444401,
+      "grad_norm": 0.9043342174046568,
+      "learning_rate": 7.005985218171825e-06,
+      "loss": 0.8298,
+      "step": 3166
+    },
+    {
+      "epoch": 0.6092727972297037,
+      "grad_norm": 0.9105373675773297,
+      "learning_rate": 7.000040779774591e-06,
+      "loss": 0.8213,
+      "step": 3167
+    },
+    {
+      "epoch": 0.6094651789149673,
+      "grad_norm": 0.8870548361100743,
+      "learning_rate": 6.994097506064812e-06,
+      "loss": 0.7731,
+      "step": 3168
+    },
+    {
+      "epoch": 0.6096575606002309,
+      "grad_norm": 0.8951965845474892,
+      "learning_rate": 6.9881553993498805e-06,
+      "loss": 0.8446,
+      "step": 3169
+    },
+    {
+      "epoch": 0.6098499422854944,
+      "grad_norm": 0.8015706311751705,
+      "learning_rate": 6.9822144619367275e-06,
+      "loss": 0.7997,
+      "step": 3170
+    },
+    {
+      "epoch": 0.610042323970758,
+      "grad_norm": 0.9740416394395843,
+      "learning_rate": 6.97627469613182e-06,
+      "loss": 0.7753,
+      "step": 3171
+    },
+    {
+      "epoch": 0.6102347056560216,
+      "grad_norm": 0.9236363410268809,
+      "learning_rate": 6.970336104241186e-06,
+      "loss": 0.7849,
+      "step": 3172
+    },
+    {
+      "epoch": 0.6104270873412851,
+      "grad_norm": 0.9421826033969919,
+      "learning_rate": 6.9643986885703955e-06,
+      "loss": 0.8121,
+      "step": 3173
+    },
+    {
+      "epoch": 0.6106194690265486,
+      "grad_norm": 0.9071042122616902,
+      "learning_rate": 6.958462451424547e-06,
+      "loss": 0.8032,
+      "step": 3174
+    },
+    {
+      "epoch": 0.6108118507118122,
+      "grad_norm": 0.9039668941925401,
+      "learning_rate": 6.952527395108302e-06,
+      "loss": 0.7991,
+      "step": 3175
+    },
+    {
+      "epoch": 0.6110042323970758,
+      "grad_norm": 0.6578580464488106,
+      "learning_rate": 6.9465935219258504e-06,
+      "loss": 0.3317,
+      "step": 3176
+    },
+    {
+      "epoch": 0.6111966140823394,
+      "grad_norm": 0.8987815195169722,
+      "learning_rate": 6.9406608341809215e-06,
+      "loss": 0.7732,
+      "step": 3177
+    },
+    {
+      "epoch": 0.611388995767603,
+      "grad_norm": 0.6043686448276163,
+      "learning_rate": 6.934729334176793e-06,
+      "loss": 0.3352,
+      "step": 3178
+    },
+    {
+      "epoch": 0.6115813774528664,
+      "grad_norm": 0.8281293000916286,
+      "learning_rate": 6.928799024216282e-06,
+      "loss": 0.7971,
+      "step": 3179
+    },
+    {
+      "epoch": 0.61177375913813,
+      "grad_norm": 0.787316992664824,
+      "learning_rate": 6.92286990660173e-06,
+      "loss": 0.8026,
+      "step": 3180
+    },
+    {
+      "epoch": 0.6119661408233936,
+      "grad_norm": 0.9150029650891822,
+      "learning_rate": 6.91694198363503e-06,
+      "loss": 0.8337,
+      "step": 3181
+    },
+    {
+      "epoch": 0.6121585225086572,
+      "grad_norm": 0.928999339918949,
+      "learning_rate": 6.911015257617606e-06,
+      "loss": 0.8479,
+      "step": 3182
+    },
+    {
+      "epoch": 0.6123509041939207,
+      "grad_norm": 0.9248540494808596,
+      "learning_rate": 6.905089730850416e-06,
+      "loss": 0.8113,
+      "step": 3183
+    },
+    {
+      "epoch": 0.6125432858791843,
+      "grad_norm": 0.8818265106947435,
+      "learning_rate": 6.8991654056339505e-06,
+      "loss": 0.7972,
+      "step": 3184
+    },
+    {
+      "epoch": 0.6127356675644479,
+      "grad_norm": 0.9822114058082378,
+      "learning_rate": 6.893242284268244e-06,
+      "loss": 0.807,
+      "step": 3185
+    },
+    {
+      "epoch": 0.6129280492497114,
+      "grad_norm": 0.9043588902710339,
+      "learning_rate": 6.887320369052848e-06,
+      "loss": 0.8618,
+      "step": 3186
+    },
+    {
+      "epoch": 0.613120430934975,
+      "grad_norm": 0.8814427723924462,
+      "learning_rate": 6.8813996622868584e-06,
+      "loss": 0.8203,
+      "step": 3187
+    },
+    {
+      "epoch": 0.6133128126202385,
+      "grad_norm": 0.9022570868851791,
+      "learning_rate": 6.8754801662688964e-06,
+      "loss": 0.8032,
+      "step": 3188
+    },
+    {
+      "epoch": 0.6135051943055021,
+      "grad_norm": 0.8238762588917637,
+      "learning_rate": 6.869561883297116e-06,
+      "loss": 0.7957,
+      "step": 3189
+    },
+    {
+      "epoch": 0.6136975759907657,
+      "grad_norm": 0.9600974124441858,
+      "learning_rate": 6.863644815669197e-06,
+      "loss": 0.8356,
+      "step": 3190
+    },
+    {
+      "epoch": 0.6138899576760293,
+      "grad_norm": 0.9758282982750709,
+      "learning_rate": 6.857728965682344e-06,
+      "loss": 0.7638,
+      "step": 3191
+    },
+    {
+      "epoch": 0.6140823393612928,
+      "grad_norm": 0.8484034413978824,
+      "learning_rate": 6.851814335633298e-06,
+      "loss": 0.8175,
+      "step": 3192
+    },
+    {
+      "epoch": 0.6142747210465563,
+      "grad_norm": 0.8665317931238101,
+      "learning_rate": 6.8459009278183275e-06,
+      "loss": 0.7794,
+      "step": 3193
+    },
+    {
+      "epoch": 0.6144671027318199,
+      "grad_norm": 0.8276333231449946,
+      "learning_rate": 6.839988744533211e-06,
+      "loss": 0.7951,
+      "step": 3194
+    },
+    {
+      "epoch": 0.6146594844170835,
+      "grad_norm": 0.8949228441491327,
+      "learning_rate": 6.834077788073268e-06,
+      "loss": 0.7734,
+      "step": 3195
+    },
+    {
+      "epoch": 0.6148518661023471,
+      "grad_norm": 0.8867414155339814,
+      "learning_rate": 6.8281680607333364e-06,
+      "loss": 0.8322,
+      "step": 3196
+    },
+    {
+      "epoch": 0.6150442477876106,
+      "grad_norm": 0.9022744107019018,
+      "learning_rate": 6.822259564807768e-06,
+      "loss": 0.8249,
+      "step": 3197
+    },
+    {
+      "epoch": 0.6152366294728742,
+      "grad_norm": 0.9197192214136798,
+      "learning_rate": 6.81635230259045e-06,
+      "loss": 0.7689,
+      "step": 3198
+    },
+    {
+      "epoch": 0.6154290111581378,
+      "grad_norm": 0.8927741062925889,
+      "learning_rate": 6.810446276374789e-06,
+      "loss": 0.8304,
+      "step": 3199
+    },
+    {
+      "epoch": 0.6156213928434013,
+      "grad_norm": 0.8431482344950866,
+      "learning_rate": 6.8045414884536975e-06,
+      "loss": 0.8055,
+      "step": 3200
+    },
+    {
+      "epoch": 0.6158137745286649,
+      "grad_norm": 0.8788465252036719,
+      "learning_rate": 6.7986379411196255e-06,
+      "loss": 0.8161,
+      "step": 3201
+    },
+    {
+      "epoch": 0.6160061562139284,
+      "grad_norm": 0.776737379692223,
+      "learning_rate": 6.7927356366645315e-06,
+      "loss": 0.7987,
+      "step": 3202
+    },
+    {
+      "epoch": 0.616198537899192,
+      "grad_norm": 0.8973694464567531,
+      "learning_rate": 6.786834577379893e-06,
+      "loss": 0.7371,
+      "step": 3203
+    },
+    {
+      "epoch": 0.6163909195844556,
+      "grad_norm": 0.8412973274047539,
+      "learning_rate": 6.780934765556702e-06,
+      "loss": 0.785,
+      "step": 3204
+    },
+    {
+      "epoch": 0.6165833012697192,
+      "grad_norm": 0.8774354172203801,
+      "learning_rate": 6.775036203485472e-06,
+      "loss": 0.7961,
+      "step": 3205
+    },
+    {
+      "epoch": 0.6167756829549826,
+      "grad_norm": 0.8700800645986232,
+      "learning_rate": 6.769138893456225e-06,
+      "loss": 0.8031,
+      "step": 3206
+    },
+    {
+      "epoch": 0.6169680646402462,
+      "grad_norm": 0.8439209000574279,
+      "learning_rate": 6.763242837758504e-06,
+      "loss": 0.8287,
+      "step": 3207
+    },
+    {
+      "epoch": 0.6171604463255098,
+      "grad_norm": 0.8598789357399323,
+      "learning_rate": 6.757348038681357e-06,
+      "loss": 0.7842,
+      "step": 3208
+    },
+    {
+      "epoch": 0.6173528280107734,
+      "grad_norm": 0.8546953014801114,
+      "learning_rate": 6.751454498513349e-06,
+      "loss": 0.8521,
+      "step": 3209
+    },
+    {
+      "epoch": 0.617545209696037,
+      "grad_norm": 0.832062168627476,
+      "learning_rate": 6.745562219542554e-06,
+      "loss": 0.745,
+      "step": 3210
+    },
+    {
+      "epoch": 0.6177375913813005,
+      "grad_norm": 0.864899078119935,
+      "learning_rate": 6.7396712040565625e-06,
+      "loss": 0.7723,
+      "step": 3211
+    },
+    {
+      "epoch": 0.617929973066564,
+      "grad_norm": 0.8780489242997199,
+      "learning_rate": 6.733781454342463e-06,
+      "loss": 0.7588,
+      "step": 3212
+    },
+    {
+      "epoch": 0.6181223547518276,
+      "grad_norm": 0.7501286954636133,
+      "learning_rate": 6.727892972686861e-06,
+      "loss": 0.3476,
+      "step": 3213
+    },
+    {
+      "epoch": 0.6183147364370912,
+      "grad_norm": 0.8497829727847006,
+      "learning_rate": 6.722005761375873e-06,
+      "loss": 0.7655,
+      "step": 3214
+    },
+    {
+      "epoch": 0.6185071181223547,
+      "grad_norm": 0.8470211952111159,
+      "learning_rate": 6.716119822695111e-06,
+      "loss": 0.8413,
+      "step": 3215
+    },
+    {
+      "epoch": 0.6186994998076183,
+      "grad_norm": 0.8064483002570961,
+      "learning_rate": 6.710235158929703e-06,
+      "loss": 0.8055,
+      "step": 3216
+    },
+    {
+      "epoch": 0.6188918814928819,
+      "grad_norm": 0.9055695067139053,
+      "learning_rate": 6.704351772364274e-06,
+      "loss": 0.8684,
+      "step": 3217
+    },
+    {
+      "epoch": 0.6190842631781455,
+      "grad_norm": 0.9278909961814427,
+      "learning_rate": 6.698469665282958e-06,
+      "loss": 0.7975,
+      "step": 3218
+    },
+    {
+      "epoch": 0.619276644863409,
+      "grad_norm": 0.8904925179085152,
+      "learning_rate": 6.692588839969397e-06,
+      "loss": 0.841,
+      "step": 3219
+    },
+    {
+      "epoch": 0.6194690265486725,
+      "grad_norm": 0.9516760357670807,
+      "learning_rate": 6.6867092987067214e-06,
+      "loss": 0.8299,
+      "step": 3220
+    },
+    {
+      "epoch": 0.6196614082339361,
+      "grad_norm": 0.8541355090281585,
+      "learning_rate": 6.680831043777579e-06,
+      "loss": 0.8669,
+      "step": 3221
+    },
+    {
+      "epoch": 0.6198537899191997,
+      "grad_norm": 0.8331384693451767,
+      "learning_rate": 6.674954077464108e-06,
+      "loss": 0.839,
+      "step": 3222
+    },
+    {
+      "epoch": 0.6200461716044633,
+      "grad_norm": 0.8670883527886091,
+      "learning_rate": 6.6690784020479484e-06,
+      "loss": 0.7976,
+      "step": 3223
+    },
+    {
+      "epoch": 0.6202385532897268,
+      "grad_norm": 0.8155510040748204,
+      "learning_rate": 6.6632040198102364e-06,
+      "loss": 0.8084,
+      "step": 3224
+    },
+    {
+      "epoch": 0.6204309349749904,
+      "grad_norm": 0.9401647822932593,
+      "learning_rate": 6.657330933031619e-06,
+      "loss": 0.8743,
+      "step": 3225
+    },
+    {
+      "epoch": 0.6206233166602539,
+      "grad_norm": 0.8688948243047673,
+      "learning_rate": 6.651459143992221e-06,
+      "loss": 0.8271,
+      "step": 3226
+    },
+    {
+      "epoch": 0.6208156983455175,
+      "grad_norm": 0.8455039082255754,
+      "learning_rate": 6.645588654971677e-06,
+      "loss": 0.8424,
+      "step": 3227
+    },
+    {
+      "epoch": 0.6210080800307811,
+      "grad_norm": 0.8581881944068434,
+      "learning_rate": 6.639719468249115e-06,
+      "loss": 0.7941,
+      "step": 3228
+    },
+    {
+      "epoch": 0.6212004617160446,
+      "grad_norm": 0.8615299184632822,
+      "learning_rate": 6.633851586103153e-06,
+      "loss": 0.7897,
+      "step": 3229
+    },
+    {
+      "epoch": 0.6213928434013082,
+      "grad_norm": 0.8538974353469998,
+      "learning_rate": 6.627985010811903e-06,
+      "loss": 0.804,
+      "step": 3230
+    },
+    {
+      "epoch": 0.6215852250865718,
+      "grad_norm": 0.7904878724070221,
+      "learning_rate": 6.622119744652977e-06,
+      "loss": 0.7774,
+      "step": 3231
+    },
+    {
+      "epoch": 0.6217776067718354,
+      "grad_norm": 0.9495319421253982,
+      "learning_rate": 6.616255789903467e-06,
+      "loss": 0.8451,
+      "step": 3232
+    },
+    {
+      "epoch": 0.6219699884570988,
+      "grad_norm": 0.884305679119242,
+      "learning_rate": 6.610393148839964e-06,
+      "loss": 0.8916,
+      "step": 3233
+    },
+    {
+      "epoch": 0.6221623701423624,
+      "grad_norm": 0.9153945146344273,
+      "learning_rate": 6.6045318237385526e-06,
+      "loss": 0.8274,
+      "step": 3234
+    },
+    {
+      "epoch": 0.622354751827626,
+      "grad_norm": 0.9965962087941118,
+      "learning_rate": 6.598671816874794e-06,
+      "loss": 0.7878,
+      "step": 3235
+    },
+    {
+      "epoch": 0.6225471335128896,
+      "grad_norm": 0.8432269387782811,
+      "learning_rate": 6.5928131305237465e-06,
+      "loss": 0.8336,
+      "step": 3236
+    },
+    {
+      "epoch": 0.6227395151981532,
+      "grad_norm": 0.9010670160239436,
+      "learning_rate": 6.586955766959958e-06,
+      "loss": 0.7638,
+      "step": 3237
+    },
+    {
+      "epoch": 0.6229318968834167,
+      "grad_norm": 0.8469635899521727,
+      "learning_rate": 6.581099728457451e-06,
+      "loss": 0.8521,
+      "step": 3238
+    },
+    {
+      "epoch": 0.6231242785686802,
+      "grad_norm": 0.9166498325707316,
+      "learning_rate": 6.5752450172897466e-06,
+      "loss": 0.8385,
+      "step": 3239
+    },
+    {
+      "epoch": 0.6233166602539438,
+      "grad_norm": 0.912813088024563,
+      "learning_rate": 6.569391635729847e-06,
+      "loss": 0.8177,
+      "step": 3240
+    },
+    {
+      "epoch": 0.6235090419392074,
+      "grad_norm": 1.0174226704872218,
+      "learning_rate": 6.563539586050233e-06,
+      "loss": 0.875,
+      "step": 3241
+    },
+    {
+      "epoch": 0.6237014236244709,
+      "grad_norm": 0.9580634599146849,
+      "learning_rate": 6.557688870522871e-06,
+      "loss": 0.7997,
+      "step": 3242
+    },
+    {
+      "epoch": 0.6238938053097345,
+      "grad_norm": 0.8823463768934486,
+      "learning_rate": 6.551839491419213e-06,
+      "loss": 0.7886,
+      "step": 3243
+    },
+    {
+      "epoch": 0.6240861869949981,
+      "grad_norm": 0.9081896490601503,
+      "learning_rate": 6.545991451010185e-06,
+      "loss": 0.7829,
+      "step": 3244
+    },
+    {
+      "epoch": 0.6242785686802617,
+      "grad_norm": 0.8181022621074394,
+      "learning_rate": 6.5401447515662065e-06,
+      "loss": 0.7631,
+      "step": 3245
+    },
+    {
+      "epoch": 0.6244709503655252,
+      "grad_norm": 0.8447177424798147,
+      "learning_rate": 6.5342993953571556e-06,
+      "loss": 0.766,
+      "step": 3246
+    },
+    {
+      "epoch": 0.6246633320507887,
+      "grad_norm": 1.0160653360067833,
+      "learning_rate": 6.52845538465241e-06,
+      "loss": 0.86,
+      "step": 3247
+    },
+    {
+      "epoch": 0.6248557137360523,
+      "grad_norm": 0.8903337894357706,
+      "learning_rate": 6.522612721720813e-06,
+      "loss": 0.8212,
+      "step": 3248
+    },
+    {
+      "epoch": 0.6250480954213159,
+      "grad_norm": 0.8022307634815511,
+      "learning_rate": 6.5167714088306865e-06,
+      "loss": 0.8,
+      "step": 3249
+    },
+    {
+      "epoch": 0.6252404771065795,
+      "grad_norm": 0.8556847634088435,
+      "learning_rate": 6.51093144824983e-06,
+      "loss": 0.8102,
+      "step": 3250
+    },
+    {
+      "epoch": 0.6254328587918431,
+      "grad_norm": 0.9337806325315741,
+      "learning_rate": 6.505092842245519e-06,
+      "loss": 0.8336,
+      "step": 3251
+    },
+    {
+      "epoch": 0.6256252404771065,
+      "grad_norm": 0.8668544708514337,
+      "learning_rate": 6.499255593084498e-06,
+      "loss": 0.8348,
+      "step": 3252
+    },
+    {
+      "epoch": 0.6258176221623701,
+      "grad_norm": 0.9303289251345446,
+      "learning_rate": 6.493419703032991e-06,
+      "loss": 0.8474,
+      "step": 3253
+    },
+    {
+      "epoch": 0.6260100038476337,
+      "grad_norm": 0.9663536425916572,
+      "learning_rate": 6.487585174356691e-06,
+      "loss": 0.8199,
+      "step": 3254
+    },
+    {
+      "epoch": 0.6262023855328973,
+      "grad_norm": 0.8758704152102418,
+      "learning_rate": 6.481752009320761e-06,
+      "loss": 0.7749,
+      "step": 3255
+    },
+    {
+      "epoch": 0.6263947672181608,
+      "grad_norm": 0.9102800370733181,
+      "learning_rate": 6.4759202101898366e-06,
+      "loss": 0.8583,
+      "step": 3256
+    },
+    {
+      "epoch": 0.6265871489034244,
+      "grad_norm": 0.842930346519657,
+      "learning_rate": 6.4700897792280285e-06,
+      "loss": 0.8076,
+      "step": 3257
+    },
+    {
+      "epoch": 0.626779530588688,
+      "grad_norm": 0.8900293430703996,
+      "learning_rate": 6.464260718698902e-06,
+      "loss": 0.806,
+      "step": 3258
+    },
+    {
+      "epoch": 0.6269719122739515,
+      "grad_norm": 0.8608757208498347,
+      "learning_rate": 6.458433030865503e-06,
+      "loss": 0.783,
+      "step": 3259
+    },
+    {
+      "epoch": 0.6271642939592151,
+      "grad_norm": 0.8232864677380176,
+      "learning_rate": 6.452606717990346e-06,
+      "loss": 0.8619,
+      "step": 3260
+    },
+    {
+      "epoch": 0.6273566756444786,
+      "grad_norm": 0.8726885734625892,
+      "learning_rate": 6.4467817823354005e-06,
+      "loss": 0.7932,
+      "step": 3261
+    },
+    {
+      "epoch": 0.6275490573297422,
+      "grad_norm": 0.84403046017706,
+      "learning_rate": 6.440958226162104e-06,
+      "loss": 0.7698,
+      "step": 3262
+    },
+    {
+      "epoch": 0.6277414390150058,
+      "grad_norm": 0.8353056218692437,
+      "learning_rate": 6.43513605173137e-06,
+      "loss": 0.7778,
+      "step": 3263
+    },
+    {
+      "epoch": 0.6279338207002694,
+      "grad_norm": 0.8200892276044632,
+      "learning_rate": 6.4293152613035594e-06,
+      "loss": 0.7902,
+      "step": 3264
+    },
+    {
+      "epoch": 0.6281262023855329,
+      "grad_norm": 0.8204864378198353,
+      "learning_rate": 6.4234958571385095e-06,
+      "loss": 0.8142,
+      "step": 3265
+    },
+    {
+      "epoch": 0.6283185840707964,
+      "grad_norm": 0.8295280740883514,
+      "learning_rate": 6.4176778414955075e-06,
+      "loss": 0.7454,
+      "step": 3266
+    },
+    {
+      "epoch": 0.62851096575606,
+      "grad_norm": 0.8518270585385435,
+      "learning_rate": 6.4118612166333124e-06,
+      "loss": 0.8077,
+      "step": 3267
+    },
+    {
+      "epoch": 0.6287033474413236,
+      "grad_norm": 0.8472690680772708,
+      "learning_rate": 6.4060459848101354e-06,
+      "loss": 0.7836,
+      "step": 3268
+    },
+    {
+      "epoch": 0.6288957291265872,
+      "grad_norm": 0.8528792588238602,
+      "learning_rate": 6.400232148283651e-06,
+      "loss": 0.7942,
+      "step": 3269
+    },
+    {
+      "epoch": 0.6290881108118507,
+      "grad_norm": 0.9037560526033779,
+      "learning_rate": 6.3944197093109885e-06,
+      "loss": 0.7552,
+      "step": 3270
+    },
+    {
+      "epoch": 0.6292804924971143,
+      "grad_norm": 0.9566575777139926,
+      "learning_rate": 6.388608670148741e-06,
+      "loss": 0.8748,
+      "step": 3271
+    },
+    {
+      "epoch": 0.6294728741823779,
+      "grad_norm": 0.867029245843893,
+      "learning_rate": 6.38279903305295e-06,
+      "loss": 0.7924,
+      "step": 3272
+    },
+    {
+      "epoch": 0.6296652558676414,
+      "grad_norm": 0.9423646267289226,
+      "learning_rate": 6.376990800279119e-06,
+      "loss": 0.8199,
+      "step": 3273
+    },
+    {
+      "epoch": 0.6298576375529049,
+      "grad_norm": 0.8542668856416233,
+      "learning_rate": 6.3711839740822035e-06,
+      "loss": 0.8408,
+      "step": 3274
+    },
+    {
+      "epoch": 0.6300500192381685,
+      "grad_norm": 0.880492884606813,
+      "learning_rate": 6.3653785567166125e-06,
+      "loss": 0.7591,
+      "step": 3275
+    },
+    {
+      "epoch": 0.6302424009234321,
+      "grad_norm": 0.8336956770194515,
+      "learning_rate": 6.359574550436209e-06,
+      "loss": 0.8287,
+      "step": 3276
+    },
+    {
+      "epoch": 0.6304347826086957,
+      "grad_norm": 0.8642848810114079,
+      "learning_rate": 6.3537719574943105e-06,
+      "loss": 0.8032,
+      "step": 3277
+    },
+    {
+      "epoch": 0.6306271642939593,
+      "grad_norm": 0.6503274362432494,
+      "learning_rate": 6.347970780143678e-06,
+      "loss": 0.3267,
+      "step": 3278
+    },
+    {
+      "epoch": 0.6308195459792227,
+      "grad_norm": 0.8392445093601094,
+      "learning_rate": 6.342171020636533e-06,
+      "loss": 0.7688,
+      "step": 3279
+    },
+    {
+      "epoch": 0.6310119276644863,
+      "grad_norm": 0.8450614345417965,
+      "learning_rate": 6.336372681224543e-06,
+      "loss": 0.8476,
+      "step": 3280
+    },
+    {
+      "epoch": 0.6312043093497499,
+      "grad_norm": 0.9852649845374464,
+      "learning_rate": 6.330575764158819e-06,
+      "loss": 0.821,
+      "step": 3281
+    },
+    {
+      "epoch": 0.6313966910350135,
+      "grad_norm": 0.9344940042387548,
+      "learning_rate": 6.324780271689923e-06,
+      "loss": 0.8716,
+      "step": 3282
+    },
+    {
+      "epoch": 0.631589072720277,
+      "grad_norm": 0.6519226845776372,
+      "learning_rate": 6.318986206067872e-06,
+      "loss": 0.353,
+      "step": 3283
+    },
+    {
+      "epoch": 0.6317814544055406,
+      "grad_norm": 0.8484010032624947,
+      "learning_rate": 6.313193569542113e-06,
+      "loss": 0.8058,
+      "step": 3284
+    },
+    {
+      "epoch": 0.6319738360908042,
+      "grad_norm": 0.8841137579636533,
+      "learning_rate": 6.30740236436155e-06,
+      "loss": 0.79,
+      "step": 3285
+    },
+    {
+      "epoch": 0.6321662177760677,
+      "grad_norm": 0.9829482204290463,
+      "learning_rate": 6.301612592774533e-06,
+      "loss": 0.7963,
+      "step": 3286
+    },
+    {
+      "epoch": 0.6323585994613313,
+      "grad_norm": 0.8531015247197092,
+      "learning_rate": 6.295824257028844e-06,
+      "loss": 0.86,
+      "step": 3287
+    },
+    {
+      "epoch": 0.6325509811465948,
+      "grad_norm": 0.8305723686894826,
+      "learning_rate": 6.290037359371717e-06,
+      "loss": 0.8545,
+      "step": 3288
+    },
+    {
+      "epoch": 0.6327433628318584,
+      "grad_norm": 0.8651097583268559,
+      "learning_rate": 6.284251902049827e-06,
+      "loss": 0.7912,
+      "step": 3289
+    },
+    {
+      "epoch": 0.632935744517122,
+      "grad_norm": 1.0192473532023132,
+      "learning_rate": 6.278467887309283e-06,
+      "loss": 0.8276,
+      "step": 3290
+    },
+    {
+      "epoch": 0.6331281262023856,
+      "grad_norm": 0.8373596439586753,
+      "learning_rate": 6.272685317395644e-06,
+      "loss": 0.8095,
+      "step": 3291
+    },
+    {
+      "epoch": 0.633320507887649,
+      "grad_norm": 0.9044483163854459,
+      "learning_rate": 6.266904194553896e-06,
+      "loss": 0.8218,
+      "step": 3292
+    },
+    {
+      "epoch": 0.6335128895729126,
+      "grad_norm": 0.8038217177123317,
+      "learning_rate": 6.261124521028477e-06,
+      "loss": 0.7973,
+      "step": 3293
+    },
+    {
+      "epoch": 0.6337052712581762,
+      "grad_norm": 0.8174540505654714,
+      "learning_rate": 6.255346299063252e-06,
+      "loss": 0.8053,
+      "step": 3294
+    },
+    {
+      "epoch": 0.6338976529434398,
+      "grad_norm": 0.9596241187255515,
+      "learning_rate": 6.249569530901525e-06,
+      "loss": 0.781,
+      "step": 3295
+    },
+    {
+      "epoch": 0.6340900346287034,
+      "grad_norm": 0.9669165422763425,
+      "learning_rate": 6.243794218786034e-06,
+      "loss": 0.8559,
+      "step": 3296
+    },
+    {
+      "epoch": 0.6342824163139669,
+      "grad_norm": 0.8705302112223647,
+      "learning_rate": 6.238020364958964e-06,
+      "loss": 0.807,
+      "step": 3297
+    },
+    {
+      "epoch": 0.6344747979992305,
+      "grad_norm": 0.9310365553249799,
+      "learning_rate": 6.232247971661912e-06,
+      "loss": 0.8256,
+      "step": 3298
+    },
+    {
+      "epoch": 0.634667179684494,
+      "grad_norm": 0.8734161487733656,
+      "learning_rate": 6.2264770411359256e-06,
+      "loss": 0.797,
+      "step": 3299
+    },
+    {
+      "epoch": 0.6348595613697576,
+      "grad_norm": 0.9197140682850001,
+      "learning_rate": 6.22070757562148e-06,
+      "loss": 0.7875,
+      "step": 3300
+    },
+    {
+      "epoch": 0.6350519430550212,
+      "grad_norm": 0.8805965953154705,
+      "learning_rate": 6.214939577358479e-06,
+      "loss": 0.8425,
+      "step": 3301
+    },
+    {
+      "epoch": 0.6352443247402847,
+      "grad_norm": 0.8845387442101851,
+      "learning_rate": 6.209173048586253e-06,
+      "loss": 0.8207,
+      "step": 3302
+    },
+    {
+      "epoch": 0.6354367064255483,
+      "grad_norm": 0.894272995468681,
+      "learning_rate": 6.203407991543577e-06,
+      "loss": 0.864,
+      "step": 3303
+    },
+    {
+      "epoch": 0.6356290881108119,
+      "grad_norm": 0.8414842017948817,
+      "learning_rate": 6.197644408468635e-06,
+      "loss": 0.8377,
+      "step": 3304
+    },
+    {
+      "epoch": 0.6358214697960755,
+      "grad_norm": 0.8524598679230792,
+      "learning_rate": 6.191882301599052e-06,
+      "loss": 0.783,
+      "step": 3305
+    },
+    {
+      "epoch": 0.6360138514813389,
+      "grad_norm": 0.8438226142920088,
+      "learning_rate": 6.186121673171882e-06,
+      "loss": 0.8523,
+      "step": 3306
+    },
+    {
+      "epoch": 0.6362062331666025,
+      "grad_norm": 0.9277434216855769,
+      "learning_rate": 6.180362525423591e-06,
+      "loss": 0.8237,
+      "step": 3307
+    },
+    {
+      "epoch": 0.6363986148518661,
+      "grad_norm": 0.8822065622776647,
+      "learning_rate": 6.174604860590081e-06,
+      "loss": 0.7883,
+      "step": 3308
+    },
+    {
+      "epoch": 0.6365909965371297,
+      "grad_norm": 0.8454787572539916,
+      "learning_rate": 6.168848680906678e-06,
+      "loss": 0.792,
+      "step": 3309
+    },
+    {
+      "epoch": 0.6367833782223933,
+      "grad_norm": 0.9043150909532154,
+      "learning_rate": 6.163093988608127e-06,
+      "loss": 0.7935,
+      "step": 3310
+    },
+    {
+      "epoch": 0.6369757599076568,
+      "grad_norm": 0.8639190083756666,
+      "learning_rate": 6.157340785928595e-06,
+      "loss": 0.8135,
+      "step": 3311
+    },
+    {
+      "epoch": 0.6371681415929203,
+      "grad_norm": 0.9131326837010276,
+      "learning_rate": 6.151589075101681e-06,
+      "loss": 0.7943,
+      "step": 3312
+    },
+    {
+      "epoch": 0.6373605232781839,
+      "grad_norm": 0.8828062960928537,
+      "learning_rate": 6.145838858360391e-06,
+      "loss": 0.7423,
+      "step": 3313
+    },
+    {
+      "epoch": 0.6375529049634475,
+      "grad_norm": 0.8562861746700261,
+      "learning_rate": 6.140090137937158e-06,
+      "loss": 0.844,
+      "step": 3314
+    },
+    {
+      "epoch": 0.637745286648711,
+      "grad_norm": 0.9308842160495651,
+      "learning_rate": 6.134342916063838e-06,
+      "loss": 0.8188,
+      "step": 3315
+    },
+    {
+      "epoch": 0.6379376683339746,
+      "grad_norm": 0.8777196298490728,
+      "learning_rate": 6.128597194971691e-06,
+      "loss": 0.8086,
+      "step": 3316
+    },
+    {
+      "epoch": 0.6381300500192382,
+      "grad_norm": 0.9175475498151552,
+      "learning_rate": 6.122852976891413e-06,
+      "loss": 0.7653,
+      "step": 3317
+    },
+    {
+      "epoch": 0.6383224317045018,
+      "grad_norm": 0.9270263535243215,
+      "learning_rate": 6.117110264053101e-06,
+      "loss": 0.7633,
+      "step": 3318
+    },
+    {
+      "epoch": 0.6385148133897653,
+      "grad_norm": 0.8935525014362838,
+      "learning_rate": 6.111369058686276e-06,
+      "loss": 0.7569,
+      "step": 3319
+    },
+    {
+      "epoch": 0.6387071950750288,
+      "grad_norm": 0.7899422882093345,
+      "learning_rate": 6.105629363019875e-06,
+      "loss": 0.6938,
+      "step": 3320
+    },
+    {
+      "epoch": 0.6388995767602924,
+      "grad_norm": 0.6485812732166312,
+      "learning_rate": 6.099891179282242e-06,
+      "loss": 0.3211,
+      "step": 3321
+    },
+    {
+      "epoch": 0.639091958445556,
+      "grad_norm": 0.8190688358345675,
+      "learning_rate": 6.094154509701133e-06,
+      "loss": 0.7446,
+      "step": 3322
+    },
+    {
+      "epoch": 0.6392843401308196,
+      "grad_norm": 0.7748032159262164,
+      "learning_rate": 6.088419356503732e-06,
+      "loss": 0.7999,
+      "step": 3323
+    },
+    {
+      "epoch": 0.6394767218160831,
+      "grad_norm": 0.8901207292892132,
+      "learning_rate": 6.082685721916612e-06,
+      "loss": 0.8785,
+      "step": 3324
+    },
+    {
+      "epoch": 0.6396691035013466,
+      "grad_norm": 0.8478939848226588,
+      "learning_rate": 6.076953608165772e-06,
+      "loss": 0.8091,
+      "step": 3325
+    },
+    {
+      "epoch": 0.6398614851866102,
+      "grad_norm": 0.9356862420097852,
+      "learning_rate": 6.07122301747662e-06,
+      "loss": 0.8322,
+      "step": 3326
+    },
+    {
+      "epoch": 0.6400538668718738,
+      "grad_norm": 0.9284375012679388,
+      "learning_rate": 6.065493952073961e-06,
+      "loss": 0.8228,
+      "step": 3327
+    },
+    {
+      "epoch": 0.6402462485571374,
+      "grad_norm": 0.8337844716287123,
+      "learning_rate": 6.0597664141820176e-06,
+      "loss": 0.7485,
+      "step": 3328
+    },
+    {
+      "epoch": 0.6404386302424009,
+      "grad_norm": 0.9587126606668859,
+      "learning_rate": 6.054040406024422e-06,
+      "loss": 0.8393,
+      "step": 3329
+    },
+    {
+      "epoch": 0.6406310119276645,
+      "grad_norm": 0.8589626554632186,
+      "learning_rate": 6.0483159298242e-06,
+      "loss": 0.7715,
+      "step": 3330
+    },
+    {
+      "epoch": 0.6408233936129281,
+      "grad_norm": 0.8378662185150687,
+      "learning_rate": 6.042592987803796e-06,
+      "loss": 0.8242,
+      "step": 3331
+    },
+    {
+      "epoch": 0.6410157752981916,
+      "grad_norm": 0.8892783882000261,
+      "learning_rate": 6.036871582185054e-06,
+      "loss": 0.8007,
+      "step": 3332
+    },
+    {
+      "epoch": 0.6412081569834551,
+      "grad_norm": 0.9093368927377729,
+      "learning_rate": 6.031151715189217e-06,
+      "loss": 0.8272,
+      "step": 3333
+    },
+    {
+      "epoch": 0.6414005386687187,
+      "grad_norm": 0.8892826663072427,
+      "learning_rate": 6.025433389036935e-06,
+      "loss": 0.8243,
+      "step": 3334
+    },
+    {
+      "epoch": 0.6415929203539823,
+      "grad_norm": 1.0164955654024477,
+      "learning_rate": 6.019716605948261e-06,
+      "loss": 0.8523,
+      "step": 3335
+    },
+    {
+      "epoch": 0.6417853020392459,
+      "grad_norm": 0.8385396200195026,
+      "learning_rate": 6.014001368142643e-06,
+      "loss": 0.7297,
+      "step": 3336
+    },
+    {
+      "epoch": 0.6419776837245095,
+      "grad_norm": 0.887065542866528,
+      "learning_rate": 6.008287677838937e-06,
+      "loss": 0.8644,
+      "step": 3337
+    },
+    {
+      "epoch": 0.642170065409773,
+      "grad_norm": 0.9194734015126724,
+      "learning_rate": 6.002575537255395e-06,
+      "loss": 0.8408,
+      "step": 3338
+    },
+    {
+      "epoch": 0.6423624470950365,
+      "grad_norm": 0.6328637996760333,
+      "learning_rate": 5.996864948609662e-06,
+      "loss": 0.3605,
+      "step": 3339
+    },
+    {
+      "epoch": 0.6425548287803001,
+      "grad_norm": 0.8049209683224757,
+      "learning_rate": 5.9911559141187924e-06,
+      "loss": 0.8503,
+      "step": 3340
+    },
+    {
+      "epoch": 0.6427472104655637,
+      "grad_norm": 0.8393806200424538,
+      "learning_rate": 5.9854484359992235e-06,
+      "loss": 0.8227,
+      "step": 3341
+    },
+    {
+      "epoch": 0.6429395921508272,
+      "grad_norm": 0.9643229223377687,
+      "learning_rate": 5.979742516466793e-06,
+      "loss": 0.8172,
+      "step": 3342
+    },
+    {
+      "epoch": 0.6431319738360908,
+      "grad_norm": 0.8036394008272545,
+      "learning_rate": 5.974038157736746e-06,
+      "loss": 0.7999,
+      "step": 3343
+    },
+    {
+      "epoch": 0.6433243555213544,
+      "grad_norm": 0.8652233806187982,
+      "learning_rate": 5.968335362023697e-06,
+      "loss": 0.8588,
+      "step": 3344
+    },
+    {
+      "epoch": 0.643516737206618,
+      "grad_norm": 0.9495401584083717,
+      "learning_rate": 5.962634131541676e-06,
+      "loss": 0.801,
+      "step": 3345
+    },
+    {
+      "epoch": 0.6437091188918815,
+      "grad_norm": 0.9646091728786461,
+      "learning_rate": 5.956934468504101e-06,
+      "loss": 0.8229,
+      "step": 3346
+    },
+    {
+      "epoch": 0.643901500577145,
+      "grad_norm": 0.889389344550636,
+      "learning_rate": 5.951236375123768e-06,
+      "loss": 0.8222,
+      "step": 3347
+    },
+    {
+      "epoch": 0.6440938822624086,
+      "grad_norm": 0.9564497647443633,
+      "learning_rate": 5.945539853612876e-06,
+      "loss": 0.8814,
+      "step": 3348
+    },
+    {
+      "epoch": 0.6442862639476722,
+      "grad_norm": 0.8685531542465685,
+      "learning_rate": 5.939844906183016e-06,
+      "loss": 0.8213,
+      "step": 3349
+    },
+    {
+      "epoch": 0.6444786456329358,
+      "grad_norm": 0.8462374324717425,
+      "learning_rate": 5.934151535045156e-06,
+      "loss": 0.7636,
+      "step": 3350
+    },
+    {
+      "epoch": 0.6446710273181993,
+      "grad_norm": 0.856804697199232,
+      "learning_rate": 5.92845974240966e-06,
+      "loss": 0.8167,
+      "step": 3351
+    },
+    {
+      "epoch": 0.6448634090034628,
+      "grad_norm": 0.8136206163875774,
+      "learning_rate": 5.922769530486283e-06,
+      "loss": 0.773,
+      "step": 3352
+    },
+    {
+      "epoch": 0.6450557906887264,
+      "grad_norm": 0.8392545191809457,
+      "learning_rate": 5.917080901484156e-06,
+      "loss": 0.7626,
+      "step": 3353
+    },
+    {
+      "epoch": 0.64524817237399,
+      "grad_norm": 0.8812308031106392,
+      "learning_rate": 5.9113938576118e-06,
+      "loss": 0.8651,
+      "step": 3354
+    },
+    {
+      "epoch": 0.6454405540592536,
+      "grad_norm": 0.8268426399611039,
+      "learning_rate": 5.905708401077128e-06,
+      "loss": 0.7948,
+      "step": 3355
+    },
+    {
+      "epoch": 0.6456329357445171,
+      "grad_norm": 0.8380468815062609,
+      "learning_rate": 5.900024534087421e-06,
+      "loss": 0.803,
+      "step": 3356
+    },
+    {
+      "epoch": 0.6458253174297807,
+      "grad_norm": 0.8298177579437777,
+      "learning_rate": 5.894342258849355e-06,
+      "loss": 0.7888,
+      "step": 3357
+    },
+    {
+      "epoch": 0.6460176991150443,
+      "grad_norm": 0.8744118054558083,
+      "learning_rate": 5.88866157756899e-06,
+      "loss": 0.8099,
+      "step": 3358
+    },
+    {
+      "epoch": 0.6462100808003078,
+      "grad_norm": 0.8619799714661638,
+      "learning_rate": 5.882982492451757e-06,
+      "loss": 0.7828,
+      "step": 3359
+    },
+    {
+      "epoch": 0.6464024624855714,
+      "grad_norm": 0.9038618763665129,
+      "learning_rate": 5.877305005702471e-06,
+      "loss": 0.8499,
+      "step": 3360
+    },
+    {
+      "epoch": 0.6465948441708349,
+      "grad_norm": 0.8039723887676707,
+      "learning_rate": 5.871629119525335e-06,
+      "loss": 0.7817,
+      "step": 3361
+    },
+    {
+      "epoch": 0.6467872258560985,
+      "grad_norm": 0.9206817160208037,
+      "learning_rate": 5.865954836123915e-06,
+      "loss": 0.7641,
+      "step": 3362
+    },
+    {
+      "epoch": 0.6469796075413621,
+      "grad_norm": 0.9142074672239355,
+      "learning_rate": 5.860282157701167e-06,
+      "loss": 0.8036,
+      "step": 3363
+    },
+    {
+      "epoch": 0.6471719892266257,
+      "grad_norm": 0.9293048166106411,
+      "learning_rate": 5.854611086459423e-06,
+      "loss": 0.7993,
+      "step": 3364
+    },
+    {
+      "epoch": 0.6473643709118891,
+      "grad_norm": 0.9131408951830778,
+      "learning_rate": 5.8489416246003814e-06,
+      "loss": 0.8165,
+      "step": 3365
+    },
+    {
+      "epoch": 0.6475567525971527,
+      "grad_norm": 0.9612734066793067,
+      "learning_rate": 5.8432737743251315e-06,
+      "loss": 0.8102,
+      "step": 3366
+    },
+    {
+      "epoch": 0.6477491342824163,
+      "grad_norm": 0.988981481708172,
+      "learning_rate": 5.8376075378341194e-06,
+      "loss": 0.7893,
+      "step": 3367
+    },
+    {
+      "epoch": 0.6479415159676799,
+      "grad_norm": 0.8353166180843536,
+      "learning_rate": 5.831942917327172e-06,
+      "loss": 0.8035,
+      "step": 3368
+    },
+    {
+      "epoch": 0.6481338976529435,
+      "grad_norm": 0.8067782255117261,
+      "learning_rate": 5.826279915003503e-06,
+      "loss": 0.841,
+      "step": 3369
+    },
+    {
+      "epoch": 0.648326279338207,
+      "grad_norm": 0.9582456531227564,
+      "learning_rate": 5.8206185330616725e-06,
+      "loss": 0.8439,
+      "step": 3370
+    },
+    {
+      "epoch": 0.6485186610234706,
+      "grad_norm": 0.8581615405100692,
+      "learning_rate": 5.814958773699625e-06,
+      "loss": 0.7667,
+      "step": 3371
+    },
+    {
+      "epoch": 0.6487110427087341,
+      "grad_norm": 0.9916848460652058,
+      "learning_rate": 5.809300639114683e-06,
+      "loss": 0.8122,
+      "step": 3372
+    },
+    {
+      "epoch": 0.6489034243939977,
+      "grad_norm": 0.8149801216976552,
+      "learning_rate": 5.803644131503516e-06,
+      "loss": 0.8339,
+      "step": 3373
+    },
+    {
+      "epoch": 0.6490958060792612,
+      "grad_norm": 0.8812249619546761,
+      "learning_rate": 5.797989253062186e-06,
+      "loss": 0.8645,
+      "step": 3374
+    },
+    {
+      "epoch": 0.6492881877645248,
+      "grad_norm": 0.8454970418961201,
+      "learning_rate": 5.792336005986105e-06,
+      "loss": 0.7393,
+      "step": 3375
+    },
+    {
+      "epoch": 0.6494805694497884,
+      "grad_norm": 0.8767169187700221,
+      "learning_rate": 5.786684392470064e-06,
+      "loss": 0.8025,
+      "step": 3376
+    },
+    {
+      "epoch": 0.649672951135052,
+      "grad_norm": 0.8670361812671556,
+      "learning_rate": 5.781034414708208e-06,
+      "loss": 0.8004,
+      "step": 3377
+    },
+    {
+      "epoch": 0.6498653328203156,
+      "grad_norm": 0.8652830843394285,
+      "learning_rate": 5.775386074894058e-06,
+      "loss": 0.8036,
+      "step": 3378
+    },
+    {
+      "epoch": 0.650057714505579,
+      "grad_norm": 0.9044367844046829,
+      "learning_rate": 5.769739375220489e-06,
+      "loss": 0.8051,
+      "step": 3379
+    },
+    {
+      "epoch": 0.6502500961908426,
+      "grad_norm": 0.9801095629810197,
+      "learning_rate": 5.7640943178797445e-06,
+      "loss": 0.794,
+      "step": 3380
+    },
+    {
+      "epoch": 0.6504424778761062,
+      "grad_norm": 0.8683924750628127,
+      "learning_rate": 5.7584509050634395e-06,
+      "loss": 0.7979,
+      "step": 3381
+    },
+    {
+      "epoch": 0.6506348595613698,
+      "grad_norm": 0.9372890464317449,
+      "learning_rate": 5.752809138962525e-06,
+      "loss": 0.7472,
+      "step": 3382
+    },
+    {
+      "epoch": 0.6508272412466333,
+      "grad_norm": 0.8060389377233684,
+      "learning_rate": 5.747169021767342e-06,
+      "loss": 0.8319,
+      "step": 3383
+    },
+    {
+      "epoch": 0.6510196229318969,
+      "grad_norm": 0.8608282094696985,
+      "learning_rate": 5.7415305556675805e-06,
+      "loss": 0.7612,
+      "step": 3384
+    },
+    {
+      "epoch": 0.6512120046171604,
+      "grad_norm": 0.923014441342459,
+      "learning_rate": 5.73589374285227e-06,
+      "loss": 0.7584,
+      "step": 3385
+    },
+    {
+      "epoch": 0.651404386302424,
+      "grad_norm": 0.8468886332538881,
+      "learning_rate": 5.730258585509832e-06,
+      "loss": 0.8095,
+      "step": 3386
+    },
+    {
+      "epoch": 0.6515967679876876,
+      "grad_norm": 0.9215857494979098,
+      "learning_rate": 5.724625085828022e-06,
+      "loss": 0.8538,
+      "step": 3387
+    },
+    {
+      "epoch": 0.6517891496729511,
+      "grad_norm": 0.9630683703919959,
+      "learning_rate": 5.718993245993958e-06,
+      "loss": 0.8165,
+      "step": 3388
+    },
+    {
+      "epoch": 0.6519815313582147,
+      "grad_norm": 0.8514373537068682,
+      "learning_rate": 5.713363068194115e-06,
+      "loss": 0.7803,
+      "step": 3389
+    },
+    {
+      "epoch": 0.6521739130434783,
+      "grad_norm": 0.8442608721777406,
+      "learning_rate": 5.7077345546143235e-06,
+      "loss": 0.8012,
+      "step": 3390
+    },
+    {
+      "epoch": 0.6523662947287419,
+      "grad_norm": 0.8959653415464214,
+      "learning_rate": 5.702107707439766e-06,
+      "loss": 0.8481,
+      "step": 3391
+    },
+    {
+      "epoch": 0.6525586764140053,
+      "grad_norm": 0.8531630525724364,
+      "learning_rate": 5.6964825288549745e-06,
+      "loss": 0.7203,
+      "step": 3392
+    },
+    {
+      "epoch": 0.6527510580992689,
+      "grad_norm": 0.8016015547845826,
+      "learning_rate": 5.690859021043842e-06,
+      "loss": 0.7806,
+      "step": 3393
+    },
+    {
+      "epoch": 0.6529434397845325,
+      "grad_norm": 0.8768006386795663,
+      "learning_rate": 5.685237186189601e-06,
+      "loss": 0.7488,
+      "step": 3394
+    },
+    {
+      "epoch": 0.6531358214697961,
+      "grad_norm": 0.850837520778442,
+      "learning_rate": 5.679617026474853e-06,
+      "loss": 0.8073,
+      "step": 3395
+    },
+    {
+      "epoch": 0.6533282031550597,
+      "grad_norm": 0.9779826813127425,
+      "learning_rate": 5.673998544081527e-06,
+      "loss": 0.8192,
+      "step": 3396
+    },
+    {
+      "epoch": 0.6535205848403232,
+      "grad_norm": 0.6520622335596368,
+      "learning_rate": 5.6683817411909114e-06,
+      "loss": 0.3513,
+      "step": 3397
+    },
+    {
+      "epoch": 0.6537129665255867,
+      "grad_norm": 0.8730142403978184,
+      "learning_rate": 5.662766619983653e-06,
+      "loss": 0.7878,
+      "step": 3398
+    },
+    {
+      "epoch": 0.6539053482108503,
+      "grad_norm": 0.9079305899692554,
+      "learning_rate": 5.65715318263972e-06,
+      "loss": 0.8046,
+      "step": 3399
+    },
+    {
+      "epoch": 0.6540977298961139,
+      "grad_norm": 0.9064988015126727,
+      "learning_rate": 5.651541431338454e-06,
+      "loss": 0.8059,
+      "step": 3400
+    },
+    {
+      "epoch": 0.6542901115813774,
+      "grad_norm": 0.9568575047559172,
+      "learning_rate": 5.645931368258527e-06,
+      "loss": 0.7563,
+      "step": 3401
+    },
+    {
+      "epoch": 0.654482493266641,
+      "grad_norm": 0.8976967109710298,
+      "learning_rate": 5.640322995577958e-06,
+      "loss": 0.8754,
+      "step": 3402
+    },
+    {
+      "epoch": 0.6546748749519046,
+      "grad_norm": 0.8791832133902907,
+      "learning_rate": 5.634716315474109e-06,
+      "loss": 0.7749,
+      "step": 3403
+    },
+    {
+      "epoch": 0.6548672566371682,
+      "grad_norm": 0.8670373637930149,
+      "learning_rate": 5.629111330123689e-06,
+      "loss": 0.7571,
+      "step": 3404
+    },
+    {
+      "epoch": 0.6550596383224317,
+      "grad_norm": 0.8576515335665034,
+      "learning_rate": 5.623508041702743e-06,
+      "loss": 0.8204,
+      "step": 3405
+    },
+    {
+      "epoch": 0.6552520200076952,
+      "grad_norm": 0.9267598891118778,
+      "learning_rate": 5.617906452386659e-06,
+      "loss": 0.8319,
+      "step": 3406
+    },
+    {
+      "epoch": 0.6554444016929588,
+      "grad_norm": 0.9417340260746901,
+      "learning_rate": 5.612306564350179e-06,
+      "loss": 0.8172,
+      "step": 3407
+    },
+    {
+      "epoch": 0.6556367833782224,
+      "grad_norm": 0.8775921919619821,
+      "learning_rate": 5.6067083797673535e-06,
+      "loss": 0.8329,
+      "step": 3408
+    },
+    {
+      "epoch": 0.655829165063486,
+      "grad_norm": 0.8581774591788369,
+      "learning_rate": 5.601111900811607e-06,
+      "loss": 0.7512,
+      "step": 3409
+    },
+    {
+      "epoch": 0.6560215467487496,
+      "grad_norm": 0.9635713510023012,
+      "learning_rate": 5.595517129655681e-06,
+      "loss": 0.8063,
+      "step": 3410
+    },
+    {
+      "epoch": 0.656213928434013,
+      "grad_norm": 0.8611527159379165,
+      "learning_rate": 5.589924068471648e-06,
+      "loss": 0.7376,
+      "step": 3411
+    },
+    {
+      "epoch": 0.6564063101192766,
+      "grad_norm": 0.9092430212397943,
+      "learning_rate": 5.58433271943094e-06,
+      "loss": 0.7997,
+      "step": 3412
+    },
+    {
+      "epoch": 0.6565986918045402,
+      "grad_norm": 0.8934372922222892,
+      "learning_rate": 5.578743084704306e-06,
+      "loss": 0.8262,
+      "step": 3413
+    },
+    {
+      "epoch": 0.6567910734898038,
+      "grad_norm": 0.915952760402876,
+      "learning_rate": 5.573155166461833e-06,
+      "loss": 0.85,
+      "step": 3414
+    },
+    {
+      "epoch": 0.6569834551750673,
+      "grad_norm": 0.8932504961093766,
+      "learning_rate": 5.567568966872947e-06,
+      "loss": 0.8431,
+      "step": 3415
+    },
+    {
+      "epoch": 0.6571758368603309,
+      "grad_norm": 0.8188686089943717,
+      "learning_rate": 5.5619844881064e-06,
+      "loss": 0.7838,
+      "step": 3416
+    },
+    {
+      "epoch": 0.6573682185455945,
+      "grad_norm": 0.8802413758508726,
+      "learning_rate": 5.556401732330281e-06,
+      "loss": 0.8582,
+      "step": 3417
+    },
+    {
+      "epoch": 0.657560600230858,
+      "grad_norm": 0.9450987389514518,
+      "learning_rate": 5.550820701712007e-06,
+      "loss": 0.8071,
+      "step": 3418
+    },
+    {
+      "epoch": 0.6577529819161216,
+      "grad_norm": 0.8636939103813792,
+      "learning_rate": 5.545241398418326e-06,
+      "loss": 0.7777,
+      "step": 3419
+    },
+    {
+      "epoch": 0.6579453636013851,
+      "grad_norm": 0.9401249896317845,
+      "learning_rate": 5.539663824615312e-06,
+      "loss": 0.7657,
+      "step": 3420
