diff --git "a/checkpoint-9000/trainer_state.json" "b/checkpoint-9000/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-9000/trainer_state.json"
@@ -0,0 +1,63033 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.5016722408026756,
+  "eval_steps": 500,
+  "global_step": 9000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 5.574136008918618e-05,
+      "grad_norm": 0.8008168339729309,
+      "learning_rate": 8e-05,
+      "loss": 2.3659,
+      "step": 1
+    },
+    {
+      "epoch": 0.00011148272017837236,
+      "grad_norm": 0.4908035099506378,
+      "learning_rate": 8e-05,
+      "loss": 1.7515,
+      "step": 2
+    },
+    {
+      "epoch": 0.00016722408026755852,
+      "grad_norm": 0.5122831463813782,
+      "learning_rate": 8e-05,
+      "loss": 1.998,
+      "step": 3
+    },
+    {
+      "epoch": 0.0002229654403567447,
+      "grad_norm": 0.47656959295272827,
+      "learning_rate": 8e-05,
+      "loss": 1.7709,
+      "step": 4
+    },
+    {
+      "epoch": 0.0002787068004459309,
+      "grad_norm": 0.47274479269981384,
+      "learning_rate": 8e-05,
+      "loss": 1.9645,
+      "step": 5
+    },
+    {
+      "epoch": 0.00033444816053511704,
+      "grad_norm": 0.42502275109291077,
+      "learning_rate": 8e-05,
+      "loss": 1.8408,
+      "step": 6
+    },
+    {
+      "epoch": 0.00039018952062430326,
+      "grad_norm": 0.4569794535636902,
+      "learning_rate": 8e-05,
+      "loss": 1.9836,
+      "step": 7
+    },
+    {
+      "epoch": 0.0004459308807134894,
+      "grad_norm": 0.43709850311279297,
+      "learning_rate": 8e-05,
+      "loss": 1.6727,
+      "step": 8
+    },
+    {
+      "epoch": 0.0005016722408026755,
+      "grad_norm": 0.5758728384971619,
+      "learning_rate": 8e-05,
+      "loss": 2.1397,
+      "step": 9
+    },
+    {
+      "epoch": 0.0005574136008918618,
+      "grad_norm": 0.5145262479782104,
+      "learning_rate": 8e-05,
+      "loss": 1.8755,
+      "step": 10
+    },
+    {
+      "epoch": 0.000613154960981048,
+      "grad_norm": 0.46036529541015625,
+      "learning_rate": 8e-05,
+      "loss": 1.8653,
+      "step": 11
+    },
+    {
+      "epoch": 0.0006688963210702341,
+      "grad_norm": 0.45718926191329956,
+      "learning_rate": 8e-05,
+      "loss": 1.7391,
+      "step": 12
+    },
+    {
+      "epoch": 0.0007246376811594203,
+      "grad_norm": 0.45198720693588257,
+      "learning_rate": 8e-05,
+      "loss": 1.8477,
+      "step": 13
+    },
+    {
+      "epoch": 0.0007803790412486065,
+      "grad_norm": 0.4350980222225189,
+      "learning_rate": 8e-05,
+      "loss": 1.8566,
+      "step": 14
+    },
+    {
+      "epoch": 0.0008361204013377926,
+      "grad_norm": 0.5104784369468689,
+      "learning_rate": 8e-05,
+      "loss": 1.8545,
+      "step": 15
+    },
+    {
+      "epoch": 0.0008918617614269788,
+      "grad_norm": 0.4408394992351532,
+      "learning_rate": 8e-05,
+      "loss": 1.9149,
+      "step": 16
+    },
+    {
+      "epoch": 0.000947603121516165,
+      "grad_norm": 0.45479220151901245,
+      "learning_rate": 8e-05,
+      "loss": 2.0529,
+      "step": 17
+    },
+    {
+      "epoch": 0.001003344481605351,
+      "grad_norm": 0.5046110153198242,
+      "learning_rate": 8e-05,
+      "loss": 1.8707,
+      "step": 18
+    },
+    {
+      "epoch": 0.0010590858416945374,
+      "grad_norm": 0.44725045561790466,
+      "learning_rate": 8e-05,
+      "loss": 1.7574,
+      "step": 19
+    },
+    {
+      "epoch": 0.0011148272017837235,
+      "grad_norm": 0.4346023499965668,
+      "learning_rate": 8e-05,
+      "loss": 1.8512,
+      "step": 20
+    },
+    {
+      "epoch": 0.0011705685618729096,
+      "grad_norm": 0.40324667096138,
+      "learning_rate": 8e-05,
+      "loss": 1.7462,
+      "step": 21
+    },
+    {
+      "epoch": 0.001226309921962096,
+      "grad_norm": 0.47196584939956665,
+      "learning_rate": 8e-05,
+      "loss": 1.9381,
+      "step": 22
+    },
+    {
+      "epoch": 0.001282051282051282,
+      "grad_norm": 0.4341256022453308,
+      "learning_rate": 8e-05,
+      "loss": 1.7059,
+      "step": 23
+    },
+    {
+      "epoch": 0.0013377926421404682,
+      "grad_norm": 0.4710449278354645,
+      "learning_rate": 8e-05,
+      "loss": 2.167,
+      "step": 24
+    },
+    {
+      "epoch": 0.0013935340022296545,
+      "grad_norm": 0.4591739773750305,
+      "learning_rate": 8e-05,
+      "loss": 2.0049,
+      "step": 25
+    },
+    {
+      "epoch": 0.0014492753623188406,
+      "grad_norm": 0.4278070330619812,
+      "learning_rate": 8e-05,
+      "loss": 1.6577,
+      "step": 26
+    },
+    {
+      "epoch": 0.0015050167224080267,
+      "grad_norm": 0.4664648771286011,
+      "learning_rate": 8e-05,
+      "loss": 2.1976,
+      "step": 27
+    },
+    {
+      "epoch": 0.001560758082497213,
+      "grad_norm": 0.44260603189468384,
+      "learning_rate": 8e-05,
+      "loss": 1.9631,
+      "step": 28
+    },
+    {
+      "epoch": 0.0016164994425863991,
+      "grad_norm": 0.4564704895019531,
+      "learning_rate": 8e-05,
+      "loss": 2.19,
+      "step": 29
+    },
+    {
+      "epoch": 0.0016722408026755853,
+      "grad_norm": 0.4133101999759674,
+      "learning_rate": 8e-05,
+      "loss": 1.6704,
+      "step": 30
+    },
+    {
+      "epoch": 0.0017279821627647714,
+      "grad_norm": 0.46999362111091614,
+      "learning_rate": 8e-05,
+      "loss": 1.8143,
+      "step": 31
+    },
+    {
+      "epoch": 0.0017837235228539577,
+      "grad_norm": 0.4743380546569824,
+      "learning_rate": 8e-05,
+      "loss": 1.9599,
+      "step": 32
+    },
+    {
+      "epoch": 0.0018394648829431438,
+      "grad_norm": 0.4114800691604614,
+      "learning_rate": 8e-05,
+      "loss": 1.7434,
+      "step": 33
+    },
+    {
+      "epoch": 0.00189520624303233,
+      "grad_norm": 0.4391994774341583,
+      "learning_rate": 8e-05,
+      "loss": 1.6625,
+      "step": 34
+    },
+    {
+      "epoch": 0.0019509476031215162,
+      "grad_norm": 0.41881778836250305,
+      "learning_rate": 8e-05,
+      "loss": 1.8337,
+      "step": 35
+    },
+    {
+      "epoch": 0.002006688963210702,
+      "grad_norm": 0.40023884177207947,
+      "learning_rate": 8e-05,
+      "loss": 1.8276,
+      "step": 36
+    },
+    {
+      "epoch": 0.0020624303232998887,
+      "grad_norm": 0.4486073851585388,
+      "learning_rate": 8e-05,
+      "loss": 2.1585,
+      "step": 37
+    },
+    {
+      "epoch": 0.002118171683389075,
+      "grad_norm": 0.442481130361557,
+      "learning_rate": 8e-05,
+      "loss": 1.8827,
+      "step": 38
+    },
+    {
+      "epoch": 0.002173913043478261,
+      "grad_norm": 0.41534623503685,
+      "learning_rate": 8e-05,
+      "loss": 2.1137,
+      "step": 39
+    },
+    {
+      "epoch": 0.002229654403567447,
+      "grad_norm": 0.40957891941070557,
+      "learning_rate": 8e-05,
+      "loss": 1.8755,
+      "step": 40
+    },
+    {
+      "epoch": 0.002285395763656633,
+      "grad_norm": 0.42762520909309387,
+      "learning_rate": 8e-05,
+      "loss": 1.9187,
+      "step": 41
+    },
+    {
+      "epoch": 0.0023411371237458192,
+      "grad_norm": 0.43732979893684387,
+      "learning_rate": 8e-05,
+      "loss": 1.8745,
+      "step": 42
+    },
+    {
+      "epoch": 0.0023968784838350058,
+      "grad_norm": 0.42588138580322266,
+      "learning_rate": 8e-05,
+      "loss": 1.79,
+      "step": 43
+    },
+    {
+      "epoch": 0.002452619843924192,
+      "grad_norm": 0.4093983769416809,
+      "learning_rate": 8e-05,
+      "loss": 1.7689,
+      "step": 44
+    },
+    {
+      "epoch": 0.002508361204013378,
+      "grad_norm": 0.39260199666023254,
+      "learning_rate": 8e-05,
+      "loss": 1.8451,
+      "step": 45
+    },
+    {
+      "epoch": 0.002564102564102564,
+      "grad_norm": 0.3898474872112274,
+      "learning_rate": 8e-05,
+      "loss": 1.6764,
+      "step": 46
+    },
+    {
+      "epoch": 0.00261984392419175,
+      "grad_norm": 0.4310983121395111,
+      "learning_rate": 8e-05,
+      "loss": 1.8254,
+      "step": 47
+    },
+    {
+      "epoch": 0.0026755852842809363,
+      "grad_norm": 0.45289134979248047,
+      "learning_rate": 8e-05,
+      "loss": 2.0347,
+      "step": 48
+    },
+    {
+      "epoch": 0.0027313266443701224,
+      "grad_norm": 0.3983175754547119,
+      "learning_rate": 8e-05,
+      "loss": 1.5513,
+      "step": 49
+    },
+    {
+      "epoch": 0.002787068004459309,
+      "grad_norm": 0.3877411186695099,
+      "learning_rate": 8e-05,
+      "loss": 1.7061,
+      "step": 50
+    },
+    {
+      "epoch": 0.002842809364548495,
+      "grad_norm": 0.4502890706062317,
+      "learning_rate": 8e-05,
+      "loss": 2.0879,
+      "step": 51
+    },
+    {
+      "epoch": 0.002898550724637681,
+      "grad_norm": 0.3804650902748108,
+      "learning_rate": 8e-05,
+      "loss": 1.5597,
+      "step": 52
+    },
+    {
+      "epoch": 0.0029542920847268673,
+      "grad_norm": 0.4189783036708832,
+      "learning_rate": 8e-05,
+      "loss": 2.0595,
+      "step": 53
+    },
+    {
+      "epoch": 0.0030100334448160534,
+      "grad_norm": 0.38596734404563904,
+      "learning_rate": 8e-05,
+      "loss": 1.8491,
+      "step": 54
+    },
+    {
+      "epoch": 0.0030657748049052395,
+      "grad_norm": 0.45324382185935974,
+      "learning_rate": 8e-05,
+      "loss": 2.12,
+      "step": 55
+    },
+    {
+      "epoch": 0.003121516164994426,
+      "grad_norm": 0.42319026589393616,
+      "learning_rate": 8e-05,
+      "loss": 1.9806,
+      "step": 56
+    },
+    {
+      "epoch": 0.003177257525083612,
+      "grad_norm": 0.3817268908023834,
+      "learning_rate": 8e-05,
+      "loss": 1.7646,
+      "step": 57
+    },
+    {
+      "epoch": 0.0032329988851727983,
+      "grad_norm": 0.4332798421382904,
+      "learning_rate": 8e-05,
+      "loss": 1.9969,
+      "step": 58
+    },
+    {
+      "epoch": 0.0032887402452619844,
+      "grad_norm": 0.43772000074386597,
+      "learning_rate": 8e-05,
+      "loss": 1.8521,
+      "step": 59
+    },
+    {
+      "epoch": 0.0033444816053511705,
+      "grad_norm": 0.4649338722229004,
+      "learning_rate": 8e-05,
+      "loss": 1.832,
+      "step": 60
+    },
+    {
+      "epoch": 0.0034002229654403566,
+      "grad_norm": 0.3952140808105469,
+      "learning_rate": 8e-05,
+      "loss": 1.8608,
+      "step": 61
+    },
+    {
+      "epoch": 0.0034559643255295427,
+      "grad_norm": 0.4187444746494293,
+      "learning_rate": 8e-05,
+      "loss": 1.8068,
+      "step": 62
+    },
+    {
+      "epoch": 0.0035117056856187293,
+      "grad_norm": 0.39462578296661377,
+      "learning_rate": 8e-05,
+      "loss": 1.7132,
+      "step": 63
+    },
+    {
+      "epoch": 0.0035674470457079154,
+      "grad_norm": 0.4504048526287079,
+      "learning_rate": 8e-05,
+      "loss": 2.0016,
+      "step": 64
+    },
+    {
+      "epoch": 0.0036231884057971015,
+      "grad_norm": 0.40093812346458435,
+      "learning_rate": 8e-05,
+      "loss": 1.7959,
+      "step": 65
+    },
+    {
+      "epoch": 0.0036789297658862876,
+      "grad_norm": 0.454746812582016,
+      "learning_rate": 8e-05,
+      "loss": 1.7046,
+      "step": 66
+    },
+    {
+      "epoch": 0.0037346711259754737,
+      "grad_norm": 0.3936002254486084,
+      "learning_rate": 8e-05,
+      "loss": 1.7267,
+      "step": 67
+    },
+    {
+      "epoch": 0.00379041248606466,
+      "grad_norm": 0.3876648247241974,
+      "learning_rate": 8e-05,
+      "loss": 1.7422,
+      "step": 68
+    },
+    {
+      "epoch": 0.0038461538461538464,
+      "grad_norm": 0.44103604555130005,
+      "learning_rate": 8e-05,
+      "loss": 1.8784,
+      "step": 69
+    },
+    {
+      "epoch": 0.0039018952062430325,
+      "grad_norm": 0.4045323431491852,
+      "learning_rate": 8e-05,
+      "loss": 1.7219,
+      "step": 70
+    },
+    {
+      "epoch": 0.003957636566332219,
+      "grad_norm": 0.45244699716567993,
+      "learning_rate": 8e-05,
+      "loss": 2.0911,
+      "step": 71
+    },
+    {
+      "epoch": 0.004013377926421404,
+      "grad_norm": 0.4084925055503845,
+      "learning_rate": 8e-05,
+      "loss": 1.8906,
+      "step": 72
+    },
+    {
+      "epoch": 0.004069119286510591,
+      "grad_norm": 0.3983491063117981,
+      "learning_rate": 8e-05,
+      "loss": 1.5774,
+      "step": 73
+    },
+    {
+      "epoch": 0.004124860646599777,
+      "grad_norm": 0.4227916896343231,
+      "learning_rate": 8e-05,
+      "loss": 1.7904,
+      "step": 74
+    },
+    {
+      "epoch": 0.004180602006688963,
+      "grad_norm": 0.41621050238609314,
+      "learning_rate": 8e-05,
+      "loss": 1.7614,
+      "step": 75
+    },
+    {
+      "epoch": 0.00423634336677815,
+      "grad_norm": 0.4000743329524994,
+      "learning_rate": 8e-05,
+      "loss": 1.8071,
+      "step": 76
+    },
+    {
+      "epoch": 0.004292084726867335,
+      "grad_norm": 0.41884997487068176,
+      "learning_rate": 8e-05,
+      "loss": 1.8346,
+      "step": 77
+    },
+    {
+      "epoch": 0.004347826086956522,
+      "grad_norm": 0.38804078102111816,
+      "learning_rate": 8e-05,
+      "loss": 1.8449,
+      "step": 78
+    },
+    {
+      "epoch": 0.004403567447045708,
+      "grad_norm": 0.46406039595603943,
+      "learning_rate": 8e-05,
+      "loss": 2.236,
+      "step": 79
+    },
+    {
+      "epoch": 0.004459308807134894,
+      "grad_norm": 0.5488149523735046,
+      "learning_rate": 8e-05,
+      "loss": 1.9053,
+      "step": 80
+    },
+    {
+      "epoch": 0.0045150501672240806,
+      "grad_norm": 0.41493290662765503,
+      "learning_rate": 8e-05,
+      "loss": 1.9203,
+      "step": 81
+    },
+    {
+      "epoch": 0.004570791527313266,
+      "grad_norm": 0.44460850954055786,
+      "learning_rate": 8e-05,
+      "loss": 1.9612,
+      "step": 82
+    },
+    {
+      "epoch": 0.004626532887402453,
+      "grad_norm": 0.42977744340896606,
+      "learning_rate": 8e-05,
+      "loss": 1.9116,
+      "step": 83
+    },
+    {
+      "epoch": 0.0046822742474916385,
+      "grad_norm": 0.39066213369369507,
+      "learning_rate": 8e-05,
+      "loss": 1.881,
+      "step": 84
+    },
+    {
+      "epoch": 0.004738015607580825,
+      "grad_norm": 0.39784467220306396,
+      "learning_rate": 8e-05,
+      "loss": 1.5992,
+      "step": 85
+    },
+    {
+      "epoch": 0.0047937569676700115,
+      "grad_norm": 0.4012242555618286,
+      "learning_rate": 8e-05,
+      "loss": 1.8036,
+      "step": 86
+    },
+    {
+      "epoch": 0.004849498327759197,
+      "grad_norm": 0.3906935751438141,
+      "learning_rate": 8e-05,
+      "loss": 1.6839,
+      "step": 87
+    },
+    {
+      "epoch": 0.004905239687848384,
+      "grad_norm": 0.47963079810142517,
+      "learning_rate": 8e-05,
+      "loss": 1.9194,
+      "step": 88
+    },
+    {
+      "epoch": 0.0049609810479375694,
+      "grad_norm": 0.4412635862827301,
+      "learning_rate": 8e-05,
+      "loss": 1.7194,
+      "step": 89
+    },
+    {
+      "epoch": 0.005016722408026756,
+      "grad_norm": 0.40161287784576416,
+      "learning_rate": 8e-05,
+      "loss": 1.9614,
+      "step": 90
+    },
+    {
+      "epoch": 0.005072463768115942,
+      "grad_norm": 0.37398388981819153,
+      "learning_rate": 8e-05,
+      "loss": 1.5069,
+      "step": 91
+    },
+    {
+      "epoch": 0.005128205128205128,
+      "grad_norm": 0.4409233629703522,
+      "learning_rate": 8e-05,
+      "loss": 1.9873,
+      "step": 92
+    },
+    {
+      "epoch": 0.005183946488294315,
+      "grad_norm": 0.3823505938053131,
+      "learning_rate": 8e-05,
+      "loss": 1.8509,
+      "step": 93
+    },
+    {
+      "epoch": 0.0052396878483835,
+      "grad_norm": 0.41670355200767517,
+      "learning_rate": 8e-05,
+      "loss": 2.0128,
+      "step": 94
+    },
+    {
+      "epoch": 0.005295429208472687,
+      "grad_norm": 0.4004131257534027,
+      "learning_rate": 8e-05,
+      "loss": 1.9371,
+      "step": 95
+    },
+    {
+      "epoch": 0.005351170568561873,
+      "grad_norm": 0.40070030093193054,
+      "learning_rate": 8e-05,
+      "loss": 1.8621,
+      "step": 96
+    },
+    {
+      "epoch": 0.005406911928651059,
+      "grad_norm": 0.4367406964302063,
+      "learning_rate": 8e-05,
+      "loss": 1.9458,
+      "step": 97
+    },
+    {
+      "epoch": 0.005462653288740245,
+      "grad_norm": 0.4184412658214569,
+      "learning_rate": 8e-05,
+      "loss": 1.9502,
+      "step": 98
+    },
+    {
+      "epoch": 0.005518394648829431,
+      "grad_norm": 0.41309577226638794,
+      "learning_rate": 8e-05,
+      "loss": 1.885,
+      "step": 99
+    },
+    {
+      "epoch": 0.005574136008918618,
+      "grad_norm": 0.4838528633117676,
+      "learning_rate": 8e-05,
+      "loss": 1.995,
+      "step": 100
+    },
+    {
+      "epoch": 0.005629877369007804,
+      "grad_norm": 0.39496132731437683,
+      "learning_rate": 8e-05,
+      "loss": 1.7633,
+      "step": 101
+    },
+    {
+      "epoch": 0.00568561872909699,
+      "grad_norm": 0.42587682604789734,
+      "learning_rate": 8e-05,
+      "loss": 1.9289,
+      "step": 102
+    },
+    {
+      "epoch": 0.005741360089186176,
+      "grad_norm": 0.4074772596359253,
+      "learning_rate": 8e-05,
+      "loss": 1.8369,
+      "step": 103
+    },
+    {
+      "epoch": 0.005797101449275362,
+      "grad_norm": 0.3786422610282898,
+      "learning_rate": 8e-05,
+      "loss": 1.6762,
+      "step": 104
+    },
+    {
+      "epoch": 0.005852842809364548,
+      "grad_norm": 0.41452324390411377,
+      "learning_rate": 8e-05,
+      "loss": 1.7729,
+      "step": 105
+    },
+    {
+      "epoch": 0.005908584169453735,
+      "grad_norm": 0.39360037446022034,
+      "learning_rate": 8e-05,
+      "loss": 1.8084,
+      "step": 106
+    },
+    {
+      "epoch": 0.005964325529542921,
+      "grad_norm": 0.44575709104537964,
+      "learning_rate": 8e-05,
+      "loss": 1.9992,
+      "step": 107
+    },
+    {
+      "epoch": 0.006020066889632107,
+      "grad_norm": 0.378961980342865,
+      "learning_rate": 8e-05,
+      "loss": 1.6555,
+      "step": 108
+    },
+    {
+      "epoch": 0.006075808249721293,
+      "grad_norm": 0.3773297667503357,
+      "learning_rate": 8e-05,
+      "loss": 1.6608,
+      "step": 109
+    },
+    {
+      "epoch": 0.006131549609810479,
+      "grad_norm": 0.3853168189525604,
+      "learning_rate": 8e-05,
+      "loss": 1.4821,
+      "step": 110
+    },
+    {
+      "epoch": 0.006187290969899666,
+      "grad_norm": 0.48528149724006653,
+      "learning_rate": 8e-05,
+      "loss": 1.9857,
+      "step": 111
+    },
+    {
+      "epoch": 0.006243032329988852,
+      "grad_norm": 0.435882031917572,
+      "learning_rate": 8e-05,
+      "loss": 1.7617,
+      "step": 112
+    },
+    {
+      "epoch": 0.006298773690078038,
+      "grad_norm": 0.39531388878822327,
+      "learning_rate": 8e-05,
+      "loss": 1.5861,
+      "step": 113
+    },
+    {
+      "epoch": 0.006354515050167224,
+      "grad_norm": 0.43698954582214355,
+      "learning_rate": 8e-05,
+      "loss": 1.9298,
+      "step": 114
+    },
+    {
+      "epoch": 0.00641025641025641,
+      "grad_norm": 0.403531938791275,
+      "learning_rate": 8e-05,
+      "loss": 1.8521,
+      "step": 115
+    },
+    {
+      "epoch": 0.006465997770345597,
+      "grad_norm": 0.4131438434123993,
+      "learning_rate": 8e-05,
+      "loss": 1.8334,
+      "step": 116
+    },
+    {
+      "epoch": 0.006521739130434782,
+      "grad_norm": 0.37669649720191956,
+      "learning_rate": 8e-05,
+      "loss": 1.7758,
+      "step": 117
+    },
+    {
+      "epoch": 0.006577480490523969,
+      "grad_norm": 0.4128376543521881,
+      "learning_rate": 8e-05,
+      "loss": 1.9285,
+      "step": 118
+    },
+    {
+      "epoch": 0.006633221850613155,
+      "grad_norm": 0.40930864214897156,
+      "learning_rate": 8e-05,
+      "loss": 1.9049,
+      "step": 119
+    },
+    {
+      "epoch": 0.006688963210702341,
+      "grad_norm": 0.42722001671791077,
+      "learning_rate": 8e-05,
+      "loss": 1.9176,
+      "step": 120
+    },
+    {
+      "epoch": 0.0067447045707915276,
+      "grad_norm": 0.4142158031463623,
+      "learning_rate": 8e-05,
+      "loss": 1.9328,
+      "step": 121
+    },
+    {
+      "epoch": 0.006800445930880713,
+      "grad_norm": 0.4357644319534302,
+      "learning_rate": 8e-05,
+      "loss": 1.7735,
+      "step": 122
+    },
+    {
+      "epoch": 0.0068561872909699,
+      "grad_norm": 0.39733514189720154,
+      "learning_rate": 8e-05,
+      "loss": 1.7758,
+      "step": 123
+    },
+    {
+      "epoch": 0.0069119286510590855,
+      "grad_norm": 0.3832526206970215,
+      "learning_rate": 8e-05,
+      "loss": 1.7294,
+      "step": 124
+    },
+    {
+      "epoch": 0.006967670011148272,
+      "grad_norm": 0.4044865369796753,
+      "learning_rate": 8e-05,
+      "loss": 1.7418,
+      "step": 125
+    },
+    {
+      "epoch": 0.0070234113712374585,
+      "grad_norm": 0.41378268599510193,
+      "learning_rate": 8e-05,
+      "loss": 1.7275,
+      "step": 126
+    },
+    {
+      "epoch": 0.007079152731326644,
+      "grad_norm": 0.4132898151874542,
+      "learning_rate": 8e-05,
+      "loss": 1.8666,
+      "step": 127
+    },
+    {
+      "epoch": 0.007134894091415831,
+      "grad_norm": 0.4443804919719696,
+      "learning_rate": 8e-05,
+      "loss": 1.9086,
+      "step": 128
+    },
+    {
+      "epoch": 0.0071906354515050164,
+      "grad_norm": 0.41749894618988037,
+      "learning_rate": 8e-05,
+      "loss": 1.8343,
+      "step": 129
+    },
+    {
+      "epoch": 0.007246376811594203,
+      "grad_norm": 0.38238924741744995,
+      "learning_rate": 8e-05,
+      "loss": 1.4543,
+      "step": 130
+    },
+    {
+      "epoch": 0.007302118171683389,
+      "grad_norm": 0.47451648116111755,
+      "learning_rate": 8e-05,
+      "loss": 1.8353,
+      "step": 131
+    },
+    {
+      "epoch": 0.007357859531772575,
+      "grad_norm": 0.4221440851688385,
+      "learning_rate": 8e-05,
+      "loss": 1.8808,
+      "step": 132
+    },
+    {
+      "epoch": 0.007413600891861762,
+      "grad_norm": 0.4281880557537079,
+      "learning_rate": 8e-05,
+      "loss": 1.9372,
+      "step": 133
+    },
+    {
+      "epoch": 0.007469342251950947,
+      "grad_norm": 0.4605168104171753,
+      "learning_rate": 8e-05,
+      "loss": 2.0274,
+      "step": 134
+    },
+    {
+      "epoch": 0.007525083612040134,
+      "grad_norm": 0.38289088010787964,
+      "learning_rate": 8e-05,
+      "loss": 1.7703,
+      "step": 135
+    },
+    {
+      "epoch": 0.00758082497212932,
+      "grad_norm": 0.40322351455688477,
+      "learning_rate": 8e-05,
+      "loss": 1.8378,
+      "step": 136
+    },
+    {
+      "epoch": 0.007636566332218506,
+      "grad_norm": 0.40321865677833557,
+      "learning_rate": 8e-05,
+      "loss": 1.6512,
+      "step": 137
+    },
+    {
+      "epoch": 0.007692307692307693,
+      "grad_norm": 0.3676503598690033,
+      "learning_rate": 8e-05,
+      "loss": 1.6513,
+      "step": 138
+    },
+    {
+      "epoch": 0.007748049052396878,
+      "grad_norm": 0.3847725987434387,
+      "learning_rate": 8e-05,
+      "loss": 1.8038,
+      "step": 139
+    },
+    {
+      "epoch": 0.007803790412486065,
+      "grad_norm": 0.4042004942893982,
+      "learning_rate": 8e-05,
+      "loss": 1.9278,
+      "step": 140
+    },
+    {
+      "epoch": 0.007859531772575251,
+      "grad_norm": 0.4066566824913025,
+      "learning_rate": 8e-05,
+      "loss": 1.7756,
+      "step": 141
+    },
+    {
+      "epoch": 0.007915273132664437,
+      "grad_norm": 0.39882540702819824,
+      "learning_rate": 8e-05,
+      "loss": 1.8299,
+      "step": 142
+    },
+    {
+      "epoch": 0.007971014492753623,
+      "grad_norm": 0.3843311071395874,
+      "learning_rate": 8e-05,
+      "loss": 1.7093,
+      "step": 143
+    },
+    {
+      "epoch": 0.008026755852842809,
+      "grad_norm": 0.39015519618988037,
+      "learning_rate": 8e-05,
+      "loss": 1.7497,
+      "step": 144
+    },
+    {
+      "epoch": 0.008082497212931996,
+      "grad_norm": 0.40187525749206543,
+      "learning_rate": 8e-05,
+      "loss": 1.9149,
+      "step": 145
+    },
+    {
+      "epoch": 0.008138238573021182,
+      "grad_norm": 0.46802017092704773,
+      "learning_rate": 8e-05,
+      "loss": 2.0147,
+      "step": 146
+    },
+    {
+      "epoch": 0.008193979933110367,
+      "grad_norm": 0.40548357367515564,
+      "learning_rate": 8e-05,
+      "loss": 1.7083,
+      "step": 147
+    },
+    {
+      "epoch": 0.008249721293199555,
+      "grad_norm": 0.4108128547668457,
+      "learning_rate": 8e-05,
+      "loss": 1.7361,
+      "step": 148
+    },
+    {
+      "epoch": 0.00830546265328874,
+      "grad_norm": 0.4061950445175171,
+      "learning_rate": 8e-05,
+      "loss": 2.0837,
+      "step": 149
+    },
+    {
+      "epoch": 0.008361204013377926,
+      "grad_norm": 0.4076997935771942,
+      "learning_rate": 8e-05,
+      "loss": 1.6702,
+      "step": 150
+    },
+    {
+      "epoch": 0.008416945373467112,
+      "grad_norm": 0.40252676606178284,
+      "learning_rate": 8e-05,
+      "loss": 1.8599,
+      "step": 151
+    },
+    {
+      "epoch": 0.0084726867335563,
+      "grad_norm": 0.41289693117141724,
+      "learning_rate": 8e-05,
+      "loss": 1.9507,
+      "step": 152
+    },
+    {
+      "epoch": 0.008528428093645485,
+      "grad_norm": 0.406189501285553,
+      "learning_rate": 8e-05,
+      "loss": 1.7856,
+      "step": 153
+    },
+    {
+      "epoch": 0.00858416945373467,
+      "grad_norm": 0.40082770586013794,
+      "learning_rate": 8e-05,
+      "loss": 1.8779,
+      "step": 154
+    },
+    {
+      "epoch": 0.008639910813823858,
+      "grad_norm": 0.4297981560230255,
+      "learning_rate": 8e-05,
+      "loss": 1.6413,
+      "step": 155
+    },
+    {
+      "epoch": 0.008695652173913044,
+      "grad_norm": 0.42399170994758606,
+      "learning_rate": 8e-05,
+      "loss": 1.6762,
+      "step": 156
+    },
+    {
+      "epoch": 0.00875139353400223,
+      "grad_norm": 0.4469023644924164,
+      "learning_rate": 8e-05,
+      "loss": 2.2177,
+      "step": 157
+    },
+    {
+      "epoch": 0.008807134894091417,
+      "grad_norm": 0.41254499554634094,
+      "learning_rate": 8e-05,
+      "loss": 1.9285,
+      "step": 158
+    },
+    {
+      "epoch": 0.008862876254180602,
+      "grad_norm": 0.37810757756233215,
+      "learning_rate": 8e-05,
+      "loss": 1.6574,
+      "step": 159
+    },
+    {
+      "epoch": 0.008918617614269788,
+      "grad_norm": 0.5044556260108948,
+      "learning_rate": 8e-05,
+      "loss": 1.4282,
+      "step": 160
+    },
+    {
+      "epoch": 0.008974358974358974,
+      "grad_norm": 0.4354773759841919,
+      "learning_rate": 8e-05,
+      "loss": 1.963,
+      "step": 161
+    },
+    {
+      "epoch": 0.009030100334448161,
+      "grad_norm": 0.4584670662879944,
+      "learning_rate": 8e-05,
+      "loss": 2.0064,
+      "step": 162
+    },
+    {
+      "epoch": 0.009085841694537347,
+      "grad_norm": 0.4361465871334076,
+      "learning_rate": 8e-05,
+      "loss": 1.8227,
+      "step": 163
+    },
+    {
+      "epoch": 0.009141583054626532,
+      "grad_norm": 0.3891424834728241,
+      "learning_rate": 8e-05,
+      "loss": 1.5972,
+      "step": 164
+    },
+    {
+      "epoch": 0.00919732441471572,
+      "grad_norm": 0.44024306535720825,
+      "learning_rate": 8e-05,
+      "loss": 1.8614,
+      "step": 165
+    },
+    {
+      "epoch": 0.009253065774804906,
+      "grad_norm": 0.38277462124824524,
+      "learning_rate": 8e-05,
+      "loss": 1.8031,
+      "step": 166
+    },
+    {
+      "epoch": 0.009308807134894091,
+      "grad_norm": 0.46869274973869324,
+      "learning_rate": 8e-05,
+      "loss": 2.156,
+      "step": 167
+    },
+    {
+      "epoch": 0.009364548494983277,
+      "grad_norm": 0.4273754358291626,
+      "learning_rate": 8e-05,
+      "loss": 1.7214,
+      "step": 168
+    },
+    {
+      "epoch": 0.009420289855072464,
+      "grad_norm": 0.4267232418060303,
+      "learning_rate": 8e-05,
+      "loss": 1.89,
+      "step": 169
+    },
+    {
+      "epoch": 0.00947603121516165,
+      "grad_norm": 0.4204559028148651,
+      "learning_rate": 8e-05,
+      "loss": 1.937,
+      "step": 170
+    },
+    {
+      "epoch": 0.009531772575250836,
+      "grad_norm": 0.4121902585029602,
+      "learning_rate": 8e-05,
+      "loss": 1.9587,
+      "step": 171
+    },
+    {
+      "epoch": 0.009587513935340023,
+      "grad_norm": 0.4139564633369446,
+      "learning_rate": 8e-05,
+      "loss": 1.7723,
+      "step": 172
+    },
+    {
+      "epoch": 0.009643255295429209,
+      "grad_norm": 0.441736102104187,
+      "learning_rate": 8e-05,
+      "loss": 1.889,
+      "step": 173
+    },
+    {
+      "epoch": 0.009698996655518394,
+      "grad_norm": 0.4170135259628296,
+      "learning_rate": 8e-05,
+      "loss": 1.9719,
+      "step": 174
+    },
+    {
+      "epoch": 0.00975473801560758,
+      "grad_norm": 0.5091986060142517,
+      "learning_rate": 8e-05,
+      "loss": 1.8272,
+      "step": 175
+    },
+    {
+      "epoch": 0.009810479375696768,
+      "grad_norm": 0.391644150018692,
+      "learning_rate": 8e-05,
+      "loss": 1.6708,
+      "step": 176
+    },
+    {
+      "epoch": 0.009866220735785953,
+      "grad_norm": 0.4127022624015808,
+      "learning_rate": 8e-05,
+      "loss": 1.7926,
+      "step": 177
+    },
+    {
+      "epoch": 0.009921962095875139,
+      "grad_norm": 0.43958815932273865,
+      "learning_rate": 8e-05,
+      "loss": 1.8704,
+      "step": 178
+    },
+    {
+      "epoch": 0.009977703455964326,
+      "grad_norm": 0.46184685826301575,
+      "learning_rate": 8e-05,
+      "loss": 1.9766,
+      "step": 179
+    },
+    {
+      "epoch": 0.010033444816053512,
+      "grad_norm": 0.4079630374908447,
+      "learning_rate": 8e-05,
+      "loss": 1.6926,
+      "step": 180
+    },
+    {
+      "epoch": 0.010089186176142698,
+      "grad_norm": 0.3960077166557312,
+      "learning_rate": 8e-05,
+      "loss": 1.599,
+      "step": 181
+    },
+    {
+      "epoch": 0.010144927536231883,
+      "grad_norm": 0.37671923637390137,
+      "learning_rate": 8e-05,
+      "loss": 1.6455,
+      "step": 182
+    },
+    {
+      "epoch": 0.01020066889632107,
+      "grad_norm": 0.4062810242176056,
+      "learning_rate": 8e-05,
+      "loss": 1.9651,
+      "step": 183
+    },
+    {
+      "epoch": 0.010256410256410256,
+      "grad_norm": 0.4247036278247833,
+      "learning_rate": 8e-05,
+      "loss": 2.0219,
+      "step": 184
+    },
+    {
+      "epoch": 0.010312151616499442,
+      "grad_norm": 0.40789371728897095,
+      "learning_rate": 8e-05,
+      "loss": 1.9062,
+      "step": 185
+    },
+    {
+      "epoch": 0.01036789297658863,
+      "grad_norm": 0.4061046540737152,
+      "learning_rate": 8e-05,
+      "loss": 1.9277,
+      "step": 186
+    },
+    {
+      "epoch": 0.010423634336677815,
+      "grad_norm": 0.396852046251297,
+      "learning_rate": 8e-05,
+      "loss": 1.8621,
+      "step": 187
+    },
+    {
+      "epoch": 0.010479375696767,
+      "grad_norm": 0.4642435610294342,
+      "learning_rate": 8e-05,
+      "loss": 1.9323,
+      "step": 188
+    },
+    {
+      "epoch": 0.010535117056856187,
+      "grad_norm": 0.43151143193244934,
+      "learning_rate": 8e-05,
+      "loss": 1.8768,
+      "step": 189
+    },
+    {
+      "epoch": 0.010590858416945374,
+      "grad_norm": 0.4810677766799927,
+      "learning_rate": 8e-05,
+      "loss": 2.1525,
+      "step": 190
+    },
+    {
+      "epoch": 0.01064659977703456,
+      "grad_norm": 0.40719544887542725,
+      "learning_rate": 8e-05,
+      "loss": 1.8136,
+      "step": 191
+    },
+    {
+      "epoch": 0.010702341137123745,
+      "grad_norm": 0.4264082908630371,
+      "learning_rate": 8e-05,
+      "loss": 2.1189,
+      "step": 192
+    },
+    {
+      "epoch": 0.010758082497212933,
+      "grad_norm": 0.40869855880737305,
+      "learning_rate": 8e-05,
+      "loss": 1.8806,
+      "step": 193
+    },
+    {
+      "epoch": 0.010813823857302118,
+      "grad_norm": 0.45739123225212097,
+      "learning_rate": 8e-05,
+      "loss": 2.0878,
+      "step": 194
+    },
+    {
+      "epoch": 0.010869565217391304,
+      "grad_norm": 0.43634286522865295,
+      "learning_rate": 8e-05,
+      "loss": 1.8352,
+      "step": 195
+    },
+    {
+      "epoch": 0.01092530657748049,
+      "grad_norm": 0.41823098063468933,
+      "learning_rate": 8e-05,
+      "loss": 1.8458,
+      "step": 196
+    },
+    {
+      "epoch": 0.010981047937569677,
+      "grad_norm": 0.4044710695743561,
+      "learning_rate": 8e-05,
+      "loss": 1.9065,
+      "step": 197
+    },
+    {
+      "epoch": 0.011036789297658863,
+      "grad_norm": 0.4063010513782501,
+      "learning_rate": 8e-05,
+      "loss": 1.7212,
+      "step": 198
+    },
+    {
+      "epoch": 0.011092530657748048,
+      "grad_norm": 0.43073466420173645,
+      "learning_rate": 8e-05,
+      "loss": 1.9087,
+      "step": 199
+    },
+    {
+      "epoch": 0.011148272017837236,
+      "grad_norm": 0.4032673239707947,
+      "learning_rate": 8e-05,
+      "loss": 1.7571,
+      "step": 200
+    },
+    {
+      "epoch": 0.011204013377926422,
+      "grad_norm": 0.39450713992118835,
+      "learning_rate": 8e-05,
+      "loss": 1.7264,
+      "step": 201
+    },
+    {
+      "epoch": 0.011259754738015607,
+      "grad_norm": 0.40192824602127075,
+      "learning_rate": 8e-05,
+      "loss": 1.8697,
+      "step": 202
+    },
+    {
+      "epoch": 0.011315496098104793,
+      "grad_norm": 0.4325474798679352,
+      "learning_rate": 8e-05,
+      "loss": 1.8927,
+      "step": 203
+    },
+    {
+      "epoch": 0.01137123745819398,
+      "grad_norm": 0.38713356852531433,
+      "learning_rate": 8e-05,
+      "loss": 1.6918,
+      "step": 204
+    },
+    {
+      "epoch": 0.011426978818283166,
+      "grad_norm": 0.39463481307029724,
+      "learning_rate": 8e-05,
+      "loss": 1.8829,
+      "step": 205
+    },
+    {
+      "epoch": 0.011482720178372352,
+      "grad_norm": 0.41305360198020935,
+      "learning_rate": 8e-05,
+      "loss": 1.7638,
+      "step": 206
+    },
+    {
+      "epoch": 0.011538461538461539,
+      "grad_norm": 0.4194389879703522,
+      "learning_rate": 8e-05,
+      "loss": 1.8702,
+      "step": 207
+    },
+    {
+      "epoch": 0.011594202898550725,
+      "grad_norm": 0.4725489020347595,
+      "learning_rate": 8e-05,
+      "loss": 2.0307,
+      "step": 208
+    },
+    {
+      "epoch": 0.01164994425863991,
+      "grad_norm": 0.45651867985725403,
+      "learning_rate": 8e-05,
+      "loss": 1.9788,
+      "step": 209
+    },
+    {
+      "epoch": 0.011705685618729096,
+      "grad_norm": 0.41888317465782166,
+      "learning_rate": 8e-05,
+      "loss": 1.9555,
+      "step": 210
+    },
+    {
+      "epoch": 0.011761426978818284,
+      "grad_norm": 0.4058757722377777,
+      "learning_rate": 8e-05,
+      "loss": 1.8253,
+      "step": 211
+    },
+    {
+      "epoch": 0.01181716833890747,
+      "grad_norm": 0.3691366910934448,
+      "learning_rate": 8e-05,
+      "loss": 1.596,
+      "step": 212
+    },
+    {
+      "epoch": 0.011872909698996655,
+      "grad_norm": 0.37606245279312134,
+      "learning_rate": 8e-05,
+      "loss": 1.587,
+      "step": 213
+    },
+    {
+      "epoch": 0.011928651059085842,
+      "grad_norm": 0.42671895027160645,
+      "learning_rate": 8e-05,
+      "loss": 1.7991,
+      "step": 214
+    },
+    {
+      "epoch": 0.011984392419175028,
+      "grad_norm": 0.40445569157600403,
+      "learning_rate": 8e-05,
+      "loss": 1.7706,
+      "step": 215
+    },
+    {
+      "epoch": 0.012040133779264214,
+      "grad_norm": 0.4408814013004303,
+      "learning_rate": 8e-05,
+      "loss": 1.8041,
+      "step": 216
+    },
+    {
+      "epoch": 0.012095875139353401,
+      "grad_norm": 0.43463078141212463,
+      "learning_rate": 8e-05,
+      "loss": 1.9364,
+      "step": 217
+    },
+    {
+      "epoch": 0.012151616499442587,
+      "grad_norm": 0.4390909671783447,
+      "learning_rate": 8e-05,
+      "loss": 1.8111,
+      "step": 218
+    },
+    {
+      "epoch": 0.012207357859531772,
+      "grad_norm": 0.4165789484977722,
+      "learning_rate": 8e-05,
+      "loss": 1.8843,
+      "step": 219
+    },
+    {
+      "epoch": 0.012263099219620958,
+      "grad_norm": 0.4711140990257263,
+      "learning_rate": 8e-05,
+      "loss": 2.1673,
+      "step": 220
+    },
+    {
+      "epoch": 0.012318840579710146,
+      "grad_norm": 0.4233494997024536,
+      "learning_rate": 8e-05,
+      "loss": 1.8285,
+      "step": 221
+    },
+    {
+      "epoch": 0.012374581939799331,
+      "grad_norm": 0.41040298342704773,
+      "learning_rate": 8e-05,
+      "loss": 1.7355,
+      "step": 222
+    },
+    {
+      "epoch": 0.012430323299888517,
+      "grad_norm": 0.4037279784679413,
+      "learning_rate": 8e-05,
+      "loss": 1.8224,
+      "step": 223
+    },
+    {
+      "epoch": 0.012486064659977704,
+      "grad_norm": 0.45015132427215576,
+      "learning_rate": 8e-05,
+      "loss": 2.0457,
+      "step": 224
+    },
+    {
+      "epoch": 0.01254180602006689,
+      "grad_norm": 0.4033258855342865,
+      "learning_rate": 8e-05,
+      "loss": 1.8827,
+      "step": 225
+    },
+    {
+      "epoch": 0.012597547380156076,
+      "grad_norm": 0.38210955262184143,
+      "learning_rate": 8e-05,
+      "loss": 1.7697,
+      "step": 226
+    },
+    {
+      "epoch": 0.012653288740245261,
+      "grad_norm": 0.39336487650871277,
+      "learning_rate": 8e-05,
+      "loss": 1.7893,
+      "step": 227
+    },
+    {
+      "epoch": 0.012709030100334449,
+      "grad_norm": 0.38320380449295044,
+      "learning_rate": 8e-05,
+      "loss": 1.7757,
+      "step": 228
+    },
+    {
+      "epoch": 0.012764771460423634,
+      "grad_norm": 0.4049006700515747,
+      "learning_rate": 8e-05,
+      "loss": 1.8953,
+      "step": 229
+    },
+    {
+      "epoch": 0.01282051282051282,
+      "grad_norm": 0.412638783454895,
+      "learning_rate": 8e-05,
+      "loss": 1.7646,
+      "step": 230
+    },
+    {
+      "epoch": 0.012876254180602007,
+      "grad_norm": 0.3987637460231781,
+      "learning_rate": 8e-05,
+      "loss": 1.7276,
+      "step": 231
+    },
+    {
+      "epoch": 0.012931995540691193,
+      "grad_norm": 0.40791159868240356,
+      "learning_rate": 8e-05,
+      "loss": 1.6983,
+      "step": 232
+    },
+    {
+      "epoch": 0.012987736900780379,
+      "grad_norm": 0.41750428080558777,
+      "learning_rate": 8e-05,
+      "loss": 2.0136,
+      "step": 233
+    },
+    {
+      "epoch": 0.013043478260869565,
+      "grad_norm": 0.4128837585449219,
+      "learning_rate": 8e-05,
+      "loss": 1.8859,
+      "step": 234
+    },
+    {
+      "epoch": 0.013099219620958752,
+      "grad_norm": 0.40263018012046814,
+      "learning_rate": 8e-05,
+      "loss": 1.8077,
+      "step": 235
+    },
+    {
+      "epoch": 0.013154960981047938,
+      "grad_norm": 0.40996524691581726,
+      "learning_rate": 8e-05,
+      "loss": 1.6886,
+      "step": 236
+    },
+    {
+      "epoch": 0.013210702341137123,
+      "grad_norm": 0.47232288122177124,
+      "learning_rate": 8e-05,
+      "loss": 1.9667,
+      "step": 237
+    },
+    {
+      "epoch": 0.01326644370122631,
+      "grad_norm": 0.4224655032157898,
+      "learning_rate": 8e-05,
+      "loss": 1.7665,
+      "step": 238
+    },
+    {
+      "epoch": 0.013322185061315496,
+      "grad_norm": 0.3976142406463623,
+      "learning_rate": 8e-05,
+      "loss": 1.7982,
+      "step": 239
+    },
+    {
+      "epoch": 0.013377926421404682,
+      "grad_norm": 0.4137797951698303,
+      "learning_rate": 8e-05,
+      "loss": 1.8543,
+      "step": 240
+    },
+    {
+      "epoch": 0.013433667781493868,
+      "grad_norm": 0.4323852062225342,
+      "learning_rate": 8e-05,
+      "loss": 1.9242,
+      "step": 241
+    },
+    {
+      "epoch": 0.013489409141583055,
+      "grad_norm": 0.4583011269569397,
+      "learning_rate": 8e-05,
+      "loss": 2.0125,
+      "step": 242
+    },
+    {
+      "epoch": 0.01354515050167224,
+      "grad_norm": 0.4027802646160126,
+      "learning_rate": 8e-05,
+      "loss": 1.8284,
+      "step": 243
+    },
+    {
+      "epoch": 0.013600891861761426,
+      "grad_norm": 0.45875123143196106,
+      "learning_rate": 8e-05,
+      "loss": 2.0507,
+      "step": 244
+    },
+    {
+      "epoch": 0.013656633221850614,
+      "grad_norm": 0.4033229351043701,
+      "learning_rate": 8e-05,
+      "loss": 1.9382,
+      "step": 245
+    },
+    {
+      "epoch": 0.0137123745819398,
+      "grad_norm": 0.3958418667316437,
+      "learning_rate": 8e-05,
+      "loss": 1.8293,
+      "step": 246
+    },
+    {
+      "epoch": 0.013768115942028985,
+      "grad_norm": 0.4176151752471924,
+      "learning_rate": 8e-05,
+      "loss": 1.8388,
+      "step": 247
+    },
+    {
+      "epoch": 0.013823857302118171,
+      "grad_norm": 0.3736697733402252,
+      "learning_rate": 8e-05,
+      "loss": 1.681,
+      "step": 248
+    },
+    {
+      "epoch": 0.013879598662207358,
+      "grad_norm": 0.40974247455596924,
+      "learning_rate": 8e-05,
+      "loss": 1.7743,
+      "step": 249
+    },
+    {
+      "epoch": 0.013935340022296544,
+      "grad_norm": 0.41332730650901794,
+      "learning_rate": 8e-05,
+      "loss": 1.6346,
+      "step": 250
+    },
+    {
+      "epoch": 0.01399108138238573,
+      "grad_norm": 0.42761102318763733,
+      "learning_rate": 8e-05,
+      "loss": 1.8236,
+      "step": 251
+    },
+    {
+      "epoch": 0.014046822742474917,
+      "grad_norm": 0.4135681986808777,
+      "learning_rate": 8e-05,
+      "loss": 1.8745,
+      "step": 252
+    },
+    {
+      "epoch": 0.014102564102564103,
+      "grad_norm": 0.38827580213546753,
+      "learning_rate": 8e-05,
+      "loss": 1.633,
+      "step": 253
+    },
+    {
+      "epoch": 0.014158305462653288,
+      "grad_norm": 0.42247524857521057,
+      "learning_rate": 8e-05,
+      "loss": 1.8305,
+      "step": 254
+    },
+    {
+      "epoch": 0.014214046822742474,
+      "grad_norm": 0.3817632794380188,
+      "learning_rate": 8e-05,
+      "loss": 1.6589,
+      "step": 255
+    },
+    {
+      "epoch": 0.014269788182831662,
+      "grad_norm": 0.41619542241096497,
+      "learning_rate": 8e-05,
+      "loss": 1.6908,
+      "step": 256
+    },
+    {
+      "epoch": 0.014325529542920847,
+      "grad_norm": 0.40807268023490906,
+      "learning_rate": 8e-05,
+      "loss": 1.8188,
+      "step": 257
+    },
+    {
+      "epoch": 0.014381270903010033,
+      "grad_norm": 0.3903752565383911,
+      "learning_rate": 8e-05,
+      "loss": 1.5068,
+      "step": 258
+    },
+    {
+      "epoch": 0.01443701226309922,
+      "grad_norm": 0.43174487352371216,
+      "learning_rate": 8e-05,
+      "loss": 2.0029,
+      "step": 259
+    },
+    {
+      "epoch": 0.014492753623188406,
+      "grad_norm": 0.40700212121009827,
+      "learning_rate": 8e-05,
+      "loss": 1.7942,
+      "step": 260
+    },
+    {
+      "epoch": 0.014548494983277592,
+      "grad_norm": 0.4008272588253021,
+      "learning_rate": 8e-05,
+      "loss": 1.8252,
+      "step": 261
+    },
+    {
+      "epoch": 0.014604236343366777,
+      "grad_norm": 0.441720187664032,
+      "learning_rate": 8e-05,
+      "loss": 1.7293,
+      "step": 262
+    },
+    {
+      "epoch": 0.014659977703455965,
+      "grad_norm": 0.37992534041404724,
+      "learning_rate": 8e-05,
+      "loss": 1.7273,
+      "step": 263
+    },
+    {
+      "epoch": 0.01471571906354515,
+      "grad_norm": 0.38580581545829773,
+      "learning_rate": 8e-05,
+      "loss": 1.7454,
+      "step": 264
+    },
+    {
+      "epoch": 0.014771460423634336,
+      "grad_norm": 0.4266432225704193,
+      "learning_rate": 8e-05,
+      "loss": 1.9173,
+      "step": 265
+    },
+    {
+      "epoch": 0.014827201783723524,
+      "grad_norm": 0.4495110809803009,
+      "learning_rate": 8e-05,
+      "loss": 1.9769,
+      "step": 266
+    },
+    {
+      "epoch": 0.01488294314381271,
+      "grad_norm": 0.4119555354118347,
+      "learning_rate": 8e-05,
+      "loss": 1.7148,
+      "step": 267
+    },
+    {
+      "epoch": 0.014938684503901895,
+      "grad_norm": 0.41930001974105835,
+      "learning_rate": 8e-05,
+      "loss": 1.9008,
+      "step": 268
+    },
+    {
+      "epoch": 0.01499442586399108,
+      "grad_norm": 0.41471660137176514,
+      "learning_rate": 8e-05,
+      "loss": 1.825,
+      "step": 269
+    },
+    {
+      "epoch": 0.015050167224080268,
+      "grad_norm": 0.4505252540111542,
+      "learning_rate": 8e-05,
+      "loss": 1.9394,
+      "step": 270
+    },
+    {
+      "epoch": 0.015105908584169454,
+      "grad_norm": 0.45706626772880554,
+      "learning_rate": 8e-05,
+      "loss": 2.0057,
+      "step": 271
+    },
+    {
+      "epoch": 0.01516164994425864,
+      "grad_norm": 0.41290906071662903,
+      "learning_rate": 8e-05,
+      "loss": 2.0293,
+      "step": 272
+    },
+    {
+      "epoch": 0.015217391304347827,
+      "grad_norm": 0.41640666127204895,
+      "learning_rate": 8e-05,
+      "loss": 1.7892,
+      "step": 273
+    },
+    {
+      "epoch": 0.015273132664437012,
+      "grad_norm": 0.4064461588859558,
+      "learning_rate": 8e-05,
+      "loss": 1.8026,
+      "step": 274
+    },
+    {
+      "epoch": 0.015328874024526198,
+      "grad_norm": 0.46918439865112305,
+      "learning_rate": 8e-05,
+      "loss": 1.8843,
+      "step": 275
+    },
+    {
+      "epoch": 0.015384615384615385,
+      "grad_norm": 0.39473360776901245,
+      "learning_rate": 8e-05,
+      "loss": 1.77,
+      "step": 276
+    },
+    {
+      "epoch": 0.015440356744704571,
+      "grad_norm": 0.36619892716407776,
+      "learning_rate": 8e-05,
+      "loss": 1.5864,
+      "step": 277
+    },
+    {
+      "epoch": 0.015496098104793757,
+      "grad_norm": 0.4101516306400299,
+      "learning_rate": 8e-05,
+      "loss": 1.7618,
+      "step": 278
+    },
+    {
+      "epoch": 0.015551839464882942,
+      "grad_norm": 0.41335687041282654,
+      "learning_rate": 8e-05,
+      "loss": 1.6014,
+      "step": 279
+    },
+    {
+      "epoch": 0.01560758082497213,
+      "grad_norm": 0.44227463006973267,
+      "learning_rate": 8e-05,
+      "loss": 2.0184,
+      "step": 280
+    },
+    {
+      "epoch": 0.015663322185061314,
+      "grad_norm": 0.40748894214630127,
+      "learning_rate": 8e-05,
+      "loss": 1.7469,
+      "step": 281
+    },
+    {
+      "epoch": 0.015719063545150503,
+      "grad_norm": 0.41847971081733704,
+      "learning_rate": 8e-05,
+      "loss": 1.7594,
+      "step": 282
+    },
+    {
+      "epoch": 0.01577480490523969,
+      "grad_norm": 0.41384992003440857,
+      "learning_rate": 8e-05,
+      "loss": 1.4822,
+      "step": 283
+    },
+    {
+      "epoch": 0.015830546265328874,
+      "grad_norm": 0.40941479802131653,
+      "learning_rate": 8e-05,
+      "loss": 1.8697,
+      "step": 284
+    },
+    {
+      "epoch": 0.01588628762541806,
+      "grad_norm": 0.42511463165283203,
+      "learning_rate": 8e-05,
+      "loss": 1.8659,
+      "step": 285
+    },
+    {
+      "epoch": 0.015942028985507246,
+      "grad_norm": 0.41972148418426514,
+      "learning_rate": 8e-05,
+      "loss": 1.7126,
+      "step": 286
+    },
+    {
+      "epoch": 0.01599777034559643,
+      "grad_norm": 0.42691537737846375,
+      "learning_rate": 8e-05,
+      "loss": 1.8377,
+      "step": 287
+    },
+    {
+      "epoch": 0.016053511705685617,
+      "grad_norm": 0.4027348756790161,
+      "learning_rate": 8e-05,
+      "loss": 1.6883,
+      "step": 288
+    },
+    {
+      "epoch": 0.016109253065774806,
+      "grad_norm": 0.41511085629463196,
+      "learning_rate": 8e-05,
+      "loss": 1.9085,
+      "step": 289
+    },
+    {
+      "epoch": 0.016164994425863992,
+      "grad_norm": 0.4326029419898987,
+      "learning_rate": 8e-05,
+      "loss": 1.8441,
+      "step": 290
+    },
+    {
+      "epoch": 0.016220735785953178,
+      "grad_norm": 0.40246617794036865,
+      "learning_rate": 8e-05,
+      "loss": 1.7686,
+      "step": 291
+    },
+    {
+      "epoch": 0.016276477146042363,
+      "grad_norm": 0.38311776518821716,
+      "learning_rate": 8e-05,
+      "loss": 1.7145,
+      "step": 292
+    },
+    {
+      "epoch": 0.01633221850613155,
+      "grad_norm": 0.3850622773170471,
+      "learning_rate": 8e-05,
+      "loss": 1.7244,
+      "step": 293
+    },
+    {
+      "epoch": 0.016387959866220735,
+      "grad_norm": 0.4168466627597809,
+      "learning_rate": 8e-05,
+      "loss": 1.8325,
+      "step": 294
+    },
+    {
+      "epoch": 0.01644370122630992,
+      "grad_norm": 0.3952409029006958,
+      "learning_rate": 8e-05,
+      "loss": 1.9131,
+      "step": 295
+    },
+    {
+      "epoch": 0.01649944258639911,
+      "grad_norm": 0.42862653732299805,
+      "learning_rate": 8e-05,
+      "loss": 1.8013,
+      "step": 296
+    },
+    {
+      "epoch": 0.016555183946488295,
+      "grad_norm": 0.42462411522865295,
+      "learning_rate": 8e-05,
+      "loss": 1.7125,
+      "step": 297
+    },
+    {
+      "epoch": 0.01661092530657748,
+      "grad_norm": 0.3978937268257141,
+      "learning_rate": 8e-05,
+      "loss": 1.7819,
+      "step": 298
+    },
+    {
+      "epoch": 0.016666666666666666,
+      "grad_norm": 0.3845030963420868,
+      "learning_rate": 8e-05,
+      "loss": 1.6577,
+      "step": 299
+    },
+    {
+      "epoch": 0.016722408026755852,
+      "grad_norm": 0.4124613404273987,
+      "learning_rate": 8e-05,
+      "loss": 1.7366,
+      "step": 300
+    },
+    {
+      "epoch": 0.016778149386845038,
+      "grad_norm": 0.45144006609916687,
+      "learning_rate": 8e-05,
+      "loss": 1.8667,
+      "step": 301
+    },
+    {
+      "epoch": 0.016833890746934223,
+      "grad_norm": 0.37532711029052734,
+      "learning_rate": 8e-05,
+      "loss": 1.5467,
+      "step": 302
+    },
+    {
+      "epoch": 0.016889632107023413,
+      "grad_norm": 0.41644760966300964,
+      "learning_rate": 8e-05,
+      "loss": 1.8618,
+      "step": 303
+    },
+    {
+      "epoch": 0.0169453734671126,
+      "grad_norm": 0.4005732834339142,
+      "learning_rate": 8e-05,
+      "loss": 1.7652,
+      "step": 304
+    },
+    {
+      "epoch": 0.017001114827201784,
+      "grad_norm": 0.384604275226593,
+      "learning_rate": 8e-05,
+      "loss": 1.7026,
+      "step": 305
+    },
+    {
+      "epoch": 0.01705685618729097,
+      "grad_norm": 0.3821338415145874,
+      "learning_rate": 8e-05,
+      "loss": 1.6649,
+      "step": 306
+    },
+    {
+      "epoch": 0.017112597547380155,
+      "grad_norm": 0.415862500667572,
+      "learning_rate": 8e-05,
+      "loss": 1.9326,
+      "step": 307
+    },
+    {
+      "epoch": 0.01716833890746934,
+      "grad_norm": 0.4307110905647278,
+      "learning_rate": 8e-05,
+      "loss": 1.7675,
+      "step": 308
+    },
+    {
+      "epoch": 0.01722408026755853,
+      "grad_norm": 0.49622923135757446,
+      "learning_rate": 8e-05,
+      "loss": 1.8899,
+      "step": 309
+    },
+    {
+      "epoch": 0.017279821627647716,
+      "grad_norm": 0.4144655466079712,
+      "learning_rate": 8e-05,
+      "loss": 1.9463,
+      "step": 310
+    },
+    {
+      "epoch": 0.0173355629877369,
+      "grad_norm": 0.42982232570648193,
+      "learning_rate": 8e-05,
+      "loss": 1.7616,
+      "step": 311
+    },
+    {
+      "epoch": 0.017391304347826087,
+      "grad_norm": 0.419792115688324,
+      "learning_rate": 8e-05,
+      "loss": 1.7261,
+      "step": 312
+    },
+    {
+      "epoch": 0.017447045707915273,
+      "grad_norm": 0.4065159261226654,
+      "learning_rate": 8e-05,
+      "loss": 1.4601,
+      "step": 313
+    },
+    {
+      "epoch": 0.01750278706800446,
+      "grad_norm": 0.4134047329425812,
+      "learning_rate": 8e-05,
+      "loss": 1.8337,
+      "step": 314
+    },
+    {
+      "epoch": 0.017558528428093644,
+      "grad_norm": 0.4063039720058441,
+      "learning_rate": 8e-05,
+      "loss": 1.8483,
+      "step": 315
+    },
+    {
+      "epoch": 0.017614269788182833,
+      "grad_norm": 0.41646286845207214,
+      "learning_rate": 8e-05,
+      "loss": 1.6276,
+      "step": 316
+    },
+    {
+      "epoch": 0.01767001114827202,
+      "grad_norm": 0.4397350549697876,
+      "learning_rate": 8e-05,
+      "loss": 1.9969,
+      "step": 317
+    },
+    {
+      "epoch": 0.017725752508361205,
+      "grad_norm": 0.40330687165260315,
+      "learning_rate": 8e-05,
+      "loss": 1.6762,
+      "step": 318
+    },
+    {
+      "epoch": 0.01778149386845039,
+      "grad_norm": 0.38782450556755066,
+      "learning_rate": 8e-05,
+      "loss": 1.8607,
+      "step": 319
+    },
+    {
+      "epoch": 0.017837235228539576,
+      "grad_norm": 0.41678085923194885,
+      "learning_rate": 8e-05,
+      "loss": 1.8923,
+      "step": 320
+    },
+    {
+      "epoch": 0.01789297658862876,
+      "grad_norm": 0.43278276920318604,
+      "learning_rate": 8e-05,
+      "loss": 1.7742,
+      "step": 321
+    },
+    {
+      "epoch": 0.017948717948717947,
+      "grad_norm": 0.43438470363616943,
+      "learning_rate": 8e-05,
+      "loss": 2.0248,
+      "step": 322
+    },
+    {
+      "epoch": 0.018004459308807137,
+      "grad_norm": 0.41625747084617615,
+      "learning_rate": 8e-05,
+      "loss": 1.8575,
+      "step": 323
+    },
+    {
+      "epoch": 0.018060200668896322,
+      "grad_norm": 0.3881988227367401,
+      "learning_rate": 8e-05,
+      "loss": 1.815,
+      "step": 324
+    },
+    {
+      "epoch": 0.018115942028985508,
+      "grad_norm": 0.462026447057724,
+      "learning_rate": 8e-05,
+      "loss": 2.1086,
+      "step": 325
+    },
+    {
+      "epoch": 0.018171683389074694,
+      "grad_norm": 0.39492854475975037,
+      "learning_rate": 8e-05,
+      "loss": 1.7288,
+      "step": 326
+    },
+    {
+      "epoch": 0.01822742474916388,
+      "grad_norm": 0.4187218248844147,
+      "learning_rate": 8e-05,
+      "loss": 1.6914,
+      "step": 327
+    },
+    {
+      "epoch": 0.018283166109253065,
+      "grad_norm": 0.37417539954185486,
+      "learning_rate": 8e-05,
+      "loss": 1.6756,
+      "step": 328
+    },
+    {
+      "epoch": 0.01833890746934225,
+      "grad_norm": 0.379098504781723,
+      "learning_rate": 8e-05,
+      "loss": 1.5272,
+      "step": 329
+    },
+    {
+      "epoch": 0.01839464882943144,
+      "grad_norm": 0.4593086838722229,
+      "learning_rate": 8e-05,
+      "loss": 1.9489,
+      "step": 330
+    },
+    {
+      "epoch": 0.018450390189520625,
+      "grad_norm": 0.4017282724380493,
+      "learning_rate": 8e-05,
+      "loss": 1.6141,
+      "step": 331
+    },
+    {
+      "epoch": 0.01850613154960981,
+      "grad_norm": 0.41741320490837097,
+      "learning_rate": 8e-05,
+      "loss": 1.9041,
+      "step": 332
+    },
+    {
+      "epoch": 0.018561872909698997,
+      "grad_norm": 0.7527666687965393,
+      "learning_rate": 8e-05,
+      "loss": 1.9661,
+      "step": 333
+    },
+    {
+      "epoch": 0.018617614269788182,
+      "grad_norm": 0.3948970437049866,
+      "learning_rate": 8e-05,
+      "loss": 1.7473,
+      "step": 334
+    },
+    {
+      "epoch": 0.018673355629877368,
+      "grad_norm": 0.46345531940460205,
+      "learning_rate": 8e-05,
+      "loss": 2.0197,
+      "step": 335
+    },
+    {
+      "epoch": 0.018729096989966554,
+      "grad_norm": 0.41160738468170166,
+      "learning_rate": 8e-05,
+      "loss": 1.8306,
+      "step": 336
+    },
+    {
+      "epoch": 0.018784838350055743,
+      "grad_norm": 0.4199187457561493,
+      "learning_rate": 8e-05,
+      "loss": 1.7609,
+      "step": 337
+    },
+    {
+      "epoch": 0.01884057971014493,
+      "grad_norm": 0.43813398480415344,
+      "learning_rate": 8e-05,
+      "loss": 1.9953,
+      "step": 338
+    },
+    {
+      "epoch": 0.018896321070234114,
+      "grad_norm": 0.3779383897781372,
+      "learning_rate": 8e-05,
+      "loss": 1.5923,
+      "step": 339
+    },
+    {
+      "epoch": 0.0189520624303233,
+      "grad_norm": 0.4910745918750763,
+      "learning_rate": 8e-05,
+      "loss": 2.1573,
+      "step": 340
+    },
+    {
+      "epoch": 0.019007803790412486,
+      "grad_norm": 0.44113677740097046,
+      "learning_rate": 8e-05,
+      "loss": 1.9559,
+      "step": 341
+    },
+    {
+      "epoch": 0.01906354515050167,
+      "grad_norm": 0.41205549240112305,
+      "learning_rate": 8e-05,
+      "loss": 1.6418,
+      "step": 342
+    },
+    {
+      "epoch": 0.019119286510590857,
+      "grad_norm": 0.41746771335601807,
+      "learning_rate": 8e-05,
+      "loss": 1.8842,
+      "step": 343
+    },
+    {
+      "epoch": 0.019175027870680046,
+      "grad_norm": 0.42498695850372314,
+      "learning_rate": 8e-05,
+      "loss": 1.8814,
+      "step": 344
+    },
+    {
+      "epoch": 0.019230769230769232,
+      "grad_norm": 0.4002072811126709,
+      "learning_rate": 8e-05,
+      "loss": 1.7794,
+      "step": 345
+    },
+    {
+      "epoch": 0.019286510590858418,
+      "grad_norm": 0.4046269953250885,
+      "learning_rate": 8e-05,
+      "loss": 1.9115,
+      "step": 346
+    },
+    {
+      "epoch": 0.019342251950947603,
+      "grad_norm": 0.38134604692459106,
+      "learning_rate": 8e-05,
+      "loss": 1.6016,
+      "step": 347
+    },
+    {
+      "epoch": 0.01939799331103679,
+      "grad_norm": 0.44585302472114563,
+      "learning_rate": 8e-05,
+      "loss": 1.7686,
+      "step": 348
+    },
+    {
+      "epoch": 0.019453734671125975,
+      "grad_norm": 0.42694950103759766,
+      "learning_rate": 8e-05,
+      "loss": 1.746,
+      "step": 349
+    },
+    {
+      "epoch": 0.01950947603121516,
+      "grad_norm": 0.3884536623954773,
+      "learning_rate": 8e-05,
+      "loss": 1.6251,
+      "step": 350
+    },
+    {
+      "epoch": 0.01956521739130435,
+      "grad_norm": 0.4125804901123047,
+      "learning_rate": 8e-05,
+      "loss": 1.6578,
+      "step": 351
+    },
+    {
+      "epoch": 0.019620958751393535,
+      "grad_norm": 0.3952336609363556,
+      "learning_rate": 8e-05,
+      "loss": 1.475,
+      "step": 352
+    },
+    {
+      "epoch": 0.01967670011148272,
+      "grad_norm": 0.40272772312164307,
+      "learning_rate": 8e-05,
+      "loss": 1.7858,
+      "step": 353
+    },
+    {
+      "epoch": 0.019732441471571906,
+      "grad_norm": 0.398934930562973,
+      "learning_rate": 8e-05,
+      "loss": 1.6765,
+      "step": 354
+    },
+    {
+      "epoch": 0.019788182831661092,
+      "grad_norm": 0.4227500557899475,
+      "learning_rate": 8e-05,
+      "loss": 1.6917,
+      "step": 355
+    },
+    {
+      "epoch": 0.019843924191750278,
+      "grad_norm": 0.44395002722740173,
+      "learning_rate": 8e-05,
+      "loss": 2.0649,
+      "step": 356
+    },
+    {
+      "epoch": 0.019899665551839463,
+      "grad_norm": 0.4013097286224365,
+      "learning_rate": 8e-05,
+      "loss": 1.7733,
+      "step": 357
+    },
+    {
+      "epoch": 0.019955406911928653,
+      "grad_norm": 0.3915567100048065,
+      "learning_rate": 8e-05,
+      "loss": 1.8218,
+      "step": 358
+    },
+    {
+      "epoch": 0.020011148272017838,
+      "grad_norm": 0.4370361268520355,
+      "learning_rate": 8e-05,
+      "loss": 1.8071,
+      "step": 359
+    },
+    {
+      "epoch": 0.020066889632107024,
+      "grad_norm": 0.4509705901145935,
+      "learning_rate": 8e-05,
+      "loss": 1.7563,
+      "step": 360
+    },
+    {
+      "epoch": 0.02012263099219621,
+      "grad_norm": 0.3994988799095154,
+      "learning_rate": 8e-05,
+      "loss": 1.7667,
+      "step": 361
+    },
+    {
+      "epoch": 0.020178372352285395,
+      "grad_norm": 0.41288429498672485,
+      "learning_rate": 8e-05,
+      "loss": 1.8398,
+      "step": 362
+    },
+    {
+      "epoch": 0.02023411371237458,
+      "grad_norm": 0.4281528890132904,
+      "learning_rate": 8e-05,
+      "loss": 1.8315,
+      "step": 363
+    },
+    {
+      "epoch": 0.020289855072463767,
+      "grad_norm": 0.5056607127189636,
+      "learning_rate": 8e-05,
+      "loss": 1.9024,
+      "step": 364
+    },
+    {
+      "epoch": 0.020345596432552956,
+      "grad_norm": 0.379406601190567,
+      "learning_rate": 8e-05,
+      "loss": 1.6976,
+      "step": 365
+    },
+    {
+      "epoch": 0.02040133779264214,
+      "grad_norm": 0.3984786868095398,
+      "learning_rate": 8e-05,
+      "loss": 1.9396,
+      "step": 366
+    },
+    {
+      "epoch": 0.020457079152731327,
+      "grad_norm": 0.4233458936214447,
+      "learning_rate": 8e-05,
+      "loss": 1.8615,
+      "step": 367
+    },
+    {
+      "epoch": 0.020512820512820513,
+      "grad_norm": 0.4357461929321289,
+      "learning_rate": 8e-05,
+      "loss": 1.8852,
+      "step": 368
+    },
+    {
+      "epoch": 0.0205685618729097,
+      "grad_norm": 0.3997776210308075,
+      "learning_rate": 8e-05,
+      "loss": 1.6763,
+      "step": 369
+    },
+    {
+      "epoch": 0.020624303232998884,
+      "grad_norm": 0.4437631070613861,
+      "learning_rate": 8e-05,
+      "loss": 1.9861,
+      "step": 370
+    },
+    {
+      "epoch": 0.02068004459308807,
+      "grad_norm": 0.41594505310058594,
+      "learning_rate": 8e-05,
+      "loss": 1.7889,
+      "step": 371
+    },
+    {
+      "epoch": 0.02073578595317726,
+      "grad_norm": 0.4146479368209839,
+      "learning_rate": 8e-05,
+      "loss": 1.8944,
+      "step": 372
+    },
+    {
+      "epoch": 0.020791527313266445,
+      "grad_norm": 0.43242311477661133,
+      "learning_rate": 8e-05,
+      "loss": 1.8777,
+      "step": 373
+    },
+    {
+      "epoch": 0.02084726867335563,
+      "grad_norm": 0.427500456571579,
+      "learning_rate": 8e-05,
+      "loss": 1.7193,
+      "step": 374
+    },
+    {
+      "epoch": 0.020903010033444816,
+      "grad_norm": 0.4136715531349182,
+      "learning_rate": 8e-05,
+      "loss": 1.8498,
+      "step": 375
+    },
+    {
+      "epoch": 0.020958751393534,
+      "grad_norm": 0.39340317249298096,
+      "learning_rate": 8e-05,
+      "loss": 1.7887,
+      "step": 376
+    },
+    {
+      "epoch": 0.021014492753623187,
+      "grad_norm": 0.46348172426223755,
+      "learning_rate": 8e-05,
+      "loss": 1.9388,
+      "step": 377
+    },
+    {
+      "epoch": 0.021070234113712373,
+      "grad_norm": 0.3817267119884491,
+      "learning_rate": 8e-05,
+      "loss": 1.5686,
+      "step": 378
+    },
+    {
+      "epoch": 0.021125975473801562,
+      "grad_norm": 0.4563877284526825,
+      "learning_rate": 8e-05,
+      "loss": 1.8118,
+      "step": 379
+    },
+    {
+      "epoch": 0.021181716833890748,
+      "grad_norm": 0.41754311323165894,
+      "learning_rate": 8e-05,
+      "loss": 1.7371,
+      "step": 380
+    },
+    {
+      "epoch": 0.021237458193979934,
+      "grad_norm": 0.3904457688331604,
+      "learning_rate": 8e-05,
+      "loss": 1.5807,
+      "step": 381
+    },
+    {
+      "epoch": 0.02129319955406912,
+      "grad_norm": 0.4270111322402954,
+      "learning_rate": 8e-05,
+      "loss": 1.7304,
+      "step": 382
+    },
+    {
+      "epoch": 0.021348940914158305,
+      "grad_norm": 0.425065279006958,
+      "learning_rate": 8e-05,
+      "loss": 1.8769,
+      "step": 383
+    },
+    {
+      "epoch": 0.02140468227424749,
+      "grad_norm": 0.4404243528842926,
+      "learning_rate": 8e-05,
+      "loss": 1.7372,
+      "step": 384
+    },
+    {
+      "epoch": 0.021460423634336676,
+      "grad_norm": 0.4347771108150482,
+      "learning_rate": 8e-05,
+      "loss": 1.9571,
+      "step": 385
+    },
+    {
+      "epoch": 0.021516164994425865,
+      "grad_norm": 0.4173373579978943,
+      "learning_rate": 8e-05,
+      "loss": 1.7546,
+      "step": 386
+    },
+    {
+      "epoch": 0.02157190635451505,
+      "grad_norm": 0.4261404275894165,
+      "learning_rate": 8e-05,
+      "loss": 1.9542,
+      "step": 387
+    },
+    {
+      "epoch": 0.021627647714604237,
+      "grad_norm": 0.4324272871017456,
+      "learning_rate": 8e-05,
+      "loss": 1.7786,
+      "step": 388
+    },
+    {
+      "epoch": 0.021683389074693422,
+      "grad_norm": 0.39340800046920776,
+      "learning_rate": 8e-05,
+      "loss": 1.6615,
+      "step": 389
+    },
+    {
+      "epoch": 0.021739130434782608,
+      "grad_norm": 0.41590434312820435,
+      "learning_rate": 8e-05,
+      "loss": 1.8427,
+      "step": 390
+    },
+    {
+      "epoch": 0.021794871794871794,
+      "grad_norm": 0.41448861360549927,
+      "learning_rate": 8e-05,
+      "loss": 1.7105,
+      "step": 391
+    },
+    {
+      "epoch": 0.02185061315496098,
+      "grad_norm": 0.42208924889564514,
+      "learning_rate": 8e-05,
+      "loss": 2.0502,
+      "step": 392
+    },
+    {
+      "epoch": 0.02190635451505017,
+      "grad_norm": 0.37751758098602295,
+      "learning_rate": 8e-05,
+      "loss": 1.65,
+      "step": 393
+    },
+    {
+      "epoch": 0.021962095875139354,
+      "grad_norm": 0.4493613839149475,
+      "learning_rate": 8e-05,
+      "loss": 2.0241,
+      "step": 394
+    },
+    {
+      "epoch": 0.02201783723522854,
+      "grad_norm": 0.3963421583175659,
+      "learning_rate": 8e-05,
+      "loss": 1.6982,
+      "step": 395
+    },
+    {
+      "epoch": 0.022073578595317726,
+      "grad_norm": 0.41097620129585266,
+      "learning_rate": 8e-05,
+      "loss": 1.894,
+      "step": 396
+    },
+    {
+      "epoch": 0.02212931995540691,
+      "grad_norm": 0.4165438711643219,
+      "learning_rate": 8e-05,
+      "loss": 1.9027,
+      "step": 397
+    },
+    {
+      "epoch": 0.022185061315496097,
+      "grad_norm": 0.42101776599884033,
+      "learning_rate": 8e-05,
+      "loss": 1.9453,
+      "step": 398
+    },
+    {
+      "epoch": 0.022240802675585283,
+      "grad_norm": 0.47843456268310547,
+      "learning_rate": 8e-05,
+      "loss": 2.0013,
+      "step": 399
+    },
+    {
+      "epoch": 0.022296544035674472,
+      "grad_norm": 0.44337159395217896,
+      "learning_rate": 8e-05,
+      "loss": 1.9636,
+      "step": 400
+    },
+    {
+      "epoch": 0.022352285395763657,
+      "grad_norm": 0.39196833968162537,
+      "learning_rate": 8e-05,
+      "loss": 1.7866,
+      "step": 401
+    },
+    {
+      "epoch": 0.022408026755852843,
+      "grad_norm": 0.45299163460731506,
+      "learning_rate": 8e-05,
+      "loss": 1.8547,
+      "step": 402
+    },
+    {
+      "epoch": 0.02246376811594203,
+      "grad_norm": 0.37776532769203186,
+      "learning_rate": 8e-05,
+      "loss": 1.6039,
+      "step": 403
+    },
+    {
+      "epoch": 0.022519509476031215,
+      "grad_norm": 0.4386959969997406,
+      "learning_rate": 8e-05,
+      "loss": 1.765,
+      "step": 404
+    },
+    {
+      "epoch": 0.0225752508361204,
+      "grad_norm": 0.4808010756969452,
+      "learning_rate": 8e-05,
+      "loss": 1.5659,
+      "step": 405
+    },
+    {
+      "epoch": 0.022630992196209586,
+      "grad_norm": 0.4042125344276428,
+      "learning_rate": 8e-05,
+      "loss": 1.7711,
+      "step": 406
+    },
+    {
+      "epoch": 0.022686733556298775,
+      "grad_norm": 0.4260920584201813,
+      "learning_rate": 8e-05,
+      "loss": 1.6221,
+      "step": 407
+    },
+    {
+      "epoch": 0.02274247491638796,
+      "grad_norm": 0.4069453477859497,
+      "learning_rate": 8e-05,
+      "loss": 2.023,
+      "step": 408
+    },
+    {
+      "epoch": 0.022798216276477146,
+      "grad_norm": 0.38290277123451233,
+      "learning_rate": 8e-05,
+      "loss": 1.7505,
+      "step": 409
+    },
+    {
+      "epoch": 0.022853957636566332,
+      "grad_norm": 0.3907026946544647,
+      "learning_rate": 8e-05,
+      "loss": 1.8634,
+      "step": 410
+    },
+    {
+      "epoch": 0.022909698996655518,
+      "grad_norm": 0.4567570686340332,
+      "learning_rate": 8e-05,
+      "loss": 1.7813,
+      "step": 411
+    },
+    {
+      "epoch": 0.022965440356744703,
+      "grad_norm": 0.4327886998653412,
+      "learning_rate": 8e-05,
+      "loss": 1.8774,
+      "step": 412
+    },
+    {
+      "epoch": 0.02302118171683389,
+      "grad_norm": 0.4102731943130493,
+      "learning_rate": 8e-05,
+      "loss": 1.6824,
+      "step": 413
+    },
+    {
+      "epoch": 0.023076923076923078,
+      "grad_norm": 0.42088595032691956,
+      "learning_rate": 8e-05,
+      "loss": 1.8037,
+      "step": 414
+    },
+    {
+      "epoch": 0.023132664437012264,
+      "grad_norm": 0.40860041975975037,
+      "learning_rate": 8e-05,
+      "loss": 1.9645,
+      "step": 415
+    },
+    {
+      "epoch": 0.02318840579710145,
+      "grad_norm": 0.4047831594944,
+      "learning_rate": 8e-05,
+      "loss": 1.6969,
+      "step": 416
+    },
+    {
+      "epoch": 0.023244147157190635,
+      "grad_norm": 0.38738128542900085,
+      "learning_rate": 8e-05,
+      "loss": 1.7507,
+      "step": 417
+    },
+    {
+      "epoch": 0.02329988851727982,
+      "grad_norm": 0.4307592809200287,
+      "learning_rate": 8e-05,
+      "loss": 1.6861,
+      "step": 418
+    },
+    {
+      "epoch": 0.023355629877369007,
+      "grad_norm": 0.44377589225769043,
+      "learning_rate": 8e-05,
+      "loss": 1.5974,
+      "step": 419
+    },
+    {
+      "epoch": 0.023411371237458192,
+      "grad_norm": 0.42999017238616943,
+      "learning_rate": 8e-05,
+      "loss": 1.8469,
+      "step": 420
+    },
+    {
+      "epoch": 0.02346711259754738,
+      "grad_norm": 0.3810324966907501,
+      "learning_rate": 8e-05,
+      "loss": 1.5995,
+      "step": 421
+    },
+    {
+      "epoch": 0.023522853957636567,
+      "grad_norm": 0.42189478874206543,
+      "learning_rate": 8e-05,
+      "loss": 1.5233,
+      "step": 422
+    },
+    {
+      "epoch": 0.023578595317725753,
+      "grad_norm": 0.40228262543678284,
+      "learning_rate": 8e-05,
+      "loss": 1.5722,
+      "step": 423
+    },
+    {
+      "epoch": 0.02363433667781494,
+      "grad_norm": 0.44540834426879883,
+      "learning_rate": 8e-05,
+      "loss": 1.8335,
+      "step": 424
+    },
+    {
+      "epoch": 0.023690078037904124,
+      "grad_norm": 0.40963396430015564,
+      "learning_rate": 8e-05,
+      "loss": 1.6149,
+      "step": 425
+    },
+    {
+      "epoch": 0.02374581939799331,
+      "grad_norm": 0.4283663332462311,
+      "learning_rate": 8e-05,
+      "loss": 1.9118,
+      "step": 426
+    },
+    {
+      "epoch": 0.0238015607580825,
+      "grad_norm": 0.41021615266799927,
+      "learning_rate": 8e-05,
+      "loss": 1.5939,
+      "step": 427
+    },
+    {
+      "epoch": 0.023857302118171685,
+      "grad_norm": 0.40412041544914246,
+      "learning_rate": 8e-05,
+      "loss": 1.8285,
+      "step": 428
+    },
+    {
+      "epoch": 0.02391304347826087,
+      "grad_norm": 0.4041658639907837,
+      "learning_rate": 8e-05,
+      "loss": 1.8908,
+      "step": 429
+    },
+    {
+      "epoch": 0.023968784838350056,
+      "grad_norm": 0.44249221682548523,
+      "learning_rate": 8e-05,
+      "loss": 1.9286,
+      "step": 430
+    },
+    {
+      "epoch": 0.02402452619843924,
+      "grad_norm": 0.40249529480934143,
+      "learning_rate": 8e-05,
+      "loss": 1.7416,
+      "step": 431
+    },
+    {
+      "epoch": 0.024080267558528427,
+      "grad_norm": 0.4116725027561188,
+      "learning_rate": 8e-05,
+      "loss": 1.8227,
+      "step": 432
+    },
+    {
+      "epoch": 0.024136008918617613,
+      "grad_norm": 0.41705870628356934,
+      "learning_rate": 8e-05,
+      "loss": 1.7292,
+      "step": 433
+    },
+    {
+      "epoch": 0.024191750278706802,
+      "grad_norm": 0.428973525762558,
+      "learning_rate": 8e-05,
+      "loss": 1.8535,
+      "step": 434
+    },
+    {
+      "epoch": 0.024247491638795988,
+      "grad_norm": 0.39946043491363525,
+      "learning_rate": 8e-05,
+      "loss": 1.6571,
+      "step": 435
+    },
+    {
+      "epoch": 0.024303232998885173,
+      "grad_norm": 0.374835729598999,
+      "learning_rate": 8e-05,
+      "loss": 1.4186,
+      "step": 436
+    },
+    {
+      "epoch": 0.02435897435897436,
+      "grad_norm": 0.4034141004085541,
+      "learning_rate": 8e-05,
+      "loss": 1.5538,
+      "step": 437
+    },
+    {
+      "epoch": 0.024414715719063545,
+      "grad_norm": 0.5284610986709595,
+      "learning_rate": 8e-05,
+      "loss": 2.1798,
+      "step": 438
+    },
+    {
+      "epoch": 0.02447045707915273,
+      "grad_norm": 0.43948811292648315,
+      "learning_rate": 8e-05,
+      "loss": 1.8571,
+      "step": 439
+    },
+    {
+      "epoch": 0.024526198439241916,
+      "grad_norm": 0.4280640780925751,
+      "learning_rate": 8e-05,
+      "loss": 1.8725,
+      "step": 440
+    },
+    {
+      "epoch": 0.024581939799331105,
+      "grad_norm": 0.45238423347473145,
+      "learning_rate": 8e-05,
+      "loss": 1.8458,
+      "step": 441
+    },
+    {
+      "epoch": 0.02463768115942029,
+      "grad_norm": 0.4630393981933594,
+      "learning_rate": 8e-05,
+      "loss": 2.1301,
+      "step": 442
+    },
+    {
+      "epoch": 0.024693422519509477,
+      "grad_norm": 0.38115811347961426,
+      "learning_rate": 8e-05,
+      "loss": 1.5022,
+      "step": 443
+    },
+    {
+      "epoch": 0.024749163879598662,
+      "grad_norm": 0.41422200202941895,
+      "learning_rate": 8e-05,
+      "loss": 1.6895,
+      "step": 444
+    },
+    {
+      "epoch": 0.024804905239687848,
+      "grad_norm": 0.398720383644104,
+      "learning_rate": 8e-05,
+      "loss": 1.7783,
+      "step": 445
+    },
+    {
+      "epoch": 0.024860646599777034,
+      "grad_norm": 0.41158318519592285,
+      "learning_rate": 8e-05,
+      "loss": 1.7808,
+      "step": 446
+    },
+    {
+      "epoch": 0.02491638795986622,
+      "grad_norm": 0.39195117354393005,
+      "learning_rate": 8e-05,
+      "loss": 1.669,
+      "step": 447
+    },
+    {
+      "epoch": 0.02497212931995541,
+      "grad_norm": 0.45583266019821167,
+      "learning_rate": 8e-05,
+      "loss": 1.8407,
+      "step": 448
+    },
+    {
+      "epoch": 0.025027870680044594,
+      "grad_norm": 0.40098410844802856,
+      "learning_rate": 8e-05,
+      "loss": 1.7442,
+      "step": 449
+    },
+    {
+      "epoch": 0.02508361204013378,
+      "grad_norm": 0.4209321439266205,
+      "learning_rate": 8e-05,
+      "loss": 1.9494,
+      "step": 450
+    },
+    {
+      "epoch": 0.025139353400222966,
+      "grad_norm": 0.42204686999320984,
+      "learning_rate": 8e-05,
+      "loss": 1.7992,
+      "step": 451
+    },
+    {
+      "epoch": 0.02519509476031215,
+      "grad_norm": 0.4665510654449463,
+      "learning_rate": 8e-05,
+      "loss": 2.1564,
+      "step": 452
+    },
+    {
+      "epoch": 0.025250836120401337,
+      "grad_norm": 0.40209999680519104,
+      "learning_rate": 8e-05,
+      "loss": 1.7804,
+      "step": 453
+    },
+    {
+      "epoch": 0.025306577480490523,
+      "grad_norm": 0.4011898636817932,
+      "learning_rate": 8e-05,
+      "loss": 1.7567,
+      "step": 454
+    },
+    {
+      "epoch": 0.025362318840579712,
+      "grad_norm": 0.41723349690437317,
+      "learning_rate": 8e-05,
+      "loss": 1.899,
+      "step": 455
+    },
+    {
+      "epoch": 0.025418060200668897,
+      "grad_norm": 0.4524818956851959,
+      "learning_rate": 8e-05,
+      "loss": 2.0513,
+      "step": 456
+    },
+    {
+      "epoch": 0.025473801560758083,
+      "grad_norm": 0.4193991422653198,
+      "learning_rate": 8e-05,
+      "loss": 1.8053,
+      "step": 457
+    },
+    {
+      "epoch": 0.02552954292084727,
+      "grad_norm": 0.5020811557769775,
+      "learning_rate": 8e-05,
+      "loss": 1.993,
+      "step": 458
+    },
+    {
+      "epoch": 0.025585284280936454,
+      "grad_norm": 0.39976030588150024,
+      "learning_rate": 8e-05,
+      "loss": 1.7527,
+      "step": 459
+    },
+    {
+      "epoch": 0.02564102564102564,
+      "grad_norm": 0.43179136514663696,
+      "learning_rate": 8e-05,
+      "loss": 1.6838,
+      "step": 460
+    },
+    {
+      "epoch": 0.025696767001114826,
+      "grad_norm": 0.4227728545665741,
+      "learning_rate": 8e-05,
+      "loss": 1.8846,
+      "step": 461
+    },
+    {
+      "epoch": 0.025752508361204015,
+      "grad_norm": 0.40155425667762756,
+      "learning_rate": 8e-05,
+      "loss": 1.662,
+      "step": 462
+    },
+    {
+      "epoch": 0.0258082497212932,
+      "grad_norm": 0.4173693358898163,
+      "learning_rate": 8e-05,
+      "loss": 1.854,
+      "step": 463
+    },
+    {
+      "epoch": 0.025863991081382386,
+      "grad_norm": 0.41311848163604736,
+      "learning_rate": 8e-05,
+      "loss": 1.7414,
+      "step": 464
+    },
+    {
+      "epoch": 0.025919732441471572,
+      "grad_norm": 0.4107280671596527,
+      "learning_rate": 8e-05,
+      "loss": 1.6538,
+      "step": 465
+    },
+    {
+      "epoch": 0.025975473801560758,
+      "grad_norm": 0.4078384339809418,
+      "learning_rate": 8e-05,
+      "loss": 1.8702,
+      "step": 466
+    },
+    {
+      "epoch": 0.026031215161649943,
+      "grad_norm": 0.41355952620506287,
+      "learning_rate": 8e-05,
+      "loss": 1.8909,
+      "step": 467
+    },
+    {
+      "epoch": 0.02608695652173913,
+      "grad_norm": 0.4284036159515381,
+      "learning_rate": 8e-05,
+      "loss": 1.8563,
+      "step": 468
+    },
+    {
+      "epoch": 0.026142697881828318,
+      "grad_norm": 0.41841962933540344,
+      "learning_rate": 8e-05,
+      "loss": 1.7837,
+      "step": 469
+    },
+    {
+      "epoch": 0.026198439241917504,
+      "grad_norm": 0.4622490406036377,
+      "learning_rate": 8e-05,
+      "loss": 1.8756,
+      "step": 470
+    },
+    {
+      "epoch": 0.02625418060200669,
+      "grad_norm": 0.4452904760837555,
+      "learning_rate": 8e-05,
+      "loss": 2.0227,
+      "step": 471
+    },
+    {
+      "epoch": 0.026309921962095875,
+      "grad_norm": 0.47404196858406067,
+      "learning_rate": 8e-05,
+      "loss": 2.0884,
+      "step": 472
+    },
+    {
+      "epoch": 0.02636566332218506,
+      "grad_norm": 0.4138879179954529,
+      "learning_rate": 8e-05,
+      "loss": 1.8301,
+      "step": 473
+    },
+    {
+      "epoch": 0.026421404682274247,
+      "grad_norm": 0.3863980174064636,
+      "learning_rate": 8e-05,
+      "loss": 1.6119,
+      "step": 474
+    },
+    {
+      "epoch": 0.026477146042363432,
+      "grad_norm": 0.42295849323272705,
+      "learning_rate": 8e-05,
+      "loss": 1.752,
+      "step": 475
+    },
+    {
+      "epoch": 0.02653288740245262,
+      "grad_norm": 0.40239834785461426,
+      "learning_rate": 8e-05,
+      "loss": 2.001,
+      "step": 476
+    },
+    {
+      "epoch": 0.026588628762541807,
+      "grad_norm": 0.4178367555141449,
+      "learning_rate": 8e-05,
+      "loss": 1.8134,
+      "step": 477
+    },
+    {
+      "epoch": 0.026644370122630993,
+      "grad_norm": 0.41569623351097107,
+      "learning_rate": 8e-05,
+      "loss": 1.7635,
+      "step": 478
+    },
+    {
+      "epoch": 0.02670011148272018,
+      "grad_norm": 0.40798434615135193,
+      "learning_rate": 8e-05,
+      "loss": 1.787,
+      "step": 479
+    },
+    {
+      "epoch": 0.026755852842809364,
+      "grad_norm": 0.3902462124824524,
+      "learning_rate": 8e-05,
+      "loss": 1.6576,
+      "step": 480
+    },
+    {
+      "epoch": 0.02681159420289855,
+      "grad_norm": 0.3853597044944763,
+      "learning_rate": 8e-05,
+      "loss": 1.5487,
+      "step": 481
+    },
+    {
+      "epoch": 0.026867335562987735,
+      "grad_norm": 0.43139299750328064,
+      "learning_rate": 8e-05,
+      "loss": 1.9277,
+      "step": 482
+    },
+    {
+      "epoch": 0.026923076923076925,
+      "grad_norm": 0.41244861483573914,
+      "learning_rate": 8e-05,
+      "loss": 1.7304,
+      "step": 483
+    },
+    {
+      "epoch": 0.02697881828316611,
+      "grad_norm": 0.4112268090248108,
+      "learning_rate": 8e-05,
+      "loss": 1.7498,
+      "step": 484
+    },
+    {
+      "epoch": 0.027034559643255296,
+      "grad_norm": 0.40720945596694946,
+      "learning_rate": 8e-05,
+      "loss": 1.7,
+      "step": 485
+    },
+    {
+      "epoch": 0.02709030100334448,
+      "grad_norm": 0.45372486114501953,
+      "learning_rate": 8e-05,
+      "loss": 2.0107,
+      "step": 486
+    },
+    {
+      "epoch": 0.027146042363433667,
+      "grad_norm": 0.429392546415329,
+      "learning_rate": 8e-05,
+      "loss": 1.7736,
+      "step": 487
+    },
+    {
+      "epoch": 0.027201783723522853,
+      "grad_norm": 0.38353031873703003,
+      "learning_rate": 8e-05,
+      "loss": 1.6865,
+      "step": 488
+    },
+    {
+      "epoch": 0.02725752508361204,
+      "grad_norm": 0.41096165776252747,
+      "learning_rate": 8e-05,
+      "loss": 1.6531,
+      "step": 489
+    },
+    {
+      "epoch": 0.027313266443701228,
+      "grad_norm": 0.395088255405426,
+      "learning_rate": 8e-05,
+      "loss": 1.6689,
+      "step": 490
+    },
+    {
+      "epoch": 0.027369007803790413,
+      "grad_norm": 0.4200167655944824,
+      "learning_rate": 8e-05,
+      "loss": 1.8258,
+      "step": 491
+    },
+    {
+      "epoch": 0.0274247491638796,
+      "grad_norm": 0.3914308249950409,
+      "learning_rate": 8e-05,
+      "loss": 1.7873,
+      "step": 492
+    },
+    {
+      "epoch": 0.027480490523968785,
+      "grad_norm": 0.4021241068840027,
+      "learning_rate": 8e-05,
+      "loss": 1.7569,
+      "step": 493
+    },
+    {
+      "epoch": 0.02753623188405797,
+      "grad_norm": 0.42652398347854614,
+      "learning_rate": 8e-05,
+      "loss": 1.913,
+      "step": 494
+    },
+    {
+      "epoch": 0.027591973244147156,
+      "grad_norm": 0.43813228607177734,
+      "learning_rate": 8e-05,
+      "loss": 2.0001,
+      "step": 495
+    },
+    {
+      "epoch": 0.027647714604236342,
+      "grad_norm": 0.42736557126045227,
+      "learning_rate": 8e-05,
+      "loss": 1.751,
+      "step": 496
+    },
+    {
+      "epoch": 0.02770345596432553,
+      "grad_norm": 0.44482421875,
+      "learning_rate": 8e-05,
+      "loss": 1.9259,
+      "step": 497
+    },
+    {
+      "epoch": 0.027759197324414717,
+      "grad_norm": 0.4331215023994446,
+      "learning_rate": 8e-05,
+      "loss": 1.9296,
+      "step": 498
+    },
+    {
+      "epoch": 0.027814938684503902,
+      "grad_norm": 0.42177051305770874,
+      "learning_rate": 8e-05,
+      "loss": 1.9159,
+      "step": 499
+    },
+    {
+      "epoch": 0.027870680044593088,
+      "grad_norm": 0.4275566339492798,
+      "learning_rate": 8e-05,
+      "loss": 1.7328,
+      "step": 500
+    },
+    {
+      "epoch": 0.027926421404682274,
+      "grad_norm": 0.4206371605396271,
+      "learning_rate": 8e-05,
+      "loss": 1.8639,
+      "step": 501
+    },
+    {
+      "epoch": 0.02798216276477146,
+      "grad_norm": 0.4210817813873291,
+      "learning_rate": 8e-05,
+      "loss": 1.772,
+      "step": 502
+    },
+    {
+      "epoch": 0.028037904124860645,
+      "grad_norm": 0.41412290930747986,
+      "learning_rate": 8e-05,
+      "loss": 1.9401,
+      "step": 503
+    },
+    {
+      "epoch": 0.028093645484949834,
+      "grad_norm": 0.44537246227264404,
+      "learning_rate": 8e-05,
+      "loss": 2.1035,
+      "step": 504
+    },
+    {
+      "epoch": 0.02814938684503902,
+      "grad_norm": 0.3925379812717438,
+      "learning_rate": 8e-05,
+      "loss": 1.7115,
+      "step": 505
+    },
+    {
+      "epoch": 0.028205128205128206,
+      "grad_norm": 0.4123174846172333,
+      "learning_rate": 8e-05,
+      "loss": 1.7238,
+      "step": 506
+    },
+    {
+      "epoch": 0.02826086956521739,
+      "grad_norm": 0.4027958810329437,
+      "learning_rate": 8e-05,
+      "loss": 1.7405,
+      "step": 507
+    },
+    {
+      "epoch": 0.028316610925306577,
+      "grad_norm": 0.3883973956108093,
+      "learning_rate": 8e-05,
+      "loss": 1.7822,
+      "step": 508
+    },
+    {
+      "epoch": 0.028372352285395763,
+      "grad_norm": 0.42848825454711914,
+      "learning_rate": 8e-05,
+      "loss": 1.7291,
+      "step": 509
+    },
+    {
+      "epoch": 0.028428093645484948,
+      "grad_norm": 0.41434234380722046,
+      "learning_rate": 8e-05,
+      "loss": 1.7974,
+      "step": 510
+    },
+    {
+      "epoch": 0.028483835005574137,
+      "grad_norm": 0.44031184911727905,
+      "learning_rate": 8e-05,
+      "loss": 1.7749,
+      "step": 511
+    },
+    {
+      "epoch": 0.028539576365663323,
+      "grad_norm": 0.40377092361450195,
+      "learning_rate": 8e-05,
+      "loss": 1.5409,
+      "step": 512
+    },
+    {
+      "epoch": 0.02859531772575251,
+      "grad_norm": 0.4525328278541565,
+      "learning_rate": 8e-05,
+      "loss": 1.9709,
+      "step": 513
+    },
+    {
+      "epoch": 0.028651059085841694,
+      "grad_norm": 0.417842298746109,
+      "learning_rate": 8e-05,
+      "loss": 1.8349,
+      "step": 514
+    },
+    {
+      "epoch": 0.02870680044593088,
+      "grad_norm": 0.40975886583328247,
+      "learning_rate": 8e-05,
+      "loss": 1.7792,
+      "step": 515
+    },
+    {
+      "epoch": 0.028762541806020066,
+      "grad_norm": 0.40695229172706604,
+      "learning_rate": 8e-05,
+      "loss": 1.7144,
+      "step": 516
+    },
+    {
+      "epoch": 0.02881828316610925,
+      "grad_norm": 0.4242783784866333,
+      "learning_rate": 8e-05,
+      "loss": 1.8216,
+      "step": 517
+    },
+    {
+      "epoch": 0.02887402452619844,
+      "grad_norm": 0.40845513343811035,
+      "learning_rate": 8e-05,
+      "loss": 1.6328,
+      "step": 518
+    },
+    {
+      "epoch": 0.028929765886287626,
+      "grad_norm": 0.444774329662323,
+      "learning_rate": 8e-05,
+      "loss": 1.9723,
+      "step": 519
+    },
+    {
+      "epoch": 0.028985507246376812,
+      "grad_norm": 0.43584856390953064,
+      "learning_rate": 8e-05,
+      "loss": 1.7096,
+      "step": 520
+    },
+    {
+      "epoch": 0.029041248606465998,
+      "grad_norm": 0.39967840909957886,
+      "learning_rate": 8e-05,
+      "loss": 1.7449,
+      "step": 521
+    },
+    {
+      "epoch": 0.029096989966555183,
+      "grad_norm": 0.46941077709198,
+      "learning_rate": 8e-05,
+      "loss": 1.8926,
+      "step": 522
+    },
+    {
+      "epoch": 0.02915273132664437,
+      "grad_norm": 0.43449297547340393,
+      "learning_rate": 8e-05,
+      "loss": 1.7825,
+      "step": 523
+    },
+    {
+      "epoch": 0.029208472686733555,
+      "grad_norm": 0.3803488314151764,
+      "learning_rate": 8e-05,
+      "loss": 1.6198,
+      "step": 524
+    },
+    {
+      "epoch": 0.029264214046822744,
+      "grad_norm": 0.4477572739124298,
+      "learning_rate": 8e-05,
+      "loss": 1.7894,
+      "step": 525
+    },
+    {
+      "epoch": 0.02931995540691193,
+      "grad_norm": 0.3807078003883362,
+      "learning_rate": 8e-05,
+      "loss": 1.576,
+      "step": 526
+    },
+    {
+      "epoch": 0.029375696767001115,
+      "grad_norm": 0.45292210578918457,
+      "learning_rate": 8e-05,
+      "loss": 1.9879,
+      "step": 527
+    },
+    {
+      "epoch": 0.0294314381270903,
+      "grad_norm": 0.460365355014801,
+      "learning_rate": 8e-05,
+      "loss": 1.924,
+      "step": 528
+    },
+    {
+      "epoch": 0.029487179487179487,
+      "grad_norm": 0.4228498339653015,
+      "learning_rate": 8e-05,
+      "loss": 1.8188,
+      "step": 529
+    },
+    {
+      "epoch": 0.029542920847268672,
+      "grad_norm": 0.3947596251964569,
+      "learning_rate": 8e-05,
+      "loss": 1.6433,
+      "step": 530
+    },
+    {
+      "epoch": 0.029598662207357858,
+      "grad_norm": 0.3465898334980011,
+      "learning_rate": 8e-05,
+      "loss": 1.1918,
+      "step": 531
+    },
+    {
+      "epoch": 0.029654403567447047,
+      "grad_norm": 0.3989419639110565,
+      "learning_rate": 8e-05,
+      "loss": 1.7479,
+      "step": 532
+    },
+    {
+      "epoch": 0.029710144927536233,
+      "grad_norm": 0.4237585663795471,
+      "learning_rate": 8e-05,
+      "loss": 1.8069,
+      "step": 533
+    },
+    {
+      "epoch": 0.02976588628762542,
+      "grad_norm": 0.39906057715415955,
+      "learning_rate": 8e-05,
+      "loss": 1.6856,
+      "step": 534
+    },
+    {
+      "epoch": 0.029821627647714604,
+      "grad_norm": 0.4093904197216034,
+      "learning_rate": 8e-05,
+      "loss": 1.7184,
+      "step": 535
+    },
+    {
+      "epoch": 0.02987736900780379,
+      "grad_norm": 0.4156496524810791,
+      "learning_rate": 8e-05,
+      "loss": 1.753,
+      "step": 536
+    },
+    {
+      "epoch": 0.029933110367892975,
+      "grad_norm": 0.4160612225532532,
+      "learning_rate": 8e-05,
+      "loss": 1.8341,
+      "step": 537
+    },
+    {
+      "epoch": 0.02998885172798216,
+      "grad_norm": 0.3792247474193573,
+      "learning_rate": 8e-05,
+      "loss": 1.6885,
+      "step": 538
+    },
+    {
+      "epoch": 0.03004459308807135,
+      "grad_norm": 0.4268282353878021,
+      "learning_rate": 8e-05,
+      "loss": 1.8001,
+      "step": 539
+    },
+    {
+      "epoch": 0.030100334448160536,
+      "grad_norm": 0.411487340927124,
+      "learning_rate": 8e-05,
+      "loss": 1.7674,
+      "step": 540
+    },
+    {
+      "epoch": 0.03015607580824972,
+      "grad_norm": 0.37929436564445496,
+      "learning_rate": 8e-05,
+      "loss": 1.5027,
+      "step": 541
+    },
+    {
+      "epoch": 0.030211817168338907,
+      "grad_norm": 0.4008893072605133,
+      "learning_rate": 8e-05,
+      "loss": 1.6987,
+      "step": 542
+    },
+    {
+      "epoch": 0.030267558528428093,
+      "grad_norm": 0.4297763407230377,
+      "learning_rate": 8e-05,
+      "loss": 1.8859,
+      "step": 543
+    },
+    {
+      "epoch": 0.03032329988851728,
+      "grad_norm": 0.3972001373767853,
+      "learning_rate": 8e-05,
+      "loss": 1.626,
+      "step": 544
+    },
+    {
+      "epoch": 0.030379041248606464,
+      "grad_norm": 0.44114306569099426,
+      "learning_rate": 8e-05,
+      "loss": 1.7328,
+      "step": 545
+    },
+    {
+      "epoch": 0.030434782608695653,
+      "grad_norm": 0.38401421904563904,
+      "learning_rate": 8e-05,
+      "loss": 1.6671,
+      "step": 546
+    },
+    {
+      "epoch": 0.03049052396878484,
+      "grad_norm": 0.4116717576980591,
+      "learning_rate": 8e-05,
+      "loss": 1.8202,
+      "step": 547
+    },
+    {
+      "epoch": 0.030546265328874025,
+      "grad_norm": 0.4649137854576111,
+      "learning_rate": 8e-05,
+      "loss": 2.0108,
+      "step": 548
+    },
+    {
+      "epoch": 0.03060200668896321,
+      "grad_norm": 0.3979105055332184,
+      "learning_rate": 8e-05,
+      "loss": 1.7386,
+      "step": 549
+    },
+    {
+      "epoch": 0.030657748049052396,
+      "grad_norm": 0.39438769221305847,
+      "learning_rate": 8e-05,
+      "loss": 1.6171,
+      "step": 550
+    },
+    {
+      "epoch": 0.030713489409141582,
+      "grad_norm": 0.41145819425582886,
+      "learning_rate": 8e-05,
+      "loss": 1.7747,
+      "step": 551
+    },
+    {
+      "epoch": 0.03076923076923077,
+      "grad_norm": 0.37144097685813904,
+      "learning_rate": 8e-05,
+      "loss": 1.6617,
+      "step": 552
+    },
+    {
+      "epoch": 0.030824972129319957,
+      "grad_norm": 0.40567418932914734,
+      "learning_rate": 8e-05,
+      "loss": 1.5851,
+      "step": 553
+    },
+    {
+      "epoch": 0.030880713489409142,
+      "grad_norm": 0.3914574980735779,
+      "learning_rate": 8e-05,
+      "loss": 1.838,
+      "step": 554
+    },
+    {
+      "epoch": 0.030936454849498328,
+      "grad_norm": 0.4121246337890625,
+      "learning_rate": 8e-05,
+      "loss": 1.8852,
+      "step": 555
+    },
+    {
+      "epoch": 0.030992196209587514,
+      "grad_norm": 0.4243144989013672,
+      "learning_rate": 8e-05,
+      "loss": 1.8316,
+      "step": 556
+    },
+    {
+      "epoch": 0.0310479375696767,
+      "grad_norm": 0.4198228716850281,
+      "learning_rate": 8e-05,
+      "loss": 1.7642,
+      "step": 557
+    },
+    {
+      "epoch": 0.031103678929765885,
+      "grad_norm": 0.41224730014801025,
+      "learning_rate": 8e-05,
+      "loss": 1.9605,
+      "step": 558
+    },
+    {
+      "epoch": 0.031159420289855074,
+      "grad_norm": 0.4147040545940399,
+      "learning_rate": 8e-05,
+      "loss": 1.7033,
+      "step": 559
+    },
+    {
+      "epoch": 0.03121516164994426,
+      "grad_norm": 0.4140629768371582,
+      "learning_rate": 8e-05,
+      "loss": 1.9389,
+      "step": 560
+    },
+    {
+      "epoch": 0.03127090301003344,
+      "grad_norm": 0.42293891310691833,
+      "learning_rate": 8e-05,
+      "loss": 1.7608,
+      "step": 561
+    },
+    {
+      "epoch": 0.03132664437012263,
+      "grad_norm": 0.398647665977478,
+      "learning_rate": 8e-05,
+      "loss": 1.7532,
+      "step": 562
+    },
+    {
+      "epoch": 0.03138238573021182,
+      "grad_norm": 0.4059430956840515,
+      "learning_rate": 8e-05,
+      "loss": 1.7975,
+      "step": 563
+    },
+    {
+      "epoch": 0.031438127090301006,
+      "grad_norm": 0.45980408787727356,
+      "learning_rate": 8e-05,
+      "loss": 1.8965,
+      "step": 564
+    },
+    {
+      "epoch": 0.03149386845039019,
+      "grad_norm": 0.4658120274543762,
+      "learning_rate": 8e-05,
+      "loss": 2.0726,
+      "step": 565
+    },
+    {
+      "epoch": 0.03154960981047938,
+      "grad_norm": 0.46347862482070923,
+      "learning_rate": 8e-05,
+      "loss": 1.8033,
+      "step": 566
+    },
+    {
+      "epoch": 0.03160535117056856,
+      "grad_norm": 0.40268826484680176,
+      "learning_rate": 8e-05,
+      "loss": 1.7909,
+      "step": 567
+    },
+    {
+      "epoch": 0.03166109253065775,
+      "grad_norm": 0.4346879720687866,
+      "learning_rate": 8e-05,
+      "loss": 1.7894,
+      "step": 568
+    },
+    {
+      "epoch": 0.031716833890746934,
+      "grad_norm": 0.423518568277359,
+      "learning_rate": 8e-05,
+      "loss": 1.8991,
+      "step": 569
+    },
+    {
+      "epoch": 0.03177257525083612,
+      "grad_norm": 0.42770105600357056,
+      "learning_rate": 8e-05,
+      "loss": 1.8731,
+      "step": 570
+    },
+    {
+      "epoch": 0.031828316610925306,
+      "grad_norm": 0.4165306091308594,
+      "learning_rate": 8e-05,
+      "loss": 1.7888,
+      "step": 571
+    },
+    {
+      "epoch": 0.03188405797101449,
+      "grad_norm": 0.40096932649612427,
+      "learning_rate": 8e-05,
+      "loss": 1.947,
+      "step": 572
+    },
+    {
+      "epoch": 0.03193979933110368,
+      "grad_norm": 0.3981882631778717,
+      "learning_rate": 8e-05,
+      "loss": 1.715,
+      "step": 573
+    },
+    {
+      "epoch": 0.03199554069119286,
+      "grad_norm": 0.4129897654056549,
+      "learning_rate": 8e-05,
+      "loss": 1.8301,
+      "step": 574
+    },
+    {
+      "epoch": 0.03205128205128205,
+      "grad_norm": 0.40707018971443176,
+      "learning_rate": 8e-05,
+      "loss": 1.4342,
+      "step": 575
+    },
+    {
+      "epoch": 0.032107023411371234,
+      "grad_norm": 0.4106016457080841,
+      "learning_rate": 8e-05,
+      "loss": 1.6461,
+      "step": 576
+    },
+    {
+      "epoch": 0.03216276477146043,
+      "grad_norm": 0.4073600769042969,
+      "learning_rate": 8e-05,
+      "loss": 1.5317,
+      "step": 577
+    },
+    {
+      "epoch": 0.03221850613154961,
+      "grad_norm": 0.4160962402820587,
+      "learning_rate": 8e-05,
+      "loss": 1.8114,
+      "step": 578
+    },
+    {
+      "epoch": 0.0322742474916388,
+      "grad_norm": 0.4713279902935028,
+      "learning_rate": 8e-05,
+      "loss": 2.0019,
+      "step": 579
+    },
+    {
+      "epoch": 0.032329988851727984,
+      "grad_norm": 0.40895307064056396,
+      "learning_rate": 8e-05,
+      "loss": 1.8948,
+      "step": 580
+    },
+    {
+      "epoch": 0.03238573021181717,
+      "grad_norm": 0.37472397089004517,
+      "learning_rate": 8e-05,
+      "loss": 1.7124,
+      "step": 581
+    },
+    {
+      "epoch": 0.032441471571906355,
+      "grad_norm": 0.3960634171962738,
+      "learning_rate": 8e-05,
+      "loss": 1.7363,
+      "step": 582
+    },
+    {
+      "epoch": 0.03249721293199554,
+      "grad_norm": 0.3730868101119995,
+      "learning_rate": 8e-05,
+      "loss": 1.7037,
+      "step": 583
+    },
+    {
+      "epoch": 0.032552954292084726,
+      "grad_norm": 0.36749905347824097,
+      "learning_rate": 8e-05,
+      "loss": 1.5133,
+      "step": 584
+    },
+    {
+      "epoch": 0.03260869565217391,
+      "grad_norm": 0.40374237298965454,
+      "learning_rate": 8e-05,
+      "loss": 1.7778,
+      "step": 585
+    },
+    {
+      "epoch": 0.0326644370122631,
+      "grad_norm": 0.43943536281585693,
+      "learning_rate": 8e-05,
+      "loss": 1.957,
+      "step": 586
+    },
+    {
+      "epoch": 0.032720178372352283,
+      "grad_norm": 0.4169772267341614,
+      "learning_rate": 8e-05,
+      "loss": 1.7456,
+      "step": 587
+    },
+    {
+      "epoch": 0.03277591973244147,
+      "grad_norm": 0.4045918583869934,
+      "learning_rate": 8e-05,
+      "loss": 1.6216,
+      "step": 588
+    },
+    {
+      "epoch": 0.032831661092530655,
+      "grad_norm": 0.41666683554649353,
+      "learning_rate": 8e-05,
+      "loss": 1.9929,
+      "step": 589
+    },
+    {
+      "epoch": 0.03288740245261984,
+      "grad_norm": 0.38340914249420166,
+      "learning_rate": 8e-05,
+      "loss": 1.6984,
+      "step": 590
+    },
+    {
+      "epoch": 0.03294314381270903,
+      "grad_norm": 0.41442227363586426,
+      "learning_rate": 8e-05,
+      "loss": 1.8554,
+      "step": 591
+    },
+    {
+      "epoch": 0.03299888517279822,
+      "grad_norm": 0.4207206666469574,
+      "learning_rate": 8e-05,
+      "loss": 1.8215,
+      "step": 592
+    },
+    {
+      "epoch": 0.033054626532887404,
+      "grad_norm": 0.4293190538883209,
+      "learning_rate": 8e-05,
+      "loss": 1.9728,
+      "step": 593
+    },
+    {
+      "epoch": 0.03311036789297659,
+      "grad_norm": 0.4169306755065918,
+      "learning_rate": 8e-05,
+      "loss": 1.8503,
+      "step": 594
+    },
+    {
+      "epoch": 0.033166109253065776,
+      "grad_norm": 0.4055793583393097,
+      "learning_rate": 8e-05,
+      "loss": 1.7387,
+      "step": 595
+    },
+    {
+      "epoch": 0.03322185061315496,
+      "grad_norm": 0.4573780596256256,
+      "learning_rate": 8e-05,
+      "loss": 1.911,
+      "step": 596
+    },
+    {
+      "epoch": 0.03327759197324415,
+      "grad_norm": 0.3761114180088043,
+      "learning_rate": 8e-05,
+      "loss": 1.6057,
+      "step": 597
+    },
+    {
+      "epoch": 0.03333333333333333,
+      "grad_norm": 0.4141940772533417,
+      "learning_rate": 8e-05,
+      "loss": 1.7041,
+      "step": 598
+    },
+    {
+      "epoch": 0.03338907469342252,
+      "grad_norm": 0.4186020791530609,
+      "learning_rate": 8e-05,
+      "loss": 1.8301,
+      "step": 599
+    },
+    {
+      "epoch": 0.033444816053511704,
+      "grad_norm": 0.42576029896736145,
+      "learning_rate": 8e-05,
+      "loss": 1.8344,
+      "step": 600
+    },
+    {
+      "epoch": 0.03350055741360089,
+      "grad_norm": 0.41014382243156433,
+      "learning_rate": 8e-05,
+      "loss": 1.8637,
+      "step": 601
+    },
+    {
+      "epoch": 0.033556298773690076,
+      "grad_norm": 0.4340376555919647,
+      "learning_rate": 8e-05,
+      "loss": 1.7651,
+      "step": 602
+    },
+    {
+      "epoch": 0.03361204013377926,
+      "grad_norm": 0.4143203794956207,
+      "learning_rate": 8e-05,
+      "loss": 1.7698,
+      "step": 603
+    },
+    {
+      "epoch": 0.03366778149386845,
+      "grad_norm": 0.42128315567970276,
+      "learning_rate": 8e-05,
+      "loss": 1.7218,
+      "step": 604
+    },
+    {
+      "epoch": 0.03372352285395764,
+      "grad_norm": 0.4313509166240692,
+      "learning_rate": 8e-05,
+      "loss": 1.8213,
+      "step": 605
+    },
+    {
+      "epoch": 0.033779264214046825,
+      "grad_norm": 0.3960542678833008,
+      "learning_rate": 8e-05,
+      "loss": 1.4856,
+      "step": 606
+    },
+    {
+      "epoch": 0.03383500557413601,
+      "grad_norm": 0.4187854528427124,
+      "learning_rate": 8e-05,
+      "loss": 1.7748,
+      "step": 607
+    },
+    {
+      "epoch": 0.0338907469342252,
+      "grad_norm": 0.4670550227165222,
+      "learning_rate": 8e-05,
+      "loss": 1.9125,
+      "step": 608
+    },
+    {
+      "epoch": 0.03394648829431438,
+      "grad_norm": 0.4087151885032654,
+      "learning_rate": 8e-05,
+      "loss": 1.6741,
+      "step": 609
+    },
+    {
+      "epoch": 0.03400222965440357,
+      "grad_norm": 0.4307301640510559,
+      "learning_rate": 8e-05,
+      "loss": 1.8006,
+      "step": 610
+    },
+    {
+      "epoch": 0.034057971014492754,
+      "grad_norm": 0.46052247285842896,
+      "learning_rate": 8e-05,
+      "loss": 1.8099,
+      "step": 611
+    },
+    {
+      "epoch": 0.03411371237458194,
+      "grad_norm": 0.41967251896858215,
+      "learning_rate": 8e-05,
+      "loss": 1.7435,
+      "step": 612
+    },
+    {
+      "epoch": 0.034169453734671125,
+      "grad_norm": 0.42859208583831787,
+      "learning_rate": 8e-05,
+      "loss": 1.8204,
+      "step": 613
+    },
+    {
+      "epoch": 0.03422519509476031,
+      "grad_norm": 0.4440454840660095,
+      "learning_rate": 8e-05,
+      "loss": 1.8071,
+      "step": 614
+    },
+    {
+      "epoch": 0.034280936454849496,
+      "grad_norm": 0.4726400673389435,
+      "learning_rate": 8e-05,
+      "loss": 2.0625,
+      "step": 615
+    },
+    {
+      "epoch": 0.03433667781493868,
+      "grad_norm": 0.43250757455825806,
+      "learning_rate": 8e-05,
+      "loss": 1.8348,
+      "step": 616
+    },
+    {
+      "epoch": 0.03439241917502787,
+      "grad_norm": 0.4797033369541168,
+      "learning_rate": 8e-05,
+      "loss": 1.9095,
+      "step": 617
+    },
+    {
+      "epoch": 0.03444816053511706,
+      "grad_norm": 0.421512633562088,
+      "learning_rate": 8e-05,
+      "loss": 1.717,
+      "step": 618
+    },
+    {
+      "epoch": 0.034503901895206246,
+      "grad_norm": 0.37292420864105225,
+      "learning_rate": 8e-05,
+      "loss": 1.651,
+      "step": 619
+    },
+    {
+      "epoch": 0.03455964325529543,
+      "grad_norm": 0.5055415630340576,
+      "learning_rate": 8e-05,
+      "loss": 1.6273,
+      "step": 620
+    },
+    {
+      "epoch": 0.03461538461538462,
+      "grad_norm": 0.4223185181617737,
+      "learning_rate": 8e-05,
+      "loss": 1.7644,
+      "step": 621
+    },
+    {
+      "epoch": 0.0346711259754738,
+      "grad_norm": 0.42007625102996826,
+      "learning_rate": 8e-05,
+      "loss": 1.7028,
+      "step": 622
+    },
+    {
+      "epoch": 0.03472686733556299,
+      "grad_norm": 0.4167858958244324,
+      "learning_rate": 8e-05,
+      "loss": 1.8177,
+      "step": 623
+    },
+    {
+      "epoch": 0.034782608695652174,
+      "grad_norm": 0.47208330035209656,
+      "learning_rate": 8e-05,
+      "loss": 1.7619,
+      "step": 624
+    },
+    {
+      "epoch": 0.03483835005574136,
+      "grad_norm": 0.3962777853012085,
+      "learning_rate": 8e-05,
+      "loss": 1.7748,
+      "step": 625
+    },
+    {
+      "epoch": 0.034894091415830546,
+      "grad_norm": 0.3653423488140106,
+      "learning_rate": 8e-05,
+      "loss": 1.5153,
+      "step": 626
+    },
+    {
+      "epoch": 0.03494983277591973,
+      "grad_norm": 0.38869357109069824,
