diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,12747 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 4.989690721649485,
+  "eval_steps": 500,
+  "global_step": 1815,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0027491408934707906,
+      "grad_norm": 6.006607135034925,
+      "learning_rate": 4.395604395604396e-07,
+      "loss": 1.1009,
+      "step": 1
+    },
+    {
+      "epoch": 0.005498281786941581,
+      "grad_norm": 6.018597667128207,
+      "learning_rate": 8.791208791208792e-07,
+      "loss": 1.1005,
+      "step": 2
+    },
+    {
+      "epoch": 0.008247422680412371,
+      "grad_norm": 6.073361462181881,
+      "learning_rate": 1.3186813186813187e-06,
+      "loss": 1.1055,
+      "step": 3
+    },
+    {
+      "epoch": 0.010996563573883162,
+      "grad_norm": 5.961879860389406,
+      "learning_rate": 1.7582417582417585e-06,
+      "loss": 1.1099,
+      "step": 4
+    },
+    {
+      "epoch": 0.013745704467353952,
+      "grad_norm": 5.617774190916827,
+      "learning_rate": 2.197802197802198e-06,
+      "loss": 1.081,
+      "step": 5
+    },
+    {
+      "epoch": 0.016494845360824743,
+      "grad_norm": 4.531649230749474,
+      "learning_rate": 2.6373626373626375e-06,
+      "loss": 1.0487,
+      "step": 6
+    },
+    {
+      "epoch": 0.019243986254295534,
+      "grad_norm": 4.217358609239622,
+      "learning_rate": 3.0769230769230774e-06,
+      "loss": 1.0556,
+      "step": 7
+    },
+    {
+      "epoch": 0.021993127147766325,
+      "grad_norm": 2.3655606096533908,
+      "learning_rate": 3.516483516483517e-06,
+      "loss": 1.0049,
+      "step": 8
+    },
+    {
+      "epoch": 0.024742268041237112,
+      "grad_norm": 2.0449739951596695,
+      "learning_rate": 3.9560439560439565e-06,
+      "loss": 0.9803,
+      "step": 9
+    },
+    {
+      "epoch": 0.027491408934707903,
+      "grad_norm": 2.8045358100308406,
+      "learning_rate": 4.395604395604396e-06,
+      "loss": 0.9589,
+      "step": 10
+    },
+    {
+      "epoch": 0.030240549828178694,
+      "grad_norm": 3.662777544001548,
+      "learning_rate": 4.8351648351648355e-06,
+      "loss": 0.9677,
+      "step": 11
+    },
+    {
+      "epoch": 0.032989690721649485,
+      "grad_norm": 3.5905401528260374,
+      "learning_rate": 5.274725274725275e-06,
+      "loss": 0.96,
+      "step": 12
+    },
+    {
+      "epoch": 0.03573883161512027,
+      "grad_norm": 3.3209223305008964,
+      "learning_rate": 5.7142857142857145e-06,
+      "loss": 0.9481,
+      "step": 13
+    },
+    {
+      "epoch": 0.03848797250859107,
+      "grad_norm": 2.3880367341332343,
+      "learning_rate": 6.153846153846155e-06,
+      "loss": 0.9057,
+      "step": 14
+    },
+    {
+      "epoch": 0.041237113402061855,
+      "grad_norm": 2.269345734711135,
+      "learning_rate": 6.5934065934065935e-06,
+      "loss": 0.9048,
+      "step": 15
+    },
+    {
+      "epoch": 0.04398625429553265,
+      "grad_norm": 1.7351295951376182,
+      "learning_rate": 7.032967032967034e-06,
+      "loss": 0.8869,
+      "step": 16
+    },
+    {
+      "epoch": 0.04673539518900344,
+      "grad_norm": 1.5080912194950085,
+      "learning_rate": 7.472527472527473e-06,
+      "loss": 0.8503,
+      "step": 17
+    },
+    {
+      "epoch": 0.049484536082474224,
+      "grad_norm": 1.4431136736456789,
+      "learning_rate": 7.912087912087913e-06,
+      "loss": 0.8405,
+      "step": 18
+    },
+    {
+      "epoch": 0.05223367697594502,
+      "grad_norm": 1.4712422848105347,
+      "learning_rate": 8.351648351648353e-06,
+      "loss": 0.8274,
+      "step": 19
+    },
+    {
+      "epoch": 0.054982817869415807,
+      "grad_norm": 1.3178930536171367,
+      "learning_rate": 8.791208791208792e-06,
+      "loss": 0.8428,
+      "step": 20
+    },
+    {
+      "epoch": 0.0577319587628866,
+      "grad_norm": 1.2703340647552794,
+      "learning_rate": 9.230769230769232e-06,
+      "loss": 0.8211,
+      "step": 21
+    },
+    {
+      "epoch": 0.06048109965635739,
+      "grad_norm": 1.2871317510857054,
+      "learning_rate": 9.670329670329671e-06,
+      "loss": 0.8053,
+      "step": 22
+    },
+    {
+      "epoch": 0.06323024054982818,
+      "grad_norm": 1.1237165728026322,
+      "learning_rate": 1.010989010989011e-05,
+      "loss": 0.8101,
+      "step": 23
+    },
+    {
+      "epoch": 0.06597938144329897,
+      "grad_norm": 0.9168689840127058,
+      "learning_rate": 1.054945054945055e-05,
+      "loss": 0.7837,
+      "step": 24
+    },
+    {
+      "epoch": 0.06872852233676977,
+      "grad_norm": 0.9355697137818322,
+      "learning_rate": 1.098901098901099e-05,
+      "loss": 0.79,
+      "step": 25
+    },
+    {
+      "epoch": 0.07147766323024055,
+      "grad_norm": 0.8919194529855764,
+      "learning_rate": 1.1428571428571429e-05,
+      "loss": 0.7927,
+      "step": 26
+    },
+    {
+      "epoch": 0.07422680412371134,
+      "grad_norm": 0.8013095068191387,
+      "learning_rate": 1.186813186813187e-05,
+      "loss": 0.7826,
+      "step": 27
+    },
+    {
+      "epoch": 0.07697594501718213,
+      "grad_norm": 0.7074089625670721,
+      "learning_rate": 1.230769230769231e-05,
+      "loss": 0.77,
+      "step": 28
+    },
+    {
+      "epoch": 0.07972508591065292,
+      "grad_norm": 0.5939851673949501,
+      "learning_rate": 1.2747252747252747e-05,
+      "loss": 0.764,
+      "step": 29
+    },
+    {
+      "epoch": 0.08247422680412371,
+      "grad_norm": 0.7291769706418423,
+      "learning_rate": 1.3186813186813187e-05,
+      "loss": 0.7667,
+      "step": 30
+    },
+    {
+      "epoch": 0.0852233676975945,
+      "grad_norm": 0.5476356706171729,
+      "learning_rate": 1.3626373626373627e-05,
+      "loss": 0.7669,
+      "step": 31
+    },
+    {
+      "epoch": 0.0879725085910653,
+      "grad_norm": 0.496258207605204,
+      "learning_rate": 1.4065934065934068e-05,
+      "loss": 0.7617,
+      "step": 32
+    },
+    {
+      "epoch": 0.09072164948453608,
+      "grad_norm": 0.5382769338196303,
+      "learning_rate": 1.4505494505494506e-05,
+      "loss": 0.7554,
+      "step": 33
+    },
+    {
+      "epoch": 0.09347079037800687,
+      "grad_norm": 0.4580628373124881,
+      "learning_rate": 1.4945054945054947e-05,
+      "loss": 0.75,
+      "step": 34
+    },
+    {
+      "epoch": 0.09621993127147767,
+      "grad_norm": 0.40847997114284523,
+      "learning_rate": 1.5384615384615387e-05,
+      "loss": 0.7467,
+      "step": 35
+    },
+    {
+      "epoch": 0.09896907216494845,
+      "grad_norm": 0.5031600722774845,
+      "learning_rate": 1.5824175824175826e-05,
+      "loss": 0.7383,
+      "step": 36
+    },
+    {
+      "epoch": 0.10171821305841924,
+      "grad_norm": 0.3529000118564782,
+      "learning_rate": 1.6263736263736265e-05,
+      "loss": 0.7524,
+      "step": 37
+    },
+    {
+      "epoch": 0.10446735395189004,
+      "grad_norm": 0.40270165373486655,
+      "learning_rate": 1.6703296703296707e-05,
+      "loss": 0.7371,
+      "step": 38
+    },
+    {
+      "epoch": 0.10721649484536082,
+      "grad_norm": 0.3743730526645748,
+      "learning_rate": 1.7142857142857142e-05,
+      "loss": 0.7332,
+      "step": 39
+    },
+    {
+      "epoch": 0.10996563573883161,
+      "grad_norm": 0.3802324817797124,
+      "learning_rate": 1.7582417582417584e-05,
+      "loss": 0.7301,
+      "step": 40
+    },
+    {
+      "epoch": 0.11271477663230241,
+      "grad_norm": 0.38122013544181327,
+      "learning_rate": 1.8021978021978023e-05,
+      "loss": 0.7481,
+      "step": 41
+    },
+    {
+      "epoch": 0.1154639175257732,
+      "grad_norm": 0.3807629715206352,
+      "learning_rate": 1.8461538461538465e-05,
+      "loss": 0.7234,
+      "step": 42
+    },
+    {
+      "epoch": 0.11821305841924398,
+      "grad_norm": 0.3690276896533298,
+      "learning_rate": 1.8901098901098903e-05,
+      "loss": 0.7382,
+      "step": 43
+    },
+    {
+      "epoch": 0.12096219931271478,
+      "grad_norm": 0.39354181368093777,
+      "learning_rate": 1.9340659340659342e-05,
+      "loss": 0.7274,
+      "step": 44
+    },
+    {
+      "epoch": 0.12371134020618557,
+      "grad_norm": 0.45930210506482494,
+      "learning_rate": 1.9780219780219784e-05,
+      "loss": 0.7222,
+      "step": 45
+    },
+    {
+      "epoch": 0.12646048109965635,
+      "grad_norm": 0.3257242749282543,
+      "learning_rate": 2.021978021978022e-05,
+      "loss": 0.7062,
+      "step": 46
+    },
+    {
+      "epoch": 0.12920962199312716,
+      "grad_norm": 0.4894322683246899,
+      "learning_rate": 2.0659340659340665e-05,
+      "loss": 0.7273,
+      "step": 47
+    },
+    {
+      "epoch": 0.13195876288659794,
+      "grad_norm": 0.4217505425727167,
+      "learning_rate": 2.10989010989011e-05,
+      "loss": 0.7144,
+      "step": 48
+    },
+    {
+      "epoch": 0.13470790378006872,
+      "grad_norm": 0.34926375879792276,
+      "learning_rate": 2.153846153846154e-05,
+      "loss": 0.7284,
+      "step": 49
+    },
+    {
+      "epoch": 0.13745704467353953,
+      "grad_norm": 0.37706775972489026,
+      "learning_rate": 2.197802197802198e-05,
+      "loss": 0.7197,
+      "step": 50
+    },
+    {
+      "epoch": 0.1402061855670103,
+      "grad_norm": 0.4461248020784725,
+      "learning_rate": 2.241758241758242e-05,
+      "loss": 0.7168,
+      "step": 51
+    },
+    {
+      "epoch": 0.1429553264604811,
+      "grad_norm": 0.45683903751388155,
+      "learning_rate": 2.2857142857142858e-05,
+      "loss": 0.7047,
+      "step": 52
+    },
+    {
+      "epoch": 0.1457044673539519,
+      "grad_norm": 0.4941785029111688,
+      "learning_rate": 2.32967032967033e-05,
+      "loss": 0.7203,
+      "step": 53
+    },
+    {
+      "epoch": 0.14845360824742268,
+      "grad_norm": 0.7101247658482969,
+      "learning_rate": 2.373626373626374e-05,
+      "loss": 0.7132,
+      "step": 54
+    },
+    {
+      "epoch": 0.15120274914089346,
+      "grad_norm": 1.3845226821923962,
+      "learning_rate": 2.4175824175824177e-05,
+      "loss": 0.7139,
+      "step": 55
+    },
+    {
+      "epoch": 0.15395189003436427,
+      "grad_norm": 0.9145522126443197,
+      "learning_rate": 2.461538461538462e-05,
+      "loss": 0.7201,
+      "step": 56
+    },
+    {
+      "epoch": 0.15670103092783505,
+      "grad_norm": 0.7316554124403013,
+      "learning_rate": 2.5054945054945058e-05,
+      "loss": 0.6997,
+      "step": 57
+    },
+    {
+      "epoch": 0.15945017182130583,
+      "grad_norm": 1.525604886016204,
+      "learning_rate": 2.5494505494505493e-05,
+      "loss": 0.708,
+      "step": 58
+    },
+    {
+      "epoch": 0.16219931271477664,
+      "grad_norm": 0.9569983399077112,
+      "learning_rate": 2.593406593406594e-05,
+      "loss": 0.7023,
+      "step": 59
+    },
+    {
+      "epoch": 0.16494845360824742,
+      "grad_norm": 1.222421966279637,
+      "learning_rate": 2.6373626373626374e-05,
+      "loss": 0.699,
+      "step": 60
+    },
+    {
+      "epoch": 0.1676975945017182,
+      "grad_norm": 1.1284003579139785,
+      "learning_rate": 2.6813186813186813e-05,
+      "loss": 0.6944,
+      "step": 61
+    },
+    {
+      "epoch": 0.170446735395189,
+      "grad_norm": 1.323080634931126,
+      "learning_rate": 2.7252747252747255e-05,
+      "loss": 0.6897,
+      "step": 62
+    },
+    {
+      "epoch": 0.1731958762886598,
+      "grad_norm": 0.9918797937768813,
+      "learning_rate": 2.7692307692307694e-05,
+      "loss": 0.6937,
+      "step": 63
+    },
+    {
+      "epoch": 0.1759450171821306,
+      "grad_norm": 1.4485583433380003,
+      "learning_rate": 2.8131868131868136e-05,
+      "loss": 0.6924,
+      "step": 64
+    },
+    {
+      "epoch": 0.17869415807560138,
+      "grad_norm": 0.9586990948210379,
+      "learning_rate": 2.8571428571428574e-05,
+      "loss": 0.7036,
+      "step": 65
+    },
+    {
+      "epoch": 0.18144329896907216,
+      "grad_norm": 1.5401125351907554,
+      "learning_rate": 2.9010989010989013e-05,
+      "loss": 0.7067,
+      "step": 66
+    },
+    {
+      "epoch": 0.18419243986254297,
+      "grad_norm": 1.149941673251316,
+      "learning_rate": 2.9450549450549455e-05,
+      "loss": 0.6846,
+      "step": 67
+    },
+    {
+      "epoch": 0.18694158075601375,
+      "grad_norm": 1.398728105276103,
+      "learning_rate": 2.9890109890109894e-05,
+      "loss": 0.7001,
+      "step": 68
+    },
+    {
+      "epoch": 0.18969072164948453,
+      "grad_norm": 1.3380685614311483,
+      "learning_rate": 3.0329670329670332e-05,
+      "loss": 0.7017,
+      "step": 69
+    },
+    {
+      "epoch": 0.19243986254295534,
+      "grad_norm": 0.966967948159332,
+      "learning_rate": 3.0769230769230774e-05,
+      "loss": 0.6945,
+      "step": 70
+    },
+    {
+      "epoch": 0.19518900343642612,
+      "grad_norm": 1.2949156130535089,
+      "learning_rate": 3.120879120879121e-05,
+      "loss": 0.6983,
+      "step": 71
+    },
+    {
+      "epoch": 0.1979381443298969,
+      "grad_norm": 0.9121501279786434,
+      "learning_rate": 3.164835164835165e-05,
+      "loss": 0.6944,
+      "step": 72
+    },
+    {
+      "epoch": 0.2006872852233677,
+      "grad_norm": 0.7744380611271706,
+      "learning_rate": 3.2087912087912094e-05,
+      "loss": 0.6797,
+      "step": 73
+    },
+    {
+      "epoch": 0.2034364261168385,
+      "grad_norm": 1.171893254170078,
+      "learning_rate": 3.252747252747253e-05,
+      "loss": 0.6916,
+      "step": 74
+    },
+    {
+      "epoch": 0.20618556701030927,
+      "grad_norm": 1.432067598703686,
+      "learning_rate": 3.296703296703297e-05,
+      "loss": 0.696,
+      "step": 75
+    },
+    {
+      "epoch": 0.20893470790378008,
+      "grad_norm": 2.015800500571526,
+      "learning_rate": 3.340659340659341e-05,
+      "loss": 0.6946,
+      "step": 76
+    },
+    {
+      "epoch": 0.21168384879725086,
+      "grad_norm": 1.1450267893677495,
+      "learning_rate": 3.384615384615385e-05,
+      "loss": 0.6938,
+      "step": 77
+    },
+    {
+      "epoch": 0.21443298969072164,
+      "grad_norm": 3.012444154989314,
+      "learning_rate": 3.4285714285714284e-05,
+      "loss": 0.708,
+      "step": 78
+    },
+    {
+      "epoch": 0.21718213058419245,
+      "grad_norm": 2.938842593069413,
+      "learning_rate": 3.4725274725274726e-05,
+      "loss": 0.7043,
+      "step": 79
+    },
+    {
+      "epoch": 0.21993127147766323,
+      "grad_norm": 1.2620210326118846,
+      "learning_rate": 3.516483516483517e-05,
+      "loss": 0.7015,
+      "step": 80
+    },
+    {
+      "epoch": 0.22268041237113403,
+      "grad_norm": 1.8136600496263213,
+      "learning_rate": 3.56043956043956e-05,
+      "loss": 0.6934,
+      "step": 81
+    },
+    {
+      "epoch": 0.22542955326460482,
+      "grad_norm": 1.4037155277378568,
+      "learning_rate": 3.6043956043956045e-05,
+      "loss": 0.6923,
+      "step": 82
+    },
+    {
+      "epoch": 0.2281786941580756,
+      "grad_norm": 1.3955889629454836,
+      "learning_rate": 3.648351648351649e-05,
+      "loss": 0.6932,
+      "step": 83
+    },
+    {
+      "epoch": 0.2309278350515464,
+      "grad_norm": 1.4401203218805492,
+      "learning_rate": 3.692307692307693e-05,
+      "loss": 0.6837,
+      "step": 84
+    },
+    {
+      "epoch": 0.23367697594501718,
+      "grad_norm": 1.1506286717804464,
+      "learning_rate": 3.7362637362637365e-05,
+      "loss": 0.6824,
+      "step": 85
+    },
+    {
+      "epoch": 0.23642611683848797,
+      "grad_norm": 1.4015178509079977,
+      "learning_rate": 3.7802197802197807e-05,
+      "loss": 0.6818,
+      "step": 86
+    },
+    {
+      "epoch": 0.23917525773195877,
+      "grad_norm": 0.8632967007668437,
+      "learning_rate": 3.824175824175825e-05,
+      "loss": 0.6881,
+      "step": 87
+    },
+    {
+      "epoch": 0.24192439862542955,
+      "grad_norm": 1.7108832289817475,
+      "learning_rate": 3.8681318681318684e-05,
+      "loss": 0.694,
+      "step": 88
+    },
+    {
+      "epoch": 0.24467353951890033,
+      "grad_norm": 1.2330487896649007,
+      "learning_rate": 3.9120879120879126e-05,
+      "loss": 0.682,
+      "step": 89
+    },
+    {
+      "epoch": 0.24742268041237114,
+      "grad_norm": 1.2807204902880718,
+      "learning_rate": 3.956043956043957e-05,
+      "loss": 0.6851,
+      "step": 90
+    },
+    {
+      "epoch": 0.2501718213058419,
+      "grad_norm": 1.327068957190879,
+      "learning_rate": 4e-05,
+      "loss": 0.6807,
+      "step": 91
+    },
+    {
+      "epoch": 0.2529209621993127,
+      "grad_norm": 1.0022125576592875,
+      "learning_rate": 4.043956043956044e-05,
+      "loss": 0.6782,
+      "step": 92
+    },
+    {
+      "epoch": 0.2556701030927835,
+      "grad_norm": 1.3221553663136452,
+      "learning_rate": 4.087912087912088e-05,
+      "loss": 0.6759,
+      "step": 93
+    },
+    {
+      "epoch": 0.2584192439862543,
+      "grad_norm": 0.9568766990099716,
+      "learning_rate": 4.131868131868133e-05,
+      "loss": 0.6873,
+      "step": 94
+    },
+    {
+      "epoch": 0.2611683848797251,
+      "grad_norm": 1.7172395359463215,
+      "learning_rate": 4.1758241758241765e-05,
+      "loss": 0.6814,
+      "step": 95
+    },
+    {
+      "epoch": 0.2639175257731959,
+      "grad_norm": 1.5142722720152748,
+      "learning_rate": 4.21978021978022e-05,
+      "loss": 0.682,
+      "step": 96
+    },
+    {
+      "epoch": 0.26666666666666666,
+      "grad_norm": 0.8979115368433407,
+      "learning_rate": 4.263736263736264e-05,
+      "loss": 0.6828,
+      "step": 97
+    },
+    {
+      "epoch": 0.26941580756013744,
+      "grad_norm": 1.3321313501935168,
+      "learning_rate": 4.307692307692308e-05,
+      "loss": 0.687,
+      "step": 98
+    },
+    {
+      "epoch": 0.2721649484536082,
+      "grad_norm": 1.6997841277714232,
+      "learning_rate": 4.351648351648352e-05,
+      "loss": 0.6759,
+      "step": 99
+    },
+    {
+      "epoch": 0.27491408934707906,
+      "grad_norm": 1.0580779198597339,
+      "learning_rate": 4.395604395604396e-05,
+      "loss": 0.6743,
+      "step": 100
+    },
+    {
+      "epoch": 0.27766323024054984,
+      "grad_norm": 2.3143610998320896,
+      "learning_rate": 4.4395604395604403e-05,
+      "loss": 0.6684,
+      "step": 101
+    },
+    {
+      "epoch": 0.2804123711340206,
+      "grad_norm": 1.370198490883026,
+      "learning_rate": 4.483516483516484e-05,
+      "loss": 0.6643,
+      "step": 102
+    },
+    {
+      "epoch": 0.2831615120274914,
+      "grad_norm": 2.7976008957802088,
+      "learning_rate": 4.527472527472528e-05,
+      "loss": 0.6878,
+      "step": 103
+    },
+    {
+      "epoch": 0.2859106529209622,
+      "grad_norm": 2.539570133020213,
+      "learning_rate": 4.5714285714285716e-05,
+      "loss": 0.6889,
+      "step": 104
+    },
+    {
+      "epoch": 0.28865979381443296,
+      "grad_norm": 1.7701837403812288,
+      "learning_rate": 4.615384615384615e-05,
+      "loss": 0.6794,
+      "step": 105
+    },
+    {
+      "epoch": 0.2914089347079038,
+      "grad_norm": 1.6244664002638955,
+      "learning_rate": 4.65934065934066e-05,
+      "loss": 0.6798,
+      "step": 106
+    },
+    {
+      "epoch": 0.2941580756013746,
+      "grad_norm": 1.658296656226508,
+      "learning_rate": 4.7032967032967035e-05,
+      "loss": 0.6852,
+      "step": 107
+    },
+    {
+      "epoch": 0.29690721649484536,
+      "grad_norm": 1.3133889484198134,
+      "learning_rate": 4.747252747252748e-05,
+      "loss": 0.6695,
+      "step": 108
+    },
+    {
+      "epoch": 0.29965635738831614,
+      "grad_norm": 1.5315701499343,
+      "learning_rate": 4.791208791208791e-05,
+      "loss": 0.6842,
+      "step": 109
+    },
+    {
+      "epoch": 0.3024054982817869,
+      "grad_norm": 1.5106029998832964,
+      "learning_rate": 4.8351648351648355e-05,
+      "loss": 0.6728,
+      "step": 110
+    },
+    {
+      "epoch": 0.30515463917525776,
+      "grad_norm": 1.2198017486153612,
+      "learning_rate": 4.87912087912088e-05,
+      "loss": 0.6804,
+      "step": 111
+    },
+    {
+      "epoch": 0.30790378006872854,
+      "grad_norm": 0.9408294254788567,
+      "learning_rate": 4.923076923076924e-05,
+      "loss": 0.6673,
+      "step": 112
+    },
+    {
+      "epoch": 0.3106529209621993,
+      "grad_norm": 1.4299436206183789,
+      "learning_rate": 4.9670329670329674e-05,
+      "loss": 0.6782,
+      "step": 113
+    },
+    {
+      "epoch": 0.3134020618556701,
+      "grad_norm": 1.3010524254683744,
+      "learning_rate": 5.0109890109890116e-05,
+      "loss": 0.6719,
+      "step": 114
+    },
+    {
+      "epoch": 0.3161512027491409,
+      "grad_norm": 1.0294736794964319,
+      "learning_rate": 5.054945054945055e-05,
+      "loss": 0.6734,
+      "step": 115
+    },
+    {
+      "epoch": 0.31890034364261166,
+      "grad_norm": 1.9844841183716888,
+      "learning_rate": 5.098901098901099e-05,
+      "loss": 0.6689,
+      "step": 116
+    },
+    {
+      "epoch": 0.3216494845360825,
+      "grad_norm": 1.0769609751269622,
+      "learning_rate": 5.1428571428571436e-05,
+      "loss": 0.6725,
+      "step": 117
+    },
+    {
+      "epoch": 0.3243986254295533,
+      "grad_norm": 1.9412353253336612,
+      "learning_rate": 5.186813186813188e-05,
+      "loss": 0.6788,
+      "step": 118
+    },
+    {
+      "epoch": 0.32714776632302406,
+      "grad_norm": 1.2983006823972247,
+      "learning_rate": 5.230769230769231e-05,
+      "loss": 0.6717,
+      "step": 119
+    },
+    {
+      "epoch": 0.32989690721649484,
+      "grad_norm": 1.78349719868111,
+      "learning_rate": 5.274725274725275e-05,
+      "loss": 0.6754,
+      "step": 120
+    },
+    {
+      "epoch": 0.3326460481099656,
+      "grad_norm": 1.2685409855209113,
+      "learning_rate": 5.318681318681319e-05,
+      "loss": 0.6689,
+      "step": 121
+    },
+    {
+      "epoch": 0.3353951890034364,
+      "grad_norm": 1.601626046067794,
+      "learning_rate": 5.3626373626373626e-05,
+      "loss": 0.6822,
+      "step": 122
+    },
+    {
+      "epoch": 0.33814432989690724,
+      "grad_norm": 1.6399160217312283,
+      "learning_rate": 5.4065934065934074e-05,
+      "loss": 0.6939,
+      "step": 123
+    },
+    {
+      "epoch": 0.340893470790378,
+      "grad_norm": 1.0496938208520434,
+      "learning_rate": 5.450549450549451e-05,
+      "loss": 0.6685,
+      "step": 124
+    },
+    {
+      "epoch": 0.3436426116838488,
+      "grad_norm": 1.8210641674833647,
+      "learning_rate": 5.494505494505495e-05,
+      "loss": 0.6726,
+      "step": 125
+    },
+    {
+      "epoch": 0.3463917525773196,
+      "grad_norm": 1.0994677896279192,
+      "learning_rate": 5.538461538461539e-05,
+      "loss": 0.6609,
+      "step": 126
+    },
+    {
+      "epoch": 0.34914089347079036,
+      "grad_norm": 1.7375889267050222,
+      "learning_rate": 5.582417582417583e-05,
+      "loss": 0.68,
+      "step": 127
+    },
+    {
+      "epoch": 0.3518900343642612,
+      "grad_norm": 1.1687845090860516,
+      "learning_rate": 5.626373626373627e-05,
+      "loss": 0.6715,
+      "step": 128
+    },
+    {
+      "epoch": 0.354639175257732,
+      "grad_norm": 1.395445256796641,
+      "learning_rate": 5.670329670329671e-05,
+      "loss": 0.6681,
+      "step": 129
+    },
+    {
+      "epoch": 0.35738831615120276,
+      "grad_norm": 1.355862476688157,
+      "learning_rate": 5.714285714285715e-05,
+      "loss": 0.6818,
+      "step": 130
+    },
+    {
+      "epoch": 0.36013745704467354,
+      "grad_norm": 1.0924489864173972,
+      "learning_rate": 5.7582417582417584e-05,
+      "loss": 0.6605,
+      "step": 131
+    },
+    {
+      "epoch": 0.3628865979381443,
+      "grad_norm": 1.9384854103627827,
+      "learning_rate": 5.8021978021978026e-05,
+      "loss": 0.6731,
+      "step": 132
+    },
+    {
+      "epoch": 0.3656357388316151,
+      "grad_norm": 1.5529834412680505,
+      "learning_rate": 5.846153846153846e-05,
+      "loss": 0.675,
+      "step": 133
+    },
+    {
+      "epoch": 0.36838487972508593,
+      "grad_norm": 0.9671293876862966,
+      "learning_rate": 5.890109890109891e-05,
+      "loss": 0.661,
+      "step": 134
+    },
+    {
+      "epoch": 0.3711340206185567,
+      "grad_norm": 1.5434547051402783,
+      "learning_rate": 5.9340659340659345e-05,
+      "loss": 0.673,
+      "step": 135
+    },
+    {
+      "epoch": 0.3738831615120275,
+      "grad_norm": 1.466260971358487,
+      "learning_rate": 5.978021978021979e-05,
+      "loss": 0.6799,
+      "step": 136
+    },
+    {
+      "epoch": 0.3766323024054983,
+      "grad_norm": 1.3025182971790092,
+      "learning_rate": 6.021978021978022e-05,
+      "loss": 0.656,
+      "step": 137
+    },
+    {
+      "epoch": 0.37938144329896906,
+      "grad_norm": 1.5407519214603846,
+      "learning_rate": 6.0659340659340665e-05,
+      "loss": 0.6703,
+      "step": 138
+    },
+    {
+      "epoch": 0.38213058419243984,
+      "grad_norm": 1.2435864649597408,
+      "learning_rate": 6.10989010989011e-05,
+      "loss": 0.6699,
+      "step": 139
+    },
+    {
+      "epoch": 0.3848797250859107,
+      "grad_norm": 1.32415071563785,
+      "learning_rate": 6.153846153846155e-05,
+      "loss": 0.6557,
+      "step": 140
+    },
+    {
+      "epoch": 0.38762886597938145,
+      "grad_norm": 1.5121087379761295,
+      "learning_rate": 6.197802197802199e-05,
+      "loss": 0.6659,
+      "step": 141
+    },
+    {
+      "epoch": 0.39037800687285223,
+      "grad_norm": 1.3570253258903069,
+      "learning_rate": 6.241758241758242e-05,
+      "loss": 0.6568,
+      "step": 142
+    },
+    {
+      "epoch": 0.393127147766323,
+      "grad_norm": 1.2775583571637077,
+      "learning_rate": 6.285714285714286e-05,
+      "loss": 0.6558,
+      "step": 143
+    },
+    {
+      "epoch": 0.3958762886597938,
+      "grad_norm": 1.029079383256909,
+      "learning_rate": 6.32967032967033e-05,
+      "loss": 0.6569,
+      "step": 144
+    },
+    {
+      "epoch": 0.39862542955326463,
+      "grad_norm": 1.379942185445747,
+      "learning_rate": 6.373626373626373e-05,
+      "loss": 0.6664,
+      "step": 145
+    },
+    {
+      "epoch": 0.4013745704467354,
+      "grad_norm": 1.7317621636752105,
+      "learning_rate": 6.417582417582419e-05,
+      "loss": 0.6675,
+      "step": 146
+    },
+    {
+      "epoch": 0.4041237113402062,
+      "grad_norm": 1.0553856530228192,
+      "learning_rate": 6.461538461538462e-05,
+      "loss": 0.6623,
+      "step": 147
+    },
+    {
+      "epoch": 0.406872852233677,
+      "grad_norm": 1.310533202428338,
+      "learning_rate": 6.505494505494506e-05,
+      "loss": 0.6658,
+      "step": 148
+    },
+    {
+      "epoch": 0.40962199312714775,
+      "grad_norm": 1.0365623836624396,
+      "learning_rate": 6.54945054945055e-05,
+      "loss": 0.6539,
+      "step": 149
+    },
+    {
+      "epoch": 0.41237113402061853,
+      "grad_norm": 1.5824321098093788,
+      "learning_rate": 6.593406593406594e-05,
+      "loss": 0.672,
+      "step": 150
+    },
+    {
+      "epoch": 0.41512027491408937,
+      "grad_norm": 1.2342253922087731,
+      "learning_rate": 6.637362637362638e-05,
+      "loss": 0.6609,
+      "step": 151
+    },
+    {
+      "epoch": 0.41786941580756015,
+      "grad_norm": 2.2157482513280797,
+      "learning_rate": 6.681318681318683e-05,
+      "loss": 0.6614,
+      "step": 152
+    },
+    {
+      "epoch": 0.42061855670103093,
+      "grad_norm": 0.8519496697821355,
+      "learning_rate": 6.725274725274725e-05,
+      "loss": 0.6558,
+      "step": 153
+    },
+    {
+      "epoch": 0.4233676975945017,
+      "grad_norm": 2.585667143290517,
+      "learning_rate": 6.76923076923077e-05,
+      "loss": 0.6712,
+      "step": 154
+    },
+    {
+      "epoch": 0.4261168384879725,
+      "grad_norm": 1.561949044082342,
+      "learning_rate": 6.813186813186814e-05,
+      "loss": 0.6606,
+      "step": 155
+    },
+    {
+      "epoch": 0.4288659793814433,
+      "grad_norm": 1.6541712978836747,
+      "learning_rate": 6.857142857142857e-05,
+      "loss": 0.6662,
+      "step": 156
+    },
+    {
+      "epoch": 0.4316151202749141,
+      "grad_norm": 1.471980620525164,
+      "learning_rate": 6.901098901098902e-05,
+      "loss": 0.6688,
+      "step": 157
+    },
+    {
+      "epoch": 0.4343642611683849,
+      "grad_norm": 1.4565326176621909,
+      "learning_rate": 6.945054945054945e-05,
+      "loss": 0.668,
+      "step": 158
+    },
+    {
+      "epoch": 0.43711340206185567,
+      "grad_norm": 1.3548150534245453,
+      "learning_rate": 6.98901098901099e-05,
+      "loss": 0.6628,
+      "step": 159
+    },
+    {
+      "epoch": 0.43986254295532645,
+      "grad_norm": 1.2045551451888672,
+      "learning_rate": 7.032967032967034e-05,
+      "loss": 0.6616,
+      "step": 160
+    },
+    {
+      "epoch": 0.44261168384879723,
+      "grad_norm": 1.7494121190993142,
+      "learning_rate": 7.076923076923078e-05,
+      "loss": 0.6581,
+      "step": 161
+    },
+    {
+      "epoch": 0.44536082474226807,
+      "grad_norm": 1.0147994550817765,
+      "learning_rate": 7.12087912087912e-05,
+      "loss": 0.6608,
+      "step": 162
+    },
+    {
+      "epoch": 0.44810996563573885,
+      "grad_norm": 1.0859192635780746,
+      "learning_rate": 7.164835164835166e-05,
+      "loss": 0.6451,
+      "step": 163
+    },
+    {
+      "epoch": 0.45085910652920963,
+      "grad_norm": 1.0492942035079653,
+      "learning_rate": 7.208791208791209e-05,
+      "loss": 0.6591,
+      "step": 164
+    },
+    {
+      "epoch": 0.4536082474226804,
+      "grad_norm": 2.379607115922538,
+      "learning_rate": 7.252747252747253e-05,
+      "loss": 0.6634,
+      "step": 165
+    },
+    {
+      "epoch": 0.4563573883161512,
+      "grad_norm": 1.112381416151378,
+      "learning_rate": 7.296703296703297e-05,
+      "loss": 0.6611,
+      "step": 166
+    },
+    {
+      "epoch": 0.45910652920962197,
+      "grad_norm": 2.0850617091598886,
+      "learning_rate": 7.34065934065934e-05,
+      "loss": 0.6569,
+      "step": 167
+    },
+    {
+      "epoch": 0.4618556701030928,
+      "grad_norm": 1.5623269496352417,
+      "learning_rate": 7.384615384615386e-05,
+      "loss": 0.672,
+      "step": 168
+    },
+    {
+      "epoch": 0.4646048109965636,
+      "grad_norm": 1.6783665454838335,
+      "learning_rate": 7.42857142857143e-05,
+      "loss": 0.6567,
+      "step": 169
+    },
+    {
+      "epoch": 0.46735395189003437,
+      "grad_norm": 1.0376996339879248,
+      "learning_rate": 7.472527472527473e-05,
+      "loss": 0.6606,
+      "step": 170
+    },
+    {
+      "epoch": 0.47010309278350515,
+      "grad_norm": 2.2216262071159356,
+      "learning_rate": 7.516483516483517e-05,
+      "loss": 0.6636,
+      "step": 171
+    },
+    {
+      "epoch": 0.47285223367697593,
+      "grad_norm": 1.3858824784808275,
+      "learning_rate": 7.560439560439561e-05,
+      "loss": 0.6673,
+      "step": 172
+    },
+    {
+      "epoch": 0.4756013745704467,
+      "grad_norm": 1.9605283100011852,
+      "learning_rate": 7.604395604395604e-05,
+      "loss": 0.6572,
+      "step": 173
+    },
+    {
+      "epoch": 0.47835051546391755,
+      "grad_norm": 1.4278083885748436,
+      "learning_rate": 7.64835164835165e-05,
+      "loss": 0.6645,
+      "step": 174
+    },
+    {
+      "epoch": 0.48109965635738833,
+      "grad_norm": 1.191754742216612,
+      "learning_rate": 7.692307692307693e-05,
+      "loss": 0.6631,
+      "step": 175
+    },
+    {
+      "epoch": 0.4838487972508591,
+      "grad_norm": 0.9597081578943443,
+      "learning_rate": 7.736263736263737e-05,
+      "loss": 0.655,
+      "step": 176
+    },
+    {
+      "epoch": 0.4865979381443299,
+      "grad_norm": 1.9487418698749246,
+      "learning_rate": 7.780219780219781e-05,
+      "loss": 0.6617,
+      "step": 177
+    },
+    {
+      "epoch": 0.48934707903780067,
+      "grad_norm": 1.4578651378155227,
+      "learning_rate": 7.824175824175825e-05,
+      "loss": 0.665,
+      "step": 178
+    },
+    {
+      "epoch": 0.49209621993127145,
+      "grad_norm": 0.9204633011470749,
+      "learning_rate": 7.868131868131868e-05,
+      "loss": 0.6559,
+      "step": 179
+    },
+    {
+      "epoch": 0.4948453608247423,
+      "grad_norm": 2.3519272780997893,
+      "learning_rate": 7.912087912087914e-05,
+      "loss": 0.661,
+      "step": 180
+    },
+    {
+      "epoch": 0.49759450171821307,
+      "grad_norm": 1.4425041211579737,
+      "learning_rate": 7.956043956043956e-05,
+      "loss": 0.6555,
+      "step": 181
+    },
+    {
+      "epoch": 0.5003436426116838,
+      "grad_norm": 2.2406732058079917,
+      "learning_rate": 8e-05,
+      "loss": 0.6672,
+      "step": 182
+    },
+    {
+      "epoch": 0.5030927835051546,
+      "grad_norm": 1.717445277482838,
+      "learning_rate": 7.999992597860977e-05,
+      "loss": 0.669,
+      "step": 183
+    },
+    {
+      "epoch": 0.5058419243986254,
+      "grad_norm": 1.9554853415214135,
+      "learning_rate": 7.999970391471297e-05,
+      "loss": 0.6484,
+      "step": 184
+    },
+    {
+      "epoch": 0.5085910652920962,
+      "grad_norm": 1.3337471142504986,
+      "learning_rate": 7.99993338091315e-05,
+      "loss": 0.6542,
+      "step": 185
+    },
+    {
+      "epoch": 0.511340206185567,
+      "grad_norm": 2.3628021636514913,
+      "learning_rate": 7.999881566323518e-05,
+      "loss": 0.6666,
+      "step": 186
+    },
+    {
+      "epoch": 0.5140893470790378,
+      "grad_norm": 1.6284990281451381,
+      "learning_rate": 7.999814947894166e-05,
+      "loss": 0.6713,
+      "step": 187
+    },
+    {
+      "epoch": 0.5168384879725086,
+      "grad_norm": 2.208780281202899,
+      "learning_rate": 7.999733525871655e-05,
+      "loss": 0.6564,
+      "step": 188
+    },
+    {
+      "epoch": 0.5195876288659794,
+      "grad_norm": 2.0324845611311653,
+      "learning_rate": 7.999637300557334e-05,
+      "loss": 0.6643,
+      "step": 189
+    },
+    {
+      "epoch": 0.5223367697594502,
+      "grad_norm": 1.411207407439057,
+      "learning_rate": 7.999526272307338e-05,
+      "loss": 0.6466,
+      "step": 190
+    },
+    {
+      "epoch": 0.525085910652921,
+      "grad_norm": 1.67795112571074,
+      "learning_rate": 7.999400441532593e-05,
+      "loss": 0.6567,
+      "step": 191
+    },
+    {
+      "epoch": 0.5278350515463918,
+      "grad_norm": 1.0171507403303963,
+      "learning_rate": 7.999259808698805e-05,
+      "loss": 0.6535,
+      "step": 192
+    },
+    {
+      "epoch": 0.5305841924398625,
+      "grad_norm": 1.8103499540778905,
+      "learning_rate": 7.999104374326465e-05,
+      "loss": 0.6546,
+      "step": 193
+    },
+    {
+      "epoch": 0.5333333333333333,
+      "grad_norm": 1.3619232333701012,
+      "learning_rate": 7.99893413899085e-05,
+      "loss": 0.6495,
+      "step": 194
+    },
+    {
+      "epoch": 0.5360824742268041,
+      "grad_norm": 1.7537074766579162,
+      "learning_rate": 7.99874910332201e-05,
+      "loss": 0.6513,
+      "step": 195
+    },
+    {
+      "epoch": 0.5388316151202749,
+      "grad_norm": 1.5125359865682328,
+      "learning_rate": 7.998549268004776e-05,
+      "loss": 0.6627,
+      "step": 196
+    },
+    {
+      "epoch": 0.5415807560137457,
+      "grad_norm": 1.8694904140870736,
+      "learning_rate": 7.998334633778752e-05,
+      "loss": 0.6552,
+      "step": 197
+    },
+    {
+      "epoch": 0.5443298969072164,
+      "grad_norm": 1.0313289192419033,
+      "learning_rate": 7.998105201438315e-05,
+      "loss": 0.6572,
+      "step": 198
+    },
+    {
+      "epoch": 0.5470790378006873,
+      "grad_norm": 1.7524234342414688,
+      "learning_rate": 7.997860971832609e-05,
+      "loss": 0.6485,
+      "step": 199
+    },
+    {
+      "epoch": 0.5498281786941581,
+      "grad_norm": 1.674466921847352,
+      "learning_rate": 7.997601945865545e-05,
+      "loss": 0.6567,
+      "step": 200
+    },
+    {
+      "epoch": 0.5525773195876289,
+      "grad_norm": 0.9044207656617541,
+      "learning_rate": 7.997328124495797e-05,
+      "loss": 0.6571,
+      "step": 201
+    },
+    {
+      "epoch": 0.5553264604810997,
+      "grad_norm": 1.5698427976284253,
+      "learning_rate": 7.997039508736794e-05,
+      "loss": 0.6552,
+      "step": 202
+    },
+    {
+      "epoch": 0.5580756013745705,
+      "grad_norm": 1.2149289312154958,
+      "learning_rate": 7.996736099656728e-05,
+      "loss": 0.6504,
+      "step": 203
+    },
+    {
+      "epoch": 0.5608247422680412,
+      "grad_norm": 1.5016916008686747,
+      "learning_rate": 7.996417898378532e-05,
+      "loss": 0.6501,
+      "step": 204
+    },
+    {
+      "epoch": 0.563573883161512,
+      "grad_norm": 1.3674415360691992,
+      "learning_rate": 7.996084906079895e-05,
+      "loss": 0.6521,
+      "step": 205
+    },
+    {
+      "epoch": 0.5663230240549828,
+      "grad_norm": 1.012486835623592,
+      "learning_rate": 7.995737123993242e-05,
+      "loss": 0.6533,
+      "step": 206
+    },
+    {
+      "epoch": 0.5690721649484536,
+      "grad_norm": 1.2333701554709846,
+      "learning_rate": 7.99537455340574e-05,
+      "loss": 0.6423,
+      "step": 207
+    },
+    {
+      "epoch": 0.5718213058419244,
+      "grad_norm": 1.432895391786843,
+      "learning_rate": 7.994997195659288e-05,
+      "loss": 0.6588,
+      "step": 208
+    },
+    {
+      "epoch": 0.5745704467353951,
+      "grad_norm": 1.3560618026372162,
+      "learning_rate": 7.994605052150512e-05,
+      "loss": 0.6419,
+      "step": 209
+    },
+    {
+      "epoch": 0.5773195876288659,
+      "grad_norm": 1.1169005294274361,
+      "learning_rate": 7.994198124330764e-05,
+      "loss": 0.643,
+      "step": 210
+    },
+    {
+      "epoch": 0.5800687285223368,
+      "grad_norm": 1.1551183527776216,
+      "learning_rate": 7.99377641370611e-05,
+      "loss": 0.6658,
+      "step": 211
+    },
+    {
+      "epoch": 0.5828178694158076,
+      "grad_norm": 1.494094722393518,
+      "learning_rate": 7.993339921837333e-05,
+      "loss": 0.6566,
+      "step": 212
+    },
+    {
+      "epoch": 0.5855670103092784,
+      "grad_norm": 1.1087501045845025,
+      "learning_rate": 7.992888650339918e-05,
+      "loss": 0.6389,
+      "step": 213
+    },
+    {
+      "epoch": 0.5883161512027492,
+      "grad_norm": 1.9002883231044971,
+      "learning_rate": 7.992422600884052e-05,
+      "loss": 0.6539,
+      "step": 214
+    },
+    {
+      "epoch": 0.5910652920962199,
+      "grad_norm": 0.7299685549898423,
+      "learning_rate": 7.991941775194619e-05,
+      "loss": 0.6465,
+      "step": 215
+    },
+    {
+      "epoch": 0.5938144329896907,
+      "grad_norm": 1.712564742362983,
+      "learning_rate": 7.991446175051184e-05,
+      "loss": 0.6474,
+      "step": 216
+    },
+    {
+      "epoch": 0.5965635738831615,
+      "grad_norm": 1.4830236426687247,
+      "learning_rate": 7.990935802288002e-05,
+      "loss": 0.6512,
+      "step": 217
+    },
+    {
+      "epoch": 0.5993127147766323,
+      "grad_norm": 1.1265182423363587,
+      "learning_rate": 7.990410658793994e-05,
+      "loss": 0.6541,
+      "step": 218
+    },
+    {
+      "epoch": 0.6020618556701031,
+      "grad_norm": 1.2971103607944785,
+      "learning_rate": 7.989870746512756e-05,
+      "loss": 0.6479,
+      "step": 219
+    },
+    {
+      "epoch": 0.6048109965635738,
+      "grad_norm": 1.2538177052036752,
+      "learning_rate": 7.989316067442539e-05,
+      "loss": 0.6497,
+      "step": 220
+    },
+    {
+      "epoch": 0.6075601374570446,
+      "grad_norm": 0.7957257777767685,
+      "learning_rate": 7.98874662363625e-05,
+      "loss": 0.6352,
+      "step": 221
+    },
+    {
+      "epoch": 0.6103092783505155,
+      "grad_norm": 0.9768965818664068,
+      "learning_rate": 7.988162417201437e-05,
+      "loss": 0.6443,
+      "step": 222
+    },
+    {
+      "epoch": 0.6130584192439863,
+      "grad_norm": 1.9580245398506602,
+      "learning_rate": 7.987563450300293e-05,
+      "loss": 0.6652,
+      "step": 223
+    },
+    {
+      "epoch": 0.6158075601374571,
+      "grad_norm": 1.065435200900762,
+      "learning_rate": 7.986949725149634e-05,
+      "loss": 0.6456,
+      "step": 224
+    },
+    {
+      "epoch": 0.6185567010309279,
+      "grad_norm": 1.9649018261160003,
+      "learning_rate": 7.986321244020901e-05,
+      "loss": 0.6479,
+      "step": 225
+    },
+    {
+      "epoch": 0.6213058419243986,
+      "grad_norm": 1.6213807682895127,
+      "learning_rate": 7.985678009240142e-05,
+      "loss": 0.6595,
+      "step": 226
+    },
+    {
+      "epoch": 0.6240549828178694,
+      "grad_norm": 1.4900550716535215,
+      "learning_rate": 7.985020023188018e-05,
+      "loss": 0.6359,
+      "step": 227
+    },