+    },
+    {
+      "epoch": 0.6581377452866487,
+      "grad_norm": 0.8817635631155128,
+      "learning_rate": 5.534087982468384e-06,
+      "loss": 0.8047,
+      "step": 3421
+    },
+    {
+      "epoch": 0.6583301269719123,
+      "grad_norm": 0.9323972125737886,
+      "learning_rate": 5.5285138741422615e-06,
+      "loss": 0.8433,
+      "step": 3422
+    },
+    {
+      "epoch": 0.6585225086571759,
+      "grad_norm": 0.849708980367468,
+      "learning_rate": 5.522941501801008e-06,
+      "loss": 0.7583,
+      "step": 3423
+    },
+    {
+      "epoch": 0.6587148903424394,
+      "grad_norm": 0.9282021039493537,
+      "learning_rate": 5.517370867608021e-06,
+      "loss": 0.7535,
+      "step": 3424
+    },
+    {
+      "epoch": 0.6589072720277029,
+      "grad_norm": 0.9001593260460555,
+      "learning_rate": 5.511801973725997e-06,
+      "loss": 0.7585,
+      "step": 3425
+    },
+    {
+      "epoch": 0.6590996537129665,
+      "grad_norm": 0.8653120678097532,
+      "learning_rate": 5.506234822316983e-06,
+      "loss": 0.8742,
+      "step": 3426
+    },
+    {
+      "epoch": 0.6592920353982301,
+      "grad_norm": 0.8974670569913294,
+      "learning_rate": 5.500669415542336e-06,
+      "loss": 0.8098,
+      "step": 3427
+    },
+    {
+      "epoch": 0.6594844170834937,
+      "grad_norm": 0.8698628843568867,
+      "learning_rate": 5.495105755562738e-06,
+      "loss": 0.7987,
+      "step": 3428
+    },
+    {
+      "epoch": 0.6596767987687572,
+      "grad_norm": 0.8520511660267934,
+      "learning_rate": 5.4895438445381945e-06,
+      "loss": 0.796,
+      "step": 3429
+    },
+    {
+      "epoch": 0.6598691804540208,
+      "grad_norm": 0.9688466594229809,
+      "learning_rate": 5.48398368462803e-06,
+      "loss": 0.8478,
+      "step": 3430
+    },
+    {
+      "epoch": 0.6600615621392844,
+      "grad_norm": 0.8958680866699011,
+      "learning_rate": 5.4784252779908905e-06,
+      "loss": 0.8085,
+      "step": 3431
+    },
+    {
+      "epoch": 0.6602539438245479,
+      "grad_norm": 0.8958501141734082,
+      "learning_rate": 5.4728686267847354e-06,
+      "loss": 0.8085,
+      "step": 3432
+    },
+    {
+      "epoch": 0.6604463255098114,
+      "grad_norm": 0.8974014005598916,
+      "learning_rate": 5.467313733166863e-06,
+      "loss": 0.772,
+      "step": 3433
+    },
+    {
+      "epoch": 0.660638707195075,
+      "grad_norm": 0.8519927522397945,
+      "learning_rate": 5.461760599293855e-06,
+      "loss": 0.8344,
+      "step": 3434
+    },
+    {
+      "epoch": 0.6608310888803386,
+      "grad_norm": 0.8699379480912242,
+      "learning_rate": 5.456209227321643e-06,
+      "loss": 0.8013,
+      "step": 3435
+    },
+    {
+      "epoch": 0.6610234705656022,
+      "grad_norm": 0.9296785907888705,
+      "learning_rate": 5.450659619405458e-06,
+      "loss": 0.7074,
+      "step": 3436
+    },
+    {
+      "epoch": 0.6612158522508658,
+      "grad_norm": 0.873430116083185,
+      "learning_rate": 5.445111777699842e-06,
+      "loss": 0.7763,
+      "step": 3437
+    },
+    {
+      "epoch": 0.6614082339361292,
+      "grad_norm": 0.8793807660009954,
+      "learning_rate": 5.439565704358667e-06,
+      "loss": 0.835,
+      "step": 3438
+    },
+    {
+      "epoch": 0.6616006156213928,
+      "grad_norm": 0.8757048483068332,
+      "learning_rate": 5.434021401535105e-06,
+      "loss": 0.7383,
+      "step": 3439
+    },
+    {
+      "epoch": 0.6617929973066564,
+      "grad_norm": 0.864569942063603,
+      "learning_rate": 5.428478871381646e-06,
+      "loss": 0.8315,
+      "step": 3440
+    },
+    {
+      "epoch": 0.66198537899192,
+      "grad_norm": 0.892409111093089,
+      "learning_rate": 5.422938116050092e-06,
+      "loss": 0.8431,
+      "step": 3441
+    },
+    {
+      "epoch": 0.6621777606771835,
+      "grad_norm": 0.6448530189781444,
+      "learning_rate": 5.417399137691552e-06,
+      "loss": 0.3693,
+      "step": 3442
+    },
+    {
+      "epoch": 0.6623701423624471,
+      "grad_norm": 0.8140353080728329,
+      "learning_rate": 5.411861938456453e-06,
+      "loss": 0.8118,
+      "step": 3443
+    },
+    {
+      "epoch": 0.6625625240477107,
+      "grad_norm": 0.9242705788117711,
+      "learning_rate": 5.406326520494522e-06,
+      "loss": 0.8091,
+      "step": 3444
+    },
+    {
+      "epoch": 0.6627549057329742,
+      "grad_norm": 0.9063442905774926,
+      "learning_rate": 5.400792885954802e-06,
+      "loss": 0.8356,
+      "step": 3445
+    },
+    {
+      "epoch": 0.6629472874182378,
+      "grad_norm": 0.9084026647725506,
+      "learning_rate": 5.395261036985635e-06,
+      "loss": 0.8055,
+      "step": 3446
+    },
+    {
+      "epoch": 0.6631396691035013,
+      "grad_norm": 0.8665331451067876,
+      "learning_rate": 5.389730975734686e-06,
+      "loss": 0.82,
+      "step": 3447
+    },
+    {
+      "epoch": 0.6633320507887649,
+      "grad_norm": 0.9222896675925749,
+      "learning_rate": 5.384202704348902e-06,
+      "loss": 0.8341,
+      "step": 3448
+    },
+    {
+      "epoch": 0.6635244324740285,
+      "grad_norm": 0.8723859888934002,
+      "learning_rate": 5.378676224974557e-06,
+      "loss": 0.7942,
+      "step": 3449
+    },
+    {
+      "epoch": 0.6637168141592921,
+      "grad_norm": 0.9459406415667972,
+      "learning_rate": 5.373151539757224e-06,
+      "loss": 0.7946,
+      "step": 3450
+    },
+    {
+      "epoch": 0.6639091958445555,
+      "grad_norm": 0.900772454328203,
+      "learning_rate": 5.367628650841761e-06,
+      "loss": 0.7933,
+      "step": 3451
+    },
+    {
+      "epoch": 0.6641015775298191,
+      "grad_norm": 0.8373295426736534,
+      "learning_rate": 5.362107560372358e-06,
+      "loss": 0.7921,
+      "step": 3452
+    },
+    {
+      "epoch": 0.6642939592150827,
+      "grad_norm": 0.925420809292765,
+      "learning_rate": 5.356588270492487e-06,
+      "loss": 0.8329,
+      "step": 3453
+    },
+    {
+      "epoch": 0.6644863409003463,
+      "grad_norm": 0.8609051991410666,
+      "learning_rate": 5.351070783344926e-06,
+      "loss": 0.8503,
+      "step": 3454
+    },
+    {
+      "epoch": 0.6646787225856099,
+      "grad_norm": 0.8944372273035616,
+      "learning_rate": 5.3455551010717545e-06,
+      "loss": 0.8463,
+      "step": 3455
+    },
+    {
+      "epoch": 0.6648711042708734,
+      "grad_norm": 0.9462118806047423,
+      "learning_rate": 5.34004122581435e-06,
+      "loss": 0.8654,
+      "step": 3456
+    },
+    {
+      "epoch": 0.665063485956137,
+      "grad_norm": 0.9215116407735443,
+      "learning_rate": 5.334529159713389e-06,
+      "loss": 0.7635,
+      "step": 3457
+    },
+    {
+      "epoch": 0.6652558676414005,
+      "grad_norm": 0.7727568990414182,
+      "learning_rate": 5.329018904908841e-06,
+      "loss": 0.7415,
+      "step": 3458
+    },
+    {
+      "epoch": 0.6654482493266641,
+      "grad_norm": 0.8072321873203201,
+      "learning_rate": 5.323510463539989e-06,
+      "loss": 0.8315,
+      "step": 3459
+    },
+    {
+      "epoch": 0.6656406310119277,
+      "grad_norm": 0.8782525262143823,
+      "learning_rate": 5.318003837745382e-06,
+      "loss": 0.8044,
+      "step": 3460
+    },
+    {
+      "epoch": 0.6658330126971912,
+      "grad_norm": 0.8698063937332698,
+      "learning_rate": 5.3124990296628974e-06,
+      "loss": 0.8458,
+      "step": 3461
+    },
+    {
+      "epoch": 0.6660253943824548,
+      "grad_norm": 0.8182830305834479,
+      "learning_rate": 5.306996041429688e-06,
+      "loss": 0.7048,
+      "step": 3462
+    },
+    {
+      "epoch": 0.6662177760677184,
+      "grad_norm": 0.6399638507193557,
+      "learning_rate": 5.301494875182192e-06,
+      "loss": 0.3662,
+      "step": 3463
+    },
+    {
+      "epoch": 0.666410157752982,
+      "grad_norm": 0.8532817520956131,
+      "learning_rate": 5.295995533056162e-06,
+      "loss": 0.838,
+      "step": 3464
+    },
+    {
+      "epoch": 0.6666025394382454,
+      "grad_norm": 0.8825055846925529,
+      "learning_rate": 5.290498017186631e-06,
+      "loss": 0.8635,
+      "step": 3465
+    },
+    {
+      "epoch": 0.666794921123509,
+      "grad_norm": 0.8725101338307265,
+      "learning_rate": 5.2850023297079235e-06,
+      "loss": 0.8135,
+      "step": 3466
+    },
+    {
+      "epoch": 0.6669873028087726,
+      "grad_norm": 0.8997876574885704,
+      "learning_rate": 5.279508472753654e-06,
+      "loss": 0.8239,
+      "step": 3467
+    },
+    {
+      "epoch": 0.6671796844940362,
+      "grad_norm": 0.8472316467188402,
+      "learning_rate": 5.274016448456725e-06,
+      "loss": 0.8218,
+      "step": 3468
+    },
+    {
+      "epoch": 0.6673720661792998,
+      "grad_norm": 0.8623613836051677,
+      "learning_rate": 5.2685262589493314e-06,
+      "loss": 0.7833,
+      "step": 3469
+    },
+    {
+      "epoch": 0.6675644478645633,
+      "grad_norm": 0.5889883151049458,
+      "learning_rate": 5.263037906362953e-06,
+      "loss": 0.341,
+      "step": 3470
+    },
+    {
+      "epoch": 0.6677568295498268,
+      "grad_norm": 0.8869149798193112,
+      "learning_rate": 5.257551392828359e-06,
+      "loss": 0.8097,
+      "step": 3471
+    },
+    {
+      "epoch": 0.6679492112350904,
+      "grad_norm": 0.8951692016075721,
+      "learning_rate": 5.252066720475597e-06,
+      "loss": 0.8458,
+      "step": 3472
+    },
+    {
+      "epoch": 0.668141592920354,
+      "grad_norm": 0.7895346833352443,
+      "learning_rate": 5.246583891434018e-06,
+      "loss": 0.7919,
+      "step": 3473
+    },
+    {
+      "epoch": 0.6683339746056175,
+      "grad_norm": 0.8160478842826477,
+      "learning_rate": 5.241102907832232e-06,
+      "loss": 0.8036,
+      "step": 3474
+    },
+    {
+      "epoch": 0.6685263562908811,
+      "grad_norm": 0.8375849015716907,
+      "learning_rate": 5.235623771798151e-06,
+      "loss": 0.8278,
+      "step": 3475
+    },
+    {
+      "epoch": 0.6687187379761447,
+      "grad_norm": 0.9003797223478527,
+      "learning_rate": 5.23014648545897e-06,
+      "loss": 0.8407,
+      "step": 3476
+    },
+    {
+      "epoch": 0.6689111196614083,
+      "grad_norm": 0.9296892699986864,
+      "learning_rate": 5.224671050941146e-06,
+      "loss": 0.7521,
+      "step": 3477
+    },
+    {
+      "epoch": 0.6691035013466718,
+      "grad_norm": 0.898672403120796,
+      "learning_rate": 5.2191974703704425e-06,
+      "loss": 0.8058,
+      "step": 3478
+    },
+    {
+      "epoch": 0.6692958830319353,
+      "grad_norm": 0.8207078126877101,
+      "learning_rate": 5.213725745871889e-06,
+      "loss": 0.7674,
+      "step": 3479
+    },
+    {
+      "epoch": 0.6694882647171989,
+      "grad_norm": 0.9093658546104043,
+      "learning_rate": 5.208255879569799e-06,
+      "loss": 0.8087,
+      "step": 3480
+    },
+    {
+      "epoch": 0.6696806464024625,
+      "grad_norm": 0.7768444019323659,
+      "learning_rate": 5.20278787358776e-06,
+      "loss": 0.7861,
+      "step": 3481
+    },
+    {
+      "epoch": 0.6698730280877261,
+      "grad_norm": 0.8379058118147227,
+      "learning_rate": 5.197321730048641e-06,
+      "loss": 0.7893,
+      "step": 3482
+    },
+    {
+      "epoch": 0.6700654097729896,
+      "grad_norm": 0.891034907830012,
+      "learning_rate": 5.1918574510745865e-06,
+      "loss": 0.856,
+      "step": 3483
+    },
+    {
+      "epoch": 0.6702577914582531,
+      "grad_norm": 0.8369067008339455,
+      "learning_rate": 5.186395038787017e-06,
+      "loss": 0.7838,
+      "step": 3484
+    },
+    {
+      "epoch": 0.6704501731435167,
+      "grad_norm": 0.8131244828807552,
+      "learning_rate": 5.180934495306638e-06,
+      "loss": 0.8098,
+      "step": 3485
+    },
+    {
+      "epoch": 0.6706425548287803,
+      "grad_norm": 0.9108690937853471,
+      "learning_rate": 5.175475822753404e-06,
+      "loss": 0.8218,
+      "step": 3486
+    },
+    {
+      "epoch": 0.6708349365140439,
+      "grad_norm": 0.8594326666681723,
+      "learning_rate": 5.170019023246574e-06,
+      "loss": 0.7641,
+      "step": 3487
+    },
+    {
+      "epoch": 0.6710273181993074,
+      "grad_norm": 0.8621661501639867,
+      "learning_rate": 5.16456409890466e-06,
+      "loss": 0.8389,
+      "step": 3488
+    },
+    {
+      "epoch": 0.671219699884571,
+      "grad_norm": 0.8082138329366039,
+      "learning_rate": 5.159111051845451e-06,
+      "loss": 0.806,
+      "step": 3489
+    },
+    {
+      "epoch": 0.6714120815698346,
+      "grad_norm": 0.8897655890671015,
+      "learning_rate": 5.153659884186013e-06,
+      "loss": 0.8172,
+      "step": 3490
+    },
+    {
+      "epoch": 0.6716044632550981,
+      "grad_norm": 0.9216149174691473,
+      "learning_rate": 5.148210598042665e-06,
+      "loss": 0.8208,
+      "step": 3491
+    },
+    {
+      "epoch": 0.6717968449403616,
+      "grad_norm": 0.9312666418457196,
+      "learning_rate": 5.142763195531017e-06,
+      "loss": 0.8119,
+      "step": 3492
+    },
+    {
+      "epoch": 0.6719892266256252,
+      "grad_norm": 0.9638720823313451,
+      "learning_rate": 5.137317678765939e-06,
+      "loss": 0.7996,
+      "step": 3493
+    },
+    {
+      "epoch": 0.6721816083108888,
+      "grad_norm": 0.8368803677586132,
+      "learning_rate": 5.131874049861563e-06,
+      "loss": 0.8294,
+      "step": 3494
+    },
+    {
+      "epoch": 0.6723739899961524,
+      "grad_norm": 0.8542322882978177,
+      "learning_rate": 5.126432310931295e-06,
+      "loss": 0.7871,
+      "step": 3495
+    },
+    {
+      "epoch": 0.672566371681416,
+      "grad_norm": 0.8993714134242115,
+      "learning_rate": 5.120992464087807e-06,
+      "loss": 0.8175,
+      "step": 3496
+    },
+    {
+      "epoch": 0.6727587533666795,
+      "grad_norm": 0.8928091417702444,
+      "learning_rate": 5.115554511443033e-06,
+      "loss": 0.7923,
+      "step": 3497
+    },
+    {
+      "epoch": 0.672951135051943,
+      "grad_norm": 0.843033263463772,
+      "learning_rate": 5.1101184551081705e-06,
+      "loss": 0.7615,
+      "step": 3498
+    },
+    {
+      "epoch": 0.6731435167372066,
+      "grad_norm": 0.9151360389039153,
+      "learning_rate": 5.104684297193694e-06,
+      "loss": 0.8135,
+      "step": 3499
+    },
+    {
+      "epoch": 0.6733358984224702,
+      "grad_norm": 0.829882229041914,
+      "learning_rate": 5.099252039809317e-06,
+      "loss": 0.8271,
+      "step": 3500
+    },
+    {
+      "epoch": 0.6735282801077337,
+      "grad_norm": 0.9521620846684223,
+      "learning_rate": 5.09382168506404e-06,
+      "loss": 0.761,
+      "step": 3501
+    },
+    {
+      "epoch": 0.6737206617929973,
+      "grad_norm": 0.8392750459139934,
+      "learning_rate": 5.088393235066114e-06,
+      "loss": 0.7772,
+      "step": 3502
+    },
+    {
+      "epoch": 0.6739130434782609,
+      "grad_norm": 0.872377262138327,
+      "learning_rate": 5.082966691923037e-06,
+      "loss": 0.7876,
+      "step": 3503
+    },
+    {
+      "epoch": 0.6741054251635245,
+      "grad_norm": 0.890769341553311,
+      "learning_rate": 5.077542057741592e-06,
+      "loss": 0.7872,
+      "step": 3504
+    },
+    {
+      "epoch": 0.674297806848788,
+      "grad_norm": 0.8888269472648895,
+      "learning_rate": 5.0721193346278066e-06,
+      "loss": 0.8757,
+      "step": 3505
+    },
+    {
+      "epoch": 0.6744901885340515,
+      "grad_norm": 0.9524378805558665,
+      "learning_rate": 5.066698524686966e-06,
+      "loss": 0.7923,
+      "step": 3506
+    },
+    {
+      "epoch": 0.6746825702193151,
+      "grad_norm": 0.8954936225156637,
+      "learning_rate": 5.061279630023618e-06,
+      "loss": 0.8185,
+      "step": 3507
+    },
+    {
+      "epoch": 0.6748749519045787,
+      "grad_norm": 0.9321481787609909,
+      "learning_rate": 5.055862652741562e-06,
+      "loss": 0.8107,
+      "step": 3508
+    },
+    {
+      "epoch": 0.6750673335898423,
+      "grad_norm": 0.8845591128468205,
+      "learning_rate": 5.050447594943856e-06,
+      "loss": 0.7805,
+      "step": 3509
+    },
+    {
+      "epoch": 0.6752597152751059,
+      "grad_norm": 0.7016011798057287,
+      "learning_rate": 5.045034458732808e-06,
+      "loss": 0.3665,
+      "step": 3510
+    },
+    {
+      "epoch": 0.6754520969603693,
+      "grad_norm": 0.9529752771244386,
+      "learning_rate": 5.0396232462099945e-06,
+      "loss": 0.8175,
+      "step": 3511
+    },
+    {
+      "epoch": 0.6756444786456329,
+      "grad_norm": 0.9637158267113911,
+      "learning_rate": 5.034213959476222e-06,
+      "loss": 0.7999,
+      "step": 3512
+    },
+    {
+      "epoch": 0.6758368603308965,
+      "grad_norm": 0.8897528584521941,
+      "learning_rate": 5.028806600631569e-06,
+      "loss": 0.7648,
+      "step": 3513
+    },
+    {
+      "epoch": 0.6760292420161601,
+      "grad_norm": 0.9264252284143452,
+      "learning_rate": 5.023401171775357e-06,
+      "loss": 0.8389,
+      "step": 3514
+    },
+    {
+      "epoch": 0.6762216237014236,
+      "grad_norm": 0.8528932381395108,
+      "learning_rate": 5.017997675006161e-06,
+      "loss": 0.7894,
+      "step": 3515
+    },
+    {
+      "epoch": 0.6764140053866872,
+      "grad_norm": 0.8691105561957181,
+      "learning_rate": 5.012596112421806e-06,
+      "loss": 0.7948,
+      "step": 3516
+    },
+    {
+      "epoch": 0.6766063870719508,
+      "grad_norm": 0.862917239500328,
+      "learning_rate": 5.007196486119355e-06,
+      "loss": 0.7697,
+      "step": 3517
+    },
+    {
+      "epoch": 0.6767987687572143,
+      "grad_norm": 0.9296188456450661,
+      "learning_rate": 5.001798798195136e-06,
+      "loss": 0.8189,
+      "step": 3518
+    },
+    {
+      "epoch": 0.6769911504424779,
+      "grad_norm": 0.9286094390761442,
+      "learning_rate": 4.996403050744719e-06,
+      "loss": 0.8095,
+      "step": 3519
+    },
+    {
+      "epoch": 0.6771835321277414,
+      "grad_norm": 0.9975050753102912,
+      "learning_rate": 4.991009245862917e-06,
+      "loss": 0.8582,
+      "step": 3520
+    },
+    {
+      "epoch": 0.677375913813005,
+      "grad_norm": 0.9437541462363932,
+      "learning_rate": 4.985617385643789e-06,
+      "loss": 0.7551,
+      "step": 3521
+    },
+    {
+      "epoch": 0.6775682954982686,
+      "grad_norm": 0.8531495352125776,
+      "learning_rate": 4.980227472180643e-06,
+      "loss": 0.7896,
+      "step": 3522
+    },
+    {
+      "epoch": 0.6777606771835322,
+      "grad_norm": 0.9473935337610477,
+      "learning_rate": 4.974839507566027e-06,
+      "loss": 0.8036,
+      "step": 3523
+    },
+    {
+      "epoch": 0.6779530588687956,
+      "grad_norm": 0.8263705866451931,
+      "learning_rate": 4.969453493891733e-06,
+      "loss": 0.7381,
+      "step": 3524
+    },
+    {
+      "epoch": 0.6781454405540592,
+      "grad_norm": 0.8817300664761127,
+      "learning_rate": 4.9640694332488075e-06,
+      "loss": 0.8387,
+      "step": 3525
+    },
+    {
+      "epoch": 0.6783378222393228,
+      "grad_norm": 0.8948846937310264,
+      "learning_rate": 4.958687327727511e-06,
+      "loss": 0.752,
+      "step": 3526
+    },
+    {
+      "epoch": 0.6785302039245864,
+      "grad_norm": 0.8697811743571703,
+      "learning_rate": 4.953307179417376e-06,
+      "loss": 0.7656,
+      "step": 3527
+    },
+    {
+      "epoch": 0.67872258560985,
+      "grad_norm": 0.9663981951426841,
+      "learning_rate": 4.947928990407156e-06,
+      "loss": 0.8443,
+      "step": 3528
+    },
+    {
+      "epoch": 0.6789149672951135,
+      "grad_norm": 0.5935470158855539,
+      "learning_rate": 4.94255276278485e-06,
+      "loss": 0.3201,
+      "step": 3529
+    },
+    {
+      "epoch": 0.6791073489803771,
+      "grad_norm": 0.9057298237313905,
+      "learning_rate": 4.937178498637696e-06,
+      "loss": 0.8184,
+      "step": 3530
+    },
+    {
+      "epoch": 0.6792997306656406,
+      "grad_norm": 0.9818442494797112,
+      "learning_rate": 4.931806200052165e-06,
+      "loss": 0.7898,
+      "step": 3531
+    },
+    {
+      "epoch": 0.6794921123509042,
+      "grad_norm": 0.6131906052037738,
+      "learning_rate": 4.926435869113971e-06,
+      "loss": 0.3257,
+      "step": 3532
+    },
+    {
+      "epoch": 0.6796844940361677,
+      "grad_norm": 0.966882458002513,
+      "learning_rate": 4.92106750790806e-06,
+      "loss": 0.8157,
+      "step": 3533
+    },
+    {
+      "epoch": 0.6798768757214313,
+      "grad_norm": 0.8685023392864447,
+      "learning_rate": 4.915701118518616e-06,
+      "loss": 0.8062,
+      "step": 3534
+    },
+    {
+      "epoch": 0.6800692574066949,
+      "grad_norm": 0.900309014265757,
+      "learning_rate": 4.910336703029055e-06,
+      "loss": 0.8494,
+      "step": 3535
+    },
+    {
+      "epoch": 0.6802616390919585,
+      "grad_norm": 0.8957131784384046,
+      "learning_rate": 4.904974263522025e-06,
+      "loss": 0.7659,
+      "step": 3536
+    },
+    {
+      "epoch": 0.6804540207772221,
+      "grad_norm": 0.8898314726814779,
+      "learning_rate": 4.899613802079419e-06,
+      "loss": 0.8475,
+      "step": 3537
+    },
+    {
+      "epoch": 0.6806464024624855,
+      "grad_norm": 0.8971210866365624,
+      "learning_rate": 4.8942553207823395e-06,
+      "loss": 0.7767,
+      "step": 3538
+    },
+    {
+      "epoch": 0.6808387841477491,
+      "grad_norm": 0.8071589423066978,
+      "learning_rate": 4.888898821711144e-06,
+      "loss": 0.8324,
+      "step": 3539
+    },
+    {
+      "epoch": 0.6810311658330127,
+      "grad_norm": 0.9888037484926163,
+      "learning_rate": 4.883544306945407e-06,
+      "loss": 0.803,
+      "step": 3540
+    },
+    {
+      "epoch": 0.6812235475182763,
+      "grad_norm": 0.83202357873997,
+      "learning_rate": 4.878191778563934e-06,
+      "loss": 0.7642,
+      "step": 3541
+    },
+    {
+      "epoch": 0.6814159292035398,
+      "grad_norm": 0.8834549859352396,
+      "learning_rate": 4.872841238644766e-06,
+      "loss": 0.8408,
+      "step": 3542
+    },
+    {
+      "epoch": 0.6816083108888034,
+      "grad_norm": 0.8456824458085211,
+      "learning_rate": 4.867492689265154e-06,
+      "loss": 0.7753,
+      "step": 3543
+    },
+    {
+      "epoch": 0.681800692574067,
+      "grad_norm": 0.8717591125816796,
+      "learning_rate": 4.8621461325016015e-06,
+      "loss": 0.852,
+      "step": 3544
+    },
+    {
+      "epoch": 0.6819930742593305,
+      "grad_norm": 0.870751263583334,
+      "learning_rate": 4.856801570429822e-06,
+      "loss": 0.7926,
+      "step": 3545
+    },
+    {
+      "epoch": 0.6821854559445941,
+      "grad_norm": 0.8320511866858467,
+      "learning_rate": 4.851459005124759e-06,
+      "loss": 0.7907,
+      "step": 3546
+    },
+    {
+      "epoch": 0.6823778376298576,
+      "grad_norm": 0.9166415841498851,
+      "learning_rate": 4.846118438660578e-06,
+      "loss": 0.8166,
+      "step": 3547
+    },
+    {
+      "epoch": 0.6825702193151212,
+      "grad_norm": 0.6671878835636267,
+      "learning_rate": 4.840779873110675e-06,
+      "loss": 0.3044,
+      "step": 3548
+    },
+    {
+      "epoch": 0.6827626010003848,
+      "grad_norm": 0.8425288398390401,
+      "learning_rate": 4.83544331054766e-06,
+      "loss": 0.7998,
+      "step": 3549
+    },
+    {
+      "epoch": 0.6829549826856484,
+      "grad_norm": 0.9038587839692973,
+      "learning_rate": 4.83010875304337e-06,
+      "loss": 0.8188,
+      "step": 3550
+    },
+    {
+      "epoch": 0.6831473643709118,
+      "grad_norm": 0.8547772607603542,
+      "learning_rate": 4.824776202668875e-06,
+      "loss": 0.7699,
+      "step": 3551
+    },
+    {
+      "epoch": 0.6833397460561754,
+      "grad_norm": 0.8599977655148052,
+      "learning_rate": 4.819445661494437e-06,
+      "loss": 0.7906,
+      "step": 3552
+    },
+    {
+      "epoch": 0.683532127741439,
+      "grad_norm": 0.8454522366232705,
+      "learning_rate": 4.8141171315895694e-06,
+      "loss": 0.8515,
+      "step": 3553
+    },
+    {
+      "epoch": 0.6837245094267026,
+      "grad_norm": 0.8786553480603423,
+      "learning_rate": 4.808790615022987e-06,
+      "loss": 0.8541,
+      "step": 3554
+    },
+    {
+      "epoch": 0.6839168911119662,
+      "grad_norm": 0.9152909860876283,
+      "learning_rate": 4.803466113862626e-06,
+      "loss": 0.7442,
+      "step": 3555
+    },
+    {
+      "epoch": 0.6841092727972297,
+      "grad_norm": 0.8544293928727181,
+      "learning_rate": 4.798143630175642e-06,
+      "loss": 0.8022,
+      "step": 3556
+    },
+    {
+      "epoch": 0.6843016544824932,
+      "grad_norm": 0.9421355045291158,
+      "learning_rate": 4.792823166028405e-06,
+      "loss": 0.8363,
+      "step": 3557
+    },
+    {
+      "epoch": 0.6844940361677568,
+      "grad_norm": 0.9225094571962386,
+      "learning_rate": 4.787504723486505e-06,
+      "loss": 0.8318,
+      "step": 3558
+    },
+    {
+      "epoch": 0.6846864178530204,
+      "grad_norm": 0.867348662503782,
+      "learning_rate": 4.7821883046147414e-06,
+      "loss": 0.8454,
+      "step": 3559
+    },
+    {
+      "epoch": 0.6848787995382839,
+      "grad_norm": 0.8467810808967196,
+      "learning_rate": 4.776873911477133e-06,
+      "loss": 0.8129,
+      "step": 3560
+    },
+    {
+      "epoch": 0.6850711812235475,
+      "grad_norm": 0.9163037562183227,
+      "learning_rate": 4.771561546136908e-06,
+      "loss": 0.7905,
+      "step": 3561
+    },
+    {
+      "epoch": 0.6852635629088111,
+      "grad_norm": 0.8758727711675632,
+      "learning_rate": 4.766251210656509e-06,
+      "loss": 0.77,
+      "step": 3562
+    },
+    {
+      "epoch": 0.6854559445940747,
+      "grad_norm": 0.6301073269736157,
+      "learning_rate": 4.760942907097601e-06,
+      "loss": 0.3422,
+      "step": 3563
+    },
+    {
+      "epoch": 0.6856483262793382,
+      "grad_norm": 0.9455175170183264,
+      "learning_rate": 4.755636637521035e-06,
+      "loss": 0.7993,
+      "step": 3564
+    },
+    {
+      "epoch": 0.6858407079646017,
+      "grad_norm": 0.8575333275957185,
+      "learning_rate": 4.750332403986902e-06,
+      "loss": 0.7853,
+      "step": 3565
+    },
+    {
+      "epoch": 0.6860330896498653,
+      "grad_norm": 0.7872551521221696,
+      "learning_rate": 4.7450302085544735e-06,
+      "loss": 0.7935,
+      "step": 3566
+    },
+    {
+      "epoch": 0.6862254713351289,
+      "grad_norm": 0.8388493290128368,
+      "learning_rate": 4.739730053282255e-06,
+      "loss": 0.8089,
+      "step": 3567
+    },
+    {
+      "epoch": 0.6864178530203925,
+      "grad_norm": 0.9490269479074909,
+      "learning_rate": 4.734431940227951e-06,
+      "loss": 0.8423,
+      "step": 3568
+    },
+    {
+      "epoch": 0.6866102347056561,
+      "grad_norm": 0.8983017927426205,
+      "learning_rate": 4.7291358714484594e-06,
+      "loss": 0.8087,
+      "step": 3569
+    },
+    {
+      "epoch": 0.6868026163909196,
+      "grad_norm": 1.114557174825818,
+      "learning_rate": 4.723841848999907e-06,
+      "loss": 0.8281,
+      "step": 3570
+    },
+    {
+      "epoch": 0.6869949980761831,
+      "grad_norm": 0.8504463305097034,
+      "learning_rate": 4.718549874937612e-06,
+      "loss": 0.7698,
+      "step": 3571
+    },
+    {
+      "epoch": 0.6871873797614467,
+      "grad_norm": 0.8431301765320685,
+      "learning_rate": 4.713259951316103e-06,
+      "loss": 0.767,
+      "step": 3572
+    },
+    {
+      "epoch": 0.6873797614467103,
+      "grad_norm": 0.9808646268481005,
+      "learning_rate": 4.707972080189106e-06,
+      "loss": 0.7813,
+      "step": 3573
+    },
+    {
+      "epoch": 0.6875721431319738,
+      "grad_norm": 0.894115630772827,
+      "learning_rate": 4.702686263609559e-06,
+      "loss": 0.7689,
+      "step": 3574
+    },
+    {
+      "epoch": 0.6877645248172374,
+      "grad_norm": 0.8827206611116349,
+      "learning_rate": 4.697402503629596e-06,
+      "loss": 0.8028,
+      "step": 3575
+    },
+    {
+      "epoch": 0.687956906502501,
+      "grad_norm": 0.9317731096004206,
+      "learning_rate": 4.69212080230055e-06,
+      "loss": 0.8416,
+      "step": 3576
+    },
+    {
+      "epoch": 0.6881492881877646,
+      "grad_norm": 0.9356645088286748,
+      "learning_rate": 4.686841161672974e-06,
+      "loss": 0.8261,
+      "step": 3577
+    },
+    {
+      "epoch": 0.6883416698730281,
+      "grad_norm": 0.8348285205546403,
+      "learning_rate": 4.681563583796587e-06,
+      "loss": 0.7532,
+      "step": 3578
+    },
+    {
+      "epoch": 0.6885340515582916,
+      "grad_norm": 0.9750603186529145,
+      "learning_rate": 4.67628807072034e-06,
+      "loss": 0.8366,
+      "step": 3579
+    },
+    {
+      "epoch": 0.6887264332435552,
+      "grad_norm": 0.8642057022140033,
+      "learning_rate": 4.6710146244923645e-06,
+      "loss": 0.7714,
+      "step": 3580
+    },
+    {
+      "epoch": 0.6889188149288188,
+      "grad_norm": 0.8153077266029984,
+      "learning_rate": 4.665743247159995e-06,
+      "loss": 0.8096,
+      "step": 3581
+    },
+    {
+      "epoch": 0.6891111966140824,
+      "grad_norm": 0.9312036916856901,
+      "learning_rate": 4.660473940769761e-06,
+      "loss": 0.7994,
+      "step": 3582
+    },
+    {
+      "epoch": 0.6893035782993459,
+      "grad_norm": 0.8515489324473289,
+      "learning_rate": 4.655206707367388e-06,
+      "loss": 0.8298,
+      "step": 3583
+    },
+    {
+      "epoch": 0.6894959599846094,
+      "grad_norm": 0.8452601053645646,
+      "learning_rate": 4.649941548997797e-06,
+      "loss": 0.7996,
+      "step": 3584
+    },
+    {
+      "epoch": 0.689688341669873,
+      "grad_norm": 0.8084165532546781,
+      "learning_rate": 4.644678467705101e-06,
+      "loss": 0.7623,
+      "step": 3585
+    },
+    {
+      "epoch": 0.6898807233551366,
+      "grad_norm": 0.8434734237542381,
+      "learning_rate": 4.639417465532622e-06,
+      "loss": 0.8025,
+      "step": 3586
+    },
+    {
+      "epoch": 0.6900731050404002,
+      "grad_norm": 0.7922353935326583,
+      "learning_rate": 4.634158544522849e-06,
+      "loss": 0.7914,
+      "step": 3587
+    },
+    {
+      "epoch": 0.6902654867256637,
+      "grad_norm": 0.9126806109227786,
+      "learning_rate": 4.628901706717476e-06,
+      "loss": 0.7729,
+      "step": 3588
+    },
+    {
+      "epoch": 0.6904578684109273,
+      "grad_norm": 0.8522587899783889,
+      "learning_rate": 4.623646954157399e-06,
+      "loss": 0.8179,
+      "step": 3589
+    },
+    {
+      "epoch": 0.6906502500961909,
+      "grad_norm": 0.8664230182867717,
+      "learning_rate": 4.618394288882681e-06,
+      "loss": 0.7954,
+      "step": 3590
+    },
+    {
+      "epoch": 0.6908426317814544,
+      "grad_norm": 0.6209913227370599,
+      "learning_rate": 4.613143712932603e-06,
+      "loss": 0.3708,
+      "step": 3591
+    },
+    {
+      "epoch": 0.6910350134667179,
+      "grad_norm": 0.8018330865845792,
+      "learning_rate": 4.607895228345603e-06,
+      "loss": 0.7761,
+      "step": 3592
+    },
+    {
+      "epoch": 0.6912273951519815,
+      "grad_norm": 0.6283436482960648,
+      "learning_rate": 4.602648837159333e-06,
+      "loss": 0.3576,
+      "step": 3593
+    },
+    {
+      "epoch": 0.6914197768372451,
+      "grad_norm": 0.9371894659772573,
+      "learning_rate": 4.597404541410622e-06,
+      "loss": 0.7877,
+      "step": 3594
+    },
+    {
+      "epoch": 0.6916121585225087,
+      "grad_norm": 0.8402902400336599,
+      "learning_rate": 4.592162343135483e-06,
+      "loss": 0.794,
+      "step": 3595
+    },
+    {
+      "epoch": 0.6918045402077723,
+      "grad_norm": 0.8943076418714645,
+      "learning_rate": 4.586922244369122e-06,
+      "loss": 0.7708,
+      "step": 3596
+    },
+    {
+      "epoch": 0.6919969218930357,
+      "grad_norm": 0.8880365377059022,
+      "learning_rate": 4.5816842471459224e-06,
+      "loss": 0.8406,
+      "step": 3597
+    },
+    {
+      "epoch": 0.6921893035782993,
+      "grad_norm": 0.8734593508151025,
+      "learning_rate": 4.576448353499457e-06,
+      "loss": 0.7414,
+      "step": 3598
+    },
+    {
+      "epoch": 0.6923816852635629,
+      "grad_norm": 0.8273968468862417,
+      "learning_rate": 4.571214565462477e-06,
+      "loss": 0.7311,
+      "step": 3599
+    },
+    {
+      "epoch": 0.6925740669488265,
+      "grad_norm": 0.8951302994997731,
+      "learning_rate": 4.565982885066923e-06,
+      "loss": 0.7863,
+      "step": 3600
+    },
+    {
+      "epoch": 0.69276644863409,
+      "grad_norm": 0.8126754787935915,
+      "learning_rate": 4.560753314343912e-06,
+      "loss": 0.7565,
+      "step": 3601
+    },
+    {
+      "epoch": 0.6929588303193536,
+      "grad_norm": 0.8201505413348938,
+      "learning_rate": 4.555525855323738e-06,
+      "loss": 0.7499,
+      "step": 3602
+    },
+    {
+      "epoch": 0.6931512120046172,
+      "grad_norm": 0.6127036526875618,
+      "learning_rate": 4.5503005100358945e-06,
+      "loss": 0.3469,
+      "step": 3603
+    },
+    {
+      "epoch": 0.6933435936898807,
+      "grad_norm": 0.581354790917173,
+      "learning_rate": 4.545077280509022e-06,
+      "loss": 0.3311,
+      "step": 3604
+    },
+    {
+      "epoch": 0.6935359753751443,
+      "grad_norm": 0.8803283320673138,
+      "learning_rate": 4.539856168770974e-06,
+      "loss": 0.8188,
+      "step": 3605
+    },
+    {
+      "epoch": 0.6937283570604078,
+      "grad_norm": 0.9306159834532279,
+      "learning_rate": 4.534637176848758e-06,
+      "loss": 0.8516,
+      "step": 3606
+    },
+    {
+      "epoch": 0.6939207387456714,
+      "grad_norm": 0.8777923400228507,
+      "learning_rate": 4.52942030676857e-06,
+      "loss": 0.7396,
+      "step": 3607
+    },
+    {
+      "epoch": 0.694113120430935,
+      "grad_norm": 0.8558811835217759,
+      "learning_rate": 4.524205560555774e-06,
+      "loss": 0.7822,
+      "step": 3608
+    },
+    {
+      "epoch": 0.6943055021161986,
+      "grad_norm": 0.8731814717609567,
+      "learning_rate": 4.5189929402349175e-06,
+      "loss": 0.8141,
+      "step": 3609
+    },
+    {
+      "epoch": 0.694497883801462,
+      "grad_norm": 0.9017981055494648,
+      "learning_rate": 4.513782447829717e-06,
+      "loss": 0.7901,
+      "step": 3610
+    },
+    {
+      "epoch": 0.6946902654867256,
+      "grad_norm": 0.9302660527535979,
+      "learning_rate": 4.508574085363065e-06,
+      "loss": 0.83,
+      "step": 3611
+    },
+    {
+      "epoch": 0.6948826471719892,
+      "grad_norm": 0.7840039610226377,
+      "learning_rate": 4.503367854857035e-06,
+      "loss": 0.7205,
+      "step": 3612
+    },
+    {
+      "epoch": 0.6950750288572528,
+      "grad_norm": 0.8954900724097835,
+      "learning_rate": 4.498163758332853e-06,
+      "loss": 0.8401,
+      "step": 3613
+    },
+    {
+      "epoch": 0.6952674105425164,
+      "grad_norm": 0.9105229430186993,
+      "learning_rate": 4.492961797810932e-06,
+      "loss": 0.75,
+      "step": 3614
+    },
+    {
+      "epoch": 0.6954597922277799,
+      "grad_norm": 0.8844582456757247,
+      "learning_rate": 4.4877619753108605e-06,
+      "loss": 0.7753,
+      "step": 3615
+    },
+    {
+      "epoch": 0.6956521739130435,
+      "grad_norm": 0.9375289234121592,
+      "learning_rate": 4.4825642928513746e-06,
+      "loss": 0.8073,
+      "step": 3616
+    },
+    {
+      "epoch": 0.695844555598307,
+      "grad_norm": 0.8728196039820794,
+      "learning_rate": 4.477368752450409e-06,
+      "loss": 0.7558,
+      "step": 3617
+    },
+    {
+      "epoch": 0.6960369372835706,
+      "grad_norm": 0.8216075015342393,
+      "learning_rate": 4.472175356125036e-06,
+      "loss": 0.7722,
+      "step": 3618
+    },
+    {
+      "epoch": 0.6962293189688342,
+      "grad_norm": 0.8376958661373642,
+      "learning_rate": 4.466984105891521e-06,
+      "loss": 0.8546,
+      "step": 3619
+    },
+    {
+      "epoch": 0.6964217006540977,
+      "grad_norm": 0.9036165126420612,
+      "learning_rate": 4.461795003765285e-06,
+      "loss": 0.8336,
+      "step": 3620
+    },
+    {
+      "epoch": 0.6966140823393613,
+      "grad_norm": 0.8352903777684222,
+      "learning_rate": 4.456608051760914e-06,
+      "loss": 0.8142,
+      "step": 3621
+    },
+    {
+      "epoch": 0.6968064640246249,
+      "grad_norm": 0.884352845008155,
+      "learning_rate": 4.45142325189216e-06,
+      "loss": 0.8166,
+      "step": 3622
+    },
+    {
+      "epoch": 0.6969988457098885,
+      "grad_norm": 0.9394435069928828,
+      "learning_rate": 4.446240606171945e-06,
+      "loss": 0.8123,
+      "step": 3623
+    },
+    {
+      "epoch": 0.6971912273951519,
+      "grad_norm": 0.6215170467039589,
+      "learning_rate": 4.4410601166123475e-06,
+      "loss": 0.3347,
+      "step": 3624
+    },
+    {
+      "epoch": 0.6973836090804155,
+      "grad_norm": 0.8434764178160978,
+      "learning_rate": 4.4358817852246124e-06,
+      "loss": 0.7632,
+      "step": 3625
+    },
+    {
+      "epoch": 0.6975759907656791,
+      "grad_norm": 0.9782534070686068,
+      "learning_rate": 4.430705614019147e-06,
+      "loss": 0.8374,
+      "step": 3626
+    },
+    {
+      "epoch": 0.6977683724509427,
+      "grad_norm": 0.8564699526778444,
+      "learning_rate": 4.425531605005519e-06,
+      "loss": 0.827,
+      "step": 3627
+    },
+    {
+      "epoch": 0.6979607541362063,
+      "grad_norm": 0.90611379979646,
+      "learning_rate": 4.420359760192452e-06,
+      "loss": 0.7657,
+      "step": 3628
+    },
+    {
+      "epoch": 0.6981531358214698,
+      "grad_norm": 0.9058050162457698,
+      "learning_rate": 4.4151900815878455e-06,
+      "loss": 0.8,
+      "step": 3629
+    },
+    {
+      "epoch": 0.6983455175067333,
+      "grad_norm": 0.861891685170595,
+      "learning_rate": 4.410022571198734e-06,
+      "loss": 0.7822,
+      "step": 3630
+    },
+    {
+      "epoch": 0.6985378991919969,
+      "grad_norm": 0.8799162931942315,
+      "learning_rate": 4.404857231031332e-06,
+      "loss": 0.8394,
+      "step": 3631
+    },
+    {
+      "epoch": 0.6987302808772605,
+      "grad_norm": 0.9056661395696834,
+      "learning_rate": 4.399694063090999e-06,
+      "loss": 0.8556,
+      "step": 3632
+    },
+    {
+      "epoch": 0.698922662562524,
+      "grad_norm": 0.8942058825698069,
+      "learning_rate": 4.394533069382255e-06,
+      "loss": 0.7657,
+      "step": 3633
+    },
+    {
+      "epoch": 0.6991150442477876,
+      "grad_norm": 0.866478182313324,
+      "learning_rate": 4.3893742519087754e-06,
+      "loss": 0.8054,
+      "step": 3634
+    },
+    {
+      "epoch": 0.6993074259330512,
+      "grad_norm": 0.8920797396878783,
+      "learning_rate": 4.3842176126733914e-06,
+      "loss": 0.7972,
+      "step": 3635
+    },
+    {
+      "epoch": 0.6994998076183148,
+      "grad_norm": 0.8666726443190469,
+      "learning_rate": 4.379063153678087e-06,
+      "loss": 0.84,
+      "step": 3636
+    },
+    {
+      "epoch": 0.6996921893035783,
+      "grad_norm": 0.8792459132974164,
+      "learning_rate": 4.373910876923997e-06,
+      "loss": 0.7717,
+      "step": 3637
+    },
+    {
+      "epoch": 0.6998845709888418,
+      "grad_norm": 0.8854084744036524,
+      "learning_rate": 4.368760784411423e-06,
+      "loss": 0.795,
+      "step": 3638
+    },
+    {
+      "epoch": 0.7000769526741054,
+      "grad_norm": 0.8881829216786261,
+      "learning_rate": 4.363612878139799e-06,
+      "loss": 0.8379,
+      "step": 3639
+    },
+    {
+      "epoch": 0.700269334359369,
+      "grad_norm": 0.9295027459541669,
+      "learning_rate": 4.3584671601077224e-06,
+      "loss": 0.8449,
+      "step": 3640
+    },
+    {
+      "epoch": 0.7004617160446326,
+      "grad_norm": 0.9349512055956816,
+      "learning_rate": 4.353323632312938e-06,
+      "loss": 0.8411,
+      "step": 3641
+    },
+    {
+      "epoch": 0.7006540977298961,
+      "grad_norm": 0.833851845581803,
+      "learning_rate": 4.348182296752336e-06,
+      "loss": 0.8904,
+      "step": 3642
+    },
+    {
+      "epoch": 0.7008464794151597,
+      "grad_norm": 0.9003278154983128,
+      "learning_rate": 4.343043155421971e-06,
+      "loss": 0.8151,
+      "step": 3643
+    },
+    {
+      "epoch": 0.7010388611004232,
+      "grad_norm": 0.9020294187171629,
+      "learning_rate": 4.3379062103170214e-06,
+      "loss": 0.7859,
+      "step": 3644
+    },
+    {
+      "epoch": 0.7012312427856868,
+      "grad_norm": 0.8198691731303236,
+      "learning_rate": 4.332771463431837e-06,
+      "loss": 0.7822,
+      "step": 3645
+    },
+    {
+      "epoch": 0.7014236244709504,
+      "grad_norm": 0.9553549631275097,
+      "learning_rate": 4.327638916759898e-06,
+      "loss": 0.7992,
+      "step": 3646
+    },
+    {
+      "epoch": 0.7016160061562139,
+      "grad_norm": 0.9052779188123469,
+      "learning_rate": 4.322508572293836e-06,
+      "loss": 0.8024,
+      "step": 3647
+    },
+    {
+      "epoch": 0.7018083878414775,
+      "grad_norm": 0.8608989930197569,
+      "learning_rate": 4.317380432025428e-06,
+      "loss": 0.6961,
+      "step": 3648
+    },
+    {
+      "epoch": 0.7020007695267411,
+      "grad_norm": 0.8547826410133589,
+      "learning_rate": 4.312254497945595e-06,
+      "loss": 0.788,
+      "step": 3649
+    },
+    {
+      "epoch": 0.7021931512120047,
+      "grad_norm": 0.8735513731928284,
+      "learning_rate": 4.3071307720444015e-06,
+      "loss": 0.7955,
+      "step": 3650
+    },
+    {
+      "epoch": 0.7023855328972681,
+      "grad_norm": 0.9604392659457294,
+      "learning_rate": 4.3020092563110485e-06,
+      "loss": 0.7642,
+      "step": 3651
+    },
+    {
+      "epoch": 0.7025779145825317,
+      "grad_norm": 0.7998162829096219,
+      "learning_rate": 4.2968899527338984e-06,
+      "loss": 0.8054,
+      "step": 3652
+    },
+    {
+      "epoch": 0.7027702962677953,
+      "grad_norm": 0.9457236354666541,
+      "learning_rate": 4.291772863300428e-06,
+      "loss": 0.8588,
+      "step": 3653
+    },
+    {
+      "epoch": 0.7029626779530589,
+      "grad_norm": 0.8771544928630766,
+      "learning_rate": 4.2866579899972686e-06,
+      "loss": 0.7473,
+      "step": 3654
+    },
+    {
+      "epoch": 0.7031550596383225,
+      "grad_norm": 0.876686920690471,
+      "learning_rate": 4.281545334810201e-06,
+      "loss": 0.8634,
+      "step": 3655
+    },
+    {
+      "epoch": 0.703347441323586,
+      "grad_norm": 0.8231908487774987,
+      "learning_rate": 4.276434899724119e-06,
+      "loss": 0.7556,
+      "step": 3656
+    },
+    {
+      "epoch": 0.7035398230088495,
+      "grad_norm": 0.7772696693537197,
+      "learning_rate": 4.27132668672308e-06,
+      "loss": 0.8131,
+      "step": 3657
+    },
+    {
+      "epoch": 0.7037322046941131,
+      "grad_norm": 0.882279541593819,
+      "learning_rate": 4.266220697790266e-06,
+      "loss": 0.8264,
+      "step": 3658
+    },
+    {
+      "epoch": 0.7039245863793767,
+      "grad_norm": 0.8400225642959428,
+      "learning_rate": 4.2611169349079985e-06,
+      "loss": 0.8045,
+      "step": 3659
+    },
+    {
+      "epoch": 0.7041169680646402,
+      "grad_norm": 0.9489275783461545,
+      "learning_rate": 4.25601540005773e-06,
+      "loss": 0.8263,
+      "step": 3660
+    },
+    {
+      "epoch": 0.7043093497499038,
+      "grad_norm": 1.0061206192051722,
+      "learning_rate": 4.250916095220056e-06,
+      "loss": 0.764,
+      "step": 3661
+    },
+    {
+      "epoch": 0.7045017314351674,
+      "grad_norm": 0.8291280832830611,
+      "learning_rate": 4.2458190223747e-06,
+      "loss": 0.8431,
+      "step": 3662
+    },
+    {
+      "epoch": 0.704694113120431,
+      "grad_norm": 0.9812407667702634,
+      "learning_rate": 4.240724183500518e-06,
+      "loss": 0.8514,
+      "step": 3663
+    },
+    {
+      "epoch": 0.7048864948056945,
+      "grad_norm": 0.8781793945098119,
+      "learning_rate": 4.2356315805755135e-06,
+      "loss": 0.8363,
+      "step": 3664
+    },
+    {
+      "epoch": 0.705078876490958,
+      "grad_norm": 0.9360706197504766,
+      "learning_rate": 4.230541215576798e-06,
+      "loss": 0.8358,
+      "step": 3665
+    },
+    {
+      "epoch": 0.7052712581762216,
+      "grad_norm": 0.6538089206805819,
+      "learning_rate": 4.225453090480631e-06,
+      "loss": 0.3697,
+      "step": 3666
+    },
+    {
+      "epoch": 0.7054636398614852,
+      "grad_norm": 0.8881745381343674,
+      "learning_rate": 4.220367207262398e-06,
+      "loss": 0.8185,
+      "step": 3667
+    },
+    {
+      "epoch": 0.7056560215467488,
+      "grad_norm": 0.8829278090782148,
+      "learning_rate": 4.21528356789661e-06,
+      "loss": 0.8373,
+      "step": 3668
+    },
+    {
+      "epoch": 0.7058484032320124,
+      "grad_norm": 0.8134865986174589,
+      "learning_rate": 4.210202174356922e-06,
+      "loss": 0.7697,
+      "step": 3669
+    },
+    {
+      "epoch": 0.7060407849172758,
+      "grad_norm": 0.9485408252677925,
+      "learning_rate": 4.20512302861609e-06,
+      "loss": 0.8137,
+      "step": 3670
+    },
+    {
+      "epoch": 0.7062331666025394,
+      "grad_norm": 0.9067440272847274,
+      "learning_rate": 4.2000461326460274e-06,
+      "loss": 0.8112,
+      "step": 3671
+    },
+    {
+      "epoch": 0.706425548287803,
+      "grad_norm": 0.9236876052499351,
+      "learning_rate": 4.194971488417753e-06,
+      "loss": 0.8587,
+      "step": 3672
+    },
+    {
+      "epoch": 0.7066179299730666,
+      "grad_norm": 0.9413846302973349,
+      "learning_rate": 4.189899097901421e-06,
+      "loss": 0.8359,
+      "step": 3673
+    },
+    {
+      "epoch": 0.7068103116583301,
+      "grad_norm": 0.9018301683199332,
+      "learning_rate": 4.184828963066305e-06,
+      "loss": 0.7991,
+      "step": 3674
+    },
+    {
+      "epoch": 0.7070026933435937,
+      "grad_norm": 0.9275459496130362,
+      "learning_rate": 4.179761085880809e-06,