+      "learning_rate": 8e-05,
+      "loss": 1.623,
+      "step": 627
+    },
+    {
+      "epoch": 0.03500557413600892,
+      "grad_norm": 0.43486788868904114,
+      "learning_rate": 8e-05,
+      "loss": 1.8753,
+      "step": 628
+    },
+    {
+      "epoch": 0.0350613154960981,
+      "grad_norm": 0.38648611307144165,
+      "learning_rate": 8e-05,
+      "loss": 1.6221,
+      "step": 629
+    },
+    {
+      "epoch": 0.03511705685618729,
+      "grad_norm": 0.41406360268592834,
+      "learning_rate": 8e-05,
+      "loss": 1.9074,
+      "step": 630
+    },
+    {
+      "epoch": 0.035172798216276474,
+      "grad_norm": 0.45213326811790466,
+      "learning_rate": 8e-05,
+      "loss": 1.9577,
+      "step": 631
+    },
+    {
+      "epoch": 0.03522853957636567,
+      "grad_norm": 0.4049549698829651,
+      "learning_rate": 8e-05,
+      "loss": 1.8268,
+      "step": 632
+    },
+    {
+      "epoch": 0.03528428093645485,
+      "grad_norm": 0.39177536964416504,
+      "learning_rate": 8e-05,
+      "loss": 1.3487,
+      "step": 633
+    },
+    {
+      "epoch": 0.03534002229654404,
+      "grad_norm": 0.42474237084388733,
+      "learning_rate": 8e-05,
+      "loss": 1.8953,
+      "step": 634
+    },
+    {
+      "epoch": 0.035395763656633224,
+      "grad_norm": 0.42674046754837036,
+      "learning_rate": 8e-05,
+      "loss": 1.8642,
+      "step": 635
+    },
+    {
+      "epoch": 0.03545150501672241,
+      "grad_norm": 0.4271129071712494,
+      "learning_rate": 8e-05,
+      "loss": 1.8885,
+      "step": 636
+    },
+    {
+      "epoch": 0.035507246376811595,
+      "grad_norm": 0.43485212326049805,
+      "learning_rate": 8e-05,
+      "loss": 1.812,
+      "step": 637
+    },
+    {
+      "epoch": 0.03556298773690078,
+      "grad_norm": 0.4561612606048584,
+      "learning_rate": 8e-05,
+      "loss": 1.8138,
+      "step": 638
+    },
+    {
+      "epoch": 0.035618729096989966,
+      "grad_norm": 0.4725930094718933,
+      "learning_rate": 8e-05,
+      "loss": 1.9059,
+      "step": 639
+    },
+    {
+      "epoch": 0.03567447045707915,
+      "grad_norm": 0.468065083026886,
+      "learning_rate": 8e-05,
+      "loss": 1.7419,
+      "step": 640
+    },
+    {
+      "epoch": 0.03573021181716834,
+      "grad_norm": 0.43373408913612366,
+      "learning_rate": 8e-05,
+      "loss": 1.9229,
+      "step": 641
+    },
+    {
+      "epoch": 0.03578595317725752,
+      "grad_norm": 0.4082108736038208,
+      "learning_rate": 8e-05,
+      "loss": 1.8043,
+      "step": 642
+    },
+    {
+      "epoch": 0.03584169453734671,
+      "grad_norm": 0.3930438756942749,
+      "learning_rate": 8e-05,
+      "loss": 1.6384,
+      "step": 643
+    },
+    {
+      "epoch": 0.035897435897435895,
+      "grad_norm": 0.39798983931541443,
+      "learning_rate": 8e-05,
+      "loss": 1.7297,
+      "step": 644
+    },
+    {
+      "epoch": 0.03595317725752508,
+      "grad_norm": 0.4810659885406494,
+      "learning_rate": 8e-05,
+      "loss": 1.8956,
+      "step": 645
+    },
+    {
+      "epoch": 0.03600891861761427,
+      "grad_norm": 0.4323861300945282,
+      "learning_rate": 8e-05,
+      "loss": 1.8167,
+      "step": 646
+    },
+    {
+      "epoch": 0.03606465997770346,
+      "grad_norm": 0.42571496963500977,
+      "learning_rate": 8e-05,
+      "loss": 1.9223,
+      "step": 647
+    },
+    {
+      "epoch": 0.036120401337792644,
+      "grad_norm": 0.4450274109840393,
+      "learning_rate": 8e-05,
+      "loss": 1.8361,
+      "step": 648
+    },
+    {
+      "epoch": 0.03617614269788183,
+      "grad_norm": 0.4110467731952667,
+      "learning_rate": 8e-05,
+      "loss": 1.6817,
+      "step": 649
+    },
+    {
+      "epoch": 0.036231884057971016,
+      "grad_norm": 0.39012429118156433,
+      "learning_rate": 8e-05,
+      "loss": 1.6442,
+      "step": 650
+    },
+    {
+      "epoch": 0.0362876254180602,
+      "grad_norm": 0.39485234022140503,
+      "learning_rate": 8e-05,
+      "loss": 1.7741,
+      "step": 651
+    },
+    {
+      "epoch": 0.03634336677814939,
+      "grad_norm": 0.4552437961101532,
+      "learning_rate": 8e-05,
+      "loss": 1.8181,
+      "step": 652
+    },
+    {
+      "epoch": 0.03639910813823857,
+      "grad_norm": 0.4294509291648865,
+      "learning_rate": 8e-05,
+      "loss": 1.6509,
+      "step": 653
+    },
+    {
+      "epoch": 0.03645484949832776,
+      "grad_norm": 0.3852769136428833,
+      "learning_rate": 8e-05,
+      "loss": 1.6274,
+      "step": 654
+    },
+    {
+      "epoch": 0.036510590858416944,
+      "grad_norm": 0.4140886962413788,
+      "learning_rate": 8e-05,
+      "loss": 1.6733,
+      "step": 655
+    },
+    {
+      "epoch": 0.03656633221850613,
+      "grad_norm": 0.4174540638923645,
+      "learning_rate": 8e-05,
+      "loss": 1.7846,
+      "step": 656
+    },
+    {
+      "epoch": 0.036622073578595316,
+      "grad_norm": 0.41302043199539185,
+      "learning_rate": 8e-05,
+      "loss": 1.6438,
+      "step": 657
+    },
+    {
+      "epoch": 0.0366778149386845,
+      "grad_norm": 0.44165873527526855,
+      "learning_rate": 8e-05,
+      "loss": 1.5783,
+      "step": 658
+    },
+    {
+      "epoch": 0.03673355629877369,
+      "grad_norm": 0.4951744079589844,
+      "learning_rate": 8e-05,
+      "loss": 1.9361,
+      "step": 659
+    },
+    {
+      "epoch": 0.03678929765886288,
+      "grad_norm": 0.4248286485671997,
+      "learning_rate": 8e-05,
+      "loss": 1.7216,
+      "step": 660
+    },
+    {
+      "epoch": 0.036845039018952065,
+      "grad_norm": 0.4633992314338684,
+      "learning_rate": 8e-05,
+      "loss": 1.7314,
+      "step": 661
+    },
+    {
+      "epoch": 0.03690078037904125,
+      "grad_norm": 0.3990238606929779,
+      "learning_rate": 8e-05,
+      "loss": 1.6677,
+      "step": 662
+    },
+    {
+      "epoch": 0.03695652173913044,
+      "grad_norm": 0.40574949979782104,
+      "learning_rate": 8e-05,
+      "loss": 1.6418,
+      "step": 663
+    },
+    {
+      "epoch": 0.03701226309921962,
+      "grad_norm": 0.37986770272254944,
+      "learning_rate": 8e-05,
+      "loss": 1.5045,
+      "step": 664
+    },
+    {
+      "epoch": 0.03706800445930881,
+      "grad_norm": 0.4152768552303314,
+      "learning_rate": 8e-05,
+      "loss": 1.6917,
+      "step": 665
+    },
+    {
+      "epoch": 0.037123745819397994,
+      "grad_norm": 0.4041266441345215,
+      "learning_rate": 8e-05,
+      "loss": 1.7146,
+      "step": 666
+    },
+    {
+      "epoch": 0.03717948717948718,
+      "grad_norm": 0.46567487716674805,
+      "learning_rate": 8e-05,
+      "loss": 1.8633,
+      "step": 667
+    },
+    {
+      "epoch": 0.037235228539576365,
+      "grad_norm": 0.4492206275463104,
+      "learning_rate": 8e-05,
+      "loss": 1.8854,
+      "step": 668
+    },
+    {
+      "epoch": 0.03729096989966555,
+      "grad_norm": 0.4595526456832886,
+      "learning_rate": 8e-05,
+      "loss": 1.458,
+      "step": 669
+    },
+    {
+      "epoch": 0.037346711259754736,
+      "grad_norm": 0.43089577555656433,
+      "learning_rate": 8e-05,
+      "loss": 1.9997,
+      "step": 670
+    },
+    {
+      "epoch": 0.03740245261984392,
+      "grad_norm": 0.4291883409023285,
+      "learning_rate": 8e-05,
+      "loss": 1.7632,
+      "step": 671
+    },
+    {
+      "epoch": 0.03745819397993311,
+      "grad_norm": 0.42062389850616455,
+      "learning_rate": 8e-05,
+      "loss": 1.7392,
+      "step": 672
+    },
+    {
+      "epoch": 0.03751393534002229,
+      "grad_norm": 0.4016992449760437,
+      "learning_rate": 8e-05,
+      "loss": 1.7487,
+      "step": 673
+    },
+    {
+      "epoch": 0.037569676700111486,
+      "grad_norm": 0.40747544169425964,
+      "learning_rate": 8e-05,
+      "loss": 1.6961,
+      "step": 674
+    },
+    {
+      "epoch": 0.03762541806020067,
+      "grad_norm": 0.40056928992271423,
+      "learning_rate": 8e-05,
+      "loss": 1.6378,
+      "step": 675
+    },
+    {
+      "epoch": 0.03768115942028986,
+      "grad_norm": 0.4546568989753723,
+      "learning_rate": 8e-05,
+      "loss": 1.928,
+      "step": 676
+    },
+    {
+      "epoch": 0.03773690078037904,
+      "grad_norm": 0.41371530294418335,
+      "learning_rate": 8e-05,
+      "loss": 1.7306,
+      "step": 677
+    },
+    {
+      "epoch": 0.03779264214046823,
+      "grad_norm": 0.4063621163368225,
+      "learning_rate": 8e-05,
+      "loss": 1.8541,
+      "step": 678
+    },
+    {
+      "epoch": 0.037848383500557414,
+      "grad_norm": 0.45284903049468994,
+      "learning_rate": 8e-05,
+      "loss": 1.7621,
+      "step": 679
+    },
+    {
+      "epoch": 0.0379041248606466,
+      "grad_norm": 0.3936507999897003,
+      "learning_rate": 8e-05,
+      "loss": 1.6523,
+      "step": 680
+    },
+    {
+      "epoch": 0.037959866220735786,
+      "grad_norm": 0.4094870984554291,
+      "learning_rate": 8e-05,
+      "loss": 1.8397,
+      "step": 681
+    },
+    {
+      "epoch": 0.03801560758082497,
+      "grad_norm": 0.4652598798274994,
+      "learning_rate": 8e-05,
+      "loss": 1.956,
+      "step": 682
+    },
+    {
+      "epoch": 0.03807134894091416,
+      "grad_norm": 0.4646802842617035,
+      "learning_rate": 8e-05,
+      "loss": 2.0511,
+      "step": 683
+    },
+    {
+      "epoch": 0.03812709030100334,
+      "grad_norm": 0.4306330680847168,
+      "learning_rate": 8e-05,
+      "loss": 1.7777,
+      "step": 684
+    },
+    {
+      "epoch": 0.03818283166109253,
+      "grad_norm": 0.4153509736061096,
+      "learning_rate": 8e-05,
+      "loss": 1.7838,
+      "step": 685
+    },
+    {
+      "epoch": 0.038238573021181714,
+      "grad_norm": 0.4056133031845093,
+      "learning_rate": 8e-05,
+      "loss": 1.6948,
+      "step": 686
+    },
+    {
+      "epoch": 0.0382943143812709,
+      "grad_norm": 0.42202329635620117,
+      "learning_rate": 8e-05,
+      "loss": 1.574,
+      "step": 687
+    },
+    {
+      "epoch": 0.03835005574136009,
+      "grad_norm": 0.4424411952495575,
+      "learning_rate": 8e-05,
+      "loss": 1.7239,
+      "step": 688
+    },
+    {
+      "epoch": 0.03840579710144928,
+      "grad_norm": 0.404994934797287,
+      "learning_rate": 8e-05,
+      "loss": 1.6564,
+      "step": 689
+    },
+    {
+      "epoch": 0.038461538461538464,
+      "grad_norm": 0.44672891497612,
+      "learning_rate": 8e-05,
+      "loss": 2.0652,
+      "step": 690
+    },
+    {
+      "epoch": 0.03851727982162765,
+      "grad_norm": 0.357619971036911,
+      "learning_rate": 8e-05,
+      "loss": 1.5158,
+      "step": 691
+    },
+    {
+      "epoch": 0.038573021181716835,
+      "grad_norm": 0.4439055621623993,
+      "learning_rate": 8e-05,
+      "loss": 1.9357,
+      "step": 692
+    },
+    {
+      "epoch": 0.03862876254180602,
+      "grad_norm": 0.4590149521827698,
+      "learning_rate": 8e-05,
+      "loss": 1.8221,
+      "step": 693
+    },
+    {
+      "epoch": 0.038684503901895206,
+      "grad_norm": 0.4029330909252167,
+      "learning_rate": 8e-05,
+      "loss": 1.5921,
+      "step": 694
+    },
+    {
+      "epoch": 0.03874024526198439,
+      "grad_norm": 0.4233595132827759,
+      "learning_rate": 8e-05,
+      "loss": 1.4612,
+      "step": 695
+    },
+    {
+      "epoch": 0.03879598662207358,
+      "grad_norm": 0.47652146220207214,
+      "learning_rate": 8e-05,
+      "loss": 1.6849,
+      "step": 696
+    },
+    {
+      "epoch": 0.03885172798216276,
+      "grad_norm": 0.45156988501548767,
+      "learning_rate": 8e-05,
+      "loss": 1.8079,
+      "step": 697
+    },
+    {
+      "epoch": 0.03890746934225195,
+      "grad_norm": 0.44426658749580383,
+      "learning_rate": 8e-05,
+      "loss": 1.7389,
+      "step": 698
+    },
+    {
+      "epoch": 0.038963210702341135,
+      "grad_norm": 0.4575849771499634,
+      "learning_rate": 8e-05,
+      "loss": 1.902,
+      "step": 699
+    },
+    {
+      "epoch": 0.03901895206243032,
+      "grad_norm": 0.4535796046257019,
+      "learning_rate": 8e-05,
+      "loss": 1.5685,
+      "step": 700
+    },
+    {
+      "epoch": 0.039074693422519506,
+      "grad_norm": 0.4648195207118988,
+      "learning_rate": 8e-05,
+      "loss": 1.8874,
+      "step": 701
+    },
+    {
+      "epoch": 0.0391304347826087,
+      "grad_norm": 0.44183456897735596,
+      "learning_rate": 8e-05,
+      "loss": 1.8103,
+      "step": 702
+    },
+    {
+      "epoch": 0.039186176142697884,
+      "grad_norm": 0.37236881256103516,
+      "learning_rate": 8e-05,
+      "loss": 1.6319,
+      "step": 703
+    },
+    {
+      "epoch": 0.03924191750278707,
+      "grad_norm": 0.4322548508644104,
+      "learning_rate": 8e-05,
+      "loss": 1.7023,
+      "step": 704
+    },
+    {
+      "epoch": 0.039297658862876256,
+      "grad_norm": 0.4289608597755432,
+      "learning_rate": 8e-05,
+      "loss": 1.9537,
+      "step": 705
+    },
+    {
+      "epoch": 0.03935340022296544,
+      "grad_norm": 0.4339289367198944,
+      "learning_rate": 8e-05,
+      "loss": 1.8794,
+      "step": 706
+    },
+    {
+      "epoch": 0.03940914158305463,
+      "grad_norm": 0.451537549495697,
+      "learning_rate": 8e-05,
+      "loss": 1.8869,
+      "step": 707
+    },
+    {
+      "epoch": 0.03946488294314381,
+      "grad_norm": 0.4093172550201416,
+      "learning_rate": 8e-05,
+      "loss": 1.491,
+      "step": 708
+    },
+    {
+      "epoch": 0.039520624303233,
+      "grad_norm": 0.5952094793319702,
+      "learning_rate": 8e-05,
+      "loss": 1.6896,
+      "step": 709
+    },
+    {
+      "epoch": 0.039576365663322184,
+      "grad_norm": 0.4200764000415802,
+      "learning_rate": 8e-05,
+      "loss": 1.6626,
+      "step": 710
+    },
+    {
+      "epoch": 0.03963210702341137,
+      "grad_norm": 0.39256078004837036,
+      "learning_rate": 8e-05,
+      "loss": 1.5719,
+      "step": 711
+    },
+    {
+      "epoch": 0.039687848383500555,
+      "grad_norm": 0.4018345773220062,
+      "learning_rate": 8e-05,
+      "loss": 1.6568,
+      "step": 712
+    },
+    {
+      "epoch": 0.03974358974358974,
+      "grad_norm": 0.39940693974494934,
+      "learning_rate": 8e-05,
+      "loss": 1.6682,
+      "step": 713
+    },
+    {
+      "epoch": 0.03979933110367893,
+      "grad_norm": 0.421329528093338,
+      "learning_rate": 8e-05,
+      "loss": 1.6526,
+      "step": 714
+    },
+    {
+      "epoch": 0.03985507246376811,
+      "grad_norm": 0.47900378704071045,
+      "learning_rate": 8e-05,
+      "loss": 2.1028,
+      "step": 715
+    },
+    {
+      "epoch": 0.039910813823857305,
+      "grad_norm": 0.4266323447227478,
+      "learning_rate": 8e-05,
+      "loss": 1.9051,
+      "step": 716
+    },
+    {
+      "epoch": 0.03996655518394649,
+      "grad_norm": 0.4186037480831146,
+      "learning_rate": 8e-05,
+      "loss": 1.6456,
+      "step": 717
+    },
+    {
+      "epoch": 0.040022296544035676,
+      "grad_norm": 0.459991991519928,
+      "learning_rate": 8e-05,
+      "loss": 1.7576,
+      "step": 718
+    },
+    {
+      "epoch": 0.04007803790412486,
+      "grad_norm": 0.5402688980102539,
+      "learning_rate": 8e-05,
+      "loss": 1.8946,
+      "step": 719
+    },
+    {
+      "epoch": 0.04013377926421405,
+      "grad_norm": 0.4141964316368103,
+      "learning_rate": 8e-05,
+      "loss": 1.7124,
+      "step": 720
+    },
+    {
+      "epoch": 0.040189520624303234,
+      "grad_norm": 0.4429500997066498,
+      "learning_rate": 8e-05,
+      "loss": 2.0645,
+      "step": 721
+    },
+    {
+      "epoch": 0.04024526198439242,
+      "grad_norm": 0.4050302803516388,
+      "learning_rate": 8e-05,
+      "loss": 1.7279,
+      "step": 722
+    },
+    {
+      "epoch": 0.040301003344481605,
+      "grad_norm": 0.42763057351112366,
+      "learning_rate": 8e-05,
+      "loss": 1.9281,
+      "step": 723
+    },
+    {
+      "epoch": 0.04035674470457079,
+      "grad_norm": 0.4697290360927582,
+      "learning_rate": 8e-05,
+      "loss": 1.9958,
+      "step": 724
+    },
+    {
+      "epoch": 0.040412486064659976,
+      "grad_norm": 0.3971403241157532,
+      "learning_rate": 8e-05,
+      "loss": 1.6868,
+      "step": 725
+    },
+    {
+      "epoch": 0.04046822742474916,
+      "grad_norm": 0.405602365732193,
+      "learning_rate": 8e-05,
+      "loss": 1.7869,
+      "step": 726
+    },
+    {
+      "epoch": 0.04052396878483835,
+      "grad_norm": 0.41436219215393066,
+      "learning_rate": 8e-05,
+      "loss": 1.8323,
+      "step": 727
+    },
+    {
+      "epoch": 0.04057971014492753,
+      "grad_norm": 0.41769540309906006,
+      "learning_rate": 8e-05,
+      "loss": 1.7949,
+      "step": 728
+    },
+    {
+      "epoch": 0.040635451505016726,
+      "grad_norm": 0.40197986364364624,
+      "learning_rate": 8e-05,
+      "loss": 1.8478,
+      "step": 729
+    },
+    {
+      "epoch": 0.04069119286510591,
+      "grad_norm": 0.3910923898220062,
+      "learning_rate": 8e-05,
+      "loss": 1.5842,
+      "step": 730
+    },
+    {
+      "epoch": 0.0407469342251951,
+      "grad_norm": 0.41100266575813293,
+      "learning_rate": 8e-05,
+      "loss": 1.8538,
+      "step": 731
+    },
+    {
+      "epoch": 0.04080267558528428,
+      "grad_norm": 0.4086783826351166,
+      "learning_rate": 8e-05,
+      "loss": 1.6857,
+      "step": 732
+    },
+    {
+      "epoch": 0.04085841694537347,
+      "grad_norm": 0.41415607929229736,
+      "learning_rate": 8e-05,
+      "loss": 1.7507,
+      "step": 733
+    },
+    {
+      "epoch": 0.040914158305462654,
+      "grad_norm": 0.3935626149177551,
+      "learning_rate": 8e-05,
+      "loss": 1.7366,
+      "step": 734
+    },
+    {
+      "epoch": 0.04096989966555184,
+      "grad_norm": 0.41125422716140747,
+      "learning_rate": 8e-05,
+      "loss": 1.9468,
+      "step": 735
+    },
+    {
+      "epoch": 0.041025641025641026,
+      "grad_norm": 0.4091278314590454,
+      "learning_rate": 8e-05,
+      "loss": 1.7788,
+      "step": 736
+    },
+    {
+      "epoch": 0.04108138238573021,
+      "grad_norm": 0.43207159638404846,
+      "learning_rate": 8e-05,
+      "loss": 2.0038,
+      "step": 737
+    },
+    {
+      "epoch": 0.0411371237458194,
+      "grad_norm": 0.43338799476623535,
+      "learning_rate": 8e-05,
+      "loss": 1.7264,
+      "step": 738
+    },
+    {
+      "epoch": 0.04119286510590858,
+      "grad_norm": 0.40661856532096863,
+      "learning_rate": 8e-05,
+      "loss": 1.7491,
+      "step": 739
+    },
+    {
+      "epoch": 0.04124860646599777,
+      "grad_norm": 0.3823534846305847,
+      "learning_rate": 8e-05,
+      "loss": 1.5044,
+      "step": 740
+    },
+    {
+      "epoch": 0.041304347826086954,
+      "grad_norm": 0.43768197298049927,
+      "learning_rate": 8e-05,
+      "loss": 1.8421,
+      "step": 741
+    },
+    {
+      "epoch": 0.04136008918617614,
+      "grad_norm": 0.43626633286476135,
+      "learning_rate": 8e-05,
+      "loss": 1.8624,
+      "step": 742
+    },
+    {
+      "epoch": 0.04141583054626533,
+      "grad_norm": 0.44751766324043274,
+      "learning_rate": 8e-05,
+      "loss": 1.8846,
+      "step": 743
+    },
+    {
+      "epoch": 0.04147157190635452,
+      "grad_norm": 0.3814850151538849,
+      "learning_rate": 8e-05,
+      "loss": 1.5948,
+      "step": 744
+    },
+    {
+      "epoch": 0.041527313266443704,
+      "grad_norm": 0.4221818447113037,
+      "learning_rate": 8e-05,
+      "loss": 1.6793,
+      "step": 745
+    },
+    {
+      "epoch": 0.04158305462653289,
+      "grad_norm": 0.4538663625717163,
+      "learning_rate": 8e-05,
+      "loss": 1.9088,
+      "step": 746
+    },
+    {
+      "epoch": 0.041638795986622075,
+      "grad_norm": 0.4213586449623108,
+      "learning_rate": 8e-05,
+      "loss": 1.7283,
+      "step": 747
+    },
+    {
+      "epoch": 0.04169453734671126,
+      "grad_norm": 0.458469033241272,
+      "learning_rate": 8e-05,
+      "loss": 1.7802,
+      "step": 748
+    },
+    {
+      "epoch": 0.041750278706800446,
+      "grad_norm": 0.43127307295799255,
+      "learning_rate": 8e-05,
+      "loss": 1.6724,
+      "step": 749
+    },
+    {
+      "epoch": 0.04180602006688963,
+      "grad_norm": 0.45649224519729614,
+      "learning_rate": 8e-05,
+      "loss": 1.9681,
+      "step": 750
+    },
+    {
+      "epoch": 0.04186176142697882,
+      "grad_norm": 0.4290968179702759,
+      "learning_rate": 8e-05,
+      "loss": 1.8333,
+      "step": 751
+    },
+    {
+      "epoch": 0.041917502787068,
+      "grad_norm": 0.4485679566860199,
+      "learning_rate": 8e-05,
+      "loss": 1.8084,
+      "step": 752
+    },
+    {
+      "epoch": 0.04197324414715719,
+      "grad_norm": 0.4158881902694702,
+      "learning_rate": 8e-05,
+      "loss": 1.7163,
+      "step": 753
+    },
+    {
+      "epoch": 0.042028985507246375,
+      "grad_norm": 0.4581759572029114,
+      "learning_rate": 8e-05,
+      "loss": 1.5889,
+      "step": 754
+    },
+    {
+      "epoch": 0.04208472686733556,
+      "grad_norm": 0.4187484085559845,
+      "learning_rate": 8e-05,
+      "loss": 1.8103,
+      "step": 755
+    },
+    {
+      "epoch": 0.042140468227424746,
+      "grad_norm": 0.41237813234329224,
+      "learning_rate": 8e-05,
+      "loss": 1.6807,
+      "step": 756
+    },
+    {
+      "epoch": 0.04219620958751394,
+      "grad_norm": 0.46197548508644104,
+      "learning_rate": 8e-05,
+      "loss": 1.9533,
+      "step": 757
+    },
+    {
+      "epoch": 0.042251950947603124,
+      "grad_norm": 0.44682711362838745,
+      "learning_rate": 8e-05,
+      "loss": 1.6977,
+      "step": 758
+    },
+    {
+      "epoch": 0.04230769230769231,
+      "grad_norm": 0.39638906717300415,
+      "learning_rate": 8e-05,
+      "loss": 1.5057,
+      "step": 759
+    },
+    {
+      "epoch": 0.042363433667781496,
+      "grad_norm": 0.45771071314811707,
+      "learning_rate": 8e-05,
+      "loss": 2.0477,
+      "step": 760
+    },
+    {
+      "epoch": 0.04241917502787068,
+      "grad_norm": 0.42211708426475525,
+      "learning_rate": 8e-05,
+      "loss": 1.8795,
+      "step": 761
+    },
+    {
+      "epoch": 0.04247491638795987,
+      "grad_norm": 0.4084051549434662,
+      "learning_rate": 8e-05,
+      "loss": 1.6568,
+      "step": 762
+    },
+    {
+      "epoch": 0.04253065774804905,
+      "grad_norm": 0.4379608631134033,
+      "learning_rate": 8e-05,
+      "loss": 1.7923,
+      "step": 763
+    },
+    {
+      "epoch": 0.04258639910813824,
+      "grad_norm": 0.40212878584861755,
+      "learning_rate": 8e-05,
+      "loss": 1.508,
+      "step": 764
+    },
+    {
+      "epoch": 0.042642140468227424,
+      "grad_norm": 0.40446737408638,
+      "learning_rate": 8e-05,
+      "loss": 1.8271,
+      "step": 765
+    },
+    {
+      "epoch": 0.04269788182831661,
+      "grad_norm": 0.41475871205329895,
+      "learning_rate": 8e-05,
+      "loss": 1.9338,
+      "step": 766
+    },
+    {
+      "epoch": 0.042753623188405795,
+      "grad_norm": 0.4280930757522583,
+      "learning_rate": 8e-05,
+      "loss": 1.9265,
+      "step": 767
+    },
+    {
+      "epoch": 0.04280936454849498,
+      "grad_norm": 0.3947325050830841,
+      "learning_rate": 8e-05,
+      "loss": 1.7675,
+      "step": 768
+    },
+    {
+      "epoch": 0.04286510590858417,
+      "grad_norm": 0.4449823796749115,
+      "learning_rate": 8e-05,
+      "loss": 1.6442,
+      "step": 769
+    },
+    {
+      "epoch": 0.04292084726867335,
+      "grad_norm": 0.40130770206451416,
+      "learning_rate": 8e-05,
+      "loss": 1.6473,
+      "step": 770
+    },
+    {
+      "epoch": 0.042976588628762545,
+      "grad_norm": 0.4342806041240692,
+      "learning_rate": 8e-05,
+      "loss": 1.852,
+      "step": 771
+    },
+    {
+      "epoch": 0.04303232998885173,
+      "grad_norm": 0.4064382016658783,
+      "learning_rate": 8e-05,
+      "loss": 1.8133,
+      "step": 772
+    },
+    {
+      "epoch": 0.043088071348940916,
+      "grad_norm": 0.3967300057411194,
+      "learning_rate": 8e-05,
+      "loss": 1.7411,
+      "step": 773
+    },
+    {
+      "epoch": 0.0431438127090301,
+      "grad_norm": 0.4405178427696228,
+      "learning_rate": 8e-05,
+      "loss": 1.9962,
+      "step": 774
+    },
+    {
+      "epoch": 0.04319955406911929,
+      "grad_norm": 0.45430299639701843,
+      "learning_rate": 8e-05,
+      "loss": 1.7633,
+      "step": 775
+    },
+    {
+      "epoch": 0.043255295429208473,
+      "grad_norm": 0.4152611494064331,
+      "learning_rate": 8e-05,
+      "loss": 1.4639,
+      "step": 776
+    },
+    {
+      "epoch": 0.04331103678929766,
+      "grad_norm": 0.4138352572917938,
+      "learning_rate": 8e-05,
+      "loss": 1.8169,
+      "step": 777
+    },
+    {
+      "epoch": 0.043366778149386845,
+      "grad_norm": 0.39472225308418274,
+      "learning_rate": 8e-05,
+      "loss": 1.5993,
+      "step": 778
+    },
+    {
+      "epoch": 0.04342251950947603,
+      "grad_norm": 0.4525393545627594,
+      "learning_rate": 8e-05,
+      "loss": 1.8683,
+      "step": 779
+    },
+    {
+      "epoch": 0.043478260869565216,
+      "grad_norm": 0.4362929165363312,
+      "learning_rate": 8e-05,
+      "loss": 1.8648,
+      "step": 780
+    },
+    {
+      "epoch": 0.0435340022296544,
+      "grad_norm": 0.40343475341796875,
+      "learning_rate": 8e-05,
+      "loss": 1.6745,
+      "step": 781
+    },
+    {
+      "epoch": 0.04358974358974359,
+      "grad_norm": 0.4135652184486389,
+      "learning_rate": 8e-05,
+      "loss": 1.753,
+      "step": 782
+    },
+    {
+      "epoch": 0.04364548494983277,
+      "grad_norm": 0.45906737446784973,
+      "learning_rate": 8e-05,
+      "loss": 1.9136,
+      "step": 783
+    },
+    {
+      "epoch": 0.04370122630992196,
+      "grad_norm": 0.42325344681739807,
+      "learning_rate": 8e-05,
+      "loss": 1.7591,
+      "step": 784
+    },
+    {
+      "epoch": 0.04375696767001115,
+      "grad_norm": 0.4207473397254944,
+      "learning_rate": 8e-05,
+      "loss": 1.7254,
+      "step": 785
+    },
+    {
+      "epoch": 0.04381270903010034,
+      "grad_norm": 0.41280874609947205,
+      "learning_rate": 8e-05,
+      "loss": 1.8155,
+      "step": 786
+    },
+    {
+      "epoch": 0.04386845039018952,
+      "grad_norm": 0.423045814037323,
+      "learning_rate": 8e-05,
+      "loss": 1.8083,
+      "step": 787
+    },
+    {
+      "epoch": 0.04392419175027871,
+      "grad_norm": 0.44817253947257996,
+      "learning_rate": 8e-05,
+      "loss": 1.9363,
+      "step": 788
+    },
+    {
+      "epoch": 0.043979933110367894,
+      "grad_norm": 0.43137675523757935,
+      "learning_rate": 8e-05,
+      "loss": 1.993,
+      "step": 789
+    },
+    {
+      "epoch": 0.04403567447045708,
+      "grad_norm": 0.4204290509223938,
+      "learning_rate": 8e-05,
+      "loss": 1.8462,
+      "step": 790
+    },
+    {
+      "epoch": 0.044091415830546266,
+      "grad_norm": 0.40246155858039856,
+      "learning_rate": 8e-05,
+      "loss": 1.5627,
+      "step": 791
+    },
+    {
+      "epoch": 0.04414715719063545,
+      "grad_norm": 0.5489731431007385,
+      "learning_rate": 8e-05,
+      "loss": 2.0595,
+      "step": 792
+    },
+    {
+      "epoch": 0.04420289855072464,
+      "grad_norm": 0.46178895235061646,
+      "learning_rate": 8e-05,
+      "loss": 1.952,
+      "step": 793
+    },
+    {
+      "epoch": 0.04425863991081382,
+      "grad_norm": 0.41195032000541687,
+      "learning_rate": 8e-05,
+      "loss": 1.6914,
+      "step": 794
+    },
+    {
+      "epoch": 0.04431438127090301,
+      "grad_norm": 0.4108409881591797,
+      "learning_rate": 8e-05,
+      "loss": 1.7034,
+      "step": 795
+    },
+    {
+      "epoch": 0.044370122630992194,
+      "grad_norm": 0.48466190695762634,
+      "learning_rate": 8e-05,
+      "loss": 2.1322,
+      "step": 796
+    },
+    {
+      "epoch": 0.04442586399108138,
+      "grad_norm": 0.4071280360221863,
+      "learning_rate": 8e-05,
+      "loss": 1.6725,
+      "step": 797
+    },
+    {
+      "epoch": 0.044481605351170565,
+      "grad_norm": 0.4744816720485687,
+      "learning_rate": 8e-05,
+      "loss": 2.0929,
+      "step": 798
+    },
+    {
+      "epoch": 0.04453734671125976,
+      "grad_norm": 0.42697998881340027,
+      "learning_rate": 8e-05,
+      "loss": 1.8765,
+      "step": 799
+    },
+    {
+      "epoch": 0.044593088071348944,
+      "grad_norm": 0.4529907703399658,
+      "learning_rate": 8e-05,
+      "loss": 1.8202,
+      "step": 800
+    },
+    {
+      "epoch": 0.04464882943143813,
+      "grad_norm": 0.4429434537887573,
+      "learning_rate": 8e-05,
+      "loss": 1.8094,
+      "step": 801
+    },
+    {
+      "epoch": 0.044704570791527315,
+      "grad_norm": 0.4543124735355377,
+      "learning_rate": 8e-05,
+      "loss": 1.7702,
+      "step": 802
+    },
+    {
+      "epoch": 0.0447603121516165,
+      "grad_norm": 0.4180499315261841,
+      "learning_rate": 8e-05,
+      "loss": 1.8529,
+      "step": 803
+    },
+    {
+      "epoch": 0.044816053511705686,
+      "grad_norm": 0.44464290142059326,
+      "learning_rate": 8e-05,
+      "loss": 1.8953,
+      "step": 804
+    },
+    {
+      "epoch": 0.04487179487179487,
+      "grad_norm": 0.399472177028656,
+      "learning_rate": 8e-05,
+      "loss": 1.643,
+      "step": 805
+    },
+    {
+      "epoch": 0.04492753623188406,
+      "grad_norm": 0.44110196828842163,
+      "learning_rate": 8e-05,
+      "loss": 2.1205,
+      "step": 806
+    },
+    {
+      "epoch": 0.04498327759197324,
+      "grad_norm": 0.42552313208580017,
+      "learning_rate": 8e-05,
+      "loss": 1.8727,
+      "step": 807
+    },
+    {
+      "epoch": 0.04503901895206243,
+      "grad_norm": 0.4377604126930237,
+      "learning_rate": 8e-05,
+      "loss": 1.8279,
+      "step": 808
+    },
+    {
+      "epoch": 0.045094760312151615,
+      "grad_norm": 0.3928232789039612,
+      "learning_rate": 8e-05,
+      "loss": 1.5861,
+      "step": 809
+    },
+    {
+      "epoch": 0.0451505016722408,
+      "grad_norm": 0.43836626410484314,
+      "learning_rate": 8e-05,
+      "loss": 1.9002,
+      "step": 810
+    },
+    {
+      "epoch": 0.045206243032329986,
+      "grad_norm": 0.423073410987854,
+      "learning_rate": 8e-05,
+      "loss": 1.6317,
+      "step": 811
+    },
+    {
+      "epoch": 0.04526198439241917,
+      "grad_norm": 0.46345505118370056,
+      "learning_rate": 8e-05,
+      "loss": 1.9068,
+      "step": 812
+    },
+    {
+      "epoch": 0.045317725752508364,
+      "grad_norm": 0.47184428572654724,
+      "learning_rate": 8e-05,
+      "loss": 1.9154,
+      "step": 813
+    },
+    {
+      "epoch": 0.04537346711259755,
+      "grad_norm": 0.43765196204185486,
+      "learning_rate": 8e-05,
+      "loss": 2.0261,
+      "step": 814
+    },
+    {
+      "epoch": 0.045429208472686736,
+      "grad_norm": 0.4310826063156128,
+      "learning_rate": 8e-05,
+      "loss": 1.8355,
+      "step": 815
+    },
+    {
+      "epoch": 0.04548494983277592,
+      "grad_norm": 0.47284650802612305,
+      "learning_rate": 8e-05,
+      "loss": 1.8096,
+      "step": 816
+    },
+    {
+      "epoch": 0.04554069119286511,
+      "grad_norm": 0.42523178458213806,
+      "learning_rate": 8e-05,
+      "loss": 1.7155,
+      "step": 817
+    },
+    {
+      "epoch": 0.04559643255295429,
+      "grad_norm": 0.3846169114112854,
+      "learning_rate": 8e-05,
+      "loss": 1.616,
+      "step": 818
+    },
+    {
+      "epoch": 0.04565217391304348,
+      "grad_norm": 0.42457297444343567,
+      "learning_rate": 8e-05,
+      "loss": 1.9693,
+      "step": 819
+    },
+    {
+      "epoch": 0.045707915273132664,
+      "grad_norm": 0.43201375007629395,
+      "learning_rate": 8e-05,
+      "loss": 1.7744,
+      "step": 820
+    },
+    {
+      "epoch": 0.04576365663322185,
+      "grad_norm": 0.39910250902175903,
+      "learning_rate": 8e-05,
+      "loss": 1.6146,
+      "step": 821
+    },
+    {
+      "epoch": 0.045819397993311035,
+      "grad_norm": 0.45629212260246277,
+      "learning_rate": 8e-05,
+      "loss": 1.8616,
+      "step": 822
+    },
+    {
+      "epoch": 0.04587513935340022,
+      "grad_norm": 0.4624870717525482,
+      "learning_rate": 8e-05,
+      "loss": 1.8472,
+      "step": 823
+    },
+    {
+      "epoch": 0.04593088071348941,
+      "grad_norm": 0.41313448548316956,
+      "learning_rate": 8e-05,
+      "loss": 1.5226,
+      "step": 824
+    },
+    {
+      "epoch": 0.04598662207357859,
+      "grad_norm": 0.3996032178401947,
+      "learning_rate": 8e-05,
+      "loss": 1.6795,
+      "step": 825
+    },
+    {
+      "epoch": 0.04604236343366778,
+      "grad_norm": 0.4157262146472931,
+      "learning_rate": 8e-05,
+      "loss": 1.7918,
+      "step": 826
+    },
+    {
+      "epoch": 0.04609810479375697,
+      "grad_norm": 0.46568799018859863,
+      "learning_rate": 8e-05,
+      "loss": 2.0148,
+      "step": 827
+    },
+    {
+      "epoch": 0.046153846153846156,
+      "grad_norm": 0.3986344337463379,
+      "learning_rate": 8e-05,
+      "loss": 1.7419,
+      "step": 828
+    },
+    {
+      "epoch": 0.04620958751393534,
+      "grad_norm": 0.38107046484947205,
+      "learning_rate": 8e-05,
+      "loss": 1.6143,
+      "step": 829
+    },
+    {
+      "epoch": 0.04626532887402453,
+      "grad_norm": 0.4358171820640564,
+      "learning_rate": 8e-05,
+      "loss": 1.7574,
+      "step": 830
+    },
+    {
+      "epoch": 0.04632107023411371,
+      "grad_norm": 0.38718435168266296,
+      "learning_rate": 8e-05,
+      "loss": 1.4672,
+      "step": 831
+    },
+    {
+      "epoch": 0.0463768115942029,
+      "grad_norm": 0.3949934244155884,
+      "learning_rate": 8e-05,
+      "loss": 1.7404,
+      "step": 832
+    },
+    {
+      "epoch": 0.046432552954292085,
+      "grad_norm": 0.39293748140335083,
+      "learning_rate": 8e-05,
+      "loss": 1.7409,
+      "step": 833
+    },
+    {
+      "epoch": 0.04648829431438127,
+      "grad_norm": 0.3997895121574402,
+      "learning_rate": 8e-05,
+      "loss": 1.68,
+      "step": 834
+    },
+    {
+      "epoch": 0.046544035674470456,
+      "grad_norm": 0.4256502091884613,
+      "learning_rate": 8e-05,
+      "loss": 1.8699,
+      "step": 835
+    },
+    {
+      "epoch": 0.04659977703455964,
+      "grad_norm": 0.4751846492290497,
+      "learning_rate": 8e-05,
+      "loss": 1.8223,
+      "step": 836
+    },
+    {
+      "epoch": 0.04665551839464883,
+      "grad_norm": 0.4109809398651123,
+      "learning_rate": 8e-05,
+      "loss": 1.6245,
+      "step": 837
+    },
+    {
+      "epoch": 0.04671125975473801,
+      "grad_norm": 0.41468480229377747,
+      "learning_rate": 8e-05,
+      "loss": 1.7512,
+      "step": 838
+    },
+    {
+      "epoch": 0.0467670011148272,
+      "grad_norm": 0.43772780895233154,
+      "learning_rate": 8e-05,
+      "loss": 1.8532,
+      "step": 839
+    },
+    {
+      "epoch": 0.046822742474916385,
+      "grad_norm": 0.4629352390766144,
+      "learning_rate": 8e-05,
+      "loss": 1.8597,
+      "step": 840
+    },
+    {
+      "epoch": 0.04687848383500558,
+      "grad_norm": 0.4625055193901062,
+      "learning_rate": 8e-05,
+      "loss": 1.9523,
+      "step": 841
+    },
+    {
+      "epoch": 0.04693422519509476,
+      "grad_norm": 0.42157262563705444,
+      "learning_rate": 8e-05,
+      "loss": 1.7857,
+      "step": 842
+    },
+    {
+      "epoch": 0.04698996655518395,
+      "grad_norm": 0.40957966446876526,
+      "learning_rate": 8e-05,
+      "loss": 1.636,
+      "step": 843
+    },
+    {
+      "epoch": 0.047045707915273134,
+      "grad_norm": 0.41058987379074097,
+      "learning_rate": 8e-05,
+      "loss": 1.8647,
+      "step": 844
+    },
+    {
+      "epoch": 0.04710144927536232,
+      "grad_norm": 0.4620298147201538,
+      "learning_rate": 8e-05,
+      "loss": 1.7832,
+      "step": 845
+    },
+    {
+      "epoch": 0.047157190635451506,
+      "grad_norm": 0.4322960674762726,
+      "learning_rate": 8e-05,
+      "loss": 1.7886,
+      "step": 846
+    },
+    {
+      "epoch": 0.04721293199554069,
+      "grad_norm": 0.45721161365509033,
+      "learning_rate": 8e-05,
+      "loss": 2.1069,
+      "step": 847
+    },
+    {
+      "epoch": 0.04726867335562988,
+      "grad_norm": 0.44434359669685364,
+      "learning_rate": 8e-05,
+      "loss": 1.881,
+      "step": 848
+    },
+    {
+      "epoch": 0.04732441471571906,
+      "grad_norm": 0.47906211018562317,
+      "learning_rate": 8e-05,
+      "loss": 1.9895,
+      "step": 849
+    },
+    {
+      "epoch": 0.04738015607580825,
+      "grad_norm": 0.4150347411632538,
+      "learning_rate": 8e-05,
+      "loss": 1.8052,
+      "step": 850
+    },
+    {
+      "epoch": 0.047435897435897434,
+      "grad_norm": 0.4207112491130829,
+      "learning_rate": 8e-05,
+      "loss": 1.8485,
+      "step": 851
+    },
+    {
+      "epoch": 0.04749163879598662,
+      "grad_norm": 0.6500937342643738,
+      "learning_rate": 8e-05,
+      "loss": 1.6529,
+      "step": 852
+    },
+    {
+      "epoch": 0.047547380156075805,
+      "grad_norm": 0.43128669261932373,
+      "learning_rate": 8e-05,
+      "loss": 1.9591,
+      "step": 853
+    },
+    {
+      "epoch": 0.047603121516165,
+      "grad_norm": 0.43956103920936584,
+      "learning_rate": 8e-05,
+      "loss": 1.7868,
+      "step": 854
+    },
+    {
+      "epoch": 0.047658862876254184,
+      "grad_norm": 0.41977569460868835,
+      "learning_rate": 8e-05,
+      "loss": 1.8002,
+      "step": 855
+    },
+    {
+      "epoch": 0.04771460423634337,
+      "grad_norm": 0.4412843883037567,
+      "learning_rate": 8e-05,
+      "loss": 1.7019,
+      "step": 856
+    },
+    {
+      "epoch": 0.047770345596432555,
+      "grad_norm": 0.4043019115924835,
+      "learning_rate": 8e-05,
+      "loss": 1.6508,
+      "step": 857
+    },
+    {
+      "epoch": 0.04782608695652174,
+      "grad_norm": 0.3964078724384308,
+      "learning_rate": 8e-05,
+      "loss": 1.7001,
+      "step": 858
+    },
+    {
+      "epoch": 0.047881828316610926,
+      "grad_norm": 0.41778528690338135,
+      "learning_rate": 8e-05,
+      "loss": 1.6894,
+      "step": 859
+    },
+    {
+      "epoch": 0.04793756967670011,
+      "grad_norm": 0.40355780720710754,
+      "learning_rate": 8e-05,
+      "loss": 1.5954,
+      "step": 860
+    },
+    {
+      "epoch": 0.0479933110367893,
+      "grad_norm": 0.4723891317844391,
+      "learning_rate": 8e-05,
+      "loss": 1.9092,
+      "step": 861
+    },
+    {
+      "epoch": 0.04804905239687848,
+      "grad_norm": 0.42808517813682556,
+      "learning_rate": 8e-05,
+      "loss": 1.8493,
+      "step": 862
+    },
+    {
+      "epoch": 0.04810479375696767,
+      "grad_norm": 0.4569825232028961,
+      "learning_rate": 8e-05,
+      "loss": 1.7595,
+      "step": 863
+    },
+    {
+      "epoch": 0.048160535117056855,
+      "grad_norm": 0.40674179792404175,
+      "learning_rate": 8e-05,
+      "loss": 1.5641,
+      "step": 864
+    },
+    {
+      "epoch": 0.04821627647714604,
+      "grad_norm": 0.41264376044273376,
+      "learning_rate": 8e-05,
+      "loss": 1.6743,
+      "step": 865
+    },
+    {
+      "epoch": 0.048272017837235226,
+      "grad_norm": 0.4213387966156006,
+      "learning_rate": 8e-05,
+      "loss": 1.5963,
+      "step": 866
+    },
+    {
+      "epoch": 0.04832775919732441,
+      "grad_norm": 0.4398549199104309,
+      "learning_rate": 8e-05,
+      "loss": 1.901,
+      "step": 867
+    },
+    {
+      "epoch": 0.048383500557413604,
+      "grad_norm": 0.4367522597312927,
+      "learning_rate": 8e-05,
+      "loss": 1.8161,
+      "step": 868
+    },
+    {
+      "epoch": 0.04843924191750279,
+      "grad_norm": 0.46793094277381897,
+      "learning_rate": 8e-05,
+      "loss": 2.0695,
+      "step": 869
+    },
+    {
+      "epoch": 0.048494983277591976,
+      "grad_norm": 0.4361442029476166,
+      "learning_rate": 8e-05,
+      "loss": 1.8895,
+      "step": 870
+    },
+    {
+      "epoch": 0.04855072463768116,
+      "grad_norm": 0.3976099193096161,
+      "learning_rate": 8e-05,
+      "loss": 1.7172,
+      "step": 871
+    },
+    {
+      "epoch": 0.04860646599777035,
+      "grad_norm": 0.4153869152069092,
+      "learning_rate": 8e-05,
+      "loss": 1.7037,
+      "step": 872
+    },
+    {
+      "epoch": 0.04866220735785953,
+      "grad_norm": 0.4478858709335327,
+      "learning_rate": 8e-05,
+      "loss": 1.7821,
+      "step": 873
+    },
+    {
+      "epoch": 0.04871794871794872,
+      "grad_norm": 0.44808489084243774,
+      "learning_rate": 8e-05,
+      "loss": 1.8322,
+      "step": 874
+    },
+    {
+      "epoch": 0.048773690078037904,
+      "grad_norm": 0.44651350378990173,
+      "learning_rate": 8e-05,
+      "loss": 1.8766,
+      "step": 875
+    },
+    {
+      "epoch": 0.04882943143812709,
+      "grad_norm": 0.4439350962638855,
+      "learning_rate": 8e-05,
+      "loss": 1.7531,
+      "step": 876
+    },
+    {
+      "epoch": 0.048885172798216275,
+      "grad_norm": 0.4228324890136719,
+      "learning_rate": 8e-05,
+      "loss": 1.7428,
+      "step": 877
+    },
+    {
+      "epoch": 0.04894091415830546,
+      "grad_norm": 0.42650389671325684,
+      "learning_rate": 8e-05,
+      "loss": 1.8161,
+      "step": 878
+    },
+    {
+      "epoch": 0.04899665551839465,
+      "grad_norm": 0.38016974925994873,
+      "learning_rate": 8e-05,
+      "loss": 1.6092,
+      "step": 879
+    },
+    {
+      "epoch": 0.04905239687848383,
+      "grad_norm": 0.4425662159919739,
+      "learning_rate": 8e-05,
+      "loss": 1.6695,
+      "step": 880
+    },
+    {
+      "epoch": 0.04910813823857302,
+      "grad_norm": 0.4162566363811493,
+      "learning_rate": 8e-05,
+      "loss": 1.7926,
+      "step": 881
+    },
+    {
+      "epoch": 0.04916387959866221,
+      "grad_norm": 0.4386153221130371,
+      "learning_rate": 8e-05,
+      "loss": 1.7272,
+      "step": 882
+    },
+    {
+      "epoch": 0.049219620958751396,
+      "grad_norm": 0.41146156191825867,
+      "learning_rate": 8e-05,
+      "loss": 1.7901,
+      "step": 883
+    },
+    {
+      "epoch": 0.04927536231884058,
+      "grad_norm": 0.38946112990379333,
+      "learning_rate": 8e-05,
+      "loss": 1.7411,
+      "step": 884
+    },
+    {
+      "epoch": 0.04933110367892977,
+      "grad_norm": 0.4616527557373047,
+      "learning_rate": 8e-05,
+      "loss": 2.0916,
+      "step": 885
+    },
+    {
+      "epoch": 0.04938684503901895,
+      "grad_norm": 0.4302521049976349,
+      "learning_rate": 8e-05,
+      "loss": 1.7878,
+      "step": 886
+    },
+    {
+      "epoch": 0.04944258639910814,
+      "grad_norm": 0.4233269691467285,
+      "learning_rate": 8e-05,
+      "loss": 1.6362,
+      "step": 887
+    },
+    {
+      "epoch": 0.049498327759197325,
+      "grad_norm": 0.4491744935512543,
+      "learning_rate": 8e-05,
+      "loss": 1.8209,
+      "step": 888
+    },
+    {
+      "epoch": 0.04955406911928651,
+      "grad_norm": 0.4061885178089142,
+      "learning_rate": 8e-05,
+      "loss": 1.7068,
+      "step": 889
+    },
+    {
+      "epoch": 0.049609810479375696,
+      "grad_norm": 0.42203137278556824,
+      "learning_rate": 8e-05,
+      "loss": 1.6899,
+      "step": 890
+    },
+    {
+      "epoch": 0.04966555183946488,
+      "grad_norm": 0.4250293970108032,
+      "learning_rate": 8e-05,
+      "loss": 1.7895,
+      "step": 891
+    },
+    {
+      "epoch": 0.04972129319955407,
+      "grad_norm": 0.43000224232673645,
+      "learning_rate": 8e-05,
+      "loss": 1.7183,
+      "step": 892
+    },
+    {
+      "epoch": 0.04977703455964325,
+      "grad_norm": 0.49140119552612305,
+      "learning_rate": 8e-05,
+      "loss": 2.0823,
+      "step": 893
+    },
+    {
+      "epoch": 0.04983277591973244,
+      "grad_norm": 0.4494474530220032,
+      "learning_rate": 8e-05,
+      "loss": 1.854,
+      "step": 894
+    },
+    {
+      "epoch": 0.049888517279821624,
+      "grad_norm": 0.45056265592575073,
+      "learning_rate": 8e-05,
+      "loss": 1.754,
+      "step": 895
+    },
+    {
+      "epoch": 0.04994425863991082,
+      "grad_norm": 0.45480015873908997,
+      "learning_rate": 8e-05,
+      "loss": 1.7431,
+      "step": 896
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4143247902393341,
+      "learning_rate": 8e-05,
+      "loss": 1.6383,
+      "step": 897
+    },
+    {
+      "epoch": 0.05005574136008919,
+      "grad_norm": 0.4520198702812195,
+      "learning_rate": 8e-05,
+      "loss": 1.7012,
+      "step": 898
+    },
+    {
+      "epoch": 0.050111482720178374,
+      "grad_norm": 0.4883075952529907,
+      "learning_rate": 8e-05,
+      "loss": 1.926,
+      "step": 899
+    },
+    {
+      "epoch": 0.05016722408026756,
+      "grad_norm": 0.46189042925834656,
+      "learning_rate": 8e-05,
+      "loss": 1.6825,
+      "step": 900
+    },
+    {
+      "epoch": 0.050222965440356745,
+      "grad_norm": 0.40424585342407227,
+      "learning_rate": 8e-05,
+      "loss": 1.6505,
+      "step": 901
+    },
+    {
+      "epoch": 0.05027870680044593,
+      "grad_norm": 0.3945220112800598,
+      "learning_rate": 8e-05,
+      "loss": 1.5049,
+      "step": 902
+    },
+    {
+      "epoch": 0.05033444816053512,
+      "grad_norm": 0.4237712025642395,
+      "learning_rate": 8e-05,
+      "loss": 1.74,
+      "step": 903
+    },
+    {
+      "epoch": 0.0503901895206243,
+      "grad_norm": 0.434099018573761,
+      "learning_rate": 8e-05,
+      "loss": 1.8923,
+      "step": 904
+    },
+    {
+      "epoch": 0.05044593088071349,
+      "grad_norm": 0.4102526903152466,
+      "learning_rate": 8e-05,
+      "loss": 1.7801,
+      "step": 905
+    },
+    {
+      "epoch": 0.050501672240802674,
+      "grad_norm": 0.4492480158805847,
+      "learning_rate": 8e-05,
+      "loss": 1.8818,
+      "step": 906
+    },
+    {
+      "epoch": 0.05055741360089186,
+      "grad_norm": 0.42090365290641785,
+      "learning_rate": 8e-05,
+      "loss": 1.6556,
+      "step": 907
+    },
+    {
+      "epoch": 0.050613154960981045,
+      "grad_norm": 0.37998881936073303,
+      "learning_rate": 8e-05,
+      "loss": 1.6473,
+      "step": 908
+    },
+    {
+      "epoch": 0.05066889632107023,
+      "grad_norm": 0.44362199306488037,
+      "learning_rate": 8e-05,
+      "loss": 1.9331,
+      "step": 909
+    },
+    {
+      "epoch": 0.050724637681159424,
+      "grad_norm": 0.42187029123306274,
+      "learning_rate": 8e-05,
+      "loss": 1.7093,
+      "step": 910
+    },
+    {
+      "epoch": 0.05078037904124861,
+      "grad_norm": 0.4547136425971985,
+      "learning_rate": 8e-05,
+      "loss": 1.8013,
+      "step": 911
+    },
+    {
+      "epoch": 0.050836120401337795,
+      "grad_norm": 0.6749727129936218,
+      "learning_rate": 8e-05,
+      "loss": 1.6621,
+      "step": 912
+    },
+    {
+      "epoch": 0.05089186176142698,
+      "grad_norm": 0.39881759881973267,
+      "learning_rate": 8e-05,
+      "loss": 1.6822,
+      "step": 913
+    },
+    {
+      "epoch": 0.050947603121516166,
+      "grad_norm": 0.40810444951057434,
+      "learning_rate": 8e-05,
+      "loss": 1.5836,
+      "step": 914
+    },
+    {
+      "epoch": 0.05100334448160535,
+      "grad_norm": 0.4395993947982788,
+      "learning_rate": 8e-05,
+      "loss": 1.8861,
+      "step": 915
+    },
+    {
+      "epoch": 0.05105908584169454,
+      "grad_norm": 0.4386335015296936,
+      "learning_rate": 8e-05,
+      "loss": 1.824,
+      "step": 916
+    },
+    {
+      "epoch": 0.05111482720178372,
+      "grad_norm": 0.3997780978679657,
+      "learning_rate": 8e-05,
+      "loss": 1.733,
+      "step": 917
+    },
+    {
+      "epoch": 0.05117056856187291,
+      "grad_norm": 0.4373658299446106,
+      "learning_rate": 8e-05,
+      "loss": 1.8932,
+      "step": 918
+    },
+    {
+      "epoch": 0.051226309921962095,
+      "grad_norm": 0.42019960284233093,
+      "learning_rate": 8e-05,
+      "loss": 1.6854,
+      "step": 919
+    },
+    {
+      "epoch": 0.05128205128205128,
+      "grad_norm": 0.4224587380886078,
+      "learning_rate": 8e-05,
+      "loss": 1.6851,
+      "step": 920
+    },
+    {
+      "epoch": 0.051337792642140466,
+      "grad_norm": 0.4506530463695526,
+      "learning_rate": 8e-05,
+      "loss": 2.0913,
+      "step": 921
+    },
+    {
+      "epoch": 0.05139353400222965,
+      "grad_norm": 0.5099965333938599,
+      "learning_rate": 8e-05,
+      "loss": 2.0435,
+      "step": 922
+    },
+    {
+      "epoch": 0.05144927536231884,
+      "grad_norm": 0.4297906756401062,
+      "learning_rate": 8e-05,
+      "loss": 1.761,
+      "step": 923
+    },
+    {
+      "epoch": 0.05150501672240803,
+      "grad_norm": 0.44647520780563354,
+      "learning_rate": 8e-05,
+      "loss": 1.8595,
+      "step": 924
+    },
+    {
+      "epoch": 0.051560758082497216,
+      "grad_norm": 0.42764410376548767,
+      "learning_rate": 8e-05,
+      "loss": 1.9272,
+      "step": 925
+    },
+    {
+      "epoch": 0.0516164994425864,
+      "grad_norm": 0.42151692509651184,
+      "learning_rate": 8e-05,
+      "loss": 1.9704,
+      "step": 926
+    },
+    {
+      "epoch": 0.05167224080267559,
+      "grad_norm": 0.432170569896698,
+      "learning_rate": 8e-05,
+      "loss": 1.7059,
+      "step": 927
+    },
+    {
+      "epoch": 0.05172798216276477,
+      "grad_norm": 0.49925047159194946,
+      "learning_rate": 8e-05,
+      "loss": 1.8333,
+      "step": 928
+    },
+    {
+      "epoch": 0.05178372352285396,
+      "grad_norm": 0.4397178292274475,
+      "learning_rate": 8e-05,
+      "loss": 1.8248,
+      "step": 929
+    },
+    {
+      "epoch": 0.051839464882943144,
+      "grad_norm": 0.390194833278656,
+      "learning_rate": 8e-05,
+      "loss": 1.7337,
+      "step": 930
+    },
+    {
+      "epoch": 0.05189520624303233,
+      "grad_norm": 0.4256782531738281,
+      "learning_rate": 8e-05,
+      "loss": 1.6734,
+      "step": 931
+    },
+    {
+      "epoch": 0.051950947603121515,
+      "grad_norm": 0.3956703245639801,
+      "learning_rate": 8e-05,
+      "loss": 1.665,
+      "step": 932
+    },
+    {
+      "epoch": 0.0520066889632107,
+      "grad_norm": 0.4501950740814209,
+      "learning_rate": 8e-05,
+      "loss": 1.8167,
+      "step": 933
+    },
+    {
+      "epoch": 0.05206243032329989,
+      "grad_norm": 0.41400691866874695,
+      "learning_rate": 8e-05,
+      "loss": 1.5718,
+      "step": 934
+    },
+    {
+      "epoch": 0.05211817168338907,
+      "grad_norm": 0.4194335639476776,
+      "learning_rate": 8e-05,
+      "loss": 1.6497,
+      "step": 935
+    },
+    {
+      "epoch": 0.05217391304347826,
+      "grad_norm": 0.41398027539253235,
+      "learning_rate": 8e-05,
+      "loss": 1.5243,
+      "step": 936
+    },
+    {
+      "epoch": 0.052229654403567444,
+      "grad_norm": 0.42045271396636963,
+      "learning_rate": 8e-05,
+      "loss": 1.7342,
+      "step": 937
+    },
+    {
+      "epoch": 0.052285395763656636,
+      "grad_norm": 0.36227548122406006,
+      "learning_rate": 8e-05,
+      "loss": 1.3389,
+      "step": 938
+    },
+    {
+      "epoch": 0.05234113712374582,
+      "grad_norm": 0.4060024321079254,
+      "learning_rate": 8e-05,
+      "loss": 1.6522,
+      "step": 939
+    },
+    {
+      "epoch": 0.05239687848383501,
+      "grad_norm": 0.3965223431587219,
+      "learning_rate": 8e-05,
+      "loss": 1.7567,
+      "step": 940
+    },
+    {
+      "epoch": 0.05245261984392419,
+      "grad_norm": 0.4708893299102783,
+      "learning_rate": 8e-05,
+      "loss": 1.7974,
+      "step": 941
+    },
+    {
+      "epoch": 0.05250836120401338,
+      "grad_norm": 0.3749980628490448,
+      "learning_rate": 8e-05,
+      "loss": 1.595,
+      "step": 942
+    },
+    {
+      "epoch": 0.052564102564102565,
+      "grad_norm": 0.4108208119869232,
+      "learning_rate": 8e-05,
+      "loss": 1.6256,
+      "step": 943
+    },
+    {
+      "epoch": 0.05261984392419175,
+      "grad_norm": 0.4239984452724457,
+      "learning_rate": 8e-05,
+      "loss": 1.7197,
+      "step": 944
+    },
+    {
+      "epoch": 0.052675585284280936,
+      "grad_norm": 0.4631035029888153,
+      "learning_rate": 8e-05,
+      "loss": 1.8541,
+      "step": 945
+    },
+    {
+      "epoch": 0.05273132664437012,
+      "grad_norm": 0.4573301374912262,
+      "learning_rate": 8e-05,
+      "loss": 1.8671,
+      "step": 946
+    },
+    {
+      "epoch": 0.05278706800445931,
+      "grad_norm": 0.4194883108139038,
+      "learning_rate": 8e-05,
+      "loss": 1.7707,
+      "step": 947
+    },
+    {
+      "epoch": 0.05284280936454849,
+      "grad_norm": 0.412008672952652,
+      "learning_rate": 8e-05,
+      "loss": 1.5866,
+      "step": 948
+    },
+    {
+      "epoch": 0.05289855072463768,
+      "grad_norm": 0.4511415660381317,
+      "learning_rate": 8e-05,
+      "loss": 1.7598,
+      "step": 949
+    },
+    {
+      "epoch": 0.052954292084726864,
+      "grad_norm": 0.4199249744415283,
+      "learning_rate": 8e-05,
+      "loss": 1.6622,
+      "step": 950
+    },
+    {
+      "epoch": 0.05301003344481605,
+      "grad_norm": 0.4607292711734772,
+      "learning_rate": 8e-05,
+      "loss": 1.7429,
+      "step": 951
+    },
+    {
+      "epoch": 0.05306577480490524,
+      "grad_norm": 0.47078394889831543,
+      "learning_rate": 8e-05,
+      "loss": 1.997,
+      "step": 952
+    },
+    {
+      "epoch": 0.05312151616499443,
+      "grad_norm": 0.4439176023006439,
+      "learning_rate": 8e-05,
+      "loss": 1.7236,
+      "step": 953
+    },
+    {
+      "epoch": 0.053177257525083614,
+      "grad_norm": 0.4908515214920044,
+      "learning_rate": 8e-05,
+      "loss": 1.5529,
+      "step": 954
+    },
+    {
+      "epoch": 0.0532329988851728,
+      "grad_norm": 0.48976606130599976,
+      "learning_rate": 8e-05,
+      "loss": 1.9534,
+      "step": 955
+    },
+    {
+      "epoch": 0.053288740245261985,
+      "grad_norm": 0.4504796266555786,
+      "learning_rate": 8e-05,
+      "loss": 1.8611,
+      "step": 956
+    },
+    {
+      "epoch": 0.05334448160535117,
+      "grad_norm": 0.4429522156715393,
+      "learning_rate": 8e-05,
+      "loss": 1.8884,
+      "step": 957
+    },
+    {
+      "epoch": 0.05340022296544036,
+      "grad_norm": 0.4269380271434784,
+      "learning_rate": 8e-05,
+      "loss": 1.6737,
+      "step": 958
+    },
+    {
+      "epoch": 0.05345596432552954,
+      "grad_norm": 0.4351010322570801,
+      "learning_rate": 8e-05,
+      "loss": 1.7113,
+      "step": 959
+    },
+    {
+      "epoch": 0.05351170568561873,
+      "grad_norm": 0.4203762114048004,
+      "learning_rate": 8e-05,
+      "loss": 1.7045,
+      "step": 960
+    },
+    {
+      "epoch": 0.053567447045707914,
+      "grad_norm": 0.4742576479911804,
+      "learning_rate": 8e-05,
+      "loss": 1.7861,
+      "step": 961
+    },
+    {
+      "epoch": 0.0536231884057971,
+      "grad_norm": 0.40854305028915405,
+      "learning_rate": 8e-05,
+      "loss": 1.7741,
+      "step": 962
+    },
+    {
+      "epoch": 0.053678929765886285,
+      "grad_norm": 0.4558371901512146,
+      "learning_rate": 8e-05,
+      "loss": 1.9133,
+      "step": 963
+    },
+    {
+      "epoch": 0.05373467112597547,
+      "grad_norm": 0.44087305665016174,
+      "learning_rate": 8e-05,
+      "loss": 1.6143,
+      "step": 964
+    },
+    {
+      "epoch": 0.053790412486064657,
+      "grad_norm": 0.4590452313423157,
+      "learning_rate": 8e-05,
+      "loss": 1.8756,
+      "step": 965
+    },
+    {
+      "epoch": 0.05384615384615385,
+      "grad_norm": 0.43718910217285156,
+      "learning_rate": 8e-05,
+      "loss": 1.8377,
+      "step": 966
+    },
+    {
+      "epoch": 0.053901895206243035,
+      "grad_norm": 0.39862263202667236,
+      "learning_rate": 8e-05,
+      "loss": 1.6154,
+      "step": 967
+    },
+    {
+      "epoch": 0.05395763656633222,
+      "grad_norm": 0.4717220366001129,
+      "learning_rate": 8e-05,
+      "loss": 1.8788,
+      "step": 968
+    },
+    {
+      "epoch": 0.054013377926421406,
+      "grad_norm": 0.43801164627075195,
+      "learning_rate": 8e-05,
+      "loss": 1.8031,
+      "step": 969
+    },
+    {
+      "epoch": 0.05406911928651059,
+      "grad_norm": 0.43170639872550964,
+      "learning_rate": 8e-05,
+      "loss": 1.8745,
+      "step": 970
+    },
+    {
+      "epoch": 0.05412486064659978,
+      "grad_norm": 0.4330897033214569,
+      "learning_rate": 8e-05,
+      "loss": 1.8022,
+      "step": 971
+    },
+    {
+      "epoch": 0.05418060200668896,
+      "grad_norm": 0.43624183535575867,
+      "learning_rate": 8e-05,
+      "loss": 1.874,
+      "step": 972
+    },
+    {
+      "epoch": 0.05423634336677815,
+      "grad_norm": 0.4237411916255951,
+      "learning_rate": 8e-05,
+      "loss": 1.7464,
+      "step": 973
+    },
+    {
+      "epoch": 0.054292084726867335,
+      "grad_norm": 0.4099571406841278,
+      "learning_rate": 8e-05,
+      "loss": 1.5987,
+      "step": 974
+    },
+    {
+      "epoch": 0.05434782608695652,
+      "grad_norm": 0.44440948963165283,
+      "learning_rate": 8e-05,
+      "loss": 1.7165,
+      "step": 975
+    },
+    {
+      "epoch": 0.054403567447045706,
+      "grad_norm": 0.43227434158325195,
+      "learning_rate": 8e-05,
+      "loss": 1.9345,
+      "step": 976
+    },
+    {
+      "epoch": 0.05445930880713489,
+      "grad_norm": 0.43842023611068726,
+      "learning_rate": 8e-05,
+      "loss": 1.6778,
+      "step": 977
+    },
+    {
+      "epoch": 0.05451505016722408,
+      "grad_norm": 0.4868525564670563,
+      "learning_rate": 8e-05,
+      "loss": 1.8353,
+      "step": 978
+    },
+    {
+      "epoch": 0.05457079152731327,
+      "grad_norm": 0.43103519082069397,
+      "learning_rate": 8e-05,
+      "loss": 1.6692,
+      "step": 979
+    },
+    {
+      "epoch": 0.054626532887402456,
+      "grad_norm": 0.465599924325943,
+      "learning_rate": 8e-05,
+      "loss": 2.117,
+      "step": 980
+    },
+    {
+      "epoch": 0.05468227424749164,
+      "grad_norm": 0.46371522545814514,
+      "learning_rate": 8e-05,
+      "loss": 1.8647,
+      "step": 981
+    },
+    {
+      "epoch": 0.05473801560758083,
+      "grad_norm": 0.4412715435028076,
+      "learning_rate": 8e-05,
+      "loss": 1.7827,
+      "step": 982
+    },
+    {
+      "epoch": 0.05479375696767001,
+      "grad_norm": 0.5162610411643982,
+      "learning_rate": 8e-05,
+      "loss": 1.969,
+      "step": 983
+    },
+    {
+      "epoch": 0.0548494983277592,
+      "grad_norm": 0.4213210940361023,
+      "learning_rate": 8e-05,
+      "loss": 1.6282,
+      "step": 984
+    },
+    {
+      "epoch": 0.054905239687848384,
+      "grad_norm": 0.4084482491016388,
+      "learning_rate": 8e-05,
+      "loss": 1.6544,
+      "step": 985
+    },
+    {
+      "epoch": 0.05496098104793757,
+      "grad_norm": 0.4904555678367615,
+      "learning_rate": 8e-05,
+      "loss": 1.9587,
+      "step": 986
+    },
+    {
+      "epoch": 0.055016722408026755,
+      "grad_norm": 0.4115452170372009,
+      "learning_rate": 8e-05,
+      "loss": 1.8093,
+      "step": 987
+    },
+    {
+      "epoch": 0.05507246376811594,
+      "grad_norm": 0.4041500389575958,
+      "learning_rate": 8e-05,
+      "loss": 1.7468,
+      "step": 988
+    },
+    {
+      "epoch": 0.05512820512820513,
+      "grad_norm": 0.42626655101776123,
+      "learning_rate": 8e-05,
+      "loss": 1.8658,
+      "step": 989
+    },
+    {
+      "epoch": 0.05518394648829431,
+      "grad_norm": 0.4490128457546234,
+      "learning_rate": 8e-05,
+      "loss": 1.9277,
+      "step": 990
+    },
+    {
+      "epoch": 0.0552396878483835,
+      "grad_norm": 0.4793890118598938,
+      "learning_rate": 8e-05,
+      "loss": 2.067,
+      "step": 991
+    },
+    {
+      "epoch": 0.055295429208472684,
+      "grad_norm": 0.4292072653770447,
+      "learning_rate": 8e-05,
+      "loss": 1.5705,
+      "step": 992
+    },
+    {
+      "epoch": 0.055351170568561876,
+      "grad_norm": 0.4330323040485382,
+      "learning_rate": 8e-05,
+      "loss": 1.9699,
+      "step": 993
+    },
+    {
+      "epoch": 0.05540691192865106,
+      "grad_norm": 0.5033552646636963,
+      "learning_rate": 8e-05,
+      "loss": 1.7123,
+      "step": 994
+    },
+    {
+      "epoch": 0.05546265328874025,
+      "grad_norm": 0.4264853298664093,
+      "learning_rate": 8e-05,
+      "loss": 1.9333,
+      "step": 995
+    },
+    {
+      "epoch": 0.05551839464882943,
+      "grad_norm": 0.43361344933509827,
+      "learning_rate": 8e-05,
+      "loss": 1.7706,
+      "step": 996
+    },
+    {
+      "epoch": 0.05557413600891862,
+      "grad_norm": 0.4839682877063751,
+      "learning_rate": 8e-05,
+      "loss": 1.9086,
+      "step": 997
+    },
+    {
+      "epoch": 0.055629877369007805,
+      "grad_norm": 0.46842366456985474,
+      "learning_rate": 8e-05,
+      "loss": 2.0759,
+      "step": 998
+    },
+    {
+      "epoch": 0.05568561872909699,
+      "grad_norm": 0.4213140308856964,
+      "learning_rate": 8e-05,
+      "loss": 1.7904,
+      "step": 999
+    },
+    {
+      "epoch": 0.055741360089186176,
+      "grad_norm": 0.4456956684589386,
+      "learning_rate": 8e-05,
+      "loss": 1.8378,
+      "step": 1000
+    },
+    {
+      "epoch": 0.05579710144927536,
+      "grad_norm": 0.4476751685142517,
+      "learning_rate": 8e-05,
+      "loss": 1.7923,
+      "step": 1001
+    },
+    {
+      "epoch": 0.05585284280936455,
+      "grad_norm": 0.4022519588470459,
+      "learning_rate": 8e-05,
+      "loss": 1.6922,
+      "step": 1002
+    },
+    {
+      "epoch": 0.05590858416945373,
+      "grad_norm": 0.4169318377971649,
+      "learning_rate": 8e-05,
+      "loss": 1.867,
+      "step": 1003
+    },
+    {
+      "epoch": 0.05596432552954292,
+      "grad_norm": 0.5523567795753479,
+      "learning_rate": 8e-05,
+      "loss": 2.3807,
+      "step": 1004
+    },
+    {
+      "epoch": 0.056020066889632104,
+      "grad_norm": 0.40831258893013,
+      "learning_rate": 8e-05,
+      "loss": 1.617,
+      "step": 1005
+    },
+    {
+      "epoch": 0.05607580824972129,
+      "grad_norm": 0.4216151535511017,
+      "learning_rate": 8e-05,
+      "loss": 1.7542,
+      "step": 1006
+    },
+    {
+      "epoch": 0.05613154960981048,
+      "grad_norm": 0.4263494610786438,
+      "learning_rate": 8e-05,
+      "loss": 1.6239,
+      "step": 1007
+    },
+    {
+      "epoch": 0.05618729096989967,
+      "grad_norm": 0.4082014858722687,
+      "learning_rate": 8e-05,
+      "loss": 1.5427,
+      "step": 1008
+    },
+    {
+      "epoch": 0.056243032329988854,
+      "grad_norm": 0.4632360339164734,
+      "learning_rate": 8e-05,
+      "loss": 1.8907,
+      "step": 1009
+    },
+    {
+      "epoch": 0.05629877369007804,
+      "grad_norm": 0.4561009109020233,
+      "learning_rate": 8e-05,
+      "loss": 1.7918,
+      "step": 1010
+    },
+    {
+      "epoch": 0.056354515050167225,
+      "grad_norm": 0.4626276195049286,
+      "learning_rate": 8e-05,
+      "loss": 1.7108,
+      "step": 1011
+    },
+    {
+      "epoch": 0.05641025641025641,
+      "grad_norm": 0.4755576252937317,
+      "learning_rate": 8e-05,
+      "loss": 2.021,
+      "step": 1012
+    },
+    {
+      "epoch": 0.0564659977703456,
+      "grad_norm": 0.4339366555213928,
+      "learning_rate": 8e-05,
+      "loss": 1.8268,
+      "step": 1013
+    },
+    {
+      "epoch": 0.05652173913043478,
+      "grad_norm": 0.4644366502761841,
+      "learning_rate": 8e-05,
+      "loss": 1.8686,
+      "step": 1014
+    },
+    {
+      "epoch": 0.05657748049052397,
+      "grad_norm": 0.47668248414993286,
+      "learning_rate": 8e-05,
+      "loss": 2.1487,
+      "step": 1015
+    },
+    {
+      "epoch": 0.056633221850613154,
+      "grad_norm": 0.46674248576164246,
+      "learning_rate": 8e-05,
+      "loss": 2.0188,
+      "step": 1016
+    },
+    {
+      "epoch": 0.05668896321070234,
+      "grad_norm": 0.45375779271125793,
+      "learning_rate": 8e-05,
+      "loss": 1.8338,
+      "step": 1017
+    },
+    {
+      "epoch": 0.056744704570791525,
+      "grad_norm": 0.40222418308258057,
+      "learning_rate": 8e-05,
+      "loss": 1.7165,
+      "step": 1018
+    },
+    {
+      "epoch": 0.05680044593088071,
+      "grad_norm": 0.5088784694671631,
+      "learning_rate": 8e-05,
+      "loss": 2.2179,
+      "step": 1019
+    },
+    {
+      "epoch": 0.056856187290969896,
+      "grad_norm": 0.4156004786491394,
+      "learning_rate": 8e-05,
+      "loss": 1.7308,
+      "step": 1020
+    },
+    {
+      "epoch": 0.05691192865105909,
+      "grad_norm": 0.43296828866004944,
+      "learning_rate": 8e-05,
+      "loss": 1.9068,
+      "step": 1021
+    },
+    {
+      "epoch": 0.056967670011148275,
+      "grad_norm": 0.42278701066970825,
+      "learning_rate": 8e-05,
+      "loss": 1.8354,
+      "step": 1022
+    },
+    {
+      "epoch": 0.05702341137123746,
+      "grad_norm": 0.44968274235725403,
+      "learning_rate": 8e-05,
+      "loss": 1.8766,
+      "step": 1023
+    },
+    {
+      "epoch": 0.057079152731326646,
+      "grad_norm": 0.4208853244781494,
+      "learning_rate": 8e-05,
+      "loss": 1.7541,
+      "step": 1024
+    },
+    {
+      "epoch": 0.05713489409141583,
+      "grad_norm": 0.42695286870002747,
+      "learning_rate": 8e-05,
+      "loss": 1.852,
+      "step": 1025
+    },
+    {
+      "epoch": 0.05719063545150502,
+      "grad_norm": 0.4679296910762787,
+      "learning_rate": 8e-05,
+      "loss": 1.9165,
+      "step": 1026
+    },
+    {
+      "epoch": 0.0572463768115942,
+      "grad_norm": 0.4364148676395416,
+      "learning_rate": 8e-05,
+      "loss": 1.7477,
+      "step": 1027
+    },
+    {
+      "epoch": 0.05730211817168339,
+      "grad_norm": 0.41188475489616394,
+      "learning_rate": 8e-05,
+      "loss": 1.5547,
+      "step": 1028
+    },
+    {
+      "epoch": 0.057357859531772575,
+      "grad_norm": 0.3878803849220276,
+      "learning_rate": 8e-05,
+      "loss": 1.6174,
+      "step": 1029
+    },
+    {
+      "epoch": 0.05741360089186176,
+      "grad_norm": 0.4230212867259979,
+      "learning_rate": 8e-05,
+      "loss": 1.7003,
+      "step": 1030
+    },
+    {
+      "epoch": 0.057469342251950946,
+      "grad_norm": 0.5208142995834351,
+      "learning_rate": 8e-05,
+      "loss": 2.1672,
+      "step": 1031
+    },
+    {
+      "epoch": 0.05752508361204013,
+      "grad_norm": 0.43669408559799194,
+      "learning_rate": 8e-05,
+      "loss": 1.7353,
+      "step": 1032
+    },
+    {
+      "epoch": 0.05758082497212932,
+      "grad_norm": 0.4216177761554718,
+      "learning_rate": 8e-05,
+      "loss": 1.6684,
+      "step": 1033
+    },
+    {
+      "epoch": 0.0576365663322185,
+      "grad_norm": 0.42164376378059387,
+      "learning_rate": 8e-05,
+      "loss": 1.6215,
+      "step": 1034
+    },
+    {
+      "epoch": 0.057692307692307696,
+      "grad_norm": 0.4818020462989807,
+      "learning_rate": 8e-05,
+      "loss": 2.0377,
+      "step": 1035
+    },
+    {
+      "epoch": 0.05774804905239688,
+      "grad_norm": 0.462981641292572,
+      "learning_rate": 8e-05,
+      "loss": 1.8978,
+      "step": 1036
+    },
+    {
+      "epoch": 0.05780379041248607,
+      "grad_norm": 0.4435645043849945,
+      "learning_rate": 8e-05,
+      "loss": 1.6334,
+      "step": 1037
+    },
+    {
+      "epoch": 0.05785953177257525,
+      "grad_norm": 0.3870157301425934,
+      "learning_rate": 8e-05,
+      "loss": 1.6838,
+      "step": 1038
+    },
+    {
+      "epoch": 0.05791527313266444,
+      "grad_norm": 0.43893805146217346,
+      "learning_rate": 8e-05,
+      "loss": 1.7563,
+      "step": 1039
+    },
+    {
+      "epoch": 0.057971014492753624,
+      "grad_norm": 0.46928489208221436,
+      "learning_rate": 8e-05,
+      "loss": 1.9366,
+      "step": 1040
+    },
+    {
+      "epoch": 0.05802675585284281,
+      "grad_norm": 0.5174750685691833,
+      "learning_rate": 8e-05,
+      "loss": 1.9242,
+      "step": 1041
+    },
+    {
+      "epoch": 0.058082497212931995,
+      "grad_norm": 0.41075706481933594,
+      "learning_rate": 8e-05,
+      "loss": 1.5632,
+      "step": 1042
+    },
+    {
+      "epoch": 0.05813823857302118,
+      "grad_norm": 0.39559102058410645,
+      "learning_rate": 8e-05,
+      "loss": 1.6224,
+      "step": 1043
+    },
+    {
+      "epoch": 0.05819397993311037,
+      "grad_norm": 0.43666520714759827,
+      "learning_rate": 8e-05,
+      "loss": 1.737,
+      "step": 1044
+    },
+    {
+      "epoch": 0.05824972129319955,
+      "grad_norm": 0.44067472219467163,
+      "learning_rate": 8e-05,
+      "loss": 1.915,
+      "step": 1045
+    },
+    {
+      "epoch": 0.05830546265328874,
+      "grad_norm": 0.5051589012145996,
+      "learning_rate": 8e-05,
+      "loss": 1.9358,
+      "step": 1046
+    },
+    {
+      "epoch": 0.058361204013377924,
+      "grad_norm": 0.4161045253276825,
+      "learning_rate": 8e-05,
+      "loss": 1.707,
+      "step": 1047
+    },
+    {
+      "epoch": 0.05841694537346711,
+      "grad_norm": 0.44676727056503296,
+      "learning_rate": 8e-05,
+      "loss": 1.8439,
+      "step": 1048
+    },
+    {
+      "epoch": 0.0584726867335563,
+      "grad_norm": 0.4336719512939453,
+      "learning_rate": 8e-05,
+      "loss": 1.8815,
+      "step": 1049
+    },
+    {
+      "epoch": 0.05852842809364549,
+      "grad_norm": 0.4447490870952606,
+      "learning_rate": 8e-05,
+      "loss": 1.6256,
+      "step": 1050
+    },
+    {
+      "epoch": 0.05858416945373467,
+      "grad_norm": 0.4244912564754486,
+      "learning_rate": 8e-05,
+      "loss": 1.7549,
+      "step": 1051
+    },
+    {
+      "epoch": 0.05863991081382386,
+      "grad_norm": 0.42827799916267395,
+      "learning_rate": 8e-05,
+      "loss": 1.9004,
+      "step": 1052
+    },
+    {
+      "epoch": 0.058695652173913045,
+      "grad_norm": 0.4283365309238434,
+      "learning_rate": 8e-05,
+      "loss": 1.8683,
+      "step": 1053
+    },
+    {
+      "epoch": 0.05875139353400223,
+      "grad_norm": 0.4388998746871948,
+      "learning_rate": 8e-05,
+      "loss": 1.9283,
+      "step": 1054
+    },
+    {
+      "epoch": 0.058807134894091416,
+      "grad_norm": 0.4022662937641144,
+      "learning_rate": 8e-05,
+      "loss": 1.4931,
+      "step": 1055
+    },
+    {
+      "epoch": 0.0588628762541806,
+      "grad_norm": 0.4228089153766632,
+      "learning_rate": 8e-05,
+      "loss": 1.7865,
+      "step": 1056
+    },
+    {
+      "epoch": 0.05891861761426979,
+      "grad_norm": 0.4698995053768158,
+      "learning_rate": 8e-05,
+      "loss": 1.8428,
+      "step": 1057
+    },
+    {
+      "epoch": 0.05897435897435897,
+      "grad_norm": 0.40131834149360657,
+      "learning_rate": 8e-05,
+      "loss": 1.5503,
+      "step": 1058
+    },
+    {
+      "epoch": 0.05903010033444816,
+      "grad_norm": 0.46206822991371155,
+      "learning_rate": 8e-05,
+      "loss": 1.8623,
+      "step": 1059
+    },
+    {
+      "epoch": 0.059085841694537344,
+      "grad_norm": 0.4545944929122925,
+      "learning_rate": 8e-05,
+      "loss": 1.688,
+      "step": 1060
+    },
+    {
+      "epoch": 0.05914158305462653,
+      "grad_norm": 0.42100989818573,
+      "learning_rate": 8e-05,
+      "loss": 1.6139,
+      "step": 1061
+    },
+    {
+      "epoch": 0.059197324414715716,
+      "grad_norm": 0.46316903829574585,
+      "learning_rate": 8e-05,
+      "loss": 1.7782,
+      "step": 1062
+    },
+    {
+      "epoch": 0.05925306577480491,
+      "grad_norm": 0.4210160970687866,
+      "learning_rate": 8e-05,
+      "loss": 1.6899,
+      "step": 1063
+    },
+    {
+      "epoch": 0.059308807134894094,
+      "grad_norm": 0.4561643898487091,
+      "learning_rate": 8e-05,
+      "loss": 1.9143,
+      "step": 1064
+    },
+    {
+      "epoch": 0.05936454849498328,
+      "grad_norm": 0.4355778098106384,
+      "learning_rate": 8e-05,
+      "loss": 1.7637,
+      "step": 1065
+    },
+    {
+      "epoch": 0.059420289855072465,
+      "grad_norm": 0.4599842429161072,
+      "learning_rate": 8e-05,
+      "loss": 1.8062,
+      "step": 1066
+    },
+    {
+      "epoch": 0.05947603121516165,
+      "grad_norm": 0.4494960904121399,
+      "learning_rate": 8e-05,
+      "loss": 1.6854,
+      "step": 1067
+    },
+    {
+      "epoch": 0.05953177257525084,
+      "grad_norm": 0.4336021840572357,
+      "learning_rate": 8e-05,
+      "loss": 1.8961,
+      "step": 1068
+    },
+    {
+      "epoch": 0.05958751393534002,
+      "grad_norm": 0.42015063762664795,
+      "learning_rate": 8e-05,
+      "loss": 1.7362,
+      "step": 1069
+    },
+    {
+      "epoch": 0.05964325529542921,
+      "grad_norm": 0.44604378938674927,
+      "learning_rate": 8e-05,
+      "loss": 1.9194,
+      "step": 1070
+    },
+    {
+      "epoch": 0.059698996655518394,
+      "grad_norm": 0.4459189474582672,
+      "learning_rate": 8e-05,
+      "loss": 1.8687,
+      "step": 1071
+    },
+    {
+      "epoch": 0.05975473801560758,
+      "grad_norm": 0.4442739486694336,
+      "learning_rate": 8e-05,
+      "loss": 1.8116,
+      "step": 1072
+    },
+    {
+      "epoch": 0.059810479375696765,
+      "grad_norm": 0.45290660858154297,
+      "learning_rate": 8e-05,
+      "loss": 1.8709,
+      "step": 1073
+    },
+    {
+      "epoch": 0.05986622073578595,
+      "grad_norm": 0.41268160939216614,
+      "learning_rate": 8e-05,
+      "loss": 1.6687,
+      "step": 1074
+    },
+    {
+      "epoch": 0.059921962095875136,
+      "grad_norm": 0.47134310007095337,
+      "learning_rate": 8e-05,
+      "loss": 2.1694,
+      "step": 1075
+    },
+    {
+      "epoch": 0.05997770345596432,
+      "grad_norm": 0.45093834400177,
+      "learning_rate": 8e-05,
+      "loss": 1.8108,
+      "step": 1076
+    },
+    {
+      "epoch": 0.060033444816053515,
+      "grad_norm": 0.40651044249534607,
+      "learning_rate": 8e-05,
+      "loss": 1.5763,
+      "step": 1077
+    },
+    {
+      "epoch": 0.0600891861761427,
+      "grad_norm": 0.42038488388061523,
+      "learning_rate": 8e-05,
+      "loss": 1.5817,
+      "step": 1078
+    },
+    {
+      "epoch": 0.060144927536231886,
+      "grad_norm": 0.4239666163921356,
+      "learning_rate": 8e-05,
+      "loss": 1.7996,
+      "step": 1079
+    },
+    {
+      "epoch": 0.06020066889632107,
+      "grad_norm": 0.43619927763938904,
+      "learning_rate": 8e-05,
+      "loss": 1.9681,
+      "step": 1080
+    },
+    {
+      "epoch": 0.06025641025641026,
+      "grad_norm": 0.41996511816978455,
+      "learning_rate": 8e-05,
+      "loss": 1.8411,
+      "step": 1081
+    },
+    {
+      "epoch": 0.06031215161649944,
+      "grad_norm": 0.4244367182254791,
+      "learning_rate": 8e-05,
+      "loss": 1.7147,
+      "step": 1082
+    },
+    {
+      "epoch": 0.06036789297658863,
+      "grad_norm": 0.44703248143196106,
+      "learning_rate": 8e-05,
+      "loss": 1.8158,
+      "step": 1083
+    },
+    {
+      "epoch": 0.060423634336677814,
+      "grad_norm": 0.45021986961364746,
+      "learning_rate": 8e-05,
+      "loss": 1.9023,
+      "step": 1084
+    },
+    {
+      "epoch": 0.060479375696767,
+      "grad_norm": 0.4354102909564972,
+      "learning_rate": 8e-05,
+      "loss": 1.9024,
+      "step": 1085
+    },
+    {
+      "epoch": 0.060535117056856186,
+      "grad_norm": 0.4334500730037689,
+      "learning_rate": 8e-05,
+      "loss": 1.8172,
+      "step": 1086
+    },
+    {
+      "epoch": 0.06059085841694537,
+      "grad_norm": 0.4097336530685425,
+      "learning_rate": 8e-05,
+      "loss": 1.6711,
+      "step": 1087
+    },
+    {
+      "epoch": 0.06064659977703456,
+      "grad_norm": 0.4830532968044281,
+      "learning_rate": 8e-05,
+      "loss": 1.824,
+      "step": 1088
+    },
+    {
+      "epoch": 0.06070234113712374,
+      "grad_norm": 0.447437047958374,
+      "learning_rate": 8e-05,
+      "loss": 1.9229,
+      "step": 1089
+    },
+    {
+      "epoch": 0.06075808249721293,
+      "grad_norm": 0.4807756543159485,
+      "learning_rate": 8e-05,
+      "loss": 1.8219,
+      "step": 1090
+    },
+    {
+      "epoch": 0.06081382385730212,
+      "grad_norm": 0.4496588408946991,
+      "learning_rate": 8e-05,
+      "loss": 2.0536,
+      "step": 1091
+    },
+    {
+      "epoch": 0.06086956521739131,
+      "grad_norm": 0.41599252820014954,
+      "learning_rate": 8e-05,
+      "loss": 1.6985,
+      "step": 1092
+    },
+    {
+      "epoch": 0.06092530657748049,
+      "grad_norm": 0.4328576922416687,
+      "learning_rate": 8e-05,
+      "loss": 1.8231,
+      "step": 1093
+    },
+    {
+      "epoch": 0.06098104793756968,
+      "grad_norm": 0.4406525790691376,
+      "learning_rate": 8e-05,
+      "loss": 1.8673,
+      "step": 1094
+    },
+    {
+      "epoch": 0.061036789297658864,
+      "grad_norm": 0.401132732629776,
+      "learning_rate": 8e-05,
+      "loss": 1.5546,
+      "step": 1095
+    },
+    {
+      "epoch": 0.06109253065774805,
+      "grad_norm": 0.42298394441604614,
+      "learning_rate": 8e-05,
+      "loss": 1.568,
+      "step": 1096
+    },
+    {
+      "epoch": 0.061148272017837235,
+      "grad_norm": 0.4252376854419708,
+      "learning_rate": 8e-05,
+      "loss": 1.752,
+      "step": 1097
+    },
+    {
+      "epoch": 0.06120401337792642,
+      "grad_norm": 0.4755648374557495,
+      "learning_rate": 8e-05,
+      "loss": 1.832,
+      "step": 1098
+    },
+    {
+      "epoch": 0.06125975473801561,
+      "grad_norm": 0.5341011881828308,
+      "learning_rate": 8e-05,
+      "loss": 2.1064,
+      "step": 1099
+    },
+    {
+      "epoch": 0.06131549609810479,
+      "grad_norm": 0.43493762612342834,
+      "learning_rate": 8e-05,
+      "loss": 1.8192,
+      "step": 1100
+    },
+    {
+      "epoch": 0.06137123745819398,
+      "grad_norm": 0.4518592059612274,
+      "learning_rate": 8e-05,
+      "loss": 1.7659,
+      "step": 1101
+    },
+    {
+      "epoch": 0.061426978818283164,
+      "grad_norm": 0.4443103075027466,
+      "learning_rate": 8e-05,
+      "loss": 1.8146,
+      "step": 1102
+    },
+    {
+      "epoch": 0.06148272017837235,
+      "grad_norm": 0.4577937424182892,
+      "learning_rate": 8e-05,
+      "loss": 1.7822,
+      "step": 1103
+    },
+    {
+      "epoch": 0.06153846153846154,
+      "grad_norm": 0.48581668734550476,
+      "learning_rate": 8e-05,
+      "loss": 1.9258,
+      "step": 1104
+    },
+    {
+      "epoch": 0.06159420289855073,
+      "grad_norm": 0.4341087341308594,
+      "learning_rate": 8e-05,
+      "loss": 1.8588,
+      "step": 1105
+    },
+    {
+      "epoch": 0.06164994425863991,
+      "grad_norm": 0.43398237228393555,
+      "learning_rate": 8e-05,
+      "loss": 1.778,
+      "step": 1106
+    },
+    {
+      "epoch": 0.0617056856187291,
+      "grad_norm": 0.40834975242614746,
+      "learning_rate": 8e-05,
+      "loss": 1.7681,
+      "step": 1107
+    },
+    {
+      "epoch": 0.061761426978818285,
+      "grad_norm": 0.41185036301612854,
+      "learning_rate": 8e-05,
+      "loss": 1.638,
+      "step": 1108
+    },
+    {
+      "epoch": 0.06181716833890747,
+      "grad_norm": 0.4246727526187897,
+      "learning_rate": 8e-05,
+      "loss": 1.7239,
+      "step": 1109
+    },
+    {
+      "epoch": 0.061872909698996656,
+      "grad_norm": 0.4513862431049347,
+      "learning_rate": 8e-05,
+      "loss": 1.8033,
+      "step": 1110
+    },
+    {
+      "epoch": 0.06192865105908584,
+      "grad_norm": 0.4458586573600769,
+      "learning_rate": 8e-05,
+      "loss": 1.9101,
+      "step": 1111
+    },
+    {
+      "epoch": 0.06198439241917503,
+      "grad_norm": 0.46575620770454407,
+      "learning_rate": 8e-05,
+      "loss": 1.8896,
+      "step": 1112
+    },
+    {
+      "epoch": 0.06204013377926421,
+      "grad_norm": 0.4389342963695526,
+      "learning_rate": 8e-05,
+      "loss": 1.7504,
+      "step": 1113
+    },
+    {
+      "epoch": 0.0620958751393534,
+      "grad_norm": 0.49950799345970154,
+      "learning_rate": 8e-05,
+      "loss": 1.9905,
+      "step": 1114
+    },
+    {
+      "epoch": 0.062151616499442584,
+      "grad_norm": 0.4360363781452179,
+      "learning_rate": 8e-05,
+      "loss": 1.2877,
+      "step": 1115
+    },
+    {
+      "epoch": 0.06220735785953177,
+      "grad_norm": 0.4367810785770416,
+      "learning_rate": 8e-05,
+      "loss": 1.7995,
+      "step": 1116
+    },
+    {
+      "epoch": 0.062263099219620956,
+      "grad_norm": 0.45174649357795715,
+      "learning_rate": 8e-05,
+      "loss": 2.0174,
+      "step": 1117
+    },
+    {
+      "epoch": 0.06231884057971015,
+      "grad_norm": 0.4332228899002075,
+      "learning_rate": 8e-05,
+      "loss": 1.7798,
+      "step": 1118
+    },
+    {
+      "epoch": 0.062374581939799334,
+      "grad_norm": 0.42212432622909546,
+      "learning_rate": 8e-05,
+      "loss": 1.6598,
+      "step": 1119
+    },
+    {
+      "epoch": 0.06243032329988852,
+      "grad_norm": 0.3963571786880493,
+      "learning_rate": 8e-05,
+      "loss": 1.5384,
+      "step": 1120
+    },
+    {
+      "epoch": 0.062486064659977705,
+      "grad_norm": 0.42810705304145813,
+      "learning_rate": 8e-05,
+      "loss": 1.7214,
+      "step": 1121
+    },
+    {
+      "epoch": 0.06254180602006688,
+      "grad_norm": 0.4167286455631256,
+      "learning_rate": 8e-05,
+      "loss": 1.7947,
+      "step": 1122
+    },
+    {
+      "epoch": 0.06259754738015608,
+      "grad_norm": 0.43644922971725464,
+      "learning_rate": 8e-05,
+      "loss": 1.8355,
+      "step": 1123
+    },
+    {
+      "epoch": 0.06265328874024526,
+      "grad_norm": 0.4328288435935974,
+      "learning_rate": 8e-05,
+      "loss": 1.6814,
+      "step": 1124
+    },
+    {
+      "epoch": 0.06270903010033445,
+      "grad_norm": 0.4443534314632416,
+      "learning_rate": 8e-05,
+      "loss": 1.7836,
+      "step": 1125
+    },
+    {
+      "epoch": 0.06276477146042364,
+      "grad_norm": 0.4291650652885437,
+      "learning_rate": 8e-05,
+      "loss": 1.654,
+      "step": 1126
+    },
+    {
+      "epoch": 0.06282051282051282,
+      "grad_norm": 0.44335997104644775,
+      "learning_rate": 8e-05,
+      "loss": 1.9186,
+      "step": 1127
+    },
+    {
+      "epoch": 0.06287625418060201,
+      "grad_norm": 0.4430725872516632,
+      "learning_rate": 8e-05,
+      "loss": 1.7812,
+      "step": 1128
+    },
+    {
+      "epoch": 0.06293199554069119,
+      "grad_norm": 0.4567205607891083,
+      "learning_rate": 8e-05,
+      "loss": 1.9849,
+      "step": 1129
+    },
+    {
+      "epoch": 0.06298773690078038,
+      "grad_norm": 0.45621874928474426,
+      "learning_rate": 8e-05,
+      "loss": 1.9115,
+      "step": 1130
+    },
+    {
+      "epoch": 0.06304347826086956,
+      "grad_norm": 0.45525816082954407,
+      "learning_rate": 8e-05,
+      "loss": 1.8807,
+      "step": 1131
+    },
+    {
+      "epoch": 0.06309921962095875,
+      "grad_norm": 0.4256766736507416,
+      "learning_rate": 8e-05,
+      "loss": 1.7896,
+      "step": 1132
+    },
+    {
+      "epoch": 0.06315496098104793,
+      "grad_norm": 0.4518592059612274,
+      "learning_rate": 8e-05,
+      "loss": 1.8965,
+      "step": 1133
+    },
+    {
+      "epoch": 0.06321070234113713,
+      "grad_norm": 0.44870826601982117,
+      "learning_rate": 8e-05,
+      "loss": 1.7877,
+      "step": 1134
+    },
+    {
+      "epoch": 0.0632664437012263,
+      "grad_norm": 0.39587363600730896,
+      "learning_rate": 8e-05,
+      "loss": 1.498,
+      "step": 1135
+    },
+    {
+      "epoch": 0.0633221850613155,
+      "grad_norm": 0.42017364501953125,
+      "learning_rate": 8e-05,
+      "loss": 1.7699,
+      "step": 1136
+    },
+    {
+      "epoch": 0.06337792642140468,
+      "grad_norm": 0.430050790309906,
+      "learning_rate": 8e-05,
+      "loss": 1.7077,
+      "step": 1137
+    },
+    {
+      "epoch": 0.06343366778149387,
+      "grad_norm": 0.4952334463596344,
+      "learning_rate": 8e-05,
+      "loss": 1.5378,
+      "step": 1138
+    },
+    {
+      "epoch": 0.06348940914158306,
+      "grad_norm": 0.39611610770225525,
+      "learning_rate": 8e-05,
+      "loss": 1.3078,
+      "step": 1139
+    },
+    {
+      "epoch": 0.06354515050167224,
+      "grad_norm": 0.4411247968673706,
+      "learning_rate": 8e-05,
+      "loss": 1.7183,
+      "step": 1140
+    },
+    {
+      "epoch": 0.06360089186176143,
+      "grad_norm": 0.441955029964447,
+      "learning_rate": 8e-05,
+      "loss": 1.943,
+      "step": 1141
+    },
+    {
+      "epoch": 0.06365663322185061,
+      "grad_norm": 0.4392283856868744,
+      "learning_rate": 8e-05,
+      "loss": 1.6887,
+      "step": 1142
+    },
+    {
+      "epoch": 0.0637123745819398,
+      "grad_norm": 0.44392138719558716,
+      "learning_rate": 8e-05,
+      "loss": 1.7069,
+      "step": 1143
+    },
+    {
+      "epoch": 0.06376811594202898,
+      "grad_norm": 0.40321075916290283,
+      "learning_rate": 8e-05,
+      "loss": 1.8176,
+      "step": 1144
+    },
+    {
+      "epoch": 0.06382385730211818,
+      "grad_norm": 0.4580495059490204,
+      "learning_rate": 8e-05,
+      "loss": 1.9481,
+      "step": 1145
+    },
+    {
+      "epoch": 0.06387959866220735,
+      "grad_norm": 0.4451245069503784,
+      "learning_rate": 8e-05,
+      "loss": 1.7982,
+      "step": 1146
+    },
+    {
+      "epoch": 0.06393534002229655,
+      "grad_norm": 0.4405924081802368,
+      "learning_rate": 8e-05,
+      "loss": 1.863,
+      "step": 1147
+    },
+    {
+      "epoch": 0.06399108138238573,
+      "grad_norm": 0.44056808948516846,
+      "learning_rate": 8e-05,
+      "loss": 1.8536,
+      "step": 1148
+    },
+    {
+      "epoch": 0.06404682274247492,
+      "grad_norm": 0.44250497221946716,
+      "learning_rate": 8e-05,
+      "loss": 1.7549,
+      "step": 1149
+    },
+    {
+      "epoch": 0.0641025641025641,
+      "grad_norm": 0.42131105065345764,
+      "learning_rate": 8e-05,
+      "loss": 1.7281,
+      "step": 1150
+    },
+    {
+      "epoch": 0.06415830546265329,
+      "grad_norm": 0.4144631624221802,
+      "learning_rate": 8e-05,
+      "loss": 1.8841,
+      "step": 1151
+    },
+    {
+      "epoch": 0.06421404682274247,
+      "grad_norm": 0.4111699163913727,
+      "learning_rate": 8e-05,
+      "loss": 1.5023,
+      "step": 1152
+    },
+    {
+      "epoch": 0.06426978818283166,
+      "grad_norm": 0.4193449020385742,
+      "learning_rate": 8e-05,
+      "loss": 1.584,
+      "step": 1153
+    },
+    {
+      "epoch": 0.06432552954292085,
+      "grad_norm": 0.44446641206741333,
+      "learning_rate": 8e-05,
+      "loss": 1.8562,
+      "step": 1154
+    },
+    {
+      "epoch": 0.06438127090301003,
+      "grad_norm": 0.4735074043273926,
+      "learning_rate": 8e-05,
+      "loss": 1.6038,
+      "step": 1155
+    },
+    {
+      "epoch": 0.06443701226309922,
+      "grad_norm": 0.4791753888130188,
+      "learning_rate": 8e-05,
+      "loss": 2.0993,
+      "step": 1156
+    },
+    {
+      "epoch": 0.0644927536231884,
+      "grad_norm": 0.42762860655784607,
+      "learning_rate": 8e-05,
+      "loss": 1.8449,
+      "step": 1157
+    },
+    {
+      "epoch": 0.0645484949832776,
+      "grad_norm": 0.428559273481369,
+      "learning_rate": 8e-05,
+      "loss": 1.6434,
+      "step": 1158
+    },
+    {
+      "epoch": 0.06460423634336677,
+      "grad_norm": 0.41851234436035156,
+      "learning_rate": 8e-05,
+      "loss": 1.7801,
+      "step": 1159
+    },
+    {
+      "epoch": 0.06465997770345597,
+      "grad_norm": 0.4329962730407715,
+      "learning_rate": 8e-05,
+      "loss": 1.7616,
+      "step": 1160
+    },
+    {
+      "epoch": 0.06471571906354515,
+      "grad_norm": 0.42422378063201904,
+      "learning_rate": 8e-05,
+      "loss": 1.7998,
+      "step": 1161
+    },
+    {
+      "epoch": 0.06477146042363434,
+      "grad_norm": 0.4383358955383301,
+      "learning_rate": 8e-05,
+      "loss": 1.9044,
+      "step": 1162
+    },
+    {
+      "epoch": 0.06482720178372352,
+      "grad_norm": 0.4317667484283447,
+      "learning_rate": 8e-05,
+      "loss": 1.8342,
+      "step": 1163
+    },
+    {
+      "epoch": 0.06488294314381271,
+      "grad_norm": 0.4675438404083252,
+      "learning_rate": 8e-05,
+      "loss": 1.7965,
+      "step": 1164
+    },
+    {
+      "epoch": 0.06493868450390189,
+      "grad_norm": 0.43177488446235657,
+      "learning_rate": 8e-05,
+      "loss": 1.8793,
+      "step": 1165
+    },
+    {
+      "epoch": 0.06499442586399108,
+      "grad_norm": 0.3976529836654663,
+      "learning_rate": 8e-05,
+      "loss": 1.8037,
+      "step": 1166
+    },
+    {
+      "epoch": 0.06505016722408027,
+      "grad_norm": 0.49942541122436523,
+      "learning_rate": 8e-05,
+      "loss": 1.9051,
+      "step": 1167
+    },
+    {
+      "epoch": 0.06510590858416945,
+      "grad_norm": 0.46481525897979736,
+      "learning_rate": 8e-05,
+      "loss": 1.8427,
+      "step": 1168
+    },
+    {
+      "epoch": 0.06516164994425865,
+      "grad_norm": 0.44631510972976685,
+      "learning_rate": 8e-05,
+      "loss": 1.8011,
+      "step": 1169
+    },
+    {
+      "epoch": 0.06521739130434782,
+      "grad_norm": 0.4932478368282318,
+      "learning_rate": 8e-05,
+      "loss": 1.7985,
+      "step": 1170
+    },
+    {
+      "epoch": 0.06527313266443702,
+      "grad_norm": 0.4396071135997772,
+      "learning_rate": 8e-05,
+      "loss": 1.7359,
+      "step": 1171
+    },
+    {
+      "epoch": 0.0653288740245262,
+      "grad_norm": 0.45359596610069275,
+      "learning_rate": 8e-05,
+      "loss": 1.8659,
+      "step": 1172
+    },
+    {
+      "epoch": 0.06538461538461539,
+      "grad_norm": 0.40468963980674744,
+      "learning_rate": 8e-05,
+      "loss": 1.4637,
+      "step": 1173
+    },
+    {
+      "epoch": 0.06544035674470457,
+      "grad_norm": 0.4339541792869568,
+      "learning_rate": 8e-05,
+      "loss": 1.8278,
+      "step": 1174
+    },
+    {
+      "epoch": 0.06549609810479376,
+      "grad_norm": 0.3990840017795563,
+      "learning_rate": 8e-05,
+      "loss": 1.6032,
+      "step": 1175
+    },
+    {
+      "epoch": 0.06555183946488294,
+      "grad_norm": 0.4356841742992401,
+      "learning_rate": 8e-05,
+      "loss": 1.7532,
+      "step": 1176
+    },
+    {
+      "epoch": 0.06560758082497213,
+      "grad_norm": 0.4568274915218353,
+      "learning_rate": 8e-05,
+      "loss": 1.9382,
+      "step": 1177
+    },
+    {
+      "epoch": 0.06566332218506131,
+      "grad_norm": 0.43120601773262024,
+      "learning_rate": 8e-05,
+      "loss": 1.8981,
+      "step": 1178
+    },
+    {
+      "epoch": 0.0657190635451505,
+      "grad_norm": 0.4385588765144348,
+      "learning_rate": 8e-05,
+      "loss": 1.7199,
+      "step": 1179
+    },
+    {
+      "epoch": 0.06577480490523968,
+      "grad_norm": 0.4216311275959015,
+      "learning_rate": 8e-05,
+      "loss": 1.7709,
+      "step": 1180
+    },
+    {
+      "epoch": 0.06583054626532887,
+      "grad_norm": 0.433501273393631,
+      "learning_rate": 8e-05,
+      "loss": 1.5969,
+      "step": 1181
+    },
+    {
+      "epoch": 0.06588628762541807,
+      "grad_norm": 0.44115445017814636,
+      "learning_rate": 8e-05,
+      "loss": 1.8182,
+      "step": 1182
+    },
+    {
+      "epoch": 0.06594202898550725,
+      "grad_norm": 0.4317324161529541,
+      "learning_rate": 8e-05,
+      "loss": 1.7448,
+      "step": 1183
+    },
+    {
+      "epoch": 0.06599777034559644,
+      "grad_norm": 0.43778157234191895,
+      "learning_rate": 8e-05,
+      "loss": 1.759,
+      "step": 1184
+    },
+    {
+      "epoch": 0.06605351170568562,
+      "grad_norm": 0.4287802278995514,
+      "learning_rate": 8e-05,
+      "loss": 1.7302,
+      "step": 1185
+    },
+    {
+      "epoch": 0.06610925306577481,
+      "grad_norm": 0.4651936888694763,
+      "learning_rate": 8e-05,
+      "loss": 1.6471,
+      "step": 1186
+    },
+    {
+      "epoch": 0.06616499442586399,
+      "grad_norm": 0.4521205425262451,
+      "learning_rate": 8e-05,
+      "loss": 1.807,
+      "step": 1187
+    },
+    {
+      "epoch": 0.06622073578595318,
+      "grad_norm": 0.43945908546447754,
+      "learning_rate": 8e-05,
+      "loss": 1.7347,
+      "step": 1188
+    },
+    {
+      "epoch": 0.06627647714604236,
+      "grad_norm": 0.4200635850429535,
+      "learning_rate": 8e-05,
+      "loss": 1.6566,
+      "step": 1189
+    },
+    {
+      "epoch": 0.06633221850613155,
+      "grad_norm": 0.43169304728507996,
+      "learning_rate": 8e-05,
+      "loss": 1.7691,
+      "step": 1190
+    },
+    {
+      "epoch": 0.06638795986622073,
+      "grad_norm": 0.4562690556049347,
+      "learning_rate": 8e-05,
+      "loss": 1.8396,
+      "step": 1191
+    },
+    {
+      "epoch": 0.06644370122630992,
+      "grad_norm": 0.43797406554222107,
+      "learning_rate": 8e-05,
+      "loss": 1.6736,
+      "step": 1192
+    },
+    {
+      "epoch": 0.0664994425863991,
+      "grad_norm": 0.4583410322666168,
+      "learning_rate": 8e-05,
+      "loss": 2.0178,
+      "step": 1193
+    },
+    {
+      "epoch": 0.0665551839464883,
+      "grad_norm": 0.49902230501174927,
+      "learning_rate": 8e-05,
+      "loss": 1.8571,
+      "step": 1194
+    },
+    {
+      "epoch": 0.06661092530657749,
+      "grad_norm": 0.43419089913368225,
+      "learning_rate": 8e-05,
+      "loss": 1.6779,
+      "step": 1195
+    },
+    {
+      "epoch": 0.06666666666666667,
+      "grad_norm": 0.4792231023311615,
+      "learning_rate": 8e-05,
+      "loss": 1.8727,
+      "step": 1196
+    },
+    {
+      "epoch": 0.06672240802675586,
+      "grad_norm": 0.43299925327301025,
+      "learning_rate": 8e-05,
+      "loss": 1.7666,
+      "step": 1197
+    },
+    {
+      "epoch": 0.06677814938684504,
+      "grad_norm": 0.4165794253349304,
+      "learning_rate": 8e-05,
+      "loss": 1.7137,
+      "step": 1198
+    },
+    {
+      "epoch": 0.06683389074693423,
+      "grad_norm": 0.4535585939884186,
+      "learning_rate": 8e-05,
+      "loss": 1.7655,
+      "step": 1199
+    },
+    {
+      "epoch": 0.06688963210702341,
+      "grad_norm": 0.4509890377521515,
+      "learning_rate": 8e-05,
+      "loss": 1.8209,
+      "step": 1200
+    },
+    {
+      "epoch": 0.0669453734671126,
+      "grad_norm": 0.4380910098552704,
+      "learning_rate": 8e-05,
+      "loss": 1.7428,
+      "step": 1201
+    },
+    {
+      "epoch": 0.06700111482720178,
+      "grad_norm": 0.4500320553779602,
+      "learning_rate": 8e-05,
+      "loss": 1.8026,
+      "step": 1202
+    },
+    {
+      "epoch": 0.06705685618729097,
+      "grad_norm": 0.47676727175712585,
+      "learning_rate": 8e-05,
+      "loss": 1.8406,
+      "step": 1203
+    },
+    {
+      "epoch": 0.06711259754738015,
+      "grad_norm": 0.43615782260894775,
+      "learning_rate": 8e-05,
+      "loss": 1.7105,
+      "step": 1204
+    },
+    {
+      "epoch": 0.06716833890746934,
+      "grad_norm": 0.4056113660335541,
+      "learning_rate": 8e-05,
+      "loss": 1.7475,
+      "step": 1205
+    },
+    {
+      "epoch": 0.06722408026755852,
+      "grad_norm": 0.43659117817878723,
+      "learning_rate": 8e-05,
+      "loss": 1.7423,
+      "step": 1206
+    },
+    {
+      "epoch": 0.06727982162764772,
+      "grad_norm": 0.4657900631427765,
+      "learning_rate": 8e-05,
+      "loss": 1.7744,
+      "step": 1207
+    },
+    {
+      "epoch": 0.0673355629877369,
+      "grad_norm": 0.47949662804603577,
+      "learning_rate": 8e-05,
+      "loss": 1.9269,
+      "step": 1208
+    },
+    {
+      "epoch": 0.06739130434782609,
+      "grad_norm": 0.45301929116249084,
+      "learning_rate": 8e-05,
+      "loss": 1.7783,
+      "step": 1209
+    },
+    {
+      "epoch": 0.06744704570791528,
+      "grad_norm": 0.48128190636634827,
+      "learning_rate": 8e-05,
+      "loss": 1.8949,
+      "step": 1210
+    },
+    {
+      "epoch": 0.06750278706800446,
+      "grad_norm": 0.4121861159801483,
+      "learning_rate": 8e-05,
+      "loss": 1.6111,
+      "step": 1211
+    },
+    {
+      "epoch": 0.06755852842809365,
+      "grad_norm": 0.5082221627235413,
+      "learning_rate": 8e-05,
+      "loss": 1.8098,
+      "step": 1212
+    },
+    {
+      "epoch": 0.06761426978818283,
+      "grad_norm": 0.45201507210731506,
+      "learning_rate": 8e-05,
+      "loss": 1.7695,
+      "step": 1213
+    },
+    {
+      "epoch": 0.06767001114827202,
+      "grad_norm": 0.40513336658477783,
+      "learning_rate": 8e-05,
+      "loss": 1.6133,
+      "step": 1214
+    },
+    {
+      "epoch": 0.0677257525083612,
+      "grad_norm": 0.45347344875335693,
+      "learning_rate": 8e-05,
+      "loss": 1.9623,
+      "step": 1215
+    },
+    {
+      "epoch": 0.0677814938684504,
+      "grad_norm": 0.42242351174354553,
+      "learning_rate": 8e-05,
+      "loss": 1.7273,
+      "step": 1216
+    },
+    {
+      "epoch": 0.06783723522853957,
+      "grad_norm": 0.42420676350593567,
+      "learning_rate": 8e-05,
+      "loss": 1.6187,
+      "step": 1217
+    },
+    {
+      "epoch": 0.06789297658862876,
+      "grad_norm": 0.45483478903770447,
+      "learning_rate": 8e-05,
+      "loss": 1.844,
+      "step": 1218
+    },
+    {
+      "epoch": 0.06794871794871794,
+      "grad_norm": 0.440166175365448,
+      "learning_rate": 8e-05,
+      "loss": 1.7197,
+      "step": 1219
+    },
+    {
+      "epoch": 0.06800445930880714,
+      "grad_norm": 0.4046355187892914,
+      "learning_rate": 8e-05,
+      "loss": 1.6218,
+      "step": 1220
+    },
+    {
+      "epoch": 0.06806020066889631,
+      "grad_norm": 0.44956326484680176,
+      "learning_rate": 8e-05,
+      "loss": 1.6526,
+      "step": 1221
+    },
+    {
+      "epoch": 0.06811594202898551,
+      "grad_norm": 0.4798918664455414,
+      "learning_rate": 8e-05,
+      "loss": 1.977,
+      "step": 1222
+    },
+    {
+      "epoch": 0.0681716833890747,
+      "grad_norm": 0.45986703038215637,
+      "learning_rate": 8e-05,
+      "loss": 1.8512,
+      "step": 1223
+    },
+    {
+      "epoch": 0.06822742474916388,
+      "grad_norm": 0.4693866968154907,
+      "learning_rate": 8e-05,
+      "loss": 1.6778,
+      "step": 1224
+    },
+    {
+      "epoch": 0.06828316610925307,
+      "grad_norm": 0.42940130829811096,
+      "learning_rate": 8e-05,
+      "loss": 1.6264,
+      "step": 1225
+    },
+    {
+      "epoch": 0.06833890746934225,
+      "grad_norm": 0.4441980719566345,
+      "learning_rate": 8e-05,
+      "loss": 1.7485,
+      "step": 1226
+    },
+    {
+      "epoch": 0.06839464882943144,
+      "grad_norm": 0.45005375146865845,
+      "learning_rate": 8e-05,
+      "loss": 2.0098,
+      "step": 1227
+    },
+    {
+      "epoch": 0.06845039018952062,
+      "grad_norm": 0.4309133291244507,
+      "learning_rate": 8e-05,
+      "loss": 1.7733,
+      "step": 1228
+    },
+    {
+      "epoch": 0.06850613154960981,
+      "grad_norm": 0.43084168434143066,
+      "learning_rate": 8e-05,
+      "loss": 1.7545,
+      "step": 1229
+    },
+    {
+      "epoch": 0.06856187290969899,
+      "grad_norm": 0.45081162452697754,
+      "learning_rate": 8e-05,
+      "loss": 1.5478,
+      "step": 1230
+    },
+    {
+      "epoch": 0.06861761426978819,
+      "grad_norm": 0.5341338515281677,
+      "learning_rate": 8e-05,
+      "loss": 1.6129,
+      "step": 1231
+    },
+    {
+      "epoch": 0.06867335562987736,
+      "grad_norm": 0.4506458640098572,
+      "learning_rate": 8e-05,
+      "loss": 1.8137,
+      "step": 1232
+    },
+    {
+      "epoch": 0.06872909698996656,
+      "grad_norm": 0.47452664375305176,
+      "learning_rate": 8e-05,
+      "loss": 1.8111,
+      "step": 1233
+    },
+    {
+      "epoch": 0.06878483835005574,
+      "grad_norm": 0.46135878562927246,
+      "learning_rate": 8e-05,
+      "loss": 1.6382,
+      "step": 1234
+    },
+    {
+      "epoch": 0.06884057971014493,
+      "grad_norm": 0.4212229549884796,
+      "learning_rate": 8e-05,
+      "loss": 1.6441,
+      "step": 1235
+    },
+    {
+      "epoch": 0.06889632107023412,
+      "grad_norm": 0.4409344494342804,
+      "learning_rate": 8e-05,
+      "loss": 1.5812,
+      "step": 1236
+    },
+    {
+      "epoch": 0.0689520624303233,
+      "grad_norm": 0.45152005553245544,
+      "learning_rate": 8e-05,
+      "loss": 1.5605,
+      "step": 1237
+    },
+    {
+      "epoch": 0.06900780379041249,
+      "grad_norm": 0.45905745029449463,
+      "learning_rate": 8e-05,
+      "loss": 1.7057,
+      "step": 1238
+    },
+    {
+      "epoch": 0.06906354515050167,
+      "grad_norm": 0.45360052585601807,
+      "learning_rate": 8e-05,
+      "loss": 1.7103,
+      "step": 1239
+    },
+    {
+      "epoch": 0.06911928651059086,
+      "grad_norm": 0.43128153681755066,
+      "learning_rate": 8e-05,
+      "loss": 1.6429,
+      "step": 1240
+    },
+    {
+      "epoch": 0.06917502787068004,
+      "grad_norm": 0.5309453010559082,
+      "learning_rate": 8e-05,
+      "loss": 1.9669,
+      "step": 1241
+    },
+    {
+      "epoch": 0.06923076923076923,
+      "grad_norm": 0.4460867941379547,
+      "learning_rate": 8e-05,
+      "loss": 1.6892,
+      "step": 1242
+    },
+    {
+      "epoch": 0.06928651059085841,
+      "grad_norm": 0.44082680344581604,
+      "learning_rate": 8e-05,
+      "loss": 1.8816,
+      "step": 1243
+    },
+    {
+      "epoch": 0.0693422519509476,
+      "grad_norm": 0.43136364221572876,
+      "learning_rate": 8e-05,
+      "loss": 1.6994,
+      "step": 1244
+    },
+    {
+      "epoch": 0.06939799331103678,
+      "grad_norm": 0.40499478578567505,
+      "learning_rate": 8e-05,
+      "loss": 1.5882,
+      "step": 1245
+    },
+    {
+      "epoch": 0.06945373467112598,
+      "grad_norm": 0.4175102710723877,
+      "learning_rate": 8e-05,
+      "loss": 1.5226,
+      "step": 1246
+    },
+    {
+      "epoch": 0.06950947603121516,
+      "grad_norm": 0.4481530487537384,
+      "learning_rate": 8e-05,
+      "loss": 1.7025,
+      "step": 1247
+    },
+    {
+      "epoch": 0.06956521739130435,
+      "grad_norm": 0.42439576983451843,
+      "learning_rate": 8e-05,
+      "loss": 1.6737,
+      "step": 1248
+    },
+    {
+      "epoch": 0.06962095875139353,
+      "grad_norm": 0.43070897459983826,
+      "learning_rate": 8e-05,
+      "loss": 1.6205,
+      "step": 1249
+    },
+    {
+      "epoch": 0.06967670011148272,
+      "grad_norm": 0.4086393415927887,
+      "learning_rate": 8e-05,
+      "loss": 1.7679,
+      "step": 1250
+    },
+    {
+      "epoch": 0.06973244147157191,
+      "grad_norm": 0.4932767450809479,
+      "learning_rate": 8e-05,
+      "loss": 1.5989,
+      "step": 1251
+    },
+    {
+      "epoch": 0.06978818283166109,
+      "grad_norm": 0.5050959587097168,
+      "learning_rate": 8e-05,
+      "loss": 2.0871,
+      "step": 1252
+    },
+    {
+      "epoch": 0.06984392419175028,
+      "grad_norm": 0.4637925326824188,
+      "learning_rate": 8e-05,
+      "loss": 1.868,
+      "step": 1253
+    },
+    {
+      "epoch": 0.06989966555183946,
+      "grad_norm": 0.41985011100769043,
+      "learning_rate": 8e-05,
+      "loss": 1.5779,
+      "step": 1254
+    },
+    {
+      "epoch": 0.06995540691192866,
+      "grad_norm": 0.4753486216068268,
+      "learning_rate": 8e-05,
+      "loss": 1.9415,
+      "step": 1255
+    },
+    {
+      "epoch": 0.07001114827201783,