+    {
+      "epoch": 0.6268041237113402,
+      "grad_norm": 1.3214648767288044,
+      "learning_rate": 7.98434728829978e-05,
+      "loss": 0.647,
+      "step": 228
+    },
+    {
+      "epoch": 0.629553264604811,
+      "grad_norm": 1.4301295600290387,
+      "learning_rate": 7.983659807065267e-05,
+      "loss": 0.6475,
+      "step": 229
+    },
+    {
+      "epoch": 0.6323024054982818,
+      "grad_norm": 1.0375560871972622,
+      "learning_rate": 7.982957582028892e-05,
+      "loss": 0.6364,
+      "step": 230
+    },
+    {
+      "epoch": 0.6350515463917525,
+      "grad_norm": 1.3894087032049172,
+      "learning_rate": 7.982240615789641e-05,
+      "loss": 0.6426,
+      "step": 231
+    },
+    {
+      "epoch": 0.6378006872852233,
+      "grad_norm": 0.9850693356026884,
+      "learning_rate": 7.981508911001057e-05,
+      "loss": 0.6459,
+      "step": 232
+    },
+    {
+      "epoch": 0.6405498281786941,
+      "grad_norm": 1.4426682351463136,
+      "learning_rate": 7.980762470371228e-05,
+      "loss": 0.6474,
+      "step": 233
+    },
+    {
+      "epoch": 0.643298969072165,
+      "grad_norm": 1.0547468177759594,
+      "learning_rate": 7.980001296662784e-05,
+      "loss": 0.6469,
+      "step": 234
+    },
+    {
+      "epoch": 0.6460481099656358,
+      "grad_norm": 1.2056014455360582,
+      "learning_rate": 7.979225392692882e-05,
+      "loss": 0.6354,
+      "step": 235
+    },
+    {
+      "epoch": 0.6487972508591066,
+      "grad_norm": 1.535046209332006,
+      "learning_rate": 7.978434761333195e-05,
+      "loss": 0.6432,
+      "step": 236
+    },
+    {
+      "epoch": 0.6515463917525773,
+      "grad_norm": 0.7757101221034997,
+      "learning_rate": 7.977629405509905e-05,
+      "loss": 0.633,
+      "step": 237
+    },
+    {
+      "epoch": 0.6542955326460481,
+      "grad_norm": 0.992595763686358,
+      "learning_rate": 7.976809328203693e-05,
+      "loss": 0.64,
+      "step": 238
+    },
+    {
+      "epoch": 0.6570446735395189,
+      "grad_norm": 1.3919630066257997,
+      "learning_rate": 7.975974532449718e-05,
+      "loss": 0.6495,
+      "step": 239
+    },
+    {
+      "epoch": 0.6597938144329897,
+      "grad_norm": 1.130072368133697,
+      "learning_rate": 7.975125021337618e-05,
+      "loss": 0.6337,
+      "step": 240
+    },
+    {
+      "epoch": 0.6625429553264605,
+      "grad_norm": 1.0428416388066186,
+      "learning_rate": 7.974260798011494e-05,
+      "loss": 0.6515,
+      "step": 241
+    },
+    {
+      "epoch": 0.6652920962199312,
+      "grad_norm": 1.4678385771525282,
+      "learning_rate": 7.973381865669897e-05,
+      "loss": 0.6446,
+      "step": 242
+    },
+    {
+      "epoch": 0.668041237113402,
+      "grad_norm": 1.1731453018255897,
+      "learning_rate": 7.972488227565814e-05,
+      "loss": 0.6424,
+      "step": 243
+    },
+    {
+      "epoch": 0.6707903780068728,
+      "grad_norm": 1.1037532875258624,
+      "learning_rate": 7.971579887006663e-05,
+      "loss": 0.6356,
+      "step": 244
+    },
+    {
+      "epoch": 0.6735395189003437,
+      "grad_norm": 1.3601555121892395,
+      "learning_rate": 7.970656847354277e-05,
+      "loss": 0.6332,
+      "step": 245
+    },
+    {
+      "epoch": 0.6762886597938145,
+      "grad_norm": 0.6674453682989576,
+      "learning_rate": 7.969719112024889e-05,
+      "loss": 0.6341,
+      "step": 246
+    },
+    {
+      "epoch": 0.6790378006872853,
+      "grad_norm": 0.8185867421910552,
+      "learning_rate": 7.968766684489122e-05,
+      "loss": 0.6351,
+      "step": 247
+    },
+    {
+      "epoch": 0.681786941580756,
+      "grad_norm": 1.2859620347833656,
+      "learning_rate": 7.967799568271978e-05,
+      "loss": 0.636,
+      "step": 248
+    },
+    {
+      "epoch": 0.6845360824742268,
+      "grad_norm": 1.0894539357769442,
+      "learning_rate": 7.96681776695282e-05,
+      "loss": 0.6483,
+      "step": 249
+    },
+    {
+      "epoch": 0.6872852233676976,
+      "grad_norm": 1.536917534466485,
+      "learning_rate": 7.965821284165362e-05,
+      "loss": 0.6358,
+      "step": 250
+    },
+    {
+      "epoch": 0.6900343642611684,
+      "grad_norm": 0.8985662521364918,
+      "learning_rate": 7.964810123597659e-05,
+      "loss": 0.6345,
+      "step": 251
+    },
+    {
+      "epoch": 0.6927835051546392,
+      "grad_norm": 1.274319886309132,
+      "learning_rate": 7.963784288992085e-05,
+      "loss": 0.6416,
+      "step": 252
+    },
+    {
+      "epoch": 0.6955326460481099,
+      "grad_norm": 0.8642801778285559,
+      "learning_rate": 7.962743784145323e-05,
+      "loss": 0.6386,
+      "step": 253
+    },
+    {
+      "epoch": 0.6982817869415807,
+      "grad_norm": 0.9360706797234668,
+      "learning_rate": 7.961688612908358e-05,
+      "loss": 0.6311,
+      "step": 254
+    },
+    {
+      "epoch": 0.7010309278350515,
+      "grad_norm": 1.2347692000194708,
+      "learning_rate": 7.96061877918645e-05,
+      "loss": 0.6342,
+      "step": 255
+    },
+    {
+      "epoch": 0.7037800687285224,
+      "grad_norm": 1.3067449184972966,
+      "learning_rate": 7.959534286939126e-05,
+      "loss": 0.6348,
+      "step": 256
+    },
+    {
+      "epoch": 0.7065292096219932,
+      "grad_norm": 1.0853797825255465,
+      "learning_rate": 7.95843514018017e-05,
+      "loss": 0.6387,
+      "step": 257
+    },
+    {
+      "epoch": 0.709278350515464,
+      "grad_norm": 1.3876365868084906,
+      "learning_rate": 7.9573213429776e-05,
+      "loss": 0.6408,
+      "step": 258
+    },
+    {
+      "epoch": 0.7120274914089347,
+      "grad_norm": 1.2720820097889098,
+      "learning_rate": 7.956192899453656e-05,
+      "loss": 0.6403,
+      "step": 259
+    },
+    {
+      "epoch": 0.7147766323024055,
+      "grad_norm": 1.083330147390025,
+      "learning_rate": 7.955049813784787e-05,
+      "loss": 0.6326,
+      "step": 260
+    },
+    {
+      "epoch": 0.7175257731958763,
+      "grad_norm": 0.8536354360684208,
+      "learning_rate": 7.953892090201633e-05,
+      "loss": 0.6289,
+      "step": 261
+    },
+    {
+      "epoch": 0.7202749140893471,
+      "grad_norm": 0.6669283801432064,
+      "learning_rate": 7.952719732989007e-05,
+      "loss": 0.6311,
+      "step": 262
+    },
+    {
+      "epoch": 0.7230240549828179,
+      "grad_norm": 0.6028131310980143,
+      "learning_rate": 7.951532746485886e-05,
+      "loss": 0.6383,
+      "step": 263
+    },
+    {
+      "epoch": 0.7257731958762886,
+      "grad_norm": 1.0830419136461447,
+      "learning_rate": 7.950331135085389e-05,
+      "loss": 0.6257,
+      "step": 264
+    },
+    {
+      "epoch": 0.7285223367697594,
+      "grad_norm": 1.7725196403596395,
+      "learning_rate": 7.949114903234766e-05,
+      "loss": 0.6431,
+      "step": 265
+    },
+    {
+      "epoch": 0.7312714776632302,
+      "grad_norm": 0.7205576501203266,
+      "learning_rate": 7.947884055435371e-05,
+      "loss": 0.636,
+      "step": 266
+    },
+    {
+      "epoch": 0.734020618556701,
+      "grad_norm": 2.0339470836036386,
+      "learning_rate": 7.946638596242661e-05,
+      "loss": 0.6435,
+      "step": 267
+    },
+    {
+      "epoch": 0.7367697594501719,
+      "grad_norm": 1.0158450970399455,
+      "learning_rate": 7.945378530266166e-05,
+      "loss": 0.6342,
+      "step": 268
+    },
+    {
+      "epoch": 0.7395189003436426,
+      "grad_norm": 2.328741839753608,
+      "learning_rate": 7.944103862169478e-05,
+      "loss": 0.6496,
+      "step": 269
+    },
+    {
+      "epoch": 0.7422680412371134,
+      "grad_norm": 1.5377779904132143,
+      "learning_rate": 7.94281459667023e-05,
+      "loss": 0.6536,
+      "step": 270
+    },
+    {
+      "epoch": 0.7450171821305842,
+      "grad_norm": 2.1537049583171064,
+      "learning_rate": 7.941510738540086e-05,
+      "loss": 0.6411,
+      "step": 271
+    },
+    {
+      "epoch": 0.747766323024055,
+      "grad_norm": 1.7517265079351534,
+      "learning_rate": 7.940192292604714e-05,
+      "loss": 0.6489,
+      "step": 272
+    },
+    {
+      "epoch": 0.7505154639175258,
+      "grad_norm": 2.1942121182211527,
+      "learning_rate": 7.938859263743776e-05,
+      "loss": 0.6345,
+      "step": 273
+    },
+    {
+      "epoch": 0.7532646048109966,
+      "grad_norm": 1.863240282026407,
+      "learning_rate": 7.937511656890903e-05,
+      "loss": 0.65,
+      "step": 274
+    },
+    {
+      "epoch": 0.7560137457044673,
+      "grad_norm": 1.4298523411886686,
+      "learning_rate": 7.936149477033682e-05,
+      "loss": 0.6391,
+      "step": 275
+    },
+    {
+      "epoch": 0.7587628865979381,
+      "grad_norm": 1.2516811961606267,
+      "learning_rate": 7.934772729213634e-05,
+      "loss": 0.638,
+      "step": 276
+    },
+    {
+      "epoch": 0.7615120274914089,
+      "grad_norm": 1.5514968401316607,
+      "learning_rate": 7.9333814185262e-05,
+      "loss": 0.6464,
+      "step": 277
+    },
+    {
+      "epoch": 0.7642611683848797,
+      "grad_norm": 1.1588148872170223,
+      "learning_rate": 7.931975550120716e-05,
+      "loss": 0.642,
+      "step": 278
+    },
+    {
+      "epoch": 0.7670103092783506,
+      "grad_norm": 2.0239434861105403,
+      "learning_rate": 7.930555129200402e-05,
+      "loss": 0.6428,
+      "step": 279
+    },
+    {
+      "epoch": 0.7697594501718213,
+      "grad_norm": 1.6949238476000754,
+      "learning_rate": 7.929120161022329e-05,
+      "loss": 0.6333,
+      "step": 280
+    },
+    {
+      "epoch": 0.7725085910652921,
+      "grad_norm": 1.2740472233054914,
+      "learning_rate": 7.927670650897421e-05,
+      "loss": 0.6392,
+      "step": 281
+    },
+    {
+      "epoch": 0.7752577319587629,
+      "grad_norm": 1.2888896268808947,
+      "learning_rate": 7.92620660419041e-05,
+      "loss": 0.6324,
+      "step": 282
+    },
+    {
+      "epoch": 0.7780068728522337,
+      "grad_norm": 1.2148413669943177,
+      "learning_rate": 7.924728026319837e-05,
+      "loss": 0.646,
+      "step": 283
+    },
+    {
+      "epoch": 0.7807560137457045,
+      "grad_norm": 0.7141756639924294,
+      "learning_rate": 7.923234922758021e-05,
+      "loss": 0.6403,
+      "step": 284
+    },
+    {
+      "epoch": 0.7835051546391752,
+      "grad_norm": 1.0058322496482315,
+      "learning_rate": 7.921727299031042e-05,
+      "loss": 0.6418,
+      "step": 285
+    },
+    {
+      "epoch": 0.786254295532646,
+      "grad_norm": 0.794851634713387,
+      "learning_rate": 7.920205160718721e-05,
+      "loss": 0.6383,
+      "step": 286
+    },
+    {
+      "epoch": 0.7890034364261168,
+      "grad_norm": 0.7763515931120876,
+      "learning_rate": 7.918668513454598e-05,
+      "loss": 0.629,
+      "step": 287
+    },
+    {
+      "epoch": 0.7917525773195876,
+      "grad_norm": 1.0206658407382474,
+      "learning_rate": 7.917117362925907e-05,
+      "loss": 0.6369,
+      "step": 288
+    },
+    {
+      "epoch": 0.7945017182130584,
+      "grad_norm": 0.9044945170274734,
+      "learning_rate": 7.915551714873571e-05,
+      "loss": 0.6265,
+      "step": 289
+    },
+    {
+      "epoch": 0.7972508591065293,
+      "grad_norm": 1.7861242632179295,
+      "learning_rate": 7.913971575092157e-05,
+      "loss": 0.6354,
+      "step": 290
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.8352377362963846,
+      "learning_rate": 7.912376949429875e-05,
+      "loss": 0.6258,
+      "step": 291
+    },
+    {
+      "epoch": 0.8027491408934708,
+      "grad_norm": 1.1285143530097073,
+      "learning_rate": 7.910767843788543e-05,
+      "loss": 0.6411,
+      "step": 292
+    },
+    {
+      "epoch": 0.8054982817869416,
+      "grad_norm": 2.0986758482850774,
+      "learning_rate": 7.909144264123575e-05,
+      "loss": 0.6432,
+      "step": 293
+    },
+    {
+      "epoch": 0.8082474226804124,
+      "grad_norm": 1.0319530318383152,
+      "learning_rate": 7.90750621644395e-05,
+      "loss": 0.6489,
+      "step": 294
+    },
+    {
+      "epoch": 0.8109965635738832,
+      "grad_norm": 2.628617525239813,
+      "learning_rate": 7.905853706812199e-05,
+      "loss": 0.6439,
+      "step": 295
+    },
+    {
+      "epoch": 0.813745704467354,
+      "grad_norm": 1.977433582449252,
+      "learning_rate": 7.904186741344373e-05,
+      "loss": 0.6482,
+      "step": 296
+    },
+    {
+      "epoch": 0.8164948453608247,
+      "grad_norm": 1.8360229102761563,
+      "learning_rate": 7.902505326210028e-05,
+      "loss": 0.65,
+      "step": 297
+    },
+    {
+      "epoch": 0.8192439862542955,
+      "grad_norm": 1.5433944283254346,
+      "learning_rate": 7.900809467632197e-05,
+      "loss": 0.6434,
+      "step": 298
+    },
+    {
+      "epoch": 0.8219931271477663,
+      "grad_norm": 1.6379518485922502,
+      "learning_rate": 7.899099171887373e-05,
+      "loss": 0.6553,
+      "step": 299
+    },
+    {
+      "epoch": 0.8247422680412371,
+      "grad_norm": 1.477099597856914,
+      "learning_rate": 7.897374445305478e-05,
+      "loss": 0.6392,
+      "step": 300
+    },
+    {
+      "epoch": 0.8274914089347079,
+      "grad_norm": 1.1857730742135408,
+      "learning_rate": 7.895635294269843e-05,
+      "loss": 0.6316,
+      "step": 301
+    },
+    {
+      "epoch": 0.8302405498281787,
+      "grad_norm": 1.361216152623904,
+      "learning_rate": 7.893881725217191e-05,
+      "loss": 0.6352,
+      "step": 302
+    },
+    {
+      "epoch": 0.8329896907216495,
+      "grad_norm": 0.8010015883602253,
+      "learning_rate": 7.892113744637599e-05,
+      "loss": 0.6382,
+      "step": 303
+    },
+    {
+      "epoch": 0.8357388316151203,
+      "grad_norm": 1.102241513039724,
+      "learning_rate": 7.890331359074488e-05,
+      "loss": 0.6447,
+      "step": 304
+    },
+    {
+      "epoch": 0.8384879725085911,
+      "grad_norm": 1.1821390144057153,
+      "learning_rate": 7.888534575124591e-05,
+      "loss": 0.6292,
+      "step": 305
+    },
+    {
+      "epoch": 0.8412371134020619,
+      "grad_norm": 0.7269296274858387,
+      "learning_rate": 7.886723399437931e-05,
+      "loss": 0.6319,
+      "step": 306
+    },
+    {
+      "epoch": 0.8439862542955326,
+      "grad_norm": 1.630523895781267,
+      "learning_rate": 7.884897838717792e-05,
+      "loss": 0.6169,
+      "step": 307
+    },
+    {
+      "epoch": 0.8467353951890034,
+      "grad_norm": 0.740638554856235,
+      "learning_rate": 7.883057899720703e-05,
+      "loss": 0.6355,
+      "step": 308
+    },
+    {
+      "epoch": 0.8494845360824742,
+      "grad_norm": 1.8497443172662162,
+      "learning_rate": 7.881203589256408e-05,
+      "loss": 0.6274,
+      "step": 309
+    },
+    {
+      "epoch": 0.852233676975945,
+      "grad_norm": 1.1910232798072693,
+      "learning_rate": 7.879334914187836e-05,
+      "loss": 0.6392,
+      "step": 310
+    },
+    {
+      "epoch": 0.8549828178694158,
+      "grad_norm": 2.1003839885669815,
+      "learning_rate": 7.877451881431086e-05,
+      "loss": 0.6296,
+      "step": 311
+    },
+    {
+      "epoch": 0.8577319587628865,
+      "grad_norm": 2.041585074702372,
+      "learning_rate": 7.87555449795539e-05,
+      "loss": 0.6352,
+      "step": 312
+    },
+    {
+      "epoch": 0.8604810996563574,
+      "grad_norm": 1.0287133005035243,
+      "learning_rate": 7.873642770783098e-05,
+      "loss": 0.6361,
+      "step": 313
+    },
+    {
+      "epoch": 0.8632302405498282,
+      "grad_norm": 1.4803932962834592,
+      "learning_rate": 7.871716706989645e-05,
+      "loss": 0.6433,
+      "step": 314
+    },
+    {
+      "epoch": 0.865979381443299,
+      "grad_norm": 0.9388113471557215,
+      "learning_rate": 7.869776313703528e-05,
+      "loss": 0.6346,
+      "step": 315
+    },
+    {
+      "epoch": 0.8687285223367698,
+      "grad_norm": 0.8575055727089447,
+      "learning_rate": 7.867821598106275e-05,
+      "loss": 0.6287,
+      "step": 316
+    },
+    {
+      "epoch": 0.8714776632302406,
+      "grad_norm": 0.9753760114729692,
+      "learning_rate": 7.865852567432428e-05,
+      "loss": 0.6386,
+      "step": 317
+    },
+    {
+      "epoch": 0.8742268041237113,
+      "grad_norm": 0.9688066608476287,
+      "learning_rate": 7.863869228969501e-05,
+      "loss": 0.631,
+      "step": 318
+    },
+    {
+      "epoch": 0.8769759450171821,
+      "grad_norm": 0.9593995798512949,
+      "learning_rate": 7.861871590057971e-05,
+      "loss": 0.6347,
+      "step": 319
+    },
+    {
+      "epoch": 0.8797250859106529,
+      "grad_norm": 1.2389383575654551,
+      "learning_rate": 7.85985965809124e-05,
+      "loss": 0.6346,
+      "step": 320
+    },
+    {
+      "epoch": 0.8824742268041237,
+      "grad_norm": 0.8827816805243792,
+      "learning_rate": 7.857833440515605e-05,
+      "loss": 0.6265,
+      "step": 321
+    },
+    {
+      "epoch": 0.8852233676975945,
+      "grad_norm": 0.6061777594451044,
+      "learning_rate": 7.85579294483024e-05,
+      "loss": 0.6286,
+      "step": 322
+    },
+    {
+      "epoch": 0.8879725085910652,
+      "grad_norm": 0.6113638623646376,
+      "learning_rate": 7.85373817858716e-05,
+      "loss": 0.6256,
+      "step": 323
+    },
+    {
+      "epoch": 0.8907216494845361,
+      "grad_norm": 0.8948079812828788,
+      "learning_rate": 7.851669149391198e-05,
+      "loss": 0.6238,
+      "step": 324
+    },
+    {
+      "epoch": 0.8934707903780069,
+      "grad_norm": 1.4121615992032686,
+      "learning_rate": 7.849585864899976e-05,
+      "loss": 0.633,
+      "step": 325
+    },
+    {
+      "epoch": 0.8962199312714777,
+      "grad_norm": 0.7940592789567055,
+      "learning_rate": 7.847488332823873e-05,
+      "loss": 0.6359,
+      "step": 326
+    },
+    {
+      "epoch": 0.8989690721649485,
+      "grad_norm": 0.5764914369379771,
+      "learning_rate": 7.845376560926002e-05,
+      "loss": 0.6375,
+      "step": 327
+    },
+    {
+      "epoch": 0.9017182130584193,
+      "grad_norm": 1.071092163645109,
+      "learning_rate": 7.843250557022177e-05,
+      "loss": 0.6196,
+      "step": 328
+    },
+    {
+      "epoch": 0.90446735395189,
+      "grad_norm": 1.6020470680476175,
+      "learning_rate": 7.841110328980887e-05,
+      "loss": 0.6293,
+      "step": 329
+    },
+    {
+      "epoch": 0.9072164948453608,
+      "grad_norm": 0.4752698709356062,
+      "learning_rate": 7.838955884723265e-05,
+      "loss": 0.626,
+      "step": 330
+    },
+    {
+      "epoch": 0.9099656357388316,
+      "grad_norm": 1.5726092706691321,
+      "learning_rate": 7.836787232223058e-05,
+      "loss": 0.6294,
+      "step": 331
+    },
+    {
+      "epoch": 0.9127147766323024,
+      "grad_norm": 1.01011634702882,
+      "learning_rate": 7.8346043795066e-05,
+      "loss": 0.6431,
+      "step": 332
+    },
+    {
+      "epoch": 0.9154639175257732,
+      "grad_norm": 0.8236758132081481,
+      "learning_rate": 7.83240733465278e-05,
+      "loss": 0.6237,
+      "step": 333
+    },
+    {
+      "epoch": 0.9182130584192439,
+      "grad_norm": 0.816087984512738,
+      "learning_rate": 7.830196105793017e-05,
+      "loss": 0.6304,
+      "step": 334
+    },
+    {
+      "epoch": 0.9209621993127147,
+      "grad_norm": 0.7514093388017495,
+      "learning_rate": 7.827970701111219e-05,
+      "loss": 0.6311,
+      "step": 335
+    },
+    {
+      "epoch": 0.9237113402061856,
+      "grad_norm": 0.9502215222777328,
+      "learning_rate": 7.825731128843762e-05,
+      "loss": 0.6317,
+      "step": 336
+    },
+    {
+      "epoch": 0.9264604810996564,
+      "grad_norm": 1.2376836455505118,
+      "learning_rate": 7.823477397279464e-05,
+      "loss": 0.6199,
+      "step": 337
+    },
+    {
+      "epoch": 0.9292096219931272,
+      "grad_norm": 0.7535199306698559,
+      "learning_rate": 7.821209514759539e-05,
+      "loss": 0.6248,
+      "step": 338
+    },
+    {
+      "epoch": 0.931958762886598,
+      "grad_norm": 0.8490031147278277,
+      "learning_rate": 7.818927489677577e-05,
+      "loss": 0.6267,
+      "step": 339
+    },
+    {
+      "epoch": 0.9347079037800687,
+      "grad_norm": 1.3764847868272945,
+      "learning_rate": 7.816631330479514e-05,
+      "loss": 0.6307,
+      "step": 340
+    },
+    {
+      "epoch": 0.9374570446735395,
+      "grad_norm": 0.6440823293174261,
+      "learning_rate": 7.814321045663594e-05,
+      "loss": 0.6366,
+      "step": 341
+    },
+    {
+      "epoch": 0.9402061855670103,
+      "grad_norm": 1.117461654132214,
+      "learning_rate": 7.811996643780339e-05,
+      "loss": 0.6284,
+      "step": 342
+    },
+    {
+      "epoch": 0.9429553264604811,
+      "grad_norm": 1.198712384928446,
+      "learning_rate": 7.809658133432526e-05,
+      "loss": 0.6179,
+      "step": 343
+    },
+    {
+      "epoch": 0.9457044673539519,
+      "grad_norm": 0.7800718683120617,
+      "learning_rate": 7.807305523275142e-05,
+      "loss": 0.6246,
+      "step": 344
+    },
+    {
+      "epoch": 0.9484536082474226,
+      "grad_norm": 1.106890600114741,
+      "learning_rate": 7.804938822015361e-05,
+      "loss": 0.6186,
+      "step": 345
+    },
+    {
+      "epoch": 0.9512027491408934,
+      "grad_norm": 1.1783968723372122,
+      "learning_rate": 7.802558038412509e-05,
+      "loss": 0.6249,
+      "step": 346
+    },
+    {
+      "epoch": 0.9539518900343643,
+      "grad_norm": 0.909115741821025,
+      "learning_rate": 7.800163181278033e-05,
+      "loss": 0.6244,
+      "step": 347
+    },
+    {
+      "epoch": 0.9567010309278351,
+      "grad_norm": 1.0133199245495836,
+      "learning_rate": 7.797754259475464e-05,
+      "loss": 0.6264,
+      "step": 348
+    },
+    {
+      "epoch": 0.9594501718213059,
+      "grad_norm": 1.195705135014382,
+      "learning_rate": 7.795331281920387e-05,
+      "loss": 0.6278,
+      "step": 349
+    },
+    {
+      "epoch": 0.9621993127147767,
+      "grad_norm": 0.7727285070098746,
+      "learning_rate": 7.792894257580415e-05,
+      "loss": 0.6322,
+      "step": 350
+    },
+    {
+      "epoch": 0.9649484536082474,
+      "grad_norm": 0.9523131612793756,
+      "learning_rate": 7.790443195475142e-05,
+      "loss": 0.6251,
+      "step": 351
+    },
+    {
+      "epoch": 0.9676975945017182,
+      "grad_norm": 1.3036288679788564,
+      "learning_rate": 7.78797810467612e-05,
+      "loss": 0.6218,
+      "step": 352
+    },
+    {
+      "epoch": 0.970446735395189,
+      "grad_norm": 0.5634490374165979,
+      "learning_rate": 7.785498994306821e-05,
+      "loss": 0.6221,
+      "step": 353
+    },
+    {
+      "epoch": 0.9731958762886598,
+      "grad_norm": 1.5260597480048004,
+      "learning_rate": 7.783005873542605e-05,
+      "loss": 0.619,
+      "step": 354
+    },
+    {
+      "epoch": 0.9759450171821306,
+      "grad_norm": 0.6409092943149695,
+      "learning_rate": 7.780498751610684e-05,
+      "loss": 0.6251,
+      "step": 355
+    },
+    {
+      "epoch": 0.9786941580756013,
+      "grad_norm": 1.6664465662420271,
+      "learning_rate": 7.777977637790092e-05,
+      "loss": 0.6357,
+      "step": 356
+    },
+    {
+      "epoch": 0.9814432989690721,
+      "grad_norm": 0.928708660543268,
+      "learning_rate": 7.775442541411647e-05,
+      "loss": 0.6218,
+      "step": 357
+    },
+    {
+      "epoch": 0.9841924398625429,
+      "grad_norm": 1.8449391562901856,
+      "learning_rate": 7.772893471857915e-05,
+      "loss": 0.6396,
+      "step": 358
+    },
+    {
+      "epoch": 0.9869415807560138,
+      "grad_norm": 1.6259312779533432,
+      "learning_rate": 7.77033043856318e-05,
+      "loss": 0.6368,
+      "step": 359
+    },
+    {
+      "epoch": 0.9896907216494846,
+      "grad_norm": 0.9340296294033434,
+      "learning_rate": 7.767753451013408e-05,
+      "loss": 0.6204,
+      "step": 360
+    },
+    {
+      "epoch": 0.9924398625429554,
+      "grad_norm": 1.2307139404659375,
+      "learning_rate": 7.765162518746207e-05,
+      "loss": 0.6177,
+      "step": 361
+    },
+    {
+      "epoch": 0.9951890034364261,
+      "grad_norm": 0.7744777080187526,
+      "learning_rate": 7.762557651350798e-05,
+      "loss": 0.6288,
+      "step": 362
+    },
+    {
+      "epoch": 0.9979381443298969,
+      "grad_norm": 1.0676481742466821,
+      "learning_rate": 7.759938858467979e-05,
+      "loss": 0.6272,
+      "step": 363
+    },
+    {
+      "epoch": 1.0006872852233677,
+      "grad_norm": 0.9232629543637857,
+      "learning_rate": 7.757306149790082e-05,
+      "loss": 0.7685,
+      "step": 364
+    },
+    {
+      "epoch": 1.0034364261168385,
+      "grad_norm": 1.0678574538408068,
+      "learning_rate": 7.754659535060943e-05,
+      "loss": 0.5916,
+      "step": 365
+    },
+    {
+      "epoch": 1.0061855670103093,
+      "grad_norm": 1.2551799774905172,
+      "learning_rate": 7.751999024075871e-05,
+      "loss": 0.5988,
+      "step": 366
+    },
+    {
+      "epoch": 1.00893470790378,
+      "grad_norm": 0.9263667600041875,
+      "learning_rate": 7.749324626681599e-05,
+      "loss": 0.5966,
+      "step": 367
+    },
+    {
+      "epoch": 1.0116838487972508,
+      "grad_norm": 0.731842781771422,
+      "learning_rate": 7.746636352776259e-05,
+      "loss": 0.5928,
+      "step": 368
+    },
+    {
+      "epoch": 1.0144329896907216,
+      "grad_norm": 0.8372632610382791,
+      "learning_rate": 7.74393421230934e-05,
+      "loss": 0.5919,
+      "step": 369
+    },
+    {
+      "epoch": 1.0171821305841924,
+      "grad_norm": 1.0188957663159135,
+      "learning_rate": 7.741218215281652e-05,
+      "loss": 0.5932,
+      "step": 370
+    },
+    {
+      "epoch": 1.0199312714776632,
+      "grad_norm": 1.3394999167844803,
+      "learning_rate": 7.738488371745287e-05,
+      "loss": 0.5879,
+      "step": 371
+    },
+    {
+      "epoch": 1.022680412371134,
+      "grad_norm": 0.8980665393748412,
+      "learning_rate": 7.735744691803588e-05,
+      "loss": 0.5942,
+      "step": 372
+    },
+    {
+      "epoch": 1.0254295532646047,
+      "grad_norm": 0.9856701598024391,
+      "learning_rate": 7.732987185611102e-05,
+      "loss": 0.5793,
+      "step": 373
+    },
+    {
+      "epoch": 1.0281786941580755,
+      "grad_norm": 0.8156719105288647,
+      "learning_rate": 7.730215863373554e-05,
+      "loss": 0.5953,
+      "step": 374
+    },
+    {
+      "epoch": 1.0309278350515463,
+      "grad_norm": 0.5095314934411239,
+      "learning_rate": 7.727430735347799e-05,
+      "loss": 0.5848,
+      "step": 375
+    },
+    {
+      "epoch": 1.0336769759450173,
+      "grad_norm": 0.5914859323410439,
+      "learning_rate": 7.724631811841789e-05,
+      "loss": 0.5877,
+      "step": 376
+    },
+    {
+      "epoch": 1.036426116838488,
+      "grad_norm": 0.6410263828936001,
+      "learning_rate": 7.721819103214536e-05,
+      "loss": 0.5852,
+      "step": 377
+    },
+    {
+      "epoch": 1.0391752577319588,
+      "grad_norm": 0.9724802661391164,
+      "learning_rate": 7.71899261987607e-05,
+      "loss": 0.5949,
+      "step": 378
+    },
+    {
+      "epoch": 1.0419243986254296,
+      "grad_norm": 1.3805105132332318,
+      "learning_rate": 7.716152372287399e-05,
+      "loss": 0.5912,
+      "step": 379
+    },
+    {
+      "epoch": 1.0446735395189004,
+      "grad_norm": 0.5961514316349491,
+      "learning_rate": 7.713298370960481e-05,
+      "loss": 0.5989,
+      "step": 380
+    },
+    {
+      "epoch": 1.0474226804123712,
+      "grad_norm": 1.0381265315374555,
+      "learning_rate": 7.710430626458171e-05,
+      "loss": 0.5853,
+      "step": 381
+    },
+    {
+      "epoch": 1.050171821305842,
+      "grad_norm": 0.88611297798641,
+      "learning_rate": 7.707549149394192e-05,
+      "loss": 0.5971,
+      "step": 382
+    },
+    {
+      "epoch": 1.0529209621993127,
+      "grad_norm": 0.9894621130663098,
+      "learning_rate": 7.704653950433092e-05,
+      "loss": 0.5839,
+      "step": 383
+    },
+    {
+      "epoch": 1.0556701030927835,
+      "grad_norm": 1.4525179754862534,
+      "learning_rate": 7.7017450402902e-05,
+      "loss": 0.5841,
+      "step": 384
+    },
+    {
+      "epoch": 1.0584192439862543,
+      "grad_norm": 0.6031744444046739,
+      "learning_rate": 7.698822429731595e-05,
+      "loss": 0.577,
+      "step": 385
+    },
+    {
+      "epoch": 1.061168384879725,
+      "grad_norm": 1.0814979553674235,
+      "learning_rate": 7.695886129574065e-05,
+      "loss": 0.5963,
+      "step": 386
+    },
+    {
+      "epoch": 1.0639175257731959,
+      "grad_norm": 1.2035521726093799,
+      "learning_rate": 7.692936150685059e-05,
+      "loss": 0.59,
+      "step": 387
+    },
+    {
+      "epoch": 1.0666666666666667,
+      "grad_norm": 1.065535271974634,
+      "learning_rate": 7.689972503982654e-05,
+      "loss": 0.59,
+      "step": 388
+    },
+    {
+      "epoch": 1.0694158075601374,
+      "grad_norm": 0.9885511702158032,
+      "learning_rate": 7.686995200435513e-05,
+      "loss": 0.5908,
+      "step": 389
+    },
+    {
+      "epoch": 1.0721649484536082,
+      "grad_norm": 1.0056851158221545,
+      "learning_rate": 7.684004251062844e-05,
+      "loss": 0.592,
+      "step": 390
+    },
+    {
+      "epoch": 1.074914089347079,
+      "grad_norm": 1.0202284999862656,
+      "learning_rate": 7.680999666934358e-05,
+      "loss": 0.5809,
+      "step": 391
+    },
+    {
+      "epoch": 1.0776632302405498,
+      "grad_norm": 1.1513878865642915,
+      "learning_rate": 7.677981459170228e-05,
+      "loss": 0.5878,
+      "step": 392
+    },
+    {
+      "epoch": 1.0804123711340206,
+      "grad_norm": 0.8699810682515629,
+      "learning_rate": 7.674949638941053e-05,
+      "loss": 0.5915,
+      "step": 393
+    },
+    {
+      "epoch": 1.0831615120274913,
+      "grad_norm": 0.4496053235798414,
+      "learning_rate": 7.671904217467808e-05,
+      "loss": 0.5801,
+      "step": 394
+    },
+    {
+      "epoch": 1.0859106529209621,
+      "grad_norm": 0.6745209087686392,
+      "learning_rate": 7.668845206021812e-05,
+      "loss": 0.5979,
+      "step": 395
+    },
+    {
+      "epoch": 1.088659793814433,
+      "grad_norm": 1.2754326592180227,
+      "learning_rate": 7.66577261592468e-05,
+      "loss": 0.5807,
+      "step": 396
+    },
+    {
+      "epoch": 1.0914089347079037,
+      "grad_norm": 0.7069250995821895,
+      "learning_rate": 7.662686458548276e-05,
+      "loss": 0.582,
+      "step": 397
+    },
+    {
+      "epoch": 1.0941580756013747,
+      "grad_norm": 0.4973387506195001,
+      "learning_rate": 7.659586745314689e-05,
+      "loss": 0.5855,
+      "step": 398
+    },
+    {
+      "epoch": 1.0969072164948455,
+      "grad_norm": 0.5388764587620251,
+      "learning_rate": 7.65647348769617e-05,
+      "loss": 0.5901,
+      "step": 399
+    },
+    {
+      "epoch": 1.0996563573883162,
+      "grad_norm": 0.5432430222716064,
+      "learning_rate": 7.653346697215103e-05,
+      "loss": 0.5852,
+      "step": 400
+    },
+    {
+      "epoch": 1.102405498281787,
+      "grad_norm": 0.7845045531047907,
+      "learning_rate": 7.650206385443958e-05,
+      "loss": 0.5815,
+      "step": 401
+    },
+    {
+      "epoch": 1.1051546391752578,
+      "grad_norm": 1.1593889092948502,
+      "learning_rate": 7.647052564005243e-05,
+      "loss": 0.5884,
+      "step": 402
+    },
+    {
+      "epoch": 1.1079037800687286,
+      "grad_norm": 0.9849246083500859,
+      "learning_rate": 7.643885244571474e-05,
+      "loss": 0.5884,
+      "step": 403
+    },
+    {
+      "epoch": 1.1106529209621994,
+      "grad_norm": 1.2327221523818575,
+      "learning_rate": 7.64070443886512e-05,
+      "loss": 0.5808,
+      "step": 404
+    },
+    {
+      "epoch": 1.1134020618556701,
+      "grad_norm": 0.8894760613915432,
+      "learning_rate": 7.637510158658563e-05,
+      "loss": 0.5808,
+      "step": 405
+    },
+    {
+      "epoch": 1.116151202749141,
+      "grad_norm": 0.8458010101721215,
+      "learning_rate": 7.634302415774057e-05,
+      "loss": 0.5921,
+      "step": 406
+    },
+    {
+      "epoch": 1.1189003436426117,
+      "grad_norm": 1.0269578881644645,
+      "learning_rate": 7.63108122208368e-05,
+      "loss": 0.5876,
+      "step": 407
+    },
+    {
+      "epoch": 1.1216494845360825,
+      "grad_norm": 0.9883024415526525,
+      "learning_rate": 7.627846589509295e-05,
+      "loss": 0.5938,
+      "step": 408
+    },
+    {
+      "epoch": 1.1243986254295533,
+      "grad_norm": 1.0374738987854506,
+      "learning_rate": 7.624598530022502e-05,
+      "loss": 0.5759,
+      "step": 409
+    },
+    {
+      "epoch": 1.127147766323024,
+      "grad_norm": 0.8968993730271081,
+      "learning_rate": 7.621337055644596e-05,
+      "loss": 0.5971,
+      "step": 410
+    },
+    {
+      "epoch": 1.1298969072164948,
+      "grad_norm": 0.645336647760953,
+      "learning_rate": 7.618062178446518e-05,
+      "loss": 0.5843,
+      "step": 411
+    },
+    {
+      "epoch": 1.1326460481099656,
+      "grad_norm": 0.8156198060430901,
+      "learning_rate": 7.614773910548816e-05,
+      "loss": 0.5834,
+      "step": 412
+    },
+    {
+      "epoch": 1.1353951890034364,
+      "grad_norm": 0.9969205762211508,
+      "learning_rate": 7.6114722641216e-05,
+      "loss": 0.5874,
+      "step": 413
+    },
+    {
+      "epoch": 1.1381443298969072,
+      "grad_norm": 0.9512507581396507,
+      "learning_rate": 7.608157251384493e-05,
+      "loss": 0.5907,
+      "step": 414
+    },
+    {
+      "epoch": 1.140893470790378,
+      "grad_norm": 0.46469337738492333,
+      "learning_rate": 7.604828884606587e-05,
+      "loss": 0.5854,
+      "step": 415
+    },
+    {
+      "epoch": 1.1436426116838487,
+      "grad_norm": 0.6875741181821637,
+      "learning_rate": 7.601487176106397e-05,
+      "loss": 0.5907,
+      "step": 416
+    },
+    {
+      "epoch": 1.1463917525773195,
+      "grad_norm": 0.8229065410063605,
+      "learning_rate": 7.59813213825182e-05,
+      "loss": 0.5835,
+      "step": 417
+    },
+    {
+      "epoch": 1.1491408934707903,
+      "grad_norm": 0.5535331052196206,
+      "learning_rate": 7.594763783460086e-05,
+      "loss": 0.5913,
+      "step": 418
+    },
+    {
+      "epoch": 1.151890034364261,
+      "grad_norm": 0.609930416365321,
+      "learning_rate": 7.591382124197708e-05,
+      "loss": 0.5861,
+      "step": 419
+    },
+    {
+      "epoch": 1.1546391752577319,
+      "grad_norm": 0.549391385583707,
+      "learning_rate": 7.587987172980443e-05,
+      "loss": 0.5774,
+      "step": 420
+    },
+    {
+      "epoch": 1.1573883161512026,
+      "grad_norm": 0.5299626474050247,
+      "learning_rate": 7.58457894237324e-05,
+      "loss": 0.5915,
+      "step": 421
+    },
+    {
+      "epoch": 1.1601374570446734,
+      "grad_norm": 0.5702316031822813,
+      "learning_rate": 7.581157444990199e-05,
+      "loss": 0.5771,
+      "step": 422
+    },
+    {
+      "epoch": 1.1628865979381444,
+      "grad_norm": 0.8908922680435231,
+      "learning_rate": 7.577722693494519e-05,
+      "loss": 0.5932,
+      "step": 423
+    },
+    {
+      "epoch": 1.1656357388316152,
+      "grad_norm": 1.4431986503442844,
+      "learning_rate": 7.574274700598453e-05,
+      "loss": 0.5813,
+      "step": 424
+    },
+    {
+      "epoch": 1.168384879725086,
+      "grad_norm": 0.6369720186684446,
+      "learning_rate": 7.570813479063265e-05,
+      "loss": 0.5845,
+      "step": 425
+    },
+    {
+      "epoch": 1.1711340206185568,
+      "grad_norm": 0.5256518804042303,
+      "learning_rate": 7.567339041699175e-05,
+      "loss": 0.5924,
+      "step": 426
+    },
+    {
+      "epoch": 1.1738831615120275,
+      "grad_norm": 0.8165966324851138,
+      "learning_rate": 7.563851401365316e-05,
+      "loss": 0.588,
+      "step": 427
+    },
+    {
+      "epoch": 1.1766323024054983,
+      "grad_norm": 1.1011809927414538,
+      "learning_rate": 7.56035057096969e-05,
+      "loss": 0.5878,
+      "step": 428
+    },
+    {
+      "epoch": 1.179381443298969,
+      "grad_norm": 1.2202082201319733,
+      "learning_rate": 7.556836563469111e-05,
+      "loss": 0.589,
+      "step": 429
+    },
+    {
+      "epoch": 1.1821305841924399,
+      "grad_norm": 0.8269982439978933,
+      "learning_rate": 7.553309391869167e-05,
+      "loss": 0.5846,
+      "step": 430
+    },
+    {
+      "epoch": 1.1848797250859107,
+      "grad_norm": 1.0285455684872933,
+      "learning_rate": 7.549769069224164e-05,
+      "loss": 0.5883,
+      "step": 431
+    },
+    {
+      "epoch": 1.1876288659793814,
+      "grad_norm": 1.4122193045161404,
+      "learning_rate": 7.546215608637083e-05,
+      "loss": 0.5832,
+      "step": 432
+    },
+    {
+      "epoch": 1.1903780068728522,
+      "grad_norm": 0.3041741502072138,
+      "learning_rate": 7.542649023259527e-05,
+      "loss": 0.5813,
+      "step": 433
+    },
+    {
+      "epoch": 1.193127147766323,
+      "grad_norm": 1.5989692268879376,
+      "learning_rate": 7.53906932629168e-05,
+      "loss": 0.5896,
+      "step": 434
+    },
+    {
+      "epoch": 1.1958762886597938,
+      "grad_norm": 0.5519101760067153,
+      "learning_rate": 7.535476530982244e-05,
+      "loss": 0.5719,
+      "step": 435
+    },
+    {
+      "epoch": 1.1986254295532646,
+      "grad_norm": 0.770417484741293,
+      "learning_rate": 7.53187065062841e-05,
+      "loss": 0.5881,
+      "step": 436
+    },
+    {
+      "epoch": 1.2013745704467353,
+      "grad_norm": 1.3484068230851232,
+      "learning_rate": 7.528251698575788e-05,
+      "loss": 0.5868,
+      "step": 437
+    },
+    {
+      "epoch": 1.2041237113402061,
+      "grad_norm": 0.5995693974246964,
+      "learning_rate": 7.524619688218372e-05,
+      "loss": 0.5812,
+      "step": 438
+    },
+    {
+      "epoch": 1.206872852233677,
+      "grad_norm": 0.9168760247232807,
+      "learning_rate": 7.520974632998485e-05,
+      "loss": 0.5877,
+      "step": 439
+    },
+    {
+      "epoch": 1.2096219931271477,
+      "grad_norm": 1.1013744178468492,
+      "learning_rate": 7.51731654640673e-05,
+      "loss": 0.592,
+      "step": 440
+    },
+    {
+      "epoch": 1.2123711340206185,
+      "grad_norm": 0.5790863138582544,
+      "learning_rate": 7.51364544198194e-05,
+      "loss": 0.5749,
+      "step": 441
+    },
+    {
+      "epoch": 1.2151202749140895,
+      "grad_norm": 0.5723305467246104,
+      "learning_rate": 7.509961333311126e-05,
+      "loss": 0.5905,
+      "step": 442
+    },
+    {
+      "epoch": 1.2178694158075603,
+      "grad_norm": 0.6893985276435765,
+      "learning_rate": 7.506264234029432e-05,
+      "loss": 0.5799,
+      "step": 443
+    },
+    {
+      "epoch": 1.220618556701031,
+      "grad_norm": 0.6306086582019484,
+      "learning_rate": 7.502554157820079e-05,
+      "loss": 0.583,
+      "step": 444
+    },
+    {
+      "epoch": 1.2233676975945018,
+      "grad_norm": 0.5828509036657884,
+      "learning_rate": 7.498831118414316e-05,
+      "loss": 0.5791,
+      "step": 445
+    },
+    {
+      "epoch": 1.2261168384879726,
+      "grad_norm": 0.9766394221436852,
+      "learning_rate": 7.495095129591373e-05,
+      "loss": 0.5876,
+      "step": 446
+    },
+    {
+      "epoch": 1.2288659793814434,
+      "grad_norm": 1.416296834720271,
+      "learning_rate": 7.4913462051784e-05,
+      "loss": 0.5876,
+      "step": 447
+    },
+    {
+      "epoch": 1.2316151202749142,
+      "grad_norm": 0.6702851111393332,
+      "learning_rate": 7.487584359050431e-05,
+      "loss": 0.5881,
+      "step": 448
+    },
+    {
+      "epoch": 1.234364261168385,
+      "grad_norm": 0.9384647234340839,
+      "learning_rate": 7.483809605130319e-05,
+      "loss": 0.5959,
+      "step": 449
+    },
+    {
+      "epoch": 1.2371134020618557,
+      "grad_norm": 1.2907110973467573,
+      "learning_rate": 7.480021957388691e-05,
+      "loss": 0.5785,
+      "step": 450
+    },
+    {
+      "epoch": 1.2398625429553265,
+      "grad_norm": 0.8298126531860124,
+      "learning_rate": 7.476221429843894e-05,
+      "loss": 0.5934,
+      "step": 451
+    },
+    {
+      "epoch": 1.2426116838487973,
+      "grad_norm": 1.1127124873762135,
+      "learning_rate": 7.472408036561942e-05,
+      "loss": 0.5833,
+      "step": 452
+    },
+    {
+      "epoch": 1.245360824742268,
+      "grad_norm": 0.7148549185439407,
+      "learning_rate": 7.468581791656473e-05,
+      "loss": 0.5927,
+      "step": 453
+    },
+    {
+      "epoch": 1.2481099656357388,
+      "grad_norm": 0.863808291677187,
+      "learning_rate": 7.464742709288683e-05,
+      "loss": 0.5955,
+      "step": 454
+    },
+    {
+      "epoch": 1.2508591065292096,
+      "grad_norm": 0.8446688366988994,
+      "learning_rate": 7.460890803667283e-05,
+      "loss": 0.5939,
+      "step": 455
+    },
+    {
+      "epoch": 1.2536082474226804,
+      "grad_norm": 1.0639580002350018,
+      "learning_rate": 7.457026089048445e-05,