+      "loss": 0.7744,
+      "step": 3675
+    },
+    {
+      "epoch": 0.7071950750288573,
+      "grad_norm": 0.853754362604457,
+      "learning_rate": 4.174695468312456e-06,
+      "loss": 0.7906,
+      "step": 3676
+    },
+    {
+      "epoch": 0.7073874567141208,
+      "grad_norm": 0.8818126541789394,
+      "learning_rate": 4.16963211232789e-06,
+      "loss": 0.8414,
+      "step": 3677
+    },
+    {
+      "epoch": 0.7075798383993844,
+      "grad_norm": 0.8884100351438186,
+      "learning_rate": 4.16457101989289e-06,
+      "loss": 0.7788,
+      "step": 3678
+    },
+    {
+      "epoch": 0.7077722200846479,
+      "grad_norm": 0.8929662255979681,
+      "learning_rate": 4.159512192972337e-06,
+      "loss": 0.7846,
+      "step": 3679
+    },
+    {
+      "epoch": 0.7079646017699115,
+      "grad_norm": 0.5730570629353559,
+      "learning_rate": 4.15445563353024e-06,
+      "loss": 0.315,
+      "step": 3680
+    },
+    {
+      "epoch": 0.7081569834551751,
+      "grad_norm": 0.7873220099071782,
+      "learning_rate": 4.149401343529742e-06,
+      "loss": 0.7522,
+      "step": 3681
+    },
+    {
+      "epoch": 0.7083493651404387,
+      "grad_norm": 0.8164350379114531,
+      "learning_rate": 4.144349324933077e-06,
+      "loss": 0.7639,
+      "step": 3682
+    },
+    {
+      "epoch": 0.7085417468257021,
+      "grad_norm": 0.9555709889087023,
+      "learning_rate": 4.139299579701623e-06,
+      "loss": 0.7385,
+      "step": 3683
+    },
+    {
+      "epoch": 0.7087341285109657,
+      "grad_norm": 0.8812863123454072,
+      "learning_rate": 4.134252109795863e-06,
+      "loss": 0.8377,
+      "step": 3684
+    },
+    {
+      "epoch": 0.7089265101962293,
+      "grad_norm": 0.8358017576236567,
+      "learning_rate": 4.129206917175397e-06,
+      "loss": 0.8262,
+      "step": 3685
+    },
+    {
+      "epoch": 0.7091188918814929,
+      "grad_norm": 0.9778135767546106,
+      "learning_rate": 4.124164003798944e-06,
+      "loss": 0.8345,
+      "step": 3686
+    },
+    {
+      "epoch": 0.7093112735667565,
+      "grad_norm": 0.8877297306813341,
+      "learning_rate": 4.119123371624335e-06,
+      "loss": 0.792,
+      "step": 3687
+    },
+    {
+      "epoch": 0.70950365525202,
+      "grad_norm": 1.0144800062800818,
+      "learning_rate": 4.114085022608517e-06,
+      "loss": 0.8595,
+      "step": 3688
+    },
+    {
+      "epoch": 0.7096960369372836,
+      "grad_norm": 1.018108770893766,
+      "learning_rate": 4.109048958707552e-06,
+      "loss": 0.7858,
+      "step": 3689
+    },
+    {
+      "epoch": 0.7098884186225471,
+      "grad_norm": 0.9118638853184164,
+      "learning_rate": 4.104015181876613e-06,
+      "loss": 0.7799,
+      "step": 3690
+    },
+    {
+      "epoch": 0.7100808003078107,
+      "grad_norm": 0.9351192700551785,
+      "learning_rate": 4.09898369406998e-06,
+      "loss": 0.8393,
+      "step": 3691
+    },
+    {
+      "epoch": 0.7102731819930742,
+      "grad_norm": 0.8086384963152423,
+      "learning_rate": 4.0939544972410636e-06,
+      "loss": 0.7904,
+      "step": 3692
+    },
+    {
+      "epoch": 0.7104655636783378,
+      "grad_norm": 0.8200093450606946,
+      "learning_rate": 4.0889275933423576e-06,
+      "loss": 0.7626,
+      "step": 3693
+    },
+    {
+      "epoch": 0.7106579453636014,
+      "grad_norm": 0.8661871096937034,
+      "learning_rate": 4.0839029843254815e-06,
+      "loss": 0.7792,
+      "step": 3694
+    },
+    {
+      "epoch": 0.710850327048865,
+      "grad_norm": 0.8672406613118202,
+      "learning_rate": 4.078880672141171e-06,
+      "loss": 0.8598,
+      "step": 3695
+    },
+    {
+      "epoch": 0.7110427087341286,
+      "grad_norm": 0.8655636200707267,
+      "learning_rate": 4.073860658739246e-06,
+      "loss": 0.7699,
+      "step": 3696
+    },
+    {
+      "epoch": 0.711235090419392,
+      "grad_norm": 0.7603700196534223,
+      "learning_rate": 4.068842946068661e-06,
+      "loss": 0.7515,
+      "step": 3697
+    },
+    {
+      "epoch": 0.7114274721046556,
+      "grad_norm": 0.8953239660650684,
+      "learning_rate": 4.063827536077459e-06,
+      "loss": 0.7849,
+      "step": 3698
+    },
+    {
+      "epoch": 0.7116198537899192,
+      "grad_norm": 0.9127938308900941,
+      "learning_rate": 4.058814430712796e-06,
+      "loss": 0.8254,
+      "step": 3699
+    },
+    {
+      "epoch": 0.7118122354751828,
+      "grad_norm": 0.8749951110893726,
+      "learning_rate": 4.0538036319209325e-06,
+      "loss": 0.8076,
+      "step": 3700
+    },
+    {
+      "epoch": 0.7120046171604463,
+      "grad_norm": 0.9039864098168661,
+      "learning_rate": 4.0487951416472324e-06,
+      "loss": 0.8455,
+      "step": 3701
+    },
+    {
+      "epoch": 0.7121969988457099,
+      "grad_norm": 0.8825925951043,
+      "learning_rate": 4.043788961836164e-06,
+      "loss": 0.8082,
+      "step": 3702
+    },
+    {
+      "epoch": 0.7123893805309734,
+      "grad_norm": 0.5853218061850628,
+      "learning_rate": 4.038785094431295e-06,
+      "loss": 0.3118,
+      "step": 3703
+    },
+    {
+      "epoch": 0.712581762216237,
+      "grad_norm": 0.8469940055863717,
+      "learning_rate": 4.0337835413753116e-06,
+      "loss": 0.816,
+      "step": 3704
+    },
+    {
+      "epoch": 0.7127741439015006,
+      "grad_norm": 0.9498589873733675,
+      "learning_rate": 4.0287843046099765e-06,
+      "loss": 0.8104,
+      "step": 3705
+    },
+    {
+      "epoch": 0.7129665255867641,
+      "grad_norm": 0.9225209245940793,
+      "learning_rate": 4.0237873860761645e-06,
+      "loss": 0.7839,
+      "step": 3706
+    },
+    {
+      "epoch": 0.7131589072720277,
+      "grad_norm": 0.8452159143873931,
+      "learning_rate": 4.018792787713865e-06,
+      "loss": 0.7885,
+      "step": 3707
+    },
+    {
+      "epoch": 0.7133512889572913,
+      "grad_norm": 0.8815967333455026,
+      "learning_rate": 4.013800511462135e-06,
+      "loss": 0.7692,
+      "step": 3708
+    },
+    {
+      "epoch": 0.7135436706425549,
+      "grad_norm": 0.7964766107787614,
+      "learning_rate": 4.008810559259162e-06,
+      "loss": 0.7949,
+      "step": 3709
+    },
+    {
+      "epoch": 0.7137360523278183,
+      "grad_norm": 0.8570852674533307,
+      "learning_rate": 4.003822933042213e-06,
+      "loss": 0.7764,
+      "step": 3710
+    },
+    {
+      "epoch": 0.7139284340130819,
+      "grad_norm": 0.8953899418876603,
+      "learning_rate": 3.998837634747655e-06,
+      "loss": 0.7856,
+      "step": 3711
+    },
+    {
+      "epoch": 0.7141208156983455,
+      "grad_norm": 0.8910887689758684,
+      "learning_rate": 3.993854666310955e-06,
+      "loss": 0.8495,
+      "step": 3712
+    },
+    {
+      "epoch": 0.7143131973836091,
+      "grad_norm": 0.8963126066621911,
+      "learning_rate": 3.98887402966667e-06,
+      "loss": 0.785,
+      "step": 3713
+    },
+    {
+      "epoch": 0.7145055790688727,
+      "grad_norm": 0.9107112969348842,
+      "learning_rate": 3.983895726748455e-06,
+      "loss": 0.8233,
+      "step": 3714
+    },
+    {
+      "epoch": 0.7146979607541362,
+      "grad_norm": 0.878432597527445,
+      "learning_rate": 3.97891975948906e-06,
+      "loss": 0.7913,
+      "step": 3715
+    },
+    {
+      "epoch": 0.7148903424393998,
+      "grad_norm": 0.8858378431396968,
+      "learning_rate": 3.973946129820326e-06,
+      "loss": 0.7123,
+      "step": 3716
+    },
+    {
+      "epoch": 0.7150827241246633,
+      "grad_norm": 0.8863505493092493,
+      "learning_rate": 3.968974839673186e-06,
+      "loss": 0.8145,
+      "step": 3717
+    },
+    {
+      "epoch": 0.7152751058099269,
+      "grad_norm": 0.5737575223937432,
+      "learning_rate": 3.964005890977672e-06,
+      "loss": 0.3283,
+      "step": 3718
+    },
+    {
+      "epoch": 0.7154674874951905,
+      "grad_norm": 0.8960222027508334,
+      "learning_rate": 3.9590392856628946e-06,
+      "loss": 0.7772,
+      "step": 3719
+    },
+    {
+      "epoch": 0.715659869180454,
+      "grad_norm": 0.8385357501282406,
+      "learning_rate": 3.954075025657058e-06,
+      "loss": 0.828,
+      "step": 3720
+    },
+    {
+      "epoch": 0.7158522508657176,
+      "grad_norm": 0.839768347911263,
+      "learning_rate": 3.949113112887471e-06,
+      "loss": 0.8734,
+      "step": 3721
+    },
+    {
+      "epoch": 0.7160446325509812,
+      "grad_norm": 0.8232233316279501,
+      "learning_rate": 3.944153549280506e-06,
+      "loss": 0.7709,
+      "step": 3722
+    },
+    {
+      "epoch": 0.7162370142362448,
+      "grad_norm": 0.8741883633474659,
+      "learning_rate": 3.939196336761645e-06,
+      "loss": 0.8346,
+      "step": 3723
+    },
+    {
+      "epoch": 0.7164293959215082,
+      "grad_norm": 0.9010833816800508,
+      "learning_rate": 3.934241477255445e-06,
+      "loss": 0.7579,
+      "step": 3724
+    },
+    {
+      "epoch": 0.7166217776067718,
+      "grad_norm": 0.9405305213772873,
+      "learning_rate": 3.929288972685555e-06,
+      "loss": 0.8567,
+      "step": 3725
+    },
+    {
+      "epoch": 0.7168141592920354,
+      "grad_norm": 0.6315964606936392,
+      "learning_rate": 3.924338824974705e-06,
+      "loss": 0.3221,
+      "step": 3726
+    },
+    {
+      "epoch": 0.717006540977299,
+      "grad_norm": 0.7995940803522966,
+      "learning_rate": 3.919391036044715e-06,
+      "loss": 0.7437,
+      "step": 3727
+    },
+    {
+      "epoch": 0.7171989226625626,
+      "grad_norm": 0.8485378191998214,
+      "learning_rate": 3.914445607816486e-06,
+      "loss": 0.8123,
+      "step": 3728
+    },
+    {
+      "epoch": 0.717391304347826,
+      "grad_norm": 0.8434066304931973,
+      "learning_rate": 3.909502542210001e-06,
+      "loss": 0.7641,
+      "step": 3729
+    },
+    {
+      "epoch": 0.7175836860330896,
+      "grad_norm": 0.8903797785458012,
+      "learning_rate": 3.904561841144338e-06,
+      "loss": 0.8052,
+      "step": 3730
+    },
+    {
+      "epoch": 0.7177760677183532,
+      "grad_norm": 0.9490456869048902,
+      "learning_rate": 3.899623506537635e-06,
+      "loss": 0.7998,
+      "step": 3731
+    },
+    {
+      "epoch": 0.7179684494036168,
+      "grad_norm": 0.8507767016934469,
+      "learning_rate": 3.894687540307127e-06,
+      "loss": 0.884,
+      "step": 3732
+    },
+    {
+      "epoch": 0.7181608310888803,
+      "grad_norm": 0.8549233010044046,
+      "learning_rate": 3.8897539443691355e-06,
+      "loss": 0.8349,
+      "step": 3733
+    },
+    {
+      "epoch": 0.7183532127741439,
+      "grad_norm": 0.8830298140721862,
+      "learning_rate": 3.884822720639036e-06,
+      "loss": 0.7658,
+      "step": 3734
+    },
+    {
+      "epoch": 0.7185455944594075,
+      "grad_norm": 0.922645618979009,
+      "learning_rate": 3.879893871031314e-06,
+      "loss": 0.784,
+      "step": 3735
+    },
+    {
+      "epoch": 0.718737976144671,
+      "grad_norm": 0.8937079493280837,
+      "learning_rate": 3.874967397459511e-06,
+      "loss": 0.7931,
+      "step": 3736
+    },
+    {
+      "epoch": 0.7189303578299346,
+      "grad_norm": 0.9407743866737908,
+      "learning_rate": 3.870043301836256e-06,
+      "loss": 0.8455,
+      "step": 3737
+    },
+    {
+      "epoch": 0.7191227395151981,
+      "grad_norm": 0.965527855179053,
+      "learning_rate": 3.86512158607325e-06,
+      "loss": 0.838,
+      "step": 3738
+    },
+    {
+      "epoch": 0.7193151212004617,
+      "grad_norm": 0.8797819538642561,
+      "learning_rate": 3.860202252081276e-06,
+      "loss": 0.7796,
+      "step": 3739
+    },
+    {
+      "epoch": 0.7195075028857253,
+      "grad_norm": 0.799443144449611,
+      "learning_rate": 3.855285301770188e-06,
+      "loss": 0.7571,
+      "step": 3740
+    },
+    {
+      "epoch": 0.7196998845709889,
+      "grad_norm": 0.8901727392958076,
+      "learning_rate": 3.850370737048913e-06,
+      "loss": 0.7575,
+      "step": 3741
+    },
+    {
+      "epoch": 0.7198922662562524,
+      "grad_norm": 0.891839635466915,
+      "learning_rate": 3.8454585598254565e-06,
+      "loss": 0.766,
+      "step": 3742
+    },
+    {
+      "epoch": 0.7200846479415159,
+      "grad_norm": 0.7801006798445932,
+      "learning_rate": 3.840548772006891e-06,
+      "loss": 0.792,
+      "step": 3743
+    },
+    {
+      "epoch": 0.7202770296267795,
+      "grad_norm": 0.7989690221048831,
+      "learning_rate": 3.835641375499375e-06,
+      "loss": 0.8215,
+      "step": 3744
+    },
+    {
+      "epoch": 0.7204694113120431,
+      "grad_norm": 0.9411837544462174,
+      "learning_rate": 3.830736372208118e-06,
+      "loss": 0.7975,
+      "step": 3745
+    },
+    {
+      "epoch": 0.7206617929973067,
+      "grad_norm": 0.9392307050231288,
+      "learning_rate": 3.8258337640374125e-06,
+      "loss": 0.7574,
+      "step": 3746
+    },
+    {
+      "epoch": 0.7208541746825702,
+      "grad_norm": 0.8453585922112955,
+      "learning_rate": 3.820933552890629e-06,
+      "loss": 0.777,
+      "step": 3747
+    },
+    {
+      "epoch": 0.7210465563678338,
+      "grad_norm": 0.8578135108433788,
+      "learning_rate": 3.816035740670185e-06,
+      "loss": 0.7924,
+      "step": 3748
+    },
+    {
+      "epoch": 0.7212389380530974,
+      "grad_norm": 0.6681794399115978,
+      "learning_rate": 3.811140329277591e-06,
+      "loss": 0.3353,
+      "step": 3749
+    },
+    {
+      "epoch": 0.7214313197383609,
+      "grad_norm": 0.8685556698472521,
+      "learning_rate": 3.8062473206134088e-06,
+      "loss": 0.8306,
+      "step": 3750
+    },
+    {
+      "epoch": 0.7216237014236244,
+      "grad_norm": 0.8389977015822477,
+      "learning_rate": 3.8013567165772735e-06,
+      "loss": 0.7858,
+      "step": 3751
+    },
+    {
+      "epoch": 0.721816083108888,
+      "grad_norm": 0.8994161315427553,
+      "learning_rate": 3.7964685190678874e-06,
+      "loss": 0.7902,
+      "step": 3752
+    },
+    {
+      "epoch": 0.7220084647941516,
+      "grad_norm": 0.8768742438108849,
+      "learning_rate": 3.7915827299830154e-06,
+      "loss": 0.7546,
+      "step": 3753
+    },
+    {
+      "epoch": 0.7222008464794152,
+      "grad_norm": 0.9613743428180657,
+      "learning_rate": 3.7866993512194895e-06,
+      "loss": 0.7974,
+      "step": 3754
+    },
+    {
+      "epoch": 0.7223932281646788,
+      "grad_norm": 0.8162862338301269,
+      "learning_rate": 3.7818183846732024e-06,
+      "loss": 0.6824,
+      "step": 3755
+    },
+    {
+      "epoch": 0.7225856098499422,
+      "grad_norm": 0.864024086924796,
+      "learning_rate": 3.776939832239125e-06,
+      "loss": 0.7387,
+      "step": 3756
+    },
+    {
+      "epoch": 0.7227779915352058,
+      "grad_norm": 0.8887126565813567,
+      "learning_rate": 3.7720636958112623e-06,
+      "loss": 0.8194,
+      "step": 3757
+    },
+    {
+      "epoch": 0.7229703732204694,
+      "grad_norm": 0.8761689521390986,
+      "learning_rate": 3.7671899772827113e-06,
+      "loss": 0.7674,
+      "step": 3758
+    },
+    {
+      "epoch": 0.723162754905733,
+      "grad_norm": 0.8353153201977478,
+      "learning_rate": 3.7623186785456156e-06,
+      "loss": 0.7575,
+      "step": 3759
+    },
+    {
+      "epoch": 0.7233551365909965,
+      "grad_norm": 0.8415976151781764,
+      "learning_rate": 3.757449801491172e-06,
+      "loss": 0.695,
+      "step": 3760
+    },
+    {
+      "epoch": 0.7235475182762601,
+      "grad_norm": 0.8977208934681827,
+      "learning_rate": 3.7525833480096575e-06,
+      "loss": 0.8246,
+      "step": 3761
+    },
+    {
+      "epoch": 0.7237398999615237,
+      "grad_norm": 0.9153155651382715,
+      "learning_rate": 3.7477193199903903e-06,
+      "loss": 0.8285,
+      "step": 3762
+    },
+    {
+      "epoch": 0.7239322816467872,
+      "grad_norm": 0.8201780284298487,
+      "learning_rate": 3.7428577193217563e-06,
+      "loss": 0.8083,
+      "step": 3763
+    },
+    {
+      "epoch": 0.7241246633320508,
+      "grad_norm": 0.8611367378272043,
+      "learning_rate": 3.737998547891195e-06,
+      "loss": 0.823,
+      "step": 3764
+    },
+    {
+      "epoch": 0.7243170450173143,
+      "grad_norm": 0.9275767902287502,
+      "learning_rate": 3.7331418075852053e-06,
+      "loss": 0.8335,
+      "step": 3765
+    },
+    {
+      "epoch": 0.7245094267025779,
+      "grad_norm": 0.9254875904163359,
+      "learning_rate": 3.728287500289339e-06,
+      "loss": 0.7952,
+      "step": 3766
+    },
+    {
+      "epoch": 0.7247018083878415,
+      "grad_norm": 0.8429393903547647,
+      "learning_rate": 3.7234356278882076e-06,
+      "loss": 0.7689,
+      "step": 3767
+    },
+    {
+      "epoch": 0.7248941900731051,
+      "grad_norm": 0.8802196877574507,
+      "learning_rate": 3.718586192265473e-06,
+      "loss": 0.8737,
+      "step": 3768
+    },
+    {
+      "epoch": 0.7250865717583687,
+      "grad_norm": 1.009239293570677,
+      "learning_rate": 3.7137391953038516e-06,
+      "loss": 0.7707,
+      "step": 3769
+    },
+    {
+      "epoch": 0.7252789534436321,
+      "grad_norm": 0.9309409448928583,
+      "learning_rate": 3.7088946388851223e-06,
+      "loss": 0.7706,
+      "step": 3770
+    },
+    {
+      "epoch": 0.7254713351288957,
+      "grad_norm": 0.9141101532140833,
+      "learning_rate": 3.7040525248901003e-06,
+      "loss": 0.8518,
+      "step": 3771
+    },
+    {
+      "epoch": 0.7256637168141593,
+      "grad_norm": 0.867571840093179,
+      "learning_rate": 3.6992128551986617e-06,
+      "loss": 0.8109,
+      "step": 3772
+    },
+    {
+      "epoch": 0.7258560984994229,
+      "grad_norm": 0.8016380827981131,
+      "learning_rate": 3.6943756316897406e-06,
+      "loss": 0.8146,
+      "step": 3773
+    },
+    {
+      "epoch": 0.7260484801846864,
+      "grad_norm": 0.8807747606169686,
+      "learning_rate": 3.6895408562413027e-06,
+      "loss": 0.8028,
+      "step": 3774
+    },
+    {
+      "epoch": 0.72624086186995,
+      "grad_norm": 0.9367963953372366,
+      "learning_rate": 3.684708530730382e-06,
+      "loss": 0.8053,
+      "step": 3775
+    },
+    {
+      "epoch": 0.7264332435552135,
+      "grad_norm": 0.8863556188354325,
+      "learning_rate": 3.6798786570330526e-06,
+      "loss": 0.7761,
+      "step": 3776
+    },
+    {
+      "epoch": 0.7266256252404771,
+      "grad_norm": 0.9341994269192864,
+      "learning_rate": 3.6750512370244363e-06,
+      "loss": 0.7942,
+      "step": 3777
+    },
+    {
+      "epoch": 0.7268180069257407,
+      "grad_norm": 0.948758209951495,
+      "learning_rate": 3.670226272578704e-06,
+      "loss": 0.7994,
+      "step": 3778
+    },
+    {
+      "epoch": 0.7270103886110042,
+      "grad_norm": 0.879926714174962,
+      "learning_rate": 3.6654037655690732e-06,
+      "loss": 0.8405,
+      "step": 3779
+    },
+    {
+      "epoch": 0.7272027702962678,
+      "grad_norm": 0.8531062627633448,
+      "learning_rate": 3.660583717867807e-06,
+      "loss": 0.8611,
+      "step": 3780
+    },
+    {
+      "epoch": 0.7273951519815314,
+      "grad_norm": 0.8965096037993485,
+      "learning_rate": 3.655766131346211e-06,
+      "loss": 0.872,
+      "step": 3781
+    },
+    {
+      "epoch": 0.727587533666795,
+      "grad_norm": 0.8695689435766633,
+      "learning_rate": 3.650951007874648e-06,
+      "loss": 0.8093,
+      "step": 3782
+    },
+    {
+      "epoch": 0.7277799153520584,
+      "grad_norm": 0.964908456623004,
+      "learning_rate": 3.6461383493225012e-06,
+      "loss": 0.7691,
+      "step": 3783
+    },
+    {
+      "epoch": 0.727972297037322,
+      "grad_norm": 0.9314081860269717,
+      "learning_rate": 3.6413281575582194e-06,
+      "loss": 0.8275,
+      "step": 3784
+    },
+    {
+      "epoch": 0.7281646787225856,
+      "grad_norm": 0.9019126900211153,
+      "learning_rate": 3.6365204344492867e-06,
+      "loss": 0.7503,
+      "step": 3785
+    },
+    {
+      "epoch": 0.7283570604078492,
+      "grad_norm": 0.8931854818321364,
+      "learning_rate": 3.6317151818622154e-06,
+      "loss": 0.7827,
+      "step": 3786
+    },
+    {
+      "epoch": 0.7285494420931128,
+      "grad_norm": 0.8879991780749673,
+      "learning_rate": 3.62691240166258e-06,
+      "loss": 0.8353,
+      "step": 3787
+    },
+    {
+      "epoch": 0.7287418237783763,
+      "grad_norm": 0.9191814193590424,
+      "learning_rate": 3.6221120957149826e-06,
+      "loss": 0.7792,
+      "step": 3788
+    },
+    {
+      "epoch": 0.7289342054636399,
+      "grad_norm": 0.9755865055471803,
+      "learning_rate": 3.617314265883066e-06,
+      "loss": 0.8113,
+      "step": 3789
+    },
+    {
+      "epoch": 0.7291265871489034,
+      "grad_norm": 0.808915069693077,
+      "learning_rate": 3.612518914029515e-06,
+      "loss": 0.8229,
+      "step": 3790
+    },
+    {
+      "epoch": 0.729318968834167,
+      "grad_norm": 0.9154201876389944,
+      "learning_rate": 3.6077260420160487e-06,
+      "loss": 0.8323,
+      "step": 3791
+    },
+    {
+      "epoch": 0.7295113505194305,
+      "grad_norm": 0.9086575796844075,
+      "learning_rate": 3.602935651703424e-06,
+      "loss": 0.7689,
+      "step": 3792
+    },
+    {
+      "epoch": 0.7297037322046941,
+      "grad_norm": 0.8797358534186978,
+      "learning_rate": 3.598147744951438e-06,
+      "loss": 0.8141,
+      "step": 3793
+    },
+    {
+      "epoch": 0.7298961138899577,
+      "grad_norm": 0.9079398893398142,
+      "learning_rate": 3.5933623236189198e-06,
+      "loss": 0.816,
+      "step": 3794
+    },
+    {
+      "epoch": 0.7300884955752213,
+      "grad_norm": 0.6452422168939768,
+      "learning_rate": 3.58857938956373e-06,
+      "loss": 0.3232,
+      "step": 3795
+    },
+    {
+      "epoch": 0.7302808772604849,
+      "grad_norm": 0.9258621636623925,
+      "learning_rate": 3.58379894464278e-06,
+      "loss": 0.8192,
+      "step": 3796
+    },
+    {
+      "epoch": 0.7304732589457483,
+      "grad_norm": 0.9865170719929216,
+      "learning_rate": 3.57902099071199e-06,
+      "loss": 0.8566,
+      "step": 3797
+    },
+    {
+      "epoch": 0.7306656406310119,
+      "grad_norm": 0.870054861759254,
+      "learning_rate": 3.5742455296263346e-06,
+      "loss": 0.8202,
+      "step": 3798
+    },
+    {
+      "epoch": 0.7308580223162755,
+      "grad_norm": 0.8010962597690916,
+      "learning_rate": 3.569472563239814e-06,
+      "loss": 0.7107,
+      "step": 3799
+    },
+    {
+      "epoch": 0.7310504040015391,
+      "grad_norm": 0.9055755926973155,
+      "learning_rate": 3.5647020934054465e-06,
+      "loss": 0.753,
+      "step": 3800
+    },
+    {
+      "epoch": 0.7312427856868026,
+      "grad_norm": 0.8902505997852312,
+      "learning_rate": 3.559934121975304e-06,
+      "loss": 0.8695,
+      "step": 3801
+    },
+    {
+      "epoch": 0.7314351673720662,
+      "grad_norm": 0.9164448384660259,
+      "learning_rate": 3.5551686508004735e-06,
+      "loss": 0.8066,
+      "step": 3802
+    },
+    {
+      "epoch": 0.7316275490573297,
+      "grad_norm": 0.8981869512216515,
+      "learning_rate": 3.550405681731074e-06,
+      "loss": 0.8195,
+      "step": 3803
+    },
+    {
+      "epoch": 0.7318199307425933,
+      "grad_norm": 0.9047100195584635,
+      "learning_rate": 3.5456452166162547e-06,
+      "loss": 0.7952,
+      "step": 3804
+    },
+    {
+      "epoch": 0.7320123124278569,
+      "grad_norm": 0.9397666211853257,
+      "learning_rate": 3.540887257304193e-06,
+      "loss": 0.8205,
+      "step": 3805
+    },
+    {
+      "epoch": 0.7322046941131204,
+      "grad_norm": 0.9987204890128047,
+      "learning_rate": 3.5361318056420925e-06,
+      "loss": 0.8029,
+      "step": 3806
+    },
+    {
+      "epoch": 0.732397075798384,
+      "grad_norm": 0.6079153858530555,
+      "learning_rate": 3.531378863476178e-06,
+      "loss": 0.324,
+      "step": 3807
+    },
+    {
+      "epoch": 0.7325894574836476,
+      "grad_norm": 0.9339020437197055,
+      "learning_rate": 3.5266284326517165e-06,
+      "loss": 0.8078,
+      "step": 3808
+    },
+    {
+      "epoch": 0.7327818391689112,
+      "grad_norm": 0.8155577503152659,
+      "learning_rate": 3.5218805150129755e-06,
+      "loss": 0.7894,
+      "step": 3809
+    },
+    {
+      "epoch": 0.7329742208541746,
+      "grad_norm": 0.5971792718469997,
+      "learning_rate": 3.5171351124032703e-06,
+      "loss": 0.3369,
+      "step": 3810
+    },
+    {
+      "epoch": 0.7331666025394382,
+      "grad_norm": 0.9472980345593237,
+      "learning_rate": 3.51239222666493e-06,
+      "loss": 0.8089,
+      "step": 3811
+    },
+    {
+      "epoch": 0.7333589842247018,
+      "grad_norm": 0.8624599726126105,
+      "learning_rate": 3.507651859639295e-06,
+      "loss": 0.7896,
+      "step": 3812
+    },
+    {
+      "epoch": 0.7335513659099654,
+      "grad_norm": 0.8830048321233243,
+      "learning_rate": 3.5029140131667493e-06,
+      "loss": 0.7804,
+      "step": 3813
+    },
+    {
+      "epoch": 0.733743747595229,
+      "grad_norm": 0.8756903700696468,
+      "learning_rate": 3.4981786890866853e-06,
+      "loss": 0.7911,
+      "step": 3814
+    },
+    {
+      "epoch": 0.7339361292804925,
+      "grad_norm": 0.8438496797055104,
+      "learning_rate": 3.493445889237518e-06,
+      "loss": 0.8006,
+      "step": 3815
+    },
+    {
+      "epoch": 0.734128510965756,
+      "grad_norm": 0.814658791021474,
+      "learning_rate": 3.4887156154566847e-06,
+      "loss": 0.8298,
+      "step": 3816
+    },
+    {
+      "epoch": 0.7343208926510196,
+      "grad_norm": 0.943150508246111,
+      "learning_rate": 3.4839878695806385e-06,
+      "loss": 0.8131,
+      "step": 3817
+    },
+    {
+      "epoch": 0.7345132743362832,
+      "grad_norm": 0.8898099861924511,
+      "learning_rate": 3.4792626534448547e-06,
+      "loss": 0.8575,
+      "step": 3818
+    },
+    {
+      "epoch": 0.7347056560215467,
+      "grad_norm": 0.879187055542594,
+      "learning_rate": 3.4745399688838243e-06,
+      "loss": 0.8041,
+      "step": 3819
+    },
+    {
+      "epoch": 0.7348980377068103,
+      "grad_norm": 0.8279240640699425,
+      "learning_rate": 3.469819817731056e-06,
+      "loss": 0.8002,
+      "step": 3820
+    },
+    {
+      "epoch": 0.7350904193920739,
+      "grad_norm": 0.8806858705339966,
+      "learning_rate": 3.4651022018190715e-06,
+      "loss": 0.8086,
+      "step": 3821
+    },
+    {
+      "epoch": 0.7352828010773375,
+      "grad_norm": 0.8382618048878822,
+      "learning_rate": 3.460387122979423e-06,
+      "loss": 0.7949,
+      "step": 3822
+    },
+    {
+      "epoch": 0.735475182762601,
+      "grad_norm": 0.6027920535245088,
+      "learning_rate": 3.455674583042652e-06,
+      "loss": 0.3541,
+      "step": 3823
+    },
+    {
+      "epoch": 0.7356675644478645,
+      "grad_norm": 0.8152918745499881,
+      "learning_rate": 3.4509645838383386e-06,
+      "loss": 0.7782,
+      "step": 3824
+    },
+    {
+      "epoch": 0.7358599461331281,
+      "grad_norm": 0.9115623013756979,
+      "learning_rate": 3.4462571271950674e-06,
+      "loss": 0.7979,
+      "step": 3825
+    },
+    {
+      "epoch": 0.7360523278183917,
+      "grad_norm": 0.84513545459806,
+      "learning_rate": 3.4415522149404233e-06,
+      "loss": 0.8232,
+      "step": 3826
+    },
+    {
+      "epoch": 0.7362447095036553,
+      "grad_norm": 0.8157876278409406,
+      "learning_rate": 3.436849848901028e-06,
+      "loss": 0.8383,
+      "step": 3827
+    },
+    {
+      "epoch": 0.7364370911889189,
+      "grad_norm": 0.847447520255758,
+      "learning_rate": 3.432150030902497e-06,
+      "loss": 0.8169,
+      "step": 3828
+    },
+    {
+      "epoch": 0.7366294728741823,
+      "grad_norm": 0.9296541942836075,
+      "learning_rate": 3.427452762769462e-06,
+      "loss": 0.7839,
+      "step": 3829
+    },
+    {
+      "epoch": 0.7368218545594459,
+      "grad_norm": 0.6358398665552776,
+      "learning_rate": 3.4227580463255628e-06,
+      "loss": 0.3407,
+      "step": 3830
+    },
+    {
+      "epoch": 0.7370142362447095,
+      "grad_norm": 0.9035342848198308,
+      "learning_rate": 3.4180658833934523e-06,
+      "loss": 0.8049,
+      "step": 3831
+    },
+    {
+      "epoch": 0.7372066179299731,
+      "grad_norm": 0.8640123717323364,
+      "learning_rate": 3.4133762757947873e-06,
+      "loss": 0.8104,
+      "step": 3832
+    },
+    {
+      "epoch": 0.7373989996152366,
+      "grad_norm": 0.8767988642373933,
+      "learning_rate": 3.4086892253502344e-06,
+      "loss": 0.749,
+      "step": 3833
+    },
+    {
+      "epoch": 0.7375913813005002,
+      "grad_norm": 0.9016192875636977,
+      "learning_rate": 3.4040047338794756e-06,
+      "loss": 0.7948,
+      "step": 3834
+    },
+    {
+      "epoch": 0.7377837629857638,
+      "grad_norm": 0.8491419988818808,
+      "learning_rate": 3.3993228032011784e-06,
+      "loss": 0.7702,
+      "step": 3835
+    },
+    {
+      "epoch": 0.7379761446710273,
+      "grad_norm": 0.6065982493331648,
+      "learning_rate": 3.3946434351330415e-06,
+      "loss": 0.3314,
+      "step": 3836
+    },
+    {
+      "epoch": 0.7381685263562909,
+      "grad_norm": 0.8317442663350035,
+      "learning_rate": 3.3899666314917512e-06,
+      "loss": 0.807,
+      "step": 3837
+    },
+    {
+      "epoch": 0.7383609080415544,
+      "grad_norm": 0.876344274554293,
+      "learning_rate": 3.385292394093006e-06,
+      "loss": 0.8217,
+      "step": 3838
+    },
+    {
+      "epoch": 0.738553289726818,
+      "grad_norm": 0.9674392958408837,
+      "learning_rate": 3.3806207247515068e-06,
+      "loss": 0.7815,
+      "step": 3839
+    },
+    {
+      "epoch": 0.7387456714120816,
+      "grad_norm": 0.9501412187563424,
+      "learning_rate": 3.375951625280948e-06,
+      "loss": 0.7653,
+      "step": 3840
+    },
+    {
+      "epoch": 0.7389380530973452,
+      "grad_norm": 0.8849987559306616,
+      "learning_rate": 3.3712850974940437e-06,
+      "loss": 0.7596,
+      "step": 3841
+    },
+    {
+      "epoch": 0.7391304347826086,
+      "grad_norm": 0.919818963070698,
+      "learning_rate": 3.3666211432024974e-06,
+      "loss": 0.8203,
+      "step": 3842
+    },
+    {
+      "epoch": 0.7393228164678722,
+      "grad_norm": 0.8801904269790997,
+      "learning_rate": 3.361959764217018e-06,
+      "loss": 0.852,
+      "step": 3843
+    },
+    {
+      "epoch": 0.7395151981531358,
+      "grad_norm": 0.8297822354742663,
+      "learning_rate": 3.357300962347313e-06,
+      "loss": 0.8063,
+      "step": 3844
+    },
+    {
+      "epoch": 0.7397075798383994,
+      "grad_norm": 0.8368712406136638,
+      "learning_rate": 3.3526447394020887e-06,
+      "loss": 0.8193,
+      "step": 3845
+    },
+    {
+      "epoch": 0.739899961523663,
+      "grad_norm": 0.8805177709035977,
+      "learning_rate": 3.3479910971890516e-06,
+      "loss": 0.8194,
+      "step": 3846
+    },
+    {
+      "epoch": 0.7400923432089265,
+      "grad_norm": 0.8928465691458287,
+      "learning_rate": 3.343340037514903e-06,
+      "loss": 0.8151,
+      "step": 3847
+    },
+    {
+      "epoch": 0.7402847248941901,
+      "grad_norm": 0.9750492425352131,
+      "learning_rate": 3.3386915621853533e-06,
+      "loss": 0.8415,
+      "step": 3848
+    },
+    {
+      "epoch": 0.7404771065794536,
+      "grad_norm": 0.9430610597210126,
+      "learning_rate": 3.3340456730050887e-06,
+      "loss": 0.8142,
+      "step": 3849
+    },
+    {
+      "epoch": 0.7406694882647172,
+      "grad_norm": 0.9435965381410739,
+      "learning_rate": 3.3294023717778122e-06,
+      "loss": 0.8377,
+      "step": 3850
+    },
+    {
+      "epoch": 0.7408618699499807,
+      "grad_norm": 0.8877023535937327,
+      "learning_rate": 3.324761660306215e-06,
+      "loss": 0.7682,
+      "step": 3851
+    },
+    {
+      "epoch": 0.7410542516352443,
+      "grad_norm": 0.6426601278394277,
+      "learning_rate": 3.3201235403919683e-06,
+      "loss": 0.3273,
+      "step": 3852
+    },
+    {
+      "epoch": 0.7412466333205079,
+      "grad_norm": 0.8446521873245759,
+      "learning_rate": 3.3154880138357626e-06,
+      "loss": 0.7907,
+      "step": 3853
+    },
+    {
+      "epoch": 0.7414390150057715,
+      "grad_norm": 0.872495078144355,
+      "learning_rate": 3.3108550824372632e-06,
+      "loss": 0.781,
+      "step": 3854
+    },
+    {
+      "epoch": 0.7416313966910351,
+      "grad_norm": 0.9196788997153229,
+      "learning_rate": 3.306224747995136e-06,
+      "loss": 0.7268,
+      "step": 3855
+    },
+    {
+      "epoch": 0.7418237783762985,
+      "grad_norm": 0.992920160192317,
+      "learning_rate": 3.301597012307034e-06,
+      "loss": 0.8372,
+      "step": 3856
+    },
+    {
+      "epoch": 0.7420161600615621,
+      "grad_norm": 0.825426674030491,
+      "learning_rate": 3.2969718771696047e-06,
+      "loss": 0.7541,
+      "step": 3857
+    },
+    {
+      "epoch": 0.7422085417468257,
+      "grad_norm": 0.9307423729428119,
+      "learning_rate": 3.292349344378486e-06,
+      "loss": 0.7726,
+      "step": 3858
+    },
+    {
+      "epoch": 0.7424009234320893,
+      "grad_norm": 0.8469356088347069,
+      "learning_rate": 3.287729415728298e-06,
+      "loss": 0.791,
+      "step": 3859
+    },
+    {
+      "epoch": 0.7425933051173528,
+      "grad_norm": 0.8232457177122736,
+      "learning_rate": 3.283112093012669e-06,
+      "loss": 0.7596,
+      "step": 3860
+    },
+    {
+      "epoch": 0.7427856868026164,
+      "grad_norm": 0.8936768068076824,
+      "learning_rate": 3.278497378024187e-06,
+      "loss": 0.8563,
+      "step": 3861
+    },
+    {
+      "epoch": 0.74297806848788,
+      "grad_norm": 0.8418390984593817,
+      "learning_rate": 3.2738852725544547e-06,
+      "loss": 0.7694,
+      "step": 3862
+    },
+    {
+      "epoch": 0.7431704501731435,
+      "grad_norm": 0.8848334707924208,
+      "learning_rate": 3.2692757783940467e-06,
+      "loss": 0.7881,
+      "step": 3863
+    },
+    {
+      "epoch": 0.7433628318584071,
+      "grad_norm": 0.9201624066362469,
+      "learning_rate": 3.264668897332527e-06,
+      "loss": 0.7785,
+      "step": 3864
+    },
+    {
+      "epoch": 0.7435552135436706,
+      "grad_norm": 0.8255798581245719,
+      "learning_rate": 3.2600646311584494e-06,
+      "loss": 0.8612,
+      "step": 3865
+    },
+    {
+      "epoch": 0.7437475952289342,
+      "grad_norm": 0.9272591409772415,
+      "learning_rate": 3.2554629816593375e-06,
+      "loss": 0.7265,
+      "step": 3866
+    },
+    {
+      "epoch": 0.7439399769141978,
+      "grad_norm": 0.8478669533703944,
+      "learning_rate": 3.250863950621721e-06,
+      "loss": 0.7119,
+      "step": 3867
+    },
+    {
+      "epoch": 0.7441323585994614,
+      "grad_norm": 0.8531714672265236,
+      "learning_rate": 3.2462675398310984e-06,
+      "loss": 0.7491,
+      "step": 3868
+    },
+    {
+      "epoch": 0.7443247402847248,
+      "grad_norm": 0.8274027301903057,
+      "learning_rate": 3.241673751071954e-06,
+      "loss": 0.772,
+      "step": 3869
+    },
+    {
+      "epoch": 0.7445171219699884,
+      "grad_norm": 0.8562827971816653,
+      "learning_rate": 3.2370825861277567e-06,
+      "loss": 0.7983,
+      "step": 3870
+    },
+    {
+      "epoch": 0.744709503655252,
+      "grad_norm": 0.9233593475000587,
+      "learning_rate": 3.2324940467809527e-06,
+      "loss": 0.798,
+      "step": 3871
+    },
+    {
+      "epoch": 0.7449018853405156,
+      "grad_norm": 0.8063646448334578,
+      "learning_rate": 3.2279081348129713e-06,
+      "loss": 0.7362,
+      "step": 3872
+    },
+    {
+      "epoch": 0.7450942670257792,
+      "grad_norm": 0.8711967505173981,
+      "learning_rate": 3.223324852004219e-06,
+      "loss": 0.8116,
+      "step": 3873
+    },
+    {
+      "epoch": 0.7452866487110427,
+      "grad_norm": 0.9923536051028559,
+      "learning_rate": 3.2187442001340942e-06,
+      "loss": 0.8219,
+      "step": 3874
+    },
+    {
+      "epoch": 0.7454790303963063,
+      "grad_norm": 0.916561863597196,
+      "learning_rate": 3.21416618098095e-06,
+      "loss": 0.7528,
+      "step": 3875
+    },
+    {
+      "epoch": 0.7456714120815698,
+      "grad_norm": 0.6458176656259352,
+      "learning_rate": 3.2095907963221396e-06,
+      "loss": 0.3409,
+      "step": 3876
+    },
+    {
+      "epoch": 0.7458637937668334,
+      "grad_norm": 0.8490721996334177,
+      "learning_rate": 3.2050180479339865e-06,
+      "loss": 0.78,
+      "step": 3877
+    },
+    {
+      "epoch": 0.746056175452097,
+      "grad_norm": 0.6074898031709546,
+      "learning_rate": 3.2004479375917783e-06,
+      "loss": 0.3492,
+      "step": 3878
+    },
+    {
+      "epoch": 0.7462485571373605,
+      "grad_norm": 0.9604075419164727,
+      "learning_rate": 3.1958804670698008e-06,
+      "loss": 0.7874,
+      "step": 3879
+    },
+    {
+      "epoch": 0.7464409388226241,
+      "grad_norm": 0.8887739959635841,
+      "learning_rate": 3.191315638141297e-06,
+      "loss": 0.7633,
+      "step": 3880
+    },
+    {
+      "epoch": 0.7466333205078877,
+      "grad_norm": 0.9671528513174097,
+      "learning_rate": 3.1867534525784937e-06,
+      "loss": 0.7963,
+      "step": 3881
+    },
+    {
+      "epoch": 0.7468257021931513,
+      "grad_norm": 0.8920435586501148,
+      "learning_rate": 3.182193912152586e-06,
+      "loss": 0.805,
+      "step": 3882
+    },
+    {
+      "epoch": 0.7470180838784147,
+      "grad_norm": 0.8459026670549139,
+      "learning_rate": 3.177637018633746e-06,
+      "loss": 0.7897,
+      "step": 3883
+    },
+    {
+      "epoch": 0.7472104655636783,
+      "grad_norm": 0.8542505551445454,
+      "learning_rate": 3.1730827737911163e-06,
+      "loss": 0.8102,
+      "step": 3884
+    },
+    {
+      "epoch": 0.7474028472489419,
+      "grad_norm": 0.8273380951798446,
+      "learning_rate": 3.1685311793928077e-06,
+      "loss": 0.7197,
+      "step": 3885
+    },
+    {
+      "epoch": 0.7475952289342055,
+      "grad_norm": 0.904215836115893,
+      "learning_rate": 3.163982237205917e-06,
+      "loss": 0.8393,
+      "step": 3886
+    },
+    {
+      "epoch": 0.7477876106194691,
+      "grad_norm": 0.8233647803339943,
+      "learning_rate": 3.1594359489964853e-06,
+      "loss": 0.7743,
+      "step": 3887
+    },
+    {
+      "epoch": 0.7479799923047326,
+      "grad_norm": 0.936458004183698,
+      "learning_rate": 3.15489231652955e-06,
+      "loss": 0.7637,
+      "step": 3888
+    },
+    {
+      "epoch": 0.7481723739899961,
+      "grad_norm": 0.8969716354216654,
+      "learning_rate": 3.150351341569101e-06,
+      "loss": 0.8557,
+      "step": 3889
+    },
+    {
+      "epoch": 0.7483647556752597,
+      "grad_norm": 0.9472500226765785,
+      "learning_rate": 3.1458130258781006e-06,
+      "loss": 0.8,
+      "step": 3890
+    },
+    {
+      "epoch": 0.7485571373605233,
+      "grad_norm": 0.9135441621781233,
+      "learning_rate": 3.141277371218484e-06,
+      "loss": 0.7786,
+      "step": 3891
+    },
+    {
+      "epoch": 0.7487495190457868,
+      "grad_norm": 0.6310305412025723,
+      "learning_rate": 3.136744379351139e-06,
+      "loss": 0.3448,
+      "step": 3892
+    },
+    {
+      "epoch": 0.7489419007310504,
+      "grad_norm": 0.8968323260150286,
+      "learning_rate": 3.1322140520359366e-06,
+      "loss": 0.8043,
+      "step": 3893
+    },
+    {
+      "epoch": 0.749134282416314,
+      "grad_norm": 0.81945277714954,
+      "learning_rate": 3.1276863910317057e-06,
+      "loss": 0.7771,
+      "step": 3894
+    },
+    {
+      "epoch": 0.7493266641015776,
+      "grad_norm": 0.9086797223006944,
+      "learning_rate": 3.1231613980962373e-06,
+      "loss": 0.8332,
+      "step": 3895
+    },
+    {
+      "epoch": 0.7495190457868411,
+      "grad_norm": 0.9376022671994613,
+      "learning_rate": 3.1186390749862904e-06,
+      "loss": 0.7736,
+      "step": 3896
+    },
+    {
+      "epoch": 0.7497114274721046,
+      "grad_norm": 0.8942953918695745,
+      "learning_rate": 3.1141194234575878e-06,
+      "loss": 0.7571,
+      "step": 3897
+    },
+    {
+      "epoch": 0.7499038091573682,
+      "grad_norm": 0.9253909991443271,
+      "learning_rate": 3.1096024452648123e-06,
+      "loss": 0.8337,
+      "step": 3898
+    },
+    {
+      "epoch": 0.7500961908426318,
+      "grad_norm": 0.7912519614133072,
+      "learning_rate": 3.1050881421616076e-06,
+      "loss": 0.7406,
+      "step": 3899
+    },
+    {
+      "epoch": 0.7502885725278954,
+      "grad_norm": 0.9429624497592659,
+      "learning_rate": 3.100576515900591e-06,
+      "loss": 0.8315,
+      "step": 3900
+    },
+    {
+      "epoch": 0.7504809542131589,
+      "grad_norm": 0.9163722146745016,
+      "learning_rate": 3.0960675682333186e-06,
+      "loss": 0.8037,
+      "step": 3901
+    },
+    {
+      "epoch": 0.7506733358984224,
+      "grad_norm": 0.8458743986732001,
+      "learning_rate": 3.0915613009103296e-06,
+      "loss": 0.7438,
+      "step": 3902
+    },
+    {
+      "epoch": 0.750865717583686,
+      "grad_norm": 0.867622567790327,
+      "learning_rate": 3.0870577156811077e-06,
+      "loss": 0.7876,
+      "step": 3903
+    },
+    {
+      "epoch": 0.7510580992689496,
+      "grad_norm": 0.9851714553348679,
+      "learning_rate": 3.0825568142940998e-06,
+      "loss": 0.7862,
+      "step": 3904
+    },
+    {
+      "epoch": 0.7512504809542132,
+      "grad_norm": 0.8685257300768066,
+      "learning_rate": 3.0780585984967113e-06,
+      "loss": 0.8177,
+      "step": 3905
+    },
+    {
+      "epoch": 0.7514428626394767,
+      "grad_norm": 0.8487212120323661,
+      "learning_rate": 3.073563070035305e-06,
+      "loss": 0.8061,
+      "step": 3906
+    },
+    {
+      "epoch": 0.7516352443247403,
+      "grad_norm": 0.883965536598361,
+      "learning_rate": 3.069070230655198e-06,
+      "loss": 0.7735,
+      "step": 3907
+    },
+    {
+      "epoch": 0.7518276260100039,
+      "grad_norm": 0.8225428435878037,
+      "learning_rate": 3.0645800821006667e-06,
+      "loss": 0.753,
+      "step": 3908
+    },
+    {
+      "epoch": 0.7520200076952674,
+      "grad_norm": 0.8565578666169182,
+      "learning_rate": 3.060092626114941e-06,
+      "loss": 0.8049,
+      "step": 3909
+    },
+    {
+      "epoch": 0.7522123893805309,
+      "grad_norm": 0.89769366032698,
+      "learning_rate": 3.0556078644402066e-06,
+      "loss": 0.7588,
+      "step": 3910
+    },
+    {
+      "epoch": 0.7524047710657945,
+      "grad_norm": 0.9089787758987465,
+      "learning_rate": 3.051125798817598e-06,
+      "loss": 0.761,
+      "step": 3911
+    },
+    {
+      "epoch": 0.7525971527510581,
+      "grad_norm": 0.8451815335593035,
+      "learning_rate": 3.0466464309872167e-06,
+      "loss": 0.7284,
+      "step": 3912
+    },
+    {
+      "epoch": 0.7527895344363217,
+      "grad_norm": 0.9041591906061932,
+      "learning_rate": 3.042169762688096e-06,
+      "loss": 0.8334,
+      "step": 3913
+    },
+    {
+      "epoch": 0.7529819161215853,
+      "grad_norm": 0.877390502239757,
+      "learning_rate": 3.0376957956582452e-06,
+      "loss": 0.8542,
+      "step": 3914
+    },
+    {
+      "epoch": 0.7531742978068487,
+      "grad_norm": 0.8930749970530174,
+      "learning_rate": 3.0332245316346e-06,
+      "loss": 0.8484,
+      "step": 3915
+    },
+    {
+      "epoch": 0.7533666794921123,
+      "grad_norm": 0.9135581022569454,
+      "learning_rate": 3.0287559723530667e-06,
+      "loss": 0.7978,
+      "step": 3916
+    },
+    {
+      "epoch": 0.7535590611773759,
+      "grad_norm": 0.8496644468686498,
+      "learning_rate": 3.024290119548495e-06,
+      "loss": 0.7355,
+      "step": 3917
+    },
+    {
+      "epoch": 0.7537514428626395,
+      "grad_norm": 0.8375915121254048,
+      "learning_rate": 3.019826974954674e-06,
+      "loss": 0.7805,
+      "step": 3918
+    },
+    {
+      "epoch": 0.753943824547903,
+      "grad_norm": 0.9073859439714186,
+      "learning_rate": 3.0153665403043586e-06,
+      "loss": 0.7726,
+      "step": 3919
+    },
+    {
+      "epoch": 0.7541362062331666,
+      "grad_norm": 0.8920864038912314,
+      "learning_rate": 3.01090881732924e-06,
+      "loss": 0.7404,
+      "step": 3920
+    },
+    {
+      "epoch": 0.7543285879184302,
+      "grad_norm": 0.8084788384416017,
+      "learning_rate": 3.0064538077599603e-06,
+      "loss": 0.7258,
+      "step": 3921
+    },
+    {
+      "epoch": 0.7545209696036937,
+      "grad_norm": 0.9346090100120188,
+      "learning_rate": 3.002001513326107e-06,
+      "loss": 0.7827,
+      "step": 3922
+    },
+    {
+      "epoch": 0.7547133512889573,
+      "grad_norm": 0.8603653960426373,
+      "learning_rate": 2.9975519357562155e-06,
+      "loss": 0.7809,
+      "step": 3923
+    },
+    {
+      "epoch": 0.7549057329742208,
+      "grad_norm": 0.7672384179900772,
+      "learning_rate": 2.9931050767777626e-06,
+      "loss": 0.6812,
+      "step": 3924
+    },
+    {
+      "epoch": 0.7550981146594844,
+      "grad_norm": 0.9238210258604374,
+      "learning_rate": 2.9886609381171703e-06,
+      "loss": 0.7635,
+      "step": 3925
+    },
+    {
+      "epoch": 0.755290496344748,
+      "grad_norm": 0.8507437932205408,
+      "learning_rate": 2.984219521499816e-06,
+      "loss": 0.7754,
+      "step": 3926
+    },
+    {
+      "epoch": 0.7554828780300116,
+      "grad_norm": 0.6383956849892184,
+      "learning_rate": 2.9797808286499976e-06,
+      "loss": 0.3083,
+      "step": 3927
+    },
+    {
+      "epoch": 0.7556752597152752,
+      "grad_norm": 0.8605176380400721,
+      "learning_rate": 2.9753448612909775e-06,
+      "loss": 0.7869,
+      "step": 3928
+    },
+    {
+      "epoch": 0.7558676414005386,
+      "grad_norm": 0.8095561016490535,
+      "learning_rate": 2.9709116211449484e-06,
+      "loss": 0.8052,
+      "step": 3929
+    },
+    {