+      "grad_norm": 0.43326473236083984,
+      "learning_rate": 8e-05,
+      "loss": 1.7659,
+      "step": 1256
+    },
+    {
+      "epoch": 0.07006688963210703,
+      "grad_norm": 0.41782644391059875,
+      "learning_rate": 8e-05,
+      "loss": 1.6882,
+      "step": 1257
+    },
+    {
+      "epoch": 0.0701226309921962,
+      "grad_norm": 0.44368311762809753,
+      "learning_rate": 8e-05,
+      "loss": 1.541,
+      "step": 1258
+    },
+    {
+      "epoch": 0.0701783723522854,
+      "grad_norm": 0.4449438750743866,
+      "learning_rate": 8e-05,
+      "loss": 1.6659,
+      "step": 1259
+    },
+    {
+      "epoch": 0.07023411371237458,
+      "grad_norm": 0.4272336959838867,
+      "learning_rate": 8e-05,
+      "loss": 1.4967,
+      "step": 1260
+    },
+    {
+      "epoch": 0.07028985507246377,
+      "grad_norm": 0.4614323377609253,
+      "learning_rate": 8e-05,
+      "loss": 1.8735,
+      "step": 1261
+    },
+    {
+      "epoch": 0.07034559643255295,
+      "grad_norm": 0.42471393942832947,
+      "learning_rate": 8e-05,
+      "loss": 1.946,
+      "step": 1262
+    },
+    {
+      "epoch": 0.07040133779264214,
+      "grad_norm": 0.4473448097705841,
+      "learning_rate": 8e-05,
+      "loss": 1.8145,
+      "step": 1263
+    },
+    {
+      "epoch": 0.07045707915273133,
+      "grad_norm": 0.582961916923523,
+      "learning_rate": 8e-05,
+      "loss": 1.7832,
+      "step": 1264
+    },
+    {
+      "epoch": 0.07051282051282051,
+      "grad_norm": 0.4432711601257324,
+      "learning_rate": 8e-05,
+      "loss": 1.9099,
+      "step": 1265
+    },
+    {
+      "epoch": 0.0705685618729097,
+      "grad_norm": 0.4061930775642395,
+      "learning_rate": 8e-05,
+      "loss": 1.6089,
+      "step": 1266
+    },
+    {
+      "epoch": 0.07062430323299888,
+      "grad_norm": 0.43295302987098694,
+      "learning_rate": 8e-05,
+      "loss": 1.9276,
+      "step": 1267
+    },
+    {
+      "epoch": 0.07068004459308808,
+      "grad_norm": 0.4516142010688782,
+      "learning_rate": 8e-05,
+      "loss": 1.7385,
+      "step": 1268
+    },
+    {
+      "epoch": 0.07073578595317725,
+      "grad_norm": 0.5671238899230957,
+      "learning_rate": 8e-05,
+      "loss": 1.9751,
+      "step": 1269
+    },
+    {
+      "epoch": 0.07079152731326645,
+      "grad_norm": 0.4582408666610718,
+      "learning_rate": 8e-05,
+      "loss": 1.9635,
+      "step": 1270
+    },
+    {
+      "epoch": 0.07084726867335563,
+      "grad_norm": 0.45998480916023254,
+      "learning_rate": 8e-05,
+      "loss": 1.7577,
+      "step": 1271
+    },
+    {
+      "epoch": 0.07090301003344482,
+      "grad_norm": 0.44361039996147156,
+      "learning_rate": 8e-05,
+      "loss": 1.7662,
+      "step": 1272
+    },
+    {
+      "epoch": 0.070958751393534,
+      "grad_norm": 0.45488402247428894,
+      "learning_rate": 8e-05,
+      "loss": 1.8307,
+      "step": 1273
+    },
+    {
+      "epoch": 0.07101449275362319,
+      "grad_norm": 0.4565194845199585,
+      "learning_rate": 8e-05,
+      "loss": 1.6827,
+      "step": 1274
+    },
+    {
+      "epoch": 0.07107023411371237,
+      "grad_norm": 0.44688448309898376,
+      "learning_rate": 8e-05,
+      "loss": 1.6676,
+      "step": 1275
+    },
+    {
+      "epoch": 0.07112597547380156,
+      "grad_norm": 0.4568803012371063,
+      "learning_rate": 8e-05,
+      "loss": 1.7335,
+      "step": 1276
+    },
+    {
+      "epoch": 0.07118171683389074,
+      "grad_norm": 0.43606460094451904,
+      "learning_rate": 8e-05,
+      "loss": 1.8572,
+      "step": 1277
+    },
+    {
+      "epoch": 0.07123745819397993,
+      "grad_norm": 0.4564160406589508,
+      "learning_rate": 8e-05,
+      "loss": 1.8938,
+      "step": 1278
+    },
+    {
+      "epoch": 0.07129319955406913,
+      "grad_norm": 0.4709673821926117,
+      "learning_rate": 8e-05,
+      "loss": 2.0716,
+      "step": 1279
+    },
+    {
+      "epoch": 0.0713489409141583,
+      "grad_norm": 0.4627903401851654,
+      "learning_rate": 8e-05,
+      "loss": 1.9952,
+      "step": 1280
+    },
+    {
+      "epoch": 0.0714046822742475,
+      "grad_norm": 0.45227548480033875,
+      "learning_rate": 8e-05,
+      "loss": 1.7264,
+      "step": 1281
+    },
+    {
+      "epoch": 0.07146042363433668,
+      "grad_norm": 0.4532218277454376,
+      "learning_rate": 8e-05,
+      "loss": 1.7318,
+      "step": 1282
+    },
+    {
+      "epoch": 0.07151616499442587,
+      "grad_norm": 0.43891414999961853,
+      "learning_rate": 8e-05,
+      "loss": 1.7138,
+      "step": 1283
+    },
+    {
+      "epoch": 0.07157190635451505,
+      "grad_norm": 0.41745471954345703,
+      "learning_rate": 8e-05,
+      "loss": 1.6738,
+      "step": 1284
+    },
+    {
+      "epoch": 0.07162764771460424,
+      "grad_norm": 0.43092063069343567,
+      "learning_rate": 8e-05,
+      "loss": 1.7968,
+      "step": 1285
+    },
+    {
+      "epoch": 0.07168338907469342,
+      "grad_norm": 0.3872428834438324,
+      "learning_rate": 8e-05,
+      "loss": 1.4228,
+      "step": 1286
+    },
+    {
+      "epoch": 0.07173913043478261,
+      "grad_norm": 0.4488355815410614,
+      "learning_rate": 8e-05,
+      "loss": 1.8599,
+      "step": 1287
+    },
+    {
+      "epoch": 0.07179487179487179,
+      "grad_norm": 0.4347779154777527,
+      "learning_rate": 8e-05,
+      "loss": 1.7125,
+      "step": 1288
+    },
+    {
+      "epoch": 0.07185061315496098,
+      "grad_norm": 0.46295249462127686,
+      "learning_rate": 8e-05,
+      "loss": 1.9316,
+      "step": 1289
+    },
+    {
+      "epoch": 0.07190635451505016,
+      "grad_norm": 0.47468453645706177,
+      "learning_rate": 8e-05,
+      "loss": 2.0974,
+      "step": 1290
+    },
+    {
+      "epoch": 0.07196209587513935,
+      "grad_norm": 0.44178271293640137,
+      "learning_rate": 8e-05,
+      "loss": 1.7739,
+      "step": 1291
+    },
+    {
+      "epoch": 0.07201783723522855,
+      "grad_norm": 0.4681906998157501,
+      "learning_rate": 8e-05,
+      "loss": 1.6988,
+      "step": 1292
+    },
+    {
+      "epoch": 0.07207357859531772,
+      "grad_norm": 0.43681320548057556,
+      "learning_rate": 8e-05,
+      "loss": 1.7113,
+      "step": 1293
+    },
+    {
+      "epoch": 0.07212931995540692,
+      "grad_norm": 0.45803403854370117,
+      "learning_rate": 8e-05,
+      "loss": 1.8703,
+      "step": 1294
+    },
+    {
+      "epoch": 0.0721850613154961,
+      "grad_norm": 0.48975130915641785,
+      "learning_rate": 8e-05,
+      "loss": 1.5916,
+      "step": 1295
+    },
+    {
+      "epoch": 0.07224080267558529,
+      "grad_norm": 0.4666009545326233,
+      "learning_rate": 8e-05,
+      "loss": 1.7433,
+      "step": 1296
+    },
+    {
+      "epoch": 0.07229654403567447,
+      "grad_norm": 0.43131640553474426,
+      "learning_rate": 8e-05,
+      "loss": 1.7401,
+      "step": 1297
+    },
+    {
+      "epoch": 0.07235228539576366,
+      "grad_norm": 0.44986674189567566,
+      "learning_rate": 8e-05,
+      "loss": 1.6756,
+      "step": 1298
+    },
+    {
+      "epoch": 0.07240802675585284,
+      "grad_norm": 0.4445556104183197,
+      "learning_rate": 8e-05,
+      "loss": 1.6817,
+      "step": 1299
+    },
+    {
+      "epoch": 0.07246376811594203,
+      "grad_norm": 0.4518899917602539,
+      "learning_rate": 8e-05,
+      "loss": 1.6919,
+      "step": 1300
+    },
+    {
+      "epoch": 0.07251950947603121,
+      "grad_norm": 0.44288474321365356,
+      "learning_rate": 8e-05,
+      "loss": 1.6887,
+      "step": 1301
+    },
+    {
+      "epoch": 0.0725752508361204,
+      "grad_norm": 0.4361318051815033,
+      "learning_rate": 8e-05,
+      "loss": 1.9064,
+      "step": 1302
+    },
+    {
+      "epoch": 0.07263099219620958,
+      "grad_norm": 0.4643789529800415,
+      "learning_rate": 8e-05,
+      "loss": 1.6757,
+      "step": 1303
+    },
+    {
+      "epoch": 0.07268673355629877,
+      "grad_norm": 0.44260480999946594,
+      "learning_rate": 8e-05,
+      "loss": 1.6959,
+      "step": 1304
+    },
+    {
+      "epoch": 0.07274247491638795,
+      "grad_norm": 0.4395084083080292,
+      "learning_rate": 8e-05,
+      "loss": 1.7706,
+      "step": 1305
+    },
+    {
+      "epoch": 0.07279821627647715,
+      "grad_norm": 0.4949529469013214,
+      "learning_rate": 8e-05,
+      "loss": 2.1344,
+      "step": 1306
+    },
+    {
+      "epoch": 0.07285395763656634,
+      "grad_norm": 0.4440799653530121,
+      "learning_rate": 8e-05,
+      "loss": 1.9293,
+      "step": 1307
+    },
+    {
+      "epoch": 0.07290969899665552,
+      "grad_norm": 0.42284902930259705,
+      "learning_rate": 8e-05,
+      "loss": 1.5556,
+      "step": 1308
+    },
+    {
+      "epoch": 0.07296544035674471,
+      "grad_norm": 0.395102322101593,
+      "learning_rate": 8e-05,
+      "loss": 1.7397,
+      "step": 1309
+    },
+    {
+      "epoch": 0.07302118171683389,
+      "grad_norm": 0.42531904578208923,
+      "learning_rate": 8e-05,
+      "loss": 1.7416,
+      "step": 1310
+    },
+    {
+      "epoch": 0.07307692307692308,
+      "grad_norm": 0.42302653193473816,
+      "learning_rate": 8e-05,
+      "loss": 1.6757,
+      "step": 1311
+    },
+    {
+      "epoch": 0.07313266443701226,
+      "grad_norm": 0.4515868127346039,
+      "learning_rate": 8e-05,
+      "loss": 1.9239,
+      "step": 1312
+    },
+    {
+      "epoch": 0.07318840579710145,
+      "grad_norm": 0.42175641655921936,
+      "learning_rate": 8e-05,
+      "loss": 1.6537,
+      "step": 1313
+    },
+    {
+      "epoch": 0.07324414715719063,
+      "grad_norm": 0.428892582654953,
+      "learning_rate": 8e-05,
+      "loss": 1.6068,
+      "step": 1314
+    },
+    {
+      "epoch": 0.07329988851727982,
+      "grad_norm": 0.4598456919193268,
+      "learning_rate": 8e-05,
+      "loss": 1.9036,
+      "step": 1315
+    },
+    {
+      "epoch": 0.073355629877369,
+      "grad_norm": 0.4194553792476654,
+      "learning_rate": 8e-05,
+      "loss": 1.7527,
+      "step": 1316
+    },
+    {
+      "epoch": 0.0734113712374582,
+      "grad_norm": 0.40796273946762085,
+      "learning_rate": 8e-05,
+      "loss": 1.7015,
+      "step": 1317
+    },
+    {
+      "epoch": 0.07346711259754737,
+      "grad_norm": 0.4502258002758026,
+      "learning_rate": 8e-05,
+      "loss": 1.5502,
+      "step": 1318
+    },
+    {
+      "epoch": 0.07352285395763657,
+      "grad_norm": 0.4401506185531616,
+      "learning_rate": 8e-05,
+      "loss": 1.7199,
+      "step": 1319
+    },
+    {
+      "epoch": 0.07357859531772576,
+      "grad_norm": 0.41750505566596985,
+      "learning_rate": 8e-05,
+      "loss": 1.8776,
+      "step": 1320
+    },
+    {
+      "epoch": 0.07363433667781494,
+      "grad_norm": 0.4231717586517334,
+      "learning_rate": 8e-05,
+      "loss": 1.6411,
+      "step": 1321
+    },
+    {
+      "epoch": 0.07369007803790413,
+      "grad_norm": 0.42720434069633484,
+      "learning_rate": 8e-05,
+      "loss": 1.536,
+      "step": 1322
+    },
+    {
+      "epoch": 0.07374581939799331,
+      "grad_norm": 0.47691041231155396,
+      "learning_rate": 8e-05,
+      "loss": 1.9963,
+      "step": 1323
+    },
+    {
+      "epoch": 0.0738015607580825,
+      "grad_norm": 0.4454549252986908,
+      "learning_rate": 8e-05,
+      "loss": 1.6888,
+      "step": 1324
+    },
+    {
+      "epoch": 0.07385730211817168,
+      "grad_norm": 0.4553287923336029,
+      "learning_rate": 8e-05,
+      "loss": 1.5597,
+      "step": 1325
+    },
+    {
+      "epoch": 0.07391304347826087,
+      "grad_norm": 0.4337655305862427,
+      "learning_rate": 8e-05,
+      "loss": 1.7658,
+      "step": 1326
+    },
+    {
+      "epoch": 0.07396878483835005,
+      "grad_norm": 0.46036162972450256,
+      "learning_rate": 8e-05,
+      "loss": 1.7755,
+      "step": 1327
+    },
+    {
+      "epoch": 0.07402452619843924,
+      "grad_norm": 0.45223167538642883,
+      "learning_rate": 8e-05,
+      "loss": 1.8664,
+      "step": 1328
+    },
+    {
+      "epoch": 0.07408026755852842,
+      "grad_norm": 0.44723761081695557,
+      "learning_rate": 8e-05,
+      "loss": 1.7667,
+      "step": 1329
+    },
+    {
+      "epoch": 0.07413600891861762,
+      "grad_norm": 0.4148390591144562,
+      "learning_rate": 8e-05,
+      "loss": 1.6549,
+      "step": 1330
+    },
+    {
+      "epoch": 0.0741917502787068,
+      "grad_norm": 0.4551730155944824,
+      "learning_rate": 8e-05,
+      "loss": 1.7353,
+      "step": 1331
+    },
+    {
+      "epoch": 0.07424749163879599,
+      "grad_norm": 0.4615274965763092,
+      "learning_rate": 8e-05,
+      "loss": 1.8675,
+      "step": 1332
+    },
+    {
+      "epoch": 0.07430323299888517,
+      "grad_norm": 0.49332278966903687,
+      "learning_rate": 8e-05,
+      "loss": 1.7342,
+      "step": 1333
+    },
+    {
+      "epoch": 0.07435897435897436,
+      "grad_norm": 0.44719934463500977,
+      "learning_rate": 8e-05,
+      "loss": 1.8286,
+      "step": 1334
+    },
+    {
+      "epoch": 0.07441471571906355,
+      "grad_norm": 0.42437857389450073,
+      "learning_rate": 8e-05,
+      "loss": 1.7739,
+      "step": 1335
+    },
+    {
+      "epoch": 0.07447045707915273,
+      "grad_norm": 0.4635609984397888,
+      "learning_rate": 8e-05,
+      "loss": 1.9338,
+      "step": 1336
+    },
+    {
+      "epoch": 0.07452619843924192,
+      "grad_norm": 0.438240110874176,
+      "learning_rate": 8e-05,
+      "loss": 1.6367,
+      "step": 1337
+    },
+    {
+      "epoch": 0.0745819397993311,
+      "grad_norm": 0.4692282974720001,
+      "learning_rate": 8e-05,
+      "loss": 1.8855,
+      "step": 1338
+    },
+    {
+      "epoch": 0.0746376811594203,
+      "grad_norm": 0.4450299143791199,
+      "learning_rate": 8e-05,
+      "loss": 1.9291,
+      "step": 1339
+    },
+    {
+      "epoch": 0.07469342251950947,
+      "grad_norm": 0.4614245593547821,
+      "learning_rate": 8e-05,
+      "loss": 1.8017,
+      "step": 1340
+    },
+    {
+      "epoch": 0.07474916387959867,
+      "grad_norm": 0.4570753574371338,
+      "learning_rate": 8e-05,
+      "loss": 1.656,
+      "step": 1341
+    },
+    {
+      "epoch": 0.07480490523968784,
+      "grad_norm": 0.47076496481895447,
+      "learning_rate": 8e-05,
+      "loss": 1.7928,
+      "step": 1342
+    },
+    {
+      "epoch": 0.07486064659977704,
+      "grad_norm": 0.47231245040893555,
+      "learning_rate": 8e-05,
+      "loss": 1.8747,
+      "step": 1343
+    },
+    {
+      "epoch": 0.07491638795986622,
+      "grad_norm": 0.4231669008731842,
+      "learning_rate": 8e-05,
+      "loss": 1.7642,
+      "step": 1344
+    },
+    {
+      "epoch": 0.07497212931995541,
+      "grad_norm": 0.45035478472709656,
+      "learning_rate": 8e-05,
+      "loss": 1.8285,
+      "step": 1345
+    },
+    {
+      "epoch": 0.07502787068004459,
+      "grad_norm": 0.4413871467113495,
+      "learning_rate": 8e-05,
+      "loss": 1.8548,
+      "step": 1346
+    },
+    {
+      "epoch": 0.07508361204013378,
+      "grad_norm": 0.4781755208969116,
+      "learning_rate": 8e-05,
+      "loss": 1.8532,
+      "step": 1347
+    },
+    {
+      "epoch": 0.07513935340022297,
+      "grad_norm": 0.41016334295272827,
+      "learning_rate": 8e-05,
+      "loss": 1.5955,
+      "step": 1348
+    },
+    {
+      "epoch": 0.07519509476031215,
+      "grad_norm": 0.4543410837650299,
+      "learning_rate": 8e-05,
+      "loss": 1.9111,
+      "step": 1349
+    },
+    {
+      "epoch": 0.07525083612040134,
+      "grad_norm": 0.43364307284355164,
+      "learning_rate": 8e-05,
+      "loss": 1.7241,
+      "step": 1350
+    },
+    {
+      "epoch": 0.07530657748049052,
+      "grad_norm": 0.44212257862091064,
+      "learning_rate": 8e-05,
+      "loss": 1.6172,
+      "step": 1351
+    },
+    {
+      "epoch": 0.07536231884057971,
+      "grad_norm": 0.777114748954773,
+      "learning_rate": 8e-05,
+      "loss": 1.8149,
+      "step": 1352
+    },
+    {
+      "epoch": 0.0754180602006689,
+      "grad_norm": 0.4357220232486725,
+      "learning_rate": 8e-05,
+      "loss": 1.7667,
+      "step": 1353
+    },
+    {
+      "epoch": 0.07547380156075809,
+      "grad_norm": 0.4471208155155182,
+      "learning_rate": 8e-05,
+      "loss": 1.5841,
+      "step": 1354
+    },
+    {
+      "epoch": 0.07552954292084726,
+      "grad_norm": 0.43049171566963196,
+      "learning_rate": 8e-05,
+      "loss": 1.7802,
+      "step": 1355
+    },
+    {
+      "epoch": 0.07558528428093646,
+      "grad_norm": 0.4427327513694763,
+      "learning_rate": 8e-05,
+      "loss": 1.897,
+      "step": 1356
+    },
+    {
+      "epoch": 0.07564102564102564,
+      "grad_norm": 0.45292583107948303,
+      "learning_rate": 8e-05,
+      "loss": 1.8114,
+      "step": 1357
+    },
+    {
+      "epoch": 0.07569676700111483,
+      "grad_norm": 0.45449328422546387,
+      "learning_rate": 8e-05,
+      "loss": 1.8064,
+      "step": 1358
+    },
+    {
+      "epoch": 0.07575250836120401,
+      "grad_norm": 0.4421929717063904,
+      "learning_rate": 8e-05,
+      "loss": 1.9081,
+      "step": 1359
+    },
+    {
+      "epoch": 0.0758082497212932,
+      "grad_norm": 0.4415033459663391,
+      "learning_rate": 8e-05,
+      "loss": 1.7954,
+      "step": 1360
+    },
+    {
+      "epoch": 0.07586399108138239,
+      "grad_norm": 0.494498074054718,
+      "learning_rate": 8e-05,
+      "loss": 1.5815,
+      "step": 1361
+    },
+    {
+      "epoch": 0.07591973244147157,
+      "grad_norm": 0.5168023109436035,
+      "learning_rate": 8e-05,
+      "loss": 1.9749,
+      "step": 1362
+    },
+    {
+      "epoch": 0.07597547380156076,
+      "grad_norm": 0.45021355152130127,
+      "learning_rate": 8e-05,
+      "loss": 1.8096,
+      "step": 1363
+    },
+    {
+      "epoch": 0.07603121516164994,
+      "grad_norm": 0.4276469647884369,
+      "learning_rate": 8e-05,
+      "loss": 1.7189,
+      "step": 1364
+    },
+    {
+      "epoch": 0.07608695652173914,
+      "grad_norm": 0.4566865861415863,
+      "learning_rate": 8e-05,
+      "loss": 1.807,
+      "step": 1365
+    },
+    {
+      "epoch": 0.07614269788182831,
+      "grad_norm": 0.445454478263855,
+      "learning_rate": 8e-05,
+      "loss": 1.8407,
+      "step": 1366
+    },
+    {
+      "epoch": 0.0761984392419175,
+      "grad_norm": 0.48000916838645935,
+      "learning_rate": 8e-05,
+      "loss": 1.8314,
+      "step": 1367
+    },
+    {
+      "epoch": 0.07625418060200669,
+      "grad_norm": 0.43749409914016724,
+      "learning_rate": 8e-05,
+      "loss": 1.6374,
+      "step": 1368
+    },
+    {
+      "epoch": 0.07630992196209588,
+      "grad_norm": 0.46065568923950195,
+      "learning_rate": 8e-05,
+      "loss": 1.7961,
+      "step": 1369
+    },
+    {
+      "epoch": 0.07636566332218506,
+      "grad_norm": 0.5014117956161499,
+      "learning_rate": 8e-05,
+      "loss": 1.9533,
+      "step": 1370
+    },
+    {
+      "epoch": 0.07642140468227425,
+      "grad_norm": 0.4392300546169281,
+      "learning_rate": 8e-05,
+      "loss": 1.7795,
+      "step": 1371
+    },
+    {
+      "epoch": 0.07647714604236343,
+      "grad_norm": 0.4479599595069885,
+      "learning_rate": 8e-05,
+      "loss": 1.6262,
+      "step": 1372
+    },
+    {
+      "epoch": 0.07653288740245262,
+      "grad_norm": 0.4235548675060272,
+      "learning_rate": 8e-05,
+      "loss": 1.6683,
+      "step": 1373
+    },
+    {
+      "epoch": 0.0765886287625418,
+      "grad_norm": 0.4663847088813782,
+      "learning_rate": 8e-05,
+      "loss": 1.8703,
+      "step": 1374
+    },
+    {
+      "epoch": 0.07664437012263099,
+      "grad_norm": 0.5052254796028137,
+      "learning_rate": 8e-05,
+      "loss": 1.7682,
+      "step": 1375
+    },
+    {
+      "epoch": 0.07670011148272018,
+      "grad_norm": 0.4608347713947296,
+      "learning_rate": 8e-05,
+      "loss": 1.8169,
+      "step": 1376
+    },
+    {
+      "epoch": 0.07675585284280936,
+      "grad_norm": 0.4845850467681885,
+      "learning_rate": 8e-05,
+      "loss": 1.9304,
+      "step": 1377
+    },
+    {
+      "epoch": 0.07681159420289856,
+      "grad_norm": 0.4543549120426178,
+      "learning_rate": 8e-05,
+      "loss": 1.9527,
+      "step": 1378
+    },
+    {
+      "epoch": 0.07686733556298773,
+      "grad_norm": 0.4503234922885895,
+      "learning_rate": 8e-05,
+      "loss": 1.7021,
+      "step": 1379
+    },
+    {
+      "epoch": 0.07692307692307693,
+      "grad_norm": 0.48853322863578796,
+      "learning_rate": 8e-05,
+      "loss": 1.7907,
+      "step": 1380
+    },
+    {
+      "epoch": 0.0769788182831661,
+      "grad_norm": 0.4230073094367981,
+      "learning_rate": 8e-05,
+      "loss": 1.7625,
+      "step": 1381
+    },
+    {
+      "epoch": 0.0770345596432553,
+      "grad_norm": 0.4668857753276825,
+      "learning_rate": 8e-05,
+      "loss": 1.7012,
+      "step": 1382
+    },
+    {
+      "epoch": 0.07709030100334448,
+      "grad_norm": 0.43255892395973206,
+      "learning_rate": 8e-05,
+      "loss": 1.7422,
+      "step": 1383
+    },
+    {
+      "epoch": 0.07714604236343367,
+      "grad_norm": 0.44348350167274475,
+      "learning_rate": 8e-05,
+      "loss": 1.6422,
+      "step": 1384
+    },
+    {
+      "epoch": 0.07720178372352285,
+      "grad_norm": 0.45581987500190735,
+      "learning_rate": 8e-05,
+      "loss": 1.8175,
+      "step": 1385
+    },
+    {
+      "epoch": 0.07725752508361204,
+      "grad_norm": 0.518310010433197,
+      "learning_rate": 8e-05,
+      "loss": 1.9449,
+      "step": 1386
+    },
+    {
+      "epoch": 0.07731326644370122,
+      "grad_norm": 0.4802246391773224,
+      "learning_rate": 8e-05,
+      "loss": 1.9455,
+      "step": 1387
+    },
+    {
+      "epoch": 0.07736900780379041,
+      "grad_norm": 0.4345766603946686,
+      "learning_rate": 8e-05,
+      "loss": 1.6838,
+      "step": 1388
+    },
+    {
+      "epoch": 0.0774247491638796,
+      "grad_norm": 0.4461798071861267,
+      "learning_rate": 8e-05,
+      "loss": 1.8256,
+      "step": 1389
+    },
+    {
+      "epoch": 0.07748049052396878,
+      "grad_norm": 0.45585009455680847,
+      "learning_rate": 8e-05,
+      "loss": 1.7255,
+      "step": 1390
+    },
+    {
+      "epoch": 0.07753623188405798,
+      "grad_norm": 0.4471147954463959,
+      "learning_rate": 8e-05,
+      "loss": 1.7223,
+      "step": 1391
+    },
+    {
+      "epoch": 0.07759197324414716,
+      "grad_norm": 0.4584360420703888,
+      "learning_rate": 8e-05,
+      "loss": 1.8044,
+      "step": 1392
+    },
+    {
+      "epoch": 0.07764771460423635,
+      "grad_norm": 0.46038952469825745,
+      "learning_rate": 8e-05,
+      "loss": 1.6906,
+      "step": 1393
+    },
+    {
+      "epoch": 0.07770345596432553,
+      "grad_norm": 0.45119255781173706,
+      "learning_rate": 8e-05,
+      "loss": 1.8722,
+      "step": 1394
+    },
+    {
+      "epoch": 0.07775919732441472,
+      "grad_norm": 0.4340799152851105,
+      "learning_rate": 8e-05,
+      "loss": 1.8374,
+      "step": 1395
+    },
+    {
+      "epoch": 0.0778149386845039,
+      "grad_norm": 0.44295990467071533,
+      "learning_rate": 8e-05,
+      "loss": 1.7464,
+      "step": 1396
+    },
+    {
+      "epoch": 0.07787068004459309,
+      "grad_norm": 0.4677167236804962,
+      "learning_rate": 8e-05,
+      "loss": 1.7631,
+      "step": 1397
+    },
+    {
+      "epoch": 0.07792642140468227,
+      "grad_norm": 0.4594052731990814,
+      "learning_rate": 8e-05,
+      "loss": 1.7957,
+      "step": 1398
+    },
+    {
+      "epoch": 0.07798216276477146,
+      "grad_norm": 0.4324091076850891,
+      "learning_rate": 8e-05,
+      "loss": 1.7465,
+      "step": 1399
+    },
+    {
+      "epoch": 0.07803790412486064,
+      "grad_norm": 0.45841190218925476,
+      "learning_rate": 8e-05,
+      "loss": 1.7334,
+      "step": 1400
+    },
+    {
+      "epoch": 0.07809364548494983,
+      "grad_norm": 0.5322265625,
+      "learning_rate": 8e-05,
+      "loss": 2.1824,
+      "step": 1401
+    },
+    {
+      "epoch": 0.07814938684503901,
+      "grad_norm": 0.36952677369117737,
+      "learning_rate": 8e-05,
+      "loss": 1.4223,
+      "step": 1402
+    },
+    {
+      "epoch": 0.0782051282051282,
+      "grad_norm": 0.486471027135849,
+      "learning_rate": 8e-05,
+      "loss": 1.8555,
+      "step": 1403
+    },
+    {
+      "epoch": 0.0782608695652174,
+      "grad_norm": 0.43990740180015564,
+      "learning_rate": 8e-05,
+      "loss": 1.8663,
+      "step": 1404
+    },
+    {
+      "epoch": 0.07831661092530658,
+      "grad_norm": 0.4258549213409424,
+      "learning_rate": 8e-05,
+      "loss": 1.8713,
+      "step": 1405
+    },
+    {
+      "epoch": 0.07837235228539577,
+      "grad_norm": 0.510369598865509,
+      "learning_rate": 8e-05,
+      "loss": 1.9984,
+      "step": 1406
+    },
+    {
+      "epoch": 0.07842809364548495,
+      "grad_norm": 0.42820146679878235,
+      "learning_rate": 8e-05,
+      "loss": 1.7277,
+      "step": 1407
+    },
+    {
+      "epoch": 0.07848383500557414,
+      "grad_norm": 0.4371456801891327,
+      "learning_rate": 8e-05,
+      "loss": 1.5951,
+      "step": 1408
+    },
+    {
+      "epoch": 0.07853957636566332,
+      "grad_norm": 0.5079208016395569,
+      "learning_rate": 8e-05,
+      "loss": 1.6506,
+      "step": 1409
+    },
+    {
+      "epoch": 0.07859531772575251,
+      "grad_norm": 0.4610532224178314,
+      "learning_rate": 8e-05,
+      "loss": 1.9321,
+      "step": 1410
+    },
+    {
+      "epoch": 0.07865105908584169,
+      "grad_norm": 0.4237179756164551,
+      "learning_rate": 8e-05,
+      "loss": 1.7152,
+      "step": 1411
+    },
+    {
+      "epoch": 0.07870680044593088,
+      "grad_norm": 0.4514097571372986,
+      "learning_rate": 8e-05,
+      "loss": 1.659,
+      "step": 1412
+    },
+    {
+      "epoch": 0.07876254180602006,
+      "grad_norm": 0.4557318389415741,
+      "learning_rate": 8e-05,
+      "loss": 1.8183,
+      "step": 1413
+    },
+    {
+      "epoch": 0.07881828316610925,
+      "grad_norm": 0.42246413230895996,
+      "learning_rate": 8e-05,
+      "loss": 1.8077,
+      "step": 1414
+    },
+    {
+      "epoch": 0.07887402452619843,
+      "grad_norm": 0.4116936922073364,
+      "learning_rate": 8e-05,
+      "loss": 1.5519,
+      "step": 1415
+    },
+    {
+      "epoch": 0.07892976588628763,
+      "grad_norm": 0.4407467842102051,
+      "learning_rate": 8e-05,
+      "loss": 1.8434,
+      "step": 1416
+    },
+    {
+      "epoch": 0.07898550724637682,
+      "grad_norm": 0.5225424766540527,
+      "learning_rate": 8e-05,
+      "loss": 1.9881,
+      "step": 1417
+    },
+    {
+      "epoch": 0.079041248606466,
+      "grad_norm": 0.4358792006969452,
+      "learning_rate": 8e-05,
+      "loss": 1.7754,
+      "step": 1418
+    },
+    {
+      "epoch": 0.07909698996655519,
+      "grad_norm": 0.4558788537979126,
+      "learning_rate": 8e-05,
+      "loss": 1.7594,
+      "step": 1419
+    },
+    {
+      "epoch": 0.07915273132664437,
+      "grad_norm": 0.532288134098053,
+      "learning_rate": 8e-05,
+      "loss": 2.1258,
+      "step": 1420
+    },
+    {
+      "epoch": 0.07920847268673356,
+      "grad_norm": 0.5065916180610657,
+      "learning_rate": 8e-05,
+      "loss": 1.8183,
+      "step": 1421
+    },
+    {
+      "epoch": 0.07926421404682274,
+      "grad_norm": 0.44754672050476074,
+      "learning_rate": 8e-05,
+      "loss": 1.6664,
+      "step": 1422
+    },
+    {
+      "epoch": 0.07931995540691193,
+      "grad_norm": 0.45572909712791443,
+      "learning_rate": 8e-05,
+      "loss": 1.915,
+      "step": 1423
+    },
+    {
+      "epoch": 0.07937569676700111,
+      "grad_norm": 0.43835529685020447,
+      "learning_rate": 8e-05,
+      "loss": 1.7239,
+      "step": 1424
+    },
+    {
+      "epoch": 0.0794314381270903,
+      "grad_norm": 0.43413209915161133,
+      "learning_rate": 8e-05,
+      "loss": 1.8262,
+      "step": 1425
+    },
+    {
+      "epoch": 0.07948717948717948,
+      "grad_norm": 0.4330446124076843,
+      "learning_rate": 8e-05,
+      "loss": 1.4287,
+      "step": 1426
+    },
+    {
+      "epoch": 0.07954292084726867,
+      "grad_norm": 0.4555690884590149,
+      "learning_rate": 8e-05,
+      "loss": 1.638,
+      "step": 1427
+    },
+    {
+      "epoch": 0.07959866220735785,
+      "grad_norm": 0.4385247826576233,
+      "learning_rate": 8e-05,
+      "loss": 1.6261,
+      "step": 1428
+    },
+    {
+      "epoch": 0.07965440356744705,
+      "grad_norm": 0.4520176649093628,
+      "learning_rate": 8e-05,
+      "loss": 1.8925,
+      "step": 1429
+    },
+    {
+      "epoch": 0.07971014492753623,
+      "grad_norm": 0.43680456280708313,
+      "learning_rate": 8e-05,
+      "loss": 1.9335,
+      "step": 1430
+    },
+    {
+      "epoch": 0.07976588628762542,
+      "grad_norm": 0.4729453921318054,
+      "learning_rate": 8e-05,
+      "loss": 1.8992,
+      "step": 1431
+    },
+    {
+      "epoch": 0.07982162764771461,
+      "grad_norm": 0.4304512143135071,
+      "learning_rate": 8e-05,
+      "loss": 1.6413,
+      "step": 1432
+    },
+    {
+      "epoch": 0.07987736900780379,
+      "grad_norm": 0.41261231899261475,
+      "learning_rate": 8e-05,
+      "loss": 1.8193,
+      "step": 1433
+    },
+    {
+      "epoch": 0.07993311036789298,
+      "grad_norm": 0.47836706042289734,
+      "learning_rate": 8e-05,
+      "loss": 1.8577,
+      "step": 1434
+    },
+    {
+      "epoch": 0.07998885172798216,
+      "grad_norm": 0.5016499161720276,
+      "learning_rate": 8e-05,
+      "loss": 1.9283,
+      "step": 1435
+    },
+    {
+      "epoch": 0.08004459308807135,
+      "grad_norm": 0.4799736440181732,
+      "learning_rate": 8e-05,
+      "loss": 1.9349,
+      "step": 1436
+    },
+    {
+      "epoch": 0.08010033444816053,
+      "grad_norm": 0.45294785499572754,
+      "learning_rate": 8e-05,
+      "loss": 1.9936,
+      "step": 1437
+    },
+    {
+      "epoch": 0.08015607580824972,
+      "grad_norm": 0.3943483233451843,
+      "learning_rate": 8e-05,
+      "loss": 1.5038,
+      "step": 1438
+    },
+    {
+      "epoch": 0.0802118171683389,
+      "grad_norm": 0.44112369418144226,
+      "learning_rate": 8e-05,
+      "loss": 1.6879,
+      "step": 1439
+    },
+    {
+      "epoch": 0.0802675585284281,
+      "grad_norm": 0.48437753319740295,
+      "learning_rate": 8e-05,
+      "loss": 2.0316,
+      "step": 1440
+    },
+    {
+      "epoch": 0.08032329988851727,
+      "grad_norm": 0.4124515652656555,
+      "learning_rate": 8e-05,
+      "loss": 1.4986,
+      "step": 1441
+    },
+    {
+      "epoch": 0.08037904124860647,
+      "grad_norm": 0.46731048822402954,
+      "learning_rate": 8e-05,
+      "loss": 1.7208,
+      "step": 1442
+    },
+    {
+      "epoch": 0.08043478260869565,
+      "grad_norm": 0.47354432940483093,
+      "learning_rate": 8e-05,
+      "loss": 1.9603,
+      "step": 1443
+    },
+    {
+      "epoch": 0.08049052396878484,
+      "grad_norm": 0.4480750262737274,
+      "learning_rate": 8e-05,
+      "loss": 1.7549,
+      "step": 1444
+    },
+    {
+      "epoch": 0.08054626532887403,
+      "grad_norm": 0.4697958528995514,
+      "learning_rate": 8e-05,
+      "loss": 1.4256,
+      "step": 1445
+    },
+    {
+      "epoch": 0.08060200668896321,
+      "grad_norm": 0.42544737458229065,
+      "learning_rate": 8e-05,
+      "loss": 1.7851,
+      "step": 1446
+    },
+    {
+      "epoch": 0.0806577480490524,
+      "grad_norm": 0.44435450434684753,
+      "learning_rate": 8e-05,
+      "loss": 1.6728,
+      "step": 1447
+    },
+    {
+      "epoch": 0.08071348940914158,
+      "grad_norm": 0.4432964026927948,
+      "learning_rate": 8e-05,
+      "loss": 1.755,
+      "step": 1448
+    },
+    {
+      "epoch": 0.08076923076923077,
+      "grad_norm": 0.4708921015262604,
+      "learning_rate": 8e-05,
+      "loss": 1.6021,
+      "step": 1449
+    },
+    {
+      "epoch": 0.08082497212931995,
+      "grad_norm": 0.4536772072315216,
+      "learning_rate": 8e-05,
+      "loss": 1.76,
+      "step": 1450
+    },
+    {
+      "epoch": 0.08088071348940915,
+      "grad_norm": 0.4945363700389862,
+      "learning_rate": 8e-05,
+      "loss": 1.9236,
+      "step": 1451
+    },
+    {
+      "epoch": 0.08093645484949832,
+      "grad_norm": 0.48008787631988525,
+      "learning_rate": 8e-05,
+      "loss": 1.9332,
+      "step": 1452
+    },
+    {
+      "epoch": 0.08099219620958752,
+      "grad_norm": 0.446678102016449,
+      "learning_rate": 8e-05,
+      "loss": 1.7992,
+      "step": 1453
+    },
+    {
+      "epoch": 0.0810479375696767,
+      "grad_norm": 0.4077160358428955,
+      "learning_rate": 8e-05,
+      "loss": 1.5497,
+      "step": 1454
+    },
+    {
+      "epoch": 0.08110367892976589,
+      "grad_norm": 0.4312504529953003,
+      "learning_rate": 8e-05,
+      "loss": 1.8776,
+      "step": 1455
+    },
+    {
+      "epoch": 0.08115942028985507,
+      "grad_norm": 0.42562606930732727,
+      "learning_rate": 8e-05,
+      "loss": 1.6336,
+      "step": 1456
+    },
+    {
+      "epoch": 0.08121516164994426,
+      "grad_norm": 0.44738060235977173,
+      "learning_rate": 8e-05,
+      "loss": 1.7712,
+      "step": 1457
+    },
+    {
+      "epoch": 0.08127090301003345,
+      "grad_norm": 0.4285328686237335,
+      "learning_rate": 8e-05,
+      "loss": 1.8588,
+      "step": 1458
+    },
+    {
+      "epoch": 0.08132664437012263,
+      "grad_norm": 0.5084940195083618,
+      "learning_rate": 8e-05,
+      "loss": 2.0677,
+      "step": 1459
+    },
+    {
+      "epoch": 0.08138238573021182,
+      "grad_norm": 0.4501677453517914,
+      "learning_rate": 8e-05,
+      "loss": 1.7812,
+      "step": 1460
+    },
+    {
+      "epoch": 0.081438127090301,
+      "grad_norm": 0.4214770793914795,
+      "learning_rate": 8e-05,
+      "loss": 1.578,
+      "step": 1461
+    },
+    {
+      "epoch": 0.0814938684503902,
+      "grad_norm": 0.4178116023540497,
+      "learning_rate": 8e-05,
+      "loss": 1.6665,
+      "step": 1462
+    },
+    {
+      "epoch": 0.08154960981047937,
+      "grad_norm": 0.46865931153297424,
+      "learning_rate": 8e-05,
+      "loss": 1.8594,
+      "step": 1463
+    },
+    {
+      "epoch": 0.08160535117056857,
+      "grad_norm": 0.4274550974369049,
+      "learning_rate": 8e-05,
+      "loss": 1.7344,
+      "step": 1464
+    },
+    {
+      "epoch": 0.08166109253065774,
+      "grad_norm": 0.47668880224227905,
+      "learning_rate": 8e-05,
+      "loss": 1.7793,
+      "step": 1465
+    },
+    {
+      "epoch": 0.08171683389074694,
+      "grad_norm": 0.45111218094825745,
+      "learning_rate": 8e-05,
+      "loss": 1.5396,
+      "step": 1466
+    },
+    {
+      "epoch": 0.08177257525083612,
+      "grad_norm": 0.46514469385147095,
+      "learning_rate": 8e-05,
+      "loss": 1.6373,
+      "step": 1467
+    },
+    {
+      "epoch": 0.08182831661092531,
+      "grad_norm": 0.4513457119464874,
+      "learning_rate": 8e-05,
+      "loss": 1.9016,
+      "step": 1468
+    },
+    {
+      "epoch": 0.08188405797101449,
+      "grad_norm": 0.4727977216243744,
+      "learning_rate": 8e-05,
+      "loss": 1.7992,
+      "step": 1469
+    },
+    {
+      "epoch": 0.08193979933110368,
+      "grad_norm": 0.4027753174304962,
+      "learning_rate": 8e-05,
+      "loss": 1.5857,
+      "step": 1470
+    },
+    {
+      "epoch": 0.08199554069119286,
+      "grad_norm": 0.4126647114753723,
+      "learning_rate": 8e-05,
+      "loss": 1.5366,
+      "step": 1471
+    },
+    {
+      "epoch": 0.08205128205128205,
+      "grad_norm": 0.40489810705184937,
+      "learning_rate": 8e-05,
+      "loss": 1.781,
+      "step": 1472
+    },
+    {
+      "epoch": 0.08210702341137124,
+      "grad_norm": 0.4544224143028259,
+      "learning_rate": 8e-05,
+      "loss": 1.7395,
+      "step": 1473
+    },
+    {
+      "epoch": 0.08216276477146042,
+      "grad_norm": 0.4425218999385834,
+      "learning_rate": 8e-05,
+      "loss": 1.8251,
+      "step": 1474
+    },
+    {
+      "epoch": 0.08221850613154962,
+      "grad_norm": 0.4314938187599182,
+      "learning_rate": 8e-05,
+      "loss": 1.6134,
+      "step": 1475
+    },
+    {
+      "epoch": 0.0822742474916388,
+      "grad_norm": 0.4418397545814514,
+      "learning_rate": 8e-05,
+      "loss": 1.7419,
+      "step": 1476
+    },
+    {
+      "epoch": 0.08232998885172799,
+      "grad_norm": 0.4508616626262665,
+      "learning_rate": 8e-05,
+      "loss": 1.5632,
+      "step": 1477
+    },
+    {
+      "epoch": 0.08238573021181717,
+      "grad_norm": 0.44871529936790466,
+      "learning_rate": 8e-05,
+      "loss": 1.5814,
+      "step": 1478
+    },
+    {
+      "epoch": 0.08244147157190636,
+      "grad_norm": 0.4989759027957916,
+      "learning_rate": 8e-05,
+      "loss": 1.8102,
+      "step": 1479
+    },
+    {
+      "epoch": 0.08249721293199554,
+      "grad_norm": 0.4352324903011322,
+      "learning_rate": 8e-05,
+      "loss": 1.8375,
+      "step": 1480
+    },
+    {
+      "epoch": 0.08255295429208473,
+      "grad_norm": 0.452763170003891,
+      "learning_rate": 8e-05,
+      "loss": 1.9005,
+      "step": 1481
+    },
+    {
+      "epoch": 0.08260869565217391,
+      "grad_norm": 0.5133951306343079,
+      "learning_rate": 8e-05,
+      "loss": 1.9717,
+      "step": 1482
+    },
+    {
+      "epoch": 0.0826644370122631,
+      "grad_norm": 0.45723623037338257,
+      "learning_rate": 8e-05,
+      "loss": 1.7423,
+      "step": 1483
+    },
+    {
+      "epoch": 0.08272017837235228,
+      "grad_norm": 0.4535520672798157,
+      "learning_rate": 8e-05,
+      "loss": 1.9049,
+      "step": 1484
+    },
+    {
+      "epoch": 0.08277591973244147,
+      "grad_norm": 0.4558051824569702,
+      "learning_rate": 8e-05,
+      "loss": 1.8305,
+      "step": 1485
+    },
+    {
+      "epoch": 0.08283166109253066,
+      "grad_norm": 0.43357986211776733,
+      "learning_rate": 8e-05,
+      "loss": 1.7504,
+      "step": 1486
+    },
+    {
+      "epoch": 0.08288740245261984,
+      "grad_norm": 0.4610099196434021,
+      "learning_rate": 8e-05,
+      "loss": 1.8672,
+      "step": 1487
+    },
+    {
+      "epoch": 0.08294314381270904,
+      "grad_norm": 0.48371803760528564,
+      "learning_rate": 8e-05,
+      "loss": 1.9017,
+      "step": 1488
+    },
+    {
+      "epoch": 0.08299888517279821,
+      "grad_norm": 0.4555537700653076,
+      "learning_rate": 8e-05,
+      "loss": 1.7175,
+      "step": 1489
+    },
+    {
+      "epoch": 0.08305462653288741,
+      "grad_norm": 0.463793009519577,
+      "learning_rate": 8e-05,
+      "loss": 1.6915,
+      "step": 1490
+    },
+    {
+      "epoch": 0.08311036789297659,
+      "grad_norm": 0.46372777223587036,
+      "learning_rate": 8e-05,
+      "loss": 1.9179,
+      "step": 1491
+    },
+    {
+      "epoch": 0.08316610925306578,
+      "grad_norm": 0.4361550509929657,
+      "learning_rate": 8e-05,
+      "loss": 1.7615,
+      "step": 1492
+    },
+    {
+      "epoch": 0.08322185061315496,
+      "grad_norm": 0.4673331677913666,
+      "learning_rate": 8e-05,
+      "loss": 1.937,
+      "step": 1493
+    },
+    {
+      "epoch": 0.08327759197324415,
+      "grad_norm": 0.41564467549324036,
+      "learning_rate": 8e-05,
+      "loss": 1.5847,
+      "step": 1494
+    },
+    {
+      "epoch": 0.08333333333333333,
+      "grad_norm": 0.4272595942020416,
+      "learning_rate": 8e-05,
+      "loss": 1.7388,
+      "step": 1495
+    },
+    {
+      "epoch": 0.08338907469342252,
+      "grad_norm": 0.44510963559150696,
+      "learning_rate": 8e-05,
+      "loss": 1.7609,
+      "step": 1496
+    },
+    {
+      "epoch": 0.0834448160535117,
+      "grad_norm": 0.4720876216888428,
+      "learning_rate": 8e-05,
+      "loss": 1.9245,
+      "step": 1497
+    },
+    {
+      "epoch": 0.08350055741360089,
+      "grad_norm": 0.4473607540130615,
+      "learning_rate": 8e-05,
+      "loss": 1.6843,
+      "step": 1498
+    },
+    {
+      "epoch": 0.08355629877369007,
+      "grad_norm": 0.4595058858394623,
+      "learning_rate": 8e-05,
+      "loss": 1.6747,
+      "step": 1499
+    },
+    {
+      "epoch": 0.08361204013377926,
+      "grad_norm": 0.43724626302719116,
+      "learning_rate": 8e-05,
+      "loss": 1.6508,
+      "step": 1500
+    },
+    {
+      "epoch": 0.08366778149386846,
+      "grad_norm": 0.4269401729106903,
+      "learning_rate": 8e-05,
+      "loss": 1.6373,
+      "step": 1501
+    },
+    {
+      "epoch": 0.08372352285395764,
+      "grad_norm": 0.4769815504550934,
+      "learning_rate": 8e-05,
+      "loss": 1.9236,
+      "step": 1502
+    },
+    {
+      "epoch": 0.08377926421404683,
+      "grad_norm": 0.4663258194923401,
+      "learning_rate": 8e-05,
+      "loss": 1.8029,
+      "step": 1503
+    },
+    {
+      "epoch": 0.083835005574136,
+      "grad_norm": 0.44873517751693726,
+      "learning_rate": 8e-05,
+      "loss": 1.965,
+      "step": 1504
+    },
+    {
+      "epoch": 0.0838907469342252,
+      "grad_norm": 0.4724174737930298,
+      "learning_rate": 8e-05,
+      "loss": 1.9627,
+      "step": 1505
+    },
+    {
+      "epoch": 0.08394648829431438,
+      "grad_norm": 0.43560653924942017,
+      "learning_rate": 8e-05,
+      "loss": 1.7095,
+      "step": 1506
+    },
+    {
+      "epoch": 0.08400222965440357,
+      "grad_norm": 0.45047271251678467,
+      "learning_rate": 8e-05,
+      "loss": 1.8491,
+      "step": 1507
+    },
+    {
+      "epoch": 0.08405797101449275,
+      "grad_norm": 0.4241717457771301,
+      "learning_rate": 8e-05,
+      "loss": 1.6695,
+      "step": 1508
+    },
+    {
+      "epoch": 0.08411371237458194,
+      "grad_norm": 0.4313092529773712,
+      "learning_rate": 8e-05,
+      "loss": 1.6094,
+      "step": 1509
+    },
+    {
+      "epoch": 0.08416945373467112,
+      "grad_norm": 0.4118594527244568,
+      "learning_rate": 8e-05,
+      "loss": 1.659,
+      "step": 1510
+    },
+    {
+      "epoch": 0.08422519509476031,
+      "grad_norm": 0.43712878227233887,
+      "learning_rate": 8e-05,
+      "loss": 1.5832,
+      "step": 1511
+    },
+    {
+      "epoch": 0.08428093645484949,
+      "grad_norm": 0.515166699886322,
+      "learning_rate": 8e-05,
+      "loss": 1.6133,
+      "step": 1512
+    },
+    {
+      "epoch": 0.08433667781493868,
+      "grad_norm": 0.4808172285556793,
+      "learning_rate": 8e-05,
+      "loss": 1.8906,
+      "step": 1513
+    },
+    {
+      "epoch": 0.08439241917502788,
+      "grad_norm": 0.451366126537323,
+      "learning_rate": 8e-05,
+      "loss": 1.8264,
+      "step": 1514
+    },
+    {
+      "epoch": 0.08444816053511706,
+      "grad_norm": 0.47188040614128113,
+      "learning_rate": 8e-05,
+      "loss": 1.863,
+      "step": 1515
+    },
+    {
+      "epoch": 0.08450390189520625,
+      "grad_norm": 0.4248305857181549,
+      "learning_rate": 8e-05,
+      "loss": 1.6737,
+      "step": 1516
+    },
+    {
+      "epoch": 0.08455964325529543,
+      "grad_norm": 0.44855618476867676,
+      "learning_rate": 8e-05,
+      "loss": 1.7684,
+      "step": 1517
+    },
+    {
+      "epoch": 0.08461538461538462,
+      "grad_norm": 0.4507581293582916,
+      "learning_rate": 8e-05,
+      "loss": 1.9199,
+      "step": 1518
+    },
+    {
+      "epoch": 0.0846711259754738,
+      "grad_norm": 0.4774366617202759,
+      "learning_rate": 8e-05,
+      "loss": 2.0712,
+      "step": 1519
+    },
+    {
+      "epoch": 0.08472686733556299,
+      "grad_norm": 0.45017415285110474,
+      "learning_rate": 8e-05,
+      "loss": 1.7392,
+      "step": 1520
+    },
+    {
+      "epoch": 0.08478260869565217,
+      "grad_norm": 0.5057975649833679,
+      "learning_rate": 8e-05,
+      "loss": 1.8693,
+      "step": 1521
+    },
+    {
+      "epoch": 0.08483835005574136,
+      "grad_norm": 0.4550042748451233,
+      "learning_rate": 8e-05,
+      "loss": 1.6986,
+      "step": 1522
+    },
+    {
+      "epoch": 0.08489409141583054,
+      "grad_norm": 0.48192504048347473,
+      "learning_rate": 8e-05,
+      "loss": 1.868,
+      "step": 1523
+    },
+    {
+      "epoch": 0.08494983277591973,
+      "grad_norm": 0.49184849858283997,
+      "learning_rate": 8e-05,
+      "loss": 1.7318,
+      "step": 1524
+    },
+    {
+      "epoch": 0.08500557413600891,
+      "grad_norm": 0.46011582016944885,
+      "learning_rate": 8e-05,
+      "loss": 1.7582,
+      "step": 1525
+    },
+    {
+      "epoch": 0.0850613154960981,
+      "grad_norm": 0.49036216735839844,
+      "learning_rate": 8e-05,
+      "loss": 1.8992,
+      "step": 1526
+    },
+    {
+      "epoch": 0.08511705685618728,
+      "grad_norm": 0.4744269549846649,
+      "learning_rate": 8e-05,
+      "loss": 1.8252,
+      "step": 1527
+    },
+    {
+      "epoch": 0.08517279821627648,
+      "grad_norm": 0.4782600998878479,
+      "learning_rate": 8e-05,
+      "loss": 1.6982,
+      "step": 1528
+    },
+    {
+      "epoch": 0.08522853957636567,
+      "grad_norm": 0.473105251789093,
+      "learning_rate": 8e-05,
+      "loss": 1.9843,
+      "step": 1529
+    },
+    {
+      "epoch": 0.08528428093645485,
+      "grad_norm": 0.4600432515144348,
+      "learning_rate": 8e-05,
+      "loss": 1.8821,
+      "step": 1530
+    },
+    {
+      "epoch": 0.08534002229654404,
+      "grad_norm": 0.4404701590538025,
+      "learning_rate": 8e-05,
+      "loss": 1.6798,
+      "step": 1531
+    },
+    {
+      "epoch": 0.08539576365663322,
+      "grad_norm": 0.4240405559539795,
+      "learning_rate": 8e-05,
+      "loss": 1.7695,
+      "step": 1532
+    },
+    {
+      "epoch": 0.08545150501672241,
+      "grad_norm": 0.4390221834182739,
+      "learning_rate": 8e-05,
+      "loss": 1.4667,
+      "step": 1533
+    },
+    {
+      "epoch": 0.08550724637681159,
+      "grad_norm": 0.5003925561904907,
+      "learning_rate": 8e-05,
+      "loss": 1.82,
+      "step": 1534
+    },
+    {
+      "epoch": 0.08556298773690078,
+      "grad_norm": 0.4676069915294647,
+      "learning_rate": 8e-05,
+      "loss": 1.9581,
+      "step": 1535
+    },
+    {
+      "epoch": 0.08561872909698996,
+      "grad_norm": 0.45873093605041504,
+      "learning_rate": 8e-05,
+      "loss": 1.9127,
+      "step": 1536
+    },
+    {
+      "epoch": 0.08567447045707915,
+      "grad_norm": 0.44906124472618103,
+      "learning_rate": 8e-05,
+      "loss": 1.6728,
+      "step": 1537
+    },
+    {
+      "epoch": 0.08573021181716833,
+      "grad_norm": 0.4520525634288788,
+      "learning_rate": 8e-05,
+      "loss": 1.6776,
+      "step": 1538
+    },
+    {
+      "epoch": 0.08578595317725753,
+      "grad_norm": 0.4146431088447571,
+      "learning_rate": 8e-05,
+      "loss": 1.6065,
+      "step": 1539
+    },
+    {
+      "epoch": 0.0858416945373467,
+      "grad_norm": 0.4170554578304291,
+      "learning_rate": 8e-05,
+      "loss": 1.6201,
+      "step": 1540
+    },
+    {
+      "epoch": 0.0858974358974359,
+      "grad_norm": 0.43560776114463806,
+      "learning_rate": 8e-05,
+      "loss": 1.4419,
+      "step": 1541
+    },
+    {
+      "epoch": 0.08595317725752509,
+      "grad_norm": 0.4761258661746979,
+      "learning_rate": 8e-05,
+      "loss": 1.8902,
+      "step": 1542
+    },
+    {
+      "epoch": 0.08600891861761427,
+      "grad_norm": 0.456225723028183,
+      "learning_rate": 8e-05,
+      "loss": 1.84,
+      "step": 1543
+    },
+    {
+      "epoch": 0.08606465997770346,
+      "grad_norm": 0.39672479033470154,
+      "learning_rate": 8e-05,
+      "loss": 1.5196,
+      "step": 1544
+    },
+    {
+      "epoch": 0.08612040133779264,
+      "grad_norm": 0.3968527019023895,
+      "learning_rate": 8e-05,
+      "loss": 1.5071,
+      "step": 1545
+    },
+    {
+      "epoch": 0.08617614269788183,
+      "grad_norm": 0.46591445803642273,
+      "learning_rate": 8e-05,
+      "loss": 1.8287,
+      "step": 1546
+    },
+    {
+      "epoch": 0.08623188405797101,
+      "grad_norm": 0.41506892442703247,
+      "learning_rate": 8e-05,
+      "loss": 1.4293,
+      "step": 1547
+    },
+    {
+      "epoch": 0.0862876254180602,
+      "grad_norm": 0.4221274256706238,
+      "learning_rate": 8e-05,
+      "loss": 1.7141,
+      "step": 1548
+    },
+    {
+      "epoch": 0.08634336677814938,
+      "grad_norm": 0.42595311999320984,
+      "learning_rate": 8e-05,
+      "loss": 1.7241,
+      "step": 1549
+    },
+    {
+      "epoch": 0.08639910813823858,
+      "grad_norm": 0.4534299373626709,
+      "learning_rate": 8e-05,
+      "loss": 1.8307,
+      "step": 1550
+    },
+    {
+      "epoch": 0.08645484949832775,
+      "grad_norm": 0.4355649948120117,
+      "learning_rate": 8e-05,
+      "loss": 1.717,
+      "step": 1551
+    },
+    {
+      "epoch": 0.08651059085841695,
+      "grad_norm": 0.4492719769477844,
+      "learning_rate": 8e-05,
+      "loss": 1.7433,
+      "step": 1552
+    },
+    {
+      "epoch": 0.08656633221850613,
+      "grad_norm": 0.4739130437374115,
+      "learning_rate": 8e-05,
+      "loss": 1.929,
+      "step": 1553
+    },
+    {
+      "epoch": 0.08662207357859532,
+      "grad_norm": 0.4500783681869507,
+      "learning_rate": 8e-05,
+      "loss": 1.7929,
+      "step": 1554
+    },
+    {
+      "epoch": 0.0866778149386845,
+      "grad_norm": 0.46833619475364685,
+      "learning_rate": 8e-05,
+      "loss": 1.7378,
+      "step": 1555
+    },
+    {
+      "epoch": 0.08673355629877369,
+      "grad_norm": 0.43054115772247314,
+      "learning_rate": 8e-05,
+      "loss": 1.6093,
+      "step": 1556
+    },
+    {
+      "epoch": 0.08678929765886288,
+      "grad_norm": 0.43833228945732117,
+      "learning_rate": 8e-05,
+      "loss": 1.7848,
+      "step": 1557
+    },
+    {
+      "epoch": 0.08684503901895206,
+      "grad_norm": 0.4455707371234894,
+      "learning_rate": 8e-05,
+      "loss": 1.8429,
+      "step": 1558
+    },
+    {
+      "epoch": 0.08690078037904125,
+      "grad_norm": 0.42907917499542236,
+      "learning_rate": 8e-05,
+      "loss": 1.6299,
+      "step": 1559
+    },
+    {
+      "epoch": 0.08695652173913043,
+      "grad_norm": 0.4292406141757965,
+      "learning_rate": 8e-05,
+      "loss": 1.7396,
+      "step": 1560
+    },
+    {
+      "epoch": 0.08701226309921962,
+      "grad_norm": 0.45911526679992676,
+      "learning_rate": 8e-05,
+      "loss": 1.8727,
+      "step": 1561
+    },
+    {
+      "epoch": 0.0870680044593088,
+      "grad_norm": 0.47614502906799316,
+      "learning_rate": 8e-05,
+      "loss": 1.7738,
+      "step": 1562
+    },
+    {
+      "epoch": 0.087123745819398,
+      "grad_norm": 0.4749417006969452,
+      "learning_rate": 8e-05,
+      "loss": 1.9041,
+      "step": 1563
+    },
+    {
+      "epoch": 0.08717948717948718,
+      "grad_norm": 0.4312790036201477,
+      "learning_rate": 8e-05,
+      "loss": 1.7095,
+      "step": 1564
+    },
+    {
+      "epoch": 0.08723522853957637,
+      "grad_norm": 0.4703930616378784,
+      "learning_rate": 8e-05,
+      "loss": 1.6906,
+      "step": 1565
+    },
+    {
+      "epoch": 0.08729096989966555,
+      "grad_norm": 0.480101615190506,
+      "learning_rate": 8e-05,
+      "loss": 1.7722,
+      "step": 1566
+    },
+    {
+      "epoch": 0.08734671125975474,
+      "grad_norm": 0.42672696709632874,
+      "learning_rate": 8e-05,
+      "loss": 1.6706,
+      "step": 1567
+    },
+    {
+      "epoch": 0.08740245261984392,
+      "grad_norm": 0.42830973863601685,
+      "learning_rate": 8e-05,
+      "loss": 1.6612,
+      "step": 1568
+    },
+    {
+      "epoch": 0.08745819397993311,
+      "grad_norm": 0.44229260087013245,
+      "learning_rate": 8e-05,
+      "loss": 1.7886,
+      "step": 1569
+    },
+    {
+      "epoch": 0.0875139353400223,
+      "grad_norm": 0.445111483335495,
+      "learning_rate": 8e-05,
+      "loss": 1.3896,
+      "step": 1570
+    },
+    {
+      "epoch": 0.08756967670011148,
+      "grad_norm": 0.47544875741004944,
+      "learning_rate": 8e-05,
+      "loss": 1.9077,
+      "step": 1571
+    },
+    {
+      "epoch": 0.08762541806020067,
+      "grad_norm": 0.4429614543914795,
+      "learning_rate": 8e-05,
+      "loss": 1.9828,
+      "step": 1572
+    },
+    {
+      "epoch": 0.08768115942028985,
+      "grad_norm": 0.4450688064098358,
+      "learning_rate": 8e-05,
+      "loss": 1.801,
+      "step": 1573
+    },
+    {
+      "epoch": 0.08773690078037905,
+      "grad_norm": 0.4217309355735779,
+      "learning_rate": 8e-05,
+      "loss": 1.543,
+      "step": 1574
+    },
+    {
+      "epoch": 0.08779264214046822,
+      "grad_norm": 0.43825316429138184,
+      "learning_rate": 8e-05,
+      "loss": 1.8347,
+      "step": 1575
+    },
+    {
+      "epoch": 0.08784838350055742,
+      "grad_norm": 0.44539615511894226,
+      "learning_rate": 8e-05,
+      "loss": 1.8869,
+      "step": 1576
+    },
+    {
+      "epoch": 0.0879041248606466,
+      "grad_norm": 0.426485151052475,
+      "learning_rate": 8e-05,
+      "loss": 1.6711,
+      "step": 1577
+    },
+    {
+      "epoch": 0.08795986622073579,
+      "grad_norm": 0.451470285654068,
+      "learning_rate": 8e-05,
+      "loss": 1.6693,
+      "step": 1578
+    },
+    {
+      "epoch": 0.08801560758082497,
+      "grad_norm": 0.43202123045921326,
+      "learning_rate": 8e-05,
+      "loss": 1.4994,
+      "step": 1579
+    },
+    {
+      "epoch": 0.08807134894091416,
+      "grad_norm": 0.4591750502586365,
+      "learning_rate": 8e-05,
+      "loss": 1.9642,
+      "step": 1580
+    },
+    {
+      "epoch": 0.08812709030100334,
+      "grad_norm": 0.43216124176979065,
+      "learning_rate": 8e-05,
+      "loss": 1.8251,
+      "step": 1581
+    },
+    {
+      "epoch": 0.08818283166109253,
+      "grad_norm": 0.4440372586250305,
+      "learning_rate": 8e-05,
+      "loss": 1.7201,
+      "step": 1582
+    },
+    {
+      "epoch": 0.08823857302118172,
+      "grad_norm": 0.46146702766418457,
+      "learning_rate": 8e-05,
+      "loss": 1.845,
+      "step": 1583
+    },
+    {
+      "epoch": 0.0882943143812709,
+      "grad_norm": 0.44598087668418884,
+      "learning_rate": 8e-05,
+      "loss": 1.5421,
+      "step": 1584
+    },
+    {
+      "epoch": 0.0883500557413601,
+      "grad_norm": 0.4281664788722992,
+      "learning_rate": 8e-05,
+      "loss": 1.7447,
+      "step": 1585
+    },
+    {
+      "epoch": 0.08840579710144927,
+      "grad_norm": 0.47506773471832275,
+      "learning_rate": 8e-05,
+      "loss": 1.7401,
+      "step": 1586
+    },
+    {
+      "epoch": 0.08846153846153847,
+      "grad_norm": 0.4095071256160736,
+      "learning_rate": 8e-05,
+      "loss": 1.5696,
+      "step": 1587
+    },
+    {
+      "epoch": 0.08851727982162765,
+      "grad_norm": 0.46955305337905884,
+      "learning_rate": 8e-05,
+      "loss": 2.009,
+      "step": 1588
+    },
+    {
+      "epoch": 0.08857302118171684,
+      "grad_norm": 0.47986504435539246,
+      "learning_rate": 8e-05,
+      "loss": 1.8714,
+      "step": 1589
+    },
+    {
+      "epoch": 0.08862876254180602,
+      "grad_norm": 0.5251606702804565,
+      "learning_rate": 8e-05,
+      "loss": 1.9946,
+      "step": 1590
+    },
+    {
+      "epoch": 0.08868450390189521,
+      "grad_norm": 0.46084538102149963,
+      "learning_rate": 8e-05,
+      "loss": 1.7821,
+      "step": 1591
+    },
+    {
+      "epoch": 0.08874024526198439,
+      "grad_norm": 0.4135066866874695,
+      "learning_rate": 8e-05,
+      "loss": 1.6247,
+      "step": 1592
+    },
+    {
+      "epoch": 0.08879598662207358,
+      "grad_norm": 0.49193426966667175,
+      "learning_rate": 8e-05,
+      "loss": 2.0588,
+      "step": 1593
+    },
+    {
+      "epoch": 0.08885172798216276,
+      "grad_norm": 0.4328749179840088,
+      "learning_rate": 8e-05,
+      "loss": 1.7467,
+      "step": 1594
+    },
+    {
+      "epoch": 0.08890746934225195,
+      "grad_norm": 0.43832674622535706,
+      "learning_rate": 8e-05,
+      "loss": 1.7046,
+      "step": 1595
+    },
+    {
+      "epoch": 0.08896321070234113,
+      "grad_norm": 0.4430064558982849,
+      "learning_rate": 8e-05,
+      "loss": 1.7493,
+      "step": 1596
+    },
+    {
+      "epoch": 0.08901895206243032,
+      "grad_norm": 0.42061346769332886,
+      "learning_rate": 8e-05,
+      "loss": 1.6184,
+      "step": 1597
+    },
+    {
+      "epoch": 0.08907469342251952,
+      "grad_norm": 0.4191155731678009,
+      "learning_rate": 8e-05,
+      "loss": 1.6722,
+      "step": 1598
+    },
+    {
+      "epoch": 0.0891304347826087,
+      "grad_norm": 0.46061015129089355,
+      "learning_rate": 8e-05,
+      "loss": 1.886,
+      "step": 1599
+    },
+    {
+      "epoch": 0.08918617614269789,
+      "grad_norm": 0.42885342240333557,
+      "learning_rate": 8e-05,
+      "loss": 1.6767,
+      "step": 1600
+    },
+    {
+      "epoch": 0.08924191750278707,
+      "grad_norm": 0.4492940902709961,
+      "learning_rate": 8e-05,
+      "loss": 1.905,
+      "step": 1601
+    },
+    {
+      "epoch": 0.08929765886287626,
+      "grad_norm": 0.4751074016094208,
+      "learning_rate": 8e-05,
+      "loss": 1.7845,
+      "step": 1602
+    },
+    {
+      "epoch": 0.08935340022296544,
+      "grad_norm": 0.4468177855014801,
+      "learning_rate": 8e-05,
+      "loss": 1.906,
+      "step": 1603
+    },
+    {
+      "epoch": 0.08940914158305463,
+      "grad_norm": 0.43063390254974365,
+      "learning_rate": 8e-05,
+      "loss": 1.8162,
+      "step": 1604
+    },
+    {
+      "epoch": 0.08946488294314381,
+      "grad_norm": 0.43314266204833984,
+      "learning_rate": 8e-05,
+      "loss": 1.6541,
+      "step": 1605
+    },
+    {
+      "epoch": 0.089520624303233,
+      "grad_norm": 0.4799504578113556,
+      "learning_rate": 8e-05,
+      "loss": 1.7615,
+      "step": 1606
+    },
+    {
+      "epoch": 0.08957636566332218,
+      "grad_norm": 0.45094433426856995,
+      "learning_rate": 8e-05,
+      "loss": 1.7955,
+      "step": 1607
+    },
+    {
+      "epoch": 0.08963210702341137,
+      "grad_norm": 0.41243717074394226,
+      "learning_rate": 8e-05,
+      "loss": 1.7514,
+      "step": 1608
+    },
+    {
+      "epoch": 0.08968784838350055,
+      "grad_norm": 0.45868125557899475,
+      "learning_rate": 8e-05,
+      "loss": 1.7971,
+      "step": 1609
+    },
+    {
+      "epoch": 0.08974358974358974,
+      "grad_norm": 0.4831272065639496,
+      "learning_rate": 8e-05,
+      "loss": 1.8472,
+      "step": 1610
+    },
+    {
+      "epoch": 0.08979933110367894,
+      "grad_norm": 0.43193748593330383,
+      "learning_rate": 8e-05,
+      "loss": 1.589,
+      "step": 1611
+    },
+    {
+      "epoch": 0.08985507246376812,
+      "grad_norm": 0.4636029303073883,
+      "learning_rate": 8e-05,
+      "loss": 1.7388,
+      "step": 1612
+    },
+    {
+      "epoch": 0.08991081382385731,
+      "grad_norm": 0.40288928151130676,
+      "learning_rate": 8e-05,
+      "loss": 1.4132,
+      "step": 1613
+    },
+    {
+      "epoch": 0.08996655518394649,
+      "grad_norm": 0.44026824831962585,
+      "learning_rate": 8e-05,
+      "loss": 1.684,
+      "step": 1614
+    },
+    {
+      "epoch": 0.09002229654403568,
+      "grad_norm": 0.4749715328216553,
+      "learning_rate": 8e-05,
+      "loss": 1.8288,
+      "step": 1615
+    },
+    {
+      "epoch": 0.09007803790412486,
+      "grad_norm": 0.4861137568950653,
+      "learning_rate": 8e-05,
+      "loss": 1.7672,
+      "step": 1616
+    },
+    {
+      "epoch": 0.09013377926421405,
+      "grad_norm": 0.4515342712402344,
+      "learning_rate": 8e-05,
+      "loss": 1.6005,
+      "step": 1617
+    },
+    {
+      "epoch": 0.09018952062430323,
+      "grad_norm": 0.4273092448711395,
+      "learning_rate": 8e-05,
+      "loss": 1.5918,
+      "step": 1618
+    },
+    {
+      "epoch": 0.09024526198439242,
+      "grad_norm": 0.43101832270622253,
+      "learning_rate": 8e-05,
+      "loss": 1.8811,
+      "step": 1619
+    },
+    {
+      "epoch": 0.0903010033444816,
+      "grad_norm": 0.4761602580547333,
+      "learning_rate": 8e-05,
+      "loss": 1.8548,
+      "step": 1620
+    },
+    {
+      "epoch": 0.0903567447045708,
+      "grad_norm": 0.4239153563976288,
+      "learning_rate": 8e-05,
+      "loss": 1.7455,
+      "step": 1621
+    },
+    {
+      "epoch": 0.09041248606465997,
+      "grad_norm": 0.44429951906204224,
+      "learning_rate": 8e-05,
+      "loss": 1.8431,
+      "step": 1622
+    },
+    {
+      "epoch": 0.09046822742474916,
+      "grad_norm": 0.4681923985481262,
+      "learning_rate": 8e-05,
+      "loss": 1.7589,
+      "step": 1623
+    },
+    {
+      "epoch": 0.09052396878483834,
+      "grad_norm": 0.41767820715904236,
+      "learning_rate": 8e-05,
+      "loss": 1.7157,
+      "step": 1624
+    },
+    {
+      "epoch": 0.09057971014492754,
+      "grad_norm": 0.6465441584587097,
+      "learning_rate": 8e-05,
+      "loss": 1.6272,
+      "step": 1625
+    },
+    {
+      "epoch": 0.09063545150501673,
+      "grad_norm": 0.48799723386764526,
+      "learning_rate": 8e-05,
+      "loss": 1.7972,
+      "step": 1626
+    },
+    {
+      "epoch": 0.09069119286510591,
+      "grad_norm": 0.45866358280181885,
+      "learning_rate": 8e-05,
+      "loss": 1.9811,
+      "step": 1627
+    },
+    {
+      "epoch": 0.0907469342251951,
+      "grad_norm": 0.4481167197227478,
+      "learning_rate": 8e-05,
+      "loss": 1.7685,
+      "step": 1628
+    },
+    {
+      "epoch": 0.09080267558528428,
+      "grad_norm": 0.4584820568561554,
+      "learning_rate": 8e-05,
+      "loss": 1.7461,
+      "step": 1629
+    },
+    {
+      "epoch": 0.09085841694537347,
+      "grad_norm": 0.45503467321395874,
+      "learning_rate": 8e-05,
+      "loss": 1.6809,
+      "step": 1630
+    },
+    {
+      "epoch": 0.09091415830546265,
+      "grad_norm": 0.4234899878501892,
+      "learning_rate": 8e-05,
+      "loss": 1.5796,
+      "step": 1631
+    },
+    {
+      "epoch": 0.09096989966555184,
+      "grad_norm": 0.47477665543556213,
+      "learning_rate": 8e-05,
+      "loss": 1.8594,
+      "step": 1632
+    },
+    {
+      "epoch": 0.09102564102564102,
+      "grad_norm": 0.44016775488853455,
+      "learning_rate": 8e-05,
+      "loss": 1.6832,
+      "step": 1633
+    },
+    {
+      "epoch": 0.09108138238573021,
+      "grad_norm": 0.4387587606906891,
+      "learning_rate": 8e-05,
+      "loss": 1.7215,
+      "step": 1634
+    },
+    {
+      "epoch": 0.09113712374581939,
+      "grad_norm": 0.4861997067928314,
+      "learning_rate": 8e-05,
+      "loss": 1.8869,
+      "step": 1635
+    },
+    {
+      "epoch": 0.09119286510590859,
+      "grad_norm": 0.42896974086761475,
+      "learning_rate": 8e-05,
+      "loss": 1.5212,
+      "step": 1636
+    },
+    {
+      "epoch": 0.09124860646599776,
+      "grad_norm": 0.43611395359039307,
+      "learning_rate": 8e-05,
+      "loss": 1.7206,
+      "step": 1637
+    },
+    {
+      "epoch": 0.09130434782608696,
+      "grad_norm": 0.45591825246810913,
+      "learning_rate": 8e-05,
+      "loss": 1.8414,
+      "step": 1638
+    },
+    {
+      "epoch": 0.09136008918617615,
+      "grad_norm": 0.40458470582962036,
+      "learning_rate": 8e-05,
+      "loss": 1.5752,
+      "step": 1639
+    },
+    {
+      "epoch": 0.09141583054626533,
+      "grad_norm": 0.43076464533805847,
+      "learning_rate": 8e-05,
+      "loss": 1.7437,
+      "step": 1640
+    },
+    {
+      "epoch": 0.09147157190635452,
+      "grad_norm": 0.4555833637714386,
+      "learning_rate": 8e-05,
+      "loss": 1.6524,
+      "step": 1641
+    },
+    {
+      "epoch": 0.0915273132664437,
+      "grad_norm": 0.4160289168357849,
+      "learning_rate": 8e-05,
+      "loss": 1.4798,
+      "step": 1642
+    },
+    {
+      "epoch": 0.09158305462653289,
+      "grad_norm": 0.4631211459636688,
+      "learning_rate": 8e-05,
+      "loss": 1.779,
+      "step": 1643
+    },
+    {
+      "epoch": 0.09163879598662207,
+      "grad_norm": 0.44678810238838196,
+      "learning_rate": 8e-05,
+      "loss": 1.6644,
+      "step": 1644
+    },
+    {
+      "epoch": 0.09169453734671126,
+      "grad_norm": 0.45467978715896606,
+      "learning_rate": 8e-05,
+      "loss": 1.7205,
+      "step": 1645
+    },
+    {
+      "epoch": 0.09175027870680044,
+      "grad_norm": 0.4594002068042755,
+      "learning_rate": 8e-05,
+      "loss": 1.7903,
+      "step": 1646
+    },
+    {
+      "epoch": 0.09180602006688963,
+      "grad_norm": 0.4318181872367859,
+      "learning_rate": 8e-05,
+      "loss": 1.5232,
+      "step": 1647
+    },
+    {
+      "epoch": 0.09186176142697881,
+      "grad_norm": 0.4873585104942322,
+      "learning_rate": 8e-05,
+      "loss": 2.0432,
+      "step": 1648
+    },
+    {
+      "epoch": 0.091917502787068,
+      "grad_norm": 0.4591080844402313,
+      "learning_rate": 8e-05,
+      "loss": 1.8974,
+      "step": 1649
+    },
+    {
+      "epoch": 0.09197324414715718,
+      "grad_norm": 0.45263177156448364,
+      "learning_rate": 8e-05,
+      "loss": 1.8263,
+      "step": 1650
+    },
+    {
+      "epoch": 0.09202898550724638,
+      "grad_norm": 0.6010552644729614,
+      "learning_rate": 8e-05,
+      "loss": 1.7563,
+      "step": 1651
+    },
+    {
+      "epoch": 0.09208472686733556,
+      "grad_norm": 0.49455901980400085,
+      "learning_rate": 8e-05,
+      "loss": 1.7556,
+      "step": 1652
+    },
+    {
+      "epoch": 0.09214046822742475,
+      "grad_norm": 0.4340011179447174,
+      "learning_rate": 8e-05,
+      "loss": 1.6308,
+      "step": 1653
+    },
+    {
+      "epoch": 0.09219620958751394,
+      "grad_norm": 0.4920664429664612,
+      "learning_rate": 8e-05,
+      "loss": 1.8942,
+      "step": 1654
+    },
+    {
+      "epoch": 0.09225195094760312,
+      "grad_norm": 0.4639732241630554,
+      "learning_rate": 8e-05,
+      "loss": 1.9255,
+      "step": 1655
+    },
+    {
+      "epoch": 0.09230769230769231,
+      "grad_norm": 0.4229355752468109,
+      "learning_rate": 8e-05,
+      "loss": 1.7046,
+      "step": 1656
+    },
+    {
+      "epoch": 0.09236343366778149,
+      "grad_norm": 0.42051416635513306,
+      "learning_rate": 8e-05,
+      "loss": 1.7474,
+      "step": 1657
+    },
+    {
+      "epoch": 0.09241917502787068,
+      "grad_norm": 0.409633070230484,
+      "learning_rate": 8e-05,
+      "loss": 1.9147,
+      "step": 1658
+    },
+    {
+      "epoch": 0.09247491638795986,
+      "grad_norm": 0.4497150480747223,
+      "learning_rate": 8e-05,
+      "loss": 1.6605,
+      "step": 1659
+    },
+    {
+      "epoch": 0.09253065774804906,
+      "grad_norm": 0.42622387409210205,
+      "learning_rate": 8e-05,
+      "loss": 1.6316,
+      "step": 1660
+    },
+    {
+      "epoch": 0.09258639910813823,
+      "grad_norm": 0.5611870288848877,
+      "learning_rate": 8e-05,
+      "loss": 1.8821,
+      "step": 1661
+    },
+    {
+      "epoch": 0.09264214046822743,
+      "grad_norm": 0.5077276229858398,
+      "learning_rate": 8e-05,
+      "loss": 1.6538,
+      "step": 1662
+    },
+    {
+      "epoch": 0.0926978818283166,
+      "grad_norm": 0.47354525327682495,
+      "learning_rate": 8e-05,
+      "loss": 1.7993,
+      "step": 1663
+    },
+    {
+      "epoch": 0.0927536231884058,
+      "grad_norm": 0.4800141751766205,
+      "learning_rate": 8e-05,
+      "loss": 1.8998,
+      "step": 1664
+    },
+    {
+      "epoch": 0.09280936454849498,
+      "grad_norm": 0.47717806696891785,
+      "learning_rate": 8e-05,
+      "loss": 1.8321,
+      "step": 1665
+    },
+    {
+      "epoch": 0.09286510590858417,
+      "grad_norm": 0.44892174005508423,
+      "learning_rate": 8e-05,
+      "loss": 1.4733,
+      "step": 1666
+    },
+    {
+      "epoch": 0.09292084726867336,
+      "grad_norm": 0.48279869556427,
+      "learning_rate": 8e-05,
+      "loss": 1.7329,
+      "step": 1667
+    },
+    {
+      "epoch": 0.09297658862876254,
+      "grad_norm": 0.4728907346725464,
+      "learning_rate": 8e-05,
+      "loss": 1.803,
+      "step": 1668
+    },
+    {
+      "epoch": 0.09303232998885173,
+      "grad_norm": 0.4132857918739319,
+      "learning_rate": 8e-05,
+      "loss": 1.6814,
+      "step": 1669
+    },
+    {
+      "epoch": 0.09308807134894091,
+      "grad_norm": 0.4486645460128784,
+      "learning_rate": 8e-05,
+      "loss": 1.5777,
+      "step": 1670
+    },
+    {
+      "epoch": 0.0931438127090301,
+      "grad_norm": 0.47105464339256287,
+      "learning_rate": 8e-05,
+      "loss": 1.7778,
+      "step": 1671
+    },
+    {
+      "epoch": 0.09319955406911928,
+      "grad_norm": 0.4360397160053253,
+      "learning_rate": 8e-05,
+      "loss": 1.7292,
+      "step": 1672
+    },
+    {
+      "epoch": 0.09325529542920848,
+      "grad_norm": 0.3917115330696106,
+      "learning_rate": 8e-05,
+      "loss": 1.728,
+      "step": 1673
+    },
+    {
+      "epoch": 0.09331103678929765,
+      "grad_norm": 0.40846002101898193,
+      "learning_rate": 8e-05,
+      "loss": 1.6767,
+      "step": 1674
+    },
+    {
+      "epoch": 0.09336677814938685,
+      "grad_norm": 0.43595507740974426,
+      "learning_rate": 8e-05,
+      "loss": 1.6926,
+      "step": 1675
+    },
+    {
+      "epoch": 0.09342251950947603,
+      "grad_norm": 0.4381343424320221,
+      "learning_rate": 8e-05,
+      "loss": 1.6388,
+      "step": 1676
+    },
+    {
+      "epoch": 0.09347826086956522,
+      "grad_norm": 0.4457002878189087,
+      "learning_rate": 8e-05,
+      "loss": 1.7086,
+      "step": 1677
+    },
+    {
+      "epoch": 0.0935340022296544,
+      "grad_norm": 0.44507187604904175,
+      "learning_rate": 8e-05,
+      "loss": 1.553,
+      "step": 1678
+    },
+    {
+      "epoch": 0.09358974358974359,
+      "grad_norm": 0.46890974044799805,
+      "learning_rate": 8e-05,
+      "loss": 1.8152,
+      "step": 1679
+    },
+    {
+      "epoch": 0.09364548494983277,
+      "grad_norm": 0.5010303258895874,
+      "learning_rate": 8e-05,
+      "loss": 1.9007,
+      "step": 1680
+    },
+    {
+      "epoch": 0.09370122630992196,
+      "grad_norm": 0.44895729422569275,
+      "learning_rate": 8e-05,
+      "loss": 1.6811,
+      "step": 1681
+    },
+    {
+      "epoch": 0.09375696767001115,
+      "grad_norm": 0.44381895661354065,
+      "learning_rate": 8e-05,
+      "loss": 1.7254,
+      "step": 1682
+    },
+    {
+      "epoch": 0.09381270903010033,
+      "grad_norm": 0.42607060074806213,
+      "learning_rate": 8e-05,
+      "loss": 1.6745,
+      "step": 1683
+    },
+    {
+      "epoch": 0.09386845039018953,
+      "grad_norm": 0.4664064347743988,
+      "learning_rate": 8e-05,
+      "loss": 1.8605,
+      "step": 1684
+    },
+    {
+      "epoch": 0.0939241917502787,
+      "grad_norm": 0.4705875813961029,
+      "learning_rate": 8e-05,
+      "loss": 1.9974,
+      "step": 1685
+    },
+    {
+      "epoch": 0.0939799331103679,
+      "grad_norm": 0.4623195230960846,
+      "learning_rate": 8e-05,
+      "loss": 1.6875,
+      "step": 1686
+    },
+    {
+      "epoch": 0.09403567447045708,
+      "grad_norm": 0.4435681998729706,
+      "learning_rate": 8e-05,
+      "loss": 1.7876,
+      "step": 1687
+    },
+    {
+      "epoch": 0.09409141583054627,
+      "grad_norm": 0.4290086627006531,
+      "learning_rate": 8e-05,
+      "loss": 1.7866,
+      "step": 1688
+    },
+    {
+      "epoch": 0.09414715719063545,
+      "grad_norm": 0.47279369831085205,
+      "learning_rate": 8e-05,
+      "loss": 1.8753,
+      "step": 1689
+    },
+    {
+      "epoch": 0.09420289855072464,
+      "grad_norm": 0.43134960532188416,
+      "learning_rate": 8e-05,
+      "loss": 1.6947,
+      "step": 1690
+    },
+    {
+      "epoch": 0.09425863991081382,
+      "grad_norm": 0.588997483253479,
+      "learning_rate": 8e-05,
+      "loss": 2.0056,
+      "step": 1691
+    },
+    {
+      "epoch": 0.09431438127090301,
+      "grad_norm": 0.48162195086479187,
+      "learning_rate": 8e-05,
+      "loss": 1.9277,
+      "step": 1692
+    },
+    {
+      "epoch": 0.09437012263099219,
+      "grad_norm": 0.42257192730903625,
+      "learning_rate": 8e-05,
+      "loss": 1.6565,
+      "step": 1693
+    },
+    {
+      "epoch": 0.09442586399108138,
+      "grad_norm": 0.41762158274650574,
+      "learning_rate": 8e-05,
+      "loss": 1.7561,
+      "step": 1694
+    },
+    {
+      "epoch": 0.09448160535117058,
+      "grad_norm": 0.44371649622917175,
+      "learning_rate": 8e-05,
+      "loss": 1.5394,
+      "step": 1695
+    },
+    {
+      "epoch": 0.09453734671125975,
+      "grad_norm": 0.4283082187175751,
+      "learning_rate": 8e-05,
+      "loss": 1.586,
+      "step": 1696
+    },
+    {
+      "epoch": 0.09459308807134895,
+      "grad_norm": 0.4454692304134369,
+      "learning_rate": 8e-05,
+      "loss": 1.6537,
+      "step": 1697
+    },
+    {
+      "epoch": 0.09464882943143813,
+      "grad_norm": 0.463825523853302,
+      "learning_rate": 8e-05,
+      "loss": 1.8293,
+      "step": 1698
+    },
+    {
+      "epoch": 0.09470457079152732,
+      "grad_norm": 0.41332781314849854,
+      "learning_rate": 8e-05,
+      "loss": 1.7457,
+      "step": 1699
+    },
+    {
+      "epoch": 0.0947603121516165,
+      "grad_norm": 0.42744752764701843,
+      "learning_rate": 8e-05,
+      "loss": 1.616,
+      "step": 1700
+    },
+    {
+      "epoch": 0.09481605351170569,
+      "grad_norm": 0.4430151581764221,
+      "learning_rate": 8e-05,
+      "loss": 1.7341,
+      "step": 1701
+    },
+    {
+      "epoch": 0.09487179487179487,
+      "grad_norm": 0.48058566451072693,
+      "learning_rate": 8e-05,
+      "loss": 1.8291,
+      "step": 1702
+    },
+    {
+      "epoch": 0.09492753623188406,
+      "grad_norm": 0.4263574481010437,
+      "learning_rate": 8e-05,
+      "loss": 1.6344,
+      "step": 1703
+    },
+    {
+      "epoch": 0.09498327759197324,
+      "grad_norm": 0.4660191237926483,
+      "learning_rate": 8e-05,
+      "loss": 1.7992,
+      "step": 1704
+    },
+    {
+      "epoch": 0.09503901895206243,
+      "grad_norm": 0.44191208481788635,
+      "learning_rate": 8e-05,
+      "loss": 1.7902,
+      "step": 1705
+    },
+    {
+      "epoch": 0.09509476031215161,
+      "grad_norm": 0.4195825159549713,
+      "learning_rate": 8e-05,
+      "loss": 1.5295,
+      "step": 1706
+    },
+    {
+      "epoch": 0.0951505016722408,
+      "grad_norm": 0.4643992483615875,
+      "learning_rate": 8e-05,
+      "loss": 1.8399,
+      "step": 1707
+    },
+    {
+      "epoch": 0.09520624303233,
+      "grad_norm": 0.4567433297634125,
+      "learning_rate": 8e-05,
+      "loss": 1.7278,
+      "step": 1708
+    },
+    {
+      "epoch": 0.09526198439241917,
+      "grad_norm": 0.4396064579486847,
+      "learning_rate": 8e-05,
+      "loss": 1.5918,
+      "step": 1709
+    },
+    {
+      "epoch": 0.09531772575250837,
+      "grad_norm": 0.47985053062438965,
+      "learning_rate": 8e-05,
+      "loss": 1.8334,
+      "step": 1710
+    },
+    {
+      "epoch": 0.09537346711259755,
+      "grad_norm": 0.46934187412261963,
+      "learning_rate": 8e-05,
+      "loss": 1.8587,
+      "step": 1711
+    },
+    {
+      "epoch": 0.09542920847268674,
+      "grad_norm": 0.4805198311805725,
+      "learning_rate": 8e-05,
+      "loss": 1.9085,
+      "step": 1712
+    },
+    {
+      "epoch": 0.09548494983277592,
+      "grad_norm": 0.47350266575813293,
+      "learning_rate": 8e-05,
+      "loss": 1.8195,
+      "step": 1713
+    },
+    {
+      "epoch": 0.09554069119286511,
+      "grad_norm": 0.46012142300605774,
+      "learning_rate": 8e-05,
+      "loss": 1.784,
+      "step": 1714
+    },
+    {
+      "epoch": 0.09559643255295429,
+      "grad_norm": 0.49588704109191895,
+      "learning_rate": 8e-05,
+      "loss": 1.9803,
+      "step": 1715
+    },
+    {
+      "epoch": 0.09565217391304348,
+      "grad_norm": 0.4539101719856262,
+      "learning_rate": 8e-05,
+      "loss": 1.8789,
+      "step": 1716
+    },
+    {
+      "epoch": 0.09570791527313266,
+      "grad_norm": 0.432472288608551,
+      "learning_rate": 8e-05,
+      "loss": 1.6145,
+      "step": 1717
+    },
+    {
+      "epoch": 0.09576365663322185,
+      "grad_norm": 0.47179466485977173,
+      "learning_rate": 8e-05,
+      "loss": 1.7848,
+      "step": 1718
+    },
+    {
+      "epoch": 0.09581939799331103,
+      "grad_norm": 0.42406386137008667,
+      "learning_rate": 8e-05,
+      "loss": 1.4585,
+      "step": 1719
+    },
+    {
+      "epoch": 0.09587513935340022,
+      "grad_norm": 0.442560613155365,
+      "learning_rate": 8e-05,
+      "loss": 1.7944,
+      "step": 1720
+    },
+    {
+      "epoch": 0.0959308807134894,
+      "grad_norm": 0.4633161425590515,
+      "learning_rate": 8e-05,
+      "loss": 1.7961,
+      "step": 1721
+    },
+    {
+      "epoch": 0.0959866220735786,
+      "grad_norm": 0.44542253017425537,
+      "learning_rate": 8e-05,
+      "loss": 1.6882,
+      "step": 1722
+    },
+    {
+      "epoch": 0.09604236343366779,
+      "grad_norm": 0.6311537623405457,
+      "learning_rate": 8e-05,
+      "loss": 1.7164,
+      "step": 1723
+    },
+    {
+      "epoch": 0.09609810479375697,
+      "grad_norm": 0.40386149287223816,
+      "learning_rate": 8e-05,
+      "loss": 1.6877,
+      "step": 1724
+    },
+    {
+      "epoch": 0.09615384615384616,
+      "grad_norm": 0.4860149621963501,
+      "learning_rate": 8e-05,
+      "loss": 1.7702,
+      "step": 1725
+    },
+    {
+      "epoch": 0.09620958751393534,
+      "grad_norm": 0.4561831057071686,
+      "learning_rate": 8e-05,
+      "loss": 1.7912,
+      "step": 1726
+    },
+    {
+      "epoch": 0.09626532887402453,
+      "grad_norm": 0.48729828000068665,
+      "learning_rate": 8e-05,
+      "loss": 1.816,
+      "step": 1727
+    },
+    {
+      "epoch": 0.09632107023411371,
+      "grad_norm": 0.4261293411254883,
+      "learning_rate": 8e-05,
+      "loss": 1.585,
+      "step": 1728
+    },
+    {
+      "epoch": 0.0963768115942029,
+      "grad_norm": 0.4322323799133301,
+      "learning_rate": 8e-05,
+      "loss": 1.5954,
+      "step": 1729
+    },
+    {
+      "epoch": 0.09643255295429208,
+      "grad_norm": 0.4254898130893707,
+      "learning_rate": 8e-05,
+      "loss": 1.5349,
+      "step": 1730
+    },
+    {
+      "epoch": 0.09648829431438127,
+      "grad_norm": 0.4253242611885071,
+      "learning_rate": 8e-05,
+      "loss": 1.8648,
+      "step": 1731
+    },
+    {
+      "epoch": 0.09654403567447045,
+      "grad_norm": 0.4364851117134094,
+      "learning_rate": 8e-05,
+      "loss": 1.5873,
+      "step": 1732
+    },
+    {
+      "epoch": 0.09659977703455964,
+      "grad_norm": 0.4467128813266754,
+      "learning_rate": 8e-05,
+      "loss": 1.8267,
+      "step": 1733
+    },
+    {
+      "epoch": 0.09665551839464882,
+      "grad_norm": 0.45185184478759766,
+      "learning_rate": 8e-05,
+      "loss": 1.604,
+      "step": 1734
+    },
+    {
+      "epoch": 0.09671125975473802,
+      "grad_norm": 0.4952625632286072,
+      "learning_rate": 8e-05,
+      "loss": 1.8941,
+      "step": 1735
+    },
+    {
+      "epoch": 0.09676700111482721,
+      "grad_norm": 0.4527432918548584,
+      "learning_rate": 8e-05,
+      "loss": 1.5787,
+      "step": 1736
+    },
+    {
+      "epoch": 0.09682274247491639,
+      "grad_norm": 0.5307126045227051,
+      "learning_rate": 8e-05,
+      "loss": 2.1902,
+      "step": 1737
+    },
+    {
+      "epoch": 0.09687848383500558,
+      "grad_norm": 0.45969170331954956,
+      "learning_rate": 8e-05,
+      "loss": 1.8296,
+      "step": 1738
+    },
+    {
+      "epoch": 0.09693422519509476,
+      "grad_norm": 0.4471434950828552,
+      "learning_rate": 8e-05,
+      "loss": 1.711,
+      "step": 1739
+    },
+    {
+      "epoch": 0.09698996655518395,
+      "grad_norm": 0.45372387766838074,
+      "learning_rate": 8e-05,
+      "loss": 1.8327,
+      "step": 1740
+    },
+    {
+      "epoch": 0.09704570791527313,
+      "grad_norm": 0.4237399995326996,
+      "learning_rate": 8e-05,
+      "loss": 1.7176,
+      "step": 1741
+    },
+    {
+      "epoch": 0.09710144927536232,
+      "grad_norm": 0.4647175967693329,
+      "learning_rate": 8e-05,
+      "loss": 1.577,
+      "step": 1742
+    },
+    {
+      "epoch": 0.0971571906354515,
+      "grad_norm": 0.46522289514541626,
+      "learning_rate": 8e-05,
+      "loss": 1.6859,
+      "step": 1743
+    },
+    {
+      "epoch": 0.0972129319955407,
+      "grad_norm": 0.4213174879550934,
+      "learning_rate": 8e-05,
+      "loss": 1.6958,
+      "step": 1744
+    },
+    {
+      "epoch": 0.09726867335562987,
+      "grad_norm": 0.4828932583332062,
+      "learning_rate": 8e-05,
+      "loss": 1.7898,
+      "step": 1745
+    },
+    {
+      "epoch": 0.09732441471571907,
+      "grad_norm": 0.4179130494594574,
+      "learning_rate": 8e-05,
+      "loss": 1.4819,
+      "step": 1746
+    },
+    {
+      "epoch": 0.09738015607580824,
+      "grad_norm": 0.4496394097805023,
+      "learning_rate": 8e-05,
+      "loss": 1.7487,
+      "step": 1747
+    },
+    {
+      "epoch": 0.09743589743589744,
+      "grad_norm": 0.4822089374065399,
+      "learning_rate": 8e-05,
+      "loss": 1.7319,
+      "step": 1748
+    },
+    {
+      "epoch": 0.09749163879598662,
+      "grad_norm": 0.454473614692688,
+      "learning_rate": 8e-05,
+      "loss": 1.6915,
+      "step": 1749
+    },
+    {
+      "epoch": 0.09754738015607581,
+      "grad_norm": 0.46822261810302734,
+      "learning_rate": 8e-05,
+      "loss": 1.9073,
+      "step": 1750
+    },
+    {
+      "epoch": 0.097603121516165,
+      "grad_norm": 0.44990289211273193,
+      "learning_rate": 8e-05,
+      "loss": 1.7303,
+      "step": 1751
+    },
+    {
+      "epoch": 0.09765886287625418,
+      "grad_norm": 0.5409207344055176,
+      "learning_rate": 8e-05,
+      "loss": 1.7165,
+      "step": 1752
+    },
+    {
+      "epoch": 0.09771460423634337,
+      "grad_norm": 0.43510881066322327,
+      "learning_rate": 8e-05,
+      "loss": 1.6771,
+      "step": 1753
+    },
+    {
+      "epoch": 0.09777034559643255,
+      "grad_norm": 0.5424200892448425,
+      "learning_rate": 8e-05,
+      "loss": 2.2344,
+      "step": 1754
+    },
+    {
+      "epoch": 0.09782608695652174,
+      "grad_norm": 0.4397199749946594,
+      "learning_rate": 8e-05,
+      "loss": 1.701,
+      "step": 1755
+    },
+    {
+      "epoch": 0.09788182831661092,
+      "grad_norm": 0.40220770239830017,
+      "learning_rate": 8e-05,
+      "loss": 1.5287,
+      "step": 1756
+    },
+    {
+      "epoch": 0.09793756967670011,
+      "grad_norm": 0.4556962847709656,
+      "learning_rate": 8e-05,
+      "loss": 1.7051,
+      "step": 1757
+    },
+    {
+      "epoch": 0.0979933110367893,
+      "grad_norm": 0.4355781376361847,
+      "learning_rate": 8e-05,
+      "loss": 1.7602,
+      "step": 1758
+    },
+    {
+      "epoch": 0.09804905239687849,
+      "grad_norm": 0.41649800539016724,
+      "learning_rate": 8e-05,
+      "loss": 1.44,
+      "step": 1759
+    },
+    {
+      "epoch": 0.09810479375696766,
+      "grad_norm": 0.4674367606639862,
+      "learning_rate": 8e-05,
+      "loss": 1.7817,
+      "step": 1760
+    },
+    {
+      "epoch": 0.09816053511705686,
+      "grad_norm": 0.45049193501472473,
+      "learning_rate": 8e-05,
+      "loss": 1.6204,
+      "step": 1761
+    },
+    {
+      "epoch": 0.09821627647714604,
+      "grad_norm": 0.45521673560142517,
+      "learning_rate": 8e-05,
+      "loss": 1.7195,
+      "step": 1762
+    },
+    {
+      "epoch": 0.09827201783723523,
+      "grad_norm": 0.4201885759830475,
+      "learning_rate": 8e-05,
+      "loss": 1.3177,
+      "step": 1763
+    },
+    {
+      "epoch": 0.09832775919732442,
+      "grad_norm": 0.4933217763900757,
+      "learning_rate": 8e-05,
+      "loss": 1.8507,
+      "step": 1764
+    },
+    {
+      "epoch": 0.0983835005574136,
+      "grad_norm": 0.4303235411643982,
+      "learning_rate": 8e-05,
+      "loss": 1.7071,
+      "step": 1765
+    },
+    {
+      "epoch": 0.09843924191750279,
+      "grad_norm": 0.4998074769973755,
+      "learning_rate": 8e-05,
+      "loss": 1.5192,
+      "step": 1766
+    },
+    {
+      "epoch": 0.09849498327759197,
+      "grad_norm": 0.4623011648654938,
+      "learning_rate": 8e-05,
+      "loss": 1.8838,
+      "step": 1767
+    },
+    {
+      "epoch": 0.09855072463768116,
+      "grad_norm": 0.486713171005249,
+      "learning_rate": 8e-05,
+      "loss": 1.9141,
+      "step": 1768
+    },
+    {
+      "epoch": 0.09860646599777034,
+      "grad_norm": 0.5125125050544739,
+      "learning_rate": 8e-05,
+      "loss": 1.9007,
+      "step": 1769
+    },
+    {
+      "epoch": 0.09866220735785954,
+      "grad_norm": 0.4394649565219879,
+      "learning_rate": 8e-05,
+      "loss": 1.669,
+      "step": 1770
+    },
+    {
+      "epoch": 0.09871794871794871,
+      "grad_norm": 0.4292559325695038,
+      "learning_rate": 8e-05,
+      "loss": 1.6525,
+      "step": 1771
+    },
+    {
+      "epoch": 0.0987736900780379,
+      "grad_norm": 0.47850099205970764,
+      "learning_rate": 8e-05,
+      "loss": 1.5841,
+      "step": 1772
+    },
+    {
+      "epoch": 0.09882943143812709,
+      "grad_norm": 0.4640037715435028,
+      "learning_rate": 8e-05,
+      "loss": 1.8342,
+      "step": 1773
+    },
+    {
+      "epoch": 0.09888517279821628,
+      "grad_norm": 0.47437646985054016,
+      "learning_rate": 8e-05,
+      "loss": 1.9839,
+      "step": 1774
+    },
+    {
+      "epoch": 0.09894091415830546,
+      "grad_norm": 0.511099636554718,
+      "learning_rate": 8e-05,
+      "loss": 1.9201,
+      "step": 1775
+    },
+    {
+      "epoch": 0.09899665551839465,
+      "grad_norm": 0.42483294010162354,
+      "learning_rate": 8e-05,
+      "loss": 1.6935,
+      "step": 1776
+    },
+    {
+      "epoch": 0.09905239687848383,
+      "grad_norm": 0.4893593490123749,
+      "learning_rate": 8e-05,
+      "loss": 1.9715,
+      "step": 1777
+    },
+    {
+      "epoch": 0.09910813823857302,
+      "grad_norm": 0.47139880061149597,
+      "learning_rate": 8e-05,
+      "loss": 1.6563,
+      "step": 1778
+    },
+    {
+      "epoch": 0.09916387959866221,
+      "grad_norm": 0.4339061379432678,
+      "learning_rate": 8e-05,
+      "loss": 1.5654,
+      "step": 1779
+    },
+    {
+      "epoch": 0.09921962095875139,
+      "grad_norm": 0.4311170279979706,
+      "learning_rate": 8e-05,
+      "loss": 1.5791,
+      "step": 1780
+    },
+    {
+      "epoch": 0.09927536231884058,
+      "grad_norm": 0.42284849286079407,
+      "learning_rate": 8e-05,
+      "loss": 1.6309,
+      "step": 1781
+    },
+    {
+      "epoch": 0.09933110367892976,
+      "grad_norm": 0.442691832780838,
+      "learning_rate": 8e-05,
+      "loss": 1.8009,
+      "step": 1782
+    },
+    {
+      "epoch": 0.09938684503901896,
+      "grad_norm": 0.4467600882053375,
+      "learning_rate": 8e-05,
+      "loss": 1.6598,
+      "step": 1783
+    },
+    {
+      "epoch": 0.09944258639910813,
+      "grad_norm": 0.48540544509887695,
+      "learning_rate": 8e-05,
+      "loss": 1.7799,
+      "step": 1784
+    },
+    {
+      "epoch": 0.09949832775919733,
+      "grad_norm": 0.48900890350341797,
+      "learning_rate": 8e-05,
+      "loss": 1.8178,
+      "step": 1785
+    },
+    {
+      "epoch": 0.0995540691192865,
+      "grad_norm": 0.4849342107772827,
+      "learning_rate": 8e-05,
+      "loss": 1.5124,
+      "step": 1786
+    },
+    {
+      "epoch": 0.0996098104793757,
+      "grad_norm": 0.43086501955986023,
+      "learning_rate": 8e-05,
+      "loss": 1.4617,
+      "step": 1787
+    },
+    {
+      "epoch": 0.09966555183946488,
+      "grad_norm": 0.4691721200942993,
+      "learning_rate": 8e-05,
+      "loss": 1.8552,
+      "step": 1788
+    },
+    {
+      "epoch": 0.09972129319955407,
+      "grad_norm": 0.5271967649459839,
+      "learning_rate": 8e-05,
+      "loss": 1.8783,
+      "step": 1789
+    },
+    {
+      "epoch": 0.09977703455964325,
+      "grad_norm": 0.4975190758705139,
+      "learning_rate": 8e-05,
+      "loss": 1.7687,
+      "step": 1790
+    },
+    {
+      "epoch": 0.09983277591973244,
+      "grad_norm": 0.4852193593978882,
+      "learning_rate": 8e-05,
+      "loss": 1.9312,
+      "step": 1791
+    },
+    {
+      "epoch": 0.09988851727982163,
+      "grad_norm": 0.4707431495189667,
+      "learning_rate": 8e-05,
+      "loss": 1.7425,
+      "step": 1792
+    },
+    {
+      "epoch": 0.09994425863991081,
+      "grad_norm": 0.4759067893028259,
+      "learning_rate": 8e-05,
+      "loss": 1.6955,
+      "step": 1793
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.43870437145233154,
+      "learning_rate": 8e-05,
+      "loss": 1.4649,
+      "step": 1794
+    },
+    {
+      "epoch": 0.10005574136008918,
+      "grad_norm": 0.5789974331855774,
+      "learning_rate": 8e-05,
+      "loss": 1.6798,
+      "step": 1795
+    },
+    {
+      "epoch": 0.10011148272017838,
+      "grad_norm": 0.45335498452186584,
+      "learning_rate": 8e-05,
+      "loss": 1.6491,
+      "step": 1796
+    },
+    {
+      "epoch": 0.10016722408026756,
+      "grad_norm": 0.5454590320587158,
+      "learning_rate": 8e-05,
+      "loss": 1.7319,
+      "step": 1797
+    },
+    {
+      "epoch": 0.10022296544035675,
+      "grad_norm": 0.6508112549781799,
+      "learning_rate": 8e-05,
+      "loss": 1.9191,
+      "step": 1798
+    },
+    {
+      "epoch": 0.10027870680044593,
+      "grad_norm": 0.4346283972263336,
+      "learning_rate": 8e-05,
+      "loss": 1.5611,
+      "step": 1799
+    },
+    {
+      "epoch": 0.10033444816053512,
+      "grad_norm": 0.46036475896835327,
+      "learning_rate": 8e-05,
+      "loss": 1.8125,
+      "step": 1800
+    },
+    {
+      "epoch": 0.1003901895206243,
+      "grad_norm": 0.44356054067611694,
+      "learning_rate": 8e-05,
+      "loss": 1.6516,
+      "step": 1801
+    },
+    {
+      "epoch": 0.10044593088071349,
+      "grad_norm": 0.4230916500091553,
+      "learning_rate": 8e-05,
+      "loss": 1.6221,
+      "step": 1802
+    },
+    {
+      "epoch": 0.10050167224080267,
+      "grad_norm": 0.4426247775554657,
+      "learning_rate": 8e-05,
+      "loss": 1.7226,
+      "step": 1803
+    },
+    {
+      "epoch": 0.10055741360089186,
+      "grad_norm": 0.42851877212524414,
+      "learning_rate": 8e-05,
+      "loss": 1.6643,
+      "step": 1804
+    },
+    {
+      "epoch": 0.10061315496098104,
+      "grad_norm": 0.47624823451042175,
+      "learning_rate": 8e-05,
+      "loss": 1.8385,
+      "step": 1805
+    },
+    {
+      "epoch": 0.10066889632107023,
+      "grad_norm": 0.4784403145313263,
+      "learning_rate": 8e-05,
+      "loss": 1.7394,
+      "step": 1806
+    },
+    {
+      "epoch": 0.10072463768115943,
+      "grad_norm": 0.45466241240501404,
+      "learning_rate": 8e-05,
+      "loss": 1.6348,
+      "step": 1807
+    },
+    {
+      "epoch": 0.1007803790412486,
+      "grad_norm": 0.44868579506874084,
+      "learning_rate": 8e-05,
+      "loss": 1.834,
+      "step": 1808
+    },
+    {
+      "epoch": 0.1008361204013378,
+      "grad_norm": 0.4959905445575714,
+      "learning_rate": 8e-05,
+      "loss": 1.7847,
+      "step": 1809
+    },
+    {
+      "epoch": 0.10089186176142698,
+      "grad_norm": 0.4354304373264313,
+      "learning_rate": 8e-05,
+      "loss": 1.7289,
+      "step": 1810
+    },
+    {
+      "epoch": 0.10094760312151617,
+      "grad_norm": 0.4733140766620636,
+      "learning_rate": 8e-05,
+      "loss": 1.8648,
+      "step": 1811
+    },
+    {
+      "epoch": 0.10100334448160535,
+      "grad_norm": 0.4682197868824005,
+      "learning_rate": 8e-05,
+      "loss": 1.8402,
+      "step": 1812
+    },
+    {
+      "epoch": 0.10105908584169454,
+      "grad_norm": 0.4327491521835327,
+      "learning_rate": 8e-05,
+      "loss": 1.8034,
+      "step": 1813
+    },
+    {
+      "epoch": 0.10111482720178372,
+      "grad_norm": 0.4196842610836029,
+      "learning_rate": 8e-05,
+      "loss": 1.5267,
+      "step": 1814
+    },
+    {
+      "epoch": 0.10117056856187291,
+      "grad_norm": 0.4950825870037079,
+      "learning_rate": 8e-05,
+      "loss": 1.7586,
+      "step": 1815
+    },
+    {
+      "epoch": 0.10122630992196209,
+      "grad_norm": 0.47089883685112,
+      "learning_rate": 8e-05,
+      "loss": 1.6667,
+      "step": 1816
+    },
+    {
+      "epoch": 0.10128205128205128,
+      "grad_norm": 0.418155312538147,
+      "learning_rate": 8e-05,
+      "loss": 1.6951,
+      "step": 1817
+    },
+    {
+      "epoch": 0.10133779264214046,
+      "grad_norm": 0.4778137803077698,
+      "learning_rate": 8e-05,
+      "loss": 1.7777,
+      "step": 1818
+    },
+    {
+      "epoch": 0.10139353400222965,
+      "grad_norm": 0.44255295395851135,
+      "learning_rate": 8e-05,
+      "loss": 1.7593,
+      "step": 1819
+    },
+    {
+      "epoch": 0.10144927536231885,
+      "grad_norm": 0.43998026847839355,
+      "learning_rate": 8e-05,
+      "loss": 1.8034,
+      "step": 1820
+    },
+    {
+      "epoch": 0.10150501672240803,
+      "grad_norm": 0.4900084435939789,
+      "learning_rate": 8e-05,
+      "loss": 1.835,
+      "step": 1821
+    },
+    {
+      "epoch": 0.10156075808249722,
+      "grad_norm": 0.44791969656944275,
+      "learning_rate": 8e-05,
+      "loss": 1.6511,
+      "step": 1822
+    },
+    {
+      "epoch": 0.1016164994425864,
+      "grad_norm": 0.4674259126186371,
+      "learning_rate": 8e-05,
+      "loss": 1.7348,
+      "step": 1823
+    },
+    {
+      "epoch": 0.10167224080267559,
+      "grad_norm": 0.44596970081329346,
+      "learning_rate": 8e-05,
+      "loss": 1.5885,
+      "step": 1824
+    },
+    {
+      "epoch": 0.10172798216276477,
+      "grad_norm": 0.4487878084182739,
+      "learning_rate": 8e-05,
+      "loss": 1.7307,
+      "step": 1825
+    },
+    {
+      "epoch": 0.10178372352285396,
+      "grad_norm": 0.4508936107158661,
+      "learning_rate": 8e-05,
+      "loss": 1.8602,
+      "step": 1826
+    },
+    {
+      "epoch": 0.10183946488294314,
+      "grad_norm": 0.44692227244377136,
+      "learning_rate": 8e-05,
+      "loss": 1.6913,
+      "step": 1827
+    },
+    {
+      "epoch": 0.10189520624303233,
+      "grad_norm": 0.47166091203689575,
+      "learning_rate": 8e-05,
+      "loss": 1.641,
+      "step": 1828
+    },
+    {
+      "epoch": 0.10195094760312151,
+      "grad_norm": 0.43342047929763794,
+      "learning_rate": 8e-05,
+      "loss": 1.6667,
+      "step": 1829
+    },
+    {
+      "epoch": 0.1020066889632107,
+      "grad_norm": 0.514275848865509,
+      "learning_rate": 8e-05,
+      "loss": 1.9541,
+      "step": 1830
+    },
+    {
+      "epoch": 0.10206243032329988,
+      "grad_norm": 0.45009151101112366,
+      "learning_rate": 8e-05,
+      "loss": 1.7608,
+      "step": 1831
+    },
+    {
+      "epoch": 0.10211817168338908,
+      "grad_norm": 0.43719324469566345,
+      "learning_rate": 8e-05,
+      "loss": 1.7147,
+      "step": 1832
+    },
+    {
+      "epoch": 0.10217391304347827,
+      "grad_norm": 0.4204999804496765,
+      "learning_rate": 8e-05,
+      "loss": 1.6842,
+      "step": 1833
+    },
+    {
+      "epoch": 0.10222965440356745,
+      "grad_norm": 0.4463106095790863,
+      "learning_rate": 8e-05,
+      "loss": 1.6066,
+      "step": 1834
+    },
+    {
+      "epoch": 0.10228539576365664,
+      "grad_norm": 0.4450172781944275,
+      "learning_rate": 8e-05,
+      "loss": 1.6528,
+      "step": 1835
+    },
+    {
+      "epoch": 0.10234113712374582,
+      "grad_norm": 0.5039206743240356,
+      "learning_rate": 8e-05,
+      "loss": 1.8365,
+      "step": 1836
+    },
+    {
+      "epoch": 0.10239687848383501,
+      "grad_norm": 0.45737943053245544,
+      "learning_rate": 8e-05,
+      "loss": 1.6945,
+      "step": 1837
+    },
+    {
+      "epoch": 0.10245261984392419,
+      "grad_norm": 0.48224934935569763,
+      "learning_rate": 8e-05,
+      "loss": 1.5769,
+      "step": 1838
+    },
+    {
+      "epoch": 0.10250836120401338,
+      "grad_norm": 0.4501650333404541,
+      "learning_rate": 8e-05,
+      "loss": 1.732,
+      "step": 1839
+    },
+    {
+      "epoch": 0.10256410256410256,
+      "grad_norm": 0.45120200514793396,
+      "learning_rate": 8e-05,
+      "loss": 1.7743,
+      "step": 1840
+    },
+    {
+      "epoch": 0.10261984392419175,
+      "grad_norm": 0.5067091584205627,
+      "learning_rate": 8e-05,
+      "loss": 1.8807,
+      "step": 1841
+    },
+    {
+      "epoch": 0.10267558528428093,
+      "grad_norm": 0.4903699457645416,
+      "learning_rate": 8e-05,
+      "loss": 1.7831,
+      "step": 1842
+    },
+    {
+      "epoch": 0.10273132664437012,
+      "grad_norm": 0.4485124945640564,
+      "learning_rate": 8e-05,
+      "loss": 1.7676,
+      "step": 1843
+    },
+    {
+      "epoch": 0.1027870680044593,
+      "grad_norm": 0.49188125133514404,
+      "learning_rate": 8e-05,
+      "loss": 1.9028,
+      "step": 1844
+    },
+    {
+      "epoch": 0.1028428093645485,
+      "grad_norm": 0.45281144976615906,
+      "learning_rate": 8e-05,
+      "loss": 1.7229,
+      "step": 1845
+    },
+    {
+      "epoch": 0.10289855072463767,
+      "grad_norm": 0.48761826753616333,
+      "learning_rate": 8e-05,
+      "loss": 1.8395,
+      "step": 1846
+    },
+    {
+      "epoch": 0.10295429208472687,
+      "grad_norm": 0.47400587797164917,
+      "learning_rate": 8e-05,
+      "loss": 1.8974,
+      "step": 1847
+    },
+    {
+      "epoch": 0.10301003344481606,
+      "grad_norm": 0.4568255543708801,
+      "learning_rate": 8e-05,
+      "loss": 1.8252,
+      "step": 1848
+    },
+    {
+      "epoch": 0.10306577480490524,
+      "grad_norm": 0.4409621059894562,
+      "learning_rate": 8e-05,
+      "loss": 1.6955,
+      "step": 1849
+    },
+    {
+      "epoch": 0.10312151616499443,
+      "grad_norm": 0.48078006505966187,
+      "learning_rate": 8e-05,
+      "loss": 1.795,
+      "step": 1850
+    },
+    {
+      "epoch": 0.10317725752508361,
+      "grad_norm": 0.48587360978126526,
+      "learning_rate": 8e-05,
+      "loss": 1.864,
+      "step": 1851
+    },
+    {
+      "epoch": 0.1032329988851728,
+      "grad_norm": 0.5020497441291809,
+      "learning_rate": 8e-05,
+      "loss": 1.8202,
+      "step": 1852
+    },
+    {
+      "epoch": 0.10328874024526198,
+      "grad_norm": 0.4517962634563446,
+      "learning_rate": 8e-05,
+      "loss": 1.5516,
+      "step": 1853
+    },
+    {
+      "epoch": 0.10334448160535117,
+      "grad_norm": 0.4889071583747864,
+      "learning_rate": 8e-05,
+      "loss": 1.9975,
+      "step": 1854
+    },
+    {
+      "epoch": 0.10340022296544035,
+      "grad_norm": 0.4774690568447113,
+      "learning_rate": 8e-05,
+      "loss": 1.8562,
+      "step": 1855
+    },
+    {
+      "epoch": 0.10345596432552955,
+      "grad_norm": 0.427127867937088,
+      "learning_rate": 8e-05,
+      "loss": 1.7057,
+      "step": 1856
+    },
+    {
+      "epoch": 0.10351170568561872,
+      "grad_norm": 0.4165712594985962,
+      "learning_rate": 8e-05,
+      "loss": 1.5408,
+      "step": 1857
+    },
+    {
+      "epoch": 0.10356744704570792,
+      "grad_norm": 0.4988797605037689,
+      "learning_rate": 8e-05,
+      "loss": 1.8549,
+      "step": 1858
+    },
+    {
+      "epoch": 0.1036231884057971,
+      "grad_norm": 0.45366349816322327,
+      "learning_rate": 8e-05,
+      "loss": 1.7291,
+      "step": 1859
+    },
+    {
+      "epoch": 0.10367892976588629,
+      "grad_norm": 0.48266640305519104,
+      "learning_rate": 8e-05,
+      "loss": 1.8506,
+      "step": 1860
+    },
+    {
+      "epoch": 0.10373467112597548,
+      "grad_norm": 0.5486199855804443,
+      "learning_rate": 8e-05,
+      "loss": 1.9328,
+      "step": 1861
+    },
+    {
+      "epoch": 0.10379041248606466,
+      "grad_norm": 0.4452139735221863,
+      "learning_rate": 8e-05,
+      "loss": 1.6458,
+      "step": 1862
+    },
+    {
+      "epoch": 0.10384615384615385,
+      "grad_norm": 0.49109047651290894,
+      "learning_rate": 8e-05,
+      "loss": 1.9918,
+      "step": 1863
+    },
+    {
+      "epoch": 0.10390189520624303,
+      "grad_norm": 0.4817899763584137,
+      "learning_rate": 8e-05,
+      "loss": 1.8768,
+      "step": 1864
+    },
+    {
+      "epoch": 0.10395763656633222,
+      "grad_norm": 0.456006795167923,
+      "learning_rate": 8e-05,
+      "loss": 1.7827,
+      "step": 1865
+    },
+    {
+      "epoch": 0.1040133779264214,
+      "grad_norm": 0.4890350103378296,
+      "learning_rate": 8e-05,
+      "loss": 1.8561,
+      "step": 1866
+    },
+    {
+      "epoch": 0.1040691192865106,
+      "grad_norm": 0.46255168318748474,
+      "learning_rate": 8e-05,
+      "loss": 1.8764,
+      "step": 1867
+    },
+    {
+      "epoch": 0.10412486064659977,
+      "grad_norm": 0.4551663398742676,
+      "learning_rate": 8e-05,
+      "loss": 1.773,
+      "step": 1868
+    },
+    {
+      "epoch": 0.10418060200668897,
+      "grad_norm": 0.4570346772670746,
+      "learning_rate": 8e-05,
+      "loss": 1.7073,
+      "step": 1869
+    },
+    {
+      "epoch": 0.10423634336677814,
+      "grad_norm": 0.4652957618236542,
+      "learning_rate": 8e-05,
+      "loss": 1.871,
+      "step": 1870
+    },
+    {
+      "epoch": 0.10429208472686734,
+      "grad_norm": 0.45328280329704285,
+      "learning_rate": 8e-05,
+      "loss": 1.6733,
+      "step": 1871
+    },
+    {
+      "epoch": 0.10434782608695652,
+      "grad_norm": 0.4582420289516449,
+      "learning_rate": 8e-05,
+      "loss": 2.0648,
+      "step": 1872
+    },
+    {
+      "epoch": 0.10440356744704571,
+      "grad_norm": 0.4341772794723511,
+      "learning_rate": 8e-05,
+      "loss": 1.7146,
+      "step": 1873
+    },
+    {
+      "epoch": 0.10445930880713489,
+      "grad_norm": 0.4439215660095215,
+      "learning_rate": 8e-05,
+      "loss": 1.7852,
+      "step": 1874
+    },
+    {
+      "epoch": 0.10451505016722408,
+      "grad_norm": 0.46234145760536194,
+      "learning_rate": 8e-05,
+      "loss": 1.7784,
+      "step": 1875
+    },
+    {
+      "epoch": 0.10457079152731327,
+      "grad_norm": 0.5071830749511719,
+      "learning_rate": 8e-05,
+      "loss": 1.9705,
+      "step": 1876
+    },
+    {
+      "epoch": 0.10462653288740245,
+      "grad_norm": 0.4649686813354492,
+      "learning_rate": 8e-05,
+      "loss": 1.6822,
+      "step": 1877
+    },
+    {
+      "epoch": 0.10468227424749164,
+      "grad_norm": 0.4438864588737488,
+      "learning_rate": 8e-05,
+      "loss": 1.8218,
+      "step": 1878
+    },
+    {
+      "epoch": 0.10473801560758082,
+      "grad_norm": 0.5034699440002441,
+      "learning_rate": 8e-05,
+      "loss": 1.8693,
+      "step": 1879
+    },
+    {
+      "epoch": 0.10479375696767002,
+      "grad_norm": 0.43608057498931885,
+      "learning_rate": 8e-05,
+      "loss": 1.7494,
+      "step": 1880
+    },
+    {
+      "epoch": 0.1048494983277592,
+      "grad_norm": 0.4907682538032532,
+      "learning_rate": 8e-05,
+      "loss": 1.9842,
+      "step": 1881
+    },
+    {
+      "epoch": 0.10490523968784839,
+      "grad_norm": 0.44603654742240906,
+      "learning_rate": 8e-05,
+      "loss": 1.6714,
+      "step": 1882
+    },
+    {
+      "epoch": 0.10496098104793757,
+      "grad_norm": 0.440072238445282,
+      "learning_rate": 8e-05,
+      "loss": 1.7894,
+      "step": 1883
+    },
+    {