+      "loss": 0.5839,
+      "step": 456
+    },
+    {
+      "epoch": 1.2563573883161512,
+      "grad_norm": 1.1091140834360853,
+      "learning_rate": 7.453148579735743e-05,
+      "loss": 0.5829,
+      "step": 457
+    },
+    {
+      "epoch": 1.259106529209622,
+      "grad_norm": 0.80462944337638,
+      "learning_rate": 7.449258290080112e-05,
+      "loss": 0.5845,
+      "step": 458
+    },
+    {
+      "epoch": 1.2618556701030927,
+      "grad_norm": 0.8412084155727457,
+      "learning_rate": 7.445355234479783e-05,
+      "loss": 0.5767,
+      "step": 459
+    },
+    {
+      "epoch": 1.2646048109965635,
+      "grad_norm": 0.6330472076770527,
+      "learning_rate": 7.441439427380235e-05,
+      "loss": 0.5816,
+      "step": 460
+    },
+    {
+      "epoch": 1.2673539518900343,
+      "grad_norm": 0.597903788175847,
+      "learning_rate": 7.437510883274144e-05,
+      "loss": 0.5729,
+      "step": 461
+    },
+    {
+      "epoch": 1.270103092783505,
+      "grad_norm": 0.4287714351702764,
+      "learning_rate": 7.433569616701324e-05,
+      "loss": 0.5932,
+      "step": 462
+    },
+    {
+      "epoch": 1.2728522336769759,
+      "grad_norm": 0.47550714789032805,
+      "learning_rate": 7.429615642248677e-05,
+      "loss": 0.5929,
+      "step": 463
+    },
+    {
+      "epoch": 1.2756013745704466,
+      "grad_norm": 0.5414468519874974,
+      "learning_rate": 7.425648974550138e-05,
+      "loss": 0.5774,
+      "step": 464
+    },
+    {
+      "epoch": 1.2783505154639174,
+      "grad_norm": 0.5640877600721679,
+      "learning_rate": 7.421669628286617e-05,
+      "loss": 0.586,
+      "step": 465
+    },
+    {
+      "epoch": 1.2810996563573882,
+      "grad_norm": 0.6048691932188638,
+      "learning_rate": 7.417677618185955e-05,
+      "loss": 0.597,
+      "step": 466
+    },
+    {
+      "epoch": 1.283848797250859,
+      "grad_norm": 0.7508265029471074,
+      "learning_rate": 7.413672959022856e-05,
+      "loss": 0.5961,
+      "step": 467
+    },
+    {
+      "epoch": 1.2865979381443298,
+      "grad_norm": 1.066042006928496,
+      "learning_rate": 7.409655665618843e-05,
+      "loss": 0.5837,
+      "step": 468
+    },
+    {
+      "epoch": 1.2893470790378008,
+      "grad_norm": 1.4460442111626333,
+      "learning_rate": 7.405625752842198e-05,
+      "loss": 0.5771,
+      "step": 469
+    },
+    {
+      "epoch": 1.2920962199312716,
+      "grad_norm": 0.40527253568707877,
+      "learning_rate": 7.401583235607908e-05,
+      "loss": 0.5769,
+      "step": 470
+    },
+    {
+      "epoch": 1.2948453608247423,
+      "grad_norm": 0.8934936601794983,
+      "learning_rate": 7.397528128877611e-05,
+      "loss": 0.585,
+      "step": 471
+    },
+    {
+      "epoch": 1.2975945017182131,
+      "grad_norm": 1.6705979909682174,
+      "learning_rate": 7.393460447659539e-05,
+      "loss": 0.5932,
+      "step": 472
+    },
+    {
+      "epoch": 1.300343642611684,
+      "grad_norm": 0.4422750143504963,
+      "learning_rate": 7.389380207008462e-05,
+      "loss": 0.5878,
+      "step": 473
+    },
+    {
+      "epoch": 1.3030927835051547,
+      "grad_norm": 1.5622517777634235,
+      "learning_rate": 7.385287422025635e-05,
+      "loss": 0.5895,
+      "step": 474
+    },
+    {
+      "epoch": 1.3058419243986255,
+      "grad_norm": 0.6900852059622544,
+      "learning_rate": 7.381182107858738e-05,
+      "loss": 0.5883,
+      "step": 475
+    },
+    {
+      "epoch": 1.3085910652920962,
+      "grad_norm": 0.9882097928385363,
+      "learning_rate": 7.377064279701827e-05,
+      "loss": 0.5894,
+      "step": 476
+    },
+    {
+      "epoch": 1.311340206185567,
+      "grad_norm": 0.9430491893537721,
+      "learning_rate": 7.37293395279527e-05,
+      "loss": 0.5921,
+      "step": 477
+    },
+    {
+      "epoch": 1.3140893470790378,
+      "grad_norm": 0.7682283056356898,
+      "learning_rate": 7.368791142425691e-05,
+      "loss": 0.5977,
+      "step": 478
+    },
+    {
+      "epoch": 1.3168384879725086,
+      "grad_norm": 1.0703629836228026,
+      "learning_rate": 7.364635863925922e-05,
+      "loss": 0.5829,
+      "step": 479
+    },
+    {
+      "epoch": 1.3195876288659794,
+      "grad_norm": 0.8976152595403926,
+      "learning_rate": 7.360468132674935e-05,
+      "loss": 0.5902,
+      "step": 480
+    },
+    {
+      "epoch": 1.3223367697594501,
+      "grad_norm": 0.7540663375222781,
+      "learning_rate": 7.356287964097795e-05,
+      "loss": 0.588,
+      "step": 481
+    },
+    {
+      "epoch": 1.325085910652921,
+      "grad_norm": 1.1037972218662753,
+      "learning_rate": 7.352095373665598e-05,
+      "loss": 0.5814,
+      "step": 482
+    },
+    {
+      "epoch": 1.3278350515463917,
+      "grad_norm": 0.6721571418198156,
+      "learning_rate": 7.347890376895407e-05,
+      "loss": 0.586,
+      "step": 483
+    },
+    {
+      "epoch": 1.3305841924398625,
+      "grad_norm": 0.6580601865881139,
+      "learning_rate": 7.343672989350214e-05,
+      "loss": 0.5726,
+      "step": 484
+    },
+    {
+      "epoch": 1.3333333333333333,
+      "grad_norm": 0.9232629737168565,
+      "learning_rate": 7.33944322663886e-05,
+      "loss": 0.5841,
+      "step": 485
+    },
+    {
+      "epoch": 1.3360824742268043,
+      "grad_norm": 0.921611093133066,
+      "learning_rate": 7.335201104415992e-05,
+      "loss": 0.5961,
+      "step": 486
+    },
+    {
+      "epoch": 1.338831615120275,
+      "grad_norm": 0.7555439012244609,
+      "learning_rate": 7.330946638381998e-05,
+      "loss": 0.5877,
+      "step": 487
+    },
+    {
+      "epoch": 1.3415807560137458,
+      "grad_norm": 0.8027350119891178,
+      "learning_rate": 7.326679844282953e-05,
+      "loss": 0.5917,
+      "step": 488
+    },
+    {
+      "epoch": 1.3443298969072166,
+      "grad_norm": 1.1401383602299868,
+      "learning_rate": 7.322400737910558e-05,
+      "loss": 0.5852,
+      "step": 489
+    },
+    {
+      "epoch": 1.3470790378006874,
+      "grad_norm": 0.805628145096471,
+      "learning_rate": 7.318109335102083e-05,
+      "loss": 0.5916,
+      "step": 490
+    },
+    {
+      "epoch": 1.3498281786941582,
+      "grad_norm": 0.5776321913613638,
+      "learning_rate": 7.31380565174031e-05,
+      "loss": 0.5846,
+      "step": 491
+    },
+    {
+      "epoch": 1.352577319587629,
+      "grad_norm": 0.730623963485602,
+      "learning_rate": 7.309489703753471e-05,
+      "loss": 0.5859,
+      "step": 492
+    },
+    {
+      "epoch": 1.3553264604810997,
+      "grad_norm": 0.6579197440168587,
+      "learning_rate": 7.305161507115185e-05,
+      "loss": 0.5796,
+      "step": 493
+    },
+    {
+      "epoch": 1.3580756013745705,
+      "grad_norm": 0.5630346265554919,
+      "learning_rate": 7.300821077844413e-05,
+      "loss": 0.5822,
+      "step": 494
+    },
+    {
+      "epoch": 1.3608247422680413,
+      "grad_norm": 0.7908819829643783,
+      "learning_rate": 7.296468432005382e-05,
+      "loss": 0.5709,
+      "step": 495
+    },
+    {
+      "epoch": 1.363573883161512,
+      "grad_norm": 1.1004873982746464,
+      "learning_rate": 7.292103585707541e-05,
+      "loss": 0.5923,
+      "step": 496
+    },
+    {
+      "epoch": 1.3663230240549828,
+      "grad_norm": 1.0680631626760493,
+      "learning_rate": 7.287726555105485e-05,
+      "loss": 0.5932,
+      "step": 497
+    },
+    {
+      "epoch": 1.3690721649484536,
+      "grad_norm": 0.9112494358591099,
+      "learning_rate": 7.283337356398911e-05,
+      "loss": 0.5785,
+      "step": 498
+    },
+    {
+      "epoch": 1.3718213058419244,
+      "grad_norm": 1.0337260282053968,
+      "learning_rate": 7.278936005832549e-05,
+      "loss": 0.5882,
+      "step": 499
+    },
+    {
+      "epoch": 1.3745704467353952,
+      "grad_norm": 1.3007806381700902,
+      "learning_rate": 7.274522519696102e-05,
+      "loss": 0.5798,
+      "step": 500
+    },
+    {
+      "epoch": 1.377319587628866,
+      "grad_norm": 0.43519904375561846,
+      "learning_rate": 7.270096914324189e-05,
+      "loss": 0.5791,
+      "step": 501
+    },
+    {
+      "epoch": 1.3800687285223368,
+      "grad_norm": 0.8544940867252295,
+      "learning_rate": 7.265659206096285e-05,
+      "loss": 0.5901,
+      "step": 502
+    },
+    {
+      "epoch": 1.3828178694158075,
+      "grad_norm": 1.2429689674231412,
+      "learning_rate": 7.261209411436654e-05,
+      "loss": 0.5809,
+      "step": 503
+    },
+    {
+      "epoch": 1.3855670103092783,
+      "grad_norm": 0.6308748621084995,
+      "learning_rate": 7.256747546814298e-05,
+      "loss": 0.5791,
+      "step": 504
+    },
+    {
+      "epoch": 1.388316151202749,
+      "grad_norm": 0.7148204479546915,
+      "learning_rate": 7.252273628742885e-05,
+      "loss": 0.5796,
+      "step": 505
+    },
+    {
+      "epoch": 1.3910652920962199,
+      "grad_norm": 0.9925558904292439,
+      "learning_rate": 7.2477876737807e-05,
+      "loss": 0.5858,
+      "step": 506
+    },
+    {
+      "epoch": 1.3938144329896907,
+      "grad_norm": 0.8578115038769086,
+      "learning_rate": 7.243289698530572e-05,
+      "loss": 0.5845,
+      "step": 507
+    },
+    {
+      "epoch": 1.3965635738831614,
+      "grad_norm": 0.7229862425355326,
+      "learning_rate": 7.23877971963982e-05,
+      "loss": 0.5755,
+      "step": 508
+    },
+    {
+      "epoch": 1.3993127147766322,
+      "grad_norm": 0.8939365007653768,
+      "learning_rate": 7.234257753800191e-05,
+      "loss": 0.5875,
+      "step": 509
+    },
+    {
+      "epoch": 1.402061855670103,
+      "grad_norm": 0.9280862590900643,
+      "learning_rate": 7.229723817747793e-05,
+      "loss": 0.5916,
+      "step": 510
+    },
+    {
+      "epoch": 1.4048109965635738,
+      "grad_norm": 0.7722106893797273,
+      "learning_rate": 7.225177928263042e-05,
+      "loss": 0.5817,
+      "step": 511
+    },
+    {
+      "epoch": 1.4075601374570446,
+      "grad_norm": 0.5590378429645152,
+      "learning_rate": 7.220620102170585e-05,
+      "loss": 0.5804,
+      "step": 512
+    },
+    {
+      "epoch": 1.4103092783505153,
+      "grad_norm": 0.5415497371424565,
+      "learning_rate": 7.21605035633926e-05,
+      "loss": 0.5829,
+      "step": 513
+    },
+    {
+      "epoch": 1.4130584192439861,
+      "grad_norm": 0.6720235507203421,
+      "learning_rate": 7.21146870768201e-05,
+      "loss": 0.5818,
+      "step": 514
+    },
+    {
+      "epoch": 1.4158075601374571,
+      "grad_norm": 0.5829950782167737,
+      "learning_rate": 7.206875173155834e-05,
+      "loss": 0.5922,
+      "step": 515
+    },
+    {
+      "epoch": 1.418556701030928,
+      "grad_norm": 0.3621229457523989,
+      "learning_rate": 7.202269769761726e-05,
+      "loss": 0.5842,
+      "step": 516
+    },
+    {
+      "epoch": 1.4213058419243987,
+      "grad_norm": 0.5023031289705603,
+      "learning_rate": 7.197652514544601e-05,
+      "loss": 0.576,
+      "step": 517
+    },
+    {
+      "epoch": 1.4240549828178695,
+      "grad_norm": 0.6099312033435136,
+      "learning_rate": 7.193023424593245e-05,
+      "loss": 0.5795,
+      "step": 518
+    },
+    {
+      "epoch": 1.4268041237113402,
+      "grad_norm": 0.5933868127441155,
+      "learning_rate": 7.188382517040237e-05,
+      "loss": 0.5736,
+      "step": 519
+    },
+    {
+      "epoch": 1.429553264604811,
+      "grad_norm": 0.582367928707779,
+      "learning_rate": 7.183729809061904e-05,
+      "loss": 0.5794,
+      "step": 520
+    },
+    {
+      "epoch": 1.4323024054982818,
+      "grad_norm": 0.7208839572689155,
+      "learning_rate": 7.179065317878237e-05,
+      "loss": 0.587,
+      "step": 521
+    },
+    {
+      "epoch": 1.4350515463917526,
+      "grad_norm": 1.0049567135040567,
+      "learning_rate": 7.174389060752845e-05,
+      "loss": 0.5863,
+      "step": 522
+    },
+    {
+      "epoch": 1.4378006872852234,
+      "grad_norm": 1.3116471278229,
+      "learning_rate": 7.16970105499288e-05,
+      "loss": 0.5873,
+      "step": 523
+    },
+    {
+      "epoch": 1.4405498281786941,
+      "grad_norm": 0.7052150815292837,
+      "learning_rate": 7.165001317948976e-05,
+      "loss": 0.5854,
+      "step": 524
+    },
+    {
+      "epoch": 1.443298969072165,
+      "grad_norm": 0.6674498164029434,
+      "learning_rate": 7.160289867015187e-05,
+      "loss": 0.58,
+      "step": 525
+    },
+    {
+      "epoch": 1.4460481099656357,
+      "grad_norm": 0.6479049706905784,
+      "learning_rate": 7.155566719628921e-05,
+      "loss": 0.583,
+      "step": 526
+    },
+    {
+      "epoch": 1.4487972508591065,
+      "grad_norm": 0.6003355447355602,
+      "learning_rate": 7.150831893270874e-05,
+      "loss": 0.5792,
+      "step": 527
+    },
+    {
+      "epoch": 1.4515463917525773,
+      "grad_norm": 0.7215636928866146,
+      "learning_rate": 7.146085405464968e-05,
+      "loss": 0.5746,
+      "step": 528
+    },
+    {
+      "epoch": 1.454295532646048,
+      "grad_norm": 0.6022527234482411,
+      "learning_rate": 7.141327273778284e-05,
+      "loss": 0.5808,
+      "step": 529
+    },
+    {
+      "epoch": 1.4570446735395188,
+      "grad_norm": 0.44820302089966574,
+      "learning_rate": 7.136557515820999e-05,
+      "loss": 0.5745,
+      "step": 530
+    },
+    {
+      "epoch": 1.4597938144329896,
+      "grad_norm": 0.42351586228242205,
+      "learning_rate": 7.13177614924632e-05,
+      "loss": 0.5858,
+      "step": 531
+    },
+    {
+      "epoch": 1.4625429553264606,
+      "grad_norm": 0.4466996033955547,
+      "learning_rate": 7.126983191750412e-05,
+      "loss": 0.5746,
+      "step": 532
+    },
+    {
+      "epoch": 1.4652920962199314,
+      "grad_norm": 0.40737004869624294,
+      "learning_rate": 7.12217866107235e-05,
+      "loss": 0.5818,
+      "step": 533
+    },
+    {
+      "epoch": 1.4680412371134022,
+      "grad_norm": 0.47191169515150533,
+      "learning_rate": 7.11736257499403e-05,
+      "loss": 0.5809,
+      "step": 534
+    },
+    {
+      "epoch": 1.470790378006873,
+      "grad_norm": 0.4813204578717719,
+      "learning_rate": 7.112534951340126e-05,
+      "loss": 0.5844,
+      "step": 535
+    },
+    {
+      "epoch": 1.4735395189003437,
+      "grad_norm": 0.5549078938525966,
+      "learning_rate": 7.107695807978007e-05,
+      "loss": 0.5812,
+      "step": 536
+    },
+    {
+      "epoch": 1.4762886597938145,
+      "grad_norm": 0.6761669063164707,
+      "learning_rate": 7.10284516281768e-05,
+      "loss": 0.5766,
+      "step": 537
+    },
+    {
+      "epoch": 1.4790378006872853,
+      "grad_norm": 0.6480701339305898,
+      "learning_rate": 7.097983033811718e-05,
+      "loss": 0.5805,
+      "step": 538
+    },
+    {
+      "epoch": 1.481786941580756,
+      "grad_norm": 0.5593193077179679,
+      "learning_rate": 7.0931094389552e-05,
+      "loss": 0.575,
+      "step": 539
+    },
+    {
+      "epoch": 1.4845360824742269,
+      "grad_norm": 0.4942110243815134,
+      "learning_rate": 7.088224396285638e-05,
+      "loss": 0.5833,
+      "step": 540
+    },
+    {
+      "epoch": 1.4872852233676976,
+      "grad_norm": 0.5797979347815979,
+      "learning_rate": 7.083327923882917e-05,
+      "loss": 0.5747,
+      "step": 541
+    },
+    {
+      "epoch": 1.4900343642611684,
+      "grad_norm": 0.9315007373980465,
+      "learning_rate": 7.078420039869218e-05,
+      "loss": 0.592,
+      "step": 542
+    },
+    {
+      "epoch": 1.4927835051546392,
+      "grad_norm": 1.616429951355777,
+      "learning_rate": 7.073500762408964e-05,
+      "loss": 0.5944,
+      "step": 543
+    },
+    {
+      "epoch": 1.49553264604811,
+      "grad_norm": 0.5055431608140836,
+      "learning_rate": 7.068570109708741e-05,
+      "loss": 0.5901,
+      "step": 544
+    },
+    {
+      "epoch": 1.4982817869415808,
+      "grad_norm": 1.0997422339745508,
+      "learning_rate": 7.06362810001724e-05,
+      "loss": 0.5902,
+      "step": 545
+    },
+    {
+      "epoch": 1.5010309278350515,
+      "grad_norm": 1.7657567775705625,
+      "learning_rate": 7.05867475162518e-05,
+      "loss": 0.5885,
+      "step": 546
+    },
+    {
+      "epoch": 1.5037800687285223,
+      "grad_norm": 0.7356302747726631,
+      "learning_rate": 7.053710082865247e-05,
+      "loss": 0.5861,
+      "step": 547
+    },
+    {
+      "epoch": 1.506529209621993,
+      "grad_norm": 2.1759109547514943,
+      "learning_rate": 7.048734112112026e-05,
+      "loss": 0.5997,
+      "step": 548
+    },
+    {
+      "epoch": 1.5092783505154639,
+      "grad_norm": 1.2978110027459113,
+      "learning_rate": 7.043746857781933e-05,
+      "loss": 0.6157,
+      "step": 549
+    },
+    {
+      "epoch": 1.5120274914089347,
+      "grad_norm": 2.219306449832288,
+      "learning_rate": 7.03874833833314e-05,
+      "loss": 0.607,
+      "step": 550
+    },
+    {
+      "epoch": 1.5147766323024054,
+      "grad_norm": 1.9591339401200079,
+      "learning_rate": 7.033738572265517e-05,
+      "loss": 0.6369,
+      "step": 551
+    },
+    {
+      "epoch": 1.5175257731958762,
+      "grad_norm": 1.478246394702369,
+      "learning_rate": 7.028717578120555e-05,
+      "loss": 0.6063,
+      "step": 552
+    },
+    {
+      "epoch": 1.520274914089347,
+      "grad_norm": 1.2579901505180804,
+      "learning_rate": 7.023685374481301e-05,
+      "loss": 0.6019,
+      "step": 553
+    },
+    {
+      "epoch": 1.5230240549828178,
+      "grad_norm": 1.46105914035749,
+      "learning_rate": 7.018641979972295e-05,
+      "loss": 0.6094,
+      "step": 554
+    },
+    {
+      "epoch": 1.5257731958762886,
+      "grad_norm": 0.8090152261985227,
+      "learning_rate": 7.013587413259486e-05,
+      "loss": 0.5913,
+      "step": 555
+    },
+    {
+      "epoch": 1.5285223367697593,
+      "grad_norm": 1.2609062811324916,
+      "learning_rate": 7.008521693050179e-05,
+      "loss": 0.5856,
+      "step": 556
+    },
+    {
+      "epoch": 1.5312714776632301,
+      "grad_norm": 1.0043747397077383,
+      "learning_rate": 7.003444838092957e-05,
+      "loss": 0.5989,
+      "step": 557
+    },
+    {
+      "epoch": 1.534020618556701,
+      "grad_norm": 0.7445889690645772,
+      "learning_rate": 6.998356867177613e-05,
+      "loss": 0.5865,
+      "step": 558
+    },
+    {
+      "epoch": 1.5367697594501717,
+      "grad_norm": 1.2566147555053582,
+      "learning_rate": 6.993257799135078e-05,
+      "loss": 0.5855,
+      "step": 559
+    },
+    {
+      "epoch": 1.5395189003436425,
+      "grad_norm": 0.623650378420435,
+      "learning_rate": 6.988147652837362e-05,
+      "loss": 0.5931,
+      "step": 560
+    },
+    {
+      "epoch": 1.5422680412371133,
+      "grad_norm": 1.0346995401552566,
+      "learning_rate": 6.983026447197469e-05,
+      "loss": 0.5914,
+      "step": 561
+    },
+    {
+      "epoch": 1.545017182130584,
+      "grad_norm": 0.9620047133783202,
+      "learning_rate": 6.977894201169336e-05,
+      "loss": 0.5915,
+      "step": 562
+    },
+    {
+      "epoch": 1.547766323024055,
+      "grad_norm": 0.6073721542597159,
+      "learning_rate": 6.972750933747765e-05,
+      "loss": 0.5834,
+      "step": 563
+    },
+    {
+      "epoch": 1.5505154639175258,
+      "grad_norm": 0.9240637148715215,
+      "learning_rate": 6.967596663968346e-05,
+      "loss": 0.5891,
+      "step": 564
+    },
+    {
+      "epoch": 1.5532646048109966,
+      "grad_norm": 0.7729700501754245,
+      "learning_rate": 6.962431410907387e-05,
+      "loss": 0.5864,
+      "step": 565
+    },
+    {
+      "epoch": 1.5560137457044674,
+      "grad_norm": 0.7023186108460778,
+      "learning_rate": 6.957255193681852e-05,
+      "loss": 0.5792,
+      "step": 566
+    },
+    {
+      "epoch": 1.5587628865979382,
+      "grad_norm": 0.8156415963129797,
+      "learning_rate": 6.952068031449278e-05,
+      "loss": 0.5804,
+      "step": 567
+    },
+    {
+      "epoch": 1.561512027491409,
+      "grad_norm": 0.4418458963414597,
+      "learning_rate": 6.946869943407714e-05,
+      "loss": 0.5796,
+      "step": 568
+    },
+    {
+      "epoch": 1.5642611683848797,
+      "grad_norm": 0.7517403935755494,
+      "learning_rate": 6.941660948795646e-05,
+      "loss": 0.5757,
+      "step": 569
+    },
+    {
+      "epoch": 1.5670103092783505,
+      "grad_norm": 0.6880319452283528,
+      "learning_rate": 6.936441066891924e-05,
+      "loss": 0.5866,
+      "step": 570
+    },
+    {
+      "epoch": 1.5697594501718213,
+      "grad_norm": 0.44431908466865444,
+      "learning_rate": 6.931210317015694e-05,
+      "loss": 0.5831,
+      "step": 571
+    },
+    {
+      "epoch": 1.572508591065292,
+      "grad_norm": 0.5206213485516508,
+      "learning_rate": 6.925968718526325e-05,
+      "loss": 0.5835,
+      "step": 572
+    },
+    {
+      "epoch": 1.5752577319587628,
+      "grad_norm": 0.4865149316480254,
+      "learning_rate": 6.920716290823337e-05,
+      "loss": 0.5872,
+      "step": 573
+    },
+    {
+      "epoch": 1.5780068728522336,
+      "grad_norm": 0.45702506236242896,
+      "learning_rate": 6.915453053346332e-05,
+      "loss": 0.5918,
+      "step": 574
+    },
+    {
+      "epoch": 1.5807560137457046,
+      "grad_norm": 0.44494079428514854,
+      "learning_rate": 6.910179025574916e-05,
+      "loss": 0.5889,
+      "step": 575
+    },
+    {
+      "epoch": 1.5835051546391754,
+      "grad_norm": 0.40161179223147797,
+      "learning_rate": 6.904894227028631e-05,
+      "loss": 0.5793,
+      "step": 576
+    },
+    {
+      "epoch": 1.5862542955326462,
+      "grad_norm": 0.40020351103635327,
+      "learning_rate": 6.899598677266888e-05,
+      "loss": 0.5791,
+      "step": 577
+    },
+    {
+      "epoch": 1.589003436426117,
+      "grad_norm": 0.5533864087207034,
+      "learning_rate": 6.89429239588888e-05,
+      "loss": 0.5686,
+      "step": 578
+    },
+    {
+      "epoch": 1.5917525773195877,
+      "grad_norm": 0.4609556358868059,
+      "learning_rate": 6.888975402533528e-05,
+      "loss": 0.5696,
+      "step": 579
+    },
+    {
+      "epoch": 1.5945017182130585,
+      "grad_norm": 0.376485016123922,
+      "learning_rate": 6.883647716879389e-05,
+      "loss": 0.5787,
+      "step": 580
+    },
+    {
+      "epoch": 1.5972508591065293,
+      "grad_norm": 0.3703464382281221,
+      "learning_rate": 6.878309358644602e-05,
+      "loss": 0.5786,
+      "step": 581
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.36242738510972944,
+      "learning_rate": 6.8729603475868e-05,
+      "loss": 0.5838,
+      "step": 582
+    },
+    {
+      "epoch": 1.6027491408934709,
+      "grad_norm": 0.4005050809946878,
+      "learning_rate": 6.867600703503044e-05,
+      "loss": 0.5828,
+      "step": 583
+    },
+    {
+      "epoch": 1.6054982817869417,
+      "grad_norm": 0.42329454686286444,
+      "learning_rate": 6.86223044622975e-05,
+      "loss": 0.5803,
+      "step": 584
+    },
+    {
+      "epoch": 1.6082474226804124,
+      "grad_norm": 0.3329212652227201,
+      "learning_rate": 6.856849595642617e-05,
+      "loss": 0.5822,
+      "step": 585
+    },
+    {
+      "epoch": 1.6109965635738832,
+      "grad_norm": 0.46886870029058014,
+      "learning_rate": 6.851458171656541e-05,
+      "loss": 0.58,
+      "step": 586
+    },
+    {
+      "epoch": 1.613745704467354,
+      "grad_norm": 0.5556659975442035,
+      "learning_rate": 6.846056194225562e-05,
+      "loss": 0.5702,
+      "step": 587
+    },
+    {
+      "epoch": 1.6164948453608248,
+      "grad_norm": 0.5260721900968394,
+      "learning_rate": 6.84064368334277e-05,
+      "loss": 0.5867,
+      "step": 588
+    },
+    {
+      "epoch": 1.6192439862542956,
+      "grad_norm": 0.40786759602060935,
+      "learning_rate": 6.835220659040246e-05,
+      "loss": 0.5834,
+      "step": 589
+    },
+    {
+      "epoch": 1.6219931271477663,
+      "grad_norm": 0.2996276604956069,
+      "learning_rate": 6.82978714138898e-05,
+      "loss": 0.5776,
+      "step": 590
+    },
+    {
+      "epoch": 1.6247422680412371,
+      "grad_norm": 0.35270615435100505,
+      "learning_rate": 6.824343150498798e-05,
+      "loss": 0.5783,
+      "step": 591
+    },
+    {
+      "epoch": 1.627491408934708,
+      "grad_norm": 0.28376863619687237,
+      "learning_rate": 6.81888870651829e-05,
+      "loss": 0.5767,
+      "step": 592
+    },
+    {
+      "epoch": 1.6302405498281787,
+      "grad_norm": 0.36443318149486764,
+      "learning_rate": 6.813423829634732e-05,
+      "loss": 0.579,
+      "step": 593
+    },
+    {
+      "epoch": 1.6329896907216495,
+      "grad_norm": 0.481183239710988,
+      "learning_rate": 6.80794854007401e-05,
+      "loss": 0.5815,
+      "step": 594
+    },
+    {
+      "epoch": 1.6357388316151202,
+      "grad_norm": 0.4948913070938149,
+      "learning_rate": 6.802462858100556e-05,
+      "loss": 0.5847,
+      "step": 595
+    },
+    {
+      "epoch": 1.638487972508591,
+      "grad_norm": 0.3815071178544923,
+      "learning_rate": 6.796966804017258e-05,
+      "loss": 0.5724,
+      "step": 596
+    },
+    {
+      "epoch": 1.6412371134020618,
+      "grad_norm": 0.4407973877140981,
+      "learning_rate": 6.791460398165393e-05,
+      "loss": 0.5811,
+      "step": 597
+    },
+    {
+      "epoch": 1.6439862542955326,
+      "grad_norm": 0.43391575615156724,
+      "learning_rate": 6.785943660924553e-05,
+      "loss": 0.5827,
+      "step": 598
+    },
+    {
+      "epoch": 1.6467353951890034,
+      "grad_norm": 0.4035737413547384,
+      "learning_rate": 6.780416612712568e-05,
+      "loss": 0.5819,
+      "step": 599
+    },
+    {
+      "epoch": 1.6494845360824741,
+      "grad_norm": 0.5281104884380362,
+      "learning_rate": 6.774879273985423e-05,
+      "loss": 0.5775,
+      "step": 600
+    },
+    {
+      "epoch": 1.652233676975945,
+      "grad_norm": 0.7760361977252223,
+      "learning_rate": 6.769331665237198e-05,
+      "loss": 0.5899,
+      "step": 601
+    },
+    {
+      "epoch": 1.6549828178694157,
+      "grad_norm": 1.09403271674242,
+      "learning_rate": 6.763773806999977e-05,
+      "loss": 0.5859,
+      "step": 602
+    },
+    {
+      "epoch": 1.6577319587628865,
+      "grad_norm": 0.9348914419022682,
+      "learning_rate": 6.75820571984378e-05,
+      "loss": 0.5918,
+      "step": 603
+    },
+    {
+      "epoch": 1.6604810996563573,
+      "grad_norm": 0.7435836588260047,
+      "learning_rate": 6.752627424376484e-05,
+      "loss": 0.582,
+      "step": 604
+    },
+    {
+      "epoch": 1.663230240549828,
+      "grad_norm": 0.5416202424064285,
+      "learning_rate": 6.747038941243748e-05,
+      "loss": 0.5845,
+      "step": 605
+    },
+    {
+      "epoch": 1.6659793814432988,
+      "grad_norm": 0.4289104135369005,
+      "learning_rate": 6.741440291128938e-05,
+      "loss": 0.5841,
+      "step": 606
+    },
+    {
+      "epoch": 1.6687285223367696,
+      "grad_norm": 0.3842777168497921,
+      "learning_rate": 6.735831494753046e-05,
+      "loss": 0.5783,
+      "step": 607
+    },
+    {
+      "epoch": 1.6714776632302404,
+      "grad_norm": 0.3676465027204242,
+      "learning_rate": 6.730212572874618e-05,
+      "loss": 0.5708,
+      "step": 608
+    },
+    {
+      "epoch": 1.6742268041237114,
+      "grad_norm": 0.3960487648888393,
+      "learning_rate": 6.724583546289672e-05,
+      "loss": 0.5724,
+      "step": 609
+    },
+    {
+      "epoch": 1.6769759450171822,
+      "grad_norm": 0.4083333422479102,
+      "learning_rate": 6.71894443583163e-05,
+      "loss": 0.5925,
+      "step": 610
+    },
+    {
+      "epoch": 1.679725085910653,
+      "grad_norm": 0.4578552218287685,
+      "learning_rate": 6.713295262371232e-05,
+      "loss": 0.5715,
+      "step": 611
+    },
+    {
+      "epoch": 1.6824742268041237,
+      "grad_norm": 0.5656117537132986,
+      "learning_rate": 6.707636046816457e-05,
+      "loss": 0.5766,
+      "step": 612
+    },
+    {
+      "epoch": 1.6852233676975945,
+      "grad_norm": 0.6850771753028508,
+      "learning_rate": 6.70196681011246e-05,
+      "loss": 0.5928,
+      "step": 613
+    },
+    {
+      "epoch": 1.6879725085910653,
+      "grad_norm": 0.698725362743093,
+      "learning_rate": 6.696287573241478e-05,
+      "loss": 0.5897,
+      "step": 614
+    },
+    {
+      "epoch": 1.690721649484536,
+      "grad_norm": 0.7382677867943621,
+      "learning_rate": 6.690598357222764e-05,
+      "loss": 0.5668,
+      "step": 615
+    },
+    {
+      "epoch": 1.6934707903780069,
+      "grad_norm": 0.7105270810835767,
+      "learning_rate": 6.684899183112498e-05,
+      "loss": 0.5736,
+      "step": 616
+    },
+    {
+      "epoch": 1.6962199312714776,
+      "grad_norm": 0.6862323082435818,
+      "learning_rate": 6.679190072003722e-05,
+      "loss": 0.5765,
+      "step": 617
+    },
+    {
+      "epoch": 1.6989690721649484,
+      "grad_norm": 0.6299104085981699,
+      "learning_rate": 6.67347104502625e-05,
+      "loss": 0.5642,
+      "step": 618
+    },
+    {
+      "epoch": 1.7017182130584192,
+      "grad_norm": 0.6026002947844302,
+      "learning_rate": 6.667742123346603e-05,
+      "loss": 0.5804,
+      "step": 619
+    },
+    {
+      "epoch": 1.7044673539518902,
+      "grad_norm": 0.4462813007489562,
+      "learning_rate": 6.662003328167917e-05,
+      "loss": 0.5807,
+      "step": 620
+    },
+    {
+      "epoch": 1.707216494845361,
+      "grad_norm": 0.29472463703813456,
+      "learning_rate": 6.65625468072987e-05,
+      "loss": 0.5702,
+      "step": 621
+    },
+    {
+      "epoch": 1.7099656357388318,
+      "grad_norm": 0.320893033856143,
+      "learning_rate": 6.650496202308607e-05,
+      "loss": 0.5835,
+      "step": 622
+    },
+    {
+      "epoch": 1.7127147766323025,
+      "grad_norm": 0.39010773443389357,
+      "learning_rate": 6.644727914216656e-05,
+      "loss": 0.576,
+      "step": 623
+    },
+    {
+      "epoch": 1.7154639175257733,
+      "grad_norm": 0.5319160826521161,
+      "learning_rate": 6.638949837802855e-05,
+      "loss": 0.5798,
+      "step": 624
+    },
+    {
+      "epoch": 1.718213058419244,
+      "grad_norm": 0.6498383386047935,
+      "learning_rate": 6.633161994452262e-05,
+      "loss": 0.5795,
+      "step": 625
+    },
+    {
+      "epoch": 1.7209621993127149,
+      "grad_norm": 0.7259267311718851,
+      "learning_rate": 6.627364405586091e-05,
+      "loss": 0.5919,
+      "step": 626
+    },
+    {
+      "epoch": 1.7237113402061857,
+      "grad_norm": 0.8162977870752441,
+      "learning_rate": 6.621557092661621e-05,
+      "loss": 0.5753,
+      "step": 627
+    },
+    {
+      "epoch": 1.7264604810996564,
+      "grad_norm": 0.7728017259091454,
+      "learning_rate": 6.61574007717212e-05,
+      "loss": 0.5843,
+      "step": 628
+    },
+    {
+      "epoch": 1.7292096219931272,
+      "grad_norm": 0.7443225127869316,
+      "learning_rate": 6.609913380646767e-05,
+      "loss": 0.5802,
+      "step": 629
+    },
+    {
+      "epoch": 1.731958762886598,
+      "grad_norm": 0.7541481509168282,
+      "learning_rate": 6.604077024650571e-05,
+      "loss": 0.5841,
+      "step": 630
+    },
+    {
+      "epoch": 1.7347079037800688,
+      "grad_norm": 0.6874877805251565,
+      "learning_rate": 6.598231030784289e-05,
+      "loss": 0.5861,
+      "step": 631
+    },
+    {
+      "epoch": 1.7374570446735396,
+      "grad_norm": 0.6302784745574853,
+      "learning_rate": 6.592375420684354e-05,
+      "loss": 0.579,
+      "step": 632
+    },
+    {
+      "epoch": 1.7402061855670103,
+      "grad_norm": 0.3605005517279551,
+      "learning_rate": 6.586510216022785e-05,
+      "loss": 0.5761,
+      "step": 633
+    },
+    {
+      "epoch": 1.7429553264604811,
+      "grad_norm": 0.40851626295602705,
+      "learning_rate": 6.58063543850711e-05,
+      "loss": 0.5809,
+      "step": 634
+    },
+    {
+      "epoch": 1.745704467353952,
+      "grad_norm": 0.5889121925134353,
+      "learning_rate": 6.574751109880295e-05,
+      "loss": 0.589,
+      "step": 635
+    },
+    {
+      "epoch": 1.7484536082474227,
+      "grad_norm": 0.6855740977784578,
+      "learning_rate": 6.568857251920641e-05,
+      "loss": 0.5791,
+      "step": 636
+    },
+    {
+      "epoch": 1.7512027491408935,
+      "grad_norm": 0.7763253623405283,
+      "learning_rate": 6.562953886441731e-05,
+      "loss": 0.5698,
+      "step": 637
+    },
+    {
+      "epoch": 1.7539518900343642,
+      "grad_norm": 0.8027574724353924,
+      "learning_rate": 6.557041035292331e-05,
+      "loss": 0.5695,
+      "step": 638
+    },
+    {
+      "epoch": 1.756701030927835,
+      "grad_norm": 0.8124682042384597,
+      "learning_rate": 6.551118720356313e-05,
+      "loss": 0.5793,
+      "step": 639
+    },
+    {
+      "epoch": 1.7594501718213058,
+      "grad_norm": 0.7285676570172558,
+      "learning_rate": 6.545186963552578e-05,
+      "loss": 0.5809,
+      "step": 640
+    },
+    {
+      "epoch": 1.7621993127147766,
+      "grad_norm": 0.5493034420656492,
+      "learning_rate": 6.539245786834965e-05,
+      "loss": 0.5779,
+      "step": 641
+    },
+    {
+      "epoch": 1.7649484536082474,
+      "grad_norm": 0.4100237013175215,
+      "learning_rate": 6.533295212192189e-05,
+      "loss": 0.5634,
+      "step": 642
+    },
+    {
+      "epoch": 1.7676975945017182,
+      "grad_norm": 0.44386016701061204,
+      "learning_rate": 6.527335261647735e-05,
+      "loss": 0.5865,
+      "step": 643
+    },
+    {
+      "epoch": 1.770446735395189,
+      "grad_norm": 0.7605693123034108,
+      "learning_rate": 6.521365957259798e-05,
+      "loss": 0.5758,
+      "step": 644
+    },
+    {
+      "epoch": 1.7731958762886597,
+      "grad_norm": 0.9531489941794372,
+      "learning_rate": 6.515387321121185e-05,
+      "loss": 0.5845,
+      "step": 645
+    },
+    {
+      "epoch": 1.7759450171821305,
+      "grad_norm": 0.9748291614843003,
+      "learning_rate": 6.509399375359247e-05,
+      "loss": 0.5719,
+      "step": 646
+    },
+    {
+      "epoch": 1.7786941580756013,
+      "grad_norm": 0.7628578645914652,
+      "learning_rate": 6.503402142135787e-05,
+      "loss": 0.5768,
+      "step": 647
+    },
+    {
+      "epoch": 1.781443298969072,
+      "grad_norm": 0.5799935111185708,
+      "learning_rate": 6.497395643646981e-05,
+      "loss": 0.5701,
+      "step": 648
+    },
+    {
+      "epoch": 1.7841924398625428,
+      "grad_norm": 0.5277145245316653,
+      "learning_rate": 6.491379902123297e-05,
+      "loss": 0.5683,
+      "step": 649
+    },
+    {
+      "epoch": 1.7869415807560136,
+      "grad_norm": 0.4736528081812741,
+      "learning_rate": 6.485354939829414e-05,
+      "loss": 0.5779,
+      "step": 650
+    },
+    {
+      "epoch": 1.7896907216494844,
+      "grad_norm": 0.5669836094412534,
+      "learning_rate": 6.479320779064134e-05,
+      "loss": 0.5764,
+      "step": 651
+    },
+    {
+      "epoch": 1.7924398625429552,
+      "grad_norm": 0.559780538593531,
+      "learning_rate": 6.473277442160309e-05,
+      "loss": 0.5832,
+      "step": 652
+    },
+    {
+      "epoch": 1.795189003436426,
+      "grad_norm": 0.5065885370576576,
+      "learning_rate": 6.467224951484746e-05,
+      "loss": 0.5783,
+      "step": 653
+    },
+    {
+      "epoch": 1.797938144329897,
+      "grad_norm": 0.4559459768927469,
+      "learning_rate": 6.461163329438135e-05,
+      "loss": 0.5715,
+      "step": 654
+    },
+    {
+      "epoch": 1.8006872852233677,
+      "grad_norm": 0.42548224770937315,
+      "learning_rate": 6.45509259845496e-05,
+      "loss": 0.5751,
+      "step": 655
+    },
+    {
+      "epoch": 1.8034364261168385,
+      "grad_norm": 0.3455007524085001,
+      "learning_rate": 6.449012781003419e-05,
+      "loss": 0.5754,
+      "step": 656
+    },
+    {
+      "epoch": 1.8061855670103093,
+      "grad_norm": 0.4400225976066889,
+      "learning_rate": 6.442923899585338e-05,
+      "loss": 0.5941,
+      "step": 657
+    },
+    {
+      "epoch": 1.80893470790378,
+      "grad_norm": 0.5084943205879741,
+      "learning_rate": 6.436825976736092e-05,
+      "loss": 0.579,
+      "step": 658
+    },
+    {
+      "epoch": 1.8116838487972509,
+      "grad_norm": 0.5358918155905726,
+      "learning_rate": 6.430719035024515e-05,
+      "loss": 0.5748,
+      "step": 659
+    },
+    {
+      "epoch": 1.8144329896907216,
+      "grad_norm": 0.503047756943303,
+      "learning_rate": 6.424603097052826e-05,
+      "loss": 0.5794,
+      "step": 660
+    },
+    {
+      "epoch": 1.8171821305841924,
+      "grad_norm": 0.44321729048134617,
+      "learning_rate": 6.418478185456534e-05,
+      "loss": 0.5865,
+      "step": 661
+    },
+    {
+      "epoch": 1.8199312714776632,
+      "grad_norm": 0.3737976145350619,
+      "learning_rate": 6.412344322904362e-05,
+      "loss": 0.5762,
+      "step": 662
+    },
+    {
+      "epoch": 1.822680412371134,
+      "grad_norm": 0.2731612660975183,
+      "learning_rate": 6.406201532098165e-05,
+      "loss": 0.5691,
+      "step": 663
+    },
+    {
+      "epoch": 1.8254295532646048,
+      "grad_norm": 0.3311008105208644,
+      "learning_rate": 6.400049835772836e-05,
+      "loss": 0.5729,
+      "step": 664
+    },
+    {
+      "epoch": 1.8281786941580758,
+      "grad_norm": 0.3627533269491876,
+      "learning_rate": 6.393889256696231e-05,
+      "loss": 0.5729,
+      "step": 665
+    },
+    {
+      "epoch": 1.8309278350515465,
+      "grad_norm": 0.3316310187075983,
+      "learning_rate": 6.387719817669081e-05,
+      "loss": 0.582,
+      "step": 666
+    },
+    {
+      "epoch": 1.8336769759450173,
+      "grad_norm": 0.3431221379909935,
+      "learning_rate": 6.381541541524911e-05,
+      "loss": 0.5775,
+      "step": 667
+    },
+    {
+      "epoch": 1.836426116838488,
+      "grad_norm": 0.308724522020036,
+      "learning_rate": 6.375354451129949e-05,
+      "loss": 0.5759,
+      "step": 668
+    },
+    {
+      "epoch": 1.839175257731959,
+      "grad_norm": 0.2799060976007718,
+      "learning_rate": 6.369158569383045e-05,
+      "loss": 0.5826,
+      "step": 669
+    },
+    {
+      "epoch": 1.8419243986254297,
+      "grad_norm": 0.3820386705421886,
+      "learning_rate": 6.362953919215591e-05,
+      "loss": 0.5778,
+      "step": 670
+    },
+    {
+      "epoch": 1.8446735395189005,
+      "grad_norm": 0.38246444479750036,
+      "learning_rate": 6.356740523591427e-05,
+      "loss": 0.5749,
+      "step": 671
+    },
+    {
+      "epoch": 1.8474226804123712,
+      "grad_norm": 0.3271098200229515,
+      "learning_rate": 6.350518405506764e-05,
+      "loss": 0.5817,
+      "step": 672
+    },
+    {
+      "epoch": 1.850171821305842,
+      "grad_norm": 0.34228419827242806,
+      "learning_rate": 6.344287587990091e-05,
+      "loss": 0.5819,
+      "step": 673
+    },
+    {
+      "epoch": 1.8529209621993128,
+      "grad_norm": 0.32655548951800334,
+      "learning_rate": 6.338048094102096e-05,
+      "loss": 0.5749,
+      "step": 674
+    },
+    {
+      "epoch": 1.8556701030927836,
+      "grad_norm": 0.32871268032009765,
+      "learning_rate": 6.331799946935583e-05,
+      "loss": 0.5685,
+      "step": 675
+    },
+    {
+      "epoch": 1.8584192439862544,
+      "grad_norm": 0.4199684721453658,
+      "learning_rate": 6.325543169615378e-05,
+      "loss": 0.5725,
+      "step": 676
+    },
+    {
+      "epoch": 1.8611683848797251,
+      "grad_norm": 0.5756048364369786,
+      "learning_rate": 6.319277785298247e-05,
+      "loss": 0.5805,
+      "step": 677
+    },
+    {
+      "epoch": 1.863917525773196,
+      "grad_norm": 0.6547196962591697,
+      "learning_rate": 6.313003817172812e-05,
+      "loss": 0.5662,
+      "step": 678
+    },
+    {
+      "epoch": 1.8666666666666667,
+      "grad_norm": 0.7656825534443144,
+      "learning_rate": 6.30672128845947e-05,
+      "loss": 0.5864,
+      "step": 679
+    },
+    {
+      "epoch": 1.8694158075601375,
+      "grad_norm": 0.9358535366007116,
+      "learning_rate": 6.300430222410292e-05,
+      "loss": 0.5798,
+      "step": 680
+    },
+    {
+      "epoch": 1.8721649484536083,
+      "grad_norm": 0.9703236167568001,
+      "learning_rate": 6.294130642308952e-05,
+      "loss": 0.5752,
+      "step": 681
+    },
+    {
+      "epoch": 1.874914089347079,
+      "grad_norm": 0.8892433252848201,
+      "learning_rate": 6.287822571470636e-05,
+      "loss": 0.5709,
+      "step": 682
+    },
+    {
+      "epoch": 1.8776632302405498,
+      "grad_norm": 0.8810114036796293,
+      "learning_rate": 6.281506033241948e-05,
+      "loss": 0.5733,
+      "step": 683
+    },
+    {
+      "epoch": 1.8804123711340206,
+      "grad_norm": 0.853242199524319,
+      "learning_rate": 6.275181051000842e-05,
+      "loss": 0.5758,
+      "step": 684
+    },
+    {
+      "epoch": 1.8831615120274914,