+      "epoch": 0.7560600230858022,
+      "grad_norm": 0.9296871974800398,
+      "learning_rate": 2.966481109933047e-06,
+      "loss": 0.7434,
+      "step": 3930
+    },
+    {
+      "epoch": 0.7562524047710658,
+      "grad_norm": 0.8884597016803714,
+      "learning_rate": 2.9620533293753495e-06,
+      "loss": 0.7625,
+      "step": 3931
+    },
+    {
+      "epoch": 0.7564447864563294,
+      "grad_norm": 0.8680335090258381,
+      "learning_rate": 2.957628281190873e-06,
+      "loss": 0.8178,
+      "step": 3932
+    },
+    {
+      "epoch": 0.7566371681415929,
+      "grad_norm": 0.9077503880401392,
+      "learning_rate": 2.9532059670975732e-06,
+      "loss": 0.755,
+      "step": 3933
+    },
+    {
+      "epoch": 0.7568295498268565,
+      "grad_norm": 0.8539037462598993,
+      "learning_rate": 2.948786388812346e-06,
+      "loss": 0.7783,
+      "step": 3934
+    },
+    {
+      "epoch": 0.75702193151212,
+      "grad_norm": 0.6045288876775105,
+      "learning_rate": 2.9443695480510225e-06,
+      "loss": 0.2999,
+      "step": 3935
+    },
+    {
+      "epoch": 0.7572143131973836,
+      "grad_norm": 0.9649581322721572,
+      "learning_rate": 2.9399554465283742e-06,
+      "loss": 0.7639,
+      "step": 3936
+    },
+    {
+      "epoch": 0.7574066948826472,
+      "grad_norm": 0.8674192204881583,
+      "learning_rate": 2.935544085958102e-06,
+      "loss": 0.8344,
+      "step": 3937
+    },
+    {
+      "epoch": 0.7575990765679107,
+      "grad_norm": 0.9742233312640968,
+      "learning_rate": 2.931135468052858e-06,
+      "loss": 0.8158,
+      "step": 3938
+    },
+    {
+      "epoch": 0.7577914582531743,
+      "grad_norm": 0.8388601833550142,
+      "learning_rate": 2.926729594524207e-06,
+      "loss": 0.8134,
+      "step": 3939
+    },
+    {
+      "epoch": 0.7579838399384379,
+      "grad_norm": 0.8377034863642825,
+      "learning_rate": 2.9223264670826746e-06,
+      "loss": 0.8073,
+      "step": 3940
+    },
+    {
+      "epoch": 0.7581762216237015,
+      "grad_norm": 0.9828794496818635,
+      "learning_rate": 2.9179260874376915e-06,
+      "loss": 0.7794,
+      "step": 3941
+    },
+    {
+      "epoch": 0.7583686033089649,
+      "grad_norm": 0.9354777578371479,
+      "learning_rate": 2.9135284572976486e-06,
+      "loss": 0.7717,
+      "step": 3942
+    },
+    {
+      "epoch": 0.7585609849942285,
+      "grad_norm": 0.8603401941404448,
+      "learning_rate": 2.9091335783698517e-06,
+      "loss": 0.7968,
+      "step": 3943
+    },
+    {
+      "epoch": 0.7587533666794921,
+      "grad_norm": 0.8421324233780078,
+      "learning_rate": 2.9047414523605467e-06,
+      "loss": 0.7859,
+      "step": 3944
+    },
+    {
+      "epoch": 0.7589457483647557,
+      "grad_norm": 0.8596333408087772,
+      "learning_rate": 2.9003520809749053e-06,
+      "loss": 0.8356,
+      "step": 3945
+    },
+    {
+      "epoch": 0.7591381300500193,
+      "grad_norm": 0.9000811465239291,
+      "learning_rate": 2.8959654659170354e-06,
+      "loss": 0.821,
+      "step": 3946
+    },
+    {
+      "epoch": 0.7593305117352828,
+      "grad_norm": 0.8062449544975521,
+      "learning_rate": 2.8915816088899696e-06,
+      "loss": 0.8151,
+      "step": 3947
+    },
+    {
+      "epoch": 0.7595228934205464,
+      "grad_norm": 0.5851683767592993,
+      "learning_rate": 2.8872005115956746e-06,
+      "loss": 0.317,
+      "step": 3948
+    },
+    {
+      "epoch": 0.7597152751058099,
+      "grad_norm": 0.851740743428255,
+      "learning_rate": 2.8828221757350406e-06,
+      "loss": 0.7832,
+      "step": 3949
+    },
+    {
+      "epoch": 0.7599076567910735,
+      "grad_norm": 0.8870757774037864,
+      "learning_rate": 2.8784466030078905e-06,
+      "loss": 0.735,
+      "step": 3950
+    },
+    {
+      "epoch": 0.760100038476337,
+      "grad_norm": 0.9801815610072047,
+      "learning_rate": 2.874073795112967e-06,
+      "loss": 0.8213,
+      "step": 3951
+    },
+    {
+      "epoch": 0.7602924201616006,
+      "grad_norm": 0.5971382795238487,
+      "learning_rate": 2.8697037537479565e-06,
+      "loss": 0.351,
+      "step": 3952
+    },
+    {
+      "epoch": 0.7604848018468642,
+      "grad_norm": 0.8685678202793361,
+      "learning_rate": 2.8653364806094454e-06,
+      "loss": 0.8362,
+      "step": 3953
+    },
+    {
+      "epoch": 0.7606771835321278,
+      "grad_norm": 0.8406755544089135,
+      "learning_rate": 2.86097197739297e-06,
+      "loss": 0.8119,
+      "step": 3954
+    },
+    {
+      "epoch": 0.7608695652173914,
+      "grad_norm": 0.8818020499690645,
+      "learning_rate": 2.856610245792976e-06,
+      "loss": 0.8086,
+      "step": 3955
+    },
+    {
+      "epoch": 0.7610619469026548,
+      "grad_norm": 0.8677598798605344,
+      "learning_rate": 2.8522512875028396e-06,
+      "loss": 0.7819,
+      "step": 3956
+    },
+    {
+      "epoch": 0.7612543285879184,
+      "grad_norm": 0.9050342655156159,
+      "learning_rate": 2.847895104214856e-06,
+      "loss": 0.7282,
+      "step": 3957
+    },
+    {
+      "epoch": 0.761446710273182,
+      "grad_norm": 0.9380056313624792,
+      "learning_rate": 2.843541697620249e-06,
+      "loss": 0.7587,
+      "step": 3958
+    },
+    {
+      "epoch": 0.7616390919584456,
+      "grad_norm": 0.9095038980911313,
+      "learning_rate": 2.8391910694091584e-06,
+      "loss": 0.8119,
+      "step": 3959
+    },
+    {
+      "epoch": 0.7618314736437091,
+      "grad_norm": 0.9098971233275468,
+      "learning_rate": 2.8348432212706443e-06,
+      "loss": 0.754,
+      "step": 3960
+    },
+    {
+      "epoch": 0.7620238553289727,
+      "grad_norm": 0.9524150559136597,
+      "learning_rate": 2.8304981548927025e-06,
+      "loss": 0.832,
+      "step": 3961
+    },
+    {
+      "epoch": 0.7622162370142362,
+      "grad_norm": 0.9253562579686191,
+      "learning_rate": 2.826155871962227e-06,
+      "loss": 0.842,
+      "step": 3962
+    },
+    {
+      "epoch": 0.7624086186994998,
+      "grad_norm": 0.9014611639115168,
+      "learning_rate": 2.8218163741650415e-06,
+      "loss": 0.8111,
+      "step": 3963
+    },
+    {
+      "epoch": 0.7626010003847634,
+      "grad_norm": 0.9501532388197814,
+      "learning_rate": 2.817479663185898e-06,
+      "loss": 0.7469,
+      "step": 3964
+    },
+    {
+      "epoch": 0.7627933820700269,
+      "grad_norm": 0.921066945742715,
+      "learning_rate": 2.813145740708445e-06,
+      "loss": 0.8549,
+      "step": 3965
+    },
+    {
+      "epoch": 0.7629857637552905,
+      "grad_norm": 0.8696604860916524,
+      "learning_rate": 2.808814608415271e-06,
+      "loss": 0.8622,
+      "step": 3966
+    },
+    {
+      "epoch": 0.7631781454405541,
+      "grad_norm": 0.9474150297072254,
+      "learning_rate": 2.8044862679878605e-06,
+      "loss": 0.7573,
+      "step": 3967
+    },
+    {
+      "epoch": 0.7633705271258177,
+      "grad_norm": 1.0135597720196137,
+      "learning_rate": 2.800160721106633e-06,
+      "loss": 0.8153,
+      "step": 3968
+    },
+    {
+      "epoch": 0.7635629088110811,
+      "grad_norm": 0.8892805640352013,
+      "learning_rate": 2.7958379694509108e-06,
+      "loss": 0.7839,
+      "step": 3969
+    },
+    {
+      "epoch": 0.7637552904963447,
+      "grad_norm": 0.9803309088823355,
+      "learning_rate": 2.791518014698935e-06,
+      "loss": 0.7599,
+      "step": 3970
+    },
+    {
+      "epoch": 0.7639476721816083,
+      "grad_norm": 0.6282410205202417,
+      "learning_rate": 2.787200858527862e-06,
+      "loss": 0.3043,
+      "step": 3971
+    },
+    {
+      "epoch": 0.7641400538668719,
+      "grad_norm": 0.852903920635151,
+      "learning_rate": 2.7828865026137584e-06,
+      "loss": 0.7996,
+      "step": 3972
+    },
+    {
+      "epoch": 0.7643324355521355,
+      "grad_norm": 0.7826463135893983,
+      "learning_rate": 2.7785749486316085e-06,
+      "loss": 0.7744,
+      "step": 3973
+    },
+    {
+      "epoch": 0.764524817237399,
+      "grad_norm": 0.9591096799525374,
+      "learning_rate": 2.774266198255303e-06,
+      "loss": 0.7689,
+      "step": 3974
+    },
+    {
+      "epoch": 0.7647171989226625,
+      "grad_norm": 0.8298619672159963,
+      "learning_rate": 2.7699602531576496e-06,
+      "loss": 0.7846,
+      "step": 3975
+    },
+    {
+      "epoch": 0.7649095806079261,
+      "grad_norm": 0.9133803116084399,
+      "learning_rate": 2.765657115010364e-06,
+      "loss": 0.828,
+      "step": 3976
+    },
+    {
+      "epoch": 0.7651019622931897,
+      "grad_norm": 0.8152717940873574,
+      "learning_rate": 2.7613567854840685e-06,
+      "loss": 0.8172,
+      "step": 3977
+    },
+    {
+      "epoch": 0.7652943439784533,
+      "grad_norm": 0.9076123067381614,
+      "learning_rate": 2.7570592662483086e-06,
+      "loss": 0.8452,
+      "step": 3978
+    },
+    {
+      "epoch": 0.7654867256637168,
+      "grad_norm": 0.8768129631134592,
+      "learning_rate": 2.752764558971517e-06,
+      "loss": 0.8677,
+      "step": 3979
+    },
+    {
+      "epoch": 0.7656791073489804,
+      "grad_norm": 0.8199372643180708,
+      "learning_rate": 2.748472665321056e-06,
+      "loss": 0.7771,
+      "step": 3980
+    },
+    {
+      "epoch": 0.765871489034244,
+      "grad_norm": 0.9381402186070591,
+      "learning_rate": 2.744183586963185e-06,
+      "loss": 0.7823,
+      "step": 3981
+    },
+    {
+      "epoch": 0.7660638707195075,
+      "grad_norm": 0.8485287485198522,
+      "learning_rate": 2.739897325563069e-06,
+      "loss": 0.8052,
+      "step": 3982
+    },
+    {
+      "epoch": 0.766256252404771,
+      "grad_norm": 0.8732035807629315,
+      "learning_rate": 2.7356138827847856e-06,
+      "loss": 0.8115,
+      "step": 3983
+    },
+    {
+      "epoch": 0.7664486340900346,
+      "grad_norm": 0.8895382410915199,
+      "learning_rate": 2.731333260291311e-06,
+      "loss": 0.8219,
+      "step": 3984
+    },
+    {
+      "epoch": 0.7666410157752982,
+      "grad_norm": 0.8672848403690884,
+      "learning_rate": 2.7270554597445343e-06,
+      "loss": 0.721,
+      "step": 3985
+    },
+    {
+      "epoch": 0.7668333974605618,
+      "grad_norm": 0.965751260671076,
+      "learning_rate": 2.7227804828052384e-06,
+      "loss": 0.8193,
+      "step": 3986
+    },
+    {
+      "epoch": 0.7670257791458254,
+      "grad_norm": 0.9348261315594508,
+      "learning_rate": 2.7185083311331283e-06,
+      "loss": 0.8234,
+      "step": 3987
+    },
+    {
+      "epoch": 0.7672181608310888,
+      "grad_norm": 0.6329328489539221,
+      "learning_rate": 2.7142390063867896e-06,
+      "loss": 0.3242,
+      "step": 3988
+    },
+    {
+      "epoch": 0.7674105425163524,
+      "grad_norm": 0.8908618413425756,
+      "learning_rate": 2.709972510223725e-06,
+      "loss": 0.818,
+      "step": 3989
+    },
+    {
+      "epoch": 0.767602924201616,
+      "grad_norm": 0.9147365291831651,
+      "learning_rate": 2.7057088443003343e-06,
+      "loss": 0.8114,
+      "step": 3990
+    },
+    {
+      "epoch": 0.7677953058868796,
+      "grad_norm": 0.8907559472174097,
+      "learning_rate": 2.7014480102719174e-06,
+      "loss": 0.8261,
+      "step": 3991
+    },
+    {
+      "epoch": 0.7679876875721431,
+      "grad_norm": 0.8805955050987004,
+      "learning_rate": 2.697190009792685e-06,
+      "loss": 0.7987,
+      "step": 3992
+    },
+    {
+      "epoch": 0.7681800692574067,
+      "grad_norm": 0.9334763361286479,
+      "learning_rate": 2.692934844515729e-06,
+      "loss": 0.8232,
+      "step": 3993
+    },
+    {
+      "epoch": 0.7683724509426703,
+      "grad_norm": 0.8644997906091039,
+      "learning_rate": 2.6886825160930587e-06,
+      "loss": 0.8273,
+      "step": 3994
+    },
+    {
+      "epoch": 0.7685648326279338,
+      "grad_norm": 0.9557820499916851,
+      "learning_rate": 2.6844330261755715e-06,
+      "loss": 0.7821,
+      "step": 3995
+    },
+    {
+      "epoch": 0.7687572143131974,
+      "grad_norm": 0.8851301020201698,
+      "learning_rate": 2.6801863764130653e-06,
+      "loss": 0.7833,
+      "step": 3996
+    },
+    {
+      "epoch": 0.7689495959984609,
+      "grad_norm": 0.599949145595607,
+      "learning_rate": 2.675942568454236e-06,
+      "loss": 0.3078,
+      "step": 3997
+    },
+    {
+      "epoch": 0.7691419776837245,
+      "grad_norm": 0.85364308537703,
+      "learning_rate": 2.671701603946678e-06,
+      "loss": 0.7829,
+      "step": 3998
+    },
+    {
+      "epoch": 0.7693343593689881,
+      "grad_norm": 0.8857350772563871,
+      "learning_rate": 2.667463484536876e-06,
+      "loss": 0.8259,
+      "step": 3999
+    },
+    {
+      "epoch": 0.7695267410542517,
+      "grad_norm": 0.8974312648004988,
+      "learning_rate": 2.6632282118702147e-06,
+      "loss": 0.8216,
+      "step": 4000
+    },
+    {
+      "epoch": 0.7697191227395151,
+      "grad_norm": 0.9432694620461326,
+      "learning_rate": 2.65899578759098e-06,
+      "loss": 0.8311,
+      "step": 4001
+    },
+    {
+      "epoch": 0.7699115044247787,
+      "grad_norm": 0.824430764470664,
+      "learning_rate": 2.654766213342335e-06,
+      "loss": 0.7707,
+      "step": 4002
+    },
+    {
+      "epoch": 0.7701038861100423,
+      "grad_norm": 0.8569850554825471,
+      "learning_rate": 2.650539490766346e-06,
+      "loss": 0.7876,
+      "step": 4003
+    },
+    {
+      "epoch": 0.7702962677953059,
+      "grad_norm": 0.9024391057552891,
+      "learning_rate": 2.646315621503983e-06,
+      "loss": 0.8098,
+      "step": 4004
+    },
+    {
+      "epoch": 0.7704886494805695,
+      "grad_norm": 0.9702149923817245,
+      "learning_rate": 2.642094607195085e-06,
+      "loss": 0.8207,
+      "step": 4005
+    },
+    {
+      "epoch": 0.770681031165833,
+      "grad_norm": 0.8338064593117002,
+      "learning_rate": 2.6378764494784027e-06,
+      "loss": 0.8041,
+      "step": 4006
+    },
+    {
+      "epoch": 0.7708734128510966,
+      "grad_norm": 0.9089008217709775,
+      "learning_rate": 2.633661149991569e-06,
+      "loss": 0.7986,
+      "step": 4007
+    },
+    {
+      "epoch": 0.7710657945363601,
+      "grad_norm": 0.9045732958435624,
+      "learning_rate": 2.6294487103711064e-06,
+      "loss": 0.803,
+      "step": 4008
+    },
+    {
+      "epoch": 0.7712581762216237,
+      "grad_norm": 0.8582370135902713,
+      "learning_rate": 2.6252391322524297e-06,
+      "loss": 0.8495,
+      "step": 4009
+    },
+    {
+      "epoch": 0.7714505579068872,
+      "grad_norm": 0.878724509499889,
+      "learning_rate": 2.6210324172698432e-06,
+      "loss": 0.8338,
+      "step": 4010
+    },
+    {
+      "epoch": 0.7716429395921508,
+      "grad_norm": 0.6429293209083587,
+      "learning_rate": 2.6168285670565374e-06,
+      "loss": 0.3609,
+      "step": 4011
+    },
+    {
+      "epoch": 0.7718353212774144,
+      "grad_norm": 0.8913127466690723,
+      "learning_rate": 2.6126275832445892e-06,
+      "loss": 0.8515,
+      "step": 4012
+    },
+    {
+      "epoch": 0.772027702962678,
+      "grad_norm": 0.9356368381280117,
+      "learning_rate": 2.6084294674649734e-06,
+      "loss": 0.7874,
+      "step": 4013
+    },
+    {
+      "epoch": 0.7722200846479416,
+      "grad_norm": 0.8231157386966805,
+      "learning_rate": 2.6042342213475346e-06,
+      "loss": 0.7872,
+      "step": 4014
+    },
+    {
+      "epoch": 0.772412466333205,
+      "grad_norm": 0.899346927182789,
+      "learning_rate": 2.6000418465210143e-06,
+      "loss": 0.7808,
+      "step": 4015
+    },
+    {
+      "epoch": 0.7726048480184686,
+      "grad_norm": 0.8023029325049612,
+      "learning_rate": 2.595852344613038e-06,
+      "loss": 0.7555,
+      "step": 4016
+    },
+    {
+      "epoch": 0.7727972297037322,
+      "grad_norm": 0.7982921750780583,
+      "learning_rate": 2.5916657172501103e-06,
+      "loss": 0.6926,
+      "step": 4017
+    },
+    {
+      "epoch": 0.7729896113889958,
+      "grad_norm": 0.9746474686160395,
+      "learning_rate": 2.587481966057633e-06,
+      "loss": 0.8243,
+      "step": 4018
+    },
+    {
+      "epoch": 0.7731819930742593,
+      "grad_norm": 0.9020623611469117,
+      "learning_rate": 2.583301092659872e-06,
+      "loss": 0.7638,
+      "step": 4019
+    },
+    {
+      "epoch": 0.7733743747595229,
+      "grad_norm": 0.9297160316657678,
+      "learning_rate": 2.5791230986799944e-06,
+      "loss": 0.7919,
+      "step": 4020
+    },
+    {
+      "epoch": 0.7735667564447865,
+      "grad_norm": 0.9659971492198461,
+      "learning_rate": 2.5749479857400383e-06,
+      "loss": 0.7745,
+      "step": 4021
+    },
+    {
+      "epoch": 0.77375913813005,
+      "grad_norm": 0.8591227584773681,
+      "learning_rate": 2.5707757554609247e-06,
+      "loss": 0.8396,
+      "step": 4022
+    },
+    {
+      "epoch": 0.7739515198153136,
+      "grad_norm": 0.6043671805380403,
+      "learning_rate": 2.56660640946246e-06,
+      "loss": 0.3461,
+      "step": 4023
+    },
+    {
+      "epoch": 0.7741439015005771,
+      "grad_norm": 0.8976979341363452,
+      "learning_rate": 2.5624399493633257e-06,
+      "loss": 0.7627,
+      "step": 4024
+    },
+    {
+      "epoch": 0.7743362831858407,
+      "grad_norm": 0.8645632196740883,
+      "learning_rate": 2.558276376781086e-06,
+      "loss": 0.8214,
+      "step": 4025
+    },
+    {
+      "epoch": 0.7745286648711043,
+      "grad_norm": 0.8223181194445588,
+      "learning_rate": 2.55411569333218e-06,
+      "loss": 0.775,
+      "step": 4026
+    },
+    {
+      "epoch": 0.7747210465563679,
+      "grad_norm": 0.8996716980714874,
+      "learning_rate": 2.5499579006319365e-06,
+      "loss": 0.7827,
+      "step": 4027
+    },
+    {
+      "epoch": 0.7749134282416313,
+      "grad_norm": 0.8573224154591396,
+      "learning_rate": 2.5458030002945457e-06,
+      "loss": 0.786,
+      "step": 4028
+    },
+    {
+      "epoch": 0.7751058099268949,
+      "grad_norm": 0.9458883358789721,
+      "learning_rate": 2.5416509939330836e-06,
+      "loss": 0.8122,
+      "step": 4029
+    },
+    {
+      "epoch": 0.7752981916121585,
+      "grad_norm": 0.8808118352865209,
+      "learning_rate": 2.537501883159509e-06,
+      "loss": 0.7909,
+      "step": 4030
+    },
+    {
+      "epoch": 0.7754905732974221,
+      "grad_norm": 0.7776301867387295,
+      "learning_rate": 2.5333556695846384e-06,
+      "loss": 0.799,
+      "step": 4031
+    },
+    {
+      "epoch": 0.7756829549826857,
+      "grad_norm": 0.9775492118525482,
+      "learning_rate": 2.5292123548181847e-06,
+      "loss": 0.8395,
+      "step": 4032
+    },
+    {
+      "epoch": 0.7758753366679492,
+      "grad_norm": 0.9029385153148063,
+      "learning_rate": 2.525071940468722e-06,
+      "loss": 0.8216,
+      "step": 4033
+    },
+    {
+      "epoch": 0.7760677183532128,
+      "grad_norm": 0.8701256082896985,
+      "learning_rate": 2.520934428143701e-06,
+      "loss": 0.75,
+      "step": 4034
+    },
+    {
+      "epoch": 0.7762601000384763,
+      "grad_norm": 0.938609360328394,
+      "learning_rate": 2.5167998194494468e-06,
+      "loss": 0.7725,
+      "step": 4035
+    },
+    {
+      "epoch": 0.7764524817237399,
+      "grad_norm": 0.8453376483566815,
+      "learning_rate": 2.5126681159911558e-06,
+      "loss": 0.8251,
+      "step": 4036
+    },
+    {
+      "epoch": 0.7766448634090035,
+      "grad_norm": 0.8690556382376977,
+      "learning_rate": 2.5085393193729e-06,
+      "loss": 0.7925,
+      "step": 4037
+    },
+    {
+      "epoch": 0.776837245094267,
+      "grad_norm": 0.9582862096166292,
+      "learning_rate": 2.5044134311976156e-06,
+      "loss": 0.73,
+      "step": 4038
+    },
+    {
+      "epoch": 0.7770296267795306,
+      "grad_norm": 0.8557402222376954,
+      "learning_rate": 2.5002904530671236e-06,
+      "loss": 0.8132,
+      "step": 4039
+    },
+    {
+      "epoch": 0.7772220084647942,
+      "grad_norm": 0.8545423381228829,
+      "learning_rate": 2.4961703865820974e-06,
+      "loss": 0.7974,
+      "step": 4040
+    },
+    {
+      "epoch": 0.7774143901500578,
+      "grad_norm": 0.8757739149160939,
+      "learning_rate": 2.492053233342091e-06,
+      "loss": 0.723,
+      "step": 4041
+    },
+    {
+      "epoch": 0.7776067718353212,
+      "grad_norm": 0.8391806613194674,
+      "learning_rate": 2.487938994945527e-06,
+      "loss": 0.7513,
+      "step": 4042
+    },
+    {
+      "epoch": 0.7777991535205848,
+      "grad_norm": 0.8884494728144692,
+      "learning_rate": 2.4838276729896884e-06,
+      "loss": 0.7875,
+      "step": 4043
+    },
+    {
+      "epoch": 0.7779915352058484,
+      "grad_norm": 0.8703664301409288,
+      "learning_rate": 2.479719269070743e-06,
+      "loss": 0.7956,
+      "step": 4044
+    },
+    {
+      "epoch": 0.778183916891112,
+      "grad_norm": 0.8613897339581262,
+      "learning_rate": 2.4756137847837025e-06,
+      "loss": 0.7771,
+      "step": 4045
+    },
+    {
+      "epoch": 0.7783762985763756,
+      "grad_norm": 0.9339022882956161,
+      "learning_rate": 2.4715112217224657e-06,
+      "loss": 0.8056,
+      "step": 4046
+    },
+    {
+      "epoch": 0.7785686802616391,
+      "grad_norm": 0.84582051194381,
+      "learning_rate": 2.467411581479786e-06,
+      "loss": 0.7034,
+      "step": 4047
+    },
+    {
+      "epoch": 0.7787610619469026,
+      "grad_norm": 0.9065223669318037,
+      "learning_rate": 2.463314865647286e-06,
+      "loss": 0.7513,
+      "step": 4048
+    },
+    {
+      "epoch": 0.7789534436321662,
+      "grad_norm": 0.9582499713767487,
+      "learning_rate": 2.45922107581545e-06,
+      "loss": 0.8608,
+      "step": 4049
+    },
+    {
+      "epoch": 0.7791458253174298,
+      "grad_norm": 0.883604354615583,
+      "learning_rate": 2.4551302135736287e-06,
+      "loss": 0.7852,
+      "step": 4050
+    },
+    {
+      "epoch": 0.7793382070026933,
+      "grad_norm": 0.8305700253643171,
+      "learning_rate": 2.4510422805100366e-06,
+      "loss": 0.7261,
+      "step": 4051
+    },
+    {
+      "epoch": 0.7795305886879569,
+      "grad_norm": 1.093938276780993,
+      "learning_rate": 2.446957278211746e-06,
+      "loss": 0.7997,
+      "step": 4052
+    },
+    {
+      "epoch": 0.7797229703732205,
+      "grad_norm": 0.8698701310867507,
+      "learning_rate": 2.4428752082647044e-06,
+      "loss": 0.7786,
+      "step": 4053
+    },
+    {
+      "epoch": 0.7799153520584841,
+      "grad_norm": 0.8766206399389889,
+      "learning_rate": 2.438796072253704e-06,
+      "loss": 0.741,
+      "step": 4054
+    },
+    {
+      "epoch": 0.7801077337437476,
+      "grad_norm": 0.9156512949498767,
+      "learning_rate": 2.4347198717624054e-06,
+      "loss": 0.804,
+      "step": 4055
+    },
+    {
+      "epoch": 0.7803001154290111,
+      "grad_norm": 0.8885389594439654,
+      "learning_rate": 2.4306466083733392e-06,
+      "loss": 0.8507,
+      "step": 4056
+    },
+    {
+      "epoch": 0.7804924971142747,
+      "grad_norm": 0.5927057508774353,
+      "learning_rate": 2.426576283667873e-06,
+      "loss": 0.3363,
+      "step": 4057
+    },
+    {
+      "epoch": 0.7806848787995383,
+      "grad_norm": 0.8876007848760398,
+      "learning_rate": 2.422508899226258e-06,
+      "loss": 0.7473,
+      "step": 4058
+    },
+    {
+      "epoch": 0.7808772604848019,
+      "grad_norm": 0.6316664332412024,
+      "learning_rate": 2.418444456627589e-06,
+      "loss": 0.3258,
+      "step": 4059
+    },
+    {
+      "epoch": 0.7810696421700654,
+      "grad_norm": 0.8555129674152326,
+      "learning_rate": 2.4143829574498224e-06,
+      "loss": 0.8085,
+      "step": 4060
+    },
+    {
+      "epoch": 0.781262023855329,
+      "grad_norm": 0.8977449047769018,
+      "learning_rate": 2.4103244032697717e-06,
+      "loss": 0.8025,
+      "step": 4061
+    },
+    {
+      "epoch": 0.7814544055405925,
+      "grad_norm": 0.8743096371722168,
+      "learning_rate": 2.406268795663108e-06,
+      "loss": 0.8087,
+      "step": 4062
+    },
+    {
+      "epoch": 0.7816467872258561,
+      "grad_norm": 0.8784540465971181,
+      "learning_rate": 2.4022161362043574e-06,
+      "loss": 0.8276,
+      "step": 4063
+    },
+    {
+      "epoch": 0.7818391689111197,
+      "grad_norm": 0.9697592878329929,
+      "learning_rate": 2.3981664264669025e-06,
+      "loss": 0.8115,
+      "step": 4064
+    },
+    {
+      "epoch": 0.7820315505963832,
+      "grad_norm": 0.9121847674612503,
+      "learning_rate": 2.3941196680229794e-06,
+      "loss": 0.7961,
+      "step": 4065
+    },
+    {
+      "epoch": 0.7822239322816468,
+      "grad_norm": 0.8636976723867479,
+      "learning_rate": 2.3900758624436772e-06,
+      "loss": 0.8239,
+      "step": 4066
+    },
+    {
+      "epoch": 0.7824163139669104,
+      "grad_norm": 0.8273874270379387,
+      "learning_rate": 2.3860350112989473e-06,
+      "loss": 0.7305,
+      "step": 4067
+    },
+    {
+      "epoch": 0.782608695652174,
+      "grad_norm": 0.9243277382939894,
+      "learning_rate": 2.3819971161575807e-06,
+      "loss": 0.7964,
+      "step": 4068
+    },
+    {
+      "epoch": 0.7828010773374374,
+      "grad_norm": 0.8236322780808236,
+      "learning_rate": 2.3779621785872252e-06,
+      "loss": 0.7432,
+      "step": 4069
+    },
+    {
+      "epoch": 0.782993459022701,
+      "grad_norm": 0.8428324371046615,
+      "learning_rate": 2.3739302001543918e-06,
+      "loss": 0.8061,
+      "step": 4070
+    },
+    {
+      "epoch": 0.7831858407079646,
+      "grad_norm": 0.9353382487782372,
+      "learning_rate": 2.3699011824244234e-06,
+      "loss": 0.7486,
+      "step": 4071
+    },
+    {
+      "epoch": 0.7833782223932282,
+      "grad_norm": 0.8175355201496345,
+      "learning_rate": 2.365875126961531e-06,
+      "loss": 0.7892,
+      "step": 4072
+    },
+    {
+      "epoch": 0.7835706040784918,
+      "grad_norm": 0.9721163266487696,
+      "learning_rate": 2.3618520353287644e-06,
+      "loss": 0.8056,
+      "step": 4073
+    },
+    {
+      "epoch": 0.7837629857637552,
+      "grad_norm": 0.82776197996111,
+      "learning_rate": 2.3578319090880263e-06,
+      "loss": 0.8097,
+      "step": 4074
+    },
+    {
+      "epoch": 0.7839553674490188,
+      "grad_norm": 0.8818075639821297,
+      "learning_rate": 2.3538147498000695e-06,
+      "loss": 0.7275,
+      "step": 4075
+    },
+    {
+      "epoch": 0.7841477491342824,
+      "grad_norm": 0.8295590862216381,
+      "learning_rate": 2.349800559024492e-06,
+      "loss": 0.7959,
+      "step": 4076
+    },
+    {
+      "epoch": 0.784340130819546,
+      "grad_norm": 0.8826322001826067,
+      "learning_rate": 2.3457893383197415e-06,
+      "loss": 0.7089,
+      "step": 4077
+    },
+    {
+      "epoch": 0.7845325125048095,
+      "grad_norm": 0.8880092917105434,
+      "learning_rate": 2.3417810892431104e-06,
+      "loss": 0.7673,
+      "step": 4078
+    },
+    {
+      "epoch": 0.7847248941900731,
+      "grad_norm": 0.9327771854753525,
+      "learning_rate": 2.3377758133507455e-06,
+      "loss": 0.8499,
+      "step": 4079
+    },
+    {
+      "epoch": 0.7849172758753367,
+      "grad_norm": 0.8826249656845416,
+      "learning_rate": 2.3337735121976247e-06,
+      "loss": 0.8263,
+      "step": 4080
+    },
+    {
+      "epoch": 0.7851096575606002,
+      "grad_norm": 0.920835811813574,
+      "learning_rate": 2.32977418733758e-06,
+      "loss": 0.7577,
+      "step": 4081
+    },
+    {
+      "epoch": 0.7853020392458638,
+      "grad_norm": 0.9305104458166056,
+      "learning_rate": 2.3257778403232954e-06,
+      "loss": 0.7941,
+      "step": 4082
+    },
+    {
+      "epoch": 0.7854944209311273,
+      "grad_norm": 0.9210915521174973,
+      "learning_rate": 2.321784472706279e-06,
+      "loss": 0.784,
+      "step": 4083
+    },
+    {
+      "epoch": 0.7856868026163909,
+      "grad_norm": 0.8121790003813983,
+      "learning_rate": 2.317794086036901e-06,
+      "loss": 0.8074,
+      "step": 4084
+    },
+    {
+      "epoch": 0.7858791843016545,
+      "grad_norm": 0.8450572134979327,
+      "learning_rate": 2.3138066818643647e-06,
+      "loss": 0.7263,
+      "step": 4085
+    },
+    {
+      "epoch": 0.7860715659869181,
+      "grad_norm": 0.8790481011945044,
+      "learning_rate": 2.3098222617367184e-06,
+      "loss": 0.7879,
+      "step": 4086
+    },
+    {
+      "epoch": 0.7862639476721817,
+      "grad_norm": 0.9255094601578092,
+      "learning_rate": 2.30584082720085e-06,
+      "loss": 0.8319,
+      "step": 4087
+    },
+    {
+      "epoch": 0.7864563293574451,
+      "grad_norm": 0.9977501172294796,
+      "learning_rate": 2.301862379802492e-06,
+      "loss": 0.763,
+      "step": 4088
+    },
+    {
+      "epoch": 0.7866487110427087,
+      "grad_norm": 0.8980269658971917,
+      "learning_rate": 2.297886921086211e-06,
+      "loss": 0.7818,
+      "step": 4089
+    },
+    {
+      "epoch": 0.7868410927279723,
+      "grad_norm": 0.9142903401991305,
+      "learning_rate": 2.2939144525954194e-06,
+      "loss": 0.7691,
+      "step": 4090
+    },
+    {
+      "epoch": 0.7870334744132359,
+      "grad_norm": 0.9195107438802298,
+      "learning_rate": 2.2899449758723657e-06,
+      "loss": 0.715,
+      "step": 4091
+    },
+    {
+      "epoch": 0.7872258560984994,
+      "grad_norm": 0.8356784533790279,
+      "learning_rate": 2.285978492458134e-06,
+      "loss": 0.7996,
+      "step": 4092
+    },
+    {
+      "epoch": 0.787418237783763,
+      "grad_norm": 0.8241786363208338,
+      "learning_rate": 2.282015003892659e-06,
+      "loss": 0.7683,
+      "step": 4093
+    },
+    {
+      "epoch": 0.7876106194690266,
+      "grad_norm": 0.8347831650887655,
+      "learning_rate": 2.2780545117146947e-06,
+      "loss": 0.7792,
+      "step": 4094
+    },
+    {
+      "epoch": 0.7878030011542901,
+      "grad_norm": 0.9069057812664216,
+      "learning_rate": 2.2740970174618405e-06,
+      "loss": 0.7515,
+      "step": 4095
+    },
+    {
+      "epoch": 0.7879953828395537,
+      "grad_norm": 0.968950607972179,
+      "learning_rate": 2.270142522670541e-06,
+      "loss": 0.7813,
+      "step": 4096
+    },
+    {
+      "epoch": 0.7881877645248172,
+      "grad_norm": 0.9039658764220956,
+      "learning_rate": 2.2661910288760545e-06,
+      "loss": 0.8149,
+      "step": 4097
+    },
+    {
+      "epoch": 0.7883801462100808,
+      "grad_norm": 0.8240272041766706,
+      "learning_rate": 2.262242537612497e-06,
+      "loss": 0.731,
+      "step": 4098
+    },
+    {
+      "epoch": 0.7885725278953444,
+      "grad_norm": 0.8405039786301076,
+      "learning_rate": 2.258297050412804e-06,
+      "loss": 0.8009,
+      "step": 4099
+    },
+    {
+      "epoch": 0.788764909580608,
+      "grad_norm": 0.8984374854287174,
+      "learning_rate": 2.254354568808752e-06,
+      "loss": 0.7263,
+      "step": 4100
+    },
+    {
+      "epoch": 0.7889572912658714,
+      "grad_norm": 0.878649435046439,
+      "learning_rate": 2.2504150943309455e-06,
+      "loss": 0.7461,
+      "step": 4101
+    },
+    {
+      "epoch": 0.789149672951135,
+      "grad_norm": 0.8475048166361143,
+      "learning_rate": 2.246478628508827e-06,
+      "loss": 0.8102,
+      "step": 4102
+    },
+    {
+      "epoch": 0.7893420546363986,
+      "grad_norm": 0.5960673601443803,
+      "learning_rate": 2.242545172870665e-06,
+      "loss": 0.2922,
+      "step": 4103
+    },
+    {
+      "epoch": 0.7895344363216622,
+      "grad_norm": 0.9594129235831389,
+      "learning_rate": 2.238614728943561e-06,
+      "loss": 0.7579,
+      "step": 4104
+    },
+    {
+      "epoch": 0.7897268180069258,
+      "grad_norm": 0.895910537477269,
+      "learning_rate": 2.2346872982534584e-06,
+      "loss": 0.7927,
+      "step": 4105
+    },
+    {
+      "epoch": 0.7899191996921893,
+      "grad_norm": 0.6043874445753018,
+      "learning_rate": 2.2307628823251083e-06,
+      "loss": 0.3253,
+      "step": 4106
+    },
+    {
+      "epoch": 0.7901115813774529,
+      "grad_norm": 0.8818347622492401,
+      "learning_rate": 2.2268414826821117e-06,
+      "loss": 0.8283,
+      "step": 4107
+    },
+    {
+      "epoch": 0.7903039630627164,
+      "grad_norm": 0.8468221875085173,
+      "learning_rate": 2.222923100846893e-06,
+      "loss": 0.7888,
+      "step": 4108
+    },
+    {
+      "epoch": 0.79049634474798,
+      "grad_norm": 0.8520789915635194,
+      "learning_rate": 2.2190077383406938e-06,
+      "loss": 0.7737,
+      "step": 4109
+    },
+    {
+      "epoch": 0.7906887264332435,
+      "grad_norm": 0.9228921341054015,
+      "learning_rate": 2.2150953966835996e-06,
+      "loss": 0.7588,
+      "step": 4110
+    },
+    {
+      "epoch": 0.7908811081185071,
+      "grad_norm": 0.8781055855792993,
+      "learning_rate": 2.211186077394516e-06,
+      "loss": 0.729,
+      "step": 4111
+    },
+    {
+      "epoch": 0.7910734898037707,
+      "grad_norm": 0.8797139144777177,
+      "learning_rate": 2.207279781991173e-06,
+      "loss": 0.7926,
+      "step": 4112
+    },
+    {
+      "epoch": 0.7912658714890343,
+      "grad_norm": 0.9841275855597901,
+      "learning_rate": 2.2033765119901294e-06,
+      "loss": 0.7892,
+      "step": 4113
+    },
+    {
+      "epoch": 0.7914582531742979,
+      "grad_norm": 0.8453391138762986,
+      "learning_rate": 2.1994762689067705e-06,
+      "loss": 0.7828,
+      "step": 4114
+    },
+    {
+      "epoch": 0.7916506348595613,
+      "grad_norm": 0.8723293515368494,
+      "learning_rate": 2.1955790542553036e-06,
+      "loss": 0.7952,
+      "step": 4115
+    },
+    {
+      "epoch": 0.7918430165448249,
+      "grad_norm": 0.8427118435862677,
+      "learning_rate": 2.1916848695487615e-06,
+      "loss": 0.7589,
+      "step": 4116
+    },
+    {
+      "epoch": 0.7920353982300885,
+      "grad_norm": 0.9127939375372199,
+      "learning_rate": 2.1877937162990015e-06,
+      "loss": 0.7526,
+      "step": 4117
+    },
+    {
+      "epoch": 0.7922277799153521,
+      "grad_norm": 0.8811646881467043,
+      "learning_rate": 2.1839055960167e-06,
+      "loss": 0.7317,
+      "step": 4118
+    },
+    {
+      "epoch": 0.7924201616006156,
+      "grad_norm": 0.9095809253083126,
+      "learning_rate": 2.180020510211367e-06,
+      "loss": 0.8404,
+      "step": 4119
+    },
+    {
+      "epoch": 0.7926125432858792,
+      "grad_norm": 0.8653882558820288,
+      "learning_rate": 2.1761384603913203e-06,
+      "loss": 0.854,
+      "step": 4120
+    },
+    {
+      "epoch": 0.7928049249711427,
+      "grad_norm": 0.9237143059855143,
+      "learning_rate": 2.172259448063704e-06,
+      "loss": 0.8601,
+      "step": 4121
+    },
+    {
+      "epoch": 0.7929973066564063,
+      "grad_norm": 0.8931786121179608,
+      "learning_rate": 2.1683834747344913e-06,
+      "loss": 0.7469,
+      "step": 4122
+    },
+    {
+      "epoch": 0.7931896883416699,
+      "grad_norm": 0.8563072427068884,
+      "learning_rate": 2.1645105419084587e-06,
+      "loss": 0.846,
+      "step": 4123
+    },
+    {
+      "epoch": 0.7933820700269334,
+      "grad_norm": 0.8875538167755972,
+      "learning_rate": 2.160640651089221e-06,
+      "loss": 0.8108,
+      "step": 4124
+    },
+    {
+      "epoch": 0.793574451712197,
+      "grad_norm": 0.9108161228415991,
+      "learning_rate": 2.1567738037791998e-06,
+      "loss": 0.8247,
+      "step": 4125
+    },
+    {
+      "epoch": 0.7937668333974606,
+      "grad_norm": 0.8397007954914608,
+      "learning_rate": 2.152910001479638e-06,
+      "loss": 0.7799,
+      "step": 4126
+    },
+    {
+      "epoch": 0.7939592150827242,
+      "grad_norm": 0.8406507560839188,
+      "learning_rate": 2.1490492456905964e-06,
+      "loss": 0.8034,
+      "step": 4127
+    },
+    {
+      "epoch": 0.7941515967679876,
+      "grad_norm": 0.8317887363443469,
+      "learning_rate": 2.1451915379109546e-06,
+      "loss": 0.8286,
+      "step": 4128
+    },
+    {
+      "epoch": 0.7943439784532512,
+      "grad_norm": 0.8973790003711448,
+      "learning_rate": 2.141336879638406e-06,
+      "loss": 0.7888,
+      "step": 4129
+    },
+    {
+      "epoch": 0.7945363601385148,
+      "grad_norm": 0.8730359261933915,
+      "learning_rate": 2.1374852723694595e-06,
+      "loss": 0.8208,
+      "step": 4130
+    },
+    {
+      "epoch": 0.7947287418237784,
+      "grad_norm": 0.7899560476653634,
+      "learning_rate": 2.133636717599451e-06,
+      "loss": 0.7483,
+      "step": 4131
+    },
+    {
+      "epoch": 0.794921123509042,
+      "grad_norm": 0.8572382187386296,
+      "learning_rate": 2.1297912168225086e-06,
+      "loss": 0.7748,
+      "step": 4132
+    },
+    {
+      "epoch": 0.7951135051943055,
+      "grad_norm": 0.5790890754782116,
+      "learning_rate": 2.1259487715316e-06,
+      "loss": 0.277,
+      "step": 4133
+    },
+    {
+      "epoch": 0.795305886879569,
+      "grad_norm": 0.9144544955726401,
+      "learning_rate": 2.1221093832184903e-06,
+      "loss": 0.8056,
+      "step": 4134
+    },
+    {
+      "epoch": 0.7954982685648326,
+      "grad_norm": 0.6027346145280669,
+      "learning_rate": 2.118273053373757e-06,
+      "loss": 0.3255,
+      "step": 4135
+    },
+    {
+      "epoch": 0.7956906502500962,
+      "grad_norm": 0.923472575823799,
+      "learning_rate": 2.1144397834868034e-06,
+      "loss": 0.7709,
+      "step": 4136
+    },
+    {
+      "epoch": 0.7958830319353598,
+      "grad_norm": 0.8038263631035573,
+      "learning_rate": 2.1106095750458332e-06,
+      "loss": 0.8237,
+      "step": 4137
+    },
+    {
+      "epoch": 0.7960754136206233,
+      "grad_norm": 0.8873228716598746,
+      "learning_rate": 2.106782429537866e-06,
+      "loss": 0.8096,
+      "step": 4138
+    },
+    {
+      "epoch": 0.7962677953058869,
+      "grad_norm": 0.901397942737722,
+      "learning_rate": 2.1029583484487315e-06,
+      "loss": 0.797,
+      "step": 4139
+    },
+    {
+      "epoch": 0.7964601769911505,
+      "grad_norm": 0.943922602664964,
+      "learning_rate": 2.0991373332630683e-06,
+      "loss": 0.7658,
+      "step": 4140
+    },
+    {
+      "epoch": 0.796652558676414,
+      "grad_norm": 0.793451850527307,
+      "learning_rate": 2.0953193854643274e-06,
+      "loss": 0.7598,
+      "step": 4141
+    },
+    {
+      "epoch": 0.7968449403616775,
+      "grad_norm": 0.8332333656063411,
+      "learning_rate": 2.0915045065347673e-06,
+      "loss": 0.7995,
+      "step": 4142
+    },
+    {
+      "epoch": 0.7970373220469411,
+      "grad_norm": 0.9539057703498552,
+      "learning_rate": 2.0876926979554545e-06,
+      "loss": 0.7627,
+      "step": 4143
+    },
+    {
+      "epoch": 0.7972297037322047,
+      "grad_norm": 0.8247177651864366,
+      "learning_rate": 2.0838839612062633e-06,
+      "loss": 0.8256,
+      "step": 4144
+    },
+    {
+      "epoch": 0.7974220854174683,
+      "grad_norm": 0.8894124415675999,
+      "learning_rate": 2.080078297765884e-06,
+      "loss": 0.8347,
+      "step": 4145
+    },
+    {
+      "epoch": 0.7976144671027319,
+      "grad_norm": 0.8853857586465443,
+      "learning_rate": 2.0762757091117937e-06,
+      "loss": 0.8186,
+      "step": 4146
+    },
+    {
+      "epoch": 0.7978068487879953,
+      "grad_norm": 0.7736692576658899,
+      "learning_rate": 2.0724761967202987e-06,
+      "loss": 0.7833,
+      "step": 4147
+    },
+    {
+      "epoch": 0.7979992304732589,
+      "grad_norm": 0.9279417603692123,
+      "learning_rate": 2.0686797620664987e-06,
+      "loss": 0.8024,
+      "step": 4148
+    },
+    {
+      "epoch": 0.7981916121585225,
+      "grad_norm": 0.8520058827613523,
+      "learning_rate": 2.0648864066242937e-06,
+      "loss": 0.8049,
+      "step": 4149
+    },
+    {
+      "epoch": 0.7983839938437861,
+      "grad_norm": 0.835025451645815,
+      "learning_rate": 2.0610961318664013e-06,
+      "loss": 0.7874,
+      "step": 4150
+    },
+    {
+      "epoch": 0.7985763755290496,
+      "grad_norm": 0.9507688054073324,
+      "learning_rate": 2.0573089392643362e-06,
+      "loss": 0.7859,
+      "step": 4151
+    },
+    {
+      "epoch": 0.7987687572143132,
+      "grad_norm": 0.8619890642041846,
+      "learning_rate": 2.0535248302884147e-06,
+      "loss": 0.7711,
+      "step": 4152
+    },
+    {
+      "epoch": 0.7989611388995768,
+      "grad_norm": 0.9624831905185841,
+      "learning_rate": 2.0497438064077603e-06,
+      "loss": 0.784,
+      "step": 4153
+    },
+    {
+      "epoch": 0.7991535205848403,
+      "grad_norm": 0.9045487483587095,
+      "learning_rate": 2.045965869090295e-06,
+      "loss": 0.7714,
+      "step": 4154
+    },
+    {
+      "epoch": 0.7993459022701039,
+      "grad_norm": 1.019933089000337,
+      "learning_rate": 2.0421910198027452e-06,
+      "loss": 0.7389,
+      "step": 4155
+    },
+    {
+      "epoch": 0.7995382839553674,
+      "grad_norm": 0.7856444609222353,
+      "learning_rate": 2.0384192600106335e-06,
+      "loss": 0.8185,
+      "step": 4156
+    },
+    {
+      "epoch": 0.799730665640631,
+      "grad_norm": 0.8933923422062937,
+      "learning_rate": 2.0346505911782956e-06,
+      "loss": 0.7026,
+      "step": 4157
+    },
+    {
+      "epoch": 0.7999230473258946,
+      "grad_norm": 0.890662880037821,
+      "learning_rate": 2.0308850147688484e-06,
+      "loss": 0.8087,
+      "step": 4158
+    },
+    {
+      "epoch": 0.8001154290111582,
+      "grad_norm": 0.8918006994276415,
+      "learning_rate": 2.0271225322442255e-06,
+      "loss": 0.7696,
+      "step": 4159
+    },
+    {
+      "epoch": 0.8003078106964217,
+      "grad_norm": 0.8380478501260057,
+      "learning_rate": 2.0233631450651525e-06,
+      "loss": 0.774,
+      "step": 4160
+    },
+    {
+      "epoch": 0.8005001923816852,
+      "grad_norm": 0.8830605323873495,
+      "learning_rate": 2.019606854691145e-06,
+      "loss": 0.7859,
+      "step": 4161
+    },
+    {
+      "epoch": 0.8006925740669488,
+      "grad_norm": 0.8382896602178727,
+      "learning_rate": 2.0158536625805325e-06,
+      "loss": 0.7718,
+      "step": 4162
+    },
+    {
+      "epoch": 0.8008849557522124,
+      "grad_norm": 0.9222294377749531,
+      "learning_rate": 2.01210357019043e-06,
+      "loss": 0.7599,
+      "step": 4163
+    },
+    {
+      "epoch": 0.801077337437476,
+      "grad_norm": 0.9024471843040992,
+      "learning_rate": 2.008356578976752e-06,
+      "loss": 0.8036,
+      "step": 4164
+    },
+    {
+      "epoch": 0.8012697191227395,
+      "grad_norm": 0.9809512286255739,
+      "learning_rate": 2.004612690394212e-06,
+      "loss": 0.7962,
+      "step": 4165
+    },
+    {
+      "epoch": 0.8014621008080031,
+      "grad_norm": 1.0019981612516338,
+      "learning_rate": 2.0008719058963144e-06,
+      "loss": 0.7869,
+      "step": 4166
+    },
+    {
+      "epoch": 0.8016544824932667,
+      "grad_norm": 0.865009196746077,
+      "learning_rate": 1.997134226935361e-06,
+      "loss": 0.7997,
+      "step": 4167
+    },
+    {
+      "epoch": 0.8018468641785302,
+      "grad_norm": 0.9729922985864741,
+      "learning_rate": 1.9933996549624468e-06,
+      "loss": 0.777,
+      "step": 4168
+    },
+    {
+      "epoch": 0.8020392458637937,
+      "grad_norm": 0.9343215240933141,
+      "learning_rate": 1.9896681914274616e-06,
+      "loss": 0.7512,
+      "step": 4169
+    },
+    {
+      "epoch": 0.8022316275490573,
+      "grad_norm": 0.6126707186243839,
+      "learning_rate": 1.9859398377790872e-06,
+      "loss": 0.3343,
+      "step": 4170
+    },
+    {
+      "epoch": 0.8024240092343209,
+      "grad_norm": 0.8859426239148759,
+      "learning_rate": 1.982214595464804e-06,
+      "loss": 0.8141,
+      "step": 4171
+    },
+    {
+      "epoch": 0.8026163909195845,
+      "grad_norm": 0.8641828398201232,
+      "learning_rate": 1.97849246593087e-06,
+      "loss": 0.7666,
+      "step": 4172
+    },
+    {
+      "epoch": 0.8028087726048481,
+      "grad_norm": 0.6084196009075267,
+      "learning_rate": 1.9747734506223525e-06,
+      "loss": 0.3518,
+      "step": 4173
+    },
+    {
+      "epoch": 0.8030011542901115,
+      "grad_norm": 0.9265143468008795,
+      "learning_rate": 1.9710575509831008e-06,
+      "loss": 0.8668,
+      "step": 4174
+    },
+    {
+      "epoch": 0.8031935359753751,
+      "grad_norm": 0.8696621551981761,
+      "learning_rate": 1.967344768455747e-06,
+      "loss": 0.7368,
+      "step": 4175
+    },
+    {
+      "epoch": 0.8033859176606387,
+      "grad_norm": 0.8722562721442059,
+      "learning_rate": 1.9636351044817292e-06,
+      "loss": 0.8378,
+      "step": 4176
+    },
+    {
+      "epoch": 0.8035782993459023,
+      "grad_norm": 0.9564218534174184,
+      "learning_rate": 1.9599285605012643e-06,
+      "loss": 0.7811,
+      "step": 4177
+    },
+    {
+      "epoch": 0.8037706810311658,
+      "grad_norm": 0.9983082723492495,
+      "learning_rate": 1.9562251379533593e-06,
+      "loss": 0.8174,
+      "step": 4178
+    },
+    {
+      "epoch": 0.8039630627164294,
+      "grad_norm": 0.8510593178658997,
+      "learning_rate": 1.952524838275811e-06,
+      "loss": 0.7648,
+      "step": 4179
+    },
+    {
+      "epoch": 0.804155444401693,
+      "grad_norm": 0.6022097601833074,
+      "learning_rate": 1.9488276629052026e-06,
+      "loss": 0.3007,
+      "step": 4180
+    },
+    {
+      "epoch": 0.8043478260869565,
+      "grad_norm": 0.9043318847270205,
+      "learning_rate": 1.945133613276907e-06,
+      "loss": 0.7629,
+      "step": 4181
+    },
+    {
+      "epoch": 0.8045402077722201,
+      "grad_norm": 0.8675701214066187,
+      "learning_rate": 1.941442690825076e-06,
+      "loss": 0.7709,
+      "step": 4182
+    },
+    {
+      "epoch": 0.8047325894574836,
+      "grad_norm": 0.841955521770211,
+      "learning_rate": 1.937754896982663e-06,
+      "loss": 0.7331,
+      "step": 4183
+    },
+    {
+      "epoch": 0.8049249711427472,
