diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,17611 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.5467114215730745,
+  "eval_steps": 500,
+  "global_step": 2511,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0002178550356283756,
+      "grad_norm": 3.296875,
+      "learning_rate": 0.0,
+      "loss": 1.8496471643447876,
+      "step": 1
+    },
+    {
+      "epoch": 0.0004357100712567512,
+      "grad_norm": 2.78125,
+      "learning_rate": 5.797101449275363e-07,
+      "loss": 1.7491246461868286,
+      "step": 2
+    },
+    {
+      "epoch": 0.0006535651068851269,
+      "grad_norm": 3.53125,
+      "learning_rate": 1.1594202898550726e-06,
+      "loss": 1.705483078956604,
+      "step": 3
+    },
+    {
+      "epoch": 0.0008714201425135024,
+      "grad_norm": 2.75,
+      "learning_rate": 1.7391304347826088e-06,
+      "loss": 1.746169090270996,
+      "step": 4
+    },
+    {
+      "epoch": 0.001089275178141878,
+      "grad_norm": 2.78125,
+      "learning_rate": 2.3188405797101453e-06,
+      "loss": 1.6889724731445312,
+      "step": 5
+    },
+    {
+      "epoch": 0.0013071302137702538,
+      "grad_norm": 2.84375,
+      "learning_rate": 2.8985507246376816e-06,
+      "loss": 1.6988413333892822,
+      "step": 6
+    },
+    {
+      "epoch": 0.0015249852493986293,
+      "grad_norm": 3.28125,
+      "learning_rate": 3.4782608695652175e-06,
+      "loss": 1.8085319995880127,
+      "step": 7
+    },
+    {
+      "epoch": 0.0017428402850270049,
+      "grad_norm": 3.125,
+      "learning_rate": 4.057971014492754e-06,
+      "loss": 1.607370138168335,
+      "step": 8
+    },
+    {
+      "epoch": 0.0019606953206553804,
+      "grad_norm": 2.75,
+      "learning_rate": 4.637681159420291e-06,
+      "loss": 1.8274322748184204,
+      "step": 9
+    },
+    {
+      "epoch": 0.002178550356283756,
+      "grad_norm": 2.453125,
+      "learning_rate": 5.2173913043478265e-06,
+      "loss": 1.7205524444580078,
+      "step": 10
+    },
+    {
+      "epoch": 0.002396405391912132,
+      "grad_norm": 2.75,
+      "learning_rate": 5.797101449275363e-06,
+      "loss": 1.7673749923706055,
+      "step": 11
+    },
+    {
+      "epoch": 0.0026142604275405075,
+      "grad_norm": 2.75,
+      "learning_rate": 6.376811594202898e-06,
+      "loss": 1.6814430952072144,
+      "step": 12
+    },
+    {
+      "epoch": 0.002832115463168883,
+      "grad_norm": 2.390625,
+      "learning_rate": 6.956521739130435e-06,
+      "loss": 1.681013822555542,
+      "step": 13
+    },
+    {
+      "epoch": 0.0030499704987972586,
+      "grad_norm": 2.078125,
+      "learning_rate": 7.536231884057972e-06,
+      "loss": 1.6181104183197021,
+      "step": 14
+    },
+    {
+      "epoch": 0.003267825534425634,
+      "grad_norm": 1.984375,
+      "learning_rate": 8.115942028985508e-06,
+      "loss": 1.742430567741394,
+      "step": 15
+    },
+    {
+      "epoch": 0.0034856805700540098,
+      "grad_norm": 2.109375,
+      "learning_rate": 8.695652173913044e-06,
+      "loss": 1.600385308265686,
+      "step": 16
+    },
+    {
+      "epoch": 0.0037035356056823853,
+      "grad_norm": 1.984375,
+      "learning_rate": 9.275362318840581e-06,
+      "loss": 1.4717581272125244,
+      "step": 17
+    },
+    {
+      "epoch": 0.003921390641310761,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.855072463768118e-06,
+      "loss": 1.7800095081329346,
+      "step": 18
+    },
+    {
+      "epoch": 0.004139245676939136,
+      "grad_norm": 2.28125,
+      "learning_rate": 1.0434782608695653e-05,
+      "loss": 1.5602055788040161,
+      "step": 19
+    },
+    {
+      "epoch": 0.004357100712567512,
+      "grad_norm": 2.015625,
+      "learning_rate": 1.101449275362319e-05,
+      "loss": 1.5334831476211548,
+      "step": 20
+    },
+    {
+      "epoch": 0.004574955748195888,
+      "grad_norm": 2.234375,
+      "learning_rate": 1.1594202898550726e-05,
+      "loss": 1.7320983409881592,
+      "step": 21
+    },
+    {
+      "epoch": 0.004792810783824264,
+      "grad_norm": 2.390625,
+      "learning_rate": 1.2173913043478263e-05,
+      "loss": 1.6446813344955444,
+      "step": 22
+    },
+    {
+      "epoch": 0.0050106658194526395,
+      "grad_norm": 1.953125,
+      "learning_rate": 1.2753623188405797e-05,
+      "loss": 1.62025785446167,
+      "step": 23
+    },
+    {
+      "epoch": 0.005228520855081015,
+      "grad_norm": 1.9375,
+      "learning_rate": 1.3333333333333333e-05,
+      "loss": 1.6385928392410278,
+      "step": 24
+    },
+    {
+      "epoch": 0.005446375890709391,
+      "grad_norm": 1.75,
+      "learning_rate": 1.391304347826087e-05,
+      "loss": 1.4344453811645508,
+      "step": 25
+    },
+    {
+      "epoch": 0.005664230926337766,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.4492753623188407e-05,
+      "loss": 1.5687377452850342,
+      "step": 26
+    },
+    {
+      "epoch": 0.005882085961966142,
+      "grad_norm": 1.34375,
+      "learning_rate": 1.5072463768115944e-05,
+      "loss": 1.5073459148406982,
+      "step": 27
+    },
+    {
+      "epoch": 0.006099940997594517,
+      "grad_norm": 1.3671875,
+      "learning_rate": 1.565217391304348e-05,
+      "loss": 1.4817824363708496,
+      "step": 28
+    },
+    {
+      "epoch": 0.006317796033222893,
+      "grad_norm": 1.4453125,
+      "learning_rate": 1.6231884057971015e-05,
+      "loss": 1.4426829814910889,
+      "step": 29
+    },
+    {
+      "epoch": 0.006535651068851268,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.681159420289855e-05,
+      "loss": 1.572144627571106,
+      "step": 30
+    },
+    {
+      "epoch": 0.006753506104479644,
+      "grad_norm": 1.25,
+      "learning_rate": 1.739130434782609e-05,
+      "loss": 1.643051266670227,
+      "step": 31
+    },
+    {
+      "epoch": 0.0069713611401080195,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.7971014492753624e-05,
+      "loss": 1.5810363292694092,
+      "step": 32
+    },
+    {
+      "epoch": 0.007189216175736395,
+      "grad_norm": 1.1328125,
+      "learning_rate": 1.8550724637681162e-05,
+      "loss": 1.4251741170883179,
+      "step": 33
+    },
+    {
+      "epoch": 0.007407071211364771,
+      "grad_norm": 1.1875,
+      "learning_rate": 1.9130434782608697e-05,
+      "loss": 1.377197504043579,
+      "step": 34
+    },
+    {
+      "epoch": 0.007624926246993147,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.9710144927536236e-05,
+      "loss": 1.5657248497009277,
+      "step": 35
+    },
+    {
+      "epoch": 0.007842781282621522,
+      "grad_norm": 1.40625,
+      "learning_rate": 2.028985507246377e-05,
+      "loss": 1.4046844244003296,
+      "step": 36
+    },
+    {
+      "epoch": 0.008060636318249898,
+      "grad_norm": 1.2734375,
+      "learning_rate": 2.0869565217391306e-05,
+      "loss": 1.413676142692566,
+      "step": 37
+    },
+    {
+      "epoch": 0.008278491353878273,
+      "grad_norm": 1.125,
+      "learning_rate": 2.1449275362318844e-05,
+      "loss": 1.656702995300293,
+      "step": 38
+    },
+    {
+      "epoch": 0.00849634638950665,
+      "grad_norm": 1.359375,
+      "learning_rate": 2.202898550724638e-05,
+      "loss": 1.4696394205093384,
+      "step": 39
+    },
+    {
+      "epoch": 0.008714201425135024,
+      "grad_norm": 1.5859375,
+      "learning_rate": 2.2608695652173914e-05,
+      "loss": 1.4343537092208862,
+      "step": 40
+    },
+    {
+      "epoch": 0.0089320564607634,
+      "grad_norm": 1.25,
+      "learning_rate": 2.3188405797101453e-05,
+      "loss": 1.4311915636062622,
+      "step": 41
+    },
+    {
+      "epoch": 0.009149911496391777,
+      "grad_norm": 1.4765625,
+      "learning_rate": 2.3768115942028988e-05,
+      "loss": 1.4660894870758057,
+      "step": 42
+    },
+    {
+      "epoch": 0.009367766532020151,
+      "grad_norm": 1.3203125,
+      "learning_rate": 2.4347826086956526e-05,
+      "loss": 1.4359508752822876,
+      "step": 43
+    },
+    {
+      "epoch": 0.009585621567648528,
+      "grad_norm": 1.1640625,
+      "learning_rate": 2.4927536231884058e-05,
+      "loss": 1.2145781517028809,
+      "step": 44
+    },
+    {
+      "epoch": 0.009803476603276903,
+      "grad_norm": 1.25,
+      "learning_rate": 2.5507246376811593e-05,
+      "loss": 1.3644131422042847,
+      "step": 45
+    },
+    {
+      "epoch": 0.010021331638905279,
+      "grad_norm": 1.1015625,
+      "learning_rate": 2.608695652173913e-05,
+      "loss": 1.4376693964004517,
+      "step": 46
+    },
+    {
+      "epoch": 0.010239186674533654,
+      "grad_norm": 1.125,
+      "learning_rate": 2.6666666666666667e-05,
+      "loss": 1.4320605993270874,
+      "step": 47
+    },
+    {
+      "epoch": 0.01045704171016203,
+      "grad_norm": 1.3203125,
+      "learning_rate": 2.7246376811594205e-05,
+      "loss": 1.4258242845535278,
+      "step": 48
+    },
+    {
+      "epoch": 0.010674896745790405,
+      "grad_norm": 1.3828125,
+      "learning_rate": 2.782608695652174e-05,
+      "loss": 1.4765112400054932,
+      "step": 49
+    },
+    {
+      "epoch": 0.010892751781418781,
+      "grad_norm": 1.125,
+      "learning_rate": 2.840579710144928e-05,
+      "loss": 1.4314082860946655,
+      "step": 50
+    },
+    {
+      "epoch": 0.011110606817047156,
+      "grad_norm": 1.6796875,
+      "learning_rate": 2.8985507246376814e-05,
+      "loss": 1.4419467449188232,
+      "step": 51
+    },
+    {
+      "epoch": 0.011328461852675532,
+      "grad_norm": 1.0078125,
+      "learning_rate": 2.956521739130435e-05,
+      "loss": 1.323533058166504,
+      "step": 52
+    },
+    {
+      "epoch": 0.011546316888303907,
+      "grad_norm": 1.078125,
+      "learning_rate": 3.0144927536231887e-05,
+      "loss": 1.4481195211410522,
+      "step": 53
+    },
+    {
+      "epoch": 0.011764171923932283,
+      "grad_norm": 1.0078125,
+      "learning_rate": 3.072463768115942e-05,
+      "loss": 1.3901069164276123,
+      "step": 54
+    },
+    {
+      "epoch": 0.011982026959560658,
+      "grad_norm": 1.3671875,
+      "learning_rate": 3.130434782608696e-05,
+      "loss": 1.226739525794983,
+      "step": 55
+    },
+    {
+      "epoch": 0.012199881995189035,
+      "grad_norm": 1.0,
+      "learning_rate": 3.188405797101449e-05,
+      "loss": 1.532025933265686,
+      "step": 56
+    },
+    {
+      "epoch": 0.012417737030817411,
+      "grad_norm": 0.96484375,
+      "learning_rate": 3.246376811594203e-05,
+      "loss": 1.3564139604568481,
+      "step": 57
+    },
+    {
+      "epoch": 0.012635592066445786,
+      "grad_norm": 0.9609375,
+      "learning_rate": 3.304347826086957e-05,
+      "loss": 1.263455867767334,
+      "step": 58
+    },
+    {
+      "epoch": 0.012853447102074162,
+      "grad_norm": 1.0078125,
+      "learning_rate": 3.36231884057971e-05,
+      "loss": 1.3263088464736938,
+      "step": 59
+    },
+    {
+      "epoch": 0.013071302137702537,
+      "grad_norm": 1.0859375,
+      "learning_rate": 3.420289855072464e-05,
+      "loss": 1.47297203540802,
+      "step": 60
+    },
+    {
+      "epoch": 0.013289157173330913,
+      "grad_norm": 0.859375,
+      "learning_rate": 3.478260869565218e-05,
+      "loss": 1.1572554111480713,
+      "step": 61
+    },
+    {
+      "epoch": 0.013507012208959288,
+      "grad_norm": 0.90234375,
+      "learning_rate": 3.5362318840579716e-05,
+      "loss": 1.3723890781402588,
+      "step": 62
+    },
+    {
+      "epoch": 0.013724867244587664,
+      "grad_norm": 0.94921875,
+      "learning_rate": 3.594202898550725e-05,
+      "loss": 1.4838632345199585,
+      "step": 63
+    },
+    {
+      "epoch": 0.013942722280216039,
+      "grad_norm": 0.90234375,
+      "learning_rate": 3.6521739130434786e-05,
+      "loss": 1.4126598834991455,
+      "step": 64
+    },
+    {
+      "epoch": 0.014160577315844415,
+      "grad_norm": 0.94140625,
+      "learning_rate": 3.7101449275362325e-05,
+      "loss": 1.207360863685608,
+      "step": 65
+    },
+    {
+      "epoch": 0.01437843235147279,
+      "grad_norm": 0.97265625,
+      "learning_rate": 3.7681159420289856e-05,
+      "loss": 1.3931212425231934,
+      "step": 66
+    },
+    {
+      "epoch": 0.014596287387101167,
+      "grad_norm": 0.875,
+      "learning_rate": 3.8260869565217395e-05,
+      "loss": 1.3188490867614746,
+      "step": 67
+    },
+    {
+      "epoch": 0.014814142422729541,
+      "grad_norm": 0.93359375,
+      "learning_rate": 3.884057971014493e-05,
+      "loss": 1.3953640460968018,
+      "step": 68
+    },
+    {
+      "epoch": 0.015031997458357918,
+      "grad_norm": 1.0078125,
+      "learning_rate": 3.942028985507247e-05,
+      "loss": 1.4130436182022095,
+      "step": 69
+    },
+    {
+      "epoch": 0.015249852493986294,
+      "grad_norm": 0.9921875,
+      "learning_rate": 4e-05,
+      "loss": 1.3816159963607788,
+      "step": 70
+    },
+    {
+      "epoch": 0.015467707529614669,
+      "grad_norm": 0.8828125,
+      "learning_rate": 4.057971014492754e-05,
+      "loss": 1.351740837097168,
+      "step": 71
+    },
+    {
+      "epoch": 0.015685562565243043,
+      "grad_norm": 1.046875,
+      "learning_rate": 4.115942028985507e-05,
+      "loss": 1.0990729331970215,
+      "step": 72
+    },
+    {
+      "epoch": 0.01590341760087142,
+      "grad_norm": 0.93359375,
+      "learning_rate": 4.173913043478261e-05,
+      "loss": 1.4037866592407227,
+      "step": 73
+    },
+    {
+      "epoch": 0.016121272636499796,
+      "grad_norm": 0.9609375,
+      "learning_rate": 4.231884057971015e-05,
+      "loss": 1.2732292413711548,
+      "step": 74
+    },
+    {
+      "epoch": 0.01633912767212817,
+      "grad_norm": 1.046875,
+      "learning_rate": 4.289855072463769e-05,
+      "loss": 1.4371072053909302,
+      "step": 75
+    },
+    {
+      "epoch": 0.016556982707756546,
+      "grad_norm": 0.92578125,
+      "learning_rate": 4.347826086956522e-05,
+      "loss": 1.5245319604873657,
+      "step": 76
+    },
+    {
+      "epoch": 0.016774837743384924,
+      "grad_norm": 0.94140625,
+      "learning_rate": 4.405797101449276e-05,
+      "loss": 1.3343805074691772,
+      "step": 77
+    },
+    {
+      "epoch": 0.0169926927790133,
+      "grad_norm": 0.9375,
+      "learning_rate": 4.46376811594203e-05,
+      "loss": 1.2328696250915527,
+      "step": 78
+    },
+    {
+      "epoch": 0.017210547814641673,
+      "grad_norm": 0.87890625,
+      "learning_rate": 4.521739130434783e-05,
+      "loss": 1.1644716262817383,
+      "step": 79
+    },
+    {
+      "epoch": 0.017428402850270048,
+      "grad_norm": 0.984375,
+      "learning_rate": 4.579710144927537e-05,
+      "loss": 1.4006528854370117,
+      "step": 80
+    },
+    {
+      "epoch": 0.017646257885898426,
+      "grad_norm": 0.91796875,
+      "learning_rate": 4.6376811594202906e-05,
+      "loss": 1.2863035202026367,
+      "step": 81
+    },
+    {
+      "epoch": 0.0178641129215268,
+      "grad_norm": 0.890625,
+      "learning_rate": 4.6956521739130444e-05,
+      "loss": 1.4619357585906982,
+      "step": 82
+    },
+    {
+      "epoch": 0.018081967957155175,
+      "grad_norm": 1.0234375,
+      "learning_rate": 4.7536231884057976e-05,
+      "loss": 1.200344204902649,
+      "step": 83
+    },
+    {
+      "epoch": 0.018299822992783554,
+      "grad_norm": 0.98828125,
+      "learning_rate": 4.8115942028985514e-05,
+      "loss": 1.4004018306732178,
+      "step": 84
+    },
+    {
+      "epoch": 0.01851767802841193,
+      "grad_norm": 0.9453125,
+      "learning_rate": 4.869565217391305e-05,
+      "loss": 1.185855746269226,
+      "step": 85
+    },
+    {
+      "epoch": 0.018735533064040303,
+      "grad_norm": 1.015625,
+      "learning_rate": 4.927536231884058e-05,
+      "loss": 1.350454568862915,
+      "step": 86
+    },
+    {
+      "epoch": 0.018953388099668678,
+      "grad_norm": 0.9375,
+      "learning_rate": 4.9855072463768116e-05,
+      "loss": 1.483351707458496,
+      "step": 87
+    },
+    {
+      "epoch": 0.019171243135297056,
+      "grad_norm": 0.87890625,
+      "learning_rate": 5.043478260869566e-05,
+      "loss": 1.224565029144287,
+      "step": 88
+    },
+    {
+      "epoch": 0.01938909817092543,
+      "grad_norm": 0.96875,
+      "learning_rate": 5.1014492753623186e-05,
+      "loss": 1.316239595413208,
+      "step": 89
+    },
+    {
+      "epoch": 0.019606953206553805,
+      "grad_norm": 0.95703125,
+      "learning_rate": 5.1594202898550725e-05,
+      "loss": 1.3671209812164307,
+      "step": 90
+    },
+    {
+      "epoch": 0.01982480824218218,
+      "grad_norm": 0.94140625,
+      "learning_rate": 5.217391304347826e-05,
+      "loss": 1.3540199995040894,
+      "step": 91
+    },
+    {
+      "epoch": 0.020042663277810558,
+      "grad_norm": 0.921875,
+      "learning_rate": 5.27536231884058e-05,
+      "loss": 1.376987338066101,
+      "step": 92
+    },
+    {
+      "epoch": 0.020260518313438933,
+      "grad_norm": 1.0,
+      "learning_rate": 5.333333333333333e-05,
+      "loss": 1.3747578859329224,
+      "step": 93
+    },
+    {
+      "epoch": 0.020478373349067307,
+      "grad_norm": 0.94921875,
+      "learning_rate": 5.391304347826087e-05,
+      "loss": 1.3860173225402832,
+      "step": 94
+    },
+    {
+      "epoch": 0.020696228384695682,
+      "grad_norm": 1.1015625,
+      "learning_rate": 5.449275362318841e-05,
+      "loss": 1.2890279293060303,
+      "step": 95
+    },
+    {
+      "epoch": 0.02091408342032406,
+      "grad_norm": 1.0546875,
+      "learning_rate": 5.507246376811594e-05,
+      "loss": 1.5444395542144775,
+      "step": 96
+    },
+    {
+      "epoch": 0.021131938455952435,
+      "grad_norm": 0.8984375,
+      "learning_rate": 5.565217391304348e-05,
+      "loss": 1.249709129333496,
+      "step": 97
+    },
+    {
+      "epoch": 0.02134979349158081,
+      "grad_norm": 0.9140625,
+      "learning_rate": 5.623188405797102e-05,
+      "loss": 1.2482976913452148,
+      "step": 98
+    },
+    {
+      "epoch": 0.021567648527209188,
+      "grad_norm": 0.890625,
+      "learning_rate": 5.681159420289856e-05,
+      "loss": 1.22403085231781,
+      "step": 99
+    },
+    {
+      "epoch": 0.021785503562837562,
+      "grad_norm": 0.9765625,
+      "learning_rate": 5.739130434782609e-05,
+      "loss": 1.3967232704162598,
+      "step": 100
+    },
+    {
+      "epoch": 0.022003358598465937,
+      "grad_norm": 0.8984375,
+      "learning_rate": 5.797101449275363e-05,
+      "loss": 1.3915894031524658,
+      "step": 101
+    },
+    {
+      "epoch": 0.022221213634094312,
+      "grad_norm": 0.875,
+      "learning_rate": 5.8550724637681166e-05,
+      "loss": 1.2861565351486206,
+      "step": 102
+    },
+    {
+      "epoch": 0.02243906866972269,
+      "grad_norm": 0.9921875,
+      "learning_rate": 5.91304347826087e-05,
+      "loss": 1.289185643196106,
+      "step": 103
+    },
+    {
+      "epoch": 0.022656923705351065,
+      "grad_norm": 0.9921875,
+      "learning_rate": 5.9710144927536236e-05,
+      "loss": 1.4855297803878784,
+      "step": 104
+    },
+    {
+      "epoch": 0.02287477874097944,
+      "grad_norm": 0.9140625,
+      "learning_rate": 6.0289855072463774e-05,
+      "loss": 1.5117406845092773,
+      "step": 105
+    },
+    {
+      "epoch": 0.023092633776607814,
+      "grad_norm": 0.9765625,
+      "learning_rate": 6.086956521739131e-05,
+      "loss": 1.3688548803329468,
+      "step": 106
+    },
+    {
+      "epoch": 0.023310488812236192,
+      "grad_norm": 1.1171875,
+      "learning_rate": 6.144927536231884e-05,
+      "loss": 1.3791358470916748,
+      "step": 107
+    },
+    {
+      "epoch": 0.023528343847864567,
+      "grad_norm": 1.0234375,
+      "learning_rate": 6.202898550724638e-05,
+      "loss": 1.4389981031417847,
+      "step": 108
+    },
+    {
+      "epoch": 0.02374619888349294,
+      "grad_norm": 1.0234375,
+      "learning_rate": 6.260869565217392e-05,
+      "loss": 1.3525632619857788,
+      "step": 109
+    },
+    {
+      "epoch": 0.023964053919121316,
+      "grad_norm": 0.890625,
+      "learning_rate": 6.318840579710145e-05,
+      "loss": 1.384903073310852,
+      "step": 110
+    },
+    {
+      "epoch": 0.024181908954749694,
+      "grad_norm": 1.0234375,
+      "learning_rate": 6.376811594202898e-05,
+      "loss": 1.35524582862854,
+      "step": 111
+    },
+    {
+      "epoch": 0.02439976399037807,
+      "grad_norm": 1.0703125,
+      "learning_rate": 6.434782608695653e-05,
+      "loss": 1.3761764764785767,
+      "step": 112
+    },
+    {
+      "epoch": 0.024617619026006444,
+      "grad_norm": 0.8984375,
+      "learning_rate": 6.492753623188406e-05,
+      "loss": 1.4163299798965454,
+      "step": 113
+    },
+    {
+      "epoch": 0.024835474061634822,
+      "grad_norm": 1.140625,
+      "learning_rate": 6.550724637681159e-05,
+      "loss": 1.3786147832870483,
+      "step": 114
+    },
+    {
+      "epoch": 0.025053329097263197,
+      "grad_norm": 1.03125,
+      "learning_rate": 6.608695652173914e-05,
+      "loss": 1.3683019876480103,
+      "step": 115
+    },
+    {
+      "epoch": 0.02527118413289157,
+      "grad_norm": 0.96484375,
+      "learning_rate": 6.666666666666667e-05,
+      "loss": 1.218237280845642,
+      "step": 116
+    },
+    {
+      "epoch": 0.025489039168519946,
+      "grad_norm": 1.0703125,
+      "learning_rate": 6.72463768115942e-05,
+      "loss": 1.352725863456726,
+      "step": 117
+    },
+    {
+      "epoch": 0.025706894204148324,
+      "grad_norm": 0.96875,
+      "learning_rate": 6.782608695652175e-05,
+      "loss": 1.310965657234192,
+      "step": 118
+    },
+    {
+      "epoch": 0.0259247492397767,
+      "grad_norm": 0.86328125,
+      "learning_rate": 6.840579710144928e-05,
+      "loss": 1.263986587524414,
+      "step": 119
+    },
+    {
+      "epoch": 0.026142604275405074,
+      "grad_norm": 0.9140625,
+      "learning_rate": 6.898550724637682e-05,
+      "loss": 1.3614460229873657,
+      "step": 120
+    },
+    {
+      "epoch": 0.02636045931103345,
+      "grad_norm": 0.9375,
+      "learning_rate": 6.956521739130436e-05,
+      "loss": 1.3661404848098755,
+      "step": 121
+    },
+    {
+      "epoch": 0.026578314346661826,
+      "grad_norm": 0.83203125,
+      "learning_rate": 7.014492753623189e-05,
+      "loss": 1.2595158815383911,
+      "step": 122
+    },
+    {
+      "epoch": 0.0267961693822902,
+      "grad_norm": 0.83203125,
+      "learning_rate": 7.072463768115943e-05,
+      "loss": 1.1677026748657227,
+      "step": 123
+    },
+    {
+      "epoch": 0.027014024417918576,
+      "grad_norm": 0.953125,
+      "learning_rate": 7.130434782608696e-05,
+      "loss": 1.3778382539749146,
+      "step": 124
+    },
+    {
+      "epoch": 0.027231879453546954,
+      "grad_norm": 0.93359375,
+      "learning_rate": 7.18840579710145e-05,
+      "loss": 1.2247304916381836,
+      "step": 125
+    },
+    {
+      "epoch": 0.02744973448917533,
+      "grad_norm": 1.0,
+      "learning_rate": 7.246376811594204e-05,
+      "loss": 1.3398704528808594,
+      "step": 126
+    },
+    {
+      "epoch": 0.027667589524803703,
+      "grad_norm": 0.9296875,
+      "learning_rate": 7.304347826086957e-05,
+      "loss": 1.3399076461791992,
+      "step": 127
+    },
+    {
+      "epoch": 0.027885444560432078,
+      "grad_norm": 0.98828125,
+      "learning_rate": 7.36231884057971e-05,
+      "loss": 1.331055760383606,
+      "step": 128
+    },
+    {
+      "epoch": 0.028103299596060456,
+      "grad_norm": 1.046875,
+      "learning_rate": 7.420289855072465e-05,
+      "loss": 1.3057575225830078,
+      "step": 129
+    },
+    {
+      "epoch": 0.02832115463168883,
+      "grad_norm": 0.98046875,
+      "learning_rate": 7.478260869565218e-05,
+      "loss": 1.1866379976272583,
+      "step": 130
+    },
+    {
+      "epoch": 0.028539009667317206,
+      "grad_norm": 0.91015625,
+      "learning_rate": 7.536231884057971e-05,
+      "loss": 1.2767364978790283,
+      "step": 131
+    },
+    {
+      "epoch": 0.02875686470294558,
+      "grad_norm": 0.89453125,
+      "learning_rate": 7.594202898550726e-05,
+      "loss": 1.3441228866577148,
+      "step": 132
+    },
+    {
+      "epoch": 0.02897471973857396,
+      "grad_norm": 1.015625,
+      "learning_rate": 7.652173913043479e-05,
+      "loss": 1.55601167678833,
+      "step": 133
+    },
+    {
+      "epoch": 0.029192574774202333,
+      "grad_norm": 0.92578125,
+      "learning_rate": 7.710144927536232e-05,
+      "loss": 1.3233695030212402,
+      "step": 134
+    },
+    {
+      "epoch": 0.029410429809830708,
+      "grad_norm": 1.046875,
+      "learning_rate": 7.768115942028987e-05,
+      "loss": 1.416664481163025,
+      "step": 135
+    },
+    {
+      "epoch": 0.029628284845459082,
+      "grad_norm": 1.0546875,
+      "learning_rate": 7.82608695652174e-05,
+      "loss": 1.4259668588638306,
+      "step": 136
+    },
+    {
+      "epoch": 0.02984613988108746,
+      "grad_norm": 0.94921875,
+      "learning_rate": 7.884057971014494e-05,
+      "loss": 1.3535914421081543,
+      "step": 137
+    },
+    {
+      "epoch": 0.030063994916715835,
+      "grad_norm": 0.9375,
+      "learning_rate": 7.942028985507246e-05,
+      "loss": 1.3752613067626953,
+      "step": 138
+    },
+    {
+      "epoch": 0.03028184995234421,
+      "grad_norm": 0.9296875,
+      "learning_rate": 8e-05,
+      "loss": 1.2196537256240845,
+      "step": 139
+    },
+    {
+      "epoch": 0.030499704987972588,
+      "grad_norm": 0.98046875,
+      "learning_rate": 7.9999993707857e-05,
+      "loss": 1.2170228958129883,
+      "step": 140
+    },
+    {
+      "epoch": 0.030717560023600963,
+      "grad_norm": 0.9609375,
+      "learning_rate": 7.999997483142998e-05,
+      "loss": 1.258544683456421,
+      "step": 141
+    },
+    {
+      "epoch": 0.030935415059229338,
+      "grad_norm": 1.0078125,
+      "learning_rate": 7.999994337072486e-05,
+      "loss": 1.3800123929977417,
+      "step": 142
+    },
+    {
+      "epoch": 0.031153270094857712,
+      "grad_norm": 0.9375,
+      "learning_rate": 7.999989932575157e-05,
+      "loss": 1.468567967414856,
+      "step": 143
+    },
+    {
+      "epoch": 0.03137112513048609,
+      "grad_norm": 0.96875,
+      "learning_rate": 7.999984269652393e-05,
+      "loss": 1.2860223054885864,
+      "step": 144
+    },
+    {
+      "epoch": 0.031588980166114465,
+      "grad_norm": 0.9140625,
+      "learning_rate": 7.999977348305978e-05,
+      "loss": 1.2457164525985718,
+      "step": 145
+    },
+    {
+      "epoch": 0.03180683520174284,
+      "grad_norm": 0.88671875,
+      "learning_rate": 7.999969168538089e-05,
+      "loss": 1.3419649600982666,
+      "step": 146
+    },
+    {
+      "epoch": 0.032024690237371214,
+      "grad_norm": 0.91015625,
+      "learning_rate": 7.999959730351298e-05,
+      "loss": 1.3648418188095093,
+      "step": 147
+    },
+    {
+      "epoch": 0.03224254527299959,
+      "grad_norm": 0.83203125,
+      "learning_rate": 7.999949033748577e-05,
+      "loss": 1.161346197128296,
+      "step": 148
+    },
+    {
+      "epoch": 0.032460400308627964,
+      "grad_norm": 1.0546875,
+      "learning_rate": 7.99993707873329e-05,
+      "loss": 1.4876608848571777,
+      "step": 149
+    },
+    {
+      "epoch": 0.03267825534425634,
+      "grad_norm": 0.9921875,
+      "learning_rate": 7.999923865309198e-05,
+      "loss": 1.3237810134887695,
+      "step": 150
+    },
+    {
+      "epoch": 0.03289611037988472,
+      "grad_norm": 0.90625,
+      "learning_rate": 7.999909393480459e-05,
+      "loss": 1.277789831161499,
+      "step": 151
+    },
+    {
+      "epoch": 0.03311396541551309,
+      "grad_norm": 1.0703125,
+      "learning_rate": 7.999893663251623e-05,
+      "loss": 1.3773994445800781,
+      "step": 152
+    },
+    {
+      "epoch": 0.03333182045114147,
+      "grad_norm": 0.87109375,
+      "learning_rate": 7.999876674627642e-05,
+      "loss": 1.133509635925293,
+      "step": 153
+    },
+    {
+      "epoch": 0.03354967548676985,
+      "grad_norm": 0.890625,
+      "learning_rate": 7.99985842761386e-05,
+      "loss": 1.2682969570159912,
+      "step": 154
+    },
+    {
+      "epoch": 0.03376753052239822,
+      "grad_norm": 0.93359375,
+      "learning_rate": 7.999838922216015e-05,
+      "loss": 1.3067295551300049,
+      "step": 155
+    },
+    {
+      "epoch": 0.0339853855580266,
+      "grad_norm": 0.97265625,
+      "learning_rate": 7.999818158440248e-05,
+      "loss": 1.4196363687515259,
+      "step": 156
+    },
+    {
+      "epoch": 0.034203240593654975,
+      "grad_norm": 0.875,
+      "learning_rate": 7.999796136293087e-05,
+      "loss": 1.349758267402649,
+      "step": 157
+    },
+    {
+      "epoch": 0.034421095629283346,
+      "grad_norm": 0.91796875,
+      "learning_rate": 7.999772855781464e-05,
+      "loss": 1.2899991273880005,
+      "step": 158
+    },
+    {
+      "epoch": 0.034638950664911725,
+      "grad_norm": 1.34375,
+      "learning_rate": 7.999748316912702e-05,
+      "loss": 1.3930132389068604,
+      "step": 159
+    },
+    {
+      "epoch": 0.034856805700540096,
+      "grad_norm": 0.84765625,
+      "learning_rate": 7.999722519694521e-05,
+      "loss": 1.1908881664276123,
+      "step": 160
+    },
+    {
+      "epoch": 0.035074660736168474,
+      "grad_norm": 0.83203125,
+      "learning_rate": 7.999695464135036e-05,
+      "loss": 1.2571446895599365,
+      "step": 161
+    },
+    {
+      "epoch": 0.03529251577179685,
+      "grad_norm": 0.87890625,
+      "learning_rate": 7.99966715024276e-05,
+      "loss": 1.3681079149246216,
+      "step": 162
+    },
+    {
+      "epoch": 0.03551037080742522,
+      "grad_norm": 0.86328125,
+      "learning_rate": 7.999637578026602e-05,
+      "loss": 1.1920379400253296,
+      "step": 163
+    },
+    {
+      "epoch": 0.0357282258430536,
+      "grad_norm": 0.81640625,
+      "learning_rate": 7.999606747495862e-05,
+      "loss": 1.402422308921814,
+      "step": 164
+    },
+    {
+      "epoch": 0.03594608087868198,
+      "grad_norm": 0.96484375,
+      "learning_rate": 7.99957465866024e-05,
+      "loss": 1.268337607383728,
+      "step": 165
+    },
+    {
+      "epoch": 0.03616393591431035,
+      "grad_norm": 0.89453125,
+      "learning_rate": 7.999541311529837e-05,
+      "loss": 1.3225244283676147,
+      "step": 166
+    },
+    {
+      "epoch": 0.03638179094993873,
+      "grad_norm": 0.984375,
+      "learning_rate": 7.999506706115137e-05,
+      "loss": 1.3034924268722534,
+      "step": 167
+    },
+    {
+      "epoch": 0.03659964598556711,
+      "grad_norm": 0.88671875,
+      "learning_rate": 7.999470842427033e-05,
+      "loss": 1.3235634565353394,
+      "step": 168
+    },
+    {
+      "epoch": 0.03681750102119548,
+      "grad_norm": 0.859375,
+      "learning_rate": 7.999433720476804e-05,
+      "loss": 1.392399549484253,
+      "step": 169
+    },
+    {
+      "epoch": 0.03703535605682386,
+      "grad_norm": 0.87890625,
+      "learning_rate": 7.99939534027613e-05,
+      "loss": 1.4103503227233887,
+      "step": 170
+    },
+    {
+      "epoch": 0.03725321109245223,
+      "grad_norm": 0.94140625,
+      "learning_rate": 7.999355701837085e-05,
+      "loss": 1.3263332843780518,
+      "step": 171
+    },
+    {
+      "epoch": 0.037471066128080606,
+      "grad_norm": 0.90625,
+      "learning_rate": 7.999314805172141e-05,
+      "loss": 1.320260763168335,
+      "step": 172
+    },
+    {
+      "epoch": 0.037688921163708984,
+      "grad_norm": 0.8515625,
+      "learning_rate": 7.999272650294163e-05,
+      "loss": 1.382466197013855,
+      "step": 173
+    },
+    {
+      "epoch": 0.037906776199337355,
+      "grad_norm": 0.86328125,
+      "learning_rate": 7.999229237216416e-05,
+      "loss": 1.4541735649108887,
+      "step": 174
+    },
+    {
+      "epoch": 0.038124631234965733,
+      "grad_norm": 0.77734375,
+      "learning_rate": 7.999184565952555e-05,
+      "loss": 1.125183343887329,
+      "step": 175
+    },
+    {
+      "epoch": 0.03834248627059411,
+      "grad_norm": 1.1171875,
+      "learning_rate": 7.999138636516634e-05,
+      "loss": 1.353598713874817,
+      "step": 176
+    },
+    {
+      "epoch": 0.03856034130622248,
+      "grad_norm": 0.875,
+      "learning_rate": 7.999091448923106e-05,
+      "loss": 1.439039707183838,
+      "step": 177
+    },
+    {
+      "epoch": 0.03877819634185086,
+      "grad_norm": 0.91015625,
+      "learning_rate": 7.999043003186812e-05,
+      "loss": 1.4423664808273315,
+      "step": 178
+    },
+    {
+      "epoch": 0.03899605137747923,
+      "grad_norm": 0.921875,
+      "learning_rate": 7.998993299322997e-05,
+      "loss": 1.25596022605896,
+      "step": 179
+    },
+    {
+      "epoch": 0.03921390641310761,
+      "grad_norm": 0.8125,
+      "learning_rate": 7.998942337347295e-05,
+      "loss": 1.2615466117858887,
+      "step": 180
+    },
+    {
+      "epoch": 0.03943176144873599,
+      "grad_norm": 0.94140625,
+      "learning_rate": 7.998890117275741e-05,
+      "loss": 1.3491063117980957,
+      "step": 181
+    },
+    {
+      "epoch": 0.03964961648436436,
+      "grad_norm": 0.90234375,
+      "learning_rate": 7.998836639124766e-05,
+      "loss": 1.248676061630249,
+      "step": 182
+    },
+    {
+      "epoch": 0.03986747151999274,
+      "grad_norm": 0.90234375,
+      "learning_rate": 7.99878190291119e-05,
+      "loss": 1.3463438749313354,
+      "step": 183
+    },
+    {
+      "epoch": 0.040085326555621116,
+      "grad_norm": 0.84375,
+      "learning_rate": 7.998725908652236e-05,
+      "loss": 1.2867635488510132,
+      "step": 184
+    },
+    {
+      "epoch": 0.04030318159124949,
+      "grad_norm": 0.828125,
+      "learning_rate": 7.998668656365519e-05,
+      "loss": 1.3887321949005127,
+      "step": 185
+    },
+    {
+      "epoch": 0.040521036626877865,
+      "grad_norm": 0.84765625,
+      "learning_rate": 7.998610146069053e-05,
+      "loss": 1.4239572286605835,
+      "step": 186
+    },
+    {
+      "epoch": 0.040738891662506244,
+      "grad_norm": 0.78515625,
+      "learning_rate": 7.998550377781244e-05,
+      "loss": 1.2752599716186523,
+      "step": 187
+    },
+    {
+      "epoch": 0.040956746698134615,
+      "grad_norm": 0.7734375,
+      "learning_rate": 7.998489351520896e-05,
+      "loss": 1.172117829322815,
+      "step": 188
+    },
+    {
+      "epoch": 0.04117460173376299,
+      "grad_norm": 0.875,
+      "learning_rate": 7.998427067307209e-05,
+      "loss": 1.4123038053512573,
+      "step": 189
+    },
+    {
+      "epoch": 0.041392456769391364,
+      "grad_norm": 0.828125,
+      "learning_rate": 7.998363525159777e-05,
+      "loss": 1.3569304943084717,
+      "step": 190
+    },
+    {
+      "epoch": 0.04161031180501974,
+      "grad_norm": 0.81640625,
+      "learning_rate": 7.998298725098591e-05,
+      "loss": 1.4077844619750977,
+      "step": 191
+    },
+    {
+      "epoch": 0.04182816684064812,
+      "grad_norm": 0.7890625,
+      "learning_rate": 7.998232667144038e-05,
+      "loss": 1.2747725248336792,
+      "step": 192
+    },
+    {
+      "epoch": 0.04204602187627649,
+      "grad_norm": 0.8203125,
+      "learning_rate": 7.9981653513169e-05,
+      "loss": 1.3591704368591309,
+      "step": 193
+    },
+    {
+      "epoch": 0.04226387691190487,
+      "grad_norm": 0.78125,
+      "learning_rate": 7.998096777638354e-05,
+      "loss": 1.309944987297058,
+      "step": 194
+    },
+    {
+      "epoch": 0.04248173194753325,
+      "grad_norm": 0.7578125,
+      "learning_rate": 7.998026946129977e-05,
+      "loss": 1.1596174240112305,
+      "step": 195
+    },
+    {
+      "epoch": 0.04269958698316162,
+      "grad_norm": 0.890625,
+      "learning_rate": 7.997955856813737e-05,
+      "loss": 1.4131667613983154,
+      "step": 196
+    },
+    {
+      "epoch": 0.04291744201879,
+      "grad_norm": 0.828125,
+      "learning_rate": 7.997883509711995e-05,
+      "loss": 1.3620524406433105,
+      "step": 197
+    },
+    {
+      "epoch": 0.043135297054418376,
+      "grad_norm": 0.8828125,
+      "learning_rate": 7.997809904847519e-05,
+      "loss": 1.374614953994751,
+      "step": 198
+    },
+    {
+      "epoch": 0.04335315209004675,
+      "grad_norm": 0.921875,
+      "learning_rate": 7.99773504224346e-05,
+      "loss": 1.282700777053833,
+      "step": 199
+    },
+    {
+      "epoch": 0.043571007125675125,
+      "grad_norm": 0.69140625,
+      "learning_rate": 7.997658921923373e-05,
+      "loss": 1.2436307668685913,
+      "step": 200
+    },
+    {
+      "epoch": 0.043788862161303496,
+      "grad_norm": 0.81640625,
+      "learning_rate": 7.997581543911206e-05,
+      "loss": 1.2132856845855713,
+      "step": 201
+    },
+    {
+      "epoch": 0.044006717196931874,
+      "grad_norm": 0.9140625,
+      "learning_rate": 7.997502908231302e-05,
+      "loss": 1.2258870601654053,
+      "step": 202
+    },
+    {
+      "epoch": 0.04422457223256025,
+      "grad_norm": 0.8125,
+      "learning_rate": 7.9974230149084e-05,
+      "loss": 1.3195312023162842,
+      "step": 203
+    },
+    {
+      "epoch": 0.044442427268188624,
+      "grad_norm": 0.95703125,
+      "learning_rate": 7.997341863967635e-05,
+      "loss": 1.288048267364502,
+      "step": 204
+    },
+    {
+      "epoch": 0.044660282303817,
+      "grad_norm": 0.859375,
+      "learning_rate": 7.997259455434539e-05,
+      "loss": 1.5641199350357056,
+      "step": 205
+    },
+    {
+      "epoch": 0.04487813733944538,
+      "grad_norm": 0.734375,
+      "learning_rate": 7.997175789335037e-05,
+      "loss": 1.0514225959777832,
+      "step": 206
+    },
+    {
+      "epoch": 0.04509599237507375,
+      "grad_norm": 1.0703125,
+      "learning_rate": 7.997090865695452e-05,
+      "loss": 1.3750542402267456,
+      "step": 207
+    },
+    {
+      "epoch": 0.04531384741070213,
+      "grad_norm": 0.78125,
+      "learning_rate": 7.997004684542501e-05,
+      "loss": 1.2399580478668213,
+      "step": 208
+    },
+    {
+      "epoch": 0.04553170244633051,
+      "grad_norm": 0.953125,
+      "learning_rate": 7.996917245903297e-05,
+      "loss": 1.592616081237793,
+      "step": 209
+    },
+    {
+      "epoch": 0.04574955748195888,
+      "grad_norm": 1.1484375,
+      "learning_rate": 7.996828549805348e-05,
+      "loss": 1.3490384817123413,
+      "step": 210
+    },
+    {
+      "epoch": 0.04596741251758726,
+      "grad_norm": 0.78515625,
+      "learning_rate": 7.99673859627656e-05,
+      "loss": 1.3069775104522705,
+      "step": 211
+    },
+    {
+      "epoch": 0.04618526755321563,
+      "grad_norm": 0.83984375,
+      "learning_rate": 7.996647385345233e-05,
+      "loss": 1.4368929862976074,
+      "step": 212
+    },
+    {
+      "epoch": 0.046403122588844006,
+      "grad_norm": 0.82421875,
+      "learning_rate": 7.996554917040063e-05,
+      "loss": 1.3372946977615356,
+      "step": 213
+    },
+    {
+      "epoch": 0.046620977624472384,
+      "grad_norm": 0.8203125,
+      "learning_rate": 7.996461191390138e-05,
+      "loss": 1.2073931694030762,
+      "step": 214
+    },
+    {
+      "epoch": 0.046838832660100756,
+      "grad_norm": 0.9140625,
+      "learning_rate": 7.996366208424948e-05,
+      "loss": 1.249418020248413,
+      "step": 215
+    },
+    {
+      "epoch": 0.047056687695729134,
+      "grad_norm": 0.8125,
+      "learning_rate": 7.996269968174375e-05,
+      "loss": 1.3863368034362793,
+      "step": 216
+    },
+    {
+      "epoch": 0.04727454273135751,
+      "grad_norm": 1.0078125,
+      "learning_rate": 7.996172470668696e-05,
+      "loss": 1.1760282516479492,
+      "step": 217
+    },
+    {
+      "epoch": 0.04749239776698588,
+      "grad_norm": 0.88671875,
+      "learning_rate": 7.996073715938583e-05,
+      "loss": 1.3660634756088257,
+      "step": 218
+    },
+    {
+      "epoch": 0.04771025280261426,
+      "grad_norm": 0.76953125,
+      "learning_rate": 7.995973704015109e-05,
+      "loss": 1.254399299621582,
+      "step": 219
+    },
+    {
+      "epoch": 0.04792810783824263,
+      "grad_norm": 0.87109375,
+      "learning_rate": 7.995872434929734e-05,
+      "loss": 1.2892965078353882,
+      "step": 220
+    },
+    {
+      "epoch": 0.04814596287387101,
+      "grad_norm": 0.796875,
+      "learning_rate": 7.995769908714322e-05,
+      "loss": 1.2437909841537476,
+      "step": 221
+    },
+    {
+      "epoch": 0.04836381790949939,
+      "grad_norm": 0.8046875,
+      "learning_rate": 7.995666125401124e-05,
+      "loss": 1.4183120727539062,
+      "step": 222
+    },
+    {
+      "epoch": 0.04858167294512776,
+      "grad_norm": 0.8359375,
+      "learning_rate": 7.995561085022795e-05,
+      "loss": 1.2279728651046753,
+      "step": 223
+    },
+    {
+      "epoch": 0.04879952798075614,
+      "grad_norm": 0.828125,
+      "learning_rate": 7.99545478761238e-05,
+      "loss": 1.310274600982666,
+      "step": 224
+    },
+    {
+      "epoch": 0.049017383016384516,
+      "grad_norm": 0.9765625,
+      "learning_rate": 7.995347233203321e-05,
+      "loss": 1.218197226524353,
+      "step": 225
+    },
+    {
+      "epoch": 0.04923523805201289,
+      "grad_norm": 0.828125,
+      "learning_rate": 7.995238421829455e-05,
+      "loss": 1.2757151126861572,
+      "step": 226
+    },
+    {
+      "epoch": 0.049453093087641266,
+      "grad_norm": 0.80859375,
+      "learning_rate": 7.995128353525015e-05,
+      "loss": 1.3127573728561401,
+      "step": 227
+    },
+    {
+      "epoch": 0.049670948123269644,
+      "grad_norm": 0.828125,
+      "learning_rate": 7.995017028324629e-05,
+      "loss": 1.2553093433380127,
+      "step": 228
+    },
+    {
+      "epoch": 0.049888803158898015,
+      "grad_norm": 0.79296875,
+      "learning_rate": 7.994904446263322e-05,
+      "loss": 1.1580580472946167,
+      "step": 229
+    },
+    {
+      "epoch": 0.05010665819452639,
+      "grad_norm": 0.87890625,
+      "learning_rate": 7.99479060737651e-05,
+      "loss": 1.4943355321884155,
+      "step": 230
+    },
+    {
+      "epoch": 0.050324513230154765,
+      "grad_norm": 0.75,
+      "learning_rate": 7.994675511700012e-05,
+      "loss": 1.1570547819137573,
+      "step": 231
+    },
+    {
+      "epoch": 0.05054236826578314,
+      "grad_norm": 0.734375,
+      "learning_rate": 7.994559159270035e-05,
+      "loss": 1.3556218147277832,
+      "step": 232
+    },
+    {
+      "epoch": 0.05076022330141152,
+      "grad_norm": 0.84375,
+      "learning_rate": 7.994441550123186e-05,
+      "loss": 1.328752040863037,
+      "step": 233
+    },
+    {
+      "epoch": 0.05097807833703989,
+      "grad_norm": 0.82421875,
+      "learning_rate": 7.994322684296462e-05,
+      "loss": 1.405084252357483,
+      "step": 234
+    },
+    {
+      "epoch": 0.05119593337266827,
+      "grad_norm": 0.7890625,
+      "learning_rate": 7.994202561827263e-05,
+      "loss": 1.271226406097412,
+      "step": 235
+    },
+    {
+      "epoch": 0.05141378840829665,
+      "grad_norm": 0.828125,
+      "learning_rate": 7.994081182753378e-05,
+      "loss": 1.2796803712844849,
+      "step": 236
+    },
+    {
+      "epoch": 0.05163164344392502,
+      "grad_norm": 0.85546875,
+      "learning_rate": 7.993958547112997e-05,
+      "loss": 1.370160698890686,
+      "step": 237
+    },
+    {
+      "epoch": 0.0518494984795534,
+      "grad_norm": 0.7421875,
+      "learning_rate": 7.993834654944698e-05,
+      "loss": 1.3411636352539062,
+      "step": 238
+    },
+    {
+      "epoch": 0.052067353515181776,
+      "grad_norm": 0.8203125,
+      "learning_rate": 7.993709506287461e-05,
+      "loss": 1.2586621046066284,
+      "step": 239
+    },
+    {
+      "epoch": 0.05228520855081015,
+      "grad_norm": 0.83203125,
+      "learning_rate": 7.993583101180658e-05,
+      "loss": 1.2061572074890137,
+      "step": 240
+    },
+    {
+      "epoch": 0.052503063586438525,
+      "grad_norm": 0.88671875,
+      "learning_rate": 7.993455439664057e-05,
+      "loss": 1.3799080848693848,
+      "step": 241
+    },
+    {
+      "epoch": 0.0527209186220669,
+      "grad_norm": 0.8359375,
+      "learning_rate": 7.993326521777821e-05,
+      "loss": 1.2434254884719849,
+      "step": 242
+    },
+    {
+      "epoch": 0.052938773657695275,
+      "grad_norm": 0.73046875,
+      "learning_rate": 7.993196347562506e-05,
+      "loss": 1.273720145225525,
+      "step": 243
+    },
+    {
+      "epoch": 0.05315662869332365,
+      "grad_norm": 0.7578125,
+      "learning_rate": 7.993064917059071e-05,
+      "loss": 1.322296380996704,
+      "step": 244
+    },
+    {
+      "epoch": 0.053374483728952024,
+      "grad_norm": 0.8125,
+      "learning_rate": 7.992932230308862e-05,
+      "loss": 1.4182085990905762,
+      "step": 245
+    },
+    {
+      "epoch": 0.0535923387645804,
+      "grad_norm": 0.796875,
+      "learning_rate": 7.992798287353623e-05,
+      "loss": 1.19390869140625,
+      "step": 246
+    },
+    {
+      "epoch": 0.05381019380020878,
+      "grad_norm": 0.75,
+      "learning_rate": 7.992663088235494e-05,
+      "loss": 1.1668815612792969,
+      "step": 247
+    },
+    {
+      "epoch": 0.05402804883583715,
+      "grad_norm": 0.75390625,
+      "learning_rate": 7.992526632997009e-05,
+      "loss": 1.245719313621521,
+      "step": 248
+    },
+    {
+      "epoch": 0.05424590387146553,
+      "grad_norm": 0.80078125,
+      "learning_rate": 7.992388921681098e-05,
+      "loss": 1.236884593963623,
+      "step": 249
+    },
+    {
+      "epoch": 0.05446375890709391,
+      "grad_norm": 0.796875,
+      "learning_rate": 7.992249954331088e-05,
+      "loss": 1.2868388891220093,
+      "step": 250
+    },
+    {
+      "epoch": 0.05468161394272228,
+      "grad_norm": 0.7890625,
+      "learning_rate": 7.992109730990694e-05,
+      "loss": 1.4079562425613403,
+      "step": 251
+    },
+    {
+      "epoch": 0.05489946897835066,
+      "grad_norm": 0.80859375,
+      "learning_rate": 7.991968251704037e-05,
+      "loss": 1.2480814456939697,
+      "step": 252
+    },
+    {
+      "epoch": 0.05511732401397903,
+      "grad_norm": 0.79296875,
+      "learning_rate": 7.991825516515625e-05,
+      "loss": 1.2520029544830322,
+      "step": 253
+    },
+    {
+      "epoch": 0.05533517904960741,
+      "grad_norm": 0.74609375,
+      "learning_rate": 7.991681525470361e-05,
+      "loss": 1.2378876209259033,
+      "step": 254
+    },
+    {
+      "epoch": 0.055553034085235785,
+      "grad_norm": 0.84765625,
+      "learning_rate": 7.99153627861355e-05,
+      "loss": 1.3548156023025513,
+      "step": 255
+    },
+    {
+      "epoch": 0.055770889120864156,
+      "grad_norm": 0.87890625,
+      "learning_rate": 7.991389775990886e-05,
+      "loss": 1.3737314939498901,
+      "step": 256
+    },
+    {
+      "epoch": 0.055988744156492534,
+      "grad_norm": 0.7265625,
+      "learning_rate": 7.99124201764846e-05,
+      "loss": 1.2327758073806763,
+      "step": 257
+    },
+    {
+      "epoch": 0.05620659919212091,
+      "grad_norm": 0.8125,
+      "learning_rate": 7.991093003632756e-05,
+      "loss": 1.3469085693359375,
+      "step": 258
+    },
+    {
+      "epoch": 0.056424454227749284,
+      "grad_norm": 0.73828125,
+      "learning_rate": 7.990942733990658e-05,
+      "loss": 1.2434908151626587,
+      "step": 259
+    },
+    {
+      "epoch": 0.05664230926337766,
+      "grad_norm": 0.73828125,
+      "learning_rate": 7.99079120876944e-05,
+      "loss": 1.2233469486236572,
+      "step": 260
+    },
+    {
+      "epoch": 0.05686016429900603,
+      "grad_norm": 0.76953125,
+      "learning_rate": 7.990638428016772e-05,
+      "loss": 1.3632105588912964,
+      "step": 261
+    },
+    {
+      "epoch": 0.05707801933463441,
+      "grad_norm": 0.74609375,
+      "learning_rate": 7.99048439178072e-05,
+      "loss": 1.3060120344161987,
+      "step": 262
+    },
+    {
+      "epoch": 0.05729587437026279,
+      "grad_norm": 0.7890625,
+      "learning_rate": 7.990329100109748e-05,
+      "loss": 1.2777645587921143,
+      "step": 263
+    },
+    {
+      "epoch": 0.05751372940589116,
+      "grad_norm": 0.81640625,
+      "learning_rate": 7.990172553052709e-05,
+      "loss": 1.2903178930282593,
+      "step": 264
+    },
+    {
+      "epoch": 0.05773158444151954,
+      "grad_norm": 0.78125,
+      "learning_rate": 7.990014750658856e-05,
+      "loss": 1.3590680360794067,
+      "step": 265
+    },
+    {
+      "epoch": 0.05794943947714792,
+      "grad_norm": 0.76953125,
+      "learning_rate": 7.989855692977831e-05,
+      "loss": 1.3560243844985962,
+      "step": 266
+    },
+    {
+      "epoch": 0.05816729451277629,
+      "grad_norm": 0.73828125,
+      "learning_rate": 7.989695380059678e-05,
+      "loss": 1.1870315074920654,
+      "step": 267
+    },
+    {
+      "epoch": 0.058385149548404666,
+      "grad_norm": 0.796875,
+      "learning_rate": 7.989533811954832e-05,
+      "loss": 1.3784286975860596,
+      "step": 268
+    },
+    {
+      "epoch": 0.058603004584033044,
+      "grad_norm": 0.921875,
+      "learning_rate": 7.989370988714122e-05,
+      "loss": 1.415964961051941,
+      "step": 269
+    },
+    {
+      "epoch": 0.058820859619661416,
+      "grad_norm": 0.83203125,
+      "learning_rate": 7.989206910388776e-05,
+      "loss": 1.406625747680664,
+      "step": 270
+    },
+    {
+      "epoch": 0.059038714655289794,
+      "grad_norm": 0.78515625,
+      "learning_rate": 7.989041577030411e-05,
+      "loss": 1.3986737728118896,
+      "step": 271
+    },
+    {
+      "epoch": 0.059256569690918165,
+      "grad_norm": 0.8125,
+      "learning_rate": 7.988874988691045e-05,
+      "loss": 1.4400463104248047,
+      "step": 272
+    },
+    {
+      "epoch": 0.05947442472654654,
+      "grad_norm": 0.80078125,
+      "learning_rate": 7.988707145423087e-05,
+      "loss": 1.2225327491760254,
+      "step": 273
+    },
+    {
+      "epoch": 0.05969227976217492,
+      "grad_norm": 0.79296875,
+      "learning_rate": 7.98853804727934e-05,
+      "loss": 1.175416111946106,
+      "step": 274
+    },
+    {
+      "epoch": 0.05991013479780329,
+      "grad_norm": 0.8515625,
+      "learning_rate": 7.988367694313006e-05,
+      "loss": 1.3557404279708862,
+      "step": 275
+    },
+    {
+      "epoch": 0.06012798983343167,
+      "grad_norm": 0.7578125,
+      "learning_rate": 7.988196086577677e-05,
+      "loss": 1.1438641548156738,
+      "step": 276
+    },
+    {
+      "epoch": 0.06034584486906005,
+      "grad_norm": 0.859375,
+      "learning_rate": 7.988023224127345e-05,
+      "loss": 1.2602871656417847,
+      "step": 277
+    },
+    {
+      "epoch": 0.06056369990468842,
+      "grad_norm": 0.7890625,
+      "learning_rate": 7.98784910701639e-05,
+      "loss": 1.2581192255020142,
+      "step": 278
+    },
+    {
+      "epoch": 0.0607815549403168,
+      "grad_norm": 0.77734375,
+      "learning_rate": 7.987673735299594e-05,
+      "loss": 1.3195080757141113,
+      "step": 279
+    },
+    {
+      "epoch": 0.060999409975945176,
+      "grad_norm": 0.79296875,
+      "learning_rate": 7.987497109032128e-05,
+      "loss": 1.2750706672668457,
+      "step": 280
+    },
+    {
+      "epoch": 0.06121726501157355,
+      "grad_norm": 0.80859375,
+      "learning_rate": 7.987319228269558e-05,
+      "loss": 1.2139368057250977,
+      "step": 281
+    },
+    {
+      "epoch": 0.061435120047201926,
+      "grad_norm": 0.74609375,
+      "learning_rate": 7.987140093067852e-05,
+      "loss": 1.370411992073059,
+      "step": 282
+    },
+    {
+      "epoch": 0.0616529750828303,
+      "grad_norm": 0.7578125,
+      "learning_rate": 7.986959703483363e-05,
+      "loss": 1.2239960432052612,
+      "step": 283
+    },
+    {
+      "epoch": 0.061870830118458675,
+      "grad_norm": 0.7734375,
+      "learning_rate": 7.986778059572845e-05,
+      "loss": 1.2411049604415894,
+      "step": 284
+    },
+    {
+      "epoch": 0.06208868515408705,
+      "grad_norm": 0.75,
+      "learning_rate": 7.986595161393442e-05,
+      "loss": 1.3169654607772827,
+      "step": 285
+    },
+    {
+      "epoch": 0.062306540189715424,
+      "grad_norm": 0.875,
+      "learning_rate": 7.986411009002697e-05,
+      "loss": 1.1896905899047852,
+      "step": 286
+    },
+    {
+      "epoch": 0.0625243952253438,
+      "grad_norm": 0.7578125,
+      "learning_rate": 7.986225602458546e-05,
+      "loss": 1.2595183849334717,
+      "step": 287
+    },
+    {
+      "epoch": 0.06274225026097217,
+      "grad_norm": 0.7578125,
+      "learning_rate": 7.986038941819318e-05,
+      "loss": 1.371169924736023,
+      "step": 288
+    },
+    {
+      "epoch": 0.06296010529660055,
+      "grad_norm": 0.8125,
+      "learning_rate": 7.985851027143739e-05,
+      "loss": 1.3598040342330933,
+      "step": 289
+    },
+    {
+      "epoch": 0.06317796033222893,
+      "grad_norm": 0.84765625,
+      "learning_rate": 7.985661858490925e-05,
+      "loss": 1.414171576499939,
+      "step": 290
+    },
+    {
+      "epoch": 0.06339581536785731,
+      "grad_norm": 0.7578125,
+      "learning_rate": 7.985471435920396e-05,
+      "loss": 1.3129502534866333,
+      "step": 291
+    },
+    {
+      "epoch": 0.06361367040348569,
+      "grad_norm": 0.81640625,
+      "learning_rate": 7.985279759492054e-05,
+      "loss": 1.1789436340332031,
+      "step": 292
+    },
+    {
+      "epoch": 0.06383152543911405,
+      "grad_norm": 0.828125,
+      "learning_rate": 7.985086829266206e-05,
+      "loss": 1.3992222547531128,
+      "step": 293
+    },
+    {
+      "epoch": 0.06404938047474243,
+      "grad_norm": 0.9296875,
+      "learning_rate": 7.984892645303546e-05,
+      "loss": 1.3122252225875854,
+      "step": 294
+    },
+    {
+      "epoch": 0.06426723551037081,
+      "grad_norm": 0.78515625,
+      "learning_rate": 7.98469720766517e-05,
+      "loss": 1.3924921751022339,
+      "step": 295
+    },
+    {
+      "epoch": 0.06448509054599919,
+      "grad_norm": 0.80078125,
+      "learning_rate": 7.984500516412559e-05,
+      "loss": 1.207013487815857,
+      "step": 296
+    },
+    {
+      "epoch": 0.06470294558162756,
+      "grad_norm": 0.79296875,
+      "learning_rate": 7.984302571607597e-05,
+      "loss": 1.3166091442108154,
+      "step": 297
+    },
+    {
+      "epoch": 0.06492080061725593,
+      "grad_norm": 0.87890625,
+      "learning_rate": 7.984103373312558e-05,
+      "loss": 1.4074251651763916,
+      "step": 298
+    },
+    {
+      "epoch": 0.0651386556528843,
+      "grad_norm": 0.76953125,
+      "learning_rate": 7.983902921590111e-05,
+      "loss": 1.423839807510376,
+      "step": 299
+    },
+    {
+      "epoch": 0.06535651068851268,
+      "grad_norm": 0.74609375,
+      "learning_rate": 7.983701216503317e-05,
+      "loss": 1.151694655418396,
+      "step": 300
+    },
+    {
+      "epoch": 0.06557436572414106,
+      "grad_norm": 0.84375,
+      "learning_rate": 7.98349825811564e-05,
+      "loss": 1.3391306400299072,
+      "step": 301
+    },
+    {
+      "epoch": 0.06579222075976944,
+      "grad_norm": 0.75390625,
+      "learning_rate": 7.983294046490927e-05,
+      "loss": 1.281900405883789,
+      "step": 302
+    },
+    {
+      "epoch": 0.06601007579539782,
+      "grad_norm": 0.75,
+      "learning_rate": 7.983088581693425e-05,
+      "loss": 1.325886607170105,
+      "step": 303
+    },
+    {
+      "epoch": 0.06622793083102618,
+      "grad_norm": 0.83203125,
+      "learning_rate": 7.982881863787776e-05,
+      "loss": 1.236878514289856,
+      "step": 304
+    },
+    {
+      "epoch": 0.06644578586665456,
+      "grad_norm": 0.80078125,
+      "learning_rate": 7.982673892839015e-05,
+      "loss": 1.301095724105835,
+      "step": 305
+    },
+    {
+      "epoch": 0.06666364090228294,
+      "grad_norm": 0.70703125,
+      "learning_rate": 7.982464668912571e-05,
+      "loss": 1.1789932250976562,
+      "step": 306
+    },
+    {
+      "epoch": 0.06688149593791132,
+      "grad_norm": 0.75,
+      "learning_rate": 7.982254192074266e-05,
+      "loss": 1.2153866291046143,
+      "step": 307
+    },
+    {
+      "epoch": 0.0670993509735397,
+      "grad_norm": 0.82421875,
+      "learning_rate": 7.982042462390321e-05,
+      "loss": 1.3160028457641602,
+      "step": 308
+    },
+    {
+      "epoch": 0.06731720600916806,
+      "grad_norm": 0.8203125,
+      "learning_rate": 7.981829479927343e-05,
+      "loss": 1.3893190622329712,
+      "step": 309
+    },
+    {
+      "epoch": 0.06753506104479644,
+      "grad_norm": 0.7265625,
+      "learning_rate": 7.98161524475234e-05,
+      "loss": 1.1918959617614746,
+      "step": 310
+    },
+    {
+      "epoch": 0.06775291608042482,
+      "grad_norm": 0.765625,
+      "learning_rate": 7.981399756932713e-05,
+      "loss": 1.187642216682434,
+      "step": 311
+    },
+    {
+      "epoch": 0.0679707711160532,
+      "grad_norm": 0.7421875,
+      "learning_rate": 7.981183016536256e-05,
+      "loss": 1.4466753005981445,
+      "step": 312
+    },
+    {
+      "epoch": 0.06818862615168157,
+      "grad_norm": 0.78125,
+      "learning_rate": 7.980965023631154e-05,
+      "loss": 1.298957109451294,
+      "step": 313
+    },
+    {
+      "epoch": 0.06840648118730995,
+      "grad_norm": 0.7421875,
+      "learning_rate": 7.980745778285992e-05,
+      "loss": 1.5032776594161987,
+      "step": 314
+    },
+    {