+      "grad_norm": 0.6715868178268443,
+      "learning_rate": 6.26884764815651e-05,
+      "loss": 0.5681,
+      "step": 685
+    },
+    {
+      "epoch": 1.8859106529209622,
+      "grad_norm": 0.41927983624654164,
+      "learning_rate": 6.26250584814932e-05,
+      "loss": 0.5706,
+      "step": 686
+    },
+    {
+      "epoch": 1.888659793814433,
+      "grad_norm": 0.35123714928190247,
+      "learning_rate": 6.256155674450712e-05,
+      "loss": 0.5742,
+      "step": 687
+    },
+    {
+      "epoch": 1.8914089347079037,
+      "grad_norm": 0.5653996664319825,
+      "learning_rate": 6.249797150563124e-05,
+      "loss": 0.5731,
+      "step": 688
+    },
+    {
+      "epoch": 1.8941580756013745,
+      "grad_norm": 0.6649259671430193,
+      "learning_rate": 6.243430300019891e-05,
+      "loss": 0.5797,
+      "step": 689
+    },
+    {
+      "epoch": 1.8969072164948453,
+      "grad_norm": 0.60181480524739,
+      "learning_rate": 6.237055146385173e-05,
+      "loss": 0.5758,
+      "step": 690
+    },
+    {
+      "epoch": 1.899656357388316,
+      "grad_norm": 0.6235277209689051,
+      "learning_rate": 6.230671713253854e-05,
+      "loss": 0.5778,
+      "step": 691
+    },
+    {
+      "epoch": 1.9024054982817868,
+      "grad_norm": 0.6196187849528243,
+      "learning_rate": 6.224280024251466e-05,
+      "loss": 0.5698,
+      "step": 692
+    },
+    {
+      "epoch": 1.9051546391752576,
+      "grad_norm": 0.42709196066527666,
+      "learning_rate": 6.217880103034091e-05,
+      "loss": 0.5736,
+      "step": 693
+    },
+    {
+      "epoch": 1.9079037800687284,
+      "grad_norm": 0.34692776304267553,
+      "learning_rate": 6.211471973288285e-05,
+      "loss": 0.5712,
+      "step": 694
+    },
+    {
+      "epoch": 1.9106529209621992,
+      "grad_norm": 0.3110542487628964,
+      "learning_rate": 6.205055658730983e-05,
+      "loss": 0.5653,
+      "step": 695
+    },
+    {
+      "epoch": 1.91340206185567,
+      "grad_norm": 0.34540933819310177,
+      "learning_rate": 6.198631183109408e-05,
+      "loss": 0.576,
+      "step": 696
+    },
+    {
+      "epoch": 1.9161512027491407,
+      "grad_norm": 0.3126265820113025,
+      "learning_rate": 6.192198570200992e-05,
+      "loss": 0.5906,
+      "step": 697
+    },
+    {
+      "epoch": 1.9189003436426115,
+      "grad_norm": 0.29077210119155705,
+      "learning_rate": 6.185757843813282e-05,
+      "loss": 0.5799,
+      "step": 698
+    },
+    {
+      "epoch": 1.9216494845360823,
+      "grad_norm": 0.43001232514386034,
+      "learning_rate": 6.179309027783856e-05,
+      "loss": 0.5784,
+      "step": 699
+    },
+    {
+      "epoch": 1.9243986254295533,
+      "grad_norm": 0.4962679375465972,
+      "learning_rate": 6.172852145980228e-05,
+      "loss": 0.5731,
+      "step": 700
+    },
+    {
+      "epoch": 1.927147766323024,
+      "grad_norm": 0.5311714605519653,
+      "learning_rate": 6.166387222299767e-05,
+      "loss": 0.5812,
+      "step": 701
+    },
+    {
+      "epoch": 1.9298969072164949,
+      "grad_norm": 0.586368926903193,
+      "learning_rate": 6.159914280669607e-05,
+      "loss": 0.5638,
+      "step": 702
+    },
+    {
+      "epoch": 1.9326460481099657,
+      "grad_norm": 0.6549081294272183,
+      "learning_rate": 6.153433345046552e-05,
+      "loss": 0.5785,
+      "step": 703
+    },
+    {
+      "epoch": 1.9353951890034364,
+      "grad_norm": 0.7252681437363886,
+      "learning_rate": 6.146944439416997e-05,
+      "loss": 0.5751,
+      "step": 704
+    },
+    {
+      "epoch": 1.9381443298969072,
+      "grad_norm": 0.7845460102188683,
+      "learning_rate": 6.140447587796832e-05,
+      "loss": 0.5646,
+      "step": 705
+    },
+    {
+      "epoch": 1.940893470790378,
+      "grad_norm": 0.8048678290944055,
+      "learning_rate": 6.133942814231357e-05,
+      "loss": 0.5742,
+      "step": 706
+    },
+    {
+      "epoch": 1.9436426116838488,
+      "grad_norm": 0.7734169093084938,
+      "learning_rate": 6.127430142795192e-05,
+      "loss": 0.5876,
+      "step": 707
+    },
+    {
+      "epoch": 1.9463917525773196,
+      "grad_norm": 0.5484007054312127,
+      "learning_rate": 6.120909597592185e-05,
+      "loss": 0.5843,
+      "step": 708
+    },
+    {
+      "epoch": 1.9491408934707903,
+      "grad_norm": 0.3242608423209931,
+      "learning_rate": 6.114381202755328e-05,
+      "loss": 0.5654,
+      "step": 709
+    },
+    {
+      "epoch": 1.9518900343642611,
+      "grad_norm": 0.3539205469330434,
+      "learning_rate": 6.107844982446662e-05,
+      "loss": 0.5833,
+      "step": 710
+    },
+    {
+      "epoch": 1.9546391752577321,
+      "grad_norm": 0.5156015064072683,
+      "learning_rate": 6.1013009608571954e-05,
+      "loss": 0.583,
+      "step": 711
+    },
+    {
+      "epoch": 1.957388316151203,
+      "grad_norm": 0.6131503596319059,
+      "learning_rate": 6.094749162206807e-05,
+      "loss": 0.5742,
+      "step": 712
+    },
+    {
+      "epoch": 1.9601374570446737,
+      "grad_norm": 0.6333732757531064,
+      "learning_rate": 6.0881896107441575e-05,
+      "loss": 0.5751,
+      "step": 713
+    },
+    {
+      "epoch": 1.9628865979381445,
+      "grad_norm": 0.456390717989884,
+      "learning_rate": 6.081622330746603e-05,
+      "loss": 0.5844,
+      "step": 714
+    },
+    {
+      "epoch": 1.9656357388316152,
+      "grad_norm": 0.41591397010146025,
+      "learning_rate": 6.075047346520105e-05,
+      "loss": 0.5756,
+      "step": 715
+    },
+    {
+      "epoch": 1.968384879725086,
+      "grad_norm": 0.5054704227601636,
+      "learning_rate": 6.0684646823991345e-05,
+      "loss": 0.5688,
+      "step": 716
+    },
+    {
+      "epoch": 1.9711340206185568,
+      "grad_norm": 0.6526494286710874,
+      "learning_rate": 6.061874362746592e-05,
+      "loss": 0.568,
+      "step": 717
+    },
+    {
+      "epoch": 1.9738831615120276,
+      "grad_norm": 0.7155431076783062,
+      "learning_rate": 6.055276411953705e-05,
+      "loss": 0.5746,
+      "step": 718
+    },
+    {
+      "epoch": 1.9766323024054984,
+      "grad_norm": 0.5467249394073967,
+      "learning_rate": 6.048670854439952e-05,
+      "loss": 0.5793,
+      "step": 719
+    },
+    {
+      "epoch": 1.9793814432989691,
+      "grad_norm": 0.44545561246216775,
+      "learning_rate": 6.042057714652958e-05,
+      "loss": 0.5793,
+      "step": 720
+    },
+    {
+      "epoch": 1.98213058419244,
+      "grad_norm": 0.43668565117704844,
+      "learning_rate": 6.035437017068413e-05,
+      "loss": 0.5711,
+      "step": 721
+    },
+    {
+      "epoch": 1.9848797250859107,
+      "grad_norm": 0.4794911786968242,
+      "learning_rate": 6.02880878618998e-05,
+      "loss": 0.5764,
+      "step": 722
+    },
+    {
+      "epoch": 1.9876288659793815,
+      "grad_norm": 0.45936527802619287,
+      "learning_rate": 6.022173046549201e-05,
+      "loss": 0.5728,
+      "step": 723
+    },
+    {
+      "epoch": 1.9903780068728523,
+      "grad_norm": 0.3710590792543285,
+      "learning_rate": 6.0155298227054125e-05,
+      "loss": 0.5764,
+      "step": 724
+    },
+    {
+      "epoch": 1.993127147766323,
+      "grad_norm": 0.3684226778884953,
+      "learning_rate": 6.008879139245645e-05,
+      "loss": 0.5711,
+      "step": 725
+    },
+    {
+      "epoch": 1.9958762886597938,
+      "grad_norm": 0.37506727135987533,
+      "learning_rate": 6.00222102078454e-05,
+      "loss": 0.5655,
+      "step": 726
+    },
+    {
+      "epoch": 1.9986254295532646,
+      "grad_norm": 0.42203860157012285,
+      "learning_rate": 5.995555491964259e-05,
+      "loss": 0.6314,
+      "step": 727
+    },
+    {
+      "epoch": 2.0013745704467354,
+      "grad_norm": 0.6121901786626351,
+      "learning_rate": 5.988882577454386e-05,
+      "loss": 0.6437,
+      "step": 728
+    },
+    {
+      "epoch": 2.004123711340206,
+      "grad_norm": 0.838029390831891,
+      "learning_rate": 5.982202301951841e-05,
+      "loss": 0.5246,
+      "step": 729
+    },
+    {
+      "epoch": 2.006872852233677,
+      "grad_norm": 1.0035111815398425,
+      "learning_rate": 5.97551469018079e-05,
+      "loss": 0.532,
+      "step": 730
+    },
+    {
+      "epoch": 2.0096219931271477,
+      "grad_norm": 1.0550172252909944,
+      "learning_rate": 5.968819766892546e-05,
+      "loss": 0.5356,
+      "step": 731
+    },
+    {
+      "epoch": 2.0123711340206185,
+      "grad_norm": 0.7212990022902317,
+      "learning_rate": 5.962117556865489e-05,
+      "loss": 0.5149,
+      "step": 732
+    },
+    {
+      "epoch": 2.0151202749140893,
+      "grad_norm": 0.5084983736290086,
+      "learning_rate": 5.955408084904962e-05,
+      "loss": 0.5248,
+      "step": 733
+    },
+    {
+      "epoch": 2.01786941580756,
+      "grad_norm": 0.508857242926003,
+      "learning_rate": 5.948691375843187e-05,
+      "loss": 0.5309,
+      "step": 734
+    },
+    {
+      "epoch": 2.020618556701031,
+      "grad_norm": 0.5715174569632097,
+      "learning_rate": 5.941967454539171e-05,
+      "loss": 0.5247,
+      "step": 735
+    },
+    {
+      "epoch": 2.0233676975945016,
+      "grad_norm": 0.5565322020849539,
+      "learning_rate": 5.935236345878616e-05,
+      "loss": 0.5175,
+      "step": 736
+    },
+    {
+      "epoch": 2.0261168384879724,
+      "grad_norm": 0.5745604042771598,
+      "learning_rate": 5.928498074773823e-05,
+      "loss": 0.517,
+      "step": 737
+    },
+    {
+      "epoch": 2.028865979381443,
+      "grad_norm": 0.500350720005458,
+      "learning_rate": 5.921752666163598e-05,
+      "loss": 0.5215,
+      "step": 738
+    },
+    {
+      "epoch": 2.031615120274914,
+      "grad_norm": 0.4067063681679997,
+      "learning_rate": 5.9150001450131716e-05,
+      "loss": 0.5212,
+      "step": 739
+    },
+    {
+      "epoch": 2.0343642611683848,
+      "grad_norm": 0.4770810449857095,
+      "learning_rate": 5.9082405363140916e-05,
+      "loss": 0.521,
+      "step": 740
+    },
+    {
+      "epoch": 2.0371134020618555,
+      "grad_norm": 0.5487285643494075,
+      "learning_rate": 5.901473865084141e-05,
+      "loss": 0.5268,
+      "step": 741
+    },
+    {
+      "epoch": 2.0398625429553263,
+      "grad_norm": 0.4395320063115494,
+      "learning_rate": 5.894700156367238e-05,
+      "loss": 0.5243,
+      "step": 742
+    },
+    {
+      "epoch": 2.042611683848797,
+      "grad_norm": 0.29982074556969307,
+      "learning_rate": 5.887919435233352e-05,
+      "loss": 0.5298,
+      "step": 743
+    },
+    {
+      "epoch": 2.045360824742268,
+      "grad_norm": 0.41119709116602937,
+      "learning_rate": 5.881131726778402e-05,
+      "loss": 0.5145,
+      "step": 744
+    },
+    {
+      "epoch": 2.0481099656357387,
+      "grad_norm": 0.384320461752947,
+      "learning_rate": 5.8743370561241715e-05,
+      "loss": 0.5253,
+      "step": 745
+    },
+    {
+      "epoch": 2.0508591065292094,
+      "grad_norm": 0.3420792544611106,
+      "learning_rate": 5.867535448418203e-05,
+      "loss": 0.5183,
+      "step": 746
+    },
+    {
+      "epoch": 2.05360824742268,
+      "grad_norm": 0.36905823023095596,
+      "learning_rate": 5.8607269288337244e-05,
+      "loss": 0.5208,
+      "step": 747
+    },
+    {
+      "epoch": 2.056357388316151,
+      "grad_norm": 0.32511574569133067,
+      "learning_rate": 5.853911522569538e-05,
+      "loss": 0.5252,
+      "step": 748
+    },
+    {
+      "epoch": 2.059106529209622,
+      "grad_norm": 0.42000247500887966,
+      "learning_rate": 5.8470892548499384e-05,
+      "loss": 0.5182,
+      "step": 749
+    },
+    {
+      "epoch": 2.0618556701030926,
+      "grad_norm": 0.5749785701633202,
+      "learning_rate": 5.840260150924609e-05,
+      "loss": 0.5172,
+      "step": 750
+    },
+    {
+      "epoch": 2.064604810996564,
+      "grad_norm": 0.6483835292108968,
+      "learning_rate": 5.833424236068541e-05,
+      "loss": 0.5336,
+      "step": 751
+    },
+    {
+      "epoch": 2.0673539518900346,
+      "grad_norm": 0.6231601301759575,
+      "learning_rate": 5.8265815355819284e-05,
+      "loss": 0.527,
+      "step": 752
+    },
+    {
+      "epoch": 2.0701030927835053,
+      "grad_norm": 0.6400482537396563,
+      "learning_rate": 5.819732074790084e-05,
+      "loss": 0.5187,
+      "step": 753
+    },
+    {
+      "epoch": 2.072852233676976,
+      "grad_norm": 0.6112696586750945,
+      "learning_rate": 5.812875879043336e-05,
+      "loss": 0.5198,
+      "step": 754
+    },
+    {
+      "epoch": 2.075601374570447,
+      "grad_norm": 0.5115706970358354,
+      "learning_rate": 5.8060129737169416e-05,
+      "loss": 0.5279,
+      "step": 755
+    },
+    {
+      "epoch": 2.0783505154639177,
+      "grad_norm": 0.4058486284297744,
+      "learning_rate": 5.799143384210992e-05,
+      "loss": 0.5268,
+      "step": 756
+    },
+    {
+      "epoch": 2.0810996563573885,
+      "grad_norm": 0.32047749300200185,
+      "learning_rate": 5.7922671359503145e-05,
+      "loss": 0.5252,
+      "step": 757
+    },
+    {
+      "epoch": 2.0838487972508593,
+      "grad_norm": 0.37004296113308727,
+      "learning_rate": 5.785384254384382e-05,
+      "loss": 0.5239,
+      "step": 758
+    },
+    {
+      "epoch": 2.08659793814433,
+      "grad_norm": 0.4458786395281163,
+      "learning_rate": 5.7784947649872176e-05,
+      "loss": 0.5294,
+      "step": 759
+    },
+    {
+      "epoch": 2.089347079037801,
+      "grad_norm": 0.4272640389491108,
+      "learning_rate": 5.7715986932573e-05,
+      "loss": 0.5249,
+      "step": 760
+    },
+    {
+      "epoch": 2.0920962199312716,
+      "grad_norm": 0.3526739356483497,
+      "learning_rate": 5.7646960647174706e-05,
+      "loss": 0.5184,
+      "step": 761
+    },
+    {
+      "epoch": 2.0948453608247424,
+      "grad_norm": 0.27188265621194263,
+      "learning_rate": 5.757786904914838e-05,
+      "loss": 0.5223,
+      "step": 762
+    },
+    {
+      "epoch": 2.097594501718213,
+      "grad_norm": 0.3457734998315455,
+      "learning_rate": 5.750871239420681e-05,
+      "loss": 0.5227,
+      "step": 763
+    },
+    {
+      "epoch": 2.100343642611684,
+      "grad_norm": 0.4222137704724379,
+      "learning_rate": 5.7439490938303604e-05,
+      "loss": 0.5273,
+      "step": 764
+    },
+    {
+      "epoch": 2.1030927835051547,
+      "grad_norm": 0.45756362105195786,
+      "learning_rate": 5.7370204937632167e-05,
+      "loss": 0.5222,
+      "step": 765
+    },
+    {
+      "epoch": 2.1058419243986255,
+      "grad_norm": 0.3435489204211452,
+      "learning_rate": 5.7300854648624824e-05,
+      "loss": 0.5102,
+      "step": 766
+    },
+    {
+      "epoch": 2.1085910652920963,
+      "grad_norm": 0.24611635871938473,
+      "learning_rate": 5.723144032795179e-05,
+      "loss": 0.5221,
+      "step": 767
+    },
+    {
+      "epoch": 2.111340206185567,
+      "grad_norm": 0.3013065227913591,
+      "learning_rate": 5.7161962232520304e-05,
+      "loss": 0.532,
+      "step": 768
+    },
+    {
+      "epoch": 2.114089347079038,
+      "grad_norm": 0.3470862495086026,
+      "learning_rate": 5.7092420619473626e-05,
+      "loss": 0.5258,
+      "step": 769
+    },
+    {
+      "epoch": 2.1168384879725086,
+      "grad_norm": 0.38943011157017127,
+      "learning_rate": 5.70228157461901e-05,
+      "loss": 0.5264,
+      "step": 770
+    },
+    {
+      "epoch": 2.1195876288659794,
+      "grad_norm": 0.39287644862912424,
+      "learning_rate": 5.69531478702822e-05,
+      "loss": 0.5114,
+      "step": 771
+    },
+    {
+      "epoch": 2.12233676975945,
+      "grad_norm": 0.2916898291723961,
+      "learning_rate": 5.688341724959557e-05,
+      "loss": 0.5275,
+      "step": 772
+    },
+    {
+      "epoch": 2.125085910652921,
+      "grad_norm": 0.22962256743492102,
+      "learning_rate": 5.681362414220811e-05,
+      "loss": 0.5214,
+      "step": 773
+    },
+    {
+      "epoch": 2.1278350515463917,
+      "grad_norm": 0.27944270631912393,
+      "learning_rate": 5.674376880642893e-05,
+      "loss": 0.5264,
+      "step": 774
+    },
+    {
+      "epoch": 2.1305841924398625,
+      "grad_norm": 0.29044286341885533,
+      "learning_rate": 5.66738515007975e-05,
+      "loss": 0.5148,
+      "step": 775
+    },
+    {
+      "epoch": 2.1333333333333333,
+      "grad_norm": 0.28701568805373207,
+      "learning_rate": 5.6603872484082614e-05,
+      "loss": 0.5218,
+      "step": 776
+    },
+    {
+      "epoch": 2.136082474226804,
+      "grad_norm": 0.33569332451853506,
+      "learning_rate": 5.653383201528151e-05,
+      "loss": 0.5244,
+      "step": 777
+    },
+    {
+      "epoch": 2.138831615120275,
+      "grad_norm": 0.2997007948052843,
+      "learning_rate": 5.6463730353618795e-05,
+      "loss": 0.5207,
+      "step": 778
+    },
+    {
+      "epoch": 2.1415807560137456,
+      "grad_norm": 0.21237174047717156,
+      "learning_rate": 5.6393567758545616e-05,
+      "loss": 0.5103,
+      "step": 779
+    },
+    {
+      "epoch": 2.1443298969072164,
+      "grad_norm": 0.2360600600158792,
+      "learning_rate": 5.63233444897386e-05,
+      "loss": 0.5293,
+      "step": 780
+    },
+    {
+      "epoch": 2.147079037800687,
+      "grad_norm": 0.23546258348715238,
+      "learning_rate": 5.625306080709895e-05,
+      "loss": 0.526,
+      "step": 781
+    },
+    {
+      "epoch": 2.149828178694158,
+      "grad_norm": 0.23742640059602604,
+      "learning_rate": 5.618271697075147e-05,
+      "loss": 0.5207,
+      "step": 782
+    },
+    {
+      "epoch": 2.1525773195876288,
+      "grad_norm": 0.2697802862240549,
+      "learning_rate": 5.611231324104358e-05,
+      "loss": 0.5185,
+      "step": 783
+    },
+    {
+      "epoch": 2.1553264604810995,
+      "grad_norm": 0.33962286687598364,
+      "learning_rate": 5.604184987854437e-05,
+      "loss": 0.5324,
+      "step": 784
+    },
+    {
+      "epoch": 2.1580756013745703,
+      "grad_norm": 0.3496574863191211,
+      "learning_rate": 5.597132714404366e-05,
+      "loss": 0.5149,
+      "step": 785
+    },
+    {
+      "epoch": 2.160824742268041,
+      "grad_norm": 0.3002103817256998,
+      "learning_rate": 5.590074529855099e-05,
+      "loss": 0.5271,
+      "step": 786
+    },
+    {
+      "epoch": 2.163573883161512,
+      "grad_norm": 0.33431676858651316,
+      "learning_rate": 5.583010460329465e-05,
+      "loss": 0.5304,
+      "step": 787
+    },
+    {
+      "epoch": 2.1663230240549827,
+      "grad_norm": 0.3298687279509311,
+      "learning_rate": 5.5759405319720806e-05,
+      "loss": 0.5167,
+      "step": 788
+    },
+    {
+      "epoch": 2.1690721649484535,
+      "grad_norm": 0.21499388088920707,
+      "learning_rate": 5.568864770949237e-05,
+      "loss": 0.536,
+      "step": 789
+    },
+    {
+      "epoch": 2.1718213058419242,
+      "grad_norm": 0.26397548444272334,
+      "learning_rate": 5.5617832034488236e-05,
+      "loss": 0.5251,
+      "step": 790
+    },
+    {
+      "epoch": 2.174570446735395,
+      "grad_norm": 0.40340132985540694,
+      "learning_rate": 5.554695855680209e-05,
+      "loss": 0.5283,
+      "step": 791
+    },
+    {
+      "epoch": 2.177319587628866,
+      "grad_norm": 0.4098335019644846,
+      "learning_rate": 5.547602753874163e-05,
+      "loss": 0.5238,
+      "step": 792
+    },
+    {
+      "epoch": 2.1800687285223366,
+      "grad_norm": 0.3531497244525965,
+      "learning_rate": 5.540503924282746e-05,
+      "loss": 0.5231,
+      "step": 793
+    },
+    {
+      "epoch": 2.1828178694158074,
+      "grad_norm": 0.38987792040546504,
+      "learning_rate": 5.5333993931792224e-05,
+      "loss": 0.5366,
+      "step": 794
+    },
+    {
+      "epoch": 2.1855670103092786,
+      "grad_norm": 0.2647915747419579,
+      "learning_rate": 5.526289186857953e-05,
+      "loss": 0.5348,
+      "step": 795
+    },
+    {
+      "epoch": 2.1883161512027494,
+      "grad_norm": 0.26788341550791417,
+      "learning_rate": 5.5191733316343074e-05,
+      "loss": 0.5253,
+      "step": 796
+    },
+    {
+      "epoch": 2.19106529209622,
+      "grad_norm": 0.31976809906665954,
+      "learning_rate": 5.5120518538445595e-05,
+      "loss": 0.5225,
+      "step": 797
+    },
+    {
+      "epoch": 2.193814432989691,
+      "grad_norm": 0.27229292264708094,
+      "learning_rate": 5.504924779845794e-05,
+      "loss": 0.5169,
+      "step": 798
+    },
+    {
+      "epoch": 2.1965635738831617,
+      "grad_norm": 0.3204542340097227,
+      "learning_rate": 5.4977921360158096e-05,
+      "loss": 0.5256,
+      "step": 799
+    },
+    {
+      "epoch": 2.1993127147766325,
+      "grad_norm": 0.33648692429827426,
+      "learning_rate": 5.490653948753013e-05,
+      "loss": 0.523,
+      "step": 800
+    },
+    {
+      "epoch": 2.2020618556701033,
+      "grad_norm": 0.25849474311106246,
+      "learning_rate": 5.483510244476332e-05,
+      "loss": 0.5345,
+      "step": 801
+    },
+    {
+      "epoch": 2.204810996563574,
+      "grad_norm": 0.2777163411097152,
+      "learning_rate": 5.476361049625115e-05,
+      "loss": 0.5217,
+      "step": 802
+    },
+    {
+      "epoch": 2.207560137457045,
+      "grad_norm": 0.259120862015438,
+      "learning_rate": 5.469206390659028e-05,
+      "loss": 0.5293,
+      "step": 803
+    },
+    {
+      "epoch": 2.2103092783505156,
+      "grad_norm": 0.25014178976390844,
+      "learning_rate": 5.46204629405796e-05,
+      "loss": 0.5269,
+      "step": 804
+    },
+    {
+      "epoch": 2.2130584192439864,
+      "grad_norm": 0.223711086431831,
+      "learning_rate": 5.454880786321928e-05,
+      "loss": 0.5187,
+      "step": 805
+    },
+    {
+      "epoch": 2.215807560137457,
+      "grad_norm": 0.2438903572639829,
+      "learning_rate": 5.447709893970974e-05,
+      "loss": 0.5209,
+      "step": 806
+    },
+    {
+      "epoch": 2.218556701030928,
+      "grad_norm": 0.2545794798823825,
+      "learning_rate": 5.4405336435450684e-05,
+      "loss": 0.527,
+      "step": 807
+    },
+    {
+      "epoch": 2.2213058419243987,
+      "grad_norm": 0.18607426816746073,
+      "learning_rate": 5.4333520616040116e-05,
+      "loss": 0.5184,
+      "step": 808
+    },
+    {
+      "epoch": 2.2240549828178695,
+      "grad_norm": 0.29001196046138,
+      "learning_rate": 5.42616517472734e-05,
+      "loss": 0.5195,
+      "step": 809
+    },
+    {
+      "epoch": 2.2268041237113403,
+      "grad_norm": 0.2664754443847999,
+      "learning_rate": 5.418973009514221e-05,
+      "loss": 0.5254,
+      "step": 810
+    },
+    {
+      "epoch": 2.229553264604811,
+      "grad_norm": 0.2763244689974836,
+      "learning_rate": 5.411775592583358e-05,
+      "loss": 0.5225,
+      "step": 811
+    },
+    {
+      "epoch": 2.232302405498282,
+      "grad_norm": 0.2952802637754747,
+      "learning_rate": 5.4045729505728884e-05,
+      "loss": 0.5227,
+      "step": 812
+    },
+    {
+      "epoch": 2.2350515463917526,
+      "grad_norm": 0.23994298354796337,
+      "learning_rate": 5.397365110140295e-05,
+      "loss": 0.5305,
+      "step": 813
+    },
+    {
+      "epoch": 2.2378006872852234,
+      "grad_norm": 0.22190622512563482,
+      "learning_rate": 5.390152097962295e-05,
+      "loss": 0.532,
+      "step": 814
+    },
+    {
+      "epoch": 2.240549828178694,
+      "grad_norm": 0.2958910982821988,
+      "learning_rate": 5.382933940734747e-05,
+      "loss": 0.5215,
+      "step": 815
+    },
+    {
+      "epoch": 2.243298969072165,
+      "grad_norm": 0.3582968358435383,
+      "learning_rate": 5.375710665172554e-05,
+      "loss": 0.5206,
+      "step": 816
+    },
+    {
+      "epoch": 2.2460481099656358,
+      "grad_norm": 0.4048685164576461,
+      "learning_rate": 5.368482298009559e-05,
+      "loss": 0.5322,
+      "step": 817
+    },
+    {
+      "epoch": 2.2487972508591065,
+      "grad_norm": 0.35928125952347756,
+      "learning_rate": 5.3612488659984534e-05,
+      "loss": 0.5256,
+      "step": 818
+    },
+    {
+      "epoch": 2.2515463917525773,
+      "grad_norm": 0.27863313688997904,
+      "learning_rate": 5.3540103959106696e-05,
+      "loss": 0.5243,
+      "step": 819
+    },
+    {
+      "epoch": 2.254295532646048,
+      "grad_norm": 0.3504201948483768,
+      "learning_rate": 5.3467669145362914e-05,
+      "loss": 0.5369,
+      "step": 820
+    },
+    {
+      "epoch": 2.257044673539519,
+      "grad_norm": 0.4374644885930723,
+      "learning_rate": 5.339518448683945e-05,
+      "loss": 0.5206,
+      "step": 821
+    },
+    {
+      "epoch": 2.2597938144329897,
+      "grad_norm": 0.2820104015984913,
+      "learning_rate": 5.332265025180707e-05,
+      "loss": 0.5201,
+      "step": 822
+    },
+    {
+      "epoch": 2.2625429553264604,
+      "grad_norm": 0.3051225307345294,
+      "learning_rate": 5.325006670872002e-05,
+      "loss": 0.5248,
+      "step": 823
+    },
+    {
+      "epoch": 2.265292096219931,
+      "grad_norm": 0.41761093729690574,
+      "learning_rate": 5.3177434126215034e-05,
+      "loss": 0.5378,
+      "step": 824
+    },
+    {
+      "epoch": 2.268041237113402,
+      "grad_norm": 0.3487478544658144,
+      "learning_rate": 5.3104752773110345e-05,
+      "loss": 0.5245,
+      "step": 825
+    },
+    {
+      "epoch": 2.270790378006873,
+      "grad_norm": 0.32194879135312154,
+      "learning_rate": 5.303202291840471e-05,
+      "loss": 0.526,
+      "step": 826
+    },
+    {
+      "epoch": 2.2735395189003436,
+      "grad_norm": 0.27245901521747345,
+      "learning_rate": 5.295924483127635e-05,
+      "loss": 0.5184,
+      "step": 827
+    },
+    {
+      "epoch": 2.2762886597938143,
+      "grad_norm": 0.26804164140224673,
+      "learning_rate": 5.2886418781082066e-05,
+      "loss": 0.5201,
+      "step": 828
+    },
+    {
+      "epoch": 2.279037800687285,
+      "grad_norm": 0.23960902170409884,
+      "learning_rate": 5.2813545037356086e-05,
+      "loss": 0.5256,
+      "step": 829
+    },
+    {
+      "epoch": 2.281786941580756,
+      "grad_norm": 0.23457140639322555,
+      "learning_rate": 5.2740623869809234e-05,
+      "loss": 0.5209,
+      "step": 830
+    },
+    {
+      "epoch": 2.2845360824742267,
+      "grad_norm": 0.27297058783320943,
+      "learning_rate": 5.2667655548327796e-05,
+      "loss": 0.5315,
+      "step": 831
+    },
+    {
+      "epoch": 2.2872852233676975,
+      "grad_norm": 0.28200029368869456,
+      "learning_rate": 5.259464034297262e-05,
+      "loss": 0.527,
+      "step": 832
+    },
+    {
+      "epoch": 2.2900343642611682,
+      "grad_norm": 0.34366587435096785,
+      "learning_rate": 5.2521578523978046e-05,
+      "loss": 0.5166,
+      "step": 833
+    },
+    {
+      "epoch": 2.292783505154639,
+      "grad_norm": 0.33716783962672386,
+      "learning_rate": 5.2448470361750955e-05,
+      "loss": 0.5271,
+      "step": 834
+    },
+    {
+      "epoch": 2.29553264604811,
+      "grad_norm": 0.2883538098562466,
+      "learning_rate": 5.237531612686973e-05,
+      "loss": 0.5326,
+      "step": 835
+    },
+    {
+      "epoch": 2.2982817869415806,
+      "grad_norm": 0.23816842715139308,
+      "learning_rate": 5.2302116090083274e-05,
+      "loss": 0.521,
+      "step": 836
+    },
+    {
+      "epoch": 2.3010309278350514,
+      "grad_norm": 0.2676295491191559,
+      "learning_rate": 5.222887052231003e-05,
+      "loss": 0.5307,
+      "step": 837
+    },
+    {
+      "epoch": 2.303780068728522,
+      "grad_norm": 0.21899952831754288,
+      "learning_rate": 5.2155579694636916e-05,
+      "loss": 0.5254,
+      "step": 838
+    },
+    {
+      "epoch": 2.306529209621993,
+      "grad_norm": 0.2725930609511096,
+      "learning_rate": 5.208224387831839e-05,
+      "loss": 0.5243,
+      "step": 839
+    },
+    {
+      "epoch": 2.3092783505154637,
+      "grad_norm": 0.2670269822849005,
+      "learning_rate": 5.200886334477541e-05,
+      "loss": 0.5263,
+      "step": 840
+    },
+    {
+      "epoch": 2.3120274914089345,
+      "grad_norm": 0.2896259602041522,
+      "learning_rate": 5.1935438365594424e-05,
+      "loss": 0.5212,
+      "step": 841
+    },
+    {
+      "epoch": 2.3147766323024053,
+      "grad_norm": 0.3040525797226736,
+      "learning_rate": 5.186196921252638e-05,
+      "loss": 0.5272,
+      "step": 842
+    },
+    {
+      "epoch": 2.317525773195876,
+      "grad_norm": 0.402527650035371,
+      "learning_rate": 5.178845615748573e-05,
+      "loss": 0.5317,
+      "step": 843
+    },
+    {
+      "epoch": 2.320274914089347,
+      "grad_norm": 0.28589840098643327,
+      "learning_rate": 5.1714899472549394e-05,
+      "loss": 0.5302,
+      "step": 844
+    },
+    {
+      "epoch": 2.323024054982818,
+      "grad_norm": 0.2746598297288049,
+      "learning_rate": 5.164129942995578e-05,
+      "loss": 0.5237,
+      "step": 845
+    },
+    {
+      "epoch": 2.325773195876289,
+      "grad_norm": 0.35895254123689857,
+      "learning_rate": 5.156765630210375e-05,
+      "loss": 0.5321,
+      "step": 846
+    },
+    {
+      "epoch": 2.3285223367697596,
+      "grad_norm": 0.3075206472125044,
+      "learning_rate": 5.1493970361551666e-05,
+      "loss": 0.5261,
+      "step": 847
+    },
+    {
+      "epoch": 2.3312714776632304,
+      "grad_norm": 0.2631726158525417,
+      "learning_rate": 5.14202418810163e-05,
+      "loss": 0.5295,
+      "step": 848
+    },
+    {
+      "epoch": 2.334020618556701,
+      "grad_norm": 0.26253313412596246,
+      "learning_rate": 5.134647113337187e-05,
+      "loss": 0.5278,
+      "step": 849
+    },
+    {
+      "epoch": 2.336769759450172,
+      "grad_norm": 0.27292981975275177,
+      "learning_rate": 5.127265839164906e-05,
+      "loss": 0.5174,
+      "step": 850
+    },
+    {
+      "epoch": 2.3395189003436427,
+      "grad_norm": 0.29830820148659387,
+      "learning_rate": 5.119880392903396e-05,
+      "loss": 0.5235,
+      "step": 851
+    },
+    {
+      "epoch": 2.3422680412371135,
+      "grad_norm": 0.319178636597951,
+      "learning_rate": 5.112490801886706e-05,
+      "loss": 0.5249,
+      "step": 852
+    },
+    {
+      "epoch": 2.3450171821305843,
+      "grad_norm": 0.39509509521849606,
+      "learning_rate": 5.105097093464225e-05,
+      "loss": 0.53,
+      "step": 853
+    },
+    {
+      "epoch": 2.347766323024055,
+      "grad_norm": 0.43481008500350815,
+      "learning_rate": 5.0976992950005836e-05,
+      "loss": 0.5222,
+      "step": 854
+    },
+    {
+      "epoch": 2.350515463917526,
+      "grad_norm": 0.35580076547200273,
+      "learning_rate": 5.090297433875549e-05,
+      "loss": 0.5341,
+      "step": 855
+    },
+    {
+      "epoch": 2.3532646048109966,
+      "grad_norm": 0.27387444348548706,
+      "learning_rate": 5.082891537483921e-05,
+      "loss": 0.527,
+      "step": 856
+    },
+    {
+      "epoch": 2.3560137457044674,
+      "grad_norm": 0.28242295705533316,
+      "learning_rate": 5.0754816332354384e-05,
+      "loss": 0.5313,
+      "step": 857
+    },
+    {
+      "epoch": 2.358762886597938,
+      "grad_norm": 0.23633783144067932,
+      "learning_rate": 5.0680677485546724e-05,
+      "loss": 0.5256,
+      "step": 858
+    },
+    {
+      "epoch": 2.361512027491409,
+      "grad_norm": 0.20611827366274735,
+      "learning_rate": 5.060649910880926e-05,
+      "loss": 0.5246,
+      "step": 859
+    },
+    {
+      "epoch": 2.3642611683848798,
+      "grad_norm": 0.24527029533975683,
+      "learning_rate": 5.0532281476681295e-05,
+      "loss": 0.526,
+      "step": 860
+    },
+    {
+      "epoch": 2.3670103092783505,
+      "grad_norm": 0.2988866489058128,
+      "learning_rate": 5.0458024863847455e-05,
+      "loss": 0.5318,
+      "step": 861
+    },
+    {
+      "epoch": 2.3697594501718213,
+      "grad_norm": 0.3278941979649635,
+      "learning_rate": 5.038372954513664e-05,
+      "loss": 0.5208,
+      "step": 862
+    },
+    {
+      "epoch": 2.372508591065292,
+      "grad_norm": 0.3590976948020781,
+      "learning_rate": 5.030939579552098e-05,
+      "loss": 0.5288,
+      "step": 863
+    },
+    {
+      "epoch": 2.375257731958763,
+      "grad_norm": 0.3257482610923462,
+      "learning_rate": 5.0235023890114844e-05,
+      "loss": 0.5342,
+      "step": 864
+    },
+    {
+      "epoch": 2.3780068728522337,
+      "grad_norm": 0.26418250148170586,
+      "learning_rate": 5.016061410417384e-05,
+      "loss": 0.534,
+      "step": 865
+    },
+    {
+      "epoch": 2.3807560137457044,
+      "grad_norm": 0.25433731862506237,
+      "learning_rate": 5.008616671309374e-05,
+      "loss": 0.5349,
+      "step": 866
+    },
+    {
+      "epoch": 2.3835051546391752,
+      "grad_norm": 0.3098360496587967,
+      "learning_rate": 5.0011681992409526e-05,
+      "loss": 0.536,
+      "step": 867
+    },
+    {
+      "epoch": 2.386254295532646,
+      "grad_norm": 0.48700691274762764,
+      "learning_rate": 4.993716021779431e-05,
+      "loss": 0.5281,
+      "step": 868
+    },
+    {
+      "epoch": 2.389003436426117,
+      "grad_norm": 0.4325007530550783,
+      "learning_rate": 4.986260166505838e-05,
+      "loss": 0.5287,
+      "step": 869
+    },
+    {
+      "epoch": 2.3917525773195876,
+      "grad_norm": 0.40355654494603005,
+      "learning_rate": 4.978800661014811e-05,
+      "loss": 0.5211,
+      "step": 870
+    },
+    {
+      "epoch": 2.3945017182130583,
+      "grad_norm": 0.26544768260770013,
+      "learning_rate": 4.9713375329145e-05,
+      "loss": 0.5195,
+      "step": 871
+    },
+    {
+      "epoch": 2.397250859106529,
+      "grad_norm": 0.2791642651045501,
+      "learning_rate": 4.963870809826458e-05,
+      "loss": 0.5265,
+      "step": 872
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.3408862865130159,
+      "learning_rate": 4.9564005193855486e-05,
+      "loss": 0.5391,
+      "step": 873
+    },
+    {
+      "epoch": 2.4027491408934707,
+      "grad_norm": 0.3331718549433266,
+      "learning_rate": 4.9489266892398346e-05,
+      "loss": 0.5216,
+      "step": 874
+    },
+    {
+      "epoch": 2.4054982817869415,
+      "grad_norm": 0.33450428391908577,
+      "learning_rate": 4.9414493470504834e-05,
+      "loss": 0.5274,
+      "step": 875
+    },
+    {
+      "epoch": 2.4082474226804123,
+      "grad_norm": 0.324321171570019,
+      "learning_rate": 4.933968520491654e-05,
+      "loss": 0.5373,
+      "step": 876
+    },
+    {
+      "epoch": 2.410996563573883,
+      "grad_norm": 0.28829591882553246,
+      "learning_rate": 4.9264842372504095e-05,
+      "loss": 0.5286,
+      "step": 877
+    },
+    {
+      "epoch": 2.413745704467354,
+      "grad_norm": 0.28746004519536006,
+      "learning_rate": 4.9189965250265994e-05,
+      "loss": 0.5199,
+      "step": 878
+    },
+    {
+      "epoch": 2.4164948453608246,
+      "grad_norm": 0.31445005880135923,
+      "learning_rate": 4.911505411532769e-05,
+      "loss": 0.5278,
+      "step": 879
+    },
+    {
+      "epoch": 2.4192439862542954,
+      "grad_norm": 0.47892737901970533,
+      "learning_rate": 4.9040109244940505e-05,
+      "loss": 0.5234,
+      "step": 880
+    },
+    {
+      "epoch": 2.421993127147766,
+      "grad_norm": 0.45215275967709107,
+      "learning_rate": 4.896513091648058e-05,
+      "loss": 0.535,
+      "step": 881
+    },
+    {
+      "epoch": 2.424742268041237,
+      "grad_norm": 0.34138582396119665,
+      "learning_rate": 4.889011940744796e-05,
+      "loss": 0.5164,
+      "step": 882
+    },
+    {
+      "epoch": 2.4274914089347077,
+      "grad_norm": 0.2682120490670279,
+      "learning_rate": 4.881507499546545e-05,
+      "loss": 0.5342,
+      "step": 883
+    },
+    {
+      "epoch": 2.430240549828179,
+      "grad_norm": 0.2692048784370494,
+      "learning_rate": 4.873999795827761e-05,
+      "loss": 0.5344,
+      "step": 884
+    },
+    {
+      "epoch": 2.4329896907216497,
+      "grad_norm": 0.3048032911190028,
+      "learning_rate": 4.866488857374979e-05,
+      "loss": 0.5131,
+      "step": 885
+    },
+    {
+      "epoch": 2.4357388316151205,
+      "grad_norm": 0.43798907642210033,
+      "learning_rate": 4.858974711986704e-05,
+      "loss": 0.5277,
+      "step": 886
+    },
+    {
+      "epoch": 2.4384879725085913,
+      "grad_norm": 0.4537544665998809,
+      "learning_rate": 4.851457387473312e-05,
+      "loss": 0.5272,
+      "step": 887
+    },
+    {
+      "epoch": 2.441237113402062,
+      "grad_norm": 0.275619974234131,
+      "learning_rate": 4.843936911656941e-05,
+      "loss": 0.5229,
+      "step": 888
+    },
+    {
+      "epoch": 2.443986254295533,
+      "grad_norm": 0.24697116802006067,
+      "learning_rate": 4.836413312371394e-05,
+      "loss": 0.5237,
+      "step": 889
+    },
+    {
+      "epoch": 2.4467353951890036,
+      "grad_norm": 0.2889992360936922,
+      "learning_rate": 4.828886617462039e-05,
+      "loss": 0.5214,
+      "step": 890
+    },
+    {
+      "epoch": 2.4494845360824744,
+      "grad_norm": 0.34236892659433343,
+      "learning_rate": 4.821356854785695e-05,
+      "loss": 0.5204,
+      "step": 891
+    },
+    {
+      "epoch": 2.452233676975945,
+      "grad_norm": 0.26986931392665453,
+      "learning_rate": 4.8138240522105365e-05,
+      "loss": 0.5285,
+      "step": 892
+    },
+    {
+      "epoch": 2.454982817869416,
+      "grad_norm": 0.2386727777106889,
+      "learning_rate": 4.806288237615989e-05,
+      "loss": 0.5258,
+      "step": 893
+    },
+    {
+      "epoch": 2.4577319587628867,
+      "grad_norm": 0.2400107390785469,
+      "learning_rate": 4.7987494388926275e-05,
+      "loss": 0.523,
+      "step": 894
+    },
+    {
+      "epoch": 2.4604810996563575,
+      "grad_norm": 0.30720815238105814,
+      "learning_rate": 4.7912076839420695e-05,
+      "loss": 0.5232,
+      "step": 895
+    },
+    {
+      "epoch": 2.4632302405498283,
+      "grad_norm": 0.25563044453886097,
+      "learning_rate": 4.7836630006768746e-05,
+      "loss": 0.5289,
+      "step": 896
+    },
+    {
+      "epoch": 2.465979381443299,
+      "grad_norm": 0.2618995944120738,
+      "learning_rate": 4.77611541702044e-05,
+      "loss": 0.5256,
+      "step": 897
+    },
+    {
+      "epoch": 2.46872852233677,
+      "grad_norm": 0.2872261399026541,
+      "learning_rate": 4.768564960906897e-05,
+      "loss": 0.527,
+      "step": 898
+    },
+    {
+      "epoch": 2.4714776632302407,
+      "grad_norm": 0.22134537204990035,
+      "learning_rate": 4.7610116602810096e-05,
+      "loss": 0.5343,
+      "step": 899
+    },
+    {
+      "epoch": 2.4742268041237114,
+      "grad_norm": 0.2565389994931295,
+      "learning_rate": 4.753455543098067e-05,
+      "loss": 0.5375,
+      "step": 900
+    },
+    {
+      "epoch": 2.476975945017182,
+      "grad_norm": 0.3740627121098048,
+      "learning_rate": 4.745896637323785e-05,
+      "loss": 0.5227,
+      "step": 901
+    },
+    {
+      "epoch": 2.479725085910653,
+      "grad_norm": 0.3714016967888814,
+      "learning_rate": 4.7383349709341994e-05,
+      "loss": 0.5216,
+      "step": 902
+    },
+    {
+      "epoch": 2.4824742268041238,
+      "grad_norm": 0.2863979507280106,
+      "learning_rate": 4.730770571915562e-05,
+      "loss": 0.5289,
+      "step": 903
+    },
+    {
+      "epoch": 2.4852233676975946,
+      "grad_norm": 0.2639569172110602,
+      "learning_rate": 4.7232034682642417e-05,
+      "loss": 0.5207,
+      "step": 904
+    },
+    {
+      "epoch": 2.4879725085910653,
+      "grad_norm": 0.26506986793397436,
+      "learning_rate": 4.715633687986613e-05,
+      "loss": 0.5276,
+      "step": 905
+    },
+    {
+      "epoch": 2.490721649484536,
+      "grad_norm": 0.3034302978268054,
+      "learning_rate": 4.7080612590989596e-05,
+      "loss": 0.5185,
+      "step": 906
+    },
+    {
+      "epoch": 2.493470790378007,
+      "grad_norm": 0.3242741496552742,
+      "learning_rate": 4.7004862096273674e-05,
+      "loss": 0.5292,
+      "step": 907
+    },
+    {
+      "epoch": 2.4962199312714777,
+      "grad_norm": 0.25198934191589384,
+      "learning_rate": 4.692908567607621e-05,
+      "loss": 0.5209,
+      "step": 908
+    },
+    {
+      "epoch": 2.4989690721649485,
+      "grad_norm": 0.2851823027369076,
+      "learning_rate": 4.6853283610851004e-05,
+      "loss": 0.5332,
+      "step": 909
+    },
+    {
+      "epoch": 2.5017182130584192,
+      "grad_norm": 0.2250586051327087,
+      "learning_rate": 4.677745618114674e-05,
+      "loss": 0.5326,
+      "step": 910
+    },
+    {
+      "epoch": 2.50446735395189,
+      "grad_norm": 0.2137916864449506,
+      "learning_rate": 4.670160366760606e-05,
+      "loss": 0.5214,
+      "step": 911
+    },
+    {
+      "epoch": 2.507216494845361,
+      "grad_norm": 0.23692573516235416,
+      "learning_rate": 4.6625726350964355e-05,
+      "loss": 0.5209,
+      "step": 912
+    },
+    {
+      "epoch": 2.5099656357388316,