+      "grad_norm": 0.9091538536576826,
+      "learning_rate": 1.9340702331813842e-06,
+      "loss": 0.7774,
+      "step": 4184
+    },
+    {
+      "epoch": 0.8051173528280108,
+      "grad_norm": 0.8825486792108282,
+      "learning_rate": 1.9303887008517618e-06,
+      "loss": 0.7759,
+      "step": 4185
+    },
+    {
+      "epoch": 0.8053097345132744,
+      "grad_norm": 0.8672241103933366,
+      "learning_rate": 1.9267103014230935e-06,
+      "loss": 0.7638,
+      "step": 4186
+    },
+    {
+      "epoch": 0.805502116198538,
+      "grad_norm": 0.9217935189617378,
+      "learning_rate": 1.923035036323452e-06,
+      "loss": 0.7906,
+      "step": 4187
+    },
+    {
+      "epoch": 0.8056944978838014,
+      "grad_norm": 0.8660645543094998,
+      "learning_rate": 1.91936290697971e-06,
+      "loss": 0.8359,
+      "step": 4188
+    },
+    {
+      "epoch": 0.805886879569065,
+      "grad_norm": 0.8507471577987193,
+      "learning_rate": 1.9156939148175125e-06,
+      "loss": 0.7472,
+      "step": 4189
+    },
+    {
+      "epoch": 0.8060792612543286,
+      "grad_norm": 0.8733227989257416,
+      "learning_rate": 1.9120280612612873e-06,
+      "loss": 0.7965,
+      "step": 4190
+    },
+    {
+      "epoch": 0.8062716429395922,
+      "grad_norm": 0.9034189094315214,
+      "learning_rate": 1.9083653477342467e-06,
+      "loss": 0.7926,
+      "step": 4191
+    },
+    {
+      "epoch": 0.8064640246248557,
+      "grad_norm": 0.9024472281635157,
+      "learning_rate": 1.904705775658381e-06,
+      "loss": 0.7978,
+      "step": 4192
+    },
+    {
+      "epoch": 0.8066564063101193,
+      "grad_norm": 0.8804151840813323,
+      "learning_rate": 1.9010493464544621e-06,
+      "loss": 0.7917,
+      "step": 4193
+    },
+    {
+      "epoch": 0.8068487879953828,
+      "grad_norm": 0.881621070215491,
+      "learning_rate": 1.8973960615420416e-06,
+      "loss": 0.8056,
+      "step": 4194
+    },
+    {
+      "epoch": 0.8070411696806464,
+      "grad_norm": 0.8727035764755112,
+      "learning_rate": 1.8937459223394517e-06,
+      "loss": 0.8016,
+      "step": 4195
+    },
+    {
+      "epoch": 0.80723355136591,
+      "grad_norm": 0.8092086996243925,
+      "learning_rate": 1.8900989302637985e-06,
+      "loss": 0.7461,
+      "step": 4196
+    },
+    {
+      "epoch": 0.8074259330511735,
+      "grad_norm": 0.9104673427935519,
+      "learning_rate": 1.8864550867309771e-06,
+      "loss": 0.8001,
+      "step": 4197
+    },
+    {
+      "epoch": 0.8076183147364371,
+      "grad_norm": 0.8136070448280946,
+      "learning_rate": 1.8828143931556442e-06,
+      "loss": 0.766,
+      "step": 4198
+    },
+    {
+      "epoch": 0.8078106964217007,
+      "grad_norm": 0.8706839574002638,
+      "learning_rate": 1.8791768509512487e-06,
+      "loss": 0.7879,
+      "step": 4199
+    },
+    {
+      "epoch": 0.8080030781069643,
+      "grad_norm": 0.9313532060871341,
+      "learning_rate": 1.875542461530011e-06,
+      "loss": 0.8053,
+      "step": 4200
+    },
+    {
+      "epoch": 0.8081954597922277,
+      "grad_norm": 0.9381957047031164,
+      "learning_rate": 1.871911226302917e-06,
+      "loss": 0.7773,
+      "step": 4201
+    },
+    {
+      "epoch": 0.8083878414774913,
+      "grad_norm": 0.9285213531950582,
+      "learning_rate": 1.868283146679747e-06,
+      "loss": 0.7437,
+      "step": 4202
+    },
+    {
+      "epoch": 0.8085802231627549,
+      "grad_norm": 0.8608532679127132,
+      "learning_rate": 1.8646582240690414e-06,
+      "loss": 0.714,
+      "step": 4203
+    },
+    {
+      "epoch": 0.8087726048480185,
+      "grad_norm": 0.9362232468887034,
+      "learning_rate": 1.8610364598781227e-06,
+      "loss": 0.8473,
+      "step": 4204
+    },
+    {
+      "epoch": 0.8089649865332821,
+      "grad_norm": 0.8195220396029873,
+      "learning_rate": 1.8574178555130818e-06,
+      "loss": 0.8456,
+      "step": 4205
+    },
+    {
+      "epoch": 0.8091573682185456,
+      "grad_norm": 0.9271030464814155,
+      "learning_rate": 1.8538024123787868e-06,
+      "loss": 0.8403,
+      "step": 4206
+    },
+    {
+      "epoch": 0.8093497499038091,
+      "grad_norm": 0.8608398437529724,
+      "learning_rate": 1.8501901318788773e-06,
+      "loss": 0.7599,
+      "step": 4207
+    },
+    {
+      "epoch": 0.8095421315890727,
+      "grad_norm": 0.9194275154503163,
+      "learning_rate": 1.8465810154157626e-06,
+      "loss": 0.7413,
+      "step": 4208
+    },
+    {
+      "epoch": 0.8097345132743363,
+      "grad_norm": 0.8328298113024317,
+      "learning_rate": 1.8429750643906331e-06,
+      "loss": 0.7895,
+      "step": 4209
+    },
+    {
+      "epoch": 0.8099268949595998,
+      "grad_norm": 0.9156064299777941,
+      "learning_rate": 1.8393722802034331e-06,
+      "loss": 0.764,
+      "step": 4210
+    },
+    {
+      "epoch": 0.8101192766448634,
+      "grad_norm": 0.8239026774508166,
+      "learning_rate": 1.835772664252895e-06,
+      "loss": 0.7511,
+      "step": 4211
+    },
+    {
+      "epoch": 0.810311658330127,
+      "grad_norm": 0.9681454347448929,
+      "learning_rate": 1.832176217936511e-06,
+      "loss": 0.8152,
+      "step": 4212
+    },
+    {
+      "epoch": 0.8105040400153906,
+      "grad_norm": 0.8838843734521495,
+      "learning_rate": 1.8285829426505453e-06,
+      "loss": 0.8265,
+      "step": 4213
+    },
+    {
+      "epoch": 0.8106964217006541,
+      "grad_norm": 0.913422488148725,
+      "learning_rate": 1.8249928397900351e-06,
+      "loss": 0.8382,
+      "step": 4214
+    },
+    {
+      "epoch": 0.8108888033859176,
+      "grad_norm": 0.9229005685459502,
+      "learning_rate": 1.8214059107487726e-06,
+      "loss": 0.8014,
+      "step": 4215
+    },
+    {
+      "epoch": 0.8110811850711812,
+      "grad_norm": 0.8718955857764944,
+      "learning_rate": 1.8178221569193343e-06,
+      "loss": 0.814,
+      "step": 4216
+    },
+    {
+      "epoch": 0.8112735667564448,
+      "grad_norm": 0.6395255080370342,
+      "learning_rate": 1.8142415796930568e-06,
+      "loss": 0.3119,
+      "step": 4217
+    },
+    {
+      "epoch": 0.8114659484417084,
+      "grad_norm": 0.8802419432943592,
+      "learning_rate": 1.8106641804600411e-06,
+      "loss": 0.8041,
+      "step": 4218
+    },
+    {
+      "epoch": 0.8116583301269719,
+      "grad_norm": 0.8994638086928565,
+      "learning_rate": 1.8070899606091586e-06,
+      "loss": 0.7673,
+      "step": 4219
+    },
+    {
+      "epoch": 0.8118507118122354,
+      "grad_norm": 0.9108557336988482,
+      "learning_rate": 1.8035189215280423e-06,
+      "loss": 0.7735,
+      "step": 4220
+    },
+    {
+      "epoch": 0.812043093497499,
+      "grad_norm": 0.8726036232413998,
+      "learning_rate": 1.799951064603095e-06,
+      "loss": 0.8429,
+      "step": 4221
+    },
+    {
+      "epoch": 0.8122354751827626,
+      "grad_norm": 0.8443960659670143,
+      "learning_rate": 1.7963863912194768e-06,
+      "loss": 0.8011,
+      "step": 4222
+    },
+    {
+      "epoch": 0.8124278568680262,
+      "grad_norm": 0.6530164393057462,
+      "learning_rate": 1.7928249027611255e-06,
+      "loss": 0.3371,
+      "step": 4223
+    },
+    {
+      "epoch": 0.8126202385532897,
+      "grad_norm": 0.9018076918863142,
+      "learning_rate": 1.789266600610724e-06,
+      "loss": 0.7626,
+      "step": 4224
+    },
+    {
+      "epoch": 0.8128126202385533,
+      "grad_norm": 0.6424114761989578,
+      "learning_rate": 1.7857114861497337e-06,
+      "loss": 0.3438,
+      "step": 4225
+    },
+    {
+      "epoch": 0.8130050019238169,
+      "grad_norm": 0.964762386747457,
+      "learning_rate": 1.782159560758373e-06,
+      "loss": 0.8666,
+      "step": 4226
+    },
+    {
+      "epoch": 0.8131973836090804,
+      "grad_norm": 0.9026900368464168,
+      "learning_rate": 1.7786108258156154e-06,
+      "loss": 0.7482,
+      "step": 4227
+    },
+    {
+      "epoch": 0.8133897652943439,
+      "grad_norm": 0.9414121650434548,
+      "learning_rate": 1.7750652826992077e-06,
+      "loss": 0.7889,
+      "step": 4228
+    },
+    {
+      "epoch": 0.8135821469796075,
+      "grad_norm": 0.8930207685636924,
+      "learning_rate": 1.7715229327856498e-06,
+      "loss": 0.7682,
+      "step": 4229
+    },
+    {
+      "epoch": 0.8137745286648711,
+      "grad_norm": 0.8388002025690184,
+      "learning_rate": 1.7679837774502052e-06,
+      "loss": 0.8189,
+      "step": 4230
+    },
+    {
+      "epoch": 0.8139669103501347,
+      "grad_norm": 0.8694688678920058,
+      "learning_rate": 1.7644478180668945e-06,
+      "loss": 0.7092,
+      "step": 4231
+    },
+    {
+      "epoch": 0.8141592920353983,
+      "grad_norm": 0.9632004125839141,
+      "learning_rate": 1.7609150560084986e-06,
+      "loss": 0.8069,
+      "step": 4232
+    },
+    {
+      "epoch": 0.8143516737206618,
+      "grad_norm": 1.0437866751804878,
+      "learning_rate": 1.7573854926465582e-06,
+      "loss": 0.7875,
+      "step": 4233
+    },
+    {
+      "epoch": 0.8145440554059253,
+      "grad_norm": 0.9069091083019756,
+      "learning_rate": 1.7538591293513685e-06,
+      "loss": 0.8086,
+      "step": 4234
+    },
+    {
+      "epoch": 0.8147364370911889,
+      "grad_norm": 0.8751978035095651,
+      "learning_rate": 1.7503359674919929e-06,
+      "loss": 0.8272,
+      "step": 4235
+    },
+    {
+      "epoch": 0.8149288187764525,
+      "grad_norm": 0.9911749157573895,
+      "learning_rate": 1.746816008436234e-06,
+      "loss": 0.7798,
+      "step": 4236
+    },
+    {
+      "epoch": 0.8151212004617161,
+      "grad_norm": 0.8818075599421732,
+      "learning_rate": 1.7432992535506687e-06,
+      "loss": 0.7702,
+      "step": 4237
+    },
+    {
+      "epoch": 0.8153135821469796,
+      "grad_norm": 0.586141827965161,
+      "learning_rate": 1.7397857042006194e-06,
+      "loss": 0.3289,
+      "step": 4238
+    },
+    {
+      "epoch": 0.8155059638322432,
+      "grad_norm": 0.9578750940088012,
+      "learning_rate": 1.736275361750167e-06,
+      "loss": 0.807,
+      "step": 4239
+    },
+    {
+      "epoch": 0.8156983455175068,
+      "grad_norm": 0.9929884186168799,
+      "learning_rate": 1.7327682275621506e-06,
+      "loss": 0.8396,
+      "step": 4240
+    },
+    {
+      "epoch": 0.8158907272027703,
+      "grad_norm": 0.9098039047384134,
+      "learning_rate": 1.7292643029981525e-06,
+      "loss": 0.7756,
+      "step": 4241
+    },
+    {
+      "epoch": 0.8160831088880338,
+      "grad_norm": 1.0024706485840764,
+      "learning_rate": 1.7257635894185232e-06,
+      "loss": 0.8265,
+      "step": 4242
+    },
+    {
+      "epoch": 0.8162754905732974,
+      "grad_norm": 0.8676634178200934,
+      "learning_rate": 1.7222660881823594e-06,
+      "loss": 0.769,
+      "step": 4243
+    },
+    {
+      "epoch": 0.816467872258561,
+      "grad_norm": 0.6184185678390774,
+      "learning_rate": 1.7187718006475117e-06,
+      "loss": 0.311,
+      "step": 4244
+    },
+    {
+      "epoch": 0.8166602539438246,
+      "grad_norm": 0.9656321413120306,
+      "learning_rate": 1.7152807281705809e-06,
+      "loss": 0.8294,
+      "step": 4245
+    },
+    {
+      "epoch": 0.8168526356290882,
+      "grad_norm": 0.6121859585095518,
+      "learning_rate": 1.7117928721069233e-06,
+      "loss": 0.3152,
+      "step": 4246
+    },
+    {
+      "epoch": 0.8170450173143516,
+      "grad_norm": 0.7868293557788892,
+      "learning_rate": 1.708308233810644e-06,
+      "loss": 0.8125,
+      "step": 4247
+    },
+    {
+      "epoch": 0.8172373989996152,
+      "grad_norm": 0.8936361794692677,
+      "learning_rate": 1.704826814634597e-06,
+      "loss": 0.8008,
+      "step": 4248
+    },
+    {
+      "epoch": 0.8174297806848788,
+      "grad_norm": 0.8259003491700087,
+      "learning_rate": 1.701348615930397e-06,
+      "loss": 0.7703,
+      "step": 4249
+    },
+    {
+      "epoch": 0.8176221623701424,
+      "grad_norm": 0.7887422164193342,
+      "learning_rate": 1.6978736390483896e-06,
+      "loss": 0.7579,
+      "step": 4250
+    },
+    {
+      "epoch": 0.8178145440554059,
+      "grad_norm": 0.9704459352585177,
+      "learning_rate": 1.6944018853376898e-06,
+      "loss": 0.776,
+      "step": 4251
+    },
+    {
+      "epoch": 0.8180069257406695,
+      "grad_norm": 0.9315696135230932,
+      "learning_rate": 1.6909333561461471e-06,
+      "loss": 0.8281,
+      "step": 4252
+    },
+    {
+      "epoch": 0.818199307425933,
+      "grad_norm": 0.9313117892751663,
+      "learning_rate": 1.6874680528203657e-06,
+      "loss": 0.8154,
+      "step": 4253
+    },
+    {
+      "epoch": 0.8183916891111966,
+      "grad_norm": 0.8380511942587682,
+      "learning_rate": 1.6840059767056949e-06,
+      "loss": 0.7499,
+      "step": 4254
+    },
+    {
+      "epoch": 0.8185840707964602,
+      "grad_norm": 0.9624132290311658,
+      "learning_rate": 1.6805471291462316e-06,
+      "loss": 0.7701,
+      "step": 4255
+    },
+    {
+      "epoch": 0.8187764524817237,
+      "grad_norm": 0.8294987025211037,
+      "learning_rate": 1.6770915114848197e-06,
+      "loss": 0.7743,
+      "step": 4256
+    },
+    {
+      "epoch": 0.8189688341669873,
+      "grad_norm": 0.8493510862349811,
+      "learning_rate": 1.67363912506305e-06,
+      "loss": 0.7715,
+      "step": 4257
+    },
+    {
+      "epoch": 0.8191612158522509,
+      "grad_norm": 0.5850017314092377,
+      "learning_rate": 1.6701899712212565e-06,
+      "loss": 0.3238,
+      "step": 4258
+    },
+    {
+      "epoch": 0.8193535975375145,
+      "grad_norm": 0.9465570103964469,
+      "learning_rate": 1.66674405129852e-06,
+      "loss": 0.8338,
+      "step": 4259
+    },
+    {
+      "epoch": 0.8195459792227779,
+      "grad_norm": 0.8616853471596508,
+      "learning_rate": 1.6633013666326636e-06,
+      "loss": 0.8244,
+      "step": 4260
+    },
+    {
+      "epoch": 0.8197383609080415,
+      "grad_norm": 0.9066856943643088,
+      "learning_rate": 1.6598619185602616e-06,
+      "loss": 0.7957,
+      "step": 4261
+    },
+    {
+      "epoch": 0.8199307425933051,
+      "grad_norm": 0.8954015794528074,
+      "learning_rate": 1.656425708416617e-06,
+      "loss": 0.8058,
+      "step": 4262
+    },
+    {
+      "epoch": 0.8201231242785687,
+      "grad_norm": 0.9100545119882483,
+      "learning_rate": 1.6529927375357957e-06,
+      "loss": 0.8247,
+      "step": 4263
+    },
+    {
+      "epoch": 0.8203155059638323,
+      "grad_norm": 0.8929041224708161,
+      "learning_rate": 1.6495630072505841e-06,
+      "loss": 0.7776,
+      "step": 4264
+    },
+    {
+      "epoch": 0.8205078876490958,
+      "grad_norm": 0.8404939221530606,
+      "learning_rate": 1.6461365188925304e-06,
+      "loss": 0.756,
+      "step": 4265
+    },
+    {
+      "epoch": 0.8207002693343594,
+      "grad_norm": 0.8350469118357706,
+      "learning_rate": 1.642713273791914e-06,
+      "loss": 0.7088,
+      "step": 4266
+    },
+    {
+      "epoch": 0.8208926510196229,
+      "grad_norm": 0.8154384049987023,
+      "learning_rate": 1.6392932732777489e-06,
+      "loss": 0.776,
+      "step": 4267
+    },
+    {
+      "epoch": 0.8210850327048865,
+      "grad_norm": 0.8026882604894876,
+      "learning_rate": 1.6358765186778057e-06,
+      "loss": 0.8079,
+      "step": 4268
+    },
+    {
+      "epoch": 0.82127741439015,
+      "grad_norm": 0.8777064323363232,
+      "learning_rate": 1.6324630113185835e-06,
+      "loss": 0.778,
+      "step": 4269
+    },
+    {
+      "epoch": 0.8214697960754136,
+      "grad_norm": 0.8320087612940796,
+      "learning_rate": 1.629052752525323e-06,
+      "loss": 0.8213,
+      "step": 4270
+    },
+    {
+      "epoch": 0.8216621777606772,
+      "grad_norm": 1.0728983918357866,
+      "learning_rate": 1.625645743622003e-06,
+      "loss": 0.8474,
+      "step": 4271
+    },
+    {
+      "epoch": 0.8218545594459408,
+      "grad_norm": 0.8886666028177266,
+      "learning_rate": 1.6222419859313443e-06,
+      "loss": 0.7853,
+      "step": 4272
+    },
+    {
+      "epoch": 0.8220469411312044,
+      "grad_norm": 0.828801878690678,
+      "learning_rate": 1.6188414807747999e-06,
+      "loss": 0.839,
+      "step": 4273
+    },
+    {
+      "epoch": 0.8222393228164678,
+      "grad_norm": 0.982096447683237,
+      "learning_rate": 1.6154442294725636e-06,
+      "loss": 0.8536,
+      "step": 4274
+    },
+    {
+      "epoch": 0.8224317045017314,
+      "grad_norm": 0.8645154190262186,
+      "learning_rate": 1.6120502333435695e-06,
+      "loss": 0.844,
+      "step": 4275
+    },
+    {
+      "epoch": 0.822624086186995,
+      "grad_norm": 0.8892336754495268,
+      "learning_rate": 1.6086594937054767e-06,
+      "loss": 0.7789,
+      "step": 4276
+    },
+    {
+      "epoch": 0.8228164678722586,
+      "grad_norm": 0.8565513603980988,
+      "learning_rate": 1.6052720118746923e-06,
+      "loss": 0.7689,
+      "step": 4277
+    },
+    {
+      "epoch": 0.8230088495575221,
+      "grad_norm": 0.7716950860041811,
+      "learning_rate": 1.6018877891663521e-06,
+      "loss": 0.7505,
+      "step": 4278
+    },
+    {
+      "epoch": 0.8232012312427857,
+      "grad_norm": 0.9327087735380407,
+      "learning_rate": 1.5985068268943283e-06,
+      "loss": 0.8017,
+      "step": 4279
+    },
+    {
+      "epoch": 0.8233936129280492,
+      "grad_norm": 0.7769637933730787,
+      "learning_rate": 1.5951291263712255e-06,
+      "loss": 0.7095,
+      "step": 4280
+    },
+    {
+      "epoch": 0.8235859946133128,
+      "grad_norm": 0.8477111263673202,
+      "learning_rate": 1.5917546889083834e-06,
+      "loss": 0.7799,
+      "step": 4281
+    },
+    {
+      "epoch": 0.8237783762985764,
+      "grad_norm": 0.821007099718049,
+      "learning_rate": 1.5883835158158767e-06,
+      "loss": 0.7755,
+      "step": 4282
+    },
+    {
+      "epoch": 0.8239707579838399,
+      "grad_norm": 0.5571973325733733,
+      "learning_rate": 1.5850156084025091e-06,
+      "loss": 0.3556,
+      "step": 4283
+    },
+    {
+      "epoch": 0.8241631396691035,
+      "grad_norm": 0.8877120085097439,
+      "learning_rate": 1.5816509679758185e-06,
+      "loss": 0.7535,
+      "step": 4284
+    },
+    {
+      "epoch": 0.8243555213543671,
+      "grad_norm": 0.5991811147993774,
+      "learning_rate": 1.578289595842074e-06,
+      "loss": 0.3214,
+      "step": 4285
+    },
+    {
+      "epoch": 0.8245479030396307,
+      "grad_norm": 0.847384965880218,
+      "learning_rate": 1.5749314933062754e-06,
+      "loss": 0.8034,
+      "step": 4286
+    },
+    {
+      "epoch": 0.8247402847248941,
+      "grad_norm": 0.8552646471201931,
+      "learning_rate": 1.5715766616721584e-06,
+      "loss": 0.7584,
+      "step": 4287
+    },
+    {
+      "epoch": 0.8249326664101577,
+      "grad_norm": 0.8048737853679329,
+      "learning_rate": 1.5682251022421757e-06,
+      "loss": 0.7697,
+      "step": 4288
+    },
+    {
+      "epoch": 0.8251250480954213,
+      "grad_norm": 0.920097898709925,
+      "learning_rate": 1.5648768163175277e-06,
+      "loss": 0.8143,
+      "step": 4289
+    },
+    {
+      "epoch": 0.8253174297806849,
+      "grad_norm": 0.8600867470435791,
+      "learning_rate": 1.5615318051981243e-06,
+      "loss": 0.7521,
+      "step": 4290
+    },
+    {
+      "epoch": 0.8255098114659485,
+      "grad_norm": 0.8629429722260675,
+      "learning_rate": 1.5581900701826226e-06,
+      "loss": 0.8287,
+      "step": 4291
+    },
+    {
+      "epoch": 0.825702193151212,
+      "grad_norm": 0.8392983923572437,
+      "learning_rate": 1.5548516125683976e-06,
+      "loss": 0.7591,
+      "step": 4292
+    },
+    {
+      "epoch": 0.8258945748364755,
+      "grad_norm": 0.8277301200627027,
+      "learning_rate": 1.5515164336515465e-06,
+      "loss": 0.7804,
+      "step": 4293
+    },
+    {
+      "epoch": 0.8260869565217391,
+      "grad_norm": 0.8523008859830253,
+      "learning_rate": 1.5481845347269077e-06,
+      "loss": 0.7563,
+      "step": 4294
+    },
+    {
+      "epoch": 0.8262793382070027,
+      "grad_norm": 0.9264393392533055,
+      "learning_rate": 1.5448559170880373e-06,
+      "loss": 0.817,
+      "step": 4295
+    },
+    {
+      "epoch": 0.8264717198922663,
+      "grad_norm": 0.9216702014133332,
+      "learning_rate": 1.5415305820272198e-06,
+      "loss": 0.8124,
+      "step": 4296
+    },
+    {
+      "epoch": 0.8266641015775298,
+      "grad_norm": 0.901978194204991,
+      "learning_rate": 1.5382085308354633e-06,
+      "loss": 0.7444,
+      "step": 4297
+    },
+    {
+      "epoch": 0.8268564832627934,
+      "grad_norm": 0.890000114166881,
+      "learning_rate": 1.534889764802503e-06,
+      "loss": 0.821,
+      "step": 4298
+    },
+    {
+      "epoch": 0.827048864948057,
+      "grad_norm": 0.9063454374924086,
+      "learning_rate": 1.5315742852167992e-06,
+      "loss": 0.8122,
+      "step": 4299
+    },
+    {
+      "epoch": 0.8272412466333205,
+      "grad_norm": 0.9263069651126002,
+      "learning_rate": 1.5282620933655312e-06,
+      "loss": 0.8333,
+      "step": 4300
+    },
+    {
+      "epoch": 0.827433628318584,
+      "grad_norm": 0.8579476956874136,
+      "learning_rate": 1.5249531905346138e-06,
+      "loss": 0.7973,
+      "step": 4301
+    },
+    {
+      "epoch": 0.8276260100038476,
+      "grad_norm": 0.9202054745456872,
+      "learning_rate": 1.521647578008667e-06,
+      "loss": 0.7722,
+      "step": 4302
+    },
+    {
+      "epoch": 0.8278183916891112,
+      "grad_norm": 0.8349872138359135,
+      "learning_rate": 1.5183452570710522e-06,
+      "loss": 0.8038,
+      "step": 4303
+    },
+    {
+      "epoch": 0.8280107733743748,
+      "grad_norm": 0.8252218943801729,
+      "learning_rate": 1.5150462290038392e-06,
+      "loss": 0.7421,
+      "step": 4304
+    },
+    {
+      "epoch": 0.8282031550596384,
+      "grad_norm": 0.938476379274415,
+      "learning_rate": 1.511750495087827e-06,
+      "loss": 0.8155,
+      "step": 4305
+    },
+    {
+      "epoch": 0.8283955367449019,
+      "grad_norm": 0.8717448710462183,
+      "learning_rate": 1.5084580566025309e-06,
+      "loss": 0.753,
+      "step": 4306
+    },
+    {
+      "epoch": 0.8285879184301654,
+      "grad_norm": 0.8351530096597882,
+      "learning_rate": 1.5051689148261895e-06,
+      "loss": 0.8012,
+      "step": 4307
+    },
+    {
+      "epoch": 0.828780300115429,
+      "grad_norm": 0.9308622440824351,
+      "learning_rate": 1.5018830710357612e-06,
+      "loss": 0.758,
+      "step": 4308
+    },
+    {
+      "epoch": 0.8289726818006926,
+      "grad_norm": 0.9257165774055186,
+      "learning_rate": 1.4986005265069204e-06,
+      "loss": 0.7921,
+      "step": 4309
+    },
+    {
+      "epoch": 0.8291650634859561,
+      "grad_norm": 0.9052880438987351,
+      "learning_rate": 1.4953212825140728e-06,
+      "loss": 0.7833,
+      "step": 4310
+    },
+    {
+      "epoch": 0.8293574451712197,
+      "grad_norm": 0.8925335021035182,
+      "learning_rate": 1.4920453403303249e-06,
+      "loss": 0.7882,
+      "step": 4311
+    },
+    {
+      "epoch": 0.8295498268564833,
+      "grad_norm": 0.8510713768932524,
+      "learning_rate": 1.4887727012275112e-06,
+      "loss": 0.7902,
+      "step": 4312
+    },
+    {
+      "epoch": 0.8297422085417469,
+      "grad_norm": 0.8471688836106497,
+      "learning_rate": 1.4855033664761898e-06,
+      "loss": 0.7546,
+      "step": 4313
+    },
+    {
+      "epoch": 0.8299345902270104,
+      "grad_norm": 0.8513378857567512,
+      "learning_rate": 1.48223733734562e-06,
+      "loss": 0.7398,
+      "step": 4314
+    },
+    {
+      "epoch": 0.8301269719122739,
+      "grad_norm": 0.9331650091922404,
+      "learning_rate": 1.4789746151037942e-06,
+      "loss": 0.8114,
+      "step": 4315
+    },
+    {
+      "epoch": 0.8303193535975375,
+      "grad_norm": 0.9024411084826682,
+      "learning_rate": 1.475715201017407e-06,
+      "loss": 0.8478,
+      "step": 4316
+    },
+    {
+      "epoch": 0.8305117352828011,
+      "grad_norm": 0.8949036242122693,
+      "learning_rate": 1.4724590963518803e-06,
+      "loss": 0.811,
+      "step": 4317
+    },
+    {
+      "epoch": 0.8307041169680647,
+      "grad_norm": 0.8511786688721773,
+      "learning_rate": 1.4692063023713444e-06,
+      "loss": 0.7492,
+      "step": 4318
+    },
+    {
+      "epoch": 0.8308964986533282,
+      "grad_norm": 0.9036216616390776,
+      "learning_rate": 1.4659568203386464e-06,
+      "loss": 0.799,
+      "step": 4319
+    },
+    {
+      "epoch": 0.8310888803385917,
+      "grad_norm": 0.9499242908526755,
+      "learning_rate": 1.4627106515153456e-06,
+      "loss": 0.779,
+      "step": 4320
+    },
+    {
+      "epoch": 0.8312812620238553,
+      "grad_norm": 0.8860988951612047,
+      "learning_rate": 1.4594677971617178e-06,
+      "loss": 0.81,
+      "step": 4321
+    },
+    {
+      "epoch": 0.8314736437091189,
+      "grad_norm": 0.8761763915693633,
+      "learning_rate": 1.4562282585367493e-06,
+      "loss": 0.7232,
+      "step": 4322
+    },
+    {
+      "epoch": 0.8316660253943825,
+      "grad_norm": 0.9063176573494522,
+      "learning_rate": 1.452992036898142e-06,
+      "loss": 0.8004,
+      "step": 4323
+    },
+    {
+      "epoch": 0.831858407079646,
+      "grad_norm": 0.8544054525950411,
+      "learning_rate": 1.4497591335023087e-06,
+      "loss": 0.7558,
+      "step": 4324
+    },
+    {
+      "epoch": 0.8320507887649096,
+      "grad_norm": 0.9655838387560247,
+      "learning_rate": 1.446529549604373e-06,
+      "loss": 0.8176,
+      "step": 4325
+    },
+    {
+      "epoch": 0.8322431704501732,
+      "grad_norm": 0.8822262252779536,
+      "learning_rate": 1.4433032864581687e-06,
+      "loss": 0.7879,
+      "step": 4326
+    },
+    {
+      "epoch": 0.8324355521354367,
+      "grad_norm": 0.8536207335176483,
+      "learning_rate": 1.4400803453162482e-06,
+      "loss": 0.8241,
+      "step": 4327
+    },
+    {
+      "epoch": 0.8326279338207002,
+      "grad_norm": 0.8398872852842305,
+      "learning_rate": 1.4368607274298596e-06,
+      "loss": 0.7815,
+      "step": 4328
+    },
+    {
+      "epoch": 0.8328203155059638,
+      "grad_norm": 0.8734530447794816,
+      "learning_rate": 1.4336444340489775e-06,
+      "loss": 0.7867,
+      "step": 4329
+    },
+    {
+      "epoch": 0.8330126971912274,
+      "grad_norm": 0.871957533410548,
+      "learning_rate": 1.430431466422273e-06,
+      "loss": 0.8236,
+      "step": 4330
+    },
+    {
+      "epoch": 0.833205078876491,
+      "grad_norm": 0.8794358907652294,
+      "learning_rate": 1.4272218257971327e-06,
+      "loss": 0.8403,
+      "step": 4331
+    },
+    {
+      "epoch": 0.8333974605617546,
+      "grad_norm": 0.8486424527110809,
+      "learning_rate": 1.4240155134196499e-06,
+      "loss": 0.795,
+      "step": 4332
+    },
+    {
+      "epoch": 0.833589842247018,
+      "grad_norm": 0.9337572083033404,
+      "learning_rate": 1.4208125305346232e-06,
+      "loss": 0.7872,
+      "step": 4333
+    },
+    {
+      "epoch": 0.8337822239322816,
+      "grad_norm": 0.8934583006601048,
+      "learning_rate": 1.4176128783855636e-06,
+      "loss": 0.7983,
+      "step": 4334
+    },
+    {
+      "epoch": 0.8339746056175452,
+      "grad_norm": 0.9198382632841325,
+      "learning_rate": 1.4144165582146819e-06,
+      "loss": 0.7906,
+      "step": 4335
+    },
+    {
+      "epoch": 0.8341669873028088,
+      "grad_norm": 0.858103704027899,
+      "learning_rate": 1.4112235712629063e-06,
+      "loss": 0.8265,
+      "step": 4336
+    },
+    {
+      "epoch": 0.8343593689880723,
+      "grad_norm": 0.9847096950407326,
+      "learning_rate": 1.40803391876986e-06,
+      "loss": 0.8049,
+      "step": 4337
+    },
+    {
+      "epoch": 0.8345517506733359,
+      "grad_norm": 0.8800854577330405,
+      "learning_rate": 1.4048476019738756e-06,
+      "loss": 0.7616,
+      "step": 4338
+    },
+    {
+      "epoch": 0.8347441323585995,
+      "grad_norm": 0.901784783697436,
+      "learning_rate": 1.4016646221119912e-06,
+      "loss": 0.7431,
+      "step": 4339
+    },
+    {
+      "epoch": 0.834936514043863,
+      "grad_norm": 0.8600285831471971,
+      "learning_rate": 1.3984849804199485e-06,
+      "loss": 0.8008,
+      "step": 4340
+    },
+    {
+      "epoch": 0.8351288957291266,
+      "grad_norm": 0.907650429491348,
+      "learning_rate": 1.395308678132199e-06,
+      "loss": 0.7676,
+      "step": 4341
+    },
+    {
+      "epoch": 0.8353212774143901,
+      "grad_norm": 0.826593300491324,
+      "learning_rate": 1.392135716481885e-06,
+      "loss": 0.7619,
+      "step": 4342
+    },
+    {
+      "epoch": 0.8355136590996537,
+      "grad_norm": 0.8336596922452981,
+      "learning_rate": 1.3889660967008656e-06,
+      "loss": 0.7758,
+      "step": 4343
+    },
+    {
+      "epoch": 0.8357060407849173,
+      "grad_norm": 0.8725593160172155,
+      "learning_rate": 1.3857998200196943e-06,
+      "loss": 0.7845,
+      "step": 4344
+    },
+    {
+      "epoch": 0.8358984224701809,
+      "grad_norm": 0.8462447872033739,
+      "learning_rate": 1.3826368876676278e-06,
+      "loss": 0.8222,
+      "step": 4345
+    },
+    {
+      "epoch": 0.8360908041554445,
+      "grad_norm": 0.8891923140054204,
+      "learning_rate": 1.379477300872626e-06,
+      "loss": 0.7871,
+      "step": 4346
+    },
+    {
+      "epoch": 0.8362831858407079,
+      "grad_norm": 0.838765956254086,
+      "learning_rate": 1.3763210608613497e-06,
+      "loss": 0.8102,
+      "step": 4347
+    },
+    {
+      "epoch": 0.8364755675259715,
+      "grad_norm": 0.8905211979797233,
+      "learning_rate": 1.3731681688591593e-06,
+      "loss": 0.762,
+      "step": 4348
+    },
+    {
+      "epoch": 0.8366679492112351,
+      "grad_norm": 0.969736553586293,
+      "learning_rate": 1.370018626090116e-06,
+      "loss": 0.8457,
+      "step": 4349
+    },
+    {
+      "epoch": 0.8368603308964987,
+      "grad_norm": 0.9201057079586143,
+      "learning_rate": 1.3668724337769823e-06,
+      "loss": 0.8194,
+      "step": 4350
+    },
+    {
+      "epoch": 0.8370527125817622,
+      "grad_norm": 0.8519578206191057,
+      "learning_rate": 1.3637295931412153e-06,
+      "loss": 0.8223,
+      "step": 4351
+    },
+    {
+      "epoch": 0.8372450942670258,
+      "grad_norm": 0.830444520133561,
+      "learning_rate": 1.3605901054029746e-06,
+      "loss": 0.788,
+      "step": 4352
+    },
+    {
+      "epoch": 0.8374374759522893,
+      "grad_norm": 0.8468434561720264,
+      "learning_rate": 1.3574539717811231e-06,
+      "loss": 0.8106,
+      "step": 4353
+    },
+    {
+      "epoch": 0.8376298576375529,
+      "grad_norm": 0.9273697408004078,
+      "learning_rate": 1.3543211934932065e-06,
+      "loss": 0.7847,
+      "step": 4354
+    },
+    {
+      "epoch": 0.8378222393228165,
+      "grad_norm": 0.908488131405535,
+      "learning_rate": 1.3511917717554846e-06,
+      "loss": 0.7629,
+      "step": 4355
+    },
+    {
+      "epoch": 0.83801462100808,
+      "grad_norm": 0.82224677193205,
+      "learning_rate": 1.348065707782904e-06,
+      "loss": 0.7885,
+      "step": 4356
+    },
+    {
+      "epoch": 0.8382070026933436,
+      "grad_norm": 0.9034074089455089,
+      "learning_rate": 1.3449430027891096e-06,
+      "loss": 0.864,
+      "step": 4357
+    },
+    {
+      "epoch": 0.8383993843786072,
+      "grad_norm": 0.9230000325390251,
+      "learning_rate": 1.3418236579864452e-06,
+      "loss": 0.8016,
+      "step": 4358
+    },
+    {
+      "epoch": 0.8385917660638708,
+      "grad_norm": 0.9427423575501956,
+      "learning_rate": 1.338707674585945e-06,
+      "loss": 0.8333,
+      "step": 4359
+    },
+    {
+      "epoch": 0.8387841477491342,
+      "grad_norm": 0.9234938283140484,
+      "learning_rate": 1.3355950537973438e-06,
+      "loss": 0.8425,
+      "step": 4360
+    },
+    {
+      "epoch": 0.8389765294343978,
+      "grad_norm": 0.5785069631289447,
+      "learning_rate": 1.332485796829065e-06,
+      "loss": 0.3208,
+      "step": 4361
+    },
+    {
+      "epoch": 0.8391689111196614,
+      "grad_norm": 0.8371762387862759,
+      "learning_rate": 1.329379904888235e-06,
+      "loss": 0.811,
+      "step": 4362
+    },
+    {
+      "epoch": 0.839361292804925,
+      "grad_norm": 0.8645542960635137,
+      "learning_rate": 1.3262773791806617e-06,
+      "loss": 0.7741,
+      "step": 4363
+    },
+    {
+      "epoch": 0.8395536744901886,
+      "grad_norm": 0.8878966867663053,
+      "learning_rate": 1.3231782209108546e-06,
+      "loss": 0.8112,
+      "step": 4364
+    },
+    {
+      "epoch": 0.8397460561754521,
+      "grad_norm": 0.8968689756531806,
+      "learning_rate": 1.3200824312820137e-06,
+      "loss": 0.8152,
+      "step": 4365
+    },
+    {
+      "epoch": 0.8399384378607156,
+      "grad_norm": 0.9111872817785706,
+      "learning_rate": 1.3169900114960298e-06,
+      "loss": 0.877,
+      "step": 4366
+    },
+    {
+      "epoch": 0.8401308195459792,
+      "grad_norm": 0.7822280251944174,
+      "learning_rate": 1.3139009627534927e-06,
+      "loss": 0.7952,
+      "step": 4367
+    },
+    {
+      "epoch": 0.8403232012312428,
+      "grad_norm": 0.934760142726033,
+      "learning_rate": 1.3108152862536683e-06,
+      "loss": 0.779,
+      "step": 4368
+    },
+    {
+      "epoch": 0.8405155829165063,
+      "grad_norm": 1.0082636070474476,
+      "learning_rate": 1.3077329831945295e-06,
+      "loss": 0.7647,
+      "step": 4369
+    },
+    {
+      "epoch": 0.8407079646017699,
+      "grad_norm": 0.9369225627931361,
+      "learning_rate": 1.3046540547727305e-06,
+      "loss": 0.8602,
+      "step": 4370
+    },
+    {
+      "epoch": 0.8409003462870335,
+      "grad_norm": 0.8581913368449883,
+      "learning_rate": 1.3015785021836159e-06,
+      "loss": 0.7869,
+      "step": 4371
+    },
+    {
+      "epoch": 0.8410927279722971,
+      "grad_norm": 0.9045760803046817,
+      "learning_rate": 1.2985063266212229e-06,
+      "loss": 0.7568,
+      "step": 4372
+    },
+    {
+      "epoch": 0.8412851096575606,
+      "grad_norm": 0.5812201721232682,
+      "learning_rate": 1.295437529278275e-06,
+      "loss": 0.3383,
+      "step": 4373
+    },
+    {
+      "epoch": 0.8414774913428241,
+      "grad_norm": 0.8382412082500551,
+      "learning_rate": 1.2923721113461852e-06,
+      "loss": 0.7677,
+      "step": 4374
+    },
+    {
+      "epoch": 0.8416698730280877,
+      "grad_norm": 0.8600313667375051,
+      "learning_rate": 1.2893100740150522e-06,
+      "loss": 0.7547,
+      "step": 4375
+    },
+    {
+      "epoch": 0.8418622547133513,
+      "grad_norm": 0.940482214155153,
+      "learning_rate": 1.2862514184736695e-06,
+      "loss": 0.7738,
+      "step": 4376
+    },
+    {
+      "epoch": 0.8420546363986149,
+      "grad_norm": 0.9261369887990596,
+      "learning_rate": 1.2831961459095088e-06,
+      "loss": 0.6931,
+      "step": 4377
+    },
+    {
+      "epoch": 0.8422470180838784,
+      "grad_norm": 0.9128374404171035,
+      "learning_rate": 1.2801442575087296e-06,
+      "loss": 0.8054,
+      "step": 4378
+    },
+    {
+      "epoch": 0.842439399769142,
+      "grad_norm": 0.8238097891127896,
+      "learning_rate": 1.2770957544561868e-06,
+      "loss": 0.7868,
+      "step": 4379
+    },
+    {
+      "epoch": 0.8426317814544055,
+      "grad_norm": 0.9041283098027844,
+      "learning_rate": 1.274050637935408e-06,
+      "loss": 0.8564,
+      "step": 4380
+    },
+    {
+      "epoch": 0.8428241631396691,
+      "grad_norm": 0.9370263973122149,
+      "learning_rate": 1.2710089091286148e-06,
+      "loss": 0.7669,
+      "step": 4381
+    },
+    {
+      "epoch": 0.8430165448249327,
+      "grad_norm": 0.8372705698001814,
+      "learning_rate": 1.2679705692167122e-06,
+      "loss": 0.7793,
+      "step": 4382
+    },
+    {
+      "epoch": 0.8432089265101962,
+      "grad_norm": 0.856313358985364,
+      "learning_rate": 1.2649356193792873e-06,
+      "loss": 0.7596,
+      "step": 4383
+    },
+    {
+      "epoch": 0.8434013081954598,
+      "grad_norm": 0.9665943559019,
+      "learning_rate": 1.261904060794612e-06,
+      "loss": 0.8349,
+      "step": 4384
+    },
+    {
+      "epoch": 0.8435936898807234,
+      "grad_norm": 0.9389552347341181,
+      "learning_rate": 1.2588758946396417e-06,
+      "loss": 0.8279,
+      "step": 4385
+    },
+    {
+      "epoch": 0.843786071565987,
+      "grad_norm": 0.9283615380948241,
+      "learning_rate": 1.2558511220900138e-06,
+      "loss": 0.7735,
+      "step": 4386
+    },
+    {
+      "epoch": 0.8439784532512504,
+      "grad_norm": 0.6181029890981897,
+      "learning_rate": 1.2528297443200489e-06,
+      "loss": 0.3077,
+      "step": 4387
+    },
+    {
+      "epoch": 0.844170834936514,
+      "grad_norm": 0.9150258177167815,
+      "learning_rate": 1.2498117625027562e-06,
+      "loss": 0.7769,
+      "step": 4388
+    },
+    {
+      "epoch": 0.8443632166217776,
+      "grad_norm": 0.8951261891064789,
+      "learning_rate": 1.246797177809812e-06,
+      "loss": 0.7747,
+      "step": 4389
+    },
+    {
+      "epoch": 0.8445555983070412,
+      "grad_norm": 0.8743116929955407,
+      "learning_rate": 1.2437859914115847e-06,
+      "loss": 0.7831,
+      "step": 4390
+    },
+    {
+      "epoch": 0.8447479799923048,
+      "grad_norm": 0.8298840292450792,
+      "learning_rate": 1.2407782044771222e-06,
+      "loss": 0.7214,
+      "step": 4391
+    },
+    {
+      "epoch": 0.8449403616775683,
+      "grad_norm": 0.8882144181710543,
+      "learning_rate": 1.237773818174146e-06,
+      "loss": 0.7596,
+      "step": 4392
+    },
+    {
+      "epoch": 0.8451327433628318,
+      "grad_norm": 0.8672125268860936,
+      "learning_rate": 1.23477283366907e-06,
+      "loss": 0.6953,
+      "step": 4393
+    },
+    {
+      "epoch": 0.8453251250480954,
+      "grad_norm": 0.9268890671690084,
+      "learning_rate": 1.2317752521269722e-06,
+      "loss": 0.7955,
+      "step": 4394
+    },
+    {
+      "epoch": 0.845517506733359,
+      "grad_norm": 0.9388811360550405,
+      "learning_rate": 1.2287810747116224e-06,
+      "loss": 0.6816,
+      "step": 4395
+    },
+    {
+      "epoch": 0.8457098884186226,
+      "grad_norm": 0.8880815467048738,
+      "learning_rate": 1.225790302585461e-06,
+      "loss": 0.8159,
+      "step": 4396
+    },
+    {
+      "epoch": 0.8459022701038861,
+      "grad_norm": 0.9263965440892908,
+      "learning_rate": 1.2228029369096094e-06,
+      "loss": 0.7934,
+      "step": 4397
+    },
+    {
+      "epoch": 0.8460946517891497,
+      "grad_norm": 0.7926879277337392,
+      "learning_rate": 1.2198189788438652e-06,
+      "loss": 0.7016,
+      "step": 4398
+    },
+    {
+      "epoch": 0.8462870334744133,
+      "grad_norm": 0.8631846170201514,
+      "learning_rate": 1.216838429546704e-06,
+      "loss": 0.798,
+      "step": 4399
+    },
+    {
+      "epoch": 0.8464794151596768,
+      "grad_norm": 0.6269409303178556,
+      "learning_rate": 1.2138612901752777e-06,
+      "loss": 0.3057,
+      "step": 4400
+    },
+    {
+      "epoch": 0.8466717968449403,
+      "grad_norm": 0.8268329659473057,
+      "learning_rate": 1.2108875618854122e-06,
+      "loss": 0.7574,
+      "step": 4401
+    },
+    {
+      "epoch": 0.8468641785302039,
+      "grad_norm": 0.9328435535663747,
+      "learning_rate": 1.2079172458316168e-06,
+      "loss": 0.8154,
+      "step": 4402
+    },
+    {
+      "epoch": 0.8470565602154675,
+      "grad_norm": 0.9168345818061228,
+      "learning_rate": 1.204950343167065e-06,
+      "loss": 0.8352,
+      "step": 4403
+    },
+    {
+      "epoch": 0.8472489419007311,
+      "grad_norm": 0.9355966371274494,
+      "learning_rate": 1.2019868550436099e-06,
+      "loss": 0.8239,
+      "step": 4404
+    },
+    {
+      "epoch": 0.8474413235859947,
+      "grad_norm": 0.8464663050318761,
+      "learning_rate": 1.1990267826117874e-06,
+      "loss": 0.7586,
+      "step": 4405
+    },
+    {
+      "epoch": 0.8476337052712581,
+      "grad_norm": 0.9330870617835246,
+      "learning_rate": 1.1960701270207885e-06,
+      "loss": 0.7719,
+      "step": 4406
+    },
+    {
+      "epoch": 0.8478260869565217,
+      "grad_norm": 0.8896999617518034,
+      "learning_rate": 1.1931168894184974e-06,
+      "loss": 0.801,
+      "step": 4407
+    },
+    {
+      "epoch": 0.8480184686417853,
+      "grad_norm": 0.9171127408696418,
+      "learning_rate": 1.19016707095146e-06,
+      "loss": 0.7798,
+      "step": 4408
+    },
+    {
+      "epoch": 0.8482108503270489,
+      "grad_norm": 0.7920934830646624,
+      "learning_rate": 1.187220672764897e-06,
+      "loss": 0.7836,
+      "step": 4409
+    },
+    {
+      "epoch": 0.8484032320123124,
+      "grad_norm": 0.8887310126043334,
+      "learning_rate": 1.1842776960027014e-06,
+      "loss": 0.8056,
+      "step": 4410
+    },
+    {
+      "epoch": 0.848595613697576,
+      "grad_norm": 0.6106220717083156,
+      "learning_rate": 1.1813381418074388e-06,
+      "loss": 0.3507,
+      "step": 4411
+    },
+    {
+      "epoch": 0.8487879953828396,
+      "grad_norm": 0.8694532925471362,
+      "learning_rate": 1.1784020113203453e-06,
+      "loss": 0.7193,
+      "step": 4412
+    },
+    {
+      "epoch": 0.8489803770681031,
+      "grad_norm": 0.8543898449733707,
+      "learning_rate": 1.1754693056813272e-06,
+      "loss": 0.7665,
+      "step": 4413
+    },
+    {
+      "epoch": 0.8491727587533667,
+      "grad_norm": 0.8223623183004528,
+      "learning_rate": 1.172540026028962e-06,
+      "loss": 0.7126,
+      "step": 4414
+    },
+    {
+      "epoch": 0.8493651404386302,
+      "grad_norm": 0.9145340042490969,
+      "learning_rate": 1.169614173500494e-06,
+      "loss": 0.7518,
+      "step": 4415
+    },
+    {
+      "epoch": 0.8495575221238938,
+      "grad_norm": 0.8632468730889142,
+      "learning_rate": 1.1666917492318486e-06,
+      "loss": 0.7799,
+      "step": 4416
+    },
+    {
+      "epoch": 0.8497499038091574,
+      "grad_norm": 0.8643093994592866,
+      "learning_rate": 1.1637727543576027e-06,
+      "loss": 0.7803,
+      "step": 4417
+    },
+    {
+      "epoch": 0.849942285494421,
+      "grad_norm": 0.897635264506733,
+      "learning_rate": 1.1608571900110122e-06,
+      "loss": 0.74,
+      "step": 4418
+    },
+    {
+      "epoch": 0.8501346671796844,
+      "grad_norm": 0.9093094389357693,
+      "learning_rate": 1.1579450573240058e-06,
+      "loss": 0.8155,
+      "step": 4419
+    },
+    {
+      "epoch": 0.850327048864948,
+      "grad_norm": 0.8740465383215266,
+      "learning_rate": 1.1550363574271638e-06,
+      "loss": 0.8182,
+      "step": 4420
+    },
+    {
+      "epoch": 0.8505194305502116,
+      "grad_norm": 0.9404886665598711,
+      "learning_rate": 1.1521310914497518e-06,
+      "loss": 0.7633,
+      "step": 4421
+    },
+    {
+      "epoch": 0.8507118122354752,
+      "grad_norm": 0.8984693163308987,
+      "learning_rate": 1.149229260519691e-06,
+      "loss": 0.8488,
+      "step": 4422
+    },
+    {
+      "epoch": 0.8509041939207388,
+      "grad_norm": 0.8779787320864925,
+      "learning_rate": 1.1463308657635718e-06,
+      "loss": 0.8213,
+      "step": 4423
+    },
+    {
+      "epoch": 0.8510965756060023,
+      "grad_norm": 0.9114679121494595,
+      "learning_rate": 1.1434359083066515e-06,
+      "loss": 0.8224,
+      "step": 4424
+    },
+    {
+      "epoch": 0.8512889572912659,
+      "grad_norm": 0.9349701207108142,
+      "learning_rate": 1.140544389272853e-06,
+      "loss": 0.735,
+      "step": 4425
+    },
+    {
+      "epoch": 0.8514813389765294,
+      "grad_norm": 0.8680820031137845,
+      "learning_rate": 1.1376563097847616e-06,
+      "loss": 0.7581,
+      "step": 4426
+    },
+    {
+      "epoch": 0.851673720661793,
+      "grad_norm": 0.8657558398836773,
+      "learning_rate": 1.1347716709636282e-06,
+      "loss": 0.7488,
+      "step": 4427
+    },
+    {
+      "epoch": 0.8518661023470565,
+      "grad_norm": 0.925103875553607,
+      "learning_rate": 1.1318904739293745e-06,
+      "loss": 0.7893,
+      "step": 4428
+    },
+    {
+      "epoch": 0.8520584840323201,
+      "grad_norm": 0.8650012891998499,
+      "learning_rate": 1.129012719800575e-06,
+      "loss": 0.7511,
+      "step": 4429
+    },
+    {
+      "epoch": 0.8522508657175837,
+      "grad_norm": 0.9607193377991963,
+      "learning_rate": 1.1261384096944728e-06,
+      "loss": 0.7643,
+      "step": 4430
+    },
+    {
+      "epoch": 0.8524432474028473,
+      "grad_norm": 0.8296295830541922,
+      "learning_rate": 1.1232675447269803e-06,
+      "loss": 0.7643,
+      "step": 4431
+    },
+    {
+      "epoch": 0.8526356290881109,
+      "grad_norm": 0.9796128742639673,
+      "learning_rate": 1.1204001260126574e-06,
+      "loss": 0.7679,
+      "step": 4432
+    },
+    {
+      "epoch": 0.8528280107733743,
+      "grad_norm": 0.8365956706381201,
+      "learning_rate": 1.1175361546647413e-06,
+      "loss": 0.7802,
+      "step": 4433
+    },
+    {
+      "epoch": 0.8530203924586379,
+      "grad_norm": 0.9467417697272875,
+      "learning_rate": 1.1146756317951224e-06,
+      "loss": 0.8119,
+      "step": 4434
+    },
+    {
+      "epoch": 0.8532127741439015,
+      "grad_norm": 0.9294341087302771,
+      "learning_rate": 1.1118185585143536e-06,
+      "loss": 0.8047,
+      "step": 4435
+    },
+    {
+      "epoch": 0.8534051558291651,
+      "grad_norm": 0.7924180973722184,
+      "learning_rate": 1.1089649359316501e-06,
+      "loss": 0.8018,
+      "step": 4436
+    },
+    {
+      "epoch": 0.8535975375144286,
+      "grad_norm": 0.9371793764558272,
+      "learning_rate": 1.1061147651548855e-06,
+      "loss": 0.764,
+      "step": 4437
+    },
+    {
+      "epoch": 0.8537899191996922,