+      "epoch": 0.06862433622293831,
+      "grad_norm": 0.73828125,
+      "learning_rate": 7.980525280569746e-05,
+      "loss": 1.327757477760315,
+      "step": 315
+    },
+    {
+      "epoch": 0.06884219125856669,
+      "grad_norm": 0.75390625,
+      "learning_rate": 7.980303530551784e-05,
+      "loss": 1.3077589273452759,
+      "step": 316
+    },
+    {
+      "epoch": 0.06906004629419507,
+      "grad_norm": 0.8046875,
+      "learning_rate": 7.980080528301874e-05,
+      "loss": 1.3015737533569336,
+      "step": 317
+    },
+    {
+      "epoch": 0.06927790132982345,
+      "grad_norm": 0.734375,
+      "learning_rate": 7.97985627389017e-05,
+      "loss": 1.267089605331421,
+      "step": 318
+    },
+    {
+      "epoch": 0.06949575636545183,
+      "grad_norm": 0.7421875,
+      "learning_rate": 7.979630767387225e-05,
+      "loss": 1.2797727584838867,
+      "step": 319
+    },
+    {
+      "epoch": 0.06971361140108019,
+      "grad_norm": 0.75390625,
+      "learning_rate": 7.979404008863988e-05,
+      "loss": 1.3705174922943115,
+      "step": 320
+    },
+    {
+      "epoch": 0.06993146643670857,
+      "grad_norm": 0.7421875,
+      "learning_rate": 7.979175998391794e-05,
+      "loss": 1.3233641386032104,
+      "step": 321
+    },
+    {
+      "epoch": 0.07014932147233695,
+      "grad_norm": 3.203125,
+      "learning_rate": 7.978946736042382e-05,
+      "loss": 1.411566972732544,
+      "step": 322
+    },
+    {
+      "epoch": 0.07036717650796533,
+      "grad_norm": 0.8046875,
+      "learning_rate": 7.978716221887874e-05,
+      "loss": 1.2296078205108643,
+      "step": 323
+    },
+    {
+      "epoch": 0.0705850315435937,
+      "grad_norm": 0.73046875,
+      "learning_rate": 7.978484456000797e-05,
+      "loss": 1.2896935939788818,
+      "step": 324
+    },
+    {
+      "epoch": 0.07080288657922208,
+      "grad_norm": 0.734375,
+      "learning_rate": 7.978251438454063e-05,
+      "loss": 1.2631288766860962,
+      "step": 325
+    },
+    {
+      "epoch": 0.07102074161485045,
+      "grad_norm": 0.71484375,
+      "learning_rate": 7.97801716932098e-05,
+      "loss": 1.3236161470413208,
+      "step": 326
+    },
+    {
+      "epoch": 0.07123859665047882,
+      "grad_norm": 0.73828125,
+      "learning_rate": 7.977781648675254e-05,
+      "loss": 1.321068286895752,
+      "step": 327
+    },
+    {
+      "epoch": 0.0714564516861072,
+      "grad_norm": 0.7734375,
+      "learning_rate": 7.97754487659098e-05,
+      "loss": 1.322007417678833,
+      "step": 328
+    },
+    {
+      "epoch": 0.07167430672173558,
+      "grad_norm": 0.76171875,
+      "learning_rate": 7.977306853142646e-05,
+      "loss": 1.374714732170105,
+      "step": 329
+    },
+    {
+      "epoch": 0.07189216175736396,
+      "grad_norm": 0.7265625,
+      "learning_rate": 7.977067578405139e-05,
+      "loss": 1.2924494743347168,
+      "step": 330
+    },
+    {
+      "epoch": 0.07211001679299232,
+      "grad_norm": 0.765625,
+      "learning_rate": 7.976827052453734e-05,
+      "loss": 1.3249881267547607,
+      "step": 331
+    },
+    {
+      "epoch": 0.0723278718286207,
+      "grad_norm": 0.765625,
+      "learning_rate": 7.976585275364105e-05,
+      "loss": 1.2810444831848145,
+      "step": 332
+    },
+    {
+      "epoch": 0.07254572686424908,
+      "grad_norm": 0.7734375,
+      "learning_rate": 7.976342247212315e-05,
+      "loss": 1.3198928833007812,
+      "step": 333
+    },
+    {
+      "epoch": 0.07276358189987746,
+      "grad_norm": 0.77734375,
+      "learning_rate": 7.976097968074822e-05,
+      "loss": 1.3211297988891602,
+      "step": 334
+    },
+    {
+      "epoch": 0.07298143693550584,
+      "grad_norm": 0.80078125,
+      "learning_rate": 7.975852438028479e-05,
+      "loss": 1.3993768692016602,
+      "step": 335
+    },
+    {
+      "epoch": 0.07319929197113421,
+      "grad_norm": 0.8515625,
+      "learning_rate": 7.975605657150533e-05,
+      "loss": 1.515588402748108,
+      "step": 336
+    },
+    {
+      "epoch": 0.07341714700676258,
+      "grad_norm": 0.7421875,
+      "learning_rate": 7.97535762551862e-05,
+      "loss": 1.190734624862671,
+      "step": 337
+    },
+    {
+      "epoch": 0.07363500204239096,
+      "grad_norm": 0.75390625,
+      "learning_rate": 7.975108343210773e-05,
+      "loss": 1.2638881206512451,
+      "step": 338
+    },
+    {
+      "epoch": 0.07385285707801933,
+      "grad_norm": 0.73828125,
+      "learning_rate": 7.974857810305419e-05,
+      "loss": 1.2987593412399292,
+      "step": 339
+    },
+    {
+      "epoch": 0.07407071211364771,
+      "grad_norm": 0.78125,
+      "learning_rate": 7.974606026881378e-05,
+      "loss": 1.3124685287475586,
+      "step": 340
+    },
+    {
+      "epoch": 0.07428856714927609,
+      "grad_norm": 0.7578125,
+      "learning_rate": 7.974352993017861e-05,
+      "loss": 1.354876160621643,
+      "step": 341
+    },
+    {
+      "epoch": 0.07450642218490446,
+      "grad_norm": 0.75390625,
+      "learning_rate": 7.974098708794475e-05,
+      "loss": 1.3209162950515747,
+      "step": 342
+    },
+    {
+      "epoch": 0.07472427722053283,
+      "grad_norm": 0.8515625,
+      "learning_rate": 7.97384317429122e-05,
+      "loss": 1.3101015090942383,
+      "step": 343
+    },
+    {
+      "epoch": 0.07494213225616121,
+      "grad_norm": 0.8046875,
+      "learning_rate": 7.973586389588492e-05,
+      "loss": 1.1449381113052368,
+      "step": 344
+    },
+    {
+      "epoch": 0.07515998729178959,
+      "grad_norm": 0.73828125,
+      "learning_rate": 7.97332835476707e-05,
+      "loss": 1.2249360084533691,
+      "step": 345
+    },
+    {
+      "epoch": 0.07537784232741797,
+      "grad_norm": 0.7734375,
+      "learning_rate": 7.97306906990814e-05,
+      "loss": 1.265030860900879,
+      "step": 346
+    },
+    {
+      "epoch": 0.07559569736304635,
+      "grad_norm": 0.76953125,
+      "learning_rate": 7.972808535093272e-05,
+      "loss": 1.3013360500335693,
+      "step": 347
+    },
+    {
+      "epoch": 0.07581355239867471,
+      "grad_norm": 0.79296875,
+      "learning_rate": 7.972546750404433e-05,
+      "loss": 1.2092686891555786,
+      "step": 348
+    },
+    {
+      "epoch": 0.07603140743430309,
+      "grad_norm": 0.79296875,
+      "learning_rate": 7.972283715923983e-05,
+      "loss": 1.3573532104492188,
+      "step": 349
+    },
+    {
+      "epoch": 0.07624926246993147,
+      "grad_norm": 0.79296875,
+      "learning_rate": 7.972019431734673e-05,
+      "loss": 1.2845659255981445,
+      "step": 350
+    },
+    {
+      "epoch": 0.07646711750555985,
+      "grad_norm": 0.703125,
+      "learning_rate": 7.97175389791965e-05,
+      "loss": 1.2901146411895752,
+      "step": 351
+    },
+    {
+      "epoch": 0.07668497254118822,
+      "grad_norm": 0.76171875,
+      "learning_rate": 7.971487114562451e-05,
+      "loss": 1.1869018077850342,
+      "step": 352
+    },
+    {
+      "epoch": 0.07690282757681659,
+      "grad_norm": 0.72265625,
+      "learning_rate": 7.97121908174701e-05,
+      "loss": 1.1580688953399658,
+      "step": 353
+    },
+    {
+      "epoch": 0.07712068261244497,
+      "grad_norm": 0.75,
+      "learning_rate": 7.97094979955765e-05,
+      "loss": 1.206789255142212,
+      "step": 354
+    },
+    {
+      "epoch": 0.07733853764807334,
+      "grad_norm": 0.7421875,
+      "learning_rate": 7.970679268079092e-05,
+      "loss": 1.2828707695007324,
+      "step": 355
+    },
+    {
+      "epoch": 0.07755639268370172,
+      "grad_norm": 0.71875,
+      "learning_rate": 7.970407487396446e-05,
+      "loss": 1.2693021297454834,
+      "step": 356
+    },
+    {
+      "epoch": 0.0777742477193301,
+      "grad_norm": 0.80078125,
+      "learning_rate": 7.970134457595214e-05,
+      "loss": 1.3986281156539917,
+      "step": 357
+    },
+    {
+      "epoch": 0.07799210275495846,
+      "grad_norm": 0.75,
+      "learning_rate": 7.969860178761294e-05,
+      "loss": 1.282415747642517,
+      "step": 358
+    },
+    {
+      "epoch": 0.07820995779058684,
+      "grad_norm": 0.73828125,
+      "learning_rate": 7.969584650980977e-05,
+      "loss": 1.2725651264190674,
+      "step": 359
+    },
+    {
+      "epoch": 0.07842781282621522,
+      "grad_norm": 0.76171875,
+      "learning_rate": 7.969307874340948e-05,
+      "loss": 1.1507782936096191,
+      "step": 360
+    },
+    {
+      "epoch": 0.0786456678618436,
+      "grad_norm": 0.69921875,
+      "learning_rate": 7.969029848928279e-05,
+      "loss": 1.1353940963745117,
+      "step": 361
+    },
+    {
+      "epoch": 0.07886352289747198,
+      "grad_norm": 0.73046875,
+      "learning_rate": 7.968750574830441e-05,
+      "loss": 1.2525478601455688,
+      "step": 362
+    },
+    {
+      "epoch": 0.07908137793310036,
+      "grad_norm": 0.67578125,
+      "learning_rate": 7.968470052135295e-05,
+      "loss": 1.2067474126815796,
+      "step": 363
+    },
+    {
+      "epoch": 0.07929923296872872,
+      "grad_norm": 0.765625,
+      "learning_rate": 7.968188280931095e-05,
+      "loss": 1.4495424032211304,
+      "step": 364
+    },
+    {
+      "epoch": 0.0795170880043571,
+      "grad_norm": 0.7109375,
+      "learning_rate": 7.96790526130649e-05,
+      "loss": 1.2542188167572021,
+      "step": 365
+    },
+    {
+      "epoch": 0.07973494303998548,
+      "grad_norm": 0.96875,
+      "learning_rate": 7.967620993350518e-05,
+      "loss": 1.2136855125427246,
+      "step": 366
+    },
+    {
+      "epoch": 0.07995279807561385,
+      "grad_norm": 0.7265625,
+      "learning_rate": 7.967335477152613e-05,
+      "loss": 1.2022827863693237,
+      "step": 367
+    },
+    {
+      "epoch": 0.08017065311124223,
+      "grad_norm": 0.7265625,
+      "learning_rate": 7.9670487128026e-05,
+      "loss": 1.2343459129333496,
+      "step": 368
+    },
+    {
+      "epoch": 0.0803885081468706,
+      "grad_norm": 0.78515625,
+      "learning_rate": 7.966760700390696e-05,
+      "loss": 1.4966288805007935,
+      "step": 369
+    },
+    {
+      "epoch": 0.08060636318249897,
+      "grad_norm": 0.7734375,
+      "learning_rate": 7.966471440007514e-05,
+      "loss": 1.2547136545181274,
+      "step": 370
+    },
+    {
+      "epoch": 0.08082421821812735,
+      "grad_norm": 0.70703125,
+      "learning_rate": 7.966180931744056e-05,
+      "loss": 1.2171647548675537,
+      "step": 371
+    },
+    {
+      "epoch": 0.08104207325375573,
+      "grad_norm": 0.73046875,
+      "learning_rate": 7.965889175691719e-05,
+      "loss": 1.359579086303711,
+      "step": 372
+    },
+    {
+      "epoch": 0.08125992828938411,
+      "grad_norm": 0.75,
+      "learning_rate": 7.965596171942288e-05,
+      "loss": 1.2641489505767822,
+      "step": 373
+    },
+    {
+      "epoch": 0.08147778332501249,
+      "grad_norm": 0.7890625,
+      "learning_rate": 7.96530192058795e-05,
+      "loss": 1.413927435874939,
+      "step": 374
+    },
+    {
+      "epoch": 0.08169563836064085,
+      "grad_norm": 0.71484375,
+      "learning_rate": 7.965006421721274e-05,
+      "loss": 1.186339259147644,
+      "step": 375
+    },
+    {
+      "epoch": 0.08191349339626923,
+      "grad_norm": 0.78125,
+      "learning_rate": 7.964709675435227e-05,
+      "loss": 1.43724524974823,
+      "step": 376
+    },
+    {
+      "epoch": 0.08213134843189761,
+      "grad_norm": 0.80078125,
+      "learning_rate": 7.96441168182317e-05,
+      "loss": 1.3339821100234985,
+      "step": 377
+    },
+    {
+      "epoch": 0.08234920346752599,
+      "grad_norm": 0.734375,
+      "learning_rate": 7.96411244097885e-05,
+      "loss": 1.2389692068099976,
+      "step": 378
+    },
+    {
+      "epoch": 0.08256705850315436,
+      "grad_norm": 0.765625,
+      "learning_rate": 7.963811952996412e-05,
+      "loss": 1.3433022499084473,
+      "step": 379
+    },
+    {
+      "epoch": 0.08278491353878273,
+      "grad_norm": 0.875,
+      "learning_rate": 7.963510217970393e-05,
+      "loss": 1.3269431591033936,
+      "step": 380
+    },
+    {
+      "epoch": 0.0830027685744111,
+      "grad_norm": 0.73046875,
+      "learning_rate": 7.963207235995719e-05,
+      "loss": 1.2242637872695923,
+      "step": 381
+    },
+    {
+      "epoch": 0.08322062361003948,
+      "grad_norm": 0.68359375,
+      "learning_rate": 7.962903007167712e-05,
+      "loss": 1.225073218345642,
+      "step": 382
+    },
+    {
+      "epoch": 0.08343847864566786,
+      "grad_norm": 0.82421875,
+      "learning_rate": 7.962597531582083e-05,
+      "loss": 1.1422075033187866,
+      "step": 383
+    },
+    {
+      "epoch": 0.08365633368129624,
+      "grad_norm": 0.74609375,
+      "learning_rate": 7.962290809334937e-05,
+      "loss": 1.3883594274520874,
+      "step": 384
+    },
+    {
+      "epoch": 0.08387418871692462,
+      "grad_norm": 0.7109375,
+      "learning_rate": 7.961982840522773e-05,
+      "loss": 1.1966588497161865,
+      "step": 385
+    },
+    {
+      "epoch": 0.08409204375255298,
+      "grad_norm": 0.82421875,
+      "learning_rate": 7.961673625242478e-05,
+      "loss": 1.3808834552764893,
+      "step": 386
+    },
+    {
+      "epoch": 0.08430989878818136,
+      "grad_norm": 0.7578125,
+      "learning_rate": 7.961363163591333e-05,
+      "loss": 1.3412374258041382,
+      "step": 387
+    },
+    {
+      "epoch": 0.08452775382380974,
+      "grad_norm": 0.7578125,
+      "learning_rate": 7.961051455667016e-05,
+      "loss": 1.261780858039856,
+      "step": 388
+    },
+    {
+      "epoch": 0.08474560885943812,
+      "grad_norm": 0.75390625,
+      "learning_rate": 7.960738501567586e-05,
+      "loss": 1.1538840532302856,
+      "step": 389
+    },
+    {
+      "epoch": 0.0849634638950665,
+      "grad_norm": 0.71875,
+      "learning_rate": 7.960424301391504e-05,
+      "loss": 1.3194695711135864,
+      "step": 390
+    },
+    {
+      "epoch": 0.08518131893069486,
+      "grad_norm": 0.72265625,
+      "learning_rate": 7.96010885523762e-05,
+      "loss": 1.1754437685012817,
+      "step": 391
+    },
+    {
+      "epoch": 0.08539917396632324,
+      "grad_norm": 0.73828125,
+      "learning_rate": 7.959792163205175e-05,
+      "loss": 1.2345529794692993,
+      "step": 392
+    },
+    {
+      "epoch": 0.08561702900195162,
+      "grad_norm": 0.72265625,
+      "learning_rate": 7.959474225393801e-05,
+      "loss": 1.2869163751602173,
+      "step": 393
+    },
+    {
+      "epoch": 0.08583488403758,
+      "grad_norm": 0.72265625,
+      "learning_rate": 7.959155041903526e-05,
+      "loss": 1.256472110748291,
+      "step": 394
+    },
+    {
+      "epoch": 0.08605273907320837,
+      "grad_norm": 0.77734375,
+      "learning_rate": 7.958834612834767e-05,
+      "loss": 1.2894366979599,
+      "step": 395
+    },
+    {
+      "epoch": 0.08627059410883675,
+      "grad_norm": 0.89453125,
+      "learning_rate": 7.958512938288332e-05,
+      "loss": 1.1717277765274048,
+      "step": 396
+    },
+    {
+      "epoch": 0.08648844914446512,
+      "grad_norm": 0.8046875,
+      "learning_rate": 7.958190018365424e-05,
+      "loss": 1.2086405754089355,
+      "step": 397
+    },
+    {
+      "epoch": 0.0867063041800935,
+      "grad_norm": 0.75390625,
+      "learning_rate": 7.957865853167634e-05,
+      "loss": 1.3303413391113281,
+      "step": 398
+    },
+    {
+      "epoch": 0.08692415921572187,
+      "grad_norm": 0.765625,
+      "learning_rate": 7.957540442796946e-05,
+      "loss": 1.3979418277740479,
+      "step": 399
+    },
+    {
+      "epoch": 0.08714201425135025,
+      "grad_norm": 0.78125,
+      "learning_rate": 7.957213787355738e-05,
+      "loss": 1.2019472122192383,
+      "step": 400
+    },
+    {
+      "epoch": 0.08735986928697863,
+      "grad_norm": 0.73828125,
+      "learning_rate": 7.95688588694678e-05,
+      "loss": 1.1940120458602905,
+      "step": 401
+    },
+    {
+      "epoch": 0.08757772432260699,
+      "grad_norm": 0.75,
+      "learning_rate": 7.956556741673228e-05,
+      "loss": 1.3304308652877808,
+      "step": 402
+    },
+    {
+      "epoch": 0.08779557935823537,
+      "grad_norm": 0.75390625,
+      "learning_rate": 7.956226351638636e-05,
+      "loss": 1.1966047286987305,
+      "step": 403
+    },
+    {
+      "epoch": 0.08801343439386375,
+      "grad_norm": 0.7265625,
+      "learning_rate": 7.955894716946946e-05,
+      "loss": 1.19243586063385,
+      "step": 404
+    },
+    {
+      "epoch": 0.08823128942949213,
+      "grad_norm": 0.83203125,
+      "learning_rate": 7.955561837702493e-05,
+      "loss": 1.4080017805099487,
+      "step": 405
+    },
+    {
+      "epoch": 0.0884491444651205,
+      "grad_norm": 0.73046875,
+      "learning_rate": 7.955227714010003e-05,
+      "loss": 1.2929046154022217,
+      "step": 406
+    },
+    {
+      "epoch": 0.08866699950074888,
+      "grad_norm": 0.69921875,
+      "learning_rate": 7.954892345974595e-05,
+      "loss": 1.1524901390075684,
+      "step": 407
+    },
+    {
+      "epoch": 0.08888485453637725,
+      "grad_norm": 0.765625,
+      "learning_rate": 7.954555733701776e-05,
+      "loss": 1.309418797492981,
+      "step": 408
+    },
+    {
+      "epoch": 0.08910270957200563,
+      "grad_norm": 0.72265625,
+      "learning_rate": 7.954217877297447e-05,
+      "loss": 1.3016916513442993,
+      "step": 409
+    },
+    {
+      "epoch": 0.089320564607634,
+      "grad_norm": 0.7421875,
+      "learning_rate": 7.953878776867902e-05,
+      "loss": 1.3607395887374878,
+      "step": 410
+    },
+    {
+      "epoch": 0.08953841964326238,
+      "grad_norm": 0.734375,
+      "learning_rate": 7.953538432519822e-05,
+      "loss": 1.235575795173645,
+      "step": 411
+    },
+    {
+      "epoch": 0.08975627467889076,
+      "grad_norm": 0.78515625,
+      "learning_rate": 7.953196844360284e-05,
+      "loss": 1.2608917951583862,
+      "step": 412
+    },
+    {
+      "epoch": 0.08997412971451912,
+      "grad_norm": 0.75390625,
+      "learning_rate": 7.952854012496751e-05,
+      "loss": 1.197023630142212,
+      "step": 413
+    },
+    {
+      "epoch": 0.0901919847501475,
+      "grad_norm": 0.796875,
+      "learning_rate": 7.952509937037084e-05,
+      "loss": 1.3533226251602173,
+      "step": 414
+    },
+    {
+      "epoch": 0.09040983978577588,
+      "grad_norm": 0.74609375,
+      "learning_rate": 7.95216461808953e-05,
+      "loss": 1.1115686893463135,
+      "step": 415
+    },
+    {
+      "epoch": 0.09062769482140426,
+      "grad_norm": 0.6796875,
+      "learning_rate": 7.951818055762727e-05,
+      "loss": 1.1842843294143677,
+      "step": 416
+    },
+    {
+      "epoch": 0.09084554985703264,
+      "grad_norm": 0.80859375,
+      "learning_rate": 7.951470250165709e-05,
+      "loss": 1.2439857721328735,
+      "step": 417
+    },
+    {
+      "epoch": 0.09106340489266102,
+      "grad_norm": 0.828125,
+      "learning_rate": 7.951121201407897e-05,
+      "loss": 1.2578680515289307,
+      "step": 418
+    },
+    {
+      "epoch": 0.09128125992828938,
+      "grad_norm": 0.734375,
+      "learning_rate": 7.950770909599104e-05,
+      "loss": 1.2519304752349854,
+      "step": 419
+    },
+    {
+      "epoch": 0.09149911496391776,
+      "grad_norm": 0.7578125,
+      "learning_rate": 7.950419374849535e-05,
+      "loss": 1.2428638935089111,
+      "step": 420
+    },
+    {
+      "epoch": 0.09171696999954614,
+      "grad_norm": 0.78125,
+      "learning_rate": 7.950066597269783e-05,
+      "loss": 1.2745386362075806,
+      "step": 421
+    },
+    {
+      "epoch": 0.09193482503517451,
+      "grad_norm": 0.75,
+      "learning_rate": 7.949712576970838e-05,
+      "loss": 1.1048963069915771,
+      "step": 422
+    },
+    {
+      "epoch": 0.09215268007080289,
+      "grad_norm": 0.76953125,
+      "learning_rate": 7.949357314064073e-05,
+      "loss": 1.0933512449264526,
+      "step": 423
+    },
+    {
+      "epoch": 0.09237053510643126,
+      "grad_norm": 0.7578125,
+      "learning_rate": 7.949000808661261e-05,
+      "loss": 1.1222467422485352,
+      "step": 424
+    },
+    {
+      "epoch": 0.09258839014205963,
+      "grad_norm": 0.75390625,
+      "learning_rate": 7.948643060874558e-05,
+      "loss": 1.1561400890350342,
+      "step": 425
+    },
+    {
+      "epoch": 0.09280624517768801,
+      "grad_norm": 0.81640625,
+      "learning_rate": 7.948284070816513e-05,
+      "loss": 1.2059650421142578,
+      "step": 426
+    },
+    {
+      "epoch": 0.09302410021331639,
+      "grad_norm": 0.7578125,
+      "learning_rate": 7.947923838600071e-05,
+      "loss": 1.3203227519989014,
+      "step": 427
+    },
+    {
+      "epoch": 0.09324195524894477,
+      "grad_norm": 0.8125,
+      "learning_rate": 7.94756236433856e-05,
+      "loss": 1.419194221496582,
+      "step": 428
+    },
+    {
+      "epoch": 0.09345981028457315,
+      "grad_norm": 0.8359375,
+      "learning_rate": 7.947199648145704e-05,
+      "loss": 1.323651909828186,
+      "step": 429
+    },
+    {
+      "epoch": 0.09367766532020151,
+      "grad_norm": 0.68359375,
+      "learning_rate": 7.946835690135617e-05,
+      "loss": 1.190460205078125,
+      "step": 430
+    },
+    {
+      "epoch": 0.09389552035582989,
+      "grad_norm": 0.81640625,
+      "learning_rate": 7.9464704904228e-05,
+      "loss": 1.3484115600585938,
+      "step": 431
+    },
+    {
+      "epoch": 0.09411337539145827,
+      "grad_norm": 0.7578125,
+      "learning_rate": 7.946104049122149e-05,
+      "loss": 1.1663109064102173,
+      "step": 432
+    },
+    {
+      "epoch": 0.09433123042708665,
+      "grad_norm": 0.72265625,
+      "learning_rate": 7.945736366348951e-05,
+      "loss": 1.2549434900283813,
+      "step": 433
+    },
+    {
+      "epoch": 0.09454908546271502,
+      "grad_norm": 0.70703125,
+      "learning_rate": 7.945367442218876e-05,
+      "loss": 1.1870343685150146,
+      "step": 434
+    },
+    {
+      "epoch": 0.09476694049834339,
+      "grad_norm": 1.8046875,
+      "learning_rate": 7.944997276847996e-05,
+      "loss": 1.2138428688049316,
+      "step": 435
+    },
+    {
+      "epoch": 0.09498479553397177,
+      "grad_norm": 0.80078125,
+      "learning_rate": 7.944625870352766e-05,
+      "loss": 1.2461137771606445,
+      "step": 436
+    },
+    {
+      "epoch": 0.09520265056960014,
+      "grad_norm": 0.73046875,
+      "learning_rate": 7.944253222850032e-05,
+      "loss": 1.395491600036621,
+      "step": 437
+    },
+    {
+      "epoch": 0.09542050560522852,
+      "grad_norm": 0.7421875,
+      "learning_rate": 7.943879334457033e-05,
+      "loss": 1.1486619710922241,
+      "step": 438
+    },
+    {
+      "epoch": 0.0956383606408569,
+      "grad_norm": 0.6875,
+      "learning_rate": 7.943504205291395e-05,
+      "loss": 1.181916356086731,
+      "step": 439
+    },
+    {
+      "epoch": 0.09585621567648527,
+      "grad_norm": 0.828125,
+      "learning_rate": 7.943127835471138e-05,
+      "loss": 1.3973501920700073,
+      "step": 440
+    },
+    {
+      "epoch": 0.09607407071211364,
+      "grad_norm": 0.72265625,
+      "learning_rate": 7.94275022511467e-05,
+      "loss": 1.177747368812561,
+      "step": 441
+    },
+    {
+      "epoch": 0.09629192574774202,
+      "grad_norm": 0.89453125,
+      "learning_rate": 7.94237137434079e-05,
+      "loss": 1.3191545009613037,
+      "step": 442
+    },
+    {
+      "epoch": 0.0965097807833704,
+      "grad_norm": 0.7578125,
+      "learning_rate": 7.941991283268687e-05,
+      "loss": 1.4455912113189697,
+      "step": 443
+    },
+    {
+      "epoch": 0.09672763581899878,
+      "grad_norm": 0.7109375,
+      "learning_rate": 7.941609952017942e-05,
+      "loss": 1.2649753093719482,
+      "step": 444
+    },
+    {
+      "epoch": 0.09694549085462716,
+      "grad_norm": 0.8125,
+      "learning_rate": 7.941227380708522e-05,
+      "loss": 1.407196283340454,
+      "step": 445
+    },
+    {
+      "epoch": 0.09716334589025552,
+      "grad_norm": 0.74609375,
+      "learning_rate": 7.940843569460788e-05,
+      "loss": 1.3769363164901733,
+      "step": 446
+    },
+    {
+      "epoch": 0.0973812009258839,
+      "grad_norm": 0.71875,
+      "learning_rate": 7.940458518395488e-05,
+      "loss": 1.3271342515945435,
+      "step": 447
+    },
+    {
+      "epoch": 0.09759905596151228,
+      "grad_norm": 0.70703125,
+      "learning_rate": 7.940072227633766e-05,
+      "loss": 1.1189401149749756,
+      "step": 448
+    },
+    {
+      "epoch": 0.09781691099714065,
+      "grad_norm": 0.7265625,
+      "learning_rate": 7.939684697297149e-05,
+      "loss": 1.176874041557312,
+      "step": 449
+    },
+    {
+      "epoch": 0.09803476603276903,
+      "grad_norm": 0.71875,
+      "learning_rate": 7.939295927507555e-05,
+      "loss": 1.2563841342926025,
+      "step": 450
+    },
+    {
+      "epoch": 0.0982526210683974,
+      "grad_norm": 0.7578125,
+      "learning_rate": 7.938905918387298e-05,
+      "loss": 1.2415258884429932,
+      "step": 451
+    },
+    {
+      "epoch": 0.09847047610402578,
+      "grad_norm": 0.7421875,
+      "learning_rate": 7.938514670059073e-05,
+      "loss": 1.4033838510513306,
+      "step": 452
+    },
+    {
+      "epoch": 0.09868833113965415,
+      "grad_norm": 0.73828125,
+      "learning_rate": 7.938122182645974e-05,
+      "loss": 1.2351109981536865,
+      "step": 453
+    },
+    {
+      "epoch": 0.09890618617528253,
+      "grad_norm": 0.71875,
+      "learning_rate": 7.937728456271478e-05,
+      "loss": 1.5063072443008423,
+      "step": 454
+    },
+    {
+      "epoch": 0.09912404121091091,
+      "grad_norm": 0.734375,
+      "learning_rate": 7.937333491059454e-05,
+      "loss": 1.3696823120117188,
+      "step": 455
+    },
+    {
+      "epoch": 0.09934189624653929,
+      "grad_norm": 0.68359375,
+      "learning_rate": 7.936937287134161e-05,
+      "loss": 1.1766464710235596,
+      "step": 456
+    },
+    {
+      "epoch": 0.09955975128216765,
+      "grad_norm": 0.70703125,
+      "learning_rate": 7.936539844620247e-05,
+      "loss": 1.2987933158874512,
+      "step": 457
+    },
+    {
+      "epoch": 0.09977760631779603,
+      "grad_norm": 0.7578125,
+      "learning_rate": 7.936141163642753e-05,
+      "loss": 1.3872061967849731,
+      "step": 458
+    },
+    {
+      "epoch": 0.09999546135342441,
+      "grad_norm": 0.72265625,
+      "learning_rate": 7.935741244327106e-05,
+      "loss": 1.290106177330017,
+      "step": 459
+    },
+    {
+      "epoch": 0.10021331638905279,
+      "grad_norm": 0.72265625,
+      "learning_rate": 7.93534008679912e-05,
+      "loss": 1.1763856410980225,
+      "step": 460
+    },
+    {
+      "epoch": 0.10043117142468116,
+      "grad_norm": 0.703125,
+      "learning_rate": 7.934937691185004e-05,
+      "loss": 1.0290653705596924,
+      "step": 461
+    },
+    {
+      "epoch": 0.10064902646030953,
+      "grad_norm": 0.69921875,
+      "learning_rate": 7.934534057611357e-05,
+      "loss": 1.2467173337936401,
+      "step": 462
+    },
+    {
+      "epoch": 0.10086688149593791,
+      "grad_norm": 0.73828125,
+      "learning_rate": 7.934129186205164e-05,
+      "loss": 1.3960658311843872,
+      "step": 463
+    },
+    {
+      "epoch": 0.10108473653156629,
+      "grad_norm": 0.7421875,
+      "learning_rate": 7.933723077093798e-05,
+      "loss": 1.347214698791504,
+      "step": 464
+    },
+    {
+      "epoch": 0.10130259156719466,
+      "grad_norm": 0.6875,
+      "learning_rate": 7.933315730405026e-05,
+      "loss": 1.241326928138733,
+      "step": 465
+    },
+    {
+      "epoch": 0.10152044660282304,
+      "grad_norm": 0.69140625,
+      "learning_rate": 7.932907146267001e-05,
+      "loss": 1.2166563272476196,
+      "step": 466
+    },
+    {
+      "epoch": 0.10173830163845142,
+      "grad_norm": 0.671875,
+      "learning_rate": 7.932497324808267e-05,
+      "loss": 1.0748162269592285,
+      "step": 467
+    },
+    {
+      "epoch": 0.10195615667407978,
+      "grad_norm": 0.7578125,
+      "learning_rate": 7.932086266157758e-05,
+      "loss": 1.4946786165237427,
+      "step": 468
+    },
+    {
+      "epoch": 0.10217401170970816,
+      "grad_norm": 0.75,
+      "learning_rate": 7.931673970444793e-05,
+      "loss": 1.2904515266418457,
+      "step": 469
+    },
+    {
+      "epoch": 0.10239186674533654,
+      "grad_norm": 0.6953125,
+      "learning_rate": 7.931260437799085e-05,
+      "loss": 1.2674866914749146,
+      "step": 470
+    },
+    {
+      "epoch": 0.10260972178096492,
+      "grad_norm": 0.74609375,
+      "learning_rate": 7.930845668350736e-05,
+      "loss": 1.2883021831512451,
+      "step": 471
+    },
+    {
+      "epoch": 0.1028275768165933,
+      "grad_norm": 0.72265625,
+      "learning_rate": 7.930429662230232e-05,
+      "loss": 1.2637059688568115,
+      "step": 472
+    },
+    {
+      "epoch": 0.10304543185222166,
+      "grad_norm": 0.6640625,
+      "learning_rate": 7.930012419568456e-05,
+      "loss": 1.1326955556869507,
+      "step": 473
+    },
+    {
+      "epoch": 0.10326328688785004,
+      "grad_norm": 0.70703125,
+      "learning_rate": 7.92959394049667e-05,
+      "loss": 1.217268705368042,
+      "step": 474
+    },
+    {
+      "epoch": 0.10348114192347842,
+      "grad_norm": 0.70703125,
+      "learning_rate": 7.929174225146534e-05,
+      "loss": 1.1191366910934448,
+      "step": 475
+    },
+    {
+      "epoch": 0.1036989969591068,
+      "grad_norm": 0.6484375,
+      "learning_rate": 7.928753273650092e-05,
+      "loss": 0.9647811651229858,
+      "step": 476
+    },
+    {
+      "epoch": 0.10391685199473517,
+      "grad_norm": 0.69140625,
+      "learning_rate": 7.92833108613978e-05,
+      "loss": 1.2803564071655273,
+      "step": 477
+    },
+    {
+      "epoch": 0.10413470703036355,
+      "grad_norm": 0.734375,
+      "learning_rate": 7.927907662748421e-05,
+      "loss": 1.32694673538208,
+      "step": 478
+    },
+    {
+      "epoch": 0.10435256206599192,
+      "grad_norm": 0.73046875,
+      "learning_rate": 7.927483003609225e-05,
+      "loss": 1.295607566833496,
+      "step": 479
+    },
+    {
+      "epoch": 0.1045704171016203,
+      "grad_norm": 0.734375,
+      "learning_rate": 7.927057108855795e-05,
+      "loss": 1.2564709186553955,
+      "step": 480
+    },
+    {
+      "epoch": 0.10478827213724867,
+      "grad_norm": 0.68359375,
+      "learning_rate": 7.92662997862212e-05,
+      "loss": 1.2449184656143188,
+      "step": 481
+    },
+    {
+      "epoch": 0.10500612717287705,
+      "grad_norm": 0.734375,
+      "learning_rate": 7.926201613042576e-05,
+      "loss": 1.296997308731079,
+      "step": 482
+    },
+    {
+      "epoch": 0.10522398220850543,
+      "grad_norm": 0.67578125,
+      "learning_rate": 7.925772012251933e-05,
+      "loss": 1.186234951019287,
+      "step": 483
+    },
+    {
+      "epoch": 0.1054418372441338,
+      "grad_norm": 0.76171875,
+      "learning_rate": 7.925341176385345e-05,
+      "loss": 1.3831301927566528,
+      "step": 484
+    },
+    {
+      "epoch": 0.10565969227976217,
+      "grad_norm": 0.734375,
+      "learning_rate": 7.924909105578356e-05,
+      "loss": 1.2782447338104248,
+      "step": 485
+    },
+    {
+      "epoch": 0.10587754731539055,
+      "grad_norm": 0.6875,
+      "learning_rate": 7.9244757999669e-05,
+      "loss": 1.3897466659545898,
+      "step": 486
+    },
+    {
+      "epoch": 0.10609540235101893,
+      "grad_norm": 0.71484375,
+      "learning_rate": 7.924041259687295e-05,
+      "loss": 1.4396719932556152,
+      "step": 487
+    },
+    {
+      "epoch": 0.1063132573866473,
+      "grad_norm": 0.71875,
+      "learning_rate": 7.923605484876253e-05,
+      "loss": 1.301027774810791,
+      "step": 488
+    },
+    {
+      "epoch": 0.10653111242227568,
+      "grad_norm": 0.6328125,
+      "learning_rate": 7.923168475670872e-05,
+      "loss": 1.1693106889724731,
+      "step": 489
+    },
+    {
+      "epoch": 0.10674896745790405,
+      "grad_norm": 0.703125,
+      "learning_rate": 7.922730232208637e-05,
+      "loss": 1.2485506534576416,
+      "step": 490
+    },
+    {
+      "epoch": 0.10696682249353243,
+      "grad_norm": 0.734375,
+      "learning_rate": 7.922290754627422e-05,
+      "loss": 1.2698620557785034,
+      "step": 491
+    },
+    {
+      "epoch": 0.1071846775291608,
+      "grad_norm": 0.7109375,
+      "learning_rate": 7.921850043065492e-05,
+      "loss": 1.2255313396453857,
+      "step": 492
+    },
+    {
+      "epoch": 0.10740253256478918,
+      "grad_norm": 0.6875,
+      "learning_rate": 7.921408097661497e-05,
+      "loss": 1.3092511892318726,
+      "step": 493
+    },
+    {
+      "epoch": 0.10762038760041756,
+      "grad_norm": 0.71484375,
+      "learning_rate": 7.920964918554475e-05,
+      "loss": 1.0415505170822144,
+      "step": 494
+    },
+    {
+      "epoch": 0.10783824263604593,
+      "grad_norm": 0.65625,
+      "learning_rate": 7.920520505883855e-05,
+      "loss": 1.2461661100387573,
+      "step": 495
+    },
+    {
+      "epoch": 0.1080560976716743,
+      "grad_norm": 0.72265625,
+      "learning_rate": 7.920074859789453e-05,
+      "loss": 1.3234269618988037,
+      "step": 496
+    },
+    {
+      "epoch": 0.10827395270730268,
+      "grad_norm": 0.69921875,
+      "learning_rate": 7.919627980411468e-05,
+      "loss": 1.2717959880828857,
+      "step": 497
+    },
+    {
+      "epoch": 0.10849180774293106,
+      "grad_norm": 0.6953125,
+      "learning_rate": 7.919179867890496e-05,
+      "loss": 1.2537888288497925,
+      "step": 498
+    },
+    {
+      "epoch": 0.10870966277855944,
+      "grad_norm": 0.734375,
+      "learning_rate": 7.918730522367515e-05,
+      "loss": 1.2870296239852905,
+      "step": 499
+    },
+    {
+      "epoch": 0.10892751781418782,
+      "grad_norm": 0.7265625,
+      "learning_rate": 7.918279943983893e-05,
+      "loss": 1.2527849674224854,
+      "step": 500
+    },
+    {
+      "epoch": 0.10914537284981618,
+      "grad_norm": 0.6640625,
+      "learning_rate": 7.917828132881383e-05,
+      "loss": 1.2299834489822388,
+      "step": 501
+    },
+    {
+      "epoch": 0.10936322788544456,
+      "grad_norm": 0.7734375,
+      "learning_rate": 7.91737508920213e-05,
+      "loss": 1.4060068130493164,
+      "step": 502
+    },
+    {
+      "epoch": 0.10958108292107294,
+      "grad_norm": 0.7578125,
+      "learning_rate": 7.916920813088664e-05,
+      "loss": 1.1786068677902222,
+      "step": 503
+    },
+    {
+      "epoch": 0.10979893795670131,
+      "grad_norm": 0.79296875,
+      "learning_rate": 7.916465304683906e-05,
+      "loss": 1.3912036418914795,
+      "step": 504
+    },
+    {
+      "epoch": 0.11001679299232969,
+      "grad_norm": 0.7109375,
+      "learning_rate": 7.916008564131158e-05,
+      "loss": 1.2980504035949707,
+      "step": 505
+    },
+    {
+      "epoch": 0.11023464802795806,
+      "grad_norm": 0.66015625,
+      "learning_rate": 7.915550591574117e-05,
+      "loss": 1.1767922639846802,
+      "step": 506
+    },
+    {
+      "epoch": 0.11045250306358644,
+      "grad_norm": 0.640625,
+      "learning_rate": 7.915091387156862e-05,
+      "loss": 1.1484233140945435,
+      "step": 507
+    },
+    {
+      "epoch": 0.11067035809921481,
+      "grad_norm": 0.67578125,
+      "learning_rate": 7.914630951023862e-05,
+      "loss": 1.169138789176941,
+      "step": 508
+    },
+    {
+      "epoch": 0.11088821313484319,
+      "grad_norm": 0.7109375,
+      "learning_rate": 7.914169283319977e-05,
+      "loss": 1.1663360595703125,
+      "step": 509
+    },
+    {
+      "epoch": 0.11110606817047157,
+      "grad_norm": 0.7734375,
+      "learning_rate": 7.913706384190449e-05,
+      "loss": 1.27113676071167,
+      "step": 510
+    },
+    {
+      "epoch": 0.11132392320609995,
+      "grad_norm": 0.6484375,
+      "learning_rate": 7.913242253780906e-05,
+      "loss": 1.1611939668655396,
+      "step": 511
+    },
+    {
+      "epoch": 0.11154177824172831,
+      "grad_norm": 0.71484375,
+      "learning_rate": 7.912776892237373e-05,
+      "loss": 1.3341511487960815,
+      "step": 512
+    },
+    {
+      "epoch": 0.11175963327735669,
+      "grad_norm": 0.79296875,
+      "learning_rate": 7.912310299706252e-05,
+      "loss": 1.1612821817398071,
+      "step": 513
+    },
+    {
+      "epoch": 0.11197748831298507,
+      "grad_norm": 1.0234375,
+      "learning_rate": 7.911842476334336e-05,
+      "loss": 1.4641658067703247,
+      "step": 514
+    },
+    {
+      "epoch": 0.11219534334861345,
+      "grad_norm": 0.71875,
+      "learning_rate": 7.911373422268809e-05,
+      "loss": 1.146370530128479,
+      "step": 515
+    },
+    {
+      "epoch": 0.11241319838424182,
+      "grad_norm": 0.74609375,
+      "learning_rate": 7.910903137657235e-05,
+      "loss": 1.225861668586731,
+      "step": 516
+    },
+    {
+      "epoch": 0.11263105341987019,
+      "grad_norm": 0.69140625,
+      "learning_rate": 7.91043162264757e-05,
+      "loss": 1.2287743091583252,
+      "step": 517
+    },
+    {
+      "epoch": 0.11284890845549857,
+      "grad_norm": 0.6953125,
+      "learning_rate": 7.909958877388159e-05,
+      "loss": 1.2347304821014404,
+      "step": 518
+    },
+    {
+      "epoch": 0.11306676349112695,
+      "grad_norm": 0.734375,
+      "learning_rate": 7.909484902027726e-05,
+      "loss": 1.3644320964813232,
+      "step": 519
+    },
+    {
+      "epoch": 0.11328461852675532,
+      "grad_norm": 0.6640625,
+      "learning_rate": 7.909009696715391e-05,
+      "loss": 1.1880377531051636,
+      "step": 520
+    },
+    {
+      "epoch": 0.1135024735623837,
+      "grad_norm": 0.70703125,
+      "learning_rate": 7.908533261600655e-05,
+      "loss": 1.2591747045516968,
+      "step": 521
+    },
+    {
+      "epoch": 0.11372032859801207,
+      "grad_norm": 0.71875,
+      "learning_rate": 7.908055596833407e-05,
+      "loss": 1.2195019721984863,
+      "step": 522
+    },
+    {
+      "epoch": 0.11393818363364044,
+      "grad_norm": 0.69921875,
+      "learning_rate": 7.907576702563928e-05,
+      "loss": 1.228622555732727,
+      "step": 523
+    },
+    {
+      "epoch": 0.11415603866926882,
+      "grad_norm": 0.75,
+      "learning_rate": 7.907096578942878e-05,
+      "loss": 1.28585946559906,
+      "step": 524
+    },
+    {
+      "epoch": 0.1143738937048972,
+      "grad_norm": 0.765625,
+      "learning_rate": 7.906615226121306e-05,
+      "loss": 1.361154556274414,
+      "step": 525
+    },
+    {
+      "epoch": 0.11459174874052558,
+      "grad_norm": 0.69140625,
+      "learning_rate": 7.906132644250652e-05,
+      "loss": 1.2284362316131592,
+      "step": 526
+    },
+    {
+      "epoch": 0.11480960377615396,
+      "grad_norm": 0.71875,
+      "learning_rate": 7.90564883348274e-05,
+      "loss": 1.3480770587921143,
+      "step": 527
+    },
+    {
+      "epoch": 0.11502745881178232,
+      "grad_norm": 0.7109375,
+      "learning_rate": 7.905163793969778e-05,
+      "loss": 1.2245632410049438,
+      "step": 528
+    },
+    {
+      "epoch": 0.1152453138474107,
+      "grad_norm": 0.72265625,
+      "learning_rate": 7.904677525864366e-05,
+      "loss": 1.224600076675415,
+      "step": 529
+    },
+    {
+      "epoch": 0.11546316888303908,
+      "grad_norm": 0.76171875,
+      "learning_rate": 7.904190029319483e-05,
+      "loss": 1.2938926219940186,
+      "step": 530
+    },
+    {
+      "epoch": 0.11568102391866746,
+      "grad_norm": 0.67578125,
+      "learning_rate": 7.903701304488502e-05,
+      "loss": 1.1516841650009155,
+      "step": 531
+    },
+    {
+      "epoch": 0.11589887895429583,
+      "grad_norm": 0.70703125,
+      "learning_rate": 7.90321135152518e-05,
+      "loss": 1.3420064449310303,
+      "step": 532
+    },
+    {
+      "epoch": 0.1161167339899242,
+      "grad_norm": 0.70703125,
+      "learning_rate": 7.902720170583656e-05,
+      "loss": 1.1904546022415161,
+      "step": 533
+    },
+    {
+      "epoch": 0.11633458902555258,
+      "grad_norm": 0.6796875,
+      "learning_rate": 7.902227761818464e-05,
+      "loss": 1.2729661464691162,
+      "step": 534
+    },
+    {
+      "epoch": 0.11655244406118095,
+      "grad_norm": 0.6953125,
+      "learning_rate": 7.901734125384515e-05,
+      "loss": 1.1212387084960938,
+      "step": 535
+    },
+    {
+      "epoch": 0.11677029909680933,
+      "grad_norm": 0.75390625,
+      "learning_rate": 7.901239261437113e-05,
+      "loss": 1.3297935724258423,
+      "step": 536
+    },
+    {
+      "epoch": 0.11698815413243771,
+      "grad_norm": 0.63671875,
+      "learning_rate": 7.900743170131945e-05,
+      "loss": 1.1069059371948242,
+      "step": 537
+    },
+    {
+      "epoch": 0.11720600916806609,
+      "grad_norm": 0.67578125,
+      "learning_rate": 7.900245851625086e-05,
+      "loss": 1.2077605724334717,
+      "step": 538
+    },
+    {
+      "epoch": 0.11742386420369445,
+      "grad_norm": 0.67578125,
+      "learning_rate": 7.899747306072993e-05,
+      "loss": 1.2002694606781006,
+      "step": 539
+    },
+    {
+      "epoch": 0.11764171923932283,
+      "grad_norm": 0.7265625,
+      "learning_rate": 7.899247533632514e-05,
+      "loss": 1.1837718486785889,
+      "step": 540
+    },
+    {
+      "epoch": 0.11785957427495121,
+      "grad_norm": 0.69921875,
+      "learning_rate": 7.898746534460883e-05,
+      "loss": 1.1894031763076782,
+      "step": 541
+    },
+    {
+      "epoch": 0.11807742931057959,
+      "grad_norm": 0.68359375,
+      "learning_rate": 7.898244308715713e-05,
+      "loss": 1.220327377319336,
+      "step": 542
+    },
+    {
+      "epoch": 0.11829528434620797,
+      "grad_norm": 0.7421875,
+      "learning_rate": 7.897740856555011e-05,
+      "loss": 1.357073187828064,
+      "step": 543
+    },
+    {
+      "epoch": 0.11851313938183633,
+      "grad_norm": 0.74609375,
+      "learning_rate": 7.897236178137167e-05,
+      "loss": 1.2702226638793945,
+      "step": 544
+    },
+    {
+      "epoch": 0.11873099441746471,
+      "grad_norm": 0.703125,
+      "learning_rate": 7.896730273620956e-05,
+      "loss": 1.1159194707870483,
+      "step": 545
+    },
+    {
+      "epoch": 0.11894884945309309,
+      "grad_norm": 0.734375,
+      "learning_rate": 7.89622314316554e-05,
+      "loss": 1.3496376276016235,
+      "step": 546
+    },
+    {
+      "epoch": 0.11916670448872146,
+      "grad_norm": 0.6796875,
+      "learning_rate": 7.895714786930462e-05,
+      "loss": 1.2019708156585693,
+      "step": 547
+    },
+    {
+      "epoch": 0.11938455952434984,
+      "grad_norm": 0.74609375,
+      "learning_rate": 7.89520520507566e-05,
+      "loss": 1.3740615844726562,
+      "step": 548
+    },
+    {
+      "epoch": 0.11960241455997822,
+      "grad_norm": 0.71484375,
+      "learning_rate": 7.894694397761447e-05,
+      "loss": 1.3080002069473267,
+      "step": 549
+    },
+    {
+      "epoch": 0.11982026959560658,
+      "grad_norm": 0.6953125,
+      "learning_rate": 7.89418236514853e-05,
+      "loss": 1.2179882526397705,
+      "step": 550
+    },
+    {
+      "epoch": 0.12003812463123496,
+      "grad_norm": 0.71875,
+      "learning_rate": 7.893669107397996e-05,
+      "loss": 1.272958517074585,
+      "step": 551
+    },
+    {
+      "epoch": 0.12025597966686334,
+      "grad_norm": 0.72265625,
+      "learning_rate": 7.893154624671322e-05,
+      "loss": 1.278180480003357,
+      "step": 552
+    },
+    {
+      "epoch": 0.12047383470249172,
+      "grad_norm": 0.6953125,
+      "learning_rate": 7.892638917130365e-05,
+      "loss": 1.1408332586288452,
+      "step": 553
+    },
+    {
+      "epoch": 0.1206916897381201,
+      "grad_norm": 0.7109375,
+      "learning_rate": 7.892121984937373e-05,
+      "loss": 1.2112058401107788,
+      "step": 554
+    },
+    {
+      "epoch": 0.12090954477374846,
+      "grad_norm": 0.74609375,
+      "learning_rate": 7.891603828254975e-05,
+      "loss": 1.3279653787612915,
+      "step": 555
+    },
+    {
+      "epoch": 0.12112739980937684,
+      "grad_norm": 0.734375,
+      "learning_rate": 7.891084447246188e-05,
+      "loss": 1.3556170463562012,
+      "step": 556
+    },
+    {
+      "epoch": 0.12134525484500522,
+      "grad_norm": 0.71484375,
+      "learning_rate": 7.890563842074412e-05,
+      "loss": 1.3505316972732544,
+      "step": 557
+    },
+    {
+      "epoch": 0.1215631098806336,
+      "grad_norm": 0.67578125,
+      "learning_rate": 7.890042012903432e-05,
+      "loss": 1.3118557929992676,
+      "step": 558
+    },
+    {
+      "epoch": 0.12178096491626197,
+      "grad_norm": 0.75390625,
+      "learning_rate": 7.889518959897421e-05,
+      "loss": 1.1980655193328857,
+      "step": 559
+    },
+    {
+      "epoch": 0.12199881995189035,
+      "grad_norm": 0.70703125,
+      "learning_rate": 7.888994683220935e-05,
+      "loss": 1.1632696390151978,
+      "step": 560
+    },
+    {
+      "epoch": 0.12221667498751872,
+      "grad_norm": 0.69140625,
+      "learning_rate": 7.888469183038916e-05,
+      "loss": 1.2610007524490356,
+      "step": 561
+    },
+    {
+      "epoch": 0.1224345300231471,
+      "grad_norm": 0.68359375,
+      "learning_rate": 7.887942459516688e-05,
+      "loss": 1.1844148635864258,
+      "step": 562
+    },
+    {
+      "epoch": 0.12265238505877547,
+      "grad_norm": 0.69140625,
+      "learning_rate": 7.887414512819962e-05,
+      "loss": 1.2360007762908936,
+      "step": 563
+    },
+    {
+      "epoch": 0.12287024009440385,
+      "grad_norm": 0.7734375,
+      "learning_rate": 7.886885343114836e-05,
+      "loss": 1.3873823881149292,
+      "step": 564
+    },
+    {
+      "epoch": 0.12308809513003223,
+      "grad_norm": 0.6953125,
+      "learning_rate": 7.886354950567789e-05,
+      "loss": 1.2969856262207031,
+      "step": 565
+    },
+    {
+      "epoch": 0.1233059501656606,
+      "grad_norm": 0.72265625,
+      "learning_rate": 7.885823335345687e-05,
+      "loss": 1.1861568689346313,
+      "step": 566
+    },
+    {
+      "epoch": 0.12352380520128897,
+      "grad_norm": 0.765625,
+      "learning_rate": 7.88529049761578e-05,
+      "loss": 1.2690879106521606,
+      "step": 567
+    },
+    {
+      "epoch": 0.12374166023691735,
+      "grad_norm": 0.69921875,
+      "learning_rate": 7.884756437545703e-05,
+      "loss": 1.1326102018356323,
+      "step": 568
+    },
+    {
+      "epoch": 0.12395951527254573,
+      "grad_norm": 0.75390625,
+      "learning_rate": 7.884221155303472e-05,
+      "loss": 1.1546456813812256,
+      "step": 569
+    },
+    {
+      "epoch": 0.1241773703081741,
+      "grad_norm": 0.74609375,
+      "learning_rate": 7.883684651057494e-05,
+      "loss": 1.1913681030273438,
+      "step": 570
+    },
+    {
+      "epoch": 0.12439522534380248,
+      "grad_norm": 0.72265625,
+      "learning_rate": 7.883146924976555e-05,
+      "loss": 1.1861920356750488,
+      "step": 571
+    },
+    {
+      "epoch": 0.12461308037943085,
+      "grad_norm": 0.66796875,
+      "learning_rate": 7.88260797722983e-05,
+      "loss": 1.1968181133270264,
+      "step": 572
+    },
+    {
+      "epoch": 0.12483093541505923,
+      "grad_norm": 0.6640625,
+      "learning_rate": 7.882067807986874e-05,
+      "loss": 0.9793790578842163,
+      "step": 573
+    },
+    {
+      "epoch": 0.1250487904506876,
+      "grad_norm": 0.76171875,
+      "learning_rate": 7.881526417417628e-05,
+      "loss": 1.266435146331787,
+      "step": 574
+    },
+    {
+      "epoch": 0.12526664548631597,
+      "grad_norm": 0.69921875,
+      "learning_rate": 7.880983805692417e-05,
+      "loss": 1.3200803995132446,
+      "step": 575
+    },
+    {
+      "epoch": 0.12548450052194435,
+      "grad_norm": 0.7109375,
+      "learning_rate": 7.880439972981952e-05,
+      "loss": 1.2245416641235352,
+      "step": 576
+    },
+    {
+      "epoch": 0.12570235555757273,
+      "grad_norm": 0.71484375,
+      "learning_rate": 7.879894919457326e-05,
+      "loss": 1.2432984113693237,
+      "step": 577
+    },
+    {
+      "epoch": 0.1259202105932011,
+      "grad_norm": 0.7421875,
+      "learning_rate": 7.879348645290017e-05,
+      "loss": 1.2656619548797607,
+      "step": 578
+    },
+    {
+      "epoch": 0.12613806562882948,
+      "grad_norm": 0.6875,
+      "learning_rate": 7.878801150651885e-05,
+      "loss": 1.3180609941482544,
+      "step": 579
+    },
+    {
+      "epoch": 0.12635592066445786,
+      "grad_norm": 0.71484375,
+      "learning_rate": 7.878252435715178e-05,
+      "loss": 1.2602907419204712,
+      "step": 580
+    },
+    {
+      "epoch": 0.12657377570008624,
+      "grad_norm": 0.78125,
+      "learning_rate": 7.877702500652523e-05,
+      "loss": 1.1314157247543335,
+      "step": 581
+    },
+    {
+      "epoch": 0.12679163073571462,
+      "grad_norm": 0.78125,
+      "learning_rate": 7.877151345636938e-05,
+      "loss": 1.3106298446655273,
+      "step": 582
+    },
+    {
+      "epoch": 0.127009485771343,
+      "grad_norm": 0.7421875,
+      "learning_rate": 7.876598970841817e-05,
+      "loss": 1.403730034828186,
+      "step": 583
+    },
+    {
+      "epoch": 0.12722734080697137,
+      "grad_norm": 0.7109375,
+      "learning_rate": 7.876045376440941e-05,
+      "loss": 1.2627794742584229,
+      "step": 584
+    },
+    {
+      "epoch": 0.12744519584259972,
+      "grad_norm": 0.67578125,
+      "learning_rate": 7.875490562608476e-05,
+      "loss": 1.1343772411346436,
+      "step": 585
+    },
+    {
+      "epoch": 0.1276630508782281,
+      "grad_norm": 0.73828125,
+      "learning_rate": 7.87493452951897e-05,
+      "loss": 1.3060775995254517,
+      "step": 586
+    },
+    {
+      "epoch": 0.12788090591385648,
+      "grad_norm": 0.6796875,
+      "learning_rate": 7.874377277347354e-05,
+      "loss": 1.3424075841903687,
+      "step": 587
+    },
+    {
+      "epoch": 0.12809876094948486,
+      "grad_norm": 0.76953125,
+      "learning_rate": 7.873818806268944e-05,
+      "loss": 1.2973421812057495,
+      "step": 588
+    },
+    {
+      "epoch": 0.12831661598511324,
+      "grad_norm": 0.734375,
+      "learning_rate": 7.873259116459439e-05,
+      "loss": 1.2930057048797607,
+      "step": 589
+    },
+    {
+      "epoch": 0.12853447102074161,
+      "grad_norm": 0.71875,
+      "learning_rate": 7.872698208094923e-05,
+      "loss": 1.3182575702667236,
+      "step": 590
+    },
+    {
+      "epoch": 0.12875232605637,
+      "grad_norm": 0.69140625,
+      "learning_rate": 7.872136081351861e-05,
+      "loss": 1.2012242078781128,
+      "step": 591
+    },
+    {
+      "epoch": 0.12897018109199837,
+      "grad_norm": 0.75,
+      "learning_rate": 7.871572736407099e-05,
+      "loss": 1.3180583715438843,
+      "step": 592
+    },
+    {
+      "epoch": 0.12918803612762675,
+      "grad_norm": 0.734375,
+      "learning_rate": 7.871008173437874e-05,
+      "loss": 1.2393542528152466,
+      "step": 593
+    },
+    {
+      "epoch": 0.12940589116325513,
+      "grad_norm": 0.68359375,
+      "learning_rate": 7.870442392621798e-05,
+      "loss": 1.2725328207015991,
+      "step": 594
+    },
+    {
+      "epoch": 0.1296237461988835,
+      "grad_norm": 0.671875,
+      "learning_rate": 7.869875394136873e-05,
+      "loss": 1.2176402807235718,
+      "step": 595
+    },
+    {
+      "epoch": 0.12984160123451186,
+      "grad_norm": 0.66796875,
+      "learning_rate": 7.869307178161478e-05,
+      "loss": 1.2924630641937256,
+      "step": 596
+    },
+    {
+      "epoch": 0.13005945627014023,
+      "grad_norm": 0.68359375,
+      "learning_rate": 7.86873774487438e-05,
+      "loss": 1.220310091972351,
+      "step": 597
+    },
+    {
+      "epoch": 0.1302773113057686,
+      "grad_norm": 0.73046875,
+      "learning_rate": 7.868167094454724e-05,
+      "loss": 1.327128529548645,
+      "step": 598
+    },
+    {
+      "epoch": 0.130495166341397,
+      "grad_norm": 0.78515625,
+      "learning_rate": 7.867595227082041e-05,
+      "loss": 1.392011046409607,
+      "step": 599
+    },
+    {
+      "epoch": 0.13071302137702537,
+      "grad_norm": 0.73828125,
+      "learning_rate": 7.867022142936248e-05,
+      "loss": 1.25203275680542,
+      "step": 600
+    },
+    {
+      "epoch": 0.13093087641265375,
+      "grad_norm": 0.7109375,
+      "learning_rate": 7.866447842197638e-05,
+      "loss": 1.2183114290237427,
+      "step": 601
+    },
+    {
+      "epoch": 0.13114873144828212,
+      "grad_norm": 0.7265625,
+      "learning_rate": 7.865872325046893e-05,
+      "loss": 1.3020896911621094,
+      "step": 602
+    },
+    {
+      "epoch": 0.1313665864839105,
+      "grad_norm": 0.703125,
+      "learning_rate": 7.86529559166507e-05,
+      "loss": 1.3516041040420532,
+      "step": 603
+    },
+    {
+      "epoch": 0.13158444151953888,
+      "grad_norm": 0.69921875,
+      "learning_rate": 7.864717642233621e-05,
+      "loss": 1.1778078079223633,
+      "step": 604
+    },
+    {
+      "epoch": 0.13180229655516726,
+      "grad_norm": 0.7734375,
+      "learning_rate": 7.864138476934365e-05,
+      "loss": 1.405396819114685,
+      "step": 605
+    },
+    {
+      "epoch": 0.13202015159079564,
+      "grad_norm": 0.7109375,
+      "learning_rate": 7.863558095949517e-05,
+      "loss": 1.4001930952072144,
+      "step": 606
+    },
+    {
+      "epoch": 0.132238006626424,
+      "grad_norm": 0.72265625,
+      "learning_rate": 7.862976499461665e-05,
+      "loss": 1.3288977146148682,
+      "step": 607
+    },
+    {
+      "epoch": 0.13245586166205237,
+      "grad_norm": 0.71875,
+      "learning_rate": 7.862393687653788e-05,
+      "loss": 1.1715888977050781,
+      "step": 608
+    },
+    {
+      "epoch": 0.13267371669768074,
+      "grad_norm": 0.6796875,
+      "learning_rate": 7.861809660709239e-05,
+      "loss": 1.3586455583572388,
+      "step": 609
+    },
+    {
+      "epoch": 0.13289157173330912,
+      "grad_norm": 0.7265625,
+      "learning_rate": 7.861224418811757e-05,
+      "loss": 1.2887132167816162,
+      "step": 610
+    },
+    {
+      "epoch": 0.1331094267689375,
+      "grad_norm": 0.75,
+      "learning_rate": 7.860637962145467e-05,
+      "loss": 1.3280686140060425,
+      "step": 611
+    },
+    {
+      "epoch": 0.13332728180456588,
+      "grad_norm": 0.72265625,
+      "learning_rate": 7.86005029089487e-05,
+      "loss": 1.3015261888504028,
+      "step": 612
+    },
+    {
+      "epoch": 0.13354513684019426,
+      "grad_norm": 0.84765625,
+      "learning_rate": 7.859461405244851e-05,
+      "loss": 1.3692591190338135,
+      "step": 613
+    },
+    {
+      "epoch": 0.13376299187582263,
+      "grad_norm": 0.7890625,
+      "learning_rate": 7.858871305380677e-05,
+      "loss": 1.3561069965362549,
+      "step": 614
+    },
+    {
+      "epoch": 0.133980846911451,
+      "grad_norm": 0.7265625,
+      "learning_rate": 7.858279991488001e-05,
+      "loss": 1.3406193256378174,
+      "step": 615
+    },
+    {
+      "epoch": 0.1341987019470794,
+      "grad_norm": 0.765625,
+      "learning_rate": 7.857687463752851e-05,
+      "loss": 1.2436710596084595,
+      "step": 616
+    },
+    {
+      "epoch": 0.13441655698270777,
+      "grad_norm": 0.703125,
+      "learning_rate": 7.857093722361643e-05,
+      "loss": 1.0043692588806152,
+      "step": 617
+    },
+    {
+      "epoch": 0.13463441201833612,
+      "grad_norm": 0.73046875,
+      "learning_rate": 7.856498767501171e-05,
+      "loss": 1.307030200958252,
+      "step": 618
+    },
+    {
+      "epoch": 0.1348522670539645,
+      "grad_norm": 0.78125,
+      "learning_rate": 7.855902599358613e-05,
+      "loss": 1.3335739374160767,
+      "step": 619
+    },
+    {
+      "epoch": 0.13507012208959288,
+      "grad_norm": 0.7734375,
+      "learning_rate": 7.855305218121526e-05,
+      "loss": 1.3600963354110718,
+      "step": 620
+    },
+    {
+      "epoch": 0.13528797712522125,
+      "grad_norm": 0.7265625,
+      "learning_rate": 7.854706623977852e-05,
+      "loss": 1.3324352502822876,
+      "step": 621
+    },
+    {
+      "epoch": 0.13550583216084963,
+      "grad_norm": 0.77734375,
+      "learning_rate": 7.854106817115913e-05,
+      "loss": 1.3265318870544434,
+      "step": 622
+    },
+    {
+      "epoch": 0.135723687196478,
+      "grad_norm": 0.73828125,
+      "learning_rate": 7.85350579772441e-05,
+      "loss": 1.20114004611969,
+      "step": 623
+    },
+    {
+      "epoch": 0.1359415422321064,
+      "grad_norm": 0.67578125,
+      "learning_rate": 7.852903565992432e-05,
+      "loss": 1.1523607969284058,
+      "step": 624
+    },
+    {
+      "epoch": 0.13615939726773477,
+      "grad_norm": 0.74609375,
+      "learning_rate": 7.852300122109443e-05,
+      "loss": 1.1193900108337402,
+      "step": 625
+    },
+    {
+      "epoch": 0.13637725230336314,
+      "grad_norm": 0.76171875,
+      "learning_rate": 7.851695466265292e-05,
+      "loss": 1.260627269744873,
+      "step": 626
+    },
+    {
+      "epoch": 0.13659510733899152,
+      "grad_norm": 0.65625,
+      "learning_rate": 7.851089598650205e-05,
+      "loss": 1.2052860260009766,
+      "step": 627
+    },
+    {
+      "epoch": 0.1368129623746199,
+      "grad_norm": 0.7890625,
+      "learning_rate": 7.850482519454795e-05,
+      "loss": 1.2288596630096436,
+      "step": 628
+    },
+    {
+      "epoch": 0.13703081741024825,
+      "grad_norm": 0.75,
+      "learning_rate": 7.849874228870055e-05,
+      "loss": 1.1968188285827637,
+      "step": 629
+    },