+      "grad_norm": 0.2503089565651227,
+      "learning_rate": 4.654982451204885e-05,
+      "loss": 0.5228,
+      "step": 913
+    },
+    {
+      "epoch": 2.5127147766323024,
+      "grad_norm": 0.187246659026728,
+      "learning_rate": 4.6473898431777535e-05,
+      "loss": 0.5279,
+      "step": 914
+    },
+    {
+      "epoch": 2.515463917525773,
+      "grad_norm": 0.22968241737897607,
+      "learning_rate": 4.6397948391158104e-05,
+      "loss": 0.5229,
+      "step": 915
+    },
+    {
+      "epoch": 2.518213058419244,
+      "grad_norm": 0.34560044485812286,
+      "learning_rate": 4.632197467128695e-05,
+      "loss": 0.521,
+      "step": 916
+    },
+    {
+      "epoch": 2.5209621993127147,
+      "grad_norm": 0.36369842042382355,
+      "learning_rate": 4.624597755334807e-05,
+      "loss": 0.5222,
+      "step": 917
+    },
+    {
+      "epoch": 2.5237113402061855,
+      "grad_norm": 0.3273459802424622,
+      "learning_rate": 4.6169957318612096e-05,
+      "loss": 0.5269,
+      "step": 918
+    },
+    {
+      "epoch": 2.5264604810996563,
+      "grad_norm": 0.27046737413984473,
+      "learning_rate": 4.609391424843519e-05,
+      "loss": 0.5207,
+      "step": 919
+    },
+    {
+      "epoch": 2.529209621993127,
+      "grad_norm": 0.23973509685168032,
+      "learning_rate": 4.601784862425807e-05,
+      "loss": 0.5305,
+      "step": 920
+    },
+    {
+      "epoch": 2.531958762886598,
+      "grad_norm": 0.207584971564237,
+      "learning_rate": 4.594176072760485e-05,
+      "loss": 0.5391,
+      "step": 921
+    },
+    {
+      "epoch": 2.5347079037800686,
+      "grad_norm": 0.26389513176009577,
+      "learning_rate": 4.586565084008217e-05,
+      "loss": 0.5257,
+      "step": 922
+    },
+    {
+      "epoch": 2.5374570446735394,
+      "grad_norm": 0.27492964134545905,
+      "learning_rate": 4.5789519243377975e-05,
+      "loss": 0.5198,
+      "step": 923
+    },
+    {
+      "epoch": 2.54020618556701,
+      "grad_norm": 0.2550669423258001,
+      "learning_rate": 4.571336621926065e-05,
+      "loss": 0.5161,
+      "step": 924
+    },
+    {
+      "epoch": 2.542955326460481,
+      "grad_norm": 0.2434323842617188,
+      "learning_rate": 4.563719204957776e-05,
+      "loss": 0.5266,
+      "step": 925
+    },
+    {
+      "epoch": 2.5457044673539517,
+      "grad_norm": 0.2381537603353079,
+      "learning_rate": 4.5560997016255265e-05,
+      "loss": 0.5306,
+      "step": 926
+    },
+    {
+      "epoch": 2.5484536082474225,
+      "grad_norm": 0.25877756360414456,
+      "learning_rate": 4.548478140129624e-05,
+      "loss": 0.5279,
+      "step": 927
+    },
+    {
+      "epoch": 2.5512027491408933,
+      "grad_norm": 0.23404565240810715,
+      "learning_rate": 4.5408545486779996e-05,
+      "loss": 0.5325,
+      "step": 928
+    },
+    {
+      "epoch": 2.553951890034364,
+      "grad_norm": 0.20244068111276767,
+      "learning_rate": 4.533228955486094e-05,
+      "loss": 0.5245,
+      "step": 929
+    },
+    {
+      "epoch": 2.556701030927835,
+      "grad_norm": 0.20414770829038065,
+      "learning_rate": 4.525601388776758e-05,
+      "loss": 0.5193,
+      "step": 930
+    },
+    {
+      "epoch": 2.5594501718213056,
+      "grad_norm": 0.22610091916102834,
+      "learning_rate": 4.517971876780147e-05,
+      "loss": 0.5335,
+      "step": 931
+    },
+    {
+      "epoch": 2.5621993127147764,
+      "grad_norm": 0.2030707662947604,
+      "learning_rate": 4.5103404477336144e-05,
+      "loss": 0.5247,
+      "step": 932
+    },
+    {
+      "epoch": 2.564948453608247,
+      "grad_norm": 0.19565686931462759,
+      "learning_rate": 4.502707129881609e-05,
+      "loss": 0.5223,
+      "step": 933
+    },
+    {
+      "epoch": 2.567697594501718,
+      "grad_norm": 0.22241975043349305,
+      "learning_rate": 4.495071951475572e-05,
+      "loss": 0.5289,
+      "step": 934
+    },
+    {
+      "epoch": 2.5704467353951888,
+      "grad_norm": 0.25405146475942325,
+      "learning_rate": 4.487434940773828e-05,
+      "loss": 0.5216,
+      "step": 935
+    },
+    {
+      "epoch": 2.5731958762886595,
+      "grad_norm": 0.24090685844792697,
+      "learning_rate": 4.479796126041487e-05,
+      "loss": 0.5347,
+      "step": 936
+    },
+    {
+      "epoch": 2.5759450171821303,
+      "grad_norm": 0.20268315657201996,
+      "learning_rate": 4.472155535550331e-05,
+      "loss": 0.5174,
+      "step": 937
+    },
+    {
+      "epoch": 2.5786941580756015,
+      "grad_norm": 0.2164104030727069,
+      "learning_rate": 4.464513197578717e-05,
+      "loss": 0.5306,
+      "step": 938
+    },
+    {
+      "epoch": 2.5814432989690723,
+      "grad_norm": 0.20675687136906504,
+      "learning_rate": 4.45686914041147e-05,
+      "loss": 0.5217,
+      "step": 939
+    },
+    {
+      "epoch": 2.584192439862543,
+      "grad_norm": 0.2495476073768104,
+      "learning_rate": 4.449223392339776e-05,
+      "loss": 0.5233,
+      "step": 940
+    },
+    {
+      "epoch": 2.586941580756014,
+      "grad_norm": 0.2188055318879783,
+      "learning_rate": 4.44157598166108e-05,
+      "loss": 0.536,
+      "step": 941
+    },
+    {
+      "epoch": 2.5896907216494847,
+      "grad_norm": 0.19661959776334886,
+      "learning_rate": 4.43392693667898e-05,
+      "loss": 0.5167,
+      "step": 942
+    },
+    {
+      "epoch": 2.5924398625429554,
+      "grad_norm": 0.21281895098022574,
+      "learning_rate": 4.426276285703125e-05,
+      "loss": 0.5241,
+      "step": 943
+    },
+    {
+      "epoch": 2.5951890034364262,
+      "grad_norm": 0.24612261354981668,
+      "learning_rate": 4.418624057049106e-05,
+      "loss": 0.5288,
+      "step": 944
+    },
+    {
+      "epoch": 2.597938144329897,
+      "grad_norm": 0.26256604053792143,
+      "learning_rate": 4.410970279038351e-05,
+      "loss": 0.528,
+      "step": 945
+    },
+    {
+      "epoch": 2.600687285223368,
+      "grad_norm": 0.2146570261022671,
+      "learning_rate": 4.4033149799980265e-05,
+      "loss": 0.5336,
+      "step": 946
+    },
+    {
+      "epoch": 2.6034364261168386,
+      "grad_norm": 0.20864992098912383,
+      "learning_rate": 4.395658188260924e-05,
+      "loss": 0.5289,
+      "step": 947
+    },
+    {
+      "epoch": 2.6061855670103093,
+      "grad_norm": 0.23475812934644366,
+      "learning_rate": 4.3879999321653664e-05,
+      "loss": 0.5294,
+      "step": 948
+    },
+    {
+      "epoch": 2.60893470790378,
+      "grad_norm": 0.21552160156786482,
+      "learning_rate": 4.380340240055087e-05,
+      "loss": 0.5225,
+      "step": 949
+    },
+    {
+      "epoch": 2.611683848797251,
+      "grad_norm": 0.21451249484405446,
+      "learning_rate": 4.3726791402791405e-05,
+      "loss": 0.5309,
+      "step": 950
+    },
+    {
+      "epoch": 2.6144329896907217,
+      "grad_norm": 0.2228499608780446,
+      "learning_rate": 4.3650166611917904e-05,
+      "loss": 0.5218,
+      "step": 951
+    },
+    {
+      "epoch": 2.6171821305841925,
+      "grad_norm": 0.19761673711898892,
+      "learning_rate": 4.3573528311524056e-05,
+      "loss": 0.5263,
+      "step": 952
+    },
+    {
+      "epoch": 2.6199312714776632,
+      "grad_norm": 0.2717760446302159,
+      "learning_rate": 4.34968767852535e-05,
+      "loss": 0.5192,
+      "step": 953
+    },
+    {
+      "epoch": 2.622680412371134,
+      "grad_norm": 0.2307774736780584,
+      "learning_rate": 4.3420212316798895e-05,
+      "loss": 0.5224,
+      "step": 954
+    },
+    {
+      "epoch": 2.625429553264605,
+      "grad_norm": 0.26499404963083745,
+      "learning_rate": 4.334353518990075e-05,
+      "loss": 0.5284,
+      "step": 955
+    },
+    {
+      "epoch": 2.6281786941580756,
+      "grad_norm": 0.33949301525073333,
+      "learning_rate": 4.326684568834647e-05,
+      "loss": 0.5215,
+      "step": 956
+    },
+    {
+      "epoch": 2.6309278350515464,
+      "grad_norm": 0.33940964847094485,
+      "learning_rate": 4.31901440959692e-05,
+      "loss": 0.527,
+      "step": 957
+    },
+    {
+      "epoch": 2.633676975945017,
+      "grad_norm": 0.3034703314779094,
+      "learning_rate": 4.311343069664688e-05,
+      "loss": 0.5315,
+      "step": 958
+    },
+    {
+      "epoch": 2.636426116838488,
+      "grad_norm": 0.3285064736514596,
+      "learning_rate": 4.303670577430111e-05,
+      "loss": 0.5328,
+      "step": 959
+    },
+    {
+      "epoch": 2.6391752577319587,
+      "grad_norm": 0.3453164976173602,
+      "learning_rate": 4.295996961289619e-05,
+      "loss": 0.5323,
+      "step": 960
+    },
+    {
+      "epoch": 2.6419243986254295,
+      "grad_norm": 0.38444181745132805,
+      "learning_rate": 4.288322249643799e-05,
+      "loss": 0.5288,
+      "step": 961
+    },
+    {
+      "epoch": 2.6446735395189003,
+      "grad_norm": 0.2686311128572819,
+      "learning_rate": 4.2806464708972905e-05,
+      "loss": 0.525,
+      "step": 962
+    },
+    {
+      "epoch": 2.647422680412371,
+      "grad_norm": 0.3232991261053769,
+      "learning_rate": 4.272969653458685e-05,
+      "loss": 0.5278,
+      "step": 963
+    },
+    {
+      "epoch": 2.650171821305842,
+      "grad_norm": 0.32877598170142874,
+      "learning_rate": 4.265291825740416e-05,
+      "loss": 0.5176,
+      "step": 964
+    },
+    {
+      "epoch": 2.6529209621993126,
+      "grad_norm": 0.30844002727607134,
+      "learning_rate": 4.257613016158661e-05,
+      "loss": 0.5326,
+      "step": 965
+    },
+    {
+      "epoch": 2.6556701030927834,
+      "grad_norm": 0.2255639863284666,
+      "learning_rate": 4.249933253133224e-05,
+      "loss": 0.5266,
+      "step": 966
+    },
+    {
+      "epoch": 2.658419243986254,
+      "grad_norm": 0.25200678515207303,
+      "learning_rate": 4.2422525650874446e-05,
+      "loss": 0.5292,
+      "step": 967
+    },
+    {
+      "epoch": 2.661168384879725,
+      "grad_norm": 0.22993658765349623,
+      "learning_rate": 4.234570980448081e-05,
+      "loss": 0.5251,
+      "step": 968
+    },
+    {
+      "epoch": 2.6639175257731957,
+      "grad_norm": 0.2642010038722434,
+      "learning_rate": 4.226888527645215e-05,
+      "loss": 0.529,
+      "step": 969
+    },
+    {
+      "epoch": 2.6666666666666665,
+      "grad_norm": 0.28025388603567647,
+      "learning_rate": 4.219205235112135e-05,
+      "loss": 0.523,
+      "step": 970
+    },
+    {
+      "epoch": 2.6694158075601373,
+      "grad_norm": 0.24286835186859662,
+      "learning_rate": 4.2115211312852435e-05,
+      "loss": 0.5241,
+      "step": 971
+    },
+    {
+      "epoch": 2.6721649484536085,
+      "grad_norm": 0.260911275304071,
+      "learning_rate": 4.203836244603941e-05,
+      "loss": 0.5251,
+      "step": 972
+    },
+    {
+      "epoch": 2.6749140893470793,
+      "grad_norm": 0.23336974876081842,
+      "learning_rate": 4.1961506035105285e-05,
+      "loss": 0.5337,
+      "step": 973
+    },
+    {
+      "epoch": 2.67766323024055,
+      "grad_norm": 0.19717415494962492,
+      "learning_rate": 4.188464236450098e-05,
+      "loss": 0.518,
+      "step": 974
+    },
+    {
+      "epoch": 2.680412371134021,
+      "grad_norm": 0.2635705526316712,
+      "learning_rate": 4.180777171870427e-05,
+      "loss": 0.519,
+      "step": 975
+    },
+    {
+      "epoch": 2.6831615120274916,
+      "grad_norm": 0.2420795496999561,
+      "learning_rate": 4.173089438221876e-05,
+      "loss": 0.5329,
+      "step": 976
+    },
+    {
+      "epoch": 2.6859106529209624,
+      "grad_norm": 0.24604439574458017,
+      "learning_rate": 4.165401063957283e-05,
+      "loss": 0.5331,
+      "step": 977
+    },
+    {
+      "epoch": 2.688659793814433,
+      "grad_norm": 0.1948605941664403,
+      "learning_rate": 4.157712077531856e-05,
+      "loss": 0.5221,
+      "step": 978
+    },
+    {
+      "epoch": 2.691408934707904,
+      "grad_norm": 0.21270792355056445,
+      "learning_rate": 4.1500225074030654e-05,
+      "loss": 0.5355,
+      "step": 979
+    },
+    {
+      "epoch": 2.6941580756013748,
+      "grad_norm": 0.251861479262255,
+      "learning_rate": 4.142332382030547e-05,
+      "loss": 0.5336,
+      "step": 980
+    },
+    {
+      "epoch": 2.6969072164948455,
+      "grad_norm": 0.2112157139724383,
+      "learning_rate": 4.13464172987599e-05,
+      "loss": 0.5243,
+      "step": 981
+    },
+    {
+      "epoch": 2.6996563573883163,
+      "grad_norm": 0.20387549860274823,
+      "learning_rate": 4.1269505794030296e-05,
+      "loss": 0.5237,
+      "step": 982
+    },
+    {
+      "epoch": 2.702405498281787,
+      "grad_norm": 0.2515717159268574,
+      "learning_rate": 4.119258959077151e-05,
+      "loss": 0.5319,
+      "step": 983
+    },
+    {
+      "epoch": 2.705154639175258,
+      "grad_norm": 0.20847504323318589,
+      "learning_rate": 4.111566897365575e-05,
+      "loss": 0.52,
+      "step": 984
+    },
+    {
+      "epoch": 2.7079037800687287,
+      "grad_norm": 0.21911840052600298,
+      "learning_rate": 4.103874422737157e-05,
+      "loss": 0.5283,
+      "step": 985
+    },
+    {
+      "epoch": 2.7106529209621995,
+      "grad_norm": 0.2049343201857368,
+      "learning_rate": 4.096181563662279e-05,
+      "loss": 0.5263,
+      "step": 986
+    },
+    {
+      "epoch": 2.7134020618556702,
+      "grad_norm": 0.1947687795273586,
+      "learning_rate": 4.088488348612748e-05,
+      "loss": 0.5239,
+      "step": 987
+    },
+    {
+      "epoch": 2.716151202749141,
+      "grad_norm": 0.23438600392965853,
+      "learning_rate": 4.080794806061688e-05,
+      "loss": 0.5235,
+      "step": 988
+    },
+    {
+      "epoch": 2.718900343642612,
+      "grad_norm": 0.19851824128921414,
+      "learning_rate": 4.073100964483435e-05,
+      "loss": 0.5236,
+      "step": 989
+    },
+    {
+      "epoch": 2.7216494845360826,
+      "grad_norm": 0.23677872375792366,
+      "learning_rate": 4.06540685235343e-05,
+      "loss": 0.5307,
+      "step": 990
+    },
+    {
+      "epoch": 2.7243986254295534,
+      "grad_norm": 0.19484452355877876,
+      "learning_rate": 4.057712498148119e-05,
+      "loss": 0.533,
+      "step": 991
+    },
+    {
+      "epoch": 2.727147766323024,
+      "grad_norm": 0.24691106342879912,
+      "learning_rate": 4.05001793034484e-05,
+      "loss": 0.5282,
+      "step": 992
+    },
+    {
+      "epoch": 2.729896907216495,
+      "grad_norm": 0.23911573502762684,
+      "learning_rate": 4.0423231774217246e-05,
+      "loss": 0.5291,
+      "step": 993
+    },
+    {
+      "epoch": 2.7326460481099657,
+      "grad_norm": 0.24784394785326042,
+      "learning_rate": 4.034628267857587e-05,
+      "loss": 0.5299,
+      "step": 994
+    },
+    {
+      "epoch": 2.7353951890034365,
+      "grad_norm": 0.2530406886888911,
+      "learning_rate": 4.026933230131823e-05,
+      "loss": 0.5155,
+      "step": 995
+    },
+    {
+      "epoch": 2.7381443298969073,
+      "grad_norm": 0.24815572290601853,
+      "learning_rate": 4.0192380927243014e-05,
+      "loss": 0.5331,
+      "step": 996
+    },
+    {
+      "epoch": 2.740893470790378,
+      "grad_norm": 0.2756774245673258,
+      "learning_rate": 4.0115428841152637e-05,
+      "loss": 0.5334,
+      "step": 997
+    },
+    {
+      "epoch": 2.743642611683849,
+      "grad_norm": 0.24077726898420207,
+      "learning_rate": 4.0038476327852065e-05,
+      "loss": 0.5337,
+      "step": 998
+    },
+    {
+      "epoch": 2.7463917525773196,
+      "grad_norm": 0.20128889819397291,
+      "learning_rate": 3.9961523672147955e-05,
+      "loss": 0.5302,
+      "step": 999
+    },
+    {
+      "epoch": 2.7491408934707904,
+      "grad_norm": 0.20443940819422327,
+      "learning_rate": 3.988457115884739e-05,
+      "loss": 0.5284,
+      "step": 1000
+    },
+    {
+      "epoch": 2.751890034364261,
+      "grad_norm": 0.2208826727313727,
+      "learning_rate": 3.9807619072757e-05,
+      "loss": 0.5271,
+      "step": 1001
+    },
+    {
+      "epoch": 2.754639175257732,
+      "grad_norm": 0.24447959308366699,
+      "learning_rate": 3.973066769868178e-05,
+      "loss": 0.5283,
+      "step": 1002
+    },
+    {
+      "epoch": 2.7573883161512027,
+      "grad_norm": 0.1966957237573264,
+      "learning_rate": 3.965371732142415e-05,
+      "loss": 0.5292,
+      "step": 1003
+    },
+    {
+      "epoch": 2.7601374570446735,
+      "grad_norm": 0.21603427189344282,
+      "learning_rate": 3.957676822578276e-05,
+      "loss": 0.5237,
+      "step": 1004
+    },
+    {
+      "epoch": 2.7628865979381443,
+      "grad_norm": 0.18664057521742963,
+      "learning_rate": 3.949982069655161e-05,
+      "loss": 0.5233,
+      "step": 1005
+    },
+    {
+      "epoch": 2.765635738831615,
+      "grad_norm": 0.2234326703320951,
+      "learning_rate": 3.942287501851881e-05,
+      "loss": 0.5256,
+      "step": 1006
+    },
+    {
+      "epoch": 2.768384879725086,
+      "grad_norm": 0.2278011506232875,
+      "learning_rate": 3.9345931476465706e-05,
+      "loss": 0.5279,
+      "step": 1007
+    },
+    {
+      "epoch": 2.7711340206185566,
+      "grad_norm": 0.19798832981596853,
+      "learning_rate": 3.9268990355165664e-05,
+      "loss": 0.5211,
+      "step": 1008
+    },
+    {
+      "epoch": 2.7738831615120274,
+      "grad_norm": 0.2183804005479682,
+      "learning_rate": 3.9192051939383126e-05,
+      "loss": 0.5239,
+      "step": 1009
+    },
+    {
+      "epoch": 2.776632302405498,
+      "grad_norm": 0.22246588980190424,
+      "learning_rate": 3.911511651387253e-05,
+      "loss": 0.527,
+      "step": 1010
+    },
+    {
+      "epoch": 2.779381443298969,
+      "grad_norm": 0.1780086146278406,
+      "learning_rate": 3.903818436337722e-05,
+      "loss": 0.5284,
+      "step": 1011
+    },
+    {
+      "epoch": 2.7821305841924397,
+      "grad_norm": 0.19604513054901485,
+      "learning_rate": 3.896125577262845e-05,
+      "loss": 0.5162,
+      "step": 1012
+    },
+    {
+      "epoch": 2.7848797250859105,
+      "grad_norm": 0.18794858233130177,
+      "learning_rate": 3.888433102634425e-05,
+      "loss": 0.5216,
+      "step": 1013
+    },
+    {
+      "epoch": 2.7876288659793813,
+      "grad_norm": 0.21435085046499355,
+      "learning_rate": 3.8807410409228496e-05,
+      "loss": 0.5305,
+      "step": 1014
+    },
+    {
+      "epoch": 2.790378006872852,
+      "grad_norm": 0.17410246023922102,
+      "learning_rate": 3.8730494205969724e-05,
+      "loss": 0.5312,
+      "step": 1015
+    },
+    {
+      "epoch": 2.793127147766323,
+      "grad_norm": 0.1822314934502394,
+      "learning_rate": 3.865358270124013e-05,
+      "loss": 0.5189,
+      "step": 1016
+    },
+    {
+      "epoch": 2.7958762886597937,
+      "grad_norm": 0.206717747798629,
+      "learning_rate": 3.857667617969454e-05,
+      "loss": 0.5269,
+      "step": 1017
+    },
+    {
+      "epoch": 2.7986254295532644,
+      "grad_norm": 0.16863448749904908,
+      "learning_rate": 3.849977492596936e-05,
+      "loss": 0.5303,
+      "step": 1018
+    },
+    {
+      "epoch": 2.801374570446735,
+      "grad_norm": 0.23449573918173383,
+      "learning_rate": 3.8422879224681456e-05,
+      "loss": 0.5322,
+      "step": 1019
+    },
+    {
+      "epoch": 2.804123711340206,
+      "grad_norm": 0.17234697630039422,
+      "learning_rate": 3.8345989360427174e-05,
+      "loss": 0.5229,
+      "step": 1020
+    },
+    {
+      "epoch": 2.8068728522336768,
+      "grad_norm": 0.21339720138725765,
+      "learning_rate": 3.826910561778124e-05,
+      "loss": 0.5271,
+      "step": 1021
+    },
+    {
+      "epoch": 2.8096219931271476,
+      "grad_norm": 0.22865589786306695,
+      "learning_rate": 3.819222828129574e-05,
+      "loss": 0.5261,
+      "step": 1022
+    },
+    {
+      "epoch": 2.8123711340206183,
+      "grad_norm": 0.19395333537771012,
+      "learning_rate": 3.8115357635499045e-05,
+      "loss": 0.5307,
+      "step": 1023
+    },
+    {
+      "epoch": 2.815120274914089,
+      "grad_norm": 0.26952248831929015,
+      "learning_rate": 3.803849396489473e-05,
+      "loss": 0.5304,
+      "step": 1024
+    },
+    {
+      "epoch": 2.81786941580756,
+      "grad_norm": 0.2078676429133822,
+      "learning_rate": 3.7961637553960605e-05,
+      "loss": 0.521,
+      "step": 1025
+    },
+    {
+      "epoch": 2.8206185567010307,
+      "grad_norm": 0.2437865253535147,
+      "learning_rate": 3.788478868714758e-05,
+      "loss": 0.5284,
+      "step": 1026
+    },
+    {
+      "epoch": 2.8233676975945015,
+      "grad_norm": 0.23244204464146223,
+      "learning_rate": 3.780794764887866e-05,
+      "loss": 0.5347,
+      "step": 1027
+    },
+    {
+      "epoch": 2.8261168384879722,
+      "grad_norm": 0.1883460034103066,
+      "learning_rate": 3.7731114723547856e-05,
+      "loss": 0.5332,
+      "step": 1028
+    },
+    {
+      "epoch": 2.8288659793814435,
+      "grad_norm": 0.18889412109996112,
+      "learning_rate": 3.7654290195519195e-05,
+      "loss": 0.527,
+      "step": 1029
+    },
+    {
+      "epoch": 2.8316151202749142,
+      "grad_norm": 0.23037525353226912,
+      "learning_rate": 3.757747434912556e-05,
+      "loss": 0.5293,
+      "step": 1030
+    },
+    {
+      "epoch": 2.834364261168385,
+      "grad_norm": 0.21033380977426863,
+      "learning_rate": 3.750066746866778e-05,
+      "loss": 0.5226,
+      "step": 1031
+    },
+    {
+      "epoch": 2.837113402061856,
+      "grad_norm": 0.17225724075961382,
+      "learning_rate": 3.742386983841341e-05,
+      "loss": 0.5244,
+      "step": 1032
+    },
+    {
+      "epoch": 2.8398625429553266,
+      "grad_norm": 0.22879502842160185,
+      "learning_rate": 3.734708174259585e-05,
+      "loss": 0.5291,
+      "step": 1033
+    },
+    {
+      "epoch": 2.8426116838487974,
+      "grad_norm": 0.1859425423671784,
+      "learning_rate": 3.727030346541317e-05,
+      "loss": 0.5207,
+      "step": 1034
+    },
+    {
+      "epoch": 2.845360824742268,
+      "grad_norm": 0.21138789242685446,
+      "learning_rate": 3.7193535291027115e-05,
+      "loss": 0.5283,
+      "step": 1035
+    },
+    {
+      "epoch": 2.848109965635739,
+      "grad_norm": 0.22384899016314974,
+      "learning_rate": 3.7116777503562016e-05,
+      "loss": 0.528,
+      "step": 1036
+    },
+    {
+      "epoch": 2.8508591065292097,
+      "grad_norm": 0.2431535434052746,
+      "learning_rate": 3.7040030387103815e-05,
+      "loss": 0.5187,
+      "step": 1037
+    },
+    {
+      "epoch": 2.8536082474226805,
+      "grad_norm": 0.24399503353207871,
+      "learning_rate": 3.696329422569889e-05,
+      "loss": 0.5179,
+      "step": 1038
+    },
+    {
+      "epoch": 2.8563573883161513,
+      "grad_norm": 0.16921161494449857,
+      "learning_rate": 3.6886569303353136e-05,
+      "loss": 0.5292,
+      "step": 1039
+    },
+    {
+      "epoch": 2.859106529209622,
+      "grad_norm": 0.26728522647583636,
+      "learning_rate": 3.680985590403082e-05,
+      "loss": 0.5262,
+      "step": 1040
+    },
+    {
+      "epoch": 2.861855670103093,
+      "grad_norm": 0.27421229932307095,
+      "learning_rate": 3.673315431165355e-05,
+      "loss": 0.5252,
+      "step": 1041
+    },
+    {
+      "epoch": 2.8646048109965636,
+      "grad_norm": 0.2601105298366,
+      "learning_rate": 3.665646481009926e-05,
+      "loss": 0.5203,
+      "step": 1042
+    },
+    {
+      "epoch": 2.8673539518900344,
+      "grad_norm": 0.2997837262368804,
+      "learning_rate": 3.657978768320111e-05,
+      "loss": 0.5266,
+      "step": 1043
+    },
+    {
+      "epoch": 2.870103092783505,
+      "grad_norm": 0.1739319697716107,
+      "learning_rate": 3.650312321474651e-05,
+      "loss": 0.5285,
+      "step": 1044
+    },
+    {
+      "epoch": 2.872852233676976,
+      "grad_norm": 0.2868154357660894,
+      "learning_rate": 3.642647168847596e-05,
+      "loss": 0.5349,
+      "step": 1045
+    },
+    {
+      "epoch": 2.8756013745704467,
+      "grad_norm": 0.3150816781576196,
+      "learning_rate": 3.6349833388082096e-05,
+      "loss": 0.5213,
+      "step": 1046
+    },
+    {
+      "epoch": 2.8783505154639175,
+      "grad_norm": 0.21703226563550534,
+      "learning_rate": 3.6273208597208595e-05,
+      "loss": 0.5302,
+      "step": 1047
+    },
+    {
+      "epoch": 2.8810996563573883,
+      "grad_norm": 0.24416409278757065,
+      "learning_rate": 3.619659759944916e-05,
+      "loss": 0.5249,
+      "step": 1048
+    },
+    {
+      "epoch": 2.883848797250859,
+      "grad_norm": 0.2807527048682551,
+      "learning_rate": 3.612000067834636e-05,
+      "loss": 0.5249,
+      "step": 1049
+    },
+    {
+      "epoch": 2.88659793814433,
+      "grad_norm": 0.2744380467215706,
+      "learning_rate": 3.604341811739077e-05,
+      "loss": 0.5305,
+      "step": 1050
+    },
+    {
+      "epoch": 2.8893470790378006,
+      "grad_norm": 0.2271292993546667,
+      "learning_rate": 3.596685020001975e-05,
+      "loss": 0.5194,
+      "step": 1051
+    },
+    {
+      "epoch": 2.8920962199312714,
+      "grad_norm": 0.21515311729133313,
+      "learning_rate": 3.5890297209616507e-05,
+      "loss": 0.5242,
+      "step": 1052
+    },
+    {
+      "epoch": 2.894845360824742,
+      "grad_norm": 0.2904407669489142,
+      "learning_rate": 3.581375942950895e-05,
+      "loss": 0.5163,
+      "step": 1053
+    },
+    {
+      "epoch": 2.897594501718213,
+      "grad_norm": 0.23987566400445187,
+      "learning_rate": 3.5737237142968755e-05,
+      "loss": 0.5321,
+      "step": 1054
+    },
+    {
+      "epoch": 2.9003436426116838,
+      "grad_norm": 0.203928269390139,
+      "learning_rate": 3.56607306332102e-05,
+      "loss": 0.5203,
+      "step": 1055
+    },
+    {
+      "epoch": 2.9030927835051545,
+      "grad_norm": 0.26314705579076864,
+      "learning_rate": 3.558424018338922e-05,
+      "loss": 0.5267,
+      "step": 1056
+    },
+    {
+      "epoch": 2.9058419243986253,
+      "grad_norm": 0.22978092693646193,
+      "learning_rate": 3.5507766076602264e-05,
+      "loss": 0.5211,
+      "step": 1057
+    },
+    {
+      "epoch": 2.908591065292096,
+      "grad_norm": 0.1831775606016915,
+      "learning_rate": 3.5431308595885316e-05,
+      "loss": 0.5186,
+      "step": 1058
+    },
+    {
+      "epoch": 2.911340206185567,
+      "grad_norm": 0.2557072250330763,
+      "learning_rate": 3.535486802421284e-05,
+      "loss": 0.5204,
+      "step": 1059
+    },
+    {
+      "epoch": 2.9140893470790377,
+      "grad_norm": 0.2285012181274199,
+      "learning_rate": 3.5278444644496695e-05,
+      "loss": 0.5273,
+      "step": 1060
+    },
+    {
+      "epoch": 2.9168384879725084,
+      "grad_norm": 0.18355449593774995,
+      "learning_rate": 3.520203873958514e-05,
+      "loss": 0.531,
+      "step": 1061
+    },
+    {
+      "epoch": 2.9195876288659792,
+      "grad_norm": 0.21328292549083275,
+      "learning_rate": 3.512565059226172e-05,
+      "loss": 0.5229,
+      "step": 1062
+    },
+    {
+      "epoch": 2.9223367697594504,
+      "grad_norm": 0.21044538382380515,
+      "learning_rate": 3.5049280485244286e-05,
+      "loss": 0.5231,
+      "step": 1063
+    },
+    {
+      "epoch": 2.9250859106529212,
+      "grad_norm": 0.17119562274819797,
+      "learning_rate": 3.4972928701183925e-05,
+      "loss": 0.522,
+      "step": 1064
+    },
+    {
+      "epoch": 2.927835051546392,
+      "grad_norm": 0.22581900383360864,
+      "learning_rate": 3.489659552266388e-05,
+      "loss": 0.5246,
+      "step": 1065
+    },
+    {
+      "epoch": 2.930584192439863,
+      "grad_norm": 0.301379881236332,
+      "learning_rate": 3.482028123219855e-05,
+      "loss": 0.529,
+      "step": 1066
+    },
+    {
+      "epoch": 2.9333333333333336,
+      "grad_norm": 0.295319936733804,
+      "learning_rate": 3.4743986112232434e-05,
+      "loss": 0.5289,
+      "step": 1067
+    },
+    {
+      "epoch": 2.9360824742268044,
+      "grad_norm": 0.2388909601572767,
+      "learning_rate": 3.466771044513907e-05,
+      "loss": 0.5252,
+      "step": 1068
+    },
+    {
+      "epoch": 2.938831615120275,
+      "grad_norm": 0.2678249461923104,
+      "learning_rate": 3.459145451322002e-05,
+      "loss": 0.5262,
+      "step": 1069
+    },
+    {
+      "epoch": 2.941580756013746,
+      "grad_norm": 0.22322825864211238,
+      "learning_rate": 3.4515218598703765e-05,
+      "loss": 0.5297,
+      "step": 1070
+    },
+    {
+      "epoch": 2.9443298969072167,
+      "grad_norm": 0.179417097733591,
+      "learning_rate": 3.443900298374475e-05,
+      "loss": 0.5249,
+      "step": 1071
+    },
+    {
+      "epoch": 2.9470790378006875,
+      "grad_norm": 0.2551012950816274,
+      "learning_rate": 3.436280795042225e-05,
+      "loss": 0.5212,
+      "step": 1072
+    },
+    {
+      "epoch": 2.9498281786941583,
+      "grad_norm": 0.25848561223138594,
+      "learning_rate": 3.428663378073937e-05,
+      "loss": 0.528,
+      "step": 1073
+    },
+    {
+      "epoch": 2.952577319587629,
+      "grad_norm": 0.17137059703558605,
+      "learning_rate": 3.421048075662203e-05,
+      "loss": 0.524,
+      "step": 1074
+    },
+    {
+      "epoch": 2.9553264604811,
+      "grad_norm": 0.21809992041709833,
+      "learning_rate": 3.413434915991784e-05,
+      "loss": 0.5225,
+      "step": 1075
+    },
+    {
+      "epoch": 2.9580756013745706,
+      "grad_norm": 0.19281232602243703,
+      "learning_rate": 3.4058239272395156e-05,
+      "loss": 0.5198,
+      "step": 1076
+    },
+    {
+      "epoch": 2.9608247422680414,
+      "grad_norm": 0.18758042777019787,
+      "learning_rate": 3.398215137574194e-05,
+      "loss": 0.5145,
+      "step": 1077
+    },
+    {
+      "epoch": 2.963573883161512,
+      "grad_norm": 0.24212160366698007,
+      "learning_rate": 3.390608575156481e-05,
+      "loss": 0.529,
+      "step": 1078
+    },
+    {
+      "epoch": 2.966323024054983,
+      "grad_norm": 0.19965723896338422,
+      "learning_rate": 3.3830042681387904e-05,
+      "loss": 0.5345,
+      "step": 1079
+    },
+    {
+      "epoch": 2.9690721649484537,
+      "grad_norm": 0.16936178177221572,
+      "learning_rate": 3.375402244665194e-05,
+      "loss": 0.5208,
+      "step": 1080
+    },
+    {
+      "epoch": 2.9718213058419245,
+      "grad_norm": 0.20843282321759193,
+      "learning_rate": 3.367802532871306e-05,
+      "loss": 0.5254,
+      "step": 1081
+    },
+    {
+      "epoch": 2.9745704467353953,
+      "grad_norm": 0.1837127741621114,
+      "learning_rate": 3.360205160884191e-05,
+      "loss": 0.5313,
+      "step": 1082
+    },
+    {
+      "epoch": 2.977319587628866,
+      "grad_norm": 0.199460376426822,
+      "learning_rate": 3.352610156822248e-05,
+      "loss": 0.5292,
+      "step": 1083
+    },
+    {
+      "epoch": 2.980068728522337,
+      "grad_norm": 0.24820207311148457,
+      "learning_rate": 3.345017548795116e-05,
+      "loss": 0.5257,
+      "step": 1084
+    },
+    {
+      "epoch": 2.9828178694158076,
+      "grad_norm": 0.24493105512545987,
+      "learning_rate": 3.337427364903565e-05,
+      "loss": 0.5224,
+      "step": 1085
+    },
+    {
+      "epoch": 2.9855670103092784,
+      "grad_norm": 0.23786424160796374,
+      "learning_rate": 3.329839633239395e-05,
+      "loss": 0.5304,
+      "step": 1086
+    },
+    {
+      "epoch": 2.988316151202749,
+      "grad_norm": 0.27502699218694276,
+      "learning_rate": 3.322254381885325e-05,
+      "loss": 0.5281,
+      "step": 1087
+    },
+    {
+      "epoch": 2.99106529209622,
+      "grad_norm": 0.27049323273784326,
+      "learning_rate": 3.314671638914902e-05,
+      "loss": 0.5301,
+      "step": 1088
+    },
+    {
+      "epoch": 2.9938144329896907,
+      "grad_norm": 0.19737402579388286,
+      "learning_rate": 3.307091432392382e-05,
+      "loss": 0.5288,
+      "step": 1089
+    },
+    {
+      "epoch": 2.9965635738831615,
+      "grad_norm": 0.26232022578272246,
+      "learning_rate": 3.299513790372634e-05,
+      "loss": 0.521,
+      "step": 1090
+    },
+    {
+      "epoch": 2.9993127147766323,
+      "grad_norm": 0.23892201559966744,
+      "learning_rate": 3.2919387409010424e-05,
+      "loss": 0.6184,
+      "step": 1091
+    },
+    {
+      "epoch": 3.002061855670103,
+      "grad_norm": 0.3594917888207816,
+      "learning_rate": 3.284366312013388e-05,
+      "loss": 0.5355,
+      "step": 1092
+    },
+    {
+      "epoch": 3.004810996563574,
+      "grad_norm": 0.25599834111682307,
+      "learning_rate": 3.27679653173576e-05,
+      "loss": 0.4746,
+      "step": 1093
+    },
+    {
+      "epoch": 3.0075601374570446,
+      "grad_norm": 0.2254148080474178,
+      "learning_rate": 3.2692294280844374e-05,
+      "loss": 0.4669,
+      "step": 1094
+    },
+    {
+      "epoch": 3.0103092783505154,
+      "grad_norm": 0.2506660939835137,
+      "learning_rate": 3.261665029065801e-05,
+      "loss": 0.4682,
+      "step": 1095
+    },
+    {
+      "epoch": 3.013058419243986,
+      "grad_norm": 0.267195831442756,
+      "learning_rate": 3.254103362676217e-05,
+      "loss": 0.4715,
+      "step": 1096
+    },
+    {
+      "epoch": 3.015807560137457,
+      "grad_norm": 0.24179014490822495,
+      "learning_rate": 3.246544456901934e-05,
+      "loss": 0.4778,
+      "step": 1097
+    },
+    {
+      "epoch": 3.0185567010309278,
+      "grad_norm": 0.2329847121267379,
+      "learning_rate": 3.238988339718992e-05,
+      "loss": 0.4747,
+      "step": 1098
+    },
+    {
+      "epoch": 3.0213058419243985,
+      "grad_norm": 0.2043691678048163,
+      "learning_rate": 3.231435039093104e-05,
+      "loss": 0.4786,
+      "step": 1099
+    },
+    {
+      "epoch": 3.0240549828178693,
+      "grad_norm": 0.22745344698265332,
+      "learning_rate": 3.223884582979561e-05,
+      "loss": 0.4657,
+      "step": 1100
+    },
+    {
+      "epoch": 3.02680412371134,
+      "grad_norm": 0.22934476133037818,
+      "learning_rate": 3.216336999323127e-05,
+      "loss": 0.4889,
+      "step": 1101
+    },
+    {
+      "epoch": 3.029553264604811,
+      "grad_norm": 0.21801654584396532,
+      "learning_rate": 3.208792316057931e-05,
+      "loss": 0.4784,
+      "step": 1102
+    },
+    {
+      "epoch": 3.0323024054982817,
+      "grad_norm": 0.2431718990003574,
+      "learning_rate": 3.201250561107374e-05,
+      "loss": 0.4797,
+      "step": 1103
+    },
+    {
+      "epoch": 3.0350515463917525,
+      "grad_norm": 0.26915847727829645,
+      "learning_rate": 3.1937117623840114e-05,
+      "loss": 0.4826,
+      "step": 1104
+    },
+    {
+      "epoch": 3.0378006872852232,
+      "grad_norm": 0.2351108919147438,
+      "learning_rate": 3.1861759477894656e-05,
+      "loss": 0.4772,
+      "step": 1105
+    },
+    {
+      "epoch": 3.040549828178694,
+      "grad_norm": 0.2904532318483722,
+      "learning_rate": 3.178643145214307e-05,
+      "loss": 0.4803,
+      "step": 1106
+    },
+    {
+      "epoch": 3.043298969072165,
+      "grad_norm": 0.19635036055007402,
+      "learning_rate": 3.171113382537962e-05,
+      "loss": 0.4701,
+      "step": 1107
+    },
+    {
+      "epoch": 3.0460481099656356,
+      "grad_norm": 0.22223894507419642,
+      "learning_rate": 3.1635866876286064e-05,
+      "loss": 0.4741,
+      "step": 1108
+    },
+    {
+      "epoch": 3.0487972508591064,
+      "grad_norm": 0.2152706605758009,
+      "learning_rate": 3.156063088343061e-05,
+      "loss": 0.4746,
+      "step": 1109
+    },
+    {
+      "epoch": 3.051546391752577,
+      "grad_norm": 0.17409361205810703,
+      "learning_rate": 3.1485426125266896e-05,
+      "loss": 0.478,
+      "step": 1110
+    },
+    {
+      "epoch": 3.054295532646048,
+      "grad_norm": 0.21539232573809372,
+      "learning_rate": 3.141025288013296e-05,
+      "loss": 0.4706,
+      "step": 1111
+    },
+    {
+      "epoch": 3.0570446735395187,
+      "grad_norm": 0.19308259974547204,
+      "learning_rate": 3.1335111426250216e-05,
+      "loss": 0.4777,
+      "step": 1112
+    },
+    {
+      "epoch": 3.0597938144329895,
+      "grad_norm": 0.2276406440928922,
+      "learning_rate": 3.1260002041722405e-05,
+      "loss": 0.4823,
+      "step": 1113
+    },
+    {
+      "epoch": 3.0625429553264603,
+      "grad_norm": 0.1977395826548686,
+      "learning_rate": 3.118492500453456e-05,
+      "loss": 0.4781,
+      "step": 1114
+    },
+    {
+      "epoch": 3.0652920962199315,
+      "grad_norm": 0.17401988491729847,
+      "learning_rate": 3.1109880592552046e-05,
+      "loss": 0.474,
+      "step": 1115
+    },
+    {
+      "epoch": 3.0680412371134023,
+      "grad_norm": 0.1912931420538338,
+      "learning_rate": 3.103486908351943e-05,
+      "loss": 0.4763,
+      "step": 1116
+    },
+    {
+      "epoch": 3.070790378006873,
+      "grad_norm": 0.16713338636093636,
+      "learning_rate": 3.0959890755059515e-05,
+      "loss": 0.4796,
+      "step": 1117
+    },
+    {
+      "epoch": 3.073539518900344,
+      "grad_norm": 0.20702313839329717,
+      "learning_rate": 3.088494588467232e-05,
+      "loss": 0.4778,
+      "step": 1118
+    },
+    {
+      "epoch": 3.0762886597938146,
+      "grad_norm": 0.1616831225744788,
+      "learning_rate": 3.081003474973401e-05,
+      "loss": 0.4756,
+      "step": 1119
+    },
+    {
+      "epoch": 3.0790378006872854,
+      "grad_norm": 0.24558159371403487,
+      "learning_rate": 3.0735157627495925e-05,
+      "loss": 0.4789,
+      "step": 1120
+    },
+    {
+      "epoch": 3.081786941580756,
+      "grad_norm": 0.17163392192032517,
+      "learning_rate": 3.0660314795083475e-05,
+      "loss": 0.4768,
+      "step": 1121
+    },
+    {
+      "epoch": 3.084536082474227,
+      "grad_norm": 0.21495669147653343,
+      "learning_rate": 3.0585506529495186e-05,
+      "loss": 0.4757,
+      "step": 1122
+    },
+    {
+      "epoch": 3.0872852233676977,
+      "grad_norm": 0.15045704205058852,
+      "learning_rate": 3.0510733107601664e-05,
+      "loss": 0.473,
+      "step": 1123
+    },
+    {
+      "epoch": 3.0900343642611685,
+      "grad_norm": 0.2020087192391845,
+      "learning_rate": 3.0435994806144524e-05,
+      "loss": 0.4735,
+      "step": 1124
+    },
+    {
+      "epoch": 3.0927835051546393,
+      "grad_norm": 0.15940495290357568,
+      "learning_rate": 3.0361291901735428e-05,
+      "loss": 0.4709,
+      "step": 1125
+    },
+    {
+      "epoch": 3.09553264604811,
+      "grad_norm": 0.18904118815363216,
+      "learning_rate": 3.028662467085501e-05,
+      "loss": 0.4838,
+      "step": 1126
+    },
+    {
+      "epoch": 3.098281786941581,
+      "grad_norm": 0.17129808608598124,
+      "learning_rate": 3.0211993389851893e-05,
+      "loss": 0.4768,
+      "step": 1127
+    },
+    {
+      "epoch": 3.1010309278350516,
+      "grad_norm": 0.18416866902814874,
+      "learning_rate": 3.013739833494162e-05,
+      "loss": 0.4718,
+      "step": 1128
+    },
+    {
+      "epoch": 3.1037800687285224,
+      "grad_norm": 0.1705293761247234,
+      "learning_rate": 3.00628397822057e-05,
+      "loss": 0.4791,
+      "step": 1129
+    },
+    {
+      "epoch": 3.106529209621993,
+      "grad_norm": 0.20209221200923272,
+      "learning_rate": 2.9988318007590494e-05,
+      "loss": 0.4823,
+      "step": 1130
+    },
+    {
+      "epoch": 3.109278350515464,
+      "grad_norm": 0.2118645657673755,
+      "learning_rate": 2.991383328690627e-05,
+      "loss": 0.4782,
+      "step": 1131
+    },
+    {
+      "epoch": 3.1120274914089348,
+      "grad_norm": 0.17556823451574127,
+      "learning_rate": 2.9839385895826175e-05,
+      "loss": 0.4783,
+      "step": 1132
+    },
+    {
+      "epoch": 3.1147766323024055,
+      "grad_norm": 0.19795461740770434,
+      "learning_rate": 2.9764976109885166e-05,
+      "loss": 0.4818,
+      "step": 1133
+    },
+    {
+      "epoch": 3.1175257731958763,
+      "grad_norm": 0.16701933368048089,
+      "learning_rate": 2.9690604204479026e-05,
+      "loss": 0.476,
+      "step": 1134
+    },
+    {
+      "epoch": 3.120274914089347,
+      "grad_norm": 0.20618027859459478,
+      "learning_rate": 2.9616270454863368e-05,
+      "loss": 0.4798,
+      "step": 1135
+    },
+    {
+      "epoch": 3.123024054982818,
+      "grad_norm": 0.1431586178012281,
+      "learning_rate": 2.9541975136152548e-05,
+      "loss": 0.478,
+      "step": 1136
+    },
+    {
+      "epoch": 3.1257731958762887,
+      "grad_norm": 0.17793879100992058,
+      "learning_rate": 2.946771852331873e-05,
+      "loss": 0.4756,
+      "step": 1137
+    },
+    {
+      "epoch": 3.1285223367697594,
+      "grad_norm": 0.16681103898774272,
+      "learning_rate": 2.939350089119077e-05,
+      "loss": 0.4715,
+      "step": 1138
+    },
+    {
+      "epoch": 3.13127147766323,
+      "grad_norm": 0.17261193531928995,
+      "learning_rate": 2.9319322514453286e-05,
+      "loss": 0.4759,
+      "step": 1139
+    },