+      "grad_norm": 1.0000262537042508,
+      "learning_rate": 1.1032680472905932e-06,
+      "loss": 0.7964,
+      "step": 4438
+    },
+    {
+      "epoch": 0.8539823008849557,
+      "grad_norm": 0.8909608003360948,
+      "learning_rate": 1.1004247834439697e-06,
+      "loss": 0.7948,
+      "step": 4439
+    },
+    {
+      "epoch": 0.8541746825702193,
+      "grad_norm": 0.8903095382095011,
+      "learning_rate": 1.097584974718866e-06,
+      "loss": 0.787,
+      "step": 4440
+    },
+    {
+      "epoch": 0.8543670642554829,
+      "grad_norm": 0.8637770200727334,
+      "learning_rate": 1.0947486222177928e-06,
+      "loss": 0.7426,
+      "step": 4441
+    },
+    {
+      "epoch": 0.8545594459407464,
+      "grad_norm": 0.8122002422513042,
+      "learning_rate": 1.0919157270419257e-06,
+      "loss": 0.7951,
+      "step": 4442
+    },
+    {
+      "epoch": 0.85475182762601,
+      "grad_norm": 0.8956135241788684,
+      "learning_rate": 1.0890862902910849e-06,
+      "loss": 0.7747,
+      "step": 4443
+    },
+    {
+      "epoch": 0.8549442093112736,
+      "grad_norm": 0.8927852919530282,
+      "learning_rate": 1.0862603130637562e-06,
+      "loss": 0.7387,
+      "step": 4444
+    },
+    {
+      "epoch": 0.8551365909965372,
+      "grad_norm": 0.890104053077436,
+      "learning_rate": 1.0834377964570863e-06,
+      "loss": 0.7735,
+      "step": 4445
+    },
+    {
+      "epoch": 0.8553289726818007,
+      "grad_norm": 0.8458243598230544,
+      "learning_rate": 1.0806187415668668e-06,
+      "loss": 0.7529,
+      "step": 4446
+    },
+    {
+      "epoch": 0.8555213543670642,
+      "grad_norm": 0.8983575332889089,
+      "learning_rate": 1.0778031494875574e-06,
+      "loss": 0.7418,
+      "step": 4447
+    },
+    {
+      "epoch": 0.8557137360523278,
+      "grad_norm": 0.8770788606455594,
+      "learning_rate": 1.0749910213122649e-06,
+      "loss": 0.7417,
+      "step": 4448
+    },
+    {
+      "epoch": 0.8559061177375914,
+      "grad_norm": 0.9038520451385039,
+      "learning_rate": 1.072182358132755e-06,
+      "loss": 0.7712,
+      "step": 4449
+    },
+    {
+      "epoch": 0.856098499422855,
+      "grad_norm": 0.9965810799904544,
+      "learning_rate": 1.0693771610394477e-06,
+      "loss": 0.8182,
+      "step": 4450
+    },
+    {
+      "epoch": 0.8562908811081185,
+      "grad_norm": 0.9179961661295949,
+      "learning_rate": 1.066575431121417e-06,
+      "loss": 0.803,
+      "step": 4451
+    },
+    {
+      "epoch": 0.856483262793382,
+      "grad_norm": 0.8738473536513838,
+      "learning_rate": 1.06377716946639e-06,
+      "loss": 0.7568,
+      "step": 4452
+    },
+    {
+      "epoch": 0.8566756444786456,
+      "grad_norm": 0.8416783683479411,
+      "learning_rate": 1.0609823771607487e-06,
+      "loss": 0.8074,
+      "step": 4453
+    },
+    {
+      "epoch": 0.8568680261639092,
+      "grad_norm": 0.8621970470829756,
+      "learning_rate": 1.0581910552895302e-06,
+      "loss": 0.816,
+      "step": 4454
+    },
+    {
+      "epoch": 0.8570604078491728,
+      "grad_norm": 0.9177079705035212,
+      "learning_rate": 1.055403204936416e-06,
+      "loss": 0.7902,
+      "step": 4455
+    },
+    {
+      "epoch": 0.8572527895344363,
+      "grad_norm": 0.8902617724865826,
+      "learning_rate": 1.0526188271837512e-06,
+      "loss": 0.8223,
+      "step": 4456
+    },
+    {
+      "epoch": 0.8574451712196999,
+      "grad_norm": 0.8884770302910331,
+      "learning_rate": 1.0498379231125278e-06,
+      "loss": 0.818,
+      "step": 4457
+    },
+    {
+      "epoch": 0.8576375529049635,
+      "grad_norm": 0.8772061747925244,
+      "learning_rate": 1.047060493802381e-06,
+      "loss": 0.7265,
+      "step": 4458
+    },
+    {
+      "epoch": 0.857829934590227,
+      "grad_norm": 0.6508103940292379,
+      "learning_rate": 1.0442865403316117e-06,
+      "loss": 0.3685,
+      "step": 4459
+    },
+    {
+      "epoch": 0.8580223162754905,
+      "grad_norm": 0.8821974268287692,
+      "learning_rate": 1.0415160637771604e-06,
+      "loss": 0.8624,
+      "step": 4460
+    },
+    {
+      "epoch": 0.8582146979607541,
+      "grad_norm": 0.9005931996831463,
+      "learning_rate": 1.0387490652146236e-06,
+      "loss": 0.7397,
+      "step": 4461
+    },
+    {
+      "epoch": 0.8584070796460177,
+      "grad_norm": 0.870724298995309,
+      "learning_rate": 1.0359855457182455e-06,
+      "loss": 0.7885,
+      "step": 4462
+    },
+    {
+      "epoch": 0.8585994613312813,
+      "grad_norm": 0.8798984258361118,
+      "learning_rate": 1.0332255063609177e-06,
+      "loss": 0.8167,
+      "step": 4463
+    },
+    {
+      "epoch": 0.8587918430165449,
+      "grad_norm": 0.8933963114095317,
+      "learning_rate": 1.0304689482141839e-06,
+      "loss": 0.779,
+      "step": 4464
+    },
+    {
+      "epoch": 0.8589842247018084,
+      "grad_norm": 0.8522662581197861,
+      "learning_rate": 1.027715872348234e-06,
+      "loss": 0.7683,
+      "step": 4465
+    },
+    {
+      "epoch": 0.8591766063870719,
+      "grad_norm": 0.964429668582689,
+      "learning_rate": 1.0249662798319072e-06,
+      "loss": 0.7621,
+      "step": 4466
+    },
+    {
+      "epoch": 0.8593689880723355,
+      "grad_norm": 0.8914688910367864,
+      "learning_rate": 1.0222201717326885e-06,
+      "loss": 0.7747,
+      "step": 4467
+    },
+    {
+      "epoch": 0.8595613697575991,
+      "grad_norm": 0.8903592930914297,
+      "learning_rate": 1.0194775491167164e-06,
+      "loss": 0.8157,
+      "step": 4468
+    },
+    {
+      "epoch": 0.8597537514428626,
+      "grad_norm": 0.9797019069223979,
+      "learning_rate": 1.0167384130487667e-06,
+      "loss": 0.7442,
+      "step": 4469
+    },
+    {
+      "epoch": 0.8599461331281262,
+      "grad_norm": 0.9642789029119208,
+      "learning_rate": 1.0140027645922656e-06,
+      "loss": 0.7921,
+      "step": 4470
+    },
+    {
+      "epoch": 0.8601385148133898,
+      "grad_norm": 0.8776784237827279,
+      "learning_rate": 1.0112706048092924e-06,
+      "loss": 0.7939,
+      "step": 4471
+    },
+    {
+      "epoch": 0.8603308964986534,
+      "grad_norm": 0.8531048978869771,
+      "learning_rate": 1.0085419347605575e-06,
+      "loss": 0.7955,
+      "step": 4472
+    },
+    {
+      "epoch": 0.8605232781839169,
+      "grad_norm": 0.9003177523982999,
+      "learning_rate": 1.00581675550543e-06,
+      "loss": 0.7294,
+      "step": 4473
+    },
+    {
+      "epoch": 0.8607156598691804,
+      "grad_norm": 0.8919042887949754,
+      "learning_rate": 1.003095068101917e-06,
+      "loss": 0.7335,
+      "step": 4474
+    },
+    {
+      "epoch": 0.860908041554444,
+      "grad_norm": 0.901304063804905,
+      "learning_rate": 1.0003768736066722e-06,
+      "loss": 0.8099,
+      "step": 4475
+    },
+    {
+      "epoch": 0.8611004232397076,
+      "grad_norm": 0.962976033053853,
+      "learning_rate": 9.976621730749892e-07,
+      "loss": 0.7771,
+      "step": 4476
+    },
+    {
+      "epoch": 0.8612928049249712,
+      "grad_norm": 0.8700289262456535,
+      "learning_rate": 9.949509675608115e-07,
+      "loss": 0.822,
+      "step": 4477
+    },
+    {
+      "epoch": 0.8614851866102347,
+      "grad_norm": 0.8679857051503158,
+      "learning_rate": 9.922432581167207e-07,
+      "loss": 0.795,
+      "step": 4478
+    },
+    {
+      "epoch": 0.8616775682954982,
+      "grad_norm": 0.85953122779688,
+      "learning_rate": 9.895390457939414e-07,
+      "loss": 0.778,
+      "step": 4479
+    },
+    {
+      "epoch": 0.8618699499807618,
+      "grad_norm": 0.9875503703559247,
+      "learning_rate": 9.86838331642348e-07,
+      "loss": 0.7815,
+      "step": 4480
+    },
+    {
+      "epoch": 0.8620623316660254,
+      "grad_norm": 0.8677427995850563,
+      "learning_rate": 9.84141116710442e-07,
+      "loss": 0.6795,
+      "step": 4481
+    },
+    {
+      "epoch": 0.862254713351289,
+      "grad_norm": 0.879883295985596,
+      "learning_rate": 9.814474020453824e-07,
+      "loss": 0.8148,
+      "step": 4482
+    },
+    {
+      "epoch": 0.8624470950365525,
+      "grad_norm": 0.8689572689197567,
+      "learning_rate": 9.787571886929604e-07,
+      "loss": 0.7564,
+      "step": 4483
+    },
+    {
+      "epoch": 0.8626394767218161,
+      "grad_norm": 0.8826907667805138,
+      "learning_rate": 9.76070477697605e-07,
+      "loss": 0.7913,
+      "step": 4484
+    },
+    {
+      "epoch": 0.8628318584070797,
+      "grad_norm": 0.9246752106935797,
+      "learning_rate": 9.733872701023938e-07,
+      "loss": 0.8083,
+      "step": 4485
+    },
+    {
+      "epoch": 0.8630242400923432,
+      "grad_norm": 0.8204625883246647,
+      "learning_rate": 9.707075669490407e-07,
+      "loss": 0.7479,
+      "step": 4486
+    },
+    {
+      "epoch": 0.8632166217776067,
+      "grad_norm": 0.9185415200717927,
+      "learning_rate": 9.680313692778976e-07,
+      "loss": 0.7258,
+      "step": 4487
+    },
+    {
+      "epoch": 0.8634090034628703,
+      "grad_norm": 0.9299113006962008,
+      "learning_rate": 9.653586781279567e-07,
+      "loss": 0.7576,
+      "step": 4488
+    },
+    {
+      "epoch": 0.8636013851481339,
+      "grad_norm": 0.8983083610292228,
+      "learning_rate": 9.626894945368492e-07,
+      "loss": 0.7994,
+      "step": 4489
+    },
+    {
+      "epoch": 0.8637937668333975,
+      "grad_norm": 0.9161459127495426,
+      "learning_rate": 9.600238195408428e-07,
+      "loss": 0.7811,
+      "step": 4490
+    },
+    {
+      "epoch": 0.8639861485186611,
+      "grad_norm": 0.8199628944285414,
+      "learning_rate": 9.573616541748464e-07,
+      "loss": 0.7361,
+      "step": 4491
+    },
+    {
+      "epoch": 0.8641785302039245,
+      "grad_norm": 0.9800547013009101,
+      "learning_rate": 9.547029994724023e-07,
+      "loss": 0.7469,
+      "step": 4492
+    },
+    {
+      "epoch": 0.8643709118891881,
+      "grad_norm": 0.8418587876354746,
+      "learning_rate": 9.520478564656898e-07,
+      "loss": 0.7824,
+      "step": 4493
+    },
+    {
+      "epoch": 0.8645632935744517,
+      "grad_norm": 0.8626127756456105,
+      "learning_rate": 9.49396226185535e-07,
+      "loss": 0.7953,
+      "step": 4494
+    },
+    {
+      "epoch": 0.8647556752597153,
+      "grad_norm": 0.9850483356220473,
+      "learning_rate": 9.467481096613829e-07,
+      "loss": 0.8297,
+      "step": 4495
+    },
+    {
+      "epoch": 0.8649480569449788,
+      "grad_norm": 0.858898737203966,
+      "learning_rate": 9.441035079213267e-07,
+      "loss": 0.7669,
+      "step": 4496
+    },
+    {
+      "epoch": 0.8651404386302424,
+      "grad_norm": 0.8422555209699688,
+      "learning_rate": 9.414624219920953e-07,
+      "loss": 0.7369,
+      "step": 4497
+    },
+    {
+      "epoch": 0.865332820315506,
+      "grad_norm": 0.8899037373495923,
+      "learning_rate": 9.38824852899043e-07,
+      "loss": 0.7898,
+      "step": 4498
+    },
+    {
+      "epoch": 0.8655252020007695,
+      "grad_norm": 0.5966999054885143,
+      "learning_rate": 9.361908016661703e-07,
+      "loss": 0.3159,
+      "step": 4499
+    },
+    {
+      "epoch": 0.8657175836860331,
+      "grad_norm": 0.8383860453396158,
+      "learning_rate": 9.335602693161039e-07,
+      "loss": 0.8051,
+      "step": 4500
+    },
+    {
+      "epoch": 0.8659099653712966,
+      "grad_norm": 0.8652000030338419,
+      "learning_rate": 9.309332568701079e-07,
+      "loss": 0.7783,
+      "step": 4501
+    },
+    {
+      "epoch": 0.8661023470565602,
+      "grad_norm": 0.8967412702033152,
+      "learning_rate": 9.283097653480788e-07,
+      "loss": 0.8523,
+      "step": 4502
+    },
+    {
+      "epoch": 0.8662947287418238,
+      "grad_norm": 0.8673663271753793,
+      "learning_rate": 9.256897957685463e-07,
+      "loss": 0.7716,
+      "step": 4503
+    },
+    {
+      "epoch": 0.8664871104270874,
+      "grad_norm": 0.8520524236794936,
+      "learning_rate": 9.230733491486721e-07,
+      "loss": 0.792,
+      "step": 4504
+    },
+    {
+      "epoch": 0.866679492112351,
+      "grad_norm": 0.832113852524758,
+      "learning_rate": 9.204604265042505e-07,
+      "loss": 0.7447,
+      "step": 4505
+    },
+    {
+      "epoch": 0.8668718737976144,
+      "grad_norm": 0.9418491988601584,
+      "learning_rate": 9.178510288497123e-07,
+      "loss": 0.7852,
+      "step": 4506
+    },
+    {
+      "epoch": 0.867064255482878,
+      "grad_norm": 0.9278960819905789,
+      "learning_rate": 9.15245157198108e-07,
+      "loss": 0.8036,
+      "step": 4507
+    },
+    {
+      "epoch": 0.8672566371681416,
+      "grad_norm": 0.8155267465574518,
+      "learning_rate": 9.126428125611342e-07,
+      "loss": 0.7285,
+      "step": 4508
+    },
+    {
+      "epoch": 0.8674490188534052,
+      "grad_norm": 0.9057795604961825,
+      "learning_rate": 9.10043995949108e-07,
+      "loss": 0.8251,
+      "step": 4509
+    },
+    {
+      "epoch": 0.8676414005386687,
+      "grad_norm": 0.8800172856296676,
+      "learning_rate": 9.074487083709759e-07,
+      "loss": 0.8338,
+      "step": 4510
+    },
+    {
+      "epoch": 0.8678337822239323,
+      "grad_norm": 0.944946732246939,
+      "learning_rate": 9.04856950834323e-07,
+      "loss": 0.8077,
+      "step": 4511
+    },
+    {
+      "epoch": 0.8680261639091958,
+      "grad_norm": 0.9182741194159008,
+      "learning_rate": 9.022687243453554e-07,
+      "loss": 0.7691,
+      "step": 4512
+    },
+    {
+      "epoch": 0.8682185455944594,
+      "grad_norm": 0.8789939729067757,
+      "learning_rate": 8.996840299089149e-07,
+      "loss": 0.7905,
+      "step": 4513
+    },
+    {
+      "epoch": 0.868410927279723,
+      "grad_norm": 0.840420315209614,
+      "learning_rate": 8.971028685284655e-07,
+      "loss": 0.7769,
+      "step": 4514
+    },
+    {
+      "epoch": 0.8686033089649865,
+      "grad_norm": 0.981615602292768,
+      "learning_rate": 8.945252412061056e-07,
+      "loss": 0.7913,
+      "step": 4515
+    },
+    {
+      "epoch": 0.8687956906502501,
+      "grad_norm": 0.9314004088782799,
+      "learning_rate": 8.91951148942557e-07,
+      "loss": 0.8055,
+      "step": 4516
+    },
+    {
+      "epoch": 0.8689880723355137,
+      "grad_norm": 0.9517091153771644,
+      "learning_rate": 8.893805927371724e-07,
+      "loss": 0.8294,
+      "step": 4517
+    },
+    {
+      "epoch": 0.8691804540207773,
+      "grad_norm": 0.9037940444825492,
+      "learning_rate": 8.868135735879291e-07,
+      "loss": 0.8452,
+      "step": 4518
+    },
+    {
+      "epoch": 0.8693728357060407,
+      "grad_norm": 0.8970846111804724,
+      "learning_rate": 8.842500924914299e-07,
+      "loss": 0.7672,
+      "step": 4519
+    },
+    {
+      "epoch": 0.8695652173913043,
+      "grad_norm": 0.8559055722490901,
+      "learning_rate": 8.816901504429143e-07,
+      "loss": 0.7829,
+      "step": 4520
+    },
+    {
+      "epoch": 0.8697575990765679,
+      "grad_norm": 0.839427327437618,
+      "learning_rate": 8.791337484362305e-07,
+      "loss": 0.7489,
+      "step": 4521
+    },
+    {
+      "epoch": 0.8699499807618315,
+      "grad_norm": 0.8072190609538022,
+      "learning_rate": 8.765808874638682e-07,
+      "loss": 0.7948,
+      "step": 4522
+    },
+    {
+      "epoch": 0.8701423624470951,
+      "grad_norm": 0.9195942116499404,
+      "learning_rate": 8.740315685169364e-07,
+      "loss": 0.7507,
+      "step": 4523
+    },
+    {
+      "epoch": 0.8703347441323586,
+      "grad_norm": 0.9030341822855418,
+      "learning_rate": 8.714857925851617e-07,
+      "loss": 0.8448,
+      "step": 4524
+    },
+    {
+      "epoch": 0.8705271258176221,
+      "grad_norm": 0.8279895449462823,
+      "learning_rate": 8.689435606569086e-07,
+      "loss": 0.8015,
+      "step": 4525
+    },
+    {
+      "epoch": 0.8707195075028857,
+      "grad_norm": 0.870024110060597,
+      "learning_rate": 8.664048737191566e-07,
+      "loss": 0.8258,
+      "step": 4526
+    },
+    {
+      "epoch": 0.8709118891881493,
+      "grad_norm": 0.8784588121734871,
+      "learning_rate": 8.638697327575108e-07,
+      "loss": 0.8229,
+      "step": 4527
+    },
+    {
+      "epoch": 0.8711042708734128,
+      "grad_norm": 0.8706433921771627,
+      "learning_rate": 8.613381387562015e-07,
+      "loss": 0.7777,
+      "step": 4528
+    },
+    {
+      "epoch": 0.8712966525586764,
+      "grad_norm": 0.9564067873105776,
+      "learning_rate": 8.588100926980802e-07,
+      "loss": 0.8146,
+      "step": 4529
+    },
+    {
+      "epoch": 0.87148903424394,
+      "grad_norm": 0.8203972255091385,
+      "learning_rate": 8.56285595564621e-07,
+      "loss": 0.7009,
+      "step": 4530
+    },
+    {
+      "epoch": 0.8716814159292036,
+      "grad_norm": 0.8613353132364849,
+      "learning_rate": 8.537646483359185e-07,
+      "loss": 0.8304,
+      "step": 4531
+    },
+    {
+      "epoch": 0.8718737976144671,
+      "grad_norm": 0.9822588331562716,
+      "learning_rate": 8.512472519906978e-07,
+      "loss": 0.8358,
+      "step": 4532
+    },
+    {
+      "epoch": 0.8720661792997306,
+      "grad_norm": 0.8144120460275345,
+      "learning_rate": 8.487334075062914e-07,
+      "loss": 0.7592,
+      "step": 4533
+    },
+    {
+      "epoch": 0.8722585609849942,
+      "grad_norm": 0.6271898369376778,
+      "learning_rate": 8.462231158586654e-07,
+      "loss": 0.3175,
+      "step": 4534
+    },
+    {
+      "epoch": 0.8724509426702578,
+      "grad_norm": 0.9024785157860814,
+      "learning_rate": 8.437163780224011e-07,
+      "loss": 0.8454,
+      "step": 4535
+    },
+    {
+      "epoch": 0.8726433243555214,
+      "grad_norm": 0.9118957102957908,
+      "learning_rate": 8.412131949706958e-07,
+      "loss": 0.7654,
+      "step": 4536
+    },
+    {
+      "epoch": 0.8728357060407849,
+      "grad_norm": 0.9205372945416935,
+      "learning_rate": 8.387135676753755e-07,
+      "loss": 0.7913,
+      "step": 4537
+    },
+    {
+      "epoch": 0.8730280877260485,
+      "grad_norm": 0.858779867425097,
+      "learning_rate": 8.362174971068804e-07,
+      "loss": 0.7689,
+      "step": 4538
+    },
+    {
+      "epoch": 0.873220469411312,
+      "grad_norm": 0.871947625589398,
+      "learning_rate": 8.337249842342721e-07,
+      "loss": 0.7637,
+      "step": 4539
+    },
+    {
+      "epoch": 0.8734128510965756,
+      "grad_norm": 0.8864340167620902,
+      "learning_rate": 8.312360300252287e-07,
+      "loss": 0.8153,
+      "step": 4540
+    },
+    {
+      "epoch": 0.8736052327818392,
+      "grad_norm": 0.8754342408723397,
+      "learning_rate": 8.287506354460484e-07,
+      "loss": 0.7893,
+      "step": 4541
+    },
+    {
+      "epoch": 0.8737976144671027,
+      "grad_norm": 0.9272286993734138,
+      "learning_rate": 8.26268801461646e-07,
+      "loss": 0.7783,
+      "step": 4542
+    },
+    {
+      "epoch": 0.8739899961523663,
+      "grad_norm": 0.8829670807636472,
+      "learning_rate": 8.237905290355563e-07,
+      "loss": 0.7724,
+      "step": 4543
+    },
+    {
+      "epoch": 0.8741823778376299,
+      "grad_norm": 0.9845100574859291,
+      "learning_rate": 8.213158191299297e-07,
+      "loss": 0.7509,
+      "step": 4544
+    },
+    {
+      "epoch": 0.8743747595228935,
+      "grad_norm": 0.9377577460727503,
+      "learning_rate": 8.188446727055311e-07,
+      "loss": 0.8063,
+      "step": 4545
+    },
+    {
+      "epoch": 0.8745671412081569,
+      "grad_norm": 0.9436227585672924,
+      "learning_rate": 8.163770907217506e-07,
+      "loss": 0.8062,
+      "step": 4546
+    },
+    {
+      "epoch": 0.8747595228934205,
+      "grad_norm": 0.8186878807866822,
+      "learning_rate": 8.139130741365819e-07,
+      "loss": 0.811,
+      "step": 4547
+    },
+    {
+      "epoch": 0.8749519045786841,
+      "grad_norm": 0.9059072334708077,
+      "learning_rate": 8.114526239066456e-07,
+      "loss": 0.7766,
+      "step": 4548
+    },
+    {
+      "epoch": 0.8751442862639477,
+      "grad_norm": 0.8335576693067923,
+      "learning_rate": 8.08995740987173e-07,
+      "loss": 0.7638,
+      "step": 4549
+    },
+    {
+      "epoch": 0.8753366679492113,
+      "grad_norm": 0.8418562373750086,
+      "learning_rate": 8.065424263320054e-07,
+      "loss": 0.7636,
+      "step": 4550
+    },
+    {
+      "epoch": 0.8755290496344748,
+      "grad_norm": 0.8232782319825807,
+      "learning_rate": 8.040926808936112e-07,
+      "loss": 0.7881,
+      "step": 4551
+    },
+    {
+      "epoch": 0.8757214313197383,
+      "grad_norm": 0.8795362715442895,
+      "learning_rate": 8.016465056230616e-07,
+      "loss": 0.775,
+      "step": 4552
+    },
+    {
+      "epoch": 0.8759138130050019,
+      "grad_norm": 0.9255186783157416,
+      "learning_rate": 7.99203901470047e-07,
+      "loss": 0.7312,
+      "step": 4553
+    },
+    {
+      "epoch": 0.8761061946902655,
+      "grad_norm": 0.7917877790774815,
+      "learning_rate": 7.967648693828712e-07,
+      "loss": 0.739,
+      "step": 4554
+    },
+    {
+      "epoch": 0.8762985763755291,
+      "grad_norm": 0.8672590216974501,
+      "learning_rate": 7.943294103084487e-07,
+      "loss": 0.7406,
+      "step": 4555
+    },
+    {
+      "epoch": 0.8764909580607926,
+      "grad_norm": 0.8481478634414983,
+      "learning_rate": 7.9189752519231e-07,
+      "loss": 0.7464,
+      "step": 4556
+    },
+    {
+      "epoch": 0.8766833397460562,
+      "grad_norm": 0.8422213732685166,
+      "learning_rate": 7.894692149785954e-07,
+      "loss": 0.7835,
+      "step": 4557
+    },
+    {
+      "epoch": 0.8768757214313198,
+      "grad_norm": 0.8420071371344874,
+      "learning_rate": 7.870444806100619e-07,
+      "loss": 0.7634,
+      "step": 4558
+    },
+    {
+      "epoch": 0.8770681031165833,
+      "grad_norm": 0.9354089990673152,
+      "learning_rate": 7.846233230280698e-07,
+      "loss": 0.8067,
+      "step": 4559
+    },
+    {
+      "epoch": 0.8772604848018468,
+      "grad_norm": 0.9126706815652169,
+      "learning_rate": 7.822057431725994e-07,
+      "loss": 0.7773,
+      "step": 4560
+    },
+    {
+      "epoch": 0.8774528664871104,
+      "grad_norm": 0.978977811473573,
+      "learning_rate": 7.797917419822377e-07,
+      "loss": 0.8018,
+      "step": 4561
+    },
+    {
+      "epoch": 0.877645248172374,
+      "grad_norm": 0.8333753665295305,
+      "learning_rate": 7.773813203941827e-07,
+      "loss": 0.7858,
+      "step": 4562
+    },
+    {
+      "epoch": 0.8778376298576376,
+      "grad_norm": 0.9292535913083386,
+      "learning_rate": 7.749744793442448e-07,
+      "loss": 0.7347,
+      "step": 4563
+    },
+    {
+      "epoch": 0.8780300115429012,
+      "grad_norm": 0.8330505278026218,
+      "learning_rate": 7.725712197668378e-07,
+      "loss": 0.7605,
+      "step": 4564
+    },
+    {
+      "epoch": 0.8782223932281646,
+      "grad_norm": 0.8232699309338427,
+      "learning_rate": 7.701715425949952e-07,
+      "loss": 0.7356,
+      "step": 4565
+    },
+    {
+      "epoch": 0.8784147749134282,
+      "grad_norm": 0.8909770670789945,
+      "learning_rate": 7.677754487603517e-07,
+      "loss": 0.658,
+      "step": 4566
+    },
+    {
+      "epoch": 0.8786071565986918,
+      "grad_norm": 0.8616579784395175,
+      "learning_rate": 7.653829391931533e-07,
+      "loss": 0.7042,
+      "step": 4567
+    },
+    {
+      "epoch": 0.8787995382839554,
+      "grad_norm": 0.8133928392147357,
+      "learning_rate": 7.629940148222559e-07,
+      "loss": 0.7575,
+      "step": 4568
+    },
+    {
+      "epoch": 0.8789919199692189,
+      "grad_norm": 0.9148781037072583,
+      "learning_rate": 7.606086765751209e-07,
+      "loss": 0.8163,
+      "step": 4569
+    },
+    {
+      "epoch": 0.8791843016544825,
+      "grad_norm": 0.9655091649230655,
+      "learning_rate": 7.582269253778185e-07,
+      "loss": 0.7949,
+      "step": 4570
+    },
+    {
+      "epoch": 0.8793766833397461,
+      "grad_norm": 0.8129721078999007,
+      "learning_rate": 7.55848762155027e-07,
+      "loss": 0.7276,
+      "step": 4571
+    },
+    {
+      "epoch": 0.8795690650250096,
+      "grad_norm": 0.8188252204567602,
+      "learning_rate": 7.534741878300333e-07,
+      "loss": 0.778,
+      "step": 4572
+    },
+    {
+      "epoch": 0.8797614467102732,
+      "grad_norm": 0.9578656241837025,
+      "learning_rate": 7.511032033247256e-07,
+      "loss": 0.8672,
+      "step": 4573
+    },
+    {
+      "epoch": 0.8799538283955367,
+      "grad_norm": 0.883649546331947,
+      "learning_rate": 7.487358095596031e-07,
+      "loss": 0.8334,
+      "step": 4574
+    },
+    {
+      "epoch": 0.8801462100808003,
+      "grad_norm": 0.8489767409233582,
+      "learning_rate": 7.463720074537728e-07,
+      "loss": 0.8008,
+      "step": 4575
+    },
+    {
+      "epoch": 0.8803385917660639,
+      "grad_norm": 0.8882187139037643,
+      "learning_rate": 7.440117979249362e-07,
+      "loss": 0.8261,
+      "step": 4576
+    },
+    {
+      "epoch": 0.8805309734513275,
+      "grad_norm": 0.7972642650691867,
+      "learning_rate": 7.416551818894158e-07,
+      "loss": 0.7786,
+      "step": 4577
+    },
+    {
+      "epoch": 0.880723355136591,
+      "grad_norm": 0.8410467122654935,
+      "learning_rate": 7.393021602621264e-07,
+      "loss": 0.7784,
+      "step": 4578
+    },
+    {
+      "epoch": 0.8809157368218545,
+      "grad_norm": 0.8629809618688792,
+      "learning_rate": 7.369527339565951e-07,
+      "loss": 0.7443,
+      "step": 4579
+    },
+    {
+      "epoch": 0.8811081185071181,
+      "grad_norm": 0.8484342797324674,
+      "learning_rate": 7.346069038849469e-07,
+      "loss": 0.7313,
+      "step": 4580
+    },
+    {
+      "epoch": 0.8813005001923817,
+      "grad_norm": 0.9119080835709563,
+      "learning_rate": 7.322646709579173e-07,
+      "loss": 0.8032,
+      "step": 4581
+    },
+    {
+      "epoch": 0.8814928818776453,
+      "grad_norm": 0.902669947166699,
+      "learning_rate": 7.299260360848382e-07,
+      "loss": 0.7965,
+      "step": 4582
+    },
+    {
+      "epoch": 0.8816852635629088,
+      "grad_norm": 0.8948449983032412,
+      "learning_rate": 7.275910001736497e-07,
+      "loss": 0.8322,
+      "step": 4583
+    },
+    {
+      "epoch": 0.8818776452481724,
+      "grad_norm": 0.9018200064885101,
+      "learning_rate": 7.252595641308957e-07,
+      "loss": 0.7713,
+      "step": 4584
+    },
+    {
+      "epoch": 0.882070026933436,
+      "grad_norm": 0.8822351390476658,
+      "learning_rate": 7.229317288617144e-07,
+      "loss": 0.7098,
+      "step": 4585
+    },
+    {
+      "epoch": 0.8822624086186995,
+      "grad_norm": 0.8446172951097765,
+      "learning_rate": 7.20607495269856e-07,
+      "loss": 0.7471,
+      "step": 4586
+    },
+    {
+      "epoch": 0.882454790303963,
+      "grad_norm": 0.9214312921004332,
+      "learning_rate": 7.182868642576679e-07,
+      "loss": 0.7739,
+      "step": 4587
+    },
+    {
+      "epoch": 0.8826471719892266,
+      "grad_norm": 0.8810332198383007,
+      "learning_rate": 7.15969836726097e-07,
+      "loss": 0.7453,
+      "step": 4588
+    },
+    {
+      "epoch": 0.8828395536744902,
+      "grad_norm": 0.8380435289345654,
+      "learning_rate": 7.13656413574696e-07,
+      "loss": 0.8124,
+      "step": 4589
+    },
+    {
+      "epoch": 0.8830319353597538,
+      "grad_norm": 0.8345321622262698,
+      "learning_rate": 7.113465957016097e-07,
+      "loss": 0.7715,
+      "step": 4590
+    },
+    {
+      "epoch": 0.8832243170450174,
+      "grad_norm": 0.8691885866085385,
+      "learning_rate": 7.090403840035942e-07,
+      "loss": 0.7759,
+      "step": 4591
+    },
+    {
+      "epoch": 0.8834166987302808,
+      "grad_norm": 0.9173869322609679,
+      "learning_rate": 7.067377793759999e-07,
+      "loss": 0.801,
+      "step": 4592
+    },
+    {
+      "epoch": 0.8836090804155444,
+      "grad_norm": 0.8385735989107319,
+      "learning_rate": 7.044387827127752e-07,
+      "loss": 0.7595,
+      "step": 4593
+    },
+    {
+      "epoch": 0.883801462100808,
+      "grad_norm": 0.9109371034594694,
+      "learning_rate": 7.021433949064704e-07,
+      "loss": 0.7735,
+      "step": 4594
+    },
+    {
+      "epoch": 0.8839938437860716,
+      "grad_norm": 0.9520680842169031,
+      "learning_rate": 6.99851616848235e-07,
+      "loss": 0.7423,
+      "step": 4595
+    },
+    {
+      "epoch": 0.8841862254713351,
+      "grad_norm": 0.8603936805210145,
+      "learning_rate": 6.975634494278149e-07,
+      "loss": 0.7655,
+      "step": 4596
+    },
+    {
+      "epoch": 0.8843786071565987,
+      "grad_norm": 0.8386802371555956,
+      "learning_rate": 6.952788935335541e-07,
+      "loss": 0.787,
+      "step": 4597
+    },
+    {
+      "epoch": 0.8845709888418622,
+      "grad_norm": 0.8666922901580413,
+      "learning_rate": 6.92997950052402e-07,
+      "loss": 0.7839,
+      "step": 4598
+    },
+    {
+      "epoch": 0.8847633705271258,
+      "grad_norm": 0.8699519691246358,
+      "learning_rate": 6.907206198698912e-07,
+      "loss": 0.7729,
+      "step": 4599
+    },
+    {
+      "epoch": 0.8849557522123894,
+      "grad_norm": 0.9012397354255274,
+      "learning_rate": 6.884469038701646e-07,
+      "loss": 0.7821,
+      "step": 4600
+    },
+    {
+      "epoch": 0.8851481338976529,
+      "grad_norm": 0.8401734838100391,
+      "learning_rate": 6.861768029359595e-07,
+      "loss": 0.8093,
+      "step": 4601
+    },
+    {
+      "epoch": 0.8853405155829165,
+      "grad_norm": 0.8614189446846865,
+      "learning_rate": 6.839103179485995e-07,
+      "loss": 0.7805,
+      "step": 4602
+    },
+    {
+      "epoch": 0.8855328972681801,
+      "grad_norm": 0.8726139456833257,
+      "learning_rate": 6.816474497880177e-07,
+      "loss": 0.7755,
+      "step": 4603
+    },
+    {
+      "epoch": 0.8857252789534437,
+      "grad_norm": 0.8599074679225721,
+      "learning_rate": 6.793881993327366e-07,
+      "loss": 0.7577,
+      "step": 4604
+    },
+    {
+      "epoch": 0.8859176606387072,
+      "grad_norm": 0.9148660859766377,
+      "learning_rate": 6.77132567459875e-07,
+      "loss": 0.7956,
+      "step": 4605
+    },
+    {
+      "epoch": 0.8861100423239707,
+      "grad_norm": 0.8741712772248217,
+      "learning_rate": 6.748805550451453e-07,
+      "loss": 0.819,
+      "step": 4606
+    },
+    {
+      "epoch": 0.8863024240092343,
+      "grad_norm": 0.905373935204237,
+      "learning_rate": 6.726321629628585e-07,
+      "loss": 0.7572,
+      "step": 4607
+    },
+    {
+      "epoch": 0.8864948056944979,
+      "grad_norm": 0.8270937534972379,
+      "learning_rate": 6.703873920859161e-07,
+      "loss": 0.7947,
+      "step": 4608
+    },
+    {
+      "epoch": 0.8866871873797615,
+      "grad_norm": 0.5837298118451213,
+      "learning_rate": 6.681462432858154e-07,
+      "loss": 0.3238,
+      "step": 4609
+    },
+    {
+      "epoch": 0.886879569065025,
+      "grad_norm": 0.8878528122353898,
+      "learning_rate": 6.659087174326506e-07,
+      "loss": 0.7715,
+      "step": 4610
+    },
+    {
+      "epoch": 0.8870719507502886,
+      "grad_norm": 0.6211922559904252,
+      "learning_rate": 6.636748153951e-07,
+      "loss": 0.3445,
+      "step": 4611
+    },
+    {
+      "epoch": 0.8872643324355521,
+      "grad_norm": 0.8332544662160154,
+      "learning_rate": 6.614445380404478e-07,
+      "loss": 0.7362,
+      "step": 4612
+    },
+    {
+      "epoch": 0.8874567141208157,
+      "grad_norm": 0.9243704984151553,
+      "learning_rate": 6.592178862345622e-07,
+      "loss": 0.789,
+      "step": 4613
+    },
+    {
+      "epoch": 0.8876490958060793,
+      "grad_norm": 0.8806948832850009,
+      "learning_rate": 6.569948608419041e-07,
+      "loss": 0.769,
+      "step": 4614
+    },
+    {
+      "epoch": 0.8878414774913428,
+      "grad_norm": 0.879124892317915,
+      "learning_rate": 6.547754627255332e-07,
+      "loss": 0.8009,
+      "step": 4615
+    },
+    {
+      "epoch": 0.8880338591766064,
+      "grad_norm": 1.0383291493810105,
+      "learning_rate": 6.52559692747089e-07,
+      "loss": 0.801,
+      "step": 4616
+    },
+    {
+      "epoch": 0.88822624086187,
+      "grad_norm": 0.898163408943583,
+      "learning_rate": 6.503475517668168e-07,
+      "loss": 0.7722,
+      "step": 4617
+    },
+    {
+      "epoch": 0.8884186225471336,
+      "grad_norm": 0.9855704245204812,
+      "learning_rate": 6.481390406435417e-07,
+      "loss": 0.802,
+      "step": 4618
+    },
+    {
+      "epoch": 0.888611004232397,
+      "grad_norm": 0.5719968982624802,
+      "learning_rate": 6.459341602346858e-07,
+      "loss": 0.3584,
+      "step": 4619
+    },
+    {
+      "epoch": 0.8888033859176606,
+      "grad_norm": 0.8377188426178106,
+      "learning_rate": 6.437329113962576e-07,
+      "loss": 0.7859,
+      "step": 4620
+    },
+    {
+      "epoch": 0.8889957676029242,
+      "grad_norm": 0.9382062487852604,
+      "learning_rate": 6.415352949828601e-07,
+      "loss": 0.7918,
+      "step": 4621
+    },
+    {
+      "epoch": 0.8891881492881878,
+      "grad_norm": 0.8637435530032973,
+      "learning_rate": 6.393413118476821e-07,
+      "loss": 0.7864,
+      "step": 4622
+    },
+    {
+      "epoch": 0.8893805309734514,
+      "grad_norm": 0.9015983535117363,
+      "learning_rate": 6.371509628425021e-07,
+      "loss": 0.8024,
+      "step": 4623
+    },
+    {
+      "epoch": 0.8895729126587149,
+      "grad_norm": 0.9253891923335142,
+      "learning_rate": 6.349642488176943e-07,
+      "loss": 0.792,
+      "step": 4624
+    },
+    {
+      "epoch": 0.8897652943439784,
+      "grad_norm": 0.8951116533908148,
+      "learning_rate": 6.327811706222097e-07,
+      "loss": 0.7765,
+      "step": 4625
+    },
+    {
+      "epoch": 0.889957676029242,
+      "grad_norm": 0.8659271006262403,
+      "learning_rate": 6.306017291035981e-07,
+      "loss": 0.72,
+      "step": 4626
+    },
+    {
+      "epoch": 0.8901500577145056,
+      "grad_norm": 0.8656296157590246,
+      "learning_rate": 6.284259251079939e-07,
+      "loss": 0.7801,
+      "step": 4627
+    },
+    {
+      "epoch": 0.8903424393997691,
+      "grad_norm": 0.9438972797539886,
+      "learning_rate": 6.262537594801177e-07,
+      "loss": 0.7581,
+      "step": 4628
+    },
+    {
+      "epoch": 0.8905348210850327,
+      "grad_norm": 0.8337840549092757,
+      "learning_rate": 6.240852330632796e-07,
+      "loss": 0.7882,
+      "step": 4629
+    },
+    {
+      "epoch": 0.8907272027702963,
+      "grad_norm": 0.8541771871692758,
+      "learning_rate": 6.219203466993762e-07,
+      "loss": 0.7755,
+      "step": 4630
+    },
+    {
+      "epoch": 0.8909195844555599,
+      "grad_norm": 0.9235497458413212,
+      "learning_rate": 6.197591012288918e-07,
+      "loss": 0.7571,
+      "step": 4631
+    },
+    {
+      "epoch": 0.8911119661408234,
+      "grad_norm": 0.8889019336638634,
+      "learning_rate": 6.17601497490895e-07,
+      "loss": 0.8177,
+      "step": 4632
+    },
+    {
+      "epoch": 0.8913043478260869,
+      "grad_norm": 0.9748439152516984,
+      "learning_rate": 6.154475363230417e-07,
+      "loss": 0.7963,
+      "step": 4633
+    },
+    {
+      "epoch": 0.8914967295113505,
+      "grad_norm": 0.8835212926458577,
+      "learning_rate": 6.132972185615749e-07,
+      "loss": 0.8325,
+      "step": 4634
+    },
+    {
+      "epoch": 0.8916891111966141,
+      "grad_norm": 0.8778082661832595,
+      "learning_rate": 6.111505450413202e-07,
+      "loss": 0.7866,
+      "step": 4635
+    },
+    {
+      "epoch": 0.8918814928818777,
+      "grad_norm": 0.822652017540183,
+      "learning_rate": 6.090075165956943e-07,
+      "loss": 0.7217,
+      "step": 4636
+    },
+    {
+      "epoch": 0.8920738745671412,
+      "grad_norm": 0.8603038997430373,
+      "learning_rate": 6.068681340566896e-07,
+      "loss": 0.7878,
+      "step": 4637
+    },
+    {
+      "epoch": 0.8922662562524047,
+      "grad_norm": 0.8937047039937795,
+      "learning_rate": 6.047323982548924e-07,
+      "loss": 0.8279,
+      "step": 4638
+    },
+    {
+      "epoch": 0.8924586379376683,
+      "grad_norm": 0.9090307009982274,
+      "learning_rate": 6.026003100194633e-07,
+      "loss": 0.7661,
+      "step": 4639
+    },
+    {
+      "epoch": 0.8926510196229319,
+      "grad_norm": 0.8257564600020505,
+      "learning_rate": 6.004718701781575e-07,
+      "loss": 0.7837,
+      "step": 4640
+    },
+    {
+      "epoch": 0.8928434013081955,
+      "grad_norm": 0.8994870831657041,
+      "learning_rate": 5.983470795573088e-07,
+      "loss": 0.7648,
+      "step": 4641
+    },
+    {
+      "epoch": 0.893035782993459,
+      "grad_norm": 0.6308061129999719,
+      "learning_rate": 5.962259389818292e-07,
+      "loss": 0.3258,
+      "step": 4642
+    },
+    {
+      "epoch": 0.8932281646787226,
+      "grad_norm": 0.8495438471504151,
+      "learning_rate": 5.941084492752236e-07,
+      "loss": 0.7723,
+      "step": 4643
+    },
+    {
+      "epoch": 0.8934205463639862,
+      "grad_norm": 0.8715867071544464,
+      "learning_rate": 5.91994611259572e-07,
+      "loss": 0.8152,
+      "step": 4644
+    },
+    {
+      "epoch": 0.8936129280492497,
+      "grad_norm": 0.9500138134719631,
+      "learning_rate": 5.898844257555392e-07,
+      "loss": 0.7595,
+      "step": 4645
+    },
+    {
+      "epoch": 0.8938053097345132,
+      "grad_norm": 0.9038044170144763,
+      "learning_rate": 5.87777893582372e-07,
+      "loss": 0.7902,
+      "step": 4646
+    },
+    {
+      "epoch": 0.8939976914197768,
+      "grad_norm": 0.8900235383377533,
+      "learning_rate": 5.856750155578983e-07,
+      "loss": 0.8123,
+      "step": 4647
+    },
+    {
+      "epoch": 0.8941900731050404,
+      "grad_norm": 0.9457121544483315,
+      "learning_rate": 5.835757924985286e-07,
+      "loss": 0.7936,
+      "step": 4648
+    },
+    {
+      "epoch": 0.894382454790304,
+      "grad_norm": 0.8713939927379611,
+      "learning_rate": 5.81480225219252e-07,
+      "loss": 0.7568,
+      "step": 4649
+    },
+    {
+      "epoch": 0.8945748364755676,
+      "grad_norm": 0.9002903233479111,
+      "learning_rate": 5.793883145336443e-07,
+      "loss": 0.8253,
+      "step": 4650
+    },
+    {
+      "epoch": 0.894767218160831,
+      "grad_norm": 0.8995329986855699,
+      "learning_rate": 5.773000612538505e-07,
+      "loss": 0.7942,
+      "step": 4651
+    },
+    {
+      "epoch": 0.8949595998460946,
+      "grad_norm": 0.902296176408974,
+      "learning_rate": 5.752154661906085e-07,
+      "loss": 0.7995,
+      "step": 4652
+    },
+    {
+      "epoch": 0.8951519815313582,
+      "grad_norm": 0.8389700083224155,
+      "learning_rate": 5.731345301532265e-07,
+      "loss": 0.7454,
+      "step": 4653
+    },
+    {
+      "epoch": 0.8953443632166218,
+      "grad_norm": 0.9049925647722517,
+      "learning_rate": 5.710572539495962e-07,
+      "loss": 0.8309,
+      "step": 4654
+    },
+    {
+      "epoch": 0.8955367449018854,
+      "grad_norm": 0.8812901619648547,
+      "learning_rate": 5.68983638386188e-07,
+      "loss": 0.8141,
+      "step": 4655
+    },
+    {
+      "epoch": 0.8957291265871489,
+      "grad_norm": 0.8874993447416563,
+      "learning_rate": 5.669136842680512e-07,
+      "loss": 0.7369,
+      "step": 4656
+    },
+    {
+      "epoch": 0.8959215082724125,
+      "grad_norm": 0.843800677611561,
+      "learning_rate": 5.648473923988129e-07,
+      "loss": 0.7376,
+      "step": 4657
+    },
+    {
+      "epoch": 0.896113889957676,
+      "grad_norm": 0.8717789207483866,
+      "learning_rate": 5.627847635806771e-07,
+      "loss": 0.767,
+      "step": 4658
+    },
+    {
+      "epoch": 0.8963062716429396,
+      "grad_norm": 0.8263674924843183,
+      "learning_rate": 5.607257986144321e-07,
+      "loss": 0.7993,
+      "step": 4659
+    },
+    {
+      "epoch": 0.8964986533282031,
+      "grad_norm": 0.8771473862170794,
+      "learning_rate": 5.58670498299434e-07,
+      "loss": 0.7856,
+      "step": 4660
+    },
+    {
+      "epoch": 0.8966910350134667,
+      "grad_norm": 0.9623684310585991,
+      "learning_rate": 5.566188634336212e-07,
+      "loss": 0.7903,
+      "step": 4661
+    },
+    {
+      "epoch": 0.8968834166987303,
+      "grad_norm": 0.9115166621515631,
+      "learning_rate": 5.545708948135142e-07,
+      "loss": 0.7906,
+      "step": 4662
+    },
+    {
+      "epoch": 0.8970757983839939,
+      "grad_norm": 0.84830874707016,
+      "learning_rate": 5.525265932341984e-07,
+      "loss": 0.8128,
+      "step": 4663
+    },
+    {
+      "epoch": 0.8972681800692575,
+      "grad_norm": 0.8627476717575089,
+      "learning_rate": 5.504859594893475e-07,
+      "loss": 0.7223,
+      "step": 4664
+    },
+    {
+      "epoch": 0.8974605617545209,
+      "grad_norm": 0.8704968929246587,
+      "learning_rate": 5.484489943712013e-07,
+      "loss": 0.8029,
+      "step": 4665
+    },
+    {
+      "epoch": 0.8976529434397845,
+      "grad_norm": 0.8875903091390805,
+      "learning_rate": 5.464156986705826e-07,
+      "loss": 0.8289,
+      "step": 4666
+    },
+    {
+      "epoch": 0.8978453251250481,
+      "grad_norm": 0.8789450315962918,
+      "learning_rate": 5.443860731768869e-07,
+      "loss": 0.8004,
+      "step": 4667
+    },
+    {
+      "epoch": 0.8980377068103117,
+      "grad_norm": 0.8192443858505912,
+      "learning_rate": 5.423601186780836e-07,
+      "loss": 0.7934,
+      "step": 4668
+    },
+    {
+      "epoch": 0.8982300884955752,
+      "grad_norm": 0.939311816629026,
+      "learning_rate": 5.403378359607181e-07,
+      "loss": 0.7924,
+      "step": 4669
+    },
+    {
+      "epoch": 0.8984224701808388,
+      "grad_norm": 0.9075520321543574,
+      "learning_rate": 5.383192258099113e-07,
+      "loss": 0.7492,
+      "step": 4670
+    },
+    {
+      "epoch": 0.8986148518661023,
+      "grad_norm": 0.8622672948291602,
+      "learning_rate": 5.36304289009355e-07,
+      "loss": 0.7394,
+      "step": 4671
+    },
+    {
+      "epoch": 0.8988072335513659,
+      "grad_norm": 0.8630827416674403,
+      "learning_rate": 5.342930263413193e-07,
+      "loss": 0.7402,
+      "step": 4672
+    },
+    {
+      "epoch": 0.8989996152366295,
+      "grad_norm": 0.9339872278310696,
+      "learning_rate": 5.322854385866439e-07,
+      "loss": 0.826,
+      "step": 4673
+    },
+    {
+      "epoch": 0.899191996921893,
+      "grad_norm": 0.9187426790160187,
+      "learning_rate": 5.302815265247452e-07,
+      "loss": 0.7349,
+      "step": 4674
+    },
+    {
+      "epoch": 0.8993843786071566,
+      "grad_norm": 0.8653642194626923,
+      "learning_rate": 5.282812909336077e-07,
+      "loss": 0.7486,
+      "step": 4675
+    },
+    {
+      "epoch": 0.8995767602924202,
+      "grad_norm": 0.8465877969381373,
+      "learning_rate": 5.262847325897968e-07,
+      "loss": 0.8017,
+      "step": 4676
+    },
+    {
+      "epoch": 0.8997691419776838,
+      "grad_norm": 0.8562929907894397,
+      "learning_rate": 5.242918522684392e-07,
+      "loss": 0.8097,
+      "step": 4677
+    },
+    {
+      "epoch": 0.8999615236629472,
+      "grad_norm": 0.8718305962481409,
+      "learning_rate": 5.22302650743245e-07,
+      "loss": 0.7823,
+      "step": 4678
+    },
+    {
+      "epoch": 0.9001539053482108,
+      "grad_norm": 0.6298552608317564,
+      "learning_rate": 5.203171287864872e-07,
+      "loss": 0.3494,
+      "step": 4679
+    },
+    {
+      "epoch": 0.9003462870334744,
+      "grad_norm": 0.9153492367083633,
+      "learning_rate": 5.183352871690162e-07,
+      "loss": 0.7787,
+      "step": 4680
+    },
+    {
+      "epoch": 0.900538668718738,
+      "grad_norm": 0.843142808262279,
+      "learning_rate": 5.163571266602485e-07,
+      "loss": 0.742,
+      "step": 4681
+    },
+    {
+      "epoch": 0.9007310504040016,
+      "grad_norm": 0.9311607205281219,
+      "learning_rate": 5.143826480281778e-07,
+      "loss": 0.8166,
+      "step": 4682
+    },
+    {
+      "epoch": 0.9009234320892651,
+      "grad_norm": 0.8906607647036828,
+      "learning_rate": 5.124118520393606e-07,
+      "loss": 0.7568,
+      "step": 4683
+    },
+    {
+      "epoch": 0.9011158137745287,
+      "grad_norm": 0.8897032432810476,
+      "learning_rate": 5.104447394589295e-07,
+      "loss": 0.816,
+      "step": 4684
+    },
+    {
+      "epoch": 0.9013081954597922,
+      "grad_norm": 0.9030446592001629,
+      "learning_rate": 5.084813110505871e-07,
+      "loss": 0.7837,
+      "step": 4685
+    },
+    {
+      "epoch": 0.9015005771450558,
+      "grad_norm": 0.8239474424807613,
+      "learning_rate": 5.065215675766023e-07,
+      "loss": 0.7641,
+      "step": 4686
+    },
+    {
+      "epoch": 0.9016929588303193,
+      "grad_norm": 0.9424974862215083,
+      "learning_rate": 5.045655097978131e-07,
+      "loss": 0.7918,
+      "step": 4687
+    },
+    {
+      "epoch": 0.9018853405155829,
+      "grad_norm": 0.8407820949107572,
+      "learning_rate": 5.026131384736321e-07,
+      "loss": 0.8374,
+      "step": 4688
+    },
+    {
+      "epoch": 0.9020777222008465,
+      "grad_norm": 0.8924436507551087,
+      "learning_rate": 5.006644543620342e-07,
+      "loss": 0.7716,
+      "step": 4689
+    },
+    {
+      "epoch": 0.9022701038861101,
+      "grad_norm": 0.8398070333022961,
+      "learning_rate": 4.987194582195687e-07,
+      "loss": 0.7804,
+      "step": 4690
+    },
+    {
+      "epoch": 0.9024624855713737,
+      "grad_norm": 0.8277665554995262,
+      "learning_rate": 4.967781508013459e-07,
+      "loss": 0.8046,
+      "step": 4691
+    },
+    {
+      "epoch": 0.9026548672566371,
+      "grad_norm": 0.9142984274093124,
+      "learning_rate": 4.948405328610506e-07,
+      "loss": 0.8487,
+      "step": 4692
+    },
+    {