+    {
+      "epoch": 0.13724867244587663,
+      "grad_norm": 0.76953125,
+      "learning_rate": 7.849264727087354e-05,
+      "loss": 1.275305151939392,
+      "step": 630
+    },
+    {
+      "epoch": 0.137466527481505,
+      "grad_norm": 0.7421875,
+      "learning_rate": 7.848654014298448e-05,
+      "loss": 1.2851362228393555,
+      "step": 631
+    },
+    {
+      "epoch": 0.13768438251713339,
+      "grad_norm": 0.6953125,
+      "learning_rate": 7.84804209069547e-05,
+      "loss": 1.185146689414978,
+      "step": 632
+    },
+    {
+      "epoch": 0.13790223755276176,
+      "grad_norm": 0.6796875,
+      "learning_rate": 7.847428956470936e-05,
+      "loss": 1.251382827758789,
+      "step": 633
+    },
+    {
+      "epoch": 0.13812009258839014,
+      "grad_norm": 0.6640625,
+      "learning_rate": 7.846814611817743e-05,
+      "loss": 1.1854121685028076,
+      "step": 634
+    },
+    {
+      "epoch": 0.13833794762401852,
+      "grad_norm": 0.69921875,
+      "learning_rate": 7.84619905692917e-05,
+      "loss": 1.263794183731079,
+      "step": 635
+    },
+    {
+      "epoch": 0.1385558026596469,
+      "grad_norm": 0.6484375,
+      "learning_rate": 7.845582291998871e-05,
+      "loss": 1.1930968761444092,
+      "step": 636
+    },
+    {
+      "epoch": 0.13877365769527528,
+      "grad_norm": 0.73046875,
+      "learning_rate": 7.844964317220886e-05,
+      "loss": 1.137607216835022,
+      "step": 637
+    },
+    {
+      "epoch": 0.13899151273090365,
+      "grad_norm": 0.6953125,
+      "learning_rate": 7.844345132789637e-05,
+      "loss": 0.9965424537658691,
+      "step": 638
+    },
+    {
+      "epoch": 0.13920936776653203,
+      "grad_norm": 0.71875,
+      "learning_rate": 7.84372473889992e-05,
+      "loss": 1.2219592332839966,
+      "step": 639
+    },
+    {
+      "epoch": 0.13942722280216038,
+      "grad_norm": 0.85546875,
+      "learning_rate": 7.843103135746916e-05,
+      "loss": 1.2363194227218628,
+      "step": 640
+    },
+    {
+      "epoch": 0.13964507783778876,
+      "grad_norm": 0.703125,
+      "learning_rate": 7.84248032352619e-05,
+      "loss": 1.1734493970870972,
+      "step": 641
+    },
+    {
+      "epoch": 0.13986293287341714,
+      "grad_norm": 0.78515625,
+      "learning_rate": 7.841856302433677e-05,
+      "loss": 1.372399926185608,
+      "step": 642
+    },
+    {
+      "epoch": 0.14008078790904552,
+      "grad_norm": 0.68359375,
+      "learning_rate": 7.841231072665705e-05,
+      "loss": 1.1826549768447876,
+      "step": 643
+    },
+    {
+      "epoch": 0.1402986429446739,
+      "grad_norm": 0.6796875,
+      "learning_rate": 7.840604634418968e-05,
+      "loss": 1.247737169265747,
+      "step": 644
+    },
+    {
+      "epoch": 0.14051649798030227,
+      "grad_norm": 0.66796875,
+      "learning_rate": 7.839976987890555e-05,
+      "loss": 1.2373405694961548,
+      "step": 645
+    },
+    {
+      "epoch": 0.14073435301593065,
+      "grad_norm": 0.7421875,
+      "learning_rate": 7.839348133277925e-05,
+      "loss": 1.4079170227050781,
+      "step": 646
+    },
+    {
+      "epoch": 0.14095220805155903,
+      "grad_norm": 0.7421875,
+      "learning_rate": 7.83871807077892e-05,
+      "loss": 1.3397151231765747,
+      "step": 647
+    },
+    {
+      "epoch": 0.1411700630871874,
+      "grad_norm": 0.75,
+      "learning_rate": 7.838086800591764e-05,
+      "loss": 1.2201815843582153,
+      "step": 648
+    },
+    {
+      "epoch": 0.1413879181228158,
+      "grad_norm": 0.71875,
+      "learning_rate": 7.837454322915055e-05,
+      "loss": 1.3436001539230347,
+      "step": 649
+    },
+    {
+      "epoch": 0.14160577315844416,
+      "grad_norm": 0.6796875,
+      "learning_rate": 7.83682063794778e-05,
+      "loss": 1.1800001859664917,
+      "step": 650
+    },
+    {
+      "epoch": 0.14182362819407252,
+      "grad_norm": 0.74609375,
+      "learning_rate": 7.836185745889298e-05,
+      "loss": 1.4116346836090088,
+      "step": 651
+    },
+    {
+      "epoch": 0.1420414832297009,
+      "grad_norm": 0.66015625,
+      "learning_rate": 7.835549646939351e-05,
+      "loss": 1.175708532333374,
+      "step": 652
+    },
+    {
+      "epoch": 0.14225933826532927,
+      "grad_norm": 0.65234375,
+      "learning_rate": 7.83491234129806e-05,
+      "loss": 1.05419921875,
+      "step": 653
+    },
+    {
+      "epoch": 0.14247719330095765,
+      "grad_norm": 0.69140625,
+      "learning_rate": 7.834273829165927e-05,
+      "loss": 1.0953856706619263,
+      "step": 654
+    },
+    {
+      "epoch": 0.14269504833658603,
+      "grad_norm": 0.75,
+      "learning_rate": 7.833634110743832e-05,
+      "loss": 1.4264826774597168,
+      "step": 655
+    },
+    {
+      "epoch": 0.1429129033722144,
+      "grad_norm": 0.69140625,
+      "learning_rate": 7.832993186233036e-05,
+      "loss": 1.2560495138168335,
+      "step": 656
+    },
+    {
+      "epoch": 0.14313075840784278,
+      "grad_norm": 0.75390625,
+      "learning_rate": 7.832351055835174e-05,
+      "loss": 1.2431122064590454,
+      "step": 657
+    },
+    {
+      "epoch": 0.14334861344347116,
+      "grad_norm": 0.68359375,
+      "learning_rate": 7.831707719752271e-05,
+      "loss": 1.3363019227981567,
+      "step": 658
+    },
+    {
+      "epoch": 0.14356646847909954,
+      "grad_norm": 0.71484375,
+      "learning_rate": 7.831063178186722e-05,
+      "loss": 1.1664048433303833,
+      "step": 659
+    },
+    {
+      "epoch": 0.14378432351472792,
+      "grad_norm": 0.69921875,
+      "learning_rate": 7.830417431341303e-05,
+      "loss": 1.2564953565597534,
+      "step": 660
+    },
+    {
+      "epoch": 0.1440021785503563,
+      "grad_norm": 0.66796875,
+      "learning_rate": 7.829770479419173e-05,
+      "loss": 1.2964409589767456,
+      "step": 661
+    },
+    {
+      "epoch": 0.14422003358598465,
+      "grad_norm": 0.69140625,
+      "learning_rate": 7.829122322623865e-05,
+      "loss": 1.2010859251022339,
+      "step": 662
+    },
+    {
+      "epoch": 0.14443788862161303,
+      "grad_norm": 0.71875,
+      "learning_rate": 7.8284729611593e-05,
+      "loss": 1.3098257780075073,
+      "step": 663
+    },
+    {
+      "epoch": 0.1446557436572414,
+      "grad_norm": 0.6875,
+      "learning_rate": 7.827822395229764e-05,
+      "loss": 1.3169783353805542,
+      "step": 664
+    },
+    {
+      "epoch": 0.14487359869286978,
+      "grad_norm": 0.7109375,
+      "learning_rate": 7.827170625039935e-05,
+      "loss": 1.2674981355667114,
+      "step": 665
+    },
+    {
+      "epoch": 0.14509145372849816,
+      "grad_norm": 0.7265625,
+      "learning_rate": 7.826517650794861e-05,
+      "loss": 1.274964690208435,
+      "step": 666
+    },
+    {
+      "epoch": 0.14530930876412654,
+      "grad_norm": 0.6171875,
+      "learning_rate": 7.825863472699974e-05,
+      "loss": 1.0965900421142578,
+      "step": 667
+    },
+    {
+      "epoch": 0.14552716379975492,
+      "grad_norm": 0.671875,
+      "learning_rate": 7.825208090961083e-05,
+      "loss": 1.2341606616973877,
+      "step": 668
+    },
+    {
+      "epoch": 0.1457450188353833,
+      "grad_norm": 0.76171875,
+      "learning_rate": 7.824551505784377e-05,
+      "loss": 1.3817017078399658,
+      "step": 669
+    },
+    {
+      "epoch": 0.14596287387101167,
+      "grad_norm": 0.68359375,
+      "learning_rate": 7.823893717376422e-05,
+      "loss": 1.2202906608581543,
+      "step": 670
+    },
+    {
+      "epoch": 0.14618072890664005,
+      "grad_norm": 0.66015625,
+      "learning_rate": 7.82323472594416e-05,
+      "loss": 1.1820346117019653,
+      "step": 671
+    },
+    {
+      "epoch": 0.14639858394226843,
+      "grad_norm": 0.6875,
+      "learning_rate": 7.822574531694918e-05,
+      "loss": 1.1202290058135986,
+      "step": 672
+    },
+    {
+      "epoch": 0.14661643897789678,
+      "grad_norm": 0.68359375,
+      "learning_rate": 7.821913134836397e-05,
+      "loss": 1.2116087675094604,
+      "step": 673
+    },
+    {
+      "epoch": 0.14683429401352516,
+      "grad_norm": 0.640625,
+      "learning_rate": 7.821250535576676e-05,
+      "loss": 1.1239010095596313,
+      "step": 674
+    },
+    {
+      "epoch": 0.14705214904915354,
+      "grad_norm": 0.6953125,
+      "learning_rate": 7.820586734124214e-05,
+      "loss": 1.2283275127410889,
+      "step": 675
+    },
+    {
+      "epoch": 0.1472700040847819,
+      "grad_norm": 0.6875,
+      "learning_rate": 7.819921730687848e-05,
+      "loss": 1.2085778713226318,
+      "step": 676
+    },
+    {
+      "epoch": 0.1474878591204103,
+      "grad_norm": 0.70703125,
+      "learning_rate": 7.819255525476792e-05,
+      "loss": 1.4613054990768433,
+      "step": 677
+    },
+    {
+      "epoch": 0.14770571415603867,
+      "grad_norm": 0.6875,
+      "learning_rate": 7.818588118700639e-05,
+      "loss": 1.3091479539871216,
+      "step": 678
+    },
+    {
+      "epoch": 0.14792356919166705,
+      "grad_norm": 0.72265625,
+      "learning_rate": 7.817919510569363e-05,
+      "loss": 1.1988037824630737,
+      "step": 679
+    },
+    {
+      "epoch": 0.14814142422729543,
+      "grad_norm": 0.72265625,
+      "learning_rate": 7.817249701293308e-05,
+      "loss": 1.4027717113494873,
+      "step": 680
+    },
+    {
+      "epoch": 0.1483592792629238,
+      "grad_norm": 0.69140625,
+      "learning_rate": 7.816578691083205e-05,
+      "loss": 1.1985427141189575,
+      "step": 681
+    },
+    {
+      "epoch": 0.14857713429855218,
+      "grad_norm": 0.68359375,
+      "learning_rate": 7.815906480150155e-05,
+      "loss": 1.2810072898864746,
+      "step": 682
+    },
+    {
+      "epoch": 0.14879498933418056,
+      "grad_norm": 0.63671875,
+      "learning_rate": 7.815233068705644e-05,
+      "loss": 1.0361299514770508,
+      "step": 683
+    },
+    {
+      "epoch": 0.1490128443698089,
+      "grad_norm": 0.6875,
+      "learning_rate": 7.81455845696153e-05,
+      "loss": 1.247094988822937,
+      "step": 684
+    },
+    {
+      "epoch": 0.1492306994054373,
+      "grad_norm": 0.65625,
+      "learning_rate": 7.813882645130049e-05,
+      "loss": 1.1444871425628662,
+      "step": 685
+    },
+    {
+      "epoch": 0.14944855444106567,
+      "grad_norm": 0.67578125,
+      "learning_rate": 7.81320563342382e-05,
+      "loss": 1.1886913776397705,
+      "step": 686
+    },
+    {
+      "epoch": 0.14966640947669405,
+      "grad_norm": 0.73046875,
+      "learning_rate": 7.812527422055832e-05,
+      "loss": 1.2559126615524292,
+      "step": 687
+    },
+    {
+      "epoch": 0.14988426451232242,
+      "grad_norm": 0.72265625,
+      "learning_rate": 7.81184801123946e-05,
+      "loss": 1.366111397743225,
+      "step": 688
+    },
+    {
+      "epoch": 0.1501021195479508,
+      "grad_norm": 0.71875,
+      "learning_rate": 7.811167401188446e-05,
+      "loss": 1.2285182476043701,
+      "step": 689
+    },
+    {
+      "epoch": 0.15031997458357918,
+      "grad_norm": 0.70703125,
+      "learning_rate": 7.810485592116918e-05,
+      "loss": 1.2753976583480835,
+      "step": 690
+    },
+    {
+      "epoch": 0.15053782961920756,
+      "grad_norm": 0.703125,
+      "learning_rate": 7.809802584239377e-05,
+      "loss": 1.25270414352417,
+      "step": 691
+    },
+    {
+      "epoch": 0.15075568465483594,
+      "grad_norm": 0.71875,
+      "learning_rate": 7.809118377770702e-05,
+      "loss": 1.212296485900879,
+      "step": 692
+    },
+    {
+      "epoch": 0.15097353969046431,
+      "grad_norm": 0.7578125,
+      "learning_rate": 7.808432972926152e-05,
+      "loss": 1.273385763168335,
+      "step": 693
+    },
+    {
+      "epoch": 0.1511913947260927,
+      "grad_norm": 0.69140625,
+      "learning_rate": 7.807746369921357e-05,
+      "loss": 1.201207160949707,
+      "step": 694
+    },
+    {
+      "epoch": 0.15140924976172104,
+      "grad_norm": 0.6796875,
+      "learning_rate": 7.807058568972327e-05,
+      "loss": 1.3076362609863281,
+      "step": 695
+    },
+    {
+      "epoch": 0.15162710479734942,
+      "grad_norm": 0.7265625,
+      "learning_rate": 7.80636957029545e-05,
+      "loss": 1.2322124242782593,
+      "step": 696
+    },
+    {
+      "epoch": 0.1518449598329778,
+      "grad_norm": 0.68359375,
+      "learning_rate": 7.805679374107491e-05,
+      "loss": 1.3673020601272583,
+      "step": 697
+    },
+    {
+      "epoch": 0.15206281486860618,
+      "grad_norm": 0.6953125,
+      "learning_rate": 7.804987980625591e-05,
+      "loss": 1.1426410675048828,
+      "step": 698
+    },
+    {
+      "epoch": 0.15228066990423456,
+      "grad_norm": 0.65234375,
+      "learning_rate": 7.804295390067265e-05,
+      "loss": 1.133246898651123,
+      "step": 699
+    },
+    {
+      "epoch": 0.15249852493986293,
+      "grad_norm": 1.0703125,
+      "learning_rate": 7.80360160265041e-05,
+      "loss": 1.2627530097961426,
+      "step": 700
+    },
+    {
+      "epoch": 0.1527163799754913,
+      "grad_norm": 0.6953125,
+      "learning_rate": 7.802906618593291e-05,
+      "loss": 1.3805313110351562,
+      "step": 701
+    },
+    {
+      "epoch": 0.1529342350111197,
+      "grad_norm": 0.7265625,
+      "learning_rate": 7.80221043811456e-05,
+      "loss": 1.2462581396102905,
+      "step": 702
+    },
+    {
+      "epoch": 0.15315209004674807,
+      "grad_norm": 0.66796875,
+      "learning_rate": 7.801513061433241e-05,
+      "loss": 1.1993404626846313,
+      "step": 703
+    },
+    {
+      "epoch": 0.15336994508237645,
+      "grad_norm": 0.7265625,
+      "learning_rate": 7.80081448876873e-05,
+      "loss": 1.2287969589233398,
+      "step": 704
+    },
+    {
+      "epoch": 0.15358780011800482,
+      "grad_norm": 0.69921875,
+      "learning_rate": 7.800114720340805e-05,
+      "loss": 1.1585383415222168,
+      "step": 705
+    },
+    {
+      "epoch": 0.15380565515363318,
+      "grad_norm": 0.64453125,
+      "learning_rate": 7.799413756369616e-05,
+      "loss": 1.086818814277649,
+      "step": 706
+    },
+    {
+      "epoch": 0.15402351018926155,
+      "grad_norm": 0.703125,
+      "learning_rate": 7.798711597075695e-05,
+      "loss": 1.214449405670166,
+      "step": 707
+    },
+    {
+      "epoch": 0.15424136522488993,
+      "grad_norm": 0.66015625,
+      "learning_rate": 7.798008242679943e-05,
+      "loss": 1.1503362655639648,
+      "step": 708
+    },
+    {
+      "epoch": 0.1544592202605183,
+      "grad_norm": 0.6875,
+      "learning_rate": 7.797303693403644e-05,
+      "loss": 1.379593014717102,
+      "step": 709
+    },
+    {
+      "epoch": 0.1546770752961467,
+      "grad_norm": 0.69921875,
+      "learning_rate": 7.79659794946845e-05,
+      "loss": 1.4684087038040161,
+      "step": 710
+    },
+    {
+      "epoch": 0.15489493033177507,
+      "grad_norm": 0.65625,
+      "learning_rate": 7.795891011096394e-05,
+      "loss": 1.1138107776641846,
+      "step": 711
+    },
+    {
+      "epoch": 0.15511278536740344,
+      "grad_norm": 0.70703125,
+      "learning_rate": 7.795182878509886e-05,
+      "loss": 1.2464227676391602,
+      "step": 712
+    },
+    {
+      "epoch": 0.15533064040303182,
+      "grad_norm": 0.77734375,
+      "learning_rate": 7.79447355193171e-05,
+      "loss": 1.3661491870880127,
+      "step": 713
+    },
+    {
+      "epoch": 0.1555484954386602,
+      "grad_norm": 0.69921875,
+      "learning_rate": 7.793763031585022e-05,
+      "loss": 1.3169022798538208,
+      "step": 714
+    },
+    {
+      "epoch": 0.15576635047428858,
+      "grad_norm": 0.63671875,
+      "learning_rate": 7.793051317693359e-05,
+      "loss": 1.1428102254867554,
+      "step": 715
+    },
+    {
+      "epoch": 0.15598420550991693,
+      "grad_norm": 0.671875,
+      "learning_rate": 7.792338410480629e-05,
+      "loss": 1.1966216564178467,
+      "step": 716
+    },
+    {
+      "epoch": 0.1562020605455453,
+      "grad_norm": 0.71875,
+      "learning_rate": 7.791624310171121e-05,
+      "loss": 1.2206647396087646,
+      "step": 717
+    },
+    {
+      "epoch": 0.15641991558117369,
+      "grad_norm": 0.6484375,
+      "learning_rate": 7.790909016989494e-05,
+      "loss": 1.1341267824172974,
+      "step": 718
+    },
+    {
+      "epoch": 0.15663777061680206,
+      "grad_norm": 0.6171875,
+      "learning_rate": 7.790192531160785e-05,
+      "loss": 0.9939894676208496,
+      "step": 719
+    },
+    {
+      "epoch": 0.15685562565243044,
+      "grad_norm": 0.6328125,
+      "learning_rate": 7.789474852910405e-05,
+      "loss": 1.0250625610351562,
+      "step": 720
+    },
+    {
+      "epoch": 0.15707348068805882,
+      "grad_norm": 0.67578125,
+      "learning_rate": 7.788755982464142e-05,
+      "loss": 1.2326240539550781,
+      "step": 721
+    },
+    {
+      "epoch": 0.1572913357236872,
+      "grad_norm": 0.7109375,
+      "learning_rate": 7.788035920048157e-05,
+      "loss": 1.0759565830230713,
+      "step": 722
+    },
+    {
+      "epoch": 0.15750919075931558,
+      "grad_norm": 0.70703125,
+      "learning_rate": 7.787314665888984e-05,
+      "loss": 1.2746435403823853,
+      "step": 723
+    },
+    {
+      "epoch": 0.15772704579494395,
+      "grad_norm": 0.69140625,
+      "learning_rate": 7.78659222021354e-05,
+      "loss": 1.377940058708191,
+      "step": 724
+    },
+    {
+      "epoch": 0.15794490083057233,
+      "grad_norm": 0.65234375,
+      "learning_rate": 7.785868583249109e-05,
+      "loss": 1.1879512071609497,
+      "step": 725
+    },
+    {
+      "epoch": 0.1581627558662007,
+      "grad_norm": 0.62109375,
+      "learning_rate": 7.785143755223351e-05,
+      "loss": 1.1852167844772339,
+      "step": 726
+    },
+    {
+      "epoch": 0.15838061090182906,
+      "grad_norm": 0.69140625,
+      "learning_rate": 7.784417736364304e-05,
+      "loss": 1.2134246826171875,
+      "step": 727
+    },
+    {
+      "epoch": 0.15859846593745744,
+      "grad_norm": 0.7109375,
+      "learning_rate": 7.783690526900378e-05,
+      "loss": 1.2192740440368652,
+      "step": 728
+    },
+    {
+      "epoch": 0.15881632097308582,
+      "grad_norm": 0.671875,
+      "learning_rate": 7.782962127060358e-05,
+      "loss": 1.2214305400848389,
+      "step": 729
+    },
+    {
+      "epoch": 0.1590341760087142,
+      "grad_norm": 0.68359375,
+      "learning_rate": 7.782232537073406e-05,
+      "loss": 1.1630748510360718,
+      "step": 730
+    },
+    {
+      "epoch": 0.15925203104434257,
+      "grad_norm": 0.671875,
+      "learning_rate": 7.781501757169052e-05,
+      "loss": 1.152377963066101,
+      "step": 731
+    },
+    {
+      "epoch": 0.15946988607997095,
+      "grad_norm": 0.78125,
+      "learning_rate": 7.780769787577209e-05,
+      "loss": 1.3253458738327026,
+      "step": 732
+    },
+    {
+      "epoch": 0.15968774111559933,
+      "grad_norm": 0.87890625,
+      "learning_rate": 7.780036628528156e-05,
+      "loss": 1.2931890487670898,
+      "step": 733
+    },
+    {
+      "epoch": 0.1599055961512277,
+      "grad_norm": 0.6875,
+      "learning_rate": 7.779302280252554e-05,
+      "loss": 1.1821870803833008,
+      "step": 734
+    },
+    {
+      "epoch": 0.1601234511868561,
+      "grad_norm": 0.734375,
+      "learning_rate": 7.77856674298143e-05,
+      "loss": 1.2504937648773193,
+      "step": 735
+    },
+    {
+      "epoch": 0.16034130622248446,
+      "grad_norm": 0.6875,
+      "learning_rate": 7.777830016946193e-05,
+      "loss": 1.225614070892334,
+      "step": 736
+    },
+    {
+      "epoch": 0.16055916125811284,
+      "grad_norm": 0.75390625,
+      "learning_rate": 7.77709210237862e-05,
+      "loss": 1.2594791650772095,
+      "step": 737
+    },
+    {
+      "epoch": 0.1607770162937412,
+      "grad_norm": 0.67578125,
+      "learning_rate": 7.776352999510867e-05,
+      "loss": 1.2317826747894287,
+      "step": 738
+    },
+    {
+      "epoch": 0.16099487132936957,
+      "grad_norm": 0.72265625,
+      "learning_rate": 7.775612708575455e-05,
+      "loss": 1.3100180625915527,
+      "step": 739
+    },
+    {
+      "epoch": 0.16121272636499795,
+      "grad_norm": 0.69140625,
+      "learning_rate": 7.774871229805293e-05,
+      "loss": 1.202368140220642,
+      "step": 740
+    },
+    {
+      "epoch": 0.16143058140062633,
+      "grad_norm": 0.671875,
+      "learning_rate": 7.774128563433648e-05,
+      "loss": 1.14451265335083,
+      "step": 741
+    },
+    {
+      "epoch": 0.1616484364362547,
+      "grad_norm": 0.6640625,
+      "learning_rate": 7.773384709694174e-05,
+      "loss": 1.2216354608535767,
+      "step": 742
+    },
+    {
+      "epoch": 0.16186629147188308,
+      "grad_norm": 0.671875,
+      "learning_rate": 7.772639668820889e-05,
+      "loss": 1.1377766132354736,
+      "step": 743
+    },
+    {
+      "epoch": 0.16208414650751146,
+      "grad_norm": 0.70703125,
+      "learning_rate": 7.77189344104819e-05,
+      "loss": 1.4658130407333374,
+      "step": 744
+    },
+    {
+      "epoch": 0.16230200154313984,
+      "grad_norm": 0.7421875,
+      "learning_rate": 7.771146026610844e-05,
+      "loss": 1.206356167793274,
+      "step": 745
+    },
+    {
+      "epoch": 0.16251985657876822,
+      "grad_norm": 0.6875,
+      "learning_rate": 7.770397425743994e-05,
+      "loss": 1.1080701351165771,
+      "step": 746
+    },
+    {
+      "epoch": 0.1627377116143966,
+      "grad_norm": 0.75390625,
+      "learning_rate": 7.769647638683156e-05,
+      "loss": 1.183564305305481,
+      "step": 747
+    },
+    {
+      "epoch": 0.16295556665002497,
+      "grad_norm": 0.71875,
+      "learning_rate": 7.768896665664216e-05,
+      "loss": 1.179029941558838,
+      "step": 748
+    },
+    {
+      "epoch": 0.16317342168565332,
+      "grad_norm": 0.640625,
+      "learning_rate": 7.768144506923439e-05,
+      "loss": 1.0771626234054565,
+      "step": 749
+    },
+    {
+      "epoch": 0.1633912767212817,
+      "grad_norm": 0.7109375,
+      "learning_rate": 7.767391162697457e-05,
+      "loss": 1.2191189527511597,
+      "step": 750
+    },
+    {
+      "epoch": 0.16360913175691008,
+      "grad_norm": 0.6328125,
+      "learning_rate": 7.766636633223275e-05,
+      "loss": 1.07988440990448,
+      "step": 751
+    },
+    {
+      "epoch": 0.16382698679253846,
+      "grad_norm": 0.671875,
+      "learning_rate": 7.76588091873828e-05,
+      "loss": 1.3188848495483398,
+      "step": 752
+    },
+    {
+      "epoch": 0.16404484182816684,
+      "grad_norm": 1.1171875,
+      "learning_rate": 7.765124019480219e-05,
+      "loss": 1.1284756660461426,
+      "step": 753
+    },
+    {
+      "epoch": 0.16426269686379522,
+      "grad_norm": 0.6953125,
+      "learning_rate": 7.764365935687222e-05,
+      "loss": 1.336605429649353,
+      "step": 754
+    },
+    {
+      "epoch": 0.1644805518994236,
+      "grad_norm": 0.6875,
+      "learning_rate": 7.763606667597785e-05,
+      "loss": 1.2390000820159912,
+      "step": 755
+    },
+    {
+      "epoch": 0.16469840693505197,
+      "grad_norm": 0.69140625,
+      "learning_rate": 7.762846215450782e-05,
+      "loss": 1.3190333843231201,
+      "step": 756
+    },
+    {
+      "epoch": 0.16491626197068035,
+      "grad_norm": 0.63671875,
+      "learning_rate": 7.762084579485452e-05,
+      "loss": 1.0820761919021606,
+      "step": 757
+    },
+    {
+      "epoch": 0.16513411700630873,
+      "grad_norm": 0.71484375,
+      "learning_rate": 7.761321759941414e-05,
+      "loss": 1.1396974325180054,
+      "step": 758
+    },
+    {
+      "epoch": 0.1653519720419371,
+      "grad_norm": 0.69921875,
+      "learning_rate": 7.760557757058659e-05,
+      "loss": 1.2285616397857666,
+      "step": 759
+    },
+    {
+      "epoch": 0.16556982707756546,
+      "grad_norm": 0.68359375,
+      "learning_rate": 7.759792571077545e-05,
+      "loss": 1.3036624193191528,
+      "step": 760
+    },
+    {
+      "epoch": 0.16578768211319383,
+      "grad_norm": 0.74609375,
+      "learning_rate": 7.759026202238803e-05,
+      "loss": 1.3950209617614746,
+      "step": 761
+    },
+    {
+      "epoch": 0.1660055371488222,
+      "grad_norm": 0.66796875,
+      "learning_rate": 7.758258650783541e-05,
+      "loss": 1.178895354270935,
+      "step": 762
+    },
+    {
+      "epoch": 0.1662233921844506,
+      "grad_norm": 0.6484375,
+      "learning_rate": 7.757489916953237e-05,
+      "loss": 1.219315767288208,
+      "step": 763
+    },
+    {
+      "epoch": 0.16644124722007897,
+      "grad_norm": 0.75390625,
+      "learning_rate": 7.756720000989737e-05,
+      "loss": 1.2343075275421143,
+      "step": 764
+    },
+    {
+      "epoch": 0.16665910225570735,
+      "grad_norm": 0.734375,
+      "learning_rate": 7.755948903135264e-05,
+      "loss": 1.2237434387207031,
+      "step": 765
+    },
+    {
+      "epoch": 0.16687695729133573,
+      "grad_norm": 0.66015625,
+      "learning_rate": 7.755176623632411e-05,
+      "loss": 1.1339753866195679,
+      "step": 766
+    },
+    {
+      "epoch": 0.1670948123269641,
+      "grad_norm": 0.6953125,
+      "learning_rate": 7.754403162724142e-05,
+      "loss": 1.256407380104065,
+      "step": 767
+    },
+    {
+      "epoch": 0.16731266736259248,
+      "grad_norm": 0.625,
+      "learning_rate": 7.753628520653794e-05,
+      "loss": 1.0507087707519531,
+      "step": 768
+    },
+    {
+      "epoch": 0.16753052239822086,
+      "grad_norm": 0.66796875,
+      "learning_rate": 7.752852697665073e-05,
+      "loss": 1.29155695438385,
+      "step": 769
+    },
+    {
+      "epoch": 0.16774837743384924,
+      "grad_norm": 0.6953125,
+      "learning_rate": 7.752075694002061e-05,
+      "loss": 1.2239818572998047,
+      "step": 770
+    },
+    {
+      "epoch": 0.1679662324694776,
+      "grad_norm": 0.703125,
+      "learning_rate": 7.751297509909209e-05,
+      "loss": 1.2349203824996948,
+      "step": 771
+    },
+    {
+      "epoch": 0.16818408750510597,
+      "grad_norm": 0.65625,
+      "learning_rate": 7.750518145631338e-05,
+      "loss": 1.1757333278656006,
+      "step": 772
+    },
+    {
+      "epoch": 0.16840194254073435,
+      "grad_norm": 0.70703125,
+      "learning_rate": 7.74973760141364e-05,
+      "loss": 1.2648236751556396,
+      "step": 773
+    },
+    {
+      "epoch": 0.16861979757636272,
+      "grad_norm": 0.70703125,
+      "learning_rate": 7.748955877501683e-05,
+      "loss": 1.2968645095825195,
+      "step": 774
+    },
+    {
+      "epoch": 0.1688376526119911,
+      "grad_norm": 0.68359375,
+      "learning_rate": 7.748172974141401e-05,
+      "loss": 1.1385730504989624,
+      "step": 775
+    },
+    {
+      "epoch": 0.16905550764761948,
+      "grad_norm": 0.66796875,
+      "learning_rate": 7.747388891579101e-05,
+      "loss": 1.2600194215774536,
+      "step": 776
+    },
+    {
+      "epoch": 0.16927336268324786,
+      "grad_norm": 0.66015625,
+      "learning_rate": 7.746603630061462e-05,
+      "loss": 1.248655915260315,
+      "step": 777
+    },
+    {
+      "epoch": 0.16949121771887624,
+      "grad_norm": 0.6484375,
+      "learning_rate": 7.745817189835532e-05,
+      "loss": 1.0766615867614746,
+      "step": 778
+    },
+    {
+      "epoch": 0.16970907275450461,
+      "grad_norm": 0.7265625,
+      "learning_rate": 7.745029571148732e-05,
+      "loss": 1.287589192390442,
+      "step": 779
+    },
+    {
+      "epoch": 0.169926927790133,
+      "grad_norm": 0.8359375,
+      "learning_rate": 7.74424077424885e-05,
+      "loss": 1.241224765777588,
+      "step": 780
+    },
+    {
+      "epoch": 0.17014478282576137,
+      "grad_norm": 0.65234375,
+      "learning_rate": 7.74345079938405e-05,
+      "loss": 1.1736308336257935,
+      "step": 781
+    },
+    {
+      "epoch": 0.17036263786138972,
+      "grad_norm": 0.6953125,
+      "learning_rate": 7.742659646802863e-05,
+      "loss": 1.2479166984558105,
+      "step": 782
+    },
+    {
+      "epoch": 0.1705804928970181,
+      "grad_norm": 0.66796875,
+      "learning_rate": 7.741867316754189e-05,
+      "loss": 1.2747901678085327,
+      "step": 783
+    },
+    {
+      "epoch": 0.17079834793264648,
+      "grad_norm": 0.65234375,
+      "learning_rate": 7.741073809487303e-05,
+      "loss": 1.1904420852661133,
+      "step": 784
+    },
+    {
+      "epoch": 0.17101620296827486,
+      "grad_norm": 0.65234375,
+      "learning_rate": 7.740279125251846e-05,
+      "loss": 1.189729928970337,
+      "step": 785
+    },
+    {
+      "epoch": 0.17123405800390323,
+      "grad_norm": 0.66796875,
+      "learning_rate": 7.739483264297835e-05,
+      "loss": 1.2796211242675781,
+      "step": 786
+    },
+    {
+      "epoch": 0.1714519130395316,
+      "grad_norm": 0.69921875,
+      "learning_rate": 7.73868622687565e-05,
+      "loss": 1.2826827764511108,
+      "step": 787
+    },
+    {
+      "epoch": 0.17166976807516,
+      "grad_norm": 0.7109375,
+      "learning_rate": 7.737888013236047e-05,
+      "loss": 1.3719871044158936,
+      "step": 788
+    },
+    {
+      "epoch": 0.17188762311078837,
+      "grad_norm": 0.66015625,
+      "learning_rate": 7.737088623630147e-05,
+      "loss": 1.217905044555664,
+      "step": 789
+    },
+    {
+      "epoch": 0.17210547814641675,
+      "grad_norm": 0.671875,
+      "learning_rate": 7.736288058309448e-05,
+      "loss": 1.3556839227676392,
+      "step": 790
+    },
+    {
+      "epoch": 0.17232333318204512,
+      "grad_norm": 0.66015625,
+      "learning_rate": 7.735486317525808e-05,
+      "loss": 1.248610019683838,
+      "step": 791
+    },
+    {
+      "epoch": 0.1725411882176735,
+      "grad_norm": 0.6875,
+      "learning_rate": 7.734683401531466e-05,
+      "loss": 1.2616955041885376,
+      "step": 792
+    },
+    {
+      "epoch": 0.17275904325330185,
+      "grad_norm": 0.77734375,
+      "learning_rate": 7.733879310579021e-05,
+      "loss": 1.2555466890335083,
+      "step": 793
+    },
+    {
+      "epoch": 0.17297689828893023,
+      "grad_norm": 0.68359375,
+      "learning_rate": 7.733074044921447e-05,
+      "loss": 1.2149146795272827,
+      "step": 794
+    },
+    {
+      "epoch": 0.1731947533245586,
+      "grad_norm": 0.67578125,
+      "learning_rate": 7.732267604812086e-05,
+      "loss": 1.2069448232650757,
+      "step": 795
+    },
+    {
+      "epoch": 0.173412608360187,
+      "grad_norm": 0.67578125,
+      "learning_rate": 7.731459990504652e-05,
+      "loss": 1.2715003490447998,
+      "step": 796
+    },
+    {
+      "epoch": 0.17363046339581537,
+      "grad_norm": 0.6484375,
+      "learning_rate": 7.730651202253225e-05,
+      "loss": 1.1273269653320312,
+      "step": 797
+    },
+    {
+      "epoch": 0.17384831843144374,
+      "grad_norm": 0.7109375,
+      "learning_rate": 7.729841240312253e-05,
+      "loss": 1.3250832557678223,
+      "step": 798
+    },
+    {
+      "epoch": 0.17406617346707212,
+      "grad_norm": 0.67578125,
+      "learning_rate": 7.72903010493656e-05,
+      "loss": 1.0802199840545654,
+      "step": 799
+    },
+    {
+      "epoch": 0.1742840285027005,
+      "grad_norm": 0.640625,
+      "learning_rate": 7.728217796381332e-05,
+      "loss": 1.2309858798980713,
+      "step": 800
+    },
+    {
+      "epoch": 0.17450188353832888,
+      "grad_norm": 0.6484375,
+      "learning_rate": 7.727404314902128e-05,
+      "loss": 1.255175232887268,
+      "step": 801
+    },
+    {
+      "epoch": 0.17471973857395726,
+      "grad_norm": 0.6796875,
+      "learning_rate": 7.726589660754876e-05,
+      "loss": 1.2474976778030396,
+      "step": 802
+    },
+    {
+      "epoch": 0.17493759360958563,
+      "grad_norm": 0.66015625,
+      "learning_rate": 7.72577383419587e-05,
+      "loss": 1.207207441329956,
+      "step": 803
+    },
+    {
+      "epoch": 0.17515544864521398,
+      "grad_norm": 0.6875,
+      "learning_rate": 7.724956835481776e-05,
+      "loss": 1.1772822141647339,
+      "step": 804
+    },
+    {
+      "epoch": 0.17537330368084236,
+      "grad_norm": 0.6796875,
+      "learning_rate": 7.724138664869628e-05,
+      "loss": 1.1280514001846313,
+      "step": 805
+    },
+    {
+      "epoch": 0.17559115871647074,
+      "grad_norm": 0.73046875,
+      "learning_rate": 7.723319322616829e-05,
+      "loss": 1.258151888847351,
+      "step": 806
+    },
+    {
+      "epoch": 0.17580901375209912,
+      "grad_norm": 0.73046875,
+      "learning_rate": 7.722498808981149e-05,
+      "loss": 1.0578664541244507,
+      "step": 807
+    },
+    {
+      "epoch": 0.1760268687877275,
+      "grad_norm": 0.6875,
+      "learning_rate": 7.721677124220727e-05,
+      "loss": 1.116953730583191,
+      "step": 808
+    },
+    {
+      "epoch": 0.17624472382335588,
+      "grad_norm": 0.6875,
+      "learning_rate": 7.72085426859407e-05,
+      "loss": 1.1710602045059204,
+      "step": 809
+    },
+    {
+      "epoch": 0.17646257885898425,
+      "grad_norm": 0.66796875,
+      "learning_rate": 7.720030242360057e-05,
+      "loss": 1.165201187133789,
+      "step": 810
+    },
+    {
+      "epoch": 0.17668043389461263,
+      "grad_norm": 0.69140625,
+      "learning_rate": 7.71920504577793e-05,
+      "loss": 1.3554954528808594,
+      "step": 811
+    },
+    {
+      "epoch": 0.176898288930241,
+      "grad_norm": 0.625,
+      "learning_rate": 7.718378679107304e-05,
+      "loss": 1.104645013809204,
+      "step": 812
+    },
+    {
+      "epoch": 0.1771161439658694,
+      "grad_norm": 0.75390625,
+      "learning_rate": 7.717551142608157e-05,
+      "loss": 1.2338337898254395,
+      "step": 813
+    },
+    {
+      "epoch": 0.17733399900149777,
+      "grad_norm": 0.73828125,
+      "learning_rate": 7.716722436540842e-05,
+      "loss": 1.1284233331680298,
+      "step": 814
+    },
+    {
+      "epoch": 0.17755185403712612,
+      "grad_norm": 0.6796875,
+      "learning_rate": 7.715892561166072e-05,
+      "loss": 1.11342453956604,
+      "step": 815
+    },
+    {
+      "epoch": 0.1777697090727545,
+      "grad_norm": 0.875,
+      "learning_rate": 7.715061516744933e-05,
+      "loss": 1.1626187562942505,
+      "step": 816
+    },
+    {
+      "epoch": 0.17798756410838287,
+      "grad_norm": 0.83203125,
+      "learning_rate": 7.714229303538877e-05,
+      "loss": 1.2421140670776367,
+      "step": 817
+    },
+    {
+      "epoch": 0.17820541914401125,
+      "grad_norm": 0.66015625,
+      "learning_rate": 7.713395921809727e-05,
+      "loss": 1.1241856813430786,
+      "step": 818
+    },
+    {
+      "epoch": 0.17842327417963963,
+      "grad_norm": 0.765625,
+      "learning_rate": 7.712561371819668e-05,
+      "loss": 1.2550634145736694,
+      "step": 819
+    },
+    {
+      "epoch": 0.178641129215268,
+      "grad_norm": 0.6484375,
+      "learning_rate": 7.711725653831255e-05,
+      "loss": 1.0861536264419556,
+      "step": 820
+    },
+    {
+      "epoch": 0.17885898425089639,
+      "grad_norm": 0.6796875,
+      "learning_rate": 7.710888768107414e-05,
+      "loss": 1.2194087505340576,
+      "step": 821
+    },
+    {
+      "epoch": 0.17907683928652476,
+      "grad_norm": 0.6484375,
+      "learning_rate": 7.710050714911433e-05,
+      "loss": 1.1688408851623535,
+      "step": 822
+    },
+    {
+      "epoch": 0.17929469432215314,
+      "grad_norm": 0.6953125,
+      "learning_rate": 7.70921149450697e-05,
+      "loss": 1.196813941001892,
+      "step": 823
+    },
+    {
+      "epoch": 0.17951254935778152,
+      "grad_norm": 0.66015625,
+      "learning_rate": 7.708371107158048e-05,
+      "loss": 1.1441730260849,
+      "step": 824
+    },
+    {
+      "epoch": 0.1797304043934099,
+      "grad_norm": 0.60546875,
+      "learning_rate": 7.70752955312906e-05,
+      "loss": 1.0849382877349854,
+      "step": 825
+    },
+    {
+      "epoch": 0.17994825942903825,
+      "grad_norm": 0.72265625,
+      "learning_rate": 7.706686832684768e-05,
+      "loss": 1.2383805513381958,
+      "step": 826
+    },
+    {
+      "epoch": 0.18016611446466663,
+      "grad_norm": 0.671875,
+      "learning_rate": 7.705842946090292e-05,
+      "loss": 1.2954448461532593,
+      "step": 827
+    },
+    {
+      "epoch": 0.180383969500295,
+      "grad_norm": 0.6640625,
+      "learning_rate": 7.704997893611129e-05,
+      "loss": 1.171963095664978,
+      "step": 828
+    },
+    {
+      "epoch": 0.18060182453592338,
+      "grad_norm": 0.72265625,
+      "learning_rate": 7.704151675513138e-05,
+      "loss": 1.2075815200805664,
+      "step": 829
+    },
+    {
+      "epoch": 0.18081967957155176,
+      "grad_norm": 0.6640625,
+      "learning_rate": 7.703304292062546e-05,
+      "loss": 1.2456014156341553,
+      "step": 830
+    },
+    {
+      "epoch": 0.18103753460718014,
+      "grad_norm": 0.796875,
+      "learning_rate": 7.702455743525944e-05,
+      "loss": 1.2049517631530762,
+      "step": 831
+    },
+    {
+      "epoch": 0.18125538964280852,
+      "grad_norm": 0.6640625,
+      "learning_rate": 7.70160603017029e-05,
+      "loss": 1.1813679933547974,
+      "step": 832
+    },
+    {
+      "epoch": 0.1814732446784369,
+      "grad_norm": 0.6796875,
+      "learning_rate": 7.700755152262915e-05,
+      "loss": 1.2362619638442993,
+      "step": 833
+    },
+    {
+      "epoch": 0.18169109971406527,
+      "grad_norm": 0.765625,
+      "learning_rate": 7.699903110071506e-05,
+      "loss": 1.147566318511963,
+      "step": 834
+    },
+    {
+      "epoch": 0.18190895474969365,
+      "grad_norm": 0.63671875,
+      "learning_rate": 7.699049903864125e-05,
+      "loss": 1.1306720972061157,
+      "step": 835
+    },
+    {
+      "epoch": 0.18212680978532203,
+      "grad_norm": 0.75,
+      "learning_rate": 7.698195533909195e-05,
+      "loss": 1.2871365547180176,
+      "step": 836
+    },
+    {
+      "epoch": 0.18234466482095038,
+      "grad_norm": 0.7109375,
+      "learning_rate": 7.697340000475508e-05,
+      "loss": 1.255440592765808,
+      "step": 837
+    },
+    {
+      "epoch": 0.18256251985657876,
+      "grad_norm": 0.71875,
+      "learning_rate": 7.69648330383222e-05,
+      "loss": 1.2212032079696655,
+      "step": 838
+    },
+    {
+      "epoch": 0.18278037489220714,
+      "grad_norm": 0.66796875,
+      "learning_rate": 7.695625444248855e-05,
+      "loss": 1.1900126934051514,
+      "step": 839
+    },
+    {
+      "epoch": 0.18299822992783552,
+      "grad_norm": 0.671875,
+      "learning_rate": 7.694766421995301e-05,
+      "loss": 1.157047986984253,
+      "step": 840
+    },
+    {
+      "epoch": 0.1832160849634639,
+      "grad_norm": 0.69921875,
+      "learning_rate": 7.693906237341812e-05,
+      "loss": 1.0393284559249878,
+      "step": 841
+    },
+    {
+      "epoch": 0.18343393999909227,
+      "grad_norm": 0.63671875,
+      "learning_rate": 7.693044890559009e-05,
+      "loss": 1.0511822700500488,
+      "step": 842
+    },
+    {
+      "epoch": 0.18365179503472065,
+      "grad_norm": 0.69140625,
+      "learning_rate": 7.692182381917879e-05,
+      "loss": 1.3760817050933838,
+      "step": 843
+    },
+    {
+      "epoch": 0.18386965007034903,
+      "grad_norm": 0.671875,
+      "learning_rate": 7.69131871168977e-05,
+      "loss": 1.1491681337356567,
+      "step": 844
+    },
+    {
+      "epoch": 0.1840875051059774,
+      "grad_norm": 0.703125,
+      "learning_rate": 7.690453880146401e-05,
+      "loss": 1.2729957103729248,
+      "step": 845
+    },
+    {
+      "epoch": 0.18430536014160578,
+      "grad_norm": 0.6796875,
+      "learning_rate": 7.689587887559856e-05,
+      "loss": 1.338852047920227,
+      "step": 846
+    },
+    {
+      "epoch": 0.18452321517723416,
+      "grad_norm": 0.68359375,
+      "learning_rate": 7.688720734202579e-05,
+      "loss": 1.1981006860733032,
+      "step": 847
+    },
+    {
+      "epoch": 0.1847410702128625,
+      "grad_norm": 0.69921875,
+      "learning_rate": 7.687852420347386e-05,
+      "loss": 1.3390406370162964,
+      "step": 848
+    },
+    {
+      "epoch": 0.1849589252484909,
+      "grad_norm": 0.6953125,
+      "learning_rate": 7.686982946267452e-05,
+      "loss": 1.1533819437026978,
+      "step": 849
+    },
+    {
+      "epoch": 0.18517678028411927,
+      "grad_norm": 0.62109375,
+      "learning_rate": 7.68611231223632e-05,
+      "loss": 1.1671589612960815,
+      "step": 850
+    },
+    {
+      "epoch": 0.18539463531974765,
+      "grad_norm": 0.69140625,
+      "learning_rate": 7.6852405185279e-05,
+      "loss": 1.1430222988128662,
+      "step": 851
+    },
+    {
+      "epoch": 0.18561249035537603,
+      "grad_norm": 0.71484375,
+      "learning_rate": 7.684367565416461e-05,
+      "loss": 1.2375198602676392,
+      "step": 852
+    },
+    {
+      "epoch": 0.1858303453910044,
+      "grad_norm": 0.6953125,
+      "learning_rate": 7.683493453176643e-05,
+      "loss": 1.1181293725967407,
+      "step": 853
+    },
+    {
+      "epoch": 0.18604820042663278,
+      "grad_norm": 0.6640625,
+      "learning_rate": 7.682618182083447e-05,
+      "loss": 1.2665832042694092,
+      "step": 854
+    },
+    {
+      "epoch": 0.18626605546226116,
+      "grad_norm": 0.66015625,
+      "learning_rate": 7.68174175241224e-05,
+      "loss": 1.2641090154647827,
+      "step": 855
+    },
+    {
+      "epoch": 0.18648391049788954,
+      "grad_norm": 0.65625,
+      "learning_rate": 7.680864164438754e-05,
+      "loss": 1.1909894943237305,
+      "step": 856
+    },
+    {
+      "epoch": 0.18670176553351792,
+      "grad_norm": 0.7109375,
+      "learning_rate": 7.679985418439083e-05,
+      "loss": 1.3348753452301025,
+      "step": 857
+    },
+    {
+      "epoch": 0.1869196205691463,
+      "grad_norm": 0.68359375,
+      "learning_rate": 7.679105514689686e-05,
+      "loss": 1.2946274280548096,
+      "step": 858
+    },
+    {
+      "epoch": 0.18713747560477464,
+      "grad_norm": 0.703125,
+      "learning_rate": 7.67822445346739e-05,
+      "loss": 1.3366048336029053,
+      "step": 859
+    },
+    {
+      "epoch": 0.18735533064040302,
+      "grad_norm": 0.6796875,
+      "learning_rate": 7.677342235049378e-05,
+      "loss": 1.1000404357910156,
+      "step": 860
+    },
+    {
+      "epoch": 0.1875731856760314,
+      "grad_norm": 0.66015625,
+      "learning_rate": 7.676458859713208e-05,
+      "loss": 1.1800968647003174,
+      "step": 861
+    },
+    {
+      "epoch": 0.18779104071165978,
+      "grad_norm": 0.66796875,
+      "learning_rate": 7.675574327736792e-05,
+      "loss": 1.310851812362671,
+      "step": 862
+    },
+    {
+      "epoch": 0.18800889574728816,
+      "grad_norm": 0.6796875,
+      "learning_rate": 7.674688639398412e-05,
+      "loss": 1.2554386854171753,
+      "step": 863
+    },
+    {
+      "epoch": 0.18822675078291654,
+      "grad_norm": 0.703125,
+      "learning_rate": 7.673801794976712e-05,
+      "loss": 1.4371528625488281,
+      "step": 864
+    },
+    {
+      "epoch": 0.1884446058185449,
+      "grad_norm": 0.6484375,
+      "learning_rate": 7.672913794750699e-05,
+      "loss": 1.1684446334838867,
+      "step": 865
+    },
+    {
+      "epoch": 0.1886624608541733,
+      "grad_norm": 0.65625,
+      "learning_rate": 7.672024638999745e-05,
+      "loss": 1.139878273010254,
+      "step": 866
+    },
+    {
+      "epoch": 0.18888031588980167,
+      "grad_norm": 0.6796875,
+      "learning_rate": 7.671134328003584e-05,
+      "loss": 1.3184255361557007,
+      "step": 867
+    },
+    {
+      "epoch": 0.18909817092543005,
+      "grad_norm": 0.71875,
+      "learning_rate": 7.670242862042314e-05,
+      "loss": 1.2221405506134033,
+      "step": 868
+    },
+    {
+      "epoch": 0.18931602596105843,
+      "grad_norm": 0.65234375,
+      "learning_rate": 7.669350241396397e-05,
+      "loss": 1.18660306930542,
+      "step": 869
+    },
+    {
+      "epoch": 0.18953388099668678,
+      "grad_norm": 0.6875,
+      "learning_rate": 7.668456466346658e-05,
+      "loss": 1.2348490953445435,
+      "step": 870
+    },
+    {
+      "epoch": 0.18975173603231515,
+      "grad_norm": 0.67578125,
+      "learning_rate": 7.667561537174284e-05,
+      "loss": 1.1676623821258545,
+      "step": 871
+    },
+    {
+      "epoch": 0.18996959106794353,
+      "grad_norm": 0.66796875,
+      "learning_rate": 7.666665454160829e-05,
+      "loss": 1.1586041450500488,
+      "step": 872
+    },
+    {
+      "epoch": 0.1901874461035719,
+      "grad_norm": 0.6796875,
+      "learning_rate": 7.665768217588202e-05,
+      "loss": 1.265610933303833,
+      "step": 873
+    },
+    {
+      "epoch": 0.1904053011392003,
+      "grad_norm": 0.66796875,
+      "learning_rate": 7.664869827738684e-05,
+      "loss": 1.2568941116333008,
+      "step": 874
+    },
+    {
+      "epoch": 0.19062315617482867,
+      "grad_norm": 0.703125,
+      "learning_rate": 7.663970284894916e-05,
+      "loss": 1.1782686710357666,
+      "step": 875
+    },
+    {
+      "epoch": 0.19084101121045705,
+      "grad_norm": 0.7109375,
+      "learning_rate": 7.663069589339897e-05,
+      "loss": 1.1598455905914307,
+      "step": 876
+    },
+    {
+      "epoch": 0.19105886624608542,
+      "grad_norm": 0.6640625,
+      "learning_rate": 7.662167741356994e-05,
+      "loss": 1.152151107788086,
+      "step": 877
+    },
+    {
+      "epoch": 0.1912767212817138,
+      "grad_norm": 0.66015625,
+      "learning_rate": 7.661264741229935e-05,
+      "loss": 1.1954257488250732,
+      "step": 878
+    },
+    {
+      "epoch": 0.19149457631734218,
+      "grad_norm": 0.64453125,
+      "learning_rate": 7.660360589242809e-05,
+      "loss": 1.1556673049926758,
+      "step": 879
+    },
+    {
+      "epoch": 0.19171243135297053,
+      "grad_norm": 0.68359375,
+      "learning_rate": 7.659455285680071e-05,
+      "loss": 1.2872236967086792,
+      "step": 880
+    },
+    {
+      "epoch": 0.1919302863885989,
+      "grad_norm": 0.70703125,
+      "learning_rate": 7.658548830826534e-05,
+      "loss": 1.2802869081497192,
+      "step": 881
+    },
+    {
+      "epoch": 0.1921481414242273,
+      "grad_norm": 0.734375,
+      "learning_rate": 7.657641224967375e-05,
+      "loss": 1.3013818264007568,
+      "step": 882
+    },
+    {
+      "epoch": 0.19236599645985566,
+      "grad_norm": 0.6796875,
+      "learning_rate": 7.656732468388134e-05,
+      "loss": 1.1496351957321167,
+      "step": 883
+    },
+    {
+      "epoch": 0.19258385149548404,
+      "grad_norm": 0.70703125,
+      "learning_rate": 7.655822561374713e-05,
+      "loss": 1.3301492929458618,
+      "step": 884
+    },
+    {
+      "epoch": 0.19280170653111242,
+      "grad_norm": 0.66796875,
+      "learning_rate": 7.654911504213375e-05,
+      "loss": 1.2236876487731934,
+      "step": 885
+    },
+    {
+      "epoch": 0.1930195615667408,
+      "grad_norm": 0.70703125,
+      "learning_rate": 7.653999297190745e-05,
+      "loss": 1.1889899969100952,
+      "step": 886
+    },
+    {
+      "epoch": 0.19323741660236918,
+      "grad_norm": 0.71875,
+      "learning_rate": 7.653085940593808e-05,
+      "loss": 1.221051812171936,
+      "step": 887
+    },
+    {
+      "epoch": 0.19345527163799756,
+      "grad_norm": 0.69921875,
+      "learning_rate": 7.652171434709915e-05,
+      "loss": 1.2867062091827393,
+      "step": 888
+    },
+    {
+      "epoch": 0.19367312667362593,
+      "grad_norm": 0.68359375,
+      "learning_rate": 7.651255779826776e-05,
+      "loss": 1.2820767164230347,
+      "step": 889
+    },
+    {
+      "epoch": 0.1938909817092543,
+      "grad_norm": 0.6875,
+      "learning_rate": 7.650338976232461e-05,
+      "loss": 1.233717441558838,
+      "step": 890
+    },
+    {
+      "epoch": 0.19410883674488266,
+      "grad_norm": 0.62890625,
+      "learning_rate": 7.649421024215404e-05,
+      "loss": 1.0859155654907227,
+      "step": 891
+    },
+    {
+      "epoch": 0.19432669178051104,
+      "grad_norm": 0.73046875,
+      "learning_rate": 7.648501924064399e-05,
+      "loss": 1.3149640560150146,
+      "step": 892
+    },
+    {
+      "epoch": 0.19454454681613942,
+      "grad_norm": 0.6484375,
+      "learning_rate": 7.647581676068601e-05,
+      "loss": 1.2315146923065186,
+      "step": 893
+    },
+    {
+      "epoch": 0.1947624018517678,
+      "grad_norm": 0.703125,
+      "learning_rate": 7.646660280517527e-05,
+      "loss": 1.187409520149231,
+      "step": 894
+    },
+    {
+      "epoch": 0.19498025688739618,
+      "grad_norm": 0.69140625,
+      "learning_rate": 7.645737737701054e-05,
+      "loss": 1.2307687997817993,
+      "step": 895
+    },
+    {
+      "epoch": 0.19519811192302455,
+      "grad_norm": 0.6484375,
+      "learning_rate": 7.644814047909424e-05,
+      "loss": 1.165853500366211,
+      "step": 896
+    },
+    {
+      "epoch": 0.19541596695865293,
+      "grad_norm": 0.671875,
+      "learning_rate": 7.643889211433232e-05,
+      "loss": 1.1534236669540405,
+      "step": 897
+    },
+    {
+      "epoch": 0.1956338219942813,
+      "grad_norm": 0.640625,
+      "learning_rate": 7.64296322856344e-05,
+      "loss": 1.276362657546997,
+      "step": 898
+    },
+    {
+      "epoch": 0.1958516770299097,
+      "grad_norm": 0.66796875,
+      "learning_rate": 7.642036099591368e-05,
+      "loss": 1.2966396808624268,
+      "step": 899
+    },
+    {
+      "epoch": 0.19606953206553807,
+      "grad_norm": 0.6328125,
+      "learning_rate": 7.641107824808699e-05,
+      "loss": 1.1673879623413086,
+      "step": 900
+    },
+    {
+      "epoch": 0.19628738710116644,
+      "grad_norm": 0.72265625,
+      "learning_rate": 7.640178404507473e-05,
+      "loss": 1.251414179801941,
+      "step": 901
+    },
+    {
+      "epoch": 0.1965052421367948,
+      "grad_norm": 0.7265625,
+      "learning_rate": 7.639247838980094e-05,
+      "loss": 1.3393563032150269,
+      "step": 902
+    },
+    {
+      "epoch": 0.19672309717242317,
+      "grad_norm": 0.671875,
+      "learning_rate": 7.638316128519324e-05,
+      "loss": 1.2596877813339233,
+      "step": 903
+    },
+    {
+      "epoch": 0.19694095220805155,
+      "grad_norm": 0.6953125,
+      "learning_rate": 7.637383273418285e-05,
+      "loss": 1.3589540719985962,
+      "step": 904
+    },
+    {
+      "epoch": 0.19715880724367993,
+      "grad_norm": 0.6875,
+      "learning_rate": 7.63644927397046e-05,
+      "loss": 1.1394997835159302,
+      "step": 905
+    },
+    {
+      "epoch": 0.1973766622793083,
+      "grad_norm": 0.6796875,
+      "learning_rate": 7.635514130469693e-05,
+      "loss": 1.2693322896957397,
+      "step": 906
+    },
+    {
+      "epoch": 0.19759451731493669,
+      "grad_norm": 0.65234375,
+      "learning_rate": 7.634577843210187e-05,
+      "loss": 1.2143785953521729,
+      "step": 907
+    },
+    {
+      "epoch": 0.19781237235056506,
+      "grad_norm": 0.62890625,
+      "learning_rate": 7.633640412486502e-05,
+      "loss": 1.1092047691345215,
+      "step": 908
+    },
+    {
+      "epoch": 0.19803022738619344,
+      "grad_norm": 0.671875,
+      "learning_rate": 7.632701838593564e-05,
+      "loss": 1.1766585111618042,
+      "step": 909
+    },
+    {
+      "epoch": 0.19824808242182182,
+      "grad_norm": 0.671875,
+      "learning_rate": 7.631762121826653e-05,
+      "loss": 1.0802528858184814,
+      "step": 910
+    },
+    {
+      "epoch": 0.1984659374574502,
+      "grad_norm": 0.67578125,
+      "learning_rate": 7.63082126248141e-05,
+      "loss": 1.2457014322280884,
+      "step": 911
+    },
+    {
+      "epoch": 0.19868379249307858,
+      "grad_norm": 0.69921875,
+      "learning_rate": 7.629879260853838e-05,
+      "loss": 1.2698745727539062,
+      "step": 912
+    },
+    {
+      "epoch": 0.19890164752870693,
+      "grad_norm": 0.71484375,
+      "learning_rate": 7.628936117240297e-05,
+      "loss": 1.2398320436477661,
+      "step": 913
+    },
+    {
+      "epoch": 0.1991195025643353,
+      "grad_norm": 0.66015625,
+      "learning_rate": 7.627991831937504e-05,
+      "loss": 1.0986076593399048,
+      "step": 914
+    },
+    {
+      "epoch": 0.19933735759996368,
+      "grad_norm": 0.6796875,
+      "learning_rate": 7.627046405242542e-05,
+      "loss": 1.2492597103118896,
+      "step": 915
+    },
+    {
+      "epoch": 0.19955521263559206,
+      "grad_norm": 0.63671875,
+      "learning_rate": 7.626099837452847e-05,
+      "loss": 1.1035375595092773,
+      "step": 916
+    },
+    {
+      "epoch": 0.19977306767122044,
+      "grad_norm": 0.6953125,
+      "learning_rate": 7.625152128866216e-05,
+      "loss": 1.3017324209213257,
+      "step": 917
+    },
+    {
+      "epoch": 0.19999092270684882,
+      "grad_norm": 0.6640625,
+      "learning_rate": 7.624203279780805e-05,
+      "loss": 1.1972295045852661,
+      "step": 918
+    },
+    {
+      "epoch": 0.2002087777424772,
+      "grad_norm": 0.671875,
+      "learning_rate": 7.623253290495128e-05,
+      "loss": 1.2857555150985718,
+      "step": 919
+    },
+    {
+      "epoch": 0.20042663277810557,
+      "grad_norm": 0.72265625,
+      "learning_rate": 7.622302161308059e-05,
+      "loss": 1.2884573936462402,
+      "step": 920
+    },
+    {
+      "epoch": 0.20064448781373395,
+      "grad_norm": 0.64453125,
+      "learning_rate": 7.621349892518831e-05,
+      "loss": 1.1916816234588623,
+      "step": 921
+    },
+    {
+      "epoch": 0.20086234284936233,
+      "grad_norm": 0.6484375,
+      "learning_rate": 7.620396484427033e-05,
+      "loss": 1.3086544275283813,
+      "step": 922
+    },
+    {
+      "epoch": 0.2010801978849907,
+      "grad_norm": 0.640625,
+      "learning_rate": 7.619441937332616e-05,
+      "loss": 1.206362247467041,
+      "step": 923
+    },
+    {
+      "epoch": 0.20129805292061906,
+      "grad_norm": 0.66015625,
+      "learning_rate": 7.618486251535885e-05,
+      "loss": 1.1797014474868774,
+      "step": 924
+    },
+    {
+      "epoch": 0.20151590795624744,
+      "grad_norm": 0.6640625,
+      "learning_rate": 7.617529427337507e-05,
+      "loss": 1.3439568281173706,
+      "step": 925
+    },
+    {
+      "epoch": 0.20173376299187581,
+      "grad_norm": 0.64453125,
+      "learning_rate": 7.616571465038505e-05,
+      "loss": 1.2023500204086304,
+      "step": 926
+    },
+    {
+      "epoch": 0.2019516180275042,
+      "grad_norm": 0.6875,
+      "learning_rate": 7.615612364940263e-05,
+      "loss": 1.346264362335205,
+      "step": 927
+    },
+    {
+      "epoch": 0.20216947306313257,
+      "grad_norm": 0.69140625,
+      "learning_rate": 7.614652127344517e-05,
+      "loss": 1.276180386543274,
+      "step": 928
+    },
+    {
+      "epoch": 0.20238732809876095,
+      "grad_norm": 0.62890625,
+      "learning_rate": 7.613690752553366e-05,
+      "loss": 1.1624594926834106,
+      "step": 929
+    },
+    {
+      "epoch": 0.20260518313438933,
+      "grad_norm": 0.6796875,
+      "learning_rate": 7.612728240869267e-05,
+      "loss": 1.1463534832000732,
+      "step": 930
+    },
+    {
+      "epoch": 0.2028230381700177,
+      "grad_norm": 0.66015625,
+      "learning_rate": 7.611764592595032e-05,
+      "loss": 1.3583242893218994,
+      "step": 931
+    },
+    {
+      "epoch": 0.20304089320564608,
+      "grad_norm": 0.68359375,
+      "learning_rate": 7.610799808033833e-05,
+      "loss": 1.2828558683395386,
+      "step": 932
+    },
+    {
+      "epoch": 0.20325874824127446,
+      "grad_norm": 0.625,
+      "learning_rate": 7.609833887489195e-05,
+      "loss": 1.0727781057357788,
+      "step": 933
+    },
+    {
+      "epoch": 0.20347660327690284,
+      "grad_norm": 0.72265625,
+      "learning_rate": 7.608866831265005e-05,
+      "loss": 1.2828859090805054,
+      "step": 934
+    },
+    {
+      "epoch": 0.2036944583125312,
+      "grad_norm": 0.6640625,
+      "learning_rate": 7.607898639665507e-05,
+      "loss": 1.2264331579208374,
+      "step": 935
+    },
+    {
+      "epoch": 0.20391231334815957,
+      "grad_norm": 0.65234375,
+      "learning_rate": 7.6069293129953e-05,
+      "loss": 1.1783287525177002,
+      "step": 936
+    },
+    {
+      "epoch": 0.20413016838378795,
+      "grad_norm": 0.71484375,
+      "learning_rate": 7.605958851559341e-05,
+      "loss": 1.3368380069732666,
+      "step": 937
+    },
+    {
+      "epoch": 0.20434802341941632,
+      "grad_norm": 0.7109375,
+      "learning_rate": 7.604987255662944e-05,
+      "loss": 1.3347231149673462,
+      "step": 938
+    },
+    {
+      "epoch": 0.2045658784550447,
+      "grad_norm": 0.71484375,
+      "learning_rate": 7.60401452561178e-05,
+      "loss": 1.397395133972168,
+      "step": 939
+    },
+    {
+      "epoch": 0.20478373349067308,
+      "grad_norm": 0.6484375,
+      "learning_rate": 7.603040661711879e-05,
+      "loss": 1.2082319259643555,
+      "step": 940
+    },
+    {
+      "epoch": 0.20500158852630146,
+      "grad_norm": 0.66796875,
+      "learning_rate": 7.602065664269621e-05,
+      "loss": 1.146630883216858,
+      "step": 941
+    },
+    {
+      "epoch": 0.20521944356192984,
+      "grad_norm": 0.734375,
+      "learning_rate": 7.60108953359175e-05,
+      "loss": 1.4700461626052856,
+      "step": 942
+    },
+    {
+      "epoch": 0.20543729859755822,
+      "grad_norm": 0.6484375,
+      "learning_rate": 7.600112269985366e-05,
+      "loss": 1.1662871837615967,
+      "step": 943
+    },
+    {
+      "epoch": 0.2056551536331866,
+      "grad_norm": 0.68359375,
+      "learning_rate": 7.59913387375792e-05,
+      "loss": 1.1711750030517578,
+      "step": 944
+    },
+    {