+    {
+      "epoch": 3.134020618556701,
+      "grad_norm": 0.1582640372762462,
+      "learning_rate": 2.9245183667645626e-05,
+      "loss": 0.4749,
+      "step": 1140
+    },
+    {
+      "epoch": 3.136769759450172,
+      "grad_norm": 0.17322436049835935,
+      "learning_rate": 2.9171084625160797e-05,
+      "loss": 0.4703,
+      "step": 1141
+    },
+    {
+      "epoch": 3.1395189003436426,
+      "grad_norm": 0.17367955729692558,
+      "learning_rate": 2.909702566124452e-05,
+      "loss": 0.4792,
+      "step": 1142
+    },
+    {
+      "epoch": 3.1422680412371133,
+      "grad_norm": 0.16722751357988883,
+      "learning_rate": 2.9023007049994157e-05,
+      "loss": 0.4717,
+      "step": 1143
+    },
+    {
+      "epoch": 3.145017182130584,
+      "grad_norm": 0.1728042548273763,
+      "learning_rate": 2.8949029065357753e-05,
+      "loss": 0.4818,
+      "step": 1144
+    },
+    {
+      "epoch": 3.147766323024055,
+      "grad_norm": 0.18069700472194267,
+      "learning_rate": 2.8875091981132963e-05,
+      "loss": 0.476,
+      "step": 1145
+    },
+    {
+      "epoch": 3.1505154639175257,
+      "grad_norm": 0.15820067646687608,
+      "learning_rate": 2.8801196070966054e-05,
+      "loss": 0.4666,
+      "step": 1146
+    },
+    {
+      "epoch": 3.1532646048109965,
+      "grad_norm": 0.15258574864023677,
+      "learning_rate": 2.8727341608350952e-05,
+      "loss": 0.4759,
+      "step": 1147
+    },
+    {
+      "epoch": 3.1560137457044672,
+      "grad_norm": 0.1960418259666068,
+      "learning_rate": 2.8653528866628132e-05,
+      "loss": 0.4769,
+      "step": 1148
+    },
+    {
+      "epoch": 3.158762886597938,
+      "grad_norm": 0.15371731621163243,
+      "learning_rate": 2.8579758118983716e-05,
+      "loss": 0.4842,
+      "step": 1149
+    },
+    {
+      "epoch": 3.161512027491409,
+      "grad_norm": 0.19068409005542658,
+      "learning_rate": 2.8506029638448347e-05,
+      "loss": 0.4776,
+      "step": 1150
+    },
+    {
+      "epoch": 3.1642611683848796,
+      "grad_norm": 0.17886606119360748,
+      "learning_rate": 2.843234369789625e-05,
+      "loss": 0.4744,
+      "step": 1151
+    },
+    {
+      "epoch": 3.1670103092783504,
+      "grad_norm": 0.19710185432909114,
+      "learning_rate": 2.8358700570044232e-05,
+      "loss": 0.4813,
+      "step": 1152
+    },
+    {
+      "epoch": 3.169759450171821,
+      "grad_norm": 0.18041901946396155,
+      "learning_rate": 2.8285100527450623e-05,
+      "loss": 0.4833,
+      "step": 1153
+    },
+    {
+      "epoch": 3.172508591065292,
+      "grad_norm": 0.16581599590174512,
+      "learning_rate": 2.8211543842514288e-05,
+      "loss": 0.4739,
+      "step": 1154
+    },
+    {
+      "epoch": 3.1752577319587627,
+      "grad_norm": 0.15872506391538035,
+      "learning_rate": 2.8138030787473635e-05,
+      "loss": 0.4854,
+      "step": 1155
+    },
+    {
+      "epoch": 3.1780068728522335,
+      "grad_norm": 0.18595712745784665,
+      "learning_rate": 2.806456163440559e-05,
+      "loss": 0.4791,
+      "step": 1156
+    },
+    {
+      "epoch": 3.1807560137457043,
+      "grad_norm": 0.1585604017012495,
+      "learning_rate": 2.7991136655224602e-05,
+      "loss": 0.4794,
+      "step": 1157
+    },
+    {
+      "epoch": 3.183505154639175,
+      "grad_norm": 0.16724675223224328,
+      "learning_rate": 2.791775612168161e-05,
+      "loss": 0.484,
+      "step": 1158
+    },
+    {
+      "epoch": 3.1862542955326463,
+      "grad_norm": 0.1598241225286718,
+      "learning_rate": 2.784442030536309e-05,
+      "loss": 0.4816,
+      "step": 1159
+    },
+    {
+      "epoch": 3.189003436426117,
+      "grad_norm": 0.16901071883801533,
+      "learning_rate": 2.7771129477689972e-05,
+      "loss": 0.473,
+      "step": 1160
+    },
+    {
+      "epoch": 3.191752577319588,
+      "grad_norm": 0.13853638340372265,
+      "learning_rate": 2.7697883909916732e-05,
+      "loss": 0.4736,
+      "step": 1161
+    },
+    {
+      "epoch": 3.1945017182130586,
+      "grad_norm": 0.16091561527631862,
+      "learning_rate": 2.7624683873130287e-05,
+      "loss": 0.4805,
+      "step": 1162
+    },
+    {
+      "epoch": 3.1972508591065294,
+      "grad_norm": 0.14598087437104398,
+      "learning_rate": 2.755152963824905e-05,
+      "loss": 0.4835,
+      "step": 1163
+    },
+    {
+      "epoch": 3.2,
+      "grad_norm": 0.175532838361529,
+      "learning_rate": 2.7478421476021968e-05,
+      "loss": 0.4743,
+      "step": 1164
+    },
+    {
+      "epoch": 3.202749140893471,
+      "grad_norm": 0.1337685529203635,
+      "learning_rate": 2.740535965702739e-05,
+      "loss": 0.4734,
+      "step": 1165
+    },
+    {
+      "epoch": 3.2054982817869417,
+      "grad_norm": 0.1735634483137809,
+      "learning_rate": 2.7332344451672214e-05,
+      "loss": 0.4709,
+      "step": 1166
+    },
+    {
+      "epoch": 3.2082474226804125,
+      "grad_norm": 0.15693421713978023,
+      "learning_rate": 2.7259376130190783e-05,
+      "loss": 0.4741,
+      "step": 1167
+    },
+    {
+      "epoch": 3.2109965635738833,
+      "grad_norm": 0.1330436144853641,
+      "learning_rate": 2.718645496264392e-05,
+      "loss": 0.4739,
+      "step": 1168
+    },
+    {
+      "epoch": 3.213745704467354,
+      "grad_norm": 0.17309959852487516,
+      "learning_rate": 2.711358121891795e-05,
+      "loss": 0.4705,
+      "step": 1169
+    },
+    {
+      "epoch": 3.216494845360825,
+      "grad_norm": 0.1642984807751716,
+      "learning_rate": 2.704075516872366e-05,
+      "loss": 0.4726,
+      "step": 1170
+    },
+    {
+      "epoch": 3.2192439862542956,
+      "grad_norm": 0.14722892838781462,
+      "learning_rate": 2.6967977081595304e-05,
+      "loss": 0.4775,
+      "step": 1171
+    },
+    {
+      "epoch": 3.2219931271477664,
+      "grad_norm": 0.16887717979369643,
+      "learning_rate": 2.6895247226889672e-05,
+      "loss": 0.4738,
+      "step": 1172
+    },
+    {
+      "epoch": 3.224742268041237,
+      "grad_norm": 0.17049853048281954,
+      "learning_rate": 2.682256587378498e-05,
+      "loss": 0.4719,
+      "step": 1173
+    },
+    {
+      "epoch": 3.227491408934708,
+      "grad_norm": 0.15138446686573354,
+      "learning_rate": 2.6749933291279994e-05,
+      "loss": 0.4856,
+      "step": 1174
+    },
+    {
+      "epoch": 3.2302405498281788,
+      "grad_norm": 0.16148723512701157,
+      "learning_rate": 2.6677349748192934e-05,
+      "loss": 0.4762,
+      "step": 1175
+    },
+    {
+      "epoch": 3.2329896907216495,
+      "grad_norm": 0.14912677660800724,
+      "learning_rate": 2.6604815513160556e-05,
+      "loss": 0.4653,
+      "step": 1176
+    },
+    {
+      "epoch": 3.2357388316151203,
+      "grad_norm": 0.13862596201937702,
+      "learning_rate": 2.6532330854637086e-05,
+      "loss": 0.4743,
+      "step": 1177
+    },
+    {
+      "epoch": 3.238487972508591,
+      "grad_norm": 0.15927565474045197,
+      "learning_rate": 2.645989604089331e-05,
+      "loss": 0.4746,
+      "step": 1178
+    },
+    {
+      "epoch": 3.241237113402062,
+      "grad_norm": 0.17434623830674337,
+      "learning_rate": 2.638751134001549e-05,
+      "loss": 0.48,
+      "step": 1179
+    },
+    {
+      "epoch": 3.2439862542955327,
+      "grad_norm": 0.1447492759368682,
+      "learning_rate": 2.6315177019904423e-05,
+      "loss": 0.4781,
+      "step": 1180
+    },
+    {
+      "epoch": 3.2467353951890034,
+      "grad_norm": 0.16927963210867344,
+      "learning_rate": 2.624289334827448e-05,
+      "loss": 0.4704,
+      "step": 1181
+    },
+    {
+      "epoch": 3.2494845360824742,
+      "grad_norm": 0.1665258090219738,
+      "learning_rate": 2.6170660592652545e-05,
+      "loss": 0.4781,
+      "step": 1182
+    },
+    {
+      "epoch": 3.252233676975945,
+      "grad_norm": 0.15855482153444905,
+      "learning_rate": 2.609847902037706e-05,
+      "loss": 0.4743,
+      "step": 1183
+    },
+    {
+      "epoch": 3.254982817869416,
+      "grad_norm": 0.17971555626146168,
+      "learning_rate": 2.6026348898597057e-05,
+      "loss": 0.4803,
+      "step": 1184
+    },
+    {
+      "epoch": 3.2577319587628866,
+      "grad_norm": 0.2816535496494584,
+      "learning_rate": 2.5954270494271116e-05,
+      "loss": 0.4761,
+      "step": 1185
+    },
+    {
+      "epoch": 3.2604810996563574,
+      "grad_norm": 0.23781070038647936,
+      "learning_rate": 2.588224407416645e-05,
+      "loss": 0.4729,
+      "step": 1186
+    },
+    {
+      "epoch": 3.263230240549828,
+      "grad_norm": 0.23075459880557803,
+      "learning_rate": 2.581026990485781e-05,
+      "loss": 0.4764,
+      "step": 1187
+    },
+    {
+      "epoch": 3.265979381443299,
+      "grad_norm": 0.14086275865969622,
+      "learning_rate": 2.5738348252726607e-05,
+      "loss": 0.4809,
+      "step": 1188
+    },
+    {
+      "epoch": 3.2687285223367697,
+      "grad_norm": 0.18978376603435007,
+      "learning_rate": 2.566647938395989e-05,
+      "loss": 0.4817,
+      "step": 1189
+    },
+    {
+      "epoch": 3.2714776632302405,
+      "grad_norm": 0.16648197235336828,
+      "learning_rate": 2.559466356454933e-05,
+      "loss": 0.4703,
+      "step": 1190
+    },
+    {
+      "epoch": 3.2742268041237113,
+      "grad_norm": 0.17731919695127468,
+      "learning_rate": 2.5522901060290272e-05,
+      "loss": 0.4757,
+      "step": 1191
+    },
+    {
+      "epoch": 3.276975945017182,
+      "grad_norm": 0.22533634711616554,
+      "learning_rate": 2.545119213678072e-05,
+      "loss": 0.4779,
+      "step": 1192
+    },
+    {
+      "epoch": 3.279725085910653,
+      "grad_norm": 0.29341047419363175,
+      "learning_rate": 2.53795370594204e-05,
+      "loss": 0.4809,
+      "step": 1193
+    },
+    {
+      "epoch": 3.2824742268041236,
+      "grad_norm": 0.17396435349550476,
+      "learning_rate": 2.530793609340974e-05,
+      "loss": 0.4796,
+      "step": 1194
+    },
+    {
+      "epoch": 3.2852233676975944,
+      "grad_norm": 0.20383849713641408,
+      "learning_rate": 2.523638950374886e-05,
+      "loss": 0.4721,
+      "step": 1195
+    },
+    {
+      "epoch": 3.287972508591065,
+      "grad_norm": 0.16601271855097022,
+      "learning_rate": 2.5164897555236686e-05,
+      "loss": 0.4804,
+      "step": 1196
+    },
+    {
+      "epoch": 3.290721649484536,
+      "grad_norm": 0.30470051683065796,
+      "learning_rate": 2.509346051246988e-05,
+      "loss": 0.4792,
+      "step": 1197
+    },
+    {
+      "epoch": 3.2934707903780067,
+      "grad_norm": 0.2812298427453436,
+      "learning_rate": 2.5022078639841918e-05,
+      "loss": 0.4782,
+      "step": 1198
+    },
+    {
+      "epoch": 3.2962199312714775,
+      "grad_norm": 0.1597994542210946,
+      "learning_rate": 2.495075220154206e-05,
+      "loss": 0.4823,
+      "step": 1199
+    },
+    {
+      "epoch": 3.2989690721649483,
+      "grad_norm": 0.3095999568945906,
+      "learning_rate": 2.4879481461554405e-05,
+      "loss": 0.4762,
+      "step": 1200
+    },
+    {
+      "epoch": 3.301718213058419,
+      "grad_norm": 0.24822013052998354,
+      "learning_rate": 2.4808266683656932e-05,
+      "loss": 0.4805,
+      "step": 1201
+    },
+    {
+      "epoch": 3.30446735395189,
+      "grad_norm": 0.19304267454670382,
+      "learning_rate": 2.473710813142049e-05,
+      "loss": 0.4709,
+      "step": 1202
+    },
+    {
+      "epoch": 3.3072164948453606,
+      "grad_norm": 0.22511180187157134,
+      "learning_rate": 2.4666006068207793e-05,
+      "loss": 0.4791,
+      "step": 1203
+    },
+    {
+      "epoch": 3.3099656357388314,
+      "grad_norm": 0.17597702188477718,
+      "learning_rate": 2.4594960757172555e-05,
+      "loss": 0.4825,
+      "step": 1204
+    },
+    {
+      "epoch": 3.312714776632302,
+      "grad_norm": 0.18231300214622423,
+      "learning_rate": 2.4523972461258386e-05,
+      "loss": 0.4794,
+      "step": 1205
+    },
+    {
+      "epoch": 3.315463917525773,
+      "grad_norm": 0.25489199823717157,
+      "learning_rate": 2.445304144319792e-05,
+      "loss": 0.4795,
+      "step": 1206
+    },
+    {
+      "epoch": 3.3182130584192437,
+      "grad_norm": 0.15105215480831594,
+      "learning_rate": 2.4382167965511774e-05,
+      "loss": 0.4902,
+      "step": 1207
+    },
+    {
+      "epoch": 3.320962199312715,
+      "grad_norm": 0.2799009342897516,
+      "learning_rate": 2.431135229050763e-05,
+      "loss": 0.4818,
+      "step": 1208
+    },
+    {
+      "epoch": 3.3237113402061857,
+      "grad_norm": 0.21951158609310864,
+      "learning_rate": 2.4240594680279204e-05,
+      "loss": 0.479,
+      "step": 1209
+    },
+    {
+      "epoch": 3.3264604810996565,
+      "grad_norm": 0.19395691144843719,
+      "learning_rate": 2.416989539670536e-05,
+      "loss": 0.479,
+      "step": 1210
+    },
+    {
+      "epoch": 3.3292096219931273,
+      "grad_norm": 0.17094551246350398,
+      "learning_rate": 2.409925470144903e-05,
+      "loss": 0.4808,
+      "step": 1211
+    },
+    {
+      "epoch": 3.331958762886598,
+      "grad_norm": 0.23964152078624051,
+      "learning_rate": 2.402867285595635e-05,
+      "loss": 0.4764,
+      "step": 1212
+    },
+    {
+      "epoch": 3.334707903780069,
+      "grad_norm": 0.2782973993431327,
+      "learning_rate": 2.3958150121455638e-05,
+      "loss": 0.4811,
+      "step": 1213
+    },
+    {
+      "epoch": 3.3374570446735397,
+      "grad_norm": 0.1640795235166955,
+      "learning_rate": 2.3887686758956425e-05,
+      "loss": 0.4793,
+      "step": 1214
+    },
+    {
+      "epoch": 3.3402061855670104,
+      "grad_norm": 0.3070072205373487,
+      "learning_rate": 2.3817283029248536e-05,
+      "loss": 0.475,
+      "step": 1215
+    },
+    {
+      "epoch": 3.342955326460481,
+      "grad_norm": 0.23275102863915093,
+      "learning_rate": 2.3746939192901052e-05,
+      "loss": 0.479,
+      "step": 1216
+    },
+    {
+      "epoch": 3.345704467353952,
+      "grad_norm": 0.23028817088895479,
+      "learning_rate": 2.3676655510261402e-05,
+      "loss": 0.4735,
+      "step": 1217
+    },
+    {
+      "epoch": 3.3484536082474228,
+      "grad_norm": 0.320766094957754,
+      "learning_rate": 2.3606432241454398e-05,
+      "loss": 0.4877,
+      "step": 1218
+    },
+    {
+      "epoch": 3.3512027491408936,
+      "grad_norm": 0.2518366256607952,
+      "learning_rate": 2.353626964638122e-05,
+      "loss": 0.4733,
+      "step": 1219
+    },
+    {
+      "epoch": 3.3539518900343643,
+      "grad_norm": 0.1904395352124866,
+      "learning_rate": 2.34661679847185e-05,
+      "loss": 0.4895,
+      "step": 1220
+    },
+    {
+      "epoch": 3.356701030927835,
+      "grad_norm": 0.3279385370633572,
+      "learning_rate": 2.3396127515917392e-05,
+      "loss": 0.4777,
+      "step": 1221
+    },
+    {
+      "epoch": 3.359450171821306,
+      "grad_norm": 0.20836286167731985,
+      "learning_rate": 2.3326148499202518e-05,
+      "loss": 0.4765,
+      "step": 1222
+    },
+    {
+      "epoch": 3.3621993127147767,
+      "grad_norm": 0.22895782643501986,
+      "learning_rate": 2.325623119357107e-05,
+      "loss": 0.4883,
+      "step": 1223
+    },
+    {
+      "epoch": 3.3649484536082475,
+      "grad_norm": 0.29237657945287854,
+      "learning_rate": 2.3186375857791896e-05,
+      "loss": 0.483,
+      "step": 1224
+    },
+    {
+      "epoch": 3.3676975945017182,
+      "grad_norm": 0.18465319808299466,
+      "learning_rate": 2.3116582750404427e-05,
+      "loss": 0.472,
+      "step": 1225
+    },
+    {
+      "epoch": 3.370446735395189,
+      "grad_norm": 0.20550527495867124,
+      "learning_rate": 2.304685212971781e-05,
+      "loss": 0.477,
+      "step": 1226
+    },
+    {
+      "epoch": 3.37319587628866,
+      "grad_norm": 0.2692822007314835,
+      "learning_rate": 2.2977184253809913e-05,
+      "loss": 0.4763,
+      "step": 1227
+    },
+    {
+      "epoch": 3.3759450171821306,
+      "grad_norm": 0.1708056032464783,
+      "learning_rate": 2.2907579380526387e-05,
+      "loss": 0.48,
+      "step": 1228
+    },
+    {
+      "epoch": 3.3786941580756014,
+      "grad_norm": 0.368849523723143,
+      "learning_rate": 2.2838037767479713e-05,
+      "loss": 0.4797,
+      "step": 1229
+    },
+    {
+      "epoch": 3.381443298969072,
+      "grad_norm": 0.21646292002270615,
+      "learning_rate": 2.2768559672048218e-05,
+      "loss": 0.474,
+      "step": 1230
+    },
+    {
+      "epoch": 3.384192439862543,
+      "grad_norm": 0.26373109002232037,
+      "learning_rate": 2.2699145351375186e-05,
+      "loss": 0.4803,
+      "step": 1231
+    },
+    {
+      "epoch": 3.3869415807560137,
+      "grad_norm": 0.2843624854578763,
+      "learning_rate": 2.262979506236784e-05,
+      "loss": 0.4826,
+      "step": 1232
+    },
+    {
+      "epoch": 3.3896907216494845,
+      "grad_norm": 0.19833041217898748,
+      "learning_rate": 2.256050906169641e-05,
+      "loss": 0.491,
+      "step": 1233
+    },
+    {
+      "epoch": 3.3924398625429553,
+      "grad_norm": 0.2800244352558394,
+      "learning_rate": 2.2491287605793204e-05,
+      "loss": 0.4707,
+      "step": 1234
+    },
+    {
+      "epoch": 3.395189003436426,
+      "grad_norm": 0.18325778358692565,
+      "learning_rate": 2.2422130950851643e-05,
+      "loss": 0.4744,
+      "step": 1235
+    },
+    {
+      "epoch": 3.397938144329897,
+      "grad_norm": 0.3008410533335858,
+      "learning_rate": 2.235303935282531e-05,
+      "loss": 0.4846,
+      "step": 1236
+    },
+    {
+      "epoch": 3.4006872852233676,
+      "grad_norm": 0.19194203147930453,
+      "learning_rate": 2.2284013067427025e-05,
+      "loss": 0.4718,
+      "step": 1237
+    },
+    {
+      "epoch": 3.4034364261168384,
+      "grad_norm": 0.34898181416099955,
+      "learning_rate": 2.2215052350127834e-05,
+      "loss": 0.484,
+      "step": 1238
+    },
+    {
+      "epoch": 3.406185567010309,
+      "grad_norm": 0.2356425803363679,
+      "learning_rate": 2.214615745615619e-05,
+      "loss": 0.4774,
+      "step": 1239
+    },
+    {
+      "epoch": 3.40893470790378,
+      "grad_norm": 0.21466246650328097,
+      "learning_rate": 2.207732864049686e-05,
+      "loss": 0.4812,
+      "step": 1240
+    },
+    {
+      "epoch": 3.4116838487972507,
+      "grad_norm": 0.26377375192350677,
+      "learning_rate": 2.200856615789009e-05,
+      "loss": 0.4912,
+      "step": 1241
+    },
+    {
+      "epoch": 3.4144329896907215,
+      "grad_norm": 0.15695872082264978,
+      "learning_rate": 2.1939870262830577e-05,
+      "loss": 0.4715,
+      "step": 1242
+    },
+    {
+      "epoch": 3.4171821305841923,
+      "grad_norm": 0.28736994518322967,
+      "learning_rate": 2.187124120956666e-05,
+      "loss": 0.4811,
+      "step": 1243
+    },
+    {
+      "epoch": 3.419931271477663,
+      "grad_norm": 0.24847231178584678,
+      "learning_rate": 2.1802679252099184e-05,
+      "loss": 0.4784,
+      "step": 1244
+    },
+    {
+      "epoch": 3.422680412371134,
+      "grad_norm": 0.17957114598770363,
+      "learning_rate": 2.1734184644180715e-05,
+      "loss": 0.4706,
+      "step": 1245
+    },
+    {
+      "epoch": 3.4254295532646046,
+      "grad_norm": 0.3386000360336158,
+      "learning_rate": 2.16657576393146e-05,
+      "loss": 0.4764,
+      "step": 1246
+    },
+    {
+      "epoch": 3.4281786941580754,
+      "grad_norm": 0.14262617741372224,
+      "learning_rate": 2.1597398490753917e-05,
+      "loss": 0.4645,
+      "step": 1247
+    },
+    {
+      "epoch": 3.4309278350515466,
+      "grad_norm": 0.2371298714043362,
+      "learning_rate": 2.152910745150063e-05,
+      "loss": 0.4778,
+      "step": 1248
+    },
+    {
+      "epoch": 3.4336769759450174,
+      "grad_norm": 0.15007549750939145,
+      "learning_rate": 2.1460884774304614e-05,
+      "loss": 0.481,
+      "step": 1249
+    },
+    {
+      "epoch": 3.436426116838488,
+      "grad_norm": 0.2666157528473445,
+      "learning_rate": 2.1392730711662755e-05,
+      "loss": 0.473,
+      "step": 1250
+    },
+    {
+      "epoch": 3.439175257731959,
+      "grad_norm": 0.19370621015791017,
+      "learning_rate": 2.1324645515817988e-05,
+      "loss": 0.4916,
+      "step": 1251
+    },
+    {
+      "epoch": 3.4419243986254298,
+      "grad_norm": 0.1363626153767201,
+      "learning_rate": 2.125662943875832e-05,
+      "loss": 0.4783,
+      "step": 1252
+    },
+    {
+      "epoch": 3.4446735395189005,
+      "grad_norm": 0.1558927651589561,
+      "learning_rate": 2.1188682732215978e-05,
+      "loss": 0.4863,
+      "step": 1253
+    },
+    {
+      "epoch": 3.4474226804123713,
+      "grad_norm": 0.1878171719282548,
+      "learning_rate": 2.1120805647666484e-05,
+      "loss": 0.483,
+      "step": 1254
+    },
+    {
+      "epoch": 3.450171821305842,
+      "grad_norm": 0.1706433093556562,
+      "learning_rate": 2.1052998436327624e-05,
+      "loss": 0.4766,
+      "step": 1255
+    },
+    {
+      "epoch": 3.452920962199313,
+      "grad_norm": 0.14594078534672697,
+      "learning_rate": 2.0985261349158606e-05,
+      "loss": 0.4701,
+      "step": 1256
+    },
+    {
+      "epoch": 3.4556701030927837,
+      "grad_norm": 0.2713964023439289,
+      "learning_rate": 2.0917594636859084e-05,
+      "loss": 0.4734,
+      "step": 1257
+    },
+    {
+      "epoch": 3.4584192439862544,
+      "grad_norm": 0.20301631548067403,
+      "learning_rate": 2.084999854986829e-05,
+      "loss": 0.4664,
+      "step": 1258
+    },
+    {
+      "epoch": 3.4611683848797252,
+      "grad_norm": 0.21017329606796895,
+      "learning_rate": 2.0782473338364034e-05,
+      "loss": 0.4801,
+      "step": 1259
+    },
+    {
+      "epoch": 3.463917525773196,
+      "grad_norm": 0.2024174569645721,
+      "learning_rate": 2.0715019252261786e-05,
+      "loss": 0.473,
+      "step": 1260
+    },
+    {
+      "epoch": 3.466666666666667,
+      "grad_norm": 0.13513087820592787,
+      "learning_rate": 2.0647636541213843e-05,
+      "loss": 0.479,
+      "step": 1261
+    },
+    {
+      "epoch": 3.4694158075601376,
+      "grad_norm": 0.15323818419427804,
+      "learning_rate": 2.0580325454608294e-05,
+      "loss": 0.479,
+      "step": 1262
+    },
+    {
+      "epoch": 3.4721649484536083,
+      "grad_norm": 0.1883180981275938,
+      "learning_rate": 2.051308624156815e-05,
+      "loss": 0.4797,
+      "step": 1263
+    },
+    {
+      "epoch": 3.474914089347079,
+      "grad_norm": 0.18535688822536886,
+      "learning_rate": 2.0445919150950387e-05,
+      "loss": 0.4828,
+      "step": 1264
+    },
+    {
+      "epoch": 3.47766323024055,
+      "grad_norm": 0.1735738790195007,
+      "learning_rate": 2.0378824431345116e-05,
+      "loss": 0.4752,
+      "step": 1265
+    },
+    {
+      "epoch": 3.4804123711340207,
+      "grad_norm": 0.25380262888772,
+      "learning_rate": 2.0311802331074543e-05,
+      "loss": 0.479,
+      "step": 1266
+    },
+    {
+      "epoch": 3.4831615120274915,
+      "grad_norm": 0.17442121730064505,
+      "learning_rate": 2.024485309819213e-05,
+      "loss": 0.4739,
+      "step": 1267
+    },
+    {
+      "epoch": 3.4859106529209622,
+      "grad_norm": 0.284320944684907,
+      "learning_rate": 2.01779769804816e-05,
+      "loss": 0.4843,
+      "step": 1268
+    },
+    {
+      "epoch": 3.488659793814433,
+      "grad_norm": 0.1625892440951168,
+      "learning_rate": 2.011117422545616e-05,
+      "loss": 0.4763,
+      "step": 1269
+    },
+    {
+      "epoch": 3.491408934707904,
+      "grad_norm": 0.28275001646900066,
+      "learning_rate": 2.004444508035743e-05,
+      "loss": 0.4699,
+      "step": 1270
+    },
+    {
+      "epoch": 3.4941580756013746,
+      "grad_norm": 0.17127917891669825,
+      "learning_rate": 1.9977789792154615e-05,
+      "loss": 0.4882,
+      "step": 1271
+    },
+    {
+      "epoch": 3.4969072164948454,
+      "grad_norm": 0.155941741872769,
+      "learning_rate": 1.991120860754356e-05,
+      "loss": 0.4692,
+      "step": 1272
+    },
+    {
+      "epoch": 3.499656357388316,
+      "grad_norm": 0.15459720154148457,
+      "learning_rate": 1.984470177294588e-05,
+      "loss": 0.4841,
+      "step": 1273
+    },
+    {
+      "epoch": 3.502405498281787,
+      "grad_norm": 0.2196341305641022,
+      "learning_rate": 1.9778269534507987e-05,
+      "loss": 0.4713,
+      "step": 1274
+    },
+    {
+      "epoch": 3.5051546391752577,
+      "grad_norm": 0.22478563840945578,
+      "learning_rate": 1.971191213810021e-05,
+      "loss": 0.4892,
+      "step": 1275
+    },
+    {
+      "epoch": 3.5079037800687285,
+      "grad_norm": 0.1739449107806518,
+      "learning_rate": 1.964562982931588e-05,
+      "loss": 0.4765,
+      "step": 1276
+    },
+    {
+      "epoch": 3.5106529209621993,
+      "grad_norm": 0.27214067372219086,
+      "learning_rate": 1.9579422853470436e-05,
+      "loss": 0.4815,
+      "step": 1277
+    },
+    {
+      "epoch": 3.51340206185567,
+      "grad_norm": 0.17400788077272825,
+      "learning_rate": 1.95132914556005e-05,
+      "loss": 0.4727,
+      "step": 1278
+    },
+    {
+      "epoch": 3.516151202749141,
+      "grad_norm": 0.19337030209527964,
+      "learning_rate": 1.9447235880462947e-05,
+      "loss": 0.4804,
+      "step": 1279
+    },
+    {
+      "epoch": 3.5189003436426116,
+      "grad_norm": 0.1514665359437226,
+      "learning_rate": 1.938125637253409e-05,
+      "loss": 0.4729,
+      "step": 1280
+    },
+    {
+      "epoch": 3.5216494845360824,
+      "grad_norm": 0.27810047698132523,
+      "learning_rate": 1.9315353176008655e-05,
+      "loss": 0.478,
+      "step": 1281
+    },
+    {
+      "epoch": 3.524398625429553,
+      "grad_norm": 0.2052552676995858,
+      "learning_rate": 1.9249526534798965e-05,
+      "loss": 0.4825,
+      "step": 1282
+    },
+    {
+      "epoch": 3.527147766323024,
+      "grad_norm": 0.1638869338895521,
+      "learning_rate": 1.9183776692533977e-05,
+      "loss": 0.4754,
+      "step": 1283
+    },
+    {
+      "epoch": 3.5298969072164947,
+      "grad_norm": 0.1623552411012285,
+      "learning_rate": 1.9118103892558438e-05,
+      "loss": 0.4721,
+      "step": 1284
+    },
+    {
+      "epoch": 3.5326460481099655,
+      "grad_norm": 0.2765823979801074,
+      "learning_rate": 1.9052508377931945e-05,
+      "loss": 0.4721,
+      "step": 1285
+    },
+    {
+      "epoch": 3.5353951890034363,
+      "grad_norm": 0.20787425872060086,
+      "learning_rate": 1.8986990391428056e-05,
+      "loss": 0.4721,
+      "step": 1286
+    },
+    {
+      "epoch": 3.538144329896907,
+      "grad_norm": 0.23103945726058686,
+      "learning_rate": 1.8921550175533384e-05,
+      "loss": 0.4804,
+      "step": 1287
+    },
+    {
+      "epoch": 3.540893470790378,
+      "grad_norm": 0.14819667238986017,
+      "learning_rate": 1.885618797244674e-05,
+      "loss": 0.4781,
+      "step": 1288
+    },
+    {
+      "epoch": 3.5436426116838486,
+      "grad_norm": 0.34832469601339244,
+      "learning_rate": 1.8790904024078166e-05,
+      "loss": 0.4853,
+      "step": 1289
+    },
+    {
+      "epoch": 3.5463917525773194,
+      "grad_norm": 0.17762443199441277,
+      "learning_rate": 1.8725698572048096e-05,
+      "loss": 0.4687,
+      "step": 1290
+    },
+    {
+      "epoch": 3.54914089347079,
+      "grad_norm": 0.220110309410904,
+      "learning_rate": 1.866057185768644e-05,
+      "loss": 0.4722,
+      "step": 1291
+    },
+    {
+      "epoch": 3.551890034364261,
+      "grad_norm": 0.1568411202207363,
+      "learning_rate": 1.8595524122031695e-05,
+      "loss": 0.477,
+      "step": 1292
+    },
+    {
+      "epoch": 3.5546391752577318,
+      "grad_norm": 0.1487231862533628,
+      "learning_rate": 1.8530555605830048e-05,
+      "loss": 0.4812,
+      "step": 1293
+    },
+    {
+      "epoch": 3.5573883161512025,
+      "grad_norm": 0.27212850028975205,
+      "learning_rate": 1.8465666549534487e-05,
+      "loss": 0.4716,
+      "step": 1294
+    },
+    {
+      "epoch": 3.5601374570446733,
+      "grad_norm": 0.26360242336986534,
+      "learning_rate": 1.840085719330394e-05,
+      "loss": 0.4794,
+      "step": 1295
+    },
+    {
+      "epoch": 3.562886597938144,
+      "grad_norm": 0.20568964864290215,
+      "learning_rate": 1.8336127777002336e-05,
+      "loss": 0.4808,
+      "step": 1296
+    },
+    {
+      "epoch": 3.565635738831615,
+      "grad_norm": 0.282051987566169,
+      "learning_rate": 1.8271478540197735e-05,
+      "loss": 0.4761,
+      "step": 1297
+    },
+    {
+      "epoch": 3.5683848797250857,
+      "grad_norm": 0.16289176663769198,
+      "learning_rate": 1.8206909722161444e-05,
+      "loss": 0.4774,
+      "step": 1298
+    },
+    {
+      "epoch": 3.5711340206185564,
+      "grad_norm": 0.2575145343236498,
+      "learning_rate": 1.814242156186718e-05,
+      "loss": 0.4795,
+      "step": 1299
+    },
+    {
+      "epoch": 3.5738831615120272,
+      "grad_norm": 0.1486906599375562,
+      "learning_rate": 1.8078014297990104e-05,
+      "loss": 0.4824,
+      "step": 1300
+    },
+    {
+      "epoch": 3.576632302405498,
+      "grad_norm": 0.1501909622248065,
+      "learning_rate": 1.8013688168905946e-05,
+      "loss": 0.4804,
+      "step": 1301
+    },
+    {
+      "epoch": 3.5793814432989692,
+      "grad_norm": 0.15230793704069007,
+      "learning_rate": 1.7949443412690182e-05,
+      "loss": 0.4751,
+      "step": 1302
+    },
+    {
+      "epoch": 3.58213058419244,
+      "grad_norm": 0.16635368754820404,
+      "learning_rate": 1.788528026711715e-05,
+      "loss": 0.4806,
+      "step": 1303
+    },
+    {
+      "epoch": 3.584879725085911,
+      "grad_norm": 0.1347568536197528,
+      "learning_rate": 1.78211989696591e-05,
+      "loss": 0.4723,
+      "step": 1304
+    },
+    {
+      "epoch": 3.5876288659793816,
+      "grad_norm": 0.19251735027744715,
+      "learning_rate": 1.7757199757485363e-05,
+      "loss": 0.4796,
+      "step": 1305
+    },
+    {
+      "epoch": 3.5903780068728524,
+      "grad_norm": 0.14511887693354616,
+      "learning_rate": 1.7693282867461464e-05,
+      "loss": 0.4875,
+      "step": 1306
+    },
+    {
+      "epoch": 3.593127147766323,
+      "grad_norm": 0.21823490371477205,
+      "learning_rate": 1.762944853614828e-05,
+      "loss": 0.4831,
+      "step": 1307
+    },
+    {
+      "epoch": 3.595876288659794,
+      "grad_norm": 0.14789140619202265,
+      "learning_rate": 1.756569699980111e-05,
+      "loss": 0.4719,
+      "step": 1308
+    },
+    {
+      "epoch": 3.5986254295532647,
+      "grad_norm": 0.1537689543052464,
+      "learning_rate": 1.7502028494368777e-05,
+      "loss": 0.4775,
+      "step": 1309
+    },
+    {
+      "epoch": 3.6013745704467355,
+      "grad_norm": 0.16438166123180578,
+      "learning_rate": 1.743844325549289e-05,
+      "loss": 0.4778,
+      "step": 1310
+    },
+    {
+      "epoch": 3.6041237113402063,
+      "grad_norm": 0.1309598142497098,
+      "learning_rate": 1.737494151850682e-05,
+      "loss": 0.4799,
+      "step": 1311
+    },
+    {
+      "epoch": 3.606872852233677,
+      "grad_norm": 0.13501349366551338,
+      "learning_rate": 1.731152351843492e-05,
+      "loss": 0.4778,
+      "step": 1312
+    },
+    {
+      "epoch": 3.609621993127148,
+      "grad_norm": 0.12271151015542832,
+      "learning_rate": 1.7248189489991594e-05,
+      "loss": 0.4804,
+      "step": 1313
+    },
+    {
+      "epoch": 3.6123711340206186,
+      "grad_norm": 0.1579684841055174,
+      "learning_rate": 1.7184939667580512e-05,
+      "loss": 0.4756,
+      "step": 1314
+    },
+    {
+      "epoch": 3.6151202749140894,
+      "grad_norm": 0.11550758392122117,
+      "learning_rate": 1.712177428529365e-05,
+      "loss": 0.4813,
+      "step": 1315
+    },
+    {
+      "epoch": 3.61786941580756,
+      "grad_norm": 0.1446519336760662,
+      "learning_rate": 1.7058693576910495e-05,
+      "loss": 0.4761,
+      "step": 1316
+    },
+    {
+      "epoch": 3.620618556701031,
+      "grad_norm": 0.1387283733233433,
+      "learning_rate": 1.6995697775897097e-05,
+      "loss": 0.4855,
+      "step": 1317
+    },
+    {
+      "epoch": 3.6233676975945017,
+      "grad_norm": 0.12450900277305127,
+      "learning_rate": 1.6932787115405318e-05,
+      "loss": 0.4768,
+      "step": 1318
+    },
+    {
+      "epoch": 3.6261168384879725,
+      "grad_norm": 0.15118692423046173,
+      "learning_rate": 1.6869961828271892e-05,
+      "loss": 0.4795,
+      "step": 1319
+    },
+    {
+      "epoch": 3.6288659793814433,
+      "grad_norm": 0.1275953061475887,
+      "learning_rate": 1.6807222147017558e-05,
+      "loss": 0.4844,
+      "step": 1320
+    },
+    {
+      "epoch": 3.631615120274914,
+      "grad_norm": 0.12357741430968726,
+      "learning_rate": 1.6744568303846237e-05,
+      "loss": 0.4718,
+      "step": 1321
+    },
+    {
+      "epoch": 3.634364261168385,
+      "grad_norm": 0.13324292495631146,
+      "learning_rate": 1.6682000530644177e-05,
+      "loss": 0.4828,
+      "step": 1322
+    },
+    {
+      "epoch": 3.6371134020618556,
+      "grad_norm": 0.12727908902135235,
+      "learning_rate": 1.6619519058979044e-05,
+      "loss": 0.4731,
+      "step": 1323
+    },
+    {
+      "epoch": 3.6398625429553264,
+      "grad_norm": 0.14031695435202224,
+      "learning_rate": 1.6557124120099108e-05,
+      "loss": 0.4723,
+      "step": 1324
+    },
+    {
+      "epoch": 3.642611683848797,
+      "grad_norm": 0.1238822517218841,
+      "learning_rate": 1.6494815944932376e-05,
+      "loss": 0.4767,
+      "step": 1325
+    },
+    {
+      "epoch": 3.645360824742268,
+      "grad_norm": 0.14383655055562664,
+      "learning_rate": 1.6432594764085738e-05,
+      "loss": 0.4744,
+      "step": 1326
+    },
+    {
+      "epoch": 3.6481099656357387,
+      "grad_norm": 0.13074212391556805,
+      "learning_rate": 1.6370460807844096e-05,
+      "loss": 0.4719,
+      "step": 1327
+    },
+    {
+      "epoch": 3.6508591065292095,
+      "grad_norm": 0.1408486433749572,
+      "learning_rate": 1.6308414306169545e-05,
+      "loss": 0.4792,
+      "step": 1328
+    },
+    {
+      "epoch": 3.6536082474226803,
+      "grad_norm": 0.1425866228634816,
+      "learning_rate": 1.6246455488700522e-05,
+      "loss": 0.4759,
+      "step": 1329
+    },
+    {
+      "epoch": 3.656357388316151,
+      "grad_norm": 0.13452543614110554,
+      "learning_rate": 1.61845845847509e-05,
+      "loss": 0.4825,
+      "step": 1330
+    },
+    {
+      "epoch": 3.659106529209622,
+      "grad_norm": 0.123447733182089,
+      "learning_rate": 1.6122801823309195e-05,
+      "loss": 0.4839,
+      "step": 1331
+    },
+    {
+      "epoch": 3.6618556701030927,
+      "grad_norm": 0.141262084894595,
+      "learning_rate": 1.6061107433037707e-05,
+      "loss": 0.4803,
+      "step": 1332
+    },
+    {
+      "epoch": 3.6646048109965634,
+      "grad_norm": 0.16335769694535568,
+      "learning_rate": 1.5999501642271664e-05,
+      "loss": 0.491,
+      "step": 1333
+    },
+    {
+      "epoch": 3.667353951890034,
+      "grad_norm": 0.13731203188970648,
+      "learning_rate": 1.593798467901837e-05,
+      "loss": 0.4764,
+      "step": 1334
+    },
+    {
+      "epoch": 3.670103092783505,
+      "grad_norm": 0.14344242124627954,
+      "learning_rate": 1.5876556770956394e-05,
+      "loss": 0.4776,
+      "step": 1335
+    },
+    {
+      "epoch": 3.672852233676976,
+      "grad_norm": 0.14287518229033996,
+      "learning_rate": 1.581521814543467e-05,
+      "loss": 0.4729,
+      "step": 1336
+    },
+    {
+      "epoch": 3.675601374570447,
+      "grad_norm": 0.13919137334446374,
+      "learning_rate": 1.575396902947175e-05,
+      "loss": 0.4803,
+      "step": 1337
+    },
+    {
+      "epoch": 3.678350515463918,
+      "grad_norm": 0.1332949080422285,
+      "learning_rate": 1.5692809649754855e-05,
+      "loss": 0.4732,
+      "step": 1338
+    },
+    {
+      "epoch": 3.6810996563573886,
+      "grad_norm": 0.15251079258406947,
+      "learning_rate": 1.563174023263909e-05,
+      "loss": 0.4721,
+      "step": 1339
+    },
+    {
+      "epoch": 3.6838487972508593,
+      "grad_norm": 0.12884291771360323,
+      "learning_rate": 1.557076100414663e-05,
+      "loss": 0.4771,
+      "step": 1340
+    },
+    {
+      "epoch": 3.68659793814433,
+      "grad_norm": 0.14077765554999186,
+      "learning_rate": 1.5509872189965826e-05,
+      "loss": 0.4765,
+      "step": 1341
+    },
+    {
+      "epoch": 3.689347079037801,
+      "grad_norm": 0.12097139782888651,
+      "learning_rate": 1.5449074015450416e-05,
+      "loss": 0.4816,
+      "step": 1342
+    },
+    {
+      "epoch": 3.6920962199312717,
+      "grad_norm": 0.12394128179144817,
+      "learning_rate": 1.5388366705618656e-05,
+      "loss": 0.4719,
+      "step": 1343
+    },
+    {
+      "epoch": 3.6948453608247425,
+      "grad_norm": 0.12482467627547454,
+      "learning_rate": 1.5327750485152546e-05,
+      "loss": 0.4763,
+      "step": 1344
+    },
+    {
+      "epoch": 3.6975945017182132,
+      "grad_norm": 0.12282644196525853,
+      "learning_rate": 1.5267225578396923e-05,
+      "loss": 0.4773,
+      "step": 1345
+    },
+    {
+      "epoch": 3.700343642611684,
+      "grad_norm": 0.12030648328844293,
+      "learning_rate": 1.5206792209358665e-05,
+      "loss": 0.486,
+      "step": 1346
+    },
+    {
+      "epoch": 3.703092783505155,
+      "grad_norm": 0.1242904271715769,
+      "learning_rate": 1.5146450601705866e-05,
+      "loss": 0.4788,
+      "step": 1347
+    },
+    {
+      "epoch": 3.7058419243986256,
+      "grad_norm": 0.12938487720592814,
+      "learning_rate": 1.5086200978767051e-05,
+      "loss": 0.482,
+      "step": 1348
+    },
+    {
+      "epoch": 3.7085910652920964,
+      "grad_norm": 0.1181010757249468,
+      "learning_rate": 1.5026043563530213e-05,
+      "loss": 0.4817,
+      "step": 1349
+    },
+    {
+      "epoch": 3.711340206185567,
+      "grad_norm": 0.12542244186954696,
+      "learning_rate": 1.4965978578642152e-05,
+      "loss": 0.4829,
+      "step": 1350
+    },
+    {
+      "epoch": 3.714089347079038,
+      "grad_norm": 0.1301001364291388,
+      "learning_rate": 1.4906006246407531e-05,
+      "loss": 0.4801,
+      "step": 1351
+    },
+    {
+      "epoch": 3.7168384879725087,
+      "grad_norm": 0.11752586529006614,
+      "learning_rate": 1.4846126788788153e-05,
+      "loss": 0.4824,
+      "step": 1352
+    },
+    {
+      "epoch": 3.7195876288659795,
+      "grad_norm": 0.1250323838865372,
+      "learning_rate": 1.4786340427402034e-05,
+      "loss": 0.4735,
+      "step": 1353
+    },
+    {
+      "epoch": 3.7223367697594503,
+      "grad_norm": 0.11762297628012565,
+      "learning_rate": 1.472664738352266e-05,
+      "loss": 0.4738,
+      "step": 1354
+    },
+    {
+      "epoch": 3.725085910652921,
+      "grad_norm": 0.1295998839343961,
+      "learning_rate": 1.4667047878078115e-05,
+      "loss": 0.4797,
+      "step": 1355
+    },
+    {
+      "epoch": 3.727835051546392,
+      "grad_norm": 0.12318111875882015,
+      "learning_rate": 1.460754213165036e-05,
+      "loss": 0.4749,
+      "step": 1356
+    },
+    {
+      "epoch": 3.7305841924398626,
+      "grad_norm": 0.1372696026783743,
+      "learning_rate": 1.4548130364474253e-05,
+      "loss": 0.4828,
+      "step": 1357
+    },
+    {
+      "epoch": 3.7333333333333334,
+      "grad_norm": 0.11717512263035454,
+      "learning_rate": 1.4488812796436871e-05,
+      "loss": 0.4672,
+      "step": 1358
+    },
+    {
+      "epoch": 3.736082474226804,
+      "grad_norm": 0.13561308791703292,
+      "learning_rate": 1.4429589647076693e-05,
+      "loss": 0.4786,
+      "step": 1359
+    },
+    {
+      "epoch": 3.738831615120275,
+      "grad_norm": 0.12422009286706612,
+      "learning_rate": 1.4370461135582691e-05,
+      "loss": 0.4851,
+      "step": 1360
+    },
+    {
+      "epoch": 3.7415807560137457,
+      "grad_norm": 0.11832155403134417,
+      "learning_rate": 1.4311427480793597e-05,
+      "loss": 0.4766,
+      "step": 1361
+    },
+    {
+      "epoch": 3.7443298969072165,
+      "grad_norm": 0.12867243461981434,
+      "learning_rate": 1.4252488901197064e-05,
+      "loss": 0.4724,
+      "step": 1362
+    },
+    {
+      "epoch": 3.7470790378006873,
+      "grad_norm": 0.1313834080728087,
+      "learning_rate": 1.4193645614928886e-05,
+      "loss": 0.4714,
+      "step": 1363
+    },
+    {
+      "epoch": 3.749828178694158,
+      "grad_norm": 0.129183260330846,
+      "learning_rate": 1.4134897839772155e-05,
+      "loss": 0.4809,
+      "step": 1364
+    },
+    {
+      "epoch": 3.752577319587629,
+      "grad_norm": 0.12556934848044315,
+      "learning_rate": 1.4076245793156477e-05,
+      "loss": 0.4769,
+      "step": 1365