+      "epoch": 0.10501672240802676,
+      "grad_norm": 0.501599133014679,
+      "learning_rate": 8e-05,
+      "loss": 1.9509,
+      "step": 1884
+    },
+    {
+      "epoch": 0.10507246376811594,
+      "grad_norm": 0.47615668177604675,
+      "learning_rate": 8e-05,
+      "loss": 1.6011,
+      "step": 1885
+    },
+    {
+      "epoch": 0.10512820512820513,
+      "grad_norm": 0.4609353840351105,
+      "learning_rate": 8e-05,
+      "loss": 1.799,
+      "step": 1886
+    },
+    {
+      "epoch": 0.10518394648829431,
+      "grad_norm": 0.4454839825630188,
+      "learning_rate": 8e-05,
+      "loss": 1.6604,
+      "step": 1887
+    },
+    {
+      "epoch": 0.1052396878483835,
+      "grad_norm": 0.44142767786979675,
+      "learning_rate": 8e-05,
+      "loss": 1.8667,
+      "step": 1888
+    },
+    {
+      "epoch": 0.1052954292084727,
+      "grad_norm": 0.4566827416419983,
+      "learning_rate": 8e-05,
+      "loss": 1.3807,
+      "step": 1889
+    },
+    {
+      "epoch": 0.10535117056856187,
+      "grad_norm": 0.4766084551811218,
+      "learning_rate": 8e-05,
+      "loss": 1.7556,
+      "step": 1890
+    },
+    {
+      "epoch": 0.10540691192865106,
+      "grad_norm": 0.44653797149658203,
+      "learning_rate": 8e-05,
+      "loss": 1.721,
+      "step": 1891
+    },
+    {
+      "epoch": 0.10546265328874024,
+      "grad_norm": 0.4749918580055237,
+      "learning_rate": 8e-05,
+      "loss": 2.0207,
+      "step": 1892
+    },
+    {
+      "epoch": 0.10551839464882944,
+      "grad_norm": 0.42253586649894714,
+      "learning_rate": 8e-05,
+      "loss": 1.4832,
+      "step": 1893
+    },
+    {
+      "epoch": 0.10557413600891861,
+      "grad_norm": 0.46958550810813904,
+      "learning_rate": 8e-05,
+      "loss": 1.5483,
+      "step": 1894
+    },
+    {
+      "epoch": 0.10562987736900781,
+      "grad_norm": 0.4402949810028076,
+      "learning_rate": 8e-05,
+      "loss": 1.6613,
+      "step": 1895
+    },
+    {
+      "epoch": 0.10568561872909699,
+      "grad_norm": 0.49999094009399414,
+      "learning_rate": 8e-05,
+      "loss": 1.6266,
+      "step": 1896
+    },
+    {
+      "epoch": 0.10574136008918618,
+      "grad_norm": 0.47692495584487915,
+      "learning_rate": 8e-05,
+      "loss": 1.8586,
+      "step": 1897
+    },
+    {
+      "epoch": 0.10579710144927536,
+      "grad_norm": 0.48545780777931213,
+      "learning_rate": 8e-05,
+      "loss": 1.8865,
+      "step": 1898
+    },
+    {
+      "epoch": 0.10585284280936455,
+      "grad_norm": 0.45592087507247925,
+      "learning_rate": 8e-05,
+      "loss": 1.8979,
+      "step": 1899
+    },
+    {
+      "epoch": 0.10590858416945373,
+      "grad_norm": 0.5070605278015137,
+      "learning_rate": 8e-05,
+      "loss": 1.7738,
+      "step": 1900
+    },
+    {
+      "epoch": 0.10596432552954292,
+      "grad_norm": 0.5075011253356934,
+      "learning_rate": 8e-05,
+      "loss": 1.7382,
+      "step": 1901
+    },
+    {
+      "epoch": 0.1060200668896321,
+      "grad_norm": 0.4672543406486511,
+      "learning_rate": 8e-05,
+      "loss": 1.7201,
+      "step": 1902
+    },
+    {
+      "epoch": 0.10607580824972129,
+      "grad_norm": 0.4875164031982422,
+      "learning_rate": 8e-05,
+      "loss": 1.8483,
+      "step": 1903
+    },
+    {
+      "epoch": 0.10613154960981049,
+      "grad_norm": 0.4881870448589325,
+      "learning_rate": 8e-05,
+      "loss": 1.6607,
+      "step": 1904
+    },
+    {
+      "epoch": 0.10618729096989966,
+      "grad_norm": 0.5196061730384827,
+      "learning_rate": 8e-05,
+      "loss": 1.8877,
+      "step": 1905
+    },
+    {
+      "epoch": 0.10624303232998886,
+      "grad_norm": 0.45339685678482056,
+      "learning_rate": 8e-05,
+      "loss": 1.8615,
+      "step": 1906
+    },
+    {
+      "epoch": 0.10629877369007804,
+      "grad_norm": 0.4748283624649048,
+      "learning_rate": 8e-05,
+      "loss": 1.6601,
+      "step": 1907
+    },
+    {
+      "epoch": 0.10635451505016723,
+      "grad_norm": 0.41848766803741455,
+      "learning_rate": 8e-05,
+      "loss": 1.6985,
+      "step": 1908
+    },
+    {
+      "epoch": 0.1064102564102564,
+      "grad_norm": 0.4328663945198059,
+      "learning_rate": 8e-05,
+      "loss": 1.6866,
+      "step": 1909
+    },
+    {
+      "epoch": 0.1064659977703456,
+      "grad_norm": 0.4594240188598633,
+      "learning_rate": 8e-05,
+      "loss": 1.831,
+      "step": 1910
+    },
+    {
+      "epoch": 0.10652173913043478,
+      "grad_norm": 0.48182612657546997,
+      "learning_rate": 8e-05,
+      "loss": 1.6941,
+      "step": 1911
+    },
+    {
+      "epoch": 0.10657748049052397,
+      "grad_norm": 0.4915633797645569,
+      "learning_rate": 8e-05,
+      "loss": 1.739,
+      "step": 1912
+    },
+    {
+      "epoch": 0.10663322185061315,
+      "grad_norm": 0.44918638467788696,
+      "learning_rate": 8e-05,
+      "loss": 1.6886,
+      "step": 1913
+    },
+    {
+      "epoch": 0.10668896321070234,
+      "grad_norm": 0.45250630378723145,
+      "learning_rate": 8e-05,
+      "loss": 1.652,
+      "step": 1914
+    },
+    {
+      "epoch": 0.10674470457079152,
+      "grad_norm": 0.4650265574455261,
+      "learning_rate": 8e-05,
+      "loss": 1.5937,
+      "step": 1915
+    },
+    {
+      "epoch": 0.10680044593088071,
+      "grad_norm": 0.4698179364204407,
+      "learning_rate": 8e-05,
+      "loss": 1.7066,
+      "step": 1916
+    },
+    {
+      "epoch": 0.1068561872909699,
+      "grad_norm": 0.46050387620925903,
+      "learning_rate": 8e-05,
+      "loss": 1.6285,
+      "step": 1917
+    },
+    {
+      "epoch": 0.10691192865105908,
+      "grad_norm": 0.4757847785949707,
+      "learning_rate": 8e-05,
+      "loss": 1.9062,
+      "step": 1918
+    },
+    {
+      "epoch": 0.10696767001114828,
+      "grad_norm": 0.48355957865715027,
+      "learning_rate": 8e-05,
+      "loss": 1.8416,
+      "step": 1919
+    },
+    {
+      "epoch": 0.10702341137123746,
+      "grad_norm": 0.4324290156364441,
+      "learning_rate": 8e-05,
+      "loss": 1.7597,
+      "step": 1920
+    },
+    {
+      "epoch": 0.10707915273132665,
+      "grad_norm": 0.48721253871917725,
+      "learning_rate": 8e-05,
+      "loss": 1.9141,
+      "step": 1921
+    },
+    {
+      "epoch": 0.10713489409141583,
+      "grad_norm": 0.5150349736213684,
+      "learning_rate": 8e-05,
+      "loss": 1.7725,
+      "step": 1922
+    },
+    {
+      "epoch": 0.10719063545150502,
+      "grad_norm": 0.4498967230319977,
+      "learning_rate": 8e-05,
+      "loss": 1.4825,
+      "step": 1923
+    },
+    {
+      "epoch": 0.1072463768115942,
+      "grad_norm": 0.7026382088661194,
+      "learning_rate": 8e-05,
+      "loss": 1.7359,
+      "step": 1924
+    },
+    {
+      "epoch": 0.10730211817168339,
+      "grad_norm": 0.42979228496551514,
+      "learning_rate": 8e-05,
+      "loss": 1.2196,
+      "step": 1925
+    },
+    {
+      "epoch": 0.10735785953177257,
+      "grad_norm": 0.4411061704158783,
+      "learning_rate": 8e-05,
+      "loss": 1.5994,
+      "step": 1926
+    },
+    {
+      "epoch": 0.10741360089186176,
+      "grad_norm": 0.4753207862377167,
+      "learning_rate": 8e-05,
+      "loss": 1.7305,
+      "step": 1927
+    },
+    {
+      "epoch": 0.10746934225195094,
+      "grad_norm": 0.4685523509979248,
+      "learning_rate": 8e-05,
+      "loss": 1.7912,
+      "step": 1928
+    },
+    {
+      "epoch": 0.10752508361204013,
+      "grad_norm": 0.4643222987651825,
+      "learning_rate": 8e-05,
+      "loss": 1.8235,
+      "step": 1929
+    },
+    {
+      "epoch": 0.10758082497212931,
+      "grad_norm": 0.44670549035072327,
+      "learning_rate": 8e-05,
+      "loss": 1.7247,
+      "step": 1930
+    },
+    {
+      "epoch": 0.1076365663322185,
+      "grad_norm": 0.4861903488636017,
+      "learning_rate": 8e-05,
+      "loss": 1.8318,
+      "step": 1931
+    },
+    {
+      "epoch": 0.1076923076923077,
+      "grad_norm": 0.5127159357070923,
+      "learning_rate": 8e-05,
+      "loss": 1.8395,
+      "step": 1932
+    },
+    {
+      "epoch": 0.10774804905239688,
+      "grad_norm": 0.46058979630470276,
+      "learning_rate": 8e-05,
+      "loss": 1.7455,
+      "step": 1933
+    },
+    {
+      "epoch": 0.10780379041248607,
+      "grad_norm": 0.5476949214935303,
+      "learning_rate": 8e-05,
+      "loss": 1.9165,
+      "step": 1934
+    },
+    {
+      "epoch": 0.10785953177257525,
+      "grad_norm": 0.4545874297618866,
+      "learning_rate": 8e-05,
+      "loss": 1.6108,
+      "step": 1935
+    },
+    {
+      "epoch": 0.10791527313266444,
+      "grad_norm": 0.4586825668811798,
+      "learning_rate": 8e-05,
+      "loss": 1.743,
+      "step": 1936
+    },
+    {
+      "epoch": 0.10797101449275362,
+      "grad_norm": 0.48396754264831543,
+      "learning_rate": 8e-05,
+      "loss": 1.513,
+      "step": 1937
+    },
+    {
+      "epoch": 0.10802675585284281,
+      "grad_norm": 0.4474707543849945,
+      "learning_rate": 8e-05,
+      "loss": 1.7162,
+      "step": 1938
+    },
+    {
+      "epoch": 0.10808249721293199,
+      "grad_norm": 0.42385002970695496,
+      "learning_rate": 8e-05,
+      "loss": 1.5032,
+      "step": 1939
+    },
+    {
+      "epoch": 0.10813823857302118,
+      "grad_norm": 0.5376535654067993,
+      "learning_rate": 8e-05,
+      "loss": 1.5726,
+      "step": 1940
+    },
+    {
+      "epoch": 0.10819397993311036,
+      "grad_norm": 0.4649655818939209,
+      "learning_rate": 8e-05,
+      "loss": 1.7542,
+      "step": 1941
+    },
+    {
+      "epoch": 0.10824972129319956,
+      "grad_norm": 0.4395988881587982,
+      "learning_rate": 8e-05,
+      "loss": 1.5953,
+      "step": 1942
+    },
+    {
+      "epoch": 0.10830546265328873,
+      "grad_norm": 0.5266790390014648,
+      "learning_rate": 8e-05,
+      "loss": 2.0656,
+      "step": 1943
+    },
+    {
+      "epoch": 0.10836120401337793,
+      "grad_norm": 0.44322502613067627,
+      "learning_rate": 8e-05,
+      "loss": 1.8313,
+      "step": 1944
+    },
+    {
+      "epoch": 0.10841694537346712,
+      "grad_norm": 0.4520188868045807,
+      "learning_rate": 8e-05,
+      "loss": 1.7102,
+      "step": 1945
+    },
+    {
+      "epoch": 0.1084726867335563,
+      "grad_norm": 0.4372086822986603,
+      "learning_rate": 8e-05,
+      "loss": 1.6014,
+      "step": 1946
+    },
+    {
+      "epoch": 0.10852842809364549,
+      "grad_norm": 0.46225085854530334,
+      "learning_rate": 8e-05,
+      "loss": 1.8784,
+      "step": 1947
+    },
+    {
+      "epoch": 0.10858416945373467,
+      "grad_norm": 0.49719974398612976,
+      "learning_rate": 8e-05,
+      "loss": 1.8832,
+      "step": 1948
+    },
+    {
+      "epoch": 0.10863991081382386,
+      "grad_norm": 0.4694206714630127,
+      "learning_rate": 8e-05,
+      "loss": 1.7833,
+      "step": 1949
+    },
+    {
+      "epoch": 0.10869565217391304,
+      "grad_norm": 0.5506940484046936,
+      "learning_rate": 8e-05,
+      "loss": 2.2823,
+      "step": 1950
+    },
+    {
+      "epoch": 0.10875139353400223,
+      "grad_norm": 0.5728739500045776,
+      "learning_rate": 8e-05,
+      "loss": 2.0483,
+      "step": 1951
+    },
+    {
+      "epoch": 0.10880713489409141,
+      "grad_norm": 0.44349315762519836,
+      "learning_rate": 8e-05,
+      "loss": 1.6299,
+      "step": 1952
+    },
+    {
+      "epoch": 0.1088628762541806,
+      "grad_norm": 0.49954578280448914,
+      "learning_rate": 8e-05,
+      "loss": 1.7425,
+      "step": 1953
+    },
+    {
+      "epoch": 0.10891861761426978,
+      "grad_norm": 0.5076021552085876,
+      "learning_rate": 8e-05,
+      "loss": 1.9896,
+      "step": 1954
+    },
+    {
+      "epoch": 0.10897435897435898,
+      "grad_norm": 0.48354780673980713,
+      "learning_rate": 8e-05,
+      "loss": 1.7661,
+      "step": 1955
+    },
+    {
+      "epoch": 0.10903010033444815,
+      "grad_norm": 0.47348281741142273,
+      "learning_rate": 8e-05,
+      "loss": 1.7341,
+      "step": 1956
+    },
+    {
+      "epoch": 0.10908584169453735,
+      "grad_norm": 0.45801085233688354,
+      "learning_rate": 8e-05,
+      "loss": 1.6604,
+      "step": 1957
+    },
+    {
+      "epoch": 0.10914158305462654,
+      "grad_norm": 0.4829603433609009,
+      "learning_rate": 8e-05,
+      "loss": 1.8312,
+      "step": 1958
+    },
+    {
+      "epoch": 0.10919732441471572,
+      "grad_norm": 0.4740946590900421,
+      "learning_rate": 8e-05,
+      "loss": 1.7841,
+      "step": 1959
+    },
+    {
+      "epoch": 0.10925306577480491,
+      "grad_norm": 0.46358588337898254,
+      "learning_rate": 8e-05,
+      "loss": 2.0598,
+      "step": 1960
+    },
+    {
+      "epoch": 0.10930880713489409,
+      "grad_norm": 0.44745883345603943,
+      "learning_rate": 8e-05,
+      "loss": 1.7173,
+      "step": 1961
+    },
+    {
+      "epoch": 0.10936454849498328,
+      "grad_norm": 0.45677393674850464,
+      "learning_rate": 8e-05,
+      "loss": 1.8382,
+      "step": 1962
+    },
+    {
+      "epoch": 0.10942028985507246,
+      "grad_norm": 0.4677621126174927,
+      "learning_rate": 8e-05,
+      "loss": 1.7491,
+      "step": 1963
+    },
+    {
+      "epoch": 0.10947603121516165,
+      "grad_norm": 0.4459719955921173,
+      "learning_rate": 8e-05,
+      "loss": 1.8215,
+      "step": 1964
+    },
+    {
+      "epoch": 0.10953177257525083,
+      "grad_norm": 0.44011497497558594,
+      "learning_rate": 8e-05,
+      "loss": 1.4185,
+      "step": 1965
+    },
+    {
+      "epoch": 0.10958751393534003,
+      "grad_norm": 0.5114625692367554,
+      "learning_rate": 8e-05,
+      "loss": 2.0072,
+      "step": 1966
+    },
+    {
+      "epoch": 0.1096432552954292,
+      "grad_norm": 0.4603794813156128,
+      "learning_rate": 8e-05,
+      "loss": 1.8177,
+      "step": 1967
+    },
+    {
+      "epoch": 0.1096989966555184,
+      "grad_norm": 0.44069552421569824,
+      "learning_rate": 8e-05,
+      "loss": 1.6253,
+      "step": 1968
+    },
+    {
+      "epoch": 0.10975473801560758,
+      "grad_norm": 0.4534465968608856,
+      "learning_rate": 8e-05,
+      "loss": 1.5732,
+      "step": 1969
+    },
+    {
+      "epoch": 0.10981047937569677,
+      "grad_norm": 0.5088835954666138,
+      "learning_rate": 8e-05,
+      "loss": 1.549,
+      "step": 1970
+    },
+    {
+      "epoch": 0.10986622073578595,
+      "grad_norm": 0.472575843334198,
+      "learning_rate": 8e-05,
+      "loss": 1.7316,
+      "step": 1971
+    },
+    {
+      "epoch": 0.10992196209587514,
+      "grad_norm": 0.4615132510662079,
+      "learning_rate": 8e-05,
+      "loss": 1.6679,
+      "step": 1972
+    },
+    {
+      "epoch": 0.10997770345596433,
+      "grad_norm": 0.5491868853569031,
+      "learning_rate": 8e-05,
+      "loss": 1.7941,
+      "step": 1973
+    },
+    {
+      "epoch": 0.11003344481605351,
+      "grad_norm": 0.45354369282722473,
+      "learning_rate": 8e-05,
+      "loss": 1.7325,
+      "step": 1974
+    },
+    {
+      "epoch": 0.1100891861761427,
+      "grad_norm": 0.4760855436325073,
+      "learning_rate": 8e-05,
+      "loss": 1.7824,
+      "step": 1975
+    },
+    {
+      "epoch": 0.11014492753623188,
+      "grad_norm": 0.4518475830554962,
+      "learning_rate": 8e-05,
+      "loss": 1.6424,
+      "step": 1976
+    },
+    {
+      "epoch": 0.11020066889632107,
+      "grad_norm": 0.44643738865852356,
+      "learning_rate": 8e-05,
+      "loss": 1.7556,
+      "step": 1977
+    },
+    {
+      "epoch": 0.11025641025641025,
+      "grad_norm": 0.46472617983818054,
+      "learning_rate": 8e-05,
+      "loss": 1.6799,
+      "step": 1978
+    },
+    {
+      "epoch": 0.11031215161649945,
+      "grad_norm": 0.46941208839416504,
+      "learning_rate": 8e-05,
+      "loss": 1.8121,
+      "step": 1979
+    },
+    {
+      "epoch": 0.11036789297658862,
+      "grad_norm": 0.47246110439300537,
+      "learning_rate": 8e-05,
+      "loss": 1.8077,
+      "step": 1980
+    },
+    {
+      "epoch": 0.11042363433667782,
+      "grad_norm": 0.4268743693828583,
+      "learning_rate": 8e-05,
+      "loss": 1.5715,
+      "step": 1981
+    },
+    {
+      "epoch": 0.110479375696767,
+      "grad_norm": 0.4658483564853668,
+      "learning_rate": 8e-05,
+      "loss": 1.9498,
+      "step": 1982
+    },
+    {
+      "epoch": 0.11053511705685619,
+      "grad_norm": 0.4351171851158142,
+      "learning_rate": 8e-05,
+      "loss": 1.6103,
+      "step": 1983
+    },
+    {
+      "epoch": 0.11059085841694537,
+      "grad_norm": 0.48690593242645264,
+      "learning_rate": 8e-05,
+      "loss": 1.8043,
+      "step": 1984
+    },
+    {
+      "epoch": 0.11064659977703456,
+      "grad_norm": 0.4481874108314514,
+      "learning_rate": 8e-05,
+      "loss": 1.7241,
+      "step": 1985
+    },
+    {
+      "epoch": 0.11070234113712375,
+      "grad_norm": 0.43319857120513916,
+      "learning_rate": 8e-05,
+      "loss": 1.7234,
+      "step": 1986
+    },
+    {
+      "epoch": 0.11075808249721293,
+      "grad_norm": 0.48339614272117615,
+      "learning_rate": 8e-05,
+      "loss": 1.7305,
+      "step": 1987
+    },
+    {
+      "epoch": 0.11081382385730212,
+      "grad_norm": 0.4544183313846588,
+      "learning_rate": 8e-05,
+      "loss": 1.8293,
+      "step": 1988
+    },
+    {
+      "epoch": 0.1108695652173913,
+      "grad_norm": 0.4810698330402374,
+      "learning_rate": 8e-05,
+      "loss": 1.7221,
+      "step": 1989
+    },
+    {
+      "epoch": 0.1109253065774805,
+      "grad_norm": 0.4192270040512085,
+      "learning_rate": 8e-05,
+      "loss": 1.6379,
+      "step": 1990
+    },
+    {
+      "epoch": 0.11098104793756967,
+      "grad_norm": 0.4315672516822815,
+      "learning_rate": 8e-05,
+      "loss": 1.543,
+      "step": 1991
+    },
+    {
+      "epoch": 0.11103678929765887,
+      "grad_norm": 0.4906083643436432,
+      "learning_rate": 8e-05,
+      "loss": 1.8446,
+      "step": 1992
+    },
+    {
+      "epoch": 0.11109253065774805,
+      "grad_norm": 0.42365047335624695,
+      "learning_rate": 8e-05,
+      "loss": 1.6205,
+      "step": 1993
+    },
+    {
+      "epoch": 0.11114827201783724,
+      "grad_norm": 0.51629239320755,
+      "learning_rate": 8e-05,
+      "loss": 1.9564,
+      "step": 1994
+    },
+    {
+      "epoch": 0.11120401337792642,
+      "grad_norm": 0.43064284324645996,
+      "learning_rate": 8e-05,
+      "loss": 1.5131,
+      "step": 1995
+    },
+    {
+      "epoch": 0.11125975473801561,
+      "grad_norm": 0.5297256112098694,
+      "learning_rate": 8e-05,
+      "loss": 2.0655,
+      "step": 1996
+    },
+    {
+      "epoch": 0.11131549609810479,
+      "grad_norm": 0.4883597493171692,
+      "learning_rate": 8e-05,
+      "loss": 1.7461,
+      "step": 1997
+    },
+    {
+      "epoch": 0.11137123745819398,
+      "grad_norm": 0.46033623814582825,
+      "learning_rate": 8e-05,
+      "loss": 1.7602,
+      "step": 1998
+    },
+    {
+      "epoch": 0.11142697881828316,
+      "grad_norm": 0.4744260311126709,
+      "learning_rate": 8e-05,
+      "loss": 1.9014,
+      "step": 1999
+    },
+    {
+      "epoch": 0.11148272017837235,
+      "grad_norm": 0.49645110964775085,
+      "learning_rate": 8e-05,
+      "loss": 1.8827,
+      "step": 2000
+    },
+    {
+      "epoch": 0.11153846153846154,
+      "grad_norm": 0.4630310833454132,
+      "learning_rate": 8e-05,
+      "loss": 1.7795,
+      "step": 2001
+    },
+    {
+      "epoch": 0.11159420289855072,
+      "grad_norm": 0.48973262310028076,
+      "learning_rate": 8e-05,
+      "loss": 1.8411,
+      "step": 2002
+    },
+    {
+      "epoch": 0.11164994425863992,
+      "grad_norm": 0.4827178716659546,
+      "learning_rate": 8e-05,
+      "loss": 1.7959,
+      "step": 2003
+    },
+    {
+      "epoch": 0.1117056856187291,
+      "grad_norm": 0.4800148904323578,
+      "learning_rate": 8e-05,
+      "loss": 1.8899,
+      "step": 2004
+    },
+    {
+      "epoch": 0.11176142697881829,
+      "grad_norm": 0.4850090444087982,
+      "learning_rate": 8e-05,
+      "loss": 1.7319,
+      "step": 2005
+    },
+    {
+      "epoch": 0.11181716833890747,
+      "grad_norm": 0.47121182084083557,
+      "learning_rate": 8e-05,
+      "loss": 1.6407,
+      "step": 2006
+    },
+    {
+      "epoch": 0.11187290969899666,
+      "grad_norm": 0.5369400382041931,
+      "learning_rate": 8e-05,
+      "loss": 1.9465,
+      "step": 2007
+    },
+    {
+      "epoch": 0.11192865105908584,
+      "grad_norm": 0.4796384871006012,
+      "learning_rate": 8e-05,
+      "loss": 1.9416,
+      "step": 2008
+    },
+    {
+      "epoch": 0.11198439241917503,
+      "grad_norm": 0.4600585699081421,
+      "learning_rate": 8e-05,
+      "loss": 1.759,
+      "step": 2009
+    },
+    {
+      "epoch": 0.11204013377926421,
+      "grad_norm": 0.4528631567955017,
+      "learning_rate": 8e-05,
+      "loss": 1.6321,
+      "step": 2010
+    },
+    {
+      "epoch": 0.1120958751393534,
+      "grad_norm": 0.48515555262565613,
+      "learning_rate": 8e-05,
+      "loss": 1.7965,
+      "step": 2011
+    },
+    {
+      "epoch": 0.11215161649944258,
+      "grad_norm": 0.46082037687301636,
+      "learning_rate": 8e-05,
+      "loss": 1.7435,
+      "step": 2012
+    },
+    {
+      "epoch": 0.11220735785953177,
+      "grad_norm": 0.4584730863571167,
+      "learning_rate": 8e-05,
+      "loss": 1.7968,
+      "step": 2013
+    },
+    {
+      "epoch": 0.11226309921962097,
+      "grad_norm": 0.47898101806640625,
+      "learning_rate": 8e-05,
+      "loss": 1.7914,
+      "step": 2014
+    },
+    {
+      "epoch": 0.11231884057971014,
+      "grad_norm": 0.4609237313270569,
+      "learning_rate": 8e-05,
+      "loss": 1.7662,
+      "step": 2015
+    },
+    {
+      "epoch": 0.11237458193979934,
+      "grad_norm": 0.4741979241371155,
+      "learning_rate": 8e-05,
+      "loss": 1.682,
+      "step": 2016
+    },
+    {
+      "epoch": 0.11243032329988852,
+      "grad_norm": 0.41993820667266846,
+      "learning_rate": 8e-05,
+      "loss": 1.7958,
+      "step": 2017
+    },
+    {
+      "epoch": 0.11248606465997771,
+      "grad_norm": 0.45935675501823425,
+      "learning_rate": 8e-05,
+      "loss": 1.7562,
+      "step": 2018
+    },
+    {
+      "epoch": 0.11254180602006689,
+      "grad_norm": 0.46503686904907227,
+      "learning_rate": 8e-05,
+      "loss": 1.7462,
+      "step": 2019
+    },
+    {
+      "epoch": 0.11259754738015608,
+      "grad_norm": 0.445574551820755,
+      "learning_rate": 8e-05,
+      "loss": 1.6958,
+      "step": 2020
+    },
+    {
+      "epoch": 0.11265328874024526,
+      "grad_norm": 0.51012122631073,
+      "learning_rate": 8e-05,
+      "loss": 1.984,
+      "step": 2021
+    },
+    {
+      "epoch": 0.11270903010033445,
+      "grad_norm": 0.4418213367462158,
+      "learning_rate": 8e-05,
+      "loss": 1.7274,
+      "step": 2022
+    },
+    {
+      "epoch": 0.11276477146042363,
+      "grad_norm": 0.4851004481315613,
+      "learning_rate": 8e-05,
+      "loss": 1.8347,
+      "step": 2023
+    },
+    {
+      "epoch": 0.11282051282051282,
+      "grad_norm": 0.44825586676597595,
+      "learning_rate": 8e-05,
+      "loss": 1.6569,
+      "step": 2024
+    },
+    {
+      "epoch": 0.112876254180602,
+      "grad_norm": 0.45392194390296936,
+      "learning_rate": 8e-05,
+      "loss": 1.7268,
+      "step": 2025
+    },
+    {
+      "epoch": 0.1129319955406912,
+      "grad_norm": 0.4423568546772003,
+      "learning_rate": 8e-05,
+      "loss": 1.5794,
+      "step": 2026
+    },
+    {
+      "epoch": 0.11298773690078037,
+      "grad_norm": 0.49717170000076294,
+      "learning_rate": 8e-05,
+      "loss": 1.8071,
+      "step": 2027
+    },
+    {
+      "epoch": 0.11304347826086956,
+      "grad_norm": 0.4636692404747009,
+      "learning_rate": 8e-05,
+      "loss": 1.9055,
+      "step": 2028
+    },
+    {
+      "epoch": 0.11309921962095876,
+      "grad_norm": 0.4840092062950134,
+      "learning_rate": 8e-05,
+      "loss": 1.8431,
+      "step": 2029
+    },
+    {
+      "epoch": 0.11315496098104794,
+      "grad_norm": 0.44109421968460083,
+      "learning_rate": 8e-05,
+      "loss": 1.8299,
+      "step": 2030
+    },
+    {
+      "epoch": 0.11321070234113713,
+      "grad_norm": 0.47095662355422974,
+      "learning_rate": 8e-05,
+      "loss": 1.6746,
+      "step": 2031
+    },
+    {
+      "epoch": 0.11326644370122631,
+      "grad_norm": 0.4372450113296509,
+      "learning_rate": 8e-05,
+      "loss": 1.8052,
+      "step": 2032
+    },
+    {
+      "epoch": 0.1133221850613155,
+      "grad_norm": 0.5086802244186401,
+      "learning_rate": 8e-05,
+      "loss": 1.9591,
+      "step": 2033
+    },
+    {
+      "epoch": 0.11337792642140468,
+      "grad_norm": 0.4743632376194,
+      "learning_rate": 8e-05,
+      "loss": 1.888,
+      "step": 2034
+    },
+    {
+      "epoch": 0.11343366778149387,
+      "grad_norm": 0.4422663152217865,
+      "learning_rate": 8e-05,
+      "loss": 1.6306,
+      "step": 2035
+    },
+    {
+      "epoch": 0.11348940914158305,
+      "grad_norm": 0.44120973348617554,
+      "learning_rate": 8e-05,
+      "loss": 1.7567,
+      "step": 2036
+    },
+    {
+      "epoch": 0.11354515050167224,
+      "grad_norm": 0.5340293049812317,
+      "learning_rate": 8e-05,
+      "loss": 1.8003,
+      "step": 2037
+    },
+    {
+      "epoch": 0.11360089186176142,
+      "grad_norm": 0.4521288573741913,
+      "learning_rate": 8e-05,
+      "loss": 1.6788,
+      "step": 2038
+    },
+    {
+      "epoch": 0.11365663322185061,
+      "grad_norm": 0.46381545066833496,
+      "learning_rate": 8e-05,
+      "loss": 1.8585,
+      "step": 2039
+    },
+    {
+      "epoch": 0.11371237458193979,
+      "grad_norm": 0.4723742604255676,
+      "learning_rate": 8e-05,
+      "loss": 1.624,
+      "step": 2040
+    },
+    {
+      "epoch": 0.11376811594202899,
+      "grad_norm": 0.48812541365623474,
+      "learning_rate": 8e-05,
+      "loss": 1.799,
+      "step": 2041
+    },
+    {
+      "epoch": 0.11382385730211818,
+      "grad_norm": 0.42873966693878174,
+      "learning_rate": 8e-05,
+      "loss": 1.4369,
+      "step": 2042
+    },
+    {
+      "epoch": 0.11387959866220736,
+      "grad_norm": 0.4686322510242462,
+      "learning_rate": 8e-05,
+      "loss": 1.7953,
+      "step": 2043
+    },
+    {
+      "epoch": 0.11393534002229655,
+      "grad_norm": 0.4819076657295227,
+      "learning_rate": 8e-05,
+      "loss": 1.9851,
+      "step": 2044
+    },
+    {
+      "epoch": 0.11399108138238573,
+      "grad_norm": 0.44066208600997925,
+      "learning_rate": 8e-05,
+      "loss": 1.8221,
+      "step": 2045
+    },
+    {
+      "epoch": 0.11404682274247492,
+      "grad_norm": 0.4927108585834503,
+      "learning_rate": 8e-05,
+      "loss": 1.8009,
+      "step": 2046
+    },
+    {
+      "epoch": 0.1141025641025641,
+      "grad_norm": 0.4504014849662781,
+      "learning_rate": 8e-05,
+      "loss": 1.5526,
+      "step": 2047
+    },
+    {
+      "epoch": 0.11415830546265329,
+      "grad_norm": 0.44212785363197327,
+      "learning_rate": 8e-05,
+      "loss": 1.7123,
+      "step": 2048
+    },
+    {
+      "epoch": 0.11421404682274247,
+      "grad_norm": 0.46705907583236694,
+      "learning_rate": 8e-05,
+      "loss": 1.7985,
+      "step": 2049
+    },
+    {
+      "epoch": 0.11426978818283166,
+      "grad_norm": 0.43998247385025024,
+      "learning_rate": 8e-05,
+      "loss": 1.4829,
+      "step": 2050
+    },
+    {
+      "epoch": 0.11432552954292084,
+      "grad_norm": 0.48790040612220764,
+      "learning_rate": 8e-05,
+      "loss": 1.9512,
+      "step": 2051
+    },
+    {
+      "epoch": 0.11438127090301003,
+      "grad_norm": 0.4267139434814453,
+      "learning_rate": 8e-05,
+      "loss": 1.5513,
+      "step": 2052
+    },
+    {
+      "epoch": 0.11443701226309921,
+      "grad_norm": 0.428985059261322,
+      "learning_rate": 8e-05,
+      "loss": 1.6513,
+      "step": 2053
+    },
+    {
+      "epoch": 0.1144927536231884,
+      "grad_norm": 0.47211799025535583,
+      "learning_rate": 8e-05,
+      "loss": 1.7737,
+      "step": 2054
+    },
+    {
+      "epoch": 0.11454849498327759,
+      "grad_norm": 0.4569706618785858,
+      "learning_rate": 8e-05,
+      "loss": 1.7369,
+      "step": 2055
+    },
+    {
+      "epoch": 0.11460423634336678,
+      "grad_norm": 0.44445621967315674,
+      "learning_rate": 8e-05,
+      "loss": 1.8191,
+      "step": 2056
+    },
+    {
+      "epoch": 0.11465997770345597,
+      "grad_norm": 0.497089147567749,
+      "learning_rate": 8e-05,
+      "loss": 1.79,
+      "step": 2057
+    },
+    {
+      "epoch": 0.11471571906354515,
+      "grad_norm": 0.46985799074172974,
+      "learning_rate": 8e-05,
+      "loss": 1.799,
+      "step": 2058
+    },
+    {
+      "epoch": 0.11477146042363434,
+      "grad_norm": 0.4886013865470886,
+      "learning_rate": 8e-05,
+      "loss": 1.9005,
+      "step": 2059
+    },
+    {
+      "epoch": 0.11482720178372352,
+      "grad_norm": 0.45066893100738525,
+      "learning_rate": 8e-05,
+      "loss": 1.3898,
+      "step": 2060
+    },
+    {
+      "epoch": 0.11488294314381271,
+      "grad_norm": 0.4732798933982849,
+      "learning_rate": 8e-05,
+      "loss": 1.7565,
+      "step": 2061
+    },
+    {
+      "epoch": 0.11493868450390189,
+      "grad_norm": 0.469097375869751,
+      "learning_rate": 8e-05,
+      "loss": 1.6626,
+      "step": 2062
+    },
+    {
+      "epoch": 0.11499442586399108,
+      "grad_norm": 0.475689560174942,
+      "learning_rate": 8e-05,
+      "loss": 1.7464,
+      "step": 2063
+    },
+    {
+      "epoch": 0.11505016722408026,
+      "grad_norm": 0.4490555226802826,
+      "learning_rate": 8e-05,
+      "loss": 1.7317,
+      "step": 2064
+    },
+    {
+      "epoch": 0.11510590858416946,
+      "grad_norm": 0.4971592426300049,
+      "learning_rate": 8e-05,
+      "loss": 1.7497,
+      "step": 2065
+    },
+    {
+      "epoch": 0.11516164994425863,
+      "grad_norm": 0.4859224855899811,
+      "learning_rate": 8e-05,
+      "loss": 1.7051,
+      "step": 2066
+    },
+    {
+      "epoch": 0.11521739130434783,
+      "grad_norm": 0.5182693600654602,
+      "learning_rate": 8e-05,
+      "loss": 1.6759,
+      "step": 2067
+    },
+    {
+      "epoch": 0.115273132664437,
+      "grad_norm": 0.5141357183456421,
+      "learning_rate": 8e-05,
+      "loss": 1.968,
+      "step": 2068
+    },
+    {
+      "epoch": 0.1153288740245262,
+      "grad_norm": 0.4426954686641693,
+      "learning_rate": 8e-05,
+      "loss": 1.6516,
+      "step": 2069
+    },
+    {
+      "epoch": 0.11538461538461539,
+      "grad_norm": 0.47207707166671753,
+      "learning_rate": 8e-05,
+      "loss": 2.1095,
+      "step": 2070
+    },
+    {
+      "epoch": 0.11544035674470457,
+      "grad_norm": 0.48400771617889404,
+      "learning_rate": 8e-05,
+      "loss": 1.6553,
+      "step": 2071
+    },
+    {
+      "epoch": 0.11549609810479376,
+      "grad_norm": 0.5170599818229675,
+      "learning_rate": 8e-05,
+      "loss": 1.9569,
+      "step": 2072
+    },
+    {
+      "epoch": 0.11555183946488294,
+      "grad_norm": 0.4554281234741211,
+      "learning_rate": 8e-05,
+      "loss": 1.7292,
+      "step": 2073
+    },
+    {
+      "epoch": 0.11560758082497213,
+      "grad_norm": 0.443513423204422,
+      "learning_rate": 8e-05,
+      "loss": 1.5986,
+      "step": 2074
+    },
+    {
+      "epoch": 0.11566332218506131,
+      "grad_norm": 0.44155359268188477,
+      "learning_rate": 8e-05,
+      "loss": 1.698,
+      "step": 2075
+    },
+    {
+      "epoch": 0.1157190635451505,
+      "grad_norm": 0.48619598150253296,
+      "learning_rate": 8e-05,
+      "loss": 1.863,
+      "step": 2076
+    },
+    {
+      "epoch": 0.11577480490523968,
+      "grad_norm": 0.42969754338264465,
+      "learning_rate": 8e-05,
+      "loss": 1.4224,
+      "step": 2077
+    },
+    {
+      "epoch": 0.11583054626532888,
+      "grad_norm": 0.44721049070358276,
+      "learning_rate": 8e-05,
+      "loss": 1.6163,
+      "step": 2078
+    },
+    {
+      "epoch": 0.11588628762541806,
+      "grad_norm": 0.474103718996048,
+      "learning_rate": 8e-05,
+      "loss": 1.8101,
+      "step": 2079
+    },
+    {
+      "epoch": 0.11594202898550725,
+      "grad_norm": 0.40180113911628723,
+      "learning_rate": 8e-05,
+      "loss": 1.2825,
+      "step": 2080
+    },
+    {
+      "epoch": 0.11599777034559643,
+      "grad_norm": 0.4595925211906433,
+      "learning_rate": 8e-05,
+      "loss": 1.5944,
+      "step": 2081
+    },
+    {
+      "epoch": 0.11605351170568562,
+      "grad_norm": 0.4501091539859772,
+      "learning_rate": 8e-05,
+      "loss": 1.6415,
+      "step": 2082
+    },
+    {
+      "epoch": 0.11610925306577481,
+      "grad_norm": 0.4637793004512787,
+      "learning_rate": 8e-05,
+      "loss": 1.9328,
+      "step": 2083
+    },
+    {
+      "epoch": 0.11616499442586399,
+      "grad_norm": 0.486345112323761,
+      "learning_rate": 8e-05,
+      "loss": 1.8168,
+      "step": 2084
+    },
+    {
+      "epoch": 0.11622073578595318,
+      "grad_norm": 0.4314326047897339,
+      "learning_rate": 8e-05,
+      "loss": 1.7082,
+      "step": 2085
+    },
+    {
+      "epoch": 0.11627647714604236,
+      "grad_norm": 0.4588695466518402,
+      "learning_rate": 8e-05,
+      "loss": 1.801,
+      "step": 2086
+    },
+    {
+      "epoch": 0.11633221850613155,
+      "grad_norm": 0.4157659709453583,
+      "learning_rate": 8e-05,
+      "loss": 1.5326,
+      "step": 2087
+    },
+    {
+      "epoch": 0.11638795986622073,
+      "grad_norm": 0.4738490879535675,
+      "learning_rate": 8e-05,
+      "loss": 1.8684,
+      "step": 2088
+    },
+    {
+      "epoch": 0.11644370122630993,
+      "grad_norm": 0.47257229685783386,
+      "learning_rate": 8e-05,
+      "loss": 1.8574,
+      "step": 2089
+    },
+    {
+      "epoch": 0.1164994425863991,
+      "grad_norm": 0.4345060884952545,
+      "learning_rate": 8e-05,
+      "loss": 1.6715,
+      "step": 2090
+    },
+    {
+      "epoch": 0.1165551839464883,
+      "grad_norm": 0.4843539893627167,
+      "learning_rate": 8e-05,
+      "loss": 1.8623,
+      "step": 2091
+    },
+    {
+      "epoch": 0.11661092530657748,
+      "grad_norm": 0.4963710904121399,
+      "learning_rate": 8e-05,
+      "loss": 2.03,
+      "step": 2092
+    },
+    {
+      "epoch": 0.11666666666666667,
+      "grad_norm": 0.4864402413368225,
+      "learning_rate": 8e-05,
+      "loss": 1.6748,
+      "step": 2093
+    },
+    {
+      "epoch": 0.11672240802675585,
+      "grad_norm": 0.4512304663658142,
+      "learning_rate": 8e-05,
+      "loss": 1.725,
+      "step": 2094
+    },
+    {
+      "epoch": 0.11677814938684504,
+      "grad_norm": 0.5019859671592712,
+      "learning_rate": 8e-05,
+      "loss": 1.8322,
+      "step": 2095
+    },
+    {
+      "epoch": 0.11683389074693422,
+      "grad_norm": 0.44922488927841187,
+      "learning_rate": 8e-05,
+      "loss": 1.8584,
+      "step": 2096
+    },
+    {
+      "epoch": 0.11688963210702341,
+      "grad_norm": 0.5823623538017273,
+      "learning_rate": 8e-05,
+      "loss": 1.9786,
+      "step": 2097
+    },
+    {
+      "epoch": 0.1169453734671126,
+      "grad_norm": 0.5128968954086304,
+      "learning_rate": 8e-05,
+      "loss": 1.3316,
+      "step": 2098
+    },
+    {
+      "epoch": 0.11700111482720178,
+      "grad_norm": 0.4815669357776642,
+      "learning_rate": 8e-05,
+      "loss": 1.7753,
+      "step": 2099
+    },
+    {
+      "epoch": 0.11705685618729098,
+      "grad_norm": 0.44617798924446106,
+      "learning_rate": 8e-05,
+      "loss": 1.4031,
+      "step": 2100
+    },
+    {
+      "epoch": 0.11711259754738015,
+      "grad_norm": 0.4319689869880676,
+      "learning_rate": 8e-05,
+      "loss": 1.6541,
+      "step": 2101
+    },
+    {
+      "epoch": 0.11716833890746935,
+      "grad_norm": 0.47626733779907227,
+      "learning_rate": 8e-05,
+      "loss": 1.8199,
+      "step": 2102
+    },
+    {
+      "epoch": 0.11722408026755853,
+      "grad_norm": 0.5096200108528137,
+      "learning_rate": 8e-05,
+      "loss": 1.9839,
+      "step": 2103
+    },
+    {
+      "epoch": 0.11727982162764772,
+      "grad_norm": 0.4460195302963257,
+      "learning_rate": 8e-05,
+      "loss": 1.6491,
+      "step": 2104
+    },
+    {
+      "epoch": 0.1173355629877369,
+      "grad_norm": 0.4559648036956787,
+      "learning_rate": 8e-05,
+      "loss": 1.6841,
+      "step": 2105
+    },
+    {
+      "epoch": 0.11739130434782609,
+      "grad_norm": 0.46103864908218384,
+      "learning_rate": 8e-05,
+      "loss": 1.7696,
+      "step": 2106
+    },
+    {
+      "epoch": 0.11744704570791527,
+      "grad_norm": 0.46135443449020386,
+      "learning_rate": 8e-05,
+      "loss": 1.7862,
+      "step": 2107
+    },
+    {
+      "epoch": 0.11750278706800446,
+      "grad_norm": 0.4671906530857086,
+      "learning_rate": 8e-05,
+      "loss": 1.8382,
+      "step": 2108
+    },
+    {
+      "epoch": 0.11755852842809364,
+      "grad_norm": 0.39926496148109436,
+      "learning_rate": 8e-05,
+      "loss": 1.4544,
+      "step": 2109
+    },
+    {
+      "epoch": 0.11761426978818283,
+      "grad_norm": 0.4120591878890991,
+      "learning_rate": 8e-05,
+      "loss": 1.5544,
+      "step": 2110
+    },
+    {
+      "epoch": 0.11767001114827202,
+      "grad_norm": 0.4811738431453705,
+      "learning_rate": 8e-05,
+      "loss": 1.8538,
+      "step": 2111
+    },
+    {
+      "epoch": 0.1177257525083612,
+      "grad_norm": 0.4601971507072449,
+      "learning_rate": 8e-05,
+      "loss": 1.8839,
+      "step": 2112
+    },
+    {
+      "epoch": 0.1177814938684504,
+      "grad_norm": 0.48387205600738525,
+      "learning_rate": 8e-05,
+      "loss": 1.9743,
+      "step": 2113
+    },
+    {
+      "epoch": 0.11783723522853957,
+      "grad_norm": 0.4420911371707916,
+      "learning_rate": 8e-05,
+      "loss": 1.5835,
+      "step": 2114
+    },
+    {
+      "epoch": 0.11789297658862877,
+      "grad_norm": 0.4618477523326874,
+      "learning_rate": 8e-05,
+      "loss": 1.7418,
+      "step": 2115
+    },
+    {
+      "epoch": 0.11794871794871795,
+      "grad_norm": 0.44071194529533386,
+      "learning_rate": 8e-05,
+      "loss": 1.6977,
+      "step": 2116
+    },
+    {
+      "epoch": 0.11800445930880714,
+      "grad_norm": 0.5029857754707336,
+      "learning_rate": 8e-05,
+      "loss": 1.6455,
+      "step": 2117
+    },
+    {
+      "epoch": 0.11806020066889632,
+      "grad_norm": 0.5218876004219055,
+      "learning_rate": 8e-05,
+      "loss": 1.7321,
+      "step": 2118
+    },
+    {
+      "epoch": 0.11811594202898551,
+      "grad_norm": 0.4479510486125946,
+      "learning_rate": 8e-05,
+      "loss": 1.772,
+      "step": 2119
+    },
+    {
+      "epoch": 0.11817168338907469,
+      "grad_norm": 0.45699721574783325,
+      "learning_rate": 8e-05,
+      "loss": 1.6361,
+      "step": 2120
+    },
+    {
+      "epoch": 0.11822742474916388,
+      "grad_norm": 0.45344969630241394,
+      "learning_rate": 8e-05,
+      "loss": 1.6322,
+      "step": 2121
+    },
+    {
+      "epoch": 0.11828316610925306,
+      "grad_norm": 0.48492100834846497,
+      "learning_rate": 8e-05,
+      "loss": 1.7463,
+      "step": 2122
+    },
+    {
+      "epoch": 0.11833890746934225,
+      "grad_norm": 0.4809948205947876,
+      "learning_rate": 8e-05,
+      "loss": 1.6205,
+      "step": 2123
+    },
+    {
+      "epoch": 0.11839464882943143,
+      "grad_norm": 0.4564463198184967,
+      "learning_rate": 8e-05,
+      "loss": 1.6662,
+      "step": 2124
+    },
+    {
+      "epoch": 0.11845039018952062,
+      "grad_norm": 0.46469205617904663,
+      "learning_rate": 8e-05,
+      "loss": 1.6511,
+      "step": 2125
+    },
+    {
+      "epoch": 0.11850613154960982,
+      "grad_norm": 0.4757656157016754,
+      "learning_rate": 8e-05,
+      "loss": 1.8471,
+      "step": 2126
+    },
+    {
+      "epoch": 0.118561872909699,
+      "grad_norm": 0.424377977848053,
+      "learning_rate": 8e-05,
+      "loss": 1.5451,
+      "step": 2127
+    },
+    {
+      "epoch": 0.11861761426978819,
+      "grad_norm": 0.46142521500587463,
+      "learning_rate": 8e-05,
+      "loss": 1.7637,
+      "step": 2128
+    },
+    {
+      "epoch": 0.11867335562987737,
+      "grad_norm": 0.4726645052433014,
+      "learning_rate": 8e-05,
+      "loss": 1.8552,
+      "step": 2129
+    },
+    {
+      "epoch": 0.11872909698996656,
+      "grad_norm": 0.4805576801300049,
+      "learning_rate": 8e-05,
+      "loss": 1.9007,
+      "step": 2130
+    },
+    {
+      "epoch": 0.11878483835005574,
+      "grad_norm": 0.4529826045036316,
+      "learning_rate": 8e-05,
+      "loss": 1.5354,
+      "step": 2131
+    },
+    {
+      "epoch": 0.11884057971014493,
+      "grad_norm": 0.4534931182861328,
+      "learning_rate": 8e-05,
+      "loss": 1.6667,
+      "step": 2132
+    },
+    {
+      "epoch": 0.11889632107023411,
+      "grad_norm": 0.4818662106990814,
+      "learning_rate": 8e-05,
+      "loss": 1.6025,
+      "step": 2133
+    },
+    {
+      "epoch": 0.1189520624303233,
+      "grad_norm": 0.4961552917957306,
+      "learning_rate": 8e-05,
+      "loss": 1.8231,
+      "step": 2134
+    },
+    {
+      "epoch": 0.11900780379041248,
+      "grad_norm": 0.5023543834686279,
+      "learning_rate": 8e-05,
+      "loss": 1.9326,
+      "step": 2135
+    },
+    {
+      "epoch": 0.11906354515050167,
+      "grad_norm": 0.4256194829940796,
+      "learning_rate": 8e-05,
+      "loss": 1.6839,
+      "step": 2136
+    },
+    {
+      "epoch": 0.11911928651059085,
+      "grad_norm": 0.4191765785217285,
+      "learning_rate": 8e-05,
+      "loss": 1.5835,
+      "step": 2137
+    },
+    {
+      "epoch": 0.11917502787068004,
+      "grad_norm": 0.44517984986305237,
+      "learning_rate": 8e-05,
+      "loss": 1.5365,
+      "step": 2138
+    },
+    {
+      "epoch": 0.11923076923076924,
+      "grad_norm": 0.4512975811958313,
+      "learning_rate": 8e-05,
+      "loss": 1.681,
+      "step": 2139
+    },
+    {
+      "epoch": 0.11928651059085842,
+      "grad_norm": 0.4347441792488098,
+      "learning_rate": 8e-05,
+      "loss": 1.7185,
+      "step": 2140
+    },
+    {
+      "epoch": 0.11934225195094761,
+      "grad_norm": 0.46845176815986633,
+      "learning_rate": 8e-05,
+      "loss": 1.663,
+      "step": 2141
+    },
+    {
+      "epoch": 0.11939799331103679,
+      "grad_norm": 0.47879666090011597,
+      "learning_rate": 8e-05,
+      "loss": 1.7167,
+      "step": 2142
+    },
+    {
+      "epoch": 0.11945373467112598,
+      "grad_norm": 0.49749916791915894,
+      "learning_rate": 8e-05,
+      "loss": 1.8381,
+      "step": 2143
+    },
+    {
+      "epoch": 0.11950947603121516,
+      "grad_norm": 0.4949910640716553,
+      "learning_rate": 8e-05,
+      "loss": 1.9321,
+      "step": 2144
+    },
+    {
+      "epoch": 0.11956521739130435,
+      "grad_norm": 0.4539514482021332,
+      "learning_rate": 8e-05,
+      "loss": 1.7162,
+      "step": 2145
+    },
+    {
+      "epoch": 0.11962095875139353,
+      "grad_norm": 0.5238786339759827,
+      "learning_rate": 8e-05,
+      "loss": 1.6846,
+      "step": 2146
+    },
+    {
+      "epoch": 0.11967670011148272,
+      "grad_norm": 0.44461992383003235,
+      "learning_rate": 8e-05,
+      "loss": 1.6656,
+      "step": 2147
+    },
+    {
+      "epoch": 0.1197324414715719,
+      "grad_norm": 0.4791272282600403,
+      "learning_rate": 8e-05,
+      "loss": 1.9857,
+      "step": 2148
+    },
+    {
+      "epoch": 0.1197881828316611,
+      "grad_norm": 0.4683381915092468,
+      "learning_rate": 8e-05,
+      "loss": 1.6514,
+      "step": 2149
+    },
+    {
+      "epoch": 0.11984392419175027,
+      "grad_norm": 0.5213828682899475,
+      "learning_rate": 8e-05,
+      "loss": 1.8802,
+      "step": 2150
+    },
+    {
+      "epoch": 0.11989966555183947,
+      "grad_norm": 0.48594242334365845,
+      "learning_rate": 8e-05,
+      "loss": 1.8658,
+      "step": 2151
+    },
+    {
+      "epoch": 0.11995540691192864,
+      "grad_norm": 0.4764782190322876,
+      "learning_rate": 8e-05,
+      "loss": 1.7479,
+      "step": 2152
+    },
+    {
+      "epoch": 0.12001114827201784,
+      "grad_norm": 0.45884522795677185,
+      "learning_rate": 8e-05,
+      "loss": 1.8825,
+      "step": 2153
+    },
+    {
+      "epoch": 0.12006688963210703,
+      "grad_norm": 0.5118260979652405,
+      "learning_rate": 8e-05,
+      "loss": 2.047,
+      "step": 2154
+    },
+    {
+      "epoch": 0.12012263099219621,
+      "grad_norm": 0.5420330166816711,
+      "learning_rate": 8e-05,
+      "loss": 2.1218,
+      "step": 2155
+    },
+    {
+      "epoch": 0.1201783723522854,
+      "grad_norm": 0.4370868504047394,
+      "learning_rate": 8e-05,
+      "loss": 1.5695,
+      "step": 2156
+    },
+    {
+      "epoch": 0.12023411371237458,
+      "grad_norm": 0.4741508960723877,
+      "learning_rate": 8e-05,
+      "loss": 1.654,
+      "step": 2157
+    },
+    {
+      "epoch": 0.12028985507246377,
+      "grad_norm": 0.4782915711402893,
+      "learning_rate": 8e-05,
+      "loss": 1.7332,
+      "step": 2158
+    },
+    {
+      "epoch": 0.12034559643255295,
+      "grad_norm": 0.5331648588180542,
+      "learning_rate": 8e-05,
+      "loss": 1.9393,
+      "step": 2159
+    },
+    {
+      "epoch": 0.12040133779264214,
+      "grad_norm": 0.4637068212032318,
+      "learning_rate": 8e-05,
+      "loss": 1.6524,
+      "step": 2160
+    },
+    {
+      "epoch": 0.12045707915273132,
+      "grad_norm": 0.4589218497276306,
+      "learning_rate": 8e-05,
+      "loss": 1.8953,
+      "step": 2161
+    },
+    {
+      "epoch": 0.12051282051282051,
+      "grad_norm": 0.4819149672985077,
+      "learning_rate": 8e-05,
+      "loss": 2.0129,
+      "step": 2162
+    },
+    {
+      "epoch": 0.1205685618729097,
+      "grad_norm": 0.437137246131897,
+      "learning_rate": 8e-05,
+      "loss": 1.5711,
+      "step": 2163
+    },
+    {
+      "epoch": 0.12062430323299889,
+      "grad_norm": 0.47422948479652405,
+      "learning_rate": 8e-05,
+      "loss": 1.7699,
+      "step": 2164
+    },
+    {
+      "epoch": 0.12068004459308806,
+      "grad_norm": 0.474104642868042,
+      "learning_rate": 8e-05,
+      "loss": 1.7519,
+      "step": 2165
+    },
+    {
+      "epoch": 0.12073578595317726,
+      "grad_norm": 0.4602581262588501,
+      "learning_rate": 8e-05,
+      "loss": 1.6553,
+      "step": 2166
+    },
+    {
+      "epoch": 0.12079152731326645,
+      "grad_norm": 0.5225976705551147,
+      "learning_rate": 8e-05,
+      "loss": 2.3767,
+      "step": 2167
+    },
+    {
+      "epoch": 0.12084726867335563,
+      "grad_norm": 0.45720773935317993,
+      "learning_rate": 8e-05,
+      "loss": 1.6674,
+      "step": 2168
+    },
+    {
+      "epoch": 0.12090301003344482,
+      "grad_norm": 0.4820588529109955,
+      "learning_rate": 8e-05,
+      "loss": 1.842,
+      "step": 2169
+    },
+    {
+      "epoch": 0.120958751393534,
+      "grad_norm": 0.4850405752658844,
+      "learning_rate": 8e-05,
+      "loss": 1.8699,
+      "step": 2170
+    },
+    {
+      "epoch": 0.12101449275362319,
+      "grad_norm": 0.4590352177619934,
+      "learning_rate": 8e-05,
+      "loss": 1.7275,
+      "step": 2171
+    },
+    {
+      "epoch": 0.12107023411371237,
+      "grad_norm": 0.4691894054412842,
+      "learning_rate": 8e-05,
+      "loss": 1.8761,
+      "step": 2172
+    },
+    {
+      "epoch": 0.12112597547380156,
+      "grad_norm": 0.467676043510437,
+      "learning_rate": 8e-05,
+      "loss": 1.7993,
+      "step": 2173
+    },
+    {
+      "epoch": 0.12118171683389074,
+      "grad_norm": 0.4780999720096588,
+      "learning_rate": 8e-05,
+      "loss": 1.7737,
+      "step": 2174
+    },
+    {
+      "epoch": 0.12123745819397994,
+      "grad_norm": 0.46493837237358093,
+      "learning_rate": 8e-05,
+      "loss": 1.7838,
+      "step": 2175
+    },
+    {
+      "epoch": 0.12129319955406911,
+      "grad_norm": 0.4486934542655945,
+      "learning_rate": 8e-05,
+      "loss": 1.5086,
+      "step": 2176
+    },
+    {
+      "epoch": 0.12134894091415831,
+      "grad_norm": 0.4756641983985901,
+      "learning_rate": 8e-05,
+      "loss": 1.6553,
+      "step": 2177
+    },
+    {
+      "epoch": 0.12140468227424749,
+      "grad_norm": 0.47148486971855164,
+      "learning_rate": 8e-05,
+      "loss": 2.0012,
+      "step": 2178
+    },
+    {
+      "epoch": 0.12146042363433668,
+      "grad_norm": 0.45895448327064514,
+      "learning_rate": 8e-05,
+      "loss": 1.7609,
+      "step": 2179
+    },
+    {
+      "epoch": 0.12151616499442586,
+      "grad_norm": 0.4958876371383667,
+      "learning_rate": 8e-05,
+      "loss": 1.8134,
+      "step": 2180
+    },
+    {
+      "epoch": 0.12157190635451505,
+      "grad_norm": 0.4782322645187378,
+      "learning_rate": 8e-05,
+      "loss": 1.7053,
+      "step": 2181
+    },
+    {
+      "epoch": 0.12162764771460424,
+      "grad_norm": 0.4557296931743622,
+      "learning_rate": 8e-05,
+      "loss": 1.6757,
+      "step": 2182
+    },
+    {
+      "epoch": 0.12168338907469342,
+      "grad_norm": 0.4677014648914337,
+      "learning_rate": 8e-05,
+      "loss": 1.7492,
+      "step": 2183
+    },
+    {
+      "epoch": 0.12173913043478261,
+      "grad_norm": 0.47127747535705566,
+      "learning_rate": 8e-05,
+      "loss": 1.8121,
+      "step": 2184
+    },
+    {
+      "epoch": 0.12179487179487179,
+      "grad_norm": 0.44277313351631165,
+      "learning_rate": 8e-05,
+      "loss": 1.7282,
+      "step": 2185
+    },
+    {
+      "epoch": 0.12185061315496098,
+      "grad_norm": 0.4506374001502991,
+      "learning_rate": 8e-05,
+      "loss": 1.874,
+      "step": 2186
+    },
+    {
+      "epoch": 0.12190635451505016,
+      "grad_norm": 0.4785115122795105,
+      "learning_rate": 8e-05,
+      "loss": 1.8468,
+      "step": 2187
+    },
+    {
+      "epoch": 0.12196209587513936,
+      "grad_norm": 0.4615698456764221,
+      "learning_rate": 8e-05,
+      "loss": 1.7637,
+      "step": 2188
+    },
+    {
+      "epoch": 0.12201783723522854,
+      "grad_norm": 0.5574464797973633,
+      "learning_rate": 8e-05,
+      "loss": 1.7617,
+      "step": 2189
+    },
+    {
+      "epoch": 0.12207357859531773,
+      "grad_norm": 0.47425100207328796,
+      "learning_rate": 8e-05,
+      "loss": 1.7472,
+      "step": 2190
+    },
+    {
+      "epoch": 0.1221293199554069,
+      "grad_norm": 0.4637100100517273,
+      "learning_rate": 8e-05,
+      "loss": 1.7061,
+      "step": 2191
+    },
+    {
+      "epoch": 0.1221850613154961,
+      "grad_norm": 0.4725492000579834,
+      "learning_rate": 8e-05,
+      "loss": 1.7959,
+      "step": 2192
+    },
+    {
+      "epoch": 0.12224080267558528,
+      "grad_norm": 0.47741201519966125,
+      "learning_rate": 8e-05,
+      "loss": 1.854,
+      "step": 2193
+    },
+    {
+      "epoch": 0.12229654403567447,
+      "grad_norm": 0.4578995108604431,
+      "learning_rate": 8e-05,
+      "loss": 1.569,
+      "step": 2194
+    },
+    {
+      "epoch": 0.12235228539576366,
+      "grad_norm": 0.4601028263568878,
+      "learning_rate": 8e-05,
+      "loss": 1.6284,
+      "step": 2195
+    },
+    {
+      "epoch": 0.12240802675585284,
+      "grad_norm": 0.45650744438171387,
+      "learning_rate": 8e-05,
+      "loss": 1.6563,
+      "step": 2196
+    },
+    {
+      "epoch": 0.12246376811594203,
+      "grad_norm": 0.4996529221534729,
+      "learning_rate": 8e-05,
+      "loss": 1.8408,
+      "step": 2197
+    },
+    {
+      "epoch": 0.12251950947603121,
+      "grad_norm": 0.4483855068683624,
+      "learning_rate": 8e-05,
+      "loss": 1.5627,
+      "step": 2198
+    },
+    {
+      "epoch": 0.1225752508361204,
+      "grad_norm": 0.489122599363327,
+      "learning_rate": 8e-05,
+      "loss": 1.7472,
+      "step": 2199
+    },
+    {
+      "epoch": 0.12263099219620958,
+      "grad_norm": 0.4574604332447052,
+      "learning_rate": 8e-05,
+      "loss": 1.801,
+      "step": 2200
+    },
+    {
+      "epoch": 0.12268673355629878,
+      "grad_norm": 0.5230749249458313,
+      "learning_rate": 8e-05,
+      "loss": 2.0084,
+      "step": 2201
+    },
+    {
+      "epoch": 0.12274247491638796,
+      "grad_norm": 0.4647735357284546,
+      "learning_rate": 8e-05,
+      "loss": 1.6136,
+      "step": 2202
+    },
+    {
+      "epoch": 0.12279821627647715,
+      "grad_norm": 0.45680564641952515,
+      "learning_rate": 8e-05,
+      "loss": 1.5722,
+      "step": 2203
+    },
+    {
+      "epoch": 0.12285395763656633,
+      "grad_norm": 0.4625045955181122,
+      "learning_rate": 8e-05,
+      "loss": 1.7592,
+      "step": 2204
+    },
+    {
+      "epoch": 0.12290969899665552,
+      "grad_norm": 0.46245506405830383,
+      "learning_rate": 8e-05,
+      "loss": 1.5329,
+      "step": 2205
+    },
+    {
+      "epoch": 0.1229654403567447,
+      "grad_norm": 0.4651133418083191,
+      "learning_rate": 8e-05,
+      "loss": 1.7119,
+      "step": 2206
+    },
+    {
+      "epoch": 0.12302118171683389,
+      "grad_norm": 0.45406287908554077,
+      "learning_rate": 8e-05,
+      "loss": 1.4837,
+      "step": 2207
+    },
+    {
+      "epoch": 0.12307692307692308,
+      "grad_norm": 0.5053935050964355,
+      "learning_rate": 8e-05,
+      "loss": 1.8821,
+      "step": 2208
+    },
+    {
+      "epoch": 0.12313266443701226,
+      "grad_norm": 0.5143939256668091,
+      "learning_rate": 8e-05,
+      "loss": 1.8583,
+      "step": 2209
+    },
+    {
+      "epoch": 0.12318840579710146,
+      "grad_norm": 0.4740615487098694,
+      "learning_rate": 8e-05,
+      "loss": 1.8297,
+      "step": 2210
+    },
+    {
+      "epoch": 0.12324414715719063,
+      "grad_norm": 0.5244475603103638,
+      "learning_rate": 8e-05,
+      "loss": 1.7763,
+      "step": 2211
+    },
+    {
+      "epoch": 0.12329988851727983,
+      "grad_norm": 0.5088525414466858,
+      "learning_rate": 8e-05,
+      "loss": 2.1075,
+      "step": 2212
+    },
+    {
+      "epoch": 0.123355629877369,
+      "grad_norm": 0.48931679129600525,
+      "learning_rate": 8e-05,
+      "loss": 1.9784,
+      "step": 2213
+    },
+    {
+      "epoch": 0.1234113712374582,
+      "grad_norm": 0.4884590804576874,
+      "learning_rate": 8e-05,
+      "loss": 1.8409,
+      "step": 2214
+    },
+    {
+      "epoch": 0.12346711259754738,
+      "grad_norm": 0.4486399292945862,
+      "learning_rate": 8e-05,
+      "loss": 1.8856,
+      "step": 2215
+    },
+    {
+      "epoch": 0.12352285395763657,
+      "grad_norm": 0.4579851031303406,
+      "learning_rate": 8e-05,
+      "loss": 1.7836,
+      "step": 2216
+    },
+    {
+      "epoch": 0.12357859531772575,
+      "grad_norm": 0.40383824706077576,
+      "learning_rate": 8e-05,
+      "loss": 1.4822,
+      "step": 2217
+    },
+    {
+      "epoch": 0.12363433667781494,
+      "grad_norm": 0.4631842374801636,
+      "learning_rate": 8e-05,
+      "loss": 1.6552,
+      "step": 2218
+    },
+    {
+      "epoch": 0.12369007803790412,
+      "grad_norm": 0.47191739082336426,
+      "learning_rate": 8e-05,
+      "loss": 1.7517,
+      "step": 2219
+    },
+    {
+      "epoch": 0.12374581939799331,
+      "grad_norm": 0.47322702407836914,
+      "learning_rate": 8e-05,
+      "loss": 1.8699,
+      "step": 2220
+    },
+    {
+      "epoch": 0.12380156075808249,
+      "grad_norm": 0.48753586411476135,
+      "learning_rate": 8e-05,
+      "loss": 1.6901,
+      "step": 2221
+    },
+    {
+      "epoch": 0.12385730211817168,
+      "grad_norm": 0.489136278629303,
+      "learning_rate": 8e-05,
+      "loss": 1.9412,
+      "step": 2222
+    },
+    {
+      "epoch": 0.12391304347826088,
+      "grad_norm": 0.45695024728775024,
+      "learning_rate": 8e-05,
+      "loss": 1.8031,
+      "step": 2223
+    },
+    {
+      "epoch": 0.12396878483835005,
+      "grad_norm": 0.4342595636844635,
+      "learning_rate": 8e-05,
+      "loss": 1.733,
+      "step": 2224
+    },
+    {
+      "epoch": 0.12402452619843925,
+      "grad_norm": 0.5584686994552612,
+      "learning_rate": 8e-05,
+      "loss": 1.6239,
+      "step": 2225
+    },
+    {
+      "epoch": 0.12408026755852843,
+      "grad_norm": 0.5097849369049072,
+      "learning_rate": 8e-05,
+      "loss": 1.7744,
+      "step": 2226
+    },
+    {
+      "epoch": 0.12413600891861762,
+      "grad_norm": 0.43732327222824097,
+      "learning_rate": 8e-05,
+      "loss": 1.5259,
+      "step": 2227
+    },
+    {
+      "epoch": 0.1241917502787068,
+      "grad_norm": 0.46310287714004517,
+      "learning_rate": 8e-05,
+      "loss": 1.6754,
+      "step": 2228
+    },
+    {
+      "epoch": 0.12424749163879599,
+      "grad_norm": 0.4878222942352295,
+      "learning_rate": 8e-05,
+      "loss": 1.8352,
+      "step": 2229
+    },
+    {
+      "epoch": 0.12430323299888517,
+      "grad_norm": 0.4622375965118408,
+      "learning_rate": 8e-05,
+      "loss": 1.7948,
+      "step": 2230
+    },
+    {
+      "epoch": 0.12435897435897436,
+      "grad_norm": 0.4793466329574585,
+      "learning_rate": 8e-05,
+      "loss": 2.0367,
+      "step": 2231
+    },
+    {
+      "epoch": 0.12441471571906354,
+      "grad_norm": 0.45333993434906006,
+      "learning_rate": 8e-05,
+      "loss": 1.5381,
+      "step": 2232
+    },
+    {
+      "epoch": 0.12447045707915273,
+      "grad_norm": 0.4307631850242615,
+      "learning_rate": 8e-05,
+      "loss": 1.5617,
+      "step": 2233
+    },
+    {
+      "epoch": 0.12452619843924191,
+      "grad_norm": 0.4964076578617096,
+      "learning_rate": 8e-05,
+      "loss": 1.9645,
+      "step": 2234
+    },
+    {
+      "epoch": 0.1245819397993311,
+      "grad_norm": 0.4543098509311676,
+      "learning_rate": 8e-05,
+      "loss": 1.4873,
+      "step": 2235
+    },
+    {
+      "epoch": 0.1246376811594203,
+      "grad_norm": 0.4989856779575348,
+      "learning_rate": 8e-05,
+      "loss": 1.8919,
+      "step": 2236
+    },
+    {
+      "epoch": 0.12469342251950948,
+      "grad_norm": 0.46211791038513184,
+      "learning_rate": 8e-05,
+      "loss": 1.745,
+      "step": 2237
+    },
+    {
+      "epoch": 0.12474916387959867,
+      "grad_norm": 0.525151789188385,
+      "learning_rate": 8e-05,
+      "loss": 2.0378,
+      "step": 2238
+    },
+    {
+      "epoch": 0.12480490523968785,
+      "grad_norm": 0.4855180084705353,
+      "learning_rate": 8e-05,
+      "loss": 1.8417,
+      "step": 2239
+    },
+    {
+      "epoch": 0.12486064659977704,
+      "grad_norm": 0.46944236755371094,
+      "learning_rate": 8e-05,
+      "loss": 1.6507,
+      "step": 2240
+    },
+    {
+      "epoch": 0.12491638795986622,
+      "grad_norm": 0.4638369679450989,
+      "learning_rate": 8e-05,
+      "loss": 1.7477,
+      "step": 2241
+    },
+    {
+      "epoch": 0.12497212931995541,
+      "grad_norm": 0.49965327978134155,
+      "learning_rate": 8e-05,
+      "loss": 1.8928,
+      "step": 2242
+    },
+    {
+      "epoch": 0.1250278706800446,
+      "grad_norm": 0.4624139964580536,
+      "learning_rate": 8e-05,
+      "loss": 1.8775,
+      "step": 2243
+    },
+    {
+      "epoch": 0.12508361204013377,
+      "grad_norm": 0.4745078384876251,
+      "learning_rate": 8e-05,
+      "loss": 2.0191,
+      "step": 2244
+    },
+    {
+      "epoch": 0.12513935340022297,
+      "grad_norm": 0.4567002058029175,
+      "learning_rate": 8e-05,
+      "loss": 1.7058,
+      "step": 2245
+    },
+    {
+      "epoch": 0.12519509476031215,
+      "grad_norm": 0.4978659749031067,
+      "learning_rate": 8e-05,
+      "loss": 1.8425,
+      "step": 2246
+    },
+    {
+      "epoch": 0.12525083612040133,
+      "grad_norm": 0.5292001366615295,
+      "learning_rate": 8e-05,
+      "loss": 1.9908,
+      "step": 2247
+    },
+    {
+      "epoch": 0.1253065774804905,
+      "grad_norm": 0.4715481698513031,
+      "learning_rate": 8e-05,
+      "loss": 1.7982,
+      "step": 2248
+    },
+    {
+      "epoch": 0.12536231884057972,
+      "grad_norm": 0.4420139491558075,
+      "learning_rate": 8e-05,
+      "loss": 1.2473,
+      "step": 2249
+    },
+    {
+      "epoch": 0.1254180602006689,
+      "grad_norm": 0.5093467235565186,
+      "learning_rate": 8e-05,
+      "loss": 1.9668,
+      "step": 2250
+    },
+    {
+      "epoch": 0.12547380156075807,
+      "grad_norm": 0.449186772108078,
+      "learning_rate": 8e-05,
+      "loss": 1.5954,
+      "step": 2251
+    },
+    {
+      "epoch": 0.12552954292084728,
+      "grad_norm": 0.45465993881225586,
+      "learning_rate": 8e-05,
+      "loss": 1.7156,
+      "step": 2252
+    },
+    {
+      "epoch": 0.12558528428093646,
+      "grad_norm": 0.46614545583724976,
+      "learning_rate": 8e-05,
+      "loss": 1.647,
+      "step": 2253
+    },
+    {
+      "epoch": 0.12564102564102564,
+      "grad_norm": 0.4224128723144531,
+      "learning_rate": 8e-05,
+      "loss": 1.6038,
+      "step": 2254
+    },
+    {
+      "epoch": 0.12569676700111482,
+      "grad_norm": 0.4402707517147064,
+      "learning_rate": 8e-05,
+      "loss": 1.748,
+      "step": 2255
+    },
+    {
+      "epoch": 0.12575250836120402,
+      "grad_norm": 0.4561031758785248,
+      "learning_rate": 8e-05,
+      "loss": 1.8052,
+      "step": 2256
+    },
+    {
+      "epoch": 0.1258082497212932,
+      "grad_norm": 0.44972988963127136,
+      "learning_rate": 8e-05,
+      "loss": 1.7521,
+      "step": 2257
+    },
+    {
+      "epoch": 0.12586399108138238,
+      "grad_norm": 0.5131308436393738,
+      "learning_rate": 8e-05,
+      "loss": 1.7484,
+      "step": 2258
+    },
+    {
+      "epoch": 0.12591973244147156,
+      "grad_norm": 0.45187079906463623,
+      "learning_rate": 8e-05,
+      "loss": 1.8119,
+      "step": 2259
+    },
+    {
+      "epoch": 0.12597547380156077,
+      "grad_norm": 0.46400967240333557,
+      "learning_rate": 8e-05,
+      "loss": 1.689,
+      "step": 2260
+    },
+    {
+      "epoch": 0.12603121516164995,
+      "grad_norm": 0.44966134428977966,
+      "learning_rate": 8e-05,
+      "loss": 1.4424,
+      "step": 2261
+    },
+    {
+      "epoch": 0.12608695652173912,
+      "grad_norm": 0.468363881111145,
+      "learning_rate": 8e-05,
+      "loss": 1.5864,
+      "step": 2262
+    },
+    {
+      "epoch": 0.1261426978818283,
+      "grad_norm": 0.4988058805465698,
+      "learning_rate": 8e-05,
+      "loss": 1.8812,
+      "step": 2263
+    },
+    {
+      "epoch": 0.1261984392419175,
+      "grad_norm": 0.5007922649383545,
+      "learning_rate": 8e-05,
+      "loss": 1.9379,
+      "step": 2264
+    },
+    {
+      "epoch": 0.1262541806020067,
+      "grad_norm": 0.4592812955379486,
+      "learning_rate": 8e-05,
+      "loss": 1.6016,
+      "step": 2265
+    },
+    {
+      "epoch": 0.12630992196209587,
+      "grad_norm": 0.5309374332427979,
+      "learning_rate": 8e-05,
+      "loss": 2.0361,
+      "step": 2266
+    },
+    {
+      "epoch": 0.12636566332218507,
+      "grad_norm": 0.49097344279289246,
+      "learning_rate": 8e-05,
+      "loss": 1.8266,
+      "step": 2267
+    },
+    {
+      "epoch": 0.12642140468227425,
+      "grad_norm": 0.48501044511795044,
+      "learning_rate": 8e-05,
+      "loss": 1.5567,
+      "step": 2268
+    },
+    {
+      "epoch": 0.12647714604236343,
+      "grad_norm": 0.4972813129425049,
+      "learning_rate": 8e-05,
+      "loss": 1.6461,
+      "step": 2269
+    },
+    {
+      "epoch": 0.1265328874024526,
+      "grad_norm": 0.544996440410614,
+      "learning_rate": 8e-05,
+      "loss": 1.8706,
+      "step": 2270
+    },
+    {
+      "epoch": 0.12658862876254182,
+      "grad_norm": 0.45850032567977905,
+      "learning_rate": 8e-05,
+      "loss": 1.3768,
+      "step": 2271
+    },
+    {
+      "epoch": 0.126644370122631,
+      "grad_norm": 0.46759000420570374,
+      "learning_rate": 8e-05,
+      "loss": 1.6794,
+      "step": 2272
+    },
+    {
+      "epoch": 0.12670011148272017,
+      "grad_norm": 0.47211045026779175,
+      "learning_rate": 8e-05,
+      "loss": 1.7303,
+      "step": 2273
+    },
+    {
+      "epoch": 0.12675585284280935,
+      "grad_norm": 0.4648776054382324,
+      "learning_rate": 8e-05,
+      "loss": 1.6901,
+      "step": 2274
+    },
+    {
+      "epoch": 0.12681159420289856,
+      "grad_norm": 0.5143617987632751,
+      "learning_rate": 8e-05,
+      "loss": 1.9032,
+      "step": 2275
+    },
+    {
+      "epoch": 0.12686733556298774,
+      "grad_norm": 0.49569982290267944,
+      "learning_rate": 8e-05,
+      "loss": 1.7533,
+      "step": 2276
+    },
+    {
+      "epoch": 0.12692307692307692,
+      "grad_norm": 0.4662966728210449,
+      "learning_rate": 8e-05,
+      "loss": 1.4075,
+      "step": 2277
+    },
+    {
+      "epoch": 0.12697881828316612,
+      "grad_norm": 0.5277073979377747,
+      "learning_rate": 8e-05,
+      "loss": 1.8739,
+      "step": 2278
+    },
+    {
+      "epoch": 0.1270345596432553,
+      "grad_norm": 0.45093291997909546,
+      "learning_rate": 8e-05,
+      "loss": 1.6588,
+      "step": 2279
+    },
+    {
+      "epoch": 0.12709030100334448,
+      "grad_norm": 0.43871331214904785,
+      "learning_rate": 8e-05,
+      "loss": 1.514,
+      "step": 2280
+    },
+    {
+      "epoch": 0.12714604236343366,
+      "grad_norm": 0.46460381150245667,
+      "learning_rate": 8e-05,
+      "loss": 1.7531,
+      "step": 2281
+    },
+    {
+      "epoch": 0.12720178372352287,
+      "grad_norm": 0.44319167733192444,
+      "learning_rate": 8e-05,
+      "loss": 1.8248,
+      "step": 2282
+    },
+    {
+      "epoch": 0.12725752508361204,
+      "grad_norm": 0.4465970993041992,
+      "learning_rate": 8e-05,
+      "loss": 1.659,
+      "step": 2283
+    },
+    {
+      "epoch": 0.12731326644370122,
+      "grad_norm": 0.466734915971756,
+      "learning_rate": 8e-05,
+      "loss": 1.9343,
+      "step": 2284
+    },
+    {
+      "epoch": 0.1273690078037904,
+      "grad_norm": 0.42454588413238525,
+      "learning_rate": 8e-05,
+      "loss": 1.6708,
+      "step": 2285
+    },
+    {
+      "epoch": 0.1274247491638796,
+      "grad_norm": 0.4505043625831604,
+      "learning_rate": 8e-05,
+      "loss": 1.6236,
+      "step": 2286
+    },
+    {
+      "epoch": 0.1274804905239688,
+      "grad_norm": 0.437162846326828,
+      "learning_rate": 8e-05,
+      "loss": 1.6453,
+      "step": 2287
+    },
+    {
+      "epoch": 0.12753623188405797,
+      "grad_norm": 0.48901355266571045,
+      "learning_rate": 8e-05,
+      "loss": 2.0439,
+      "step": 2288
+    },
+    {
+      "epoch": 0.12759197324414714,
+      "grad_norm": 0.44122645258903503,
+      "learning_rate": 8e-05,
+      "loss": 1.4767,
+      "step": 2289
+    },
+    {
+      "epoch": 0.12764771460423635,
+      "grad_norm": 0.4454699158668518,
+      "learning_rate": 8e-05,
+      "loss": 1.6566,
+      "step": 2290
+    },
+    {
+      "epoch": 0.12770345596432553,
+      "grad_norm": 0.444238543510437,
+      "learning_rate": 8e-05,
+      "loss": 1.6981,
+      "step": 2291
+    },
+    {
+      "epoch": 0.1277591973244147,
+      "grad_norm": 0.47556188702583313,
+      "learning_rate": 8e-05,
+      "loss": 1.7683,
+      "step": 2292
+    },
+    {
+      "epoch": 0.12781493868450391,
+      "grad_norm": 0.45707184076309204,
+      "learning_rate": 8e-05,
+      "loss": 1.6876,
+      "step": 2293
+    },
+    {
+      "epoch": 0.1278706800445931,
+      "grad_norm": 0.5256324410438538,
+      "learning_rate": 8e-05,
+      "loss": 1.8879,
+      "step": 2294
+    },
+    {
+      "epoch": 0.12792642140468227,
+      "grad_norm": 0.4542886018753052,
+      "learning_rate": 8e-05,
+      "loss": 1.6595,
+      "step": 2295
+    },
+    {
+      "epoch": 0.12798216276477145,
+      "grad_norm": 0.4788365662097931,
+      "learning_rate": 8e-05,
+      "loss": 1.5935,
+      "step": 2296
+    },
+    {
+      "epoch": 0.12803790412486066,
+      "grad_norm": 0.4697909653186798,
+      "learning_rate": 8e-05,
+      "loss": 1.8354,
+      "step": 2297
+    },
+    {
+      "epoch": 0.12809364548494984,
+      "grad_norm": 0.4855862557888031,
+      "learning_rate": 8e-05,
+      "loss": 1.5338,
+      "step": 2298
+    },
+    {
+      "epoch": 0.12814938684503902,
+      "grad_norm": 0.4692869484424591,
+      "learning_rate": 8e-05,
+      "loss": 1.7148,
+      "step": 2299
+    },
+    {
+      "epoch": 0.1282051282051282,
+      "grad_norm": 0.4679480493068695,
+      "learning_rate": 8e-05,
+      "loss": 1.6737,
+      "step": 2300
+    },
+    {
+      "epoch": 0.1282608695652174,
+      "grad_norm": 0.4384745955467224,
+      "learning_rate": 8e-05,
+      "loss": 1.5909,
+      "step": 2301
+    },
+    {
+      "epoch": 0.12831661092530658,
+      "grad_norm": 0.46700745820999146,
+      "learning_rate": 8e-05,
+      "loss": 1.8678,
+      "step": 2302
+    },
+    {
+      "epoch": 0.12837235228539576,
+      "grad_norm": 0.42763611674308777,
+      "learning_rate": 8e-05,
+      "loss": 1.6465,
+      "step": 2303
+    },
+    {
+      "epoch": 0.12842809364548494,
+      "grad_norm": 0.6245459318161011,
+      "learning_rate": 8e-05,
+      "loss": 2.2235,
+      "step": 2304
+    },
+    {
+      "epoch": 0.12848383500557414,
+      "grad_norm": 0.5192124247550964,
+      "learning_rate": 8e-05,
+      "loss": 1.802,
+      "step": 2305
+    },
+    {
+      "epoch": 0.12853957636566332,
+      "grad_norm": 0.45005589723587036,
+      "learning_rate": 8e-05,
+      "loss": 1.6594,
+      "step": 2306
+    },
+    {
+      "epoch": 0.1285953177257525,
+      "grad_norm": 0.4243832230567932,
+      "learning_rate": 8e-05,
+      "loss": 1.677,
+      "step": 2307
+    },
+    {
+      "epoch": 0.1286510590858417,
+      "grad_norm": 0.44305911660194397,
+      "learning_rate": 8e-05,
+      "loss": 1.5636,
+      "step": 2308
+    },
+    {
+      "epoch": 0.12870680044593089,
+      "grad_norm": 0.4739604592323303,
+      "learning_rate": 8e-05,
+      "loss": 1.8299,
+      "step": 2309
+    },
+    {
+      "epoch": 0.12876254180602006,
+      "grad_norm": 0.46584707498550415,
+      "learning_rate": 8e-05,
+      "loss": 1.6265,
+      "step": 2310
+    },
+    {
+      "epoch": 0.12881828316610924,
+      "grad_norm": 0.46699920296669006,
+      "learning_rate": 8e-05,
+      "loss": 1.7646,
+      "step": 2311
+    },
+    {
+      "epoch": 0.12887402452619845,
+      "grad_norm": 0.44824424386024475,
+      "learning_rate": 8e-05,
+      "loss": 1.5775,
+      "step": 2312
+    },
+    {
+      "epoch": 0.12892976588628763,
+      "grad_norm": 0.48826146125793457,
+      "learning_rate": 8e-05,
+      "loss": 1.8624,
+      "step": 2313
+    },
+    {
+      "epoch": 0.1289855072463768,
+      "grad_norm": 0.44969749450683594,
+      "learning_rate": 8e-05,
+      "loss": 1.5919,
+      "step": 2314
+    },
+    {
+      "epoch": 0.12904124860646599,
+      "grad_norm": 0.4631759822368622,
+      "learning_rate": 8e-05,
+      "loss": 1.6601,
+      "step": 2315
+    },
+    {
+      "epoch": 0.1290969899665552,
+      "grad_norm": 0.48021945357322693,
+      "learning_rate": 8e-05,
+      "loss": 1.6901,
+      "step": 2316
+    },
+    {
+      "epoch": 0.12915273132664437,
+      "grad_norm": 0.46702641248703003,
+      "learning_rate": 8e-05,
+      "loss": 1.6362,
+      "step": 2317
+    },
+    {
+      "epoch": 0.12920847268673355,
+      "grad_norm": 0.451192170381546,
+      "learning_rate": 8e-05,
+      "loss": 1.724,
+      "step": 2318
+    },
+    {
+      "epoch": 0.12926421404682276,
+      "grad_norm": 0.5693803429603577,
+      "learning_rate": 8e-05,
+      "loss": 2.0528,
+      "step": 2319
+    },
+    {
+      "epoch": 0.12931995540691194,
+      "grad_norm": 0.46649542450904846,
+      "learning_rate": 8e-05,
+      "loss": 1.5328,
+      "step": 2320
+    },
+    {
+      "epoch": 0.1293756967670011,
+      "grad_norm": 0.45501086115837097,
+      "learning_rate": 8e-05,
+      "loss": 1.6744,
+      "step": 2321
+    },
+    {
+      "epoch": 0.1294314381270903,
+      "grad_norm": 0.4884272813796997,
+      "learning_rate": 8e-05,
+      "loss": 1.7933,
+      "step": 2322
+    },
+    {
+      "epoch": 0.1294871794871795,
+      "grad_norm": 0.45968663692474365,
+      "learning_rate": 8e-05,
+      "loss": 1.8927,
+      "step": 2323
+    },
+    {
+      "epoch": 0.12954292084726868,
+      "grad_norm": 0.4407302141189575,
+      "learning_rate": 8e-05,
+      "loss": 1.7062,
+      "step": 2324
+    },
+    {
+      "epoch": 0.12959866220735786,
+      "grad_norm": 0.47550809383392334,
+      "learning_rate": 8e-05,
+      "loss": 1.8064,
+      "step": 2325
+    },
+    {
+      "epoch": 0.12965440356744704,
+      "grad_norm": 0.4188564121723175,
+      "learning_rate": 8e-05,
+      "loss": 1.5301,
+      "step": 2326
+    },
+    {
+      "epoch": 0.12971014492753624,
+      "grad_norm": 0.4824972450733185,
+      "learning_rate": 8e-05,
+      "loss": 1.6146,
+      "step": 2327
+    },
+    {
+      "epoch": 0.12976588628762542,
+      "grad_norm": 0.5340510606765747,
+      "learning_rate": 8e-05,
+      "loss": 1.9268,
+      "step": 2328
+    },
+    {
+      "epoch": 0.1298216276477146,
+      "grad_norm": 0.46250587701797485,
+      "learning_rate": 8e-05,
+      "loss": 1.7372,
+      "step": 2329
+    },
+    {
+      "epoch": 0.12987736900780378,
+      "grad_norm": 0.46774420142173767,
+      "learning_rate": 8e-05,
+      "loss": 1.7521,
+      "step": 2330
+    },
+    {
+      "epoch": 0.12993311036789298,
+      "grad_norm": 0.47270211577415466,
+      "learning_rate": 8e-05,
+      "loss": 1.7448,
+      "step": 2331
+    },
+    {
+      "epoch": 0.12998885172798216,
+      "grad_norm": 0.46113133430480957,
+      "learning_rate": 8e-05,
+      "loss": 1.7026,
+      "step": 2332
+    },
+    {
+      "epoch": 0.13004459308807134,
+      "grad_norm": 0.46038854122161865,
+      "learning_rate": 8e-05,
+      "loss": 1.8473,
+      "step": 2333
+    },
+    {
+      "epoch": 0.13010033444816055,
+      "grad_norm": 0.5171126127243042,
+      "learning_rate": 8e-05,
+      "loss": 1.9117,
+      "step": 2334
+    },
+    {
+      "epoch": 0.13015607580824973,
+      "grad_norm": 0.462720662355423,
+      "learning_rate": 8e-05,
+      "loss": 2.0309,
+      "step": 2335
+    },
+    {
+      "epoch": 0.1302118171683389,
+      "grad_norm": 0.45984727144241333,
+      "learning_rate": 8e-05,
+      "loss": 1.618,
+      "step": 2336
+    },
+    {
+      "epoch": 0.13026755852842808,
+      "grad_norm": 0.476755827665329,
+      "learning_rate": 8e-05,
+      "loss": 1.8149,
+      "step": 2337
+    },
+    {
+      "epoch": 0.1303232998885173,
+      "grad_norm": 0.44505324959754944,
+      "learning_rate": 8e-05,
+      "loss": 1.7122,
+      "step": 2338
+    },
+    {
+      "epoch": 0.13037904124860647,
+      "grad_norm": 0.4444805681705475,
+      "learning_rate": 8e-05,
+      "loss": 1.5229,
+      "step": 2339
+    },
+    {
+      "epoch": 0.13043478260869565,
+      "grad_norm": 0.4566769599914551,
+      "learning_rate": 8e-05,
+      "loss": 1.6963,
+      "step": 2340
+    },
+    {
+      "epoch": 0.13049052396878483,
+      "grad_norm": 0.48411083221435547,
+      "learning_rate": 8e-05,
+      "loss": 1.7126,
+      "step": 2341
+    },
+    {
+      "epoch": 0.13054626532887403,
+      "grad_norm": 0.485125333070755,
+      "learning_rate": 8e-05,
+      "loss": 1.6481,
+      "step": 2342
+    },
+    {
+      "epoch": 0.1306020066889632,
+      "grad_norm": 0.4920346140861511,
+      "learning_rate": 8e-05,
+      "loss": 1.8729,
+      "step": 2343
+    },
+    {
+      "epoch": 0.1306577480490524,
+      "grad_norm": 0.5254184603691101,
+      "learning_rate": 8e-05,
+      "loss": 1.9184,
+      "step": 2344
+    },
+    {
+      "epoch": 0.13071348940914157,
+      "grad_norm": 0.475009560585022,
+      "learning_rate": 8e-05,
+      "loss": 1.6471,
+      "step": 2345
+    },
+    {
+      "epoch": 0.13076923076923078,
+      "grad_norm": 0.41936373710632324,
+      "learning_rate": 8e-05,
+      "loss": 1.2888,
+      "step": 2346
+    },
+    {
+      "epoch": 0.13082497212931996,
+      "grad_norm": 0.49300825595855713,
+      "learning_rate": 8e-05,
+      "loss": 1.5936,
+      "step": 2347
+    },
+    {
+      "epoch": 0.13088071348940913,
+      "grad_norm": 0.4508037865161896,
+      "learning_rate": 8e-05,
+      "loss": 1.6086,
+      "step": 2348
+    },
+    {
+      "epoch": 0.13093645484949834,
+      "grad_norm": 0.4737650156021118,
+      "learning_rate": 8e-05,
+      "loss": 1.7373,
+      "step": 2349
+    },
+    {
+      "epoch": 0.13099219620958752,
+      "grad_norm": 0.48412951827049255,
+      "learning_rate": 8e-05,
+      "loss": 1.5208,
+      "step": 2350
+    },
+    {
+      "epoch": 0.1310479375696767,
+      "grad_norm": 0.5189991593360901,
+      "learning_rate": 8e-05,
+      "loss": 2.0046,
+      "step": 2351
+    },
+    {
+      "epoch": 0.13110367892976588,
+      "grad_norm": 0.49417969584465027,
+      "learning_rate": 8e-05,
+      "loss": 1.7563,
+      "step": 2352
+    },
+    {
+      "epoch": 0.13115942028985508,
+      "grad_norm": 0.5066606998443604,
+      "learning_rate": 8e-05,
+      "loss": 1.8102,
+      "step": 2353
+    },
+    {
+      "epoch": 0.13121516164994426,
+      "grad_norm": 0.47830361127853394,
+      "learning_rate": 8e-05,
+      "loss": 1.6611,
+      "step": 2354
+    },
+    {
+      "epoch": 0.13127090301003344,
+      "grad_norm": 0.4630042016506195,
+      "learning_rate": 8e-05,
+      "loss": 1.9327,
+      "step": 2355
+    },
+    {
+      "epoch": 0.13132664437012262,
+      "grad_norm": 0.4288950562477112,
+      "learning_rate": 8e-05,
+      "loss": 1.461,
+      "step": 2356
+    },
+    {
+      "epoch": 0.13138238573021183,
+      "grad_norm": 0.47212231159210205,
+      "learning_rate": 8e-05,
+      "loss": 1.831,
+      "step": 2357
+    },
+    {
+      "epoch": 0.131438127090301,
+      "grad_norm": 0.5282948017120361,
+      "learning_rate": 8e-05,
+      "loss": 2.0361,
+      "step": 2358
+    },
+    {
+      "epoch": 0.13149386845039018,
+      "grad_norm": 0.4551253616809845,
+      "learning_rate": 8e-05,
+      "loss": 1.7594,
+      "step": 2359
+    },
+    {
+      "epoch": 0.13154960981047936,
+      "grad_norm": 0.47356507182121277,
+      "learning_rate": 8e-05,
+      "loss": 1.7755,
+      "step": 2360
+    },
+    {
+      "epoch": 0.13160535117056857,
+      "grad_norm": 0.45470499992370605,
+      "learning_rate": 8e-05,
+      "loss": 1.8588,
+      "step": 2361
+    },
+    {
+      "epoch": 0.13166109253065775,
+      "grad_norm": 0.4819483160972595,
+      "learning_rate": 8e-05,
+      "loss": 1.6541,
+      "step": 2362
+    },
+    {
+      "epoch": 0.13171683389074693,
+      "grad_norm": 0.43687912821769714,
+      "learning_rate": 8e-05,
+      "loss": 1.658,
+      "step": 2363
+    },
+    {
+      "epoch": 0.13177257525083613,
+      "grad_norm": 0.48129919171333313,
+      "learning_rate": 8e-05,
+      "loss": 1.886,
+      "step": 2364
+    },
+    {
+      "epoch": 0.1318283166109253,
+      "grad_norm": 0.49660441279411316,
+      "learning_rate": 8e-05,
+      "loss": 1.7364,
+      "step": 2365
+    },
+    {
+      "epoch": 0.1318840579710145,
+      "grad_norm": 0.4926661550998688,
+      "learning_rate": 8e-05,
+      "loss": 2.0036,
+      "step": 2366
+    },
+    {
+      "epoch": 0.13193979933110367,
+      "grad_norm": 0.46969401836395264,
+      "learning_rate": 8e-05,
+      "loss": 1.6531,
+      "step": 2367
+    },
+    {
+      "epoch": 0.13199554069119288,
+      "grad_norm": 0.46234428882598877,
+      "learning_rate": 8e-05,
+      "loss": 1.7915,
+      "step": 2368
+    },
+    {
+      "epoch": 0.13205128205128205,
+      "grad_norm": 0.42720118165016174,
+      "learning_rate": 8e-05,
+      "loss": 1.6161,
+      "step": 2369
+    },
+    {
+      "epoch": 0.13210702341137123,
+      "grad_norm": 0.48308509588241577,
+      "learning_rate": 8e-05,
+      "loss": 1.7814,
+      "step": 2370
+    },
+    {
+      "epoch": 0.1321627647714604,
+      "grad_norm": 0.4851537048816681,
+      "learning_rate": 8e-05,
+      "loss": 1.7679,
+      "step": 2371
+    },
+    {
+      "epoch": 0.13221850613154962,
+      "grad_norm": 0.4548836052417755,
+      "learning_rate": 8e-05,
+      "loss": 1.5165,
+      "step": 2372
+    },
+    {
+      "epoch": 0.1322742474916388,
+      "grad_norm": 0.47316473722457886,
+      "learning_rate": 8e-05,
+      "loss": 1.7382,
+      "step": 2373
+    },
+    {
+      "epoch": 0.13232998885172798,
+      "grad_norm": 0.5082695484161377,
+      "learning_rate": 8e-05,
+      "loss": 1.7757,
+      "step": 2374
+    },
+    {
+      "epoch": 0.13238573021181718,
+      "grad_norm": 0.42562705278396606,
+      "learning_rate": 8e-05,
+      "loss": 1.3878,
+      "step": 2375
+    },
+    {
+      "epoch": 0.13244147157190636,
+      "grad_norm": 0.4784887135028839,
+      "learning_rate": 8e-05,
+      "loss": 1.5597,
+      "step": 2376
+    },
+    {
+      "epoch": 0.13249721293199554,
+      "grad_norm": 0.45655760169029236,
+      "learning_rate": 8e-05,
+      "loss": 1.3887,
+      "step": 2377
+    },
+    {
+      "epoch": 0.13255295429208472,
+      "grad_norm": 0.4548302888870239,
+      "learning_rate": 8e-05,
+      "loss": 1.7282,
+      "step": 2378
+    },
+    {
+      "epoch": 0.13260869565217392,
+      "grad_norm": 0.5463017821311951,
+      "learning_rate": 8e-05,
+      "loss": 1.93,
+      "step": 2379
+    },
+    {
+      "epoch": 0.1326644370122631,
+      "grad_norm": 0.5709219574928284,
+      "learning_rate": 8e-05,
+      "loss": 1.7358,
+      "step": 2380
+    },
+    {
+      "epoch": 0.13272017837235228,
+      "grad_norm": 0.4988589882850647,
+      "learning_rate": 8e-05,
+      "loss": 1.9047,
+      "step": 2381
+    },
+    {
+      "epoch": 0.13277591973244146,
+      "grad_norm": 0.47514599561691284,
+      "learning_rate": 8e-05,
+      "loss": 1.7282,
+      "step": 2382
+    },
+    {
+      "epoch": 0.13283166109253067,
+      "grad_norm": 0.47735369205474854,
+      "learning_rate": 8e-05,
+      "loss": 1.7754,
+      "step": 2383
+    },
+    {
+      "epoch": 0.13288740245261985,
+      "grad_norm": 0.4829070568084717,
+      "learning_rate": 8e-05,
+      "loss": 1.804,
+      "step": 2384
+    },
+    {
+      "epoch": 0.13294314381270902,
+      "grad_norm": 0.4622078537940979,
+      "learning_rate": 8e-05,
+      "loss": 1.6201,
+      "step": 2385
+    },
+    {
+      "epoch": 0.1329988851727982,
+      "grad_norm": 0.4724544286727905,
+      "learning_rate": 8e-05,
+      "loss": 1.6616,
+      "step": 2386
+    },
+    {
+      "epoch": 0.1330546265328874,
+      "grad_norm": 0.4968656301498413,
+      "learning_rate": 8e-05,
+      "loss": 1.8219,
+      "step": 2387
+    },
+    {
+      "epoch": 0.1331103678929766,
+      "grad_norm": 0.4678482413291931,
+      "learning_rate": 8e-05,
+      "loss": 1.7585,
+      "step": 2388
+    },
+    {
+      "epoch": 0.13316610925306577,
+      "grad_norm": 0.42720043659210205,
+      "learning_rate": 8e-05,
+      "loss": 1.6677,
+      "step": 2389
+    },
+    {
+      "epoch": 0.13322185061315497,
+      "grad_norm": 0.4516279995441437,
+      "learning_rate": 8e-05,
+      "loss": 1.8366,
+      "step": 2390
+    },
+    {
+      "epoch": 0.13327759197324415,
+      "grad_norm": 0.44496116042137146,
+      "learning_rate": 8e-05,
+      "loss": 1.603,
+      "step": 2391
+    },
+    {
+      "epoch": 0.13333333333333333,
+      "grad_norm": 0.4435427188873291,
+      "learning_rate": 8e-05,
+      "loss": 1.7424,
+      "step": 2392
+    },
+    {
+      "epoch": 0.1333890746934225,
+      "grad_norm": 0.492079496383667,
+      "learning_rate": 8e-05,
+      "loss": 1.8003,
+      "step": 2393
+    },
+    {
+      "epoch": 0.13344481605351172,
+      "grad_norm": 0.47527799010276794,
+      "learning_rate": 8e-05,
+      "loss": 1.8065,
+      "step": 2394
+    },
+    {
+      "epoch": 0.1335005574136009,
+      "grad_norm": 0.48369163274765015,
+      "learning_rate": 8e-05,
+      "loss": 1.7151,
+      "step": 2395
+    },
+    {
+      "epoch": 0.13355629877369007,
+      "grad_norm": 0.47162288427352905,
+      "learning_rate": 8e-05,
+      "loss": 1.7446,
+      "step": 2396
+    },
+    {
+      "epoch": 0.13361204013377925,
+      "grad_norm": 0.45104944705963135,
+      "learning_rate": 8e-05,
+      "loss": 1.8041,
+      "step": 2397
+    },
+    {
+      "epoch": 0.13366778149386846,
+      "grad_norm": 0.48598307371139526,
+      "learning_rate": 8e-05,
+      "loss": 1.6556,
+      "step": 2398
+    },
+    {
+      "epoch": 0.13372352285395764,
+      "grad_norm": 0.4734358787536621,
+      "learning_rate": 8e-05,
+      "loss": 1.7203,
+      "step": 2399
+    },
+    {
+      "epoch": 0.13377926421404682,
+      "grad_norm": 0.46318376064300537,
+      "learning_rate": 8e-05,
+      "loss": 1.5928,
+      "step": 2400
+    },
+    {
+      "epoch": 0.133835005574136,
+      "grad_norm": 0.48137640953063965,
+      "learning_rate": 8e-05,
+      "loss": 1.8884,
+      "step": 2401
+    },
+    {
+      "epoch": 0.1338907469342252,
+      "grad_norm": 0.4602339565753937,
+      "learning_rate": 8e-05,
+      "loss": 1.9138,
+      "step": 2402
+    },
+    {
+      "epoch": 0.13394648829431438,
+      "grad_norm": 0.4798427224159241,
+      "learning_rate": 8e-05,
+      "loss": 1.7582,
+      "step": 2403
+    },
+    {
+      "epoch": 0.13400222965440356,
+      "grad_norm": 0.464205265045166,
+      "learning_rate": 8e-05,
+      "loss": 1.5574,
+      "step": 2404
+    },
+    {
+      "epoch": 0.13405797101449277,
+      "grad_norm": 0.4638040363788605,
+      "learning_rate": 8e-05,
+      "loss": 1.9085,
+      "step": 2405
+    },
+    {
+      "epoch": 0.13411371237458194,
+      "grad_norm": 0.4874891936779022,
+      "learning_rate": 8e-05,
+      "loss": 1.8563,
+      "step": 2406
+    },
+    {
+      "epoch": 0.13416945373467112,
+      "grad_norm": 0.4475204646587372,
+      "learning_rate": 8e-05,
+      "loss": 1.6447,
+      "step": 2407
+    },
+    {
+      "epoch": 0.1342251950947603,
+      "grad_norm": 0.4701690673828125,
+      "learning_rate": 8e-05,
+      "loss": 1.7522,
+      "step": 2408
+    },
+    {
+      "epoch": 0.1342809364548495,
+      "grad_norm": 0.44603896141052246,
+      "learning_rate": 8e-05,
+      "loss": 1.6628,
+      "step": 2409
+    },
+    {
+      "epoch": 0.1343366778149387,
+      "grad_norm": 0.41180136799812317,
+      "learning_rate": 8e-05,
+      "loss": 1.5771,
+      "step": 2410
+    },
+    {
+      "epoch": 0.13439241917502787,
+      "grad_norm": 0.4283235967159271,
+      "learning_rate": 8e-05,
+      "loss": 1.5256,
+      "step": 2411
+    },
+    {
+      "epoch": 0.13444816053511705,
+      "grad_norm": 0.44074901938438416,
+      "learning_rate": 8e-05,
+      "loss": 1.5656,
+      "step": 2412
+    },
+    {
+      "epoch": 0.13450390189520625,
+      "grad_norm": 0.4715913236141205,
+      "learning_rate": 8e-05,
+      "loss": 1.7263,
+      "step": 2413
+    },
+    {
+      "epoch": 0.13455964325529543,
+      "grad_norm": 0.4731295704841614,
+      "learning_rate": 8e-05,
+      "loss": 1.6281,
+      "step": 2414
+    },
+    {
+      "epoch": 0.1346153846153846,
+      "grad_norm": 0.49616292119026184,
+      "learning_rate": 8e-05,
+      "loss": 1.7264,
+      "step": 2415
+    },
+    {
+      "epoch": 0.1346711259754738,
+      "grad_norm": 0.4744231700897217,
+      "learning_rate": 8e-05,
+      "loss": 1.5563,
+      "step": 2416
+    },
+    {
+      "epoch": 0.134726867335563,
+      "grad_norm": 0.5391608476638794,
+      "learning_rate": 8e-05,
+      "loss": 1.9143,
+      "step": 2417
+    },
+    {
+      "epoch": 0.13478260869565217,
+      "grad_norm": 0.47722572088241577,
+      "learning_rate": 8e-05,
+      "loss": 1.7501,
+      "step": 2418
+    },
+    {
+      "epoch": 0.13483835005574135,
+      "grad_norm": 0.45435523986816406,
+      "learning_rate": 8e-05,
+      "loss": 1.5783,
+      "step": 2419
+    },
+    {
+      "epoch": 0.13489409141583056,
+      "grad_norm": 0.49873173236846924,
+      "learning_rate": 8e-05,
+      "loss": 1.4969,
+      "step": 2420
+    },
+    {
+      "epoch": 0.13494983277591974,
+      "grad_norm": 0.4588122069835663,
+      "learning_rate": 8e-05,
+      "loss": 1.6821,
+      "step": 2421
+    },
+    {
+      "epoch": 0.13500557413600892,
+      "grad_norm": 0.5086076259613037,
+      "learning_rate": 8e-05,
+      "loss": 1.6913,
+      "step": 2422
+    },
+    {
+      "epoch": 0.1350613154960981,
+      "grad_norm": 0.49438467621803284,
+      "learning_rate": 8e-05,
+      "loss": 1.7362,
+      "step": 2423
+    },
+    {
+      "epoch": 0.1351170568561873,
+      "grad_norm": 0.44089025259017944,
+      "learning_rate": 8e-05,
+      "loss": 1.5156,
+      "step": 2424
+    },
+    {
+      "epoch": 0.13517279821627648,
+      "grad_norm": 0.4743707776069641,
+      "learning_rate": 8e-05,
+      "loss": 1.725,
+      "step": 2425
+    },
+    {
+      "epoch": 0.13522853957636566,
+      "grad_norm": 0.47016942501068115,
+      "learning_rate": 8e-05,
+      "loss": 1.6863,
+      "step": 2426
+    },
+    {
+      "epoch": 0.13528428093645484,
+      "grad_norm": 0.46018049120903015,
+      "learning_rate": 8e-05,
+      "loss": 1.822,
+      "step": 2427
+    },
+    {
+      "epoch": 0.13534002229654404,
+      "grad_norm": 0.49280330538749695,
+      "learning_rate": 8e-05,
+      "loss": 1.7777,
+      "step": 2428
+    },
+    {
+      "epoch": 0.13539576365663322,
+      "grad_norm": 0.4748382270336151,
+      "learning_rate": 8e-05,
+      "loss": 1.7763,
+      "step": 2429
+    },
+    {
+      "epoch": 0.1354515050167224,
+      "grad_norm": 0.49057191610336304,
+      "learning_rate": 8e-05,
+      "loss": 1.7156,
+      "step": 2430
+    },
+    {
+      "epoch": 0.1355072463768116,
+      "grad_norm": 0.4892786145210266,
+      "learning_rate": 8e-05,
+      "loss": 1.7613,
+      "step": 2431
+    },
+    {
+      "epoch": 0.1355629877369008,
+      "grad_norm": 0.48499733209609985,
+      "learning_rate": 8e-05,
+      "loss": 1.9075,
+      "step": 2432
+    },
+    {
+      "epoch": 0.13561872909698997,
+      "grad_norm": 0.49811092019081116,
+      "learning_rate": 8e-05,
+      "loss": 1.8027,
+      "step": 2433
+    },
+    {
+      "epoch": 0.13567447045707914,
+      "grad_norm": 0.46384453773498535,
+      "learning_rate": 8e-05,
+      "loss": 1.7716,
+      "step": 2434
+    },
+    {
+      "epoch": 0.13573021181716835,
+      "grad_norm": 0.4836145341396332,
+      "learning_rate": 8e-05,
+      "loss": 1.7956,
+      "step": 2435
+    },
+    {
+      "epoch": 0.13578595317725753,
+      "grad_norm": 0.5062441229820251,
+      "learning_rate": 8e-05,
+      "loss": 2.0621,
+      "step": 2436
+    },
+    {
+      "epoch": 0.1358416945373467,
+      "grad_norm": 0.4342644512653351,
+      "learning_rate": 8e-05,
+      "loss": 1.5002,
+      "step": 2437
+    },
+    {
+      "epoch": 0.1358974358974359,
+      "grad_norm": 0.45354142785072327,
+      "learning_rate": 8e-05,
+      "loss": 1.6726,
+      "step": 2438
+    },
+    {
+      "epoch": 0.1359531772575251,
+      "grad_norm": 0.4888480007648468,
+      "learning_rate": 8e-05,
+      "loss": 2.021,
+      "step": 2439
+    },
+    {
+      "epoch": 0.13600891861761427,
+      "grad_norm": 0.4443151652812958,
+      "learning_rate": 8e-05,
+      "loss": 1.6286,
+      "step": 2440
+    },
+    {
+      "epoch": 0.13606465997770345,
+      "grad_norm": 0.4630712866783142,
+      "learning_rate": 8e-05,
+      "loss": 1.8708,
+      "step": 2441
+    },
+    {
+      "epoch": 0.13612040133779263,
+      "grad_norm": 0.4247455298900604,
+      "learning_rate": 8e-05,
+      "loss": 1.5506,
+      "step": 2442
+    },
+    {
+      "epoch": 0.13617614269788184,
+      "grad_norm": 0.47813916206359863,
+      "learning_rate": 8e-05,
+      "loss": 1.6775,
+      "step": 2443
+    },
+    {
+      "epoch": 0.13623188405797101,
+      "grad_norm": 0.5136623382568359,
+      "learning_rate": 8e-05,
+      "loss": 1.6883,
+      "step": 2444
+    },
+    {
+      "epoch": 0.1362876254180602,
+      "grad_norm": 0.448414146900177,
+      "learning_rate": 8e-05,
+      "loss": 1.7714,
+      "step": 2445
+    },
+    {
+      "epoch": 0.1363433667781494,
+      "grad_norm": 0.4655492603778839,
+      "learning_rate": 8e-05,
+      "loss": 1.7398,
+      "step": 2446
+    },
+    {
+      "epoch": 0.13639910813823858,
+      "grad_norm": 0.4698059856891632,
+      "learning_rate": 8e-05,
+      "loss": 1.6876,
+      "step": 2447
+    },
+    {
+      "epoch": 0.13645484949832776,
+      "grad_norm": 0.4656384587287903,
+      "learning_rate": 8e-05,
+      "loss": 1.7179,
+      "step": 2448
+    },
+    {
+      "epoch": 0.13651059085841694,
+      "grad_norm": 0.4606095552444458,
+      "learning_rate": 8e-05,
+      "loss": 1.6952,
+      "step": 2449
+    },
+    {
+      "epoch": 0.13656633221850614,
+      "grad_norm": 0.4883759617805481,
+      "learning_rate": 8e-05,
+      "loss": 1.7862,
+      "step": 2450
+    },
+    {
+      "epoch": 0.13662207357859532,
+      "grad_norm": 0.4531194567680359,
+      "learning_rate": 8e-05,
+      "loss": 1.6778,
+      "step": 2451
+    },
+    {
+      "epoch": 0.1366778149386845,
+      "grad_norm": 0.4543571472167969,
+      "learning_rate": 8e-05,
+      "loss": 1.5263,
+      "step": 2452
+    },
+    {
+      "epoch": 0.13673355629877368,
+      "grad_norm": 0.49247556924819946,
+      "learning_rate": 8e-05,
+      "loss": 1.6838,
+      "step": 2453
+    },
+    {
+      "epoch": 0.13678929765886289,
+      "grad_norm": 0.46396002173423767,
+      "learning_rate": 8e-05,
+      "loss": 1.6609,
+      "step": 2454
+    },
+    {
+      "epoch": 0.13684503901895206,
+      "grad_norm": 0.46788784861564636,
+      "learning_rate": 8e-05,
+      "loss": 1.7977,
+      "step": 2455
+    },
+    {
+      "epoch": 0.13690078037904124,
+      "grad_norm": 0.4540233910083771,
+      "learning_rate": 8e-05,
+      "loss": 1.4826,
+      "step": 2456
+    },
+    {
+      "epoch": 0.13695652173913042,
+      "grad_norm": 0.4384143650531769,
+      "learning_rate": 8e-05,
+      "loss": 1.5633,
+      "step": 2457
+    },
+    {
+      "epoch": 0.13701226309921963,
+      "grad_norm": 0.4648273289203644,
+      "learning_rate": 8e-05,
+      "loss": 1.7519,
+      "step": 2458
+    },
+    {
+      "epoch": 0.1370680044593088,
+      "grad_norm": 0.4895458221435547,
+      "learning_rate": 8e-05,
+      "loss": 1.7789,
+      "step": 2459
+    },
+    {
+      "epoch": 0.13712374581939799,
+      "grad_norm": 0.4567234516143799,
+      "learning_rate": 8e-05,
+      "loss": 1.7469,
+      "step": 2460
+    },
+    {
+      "epoch": 0.1371794871794872,
+      "grad_norm": 0.48818498849868774,
+      "learning_rate": 8e-05,
+      "loss": 1.8405,
+      "step": 2461
+    },
+    {
+      "epoch": 0.13723522853957637,
+      "grad_norm": 0.4511561393737793,
+      "learning_rate": 8e-05,
+      "loss": 1.7594,
+      "step": 2462
+    },
+    {
+      "epoch": 0.13729096989966555,
+      "grad_norm": 0.4609898626804352,
+      "learning_rate": 8e-05,
+      "loss": 1.5258,
+      "step": 2463
+    },
+    {
+      "epoch": 0.13734671125975473,
+      "grad_norm": 0.47602224349975586,
+      "learning_rate": 8e-05,
+      "loss": 1.7372,
+      "step": 2464
+    },
+    {
+      "epoch": 0.13740245261984393,
+      "grad_norm": 0.49551478028297424,
+      "learning_rate": 8e-05,
+      "loss": 1.8406,
+      "step": 2465
+    },
+    {
+      "epoch": 0.1374581939799331,
+      "grad_norm": 0.49049773812294006,
+      "learning_rate": 8e-05,
+      "loss": 1.7249,
+      "step": 2466
+    },
+    {
+      "epoch": 0.1375139353400223,
+      "grad_norm": 0.4555240869522095,
+      "learning_rate": 8e-05,
+      "loss": 1.6711,
+      "step": 2467
+    },
+    {
+      "epoch": 0.13756967670011147,
+      "grad_norm": 0.5260162949562073,
+      "learning_rate": 8e-05,
+      "loss": 1.6192,
+      "step": 2468
+    },
+    {
+      "epoch": 0.13762541806020068,
+      "grad_norm": 0.44580748677253723,
+      "learning_rate": 8e-05,
+      "loss": 1.7445,
+      "step": 2469
+    },
+    {
+      "epoch": 0.13768115942028986,
+      "grad_norm": 0.5073334574699402,
+      "learning_rate": 8e-05,
+      "loss": 1.7509,
+      "step": 2470
+    },
+    {
+      "epoch": 0.13773690078037903,
+      "grad_norm": 0.44060927629470825,
+      "learning_rate": 8e-05,
+      "loss": 1.7038,
+      "step": 2471
+    },
+    {
+      "epoch": 0.13779264214046824,
+      "grad_norm": 0.4801987409591675,
+      "learning_rate": 8e-05,
+      "loss": 1.8005,
+      "step": 2472
+    },
+    {
+      "epoch": 0.13784838350055742,
+      "grad_norm": 0.47730058431625366,
+      "learning_rate": 8e-05,
+      "loss": 1.7899,
+      "step": 2473
+    },
+    {
+      "epoch": 0.1379041248606466,
+      "grad_norm": 0.4480297863483429,
+      "learning_rate": 8e-05,
+      "loss": 1.6313,
+      "step": 2474
+    },
+    {
+      "epoch": 0.13795986622073578,
+      "grad_norm": 0.4790990948677063,
+      "learning_rate": 8e-05,
+      "loss": 1.7961,
+      "step": 2475
+    },
+    {
+      "epoch": 0.13801560758082498,
+      "grad_norm": 0.4796399176120758,
+      "learning_rate": 8e-05,
+      "loss": 1.9536,
+      "step": 2476
+    },
+    {
+      "epoch": 0.13807134894091416,
+      "grad_norm": 0.43935978412628174,
+      "learning_rate": 8e-05,
+      "loss": 1.7373,
+      "step": 2477
+    },
+    {
+      "epoch": 0.13812709030100334,
+      "grad_norm": 0.48501768708229065,
+      "learning_rate": 8e-05,
+      "loss": 1.93,
+      "step": 2478
+    },
+    {
+      "epoch": 0.13818283166109252,
+      "grad_norm": 0.43100371956825256,
+      "learning_rate": 8e-05,
+      "loss": 1.6818,
+      "step": 2479
+    },
+    {
+      "epoch": 0.13823857302118173,
+      "grad_norm": 0.4330509901046753,
+      "learning_rate": 8e-05,
+      "loss": 1.4093,
+      "step": 2480
+    },
+    {
+      "epoch": 0.1382943143812709,
+      "grad_norm": 0.4759175181388855,
+      "learning_rate": 8e-05,
+      "loss": 1.9259,
+      "step": 2481
+    },
+    {
+      "epoch": 0.13835005574136008,
+      "grad_norm": 0.45059260725975037,
+      "learning_rate": 8e-05,
+      "loss": 1.7486,
+      "step": 2482
+    },
+    {
+      "epoch": 0.13840579710144926,
+      "grad_norm": 0.4803008735179901,
+      "learning_rate": 8e-05,
+      "loss": 1.7893,
+      "step": 2483
+    },
+    {
+      "epoch": 0.13846153846153847,
+      "grad_norm": 0.46275773644447327,
+      "learning_rate": 8e-05,
+      "loss": 1.7574,
+      "step": 2484
+    },
+    {
+      "epoch": 0.13851727982162765,
+      "grad_norm": 0.5310719609260559,
+      "learning_rate": 8e-05,
+      "loss": 1.8372,
+      "step": 2485
+    },
+    {
+      "epoch": 0.13857302118171683,
+      "grad_norm": 0.4447019398212433,
+      "learning_rate": 8e-05,
+      "loss": 1.6102,
+      "step": 2486
+    },
+    {
+      "epoch": 0.13862876254180603,
+      "grad_norm": 0.48079508543014526,
+      "learning_rate": 8e-05,
+      "loss": 1.745,
+      "step": 2487
+    },
+    {
+      "epoch": 0.1386845039018952,
+      "grad_norm": 0.4743659496307373,
+      "learning_rate": 8e-05,
+      "loss": 1.6541,
+      "step": 2488
+    },
+    {
+      "epoch": 0.1387402452619844,
+      "grad_norm": 0.49217358231544495,
+      "learning_rate": 8e-05,
+      "loss": 1.6408,
+      "step": 2489
+    },
+    {
+      "epoch": 0.13879598662207357,
+      "grad_norm": 0.4995599091053009,
+      "learning_rate": 8e-05,
+      "loss": 1.7086,
+      "step": 2490
+    },
+    {
+      "epoch": 0.13885172798216278,
+      "grad_norm": 0.4510442912578583,
+      "learning_rate": 8e-05,
+      "loss": 1.5491,
+      "step": 2491
+    },
+    {
+      "epoch": 0.13890746934225195,
+      "grad_norm": 0.461445152759552,
+      "learning_rate": 8e-05,
+      "loss": 1.7203,
+      "step": 2492
+    },
+    {
+      "epoch": 0.13896321070234113,
+      "grad_norm": 0.4649452269077301,
+      "learning_rate": 8e-05,
+      "loss": 1.7949,
+      "step": 2493
+    },
+    {
+      "epoch": 0.1390189520624303,
+      "grad_norm": 0.48574158549308777,
+      "learning_rate": 8e-05,
+      "loss": 1.8664,
+      "step": 2494
+    },
+    {
+      "epoch": 0.13907469342251952,
+      "grad_norm": 0.47030866146087646,
+      "learning_rate": 8e-05,
+      "loss": 1.4127,
+      "step": 2495
+    },
+    {
+      "epoch": 0.1391304347826087,
+      "grad_norm": 0.44855526089668274,
+      "learning_rate": 8e-05,
+      "loss": 1.7595,
+      "step": 2496
+    },
+    {
+      "epoch": 0.13918617614269788,
+      "grad_norm": 0.46104398369789124,
+      "learning_rate": 8e-05,
+      "loss": 1.8975,
+      "step": 2497
+    },
+    {
+      "epoch": 0.13924191750278705,
+      "grad_norm": 0.4790503978729248,
+      "learning_rate": 8e-05,
+      "loss": 1.7316,
+      "step": 2498
+    },
+    {
+      "epoch": 0.13929765886287626,
+      "grad_norm": 0.4176745116710663,
+      "learning_rate": 8e-05,
+      "loss": 1.7592,
+      "step": 2499
+    },
+    {
+      "epoch": 0.13935340022296544,
+      "grad_norm": 0.5315077900886536,
+      "learning_rate": 8e-05,
+      "loss": 2.0945,
+      "step": 2500
+    },
+    {
+      "epoch": 0.13940914158305462,
+      "grad_norm": 0.47703084349632263,
+      "learning_rate": 8e-05,
+      "loss": 1.6344,
+      "step": 2501
+    },
+    {
+      "epoch": 0.13946488294314383,
+      "grad_norm": 0.42754313349723816,
+      "learning_rate": 8e-05,
+      "loss": 1.5776,
+      "step": 2502
+    },
+    {
+      "epoch": 0.139520624303233,
+      "grad_norm": 0.4698762893676758,
+      "learning_rate": 8e-05,
+      "loss": 1.7629,
+      "step": 2503
+    },
+    {
+      "epoch": 0.13957636566332218,
+      "grad_norm": 0.4266977906227112,
+      "learning_rate": 8e-05,
+      "loss": 1.5624,
+      "step": 2504
+    },
+    {
+      "epoch": 0.13963210702341136,
+      "grad_norm": 0.48394453525543213,
+      "learning_rate": 8e-05,
+      "loss": 1.9536,
+      "step": 2505
+    },
+    {
+      "epoch": 0.13968784838350057,
+      "grad_norm": 0.4214484691619873,
+      "learning_rate": 8e-05,
+      "loss": 1.5147,
+      "step": 2506
+    },
+    {
+      "epoch": 0.13974358974358975,
+      "grad_norm": 0.4360334575176239,
+      "learning_rate": 8e-05,
+      "loss": 1.5634,
+      "step": 2507
+    },
+    {
+      "epoch": 0.13979933110367893,
+      "grad_norm": 0.4772351086139679,
+      "learning_rate": 8e-05,
+      "loss": 1.8834,
+      "step": 2508
+    },
+    {
+      "epoch": 0.1398550724637681,
+      "grad_norm": 0.42568275332450867,
+      "learning_rate": 8e-05,
+      "loss": 1.4775,
+      "step": 2509
+    },
+    {
+      "epoch": 0.1399108138238573,
+      "grad_norm": 0.4546456038951874,
+      "learning_rate": 8e-05,
+      "loss": 1.6071,
+      "step": 2510
+    },
+    {
+      "epoch": 0.1399665551839465,
+      "grad_norm": 0.4755224883556366,
+      "learning_rate": 8e-05,
+      "loss": 1.7269,
+      "step": 2511
+    },
+    {