+      "epoch": 0.20587300866881497,
+      "grad_norm": 0.65625,
+      "learning_rate": 7.598154345217222e-05,
+      "loss": 1.1843912601470947,
+      "step": 945
+    },
+    {
+      "epoch": 0.20609086370444332,
+      "grad_norm": 0.640625,
+      "learning_rate": 7.59717368467144e-05,
+      "loss": 1.1729689836502075,
+      "step": 946
+    },
+    {
+      "epoch": 0.2063087187400717,
+      "grad_norm": 0.65234375,
+      "learning_rate": 7.596191892429096e-05,
+      "loss": 1.1694141626358032,
+      "step": 947
+    },
+    {
+      "epoch": 0.20652657377570008,
+      "grad_norm": 0.62109375,
+      "learning_rate": 7.595208968799071e-05,
+      "loss": 1.226089596748352,
+      "step": 948
+    },
+    {
+      "epoch": 0.20674442881132846,
+      "grad_norm": 0.65234375,
+      "learning_rate": 7.594224914090597e-05,
+      "loss": 1.088263988494873,
+      "step": 949
+    },
+    {
+      "epoch": 0.20696228384695683,
+      "grad_norm": 0.640625,
+      "learning_rate": 7.593239728613267e-05,
+      "loss": 1.2099326848983765,
+      "step": 950
+    },
+    {
+      "epoch": 0.2071801388825852,
+      "grad_norm": 0.6484375,
+      "learning_rate": 7.592253412677025e-05,
+      "loss": 1.2227391004562378,
+      "step": 951
+    },
+    {
+      "epoch": 0.2073979939182136,
+      "grad_norm": 0.67578125,
+      "learning_rate": 7.591265966592176e-05,
+      "loss": 1.1606372594833374,
+      "step": 952
+    },
+    {
+      "epoch": 0.20761584895384197,
+      "grad_norm": 0.66796875,
+      "learning_rate": 7.590277390669374e-05,
+      "loss": 1.1595875024795532,
+      "step": 953
+    },
+    {
+      "epoch": 0.20783370398947035,
+      "grad_norm": 0.625,
+      "learning_rate": 7.589287685219635e-05,
+      "loss": 1.109031081199646,
+      "step": 954
+    },
+    {
+      "epoch": 0.20805155902509873,
+      "grad_norm": 0.6640625,
+      "learning_rate": 7.588296850554326e-05,
+      "loss": 1.2290798425674438,
+      "step": 955
+    },
+    {
+      "epoch": 0.2082694140607271,
+      "grad_norm": 0.71875,
+      "learning_rate": 7.587304886985171e-05,
+      "loss": 1.3113021850585938,
+      "step": 956
+    },
+    {
+      "epoch": 0.20848726909635545,
+      "grad_norm": 0.640625,
+      "learning_rate": 7.58631179482425e-05,
+      "loss": 1.2492119073867798,
+      "step": 957
+    },
+    {
+      "epoch": 0.20870512413198383,
+      "grad_norm": 0.65625,
+      "learning_rate": 7.585317574383995e-05,
+      "loss": 1.2468491792678833,
+      "step": 958
+    },
+    {
+      "epoch": 0.2089229791676122,
+      "grad_norm": 0.6640625,
+      "learning_rate": 7.584322225977196e-05,
+      "loss": 1.128800630569458,
+      "step": 959
+    },
+    {
+      "epoch": 0.2091408342032406,
+      "grad_norm": 0.67578125,
+      "learning_rate": 7.583325749916997e-05,
+      "loss": 1.2222084999084473,
+      "step": 960
+    },
+    {
+      "epoch": 0.20935868923886897,
+      "grad_norm": 0.6484375,
+      "learning_rate": 7.582328146516894e-05,
+      "loss": 1.1169497966766357,
+      "step": 961
+    },
+    {
+      "epoch": 0.20957654427449735,
+      "grad_norm": 0.6796875,
+      "learning_rate": 7.581329416090743e-05,
+      "loss": 1.2719008922576904,
+      "step": 962
+    },
+    {
+      "epoch": 0.20979439931012572,
+      "grad_norm": 0.6484375,
+      "learning_rate": 7.580329558952751e-05,
+      "loss": 1.1502810716629028,
+      "step": 963
+    },
+    {
+      "epoch": 0.2100122543457541,
+      "grad_norm": 0.6328125,
+      "learning_rate": 7.57932857541748e-05,
+      "loss": 1.1090174913406372,
+      "step": 964
+    },
+    {
+      "epoch": 0.21023010938138248,
+      "grad_norm": 0.63671875,
+      "learning_rate": 7.578326465799845e-05,
+      "loss": 1.117526888847351,
+      "step": 965
+    },
+    {
+      "epoch": 0.21044796441701086,
+      "grad_norm": 0.62890625,
+      "learning_rate": 7.577323230415121e-05,
+      "loss": 1.1834304332733154,
+      "step": 966
+    },
+    {
+      "epoch": 0.21066581945263924,
+      "grad_norm": 0.66015625,
+      "learning_rate": 7.576318869578929e-05,
+      "loss": 1.1292304992675781,
+      "step": 967
+    },
+    {
+      "epoch": 0.2108836744882676,
+      "grad_norm": 0.72265625,
+      "learning_rate": 7.57531338360725e-05,
+      "loss": 1.3212367296218872,
+      "step": 968
+    },
+    {
+      "epoch": 0.21110152952389596,
+      "grad_norm": 0.63671875,
+      "learning_rate": 7.574306772816416e-05,
+      "loss": 1.3200429677963257,
+      "step": 969
+    },
+    {
+      "epoch": 0.21131938455952434,
+      "grad_norm": 0.65625,
+      "learning_rate": 7.573299037523115e-05,
+      "loss": 1.1937294006347656,
+      "step": 970
+    },
+    {
+      "epoch": 0.21153723959515272,
+      "grad_norm": 0.69921875,
+      "learning_rate": 7.572290178044387e-05,
+      "loss": 1.323574185371399,
+      "step": 971
+    },
+    {
+      "epoch": 0.2117550946307811,
+      "grad_norm": 0.65625,
+      "learning_rate": 7.571280194697627e-05,
+      "loss": 1.2618815898895264,
+      "step": 972
+    },
+    {
+      "epoch": 0.21197294966640948,
+      "grad_norm": 0.65234375,
+      "learning_rate": 7.570269087800583e-05,
+      "loss": 1.181321382522583,
+      "step": 973
+    },
+    {
+      "epoch": 0.21219080470203786,
+      "grad_norm": 0.69140625,
+      "learning_rate": 7.569256857671356e-05,
+      "loss": 1.1240822076797485,
+      "step": 974
+    },
+    {
+      "epoch": 0.21240865973766623,
+      "grad_norm": 0.640625,
+      "learning_rate": 7.5682435046284e-05,
+      "loss": 1.0895881652832031,
+      "step": 975
+    },
+    {
+      "epoch": 0.2126265147732946,
+      "grad_norm": 0.6484375,
+      "learning_rate": 7.567229028990525e-05,
+      "loss": 1.1416484117507935,
+      "step": 976
+    },
+    {
+      "epoch": 0.212844369808923,
+      "grad_norm": 0.6640625,
+      "learning_rate": 7.566213431076891e-05,
+      "loss": 1.0759592056274414,
+      "step": 977
+    },
+    {
+      "epoch": 0.21306222484455137,
+      "grad_norm": 0.64453125,
+      "learning_rate": 7.565196711207014e-05,
+      "loss": 1.2216095924377441,
+      "step": 978
+    },
+    {
+      "epoch": 0.21328007988017972,
+      "grad_norm": 0.67578125,
+      "learning_rate": 7.564178869700758e-05,
+      "loss": 1.3798038959503174,
+      "step": 979
+    },
+    {
+      "epoch": 0.2134979349158081,
+      "grad_norm": 0.67578125,
+      "learning_rate": 7.563159906878347e-05,
+      "loss": 1.336762547492981,
+      "step": 980
+    },
+    {
+      "epoch": 0.21371578995143647,
+      "grad_norm": 0.66796875,
+      "learning_rate": 7.562139823060351e-05,
+      "loss": 1.178354263305664,
+      "step": 981
+    },
+    {
+      "epoch": 0.21393364498706485,
+      "grad_norm": 0.6640625,
+      "learning_rate": 7.561118618567699e-05,
+      "loss": 1.3247089385986328,
+      "step": 982
+    },
+    {
+      "epoch": 0.21415150002269323,
+      "grad_norm": 0.6796875,
+      "learning_rate": 7.560096293721665e-05,
+      "loss": 1.1169089078903198,
+      "step": 983
+    },
+    {
+      "epoch": 0.2143693550583216,
+      "grad_norm": 0.6328125,
+      "learning_rate": 7.559072848843883e-05,
+      "loss": 1.0922695398330688,
+      "step": 984
+    },
+    {
+      "epoch": 0.21458721009395,
+      "grad_norm": 0.66015625,
+      "learning_rate": 7.558048284256334e-05,
+      "loss": 1.1527273654937744,
+      "step": 985
+    },
+    {
+      "epoch": 0.21480506512957837,
+      "grad_norm": 0.66015625,
+      "learning_rate": 7.557022600281356e-05,
+      "loss": 1.4035717248916626,
+      "step": 986
+    },
+    {
+      "epoch": 0.21502292016520674,
+      "grad_norm": 0.63671875,
+      "learning_rate": 7.555995797241633e-05,
+      "loss": 1.179969310760498,
+      "step": 987
+    },
+    {
+      "epoch": 0.21524077520083512,
+      "grad_norm": 0.640625,
+      "learning_rate": 7.554967875460208e-05,
+      "loss": 1.1839473247528076,
+      "step": 988
+    },
+    {
+      "epoch": 0.2154586302364635,
+      "grad_norm": 0.640625,
+      "learning_rate": 7.55393883526047e-05,
+      "loss": 1.2991472482681274,
+      "step": 989
+    },
+    {
+      "epoch": 0.21567648527209185,
+      "grad_norm": 0.65625,
+      "learning_rate": 7.552908676966162e-05,
+      "loss": 1.1108072996139526,
+      "step": 990
+    },
+    {
+      "epoch": 0.21589434030772023,
+      "grad_norm": 2.1875,
+      "learning_rate": 7.551877400901383e-05,
+      "loss": 1.2225161790847778,
+      "step": 991
+    },
+    {
+      "epoch": 0.2161121953433486,
+      "grad_norm": 0.65625,
+      "learning_rate": 7.550845007390577e-05,
+      "loss": 1.0939679145812988,
+      "step": 992
+    },
+    {
+      "epoch": 0.21633005037897698,
+      "grad_norm": 0.671875,
+      "learning_rate": 7.549811496758541e-05,
+      "loss": 1.2206825017929077,
+      "step": 993
+    },
+    {
+      "epoch": 0.21654790541460536,
+      "grad_norm": 0.65625,
+      "learning_rate": 7.548776869330427e-05,
+      "loss": 1.1353485584259033,
+      "step": 994
+    },
+    {
+      "epoch": 0.21676576045023374,
+      "grad_norm": 0.6796875,
+      "learning_rate": 7.547741125431737e-05,
+      "loss": 1.3557764291763306,
+      "step": 995
+    },
+    {
+      "epoch": 0.21698361548586212,
+      "grad_norm": 0.67578125,
+      "learning_rate": 7.546704265388322e-05,
+      "loss": 1.274658441543579,
+      "step": 996
+    },
+    {
+      "epoch": 0.2172014705214905,
+      "grad_norm": 0.7109375,
+      "learning_rate": 7.545666289526385e-05,
+      "loss": 1.3172755241394043,
+      "step": 997
+    },
+    {
+      "epoch": 0.21741932555711888,
+      "grad_norm": 0.640625,
+      "learning_rate": 7.544627198172482e-05,
+      "loss": 1.206583857536316,
+      "step": 998
+    },
+    {
+      "epoch": 0.21763718059274725,
+      "grad_norm": 0.66015625,
+      "learning_rate": 7.543586991653517e-05,
+      "loss": 1.1466091871261597,
+      "step": 999
+    },
+    {
+      "epoch": 0.21785503562837563,
+      "grad_norm": 0.67578125,
+      "learning_rate": 7.542545670296747e-05,
+      "loss": 1.1396690607070923,
+      "step": 1000
+    },
+    {
+      "epoch": 0.21807289066400398,
+      "grad_norm": 0.65625,
+      "learning_rate": 7.541503234429782e-05,
+      "loss": 1.0955028533935547,
+      "step": 1001
+    },
+    {
+      "epoch": 0.21829074569963236,
+      "grad_norm": 0.62890625,
+      "learning_rate": 7.540459684380575e-05,
+      "loss": 1.2379428148269653,
+      "step": 1002
+    },
+    {
+      "epoch": 0.21850860073526074,
+      "grad_norm": 0.6796875,
+      "learning_rate": 7.539415020477438e-05,
+      "loss": 1.248972773551941,
+      "step": 1003
+    },
+    {
+      "epoch": 0.21872645577088912,
+      "grad_norm": 0.6875,
+      "learning_rate": 7.538369243049027e-05,
+      "loss": 1.1923973560333252,
+      "step": 1004
+    },
+    {
+      "epoch": 0.2189443108065175,
+      "grad_norm": 0.7265625,
+      "learning_rate": 7.537322352424353e-05,
+      "loss": 1.1802327632904053,
+      "step": 1005
+    },
+    {
+      "epoch": 0.21916216584214587,
+      "grad_norm": 0.65625,
+      "learning_rate": 7.536274348932775e-05,
+      "loss": 1.235487461090088,
+      "step": 1006
+    },
+    {
+      "epoch": 0.21938002087777425,
+      "grad_norm": 0.6953125,
+      "learning_rate": 7.535225232904002e-05,
+      "loss": 1.3099085092544556,
+      "step": 1007
+    },
+    {
+      "epoch": 0.21959787591340263,
+      "grad_norm": 0.74609375,
+      "learning_rate": 7.534175004668093e-05,
+      "loss": 1.2981946468353271,
+      "step": 1008
+    },
+    {
+      "epoch": 0.219815730949031,
+      "grad_norm": 0.67578125,
+      "learning_rate": 7.533123664555458e-05,
+      "loss": 1.3830394744873047,
+      "step": 1009
+    },
+    {
+      "epoch": 0.22003358598465939,
+      "grad_norm": 0.64453125,
+      "learning_rate": 7.532071212896856e-05,
+      "loss": 1.3395079374313354,
+      "step": 1010
+    },
+    {
+      "epoch": 0.22025144102028776,
+      "grad_norm": 0.64453125,
+      "learning_rate": 7.531017650023397e-05,
+      "loss": 1.1506760120391846,
+      "step": 1011
+    },
+    {
+      "epoch": 0.22046929605591611,
+      "grad_norm": 0.68359375,
+      "learning_rate": 7.529962976266535e-05,
+      "loss": 1.3540916442871094,
+      "step": 1012
+    },
+    {
+      "epoch": 0.2206871510915445,
+      "grad_norm": 0.6015625,
+      "learning_rate": 7.528907191958084e-05,
+      "loss": 1.12588369846344,
+      "step": 1013
+    },
+    {
+      "epoch": 0.22090500612717287,
+      "grad_norm": 0.68359375,
+      "learning_rate": 7.527850297430196e-05,
+      "loss": 1.3341773748397827,
+      "step": 1014
+    },
+    {
+      "epoch": 0.22112286116280125,
+      "grad_norm": 0.671875,
+      "learning_rate": 7.526792293015382e-05,
+      "loss": 1.2608731985092163,
+      "step": 1015
+    },
+    {
+      "epoch": 0.22134071619842963,
+      "grad_norm": 0.6484375,
+      "learning_rate": 7.525733179046494e-05,
+      "loss": 1.1986945867538452,
+      "step": 1016
+    },
+    {
+      "epoch": 0.221558571234058,
+      "grad_norm": 0.62890625,
+      "learning_rate": 7.524672955856738e-05,
+      "loss": 1.1029630899429321,
+      "step": 1017
+    },
+    {
+      "epoch": 0.22177642626968638,
+      "grad_norm": 0.62890625,
+      "learning_rate": 7.52361162377967e-05,
+      "loss": 1.1824685335159302,
+      "step": 1018
+    },
+    {
+      "epoch": 0.22199428130531476,
+      "grad_norm": 0.61328125,
+      "learning_rate": 7.52254918314919e-05,
+      "loss": 1.1138750314712524,
+      "step": 1019
+    },
+    {
+      "epoch": 0.22221213634094314,
+      "grad_norm": 0.66796875,
+      "learning_rate": 7.521485634299551e-05,
+      "loss": 1.3057788610458374,
+      "step": 1020
+    },
+    {
+      "epoch": 0.22242999137657152,
+      "grad_norm": 0.68359375,
+      "learning_rate": 7.520420977565352e-05,
+      "loss": 1.2620081901550293,
+      "step": 1021
+    },
+    {
+      "epoch": 0.2226478464121999,
+      "grad_norm": 0.67578125,
+      "learning_rate": 7.519355213281542e-05,
+      "loss": 1.2471615076065063,
+      "step": 1022
+    },
+    {
+      "epoch": 0.22286570144782825,
+      "grad_norm": 0.61328125,
+      "learning_rate": 7.518288341783418e-05,
+      "loss": 1.2310640811920166,
+      "step": 1023
+    },
+    {
+      "epoch": 0.22308355648345662,
+      "grad_norm": 0.7109375,
+      "learning_rate": 7.517220363406627e-05,
+      "loss": 1.2594026327133179,
+      "step": 1024
+    },
+    {
+      "epoch": 0.223301411519085,
+      "grad_norm": 0.69140625,
+      "learning_rate": 7.516151278487158e-05,
+      "loss": 1.2466851472854614,
+      "step": 1025
+    },
+    {
+      "epoch": 0.22351926655471338,
+      "grad_norm": 0.71484375,
+      "learning_rate": 7.515081087361359e-05,
+      "loss": 1.2096786499023438,
+      "step": 1026
+    },
+    {
+      "epoch": 0.22373712159034176,
+      "grad_norm": 0.71875,
+      "learning_rate": 7.514009790365915e-05,
+      "loss": 1.1822397708892822,
+      "step": 1027
+    },
+    {
+      "epoch": 0.22395497662597014,
+      "grad_norm": 0.64453125,
+      "learning_rate": 7.512937387837866e-05,
+      "loss": 1.2193036079406738,
+      "step": 1028
+    },
+    {
+      "epoch": 0.22417283166159852,
+      "grad_norm": 0.74609375,
+      "learning_rate": 7.511863880114595e-05,
+      "loss": 1.1574735641479492,
+      "step": 1029
+    },
+    {
+      "epoch": 0.2243906866972269,
+      "grad_norm": 0.66015625,
+      "learning_rate": 7.51078926753384e-05,
+      "loss": 1.1691049337387085,
+      "step": 1030
+    },
+    {
+      "epoch": 0.22460854173285527,
+      "grad_norm": 0.6484375,
+      "learning_rate": 7.509713550433677e-05,
+      "loss": 1.1394481658935547,
+      "step": 1031
+    },
+    {
+      "epoch": 0.22482639676848365,
+      "grad_norm": 0.64453125,
+      "learning_rate": 7.508636729152538e-05,
+      "loss": 1.0755339860916138,
+      "step": 1032
+    },
+    {
+      "epoch": 0.22504425180411203,
+      "grad_norm": 0.70703125,
+      "learning_rate": 7.507558804029194e-05,
+      "loss": 1.2174638509750366,
+      "step": 1033
+    },
+    {
+      "epoch": 0.22526210683974038,
+      "grad_norm": 0.7109375,
+      "learning_rate": 7.506479775402774e-05,
+      "loss": 1.1659479141235352,
+      "step": 1034
+    },
+    {
+      "epoch": 0.22547996187536876,
+      "grad_norm": 0.71484375,
+      "learning_rate": 7.505399643612743e-05,
+      "loss": 1.360796332359314,
+      "step": 1035
+    },
+    {
+      "epoch": 0.22569781691099713,
+      "grad_norm": 0.67578125,
+      "learning_rate": 7.50431840899892e-05,
+      "loss": 1.377150058746338,
+      "step": 1036
+    },
+    {
+      "epoch": 0.2259156719466255,
+      "grad_norm": 0.65234375,
+      "learning_rate": 7.50323607190147e-05,
+      "loss": 1.1639362573623657,
+      "step": 1037
+    },
+    {
+      "epoch": 0.2261335269822539,
+      "grad_norm": 0.6875,
+      "learning_rate": 7.502152632660903e-05,
+      "loss": 1.226845383644104,
+      "step": 1038
+    },
+    {
+      "epoch": 0.22635138201788227,
+      "grad_norm": 0.65625,
+      "learning_rate": 7.501068091618077e-05,
+      "loss": 1.3512401580810547,
+      "step": 1039
+    },
+    {
+      "epoch": 0.22656923705351065,
+      "grad_norm": 0.65625,
+      "learning_rate": 7.499982449114195e-05,
+      "loss": 1.1122756004333496,
+      "step": 1040
+    },
+    {
+      "epoch": 0.22678709208913903,
+      "grad_norm": 0.66015625,
+      "learning_rate": 7.498895705490811e-05,
+      "loss": 1.198566198348999,
+      "step": 1041
+    },
+    {
+      "epoch": 0.2270049471247674,
+      "grad_norm": 0.64453125,
+      "learning_rate": 7.49780786108982e-05,
+      "loss": 1.2317081689834595,
+      "step": 1042
+    },
+    {
+      "epoch": 0.22722280216039578,
+      "grad_norm": 0.65625,
+      "learning_rate": 7.496718916253466e-05,
+      "loss": 1.2326804399490356,
+      "step": 1043
+    },
+    {
+      "epoch": 0.22744065719602413,
+      "grad_norm": 0.6015625,
+      "learning_rate": 7.49562887132434e-05,
+      "loss": 1.0571157932281494,
+      "step": 1044
+    },
+    {
+      "epoch": 0.2276585122316525,
+      "grad_norm": 0.66796875,
+      "learning_rate": 7.494537726645375e-05,
+      "loss": 1.262416958808899,
+      "step": 1045
+    },
+    {
+      "epoch": 0.2278763672672809,
+      "grad_norm": 0.85546875,
+      "learning_rate": 7.493445482559857e-05,
+      "loss": 1.2508392333984375,
+      "step": 1046
+    },
+    {
+      "epoch": 0.22809422230290927,
+      "grad_norm": 0.6328125,
+      "learning_rate": 7.49235213941141e-05,
+      "loss": 1.0722736120224,
+      "step": 1047
+    },
+    {
+      "epoch": 0.22831207733853764,
+      "grad_norm": 0.69921875,
+      "learning_rate": 7.49125769754401e-05,
+      "loss": 1.3226914405822754,
+      "step": 1048
+    },
+    {
+      "epoch": 0.22852993237416602,
+      "grad_norm": 0.6875,
+      "learning_rate": 7.490162157301975e-05,
+      "loss": 1.4083926677703857,
+      "step": 1049
+    },
+    {
+      "epoch": 0.2287477874097944,
+      "grad_norm": 0.6484375,
+      "learning_rate": 7.48906551902997e-05,
+      "loss": 1.1254980564117432,
+      "step": 1050
+    },
+    {
+      "epoch": 0.22896564244542278,
+      "grad_norm": 0.6796875,
+      "learning_rate": 7.487967783073006e-05,
+      "loss": 1.2946972846984863,
+      "step": 1051
+    },
+    {
+      "epoch": 0.22918349748105116,
+      "grad_norm": 0.66796875,
+      "learning_rate": 7.486868949776439e-05,
+      "loss": 1.2502243518829346,
+      "step": 1052
+    },
+    {
+      "epoch": 0.22940135251667954,
+      "grad_norm": 0.7109375,
+      "learning_rate": 7.485769019485967e-05,
+      "loss": 1.2442363500595093,
+      "step": 1053
+    },
+    {
+      "epoch": 0.2296192075523079,
+      "grad_norm": 0.65234375,
+      "learning_rate": 7.484667992547639e-05,
+      "loss": 1.1876962184906006,
+      "step": 1054
+    },
+    {
+      "epoch": 0.22983706258793626,
+      "grad_norm": 0.6640625,
+      "learning_rate": 7.483565869307845e-05,
+      "loss": 1.2130298614501953,
+      "step": 1055
+    },
+    {
+      "epoch": 0.23005491762356464,
+      "grad_norm": 0.640625,
+      "learning_rate": 7.48246265011332e-05,
+      "loss": 1.1834275722503662,
+      "step": 1056
+    },
+    {
+      "epoch": 0.23027277265919302,
+      "grad_norm": 0.66015625,
+      "learning_rate": 7.481358335311145e-05,
+      "loss": 1.1670516729354858,
+      "step": 1057
+    },
+    {
+      "epoch": 0.2304906276948214,
+      "grad_norm": 0.66015625,
+      "learning_rate": 7.480252925248746e-05,
+      "loss": 1.0916193723678589,
+      "step": 1058
+    },
+    {
+      "epoch": 0.23070848273044978,
+      "grad_norm": 0.67578125,
+      "learning_rate": 7.479146420273893e-05,
+      "loss": 1.229089379310608,
+      "step": 1059
+    },
+    {
+      "epoch": 0.23092633776607815,
+      "grad_norm": 0.703125,
+      "learning_rate": 7.478038820734698e-05,
+      "loss": 1.3365142345428467,
+      "step": 1060
+    },
+    {
+      "epoch": 0.23114419280170653,
+      "grad_norm": 0.71484375,
+      "learning_rate": 7.476930126979623e-05,
+      "loss": 1.1554499864578247,
+      "step": 1061
+    },
+    {
+      "epoch": 0.2313620478373349,
+      "grad_norm": 0.66015625,
+      "learning_rate": 7.47582033935747e-05,
+      "loss": 1.1742740869522095,
+      "step": 1062
+    },
+    {
+      "epoch": 0.2315799028729633,
+      "grad_norm": 0.66015625,
+      "learning_rate": 7.474709458217385e-05,
+      "loss": 1.184830665588379,
+      "step": 1063
+    },
+    {
+      "epoch": 0.23179775790859167,
+      "grad_norm": 0.671875,
+      "learning_rate": 7.473597483908861e-05,
+      "loss": 1.3750882148742676,
+      "step": 1064
+    },
+    {
+      "epoch": 0.23201561294422005,
+      "grad_norm": 0.59765625,
+      "learning_rate": 7.472484416781731e-05,
+      "loss": 0.9657297134399414,
+      "step": 1065
+    },
+    {
+      "epoch": 0.2322334679798484,
+      "grad_norm": 0.69921875,
+      "learning_rate": 7.471370257186175e-05,
+      "loss": 1.2051687240600586,
+      "step": 1066
+    },
+    {
+      "epoch": 0.23245132301547677,
+      "grad_norm": 0.67578125,
+      "learning_rate": 7.470255005472715e-05,
+      "loss": 1.2893788814544678,
+      "step": 1067
+    },
+    {
+      "epoch": 0.23266917805110515,
+      "grad_norm": 0.66015625,
+      "learning_rate": 7.469138661992219e-05,
+      "loss": 1.0680146217346191,
+      "step": 1068
+    },
+    {
+      "epoch": 0.23288703308673353,
+      "grad_norm": 0.671875,
+      "learning_rate": 7.468021227095893e-05,
+      "loss": 1.216873049736023,
+      "step": 1069
+    },
+    {
+      "epoch": 0.2331048881223619,
+      "grad_norm": 0.67578125,
+      "learning_rate": 7.466902701135294e-05,
+      "loss": 1.2305889129638672,
+      "step": 1070
+    },
+    {
+      "epoch": 0.2333227431579903,
+      "grad_norm": 0.671875,
+      "learning_rate": 7.465783084462315e-05,
+      "loss": 1.2332993745803833,
+      "step": 1071
+    },
+    {
+      "epoch": 0.23354059819361866,
+      "grad_norm": 0.6796875,
+      "learning_rate": 7.464662377429197e-05,
+      "loss": 1.2330303192138672,
+      "step": 1072
+    },
+    {
+      "epoch": 0.23375845322924704,
+      "grad_norm": 0.703125,
+      "learning_rate": 7.463540580388524e-05,
+      "loss": 1.2257293462753296,
+      "step": 1073
+    },
+    {
+      "epoch": 0.23397630826487542,
+      "grad_norm": 0.6953125,
+      "learning_rate": 7.462417693693217e-05,
+      "loss": 1.248563528060913,
+      "step": 1074
+    },
+    {
+      "epoch": 0.2341941633005038,
+      "grad_norm": 0.7109375,
+      "learning_rate": 7.461293717696549e-05,
+      "loss": 1.248349905014038,
+      "step": 1075
+    },
+    {
+      "epoch": 0.23441201833613218,
+      "grad_norm": 0.69140625,
+      "learning_rate": 7.460168652752129e-05,
+      "loss": 1.173666000366211,
+      "step": 1076
+    },
+    {
+      "epoch": 0.23462987337176053,
+      "grad_norm": 0.640625,
+      "learning_rate": 7.45904249921391e-05,
+      "loss": 1.1662412881851196,
+      "step": 1077
+    },
+    {
+      "epoch": 0.2348477284073889,
+      "grad_norm": 0.6875,
+      "learning_rate": 7.457915257436187e-05,
+      "loss": 1.1980122327804565,
+      "step": 1078
+    },
+    {
+      "epoch": 0.23506558344301728,
+      "grad_norm": 0.68359375,
+      "learning_rate": 7.4567869277736e-05,
+      "loss": 1.2592887878417969,
+      "step": 1079
+    },
+    {
+      "epoch": 0.23528343847864566,
+      "grad_norm": 0.6484375,
+      "learning_rate": 7.455657510581129e-05,
+      "loss": 1.2540268898010254,
+      "step": 1080
+    },
+    {
+      "epoch": 0.23550129351427404,
+      "grad_norm": 0.62890625,
+      "learning_rate": 7.454527006214097e-05,
+      "loss": 1.1521363258361816,
+      "step": 1081
+    },
+    {
+      "epoch": 0.23571914854990242,
+      "grad_norm": 0.6328125,
+      "learning_rate": 7.453395415028169e-05,
+      "loss": 1.1716868877410889,
+      "step": 1082
+    },
+    {
+      "epoch": 0.2359370035855308,
+      "grad_norm": 0.62109375,
+      "learning_rate": 7.45226273737935e-05,
+      "loss": 1.093170404434204,
+      "step": 1083
+    },
+    {
+      "epoch": 0.23615485862115917,
+      "grad_norm": 0.66796875,
+      "learning_rate": 7.451128973623992e-05,
+      "loss": 1.1784939765930176,
+      "step": 1084
+    },
+    {
+      "epoch": 0.23637271365678755,
+      "grad_norm": 0.6171875,
+      "learning_rate": 7.449994124118781e-05,
+      "loss": 1.0779354572296143,
+      "step": 1085
+    },
+    {
+      "epoch": 0.23659056869241593,
+      "grad_norm": 0.6328125,
+      "learning_rate": 7.448858189220751e-05,
+      "loss": 1.021164894104004,
+      "step": 1086
+    },
+    {
+      "epoch": 0.2368084237280443,
+      "grad_norm": 0.62890625,
+      "learning_rate": 7.447721169287276e-05,
+      "loss": 1.153201937675476,
+      "step": 1087
+    },
+    {
+      "epoch": 0.23702627876367266,
+      "grad_norm": 0.6328125,
+      "learning_rate": 7.446583064676068e-05,
+      "loss": 1.100892186164856,
+      "step": 1088
+    },
+    {
+      "epoch": 0.23724413379930104,
+      "grad_norm": 0.6640625,
+      "learning_rate": 7.445443875745185e-05,
+      "loss": 1.2319788932800293,
+      "step": 1089
+    },
+    {
+      "epoch": 0.23746198883492942,
+      "grad_norm": 0.6953125,
+      "learning_rate": 7.444303602853024e-05,
+      "loss": 1.2787885665893555,
+      "step": 1090
+    },
+    {
+      "epoch": 0.2376798438705578,
+      "grad_norm": 0.69140625,
+      "learning_rate": 7.443162246358322e-05,
+      "loss": 1.3306525945663452,
+      "step": 1091
+    },
+    {
+      "epoch": 0.23789769890618617,
+      "grad_norm": 0.66015625,
+      "learning_rate": 7.442019806620158e-05,
+      "loss": 1.3167544603347778,
+      "step": 1092
+    },
+    {
+      "epoch": 0.23811555394181455,
+      "grad_norm": 0.625,
+      "learning_rate": 7.440876283997952e-05,
+      "loss": 1.1668832302093506,
+      "step": 1093
+    },
+    {
+      "epoch": 0.23833340897744293,
+      "grad_norm": 0.625,
+      "learning_rate": 7.439731678851465e-05,
+      "loss": 1.1132535934448242,
+      "step": 1094
+    },
+    {
+      "epoch": 0.2385512640130713,
+      "grad_norm": 0.6640625,
+      "learning_rate": 7.438585991540796e-05,
+      "loss": 1.269753336906433,
+      "step": 1095
+    },
+    {
+      "epoch": 0.23876911904869969,
+      "grad_norm": 0.671875,
+      "learning_rate": 7.437439222426388e-05,
+      "loss": 1.1666823625564575,
+      "step": 1096
+    },
+    {
+      "epoch": 0.23898697408432806,
+      "grad_norm": 0.7265625,
+      "learning_rate": 7.436291371869025e-05,
+      "loss": 1.2803019285202026,
+      "step": 1097
+    },
+    {
+      "epoch": 0.23920482911995644,
+      "grad_norm": 0.640625,
+      "learning_rate": 7.435142440229824e-05,
+      "loss": 1.2915737628936768,
+      "step": 1098
+    },
+    {
+      "epoch": 0.2394226841555848,
+      "grad_norm": 0.69140625,
+      "learning_rate": 7.43399242787025e-05,
+      "loss": 1.2518784999847412,
+      "step": 1099
+    },
+    {
+      "epoch": 0.23964053919121317,
+      "grad_norm": 0.6328125,
+      "learning_rate": 7.432841335152106e-05,
+      "loss": 1.1555505990982056,
+      "step": 1100
+    },
+    {
+      "epoch": 0.23985839422684155,
+      "grad_norm": 0.6328125,
+      "learning_rate": 7.431689162437531e-05,
+      "loss": 1.2523589134216309,
+      "step": 1101
+    },
+    {
+      "epoch": 0.24007624926246993,
+      "grad_norm": 0.828125,
+      "learning_rate": 7.43053591008901e-05,
+      "loss": 1.3032128810882568,
+      "step": 1102
+    },
+    {
+      "epoch": 0.2402941042980983,
+      "grad_norm": 0.640625,
+      "learning_rate": 7.429381578469362e-05,
+      "loss": 1.2056561708450317,
+      "step": 1103
+    },
+    {
+      "epoch": 0.24051195933372668,
+      "grad_norm": 0.6484375,
+      "learning_rate": 7.428226167941749e-05,
+      "loss": 1.089126706123352,
+      "step": 1104
+    },
+    {
+      "epoch": 0.24072981436935506,
+      "grad_norm": 0.5859375,
+      "learning_rate": 7.427069678869672e-05,
+      "loss": 1.127699613571167,
+      "step": 1105
+    },
+    {
+      "epoch": 0.24094766940498344,
+      "grad_norm": 0.69921875,
+      "learning_rate": 7.425912111616969e-05,
+      "loss": 1.303905963897705,
+      "step": 1106
+    },
+    {
+      "epoch": 0.24116552444061182,
+      "grad_norm": 0.69140625,
+      "learning_rate": 7.424753466547821e-05,
+      "loss": 1.2440252304077148,
+      "step": 1107
+    },
+    {
+      "epoch": 0.2413833794762402,
+      "grad_norm": 0.66796875,
+      "learning_rate": 7.423593744026746e-05,
+      "loss": 1.1817470788955688,
+      "step": 1108
+    },
+    {
+      "epoch": 0.24160123451186857,
+      "grad_norm": 0.671875,
+      "learning_rate": 7.422432944418597e-05,
+      "loss": 1.3318527936935425,
+      "step": 1109
+    },
+    {
+      "epoch": 0.24181908954749692,
+      "grad_norm": 0.61328125,
+      "learning_rate": 7.421271068088575e-05,
+      "loss": 1.0726319551467896,
+      "step": 1110
+    },
+    {
+      "epoch": 0.2420369445831253,
+      "grad_norm": 0.62109375,
+      "learning_rate": 7.420108115402213e-05,
+      "loss": 1.1061347723007202,
+      "step": 1111
+    },
+    {
+      "epoch": 0.24225479961875368,
+      "grad_norm": 0.6328125,
+      "learning_rate": 7.418944086725383e-05,
+      "loss": 1.1564338207244873,
+      "step": 1112
+    },
+    {
+      "epoch": 0.24247265465438206,
+      "grad_norm": 0.6640625,
+      "learning_rate": 7.417778982424297e-05,
+      "loss": 1.2411408424377441,
+      "step": 1113
+    },
+    {
+      "epoch": 0.24269050969001044,
+      "grad_norm": 0.65234375,
+      "learning_rate": 7.416612802865505e-05,
+      "loss": 1.194381594657898,
+      "step": 1114
+    },
+    {
+      "epoch": 0.24290836472563881,
+      "grad_norm": 0.6328125,
+      "learning_rate": 7.415445548415898e-05,
+      "loss": 1.2441036701202393,
+      "step": 1115
+    },
+    {
+      "epoch": 0.2431262197612672,
+      "grad_norm": 0.625,
+      "learning_rate": 7.4142772194427e-05,
+      "loss": 1.1813486814498901,
+      "step": 1116
+    },
+    {
+      "epoch": 0.24334407479689557,
+      "grad_norm": 0.67578125,
+      "learning_rate": 7.413107816313475e-05,
+      "loss": 1.26551353931427,
+      "step": 1117
+    },
+    {
+      "epoch": 0.24356192983252395,
+      "grad_norm": 0.69921875,
+      "learning_rate": 7.411937339396127e-05,
+      "loss": 0.8941237330436707,
+      "step": 1118
+    },
+    {
+      "epoch": 0.24377978486815233,
+      "grad_norm": 0.6796875,
+      "learning_rate": 7.410765789058899e-05,
+      "loss": 1.061943769454956,
+      "step": 1119
+    },
+    {
+      "epoch": 0.2439976399037807,
+      "grad_norm": 0.625,
+      "learning_rate": 7.409593165670364e-05,
+      "loss": 1.1157712936401367,
+      "step": 1120
+    },
+    {
+      "epoch": 0.24421549493940906,
+      "grad_norm": 0.6328125,
+      "learning_rate": 7.40841946959944e-05,
+      "loss": 1.1936880350112915,
+      "step": 1121
+    },
+    {
+      "epoch": 0.24443334997503743,
+      "grad_norm": 0.6640625,
+      "learning_rate": 7.40724470121538e-05,
+      "loss": 1.156053066253662,
+      "step": 1122
+    },
+    {
+      "epoch": 0.2446512050106658,
+      "grad_norm": 0.65625,
+      "learning_rate": 7.406068860887777e-05,
+      "loss": 1.238749384880066,
+      "step": 1123
+    },
+    {
+      "epoch": 0.2448690600462942,
+      "grad_norm": 0.671875,
+      "learning_rate": 7.404891948986555e-05,
+      "loss": 1.2391966581344604,
+      "step": 1124
+    },
+    {
+      "epoch": 0.24508691508192257,
+      "grad_norm": 0.6875,
+      "learning_rate": 7.403713965881981e-05,
+      "loss": 1.276589035987854,
+      "step": 1125
+    },
+    {
+      "epoch": 0.24530477011755095,
+      "grad_norm": 0.6328125,
+      "learning_rate": 7.402534911944655e-05,
+      "loss": 1.1045472621917725,
+      "step": 1126
+    },
+    {
+      "epoch": 0.24552262515317932,
+      "grad_norm": 0.69921875,
+      "learning_rate": 7.401354787545519e-05,
+      "loss": 1.2258763313293457,
+      "step": 1127
+    },
+    {
+      "epoch": 0.2457404801888077,
+      "grad_norm": 0.65234375,
+      "learning_rate": 7.400173593055846e-05,
+      "loss": 1.1342498064041138,
+      "step": 1128
+    },
+    {
+      "epoch": 0.24595833522443608,
+      "grad_norm": 0.62890625,
+      "learning_rate": 7.398991328847249e-05,
+      "loss": 1.0424692630767822,
+      "step": 1129
+    },
+    {
+      "epoch": 0.24617619026006446,
+      "grad_norm": 0.65625,
+      "learning_rate": 7.397807995291677e-05,
+      "loss": 1.1807204484939575,
+      "step": 1130
+    },
+    {
+      "epoch": 0.24639404529569284,
+      "grad_norm": 0.69140625,
+      "learning_rate": 7.396623592761415e-05,
+      "loss": 1.3167810440063477,
+      "step": 1131
+    },
+    {
+      "epoch": 0.2466119003313212,
+      "grad_norm": 0.6015625,
+      "learning_rate": 7.395438121629082e-05,
+      "loss": 1.0902800559997559,
+      "step": 1132
+    },
+    {
+      "epoch": 0.24682975536694957,
+      "grad_norm": 0.64453125,
+      "learning_rate": 7.394251582267642e-05,
+      "loss": 1.1612921953201294,
+      "step": 1133
+    },
+    {
+      "epoch": 0.24704761040257794,
+      "grad_norm": 0.66796875,
+      "learning_rate": 7.393063975050383e-05,
+      "loss": 1.2575225830078125,
+      "step": 1134
+    },
+    {
+      "epoch": 0.24726546543820632,
+      "grad_norm": 0.63671875,
+      "learning_rate": 7.391875300350937e-05,
+      "loss": 1.2239922285079956,
+      "step": 1135
+    },
+    {
+      "epoch": 0.2474833204738347,
+      "grad_norm": 0.62109375,
+      "learning_rate": 7.390685558543268e-05,
+      "loss": 1.197227120399475,
+      "step": 1136
+    },
+    {
+      "epoch": 0.24770117550946308,
+      "grad_norm": 0.671875,
+      "learning_rate": 7.38949475000168e-05,
+      "loss": 1.1977479457855225,
+      "step": 1137
+    },
+    {
+      "epoch": 0.24791903054509146,
+      "grad_norm": 0.671875,
+      "learning_rate": 7.388302875100808e-05,
+      "loss": 1.278836727142334,
+      "step": 1138
+    },
+    {
+      "epoch": 0.24813688558071983,
+      "grad_norm": 0.62890625,
+      "learning_rate": 7.387109934215626e-05,
+      "loss": 1.1334035396575928,
+      "step": 1139
+    },
+    {
+      "epoch": 0.2483547406163482,
+      "grad_norm": 0.6796875,
+      "learning_rate": 7.385915927721438e-05,
+      "loss": 1.1485494375228882,
+      "step": 1140
+    },
+    {
+      "epoch": 0.2485725956519766,
+      "grad_norm": 0.6796875,
+      "learning_rate": 7.384720855993892e-05,
+      "loss": 1.250815510749817,
+      "step": 1141
+    },
+    {
+      "epoch": 0.24879045068760497,
+      "grad_norm": 0.68359375,
+      "learning_rate": 7.38352471940896e-05,
+      "loss": 1.3503577709197998,
+      "step": 1142
+    },
+    {
+      "epoch": 0.24900830572323332,
+      "grad_norm": 0.66796875,
+      "learning_rate": 7.382327518342962e-05,
+      "loss": 1.3607208728790283,
+      "step": 1143
+    },
+    {
+      "epoch": 0.2492261607588617,
+      "grad_norm": 0.6484375,
+      "learning_rate": 7.381129253172541e-05,
+      "loss": 1.0941743850708008,
+      "step": 1144
+    },
+    {
+      "epoch": 0.24944401579449008,
+      "grad_norm": 0.66796875,
+      "learning_rate": 7.379929924274682e-05,
+      "loss": 1.2399653196334839,
+      "step": 1145
+    },
+    {
+      "epoch": 0.24966187083011845,
+      "grad_norm": 0.61328125,
+      "learning_rate": 7.378729532026702e-05,
+      "loss": 1.0081626176834106,
+      "step": 1146
+    },
+    {
+      "epoch": 0.24987972586574683,
+      "grad_norm": 0.640625,
+      "learning_rate": 7.377528076806253e-05,
+      "loss": 1.2834120988845825,
+      "step": 1147
+    },
+    {
+      "epoch": 0.2500975809013752,
+      "grad_norm": 0.640625,
+      "learning_rate": 7.376325558991321e-05,
+      "loss": 1.1639472246170044,
+      "step": 1148
+    },
+    {
+      "epoch": 0.25031543593700356,
+      "grad_norm": 0.6796875,
+      "learning_rate": 7.375121978960227e-05,
+      "loss": 1.1719928979873657,
+      "step": 1149
+    },
+    {
+      "epoch": 0.25053329097263194,
+      "grad_norm": 0.64453125,
+      "learning_rate": 7.373917337091626e-05,
+      "loss": 1.1508649587631226,
+      "step": 1150
+    },
+    {
+      "epoch": 0.2507511460082603,
+      "grad_norm": 0.60546875,
+      "learning_rate": 7.372711633764508e-05,
+      "loss": 1.0626050233840942,
+      "step": 1151
+    },
+    {
+      "epoch": 0.2509690010438887,
+      "grad_norm": 0.65625,
+      "learning_rate": 7.371504869358193e-05,
+      "loss": 1.1858652830123901,
+      "step": 1152
+    },
+    {
+      "epoch": 0.2511868560795171,
+      "grad_norm": 0.61328125,
+      "learning_rate": 7.37029704425234e-05,
+      "loss": 1.045055627822876,
+      "step": 1153
+    },
+    {
+      "epoch": 0.25140471111514545,
+      "grad_norm": 0.71875,
+      "learning_rate": 7.36908815882694e-05,
+      "loss": 1.2046691179275513,
+      "step": 1154
+    },
+    {
+      "epoch": 0.25162256615077383,
+      "grad_norm": 0.65234375,
+      "learning_rate": 7.367878213462315e-05,
+      "loss": 1.1853634119033813,
+      "step": 1155
+    },
+    {
+      "epoch": 0.2518404211864022,
+      "grad_norm": 0.61328125,
+      "learning_rate": 7.366667208539123e-05,
+      "loss": 1.161887764930725,
+      "step": 1156
+    },
+    {
+      "epoch": 0.2520582762220306,
+      "grad_norm": 0.5859375,
+      "learning_rate": 7.365455144438355e-05,
+      "loss": 1.0042407512664795,
+      "step": 1157
+    },
+    {
+      "epoch": 0.25227613125765896,
+      "grad_norm": 0.95703125,
+      "learning_rate": 7.364242021541335e-05,
+      "loss": 1.1904877424240112,
+      "step": 1158
+    },
+    {
+      "epoch": 0.25249398629328734,
+      "grad_norm": 0.6796875,
+      "learning_rate": 7.363027840229721e-05,
+      "loss": 1.0873665809631348,
+      "step": 1159
+    },
+    {
+      "epoch": 0.2527118413289157,
+      "grad_norm": 0.625,
+      "learning_rate": 7.361812600885503e-05,
+      "loss": 1.209153652191162,
+      "step": 1160
+    },
+    {
+      "epoch": 0.2529296963645441,
+      "grad_norm": 0.65234375,
+      "learning_rate": 7.360596303891002e-05,
+      "loss": 1.196313738822937,
+      "step": 1161
+    },
+    {
+      "epoch": 0.2531475514001725,
+      "grad_norm": 0.64453125,
+      "learning_rate": 7.359378949628875e-05,
+      "loss": 1.1550343036651611,
+      "step": 1162
+    },
+    {
+      "epoch": 0.25336540643580086,
+      "grad_norm": 0.6484375,
+      "learning_rate": 7.358160538482113e-05,
+      "loss": 1.1793404817581177,
+      "step": 1163
+    },
+    {
+      "epoch": 0.25358326147142923,
+      "grad_norm": 0.65625,
+      "learning_rate": 7.356941070834032e-05,
+      "loss": 1.1648451089859009,
+      "step": 1164
+    },
+    {
+      "epoch": 0.2538011165070576,
+      "grad_norm": 0.6796875,
+      "learning_rate": 7.355720547068288e-05,
+      "loss": 1.0461862087249756,
+      "step": 1165
+    },
+    {
+      "epoch": 0.254018971542686,
+      "grad_norm": 0.67578125,
+      "learning_rate": 7.354498967568866e-05,
+      "loss": 1.0747504234313965,
+      "step": 1166
+    },
+    {
+      "epoch": 0.25423682657831437,
+      "grad_norm": 0.71875,
+      "learning_rate": 7.353276332720085e-05,
+      "loss": 1.264075517654419,
+      "step": 1167
+    },
+    {
+      "epoch": 0.25445468161394275,
+      "grad_norm": 0.64453125,
+      "learning_rate": 7.352052642906591e-05,
+      "loss": 1.0863678455352783,
+      "step": 1168
+    },
+    {
+      "epoch": 0.2546725366495711,
+      "grad_norm": 0.625,
+      "learning_rate": 7.35082789851337e-05,
+      "loss": 1.048417568206787,
+      "step": 1169
+    },
+    {
+      "epoch": 0.25489039168519945,
+      "grad_norm": 0.6484375,
+      "learning_rate": 7.349602099925733e-05,
+      "loss": 1.1720982789993286,
+      "step": 1170
+    },
+    {
+      "epoch": 0.2551082467208278,
+      "grad_norm": 0.65234375,
+      "learning_rate": 7.348375247529324e-05,
+      "loss": 1.1866259574890137,
+      "step": 1171
+    },
+    {
+      "epoch": 0.2553261017564562,
+      "grad_norm": 0.6796875,
+      "learning_rate": 7.347147341710123e-05,
+      "loss": 1.1172094345092773,
+      "step": 1172
+    },
+    {
+      "epoch": 0.2555439567920846,
+      "grad_norm": 0.65625,
+      "learning_rate": 7.345918382854435e-05,
+      "loss": 1.336367130279541,
+      "step": 1173
+    },
+    {
+      "epoch": 0.25576181182771296,
+      "grad_norm": 0.6640625,
+      "learning_rate": 7.3446883713489e-05,
+      "loss": 1.1751132011413574,
+      "step": 1174
+    },
+    {
+      "epoch": 0.25597966686334134,
+      "grad_norm": 0.62109375,
+      "learning_rate": 7.343457307580488e-05,
+      "loss": 1.2584152221679688,
+      "step": 1175
+    },
+    {
+      "epoch": 0.2561975218989697,
+      "grad_norm": 0.609375,
+      "learning_rate": 7.342225191936501e-05,
+      "loss": 0.9509475231170654,
+      "step": 1176
+    },
+    {
+      "epoch": 0.2564153769345981,
+      "grad_norm": 0.72265625,
+      "learning_rate": 7.340992024804571e-05,
+      "loss": 1.286073088645935,
+      "step": 1177
+    },
+    {
+      "epoch": 0.25663323197022647,
+      "grad_norm": 0.6484375,
+      "learning_rate": 7.339757806572662e-05,
+      "loss": 1.2196329832077026,
+      "step": 1178
+    },
+    {
+      "epoch": 0.25685108700585485,
+      "grad_norm": 0.65234375,
+      "learning_rate": 7.338522537629069e-05,
+      "loss": 1.2197144031524658,
+      "step": 1179
+    },
+    {
+      "epoch": 0.25706894204148323,
+      "grad_norm": 0.66015625,
+      "learning_rate": 7.337286218362413e-05,
+      "loss": 1.2242324352264404,
+      "step": 1180
+    },
+    {
+      "epoch": 0.2572867970771116,
+      "grad_norm": 0.640625,
+      "learning_rate": 7.33604884916165e-05,
+      "loss": 1.0997776985168457,
+      "step": 1181
+    },
+    {
+      "epoch": 0.25750465211274,
+      "grad_norm": 0.6640625,
+      "learning_rate": 7.334810430416068e-05,
+      "loss": 1.2055772542953491,
+      "step": 1182
+    },
+    {
+      "epoch": 0.25772250714836836,
+      "grad_norm": 0.69140625,
+      "learning_rate": 7.333570962515279e-05,
+      "loss": 1.2880114316940308,
+      "step": 1183
+    },
+    {
+      "epoch": 0.25794036218399674,
+      "grad_norm": 0.66015625,
+      "learning_rate": 7.332330445849229e-05,
+      "loss": 1.20735764503479,
+      "step": 1184
+    },
+    {
+      "epoch": 0.2581582172196251,
+      "grad_norm": 0.6796875,
+      "learning_rate": 7.331088880808196e-05,
+      "loss": 1.196286678314209,
+      "step": 1185
+    },
+    {
+      "epoch": 0.2583760722552535,
+      "grad_norm": 0.65234375,
+      "learning_rate": 7.329846267782782e-05,
+      "loss": 1.2637931108474731,
+      "step": 1186
+    },
+    {
+      "epoch": 0.2585939272908819,
+      "grad_norm": 0.6796875,
+      "learning_rate": 7.328602607163924e-05,
+      "loss": 1.1651897430419922,
+      "step": 1187
+    },
+    {
+      "epoch": 0.25881178232651025,
+      "grad_norm": 0.6875,
+      "learning_rate": 7.327357899342886e-05,
+      "loss": 1.2561732530593872,
+      "step": 1188
+    },
+    {
+      "epoch": 0.25902963736213863,
+      "grad_norm": 0.6640625,
+      "learning_rate": 7.326112144711261e-05,
+      "loss": 1.3208260536193848,
+      "step": 1189
+    },
+    {
+      "epoch": 0.259247492397767,
+      "grad_norm": 0.65625,
+      "learning_rate": 7.324865343660974e-05,
+      "loss": 1.132535696029663,
+      "step": 1190
+    },
+    {
+      "epoch": 0.2594653474333954,
+      "grad_norm": 0.6015625,
+      "learning_rate": 7.323617496584277e-05,
+      "loss": 1.0516167879104614,
+      "step": 1191
+    },
+    {
+      "epoch": 0.2596832024690237,
+      "grad_norm": 0.6328125,
+      "learning_rate": 7.322368603873751e-05,
+      "loss": 1.0627448558807373,
+      "step": 1192
+    },
+    {
+      "epoch": 0.2599010575046521,
+      "grad_norm": 0.65234375,
+      "learning_rate": 7.321118665922307e-05,
+      "loss": 1.0868852138519287,
+      "step": 1193
+    },
+    {
+      "epoch": 0.26011891254028047,
+      "grad_norm": 0.65234375,
+      "learning_rate": 7.319867683123183e-05,
+      "loss": 1.1963251829147339,
+      "step": 1194
+    },
+    {
+      "epoch": 0.26033676757590885,
+      "grad_norm": 0.73046875,
+      "learning_rate": 7.31861565586995e-05,
+      "loss": 1.1008213758468628,
+      "step": 1195
+    },
+    {
+      "epoch": 0.2605546226115372,
+      "grad_norm": 0.68359375,
+      "learning_rate": 7.317362584556503e-05,
+      "loss": 1.1906739473342896,
+      "step": 1196
+    },
+    {
+      "epoch": 0.2607724776471656,
+      "grad_norm": 0.66015625,
+      "learning_rate": 7.316108469577067e-05,
+      "loss": 1.1522700786590576,
+      "step": 1197
+    },
+    {
+      "epoch": 0.260990332682794,
+      "grad_norm": 0.68359375,
+      "learning_rate": 7.314853311326196e-05,
+      "loss": 1.2402634620666504,
+      "step": 1198
+    },
+    {
+      "epoch": 0.26120818771842236,
+      "grad_norm": 0.671875,
+      "learning_rate": 7.313597110198772e-05,
+      "loss": 1.2574983835220337,
+      "step": 1199
+    },
+    {
+      "epoch": 0.26142604275405074,
+      "grad_norm": 0.69140625,
+      "learning_rate": 7.312339866590004e-05,
+      "loss": 1.2076903581619263,
+      "step": 1200
+    },
+    {
+      "epoch": 0.2616438977896791,
+      "grad_norm": 0.6640625,
+      "learning_rate": 7.311081580895431e-05,
+      "loss": 1.0989799499511719,
+      "step": 1201
+    },
+    {
+      "epoch": 0.2618617528253075,
+      "grad_norm": 0.67578125,
+      "learning_rate": 7.309822253510918e-05,
+      "loss": 1.0358539819717407,
+      "step": 1202
+    },
+    {
+      "epoch": 0.26207960786093587,
+      "grad_norm": 0.8359375,
+      "learning_rate": 7.308561884832657e-05,
+      "loss": 1.2911628484725952,
+      "step": 1203
+    },
+    {
+      "epoch": 0.26229746289656425,
+      "grad_norm": 0.62890625,
+      "learning_rate": 7.307300475257172e-05,
+      "loss": 1.1763907670974731,
+      "step": 1204
+    },
+    {
+      "epoch": 0.2625153179321926,
+      "grad_norm": 0.64453125,
+      "learning_rate": 7.306038025181309e-05,
+      "loss": 1.207767128944397,
+      "step": 1205
+    },
+    {
+      "epoch": 0.262733172967821,
+      "grad_norm": 0.65234375,
+      "learning_rate": 7.304774535002246e-05,
+      "loss": 1.1671621799468994,
+      "step": 1206
+    },
+    {
+      "epoch": 0.2629510280034494,
+      "grad_norm": 0.70703125,
+      "learning_rate": 7.303510005117483e-05,
+      "loss": 1.3397165536880493,
+      "step": 1207
+    },
+    {
+      "epoch": 0.26316888303907776,
+      "grad_norm": 0.6640625,
+      "learning_rate": 7.302244435924853e-05,
+      "loss": 1.2031677961349487,
+      "step": 1208
+    },
+    {
+      "epoch": 0.26338673807470614,
+      "grad_norm": 0.6640625,
+      "learning_rate": 7.300977827822512e-05,
+      "loss": 1.1901518106460571,
+      "step": 1209
+    },
+    {
+      "epoch": 0.2636045931103345,
+      "grad_norm": 0.77734375,
+      "learning_rate": 7.299710181208943e-05,
+      "loss": 1.2355661392211914,
+      "step": 1210
+    },
+    {
+      "epoch": 0.2638224481459629,
+      "grad_norm": 0.71484375,
+      "learning_rate": 7.298441496482959e-05,
+      "loss": 1.2124489545822144,
+      "step": 1211
+    },
+    {
+      "epoch": 0.2640403031815913,
+      "grad_norm": 0.6328125,
+      "learning_rate": 7.297171774043697e-05,
+      "loss": 1.1020777225494385,
+      "step": 1212
+    },
+    {
+      "epoch": 0.26425815821721965,
+      "grad_norm": 0.6875,
+      "learning_rate": 7.295901014290616e-05,
+      "loss": 1.175908088684082,
+      "step": 1213
+    },
+    {
+      "epoch": 0.264476013252848,
+      "grad_norm": 0.64453125,
+      "learning_rate": 7.294629217623513e-05,
+      "loss": 1.1506836414337158,
+      "step": 1214
+    },
+    {
+      "epoch": 0.26469386828847635,
+      "grad_norm": 0.7109375,
+      "learning_rate": 7.2933563844425e-05,
+      "loss": 1.3017789125442505,
+      "step": 1215
+    },
+    {
+      "epoch": 0.26491172332410473,
+      "grad_norm": 0.65234375,
+      "learning_rate": 7.292082515148021e-05,
+      "loss": 1.242101788520813,
+      "step": 1216
+    },
+    {
+      "epoch": 0.2651295783597331,
+      "grad_norm": 0.72265625,
+      "learning_rate": 7.290807610140843e-05,
+      "loss": 1.2058461904525757,
+      "step": 1217
+    },
+    {
+      "epoch": 0.2653474333953615,
+      "grad_norm": 0.6484375,
+      "learning_rate": 7.289531669822061e-05,
+      "loss": 1.0876660346984863,
+      "step": 1218
+    },
+    {
+      "epoch": 0.26556528843098987,
+      "grad_norm": 0.6484375,
+      "learning_rate": 7.288254694593097e-05,
+      "loss": 1.2918983697891235,
+      "step": 1219
+    },
+    {
+      "epoch": 0.26578314346661824,
+      "grad_norm": 0.70703125,
+      "learning_rate": 7.286976684855694e-05,
+      "loss": 1.1994441747665405,
+      "step": 1220
+    },
+    {
+      "epoch": 0.2660009985022466,
+      "grad_norm": 0.65625,
+      "learning_rate": 7.285697641011923e-05,
+      "loss": 1.1494208574295044,
+      "step": 1221
+    },
+    {
+      "epoch": 0.266218853537875,
+      "grad_norm": 0.69140625,
+      "learning_rate": 7.284417563464182e-05,
+      "loss": 1.3030284643173218,
+      "step": 1222
+    },
+    {
+      "epoch": 0.2664367085735034,
+      "grad_norm": 0.6953125,
+      "learning_rate": 7.28313645261519e-05,
+      "loss": 1.2473927736282349,
+      "step": 1223
+    },
+    {
+      "epoch": 0.26665456360913176,
+      "grad_norm": 0.6484375,
+      "learning_rate": 7.281854308867996e-05,
+      "loss": 1.1656609773635864,
+      "step": 1224
+    },
+    {
+      "epoch": 0.26687241864476013,
+      "grad_norm": 0.60546875,
+      "learning_rate": 7.280571132625971e-05,
+      "loss": 0.964602530002594,
+      "step": 1225
+    },
+    {
+      "epoch": 0.2670902736803885,
+      "grad_norm": 0.66796875,
+      "learning_rate": 7.279286924292813e-05,
+      "loss": 1.142017126083374,
+      "step": 1226
+    },
+    {
+      "epoch": 0.2673081287160169,
+      "grad_norm": 0.625,
+      "learning_rate": 7.278001684272538e-05,
+      "loss": 1.1290096044540405,
+      "step": 1227
+    },
+    {
+      "epoch": 0.26752598375164527,
+      "grad_norm": 0.65625,
+      "learning_rate": 7.276715412969497e-05,
+      "loss": 1.2729934453964233,
+      "step": 1228
+    },
+    {
+      "epoch": 0.26774383878727365,
+      "grad_norm": 0.640625,
+      "learning_rate": 7.275428110788358e-05,
+      "loss": 1.299378752708435,
+      "step": 1229
+    },
+    {
+      "epoch": 0.267961693822902,
+      "grad_norm": 0.65625,
+      "learning_rate": 7.274139778134115e-05,
+      "loss": 1.223198413848877,
+      "step": 1230
+    },
+    {
+      "epoch": 0.2681795488585304,
+      "grad_norm": 0.6875,
+      "learning_rate": 7.272850415412087e-05,
+      "loss": 1.3081732988357544,
+      "step": 1231
+    },
+    {
+      "epoch": 0.2683974038941588,
+      "grad_norm": 0.67578125,
+      "learning_rate": 7.271560023027919e-05,
+      "loss": 1.0814098119735718,
+      "step": 1232
+    },
+    {
+      "epoch": 0.26861525892978716,
+      "grad_norm": 0.62109375,
+      "learning_rate": 7.270268601387573e-05,
+      "loss": 1.192091941833496,
+      "step": 1233
+    },
+    {
+      "epoch": 0.26883311396541554,
+      "grad_norm": 0.671875,
+      "learning_rate": 7.268976150897345e-05,
+      "loss": 1.235903263092041,
+      "step": 1234
+    },
+    {
+      "epoch": 0.2690509690010439,
+      "grad_norm": 0.72265625,
+      "learning_rate": 7.267682671963845e-05,
+      "loss": 1.1973187923431396,
+      "step": 1235
+    },
+    {
+      "epoch": 0.26926882403667224,
+      "grad_norm": 0.65234375,
+      "learning_rate": 7.266388164994011e-05,
+      "loss": 1.2068250179290771,
+      "step": 1236
+    },
+    {
+      "epoch": 0.2694866790723006,
+      "grad_norm": 0.6015625,
+      "learning_rate": 7.265092630395107e-05,
+      "loss": 1.046676754951477,
+      "step": 1237
+    },
+    {
+      "epoch": 0.269704534107929,
+      "grad_norm": 0.64453125,
+      "learning_rate": 7.263796068574716e-05,
+      "loss": 1.1800862550735474,
+      "step": 1238
+    },
+    {
+      "epoch": 0.2699223891435574,
+      "grad_norm": 0.703125,
+      "learning_rate": 7.262498479940743e-05,
+      "loss": 1.1856294870376587,
+      "step": 1239
+    },
+    {
+      "epoch": 0.27014024417918575,
+      "grad_norm": 0.60546875,
+      "learning_rate": 7.261199864901423e-05,
+      "loss": 1.029008150100708,
+      "step": 1240
+    },
+    {
+      "epoch": 0.27035809921481413,
+      "grad_norm": 0.5859375,
+      "learning_rate": 7.259900223865307e-05,
+      "loss": 1.0992181301116943,
+      "step": 1241
+    },
+    {
+      "epoch": 0.2705759542504425,
+      "grad_norm": 0.6796875,
+      "learning_rate": 7.258599557241271e-05,
+      "loss": 1.239281415939331,
+      "step": 1242
+    },
+    {
+      "epoch": 0.2707938092860709,
+      "grad_norm": 0.6328125,
+      "learning_rate": 7.257297865438516e-05,
+      "loss": 1.1648595333099365,
+      "step": 1243
+    },
+    {
+      "epoch": 0.27101166432169926,
+      "grad_norm": 0.6875,
+      "learning_rate": 7.255995148866563e-05,
+      "loss": 1.2446800470352173,
+      "step": 1244
+    },
+    {
+      "epoch": 0.27122951935732764,
+      "grad_norm": 0.62890625,
+      "learning_rate": 7.254691407935254e-05,
+      "loss": 1.2320107221603394,
+      "step": 1245
+    },
+    {
+      "epoch": 0.271447374392956,
+      "grad_norm": 0.6171875,
+      "learning_rate": 7.253386643054758e-05,
+      "loss": 1.131126046180725,
+      "step": 1246
+    },
+    {
+      "epoch": 0.2716652294285844,
+      "grad_norm": 0.6953125,
+      "learning_rate": 7.25208085463556e-05,
+      "loss": 1.429965615272522,
+      "step": 1247
+    },
+    {
+      "epoch": 0.2718830844642128,
+      "grad_norm": 0.6640625,
+      "learning_rate": 7.250774043088474e-05,
+      "loss": 1.228381872177124,
+      "step": 1248
+    },
+    {
+      "epoch": 0.27210093949984115,
+      "grad_norm": 0.59375,
+      "learning_rate": 7.24946620882463e-05,
+      "loss": 1.043805480003357,
+      "step": 1249
+    },
+    {
+      "epoch": 0.27231879453546953,
+      "grad_norm": 0.6796875,
+      "learning_rate": 7.248157352255483e-05,
+      "loss": 1.2669799327850342,
+      "step": 1250
+    },
+    {
+      "epoch": 0.2725366495710979,
+      "grad_norm": 0.6015625,
+      "learning_rate": 7.246847473792808e-05,
+      "loss": 1.0299328565597534,
+      "step": 1251
+    },
+    {
+      "epoch": 0.2727545046067263,
+      "grad_norm": 0.6640625,
+      "learning_rate": 7.245536573848703e-05,
+      "loss": 1.1852364540100098,
+      "step": 1252
+    },
+    {
+      "epoch": 0.27297235964235467,
+      "grad_norm": 0.66796875,
+      "learning_rate": 7.244224652835586e-05,
+      "loss": 1.2790281772613525,
+      "step": 1253
+    },
+    {
+      "epoch": 0.27319021467798305,
+      "grad_norm": 0.66796875,
+      "learning_rate": 7.242911711166196e-05,
+      "loss": 1.3223212957382202,
+      "step": 1254
+    },
+    {
+      "epoch": 0.2734080697136114,
+      "grad_norm": 0.66015625,
+      "learning_rate": 7.241597749253593e-05,
+      "loss": 1.0831646919250488,
+      "step": 1255
+    },
+    {
+      "epoch": 0.2736259247492398,
+      "grad_norm": 0.60546875,
+      "learning_rate": 7.240282767511162e-05,
+      "loss": 1.0070009231567383,
+      "step": 1256
+    },
+    {
+      "epoch": 0.2738437797848682,
+      "grad_norm": 0.73046875,
+      "learning_rate": 7.238966766352604e-05,
+      "loss": 1.3019137382507324,
+      "step": 1257
+    },
+    {
+      "epoch": 0.2740616348204965,
+      "grad_norm": 0.59375,
+      "learning_rate": 7.237649746191942e-05,
+      "loss": 1.122825026512146,
+      "step": 1258