+      "epoch": 0.9028472489419007,
+      "grad_norm": 0.9183036838824703,
+      "learning_rate": 4.929066051509346e-07,
+      "loss": 0.7227,
+      "step": 4693
+    },
+    {
+      "epoch": 0.9030396306271643,
+      "grad_norm": 0.8912069234896757,
+      "learning_rate": 4.909763684218116e-07,
+      "loss": 0.7703,
+      "step": 4694
+    },
+    {
+      "epoch": 0.9032320123124279,
+      "grad_norm": 0.8863033607100497,
+      "learning_rate": 4.890498234230689e-07,
+      "loss": 0.7956,
+      "step": 4695
+    },
+    {
+      "epoch": 0.9034243939976914,
+      "grad_norm": 0.9260399419775517,
+      "learning_rate": 4.871269709026561e-07,
+      "loss": 0.7218,
+      "step": 4696
+    },
+    {
+      "epoch": 0.903616775682955,
+      "grad_norm": 0.8944538543089001,
+      "learning_rate": 4.852078116070902e-07,
+      "loss": 0.7876,
+      "step": 4697
+    },
+    {
+      "epoch": 0.9038091573682185,
+      "grad_norm": 0.8529935807824636,
+      "learning_rate": 4.832923462814565e-07,
+      "loss": 0.7712,
+      "step": 4698
+    },
+    {
+      "epoch": 0.9040015390534821,
+      "grad_norm": 0.9933777614604952,
+      "learning_rate": 4.813805756694035e-07,
+      "loss": 0.7928,
+      "step": 4699
+    },
+    {
+      "epoch": 0.9041939207387457,
+      "grad_norm": 0.9284075820148617,
+      "learning_rate": 4.794725005131462e-07,
+      "loss": 0.768,
+      "step": 4700
+    },
+    {
+      "epoch": 0.9043863024240092,
+      "grad_norm": 0.9622376054105901,
+      "learning_rate": 4.775681215534656e-07,
+      "loss": 0.7673,
+      "step": 4701
+    },
+    {
+      "epoch": 0.9045786841092728,
+      "grad_norm": 0.5905756900242728,
+      "learning_rate": 4.7566743952970894e-07,
+      "loss": 0.303,
+      "step": 4702
+    },
+    {
+      "epoch": 0.9047710657945364,
+      "grad_norm": 0.8791251733636911,
+      "learning_rate": 4.7377045517978173e-07,
+      "loss": 0.7996,
+      "step": 4703
+    },
+    {
+      "epoch": 0.9049634474798,
+      "grad_norm": 0.8532941838673056,
+      "learning_rate": 4.7187716924016355e-07,
+      "loss": 0.7613,
+      "step": 4704
+    },
+    {
+      "epoch": 0.9051558291650635,
+      "grad_norm": 0.8767914387862712,
+      "learning_rate": 4.6998758244588995e-07,
+      "loss": 0.6998,
+      "step": 4705
+    },
+    {
+      "epoch": 0.905348210850327,
+      "grad_norm": 0.9005144379881396,
+      "learning_rate": 4.6810169553056616e-07,
+      "loss": 0.7702,
+      "step": 4706
+    },
+    {
+      "epoch": 0.9055405925355906,
+      "grad_norm": 0.8956658891064988,
+      "learning_rate": 4.662195092263566e-07,
+      "loss": 0.7835,
+      "step": 4707
+    },
+    {
+      "epoch": 0.9057329742208542,
+      "grad_norm": 0.8051644839756437,
+      "learning_rate": 4.643410242639912e-07,
+      "loss": 0.8226,
+      "step": 4708
+    },
+    {
+      "epoch": 0.9059253559061178,
+      "grad_norm": 0.8846904091276254,
+      "learning_rate": 4.6246624137276206e-07,
+      "loss": 0.7768,
+      "step": 4709
+    },
+    {
+      "epoch": 0.9061177375913813,
+      "grad_norm": 1.0123862484751038,
+      "learning_rate": 4.605951612805237e-07,
+      "loss": 0.7496,
+      "step": 4710
+    },
+    {
+      "epoch": 0.9063101192766448,
+      "grad_norm": 0.9155849021322697,
+      "learning_rate": 4.587277847136984e-07,
+      "loss": 0.7991,
+      "step": 4711
+    },
+    {
+      "epoch": 0.9065025009619084,
+      "grad_norm": 0.9349001284598418,
+      "learning_rate": 4.568641123972606e-07,
+      "loss": 0.7678,
+      "step": 4712
+    },
+    {
+      "epoch": 0.906694882647172,
+      "grad_norm": 0.9872080041332563,
+      "learning_rate": 4.550041450547549e-07,
+      "loss": 0.788,
+      "step": 4713
+    },
+    {
+      "epoch": 0.9068872643324356,
+      "grad_norm": 0.85886602213367,
+      "learning_rate": 4.5314788340828365e-07,
+      "loss": 0.8109,
+      "step": 4714
+    },
+    {
+      "epoch": 0.9070796460176991,
+      "grad_norm": 0.8759068004078431,
+      "learning_rate": 4.512953281785104e-07,
+      "loss": 0.7673,
+      "step": 4715
+    },
+    {
+      "epoch": 0.9072720277029627,
+      "grad_norm": 0.8599224940106512,
+      "learning_rate": 4.494464800846654e-07,
+      "loss": 0.7171,
+      "step": 4716
+    },
+    {
+      "epoch": 0.9074644093882263,
+      "grad_norm": 0.8340055065698861,
+      "learning_rate": 4.476013398445289e-07,
+      "loss": 0.8569,
+      "step": 4717
+    },
+    {
+      "epoch": 0.9076567910734898,
+      "grad_norm": 0.9130352489104482,
+      "learning_rate": 4.4575990817445234e-07,
+      "loss": 0.8083,
+      "step": 4718
+    },
+    {
+      "epoch": 0.9078491727587533,
+      "grad_norm": 0.8427394487720249,
+      "learning_rate": 4.4392218578934164e-07,
+      "loss": 0.8156,
+      "step": 4719
+    },
+    {
+      "epoch": 0.9080415544440169,
+      "grad_norm": 0.8764565362528305,
+      "learning_rate": 4.4208817340266385e-07,
+      "loss": 0.8251,
+      "step": 4720
+    },
+    {
+      "epoch": 0.9082339361292805,
+      "grad_norm": 0.8607462635861213,
+      "learning_rate": 4.4025787172644495e-07,
+      "loss": 0.7532,
+      "step": 4721
+    },
+    {
+      "epoch": 0.9084263178145441,
+      "grad_norm": 0.8963976839151161,
+      "learning_rate": 4.384312814712721e-07,
+      "loss": 0.8138,
+      "step": 4722
+    },
+    {
+      "epoch": 0.9086186994998077,
+      "grad_norm": 0.9015702435842757,
+      "learning_rate": 4.366084033462914e-07,
+      "loss": 0.7755,
+      "step": 4723
+    },
+    {
+      "epoch": 0.9088110811850711,
+      "grad_norm": 0.8936271225084896,
+      "learning_rate": 4.3478923805920335e-07,
+      "loss": 0.7475,
+      "step": 4724
+    },
+    {
+      "epoch": 0.9090034628703347,
+      "grad_norm": 0.8235054754564871,
+      "learning_rate": 4.329737863162753e-07,
+      "loss": 0.7391,
+      "step": 4725
+    },
+    {
+      "epoch": 0.9091958445555983,
+      "grad_norm": 0.8563233115279046,
+      "learning_rate": 4.311620488223256e-07,
+      "loss": 0.7993,
+      "step": 4726
+    },
+    {
+      "epoch": 0.9093882262408619,
+      "grad_norm": 0.8955690842941961,
+      "learning_rate": 4.2935402628073166e-07,
+      "loss": 0.7869,
+      "step": 4727
+    },
+    {
+      "epoch": 0.9095806079261254,
+      "grad_norm": 0.8565518735371742,
+      "learning_rate": 4.27549719393433e-07,
+      "loss": 0.7857,
+      "step": 4728
+    },
+    {
+      "epoch": 0.909772989611389,
+      "grad_norm": 0.8843707339729826,
+      "learning_rate": 4.2574912886092166e-07,
+      "loss": 0.8063,
+      "step": 4729
+    },
+    {
+      "epoch": 0.9099653712966526,
+      "grad_norm": 0.8478497495554939,
+      "learning_rate": 4.239522553822495e-07,
+      "loss": 0.7574,
+      "step": 4730
+    },
+    {
+      "epoch": 0.9101577529819161,
+      "grad_norm": 0.9440845267083149,
+      "learning_rate": 4.221590996550251e-07,
+      "loss": 0.761,
+      "step": 4731
+    },
+    {
+      "epoch": 0.9103501346671797,
+      "grad_norm": 0.9594460579944086,
+      "learning_rate": 4.203696623754139e-07,
+      "loss": 0.7914,
+      "step": 4732
+    },
+    {
+      "epoch": 0.9105425163524432,
+      "grad_norm": 0.9336940456861939,
+      "learning_rate": 4.1858394423813563e-07,
+      "loss": 0.7663,
+      "step": 4733
+    },
+    {
+      "epoch": 0.9107348980377068,
+      "grad_norm": 0.8530636007712875,
+      "learning_rate": 4.1680194593646696e-07,
+      "loss": 0.8036,
+      "step": 4734
+    },
+    {
+      "epoch": 0.9109272797229704,
+      "grad_norm": 0.8955481093729514,
+      "learning_rate": 4.1502366816224327e-07,
+      "loss": 0.8084,
+      "step": 4735
+    },
+    {
+      "epoch": 0.911119661408234,
+      "grad_norm": 0.8861039726074051,
+      "learning_rate": 4.1324911160585014e-07,
+      "loss": 0.7941,
+      "step": 4736
+    },
+    {
+      "epoch": 0.9113120430934974,
+      "grad_norm": 0.9103401986928893,
+      "learning_rate": 4.1147827695623643e-07,
+      "loss": 0.8072,
+      "step": 4737
+    },
+    {
+      "epoch": 0.911504424778761,
+      "grad_norm": 0.8491084071042605,
+      "learning_rate": 4.097111649008967e-07,
+      "loss": 0.7663,
+      "step": 4738
+    },
+    {
+      "epoch": 0.9116968064640246,
+      "grad_norm": 0.8774728816181154,
+      "learning_rate": 4.0794777612588543e-07,
+      "loss": 0.8069,
+      "step": 4739
+    },
+    {
+      "epoch": 0.9118891881492882,
+      "grad_norm": 0.8373598113826906,
+      "learning_rate": 4.061881113158117e-07,
+      "loss": 0.8019,
+      "step": 4740
+    },
+    {
+      "epoch": 0.9120815698345518,
+      "grad_norm": 0.9126235053385477,
+      "learning_rate": 4.044321711538368e-07,
+      "loss": 0.7669,
+      "step": 4741
+    },
+    {
+      "epoch": 0.9122739515198153,
+      "grad_norm": 0.9012013881588402,
+      "learning_rate": 4.02679956321681e-07,
+      "loss": 0.8066,
+      "step": 4742
+    },
+    {
+      "epoch": 0.9124663332050789,
+      "grad_norm": 0.9306433579075384,
+      "learning_rate": 4.00931467499609e-07,
+      "loss": 0.8294,
+      "step": 4743
+    },
+    {
+      "epoch": 0.9126587148903424,
+      "grad_norm": 0.8934019897712258,
+      "learning_rate": 3.9918670536644776e-07,
+      "loss": 0.7928,
+      "step": 4744
+    },
+    {
+      "epoch": 0.912851096575606,
+      "grad_norm": 0.9105175453279811,
+      "learning_rate": 3.974456705995733e-07,
+      "loss": 0.7638,
+      "step": 4745
+    },
+    {
+      "epoch": 0.9130434782608695,
+      "grad_norm": 0.9256776477837229,
+      "learning_rate": 3.9570836387491487e-07,
+      "loss": 0.7958,
+      "step": 4746
+    },
+    {
+      "epoch": 0.9132358599461331,
+      "grad_norm": 0.8592042740659666,
+      "learning_rate": 3.9397478586695513e-07,
+      "loss": 0.8019,
+      "step": 4747
+    },
+    {
+      "epoch": 0.9134282416313967,
+      "grad_norm": 0.8748767409334469,
+      "learning_rate": 3.9224493724872915e-07,
+      "loss": 0.7637,
+      "step": 4748
+    },
+    {
+      "epoch": 0.9136206233166603,
+      "grad_norm": 0.8731012284305932,
+      "learning_rate": 3.90518818691823e-07,
+      "loss": 0.8028,
+      "step": 4749
+    },
+    {
+      "epoch": 0.9138130050019239,
+      "grad_norm": 0.9164911093538245,
+      "learning_rate": 3.8879643086637384e-07,
+      "loss": 0.806,
+      "step": 4750
+    },
+    {
+      "epoch": 0.9140053866871873,
+      "grad_norm": 0.8977253696156583,
+      "learning_rate": 3.8707777444107697e-07,
+      "loss": 0.8839,
+      "step": 4751
+    },
+    {
+      "epoch": 0.9141977683724509,
+      "grad_norm": 0.8842932808173924,
+      "learning_rate": 3.8536285008316854e-07,
+      "loss": 0.8135,
+      "step": 4752
+    },
+    {
+      "epoch": 0.9143901500577145,
+      "grad_norm": 0.8671658457908289,
+      "learning_rate": 3.8365165845844266e-07,
+      "loss": 0.7419,
+      "step": 4753
+    },
+    {
+      "epoch": 0.9145825317429781,
+      "grad_norm": 0.9144944589418328,
+      "learning_rate": 3.819442002312457e-07,
+      "loss": 0.7753,
+      "step": 4754
+    },
+    {
+      "epoch": 0.9147749134282416,
+      "grad_norm": 0.8333827062881418,
+      "learning_rate": 3.8024047606446736e-07,
+      "loss": 0.748,
+      "step": 4755
+    },
+    {
+      "epoch": 0.9149672951135052,
+      "grad_norm": 0.8408771880553657,
+      "learning_rate": 3.785404866195552e-07,
+      "loss": 0.8281,
+      "step": 4756
+    },
+    {
+      "epoch": 0.9151596767987688,
+      "grad_norm": 0.8744140255752161,
+      "learning_rate": 3.768442325565036e-07,
+      "loss": 0.7929,
+      "step": 4757
+    },
+    {
+      "epoch": 0.9153520584840323,
+      "grad_norm": 0.8222377395501438,
+      "learning_rate": 3.751517145338546e-07,
+      "loss": 0.7881,
+      "step": 4758
+    },
+    {
+      "epoch": 0.9155444401692959,
+      "grad_norm": 0.9992393828259566,
+      "learning_rate": 3.7346293320870363e-07,
+      "loss": 0.7905,
+      "step": 4759
+    },
+    {
+      "epoch": 0.9157368218545594,
+      "grad_norm": 0.5974612893851146,
+      "learning_rate": 3.717778892366941e-07,
+      "loss": 0.3162,
+      "step": 4760
+    },
+    {
+      "epoch": 0.915929203539823,
+      "grad_norm": 0.623574968133631,
+      "learning_rate": 3.700965832720171e-07,
+      "loss": 0.3099,
+      "step": 4761
+    },
+    {
+      "epoch": 0.9161215852250866,
+      "grad_norm": 0.855023916696797,
+      "learning_rate": 3.684190159674117e-07,
+      "loss": 0.772,
+      "step": 4762
+    },
+    {
+      "epoch": 0.9163139669103502,
+      "grad_norm": 0.9037472616048626,
+      "learning_rate": 3.6674518797417236e-07,
+      "loss": 0.7795,
+      "step": 4763
+    },
+    {
+      "epoch": 0.9165063485956138,
+      "grad_norm": 0.911259763990568,
+      "learning_rate": 3.6507509994213155e-07,
+      "loss": 0.8234,
+      "step": 4764
+    },
+    {
+      "epoch": 0.9166987302808772,
+      "grad_norm": 0.9124563913247982,
+      "learning_rate": 3.6340875251967946e-07,
+      "loss": 0.779,
+      "step": 4765
+    },
+    {
+      "epoch": 0.9168911119661408,
+      "grad_norm": 0.8206120528204237,
+      "learning_rate": 3.617461463537464e-07,
+      "loss": 0.8164,
+      "step": 4766
+    },
+    {
+      "epoch": 0.9170834936514044,
+      "grad_norm": 0.8916938574497841,
+      "learning_rate": 3.6008728208981157e-07,
+      "loss": 0.8249,
+      "step": 4767
+    },
+    {
+      "epoch": 0.917275875336668,
+      "grad_norm": 0.8835561025353649,
+      "learning_rate": 3.5843216037190873e-07,
+      "loss": 0.7707,
+      "step": 4768
+    },
+    {
+      "epoch": 0.9174682570219315,
+      "grad_norm": 0.5744416109588978,
+      "learning_rate": 3.5678078184260834e-07,
+      "loss": 0.3296,
+      "step": 4769
+    },
+    {
+      "epoch": 0.917660638707195,
+      "grad_norm": 0.8576003072987293,
+      "learning_rate": 3.5513314714303524e-07,
+      "loss": 0.8014,
+      "step": 4770
+    },
+    {
+      "epoch": 0.9178530203924586,
+      "grad_norm": 0.9310193226379506,
+      "learning_rate": 3.5348925691285675e-07,
+      "loss": 0.8198,
+      "step": 4771
+    },
+    {
+      "epoch": 0.9180454020777222,
+      "grad_norm": 0.9361007325147648,
+      "learning_rate": 3.518491117902878e-07,
+      "loss": 0.7893,
+      "step": 4772
+    },
+    {
+      "epoch": 0.9182377837629858,
+      "grad_norm": 0.853328413977008,
+      "learning_rate": 3.502127124120891e-07,
+      "loss": 0.8372,
+      "step": 4773
+    },
+    {
+      "epoch": 0.9184301654482493,
+      "grad_norm": 0.8142168650022819,
+      "learning_rate": 3.48580059413568e-07,
+      "loss": 0.7634,
+      "step": 4774
+    },
+    {
+      "epoch": 0.9186225471335129,
+      "grad_norm": 0.8286809704759374,
+      "learning_rate": 3.4695115342857524e-07,
+      "loss": 0.8132,
+      "step": 4775
+    },
+    {
+      "epoch": 0.9188149288187765,
+      "grad_norm": 0.8787152830710876,
+      "learning_rate": 3.4532599508950826e-07,
+      "loss": 0.7232,
+      "step": 4776
+    },
+    {
+      "epoch": 0.91900731050404,
+      "grad_norm": 0.873956128931072,
+      "learning_rate": 3.437045850273113e-07,
+      "loss": 0.8074,
+      "step": 4777
+    },
+    {
+      "epoch": 0.9191996921893035,
+      "grad_norm": 0.9730388510959023,
+      "learning_rate": 3.420869238714708e-07,
+      "loss": 0.7763,
+      "step": 4778
+    },
+    {
+      "epoch": 0.9193920738745671,
+      "grad_norm": 0.7983188221262135,
+      "learning_rate": 3.404730122500155e-07,
+      "loss": 0.7448,
+      "step": 4779
+    },
+    {
+      "epoch": 0.9195844555598307,
+      "grad_norm": 0.6439139612397974,
+      "learning_rate": 3.3886285078952753e-07,
+      "loss": 0.3351,
+      "step": 4780
+    },
+    {
+      "epoch": 0.9197768372450943,
+      "grad_norm": 0.8993474698827857,
+      "learning_rate": 3.3725644011512125e-07,
+      "loss": 0.7522,
+      "step": 4781
+    },
+    {
+      "epoch": 0.9199692189303579,
+      "grad_norm": 0.8936350556147636,
+      "learning_rate": 3.356537808504634e-07,
+      "loss": 0.7937,
+      "step": 4782
+    },
+    {
+      "epoch": 0.9201616006156214,
+      "grad_norm": 0.8578325007480772,
+      "learning_rate": 3.3405487361776177e-07,
+      "loss": 0.7536,
+      "step": 4783
+    },
+    {
+      "epoch": 0.9203539823008849,
+      "grad_norm": 0.971098164097305,
+      "learning_rate": 3.3245971903776654e-07,
+      "loss": 0.8147,
+      "step": 4784
+    },
+    {
+      "epoch": 0.9205463639861485,
+      "grad_norm": 0.8612291302740887,
+      "learning_rate": 3.308683177297711e-07,
+      "loss": 0.7907,
+      "step": 4785
+    },
+    {
+      "epoch": 0.9207387456714121,
+      "grad_norm": 0.8975851226792488,
+      "learning_rate": 3.292806703116125e-07,
+      "loss": 0.7495,
+      "step": 4786
+    },
+    {
+      "epoch": 0.9209311273566756,
+      "grad_norm": 0.9221121920594769,
+      "learning_rate": 3.2769677739966975e-07,
+      "loss": 0.7636,
+      "step": 4787
+    },
+    {
+      "epoch": 0.9211235090419392,
+      "grad_norm": 0.8725391195706039,
+      "learning_rate": 3.2611663960886665e-07,
+      "loss": 0.7949,
+      "step": 4788
+    },
+    {
+      "epoch": 0.9213158907272028,
+      "grad_norm": 0.8483177656997632,
+      "learning_rate": 3.245402575526646e-07,
+      "loss": 0.796,
+      "step": 4789
+    },
+    {
+      "epoch": 0.9215082724124664,
+      "grad_norm": 0.878289670221381,
+      "learning_rate": 3.2296763184306965e-07,
+      "loss": 0.7319,
+      "step": 4790
+    },
+    {
+      "epoch": 0.9217006540977299,
+      "grad_norm": 0.5843651829714616,
+      "learning_rate": 3.2139876309063233e-07,
+      "loss": 0.3354,
+      "step": 4791
+    },
+    {
+      "epoch": 0.9218930357829934,
+      "grad_norm": 0.8269932644123661,
+      "learning_rate": 3.198336519044376e-07,
+      "loss": 0.7263,
+      "step": 4792
+    },
+    {
+      "epoch": 0.922085417468257,
+      "grad_norm": 0.9335712051821067,
+      "learning_rate": 3.182722988921161e-07,
+      "loss": 0.77,
+      "step": 4793
+    },
+    {
+      "epoch": 0.9222777991535206,
+      "grad_norm": 0.9370221292182082,
+      "learning_rate": 3.167147046598418e-07,
+      "loss": 0.8229,
+      "step": 4794
+    },
+    {
+      "epoch": 0.9224701808387842,
+      "grad_norm": 0.8886311673810519,
+      "learning_rate": 3.151608698123232e-07,
+      "loss": 0.7856,
+      "step": 4795
+    },
+    {
+      "epoch": 0.9226625625240477,
+      "grad_norm": 0.8183441383325959,
+      "learning_rate": 3.1361079495281443e-07,
+      "loss": 0.7367,
+      "step": 4796
+    },
+    {
+      "epoch": 0.9228549442093112,
+      "grad_norm": 0.8893406573849837,
+      "learning_rate": 3.1206448068310635e-07,
+      "loss": 0.8377,
+      "step": 4797
+    },
+    {
+      "epoch": 0.9230473258945748,
+      "grad_norm": 0.859425890975106,
+      "learning_rate": 3.1052192760353316e-07,
+      "loss": 0.8261,
+      "step": 4798
+    },
+    {
+      "epoch": 0.9232397075798384,
+      "grad_norm": 0.8463427767781472,
+      "learning_rate": 3.0898313631296586e-07,
+      "loss": 0.7512,
+      "step": 4799
+    },
+    {
+      "epoch": 0.923432089265102,
+      "grad_norm": 0.8547123632599055,
+      "learning_rate": 3.0744810740881646e-07,
+      "loss": 0.8138,
+      "step": 4800
+    },
+    {
+      "epoch": 0.9236244709503655,
+      "grad_norm": 0.8425174545003049,
+      "learning_rate": 3.0591684148703617e-07,
+      "loss": 0.7241,
+      "step": 4801
+    },
+    {
+      "epoch": 0.9238168526356291,
+      "grad_norm": 0.9034891764572637,
+      "learning_rate": 3.043893391421149e-07,
+      "loss": 0.8201,
+      "step": 4802
+    },
+    {
+      "epoch": 0.9240092343208927,
+      "grad_norm": 0.8967451971263096,
+      "learning_rate": 3.0286560096708275e-07,
+      "loss": 0.7883,
+      "step": 4803
+    },
+    {
+      "epoch": 0.9242016160061562,
+      "grad_norm": 0.8825149240056192,
+      "learning_rate": 3.013456275535054e-07,
+      "loss": 0.6918,
+      "step": 4804
+    },
+    {
+      "epoch": 0.9243939976914197,
+      "grad_norm": 0.8827651441503394,
+      "learning_rate": 2.998294194914897e-07,
+      "loss": 0.7638,
+      "step": 4805
+    },
+    {
+      "epoch": 0.9245863793766833,
+      "grad_norm": 0.8889990844805005,
+      "learning_rate": 2.983169773696815e-07,
+      "loss": 0.8757,
+      "step": 4806
+    },
+    {
+      "epoch": 0.9247787610619469,
+      "grad_norm": 0.8128852635913514,
+      "learning_rate": 2.968083017752599e-07,
+      "loss": 0.7368,
+      "step": 4807
+    },
+    {
+      "epoch": 0.9249711427472105,
+      "grad_norm": 0.8038793734661661,
+      "learning_rate": 2.953033932939464e-07,
+      "loss": 0.8186,
+      "step": 4808
+    },
+    {
+      "epoch": 0.9251635244324741,
+      "grad_norm": 0.9521858755345305,
+      "learning_rate": 2.938022525099982e-07,
+      "loss": 0.8065,
+      "step": 4809
+    },
+    {
+      "epoch": 0.9253559061177375,
+      "grad_norm": 0.9360783675714339,
+      "learning_rate": 2.9230488000621003e-07,
+      "loss": 0.7652,
+      "step": 4810
+    },
+    {
+      "epoch": 0.9255482878030011,
+      "grad_norm": 0.6123503680502191,
+      "learning_rate": 2.908112763639137e-07,
+      "loss": 0.2992,
+      "step": 4811
+    },
+    {
+      "epoch": 0.9257406694882647,
+      "grad_norm": 0.9373251229310688,
+      "learning_rate": 2.8932144216297643e-07,
+      "loss": 0.7864,
+      "step": 4812
+    },
+    {
+      "epoch": 0.9259330511735283,
+      "grad_norm": 0.938750526524058,
+      "learning_rate": 2.878353779818044e-07,
+      "loss": 0.8054,
+      "step": 4813
+    },
+    {
+      "epoch": 0.9261254328587919,
+      "grad_norm": 0.8579729254298669,
+      "learning_rate": 2.863530843973372e-07,
+      "loss": 0.7536,
+      "step": 4814
+    },
+    {
+      "epoch": 0.9263178145440554,
+      "grad_norm": 0.8818069673404321,
+      "learning_rate": 2.848745619850546e-07,
+      "loss": 0.8184,
+      "step": 4815
+    },
+    {
+      "epoch": 0.926510196229319,
+      "grad_norm": 0.8387366579343567,
+      "learning_rate": 2.833998113189662e-07,
+      "loss": 0.8094,
+      "step": 4816
+    },
+    {
+      "epoch": 0.9267025779145825,
+      "grad_norm": 0.9436774426369852,
+      "learning_rate": 2.8192883297162634e-07,
+      "loss": 0.8114,
+      "step": 4817
+    },
+    {
+      "epoch": 0.9268949595998461,
+      "grad_norm": 0.8677320934324854,
+      "learning_rate": 2.804616275141148e-07,
+      "loss": 0.7869,
+      "step": 4818
+    },
+    {
+      "epoch": 0.9270873412851096,
+      "grad_norm": 0.9178973280879288,
+      "learning_rate": 2.7899819551605256e-07,
+      "loss": 0.8019,
+      "step": 4819
+    },
+    {
+      "epoch": 0.9272797229703732,
+      "grad_norm": 0.9556209099659226,
+      "learning_rate": 2.7753853754559634e-07,
+      "loss": 0.8257,
+      "step": 4820
+    },
+    {
+      "epoch": 0.9274721046556368,
+      "grad_norm": 0.9205030610605037,
+      "learning_rate": 2.760826541694328e-07,
+      "loss": 0.752,
+      "step": 4821
+    },
+    {
+      "epoch": 0.9276644863409004,
+      "grad_norm": 0.6206076714358484,
+      "learning_rate": 2.746305459527876e-07,
+      "loss": 0.3499,
+      "step": 4822
+    },
+    {
+      "epoch": 0.927856868026164,
+      "grad_norm": 0.8868088545845979,
+      "learning_rate": 2.7318221345941865e-07,
+      "loss": 0.7762,
+      "step": 4823
+    },
+    {
+      "epoch": 0.9280492497114274,
+      "grad_norm": 0.8845025977598391,
+      "learning_rate": 2.717376572516184e-07,
+      "loss": 0.7197,
+      "step": 4824
+    },
+    {
+      "epoch": 0.928241631396691,
+      "grad_norm": 0.8942930633513787,
+      "learning_rate": 2.7029687789021377e-07,
+      "loss": 0.8015,
+      "step": 4825
+    },
+    {
+      "epoch": 0.9284340130819546,
+      "grad_norm": 0.8712318353002876,
+      "learning_rate": 2.688598759345651e-07,
+      "loss": 0.7808,
+      "step": 4826
+    },
+    {
+      "epoch": 0.9286263947672182,
+      "grad_norm": 0.9138866742399898,
+      "learning_rate": 2.67426651942565e-07,
+      "loss": 0.8105,
+      "step": 4827
+    },
+    {
+      "epoch": 0.9288187764524817,
+      "grad_norm": 0.8400705763184247,
+      "learning_rate": 2.659972064706406e-07,
+      "loss": 0.7588,
+      "step": 4828
+    },
+    {
+      "epoch": 0.9290111581377453,
+      "grad_norm": 0.8794261137135524,
+      "learning_rate": 2.645715400737536e-07,
+      "loss": 0.8156,
+      "step": 4829
+    },
+    {
+      "epoch": 0.9292035398230089,
+      "grad_norm": 0.5918879714681047,
+      "learning_rate": 2.631496533053934e-07,
+      "loss": 0.3219,
+      "step": 4830
+    },
+    {
+      "epoch": 0.9293959215082724,
+      "grad_norm": 0.6278052861145778,
+      "learning_rate": 2.6173154671758847e-07,
+      "loss": 0.3214,
+      "step": 4831
+    },
+    {
+      "epoch": 0.929588303193536,
+      "grad_norm": 1.01987370536754,
+      "learning_rate": 2.603172208608962e-07,
+      "loss": 0.8289,
+      "step": 4832
+    },
+    {
+      "epoch": 0.9297806848787995,
+      "grad_norm": 0.9491417185242587,
+      "learning_rate": 2.589066762844039e-07,
+      "loss": 0.7772,
+      "step": 4833
+    },
+    {
+      "epoch": 0.9299730665640631,
+      "grad_norm": 0.7988237079545745,
+      "learning_rate": 2.57499913535737e-07,
+      "loss": 0.754,
+      "step": 4834
+    },
+    {
+      "epoch": 0.9301654482493267,
+      "grad_norm": 0.8619654180382375,
+      "learning_rate": 2.5609693316104745e-07,
+      "loss": 0.7905,
+      "step": 4835
+    },
+    {
+      "epoch": 0.9303578299345903,
+      "grad_norm": 0.6111040129878275,
+      "learning_rate": 2.5469773570502063e-07,
+      "loss": 0.3737,
+      "step": 4836
+    },
+    {
+      "epoch": 0.9305502116198537,
+      "grad_norm": 0.8403928006788132,
+      "learning_rate": 2.5330232171087433e-07,
+      "loss": 0.7955,
+      "step": 4837
+    },
+    {
+      "epoch": 0.9307425933051173,
+      "grad_norm": 0.9901995484351438,
+      "learning_rate": 2.51910691720354e-07,
+      "loss": 0.7742,
+      "step": 4838
+    },
+    {
+      "epoch": 0.9309349749903809,
+      "grad_norm": 0.8881611065128514,
+      "learning_rate": 2.5052284627374077e-07,
+      "loss": 0.8151,
+      "step": 4839
+    },
+    {
+      "epoch": 0.9311273566756445,
+      "grad_norm": 0.8131763648753264,
+      "learning_rate": 2.491387859098426e-07,
+      "loss": 0.7351,
+      "step": 4840
+    },
+    {
+      "epoch": 0.9313197383609081,
+      "grad_norm": 0.9188000542599124,
+      "learning_rate": 2.477585111659997e-07,
+      "loss": 0.7973,
+      "step": 4841
+    },
+    {
+      "epoch": 0.9315121200461716,
+      "grad_norm": 1.0147903910065712,
+      "learning_rate": 2.463820225780811e-07,
+      "loss": 0.7853,
+      "step": 4842
+    },
+    {
+      "epoch": 0.9317045017314352,
+      "grad_norm": 0.9553225961085251,
+      "learning_rate": 2.4500932068049046e-07,
+      "loss": 0.7529,
+      "step": 4843
+    },
+    {
+      "epoch": 0.9318968834166987,
+      "grad_norm": 0.8902502922849566,
+      "learning_rate": 2.4364040600615477e-07,
+      "loss": 0.7405,
+      "step": 4844
+    },
+    {
+      "epoch": 0.9320892651019623,
+      "grad_norm": 0.8698181895721878,
+      "learning_rate": 2.422752790865346e-07,
+      "loss": 0.785,
+      "step": 4845
+    },
+    {
+      "epoch": 0.9322816467872258,
+      "grad_norm": 0.8267127736439158,
+      "learning_rate": 2.409139404516203e-07,
+      "loss": 0.77,
+      "step": 4846
+    },
+    {
+      "epoch": 0.9324740284724894,
+      "grad_norm": 0.955100882979766,
+      "learning_rate": 2.3955639062992696e-07,
+      "loss": 0.7898,
+      "step": 4847
+    },
+    {
+      "epoch": 0.932666410157753,
+      "grad_norm": 0.8906077426766357,
+      "learning_rate": 2.3820263014850741e-07,
+      "loss": 0.7978,
+      "step": 4848
+    },
+    {
+      "epoch": 0.9328587918430166,
+      "grad_norm": 0.8808140977686274,
+      "learning_rate": 2.3685265953293345e-07,
+      "loss": 0.7721,
+      "step": 4849
+    },
+    {
+      "epoch": 0.9330511735282802,
+      "grad_norm": 0.9124775641182888,
+      "learning_rate": 2.3550647930731362e-07,
+      "loss": 0.7987,
+      "step": 4850
+    },
+    {
+      "epoch": 0.9332435552135436,
+      "grad_norm": 0.8989928148037185,
+      "learning_rate": 2.3416408999427876e-07,
+      "loss": 0.8255,
+      "step": 4851
+    },
+    {
+      "epoch": 0.9334359368988072,
+      "grad_norm": 0.8869092157235258,
+      "learning_rate": 2.3282549211499307e-07,
+      "loss": 0.7718,
+      "step": 4852
+    },
+    {
+      "epoch": 0.9336283185840708,
+      "grad_norm": 0.7874376361970187,
+      "learning_rate": 2.3149068618914417e-07,
+      "loss": 0.7916,
+      "step": 4853
+    },
+    {
+      "epoch": 0.9338207002693344,
+      "grad_norm": 0.6524764801948902,
+      "learning_rate": 2.3015967273494867e-07,
+      "loss": 0.3404,
+      "step": 4854
+    },
+    {
+      "epoch": 0.9340130819545979,
+      "grad_norm": 0.8253660541763395,
+      "learning_rate": 2.2883245226915652e-07,
+      "loss": 0.797,
+      "step": 4855
+    },
+    {
+      "epoch": 0.9342054636398615,
+      "grad_norm": 0.9128441429518032,
+      "learning_rate": 2.2750902530703667e-07,
+      "loss": 0.7341,
+      "step": 4856
+    },
+    {
+      "epoch": 0.934397845325125,
+      "grad_norm": 0.8316204352591283,
+      "learning_rate": 2.2618939236238924e-07,
+      "loss": 0.8078,
+      "step": 4857
+    },
+    {
+      "epoch": 0.9345902270103886,
+      "grad_norm": 0.866766020154056,
+      "learning_rate": 2.2487355394754328e-07,
+      "loss": 0.7784,
+      "step": 4858
+    },
+    {
+      "epoch": 0.9347826086956522,
+      "grad_norm": 0.6201279737800356,
+      "learning_rate": 2.2356151057334908e-07,
+      "loss": 0.3292,
+      "step": 4859
+    },
+    {
+      "epoch": 0.9349749903809157,
+      "grad_norm": 0.8188170512938197,
+      "learning_rate": 2.2225326274919135e-07,
+      "loss": 0.7808,
+      "step": 4860
+    },
+    {
+      "epoch": 0.9351673720661793,
+      "grad_norm": 1.0044405093555755,
+      "learning_rate": 2.209488109829727e-07,
+      "loss": 0.7382,
+      "step": 4861
+    },
+    {
+      "epoch": 0.9353597537514429,
+      "grad_norm": 0.9090532538817557,
+      "learning_rate": 2.196481557811303e-07,
+      "loss": 0.8298,
+      "step": 4862
+    },
+    {
+      "epoch": 0.9355521354367065,
+      "grad_norm": 0.8454600353074784,
+      "learning_rate": 2.1835129764861907e-07,
+      "loss": 0.7976,
+      "step": 4863
+    },
+    {
+      "epoch": 0.93574451712197,
+      "grad_norm": 0.8808833040552445,
+      "learning_rate": 2.1705823708892737e-07,
+      "loss": 0.7615,
+      "step": 4864
+    },
+    {
+      "epoch": 0.9359368988072335,
+      "grad_norm": 0.9972209781986677,
+      "learning_rate": 2.1576897460406477e-07,
+      "loss": 0.7605,
+      "step": 4865
+    },
+    {
+      "epoch": 0.9361292804924971,
+      "grad_norm": 0.9405049889031838,
+      "learning_rate": 2.144835106945664e-07,
+      "loss": 0.8357,
+      "step": 4866
+    },
+    {
+      "epoch": 0.9363216621777607,
+      "grad_norm": 0.848505143620877,
+      "learning_rate": 2.1320184585949532e-07,
+      "loss": 0.8004,
+      "step": 4867
+    },
+    {
+      "epoch": 0.9365140438630243,
+      "grad_norm": 0.6041430571627274,
+      "learning_rate": 2.119239805964357e-07,
+      "loss": 0.3246,
+      "step": 4868
+    },
+    {
+      "epoch": 0.9367064255482878,
+      "grad_norm": 0.861623941278494,
+      "learning_rate": 2.106499154015018e-07,
+      "loss": 0.7626,
+      "step": 4869
+    },
+    {
+      "epoch": 0.9368988072335513,
+      "grad_norm": 0.9382389519090573,
+      "learning_rate": 2.0937965076932576e-07,
+      "loss": 0.797,
+      "step": 4870
+    },
+    {
+      "epoch": 0.9370911889188149,
+      "grad_norm": 0.8236491302938501,
+      "learning_rate": 2.0811318719307194e-07,
+      "loss": 0.7438,
+      "step": 4871
+    },
+    {
+      "epoch": 0.9372835706040785,
+      "grad_norm": 0.8394561677645125,
+      "learning_rate": 2.0685052516442373e-07,
+      "loss": 0.7539,
+      "step": 4872
+    },
+    {
+      "epoch": 0.9374759522893421,
+      "grad_norm": 0.888226436902324,
+      "learning_rate": 2.0559166517358787e-07,
+      "loss": 0.7906,
+      "step": 4873
+    },
+    {
+      "epoch": 0.9376683339746056,
+      "grad_norm": 0.924010137781753,
+      "learning_rate": 2.0433660770930009e-07,
+      "loss": 0.7861,
+      "step": 4874
+    },
+    {
+      "epoch": 0.9378607156598692,
+      "grad_norm": 0.9528445624892278,
+      "learning_rate": 2.0308535325881616e-07,
+      "loss": 0.7951,
+      "step": 4875
+    },
+    {
+      "epoch": 0.9380530973451328,
+      "grad_norm": 0.8889928181027642,
+      "learning_rate": 2.0183790230791532e-07,
+      "loss": 0.7668,
+      "step": 4876
+    },
+    {
+      "epoch": 0.9382454790303963,
+      "grad_norm": 0.9121440564077309,
+      "learning_rate": 2.0059425534090128e-07,
+      "loss": 0.8174,
+      "step": 4877
+    },
+    {
+      "epoch": 0.9384378607156598,
+      "grad_norm": 0.9261955508269615,
+      "learning_rate": 1.9935441284059998e-07,
+      "loss": 0.77,
+      "step": 4878
+    },
+    {
+      "epoch": 0.9386302424009234,
+      "grad_norm": 0.9691447426747313,
+      "learning_rate": 1.981183752883631e-07,
+      "loss": 0.7821,
+      "step": 4879
+    },
+    {
+      "epoch": 0.938822624086187,
+      "grad_norm": 0.9359508578453345,
+      "learning_rate": 1.9688614316406006e-07,
+      "loss": 0.7665,
+      "step": 4880
+    },
+    {
+      "epoch": 0.9390150057714506,
+      "grad_norm": 0.8514648532482322,
+      "learning_rate": 1.9565771694608937e-07,
+      "loss": 0.7818,
+      "step": 4881
+    },
+    {
+      "epoch": 0.9392073874567142,
+      "grad_norm": 0.8514710658117453,
+      "learning_rate": 1.9443309711136393e-07,
+      "loss": 0.7493,
+      "step": 4882
+    },
+    {
+      "epoch": 0.9393997691419776,
+      "grad_norm": 0.9479495537817948,
+      "learning_rate": 1.9321228413532788e-07,
+      "loss": 0.7674,
+      "step": 4883
+    },
+    {
+      "epoch": 0.9395921508272412,
+      "grad_norm": 0.6499195705650884,
+      "learning_rate": 1.9199527849194098e-07,
+      "loss": 0.3348,
+      "step": 4884
+    },
+    {
+      "epoch": 0.9397845325125048,
+      "grad_norm": 0.8569642165216056,
+      "learning_rate": 1.907820806536842e-07,
+      "loss": 0.7306,
+      "step": 4885
+    },
+    {
+      "epoch": 0.9399769141977684,
+      "grad_norm": 0.891715348266224,
+      "learning_rate": 1.895726910915663e-07,
+      "loss": 0.8071,
+      "step": 4886
+    },
+    {
+      "epoch": 0.9401692958830319,
+      "grad_norm": 0.5836074602570475,
+      "learning_rate": 1.883671102751128e-07,
+      "loss": 0.3315,
+      "step": 4887
+    },
+    {
+      "epoch": 0.9403616775682955,
+      "grad_norm": 0.808727221322969,
+      "learning_rate": 1.8716533867237153e-07,
+      "loss": 0.8286,
+      "step": 4888
+    },
+    {
+      "epoch": 0.9405540592535591,
+      "grad_norm": 0.5992564256877513,
+      "learning_rate": 1.859673767499115e-07,
+      "loss": 0.3098,
+      "step": 4889
+    },
+    {
+      "epoch": 0.9407464409388226,
+      "grad_norm": 0.9506623599218275,
+      "learning_rate": 1.847732249728218e-07,
+      "loss": 0.8132,
+      "step": 4890
+    },
+    {
+      "epoch": 0.9409388226240862,
+      "grad_norm": 0.8157604623251068,
+      "learning_rate": 1.83582883804716e-07,
+      "loss": 0.7406,
+      "step": 4891
+    },
+    {
+      "epoch": 0.9411312043093497,
+      "grad_norm": 0.6232175602025368,
+      "learning_rate": 1.8239635370772223e-07,
+      "loss": 0.3288,
+      "step": 4892
+    },
+    {
+      "epoch": 0.9413235859946133,
+      "grad_norm": 0.9169568087579383,
+      "learning_rate": 1.8121363514249534e-07,
+      "loss": 0.7558,
+      "step": 4893
+    },
+    {
+      "epoch": 0.9415159676798769,
+      "grad_norm": 0.8915623821249192,
+      "learning_rate": 1.8003472856820469e-07,
+      "loss": 0.7736,
+      "step": 4894
+    },
+    {
+      "epoch": 0.9417083493651405,
+      "grad_norm": 0.8052355392987692,
+      "learning_rate": 1.7885963444254528e-07,
+      "loss": 0.7651,
+      "step": 4895
+    },
+    {
+      "epoch": 0.941900731050404,
+      "grad_norm": 0.9021537670251812,
+      "learning_rate": 1.7768835322172552e-07,
+      "loss": 0.8251,
+      "step": 4896
+    },
+    {
+      "epoch": 0.9420931127356675,
+      "grad_norm": 0.8850033217133381,
+      "learning_rate": 1.7652088536048052e-07,
+      "loss": 0.7774,
+      "step": 4897
+    },
+    {
+      "epoch": 0.9422854944209311,
+      "grad_norm": 0.9709082107235305,
+      "learning_rate": 1.7535723131206106e-07,
+      "loss": 0.8263,
+      "step": 4898
+    },
+    {
+      "epoch": 0.9424778761061947,
+      "grad_norm": 0.8369615007055918,
+      "learning_rate": 1.7419739152823468e-07,
+      "loss": 0.8095,
+      "step": 4899
+    },
+    {
+      "epoch": 0.9426702577914583,
+      "grad_norm": 0.9366144753317366,
+      "learning_rate": 1.7304136645929448e-07,
+      "loss": 0.7618,
+      "step": 4900
+    },
+    {
+      "epoch": 0.9428626394767218,
+      "grad_norm": 0.9069323817526947,
+      "learning_rate": 1.7188915655404814e-07,
+      "loss": 0.7263,
+      "step": 4901
+    },
+    {
+      "epoch": 0.9430550211619854,
+      "grad_norm": 0.920816028879399,
+      "learning_rate": 1.707407622598223e-07,
+      "loss": 0.7926,
+      "step": 4902
+    },
+    {
+      "epoch": 0.943247402847249,
+      "grad_norm": 0.9100787830775261,
+      "learning_rate": 1.695961840224636e-07,
+      "loss": 0.7675,
+      "step": 4903
+    },
+    {
+      "epoch": 0.9434397845325125,
+      "grad_norm": 0.9523947572650433,
+      "learning_rate": 1.6845542228633772e-07,
+      "loss": 0.7662,
+      "step": 4904
+    },
+    {
+      "epoch": 0.943632166217776,
+      "grad_norm": 0.8689114539017175,
+      "learning_rate": 1.6731847749432705e-07,
+      "loss": 0.7993,
+      "step": 4905
+    },
+    {
+      "epoch": 0.9438245479030396,
+      "grad_norm": 0.8164573823612615,
+      "learning_rate": 1.6618535008783075e-07,
+      "loss": 0.8079,
+      "step": 4906
+    },
+    {
+      "epoch": 0.9440169295883032,
+      "grad_norm": 0.8981890505749551,
+      "learning_rate": 1.6505604050677249e-07,
+      "loss": 0.794,
+      "step": 4907
+    },
+    {
+      "epoch": 0.9442093112735668,
+      "grad_norm": 0.8292654439994808,
+      "learning_rate": 1.6393054918958373e-07,
+      "loss": 0.7363,
+      "step": 4908
+    },
+    {
+      "epoch": 0.9444016929588304,
+      "grad_norm": 0.8764618538060301,
+      "learning_rate": 1.6280887657322276e-07,
+      "loss": 0.7784,
+      "step": 4909
+    },
+    {
+      "epoch": 0.9445940746440938,
+      "grad_norm": 0.869522586365224,
+      "learning_rate": 1.616910230931612e-07,
+      "loss": 0.8066,
+      "step": 4910
+    },
+    {
+      "epoch": 0.9447864563293574,
+      "grad_norm": 0.926868886357471,
+      "learning_rate": 1.6057698918338526e-07,
+      "loss": 0.7728,
+      "step": 4911
+    },
+    {
+      "epoch": 0.944978838014621,
+      "grad_norm": 0.9279226814261857,
+      "learning_rate": 1.5946677527640563e-07,
+      "loss": 0.7615,
+      "step": 4912
+    },
+    {
+      "epoch": 0.9451712196998846,
+      "grad_norm": 0.5998612572908801,
+      "learning_rate": 1.5836038180324198e-07,
+      "loss": 0.3561,
+      "step": 4913
+    },
+    {
+      "epoch": 0.9453636013851482,
+      "grad_norm": 0.8744528097126124,
+      "learning_rate": 1.5725780919343624e-07,
+      "loss": 0.8666,
+      "step": 4914
+    },
+    {
+      "epoch": 0.9455559830704117,
+      "grad_norm": 0.5881370299117944,
+      "learning_rate": 1.561590578750438e-07,
+      "loss": 0.3543,
+      "step": 4915
+    },
+    {
+      "epoch": 0.9457483647556753,
+      "grad_norm": 0.9023008320254763,
+      "learning_rate": 1.55064128274639e-07,
+      "loss": 0.7936,
+      "step": 4916
+    },
+    {
+      "epoch": 0.9459407464409388,
+      "grad_norm": 0.8697897485565973,
+      "learning_rate": 1.5397302081731069e-07,
+      "loss": 0.7954,
+      "step": 4917
+    },
+    {
+      "epoch": 0.9461331281262024,
+      "grad_norm": 0.8553563046351385,
+      "learning_rate": 1.5288573592666445e-07,
+      "loss": 0.7779,
+      "step": 4918
+    },
+    {
+      "epoch": 0.9463255098114659,
+      "grad_norm": 0.8978048668090456,
+      "learning_rate": 1.518022740248215e-07,
+      "loss": 0.7561,
+      "step": 4919
+    },
+    {
+      "epoch": 0.9465178914967295,
+      "grad_norm": 0.9846456733861073,
+      "learning_rate": 1.5072263553241872e-07,
+      "loss": 0.7902,
+      "step": 4920
+    },
+    {
+      "epoch": 0.9467102731819931,
+      "grad_norm": 1.015667564830783,
+      "learning_rate": 1.4964682086861082e-07,
+      "loss": 0.8353,
+      "step": 4921
+    },
+    {
+      "epoch": 0.9469026548672567,
+      "grad_norm": 0.9175524513815987,
+      "learning_rate": 1.4857483045106258e-07,
+      "loss": 0.8643,
+      "step": 4922
+    },
+    {
+      "epoch": 0.9470950365525203,
+      "grad_norm": 0.8826021156015369,
+      "learning_rate": 1.475066646959611e-07,
+      "loss": 0.7561,
+      "step": 4923
+    },
+    {
+      "epoch": 0.9472874182377837,
+      "grad_norm": 0.8279248388051379,
+      "learning_rate": 1.4644232401800352e-07,
+      "loss": 0.7905,
+      "step": 4924
+    },
+    {
+      "epoch": 0.9474797999230473,
+      "grad_norm": 0.845235591321461,
+      "learning_rate": 1.4538180883040264e-07,
+      "loss": 0.777,
+      "step": 4925
+    },
+    {
+      "epoch": 0.9476721816083109,
+      "grad_norm": 0.9248288960217717,
+      "learning_rate": 1.4432511954488915e-07,
+      "loss": 0.8154,
+      "step": 4926
+    },
+    {
+      "epoch": 0.9478645632935745,
+      "grad_norm": 0.9395792244285891,
+      "learning_rate": 1.4327225657170485e-07,
+      "loss": 0.7636,
+      "step": 4927
+    },
+    {
+      "epoch": 0.948056944978838,
+      "grad_norm": 0.9536079410280032,
+      "learning_rate": 1.4222322031960723e-07,
+      "loss": 0.747,
+      "step": 4928
+    },
+    {
+      "epoch": 0.9482493266641016,
+      "grad_norm": 0.965566943548885,
+      "learning_rate": 1.411780111958694e-07,
+      "loss": 0.8105,
+      "step": 4929
+    },
+    {
+      "epoch": 0.9484417083493651,
+      "grad_norm": 0.9414245492043505,
+      "learning_rate": 1.4013662960627562e-07,
+      "loss": 0.7686,
+      "step": 4930
+    },
+    {
+      "epoch": 0.9486340900346287,
+      "grad_norm": 0.8404729561526275,
+      "learning_rate": 1.3909907595512806e-07,
+      "loss": 0.7535,
+      "step": 4931
+    },
+    {
+      "epoch": 0.9488264717198923,
+      "grad_norm": 0.9689733853693594,
+      "learning_rate": 1.3806535064524006e-07,
+      "loss": 0.819,
+      "step": 4932
+    },
+    {
+      "epoch": 0.9490188534051558,
+      "grad_norm": 0.9937615947171681,
+      "learning_rate": 1.3703545407793951e-07,
+      "loss": 0.8538,
+      "step": 4933
+    },
+    {
+      "epoch": 0.9492112350904194,
+      "grad_norm": 0.813554239387415,
+      "learning_rate": 1.360093866530665e-07,
+      "loss": 0.777,
+      "step": 4934
+    },
+    {
+      "epoch": 0.949403616775683,
+      "grad_norm": 0.8679024629926823,
+      "learning_rate": 1.34987148768978e-07,
+      "loss": 0.8002,
+      "step": 4935
+    },
+    {
+      "epoch": 0.9495959984609466,
+      "grad_norm": 0.8458585788695936,
+      "learning_rate": 1.3396874082253986e-07,
+      "loss": 0.7933,
+      "step": 4936
+    },
+    {
+      "epoch": 0.94978838014621,
+      "grad_norm": 0.8911179679504925,
+      "learning_rate": 1.3295416320913357e-07,
+      "loss": 0.7965,
+      "step": 4937
+    },
+    {
+      "epoch": 0.9499807618314736,
+      "grad_norm": 0.8711921858885162,
+      "learning_rate": 1.3194341632265518e-07,
+      "loss": 0.838,
+      "step": 4938
+    },
+    {
+      "epoch": 0.9501731435167372,
+      "grad_norm": 0.8440768301080301,
+      "learning_rate": 1.3093650055550855e-07,
+      "loss": 0.7961,
+      "step": 4939
+    },
+    {
+      "epoch": 0.9503655252020008,
+      "grad_norm": 0.9231679331917274,
+      "learning_rate": 1.2993341629861432e-07,
+      "loss": 0.7725,
+      "step": 4940
+    },
+    {
+      "epoch": 0.9505579068872644,
+      "grad_norm": 0.8326257259229253,
+      "learning_rate": 1.2893416394140323e-07,
+      "loss": 0.7993,
+      "step": 4941
+    },
+    {
+      "epoch": 0.9507502885725279,
+      "grad_norm": 0.8578472794194749,
+      "learning_rate": 1.279387438718216e-07,
+      "loss": 0.7805,
+      "step": 4942
+    },
+    {
+      "epoch": 0.9509426702577914,
+      "grad_norm": 0.8935435462298107,
+      "learning_rate": 1.269471564763247e-07,
+      "loss": 0.7756,
+      "step": 4943
+    },
+    {
+      "epoch": 0.951135051943055,
+      "grad_norm": 0.8793614891846315,
+      "learning_rate": 1.2595940213988024e-07,
+      "loss": 0.7961,
+      "step": 4944
+    },
+    {
+      "epoch": 0.9513274336283186,
+      "grad_norm": 0.8196582689739138,
+      "learning_rate": 1.2497548124597026e-07,
+      "loss": 0.782,
+      "step": 4945
+    },
+    {
+      "epoch": 0.9515198153135821,
+      "grad_norm": 0.9209999791769309,
+      "learning_rate": 1.2399539417658368e-07,
+      "loss": 0.7748,
+      "step": 4946
+    },
+    {
+      "epoch": 0.9517121969988457,