+      "epoch": 0.14002229654403567,
+      "grad_norm": 0.4712219834327698,
+      "learning_rate": 8e-05,
+      "loss": 1.6659,
+      "step": 2512
+    },
+    {
+      "epoch": 0.14007803790412485,
+      "grad_norm": 0.4919564127922058,
+      "learning_rate": 8e-05,
+      "loss": 1.702,
+      "step": 2513
+    },
+    {
+      "epoch": 0.14013377926421405,
+      "grad_norm": 0.5029149055480957,
+      "learning_rate": 8e-05,
+      "loss": 1.8241,
+      "step": 2514
+    },
+    {
+      "epoch": 0.14018952062430323,
+      "grad_norm": 0.45856478810310364,
+      "learning_rate": 8e-05,
+      "loss": 1.7345,
+      "step": 2515
+    },
+    {
+      "epoch": 0.1402452619843924,
+      "grad_norm": 0.44365623593330383,
+      "learning_rate": 8e-05,
+      "loss": 1.5347,
+      "step": 2516
+    },
+    {
+      "epoch": 0.14030100334448162,
+      "grad_norm": 0.4528595507144928,
+      "learning_rate": 8e-05,
+      "loss": 1.7813,
+      "step": 2517
+    },
+    {
+      "epoch": 0.1403567447045708,
+      "grad_norm": 0.44543299078941345,
+      "learning_rate": 8e-05,
+      "loss": 1.4169,
+      "step": 2518
+    },
+    {
+      "epoch": 0.14041248606465997,
+      "grad_norm": 0.5014076828956604,
+      "learning_rate": 8e-05,
+      "loss": 1.9818,
+      "step": 2519
+    },
+    {
+      "epoch": 0.14046822742474915,
+      "grad_norm": 0.4682339131832123,
+      "learning_rate": 8e-05,
+      "loss": 1.4848,
+      "step": 2520
+    },
+    {
+      "epoch": 0.14052396878483836,
+      "grad_norm": 0.4805692434310913,
+      "learning_rate": 8e-05,
+      "loss": 1.7424,
+      "step": 2521
+    },
+    {
+      "epoch": 0.14057971014492754,
+      "grad_norm": 0.4191701114177704,
+      "learning_rate": 8e-05,
+      "loss": 1.6192,
+      "step": 2522
+    },
+    {
+      "epoch": 0.14063545150501672,
+      "grad_norm": 0.45228466391563416,
+      "learning_rate": 8e-05,
+      "loss": 1.5186,
+      "step": 2523
+    },
+    {
+      "epoch": 0.1406911928651059,
+      "grad_norm": 0.4798497259616852,
+      "learning_rate": 8e-05,
+      "loss": 1.7834,
+      "step": 2524
+    },
+    {
+      "epoch": 0.1407469342251951,
+      "grad_norm": 0.4433005452156067,
+      "learning_rate": 8e-05,
+      "loss": 1.7957,
+      "step": 2525
+    },
+    {
+      "epoch": 0.14080267558528428,
+      "grad_norm": 0.4283502995967865,
+      "learning_rate": 8e-05,
+      "loss": 1.5706,
+      "step": 2526
+    },
+    {
+      "epoch": 0.14085841694537346,
+      "grad_norm": 0.41895848512649536,
+      "learning_rate": 8e-05,
+      "loss": 1.6454,
+      "step": 2527
+    },
+    {
+      "epoch": 0.14091415830546267,
+      "grad_norm": 0.48576539754867554,
+      "learning_rate": 8e-05,
+      "loss": 1.7753,
+      "step": 2528
+    },
+    {
+      "epoch": 0.14096989966555185,
+      "grad_norm": 0.44927990436553955,
+      "learning_rate": 8e-05,
+      "loss": 1.5163,
+      "step": 2529
+    },
+    {
+      "epoch": 0.14102564102564102,
+      "grad_norm": 0.48466259241104126,
+      "learning_rate": 8e-05,
+      "loss": 1.8365,
+      "step": 2530
+    },
+    {
+      "epoch": 0.1410813823857302,
+      "grad_norm": 0.45357945561408997,
+      "learning_rate": 8e-05,
+      "loss": 1.7663,
+      "step": 2531
+    },
+    {
+      "epoch": 0.1411371237458194,
+      "grad_norm": 0.49716877937316895,
+      "learning_rate": 8e-05,
+      "loss": 1.7604,
+      "step": 2532
+    },
+    {
+      "epoch": 0.1411928651059086,
+      "grad_norm": 0.49428150057792664,
+      "learning_rate": 8e-05,
+      "loss": 1.996,
+      "step": 2533
+    },
+    {
+      "epoch": 0.14124860646599777,
+      "grad_norm": 0.4324924945831299,
+      "learning_rate": 8e-05,
+      "loss": 1.5958,
+      "step": 2534
+    },
+    {
+      "epoch": 0.14130434782608695,
+      "grad_norm": 0.4904332160949707,
+      "learning_rate": 8e-05,
+      "loss": 1.8865,
+      "step": 2535
+    },
+    {
+      "epoch": 0.14136008918617615,
+      "grad_norm": 0.436411052942276,
+      "learning_rate": 8e-05,
+      "loss": 1.6264,
+      "step": 2536
+    },
+    {
+      "epoch": 0.14141583054626533,
+      "grad_norm": 0.5399770140647888,
+      "learning_rate": 8e-05,
+      "loss": 1.897,
+      "step": 2537
+    },
+    {
+      "epoch": 0.1414715719063545,
+      "grad_norm": 0.4226623475551605,
+      "learning_rate": 8e-05,
+      "loss": 1.5677,
+      "step": 2538
+    },
+    {
+      "epoch": 0.1415273132664437,
+      "grad_norm": 0.48872801661491394,
+      "learning_rate": 8e-05,
+      "loss": 1.9442,
+      "step": 2539
+    },
+    {
+      "epoch": 0.1415830546265329,
+      "grad_norm": 0.4366551637649536,
+      "learning_rate": 8e-05,
+      "loss": 1.6084,
+      "step": 2540
+    },
+    {
+      "epoch": 0.14163879598662207,
+      "grad_norm": 0.4629233479499817,
+      "learning_rate": 8e-05,
+      "loss": 1.6951,
+      "step": 2541
+    },
+    {
+      "epoch": 0.14169453734671125,
+      "grad_norm": 0.43385395407676697,
+      "learning_rate": 8e-05,
+      "loss": 1.548,
+      "step": 2542
+    },
+    {
+      "epoch": 0.14175027870680046,
+      "grad_norm": 0.4846147298812866,
+      "learning_rate": 8e-05,
+      "loss": 1.8151,
+      "step": 2543
+    },
+    {
+      "epoch": 0.14180602006688964,
+      "grad_norm": 0.4602271616458893,
+      "learning_rate": 8e-05,
+      "loss": 1.5667,
+      "step": 2544
+    },
+    {
+      "epoch": 0.14186176142697882,
+      "grad_norm": 0.5012955069541931,
+      "learning_rate": 8e-05,
+      "loss": 1.9807,
+      "step": 2545
+    },
+    {
+      "epoch": 0.141917502787068,
+      "grad_norm": 0.45967939496040344,
+      "learning_rate": 8e-05,
+      "loss": 1.7745,
+      "step": 2546
+    },
+    {
+      "epoch": 0.1419732441471572,
+      "grad_norm": 0.46281909942626953,
+      "learning_rate": 8e-05,
+      "loss": 1.7764,
+      "step": 2547
+    },
+    {
+      "epoch": 0.14202898550724638,
+      "grad_norm": 0.4586803913116455,
+      "learning_rate": 8e-05,
+      "loss": 1.822,
+      "step": 2548
+    },
+    {
+      "epoch": 0.14208472686733556,
+      "grad_norm": 0.4685911536216736,
+      "learning_rate": 8e-05,
+      "loss": 1.6216,
+      "step": 2549
+    },
+    {
+      "epoch": 0.14214046822742474,
+      "grad_norm": 0.4778120517730713,
+      "learning_rate": 8e-05,
+      "loss": 1.8606,
+      "step": 2550
+    },
+    {
+      "epoch": 0.14219620958751394,
+      "grad_norm": 0.4755409359931946,
+      "learning_rate": 8e-05,
+      "loss": 1.7405,
+      "step": 2551
+    },
+    {
+      "epoch": 0.14225195094760312,
+      "grad_norm": 0.5101373791694641,
+      "learning_rate": 8e-05,
+      "loss": 1.9016,
+      "step": 2552
+    },
+    {
+      "epoch": 0.1423076923076923,
+      "grad_norm": 0.4830615520477295,
+      "learning_rate": 8e-05,
+      "loss": 1.8319,
+      "step": 2553
+    },
+    {
+      "epoch": 0.14236343366778148,
+      "grad_norm": 0.4818636476993561,
+      "learning_rate": 8e-05,
+      "loss": 1.7987,
+      "step": 2554
+    },
+    {
+      "epoch": 0.1424191750278707,
+      "grad_norm": 0.5467934012413025,
+      "learning_rate": 8e-05,
+      "loss": 2.1335,
+      "step": 2555
+    },
+    {
+      "epoch": 0.14247491638795987,
+      "grad_norm": 0.48569151759147644,
+      "learning_rate": 8e-05,
+      "loss": 1.7904,
+      "step": 2556
+    },
+    {
+      "epoch": 0.14253065774804904,
+      "grad_norm": 0.5114737749099731,
+      "learning_rate": 8e-05,
+      "loss": 1.8073,
+      "step": 2557
+    },
+    {
+      "epoch": 0.14258639910813825,
+      "grad_norm": 0.47948169708251953,
+      "learning_rate": 8e-05,
+      "loss": 1.8756,
+      "step": 2558
+    },
+    {
+      "epoch": 0.14264214046822743,
+      "grad_norm": 0.4516530930995941,
+      "learning_rate": 8e-05,
+      "loss": 1.8219,
+      "step": 2559
+    },
+    {
+      "epoch": 0.1426978818283166,
+      "grad_norm": 0.485946923494339,
+      "learning_rate": 8e-05,
+      "loss": 1.703,
+      "step": 2560
+    },
+    {
+      "epoch": 0.1427536231884058,
+      "grad_norm": 0.4625237286090851,
+      "learning_rate": 8e-05,
+      "loss": 1.7076,
+      "step": 2561
+    },
+    {
+      "epoch": 0.142809364548495,
+      "grad_norm": 0.4589962065219879,
+      "learning_rate": 8e-05,
+      "loss": 1.754,
+      "step": 2562
+    },
+    {
+      "epoch": 0.14286510590858417,
+      "grad_norm": 0.45477667450904846,
+      "learning_rate": 8e-05,
+      "loss": 1.6463,
+      "step": 2563
+    },
+    {
+      "epoch": 0.14292084726867335,
+      "grad_norm": 0.46879589557647705,
+      "learning_rate": 8e-05,
+      "loss": 1.6575,
+      "step": 2564
+    },
+    {
+      "epoch": 0.14297658862876253,
+      "grad_norm": 0.4632941484451294,
+      "learning_rate": 8e-05,
+      "loss": 1.9237,
+      "step": 2565
+    },
+    {
+      "epoch": 0.14303232998885174,
+      "grad_norm": 0.5111912488937378,
+      "learning_rate": 8e-05,
+      "loss": 1.7429,
+      "step": 2566
+    },
+    {
+      "epoch": 0.14308807134894092,
+      "grad_norm": 0.5094893574714661,
+      "learning_rate": 8e-05,
+      "loss": 1.8616,
+      "step": 2567
+    },
+    {
+      "epoch": 0.1431438127090301,
+      "grad_norm": 0.5143493413925171,
+      "learning_rate": 8e-05,
+      "loss": 1.4942,
+      "step": 2568
+    },
+    {
+      "epoch": 0.1431995540691193,
+      "grad_norm": 0.44312921166419983,
+      "learning_rate": 8e-05,
+      "loss": 1.6771,
+      "step": 2569
+    },
+    {
+      "epoch": 0.14325529542920848,
+      "grad_norm": 0.46118608117103577,
+      "learning_rate": 8e-05,
+      "loss": 1.6036,
+      "step": 2570
+    },
+    {
+      "epoch": 0.14331103678929766,
+      "grad_norm": 0.44793501496315,
+      "learning_rate": 8e-05,
+      "loss": 1.5554,
+      "step": 2571
+    },
+    {
+      "epoch": 0.14336677814938684,
+      "grad_norm": 0.5015727281570435,
+      "learning_rate": 8e-05,
+      "loss": 1.7773,
+      "step": 2572
+    },
+    {
+      "epoch": 0.14342251950947604,
+      "grad_norm": 0.4987427890300751,
+      "learning_rate": 8e-05,
+      "loss": 2.0529,
+      "step": 2573
+    },
+    {
+      "epoch": 0.14347826086956522,
+      "grad_norm": 0.4836692810058594,
+      "learning_rate": 8e-05,
+      "loss": 1.6583,
+      "step": 2574
+    },
+    {
+      "epoch": 0.1435340022296544,
+      "grad_norm": 0.4915635585784912,
+      "learning_rate": 8e-05,
+      "loss": 1.5852,
+      "step": 2575
+    },
+    {
+      "epoch": 0.14358974358974358,
+      "grad_norm": 0.4204218089580536,
+      "learning_rate": 8e-05,
+      "loss": 1.4971,
+      "step": 2576
+    },
+    {
+      "epoch": 0.14364548494983279,
+      "grad_norm": 0.5908351540565491,
+      "learning_rate": 8e-05,
+      "loss": 1.9646,
+      "step": 2577
+    },
+    {
+      "epoch": 0.14370122630992196,
+      "grad_norm": 0.4868652820587158,
+      "learning_rate": 8e-05,
+      "loss": 1.9946,
+      "step": 2578
+    },
+    {
+      "epoch": 0.14375696767001114,
+      "grad_norm": 0.5083109736442566,
+      "learning_rate": 8e-05,
+      "loss": 1.9526,
+      "step": 2579
+    },
+    {
+      "epoch": 0.14381270903010032,
+      "grad_norm": 0.45809340476989746,
+      "learning_rate": 8e-05,
+      "loss": 1.6082,
+      "step": 2580
+    },
+    {
+      "epoch": 0.14386845039018953,
+      "grad_norm": 0.4436138868331909,
+      "learning_rate": 8e-05,
+      "loss": 1.5737,
+      "step": 2581
+    },
+    {
+      "epoch": 0.1439241917502787,
+      "grad_norm": 0.5178206562995911,
+      "learning_rate": 8e-05,
+      "loss": 1.7271,
+      "step": 2582
+    },
+    {
+      "epoch": 0.14397993311036789,
+      "grad_norm": 0.4951595366001129,
+      "learning_rate": 8e-05,
+      "loss": 1.8733,
+      "step": 2583
+    },
+    {
+      "epoch": 0.1440356744704571,
+      "grad_norm": 0.47114184498786926,
+      "learning_rate": 8e-05,
+      "loss": 1.8224,
+      "step": 2584
+    },
+    {
+      "epoch": 0.14409141583054627,
+      "grad_norm": 0.4795887768268585,
+      "learning_rate": 8e-05,
+      "loss": 1.7847,
+      "step": 2585
+    },
+    {
+      "epoch": 0.14414715719063545,
+      "grad_norm": 0.43819573521614075,
+      "learning_rate": 8e-05,
+      "loss": 1.5203,
+      "step": 2586
+    },
+    {
+      "epoch": 0.14420289855072463,
+      "grad_norm": 0.4707789123058319,
+      "learning_rate": 8e-05,
+      "loss": 1.622,
+      "step": 2587
+    },
+    {
+      "epoch": 0.14425863991081384,
+      "grad_norm": 0.5049995183944702,
+      "learning_rate": 8e-05,
+      "loss": 1.7843,
+      "step": 2588
+    },
+    {
+      "epoch": 0.144314381270903,
+      "grad_norm": 0.5321909785270691,
+      "learning_rate": 8e-05,
+      "loss": 1.9627,
+      "step": 2589
+    },
+    {
+      "epoch": 0.1443701226309922,
+      "grad_norm": 0.49330127239227295,
+      "learning_rate": 8e-05,
+      "loss": 1.8805,
+      "step": 2590
+    },
+    {
+      "epoch": 0.14442586399108137,
+      "grad_norm": 0.5383540391921997,
+      "learning_rate": 8e-05,
+      "loss": 1.8618,
+      "step": 2591
+    },
+    {
+      "epoch": 0.14448160535117058,
+      "grad_norm": 0.45850417017936707,
+      "learning_rate": 8e-05,
+      "loss": 1.7167,
+      "step": 2592
+    },
+    {
+      "epoch": 0.14453734671125976,
+      "grad_norm": 0.475832998752594,
+      "learning_rate": 8e-05,
+      "loss": 1.7077,
+      "step": 2593
+    },
+    {
+      "epoch": 0.14459308807134894,
+      "grad_norm": 0.5012633800506592,
+      "learning_rate": 8e-05,
+      "loss": 1.7972,
+      "step": 2594
+    },
+    {
+      "epoch": 0.14464882943143811,
+      "grad_norm": 0.491800457239151,
+      "learning_rate": 8e-05,
+      "loss": 1.7677,
+      "step": 2595
+    },
+    {
+      "epoch": 0.14470457079152732,
+      "grad_norm": 0.4740968346595764,
+      "learning_rate": 8e-05,
+      "loss": 1.7764,
+      "step": 2596
+    },
+    {
+      "epoch": 0.1447603121516165,
+      "grad_norm": 0.44466421008110046,
+      "learning_rate": 8e-05,
+      "loss": 1.7242,
+      "step": 2597
+    },
+    {
+      "epoch": 0.14481605351170568,
+      "grad_norm": 0.4517992436885834,
+      "learning_rate": 8e-05,
+      "loss": 1.6083,
+      "step": 2598
+    },
+    {
+      "epoch": 0.14487179487179488,
+      "grad_norm": 0.48363086581230164,
+      "learning_rate": 8e-05,
+      "loss": 1.8892,
+      "step": 2599
+    },
+    {
+      "epoch": 0.14492753623188406,
+      "grad_norm": 0.4761711359024048,
+      "learning_rate": 8e-05,
+      "loss": 1.8833,
+      "step": 2600
+    },
+    {
+      "epoch": 0.14498327759197324,
+      "grad_norm": 0.47965916991233826,
+      "learning_rate": 8e-05,
+      "loss": 1.9019,
+      "step": 2601
+    },
+    {
+      "epoch": 0.14503901895206242,
+      "grad_norm": 0.4785793125629425,
+      "learning_rate": 8e-05,
+      "loss": 1.7075,
+      "step": 2602
+    },
+    {
+      "epoch": 0.14509476031215163,
+      "grad_norm": 0.522576093673706,
+      "learning_rate": 8e-05,
+      "loss": 2.017,
+      "step": 2603
+    },
+    {
+      "epoch": 0.1451505016722408,
+      "grad_norm": 0.5296246409416199,
+      "learning_rate": 8e-05,
+      "loss": 1.6347,
+      "step": 2604
+    },
+    {
+      "epoch": 0.14520624303232998,
+      "grad_norm": 0.40535768866539,
+      "learning_rate": 8e-05,
+      "loss": 1.4272,
+      "step": 2605
+    },
+    {
+      "epoch": 0.14526198439241916,
+      "grad_norm": 0.44816961884498596,
+      "learning_rate": 8e-05,
+      "loss": 1.5196,
+      "step": 2606
+    },
+    {
+      "epoch": 0.14531772575250837,
+      "grad_norm": 0.5010104179382324,
+      "learning_rate": 8e-05,
+      "loss": 1.6105,
+      "step": 2607
+    },
+    {
+      "epoch": 0.14537346711259755,
+      "grad_norm": 0.5008467435836792,
+      "learning_rate": 8e-05,
+      "loss": 2.0679,
+      "step": 2608
+    },
+    {
+      "epoch": 0.14542920847268673,
+      "grad_norm": 0.562042772769928,
+      "learning_rate": 8e-05,
+      "loss": 2.1236,
+      "step": 2609
+    },
+    {
+      "epoch": 0.1454849498327759,
+      "grad_norm": 0.5288259387016296,
+      "learning_rate": 8e-05,
+      "loss": 1.8061,
+      "step": 2610
+    },
+    {
+      "epoch": 0.1455406911928651,
+      "grad_norm": 0.5157101154327393,
+      "learning_rate": 8e-05,
+      "loss": 1.8109,
+      "step": 2611
+    },
+    {
+      "epoch": 0.1455964325529543,
+      "grad_norm": 0.5012960433959961,
+      "learning_rate": 8e-05,
+      "loss": 1.7315,
+      "step": 2612
+    },
+    {
+      "epoch": 0.14565217391304347,
+      "grad_norm": 0.5298734903335571,
+      "learning_rate": 8e-05,
+      "loss": 1.8804,
+      "step": 2613
+    },
+    {
+      "epoch": 0.14570791527313268,
+      "grad_norm": 0.49280276894569397,
+      "learning_rate": 8e-05,
+      "loss": 1.809,
+      "step": 2614
+    },
+    {
+      "epoch": 0.14576365663322186,
+      "grad_norm": 0.4701242744922638,
+      "learning_rate": 8e-05,
+      "loss": 1.6592,
+      "step": 2615
+    },
+    {
+      "epoch": 0.14581939799331103,
+      "grad_norm": 0.4502270817756653,
+      "learning_rate": 8e-05,
+      "loss": 1.5762,
+      "step": 2616
+    },
+    {
+      "epoch": 0.1458751393534002,
+      "grad_norm": 0.4606007933616638,
+      "learning_rate": 8e-05,
+      "loss": 1.6542,
+      "step": 2617
+    },
+    {
+      "epoch": 0.14593088071348942,
+      "grad_norm": 0.4372333586215973,
+      "learning_rate": 8e-05,
+      "loss": 1.4818,
+      "step": 2618
+    },
+    {
+      "epoch": 0.1459866220735786,
+      "grad_norm": 0.5123947858810425,
+      "learning_rate": 8e-05,
+      "loss": 1.7565,
+      "step": 2619
+    },
+    {
+      "epoch": 0.14604236343366778,
+      "grad_norm": 0.46434301137924194,
+      "learning_rate": 8e-05,
+      "loss": 1.6585,
+      "step": 2620
+    },
+    {
+      "epoch": 0.14609810479375696,
+      "grad_norm": 0.48808830976486206,
+      "learning_rate": 8e-05,
+      "loss": 1.6373,
+      "step": 2621
+    },
+    {
+      "epoch": 0.14615384615384616,
+      "grad_norm": 0.5246288180351257,
+      "learning_rate": 8e-05,
+      "loss": 1.9852,
+      "step": 2622
+    },
+    {
+      "epoch": 0.14620958751393534,
+      "grad_norm": 0.45331263542175293,
+      "learning_rate": 8e-05,
+      "loss": 1.739,
+      "step": 2623
+    },
+    {
+      "epoch": 0.14626532887402452,
+      "grad_norm": 0.4503915011882782,
+      "learning_rate": 8e-05,
+      "loss": 1.7201,
+      "step": 2624
+    },
+    {
+      "epoch": 0.14632107023411373,
+      "grad_norm": 0.456082820892334,
+      "learning_rate": 8e-05,
+      "loss": 1.5943,
+      "step": 2625
+    },
+    {
+      "epoch": 0.1463768115942029,
+      "grad_norm": 0.43923071026802063,
+      "learning_rate": 8e-05,
+      "loss": 1.5383,
+      "step": 2626
+    },
+    {
+      "epoch": 0.14643255295429208,
+      "grad_norm": 0.5226138830184937,
+      "learning_rate": 8e-05,
+      "loss": 1.822,
+      "step": 2627
+    },
+    {
+      "epoch": 0.14648829431438126,
+      "grad_norm": 0.4432237148284912,
+      "learning_rate": 8e-05,
+      "loss": 1.6528,
+      "step": 2628
+    },
+    {
+      "epoch": 0.14654403567447047,
+      "grad_norm": 0.44176673889160156,
+      "learning_rate": 8e-05,
+      "loss": 1.4834,
+      "step": 2629
+    },
+    {
+      "epoch": 0.14659977703455965,
+      "grad_norm": 0.4683099389076233,
+      "learning_rate": 8e-05,
+      "loss": 1.6887,
+      "step": 2630
+    },
+    {
+      "epoch": 0.14665551839464883,
+      "grad_norm": 0.44806182384490967,
+      "learning_rate": 8e-05,
+      "loss": 1.6024,
+      "step": 2631
+    },
+    {
+      "epoch": 0.146711259754738,
+      "grad_norm": 0.488511323928833,
+      "learning_rate": 8e-05,
+      "loss": 1.9064,
+      "step": 2632
+    },
+    {
+      "epoch": 0.1467670011148272,
+      "grad_norm": 0.4772742986679077,
+      "learning_rate": 8e-05,
+      "loss": 1.8026,
+      "step": 2633
+    },
+    {
+      "epoch": 0.1468227424749164,
+      "grad_norm": 0.4587347209453583,
+      "learning_rate": 8e-05,
+      "loss": 1.6626,
+      "step": 2634
+    },
+    {
+      "epoch": 0.14687848383500557,
+      "grad_norm": 0.5513268113136292,
+      "learning_rate": 8e-05,
+      "loss": 2.0832,
+      "step": 2635
+    },
+    {
+      "epoch": 0.14693422519509475,
+      "grad_norm": 0.44140949845314026,
+      "learning_rate": 8e-05,
+      "loss": 1.6923,
+      "step": 2636
+    },
+    {
+      "epoch": 0.14698996655518395,
+      "grad_norm": 0.4342259168624878,
+      "learning_rate": 8e-05,
+      "loss": 1.5283,
+      "step": 2637
+    },
+    {
+      "epoch": 0.14704570791527313,
+      "grad_norm": 0.5041188597679138,
+      "learning_rate": 8e-05,
+      "loss": 1.8199,
+      "step": 2638
+    },
+    {
+      "epoch": 0.1471014492753623,
+      "grad_norm": 0.4816203713417053,
+      "learning_rate": 8e-05,
+      "loss": 1.8403,
+      "step": 2639
+    },
+    {
+      "epoch": 0.14715719063545152,
+      "grad_norm": 0.5168691873550415,
+      "learning_rate": 8e-05,
+      "loss": 1.9105,
+      "step": 2640
+    },
+    {
+      "epoch": 0.1472129319955407,
+      "grad_norm": 0.47315770387649536,
+      "learning_rate": 8e-05,
+      "loss": 1.7493,
+      "step": 2641
+    },
+    {
+      "epoch": 0.14726867335562988,
+      "grad_norm": 0.4396923780441284,
+      "learning_rate": 8e-05,
+      "loss": 1.4694,
+      "step": 2642
+    },
+    {
+      "epoch": 0.14732441471571905,
+      "grad_norm": 0.4882901906967163,
+      "learning_rate": 8e-05,
+      "loss": 1.6426,
+      "step": 2643
+    },
+    {
+      "epoch": 0.14738015607580826,
+      "grad_norm": 0.5237675905227661,
+      "learning_rate": 8e-05,
+      "loss": 1.9971,
+      "step": 2644
+    },
+    {
+      "epoch": 0.14743589743589744,
+      "grad_norm": 0.4680856466293335,
+      "learning_rate": 8e-05,
+      "loss": 1.6813,
+      "step": 2645
+    },
+    {
+      "epoch": 0.14749163879598662,
+      "grad_norm": 0.4802532494068146,
+      "learning_rate": 8e-05,
+      "loss": 1.6923,
+      "step": 2646
+    },
+    {
+      "epoch": 0.1475473801560758,
+      "grad_norm": 0.4531373381614685,
+      "learning_rate": 8e-05,
+      "loss": 1.6865,
+      "step": 2647
+    },
+    {
+      "epoch": 0.147603121516165,
+      "grad_norm": 0.49141162633895874,
+      "learning_rate": 8e-05,
+      "loss": 1.8942,
+      "step": 2648
+    },
+    {
+      "epoch": 0.14765886287625418,
+      "grad_norm": 0.40621107816696167,
+      "learning_rate": 8e-05,
+      "loss": 1.5562,
+      "step": 2649
+    },
+    {
+      "epoch": 0.14771460423634336,
+      "grad_norm": 0.48760560154914856,
+      "learning_rate": 8e-05,
+      "loss": 1.8392,
+      "step": 2650
+    },
+    {
+      "epoch": 0.14777034559643254,
+      "grad_norm": 0.4867985248565674,
+      "learning_rate": 8e-05,
+      "loss": 1.8491,
+      "step": 2651
+    },
+    {
+      "epoch": 0.14782608695652175,
+      "grad_norm": 0.5120669603347778,
+      "learning_rate": 8e-05,
+      "loss": 1.8304,
+      "step": 2652
+    },
+    {
+      "epoch": 0.14788182831661092,
+      "grad_norm": 0.4798111915588379,
+      "learning_rate": 8e-05,
+      "loss": 1.8607,
+      "step": 2653
+    },
+    {
+      "epoch": 0.1479375696767001,
+      "grad_norm": 0.47370168566703796,
+      "learning_rate": 8e-05,
+      "loss": 1.7137,
+      "step": 2654
+    },
+    {
+      "epoch": 0.1479933110367893,
+      "grad_norm": 0.6196955442428589,
+      "learning_rate": 8e-05,
+      "loss": 1.6156,
+      "step": 2655
+    },
+    {
+      "epoch": 0.1480490523968785,
+      "grad_norm": 0.4761423170566559,
+      "learning_rate": 8e-05,
+      "loss": 1.7644,
+      "step": 2656
+    },
+    {
+      "epoch": 0.14810479375696767,
+      "grad_norm": 0.49045848846435547,
+      "learning_rate": 8e-05,
+      "loss": 1.5832,
+      "step": 2657
+    },
+    {
+      "epoch": 0.14816053511705685,
+      "grad_norm": 0.45145368576049805,
+      "learning_rate": 8e-05,
+      "loss": 1.4705,
+      "step": 2658
+    },
+    {
+      "epoch": 0.14821627647714605,
+      "grad_norm": 0.49962612986564636,
+      "learning_rate": 8e-05,
+      "loss": 1.7623,
+      "step": 2659
+    },
+    {
+      "epoch": 0.14827201783723523,
+      "grad_norm": 0.4893467426300049,
+      "learning_rate": 8e-05,
+      "loss": 1.9686,
+      "step": 2660
+    },
+    {
+      "epoch": 0.1483277591973244,
+      "grad_norm": 0.48408836126327515,
+      "learning_rate": 8e-05,
+      "loss": 1.7615,
+      "step": 2661
+    },
+    {
+      "epoch": 0.1483835005574136,
+      "grad_norm": 0.48030319809913635,
+      "learning_rate": 8e-05,
+      "loss": 2.0596,
+      "step": 2662
+    },
+    {
+      "epoch": 0.1484392419175028,
+      "grad_norm": 0.46270525455474854,
+      "learning_rate": 8e-05,
+      "loss": 1.7686,
+      "step": 2663
+    },
+    {
+      "epoch": 0.14849498327759197,
+      "grad_norm": 0.41890811920166016,
+      "learning_rate": 8e-05,
+      "loss": 1.6958,
+      "step": 2664
+    },
+    {
+      "epoch": 0.14855072463768115,
+      "grad_norm": 0.49203094840049744,
+      "learning_rate": 8e-05,
+      "loss": 1.8678,
+      "step": 2665
+    },
+    {
+      "epoch": 0.14860646599777033,
+      "grad_norm": 0.4484955072402954,
+      "learning_rate": 8e-05,
+      "loss": 1.8809,
+      "step": 2666
+    },
+    {
+      "epoch": 0.14866220735785954,
+      "grad_norm": 0.5041595697402954,
+      "learning_rate": 8e-05,
+      "loss": 1.9027,
+      "step": 2667
+    },
+    {
+      "epoch": 0.14871794871794872,
+      "grad_norm": 0.4210891127586365,
+      "learning_rate": 8e-05,
+      "loss": 1.4913,
+      "step": 2668
+    },
+    {
+      "epoch": 0.1487736900780379,
+      "grad_norm": 0.47749096155166626,
+      "learning_rate": 8e-05,
+      "loss": 1.5191,
+      "step": 2669
+    },
+    {
+      "epoch": 0.1488294314381271,
+      "grad_norm": 0.45108339190483093,
+      "learning_rate": 8e-05,
+      "loss": 1.695,
+      "step": 2670
+    },
+    {
+      "epoch": 0.14888517279821628,
+      "grad_norm": 0.5282843112945557,
+      "learning_rate": 8e-05,
+      "loss": 1.4526,
+      "step": 2671
+    },
+    {
+      "epoch": 0.14894091415830546,
+      "grad_norm": 0.4593731164932251,
+      "learning_rate": 8e-05,
+      "loss": 1.7042,
+      "step": 2672
+    },
+    {
+      "epoch": 0.14899665551839464,
+      "grad_norm": 0.5506393313407898,
+      "learning_rate": 8e-05,
+      "loss": 1.935,
+      "step": 2673
+    },
+    {
+      "epoch": 0.14905239687848384,
+      "grad_norm": 0.5012826919555664,
+      "learning_rate": 8e-05,
+      "loss": 1.45,
+      "step": 2674
+    },
+    {
+      "epoch": 0.14910813823857302,
+      "grad_norm": 0.5560808777809143,
+      "learning_rate": 8e-05,
+      "loss": 2.025,
+      "step": 2675
+    },
+    {
+      "epoch": 0.1491638795986622,
+      "grad_norm": 0.5152798891067505,
+      "learning_rate": 8e-05,
+      "loss": 1.8135,
+      "step": 2676
+    },
+    {
+      "epoch": 0.14921962095875138,
+      "grad_norm": 0.5184229016304016,
+      "learning_rate": 8e-05,
+      "loss": 1.7617,
+      "step": 2677
+    },
+    {
+      "epoch": 0.1492753623188406,
+      "grad_norm": 0.49748098850250244,
+      "learning_rate": 8e-05,
+      "loss": 1.6876,
+      "step": 2678
+    },
+    {
+      "epoch": 0.14933110367892977,
+      "grad_norm": 0.5202342867851257,
+      "learning_rate": 8e-05,
+      "loss": 1.7849,
+      "step": 2679
+    },
+    {
+      "epoch": 0.14938684503901895,
+      "grad_norm": 0.47590065002441406,
+      "learning_rate": 8e-05,
+      "loss": 1.7537,
+      "step": 2680
+    },
+    {
+      "epoch": 0.14944258639910815,
+      "grad_norm": 0.5148934125900269,
+      "learning_rate": 8e-05,
+      "loss": 1.8217,
+      "step": 2681
+    },
+    {
+      "epoch": 0.14949832775919733,
+      "grad_norm": 0.487512469291687,
+      "learning_rate": 8e-05,
+      "loss": 1.8185,
+      "step": 2682
+    },
+    {
+      "epoch": 0.1495540691192865,
+      "grad_norm": 0.505423367023468,
+      "learning_rate": 8e-05,
+      "loss": 1.8296,
+      "step": 2683
+    },
+    {
+      "epoch": 0.1496098104793757,
+      "grad_norm": 0.4880646765232086,
+      "learning_rate": 8e-05,
+      "loss": 1.797,
+      "step": 2684
+    },
+    {
+      "epoch": 0.1496655518394649,
+      "grad_norm": 0.45995259284973145,
+      "learning_rate": 8e-05,
+      "loss": 1.548,
+      "step": 2685
+    },
+    {
+      "epoch": 0.14972129319955407,
+      "grad_norm": 0.48819291591644287,
+      "learning_rate": 8e-05,
+      "loss": 1.6364,
+      "step": 2686
+    },
+    {
+      "epoch": 0.14977703455964325,
+      "grad_norm": 0.47882404923439026,
+      "learning_rate": 8e-05,
+      "loss": 1.5055,
+      "step": 2687
+    },
+    {
+      "epoch": 0.14983277591973243,
+      "grad_norm": 0.437224805355072,
+      "learning_rate": 8e-05,
+      "loss": 1.6657,
+      "step": 2688
+    },
+    {
+      "epoch": 0.14988851727982164,
+      "grad_norm": 0.49094241857528687,
+      "learning_rate": 8e-05,
+      "loss": 1.8073,
+      "step": 2689
+    },
+    {
+      "epoch": 0.14994425863991082,
+      "grad_norm": 0.45183655619621277,
+      "learning_rate": 8e-05,
+      "loss": 1.5465,
+      "step": 2690
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.48996084928512573,
+      "learning_rate": 8e-05,
+      "loss": 1.901,
+      "step": 2691
+    },
+    {
+      "epoch": 0.15005574136008917,
+      "grad_norm": 0.43467774987220764,
+      "learning_rate": 8e-05,
+      "loss": 1.615,
+      "step": 2692
+    },
+    {
+      "epoch": 0.15011148272017838,
+      "grad_norm": 0.460771769285202,
+      "learning_rate": 8e-05,
+      "loss": 1.8058,
+      "step": 2693
+    },
+    {
+      "epoch": 0.15016722408026756,
+      "grad_norm": 0.49444660544395447,
+      "learning_rate": 8e-05,
+      "loss": 1.6476,
+      "step": 2694
+    },
+    {
+      "epoch": 0.15022296544035674,
+      "grad_norm": 0.4313269257545471,
+      "learning_rate": 8e-05,
+      "loss": 1.4598,
+      "step": 2695
+    },
+    {
+      "epoch": 0.15027870680044594,
+      "grad_norm": 0.5513227581977844,
+      "learning_rate": 8e-05,
+      "loss": 2.0741,
+      "step": 2696
+    },
+    {
+      "epoch": 0.15033444816053512,
+      "grad_norm": 0.46913573145866394,
+      "learning_rate": 8e-05,
+      "loss": 1.6384,
+      "step": 2697
+    },
+    {
+      "epoch": 0.1503901895206243,
+      "grad_norm": 0.4985828399658203,
+      "learning_rate": 8e-05,
+      "loss": 1.4213,
+      "step": 2698
+    },
+    {
+      "epoch": 0.15044593088071348,
+      "grad_norm": 0.4544661045074463,
+      "learning_rate": 8e-05,
+      "loss": 1.7351,
+      "step": 2699
+    },
+    {
+      "epoch": 0.1505016722408027,
+      "grad_norm": 0.5336171984672546,
+      "learning_rate": 8e-05,
+      "loss": 1.9037,
+      "step": 2700
+    },
+    {
+      "epoch": 0.15055741360089187,
+      "grad_norm": 0.4664371609687805,
+      "learning_rate": 8e-05,
+      "loss": 1.6724,
+      "step": 2701
+    },
+    {
+      "epoch": 0.15061315496098104,
+      "grad_norm": 0.4967934489250183,
+      "learning_rate": 8e-05,
+      "loss": 1.9133,
+      "step": 2702
+    },
+    {
+      "epoch": 0.15066889632107022,
+      "grad_norm": 0.4723069667816162,
+      "learning_rate": 8e-05,
+      "loss": 1.6718,
+      "step": 2703
+    },
+    {
+      "epoch": 0.15072463768115943,
+      "grad_norm": 0.5051380395889282,
+      "learning_rate": 8e-05,
+      "loss": 1.6417,
+      "step": 2704
+    },
+    {
+      "epoch": 0.1507803790412486,
+      "grad_norm": 0.49244359135627747,
+      "learning_rate": 8e-05,
+      "loss": 1.6002,
+      "step": 2705
+    },
+    {
+      "epoch": 0.1508361204013378,
+      "grad_norm": 0.4505351483821869,
+      "learning_rate": 8e-05,
+      "loss": 1.5397,
+      "step": 2706
+    },
+    {
+      "epoch": 0.15089186176142697,
+      "grad_norm": 0.5007824897766113,
+      "learning_rate": 8e-05,
+      "loss": 1.6409,
+      "step": 2707
+    },
+    {
+      "epoch": 0.15094760312151617,
+      "grad_norm": 0.5030344724655151,
+      "learning_rate": 8e-05,
+      "loss": 1.7374,
+      "step": 2708
+    },
+    {
+      "epoch": 0.15100334448160535,
+      "grad_norm": 0.5079347491264343,
+      "learning_rate": 8e-05,
+      "loss": 1.9303,
+      "step": 2709
+    },
+    {
+      "epoch": 0.15105908584169453,
+      "grad_norm": 0.5016282200813293,
+      "learning_rate": 8e-05,
+      "loss": 2.0209,
+      "step": 2710
+    },
+    {
+      "epoch": 0.15111482720178374,
+      "grad_norm": 0.4660756289958954,
+      "learning_rate": 8e-05,
+      "loss": 1.5367,
+      "step": 2711
+    },
+    {
+      "epoch": 0.15117056856187291,
+      "grad_norm": 0.536495566368103,
+      "learning_rate": 8e-05,
+      "loss": 1.8843,
+      "step": 2712
+    },
+    {
+      "epoch": 0.1512263099219621,
+      "grad_norm": 0.49942296743392944,
+      "learning_rate": 8e-05,
+      "loss": 1.7927,
+      "step": 2713
+    },
+    {
+      "epoch": 0.15128205128205127,
+      "grad_norm": 0.5121089220046997,
+      "learning_rate": 8e-05,
+      "loss": 1.9474,
+      "step": 2714
+    },
+    {
+      "epoch": 0.15133779264214048,
+      "grad_norm": 0.49932828545570374,
+      "learning_rate": 8e-05,
+      "loss": 1.9303,
+      "step": 2715
+    },
+    {
+      "epoch": 0.15139353400222966,
+      "grad_norm": 0.43620043992996216,
+      "learning_rate": 8e-05,
+      "loss": 1.638,
+      "step": 2716
+    },
+    {
+      "epoch": 0.15144927536231884,
+      "grad_norm": 0.491070032119751,
+      "learning_rate": 8e-05,
+      "loss": 1.8854,
+      "step": 2717
+    },
+    {
+      "epoch": 0.15150501672240801,
+      "grad_norm": 0.47080928087234497,
+      "learning_rate": 8e-05,
+      "loss": 1.8339,
+      "step": 2718
+    },
+    {
+      "epoch": 0.15156075808249722,
+      "grad_norm": 0.4873843193054199,
+      "learning_rate": 8e-05,
+      "loss": 1.9661,
+      "step": 2719
+    },
+    {
+      "epoch": 0.1516164994425864,
+      "grad_norm": 0.5007215738296509,
+      "learning_rate": 8e-05,
+      "loss": 1.8296,
+      "step": 2720
+    },
+    {
+      "epoch": 0.15167224080267558,
+      "grad_norm": 0.47283855080604553,
+      "learning_rate": 8e-05,
+      "loss": 1.7406,
+      "step": 2721
+    },
+    {
+      "epoch": 0.15172798216276479,
+      "grad_norm": 0.4671013653278351,
+      "learning_rate": 8e-05,
+      "loss": 1.6189,
+      "step": 2722
+    },
+    {
+      "epoch": 0.15178372352285396,
+      "grad_norm": 0.45656901597976685,
+      "learning_rate": 8e-05,
+      "loss": 1.6288,
+      "step": 2723
+    },
+    {
+      "epoch": 0.15183946488294314,
+      "grad_norm": 0.45888885855674744,
+      "learning_rate": 8e-05,
+      "loss": 1.6675,
+      "step": 2724
+    },
+    {
+      "epoch": 0.15189520624303232,
+      "grad_norm": 0.4929250180721283,
+      "learning_rate": 8e-05,
+      "loss": 1.6704,
+      "step": 2725
+    },
+    {
+      "epoch": 0.15195094760312153,
+      "grad_norm": 0.5203201770782471,
+      "learning_rate": 8e-05,
+      "loss": 1.8276,
+      "step": 2726
+    },
+    {
+      "epoch": 0.1520066889632107,
+      "grad_norm": 0.45683154463768005,
+      "learning_rate": 8e-05,
+      "loss": 1.5656,
+      "step": 2727
+    },
+    {
+      "epoch": 0.15206243032329989,
+      "grad_norm": 0.4747931659221649,
+      "learning_rate": 8e-05,
+      "loss": 1.6466,
+      "step": 2728
+    },
+    {
+      "epoch": 0.15211817168338906,
+      "grad_norm": 0.49119818210601807,
+      "learning_rate": 8e-05,
+      "loss": 1.7029,
+      "step": 2729
+    },
+    {
+      "epoch": 0.15217391304347827,
+      "grad_norm": 0.47732195258140564,
+      "learning_rate": 8e-05,
+      "loss": 1.6918,
+      "step": 2730
+    },
+    {
+      "epoch": 0.15222965440356745,
+      "grad_norm": 0.490499347448349,
+      "learning_rate": 8e-05,
+      "loss": 1.7927,
+      "step": 2731
+    },
+    {
+      "epoch": 0.15228539576365663,
+      "grad_norm": 0.5020274519920349,
+      "learning_rate": 8e-05,
+      "loss": 1.893,
+      "step": 2732
+    },
+    {
+      "epoch": 0.1523411371237458,
+      "grad_norm": 0.46516138315200806,
+      "learning_rate": 8e-05,
+      "loss": 1.6328,
+      "step": 2733
+    },
+    {
+      "epoch": 0.152396878483835,
+      "grad_norm": 0.5290771722793579,
+      "learning_rate": 8e-05,
+      "loss": 2.1387,
+      "step": 2734
+    },
+    {
+      "epoch": 0.1524526198439242,
+      "grad_norm": 0.4662680923938751,
+      "learning_rate": 8e-05,
+      "loss": 1.7573,
+      "step": 2735
+    },
+    {
+      "epoch": 0.15250836120401337,
+      "grad_norm": 0.4540351331233978,
+      "learning_rate": 8e-05,
+      "loss": 1.7422,
+      "step": 2736
+    },
+    {
+      "epoch": 0.15256410256410258,
+      "grad_norm": 0.4451354444026947,
+      "learning_rate": 8e-05,
+      "loss": 1.4756,
+      "step": 2737
+    },
+    {
+      "epoch": 0.15261984392419176,
+      "grad_norm": 0.4653637409210205,
+      "learning_rate": 8e-05,
+      "loss": 1.782,
+      "step": 2738
+    },
+    {
+      "epoch": 0.15267558528428093,
+      "grad_norm": 0.48152562975883484,
+      "learning_rate": 8e-05,
+      "loss": 1.7043,
+      "step": 2739
+    },
+    {
+      "epoch": 0.1527313266443701,
+      "grad_norm": 0.5343277454376221,
+      "learning_rate": 8e-05,
+      "loss": 1.834,
+      "step": 2740
+    },
+    {
+      "epoch": 0.15278706800445932,
+      "grad_norm": 0.48790308833122253,
+      "learning_rate": 8e-05,
+      "loss": 1.7217,
+      "step": 2741
+    },
+    {
+      "epoch": 0.1528428093645485,
+      "grad_norm": 0.4877580404281616,
+      "learning_rate": 8e-05,
+      "loss": 1.771,
+      "step": 2742
+    },
+    {
+      "epoch": 0.15289855072463768,
+      "grad_norm": 0.4496322274208069,
+      "learning_rate": 8e-05,
+      "loss": 1.7451,
+      "step": 2743
+    },
+    {
+      "epoch": 0.15295429208472686,
+      "grad_norm": 0.4950348734855652,
+      "learning_rate": 8e-05,
+      "loss": 1.7627,
+      "step": 2744
+    },
+    {
+      "epoch": 0.15301003344481606,
+      "grad_norm": 0.4993313252925873,
+      "learning_rate": 8e-05,
+      "loss": 1.7715,
+      "step": 2745
+    },
+    {
+      "epoch": 0.15306577480490524,
+      "grad_norm": 0.45644307136535645,
+      "learning_rate": 8e-05,
+      "loss": 1.7968,
+      "step": 2746
+    },
+    {
+      "epoch": 0.15312151616499442,
+      "grad_norm": 0.4572172164916992,
+      "learning_rate": 8e-05,
+      "loss": 1.6494,
+      "step": 2747
+    },
+    {
+      "epoch": 0.1531772575250836,
+      "grad_norm": 0.48790818452835083,
+      "learning_rate": 8e-05,
+      "loss": 1.8427,
+      "step": 2748
+    },
+    {
+      "epoch": 0.1532329988851728,
+      "grad_norm": 0.4811917841434479,
+      "learning_rate": 8e-05,
+      "loss": 1.7571,
+      "step": 2749
+    },
+    {
+      "epoch": 0.15328874024526198,
+      "grad_norm": 0.4466029107570648,
+      "learning_rate": 8e-05,
+      "loss": 1.6303,
+      "step": 2750
+    },
+    {
+      "epoch": 0.15334448160535116,
+      "grad_norm": 0.47539398074150085,
+      "learning_rate": 8e-05,
+      "loss": 1.8565,
+      "step": 2751
+    },
+    {
+      "epoch": 0.15340022296544037,
+      "grad_norm": 0.45929116010665894,
+      "learning_rate": 8e-05,
+      "loss": 1.9393,
+      "step": 2752
+    },
+    {
+      "epoch": 0.15345596432552955,
+      "grad_norm": 0.5932511687278748,
+      "learning_rate": 8e-05,
+      "loss": 1.7947,
+      "step": 2753
+    },
+    {
+      "epoch": 0.15351170568561873,
+      "grad_norm": 0.4963116943836212,
+      "learning_rate": 8e-05,
+      "loss": 1.84,
+      "step": 2754
+    },
+    {
+      "epoch": 0.1535674470457079,
+      "grad_norm": 0.4386138617992401,
+      "learning_rate": 8e-05,
+      "loss": 1.5813,
+      "step": 2755
+    },
+    {
+      "epoch": 0.1536231884057971,
+      "grad_norm": 0.5233179330825806,
+      "learning_rate": 8e-05,
+      "loss": 1.7951,
+      "step": 2756
+    },
+    {
+      "epoch": 0.1536789297658863,
+      "grad_norm": 0.42530587315559387,
+      "learning_rate": 8e-05,
+      "loss": 1.4469,
+      "step": 2757
+    },
+    {
+      "epoch": 0.15373467112597547,
+      "grad_norm": 0.45138615369796753,
+      "learning_rate": 8e-05,
+      "loss": 1.6971,
+      "step": 2758
+    },
+    {
+      "epoch": 0.15379041248606465,
+      "grad_norm": 0.4747151732444763,
+      "learning_rate": 8e-05,
+      "loss": 1.6825,
+      "step": 2759
+    },
+    {
+      "epoch": 0.15384615384615385,
+      "grad_norm": 0.4871062934398651,
+      "learning_rate": 8e-05,
+      "loss": 1.8835,
+      "step": 2760
+    },
+    {
+      "epoch": 0.15390189520624303,
+      "grad_norm": 0.4672834277153015,
+      "learning_rate": 8e-05,
+      "loss": 1.5864,
+      "step": 2761
+    },
+    {
+      "epoch": 0.1539576365663322,
+      "grad_norm": 0.5292240381240845,
+      "learning_rate": 8e-05,
+      "loss": 2.1529,
+      "step": 2762
+    },
+    {
+      "epoch": 0.1540133779264214,
+      "grad_norm": 0.5465215444564819,
+      "learning_rate": 8e-05,
+      "loss": 1.8391,
+      "step": 2763
+    },
+    {
+      "epoch": 0.1540691192865106,
+      "grad_norm": 0.4289812445640564,
+      "learning_rate": 8e-05,
+      "loss": 1.3537,
+      "step": 2764
+    },
+    {
+      "epoch": 0.15412486064659978,
+      "grad_norm": 0.4715195596218109,
+      "learning_rate": 8e-05,
+      "loss": 1.7668,
+      "step": 2765
+    },
+    {
+      "epoch": 0.15418060200668895,
+      "grad_norm": 0.5369305610656738,
+      "learning_rate": 8e-05,
+      "loss": 1.9698,
+      "step": 2766
+    },
+    {
+      "epoch": 0.15423634336677816,
+      "grad_norm": 0.4952063262462616,
+      "learning_rate": 8e-05,
+      "loss": 1.8069,
+      "step": 2767
+    },
+    {
+      "epoch": 0.15429208472686734,
+      "grad_norm": 0.46554696559906006,
+      "learning_rate": 8e-05,
+      "loss": 1.7292,
+      "step": 2768
+    },
+    {
+      "epoch": 0.15434782608695652,
+      "grad_norm": 0.48321375250816345,
+      "learning_rate": 8e-05,
+      "loss": 1.7208,
+      "step": 2769
+    },
+    {
+      "epoch": 0.1544035674470457,
+      "grad_norm": 0.45590123534202576,
+      "learning_rate": 8e-05,
+      "loss": 1.7738,
+      "step": 2770
+    },
+    {
+      "epoch": 0.1544593088071349,
+      "grad_norm": 0.558228075504303,
+      "learning_rate": 8e-05,
+      "loss": 2.0212,
+      "step": 2771
+    },
+    {
+      "epoch": 0.15451505016722408,
+      "grad_norm": 0.44301968812942505,
+      "learning_rate": 8e-05,
+      "loss": 1.6274,
+      "step": 2772
+    },
+    {
+      "epoch": 0.15457079152731326,
+      "grad_norm": 0.5060457587242126,
+      "learning_rate": 8e-05,
+      "loss": 1.9058,
+      "step": 2773
+    },
+    {
+      "epoch": 0.15462653288740244,
+      "grad_norm": 0.4313315153121948,
+      "learning_rate": 8e-05,
+      "loss": 1.68,
+      "step": 2774
+    },
+    {
+      "epoch": 0.15468227424749165,
+      "grad_norm": 0.47403407096862793,
+      "learning_rate": 8e-05,
+      "loss": 1.7216,
+      "step": 2775
+    },
+    {
+      "epoch": 0.15473801560758083,
+      "grad_norm": 0.48227450251579285,
+      "learning_rate": 8e-05,
+      "loss": 1.8895,
+      "step": 2776
+    },
+    {
+      "epoch": 0.15479375696767,
+      "grad_norm": 0.5059760212898254,
+      "learning_rate": 8e-05,
+      "loss": 1.9685,
+      "step": 2777
+    },
+    {
+      "epoch": 0.1548494983277592,
+      "grad_norm": 0.5228263139724731,
+      "learning_rate": 8e-05,
+      "loss": 2.211,
+      "step": 2778
+    },
+    {
+      "epoch": 0.1549052396878484,
+      "grad_norm": 0.46741005778312683,
+      "learning_rate": 8e-05,
+      "loss": 1.6471,
+      "step": 2779
+    },
+    {
+      "epoch": 0.15496098104793757,
+      "grad_norm": 0.48850780725479126,
+      "learning_rate": 8e-05,
+      "loss": 1.8356,
+      "step": 2780
+    },
+    {
+      "epoch": 0.15501672240802675,
+      "grad_norm": 0.5071626305580139,
+      "learning_rate": 8e-05,
+      "loss": 1.6518,
+      "step": 2781
+    },
+    {
+      "epoch": 0.15507246376811595,
+      "grad_norm": 0.47851163148880005,
+      "learning_rate": 8e-05,
+      "loss": 1.7904,
+      "step": 2782
+    },
+    {
+      "epoch": 0.15512820512820513,
+      "grad_norm": 0.45001041889190674,
+      "learning_rate": 8e-05,
+      "loss": 1.672,
+      "step": 2783
+    },
+    {
+      "epoch": 0.1551839464882943,
+      "grad_norm": 0.4666180908679962,
+      "learning_rate": 8e-05,
+      "loss": 1.6842,
+      "step": 2784
+    },
+    {
+      "epoch": 0.1552396878483835,
+      "grad_norm": 0.4250207841396332,
+      "learning_rate": 8e-05,
+      "loss": 1.4794,
+      "step": 2785
+    },
+    {
+      "epoch": 0.1552954292084727,
+      "grad_norm": 0.5495896935462952,
+      "learning_rate": 8e-05,
+      "loss": 1.9165,
+      "step": 2786
+    },
+    {
+      "epoch": 0.15535117056856187,
+      "grad_norm": 0.48266035318374634,
+      "learning_rate": 8e-05,
+      "loss": 1.8742,
+      "step": 2787
+    },
+    {
+      "epoch": 0.15540691192865105,
+      "grad_norm": 0.47781747579574585,
+      "learning_rate": 8e-05,
+      "loss": 1.8515,
+      "step": 2788
+    },
+    {
+      "epoch": 0.15546265328874023,
+      "grad_norm": 0.4679318368434906,
+      "learning_rate": 8e-05,
+      "loss": 1.7413,
+      "step": 2789
+    },
+    {
+      "epoch": 0.15551839464882944,
+      "grad_norm": 0.4929085969924927,
+      "learning_rate": 8e-05,
+      "loss": 1.8921,
+      "step": 2790
+    },
+    {
+      "epoch": 0.15557413600891862,
+      "grad_norm": 0.4672272205352783,
+      "learning_rate": 8e-05,
+      "loss": 1.6709,
+      "step": 2791
+    },
+    {
+      "epoch": 0.1556298773690078,
+      "grad_norm": 0.4650382399559021,
+      "learning_rate": 8e-05,
+      "loss": 1.504,
+      "step": 2792
+    },
+    {
+      "epoch": 0.155685618729097,
+      "grad_norm": 0.43125632405281067,
+      "learning_rate": 8e-05,
+      "loss": 1.4325,
+      "step": 2793
+    },
+    {
+      "epoch": 0.15574136008918618,
+      "grad_norm": 0.446275919675827,
+      "learning_rate": 8e-05,
+      "loss": 1.6414,
+      "step": 2794
+    },
+    {
+      "epoch": 0.15579710144927536,
+      "grad_norm": 0.449980229139328,
+      "learning_rate": 8e-05,
+      "loss": 1.6934,
+      "step": 2795
+    },
+    {
+      "epoch": 0.15585284280936454,
+      "grad_norm": 0.4828239977359772,
+      "learning_rate": 8e-05,
+      "loss": 1.7457,
+      "step": 2796
+    },
+    {
+      "epoch": 0.15590858416945375,
+      "grad_norm": 0.5396363735198975,
+      "learning_rate": 8e-05,
+      "loss": 1.8686,
+      "step": 2797
+    },
+    {
+      "epoch": 0.15596432552954292,
+      "grad_norm": 0.4906916320323944,
+      "learning_rate": 8e-05,
+      "loss": 1.7499,
+      "step": 2798
+    },
+    {
+      "epoch": 0.1560200668896321,
+      "grad_norm": 0.502632200717926,
+      "learning_rate": 8e-05,
+      "loss": 1.9716,
+      "step": 2799
+    },
+    {
+      "epoch": 0.15607580824972128,
+      "grad_norm": 0.5308877229690552,
+      "learning_rate": 8e-05,
+      "loss": 1.694,
+      "step": 2800
+    },
+    {
+      "epoch": 0.1561315496098105,
+      "grad_norm": 0.45435258746147156,
+      "learning_rate": 8e-05,
+      "loss": 1.573,
+      "step": 2801
+    },
+    {
+      "epoch": 0.15618729096989967,
+      "grad_norm": 0.4956980049610138,
+      "learning_rate": 8e-05,
+      "loss": 1.8578,
+      "step": 2802
+    },
+    {
+      "epoch": 0.15624303232998885,
+      "grad_norm": 0.5421011447906494,
+      "learning_rate": 8e-05,
+      "loss": 1.6132,
+      "step": 2803
+    },
+    {
+      "epoch": 0.15629877369007802,
+      "grad_norm": 0.4681341350078583,
+      "learning_rate": 8e-05,
+      "loss": 1.6352,
+      "step": 2804
+    },
+    {
+      "epoch": 0.15635451505016723,
+      "grad_norm": 0.44525671005249023,
+      "learning_rate": 8e-05,
+      "loss": 1.6762,
+      "step": 2805
+    },
+    {
+      "epoch": 0.1564102564102564,
+      "grad_norm": 0.4492572247982025,
+      "learning_rate": 8e-05,
+      "loss": 1.8298,
+      "step": 2806
+    },
+    {
+      "epoch": 0.1564659977703456,
+      "grad_norm": 0.45993149280548096,
+      "learning_rate": 8e-05,
+      "loss": 1.6447,
+      "step": 2807
+    },
+    {
+      "epoch": 0.1565217391304348,
+      "grad_norm": 0.4706052541732788,
+      "learning_rate": 8e-05,
+      "loss": 1.5694,
+      "step": 2808
+    },
+    {
+      "epoch": 0.15657748049052397,
+      "grad_norm": 0.4705091118812561,
+      "learning_rate": 8e-05,
+      "loss": 1.6959,
+      "step": 2809
+    },
+    {
+      "epoch": 0.15663322185061315,
+      "grad_norm": 0.47114166617393494,
+      "learning_rate": 8e-05,
+      "loss": 1.6371,
+      "step": 2810
+    },
+    {
+      "epoch": 0.15668896321070233,
+      "grad_norm": 0.5016292333602905,
+      "learning_rate": 8e-05,
+      "loss": 1.7122,
+      "step": 2811
+    },
+    {
+      "epoch": 0.15674470457079154,
+      "grad_norm": 0.48519521951675415,
+      "learning_rate": 8e-05,
+      "loss": 1.7304,
+      "step": 2812
+    },
+    {
+      "epoch": 0.15680044593088072,
+      "grad_norm": 0.48272377252578735,
+      "learning_rate": 8e-05,
+      "loss": 1.7092,
+      "step": 2813
+    },
+    {
+      "epoch": 0.1568561872909699,
+      "grad_norm": 0.505794882774353,
+      "learning_rate": 8e-05,
+      "loss": 1.701,
+      "step": 2814
+    },
+    {
+      "epoch": 0.15691192865105907,
+      "grad_norm": 0.44441741704940796,
+      "learning_rate": 8e-05,
+      "loss": 1.702,
+      "step": 2815
+    },
+    {
+      "epoch": 0.15696767001114828,
+      "grad_norm": 0.47186487913131714,
+      "learning_rate": 8e-05,
+      "loss": 1.7099,
+      "step": 2816
+    },
+    {
+      "epoch": 0.15702341137123746,
+      "grad_norm": 0.5894055366516113,
+      "learning_rate": 8e-05,
+      "loss": 1.6043,
+      "step": 2817
+    },
+    {
+      "epoch": 0.15707915273132664,
+      "grad_norm": 0.5004922151565552,
+      "learning_rate": 8e-05,
+      "loss": 1.7105,
+      "step": 2818
+    },
+    {
+      "epoch": 0.15713489409141584,
+      "grad_norm": 0.46560198068618774,
+      "learning_rate": 8e-05,
+      "loss": 1.7297,
+      "step": 2819
+    },
+    {
+      "epoch": 0.15719063545150502,
+      "grad_norm": 0.4759940505027771,
+      "learning_rate": 8e-05,
+      "loss": 1.7775,
+      "step": 2820
+    },
+    {
+      "epoch": 0.1572463768115942,
+      "grad_norm": 0.5159291625022888,
+      "learning_rate": 8e-05,
+      "loss": 1.9294,
+      "step": 2821
+    },
+    {
+      "epoch": 0.15730211817168338,
+      "grad_norm": 0.5078474879264832,
+      "learning_rate": 8e-05,
+      "loss": 1.9248,
+      "step": 2822
+    },
+    {
+      "epoch": 0.1573578595317726,
+      "grad_norm": 0.4777558445930481,
+      "learning_rate": 8e-05,
+      "loss": 1.7751,
+      "step": 2823
+    },
+    {
+      "epoch": 0.15741360089186177,
+      "grad_norm": 0.4959481656551361,
+      "learning_rate": 8e-05,
+      "loss": 1.6357,
+      "step": 2824
+    },
+    {
+      "epoch": 0.15746934225195094,
+      "grad_norm": 0.4931829273700714,
+      "learning_rate": 8e-05,
+      "loss": 1.5699,
+      "step": 2825
+    },
+    {
+      "epoch": 0.15752508361204012,
+      "grad_norm": 0.4662152826786041,
+      "learning_rate": 8e-05,
+      "loss": 1.6673,
+      "step": 2826
+    },
+    {
+      "epoch": 0.15758082497212933,
+      "grad_norm": 0.4971119165420532,
+      "learning_rate": 8e-05,
+      "loss": 1.6575,
+      "step": 2827
+    },
+    {
+      "epoch": 0.1576365663322185,
+      "grad_norm": 0.478135347366333,
+      "learning_rate": 8e-05,
+      "loss": 1.7929,
+      "step": 2828
+    },
+    {
+      "epoch": 0.1576923076923077,
+      "grad_norm": 0.5057351589202881,
+      "learning_rate": 8e-05,
+      "loss": 1.8841,
+      "step": 2829
+    },
+    {
+      "epoch": 0.15774804905239687,
+      "grad_norm": 0.5114669799804688,
+      "learning_rate": 8e-05,
+      "loss": 1.6789,
+      "step": 2830
+    },
+    {
+      "epoch": 0.15780379041248607,
+      "grad_norm": 0.4946887791156769,
+      "learning_rate": 8e-05,
+      "loss": 1.9107,
+      "step": 2831
+    },
+    {
+      "epoch": 0.15785953177257525,
+      "grad_norm": 0.4401973485946655,
+      "learning_rate": 8e-05,
+      "loss": 1.4577,
+      "step": 2832
+    },
+    {
+      "epoch": 0.15791527313266443,
+      "grad_norm": 0.46898627281188965,
+      "learning_rate": 8e-05,
+      "loss": 1.7578,
+      "step": 2833
+    },
+    {
+      "epoch": 0.15797101449275364,
+      "grad_norm": 0.5002691745758057,
+      "learning_rate": 8e-05,
+      "loss": 1.5825,
+      "step": 2834
+    },
+    {
+      "epoch": 0.15802675585284282,
+      "grad_norm": 0.5099799633026123,
+      "learning_rate": 8e-05,
+      "loss": 1.7851,
+      "step": 2835
+    },
+    {
+      "epoch": 0.158082497212932,
+      "grad_norm": 0.49243247509002686,
+      "learning_rate": 8e-05,
+      "loss": 1.8366,
+      "step": 2836
+    },
+    {
+      "epoch": 0.15813823857302117,
+      "grad_norm": 0.547187864780426,
+      "learning_rate": 8e-05,
+      "loss": 1.792,
+      "step": 2837
+    },
+    {
+      "epoch": 0.15819397993311038,
+      "grad_norm": 0.4610172212123871,
+      "learning_rate": 8e-05,
+      "loss": 1.6167,
+      "step": 2838
+    },
+    {
+      "epoch": 0.15824972129319956,
+      "grad_norm": 0.5065680742263794,
+      "learning_rate": 8e-05,
+      "loss": 2.005,
+      "step": 2839
+    },
+    {
+      "epoch": 0.15830546265328874,
+      "grad_norm": 0.5296475887298584,
+      "learning_rate": 8e-05,
+      "loss": 1.9063,
+      "step": 2840
+    },
+    {
+      "epoch": 0.15836120401337792,
+      "grad_norm": 0.47584494948387146,
+      "learning_rate": 8e-05,
+      "loss": 1.6319,
+      "step": 2841
+    },
+    {
+      "epoch": 0.15841694537346712,
+      "grad_norm": 0.482562780380249,
+      "learning_rate": 8e-05,
+      "loss": 1.6442,
+      "step": 2842
+    },
+    {
+      "epoch": 0.1584726867335563,
+      "grad_norm": 0.5090284943580627,
+      "learning_rate": 8e-05,
+      "loss": 1.7997,
+      "step": 2843
+    },
+    {
+      "epoch": 0.15852842809364548,
+      "grad_norm": 0.4627036452293396,
+      "learning_rate": 8e-05,
+      "loss": 1.7175,
+      "step": 2844
+    },
+    {
+      "epoch": 0.15858416945373466,
+      "grad_norm": 0.48607441782951355,
+      "learning_rate": 8e-05,
+      "loss": 1.7816,
+      "step": 2845
+    },
+    {
+      "epoch": 0.15863991081382386,
+      "grad_norm": 0.4424927532672882,
+      "learning_rate": 8e-05,
+      "loss": 1.4956,
+      "step": 2846
+    },
+    {
+      "epoch": 0.15869565217391304,
+      "grad_norm": 0.5370016098022461,
+      "learning_rate": 8e-05,
+      "loss": 1.8462,
+      "step": 2847
+    },
+    {
+      "epoch": 0.15875139353400222,
+      "grad_norm": 0.5331100225448608,
+      "learning_rate": 8e-05,
+      "loss": 1.8603,
+      "step": 2848
+    },
+    {
+      "epoch": 0.15880713489409143,
+      "grad_norm": 0.4674780070781708,
+      "learning_rate": 8e-05,
+      "loss": 1.8825,
+      "step": 2849
+    },
+    {
+      "epoch": 0.1588628762541806,
+      "grad_norm": 0.4758144021034241,
+      "learning_rate": 8e-05,
+      "loss": 1.7723,
+      "step": 2850
+    },
+    {
+      "epoch": 0.15891861761426979,
+      "grad_norm": 0.4641381502151489,
+      "learning_rate": 8e-05,
+      "loss": 1.8323,
+      "step": 2851
+    },
+    {
+      "epoch": 0.15897435897435896,
+      "grad_norm": 0.4850490689277649,
+      "learning_rate": 8e-05,
+      "loss": 1.5911,
+      "step": 2852
+    },
+    {
+      "epoch": 0.15903010033444817,
+      "grad_norm": 0.48367488384246826,
+      "learning_rate": 8e-05,
+      "loss": 1.8259,
+      "step": 2853
+    },
+    {
+      "epoch": 0.15908584169453735,
+      "grad_norm": 0.49388548731803894,
+      "learning_rate": 8e-05,
+      "loss": 1.6125,
+      "step": 2854
+    },
+    {
+      "epoch": 0.15914158305462653,
+      "grad_norm": 0.6949816346168518,
+      "learning_rate": 8e-05,
+      "loss": 1.8107,
+      "step": 2855
+    },
+    {
+      "epoch": 0.1591973244147157,
+      "grad_norm": 0.4608065187931061,
+      "learning_rate": 8e-05,
+      "loss": 1.6638,
+      "step": 2856
+    },
+    {
+      "epoch": 0.15925306577480491,
+      "grad_norm": 0.5375053882598877,
+      "learning_rate": 8e-05,
+      "loss": 1.8655,
+      "step": 2857
+    },
+    {
+      "epoch": 0.1593088071348941,
+      "grad_norm": 0.4775127172470093,
+      "learning_rate": 8e-05,
+      "loss": 1.6442,
+      "step": 2858
+    },
+    {
+      "epoch": 0.15936454849498327,
+      "grad_norm": 0.5275262594223022,
+      "learning_rate": 8e-05,
+      "loss": 1.823,
+      "step": 2859
+    },
+    {
+      "epoch": 0.15942028985507245,
+      "grad_norm": 0.4744328558444977,
+      "learning_rate": 8e-05,
+      "loss": 1.7652,
+      "step": 2860
+    },
+    {
+      "epoch": 0.15947603121516166,
+      "grad_norm": 0.4801606237888336,
+      "learning_rate": 8e-05,
+      "loss": 1.8581,
+      "step": 2861
+    },
+    {
+      "epoch": 0.15953177257525084,
+      "grad_norm": 0.49865031242370605,
+      "learning_rate": 8e-05,
+      "loss": 1.5966,
+      "step": 2862
+    },
+    {
+      "epoch": 0.15958751393534001,
+      "grad_norm": 0.5179030895233154,
+      "learning_rate": 8e-05,
+      "loss": 1.917,
+      "step": 2863
+    },
+    {
+      "epoch": 0.15964325529542922,
+      "grad_norm": 0.45887720584869385,
+      "learning_rate": 8e-05,
+      "loss": 1.7588,
+      "step": 2864
+    },
+    {
+      "epoch": 0.1596989966555184,
+      "grad_norm": 0.4781455099582672,
+      "learning_rate": 8e-05,
+      "loss": 1.5995,
+      "step": 2865
+    },
+    {
+      "epoch": 0.15975473801560758,
+      "grad_norm": 0.5103084444999695,
+      "learning_rate": 8e-05,
+      "loss": 1.7247,
+      "step": 2866
+    },
+    {
+      "epoch": 0.15981047937569676,
+      "grad_norm": 0.4783124029636383,
+      "learning_rate": 8e-05,
+      "loss": 1.8298,
+      "step": 2867
+    },
+    {
+      "epoch": 0.15986622073578596,
+      "grad_norm": 0.46342748403549194,
+      "learning_rate": 8e-05,
+      "loss": 1.7312,
+      "step": 2868
+    },
+    {
+      "epoch": 0.15992196209587514,
+      "grad_norm": 0.496513694524765,
+      "learning_rate": 8e-05,
+      "loss": 1.4232,
+      "step": 2869
+    },
+    {
+      "epoch": 0.15997770345596432,
+      "grad_norm": 0.4710693955421448,
+      "learning_rate": 8e-05,
+      "loss": 1.6533,
+      "step": 2870
+    },
+    {
+      "epoch": 0.1600334448160535,
+      "grad_norm": 0.4609511196613312,
+      "learning_rate": 8e-05,
+      "loss": 1.5186,
+      "step": 2871
+    },
+    {
+      "epoch": 0.1600891861761427,
+      "grad_norm": 0.5193403959274292,
+      "learning_rate": 8e-05,
+      "loss": 1.7667,
+      "step": 2872
+    },
+    {
+      "epoch": 0.16014492753623188,
+      "grad_norm": 0.4479955732822418,
+      "learning_rate": 8e-05,
+      "loss": 1.6617,
+      "step": 2873
+    },
+    {
+      "epoch": 0.16020066889632106,
+      "grad_norm": 0.47863709926605225,
+      "learning_rate": 8e-05,
+      "loss": 1.6544,
+      "step": 2874
+    },
+    {
+      "epoch": 0.16025641025641027,
+      "grad_norm": 0.5270398259162903,
+      "learning_rate": 8e-05,
+      "loss": 1.9415,
+      "step": 2875
+    },
+    {
+      "epoch": 0.16031215161649945,
+      "grad_norm": 0.46953439712524414,
+      "learning_rate": 8e-05,
+      "loss": 1.8319,
+      "step": 2876
+    },
+    {
+      "epoch": 0.16036789297658863,
+      "grad_norm": 0.46271249651908875,
+      "learning_rate": 8e-05,
+      "loss": 1.6766,
+      "step": 2877
+    },
+    {
+      "epoch": 0.1604236343366778,
+      "grad_norm": 0.4576559066772461,
+      "learning_rate": 8e-05,
+      "loss": 1.857,
+      "step": 2878
+    },
+    {
+      "epoch": 0.160479375696767,
+      "grad_norm": 0.4797700047492981,
+      "learning_rate": 8e-05,
+      "loss": 1.6108,
+      "step": 2879
+    },
+    {
+      "epoch": 0.1605351170568562,
+      "grad_norm": 0.47959211468696594,
+      "learning_rate": 8e-05,
+      "loss": 1.9271,
+      "step": 2880
+    },
+    {
+      "epoch": 0.16059085841694537,
+      "grad_norm": 0.4791256785392761,
+      "learning_rate": 8e-05,
+      "loss": 1.9178,
+      "step": 2881
+    },
+    {
+      "epoch": 0.16064659977703455,
+      "grad_norm": 0.47225552797317505,
+      "learning_rate": 8e-05,
+      "loss": 1.598,
+      "step": 2882
+    },
+    {
+      "epoch": 0.16070234113712376,
+      "grad_norm": 0.4455716609954834,
+      "learning_rate": 8e-05,
+      "loss": 1.5567,
+      "step": 2883
+    },
+    {
+      "epoch": 0.16075808249721293,
+      "grad_norm": 0.48891139030456543,
+      "learning_rate": 8e-05,
+      "loss": 1.9012,
+      "step": 2884
+    },
+    {
+      "epoch": 0.1608138238573021,
+      "grad_norm": 0.5016549825668335,
+      "learning_rate": 8e-05,
+      "loss": 1.816,
+      "step": 2885
+    },
+    {
+      "epoch": 0.1608695652173913,
+      "grad_norm": 0.4449464976787567,
+      "learning_rate": 8e-05,
+      "loss": 1.5622,
+      "step": 2886
+    },
+    {
+      "epoch": 0.1609253065774805,
+      "grad_norm": 0.5120397210121155,
+      "learning_rate": 8e-05,
+      "loss": 1.5982,
+      "step": 2887
+    },
+    {
+      "epoch": 0.16098104793756968,
+      "grad_norm": 0.5218576788902283,
+      "learning_rate": 8e-05,
+      "loss": 1.7747,
+      "step": 2888
+    },
+    {
+      "epoch": 0.16103678929765886,
+      "grad_norm": 0.5918441414833069,
+      "learning_rate": 8e-05,
+      "loss": 2.2052,
+      "step": 2889
+    },
+    {
+      "epoch": 0.16109253065774806,
+      "grad_norm": 0.4566386044025421,
+      "learning_rate": 8e-05,
+      "loss": 1.701,
+      "step": 2890
+    },
+    {
+      "epoch": 0.16114827201783724,
+      "grad_norm": 0.45735636353492737,
+      "learning_rate": 8e-05,
+      "loss": 1.7771,
+      "step": 2891
+    },
+    {
+      "epoch": 0.16120401337792642,
+      "grad_norm": 0.45574647188186646,
+      "learning_rate": 8e-05,
+      "loss": 1.5037,
+      "step": 2892
+    },
+    {
+      "epoch": 0.1612597547380156,
+      "grad_norm": 0.4675073027610779,
+      "learning_rate": 8e-05,
+      "loss": 1.5138,
+      "step": 2893
+    },
+    {
+      "epoch": 0.1613154960981048,
+      "grad_norm": 0.4805678427219391,
+      "learning_rate": 8e-05,
+      "loss": 1.7286,
+      "step": 2894
+    },
+    {
+      "epoch": 0.16137123745819398,
+      "grad_norm": 0.46076643466949463,
+      "learning_rate": 8e-05,
+      "loss": 1.7129,
+      "step": 2895
+    },
+    {
+      "epoch": 0.16142697881828316,
+      "grad_norm": 0.4653468132019043,
+      "learning_rate": 8e-05,
+      "loss": 1.7953,
+      "step": 2896
+    },
+    {
+      "epoch": 0.16148272017837234,
+      "grad_norm": 0.6508491635322571,
+      "learning_rate": 8e-05,
+      "loss": 1.808,
+      "step": 2897
+    },
+    {
+      "epoch": 0.16153846153846155,
+      "grad_norm": 0.4713476598262787,
+      "learning_rate": 8e-05,
+      "loss": 1.6684,
+      "step": 2898
+    },
+    {
+      "epoch": 0.16159420289855073,
+      "grad_norm": 0.511515200138092,
+      "learning_rate": 8e-05,
+      "loss": 1.8001,
+      "step": 2899
+    },
+    {
+      "epoch": 0.1616499442586399,
+      "grad_norm": 0.4759061932563782,
+      "learning_rate": 8e-05,
+      "loss": 1.6275,
+      "step": 2900
+    },
+    {
+      "epoch": 0.16170568561872908,
+      "grad_norm": 0.47947847843170166,
+      "learning_rate": 8e-05,
+      "loss": 1.9198,
+      "step": 2901
+    },
+    {
+      "epoch": 0.1617614269788183,
+      "grad_norm": 0.4760971665382385,
+      "learning_rate": 8e-05,
+      "loss": 1.7911,
+      "step": 2902
+    },
+    {
+      "epoch": 0.16181716833890747,
+      "grad_norm": 0.48154783248901367,
+      "learning_rate": 8e-05,
+      "loss": 1.6896,
+      "step": 2903
+    },
+    {
+      "epoch": 0.16187290969899665,
+      "grad_norm": 0.4559423625469208,
+      "learning_rate": 8e-05,
+      "loss": 1.6635,
+      "step": 2904
+    },
+    {
+      "epoch": 0.16192865105908585,
+      "grad_norm": 0.47385692596435547,
+      "learning_rate": 8e-05,
+      "loss": 1.7711,
+      "step": 2905
+    },
+    {
+      "epoch": 0.16198439241917503,
+      "grad_norm": 0.4514434337615967,
+      "learning_rate": 8e-05,
+      "loss": 1.7176,
+      "step": 2906
+    },
+    {
+      "epoch": 0.1620401337792642,
+      "grad_norm": 0.441662460565567,
+      "learning_rate": 8e-05,
+      "loss": 1.655,
+      "step": 2907
+    },
+    {
+      "epoch": 0.1620958751393534,
+      "grad_norm": 0.4862743020057678,
+      "learning_rate": 8e-05,
+      "loss": 1.9833,
+      "step": 2908
+    },
+    {
+      "epoch": 0.1621516164994426,
+      "grad_norm": 0.4469553232192993,
+      "learning_rate": 8e-05,
+      "loss": 1.4799,
+      "step": 2909
+    },
+    {
+      "epoch": 0.16220735785953178,
+      "grad_norm": 0.4639607071876526,
+      "learning_rate": 8e-05,
+      "loss": 1.7082,
+      "step": 2910
+    },
+    {
+      "epoch": 0.16226309921962095,
+      "grad_norm": 0.5053742527961731,
+      "learning_rate": 8e-05,
+      "loss": 1.8838,
+      "step": 2911
+    },
+    {
+      "epoch": 0.16231884057971013,
+      "grad_norm": 0.4670344293117523,
+      "learning_rate": 8e-05,
+      "loss": 1.812,
+      "step": 2912
+    },
+    {
+      "epoch": 0.16237458193979934,
+      "grad_norm": 0.47207584977149963,
+      "learning_rate": 8e-05,
+      "loss": 1.6914,
+      "step": 2913
+    },
+    {
+      "epoch": 0.16243032329988852,
+      "grad_norm": 0.4806153476238251,
+      "learning_rate": 8e-05,
+      "loss": 1.7662,
+      "step": 2914
+    },
+    {
+      "epoch": 0.1624860646599777,
+      "grad_norm": 0.4803650379180908,
+      "learning_rate": 8e-05,
+      "loss": 1.7332,
+      "step": 2915
+    },
+    {
+      "epoch": 0.1625418060200669,
+      "grad_norm": 0.40704938769340515,
+      "learning_rate": 8e-05,
+      "loss": 1.3805,
+      "step": 2916
+    },
+    {
+      "epoch": 0.16259754738015608,
+      "grad_norm": 0.48974987864494324,
+      "learning_rate": 8e-05,
+      "loss": 1.7292,
+      "step": 2917
+    },
+    {
+      "epoch": 0.16265328874024526,
+      "grad_norm": 0.45540478825569153,
+      "learning_rate": 8e-05,
+      "loss": 1.6662,
+      "step": 2918
+    },
+    {
+      "epoch": 0.16270903010033444,
+      "grad_norm": 0.4781588912010193,
+      "learning_rate": 8e-05,
+      "loss": 1.6925,
+      "step": 2919
+    },
+    {
+      "epoch": 0.16276477146042365,
+      "grad_norm": 0.5591298341751099,
+      "learning_rate": 8e-05,
+      "loss": 1.3078,
+      "step": 2920
+    },
+    {
+      "epoch": 0.16282051282051282,
+      "grad_norm": 0.49242156744003296,
+      "learning_rate": 8e-05,
+      "loss": 1.6533,
+      "step": 2921
+    },
+    {
+      "epoch": 0.162876254180602,
+      "grad_norm": 0.5097176432609558,
+      "learning_rate": 8e-05,
+      "loss": 1.8765,
+      "step": 2922
+    },
+    {
+      "epoch": 0.16293199554069118,
+      "grad_norm": 0.5155987739562988,
+      "learning_rate": 8e-05,
+      "loss": 1.743,
+      "step": 2923
+    },
+    {
+      "epoch": 0.1629877369007804,
+      "grad_norm": 0.5098856687545776,
+      "learning_rate": 8e-05,
+      "loss": 1.7513,
+      "step": 2924
+    },
+    {
+      "epoch": 0.16304347826086957,
+      "grad_norm": 0.5157544612884521,
+      "learning_rate": 8e-05,
+      "loss": 1.8728,
+      "step": 2925
+    },
+    {
+      "epoch": 0.16309921962095875,
+      "grad_norm": 0.4856857359409332,
+      "learning_rate": 8e-05,
+      "loss": 1.4589,
+      "step": 2926
+    },
+    {
+      "epoch": 0.16315496098104793,
+      "grad_norm": 0.4530739188194275,
+      "learning_rate": 8e-05,
+      "loss": 1.7018,
+      "step": 2927
+    },
+    {
+      "epoch": 0.16321070234113713,
+      "grad_norm": 0.4573028087615967,
+      "learning_rate": 8e-05,
+      "loss": 1.6762,
+      "step": 2928
+    },
+    {
+      "epoch": 0.1632664437012263,
+      "grad_norm": 0.4893021583557129,
+      "learning_rate": 8e-05,
+      "loss": 1.8043,
+      "step": 2929
+    },
+    {
+      "epoch": 0.1633221850613155,
+      "grad_norm": 0.557016909122467,
+      "learning_rate": 8e-05,
+      "loss": 2.0862,
+      "step": 2930
+    },
+    {
+      "epoch": 0.1633779264214047,
+      "grad_norm": 0.5262312889099121,
+      "learning_rate": 8e-05,
+      "loss": 1.9518,
+      "step": 2931
+    },
+    {
+      "epoch": 0.16343366778149387,
+      "grad_norm": 0.422815203666687,
+      "learning_rate": 8e-05,
+      "loss": 1.5857,
+      "step": 2932
+    },
+    {
+      "epoch": 0.16348940914158305,
+      "grad_norm": 0.4872015416622162,
+      "learning_rate": 8e-05,
+      "loss": 1.8183,
+      "step": 2933
+    },
+    {
+      "epoch": 0.16354515050167223,
+      "grad_norm": 0.45688596367836,
+      "learning_rate": 8e-05,
+      "loss": 1.6444,
+      "step": 2934
+    },
+    {
+      "epoch": 0.16360089186176144,
+      "grad_norm": 0.4974451959133148,
+      "learning_rate": 8e-05,
+      "loss": 1.8459,
+      "step": 2935
+    },
+    {
+      "epoch": 0.16365663322185062,
+      "grad_norm": 0.49110689759254456,
+      "learning_rate": 8e-05,
+      "loss": 1.9244,
+      "step": 2936
+    },
+    {
+      "epoch": 0.1637123745819398,
+      "grad_norm": 0.49117761850357056,
+      "learning_rate": 8e-05,
+      "loss": 1.6758,
+      "step": 2937
+    },
+    {
+      "epoch": 0.16376811594202897,
+      "grad_norm": 0.4526088833808899,
+      "learning_rate": 8e-05,
+      "loss": 1.7698,
+      "step": 2938
+    },
+    {
+      "epoch": 0.16382385730211818,
+      "grad_norm": 0.49112263321876526,
+      "learning_rate": 8e-05,
+      "loss": 1.9145,
+      "step": 2939
+    },
+    {
+      "epoch": 0.16387959866220736,
+      "grad_norm": 0.5060984492301941,
+      "learning_rate": 8e-05,
+      "loss": 1.734,
+      "step": 2940
+    },
+    {
+      "epoch": 0.16393534002229654,
+      "grad_norm": 0.45045602321624756,
+      "learning_rate": 8e-05,
+      "loss": 1.6977,
+      "step": 2941
+    },
+    {
+      "epoch": 0.16399108138238572,
+      "grad_norm": 0.4900680482387543,
+      "learning_rate": 8e-05,
+      "loss": 1.7354,
+      "step": 2942
+    },
+    {
+      "epoch": 0.16404682274247492,
+      "grad_norm": 0.4850306808948517,
+      "learning_rate": 8e-05,
+      "loss": 1.8396,
+      "step": 2943
+    },
+    {
+      "epoch": 0.1641025641025641,
+      "grad_norm": 0.4529063403606415,
+      "learning_rate": 8e-05,
+      "loss": 1.744,
+      "step": 2944
+    },
+    {
+      "epoch": 0.16415830546265328,
+      "grad_norm": 0.4494536817073822,
+      "learning_rate": 8e-05,
+      "loss": 1.8157,
+      "step": 2945
+    },
+    {
+      "epoch": 0.1642140468227425,
+      "grad_norm": 0.4689137935638428,
+      "learning_rate": 8e-05,
+      "loss": 1.6594,
+      "step": 2946
+    },
+    {
+      "epoch": 0.16426978818283167,
+      "grad_norm": 0.480871319770813,
+      "learning_rate": 8e-05,
+      "loss": 1.8693,
+      "step": 2947
+    },
+    {
+      "epoch": 0.16432552954292085,
+      "grad_norm": 0.49745580554008484,
+      "learning_rate": 8e-05,
+      "loss": 1.8757,
+      "step": 2948
+    },
+    {
+      "epoch": 0.16438127090301002,
+      "grad_norm": 0.49291250109672546,
+      "learning_rate": 8e-05,
+      "loss": 1.7641,
+      "step": 2949
+    },
+    {
+      "epoch": 0.16443701226309923,
+      "grad_norm": 0.47803646326065063,
+      "learning_rate": 8e-05,
+      "loss": 1.7651,
+      "step": 2950
+    },
+    {
+      "epoch": 0.1644927536231884,
+      "grad_norm": 0.44731417298316956,
+      "learning_rate": 8e-05,
+      "loss": 1.6245,
+      "step": 2951
+    },
+    {
+      "epoch": 0.1645484949832776,
+      "grad_norm": 0.46639564633369446,
+      "learning_rate": 8e-05,
+      "loss": 1.6166,
+      "step": 2952
+    },
+    {
+      "epoch": 0.16460423634336677,
+      "grad_norm": 0.460929274559021,
+      "learning_rate": 8e-05,
+      "loss": 1.9163,
+      "step": 2953
+    },
+    {
+      "epoch": 0.16465997770345597,
+      "grad_norm": 0.44395554065704346,
+      "learning_rate": 8e-05,
+      "loss": 1.6165,
+      "step": 2954
+    },
+    {
+      "epoch": 0.16471571906354515,
+      "grad_norm": 0.5062477588653564,
+      "learning_rate": 8e-05,
+      "loss": 1.8948,
+      "step": 2955
+    },
+    {
+      "epoch": 0.16477146042363433,
+      "grad_norm": 0.5431268215179443,
+      "learning_rate": 8e-05,
+      "loss": 1.5934,
+      "step": 2956
+    },
+    {
+      "epoch": 0.1648272017837235,
+      "grad_norm": 0.4659595489501953,
+      "learning_rate": 8e-05,
+      "loss": 1.7985,
+      "step": 2957
+    },
+    {
+      "epoch": 0.16488294314381272,
+      "grad_norm": 0.5105186104774475,
+      "learning_rate": 8e-05,
+      "loss": 1.7925,
+      "step": 2958
+    },
+    {
+      "epoch": 0.1649386845039019,
+      "grad_norm": 0.5099925994873047,
+      "learning_rate": 8e-05,
+      "loss": 1.8497,
+      "step": 2959
+    },
+    {
+      "epoch": 0.16499442586399107,
+      "grad_norm": 0.5125576853752136,
+      "learning_rate": 8e-05,
+      "loss": 2.023,
+      "step": 2960
+    },
+    {
+      "epoch": 0.16505016722408028,
+      "grad_norm": 0.47445541620254517,
+      "learning_rate": 8e-05,
+      "loss": 1.7916,
+      "step": 2961
+    },
+    {
+      "epoch": 0.16510590858416946,
+      "grad_norm": 0.5191357135772705,
+      "learning_rate": 8e-05,
+      "loss": 1.8498,
+      "step": 2962
+    },
+    {
+      "epoch": 0.16516164994425864,
+      "grad_norm": 0.48971402645111084,
+      "learning_rate": 8e-05,
+      "loss": 1.6435,
+      "step": 2963
+    },
+    {
+      "epoch": 0.16521739130434782,
+      "grad_norm": 0.5164715051651001,
+      "learning_rate": 8e-05,
+      "loss": 1.7424,
+      "step": 2964
+    },
+    {
+      "epoch": 0.16527313266443702,
+      "grad_norm": 0.5232493281364441,
+      "learning_rate": 8e-05,
+      "loss": 2.0409,
+      "step": 2965
+    },
+    {
+      "epoch": 0.1653288740245262,
+      "grad_norm": 0.46612119674682617,
+      "learning_rate": 8e-05,
+      "loss": 1.7034,
+      "step": 2966
+    },
+    {
+      "epoch": 0.16538461538461538,
+      "grad_norm": 0.45249107480049133,
+      "learning_rate": 8e-05,
+      "loss": 1.8216,
+      "step": 2967
+    },
+    {
+      "epoch": 0.16544035674470456,
+      "grad_norm": 0.5156910419464111,
+      "learning_rate": 8e-05,
+      "loss": 1.7517,
+      "step": 2968
+    },
+    {
+      "epoch": 0.16549609810479377,
+      "grad_norm": 0.4460727274417877,
+      "learning_rate": 8e-05,
+      "loss": 1.6204,
+      "step": 2969
+    },
+    {
+      "epoch": 0.16555183946488294,
+      "grad_norm": 0.47430551052093506,
+      "learning_rate": 8e-05,
+      "loss": 1.6935,
+      "step": 2970
+    },
+    {
+      "epoch": 0.16560758082497212,
+      "grad_norm": 0.5011522173881531,
+      "learning_rate": 8e-05,
+      "loss": 2.0306,
+      "step": 2971
+    },
+    {
+      "epoch": 0.16566332218506133,
+      "grad_norm": 0.47904592752456665,
+      "learning_rate": 8e-05,
+      "loss": 1.7152,
+      "step": 2972
+    },
+    {
+      "epoch": 0.1657190635451505,
+      "grad_norm": 0.5341163277626038,
+      "learning_rate": 8e-05,
+      "loss": 1.8604,
+      "step": 2973
+    },
+    {
+      "epoch": 0.1657748049052397,
+      "grad_norm": 0.46354594826698303,
+      "learning_rate": 8e-05,
+      "loss": 1.7609,
+      "step": 2974
+    },
+    {
+      "epoch": 0.16583054626532887,
+      "grad_norm": 0.4791519045829773,
+      "learning_rate": 8e-05,
+      "loss": 1.6751,
+      "step": 2975
+    },
+    {
+      "epoch": 0.16588628762541807,
+      "grad_norm": 0.470883309841156,
+      "learning_rate": 8e-05,
+      "loss": 1.7161,
+      "step": 2976
+    },
+    {
+      "epoch": 0.16594202898550725,
+      "grad_norm": 0.5448764562606812,
+      "learning_rate": 8e-05,
+      "loss": 1.8716,
+      "step": 2977
+    },
+    {
+      "epoch": 0.16599777034559643,
+      "grad_norm": 0.44938236474990845,
+      "learning_rate": 8e-05,
+      "loss": 1.5895,
+      "step": 2978
+    },
+    {
+      "epoch": 0.1660535117056856,
+      "grad_norm": 0.5080894231796265,
+      "learning_rate": 8e-05,
+      "loss": 1.9543,
+      "step": 2979
+    },
+    {
+      "epoch": 0.16610925306577481,
+      "grad_norm": 0.5100005865097046,
+      "learning_rate": 8e-05,
+      "loss": 2.0521,
+      "step": 2980
+    },
+    {
+      "epoch": 0.166164994425864,
+      "grad_norm": 0.49559685587882996,
+      "learning_rate": 8e-05,
+      "loss": 1.7063,
+      "step": 2981
+    },
+    {
+      "epoch": 0.16622073578595317,
+      "grad_norm": 0.4607832431793213,
+      "learning_rate": 8e-05,
+      "loss": 1.6359,
+      "step": 2982
+    },
+    {
+      "epoch": 0.16627647714604235,
+      "grad_norm": 0.4844558835029602,
+      "learning_rate": 8e-05,
+      "loss": 1.5046,
+      "step": 2983
+    },
+    {
+      "epoch": 0.16633221850613156,
+      "grad_norm": 0.4657902717590332,
+      "learning_rate": 8e-05,
+      "loss": 1.7483,
+      "step": 2984
+    },
+    {
+      "epoch": 0.16638795986622074,
+      "grad_norm": 0.47202810645103455,
+      "learning_rate": 8e-05,
+      "loss": 1.5051,
+      "step": 2985
+    },
+    {
+      "epoch": 0.16644370122630991,
+      "grad_norm": 0.45511099696159363,
+      "learning_rate": 8e-05,
+      "loss": 1.8077,
+      "step": 2986
+    },
+    {
+      "epoch": 0.16649944258639912,
+      "grad_norm": 0.462038516998291,
+      "learning_rate": 8e-05,
+      "loss": 1.6827,
+      "step": 2987
+    },
+    {
+      "epoch": 0.1665551839464883,
+      "grad_norm": 0.4840214252471924,
+      "learning_rate": 8e-05,
+      "loss": 1.5835,
+      "step": 2988
+    },
+    {
+      "epoch": 0.16661092530657748,
+      "grad_norm": 0.5105730295181274,
+      "learning_rate": 8e-05,
+      "loss": 1.9062,
+      "step": 2989
+    },
+    {
+      "epoch": 0.16666666666666666,
+      "grad_norm": 0.4960428476333618,
+      "learning_rate": 8e-05,
+      "loss": 1.8725,
+      "step": 2990
+    },
+    {
+      "epoch": 0.16672240802675586,
+      "grad_norm": 0.479893296957016,
+      "learning_rate": 8e-05,
+      "loss": 1.5747,
+      "step": 2991
+    },
+    {
+      "epoch": 0.16677814938684504,
+      "grad_norm": 0.4770605266094208,
+      "learning_rate": 8e-05,
+      "loss": 1.7569,
+      "step": 2992
+    },
+    {
+      "epoch": 0.16683389074693422,
+      "grad_norm": 0.44722259044647217,
+      "learning_rate": 8e-05,
+      "loss": 1.6787,
+      "step": 2993
+    },
+    {
+      "epoch": 0.1668896321070234,
+      "grad_norm": 0.5155320763587952,
+      "learning_rate": 8e-05,
+      "loss": 1.6401,
+      "step": 2994
+    },
+    {
+      "epoch": 0.1669453734671126,
+      "grad_norm": 0.5102251172065735,
+      "learning_rate": 8e-05,
+      "loss": 1.7707,
+      "step": 2995
+    },
+    {
+      "epoch": 0.16700111482720179,
+      "grad_norm": 0.5326622724533081,
+      "learning_rate": 8e-05,
+      "loss": 1.9579,
+      "step": 2996
+    },
+    {
+      "epoch": 0.16705685618729096,
+      "grad_norm": 0.4790048897266388,
+      "learning_rate": 8e-05,
+      "loss": 1.5006,
+      "step": 2997
+    },
+    {
+      "epoch": 0.16711259754738014,
+      "grad_norm": 0.4893651604652405,
+      "learning_rate": 8e-05,
+      "loss": 1.7729,
+      "step": 2998
+    },
+    {
+      "epoch": 0.16716833890746935,
+      "grad_norm": 0.50043785572052,
+      "learning_rate": 8e-05,
+      "loss": 1.7786,
+      "step": 2999
+    },
+    {
+      "epoch": 0.16722408026755853,
+      "grad_norm": 0.48809880018234253,
+      "learning_rate": 8e-05,
+      "loss": 1.8332,
+      "step": 3000
+    },
+    {
+      "epoch": 0.1672798216276477,
+      "grad_norm": 0.4811587333679199,
+      "learning_rate": 8e-05,
+      "loss": 1.6124,
+      "step": 3001
+    },
+    {
+      "epoch": 0.1673355629877369,
+      "grad_norm": 0.5150866508483887,
+      "learning_rate": 8e-05,
+      "loss": 1.6237,
+      "step": 3002
+    },
+    {
+      "epoch": 0.1673913043478261,
+      "grad_norm": 0.5085856914520264,
+      "learning_rate": 8e-05,
+      "loss": 1.8273,
+      "step": 3003
+    },
+    {
+      "epoch": 0.16744704570791527,
+      "grad_norm": 0.5587477684020996,
+      "learning_rate": 8e-05,
+      "loss": 1.8274,
+      "step": 3004
+    },
+    {
+      "epoch": 0.16750278706800445,
+      "grad_norm": 0.4674791693687439,
+      "learning_rate": 8e-05,
+      "loss": 1.74,
+      "step": 3005
+    },
+    {
+      "epoch": 0.16755852842809366,
+      "grad_norm": 0.4883214235305786,
+      "learning_rate": 8e-05,
+      "loss": 1.6489,
+      "step": 3006
+    },
+    {
+      "epoch": 0.16761426978818283,
+      "grad_norm": 0.5089623928070068,
+      "learning_rate": 8e-05,
+      "loss": 1.7817,
+      "step": 3007
+    },
+    {
+      "epoch": 0.167670011148272,
+      "grad_norm": 0.5148787498474121,
+      "learning_rate": 8e-05,
+      "loss": 1.6104,
+      "step": 3008
+    },
+    {
+      "epoch": 0.1677257525083612,
+      "grad_norm": 0.4540792405605316,
+      "learning_rate": 8e-05,
+      "loss": 1.6607,
+      "step": 3009
+    },
+    {
+      "epoch": 0.1677814938684504,
+      "grad_norm": 0.4619336724281311,
+      "learning_rate": 8e-05,
+      "loss": 1.6425,
+      "step": 3010
+    },
+    {
+      "epoch": 0.16783723522853958,
+      "grad_norm": 0.44761553406715393,
+      "learning_rate": 8e-05,
+      "loss": 1.5614,
+      "step": 3011
+    },
+    {
+      "epoch": 0.16789297658862876,
+      "grad_norm": 0.471306174993515,
+      "learning_rate": 8e-05,
+      "loss": 1.5586,
+      "step": 3012
+    },
+    {
+      "epoch": 0.16794871794871793,
+      "grad_norm": 0.5210175514221191,
+      "learning_rate": 8e-05,
+      "loss": 1.8424,
+      "step": 3013
+    },
+    {
+      "epoch": 0.16800445930880714,
+      "grad_norm": 0.4540490508079529,
+      "learning_rate": 8e-05,
+      "loss": 1.5963,
+      "step": 3014
+    },
+    {
+      "epoch": 0.16806020066889632,
+      "grad_norm": 0.47507572174072266,
+      "learning_rate": 8e-05,
+      "loss": 1.632,
+      "step": 3015
+    },
+    {
+      "epoch": 0.1681159420289855,
+      "grad_norm": 0.47896212339401245,
+      "learning_rate": 8e-05,
+      "loss": 1.7958,
+      "step": 3016
+    },
+    {
+      "epoch": 0.1681716833890747,
+      "grad_norm": 0.48261168599128723,
+      "learning_rate": 8e-05,
+      "loss": 1.5484,
+      "step": 3017
+    },
+    {
+      "epoch": 0.16822742474916388,
+      "grad_norm": 0.47131311893463135,
+      "learning_rate": 8e-05,
+      "loss": 1.6909,
+      "step": 3018
+    },
+    {
+      "epoch": 0.16828316610925306,
+      "grad_norm": 0.47304922342300415,
+      "learning_rate": 8e-05,
+      "loss": 1.7071,
+      "step": 3019
+    },
+    {
+      "epoch": 0.16833890746934224,
+      "grad_norm": 0.4762814939022064,
+      "learning_rate": 8e-05,
+      "loss": 1.7035,
+      "step": 3020
+    },
+    {
+      "epoch": 0.16839464882943145,
+      "grad_norm": 0.5088526606559753,
+      "learning_rate": 8e-05,
+      "loss": 1.8491,
+      "step": 3021
+    },
+    {
+      "epoch": 0.16845039018952063,
+      "grad_norm": 0.48544204235076904,
+      "learning_rate": 8e-05,
+      "loss": 1.8434,
+      "step": 3022
+    },
+    {
+      "epoch": 0.1685061315496098,
+      "grad_norm": 0.49382224678993225,
+      "learning_rate": 8e-05,
+      "loss": 1.6825,
+      "step": 3023
+    },
+    {
+      "epoch": 0.16856187290969898,
+      "grad_norm": 0.4782942235469818,
+      "learning_rate": 8e-05,
+      "loss": 1.7259,
+      "step": 3024
+    },
+    {
+      "epoch": 0.1686176142697882,
+      "grad_norm": 0.555604875087738,
+      "learning_rate": 8e-05,
+      "loss": 1.9323,
+      "step": 3025
+    },
+    {
+      "epoch": 0.16867335562987737,
+      "grad_norm": 0.5326486825942993,
+      "learning_rate": 8e-05,
+      "loss": 1.612,
+      "step": 3026
+    },
+    {
+      "epoch": 0.16872909698996655,
+      "grad_norm": 0.5532233715057373,
+      "learning_rate": 8e-05,
+      "loss": 1.9169,
+      "step": 3027
+    },
+    {
+      "epoch": 0.16878483835005575,
+      "grad_norm": 0.8552114367485046,
+      "learning_rate": 8e-05,
+      "loss": 1.8408,
+      "step": 3028
+    },
+    {
+      "epoch": 0.16884057971014493,
+      "grad_norm": 0.544897198677063,
+      "learning_rate": 8e-05,
+      "loss": 2.0083,
+      "step": 3029
+    },
+    {
+      "epoch": 0.1688963210702341,
+      "grad_norm": 0.4543430805206299,
+      "learning_rate": 8e-05,
+      "loss": 1.5772,
+      "step": 3030
+    },
+    {
+      "epoch": 0.1689520624303233,
+      "grad_norm": 0.49418988823890686,
+      "learning_rate": 8e-05,
+      "loss": 1.7765,
+      "step": 3031
+    },
+    {
+      "epoch": 0.1690078037904125,
+      "grad_norm": 0.4817911386489868,
+      "learning_rate": 8e-05,
+      "loss": 1.6769,
+      "step": 3032
+    },
+    {
+      "epoch": 0.16906354515050168,
+      "grad_norm": 0.5358490943908691,
+      "learning_rate": 8e-05,
+      "loss": 1.6488,
+      "step": 3033
+    },
+    {
+      "epoch": 0.16911928651059085,
+      "grad_norm": 0.4769594669342041,
+      "learning_rate": 8e-05,
+      "loss": 1.7918,
+      "step": 3034
+    },
+    {
+      "epoch": 0.16917502787068003,
+      "grad_norm": 0.48201489448547363,
+      "learning_rate": 8e-05,
+      "loss": 1.8778,
+      "step": 3035
+    },
+    {
+      "epoch": 0.16923076923076924,
+      "grad_norm": 0.470387727022171,
+      "learning_rate": 8e-05,
+      "loss": 1.7893,
+      "step": 3036
+    },
+    {
+      "epoch": 0.16928651059085842,
+      "grad_norm": 0.49865806102752686,
+      "learning_rate": 8e-05,
+      "loss": 1.739,
+      "step": 3037
+    },
+    {
+      "epoch": 0.1693422519509476,
+      "grad_norm": 0.5022916793823242,
+      "learning_rate": 8e-05,
+      "loss": 1.6438,
+      "step": 3038
+    },
+    {
+      "epoch": 0.16939799331103678,
+      "grad_norm": 0.4783816337585449,
+      "learning_rate": 8e-05,
+      "loss": 1.6419,
+      "step": 3039
+    },
+    {
+      "epoch": 0.16945373467112598,
+      "grad_norm": 0.4603171646595001,
+      "learning_rate": 8e-05,
+      "loss": 1.5285,
+      "step": 3040
+    },
+    {
+      "epoch": 0.16950947603121516,
+      "grad_norm": 0.5255218148231506,
+      "learning_rate": 8e-05,
+      "loss": 2.0596,
+      "step": 3041
+    },
+    {
+      "epoch": 0.16956521739130434,
+      "grad_norm": 0.4812007546424866,
+      "learning_rate": 8e-05,
+      "loss": 1.892,
+      "step": 3042
+    },
+    {
+      "epoch": 0.16962095875139355,
+      "grad_norm": 0.47994357347488403,
+      "learning_rate": 8e-05,
+      "loss": 1.8251,
+      "step": 3043
+    },
+    {
+      "epoch": 0.16967670011148273,
+      "grad_norm": 0.43760350346565247,
+      "learning_rate": 8e-05,
+      "loss": 1.4635,
+      "step": 3044
+    },
+    {
+      "epoch": 0.1697324414715719,
+      "grad_norm": 0.465116411447525,
+      "learning_rate": 8e-05,
+      "loss": 1.8058,
+      "step": 3045
+    },
+    {
+      "epoch": 0.16978818283166108,
+      "grad_norm": 0.4570809006690979,
+      "learning_rate": 8e-05,
+      "loss": 1.5542,
+      "step": 3046
+    },
+    {
+      "epoch": 0.1698439241917503,
+      "grad_norm": 0.4452064335346222,
+      "learning_rate": 8e-05,
+      "loss": 1.6527,
+      "step": 3047
+    },
+    {
+      "epoch": 0.16989966555183947,
+      "grad_norm": 0.4564603269100189,
+      "learning_rate": 8e-05,
+      "loss": 1.5997,
+      "step": 3048
+    },
+    {
+      "epoch": 0.16995540691192865,
+      "grad_norm": 0.4724256694316864,
+      "learning_rate": 8e-05,
+      "loss": 1.6461,
+      "step": 3049
+    },
+    {
+      "epoch": 0.17001114827201783,
+      "grad_norm": 0.4666452705860138,
+      "learning_rate": 8e-05,
+      "loss": 1.8152,
+      "step": 3050
+    },
+    {
+      "epoch": 0.17006688963210703,
+      "grad_norm": 0.46496039628982544,
+      "learning_rate": 8e-05,
+      "loss": 1.6681,
+      "step": 3051
+    },
+    {
+      "epoch": 0.1701226309921962,
+      "grad_norm": 0.4887557923793793,
+      "learning_rate": 8e-05,
+      "loss": 1.6887,
+      "step": 3052
+    },
+    {
+      "epoch": 0.1701783723522854,
+      "grad_norm": 0.4998195767402649,
+      "learning_rate": 8e-05,
+      "loss": 1.8402,
+      "step": 3053
+    },
+    {
+      "epoch": 0.17023411371237457,
+      "grad_norm": 0.5062663555145264,
+      "learning_rate": 8e-05,
+      "loss": 1.8232,
+      "step": 3054
+    },
+    {
+      "epoch": 0.17028985507246377,
+      "grad_norm": 0.5044751167297363,
+      "learning_rate": 8e-05,
+      "loss": 1.506,
+      "step": 3055
+    },
+    {
+      "epoch": 0.17034559643255295,
+      "grad_norm": 0.49747434258461,
+      "learning_rate": 8e-05,
+      "loss": 1.7674,
+      "step": 3056
+    },
+    {
+      "epoch": 0.17040133779264213,
+      "grad_norm": 0.463445246219635,
+      "learning_rate": 8e-05,
+      "loss": 1.5947,
+      "step": 3057
+    },
+    {
+      "epoch": 0.17045707915273134,
+      "grad_norm": 0.46119794249534607,
+      "learning_rate": 8e-05,
+      "loss": 1.6294,
+      "step": 3058
+    },
+    {
+      "epoch": 0.17051282051282052,
+      "grad_norm": 0.44738781452178955,
+      "learning_rate": 8e-05,
+      "loss": 1.7148,
+      "step": 3059
+    },
+    {
+      "epoch": 0.1705685618729097,
+      "grad_norm": 0.47132405638694763,
+      "learning_rate": 8e-05,
+      "loss": 1.7297,
+      "step": 3060
+    },
+    {
+      "epoch": 0.17062430323299888,
+      "grad_norm": 0.523993968963623,
+      "learning_rate": 8e-05,
+      "loss": 1.8783,
+      "step": 3061
+    },
+    {
+      "epoch": 0.17068004459308808,
+      "grad_norm": 0.5431890487670898,
+      "learning_rate": 8e-05,
+      "loss": 1.8269,
+      "step": 3062
+    },
+    {
+      "epoch": 0.17073578595317726,
+      "grad_norm": 0.4824149012565613,
+      "learning_rate": 8e-05,
+      "loss": 1.4909,
+      "step": 3063
+    },
+    {
+      "epoch": 0.17079152731326644,
+      "grad_norm": 0.47434085607528687,
+      "learning_rate": 8e-05,
+      "loss": 1.7974,
+      "step": 3064
+    },
+    {
+      "epoch": 0.17084726867335562,
+      "grad_norm": 0.5017184019088745,
+      "learning_rate": 8e-05,
+      "loss": 1.6937,
+      "step": 3065
+    },
+    {
+      "epoch": 0.17090301003344482,
+      "grad_norm": 0.4829297363758087,
+      "learning_rate": 8e-05,
+      "loss": 1.656,
+      "step": 3066
+    },
+    {
+      "epoch": 0.170958751393534,
+      "grad_norm": 0.4362059235572815,
+      "learning_rate": 8e-05,
+      "loss": 1.3688,
+      "step": 3067
+    },
+    {
+      "epoch": 0.17101449275362318,
+      "grad_norm": 0.4372815489768982,
+      "learning_rate": 8e-05,
+      "loss": 1.4497,
+      "step": 3068
+    },
+    {
+      "epoch": 0.1710702341137124,
+      "grad_norm": 0.48382410407066345,
+      "learning_rate": 8e-05,
+      "loss": 1.8213,
+      "step": 3069
+    },
+    {
+      "epoch": 0.17112597547380157,
+      "grad_norm": 0.49431711435317993,
+      "learning_rate": 8e-05,
+      "loss": 1.7322,
+      "step": 3070
+    },
+    {
+      "epoch": 0.17118171683389075,
+      "grad_norm": 0.5378329157829285,
+      "learning_rate": 8e-05,
+      "loss": 1.7495,
+      "step": 3071
+    },
+    {
+      "epoch": 0.17123745819397992,
+      "grad_norm": 0.4819210469722748,
+      "learning_rate": 8e-05,
+      "loss": 1.4963,
+      "step": 3072
+    },
+    {
+      "epoch": 0.17129319955406913,
+      "grad_norm": 0.4515661597251892,
+      "learning_rate": 8e-05,
+      "loss": 1.6581,
+      "step": 3073
+    },
+    {
+      "epoch": 0.1713489409141583,
+      "grad_norm": 0.40329015254974365,
+      "learning_rate": 8e-05,
+      "loss": 1.5052,
+      "step": 3074
+    },
+    {
+      "epoch": 0.1714046822742475,
+      "grad_norm": 0.48003634810447693,
+      "learning_rate": 8e-05,
+      "loss": 1.8071,
+      "step": 3075
+    },
+    {
+      "epoch": 0.17146042363433667,
+      "grad_norm": 0.4498198628425598,
+      "learning_rate": 8e-05,
+      "loss": 1.6493,
+      "step": 3076
+    },
+    {
+      "epoch": 0.17151616499442587,
+      "grad_norm": 0.4563552737236023,
+      "learning_rate": 8e-05,
+      "loss": 1.5818,
+      "step": 3077
+    },
+    {
+      "epoch": 0.17157190635451505,
+      "grad_norm": 0.5010289549827576,
+      "learning_rate": 8e-05,
+      "loss": 1.8007,
+      "step": 3078
+    },
+    {
+      "epoch": 0.17162764771460423,
+      "grad_norm": 0.4739150106906891,
+      "learning_rate": 8e-05,
+      "loss": 1.6223,
+      "step": 3079
+    },
+    {
+      "epoch": 0.1716833890746934,
+      "grad_norm": 0.48382750153541565,
+      "learning_rate": 8e-05,
+      "loss": 1.635,
+      "step": 3080
+    },
+    {
+      "epoch": 0.17173913043478262,
+      "grad_norm": 0.49976804852485657,
+      "learning_rate": 8e-05,
+      "loss": 1.6488,
+      "step": 3081
+    },
+    {
+      "epoch": 0.1717948717948718,
+      "grad_norm": 0.4661661386489868,
+      "learning_rate": 8e-05,
+      "loss": 1.6358,
+      "step": 3082
+    },
+    {
+      "epoch": 0.17185061315496097,
+      "grad_norm": 0.4675019085407257,
+      "learning_rate": 8e-05,
+      "loss": 1.7324,
+      "step": 3083
+    },
+    {
+      "epoch": 0.17190635451505018,
+      "grad_norm": 0.497050940990448,
+      "learning_rate": 8e-05,
+      "loss": 1.8559,
+      "step": 3084
+    },
+    {
+      "epoch": 0.17196209587513936,
+      "grad_norm": 0.4740678071975708,
+      "learning_rate": 8e-05,
+      "loss": 1.7249,
+      "step": 3085
+    },
+    {
+      "epoch": 0.17201783723522854,
+      "grad_norm": 0.5225610136985779,
+      "learning_rate": 8e-05,
+      "loss": 1.8596,
+      "step": 3086
+    },
+    {
+      "epoch": 0.17207357859531772,
+      "grad_norm": 0.4995535612106323,
+      "learning_rate": 8e-05,
+      "loss": 1.7586,
+      "step": 3087
+    },
+    {
+      "epoch": 0.17212931995540692,
+      "grad_norm": 0.5729968547821045,
+      "learning_rate": 8e-05,
+      "loss": 1.6392,
+      "step": 3088
+    },
+    {
+      "epoch": 0.1721850613154961,
+      "grad_norm": 0.4812772870063782,
+      "learning_rate": 8e-05,
+      "loss": 1.8687,
+      "step": 3089
+    },
+    {
+      "epoch": 0.17224080267558528,
+      "grad_norm": 0.4795670211315155,
+      "learning_rate": 8e-05,
+      "loss": 1.7171,
+      "step": 3090
+    },
+    {
+      "epoch": 0.17229654403567446,
+      "grad_norm": 0.49664998054504395,
+      "learning_rate": 8e-05,
+      "loss": 1.553,
+      "step": 3091
+    },
+    {
+      "epoch": 0.17235228539576367,
+      "grad_norm": 0.4796065092086792,
+      "learning_rate": 8e-05,
+      "loss": 1.9998,
+      "step": 3092
+    },
+    {
+      "epoch": 0.17240802675585284,
+      "grad_norm": 0.5262255072593689,
+      "learning_rate": 8e-05,
+      "loss": 1.8668,
+      "step": 3093
+    },
+    {
+      "epoch": 0.17246376811594202,
+      "grad_norm": 0.46140360832214355,
+      "learning_rate": 8e-05,
+      "loss": 1.5494,
+      "step": 3094
+    },
+    {
+      "epoch": 0.1725195094760312,
+      "grad_norm": 0.4980575144290924,
+      "learning_rate": 8e-05,
+      "loss": 1.7239,
+      "step": 3095
+    },
+    {
+      "epoch": 0.1725752508361204,
+      "grad_norm": 0.49190661311149597,
+      "learning_rate": 8e-05,
+      "loss": 1.7526,
+      "step": 3096
+    },
+    {
+      "epoch": 0.1726309921962096,
+      "grad_norm": 0.4727287292480469,
+      "learning_rate": 8e-05,
+      "loss": 1.9271,
+      "step": 3097
+    },
+    {
+      "epoch": 0.17268673355629877,
+      "grad_norm": 0.4625808894634247,
+      "learning_rate": 8e-05,
+      "loss": 1.6226,
+      "step": 3098
+    },
+    {
+      "epoch": 0.17274247491638797,
+      "grad_norm": 0.534032940864563,
+      "learning_rate": 8e-05,
+      "loss": 1.8686,
+      "step": 3099
+    },
+    {
+      "epoch": 0.17279821627647715,
+      "grad_norm": 0.47960618138313293,
+      "learning_rate": 8e-05,
+      "loss": 1.6062,
+      "step": 3100
+    },
+    {
+      "epoch": 0.17285395763656633,
+      "grad_norm": 0.47547629475593567,
+      "learning_rate": 8e-05,
+      "loss": 1.7415,
+      "step": 3101
+    },
+    {
+      "epoch": 0.1729096989966555,
+      "grad_norm": 0.5057584643363953,
+      "learning_rate": 8e-05,
+      "loss": 1.8651,
+      "step": 3102
+    },
+    {
+      "epoch": 0.17296544035674472,
+      "grad_norm": 0.4372791647911072,
+      "learning_rate": 8e-05,
+      "loss": 1.6305,
+      "step": 3103
+    },
+    {
+      "epoch": 0.1730211817168339,
+      "grad_norm": 0.4549502432346344,
+      "learning_rate": 8e-05,
+      "loss": 1.5812,
+      "step": 3104
+    },
+    {
+      "epoch": 0.17307692307692307,
+      "grad_norm": 0.5240883827209473,
+      "learning_rate": 8e-05,
+      "loss": 1.8658,
+      "step": 3105
+    },
+    {
+      "epoch": 0.17313266443701225,
+      "grad_norm": 0.5285409092903137,
+      "learning_rate": 8e-05,
+      "loss": 1.7676,
+      "step": 3106
+    },
+    {
+      "epoch": 0.17318840579710146,
+      "grad_norm": 0.5472353100776672,
+      "learning_rate": 8e-05,
+      "loss": 1.8128,
+      "step": 3107
+    },
+    {
+      "epoch": 0.17324414715719064,
+      "grad_norm": 0.48296090960502625,
+      "learning_rate": 8e-05,
+      "loss": 1.644,
+      "step": 3108
+    },
+    {
+      "epoch": 0.17329988851727982,
+      "grad_norm": 0.5111985802650452,
+      "learning_rate": 8e-05,
+      "loss": 1.7872,
+      "step": 3109
+    },
+    {
+      "epoch": 0.173355629877369,
+      "grad_norm": 0.5271138548851013,
+      "learning_rate": 8e-05,
+      "loss": 1.9658,
+      "step": 3110
+    },
+    {
+      "epoch": 0.1734113712374582,
+      "grad_norm": 0.4623909294605255,
+      "learning_rate": 8e-05,
+      "loss": 1.7587,
+      "step": 3111
+    },
+    {
+      "epoch": 0.17346711259754738,
+      "grad_norm": 0.5246699452400208,
+      "learning_rate": 8e-05,
+      "loss": 1.8177,
+      "step": 3112
+    },
+    {
+      "epoch": 0.17352285395763656,
+      "grad_norm": 0.4383607506752014,
+      "learning_rate": 8e-05,
+      "loss": 1.6104,
+      "step": 3113
+    },
+    {
+      "epoch": 0.17357859531772576,
+      "grad_norm": 0.5097667574882507,
+      "learning_rate": 8e-05,
+      "loss": 1.5114,
+      "step": 3114
+    },
+    {
+      "epoch": 0.17363433667781494,
+      "grad_norm": 0.4617764353752136,
+      "learning_rate": 8e-05,
+      "loss": 1.6502,
+      "step": 3115
+    },
+    {
+      "epoch": 0.17369007803790412,
+      "grad_norm": 0.501473605632782,
+      "learning_rate": 8e-05,
+      "loss": 1.9709,
+      "step": 3116
+    },
+    {
+      "epoch": 0.1737458193979933,
+      "grad_norm": 0.4450080990791321,
+      "learning_rate": 8e-05,
+      "loss": 1.6331,
+      "step": 3117
+    },
+    {
+      "epoch": 0.1738015607580825,
+      "grad_norm": 0.4507881999015808,
+      "learning_rate": 8e-05,
+      "loss": 1.6415,
+      "step": 3118
+    },
+    {
+      "epoch": 0.17385730211817169,
+      "grad_norm": 0.48858508467674255,
+      "learning_rate": 8e-05,
+      "loss": 1.7731,
+      "step": 3119
+    },
+    {
+      "epoch": 0.17391304347826086,
+      "grad_norm": 0.46448734402656555,
+      "learning_rate": 8e-05,
+      "loss": 1.5604,
+      "step": 3120
+    },
+    {
+      "epoch": 0.17396878483835004,
+      "grad_norm": 0.4925326108932495,
+      "learning_rate": 8e-05,
+      "loss": 1.7302,
+      "step": 3121
+    },
+    {
+      "epoch": 0.17402452619843925,
+      "grad_norm": 0.45742926001548767,
+      "learning_rate": 8e-05,
+      "loss": 1.6341,
+      "step": 3122
+    },
+    {
+      "epoch": 0.17408026755852843,
+      "grad_norm": 0.4373356103897095,
+      "learning_rate": 8e-05,
+      "loss": 1.5657,
+      "step": 3123
+    },
+    {
+      "epoch": 0.1741360089186176,
+      "grad_norm": 0.4854992628097534,
+      "learning_rate": 8e-05,
+      "loss": 1.5725,
+      "step": 3124
+    },
+    {
+      "epoch": 0.17419175027870681,
+      "grad_norm": 0.5004236102104187,
+      "learning_rate": 8e-05,
+      "loss": 1.7323,
+      "step": 3125
+    },
+    {
+      "epoch": 0.174247491638796,
+      "grad_norm": 0.47592586278915405,
+      "learning_rate": 8e-05,
+      "loss": 1.7518,
+      "step": 3126
+    },
+    {
+      "epoch": 0.17430323299888517,
+      "grad_norm": 0.4757702350616455,
+      "learning_rate": 8e-05,
+      "loss": 1.6715,
+      "step": 3127
+    },
+    {
+      "epoch": 0.17435897435897435,
+      "grad_norm": 0.5010035634040833,
+      "learning_rate": 8e-05,
+      "loss": 1.7594,
+      "step": 3128
+    },
+    {
+      "epoch": 0.17441471571906356,
+      "grad_norm": 0.48768213391304016,
+      "learning_rate": 8e-05,
+      "loss": 1.7032,
+      "step": 3129
+    },
+    {
+      "epoch": 0.17447045707915274,
+      "grad_norm": 0.4932835102081299,
+      "learning_rate": 8e-05,
+      "loss": 1.8724,
+      "step": 3130
+    },
+    {
+      "epoch": 0.17452619843924191,
+      "grad_norm": 0.5223470330238342,
+      "learning_rate": 8e-05,
+      "loss": 1.8051,
+      "step": 3131
+    },
+    {
+      "epoch": 0.1745819397993311,
+      "grad_norm": 0.5153441429138184,
+      "learning_rate": 8e-05,
+      "loss": 1.9186,
+      "step": 3132
+    },
+    {
+      "epoch": 0.1746376811594203,
+      "grad_norm": 0.4976043701171875,
+      "learning_rate": 8e-05,
+      "loss": 1.7093,
+      "step": 3133
+    },
+    {
+      "epoch": 0.17469342251950948,
+      "grad_norm": 0.5194041132926941,
+      "learning_rate": 8e-05,
+      "loss": 1.8921,
+      "step": 3134
+    },
+    {
+      "epoch": 0.17474916387959866,
+      "grad_norm": 0.48766300082206726,
+      "learning_rate": 8e-05,
+      "loss": 1.8425,
+      "step": 3135
+    },
+    {
+      "epoch": 0.17480490523968784,
+      "grad_norm": 0.4757583439350128,
+      "learning_rate": 8e-05,
+      "loss": 1.7665,
+      "step": 3136
+    },
+    {
+      "epoch": 0.17486064659977704,
+      "grad_norm": 0.4470495879650116,
+      "learning_rate": 8e-05,
+      "loss": 1.5407,
+      "step": 3137
+    },
+    {
+      "epoch": 0.17491638795986622,
+      "grad_norm": 0.5015825033187866,
+      "learning_rate": 8e-05,
+      "loss": 1.5779,
+      "step": 3138
+    },
+    {
+      "epoch": 0.1749721293199554,
+      "grad_norm": 0.49691611528396606,
+      "learning_rate": 8e-05,
+      "loss": 1.9104,
+      "step": 3139
+    },
+    {
+      "epoch": 0.1750278706800446,
+      "grad_norm": 0.46983975172042847,
+      "learning_rate": 8e-05,
+      "loss": 1.6565,