+    },
+    {
+      "epoch": 0.2742794898561249,
+      "grad_norm": 0.67578125,
+      "learning_rate": 7.236331707443521e-05,
+      "loss": 1.2877912521362305,
+      "step": 1259
+    },
+    {
+      "epoch": 0.27449734489175326,
+      "grad_norm": 0.69921875,
+      "learning_rate": 7.235012650522005e-05,
+      "loss": 1.233075737953186,
+      "step": 1260
+    },
+    {
+      "epoch": 0.27471519992738164,
+      "grad_norm": 0.73046875,
+      "learning_rate": 7.233692575842377e-05,
+      "loss": 1.2717688083648682,
+      "step": 1261
+    },
+    {
+      "epoch": 0.27493305496301,
+      "grad_norm": 0.66796875,
+      "learning_rate": 7.232371483819944e-05,
+      "loss": 1.2347332239151,
+      "step": 1262
+    },
+    {
+      "epoch": 0.2751509099986384,
+      "grad_norm": 0.62109375,
+      "learning_rate": 7.231049374870329e-05,
+      "loss": 1.116786241531372,
+      "step": 1263
+    },
+    {
+      "epoch": 0.27536876503426677,
+      "grad_norm": 0.625,
+      "learning_rate": 7.22972624940948e-05,
+      "loss": 1.0687594413757324,
+      "step": 1264
+    },
+    {
+      "epoch": 0.27558662006989515,
+      "grad_norm": 0.6484375,
+      "learning_rate": 7.22840210785366e-05,
+      "loss": 1.0891162157058716,
+      "step": 1265
+    },
+    {
+      "epoch": 0.2758044751055235,
+      "grad_norm": 0.65234375,
+      "learning_rate": 7.227076950619452e-05,
+      "loss": 1.1440852880477905,
+      "step": 1266
+    },
+    {
+      "epoch": 0.2760223301411519,
+      "grad_norm": 0.63671875,
+      "learning_rate": 7.225750778123763e-05,
+      "loss": 1.2287821769714355,
+      "step": 1267
+    },
+    {
+      "epoch": 0.2762401851767803,
+      "grad_norm": 0.6640625,
+      "learning_rate": 7.224423590783814e-05,
+      "loss": 1.265513300895691,
+      "step": 1268
+    },
+    {
+      "epoch": 0.27645804021240866,
+      "grad_norm": 0.6796875,
+      "learning_rate": 7.223095389017149e-05,
+      "loss": 1.20741605758667,
+      "step": 1269
+    },
+    {
+      "epoch": 0.27667589524803704,
+      "grad_norm": 0.609375,
+      "learning_rate": 7.221766173241629e-05,
+      "loss": 1.1178297996520996,
+      "step": 1270
+    },
+    {
+      "epoch": 0.2768937502836654,
+      "grad_norm": 0.67578125,
+      "learning_rate": 7.220435943875434e-05,
+      "loss": 1.365719199180603,
+      "step": 1271
+    },
+    {
+      "epoch": 0.2771116053192938,
+      "grad_norm": 0.69140625,
+      "learning_rate": 7.219104701337064e-05,
+      "loss": 1.2471543550491333,
+      "step": 1272
+    },
+    {
+      "epoch": 0.2773294603549222,
+      "grad_norm": 0.671875,
+      "learning_rate": 7.217772446045339e-05,
+      "loss": 1.1604433059692383,
+      "step": 1273
+    },
+    {
+      "epoch": 0.27754731539055055,
+      "grad_norm": 0.609375,
+      "learning_rate": 7.216439178419395e-05,
+      "loss": 1.0599409341812134,
+      "step": 1274
+    },
+    {
+      "epoch": 0.27776517042617893,
+      "grad_norm": 0.6015625,
+      "learning_rate": 7.215104898878688e-05,
+      "loss": 1.0941181182861328,
+      "step": 1275
+    },
+    {
+      "epoch": 0.2779830254618073,
+      "grad_norm": 0.65625,
+      "learning_rate": 7.21376960784299e-05,
+      "loss": 1.0786826610565186,
+      "step": 1276
+    },
+    {
+      "epoch": 0.2782008804974357,
+      "grad_norm": 0.64453125,
+      "learning_rate": 7.212433305732395e-05,
+      "loss": 1.196561574935913,
+      "step": 1277
+    },
+    {
+      "epoch": 0.27841873553306407,
+      "grad_norm": 0.67578125,
+      "learning_rate": 7.211095992967314e-05,
+      "loss": 1.1972441673278809,
+      "step": 1278
+    },
+    {
+      "epoch": 0.2786365905686924,
+      "grad_norm": 0.6640625,
+      "learning_rate": 7.209757669968474e-05,
+      "loss": 1.1630479097366333,
+      "step": 1279
+    },
+    {
+      "epoch": 0.27885444560432077,
+      "grad_norm": 0.65234375,
+      "learning_rate": 7.20841833715692e-05,
+      "loss": 1.240605354309082,
+      "step": 1280
+    },
+    {
+      "epoch": 0.27907230063994914,
+      "grad_norm": 0.66015625,
+      "learning_rate": 7.207077994954015e-05,
+      "loss": 1.1704919338226318,
+      "step": 1281
+    },
+    {
+      "epoch": 0.2792901556755775,
+      "grad_norm": 0.66015625,
+      "learning_rate": 7.205736643781445e-05,
+      "loss": 1.1384356021881104,
+      "step": 1282
+    },
+    {
+      "epoch": 0.2795080107112059,
+      "grad_norm": 0.65234375,
+      "learning_rate": 7.204394284061204e-05,
+      "loss": 1.1318368911743164,
+      "step": 1283
+    },
+    {
+      "epoch": 0.2797258657468343,
+      "grad_norm": 0.6171875,
+      "learning_rate": 7.203050916215608e-05,
+      "loss": 1.038714051246643,
+      "step": 1284
+    },
+    {
+      "epoch": 0.27994372078246266,
+      "grad_norm": 0.68359375,
+      "learning_rate": 7.201706540667293e-05,
+      "loss": 1.403035283088684,
+      "step": 1285
+    },
+    {
+      "epoch": 0.28016157581809104,
+      "grad_norm": 0.65625,
+      "learning_rate": 7.200361157839208e-05,
+      "loss": 1.3605185747146606,
+      "step": 1286
+    },
+    {
+      "epoch": 0.2803794308537194,
+      "grad_norm": 0.65234375,
+      "learning_rate": 7.19901476815462e-05,
+      "loss": 1.2092698812484741,
+      "step": 1287
+    },
+    {
+      "epoch": 0.2805972858893478,
+      "grad_norm": 0.56640625,
+      "learning_rate": 7.197667372037112e-05,
+      "loss": 1.0797160863876343,
+      "step": 1288
+    },
+    {
+      "epoch": 0.28081514092497617,
+      "grad_norm": 0.61328125,
+      "learning_rate": 7.196318969910585e-05,
+      "loss": 1.0315220355987549,
+      "step": 1289
+    },
+    {
+      "epoch": 0.28103299596060455,
+      "grad_norm": 0.6484375,
+      "learning_rate": 7.194969562199256e-05,
+      "loss": 1.2441558837890625,
+      "step": 1290
+    },
+    {
+      "epoch": 0.2812508509962329,
+      "grad_norm": 0.6171875,
+      "learning_rate": 7.193619149327658e-05,
+      "loss": 1.163487434387207,
+      "step": 1291
+    },
+    {
+      "epoch": 0.2814687060318613,
+      "grad_norm": 0.65625,
+      "learning_rate": 7.192267731720641e-05,
+      "loss": 1.2346569299697876,
+      "step": 1292
+    },
+    {
+      "epoch": 0.2816865610674897,
+      "grad_norm": 0.70703125,
+      "learning_rate": 7.190915309803371e-05,
+      "loss": 1.2528342008590698,
+      "step": 1293
+    },
+    {
+      "epoch": 0.28190441610311806,
+      "grad_norm": 0.65625,
+      "learning_rate": 7.189561884001329e-05,
+      "loss": 1.290441870689392,
+      "step": 1294
+    },
+    {
+      "epoch": 0.28212227113874644,
+      "grad_norm": 0.63671875,
+      "learning_rate": 7.188207454740312e-05,
+      "loss": 1.2861841917037964,
+      "step": 1295
+    },
+    {
+      "epoch": 0.2823401261743748,
+      "grad_norm": 0.6328125,
+      "learning_rate": 7.186852022446435e-05,
+      "loss": 1.2937500476837158,
+      "step": 1296
+    },
+    {
+      "epoch": 0.2825579812100032,
+      "grad_norm": 0.671875,
+      "learning_rate": 7.185495587546124e-05,
+      "loss": 1.113438367843628,
+      "step": 1297
+    },
+    {
+      "epoch": 0.2827758362456316,
+      "grad_norm": 0.64453125,
+      "learning_rate": 7.184138150466124e-05,
+      "loss": 1.2728081941604614,
+      "step": 1298
+    },
+    {
+      "epoch": 0.28299369128125995,
+      "grad_norm": 0.640625,
+      "learning_rate": 7.182779711633498e-05,
+      "loss": 1.121742844581604,
+      "step": 1299
+    },
+    {
+      "epoch": 0.28321154631688833,
+      "grad_norm": 0.65234375,
+      "learning_rate": 7.181420271475616e-05,
+      "loss": 1.3175054788589478,
+      "step": 1300
+    },
+    {
+      "epoch": 0.28342940135251665,
+      "grad_norm": 0.70703125,
+      "learning_rate": 7.180059830420167e-05,
+      "loss": 1.161149501800537,
+      "step": 1301
+    },
+    {
+      "epoch": 0.28364725638814503,
+      "grad_norm": 0.6484375,
+      "learning_rate": 7.178698388895159e-05,
+      "loss": 1.2229605913162231,
+      "step": 1302
+    },
+    {
+      "epoch": 0.2838651114237734,
+      "grad_norm": 0.66015625,
+      "learning_rate": 7.17733594732891e-05,
+      "loss": 1.156973123550415,
+      "step": 1303
+    },
+    {
+      "epoch": 0.2840829664594018,
+      "grad_norm": 0.62890625,
+      "learning_rate": 7.175972506150053e-05,
+      "loss": 1.2616362571716309,
+      "step": 1304
+    },
+    {
+      "epoch": 0.28430082149503016,
+      "grad_norm": 0.625,
+      "learning_rate": 7.174608065787539e-05,
+      "loss": 1.2470351457595825,
+      "step": 1305
+    },
+    {
+      "epoch": 0.28451867653065854,
+      "grad_norm": 0.6796875,
+      "learning_rate": 7.173242626670626e-05,
+      "loss": 1.2407808303833008,
+      "step": 1306
+    },
+    {
+      "epoch": 0.2847365315662869,
+      "grad_norm": 0.66015625,
+      "learning_rate": 7.171876189228894e-05,
+      "loss": 1.0952699184417725,
+      "step": 1307
+    },
+    {
+      "epoch": 0.2849543866019153,
+      "grad_norm": 0.70703125,
+      "learning_rate": 7.170508753892235e-05,
+      "loss": 1.1929556131362915,
+      "step": 1308
+    },
+    {
+      "epoch": 0.2851722416375437,
+      "grad_norm": 0.61328125,
+      "learning_rate": 7.169140321090852e-05,
+      "loss": 1.027903437614441,
+      "step": 1309
+    },
+    {
+      "epoch": 0.28539009667317206,
+      "grad_norm": 0.6171875,
+      "learning_rate": 7.167770891255264e-05,
+      "loss": 1.059259057044983,
+      "step": 1310
+    },
+    {
+      "epoch": 0.28560795170880043,
+      "grad_norm": 0.75390625,
+      "learning_rate": 7.166400464816304e-05,
+      "loss": 1.3059710264205933,
+      "step": 1311
+    },
+    {
+      "epoch": 0.2858258067444288,
+      "grad_norm": 0.61328125,
+      "learning_rate": 7.165029042205117e-05,
+      "loss": 1.1457419395446777,
+      "step": 1312
+    },
+    {
+      "epoch": 0.2860436617800572,
+      "grad_norm": 0.62109375,
+      "learning_rate": 7.163656623853164e-05,
+      "loss": 1.054082989692688,
+      "step": 1313
+    },
+    {
+      "epoch": 0.28626151681568557,
+      "grad_norm": 0.640625,
+      "learning_rate": 7.162283210192216e-05,
+      "loss": 1.2969257831573486,
+      "step": 1314
+    },
+    {
+      "epoch": 0.28647937185131395,
+      "grad_norm": 0.6484375,
+      "learning_rate": 7.160908801654358e-05,
+      "loss": 1.057404637336731,
+      "step": 1315
+    },
+    {
+      "epoch": 0.2866972268869423,
+      "grad_norm": 0.66015625,
+      "learning_rate": 7.159533398671993e-05,
+      "loss": 1.296829342842102,
+      "step": 1316
+    },
+    {
+      "epoch": 0.2869150819225707,
+      "grad_norm": 0.67578125,
+      "learning_rate": 7.158157001677828e-05,
+      "loss": 1.1640996932983398,
+      "step": 1317
+    },
+    {
+      "epoch": 0.2871329369581991,
+      "grad_norm": 0.66015625,
+      "learning_rate": 7.15677961110489e-05,
+      "loss": 1.2425172328948975,
+      "step": 1318
+    },
+    {
+      "epoch": 0.28735079199382746,
+      "grad_norm": 0.6328125,
+      "learning_rate": 7.155401227386514e-05,
+      "loss": 1.0855642557144165,
+      "step": 1319
+    },
+    {
+      "epoch": 0.28756864702945584,
+      "grad_norm": 0.609375,
+      "learning_rate": 7.154021850956352e-05,
+      "loss": 1.2285504341125488,
+      "step": 1320
+    },
+    {
+      "epoch": 0.2877865020650842,
+      "grad_norm": 0.5703125,
+      "learning_rate": 7.152641482248362e-05,
+      "loss": 1.0136677026748657,
+      "step": 1321
+    },
+    {
+      "epoch": 0.2880043571007126,
+      "grad_norm": 0.6640625,
+      "learning_rate": 7.151260121696821e-05,
+      "loss": 1.2524278163909912,
+      "step": 1322
+    },
+    {
+      "epoch": 0.2882222121363409,
+      "grad_norm": 0.6328125,
+      "learning_rate": 7.149877769736313e-05,
+      "loss": 1.240539312362671,
+      "step": 1323
+    },
+    {
+      "epoch": 0.2884400671719693,
+      "grad_norm": 0.625,
+      "learning_rate": 7.148494426801737e-05,
+      "loss": 1.1506719589233398,
+      "step": 1324
+    },
+    {
+      "epoch": 0.2886579222075977,
+      "grad_norm": 0.703125,
+      "learning_rate": 7.147110093328302e-05,
+      "loss": 1.2524564266204834,
+      "step": 1325
+    },
+    {
+      "epoch": 0.28887577724322605,
+      "grad_norm": 0.6484375,
+      "learning_rate": 7.145724769751531e-05,
+      "loss": 1.2418291568756104,
+      "step": 1326
+    },
+    {
+      "epoch": 0.28909363227885443,
+      "grad_norm": 0.66015625,
+      "learning_rate": 7.144338456507254e-05,
+      "loss": 1.2193326950073242,
+      "step": 1327
+    },
+    {
+      "epoch": 0.2893114873144828,
+      "grad_norm": 0.61328125,
+      "learning_rate": 7.142951154031617e-05,
+      "loss": 1.192091464996338,
+      "step": 1328
+    },
+    {
+      "epoch": 0.2895293423501112,
+      "grad_norm": 0.66015625,
+      "learning_rate": 7.141562862761072e-05,
+      "loss": 1.316902995109558,
+      "step": 1329
+    },
+    {
+      "epoch": 0.28974719738573956,
+      "grad_norm": 0.64453125,
+      "learning_rate": 7.14017358313239e-05,
+      "loss": 1.2109570503234863,
+      "step": 1330
+    },
+    {
+      "epoch": 0.28996505242136794,
+      "grad_norm": 0.64453125,
+      "learning_rate": 7.138783315582646e-05,
+      "loss": 1.224020004272461,
+      "step": 1331
+    },
+    {
+      "epoch": 0.2901829074569963,
+      "grad_norm": 0.6328125,
+      "learning_rate": 7.137392060549228e-05,
+      "loss": 1.114842414855957,
+      "step": 1332
+    },
+    {
+      "epoch": 0.2904007624926247,
+      "grad_norm": 0.609375,
+      "learning_rate": 7.135999818469834e-05,
+      "loss": 1.133542537689209,
+      "step": 1333
+    },
+    {
+      "epoch": 0.2906186175282531,
+      "grad_norm": 0.625,
+      "learning_rate": 7.134606589782476e-05,
+      "loss": 1.1622140407562256,
+      "step": 1334
+    },
+    {
+      "epoch": 0.29083647256388145,
+      "grad_norm": 0.62109375,
+      "learning_rate": 7.133212374925471e-05,
+      "loss": 1.0408300161361694,
+      "step": 1335
+    },
+    {
+      "epoch": 0.29105432759950983,
+      "grad_norm": 0.640625,
+      "learning_rate": 7.131817174337451e-05,
+      "loss": 1.1643146276474,
+      "step": 1336
+    },
+    {
+      "epoch": 0.2912721826351382,
+      "grad_norm": 0.65625,
+      "learning_rate": 7.130420988457354e-05,
+      "loss": 1.1958473920822144,
+      "step": 1337
+    },
+    {
+      "epoch": 0.2914900376707666,
+      "grad_norm": 0.66015625,
+      "learning_rate": 7.129023817724433e-05,
+      "loss": 1.2120956182479858,
+      "step": 1338
+    },
+    {
+      "epoch": 0.29170789270639497,
+      "grad_norm": 0.63671875,
+      "learning_rate": 7.127625662578243e-05,
+      "loss": 1.318742036819458,
+      "step": 1339
+    },
+    {
+      "epoch": 0.29192574774202334,
+      "grad_norm": 0.640625,
+      "learning_rate": 7.12622652345866e-05,
+      "loss": 1.3126004934310913,
+      "step": 1340
+    },
+    {
+      "epoch": 0.2921436027776517,
+      "grad_norm": 0.61328125,
+      "learning_rate": 7.124826400805857e-05,
+      "loss": 1.1231520175933838,
+      "step": 1341
+    },
+    {
+      "epoch": 0.2923614578132801,
+      "grad_norm": 0.64453125,
+      "learning_rate": 7.123425295060328e-05,
+      "loss": 1.3046287298202515,
+      "step": 1342
+    },
+    {
+      "epoch": 0.2925793128489085,
+      "grad_norm": 0.62109375,
+      "learning_rate": 7.122023206662866e-05,
+      "loss": 1.090712308883667,
+      "step": 1343
+    },
+    {
+      "epoch": 0.29279716788453686,
+      "grad_norm": 0.6484375,
+      "learning_rate": 7.120620136054583e-05,
+      "loss": 1.293134093284607,
+      "step": 1344
+    },
+    {
+      "epoch": 0.2930150229201652,
+      "grad_norm": 0.578125,
+      "learning_rate": 7.11921608367689e-05,
+      "loss": 1.026340365409851,
+      "step": 1345
+    },
+    {
+      "epoch": 0.29323287795579356,
+      "grad_norm": 0.62109375,
+      "learning_rate": 7.117811049971515e-05,
+      "loss": 1.0833635330200195,
+      "step": 1346
+    },
+    {
+      "epoch": 0.29345073299142194,
+      "grad_norm": 0.6953125,
+      "learning_rate": 7.11640503538049e-05,
+      "loss": 1.1696686744689941,
+      "step": 1347
+    },
+    {
+      "epoch": 0.2936685880270503,
+      "grad_norm": 0.62109375,
+      "learning_rate": 7.11499804034616e-05,
+      "loss": 1.1921495199203491,
+      "step": 1348
+    },
+    {
+      "epoch": 0.2938864430626787,
+      "grad_norm": 0.671875,
+      "learning_rate": 7.113590065311173e-05,
+      "loss": 1.1036195755004883,
+      "step": 1349
+    },
+    {
+      "epoch": 0.29410429809830707,
+      "grad_norm": 0.6015625,
+      "learning_rate": 7.112181110718489e-05,
+      "loss": 1.0858298540115356,
+      "step": 1350
+    },
+    {
+      "epoch": 0.29432215313393545,
+      "grad_norm": 0.62890625,
+      "learning_rate": 7.110771177011375e-05,
+      "loss": 1.1644235849380493,
+      "step": 1351
+    },
+    {
+      "epoch": 0.2945400081695638,
+      "grad_norm": 0.6875,
+      "learning_rate": 7.109360264633406e-05,
+      "loss": 1.2473119497299194,
+      "step": 1352
+    },
+    {
+      "epoch": 0.2947578632051922,
+      "grad_norm": 0.640625,
+      "learning_rate": 7.107948374028467e-05,
+      "loss": 1.3091998100280762,
+      "step": 1353
+    },
+    {
+      "epoch": 0.2949757182408206,
+      "grad_norm": 0.6015625,
+      "learning_rate": 7.106535505640746e-05,
+      "loss": 1.1931923627853394,
+      "step": 1354
+    },
+    {
+      "epoch": 0.29519357327644896,
+      "grad_norm": 0.69921875,
+      "learning_rate": 7.105121659914743e-05,
+      "loss": 1.188901662826538,
+      "step": 1355
+    },
+    {
+      "epoch": 0.29541142831207734,
+      "grad_norm": 0.671875,
+      "learning_rate": 7.103706837295264e-05,
+      "loss": 1.152740716934204,
+      "step": 1356
+    },
+    {
+      "epoch": 0.2956292833477057,
+      "grad_norm": 0.6484375,
+      "learning_rate": 7.102291038227423e-05,
+      "loss": 1.2586299180984497,
+      "step": 1357
+    },
+    {
+      "epoch": 0.2958471383833341,
+      "grad_norm": 0.66015625,
+      "learning_rate": 7.100874263156637e-05,
+      "loss": 1.229476809501648,
+      "step": 1358
+    },
+    {
+      "epoch": 0.2960649934189625,
+      "grad_norm": 0.66015625,
+      "learning_rate": 7.099456512528639e-05,
+      "loss": 1.2067805528640747,
+      "step": 1359
+    },
+    {
+      "epoch": 0.29628284845459085,
+      "grad_norm": 0.5703125,
+      "learning_rate": 7.09803778678946e-05,
+      "loss": 1.0319379568099976,
+      "step": 1360
+    },
+    {
+      "epoch": 0.29650070349021923,
+      "grad_norm": 0.625,
+      "learning_rate": 7.096618086385441e-05,
+      "loss": 1.1077347993850708,
+      "step": 1361
+    },
+    {
+      "epoch": 0.2967185585258476,
+      "grad_norm": 0.67578125,
+      "learning_rate": 7.095197411763231e-05,
+      "loss": 1.1801555156707764,
+      "step": 1362
+    },
+    {
+      "epoch": 0.296936413561476,
+      "grad_norm": 0.6484375,
+      "learning_rate": 7.093775763369785e-05,
+      "loss": 1.13539457321167,
+      "step": 1363
+    },
+    {
+      "epoch": 0.29715426859710437,
+      "grad_norm": 0.625,
+      "learning_rate": 7.092353141652363e-05,
+      "loss": 1.2413476705551147,
+      "step": 1364
+    },
+    {
+      "epoch": 0.29737212363273274,
+      "grad_norm": 0.63671875,
+      "learning_rate": 7.090929547058532e-05,
+      "loss": 1.17805814743042,
+      "step": 1365
+    },
+    {
+      "epoch": 0.2975899786683611,
+      "grad_norm": 0.640625,
+      "learning_rate": 7.089504980036163e-05,
+      "loss": 1.188424825668335,
+      "step": 1366
+    },
+    {
+      "epoch": 0.29780783370398944,
+      "grad_norm": 0.65234375,
+      "learning_rate": 7.088079441033439e-05,
+      "loss": 1.152357578277588,
+      "step": 1367
+    },
+    {
+      "epoch": 0.2980256887396178,
+      "grad_norm": 0.66015625,
+      "learning_rate": 7.086652930498843e-05,
+      "loss": 1.112777590751648,
+      "step": 1368
+    },
+    {
+      "epoch": 0.2982435437752462,
+      "grad_norm": 0.66015625,
+      "learning_rate": 7.085225448881163e-05,
+      "loss": 1.1788599491119385,
+      "step": 1369
+    },
+    {
+      "epoch": 0.2984613988108746,
+      "grad_norm": 0.671875,
+      "learning_rate": 7.083796996629499e-05,
+      "loss": 1.3916118144989014,
+      "step": 1370
+    },
+    {
+      "epoch": 0.29867925384650296,
+      "grad_norm": 0.64453125,
+      "learning_rate": 7.08236757419325e-05,
+      "loss": 1.1751484870910645,
+      "step": 1371
+    },
+    {
+      "epoch": 0.29889710888213133,
+      "grad_norm": 0.6875,
+      "learning_rate": 7.080937182022122e-05,
+      "loss": 1.0790146589279175,
+      "step": 1372
+    },
+    {
+      "epoch": 0.2991149639177597,
+      "grad_norm": 0.63671875,
+      "learning_rate": 7.079505820566128e-05,
+      "loss": 1.2723995447158813,
+      "step": 1373
+    },
+    {
+      "epoch": 0.2993328189533881,
+      "grad_norm": 0.62890625,
+      "learning_rate": 7.078073490275585e-05,
+      "loss": 1.1541448831558228,
+      "step": 1374
+    },
+    {
+      "epoch": 0.29955067398901647,
+      "grad_norm": 0.6640625,
+      "learning_rate": 7.076640191601113e-05,
+      "loss": 1.1797406673431396,
+      "step": 1375
+    },
+    {
+      "epoch": 0.29976852902464485,
+      "grad_norm": 0.65234375,
+      "learning_rate": 7.075205924993638e-05,
+      "loss": 1.209267497062683,
+      "step": 1376
+    },
+    {
+      "epoch": 0.2999863840602732,
+      "grad_norm": 0.625,
+      "learning_rate": 7.07377069090439e-05,
+      "loss": 1.0449070930480957,
+      "step": 1377
+    },
+    {
+      "epoch": 0.3002042390959016,
+      "grad_norm": 0.61328125,
+      "learning_rate": 7.072334489784906e-05,
+      "loss": 1.0917119979858398,
+      "step": 1378
+    },
+    {
+      "epoch": 0.30042209413153,
+      "grad_norm": 0.64453125,
+      "learning_rate": 7.070897322087024e-05,
+      "loss": 1.228953242301941,
+      "step": 1379
+    },
+    {
+      "epoch": 0.30063994916715836,
+      "grad_norm": 0.65234375,
+      "learning_rate": 7.069459188262887e-05,
+      "loss": 1.2287639379501343,
+      "step": 1380
+    },
+    {
+      "epoch": 0.30085780420278674,
+      "grad_norm": 0.671875,
+      "learning_rate": 7.068020088764942e-05,
+      "loss": 1.24830162525177,
+      "step": 1381
+    },
+    {
+      "epoch": 0.3010756592384151,
+      "grad_norm": 0.63671875,
+      "learning_rate": 7.06658002404594e-05,
+      "loss": 1.2582188844680786,
+      "step": 1382
+    },
+    {
+      "epoch": 0.3012935142740435,
+      "grad_norm": 0.640625,
+      "learning_rate": 7.065138994558937e-05,
+      "loss": 1.1680468320846558,
+      "step": 1383
+    },
+    {
+      "epoch": 0.3015113693096719,
+      "grad_norm": 0.62109375,
+      "learning_rate": 7.063697000757289e-05,
+      "loss": 1.1224597692489624,
+      "step": 1384
+    },
+    {
+      "epoch": 0.30172922434530025,
+      "grad_norm": 0.6328125,
+      "learning_rate": 7.06225404309466e-05,
+      "loss": 1.1221766471862793,
+      "step": 1385
+    },
+    {
+      "epoch": 0.30194707938092863,
+      "grad_norm": 0.6171875,
+      "learning_rate": 7.060810122025013e-05,
+      "loss": 1.1782262325286865,
+      "step": 1386
+    },
+    {
+      "epoch": 0.302164934416557,
+      "grad_norm": 0.6640625,
+      "learning_rate": 7.059365238002617e-05,
+      "loss": 1.18111252784729,
+      "step": 1387
+    },
+    {
+      "epoch": 0.3023827894521854,
+      "grad_norm": 0.65234375,
+      "learning_rate": 7.057919391482042e-05,
+      "loss": 1.2563743591308594,
+      "step": 1388
+    },
+    {
+      "epoch": 0.3026006444878137,
+      "grad_norm": 0.6875,
+      "learning_rate": 7.056472582918161e-05,
+      "loss": 1.2916724681854248,
+      "step": 1389
+    },
+    {
+      "epoch": 0.3028184995234421,
+      "grad_norm": 0.59765625,
+      "learning_rate": 7.055024812766153e-05,
+      "loss": 1.1345653533935547,
+      "step": 1390
+    },
+    {
+      "epoch": 0.30303635455907046,
+      "grad_norm": 0.625,
+      "learning_rate": 7.053576081481494e-05,
+      "loss": 1.1846600770950317,
+      "step": 1391
+    },
+    {
+      "epoch": 0.30325420959469884,
+      "grad_norm": 0.6171875,
+      "learning_rate": 7.052126389519966e-05,
+      "loss": 1.176259994506836,
+      "step": 1392
+    },
+    {
+      "epoch": 0.3034720646303272,
+      "grad_norm": 0.69140625,
+      "learning_rate": 7.050675737337654e-05,
+      "loss": 1.1991217136383057,
+      "step": 1393
+    },
+    {
+      "epoch": 0.3036899196659556,
+      "grad_norm": 0.63671875,
+      "learning_rate": 7.049224125390943e-05,
+      "loss": 1.2026718854904175,
+      "step": 1394
+    },
+    {
+      "epoch": 0.303907774701584,
+      "grad_norm": 0.6484375,
+      "learning_rate": 7.047771554136518e-05,
+      "loss": 1.1924123764038086,
+      "step": 1395
+    },
+    {
+      "epoch": 0.30412562973721236,
+      "grad_norm": 0.65625,
+      "learning_rate": 7.04631802403137e-05,
+      "loss": 1.233344554901123,
+      "step": 1396
+    },
+    {
+      "epoch": 0.30434348477284073,
+      "grad_norm": 0.5703125,
+      "learning_rate": 7.044863535532792e-05,
+      "loss": 0.900428056716919,
+      "step": 1397
+    },
+    {
+      "epoch": 0.3045613398084691,
+      "grad_norm": 0.73046875,
+      "learning_rate": 7.043408089098373e-05,
+      "loss": 1.2982444763183594,
+      "step": 1398
+    },
+    {
+      "epoch": 0.3047791948440975,
+      "grad_norm": 0.62109375,
+      "learning_rate": 7.041951685186009e-05,
+      "loss": 1.130938172340393,
+      "step": 1399
+    },
+    {
+      "epoch": 0.30499704987972587,
+      "grad_norm": 0.62109375,
+      "learning_rate": 7.040494324253894e-05,
+      "loss": 1.233506202697754,
+      "step": 1400
+    },
+    {
+      "epoch": 0.30521490491535425,
+      "grad_norm": 0.671875,
+      "learning_rate": 7.039036006760526e-05,
+      "loss": 1.2724239826202393,
+      "step": 1401
+    },
+    {
+      "epoch": 0.3054327599509826,
+      "grad_norm": 0.6484375,
+      "learning_rate": 7.0375767331647e-05,
+      "loss": 1.1857662200927734,
+      "step": 1402
+    },
+    {
+      "epoch": 0.305650614986611,
+      "grad_norm": 0.67578125,
+      "learning_rate": 7.036116503925514e-05,
+      "loss": 1.2770614624023438,
+      "step": 1403
+    },
+    {
+      "epoch": 0.3058684700222394,
+      "grad_norm": 0.62109375,
+      "learning_rate": 7.034655319502369e-05,
+      "loss": 1.094124436378479,
+      "step": 1404
+    },
+    {
+      "epoch": 0.30608632505786776,
+      "grad_norm": 0.640625,
+      "learning_rate": 7.03319318035496e-05,
+      "loss": 1.2934670448303223,
+      "step": 1405
+    },
+    {
+      "epoch": 0.30630418009349614,
+      "grad_norm": 0.609375,
+      "learning_rate": 7.031730086943291e-05,
+      "loss": 1.1227643489837646,
+      "step": 1406
+    },
+    {
+      "epoch": 0.3065220351291245,
+      "grad_norm": 0.67578125,
+      "learning_rate": 7.030266039727658e-05,
+      "loss": 1.2326757907867432,
+      "step": 1407
+    },
+    {
+      "epoch": 0.3067398901647529,
+      "grad_norm": 0.64453125,
+      "learning_rate": 7.028801039168663e-05,
+      "loss": 1.1312904357910156,
+      "step": 1408
+    },
+    {
+      "epoch": 0.30695774520038127,
+      "grad_norm": 0.625,
+      "learning_rate": 7.027335085727205e-05,
+      "loss": 1.2237826585769653,
+      "step": 1409
+    },
+    {
+      "epoch": 0.30717560023600965,
+      "grad_norm": 0.640625,
+      "learning_rate": 7.025868179864482e-05,
+      "loss": 1.0930768251419067,
+      "step": 1410
+    },
+    {
+      "epoch": 0.30739345527163797,
+      "grad_norm": 0.66015625,
+      "learning_rate": 7.024400322041996e-05,
+      "loss": 1.1850069761276245,
+      "step": 1411
+    },
+    {
+      "epoch": 0.30761131030726635,
+      "grad_norm": 0.640625,
+      "learning_rate": 7.022931512721541e-05,
+      "loss": 1.1926578283309937,
+      "step": 1412
+    },
+    {
+      "epoch": 0.30782916534289473,
+      "grad_norm": 0.65625,
+      "learning_rate": 7.021461752365221e-05,
+      "loss": 1.214386224746704,
+      "step": 1413
+    },
+    {
+      "epoch": 0.3080470203785231,
+      "grad_norm": 0.62109375,
+      "learning_rate": 7.01999104143543e-05,
+      "loss": 1.0641000270843506,
+      "step": 1414
+    },
+    {
+      "epoch": 0.3082648754141515,
+      "grad_norm": 0.60546875,
+      "learning_rate": 7.01851938039486e-05,
+      "loss": 1.011061668395996,
+      "step": 1415
+    },
+    {
+      "epoch": 0.30848273044977986,
+      "grad_norm": 0.65625,
+      "learning_rate": 7.017046769706515e-05,
+      "loss": 1.1176605224609375,
+      "step": 1416
+    },
+    {
+      "epoch": 0.30870058548540824,
+      "grad_norm": 0.61328125,
+      "learning_rate": 7.015573209833681e-05,
+      "loss": 1.1254489421844482,
+      "step": 1417
+    },
+    {
+      "epoch": 0.3089184405210366,
+      "grad_norm": 0.625,
+      "learning_rate": 7.014098701239955e-05,
+      "loss": 1.0963956117630005,
+      "step": 1418
+    },
+    {
+      "epoch": 0.309136295556665,
+      "grad_norm": 0.60546875,
+      "learning_rate": 7.012623244389224e-05,
+      "loss": 1.1628215312957764,
+      "step": 1419
+    },
+    {
+      "epoch": 0.3093541505922934,
+      "grad_norm": 0.62890625,
+      "learning_rate": 7.011146839745681e-05,
+      "loss": 1.129587173461914,
+      "step": 1420
+    },
+    {
+      "epoch": 0.30957200562792175,
+      "grad_norm": 0.6796875,
+      "learning_rate": 7.009669487773814e-05,
+      "loss": 1.2179979085922241,
+      "step": 1421
+    },
+    {
+      "epoch": 0.30978986066355013,
+      "grad_norm": 0.59765625,
+      "learning_rate": 7.008191188938405e-05,
+      "loss": 0.954366147518158,
+      "step": 1422
+    },
+    {
+      "epoch": 0.3100077156991785,
+      "grad_norm": 0.66796875,
+      "learning_rate": 7.00671194370454e-05,
+      "loss": 1.2386715412139893,
+      "step": 1423
+    },
+    {
+      "epoch": 0.3102255707348069,
+      "grad_norm": 0.578125,
+      "learning_rate": 7.005231752537597e-05,
+      "loss": 0.8944972157478333,
+      "step": 1424
+    },
+    {
+      "epoch": 0.31044342577043527,
+      "grad_norm": 0.6015625,
+      "learning_rate": 7.00375061590326e-05,
+      "loss": 1.189338207244873,
+      "step": 1425
+    },
+    {
+      "epoch": 0.31066128080606364,
+      "grad_norm": 0.625,
+      "learning_rate": 7.002268534267499e-05,
+      "loss": 1.2029749155044556,
+      "step": 1426
+    },
+    {
+      "epoch": 0.310879135841692,
+      "grad_norm": 0.640625,
+      "learning_rate": 7.000785508096592e-05,
+      "loss": 1.2596606016159058,
+      "step": 1427
+    },
+    {
+      "epoch": 0.3110969908773204,
+      "grad_norm": 0.62109375,
+      "learning_rate": 6.999301537857109e-05,
+      "loss": 1.052743673324585,
+      "step": 1428
+    },
+    {
+      "epoch": 0.3113148459129488,
+      "grad_norm": 0.625,
+      "learning_rate": 6.997816624015916e-05,
+      "loss": 1.196608543395996,
+      "step": 1429
+    },
+    {
+      "epoch": 0.31153270094857716,
+      "grad_norm": 0.64453125,
+      "learning_rate": 6.996330767040178e-05,
+      "loss": 1.2254507541656494,
+      "step": 1430
+    },
+    {
+      "epoch": 0.31175055598420554,
+      "grad_norm": 0.625,
+      "learning_rate": 6.994843967397357e-05,
+      "loss": 1.2161465883255005,
+      "step": 1431
+    },
+    {
+      "epoch": 0.31196841101983386,
+      "grad_norm": 0.625,
+      "learning_rate": 6.99335622555521e-05,
+      "loss": 1.1883286237716675,
+      "step": 1432
+    },
+    {
+      "epoch": 0.31218626605546224,
+      "grad_norm": 0.66015625,
+      "learning_rate": 6.991867541981792e-05,
+      "loss": 1.2944600582122803,
+      "step": 1433
+    },
+    {
+      "epoch": 0.3124041210910906,
+      "grad_norm": 0.66796875,
+      "learning_rate": 6.990377917145454e-05,
+      "loss": 1.3665403127670288,
+      "step": 1434
+    },
+    {
+      "epoch": 0.312621976126719,
+      "grad_norm": 0.64453125,
+      "learning_rate": 6.98888735151484e-05,
+      "loss": 1.1744788885116577,
+      "step": 1435
+    },
+    {
+      "epoch": 0.31283983116234737,
+      "grad_norm": 0.7109375,
+      "learning_rate": 6.987395845558896e-05,
+      "loss": 1.3728575706481934,
+      "step": 1436
+    },
+    {
+      "epoch": 0.31305768619797575,
+      "grad_norm": 0.671875,
+      "learning_rate": 6.985903399746857e-05,
+      "loss": 1.3145207166671753,
+      "step": 1437
+    },
+    {
+      "epoch": 0.3132755412336041,
+      "grad_norm": 0.640625,
+      "learning_rate": 6.98441001454826e-05,
+      "loss": 1.3044538497924805,
+      "step": 1438
+    },
+    {
+      "epoch": 0.3134933962692325,
+      "grad_norm": 0.65234375,
+      "learning_rate": 6.982915690432933e-05,
+      "loss": 1.1244580745697021,
+      "step": 1439
+    },
+    {
+      "epoch": 0.3137112513048609,
+      "grad_norm": 0.66796875,
+      "learning_rate": 6.981420427871002e-05,
+      "loss": 1.3189500570297241,
+      "step": 1440
+    },
+    {
+      "epoch": 0.31392910634048926,
+      "grad_norm": 0.60546875,
+      "learning_rate": 6.979924227332887e-05,
+      "loss": 1.132102608680725,
+      "step": 1441
+    },
+    {
+      "epoch": 0.31414696137611764,
+      "grad_norm": 0.6640625,
+      "learning_rate": 6.978427089289303e-05,
+      "loss": 1.2660733461380005,
+      "step": 1442
+    },
+    {
+      "epoch": 0.314364816411746,
+      "grad_norm": 0.65234375,
+      "learning_rate": 6.976929014211261e-05,
+      "loss": 1.1336172819137573,
+      "step": 1443
+    },
+    {
+      "epoch": 0.3145826714473744,
+      "grad_norm": 0.64453125,
+      "learning_rate": 6.975430002570065e-05,
+      "loss": 1.1576817035675049,
+      "step": 1444
+    },
+    {
+      "epoch": 0.3148005264830028,
+      "grad_norm": 0.72265625,
+      "learning_rate": 6.973930054837315e-05,
+      "loss": 1.226209044456482,
+      "step": 1445
+    },
+    {
+      "epoch": 0.31501838151863115,
+      "grad_norm": 0.640625,
+      "learning_rate": 6.972429171484907e-05,
+      "loss": 1.2378886938095093,
+      "step": 1446
+    },
+    {
+      "epoch": 0.31523623655425953,
+      "grad_norm": 0.60546875,
+      "learning_rate": 6.970927352985028e-05,
+      "loss": 1.0881059169769287,
+      "step": 1447
+    },
+    {
+      "epoch": 0.3154540915898879,
+      "grad_norm": 0.6640625,
+      "learning_rate": 6.969424599810161e-05,
+      "loss": 1.0765888690948486,
+      "step": 1448
+    },
+    {
+      "epoch": 0.3156719466255163,
+      "grad_norm": 0.6640625,
+      "learning_rate": 6.967920912433084e-05,
+      "loss": 1.1700553894042969,
+      "step": 1449
+    },
+    {
+      "epoch": 0.31588980166114466,
+      "grad_norm": 0.6640625,
+      "learning_rate": 6.966416291326867e-05,
+      "loss": 1.2773274183273315,
+      "step": 1450
+    },
+    {
+      "epoch": 0.31610765669677304,
+      "grad_norm": 0.66796875,
+      "learning_rate": 6.964910736964873e-05,
+      "loss": 1.1937968730926514,
+      "step": 1451
+    },
+    {
+      "epoch": 0.3163255117324014,
+      "grad_norm": 0.640625,
+      "learning_rate": 6.963404249820763e-05,
+      "loss": 1.0765032768249512,
+      "step": 1452
+    },
+    {
+      "epoch": 0.3165433667680298,
+      "grad_norm": 0.671875,
+      "learning_rate": 6.961896830368487e-05,
+      "loss": 1.103598952293396,
+      "step": 1453
+    },
+    {
+      "epoch": 0.3167612218036581,
+      "grad_norm": 0.67578125,
+      "learning_rate": 6.960388479082291e-05,
+      "loss": 1.286558747291565,
+      "step": 1454
+    },
+    {
+      "epoch": 0.3169790768392865,
+      "grad_norm": 0.578125,
+      "learning_rate": 6.958879196436712e-05,
+      "loss": 0.923491358757019,
+      "step": 1455
+    },
+    {
+      "epoch": 0.3171969318749149,
+      "grad_norm": 0.625,
+      "learning_rate": 6.957368982906581e-05,
+      "loss": 1.015873908996582,
+      "step": 1456
+    },
+    {
+      "epoch": 0.31741478691054326,
+      "grad_norm": 0.6328125,
+      "learning_rate": 6.955857838967022e-05,
+      "loss": 1.0527268648147583,
+      "step": 1457
+    },
+    {
+      "epoch": 0.31763264194617163,
+      "grad_norm": 0.65234375,
+      "learning_rate": 6.954345765093453e-05,
+      "loss": 1.2369741201400757,
+      "step": 1458
+    },
+    {
+      "epoch": 0.3178504969818,
+      "grad_norm": 0.65234375,
+      "learning_rate": 6.952832761761582e-05,
+      "loss": 1.17364501953125,
+      "step": 1459
+    },
+    {
+      "epoch": 0.3180683520174284,
+      "grad_norm": 0.6328125,
+      "learning_rate": 6.951318829447411e-05,
+      "loss": 1.2428830862045288,
+      "step": 1460
+    },
+    {
+      "epoch": 0.31828620705305677,
+      "grad_norm": 0.60546875,
+      "learning_rate": 6.949803968627234e-05,
+      "loss": 1.1000831127166748,
+      "step": 1461
+    },
+    {
+      "epoch": 0.31850406208868515,
+      "grad_norm": 0.65234375,
+      "learning_rate": 6.948288179777638e-05,
+      "loss": 1.2322677373886108,
+      "step": 1462
+    },
+    {
+      "epoch": 0.3187219171243135,
+      "grad_norm": 0.6171875,
+      "learning_rate": 6.946771463375498e-05,
+      "loss": 1.0844446420669556,
+      "step": 1463
+    },
+    {
+      "epoch": 0.3189397721599419,
+      "grad_norm": 0.62890625,
+      "learning_rate": 6.945253819897987e-05,
+      "loss": 1.0390617847442627,
+      "step": 1464
+    },
+    {
+      "epoch": 0.3191576271955703,
+      "grad_norm": 0.6328125,
+      "learning_rate": 6.943735249822567e-05,
+      "loss": 1.110557198524475,
+      "step": 1465
+    },
+    {
+      "epoch": 0.31937548223119866,
+      "grad_norm": 0.61328125,
+      "learning_rate": 6.942215753626988e-05,
+      "loss": 1.046217679977417,
+      "step": 1466
+    },
+    {
+      "epoch": 0.31959333726682704,
+      "grad_norm": 0.66015625,
+      "learning_rate": 6.940695331789295e-05,
+      "loss": 1.3159911632537842,
+      "step": 1467
+    },
+    {
+      "epoch": 0.3198111923024554,
+      "grad_norm": 0.59765625,
+      "learning_rate": 6.939173984787825e-05,
+      "loss": 1.013457179069519,
+      "step": 1468
+    },
+    {
+      "epoch": 0.3200290473380838,
+      "grad_norm": 0.61328125,
+      "learning_rate": 6.937651713101204e-05,
+      "loss": 1.0636940002441406,
+      "step": 1469
+    },
+    {
+      "epoch": 0.3202469023737122,
+      "grad_norm": 0.6484375,
+      "learning_rate": 6.936128517208349e-05,
+      "loss": 1.2765603065490723,
+      "step": 1470
+    },
+    {
+      "epoch": 0.32046475740934055,
+      "grad_norm": 0.6875,
+      "learning_rate": 6.934604397588468e-05,
+      "loss": 1.2509387731552124,
+      "step": 1471
+    },
+    {
+      "epoch": 0.32068261244496893,
+      "grad_norm": 0.62109375,
+      "learning_rate": 6.933079354721064e-05,
+      "loss": 1.1808905601501465,
+      "step": 1472
+    },
+    {
+      "epoch": 0.3209004674805973,
+      "grad_norm": 0.671875,
+      "learning_rate": 6.93155338908592e-05,
+      "loss": 1.2029526233673096,
+      "step": 1473
+    },
+    {
+      "epoch": 0.3211183225162257,
+      "grad_norm": 0.65625,
+      "learning_rate": 6.930026501163119e-05,
+      "loss": 1.2852375507354736,
+      "step": 1474
+    },
+    {
+      "epoch": 0.32133617755185406,
+      "grad_norm": 0.61328125,
+      "learning_rate": 6.928498691433031e-05,
+      "loss": 1.0723515748977661,
+      "step": 1475
+    },
+    {
+      "epoch": 0.3215540325874824,
+      "grad_norm": 0.6484375,
+      "learning_rate": 6.926969960376318e-05,
+      "loss": 1.080534815788269,
+      "step": 1476
+    },
+    {
+      "epoch": 0.32177188762311076,
+      "grad_norm": 0.66015625,
+      "learning_rate": 6.925440308473925e-05,
+      "loss": 1.0148671865463257,
+      "step": 1477
+    },
+    {
+      "epoch": 0.32198974265873914,
+      "grad_norm": 0.65625,
+      "learning_rate": 6.923909736207094e-05,
+      "loss": 1.1945472955703735,
+      "step": 1478
+    },
+    {
+      "epoch": 0.3222075976943675,
+      "grad_norm": 0.64453125,
+      "learning_rate": 6.922378244057354e-05,
+      "loss": 1.1598718166351318,
+      "step": 1479
+    },
+    {
+      "epoch": 0.3224254527299959,
+      "grad_norm": 0.72265625,
+      "learning_rate": 6.920845832506523e-05,
+      "loss": 1.263459324836731,
+      "step": 1480
+    },
+    {
+      "epoch": 0.3226433077656243,
+      "grad_norm": 0.7109375,
+      "learning_rate": 6.919312502036708e-05,
+      "loss": 1.129848599433899,
+      "step": 1481
+    },
+    {
+      "epoch": 0.32286116280125265,
+      "grad_norm": 0.6796875,
+      "learning_rate": 6.917778253130307e-05,
+      "loss": 1.260462760925293,
+      "step": 1482
+    },
+    {
+      "epoch": 0.32307901783688103,
+      "grad_norm": 0.70703125,
+      "learning_rate": 6.916243086270005e-05,
+      "loss": 1.2294589281082153,
+      "step": 1483
+    },
+    {
+      "epoch": 0.3232968728725094,
+      "grad_norm": 0.6484375,
+      "learning_rate": 6.914707001938777e-05,
+      "loss": 1.0783168077468872,
+      "step": 1484
+    },
+    {
+      "epoch": 0.3235147279081378,
+      "grad_norm": 0.65625,
+      "learning_rate": 6.913170000619887e-05,
+      "loss": 1.3684730529785156,
+      "step": 1485
+    },
+    {
+      "epoch": 0.32373258294376617,
+      "grad_norm": 0.640625,
+      "learning_rate": 6.911632082796881e-05,
+      "loss": 1.289169192314148,
+      "step": 1486
+    },
+    {
+      "epoch": 0.32395043797939455,
+      "grad_norm": 0.6875,
+      "learning_rate": 6.910093248953607e-05,
+      "loss": 1.1558043956756592,
+      "step": 1487
+    },
+    {
+      "epoch": 0.3241682930150229,
+      "grad_norm": 0.7734375,
+      "learning_rate": 6.908553499574189e-05,
+      "loss": 1.1722694635391235,
+      "step": 1488
+    },
+    {
+      "epoch": 0.3243861480506513,
+      "grad_norm": 0.63671875,
+      "learning_rate": 6.907012835143042e-05,
+      "loss": 0.9722030758857727,
+      "step": 1489
+    },
+    {
+      "epoch": 0.3246040030862797,
+      "grad_norm": 0.765625,
+      "learning_rate": 6.905471256144874e-05,
+      "loss": 1.1439220905303955,
+      "step": 1490
+    },
+    {
+      "epoch": 0.32482185812190806,
+      "grad_norm": 0.6953125,
+      "learning_rate": 6.903928763064672e-05,
+      "loss": 1.2399977445602417,
+      "step": 1491
+    },
+    {
+      "epoch": 0.32503971315753644,
+      "grad_norm": 0.58203125,
+      "learning_rate": 6.902385356387719e-05,
+      "loss": 1.0607490539550781,
+      "step": 1492
+    },
+    {
+      "epoch": 0.3252575681931648,
+      "grad_norm": 0.62890625,
+      "learning_rate": 6.90084103659958e-05,
+      "loss": 1.1313014030456543,
+      "step": 1493
+    },
+    {
+      "epoch": 0.3254754232287932,
+      "grad_norm": 0.63671875,
+      "learning_rate": 6.899295804186109e-05,
+      "loss": 1.0860681533813477,
+      "step": 1494
+    },
+    {
+      "epoch": 0.32569327826442157,
+      "grad_norm": 0.6640625,
+      "learning_rate": 6.897749659633448e-05,
+      "loss": 1.2622904777526855,
+      "step": 1495
+    },
+    {
+      "epoch": 0.32591113330004995,
+      "grad_norm": 0.6484375,
+      "learning_rate": 6.896202603428025e-05,
+      "loss": 1.1454192399978638,
+      "step": 1496
+    },
+    {
+      "epoch": 0.3261289883356783,
+      "grad_norm": 0.6640625,
+      "learning_rate": 6.894654636056554e-05,
+      "loss": 1.2351009845733643,
+      "step": 1497
+    },
+    {
+      "epoch": 0.32634684337130665,
+      "grad_norm": 0.625,
+      "learning_rate": 6.89310575800604e-05,
+      "loss": 1.2302058935165405,
+      "step": 1498
+    },
+    {
+      "epoch": 0.32656469840693503,
+      "grad_norm": 0.6328125,
+      "learning_rate": 6.891555969763766e-05,
+      "loss": 1.1019575595855713,
+      "step": 1499
+    },
+    {
+      "epoch": 0.3267825534425634,
+      "grad_norm": 0.6875,
+      "learning_rate": 6.89000527181731e-05,
+      "loss": 1.2814207077026367,
+      "step": 1500
+    },
+    {
+      "epoch": 0.3270004084781918,
+      "grad_norm": 0.62890625,
+      "learning_rate": 6.888453664654532e-05,
+      "loss": 1.0887812376022339,
+      "step": 1501
+    },
+    {
+      "epoch": 0.32721826351382016,
+      "grad_norm": 0.6171875,
+      "learning_rate": 6.886901148763577e-05,
+      "loss": 1.1240389347076416,
+      "step": 1502
+    },
+    {
+      "epoch": 0.32743611854944854,
+      "grad_norm": 0.6171875,
+      "learning_rate": 6.88534772463288e-05,
+      "loss": 1.0746996402740479,
+      "step": 1503
+    },
+    {
+      "epoch": 0.3276539735850769,
+      "grad_norm": 0.640625,
+      "learning_rate": 6.883793392751157e-05,
+      "loss": 1.157960057258606,
+      "step": 1504
+    },
+    {
+      "epoch": 0.3278718286207053,
+      "grad_norm": 0.609375,
+      "learning_rate": 6.882238153607415e-05,
+      "loss": 1.220470905303955,
+      "step": 1505
+    },
+    {
+      "epoch": 0.3280896836563337,
+      "grad_norm": 0.64453125,
+      "learning_rate": 6.880682007690941e-05,
+      "loss": 1.2210396528244019,
+      "step": 1506
+    },
+    {
+      "epoch": 0.32830753869196205,
+      "grad_norm": 0.63671875,
+      "learning_rate": 6.879124955491308e-05,
+      "loss": 1.1004934310913086,
+      "step": 1507
+    },
+    {
+      "epoch": 0.32852539372759043,
+      "grad_norm": 0.6484375,
+      "learning_rate": 6.877566997498381e-05,
+      "loss": 1.280647873878479,
+      "step": 1508
+    },
+    {
+      "epoch": 0.3287432487632188,
+      "grad_norm": 0.63671875,
+      "learning_rate": 6.876008134202299e-05,
+      "loss": 1.313375473022461,
+      "step": 1509
+    },
+    {
+      "epoch": 0.3289611037988472,
+      "grad_norm": 0.609375,
+      "learning_rate": 6.874448366093497e-05,
+      "loss": 1.2042105197906494,
+      "step": 1510
+    },
+    {
+      "epoch": 0.32917895883447557,
+      "grad_norm": 0.6171875,
+      "learning_rate": 6.872887693662684e-05,
+      "loss": 1.157981276512146,
+      "step": 1511
+    },
+    {
+      "epoch": 0.32939681387010394,
+      "grad_norm": 0.56640625,
+      "learning_rate": 6.871326117400861e-05,
+      "loss": 0.9306968450546265,
+      "step": 1512
+    },
+    {
+      "epoch": 0.3296146689057323,
+      "grad_norm": 0.59765625,
+      "learning_rate": 6.869763637799312e-05,
+      "loss": 0.9734666347503662,
+      "step": 1513
+    },
+    {
+      "epoch": 0.3298325239413607,
+      "grad_norm": 0.65234375,
+      "learning_rate": 6.868200255349605e-05,
+      "loss": 1.236253023147583,
+      "step": 1514
+    },
+    {
+      "epoch": 0.3300503789769891,
+      "grad_norm": 0.6015625,
+      "learning_rate": 6.866635970543588e-05,
+      "loss": 1.1675828695297241,
+      "step": 1515
+    },
+    {
+      "epoch": 0.33026823401261746,
+      "grad_norm": 0.6328125,
+      "learning_rate": 6.865070783873399e-05,
+      "loss": 1.2319984436035156,
+      "step": 1516
+    },
+    {
+      "epoch": 0.33048608904824583,
+      "grad_norm": 0.6796875,
+      "learning_rate": 6.863504695831454e-05,
+      "loss": 1.2070317268371582,
+      "step": 1517
+    },
+    {
+      "epoch": 0.3307039440838742,
+      "grad_norm": 0.6484375,
+      "learning_rate": 6.861937706910457e-05,
+      "loss": 1.210852861404419,
+      "step": 1518
+    },
+    {
+      "epoch": 0.3309217991195026,
+      "grad_norm": 0.640625,
+      "learning_rate": 6.860369817603396e-05,
+      "loss": 1.3843129873275757,
+      "step": 1519
+    },
+    {
+      "epoch": 0.3311396541551309,
+      "grad_norm": 0.6171875,
+      "learning_rate": 6.858801028403537e-05,
+      "loss": 1.188402533531189,
+      "step": 1520
+    },
+    {
+      "epoch": 0.3313575091907593,
+      "grad_norm": 0.60546875,
+      "learning_rate": 6.857231339804434e-05,
+      "loss": 1.1312038898468018,
+      "step": 1521
+    },
+    {
+      "epoch": 0.33157536422638767,
+      "grad_norm": 0.625,
+      "learning_rate": 6.855660752299923e-05,
+      "loss": 1.2426540851593018,
+      "step": 1522
+    },
+    {
+      "epoch": 0.33179321926201605,
+      "grad_norm": 0.64453125,
+      "learning_rate": 6.85408926638412e-05,
+      "loss": 1.2114938497543335,
+      "step": 1523
+    },
+    {
+      "epoch": 0.3320110742976444,
+      "grad_norm": 0.66015625,
+      "learning_rate": 6.852516882551428e-05,
+      "loss": 1.1735560894012451,
+      "step": 1524
+    },
+    {
+      "epoch": 0.3322289293332728,
+      "grad_norm": 0.6328125,
+      "learning_rate": 6.850943601296529e-05,
+      "loss": 1.1294833421707153,
+      "step": 1525
+    },
+    {
+      "epoch": 0.3324467843689012,
+      "grad_norm": 0.65234375,
+      "learning_rate": 6.849369423114387e-05,
+      "loss": 1.1633055210113525,
+      "step": 1526
+    },
+    {
+      "epoch": 0.33266463940452956,
+      "grad_norm": 0.6953125,
+      "learning_rate": 6.847794348500252e-05,
+      "loss": 1.1836817264556885,
+      "step": 1527
+    },
+    {
+      "epoch": 0.33288249444015794,
+      "grad_norm": 0.6328125,
+      "learning_rate": 6.846218377949653e-05,
+      "loss": 1.144059181213379,
+      "step": 1528
+    },
+    {
+      "epoch": 0.3331003494757863,
+      "grad_norm": 0.6328125,
+      "learning_rate": 6.844641511958401e-05,
+      "loss": 1.1378378868103027,
+      "step": 1529
+    },
+    {
+      "epoch": 0.3333182045114147,
+      "grad_norm": 0.640625,
+      "learning_rate": 6.843063751022589e-05,
+      "loss": 1.1751389503479004,
+      "step": 1530
+    },
+    {
+      "epoch": 0.3335360595470431,
+      "grad_norm": 0.6484375,
+      "learning_rate": 6.841485095638595e-05,
+      "loss": 1.298961877822876,
+      "step": 1531
+    },
+    {
+      "epoch": 0.33375391458267145,
+      "grad_norm": 0.61328125,
+      "learning_rate": 6.839905546303071e-05,
+      "loss": 1.1565808057785034,
+      "step": 1532
+    },
+    {
+      "epoch": 0.33397176961829983,
+      "grad_norm": 0.59765625,
+      "learning_rate": 6.838325103512957e-05,
+      "loss": 1.0357097387313843,
+      "step": 1533
+    },
+    {
+      "epoch": 0.3341896246539282,
+      "grad_norm": 0.640625,
+      "learning_rate": 6.836743767765472e-05,
+      "loss": 1.2824678421020508,
+      "step": 1534
+    },
+    {
+      "epoch": 0.3344074796895566,
+      "grad_norm": 0.6328125,
+      "learning_rate": 6.835161539558114e-05,
+      "loss": 1.10028874874115,
+      "step": 1535
+    },
+    {
+      "epoch": 0.33462533472518496,
+      "grad_norm": 0.625,
+      "learning_rate": 6.833578419388663e-05,
+      "loss": 1.1438976526260376,
+      "step": 1536
+    },
+    {
+      "epoch": 0.33484318976081334,
+      "grad_norm": 0.61328125,
+      "learning_rate": 6.83199440775518e-05,
+      "loss": 1.1222081184387207,
+      "step": 1537
+    },
+    {
+      "epoch": 0.3350610447964417,
+      "grad_norm": 0.58203125,
+      "learning_rate": 6.830409505156009e-05,
+      "loss": 1.096771478652954,
+      "step": 1538
+    },
+    {
+      "epoch": 0.3352788998320701,
+      "grad_norm": 0.6640625,
+      "learning_rate": 6.828823712089771e-05,
+      "loss": 1.1540428400039673,
+      "step": 1539
+    },
+    {
+      "epoch": 0.3354967548676985,
+      "grad_norm": 0.6484375,
+      "learning_rate": 6.827237029055364e-05,
+      "loss": 1.2097247838974,
+      "step": 1540
+    },
+    {
+      "epoch": 0.33571460990332685,
+      "grad_norm": 0.6171875,
+      "learning_rate": 6.825649456551975e-05,
+      "loss": 1.076340913772583,
+      "step": 1541
+    },
+    {
+      "epoch": 0.3359324649389552,
+      "grad_norm": 0.61328125,
+      "learning_rate": 6.824060995079062e-05,
+      "loss": 1.0958021879196167,
+      "step": 1542
+    },
+    {
+      "epoch": 0.33615031997458356,
+      "grad_norm": 0.60546875,
+      "learning_rate": 6.822471645136368e-05,
+      "loss": 1.0136560201644897,
+      "step": 1543
+    },
+    {
+      "epoch": 0.33636817501021193,
+      "grad_norm": 0.6484375,
+      "learning_rate": 6.820881407223914e-05,
+      "loss": 1.352805495262146,
+      "step": 1544
+    },
+    {
+      "epoch": 0.3365860300458403,
+      "grad_norm": 0.578125,
+      "learning_rate": 6.819290281842e-05,
+      "loss": 1.046970248222351,
+      "step": 1545
+    },
+    {
+      "epoch": 0.3368038850814687,
+      "grad_norm": 0.6328125,
+      "learning_rate": 6.817698269491204e-05,
+      "loss": 1.1989794969558716,
+      "step": 1546
+    },
+    {
+      "epoch": 0.33702174011709707,
+      "grad_norm": 0.6484375,
+      "learning_rate": 6.816105370672384e-05,
+      "loss": 1.207699179649353,
+      "step": 1547
+    },
+    {
+      "epoch": 0.33723959515272545,
+      "grad_norm": 0.62890625,
+      "learning_rate": 6.814511585886682e-05,
+      "loss": 1.1781643629074097,
+      "step": 1548
+    },
+    {
+      "epoch": 0.3374574501883538,
+      "grad_norm": 0.62890625,
+      "learning_rate": 6.812916915635509e-05,
+      "loss": 1.1199208498001099,
+      "step": 1549
+    },
+    {
+      "epoch": 0.3376753052239822,
+      "grad_norm": 0.62890625,
+      "learning_rate": 6.811321360420562e-05,
+      "loss": 1.1363637447357178,
+      "step": 1550
+    },
+    {
+      "epoch": 0.3378931602596106,
+      "grad_norm": 0.6171875,
+      "learning_rate": 6.809724920743813e-05,
+      "loss": 1.037628173828125,
+      "step": 1551
+    },
+    {
+      "epoch": 0.33811101529523896,
+      "grad_norm": 0.62890625,
+      "learning_rate": 6.808127597107516e-05,
+      "loss": 1.0938949584960938,
+      "step": 1552
+    },
+    {
+      "epoch": 0.33832887033086734,
+      "grad_norm": 0.6484375,
+      "learning_rate": 6.806529390014198e-05,
+      "loss": 1.1392885446548462,
+      "step": 1553
+    },
+    {
+      "epoch": 0.3385467253664957,
+      "grad_norm": 0.65625,
+      "learning_rate": 6.804930299966666e-05,
+      "loss": 1.2082873582839966,
+      "step": 1554
+    },
+    {
+      "epoch": 0.3387645804021241,
+      "grad_norm": 0.62109375,
+      "learning_rate": 6.803330327468006e-05,
+      "loss": 1.195794701576233,
+      "step": 1555
+    },
+    {
+      "epoch": 0.33898243543775247,
+      "grad_norm": 0.578125,
+      "learning_rate": 6.801729473021581e-05,
+      "loss": 0.9998102188110352,
+      "step": 1556
+    },
+    {
+      "epoch": 0.33920029047338085,
+      "grad_norm": 0.61328125,
+      "learning_rate": 6.800127737131032e-05,
+      "loss": 1.169172763824463,
+      "step": 1557
+    },
+    {
+      "epoch": 0.33941814550900923,
+      "grad_norm": 0.61328125,
+      "learning_rate": 6.798525120300274e-05,
+      "loss": 1.133528709411621,
+      "step": 1558
+    },
+    {
+      "epoch": 0.3396360005446376,
+      "grad_norm": 0.62890625,
+      "learning_rate": 6.796921623033506e-05,
+      "loss": 1.0910875797271729,
+      "step": 1559
+    },
+    {
+      "epoch": 0.339853855580266,
+      "grad_norm": 0.67578125,
+      "learning_rate": 6.795317245835195e-05,
+      "loss": 1.0966581106185913,
+      "step": 1560
+    },
+    {
+      "epoch": 0.34007171061589436,
+      "grad_norm": 0.64453125,
+      "learning_rate": 6.793711989210092e-05,
+      "loss": 1.2105357646942139,
+      "step": 1561
+    },
+    {
+      "epoch": 0.34028956565152274,
+      "grad_norm": 0.60546875,
+      "learning_rate": 6.792105853663222e-05,
+      "loss": 1.0998790264129639,
+      "step": 1562
+    },
+    {
+      "epoch": 0.3405074206871511,
+      "grad_norm": 0.66796875,
+      "learning_rate": 6.790498839699887e-05,
+      "loss": 1.1712121963500977,
+      "step": 1563
+    },
+    {
+      "epoch": 0.34072527572277944,
+      "grad_norm": 0.625,
+      "learning_rate": 6.788890947825665e-05,
+      "loss": 1.0399796962738037,
+      "step": 1564
+    },
+    {
+      "epoch": 0.3409431307584078,
+      "grad_norm": 0.63671875,
+      "learning_rate": 6.787282178546408e-05,
+      "loss": 1.1925917863845825,
+      "step": 1565
+    },
+    {
+      "epoch": 0.3411609857940362,
+      "grad_norm": 0.66015625,
+      "learning_rate": 6.78567253236825e-05,
+      "loss": 1.2144440412521362,
+      "step": 1566
+    },
+    {
+      "epoch": 0.3413788408296646,
+      "grad_norm": 0.625,
+      "learning_rate": 6.784062009797594e-05,
+      "loss": 1.2205430269241333,
+      "step": 1567
+    },
+    {
+      "epoch": 0.34159669586529295,
+      "grad_norm": 0.62890625,
+      "learning_rate": 6.782450611341124e-05,
+      "loss": 1.124921441078186,
+      "step": 1568
+    },
+    {
+      "epoch": 0.34181455090092133,
+      "grad_norm": 0.5859375,
+      "learning_rate": 6.780838337505798e-05,
+      "loss": 0.9436802864074707,
+      "step": 1569
+    },
+    {
+      "epoch": 0.3420324059365497,
+      "grad_norm": 0.66796875,
+      "learning_rate": 6.779225188798846e-05,
+      "loss": 1.2180217504501343,
+      "step": 1570
+    },
+    {
+      "epoch": 0.3422502609721781,
+      "grad_norm": 0.671875,
+      "learning_rate": 6.777611165727778e-05,
+      "loss": 1.176059603691101,
+      "step": 1571
+    },
+    {
+      "epoch": 0.34246811600780647,
+      "grad_norm": 0.6171875,
+      "learning_rate": 6.775996268800378e-05,
+      "loss": 1.1272320747375488,
+      "step": 1572
+    },