+      "grad_norm": 0.8587458610614921,
+      "learning_rate": 1.2301914131222726e-07,
+      "loss": 0.7529,
+      "step": 4947
+    },
+    {
+      "epoch": 0.9519045786841093,
+      "grad_norm": 0.939458148619404,
+      "learning_rate": 1.2204672303191335e-07,
+      "loss": 0.7308,
+      "step": 4948
+    },
+    {
+      "epoch": 0.9520969603693729,
+      "grad_norm": 0.9601922244888209,
+      "learning_rate": 1.2107813971317106e-07,
+      "loss": 0.7581,
+      "step": 4949
+    },
+    {
+      "epoch": 0.9522893420546364,
+      "grad_norm": 0.8576990515236218,
+      "learning_rate": 1.201133917320363e-07,
+      "loss": 0.8186,
+      "step": 4950
+    },
+    {
+      "epoch": 0.9524817237398999,
+      "grad_norm": 0.8653518783044469,
+      "learning_rate": 1.1915247946305498e-07,
+      "loss": 0.7641,
+      "step": 4951
+    },
+    {
+      "epoch": 0.9526741054251635,
+      "grad_norm": 0.9285702054591581,
+      "learning_rate": 1.1819540327929092e-07,
+      "loss": 0.7934,
+      "step": 4952
+    },
+    {
+      "epoch": 0.9528664871104271,
+      "grad_norm": 0.9324876256423614,
+      "learning_rate": 1.1724216355231022e-07,
+      "loss": 0.7898,
+      "step": 4953
+    },
+    {
+      "epoch": 0.9530588687956907,
+      "grad_norm": 0.9017764953597712,
+      "learning_rate": 1.1629276065219575e-07,
+      "loss": 0.8247,
+      "step": 4954
+    },
+    {
+      "epoch": 0.9532512504809542,
+      "grad_norm": 0.86601877417719,
+      "learning_rate": 1.1534719494753821e-07,
+      "loss": 0.7908,
+      "step": 4955
+    },
+    {
+      "epoch": 0.9534436321662177,
+      "grad_norm": 0.8591376600052064,
+      "learning_rate": 1.144054668054373e-07,
+      "loss": 0.7957,
+      "step": 4956
+    },
+    {
+      "epoch": 0.9536360138514813,
+      "grad_norm": 0.851792102407842,
+      "learning_rate": 1.1346757659150498e-07,
+      "loss": 0.7814,
+      "step": 4957
+    },
+    {
+      "epoch": 0.9538283955367449,
+      "grad_norm": 0.8453075024525436,
+      "learning_rate": 1.1253352466986334e-07,
+      "loss": 0.7932,
+      "step": 4958
+    },
+    {
+      "epoch": 0.9540207772220085,
+      "grad_norm": 0.7934927442461515,
+      "learning_rate": 1.116033114031434e-07,
+      "loss": 0.7636,
+      "step": 4959
+    },
+    {
+      "epoch": 0.954213158907272,
+      "grad_norm": 0.8721999182473338,
+      "learning_rate": 1.1067693715248406e-07,
+      "loss": 0.8073,
+      "step": 4960
+    },
+    {
+      "epoch": 0.9544055405925356,
+      "grad_norm": 0.8858415459342593,
+      "learning_rate": 1.0975440227753764e-07,
+      "loss": 0.784,
+      "step": 4961
+    },
+    {
+      "epoch": 0.9545979222777992,
+      "grad_norm": 0.923105486047575,
+      "learning_rate": 1.0883570713646318e-07,
+      "loss": 0.7892,
+      "step": 4962
+    },
+    {
+      "epoch": 0.9547903039630627,
+      "grad_norm": 0.9185515104903339,
+      "learning_rate": 1.0792085208593095e-07,
+      "loss": 0.7383,
+      "step": 4963
+    },
+    {
+      "epoch": 0.9549826856483262,
+      "grad_norm": 0.9393382795685349,
+      "learning_rate": 1.0700983748111792e-07,
+      "loss": 0.7823,
+      "step": 4964
+    },
+    {
+      "epoch": 0.9551750673335898,
+      "grad_norm": 0.8339212523585086,
+      "learning_rate": 1.061026636757101e-07,
+      "loss": 0.7903,
+      "step": 4965
+    },
+    {
+      "epoch": 0.9553674490188534,
+      "grad_norm": 0.8569362855827806,
+      "learning_rate": 1.0519933102190682e-07,
+      "loss": 0.8029,
+      "step": 4966
+    },
+    {
+      "epoch": 0.955559830704117,
+      "grad_norm": 0.9369321207446158,
+      "learning_rate": 1.0429983987041092e-07,
+      "loss": 0.7985,
+      "step": 4967
+    },
+    {
+      "epoch": 0.9557522123893806,
+      "grad_norm": 0.9479783213421658,
+      "learning_rate": 1.0340419057043527e-07,
+      "loss": 0.7995,
+      "step": 4968
+    },
+    {
+      "epoch": 0.955944594074644,
+      "grad_norm": 0.8015651312501346,
+      "learning_rate": 1.0251238346970393e-07,
+      "loss": 0.792,
+      "step": 4969
+    },
+    {
+      "epoch": 0.9561369757599076,
+      "grad_norm": 0.8949691064370727,
+      "learning_rate": 1.0162441891444441e-07,
+      "loss": 0.7714,
+      "step": 4970
+    },
+    {
+      "epoch": 0.9563293574451712,
+      "grad_norm": 0.8922305542397799,
+      "learning_rate": 1.007402972493976e-07,
+      "loss": 0.7453,
+      "step": 4971
+    },
+    {
+      "epoch": 0.9565217391304348,
+      "grad_norm": 0.628066508988732,
+      "learning_rate": 9.986001881780783e-08,
+      "loss": 0.3071,
+      "step": 4972
+    },
+    {
+      "epoch": 0.9567141208156984,
+      "grad_norm": 0.8164177865162465,
+      "learning_rate": 9.898358396143171e-08,
+      "loss": 0.7605,
+      "step": 4973
+    },
+    {
+      "epoch": 0.9569065025009619,
+      "grad_norm": 0.9467597017050703,
+      "learning_rate": 9.811099302052928e-08,
+      "loss": 0.7629,
+      "step": 4974
+    },
+    {
+      "epoch": 0.9570988841862255,
+      "grad_norm": 0.8185106869900445,
+      "learning_rate": 9.72422463338718e-08,
+      "loss": 0.7382,
+      "step": 4975
+    },
+    {
+      "epoch": 0.957291265871489,
+      "grad_norm": 0.9704648483888862,
+      "learning_rate": 9.637734423873612e-08,
+      "loss": 0.7317,
+      "step": 4976
+    },
+    {
+      "epoch": 0.9574836475567526,
+      "grad_norm": 0.9433878963441633,
+      "learning_rate": 9.55162870709081e-08,
+      "loss": 0.8044,
+      "step": 4977
+    },
+    {
+      "epoch": 0.9576760292420161,
+      "grad_norm": 0.5983605423854327,
+      "learning_rate": 9.465907516467698e-08,
+      "loss": 0.3306,
+      "step": 4978
+    },
+    {
+      "epoch": 0.9578684109272797,
+      "grad_norm": 0.8272197442671191,
+      "learning_rate": 9.380570885284546e-08,
+      "loss": 0.836,
+      "step": 4979
+    },
+    {
+      "epoch": 0.9580607926125433,
+      "grad_norm": 0.9603048987787275,
+      "learning_rate": 9.295618846671739e-08,
+      "loss": 0.8114,
+      "step": 4980
+    },
+    {
+      "epoch": 0.9582531742978069,
+      "grad_norm": 0.8836836594942532,
+      "learning_rate": 9.211051433610674e-08,
+      "loss": 0.7733,
+      "step": 4981
+    },
+    {
+      "epoch": 0.9584455559830705,
+      "grad_norm": 0.817814776276063,
+      "learning_rate": 9.126868678933198e-08,
+      "loss": 0.7621,
+      "step": 4982
+    },
+    {
+      "epoch": 0.9586379376683339,
+      "grad_norm": 0.8914591692965085,
+      "learning_rate": 9.04307061532217e-08,
+      "loss": 0.7855,
+      "step": 4983
+    },
+    {
+      "epoch": 0.9588303193535975,
+      "grad_norm": 0.902250012476356,
+      "learning_rate": 8.959657275310674e-08,
+      "loss": 0.8059,
+      "step": 4984
+    },
+    {
+      "epoch": 0.9590227010388611,
+      "grad_norm": 0.8307882704697991,
+      "learning_rate": 8.876628691282918e-08,
+      "loss": 0.7559,
+      "step": 4985
+    },
+    {
+      "epoch": 0.9592150827241247,
+      "grad_norm": 0.9616257845204815,
+      "learning_rate": 8.793984895473117e-08,
+      "loss": 0.7832,
+      "step": 4986
+    },
+    {
+      "epoch": 0.9594074644093882,
+      "grad_norm": 0.8785716143035189,
+      "learning_rate": 8.711725919966718e-08,
+      "loss": 0.7205,
+      "step": 4987
+    },
+    {
+      "epoch": 0.9595998460946518,
+      "grad_norm": 0.8784365762368678,
+      "learning_rate": 8.629851796699284e-08,
+      "loss": 0.7492,
+      "step": 4988
+    },
+    {
+      "epoch": 0.9597922277799154,
+      "grad_norm": 0.9440054070915694,
+      "learning_rate": 8.54836255745728e-08,
+      "loss": 0.742,
+      "step": 4989
+    },
+    {
+      "epoch": 0.9599846094651789,
+      "grad_norm": 0.977814224357969,
+      "learning_rate": 8.467258233877728e-08,
+      "loss": 0.8357,
+      "step": 4990
+    },
+    {
+      "epoch": 0.9601769911504425,
+      "grad_norm": 0.9427308805714572,
+      "learning_rate": 8.386538857447779e-08,
+      "loss": 0.7765,
+      "step": 4991
+    },
+    {
+      "epoch": 0.960369372835706,
+      "grad_norm": 0.8081139309136687,
+      "learning_rate": 8.306204459505807e-08,
+      "loss": 0.7645,
+      "step": 4992
+    },
+    {
+      "epoch": 0.9605617545209696,
+      "grad_norm": 0.8697368035913701,
+      "learning_rate": 8.226255071240308e-08,
+      "loss": 0.8046,
+      "step": 4993
+    },
+    {
+      "epoch": 0.9607541362062332,
+      "grad_norm": 0.8245374543277172,
+      "learning_rate": 8.146690723690342e-08,
+      "loss": 0.8162,
+      "step": 4994
+    },
+    {
+      "epoch": 0.9609465178914968,
+      "grad_norm": 0.8153974973038656,
+      "learning_rate": 8.067511447745535e-08,
+      "loss": 0.7399,
+      "step": 4995
+    },
+    {
+      "epoch": 0.9611388995767602,
+      "grad_norm": 0.9202290769547941,
+      "learning_rate": 7.988717274146074e-08,
+      "loss": 0.7981,
+      "step": 4996
+    },
+    {
+      "epoch": 0.9613312812620238,
+      "grad_norm": 0.9243371285188634,
+      "learning_rate": 7.910308233482488e-08,
+      "loss": 0.8062,
+      "step": 4997
+    },
+    {
+      "epoch": 0.9615236629472874,
+      "grad_norm": 0.8623529075158792,
+      "learning_rate": 7.832284356195764e-08,
+      "loss": 0.7502,
+      "step": 4998
+    },
+    {
+      "epoch": 0.961716044632551,
+      "grad_norm": 0.8594177826795781,
+      "learning_rate": 7.754645672577776e-08,
+      "loss": 0.7155,
+      "step": 4999
+    },
+    {
+      "epoch": 0.9619084263178146,
+      "grad_norm": 0.8811822717710149,
+      "learning_rate": 7.677392212770196e-08,
+      "loss": 0.772,
+      "step": 5000
+    },
+    {
+      "epoch": 0.9621008080030781,
+      "grad_norm": 0.8499887428350209,
+      "learning_rate": 7.600524006765808e-08,
+      "loss": 0.7249,
+      "step": 5001
+    },
+    {
+      "epoch": 0.9622931896883417,
+      "grad_norm": 0.9331682912687291,
+      "learning_rate": 7.524041084407185e-08,
+      "loss": 0.791,
+      "step": 5002
+    },
+    {
+      "epoch": 0.9624855713736052,
+      "grad_norm": 0.8785651743272676,
+      "learning_rate": 7.447943475387797e-08,
+      "loss": 0.7629,
+      "step": 5003
+    },
+    {
+      "epoch": 0.9626779530588688,
+      "grad_norm": 0.9637904855783187,
+      "learning_rate": 7.372231209251346e-08,
+      "loss": 0.8056,
+      "step": 5004
+    },
+    {
+      "epoch": 0.9628703347441323,
+      "grad_norm": 0.9074317399035945,
+      "learning_rate": 7.296904315391873e-08,
+      "loss": 0.7877,
+      "step": 5005
+    },
+    {
+      "epoch": 0.9630627164293959,
+      "grad_norm": 0.8501322989007611,
+      "learning_rate": 7.221962823053874e-08,
+      "loss": 0.7436,
+      "step": 5006
+    },
+    {
+      "epoch": 0.9632550981146595,
+      "grad_norm": 0.8820905369368142,
+      "learning_rate": 7.147406761332298e-08,
+      "loss": 0.7673,
+      "step": 5007
+    },
+    {
+      "epoch": 0.9634474797999231,
+      "grad_norm": 0.8561472629472588,
+      "learning_rate": 7.073236159172325e-08,
+      "loss": 0.7717,
+      "step": 5008
+    },
+    {
+      "epoch": 0.9636398614851867,
+      "grad_norm": 1.05868745398584,
+      "learning_rate": 6.999451045369587e-08,
+      "loss": 0.8281,
+      "step": 5009
+    },
+    {
+      "epoch": 0.9638322431704501,
+      "grad_norm": 0.9098749084420481,
+      "learning_rate": 6.926051448569948e-08,
+      "loss": 0.7379,
+      "step": 5010
+    },
+    {
+      "epoch": 0.9640246248557137,
+      "grad_norm": 0.9107432790139404,
+      "learning_rate": 6.853037397269724e-08,
+      "loss": 0.8058,
+      "step": 5011
+    },
+    {
+      "epoch": 0.9642170065409773,
+      "grad_norm": 0.9047215005163584,
+      "learning_rate": 6.78040891981524e-08,
+      "loss": 0.836,
+      "step": 5012
+    },
+    {
+      "epoch": 0.9644093882262409,
+      "grad_norm": 0.8654360284259628,
+      "learning_rate": 6.70816604440383e-08,
+      "loss": 0.8097,
+      "step": 5013
+    },
+    {
+      "epoch": 0.9646017699115044,
+      "grad_norm": 0.8161395144339438,
+      "learning_rate": 6.63630879908217e-08,
+      "loss": 0.7414,
+      "step": 5014
+    },
+    {
+      "epoch": 0.964794151596768,
+      "grad_norm": 0.8953987742945317,
+      "learning_rate": 6.564837211748054e-08,
+      "loss": 0.8162,
+      "step": 5015
+    },
+    {
+      "epoch": 0.9649865332820315,
+      "grad_norm": 0.8794403949557966,
+      "learning_rate": 6.493751310149177e-08,
+      "loss": 0.8209,
+      "step": 5016
+    },
+    {
+      "epoch": 0.9651789149672951,
+      "grad_norm": 0.9035781833759973,
+      "learning_rate": 6.42305112188335e-08,
+      "loss": 0.7856,
+      "step": 5017
+    },
+    {
+      "epoch": 0.9653712966525587,
+      "grad_norm": 0.8635297799288215,
+      "learning_rate": 6.352736674398951e-08,
+      "loss": 0.7726,
+      "step": 5018
+    },
+    {
+      "epoch": 0.9655636783378222,
+      "grad_norm": 0.8628851089781387,
+      "learning_rate": 6.282807994994477e-08,
+      "loss": 0.8071,
+      "step": 5019
+    },
+    {
+      "epoch": 0.9657560600230858,
+      "grad_norm": 0.8287386676307013,
+      "learning_rate": 6.213265110818656e-08,
+      "loss": 0.7881,
+      "step": 5020
+    },
+    {
+      "epoch": 0.9659484417083494,
+      "grad_norm": 0.937380961299943,
+      "learning_rate": 6.144108048870335e-08,
+      "loss": 0.8146,
+      "step": 5021
+    },
+    {
+      "epoch": 0.966140823393613,
+      "grad_norm": 0.6580756853089993,
+      "learning_rate": 6.075336835998813e-08,
+      "loss": 0.3374,
+      "step": 5022
+    },
+    {
+      "epoch": 0.9663332050788765,
+      "grad_norm": 0.9593153927315138,
+      "learning_rate": 6.00695149890329e-08,
+      "loss": 0.7701,
+      "step": 5023
+    },
+    {
+      "epoch": 0.96652558676414,
+      "grad_norm": 0.9142215017865216,
+      "learning_rate": 5.938952064133419e-08,
+      "loss": 0.8188,
+      "step": 5024
+    },
+    {
+      "epoch": 0.9667179684494036,
+      "grad_norm": 0.9726437123603898,
+      "learning_rate": 5.871338558088857e-08,
+      "loss": 0.803,
+      "step": 5025
+    },
+    {
+      "epoch": 0.9669103501346672,
+      "grad_norm": 0.8243777205166778,
+      "learning_rate": 5.8041110070194976e-08,
+      "loss": 0.8287,
+      "step": 5026
+    },
+    {
+      "epoch": 0.9671027318199308,
+      "grad_norm": 0.9596609687041973,
+      "learning_rate": 5.7372694370254614e-08,
+      "loss": 0.7253,
+      "step": 5027
+    },
+    {
+      "epoch": 0.9672951135051943,
+      "grad_norm": 0.953784932772887,
+      "learning_rate": 5.67081387405688e-08,
+      "loss": 0.7874,
+      "step": 5028
+    },
+    {
+      "epoch": 0.9674874951904578,
+      "grad_norm": 0.899029097775021,
+      "learning_rate": 5.6047443439141146e-08,
+      "loss": 0.7948,
+      "step": 5029
+    },
+    {
+      "epoch": 0.9676798768757214,
+      "grad_norm": 0.8187182822974759,
+      "learning_rate": 5.539060872247537e-08,
+      "loss": 0.7827,
+      "step": 5030
+    },
+    {
+      "epoch": 0.967872258560985,
+      "grad_norm": 0.6012157989391343,
+      "learning_rate": 5.47376348455797e-08,
+      "loss": 0.335,
+      "step": 5031
+    },
+    {
+      "epoch": 0.9680646402462486,
+      "grad_norm": 0.9057319828176345,
+      "learning_rate": 5.408852206195914e-08,
+      "loss": 0.7373,
+      "step": 5032
+    },
+    {
+      "epoch": 0.9682570219315121,
+      "grad_norm": 0.572577362253172,
+      "learning_rate": 5.344327062362098e-08,
+      "loss": 0.2964,
+      "step": 5033
+    },
+    {
+      "epoch": 0.9684494036167757,
+      "grad_norm": 0.87354526520116,
+      "learning_rate": 5.2801880781075954e-08,
+      "loss": 0.7736,
+      "step": 5034
+    },
+    {
+      "epoch": 0.9686417853020393,
+      "grad_norm": 0.8699311012773862,
+      "learning_rate": 5.216435278333376e-08,
+      "loss": 0.8043,
+      "step": 5035
+    },
+    {
+      "epoch": 0.9688341669873028,
+      "grad_norm": 0.9536227464594137,
+      "learning_rate": 5.153068687790197e-08,
+      "loss": 0.8057,
+      "step": 5036
+    },
+    {
+      "epoch": 0.9690265486725663,
+      "grad_norm": 0.9248896613289947,
+      "learning_rate": 5.0900883310794903e-08,
+      "loss": 0.775,
+      "step": 5037
+    },
+    {
+      "epoch": 0.9692189303578299,
+      "grad_norm": 0.9361002478709752,
+      "learning_rate": 5.0274942326521414e-08,
+      "loss": 0.7362,
+      "step": 5038
+    },
+    {
+      "epoch": 0.9694113120430935,
+      "grad_norm": 0.9785408062622237,
+      "learning_rate": 4.9652864168096e-08,
+      "loss": 0.7396,
+      "step": 5039
+    },
+    {
+      "epoch": 0.9696036937283571,
+      "grad_norm": 0.9130226259196453,
+      "learning_rate": 4.9034649077027706e-08,
+      "loss": 0.7918,
+      "step": 5040
+    },
+    {
+      "epoch": 0.9697960754136207,
+      "grad_norm": 0.9659872005071725,
+      "learning_rate": 4.84202972933312e-08,
+      "loss": 0.8514,
+      "step": 5041
+    },
+    {
+      "epoch": 0.9699884570988841,
+      "grad_norm": 0.9301281548014028,
+      "learning_rate": 4.7809809055517906e-08,
+      "loss": 0.8327,
+      "step": 5042
+    },
+    {
+      "epoch": 0.9701808387841477,
+      "grad_norm": 0.8859018607257724,
+      "learning_rate": 4.720318460060047e-08,
+      "loss": 0.7698,
+      "step": 5043
+    },
+    {
+      "epoch": 0.9703732204694113,
+      "grad_norm": 0.8979826167778971,
+      "learning_rate": 4.6600424164091606e-08,
+      "loss": 0.7396,
+      "step": 5044
+    },
+    {
+      "epoch": 0.9705656021546749,
+      "grad_norm": 0.8298627254817487,
+      "learning_rate": 4.6001527980004125e-08,
+      "loss": 0.7758,
+      "step": 5045
+    },
+    {
+      "epoch": 0.9707579838399384,
+      "grad_norm": 0.8204019384822165,
+      "learning_rate": 4.54064962808487e-08,
+      "loss": 0.7808,
+      "step": 5046
+    },
+    {
+      "epoch": 0.970950365525202,
+      "grad_norm": 0.890443333285141,
+      "learning_rate": 4.4815329297639434e-08,
+      "loss": 0.7613,
+      "step": 5047
+    },
+    {
+      "epoch": 0.9711427472104656,
+      "grad_norm": 0.9752770861686931,
+      "learning_rate": 4.422802725988606e-08,
+      "loss": 0.772,
+      "step": 5048
+    },
+    {
+      "epoch": 0.9713351288957291,
+      "grad_norm": 0.8863174606545253,
+      "learning_rate": 4.364459039559843e-08,
+      "loss": 0.8138,
+      "step": 5049
+    },
+    {
+      "epoch": 0.9715275105809927,
+      "grad_norm": 0.8666446667383483,
+      "learning_rate": 4.3065018931289784e-08,
+      "loss": 0.7854,
+      "step": 5050
+    },
+    {
+      "epoch": 0.9717198922662562,
+      "grad_norm": 0.8422756522684829,
+      "learning_rate": 4.248931309196791e-08,
+      "loss": 0.7541,
+      "step": 5051
+    },
+    {
+      "epoch": 0.9719122739515198,
+      "grad_norm": 0.8117138039534505,
+      "learning_rate": 4.1917473101140696e-08,
+      "loss": 0.7517,
+      "step": 5052
+    },
+    {
+      "epoch": 0.9721046556367834,
+      "grad_norm": 0.8547799747916205,
+      "learning_rate": 4.134949918081832e-08,
+      "loss": 0.7696,
+      "step": 5053
+    },
+    {
+      "epoch": 0.972297037322047,
+      "grad_norm": 0.907309349337728,
+      "learning_rate": 4.0785391551506626e-08,
+      "loss": 0.7561,
+      "step": 5054
+    },
+    {
+      "epoch": 0.9724894190073105,
+      "grad_norm": 0.8809941712391312,
+      "learning_rate": 4.022515043221154e-08,
+      "loss": 0.754,
+      "step": 5055
+    },
+    {
+      "epoch": 0.972681800692574,
+      "grad_norm": 0.9714434508605624,
+      "learning_rate": 3.966877604043795e-08,
+      "loss": 0.7982,
+      "step": 5056
+    },
+    {
+      "epoch": 0.9728741823778376,
+      "grad_norm": 0.8825467363927891,
+      "learning_rate": 3.9116268592189755e-08,
+      "loss": 0.8174,
+      "step": 5057
+    },
+    {
+      "epoch": 0.9730665640631012,
+      "grad_norm": 0.8328622180372023,
+      "learning_rate": 3.8567628301969806e-08,
+      "loss": 0.8018,
+      "step": 5058
+    },
+    {
+      "epoch": 0.9732589457483648,
+      "grad_norm": 0.8783191612293341,
+      "learning_rate": 3.802285538277772e-08,
+      "loss": 0.7381,
+      "step": 5059
+    },
+    {
+      "epoch": 0.9734513274336283,
+      "grad_norm": 0.8804359679394334,
+      "learning_rate": 3.748195004611543e-08,
+      "loss": 0.7646,
+      "step": 5060
+    },
+    {
+      "epoch": 0.9736437091188919,
+      "grad_norm": 0.8526590356124019,
+      "learning_rate": 3.69449125019794e-08,
+      "loss": 0.7419,
+      "step": 5061
+    },
+    {
+      "epoch": 0.9738360908041555,
+      "grad_norm": 0.9797992137717609,
+      "learning_rate": 3.6411742958866184e-08,
+      "loss": 0.8006,
+      "step": 5062
+    },
+    {
+      "epoch": 0.974028472489419,
+      "grad_norm": 0.5974263174858027,
+      "learning_rate": 3.588244162377019e-08,
+      "loss": 0.3229,
+      "step": 5063
+    },
+    {
+      "epoch": 0.9742208541746825,
+      "grad_norm": 0.6141648309491364,
+      "learning_rate": 3.5357008702185945e-08,
+      "loss": 0.3965,
+      "step": 5064
+    },
+    {
+      "epoch": 0.9744132358599461,
+      "grad_norm": 0.8676057550566596,
+      "learning_rate": 3.483544439810249e-08,
+      "loss": 0.7411,
+      "step": 5065
+    },
+    {
+      "epoch": 0.9746056175452097,
+      "grad_norm": 0.8375763388020031,
+      "learning_rate": 3.4317748914011187e-08,
+      "loss": 0.7327,
+      "step": 5066
+    },
+    {
+      "epoch": 0.9747979992304733,
+      "grad_norm": 0.8638920156737381,
+      "learning_rate": 3.3803922450897917e-08,
+      "loss": 0.785,
+      "step": 5067
+    },
+    {
+      "epoch": 0.9749903809157369,
+      "grad_norm": 0.9998482378682032,
+      "learning_rate": 3.329396520824757e-08,
+      "loss": 0.7276,
+      "step": 5068
+    },
+    {
+      "epoch": 0.9751827626010003,
+      "grad_norm": 0.907482124088877,
+      "learning_rate": 3.2787877384045095e-08,
+      "loss": 0.804,
+      "step": 5069
+    },
+    {
+      "epoch": 0.9753751442862639,
+      "grad_norm": 0.9059220270949934,
+      "learning_rate": 3.228565917476889e-08,
+      "loss": 0.8109,
+      "step": 5070
+    },
+    {
+      "epoch": 0.9755675259715275,
+      "grad_norm": 0.8415613186551881,
+      "learning_rate": 3.178731077539743e-08,
+      "loss": 0.802,
+      "step": 5071
+    },
+    {
+      "epoch": 0.9757599076567911,
+      "grad_norm": 0.9753740587425018,
+      "learning_rate": 3.129283237940928e-08,
+      "loss": 0.7649,
+      "step": 5072
+    },
+    {
+      "epoch": 0.9759522893420547,
+      "grad_norm": 0.893284826746859,
+      "learning_rate": 3.080222417877421e-08,
+      "loss": 0.7947,
+      "step": 5073
+    },
+    {
+      "epoch": 0.9761446710273182,
+      "grad_norm": 0.9860124107413918,
+      "learning_rate": 3.031548636396764e-08,
+      "loss": 0.8026,
+      "step": 5074
+    },
+    {
+      "epoch": 0.9763370527125818,
+      "grad_norm": 0.9343208189357735,
+      "learning_rate": 2.983261912395397e-08,
+      "loss": 0.7757,
+      "step": 5075
+    },
+    {
+      "epoch": 0.9765294343978453,
+      "grad_norm": 0.8427782327675692,
+      "learning_rate": 2.9353622646199898e-08,
+      "loss": 0.7619,
+      "step": 5076
+    },
+    {
+      "epoch": 0.9767218160831089,
+      "grad_norm": 0.9616740889215901,
+      "learning_rate": 2.8878497116671124e-08,
+      "loss": 0.8359,
+      "step": 5077
+    },
+    {
+      "epoch": 0.9769141977683724,
+      "grad_norm": 0.8619847516648865,
+      "learning_rate": 2.8407242719823424e-08,
+      "loss": 0.8071,
+      "step": 5078
+    },
+    {
+      "epoch": 0.977106579453636,
+      "grad_norm": 0.9084951041546999,
+      "learning_rate": 2.7939859638617118e-08,
+      "loss": 0.7603,
+      "step": 5079
+    },
+    {
+      "epoch": 0.9772989611388996,
+      "grad_norm": 0.8981799388904826,
+      "learning_rate": 2.7476348054504832e-08,
+      "loss": 0.7826,
+      "step": 5080
+    },
+    {
+      "epoch": 0.9774913428241632,
+      "grad_norm": 0.928737548946432,
+      "learning_rate": 2.7016708147439285e-08,
+      "loss": 0.8141,
+      "step": 5081
+    },
+    {
+      "epoch": 0.9776837245094268,
+      "grad_norm": 0.5788064436488111,
+      "learning_rate": 2.6560940095866626e-08,
+      "loss": 0.3335,
+      "step": 5082
+    },
+    {
+      "epoch": 0.9778761061946902,
+      "grad_norm": 0.994323466866682,
+      "learning_rate": 2.6109044076733092e-08,
+      "loss": 0.8069,
+      "step": 5083
+    },
+    {
+      "epoch": 0.9780684878799538,
+      "grad_norm": 0.8856211903228551,
+      "learning_rate": 2.5661020265479452e-08,
+      "loss": 0.7745,
+      "step": 5084
+    },
+    {
+      "epoch": 0.9782608695652174,
+      "grad_norm": 0.9317452943052241,
+      "learning_rate": 2.5216868836043242e-08,
+      "loss": 0.8019,
+      "step": 5085
+    },
+    {
+      "epoch": 0.978453251250481,
+      "grad_norm": 0.9068789824614535,
+      "learning_rate": 2.4776589960862074e-08,
+      "loss": 0.7813,
+      "step": 5086
+    },
+    {
+      "epoch": 0.9786456329357445,
+      "grad_norm": 0.821675649606347,
+      "learning_rate": 2.434018381086589e-08,
+      "loss": 0.7573,
+      "step": 5087
+    },
+    {
+      "epoch": 0.9788380146210081,
+      "grad_norm": 0.8876277624720081,
+      "learning_rate": 2.3907650555481387e-08,
+      "loss": 0.7711,
+      "step": 5088
+    },
+    {
+      "epoch": 0.9790303963062716,
+      "grad_norm": 0.9046175534664419,
+      "learning_rate": 2.3478990362634235e-08,
+      "loss": 0.8103,
+      "step": 5089
+    },
+    {
+      "epoch": 0.9792227779915352,
+      "grad_norm": 0.9593543619826154,
+      "learning_rate": 2.3054203398743537e-08,
+      "loss": 0.8424,
+      "step": 5090
+    },
+    {
+      "epoch": 0.9794151596767988,
+      "grad_norm": 0.9617619785180253,
+      "learning_rate": 2.263328982872959e-08,
+      "loss": 0.8002,
+      "step": 5091
+    },
+    {
+      "epoch": 0.9796075413620623,
+      "grad_norm": 0.9215935412375413,
+      "learning_rate": 2.221624981600168e-08,
+      "loss": 0.8269,
+      "step": 5092
+    },
+    {
+      "epoch": 0.9797999230473259,
+      "grad_norm": 0.9278645399349733,
+      "learning_rate": 2.1803083522471402e-08,
+      "loss": 0.8055,
+      "step": 5093
+    },
+    {
+      "epoch": 0.9799923047325895,
+      "grad_norm": 0.8755559003066465,
+      "learning_rate": 2.1393791108542672e-08,
+      "loss": 0.7939,
+      "step": 5094
+    },
+    {
+      "epoch": 0.9801846864178531,
+      "grad_norm": 0.84689856063013,
+      "learning_rate": 2.098837273311838e-08,
+      "loss": 0.8005,
+      "step": 5095
+    },
+    {
+      "epoch": 0.9803770681031165,
+      "grad_norm": 0.8463199014579492,
+      "learning_rate": 2.058682855359595e-08,
+      "loss": 0.763,
+      "step": 5096
+    },
+    {
+      "epoch": 0.9805694497883801,
+      "grad_norm": 0.8240811583731107,
+      "learning_rate": 2.0189158725867353e-08,
+      "loss": 0.7603,
+      "step": 5097
+    },
+    {
+      "epoch": 0.9807618314736437,
+      "grad_norm": 0.8388161661761047,
+      "learning_rate": 1.979536340432131e-08,
+      "loss": 0.6889,
+      "step": 5098
+    },
+    {
+      "epoch": 0.9809542131589073,
+      "grad_norm": 0.9382657294578419,
+      "learning_rate": 1.9405442741844415e-08,
+      "loss": 0.7544,
+      "step": 5099
+    },
+    {
+      "epoch": 0.9811465948441709,
+      "grad_norm": 0.8996092027173951,
+      "learning_rate": 1.9019396889816688e-08,
+      "loss": 0.7737,
+      "step": 5100
+    },
+    {
+      "epoch": 0.9813389765294344,
+      "grad_norm": 0.8962076183778311,
+      "learning_rate": 1.8637225998114904e-08,
+      "loss": 0.7682,
+      "step": 5101
+    },
+    {
+      "epoch": 0.981531358214698,
+      "grad_norm": 0.8972182432691683,
+      "learning_rate": 1.825893021510927e-08,
+      "loss": 0.8109,
+      "step": 5102
+    },
+    {
+      "epoch": 0.9817237398999615,
+      "grad_norm": 0.8742030962268361,
+      "learning_rate": 1.7884509687668972e-08,
+      "loss": 0.7856,
+      "step": 5103
+    },
+    {
+      "epoch": 0.9819161215852251,
+      "grad_norm": 0.896611962356364,
+      "learning_rate": 1.7513964561156617e-08,
+      "loss": 0.7875,
+      "step": 5104
+    },
+    {
+      "epoch": 0.9821085032704886,
+      "grad_norm": 0.8615557544506284,
+      "learning_rate": 1.714729497942935e-08,
+      "loss": 0.8239,
+      "step": 5105
+    },
+    {
+      "epoch": 0.9823008849557522,
+      "grad_norm": 0.9052065812717185,
+      "learning_rate": 1.6784501084843307e-08,
+      "loss": 0.7887,
+      "step": 5106
+    },
+    {
+      "epoch": 0.9824932666410158,
+      "grad_norm": 0.962963434755425,
+      "learning_rate": 1.6425583018244706e-08,
+      "loss": 0.7114,
+      "step": 5107
+    },
+    {
+      "epoch": 0.9826856483262794,
+      "grad_norm": 0.8236773117583839,
+      "learning_rate": 1.607054091897986e-08,
+      "loss": 0.7871,
+      "step": 5108
+    },
+    {
+      "epoch": 0.982878030011543,
+      "grad_norm": 0.8657285512596573,
+      "learning_rate": 1.57193749248874e-08,
+      "loss": 0.8162,
+      "step": 5109
+    },
+    {
+      "epoch": 0.9830704116968064,
+      "grad_norm": 0.8476783817734235,
+      "learning_rate": 1.537208517230271e-08,
+      "loss": 0.7954,
+      "step": 5110
+    },
+    {
+      "epoch": 0.98326279338207,
+      "grad_norm": 0.9141693038386359,
+      "learning_rate": 1.5028671796055715e-08,
+      "loss": 0.8166,
+      "step": 5111
+    },
+    {
+      "epoch": 0.9834551750673336,
+      "grad_norm": 0.8952321716933607,
+      "learning_rate": 1.4689134929470884e-08,
+      "loss": 0.7679,
+      "step": 5112
+    },
+    {
+      "epoch": 0.9836475567525972,
+      "grad_norm": 0.8566893469107928,
+      "learning_rate": 1.435347470436832e-08,
+      "loss": 0.7471,
+      "step": 5113
+    },
+    {
+      "epoch": 0.9838399384378607,
+      "grad_norm": 0.9598631401550527,
+      "learning_rate": 1.4021691251062675e-08,
+      "loss": 0.7234,
+      "step": 5114
+    },
+    {
+      "epoch": 0.9840323201231242,
+      "grad_norm": 0.8674652034594739,
+      "learning_rate": 1.3693784698363133e-08,
+      "loss": 0.7598,
+      "step": 5115
+    },
+    {
+      "epoch": 0.9842247018083878,
+      "grad_norm": 0.8959771641155367,
+      "learning_rate": 1.3369755173575639e-08,
+      "loss": 0.8325,
+      "step": 5116
+    },
+    {
+      "epoch": 0.9844170834936514,
+      "grad_norm": 0.8895807325587292,
+      "learning_rate": 1.3049602802498451e-08,
+      "loss": 0.8035,
+      "step": 5117
+    },
+    {
+      "epoch": 0.984609465178915,
+      "grad_norm": 0.6260198949327267,
+      "learning_rate": 1.273332770942659e-08,
+      "loss": 0.3346,
+      "step": 5118
+    },
+    {
+      "epoch": 0.9848018468641785,
+      "grad_norm": 0.9739607485608952,
+      "learning_rate": 1.2420930017148503e-08,
+      "loss": 0.8011,
+      "step": 5119
+    },
+    {
+      "epoch": 0.9849942285494421,
+      "grad_norm": 0.8803116568771583,
+      "learning_rate": 1.2112409846947171e-08,
+      "loss": 0.7881,
+      "step": 5120
+    },
+    {
+      "epoch": 0.9851866102347057,
+      "grad_norm": 0.8699080781293109,
+      "learning_rate": 1.1807767318602337e-08,
+      "loss": 0.7803,
+      "step": 5121
+    },
+    {
+      "epoch": 0.9853789919199692,
+      "grad_norm": 0.8810415946158836,
+      "learning_rate": 1.150700255038606e-08,
+      "loss": 0.7384,
+      "step": 5122
+    },
+    {
+      "epoch": 0.9855713736052328,
+      "grad_norm": 0.8443513068341612,
+      "learning_rate": 1.1210115659063825e-08,
+      "loss": 0.7215,
+      "step": 5123
+    },
+    {
+      "epoch": 0.9857637552904963,
+      "grad_norm": 0.9013926825768803,
+      "learning_rate": 1.0917106759900097e-08,
+      "loss": 0.7958,
+      "step": 5124
+    },
+    {
+      "epoch": 0.9859561369757599,
+      "grad_norm": 0.8886666473796435,
+      "learning_rate": 1.0627975966649439e-08,
+      "loss": 0.7975,
+      "step": 5125
+    },
+    {
+      "epoch": 0.9861485186610235,
+      "grad_norm": 0.8396190772526052,
+      "learning_rate": 1.034272339156206e-08,
+      "loss": 0.7649,
+      "step": 5126
+    },
+    {
+      "epoch": 0.9863409003462871,
+      "grad_norm": 0.8811224954163122,
+      "learning_rate": 1.0061349145383814e-08,
+      "loss": 0.8118,
+      "step": 5127
+    },
+    {
+      "epoch": 0.9865332820315506,
+      "grad_norm": 0.5813018443962166,
+      "learning_rate": 9.783853337353987e-09,
+      "loss": 0.3288,
+      "step": 5128
+    },
+    {
+      "epoch": 0.9867256637168141,
+      "grad_norm": 0.8487164354459856,
+      "learning_rate": 9.510236075205292e-09,
+      "loss": 0.7542,
+      "step": 5129
+    },
+    {
+      "epoch": 0.9869180454020777,
+      "grad_norm": 0.9028927202566389,
+      "learning_rate": 9.240497465164978e-09,
+      "loss": 0.8264,
+      "step": 5130
+    },
+    {
+      "epoch": 0.9871104270873413,
+      "grad_norm": 0.887642343935252,
+      "learning_rate": 8.974637611955939e-09,
+      "loss": 0.7814,
+      "step": 5131
+    },
+    {
+      "epoch": 0.9873028087726049,
+      "grad_norm": 0.923413043314393,
+      "learning_rate": 8.712656618793391e-09,
+      "loss": 0.7518,
+      "step": 5132
+    },
+    {
+      "epoch": 0.9874951904578684,
+      "grad_norm": 0.8991793558086696,
+      "learning_rate": 8.454554587388198e-09,
+      "loss": 0.8214,
+      "step": 5133
+    },
+    {
+      "epoch": 0.987687572143132,
+      "grad_norm": 0.8650295257856636,
+      "learning_rate": 8.200331617943535e-09,
+      "loss": 0.7952,
+      "step": 5134
+    },
+    {
+      "epoch": 0.9878799538283956,
+      "grad_norm": 0.5958822637939898,
+      "learning_rate": 7.949987809158232e-09,
+      "loss": 0.3342,
+      "step": 5135
+    },
+    {
+      "epoch": 0.9880723355136591,
+      "grad_norm": 0.860976216028565,
+      "learning_rate": 7.703523258223433e-09,
+      "loss": 0.7926,
+      "step": 5136
+    },
+    {
+      "epoch": 0.9882647171989226,
+      "grad_norm": 0.8206106489906109,
+      "learning_rate": 7.460938060825929e-09,
+      "loss": 0.8249,
+      "step": 5137
+    },
+    {
+      "epoch": 0.9884570988841862,
+      "grad_norm": 0.885376249045719,
+      "learning_rate": 7.222232311145938e-09,
+      "loss": 0.7898,
+      "step": 5138
+    },
+    {
+      "epoch": 0.9886494805694498,
+      "grad_norm": 0.9771606374743358,
+      "learning_rate": 6.987406101855998e-09,
+      "loss": 0.831,
+      "step": 5139
+    },
+    {
+      "epoch": 0.9888418622547134,
+      "grad_norm": 0.9196110033023569,
+      "learning_rate": 6.756459524125403e-09,
+      "loss": 0.8567,
+      "step": 5140
+    },
+    {
+      "epoch": 0.989034243939977,
+      "grad_norm": 0.923389311066732,
+      "learning_rate": 6.5293926676135434e-09,
+      "loss": 0.7693,
+      "step": 5141
+    },
+    {
+      "epoch": 0.9892266256252404,
+      "grad_norm": 0.8993640332765271,
+      "learning_rate": 6.306205620477679e-09,
+      "loss": 0.7773,
+      "step": 5142
+    },
+    {
+      "epoch": 0.989419007310504,
+      "grad_norm": 0.9070360965714515,
+      "learning_rate": 6.086898469365166e-09,
+      "loss": 0.8197,
+      "step": 5143
+    },
+    {
+      "epoch": 0.9896113889957676,
+      "grad_norm": 0.9205375731542135,
+      "learning_rate": 5.871471299419007e-09,
+      "loss": 0.7905,
+      "step": 5144
+    },
+    {
+      "epoch": 0.9898037706810312,
+      "grad_norm": 0.908183809648122,
+      "learning_rate": 5.6599241942767445e-09,
+      "loss": 0.7845,
+      "step": 5145
+    },
+    {
+      "epoch": 0.9899961523662947,
+      "grad_norm": 0.5993210135735106,
+      "learning_rate": 5.452257236066017e-09,
+      "loss": 0.3299,
+      "step": 5146
+    },
+    {
+      "epoch": 0.9901885340515583,
+      "grad_norm": 0.9086580978559785,
+      "learning_rate": 5.248470505412328e-09,
+      "loss": 0.8417,
+      "step": 5147
+    },
+    {
+      "epoch": 0.9903809157368219,
+      "grad_norm": 0.9399142343907193,
+      "learning_rate": 5.0485640814312844e-09,
+      "loss": 0.7858,
+      "step": 5148
+    },
+    {
+      "epoch": 0.9905732974220854,
+      "grad_norm": 0.8471834528567551,
+      "learning_rate": 4.8525380417330234e-09,
+      "loss": 0.7752,
+      "step": 5149
+    },
+    {
+      "epoch": 0.990765679107349,
+      "grad_norm": 0.9816601236318335,
+      "learning_rate": 4.660392462424446e-09,
+      "loss": 0.7796,
+      "step": 5150
+    },
+    {
+      "epoch": 0.9909580607926125,
+      "grad_norm": 0.9460640824721142,
+      "learning_rate": 4.472127418099215e-09,
+      "loss": 0.7843,
+      "step": 5151
+    },
+    {
+      "epoch": 0.9911504424778761,
+      "grad_norm": 0.8171744472976223,
+      "learning_rate": 4.287742981851084e-09,
+      "loss": 0.7609,
+      "step": 5152
+    },
+    {
+      "epoch": 0.9913428241631397,
+      "grad_norm": 0.8892466379823025,
+      "learning_rate": 4.1072392252639034e-09,
+      "loss": 0.7935,
+      "step": 5153
+    },
+    {
+      "epoch": 0.9915352058484033,
+      "grad_norm": 0.8440509255541729,
+      "learning_rate": 3.930616218414951e-09,
+      "loss": 0.831,
+      "step": 5154
+    },
+    {
+      "epoch": 0.9917275875336667,
+      "grad_norm": 0.912912744088773,
+      "learning_rate": 3.757874029874931e-09,
+      "loss": 0.8071,
+      "step": 5155
+    },
+    {
+      "epoch": 0.9919199692189303,
+      "grad_norm": 0.8440519754906952,
+      "learning_rate": 3.5890127267090844e-09,
+      "loss": 0.7418,
+      "step": 5156
+    },
+    {
+      "epoch": 0.9921123509041939,
+      "grad_norm": 0.8479844074400885,
+      "learning_rate": 3.424032374476083e-09,
+      "loss": 0.7983,
+      "step": 5157
+    },
+    {
+      "epoch": 0.9923047325894575,
+      "grad_norm": 0.8998714448238606,
+      "learning_rate": 3.2629330372246915e-09,
+      "loss": 0.7478,
+      "step": 5158
+    },
+    {
+      "epoch": 0.9924971142747211,
+      "grad_norm": 0.8656077726685144,
+      "learning_rate": 3.105714777501545e-09,
+      "loss": 0.779,
+      "step": 5159
+    },
+    {
+      "epoch": 0.9926894959599846,
+      "grad_norm": 0.9128941639500899,
+      "learning_rate": 2.9523776563422644e-09,
+      "loss": 0.7908,
+      "step": 5160
+    },
+    {
+      "epoch": 0.9928818776452482,
+      "grad_norm": 0.932356876253163,
+      "learning_rate": 2.802921733278119e-09,
+      "loss": 0.8002,
+      "step": 5161
+    },
+    {
+      "epoch": 0.9930742593305117,
+      "grad_norm": 0.9305915471744044,
+      "learning_rate": 2.657347066333804e-09,
+      "loss": 0.8143,
+      "step": 5162
+    },
+    {
+      "epoch": 0.9932666410157753,
+      "grad_norm": 0.9351556570843815,
+      "learning_rate": 2.5156537120263335e-09,
+      "loss": 0.8511,
+      "step": 5163
+    },
+    {
+      "epoch": 0.9934590227010388,
+      "grad_norm": 0.8712562718559272,
+      "learning_rate": 2.3778417253650376e-09,
+      "loss": 0.8114,
+      "step": 5164
+    },
+    {
+      "epoch": 0.9936514043863024,
+      "grad_norm": 0.9202816401201681,
+      "learning_rate": 2.2439111598537844e-09,
+      "loss": 0.8482,
+      "step": 5165
+    },
+    {
+      "epoch": 0.993843786071566,
+      "grad_norm": 0.8128708359775251,
+      "learning_rate": 2.113862067488759e-09,
+      "loss": 0.7757,
+      "step": 5166
+    },
+    {
+      "epoch": 0.9940361677568296,
+      "grad_norm": 0.890997257312928,
+      "learning_rate": 1.987694498760684e-09,
+      "loss": 0.7867,
+      "step": 5167
+    },
+    {
+      "epoch": 0.9942285494420932,
+      "grad_norm": 0.9259089578521759,
+      "learning_rate": 1.865408502650379e-09,
+      "loss": 0.7681,
+      "step": 5168
+    },
+    {
+      "epoch": 0.9944209311273566,
+      "grad_norm": 0.8105285287560157,
+      "learning_rate": 1.747004126635421e-09,
+      "loss": 0.7897,
+      "step": 5169
+    },
+    {
+      "epoch": 0.9946133128126202,
+      "grad_norm": 0.8426059695669351,
+      "learning_rate": 1.6324814166823744e-09,
+      "loss": 0.7943,
+      "step": 5170
+    },
+    {
+      "epoch": 0.9948056944978838,
+      "grad_norm": 0.9592766560768253,
+      "learning_rate": 1.5218404172545609e-09,
+      "loss": 0.7857,
+      "step": 5171
+    },
+    {
+      "epoch": 0.9949980761831474,
+      "grad_norm": 0.9337419979963489,
+      "learning_rate": 1.415081171305399e-09,
+      "loss": 0.8384,
+      "step": 5172
+    },
+    {
+      "epoch": 0.995190457868411,
+      "grad_norm": 0.8760197191419367,
+      "learning_rate": 1.3122037202828452e-09,
+      "loss": 0.8306,
+      "step": 5173
+    },
+    {
+      "epoch": 0.9953828395536745,
+      "grad_norm": 0.9268959483723297,
+      "learning_rate": 1.2132081041282829e-09,
+      "loss": 0.7812,
+      "step": 5174
+    },
+    {
+      "epoch": 0.995575221238938,
+      "grad_norm": 0.9453343525806948,
+      "learning_rate": 1.1180943612754124e-09,
+      "loss": 0.8065,
+      "step": 5175
+    },
+    {
+      "epoch": 0.9957676029242016,
+      "grad_norm": 0.9197186401862288,
+      "learning_rate": 1.026862528649142e-09,
+      "loss": 0.7424,
+      "step": 5176
+    },
+    {
+      "epoch": 0.9959599846094652,
+      "grad_norm": 0.905246758421038,
+      "learning_rate": 9.39512641668916e-10,
+      "loss": 0.7709,
+      "step": 5177
+    },
+    {
+      "epoch": 0.9961523662947287,
+      "grad_norm": 0.8223553603677267,
+      "learning_rate": 8.560447342487177e-10,
+      "loss": 0.7749,
+      "step": 5178
+    },
+    {
+      "epoch": 0.9963447479799923,
+      "grad_norm": 0.8034172101473165,
+      "learning_rate": 7.764588387915161e-10,
+      "loss": 0.708,
+      "step": 5179
+    },
+    {
+      "epoch": 0.9965371296652559,
+      "grad_norm": 0.9978369594306293,
+      "learning_rate": 7.007549861970387e-10,
+      "loss": 0.8394,
+      "step": 5180
+    },
+    {
+      "epoch": 0.9967295113505195,
+      "grad_norm": 0.8357362904954584,
+      "learning_rate": 6.289332058551089e-10,
+      "loss": 0.799,
+      "step": 5181
+    },
+    {
+      "epoch": 0.996921893035783,
+      "grad_norm": 0.9164566332012781,
+      "learning_rate": 5.609935256500887e-10,
+      "loss": 0.7458,
+      "step": 5182
+    },
+    {
+      "epoch": 0.9971142747210465,
+      "grad_norm": 0.8720573910962935,
+      "learning_rate": 4.969359719586563e-10,
+      "loss": 0.7719,
+      "step": 5183
+    },
+    {
+      "epoch": 0.9973066564063101,
+      "grad_norm": 0.8605228789033751,
+      "learning_rate": 4.3676056964869764e-10,
+      "loss": 0.8067,
+      "step": 5184
+    },
+    {
+      "epoch": 0.9974990380915737,
+      "grad_norm": 0.8131195996189433,
+      "learning_rate": 3.804673420837457e-10,
+      "loss": 0.7898,
+      "step": 5185
+    },
+    {
+      "epoch": 0.9976914197768373,
+      "grad_norm": 0.7569054121075874,
+      "learning_rate": 3.2805631111743064e-10,
+      "loss": 0.7348,
+      "step": 5186
+    },
+    {
+      "epoch": 0.9978838014621008,
+      "grad_norm": 0.8784725562659086,
+      "learning_rate": 2.795274971001405e-10,
+      "loss": 0.7882,
+      "step": 5187
+    },
+    {
+      "epoch": 0.9980761831473643,
+      "grad_norm": 0.853051904268405,
+      "learning_rate": 2.3488091886902933e-10,
+      "loss": 0.8182,
+      "step": 5188
+    },
+    {
+      "epoch": 0.9982685648326279,
+      "grad_norm": 0.8604978660613666,
+      "learning_rate": 1.941165937602296e-10,
+      "loss": 0.8666,
+      "step": 5189
+    },
+    {
+      "epoch": 0.9984609465178915,
+      "grad_norm": 0.8798054692613018,
+      "learning_rate": 1.5723453759886042e-10,
+      "loss": 0.7529,
+      "step": 5190
+    },
+    {
+      "epoch": 0.9986533282031551,
+      "grad_norm": 0.8877956102959825,
+      "learning_rate": 1.2423476470346808e-10,
+      "loss": 0.7941,
+      "step": 5191
+    },
+    {
+      "epoch": 0.9988457098884186,
+      "grad_norm": 0.9009270728191574,
+      "learning_rate": 9.511728788602625e-11,
+      "loss": 0.8269,
+      "step": 5192
+    },
+    {
+      "epoch": 0.9990380915736822,
+      "grad_norm": 0.956901751836482,
+      "learning_rate": 6.988211845082582e-11,
+      "loss": 0.8527,
+      "step": 5193
+    },
+    {
+      "epoch": 0.9992304732589458,
+      "grad_norm": 0.9271535246995155,
+      "learning_rate": 4.852926619447473e-11,
+      "loss": 0.8291,
+      "step": 5194
+    },
+    {
+      "epoch": 0.9994228549442093,
+      "grad_norm": 0.9664668558994455,
+      "learning_rate": 3.105873940811854e-11,
+      "loss": 0.7585,
+      "step": 5195
+    },
+    {
+      "epoch": 0.9996152366294728,
+      "grad_norm": 0.8586546380829696,
+      "learning_rate": 1.7470544874109706e-11,
+      "loss": 0.7774,
+      "step": 5196
+    },
+    {
+      "epoch": 0.9998076183147364,
+      "grad_norm": 1.1514245282646858,
+      "learning_rate": 7.764687866007592e-12,
+      "loss": 0.532,
+      "step": 5197
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.7100707359385027,
+      "learning_rate": 1.9411721552398123e-12,
+      "loss": 0.4757,
+      "step": 5198
+    },
+    {
+      "epoch": 1.0,
+      "step": 5198,
+      "total_flos": 6736564352974848.0,
+      "train_loss": 0.8072307536399287,
+      "train_runtime": 70907.0142,
+      "train_samples_per_second": 9.383,
+      "train_steps_per_second": 0.073
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 5198,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 50000,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 6736564352974848.0,
+  "train_batch_size": 8,
+  "trial_name": null,
+  "trial_params": null
+}