+      "step": 3140
+    },
+    {
+      "epoch": 0.17508361204013378,
+      "grad_norm": 0.4764535129070282,
+      "learning_rate": 8e-05,
+      "loss": 1.8129,
+      "step": 3141
+    },
+    {
+      "epoch": 0.17513935340022296,
+      "grad_norm": 0.5874860286712646,
+      "learning_rate": 8e-05,
+      "loss": 2.2321,
+      "step": 3142
+    },
+    {
+      "epoch": 0.17519509476031214,
+      "grad_norm": 0.48502081632614136,
+      "learning_rate": 8e-05,
+      "loss": 1.7636,
+      "step": 3143
+    },
+    {
+      "epoch": 0.17525083612040135,
+      "grad_norm": 0.46429380774497986,
+      "learning_rate": 8e-05,
+      "loss": 1.7481,
+      "step": 3144
+    },
+    {
+      "epoch": 0.17530657748049053,
+      "grad_norm": 0.4960067570209503,
+      "learning_rate": 8e-05,
+      "loss": 1.7872,
+      "step": 3145
+    },
+    {
+      "epoch": 0.1753623188405797,
+      "grad_norm": 0.46121087670326233,
+      "learning_rate": 8e-05,
+      "loss": 1.6306,
+      "step": 3146
+    },
+    {
+      "epoch": 0.17541806020066888,
+      "grad_norm": 0.497938871383667,
+      "learning_rate": 8e-05,
+      "loss": 1.76,
+      "step": 3147
+    },
+    {
+      "epoch": 0.1754738015607581,
+      "grad_norm": 0.45689964294433594,
+      "learning_rate": 8e-05,
+      "loss": 1.5589,
+      "step": 3148
+    },
+    {
+      "epoch": 0.17552954292084727,
+      "grad_norm": 0.4511463940143585,
+      "learning_rate": 8e-05,
+      "loss": 1.7215,
+      "step": 3149
+    },
+    {
+      "epoch": 0.17558528428093645,
+      "grad_norm": 0.48650410771369934,
+      "learning_rate": 8e-05,
+      "loss": 1.6654,
+      "step": 3150
+    },
+    {
+      "epoch": 0.17564102564102563,
+      "grad_norm": 0.5025672316551208,
+      "learning_rate": 8e-05,
+      "loss": 1.638,
+      "step": 3151
+    },
+    {
+      "epoch": 0.17569676700111483,
+      "grad_norm": 0.4942224621772766,
+      "learning_rate": 8e-05,
+      "loss": 1.6517,
+      "step": 3152
+    },
+    {
+      "epoch": 0.175752508361204,
+      "grad_norm": 0.471215158700943,
+      "learning_rate": 8e-05,
+      "loss": 1.7378,
+      "step": 3153
+    },
+    {
+      "epoch": 0.1758082497212932,
+      "grad_norm": 0.5158399343490601,
+      "learning_rate": 8e-05,
+      "loss": 1.6319,
+      "step": 3154
+    },
+    {
+      "epoch": 0.1758639910813824,
+      "grad_norm": 0.47845038771629333,
+      "learning_rate": 8e-05,
+      "loss": 1.7183,
+      "step": 3155
+    },
+    {
+      "epoch": 0.17591973244147158,
+      "grad_norm": 0.5168014764785767,
+      "learning_rate": 8e-05,
+      "loss": 1.7709,
+      "step": 3156
+    },
+    {
+      "epoch": 0.17597547380156076,
+      "grad_norm": 0.4960615336894989,
+      "learning_rate": 8e-05,
+      "loss": 1.7438,
+      "step": 3157
+    },
+    {
+      "epoch": 0.17603121516164993,
+      "grad_norm": 0.4402819871902466,
+      "learning_rate": 8e-05,
+      "loss": 1.3315,
+      "step": 3158
+    },
+    {
+      "epoch": 0.17608695652173914,
+      "grad_norm": 0.4752297103404999,
+      "learning_rate": 8e-05,
+      "loss": 1.7649,
+      "step": 3159
+    },
+    {
+      "epoch": 0.17614269788182832,
+      "grad_norm": 0.4856758713722229,
+      "learning_rate": 8e-05,
+      "loss": 1.678,
+      "step": 3160
+    },
+    {
+      "epoch": 0.1761984392419175,
+      "grad_norm": 0.4814453721046448,
+      "learning_rate": 8e-05,
+      "loss": 1.7371,
+      "step": 3161
+    },
+    {
+      "epoch": 0.17625418060200668,
+      "grad_norm": 0.4642793536186218,
+      "learning_rate": 8e-05,
+      "loss": 1.6614,
+      "step": 3162
+    },
+    {
+      "epoch": 0.17630992196209588,
+      "grad_norm": 0.4925251305103302,
+      "learning_rate": 8e-05,
+      "loss": 1.5645,
+      "step": 3163
+    },
+    {
+      "epoch": 0.17636566332218506,
+      "grad_norm": 0.52099609375,
+      "learning_rate": 8e-05,
+      "loss": 1.6317,
+      "step": 3164
+    },
+    {
+      "epoch": 0.17642140468227424,
+      "grad_norm": 0.4923543930053711,
+      "learning_rate": 8e-05,
+      "loss": 1.6023,
+      "step": 3165
+    },
+    {
+      "epoch": 0.17647714604236345,
+      "grad_norm": 0.522320568561554,
+      "learning_rate": 8e-05,
+      "loss": 1.7948,
+      "step": 3166
+    },
+    {
+      "epoch": 0.17653288740245263,
+      "grad_norm": 0.48389363288879395,
+      "learning_rate": 8e-05,
+      "loss": 1.7495,
+      "step": 3167
+    },
+    {
+      "epoch": 0.1765886287625418,
+      "grad_norm": 0.5073885321617126,
+      "learning_rate": 8e-05,
+      "loss": 1.9295,
+      "step": 3168
+    },
+    {
+      "epoch": 0.17664437012263098,
+      "grad_norm": 0.45694851875305176,
+      "learning_rate": 8e-05,
+      "loss": 1.6847,
+      "step": 3169
+    },
+    {
+      "epoch": 0.1767001114827202,
+      "grad_norm": 0.48219072818756104,
+      "learning_rate": 8e-05,
+      "loss": 1.8583,
+      "step": 3170
+    },
+    {
+      "epoch": 0.17675585284280937,
+      "grad_norm": 0.4569888710975647,
+      "learning_rate": 8e-05,
+      "loss": 1.8423,
+      "step": 3171
+    },
+    {
+      "epoch": 0.17681159420289855,
+      "grad_norm": 0.49059024453163147,
+      "learning_rate": 8e-05,
+      "loss": 1.798,
+      "step": 3172
+    },
+    {
+      "epoch": 0.17686733556298773,
+      "grad_norm": 0.5194540619850159,
+      "learning_rate": 8e-05,
+      "loss": 1.5778,
+      "step": 3173
+    },
+    {
+      "epoch": 0.17692307692307693,
+      "grad_norm": 0.5097144842147827,
+      "learning_rate": 8e-05,
+      "loss": 1.5033,
+      "step": 3174
+    },
+    {
+      "epoch": 0.1769788182831661,
+      "grad_norm": 0.48274803161621094,
+      "learning_rate": 8e-05,
+      "loss": 1.7761,
+      "step": 3175
+    },
+    {
+      "epoch": 0.1770345596432553,
+      "grad_norm": 0.5038658976554871,
+      "learning_rate": 8e-05,
+      "loss": 1.7308,
+      "step": 3176
+    },
+    {
+      "epoch": 0.17709030100334447,
+      "grad_norm": 0.46520641446113586,
+      "learning_rate": 8e-05,
+      "loss": 1.6843,
+      "step": 3177
+    },
+    {
+      "epoch": 0.17714604236343368,
+      "grad_norm": 0.47753003239631653,
+      "learning_rate": 8e-05,
+      "loss": 1.7283,
+      "step": 3178
+    },
+    {
+      "epoch": 0.17720178372352285,
+      "grad_norm": 0.5261279940605164,
+      "learning_rate": 8e-05,
+      "loss": 1.86,
+      "step": 3179
+    },
+    {
+      "epoch": 0.17725752508361203,
+      "grad_norm": 0.497805655002594,
+      "learning_rate": 8e-05,
+      "loss": 1.6267,
+      "step": 3180
+    },
+    {
+      "epoch": 0.17731326644370124,
+      "grad_norm": 0.464225172996521,
+      "learning_rate": 8e-05,
+      "loss": 1.7415,
+      "step": 3181
+    },
+    {
+      "epoch": 0.17736900780379042,
+      "grad_norm": 0.4999517798423767,
+      "learning_rate": 8e-05,
+      "loss": 1.6564,
+      "step": 3182
+    },
+    {
+      "epoch": 0.1774247491638796,
+      "grad_norm": 0.5199729800224304,
+      "learning_rate": 8e-05,
+      "loss": 1.9543,
+      "step": 3183
+    },
+    {
+      "epoch": 0.17748049052396878,
+      "grad_norm": 0.48327144980430603,
+      "learning_rate": 8e-05,
+      "loss": 1.5824,
+      "step": 3184
+    },
+    {
+      "epoch": 0.17753623188405798,
+      "grad_norm": 0.5581362843513489,
+      "learning_rate": 8e-05,
+      "loss": 1.9749,
+      "step": 3185
+    },
+    {
+      "epoch": 0.17759197324414716,
+      "grad_norm": 0.4988551437854767,
+      "learning_rate": 8e-05,
+      "loss": 1.9346,
+      "step": 3186
+    },
+    {
+      "epoch": 0.17764771460423634,
+      "grad_norm": 0.5109734535217285,
+      "learning_rate": 8e-05,
+      "loss": 1.7523,
+      "step": 3187
+    },
+    {
+      "epoch": 0.17770345596432552,
+      "grad_norm": 0.4862396717071533,
+      "learning_rate": 8e-05,
+      "loss": 1.5092,
+      "step": 3188
+    },
+    {
+      "epoch": 0.17775919732441472,
+      "grad_norm": 0.44461333751678467,
+      "learning_rate": 8e-05,
+      "loss": 1.6452,
+      "step": 3189
+    },
+    {
+      "epoch": 0.1778149386845039,
+      "grad_norm": 0.4702376425266266,
+      "learning_rate": 8e-05,
+      "loss": 1.6332,
+      "step": 3190
+    },
+    {
+      "epoch": 0.17787068004459308,
+      "grad_norm": 0.48019805550575256,
+      "learning_rate": 8e-05,
+      "loss": 1.5162,
+      "step": 3191
+    },
+    {
+      "epoch": 0.17792642140468226,
+      "grad_norm": 0.49448859691619873,
+      "learning_rate": 8e-05,
+      "loss": 1.6892,
+      "step": 3192
+    },
+    {
+      "epoch": 0.17798216276477147,
+      "grad_norm": 0.4584326446056366,
+      "learning_rate": 8e-05,
+      "loss": 1.6477,
+      "step": 3193
+    },
+    {
+      "epoch": 0.17803790412486065,
+      "grad_norm": 0.49770230054855347,
+      "learning_rate": 8e-05,
+      "loss": 1.7618,
+      "step": 3194
+    },
+    {
+      "epoch": 0.17809364548494983,
+      "grad_norm": 0.4516306221485138,
+      "learning_rate": 8e-05,
+      "loss": 1.7175,
+      "step": 3195
+    },
+    {
+      "epoch": 0.17814938684503903,
+      "grad_norm": 0.4722249507904053,
+      "learning_rate": 8e-05,
+      "loss": 1.4028,
+      "step": 3196
+    },
+    {
+      "epoch": 0.1782051282051282,
+      "grad_norm": 0.5114902853965759,
+      "learning_rate": 8e-05,
+      "loss": 1.7978,
+      "step": 3197
+    },
+    {
+      "epoch": 0.1782608695652174,
+      "grad_norm": 0.5099138021469116,
+      "learning_rate": 8e-05,
+      "loss": 2.0185,
+      "step": 3198
+    },
+    {
+      "epoch": 0.17831661092530657,
+      "grad_norm": 0.5226367712020874,
+      "learning_rate": 8e-05,
+      "loss": 1.8694,
+      "step": 3199
+    },
+    {
+      "epoch": 0.17837235228539577,
+      "grad_norm": 0.5212682485580444,
+      "learning_rate": 8e-05,
+      "loss": 1.9197,
+      "step": 3200
+    },
+    {
+      "epoch": 0.17842809364548495,
+      "grad_norm": 0.5298042893409729,
+      "learning_rate": 8e-05,
+      "loss": 1.9656,
+      "step": 3201
+    },
+    {
+      "epoch": 0.17848383500557413,
+      "grad_norm": 0.5426698327064514,
+      "learning_rate": 8e-05,
+      "loss": 1.6831,
+      "step": 3202
+    },
+    {
+      "epoch": 0.1785395763656633,
+      "grad_norm": 0.47143349051475525,
+      "learning_rate": 8e-05,
+      "loss": 1.6756,
+      "step": 3203
+    },
+    {
+      "epoch": 0.17859531772575252,
+      "grad_norm": 0.4910406470298767,
+      "learning_rate": 8e-05,
+      "loss": 1.9391,
+      "step": 3204
+    },
+    {
+      "epoch": 0.1786510590858417,
+      "grad_norm": 0.5471988320350647,
+      "learning_rate": 8e-05,
+      "loss": 2.0599,
+      "step": 3205
+    },
+    {
+      "epoch": 0.17870680044593087,
+      "grad_norm": 0.4846929907798767,
+      "learning_rate": 8e-05,
+      "loss": 1.6376,
+      "step": 3206
+    },
+    {
+      "epoch": 0.17876254180602005,
+      "grad_norm": 0.5019040703773499,
+      "learning_rate": 8e-05,
+      "loss": 1.8373,
+      "step": 3207
+    },
+    {
+      "epoch": 0.17881828316610926,
+      "grad_norm": 0.5033964514732361,
+      "learning_rate": 8e-05,
+      "loss": 1.8035,
+      "step": 3208
+    },
+    {
+      "epoch": 0.17887402452619844,
+      "grad_norm": 0.5568781495094299,
+      "learning_rate": 8e-05,
+      "loss": 1.7519,
+      "step": 3209
+    },
+    {
+      "epoch": 0.17892976588628762,
+      "grad_norm": 0.5149691700935364,
+      "learning_rate": 8e-05,
+      "loss": 1.7481,
+      "step": 3210
+    },
+    {
+      "epoch": 0.17898550724637682,
+      "grad_norm": 0.47844332456588745,
+      "learning_rate": 8e-05,
+      "loss": 1.8602,
+      "step": 3211
+    },
+    {
+      "epoch": 0.179041248606466,
+      "grad_norm": 0.5395461916923523,
+      "learning_rate": 8e-05,
+      "loss": 1.764,
+      "step": 3212
+    },
+    {
+      "epoch": 0.17909698996655518,
+      "grad_norm": 0.4720330238342285,
+      "learning_rate": 8e-05,
+      "loss": 1.7078,
+      "step": 3213
+    },
+    {
+      "epoch": 0.17915273132664436,
+      "grad_norm": 0.458877295255661,
+      "learning_rate": 8e-05,
+      "loss": 1.7795,
+      "step": 3214
+    },
+    {
+      "epoch": 0.17920847268673357,
+      "grad_norm": 0.500705361366272,
+      "learning_rate": 8e-05,
+      "loss": 1.7592,
+      "step": 3215
+    },
+    {
+      "epoch": 0.17926421404682275,
+      "grad_norm": 0.5020824074745178,
+      "learning_rate": 8e-05,
+      "loss": 1.7152,
+      "step": 3216
+    },
+    {
+      "epoch": 0.17931995540691192,
+      "grad_norm": 0.4906066060066223,
+      "learning_rate": 8e-05,
+      "loss": 1.8397,
+      "step": 3217
+    },
+    {
+      "epoch": 0.1793756967670011,
+      "grad_norm": 0.4836278557777405,
+      "learning_rate": 8e-05,
+      "loss": 1.6074,
+      "step": 3218
+    },
+    {
+      "epoch": 0.1794314381270903,
+      "grad_norm": 0.49199914932250977,
+      "learning_rate": 8e-05,
+      "loss": 1.732,
+      "step": 3219
+    },
+    {
+      "epoch": 0.1794871794871795,
+      "grad_norm": 0.5196998715400696,
+      "learning_rate": 8e-05,
+      "loss": 1.7605,
+      "step": 3220
+    },
+    {
+      "epoch": 0.17954292084726867,
+      "grad_norm": 0.453242689371109,
+      "learning_rate": 8e-05,
+      "loss": 1.5641,
+      "step": 3221
+    },
+    {
+      "epoch": 0.17959866220735787,
+      "grad_norm": 0.43356654047966003,
+      "learning_rate": 8e-05,
+      "loss": 1.5325,
+      "step": 3222
+    },
+    {
+      "epoch": 0.17965440356744705,
+      "grad_norm": 0.4988141655921936,
+      "learning_rate": 8e-05,
+      "loss": 1.5707,
+      "step": 3223
+    },
+    {
+      "epoch": 0.17971014492753623,
+      "grad_norm": 0.49998071789741516,
+      "learning_rate": 8e-05,
+      "loss": 1.849,
+      "step": 3224
+    },
+    {
+      "epoch": 0.1797658862876254,
+      "grad_norm": 0.5347229838371277,
+      "learning_rate": 8e-05,
+      "loss": 1.9297,
+      "step": 3225
+    },
+    {
+      "epoch": 0.17982162764771462,
+      "grad_norm": 0.49117156863212585,
+      "learning_rate": 8e-05,
+      "loss": 1.6969,
+      "step": 3226
+    },
+    {
+      "epoch": 0.1798773690078038,
+      "grad_norm": 0.5386723875999451,
+      "learning_rate": 8e-05,
+      "loss": 1.8024,
+      "step": 3227
+    },
+    {
+      "epoch": 0.17993311036789297,
+      "grad_norm": 0.48068922758102417,
+      "learning_rate": 8e-05,
+      "loss": 1.6902,
+      "step": 3228
+    },
+    {
+      "epoch": 0.17998885172798215,
+      "grad_norm": 0.46101266145706177,
+      "learning_rate": 8e-05,
+      "loss": 1.5828,
+      "step": 3229
+    },
+    {
+      "epoch": 0.18004459308807136,
+      "grad_norm": 0.4808231294155121,
+      "learning_rate": 8e-05,
+      "loss": 1.6905,
+      "step": 3230
+    },
+    {
+      "epoch": 0.18010033444816054,
+      "grad_norm": 0.5057591795921326,
+      "learning_rate": 8e-05,
+      "loss": 1.8851,
+      "step": 3231
+    },
+    {
+      "epoch": 0.18015607580824972,
+      "grad_norm": 0.47574618458747864,
+      "learning_rate": 8e-05,
+      "loss": 1.6438,
+      "step": 3232
+    },
+    {
+      "epoch": 0.1802118171683389,
+      "grad_norm": 0.4971584677696228,
+      "learning_rate": 8e-05,
+      "loss": 1.8116,
+      "step": 3233
+    },
+    {
+      "epoch": 0.1802675585284281,
+      "grad_norm": 0.5194235444068909,
+      "learning_rate": 8e-05,
+      "loss": 1.8604,
+      "step": 3234
+    },
+    {
+      "epoch": 0.18032329988851728,
+      "grad_norm": 0.4879535138607025,
+      "learning_rate": 8e-05,
+      "loss": 1.7534,
+      "step": 3235
+    },
+    {
+      "epoch": 0.18037904124860646,
+      "grad_norm": 0.5174830555915833,
+      "learning_rate": 8e-05,
+      "loss": 1.775,
+      "step": 3236
+    },
+    {
+      "epoch": 0.18043478260869567,
+      "grad_norm": 0.5025219917297363,
+      "learning_rate": 8e-05,
+      "loss": 1.911,
+      "step": 3237
+    },
+    {
+      "epoch": 0.18049052396878484,
+      "grad_norm": 0.5307998061180115,
+      "learning_rate": 8e-05,
+      "loss": 1.7317,
+      "step": 3238
+    },
+    {
+      "epoch": 0.18054626532887402,
+      "grad_norm": 0.47522616386413574,
+      "learning_rate": 8e-05,
+      "loss": 1.7512,
+      "step": 3239
+    },
+    {
+      "epoch": 0.1806020066889632,
+      "grad_norm": 0.4941982626914978,
+      "learning_rate": 8e-05,
+      "loss": 1.8178,
+      "step": 3240
+    },
+    {
+      "epoch": 0.1806577480490524,
+      "grad_norm": 0.5120488405227661,
+      "learning_rate": 8e-05,
+      "loss": 1.7547,
+      "step": 3241
+    },
+    {
+      "epoch": 0.1807134894091416,
+      "grad_norm": 0.5394561290740967,
+      "learning_rate": 8e-05,
+      "loss": 2.0231,
+      "step": 3242
+    },
+    {
+      "epoch": 0.18076923076923077,
+      "grad_norm": 0.45192497968673706,
+      "learning_rate": 8e-05,
+      "loss": 1.4677,
+      "step": 3243
+    },
+    {
+      "epoch": 0.18082497212931994,
+      "grad_norm": 0.4537535011768341,
+      "learning_rate": 8e-05,
+      "loss": 1.6961,
+      "step": 3244
+    },
+    {
+      "epoch": 0.18088071348940915,
+      "grad_norm": 0.5035440325737,
+      "learning_rate": 8e-05,
+      "loss": 1.7087,
+      "step": 3245
+    },
+    {
+      "epoch": 0.18093645484949833,
+      "grad_norm": 0.4700638949871063,
+      "learning_rate": 8e-05,
+      "loss": 1.7097,
+      "step": 3246
+    },
+    {
+      "epoch": 0.1809921962095875,
+      "grad_norm": 0.47896021604537964,
+      "learning_rate": 8e-05,
+      "loss": 1.724,
+      "step": 3247
+    },
+    {
+      "epoch": 0.1810479375696767,
+      "grad_norm": 0.5172485113143921,
+      "learning_rate": 8e-05,
+      "loss": 1.8554,
+      "step": 3248
+    },
+    {
+      "epoch": 0.1811036789297659,
+      "grad_norm": 0.4391520023345947,
+      "learning_rate": 8e-05,
+      "loss": 1.6714,
+      "step": 3249
+    },
+    {
+      "epoch": 0.18115942028985507,
+      "grad_norm": 0.4664669632911682,
+      "learning_rate": 8e-05,
+      "loss": 1.5275,
+      "step": 3250
+    },
+    {
+      "epoch": 0.18121516164994425,
+      "grad_norm": 0.44695574045181274,
+      "learning_rate": 8e-05,
+      "loss": 1.5913,
+      "step": 3251
+    },
+    {
+      "epoch": 0.18127090301003346,
+      "grad_norm": 0.52314692735672,
+      "learning_rate": 8e-05,
+      "loss": 1.8208,
+      "step": 3252
+    },
+    {
+      "epoch": 0.18132664437012264,
+      "grad_norm": 0.4656141698360443,
+      "learning_rate": 8e-05,
+      "loss": 1.6378,
+      "step": 3253
+    },
+    {
+      "epoch": 0.18138238573021181,
+      "grad_norm": 0.4591492712497711,
+      "learning_rate": 8e-05,
+      "loss": 1.481,
+      "step": 3254
+    },
+    {
+      "epoch": 0.181438127090301,
+      "grad_norm": 0.4969112277030945,
+      "learning_rate": 8e-05,
+      "loss": 1.6094,
+      "step": 3255
+    },
+    {
+      "epoch": 0.1814938684503902,
+      "grad_norm": 0.4783157408237457,
+      "learning_rate": 8e-05,
+      "loss": 1.5105,
+      "step": 3256
+    },
+    {
+      "epoch": 0.18154960981047938,
+      "grad_norm": 0.5113158226013184,
+      "learning_rate": 8e-05,
+      "loss": 1.7022,
+      "step": 3257
+    },
+    {
+      "epoch": 0.18160535117056856,
+      "grad_norm": 0.48434895277023315,
+      "learning_rate": 8e-05,
+      "loss": 1.6306,
+      "step": 3258
+    },
+    {
+      "epoch": 0.18166109253065774,
+      "grad_norm": 0.6748455762863159,
+      "learning_rate": 8e-05,
+      "loss": 1.6323,
+      "step": 3259
+    },
+    {
+      "epoch": 0.18171683389074694,
+      "grad_norm": 0.47873955965042114,
+      "learning_rate": 8e-05,
+      "loss": 1.6777,
+      "step": 3260
+    },
+    {
+      "epoch": 0.18177257525083612,
+      "grad_norm": 0.4789445102214813,
+      "learning_rate": 8e-05,
+      "loss": 1.8353,
+      "step": 3261
+    },
+    {
+      "epoch": 0.1818283166109253,
+      "grad_norm": 0.538105309009552,
+      "learning_rate": 8e-05,
+      "loss": 1.8148,
+      "step": 3262
+    },
+    {
+      "epoch": 0.18188405797101448,
+      "grad_norm": 0.4994378685951233,
+      "learning_rate": 8e-05,
+      "loss": 1.7498,
+      "step": 3263
+    },
+    {
+      "epoch": 0.18193979933110369,
+      "grad_norm": 0.4745805561542511,
+      "learning_rate": 8e-05,
+      "loss": 1.7025,
+      "step": 3264
+    },
+    {
+      "epoch": 0.18199554069119286,
+      "grad_norm": 0.46566084027290344,
+      "learning_rate": 8e-05,
+      "loss": 1.7204,
+      "step": 3265
+    },
+    {
+      "epoch": 0.18205128205128204,
+      "grad_norm": 0.475907564163208,
+      "learning_rate": 8e-05,
+      "loss": 1.6802,
+      "step": 3266
+    },
+    {
+      "epoch": 0.18210702341137125,
+      "grad_norm": 0.475202739238739,
+      "learning_rate": 8e-05,
+      "loss": 1.649,
+      "step": 3267
+    },
+    {
+      "epoch": 0.18216276477146043,
+      "grad_norm": 0.5017547011375427,
+      "learning_rate": 8e-05,
+      "loss": 1.7254,
+      "step": 3268
+    },
+    {
+      "epoch": 0.1822185061315496,
+      "grad_norm": 0.5284735560417175,
+      "learning_rate": 8e-05,
+      "loss": 1.9484,
+      "step": 3269
+    },
+    {
+      "epoch": 0.18227424749163879,
+      "grad_norm": 0.4836098253726959,
+      "learning_rate": 8e-05,
+      "loss": 1.7622,
+      "step": 3270
+    },
+    {
+      "epoch": 0.182329988851728,
+      "grad_norm": 0.5028160810470581,
+      "learning_rate": 8e-05,
+      "loss": 1.8658,
+      "step": 3271
+    },
+    {
+      "epoch": 0.18238573021181717,
+      "grad_norm": 0.5013949275016785,
+      "learning_rate": 8e-05,
+      "loss": 1.6311,
+      "step": 3272
+    },
+    {
+      "epoch": 0.18244147157190635,
+      "grad_norm": 0.5350913405418396,
+      "learning_rate": 8e-05,
+      "loss": 1.9957,
+      "step": 3273
+    },
+    {
+      "epoch": 0.18249721293199553,
+      "grad_norm": 0.49767985939979553,
+      "learning_rate": 8e-05,
+      "loss": 1.6772,
+      "step": 3274
+    },
+    {
+      "epoch": 0.18255295429208473,
+      "grad_norm": 0.5184692740440369,
+      "learning_rate": 8e-05,
+      "loss": 2.0062,
+      "step": 3275
+    },
+    {
+      "epoch": 0.1826086956521739,
+      "grad_norm": 0.4783124029636383,
+      "learning_rate": 8e-05,
+      "loss": 1.7281,
+      "step": 3276
+    },
+    {
+      "epoch": 0.1826644370122631,
+      "grad_norm": 0.4819895923137665,
+      "learning_rate": 8e-05,
+      "loss": 1.5692,
+      "step": 3277
+    },
+    {
+      "epoch": 0.1827201783723523,
+      "grad_norm": 0.48155200481414795,
+      "learning_rate": 8e-05,
+      "loss": 1.6602,
+      "step": 3278
+    },
+    {
+      "epoch": 0.18277591973244148,
+      "grad_norm": 0.492043673992157,
+      "learning_rate": 8e-05,
+      "loss": 1.7576,
+      "step": 3279
+    },
+    {
+      "epoch": 0.18283166109253066,
+      "grad_norm": 0.4913448095321655,
+      "learning_rate": 8e-05,
+      "loss": 1.8854,
+      "step": 3280
+    },
+    {
+      "epoch": 0.18288740245261983,
+      "grad_norm": 0.4953262507915497,
+      "learning_rate": 8e-05,
+      "loss": 1.7527,
+      "step": 3281
+    },
+    {
+      "epoch": 0.18294314381270904,
+      "grad_norm": 0.482516884803772,
+      "learning_rate": 8e-05,
+      "loss": 1.8599,
+      "step": 3282
+    },
+    {
+      "epoch": 0.18299888517279822,
+      "grad_norm": 0.4654371440410614,
+      "learning_rate": 8e-05,
+      "loss": 1.8008,
+      "step": 3283
+    },
+    {
+      "epoch": 0.1830546265328874,
+      "grad_norm": 0.44700589776039124,
+      "learning_rate": 8e-05,
+      "loss": 1.7483,
+      "step": 3284
+    },
+    {
+      "epoch": 0.18311036789297658,
+      "grad_norm": 0.49507349729537964,
+      "learning_rate": 8e-05,
+      "loss": 1.9131,
+      "step": 3285
+    },
+    {
+      "epoch": 0.18316610925306578,
+      "grad_norm": 0.46997302770614624,
+      "learning_rate": 8e-05,
+      "loss": 1.6543,
+      "step": 3286
+    },
+    {
+      "epoch": 0.18322185061315496,
+      "grad_norm": 0.5000645518302917,
+      "learning_rate": 8e-05,
+      "loss": 1.868,
+      "step": 3287
+    },
+    {
+      "epoch": 0.18327759197324414,
+      "grad_norm": 0.5379271507263184,
+      "learning_rate": 8e-05,
+      "loss": 1.9267,
+      "step": 3288
+    },
+    {
+      "epoch": 0.18333333333333332,
+      "grad_norm": 0.4435328543186188,
+      "learning_rate": 8e-05,
+      "loss": 1.5935,
+      "step": 3289
+    },
+    {
+      "epoch": 0.18338907469342253,
+      "grad_norm": 0.4761490523815155,
+      "learning_rate": 8e-05,
+      "loss": 1.828,
+      "step": 3290
+    },
+    {
+      "epoch": 0.1834448160535117,
+      "grad_norm": 0.5158578753471375,
+      "learning_rate": 8e-05,
+      "loss": 1.6859,
+      "step": 3291
+    },
+    {
+      "epoch": 0.18350055741360088,
+      "grad_norm": 0.45683667063713074,
+      "learning_rate": 8e-05,
+      "loss": 1.5538,
+      "step": 3292
+    },
+    {
+      "epoch": 0.1835562987736901,
+      "grad_norm": 0.460464209318161,
+      "learning_rate": 8e-05,
+      "loss": 1.775,
+      "step": 3293
+    },
+    {
+      "epoch": 0.18361204013377927,
+      "grad_norm": 0.4554464817047119,
+      "learning_rate": 8e-05,
+      "loss": 1.723,
+      "step": 3294
+    },
+    {
+      "epoch": 0.18366778149386845,
+      "grad_norm": 0.49409928917884827,
+      "learning_rate": 8e-05,
+      "loss": 1.8673,
+      "step": 3295
+    },
+    {
+      "epoch": 0.18372352285395763,
+      "grad_norm": 0.5223574638366699,
+      "learning_rate": 8e-05,
+      "loss": 1.7798,
+      "step": 3296
+    },
+    {
+      "epoch": 0.18377926421404683,
+      "grad_norm": 0.4630155563354492,
+      "learning_rate": 8e-05,
+      "loss": 1.6178,
+      "step": 3297
+    },
+    {
+      "epoch": 0.183835005574136,
+      "grad_norm": 0.5209290981292725,
+      "learning_rate": 8e-05,
+      "loss": 1.9866,
+      "step": 3298
+    },
+    {
+      "epoch": 0.1838907469342252,
+      "grad_norm": 0.48574021458625793,
+      "learning_rate": 8e-05,
+      "loss": 1.7712,
+      "step": 3299
+    },
+    {
+      "epoch": 0.18394648829431437,
+      "grad_norm": 0.5119464993476868,
+      "learning_rate": 8e-05,
+      "loss": 1.9065,
+      "step": 3300
+    },
+    {
+      "epoch": 0.18400222965440358,
+      "grad_norm": 0.4706062078475952,
+      "learning_rate": 8e-05,
+      "loss": 1.7841,
+      "step": 3301
+    },
+    {
+      "epoch": 0.18405797101449275,
+      "grad_norm": 0.47743088006973267,
+      "learning_rate": 8e-05,
+      "loss": 1.722,
+      "step": 3302
+    },
+    {
+      "epoch": 0.18411371237458193,
+      "grad_norm": 0.4631895422935486,
+      "learning_rate": 8e-05,
+      "loss": 1.7439,
+      "step": 3303
+    },
+    {
+      "epoch": 0.1841694537346711,
+      "grad_norm": 0.4625717103481293,
+      "learning_rate": 8e-05,
+      "loss": 1.62,
+      "step": 3304
+    },
+    {
+      "epoch": 0.18422519509476032,
+      "grad_norm": 0.5006458759307861,
+      "learning_rate": 8e-05,
+      "loss": 1.9197,
+      "step": 3305
+    },
+    {
+      "epoch": 0.1842809364548495,
+      "grad_norm": 0.5063039064407349,
+      "learning_rate": 8e-05,
+      "loss": 2.0728,
+      "step": 3306
+    },
+    {
+      "epoch": 0.18433667781493868,
+      "grad_norm": 0.4803706109523773,
+      "learning_rate": 8e-05,
+      "loss": 1.7504,
+      "step": 3307
+    },
+    {
+      "epoch": 0.18439241917502788,
+      "grad_norm": 0.47429147362709045,
+      "learning_rate": 8e-05,
+      "loss": 1.8056,
+      "step": 3308
+    },
+    {
+      "epoch": 0.18444816053511706,
+      "grad_norm": 0.4567185342311859,
+      "learning_rate": 8e-05,
+      "loss": 1.721,
+      "step": 3309
+    },
+    {
+      "epoch": 0.18450390189520624,
+      "grad_norm": 0.4384846091270447,
+      "learning_rate": 8e-05,
+      "loss": 1.7278,
+      "step": 3310
+    },
+    {
+      "epoch": 0.18455964325529542,
+      "grad_norm": 0.4725017845630646,
+      "learning_rate": 8e-05,
+      "loss": 1.5995,
+      "step": 3311
+    },
+    {
+      "epoch": 0.18461538461538463,
+      "grad_norm": 0.46991947293281555,
+      "learning_rate": 8e-05,
+      "loss": 1.3783,
+      "step": 3312
+    },
+    {
+      "epoch": 0.1846711259754738,
+      "grad_norm": 0.5544723868370056,
+      "learning_rate": 8e-05,
+      "loss": 1.9191,
+      "step": 3313
+    },
+    {
+      "epoch": 0.18472686733556298,
+      "grad_norm": 0.46554556488990784,
+      "learning_rate": 8e-05,
+      "loss": 1.7393,
+      "step": 3314
+    },
+    {
+      "epoch": 0.18478260869565216,
+      "grad_norm": 0.4672147035598755,
+      "learning_rate": 8e-05,
+      "loss": 1.7015,
+      "step": 3315
+    },
+    {
+      "epoch": 0.18483835005574137,
+      "grad_norm": 0.47809523344039917,
+      "learning_rate": 8e-05,
+      "loss": 1.7469,
+      "step": 3316
+    },
+    {
+      "epoch": 0.18489409141583055,
+      "grad_norm": 0.4963299334049225,
+      "learning_rate": 8e-05,
+      "loss": 1.9078,
+      "step": 3317
+    },
+    {
+      "epoch": 0.18494983277591973,
+      "grad_norm": 0.48992517590522766,
+      "learning_rate": 8e-05,
+      "loss": 1.5713,
+      "step": 3318
+    },
+    {
+      "epoch": 0.18500557413600893,
+      "grad_norm": 0.49141886830329895,
+      "learning_rate": 8e-05,
+      "loss": 1.847,
+      "step": 3319
+    },
+    {
+      "epoch": 0.1850613154960981,
+      "grad_norm": 0.4902811646461487,
+      "learning_rate": 8e-05,
+      "loss": 1.675,
+      "step": 3320
+    },
+    {
+      "epoch": 0.1851170568561873,
+      "grad_norm": 0.5069705843925476,
+      "learning_rate": 8e-05,
+      "loss": 1.9477,
+      "step": 3321
+    },
+    {
+      "epoch": 0.18517279821627647,
+      "grad_norm": 0.5962790846824646,
+      "learning_rate": 8e-05,
+      "loss": 1.8071,
+      "step": 3322
+    },
+    {
+      "epoch": 0.18522853957636568,
+      "grad_norm": 0.4733704626560211,
+      "learning_rate": 8e-05,
+      "loss": 1.4901,
+      "step": 3323
+    },
+    {
+      "epoch": 0.18528428093645485,
+      "grad_norm": 0.514161229133606,
+      "learning_rate": 8e-05,
+      "loss": 1.7563,
+      "step": 3324
+    },
+    {
+      "epoch": 0.18534002229654403,
+      "grad_norm": 0.5042579770088196,
+      "learning_rate": 8e-05,
+      "loss": 1.7477,
+      "step": 3325
+    },
+    {
+      "epoch": 0.1853957636566332,
+      "grad_norm": 0.503969132900238,
+      "learning_rate": 8e-05,
+      "loss": 1.7513,
+      "step": 3326
+    },
+    {
+      "epoch": 0.18545150501672242,
+      "grad_norm": 0.503426730632782,
+      "learning_rate": 8e-05,
+      "loss": 1.6288,
+      "step": 3327
+    },
+    {
+      "epoch": 0.1855072463768116,
+      "grad_norm": 0.48891016840934753,
+      "learning_rate": 8e-05,
+      "loss": 1.7253,
+      "step": 3328
+    },
+    {
+      "epoch": 0.18556298773690078,
+      "grad_norm": 0.5385537147521973,
+      "learning_rate": 8e-05,
+      "loss": 1.6348,
+      "step": 3329
+    },
+    {
+      "epoch": 0.18561872909698995,
+      "grad_norm": 0.4645160138607025,
+      "learning_rate": 8e-05,
+      "loss": 1.7895,
+      "step": 3330
+    },
+    {
+      "epoch": 0.18567447045707916,
+      "grad_norm": 0.49372631311416626,
+      "learning_rate": 8e-05,
+      "loss": 1.5114,
+      "step": 3331
+    },
+    {
+      "epoch": 0.18573021181716834,
+      "grad_norm": 0.46734169125556946,
+      "learning_rate": 8e-05,
+      "loss": 1.5832,
+      "step": 3332
+    },
+    {
+      "epoch": 0.18578595317725752,
+      "grad_norm": 0.5161205530166626,
+      "learning_rate": 8e-05,
+      "loss": 1.863,
+      "step": 3333
+    },
+    {
+      "epoch": 0.18584169453734672,
+      "grad_norm": 0.47789087891578674,
+      "learning_rate": 8e-05,
+      "loss": 1.5687,
+      "step": 3334
+    },
+    {
+      "epoch": 0.1858974358974359,
+      "grad_norm": 0.5009278655052185,
+      "learning_rate": 8e-05,
+      "loss": 1.891,
+      "step": 3335
+    },
+    {
+      "epoch": 0.18595317725752508,
+      "grad_norm": 0.44556164741516113,
+      "learning_rate": 8e-05,
+      "loss": 1.7084,
+      "step": 3336
+    },
+    {
+      "epoch": 0.18600891861761426,
+      "grad_norm": 0.45805823802948,
+      "learning_rate": 8e-05,
+      "loss": 1.5131,
+      "step": 3337
+    },
+    {
+      "epoch": 0.18606465997770347,
+      "grad_norm": 0.48302045464515686,
+      "learning_rate": 8e-05,
+      "loss": 1.7633,
+      "step": 3338
+    },
+    {
+      "epoch": 0.18612040133779265,
+      "grad_norm": 0.4615001380443573,
+      "learning_rate": 8e-05,
+      "loss": 1.4623,
+      "step": 3339
+    },
+    {
+      "epoch": 0.18617614269788182,
+      "grad_norm": 0.49533554911613464,
+      "learning_rate": 8e-05,
+      "loss": 1.6608,
+      "step": 3340
+    },
+    {
+      "epoch": 0.186231884057971,
+      "grad_norm": 0.43167346715927124,
+      "learning_rate": 8e-05,
+      "loss": 1.4441,
+      "step": 3341
+    },
+    {
+      "epoch": 0.1862876254180602,
+      "grad_norm": 0.527940571308136,
+      "learning_rate": 8e-05,
+      "loss": 1.5926,
+      "step": 3342
+    },
+    {
+      "epoch": 0.1863433667781494,
+      "grad_norm": 0.5032770037651062,
+      "learning_rate": 8e-05,
+      "loss": 1.6924,
+      "step": 3343
+    },
+    {
+      "epoch": 0.18639910813823857,
+      "grad_norm": 0.5254105925559998,
+      "learning_rate": 8e-05,
+      "loss": 1.9019,
+      "step": 3344
+    },
+    {
+      "epoch": 0.18645484949832775,
+      "grad_norm": 0.5116928815841675,
+      "learning_rate": 8e-05,
+      "loss": 1.8028,
+      "step": 3345
+    },
+    {
+      "epoch": 0.18651059085841695,
+      "grad_norm": 0.45237305760383606,
+      "learning_rate": 8e-05,
+      "loss": 1.5661,
+      "step": 3346
+    },
+    {
+      "epoch": 0.18656633221850613,
+      "grad_norm": 0.4751761853694916,
+      "learning_rate": 8e-05,
+      "loss": 1.6603,
+      "step": 3347
+    },
+    {
+      "epoch": 0.1866220735785953,
+      "grad_norm": 0.5808463096618652,
+      "learning_rate": 8e-05,
+      "loss": 1.9472,
+      "step": 3348
+    },
+    {
+      "epoch": 0.18667781493868452,
+      "grad_norm": 0.48247790336608887,
+      "learning_rate": 8e-05,
+      "loss": 1.6478,
+      "step": 3349
+    },
+    {
+      "epoch": 0.1867335562987737,
+      "grad_norm": 0.49884429574012756,
+      "learning_rate": 8e-05,
+      "loss": 1.9385,
+      "step": 3350
+    },
+    {
+      "epoch": 0.18678929765886287,
+      "grad_norm": 0.49466973543167114,
+      "learning_rate": 8e-05,
+      "loss": 1.7743,
+      "step": 3351
+    },
+    {
+      "epoch": 0.18684503901895205,
+      "grad_norm": 0.47581541538238525,
+      "learning_rate": 8e-05,
+      "loss": 1.552,
+      "step": 3352
+    },
+    {
+      "epoch": 0.18690078037904126,
+      "grad_norm": 0.448135107755661,
+      "learning_rate": 8e-05,
+      "loss": 1.5884,
+      "step": 3353
+    },
+    {
+      "epoch": 0.18695652173913044,
+      "grad_norm": 0.560992956161499,
+      "learning_rate": 8e-05,
+      "loss": 1.9356,
+      "step": 3354
+    },
+    {
+      "epoch": 0.18701226309921962,
+      "grad_norm": 0.5413427352905273,
+      "learning_rate": 8e-05,
+      "loss": 1.7125,
+      "step": 3355
+    },
+    {
+      "epoch": 0.1870680044593088,
+      "grad_norm": 0.48547688126564026,
+      "learning_rate": 8e-05,
+      "loss": 1.7948,
+      "step": 3356
+    },
+    {
+      "epoch": 0.187123745819398,
+      "grad_norm": 0.4625580906867981,
+      "learning_rate": 8e-05,
+      "loss": 1.4295,
+      "step": 3357
+    },
+    {
+      "epoch": 0.18717948717948718,
+      "grad_norm": 0.49883168935775757,
+      "learning_rate": 8e-05,
+      "loss": 1.6727,
+      "step": 3358
+    },
+    {
+      "epoch": 0.18723522853957636,
+      "grad_norm": 0.4742003381252289,
+      "learning_rate": 8e-05,
+      "loss": 1.52,
+      "step": 3359
+    },
+    {
+      "epoch": 0.18729096989966554,
+      "grad_norm": 0.5026939511299133,
+      "learning_rate": 8e-05,
+      "loss": 1.6116,
+      "step": 3360
+    },
+    {
+      "epoch": 0.18734671125975474,
+      "grad_norm": 0.48718181252479553,
+      "learning_rate": 8e-05,
+      "loss": 1.7806,
+      "step": 3361
+    },
+    {
+      "epoch": 0.18740245261984392,
+      "grad_norm": 0.5140118598937988,
+      "learning_rate": 8e-05,
+      "loss": 1.7943,
+      "step": 3362
+    },
+    {
+      "epoch": 0.1874581939799331,
+      "grad_norm": 0.506673276424408,
+      "learning_rate": 8e-05,
+      "loss": 1.7472,
+      "step": 3363
+    },
+    {
+      "epoch": 0.1875139353400223,
+      "grad_norm": 0.5099788904190063,
+      "learning_rate": 8e-05,
+      "loss": 1.7816,
+      "step": 3364
+    },
+    {
+      "epoch": 0.1875696767001115,
+      "grad_norm": 0.5400450229644775,
+      "learning_rate": 8e-05,
+      "loss": 1.7322,
+      "step": 3365
+    },
+    {
+      "epoch": 0.18762541806020067,
+      "grad_norm": 0.48514536023139954,
+      "learning_rate": 8e-05,
+      "loss": 1.7965,
+      "step": 3366
+    },
+    {
+      "epoch": 0.18768115942028984,
+      "grad_norm": 0.47433406114578247,
+      "learning_rate": 8e-05,
+      "loss": 1.6121,
+      "step": 3367
+    },
+    {
+      "epoch": 0.18773690078037905,
+      "grad_norm": 0.46342793107032776,
+      "learning_rate": 8e-05,
+      "loss": 1.7103,
+      "step": 3368
+    },
+    {
+      "epoch": 0.18779264214046823,
+      "grad_norm": 0.44723591208457947,
+      "learning_rate": 8e-05,
+      "loss": 1.4898,
+      "step": 3369
+    },
+    {
+      "epoch": 0.1878483835005574,
+      "grad_norm": 0.4659675657749176,
+      "learning_rate": 8e-05,
+      "loss": 1.5409,
+      "step": 3370
+    },
+    {
+      "epoch": 0.1879041248606466,
+      "grad_norm": 0.5222867131233215,
+      "learning_rate": 8e-05,
+      "loss": 1.7159,
+      "step": 3371
+    },
+    {
+      "epoch": 0.1879598662207358,
+      "grad_norm": 0.49539661407470703,
+      "learning_rate": 8e-05,
+      "loss": 1.8474,
+      "step": 3372
+    },
+    {
+      "epoch": 0.18801560758082497,
+      "grad_norm": 0.5112997889518738,
+      "learning_rate": 8e-05,
+      "loss": 1.9201,
+      "step": 3373
+    },
+    {
+      "epoch": 0.18807134894091415,
+      "grad_norm": 0.44712772965431213,
+      "learning_rate": 8e-05,
+      "loss": 1.4429,
+      "step": 3374
+    },
+    {
+      "epoch": 0.18812709030100336,
+      "grad_norm": 0.48310035467147827,
+      "learning_rate": 8e-05,
+      "loss": 1.689,
+      "step": 3375
+    },
+    {
+      "epoch": 0.18818283166109254,
+      "grad_norm": 0.4607887864112854,
+      "learning_rate": 8e-05,
+      "loss": 1.3945,
+      "step": 3376
+    },
+    {
+      "epoch": 0.18823857302118172,
+      "grad_norm": 0.5434450507164001,
+      "learning_rate": 8e-05,
+      "loss": 1.9745,
+      "step": 3377
+    },
+    {
+      "epoch": 0.1882943143812709,
+      "grad_norm": 0.48917487263679504,
+      "learning_rate": 8e-05,
+      "loss": 1.6444,
+      "step": 3378
+    },
+    {
+      "epoch": 0.1883500557413601,
+      "grad_norm": 0.5110101103782654,
+      "learning_rate": 8e-05,
+      "loss": 2.0023,
+      "step": 3379
+    },
+    {
+      "epoch": 0.18840579710144928,
+      "grad_norm": 0.489247590303421,
+      "learning_rate": 8e-05,
+      "loss": 1.7046,
+      "step": 3380
+    },
+    {
+      "epoch": 0.18846153846153846,
+      "grad_norm": 0.5191060304641724,
+      "learning_rate": 8e-05,
+      "loss": 1.8036,
+      "step": 3381
+    },
+    {
+      "epoch": 0.18851727982162764,
+      "grad_norm": 0.5540537238121033,
+      "learning_rate": 8e-05,
+      "loss": 1.7418,
+      "step": 3382
+    },
+    {
+      "epoch": 0.18857302118171684,
+      "grad_norm": 0.575836718082428,
+      "learning_rate": 8e-05,
+      "loss": 2.0404,
+      "step": 3383
+    },
+    {
+      "epoch": 0.18862876254180602,
+      "grad_norm": 0.5042715072631836,
+      "learning_rate": 8e-05,
+      "loss": 1.6793,
+      "step": 3384
+    },
+    {
+      "epoch": 0.1886845039018952,
+      "grad_norm": 0.45747798681259155,
+      "learning_rate": 8e-05,
+      "loss": 1.6696,
+      "step": 3385
+    },
+    {
+      "epoch": 0.18874024526198438,
+      "grad_norm": 0.4825889468193054,
+      "learning_rate": 8e-05,
+      "loss": 1.8729,
+      "step": 3386
+    },
+    {
+      "epoch": 0.18879598662207359,
+      "grad_norm": 0.5231027603149414,
+      "learning_rate": 8e-05,
+      "loss": 1.7968,
+      "step": 3387
+    },
+    {
+      "epoch": 0.18885172798216276,
+      "grad_norm": 0.4612344205379486,
+      "learning_rate": 8e-05,
+      "loss": 1.6421,
+      "step": 3388
+    },
+    {
+      "epoch": 0.18890746934225194,
+      "grad_norm": 0.44895699620246887,
+      "learning_rate": 8e-05,
+      "loss": 1.6622,
+      "step": 3389
+    },
+    {
+      "epoch": 0.18896321070234115,
+      "grad_norm": 0.44689661264419556,
+      "learning_rate": 8e-05,
+      "loss": 1.3054,
+      "step": 3390
+    },
+    {
+      "epoch": 0.18901895206243033,
+      "grad_norm": 0.5180425047874451,
+      "learning_rate": 8e-05,
+      "loss": 2.0898,
+      "step": 3391
+    },
+    {
+      "epoch": 0.1890746934225195,
+      "grad_norm": 0.44576889276504517,
+      "learning_rate": 8e-05,
+      "loss": 1.6609,
+      "step": 3392
+    },
+    {
+      "epoch": 0.1891304347826087,
+      "grad_norm": 0.47959277033805847,
+      "learning_rate": 8e-05,
+      "loss": 1.7771,
+      "step": 3393
+    },
+    {
+      "epoch": 0.1891861761426979,
+      "grad_norm": 0.4864841401576996,
+      "learning_rate": 8e-05,
+      "loss": 1.8195,
+      "step": 3394
+    },
+    {
+      "epoch": 0.18924191750278707,
+      "grad_norm": 0.4945772588253021,
+      "learning_rate": 8e-05,
+      "loss": 1.7601,
+      "step": 3395
+    },
+    {
+      "epoch": 0.18929765886287625,
+      "grad_norm": 0.4611192047595978,
+      "learning_rate": 8e-05,
+      "loss": 1.6278,
+      "step": 3396
+    },
+    {
+      "epoch": 0.18935340022296543,
+      "grad_norm": 0.5064223408699036,
+      "learning_rate": 8e-05,
+      "loss": 1.9821,
+      "step": 3397
+    },
+    {
+      "epoch": 0.18940914158305464,
+      "grad_norm": 0.49804046750068665,
+      "learning_rate": 8e-05,
+      "loss": 1.8064,
+      "step": 3398
+    },
+    {
+      "epoch": 0.18946488294314381,
+      "grad_norm": 0.4759809076786041,
+      "learning_rate": 8e-05,
+      "loss": 1.8248,
+      "step": 3399
+    },
+    {
+      "epoch": 0.189520624303233,
+      "grad_norm": 0.462550550699234,
+      "learning_rate": 8e-05,
+      "loss": 1.6176,
+      "step": 3400
+    },
+    {
+      "epoch": 0.18957636566332217,
+      "grad_norm": 0.4693813920021057,
+      "learning_rate": 8e-05,
+      "loss": 1.6927,
+      "step": 3401
+    },
+    {
+      "epoch": 0.18963210702341138,
+      "grad_norm": 0.48910364508628845,
+      "learning_rate": 8e-05,
+      "loss": 1.6771,
+      "step": 3402
+    },
+    {
+      "epoch": 0.18968784838350056,
+      "grad_norm": 0.4354751706123352,
+      "learning_rate": 8e-05,
+      "loss": 1.5138,
+      "step": 3403
+    },
+    {
+      "epoch": 0.18974358974358974,
+      "grad_norm": 0.4465082883834839,
+      "learning_rate": 8e-05,
+      "loss": 1.6123,
+      "step": 3404
+    },
+    {
+      "epoch": 0.18979933110367894,
+      "grad_norm": 0.5002762079238892,
+      "learning_rate": 8e-05,
+      "loss": 1.8652,
+      "step": 3405
+    },
+    {
+      "epoch": 0.18985507246376812,
+      "grad_norm": 0.46198657155036926,
+      "learning_rate": 8e-05,
+      "loss": 1.4936,
+      "step": 3406
+    },
+    {
+      "epoch": 0.1899108138238573,
+      "grad_norm": 0.5062534213066101,
+      "learning_rate": 8e-05,
+      "loss": 1.8373,
+      "step": 3407
+    },
+    {
+      "epoch": 0.18996655518394648,
+      "grad_norm": 0.4888454079627991,
+      "learning_rate": 8e-05,
+      "loss": 1.6433,
+      "step": 3408
+    },
+    {
+      "epoch": 0.19002229654403568,
+      "grad_norm": 0.5151571035385132,
+      "learning_rate": 8e-05,
+      "loss": 1.7548,
+      "step": 3409
+    },
+    {
+      "epoch": 0.19007803790412486,
+      "grad_norm": 0.5547520518302917,
+      "learning_rate": 8e-05,
+      "loss": 2.0664,
+      "step": 3410
+    },
+    {
+      "epoch": 0.19013377926421404,
+      "grad_norm": 0.4662136733531952,
+      "learning_rate": 8e-05,
+      "loss": 1.6069,
+      "step": 3411
+    },
+    {
+      "epoch": 0.19018952062430322,
+      "grad_norm": 0.48590052127838135,
+      "learning_rate": 8e-05,
+      "loss": 1.776,
+      "step": 3412
+    },
+    {
+      "epoch": 0.19024526198439243,
+      "grad_norm": 0.4569084048271179,
+      "learning_rate": 8e-05,
+      "loss": 1.6613,
+      "step": 3413
+    },
+    {
+      "epoch": 0.1903010033444816,
+      "grad_norm": 0.5148277282714844,
+      "learning_rate": 8e-05,
+      "loss": 1.5761,
+      "step": 3414
+    },
+    {
+      "epoch": 0.19035674470457078,
+      "grad_norm": 0.4618729054927826,
+      "learning_rate": 8e-05,
+      "loss": 1.7345,
+      "step": 3415
+    },
+    {
+      "epoch": 0.19041248606466,
+      "grad_norm": 0.44613951444625854,
+      "learning_rate": 8e-05,
+      "loss": 1.6473,
+      "step": 3416
+    },
+    {
+      "epoch": 0.19046822742474917,
+      "grad_norm": 0.49018537998199463,
+      "learning_rate": 8e-05,
+      "loss": 1.8905,
+      "step": 3417
+    },
+    {
+      "epoch": 0.19052396878483835,
+      "grad_norm": 0.47118470072746277,
+      "learning_rate": 8e-05,
+      "loss": 1.7407,
+      "step": 3418
+    },
+    {
+      "epoch": 0.19057971014492753,
+      "grad_norm": 0.46336379647254944,
+      "learning_rate": 8e-05,
+      "loss": 1.5438,
+      "step": 3419
+    },
+    {
+      "epoch": 0.19063545150501673,
+      "grad_norm": 0.49951037764549255,
+      "learning_rate": 8e-05,
+      "loss": 1.7482,
+      "step": 3420
+    },
+    {
+      "epoch": 0.1906911928651059,
+      "grad_norm": 0.5517299771308899,
+      "learning_rate": 8e-05,
+      "loss": 2.0162,
+      "step": 3421
+    },
+    {
+      "epoch": 0.1907469342251951,
+      "grad_norm": 0.4926646053791046,
+      "learning_rate": 8e-05,
+      "loss": 1.6469,
+      "step": 3422
+    },
+    {
+      "epoch": 0.19080267558528427,
+      "grad_norm": 0.524966299533844,
+      "learning_rate": 8e-05,
+      "loss": 1.6917,
+      "step": 3423
+    },
+    {
+      "epoch": 0.19085841694537348,
+      "grad_norm": 0.469455361366272,
+      "learning_rate": 8e-05,
+      "loss": 1.6371,
+      "step": 3424
+    },
+    {
+      "epoch": 0.19091415830546266,
+      "grad_norm": 0.4847599267959595,
+      "learning_rate": 8e-05,
+      "loss": 1.6893,
+      "step": 3425
+    },
+    {
+      "epoch": 0.19096989966555183,
+      "grad_norm": 0.5023056268692017,
+      "learning_rate": 8e-05,
+      "loss": 1.7794,
+      "step": 3426
+    },
+    {
+      "epoch": 0.191025641025641,
+      "grad_norm": 0.48208606243133545,
+      "learning_rate": 8e-05,
+      "loss": 1.8087,
+      "step": 3427
+    },
+    {
+      "epoch": 0.19108138238573022,
+      "grad_norm": 0.4761766493320465,
+      "learning_rate": 8e-05,
+      "loss": 1.7857,
+      "step": 3428
+    },
+    {
+      "epoch": 0.1911371237458194,
+      "grad_norm": 0.43917298316955566,
+      "learning_rate": 8e-05,
+      "loss": 1.4967,
+      "step": 3429
+    },
+    {
+      "epoch": 0.19119286510590858,
+      "grad_norm": 0.5421159267425537,
+      "learning_rate": 8e-05,
+      "loss": 1.8842,
+      "step": 3430
+    },
+    {
+      "epoch": 0.19124860646599778,
+      "grad_norm": 0.4713871479034424,
+      "learning_rate": 8e-05,
+      "loss": 1.6097,
+      "step": 3431
+    },
+    {
+      "epoch": 0.19130434782608696,
+      "grad_norm": 0.5096777081489563,
+      "learning_rate": 8e-05,
+      "loss": 1.7582,
+      "step": 3432
+    },
+    {
+      "epoch": 0.19136008918617614,
+      "grad_norm": 0.5059357285499573,
+      "learning_rate": 8e-05,
+      "loss": 1.8719,
+      "step": 3433
+    },
+    {
+      "epoch": 0.19141583054626532,
+      "grad_norm": 0.4401608109474182,
+      "learning_rate": 8e-05,
+      "loss": 1.6274,
+      "step": 3434
+    },
+    {
+      "epoch": 0.19147157190635453,
+      "grad_norm": 0.4493738114833832,
+      "learning_rate": 8e-05,
+      "loss": 1.6411,
+      "step": 3435
+    },
+    {
+      "epoch": 0.1915273132664437,
+      "grad_norm": 0.47014740109443665,
+      "learning_rate": 8e-05,
+      "loss": 1.7378,
+      "step": 3436
+    },
+    {
+      "epoch": 0.19158305462653288,
+      "grad_norm": 0.4577510952949524,
+      "learning_rate": 8e-05,
+      "loss": 1.6463,
+      "step": 3437
+    },
+    {
+      "epoch": 0.19163879598662206,
+      "grad_norm": 0.44429123401641846,
+      "learning_rate": 8e-05,
+      "loss": 1.6286,
+      "step": 3438
+    },
+    {
+      "epoch": 0.19169453734671127,
+      "grad_norm": 0.557929515838623,
+      "learning_rate": 8e-05,
+      "loss": 1.4677,
+      "step": 3439
+    },
+    {
+      "epoch": 0.19175027870680045,
+      "grad_norm": 0.5293607711791992,
+      "learning_rate": 8e-05,
+      "loss": 1.7674,
+      "step": 3440
+    },
+    {
+      "epoch": 0.19180602006688963,
+      "grad_norm": 0.5156054496765137,
+      "learning_rate": 8e-05,
+      "loss": 1.9326,
+      "step": 3441
+    },
+    {
+      "epoch": 0.1918617614269788,
+      "grad_norm": 0.576087474822998,
+      "learning_rate": 8e-05,
+      "loss": 1.8974,
+      "step": 3442
+    },
+    {
+      "epoch": 0.191917502787068,
+      "grad_norm": 0.435739129781723,
+      "learning_rate": 8e-05,
+      "loss": 1.5585,
+      "step": 3443
+    },
+    {
+      "epoch": 0.1919732441471572,
+      "grad_norm": 0.4967978894710541,
+      "learning_rate": 8e-05,
+      "loss": 2.0174,
+      "step": 3444
+    },
+    {
+      "epoch": 0.19202898550724637,
+      "grad_norm": 0.4740888178348541,
+      "learning_rate": 8e-05,
+      "loss": 1.7055,
+      "step": 3445
+    },
+    {
+      "epoch": 0.19208472686733558,
+      "grad_norm": 0.5190867185592651,
+      "learning_rate": 8e-05,
+      "loss": 1.8603,
+      "step": 3446
+    },
+    {
+      "epoch": 0.19214046822742475,
+      "grad_norm": 0.5093804001808167,
+      "learning_rate": 8e-05,
+      "loss": 1.692,
+      "step": 3447
+    },
+    {
+      "epoch": 0.19219620958751393,
+      "grad_norm": 0.6346522569656372,
+      "learning_rate": 8e-05,
+      "loss": 1.709,
+      "step": 3448
+    },
+    {
+      "epoch": 0.1922519509476031,
+      "grad_norm": 0.5031291842460632,
+      "learning_rate": 8e-05,
+      "loss": 1.7797,
+      "step": 3449
+    },
+    {
+      "epoch": 0.19230769230769232,
+      "grad_norm": 0.5159427523612976,
+      "learning_rate": 8e-05,
+      "loss": 1.6472,
+      "step": 3450
+    },
+    {
+      "epoch": 0.1923634336677815,
+      "grad_norm": 0.5027660727500916,
+      "learning_rate": 8e-05,
+      "loss": 1.7745,
+      "step": 3451
+    },
+    {
+      "epoch": 0.19241917502787068,
+      "grad_norm": 0.5181394219398499,
+      "learning_rate": 8e-05,
+      "loss": 1.8142,
+      "step": 3452
+    },
+    {
+      "epoch": 0.19247491638795985,
+      "grad_norm": 0.47508713603019714,
+      "learning_rate": 8e-05,
+      "loss": 1.723,
+      "step": 3453
+    },
+    {
+      "epoch": 0.19253065774804906,
+      "grad_norm": 0.49145597219467163,
+      "learning_rate": 8e-05,
+      "loss": 1.8007,
+      "step": 3454
+    },
+    {
+      "epoch": 0.19258639910813824,
+      "grad_norm": 0.4852198660373688,
+      "learning_rate": 8e-05,
+      "loss": 1.7551,
+      "step": 3455
+    },
+    {
+      "epoch": 0.19264214046822742,
+      "grad_norm": 0.502672553062439,
+      "learning_rate": 8e-05,
+      "loss": 1.81,
+      "step": 3456
+    },
+    {
+      "epoch": 0.1926978818283166,
+      "grad_norm": 0.47326627373695374,
+      "learning_rate": 8e-05,
+      "loss": 1.6708,
+      "step": 3457
+    },
+    {
+      "epoch": 0.1927536231884058,
+      "grad_norm": 0.4441857933998108,
+      "learning_rate": 8e-05,
+      "loss": 1.5861,
+      "step": 3458
+    },
+    {
+      "epoch": 0.19280936454849498,
+      "grad_norm": 0.4696533977985382,
+      "learning_rate": 8e-05,
+      "loss": 1.6572,
+      "step": 3459
+    },
+    {
+      "epoch": 0.19286510590858416,
+      "grad_norm": 0.444561243057251,
+      "learning_rate": 8e-05,
+      "loss": 1.5593,
+      "step": 3460
+    },
+    {
+      "epoch": 0.19292084726867337,
+      "grad_norm": 0.44416651129722595,
+      "learning_rate": 8e-05,
+      "loss": 1.4879,
+      "step": 3461
+    },
+    {
+      "epoch": 0.19297658862876255,
+      "grad_norm": 0.5351315140724182,
+      "learning_rate": 8e-05,
+      "loss": 1.8234,
+      "step": 3462
+    },
+    {
+      "epoch": 0.19303232998885173,
+      "grad_norm": 0.5529216527938843,
+      "learning_rate": 8e-05,
+      "loss": 1.623,
+      "step": 3463
+    },
+    {
+      "epoch": 0.1930880713489409,
+      "grad_norm": 0.4560251235961914,
+      "learning_rate": 8e-05,
+      "loss": 1.5238,
+      "step": 3464
+    },
+    {
+      "epoch": 0.1931438127090301,
+      "grad_norm": 0.4910484254360199,
+      "learning_rate": 8e-05,
+      "loss": 1.8426,
+      "step": 3465
+    },
+    {
+      "epoch": 0.1931995540691193,
+      "grad_norm": 0.5156372785568237,
+      "learning_rate": 8e-05,
+      "loss": 1.8463,
+      "step": 3466
+    },
+    {
+      "epoch": 0.19325529542920847,
+      "grad_norm": 0.5058777332305908,
+      "learning_rate": 8e-05,
+      "loss": 1.7118,
+      "step": 3467
+    },
+    {
+      "epoch": 0.19331103678929765,
+      "grad_norm": 0.5257939100265503,
+      "learning_rate": 8e-05,
+      "loss": 1.8323,
+      "step": 3468
+    },
+    {
+      "epoch": 0.19336677814938685,
+      "grad_norm": 0.5075682401657104,
+      "learning_rate": 8e-05,
+      "loss": 1.884,
+      "step": 3469
+    },
+    {
+      "epoch": 0.19342251950947603,
+      "grad_norm": 0.5267577171325684,
+      "learning_rate": 8e-05,
+      "loss": 1.7791,
+      "step": 3470
+    },
+    {
+      "epoch": 0.1934782608695652,
+      "grad_norm": 0.44919127225875854,
+      "learning_rate": 8e-05,
+      "loss": 1.6597,
+      "step": 3471
+    },
+    {
+      "epoch": 0.19353400222965442,
+      "grad_norm": 0.4594336450099945,
+      "learning_rate": 8e-05,
+      "loss": 1.7039,
+      "step": 3472
+    },
+    {
+      "epoch": 0.1935897435897436,
+      "grad_norm": 0.4646263122558594,
+      "learning_rate": 8e-05,
+      "loss": 1.5838,
+      "step": 3473
+    },
+    {
+      "epoch": 0.19364548494983277,
+      "grad_norm": 0.5099480748176575,
+      "learning_rate": 8e-05,
+      "loss": 1.6858,
+      "step": 3474
+    },
+    {
+      "epoch": 0.19370122630992195,
+      "grad_norm": 0.5205021500587463,
+      "learning_rate": 8e-05,
+      "loss": 1.853,
+      "step": 3475
+    },
+    {
+      "epoch": 0.19375696767001116,
+      "grad_norm": 0.450195848941803,
+      "learning_rate": 8e-05,
+      "loss": 1.5248,
+      "step": 3476
+    },
+    {
+      "epoch": 0.19381270903010034,
+      "grad_norm": 0.49337515234947205,
+      "learning_rate": 8e-05,
+      "loss": 1.662,
+      "step": 3477
+    },
+    {
+      "epoch": 0.19386845039018952,
+      "grad_norm": 0.5189906358718872,
+      "learning_rate": 8e-05,
+      "loss": 1.8305,
+      "step": 3478
+    },
+    {
+      "epoch": 0.1939241917502787,
+      "grad_norm": 0.5012702345848083,
+      "learning_rate": 8e-05,
+      "loss": 1.656,
+      "step": 3479
+    },
+    {
+      "epoch": 0.1939799331103679,
+      "grad_norm": 0.4817003011703491,
+      "learning_rate": 8e-05,
+      "loss": 1.7924,
+      "step": 3480
+    },
+    {
+      "epoch": 0.19403567447045708,
+      "grad_norm": 0.45084530115127563,
+      "learning_rate": 8e-05,
+      "loss": 1.5779,
+      "step": 3481
+    },
+    {
+      "epoch": 0.19409141583054626,
+      "grad_norm": 0.5064246654510498,
+      "learning_rate": 8e-05,
+      "loss": 1.6494,
+      "step": 3482
+    },
+    {
+      "epoch": 0.19414715719063544,
+      "grad_norm": 0.515352725982666,
+      "learning_rate": 8e-05,
+      "loss": 1.7608,
+      "step": 3483
+    },
+    {
+      "epoch": 0.19420289855072465,
+      "grad_norm": 0.4773847162723541,
+      "learning_rate": 8e-05,
+      "loss": 1.9209,
+      "step": 3484
+    },
+    {
+      "epoch": 0.19425863991081382,
+      "grad_norm": 0.4580041766166687,
+      "learning_rate": 8e-05,
+      "loss": 1.6939,
+      "step": 3485
+    },
+    {
+      "epoch": 0.194314381270903,
+      "grad_norm": 0.5036474466323853,
+      "learning_rate": 8e-05,
+      "loss": 1.7424,
+      "step": 3486
+    },
+    {
+      "epoch": 0.1943701226309922,
+      "grad_norm": 0.5263992547988892,
+      "learning_rate": 8e-05,
+      "loss": 1.929,
+      "step": 3487
+    },
+    {
+      "epoch": 0.1944258639910814,
+      "grad_norm": 0.5149288177490234,
+      "learning_rate": 8e-05,
+      "loss": 1.7964,
+      "step": 3488
+    },
+    {
+      "epoch": 0.19448160535117057,
+      "grad_norm": 0.4722621738910675,
+      "learning_rate": 8e-05,
+      "loss": 1.6756,
+      "step": 3489
+    },
+    {
+      "epoch": 0.19453734671125975,
+      "grad_norm": 0.46639785170555115,
+      "learning_rate": 8e-05,
+      "loss": 1.6602,
+      "step": 3490
+    },
+    {
+      "epoch": 0.19459308807134895,
+      "grad_norm": 0.5215376019477844,
+      "learning_rate": 8e-05,
+      "loss": 1.8744,
+      "step": 3491
+    },
+    {
+      "epoch": 0.19464882943143813,
+      "grad_norm": 0.4881868362426758,
+      "learning_rate": 8e-05,
+      "loss": 1.8877,
+      "step": 3492
+    },
+    {
+      "epoch": 0.1947045707915273,
+      "grad_norm": 0.5139914751052856,
+      "learning_rate": 8e-05,
+      "loss": 1.8469,
+      "step": 3493
+    },
+    {
+      "epoch": 0.1947603121516165,
+      "grad_norm": 0.44681939482688904,
+      "learning_rate": 8e-05,
+      "loss": 1.5248,
+      "step": 3494
+    },
+    {
+      "epoch": 0.1948160535117057,
+      "grad_norm": 0.44363757967948914,
+      "learning_rate": 8e-05,
+      "loss": 1.6206,
+      "step": 3495
+    },
+    {
+      "epoch": 0.19487179487179487,
+      "grad_norm": 0.461874783039093,
+      "learning_rate": 8e-05,
+      "loss": 1.645,
+      "step": 3496
+    },
+    {
+      "epoch": 0.19492753623188405,
+      "grad_norm": 0.46725091338157654,
+      "learning_rate": 8e-05,
+      "loss": 1.7982,
+      "step": 3497
+    },
+    {
+      "epoch": 0.19498327759197323,
+      "grad_norm": 0.4530015289783478,
+      "learning_rate": 8e-05,
+      "loss": 1.5392,
+      "step": 3498
+    },
+    {
+      "epoch": 0.19503901895206244,
+      "grad_norm": 0.5092223882675171,
+      "learning_rate": 8e-05,
+      "loss": 1.7695,
+      "step": 3499
+    },
+    {
+      "epoch": 0.19509476031215162,
+      "grad_norm": 0.4746079444885254,
+      "learning_rate": 8e-05,
+      "loss": 1.7414,
+      "step": 3500
+    },
+    {
+      "epoch": 0.1951505016722408,
+      "grad_norm": 0.5035505294799805,
+      "learning_rate": 8e-05,
+      "loss": 1.8174,
+      "step": 3501
+    },
+    {
+      "epoch": 0.19520624303233,
+      "grad_norm": 0.49879980087280273,
+      "learning_rate": 8e-05,
+      "loss": 1.7143,
+      "step": 3502
+    },
+    {
+      "epoch": 0.19526198439241918,
+      "grad_norm": 0.4672737419605255,
+      "learning_rate": 8e-05,
+      "loss": 1.5076,
+      "step": 3503
+    },
+    {
+      "epoch": 0.19531772575250836,
+      "grad_norm": 0.4909713864326477,
+      "learning_rate": 8e-05,
+      "loss": 1.9685,
+      "step": 3504
+    },
+    {
+      "epoch": 0.19537346711259754,
+      "grad_norm": 0.5099307298660278,
+      "learning_rate": 8e-05,
+      "loss": 1.6727,
+      "step": 3505
+    },
+    {
+      "epoch": 0.19542920847268674,
+      "grad_norm": 0.49376577138900757,
+      "learning_rate": 8e-05,
+      "loss": 1.8202,
+      "step": 3506
+    },
+    {
+      "epoch": 0.19548494983277592,
+      "grad_norm": 0.4841322898864746,
+      "learning_rate": 8e-05,
+      "loss": 1.9417,
+      "step": 3507
+    },
+    {
+      "epoch": 0.1955406911928651,
+      "grad_norm": 0.5126500725746155,
+      "learning_rate": 8e-05,
+      "loss": 1.7245,
+      "step": 3508
+    },
+    {
+      "epoch": 0.19559643255295428,
+      "grad_norm": 0.5002952814102173,
+      "learning_rate": 8e-05,
+      "loss": 1.8467,
+      "step": 3509
+    },
+    {
+      "epoch": 0.1956521739130435,
+      "grad_norm": 0.5159156322479248,
+      "learning_rate": 8e-05,
+      "loss": 1.6416,
+      "step": 3510
+    },
+    {
+      "epoch": 0.19570791527313267,
+      "grad_norm": 0.5152454972267151,
+      "learning_rate": 8e-05,
+      "loss": 1.5334,
+      "step": 3511
+    },
+    {
+      "epoch": 0.19576365663322184,
+      "grad_norm": 0.47474539279937744,
+      "learning_rate": 8e-05,
+      "loss": 1.6742,
+      "step": 3512
+    },
+    {
+      "epoch": 0.19581939799331102,
+      "grad_norm": 0.4961521327495575,
+      "learning_rate": 8e-05,
+      "loss": 1.6075,
+      "step": 3513
+    },
+    {
+      "epoch": 0.19587513935340023,
+      "grad_norm": 0.4723234176635742,
+      "learning_rate": 8e-05,
+      "loss": 1.6282,
+      "step": 3514
+    },
+    {
+      "epoch": 0.1959308807134894,
+      "grad_norm": 0.48595571517944336,
+      "learning_rate": 8e-05,
+      "loss": 1.7194,
+      "step": 3515
+    },
+    {
+      "epoch": 0.1959866220735786,
+      "grad_norm": 0.4401930868625641,
+      "learning_rate": 8e-05,
+      "loss": 1.5628,
+      "step": 3516
+    },
+    {
+      "epoch": 0.1960423634336678,
+      "grad_norm": 0.493459016084671,
+      "learning_rate": 8e-05,
+      "loss": 1.6753,
+      "step": 3517
+    },
+    {
+      "epoch": 0.19609810479375697,
+      "grad_norm": 0.4308880567550659,
+      "learning_rate": 8e-05,
+      "loss": 1.4897,
+      "step": 3518
+    },
+    {
+      "epoch": 0.19615384615384615,
+      "grad_norm": 0.4599229097366333,
+      "learning_rate": 8e-05,
+      "loss": 1.4711,
+      "step": 3519
+    },
+    {
+      "epoch": 0.19620958751393533,
+      "grad_norm": 0.47723254561424255,
+      "learning_rate": 8e-05,
+      "loss": 1.6434,
+      "step": 3520
+    },
+    {
+      "epoch": 0.19626532887402454,
+      "grad_norm": 0.4884144365787506,
+      "learning_rate": 8e-05,
+      "loss": 1.6371,
+      "step": 3521
+    },
+    {
+      "epoch": 0.19632107023411371,
+      "grad_norm": 0.5237398147583008,
+      "learning_rate": 8e-05,
+      "loss": 1.8574,
+      "step": 3522
+    },
+    {
+      "epoch": 0.1963768115942029,
+      "grad_norm": 0.476162850856781,
+      "learning_rate": 8e-05,
+      "loss": 1.4318,
+      "step": 3523
+    },
+    {
+      "epoch": 0.19643255295429207,
+      "grad_norm": 0.48270082473754883,
+      "learning_rate": 8e-05,
+      "loss": 1.6951,
+      "step": 3524
+    },
+    {
+      "epoch": 0.19648829431438128,
+      "grad_norm": 0.4689558446407318,
+      "learning_rate": 8e-05,
+      "loss": 1.5397,
+      "step": 3525
+    },
+    {
+      "epoch": 0.19654403567447046,
+      "grad_norm": 0.4758528470993042,
+      "learning_rate": 8e-05,
+      "loss": 1.4821,
+      "step": 3526
+    },
+    {
+      "epoch": 0.19659977703455964,
+      "grad_norm": 0.5153961777687073,
+      "learning_rate": 8e-05,
+      "loss": 1.918,
+      "step": 3527
+    },
+    {
+      "epoch": 0.19665551839464884,
+      "grad_norm": 0.5305852890014648,
+      "learning_rate": 8e-05,
+      "loss": 1.8169,
+      "step": 3528
+    },
+    {
+      "epoch": 0.19671125975473802,
+      "grad_norm": 0.5135340094566345,
+      "learning_rate": 8e-05,
+      "loss": 1.8728,
+      "step": 3529
+    },
+    {
+      "epoch": 0.1967670011148272,
+      "grad_norm": 0.48879870772361755,
+      "learning_rate": 8e-05,
+      "loss": 1.7221,
+      "step": 3530
+    },
+    {
+      "epoch": 0.19682274247491638,
+      "grad_norm": 0.4915723204612732,
+      "learning_rate": 8e-05,
+      "loss": 1.7132,
+      "step": 3531
+    },
+    {
+      "epoch": 0.19687848383500559,
+      "grad_norm": 0.49147656559944153,
+      "learning_rate": 8e-05,
+      "loss": 1.6383,
+      "step": 3532
+    },
+    {
+      "epoch": 0.19693422519509476,
+      "grad_norm": 0.5483487248420715,
+      "learning_rate": 8e-05,
+      "loss": 1.8147,
+      "step": 3533
+    },
+    {
+      "epoch": 0.19698996655518394,
+      "grad_norm": 0.4494275748729706,
+      "learning_rate": 8e-05,
+      "loss": 1.4576,
+      "step": 3534
+    },
+    {
+      "epoch": 0.19704570791527312,
+      "grad_norm": 0.5068910121917725,
+      "learning_rate": 8e-05,
+      "loss": 1.7459,
+      "step": 3535
+    },
+    {
+      "epoch": 0.19710144927536233,
+      "grad_norm": 0.4981747269630432,
+      "learning_rate": 8e-05,
+      "loss": 1.5986,
+      "step": 3536
+    },
+    {
+      "epoch": 0.1971571906354515,
+      "grad_norm": 0.47614431381225586,
+      "learning_rate": 8e-05,
+      "loss": 1.3079,
+      "step": 3537
+    },
+    {
+      "epoch": 0.19721293199554069,
+      "grad_norm": 0.48696333169937134,
+      "learning_rate": 8e-05,
+      "loss": 1.6743,
+      "step": 3538
+    },
+    {
+      "epoch": 0.19726867335562986,
+      "grad_norm": 0.4971964359283447,
+      "learning_rate": 8e-05,
+      "loss": 1.8352,
+      "step": 3539
+    },
+    {
+      "epoch": 0.19732441471571907,
+      "grad_norm": 0.5399619936943054,
+      "learning_rate": 8e-05,
+      "loss": 1.8265,
+      "step": 3540
+    },
+    {
+      "epoch": 0.19738015607580825,
+      "grad_norm": 0.50235515832901,
+      "learning_rate": 8e-05,
+      "loss": 1.7268,
+      "step": 3541
+    },
+    {
+      "epoch": 0.19743589743589743,
+      "grad_norm": 0.5028350353240967,
+      "learning_rate": 8e-05,
+      "loss": 1.6568,
+      "step": 3542
+    },
+    {
+      "epoch": 0.19749163879598663,
+      "grad_norm": 0.5096933841705322,
+      "learning_rate": 8e-05,
+      "loss": 1.6951,
+      "step": 3543
+    },
+    {
+      "epoch": 0.1975473801560758,
+      "grad_norm": 0.5013551115989685,
+      "learning_rate": 8e-05,
+      "loss": 1.5317,
+      "step": 3544
+    },
+    {
+      "epoch": 0.197603121516165,
+      "grad_norm": 0.5122877955436707,
+      "learning_rate": 8e-05,
+      "loss": 1.8456,
+      "step": 3545
+    },
+    {
+      "epoch": 0.19765886287625417,
+      "grad_norm": 0.5154286026954651,
+      "learning_rate": 8e-05,
+      "loss": 1.7878,
+      "step": 3546
+    },
+    {
+      "epoch": 0.19771460423634338,
+      "grad_norm": 0.5381590127944946,
+      "learning_rate": 8e-05,
+      "loss": 1.8982,
+      "step": 3547
+    },
+    {
+      "epoch": 0.19777034559643256,
+      "grad_norm": 0.5874319672584534,
+      "learning_rate": 8e-05,
+      "loss": 1.6731,
+      "step": 3548
+    },
+    {
+      "epoch": 0.19782608695652174,
+      "grad_norm": 0.5069901347160339,
+      "learning_rate": 8e-05,
+      "loss": 1.7125,
+      "step": 3549
+    },
+    {
+      "epoch": 0.1978818283166109,
+      "grad_norm": 0.4645526111125946,
+      "learning_rate": 8e-05,
+      "loss": 1.3551,
+      "step": 3550
+    },
+    {
+      "epoch": 0.19793756967670012,
+      "grad_norm": 0.4883759021759033,
+      "learning_rate": 8e-05,
+      "loss": 1.6016,
+      "step": 3551
+    },
+    {
+      "epoch": 0.1979933110367893,
+      "grad_norm": 0.49408966302871704,
+      "learning_rate": 8e-05,
+      "loss": 1.8289,
+      "step": 3552
+    },
+    {
+      "epoch": 0.19804905239687848,
+      "grad_norm": 0.5287270545959473,
+      "learning_rate": 8e-05,
+      "loss": 1.6608,
+      "step": 3553
+    },
+    {
+      "epoch": 0.19810479375696766,
+      "grad_norm": 0.48724809288978577,
+      "learning_rate": 8e-05,
+      "loss": 1.7525,
+      "step": 3554
+    },
+    {
+      "epoch": 0.19816053511705686,
+      "grad_norm": 0.48583024740219116,
+      "learning_rate": 8e-05,
+      "loss": 1.8145,
+      "step": 3555
+    },
+    {
+      "epoch": 0.19821627647714604,
+      "grad_norm": 0.5329705476760864,
+      "learning_rate": 8e-05,
+      "loss": 1.7537,
+      "step": 3556
+    },
+    {
+      "epoch": 0.19827201783723522,
+      "grad_norm": 0.5065861940383911,
+      "learning_rate": 8e-05,
+      "loss": 1.7054,
+      "step": 3557
+    },
+    {
+      "epoch": 0.19832775919732443,
+      "grad_norm": 0.48774999380111694,
+      "learning_rate": 8e-05,
+      "loss": 1.8509,
+      "step": 3558
+    },
+    {
+      "epoch": 0.1983835005574136,
+      "grad_norm": 0.5659329295158386,
+      "learning_rate": 8e-05,
+      "loss": 2.0576,
+      "step": 3559
+    },
+    {
+      "epoch": 0.19843924191750278,
+      "grad_norm": 0.4959155023097992,
+      "learning_rate": 8e-05,
+      "loss": 1.6683,
+      "step": 3560
+    },
+    {
+      "epoch": 0.19849498327759196,
+      "grad_norm": 0.5472777485847473,
+      "learning_rate": 8e-05,
+      "loss": 1.5919,
+      "step": 3561
+    },
+    {
+      "epoch": 0.19855072463768117,
+      "grad_norm": 0.49425390362739563,
+      "learning_rate": 8e-05,
+      "loss": 1.8153,
+      "step": 3562
+    },
+    {
+      "epoch": 0.19860646599777035,
+      "grad_norm": 0.5090674161911011,
+      "learning_rate": 8e-05,
+      "loss": 1.8033,
+      "step": 3563
+    },
+    {
+      "epoch": 0.19866220735785953,
+      "grad_norm": 0.45376265048980713,
+      "learning_rate": 8e-05,
+      "loss": 1.8647,
+      "step": 3564
+    },
+    {
+      "epoch": 0.1987179487179487,
+      "grad_norm": 0.5247412323951721,
+      "learning_rate": 8e-05,
+      "loss": 2.0173,
+      "step": 3565
+    },
+    {
+      "epoch": 0.1987736900780379,
+      "grad_norm": 0.5044318437576294,
+      "learning_rate": 8e-05,
+      "loss": 1.8077,
+      "step": 3566
+    },
+    {
+      "epoch": 0.1988294314381271,
+      "grad_norm": 0.49488723278045654,
+      "learning_rate": 8e-05,
+      "loss": 1.7759,
+      "step": 3567
+    },
+    {
+      "epoch": 0.19888517279821627,
+      "grad_norm": 0.46972551941871643,
+      "learning_rate": 8e-05,
+      "loss": 1.6848,
+      "step": 3568
+    },
+    {
+      "epoch": 0.19894091415830548,
+      "grad_norm": 0.49496331810951233,
+      "learning_rate": 8e-05,
+      "loss": 1.5929,
+      "step": 3569
+    },
+    {
+      "epoch": 0.19899665551839466,
+      "grad_norm": 0.5036976337432861,
+      "learning_rate": 8e-05,
+      "loss": 1.9322,
+      "step": 3570
+    },
+    {
+      "epoch": 0.19905239687848383,
+      "grad_norm": 0.5012423992156982,
+      "learning_rate": 8e-05,
+      "loss": 1.7887,
+      "step": 3571
+    },
+    {
+      "epoch": 0.199108138238573,
+      "grad_norm": 0.47637394070625305,
+      "learning_rate": 8e-05,
+      "loss": 1.5631,
+      "step": 3572
+    },
+    {
+      "epoch": 0.19916387959866222,
+      "grad_norm": 0.47292307019233704,
+      "learning_rate": 8e-05,
+      "loss": 1.6106,
+      "step": 3573
+    },
+    {
+      "epoch": 0.1992196209587514,
+      "grad_norm": 0.48646533489227295,
+      "learning_rate": 8e-05,
+      "loss": 1.7756,
+      "step": 3574
+    },
+    {
+      "epoch": 0.19927536231884058,
+      "grad_norm": 0.4798130691051483,
+      "learning_rate": 8e-05,
+      "loss": 1.5798,
+      "step": 3575
+    },
+    {
+      "epoch": 0.19933110367892976,
+      "grad_norm": 0.4687424600124359,
+      "learning_rate": 8e-05,
+      "loss": 1.6849,
+      "step": 3576
+    },
+    {
+      "epoch": 0.19938684503901896,
+      "grad_norm": 0.49993613362312317,
+      "learning_rate": 8e-05,
+      "loss": 1.7479,
+      "step": 3577
+    },
+    {
+      "epoch": 0.19944258639910814,
+      "grad_norm": 0.5133943557739258,
+      "learning_rate": 8e-05,
+      "loss": 1.7845,
+      "step": 3578
+    },
+    {
+      "epoch": 0.19949832775919732,
+      "grad_norm": 0.4577353596687317,
+      "learning_rate": 8e-05,
+      "loss": 1.6277,
+      "step": 3579
+    },
+    {
+      "epoch": 0.1995540691192865,
+      "grad_norm": 0.44345855712890625,
+      "learning_rate": 8e-05,
+      "loss": 1.5042,
+      "step": 3580
+    },
+    {
+      "epoch": 0.1996098104793757,
+      "grad_norm": 0.5210535526275635,
+      "learning_rate": 8e-05,
+      "loss": 1.8266,
+      "step": 3581
+    },
+    {
+      "epoch": 0.19966555183946488,
+      "grad_norm": 0.534353494644165,
+      "learning_rate": 8e-05,
+      "loss": 1.7317,
+      "step": 3582
+    },
+    {
+      "epoch": 0.19972129319955406,
+      "grad_norm": 0.531521737575531,
+      "learning_rate": 8e-05,
+      "loss": 1.6291,
+      "step": 3583
+    },
+    {
+      "epoch": 0.19977703455964327,
+      "grad_norm": 0.4943453073501587,
+      "learning_rate": 8e-05,
+      "loss": 1.5845,
+      "step": 3584
+    },
+    {
+      "epoch": 0.19983277591973245,
+      "grad_norm": 0.4547666907310486,
+      "learning_rate": 8e-05,
+      "loss": 1.7167,
+      "step": 3585
+    },
+    {
+      "epoch": 0.19988851727982163,
+      "grad_norm": 0.45870059728622437,
+      "learning_rate": 8e-05,
+      "loss": 1.646,
+      "step": 3586
+    },
+    {
+      "epoch": 0.1999442586399108,
+      "grad_norm": 0.5779637098312378,
+      "learning_rate": 8e-05,
+      "loss": 1.5718,
+      "step": 3587
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5157275795936584,
+      "learning_rate": 8e-05,
+      "loss": 1.7512,
+      "step": 3588
+    },
+    {
+      "epoch": 0.2000557413600892,
+      "grad_norm": 0.4938097596168518,
+      "learning_rate": 8e-05,
+      "loss": 1.701,
+      "step": 3589
+    },
+    {
+      "epoch": 0.20011148272017837,
+      "grad_norm": 0.5136789083480835,
+      "learning_rate": 8e-05,
+      "loss": 1.7808,
+      "step": 3590
+    },
+    {
+      "epoch": 0.20016722408026755,
+      "grad_norm": 0.49631932377815247,
+      "learning_rate": 8e-05,
+      "loss": 1.6969,
+      "step": 3591
+    },
+    {
+      "epoch": 0.20022296544035675,
+      "grad_norm": 0.5018367767333984,
+      "learning_rate": 8e-05,
+      "loss": 1.7466,
+      "step": 3592
+    },
+    {
+      "epoch": 0.20027870680044593,
+      "grad_norm": 0.5183612704277039,
+      "learning_rate": 8e-05,
+      "loss": 1.8397,
+      "step": 3593
+    },
+    {
+      "epoch": 0.2003344481605351,
+      "grad_norm": 0.48466330766677856,
+      "learning_rate": 8e-05,
+      "loss": 1.7564,
+      "step": 3594
+    },
+    {
+      "epoch": 0.2003901895206243,
+      "grad_norm": 0.448240727186203,
+      "learning_rate": 8e-05,
+      "loss": 1.4317,
+      "step": 3595
+    },
+    {
+      "epoch": 0.2004459308807135,
+      "grad_norm": 0.51080721616745,
+      "learning_rate": 8e-05,
+      "loss": 1.743,
+      "step": 3596
+    },
+    {
+      "epoch": 0.20050167224080268,
+      "grad_norm": 0.5205346941947937,
+      "learning_rate": 8e-05,
+      "loss": 1.6999,
+      "step": 3597
+    },
+    {
+      "epoch": 0.20055741360089185,
+      "grad_norm": 0.4996013641357422,
+      "learning_rate": 8e-05,
+      "loss": 1.5763,
+      "step": 3598
+    },
+    {
+      "epoch": 0.20061315496098106,
+      "grad_norm": 0.465975284576416,
+      "learning_rate": 8e-05,
+      "loss": 1.7167,
+      "step": 3599
+    },
+    {
+      "epoch": 0.20066889632107024,
+      "grad_norm": 0.5416645407676697,
+      "learning_rate": 8e-05,
+      "loss": 1.8032,
+      "step": 3600
+    },
+    {
+      "epoch": 0.20072463768115942,
+      "grad_norm": 0.49942782521247864,
+      "learning_rate": 8e-05,
+      "loss": 1.4552,
+      "step": 3601
+    },
+    {
+      "epoch": 0.2007803790412486,
+      "grad_norm": 0.4765155017375946,
+      "learning_rate": 8e-05,
+      "loss": 1.5083,
+      "step": 3602
+    },
+    {
+      "epoch": 0.2008361204013378,
+      "grad_norm": 0.5089762806892395,
+      "learning_rate": 8e-05,
+      "loss": 1.8018,
+      "step": 3603
+    },
+    {
+      "epoch": 0.20089186176142698,
+      "grad_norm": 0.5311180949211121,
+      "learning_rate": 8e-05,
+      "loss": 1.3827,
+      "step": 3604
+    },
+    {
+      "epoch": 0.20094760312151616,
+      "grad_norm": 0.4367247521877289,
+      "learning_rate": 8e-05,
+      "loss": 1.433,
+      "step": 3605
+    },
+    {
+      "epoch": 0.20100334448160534,
+      "grad_norm": 0.5144723653793335,
+      "learning_rate": 8e-05,
+      "loss": 1.4288,
+      "step": 3606
+    },
+    {
+      "epoch": 0.20105908584169455,
+      "grad_norm": 0.5270749926567078,
+      "learning_rate": 8e-05,
+      "loss": 1.7999,
+      "step": 3607
+    },
+    {
+      "epoch": 0.20111482720178372,
+      "grad_norm": 0.4606470763683319,
+      "learning_rate": 8e-05,
+      "loss": 1.5438,
+      "step": 3608
+    },
+    {
+      "epoch": 0.2011705685618729,
+      "grad_norm": 0.5114845037460327,
+      "learning_rate": 8e-05,
+      "loss": 1.7967,
+      "step": 3609
+    },
+    {
+      "epoch": 0.20122630992196208,
+      "grad_norm": 0.4875860810279846,
+      "learning_rate": 8e-05,
+      "loss": 1.8787,
+      "step": 3610
+    },
+    {
+      "epoch": 0.2012820512820513,
+      "grad_norm": 0.5535544157028198,
+      "learning_rate": 8e-05,
+      "loss": 1.5494,
+      "step": 3611
+    },
+    {
+      "epoch": 0.20133779264214047,
+      "grad_norm": 0.47383052110671997,
+      "learning_rate": 8e-05,
+      "loss": 1.6689,
+      "step": 3612
+    },
+    {
+      "epoch": 0.20139353400222965,
+      "grad_norm": 0.4879821836948395,
+      "learning_rate": 8e-05,
+      "loss": 1.5114,
+      "step": 3613
+    },
+    {
+      "epoch": 0.20144927536231885,
+      "grad_norm": 0.5040041208267212,
+      "learning_rate": 8e-05,
+      "loss": 1.6116,
+      "step": 3614
+    },
+    {
+      "epoch": 0.20150501672240803,
+      "grad_norm": 0.4656704068183899,
+      "learning_rate": 8e-05,
+      "loss": 1.6972,
+      "step": 3615
+    },
+    {
+      "epoch": 0.2015607580824972,
+      "grad_norm": 0.47616857290267944,
+      "learning_rate": 8e-05,
+      "loss": 1.8055,
+      "step": 3616
+    },
+    {
+      "epoch": 0.2016164994425864,
+      "grad_norm": 0.4548301696777344,
+      "learning_rate": 8e-05,
+      "loss": 1.742,
+      "step": 3617
+    },
+    {
+      "epoch": 0.2016722408026756,
+      "grad_norm": 0.4619283676147461,
+      "learning_rate": 8e-05,
+      "loss": 1.6631,
+      "step": 3618
+    },
+    {
+      "epoch": 0.20172798216276477,
+      "grad_norm": 0.4462347626686096,
+      "learning_rate": 8e-05,
+      "loss": 1.5333,
+      "step": 3619
+    },
+    {
+      "epoch": 0.20178372352285395,
+      "grad_norm": 0.523777961730957,
+      "learning_rate": 8e-05,
+      "loss": 1.734,
+      "step": 3620
+    },
+    {
+      "epoch": 0.20183946488294313,
+      "grad_norm": 0.4466857612133026,
+      "learning_rate": 8e-05,
+      "loss": 1.4318,
+      "step": 3621
+    },
+    {
+      "epoch": 0.20189520624303234,
+      "grad_norm": 0.5048567056655884,
+      "learning_rate": 8e-05,
+      "loss": 1.7637,
+      "step": 3622
+    },
+    {
+      "epoch": 0.20195094760312152,
+      "grad_norm": 0.5019554495811462,
+      "learning_rate": 8e-05,
+      "loss": 1.6489,
+      "step": 3623
+    },
+    {
+      "epoch": 0.2020066889632107,
+      "grad_norm": 0.5542216897010803,
+      "learning_rate": 8e-05,
+      "loss": 1.9606,
+      "step": 3624
+    },
+    {
+      "epoch": 0.2020624303232999,
+      "grad_norm": 0.4789694547653198,
+      "learning_rate": 8e-05,
+      "loss": 1.786,
+      "step": 3625
+    },
+    {
+      "epoch": 0.20211817168338908,
+      "grad_norm": 0.5140082240104675,
+      "learning_rate": 8e-05,
+      "loss": 1.8251,
+      "step": 3626
+    },
+    {
+      "epoch": 0.20217391304347826,
+      "grad_norm": 0.4956878125667572,
+      "learning_rate": 8e-05,
+      "loss": 1.8947,
+      "step": 3627
+    },
+    {
+      "epoch": 0.20222965440356744,
+      "grad_norm": 0.541315495967865,
+      "learning_rate": 8e-05,
+      "loss": 1.9393,
+      "step": 3628
+    },
+    {
+      "epoch": 0.20228539576365664,
+      "grad_norm": 0.4940366744995117,
+      "learning_rate": 8e-05,
+      "loss": 1.8449,
+      "step": 3629
+    },
+    {
+      "epoch": 0.20234113712374582,
+      "grad_norm": 0.5367388725280762,
+      "learning_rate": 8e-05,
+      "loss": 1.9762,
+      "step": 3630
+    },
+    {
+      "epoch": 0.202396878483835,
+      "grad_norm": 0.4210890829563141,
+      "learning_rate": 8e-05,
+      "loss": 1.4644,
+      "step": 3631
+    },
+    {
+      "epoch": 0.20245261984392418,
+      "grad_norm": 0.4935843348503113,
+      "learning_rate": 8e-05,
+      "loss": 1.8778,
+      "step": 3632
+    },
+    {
+      "epoch": 0.2025083612040134,
+      "grad_norm": 0.4714401960372925,
+      "learning_rate": 8e-05,
+      "loss": 1.6896,
+      "step": 3633
+    },
+    {
+      "epoch": 0.20256410256410257,
+      "grad_norm": 0.5045357346534729,
+      "learning_rate": 8e-05,
+      "loss": 1.8094,
+      "step": 3634
+    },
+    {
+      "epoch": 0.20261984392419174,
+      "grad_norm": 0.5472859740257263,
+      "learning_rate": 8e-05,
+      "loss": 1.8484,
+      "step": 3635
+    },
+    {
+      "epoch": 0.20267558528428092,
+      "grad_norm": 0.4495568573474884,
+      "learning_rate": 8e-05,
+      "loss": 1.3515,
+      "step": 3636
+    },
+    {
+      "epoch": 0.20273132664437013,
+      "grad_norm": 0.478530615568161,
+      "learning_rate": 8e-05,
+      "loss": 1.716,
+      "step": 3637
+    },
+    {
+      "epoch": 0.2027870680044593,
+      "grad_norm": 0.4960003197193146,
+      "learning_rate": 8e-05,
+      "loss": 1.7212,
+      "step": 3638
+    },
+    {
+      "epoch": 0.2028428093645485,
+      "grad_norm": 0.44264334440231323,
+      "learning_rate": 8e-05,
+      "loss": 1.5679,
+      "step": 3639
+    },
+    {
+      "epoch": 0.2028985507246377,
+      "grad_norm": 0.5412372350692749,
+      "learning_rate": 8e-05,
+      "loss": 2.0983,
+      "step": 3640
+    },
+    {
+      "epoch": 0.20295429208472687,
+      "grad_norm": 0.5256844758987427,
+      "learning_rate": 8e-05,
+      "loss": 1.5816,
+      "step": 3641
+    },
+    {
+      "epoch": 0.20301003344481605,
+      "grad_norm": 0.47482818365097046,
+      "learning_rate": 8e-05,
+      "loss": 1.4953,
+      "step": 3642
+    },
+    {
+      "epoch": 0.20306577480490523,
+      "grad_norm": 0.4723987281322479,
+      "learning_rate": 8e-05,
+      "loss": 1.7102,
+      "step": 3643
+    },
+    {
+      "epoch": 0.20312151616499444,
+      "grad_norm": 0.454328328371048,
+      "learning_rate": 8e-05,
+      "loss": 1.6628,
+      "step": 3644
+    },
+    {
+      "epoch": 0.20317725752508362,
+      "grad_norm": 0.46709755063056946,
+      "learning_rate": 8e-05,
+      "loss": 1.6382,
+      "step": 3645
+    },
+    {
+      "epoch": 0.2032329988851728,
+      "grad_norm": 0.48052144050598145,
+      "learning_rate": 8e-05,
+      "loss": 1.4057,
+      "step": 3646
+    },
+    {
+      "epoch": 0.20328874024526197,
+      "grad_norm": 0.5479748845100403,
+      "learning_rate": 8e-05,
+      "loss": 1.8707,
+      "step": 3647
+    },
+    {
+      "epoch": 0.20334448160535118,
+      "grad_norm": 0.4681054651737213,
+      "learning_rate": 8e-05,
+      "loss": 1.4069,
+      "step": 3648
+    },
+    {
+      "epoch": 0.20340022296544036,
+      "grad_norm": 0.4966769516468048,
+      "learning_rate": 8e-05,
+      "loss": 1.6872,
+      "step": 3649
+    },
+    {
+      "epoch": 0.20345596432552954,
+      "grad_norm": 0.5099707245826721,
+      "learning_rate": 8e-05,
+      "loss": 1.5968,
+      "step": 3650
+    },
+    {
+      "epoch": 0.20351170568561872,
+      "grad_norm": 0.5672361254692078,
+      "learning_rate": 8e-05,
+      "loss": 1.8859,
+      "step": 3651
+    },
+    {
+      "epoch": 0.20356744704570792,
+      "grad_norm": 0.5006271004676819,
+      "learning_rate": 8e-05,
+      "loss": 1.6707,
+      "step": 3652
+    },
+    {
+      "epoch": 0.2036231884057971,
+      "grad_norm": 0.5286397337913513,
+      "learning_rate": 8e-05,
+      "loss": 1.7566,
+      "step": 3653
+    },
+    {
+      "epoch": 0.20367892976588628,
+      "grad_norm": 0.5863698720932007,
+      "learning_rate": 8e-05,
+      "loss": 1.883,
+      "step": 3654
+    },
+    {
+      "epoch": 0.20373467112597549,
+      "grad_norm": 0.47385674715042114,
+      "learning_rate": 8e-05,
+      "loss": 1.6727,
+      "step": 3655
+    },
+    {
+      "epoch": 0.20379041248606466,
+      "grad_norm": 0.490665465593338,
+      "learning_rate": 8e-05,
+      "loss": 1.7301,
+      "step": 3656
+    },
+    {
+      "epoch": 0.20384615384615384,
+      "grad_norm": 0.49912935495376587,
+      "learning_rate": 8e-05,
+      "loss": 1.6501,
+      "step": 3657
+    },
+    {
+      "epoch": 0.20390189520624302,
+      "grad_norm": 0.4729326367378235,
+      "learning_rate": 8e-05,
+      "loss": 1.5658,
+      "step": 3658
+    },
+    {
+      "epoch": 0.20395763656633223,
+      "grad_norm": 0.4842578172683716,
+      "learning_rate": 8e-05,
+      "loss": 1.5118,
+      "step": 3659
+    },
+    {
+      "epoch": 0.2040133779264214,
+      "grad_norm": 0.494857519865036,
+      "learning_rate": 8e-05,
+      "loss": 1.7177,
+      "step": 3660
+    },
+    {
+      "epoch": 0.2040691192865106,
+      "grad_norm": 0.4974959194660187,
+      "learning_rate": 8e-05,
+      "loss": 1.668,
+      "step": 3661
+    },
+    {
+      "epoch": 0.20412486064659977,
+      "grad_norm": 0.5402908325195312,
+      "learning_rate": 8e-05,
+      "loss": 1.9019,
+      "step": 3662
+    },
+    {
+      "epoch": 0.20418060200668897,
+      "grad_norm": 0.5406147241592407,
+      "learning_rate": 8e-05,
+      "loss": 1.7114,
+      "step": 3663
+    },
+    {
+      "epoch": 0.20423634336677815,
+      "grad_norm": 0.47465139627456665,
+      "learning_rate": 8e-05,
+      "loss": 1.707,
+      "step": 3664
+    },
+    {
+      "epoch": 0.20429208472686733,
+      "grad_norm": 0.5029194355010986,
+      "learning_rate": 8e-05,
+      "loss": 1.6362,
+      "step": 3665
+    },
+    {
+      "epoch": 0.20434782608695654,
+      "grad_norm": 0.49449896812438965,
+      "learning_rate": 8e-05,
+      "loss": 1.6479,
+      "step": 3666
+    },
+    {
+      "epoch": 0.20440356744704571,
+      "grad_norm": 0.5273813009262085,
+      "learning_rate": 8e-05,
+      "loss": 1.8518,
+      "step": 3667
+    },
+    {
+      "epoch": 0.2044593088071349,
+      "grad_norm": 0.4844883382320404,
+      "learning_rate": 8e-05,
+      "loss": 1.8226,
+      "step": 3668
+    },
+    {
+      "epoch": 0.20451505016722407,
+      "grad_norm": 0.5077101588249207,
+      "learning_rate": 8e-05,
+      "loss": 1.8068,
+      "step": 3669
+    },
+    {
+      "epoch": 0.20457079152731328,
+      "grad_norm": 0.48901060223579407,
+      "learning_rate": 8e-05,
+      "loss": 1.5398,
+      "step": 3670
+    },
+    {
+      "epoch": 0.20462653288740246,
+      "grad_norm": 0.43883004784584045,
+      "learning_rate": 8e-05,
+      "loss": 1.3666,
+      "step": 3671
+    },
+    {
+      "epoch": 0.20468227424749164,
+      "grad_norm": 0.4607550799846649,
+      "learning_rate": 8e-05,
+      "loss": 1.4984,
+      "step": 3672
+    },
+    {
+      "epoch": 0.20473801560758081,
+      "grad_norm": 0.4580022096633911,
+      "learning_rate": 8e-05,
+      "loss": 1.725,
+      "step": 3673
+    },
+    {
+      "epoch": 0.20479375696767002,
+      "grad_norm": 0.47610732913017273,
+      "learning_rate": 8e-05,
+      "loss": 1.7055,
+      "step": 3674
+    },
+    {
+      "epoch": 0.2048494983277592,
+      "grad_norm": 0.48338666558265686,
+      "learning_rate": 8e-05,
+      "loss": 1.8632,
+      "step": 3675
+    },
+    {
+      "epoch": 0.20490523968784838,
+      "grad_norm": 0.4903329312801361,
+      "learning_rate": 8e-05,
+      "loss": 1.4769,
+      "step": 3676
+    },
+    {
+      "epoch": 0.20496098104793756,
+      "grad_norm": 0.5103690028190613,
+      "learning_rate": 8e-05,
+      "loss": 1.6833,
+      "step": 3677
+    },
+    {
+      "epoch": 0.20501672240802676,
+      "grad_norm": 0.5390543937683105,
+      "learning_rate": 8e-05,
+      "loss": 1.8047,
+      "step": 3678
+    },
+    {
+      "epoch": 0.20507246376811594,
+      "grad_norm": 0.4767552614212036,
+      "learning_rate": 8e-05,
+      "loss": 1.7662,
+      "step": 3679
+    },
+    {
+      "epoch": 0.20512820512820512,
+      "grad_norm": 0.5675135254859924,
+      "learning_rate": 8e-05,
+      "loss": 1.7962,
+      "step": 3680
+    },
+    {
+      "epoch": 0.20518394648829433,
+      "grad_norm": 0.521496593952179,
+      "learning_rate": 8e-05,
+      "loss": 1.7376,
+      "step": 3681
+    },
+    {
+      "epoch": 0.2052396878483835,
+      "grad_norm": 0.5281379818916321,
+      "learning_rate": 8e-05,
+      "loss": 1.6355,
+      "step": 3682
+    },
+    {
+      "epoch": 0.20529542920847269,
+      "grad_norm": 0.47564199566841125,
+      "learning_rate": 8e-05,
+      "loss": 1.5621,
+      "step": 3683
+    },
+    {
+      "epoch": 0.20535117056856186,
+      "grad_norm": 0.4812081754207611,
+      "learning_rate": 8e-05,
+      "loss": 1.6206,
+      "step": 3684
+    },
+    {
+      "epoch": 0.20540691192865107,
+      "grad_norm": 0.49556106328964233,
+      "learning_rate": 8e-05,
+      "loss": 1.7597,
+      "step": 3685
+    },
+    {
+      "epoch": 0.20546265328874025,
+      "grad_norm": 0.5186765789985657,
+      "learning_rate": 8e-05,
+      "loss": 1.6219,
+      "step": 3686
+    },
+    {
+      "epoch": 0.20551839464882943,
+      "grad_norm": 0.5270774960517883,
+      "learning_rate": 8e-05,
+      "loss": 1.7674,
+      "step": 3687
+    },
+    {
+      "epoch": 0.2055741360089186,
+      "grad_norm": 0.5229092240333557,
+      "learning_rate": 8e-05,
+      "loss": 1.9134,
+      "step": 3688
+    },
+    {
+      "epoch": 0.2056298773690078,
+      "grad_norm": 0.5199306607246399,
+      "learning_rate": 8e-05,
+      "loss": 1.9065,
+      "step": 3689
+    },
+    {
+      "epoch": 0.205685618729097,
+      "grad_norm": 0.5622035264968872,
+      "learning_rate": 8e-05,
+      "loss": 1.7641,
+      "step": 3690
+    },
+    {
+      "epoch": 0.20574136008918617,
+      "grad_norm": 0.5008182525634766,
+      "learning_rate": 8e-05,
+      "loss": 1.6597,
+      "step": 3691
+    },
+    {
+      "epoch": 0.20579710144927535,
+      "grad_norm": 0.46879804134368896,
+      "learning_rate": 8e-05,
+      "loss": 1.4784,
+      "step": 3692
+    },
+    {
+      "epoch": 0.20585284280936456,
+      "grad_norm": 0.4561975300312042,
+      "learning_rate": 8e-05,
+      "loss": 1.5148,
+      "step": 3693
+    },
+    {
+      "epoch": 0.20590858416945373,
+      "grad_norm": 0.5250406265258789,
+      "learning_rate": 8e-05,
+      "loss": 1.8827,
+      "step": 3694
+    },
+    {
+      "epoch": 0.2059643255295429,
+      "grad_norm": 0.48637449741363525,
+      "learning_rate": 8e-05,
+      "loss": 1.6882,
+      "step": 3695
+    },
+    {
+      "epoch": 0.20602006688963212,
+      "grad_norm": 0.5028102397918701,
+      "learning_rate": 8e-05,
+      "loss": 1.9182,
+      "step": 3696
+    },
+    {
+      "epoch": 0.2060758082497213,
+      "grad_norm": 0.48546063899993896,
+      "learning_rate": 8e-05,
+      "loss": 1.7412,
+      "step": 3697
+    },
+    {
+      "epoch": 0.20613154960981048,
+      "grad_norm": 0.46001705527305603,
+      "learning_rate": 8e-05,
+      "loss": 1.765,
+      "step": 3698
+    },
+    {
+      "epoch": 0.20618729096989966,
+      "grad_norm": 0.4724488854408264,
+      "learning_rate": 8e-05,
+      "loss": 1.6078,
+      "step": 3699
+    },
+    {
+      "epoch": 0.20624303232998886,
+      "grad_norm": 0.51436448097229,
+      "learning_rate": 8e-05,
+      "loss": 1.7974,
+      "step": 3700
+    },
+    {
+      "epoch": 0.20629877369007804,
+      "grad_norm": 0.5569310784339905,
+      "learning_rate": 8e-05,
+      "loss": 1.9332,
+      "step": 3701
+    },
+    {
+      "epoch": 0.20635451505016722,
+      "grad_norm": 0.5193493366241455,
+      "learning_rate": 8e-05,
+      "loss": 1.7944,
+      "step": 3702
+    },
+    {
+      "epoch": 0.2064102564102564,
+      "grad_norm": 0.5052384734153748,
+      "learning_rate": 8e-05,
+      "loss": 1.84,
+      "step": 3703
+    },
+    {
+      "epoch": 0.2064659977703456,
+      "grad_norm": 0.4906817674636841,
+      "learning_rate": 8e-05,
+      "loss": 1.6263,
+      "step": 3704
+    },
+    {
+      "epoch": 0.20652173913043478,
+      "grad_norm": 0.48959335684776306,
+      "learning_rate": 8e-05,
+      "loss": 1.6381,
+      "step": 3705
+    },
+    {
+      "epoch": 0.20657748049052396,
+      "grad_norm": 0.4935573935508728,
+      "learning_rate": 8e-05,
+      "loss": 1.7671,
+      "step": 3706
+    },
+    {
+      "epoch": 0.20663322185061314,
+      "grad_norm": 0.5062634348869324,
+      "learning_rate": 8e-05,
+      "loss": 1.8406,
+      "step": 3707
+    },
+    {
+      "epoch": 0.20668896321070235,
+      "grad_norm": 0.5000908970832825,
+      "learning_rate": 8e-05,
+      "loss": 1.7604,
+      "step": 3708
+    },
+    {
+      "epoch": 0.20674470457079153,
+      "grad_norm": 0.5079050660133362,
+      "learning_rate": 8e-05,
+      "loss": 1.8167,
+      "step": 3709
+    },
+    {
+      "epoch": 0.2068004459308807,
+      "grad_norm": 0.48654088377952576,
+      "learning_rate": 8e-05,
+      "loss": 1.7648,
+      "step": 3710
+    },
+    {
+      "epoch": 0.2068561872909699,
+      "grad_norm": 0.5441576838493347,
+      "learning_rate": 8e-05,
+      "loss": 2.0396,
+      "step": 3711
+    },
+    {
+      "epoch": 0.2069119286510591,
+      "grad_norm": 0.48717156052589417,
+      "learning_rate": 8e-05,
+      "loss": 1.8172,
+      "step": 3712
+    },
+    {
+      "epoch": 0.20696767001114827,
+      "grad_norm": 0.4929221272468567,
+      "learning_rate": 8e-05,
+      "loss": 1.5783,
+      "step": 3713
+    },
+    {
+      "epoch": 0.20702341137123745,
+      "grad_norm": 0.5209141969680786,
+      "learning_rate": 8e-05,
+      "loss": 1.9343,
+      "step": 3714
+    },
+    {
+      "epoch": 0.20707915273132665,
+      "grad_norm": 0.4781465530395508,
+      "learning_rate": 8e-05,
+      "loss": 1.8023,
+      "step": 3715
+    },
+    {
+      "epoch": 0.20713489409141583,
+      "grad_norm": 0.475308358669281,
+      "learning_rate": 8e-05,
+      "loss": 1.7513,
+      "step": 3716
+    },
+    {
+      "epoch": 0.207190635451505,
+      "grad_norm": 0.48245689272880554,
+      "learning_rate": 8e-05,
+      "loss": 1.6194,
+      "step": 3717
+    },
+    {
+      "epoch": 0.2072463768115942,
+      "grad_norm": 0.48839378356933594,
+      "learning_rate": 8e-05,
+      "loss": 1.6131,
+      "step": 3718
+    },
+    {
+      "epoch": 0.2073021181716834,
+      "grad_norm": 0.49294933676719666,
+      "learning_rate": 8e-05,
+      "loss": 1.8494,
+      "step": 3719
+    },
+    {
+      "epoch": 0.20735785953177258,
+      "grad_norm": 0.46077826619148254,
+      "learning_rate": 8e-05,
+      "loss": 1.6807,
+      "step": 3720
+    },
+    {
+      "epoch": 0.20741360089186175,
+      "grad_norm": 0.49710530042648315,
+      "learning_rate": 8e-05,
+      "loss": 1.8692,
+      "step": 3721
+    },
+    {
+      "epoch": 0.20746934225195096,
+      "grad_norm": 0.5080289244651794,
+      "learning_rate": 8e-05,
+      "loss": 1.8348,
+      "step": 3722
+    },
+    {
+      "epoch": 0.20752508361204014,
+      "grad_norm": 0.47977977991104126,
+      "learning_rate": 8e-05,
+      "loss": 1.8032,
+      "step": 3723
+    },
+    {
+      "epoch": 0.20758082497212932,
+      "grad_norm": 0.5408175587654114,
+      "learning_rate": 8e-05,
+      "loss": 1.8521,
+      "step": 3724
+    },
+    {
+      "epoch": 0.2076365663322185,
+      "grad_norm": 0.45105478167533875,
+      "learning_rate": 8e-05,
+      "loss": 1.5664,
+      "step": 3725
+    },
+    {
+      "epoch": 0.2076923076923077,
+      "grad_norm": 0.5026746988296509,
+      "learning_rate": 8e-05,
+      "loss": 1.9063,
+      "step": 3726
+    },
+    {
+      "epoch": 0.20774804905239688,
+      "grad_norm": 0.493769109249115,
+      "learning_rate": 8e-05,
+      "loss": 1.7472,
+      "step": 3727
+    },
+    {
+      "epoch": 0.20780379041248606,
+      "grad_norm": 0.4682243764400482,
+      "learning_rate": 8e-05,
+      "loss": 1.5133,
+      "step": 3728
+    },
+    {
+      "epoch": 0.20785953177257524,
+      "grad_norm": 0.5349406599998474,
+      "learning_rate": 8e-05,
+      "loss": 1.7283,
+      "step": 3729
+    },
+    {
+      "epoch": 0.20791527313266445,
+      "grad_norm": 0.46482259035110474,
+      "learning_rate": 8e-05,
+      "loss": 1.3779,
+      "step": 3730
+    },
+    {
+      "epoch": 0.20797101449275363,
+      "grad_norm": 0.5020154714584351,
+      "learning_rate": 8e-05,
+      "loss": 1.8065,
+      "step": 3731
+    },
+    {
+      "epoch": 0.2080267558528428,
+      "grad_norm": 0.5125617980957031,
+      "learning_rate": 8e-05,
+      "loss": 1.818,
+      "step": 3732
+    },
+    {
+      "epoch": 0.20808249721293198,
+      "grad_norm": 0.4770820736885071,
+      "learning_rate": 8e-05,
+      "loss": 1.4732,
+      "step": 3733
+    },
+    {
+      "epoch": 0.2081382385730212,
+      "grad_norm": 0.5883281230926514,
+      "learning_rate": 8e-05,
+      "loss": 2.0237,
+      "step": 3734
+    },
+    {
+      "epoch": 0.20819397993311037,
+      "grad_norm": 0.5591292977333069,
+      "learning_rate": 8e-05,
+      "loss": 1.7542,
+      "step": 3735
+    },
+    {
+      "epoch": 0.20824972129319955,
+      "grad_norm": 0.5060874223709106,
+      "learning_rate": 8e-05,
+      "loss": 1.866,
+      "step": 3736
+    },
+    {
+      "epoch": 0.20830546265328875,
+      "grad_norm": 0.5743696093559265,
+      "learning_rate": 8e-05,
+      "loss": 1.8753,
+      "step": 3737
+    },
+    {
+      "epoch": 0.20836120401337793,
+      "grad_norm": 0.48902419209480286,
+      "learning_rate": 8e-05,
+      "loss": 1.6463,
+      "step": 3738
+    },
+    {
+      "epoch": 0.2084169453734671,
+      "grad_norm": 0.5021654367446899,
+      "learning_rate": 8e-05,
+      "loss": 1.5628,
+      "step": 3739
+    },
+    {
+      "epoch": 0.2084726867335563,
+      "grad_norm": 0.5163289308547974,
+      "learning_rate": 8e-05,
+      "loss": 1.8731,
+      "step": 3740
+    },
+    {
+      "epoch": 0.2085284280936455,
+      "grad_norm": 0.4969714283943176,
+      "learning_rate": 8e-05,
+      "loss": 1.5473,
+      "step": 3741
+    },
+    {
+      "epoch": 0.20858416945373467,
+      "grad_norm": 0.4648710787296295,
+      "learning_rate": 8e-05,
+      "loss": 1.4971,
+      "step": 3742
+    },
+    {
+      "epoch": 0.20863991081382385,
+      "grad_norm": 0.5347151160240173,
+      "learning_rate": 8e-05,
+      "loss": 1.7845,
+      "step": 3743
+    },
+    {
+      "epoch": 0.20869565217391303,
+      "grad_norm": 0.4412599503993988,
+      "learning_rate": 8e-05,
+      "loss": 1.4139,
+      "step": 3744
+    },
+    {
+      "epoch": 0.20875139353400224,
+      "grad_norm": 0.49814075231552124,
+      "learning_rate": 8e-05,
+      "loss": 1.7027,
+      "step": 3745
+    },
+    {
+      "epoch": 0.20880713489409142,
+      "grad_norm": 0.5593386888504028,
+      "learning_rate": 8e-05,
+      "loss": 1.7623,
+      "step": 3746
+    },
+    {
+      "epoch": 0.2088628762541806,
+      "grad_norm": 0.5205774307250977,
+      "learning_rate": 8e-05,
+      "loss": 1.8267,
+      "step": 3747
+    },
+    {
+      "epoch": 0.20891861761426977,
+      "grad_norm": 0.5357809662818909,
+      "learning_rate": 8e-05,
+      "loss": 1.8368,
+      "step": 3748
+    },
+    {
+      "epoch": 0.20897435897435898,
+      "grad_norm": 0.530596911907196,
+      "learning_rate": 8e-05,
+      "loss": 1.5877,
+      "step": 3749
+    },
+    {
+      "epoch": 0.20903010033444816,
+      "grad_norm": 0.5513470768928528,
+      "learning_rate": 8e-05,
+      "loss": 1.9045,
+      "step": 3750
+    },
+    {
+      "epoch": 0.20908584169453734,
+      "grad_norm": 0.5307799577713013,
+      "learning_rate": 8e-05,
+      "loss": 1.8905,
+      "step": 3751
+    },
+    {
+      "epoch": 0.20914158305462655,
+      "grad_norm": 0.45953071117401123,
+      "learning_rate": 8e-05,
+      "loss": 1.6001,
+      "step": 3752
+    },
+    {
+      "epoch": 0.20919732441471572,
+      "grad_norm": 0.5116471648216248,
+      "learning_rate": 8e-05,
+      "loss": 1.4961,
+      "step": 3753
+    },
+    {
+      "epoch": 0.2092530657748049,
+      "grad_norm": 0.4981320798397064,
+      "learning_rate": 8e-05,
+      "loss": 1.6046,
+      "step": 3754
+    },
+    {
+      "epoch": 0.20930880713489408,
+      "grad_norm": 0.5089046955108643,
+      "learning_rate": 8e-05,
+      "loss": 1.7513,
+      "step": 3755
+    },
+    {
+      "epoch": 0.2093645484949833,
+      "grad_norm": 0.6006644368171692,
+      "learning_rate": 8e-05,
+      "loss": 1.914,
+      "step": 3756
+    },
+    {
+      "epoch": 0.20942028985507247,
+      "grad_norm": 0.534820020198822,
+      "learning_rate": 8e-05,
+      "loss": 1.6961,
+      "step": 3757
+    },
+    {
+      "epoch": 0.20947603121516165,
+      "grad_norm": 0.499206006526947,
+      "learning_rate": 8e-05,
+      "loss": 1.7807,
+      "step": 3758
+    },
+    {
+      "epoch": 0.20953177257525082,
+      "grad_norm": 0.47656241059303284,
+      "learning_rate": 8e-05,
+      "loss": 1.6846,
+      "step": 3759
+    },
+    {
+      "epoch": 0.20958751393534003,
+      "grad_norm": 0.4986822307109833,
+      "learning_rate": 8e-05,
+      "loss": 1.685,
+      "step": 3760
+    },
+    {
+      "epoch": 0.2096432552954292,
+      "grad_norm": 0.5661086440086365,
+      "learning_rate": 8e-05,
+      "loss": 1.8754,
+      "step": 3761
+    },
+    {
+      "epoch": 0.2096989966555184,
+      "grad_norm": 0.47916561365127563,
+      "learning_rate": 8e-05,
+      "loss": 1.6673,
+      "step": 3762
+    },
+    {
+      "epoch": 0.2097547380156076,
+      "grad_norm": 0.5194079875946045,
+      "learning_rate": 8e-05,
+      "loss": 1.8266,
+      "step": 3763
+    },
+    {
+      "epoch": 0.20981047937569677,
+      "grad_norm": 0.5287598967552185,
+      "learning_rate": 8e-05,
+      "loss": 1.8464,
+      "step": 3764
+    },
+    {
+      "epoch": 0.20986622073578595,
+      "grad_norm": 0.5045037269592285,
+      "learning_rate": 8e-05,
+      "loss": 1.8192,
+      "step": 3765
+    },
+    {
+      "epoch": 0.20992196209587513,
+      "grad_norm": 0.4872746169567108,
+      "learning_rate": 8e-05,
+      "loss": 1.6454,
+      "step": 3766
+    },
+    {
+      "epoch": 0.20997770345596434,
+      "grad_norm": 0.5383753180503845,
+      "learning_rate": 8e-05,
+      "loss": 2.1398,
+      "step": 3767
+    },
+    {
+      "epoch": 0.21003344481605352,
+      "grad_norm": 0.5127830505371094,
+      "learning_rate": 8e-05,
+      "loss": 1.6806,
+      "step": 3768
+    },
+    {
+      "epoch": 0.2100891861761427,
+      "grad_norm": 0.4865786135196686,
+      "learning_rate": 8e-05,
+      "loss": 1.8012,
+      "step": 3769