+    },
+    {
+      "epoch": 3.7553264604810996,
+      "grad_norm": 0.12147447097394133,
+      "learning_rate": 1.4017689692157114e-05,
+      "loss": 0.4821,
+      "step": 1366
+    },
+    {
+      "epoch": 3.7580756013745704,
+      "grad_norm": 0.13051373867173985,
+      "learning_rate": 1.3959229753494312e-05,
+      "loss": 0.472,
+      "step": 1367
+    },
+    {
+      "epoch": 3.760824742268041,
+      "grad_norm": 0.1294424563340561,
+      "learning_rate": 1.3900866193532347e-05,
+      "loss": 0.4822,
+      "step": 1368
+    },
+    {
+      "epoch": 3.763573883161512,
+      "grad_norm": 0.1319274091460674,
+      "learning_rate": 1.3842599228278819e-05,
+      "loss": 0.4808,
+      "step": 1369
+    },
+    {
+      "epoch": 3.7663230240549828,
+      "grad_norm": 0.14039131963053228,
+      "learning_rate": 1.3784429073383794e-05,
+      "loss": 0.4878,
+      "step": 1370
+    },
+    {
+      "epoch": 3.7690721649484535,
+      "grad_norm": 0.15726224881254797,
+      "learning_rate": 1.3726355944139091e-05,
+      "loss": 0.4858,
+      "step": 1371
+    },
+    {
+      "epoch": 3.7718213058419243,
+      "grad_norm": 0.1414080124007642,
+      "learning_rate": 1.3668380055477383e-05,
+      "loss": 0.4843,
+      "step": 1372
+    },
+    {
+      "epoch": 3.774570446735395,
+      "grad_norm": 0.14834309789500114,
+      "learning_rate": 1.3610501621971466e-05,
+      "loss": 0.4817,
+      "step": 1373
+    },
+    {
+      "epoch": 3.777319587628866,
+      "grad_norm": 0.17635910401341698,
+      "learning_rate": 1.3552720857833449e-05,
+      "loss": 0.4791,
+      "step": 1374
+    },
+    {
+      "epoch": 3.7800687285223367,
+      "grad_norm": 0.12924512390652784,
+      "learning_rate": 1.3495037976913947e-05,
+      "loss": 0.4786,
+      "step": 1375
+    },
+    {
+      "epoch": 3.7828178694158074,
+      "grad_norm": 0.15614756263557175,
+      "learning_rate": 1.3437453192701315e-05,
+      "loss": 0.4805,
+      "step": 1376
+    },
+    {
+      "epoch": 3.7855670103092782,
+      "grad_norm": 0.13352957029862794,
+      "learning_rate": 1.3379966718320839e-05,
+      "loss": 0.476,
+      "step": 1377
+    },
+    {
+      "epoch": 3.788316151202749,
+      "grad_norm": 0.14136665810120994,
+      "learning_rate": 1.3322578766533968e-05,
+      "loss": 0.473,
+      "step": 1378
+    },
+    {
+      "epoch": 3.79106529209622,
+      "grad_norm": 0.1527290850046059,
+      "learning_rate": 1.3265289549737501e-05,
+      "loss": 0.4826,
+      "step": 1379
+    },
+    {
+      "epoch": 3.7938144329896906,
+      "grad_norm": 0.13142189843945631,
+      "learning_rate": 1.32080992799628e-05,
+      "loss": 0.4807,
+      "step": 1380
+    },
+    {
+      "epoch": 3.7965635738831613,
+      "grad_norm": 0.13227897526677368,
+      "learning_rate": 1.3151008168875037e-05,
+      "loss": 0.47,
+      "step": 1381
+    },
+    {
+      "epoch": 3.799312714776632,
+      "grad_norm": 0.1292721449879669,
+      "learning_rate": 1.309401642777238e-05,
+      "loss": 0.4747,
+      "step": 1382
+    },
+    {
+      "epoch": 3.802061855670103,
+      "grad_norm": 0.11592891333671732,
+      "learning_rate": 1.3037124267585228e-05,
+      "loss": 0.4795,
+      "step": 1383
+    },
+    {
+      "epoch": 3.8048109965635737,
+      "grad_norm": 0.12251003727405267,
+      "learning_rate": 1.298033189887541e-05,
+      "loss": 0.4841,
+      "step": 1384
+    },
+    {
+      "epoch": 3.8075601374570445,
+      "grad_norm": 0.11989246216482814,
+      "learning_rate": 1.2923639531835433e-05,
+      "loss": 0.4688,
+      "step": 1385
+    },
+    {
+      "epoch": 3.8103092783505152,
+      "grad_norm": 0.12756974223833703,
+      "learning_rate": 1.28670473762877e-05,
+      "loss": 0.4746,
+      "step": 1386
+    },
+    {
+      "epoch": 3.813058419243986,
+      "grad_norm": 0.1180466527779192,
+      "learning_rate": 1.2810555641683706e-05,
+      "loss": 0.4779,
+      "step": 1387
+    },
+    {
+      "epoch": 3.815807560137457,
+      "grad_norm": 0.11529174071752626,
+      "learning_rate": 1.2754164537103289e-05,
+      "loss": 0.4721,
+      "step": 1388
+    },
+    {
+      "epoch": 3.8185567010309276,
+      "grad_norm": 0.11876628327578556,
+      "learning_rate": 1.2697874271253844e-05,
+      "loss": 0.4719,
+      "step": 1389
+    },
+    {
+      "epoch": 3.8213058419243984,
+      "grad_norm": 0.1212315987997364,
+      "learning_rate": 1.2641685052469556e-05,
+      "loss": 0.4901,
+      "step": 1390
+    },
+    {
+      "epoch": 3.824054982817869,
+      "grad_norm": 0.11790905831570075,
+      "learning_rate": 1.2585597088710637e-05,
+      "loss": 0.4755,
+      "step": 1391
+    },
+    {
+      "epoch": 3.82680412371134,
+      "grad_norm": 0.11385644404016117,
+      "learning_rate": 1.252961058756252e-05,
+      "loss": 0.4844,
+      "step": 1392
+    },
+    {
+      "epoch": 3.829553264604811,
+      "grad_norm": 0.11530386738039233,
+      "learning_rate": 1.2473725756235165e-05,
+      "loss": 0.4756,
+      "step": 1393
+    },
+    {
+      "epoch": 3.832302405498282,
+      "grad_norm": 0.12379904902641005,
+      "learning_rate": 1.2417942801562201e-05,
+      "loss": 0.4702,
+      "step": 1394
+    },
+    {
+      "epoch": 3.8350515463917527,
+      "grad_norm": 0.12225777044844523,
+      "learning_rate": 1.236226193000023e-05,
+      "loss": 0.4734,
+      "step": 1395
+    },
+    {
+      "epoch": 3.8378006872852235,
+      "grad_norm": 0.12161313208268175,
+      "learning_rate": 1.2306683347628012e-05,
+      "loss": 0.4809,
+      "step": 1396
+    },
+    {
+      "epoch": 3.8405498281786943,
+      "grad_norm": 0.13058946315743358,
+      "learning_rate": 1.2251207260145775e-05,
+      "loss": 0.4787,
+      "step": 1397
+    },
+    {
+      "epoch": 3.843298969072165,
+      "grad_norm": 0.11523497518057338,
+      "learning_rate": 1.2195833872874343e-05,
+      "loss": 0.4723,
+      "step": 1398
+    },
+    {
+      "epoch": 3.846048109965636,
+      "grad_norm": 0.13253286053473112,
+      "learning_rate": 1.2140563390754485e-05,
+      "loss": 0.4765,
+      "step": 1399
+    },
+    {
+      "epoch": 3.8487972508591066,
+      "grad_norm": 0.12147651056509255,
+      "learning_rate": 1.208539601834608e-05,
+      "loss": 0.4715,
+      "step": 1400
+    },
+    {
+      "epoch": 3.8515463917525774,
+      "grad_norm": 0.12552379756828472,
+      "learning_rate": 1.203033195982744e-05,
+      "loss": 0.4785,
+      "step": 1401
+    },
+    {
+      "epoch": 3.854295532646048,
+      "grad_norm": 0.12156559670865116,
+      "learning_rate": 1.1975371418994457e-05,
+      "loss": 0.4865,
+      "step": 1402
+    },
+    {
+      "epoch": 3.857044673539519,
+      "grad_norm": 0.13699296116050447,
+      "learning_rate": 1.1920514599259908e-05,
+      "loss": 0.4754,
+      "step": 1403
+    },
+    {
+      "epoch": 3.8597938144329897,
+      "grad_norm": 0.14619958749768325,
+      "learning_rate": 1.1865761703652691e-05,
+      "loss": 0.4896,
+      "step": 1404
+    },
+    {
+      "epoch": 3.8625429553264605,
+      "grad_norm": 0.11322022044471265,
+      "learning_rate": 1.1811112934817115e-05,
+      "loss": 0.4809,
+      "step": 1405
+    },
+    {
+      "epoch": 3.8652920962199313,
+      "grad_norm": 0.12792071165635296,
+      "learning_rate": 1.1756568495012032e-05,
+      "loss": 0.4777,
+      "step": 1406
+    },
+    {
+      "epoch": 3.868041237113402,
+      "grad_norm": 0.12840707488542488,
+      "learning_rate": 1.1702128586110204e-05,
+      "loss": 0.4704,
+      "step": 1407
+    },
+    {
+      "epoch": 3.870790378006873,
+      "grad_norm": 0.12236846928122713,
+      "learning_rate": 1.164779340959755e-05,
+      "loss": 0.4734,
+      "step": 1408
+    },
+    {
+      "epoch": 3.8735395189003436,
+      "grad_norm": 0.12828972122388524,
+      "learning_rate": 1.1593563166572315e-05,
+      "loss": 0.4728,
+      "step": 1409
+    },
+    {
+      "epoch": 3.8762886597938144,
+      "grad_norm": 0.11963041349113263,
+      "learning_rate": 1.1539438057744405e-05,
+      "loss": 0.4798,
+      "step": 1410
+    },
+    {
+      "epoch": 3.879037800687285,
+      "grad_norm": 0.11889755394099005,
+      "learning_rate": 1.1485418283434591e-05,
+      "loss": 0.4765,
+      "step": 1411
+    },
+    {
+      "epoch": 3.881786941580756,
+      "grad_norm": 0.1220924766559957,
+      "learning_rate": 1.1431504043573845e-05,
+      "loss": 0.4792,
+      "step": 1412
+    },
+    {
+      "epoch": 3.8845360824742268,
+      "grad_norm": 0.10991608717200368,
+      "learning_rate": 1.1377695537702506e-05,
+      "loss": 0.4758,
+      "step": 1413
+    },
+    {
+      "epoch": 3.8872852233676976,
+      "grad_norm": 0.1185758525402704,
+      "learning_rate": 1.1323992964969568e-05,
+      "loss": 0.4764,
+      "step": 1414
+    },
+    {
+      "epoch": 3.8900343642611683,
+      "grad_norm": 0.1285483122063387,
+      "learning_rate": 1.1270396524132016e-05,
+      "loss": 0.4726,
+      "step": 1415
+    },
+    {
+      "epoch": 3.892783505154639,
+      "grad_norm": 0.11257438797691342,
+      "learning_rate": 1.1216906413553996e-05,
+      "loss": 0.4853,
+      "step": 1416
+    },
+    {
+      "epoch": 3.89553264604811,
+      "grad_norm": 0.1265940578758485,
+      "learning_rate": 1.1163522831206119e-05,
+      "loss": 0.4853,
+      "step": 1417
+    },
+    {
+      "epoch": 3.8982817869415807,
+      "grad_norm": 0.12742498500479152,
+      "learning_rate": 1.1110245974664741e-05,
+      "loss": 0.4848,
+      "step": 1418
+    },
+    {
+      "epoch": 3.9010309278350515,
+      "grad_norm": 0.11699707249742979,
+      "learning_rate": 1.1057076041111197e-05,
+      "loss": 0.4728,
+      "step": 1419
+    },
+    {
+      "epoch": 3.9037800687285222,
+      "grad_norm": 0.12027686818075006,
+      "learning_rate": 1.100401322733113e-05,
+      "loss": 0.4697,
+      "step": 1420
+    },
+    {
+      "epoch": 3.906529209621993,
+      "grad_norm": 0.12028837431785216,
+      "learning_rate": 1.0951057729713698e-05,
+      "loss": 0.4735,
+      "step": 1421
+    },
+    {
+      "epoch": 3.909278350515464,
+      "grad_norm": 0.11366367375993695,
+      "learning_rate": 1.0898209744250852e-05,
+      "loss": 0.4798,
+      "step": 1422
+    },
+    {
+      "epoch": 3.9120274914089346,
+      "grad_norm": 0.12271108404053097,
+      "learning_rate": 1.0845469466536689e-05,
+      "loss": 0.4767,
+      "step": 1423
+    },
+    {
+      "epoch": 3.9147766323024054,
+      "grad_norm": 0.11169948199394023,
+      "learning_rate": 1.0792837091766635e-05,
+      "loss": 0.4773,
+      "step": 1424
+    },
+    {
+      "epoch": 3.917525773195876,
+      "grad_norm": 0.12185308524300838,
+      "learning_rate": 1.0740312814736766e-05,
+      "loss": 0.481,
+      "step": 1425
+    },
+    {
+      "epoch": 3.920274914089347,
+      "grad_norm": 0.11498767957292035,
+      "learning_rate": 1.0687896829843068e-05,
+      "loss": 0.4815,
+      "step": 1426
+    },
+    {
+      "epoch": 3.923024054982818,
+      "grad_norm": 0.10760167455092437,
+      "learning_rate": 1.0635589331080771e-05,
+      "loss": 0.4716,
+      "step": 1427
+    },
+    {
+      "epoch": 3.925773195876289,
+      "grad_norm": 0.11652282613458785,
+      "learning_rate": 1.0583390512043557e-05,
+      "loss": 0.4716,
+      "step": 1428
+    },
+    {
+      "epoch": 3.9285223367697597,
+      "grad_norm": 0.11095078761860352,
+      "learning_rate": 1.0531300565922873e-05,
+      "loss": 0.4773,
+      "step": 1429
+    },
+    {
+      "epoch": 3.9312714776632305,
+      "grad_norm": 0.10567263314079245,
+      "learning_rate": 1.0479319685507234e-05,
+      "loss": 0.4818,
+      "step": 1430
+    },
+    {
+      "epoch": 3.9340206185567013,
+      "grad_norm": 0.11290338751098791,
+      "learning_rate": 1.04274480631815e-05,
+      "loss": 0.4736,
+      "step": 1431
+    },
+    {
+      "epoch": 3.936769759450172,
+      "grad_norm": 0.11142975595031625,
+      "learning_rate": 1.0375685890926142e-05,
+      "loss": 0.4759,
+      "step": 1432
+    },
+    {
+      "epoch": 3.939518900343643,
+      "grad_norm": 0.12580032061681007,
+      "learning_rate": 1.0324033360316559e-05,
+      "loss": 0.4787,
+      "step": 1433
+    },
+    {
+      "epoch": 3.9422680412371136,
+      "grad_norm": 0.10935541303336552,
+      "learning_rate": 1.0272490662522347e-05,
+      "loss": 0.4771,
+      "step": 1434
+    },
+    {
+      "epoch": 3.9450171821305844,
+      "grad_norm": 0.11261196339303021,
+      "learning_rate": 1.0221057988306637e-05,
+      "loss": 0.4711,
+      "step": 1435
+    },
+    {
+      "epoch": 3.947766323024055,
+      "grad_norm": 0.11321244000910012,
+      "learning_rate": 1.0169735528025319e-05,
+      "loss": 0.4851,
+      "step": 1436
+    },
+    {
+      "epoch": 3.950515463917526,
+      "grad_norm": 0.11089992172259956,
+      "learning_rate": 1.011852347162639e-05,
+      "loss": 0.4751,
+      "step": 1437
+    },
+    {
+      "epoch": 3.9532646048109967,
+      "grad_norm": 0.11766021872483373,
+      "learning_rate": 1.0067422008649221e-05,
+      "loss": 0.4797,
+      "step": 1438
+    },
+    {
+      "epoch": 3.9560137457044675,
+      "grad_norm": 0.10628228690278965,
+      "learning_rate": 1.0016431328223887e-05,
+      "loss": 0.4782,
+      "step": 1439
+    },
+    {
+      "epoch": 3.9587628865979383,
+      "grad_norm": 0.11378648141949942,
+      "learning_rate": 9.96555161907044e-06,
+      "loss": 0.4766,
+      "step": 1440
+    },
+    {
+      "epoch": 3.961512027491409,
+      "grad_norm": 0.11864702105913649,
+      "learning_rate": 9.91478306949821e-06,
+      "loss": 0.4786,
+      "step": 1441
+    },
+    {
+      "epoch": 3.96426116838488,
+      "grad_norm": 0.12126170054240307,
+      "learning_rate": 9.864125867405146e-06,
+      "loss": 0.4797,
+      "step": 1442
+    },
+    {
+      "epoch": 3.9670103092783506,
+      "grad_norm": 0.11758636840862939,
+      "learning_rate": 9.813580200277063e-06,
+      "loss": 0.4768,
+      "step": 1443
+    },
+    {
+      "epoch": 3.9697594501718214,
+      "grad_norm": 0.11275441649847655,
+      "learning_rate": 9.76314625518699e-06,
+      "loss": 0.4668,
+      "step": 1444
+    },
+    {
+      "epoch": 3.972508591065292,
+      "grad_norm": 0.1131262643697182,
+      "learning_rate": 9.712824218794457e-06,
+      "loss": 0.4733,
+      "step": 1445
+    },
+    {
+      "epoch": 3.975257731958763,
+      "grad_norm": 0.10279589610315747,
+      "learning_rate": 9.662614277344846e-06,
+      "loss": 0.479,
+      "step": 1446
+    },
+    {
+      "epoch": 3.9780068728522338,
+      "grad_norm": 0.11106926427912463,
+      "learning_rate": 9.612516616668612e-06,
+      "loss": 0.4822,
+      "step": 1447
+    },
+    {
+      "epoch": 3.9807560137457045,
+      "grad_norm": 0.10564345176951702,
+      "learning_rate": 9.562531422180674e-06,
+      "loss": 0.4745,
+      "step": 1448
+    },
+    {
+      "epoch": 3.9835051546391753,
+      "grad_norm": 0.10829788647274623,
+      "learning_rate": 9.512658878879737e-06,
+      "loss": 0.4846,
+      "step": 1449
+    },
+    {
+      "epoch": 3.986254295532646,
+      "grad_norm": 0.11018495327771764,
+      "learning_rate": 9.462899171347538e-06,
+      "loss": 0.4802,
+      "step": 1450
+    },
+    {
+      "epoch": 3.989003436426117,
+      "grad_norm": 0.12608551718647237,
+      "learning_rate": 9.413252483748216e-06,
+      "loss": 0.4766,
+      "step": 1451
+    },
+    {
+      "epoch": 3.9917525773195877,
+      "grad_norm": 0.11337813442101644,
+      "learning_rate": 9.363718999827612e-06,
+      "loss": 0.479,
+      "step": 1452
+    },
+    {
+      "epoch": 3.9945017182130584,
+      "grad_norm": 0.12372671414303754,
+      "learning_rate": 9.314298902912583e-06,
+      "loss": 0.479,
+      "step": 1453
+    },
+    {
+      "epoch": 3.997250859106529,
+      "grad_norm": 0.11011001077642653,
+      "learning_rate": 9.26499237591037e-06,
+      "loss": 0.4695,
+      "step": 1454
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 0.14840181176358444,
+      "learning_rate": 9.215799601307828e-06,
+      "loss": 0.6096,
+      "step": 1455
+    },
+    {
+      "epoch": 4.002749140893471,
+      "grad_norm": 0.19421057731547003,
+      "learning_rate": 9.16672076117084e-06,
+      "loss": 0.4433,
+      "step": 1456
+    },
+    {
+      "epoch": 4.005498281786942,
+      "grad_norm": 0.15015912458886393,
+      "learning_rate": 9.117756037143622e-06,
+      "loss": 0.4425,
+      "step": 1457
+    },
+    {
+      "epoch": 4.008247422680412,
+      "grad_norm": 0.1383285525044493,
+      "learning_rate": 9.068905610448011e-06,
+      "loss": 0.4562,
+      "step": 1458
+    },
+    {
+      "epoch": 4.010996563573883,
+      "grad_norm": 0.15430721574490833,
+      "learning_rate": 9.02016966188283e-06,
+      "loss": 0.4484,
+      "step": 1459
+    },
+    {
+      "epoch": 4.013745704467354,
+      "grad_norm": 0.14905855161830828,
+      "learning_rate": 8.971548371823205e-06,
+      "loss": 0.446,
+      "step": 1460
+    },
+    {
+      "epoch": 4.016494845360825,
+      "grad_norm": 0.15470678416362293,
+      "learning_rate": 8.923041920219927e-06,
+      "loss": 0.4412,
+      "step": 1461
+    },
+    {
+      "epoch": 4.0192439862542955,
+      "grad_norm": 0.14564788987737176,
+      "learning_rate": 8.87465048659875e-06,
+      "loss": 0.4393,
+      "step": 1462
+    },
+    {
+      "epoch": 4.021993127147766,
+      "grad_norm": 0.13342656861373697,
+      "learning_rate": 8.826374250059704e-06,
+      "loss": 0.4498,
+      "step": 1463
+    },
+    {
+      "epoch": 4.024742268041237,
+      "grad_norm": 0.1302414374832906,
+      "learning_rate": 8.778213389276523e-06,
+      "loss": 0.4383,
+      "step": 1464
+    },
+    {
+      "epoch": 4.027491408934708,
+      "grad_norm": 0.14007280746183526,
+      "learning_rate": 8.730168082495885e-06,
+      "loss": 0.4474,
+      "step": 1465
+    },
+    {
+      "epoch": 4.030240549828179,
+      "grad_norm": 0.13171480431395038,
+      "learning_rate": 8.682238507536823e-06,
+      "loss": 0.4388,
+      "step": 1466
+    },
+    {
+      "epoch": 4.032989690721649,
+      "grad_norm": 0.12287866481381172,
+      "learning_rate": 8.634424841790014e-06,
+      "loss": 0.4384,
+      "step": 1467
+    },
+    {
+      "epoch": 4.03573883161512,
+      "grad_norm": 0.12293190753633977,
+      "learning_rate": 8.586727262217156e-06,
+      "loss": 0.4444,
+      "step": 1468
+    },
+    {
+      "epoch": 4.038487972508591,
+      "grad_norm": 0.12156568952497382,
+      "learning_rate": 8.539145945350324e-06,
+      "loss": 0.4393,
+      "step": 1469
+    },
+    {
+      "epoch": 4.041237113402062,
+      "grad_norm": 0.13520454469985085,
+      "learning_rate": 8.491681067291279e-06,
+      "loss": 0.4438,
+      "step": 1470
+    },
+    {
+      "epoch": 4.0439862542955325,
+      "grad_norm": 0.12818857279590956,
+      "learning_rate": 8.444332803710806e-06,
+      "loss": 0.4443,
+      "step": 1471
+    },
+    {
+      "epoch": 4.046735395189003,
+      "grad_norm": 0.11732641042539134,
+      "learning_rate": 8.397101329848146e-06,
+      "loss": 0.4443,
+      "step": 1472
+    },
+    {
+      "epoch": 4.049484536082474,
+      "grad_norm": 0.12736631288786282,
+      "learning_rate": 8.349986820510257e-06,
+      "loss": 0.4427,
+      "step": 1473
+    },
+    {
+      "epoch": 4.052233676975945,
+      "grad_norm": 0.1304818919917951,
+      "learning_rate": 8.302989450071219e-06,
+      "loss": 0.4478,
+      "step": 1474
+    },
+    {
+      "epoch": 4.054982817869416,
+      "grad_norm": 0.13641319860640636,
+      "learning_rate": 8.256109392471549e-06,
+      "loss": 0.4595,
+      "step": 1475
+    },
+    {
+      "epoch": 4.057731958762886,
+      "grad_norm": 0.11825978977034925,
+      "learning_rate": 8.20934682121763e-06,
+      "loss": 0.4461,
+      "step": 1476
+    },
+    {
+      "epoch": 4.060481099656357,
+      "grad_norm": 0.13030109067712098,
+      "learning_rate": 8.162701909380967e-06,
+      "loss": 0.4473,
+      "step": 1477
+    },
+    {
+      "epoch": 4.063230240549828,
+      "grad_norm": 0.12977389788164925,
+      "learning_rate": 8.116174829597625e-06,
+      "loss": 0.4506,
+      "step": 1478
+    },
+    {
+      "epoch": 4.065979381443299,
+      "grad_norm": 0.12148360180077224,
+      "learning_rate": 8.069765754067562e-06,
+      "loss": 0.4454,
+      "step": 1479
+    },
+    {
+      "epoch": 4.0687285223367695,
+      "grad_norm": 0.13076375704534893,
+      "learning_rate": 8.023474854553996e-06,
+      "loss": 0.4459,
+      "step": 1480
+    },
+    {
+      "epoch": 4.07147766323024,
+      "grad_norm": 0.13403343398060516,
+      "learning_rate": 7.977302302382757e-06,
+      "loss": 0.4392,
+      "step": 1481
+    },
+    {
+      "epoch": 4.074226804123711,
+      "grad_norm": 0.11166777845488875,
+      "learning_rate": 7.93124826844167e-06,
+      "loss": 0.4411,
+      "step": 1482
+    },
+    {
+      "epoch": 4.076975945017182,
+      "grad_norm": 0.13721899419089614,
+      "learning_rate": 7.885312923179912e-06,
+      "loss": 0.444,
+      "step": 1483
+    },
+    {
+      "epoch": 4.079725085910653,
+      "grad_norm": 0.13263433098857536,
+      "learning_rate": 7.839496436607406e-06,
+      "loss": 0.4459,
+      "step": 1484
+    },
+    {
+      "epoch": 4.082474226804123,
+      "grad_norm": 0.11590942502418243,
+      "learning_rate": 7.793798978294144e-06,
+      "loss": 0.4453,
+      "step": 1485
+    },
+    {
+      "epoch": 4.085223367697594,
+      "grad_norm": 0.13293144505263688,
+      "learning_rate": 7.748220717369594e-06,
+      "loss": 0.4413,
+      "step": 1486
+    },
+    {
+      "epoch": 4.087972508591065,
+      "grad_norm": 0.13483699801148805,
+      "learning_rate": 7.702761822522072e-06,
+      "loss": 0.4366,
+      "step": 1487
+    },
+    {
+      "epoch": 4.090721649484536,
+      "grad_norm": 0.11307879927149218,
+      "learning_rate": 7.6574224619981e-06,
+      "loss": 0.4352,
+      "step": 1488
+    },
+    {
+      "epoch": 4.0934707903780065,
+      "grad_norm": 0.11327258461306543,
+      "learning_rate": 7.612202803601812e-06,
+      "loss": 0.4513,
+      "step": 1489
+    },
+    {
+      "epoch": 4.096219931271477,
+      "grad_norm": 0.11998728903473971,
+      "learning_rate": 7.56710301469429e-06,
+      "loss": 0.4462,
+      "step": 1490
+    },
+    {
+      "epoch": 4.098969072164948,
+      "grad_norm": 0.11148518956657245,
+      "learning_rate": 7.522123262193011e-06,
+      "loss": 0.4414,
+      "step": 1491
+    },
+    {
+      "epoch": 4.101718213058419,
+      "grad_norm": 0.10980782127466164,
+      "learning_rate": 7.477263712571154e-06,
+      "loss": 0.4505,
+      "step": 1492
+    },
+    {
+      "epoch": 4.10446735395189,
+      "grad_norm": 0.13007209291051053,
+      "learning_rate": 7.432524531857037e-06,
+      "loss": 0.4404,
+      "step": 1493
+    },
+    {
+      "epoch": 4.10721649484536,
+      "grad_norm": 0.10633901487852752,
+      "learning_rate": 7.38790588563346e-06,
+      "loss": 0.4432,
+      "step": 1494
+    },
+    {
+      "epoch": 4.109965635738831,
+      "grad_norm": 0.10337495591575992,
+      "learning_rate": 7.343407939037166e-06,
+      "loss": 0.4442,
+      "step": 1495
+    },
+    {
+      "epoch": 4.112714776632302,
+      "grad_norm": 0.11699522345473407,
+      "learning_rate": 7.29903085675812e-06,
+      "loss": 0.4404,
+      "step": 1496
+    },
+    {
+      "epoch": 4.115463917525773,
+      "grad_norm": 0.1071569023006855,
+      "learning_rate": 7.254774803038986e-06,
+      "loss": 0.4448,
+      "step": 1497
+    },
+    {
+      "epoch": 4.118213058419244,
+      "grad_norm": 0.10590236218912877,
+      "learning_rate": 7.210639941674515e-06,
+      "loss": 0.4429,
+      "step": 1498
+    },
+    {
+      "epoch": 4.120962199312714,
+      "grad_norm": 0.10210210146806667,
+      "learning_rate": 7.166626436010893e-06,
+      "loss": 0.4415,
+      "step": 1499
+    },
+    {
+      "epoch": 4.123711340206185,
+      "grad_norm": 0.11125955970384037,
+      "learning_rate": 7.122734448945157e-06,
+      "loss": 0.4526,
+      "step": 1500
+    },
+    {
+      "epoch": 4.126460481099656,
+      "grad_norm": 0.12013542728499116,
+      "learning_rate": 7.078964142924607e-06,
+      "loss": 0.4419,
+      "step": 1501
+    },
+    {
+      "epoch": 4.129209621993128,
+      "grad_norm": 0.10484380190061886,
+      "learning_rate": 7.035315679946176e-06,
+      "loss": 0.4489,
+      "step": 1502
+    },
+    {
+      "epoch": 4.131958762886598,
+      "grad_norm": 0.09567503999966823,
+      "learning_rate": 6.991789221555887e-06,
+      "loss": 0.4372,
+      "step": 1503
+    },
+    {
+      "epoch": 4.134707903780069,
+      "grad_norm": 0.10540468880005585,
+      "learning_rate": 6.94838492884816e-06,
+      "loss": 0.4464,
+      "step": 1504
+    },
+    {
+      "epoch": 4.13745704467354,
+      "grad_norm": 0.11225166249471208,
+      "learning_rate": 6.905102962465302e-06,
+      "loss": 0.4439,
+      "step": 1505
+    },
+    {
+      "epoch": 4.140206185567011,
+      "grad_norm": 0.10759681388030853,
+      "learning_rate": 6.861943482596896e-06,
+      "loss": 0.4436,
+      "step": 1506
+    },
+    {
+      "epoch": 4.1429553264604815,
+      "grad_norm": 0.10130359403512029,
+      "learning_rate": 6.818906648979169e-06,
+      "loss": 0.4388,
+      "step": 1507
+    },
+    {
+      "epoch": 4.145704467353952,
+      "grad_norm": 0.1121174956770293,
+      "learning_rate": 6.775992620894434e-06,
+      "loss": 0.4507,
+      "step": 1508
+    },
+    {
+      "epoch": 4.148453608247423,
+      "grad_norm": 0.1014146829605852,
+      "learning_rate": 6.733201557170481e-06,
+      "loss": 0.4432,
+      "step": 1509
+    },
+    {
+      "epoch": 4.151202749140894,
+      "grad_norm": 0.09837158537375995,
+      "learning_rate": 6.690533616180031e-06,
+      "loss": 0.4416,
+      "step": 1510
+    },
+    {
+      "epoch": 4.153951890034365,
+      "grad_norm": 0.10857273055540849,
+      "learning_rate": 6.647988955840099e-06,
+      "loss": 0.4283,
+      "step": 1511
+    },
+    {
+      "epoch": 4.156701030927835,
+      "grad_norm": 0.11827148688247635,
+      "learning_rate": 6.6055677336114024e-06,
+      "loss": 0.4435,
+      "step": 1512
+    },
+    {
+      "epoch": 4.159450171821306,
+      "grad_norm": 0.10661184402237826,
+      "learning_rate": 6.563270106497866e-06,
+      "loss": 0.4493,
+      "step": 1513
+    },
+    {
+      "epoch": 4.162199312714777,
+      "grad_norm": 0.10027884158801532,
+      "learning_rate": 6.521096231045927e-06,
+      "loss": 0.4424,
+      "step": 1514
+    },
+    {
+      "epoch": 4.164948453608248,
+      "grad_norm": 0.10677714613702075,
+      "learning_rate": 6.479046263344044e-06,
+      "loss": 0.4407,
+      "step": 1515
+    },
+    {
+      "epoch": 4.1676975945017185,
+      "grad_norm": 0.10120025403165747,
+      "learning_rate": 6.43712035902206e-06,
+      "loss": 0.4482,
+      "step": 1516
+    },
+    {
+      "epoch": 4.170446735395189,
+      "grad_norm": 0.09576847158285014,
+      "learning_rate": 6.395318673250655e-06,
+      "loss": 0.4455,
+      "step": 1517
+    },
+    {
+      "epoch": 4.17319587628866,
+      "grad_norm": 0.10456502041919041,
+      "learning_rate": 6.353641360740796e-06,
+      "loss": 0.4442,
+      "step": 1518
+    },
+    {
+      "epoch": 4.175945017182131,
+      "grad_norm": 0.10975215070737157,
+      "learning_rate": 6.312088575743112e-06,
+      "loss": 0.4421,
+      "step": 1519
+    },
+    {
+      "epoch": 4.178694158075602,
+      "grad_norm": 0.09670376981272215,
+      "learning_rate": 6.270660472047318e-06,
+      "loss": 0.4357,
+      "step": 1520
+    },
+    {
+      "epoch": 4.181443298969072,
+      "grad_norm": 0.0989487056180283,
+      "learning_rate": 6.229357202981736e-06,
+      "loss": 0.4446,
+      "step": 1521
+    },
+    {
+      "epoch": 4.184192439862543,
+      "grad_norm": 0.11575820428367671,
+      "learning_rate": 6.188178921412622e-06,
+      "loss": 0.4376,
+      "step": 1522
+    },
+    {
+      "epoch": 4.186941580756014,
+      "grad_norm": 0.09506731515085455,
+      "learning_rate": 6.147125779743666e-06,
+      "loss": 0.4399,
+      "step": 1523
+    },
+    {
+      "epoch": 4.189690721649485,
+      "grad_norm": 0.10319725228868595,
+      "learning_rate": 6.106197929915385e-06,
+      "loss": 0.4494,
+      "step": 1524
+    },
+    {
+      "epoch": 4.1924398625429555,
+      "grad_norm": 0.10930094347383024,
+      "learning_rate": 6.065395523404616e-06,
+      "loss": 0.4416,
+      "step": 1525
+    },
+    {
+      "epoch": 4.195189003436426,
+      "grad_norm": 0.10333832350838476,
+      "learning_rate": 6.0247187112238936e-06,
+      "loss": 0.4419,
+      "step": 1526
+    },
+    {
+      "epoch": 4.197938144329897,
+      "grad_norm": 0.09863833187141094,
+      "learning_rate": 5.984167643920926e-06,
+      "loss": 0.45,
+      "step": 1527
+    },
+    {
+      "epoch": 4.200687285223368,
+      "grad_norm": 0.10338899938814769,
+      "learning_rate": 5.943742471578029e-06,
+      "loss": 0.4432,
+      "step": 1528
+    },
+    {
+      "epoch": 4.203436426116839,
+      "grad_norm": 0.10080408781457796,
+      "learning_rate": 5.903443343811583e-06,
+      "loss": 0.4498,
+      "step": 1529
+    },
+    {
+      "epoch": 4.206185567010309,
+      "grad_norm": 0.0995238925477759,
+      "learning_rate": 5.863270409771451e-06,
+      "loss": 0.4441,
+      "step": 1530
+    },
+    {
+      "epoch": 4.20893470790378,
+      "grad_norm": 0.09871347524798951,
+      "learning_rate": 5.823223818140458e-06,
+      "loss": 0.4465,
+      "step": 1531
+    },
+    {
+      "epoch": 4.211683848797251,
+      "grad_norm": 0.09741808192782717,
+      "learning_rate": 5.78330371713383e-06,
+      "loss": 0.4521,
+      "step": 1532
+    },
+    {
+      "epoch": 4.214432989690722,
+      "grad_norm": 0.10108809287384478,
+      "learning_rate": 5.7435102544986325e-06,
+      "loss": 0.4476,
+      "step": 1533
+    },
+    {
+      "epoch": 4.217182130584193,
+      "grad_norm": 0.09586236341789252,
+      "learning_rate": 5.70384357751323e-06,
+      "loss": 0.4417,
+      "step": 1534
+    },
+    {
+      "epoch": 4.219931271477663,
+      "grad_norm": 0.0982846297983466,
+      "learning_rate": 5.664303832986764e-06,
+      "loss": 0.4496,
+      "step": 1535
+    },
+    {
+      "epoch": 4.222680412371134,
+      "grad_norm": 0.09404435584314635,
+      "learning_rate": 5.624891167258569e-06,
+      "loss": 0.4415,
+      "step": 1536
+    },
+    {
+      "epoch": 4.225429553264605,
+      "grad_norm": 0.09792884175585087,
+      "learning_rate": 5.585605726197663e-06,
+      "loss": 0.4364,
+      "step": 1537
+    },
+    {
+      "epoch": 4.228178694158076,
+      "grad_norm": 0.0966404647926693,
+      "learning_rate": 5.54644765520219e-06,
+      "loss": 0.4464,
+      "step": 1538
+    },
+    {
+      "epoch": 4.2309278350515465,
+      "grad_norm": 0.09803372963199505,
+      "learning_rate": 5.507417099198886e-06,
+      "loss": 0.443,
+      "step": 1539
+    },
+    {
+      "epoch": 4.233676975945017,
+      "grad_norm": 0.10220414170891072,
+      "learning_rate": 5.468514202642574e-06,
+      "loss": 0.4416,
+      "step": 1540
+    },
+    {
+      "epoch": 4.236426116838488,
+      "grad_norm": 0.09940583554566296,
+      "learning_rate": 5.429739109515564e-06,
+      "loss": 0.4483,
+      "step": 1541
+    },
+    {
+      "epoch": 4.239175257731959,
+      "grad_norm": 0.09639208839569098,
+      "learning_rate": 5.3910919633271755e-06,
+      "loss": 0.442,
+      "step": 1542
+    },
+    {
+      "epoch": 4.24192439862543,
+      "grad_norm": 0.1117369320333374,
+      "learning_rate": 5.352572907113178e-06,
+      "loss": 0.4461,
+      "step": 1543
+    },
+    {
+      "epoch": 4.2446735395189,
+      "grad_norm": 0.09611791874864287,
+      "learning_rate": 5.314182083435282e-06,
+      "loss": 0.4346,
+      "step": 1544
+    },
+    {
+      "epoch": 4.247422680412371,
+      "grad_norm": 0.10794580998258632,
+      "learning_rate": 5.2759196343805885e-06,
+      "loss": 0.4415,
+      "step": 1545
+    },
+    {
+      "epoch": 4.250171821305842,
+      "grad_norm": 0.0952864724999734,
+      "learning_rate": 5.237785701561078e-06,
+      "loss": 0.4443,
+      "step": 1546
+    },
+    {
+      "epoch": 4.252920962199313,
+      "grad_norm": 0.10156924345931388,
+      "learning_rate": 5.1997804261131015e-06,
+      "loss": 0.449,
+      "step": 1547
+    },
+    {
+      "epoch": 4.2556701030927835,
+      "grad_norm": 0.09725478784660825,
+      "learning_rate": 5.161903948696813e-06,
+      "loss": 0.439,
+      "step": 1548
+    },
+    {
+      "epoch": 4.258419243986254,
+      "grad_norm": 0.10252085604277364,
+      "learning_rate": 5.124156409495693e-06,
+      "loss": 0.4385,
+      "step": 1549
+    },
+    {
+      "epoch": 4.261168384879725,
+      "grad_norm": 0.10131443009760514,
+      "learning_rate": 5.086537948216008e-06,
+      "loss": 0.4407,
+      "step": 1550
+    },
+    {
+      "epoch": 4.263917525773196,
+      "grad_norm": 0.09875784055342578,
+      "learning_rate": 5.049048704086295e-06,
+      "loss": 0.4469,
+      "step": 1551
+    },
+    {
+      "epoch": 4.266666666666667,
+      "grad_norm": 0.10157130046545569,
+      "learning_rate": 5.011688815856856e-06,
+      "loss": 0.4502,
+      "step": 1552
+    },
+    {
+      "epoch": 4.269415807560137,
+      "grad_norm": 0.10941012381717943,
+      "learning_rate": 4.974458421799231e-06,
+      "loss": 0.4411,
+      "step": 1553
+    },
+    {
+      "epoch": 4.272164948453608,
+      "grad_norm": 0.11064889415177875,
+      "learning_rate": 4.937357659705688e-06,
+      "loss": 0.4419,
+      "step": 1554
+    },
+    {
+      "epoch": 4.274914089347079,
+      "grad_norm": 0.10002641057536234,
+      "learning_rate": 4.90038666688875e-06,
+      "loss": 0.4558,
+      "step": 1555
+    },
+    {
+      "epoch": 4.27766323024055,
+      "grad_norm": 0.10016807122767976,
+      "learning_rate": 4.863545580180615e-06,
+      "loss": 0.4435,
+      "step": 1556
+    },
+    {
+      "epoch": 4.2804123711340205,
+      "grad_norm": 0.10851136506221658,
+      "learning_rate": 4.8268345359327075e-06,
+      "loss": 0.4398,
+      "step": 1557
+    },
+    {
+      "epoch": 4.283161512027491,
+      "grad_norm": 0.09488321817571037,
+      "learning_rate": 4.7902536700151504e-06,
+      "loss": 0.4429,
+      "step": 1558
+    },
+    {
+      "epoch": 4.285910652920962,
+      "grad_norm": 0.09962394867509372,
+      "learning_rate": 4.75380311781628e-06,
+      "loss": 0.4435,
+      "step": 1559
+    },
+    {
+      "epoch": 4.288659793814433,
+      "grad_norm": 0.10525730743423926,
+      "learning_rate": 4.717483014242134e-06,
+      "loss": 0.4488,
+      "step": 1560
+    },
+    {
+      "epoch": 4.291408934707904,
+      "grad_norm": 0.09444016052953329,
+      "learning_rate": 4.681293493715906e-06,
+      "loss": 0.4383,
+      "step": 1561
+    },
+    {
+      "epoch": 4.294158075601374,
+      "grad_norm": 0.0935555125757194,
+      "learning_rate": 4.645234690177556e-06,
+      "loss": 0.441,
+      "step": 1562
+    },
+    {
+      "epoch": 4.296907216494845,
+      "grad_norm": 0.09875362647621048,
+      "learning_rate": 4.6093067370832145e-06,
+      "loss": 0.4467,
+      "step": 1563
+    },
+    {
+      "epoch": 4.299656357388316,
+      "grad_norm": 0.10106687338028049,
+      "learning_rate": 4.573509767404733e-06,
+      "loss": 0.4398,
+      "step": 1564
+    },
+    {
+      "epoch": 4.302405498281787,
+      "grad_norm": 0.09430960821120914,
+      "learning_rate": 4.537843913629178e-06,
+      "loss": 0.4498,
+      "step": 1565
+    },
+    {
+      "epoch": 4.3051546391752575,
+      "grad_norm": 0.09341919086027475,
+      "learning_rate": 4.502309307758368e-06,
+      "loss": 0.4427,
+      "step": 1566
+    },
+    {
+      "epoch": 4.307903780068728,
+      "grad_norm": 0.09579646490730007,
+      "learning_rate": 4.46690608130834e-06,
+      "loss": 0.4504,
+      "step": 1567
+    },
+    {
+      "epoch": 4.310652920962199,
+      "grad_norm": 0.10140215061324537,
+      "learning_rate": 4.431634365308904e-06,
+      "loss": 0.4445,
+      "step": 1568
+    },
+    {
+      "epoch": 4.31340206185567,
+      "grad_norm": 0.09687704144452364,
+      "learning_rate": 4.39649429030311e-06,
+      "loss": 0.4478,
+      "step": 1569
+    },
+    {
+      "epoch": 4.316151202749141,
+      "grad_norm": 0.09438778342768989,
+      "learning_rate": 4.3614859863468425e-06,
+      "loss": 0.4386,
+      "step": 1570
+    },
+    {
+      "epoch": 4.318900343642611,
+      "grad_norm": 0.09156131739398227,
+      "learning_rate": 4.326609583008261e-06,
+      "loss": 0.4338,
+      "step": 1571
+    },
+    {
+      "epoch": 4.321649484536082,
+      "grad_norm": 0.09078660520618338,
+      "learning_rate": 4.2918652093673606e-06,
+      "loss": 0.442,
+      "step": 1572
+    },
+    {
+      "epoch": 4.324398625429553,
+      "grad_norm": 0.10260056112083875,
+      "learning_rate": 4.257252994015466e-06,
+      "loss": 0.4441,
+      "step": 1573
+    },
+    {
+      "epoch": 4.327147766323024,
+      "grad_norm": 0.10384708279251556,
+      "learning_rate": 4.222773065054817e-06,
+      "loss": 0.4466,
+      "step": 1574
+    },
+    {
+      "epoch": 4.329896907216495,
+      "grad_norm": 0.09444726168415894,
+      "learning_rate": 4.188425550098019e-06,
+      "loss": 0.444,
+      "step": 1575
+    },
+    {
+      "epoch": 4.332646048109965,
+      "grad_norm": 0.1139460077036934,
+      "learning_rate": 4.15421057626761e-06,
+      "loss": 0.4446,
+      "step": 1576
+    },
+    {
+      "epoch": 4.335395189003436,
+      "grad_norm": 0.09822740636361162,
+      "learning_rate": 4.120128270195585e-06,
+      "loss": 0.4424,
+      "step": 1577
+    },
+    {
+      "epoch": 4.338144329896907,
+      "grad_norm": 0.09913730861031846,
+      "learning_rate": 4.086178758022929e-06,
+      "loss": 0.4477,
+      "step": 1578
+    },
+    {
+      "epoch": 4.340893470790378,
+      "grad_norm": 0.09892692023825955,
+      "learning_rate": 4.052362165399148e-06,
+      "loss": 0.4462,
+      "step": 1579
+    },
+    {
+      "epoch": 4.3436426116838485,
+      "grad_norm": 0.09559985988178789,
+      "learning_rate": 4.018678617481797e-06,
+      "loss": 0.4518,
+      "step": 1580
+    },
+    {
+      "epoch": 4.346391752577319,
+      "grad_norm": 0.09645896728376689,
+      "learning_rate": 3.9851282389360336e-06,
+      "loss": 0.4411,
+      "step": 1581
+    },
+    {
+      "epoch": 4.34914089347079,
+      "grad_norm": 0.09621093774200687,
+      "learning_rate": 3.951711153934143e-06,
+      "loss": 0.4455,
+      "step": 1582
+    },
+    {
+      "epoch": 4.351890034364261,
+      "grad_norm": 0.09860601420290797,
+      "learning_rate": 3.918427486155079e-06,
+      "loss": 0.4452,
+      "step": 1583
+    },
+    {
+      "epoch": 4.354639175257732,
+      "grad_norm": 0.09460049776906895,
+      "learning_rate": 3.885277358784003e-06,
+      "loss": 0.4476,
+      "step": 1584
+    },
+    {
+      "epoch": 4.357388316151202,
+      "grad_norm": 0.0970914145620547,
+      "learning_rate": 3.85226089451185e-06,
+      "loss": 0.4486,
+      "step": 1585
+    },
+    {
+      "epoch": 4.360137457044673,
+      "grad_norm": 0.10970056178061441,
+      "learning_rate": 3.819378215534842e-06,
+      "loss": 0.4404,
+      "step": 1586
+    },
+    {
+      "epoch": 4.362886597938144,
+      "grad_norm": 0.09671374353988552,
+      "learning_rate": 3.7866294435540574e-06,
+      "loss": 0.4534,
+      "step": 1587
+    },
+    {
+      "epoch": 4.365635738831615,
+      "grad_norm": 0.09276644771447388,
+      "learning_rate": 3.7540146997749793e-06,
+      "loss": 0.4489,
+      "step": 1588
+    },
+    {
+      "epoch": 4.368384879725086,
+      "grad_norm": 0.09400993795205304,
+      "learning_rate": 3.72153410490705e-06,
+      "loss": 0.4467,
+      "step": 1589
+    },
+    {
+      "epoch": 4.371134020618557,
+      "grad_norm": 0.10312570233843393,
+      "learning_rate": 3.689187779163206e-06,
+      "loss": 0.4491,
+      "step": 1590
+    },
+    {
+      "epoch": 4.373883161512028,
+      "grad_norm": 0.10124141768595926,
+      "learning_rate": 3.6569758422594446e-06,
+      "loss": 0.4528,
+      "step": 1591
+    },
+    {
+      "epoch": 4.376632302405499,
+      "grad_norm": 0.0947775184692594,