+    {
+      "epoch": 0.34268597104343484,
+      "grad_norm": 0.6328125,
+      "learning_rate": 6.774380498524702e-05,
+      "loss": 1.0977709293365479,
+      "step": 1573
+    },
+    {
+      "epoch": 0.3429038260790632,
+      "grad_norm": 0.59375,
+      "learning_rate": 6.772763855409084e-05,
+      "loss": 1.0731773376464844,
+      "step": 1574
+    },
+    {
+      "epoch": 0.3431216811146916,
+      "grad_norm": 0.60546875,
+      "learning_rate": 6.771146339962133e-05,
+      "loss": 1.0730558633804321,
+      "step": 1575
+    },
+    {
+      "epoch": 0.34333953615032,
+      "grad_norm": 0.64453125,
+      "learning_rate": 6.769527952692729e-05,
+      "loss": 1.09489905834198,
+      "step": 1576
+    },
+    {
+      "epoch": 0.34355739118594836,
+      "grad_norm": 0.625,
+      "learning_rate": 6.767908694110027e-05,
+      "loss": 1.101222276687622,
+      "step": 1577
+    },
+    {
+      "epoch": 0.34377524622157674,
+      "grad_norm": 0.6171875,
+      "learning_rate": 6.76628856472346e-05,
+      "loss": 0.9394668340682983,
+      "step": 1578
+    },
+    {
+      "epoch": 0.3439931012572051,
+      "grad_norm": 0.63671875,
+      "learning_rate": 6.764667565042733e-05,
+      "loss": 1.2335200309753418,
+      "step": 1579
+    },
+    {
+      "epoch": 0.3442109562928335,
+      "grad_norm": 0.6484375,
+      "learning_rate": 6.76304569557782e-05,
+      "loss": 1.243628740310669,
+      "step": 1580
+    },
+    {
+      "epoch": 0.34442881132846187,
+      "grad_norm": 0.6875,
+      "learning_rate": 6.761422956838975e-05,
+      "loss": 1.1526364088058472,
+      "step": 1581
+    },
+    {
+      "epoch": 0.34464666636409025,
+      "grad_norm": 0.66796875,
+      "learning_rate": 6.759799349336724e-05,
+      "loss": 1.2424378395080566,
+      "step": 1582
+    },
+    {
+      "epoch": 0.3448645213997186,
+      "grad_norm": 0.6796875,
+      "learning_rate": 6.758174873581865e-05,
+      "loss": 1.3820230960845947,
+      "step": 1583
+    },
+    {
+      "epoch": 0.345082376435347,
+      "grad_norm": 0.65625,
+      "learning_rate": 6.75654953008547e-05,
+      "loss": 1.2843503952026367,
+      "step": 1584
+    },
+    {
+      "epoch": 0.3453002314709754,
+      "grad_norm": 0.65234375,
+      "learning_rate": 6.754923319358883e-05,
+      "loss": 1.1074090003967285,
+      "step": 1585
+    },
+    {
+      "epoch": 0.3455180865066037,
+      "grad_norm": 0.63671875,
+      "learning_rate": 6.75329624191372e-05,
+      "loss": 1.3198845386505127,
+      "step": 1586
+    },
+    {
+      "epoch": 0.3457359415422321,
+      "grad_norm": 0.61328125,
+      "learning_rate": 6.751668298261875e-05,
+      "loss": 1.1334837675094604,
+      "step": 1587
+    },
+    {
+      "epoch": 0.34595379657786046,
+      "grad_norm": 0.66015625,
+      "learning_rate": 6.750039488915508e-05,
+      "loss": 1.1654767990112305,
+      "step": 1588
+    },
+    {
+      "epoch": 0.34617165161348884,
+      "grad_norm": 0.6484375,
+      "learning_rate": 6.748409814387055e-05,
+      "loss": 1.28505277633667,
+      "step": 1589
+    },
+    {
+      "epoch": 0.3463895066491172,
+      "grad_norm": 0.6171875,
+      "learning_rate": 6.746779275189222e-05,
+      "loss": 1.184356689453125,
+      "step": 1590
+    },
+    {
+      "epoch": 0.3466073616847456,
+      "grad_norm": 0.61328125,
+      "learning_rate": 6.74514787183499e-05,
+      "loss": 1.0435212850570679,
+      "step": 1591
+    },
+    {
+      "epoch": 0.346825216720374,
+      "grad_norm": 0.69140625,
+      "learning_rate": 6.743515604837608e-05,
+      "loss": 1.3667446374893188,
+      "step": 1592
+    },
+    {
+      "epoch": 0.34704307175600235,
+      "grad_norm": 0.6328125,
+      "learning_rate": 6.741882474710603e-05,
+      "loss": 1.0531940460205078,
+      "step": 1593
+    },
+    {
+      "epoch": 0.34726092679163073,
+      "grad_norm": 0.65625,
+      "learning_rate": 6.740248481967764e-05,
+      "loss": 1.182543158531189,
+      "step": 1594
+    },
+    {
+      "epoch": 0.3474787818272591,
+      "grad_norm": 0.62890625,
+      "learning_rate": 6.73861362712316e-05,
+      "loss": 1.172400951385498,
+      "step": 1595
+    },
+    {
+      "epoch": 0.3476966368628875,
+      "grad_norm": 0.6484375,
+      "learning_rate": 6.736977910691128e-05,
+      "loss": 1.242552399635315,
+      "step": 1596
+    },
+    {
+      "epoch": 0.34791449189851587,
+      "grad_norm": 0.6640625,
+      "learning_rate": 6.735341333186276e-05,
+      "loss": 1.1912729740142822,
+      "step": 1597
+    },
+    {
+      "epoch": 0.34813234693414424,
+      "grad_norm": 0.65234375,
+      "learning_rate": 6.733703895123483e-05,
+      "loss": 1.1364104747772217,
+      "step": 1598
+    },
+    {
+      "epoch": 0.3483502019697726,
+      "grad_norm": 0.6328125,
+      "learning_rate": 6.732065597017897e-05,
+      "loss": 1.059836506843567,
+      "step": 1599
+    },
+    {
+      "epoch": 0.348568057005401,
+      "grad_norm": 0.65625,
+      "learning_rate": 6.730426439384939e-05,
+      "loss": 1.0784817934036255,
+      "step": 1600
+    },
+    {
+      "epoch": 0.3487859120410294,
+      "grad_norm": 0.703125,
+      "learning_rate": 6.728786422740301e-05,
+      "loss": 1.2877088785171509,
+      "step": 1601
+    },
+    {
+      "epoch": 0.34900376707665776,
+      "grad_norm": 0.62109375,
+      "learning_rate": 6.727145547599945e-05,
+      "loss": 1.079934000968933,
+      "step": 1602
+    },
+    {
+      "epoch": 0.34922162211228613,
+      "grad_norm": 0.61328125,
+      "learning_rate": 6.725503814480099e-05,
+      "loss": 1.1356475353240967,
+      "step": 1603
+    },
+    {
+      "epoch": 0.3494394771479145,
+      "grad_norm": 0.6015625,
+      "learning_rate": 6.723861223897265e-05,
+      "loss": 1.0111593008041382,
+      "step": 1604
+    },
+    {
+      "epoch": 0.3496573321835429,
+      "grad_norm": 0.6328125,
+      "learning_rate": 6.722217776368214e-05,
+      "loss": 1.1337964534759521,
+      "step": 1605
+    },
+    {
+      "epoch": 0.34987518721917127,
+      "grad_norm": 0.6640625,
+      "learning_rate": 6.720573472409987e-05,
+      "loss": 1.2811614274978638,
+      "step": 1606
+    },
+    {
+      "epoch": 0.3500930422547996,
+      "grad_norm": 0.6328125,
+      "learning_rate": 6.718928312539894e-05,
+      "loss": 1.3193395137786865,
+      "step": 1607
+    },
+    {
+      "epoch": 0.35031089729042797,
+      "grad_norm": 0.62890625,
+      "learning_rate": 6.71728229727551e-05,
+      "loss": 1.1722017526626587,
+      "step": 1608
+    },
+    {
+      "epoch": 0.35052875232605635,
+      "grad_norm": 0.6796875,
+      "learning_rate": 6.715635427134688e-05,
+      "loss": 1.2832038402557373,
+      "step": 1609
+    },
+    {
+      "epoch": 0.3507466073616847,
+      "grad_norm": 0.66796875,
+      "learning_rate": 6.713987702635544e-05,
+      "loss": 1.2275563478469849,
+      "step": 1610
+    },
+    {
+      "epoch": 0.3509644623973131,
+      "grad_norm": 0.64453125,
+      "learning_rate": 6.712339124296465e-05,
+      "loss": 1.2348490953445435,
+      "step": 1611
+    },
+    {
+      "epoch": 0.3511823174329415,
+      "grad_norm": 0.63671875,
+      "learning_rate": 6.710689692636101e-05,
+      "loss": 1.2363656759262085,
+      "step": 1612
+    },
+    {
+      "epoch": 0.35140017246856986,
+      "grad_norm": 0.6640625,
+      "learning_rate": 6.70903940817338e-05,
+      "loss": 1.2458643913269043,
+      "step": 1613
+    },
+    {
+      "epoch": 0.35161802750419824,
+      "grad_norm": 0.640625,
+      "learning_rate": 6.707388271427491e-05,
+      "loss": 1.1681312322616577,
+      "step": 1614
+    },
+    {
+      "epoch": 0.3518358825398266,
+      "grad_norm": 0.64453125,
+      "learning_rate": 6.705736282917892e-05,
+      "loss": 1.2236526012420654,
+      "step": 1615
+    },
+    {
+      "epoch": 0.352053737575455,
+      "grad_norm": 0.6015625,
+      "learning_rate": 6.704083443164313e-05,
+      "loss": 1.1403931379318237,
+      "step": 1616
+    },
+    {
+      "epoch": 0.3522715926110834,
+      "grad_norm": 0.6328125,
+      "learning_rate": 6.70242975268675e-05,
+      "loss": 1.2157894372940063,
+      "step": 1617
+    },
+    {
+      "epoch": 0.35248944764671175,
+      "grad_norm": 0.6328125,
+      "learning_rate": 6.700775212005462e-05,
+      "loss": 1.169590950012207,
+      "step": 1618
+    },
+    {
+      "epoch": 0.35270730268234013,
+      "grad_norm": 0.62109375,
+      "learning_rate": 6.699119821640981e-05,
+      "loss": 1.0481679439544678,
+      "step": 1619
+    },
+    {
+      "epoch": 0.3529251577179685,
+      "grad_norm": 0.63671875,
+      "learning_rate": 6.697463582114107e-05,
+      "loss": 1.205012321472168,
+      "step": 1620
+    },
+    {
+      "epoch": 0.3531430127535969,
+      "grad_norm": 0.63671875,
+      "learning_rate": 6.695806493945902e-05,
+      "loss": 1.1161001920700073,
+      "step": 1621
+    },
+    {
+      "epoch": 0.35336086778922526,
+      "grad_norm": 0.65625,
+      "learning_rate": 6.694148557657699e-05,
+      "loss": 1.20137619972229,
+      "step": 1622
+    },
+    {
+      "epoch": 0.35357872282485364,
+      "grad_norm": 0.640625,
+      "learning_rate": 6.692489773771096e-05,
+      "loss": 1.1343011856079102,
+      "step": 1623
+    },
+    {
+      "epoch": 0.353796577860482,
+      "grad_norm": 0.65234375,
+      "learning_rate": 6.690830142807958e-05,
+      "loss": 1.0812262296676636,
+      "step": 1624
+    },
+    {
+      "epoch": 0.3540144328961104,
+      "grad_norm": 0.58203125,
+      "learning_rate": 6.689169665290418e-05,
+      "loss": 1.0995793342590332,
+      "step": 1625
+    },
+    {
+      "epoch": 0.3542322879317388,
+      "grad_norm": 0.6328125,
+      "learning_rate": 6.687508341740874e-05,
+      "loss": 1.2122786045074463,
+      "step": 1626
+    },
+    {
+      "epoch": 0.35445014296736715,
+      "grad_norm": 0.62109375,
+      "learning_rate": 6.685846172681988e-05,
+      "loss": 1.0743293762207031,
+      "step": 1627
+    },
+    {
+      "epoch": 0.35466799800299553,
+      "grad_norm": 0.69140625,
+      "learning_rate": 6.684183158636694e-05,
+      "loss": 1.2049319744110107,
+      "step": 1628
+    },
+    {
+      "epoch": 0.35488585303862386,
+      "grad_norm": 0.703125,
+      "learning_rate": 6.682519300128185e-05,
+      "loss": 1.208034634590149,
+      "step": 1629
+    },
+    {
+      "epoch": 0.35510370807425223,
+      "grad_norm": 0.640625,
+      "learning_rate": 6.680854597679924e-05,
+      "loss": 1.0858913660049438,
+      "step": 1630
+    },
+    {
+      "epoch": 0.3553215631098806,
+      "grad_norm": 0.66015625,
+      "learning_rate": 6.679189051815637e-05,
+      "loss": 1.2026996612548828,
+      "step": 1631
+    },
+    {
+      "epoch": 0.355539418145509,
+      "grad_norm": 0.63671875,
+      "learning_rate": 6.67752266305932e-05,
+      "loss": 1.1412070989608765,
+      "step": 1632
+    },
+    {
+      "epoch": 0.35575727318113737,
+      "grad_norm": 0.6796875,
+      "learning_rate": 6.675855431935227e-05,
+      "loss": 1.2182236909866333,
+      "step": 1633
+    },
+    {
+      "epoch": 0.35597512821676575,
+      "grad_norm": 0.60546875,
+      "learning_rate": 6.674187358967883e-05,
+      "loss": 1.249222993850708,
+      "step": 1634
+    },
+    {
+      "epoch": 0.3561929832523941,
+      "grad_norm": 0.6640625,
+      "learning_rate": 6.672518444682075e-05,
+      "loss": 1.1467419862747192,
+      "step": 1635
+    },
+    {
+      "epoch": 0.3564108382880225,
+      "grad_norm": 0.69921875,
+      "learning_rate": 6.670848689602855e-05,
+      "loss": 1.064927101135254,
+      "step": 1636
+    },
+    {
+      "epoch": 0.3566286933236509,
+      "grad_norm": 0.66796875,
+      "learning_rate": 6.669178094255542e-05,
+      "loss": 1.2521836757659912,
+      "step": 1637
+    },
+    {
+      "epoch": 0.35684654835927926,
+      "grad_norm": 0.671875,
+      "learning_rate": 6.667506659165714e-05,
+      "loss": 1.112594723701477,
+      "step": 1638
+    },
+    {
+      "epoch": 0.35706440339490764,
+      "grad_norm": 0.65234375,
+      "learning_rate": 6.665834384859219e-05,
+      "loss": 1.0478167533874512,
+      "step": 1639
+    },
+    {
+      "epoch": 0.357282258430536,
+      "grad_norm": 0.64453125,
+      "learning_rate": 6.664161271862165e-05,
+      "loss": 1.217670202255249,
+      "step": 1640
+    },
+    {
+      "epoch": 0.3575001134661644,
+      "grad_norm": 0.69140625,
+      "learning_rate": 6.662487320700924e-05,
+      "loss": 1.2712846994400024,
+      "step": 1641
+    },
+    {
+      "epoch": 0.35771796850179277,
+      "grad_norm": 0.64453125,
+      "learning_rate": 6.660812531902136e-05,
+      "loss": 1.1131709814071655,
+      "step": 1642
+    },
+    {
+      "epoch": 0.35793582353742115,
+      "grad_norm": 0.625,
+      "learning_rate": 6.659136905992701e-05,
+      "loss": 1.054059386253357,
+      "step": 1643
+    },
+    {
+      "epoch": 0.3581536785730495,
+      "grad_norm": 0.625,
+      "learning_rate": 6.657460443499784e-05,
+      "loss": 1.0454964637756348,
+      "step": 1644
+    },
+    {
+      "epoch": 0.3583715336086779,
+      "grad_norm": 0.640625,
+      "learning_rate": 6.655783144950809e-05,
+      "loss": 1.188522458076477,
+      "step": 1645
+    },
+    {
+      "epoch": 0.3585893886443063,
+      "grad_norm": 0.734375,
+      "learning_rate": 6.654105010873465e-05,
+      "loss": 1.3010597229003906,
+      "step": 1646
+    },
+    {
+      "epoch": 0.35880724367993466,
+      "grad_norm": 0.640625,
+      "learning_rate": 6.652426041795711e-05,
+      "loss": 1.1448986530303955,
+      "step": 1647
+    },
+    {
+      "epoch": 0.35902509871556304,
+      "grad_norm": 0.64453125,
+      "learning_rate": 6.650746238245758e-05,
+      "loss": 1.1921846866607666,
+      "step": 1648
+    },
+    {
+      "epoch": 0.3592429537511914,
+      "grad_norm": 0.68359375,
+      "learning_rate": 6.649065600752083e-05,
+      "loss": 1.2058870792388916,
+      "step": 1649
+    },
+    {
+      "epoch": 0.3594608087868198,
+      "grad_norm": 0.62109375,
+      "learning_rate": 6.647384129843432e-05,
+      "loss": 1.2431720495224,
+      "step": 1650
+    },
+    {
+      "epoch": 0.3596786638224481,
+      "grad_norm": 0.640625,
+      "learning_rate": 6.645701826048803e-05,
+      "loss": 1.0887579917907715,
+      "step": 1651
+    },
+    {
+      "epoch": 0.3598965188580765,
+      "grad_norm": 0.60546875,
+      "learning_rate": 6.644018689897463e-05,
+      "loss": 0.915355920791626,
+      "step": 1652
+    },
+    {
+      "epoch": 0.3601143738937049,
+      "grad_norm": 0.6171875,
+      "learning_rate": 6.642334721918938e-05,
+      "loss": 1.1903151273727417,
+      "step": 1653
+    },
+    {
+      "epoch": 0.36033222892933325,
+      "grad_norm": 0.625,
+      "learning_rate": 6.640649922643016e-05,
+      "loss": 1.204622507095337,
+      "step": 1654
+    },
+    {
+      "epoch": 0.36055008396496163,
+      "grad_norm": 0.62109375,
+      "learning_rate": 6.638964292599748e-05,
+      "loss": 1.2136119604110718,
+      "step": 1655
+    },
+    {
+      "epoch": 0.36076793900059,
+      "grad_norm": 0.60546875,
+      "learning_rate": 6.637277832319444e-05,
+      "loss": 1.0511724948883057,
+      "step": 1656
+    },
+    {
+      "epoch": 0.3609857940362184,
+      "grad_norm": 0.640625,
+      "learning_rate": 6.635590542332678e-05,
+      "loss": 1.1930886507034302,
+      "step": 1657
+    },
+    {
+      "epoch": 0.36120364907184677,
+      "grad_norm": 0.6640625,
+      "learning_rate": 6.633902423170282e-05,
+      "loss": 1.1310018301010132,
+      "step": 1658
+    },
+    {
+      "epoch": 0.36142150410747514,
+      "grad_norm": 0.61328125,
+      "learning_rate": 6.632213475363351e-05,
+      "loss": 1.133131742477417,
+      "step": 1659
+    },
+    {
+      "epoch": 0.3616393591431035,
+      "grad_norm": 0.5703125,
+      "learning_rate": 6.63052369944324e-05,
+      "loss": 1.0036816596984863,
+      "step": 1660
+    },
+    {
+      "epoch": 0.3618572141787319,
+      "grad_norm": 0.625,
+      "learning_rate": 6.628833095941565e-05,
+      "loss": 1.2591612339019775,
+      "step": 1661
+    },
+    {
+      "epoch": 0.3620750692143603,
+      "grad_norm": 0.64453125,
+      "learning_rate": 6.6271416653902e-05,
+      "loss": 1.2765491008758545,
+      "step": 1662
+    },
+    {
+      "epoch": 0.36229292424998866,
+      "grad_norm": 0.625,
+      "learning_rate": 6.625449408321284e-05,
+      "loss": 1.2546664476394653,
+      "step": 1663
+    },
+    {
+      "epoch": 0.36251077928561704,
+      "grad_norm": 0.5625,
+      "learning_rate": 6.623756325267213e-05,
+      "loss": 0.9466208219528198,
+      "step": 1664
+    },
+    {
+      "epoch": 0.3627286343212454,
+      "grad_norm": 0.6484375,
+      "learning_rate": 6.622062416760641e-05,
+      "loss": 1.1330124139785767,
+      "step": 1665
+    },
+    {
+      "epoch": 0.3629464893568738,
+      "grad_norm": 0.61328125,
+      "learning_rate": 6.620367683334484e-05,
+      "loss": 1.0363385677337646,
+      "step": 1666
+    },
+    {
+      "epoch": 0.36316434439250217,
+      "grad_norm": 0.65625,
+      "learning_rate": 6.618672125521918e-05,
+      "loss": 1.0914332866668701,
+      "step": 1667
+    },
+    {
+      "epoch": 0.36338219942813055,
+      "grad_norm": 0.66015625,
+      "learning_rate": 6.616975743856378e-05,
+      "loss": 1.1961499452590942,
+      "step": 1668
+    },
+    {
+      "epoch": 0.3636000544637589,
+      "grad_norm": 0.6640625,
+      "learning_rate": 6.615278538871556e-05,
+      "loss": 1.2134696245193481,
+      "step": 1669
+    },
+    {
+      "epoch": 0.3638179094993873,
+      "grad_norm": 0.62109375,
+      "learning_rate": 6.613580511101407e-05,
+      "loss": 1.083092212677002,
+      "step": 1670
+    },
+    {
+      "epoch": 0.3640357645350157,
+      "grad_norm": 0.6796875,
+      "learning_rate": 6.611881661080142e-05,
+      "loss": 1.2902470827102661,
+      "step": 1671
+    },
+    {
+      "epoch": 0.36425361957064406,
+      "grad_norm": 0.68359375,
+      "learning_rate": 6.610181989342233e-05,
+      "loss": 1.3050168752670288,
+      "step": 1672
+    },
+    {
+      "epoch": 0.3644714746062724,
+      "grad_norm": 0.64453125,
+      "learning_rate": 6.608481496422403e-05,
+      "loss": 1.02991783618927,
+      "step": 1673
+    },
+    {
+      "epoch": 0.36468932964190076,
+      "grad_norm": 0.63671875,
+      "learning_rate": 6.606780182855646e-05,
+      "loss": 1.151662826538086,
+      "step": 1674
+    },
+    {
+      "epoch": 0.36490718467752914,
+      "grad_norm": 0.64453125,
+      "learning_rate": 6.605078049177205e-05,
+      "loss": 1.0999902486801147,
+      "step": 1675
+    },
+    {
+      "epoch": 0.3651250397131575,
+      "grad_norm": 0.6171875,
+      "learning_rate": 6.603375095922582e-05,
+      "loss": 1.063581943511963,
+      "step": 1676
+    },
+    {
+      "epoch": 0.3653428947487859,
+      "grad_norm": 0.6796875,
+      "learning_rate": 6.601671323627539e-05,
+      "loss": 1.20530104637146,
+      "step": 1677
+    },
+    {
+      "epoch": 0.3655607497844143,
+      "grad_norm": 0.6328125,
+      "learning_rate": 6.599966732828096e-05,
+      "loss": 1.0195565223693848,
+      "step": 1678
+    },
+    {
+      "epoch": 0.36577860482004265,
+      "grad_norm": 0.671875,
+      "learning_rate": 6.598261324060529e-05,
+      "loss": 1.154636263847351,
+      "step": 1679
+    },
+    {
+      "epoch": 0.36599645985567103,
+      "grad_norm": 0.6484375,
+      "learning_rate": 6.596555097861372e-05,
+      "loss": 1.2092653512954712,
+      "step": 1680
+    },
+    {
+      "epoch": 0.3662143148912994,
+      "grad_norm": 0.63671875,
+      "learning_rate": 6.594848054767415e-05,
+      "loss": 1.1064784526824951,
+      "step": 1681
+    },
+    {
+      "epoch": 0.3664321699269278,
+      "grad_norm": 0.62890625,
+      "learning_rate": 6.593140195315707e-05,
+      "loss": 1.1782898902893066,
+      "step": 1682
+    },
+    {
+      "epoch": 0.36665002496255616,
+      "grad_norm": 0.62890625,
+      "learning_rate": 6.591431520043551e-05,
+      "loss": 0.9711118340492249,
+      "step": 1683
+    },
+    {
+      "epoch": 0.36686787999818454,
+      "grad_norm": 0.5859375,
+      "learning_rate": 6.58972202948851e-05,
+      "loss": 0.9961894154548645,
+      "step": 1684
+    },
+    {
+      "epoch": 0.3670857350338129,
+      "grad_norm": 0.62890625,
+      "learning_rate": 6.588011724188402e-05,
+      "loss": 1.2505024671554565,
+      "step": 1685
+    },
+    {
+      "epoch": 0.3673035900694413,
+      "grad_norm": 0.640625,
+      "learning_rate": 6.586300604681301e-05,
+      "loss": 1.1594583988189697,
+      "step": 1686
+    },
+    {
+      "epoch": 0.3675214451050697,
+      "grad_norm": 0.67578125,
+      "learning_rate": 6.584588671505538e-05,
+      "loss": 1.2327247858047485,
+      "step": 1687
+    },
+    {
+      "epoch": 0.36773930014069806,
+      "grad_norm": 0.63671875,
+      "learning_rate": 6.582875925199699e-05,
+      "loss": 1.1252005100250244,
+      "step": 1688
+    },
+    {
+      "epoch": 0.36795715517632643,
+      "grad_norm": 0.6328125,
+      "learning_rate": 6.581162366302625e-05,
+      "loss": 1.1864293813705444,
+      "step": 1689
+    },
+    {
+      "epoch": 0.3681750102119548,
+      "grad_norm": 0.65234375,
+      "learning_rate": 6.579447995353416e-05,
+      "loss": 1.0536376237869263,
+      "step": 1690
+    },
+    {
+      "epoch": 0.3683928652475832,
+      "grad_norm": 0.6328125,
+      "learning_rate": 6.577732812891423e-05,
+      "loss": 1.1955292224884033,
+      "step": 1691
+    },
+    {
+      "epoch": 0.36861072028321157,
+      "grad_norm": 0.58984375,
+      "learning_rate": 6.576016819456257e-05,
+      "loss": 1.1023874282836914,
+      "step": 1692
+    },
+    {
+      "epoch": 0.36882857531883995,
+      "grad_norm": 0.609375,
+      "learning_rate": 6.57430001558778e-05,
+      "loss": 1.1011821031570435,
+      "step": 1693
+    },
+    {
+      "epoch": 0.3690464303544683,
+      "grad_norm": 0.66015625,
+      "learning_rate": 6.572582401826113e-05,
+      "loss": 1.3094773292541504,
+      "step": 1694
+    },
+    {
+      "epoch": 0.36926428539009665,
+      "grad_norm": 0.60546875,
+      "learning_rate": 6.570863978711627e-05,
+      "loss": 1.1376944780349731,
+      "step": 1695
+    },
+    {
+      "epoch": 0.369482140425725,
+      "grad_norm": 0.6328125,
+      "learning_rate": 6.569144746784951e-05,
+      "loss": 1.3316456079483032,
+      "step": 1696
+    },
+    {
+      "epoch": 0.3696999954613534,
+      "grad_norm": 0.67578125,
+      "learning_rate": 6.567424706586968e-05,
+      "loss": 1.1644320487976074,
+      "step": 1697
+    },
+    {
+      "epoch": 0.3699178504969818,
+      "grad_norm": 0.62109375,
+      "learning_rate": 6.565703858658816e-05,
+      "loss": 1.1114602088928223,
+      "step": 1698
+    },
+    {
+      "epoch": 0.37013570553261016,
+      "grad_norm": 0.63671875,
+      "learning_rate": 6.563982203541886e-05,
+      "loss": 1.010169506072998,
+      "step": 1699
+    },
+    {
+      "epoch": 0.37035356056823854,
+      "grad_norm": 0.625,
+      "learning_rate": 6.562259741777819e-05,
+      "loss": 1.1549890041351318,
+      "step": 1700
+    },
+    {
+      "epoch": 0.3705714156038669,
+      "grad_norm": 0.61328125,
+      "learning_rate": 6.560536473908519e-05,
+      "loss": 1.1141762733459473,
+      "step": 1701
+    },
+    {
+      "epoch": 0.3707892706394953,
+      "grad_norm": 0.61328125,
+      "learning_rate": 6.558812400476134e-05,
+      "loss": 1.0172152519226074,
+      "step": 1702
+    },
+    {
+      "epoch": 0.3710071256751237,
+      "grad_norm": 0.62890625,
+      "learning_rate": 6.557087522023075e-05,
+      "loss": 1.1427836418151855,
+      "step": 1703
+    },
+    {
+      "epoch": 0.37122498071075205,
+      "grad_norm": 0.6484375,
+      "learning_rate": 6.555361839091995e-05,
+      "loss": 1.211381196975708,
+      "step": 1704
+    },
+    {
+      "epoch": 0.37144283574638043,
+      "grad_norm": 0.63671875,
+      "learning_rate": 6.553635352225812e-05,
+      "loss": 1.2453097105026245,
+      "step": 1705
+    },
+    {
+      "epoch": 0.3716606907820088,
+      "grad_norm": 0.63671875,
+      "learning_rate": 6.551908061967686e-05,
+      "loss": 1.2755028009414673,
+      "step": 1706
+    },
+    {
+      "epoch": 0.3718785458176372,
+      "grad_norm": 0.61328125,
+      "learning_rate": 6.550179968861038e-05,
+      "loss": 1.098515510559082,
+      "step": 1707
+    },
+    {
+      "epoch": 0.37209640085326556,
+      "grad_norm": 0.62109375,
+      "learning_rate": 6.548451073449538e-05,
+      "loss": 1.2416671514511108,
+      "step": 1708
+    },
+    {
+      "epoch": 0.37231425588889394,
+      "grad_norm": 0.671875,
+      "learning_rate": 6.546721376277109e-05,
+      "loss": 1.017462134361267,
+      "step": 1709
+    },
+    {
+      "epoch": 0.3725321109245223,
+      "grad_norm": 0.6171875,
+      "learning_rate": 6.544990877887926e-05,
+      "loss": 1.1157747507095337,
+      "step": 1710
+    },
+    {
+      "epoch": 0.3727499659601507,
+      "grad_norm": 0.61328125,
+      "learning_rate": 6.543259578826416e-05,
+      "loss": 1.163440227508545,
+      "step": 1711
+    },
+    {
+      "epoch": 0.3729678209957791,
+      "grad_norm": 0.65625,
+      "learning_rate": 6.541527479637256e-05,
+      "loss": 1.176387906074524,
+      "step": 1712
+    },
+    {
+      "epoch": 0.37318567603140745,
+      "grad_norm": 0.6796875,
+      "learning_rate": 6.53979458086538e-05,
+      "loss": 1.1682159900665283,
+      "step": 1713
+    },
+    {
+      "epoch": 0.37340353106703583,
+      "grad_norm": 0.625,
+      "learning_rate": 6.538060883055969e-05,
+      "loss": 1.2080447673797607,
+      "step": 1714
+    },
+    {
+      "epoch": 0.3736213861026642,
+      "grad_norm": 0.60546875,
+      "learning_rate": 6.536326386754456e-05,
+      "loss": 1.084771990776062,
+      "step": 1715
+    },
+    {
+      "epoch": 0.3738392411382926,
+      "grad_norm": 0.62890625,
+      "learning_rate": 6.534591092506528e-05,
+      "loss": 1.157517910003662,
+      "step": 1716
+    },
+    {
+      "epoch": 0.3740570961739209,
+      "grad_norm": 0.6640625,
+      "learning_rate": 6.53285500085812e-05,
+      "loss": 1.2714173793792725,
+      "step": 1717
+    },
+    {
+      "epoch": 0.3742749512095493,
+      "grad_norm": 0.6640625,
+      "learning_rate": 6.531118112355416e-05,
+      "loss": 1.2683064937591553,
+      "step": 1718
+    },
+    {
+      "epoch": 0.37449280624517767,
+      "grad_norm": 0.61328125,
+      "learning_rate": 6.529380427544858e-05,
+      "loss": 1.1400947570800781,
+      "step": 1719
+    },
+    {
+      "epoch": 0.37471066128080605,
+      "grad_norm": 0.60546875,
+      "learning_rate": 6.52764194697313e-05,
+      "loss": 1.049367904663086,
+      "step": 1720
+    },
+    {
+      "epoch": 0.3749285163164344,
+      "grad_norm": 0.73828125,
+      "learning_rate": 6.525902671187175e-05,
+      "loss": 1.1254351139068604,
+      "step": 1721
+    },
+    {
+      "epoch": 0.3751463713520628,
+      "grad_norm": 0.6640625,
+      "learning_rate": 6.524162600734178e-05,
+      "loss": 1.1118007898330688,
+      "step": 1722
+    },
+    {
+      "epoch": 0.37538467124792346,
+      "grad_norm": 0.6328125,
+      "learning_rate": 6.522421736161579e-05,
+      "loss": 1.0529565811157227,
+      "step": 1723
+    },
+    {
+      "epoch": 0.37560253814940225,
+      "grad_norm": 0.67578125,
+      "learning_rate": 6.520186982544992e-05,
+      "loss": 1.0052276849746704,
+      "step": 1724
+    },
+    {
+      "epoch": 0.375820405050881,
+      "grad_norm": 0.61328125,
+      "learning_rate": 6.518443995759093e-05,
+      "loss": 1.0485960245132446,
+      "step": 1725
+    },
+    {
+      "epoch": 0.3760382719523598,
+      "grad_norm": 0.77734375,
+      "learning_rate": 6.516700216369706e-05,
+      "loss": 1.1342464685440063,
+      "step": 1726
+    },
+    {
+      "epoch": 0.3762561388538385,
+      "grad_norm": 0.63671875,
+      "learning_rate": 6.514955644925638e-05,
+      "loss": 1.005650281906128,
+      "step": 1727
+    },
+    {
+      "epoch": 0.3764740057553173,
+      "grad_norm": 0.65625,
+      "learning_rate": 6.513210281975935e-05,
+      "loss": 1.1308025121688843,
+      "step": 1728
+    },
+    {
+      "epoch": 0.3766918726567961,
+      "grad_norm": 0.6484375,
+      "learning_rate": 6.5114641280699e-05,
+      "loss": 1.1613308191299438,
+      "step": 1729
+    },
+    {
+      "epoch": 0.37690973955827484,
+      "grad_norm": 0.69921875,
+      "learning_rate": 6.50971718375708e-05,
+      "loss": 1.3030163049697876,
+      "step": 1730
+    },
+    {
+      "epoch": 0.37712760645975363,
+      "grad_norm": 0.6875,
+      "learning_rate": 6.507969449587272e-05,
+      "loss": 1.0774773359298706,
+      "step": 1731
+    },
+    {
+      "epoch": 0.3773454733612324,
+      "grad_norm": 0.62890625,
+      "learning_rate": 6.506220926110525e-05,
+      "loss": 1.1412074565887451,
+      "step": 1732
+    },
+    {
+      "epoch": 0.37756334026271116,
+      "grad_norm": 0.6171875,
+      "learning_rate": 6.504471613877131e-05,
+      "loss": 1.0989940166473389,
+      "step": 1733
+    },
+    {
+      "epoch": 0.37778120716418995,
+      "grad_norm": 0.6484375,
+      "learning_rate": 6.502721513437632e-05,
+      "loss": 0.9688776135444641,
+      "step": 1734
+    },
+    {
+      "epoch": 0.37799907406566874,
+      "grad_norm": 0.69140625,
+      "learning_rate": 6.500970625342822e-05,
+      "loss": 1.0857795476913452,
+      "step": 1735
+    },
+    {
+      "epoch": 0.3782169409671475,
+      "grad_norm": 0.703125,
+      "learning_rate": 6.499218950143736e-05,
+      "loss": 1.1874654293060303,
+      "step": 1736
+    },
+    {
+      "epoch": 0.37843480786862627,
+      "grad_norm": 0.71875,
+      "learning_rate": 6.497466488391663e-05,
+      "loss": 1.0694936513900757,
+      "step": 1737
+    },
+    {
+      "epoch": 0.37865267477010506,
+      "grad_norm": 0.64453125,
+      "learning_rate": 6.495713240638135e-05,
+      "loss": 1.0289191007614136,
+      "step": 1738
+    },
+    {
+      "epoch": 0.3788705416715838,
+      "grad_norm": 0.7421875,
+      "learning_rate": 6.493959207434934e-05,
+      "loss": 1.1209511756896973,
+      "step": 1739
+    },
+    {
+      "epoch": 0.3790884085730626,
+      "grad_norm": 0.6875,
+      "learning_rate": 6.49220438933409e-05,
+      "loss": 1.232471227645874,
+      "step": 1740
+    },
+    {
+      "epoch": 0.3793062754745413,
+      "grad_norm": 0.64453125,
+      "learning_rate": 6.490448786887877e-05,
+      "loss": 1.1036521196365356,
+      "step": 1741
+    },
+    {
+      "epoch": 0.3795241423760201,
+      "grad_norm": 0.7109375,
+      "learning_rate": 6.488692400648817e-05,
+      "loss": 1.1257684230804443,
+      "step": 1742
+    },
+    {
+      "epoch": 0.3797420092774989,
+      "grad_norm": 0.6484375,
+      "learning_rate": 6.48693523116968e-05,
+      "loss": 1.1527761220932007,
+      "step": 1743
+    },
+    {
+      "epoch": 0.37995987617897764,
+      "grad_norm": 0.62109375,
+      "learning_rate": 6.485177279003482e-05,
+      "loss": 1.029383897781372,
+      "step": 1744
+    },
+    {
+      "epoch": 0.38017774308045643,
+      "grad_norm": 0.65234375,
+      "learning_rate": 6.483418544703483e-05,
+      "loss": 1.1841394901275635,
+      "step": 1745
+    },
+    {
+      "epoch": 0.3803956099819352,
+      "grad_norm": 0.5703125,
+      "learning_rate": 6.481659028823192e-05,
+      "loss": 0.8331100344657898,
+      "step": 1746
+    },
+    {
+      "epoch": 0.38061347688341396,
+      "grad_norm": 0.6875,
+      "learning_rate": 6.479898731916363e-05,
+      "loss": 1.1094539165496826,
+      "step": 1747
+    },
+    {
+      "epoch": 0.38083134378489275,
+      "grad_norm": 0.66015625,
+      "learning_rate": 6.478137654536996e-05,
+      "loss": 1.1497067213058472,
+      "step": 1748
+    },
+    {
+      "epoch": 0.38104921068637154,
+      "grad_norm": 0.640625,
+      "learning_rate": 6.476375797239338e-05,
+      "loss": 1.1395723819732666,
+      "step": 1749
+    },
+    {
+      "epoch": 0.3812670775878503,
+      "grad_norm": 0.6171875,
+      "learning_rate": 6.474613160577874e-05,
+      "loss": 1.0497633218765259,
+      "step": 1750
+    },
+    {
+      "epoch": 0.38148494448932907,
+      "grad_norm": 0.63671875,
+      "learning_rate": 6.472849745107348e-05,
+      "loss": 1.0800789594650269,
+      "step": 1751
+    },
+    {
+      "epoch": 0.3817028113908078,
+      "grad_norm": 0.6796875,
+      "learning_rate": 6.471085551382736e-05,
+      "loss": 1.2812895774841309,
+      "step": 1752
+    },
+    {
+      "epoch": 0.3819206782922866,
+      "grad_norm": 0.6328125,
+      "learning_rate": 6.469320579959267e-05,
+      "loss": 1.0400524139404297,
+      "step": 1753
+    },
+    {
+      "epoch": 0.3821385451937654,
+      "grad_norm": 0.703125,
+      "learning_rate": 6.467554831392409e-05,
+      "loss": 1.2420376539230347,
+      "step": 1754
+    },
+    {
+      "epoch": 0.3823564120952441,
+      "grad_norm": 0.62109375,
+      "learning_rate": 6.46578830623788e-05,
+      "loss": 1.0350443124771118,
+      "step": 1755
+    },
+    {
+      "epoch": 0.3825742789967229,
+      "grad_norm": 0.60546875,
+      "learning_rate": 6.46402100505164e-05,
+      "loss": 1.0479007959365845,
+      "step": 1756
+    },
+    {
+      "epoch": 0.3827921458982017,
+      "grad_norm": 0.671875,
+      "learning_rate": 6.462252928389891e-05,
+      "loss": 1.0888047218322754,
+      "step": 1757
+    },
+    {
+      "epoch": 0.38301001279968044,
+      "grad_norm": 0.6796875,
+      "learning_rate": 6.460484076809084e-05,
+      "loss": 1.1251733303070068,
+      "step": 1758
+    },
+    {
+      "epoch": 0.38322787970115924,
+      "grad_norm": 0.68359375,
+      "learning_rate": 6.458714450865909e-05,
+      "loss": 1.1459622383117676,
+      "step": 1759
+    },
+    {
+      "epoch": 0.383445746602638,
+      "grad_norm": 0.6640625,
+      "learning_rate": 6.456944051117303e-05,
+      "loss": 1.1238489151000977,
+      "step": 1760
+    },
+    {
+      "epoch": 0.38366361350411676,
+      "grad_norm": 0.69921875,
+      "learning_rate": 6.455172878120445e-05,
+      "loss": 1.2678658962249756,
+      "step": 1761
+    },
+    {
+      "epoch": 0.38388148040559555,
+      "grad_norm": 0.64453125,
+      "learning_rate": 6.453400932432758e-05,
+      "loss": 1.2754414081573486,
+      "step": 1762
+    },
+    {
+      "epoch": 0.38409934730707435,
+      "grad_norm": 0.7109375,
+      "learning_rate": 6.451628214611907e-05,
+      "loss": 1.1332813501358032,
+      "step": 1763
+    },
+    {
+      "epoch": 0.3843172142085531,
+      "grad_norm": 0.6328125,
+      "learning_rate": 6.449854725215801e-05,
+      "loss": 1.0112695693969727,
+      "step": 1764
+    },
+    {
+      "epoch": 0.3845350811100319,
+      "grad_norm": 0.671875,
+      "learning_rate": 6.448080464802592e-05,
+      "loss": 1.0047507286071777,
+      "step": 1765
+    },
+    {
+      "epoch": 0.3847529480115106,
+      "grad_norm": 0.65234375,
+      "learning_rate": 6.446305433930676e-05,
+      "loss": 1.2956366539001465,
+      "step": 1766
+    },
+    {
+      "epoch": 0.3849708149129894,
+      "grad_norm": 0.6640625,
+      "learning_rate": 6.444529633158688e-05,
+      "loss": 1.1211198568344116,
+      "step": 1767
+    },
+    {
+      "epoch": 0.3851886818144682,
+      "grad_norm": 0.67578125,
+      "learning_rate": 6.442753063045507e-05,
+      "loss": 1.3321913480758667,
+      "step": 1768
+    },
+    {
+      "epoch": 0.38540654871594693,
+      "grad_norm": 0.62109375,
+      "learning_rate": 6.440975724150255e-05,
+      "loss": 1.0056191682815552,
+      "step": 1769
+    },
+    {
+      "epoch": 0.3856244156174257,
+      "grad_norm": 0.6484375,
+      "learning_rate": 6.439197617032296e-05,
+      "loss": 0.967913031578064,
+      "step": 1770
+    },
+    {
+      "epoch": 0.3858422825189045,
+      "grad_norm": 0.640625,
+      "learning_rate": 6.437418742251234e-05,
+      "loss": 1.1309343576431274,
+      "step": 1771
+    },
+    {
+      "epoch": 0.38606014942038325,
+      "grad_norm": 0.63671875,
+      "learning_rate": 6.435639100366916e-05,
+      "loss": 1.0930434465408325,
+      "step": 1772
+    },
+    {
+      "epoch": 0.38627801632186204,
+      "grad_norm": 0.6484375,
+      "learning_rate": 6.433858691939428e-05,
+      "loss": 1.1312897205352783,
+      "step": 1773
+    },
+    {
+      "epoch": 0.38649588322334083,
+      "grad_norm": 0.671875,
+      "learning_rate": 6.432077517529103e-05,
+      "loss": 1.0609943866729736,
+      "step": 1774
+    },
+    {
+      "epoch": 0.38671375012481957,
+      "grad_norm": 0.66015625,
+      "learning_rate": 6.430295577696507e-05,
+      "loss": 1.2326257228851318,
+      "step": 1775
+    },
+    {
+      "epoch": 0.38693161702629836,
+      "grad_norm": 0.65234375,
+      "learning_rate": 6.428512873002456e-05,
+      "loss": 1.1246012449264526,
+      "step": 1776
+    },
+    {
+      "epoch": 0.38714948392777715,
+      "grad_norm": 0.66796875,
+      "learning_rate": 6.426729404007999e-05,
+      "loss": 1.070870041847229,
+      "step": 1777
+    },
+    {
+      "epoch": 0.3873673508292559,
+      "grad_norm": 0.62109375,
+      "learning_rate": 6.424945171274429e-05,
+      "loss": 1.02096688747406,
+      "step": 1778
+    },
+    {
+      "epoch": 0.3875852177307347,
+      "grad_norm": 0.71484375,
+      "learning_rate": 6.42316017536328e-05,
+      "loss": 1.0925140380859375,
+      "step": 1779
+    },
+    {
+      "epoch": 0.3878030846322134,
+      "grad_norm": 0.64453125,
+      "learning_rate": 6.421374416836323e-05,
+      "loss": 1.118484377861023,
+      "step": 1780
+    },
+    {
+      "epoch": 0.3880209515336922,
+      "grad_norm": 0.62109375,
+      "learning_rate": 6.419587896255574e-05,
+      "loss": 1.04289710521698,
+      "step": 1781
+    },
+    {
+      "epoch": 0.388238818435171,
+      "grad_norm": 0.72265625,
+      "learning_rate": 6.417800614183282e-05,
+      "loss": 1.1482467651367188,
+      "step": 1782
+    },
+    {
+      "epoch": 0.38845668533664973,
+      "grad_norm": 0.62109375,
+      "learning_rate": 6.416012571181942e-05,
+      "loss": 1.0328640937805176,
+      "step": 1783
+    },
+    {
+      "epoch": 0.3886745522381285,
+      "grad_norm": 0.64453125,
+      "learning_rate": 6.414223767814286e-05,
+      "loss": 1.0286579132080078,
+      "step": 1784
+    },
+    {
+      "epoch": 0.3888924191396073,
+      "grad_norm": 0.63671875,
+      "learning_rate": 6.412434204643286e-05,
+      "loss": 0.9707363843917847,
+      "step": 1785
+    },
+    {
+      "epoch": 0.38911028604108605,
+      "grad_norm": 0.65625,
+      "learning_rate": 6.41064388223215e-05,
+      "loss": 0.9243696331977844,
+      "step": 1786
+    },
+    {
+      "epoch": 0.38932815294256484,
+      "grad_norm": 0.64453125,
+      "learning_rate": 6.40885280114433e-05,
+      "loss": 0.9334943890571594,
+      "step": 1787
+    },
+    {
+      "epoch": 0.38954601984404363,
+      "grad_norm": 0.671875,
+      "learning_rate": 6.407060961943512e-05,
+      "loss": 1.1309529542922974,
+      "step": 1788
+    },
+    {
+      "epoch": 0.38976388674552237,
+      "grad_norm": 0.65625,
+      "learning_rate": 6.405268365193624e-05,
+      "loss": 1.164347529411316,
+      "step": 1789
+    },
+    {
+      "epoch": 0.38998175364700116,
+      "grad_norm": 0.62109375,
+      "learning_rate": 6.403475011458831e-05,
+      "loss": 1.1301448345184326,
+      "step": 1790
+    },
+    {
+      "epoch": 0.3901996205484799,
+      "grad_norm": 0.65625,
+      "learning_rate": 6.401680901303537e-05,
+      "loss": 1.1017448902130127,
+      "step": 1791
+    },
+    {
+      "epoch": 0.3904174874499587,
+      "grad_norm": 0.6953125,
+      "learning_rate": 6.399886035292384e-05,
+      "loss": 1.1182204484939575,
+      "step": 1792
+    },
+    {
+      "epoch": 0.3906353543514375,
+      "grad_norm": 0.703125,
+      "learning_rate": 6.398090413990247e-05,
+      "loss": 1.005313754081726,
+      "step": 1793
+    },
+    {
+      "epoch": 0.3908532212529162,
+      "grad_norm": 0.69921875,
+      "learning_rate": 6.396294037962249e-05,
+      "loss": 1.1856545209884644,
+      "step": 1794
+    },
+    {
+      "epoch": 0.391071088154395,
+      "grad_norm": 0.7109375,
+      "learning_rate": 6.394496907773742e-05,
+      "loss": 1.2885886430740356,
+      "step": 1795
+    },
+    {
+      "epoch": 0.3912889550558738,
+      "grad_norm": 0.7109375,
+      "learning_rate": 6.392699023990316e-05,
+      "loss": 1.1432723999023438,
+      "step": 1796
+    },
+    {
+      "epoch": 0.39150682195735254,
+      "grad_norm": 0.75,
+      "learning_rate": 6.390900387177803e-05,
+      "loss": 1.1251109838485718,
+      "step": 1797
+    },
+    {
+      "epoch": 0.3917246888588313,
+      "grad_norm": 0.68359375,
+      "learning_rate": 6.389100997902268e-05,
+      "loss": 1.0353763103485107,
+      "step": 1798
+    },
+    {
+      "epoch": 0.3919425557603101,
+      "grad_norm": 0.66796875,
+      "learning_rate": 6.387300856730015e-05,
+      "loss": 1.1555976867675781,
+      "step": 1799
+    },
+    {
+      "epoch": 0.39216042266178885,
+      "grad_norm": 0.62109375,
+      "learning_rate": 6.385499964227583e-05,
+      "loss": 0.9700635671615601,
+      "step": 1800
+    },
+    {
+      "epoch": 0.39237828956326765,
+      "grad_norm": 0.80859375,
+      "learning_rate": 6.383698320961746e-05,
+      "loss": 1.1430225372314453,
+      "step": 1801
+    },
+    {
+      "epoch": 0.39259615646474644,
+      "grad_norm": 0.78515625,
+      "learning_rate": 6.381895927499519e-05,
+      "loss": 1.1007705926895142,
+      "step": 1802
+    },
+    {
+      "epoch": 0.3928140233662252,
+      "grad_norm": 0.71875,
+      "learning_rate": 6.380092784408149e-05,
+      "loss": 1.0586771965026855,
+      "step": 1803
+    },
+    {
+      "epoch": 0.39303189026770396,
+      "grad_norm": 0.66796875,
+      "learning_rate": 6.378288892255122e-05,
+      "loss": 1.0912443399429321,
+      "step": 1804
+    },
+    {
+      "epoch": 0.3932497571691827,
+      "grad_norm": 0.64453125,
+      "learning_rate": 6.376484251608155e-05,
+      "loss": 0.9680395722389221,
+      "step": 1805
+    },
+    {
+      "epoch": 0.3934676240706615,
+      "grad_norm": 0.6640625,
+      "learning_rate": 6.374678863035206e-05,
+      "loss": 0.9774947166442871,
+      "step": 1806
+    },
+    {
+      "epoch": 0.3936854909721403,
+      "grad_norm": 0.67578125,
+      "learning_rate": 6.372872727104465e-05,
+      "loss": 1.0916672945022583,
+      "step": 1807
+    },
+    {
+      "epoch": 0.393903357873619,
+      "grad_norm": 0.734375,
+      "learning_rate": 6.37106584438436e-05,
+      "loss": 1.240972638130188,
+      "step": 1808
+    },
+    {
+      "epoch": 0.3941212247750978,
+      "grad_norm": 0.66796875,
+      "learning_rate": 6.36925821544355e-05,
+      "loss": 1.1421629190444946,
+      "step": 1809
+    },
+    {
+      "epoch": 0.3943390916765766,
+      "grad_norm": 0.69140625,
+      "learning_rate": 6.367449840850932e-05,
+      "loss": 1.0104838609695435,
+      "step": 1810
+    },
+    {
+      "epoch": 0.39455695857805534,
+      "grad_norm": 0.6484375,
+      "learning_rate": 6.365640721175635e-05,
+      "loss": 1.1346025466918945,
+      "step": 1811
+    },
+    {
+      "epoch": 0.39477482547953413,
+      "grad_norm": 0.62890625,
+      "learning_rate": 6.36383085698703e-05,
+      "loss": 1.0957621335983276,
+      "step": 1812
+    },
+    {
+      "epoch": 0.3949926923810129,
+      "grad_norm": 0.66015625,
+      "learning_rate": 6.36202024885471e-05,
+      "loss": 1.098905086517334,
+      "step": 1813
+    },
+    {
+      "epoch": 0.39521055928249166,
+      "grad_norm": 0.6484375,
+      "learning_rate": 6.360208897348513e-05,
+      "loss": 1.0912461280822754,
+      "step": 1814
+    },
+    {
+      "epoch": 0.39542842618397045,
+      "grad_norm": 0.625,
+      "learning_rate": 6.358396803038505e-05,
+      "loss": 0.9739792943000793,
+      "step": 1815
+    },
+    {
+      "epoch": 0.39564629308544924,
+      "grad_norm": 0.64453125,
+      "learning_rate": 6.356583966494987e-05,
+      "loss": 1.2624092102050781,
+      "step": 1816
+    },
+    {
+      "epoch": 0.395864159986928,
+      "grad_norm": 0.640625,
+      "learning_rate": 6.354770388288494e-05,
+      "loss": 1.1486108303070068,
+      "step": 1817
+    },
+    {
+      "epoch": 0.39608202688840677,
+      "grad_norm": 0.6328125,
+      "learning_rate": 6.352956068989796e-05,
+      "loss": 1.087213397026062,
+      "step": 1818
+    },
+    {
+      "epoch": 0.3962998937898855,
+      "grad_norm": 0.64453125,
+      "learning_rate": 6.351141009169893e-05,
+      "loss": 1.1372548341751099,
+      "step": 1819
+    },
+    {
+      "epoch": 0.3965177606913643,
+      "grad_norm": 0.6640625,
+      "learning_rate": 6.349325209400021e-05,
+      "loss": 1.2276948690414429,
+      "step": 1820
+    },
+    {
+      "epoch": 0.3967356275928431,
+      "grad_norm": 0.640625,
+      "learning_rate": 6.347508670251646e-05,
+      "loss": 1.1603329181671143,
+      "step": 1821
+    },
+    {
+      "epoch": 0.3969534944943218,
+      "grad_norm": 0.62890625,
+      "learning_rate": 6.345691392296472e-05,
+      "loss": 1.0147939920425415,
+      "step": 1822
+    },
+    {
+      "epoch": 0.3971713613958006,
+      "grad_norm": 0.69140625,
+      "learning_rate": 6.343873376106426e-05,
+      "loss": 1.0370889902114868,
+      "step": 1823
+    },
+    {
+      "epoch": 0.3973892282972794,
+      "grad_norm": 0.69921875,
+      "learning_rate": 6.342054622253679e-05,
+      "loss": 1.245137333869934,
+      "step": 1824
+    },
+    {
+      "epoch": 0.39760709519875814,
+      "grad_norm": 0.6015625,
+      "learning_rate": 6.340235131310626e-05,
+      "loss": 0.9462807774543762,
+      "step": 1825
+    },
+    {
+      "epoch": 0.39782496210023693,
+      "grad_norm": 0.6875,
+      "learning_rate": 6.338414903849895e-05,
+      "loss": 1.1710662841796875,
+      "step": 1826
+    },
+    {
+      "epoch": 0.3980428290017157,
+      "grad_norm": 0.66015625,
+      "learning_rate": 6.336593940444349e-05,
+      "loss": 1.0835802555084229,
+      "step": 1827
+    },
+    {
+      "epoch": 0.39826069590319446,
+      "grad_norm": 0.6015625,
+      "learning_rate": 6.334772241667078e-05,
+      "loss": 0.9631011486053467,
+      "step": 1828
+    },
+    {
+      "epoch": 0.39847856280467325,
+      "grad_norm": 0.66796875,
+      "learning_rate": 6.332949808091411e-05,
+      "loss": 1.0388270616531372,
+      "step": 1829
+    },
+    {
+      "epoch": 0.398696429706152,
+      "grad_norm": 0.6640625,
+      "learning_rate": 6.331126640290899e-05,
+      "loss": 1.2379705905914307,
+      "step": 1830
+    },
+    {
+      "epoch": 0.3989142966076308,
+      "grad_norm": 0.7578125,
+      "learning_rate": 6.32930273883933e-05,
+      "loss": 1.2046443223953247,
+      "step": 1831
+    },
+    {
+      "epoch": 0.39913216350910957,
+      "grad_norm": 0.6796875,
+      "learning_rate": 6.327478104310723e-05,
+      "loss": 1.1935940980911255,
+      "step": 1832
+    },
+    {
+      "epoch": 0.3993500304105883,
+      "grad_norm": 0.6796875,
+      "learning_rate": 6.325652737279323e-05,
+      "loss": 1.14521062374115,
+      "step": 1833
+    },
+    {
+      "epoch": 0.3995678973120671,
+      "grad_norm": 0.69921875,
+      "learning_rate": 6.323826638319612e-05,
+      "loss": 1.0924419164657593,
+      "step": 1834
+    },
+    {
+      "epoch": 0.3997857642135459,
+      "grad_norm": 0.66015625,
+      "learning_rate": 6.321999808006294e-05,
+      "loss": 1.150847315788269,
+      "step": 1835
+    },
+    {
+      "epoch": 0.4000036311150246,
+      "grad_norm": 0.62890625,
+      "learning_rate": 6.320172246914312e-05,
+      "loss": 0.8994224071502686,
+      "step": 1836
+    },
+    {
+      "epoch": 0.4002214980165034,
+      "grad_norm": 0.66015625,
+      "learning_rate": 6.318343955618837e-05,
+      "loss": 1.0981417894363403,
+      "step": 1837
+    },
+    {
+      "epoch": 0.4004393649179822,
+      "grad_norm": 0.65625,
+      "learning_rate": 6.316514934695265e-05,
+      "loss": 1.0792003870010376,
+      "step": 1838
+    },
+    {
+      "epoch": 0.40065723181946095,
+      "grad_norm": 0.69140625,
+      "learning_rate": 6.314685184719224e-05,
+      "loss": 1.1621711254119873,
+      "step": 1839
+    },
+    {
+      "epoch": 0.40087509872093974,
+      "grad_norm": 0.64453125,
+      "learning_rate": 6.312854706266573e-05,
+      "loss": 1.0330157279968262,
+      "step": 1840
+    },
+    {
+      "epoch": 0.40109296562241853,
+      "grad_norm": 0.63671875,
+      "learning_rate": 6.3110234999134e-05,
+      "loss": 1.0749393701553345,
+      "step": 1841
+    },
+    {
+      "epoch": 0.40131083252389727,
+      "grad_norm": 0.66796875,
+      "learning_rate": 6.30919156623602e-05,
+      "loss": 1.1624059677124023,
+      "step": 1842
+    },
+    {
+      "epoch": 0.40152869942537606,
+      "grad_norm": 0.609375,
+      "learning_rate": 6.30735890581098e-05,
+      "loss": 1.0758841037750244,
+      "step": 1843
+    },
+    {
+      "epoch": 0.4017465663268548,
+      "grad_norm": 0.65234375,
+      "learning_rate": 6.305525519215052e-05,
+      "loss": 1.01979398727417,
+      "step": 1844
+    },
+    {
+      "epoch": 0.4019644332283336,
+      "grad_norm": 0.609375,
+      "learning_rate": 6.303691407025239e-05,
+      "loss": 1.057625412940979,
+      "step": 1845
+    },
+    {
+      "epoch": 0.4021823001298124,
+      "grad_norm": 0.6875,
+      "learning_rate": 6.301856569818774e-05,
+      "loss": 1.2921137809753418,
+      "step": 1846
+    },
+    {
+      "epoch": 0.4024001670312911,
+      "grad_norm": 0.671875,
+      "learning_rate": 6.300021008173116e-05,
+      "loss": 1.0984992980957031,
+      "step": 1847
+    },
+    {
+      "epoch": 0.4026180339327699,
+      "grad_norm": 0.625,
+      "learning_rate": 6.298184722665949e-05,
+      "loss": 1.110878825187683,
+      "step": 1848
+    },
+    {
+      "epoch": 0.4028359008342487,
+      "grad_norm": 0.6875,
+      "learning_rate": 6.296347713875188e-05,
+      "loss": 1.063165545463562,
+      "step": 1849
+    },
+    {
+      "epoch": 0.40305376773572743,
+      "grad_norm": 0.65234375,
+      "learning_rate": 6.294509982378978e-05,
+      "loss": 1.0200309753417969,
+      "step": 1850
+    },
+    {
+      "epoch": 0.4032716346372062,
+      "grad_norm": 0.671875,
+      "learning_rate": 6.29267152875569e-05,
+      "loss": 1.1148045063018799,
+      "step": 1851
+    },
+    {
+      "epoch": 0.403489501538685,
+      "grad_norm": 0.65234375,
+      "learning_rate": 6.290832353583917e-05,
+      "loss": 1.0960549116134644,
+      "step": 1852
+    },
+    {
+      "epoch": 0.40370736844016375,
+      "grad_norm": 0.69140625,
+      "learning_rate": 6.288992457442487e-05,
+      "loss": 1.1292675733566284,
+      "step": 1853
+    },
+    {
+      "epoch": 0.40392523534164254,
+      "grad_norm": 0.68359375,
+      "learning_rate": 6.287151840910449e-05,
+      "loss": 1.1491318941116333,
+      "step": 1854
+    },
+    {
+      "epoch": 0.40414310224312133,
+      "grad_norm": 0.6328125,
+      "learning_rate": 6.285310504567083e-05,
+      "loss": 1.027625322341919,
+      "step": 1855
+    },
+    {
+      "epoch": 0.40436096914460007,
+      "grad_norm": 0.66796875,
+      "learning_rate": 6.283468448991892e-05,
+      "loss": 1.114548921585083,
+      "step": 1856
+    },
+    {
+      "epoch": 0.40457883604607886,
+      "grad_norm": 0.63671875,
+      "learning_rate": 6.281625674764607e-05,
+      "loss": 1.099264144897461,
+      "step": 1857
+    },
+    {
+      "epoch": 0.4047967029475576,
+      "grad_norm": 0.66015625,
+      "learning_rate": 6.279782182465184e-05,
+      "loss": 1.086155652999878,
+      "step": 1858
+    },
+    {
+      "epoch": 0.4050145698490364,
+      "grad_norm": 0.6640625,
+      "learning_rate": 6.277937972673808e-05,
+      "loss": 1.109994649887085,
+      "step": 1859
+    },
+    {
+      "epoch": 0.4052324367505152,
+      "grad_norm": 0.65234375,
+      "learning_rate": 6.276093045970888e-05,
+      "loss": 1.0587780475616455,
+      "step": 1860
+    },
+    {
+      "epoch": 0.4054503036519939,
+      "grad_norm": 0.63671875,
+      "learning_rate": 6.274247402937057e-05,
+      "loss": 1.1804660558700562,
+      "step": 1861
+    },
+    {
+      "epoch": 0.4056681705534727,
+      "grad_norm": 0.67578125,
+      "learning_rate": 6.272401044153174e-05,
+      "loss": 1.0779935121536255,
+      "step": 1862
+    },
+    {
+      "epoch": 0.4058860374549515,
+      "grad_norm": 0.67578125,
+      "learning_rate": 6.270553970200328e-05,
+      "loss": 1.0518529415130615,
+      "step": 1863
+    },
+    {
+      "epoch": 0.40610390435643023,
+      "grad_norm": 0.6484375,
+      "learning_rate": 6.268706181659823e-05,
+      "loss": 1.0569307804107666,
+      "step": 1864
+    },
+    {
+      "epoch": 0.406321771257909,
+      "grad_norm": 0.6640625,
+      "learning_rate": 6.266857679113202e-05,
+      "loss": 0.9633741974830627,
+      "step": 1865
+    },
+    {
+      "epoch": 0.4065396381593878,
+      "grad_norm": 0.63671875,
+      "learning_rate": 6.265008463142218e-05,
+      "loss": 1.0988106727600098,
+      "step": 1866
+    },
+    {
+      "epoch": 0.40675750506086655,
+      "grad_norm": 0.6953125,
+      "learning_rate": 6.263158534328859e-05,
+      "loss": 1.0015345811843872,
+      "step": 1867
+    },
+    {
+      "epoch": 0.40697537196234534,
+      "grad_norm": 0.64453125,
+      "learning_rate": 6.261307893255332e-05,
+      "loss": 1.1972092390060425,
+      "step": 1868
+    },
+    {
+      "epoch": 0.40719323886382414,
+      "grad_norm": 0.64453125,
+      "learning_rate": 6.259456540504072e-05,
+      "loss": 1.0276159048080444,
+      "step": 1869
+    },
+    {
+      "epoch": 0.40741110576530287,
+      "grad_norm": 0.63671875,
+      "learning_rate": 6.257604476657733e-05,
+      "loss": 0.980731725692749,
+      "step": 1870
+    },
+    {
+      "epoch": 0.40762897266678166,
+      "grad_norm": 0.77734375,
+      "learning_rate": 6.255751702299197e-05,
+      "loss": 1.1949776411056519,
+      "step": 1871
+    },
+    {
+      "epoch": 0.4078468395682604,
+      "grad_norm": 0.66796875,
+      "learning_rate": 6.25389821801157e-05,
+      "loss": 0.9800863862037659,
+      "step": 1872
+    },
+    {
+      "epoch": 0.4080647064697392,
+      "grad_norm": 0.7265625,
+      "learning_rate": 6.252044024378176e-05,
+      "loss": 1.2662657499313354,
+      "step": 1873
+    },
+    {
+      "epoch": 0.408282573371218,
+      "grad_norm": 0.73046875,
+      "learning_rate": 6.25018912198257e-05,
+      "loss": 1.1021242141723633,
+      "step": 1874
+    },
+    {
+      "epoch": 0.4085004402726967,
+      "grad_norm": 0.66796875,
+      "learning_rate": 6.248333511408522e-05,
+      "loss": 1.0286906957626343,
+      "step": 1875
+    },
+    {
+      "epoch": 0.4087183071741755,
+      "grad_norm": 0.6875,
+      "learning_rate": 6.246477193240033e-05,
+      "loss": 1.0001877546310425,
+      "step": 1876
+    },
+    {
+      "epoch": 0.4089361740756543,
+      "grad_norm": 0.74609375,
+      "learning_rate": 6.24462016806132e-05,
+      "loss": 1.0645179748535156,
+      "step": 1877
+    },
+    {
+      "epoch": 0.40915404097713304,
+      "grad_norm": 0.66796875,
+      "learning_rate": 6.242762436456824e-05,
+      "loss": 1.069293737411499,
+      "step": 1878
+    },
+    {
+      "epoch": 0.40937190787861183,
+      "grad_norm": 0.6953125,
+      "learning_rate": 6.240903999011211e-05,
+      "loss": 1.1764706373214722,
+      "step": 1879
+    },
+    {
+      "epoch": 0.4095897747800906,
+      "grad_norm": 0.66015625,
+      "learning_rate": 6.239044856309368e-05,
+      "loss": 1.0981297492980957,
+      "step": 1880
+    },
+    {
+      "epoch": 0.40980764168156936,
+      "grad_norm": 0.6796875,
+      "learning_rate": 6.237185008936404e-05,
+      "loss": 1.0748274326324463,
+      "step": 1881
+    },
+    {
+      "epoch": 0.41002550858304815,
+      "grad_norm": 0.6875,
+      "learning_rate": 6.235324457477646e-05,
+      "loss": 1.0155075788497925,
+      "step": 1882
+    },
+    {
+      "epoch": 0.4102433754845269,
+      "grad_norm": 0.66015625,
+      "learning_rate": 6.23346320251865e-05,
+      "loss": 1.053336262702942,
+      "step": 1883
+    },
+    {
+      "epoch": 0.4104612423860057,
+      "grad_norm": 0.6796875,
+      "learning_rate": 6.231601244645187e-05,
+      "loss": 0.9894386529922485,
+      "step": 1884
+    },
+    {
+      "epoch": 0.41067910928748447,
+      "grad_norm": 0.68359375,
+      "learning_rate": 6.229738584443251e-05,
+      "loss": 1.0388132333755493,
+      "step": 1885
+    },
+    {
+      "epoch": 0.4108969761889632,
+      "grad_norm": 0.71875,
+      "learning_rate": 6.227875222499059e-05,