+    },
+    {
+      "epoch": 0.21014492753623187,
+      "grad_norm": 0.5016859173774719,
+      "learning_rate": 8e-05,
+      "loss": 1.7015,
+      "step": 3770
+    },
+    {
+      "epoch": 0.21020066889632108,
+      "grad_norm": 0.5375359654426575,
+      "learning_rate": 8e-05,
+      "loss": 2.0587,
+      "step": 3771
+    },
+    {
+      "epoch": 0.21025641025641026,
+      "grad_norm": 0.5184040665626526,
+      "learning_rate": 8e-05,
+      "loss": 1.7752,
+      "step": 3772
+    },
+    {
+      "epoch": 0.21031215161649944,
+      "grad_norm": 0.5094031095504761,
+      "learning_rate": 8e-05,
+      "loss": 1.7054,
+      "step": 3773
+    },
+    {
+      "epoch": 0.21036789297658862,
+      "grad_norm": 0.4853542745113373,
+      "learning_rate": 8e-05,
+      "loss": 1.7475,
+      "step": 3774
+    },
+    {
+      "epoch": 0.21042363433667782,
+      "grad_norm": 0.45781707763671875,
+      "learning_rate": 8e-05,
+      "loss": 1.7281,
+      "step": 3775
+    },
+    {
+      "epoch": 0.210479375696767,
+      "grad_norm": 0.48402369022369385,
+      "learning_rate": 8e-05,
+      "loss": 1.7368,
+      "step": 3776
+    },
+    {
+      "epoch": 0.21053511705685618,
+      "grad_norm": 0.46384137868881226,
+      "learning_rate": 8e-05,
+      "loss": 1.6851,
+      "step": 3777
+    },
+    {
+      "epoch": 0.2105908584169454,
+      "grad_norm": 0.5177857875823975,
+      "learning_rate": 8e-05,
+      "loss": 1.7237,
+      "step": 3778
+    },
+    {
+      "epoch": 0.21064659977703457,
+      "grad_norm": 0.5443496108055115,
+      "learning_rate": 8e-05,
+      "loss": 1.7854,
+      "step": 3779
+    },
+    {
+      "epoch": 0.21070234113712374,
+      "grad_norm": 0.5036930441856384,
+      "learning_rate": 8e-05,
+      "loss": 1.7195,
+      "step": 3780
+    },
+    {
+      "epoch": 0.21075808249721292,
+      "grad_norm": 0.5116427540779114,
+      "learning_rate": 8e-05,
+      "loss": 1.6763,
+      "step": 3781
+    },
+    {
+      "epoch": 0.21081382385730213,
+      "grad_norm": 0.5227110981941223,
+      "learning_rate": 8e-05,
+      "loss": 1.8153,
+      "step": 3782
+    },
+    {
+      "epoch": 0.2108695652173913,
+      "grad_norm": 0.49027928709983826,
+      "learning_rate": 8e-05,
+      "loss": 1.7977,
+      "step": 3783
+    },
+    {
+      "epoch": 0.2109253065774805,
+      "grad_norm": 0.4768216609954834,
+      "learning_rate": 8e-05,
+      "loss": 1.6186,
+      "step": 3784
+    },
+    {
+      "epoch": 0.21098104793756967,
+      "grad_norm": 0.47161638736724854,
+      "learning_rate": 8e-05,
+      "loss": 1.7054,
+      "step": 3785
+    },
+    {
+      "epoch": 0.21103678929765887,
+      "grad_norm": 0.46826526522636414,
+      "learning_rate": 8e-05,
+      "loss": 1.5809,
+      "step": 3786
+    },
+    {
+      "epoch": 0.21109253065774805,
+      "grad_norm": 0.5086841583251953,
+      "learning_rate": 8e-05,
+      "loss": 1.7624,
+      "step": 3787
+    },
+    {
+      "epoch": 0.21114827201783723,
+      "grad_norm": 0.5066002607345581,
+      "learning_rate": 8e-05,
+      "loss": 2.0431,
+      "step": 3788
+    },
+    {
+      "epoch": 0.2112040133779264,
+      "grad_norm": 0.49868473410606384,
+      "learning_rate": 8e-05,
+      "loss": 1.7682,
+      "step": 3789
+    },
+    {
+      "epoch": 0.21125975473801561,
+      "grad_norm": 0.5334035158157349,
+      "learning_rate": 8e-05,
+      "loss": 1.8199,
+      "step": 3790
+    },
+    {
+      "epoch": 0.2113154960981048,
+      "grad_norm": 0.5027373433113098,
+      "learning_rate": 8e-05,
+      "loss": 1.7154,
+      "step": 3791
+    },
+    {
+      "epoch": 0.21137123745819397,
+      "grad_norm": 0.48657071590423584,
+      "learning_rate": 8e-05,
+      "loss": 1.6964,
+      "step": 3792
+    },
+    {
+      "epoch": 0.21142697881828318,
+      "grad_norm": 0.4769059717655182,
+      "learning_rate": 8e-05,
+      "loss": 1.4626,
+      "step": 3793
+    },
+    {
+      "epoch": 0.21148272017837236,
+      "grad_norm": 0.47266313433647156,
+      "learning_rate": 8e-05,
+      "loss": 1.7482,
+      "step": 3794
+    },
+    {
+      "epoch": 0.21153846153846154,
+      "grad_norm": 0.5105955004692078,
+      "learning_rate": 8e-05,
+      "loss": 1.7192,
+      "step": 3795
+    },
+    {
+      "epoch": 0.21159420289855072,
+      "grad_norm": 0.474439799785614,
+      "learning_rate": 8e-05,
+      "loss": 1.6764,
+      "step": 3796
+    },
+    {
+      "epoch": 0.21164994425863992,
+      "grad_norm": 0.5234792232513428,
+      "learning_rate": 8e-05,
+      "loss": 1.9227,
+      "step": 3797
+    },
+    {
+      "epoch": 0.2117056856187291,
+      "grad_norm": 0.47840628027915955,
+      "learning_rate": 8e-05,
+      "loss": 1.7192,
+      "step": 3798
+    },
+    {
+      "epoch": 0.21176142697881828,
+      "grad_norm": 0.5176959037780762,
+      "learning_rate": 8e-05,
+      "loss": 1.8396,
+      "step": 3799
+    },
+    {
+      "epoch": 0.21181716833890746,
+      "grad_norm": 0.5095927119255066,
+      "learning_rate": 8e-05,
+      "loss": 1.7745,
+      "step": 3800
+    },
+    {
+      "epoch": 0.21187290969899666,
+      "grad_norm": 0.47420892119407654,
+      "learning_rate": 8e-05,
+      "loss": 1.5153,
+      "step": 3801
+    },
+    {
+      "epoch": 0.21192865105908584,
+      "grad_norm": 0.5116045475006104,
+      "learning_rate": 8e-05,
+      "loss": 1.7009,
+      "step": 3802
+    },
+    {
+      "epoch": 0.21198439241917502,
+      "grad_norm": 0.4922579228878021,
+      "learning_rate": 8e-05,
+      "loss": 1.8566,
+      "step": 3803
+    },
+    {
+      "epoch": 0.2120401337792642,
+      "grad_norm": 0.46212583780288696,
+      "learning_rate": 8e-05,
+      "loss": 1.5053,
+      "step": 3804
+    },
+    {
+      "epoch": 0.2120958751393534,
+      "grad_norm": 0.45981431007385254,
+      "learning_rate": 8e-05,
+      "loss": 1.6856,
+      "step": 3805
+    },
+    {
+      "epoch": 0.21215161649944259,
+      "grad_norm": 0.48287737369537354,
+      "learning_rate": 8e-05,
+      "loss": 1.7064,
+      "step": 3806
+    },
+    {
+      "epoch": 0.21220735785953176,
+      "grad_norm": 0.5647776126861572,
+      "learning_rate": 8e-05,
+      "loss": 2.1445,
+      "step": 3807
+    },
+    {
+      "epoch": 0.21226309921962097,
+      "grad_norm": 0.5176706314086914,
+      "learning_rate": 8e-05,
+      "loss": 1.7832,
+      "step": 3808
+    },
+    {
+      "epoch": 0.21231884057971015,
+      "grad_norm": 0.4690020978450775,
+      "learning_rate": 8e-05,
+      "loss": 1.7234,
+      "step": 3809
+    },
+    {
+      "epoch": 0.21237458193979933,
+      "grad_norm": 0.5125473141670227,
+      "learning_rate": 8e-05,
+      "loss": 1.616,
+      "step": 3810
+    },
+    {
+      "epoch": 0.2124303232998885,
+      "grad_norm": 0.5069548487663269,
+      "learning_rate": 8e-05,
+      "loss": 1.7209,
+      "step": 3811
+    },
+    {
+      "epoch": 0.2124860646599777,
+      "grad_norm": 0.4834950268268585,
+      "learning_rate": 8e-05,
+      "loss": 1.7798,
+      "step": 3812
+    },
+    {
+      "epoch": 0.2125418060200669,
+      "grad_norm": 0.477280855178833,
+      "learning_rate": 8e-05,
+      "loss": 1.7212,
+      "step": 3813
+    },
+    {
+      "epoch": 0.21259754738015607,
+      "grad_norm": 0.47901955246925354,
+      "learning_rate": 8e-05,
+      "loss": 1.8597,
+      "step": 3814
+    },
+    {
+      "epoch": 0.21265328874024525,
+      "grad_norm": 0.4886859953403473,
+      "learning_rate": 8e-05,
+      "loss": 1.6653,
+      "step": 3815
+    },
+    {
+      "epoch": 0.21270903010033446,
+      "grad_norm": 0.5218357443809509,
+      "learning_rate": 8e-05,
+      "loss": 1.5118,
+      "step": 3816
+    },
+    {
+      "epoch": 0.21276477146042364,
+      "grad_norm": 0.4620465934276581,
+      "learning_rate": 8e-05,
+      "loss": 1.6675,
+      "step": 3817
+    },
+    {
+      "epoch": 0.2128205128205128,
+      "grad_norm": 0.4797252416610718,
+      "learning_rate": 8e-05,
+      "loss": 1.8343,
+      "step": 3818
+    },
+    {
+      "epoch": 0.21287625418060202,
+      "grad_norm": 0.4706561863422394,
+      "learning_rate": 8e-05,
+      "loss": 1.6569,
+      "step": 3819
+    },
+    {
+      "epoch": 0.2129319955406912,
+      "grad_norm": 1.0401971340179443,
+      "learning_rate": 8e-05,
+      "loss": 1.8073,
+      "step": 3820
+    },
+    {
+      "epoch": 0.21298773690078038,
+      "grad_norm": 0.4873477816581726,
+      "learning_rate": 8e-05,
+      "loss": 1.7037,
+      "step": 3821
+    },
+    {
+      "epoch": 0.21304347826086956,
+      "grad_norm": 0.49408429861068726,
+      "learning_rate": 8e-05,
+      "loss": 1.7433,
+      "step": 3822
+    },
+    {
+      "epoch": 0.21309921962095876,
+      "grad_norm": 0.5126897096633911,
+      "learning_rate": 8e-05,
+      "loss": 1.7609,
+      "step": 3823
+    },
+    {
+      "epoch": 0.21315496098104794,
+      "grad_norm": 0.5350395441055298,
+      "learning_rate": 8e-05,
+      "loss": 1.8235,
+      "step": 3824
+    },
+    {
+      "epoch": 0.21321070234113712,
+      "grad_norm": 0.5210526585578918,
+      "learning_rate": 8e-05,
+      "loss": 1.8589,
+      "step": 3825
+    },
+    {
+      "epoch": 0.2132664437012263,
+      "grad_norm": 0.5119580030441284,
+      "learning_rate": 8e-05,
+      "loss": 1.9103,
+      "step": 3826
+    },
+    {
+      "epoch": 0.2133221850613155,
+      "grad_norm": 0.5356841087341309,
+      "learning_rate": 8e-05,
+      "loss": 1.9764,
+      "step": 3827
+    },
+    {
+      "epoch": 0.21337792642140468,
+      "grad_norm": 0.5229792594909668,
+      "learning_rate": 8e-05,
+      "loss": 1.9064,
+      "step": 3828
+    },
+    {
+      "epoch": 0.21343366778149386,
+      "grad_norm": 0.5162675380706787,
+      "learning_rate": 8e-05,
+      "loss": 1.8165,
+      "step": 3829
+    },
+    {
+      "epoch": 0.21348940914158304,
+      "grad_norm": 0.49807944893836975,
+      "learning_rate": 8e-05,
+      "loss": 1.7083,
+      "step": 3830
+    },
+    {
+      "epoch": 0.21354515050167225,
+      "grad_norm": 0.5183794498443604,
+      "learning_rate": 8e-05,
+      "loss": 1.5318,
+      "step": 3831
+    },
+    {
+      "epoch": 0.21360089186176143,
+      "grad_norm": 0.4966142773628235,
+      "learning_rate": 8e-05,
+      "loss": 1.7479,
+      "step": 3832
+    },
+    {
+      "epoch": 0.2136566332218506,
+      "grad_norm": 0.5268084406852722,
+      "learning_rate": 8e-05,
+      "loss": 1.8522,
+      "step": 3833
+    },
+    {
+      "epoch": 0.2137123745819398,
+      "grad_norm": 0.5724077820777893,
+      "learning_rate": 8e-05,
+      "loss": 2.0082,
+      "step": 3834
+    },
+    {
+      "epoch": 0.213768115942029,
+      "grad_norm": 0.4624224603176117,
+      "learning_rate": 8e-05,
+      "loss": 1.6463,
+      "step": 3835
+    },
+    {
+      "epoch": 0.21382385730211817,
+      "grad_norm": 0.4890481233596802,
+      "learning_rate": 8e-05,
+      "loss": 1.6399,
+      "step": 3836
+    },
+    {
+      "epoch": 0.21387959866220735,
+      "grad_norm": 0.5407193899154663,
+      "learning_rate": 8e-05,
+      "loss": 1.8587,
+      "step": 3837
+    },
+    {
+      "epoch": 0.21393534002229656,
+      "grad_norm": 0.48641809821128845,
+      "learning_rate": 8e-05,
+      "loss": 1.729,
+      "step": 3838
+    },
+    {
+      "epoch": 0.21399108138238573,
+      "grad_norm": 0.4703226387500763,
+      "learning_rate": 8e-05,
+      "loss": 1.6493,
+      "step": 3839
+    },
+    {
+      "epoch": 0.2140468227424749,
+      "grad_norm": 0.52630615234375,
+      "learning_rate": 8e-05,
+      "loss": 1.864,
+      "step": 3840
+    },
+    {
+      "epoch": 0.2141025641025641,
+      "grad_norm": 0.4883647859096527,
+      "learning_rate": 8e-05,
+      "loss": 1.8115,
+      "step": 3841
+    },
+    {
+      "epoch": 0.2141583054626533,
+      "grad_norm": 0.4749646782875061,
+      "learning_rate": 8e-05,
+      "loss": 1.5468,
+      "step": 3842
+    },
+    {
+      "epoch": 0.21421404682274248,
+      "grad_norm": 0.47765839099884033,
+      "learning_rate": 8e-05,
+      "loss": 1.7855,
+      "step": 3843
+    },
+    {
+      "epoch": 0.21426978818283166,
+      "grad_norm": 0.5097349286079407,
+      "learning_rate": 8e-05,
+      "loss": 1.8423,
+      "step": 3844
+    },
+    {
+      "epoch": 0.21432552954292083,
+      "grad_norm": 0.49306657910346985,
+      "learning_rate": 8e-05,
+      "loss": 1.785,
+      "step": 3845
+    },
+    {
+      "epoch": 0.21438127090301004,
+      "grad_norm": 0.47584104537963867,
+      "learning_rate": 8e-05,
+      "loss": 1.6606,
+      "step": 3846
+    },
+    {
+      "epoch": 0.21443701226309922,
+      "grad_norm": 0.5335642099380493,
+      "learning_rate": 8e-05,
+      "loss": 1.7793,
+      "step": 3847
+    },
+    {
+      "epoch": 0.2144927536231884,
+      "grad_norm": 0.5167551636695862,
+      "learning_rate": 8e-05,
+      "loss": 1.7556,
+      "step": 3848
+    },
+    {
+      "epoch": 0.2145484949832776,
+      "grad_norm": 0.528070867061615,
+      "learning_rate": 8e-05,
+      "loss": 1.943,
+      "step": 3849
+    },
+    {
+      "epoch": 0.21460423634336678,
+      "grad_norm": 0.5179917812347412,
+      "learning_rate": 8e-05,
+      "loss": 1.752,
+      "step": 3850
+    },
+    {
+      "epoch": 0.21465997770345596,
+      "grad_norm": 0.45794737339019775,
+      "learning_rate": 8e-05,
+      "loss": 1.6796,
+      "step": 3851
+    },
+    {
+      "epoch": 0.21471571906354514,
+      "grad_norm": 0.486577570438385,
+      "learning_rate": 8e-05,
+      "loss": 1.7582,
+      "step": 3852
+    },
+    {
+      "epoch": 0.21477146042363435,
+      "grad_norm": 0.5513952374458313,
+      "learning_rate": 8e-05,
+      "loss": 1.7578,
+      "step": 3853
+    },
+    {
+      "epoch": 0.21482720178372353,
+      "grad_norm": 0.4476759135723114,
+      "learning_rate": 8e-05,
+      "loss": 1.5636,
+      "step": 3854
+    },
+    {
+      "epoch": 0.2148829431438127,
+      "grad_norm": 0.5062263607978821,
+      "learning_rate": 8e-05,
+      "loss": 1.7727,
+      "step": 3855
+    },
+    {
+      "epoch": 0.21493868450390188,
+      "grad_norm": 0.47806316614151,
+      "learning_rate": 8e-05,
+      "loss": 1.6939,
+      "step": 3856
+    },
+    {
+      "epoch": 0.2149944258639911,
+      "grad_norm": 0.5024604201316833,
+      "learning_rate": 8e-05,
+      "loss": 1.7159,
+      "step": 3857
+    },
+    {
+      "epoch": 0.21505016722408027,
+      "grad_norm": 0.47938990592956543,
+      "learning_rate": 8e-05,
+      "loss": 1.723,
+      "step": 3858
+    },
+    {
+      "epoch": 0.21510590858416945,
+      "grad_norm": 0.5239027142524719,
+      "learning_rate": 8e-05,
+      "loss": 1.4392,
+      "step": 3859
+    },
+    {
+      "epoch": 0.21516164994425863,
+      "grad_norm": 0.4923347234725952,
+      "learning_rate": 8e-05,
+      "loss": 1.5475,
+      "step": 3860
+    },
+    {
+      "epoch": 0.21521739130434783,
+      "grad_norm": 0.5067053437232971,
+      "learning_rate": 8e-05,
+      "loss": 1.5485,
+      "step": 3861
+    },
+    {
+      "epoch": 0.215273132664437,
+      "grad_norm": 0.5009576082229614,
+      "learning_rate": 8e-05,
+      "loss": 1.6084,
+      "step": 3862
+    },
+    {
+      "epoch": 0.2153288740245262,
+      "grad_norm": 0.5134173631668091,
+      "learning_rate": 8e-05,
+      "loss": 1.8511,
+      "step": 3863
+    },
+    {
+      "epoch": 0.2153846153846154,
+      "grad_norm": 0.4962582290172577,
+      "learning_rate": 8e-05,
+      "loss": 1.7865,
+      "step": 3864
+    },
+    {
+      "epoch": 0.21544035674470458,
+      "grad_norm": 0.4840092957019806,
+      "learning_rate": 8e-05,
+      "loss": 1.6366,
+      "step": 3865
+    },
+    {
+      "epoch": 0.21549609810479375,
+      "grad_norm": 0.4821156859397888,
+      "learning_rate": 8e-05,
+      "loss": 1.6677,
+      "step": 3866
+    },
+    {
+      "epoch": 0.21555183946488293,
+      "grad_norm": 0.5027958750724792,
+      "learning_rate": 8e-05,
+      "loss": 1.6693,
+      "step": 3867
+    },
+    {
+      "epoch": 0.21560758082497214,
+      "grad_norm": 0.4980306625366211,
+      "learning_rate": 8e-05,
+      "loss": 1.8091,
+      "step": 3868
+    },
+    {
+      "epoch": 0.21566332218506132,
+      "grad_norm": 0.5521800518035889,
+      "learning_rate": 8e-05,
+      "loss": 2.0795,
+      "step": 3869
+    },
+    {
+      "epoch": 0.2157190635451505,
+      "grad_norm": 0.48513853549957275,
+      "learning_rate": 8e-05,
+      "loss": 1.7454,
+      "step": 3870
+    },
+    {
+      "epoch": 0.21577480490523968,
+      "grad_norm": 0.5031861662864685,
+      "learning_rate": 8e-05,
+      "loss": 1.7551,
+      "step": 3871
+    },
+    {
+      "epoch": 0.21583054626532888,
+      "grad_norm": 0.4629993140697479,
+      "learning_rate": 8e-05,
+      "loss": 1.7578,
+      "step": 3872
+    },
+    {
+      "epoch": 0.21588628762541806,
+      "grad_norm": 0.5550334453582764,
+      "learning_rate": 8e-05,
+      "loss": 1.7309,
+      "step": 3873
+    },
+    {
+      "epoch": 0.21594202898550724,
+      "grad_norm": 0.5278251767158508,
+      "learning_rate": 8e-05,
+      "loss": 1.8432,
+      "step": 3874
+    },
+    {
+      "epoch": 0.21599777034559645,
+      "grad_norm": 0.5503281950950623,
+      "learning_rate": 8e-05,
+      "loss": 1.9233,
+      "step": 3875
+    },
+    {
+      "epoch": 0.21605351170568562,
+      "grad_norm": 0.5020831823348999,
+      "learning_rate": 8e-05,
+      "loss": 1.9928,
+      "step": 3876
+    },
+    {
+      "epoch": 0.2161092530657748,
+      "grad_norm": 0.4613478481769562,
+      "learning_rate": 8e-05,
+      "loss": 1.4694,
+      "step": 3877
+    },
+    {
+      "epoch": 0.21616499442586398,
+      "grad_norm": 0.5243741869926453,
+      "learning_rate": 8e-05,
+      "loss": 1.8008,
+      "step": 3878
+    },
+    {
+      "epoch": 0.2162207357859532,
+      "grad_norm": 0.467386931180954,
+      "learning_rate": 8e-05,
+      "loss": 1.5784,
+      "step": 3879
+    },
+    {
+      "epoch": 0.21627647714604237,
+      "grad_norm": 0.4884064197540283,
+      "learning_rate": 8e-05,
+      "loss": 1.644,
+      "step": 3880
+    },
+    {
+      "epoch": 0.21633221850613155,
+      "grad_norm": 0.5259974002838135,
+      "learning_rate": 8e-05,
+      "loss": 1.8779,
+      "step": 3881
+    },
+    {
+      "epoch": 0.21638795986622072,
+      "grad_norm": 0.5110856294631958,
+      "learning_rate": 8e-05,
+      "loss": 1.8159,
+      "step": 3882
+    },
+    {
+      "epoch": 0.21644370122630993,
+      "grad_norm": 0.5513606071472168,
+      "learning_rate": 8e-05,
+      "loss": 2.0017,
+      "step": 3883
+    },
+    {
+      "epoch": 0.2164994425863991,
+      "grad_norm": 0.49897506833076477,
+      "learning_rate": 8e-05,
+      "loss": 1.9742,
+      "step": 3884
+    },
+    {
+      "epoch": 0.2165551839464883,
+      "grad_norm": 0.512783408164978,
+      "learning_rate": 8e-05,
+      "loss": 1.4606,
+      "step": 3885
+    },
+    {
+      "epoch": 0.21661092530657747,
+      "grad_norm": 0.47280240058898926,
+      "learning_rate": 8e-05,
+      "loss": 1.7028,
+      "step": 3886
+    },
+    {
+      "epoch": 0.21666666666666667,
+      "grad_norm": 0.5042930841445923,
+      "learning_rate": 8e-05,
+      "loss": 1.738,
+      "step": 3887
+    },
+    {
+      "epoch": 0.21672240802675585,
+      "grad_norm": 0.49630722403526306,
+      "learning_rate": 8e-05,
+      "loss": 1.7362,
+      "step": 3888
+    },
+    {
+      "epoch": 0.21677814938684503,
+      "grad_norm": 0.4888682961463928,
+      "learning_rate": 8e-05,
+      "loss": 1.735,
+      "step": 3889
+    },
+    {
+      "epoch": 0.21683389074693424,
+      "grad_norm": 0.4728870689868927,
+      "learning_rate": 8e-05,
+      "loss": 1.8203,
+      "step": 3890
+    },
+    {
+      "epoch": 0.21688963210702342,
+      "grad_norm": 0.5488735437393188,
+      "learning_rate": 8e-05,
+      "loss": 1.7903,
+      "step": 3891
+    },
+    {
+      "epoch": 0.2169453734671126,
+      "grad_norm": 0.48417264223098755,
+      "learning_rate": 8e-05,
+      "loss": 1.7028,
+      "step": 3892
+    },
+    {
+      "epoch": 0.21700111482720177,
+      "grad_norm": 0.5011885762214661,
+      "learning_rate": 8e-05,
+      "loss": 1.5675,
+      "step": 3893
+    },
+    {
+      "epoch": 0.21705685618729098,
+      "grad_norm": 0.503288745880127,
+      "learning_rate": 8e-05,
+      "loss": 1.5926,
+      "step": 3894
+    },
+    {
+      "epoch": 0.21711259754738016,
+      "grad_norm": 0.5019866824150085,
+      "learning_rate": 8e-05,
+      "loss": 1.7281,
+      "step": 3895
+    },
+    {
+      "epoch": 0.21716833890746934,
+      "grad_norm": 0.4990195631980896,
+      "learning_rate": 8e-05,
+      "loss": 1.8422,
+      "step": 3896
+    },
+    {
+      "epoch": 0.21722408026755852,
+      "grad_norm": 0.49024635553359985,
+      "learning_rate": 8e-05,
+      "loss": 2.0611,
+      "step": 3897
+    },
+    {
+      "epoch": 0.21727982162764772,
+      "grad_norm": 0.44876179099082947,
+      "learning_rate": 8e-05,
+      "loss": 1.3315,
+      "step": 3898
+    },
+    {
+      "epoch": 0.2173355629877369,
+      "grad_norm": 0.45446836948394775,
+      "learning_rate": 8e-05,
+      "loss": 1.6923,
+      "step": 3899
+    },
+    {
+      "epoch": 0.21739130434782608,
+      "grad_norm": 0.4820328950881958,
+      "learning_rate": 8e-05,
+      "loss": 1.5184,
+      "step": 3900
+    },
+    {
+      "epoch": 0.21744704570791526,
+      "grad_norm": 0.46806085109710693,
+      "learning_rate": 8e-05,
+      "loss": 1.5763,
+      "step": 3901
+    },
+    {
+      "epoch": 0.21750278706800447,
+      "grad_norm": 0.49191901087760925,
+      "learning_rate": 8e-05,
+      "loss": 1.7162,
+      "step": 3902
+    },
+    {
+      "epoch": 0.21755852842809364,
+      "grad_norm": 0.4780477285385132,
+      "learning_rate": 8e-05,
+      "loss": 1.6722,
+      "step": 3903
+    },
+    {
+      "epoch": 0.21761426978818282,
+      "grad_norm": 0.5140283107757568,
+      "learning_rate": 8e-05,
+      "loss": 1.7773,
+      "step": 3904
+    },
+    {
+      "epoch": 0.21767001114827203,
+      "grad_norm": 0.5415761470794678,
+      "learning_rate": 8e-05,
+      "loss": 1.8011,
+      "step": 3905
+    },
+    {
+      "epoch": 0.2177257525083612,
+      "grad_norm": 0.49452102184295654,
+      "learning_rate": 8e-05,
+      "loss": 1.6914,
+      "step": 3906
+    },
+    {
+      "epoch": 0.2177814938684504,
+      "grad_norm": 0.4950968623161316,
+      "learning_rate": 8e-05,
+      "loss": 1.6263,
+      "step": 3907
+    },
+    {
+      "epoch": 0.21783723522853957,
+      "grad_norm": 0.4706718325614929,
+      "learning_rate": 8e-05,
+      "loss": 1.6071,
+      "step": 3908
+    },
+    {
+      "epoch": 0.21789297658862877,
+      "grad_norm": 0.5482810139656067,
+      "learning_rate": 8e-05,
+      "loss": 1.9042,
+      "step": 3909
+    },
+    {
+      "epoch": 0.21794871794871795,
+      "grad_norm": 0.5073991417884827,
+      "learning_rate": 8e-05,
+      "loss": 1.7979,
+      "step": 3910
+    },
+    {
+      "epoch": 0.21800445930880713,
+      "grad_norm": 0.5709139108657837,
+      "learning_rate": 8e-05,
+      "loss": 1.3266,
+      "step": 3911
+    },
+    {
+      "epoch": 0.2180602006688963,
+      "grad_norm": 0.5192441344261169,
+      "learning_rate": 8e-05,
+      "loss": 1.7538,
+      "step": 3912
+    },
+    {
+      "epoch": 0.21811594202898552,
+      "grad_norm": 0.44952622056007385,
+      "learning_rate": 8e-05,
+      "loss": 1.5676,
+      "step": 3913
+    },
+    {
+      "epoch": 0.2181716833890747,
+      "grad_norm": 0.48135310411453247,
+      "learning_rate": 8e-05,
+      "loss": 1.6884,
+      "step": 3914
+    },
+    {
+      "epoch": 0.21822742474916387,
+      "grad_norm": 0.45865342020988464,
+      "learning_rate": 8e-05,
+      "loss": 1.5379,
+      "step": 3915
+    },
+    {
+      "epoch": 0.21828316610925308,
+      "grad_norm": 0.47076812386512756,
+      "learning_rate": 8e-05,
+      "loss": 1.7358,
+      "step": 3916
+    },
+    {
+      "epoch": 0.21833890746934226,
+      "grad_norm": 0.45950251817703247,
+      "learning_rate": 8e-05,
+      "loss": 1.7064,
+      "step": 3917
+    },
+    {
+      "epoch": 0.21839464882943144,
+      "grad_norm": 0.4762863516807556,
+      "learning_rate": 8e-05,
+      "loss": 1.6946,
+      "step": 3918
+    },
+    {
+      "epoch": 0.21845039018952062,
+      "grad_norm": 0.5255429148674011,
+      "learning_rate": 8e-05,
+      "loss": 1.6414,
+      "step": 3919
+    },
+    {
+      "epoch": 0.21850613154960982,
+      "grad_norm": 0.4682084321975708,
+      "learning_rate": 8e-05,
+      "loss": 1.8241,
+      "step": 3920
+    },
+    {
+      "epoch": 0.218561872909699,
+      "grad_norm": 0.4832845628261566,
+      "learning_rate": 8e-05,
+      "loss": 1.7665,
+      "step": 3921
+    },
+    {
+      "epoch": 0.21861761426978818,
+      "grad_norm": 0.4966776669025421,
+      "learning_rate": 8e-05,
+      "loss": 1.7103,
+      "step": 3922
+    },
+    {
+      "epoch": 0.21867335562987736,
+      "grad_norm": 0.5141316652297974,
+      "learning_rate": 8e-05,
+      "loss": 1.8795,
+      "step": 3923
+    },
+    {
+      "epoch": 0.21872909698996656,
+      "grad_norm": 0.493316650390625,
+      "learning_rate": 8e-05,
+      "loss": 1.665,
+      "step": 3924
+    },
+    {
+      "epoch": 0.21878483835005574,
+      "grad_norm": 0.49819833040237427,
+      "learning_rate": 8e-05,
+      "loss": 1.9585,
+      "step": 3925
+    },
+    {
+      "epoch": 0.21884057971014492,
+      "grad_norm": 0.4498457610607147,
+      "learning_rate": 8e-05,
+      "loss": 1.4625,
+      "step": 3926
+    },
+    {
+      "epoch": 0.2188963210702341,
+      "grad_norm": 0.5001556277275085,
+      "learning_rate": 8e-05,
+      "loss": 1.8738,
+      "step": 3927
+    },
+    {
+      "epoch": 0.2189520624303233,
+      "grad_norm": 0.49394869804382324,
+      "learning_rate": 8e-05,
+      "loss": 1.771,
+      "step": 3928
+    },
+    {
+      "epoch": 0.2190078037904125,
+      "grad_norm": 0.4962061941623688,
+      "learning_rate": 8e-05,
+      "loss": 1.6487,
+      "step": 3929
+    },
+    {
+      "epoch": 0.21906354515050167,
+      "grad_norm": 0.5592943429946899,
+      "learning_rate": 8e-05,
+      "loss": 1.9334,
+      "step": 3930
+    },
+    {
+      "epoch": 0.21911928651059087,
+      "grad_norm": 0.5389150381088257,
+      "learning_rate": 8e-05,
+      "loss": 1.9751,
+      "step": 3931
+    },
+    {
+      "epoch": 0.21917502787068005,
+      "grad_norm": 0.5096388459205627,
+      "learning_rate": 8e-05,
+      "loss": 1.9468,
+      "step": 3932
+    },
+    {
+      "epoch": 0.21923076923076923,
+      "grad_norm": 0.561657726764679,
+      "learning_rate": 8e-05,
+      "loss": 1.84,
+      "step": 3933
+    },
+    {
+      "epoch": 0.2192865105908584,
+      "grad_norm": 0.48407647013664246,
+      "learning_rate": 8e-05,
+      "loss": 1.5551,
+      "step": 3934
+    },
+    {
+      "epoch": 0.21934225195094761,
+      "grad_norm": 0.4679587185382843,
+      "learning_rate": 8e-05,
+      "loss": 1.6836,
+      "step": 3935
+    },
+    {
+      "epoch": 0.2193979933110368,
+      "grad_norm": 0.5219367742538452,
+      "learning_rate": 8e-05,
+      "loss": 1.7788,
+      "step": 3936
+    },
+    {
+      "epoch": 0.21945373467112597,
+      "grad_norm": 0.4916120171546936,
+      "learning_rate": 8e-05,
+      "loss": 1.6,
+      "step": 3937
+    },
+    {
+      "epoch": 0.21950947603121515,
+      "grad_norm": 0.4261520206928253,
+      "learning_rate": 8e-05,
+      "loss": 1.3857,
+      "step": 3938
+    },
+    {
+      "epoch": 0.21956521739130436,
+      "grad_norm": 0.48919540643692017,
+      "learning_rate": 8e-05,
+      "loss": 1.7356,
+      "step": 3939
+    },
+    {
+      "epoch": 0.21962095875139354,
+      "grad_norm": 0.49579474329948425,
+      "learning_rate": 8e-05,
+      "loss": 1.7985,
+      "step": 3940
+    },
+    {
+      "epoch": 0.21967670011148271,
+      "grad_norm": 0.47610533237457275,
+      "learning_rate": 8e-05,
+      "loss": 1.6159,
+      "step": 3941
+    },
+    {
+      "epoch": 0.2197324414715719,
+      "grad_norm": 0.4687567949295044,
+      "learning_rate": 8e-05,
+      "loss": 1.2385,
+      "step": 3942
+    },
+    {
+      "epoch": 0.2197881828316611,
+      "grad_norm": 0.4921078085899353,
+      "learning_rate": 8e-05,
+      "loss": 1.8218,
+      "step": 3943
+    },
+    {
+      "epoch": 0.21984392419175028,
+      "grad_norm": 0.45393243432044983,
+      "learning_rate": 8e-05,
+      "loss": 1.5897,
+      "step": 3944
+    },
+    {
+      "epoch": 0.21989966555183946,
+      "grad_norm": 0.503121018409729,
+      "learning_rate": 8e-05,
+      "loss": 1.9286,
+      "step": 3945
+    },
+    {
+      "epoch": 0.21995540691192866,
+      "grad_norm": 0.4986448585987091,
+      "learning_rate": 8e-05,
+      "loss": 1.8007,
+      "step": 3946
+    },
+    {
+      "epoch": 0.22001114827201784,
+      "grad_norm": 0.4901581406593323,
+      "learning_rate": 8e-05,
+      "loss": 1.6872,
+      "step": 3947
+    },
+    {
+      "epoch": 0.22006688963210702,
+      "grad_norm": 0.48810702562332153,
+      "learning_rate": 8e-05,
+      "loss": 1.5888,
+      "step": 3948
+    },
+    {
+      "epoch": 0.2201226309921962,
+      "grad_norm": 0.5303023457527161,
+      "learning_rate": 8e-05,
+      "loss": 1.5916,
+      "step": 3949
+    },
+    {
+      "epoch": 0.2201783723522854,
+      "grad_norm": 0.4808427095413208,
+      "learning_rate": 8e-05,
+      "loss": 1.6419,
+      "step": 3950
+    },
+    {
+      "epoch": 0.22023411371237459,
+      "grad_norm": 0.462200790643692,
+      "learning_rate": 8e-05,
+      "loss": 1.5292,
+      "step": 3951
+    },
+    {
+      "epoch": 0.22028985507246376,
+      "grad_norm": 0.5043500661849976,
+      "learning_rate": 8e-05,
+      "loss": 1.5486,
+      "step": 3952
+    },
+    {
+      "epoch": 0.22034559643255294,
+      "grad_norm": 0.5242111086845398,
+      "learning_rate": 8e-05,
+      "loss": 1.7288,
+      "step": 3953
+    },
+    {
+      "epoch": 0.22040133779264215,
+      "grad_norm": 0.47073057293891907,
+      "learning_rate": 8e-05,
+      "loss": 1.5865,
+      "step": 3954
+    },
+    {
+      "epoch": 0.22045707915273133,
+      "grad_norm": 0.4670935869216919,
+      "learning_rate": 8e-05,
+      "loss": 1.6749,
+      "step": 3955
+    },
+    {
+      "epoch": 0.2205128205128205,
+      "grad_norm": 0.49966269731521606,
+      "learning_rate": 8e-05,
+      "loss": 1.6227,
+      "step": 3956
+    },
+    {
+      "epoch": 0.22056856187290969,
+      "grad_norm": 0.5078837275505066,
+      "learning_rate": 8e-05,
+      "loss": 1.6791,
+      "step": 3957
+    },
+    {
+      "epoch": 0.2206243032329989,
+      "grad_norm": 0.4879036545753479,
+      "learning_rate": 8e-05,
+      "loss": 1.6912,
+      "step": 3958
+    },
+    {
+      "epoch": 0.22068004459308807,
+      "grad_norm": 0.4623623490333557,
+      "learning_rate": 8e-05,
+      "loss": 1.5699,
+      "step": 3959
+    },
+    {
+      "epoch": 0.22073578595317725,
+      "grad_norm": 0.42267879843711853,
+      "learning_rate": 8e-05,
+      "loss": 1.5417,
+      "step": 3960
+    },
+    {
+      "epoch": 0.22079152731326646,
+      "grad_norm": 0.5437389612197876,
+      "learning_rate": 8e-05,
+      "loss": 2.0238,
+      "step": 3961
+    },
+    {
+      "epoch": 0.22084726867335563,
+      "grad_norm": 0.5264094471931458,
+      "learning_rate": 8e-05,
+      "loss": 1.8531,
+      "step": 3962
+    },
+    {
+      "epoch": 0.2209030100334448,
+      "grad_norm": 0.531224250793457,
+      "learning_rate": 8e-05,
+      "loss": 1.9189,
+      "step": 3963
+    },
+    {
+      "epoch": 0.220958751393534,
+      "grad_norm": 0.4769847095012665,
+      "learning_rate": 8e-05,
+      "loss": 1.6585,
+      "step": 3964
+    },
+    {
+      "epoch": 0.2210144927536232,
+      "grad_norm": 0.4706733822822571,
+      "learning_rate": 8e-05,
+      "loss": 1.5896,
+      "step": 3965
+    },
+    {
+      "epoch": 0.22107023411371238,
+      "grad_norm": 0.46875301003456116,
+      "learning_rate": 8e-05,
+      "loss": 1.7306,
+      "step": 3966
+    },
+    {
+      "epoch": 0.22112597547380156,
+      "grad_norm": 0.5377287268638611,
+      "learning_rate": 8e-05,
+      "loss": 1.7412,
+      "step": 3967
+    },
+    {
+      "epoch": 0.22118171683389073,
+      "grad_norm": 0.5149115920066833,
+      "learning_rate": 8e-05,
+      "loss": 1.7949,
+      "step": 3968
+    },
+    {
+      "epoch": 0.22123745819397994,
+      "grad_norm": 0.47132450342178345,
+      "learning_rate": 8e-05,
+      "loss": 1.4499,
+      "step": 3969
+    },
+    {
+      "epoch": 0.22129319955406912,
+      "grad_norm": 0.46263396739959717,
+      "learning_rate": 8e-05,
+      "loss": 1.5182,
+      "step": 3970
+    },
+    {
+      "epoch": 0.2213489409141583,
+      "grad_norm": 0.49939507246017456,
+      "learning_rate": 8e-05,
+      "loss": 1.7148,
+      "step": 3971
+    },
+    {
+      "epoch": 0.2214046822742475,
+      "grad_norm": 0.4451838433742523,
+      "learning_rate": 8e-05,
+      "loss": 1.5621,
+      "step": 3972
+    },
+    {
+      "epoch": 0.22146042363433668,
+      "grad_norm": 0.5174611210823059,
+      "learning_rate": 8e-05,
+      "loss": 1.6537,
+      "step": 3973
+    },
+    {
+      "epoch": 0.22151616499442586,
+      "grad_norm": 0.5026767253875732,
+      "learning_rate": 8e-05,
+      "loss": 1.7238,
+      "step": 3974
+    },
+    {
+      "epoch": 0.22157190635451504,
+      "grad_norm": 0.5344337821006775,
+      "learning_rate": 8e-05,
+      "loss": 1.7551,
+      "step": 3975
+    },
+    {
+      "epoch": 0.22162764771460425,
+      "grad_norm": 0.48642006516456604,
+      "learning_rate": 8e-05,
+      "loss": 1.6268,
+      "step": 3976
+    },
+    {
+      "epoch": 0.22168338907469343,
+      "grad_norm": 0.4529288113117218,
+      "learning_rate": 8e-05,
+      "loss": 1.6872,
+      "step": 3977
+    },
+    {
+      "epoch": 0.2217391304347826,
+      "grad_norm": 0.46826764941215515,
+      "learning_rate": 8e-05,
+      "loss": 1.5043,
+      "step": 3978
+    },
+    {
+      "epoch": 0.22179487179487178,
+      "grad_norm": 0.467683881521225,
+      "learning_rate": 8e-05,
+      "loss": 1.6435,
+      "step": 3979
+    },
+    {
+      "epoch": 0.221850613154961,
+      "grad_norm": 0.4558468461036682,
+      "learning_rate": 8e-05,
+      "loss": 1.4534,
+      "step": 3980
+    },
+    {
+      "epoch": 0.22190635451505017,
+      "grad_norm": 0.46639105677604675,
+      "learning_rate": 8e-05,
+      "loss": 1.8479,
+      "step": 3981
+    },
+    {
+      "epoch": 0.22196209587513935,
+      "grad_norm": 0.45973077416419983,
+      "learning_rate": 8e-05,
+      "loss": 1.7484,
+      "step": 3982
+    },
+    {
+      "epoch": 0.22201783723522853,
+      "grad_norm": 0.5095018148422241,
+      "learning_rate": 8e-05,
+      "loss": 1.7441,
+      "step": 3983
+    },
+    {
+      "epoch": 0.22207357859531773,
+      "grad_norm": 0.4564839005470276,
+      "learning_rate": 8e-05,
+      "loss": 1.5841,
+      "step": 3984
+    },
+    {
+      "epoch": 0.2221293199554069,
+      "grad_norm": 0.5045386552810669,
+      "learning_rate": 8e-05,
+      "loss": 1.9361,
+      "step": 3985
+    },
+    {
+      "epoch": 0.2221850613154961,
+      "grad_norm": 0.512732982635498,
+      "learning_rate": 8e-05,
+      "loss": 1.8255,
+      "step": 3986
+    },
+    {
+      "epoch": 0.2222408026755853,
+      "grad_norm": 0.47748759388923645,
+      "learning_rate": 8e-05,
+      "loss": 1.6961,
+      "step": 3987
+    },
+    {
+      "epoch": 0.22229654403567448,
+      "grad_norm": 0.4907570481300354,
+      "learning_rate": 8e-05,
+      "loss": 1.6522,
+      "step": 3988
+    },
+    {
+      "epoch": 0.22235228539576365,
+      "grad_norm": 0.49349287152290344,
+      "learning_rate": 8e-05,
+      "loss": 1.5232,
+      "step": 3989
+    },
+    {
+      "epoch": 0.22240802675585283,
+      "grad_norm": 0.5386398434638977,
+      "learning_rate": 8e-05,
+      "loss": 1.7628,
+      "step": 3990
+    },
+    {
+      "epoch": 0.22246376811594204,
+      "grad_norm": 0.5197376012802124,
+      "learning_rate": 8e-05,
+      "loss": 1.742,
+      "step": 3991
+    },
+    {
+      "epoch": 0.22251950947603122,
+      "grad_norm": 0.45318424701690674,
+      "learning_rate": 8e-05,
+      "loss": 1.52,
+      "step": 3992
+    },
+    {
+      "epoch": 0.2225752508361204,
+      "grad_norm": 0.4926500618457794,
+      "learning_rate": 8e-05,
+      "loss": 1.7327,
+      "step": 3993
+    },
+    {
+      "epoch": 0.22263099219620958,
+      "grad_norm": 0.4656613767147064,
+      "learning_rate": 8e-05,
+      "loss": 1.7519,
+      "step": 3994
+    },
+    {
+      "epoch": 0.22268673355629878,
+      "grad_norm": 0.48848897218704224,
+      "learning_rate": 8e-05,
+      "loss": 1.7931,
+      "step": 3995
+    },
+    {
+      "epoch": 0.22274247491638796,
+      "grad_norm": 0.4615493416786194,
+      "learning_rate": 8e-05,
+      "loss": 1.6964,
+      "step": 3996
+    },
+    {
+      "epoch": 0.22279821627647714,
+      "grad_norm": 0.48456570506095886,
+      "learning_rate": 8e-05,
+      "loss": 1.8696,
+      "step": 3997
+    },
+    {
+      "epoch": 0.22285395763656632,
+      "grad_norm": 0.492690771818161,
+      "learning_rate": 8e-05,
+      "loss": 1.8145,
+      "step": 3998
+    },
+    {
+      "epoch": 0.22290969899665553,
+      "grad_norm": 0.48878180980682373,
+      "learning_rate": 8e-05,
+      "loss": 1.666,
+      "step": 3999
+    },
+    {
+      "epoch": 0.2229654403567447,
+      "grad_norm": 0.46159079670906067,
+      "learning_rate": 8e-05,
+      "loss": 1.4359,
+      "step": 4000
+    },
+    {
+      "epoch": 0.22302118171683388,
+      "grad_norm": 0.5181668996810913,
+      "learning_rate": 8e-05,
+      "loss": 1.9858,
+      "step": 4001
+    },
+    {
+      "epoch": 0.2230769230769231,
+      "grad_norm": 0.5265150666236877,
+      "learning_rate": 8e-05,
+      "loss": 1.7475,
+      "step": 4002
+    },
+    {
+      "epoch": 0.22313266443701227,
+      "grad_norm": 0.4892853796482086,
+      "learning_rate": 8e-05,
+      "loss": 1.6459,
+      "step": 4003
+    },
+    {
+      "epoch": 0.22318840579710145,
+      "grad_norm": 0.5201447606086731,
+      "learning_rate": 8e-05,
+      "loss": 1.8574,
+      "step": 4004
+    },
+    {
+      "epoch": 0.22324414715719063,
+      "grad_norm": 0.49057894945144653,
+      "learning_rate": 8e-05,
+      "loss": 1.8647,
+      "step": 4005
+    },
+    {
+      "epoch": 0.22329988851727983,
+      "grad_norm": 0.4769148528575897,
+      "learning_rate": 8e-05,
+      "loss": 1.7238,
+      "step": 4006
+    },
+    {
+      "epoch": 0.223355629877369,
+      "grad_norm": 0.4808587431907654,
+      "learning_rate": 8e-05,
+      "loss": 1.8717,
+      "step": 4007
+    },
+    {
+      "epoch": 0.2234113712374582,
+      "grad_norm": 0.48524975776672363,
+      "learning_rate": 8e-05,
+      "loss": 1.7951,
+      "step": 4008
+    },
+    {
+      "epoch": 0.22346711259754737,
+      "grad_norm": 0.4786055088043213,
+      "learning_rate": 8e-05,
+      "loss": 1.5853,
+      "step": 4009
+    },
+    {
+      "epoch": 0.22352285395763657,
+      "grad_norm": 0.4943832755088806,
+      "learning_rate": 8e-05,
+      "loss": 1.8111,
+      "step": 4010
+    },
+    {
+      "epoch": 0.22357859531772575,
+      "grad_norm": 0.5508502125740051,
+      "learning_rate": 8e-05,
+      "loss": 1.9332,
+      "step": 4011
+    },
+    {
+      "epoch": 0.22363433667781493,
+      "grad_norm": 0.5170157551765442,
+      "learning_rate": 8e-05,
+      "loss": 1.9951,
+      "step": 4012
+    },
+    {
+      "epoch": 0.22369007803790414,
+      "grad_norm": 0.4975753128528595,
+      "learning_rate": 8e-05,
+      "loss": 1.7051,
+      "step": 4013
+    },
+    {
+      "epoch": 0.22374581939799332,
+      "grad_norm": 0.5136903524398804,
+      "learning_rate": 8e-05,
+      "loss": 1.7638,
+      "step": 4014
+    },
+    {
+      "epoch": 0.2238015607580825,
+      "grad_norm": 0.5532751083374023,
+      "learning_rate": 8e-05,
+      "loss": 1.8038,
+      "step": 4015
+    },
+    {
+      "epoch": 0.22385730211817167,
+      "grad_norm": 0.525316059589386,
+      "learning_rate": 8e-05,
+      "loss": 1.877,
+      "step": 4016
+    },
+    {
+      "epoch": 0.22391304347826088,
+      "grad_norm": 0.5056881308555603,
+      "learning_rate": 8e-05,
+      "loss": 1.6853,
+      "step": 4017
+    },
+    {
+      "epoch": 0.22396878483835006,
+      "grad_norm": 0.48879149556159973,
+      "learning_rate": 8e-05,
+      "loss": 1.7952,
+      "step": 4018
+    },
+    {
+      "epoch": 0.22402452619843924,
+      "grad_norm": 0.5147798657417297,
+      "learning_rate": 8e-05,
+      "loss": 1.945,
+      "step": 4019
+    },
+    {
+      "epoch": 0.22408026755852842,
+      "grad_norm": 0.45114102959632874,
+      "learning_rate": 8e-05,
+      "loss": 1.4624,
+      "step": 4020
+    },
+    {
+      "epoch": 0.22413600891861762,
+      "grad_norm": 0.4589650630950928,
+      "learning_rate": 8e-05,
+      "loss": 1.6146,
+      "step": 4021
+    },
+    {
+      "epoch": 0.2241917502787068,
+      "grad_norm": 0.5016673803329468,
+      "learning_rate": 8e-05,
+      "loss": 1.7754,
+      "step": 4022
+    },
+    {
+      "epoch": 0.22424749163879598,
+      "grad_norm": 0.47358280420303345,
+      "learning_rate": 8e-05,
+      "loss": 1.3906,
+      "step": 4023
+    },
+    {
+      "epoch": 0.22430323299888516,
+      "grad_norm": 0.4887877404689789,
+      "learning_rate": 8e-05,
+      "loss": 1.744,
+      "step": 4024
+    },
+    {
+      "epoch": 0.22435897435897437,
+      "grad_norm": 0.5506033301353455,
+      "learning_rate": 8e-05,
+      "loss": 1.7464,
+      "step": 4025
+    },
+    {
+      "epoch": 0.22441471571906355,
+      "grad_norm": 0.5490965247154236,
+      "learning_rate": 8e-05,
+      "loss": 1.965,
+      "step": 4026
+    },
+    {
+      "epoch": 0.22447045707915272,
+      "grad_norm": 0.4694077670574188,
+      "learning_rate": 8e-05,
+      "loss": 1.638,
+      "step": 4027
+    },
+    {
+      "epoch": 0.22452619843924193,
+      "grad_norm": 0.4980950951576233,
+      "learning_rate": 8e-05,
+      "loss": 1.7362,
+      "step": 4028
+    },
+    {
+      "epoch": 0.2245819397993311,
+      "grad_norm": 0.47829583287239075,
+      "learning_rate": 8e-05,
+      "loss": 1.4633,
+      "step": 4029
+    },
+    {
+      "epoch": 0.2246376811594203,
+      "grad_norm": 0.47786465287208557,
+      "learning_rate": 8e-05,
+      "loss": 1.7009,
+      "step": 4030
+    },
+    {
+      "epoch": 0.22469342251950947,
+      "grad_norm": 0.5337238907814026,
+      "learning_rate": 8e-05,
+      "loss": 1.9198,
+      "step": 4031
+    },
+    {
+      "epoch": 0.22474916387959867,
+      "grad_norm": 0.4835289418697357,
+      "learning_rate": 8e-05,
+      "loss": 1.7093,
+      "step": 4032
+    },
+    {
+      "epoch": 0.22480490523968785,
+      "grad_norm": 0.4613661766052246,
+      "learning_rate": 8e-05,
+      "loss": 1.5062,
+      "step": 4033
+    },
+    {
+      "epoch": 0.22486064659977703,
+      "grad_norm": 0.47875353693962097,
+      "learning_rate": 8e-05,
+      "loss": 1.5384,
+      "step": 4034
+    },
+    {
+      "epoch": 0.2249163879598662,
+      "grad_norm": 0.5129967927932739,
+      "learning_rate": 8e-05,
+      "loss": 1.8924,
+      "step": 4035
+    },
+    {
+      "epoch": 0.22497212931995542,
+      "grad_norm": 0.5353673100471497,
+      "learning_rate": 8e-05,
+      "loss": 1.6436,
+      "step": 4036
+    },
+    {
+      "epoch": 0.2250278706800446,
+      "grad_norm": 0.5585724115371704,
+      "learning_rate": 8e-05,
+      "loss": 1.9538,
+      "step": 4037
+    },
+    {
+      "epoch": 0.22508361204013377,
+      "grad_norm": 0.5115454792976379,
+      "learning_rate": 8e-05,
+      "loss": 1.7637,
+      "step": 4038
+    },
+    {
+      "epoch": 0.22513935340022295,
+      "grad_norm": 0.5760135650634766,
+      "learning_rate": 8e-05,
+      "loss": 2.0288,
+      "step": 4039
+    },
+    {
+      "epoch": 0.22519509476031216,
+      "grad_norm": 0.5313804149627686,
+      "learning_rate": 8e-05,
+      "loss": 1.6336,
+      "step": 4040
+    },
+    {
+      "epoch": 0.22525083612040134,
+      "grad_norm": 0.4843834638595581,
+      "learning_rate": 8e-05,
+      "loss": 1.7583,
+      "step": 4041
+    },
+    {
+      "epoch": 0.22530657748049052,
+      "grad_norm": 0.45977699756622314,
+      "learning_rate": 8e-05,
+      "loss": 1.6708,
+      "step": 4042
+    },
+    {
+      "epoch": 0.22536231884057972,
+      "grad_norm": 0.44780343770980835,
+      "learning_rate": 8e-05,
+      "loss": 1.5733,
+      "step": 4043
+    },
+    {
+      "epoch": 0.2254180602006689,
+      "grad_norm": 0.5017595887184143,
+      "learning_rate": 8e-05,
+      "loss": 1.7078,
+      "step": 4044
+    },
+    {
+      "epoch": 0.22547380156075808,
+      "grad_norm": 0.46054017543792725,
+      "learning_rate": 8e-05,
+      "loss": 1.8015,
+      "step": 4045
+    },
+    {
+      "epoch": 0.22552954292084726,
+      "grad_norm": 0.5294150710105896,
+      "learning_rate": 8e-05,
+      "loss": 1.8796,
+      "step": 4046
+    },
+    {
+      "epoch": 0.22558528428093647,
+      "grad_norm": 0.4661663770675659,
+      "learning_rate": 8e-05,
+      "loss": 1.8345,
+      "step": 4047
+    },
+    {
+      "epoch": 0.22564102564102564,
+      "grad_norm": 0.5371278524398804,
+      "learning_rate": 8e-05,
+      "loss": 1.7545,
+      "step": 4048
+    },
+    {
+      "epoch": 0.22569676700111482,
+      "grad_norm": 0.4482397139072418,
+      "learning_rate": 8e-05,
+      "loss": 1.5857,
+      "step": 4049
+    },
+    {
+      "epoch": 0.225752508361204,
+      "grad_norm": 0.4550190567970276,
+      "learning_rate": 8e-05,
+      "loss": 1.6947,
+      "step": 4050
+    },
+    {
+      "epoch": 0.2258082497212932,
+      "grad_norm": 0.4869641661643982,
+      "learning_rate": 8e-05,
+      "loss": 1.6888,
+      "step": 4051
+    },
+    {
+      "epoch": 0.2258639910813824,
+      "grad_norm": 0.5037266612052917,
+      "learning_rate": 8e-05,
+      "loss": 1.8179,
+      "step": 4052
+    },
+    {
+      "epoch": 0.22591973244147157,
+      "grad_norm": 0.5212584137916565,
+      "learning_rate": 8e-05,
+      "loss": 1.8177,
+      "step": 4053
+    },
+    {
+      "epoch": 0.22597547380156074,
+      "grad_norm": 0.4695870876312256,
+      "learning_rate": 8e-05,
+      "loss": 1.6418,
+      "step": 4054
+    },
+    {
+      "epoch": 0.22603121516164995,
+      "grad_norm": 0.5005946159362793,
+      "learning_rate": 8e-05,
+      "loss": 1.8869,
+      "step": 4055
+    },
+    {
+      "epoch": 0.22608695652173913,
+      "grad_norm": 0.5524263381958008,
+      "learning_rate": 8e-05,
+      "loss": 1.7914,
+      "step": 4056
+    },
+    {
+      "epoch": 0.2261426978818283,
+      "grad_norm": 0.5313774347305298,
+      "learning_rate": 8e-05,
+      "loss": 1.9922,
+      "step": 4057
+    },
+    {
+      "epoch": 0.22619843924191751,
+      "grad_norm": 0.5394706726074219,
+      "learning_rate": 8e-05,
+      "loss": 1.5808,
+      "step": 4058
+    },
+    {
+      "epoch": 0.2262541806020067,
+      "grad_norm": 0.4736047089099884,
+      "learning_rate": 8e-05,
+      "loss": 1.7312,
+      "step": 4059
+    },
+    {
+      "epoch": 0.22630992196209587,
+      "grad_norm": 0.4683007299900055,
+      "learning_rate": 8e-05,
+      "loss": 1.7285,
+      "step": 4060
+    },
+    {
+      "epoch": 0.22636566332218505,
+      "grad_norm": 0.4946795701980591,
+      "learning_rate": 8e-05,
+      "loss": 1.7556,
+      "step": 4061
+    },
+    {
+      "epoch": 0.22642140468227426,
+      "grad_norm": 0.5438491702079773,
+      "learning_rate": 8e-05,
+      "loss": 2.0871,
+      "step": 4062
+    },
+    {
+      "epoch": 0.22647714604236344,
+      "grad_norm": 0.4584484100341797,
+      "learning_rate": 8e-05,
+      "loss": 1.5073,
+      "step": 4063
+    },
+    {
+      "epoch": 0.22653288740245262,
+      "grad_norm": 0.4815159738063812,
+      "learning_rate": 8e-05,
+      "loss": 1.6304,
+      "step": 4064
+    },
+    {
+      "epoch": 0.2265886287625418,
+      "grad_norm": 0.5199421048164368,
+      "learning_rate": 8e-05,
+      "loss": 1.8117,
+      "step": 4065
+    },
+    {
+      "epoch": 0.226644370122631,
+      "grad_norm": 0.5211115479469299,
+      "learning_rate": 8e-05,
+      "loss": 1.6965,
+      "step": 4066
+    },
+    {
+      "epoch": 0.22670011148272018,
+      "grad_norm": 0.5173338651657104,
+      "learning_rate": 8e-05,
+      "loss": 1.695,
+      "step": 4067
+    },
+    {
+      "epoch": 0.22675585284280936,
+      "grad_norm": 0.49385011196136475,
+      "learning_rate": 8e-05,
+      "loss": 1.5888,
+      "step": 4068
+    },
+    {
+      "epoch": 0.22681159420289856,
+      "grad_norm": 0.5200061202049255,
+      "learning_rate": 8e-05,
+      "loss": 1.9021,
+      "step": 4069
+    },
+    {
+      "epoch": 0.22686733556298774,
+      "grad_norm": 0.5064250826835632,
+      "learning_rate": 8e-05,
+      "loss": 1.8157,
+      "step": 4070
+    },
+    {
+      "epoch": 0.22692307692307692,
+      "grad_norm": 0.49232232570648193,
+      "learning_rate": 8e-05,
+      "loss": 1.8202,
+      "step": 4071
+    },
+    {
+      "epoch": 0.2269788182831661,
+      "grad_norm": 0.5013061761856079,
+      "learning_rate": 8e-05,
+      "loss": 1.7978,
+      "step": 4072
+    },
+    {
+      "epoch": 0.2270345596432553,
+      "grad_norm": 0.4811973571777344,
+      "learning_rate": 8e-05,
+      "loss": 1.7333,
+      "step": 4073
+    },
+    {
+      "epoch": 0.22709030100334449,
+      "grad_norm": 0.5344292521476746,
+      "learning_rate": 8e-05,
+      "loss": 1.8552,
+      "step": 4074
+    },
+    {
+      "epoch": 0.22714604236343366,
+      "grad_norm": 0.49985915422439575,
+      "learning_rate": 8e-05,
+      "loss": 1.5754,
+      "step": 4075
+    },
+    {
+      "epoch": 0.22720178372352284,
+      "grad_norm": 0.5089527368545532,
+      "learning_rate": 8e-05,
+      "loss": 1.7264,
+      "step": 4076
+    },
+    {
+      "epoch": 0.22725752508361205,
+      "grad_norm": 0.49516212940216064,
+      "learning_rate": 8e-05,
+      "loss": 1.6891,
+      "step": 4077
+    },
+    {
+      "epoch": 0.22731326644370123,
+      "grad_norm": 0.49753841757774353,
+      "learning_rate": 8e-05,
+      "loss": 1.6529,
+      "step": 4078
+    },
+    {
+      "epoch": 0.2273690078037904,
+      "grad_norm": 0.5481187105178833,
+      "learning_rate": 8e-05,
+      "loss": 1.8607,
+      "step": 4079
+    },
+    {
+      "epoch": 0.22742474916387959,
+      "grad_norm": 0.47371983528137207,
+      "learning_rate": 8e-05,
+      "loss": 1.4895,
+      "step": 4080
+    },
+    {
+      "epoch": 0.2274804905239688,
+      "grad_norm": 0.5114725232124329,
+      "learning_rate": 8e-05,
+      "loss": 1.8095,
+      "step": 4081
+    },
+    {
+      "epoch": 0.22753623188405797,
+      "grad_norm": 0.4610219895839691,
+      "learning_rate": 8e-05,
+      "loss": 1.6146,
+      "step": 4082
+    },
+    {
+      "epoch": 0.22759197324414715,
+      "grad_norm": 0.5723327398300171,
+      "learning_rate": 8e-05,
+      "loss": 1.8573,
+      "step": 4083
+    },
+    {
+      "epoch": 0.22764771460423636,
+      "grad_norm": 0.4931768476963043,
+      "learning_rate": 8e-05,
+      "loss": 1.8444,
+      "step": 4084
+    },
+    {
+      "epoch": 0.22770345596432554,
+      "grad_norm": 0.504690945148468,
+      "learning_rate": 8e-05,
+      "loss": 1.6546,
+      "step": 4085
+    },
+    {
+      "epoch": 0.22775919732441471,
+      "grad_norm": 0.49038875102996826,
+      "learning_rate": 8e-05,
+      "loss": 1.6053,
+      "step": 4086
+    },
+    {
+      "epoch": 0.2278149386845039,
+      "grad_norm": 0.5072624087333679,
+      "learning_rate": 8e-05,
+      "loss": 1.6051,
+      "step": 4087
+    },
+    {
+      "epoch": 0.2278706800445931,
+      "grad_norm": 0.5469437837600708,
+      "learning_rate": 8e-05,
+      "loss": 1.826,
+      "step": 4088
+    },
+    {
+      "epoch": 0.22792642140468228,
+      "grad_norm": 0.4576040208339691,
+      "learning_rate": 8e-05,
+      "loss": 1.5462,
+      "step": 4089
+    },
+    {
+      "epoch": 0.22798216276477146,
+      "grad_norm": 0.5090271830558777,
+      "learning_rate": 8e-05,
+      "loss": 1.649,
+      "step": 4090
+    },
+    {
+      "epoch": 0.22803790412486064,
+      "grad_norm": 0.5030748248100281,
+      "learning_rate": 8e-05,
+      "loss": 1.6206,
+      "step": 4091
+    },
+    {
+      "epoch": 0.22809364548494984,
+      "grad_norm": 0.5506917238235474,
+      "learning_rate": 8e-05,
+      "loss": 1.7736,
+      "step": 4092
+    },
+    {
+      "epoch": 0.22814938684503902,
+      "grad_norm": 0.5064511299133301,
+      "learning_rate": 8e-05,
+      "loss": 1.7259,
+      "step": 4093
+    },
+    {
+      "epoch": 0.2282051282051282,
+      "grad_norm": 0.5426530241966248,
+      "learning_rate": 8e-05,
+      "loss": 1.6866,
+      "step": 4094
+    },
+    {
+      "epoch": 0.22826086956521738,
+      "grad_norm": 0.5241387486457825,
+      "learning_rate": 8e-05,
+      "loss": 1.5815,
+      "step": 4095
+    },
+    {
+      "epoch": 0.22831661092530658,
+      "grad_norm": 0.458842933177948,
+      "learning_rate": 8e-05,
+      "loss": 1.5426,
+      "step": 4096
+    },
+    {
+      "epoch": 0.22837235228539576,
+      "grad_norm": 0.4742090106010437,
+      "learning_rate": 8e-05,
+      "loss": 1.6251,
+      "step": 4097
+    },
+    {
+      "epoch": 0.22842809364548494,
+      "grad_norm": 0.47541600465774536,
+      "learning_rate": 8e-05,
+      "loss": 1.5276,
+      "step": 4098
+    },
+    {
+      "epoch": 0.22848383500557415,
+      "grad_norm": 0.4711078405380249,
+      "learning_rate": 8e-05,
+      "loss": 1.8478,
+      "step": 4099
+    },
+    {
+      "epoch": 0.22853957636566333,
+      "grad_norm": 0.474153071641922,
+      "learning_rate": 8e-05,
+      "loss": 1.7794,
+      "step": 4100
+    },
+    {
+      "epoch": 0.2285953177257525,
+      "grad_norm": 0.46833422780036926,
+      "learning_rate": 8e-05,
+      "loss": 1.6416,
+      "step": 4101
+    },
+    {
+      "epoch": 0.22865105908584168,
+      "grad_norm": 0.5075449347496033,
+      "learning_rate": 8e-05,
+      "loss": 1.4074,
+      "step": 4102
+    },
+    {
+      "epoch": 0.2287068004459309,
+      "grad_norm": 0.4985544979572296,
+      "learning_rate": 8e-05,
+      "loss": 1.491,
+      "step": 4103
+    },
+    {
+      "epoch": 0.22876254180602007,
+      "grad_norm": 0.4538790285587311,
+      "learning_rate": 8e-05,
+      "loss": 1.444,
+      "step": 4104
+    },
+    {
+      "epoch": 0.22881828316610925,
+      "grad_norm": 0.5139552354812622,
+      "learning_rate": 8e-05,
+      "loss": 1.9494,
+      "step": 4105
+    },
+    {
+      "epoch": 0.22887402452619843,
+      "grad_norm": 0.507438063621521,
+      "learning_rate": 8e-05,
+      "loss": 1.8153,
+      "step": 4106
+    },
+    {
+      "epoch": 0.22892976588628763,
+      "grad_norm": 0.5059213638305664,
+      "learning_rate": 8e-05,
+      "loss": 1.52,
+      "step": 4107
+    },
+    {
+      "epoch": 0.2289855072463768,
+      "grad_norm": 0.49112826585769653,
+      "learning_rate": 8e-05,
+      "loss": 1.8568,
+      "step": 4108
+    },
+    {
+      "epoch": 0.229041248606466,
+      "grad_norm": 0.5132395029067993,
+      "learning_rate": 8e-05,
+      "loss": 1.8098,
+      "step": 4109
+    },
+    {
+      "epoch": 0.22909698996655517,
+      "grad_norm": 0.47251254320144653,
+      "learning_rate": 8e-05,
+      "loss": 1.7584,
+      "step": 4110
+    },
+    {
+      "epoch": 0.22915273132664438,
+      "grad_norm": 0.4928188920021057,
+      "learning_rate": 8e-05,
+      "loss": 1.5482,
+      "step": 4111
+    },
+    {
+      "epoch": 0.22920847268673356,
+      "grad_norm": 0.5108476877212524,
+      "learning_rate": 8e-05,
+      "loss": 1.876,
+      "step": 4112
+    },
+    {
+      "epoch": 0.22926421404682273,
+      "grad_norm": 0.4915235936641693,
+      "learning_rate": 8e-05,
+      "loss": 1.852,
+      "step": 4113
+    },
+    {
+      "epoch": 0.22931995540691194,
+      "grad_norm": 0.5093597173690796,
+      "learning_rate": 8e-05,
+      "loss": 1.6885,
+      "step": 4114
+    },
+    {
+      "epoch": 0.22937569676700112,
+      "grad_norm": 0.4863458573818207,
+      "learning_rate": 8e-05,
+      "loss": 1.6503,
+      "step": 4115
+    },
+    {
+      "epoch": 0.2294314381270903,
+      "grad_norm": 0.5055260062217712,
+      "learning_rate": 8e-05,
+      "loss": 1.8135,
+      "step": 4116
+    },
+    {
+      "epoch": 0.22948717948717948,
+      "grad_norm": 0.48014530539512634,
+      "learning_rate": 8e-05,
+      "loss": 1.6573,
+      "step": 4117
+    },
+    {
+      "epoch": 0.22954292084726868,
+      "grad_norm": 0.5218930244445801,
+      "learning_rate": 8e-05,
+      "loss": 1.8912,
+      "step": 4118
+    },
+    {
+      "epoch": 0.22959866220735786,
+      "grad_norm": 0.5085465908050537,
+      "learning_rate": 8e-05,
+      "loss": 1.6765,
+      "step": 4119
+    },
+    {
+      "epoch": 0.22965440356744704,
+      "grad_norm": 0.46710649132728577,
+      "learning_rate": 8e-05,
+      "loss": 1.6694,
+      "step": 4120
+    },
+    {
+      "epoch": 0.22971014492753622,
+      "grad_norm": 0.4992454946041107,
+      "learning_rate": 8e-05,
+      "loss": 1.6592,
+      "step": 4121
+    },
+    {
+      "epoch": 0.22976588628762543,
+      "grad_norm": 0.526770293712616,
+      "learning_rate": 8e-05,
+      "loss": 1.738,
+      "step": 4122
+    },
+    {
+      "epoch": 0.2298216276477146,
+      "grad_norm": 0.4905931055545807,
+      "learning_rate": 8e-05,
+      "loss": 1.7881,
+      "step": 4123
+    },
+    {
+      "epoch": 0.22987736900780378,
+      "grad_norm": 0.5293172597885132,
+      "learning_rate": 8e-05,
+      "loss": 1.9267,
+      "step": 4124
+    },
+    {
+      "epoch": 0.229933110367893,
+      "grad_norm": 0.5025931000709534,
+      "learning_rate": 8e-05,
+      "loss": 1.7261,
+      "step": 4125
+    },
+    {
+      "epoch": 0.22998885172798217,
+      "grad_norm": 0.5392381548881531,
+      "learning_rate": 8e-05,
+      "loss": 1.7726,
+      "step": 4126
+    },
+    {
+      "epoch": 0.23004459308807135,
+      "grad_norm": 0.5503458976745605,
+      "learning_rate": 8e-05,
+      "loss": 1.9244,
+      "step": 4127
+    },
+    {
+      "epoch": 0.23010033444816053,
+      "grad_norm": 0.5248586535453796,
+      "learning_rate": 8e-05,
+      "loss": 1.6884,
+      "step": 4128
+    },
+    {
+      "epoch": 0.23015607580824973,
+      "grad_norm": 0.5418709516525269,
+      "learning_rate": 8e-05,
+      "loss": 1.7191,
+      "step": 4129
+    },
+    {
+      "epoch": 0.2302118171683389,
+      "grad_norm": 0.476408451795578,
+      "learning_rate": 8e-05,
+      "loss": 1.433,
+      "step": 4130
+    },
+    {
+      "epoch": 0.2302675585284281,
+      "grad_norm": 0.4656904935836792,
+      "learning_rate": 8e-05,
+      "loss": 1.6336,
+      "step": 4131
+    },
+    {
+      "epoch": 0.23032329988851727,
+      "grad_norm": 0.5517549514770508,
+      "learning_rate": 8e-05,
+      "loss": 1.6893,
+      "step": 4132
+    },
+    {
+      "epoch": 0.23037904124860648,
+      "grad_norm": 0.473292738199234,
+      "learning_rate": 8e-05,
+      "loss": 1.6653,
+      "step": 4133
+    },
+    {
+      "epoch": 0.23043478260869565,
+      "grad_norm": 0.48487353324890137,
+      "learning_rate": 8e-05,
+      "loss": 1.5773,
+      "step": 4134
+    },
+    {
+      "epoch": 0.23049052396878483,
+      "grad_norm": 0.4657806158065796,
+      "learning_rate": 8e-05,
+      "loss": 1.7153,
+      "step": 4135
+    },
+    {
+      "epoch": 0.230546265328874,
+      "grad_norm": 0.4734620153903961,
+      "learning_rate": 8e-05,
+      "loss": 1.6555,
+      "step": 4136
+    },
+    {
+      "epoch": 0.23060200668896322,
+      "grad_norm": 0.5308153629302979,
+      "learning_rate": 8e-05,
+      "loss": 1.7279,
+      "step": 4137
+    },
+    {
+      "epoch": 0.2306577480490524,
+      "grad_norm": 0.5166023969650269,
+      "learning_rate": 8e-05,
+      "loss": 1.9491,
+      "step": 4138
+    },
+    {
+      "epoch": 0.23071348940914158,
+      "grad_norm": 0.46642354130744934,
+      "learning_rate": 8e-05,
+      "loss": 1.6976,
+      "step": 4139
+    },
+    {
+      "epoch": 0.23076923076923078,
+      "grad_norm": 0.5172951817512512,
+      "learning_rate": 8e-05,
+      "loss": 1.889,
+      "step": 4140
+    },
+    {
+      "epoch": 0.23082497212931996,
+      "grad_norm": 0.48462292551994324,
+      "learning_rate": 8e-05,
+      "loss": 1.6684,
+      "step": 4141
+    },
+    {
+      "epoch": 0.23088071348940914,
+      "grad_norm": 0.5029516816139221,
+      "learning_rate": 8e-05,
+      "loss": 1.7295,
+      "step": 4142
+    },
+    {
+      "epoch": 0.23093645484949832,
+      "grad_norm": 0.4953778386116028,
+      "learning_rate": 8e-05,
+      "loss": 1.7833,
+      "step": 4143
+    },
+    {
+      "epoch": 0.23099219620958752,
+      "grad_norm": 0.5369850397109985,
+      "learning_rate": 8e-05,
+      "loss": 1.6728,
+      "step": 4144
+    },
+    {
+      "epoch": 0.2310479375696767,
+      "grad_norm": 0.45318469405174255,
+      "learning_rate": 8e-05,
+      "loss": 1.3945,
+      "step": 4145
+    },
+    {
+      "epoch": 0.23110367892976588,
+      "grad_norm": 0.4903089106082916,
+      "learning_rate": 8e-05,
+      "loss": 1.8122,
+      "step": 4146
+    },
+    {
+      "epoch": 0.23115942028985506,
+      "grad_norm": 0.48141953349113464,
+      "learning_rate": 8e-05,
+      "loss": 1.869,
+      "step": 4147
+    },
+    {
+      "epoch": 0.23121516164994427,
+      "grad_norm": 0.5402095913887024,
+      "learning_rate": 8e-05,
+      "loss": 1.9276,
+      "step": 4148
+    },
+    {
+      "epoch": 0.23127090301003345,
+      "grad_norm": 0.5884537100791931,
+      "learning_rate": 8e-05,
+      "loss": 2.0484,
+      "step": 4149
+    },
+    {
+      "epoch": 0.23132664437012262,
+      "grad_norm": 0.47551897168159485,
+      "learning_rate": 8e-05,
+      "loss": 1.5622,
+      "step": 4150
+    },
+    {
+      "epoch": 0.2313823857302118,
+      "grad_norm": 0.5315403342247009,
+      "learning_rate": 8e-05,
+      "loss": 1.9436,
+      "step": 4151
+    },
+    {
+      "epoch": 0.231438127090301,
+      "grad_norm": 0.4857591390609741,
+      "learning_rate": 8e-05,
+      "loss": 1.6029,
+      "step": 4152
+    },
+    {
+      "epoch": 0.2314938684503902,
+      "grad_norm": 0.46490103006362915,
+      "learning_rate": 8e-05,
+      "loss": 1.6518,
+      "step": 4153
+    },
+    {
+      "epoch": 0.23154960981047937,
+      "grad_norm": 0.4752138555049896,
+      "learning_rate": 8e-05,
+      "loss": 1.761,
+      "step": 4154
+    },
+    {
+      "epoch": 0.23160535117056857,
+      "grad_norm": 0.4420930743217468,
+      "learning_rate": 8e-05,
+      "loss": 1.5482,
+      "step": 4155
+    },
+    {
+      "epoch": 0.23166109253065775,
+      "grad_norm": 0.44526705145835876,
+      "learning_rate": 8e-05,
+      "loss": 1.4529,
+      "step": 4156
+    },
+    {
+      "epoch": 0.23171683389074693,
+      "grad_norm": 0.5382043719291687,
+      "learning_rate": 8e-05,
+      "loss": 1.729,
+      "step": 4157
+    },
+    {
+      "epoch": 0.2317725752508361,
+      "grad_norm": 0.5094907283782959,
+      "learning_rate": 8e-05,
+      "loss": 1.8522,
+      "step": 4158
+    },
+    {
+      "epoch": 0.23182831661092532,
+      "grad_norm": 0.48272833228111267,
+      "learning_rate": 8e-05,
+      "loss": 1.8283,
+      "step": 4159
+    },
+    {
+      "epoch": 0.2318840579710145,
+      "grad_norm": 0.47672557830810547,
+      "learning_rate": 8e-05,
+      "loss": 1.6897,
+      "step": 4160
+    },
+    {
+      "epoch": 0.23193979933110367,
+      "grad_norm": 0.5173724889755249,
+      "learning_rate": 8e-05,
+      "loss": 1.8552,
+      "step": 4161
+    },
+    {
+      "epoch": 0.23199554069119285,
+      "grad_norm": 0.5262453556060791,
+      "learning_rate": 8e-05,
+      "loss": 1.6888,
+      "step": 4162
+    },
+    {
+      "epoch": 0.23205128205128206,
+      "grad_norm": 0.5105573534965515,
+      "learning_rate": 8e-05,
+      "loss": 1.7991,
+      "step": 4163
+    },
+    {
+      "epoch": 0.23210702341137124,
+      "grad_norm": 0.5028561949729919,
+      "learning_rate": 8e-05,
+      "loss": 1.6346,
+      "step": 4164
+    },
+    {
+      "epoch": 0.23216276477146042,
+      "grad_norm": 0.4426646828651428,
+      "learning_rate": 8e-05,
+      "loss": 1.6081,
+      "step": 4165
+    },
+    {
+      "epoch": 0.23221850613154962,
+      "grad_norm": 0.5555811524391174,
+      "learning_rate": 8e-05,
+      "loss": 1.8258,
+      "step": 4166
+    },
+    {
+      "epoch": 0.2322742474916388,
+      "grad_norm": 0.47089752554893494,
+      "learning_rate": 8e-05,
+      "loss": 1.7385,
+      "step": 4167
+    },
+    {
+      "epoch": 0.23232998885172798,
+      "grad_norm": 0.48612332344055176,
+      "learning_rate": 8e-05,
+      "loss": 1.7468,
+      "step": 4168
+    },
+    {
+      "epoch": 0.23238573021181716,
+      "grad_norm": 0.4978078305721283,
+      "learning_rate": 8e-05,
+      "loss": 1.74,
+      "step": 4169
+    },
+    {
+      "epoch": 0.23244147157190637,
+      "grad_norm": 1.1125239133834839,
+      "learning_rate": 8e-05,
+      "loss": 1.83,
+      "step": 4170
+    },
+    {
+      "epoch": 0.23249721293199554,
+      "grad_norm": 0.5372199416160583,
+      "learning_rate": 8e-05,
+      "loss": 1.8458,
+      "step": 4171
+    },
+    {
+      "epoch": 0.23255295429208472,
+      "grad_norm": 0.5097193717956543,
+      "learning_rate": 8e-05,
+      "loss": 1.7344,
+      "step": 4172
+    },
+    {
+      "epoch": 0.2326086956521739,
+      "grad_norm": 0.48618727922439575,
+      "learning_rate": 8e-05,
+      "loss": 1.692,
+      "step": 4173
+    },
+    {
+      "epoch": 0.2326644370122631,
+      "grad_norm": 0.45250874757766724,
+      "learning_rate": 8e-05,
+      "loss": 1.719,
+      "step": 4174
+    },
+    {
+      "epoch": 0.2327201783723523,
+      "grad_norm": 0.4872897267341614,
+      "learning_rate": 8e-05,
+      "loss": 1.6807,
+      "step": 4175
+    },
+    {
+      "epoch": 0.23277591973244147,
+      "grad_norm": 0.4631511867046356,
+      "learning_rate": 8e-05,
+      "loss": 1.6109,
+      "step": 4176
+    },
+    {
+      "epoch": 0.23283166109253065,
+      "grad_norm": 0.5219497084617615,
+      "learning_rate": 8e-05,
+      "loss": 1.7462,
+      "step": 4177
+    },
+    {
+      "epoch": 0.23288740245261985,
+      "grad_norm": 0.4476414620876312,
+      "learning_rate": 8e-05,
+      "loss": 1.422,
+      "step": 4178
+    },
+    {
+      "epoch": 0.23294314381270903,
+      "grad_norm": 0.4813176095485687,
+      "learning_rate": 8e-05,
+      "loss": 1.6695,
+      "step": 4179
+    },
+    {
+      "epoch": 0.2329988851727982,
+      "grad_norm": 0.4781123101711273,
+      "learning_rate": 8e-05,
+      "loss": 1.4277,
+      "step": 4180
+    },
+    {
+      "epoch": 0.23305462653288742,
+      "grad_norm": 0.5305860638618469,
+      "learning_rate": 8e-05,
+      "loss": 1.9108,
+      "step": 4181
+    },
+    {
+      "epoch": 0.2331103678929766,
+      "grad_norm": 0.4862423241138458,
+      "learning_rate": 8e-05,
+      "loss": 1.5683,
+      "step": 4182
+    },
+    {
+      "epoch": 0.23316610925306577,
+      "grad_norm": 0.5205289721488953,
+      "learning_rate": 8e-05,
+      "loss": 1.8538,
+      "step": 4183
+    },
+    {
+      "epoch": 0.23322185061315495,
+      "grad_norm": 0.512668251991272,
+      "learning_rate": 8e-05,
+      "loss": 1.9319,
+      "step": 4184
+    },
+    {
+      "epoch": 0.23327759197324416,
+      "grad_norm": 0.4808613061904907,
+      "learning_rate": 8e-05,
+      "loss": 1.7602,
+      "step": 4185
+    },
+    {
+      "epoch": 0.23333333333333334,
+      "grad_norm": 0.506522536277771,
+      "learning_rate": 8e-05,
+      "loss": 1.686,
+      "step": 4186
+    },
+    {
+      "epoch": 0.23338907469342252,
+      "grad_norm": 0.4858236014842987,
+      "learning_rate": 8e-05,
+      "loss": 1.5494,
+      "step": 4187
+    },
+    {
+      "epoch": 0.2334448160535117,
+      "grad_norm": 0.5407388210296631,
+      "learning_rate": 8e-05,
+      "loss": 2.1236,
+      "step": 4188
+    },
+    {
+      "epoch": 0.2335005574136009,
+      "grad_norm": 0.47253403067588806,
+      "learning_rate": 8e-05,
+      "loss": 1.6083,
+      "step": 4189
+    },
+    {
+      "epoch": 0.23355629877369008,
+      "grad_norm": 0.4397057592868805,
+      "learning_rate": 8e-05,
+      "loss": 1.3396,
+      "step": 4190
+    },
+    {
+      "epoch": 0.23361204013377926,
+      "grad_norm": 0.4820716977119446,
+      "learning_rate": 8e-05,
+      "loss": 1.6484,
+      "step": 4191
+    },
+    {
+      "epoch": 0.23366778149386844,
+      "grad_norm": 0.4903406500816345,
+      "learning_rate": 8e-05,
+      "loss": 1.5289,
+      "step": 4192
+    },
+    {
+      "epoch": 0.23372352285395764,
+      "grad_norm": 0.49805960059165955,
+      "learning_rate": 8e-05,
+      "loss": 1.7073,
+      "step": 4193
+    },
+    {
+      "epoch": 0.23377926421404682,
+      "grad_norm": 0.44631659984588623,
+      "learning_rate": 8e-05,
+      "loss": 1.3544,
+      "step": 4194
+    },
+    {
+      "epoch": 0.233835005574136,
+      "grad_norm": 0.5055618286132812,
+      "learning_rate": 8e-05,
+      "loss": 1.8731,
+      "step": 4195
+    },
+    {
+      "epoch": 0.2338907469342252,
+      "grad_norm": 0.4882362186908722,
+      "learning_rate": 8e-05,
+      "loss": 1.6237,
+      "step": 4196
+    },
+    {
+      "epoch": 0.2339464882943144,
+      "grad_norm": 0.6776293516159058,
+      "learning_rate": 8e-05,
+      "loss": 1.8226,
+      "step": 4197
+    },
+    {
+      "epoch": 0.23400222965440357,
+      "grad_norm": 0.4881233274936676,
+      "learning_rate": 8e-05,
+      "loss": 1.7181,
+      "step": 4198
+    },
+    {
+      "epoch": 0.23405797101449274,
+      "grad_norm": 0.4813898801803589,
+      "learning_rate": 8e-05,
+      "loss": 1.5247,
+      "step": 4199
+    },
+    {
+      "epoch": 0.23411371237458195,
+      "grad_norm": 0.5439554452896118,
+      "learning_rate": 8e-05,
+      "loss": 1.8405,
+      "step": 4200
+    },
+    {
+      "epoch": 0.23416945373467113,
+      "grad_norm": 0.5290339589118958,
+      "learning_rate": 8e-05,
+      "loss": 1.6565,
+      "step": 4201
+    },
+    {
+      "epoch": 0.2342251950947603,
+      "grad_norm": 0.5204188823699951,
+      "learning_rate": 8e-05,
+      "loss": 1.8186,
+      "step": 4202
+    },
+    {
+      "epoch": 0.2342809364548495,
+      "grad_norm": 0.5180361866950989,
+      "learning_rate": 8e-05,
+      "loss": 1.7259,
+      "step": 4203
+    },
+    {
+      "epoch": 0.2343366778149387,
+      "grad_norm": 0.49083536863327026,
+      "learning_rate": 8e-05,
+      "loss": 1.5448,
+      "step": 4204
+    },
+    {
+      "epoch": 0.23439241917502787,
+      "grad_norm": 0.5805219411849976,
+      "learning_rate": 8e-05,
+      "loss": 1.8847,
+      "step": 4205
+    },
+    {
+      "epoch": 0.23444816053511705,
+      "grad_norm": 0.5516480207443237,
+      "learning_rate": 8e-05,
+      "loss": 1.9711,
+      "step": 4206
+    },
+    {
+      "epoch": 0.23450390189520623,
+      "grad_norm": 0.5432198643684387,
+      "learning_rate": 8e-05,
+      "loss": 1.7182,
+      "step": 4207
+    },
+    {
+      "epoch": 0.23455964325529544,
+      "grad_norm": 0.4973669648170471,
+      "learning_rate": 8e-05,
+      "loss": 1.7876,
+      "step": 4208
+    },
+    {
+      "epoch": 0.23461538461538461,
+      "grad_norm": 0.5608261227607727,
+      "learning_rate": 8e-05,
+      "loss": 1.932,
+      "step": 4209
+    },
+    {
+      "epoch": 0.2346711259754738,
+      "grad_norm": 0.49797818064689636,
+      "learning_rate": 8e-05,
+      "loss": 1.6952,
+      "step": 4210
+    },
+    {
+      "epoch": 0.234726867335563,
+      "grad_norm": 0.4905753433704376,
+      "learning_rate": 8e-05,
+      "loss": 1.6302,
+      "step": 4211
+    },
+    {
+      "epoch": 0.23478260869565218,
+      "grad_norm": 0.542640209197998,
+      "learning_rate": 8e-05,
+      "loss": 1.8917,
+      "step": 4212
+    },
+    {
+      "epoch": 0.23483835005574136,
+      "grad_norm": 0.5524635314941406,
+      "learning_rate": 8e-05,
+      "loss": 1.9545,
+      "step": 4213
+    },
+    {
+      "epoch": 0.23489409141583054,
+      "grad_norm": 0.47688010334968567,
+      "learning_rate": 8e-05,
+      "loss": 1.8556,
+      "step": 4214
+    },
+    {
+      "epoch": 0.23494983277591974,
+      "grad_norm": 0.5069547891616821,
+      "learning_rate": 8e-05,
+      "loss": 1.7025,
+      "step": 4215
+    },
+    {
+      "epoch": 0.23500557413600892,
+      "grad_norm": 0.5091939568519592,
+      "learning_rate": 8e-05,
+      "loss": 1.8823,
+      "step": 4216
+    },
+    {
+      "epoch": 0.2350613154960981,
+      "grad_norm": 0.46288204193115234,
+      "learning_rate": 8e-05,
+      "loss": 1.6151,
+      "step": 4217
+    },
+    {
+      "epoch": 0.23511705685618728,
+      "grad_norm": 0.502398669719696,
+      "learning_rate": 8e-05,
+      "loss": 1.9325,
+      "step": 4218
+    },
+    {
+      "epoch": 0.23517279821627649,
+      "grad_norm": 0.5230593085289001,
+      "learning_rate": 8e-05,
+      "loss": 1.684,
+      "step": 4219
+    },
+    {
+      "epoch": 0.23522853957636566,
+      "grad_norm": 0.4896896481513977,
+      "learning_rate": 8e-05,
+      "loss": 1.5547,
+      "step": 4220
+    },
+    {
+      "epoch": 0.23528428093645484,
+      "grad_norm": 0.4928077459335327,
+      "learning_rate": 8e-05,
+      "loss": 1.6465,
+      "step": 4221
+    },
+    {
+      "epoch": 0.23534002229654405,
+      "grad_norm": 0.48584556579589844,
+      "learning_rate": 8e-05,
+      "loss": 1.7758,
+      "step": 4222
+    },
+    {
+      "epoch": 0.23539576365663323,
+      "grad_norm": 0.48859500885009766,
+      "learning_rate": 8e-05,
+      "loss": 1.5394,
+      "step": 4223
+    },
+    {
+      "epoch": 0.2354515050167224,
+      "grad_norm": 0.5136126279830933,
+      "learning_rate": 8e-05,
+      "loss": 1.7458,
+      "step": 4224
+    },
+    {
+      "epoch": 0.23550724637681159,
+      "grad_norm": 0.5393681526184082,
+      "learning_rate": 8e-05,
+      "loss": 1.7832,
+      "step": 4225
+    },
+    {
+      "epoch": 0.2355629877369008,
+      "grad_norm": 0.5175178050994873,
+      "learning_rate": 8e-05,
+      "loss": 1.8735,
+      "step": 4226
+    },
+    {
+      "epoch": 0.23561872909698997,
+      "grad_norm": 0.4961029887199402,
+      "learning_rate": 8e-05,
+      "loss": 1.5048,
+      "step": 4227
+    },
+    {
+      "epoch": 0.23567447045707915,
+      "grad_norm": 0.5005576610565186,
+      "learning_rate": 8e-05,
+      "loss": 1.8192,
+      "step": 4228
+    },
+    {
+      "epoch": 0.23573021181716833,
+      "grad_norm": 0.49185097217559814,
+      "learning_rate": 8e-05,
+      "loss": 1.5498,
+      "step": 4229
+    },
+    {
+      "epoch": 0.23578595317725753,
+      "grad_norm": 0.48157772421836853,
+      "learning_rate": 8e-05,
+      "loss": 1.8148,
+      "step": 4230
+    },
+    {
+      "epoch": 0.2358416945373467,
+      "grad_norm": 0.5232310891151428,
+      "learning_rate": 8e-05,
+      "loss": 2.0586,
+      "step": 4231
+    },
+    {
+      "epoch": 0.2358974358974359,
+      "grad_norm": 0.5013419389724731,
+      "learning_rate": 8e-05,
+      "loss": 1.7316,
+      "step": 4232
+    },
+    {
+      "epoch": 0.23595317725752507,
+      "grad_norm": 0.5057625770568848,
+      "learning_rate": 8e-05,
+      "loss": 1.7509,
+      "step": 4233
+    },
+    {
+      "epoch": 0.23600891861761428,
+      "grad_norm": 0.45670923590660095,
+      "learning_rate": 8e-05,
+      "loss": 1.7861,
+      "step": 4234
+    },
+    {
+      "epoch": 0.23606465997770346,
+      "grad_norm": 0.497193306684494,
+      "learning_rate": 8e-05,
+      "loss": 1.7191,
+      "step": 4235
+    },
+    {
+      "epoch": 0.23612040133779263,
+      "grad_norm": 0.4746546745300293,
+      "learning_rate": 8e-05,
+      "loss": 1.8867,
+      "step": 4236
+    },
+    {
+      "epoch": 0.23617614269788184,
+      "grad_norm": 0.48446911573410034,
+      "learning_rate": 8e-05,
+      "loss": 1.6605,
+      "step": 4237
+    },
+    {
+      "epoch": 0.23623188405797102,
+      "grad_norm": 0.49240419268608093,
+      "learning_rate": 8e-05,
+      "loss": 1.7534,
+      "step": 4238
+    },
+    {
+      "epoch": 0.2362876254180602,
+      "grad_norm": 0.512628436088562,
+      "learning_rate": 8e-05,
+      "loss": 1.51,
+      "step": 4239
+    },
+    {
+      "epoch": 0.23634336677814938,
+      "grad_norm": 0.4205702841281891,
+      "learning_rate": 8e-05,
+      "loss": 1.3157,
+      "step": 4240
+    },
+    {
+      "epoch": 0.23639910813823858,
+      "grad_norm": 0.5130642652511597,
+      "learning_rate": 8e-05,
+      "loss": 1.7798,
+      "step": 4241
+    },
+    {
+      "epoch": 0.23645484949832776,
+      "grad_norm": 0.5004068613052368,
+      "learning_rate": 8e-05,
+      "loss": 1.694,
+      "step": 4242
+    },
+    {
+      "epoch": 0.23651059085841694,
+      "grad_norm": 0.5145587921142578,
+      "learning_rate": 8e-05,
+      "loss": 1.742,
+      "step": 4243
+    },
+    {
+      "epoch": 0.23656633221850612,
+      "grad_norm": 0.4868265688419342,
+      "learning_rate": 8e-05,
+      "loss": 1.7886,
+      "step": 4244
+    },
+    {
+      "epoch": 0.23662207357859533,
+      "grad_norm": 0.5196471214294434,
+      "learning_rate": 8e-05,
+      "loss": 1.7996,
+      "step": 4245
+    },
+    {
+      "epoch": 0.2366778149386845,
+      "grad_norm": 0.47576624155044556,
+      "learning_rate": 8e-05,
+      "loss": 1.6135,
+      "step": 4246
+    },
+    {
+      "epoch": 0.23673355629877368,
+      "grad_norm": 0.5461819767951965,
+      "learning_rate": 8e-05,
+      "loss": 1.8768,
+      "step": 4247
+    },
+    {
+      "epoch": 0.23678929765886286,
+      "grad_norm": 0.522162914276123,
+      "learning_rate": 8e-05,
+      "loss": 1.8317,
+      "step": 4248
+    },
+    {
+      "epoch": 0.23684503901895207,
+      "grad_norm": 0.5273149013519287,
+      "learning_rate": 8e-05,
+      "loss": 1.8237,
+      "step": 4249
+    },
+    {
+      "epoch": 0.23690078037904125,
+      "grad_norm": 0.4362734258174896,
+      "learning_rate": 8e-05,
+      "loss": 1.4484,
+      "step": 4250
+    },
+    {
+      "epoch": 0.23695652173913043,
+      "grad_norm": 0.46066081523895264,
+      "learning_rate": 8e-05,
+      "loss": 1.5555,
+      "step": 4251
+    },
+    {
+      "epoch": 0.23701226309921963,
+      "grad_norm": 0.5295606255531311,
+      "learning_rate": 8e-05,
+      "loss": 1.7134,
+      "step": 4252
+    },
+    {
+      "epoch": 0.2370680044593088,
+      "grad_norm": 0.5264022946357727,
+      "learning_rate": 8e-05,
+      "loss": 1.8476,
+      "step": 4253
+    },
+    {
+      "epoch": 0.237123745819398,
+      "grad_norm": 0.4991813600063324,
+      "learning_rate": 8e-05,
+      "loss": 1.8835,
+      "step": 4254
+    },
+    {
+      "epoch": 0.23717948717948717,
+      "grad_norm": 0.48692673444747925,
+      "learning_rate": 8e-05,
+      "loss": 1.6272,
+      "step": 4255
+    },
+    {
+      "epoch": 0.23723522853957638,
+      "grad_norm": 0.5100888013839722,
+      "learning_rate": 8e-05,
+      "loss": 1.8114,
+      "step": 4256
+    },
+    {
+      "epoch": 0.23729096989966555,
+      "grad_norm": 0.5449693202972412,
+      "learning_rate": 8e-05,
+      "loss": 1.89,
+      "step": 4257
+    },
+    {
+      "epoch": 0.23734671125975473,
+      "grad_norm": 0.479178786277771,
+      "learning_rate": 8e-05,
+      "loss": 1.5277,
+      "step": 4258
+    },
+    {
+      "epoch": 0.2374024526198439,
+      "grad_norm": 0.4832587242126465,
+      "learning_rate": 8e-05,
+      "loss": 1.6632,
+      "step": 4259
+    },
+    {
+      "epoch": 0.23745819397993312,
+      "grad_norm": 0.45674172043800354,
+      "learning_rate": 8e-05,
+      "loss": 1.6384,
+      "step": 4260
+    },
+    {
+      "epoch": 0.2375139353400223,
+      "grad_norm": 0.4978010654449463,
+      "learning_rate": 8e-05,
+      "loss": 1.6647,
+      "step": 4261
+    },
+    {
+      "epoch": 0.23756967670011148,
+      "grad_norm": 0.5177193880081177,
+      "learning_rate": 8e-05,
+      "loss": 1.8929,
+      "step": 4262
+    },
+    {
+      "epoch": 0.23762541806020068,
+      "grad_norm": 0.477351576089859,
+      "learning_rate": 8e-05,
+      "loss": 1.7562,
+      "step": 4263
+    },
+    {
+      "epoch": 0.23768115942028986,
+      "grad_norm": 0.5086522698402405,
+      "learning_rate": 8e-05,
+      "loss": 1.751,
+      "step": 4264
+    },
+    {
+      "epoch": 0.23773690078037904,
+      "grad_norm": 0.4827437102794647,
+      "learning_rate": 8e-05,
+      "loss": 1.6273,
+      "step": 4265
+    },
+    {
+      "epoch": 0.23779264214046822,
+      "grad_norm": 0.5645084977149963,
+      "learning_rate": 8e-05,
+      "loss": 1.8631,
+      "step": 4266
+    },
+    {
+      "epoch": 0.23784838350055743,
+      "grad_norm": 0.48042717576026917,
+      "learning_rate": 8e-05,
+      "loss": 1.5488,
+      "step": 4267
+    },
+    {
+      "epoch": 0.2379041248606466,
+      "grad_norm": 0.5094911456108093,
+      "learning_rate": 8e-05,
+      "loss": 1.8086,
+      "step": 4268
+    },
+    {
+      "epoch": 0.23795986622073578,
+      "grad_norm": 0.533108651638031,
+      "learning_rate": 8e-05,
+      "loss": 1.8615,
+      "step": 4269
+    },
+    {
+      "epoch": 0.23801560758082496,
+      "grad_norm": 0.4914109706878662,
+      "learning_rate": 8e-05,
+      "loss": 1.6501,
+      "step": 4270
+    },
+    {
+      "epoch": 0.23807134894091417,
+      "grad_norm": 0.45172661542892456,
+      "learning_rate": 8e-05,
+      "loss": 1.6222,
+      "step": 4271
+    },
+    {
+      "epoch": 0.23812709030100335,
+      "grad_norm": 0.5230309963226318,
+      "learning_rate": 8e-05,
+      "loss": 1.9301,
+      "step": 4272
+    },
+    {
+      "epoch": 0.23818283166109253,
+      "grad_norm": 0.5283030867576599,
+      "learning_rate": 8e-05,
+      "loss": 1.7618,
+      "step": 4273
+    },
+    {
+      "epoch": 0.2382385730211817,
+      "grad_norm": 0.529082179069519,
+      "learning_rate": 8e-05,
+      "loss": 1.7466,
+      "step": 4274
+    },
+    {
+      "epoch": 0.2382943143812709,
+      "grad_norm": 0.5190643072128296,
+      "learning_rate": 8e-05,
+      "loss": 1.9178,
+      "step": 4275
+    },
+    {
+      "epoch": 0.2383500557413601,
+      "grad_norm": 0.5080832839012146,
+      "learning_rate": 8e-05,
+      "loss": 1.6613,
+      "step": 4276
+    },
+    {
+      "epoch": 0.23840579710144927,
+      "grad_norm": 0.574957549571991,
+      "learning_rate": 8e-05,
+      "loss": 1.8489,
+      "step": 4277
+    },
+    {
+      "epoch": 0.23846153846153847,
+      "grad_norm": 0.5026995539665222,
+      "learning_rate": 8e-05,
+      "loss": 1.5338,
+      "step": 4278
+    },
+    {
+      "epoch": 0.23851727982162765,
+      "grad_norm": 0.5039991140365601,
+      "learning_rate": 8e-05,
+      "loss": 1.7046,
+      "step": 4279
+    },
+    {
+      "epoch": 0.23857302118171683,
+      "grad_norm": 0.44856736063957214,
+      "learning_rate": 8e-05,
+      "loss": 1.551,
+      "step": 4280
+    },
+    {
+      "epoch": 0.238628762541806,
+      "grad_norm": 0.5068076252937317,
+      "learning_rate": 8e-05,
+      "loss": 1.7688,
+      "step": 4281
+    },
+    {
+      "epoch": 0.23868450390189522,
+      "grad_norm": 0.5426245331764221,
+      "learning_rate": 8e-05,
+      "loss": 1.8139,
+      "step": 4282
+    },
+    {
+      "epoch": 0.2387402452619844,
+      "grad_norm": 0.4990077316761017,
+      "learning_rate": 8e-05,
+      "loss": 1.8424,
+      "step": 4283
+    },
+    {
+      "epoch": 0.23879598662207357,
+      "grad_norm": 0.48569685220718384,
+      "learning_rate": 8e-05,
+      "loss": 1.7252,
+      "step": 4284
+    },
+    {
+      "epoch": 0.23885172798216275,
+      "grad_norm": 0.4463791847229004,
+      "learning_rate": 8e-05,
+      "loss": 1.5896,
+      "step": 4285
+    },
+    {
+      "epoch": 0.23890746934225196,
+      "grad_norm": 0.4885939657688141,
+      "learning_rate": 8e-05,
+      "loss": 1.6246,
+      "step": 4286
+    },
+    {
+      "epoch": 0.23896321070234114,
+      "grad_norm": 0.5006809234619141,
+      "learning_rate": 8e-05,
+      "loss": 1.5473,
+      "step": 4287
+    },
+    {
+      "epoch": 0.23901895206243032,
+      "grad_norm": 0.4943895637989044,
+      "learning_rate": 8e-05,
+      "loss": 1.7773,
+      "step": 4288
+    },
+    {
+      "epoch": 0.2390746934225195,
+      "grad_norm": 0.4957701861858368,
+      "learning_rate": 8e-05,
+      "loss": 1.8107,
+      "step": 4289
+    },
+    {
+      "epoch": 0.2391304347826087,
+      "grad_norm": 0.4890899360179901,
+      "learning_rate": 8e-05,
+      "loss": 1.6751,
+      "step": 4290
+    },
+    {
+      "epoch": 0.23918617614269788,
+      "grad_norm": 0.512084424495697,
+      "learning_rate": 8e-05,
+      "loss": 1.6399,
+      "step": 4291
+    },
+    {
+      "epoch": 0.23924191750278706,
+      "grad_norm": 0.5295706391334534,
+      "learning_rate": 8e-05,
+      "loss": 1.8503,
+      "step": 4292
+    },
+    {
+      "epoch": 0.23929765886287627,
+      "grad_norm": 0.49469539523124695,
+      "learning_rate": 8e-05,
+      "loss": 1.5238,
+      "step": 4293
+    },
+    {
+      "epoch": 0.23935340022296545,
+      "grad_norm": 0.48040327429771423,
+      "learning_rate": 8e-05,
+      "loss": 1.7763,
+      "step": 4294
+    },
+    {
+      "epoch": 0.23940914158305462,
+      "grad_norm": 0.4731602668762207,
+      "learning_rate": 8e-05,
+      "loss": 1.7902,
+      "step": 4295
+    },
+    {
+      "epoch": 0.2394648829431438,
+      "grad_norm": 0.5192638635635376,
+      "learning_rate": 8e-05,
+      "loss": 1.8651,
+      "step": 4296
+    },
+    {
+      "epoch": 0.239520624303233,
+      "grad_norm": 0.5231795310974121,
+      "learning_rate": 8e-05,
+      "loss": 1.8744,
+      "step": 4297
+    },
+    {
+      "epoch": 0.2395763656633222,
+      "grad_norm": 0.5153510570526123,
+      "learning_rate": 8e-05,
+      "loss": 1.9705,
+      "step": 4298
+    },
+    {
+      "epoch": 0.23963210702341137,
+      "grad_norm": 0.48195499181747437,
+      "learning_rate": 8e-05,
+      "loss": 1.7613,
+      "step": 4299
+    },
+    {
+      "epoch": 0.23968784838350055,
+      "grad_norm": 0.5296260118484497,
+      "learning_rate": 8e-05,
+      "loss": 1.9392,
+      "step": 4300
+    },
+    {
+      "epoch": 0.23974358974358975,
+      "grad_norm": 0.4738166928291321,
+      "learning_rate": 8e-05,
+      "loss": 1.6396,
+      "step": 4301
+    },
+    {
+      "epoch": 0.23979933110367893,
+      "grad_norm": 0.4884643256664276,
+      "learning_rate": 8e-05,
+      "loss": 1.8827,
+      "step": 4302
+    },
+    {
+      "epoch": 0.2398550724637681,
+      "grad_norm": 0.4728654623031616,
+      "learning_rate": 8e-05,
+      "loss": 1.7101,
+      "step": 4303
+    },
+    {
+      "epoch": 0.2399108138238573,
+      "grad_norm": 0.4706076383590698,
+      "learning_rate": 8e-05,
+      "loss": 1.7652,
+      "step": 4304
+    },
+    {
+      "epoch": 0.2399665551839465,
+      "grad_norm": 0.4969691336154938,
+      "learning_rate": 8e-05,
+      "loss": 1.942,
+      "step": 4305
+    },
+    {
+      "epoch": 0.24002229654403567,
+      "grad_norm": 0.47131776809692383,
+      "learning_rate": 8e-05,
+      "loss": 1.725,
+      "step": 4306
+    },
+    {
+      "epoch": 0.24007803790412485,
+      "grad_norm": 0.5014018416404724,
+      "learning_rate": 8e-05,
+      "loss": 1.737,
+      "step": 4307
+    },
+    {
+      "epoch": 0.24013377926421406,
+      "grad_norm": 0.54756098985672,
+      "learning_rate": 8e-05,
+      "loss": 1.7052,
+      "step": 4308
+    },
+    {
+      "epoch": 0.24018952062430324,
+      "grad_norm": 0.45123088359832764,
+      "learning_rate": 8e-05,
+      "loss": 1.4678,
+      "step": 4309
+    },
+    {
+      "epoch": 0.24024526198439242,
+      "grad_norm": 0.4705629050731659,
+      "learning_rate": 8e-05,
+      "loss": 1.6367,
+      "step": 4310
+    },
+    {
+      "epoch": 0.2403010033444816,
+      "grad_norm": 0.4728585481643677,
+      "learning_rate": 8e-05,
+      "loss": 1.4548,
+      "step": 4311
+    },
+    {
+      "epoch": 0.2403567447045708,
+      "grad_norm": 0.4710177481174469,
+      "learning_rate": 8e-05,
+      "loss": 1.6581,
+      "step": 4312
+    },
+    {
+      "epoch": 0.24041248606465998,
+      "grad_norm": 0.5417190790176392,
+      "learning_rate": 8e-05,
+      "loss": 1.796,
+      "step": 4313
+    },
+    {
+      "epoch": 0.24046822742474916,
+      "grad_norm": 0.47374585270881653,
+      "learning_rate": 8e-05,
+      "loss": 1.6996,
+      "step": 4314
+    },
+    {
+      "epoch": 0.24052396878483834,
+      "grad_norm": 0.5024921298027039,
+      "learning_rate": 8e-05,
+      "loss": 1.4732,
+      "step": 4315
+    },
+    {
+      "epoch": 0.24057971014492754,
+      "grad_norm": 0.5105860829353333,
+      "learning_rate": 8e-05,
+      "loss": 1.8446,
+      "step": 4316
+    },
+    {
+      "epoch": 0.24063545150501672,
+      "grad_norm": 0.4570033550262451,
+      "learning_rate": 8e-05,
+      "loss": 1.614,
+      "step": 4317
+    },
+    {
+      "epoch": 0.2406911928651059,
+      "grad_norm": 0.4843344986438751,
+      "learning_rate": 8e-05,
+      "loss": 1.6607,
+      "step": 4318
+    },
+    {
+      "epoch": 0.2407469342251951,
+      "grad_norm": 0.5121851563453674,
+      "learning_rate": 8e-05,
+      "loss": 1.848,
+      "step": 4319
+    },
+    {
+      "epoch": 0.2408026755852843,
+      "grad_norm": 0.5036677122116089,
+      "learning_rate": 8e-05,
+      "loss": 1.633,
+      "step": 4320
+    },
+    {
+      "epoch": 0.24085841694537347,
+      "grad_norm": 0.497364342212677,
+      "learning_rate": 8e-05,
+      "loss": 1.778,
+      "step": 4321
+    },
+    {
+      "epoch": 0.24091415830546264,
+      "grad_norm": 0.48959243297576904,
+      "learning_rate": 8e-05,
+      "loss": 1.7664,
+      "step": 4322
+    },
+    {
+      "epoch": 0.24096989966555185,
+      "grad_norm": 0.5679601430892944,
+      "learning_rate": 8e-05,
+      "loss": 2.0067,
+      "step": 4323
+    },
+    {
+      "epoch": 0.24102564102564103,
+      "grad_norm": 0.48870909214019775,
+      "learning_rate": 8e-05,
+      "loss": 1.6651,
+      "step": 4324
+    },
+    {
+      "epoch": 0.2410813823857302,
+      "grad_norm": 0.5037427544593811,
+      "learning_rate": 8e-05,
+      "loss": 1.6614,
+      "step": 4325
+    },
+    {
+      "epoch": 0.2411371237458194,
+      "grad_norm": 0.4996836185455322,
+      "learning_rate": 8e-05,
+      "loss": 1.5968,
+      "step": 4326
+    },
+    {
+      "epoch": 0.2411928651059086,
+      "grad_norm": 0.47124508023262024,
+      "learning_rate": 8e-05,
+      "loss": 1.7196,
+      "step": 4327
+    },
+    {
+      "epoch": 0.24124860646599777,
+      "grad_norm": 0.45888665318489075,
+      "learning_rate": 8e-05,
+      "loss": 1.5474,
+      "step": 4328
+    },
+    {
+      "epoch": 0.24130434782608695,
+      "grad_norm": 0.48495161533355713,
+      "learning_rate": 8e-05,
+      "loss": 1.6149,
+      "step": 4329
+    },
+    {
+      "epoch": 0.24136008918617613,
+      "grad_norm": 0.5664768218994141,
+      "learning_rate": 8e-05,
+      "loss": 1.5396,
+      "step": 4330
+    },
+    {
+      "epoch": 0.24141583054626534,
+      "grad_norm": 0.4835517704486847,
+      "learning_rate": 8e-05,
+      "loss": 1.6441,
+      "step": 4331
+    },
+    {
+      "epoch": 0.24147157190635452,
+      "grad_norm": 0.4654442071914673,
+      "learning_rate": 8e-05,
+      "loss": 1.5344,
+      "step": 4332
+    },
+    {
+      "epoch": 0.2415273132664437,
+      "grad_norm": 0.5004253387451172,
+      "learning_rate": 8e-05,
+      "loss": 1.7656,
+      "step": 4333
+    },
+    {
+      "epoch": 0.2415830546265329,
+      "grad_norm": 0.4810848832130432,
+      "learning_rate": 8e-05,
+      "loss": 1.7206,
+      "step": 4334
+    },
+    {
+      "epoch": 0.24163879598662208,
+      "grad_norm": 0.4527301490306854,
+      "learning_rate": 8e-05,
+      "loss": 1.5351,
+      "step": 4335
+    },
+    {
+      "epoch": 0.24169453734671126,
+      "grad_norm": 0.504363477230072,
+      "learning_rate": 8e-05,
+      "loss": 1.8366,
+      "step": 4336
+    },
+    {
+      "epoch": 0.24175027870680044,
+      "grad_norm": 0.49054351449012756,
+      "learning_rate": 8e-05,
+      "loss": 1.5619,
+      "step": 4337
+    },
+    {
+      "epoch": 0.24180602006688964,
+      "grad_norm": 0.5057128071784973,
+      "learning_rate": 8e-05,
+      "loss": 1.7013,
+      "step": 4338
+    },
+    {
+      "epoch": 0.24186176142697882,
+      "grad_norm": 0.5352655649185181,
+      "learning_rate": 8e-05,
+      "loss": 1.9699,
+      "step": 4339
+    },
+    {
+      "epoch": 0.241917502787068,
+      "grad_norm": 0.5110335946083069,
+      "learning_rate": 8e-05,
+      "loss": 1.8174,
+      "step": 4340
+    },
+    {
+      "epoch": 0.24197324414715718,
+      "grad_norm": 0.4975016415119171,
+      "learning_rate": 8e-05,
+      "loss": 1.6685,
+      "step": 4341
+    },
+    {
+      "epoch": 0.24202898550724639,
+      "grad_norm": 0.4963131248950958,
+      "learning_rate": 8e-05,
+      "loss": 1.5433,
+      "step": 4342
+    },
+    {
+      "epoch": 0.24208472686733556,
+      "grad_norm": 0.5176040530204773,
+      "learning_rate": 8e-05,
+      "loss": 1.7958,
+      "step": 4343
+    },
+    {
+      "epoch": 0.24214046822742474,
+      "grad_norm": 0.49720054864883423,
+      "learning_rate": 8e-05,
+      "loss": 1.7171,
+      "step": 4344
+    },
+    {
+      "epoch": 0.24219620958751392,
+      "grad_norm": 0.4979418218135834,
+      "learning_rate": 8e-05,
+      "loss": 1.8272,
+      "step": 4345
+    },
+    {
+      "epoch": 0.24225195094760313,
+      "grad_norm": 0.4736308753490448,
+      "learning_rate": 8e-05,
+      "loss": 1.4816,
+      "step": 4346
+    },
+    {
+      "epoch": 0.2423076923076923,
+      "grad_norm": 0.48848211765289307,
+      "learning_rate": 8e-05,
+      "loss": 1.7337,
+      "step": 4347
+    },
+    {
+      "epoch": 0.24236343366778149,
+      "grad_norm": 0.4813627302646637,
+      "learning_rate": 8e-05,
+      "loss": 1.6464,
+      "step": 4348
+    },
+    {
+      "epoch": 0.2424191750278707,
+      "grad_norm": 0.5645063519477844,
+      "learning_rate": 8e-05,
+      "loss": 1.801,
+      "step": 4349
+    },
+    {
+      "epoch": 0.24247491638795987,
+      "grad_norm": 0.5068780779838562,
+      "learning_rate": 8e-05,
+      "loss": 1.6917,
+      "step": 4350
+    },
+    {
+      "epoch": 0.24253065774804905,
+      "grad_norm": 0.5032528042793274,
+      "learning_rate": 8e-05,
+      "loss": 1.8905,
+      "step": 4351
+    },
+    {
+      "epoch": 0.24258639910813823,
+      "grad_norm": 0.4853366017341614,
+      "learning_rate": 8e-05,
+      "loss": 1.6199,
+      "step": 4352
+    },
+    {
+      "epoch": 0.24264214046822744,
+      "grad_norm": 0.49789735674858093,
+      "learning_rate": 8e-05,
+      "loss": 1.6751,
+      "step": 4353
+    },
+    {
+      "epoch": 0.24269788182831661,
+      "grad_norm": 0.48519742488861084,
+      "learning_rate": 8e-05,
+      "loss": 1.6459,
+      "step": 4354
+    },
+    {
+      "epoch": 0.2427536231884058,
+      "grad_norm": 0.48001378774642944,
+      "learning_rate": 8e-05,
+      "loss": 1.5651,
+      "step": 4355
+    },
+    {
+      "epoch": 0.24280936454849497,
+      "grad_norm": 0.48040971159935,
+      "learning_rate": 8e-05,
+      "loss": 1.7485,
+      "step": 4356
+    },
+    {
+      "epoch": 0.24286510590858418,
+      "grad_norm": 0.48989880084991455,
+      "learning_rate": 8e-05,
+      "loss": 1.6495,
+      "step": 4357
+    },
+    {
+      "epoch": 0.24292084726867336,
+      "grad_norm": 0.4950062334537506,
+      "learning_rate": 8e-05,
+      "loss": 1.5592,
+      "step": 4358
+    },
+    {
+      "epoch": 0.24297658862876254,
+      "grad_norm": 0.5052546262741089,
+      "learning_rate": 8e-05,
+      "loss": 1.7054,
+      "step": 4359
+    },
+    {
+      "epoch": 0.24303232998885171,
+      "grad_norm": 0.49012768268585205,
+      "learning_rate": 8e-05,
+      "loss": 1.8105,
+      "step": 4360
+    },
+    {
+      "epoch": 0.24308807134894092,
+      "grad_norm": 0.5038552284240723,
+      "learning_rate": 8e-05,
+      "loss": 1.7451,
+      "step": 4361
+    },
+    {
+      "epoch": 0.2431438127090301,
+      "grad_norm": 0.5328361392021179,
+      "learning_rate": 8e-05,
+      "loss": 1.8975,
+      "step": 4362
+    },
+    {
+      "epoch": 0.24319955406911928,
+      "grad_norm": 0.5039812326431274,
+      "learning_rate": 8e-05,
+      "loss": 1.6608,
+      "step": 4363
+    },
+    {
+      "epoch": 0.24325529542920848,
+      "grad_norm": 0.497497022151947,
+      "learning_rate": 8e-05,
+      "loss": 1.6421,
+      "step": 4364
+    },
+    {
+      "epoch": 0.24331103678929766,
+      "grad_norm": 0.5225549340248108,
+      "learning_rate": 8e-05,
+      "loss": 1.6948,
+      "step": 4365
+    },
+    {
+      "epoch": 0.24336677814938684,
+      "grad_norm": 0.4995681643486023,
+      "learning_rate": 8e-05,
+      "loss": 1.4985,
+      "step": 4366
+    },
+    {
+      "epoch": 0.24342251950947602,
+      "grad_norm": 0.48537102341651917,
+      "learning_rate": 8e-05,
+      "loss": 1.5165,
+      "step": 4367
+    },
+    {
+      "epoch": 0.24347826086956523,
+      "grad_norm": 0.511480987071991,
+      "learning_rate": 8e-05,
+      "loss": 1.7659,
+      "step": 4368
+    },
+    {
+      "epoch": 0.2435340022296544,
+      "grad_norm": 0.4678894281387329,
+      "learning_rate": 8e-05,
+      "loss": 1.6623,
+      "step": 4369
+    },
+    {
+      "epoch": 0.24358974358974358,
+      "grad_norm": 0.4894356429576874,
+      "learning_rate": 8e-05,
+      "loss": 1.7462,
+      "step": 4370
+    },
+    {
+      "epoch": 0.24364548494983276,
+      "grad_norm": 0.5064272880554199,
+      "learning_rate": 8e-05,
+      "loss": 1.7033,
+      "step": 4371
+    },
+    {
+      "epoch": 0.24370122630992197,
+      "grad_norm": 0.5376918911933899,
+      "learning_rate": 8e-05,
+      "loss": 1.7259,
+      "step": 4372
+    },
+    {
+      "epoch": 0.24375696767001115,
+      "grad_norm": 0.5779525637626648,
+      "learning_rate": 8e-05,
+      "loss": 1.819,
+      "step": 4373
+    },
+    {
+      "epoch": 0.24381270903010033,
+      "grad_norm": 0.4810136556625366,
+      "learning_rate": 8e-05,
+      "loss": 1.5904,
+      "step": 4374
+    },
+    {
+      "epoch": 0.24386845039018953,
+      "grad_norm": 0.5844541192054749,
+      "learning_rate": 8e-05,
+      "loss": 1.7991,
+      "step": 4375
+    },
+    {
+      "epoch": 0.2439241917502787,
+      "grad_norm": 0.5359873175621033,
+      "learning_rate": 8e-05,
+      "loss": 1.8402,
+      "step": 4376
+    },
+    {
+      "epoch": 0.2439799331103679,
+      "grad_norm": 0.5614349842071533,
+      "learning_rate": 8e-05,
+      "loss": 1.6488,
+      "step": 4377
+    },
+    {
+      "epoch": 0.24403567447045707,
+      "grad_norm": 0.5228754281997681,
+      "learning_rate": 8e-05,
+      "loss": 1.7735,
+      "step": 4378
+    },
+    {
+      "epoch": 0.24409141583054628,
+      "grad_norm": 0.5935977697372437,
+      "learning_rate": 8e-05,
+      "loss": 2.014,
+      "step": 4379
+    },
+    {
+      "epoch": 0.24414715719063546,
+      "grad_norm": 0.510930597782135,
+      "learning_rate": 8e-05,
+      "loss": 1.843,
+      "step": 4380
+    },
+    {
+      "epoch": 0.24420289855072463,
+      "grad_norm": 0.4812546670436859,
+      "learning_rate": 8e-05,
+      "loss": 1.6445,
+      "step": 4381
+    },
+    {
+      "epoch": 0.2442586399108138,
+      "grad_norm": 0.49154365062713623,
+      "learning_rate": 8e-05,
+      "loss": 1.6832,
+      "step": 4382
+    },
+    {
+      "epoch": 0.24431438127090302,
+      "grad_norm": 0.4897598624229431,
+      "learning_rate": 8e-05,
+      "loss": 1.709,
+      "step": 4383
+    },
+    {
+      "epoch": 0.2443701226309922,
+      "grad_norm": 0.51563960313797,
+      "learning_rate": 8e-05,
+      "loss": 1.7739,
+      "step": 4384
+    },
+    {
+      "epoch": 0.24442586399108138,
+      "grad_norm": 0.5068482756614685,
+      "learning_rate": 8e-05,
+      "loss": 1.8145,
+      "step": 4385
+    },
+    {
+      "epoch": 0.24448160535117056,
+      "grad_norm": 0.47429415583610535,
+      "learning_rate": 8e-05,
+      "loss": 1.5825,
+      "step": 4386
+    },
+    {
+      "epoch": 0.24453734671125976,
+      "grad_norm": 0.5053471922874451,
+      "learning_rate": 8e-05,
+      "loss": 1.8622,
+      "step": 4387
+    },
+    {
+      "epoch": 0.24459308807134894,
+      "grad_norm": 0.479079008102417,
+      "learning_rate": 8e-05,
+      "loss": 1.6088,
+      "step": 4388
+    },
+    {
+      "epoch": 0.24464882943143812,
+      "grad_norm": 0.4736523926258087,
+      "learning_rate": 8e-05,
+      "loss": 1.6683,
+      "step": 4389
+    },
+    {
+      "epoch": 0.24470457079152733,
+      "grad_norm": 0.5143937468528748,
+      "learning_rate": 8e-05,
+      "loss": 1.7335,
+      "step": 4390
+    },
+    {
+      "epoch": 0.2447603121516165,
+      "grad_norm": 0.5011566877365112,
+      "learning_rate": 8e-05,
+      "loss": 1.8322,
+      "step": 4391
+    },
+    {
+      "epoch": 0.24481605351170568,
+      "grad_norm": 0.4936258792877197,
+      "learning_rate": 8e-05,
+      "loss": 1.5957,
+      "step": 4392
+    },
+    {
+      "epoch": 0.24487179487179486,
+      "grad_norm": 0.49569499492645264,
+      "learning_rate": 8e-05,
+      "loss": 1.7342,
+      "step": 4393
+    },
+    {
+      "epoch": 0.24492753623188407,
+      "grad_norm": 0.5097708702087402,
+      "learning_rate": 8e-05,
+      "loss": 1.7159,
+      "step": 4394
+    },
+    {
+      "epoch": 0.24498327759197325,
+      "grad_norm": 0.500952959060669,
+      "learning_rate": 8e-05,
+      "loss": 1.7111,
+      "step": 4395
+    },
+    {
+      "epoch": 0.24503901895206243,
+      "grad_norm": 0.5259371995925903,
+      "learning_rate": 8e-05,
+      "loss": 1.7677,
+      "step": 4396
+    },
+    {
+      "epoch": 0.2450947603121516,
+      "grad_norm": 0.47353237867355347,
+      "learning_rate": 8e-05,
+      "loss": 1.6109,
+      "step": 4397
+    },
+    {
+      "epoch": 0.2451505016722408,
+      "grad_norm": 0.5153634548187256,
+      "learning_rate": 8e-05,
+      "loss": 1.8373,