+      "learning_rate": 3.6248984134143794e-06,
+      "loss": 0.4493,
+      "step": 1592
+    },
+    {
+      "epoch": 4.3793814432989695,
+      "grad_norm": 0.0949068368311883,
+      "learning_rate": 3.5929556113488117e-06,
+      "loss": 0.4502,
+      "step": 1593
+    },
+    {
+      "epoch": 4.38213058419244,
+      "grad_norm": 0.09225617183677981,
+      "learning_rate": 3.5611475542852714e-06,
+      "loss": 0.4378,
+      "step": 1594
+    },
+    {
+      "epoch": 4.384879725085911,
+      "grad_norm": 0.09712650449251893,
+      "learning_rate": 3.5294743599475754e-06,
+      "loss": 0.446,
+      "step": 1595
+    },
+    {
+      "epoch": 4.387628865979382,
+      "grad_norm": 0.10510272619911021,
+      "learning_rate": 3.4979361455604386e-06,
+      "loss": 0.4438,
+      "step": 1596
+    },
+    {
+      "epoch": 4.390378006872853,
+      "grad_norm": 0.10034299553088664,
+      "learning_rate": 3.4665330278489708e-06,
+      "loss": 0.4487,
+      "step": 1597
+    },
+    {
+      "epoch": 4.393127147766323,
+      "grad_norm": 0.09360138161898929,
+      "learning_rate": 3.4352651230383026e-06,
+      "loss": 0.4409,
+      "step": 1598
+    },
+    {
+      "epoch": 4.395876288659794,
+      "grad_norm": 0.09887187647716278,
+      "learning_rate": 3.4041325468531094e-06,
+      "loss": 0.4409,
+      "step": 1599
+    },
+    {
+      "epoch": 4.398625429553265,
+      "grad_norm": 0.10473071373698492,
+      "learning_rate": 3.3731354145172435e-06,
+      "loss": 0.4516,
+      "step": 1600
+    },
+    {
+      "epoch": 4.401374570446736,
+      "grad_norm": 0.1007375745415408,
+      "learning_rate": 3.342273840753221e-06,
+      "loss": 0.4383,
+      "step": 1601
+    },
+    {
+      "epoch": 4.4041237113402065,
+      "grad_norm": 0.09735792228082213,
+      "learning_rate": 3.311547939781887e-06,
+      "loss": 0.4409,
+      "step": 1602
+    },
+    {
+      "epoch": 4.406872852233677,
+      "grad_norm": 0.09459778267684918,
+      "learning_rate": 3.280957825321922e-06,
+      "loss": 0.4442,
+      "step": 1603
+    },
+    {
+      "epoch": 4.409621993127148,
+      "grad_norm": 0.09281792628016262,
+      "learning_rate": 3.250503610589482e-06,
+      "loss": 0.4491,
+      "step": 1604
+    },
+    {
+      "epoch": 4.412371134020619,
+      "grad_norm": 0.1007767987975053,
+      "learning_rate": 3.2201854082977292e-06,
+      "loss": 0.4517,
+      "step": 1605
+    },
+    {
+      "epoch": 4.41512027491409,
+      "grad_norm": 0.09789821294262532,
+      "learning_rate": 3.190003330656435e-06,
+      "loss": 0.4451,
+      "step": 1606
+    },
+    {
+      "epoch": 4.41786941580756,
+      "grad_norm": 0.10028034923814605,
+      "learning_rate": 3.159957489371559e-06,
+      "loss": 0.4367,
+      "step": 1607
+    },
+    {
+      "epoch": 4.420618556701031,
+      "grad_norm": 0.09081132192790976,
+      "learning_rate": 3.1300479956448693e-06,
+      "loss": 0.4387,
+      "step": 1608
+    },
+    {
+      "epoch": 4.423367697594502,
+      "grad_norm": 0.09711016308139389,
+      "learning_rate": 3.1002749601734618e-06,
+      "loss": 0.4428,
+      "step": 1609
+    },
+    {
+      "epoch": 4.426116838487973,
+      "grad_norm": 0.09435493012986142,
+      "learning_rate": 3.0706384931494137e-06,
+      "loss": 0.4328,
+      "step": 1610
+    },
+    {
+      "epoch": 4.4288659793814436,
+      "grad_norm": 0.09951785033015637,
+      "learning_rate": 3.0411387042593545e-06,
+      "loss": 0.4512,
+      "step": 1611
+    },
+    {
+      "epoch": 4.431615120274914,
+      "grad_norm": 0.0960124562000595,
+      "learning_rate": 3.0117757026840543e-06,
+      "loss": 0.4485,
+      "step": 1612
+    },
+    {
+      "epoch": 4.434364261168385,
+      "grad_norm": 0.09374569219824876,
+      "learning_rate": 2.9825495970980234e-06,
+      "loss": 0.4448,
+      "step": 1613
+    },
+    {
+      "epoch": 4.437113402061856,
+      "grad_norm": 0.09287833110723459,
+      "learning_rate": 2.953460495669096e-06,
+      "loss": 0.4473,
+      "step": 1614
+    },
+    {
+      "epoch": 4.439862542955327,
+      "grad_norm": 0.09330518761309428,
+      "learning_rate": 2.924508506058077e-06,
+      "loss": 0.4368,
+      "step": 1615
+    },
+    {
+      "epoch": 4.4426116838487975,
+      "grad_norm": 0.09851414675120082,
+      "learning_rate": 2.8956937354182923e-06,
+      "loss": 0.4438,
+      "step": 1616
+    },
+    {
+      "epoch": 4.445360824742268,
+      "grad_norm": 0.09503089697922397,
+      "learning_rate": 2.867016290395199e-06,
+      "loss": 0.4557,
+      "step": 1617
+    },
+    {
+      "epoch": 4.448109965635739,
+      "grad_norm": 0.09322925477662775,
+      "learning_rate": 2.838476277126012e-06,
+      "loss": 0.4428,
+      "step": 1618
+    },
+    {
+      "epoch": 4.45085910652921,
+      "grad_norm": 0.09242195549495436,
+      "learning_rate": 2.810073801239317e-06,
+      "loss": 0.446,
+      "step": 1619
+    },
+    {
+      "epoch": 4.453608247422681,
+      "grad_norm": 0.08945993365343624,
+      "learning_rate": 2.781808967854649e-06,
+      "loss": 0.4412,
+      "step": 1620
+    },
+    {
+      "epoch": 4.456357388316151,
+      "grad_norm": 0.09428643807864204,
+      "learning_rate": 2.753681881582115e-06,
+      "loss": 0.4416,
+      "step": 1621
+    },
+    {
+      "epoch": 4.459106529209622,
+      "grad_norm": 0.0931736838114746,
+      "learning_rate": 2.7256926465220177e-06,
+      "loss": 0.4472,
+      "step": 1622
+    },
+    {
+      "epoch": 4.461855670103093,
+      "grad_norm": 0.09181260953087753,
+      "learning_rate": 2.697841366264471e-06,
+      "loss": 0.4481,
+      "step": 1623
+    },
+    {
+      "epoch": 4.464604810996564,
+      "grad_norm": 0.09223991013911755,
+      "learning_rate": 2.670128143888988e-06,
+      "loss": 0.4417,
+      "step": 1624
+    },
+    {
+      "epoch": 4.4673539518900345,
+      "grad_norm": 0.09242485950713865,
+      "learning_rate": 2.6425530819641364e-06,
+      "loss": 0.4501,
+      "step": 1625
+    },
+    {
+      "epoch": 4.470103092783505,
+      "grad_norm": 0.09113105402491876,
+      "learning_rate": 2.6151162825471364e-06,
+      "loss": 0.4442,
+      "step": 1626
+    },
+    {
+      "epoch": 4.472852233676976,
+      "grad_norm": 0.09212808247265639,
+      "learning_rate": 2.587817847183489e-06,
+      "loss": 0.4429,
+      "step": 1627
+    },
+    {
+      "epoch": 4.475601374570447,
+      "grad_norm": 0.09767261371119439,
+      "learning_rate": 2.5606578769066026e-06,
+      "loss": 0.4432,
+      "step": 1628
+    },
+    {
+      "epoch": 4.478350515463918,
+      "grad_norm": 0.09743572389157769,
+      "learning_rate": 2.5336364722374106e-06,
+      "loss": 0.4487,
+      "step": 1629
+    },
+    {
+      "epoch": 4.481099656357388,
+      "grad_norm": 0.09102922314529087,
+      "learning_rate": 2.506753733184013e-06,
+      "loss": 0.4503,
+      "step": 1630
+    },
+    {
+      "epoch": 4.483848797250859,
+      "grad_norm": 0.09266255759834628,
+      "learning_rate": 2.480009759241302e-06,
+      "loss": 0.4405,
+      "step": 1631
+    },
+    {
+      "epoch": 4.48659793814433,
+      "grad_norm": 0.0925039627946117,
+      "learning_rate": 2.453404649390576e-06,
+      "loss": 0.4398,
+      "step": 1632
+    },
+    {
+      "epoch": 4.489347079037801,
+      "grad_norm": 0.09521064172876387,
+      "learning_rate": 2.4269385020991987e-06,
+      "loss": 0.4486,
+      "step": 1633
+    },
+    {
+      "epoch": 4.4920962199312715,
+      "grad_norm": 0.09113818118580695,
+      "learning_rate": 2.40061141532022e-06,
+      "loss": 0.4431,
+      "step": 1634
+    },
+    {
+      "epoch": 4.494845360824742,
+      "grad_norm": 0.10000887227672747,
+      "learning_rate": 2.374423486492021e-06,
+      "loss": 0.4388,
+      "step": 1635
+    },
+    {
+      "epoch": 4.497594501718213,
+      "grad_norm": 0.09939158503866538,
+      "learning_rate": 2.3483748125379434e-06,
+      "loss": 0.4526,
+      "step": 1636
+    },
+    {
+      "epoch": 4.500343642611684,
+      "grad_norm": 0.08865852194600819,
+      "learning_rate": 2.3224654898659348e-06,
+      "loss": 0.4458,
+      "step": 1637
+    },
+    {
+      "epoch": 4.503092783505155,
+      "grad_norm": 0.09118943093226906,
+      "learning_rate": 2.2966956143682094e-06,
+      "loss": 0.444,
+      "step": 1638
+    },
+    {
+      "epoch": 4.505841924398625,
+      "grad_norm": 0.09085329322674768,
+      "learning_rate": 2.2710652814208656e-06,
+      "loss": 0.4516,
+      "step": 1639
+    },
+    {
+      "epoch": 4.508591065292096,
+      "grad_norm": 0.09474948162033754,
+      "learning_rate": 2.2455745858835474e-06,
+      "loss": 0.4439,
+      "step": 1640
+    },
+    {
+      "epoch": 4.511340206185567,
+      "grad_norm": 0.09100873289075298,
+      "learning_rate": 2.2202236220990913e-06,
+      "loss": 0.4441,
+      "step": 1641
+    },
+    {
+      "epoch": 4.514089347079038,
+      "grad_norm": 0.09556416343413056,
+      "learning_rate": 2.1950124838931685e-06,
+      "loss": 0.4426,
+      "step": 1642
+    },
+    {
+      "epoch": 4.5168384879725085,
+      "grad_norm": 0.09124176627878805,
+      "learning_rate": 2.169941264573967e-06,
+      "loss": 0.4515,
+      "step": 1643
+    },
+    {
+      "epoch": 4.519587628865979,
+      "grad_norm": 0.08967523905753207,
+      "learning_rate": 2.1450100569317954e-06,
+      "loss": 0.4427,
+      "step": 1644
+    },
+    {
+      "epoch": 4.52233676975945,
+      "grad_norm": 0.08908910909605862,
+      "learning_rate": 2.1202189532387994e-06,
+      "loss": 0.4421,
+      "step": 1645
+    },
+    {
+      "epoch": 4.525085910652921,
+      "grad_norm": 0.08930544495492708,
+      "learning_rate": 2.0955680452485795e-06,
+      "loss": 0.4386,
+      "step": 1646
+    },
+    {
+      "epoch": 4.527835051546392,
+      "grad_norm": 0.09705957099996088,
+      "learning_rate": 2.0710574241958527e-06,
+      "loss": 0.4499,
+      "step": 1647
+    },
+    {
+      "epoch": 4.530584192439862,
+      "grad_norm": 0.09629213253097813,
+      "learning_rate": 2.0466871807961253e-06,
+      "loss": 0.4483,
+      "step": 1648
+    },
+    {
+      "epoch": 4.533333333333333,
+      "grad_norm": 0.08724520563465142,
+      "learning_rate": 2.0224574052453817e-06,
+      "loss": 0.4449,
+      "step": 1649
+    },
+    {
+      "epoch": 4.536082474226804,
+      "grad_norm": 0.09135773175337733,
+      "learning_rate": 1.9983681872196836e-06,
+      "loss": 0.4399,
+      "step": 1650
+    },
+    {
+      "epoch": 4.538831615120275,
+      "grad_norm": 0.09009008354611814,
+      "learning_rate": 1.9744196158749183e-06,
+      "loss": 0.4418,
+      "step": 1651
+    },
+    {
+      "epoch": 4.541580756013746,
+      "grad_norm": 0.09186018004376852,
+      "learning_rate": 1.9506117798463944e-06,
+      "loss": 0.4478,
+      "step": 1652
+    },
+    {
+      "epoch": 4.544329896907216,
+      "grad_norm": 0.08971975311980089,
+      "learning_rate": 1.9269447672485864e-06,
+      "loss": 0.4414,
+      "step": 1653
+    },
+    {
+      "epoch": 4.547079037800687,
+      "grad_norm": 0.08979949876575676,
+      "learning_rate": 1.9034186656747478e-06,
+      "loss": 0.4452,
+      "step": 1654
+    },
+    {
+      "epoch": 4.549828178694158,
+      "grad_norm": 0.0936250091293909,
+      "learning_rate": 1.8800335621966148e-06,
+      "loss": 0.4482,
+      "step": 1655
+    },
+    {
+      "epoch": 4.552577319587629,
+      "grad_norm": 0.08981143391720423,
+      "learning_rate": 1.8567895433640705e-06,
+      "loss": 0.4448,
+      "step": 1656
+    },
+    {
+      "epoch": 4.5553264604810995,
+      "grad_norm": 0.08587438163737723,
+      "learning_rate": 1.8336866952048683e-06,
+      "loss": 0.435,
+      "step": 1657
+    },
+    {
+      "epoch": 4.55807560137457,
+      "grad_norm": 0.09326486613993211,
+      "learning_rate": 1.8107251032242335e-06,
+      "loss": 0.4508,
+      "step": 1658
+    },
+    {
+      "epoch": 4.560824742268041,
+      "grad_norm": 0.09331518445515743,
+      "learning_rate": 1.7879048524046182e-06,
+      "loss": 0.4499,
+      "step": 1659
+    },
+    {
+      "epoch": 4.563573883161512,
+      "grad_norm": 0.08976792064689015,
+      "learning_rate": 1.765226027205369e-06,
+      "loss": 0.4466,
+      "step": 1660
+    },
+    {
+      "epoch": 4.566323024054983,
+      "grad_norm": 0.0875193893760576,
+      "learning_rate": 1.7426887115623791e-06,
+      "loss": 0.4452,
+      "step": 1661
+    },
+    {
+      "epoch": 4.569072164948453,
+      "grad_norm": 0.08920117183529058,
+      "learning_rate": 1.7202929888878329e-06,
+      "loss": 0.4401,
+      "step": 1662
+    },
+    {
+      "epoch": 4.571821305841924,
+      "grad_norm": 0.08832595498688149,
+      "learning_rate": 1.6980389420698395e-06,
+      "loss": 0.4451,
+      "step": 1663
+    },
+    {
+      "epoch": 4.574570446735395,
+      "grad_norm": 0.08859855283203406,
+      "learning_rate": 1.6759266534721952e-06,
+      "loss": 0.4431,
+      "step": 1664
+    },
+    {
+      "epoch": 4.577319587628866,
+      "grad_norm": 0.09152185332529074,
+      "learning_rate": 1.6539562049340085e-06,
+      "loss": 0.4402,
+      "step": 1665
+    },
+    {
+      "epoch": 4.5800687285223365,
+      "grad_norm": 0.08871925909133102,
+      "learning_rate": 1.6321276777694307e-06,
+      "loss": 0.4487,
+      "step": 1666
+    },
+    {
+      "epoch": 4.582817869415807,
+      "grad_norm": 0.09057802642441674,
+      "learning_rate": 1.6104411527673613e-06,
+      "loss": 0.4435,
+      "step": 1667
+    },
+    {
+      "epoch": 4.585567010309278,
+      "grad_norm": 0.09402629602885212,
+      "learning_rate": 1.588896710191139e-06,
+      "loss": 0.4489,
+      "step": 1668
+    },
+    {
+      "epoch": 4.588316151202749,
+      "grad_norm": 0.08990522146704653,
+      "learning_rate": 1.567494429778238e-06,
+      "loss": 0.4427,
+      "step": 1669
+    },
+    {
+      "epoch": 4.59106529209622,
+      "grad_norm": 0.09102458376196323,
+      "learning_rate": 1.546234390739998e-06,
+      "loss": 0.4498,
+      "step": 1670
+    },
+    {
+      "epoch": 4.59381443298969,
+      "grad_norm": 0.09383015002767303,
+      "learning_rate": 1.525116671761282e-06,
+      "loss": 0.4456,
+      "step": 1671
+    },
+    {
+      "epoch": 4.596563573883161,
+      "grad_norm": 0.087955107242829,
+      "learning_rate": 1.5041413510002544e-06,
+      "loss": 0.4495,
+      "step": 1672
+    },
+    {
+      "epoch": 4.599312714776632,
+      "grad_norm": 0.08874300385624094,
+      "learning_rate": 1.4833085060880349e-06,
+      "loss": 0.4412,
+      "step": 1673
+    },
+    {
+      "epoch": 4.602061855670103,
+      "grad_norm": 0.08466557475239722,
+      "learning_rate": 1.4626182141284085e-06,
+      "loss": 0.4455,
+      "step": 1674
+    },
+    {
+      "epoch": 4.6048109965635735,
+      "grad_norm": 0.08895580241312337,
+      "learning_rate": 1.4420705516976097e-06,
+      "loss": 0.4406,
+      "step": 1675
+    },
+    {
+      "epoch": 4.607560137457044,
+      "grad_norm": 0.08951149546757349,
+      "learning_rate": 1.421665594843953e-06,
+      "loss": 0.4485,
+      "step": 1676
+    },
+    {
+      "epoch": 4.610309278350515,
+      "grad_norm": 0.09239895115644584,
+      "learning_rate": 1.4014034190876057e-06,
+      "loss": 0.448,
+      "step": 1677
+    },
+    {
+      "epoch": 4.613058419243986,
+      "grad_norm": 0.08778403074207193,
+      "learning_rate": 1.3812840994202792e-06,
+      "loss": 0.44,
+      "step": 1678
+    },
+    {
+      "epoch": 4.615807560137457,
+      "grad_norm": 0.08817137101596591,
+      "learning_rate": 1.361307710304991e-06,
+      "loss": 0.4394,
+      "step": 1679
+    },
+    {
+      "epoch": 4.618556701030927,
+      "grad_norm": 0.09227710988150696,
+      "learning_rate": 1.3414743256757334e-06,
+      "loss": 0.4501,
+      "step": 1680
+    },
+    {
+      "epoch": 4.621305841924398,
+      "grad_norm": 0.08760382434561047,
+      "learning_rate": 1.3217840189372555e-06,
+      "loss": 0.4461,
+      "step": 1681
+    },
+    {
+      "epoch": 4.624054982817869,
+      "grad_norm": 0.09031853519071162,
+      "learning_rate": 1.3022368629647253e-06,
+      "loss": 0.4496,
+      "step": 1682
+    },
+    {
+      "epoch": 4.62680412371134,
+      "grad_norm": 0.09206239051131887,
+      "learning_rate": 1.2828329301035481e-06,
+      "loss": 0.4432,
+      "step": 1683
+    },
+    {
+      "epoch": 4.6295532646048105,
+      "grad_norm": 0.08843350392995919,
+      "learning_rate": 1.263572292169024e-06,
+      "loss": 0.4501,
+      "step": 1684
+    },
+    {
+      "epoch": 4.632302405498281,
+      "grad_norm": 0.09481473881768035,
+      "learning_rate": 1.2444550204461092e-06,
+      "loss": 0.4403,
+      "step": 1685
+    },
+    {
+      "epoch": 4.635051546391752,
+      "grad_norm": 0.08689601784166971,
+      "learning_rate": 1.2254811856891524e-06,
+      "loss": 0.4435,
+      "step": 1686
+    },
+    {
+      "epoch": 4.637800687285223,
+      "grad_norm": 0.08893870660419893,
+      "learning_rate": 1.2066508581216429e-06,
+      "loss": 0.4503,
+      "step": 1687
+    },
+    {
+      "epoch": 4.640549828178694,
+      "grad_norm": 0.08937431163006333,
+      "learning_rate": 1.187964107435926e-06,
+      "loss": 0.4456,
+      "step": 1688
+    },
+    {
+      "epoch": 4.643298969072165,
+      "grad_norm": 0.08865313646975571,
+      "learning_rate": 1.169421002792972e-06,
+      "loss": 0.4433,
+      "step": 1689
+    },
+    {
+      "epoch": 4.646048109965636,
+      "grad_norm": 0.08865228962259511,
+      "learning_rate": 1.151021612822092e-06,
+      "loss": 0.4491,
+      "step": 1690
+    },
+    {
+      "epoch": 4.648797250859107,
+      "grad_norm": 0.08885659723955662,
+      "learning_rate": 1.1327660056207113e-06,
+      "loss": 0.4368,
+      "step": 1691
+    },
+    {
+      "epoch": 4.651546391752578,
+      "grad_norm": 0.08962612308335349,
+      "learning_rate": 1.114654248754099e-06,
+      "loss": 0.4495,
+      "step": 1692
+    },
+    {
+      "epoch": 4.6542955326460484,
+      "grad_norm": 0.08766658442624285,
+      "learning_rate": 1.0966864092551233e-06,
+      "loss": 0.445,
+      "step": 1693
+    },
+    {
+      "epoch": 4.657044673539519,
+      "grad_norm": 0.08709960621618447,
+      "learning_rate": 1.0788625536240206e-06,
+      "loss": 0.4471,
+      "step": 1694
+    },
+    {
+      "epoch": 4.65979381443299,
+      "grad_norm": 0.09261985426164675,
+      "learning_rate": 1.0611827478281067e-06,
+      "loss": 0.4456,
+      "step": 1695
+    },
+    {
+      "epoch": 4.662542955326461,
+      "grad_norm": 0.08650196395240495,
+      "learning_rate": 1.0436470573015733e-06,
+      "loss": 0.4403,
+      "step": 1696
+    },
+    {
+      "epoch": 4.665292096219932,
+      "grad_norm": 0.08590676857219628,
+      "learning_rate": 1.026255546945234e-06,
+      "loss": 0.4449,
+      "step": 1697
+    },
+    {
+      "epoch": 4.668041237113402,
+      "grad_norm": 0.08741332778858373,
+      "learning_rate": 1.0090082811262802e-06,
+      "loss": 0.4492,
+      "step": 1698
+    },
+    {
+      "epoch": 4.670790378006873,
+      "grad_norm": 0.08738071184604224,
+      "learning_rate": 9.919053236780328e-07,
+      "loss": 0.444,
+      "step": 1699
+    },
+    {
+      "epoch": 4.673539518900344,
+      "grad_norm": 0.08949580602300157,
+      "learning_rate": 9.74946737899729e-07,
+      "loss": 0.4376,
+      "step": 1700
+    },
+    {
+      "epoch": 4.676288659793815,
+      "grad_norm": 0.08859670553407732,
+      "learning_rate": 9.581325865562775e-07,
+      "loss": 0.4486,
+      "step": 1701
+    },
+    {
+      "epoch": 4.6790378006872855,
+      "grad_norm": 0.08694870133225886,
+      "learning_rate": 9.414629318780189e-07,
+      "loss": 0.4483,
+      "step": 1702
+    },
+    {
+      "epoch": 4.681786941580756,
+      "grad_norm": 0.08702985035490321,
+      "learning_rate": 9.249378355605043e-07,
+      "loss": 0.4419,
+      "step": 1703
+    },
+    {
+      "epoch": 4.684536082474227,
+      "grad_norm": 0.08584597159667766,
+      "learning_rate": 9.085573587642637e-07,
+      "loss": 0.4517,
+      "step": 1704
+    },
+    {
+      "epoch": 4.687285223367698,
+      "grad_norm": 0.08578695182725778,
+      "learning_rate": 8.923215621145753e-07,
+      "loss": 0.4399,
+      "step": 1705
+    },
+    {
+      "epoch": 4.690034364261169,
+      "grad_norm": 0.08592571403893629,
+      "learning_rate": 8.762305057012654e-07,
+      "loss": 0.4515,
+      "step": 1706
+    },
+    {
+      "epoch": 4.692783505154639,
+      "grad_norm": 0.08693493065817048,
+      "learning_rate": 8.60284249078438e-07,
+      "loss": 0.4461,
+      "step": 1707
+    },
+    {
+      "epoch": 4.69553264604811,
+      "grad_norm": 0.08872474686819293,
+      "learning_rate": 8.444828512642966e-07,
+      "loss": 0.4545,
+      "step": 1708
+    },
+    {
+      "epoch": 4.698281786941581,
+      "grad_norm": 0.09336681238214921,
+      "learning_rate": 8.288263707409272e-07,
+      "loss": 0.4543,
+      "step": 1709
+    },
+    {
+      "epoch": 4.701030927835052,
+      "grad_norm": 0.0862849071211661,
+      "learning_rate": 8.133148654540402e-07,
+      "loss": 0.4409,
+      "step": 1710
+    },
+    {
+      "epoch": 4.7037800687285225,
+      "grad_norm": 0.08478357720964781,
+      "learning_rate": 7.979483928127974e-07,
+      "loss": 0.4422,
+      "step": 1711
+    },
+    {
+      "epoch": 4.706529209621993,
+      "grad_norm": 0.08654065268244254,
+      "learning_rate": 7.827270096895811e-07,
+      "loss": 0.4479,
+      "step": 1712
+    },
+    {
+      "epoch": 4.709278350515464,
+      "grad_norm": 0.08647792501003562,
+      "learning_rate": 7.676507724197946e-07,
+      "loss": 0.4484,
+      "step": 1713
+    },
+    {
+      "epoch": 4.712027491408935,
+      "grad_norm": 0.08597489991064104,
+      "learning_rate": 7.527197368016437e-07,
+      "loss": 0.4417,
+      "step": 1714
+    },
+    {
+      "epoch": 4.714776632302406,
+      "grad_norm": 0.0898277616665661,
+      "learning_rate": 7.37933958095911e-07,
+      "loss": 0.4514,
+      "step": 1715
+    },
+    {
+      "epoch": 4.717525773195876,
+      "grad_norm": 0.08697107341130339,
+      "learning_rate": 7.232934910258004e-07,
+      "loss": 0.4456,
+      "step": 1716
+    },
+    {
+      "epoch": 4.720274914089347,
+      "grad_norm": 0.08637312554758676,
+      "learning_rate": 7.087983897767059e-07,
+      "loss": 0.4477,
+      "step": 1717
+    },
+    {
+      "epoch": 4.723024054982818,
+      "grad_norm": 0.0882085833054912,
+      "learning_rate": 6.944487079959982e-07,
+      "loss": 0.4395,
+      "step": 1718
+    },
+    {
+      "epoch": 4.725773195876289,
+      "grad_norm": 0.085015391595313,
+      "learning_rate": 6.802444987928436e-07,
+      "loss": 0.4395,
+      "step": 1719
+    },
+    {
+      "epoch": 4.7285223367697595,
+      "grad_norm": 0.08451996713089384,
+      "learning_rate": 6.661858147380118e-07,
+      "loss": 0.437,
+      "step": 1720
+    },
+    {
+      "epoch": 4.73127147766323,
+      "grad_norm": 0.08738636017583253,
+      "learning_rate": 6.522727078636681e-07,
+      "loss": 0.4508,
+      "step": 1721
+    },
+    {
+      "epoch": 4.734020618556701,
+      "grad_norm": 0.08558858353740155,
+      "learning_rate": 6.385052296631955e-07,
+      "loss": 0.4464,
+      "step": 1722
+    },
+    {
+      "epoch": 4.736769759450172,
+      "grad_norm": 0.08661882774459452,
+      "learning_rate": 6.248834310909768e-07,
+      "loss": 0.4464,
+      "step": 1723
+    },
+    {
+      "epoch": 4.739518900343643,
+      "grad_norm": 0.08559450787680475,
+      "learning_rate": 6.114073625622396e-07,
+      "loss": 0.4372,
+      "step": 1724
+    },
+    {
+      "epoch": 4.742268041237113,
+      "grad_norm": 0.0894381151555916,
+      "learning_rate": 5.980770739528563e-07,
+      "loss": 0.4394,
+      "step": 1725
+    },
+    {
+      "epoch": 4.745017182130584,
+      "grad_norm": 0.0897214937051112,
+      "learning_rate": 5.84892614599144e-07,
+      "loss": 0.4511,
+      "step": 1726
+    },
+    {
+      "epoch": 4.747766323024055,
+      "grad_norm": 0.08890953858034158,
+      "learning_rate": 5.718540332977007e-07,
+      "loss": 0.4504,
+      "step": 1727
+    },
+    {
+      "epoch": 4.750515463917526,
+      "grad_norm": 0.08567497991113034,
+      "learning_rate": 5.589613783052317e-07,
+      "loss": 0.4422,
+      "step": 1728
+    },
+    {
+      "epoch": 4.7532646048109966,
+      "grad_norm": 0.08728822000081427,
+      "learning_rate": 5.462146973383453e-07,
+      "loss": 0.4589,
+      "step": 1729
+    },
+    {
+      "epoch": 4.756013745704467,
+      "grad_norm": 0.08723350175277615,
+      "learning_rate": 5.336140375733934e-07,
+      "loss": 0.4444,
+      "step": 1730
+    },
+    {
+      "epoch": 4.758762886597938,
+      "grad_norm": 0.0895784856053207,
+      "learning_rate": 5.211594456462932e-07,
+      "loss": 0.4502,
+      "step": 1731
+    },
+    {
+      "epoch": 4.761512027491409,
+      "grad_norm": 0.08733169216974808,
+      "learning_rate": 5.088509676523545e-07,
+      "loss": 0.4482,
+      "step": 1732
+    },
+    {
+      "epoch": 4.76426116838488,
+      "grad_norm": 0.08877952523971627,
+      "learning_rate": 4.966886491461109e-07,
+      "loss": 0.4478,
+      "step": 1733
+    },
+    {
+      "epoch": 4.7670103092783505,
+      "grad_norm": 0.08664355725321998,
+      "learning_rate": 4.846725351411507e-07,
+      "loss": 0.4463,
+      "step": 1734
+    },
+    {
+      "epoch": 4.769759450171821,
+      "grad_norm": 0.08722460501048601,
+      "learning_rate": 4.7280267010993974e-07,
+      "loss": 0.4484,
+      "step": 1735
+    },
+    {
+      "epoch": 4.772508591065292,
+      "grad_norm": 0.0869689702035606,
+      "learning_rate": 4.6107909798368324e-07,
+      "loss": 0.4449,
+      "step": 1736
+    },
+    {
+      "epoch": 4.775257731958763,
+      "grad_norm": 0.09408455047074887,
+      "learning_rate": 4.495018621521352e-07,
+      "loss": 0.4436,
+      "step": 1737
+    },
+    {
+      "epoch": 4.778006872852234,
+      "grad_norm": 0.09000664206402534,
+      "learning_rate": 4.3807100546344296e-07,
+      "loss": 0.44,
+      "step": 1738
+    },
+    {
+      "epoch": 4.780756013745704,
+      "grad_norm": 0.08893238330077334,
+      "learning_rate": 4.267865702240048e-07,
+      "loss": 0.4418,
+      "step": 1739
+    },
+    {
+      "epoch": 4.783505154639175,
+      "grad_norm": 0.08753276685918031,
+      "learning_rate": 4.1564859819830607e-07,
+      "loss": 0.4407,
+      "step": 1740
+    },
+    {
+      "epoch": 4.786254295532646,
+      "grad_norm": 0.08418852751539188,
+      "learning_rate": 4.046571306087499e-07,
+      "loss": 0.4429,
+      "step": 1741
+    },
+    {
+      "epoch": 4.789003436426117,
+      "grad_norm": 0.0880689357278319,
+      "learning_rate": 3.9381220813551555e-07,
+      "loss": 0.4517,
+      "step": 1742
+    },
+    {
+      "epoch": 4.7917525773195875,
+      "grad_norm": 0.08935991058559813,
+      "learning_rate": 3.831138709164295e-07,
+      "loss": 0.4464,
+      "step": 1743
+    },
+    {
+      "epoch": 4.794501718213058,
+      "grad_norm": 0.086751085582121,
+      "learning_rate": 3.725621585467698e-07,
+      "loss": 0.4504,
+      "step": 1744
+    },
+    {
+      "epoch": 4.797250859106529,
+      "grad_norm": 0.08660059471433466,
+      "learning_rate": 3.6215711007916434e-07,
+      "loss": 0.4457,
+      "step": 1745
+    },
+    {
+      "epoch": 4.8,
+      "grad_norm": 0.08795409500802658,
+      "learning_rate": 3.5189876402341727e-07,
+      "loss": 0.4487,
+      "step": 1746
+    },
+    {
+      "epoch": 4.802749140893471,
+      "grad_norm": 0.08517686102366333,
+      "learning_rate": 3.417871583463805e-07,
+      "loss": 0.442,
+      "step": 1747
+    },
+    {
+      "epoch": 4.805498281786941,
+      "grad_norm": 0.08656026625325718,
+      "learning_rate": 3.3182233047181154e-07,
+      "loss": 0.4482,
+      "step": 1748
+    },
+    {
+      "epoch": 4.808247422680412,
+      "grad_norm": 0.08742296055893738,
+      "learning_rate": 3.2200431728022676e-07,
+      "loss": 0.4497,
+      "step": 1749
+    },
+    {
+      "epoch": 4.810996563573883,
+      "grad_norm": 0.08706269427544457,
+      "learning_rate": 3.1233315510877714e-07,
+      "loss": 0.4399,
+      "step": 1750
+    },
+    {
+      "epoch": 4.813745704467354,
+      "grad_norm": 0.08541652658402427,
+      "learning_rate": 3.0280887975111087e-07,
+      "loss": 0.4463,
+      "step": 1751
+    },
+    {
+      "epoch": 4.8164948453608245,
+      "grad_norm": 0.08561518369119724,
+      "learning_rate": 2.9343152645723075e-07,
+      "loss": 0.4496,
+      "step": 1752
+    },
+    {
+      "epoch": 4.819243986254295,
+      "grad_norm": 0.08671209749993644,
+      "learning_rate": 2.8420112993337026e-07,
+      "loss": 0.4441,
+      "step": 1753
+    },
+    {
+      "epoch": 4.821993127147766,
+      "grad_norm": 0.0868113057408611,
+      "learning_rate": 2.75117724341869e-07,
+      "loss": 0.4539,
+      "step": 1754
+    },
+    {
+      "epoch": 4.824742268041237,
+      "grad_norm": 0.08787299513468588,
+      "learning_rate": 2.661813433010485e-07,
+      "loss": 0.4469,
+      "step": 1755
+    },
+    {
+      "epoch": 4.827491408934708,
+      "grad_norm": 0.08426954798208526,
+      "learning_rate": 2.5739201988506544e-07,
+      "loss": 0.4452,
+      "step": 1756
+    },
+    {
+      "epoch": 4.830240549828178,
+      "grad_norm": 0.08925999331930048,
+      "learning_rate": 2.487497866238231e-07,
+      "loss": 0.4431,
+      "step": 1757
+    },
+    {
+      "epoch": 4.832989690721649,
+      "grad_norm": 0.09009930022107024,
+      "learning_rate": 2.4025467550283345e-07,
+      "loss": 0.4428,
+      "step": 1758
+    },
+    {
+      "epoch": 4.83573883161512,
+      "grad_norm": 0.08903112575869974,
+      "learning_rate": 2.3190671796307963e-07,
+      "loss": 0.4376,
+      "step": 1759
+    },
+    {
+      "epoch": 4.838487972508591,
+      "grad_norm": 0.087097271598177,
+      "learning_rate": 2.237059449009449e-07,
+      "loss": 0.4514,
+      "step": 1760
+    },
+    {
+      "epoch": 4.8412371134020615,
+      "grad_norm": 0.08665721242897642,
+      "learning_rate": 2.1565238666805266e-07,
+      "loss": 0.4507,
+      "step": 1761
+    },
+    {
+      "epoch": 4.843986254295532,
+      "grad_norm": 0.08570901423189047,
+      "learning_rate": 2.0774607307118665e-07,
+      "loss": 0.444,
+      "step": 1762
+    },
+    {
+      "epoch": 4.846735395189003,
+      "grad_norm": 0.08855478539796897,
+      "learning_rate": 1.9998703337216207e-07,
+      "loss": 0.4459,
+      "step": 1763
+    },
+    {
+      "epoch": 4.849484536082474,
+      "grad_norm": 0.08830013933034192,
+      "learning_rate": 1.9237529628772345e-07,
+      "loss": 0.4471,
+      "step": 1764
+    },
+    {
+      "epoch": 4.852233676975945,
+      "grad_norm": 0.08632909822266122,
+      "learning_rate": 1.8491088998943806e-07,
+      "loss": 0.4443,
+      "step": 1765
+    },
+    {
+      "epoch": 4.854982817869415,
+      "grad_norm": 0.08788988835658808,
+      "learning_rate": 1.7759384210358943e-07,
+      "loss": 0.4495,
+      "step": 1766
+    },
+    {
+      "epoch": 4.857731958762886,
+      "grad_norm": 0.08641182871335162,
+      "learning_rate": 1.7042417971108837e-07,
+      "loss": 0.448,
+      "step": 1767
+    },
+    {
+      "epoch": 4.860481099656358,
+      "grad_norm": 0.08684528157216327,
+      "learning_rate": 1.6340192934734879e-07,
+      "loss": 0.4503,
+      "step": 1768
+    },
+    {
+      "epoch": 4.863230240549829,
+      "grad_norm": 0.08740681191652898,
+      "learning_rate": 1.5652711700220756e-07,
+      "loss": 0.4476,
+      "step": 1769
+    },
+    {
+      "epoch": 4.8659793814432994,
+      "grad_norm": 0.08455281840311087,
+      "learning_rate": 1.4979976811982267e-07,
+      "loss": 0.4501,
+      "step": 1770
+    },
+    {
+      "epoch": 4.86872852233677,
+      "grad_norm": 0.08621142637068548,
+      "learning_rate": 1.4321990759858406e-07,
+      "loss": 0.4427,
+      "step": 1771
+    },
+    {
+      "epoch": 4.871477663230241,
+      "grad_norm": 0.09048418177504676,
+      "learning_rate": 1.3678755979100734e-07,
+      "loss": 0.4414,
+      "step": 1772
+    },
+    {
+      "epoch": 4.874226804123712,
+      "grad_norm": 0.08552997152050767,
+      "learning_rate": 1.3050274850366252e-07,
+      "loss": 0.4452,
+      "step": 1773
+    },
+    {
+      "epoch": 4.876975945017183,
+      "grad_norm": 0.09163336347075388,
+      "learning_rate": 1.2436549699706756e-07,
+      "loss": 0.4421,
+      "step": 1774
+    },
+    {
+      "epoch": 4.879725085910653,
+      "grad_norm": 0.08581324177498632,
+      "learning_rate": 1.183758279856262e-07,
+      "loss": 0.445,
+      "step": 1775
+    },
+    {
+      "epoch": 4.882474226804124,
+      "grad_norm": 0.08520078540460027,
+      "learning_rate": 1.1253376363751234e-07,
+      "loss": 0.4398,
+      "step": 1776
+    },
+    {
+      "epoch": 4.885223367697595,
+      "grad_norm": 0.08389381645053694,
+      "learning_rate": 1.0683932557461696e-07,
+      "loss": 0.4509,
+      "step": 1777
+    },
+    {
+      "epoch": 4.887972508591066,
+      "grad_norm": 0.08463720514626481,
+      "learning_rate": 1.0129253487244584e-07,
+      "loss": 0.444,
+      "step": 1778
+    },
+    {
+      "epoch": 4.8907216494845365,
+      "grad_norm": 0.08376091905777829,
+      "learning_rate": 9.589341206006186e-08,
+      "loss": 0.4424,
+      "step": 1779
+    },
+    {
+      "epoch": 4.893470790378007,
+      "grad_norm": 0.0864473265773848,
+      "learning_rate": 9.064197711999178e-08,
+      "loss": 0.436,
+      "step": 1780
+    },
+    {
+      "epoch": 4.896219931271478,
+      "grad_norm": 0.08780409654999521,
+      "learning_rate": 8.553824948816402e-08,
+      "loss": 0.4439,
+      "step": 1781
+    },
+    {
+      "epoch": 4.898969072164949,
+      "grad_norm": 0.08643060928003499,
+      "learning_rate": 8.058224805382431e-08,
+      "loss": 0.457,
+      "step": 1782
+    },
+    {
+      "epoch": 4.90171821305842,
+      "grad_norm": 0.08999003837315114,
+      "learning_rate": 7.577399115948236e-08,
+      "loss": 0.4536,
+      "step": 1783
+    },
+    {
+      "epoch": 4.90446735395189,
+      "grad_norm": 0.08500920989818493,
+      "learning_rate": 7.111349660082756e-08,
+      "loss": 0.4524,
+      "step": 1784
+    },
+    {
+      "epoch": 4.907216494845361,
+      "grad_norm": 0.08713394706241256,
+      "learning_rate": 6.660078162667561e-08,
+      "loss": 0.4474,
+      "step": 1785
+    },
+    {
+      "epoch": 4.909965635738832,
+      "grad_norm": 0.08415301103290775,
+      "learning_rate": 6.223586293890193e-08,
+      "loss": 0.4388,
+      "step": 1786
+    },
+    {
+      "epoch": 4.912714776632303,
+      "grad_norm": 0.08641909751480724,
+      "learning_rate": 5.8018756692370673e-08,
+      "loss": 0.4452,
+      "step": 1787
+    },
+    {
+      "epoch": 4.9154639175257735,
+      "grad_norm": 0.085407296137395,
+      "learning_rate": 5.3949478494885745e-08,
+      "loss": 0.4442,
+      "step": 1788
+    },
+    {
+      "epoch": 4.918213058419244,
+      "grad_norm": 0.08620621228120745,
+      "learning_rate": 5.002804340712875e-08,
+      "loss": 0.4395,
+      "step": 1789
+    },
+    {
+      "epoch": 4.920962199312715,
+      "grad_norm": 0.08440435770555102,
+      "learning_rate": 4.625446594260563e-08,
+      "loss": 0.4475,
+      "step": 1790
+    },
+    {
+      "epoch": 4.923711340206186,
+      "grad_norm": 0.08498966775862096,
+      "learning_rate": 4.262876006758454e-08,
+      "loss": 0.4461,
+      "step": 1791
+    },
+    {
+      "epoch": 4.926460481099657,
+      "grad_norm": 0.0852940765401518,
+      "learning_rate": 3.915093920105584e-08,
+      "loss": 0.4508,
+      "step": 1792
+    },
+    {
+      "epoch": 4.929209621993127,
+      "grad_norm": 0.08708725826340913,
+      "learning_rate": 3.582101621467882e-08,
+      "loss": 0.4478,
+      "step": 1793
+    },
+    {
+      "epoch": 4.931958762886598,
+      "grad_norm": 0.08563954373853248,
+      "learning_rate": 3.263900343272841e-08,
+      "loss": 0.4506,
+      "step": 1794
+    },
+    {
+      "epoch": 4.934707903780069,
+      "grad_norm": 0.08351208793444787,
+      "learning_rate": 2.960491263205523e-08,
+      "loss": 0.4358,
+      "step": 1795
+    },
+    {
+      "epoch": 4.93745704467354,
+      "grad_norm": 0.08968304296903096,
+      "learning_rate": 2.6718755042041134e-08,
+      "loss": 0.4408,
+      "step": 1796
+    },
+    {
+      "epoch": 4.9402061855670105,
+      "grad_norm": 0.08687120751973657,
+      "learning_rate": 2.3980541344554852e-08,
+      "loss": 0.4472,
+      "step": 1797
+    },
+    {
+      "epoch": 4.942955326460481,
+      "grad_norm": 0.0843041818695698,
+      "learning_rate": 2.1390281673916435e-08,
+      "loss": 0.4433,
+      "step": 1798
+    },
+    {
+      "epoch": 4.945704467353952,
+      "grad_norm": 0.08678186311343745,
+      "learning_rate": 1.894798561685729e-08,
+      "loss": 0.4401,
+      "step": 1799
+    },
+    {
+      "epoch": 4.948453608247423,
+      "grad_norm": 0.08552567460187925,
+      "learning_rate": 1.6653662212484656e-08,
+      "loss": 0.4463,
+      "step": 1800
+    },
+    {
+      "epoch": 4.951202749140894,
+      "grad_norm": 0.08327662566119383,
+      "learning_rate": 1.4507319952246079e-08,
+      "loss": 0.4428,
+      "step": 1801
+    },
+    {
+      "epoch": 4.953951890034364,
+      "grad_norm": 0.08558816066436331,
+      "learning_rate": 1.2508966779907206e-08,
+      "loss": 0.4439,
+      "step": 1802
+    },
+    {
+      "epoch": 4.956701030927835,
+      "grad_norm": 0.08685318316335923,
+      "learning_rate": 1.0658610091507371e-08,
+      "loss": 0.4433,
+      "step": 1803
+    },
+    {
+      "epoch": 4.959450171821306,
+      "grad_norm": 0.08552301347515416,
+      "learning_rate": 8.956256735355162e-09,
+      "loss": 0.4462,
+      "step": 1804
+    },
+    {
+      "epoch": 4.962199312714777,
+      "grad_norm": 0.08918690199355128,
+      "learning_rate": 7.401913011966244e-09,
+      "loss": 0.4516,
+      "step": 1805
+    },
+    {
+      "epoch": 4.9649484536082475,
+      "grad_norm": 0.08439607746487938,
+      "learning_rate": 5.995584674085564e-09,
+      "loss": 0.4443,
+      "step": 1806
+    },
+    {
+      "epoch": 4.967697594501718,
+      "grad_norm": 0.0859086823143405,
+      "learning_rate": 4.737276926620738e-09,
+      "loss": 0.4477,
+      "step": 1807
+    },
+    {
+      "epoch": 4.970446735395189,
+      "grad_norm": 0.08603802799919767,
+      "learning_rate": 3.6269944266686953e-09,
+      "loss": 0.4398,
+      "step": 1808
+    },
+    {
+      "epoch": 4.97319587628866,
+      "grad_norm": 0.08797273612741212,
+      "learning_rate": 2.664741283453509e-09,
+      "loss": 0.4455,
+      "step": 1809
+    },
+    {
+      "epoch": 4.975945017182131,
+      "grad_norm": 0.08498147628736628,
+      "learning_rate": 1.8505210583441568e-09,
+      "loss": 0.4417,
+      "step": 1810
+    },
+    {
+      "epoch": 4.9786941580756015,
+      "grad_norm": 0.08573341738089828,
+      "learning_rate": 1.184336764823435e-09,
+      "loss": 0.4457,
+      "step": 1811
+    },
+    {
+      "epoch": 4.981443298969072,
+      "grad_norm": 0.08322231115466387,
+      "learning_rate": 6.661908684924002e-10,
+      "loss": 0.454,
+      "step": 1812
+    },
+    {
+      "epoch": 4.984192439862543,
+      "grad_norm": 0.08656315765889926,
+      "learning_rate": 2.9608528703928274e-10,
+      "loss": 0.4449,
+      "step": 1813
+    },
+    {
+      "epoch": 4.986941580756014,
+      "grad_norm": 0.08364091527237393,
+      "learning_rate": 7.402139024836886e-11,
+      "loss": 0.455,
+      "step": 1814
+    },
+    {
+      "epoch": 4.989690721649485,
+      "grad_norm": 0.08519652883753787,
+      "learning_rate": 0.0,
+      "loss": 0.4445,
+      "step": 1815
+    },
+    {
+      "epoch": 4.989690721649485,
+      "step": 1815,
+      "total_flos": 3.73703706563429e+19,
+      "train_loss": 0.5429748183767986,
+      "train_runtime": 94067.7478,
+      "train_samples_per_second": 9.898,
+      "train_steps_per_second": 0.019
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 1815,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 5,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 3.73703706563429e+19,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}