+      "loss": 1.1629416942596436,
+      "step": 1886
+    },
+    {
+      "epoch": 0.411114843090442,
+      "grad_norm": 0.70703125,
+      "learning_rate": 6.226011159399047e-05,
+      "loss": 1.1653544902801514,
+      "step": 1887
+    },
+    {
+      "epoch": 0.4113327099919208,
+      "grad_norm": 0.6640625,
+      "learning_rate": 6.22414639572987e-05,
+      "loss": 1.1374397277832031,
+      "step": 1888
+    },
+    {
+      "epoch": 0.4115505768933995,
+      "grad_norm": 0.69140625,
+      "learning_rate": 6.22228093207841e-05,
+      "loss": 1.0541682243347168,
+      "step": 1889
+    },
+    {
+      "epoch": 0.4117684437948783,
+      "grad_norm": 0.6484375,
+      "learning_rate": 6.22041476903176e-05,
+      "loss": 1.0161957740783691,
+      "step": 1890
+    },
+    {
+      "epoch": 0.4119863106963571,
+      "grad_norm": 0.6328125,
+      "learning_rate": 6.218547907177242e-05,
+      "loss": 0.948951780796051,
+      "step": 1891
+    },
+    {
+      "epoch": 0.41220417759783584,
+      "grad_norm": 0.6875,
+      "learning_rate": 6.21668034710239e-05,
+      "loss": 1.2530971765518188,
+      "step": 1892
+    },
+    {
+      "epoch": 0.41242204449931463,
+      "grad_norm": 0.640625,
+      "learning_rate": 6.214812089394964e-05,
+      "loss": 1.100982904434204,
+      "step": 1893
+    },
+    {
+      "epoch": 0.4126399114007934,
+      "grad_norm": 0.60546875,
+      "learning_rate": 6.21294313464294e-05,
+      "loss": 0.8721302151679993,
+      "step": 1894
+    },
+    {
+      "epoch": 0.41285777830227216,
+      "grad_norm": 0.69140625,
+      "learning_rate": 6.211073483434517e-05,
+      "loss": 1.18142569065094,
+      "step": 1895
+    },
+    {
+      "epoch": 0.41307564520375095,
+      "grad_norm": 0.65234375,
+      "learning_rate": 6.209203136358109e-05,
+      "loss": 1.1672581434249878,
+      "step": 1896
+    },
+    {
+      "epoch": 0.4132935121052297,
+      "grad_norm": 0.64453125,
+      "learning_rate": 6.207332094002347e-05,
+      "loss": 0.9802240133285522,
+      "step": 1897
+    },
+    {
+      "epoch": 0.4135113790067085,
+      "grad_norm": 0.65234375,
+      "learning_rate": 6.205460356956092e-05,
+      "loss": 1.0687721967697144,
+      "step": 1898
+    },
+    {
+      "epoch": 0.41372924590818727,
+      "grad_norm": 0.671875,
+      "learning_rate": 6.20358792580841e-05,
+      "loss": 1.0272932052612305,
+      "step": 1899
+    },
+    {
+      "epoch": 0.413947112809666,
+      "grad_norm": 0.66796875,
+      "learning_rate": 6.201714801148596e-05,
+      "loss": 0.9601287841796875,
+      "step": 1900
+    },
+    {
+      "epoch": 0.4141649797111448,
+      "grad_norm": 0.6484375,
+      "learning_rate": 6.199840983566157e-05,
+      "loss": 1.14360773563385,
+      "step": 1901
+    },
+    {
+      "epoch": 0.4143828466126236,
+      "grad_norm": 0.67578125,
+      "learning_rate": 6.197966473650819e-05,
+      "loss": 1.1436846256256104,
+      "step": 1902
+    },
+    {
+      "epoch": 0.4146007135141023,
+      "grad_norm": 0.640625,
+      "learning_rate": 6.196091271992528e-05,
+      "loss": 1.1407570838928223,
+      "step": 1903
+    },
+    {
+      "epoch": 0.4148185804155811,
+      "grad_norm": 0.63671875,
+      "learning_rate": 6.194215379181446e-05,
+      "loss": 1.1188223361968994,
+      "step": 1904
+    },
+    {
+      "epoch": 0.4150364473170599,
+      "grad_norm": 0.64453125,
+      "learning_rate": 6.192338795807954e-05,
+      "loss": 0.9962555766105652,
+      "step": 1905
+    },
+    {
+      "epoch": 0.41525431421853864,
+      "grad_norm": 0.62109375,
+      "learning_rate": 6.190461522462649e-05,
+      "loss": 1.0205055475234985,
+      "step": 1906
+    },
+    {
+      "epoch": 0.41547218112001744,
+      "grad_norm": 0.65234375,
+      "learning_rate": 6.188583559736345e-05,
+      "loss": 1.1042959690093994,
+      "step": 1907
+    },
+    {
+      "epoch": 0.41569004802149623,
+      "grad_norm": 0.66015625,
+      "learning_rate": 6.186704908220074e-05,
+      "loss": 1.066254734992981,
+      "step": 1908
+    },
+    {
+      "epoch": 0.41590791492297496,
+      "grad_norm": 0.72265625,
+      "learning_rate": 6.184825568505083e-05,
+      "loss": 1.1726858615875244,
+      "step": 1909
+    },
+    {
+      "epoch": 0.41612578182445376,
+      "grad_norm": 0.734375,
+      "learning_rate": 6.18294554118284e-05,
+      "loss": 1.1941561698913574,
+      "step": 1910
+    },
+    {
+      "epoch": 0.4163436487259325,
+      "grad_norm": 0.625,
+      "learning_rate": 6.181064826845024e-05,
+      "loss": 1.0128939151763916,
+      "step": 1911
+    },
+    {
+      "epoch": 0.4165615156274113,
+      "grad_norm": 0.609375,
+      "learning_rate": 6.179183426083533e-05,
+      "loss": 0.9807396531105042,
+      "step": 1912
+    },
+    {
+      "epoch": 0.4167793825288901,
+      "grad_norm": 0.65234375,
+      "learning_rate": 6.177301339490481e-05,
+      "loss": 1.0619124174118042,
+      "step": 1913
+    },
+    {
+      "epoch": 0.4169972494303688,
+      "grad_norm": 0.66015625,
+      "learning_rate": 6.175418567658196e-05,
+      "loss": 1.1435142755508423,
+      "step": 1914
+    },
+    {
+      "epoch": 0.4172151163318476,
+      "grad_norm": 0.6796875,
+      "learning_rate": 6.173535111179223e-05,
+      "loss": 1.0681767463684082,
+      "step": 1915
+    },
+    {
+      "epoch": 0.4174329832333264,
+      "grad_norm": 0.68359375,
+      "learning_rate": 6.171650970646324e-05,
+      "loss": 1.252109408378601,
+      "step": 1916
+    },
+    {
+      "epoch": 0.41765085013480513,
+      "grad_norm": 0.6640625,
+      "learning_rate": 6.169766146652476e-05,
+      "loss": 1.0513243675231934,
+      "step": 1917
+    },
+    {
+      "epoch": 0.4178687170362839,
+      "grad_norm": 0.71875,
+      "learning_rate": 6.167880639790867e-05,
+      "loss": 1.121030330657959,
+      "step": 1918
+    },
+    {
+      "epoch": 0.4180865839377627,
+      "grad_norm": 0.67578125,
+      "learning_rate": 6.165994450654904e-05,
+      "loss": 1.165393352508545,
+      "step": 1919
+    },
+    {
+      "epoch": 0.41830445083924145,
+      "grad_norm": 0.6953125,
+      "learning_rate": 6.164107579838207e-05,
+      "loss": 1.161364197731018,
+      "step": 1920
+    },
+    {
+      "epoch": 0.41852231774072024,
+      "grad_norm": 0.72265625,
+      "learning_rate": 6.162220027934611e-05,
+      "loss": 1.2122802734375,
+      "step": 1921
+    },
+    {
+      "epoch": 0.418740184642199,
+      "grad_norm": 0.65625,
+      "learning_rate": 6.160331795538168e-05,
+      "loss": 1.049220085144043,
+      "step": 1922
+    },
+    {
+      "epoch": 0.41895805154367777,
+      "grad_norm": 0.69140625,
+      "learning_rate": 6.158442883243138e-05,
+      "loss": 1.1243782043457031,
+      "step": 1923
+    },
+    {
+      "epoch": 0.41917591844515656,
+      "grad_norm": 0.65625,
+      "learning_rate": 6.156553291644002e-05,
+      "loss": 1.0794941186904907,
+      "step": 1924
+    },
+    {
+      "epoch": 0.4193937853466353,
+      "grad_norm": 0.671875,
+      "learning_rate": 6.154663021335448e-05,
+      "loss": 1.1088992357254028,
+      "step": 1925
+    },
+    {
+      "epoch": 0.4196116522481141,
+      "grad_norm": 0.69140625,
+      "learning_rate": 6.152772072912382e-05,
+      "loss": 1.0468710660934448,
+      "step": 1926
+    },
+    {
+      "epoch": 0.4198295191495929,
+      "grad_norm": 0.6796875,
+      "learning_rate": 6.150880446969925e-05,
+      "loss": 0.9903600215911865,
+      "step": 1927
+    },
+    {
+      "epoch": 0.4200473860510716,
+      "grad_norm": 0.671875,
+      "learning_rate": 6.148988144103405e-05,
+      "loss": 1.0827951431274414,
+      "step": 1928
+    },
+    {
+      "epoch": 0.4202652529525504,
+      "grad_norm": 0.68359375,
+      "learning_rate": 6.147095164908368e-05,
+      "loss": 1.115965723991394,
+      "step": 1929
+    },
+    {
+      "epoch": 0.4204831198540292,
+      "grad_norm": 0.65625,
+      "learning_rate": 6.145201509980572e-05,
+      "loss": 1.0705498456954956,
+      "step": 1930
+    },
+    {
+      "epoch": 0.42070098675550793,
+      "grad_norm": 0.69140625,
+      "learning_rate": 6.143307179915987e-05,
+      "loss": 1.151672124862671,
+      "step": 1931
+    },
+    {
+      "epoch": 0.4209188536569867,
+      "grad_norm": 0.68359375,
+      "learning_rate": 6.141412175310795e-05,
+      "loss": 0.9708853960037231,
+      "step": 1932
+    },
+    {
+      "epoch": 0.4211367205584655,
+      "grad_norm": 0.6640625,
+      "learning_rate": 6.139516496761391e-05,
+      "loss": 1.0335670709609985,
+      "step": 1933
+    },
+    {
+      "epoch": 0.42135458745994425,
+      "grad_norm": 0.6484375,
+      "learning_rate": 6.137620144864385e-05,
+      "loss": 1.1199465990066528,
+      "step": 1934
+    },
+    {
+      "epoch": 0.42157245436142304,
+      "grad_norm": 0.69140625,
+      "learning_rate": 6.135723120216591e-05,
+      "loss": 1.0102906227111816,
+      "step": 1935
+    },
+    {
+      "epoch": 0.4217903212629018,
+      "grad_norm": 0.66015625,
+      "learning_rate": 6.133825423415042e-05,
+      "loss": 0.9960277080535889,
+      "step": 1936
+    },
+    {
+      "epoch": 0.42200818816438057,
+      "grad_norm": 0.7109375,
+      "learning_rate": 6.131927055056982e-05,
+      "loss": 1.1750762462615967,
+      "step": 1937
+    },
+    {
+      "epoch": 0.42222605506585936,
+      "grad_norm": 0.6796875,
+      "learning_rate": 6.130028015739862e-05,
+      "loss": 1.1021322011947632,
+      "step": 1938
+    },
+    {
+      "epoch": 0.4224439219673381,
+      "grad_norm": 0.703125,
+      "learning_rate": 6.128128306061347e-05,
+      "loss": 1.0870922803878784,
+      "step": 1939
+    },
+    {
+      "epoch": 0.4226617888688169,
+      "grad_norm": 0.63671875,
+      "learning_rate": 6.126227926619313e-05,
+      "loss": 1.1016945838928223,
+      "step": 1940
+    },
+    {
+      "epoch": 0.4228796557702957,
+      "grad_norm": 0.734375,
+      "learning_rate": 6.124326878011848e-05,
+      "loss": 1.2612463235855103,
+      "step": 1941
+    },
+    {
+      "epoch": 0.4230975226717744,
+      "grad_norm": 0.6796875,
+      "learning_rate": 6.122425160837248e-05,
+      "loss": 0.8206098675727844,
+      "step": 1942
+    },
+    {
+      "epoch": 0.4233153895732532,
+      "grad_norm": 0.6484375,
+      "learning_rate": 6.120522775694021e-05,
+      "loss": 1.0014207363128662,
+      "step": 1943
+    },
+    {
+      "epoch": 0.423533256474732,
+      "grad_norm": 0.734375,
+      "learning_rate": 6.118619723180881e-05,
+      "loss": 1.1247879266738892,
+      "step": 1944
+    },
+    {
+      "epoch": 0.42375112337621074,
+      "grad_norm": 0.72265625,
+      "learning_rate": 6.116716003896763e-05,
+      "loss": 1.2366496324539185,
+      "step": 1945
+    },
+    {
+      "epoch": 0.42396899027768953,
+      "grad_norm": 0.7265625,
+      "learning_rate": 6.114811618440799e-05,
+      "loss": 1.1159974336624146,
+      "step": 1946
+    },
+    {
+      "epoch": 0.4241868571791683,
+      "grad_norm": 0.68359375,
+      "learning_rate": 6.112906567412338e-05,
+      "loss": 0.9090871214866638,
+      "step": 1947
+    },
+    {
+      "epoch": 0.42440472408064706,
+      "grad_norm": 0.640625,
+      "learning_rate": 6.111000851410937e-05,
+      "loss": 1.075299620628357,
+      "step": 1948
+    },
+    {
+      "epoch": 0.42462259098212585,
+      "grad_norm": 0.66015625,
+      "learning_rate": 6.109094471036361e-05,
+      "loss": 0.9822022318840027,
+      "step": 1949
+    },
+    {
+      "epoch": 0.4248404578836046,
+      "grad_norm": 0.671875,
+      "learning_rate": 6.107187426888585e-05,
+      "loss": 1.0201077461242676,
+      "step": 1950
+    },
+    {
+      "epoch": 0.4250583247850834,
+      "grad_norm": 0.671875,
+      "learning_rate": 6.105279719567796e-05,
+      "loss": 0.9946369528770447,
+      "step": 1951
+    },
+    {
+      "epoch": 0.42527619168656217,
+      "grad_norm": 0.70703125,
+      "learning_rate": 6.103371349674383e-05,
+      "loss": 1.1058839559555054,
+      "step": 1952
+    },
+    {
+      "epoch": 0.4254940585880409,
+      "grad_norm": 0.640625,
+      "learning_rate": 6.1014623178089477e-05,
+      "loss": 1.1300311088562012,
+      "step": 1953
+    },
+    {
+      "epoch": 0.4257119254895197,
+      "grad_norm": 0.671875,
+      "learning_rate": 6.099552624572303e-05,
+      "loss": 1.0840481519699097,
+      "step": 1954
+    },
+    {
+      "epoch": 0.4259297923909985,
+      "grad_norm": 0.81640625,
+      "learning_rate": 6.097642270565463e-05,
+      "loss": 1.1458979845046997,
+      "step": 1955
+    },
+    {
+      "epoch": 0.4261476592924772,
+      "grad_norm": 0.6875,
+      "learning_rate": 6.095731256389656e-05,
+      "loss": 1.0463526248931885,
+      "step": 1956
+    },
+    {
+      "epoch": 0.426365526193956,
+      "grad_norm": 0.625,
+      "learning_rate": 6.093819582646314e-05,
+      "loss": 0.9988933801651001,
+      "step": 1957
+    },
+    {
+      "epoch": 0.4265833930954348,
+      "grad_norm": 0.6484375,
+      "learning_rate": 6.091907249937077e-05,
+      "loss": 1.0161174535751343,
+      "step": 1958
+    },
+    {
+      "epoch": 0.42680125999691354,
+      "grad_norm": 0.66796875,
+      "learning_rate": 6.0899942588637955e-05,
+      "loss": 1.214343547821045,
+      "step": 1959
+    },
+    {
+      "epoch": 0.42701912689839233,
+      "grad_norm": 0.6875,
+      "learning_rate": 6.0880806100285257e-05,
+      "loss": 1.2266663312911987,
+      "step": 1960
+    },
+    {
+      "epoch": 0.4272369937998711,
+      "grad_norm": 0.65625,
+      "learning_rate": 6.086166304033528e-05,
+      "loss": 1.2497063875198364,
+      "step": 1961
+    },
+    {
+      "epoch": 0.42745486070134986,
+      "grad_norm": 0.65234375,
+      "learning_rate": 6.0842513414812723e-05,
+      "loss": 1.0278915166854858,
+      "step": 1962
+    },
+    {
+      "epoch": 0.42767272760282865,
+      "grad_norm": 0.65234375,
+      "learning_rate": 6.082335722974435e-05,
+      "loss": 1.1042577028274536,
+      "step": 1963
+    },
+    {
+      "epoch": 0.4278905945043074,
+      "grad_norm": 0.6875,
+      "learning_rate": 6.0804194491158987e-05,
+      "loss": 1.2182810306549072,
+      "step": 1964
+    },
+    {
+      "epoch": 0.4281084614057862,
+      "grad_norm": 0.66015625,
+      "learning_rate": 6.078502520508753e-05,
+      "loss": 1.0657992362976074,
+      "step": 1965
+    },
+    {
+      "epoch": 0.42832632830726497,
+      "grad_norm": 0.6796875,
+      "learning_rate": 6.0765849377562925e-05,
+      "loss": 1.096466302871704,
+      "step": 1966
+    },
+    {
+      "epoch": 0.4285441952087437,
+      "grad_norm": 0.70703125,
+      "learning_rate": 6.0746667014620163e-05,
+      "loss": 1.2402524948120117,
+      "step": 1967
+    },
+    {
+      "epoch": 0.4287620621102225,
+      "grad_norm": 0.640625,
+      "learning_rate": 6.072747812229633e-05,
+      "loss": 0.9342478513717651,
+      "step": 1968
+    },
+    {
+      "epoch": 0.4289799290117013,
+      "grad_norm": 0.703125,
+      "learning_rate": 6.070828270663053e-05,
+      "loss": 1.2121024131774902,
+      "step": 1969
+    },
+    {
+      "epoch": 0.42919779591318,
+      "grad_norm": 0.66015625,
+      "learning_rate": 6.068908077366394e-05,
+      "loss": 1.1286330223083496,
+      "step": 1970
+    },
+    {
+      "epoch": 0.4294156628146588,
+      "grad_norm": 0.69140625,
+      "learning_rate": 6.066987232943979e-05,
+      "loss": 1.2695151567459106,
+      "step": 1971
+    },
+    {
+      "epoch": 0.4296335297161376,
+      "grad_norm": 0.6640625,
+      "learning_rate": 6.065065738000333e-05,
+      "loss": 1.1114144325256348,
+      "step": 1972
+    },
+    {
+      "epoch": 0.42985139661761634,
+      "grad_norm": 0.7109375,
+      "learning_rate": 6.06314359314019e-05,
+      "loss": 1.2311456203460693,
+      "step": 1973
+    },
+    {
+      "epoch": 0.43006926351909514,
+      "grad_norm": 0.61328125,
+      "learning_rate": 6.061220798968487e-05,
+      "loss": 0.9795138239860535,
+      "step": 1974
+    },
+    {
+      "epoch": 0.43028713042057387,
+      "grad_norm": 0.63671875,
+      "learning_rate": 6.059297356090363e-05,
+      "loss": 1.0935685634613037,
+      "step": 1975
+    },
+    {
+      "epoch": 0.43050499732205266,
+      "grad_norm": 0.6484375,
+      "learning_rate": 6.057373265111164e-05,
+      "loss": 1.0656893253326416,
+      "step": 1976
+    },
+    {
+      "epoch": 0.43072286422353145,
+      "grad_norm": 0.65234375,
+      "learning_rate": 6.055448526636438e-05,
+      "loss": 1.0138572454452515,
+      "step": 1977
+    },
+    {
+      "epoch": 0.4309407311250102,
+      "grad_norm": 0.76171875,
+      "learning_rate": 6.0535231412719407e-05,
+      "loss": 0.9483062028884888,
+      "step": 1978
+    },
+    {
+      "epoch": 0.431158598026489,
+      "grad_norm": 0.6953125,
+      "learning_rate": 6.0515971096236253e-05,
+      "loss": 1.0323946475982666,
+      "step": 1979
+    },
+    {
+      "epoch": 0.4313764649279678,
+      "grad_norm": 0.6796875,
+      "learning_rate": 6.049670432297653e-05,
+      "loss": 1.001125693321228,
+      "step": 1980
+    },
+    {
+      "epoch": 0.4315943318294465,
+      "grad_norm": 0.72265625,
+      "learning_rate": 6.047743109900387e-05,
+      "loss": 1.1067814826965332,
+      "step": 1981
+    },
+    {
+      "epoch": 0.4318121987309253,
+      "grad_norm": 0.65234375,
+      "learning_rate": 6.04581514303839e-05,
+      "loss": 1.1367050409317017,
+      "step": 1982
+    },
+    {
+      "epoch": 0.4320300656324041,
+      "grad_norm": 0.63671875,
+      "learning_rate": 6.0438865323184356e-05,
+      "loss": 1.1320621967315674,
+      "step": 1983
+    },
+    {
+      "epoch": 0.43224793253388283,
+      "grad_norm": 0.6328125,
+      "learning_rate": 6.041957278347493e-05,
+      "loss": 0.984647274017334,
+      "step": 1984
+    },
+    {
+      "epoch": 0.4324657994353616,
+      "grad_norm": 0.68359375,
+      "learning_rate": 6.040027381732735e-05,
+      "loss": 1.0180809497833252,
+      "step": 1985
+    },
+    {
+      "epoch": 0.4326836663368404,
+      "grad_norm": 0.671875,
+      "learning_rate": 6.0380968430815407e-05,
+      "loss": 1.1544015407562256,
+      "step": 1986
+    },
+    {
+      "epoch": 0.43290153323831915,
+      "grad_norm": 0.6328125,
+      "learning_rate": 6.036165663001486e-05,
+      "loss": 1.0215165615081787,
+      "step": 1987
+    },
+    {
+      "epoch": 0.43311940013979794,
+      "grad_norm": 0.69140625,
+      "learning_rate": 6.0342338421003514e-05,
+      "loss": 1.1294870376586914,
+      "step": 1988
+    },
+    {
+      "epoch": 0.4333372670412767,
+      "grad_norm": 0.73828125,
+      "learning_rate": 6.0323013809861185e-05,
+      "loss": 1.079067349433899,
+      "step": 1989
+    },
+    {
+      "epoch": 0.43355513394275547,
+      "grad_norm": 0.6796875,
+      "learning_rate": 6.030368280266971e-05,
+      "loss": 1.051815390586853,
+      "step": 1990
+    },
+    {
+      "epoch": 0.43377300084423426,
+      "grad_norm": 0.6171875,
+      "learning_rate": 6.028434540551293e-05,
+      "loss": 0.9883132576942444,
+      "step": 1991
+    },
+    {
+      "epoch": 0.433990867745713,
+      "grad_norm": 0.6328125,
+      "learning_rate": 6.02650016244767e-05,
+      "loss": 1.0192313194274902,
+      "step": 1992
+    },
+    {
+      "epoch": 0.4342087346471918,
+      "grad_norm": 0.62890625,
+      "learning_rate": 6.0245651465648906e-05,
+      "loss": 1.070325493812561,
+      "step": 1993
+    },
+    {
+      "epoch": 0.4344266015486706,
+      "grad_norm": 0.640625,
+      "learning_rate": 6.022629493511939e-05,
+      "loss": 0.9959108829498291,
+      "step": 1994
+    },
+    {
+      "epoch": 0.4346444684501493,
+      "grad_norm": 0.65234375,
+      "learning_rate": 6.020693203898005e-05,
+      "loss": 1.1533362865447998,
+      "step": 1995
+    },
+    {
+      "epoch": 0.4348623353516281,
+      "grad_norm": 0.63671875,
+      "learning_rate": 6.0187562783324756e-05,
+      "loss": 1.128419041633606,
+      "step": 1996
+    },
+    {
+      "epoch": 0.4350802022531069,
+      "grad_norm": 0.66015625,
+      "learning_rate": 6.016818717424939e-05,
+      "loss": 1.2142513990402222,
+      "step": 1997
+    },
+    {
+      "epoch": 0.43529806915458563,
+      "grad_norm": 0.64453125,
+      "learning_rate": 6.014880521785186e-05,
+      "loss": 1.1398576498031616,
+      "step": 1998
+    },
+    {
+      "epoch": 0.4355159360560644,
+      "grad_norm": 0.66015625,
+      "learning_rate": 6.0129416920232014e-05,
+      "loss": 1.2109894752502441,
+      "step": 1999
+    },
+    {
+      "epoch": 0.4357338029575432,
+      "grad_norm": 0.703125,
+      "learning_rate": 6.011002228749174e-05,
+      "loss": 1.2901973724365234,
+      "step": 2000
+    },
+    {
+      "epoch": 0.43595166985902195,
+      "grad_norm": 0.671875,
+      "learning_rate": 6.00906213257349e-05,
+      "loss": 1.145534634590149,
+      "step": 2001
+    },
+    {
+      "epoch": 0.43616953676050074,
+      "grad_norm": 0.625,
+      "learning_rate": 6.0071214041067376e-05,
+      "loss": 1.0444122552871704,
+      "step": 2002
+    },
+    {
+      "epoch": 0.4363874036619795,
+      "grad_norm": 0.65625,
+      "learning_rate": 6.005180043959701e-05,
+      "loss": 1.1132267713546753,
+      "step": 2003
+    },
+    {
+      "epoch": 0.43660527056345827,
+      "grad_norm": 0.64453125,
+      "learning_rate": 6.003238052743363e-05,
+      "loss": 1.077405333518982,
+      "step": 2004
+    },
+    {
+      "epoch": 0.43682313746493706,
+      "grad_norm": 0.62109375,
+      "learning_rate": 6.001295431068906e-05,
+      "loss": 1.067704439163208,
+      "step": 2005
+    },
+    {
+      "epoch": 0.4370410043664158,
+      "grad_norm": 0.65625,
+      "learning_rate": 5.999352179547713e-05,
+      "loss": 1.0475717782974243,
+      "step": 2006
+    },
+    {
+      "epoch": 0.4372588712678946,
+      "grad_norm": 0.609375,
+      "learning_rate": 5.997408298791362e-05,
+      "loss": 1.0155912637710571,
+      "step": 2007
+    },
+    {
+      "epoch": 0.4374767381693734,
+      "grad_norm": 0.671875,
+      "learning_rate": 5.99546378941163e-05,
+      "loss": 1.1573059558868408,
+      "step": 2008
+    },
+    {
+      "epoch": 0.4376946050708521,
+      "grad_norm": 0.63671875,
+      "learning_rate": 5.993518652020492e-05,
+      "loss": 0.9456545114517212,
+      "step": 2009
+    },
+    {
+      "epoch": 0.4379124719723309,
+      "grad_norm": 0.6640625,
+      "learning_rate": 5.991572887230122e-05,
+      "loss": 1.0683928728103638,
+      "step": 2010
+    },
+    {
+      "epoch": 0.4381303388738097,
+      "grad_norm": 0.71484375,
+      "learning_rate": 5.989626495652888e-05,
+      "loss": 1.11894690990448,
+      "step": 2011
+    },
+    {
+      "epoch": 0.43834820577528844,
+      "grad_norm": 0.6484375,
+      "learning_rate": 5.98767947790136e-05,
+      "loss": 1.0579040050506592,
+      "step": 2012
+    },
+    {
+      "epoch": 0.4385660726767672,
+      "grad_norm": 0.62890625,
+      "learning_rate": 5.985731834588301e-05,
+      "loss": 1.054296612739563,
+      "step": 2013
+    },
+    {
+      "epoch": 0.43878393957824596,
+      "grad_norm": 0.6953125,
+      "learning_rate": 5.9837835663266724e-05,
+      "loss": 1.02498197555542,
+      "step": 2014
+    },
+    {
+      "epoch": 0.43900180647972475,
+      "grad_norm": 0.63671875,
+      "learning_rate": 5.981834673729631e-05,
+      "loss": 1.036549687385559,
+      "step": 2015
+    },
+    {
+      "epoch": 0.43921967338120355,
+      "grad_norm": 0.6171875,
+      "learning_rate": 5.979885157410533e-05,
+      "loss": 1.0050774812698364,
+      "step": 2016
+    },
+    {
+      "epoch": 0.4394375402826823,
+      "grad_norm": 0.70703125,
+      "learning_rate": 5.97793501798293e-05,
+      "loss": 1.0556309223175049,
+      "step": 2017
+    },
+    {
+      "epoch": 0.4396554071841611,
+      "grad_norm": 0.71484375,
+      "learning_rate": 5.975984256060567e-05,
+      "loss": 1.0926799774169922,
+      "step": 2018
+    },
+    {
+      "epoch": 0.43987327408563986,
+      "grad_norm": 0.6484375,
+      "learning_rate": 5.974032872257388e-05,
+      "loss": 0.9672869443893433,
+      "step": 2019
+    },
+    {
+      "epoch": 0.4400911409871186,
+      "grad_norm": 0.66796875,
+      "learning_rate": 5.972080867187531e-05,
+      "loss": 1.1118661165237427,
+      "step": 2020
+    },
+    {
+      "epoch": 0.4403090078885974,
+      "grad_norm": 0.64453125,
+      "learning_rate": 5.9701282414653305e-05,
+      "loss": 1.0718692541122437,
+      "step": 2021
+    },
+    {
+      "epoch": 0.4405268747900762,
+      "grad_norm": 0.72265625,
+      "learning_rate": 5.9681749957053165e-05,
+      "loss": 1.0343785285949707,
+      "step": 2022
+    },
+    {
+      "epoch": 0.4407447416915549,
+      "grad_norm": 0.63671875,
+      "learning_rate": 5.966221130522213e-05,
+      "loss": 1.047936201095581,
+      "step": 2023
+    },
+    {
+      "epoch": 0.4409626085930337,
+      "grad_norm": 0.63671875,
+      "learning_rate": 5.964266646530938e-05,
+      "loss": 1.0070019960403442,
+      "step": 2024
+    },
+    {
+      "epoch": 0.4411804754945125,
+      "grad_norm": 0.6015625,
+      "learning_rate": 5.962311544346608e-05,
+      "loss": 0.859718918800354,
+      "step": 2025
+    },
+    {
+      "epoch": 0.44139834239599124,
+      "grad_norm": 0.8125,
+      "learning_rate": 5.960355824584531e-05,
+      "loss": 1.1306184530258179,
+      "step": 2026
+    },
+    {
+      "epoch": 0.44161620929747003,
+      "grad_norm": 0.6015625,
+      "learning_rate": 5.95839948786021e-05,
+      "loss": 0.9593235850334167,
+      "step": 2027
+    },
+    {
+      "epoch": 0.44183407619894877,
+      "grad_norm": 0.61328125,
+      "learning_rate": 5.956442534789343e-05,
+      "loss": 1.0026582479476929,
+      "step": 2028
+    },
+    {
+      "epoch": 0.44205194310042756,
+      "grad_norm": 0.66015625,
+      "learning_rate": 5.9544849659878206e-05,
+      "loss": 1.1355586051940918,
+      "step": 2029
+    },
+    {
+      "epoch": 0.44226981000190635,
+      "grad_norm": 0.62109375,
+      "learning_rate": 5.9525267820717274e-05,
+      "loss": 0.9556364417076111,
+      "step": 2030
+    },
+    {
+      "epoch": 0.4424876769033851,
+      "grad_norm": 0.640625,
+      "learning_rate": 5.950567983657344e-05,
+      "loss": 1.0515414476394653,
+      "step": 2031
+    },
+    {
+      "epoch": 0.4427055438048639,
+      "grad_norm": 0.62890625,
+      "learning_rate": 5.9486085713611406e-05,
+      "loss": 1.071217656135559,
+      "step": 2032
+    },
+    {
+      "epoch": 0.44292341070634267,
+      "grad_norm": 0.65234375,
+      "learning_rate": 5.946648545799784e-05,
+      "loss": 1.088335633277893,
+      "step": 2033
+    },
+    {
+      "epoch": 0.4431412776078214,
+      "grad_norm": 0.59765625,
+      "learning_rate": 5.9446879075901305e-05,
+      "loss": 0.9300054907798767,
+      "step": 2034
+    },
+    {
+      "epoch": 0.4433591445093002,
+      "grad_norm": 0.60546875,
+      "learning_rate": 5.942726657349234e-05,
+      "loss": 1.0635448694229126,
+      "step": 2035
+    },
+    {
+      "epoch": 0.443577011410779,
+      "grad_norm": 0.6640625,
+      "learning_rate": 5.940764795694335e-05,
+      "loss": 1.205702543258667,
+      "step": 2036
+    },
+    {
+      "epoch": 0.4437948783122577,
+      "grad_norm": 0.6953125,
+      "learning_rate": 5.938802323242874e-05,
+      "loss": 1.1725873947143555,
+      "step": 2037
+    },
+    {
+      "epoch": 0.4440127452137365,
+      "grad_norm": 0.66015625,
+      "learning_rate": 5.936839240612474e-05,
+      "loss": 1.1911345720291138,
+      "step": 2038
+    },
+    {
+      "epoch": 0.4442306121152153,
+      "grad_norm": 0.6875,
+      "learning_rate": 5.9348755484209597e-05,
+      "loss": 1.2115331888198853,
+      "step": 2039
+    },
+    {
+      "epoch": 0.44444847901669404,
+      "grad_norm": 0.62890625,
+      "learning_rate": 5.932911247286342e-05,
+      "loss": 0.9917274713516235,
+      "step": 2040
+    },
+    {
+      "epoch": 0.44466634591817283,
+      "grad_norm": 0.67578125,
+      "learning_rate": 5.930946337826824e-05,
+      "loss": 1.095950961112976,
+      "step": 2041
+    },
+    {
+      "epoch": 0.44488421281965157,
+      "grad_norm": 0.6328125,
+      "learning_rate": 5.928980820660801e-05,
+      "loss": 1.0243055820465088,
+      "step": 2042
+    },
+    {
+      "epoch": 0.44510207972113036,
+      "grad_norm": 0.66015625,
+      "learning_rate": 5.9270146964068614e-05,
+      "loss": 0.9234710335731506,
+      "step": 2043
+    },
+    {
+      "epoch": 0.44531994662260915,
+      "grad_norm": 0.61328125,
+      "learning_rate": 5.925047965683782e-05,
+      "loss": 0.8920939564704895,
+      "step": 2044
+    },
+    {
+      "epoch": 0.4455378135240879,
+      "grad_norm": 0.66796875,
+      "learning_rate": 5.92308062911053e-05,
+      "loss": 1.1351994276046753,
+      "step": 2045
+    },
+    {
+      "epoch": 0.4457556804255667,
+      "grad_norm": 0.69140625,
+      "learning_rate": 5.9211126873062674e-05,
+      "loss": 1.0701953172683716,
+      "step": 2046
+    },
+    {
+      "epoch": 0.44597354732704547,
+      "grad_norm": 0.61328125,
+      "learning_rate": 5.9191441408903416e-05,
+      "loss": 0.8634197115898132,
+      "step": 2047
+    },
+    {
+      "epoch": 0.4461914142285242,
+      "grad_norm": 0.6796875,
+      "learning_rate": 5.9171749904822924e-05,
+      "loss": 1.010629415512085,
+      "step": 2048
+    },
+    {
+      "epoch": 0.446409281130003,
+      "grad_norm": 0.65625,
+      "learning_rate": 5.9152052367018517e-05,
+      "loss": 0.9668979048728943,
+      "step": 2049
+    },
+    {
+      "epoch": 0.4466271480314818,
+      "grad_norm": 0.6640625,
+      "learning_rate": 5.9132348801689384e-05,
+      "loss": 1.1860495805740356,
+      "step": 2050
+    },
+    {
+      "epoch": 0.4468450149329605,
+      "grad_norm": 0.68359375,
+      "learning_rate": 5.9112639215036634e-05,
+      "loss": 1.1523436307907104,
+      "step": 2051
+    },
+    {
+      "epoch": 0.4470628818344393,
+      "grad_norm": 0.671875,
+      "learning_rate": 5.909292361326322e-05,
+      "loss": 1.1051534414291382,
+      "step": 2052
+    },
+    {
+      "epoch": 0.44728074873591805,
+      "grad_norm": 0.69921875,
+      "learning_rate": 5.907320200257407e-05,
+      "loss": 0.9340056777000427,
+      "step": 2053
+    },
+    {
+      "epoch": 0.44749861563739685,
+      "grad_norm": 0.6328125,
+      "learning_rate": 5.9053474389175944e-05,
+      "loss": 1.1073839664459229,
+      "step": 2054
+    },
+    {
+      "epoch": 0.44771648253887564,
+      "grad_norm": 0.671875,
+      "learning_rate": 5.90337407792775e-05,
+      "loss": 1.1355162858963013,
+      "step": 2055
+    },
+    {
+      "epoch": 0.4479343494403544,
+      "grad_norm": 0.68359375,
+      "learning_rate": 5.901400117908931e-05,
+      "loss": 0.9775420427322388,
+      "step": 2056
+    },
+    {
+      "epoch": 0.44815221634183317,
+      "grad_norm": 0.72265625,
+      "learning_rate": 5.899425559482378e-05,
+      "loss": 1.0651700496673584,
+      "step": 2057
+    },
+    {
+      "epoch": 0.44837008324331196,
+      "grad_norm": 0.67578125,
+      "learning_rate": 5.897450403269527e-05,
+      "loss": 1.0329807996749878,
+      "step": 2058
+    },
+    {
+      "epoch": 0.4485879501447907,
+      "grad_norm": 0.64453125,
+      "learning_rate": 5.895474649891995e-05,
+      "loss": 1.0401101112365723,
+      "step": 2059
+    },
+    {
+      "epoch": 0.4488058170462695,
+      "grad_norm": 0.69140625,
+      "learning_rate": 5.893498299971592e-05,
+      "loss": 1.2177573442459106,
+      "step": 2060
+    },
+    {
+      "epoch": 0.4490236839477483,
+      "grad_norm": 0.70703125,
+      "learning_rate": 5.891521354130313e-05,
+      "loss": 1.193304181098938,
+      "step": 2061
+    },
+    {
+      "epoch": 0.449241550849227,
+      "grad_norm": 0.58984375,
+      "learning_rate": 5.889543812990341e-05,
+      "loss": 0.9251775145530701,
+      "step": 2062
+    },
+    {
+      "epoch": 0.4494594177507058,
+      "grad_norm": 0.6484375,
+      "learning_rate": 5.887565677174048e-05,
+      "loss": 1.074304461479187,
+      "step": 2063
+    },
+    {
+      "epoch": 0.4496772846521846,
+      "grad_norm": 0.6640625,
+      "learning_rate": 5.8855869473039924e-05,
+      "loss": 1.0762293338775635,
+      "step": 2064
+    },
+    {
+      "epoch": 0.44989515155366333,
+      "grad_norm": 0.66796875,
+      "learning_rate": 5.883607624002915e-05,
+      "loss": 1.019173502922058,
+      "step": 2065
+    },
+    {
+      "epoch": 0.4501130184551421,
+      "grad_norm": 0.62890625,
+      "learning_rate": 5.881627707893754e-05,
+      "loss": 0.9566801190376282,
+      "step": 2066
+    },
+    {
+      "epoch": 0.45033088535662086,
+      "grad_norm": 0.6484375,
+      "learning_rate": 5.879647199599622e-05,
+      "loss": 1.1010868549346924,
+      "step": 2067
+    },
+    {
+      "epoch": 0.45054875225809965,
+      "grad_norm": 0.67578125,
+      "learning_rate": 5.877666099743826e-05,
+      "loss": 1.0144315958023071,
+      "step": 2068
+    },
+    {
+      "epoch": 0.45076661915957844,
+      "grad_norm": 0.63671875,
+      "learning_rate": 5.875684408949858e-05,
+      "loss": 1.0331202745437622,
+      "step": 2069
+    },
+    {
+      "epoch": 0.4509844860610572,
+      "grad_norm": 0.64453125,
+      "learning_rate": 5.873702127841394e-05,
+      "loss": 1.0669623613357544,
+      "step": 2070
+    },
+    {
+      "epoch": 0.45120235296253597,
+      "grad_norm": 0.64453125,
+      "learning_rate": 5.871719257042294e-05,
+      "loss": 1.055669903755188,
+      "step": 2071
+    },
+    {
+      "epoch": 0.45142021986401476,
+      "grad_norm": 0.62890625,
+      "learning_rate": 5.869735797176609e-05,
+      "loss": 0.9666656851768494,
+      "step": 2072
+    },
+    {
+      "epoch": 0.4516380867654935,
+      "grad_norm": 0.63671875,
+      "learning_rate": 5.867751748868572e-05,
+      "loss": 0.9973516464233398,
+      "step": 2073
+    },
+    {
+      "epoch": 0.4518559536669723,
+      "grad_norm": 0.609375,
+      "learning_rate": 5.865767112742602e-05,
+      "loss": 0.8883056044578552,
+      "step": 2074
+    },
+    {
+      "epoch": 0.4517826363059058,
+      "grad_norm": 0.64453125,
+      "learning_rate": 5.8637818894232996e-05,
+      "loss": 1.1140676736831665,
+      "step": 2075
+    },
+    {
+      "epoch": 0.4520003628776195,
+      "grad_norm": 0.6640625,
+      "learning_rate": 5.864541473152664e-05,
+      "loss": 1.1264996528625488,
+      "step": 2076
+    },
+    {
+      "epoch": 0.4522180894493332,
+      "grad_norm": 0.69921875,
+      "learning_rate": 5.862557305074869e-05,
+      "loss": 1.2169564962387085,
+      "step": 2077
+    },
+    {
+      "epoch": 0.4524358160210469,
+      "grad_norm": 0.62109375,
+      "learning_rate": 5.860572551650442e-05,
+      "loss": 0.9224720597267151,
+      "step": 2078
+    },
+    {
+      "epoch": 0.4526535425927606,
+      "grad_norm": 0.64453125,
+      "learning_rate": 5.85858721350313e-05,
+      "loss": 1.043272614479065,
+      "step": 2079
+    },
+    {
+      "epoch": 0.4528712691644743,
+      "grad_norm": 0.62109375,
+      "learning_rate": 5.8566012912568694e-05,
+      "loss": 1.0768539905548096,
+      "step": 2080
+    },
+    {
+      "epoch": 0.453088995736188,
+      "grad_norm": 0.6796875,
+      "learning_rate": 5.8546147855357736e-05,
+      "loss": 1.0749694108963013,
+      "step": 2081
+    },
+    {
+      "epoch": 0.45330672230790164,
+      "grad_norm": 0.65625,
+      "learning_rate": 5.852627696964144e-05,
+      "loss": 1.0232985019683838,
+      "step": 2082
+    },
+    {
+      "epoch": 0.45352444887961535,
+      "grad_norm": 0.6640625,
+      "learning_rate": 5.8506400261664645e-05,
+      "loss": 1.122044563293457,
+      "step": 2083
+    },
+    {
+      "epoch": 0.45374217545132906,
+      "grad_norm": 0.6484375,
+      "learning_rate": 5.8486517737674004e-05,
+      "loss": 1.1187975406646729,
+      "step": 2084
+    },
+    {
+      "epoch": 0.4539599020230427,
+      "grad_norm": 0.59375,
+      "learning_rate": 5.8466629403918e-05,
+      "loss": 0.9396539330482483,
+      "step": 2085
+    },
+    {
+      "epoch": 0.4541776285947564,
+      "grad_norm": 0.6484375,
+      "learning_rate": 5.8446735266646954e-05,
+      "loss": 0.9884230494499207,
+      "step": 2086
+    },
+    {
+      "epoch": 0.4543953551664701,
+      "grad_norm": 0.62890625,
+      "learning_rate": 5.8426835332113e-05,
+      "loss": 0.9224221110343933,
+      "step": 2087
+    },
+    {
+      "epoch": 0.4546130817381838,
+      "grad_norm": 0.65625,
+      "learning_rate": 5.840692960657011e-05,
+      "loss": 1.1308064460754395,
+      "step": 2088
+    },
+    {
+      "epoch": 0.45483080830989747,
+      "grad_norm": 0.640625,
+      "learning_rate": 5.838701809627406e-05,
+      "loss": 1.08757483959198,
+      "step": 2089
+    },
+    {
+      "epoch": 0.4550485348816112,
+      "grad_norm": 0.6640625,
+      "learning_rate": 5.8367100807482444e-05,
+      "loss": 1.0933618545532227,
+      "step": 2090
+    },
+    {
+      "epoch": 0.4552662614533249,
+      "grad_norm": 0.7734375,
+      "learning_rate": 5.8347177746454676e-05,
+      "loss": 1.0601389408111572,
+      "step": 2091
+    },
+    {
+      "epoch": 0.45548398802503853,
+      "grad_norm": 0.6484375,
+      "learning_rate": 5.832724891945197e-05,
+      "loss": 1.0301694869995117,
+      "step": 2092
+    },
+    {
+      "epoch": 0.45570171459675224,
+      "grad_norm": 0.6875,
+      "learning_rate": 5.83073143327374e-05,
+      "loss": 0.9762527942657471,
+      "step": 2093
+    },
+    {
+      "epoch": 0.45591944116846594,
+      "grad_norm": 0.67578125,
+      "learning_rate": 5.82873739925758e-05,
+      "loss": 1.2493689060211182,
+      "step": 2094
+    },
+    {
+      "epoch": 0.45613716774017965,
+      "grad_norm": 0.74609375,
+      "learning_rate": 5.826742790523382e-05,
+      "loss": 0.9570315480232239,
+      "step": 2095
+    },
+    {
+      "epoch": 0.4563548943118933,
+      "grad_norm": 0.65234375,
+      "learning_rate": 5.824747607697992e-05,
+      "loss": 1.0393177270889282,
+      "step": 2096
+    },
+    {
+      "epoch": 0.456572620883607,
+      "grad_norm": 0.69921875,
+      "learning_rate": 5.8227518514084415e-05,
+      "loss": 1.1568523645401,
+      "step": 2097
+    },
+    {
+      "epoch": 0.4567903474553207,
+      "grad_norm": 0.7890625,
+      "learning_rate": 5.820755522281932e-05,
+      "loss": 1.065783977508545,
+      "step": 2098
+    },
+    {
+      "epoch": 0.45700807402703436,
+      "grad_norm": 0.6796875,
+      "learning_rate": 5.8187586209458536e-05,
+      "loss": 0.996212363243103,
+      "step": 2099
+    },
+    {
+      "epoch": 0.45722580059874807,
+      "grad_norm": 0.6875,
+      "learning_rate": 5.8167611480277715e-05,
+      "loss": 1.1269118785858154,
+      "step": 2100
+    },
+    {
+      "epoch": 0.4574435271704618,
+      "grad_norm": 0.6484375,
+      "learning_rate": 5.814763104155434e-05,
+      "loss": 1.0951814651489258,
+      "step": 2101
+    },
+    {
+      "epoch": 0.4576612537421755,
+      "grad_norm": 0.62890625,
+      "learning_rate": 5.8127644899567675e-05,
+      "loss": 1.0418212413787842,
+      "step": 2102
+    },
+    {
+      "epoch": 0.45787898031388913,
+      "grad_norm": 0.66015625,
+      "learning_rate": 5.8107653060598754e-05,
+      "loss": 0.924680769443512,
+      "step": 2103
+    },
+    {
+      "epoch": 0.45809670688560283,
+      "grad_norm": 0.65625,
+      "learning_rate": 5.808765553093044e-05,
+      "loss": 1.129743218421936,
+      "step": 2104
+    },
+    {
+      "epoch": 0.45831443345731654,
+      "grad_norm": 0.59765625,
+      "learning_rate": 5.8067652316847346e-05,
+      "loss": 0.8876528143882751,
+      "step": 2105
+    },
+    {
+      "epoch": 0.4585321600290302,
+      "grad_norm": 0.66015625,
+      "learning_rate": 5.80476434246359e-05,
+      "loss": 1.0566883087158203,
+      "step": 2106
+    },
+    {
+      "epoch": 0.4587498866007439,
+      "grad_norm": 0.70703125,
+      "learning_rate": 5.80276288605843e-05,
+      "loss": 0.9911318421363831,
+      "step": 2107
+    },
+    {
+      "epoch": 0.4589676131724576,
+      "grad_norm": 0.6796875,
+      "learning_rate": 5.800760863098254e-05,
+      "loss": 1.0184762477874756,
+      "step": 2108
+    },
+    {
+      "epoch": 0.4591853397441713,
+      "grad_norm": 0.66015625,
+      "learning_rate": 5.7987582742122385e-05,
+      "loss": 1.0505367517471313,
+      "step": 2109
+    },
+    {
+      "epoch": 0.45940306631588496,
+      "grad_norm": 0.66015625,
+      "learning_rate": 5.796755120029737e-05,
+      "loss": 1.1524643898010254,
+      "step": 2110
+    },
+    {
+      "epoch": 0.45962079288759866,
+      "grad_norm": 0.625,
+      "learning_rate": 5.794751401180282e-05,
+      "loss": 0.9461578726768494,
+      "step": 2111
+    },
+    {
+      "epoch": 0.45983851945931237,
+      "grad_norm": 0.72265625,
+      "learning_rate": 5.792747118293582e-05,
+      "loss": 1.0850226879119873,
+      "step": 2112
+    },
+    {
+      "epoch": 0.460056246031026,
+      "grad_norm": 0.70703125,
+      "learning_rate": 5.790742271999527e-05,
+      "loss": 1.134536623954773,
+      "step": 2113
+    },
+    {
+      "epoch": 0.4602739726027397,
+      "grad_norm": 0.671875,
+      "learning_rate": 5.788736862928177e-05,
+      "loss": 1.0808788537979126,
+      "step": 2114
+    },
+    {
+      "epoch": 0.46049169917445343,
+      "grad_norm": 0.765625,
+      "learning_rate": 5.786730891709775e-05,
+      "loss": 1.0658330917358398,
+      "step": 2115
+    },
+    {
+      "epoch": 0.4607094257461671,
+      "grad_norm": 0.875,
+      "learning_rate": 5.7847243589747376e-05,
+      "loss": 1.0823721885681152,
+      "step": 2116
+    },
+    {
+      "epoch": 0.4609271523178808,
+      "grad_norm": 0.703125,
+      "learning_rate": 5.782717265353658e-05,
+      "loss": 1.0344072580337524,
+      "step": 2117
+    },
+    {
+      "epoch": 0.4611448788895945,
+      "grad_norm": 0.68359375,
+      "learning_rate": 5.780709611477308e-05,
+      "loss": 1.1921982765197754,
+      "step": 2118
+    },
+    {
+      "epoch": 0.4613626054613082,
+      "grad_norm": 0.66796875,
+      "learning_rate": 5.7787013979766334e-05,
+      "loss": 1.1574395895004272,
+      "step": 2119
+    },
+    {
+      "epoch": 0.46158033203302185,
+      "grad_norm": 0.70703125,
+      "learning_rate": 5.776692625482755e-05,
+      "loss": 1.219463586807251,
+      "step": 2120
+    },
+    {
+      "epoch": 0.46179805860473555,
+      "grad_norm": 0.609375,
+      "learning_rate": 5.7746832946269716e-05,
+      "loss": 1.0288102626800537,
+      "step": 2121
+    },
+    {
+      "epoch": 0.46201578517644926,
+      "grad_norm": 0.671875,
+      "learning_rate": 5.772673406040756e-05,
+      "loss": 1.0775922536849976,
+      "step": 2122
+    },
+    {
+      "epoch": 0.4622335117481629,
+      "grad_norm": 0.6484375,
+      "learning_rate": 5.770662960355757e-05,
+      "loss": 0.9380577802658081,
+      "step": 2123
+    },
+    {
+      "epoch": 0.4624512383198766,
+      "grad_norm": 0.7421875,
+      "learning_rate": 5.768651958203798e-05,
+      "loss": 1.2333927154541016,
+      "step": 2124
+    },
+    {
+      "epoch": 0.4626689648915903,
+      "grad_norm": 0.62890625,
+      "learning_rate": 5.7666404002168775e-05,
+      "loss": 0.9679368138313293,
+      "step": 2125
+    },
+    {
+      "epoch": 0.462886691463304,
+      "grad_norm": 0.6484375,
+      "learning_rate": 5.764628287027168e-05,
+      "loss": 1.0177321434020996,
+      "step": 2126
+    },
+    {
+      "epoch": 0.4631044180350177,
+      "grad_norm": 0.73828125,
+      "learning_rate": 5.762615619267019e-05,
+      "loss": 1.1516188383102417,
+      "step": 2127
+    },
+    {
+      "epoch": 0.4633221446067314,
+      "grad_norm": 0.671875,
+      "learning_rate": 5.760602397568949e-05,
+      "loss": 1.1720983982086182,
+      "step": 2128
+    },
+    {
+      "epoch": 0.4635398711784451,
+      "grad_norm": 0.6328125,
+      "learning_rate": 5.758588622565658e-05,
+      "loss": 1.098969578742981,
+      "step": 2129
+    },
+    {
+      "epoch": 0.46375759775015873,
+      "grad_norm": 0.72265625,
+      "learning_rate": 5.7565742948900136e-05,
+      "loss": 1.106179118156433,
+      "step": 2130
+    },
+    {
+      "epoch": 0.46397532432187244,
+      "grad_norm": 0.671875,
+      "learning_rate": 5.75455941517506e-05,
+      "loss": 0.878695011138916,
+      "step": 2131
+    },
+    {
+      "epoch": 0.46419305089358615,
+      "grad_norm": 0.67578125,
+      "learning_rate": 5.752543984054013e-05,
+      "loss": 1.1113193035125732,
+      "step": 2132
+    },
+    {
+      "epoch": 0.46441077746529985,
+      "grad_norm": 0.6640625,
+      "learning_rate": 5.750528002160265e-05,
+      "loss": 1.0944639444351196,
+      "step": 2133
+    },
+    {
+      "epoch": 0.4646285040370135,
+      "grad_norm": 0.6953125,
+      "learning_rate": 5.748511470127377e-05,
+      "loss": 1.161794900894165,
+      "step": 2134
+    },
+    {
+      "epoch": 0.4648462306087272,
+      "grad_norm": 0.65625,
+      "learning_rate": 5.7464943885890886e-05,
+      "loss": 1.1001379489898682,
+      "step": 2135
+    },
+    {
+      "epoch": 0.4650639571804409,
+      "grad_norm": 0.66015625,
+      "learning_rate": 5.744476758179306e-05,
+      "loss": 1.2179656028747559,
+      "step": 2136
+    },
+    {
+      "epoch": 0.46528168375215456,
+      "grad_norm": 0.609375,
+      "learning_rate": 5.742458579532111e-05,
+      "loss": 0.9296327233314514,
+      "step": 2137
+    },
+    {
+      "epoch": 0.46549941032386827,
+      "grad_norm": 0.6171875,
+      "learning_rate": 5.7404398532817595e-05,
+      "loss": 1.0246578454971313,
+      "step": 2138
+    },
+    {
+      "epoch": 0.465717136895582,
+      "grad_norm": 0.640625,
+      "learning_rate": 5.7384205800626755e-05,
+      "loss": 1.109699010848999,
+      "step": 2139
+    },
+    {
+      "epoch": 0.4659348634672957,
+      "grad_norm": 0.6484375,
+      "learning_rate": 5.7364007605094575e-05,
+      "loss": 1.0902677774429321,
+      "step": 2140
+    },
+    {
+      "epoch": 0.46615259003900933,
+      "grad_norm": 0.64453125,
+      "learning_rate": 5.7343803952568746e-05,
+      "loss": 1.1264792680740356,
+      "step": 2141
+    },
+    {
+      "epoch": 0.46637031661072303,
+      "grad_norm": 0.64453125,
+      "learning_rate": 5.7323594849398674e-05,
+      "loss": 1.137683391571045,
+      "step": 2142
+    },
+    {
+      "epoch": 0.46658804318243674,
+      "grad_norm": 0.66796875,
+      "learning_rate": 5.7303380301935495e-05,
+      "loss": 1.0009865760803223,
+      "step": 2143
+    },
+    {
+      "epoch": 0.4668057697541504,
+      "grad_norm": 0.66015625,
+      "learning_rate": 5.728316031653203e-05,
+      "loss": 1.04068922996521,
+      "step": 2144
+    },
+    {
+      "epoch": 0.4670234963258641,
+      "grad_norm": 0.6328125,
+      "learning_rate": 5.726293489954284e-05,
+      "loss": 1.0169769525527954,
+      "step": 2145
+    },
+    {
+      "epoch": 0.4672412228975778,
+      "grad_norm": 0.671875,
+      "learning_rate": 5.724270405732416e-05,
+      "loss": 1.186317801475525,
+      "step": 2146
+    },
+    {
+      "epoch": 0.4674589494692915,
+      "grad_norm": 0.62109375,
+      "learning_rate": 5.7222467796233935e-05,
+      "loss": 0.9356879591941833,
+      "step": 2147
+    },
+    {
+      "epoch": 0.46767667604100516,
+      "grad_norm": 0.69140625,
+      "learning_rate": 5.720222612263184e-05,
+      "loss": 1.2047979831695557,
+      "step": 2148
+    },
+    {
+      "epoch": 0.46789440261271886,
+      "grad_norm": 0.65234375,
+      "learning_rate": 5.718197904287924e-05,
+      "loss": 1.0378475189208984,
+      "step": 2149
+    },
+    {
+      "epoch": 0.46811212918443257,
+      "grad_norm": 0.6484375,
+      "learning_rate": 5.7161726563339167e-05,
+      "loss": 1.060961365699768,
+      "step": 2150
+    },
+    {
+      "epoch": 0.4683298557561462,
+      "grad_norm": 0.671875,
+      "learning_rate": 5.71414686903764e-05,
+      "loss": 1.098639726638794,
+      "step": 2151
+    },
+    {
+      "epoch": 0.4685475823278599,
+      "grad_norm": 0.70703125,
+      "learning_rate": 5.7121205430357395e-05,
+      "loss": 1.170255184173584,
+      "step": 2152
+    },
+    {
+      "epoch": 0.46876530889957363,
+      "grad_norm": 0.671875,
+      "learning_rate": 5.7100936789650266e-05,
+      "loss": 1.0370244979858398,
+      "step": 2153
+    },
+    {
+      "epoch": 0.46898303547128734,
+      "grad_norm": 0.65625,
+      "learning_rate": 5.708066277462486e-05,
+      "loss": 1.1184885501861572,
+      "step": 2154
+    },
+    {
+      "epoch": 0.469200762043001,
+      "grad_norm": 0.68359375,
+      "learning_rate": 5.70603833916527e-05,
+      "loss": 1.1816723346710205,
+      "step": 2155
+    },
+    {
+      "epoch": 0.4694184886147147,
+      "grad_norm": 0.64453125,
+      "learning_rate": 5.7040098647107e-05,
+      "loss": 1.0014655590057373,
+      "step": 2156
+    },
+    {
+      "epoch": 0.4696362151864284,
+      "grad_norm": 0.62890625,
+      "learning_rate": 5.701980854736265e-05,
+      "loss": 1.0436506271362305,
+      "step": 2157
+    },
+    {
+      "epoch": 0.46985394175814205,
+      "grad_norm": 0.671875,
+      "learning_rate": 5.699951309879623e-05,
+      "loss": 0.9424639940261841,
+      "step": 2158
+    },
+    {
+      "epoch": 0.47007166832985575,
+      "grad_norm": 0.63671875,
+      "learning_rate": 5.6979212307785994e-05,
+      "loss": 1.03534996509552,
+      "step": 2159
+    },
+    {
+      "epoch": 0.47028939490156946,
+      "grad_norm": 0.61328125,
+      "learning_rate": 5.695890618071189e-05,
+      "loss": 1.1146299839019775,
+      "step": 2160
+    },
+    {
+      "epoch": 0.4705071214732831,
+      "grad_norm": 0.64453125,
+      "learning_rate": 5.6938594723955524e-05,
+      "loss": 1.0892068147659302,
+      "step": 2161
+    },
+    {
+      "epoch": 0.4707248480449968,
+      "grad_norm": 0.73046875,
+      "learning_rate": 5.691827794390019e-05,
+      "loss": 1.218643307685852,
+      "step": 2162
+    },
+    {
+      "epoch": 0.4709425746167105,
+      "grad_norm": 0.625,
+      "learning_rate": 5.689795584693085e-05,
+      "loss": 1.0001274347305298,
+      "step": 2163
+    },
+    {
+      "epoch": 0.4711603011884242,
+      "grad_norm": 0.67578125,
+      "learning_rate": 5.687762843943413e-05,
+      "loss": 1.1111476421356201,
+      "step": 2164
+    },
+    {
+      "epoch": 0.4713780277601379,
+      "grad_norm": 0.69921875,
+      "learning_rate": 5.6857295727798343e-05,
+      "loss": 1.0545016527175903,
+      "step": 2165
+    },
+    {
+      "epoch": 0.4715957543318516,
+      "grad_norm": 0.69140625,
+      "learning_rate": 5.683695771841346e-05,
+      "loss": 1.1070420742034912,
+      "step": 2166
+    },
+    {
+      "epoch": 0.4718134809035653,
+      "grad_norm": 0.6328125,
+      "learning_rate": 5.68166144176711e-05,
+      "loss": 1.0939453840255737,
+      "step": 2167
+    },
+    {
+      "epoch": 0.47203120747527894,
+      "grad_norm": 0.65234375,
+      "learning_rate": 5.679626583196458e-05,
+      "loss": 0.9351567625999451,
+      "step": 2168
+    },
+    {
+      "epoch": 0.47224893404699264,
+      "grad_norm": 0.6640625,
+      "learning_rate": 5.6775911967688836e-05,
+      "loss": 1.0471183061599731,
+      "step": 2169
+    },
+    {
+      "epoch": 0.47246666061870635,
+      "grad_norm": 0.69140625,
+      "learning_rate": 5.6755552831240504e-05,
+      "loss": 1.1822692155838013,
+      "step": 2170
+    },
+    {
+      "epoch": 0.47268438719042005,
+      "grad_norm": 0.6328125,
+      "learning_rate": 5.673518842901786e-05,
+      "loss": 1.0347946882247925,
+      "step": 2171
+    },
+    {
+      "epoch": 0.4729021137621337,
+      "grad_norm": 0.625,
+      "learning_rate": 5.671481876742081e-05,
+      "loss": 1.0629960298538208,
+      "step": 2172
+    },
+    {
+      "epoch": 0.4731198403338474,
+      "grad_norm": 0.63671875,
+      "learning_rate": 5.669444385285096e-05,
+      "loss": 0.8675985336303711,
+      "step": 2173
+    },
+    {
+      "epoch": 0.4733375669055611,
+      "grad_norm": 0.6953125,
+      "learning_rate": 5.6674063691711525e-05,
+      "loss": 1.2270151376724243,
+      "step": 2174
+    },
+    {
+      "epoch": 0.47355529347727476,
+      "grad_norm": 0.640625,
+      "learning_rate": 5.66536782904074e-05,
+      "loss": 1.0627692937850952,
+      "step": 2175
+    },
+    {
+      "epoch": 0.47377302004898847,
+      "grad_norm": 0.6953125,
+      "learning_rate": 5.6633287655345105e-05,
+      "loss": 1.0553404092788696,
+      "step": 2176
+    },
+    {
+      "epoch": 0.4739907466207022,
+      "grad_norm": 0.60546875,
+      "learning_rate": 5.6612891792932814e-05,
+      "loss": 0.9436278939247131,
+      "step": 2177
+    },
+    {
+      "epoch": 0.4742084731924159,
+      "grad_norm": 0.61328125,
+      "learning_rate": 5.659249070958035e-05,
+      "loss": 1.0463228225708008,
+      "step": 2178
+    },
+    {
+      "epoch": 0.47442619976412953,
+      "grad_norm": 0.640625,
+      "learning_rate": 5.657208441169916e-05,
+      "loss": 0.9345570206642151,
+      "step": 2179
+    },
+    {
+      "epoch": 0.47464392633584324,
+      "grad_norm": 0.875,
+      "learning_rate": 5.655167290570234e-05,
+      "loss": 1.0630699396133423,
+      "step": 2180
+    },
+    {
+      "epoch": 0.47486165290755694,
+      "grad_norm": 0.6328125,
+      "learning_rate": 5.653125619800463e-05,
+      "loss": 1.0004504919052124,
+      "step": 2181
+    },
+    {
+      "epoch": 0.4750793794792706,
+      "grad_norm": 0.62109375,
+      "learning_rate": 5.6510834295022396e-05,
+      "loss": 1.1501023769378662,
+      "step": 2182
+    },
+    {
+      "epoch": 0.4752971060509843,
+      "grad_norm": 0.6484375,
+      "learning_rate": 5.649040720317362e-05,
+      "loss": 1.0349427461624146,
+      "step": 2183
+    },
+    {
+      "epoch": 0.475514832622698,
+      "grad_norm": 0.6796875,
+      "learning_rate": 5.6469974928877955e-05,
+      "loss": 1.1747218370437622,
+      "step": 2184
+    },
+    {
+      "epoch": 0.4757325591944117,
+      "grad_norm": 0.640625,
+      "learning_rate": 5.644953747855666e-05,
+      "loss": 1.0315241813659668,
+      "step": 2185
+    },
+    {
+      "epoch": 0.47595028576612536,
+      "grad_norm": 0.66015625,
+      "learning_rate": 5.64290948586326e-05,
+      "loss": 1.2199997901916504,
+      "step": 2186
+    },
+    {
+      "epoch": 0.47616801233783906,
+      "grad_norm": 0.63671875,
+      "learning_rate": 5.64086470755303e-05,
+      "loss": 0.9179491996765137,
+      "step": 2187
+    },
+    {
+      "epoch": 0.47638573890955277,
+      "grad_norm": 0.66015625,
+      "learning_rate": 5.63881941356759e-05,
+      "loss": 1.1007548570632935,
+      "step": 2188
+    },
+    {
+      "epoch": 0.4766034654812664,
+      "grad_norm": 0.64453125,
+      "learning_rate": 5.6367736045497125e-05,
+      "loss": 1.0421907901763916,
+      "step": 2189
+    },
+    {
+      "epoch": 0.4768211920529801,
+      "grad_norm": 0.640625,
+      "learning_rate": 5.634727281142339e-05,
+      "loss": 1.0879168510437012,
+      "step": 2190
+    },
+    {
+      "epoch": 0.47703891862469383,
+      "grad_norm": 0.61328125,
+      "learning_rate": 5.6326804439885645e-05,
+      "loss": 1.022667407989502,
+      "step": 2191
+    },
+    {
+      "epoch": 0.47725664519640754,
+      "grad_norm": 0.6484375,
+      "learning_rate": 5.630633093731651e-05,
+      "loss": 1.056269645690918,
+      "step": 2192
+    },
+    {
+      "epoch": 0.4774743717681212,
+      "grad_norm": 0.6484375,
+      "learning_rate": 5.62858523101502e-05,
+      "loss": 1.013848900794983,
+      "step": 2193
+    },
+    {
+      "epoch": 0.4776920983398349,
+      "grad_norm": 0.6875,
+      "learning_rate": 5.626536856482255e-05,
+      "loss": 1.0126897096633911,
+      "step": 2194
+    },
+    {
+      "epoch": 0.4779098249115486,
+      "grad_norm": 0.69921875,
+      "learning_rate": 5.6244879707770976e-05,
+      "loss": 1.2265081405639648,
+      "step": 2195
+    },
+    {
+      "epoch": 0.47812755148326225,
+      "grad_norm": 0.63671875,
+      "learning_rate": 5.622438574543453e-05,
+      "loss": 1.0425139665603638,
+      "step": 2196
+    },
+    {
+      "epoch": 0.47834527805497595,
+      "grad_norm": 0.6171875,
+      "learning_rate": 5.6203886684253856e-05,
+      "loss": 1.028672695159912,
+      "step": 2197
+    },
+    {
+      "epoch": 0.47856300462668966,
+      "grad_norm": 0.65625,
+      "learning_rate": 5.618338253067121e-05,
+      "loss": 1.0065710544586182,
+      "step": 2198
+    },
+    {
+      "epoch": 0.47878073119840336,
+      "grad_norm": 0.66015625,
+      "learning_rate": 5.616287329113043e-05,
+      "loss": 1.096346378326416,
+      "step": 2199
+    },
+    {
+      "epoch": 0.478998457770117,
+      "grad_norm": 0.63671875,
+      "learning_rate": 5.614235897207697e-05,
+      "loss": 0.9678300023078918,
+      "step": 2200
+    },
+    {
+      "epoch": 0.4792161843418307,
+      "grad_norm": 0.62890625,
+      "learning_rate": 5.612183957995787e-05,
+      "loss": 0.9559441804885864,
+      "step": 2201
+    },
+    {
+      "epoch": 0.4794339109135444,
+      "grad_norm": 0.66796875,
+      "learning_rate": 5.610131512122175e-05,
+      "loss": 1.127706527709961,
+      "step": 2202
+    },
+    {
+      "epoch": 0.4796516374852581,
+      "grad_norm": 0.6640625,
+      "learning_rate": 5.608078560231888e-05,
+      "loss": 0.9928770661354065,
+      "step": 2203
+    },
+    {
+      "epoch": 0.4798693640569718,
+      "grad_norm": 0.7109375,
+      "learning_rate": 5.6060251029701056e-05,
+      "loss": 1.0318801403045654,
+      "step": 2204
+    },
+    {
+      "epoch": 0.4800870906286855,
+      "grad_norm": 0.6328125,
+      "learning_rate": 5.603971140982167e-05,
+      "loss": 1.1361167430877686,
+      "step": 2205
+    },
+    {
+      "epoch": 0.4803048172003992,
+      "grad_norm": 0.625,
+      "learning_rate": 5.6019166749135747e-05,
+      "loss": 0.9223031997680664,
+      "step": 2206
+    },
+    {
+      "epoch": 0.48052254377211284,
+      "grad_norm": 0.62890625,
+      "learning_rate": 5.5998617054099855e-05,
+      "loss": 0.9742498993873596,
+      "step": 2207
+    },
+    {
+      "epoch": 0.48074027034382655,
+      "grad_norm": 0.68359375,
+      "learning_rate": 5.597806233117215e-05,
+      "loss": 1.0258557796478271,
+      "step": 2208
+    },
+    {
+      "epoch": 0.48095799691554025,
+      "grad_norm": 0.63671875,
+      "learning_rate": 5.595750258681238e-05,
+      "loss": 1.2224942445755005,
+      "step": 2209
+    },
+    {
+      "epoch": 0.4811757234872539,
+      "grad_norm": 0.6484375,
+      "learning_rate": 5.593693782748185e-05,
+      "loss": 1.076530933380127,
+      "step": 2210
+    },
+    {
+      "epoch": 0.4813934500589676,
+      "grad_norm": 0.61328125,
+      "learning_rate": 5.591636805964348e-05,
+      "loss": 1.1272203922271729,
+      "step": 2211
+    },
+    {
+      "epoch": 0.4816111766306813,
+      "grad_norm": 0.6796875,
+      "learning_rate": 5.5895793289761706e-05,
+      "loss": 1.120582938194275,
+      "step": 2212
+    },
+    {
+      "epoch": 0.48182890320239496,
+      "grad_norm": 0.6015625,
+      "learning_rate": 5.5875213524302605e-05,
+      "loss": 1.0565608739852905,
+      "step": 2213
+    },
+    {
+      "epoch": 0.48204662977410867,
+      "grad_norm": 0.62890625,
+      "learning_rate": 5.585462876973376e-05,
+      "loss": 0.9601327180862427,
+      "step": 2214
+    },
+    {
+      "epoch": 0.4822643563458224,
+      "grad_norm": 0.6484375,
+      "learning_rate": 5.583403903252437e-05,
+      "loss": 1.1515915393829346,
+      "step": 2215
+    },
+    {
+      "epoch": 0.4824820829175361,
+      "grad_norm": 0.71484375,
+      "learning_rate": 5.5813444319145166e-05,
+      "loss": 1.1375643014907837,
+      "step": 2216
+    },
+    {
+      "epoch": 0.48269980948924973,
+      "grad_norm": 0.65234375,
+      "learning_rate": 5.579284463606846e-05,
+      "loss": 1.1858837604522705,
+      "step": 2217
+    },
+    {
+      "epoch": 0.48291753606096344,
+      "grad_norm": 0.66796875,
+      "learning_rate": 5.577223998976813e-05,
+      "loss": 1.1176536083221436,
+      "step": 2218
+    },
+    {
+      "epoch": 0.48313526263267714,
+      "grad_norm": 0.6015625,
+      "learning_rate": 5.57516303867196e-05,
+      "loss": 0.9239514470100403,
+      "step": 2219
+    },
+    {
+      "epoch": 0.4833529892043908,
+      "grad_norm": 0.62109375,
+      "learning_rate": 5.573101583339987e-05,
+      "loss": 0.9417582154273987,
+      "step": 2220
+    },
+    {
+      "epoch": 0.4835707157761045,
+      "grad_norm": 0.6328125,
+      "learning_rate": 5.571039633628746e-05,
+      "loss": 0.9900345206260681,
+      "step": 2221
+    },
+    {
+      "epoch": 0.4837884423478182,
+      "grad_norm": 0.65234375,
+      "learning_rate": 5.568977190186248e-05,
+      "loss": 1.0250309705734253,
+      "step": 2222
+    },
+    {
+      "epoch": 0.4840061689195319,
+      "grad_norm": 0.63671875,
+      "learning_rate": 5.566914253660659e-05,
+      "loss": 1.0679841041564941,
+      "step": 2223
+    },
+    {
+      "epoch": 0.48422389549124556,
+      "grad_norm": 0.65625,
+      "learning_rate": 5.564850824700297e-05,
+      "loss": 1.1344422101974487,
+      "step": 2224
+    },
+    {
+      "epoch": 0.48444162206295927,
+      "grad_norm": 0.68359375,
+      "learning_rate": 5.5627869039536366e-05,
+      "loss": 1.0726999044418335,
+      "step": 2225
+    },
+    {
+      "epoch": 0.48465934863467297,
+      "grad_norm": 0.6796875,
+      "learning_rate": 5.5607224920693105e-05,
+      "loss": 1.0213217735290527,
+      "step": 2226
+    },
+    {
+      "epoch": 0.4848770752063866,
+      "grad_norm": 0.6171875,
+      "learning_rate": 5.558657589696098e-05,
+      "loss": 1.029154658317566,
+      "step": 2227
+    },
+    {
+      "epoch": 0.4850948017781003,
+      "grad_norm": 0.63671875,
+      "learning_rate": 5.556592197482939e-05,
+      "loss": 1.0672905445098877,
+      "step": 2228
+    },
+    {
+      "epoch": 0.48531252834981403,
+      "grad_norm": 0.703125,
+      "learning_rate": 5.5545263160789246e-05,
+      "loss": 0.9258766770362854,
+      "step": 2229
+    },
+    {
+      "epoch": 0.48553025492152774,
+      "grad_norm": 0.64453125,
+      "learning_rate": 5.552459946133301e-05,
+      "loss": 1.083539605140686,
+      "step": 2230
+    },
+    {
+      "epoch": 0.4857479814932414,
+      "grad_norm": 0.6171875,
+      "learning_rate": 5.550393088295465e-05,
+      "loss": 0.9952678680419922,
+      "step": 2231
+    },
+    {
+      "epoch": 0.4859657080649551,
+      "grad_norm": 0.63671875,
+      "learning_rate": 5.548325743214971e-05,
+      "loss": 1.024507761001587,
+      "step": 2232
+    },
+    {
+      "epoch": 0.4861834346366688,
+      "grad_norm": 0.78125,
+      "learning_rate": 5.5462579115415235e-05,
+      "loss": 1.13560152053833,
+      "step": 2233
+    },
+    {
+      "epoch": 0.48640116120838245,
+      "grad_norm": 0.7578125,
+      "learning_rate": 5.544189593924981e-05,
+      "loss": 1.0830433368682861,
+      "step": 2234
+    },
+    {
+      "epoch": 0.48661888778009615,
+      "grad_norm": 0.63671875,
+      "learning_rate": 5.542120791015354e-05,
+      "loss": 1.008779525756836,
+      "step": 2235
+    },
+    {
+      "epoch": 0.48683661435180986,
+      "grad_norm": 0.65234375,
+      "learning_rate": 5.540051503462806e-05,
+      "loss": 1.0191333293914795,
+      "step": 2236
+    },
+    {
+      "epoch": 0.48705434092352357,
+      "grad_norm": 0.6640625,
+      "learning_rate": 5.5379817319176544e-05,
+      "loss": 1.0880435705184937,
+      "step": 2237
+    },
+    {
+      "epoch": 0.4872720674952372,
+      "grad_norm": 0.703125,
+      "learning_rate": 5.535911477030364e-05,
+      "loss": 1.0714598894119263,
+      "step": 2238
+    },
+    {
+      "epoch": 0.4874897940669509,
+      "grad_norm": 0.6171875,
+      "learning_rate": 5.533840739451559e-05,
+      "loss": 1.0297578573226929,
+      "step": 2239
+    },
+    {
+      "epoch": 0.4877075206386646,
+      "grad_norm": 0.7265625,
+      "learning_rate": 5.531769519832008e-05,
+      "loss": 1.2005749940872192,
+      "step": 2240
+    },
+    {
+      "epoch": 0.4879252472103783,
+      "grad_norm": 0.7109375,
+      "learning_rate": 5.529697818822633e-05,
+      "loss": 1.0536985397338867,
+      "step": 2241
+    },
+    {
+      "epoch": 0.488142973782092,
+      "grad_norm": 0.7734375,
+      "learning_rate": 5.527625637074511e-05,
+      "loss": 1.1556438207626343,
+      "step": 2242
+    },
+    {
+      "epoch": 0.4883607003538057,
+      "grad_norm": 0.69140625,
+      "learning_rate": 5.525552975238867e-05,
+      "loss": 1.1878830194473267,
+      "step": 2243
+    },
+    {
+      "epoch": 0.4885784269255194,
+      "grad_norm": 0.68359375,
+      "learning_rate": 5.5234798339670745e-05,
+      "loss": 1.060503363609314,
+      "step": 2244
+    },
+    {
+      "epoch": 0.48879615349723304,
+      "grad_norm": 0.7421875,
+      "learning_rate": 5.521406213910665e-05,
+      "loss": 1.1544110774993896,
+      "step": 2245
+    },
+    {
+      "epoch": 0.48901388006894675,
+      "grad_norm": 0.6484375,
+      "learning_rate": 5.519332115721311e-05,
+      "loss": 0.8744109272956848,
+      "step": 2246
+    },
+    {
+      "epoch": 0.48923160664066045,
+      "grad_norm": 0.65234375,
+      "learning_rate": 5.517257540050843e-05,
+      "loss": 1.0792778730392456,
+      "step": 2247
+    },
+    {
+      "epoch": 0.4894493332123741,
+      "grad_norm": 0.61328125,
+      "learning_rate": 5.515182487551239e-05,
+      "loss": 0.9443598389625549,
+      "step": 2248
+    },
+    {
+      "epoch": 0.4896670597840878,
+      "grad_norm": 0.7109375,
+      "learning_rate": 5.513106958874626e-05,
+      "loss": 1.0718337297439575,
+      "step": 2249
+    },
+    {
+      "epoch": 0.4898847863558015,
+      "grad_norm": 0.7421875,
+      "learning_rate": 5.51103095467328e-05,
+      "loss": 1.1133700609207153,
+      "step": 2250
+    },
+    {
+      "epoch": 0.4901025129275152,
+      "grad_norm": 0.7265625,
+      "learning_rate": 5.5089544755996294e-05,
+      "loss": 1.0058363676071167,
+      "step": 2251
+    },
+    {
+      "epoch": 0.49032023949922887,
+      "grad_norm": 0.6640625,
+      "learning_rate": 5.5068775223062497e-05,
+      "loss": 1.054334282875061,
+      "step": 2252
+    },
+    {
+      "epoch": 0.4905379660709426,
+      "grad_norm": 0.6328125,
+      "learning_rate": 5.504800095445865e-05,
+      "loss": 1.0661505460739136,
+      "step": 2253
+    },
+    {
+      "epoch": 0.4907556926426563,
+      "grad_norm": 0.61328125,
+      "learning_rate": 5.50272219567135e-05,
+      "loss": 0.9563264846801758,
+      "step": 2254
+    },
+    {
+      "epoch": 0.49097341921436993,
+      "grad_norm": 0.6328125,
+      "learning_rate": 5.500643823635727e-05,
+      "loss": 0.9690364599227905,
+      "step": 2255
+    },
+    {
+      "epoch": 0.49119114578608364,
+      "grad_norm": 0.65234375,
+      "learning_rate": 5.498564979992168e-05,
+      "loss": 0.9046138525009155,
+      "step": 2256
+    },
+    {
+      "epoch": 0.49140887235779734,
+      "grad_norm": 0.66796875,
+      "learning_rate": 5.4964856653939873e-05,
+      "loss": 0.9679109454154968,
+      "step": 2257
+    },
+    {
+      "epoch": 0.49162659892951105,
+      "grad_norm": 0.65234375,
+      "learning_rate": 5.494405880494658e-05,
+      "loss": 1.1506329774856567,
+      "step": 2258
+    },
+    {
+      "epoch": 0.4918443255012247,
+      "grad_norm": 0.765625,
+      "learning_rate": 5.492325625947791e-05,
+      "loss": 1.0352656841278076,
+      "step": 2259
+    },
+    {
+      "epoch": 0.4920620520729384,
+      "grad_norm": 0.671875,
+      "learning_rate": 5.490244902407149e-05,
+      "loss": 1.133621096611023,
+      "step": 2260
+    },
+    {
+      "epoch": 0.4922797786446521,
+      "grad_norm": 0.625,
+      "learning_rate": 5.4881637105266444e-05,
+      "loss": 1.0800414085388184,
+      "step": 2261
+    },
+    {
+      "epoch": 0.49249750521636576,
+      "grad_norm": 0.6171875,
+      "learning_rate": 5.486082050960333e-05,
+      "loss": 1.0747627019882202,
+      "step": 2262
+    },
+    {
+      "epoch": 0.49271523178807947,
+      "grad_norm": 0.65625,
+      "learning_rate": 5.483999924362417e-05,
+      "loss": 0.8609284162521362,
+      "step": 2263
+    },
+    {
+      "epoch": 0.49293295835979317,
+      "grad_norm": 0.61328125,
+      "learning_rate": 5.481917331387249e-05,
+      "loss": 0.9727290868759155,
+      "step": 2264
+    },
+    {
+      "epoch": 0.4931506849315068,
+      "grad_norm": 0.62109375,
+      "learning_rate": 5.479834272689326e-05,
+      "loss": 1.012024164199829,
+      "step": 2265
+    },
+    {
+      "epoch": 0.4933684115032205,
+      "grad_norm": 0.67578125,
+      "learning_rate": 5.47775074892329e-05,
+      "loss": 1.0493295192718506,
+      "step": 2266
+    },
+    {
+      "epoch": 0.49358613807493423,
+      "grad_norm": 0.6875,
+      "learning_rate": 5.4756667607439344e-05,
+      "loss": 1.112945318222046,
+      "step": 2267
+    },
+    {
+      "epoch": 0.49380386464664794,
+      "grad_norm": 0.65234375,
+      "learning_rate": 5.4735823088061924e-05,
+      "loss": 1.0655453205108643,
+      "step": 2268
+    },
+    {
+      "epoch": 0.4940215912183616,
+      "grad_norm": 0.60546875,
+      "learning_rate": 5.471497393765146e-05,
+      "loss": 0.8661054372787476,
+      "step": 2269
+    },
+    {
+      "epoch": 0.4942393177900753,
+      "grad_norm": 0.60546875,
+      "learning_rate": 5.4694120162760224e-05,
+      "loss": 0.9765713810920715,
+      "step": 2270
+    },
+    {
+      "epoch": 0.494457044361789,
+      "grad_norm": 0.65625,
+      "learning_rate": 5.467326176994194e-05,
+      "loss": 1.2374467849731445,
+      "step": 2271
+    },
+    {
+      "epoch": 0.49467477093350265,
+      "grad_norm": 0.63671875,
+      "learning_rate": 5.465239876575178e-05,
+      "loss": 0.9333547949790955,
+      "step": 2272
+    },
+    {
+      "epoch": 0.49489249750521636,
+      "grad_norm": 0.67578125,
+      "learning_rate": 5.463153115674637e-05,
+      "loss": 1.0719037055969238,
+      "step": 2273
+    },
+    {
+      "epoch": 0.49511022407693006,
+      "grad_norm": 0.61328125,
+      "learning_rate": 5.461065894948379e-05,
+      "loss": 0.9709722995758057,
+      "step": 2274
+    },
+    {
+      "epoch": 0.49532795064864377,
+      "grad_norm": 0.625,
+      "learning_rate": 5.458978215052357e-05,
+      "loss": 0.9057753086090088,
+      "step": 2275
+    },
+    {
+      "epoch": 0.4955456772203574,
+      "grad_norm": 0.640625,
+      "learning_rate": 5.456890076642663e-05,
+      "loss": 1.139663577079773,
+      "step": 2276
+    },
+    {
+      "epoch": 0.4957634037920711,
+      "grad_norm": 0.65625,
+      "learning_rate": 5.454801480375539e-05,
+      "loss": 1.0652050971984863,
+      "step": 2277
+    },
+    {
+      "epoch": 0.49598113036378483,
+      "grad_norm": 0.66796875,
+      "learning_rate": 5.4527124269073705e-05,
+      "loss": 0.9992473125457764,
+      "step": 2278
+    },
+    {
+      "epoch": 0.4961988569354985,
+      "grad_norm": 1.3046875,
+      "learning_rate": 5.4506229168946836e-05,
+      "loss": 1.2360562086105347,
+      "step": 2279
+    },
+    {
+      "epoch": 0.4964165835072122,
+      "grad_norm": 0.64453125,
+      "learning_rate": 5.448532950994148e-05,
+      "loss": 1.041537880897522,
+      "step": 2280
+    },
+    {
+      "epoch": 0.4966343100789259,
+      "grad_norm": 0.69140625,
+      "learning_rate": 5.446442529862582e-05,
+      "loss": 1.0250085592269897,
+      "step": 2281
+    },
+    {
+      "epoch": 0.4968520366506396,
+      "grad_norm": 0.75,
+      "learning_rate": 5.4443516541569406e-05,
+      "loss": 1.0225765705108643,
+      "step": 2282
+    },
+    {
+      "epoch": 0.49706976322235324,
+      "grad_norm": 0.671875,
+      "learning_rate": 5.442260324534325e-05,
+      "loss": 1.0208635330200195,
+      "step": 2283
+    },
+    {
+      "epoch": 0.49728748979406695,
+      "grad_norm": 0.6875,
+      "learning_rate": 5.440168541651976e-05,
+      "loss": 1.265426754951477,
+      "step": 2284
+    },
+    {
+      "epoch": 0.49750521636578066,
+      "grad_norm": 0.68359375,
+      "learning_rate": 5.438076306167282e-05,
+      "loss": 1.072894811630249,
+      "step": 2285
+    },
+    {
+      "epoch": 0.4977229429374943,
+      "grad_norm": 0.70703125,
+      "learning_rate": 5.435983618737769e-05,
+      "loss": 1.1285319328308105,
+      "step": 2286
+    },
+    {
+      "epoch": 0.497940669509208,
+      "grad_norm": 0.6328125,
+      "learning_rate": 5.433890480021107e-05,
+      "loss": 1.054388165473938,
+      "step": 2287
+    },
+    {
+      "epoch": 0.4981583960809217,
+      "grad_norm": 0.6640625,
+      "learning_rate": 5.431796890675108e-05,
+      "loss": 1.1550697088241577,
+      "step": 2288
+    },
+    {
+      "epoch": 0.4983761226526354,
+      "grad_norm": 0.63671875,
+      "learning_rate": 5.429702851357726e-05,
+      "loss": 1.0448004007339478,
+      "step": 2289
+    },
+    {
+      "epoch": 0.4985938492243491,
+      "grad_norm": 0.7578125,
+      "learning_rate": 5.4276083627270514e-05,
+      "loss": 1.1140239238739014,
+      "step": 2290
+    },
+    {
+      "epoch": 0.4988115757960628,
+      "grad_norm": 0.6953125,
+      "learning_rate": 5.425513425441326e-05,
+      "loss": 1.0522226095199585,
+      "step": 2291
+    },
+    {
+      "epoch": 0.4990293023677765,
+      "grad_norm": 0.66015625,
+      "learning_rate": 5.4234180401589225e-05,
+      "loss": 1.0664644241333008,
+      "step": 2292
+    },
+    {
+      "epoch": 0.49924702893949013,
+      "grad_norm": 0.6640625,
+      "learning_rate": 5.421322207538359e-05,
+      "loss": 1.1129908561706543,
+      "step": 2293
+    },
+    {
+      "epoch": 0.49946475551120384,
+      "grad_norm": 0.67578125,
+      "learning_rate": 5.419225928238296e-05,
+      "loss": 1.0959023237228394,
+      "step": 2294
+    },
+    {
+      "epoch": 0.49968248208291755,
+      "grad_norm": 0.625,
+      "learning_rate": 5.41712920291753e-05,
+      "loss": 1.0543608665466309,
+      "step": 2295
+    },
+    {
+      "epoch": 0.49990020865463125,
+      "grad_norm": 0.6484375,
+      "learning_rate": 5.415032032234997e-05,
+      "loss": 1.0442043542861938,
+      "step": 2296
+    },
+    {
+      "epoch": 0.500117935226345,
+      "grad_norm": 0.70703125,
+      "learning_rate": 5.412934416849783e-05,
+      "loss": 1.1262917518615723,
+      "step": 2297
+    },
+    {
+      "epoch": 0.5003356617980586,
+      "grad_norm": 0.7109375,
+      "learning_rate": 5.410836357421099e-05,
+      "loss": 1.0340837240219116,
+      "step": 2298
+    },
+    {
+      "epoch": 0.5005533883697723,
+      "grad_norm": 0.6484375,
+      "learning_rate": 5.408737854608307e-05,
+      "loss": 1.0611491203308105,
+      "step": 2299
+    },
+    {
+      "epoch": 0.500771114941486,
+      "grad_norm": 0.625,
+      "learning_rate": 5.4066389090709034e-05,
+      "loss": 1.0868473052978516,
+      "step": 2300
+    },
+    {
+      "epoch": 0.5009888415131997,
+      "grad_norm": 0.6875,
+      "learning_rate": 5.404539521468525e-05,
+      "loss": 1.1099257469177246,
+      "step": 2301
+    },
+    {
+      "epoch": 0.5012065680849134,
+      "grad_norm": 0.68359375,
+      "learning_rate": 5.402439692460946e-05,
+      "loss": 1.1413013935089111,
+      "step": 2302
+    },
+    {
+      "epoch": 0.5014242946566271,
+      "grad_norm": 0.6640625,
+      "learning_rate": 5.400339422708083e-05,
+      "loss": 1.1738660335540771,
+      "step": 2303
+    },
+    {
+      "epoch": 0.5016420212283408,
+      "grad_norm": 0.6484375,
+      "learning_rate": 5.398238712869986e-05,
+      "loss": 0.9392592310905457,
+      "step": 2304
+    },
+    {
+      "epoch": 0.5018597478000544,
+      "grad_norm": 0.68359375,
+      "learning_rate": 5.396137563606847e-05,
+      "loss": 0.9533593654632568,
+      "step": 2305
+    },
+    {
+      "epoch": 0.5020774743717681,
+      "grad_norm": 0.671875,
+      "learning_rate": 5.3940359755789946e-05,
+      "loss": 1.0576871633529663,
+      "step": 2306
+    },
+    {
+      "epoch": 0.5022952009434818,
+      "grad_norm": 0.64453125,
+      "learning_rate": 5.3919339494468965e-05,
+      "loss": 0.8642014861106873,
+      "step": 2307
+    },
+    {
+      "epoch": 0.5025129275151955,
+      "grad_norm": 0.7265625,
+      "learning_rate": 5.389831485871156e-05,
+      "loss": 0.98480623960495,
+      "step": 2308
+    },
+    {
+      "epoch": 0.5027306540869092,
+      "grad_norm": 0.69140625,
+      "learning_rate": 5.387728585512516e-05,
+      "loss": 1.069881558418274,
+      "step": 2309
+    },
+    {
+      "epoch": 0.5029483806586229,
+      "grad_norm": 0.67578125,
+      "learning_rate": 5.385625249031857e-05,
+      "loss": 1.1423238515853882,
+      "step": 2310
+    },
+    {
+      "epoch": 0.5031661072303366,
+      "grad_norm": 0.65234375,
+      "learning_rate": 5.3835214770901945e-05,
+      "loss": 1.0471490621566772,
+      "step": 2311
+    },
+    {
+      "epoch": 0.5033838338020502,
+      "grad_norm": 0.66015625,
+      "learning_rate": 5.3814172703486796e-05,
+      "loss": 0.9343581199645996,
+      "step": 2312
+    },
+    {
+      "epoch": 0.5036015603737639,
+      "grad_norm": 0.65234375,
+      "learning_rate": 5.379312629468605e-05,
+      "loss": 1.0963847637176514,
+      "step": 2313
+    },
+    {
+      "epoch": 0.5038192869454776,
+      "grad_norm": 0.67578125,
+      "learning_rate": 5.3772075551113976e-05,
+      "loss": 1.2260597944259644,
+      "step": 2314
+    },
+    {
+      "epoch": 0.5040370135171913,
+      "grad_norm": 0.73046875,
+      "learning_rate": 5.375102047938616e-05,
+      "loss": 1.0102620124816895,
+      "step": 2315
+    },
+    {
+      "epoch": 0.504254740088905,
+      "grad_norm": 0.6796875,
+      "learning_rate": 5.372996108611964e-05,
+      "loss": 1.0200021266937256,
+      "step": 2316
+    },
+    {
+      "epoch": 0.5044724666606187,
+      "grad_norm": 0.63671875,
+      "learning_rate": 5.370889737793271e-05,
+      "loss": 0.9822466373443604,
+      "step": 2317
+    },
+    {
+      "epoch": 0.5046901932323324,
+      "grad_norm": 0.640625,
+      "learning_rate": 5.3687829361445104e-05,
+      "loss": 1.0485734939575195,
+      "step": 2318
+    },
+    {
+      "epoch": 0.504907919804046,
+      "grad_norm": 0.625,
+      "learning_rate": 5.366675704327787e-05,
+      "loss": 0.9300712943077087,
+      "step": 2319
+    },
+    {
+      "epoch": 0.5051256463757597,
+      "grad_norm": 0.640625,
+      "learning_rate": 5.364568043005339e-05,
+      "loss": 1.1386038064956665,
+      "step": 2320
+    },
+    {
+      "epoch": 0.5053433729474734,
+      "grad_norm": 0.69140625,
+      "learning_rate": 5.362459952839545e-05,
+      "loss": 1.0501887798309326,
+      "step": 2321
+    },
+    {
+      "epoch": 0.5055610995191872,
+      "grad_norm": 0.59375,
+      "learning_rate": 5.3603514344929144e-05,
+      "loss": 0.9779641032218933,
+      "step": 2322
+    },
+    {
+      "epoch": 0.5057788260909009,
+      "grad_norm": 0.62890625,
+      "learning_rate": 5.358242488628092e-05,
+      "loss": 1.02035653591156,
+      "step": 2323
+    },
+    {
+      "epoch": 0.5059965526626146,
+      "grad_norm": 0.66015625,
+      "learning_rate": 5.356133115907856e-05,
+      "loss": 1.1288264989852905,
+      "step": 2324
+    },
+    {
+      "epoch": 0.5062142792343283,
+      "grad_norm": 0.62890625,
+      "learning_rate": 5.354023316995121e-05,
+      "loss": 0.928952157497406,
+      "step": 2325
+    },
+    {
+      "epoch": 0.5064320058060419,
+      "grad_norm": 0.71875,
+      "learning_rate": 5.351913092552934e-05,
+      "loss": 0.9337797164916992,
+      "step": 2326
+    },
+    {
+      "epoch": 0.5066497323777556,
+      "grad_norm": 0.6796875,
+      "learning_rate": 5.349802443244478e-05,
+      "loss": 0.9984951019287109,
+      "step": 2327
+    },
+    {
+      "epoch": 0.5068674589494693,
+      "grad_norm": 0.62890625,
+      "learning_rate": 5.347691369733063e-05,
+      "loss": 1.0399587154388428,
+      "step": 2328
+    },
+    {
+      "epoch": 0.507085185521183,
+      "grad_norm": 0.66796875,
+      "learning_rate": 5.345579872682142e-05,
+      "loss": 1.120489239692688,
+      "step": 2329
+    },
+    {
+      "epoch": 0.5073029120928967,
+      "grad_norm": 0.64453125,
+      "learning_rate": 5.3434679527552944e-05,
+      "loss": 0.9569326639175415,
+      "step": 2330
+    },
+    {
+      "epoch": 0.5075206386646104,
+      "grad_norm": 0.66796875,
+      "learning_rate": 5.341355610616232e-05,
+      "loss": 1.143354892730713,
+      "step": 2331
+    },
+    {
+      "epoch": 0.5077383652363241,
+      "grad_norm": 0.6484375,
+      "learning_rate": 5.339242846928802e-05,
+      "loss": 1.0220927000045776,
+      "step": 2332
+    },
+    {
+      "epoch": 0.5079560918080377,
+      "grad_norm": 0.65234375,
+      "learning_rate": 5.337129662356986e-05,
+      "loss": 1.089849591255188,
+      "step": 2333
+    },
+    {
+      "epoch": 0.5081738183797514,
+      "grad_norm": 0.671875,
+      "learning_rate": 5.335016057564893e-05,
+      "loss": 1.1145694255828857,
+      "step": 2334
+    },
+    {
+      "epoch": 0.5083915449514651,
+      "grad_norm": 0.609375,
+      "learning_rate": 5.332902033216766e-05,
+      "loss": 0.879360020160675,
+      "step": 2335
+    },
+    {
+      "epoch": 0.5086092715231788,
+      "grad_norm": 0.64453125,
+      "learning_rate": 5.330787589976984e-05,
+      "loss": 1.0821082592010498,
+      "step": 2336
+    },
+    {
+      "epoch": 0.5088269980948925,
+      "grad_norm": 0.640625,
+      "learning_rate": 5.32867272851005e-05,
+      "loss": 1.0466744899749756,
+      "step": 2337
+    },
+    {
+      "epoch": 0.5090447246666062,
+      "grad_norm": 0.65234375,
+      "learning_rate": 5.3265574494806035e-05,
+      "loss": 1.051186442375183,
+      "step": 2338
+    },
+    {
+      "epoch": 0.5092624512383199,
+      "grad_norm": 0.68359375,
+      "learning_rate": 5.324441753553415e-05,
+      "loss": 1.1122018098831177,
+      "step": 2339
+    },
+    {
+      "epoch": 0.5094801778100335,
+      "grad_norm": 0.62890625,
+      "learning_rate": 5.322325641393384e-05,
+      "loss": 1.1482553482055664,
+      "step": 2340
+    },
+    {
+      "epoch": 0.5096979043817472,
+      "grad_norm": 0.66015625,
+      "learning_rate": 5.3202091136655436e-05,
+      "loss": 1.0701849460601807,
+      "step": 2341
+    },
+    {
+      "epoch": 0.5099156309534609,
+      "grad_norm": 0.6875,
+      "learning_rate": 5.3180921710350536e-05,
+      "loss": 1.0480815172195435,
+      "step": 2342
+    },
+    {
+      "epoch": 0.5101333575251746,
+      "grad_norm": 0.625,
+      "learning_rate": 5.3159748141672064e-05,
+      "loss": 1.1133004426956177,
+      "step": 2343
+    },
+    {
+      "epoch": 0.5103510840968883,
+      "grad_norm": 0.66015625,
+      "learning_rate": 5.313857043727428e-05,
+      "loss": 1.0415581464767456,
+      "step": 2344
+    },
+    {
+      "epoch": 0.510568810668602,
+      "grad_norm": 0.65234375,
+      "learning_rate": 5.311738860381266e-05,
+      "loss": 0.9631825089454651,
+      "step": 2345
+    },
+    {
+      "epoch": 0.5107865372403158,
+      "grad_norm": 0.55859375,
+      "learning_rate": 5.309620264794406e-05,
+      "loss": 0.8505051136016846,
+      "step": 2346
+    },
+    {
+      "epoch": 0.5110042638120293,
+      "grad_norm": 0.62109375,
+      "learning_rate": 5.3075012576326607e-05,
+      "loss": 1.0602531433105469,
+      "step": 2347
+    },
+    {
+      "epoch": 0.511221990383743,
+      "grad_norm": 0.625,
+      "learning_rate": 5.305381839561969e-05,
+      "loss": 1.0675511360168457,
+      "step": 2348
+    },
+    {
+      "epoch": 0.5114397169554568,
+      "grad_norm": 0.62109375,
+      "learning_rate": 5.303262011248401e-05,
+      "loss": 0.9242600202560425,
+      "step": 2349
+    },
+    {
+      "epoch": 0.5116574435271705,
+      "grad_norm": 0.71484375,
+      "learning_rate": 5.301141773358159e-05,
+      "loss": 1.0865094661712646,
+      "step": 2350
+    },
+    {
+      "epoch": 0.5118751700988842,
+      "grad_norm": 0.67578125,
+      "learning_rate": 5.299021126557567e-05,
+      "loss": 1.048385739326477,
+      "step": 2351
+    },
+    {
+      "epoch": 0.5120928966705979,
+      "grad_norm": 0.6953125,
+      "learning_rate": 5.2969000715130856e-05,
+      "loss": 1.101537823677063,
+      "step": 2352
+    },
+    {
+      "epoch": 0.5123106232423116,
+      "grad_norm": 0.62890625,
+      "learning_rate": 5.294778608891297e-05,
+      "loss": 1.0104236602783203,
+      "step": 2353
+    },
+    {
+      "epoch": 0.5125283498140252,
+      "grad_norm": 0.640625,
+      "learning_rate": 5.292656739358913e-05,
+      "loss": 0.9523093104362488,
+      "step": 2354
+    },
+    {
+      "epoch": 0.5127460763857389,
+      "grad_norm": 0.671875,
+      "learning_rate": 5.2905344635827786e-05,
+      "loss": 1.1295146942138672,
+      "step": 2355
+    },
+    {
+      "epoch": 0.5129638029574526,
+      "grad_norm": 0.67578125,
+      "learning_rate": 5.288411782229859e-05,
+      "loss": 1.142831563949585,
+      "step": 2356
+    },
+    {
+      "epoch": 0.5131815295291663,
+      "grad_norm": 0.609375,
+      "learning_rate": 5.286288695967251e-05,
+      "loss": 1.1294299364089966,
+      "step": 2357
+    },
+    {
+      "epoch": 0.51339925610088,
+      "grad_norm": 0.66015625,
+      "learning_rate": 5.2841652054621775e-05,
+      "loss": 0.9770073890686035,
+      "step": 2358
+    },
+    {
+      "epoch": 0.5136169826725937,
+      "grad_norm": 0.64453125,
+      "learning_rate": 5.2820413113819885e-05,
+      "loss": 1.0592797994613647,
+      "step": 2359
+    },
+    {
+      "epoch": 0.5138347092443074,
+      "grad_norm": 0.6640625,
+      "learning_rate": 5.279917014394161e-05,
+      "loss": 1.1141693592071533,
+      "step": 2360
+    },
+    {
+      "epoch": 0.514052435816021,
+      "grad_norm": 0.71484375,
+      "learning_rate": 5.2777923151663e-05,
+      "loss": 1.1165754795074463,
+      "step": 2361
+    },
+    {
+      "epoch": 0.5142701623877347,
+      "grad_norm": 0.63671875,
+      "learning_rate": 5.275667214366135e-05,
+      "loss": 1.01983642578125,
+      "step": 2362
+    },
+    {
+      "epoch": 0.5144878889594484,
+      "grad_norm": 0.67578125,
+      "learning_rate": 5.273541712661522e-05,
+      "loss": 1.135621428489685,
+      "step": 2363
+    },
+    {
+      "epoch": 0.5147056155311621,
+      "grad_norm": 0.7109375,
+      "learning_rate": 5.271415810720442e-05,
+      "loss": 0.9678177833557129,
+      "step": 2364
+    },
+    {
+      "epoch": 0.5149233421028758,
+      "grad_norm": 0.65234375,
+      "learning_rate": 5.269289509211005e-05,
+      "loss": 1.0072041749954224,
+      "step": 2365
+    },
+    {
+      "epoch": 0.5151410686745895,
+      "grad_norm": 0.6484375,
+      "learning_rate": 5.267162808801445e-05,
+      "loss": 1.037602424621582,
+      "step": 2366
+    },
+    {
+      "epoch": 0.5153587952463032,
+      "grad_norm": 0.6640625,
+      "learning_rate": 5.265035710160119e-05,
+      "loss": 0.9858287572860718,
+      "step": 2367
+    },
+    {
+      "epoch": 0.5155765218180168,
+      "grad_norm": 0.66796875,
+      "learning_rate": 5.262908213955511e-05,
+      "loss": 0.9750328063964844,
+      "step": 2368
+    },
+    {
+      "epoch": 0.5157942483897305,
+      "grad_norm": 0.6640625,
+      "learning_rate": 5.260780320856233e-05,
+      "loss": 1.0424320697784424,
+      "step": 2369
+    },
+    {
+      "epoch": 0.5160119749614442,
+      "grad_norm": 0.66015625,
+      "learning_rate": 5.258652031531016e-05,
+      "loss": 1.1035979986190796,
+      "step": 2370
+    },
+    {
+      "epoch": 0.516229701533158,
+      "grad_norm": 0.75390625,
+      "learning_rate": 5.256523346648721e-05,
+      "loss": 1.110412359237671,
+      "step": 2371
+    },
+    {
+      "epoch": 0.5164474281048717,
+      "grad_norm": 0.65625,
+      "learning_rate": 5.254394266878327e-05,
+      "loss": 0.9439858794212341,
+      "step": 2372
+    },
+    {
+      "epoch": 0.5166651546765854,
+      "grad_norm": 0.65234375,
+      "learning_rate": 5.252264792888944e-05,
+      "loss": 1.0794421434402466,
+      "step": 2373
+    },
+    {
+      "epoch": 0.5168828812482991,
+      "grad_norm": 0.8203125,
+      "learning_rate": 5.2501349253498e-05,
+      "loss": 1.1100918054580688,
+      "step": 2374
+    },
+    {
+      "epoch": 0.5171006078200127,
+      "grad_norm": 0.68359375,
+      "learning_rate": 5.2480046649302516e-05,
+      "loss": 0.9340540170669556,
+      "step": 2375
+    },
+    {
+      "epoch": 0.5173183343917264,
+      "grad_norm": 0.59375,
+      "learning_rate": 5.245874012299776e-05,
+      "loss": 0.945186197757721,
+      "step": 2376
+    },
+    {
+      "epoch": 0.5175360609634401,
+      "grad_norm": 0.7265625,
+      "learning_rate": 5.243742968127973e-05,
+      "loss": 1.1017028093338013,
+      "step": 2377
+    },
+    {
+      "epoch": 0.5177537875351538,
+      "grad_norm": 0.68359375,
+      "learning_rate": 5.241611533084568e-05,
+      "loss": 0.9315423369407654,
+      "step": 2378
+    },
+    {
+      "epoch": 0.5179715141068675,
+      "grad_norm": 0.6484375,
+      "learning_rate": 5.239479707839408e-05,
+      "loss": 1.110782504081726,
+      "step": 2379
+    },
+    {
+      "epoch": 0.5181892406785812,
+      "grad_norm": 0.62890625,
+      "learning_rate": 5.237347493062462e-05,
+      "loss": 1.0655083656311035,
+      "step": 2380
+    },
+    {
+      "epoch": 0.5184069672502949,
+      "grad_norm": 0.64453125,
+      "learning_rate": 5.235214889423822e-05,
+      "loss": 1.0728062391281128,
+      "step": 2381
+    },
+    {
+      "epoch": 0.5186246938220085,
+      "grad_norm": 0.60546875,
+      "learning_rate": 5.233081897593702e-05,
+      "loss": 0.9030443429946899,
+      "step": 2382
+    },
+    {
+      "epoch": 0.5188424203937222,
+      "grad_norm": 0.6796875,
+      "learning_rate": 5.230948518242439e-05,
+      "loss": 1.0605692863464355,
+      "step": 2383
+    },
+    {
+      "epoch": 0.5190601469654359,
+      "grad_norm": 0.65234375,
+      "learning_rate": 5.2288147520404914e-05,
+      "loss": 0.961175262928009,
+      "step": 2384
+    },
+    {
+      "epoch": 0.5192778735371496,
+      "grad_norm": 0.65234375,
+      "learning_rate": 5.226680599658438e-05,
+      "loss": 1.0372780561447144,
+      "step": 2385
+    },
+    {
+      "epoch": 0.5194956001088633,
+      "grad_norm": 0.6484375,
+      "learning_rate": 5.22454606176698e-05,
+      "loss": 1.0863866806030273,
+      "step": 2386
+    },
+    {
+      "epoch": 0.519713326680577,
+      "grad_norm": 0.62890625,
+      "learning_rate": 5.222411139036936e-05,
+      "loss": 0.9795717597007751,
+      "step": 2387
+    },
+    {
+      "epoch": 0.5199310532522907,
+      "grad_norm": 0.65234375,
+      "learning_rate": 5.220275832139256e-05,
+      "loss": 1.0077180862426758,
+      "step": 2388
+    },
+    {
+      "epoch": 0.5201487798240043,
+      "grad_norm": 0.6328125,
+      "learning_rate": 5.218140141744998e-05,
+      "loss": 1.0163359642028809,
+      "step": 2389
+    },
+    {
+      "epoch": 0.520366506395718,
+      "grad_norm": 0.65625,
+      "learning_rate": 5.2160040685253474e-05,
+      "loss": 1.0457078218460083,
+      "step": 2390
+    },
+    {
+      "epoch": 0.5205842329674317,
+      "grad_norm": 0.67578125,
+      "learning_rate": 5.213867613151611e-05,
+      "loss": 1.0830796957015991,
+      "step": 2391
+    },
+    {
+      "epoch": 0.5208019595391454,
+      "grad_norm": 0.62890625,
+      "learning_rate": 5.211730776295213e-05,
+      "loss": 1.0148859024047852,
+      "step": 2392
+    },
+    {
+      "epoch": 0.5210196861108591,
+      "grad_norm": 0.62109375,
+      "learning_rate": 5.209593558627696e-05,
+      "loss": 1.0398069620132446,
+      "step": 2393
+    },
+    {
+      "epoch": 0.5212374126825728,
+      "grad_norm": 0.609375,
+      "learning_rate": 5.207455960820724e-05,
+      "loss": 0.9735275506973267,
+      "step": 2394
+    },
+    {
+      "epoch": 0.5214551392542864,
+      "grad_norm": 0.703125,
+      "learning_rate": 5.205317983546083e-05,
+      "loss": 1.1131062507629395,
+      "step": 2395
+    },
+    {
+      "epoch": 0.5216728658260001,
+      "grad_norm": 0.69140625,
+      "learning_rate": 5.203179627475676e-05,
+      "loss": 1.1842961311340332,
+      "step": 2396
+    },
+    {
+      "epoch": 0.5218905923977138,
+      "grad_norm": 0.66015625,
+      "learning_rate": 5.201040893281523e-05,
+      "loss": 1.0165462493896484,
+      "step": 2397
+    },
+    {
+      "epoch": 0.5221083189694276,
+      "grad_norm": 0.60546875,
+      "learning_rate": 5.1989017816357655e-05,
+      "loss": 0.9637759923934937,
+      "step": 2398
+    },
+    {
+      "epoch": 0.5223260455411413,
+      "grad_norm": 0.62109375,
+      "learning_rate": 5.196762293210665e-05,
+      "loss": 1.0319578647613525,
+      "step": 2399
+    },
+    {
+      "epoch": 0.522543772112855,
+      "grad_norm": 0.6484375,
+      "learning_rate": 5.194622428678596e-05,
+      "loss": 1.0897579193115234,
+      "step": 2400
+    },
+    {
+      "epoch": 0.5227614986845687,
+      "grad_norm": 0.63671875,
+      "learning_rate": 5.192482188712057e-05,
+      "loss": 1.0550191402435303,
+      "step": 2401
+    },
+    {
+      "epoch": 0.5229792252562823,
+      "grad_norm": 0.6015625,
+      "learning_rate": 5.190341573983663e-05,
+      "loss": 0.8960920572280884,
+      "step": 2402
+    },
+    {
+      "epoch": 0.523196951827996,
+      "grad_norm": 0.6328125,
+      "learning_rate": 5.1882005851661416e-05,
+      "loss": 1.1308008432388306,
+      "step": 2403
+    },
+    {
+      "epoch": 0.5234146783997097,
+      "grad_norm": 0.58984375,
+      "learning_rate": 5.186059222932347e-05,
+      "loss": 0.881718635559082,
+      "step": 2404
+    },
+    {
+      "epoch": 0.5236324049714234,
+      "grad_norm": 0.65625,
+      "learning_rate": 5.183917487955243e-05,
+      "loss": 1.0711826086044312,
+      "step": 2405
+    },
+    {
+      "epoch": 0.5238501315431371,
+      "grad_norm": 0.68359375,
+      "learning_rate": 5.181775380907914e-05,
+      "loss": 1.0688077211380005,
+      "step": 2406
+    },
+    {
+      "epoch": 0.5240678581148508,
+      "grad_norm": 0.640625,
+      "learning_rate": 5.179632902463562e-05,
+      "loss": 1.0407114028930664,
+      "step": 2407
+    },
+    {
+      "epoch": 0.5242855846865645,
+      "grad_norm": 0.6328125,
+      "learning_rate": 5.177490053295503e-05,
+      "loss": 1.114784598350525,
+      "step": 2408
+    },
+    {
+      "epoch": 0.5245033112582781,
+      "grad_norm": 0.625,
+      "learning_rate": 5.175346834077171e-05,
+      "loss": 0.9953246116638184,
+      "step": 2409
+    },
+    {
+      "epoch": 0.5247210378299918,
+      "grad_norm": 0.63671875,
+      "learning_rate": 5.17320324548212e-05,
+      "loss": 1.0019463300704956,
+      "step": 2410
+    },
+    {
+      "epoch": 0.5249387644017055,
+      "grad_norm": 0.62890625,
+      "learning_rate": 5.171059288184012e-05,
+      "loss": 1.0004733800888062,
+      "step": 2411
+    },
+    {
+      "epoch": 0.5251564909734192,
+      "grad_norm": 0.59765625,
+      "learning_rate": 5.1689149628566305e-05,
+      "loss": 0.9236584305763245,
+      "step": 2412
+    },
+    {
+      "epoch": 0.5253742175451329,
+      "grad_norm": 0.63671875,
+      "learning_rate": 5.166770270173874e-05,
+      "loss": 1.0335029363632202,
+      "step": 2413
+    },
+    {
+      "epoch": 0.5255919441168466,
+      "grad_norm": 0.59765625,
+      "learning_rate": 5.164625210809757e-05,
+      "loss": 0.9341624975204468,
+      "step": 2414
+    },
+    {
+      "epoch": 0.5258096706885603,
+      "grad_norm": 0.63671875,
+      "learning_rate": 5.1624797854384054e-05,
+      "loss": 1.0771796703338623,
+      "step": 2415
+    },
+    {
+      "epoch": 0.5260273972602739,
+      "grad_norm": 0.65234375,
+      "learning_rate": 5.160333994734066e-05,
+      "loss": 1.02707040309906,
+      "step": 2416
+    },
+    {
+      "epoch": 0.5262451238319876,
+      "grad_norm": 0.63671875,
+      "learning_rate": 5.158187839371096e-05,
+      "loss": 1.1355206966400146,
+      "step": 2417
+    },
+    {
+      "epoch": 0.5264628504037013,
+      "grad_norm": 0.625,
+      "learning_rate": 5.156041320023969e-05,
+      "loss": 1.0299577713012695,
+      "step": 2418
+    },
+    {
+      "epoch": 0.526680576975415,
+      "grad_norm": 0.65234375,
+      "learning_rate": 5.1538944373672694e-05,
+      "loss": 1.091135859489441,
+      "step": 2419
+    },
+    {
+      "epoch": 0.5268983035471287,
+      "grad_norm": 0.7734375,
+      "learning_rate": 5.151747192075702e-05,
+      "loss": 0.9809549450874329,
+      "step": 2420
+    },
+    {
+      "epoch": 0.5271160301188424,
+      "grad_norm": 0.65234375,
+      "learning_rate": 5.149599584824083e-05,
+      "loss": 1.058118462562561,
+      "step": 2421
+    },
+    {
+      "epoch": 0.5273337566905562,
+      "grad_norm": 0.65234375,
+      "learning_rate": 5.14745161628734e-05,
+      "loss": 1.000205159187317,
+      "step": 2422
+    },
+    {
+      "epoch": 0.5275514832622697,
+      "grad_norm": 0.65625,
+      "learning_rate": 5.1453032871405166e-05,
+      "loss": 1.0813053846359253,
+      "step": 2423
+    },
+    {
+      "epoch": 0.5277692098339835,
+      "grad_norm": 0.6953125,
+      "learning_rate": 5.143154598058769e-05,
+      "loss": 1.0957592725753784,
+      "step": 2424
+    },
+    {
+      "epoch": 0.5279869364056972,
+      "grad_norm": 0.6328125,
+      "learning_rate": 5.1410055497173654e-05,
+      "loss": 1.11738121509552,
+      "step": 2425
+    },
+    {
+      "epoch": 0.5282046629774109,
+      "grad_norm": 0.65625,
+      "learning_rate": 5.1388561427916916e-05,
+      "loss": 1.0921342372894287,
+      "step": 2426
+    },
+    {
+      "epoch": 0.5284223895491246,
+      "grad_norm": 0.6328125,
+      "learning_rate": 5.136706377957239e-05,
+      "loss": 1.0667980909347534,
+      "step": 2427
+    },
+    {
+      "epoch": 0.5286401161208383,
+      "grad_norm": 0.66796875,
+      "learning_rate": 5.134556255889616e-05,
+      "loss": 1.1588208675384521,
+      "step": 2428
+    },
+    {
+      "epoch": 0.528857842692552,
+      "grad_norm": 0.6328125,
+      "learning_rate": 5.1324057772645424e-05,
+      "loss": 0.9549190402030945,
+      "step": 2429
+    },
+    {
+      "epoch": 0.5290755692642656,
+      "grad_norm": 0.63671875,
+      "learning_rate": 5.13025494275785e-05,
+      "loss": 0.9805537462234497,
+      "step": 2430
+    },
+    {
+      "epoch": 0.5292932958359793,
+      "grad_norm": 0.640625,
+      "learning_rate": 5.128103753045484e-05,
+      "loss": 1.0451139211654663,
+      "step": 2431
+    },
+    {
+      "epoch": 0.529511022407693,
+      "grad_norm": 0.640625,
+      "learning_rate": 5.1259522088034966e-05,
+      "loss": 0.9699036478996277,
+      "step": 2432
+    },
+    {
+      "epoch": 0.5297287489794067,
+      "grad_norm": 0.60546875,
+      "learning_rate": 5.123800310708057e-05,
+      "loss": 1.0419772863388062,
+      "step": 2433
+    },
+    {
+      "epoch": 0.5299464755511204,
+      "grad_norm": 0.62890625,
+      "learning_rate": 5.1216480594354425e-05,
+      "loss": 1.0681947469711304,
+      "step": 2434
+    },
+    {
+      "epoch": 0.5301642021228341,
+      "grad_norm": 0.671875,
+      "learning_rate": 5.1194954556620404e-05,
+      "loss": 1.1232916116714478,
+      "step": 2435
+    },
+    {
+      "epoch": 0.5303819286945478,
+      "grad_norm": 0.65625,
+      "learning_rate": 5.117342500064353e-05,
+      "loss": 0.9915422201156616,
+      "step": 2436
+    },
+    {
+      "epoch": 0.5305996552662614,
+      "grad_norm": 0.58984375,
+      "learning_rate": 5.115189193318989e-05,
+      "loss": 0.9690672159194946,
+      "step": 2437
+    },
+    {
+      "epoch": 0.5308173818379751,
+      "grad_norm": 0.69921875,
+      "learning_rate": 5.1130355361026683e-05,
+      "loss": 1.1263179779052734,
+      "step": 2438
+    },
+    {
+      "epoch": 0.5310351084096888,
+      "grad_norm": 0.65625,
+      "learning_rate": 5.110881529092223e-05,
+      "loss": 0.9207851886749268,
+      "step": 2439
+    },
+    {
+      "epoch": 0.5312528349814025,
+      "grad_norm": 0.62109375,
+      "learning_rate": 5.108727172964594e-05,
+      "loss": 1.0136194229125977,
+      "step": 2440
+    },
+    {
+      "epoch": 0.5314705615531162,
+      "grad_norm": 0.6796875,
+      "learning_rate": 5.1065724683968304e-05,
+      "loss": 1.0676617622375488,
+      "step": 2441
+    },
+    {
+      "epoch": 0.5316882881248299,
+      "grad_norm": 0.64453125,
+      "learning_rate": 5.104417416066091e-05,
+      "loss": 0.9740859866142273,
+      "step": 2442
+    },
+    {
+      "epoch": 0.5319060146965436,
+      "grad_norm": 0.69140625,
+      "learning_rate": 5.1022620166496486e-05,
+      "loss": 1.0161793231964111,
+      "step": 2443
+    },
+    {
+      "epoch": 0.5321237412682572,
+      "grad_norm": 0.6875,
+      "learning_rate": 5.1001062708248786e-05,
+      "loss": 0.9609599709510803,
+      "step": 2444
+    },
+    {
+      "epoch": 0.5323414678399709,
+      "grad_norm": 0.6484375,
+      "learning_rate": 5.097950179269269e-05,
+      "loss": 1.03477144241333,
+      "step": 2445
+    },
+    {
+      "epoch": 0.5325591944116846,
+      "grad_norm": 0.62890625,
+      "learning_rate": 5.095793742660415e-05,
+      "loss": 1.052768588066101,
+      "step": 2446
+    },
+    {
+      "epoch": 0.5327769209833983,
+      "grad_norm": 0.6796875,
+      "learning_rate": 5.093636961676021e-05,
+      "loss": 1.1194722652435303,
+      "step": 2447
+    },
+    {
+      "epoch": 0.532994647555112,
+      "grad_norm": 0.66796875,
+      "learning_rate": 5.0914798369939e-05,
+      "loss": 0.983085036277771,
+      "step": 2448
+    },
+    {
+      "epoch": 0.5332123741268258,
+      "grad_norm": 0.625,
+      "learning_rate": 5.0893223692919726e-05,
+      "loss": 1.049306869506836,
+      "step": 2449
+    },
+    {
+      "epoch": 0.5334301006985395,
+      "grad_norm": 0.68359375,
+      "learning_rate": 5.087164559248265e-05,
+      "loss": 1.104406476020813,
+      "step": 2450
+    },
+    {
+      "epoch": 0.5336478272702531,
+      "grad_norm": 0.67578125,
+      "learning_rate": 5.085006407540916e-05,
+      "loss": 1.1639848947525024,
+      "step": 2451
+    },
+    {
+      "epoch": 0.5338655538419668,
+      "grad_norm": 0.63671875,
+      "learning_rate": 5.082847914848167e-05,
+      "loss": 0.8926988840103149,
+      "step": 2452
+    },
+    {
+      "epoch": 0.5340832804136805,
+      "grad_norm": 0.69140625,
+      "learning_rate": 5.080689081848369e-05,
+      "loss": 0.974684476852417,
+      "step": 2453
+    },
+    {
+      "epoch": 0.5343010069853942,
+      "grad_norm": 0.64453125,
+      "learning_rate": 5.0785299092199794e-05,
+      "loss": 1.014963984489441,
+      "step": 2454
+    },
+    {
+      "epoch": 0.5345187335571079,
+      "grad_norm": 0.6796875,
+      "learning_rate": 5.07637039764156e-05,
+      "loss": 1.1440261602401733,
+      "step": 2455
+    },
+    {
+      "epoch": 0.5347364601288216,
+      "grad_norm": 0.671875,
+      "learning_rate": 5.0742105477917846e-05,
+      "loss": 1.09966242313385,
+      "step": 2456
+    },
+    {
+      "epoch": 0.5349541867005353,
+      "grad_norm": 0.66796875,
+      "learning_rate": 5.072050360349429e-05,
+      "loss": 1.0667719841003418,
+      "step": 2457
+    },
+    {
+      "epoch": 0.5351719132722489,
+      "grad_norm": 0.640625,
+      "learning_rate": 5.069889835993375e-05,
+      "loss": 1.0512192249298096,
+      "step": 2458
+    },
+    {
+      "epoch": 0.5353896398439626,
+      "grad_norm": 0.6484375,
+      "learning_rate": 5.0677289754026134e-05,
+      "loss": 1.0711238384246826,
+      "step": 2459
+    },
+    {
+      "epoch": 0.5356073664156763,
+      "grad_norm": 0.61328125,
+      "learning_rate": 5.0655677792562366e-05,
+      "loss": 0.9013359546661377,
+      "step": 2460
+    },
+    {
+      "epoch": 0.53582509298739,
+      "grad_norm": 0.66796875,
+      "learning_rate": 5.0634062482334455e-05,
+      "loss": 1.0412367582321167,
+      "step": 2461
+    },
+    {
+      "epoch": 0.5360428195591037,
+      "grad_norm": 0.63671875,
+      "learning_rate": 5.061244383013547e-05,
+      "loss": 0.9999368786811829,
+      "step": 2462
+    },
+    {
+      "epoch": 0.5362605461308174,
+      "grad_norm": 0.8515625,
+      "learning_rate": 5.059082184275949e-05,
+      "loss": 1.0772955417633057,
+      "step": 2463
+    },
+    {
+      "epoch": 0.5364782727025311,
+      "grad_norm": 0.64453125,
+      "learning_rate": 5.0569196527001666e-05,
+      "loss": 1.1339327096939087,
+      "step": 2464
+    },
+    {
+      "epoch": 0.5366959992742447,
+      "grad_norm": 0.63671875,
+      "learning_rate": 5.054756788965822e-05,
+      "loss": 1.1122698783874512,
+      "step": 2465
+    },
+    {
+      "epoch": 0.5369137258459584,
+      "grad_norm": 0.61328125,
+      "learning_rate": 5.052593593752637e-05,
+      "loss": 0.9101467132568359,
+      "step": 2466
+    },
+    {
+      "epoch": 0.5371314524176721,
+      "grad_norm": 0.67578125,
+      "learning_rate": 5.05043006774044e-05,
+      "loss": 1.0912138223648071,
+      "step": 2467
+    },
+    {
+      "epoch": 0.5373491789893858,
+      "grad_norm": 0.703125,
+      "learning_rate": 5.0482662116091625e-05,
+      "loss": 0.9888273477554321,
+      "step": 2468
+    },
+    {
+      "epoch": 0.5375669055610995,
+      "grad_norm": 0.6640625,
+      "learning_rate": 5.046102026038843e-05,
+      "loss": 1.1312240362167358,
+      "step": 2469
+    },
+    {
+      "epoch": 0.5377846321328132,
+      "grad_norm": 0.75,
+      "learning_rate": 5.043937511709619e-05,
+      "loss": 1.107961654663086,
+      "step": 2470
+    },
+    {
+      "epoch": 0.538002358704527,
+      "grad_norm": 0.703125,
+      "learning_rate": 5.0417726693017334e-05,
+      "loss": 0.9025374054908752,
+      "step": 2471
+    },
+    {
+      "epoch": 0.5382200852762405,
+      "grad_norm": 0.640625,
+      "learning_rate": 5.0396074994955324e-05,
+      "loss": 1.0713956356048584,
+      "step": 2472
+    },
+    {
+      "epoch": 0.5384378118479543,
+      "grad_norm": 0.640625,
+      "learning_rate": 5.0374420029714666e-05,
+      "loss": 1.0674488544464111,
+      "step": 2473
+    },
+    {
+      "epoch": 0.538655538419668,
+      "grad_norm": 0.58984375,
+      "learning_rate": 5.0352761804100835e-05,
+      "loss": 0.9096345901489258,
+      "step": 2474
+    },
+    {
+      "epoch": 0.5388732649913817,
+      "grad_norm": 0.7109375,
+      "learning_rate": 5.0331100324920405e-05,
+      "loss": 1.1243880987167358,
+      "step": 2475
+    },
+    {
+      "epoch": 0.5390909915630954,
+      "grad_norm": 0.671875,
+      "learning_rate": 5.0309435598980926e-05,
+      "loss": 1.0209523439407349,
+      "step": 2476
+    },
+    {
+      "epoch": 0.5393087181348091,
+      "grad_norm": 0.65625,
+      "learning_rate": 5.028776763309096e-05,
+      "loss": 0.9685603976249695,
+      "step": 2477
+    },
+    {
+      "epoch": 0.5395264447065228,
+      "grad_norm": 0.65234375,
+      "learning_rate": 5.026609643406014e-05,
+      "loss": 1.0635377168655396,
+      "step": 2478
+    },
+    {
+      "epoch": 0.5397441712782364,
+      "grad_norm": 0.69921875,
+      "learning_rate": 5.024442200869907e-05,
+      "loss": 0.9648100137710571,
+      "step": 2479
+    },
+    {
+      "epoch": 0.5399618978499501,
+      "grad_norm": 0.6328125,
+      "learning_rate": 5.022274436381938e-05,
+      "loss": 1.0243256092071533,
+      "step": 2480
+    },
+    {
+      "epoch": 0.5401796244216638,
+      "grad_norm": 0.63671875,
+      "learning_rate": 5.020106350623371e-05,
+      "loss": 1.0216054916381836,
+      "step": 2481
+    },
+    {
+      "epoch": 0.5403973509933775,
+      "grad_norm": 0.84375,
+      "learning_rate": 5.017937944275572e-05,
+      "loss": 1.0138481855392456,
+      "step": 2482
+    },
+    {
+      "epoch": 0.5406150775650912,
+      "grad_norm": 0.65625,
+      "learning_rate": 5.0157692180200046e-05,
+      "loss": 0.8821280002593994,
+      "step": 2483
+    },
+    {
+      "epoch": 0.5408328041368049,
+      "grad_norm": 0.63671875,
+      "learning_rate": 5.013600172538238e-05,
+      "loss": 0.9896941184997559,
+      "step": 2484
+    },
+    {
+      "epoch": 0.5410505307085186,
+      "grad_norm": 0.6796875,
+      "learning_rate": 5.0114308085119375e-05,
+      "loss": 0.9831025004386902,
+      "step": 2485
+    },
+    {
+      "epoch": 0.5412682572802322,
+      "grad_norm": 0.67578125,
+      "learning_rate": 5.00926112662287e-05,
+      "loss": 1.0351450443267822,
+      "step": 2486
+    },
+    {
+      "epoch": 0.5414859838519459,
+      "grad_norm": 0.64453125,
+      "learning_rate": 5.0070911275529035e-05,
+      "loss": 1.0529847145080566,
+      "step": 2487
+    },
+    {
+      "epoch": 0.5417037104236596,
+      "grad_norm": 0.61328125,
+      "learning_rate": 5.004920811984004e-05,
+      "loss": 0.9695937633514404,
+      "step": 2488
+    },
+    {
+      "epoch": 0.5419214369953733,
+      "grad_norm": 0.6484375,
+      "learning_rate": 5.002750180598236e-05,
+      "loss": 0.9636406302452087,
+      "step": 2489
+    },
+    {
+      "epoch": 0.542139163567087,
+      "grad_norm": 0.64453125,
+      "learning_rate": 5.0005792340777665e-05,
+      "loss": 1.0438737869262695,
+      "step": 2490
+    },
+    {
+      "epoch": 0.5423568901388007,
+      "grad_norm": 0.66796875,
+      "learning_rate": 4.9984079731048594e-05,
+      "loss": 1.0804439783096313,
+      "step": 2491
+    },
+    {
+      "epoch": 0.5425746167105143,
+      "grad_norm": 0.63671875,
+      "learning_rate": 4.996236398361878e-05,
+      "loss": 1.049043893814087,
+      "step": 2492
+    },
+    {
+      "epoch": 0.542792343282228,
+      "grad_norm": 0.6484375,
+      "learning_rate": 4.9940645105312817e-05,
+      "loss": 1.1245770454406738,
+      "step": 2493
+    },
+    {
+      "epoch": 0.5430100698539417,
+      "grad_norm": 0.63671875,
+      "learning_rate": 4.991892310295633e-05,
+      "loss": 1.0100719928741455,
+      "step": 2494
+    },
+    {
+      "epoch": 0.5432277964256554,
+      "grad_norm": 0.671875,
+      "learning_rate": 4.98971979833759e-05,
+      "loss": 1.0663225650787354,
+      "step": 2495
+    },
+    {
+      "epoch": 0.5434455229973691,
+      "grad_norm": 0.62890625,
+      "learning_rate": 4.987546975339908e-05,
+      "loss": 0.9604055285453796,
+      "step": 2496
+    },
+    {
+      "epoch": 0.5436632495690829,
+      "grad_norm": 0.65625,
+      "learning_rate": 4.98537384198544e-05,
+      "loss": 1.0888320207595825,
+      "step": 2497
+    },
+    {
+      "epoch": 0.5438809761407966,
+      "grad_norm": 0.69140625,
+      "learning_rate": 4.983200398957141e-05,
+      "loss": 1.1172970533370972,
+      "step": 2498
+    },
+    {
+      "epoch": 0.5440987027125102,
+      "grad_norm": 0.62109375,
+      "learning_rate": 4.9810266469380563e-05,
+      "loss": 0.8975315093994141,
+      "step": 2499
+    },
+    {
+      "epoch": 0.5443164292842239,
+      "grad_norm": 0.63671875,
+      "learning_rate": 4.978852586611333e-05,
+      "loss": 1.107672095298767,
+      "step": 2500
+    },
+    {
+      "epoch": 0.5445341558559376,
+      "grad_norm": 0.609375,
+      "learning_rate": 4.976678218660215e-05,
+      "loss": 1.0678699016571045,
+      "step": 2501
+    },
+    {
+      "epoch": 0.5447518824276513,
+      "grad_norm": 0.6875,
+      "learning_rate": 4.974503543768039e-05,
+      "loss": 1.0138591527938843,
+      "step": 2502
+    },
+    {
+      "epoch": 0.544969608999365,
+      "grad_norm": 0.625,
+      "learning_rate": 4.972328562618244e-05,
+      "loss": 0.9653570055961609,
+      "step": 2503
+    },
+    {
+      "epoch": 0.5451873355710787,
+      "grad_norm": 0.609375,
+      "learning_rate": 4.970153275894361e-05,
+      "loss": 0.867462694644928,
+      "step": 2504
+    },
+    {
+      "epoch": 0.5454050621427924,
+      "grad_norm": 0.65234375,
+      "learning_rate": 4.9679776842800174e-05,
+      "loss": 0.9209493398666382,
+      "step": 2505
+    },
+    {
+      "epoch": 0.545622788714506,
+      "grad_norm": 0.65234375,
+      "learning_rate": 4.965801788458938e-05,
+      "loss": 1.0316194295883179,
+      "step": 2506
+    },
+    {
+      "epoch": 0.5458405152862197,
+      "grad_norm": 0.63671875,
+      "learning_rate": 4.963625589114943e-05,
+      "loss": 1.1216145753860474,
+      "step": 2507
+    },
+    {
+      "epoch": 0.5460582418579334,
+      "grad_norm": 0.65625,
+      "learning_rate": 4.961449086931947e-05,
+      "loss": 1.1638507843017578,
+      "step": 2508
+    },
+    {
+      "epoch": 0.5462759684296471,
+      "grad_norm": 0.6328125,
+      "learning_rate": 4.9592722825939615e-05,
+      "loss": 0.9778975248336792,
+      "step": 2509
+    },
+    {
+      "epoch": 0.5464936950013608,
+      "grad_norm": 0.6328125,
+      "learning_rate": 4.957095176785088e-05,
+      "loss": 1.1465274095535278,
+      "step": 2510
+    },
+    {
+      "epoch": 0.5467114215730745,
+      "grad_norm": 0.6484375,
+      "learning_rate": 4.9549177701895295e-05,
+      "loss": 1.1956334114074707,
+      "step": 2511
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 5742,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 2,
+  "save_steps": 1722,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": false,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 5.260339867623933e+18,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}