+      "step": 4398
+    },
+    {
+      "epoch": 0.24520624303233,
+      "grad_norm": 0.5812398791313171,
+      "learning_rate": 8e-05,
+      "loss": 1.977,
+      "step": 4399
+    },
+    {
+      "epoch": 0.24526198439241917,
+      "grad_norm": 0.48786085844039917,
+      "learning_rate": 8e-05,
+      "loss": 1.4989,
+      "step": 4400
+    },
+    {
+      "epoch": 0.24531772575250835,
+      "grad_norm": 0.46562907099723816,
+      "learning_rate": 8e-05,
+      "loss": 1.5978,
+      "step": 4401
+    },
+    {
+      "epoch": 0.24537346711259755,
+      "grad_norm": 0.5388596653938293,
+      "learning_rate": 8e-05,
+      "loss": 1.8173,
+      "step": 4402
+    },
+    {
+      "epoch": 0.24542920847268673,
+      "grad_norm": 0.4896782636642456,
+      "learning_rate": 8e-05,
+      "loss": 1.5591,
+      "step": 4403
+    },
+    {
+      "epoch": 0.2454849498327759,
+      "grad_norm": 0.5358275771141052,
+      "learning_rate": 8e-05,
+      "loss": 1.979,
+      "step": 4404
+    },
+    {
+      "epoch": 0.24554069119286512,
+      "grad_norm": 0.4867069125175476,
+      "learning_rate": 8e-05,
+      "loss": 1.7874,
+      "step": 4405
+    },
+    {
+      "epoch": 0.2455964325529543,
+      "grad_norm": 0.523019015789032,
+      "learning_rate": 8e-05,
+      "loss": 1.7394,
+      "step": 4406
+    },
+    {
+      "epoch": 0.24565217391304348,
+      "grad_norm": 0.5033044219017029,
+      "learning_rate": 8e-05,
+      "loss": 1.7044,
+      "step": 4407
+    },
+    {
+      "epoch": 0.24570791527313265,
+      "grad_norm": 0.5341840982437134,
+      "learning_rate": 8e-05,
+      "loss": 1.7647,
+      "step": 4408
+    },
+    {
+      "epoch": 0.24576365663322186,
+      "grad_norm": 0.52601158618927,
+      "learning_rate": 8e-05,
+      "loss": 1.8909,
+      "step": 4409
+    },
+    {
+      "epoch": 0.24581939799331104,
+      "grad_norm": 0.46190014481544495,
+      "learning_rate": 8e-05,
+      "loss": 1.5464,
+      "step": 4410
+    },
+    {
+      "epoch": 0.24587513935340022,
+      "grad_norm": 0.4728126525878906,
+      "learning_rate": 8e-05,
+      "loss": 1.5892,
+      "step": 4411
+    },
+    {
+      "epoch": 0.2459308807134894,
+      "grad_norm": 0.5146960616111755,
+      "learning_rate": 8e-05,
+      "loss": 1.8503,
+      "step": 4412
+    },
+    {
+      "epoch": 0.2459866220735786,
+      "grad_norm": 0.4900292754173279,
+      "learning_rate": 8e-05,
+      "loss": 1.6651,
+      "step": 4413
+    },
+    {
+      "epoch": 0.24604236343366778,
+      "grad_norm": 0.5353087782859802,
+      "learning_rate": 8e-05,
+      "loss": 1.9005,
+      "step": 4414
+    },
+    {
+      "epoch": 0.24609810479375696,
+      "grad_norm": 0.5150728821754456,
+      "learning_rate": 8e-05,
+      "loss": 1.8399,
+      "step": 4415
+    },
+    {
+      "epoch": 0.24615384615384617,
+      "grad_norm": 0.46863245964050293,
+      "learning_rate": 8e-05,
+      "loss": 1.639,
+      "step": 4416
+    },
+    {
+      "epoch": 0.24620958751393535,
+      "grad_norm": 0.47680985927581787,
+      "learning_rate": 8e-05,
+      "loss": 1.7233,
+      "step": 4417
+    },
+    {
+      "epoch": 0.24626532887402452,
+      "grad_norm": 0.5293520092964172,
+      "learning_rate": 8e-05,
+      "loss": 1.6824,
+      "step": 4418
+    },
+    {
+      "epoch": 0.2463210702341137,
+      "grad_norm": 0.4874856770038605,
+      "learning_rate": 8e-05,
+      "loss": 1.5342,
+      "step": 4419
+    },
+    {
+      "epoch": 0.2463768115942029,
+      "grad_norm": 0.5101845860481262,
+      "learning_rate": 8e-05,
+      "loss": 1.4651,
+      "step": 4420
+    },
+    {
+      "epoch": 0.2464325529542921,
+      "grad_norm": 0.502173900604248,
+      "learning_rate": 8e-05,
+      "loss": 1.6227,
+      "step": 4421
+    },
+    {
+      "epoch": 0.24648829431438127,
+      "grad_norm": 0.49891185760498047,
+      "learning_rate": 8e-05,
+      "loss": 1.7795,
+      "step": 4422
+    },
+    {
+      "epoch": 0.24654403567447045,
+      "grad_norm": 0.5041410326957703,
+      "learning_rate": 8e-05,
+      "loss": 1.728,
+      "step": 4423
+    },
+    {
+      "epoch": 0.24659977703455965,
+      "grad_norm": 0.5070465803146362,
+      "learning_rate": 8e-05,
+      "loss": 1.5741,
+      "step": 4424
+    },
+    {
+      "epoch": 0.24665551839464883,
+      "grad_norm": 0.5117592215538025,
+      "learning_rate": 8e-05,
+      "loss": 1.6389,
+      "step": 4425
+    },
+    {
+      "epoch": 0.246711259754738,
+      "grad_norm": 0.486873060464859,
+      "learning_rate": 8e-05,
+      "loss": 1.5909,
+      "step": 4426
+    },
+    {
+      "epoch": 0.2467670011148272,
+      "grad_norm": 0.5160625576972961,
+      "learning_rate": 8e-05,
+      "loss": 1.9141,
+      "step": 4427
+    },
+    {
+      "epoch": 0.2468227424749164,
+      "grad_norm": 0.4876449704170227,
+      "learning_rate": 8e-05,
+      "loss": 1.7166,
+      "step": 4428
+    },
+    {
+      "epoch": 0.24687848383500557,
+      "grad_norm": 0.4946858882904053,
+      "learning_rate": 8e-05,
+      "loss": 1.6845,
+      "step": 4429
+    },
+    {
+      "epoch": 0.24693422519509475,
+      "grad_norm": 0.4787217080593109,
+      "learning_rate": 8e-05,
+      "loss": 1.5846,
+      "step": 4430
+    },
+    {
+      "epoch": 0.24698996655518396,
+      "grad_norm": 0.4708745777606964,
+      "learning_rate": 8e-05,
+      "loss": 1.4559,
+      "step": 4431
+    },
+    {
+      "epoch": 0.24704570791527314,
+      "grad_norm": 0.5231176614761353,
+      "learning_rate": 8e-05,
+      "loss": 1.7047,
+      "step": 4432
+    },
+    {
+      "epoch": 0.24710144927536232,
+      "grad_norm": 0.4752837121486664,
+      "learning_rate": 8e-05,
+      "loss": 1.7061,
+      "step": 4433
+    },
+    {
+      "epoch": 0.2471571906354515,
+      "grad_norm": 0.5410003662109375,
+      "learning_rate": 8e-05,
+      "loss": 1.8634,
+      "step": 4434
+    },
+    {
+      "epoch": 0.2472129319955407,
+      "grad_norm": 0.48589983582496643,
+      "learning_rate": 8e-05,
+      "loss": 1.6513,
+      "step": 4435
+    },
+    {
+      "epoch": 0.24726867335562988,
+      "grad_norm": 0.561657190322876,
+      "learning_rate": 8e-05,
+      "loss": 1.9844,
+      "step": 4436
+    },
+    {
+      "epoch": 0.24732441471571906,
+      "grad_norm": 0.4805437922477722,
+      "learning_rate": 8e-05,
+      "loss": 1.5661,
+      "step": 4437
+    },
+    {
+      "epoch": 0.24738015607580824,
+      "grad_norm": 0.5096524953842163,
+      "learning_rate": 8e-05,
+      "loss": 1.809,
+      "step": 4438
+    },
+    {
+      "epoch": 0.24743589743589745,
+      "grad_norm": 0.5464960336685181,
+      "learning_rate": 8e-05,
+      "loss": 1.9204,
+      "step": 4439
+    },
+    {
+      "epoch": 0.24749163879598662,
+      "grad_norm": 0.4835813641548157,
+      "learning_rate": 8e-05,
+      "loss": 1.6065,
+      "step": 4440
+    },
+    {
+      "epoch": 0.2475473801560758,
+      "grad_norm": 0.4568433165550232,
+      "learning_rate": 8e-05,
+      "loss": 1.2585,
+      "step": 4441
+    },
+    {
+      "epoch": 0.24760312151616498,
+      "grad_norm": 0.5256950855255127,
+      "learning_rate": 8e-05,
+      "loss": 1.6795,
+      "step": 4442
+    },
+    {
+      "epoch": 0.2476588628762542,
+      "grad_norm": 0.4812221825122833,
+      "learning_rate": 8e-05,
+      "loss": 2.0234,
+      "step": 4443
+    },
+    {
+      "epoch": 0.24771460423634337,
+      "grad_norm": 0.49150028824806213,
+      "learning_rate": 8e-05,
+      "loss": 1.7599,
+      "step": 4444
+    },
+    {
+      "epoch": 0.24777034559643255,
+      "grad_norm": 0.5113860368728638,
+      "learning_rate": 8e-05,
+      "loss": 1.5278,
+      "step": 4445
+    },
+    {
+      "epoch": 0.24782608695652175,
+      "grad_norm": 0.497530460357666,
+      "learning_rate": 8e-05,
+      "loss": 1.6159,
+      "step": 4446
+    },
+    {
+      "epoch": 0.24788182831661093,
+      "grad_norm": 0.5291724801063538,
+      "learning_rate": 8e-05,
+      "loss": 1.4563,
+      "step": 4447
+    },
+    {
+      "epoch": 0.2479375696767001,
+      "grad_norm": 0.52456134557724,
+      "learning_rate": 8e-05,
+      "loss": 1.8435,
+      "step": 4448
+    },
+    {
+      "epoch": 0.2479933110367893,
+      "grad_norm": 0.4777376353740692,
+      "learning_rate": 8e-05,
+      "loss": 1.7825,
+      "step": 4449
+    },
+    {
+      "epoch": 0.2480490523968785,
+      "grad_norm": 0.4933377504348755,
+      "learning_rate": 8e-05,
+      "loss": 1.8244,
+      "step": 4450
+    },
+    {
+      "epoch": 0.24810479375696767,
+      "grad_norm": 0.4991428256034851,
+      "learning_rate": 8e-05,
+      "loss": 1.6007,
+      "step": 4451
+    },
+    {
+      "epoch": 0.24816053511705685,
+      "grad_norm": 0.4872720241546631,
+      "learning_rate": 8e-05,
+      "loss": 1.7075,
+      "step": 4452
+    },
+    {
+      "epoch": 0.24821627647714603,
+      "grad_norm": 0.511841893196106,
+      "learning_rate": 8e-05,
+      "loss": 1.9252,
+      "step": 4453
+    },
+    {
+      "epoch": 0.24827201783723524,
+      "grad_norm": 0.51175457239151,
+      "learning_rate": 8e-05,
+      "loss": 1.6001,
+      "step": 4454
+    },
+    {
+      "epoch": 0.24832775919732442,
+      "grad_norm": 0.48516520857810974,
+      "learning_rate": 8e-05,
+      "loss": 1.6602,
+      "step": 4455
+    },
+    {
+      "epoch": 0.2483835005574136,
+      "grad_norm": 0.4751116931438446,
+      "learning_rate": 8e-05,
+      "loss": 1.6331,
+      "step": 4456
+    },
+    {
+      "epoch": 0.24843924191750277,
+      "grad_norm": 0.4790758490562439,
+      "learning_rate": 8e-05,
+      "loss": 1.6961,
+      "step": 4457
+    },
+    {
+      "epoch": 0.24849498327759198,
+      "grad_norm": 0.536362886428833,
+      "learning_rate": 8e-05,
+      "loss": 1.8841,
+      "step": 4458
+    },
+    {
+      "epoch": 0.24855072463768116,
+      "grad_norm": 0.4993929862976074,
+      "learning_rate": 8e-05,
+      "loss": 1.6816,
+      "step": 4459
+    },
+    {
+      "epoch": 0.24860646599777034,
+      "grad_norm": 0.5037083029747009,
+      "learning_rate": 8e-05,
+      "loss": 1.877,
+      "step": 4460
+    },
+    {
+      "epoch": 0.24866220735785954,
+      "grad_norm": 0.5280500650405884,
+      "learning_rate": 8e-05,
+      "loss": 1.884,
+      "step": 4461
+    },
+    {
+      "epoch": 0.24871794871794872,
+      "grad_norm": 0.5553090572357178,
+      "learning_rate": 8e-05,
+      "loss": 1.8299,
+      "step": 4462
+    },
+    {
+      "epoch": 0.2487736900780379,
+      "grad_norm": 0.49785253405570984,
+      "learning_rate": 8e-05,
+      "loss": 1.7691,
+      "step": 4463
+    },
+    {
+      "epoch": 0.24882943143812708,
+      "grad_norm": 0.5308324694633484,
+      "learning_rate": 8e-05,
+      "loss": 1.8467,
+      "step": 4464
+    },
+    {
+      "epoch": 0.2488851727982163,
+      "grad_norm": 0.5013288259506226,
+      "learning_rate": 8e-05,
+      "loss": 1.714,
+      "step": 4465
+    },
+    {
+      "epoch": 0.24894091415830547,
+      "grad_norm": 0.49870118498802185,
+      "learning_rate": 8e-05,
+      "loss": 1.8183,
+      "step": 4466
+    },
+    {
+      "epoch": 0.24899665551839464,
+      "grad_norm": 0.49616575241088867,
+      "learning_rate": 8e-05,
+      "loss": 1.5885,
+      "step": 4467
+    },
+    {
+      "epoch": 0.24905239687848382,
+      "grad_norm": 0.5420321226119995,
+      "learning_rate": 8e-05,
+      "loss": 1.8196,
+      "step": 4468
+    },
+    {
+      "epoch": 0.24910813823857303,
+      "grad_norm": 0.5438189506530762,
+      "learning_rate": 8e-05,
+      "loss": 1.8031,
+      "step": 4469
+    },
+    {
+      "epoch": 0.2491638795986622,
+      "grad_norm": 0.5136116743087769,
+      "learning_rate": 8e-05,
+      "loss": 1.7156,
+      "step": 4470
+    },
+    {
+      "epoch": 0.2492196209587514,
+      "grad_norm": 0.4942511320114136,
+      "learning_rate": 8e-05,
+      "loss": 1.6699,
+      "step": 4471
+    },
+    {
+      "epoch": 0.2492753623188406,
+      "grad_norm": 0.4945601224899292,
+      "learning_rate": 8e-05,
+      "loss": 1.6535,
+      "step": 4472
+    },
+    {
+      "epoch": 0.24933110367892977,
+      "grad_norm": 0.48275116086006165,
+      "learning_rate": 8e-05,
+      "loss": 1.583,
+      "step": 4473
+    },
+    {
+      "epoch": 0.24938684503901895,
+      "grad_norm": 0.4992966651916504,
+      "learning_rate": 8e-05,
+      "loss": 1.6477,
+      "step": 4474
+    },
+    {
+      "epoch": 0.24944258639910813,
+      "grad_norm": 0.4911842942237854,
+      "learning_rate": 8e-05,
+      "loss": 1.5971,
+      "step": 4475
+    },
+    {
+      "epoch": 0.24949832775919734,
+      "grad_norm": 0.5159841179847717,
+      "learning_rate": 8e-05,
+      "loss": 1.782,
+      "step": 4476
+    },
+    {
+      "epoch": 0.24955406911928651,
+      "grad_norm": 0.5172543525695801,
+      "learning_rate": 8e-05,
+      "loss": 1.8104,
+      "step": 4477
+    },
+    {
+      "epoch": 0.2496098104793757,
+      "grad_norm": 0.5436558723449707,
+      "learning_rate": 8e-05,
+      "loss": 1.6937,
+      "step": 4478
+    },
+    {
+      "epoch": 0.24966555183946487,
+      "grad_norm": 0.5378687977790833,
+      "learning_rate": 8e-05,
+      "loss": 1.9014,
+      "step": 4479
+    },
+    {
+      "epoch": 0.24972129319955408,
+      "grad_norm": 0.5014615058898926,
+      "learning_rate": 8e-05,
+      "loss": 1.5676,
+      "step": 4480
+    },
+    {
+      "epoch": 0.24977703455964326,
+      "grad_norm": 0.47929060459136963,
+      "learning_rate": 8e-05,
+      "loss": 1.8378,
+      "step": 4481
+    },
+    {
+      "epoch": 0.24983277591973244,
+      "grad_norm": 0.47527262568473816,
+      "learning_rate": 8e-05,
+      "loss": 1.6668,
+      "step": 4482
+    },
+    {
+      "epoch": 0.24988851727982161,
+      "grad_norm": 0.49721941351890564,
+      "learning_rate": 8e-05,
+      "loss": 1.6998,
+      "step": 4483
+    },
+    {
+      "epoch": 0.24994425863991082,
+      "grad_norm": 0.4377490282058716,
+      "learning_rate": 8e-05,
+      "loss": 1.4564,
+      "step": 4484
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.531048059463501,
+      "learning_rate": 8e-05,
+      "loss": 1.9123,
+      "step": 4485
+    },
+    {
+      "epoch": 0.2500557413600892,
+      "grad_norm": 0.5143203139305115,
+      "learning_rate": 8e-05,
+      "loss": 1.8096,
+      "step": 4486
+    },
+    {
+      "epoch": 0.25011148272017836,
+      "grad_norm": 0.44039544463157654,
+      "learning_rate": 8e-05,
+      "loss": 1.6269,
+      "step": 4487
+    },
+    {
+      "epoch": 0.25016722408026754,
+      "grad_norm": 0.493327796459198,
+      "learning_rate": 8e-05,
+      "loss": 1.6879,
+      "step": 4488
+    },
+    {
+      "epoch": 0.25022296544035677,
+      "grad_norm": 0.47744813561439514,
+      "learning_rate": 8e-05,
+      "loss": 1.8025,
+      "step": 4489
+    },
+    {
+      "epoch": 0.25027870680044595,
+      "grad_norm": 0.5044304132461548,
+      "learning_rate": 8e-05,
+      "loss": 1.7285,
+      "step": 4490
+    },
+    {
+      "epoch": 0.25033444816053513,
+      "grad_norm": 0.5493786931037903,
+      "learning_rate": 8e-05,
+      "loss": 1.889,
+      "step": 4491
+    },
+    {
+      "epoch": 0.2503901895206243,
+      "grad_norm": 0.4919975996017456,
+      "learning_rate": 8e-05,
+      "loss": 1.6406,
+      "step": 4492
+    },
+    {
+      "epoch": 0.2504459308807135,
+      "grad_norm": 0.5749701857566833,
+      "learning_rate": 8e-05,
+      "loss": 1.8115,
+      "step": 4493
+    },
+    {
+      "epoch": 0.25050167224080266,
+      "grad_norm": 0.5213183760643005,
+      "learning_rate": 8e-05,
+      "loss": 1.8957,
+      "step": 4494
+    },
+    {
+      "epoch": 0.25055741360089184,
+      "grad_norm": 0.5226160287857056,
+      "learning_rate": 8e-05,
+      "loss": 1.6662,
+      "step": 4495
+    },
+    {
+      "epoch": 0.250613154960981,
+      "grad_norm": 0.5784220695495605,
+      "learning_rate": 8e-05,
+      "loss": 2.0499,
+      "step": 4496
+    },
+    {
+      "epoch": 0.25066889632107026,
+      "grad_norm": 0.435458242893219,
+      "learning_rate": 8e-05,
+      "loss": 1.4048,
+      "step": 4497
+    },
+    {
+      "epoch": 0.25072463768115943,
+      "grad_norm": 0.5050778388977051,
+      "learning_rate": 8e-05,
+      "loss": 1.6978,
+      "step": 4498
+    },
+    {
+      "epoch": 0.2507803790412486,
+      "grad_norm": 0.504496157169342,
+      "learning_rate": 8e-05,
+      "loss": 1.6795,
+      "step": 4499
+    },
+    {
+      "epoch": 0.2508361204013378,
+      "grad_norm": 0.5178218483924866,
+      "learning_rate": 8e-05,
+      "loss": 1.8451,
+      "step": 4500
+    },
+    {
+      "epoch": 0.25089186176142697,
+      "grad_norm": 0.4855165183544159,
+      "learning_rate": 8e-05,
+      "loss": 1.6222,
+      "step": 4501
+    },
+    {
+      "epoch": 0.25094760312151615,
+      "grad_norm": 0.5280212759971619,
+      "learning_rate": 8e-05,
+      "loss": 1.9271,
+      "step": 4502
+    },
+    {
+      "epoch": 0.25100334448160533,
+      "grad_norm": 0.5055220723152161,
+      "learning_rate": 8e-05,
+      "loss": 1.7128,
+      "step": 4503
+    },
+    {
+      "epoch": 0.25105908584169456,
+      "grad_norm": 0.5127941370010376,
+      "learning_rate": 8e-05,
+      "loss": 1.8969,
+      "step": 4504
+    },
+    {
+      "epoch": 0.25111482720178374,
+      "grad_norm": 0.47817113995552063,
+      "learning_rate": 8e-05,
+      "loss": 1.5779,
+      "step": 4505
+    },
+    {
+      "epoch": 0.2511705685618729,
+      "grad_norm": 0.5066494345664978,
+      "learning_rate": 8e-05,
+      "loss": 1.6365,
+      "step": 4506
+    },
+    {
+      "epoch": 0.2512263099219621,
+      "grad_norm": 0.5416378974914551,
+      "learning_rate": 8e-05,
+      "loss": 1.875,
+      "step": 4507
+    },
+    {
+      "epoch": 0.2512820512820513,
+      "grad_norm": 0.48432135581970215,
+      "learning_rate": 8e-05,
+      "loss": 1.5935,
+      "step": 4508
+    },
+    {
+      "epoch": 0.25133779264214046,
+      "grad_norm": 0.45147547125816345,
+      "learning_rate": 8e-05,
+      "loss": 1.5384,
+      "step": 4509
+    },
+    {
+      "epoch": 0.25139353400222963,
+      "grad_norm": 0.5376551747322083,
+      "learning_rate": 8e-05,
+      "loss": 1.9072,
+      "step": 4510
+    },
+    {
+      "epoch": 0.2514492753623188,
+      "grad_norm": 0.504919171333313,
+      "learning_rate": 8e-05,
+      "loss": 1.7057,
+      "step": 4511
+    },
+    {
+      "epoch": 0.25150501672240805,
+      "grad_norm": 0.5230275988578796,
+      "learning_rate": 8e-05,
+      "loss": 1.7698,
+      "step": 4512
+    },
+    {
+      "epoch": 0.2515607580824972,
+      "grad_norm": 0.4875522553920746,
+      "learning_rate": 8e-05,
+      "loss": 1.6571,
+      "step": 4513
+    },
+    {
+      "epoch": 0.2516164994425864,
+      "grad_norm": 0.5083442330360413,
+      "learning_rate": 8e-05,
+      "loss": 1.7291,
+      "step": 4514
+    },
+    {
+      "epoch": 0.2516722408026756,
+      "grad_norm": 0.46802598237991333,
+      "learning_rate": 8e-05,
+      "loss": 1.6689,
+      "step": 4515
+    },
+    {
+      "epoch": 0.25172798216276476,
+      "grad_norm": 0.4877402186393738,
+      "learning_rate": 8e-05,
+      "loss": 1.7509,
+      "step": 4516
+    },
+    {
+      "epoch": 0.25178372352285394,
+      "grad_norm": 0.5294947624206543,
+      "learning_rate": 8e-05,
+      "loss": 1.7165,
+      "step": 4517
+    },
+    {
+      "epoch": 0.2518394648829431,
+      "grad_norm": 0.47198450565338135,
+      "learning_rate": 8e-05,
+      "loss": 1.4383,
+      "step": 4518
+    },
+    {
+      "epoch": 0.25189520624303235,
+      "grad_norm": 0.55704665184021,
+      "learning_rate": 8e-05,
+      "loss": 2.1062,
+      "step": 4519
+    },
+    {
+      "epoch": 0.25195094760312153,
+      "grad_norm": 0.5346017479896545,
+      "learning_rate": 8e-05,
+      "loss": 1.6289,
+      "step": 4520
+    },
+    {
+      "epoch": 0.2520066889632107,
+      "grad_norm": 0.5310443639755249,
+      "learning_rate": 8e-05,
+      "loss": 1.7001,
+      "step": 4521
+    },
+    {
+      "epoch": 0.2520624303232999,
+      "grad_norm": 0.48790326714515686,
+      "learning_rate": 8e-05,
+      "loss": 1.6094,
+      "step": 4522
+    },
+    {
+      "epoch": 0.25211817168338907,
+      "grad_norm": 0.55752032995224,
+      "learning_rate": 8e-05,
+      "loss": 1.9714,
+      "step": 4523
+    },
+    {
+      "epoch": 0.25217391304347825,
+      "grad_norm": 0.48758000135421753,
+      "learning_rate": 8e-05,
+      "loss": 1.7153,
+      "step": 4524
+    },
+    {
+      "epoch": 0.2522296544035674,
+      "grad_norm": 0.4836397171020508,
+      "learning_rate": 8e-05,
+      "loss": 1.7017,
+      "step": 4525
+    },
+    {
+      "epoch": 0.2522853957636566,
+      "grad_norm": 0.48506078124046326,
+      "learning_rate": 8e-05,
+      "loss": 1.7446,
+      "step": 4526
+    },
+    {
+      "epoch": 0.25234113712374584,
+      "grad_norm": 0.4948675334453583,
+      "learning_rate": 8e-05,
+      "loss": 1.5827,
+      "step": 4527
+    },
+    {
+      "epoch": 0.252396878483835,
+      "grad_norm": 0.5153377652168274,
+      "learning_rate": 8e-05,
+      "loss": 1.8249,
+      "step": 4528
+    },
+    {
+      "epoch": 0.2524526198439242,
+      "grad_norm": 0.46015942096710205,
+      "learning_rate": 8e-05,
+      "loss": 1.4809,
+      "step": 4529
+    },
+    {
+      "epoch": 0.2525083612040134,
+      "grad_norm": 0.556275486946106,
+      "learning_rate": 8e-05,
+      "loss": 1.8923,
+      "step": 4530
+    },
+    {
+      "epoch": 0.25256410256410255,
+      "grad_norm": 0.5005241632461548,
+      "learning_rate": 8e-05,
+      "loss": 1.7709,
+      "step": 4531
+    },
+    {
+      "epoch": 0.25261984392419173,
+      "grad_norm": 0.4980381429195404,
+      "learning_rate": 8e-05,
+      "loss": 1.7674,
+      "step": 4532
+    },
+    {
+      "epoch": 0.2526755852842809,
+      "grad_norm": 0.48650914430618286,
+      "learning_rate": 8e-05,
+      "loss": 1.8241,
+      "step": 4533
+    },
+    {
+      "epoch": 0.25273132664437015,
+      "grad_norm": 0.4514947235584259,
+      "learning_rate": 8e-05,
+      "loss": 1.5061,
+      "step": 4534
+    },
+    {
+      "epoch": 0.2527870680044593,
+      "grad_norm": 0.49436303973197937,
+      "learning_rate": 8e-05,
+      "loss": 1.5731,
+      "step": 4535
+    },
+    {
+      "epoch": 0.2528428093645485,
+      "grad_norm": 0.5584789514541626,
+      "learning_rate": 8e-05,
+      "loss": 1.9422,
+      "step": 4536
+    },
+    {
+      "epoch": 0.2528985507246377,
+      "grad_norm": 0.4882524609565735,
+      "learning_rate": 8e-05,
+      "loss": 1.5859,
+      "step": 4537
+    },
+    {
+      "epoch": 0.25295429208472686,
+      "grad_norm": 0.478646844625473,
+      "learning_rate": 8e-05,
+      "loss": 1.6171,
+      "step": 4538
+    },
+    {
+      "epoch": 0.25301003344481604,
+      "grad_norm": 0.48294344544410706,
+      "learning_rate": 8e-05,
+      "loss": 1.652,
+      "step": 4539
+    },
+    {
+      "epoch": 0.2530657748049052,
+      "grad_norm": 0.5388214588165283,
+      "learning_rate": 8e-05,
+      "loss": 1.7317,
+      "step": 4540
+    },
+    {
+      "epoch": 0.25312151616499445,
+      "grad_norm": 0.5226413011550903,
+      "learning_rate": 8e-05,
+      "loss": 1.7449,
+      "step": 4541
+    },
+    {
+      "epoch": 0.25317725752508363,
+      "grad_norm": 0.5220893621444702,
+      "learning_rate": 8e-05,
+      "loss": 1.718,
+      "step": 4542
+    },
+    {
+      "epoch": 0.2532329988851728,
+      "grad_norm": 0.5089812874794006,
+      "learning_rate": 8e-05,
+      "loss": 1.8204,
+      "step": 4543
+    },
+    {
+      "epoch": 0.253288740245262,
+      "grad_norm": 0.4798874855041504,
+      "learning_rate": 8e-05,
+      "loss": 1.6559,
+      "step": 4544
+    },
+    {
+      "epoch": 0.25334448160535117,
+      "grad_norm": 0.543845534324646,
+      "learning_rate": 8e-05,
+      "loss": 1.7922,
+      "step": 4545
+    },
+    {
+      "epoch": 0.25340022296544035,
+      "grad_norm": 0.5123974680900574,
+      "learning_rate": 8e-05,
+      "loss": 1.7801,
+      "step": 4546
+    },
+    {
+      "epoch": 0.2534559643255295,
+      "grad_norm": 0.5205566883087158,
+      "learning_rate": 8e-05,
+      "loss": 1.7462,
+      "step": 4547
+    },
+    {
+      "epoch": 0.2535117056856187,
+      "grad_norm": 0.49157705903053284,
+      "learning_rate": 8e-05,
+      "loss": 1.6683,
+      "step": 4548
+    },
+    {
+      "epoch": 0.25356744704570794,
+      "grad_norm": 0.5303407907485962,
+      "learning_rate": 8e-05,
+      "loss": 1.7813,
+      "step": 4549
+    },
+    {
+      "epoch": 0.2536231884057971,
+      "grad_norm": 0.520340085029602,
+      "learning_rate": 8e-05,
+      "loss": 1.6946,
+      "step": 4550
+    },
+    {
+      "epoch": 0.2536789297658863,
+      "grad_norm": 0.528924286365509,
+      "learning_rate": 8e-05,
+      "loss": 1.3523,
+      "step": 4551
+    },
+    {
+      "epoch": 0.2537346711259755,
+      "grad_norm": 0.4791257083415985,
+      "learning_rate": 8e-05,
+      "loss": 1.5945,
+      "step": 4552
+    },
+    {
+      "epoch": 0.25379041248606465,
+      "grad_norm": 0.48630478978157043,
+      "learning_rate": 8e-05,
+      "loss": 1.8585,
+      "step": 4553
+    },
+    {
+      "epoch": 0.25384615384615383,
+      "grad_norm": 0.4839245676994324,
+      "learning_rate": 8e-05,
+      "loss": 1.6635,
+      "step": 4554
+    },
+    {
+      "epoch": 0.253901895206243,
+      "grad_norm": 0.5476188659667969,
+      "learning_rate": 8e-05,
+      "loss": 2.004,
+      "step": 4555
+    },
+    {
+      "epoch": 0.25395763656633225,
+      "grad_norm": 0.5112416744232178,
+      "learning_rate": 8e-05,
+      "loss": 1.6964,
+      "step": 4556
+    },
+    {
+      "epoch": 0.2540133779264214,
+      "grad_norm": 0.4688325822353363,
+      "learning_rate": 8e-05,
+      "loss": 1.6512,
+      "step": 4557
+    },
+    {
+      "epoch": 0.2540691192865106,
+      "grad_norm": 0.49445050954818726,
+      "learning_rate": 8e-05,
+      "loss": 1.8496,
+      "step": 4558
+    },
+    {
+      "epoch": 0.2541248606465998,
+      "grad_norm": 0.5357611775398254,
+      "learning_rate": 8e-05,
+      "loss": 1.7822,
+      "step": 4559
+    },
+    {
+      "epoch": 0.25418060200668896,
+      "grad_norm": 0.5030912756919861,
+      "learning_rate": 8e-05,
+      "loss": 1.5574,
+      "step": 4560
+    },
+    {
+      "epoch": 0.25423634336677814,
+      "grad_norm": 0.5398586988449097,
+      "learning_rate": 8e-05,
+      "loss": 1.9175,
+      "step": 4561
+    },
+    {
+      "epoch": 0.2542920847268673,
+      "grad_norm": 0.49171751737594604,
+      "learning_rate": 8e-05,
+      "loss": 1.814,
+      "step": 4562
+    },
+    {
+      "epoch": 0.2543478260869565,
+      "grad_norm": 0.4866318106651306,
+      "learning_rate": 8e-05,
+      "loss": 1.6577,
+      "step": 4563
+    },
+    {
+      "epoch": 0.25440356744704573,
+      "grad_norm": 0.4502681791782379,
+      "learning_rate": 8e-05,
+      "loss": 1.5948,
+      "step": 4564
+    },
+    {
+      "epoch": 0.2544593088071349,
+      "grad_norm": 0.5153110027313232,
+      "learning_rate": 8e-05,
+      "loss": 1.6973,
+      "step": 4565
+    },
+    {
+      "epoch": 0.2545150501672241,
+      "grad_norm": 0.4543468654155731,
+      "learning_rate": 8e-05,
+      "loss": 1.4745,
+      "step": 4566
+    },
+    {
+      "epoch": 0.25457079152731327,
+      "grad_norm": 0.4921804368495941,
+      "learning_rate": 8e-05,
+      "loss": 1.7234,
+      "step": 4567
+    },
+    {
+      "epoch": 0.25462653288740245,
+      "grad_norm": 0.5395398139953613,
+      "learning_rate": 8e-05,
+      "loss": 1.6885,
+      "step": 4568
+    },
+    {
+      "epoch": 0.2546822742474916,
+      "grad_norm": 0.49862754344940186,
+      "learning_rate": 8e-05,
+      "loss": 1.6302,
+      "step": 4569
+    },
+    {
+      "epoch": 0.2547380156075808,
+      "grad_norm": 0.5175421833992004,
+      "learning_rate": 8e-05,
+      "loss": 1.8645,
+      "step": 4570
+    },
+    {
+      "epoch": 0.25479375696767004,
+      "grad_norm": 0.5347442626953125,
+      "learning_rate": 8e-05,
+      "loss": 1.6279,
+      "step": 4571
+    },
+    {
+      "epoch": 0.2548494983277592,
+      "grad_norm": 0.5138886570930481,
+      "learning_rate": 8e-05,
+      "loss": 1.4774,
+      "step": 4572
+    },
+    {
+      "epoch": 0.2549052396878484,
+      "grad_norm": 0.450295090675354,
+      "learning_rate": 8e-05,
+      "loss": 1.0234,
+      "step": 4573
+    },
+    {
+      "epoch": 0.2549609810479376,
+      "grad_norm": 0.5415732860565186,
+      "learning_rate": 8e-05,
+      "loss": 1.7688,
+      "step": 4574
+    },
+    {
+      "epoch": 0.25501672240802675,
+      "grad_norm": 0.5672006607055664,
+      "learning_rate": 8e-05,
+      "loss": 1.8689,
+      "step": 4575
+    },
+    {
+      "epoch": 0.25507246376811593,
+      "grad_norm": 0.5203934907913208,
+      "learning_rate": 8e-05,
+      "loss": 1.6057,
+      "step": 4576
+    },
+    {
+      "epoch": 0.2551282051282051,
+      "grad_norm": 0.5594499111175537,
+      "learning_rate": 8e-05,
+      "loss": 1.7061,
+      "step": 4577
+    },
+    {
+      "epoch": 0.2551839464882943,
+      "grad_norm": 0.5058487057685852,
+      "learning_rate": 8e-05,
+      "loss": 1.7657,
+      "step": 4578
+    },
+    {
+      "epoch": 0.2552396878483835,
+      "grad_norm": 0.4886641204357147,
+      "learning_rate": 8e-05,
+      "loss": 1.615,
+      "step": 4579
+    },
+    {
+      "epoch": 0.2552954292084727,
+      "grad_norm": 0.5026295781135559,
+      "learning_rate": 8e-05,
+      "loss": 1.6597,
+      "step": 4580
+    },
+    {
+      "epoch": 0.2553511705685619,
+      "grad_norm": 0.4680199921131134,
+      "learning_rate": 8e-05,
+      "loss": 1.7476,
+      "step": 4581
+    },
+    {
+      "epoch": 0.25540691192865106,
+      "grad_norm": 0.5621483325958252,
+      "learning_rate": 8e-05,
+      "loss": 1.8504,
+      "step": 4582
+    },
+    {
+      "epoch": 0.25546265328874024,
+      "grad_norm": 0.5055748820304871,
+      "learning_rate": 8e-05,
+      "loss": 1.9231,
+      "step": 4583
+    },
+    {
+      "epoch": 0.2555183946488294,
+      "grad_norm": 0.5216691493988037,
+      "learning_rate": 8e-05,
+      "loss": 1.6375,
+      "step": 4584
+    },
+    {
+      "epoch": 0.2555741360089186,
+      "grad_norm": 0.4793626070022583,
+      "learning_rate": 8e-05,
+      "loss": 1.5565,
+      "step": 4585
+    },
+    {
+      "epoch": 0.25562987736900783,
+      "grad_norm": 0.5235403776168823,
+      "learning_rate": 8e-05,
+      "loss": 1.9795,
+      "step": 4586
+    },
+    {
+      "epoch": 0.255685618729097,
+      "grad_norm": 0.5455905199050903,
+      "learning_rate": 8e-05,
+      "loss": 2.0298,
+      "step": 4587
+    },
+    {
+      "epoch": 0.2557413600891862,
+      "grad_norm": 0.5437529683113098,
+      "learning_rate": 8e-05,
+      "loss": 1.7714,
+      "step": 4588
+    },
+    {
+      "epoch": 0.25579710144927537,
+      "grad_norm": 0.4438577890396118,
+      "learning_rate": 8e-05,
+      "loss": 1.5776,
+      "step": 4589
+    },
+    {
+      "epoch": 0.25585284280936454,
+      "grad_norm": 0.4744518995285034,
+      "learning_rate": 8e-05,
+      "loss": 1.5394,
+      "step": 4590
+    },
+    {
+      "epoch": 0.2559085841694537,
+      "grad_norm": 0.49256959557533264,
+      "learning_rate": 8e-05,
+      "loss": 1.72,
+      "step": 4591
+    },
+    {
+      "epoch": 0.2559643255295429,
+      "grad_norm": 0.4950970411300659,
+      "learning_rate": 8e-05,
+      "loss": 1.6956,
+      "step": 4592
+    },
+    {
+      "epoch": 0.2560200668896321,
+      "grad_norm": 0.5208807587623596,
+      "learning_rate": 8e-05,
+      "loss": 1.7964,
+      "step": 4593
+    },
+    {
+      "epoch": 0.2560758082497213,
+      "grad_norm": 0.47823184728622437,
+      "learning_rate": 8e-05,
+      "loss": 1.6011,
+      "step": 4594
+    },
+    {
+      "epoch": 0.2561315496098105,
+      "grad_norm": 0.5411269664764404,
+      "learning_rate": 8e-05,
+      "loss": 1.7741,
+      "step": 4595
+    },
+    {
+      "epoch": 0.2561872909698997,
+      "grad_norm": 0.4891355633735657,
+      "learning_rate": 8e-05,
+      "loss": 1.5656,
+      "step": 4596
+    },
+    {
+      "epoch": 0.25624303232998885,
+      "grad_norm": 0.5012332201004028,
+      "learning_rate": 8e-05,
+      "loss": 1.6128,
+      "step": 4597
+    },
+    {
+      "epoch": 0.25629877369007803,
+      "grad_norm": 0.49547815322875977,
+      "learning_rate": 8e-05,
+      "loss": 1.7545,
+      "step": 4598
+    },
+    {
+      "epoch": 0.2563545150501672,
+      "grad_norm": 0.5473471283912659,
+      "learning_rate": 8e-05,
+      "loss": 1.9162,
+      "step": 4599
+    },
+    {
+      "epoch": 0.2564102564102564,
+      "grad_norm": 0.5002238750457764,
+      "learning_rate": 8e-05,
+      "loss": 1.6341,
+      "step": 4600
+    },
+    {
+      "epoch": 0.2564659977703456,
+      "grad_norm": 0.5576561689376831,
+      "learning_rate": 8e-05,
+      "loss": 1.78,
+      "step": 4601
+    },
+    {
+      "epoch": 0.2565217391304348,
+      "grad_norm": 0.5200320482254028,
+      "learning_rate": 8e-05,
+      "loss": 1.8138,
+      "step": 4602
+    },
+    {
+      "epoch": 0.256577480490524,
+      "grad_norm": 0.5285021662712097,
+      "learning_rate": 8e-05,
+      "loss": 1.8028,
+      "step": 4603
+    },
+    {
+      "epoch": 0.25663322185061316,
+      "grad_norm": 0.46985504031181335,
+      "learning_rate": 8e-05,
+      "loss": 1.5906,
+      "step": 4604
+    },
+    {
+      "epoch": 0.25668896321070234,
+      "grad_norm": 0.5084950923919678,
+      "learning_rate": 8e-05,
+      "loss": 1.8134,
+      "step": 4605
+    },
+    {
+      "epoch": 0.2567447045707915,
+      "grad_norm": 0.4895882308483124,
+      "learning_rate": 8e-05,
+      "loss": 1.745,
+      "step": 4606
+    },
+    {
+      "epoch": 0.2568004459308807,
+      "grad_norm": 0.5340995788574219,
+      "learning_rate": 8e-05,
+      "loss": 1.7095,
+      "step": 4607
+    },
+    {
+      "epoch": 0.2568561872909699,
+      "grad_norm": 0.47565460205078125,
+      "learning_rate": 8e-05,
+      "loss": 1.5556,
+      "step": 4608
+    },
+    {
+      "epoch": 0.2569119286510591,
+      "grad_norm": 0.4609222412109375,
+      "learning_rate": 8e-05,
+      "loss": 1.6959,
+      "step": 4609
+    },
+    {
+      "epoch": 0.2569676700111483,
+      "grad_norm": 0.4868863523006439,
+      "learning_rate": 8e-05,
+      "loss": 1.6147,
+      "step": 4610
+    },
+    {
+      "epoch": 0.25702341137123746,
+      "grad_norm": 0.480629563331604,
+      "learning_rate": 8e-05,
+      "loss": 1.6158,
+      "step": 4611
+    },
+    {
+      "epoch": 0.25707915273132664,
+      "grad_norm": 0.5915615558624268,
+      "learning_rate": 8e-05,
+      "loss": 1.833,
+      "step": 4612
+    },
+    {
+      "epoch": 0.2571348940914158,
+      "grad_norm": 0.5114143490791321,
+      "learning_rate": 8e-05,
+      "loss": 1.7402,
+      "step": 4613
+    },
+    {
+      "epoch": 0.257190635451505,
+      "grad_norm": 0.46387219429016113,
+      "learning_rate": 8e-05,
+      "loss": 1.6528,
+      "step": 4614
+    },
+    {
+      "epoch": 0.2572463768115942,
+      "grad_norm": 0.5148956775665283,
+      "learning_rate": 8e-05,
+      "loss": 1.7891,
+      "step": 4615
+    },
+    {
+      "epoch": 0.2573021181716834,
+      "grad_norm": 0.5419014692306519,
+      "learning_rate": 8e-05,
+      "loss": 1.9276,
+      "step": 4616
+    },
+    {
+      "epoch": 0.2573578595317726,
+      "grad_norm": 0.5739691257476807,
+      "learning_rate": 8e-05,
+      "loss": 1.8287,
+      "step": 4617
+    },
+    {
+      "epoch": 0.25741360089186177,
+      "grad_norm": 0.5347944498062134,
+      "learning_rate": 8e-05,
+      "loss": 1.8231,
+      "step": 4618
+    },
+    {
+      "epoch": 0.25746934225195095,
+      "grad_norm": 0.47412458062171936,
+      "learning_rate": 8e-05,
+      "loss": 1.7517,
+      "step": 4619
+    },
+    {
+      "epoch": 0.25752508361204013,
+      "grad_norm": 0.5531072616577148,
+      "learning_rate": 8e-05,
+      "loss": 1.8875,
+      "step": 4620
+    },
+    {
+      "epoch": 0.2575808249721293,
+      "grad_norm": 0.4874540865421295,
+      "learning_rate": 8e-05,
+      "loss": 1.5426,
+      "step": 4621
+    },
+    {
+      "epoch": 0.2576365663322185,
+      "grad_norm": 0.5075742602348328,
+      "learning_rate": 8e-05,
+      "loss": 1.6955,
+      "step": 4622
+    },
+    {
+      "epoch": 0.25769230769230766,
+      "grad_norm": 0.5765818357467651,
+      "learning_rate": 8e-05,
+      "loss": 2.1356,
+      "step": 4623
+    },
+    {
+      "epoch": 0.2577480490523969,
+      "grad_norm": 0.5081427693367004,
+      "learning_rate": 8e-05,
+      "loss": 1.7109,
+      "step": 4624
+    },
+    {
+      "epoch": 0.2578037904124861,
+      "grad_norm": 0.5203375816345215,
+      "learning_rate": 8e-05,
+      "loss": 1.8658,
+      "step": 4625
+    },
+    {
+      "epoch": 0.25785953177257526,
+      "grad_norm": 0.49736112356185913,
+      "learning_rate": 8e-05,
+      "loss": 1.63,
+      "step": 4626
+    },
+    {
+      "epoch": 0.25791527313266444,
+      "grad_norm": 0.495447039604187,
+      "learning_rate": 8e-05,
+      "loss": 1.8224,
+      "step": 4627
+    },
+    {
+      "epoch": 0.2579710144927536,
+      "grad_norm": 0.4577196538448334,
+      "learning_rate": 8e-05,
+      "loss": 1.6015,
+      "step": 4628
+    },
+    {
+      "epoch": 0.2580267558528428,
+      "grad_norm": 0.47778797149658203,
+      "learning_rate": 8e-05,
+      "loss": 1.6896,
+      "step": 4629
+    },
+    {
+      "epoch": 0.25808249721293197,
+      "grad_norm": 0.4753131866455078,
+      "learning_rate": 8e-05,
+      "loss": 1.7952,
+      "step": 4630
+    },
+    {
+      "epoch": 0.2581382385730212,
+      "grad_norm": 0.5390071272850037,
+      "learning_rate": 8e-05,
+      "loss": 1.6938,
+      "step": 4631
+    },
+    {
+      "epoch": 0.2581939799331104,
+      "grad_norm": 0.5110119581222534,
+      "learning_rate": 8e-05,
+      "loss": 1.5397,
+      "step": 4632
+    },
+    {
+      "epoch": 0.25824972129319956,
+      "grad_norm": 0.4993075430393219,
+      "learning_rate": 8e-05,
+      "loss": 1.8401,
+      "step": 4633
+    },
+    {
+      "epoch": 0.25830546265328874,
+      "grad_norm": 0.5148491263389587,
+      "learning_rate": 8e-05,
+      "loss": 1.7549,
+      "step": 4634
+    },
+    {
+      "epoch": 0.2583612040133779,
+      "grad_norm": 0.5420222282409668,
+      "learning_rate": 8e-05,
+      "loss": 1.7288,
+      "step": 4635
+    },
+    {
+      "epoch": 0.2584169453734671,
+      "grad_norm": 0.5096121430397034,
+      "learning_rate": 8e-05,
+      "loss": 1.7989,
+      "step": 4636
+    },
+    {
+      "epoch": 0.2584726867335563,
+      "grad_norm": 0.5139524936676025,
+      "learning_rate": 8e-05,
+      "loss": 1.7249,
+      "step": 4637
+    },
+    {
+      "epoch": 0.2585284280936455,
+      "grad_norm": 0.5044840574264526,
+      "learning_rate": 8e-05,
+      "loss": 1.6603,
+      "step": 4638
+    },
+    {
+      "epoch": 0.2585841694537347,
+      "grad_norm": 0.5093978047370911,
+      "learning_rate": 8e-05,
+      "loss": 1.7897,
+      "step": 4639
+    },
+    {
+      "epoch": 0.25863991081382387,
+      "grad_norm": 0.5032713413238525,
+      "learning_rate": 8e-05,
+      "loss": 1.672,
+      "step": 4640
+    },
+    {
+      "epoch": 0.25869565217391305,
+      "grad_norm": 0.5168723464012146,
+      "learning_rate": 8e-05,
+      "loss": 1.5844,
+      "step": 4641
+    },
+    {
+      "epoch": 0.2587513935340022,
+      "grad_norm": 0.5161342620849609,
+      "learning_rate": 8e-05,
+      "loss": 1.7342,
+      "step": 4642
+    },
+    {
+      "epoch": 0.2588071348940914,
+      "grad_norm": 0.5017462968826294,
+      "learning_rate": 8e-05,
+      "loss": 1.7623,
+      "step": 4643
+    },
+    {
+      "epoch": 0.2588628762541806,
+      "grad_norm": 0.49921777844429016,
+      "learning_rate": 8e-05,
+      "loss": 1.6982,
+      "step": 4644
+    },
+    {
+      "epoch": 0.25891861761426976,
+      "grad_norm": 0.47593289613723755,
+      "learning_rate": 8e-05,
+      "loss": 1.6082,
+      "step": 4645
+    },
+    {
+      "epoch": 0.258974358974359,
+      "grad_norm": 0.5113307237625122,
+      "learning_rate": 8e-05,
+      "loss": 1.7108,
+      "step": 4646
+    },
+    {
+      "epoch": 0.2590301003344482,
+      "grad_norm": 0.49452194571495056,
+      "learning_rate": 8e-05,
+      "loss": 1.699,
+      "step": 4647
+    },
+    {
+      "epoch": 0.25908584169453736,
+      "grad_norm": 0.4883759319782257,
+      "learning_rate": 8e-05,
+      "loss": 1.6703,
+      "step": 4648
+    },
+    {
+      "epoch": 0.25914158305462653,
+      "grad_norm": 0.5027516484260559,
+      "learning_rate": 8e-05,
+      "loss": 1.713,
+      "step": 4649
+    },
+    {
+      "epoch": 0.2591973244147157,
+      "grad_norm": 0.4828946888446808,
+      "learning_rate": 8e-05,
+      "loss": 1.5637,
+      "step": 4650
+    },
+    {
+      "epoch": 0.2592530657748049,
+      "grad_norm": 0.4641382396221161,
+      "learning_rate": 8e-05,
+      "loss": 1.4836,
+      "step": 4651
+    },
+    {
+      "epoch": 0.25930880713489407,
+      "grad_norm": 0.5482789278030396,
+      "learning_rate": 8e-05,
+      "loss": 1.885,
+      "step": 4652
+    },
+    {
+      "epoch": 0.2593645484949833,
+      "grad_norm": 0.4978908896446228,
+      "learning_rate": 8e-05,
+      "loss": 1.5631,
+      "step": 4653
+    },
+    {
+      "epoch": 0.2594202898550725,
+      "grad_norm": 0.4770525097846985,
+      "learning_rate": 8e-05,
+      "loss": 1.6258,
+      "step": 4654
+    },
+    {
+      "epoch": 0.25947603121516166,
+      "grad_norm": 0.5395385026931763,
+      "learning_rate": 8e-05,
+      "loss": 1.7133,
+      "step": 4655
+    },
+    {
+      "epoch": 0.25953177257525084,
+      "grad_norm": 0.5239088535308838,
+      "learning_rate": 8e-05,
+      "loss": 1.8705,
+      "step": 4656
+    },
+    {
+      "epoch": 0.25958751393534,
+      "grad_norm": 0.45129743218421936,
+      "learning_rate": 8e-05,
+      "loss": 1.5612,
+      "step": 4657
+    },
+    {
+      "epoch": 0.2596432552954292,
+      "grad_norm": 0.5843803286552429,
+      "learning_rate": 8e-05,
+      "loss": 1.9104,
+      "step": 4658
+    },
+    {
+      "epoch": 0.2596989966555184,
+      "grad_norm": 0.5215731263160706,
+      "learning_rate": 8e-05,
+      "loss": 1.7122,
+      "step": 4659
+    },
+    {
+      "epoch": 0.25975473801560756,
+      "grad_norm": 0.4720430076122284,
+      "learning_rate": 8e-05,
+      "loss": 1.7054,
+      "step": 4660
+    },
+    {
+      "epoch": 0.2598104793756968,
+      "grad_norm": 0.5192232728004456,
+      "learning_rate": 8e-05,
+      "loss": 1.6112,
+      "step": 4661
+    },
+    {
+      "epoch": 0.25986622073578597,
+      "grad_norm": 0.5454217195510864,
+      "learning_rate": 8e-05,
+      "loss": 1.7424,
+      "step": 4662
+    },
+    {
+      "epoch": 0.25992196209587515,
+      "grad_norm": 0.5033514499664307,
+      "learning_rate": 8e-05,
+      "loss": 1.6306,
+      "step": 4663
+    },
+    {
+      "epoch": 0.2599777034559643,
+      "grad_norm": 0.48574110865592957,
+      "learning_rate": 8e-05,
+      "loss": 1.6805,
+      "step": 4664
+    },
+    {
+      "epoch": 0.2600334448160535,
+      "grad_norm": 0.5283193588256836,
+      "learning_rate": 8e-05,
+      "loss": 1.5043,
+      "step": 4665
+    },
+    {
+      "epoch": 0.2600891861761427,
+      "grad_norm": 0.4874545931816101,
+      "learning_rate": 8e-05,
+      "loss": 1.552,
+      "step": 4666
+    },
+    {
+      "epoch": 0.26014492753623186,
+      "grad_norm": 0.536874532699585,
+      "learning_rate": 8e-05,
+      "loss": 1.7943,
+      "step": 4667
+    },
+    {
+      "epoch": 0.2602006688963211,
+      "grad_norm": 0.5158773064613342,
+      "learning_rate": 8e-05,
+      "loss": 1.7771,
+      "step": 4668
+    },
+    {
+      "epoch": 0.2602564102564103,
+      "grad_norm": 0.5112432837486267,
+      "learning_rate": 8e-05,
+      "loss": 1.9079,
+      "step": 4669
+    },
+    {
+      "epoch": 0.26031215161649945,
+      "grad_norm": 0.5485742092132568,
+      "learning_rate": 8e-05,
+      "loss": 1.8392,
+      "step": 4670
+    },
+    {
+      "epoch": 0.26036789297658863,
+      "grad_norm": 0.5091162919998169,
+      "learning_rate": 8e-05,
+      "loss": 1.5834,
+      "step": 4671
+    },
+    {
+      "epoch": 0.2604236343366778,
+      "grad_norm": 0.45446982979774475,
+      "learning_rate": 8e-05,
+      "loss": 1.6046,
+      "step": 4672
+    },
+    {
+      "epoch": 0.260479375696767,
+      "grad_norm": 0.5022279620170593,
+      "learning_rate": 8e-05,
+      "loss": 1.6982,
+      "step": 4673
+    },
+    {
+      "epoch": 0.26053511705685617,
+      "grad_norm": 0.4997698962688446,
+      "learning_rate": 8e-05,
+      "loss": 1.6421,
+      "step": 4674
+    },
+    {
+      "epoch": 0.26059085841694535,
+      "grad_norm": 0.49530526995658875,
+      "learning_rate": 8e-05,
+      "loss": 1.6325,
+      "step": 4675
+    },
+    {
+      "epoch": 0.2606465997770346,
+      "grad_norm": 0.490505576133728,
+      "learning_rate": 8e-05,
+      "loss": 1.6641,
+      "step": 4676
+    },
+    {
+      "epoch": 0.26070234113712376,
+      "grad_norm": 0.5120657086372375,
+      "learning_rate": 8e-05,
+      "loss": 1.6395,
+      "step": 4677
+    },
+    {
+      "epoch": 0.26075808249721294,
+      "grad_norm": 0.5384310483932495,
+      "learning_rate": 8e-05,
+      "loss": 1.7659,
+      "step": 4678
+    },
+    {
+      "epoch": 0.2608138238573021,
+      "grad_norm": 0.5185050964355469,
+      "learning_rate": 8e-05,
+      "loss": 1.6108,
+      "step": 4679
+    },
+    {
+      "epoch": 0.2608695652173913,
+      "grad_norm": 0.5490922331809998,
+      "learning_rate": 8e-05,
+      "loss": 1.8128,
+      "step": 4680
+    },
+    {
+      "epoch": 0.2609253065774805,
+      "grad_norm": 0.4805544316768646,
+      "learning_rate": 8e-05,
+      "loss": 1.6368,
+      "step": 4681
+    },
+    {
+      "epoch": 0.26098104793756965,
+      "grad_norm": 0.4884585440158844,
+      "learning_rate": 8e-05,
+      "loss": 1.534,
+      "step": 4682
+    },
+    {
+      "epoch": 0.2610367892976589,
+      "grad_norm": 0.4098069667816162,
+      "learning_rate": 8e-05,
+      "loss": 1.2498,
+      "step": 4683
+    },
+    {
+      "epoch": 0.26109253065774807,
+      "grad_norm": 0.546772301197052,
+      "learning_rate": 8e-05,
+      "loss": 1.673,
+      "step": 4684
+    },
+    {
+      "epoch": 0.26114827201783725,
+      "grad_norm": 0.5115727186203003,
+      "learning_rate": 8e-05,
+      "loss": 1.6959,
+      "step": 4685
+    },
+    {
+      "epoch": 0.2612040133779264,
+      "grad_norm": 0.5167112350463867,
+      "learning_rate": 8e-05,
+      "loss": 1.9362,
+      "step": 4686
+    },
+    {
+      "epoch": 0.2612597547380156,
+      "grad_norm": 0.5140003561973572,
+      "learning_rate": 8e-05,
+      "loss": 1.6821,
+      "step": 4687
+    },
+    {
+      "epoch": 0.2613154960981048,
+      "grad_norm": 0.49699273705482483,
+      "learning_rate": 8e-05,
+      "loss": 1.8037,
+      "step": 4688
+    },
+    {
+      "epoch": 0.26137123745819396,
+      "grad_norm": 0.5371155142784119,
+      "learning_rate": 8e-05,
+      "loss": 1.8324,
+      "step": 4689
+    },
+    {
+      "epoch": 0.26142697881828314,
+      "grad_norm": 0.5561757683753967,
+      "learning_rate": 8e-05,
+      "loss": 1.8558,
+      "step": 4690
+    },
+    {
+      "epoch": 0.2614827201783724,
+      "grad_norm": 0.5155370831489563,
+      "learning_rate": 8e-05,
+      "loss": 1.6131,
+      "step": 4691
+    },
+    {
+      "epoch": 0.26153846153846155,
+      "grad_norm": 0.467456191778183,
+      "learning_rate": 8e-05,
+      "loss": 1.515,
+      "step": 4692
+    },
+    {
+      "epoch": 0.26159420289855073,
+      "grad_norm": 0.5295583009719849,
+      "learning_rate": 8e-05,
+      "loss": 1.5421,
+      "step": 4693
+    },
+    {
+      "epoch": 0.2616499442586399,
+      "grad_norm": 0.5042378306388855,
+      "learning_rate": 8e-05,
+      "loss": 1.6345,
+      "step": 4694
+    },
+    {
+      "epoch": 0.2617056856187291,
+      "grad_norm": 0.5297505855560303,
+      "learning_rate": 8e-05,
+      "loss": 1.9112,
+      "step": 4695
+    },
+    {
+      "epoch": 0.26176142697881827,
+      "grad_norm": 0.5251696109771729,
+      "learning_rate": 8e-05,
+      "loss": 2.0348,
+      "step": 4696
+    },
+    {
+      "epoch": 0.26181716833890745,
+      "grad_norm": 0.5099309086799622,
+      "learning_rate": 8e-05,
+      "loss": 1.8421,
+      "step": 4697
+    },
+    {
+      "epoch": 0.2618729096989967,
+      "grad_norm": 0.502705454826355,
+      "learning_rate": 8e-05,
+      "loss": 1.5744,
+      "step": 4698
+    },
+    {
+      "epoch": 0.26192865105908586,
+      "grad_norm": 0.5405106544494629,
+      "learning_rate": 8e-05,
+      "loss": 2.0006,
+      "step": 4699
+    },
+    {
+      "epoch": 0.26198439241917504,
+      "grad_norm": 0.5157017707824707,
+      "learning_rate": 8e-05,
+      "loss": 1.8455,
+      "step": 4700
+    },
+    {
+      "epoch": 0.2620401337792642,
+      "grad_norm": 0.5418422222137451,
+      "learning_rate": 8e-05,
+      "loss": 1.9059,
+      "step": 4701
+    },
+    {
+      "epoch": 0.2620958751393534,
+      "grad_norm": 0.5688309669494629,
+      "learning_rate": 8e-05,
+      "loss": 2.0853,
+      "step": 4702
+    },
+    {
+      "epoch": 0.2621516164994426,
+      "grad_norm": 0.5489581227302551,
+      "learning_rate": 8e-05,
+      "loss": 2.0183,
+      "step": 4703
+    },
+    {
+      "epoch": 0.26220735785953175,
+      "grad_norm": 0.49559980630874634,
+      "learning_rate": 8e-05,
+      "loss": 1.7079,
+      "step": 4704
+    },
+    {
+      "epoch": 0.26226309921962093,
+      "grad_norm": 0.5063319802284241,
+      "learning_rate": 8e-05,
+      "loss": 1.7444,
+      "step": 4705
+    },
+    {
+      "epoch": 0.26231884057971017,
+      "grad_norm": 0.47826236486434937,
+      "learning_rate": 8e-05,
+      "loss": 1.7103,
+      "step": 4706
+    },
+    {
+      "epoch": 0.26237458193979935,
+      "grad_norm": 0.5195013880729675,
+      "learning_rate": 8e-05,
+      "loss": 1.5908,
+      "step": 4707
+    },
+    {
+      "epoch": 0.2624303232998885,
+      "grad_norm": 0.5257779359817505,
+      "learning_rate": 8e-05,
+      "loss": 1.856,
+      "step": 4708
+    },
+    {
+      "epoch": 0.2624860646599777,
+      "grad_norm": 0.5113489627838135,
+      "learning_rate": 8e-05,
+      "loss": 1.9632,
+      "step": 4709
+    },
+    {
+      "epoch": 0.2625418060200669,
+      "grad_norm": 0.45053985714912415,
+      "learning_rate": 8e-05,
+      "loss": 1.5724,
+      "step": 4710
+    },
+    {
+      "epoch": 0.26259754738015606,
+      "grad_norm": 0.5337340831756592,
+      "learning_rate": 8e-05,
+      "loss": 1.8555,
+      "step": 4711
+    },
+    {
+      "epoch": 0.26265328874024524,
+      "grad_norm": 0.511391282081604,
+      "learning_rate": 8e-05,
+      "loss": 1.7864,
+      "step": 4712
+    },
+    {
+      "epoch": 0.2627090301003345,
+      "grad_norm": 0.47653481364250183,
+      "learning_rate": 8e-05,
+      "loss": 1.5625,
+      "step": 4713
+    },
+    {
+      "epoch": 0.26276477146042365,
+      "grad_norm": 0.4452597200870514,
+      "learning_rate": 8e-05,
+      "loss": 1.7052,
+      "step": 4714
+    },
+    {
+      "epoch": 0.26282051282051283,
+      "grad_norm": 0.5186541676521301,
+      "learning_rate": 8e-05,
+      "loss": 1.7116,
+      "step": 4715
+    },
+    {
+      "epoch": 0.262876254180602,
+      "grad_norm": 0.5272562503814697,
+      "learning_rate": 8e-05,
+      "loss": 1.497,
+      "step": 4716
+    },
+    {
+      "epoch": 0.2629319955406912,
+      "grad_norm": 0.5084412097930908,
+      "learning_rate": 8e-05,
+      "loss": 1.6309,
+      "step": 4717
+    },
+    {
+      "epoch": 0.26298773690078037,
+      "grad_norm": 0.46219387650489807,
+      "learning_rate": 8e-05,
+      "loss": 1.6597,
+      "step": 4718
+    },
+    {
+      "epoch": 0.26304347826086955,
+      "grad_norm": 0.6059833765029907,
+      "learning_rate": 8e-05,
+      "loss": 1.8823,
+      "step": 4719
+    },
+    {
+      "epoch": 0.2630992196209587,
+      "grad_norm": 0.5173678994178772,
+      "learning_rate": 8e-05,
+      "loss": 1.9587,
+      "step": 4720
+    },
+    {
+      "epoch": 0.26315496098104796,
+      "grad_norm": 0.488073468208313,
+      "learning_rate": 8e-05,
+      "loss": 1.6157,
+      "step": 4721
+    },
+    {
+      "epoch": 0.26321070234113714,
+      "grad_norm": 0.5277428030967712,
+      "learning_rate": 8e-05,
+      "loss": 1.9368,
+      "step": 4722
+    },
+    {
+      "epoch": 0.2632664437012263,
+      "grad_norm": 0.5465556383132935,
+      "learning_rate": 8e-05,
+      "loss": 1.766,
+      "step": 4723
+    },
+    {
+      "epoch": 0.2633221850613155,
+      "grad_norm": 0.5116319060325623,
+      "learning_rate": 8e-05,
+      "loss": 1.7194,
+      "step": 4724
+    },
+    {
+      "epoch": 0.2633779264214047,
+      "grad_norm": 0.5023199915885925,
+      "learning_rate": 8e-05,
+      "loss": 1.5882,
+      "step": 4725
+    },
+    {
+      "epoch": 0.26343366778149385,
+      "grad_norm": 0.4644854664802551,
+      "learning_rate": 8e-05,
+      "loss": 1.7469,
+      "step": 4726
+    },
+    {
+      "epoch": 0.26348940914158303,
+      "grad_norm": 0.5060609579086304,
+      "learning_rate": 8e-05,
+      "loss": 1.7745,
+      "step": 4727
+    },
+    {
+      "epoch": 0.26354515050167227,
+      "grad_norm": 0.5291848182678223,
+      "learning_rate": 8e-05,
+      "loss": 1.8109,
+      "step": 4728
+    },
+    {
+      "epoch": 0.26360089186176144,
+      "grad_norm": 0.5288201570510864,
+      "learning_rate": 8e-05,
+      "loss": 1.7741,
+      "step": 4729
+    },
+    {
+      "epoch": 0.2636566332218506,
+      "grad_norm": 0.5466488599777222,
+      "learning_rate": 8e-05,
+      "loss": 1.901,
+      "step": 4730
+    },
+    {
+      "epoch": 0.2637123745819398,
+      "grad_norm": 0.5276413559913635,
+      "learning_rate": 8e-05,
+      "loss": 1.5624,
+      "step": 4731
+    },
+    {
+      "epoch": 0.263768115942029,
+      "grad_norm": 0.5265539288520813,
+      "learning_rate": 8e-05,
+      "loss": 1.8041,
+      "step": 4732
+    },
+    {
+      "epoch": 0.26382385730211816,
+      "grad_norm": 0.4656563997268677,
+      "learning_rate": 8e-05,
+      "loss": 1.5169,
+      "step": 4733
+    },
+    {
+      "epoch": 0.26387959866220734,
+      "grad_norm": 0.5088443160057068,
+      "learning_rate": 8e-05,
+      "loss": 1.7473,
+      "step": 4734
+    },
+    {
+      "epoch": 0.26393534002229657,
+      "grad_norm": 0.5077308416366577,
+      "learning_rate": 8e-05,
+      "loss": 1.6796,
+      "step": 4735
+    },
+    {
+      "epoch": 0.26399108138238575,
+      "grad_norm": 0.45367294549942017,
+      "learning_rate": 8e-05,
+      "loss": 1.5232,
+      "step": 4736
+    },
+    {
+      "epoch": 0.26404682274247493,
+      "grad_norm": 0.5136170983314514,
+      "learning_rate": 8e-05,
+      "loss": 1.7879,
+      "step": 4737
+    },
+    {
+      "epoch": 0.2641025641025641,
+      "grad_norm": 0.5601361393928528,
+      "learning_rate": 8e-05,
+      "loss": 1.8539,
+      "step": 4738
+    },
+    {
+      "epoch": 0.2641583054626533,
+      "grad_norm": 0.49229079484939575,
+      "learning_rate": 8e-05,
+      "loss": 1.6867,
+      "step": 4739
+    },
+    {
+      "epoch": 0.26421404682274247,
+      "grad_norm": 0.4874022305011749,
+      "learning_rate": 8e-05,
+      "loss": 1.6835,
+      "step": 4740
+    },
+    {
+      "epoch": 0.26426978818283164,
+      "grad_norm": 0.4711064398288727,
+      "learning_rate": 8e-05,
+      "loss": 1.4715,
+      "step": 4741
+    },
+    {
+      "epoch": 0.2643255295429208,
+      "grad_norm": 0.4626575708389282,
+      "learning_rate": 8e-05,
+      "loss": 1.3407,
+      "step": 4742
+    },
+    {
+      "epoch": 0.26438127090301006,
+      "grad_norm": 0.5319688320159912,
+      "learning_rate": 8e-05,
+      "loss": 1.7915,
+      "step": 4743
+    },
+    {
+      "epoch": 0.26443701226309924,
+      "grad_norm": 0.5631923079490662,
+      "learning_rate": 8e-05,
+      "loss": 1.8325,
+      "step": 4744
+    },
+    {
+      "epoch": 0.2644927536231884,
+      "grad_norm": 0.5148433446884155,
+      "learning_rate": 8e-05,
+      "loss": 1.6658,
+      "step": 4745
+    },
+    {
+      "epoch": 0.2645484949832776,
+      "grad_norm": 0.5387375950813293,
+      "learning_rate": 8e-05,
+      "loss": 1.7299,
+      "step": 4746
+    },
+    {
+      "epoch": 0.26460423634336677,
+      "grad_norm": 0.48994046449661255,
+      "learning_rate": 8e-05,
+      "loss": 1.7253,
+      "step": 4747
+    },
+    {
+      "epoch": 0.26465997770345595,
+      "grad_norm": 0.4521850049495697,
+      "learning_rate": 8e-05,
+      "loss": 1.5103,
+      "step": 4748
+    },
+    {
+      "epoch": 0.26471571906354513,
+      "grad_norm": 0.49143075942993164,
+      "learning_rate": 8e-05,
+      "loss": 1.5406,
+      "step": 4749
+    },
+    {
+      "epoch": 0.26477146042363436,
+      "grad_norm": 0.5273923873901367,
+      "learning_rate": 8e-05,
+      "loss": 1.7373,
+      "step": 4750
+    },
+    {
+      "epoch": 0.26482720178372354,
+      "grad_norm": 0.513289749622345,
+      "learning_rate": 8e-05,
+      "loss": 1.6385,
+      "step": 4751
+    },
+    {
+      "epoch": 0.2648829431438127,
+      "grad_norm": 0.5224297642707825,
+      "learning_rate": 8e-05,
+      "loss": 1.8056,
+      "step": 4752
+    },
+    {
+      "epoch": 0.2649386845039019,
+      "grad_norm": 0.449360728263855,
+      "learning_rate": 8e-05,
+      "loss": 1.3766,
+      "step": 4753
+    },
+    {
+      "epoch": 0.2649944258639911,
+      "grad_norm": 0.5372068285942078,
+      "learning_rate": 8e-05,
+      "loss": 1.7699,
+      "step": 4754
+    },
+    {
+      "epoch": 0.26505016722408026,
+      "grad_norm": 0.523019552230835,
+      "learning_rate": 8e-05,
+      "loss": 1.6648,
+      "step": 4755
+    },
+    {
+      "epoch": 0.26510590858416944,
+      "grad_norm": 0.5103756785392761,
+      "learning_rate": 8e-05,
+      "loss": 1.7959,
+      "step": 4756
+    },
+    {
+      "epoch": 0.2651616499442586,
+      "grad_norm": 0.8579369783401489,
+      "learning_rate": 8e-05,
+      "loss": 1.6392,
+      "step": 4757
+    },
+    {
+      "epoch": 0.26521739130434785,
+      "grad_norm": 0.550744891166687,
+      "learning_rate": 8e-05,
+      "loss": 1.7364,
+      "step": 4758
+    },
+    {
+      "epoch": 0.26527313266443703,
+      "grad_norm": 0.5084554553031921,
+      "learning_rate": 8e-05,
+      "loss": 1.6298,
+      "step": 4759
+    },
+    {
+      "epoch": 0.2653288740245262,
+      "grad_norm": 0.6770187616348267,
+      "learning_rate": 8e-05,
+      "loss": 1.5945,
+      "step": 4760
+    },
+    {
+      "epoch": 0.2653846153846154,
+      "grad_norm": 0.49051225185394287,
+      "learning_rate": 8e-05,
+      "loss": 1.5087,
+      "step": 4761
+    },
+    {
+      "epoch": 0.26544035674470456,
+      "grad_norm": 0.5356156826019287,
+      "learning_rate": 8e-05,
+      "loss": 1.64,
+      "step": 4762
+    },
+    {
+      "epoch": 0.26549609810479374,
+      "grad_norm": 0.5299071669578552,
+      "learning_rate": 8e-05,
+      "loss": 1.6133,
+      "step": 4763
+    },
+    {
+      "epoch": 0.2655518394648829,
+      "grad_norm": 0.48002979159355164,
+      "learning_rate": 8e-05,
+      "loss": 1.4577,
+      "step": 4764
+    },
+    {
+      "epoch": 0.26560758082497216,
+      "grad_norm": 0.48105019330978394,
+      "learning_rate": 8e-05,
+      "loss": 1.6341,
+      "step": 4765
+    },
+    {
+      "epoch": 0.26566332218506133,
+      "grad_norm": 0.5923892259597778,
+      "learning_rate": 8e-05,
+      "loss": 2.0087,
+      "step": 4766
+    },
+    {
+      "epoch": 0.2657190635451505,
+      "grad_norm": 0.48609718680381775,
+      "learning_rate": 8e-05,
+      "loss": 1.5758,
+      "step": 4767
+    },
+    {
+      "epoch": 0.2657748049052397,
+      "grad_norm": 0.49808210134506226,
+      "learning_rate": 8e-05,
+      "loss": 1.6638,
+      "step": 4768
+    },
+    {
+      "epoch": 0.26583054626532887,
+      "grad_norm": 0.4882586598396301,
+      "learning_rate": 8e-05,
+      "loss": 1.579,
+      "step": 4769
+    },
+    {
+      "epoch": 0.26588628762541805,
+      "grad_norm": 0.5722583532333374,
+      "learning_rate": 8e-05,
+      "loss": 1.9595,
+      "step": 4770
+    },
+    {
+      "epoch": 0.26594202898550723,
+      "grad_norm": 0.47601956129074097,
+      "learning_rate": 8e-05,
+      "loss": 1.5847,
+      "step": 4771
+    },
+    {
+      "epoch": 0.2659977703455964,
+      "grad_norm": 0.4781881868839264,
+      "learning_rate": 8e-05,
+      "loss": 1.61,
+      "step": 4772
+    },
+    {
+      "epoch": 0.26605351170568564,
+      "grad_norm": 0.5992628931999207,
+      "learning_rate": 8e-05,
+      "loss": 1.7718,
+      "step": 4773
+    },
+    {
+      "epoch": 0.2661092530657748,
+      "grad_norm": 0.49576088786125183,
+      "learning_rate": 8e-05,
+      "loss": 1.7456,
+      "step": 4774
+    },
+    {
+      "epoch": 0.266164994425864,
+      "grad_norm": 0.49744564294815063,
+      "learning_rate": 8e-05,
+      "loss": 1.6193,
+      "step": 4775
+    },
+    {
+      "epoch": 0.2662207357859532,
+      "grad_norm": 0.518963634967804,
+      "learning_rate": 8e-05,
+      "loss": 1.7641,
+      "step": 4776
+    },
+    {
+      "epoch": 0.26627647714604236,
+      "grad_norm": 0.48951300978660583,
+      "learning_rate": 8e-05,
+      "loss": 1.6035,
+      "step": 4777
+    },
+    {
+      "epoch": 0.26633221850613154,
+      "grad_norm": 0.5045288801193237,
+      "learning_rate": 8e-05,
+      "loss": 1.7325,
+      "step": 4778
+    },
+    {
+      "epoch": 0.2663879598662207,
+      "grad_norm": 0.5186219811439514,
+      "learning_rate": 8e-05,
+      "loss": 1.7007,
+      "step": 4779
+    },
+    {
+      "epoch": 0.26644370122630995,
+      "grad_norm": 0.5073248744010925,
+      "learning_rate": 8e-05,
+      "loss": 1.7715,
+      "step": 4780
+    },
+    {
+      "epoch": 0.2664994425863991,
+      "grad_norm": 0.5213096141815186,
+      "learning_rate": 8e-05,
+      "loss": 1.684,
+      "step": 4781
+    },
+    {
+      "epoch": 0.2665551839464883,
+      "grad_norm": 0.47361043095588684,
+      "learning_rate": 8e-05,
+      "loss": 1.6807,
+      "step": 4782
+    },
+    {
+      "epoch": 0.2666109253065775,
+      "grad_norm": 0.48980844020843506,
+      "learning_rate": 8e-05,
+      "loss": 1.7931,
+      "step": 4783
+    },
+    {
+      "epoch": 0.26666666666666666,
+      "grad_norm": 0.48362529277801514,
+      "learning_rate": 8e-05,
+      "loss": 1.639,
+      "step": 4784
+    },
+    {
+      "epoch": 0.26672240802675584,
+      "grad_norm": 0.5443851947784424,
+      "learning_rate": 8e-05,
+      "loss": 1.8447,
+      "step": 4785
+    },
+    {
+      "epoch": 0.266778149386845,
+      "grad_norm": 0.4934808015823364,
+      "learning_rate": 8e-05,
+      "loss": 1.8289,
+      "step": 4786
+    },
+    {
+      "epoch": 0.2668338907469342,
+      "grad_norm": 0.5453766584396362,
+      "learning_rate": 8e-05,
+      "loss": 1.9642,
+      "step": 4787
+    },
+    {
+      "epoch": 0.26688963210702343,
+      "grad_norm": 0.46825507283210754,
+      "learning_rate": 8e-05,
+      "loss": 1.5369,
+      "step": 4788
+    },
+    {
+      "epoch": 0.2669453734671126,
+      "grad_norm": 0.5072756409645081,
+      "learning_rate": 8e-05,
+      "loss": 1.5771,
+      "step": 4789
+    },
+    {
+      "epoch": 0.2670011148272018,
+      "grad_norm": 0.5206819176673889,
+      "learning_rate": 8e-05,
+      "loss": 1.8058,
+      "step": 4790
+    },
+    {
+      "epoch": 0.26705685618729097,
+      "grad_norm": 0.47623395919799805,
+      "learning_rate": 8e-05,
+      "loss": 1.7826,
+      "step": 4791
+    },
+    {
+      "epoch": 0.26711259754738015,
+      "grad_norm": 0.5042856335639954,
+      "learning_rate": 8e-05,
+      "loss": 1.8121,
+      "step": 4792
+    },
+    {
+      "epoch": 0.2671683389074693,
+      "grad_norm": 0.48193052411079407,
+      "learning_rate": 8e-05,
+      "loss": 1.7285,
+      "step": 4793
+    },
+    {
+      "epoch": 0.2672240802675585,
+      "grad_norm": 0.4764842689037323,
+      "learning_rate": 8e-05,
+      "loss": 1.7433,
+      "step": 4794
+    },
+    {
+      "epoch": 0.26727982162764774,
+      "grad_norm": 0.5621132254600525,
+      "learning_rate": 8e-05,
+      "loss": 1.6013,
+      "step": 4795
+    },
+    {
+      "epoch": 0.2673355629877369,
+      "grad_norm": 0.48127683997154236,
+      "learning_rate": 8e-05,
+      "loss": 1.8514,
+      "step": 4796
+    },
+    {
+      "epoch": 0.2673913043478261,
+      "grad_norm": 0.46192049980163574,
+      "learning_rate": 8e-05,
+      "loss": 1.7205,
+      "step": 4797
+    },
+    {
+      "epoch": 0.2674470457079153,
+      "grad_norm": 0.46006837487220764,
+      "learning_rate": 8e-05,
+      "loss": 1.5512,
+      "step": 4798
+    },
+    {
+      "epoch": 0.26750278706800446,
+      "grad_norm": 0.4827045500278473,
+      "learning_rate": 8e-05,
+      "loss": 1.6037,
+      "step": 4799
+    },
+    {
+      "epoch": 0.26755852842809363,
+      "grad_norm": 0.534766435623169,
+      "learning_rate": 8e-05,
+      "loss": 1.4909,
+      "step": 4800
+    },
+    {
+      "epoch": 0.2676142697881828,
+      "grad_norm": 0.516715943813324,
+      "learning_rate": 8e-05,
+      "loss": 1.8948,
+      "step": 4801
+    },
+    {
+      "epoch": 0.267670011148272,
+      "grad_norm": 0.5046938061714172,
+      "learning_rate": 8e-05,
+      "loss": 1.7482,
+      "step": 4802
+    },
+    {
+      "epoch": 0.2677257525083612,
+      "grad_norm": 0.4876144826412201,
+      "learning_rate": 8e-05,
+      "loss": 1.6736,
+      "step": 4803
+    },
+    {
+      "epoch": 0.2677814938684504,
+      "grad_norm": 0.5733070969581604,
+      "learning_rate": 8e-05,
+      "loss": 1.9056,
+      "step": 4804
+    },
+    {
+      "epoch": 0.2678372352285396,
+      "grad_norm": 0.5142390727996826,
+      "learning_rate": 8e-05,
+      "loss": 1.648,
+      "step": 4805
+    },
+    {
+      "epoch": 0.26789297658862876,
+      "grad_norm": 0.49695613980293274,
+      "learning_rate": 8e-05,
+      "loss": 1.715,
+      "step": 4806
+    },
+    {
+      "epoch": 0.26794871794871794,
+      "grad_norm": 0.4340948462486267,
+      "learning_rate": 8e-05,
+      "loss": 1.156,
+      "step": 4807
+    },
+    {
+      "epoch": 0.2680044593088071,
+      "grad_norm": 0.47531047463417053,
+      "learning_rate": 8e-05,
+      "loss": 1.5231,
+      "step": 4808
+    },
+    {
+      "epoch": 0.2680602006688963,
+      "grad_norm": 0.48843592405319214,
+      "learning_rate": 8e-05,
+      "loss": 1.5196,
+      "step": 4809
+    },
+    {
+      "epoch": 0.26811594202898553,
+      "grad_norm": 0.5055053234100342,
+      "learning_rate": 8e-05,
+      "loss": 1.5967,
+      "step": 4810
+    },
+    {
+      "epoch": 0.2681716833890747,
+      "grad_norm": 0.5026859641075134,
+      "learning_rate": 8e-05,
+      "loss": 1.8444,
+      "step": 4811
+    },
+    {
+      "epoch": 0.2682274247491639,
+      "grad_norm": 0.4923660159111023,
+      "learning_rate": 8e-05,
+      "loss": 1.6562,
+      "step": 4812
+    },
+    {
+      "epoch": 0.26828316610925307,
+      "grad_norm": 0.5042104721069336,
+      "learning_rate": 8e-05,
+      "loss": 1.3715,
+      "step": 4813
+    },
+    {
+      "epoch": 0.26833890746934225,
+      "grad_norm": 0.4890134930610657,
+      "learning_rate": 8e-05,
+      "loss": 1.7762,
+      "step": 4814
+    },
+    {
+      "epoch": 0.2683946488294314,
+      "grad_norm": 0.5024535655975342,
+      "learning_rate": 8e-05,
+      "loss": 1.4346,
+      "step": 4815
+    },
+    {
+      "epoch": 0.2684503901895206,
+      "grad_norm": 0.5002022981643677,
+      "learning_rate": 8e-05,
+      "loss": 1.6237,
+      "step": 4816
+    },
+    {
+      "epoch": 0.2685061315496098,
+      "grad_norm": 0.4712158441543579,
+      "learning_rate": 8e-05,
+      "loss": 1.5621,
+      "step": 4817
+    },
+    {
+      "epoch": 0.268561872909699,
+      "grad_norm": 0.553865909576416,
+      "learning_rate": 8e-05,
+      "loss": 1.8666,
+      "step": 4818
+    },
+    {
+      "epoch": 0.2686176142697882,
+      "grad_norm": 0.6038681268692017,
+      "learning_rate": 8e-05,
+      "loss": 1.5174,
+      "step": 4819
+    },
+    {
+      "epoch": 0.2686733556298774,
+      "grad_norm": 0.5450118184089661,
+      "learning_rate": 8e-05,
+      "loss": 1.9144,
+      "step": 4820
+    },
+    {
+      "epoch": 0.26872909698996655,
+      "grad_norm": 0.517101526260376,
+      "learning_rate": 8e-05,
+      "loss": 1.7423,
+      "step": 4821
+    },
+    {
+      "epoch": 0.26878483835005573,
+      "grad_norm": 0.5435220003128052,
+      "learning_rate": 8e-05,
+      "loss": 1.7849,
+      "step": 4822
+    },
+    {
+      "epoch": 0.2688405797101449,
+      "grad_norm": 0.5636661648750305,
+      "learning_rate": 8e-05,
+      "loss": 1.9104,
+      "step": 4823
+    },
+    {
+      "epoch": 0.2688963210702341,
+      "grad_norm": 0.4907018840312958,
+      "learning_rate": 8e-05,
+      "loss": 1.8114,
+      "step": 4824
+    },
+    {
+      "epoch": 0.2689520624303233,
+      "grad_norm": 0.4901597499847412,
+      "learning_rate": 8e-05,
+      "loss": 1.4494,
+      "step": 4825
+    },
+    {
+      "epoch": 0.2690078037904125,
+      "grad_norm": 0.5196972489356995,
+      "learning_rate": 8e-05,
+      "loss": 1.8353,
+      "step": 4826
+    },
+    {
+      "epoch": 0.2690635451505017,
+      "grad_norm": 0.4989320933818817,
+      "learning_rate": 8e-05,
+      "loss": 1.7714,
+      "step": 4827
+    },
+    {
+      "epoch": 0.26911928651059086,
+      "grad_norm": 0.4761400818824768,
+      "learning_rate": 8e-05,
+      "loss": 1.6661,
+      "step": 4828
+    },
+    {
+      "epoch": 0.26917502787068004,
+      "grad_norm": 0.4977497160434723,
+      "learning_rate": 8e-05,
+      "loss": 1.5345,
+      "step": 4829
+    },
+    {
+      "epoch": 0.2692307692307692,
+      "grad_norm": 0.6041051745414734,
+      "learning_rate": 8e-05,
+      "loss": 1.8069,
+      "step": 4830
+    },
+    {
+      "epoch": 0.2692865105908584,
+      "grad_norm": 0.5953145027160645,
+      "learning_rate": 8e-05,
+      "loss": 2.0057,
+      "step": 4831
+    },
+    {
+      "epoch": 0.2693422519509476,
+      "grad_norm": 0.4859643280506134,
+      "learning_rate": 8e-05,
+      "loss": 1.8089,
+      "step": 4832
+    },
+    {
+      "epoch": 0.2693979933110368,
+      "grad_norm": 0.5140557289123535,
+      "learning_rate": 8e-05,
+      "loss": 1.7321,
+      "step": 4833
+    },
+    {
+      "epoch": 0.269453734671126,
+      "grad_norm": 0.561446487903595,
+      "learning_rate": 8e-05,
+      "loss": 1.7825,
+      "step": 4834
+    },
+    {
+      "epoch": 0.26950947603121517,
+      "grad_norm": 0.5055059194564819,
+      "learning_rate": 8e-05,
+      "loss": 1.7342,
+      "step": 4835
+    },
+    {
+      "epoch": 0.26956521739130435,
+      "grad_norm": 0.5057458281517029,
+      "learning_rate": 8e-05,
+      "loss": 1.6527,
+      "step": 4836
+    },
+    {
+      "epoch": 0.2696209587513935,
+      "grad_norm": 0.4766155183315277,
+      "learning_rate": 8e-05,
+      "loss": 1.5412,
+      "step": 4837
+    },
+    {
+      "epoch": 0.2696767001114827,
+      "grad_norm": 0.46934545040130615,
+      "learning_rate": 8e-05,
+      "loss": 1.5501,
+      "step": 4838
+    },
+    {
+      "epoch": 0.2697324414715719,
+      "grad_norm": 0.4834466278553009,
+      "learning_rate": 8e-05,
+      "loss": 1.4659,
+      "step": 4839
+    },
+    {
+      "epoch": 0.2697881828316611,
+      "grad_norm": 0.46945199370384216,
+      "learning_rate": 8e-05,
+      "loss": 1.4448,
+      "step": 4840
+    },
+    {
+      "epoch": 0.2698439241917503,
+      "grad_norm": 0.5626839995384216,
+      "learning_rate": 8e-05,
+      "loss": 1.7148,
+      "step": 4841
+    },
+    {
+      "epoch": 0.2698996655518395,
+      "grad_norm": 0.4671670198440552,
+      "learning_rate": 8e-05,
+      "loss": 1.5116,
+      "step": 4842
+    },
+    {
+      "epoch": 0.26995540691192865,
+      "grad_norm": 0.46738243103027344,
+      "learning_rate": 8e-05,
+      "loss": 1.624,
+      "step": 4843
+    },
+    {
+      "epoch": 0.27001114827201783,
+      "grad_norm": 0.49982571601867676,
+      "learning_rate": 8e-05,
+      "loss": 1.6875,
+      "step": 4844
+    },
+    {
+      "epoch": 0.270066889632107,
+      "grad_norm": 0.45273247361183167,
+      "learning_rate": 8e-05,
+      "loss": 1.5005,
+      "step": 4845
+    },
+    {
+      "epoch": 0.2701226309921962,
+      "grad_norm": 0.44871610403060913,
+      "learning_rate": 8e-05,
+      "loss": 1.4288,
+      "step": 4846
+    },
+    {
+      "epoch": 0.2701783723522854,
+      "grad_norm": 0.5660805106163025,
+      "learning_rate": 8e-05,
+      "loss": 1.8574,
+      "step": 4847
+    },
+    {
+      "epoch": 0.2702341137123746,
+      "grad_norm": 0.49467673897743225,
+      "learning_rate": 8e-05,
+      "loss": 1.6284,
+      "step": 4848
+    },
+    {
+      "epoch": 0.2702898550724638,
+      "grad_norm": 0.5443269610404968,
+      "learning_rate": 8e-05,
+      "loss": 1.681,
+      "step": 4849
+    },
+    {
+      "epoch": 0.27034559643255296,
+      "grad_norm": 0.5486834049224854,
+      "learning_rate": 8e-05,
+      "loss": 1.7333,
+      "step": 4850
+    },
+    {
+      "epoch": 0.27040133779264214,
+      "grad_norm": 0.4856089651584625,
+      "learning_rate": 8e-05,
+      "loss": 1.545,
+      "step": 4851
+    },
+    {
+      "epoch": 0.2704570791527313,
+      "grad_norm": 0.5295234322547913,
+      "learning_rate": 8e-05,
+      "loss": 1.6445,
+      "step": 4852
+    },
+    {
+      "epoch": 0.2705128205128205,
+      "grad_norm": 0.5542267560958862,
+      "learning_rate": 8e-05,
+      "loss": 1.7465,
+      "step": 4853
+    },
+    {
+      "epoch": 0.2705685618729097,
+      "grad_norm": 0.43070581555366516,
+      "learning_rate": 8e-05,
+      "loss": 1.286,
+      "step": 4854
+    },
+    {
+      "epoch": 0.2706243032329989,
+      "grad_norm": 0.5551098585128784,
+      "learning_rate": 8e-05,
+      "loss": 1.7018,
+      "step": 4855
+    },
+    {
+      "epoch": 0.2706800445930881,
+      "grad_norm": 0.5335925817489624,
+      "learning_rate": 8e-05,
+      "loss": 1.7035,
+      "step": 4856
+    },
+    {
+      "epoch": 0.27073578595317727,
+      "grad_norm": 0.48526740074157715,
+      "learning_rate": 8e-05,
+      "loss": 1.6353,
+      "step": 4857
+    },
+    {
+      "epoch": 0.27079152731326644,
+      "grad_norm": 0.5087800025939941,
+      "learning_rate": 8e-05,
+      "loss": 1.8624,
+      "step": 4858
+    },
+    {
+      "epoch": 0.2708472686733556,
+      "grad_norm": 0.4876925051212311,
+      "learning_rate": 8e-05,
+      "loss": 1.6672,
+      "step": 4859
+    },
+    {
+      "epoch": 0.2709030100334448,
+      "grad_norm": 0.5388376712799072,
+      "learning_rate": 8e-05,
+      "loss": 1.9496,
+      "step": 4860
+    },
+    {
+      "epoch": 0.270958751393534,
+      "grad_norm": 0.46098411083221436,
+      "learning_rate": 8e-05,
+      "loss": 1.5343,
+      "step": 4861
+    },
+    {
+      "epoch": 0.2710144927536232,
+      "grad_norm": 0.4828426241874695,
+      "learning_rate": 8e-05,
+      "loss": 1.5259,
+      "step": 4862
+    },
+    {
+      "epoch": 0.2710702341137124,
+      "grad_norm": 0.5513729453086853,
+      "learning_rate": 8e-05,
+      "loss": 2.01,
+      "step": 4863
+    },
+    {
+      "epoch": 0.2711259754738016,
+      "grad_norm": 0.49254027009010315,
+      "learning_rate": 8e-05,
+      "loss": 1.6194,
+      "step": 4864
+    },
+    {
+      "epoch": 0.27118171683389075,
+      "grad_norm": 0.5209529995918274,
+      "learning_rate": 8e-05,
+      "loss": 1.6279,
+      "step": 4865
+    },
+    {
+      "epoch": 0.27123745819397993,
+      "grad_norm": 0.5383169054985046,
+      "learning_rate": 8e-05,
+      "loss": 1.8516,
+      "step": 4866
+    },
+    {
+      "epoch": 0.2712931995540691,
+      "grad_norm": 0.4927859306335449,
+      "learning_rate": 8e-05,
+      "loss": 1.585,
+      "step": 4867
+    },
+    {
+      "epoch": 0.2713489409141583,
+      "grad_norm": 0.5537318587303162,
+      "learning_rate": 8e-05,
+      "loss": 1.9545,
+      "step": 4868
+    },
+    {
+      "epoch": 0.27140468227424747,
+      "grad_norm": 0.5104342103004456,
+      "learning_rate": 8e-05,
+      "loss": 1.6882,
+      "step": 4869
+    },
+    {
+      "epoch": 0.2714604236343367,
+      "grad_norm": 0.4823477268218994,
+      "learning_rate": 8e-05,
+      "loss": 1.8651,
+      "step": 4870
+    },
+    {
+      "epoch": 0.2715161649944259,
+      "grad_norm": 0.5014148950576782,
+      "learning_rate": 8e-05,
+      "loss": 1.9006,
+      "step": 4871
+    },
+    {
+      "epoch": 0.27157190635451506,
+      "grad_norm": 0.4821976125240326,
+      "learning_rate": 8e-05,
+      "loss": 1.6347,
+      "step": 4872
+    },
+    {
+      "epoch": 0.27162764771460424,
+      "grad_norm": 0.5137785077095032,
+      "learning_rate": 8e-05,
+      "loss": 1.8702,
+      "step": 4873
+    },
+    {
+      "epoch": 0.2716833890746934,
+      "grad_norm": 0.4899905323982239,
+      "learning_rate": 8e-05,
+      "loss": 1.7127,
+      "step": 4874
+    },
+    {
+      "epoch": 0.2717391304347826,
+      "grad_norm": 0.5382660627365112,
+      "learning_rate": 8e-05,
+      "loss": 1.683,
+      "step": 4875
+    },
+    {
+      "epoch": 0.2717948717948718,
+      "grad_norm": 0.47894027829170227,
+      "learning_rate": 8e-05,
+      "loss": 1.5703,
+      "step": 4876
+    },
+    {
+      "epoch": 0.271850613154961,
+      "grad_norm": 0.46718770265579224,
+      "learning_rate": 8e-05,
+      "loss": 1.6388,
+      "step": 4877
+    },
+    {
+      "epoch": 0.2719063545150502,
+      "grad_norm": 0.5733405947685242,
+      "learning_rate": 8e-05,
+      "loss": 1.9859,
+      "step": 4878
+    },
+    {
+      "epoch": 0.27196209587513936,
+      "grad_norm": 0.4837845265865326,
+      "learning_rate": 8e-05,
+      "loss": 1.5767,
+      "step": 4879
+    },
+    {
+      "epoch": 0.27201783723522854,
+      "grad_norm": 0.486235111951828,
+      "learning_rate": 8e-05,
+      "loss": 1.4791,
+      "step": 4880
+    },
+    {
+      "epoch": 0.2720735785953177,
+      "grad_norm": 0.48340904712677,
+      "learning_rate": 8e-05,
+      "loss": 1.7882,
+      "step": 4881
+    },
+    {
+      "epoch": 0.2721293199554069,
+      "grad_norm": 0.6102102994918823,
+      "learning_rate": 8e-05,
+      "loss": 1.6115,
+      "step": 4882
+    },
+    {
+      "epoch": 0.2721850613154961,
+      "grad_norm": 0.5262008309364319,
+      "learning_rate": 8e-05,
+      "loss": 1.5027,
+      "step": 4883
+    },
+    {
+      "epoch": 0.27224080267558526,
+      "grad_norm": 0.5147656202316284,
+      "learning_rate": 8e-05,
+      "loss": 1.5817,
+      "step": 4884
+    },
+    {
+      "epoch": 0.2722965440356745,
+      "grad_norm": 0.4796406328678131,
+      "learning_rate": 8e-05,
+      "loss": 1.6345,
+      "step": 4885
+    },
+    {
+      "epoch": 0.27235228539576367,
+      "grad_norm": 0.500372588634491,
+      "learning_rate": 8e-05,
+      "loss": 1.8484,
+      "step": 4886
+    },
+    {
+      "epoch": 0.27240802675585285,
+      "grad_norm": 0.49836209416389465,
+      "learning_rate": 8e-05,
+      "loss": 1.7545,
+      "step": 4887
+    },
+    {
+      "epoch": 0.27246376811594203,
+      "grad_norm": 0.5067594051361084,
+      "learning_rate": 8e-05,
+      "loss": 1.7822,
+      "step": 4888
+    },
+    {
+      "epoch": 0.2725195094760312,
+      "grad_norm": 0.5147234201431274,
+      "learning_rate": 8e-05,
+      "loss": 1.7458,
+      "step": 4889
+    },
+    {
+      "epoch": 0.2725752508361204,
+      "grad_norm": 0.5145500302314758,
+      "learning_rate": 8e-05,
+      "loss": 1.6601,
+      "step": 4890
+    },
+    {
+      "epoch": 0.27263099219620957,
+      "grad_norm": 0.4750419855117798,
+      "learning_rate": 8e-05,
+      "loss": 1.5363,
+      "step": 4891
+    },
+    {
+      "epoch": 0.2726867335562988,
+      "grad_norm": 0.4774273931980133,
+      "learning_rate": 8e-05,
+      "loss": 1.7132,
+      "step": 4892
+    },
+    {
+      "epoch": 0.272742474916388,
+      "grad_norm": 0.5947921276092529,
+      "learning_rate": 8e-05,
+      "loss": 1.5743,
+      "step": 4893
+    },
+    {
+      "epoch": 0.27279821627647716,
+      "grad_norm": 0.5759541988372803,
+      "learning_rate": 8e-05,
+      "loss": 1.9485,
+      "step": 4894
+    },
+    {
+      "epoch": 0.27285395763656634,
+      "grad_norm": 0.5098536014556885,
+      "learning_rate": 8e-05,
+      "loss": 1.8144,
+      "step": 4895
+    },
+    {
+      "epoch": 0.2729096989966555,
+      "grad_norm": 0.48607727885246277,
+      "learning_rate": 8e-05,
+      "loss": 1.6344,
+      "step": 4896
+    },
+    {
+      "epoch": 0.2729654403567447,
+      "grad_norm": 0.48327797651290894,
+      "learning_rate": 8e-05,
+      "loss": 1.5865,
+      "step": 4897
+    },
+    {
+      "epoch": 0.27302118171683387,
+      "grad_norm": 0.5167771577835083,
+      "learning_rate": 8e-05,
+      "loss": 1.5991,
+      "step": 4898
+    },
+    {
+      "epoch": 0.27307692307692305,
+      "grad_norm": 0.5023355484008789,
+      "learning_rate": 8e-05,
+      "loss": 1.5283,
+      "step": 4899
+    },
+    {
+      "epoch": 0.2731326644370123,
+      "grad_norm": 0.49115991592407227,
+      "learning_rate": 8e-05,
+      "loss": 1.6185,
+      "step": 4900
+    },
+    {
+      "epoch": 0.27318840579710146,
+      "grad_norm": 0.49574774503707886,
+      "learning_rate": 8e-05,
+      "loss": 1.6107,
+      "step": 4901
+    },
+    {
+      "epoch": 0.27324414715719064,
+      "grad_norm": 0.5031110048294067,
+      "learning_rate": 8e-05,
+      "loss": 1.804,
+      "step": 4902
+    },
+    {
+      "epoch": 0.2732998885172798,
+      "grad_norm": 0.5140913724899292,
+      "learning_rate": 8e-05,
+      "loss": 1.633,
+      "step": 4903
+    },
+    {
+      "epoch": 0.273355629877369,
+      "grad_norm": 0.5243213176727295,
+      "learning_rate": 8e-05,
+      "loss": 1.7039,
+      "step": 4904
+    },
+    {
+      "epoch": 0.2734113712374582,
+      "grad_norm": 0.5083054304122925,
+      "learning_rate": 8e-05,
+      "loss": 1.7294,
+      "step": 4905
+    },
+    {
+      "epoch": 0.27346711259754736,
+      "grad_norm": 0.5570263862609863,
+      "learning_rate": 8e-05,
+      "loss": 1.5553,
+      "step": 4906
+    },
+    {
+      "epoch": 0.2735228539576366,
+      "grad_norm": 0.5495603680610657,
+      "learning_rate": 8e-05,
+      "loss": 1.4761,
+      "step": 4907
+    },
+    {
+      "epoch": 0.27357859531772577,
+      "grad_norm": 0.5350089073181152,
+      "learning_rate": 8e-05,
+      "loss": 1.8056,
+      "step": 4908
+    },
+    {
+      "epoch": 0.27363433667781495,
+      "grad_norm": 0.4990091025829315,
+      "learning_rate": 8e-05,
+      "loss": 1.7132,
+      "step": 4909
+    },
+    {
+      "epoch": 0.2736900780379041,
+      "grad_norm": 0.5004948973655701,
+      "learning_rate": 8e-05,
+      "loss": 1.6129,
+      "step": 4910
+    },
+    {
+      "epoch": 0.2737458193979933,
+      "grad_norm": 0.48411688208580017,
+      "learning_rate": 8e-05,
+      "loss": 1.5139,
+      "step": 4911
+    },
+    {
+      "epoch": 0.2738015607580825,
+      "grad_norm": 0.511704683303833,
+      "learning_rate": 8e-05,
+      "loss": 1.7031,
+      "step": 4912
+    },
+    {
+      "epoch": 0.27385730211817166,
+      "grad_norm": 0.5102469325065613,
+      "learning_rate": 8e-05,
+      "loss": 1.6607,
+      "step": 4913
+    },
+    {
+      "epoch": 0.27391304347826084,
+      "grad_norm": 0.5354444980621338,
+      "learning_rate": 8e-05,
+      "loss": 1.6374,
+      "step": 4914
+    },
+    {
+      "epoch": 0.2739687848383501,
+      "grad_norm": 0.5184294581413269,
+      "learning_rate": 8e-05,
+      "loss": 1.7372,
+      "step": 4915
+    },
+    {
+      "epoch": 0.27402452619843926,
+      "grad_norm": 0.5330214500427246,
+      "learning_rate": 8e-05,
+      "loss": 1.8916,
+      "step": 4916
+    },
+    {
+      "epoch": 0.27408026755852843,
+      "grad_norm": 0.5457834601402283,
+      "learning_rate": 8e-05,
+      "loss": 1.9441,
+      "step": 4917
+    },
+    {
+      "epoch": 0.2741360089186176,
+      "grad_norm": 0.5252987146377563,
+      "learning_rate": 8e-05,
+      "loss": 1.6714,
+      "step": 4918
+    },
+    {
+      "epoch": 0.2741917502787068,
+      "grad_norm": 0.4930163025856018,
+      "learning_rate": 8e-05,
+      "loss": 1.8325,
+      "step": 4919
+    },
+    {
+      "epoch": 0.27424749163879597,
+      "grad_norm": 0.49244213104248047,
+      "learning_rate": 8e-05,
+      "loss": 1.8733,
+      "step": 4920
+    },
+    {
+      "epoch": 0.27430323299888515,
+      "grad_norm": 0.7317171692848206,
+      "learning_rate": 8e-05,
+      "loss": 1.5259,
+      "step": 4921
+    },
+    {
+      "epoch": 0.2743589743589744,
+      "grad_norm": 0.5334247946739197,
+      "learning_rate": 8e-05,
+      "loss": 1.9318,
+      "step": 4922
+    },
+    {
+      "epoch": 0.27441471571906356,
+      "grad_norm": 0.4989931881427765,
+      "learning_rate": 8e-05,
+      "loss": 1.7103,
+      "step": 4923
+    },
+    {
+      "epoch": 0.27447045707915274,
+      "grad_norm": 0.4885168969631195,
+      "learning_rate": 8e-05,
+      "loss": 1.667,
+      "step": 4924
+    },
+    {
+      "epoch": 0.2745261984392419,
+      "grad_norm": 0.5465773940086365,
+      "learning_rate": 8e-05,
+      "loss": 1.7775,
+      "step": 4925
+    },
+    {
+      "epoch": 0.2745819397993311,
+      "grad_norm": 0.4521493911743164,
+      "learning_rate": 8e-05,
+      "loss": 1.3499,
+      "step": 4926
+    },
+    {
+      "epoch": 0.2746376811594203,
+      "grad_norm": 0.5232551693916321,
+      "learning_rate": 8e-05,
+      "loss": 1.889,
+      "step": 4927
+    },
+    {
+      "epoch": 0.27469342251950946,
+      "grad_norm": 0.5108181834220886,
+      "learning_rate": 8e-05,
+      "loss": 1.9147,
+      "step": 4928
+    },
+    {
+      "epoch": 0.27474916387959863,
+      "grad_norm": 0.500938355922699,
+      "learning_rate": 8e-05,
+      "loss": 1.5475,
+      "step": 4929
+    },
+    {
+      "epoch": 0.27480490523968787,
+      "grad_norm": 0.5091776847839355,
+      "learning_rate": 8e-05,
+      "loss": 1.7221,
+      "step": 4930
+    },
+    {
+      "epoch": 0.27486064659977705,
+      "grad_norm": 0.5064066052436829,
+      "learning_rate": 8e-05,
+      "loss": 1.5555,
+      "step": 4931
+    },
+    {
+      "epoch": 0.2749163879598662,
+      "grad_norm": 0.5267865657806396,
+      "learning_rate": 8e-05,
+      "loss": 1.7589,
+      "step": 4932
+    },
+    {
+      "epoch": 0.2749721293199554,
+      "grad_norm": 0.4712657630443573,
+      "learning_rate": 8e-05,
+      "loss": 1.5217,
+      "step": 4933
+    },
+    {
+      "epoch": 0.2750278706800446,
+      "grad_norm": 0.45334136486053467,
+      "learning_rate": 8e-05,
+      "loss": 1.6751,
+      "step": 4934
+    },
+    {
+      "epoch": 0.27508361204013376,
+      "grad_norm": 0.5091070532798767,
+      "learning_rate": 8e-05,
+      "loss": 1.6838,
+      "step": 4935
+    },
+    {
+      "epoch": 0.27513935340022294,
+      "grad_norm": 0.5040649175643921,
+      "learning_rate": 8e-05,
+      "loss": 1.7316,
+      "step": 4936
+    },
+    {
+      "epoch": 0.2751950947603122,
+      "grad_norm": 0.5236833691596985,
+      "learning_rate": 8e-05,
+      "loss": 1.7511,
+      "step": 4937
+    },
+    {
+      "epoch": 0.27525083612040135,
+      "grad_norm": 0.4728735089302063,
+      "learning_rate": 8e-05,
+      "loss": 1.5224,
+      "step": 4938
+    },
+    {
+      "epoch": 0.27530657748049053,
+      "grad_norm": 0.4994482100009918,
+      "learning_rate": 8e-05,
+      "loss": 1.7859,
+      "step": 4939
+    },
+    {
+      "epoch": 0.2753623188405797,
+      "grad_norm": 0.5201264023780823,
+      "learning_rate": 8e-05,
+      "loss": 1.759,
+      "step": 4940
+    },
+    {
+      "epoch": 0.2754180602006689,
+      "grad_norm": 0.4370453655719757,
+      "learning_rate": 8e-05,
+      "loss": 1.4542,
+      "step": 4941
+    },
+    {
+      "epoch": 0.27547380156075807,
+      "grad_norm": 0.5194417834281921,
+      "learning_rate": 8e-05,
+      "loss": 1.9518,
+      "step": 4942
+    },
+    {
+      "epoch": 0.27552954292084725,
+      "grad_norm": 0.5058545470237732,
+      "learning_rate": 8e-05,
+      "loss": 1.6761,
+      "step": 4943
+    },
+    {
+      "epoch": 0.2755852842809365,
+      "grad_norm": 0.5490286946296692,
+      "learning_rate": 8e-05,
+      "loss": 1.8641,
+      "step": 4944
+    },
+    {
+      "epoch": 0.27564102564102566,
+      "grad_norm": 0.47824475169181824,
+      "learning_rate": 8e-05,
+      "loss": 1.7207,
+      "step": 4945
+    },
+    {
+      "epoch": 0.27569676700111484,
+      "grad_norm": 0.5228742957115173,
+      "learning_rate": 8e-05,
+      "loss": 1.729,
+      "step": 4946
+    },
+    {
+      "epoch": 0.275752508361204,
+      "grad_norm": 0.5374174118041992,
+      "learning_rate": 8e-05,
+      "loss": 1.6857,
+      "step": 4947
+    },
+    {
+      "epoch": 0.2758082497212932,
+      "grad_norm": 0.5289243459701538,
+      "learning_rate": 8e-05,
+      "loss": 1.7642,
+      "step": 4948
+    },
+    {
+      "epoch": 0.2758639910813824,
+      "grad_norm": 0.4823203980922699,
+      "learning_rate": 8e-05,
+      "loss": 1.74,
+      "step": 4949
+    },
+    {
+      "epoch": 0.27591973244147155,
+      "grad_norm": 0.4666809141635895,
+      "learning_rate": 8e-05,
+      "loss": 1.5925,
+      "step": 4950
+    },
+    {
+      "epoch": 0.27597547380156073,
+      "grad_norm": 0.4994995594024658,
+      "learning_rate": 8e-05,
+      "loss": 1.6443,
+      "step": 4951
+    },
+    {
+      "epoch": 0.27603121516164997,
+      "grad_norm": 0.47002774477005005,
+      "learning_rate": 8e-05,
+      "loss": 1.4266,
+      "step": 4952
+    },
+    {
+      "epoch": 0.27608695652173915,
+      "grad_norm": 0.5007594227790833,
+      "learning_rate": 8e-05,
+      "loss": 1.7219,
+      "step": 4953
+    },
+    {
+      "epoch": 0.2761426978818283,
+      "grad_norm": 0.49902820587158203,
+      "learning_rate": 8e-05,
+      "loss": 1.7606,
+      "step": 4954
+    },
+    {
+      "epoch": 0.2761984392419175,
+      "grad_norm": 0.5196755528450012,
+      "learning_rate": 8e-05,
+      "loss": 1.7113,
+      "step": 4955
+    },
+    {
+      "epoch": 0.2762541806020067,
+      "grad_norm": 0.4996965825557709,
+      "learning_rate": 8e-05,
+      "loss": 1.7651,
+      "step": 4956
+    },
+    {
+      "epoch": 0.27630992196209586,
+      "grad_norm": 0.52850741147995,
+      "learning_rate": 8e-05,
+      "loss": 1.7883,
+      "step": 4957
+    },
+    {
+      "epoch": 0.27636566332218504,
+      "grad_norm": 0.5271864533424377,
+      "learning_rate": 8e-05,
+      "loss": 1.9402,
+      "step": 4958
+    },
+    {
+      "epoch": 0.2764214046822743,
+      "grad_norm": 0.4749406576156616,
+      "learning_rate": 8e-05,
+      "loss": 1.5498,
+      "step": 4959
+    },
+    {
+      "epoch": 0.27647714604236345,
+      "grad_norm": 0.5235247611999512,
+      "learning_rate": 8e-05,
+      "loss": 1.8308,
+      "step": 4960
+    },
+    {
+      "epoch": 0.27653288740245263,
+      "grad_norm": 0.45240989327430725,
+      "learning_rate": 8e-05,
+      "loss": 1.1512,
+      "step": 4961
+    },
+    {
+      "epoch": 0.2765886287625418,
+      "grad_norm": 0.5295997262001038,
+      "learning_rate": 8e-05,
+      "loss": 1.7245,
+      "step": 4962
+    },
+    {
+      "epoch": 0.276644370122631,
+      "grad_norm": 0.5398911237716675,
+      "learning_rate": 8e-05,
+      "loss": 1.5923,
+      "step": 4963
+    },
+    {
+      "epoch": 0.27670011148272017,
+      "grad_norm": 0.5286577343940735,
+      "learning_rate": 8e-05,
+      "loss": 1.722,
+      "step": 4964
+    },
+    {
+      "epoch": 0.27675585284280935,
+      "grad_norm": 0.5557246208190918,
+      "learning_rate": 8e-05,
+      "loss": 1.7281,
+      "step": 4965
+    },
+    {
+      "epoch": 0.2768115942028985,
+      "grad_norm": 0.5491385459899902,
+      "learning_rate": 8e-05,
+      "loss": 1.9146,
+      "step": 4966
+    },
+    {
+      "epoch": 0.27686733556298776,
+      "grad_norm": 0.5278223156929016,
+      "learning_rate": 8e-05,
+      "loss": 1.7702,
+      "step": 4967
+    },
+    {
+      "epoch": 0.27692307692307694,
+      "grad_norm": 0.5233393311500549,
+      "learning_rate": 8e-05,
+      "loss": 1.663,
+      "step": 4968
+    },
+    {
+      "epoch": 0.2769788182831661,
+      "grad_norm": 0.5258099436759949,
+      "learning_rate": 8e-05,
+      "loss": 1.6163,
+      "step": 4969
+    },
+    {
+      "epoch": 0.2770345596432553,
+      "grad_norm": 0.46531739830970764,
+      "learning_rate": 8e-05,
+      "loss": 1.5218,
+      "step": 4970
+    },
+    {
+      "epoch": 0.2770903010033445,
+      "grad_norm": 0.4865444004535675,
+      "learning_rate": 8e-05,
+      "loss": 1.7111,
+      "step": 4971
+    },
+    {
+      "epoch": 0.27714604236343365,
+      "grad_norm": 0.5154854655265808,
+      "learning_rate": 8e-05,
+      "loss": 1.6215,
+      "step": 4972
+    },
+    {
+      "epoch": 0.27720178372352283,
+      "grad_norm": 0.5724713802337646,
+      "learning_rate": 8e-05,
+      "loss": 1.7638,
+      "step": 4973
+    },
+    {
+      "epoch": 0.27725752508361207,
+      "grad_norm": 0.4830567240715027,
+      "learning_rate": 8e-05,
+      "loss": 1.6199,
+      "step": 4974
+    },
+    {
+      "epoch": 0.27731326644370125,
+      "grad_norm": 0.528709888458252,
+      "learning_rate": 8e-05,
+      "loss": 1.8588,
+      "step": 4975
+    },
+    {
+      "epoch": 0.2773690078037904,
+      "grad_norm": 0.4855372905731201,
+      "learning_rate": 8e-05,
+      "loss": 1.7285,
+      "step": 4976
+    },
+    {
+      "epoch": 0.2774247491638796,
+      "grad_norm": 0.5438281893730164,
+      "learning_rate": 8e-05,
+      "loss": 1.658,
+      "step": 4977
+    },
+    {
+      "epoch": 0.2774804905239688,
+      "grad_norm": 0.48184525966644287,
+      "learning_rate": 8e-05,
+      "loss": 1.6383,
+      "step": 4978
+    },
+    {
+      "epoch": 0.27753623188405796,
+      "grad_norm": 0.48708274960517883,
+      "learning_rate": 8e-05,
+      "loss": 1.7222,
+      "step": 4979
+    },
+    {
+      "epoch": 0.27759197324414714,
+      "grad_norm": 0.4790375828742981,
+      "learning_rate": 8e-05,
+      "loss": 1.6565,
+      "step": 4980
+    },
+    {
+      "epoch": 0.2776477146042363,
+      "grad_norm": 0.4739849269390106,
+      "learning_rate": 8e-05,
+      "loss": 1.6659,
+      "step": 4981
+    },
+    {
+      "epoch": 0.27770345596432555,
+      "grad_norm": 0.480943888425827,
+      "learning_rate": 8e-05,
+      "loss": 1.6233,
+      "step": 4982
+    },
+    {
+      "epoch": 0.27775919732441473,
+      "grad_norm": 0.47466662526130676,
+      "learning_rate": 8e-05,
+      "loss": 1.7093,
+      "step": 4983
+    },
+    {
+      "epoch": 0.2778149386845039,
+      "grad_norm": 0.49456772208213806,
+      "learning_rate": 8e-05,
+      "loss": 1.6705,
+      "step": 4984
+    },
+    {
+      "epoch": 0.2778706800445931,
+      "grad_norm": 0.5678805112838745,
+      "learning_rate": 8e-05,
+      "loss": 1.76,
+      "step": 4985
+    },
+    {
+      "epoch": 0.27792642140468227,
+      "grad_norm": 0.4886873960494995,
+      "learning_rate": 8e-05,
+      "loss": 1.6806,
+      "step": 4986
+    },
+    {
+      "epoch": 0.27798216276477145,
+      "grad_norm": 0.5274010300636292,
+      "learning_rate": 8e-05,
+      "loss": 1.9928,
+      "step": 4987
+    },
+    {
+      "epoch": 0.2780379041248606,
+      "grad_norm": 0.5529814958572388,
+      "learning_rate": 8e-05,
+      "loss": 1.7622,
+      "step": 4988
+    },
+    {
+      "epoch": 0.27809364548494986,
+      "grad_norm": 0.4791403114795685,
+      "learning_rate": 8e-05,
+      "loss": 1.6132,
+      "step": 4989
+    },
+    {
+      "epoch": 0.27814938684503904,
+      "grad_norm": 0.5282229781150818,
+      "learning_rate": 8e-05,
+      "loss": 1.7558,
+      "step": 4990
+    },
+    {
+      "epoch": 0.2782051282051282,
+      "grad_norm": 0.47779935598373413,
+      "learning_rate": 8e-05,
+      "loss": 1.6692,
+      "step": 4991
+    },
+    {
+      "epoch": 0.2782608695652174,
+      "grad_norm": 0.5005864500999451,
+      "learning_rate": 8e-05,
+      "loss": 1.8308,
+      "step": 4992
+    },
+    {
+      "epoch": 0.2783166109253066,
+      "grad_norm": 0.49798935651779175,
+      "learning_rate": 8e-05,
+      "loss": 1.5786,
+      "step": 4993
+    },
+    {
+      "epoch": 0.27837235228539575,
+      "grad_norm": 0.5125159621238708,
+      "learning_rate": 8e-05,
+      "loss": 1.5815,
+      "step": 4994
+    },
+    {
+      "epoch": 0.27842809364548493,
+      "grad_norm": 0.5537694692611694,
+      "learning_rate": 8e-05,
+      "loss": 1.8341,
+      "step": 4995
+    },
+    {
+      "epoch": 0.2784838350055741,
+      "grad_norm": 0.552141547203064,
+      "learning_rate": 8e-05,
+      "loss": 1.4153,
+      "step": 4996
+    },
+    {
+      "epoch": 0.27853957636566334,
+      "grad_norm": 0.4947458803653717,
+      "learning_rate": 8e-05,
+      "loss": 1.6776,
+      "step": 4997
+    },
+    {
+      "epoch": 0.2785953177257525,
+      "grad_norm": 0.4945355951786041,
+      "learning_rate": 8e-05,
+      "loss": 1.5679,
+      "step": 4998
+    },
+    {
+      "epoch": 0.2786510590858417,
+      "grad_norm": 0.4887459874153137,
+      "learning_rate": 8e-05,
+      "loss": 1.5423,
+      "step": 4999
+    },
+    {
+      "epoch": 0.2787068004459309,
+      "grad_norm": 0.47073450684547424,
+      "learning_rate": 8e-05,
+      "loss": 1.5159,
+      "step": 5000
+    },
+    {
+      "epoch": 0.27876254180602006,
+      "grad_norm": 0.4986558258533478,
+      "learning_rate": 8e-05,
+      "loss": 1.523,
+      "step": 5001
+    },
+    {
+      "epoch": 0.27881828316610924,
+      "grad_norm": 0.5027152895927429,
+      "learning_rate": 8e-05,
+      "loss": 1.6087,
+      "step": 5002
+    },
+    {
+      "epoch": 0.2788740245261984,
+      "grad_norm": 0.4802919328212738,
+      "learning_rate": 8e-05,
+      "loss": 1.5582,
+      "step": 5003
+    },
+    {
+      "epoch": 0.27892976588628765,
+      "grad_norm": 0.5270285606384277,
+      "learning_rate": 8e-05,
+      "loss": 1.7817,
+      "step": 5004
+    },
+    {
+      "epoch": 0.27898550724637683,
+      "grad_norm": 0.49758392572402954,
+      "learning_rate": 8e-05,
+      "loss": 1.7247,
+      "step": 5005
+    },
+    {
+      "epoch": 0.279041248606466,
+      "grad_norm": 0.49667099118232727,
+      "learning_rate": 8e-05,
+      "loss": 1.6441,
+      "step": 5006
+    },
+    {
+      "epoch": 0.2790969899665552,
+      "grad_norm": 0.5154736638069153,
+      "learning_rate": 8e-05,
+      "loss": 1.7622,
+      "step": 5007
+    },
+    {
+      "epoch": 0.27915273132664437,
+      "grad_norm": 0.53494793176651,
+      "learning_rate": 8e-05,
+      "loss": 1.4958,
+      "step": 5008
+    },
+    {
+      "epoch": 0.27920847268673354,
+      "grad_norm": 0.5931455492973328,
+      "learning_rate": 8e-05,
+      "loss": 1.5992,
+      "step": 5009
+    },
+    {
+      "epoch": 0.2792642140468227,
+      "grad_norm": 0.4625272750854492,
+      "learning_rate": 8e-05,
+      "loss": 1.3788,
+      "step": 5010
+    },
+    {
+      "epoch": 0.2793199554069119,
+      "grad_norm": 0.47783440351486206,
+      "learning_rate": 8e-05,
+      "loss": 1.4587,
+      "step": 5011
+    },
+    {
+      "epoch": 0.27937569676700114,
+      "grad_norm": 0.5056319832801819,
+      "learning_rate": 8e-05,
+      "loss": 1.7659,
+      "step": 5012
+    },
+    {
+      "epoch": 0.2794314381270903,
+      "grad_norm": 0.503319501876831,
+      "learning_rate": 8e-05,
+      "loss": 1.8524,
+      "step": 5013
+    },
+    {
+      "epoch": 0.2794871794871795,
+      "grad_norm": 0.49347031116485596,
+      "learning_rate": 8e-05,
+      "loss": 1.6201,
+      "step": 5014
+    },
+    {
+      "epoch": 0.27954292084726867,
+      "grad_norm": 0.4896809756755829,
+      "learning_rate": 8e-05,
+      "loss": 1.7937,
+      "step": 5015
+    },
+    {
+      "epoch": 0.27959866220735785,
+      "grad_norm": 0.516046404838562,
+      "learning_rate": 8e-05,
+      "loss": 1.794,
+      "step": 5016
+    },
+    {
+      "epoch": 0.27965440356744703,
+      "grad_norm": 0.46880456805229187,
+      "learning_rate": 8e-05,
+      "loss": 1.6199,
+      "step": 5017
+    },
+    {
+      "epoch": 0.2797101449275362,
+      "grad_norm": 0.49637213349342346,
+      "learning_rate": 8e-05,
+      "loss": 1.6879,
+      "step": 5018
+    },
+    {
+      "epoch": 0.27976588628762544,
+      "grad_norm": 0.5818153023719788,
+      "learning_rate": 8e-05,
+      "loss": 2.124,
+      "step": 5019
+    },
+    {
+      "epoch": 0.2798216276477146,
+      "grad_norm": 0.49280956387519836,
+      "learning_rate": 8e-05,
+      "loss": 1.7206,
+      "step": 5020
+    },
+    {
+      "epoch": 0.2798773690078038,
+      "grad_norm": 0.5034249424934387,
+      "learning_rate": 8e-05,
+      "loss": 1.7388,
+      "step": 5021
+    },
+    {
+      "epoch": 0.279933110367893,
+      "grad_norm": 0.4745542109012604,
+      "learning_rate": 8e-05,
+      "loss": 1.6114,
+      "step": 5022
+    },
+    {
+      "epoch": 0.27998885172798216,
+      "grad_norm": 0.5141556859016418,
+      "learning_rate": 8e-05,
+      "loss": 1.867,
+      "step": 5023
+    },
+    {
+      "epoch": 0.28004459308807134,
+      "grad_norm": 0.5179420113563538,
+      "learning_rate": 8e-05,
+      "loss": 1.8646,
+      "step": 5024
+    },
+    {
+      "epoch": 0.2801003344481605,
+      "grad_norm": 0.4926811158657074,
+      "learning_rate": 8e-05,
+      "loss": 1.6828,
+      "step": 5025
+    },
+    {
+      "epoch": 0.2801560758082497,
+      "grad_norm": 0.5116769671440125,
+      "learning_rate": 8e-05,
+      "loss": 1.576,
+      "step": 5026
+    },
+    {
+      "epoch": 0.28021181716833893,
+      "grad_norm": 0.42399269342422485,
+      "learning_rate": 8e-05,
+      "loss": 1.3601,
+      "step": 5027
+    },
+    {
+      "epoch": 0.2802675585284281,
+      "grad_norm": 0.5206683278083801,
+      "learning_rate": 8e-05,
+      "loss": 1.7806,
+      "step": 5028
+    },
+    {
+      "epoch": 0.2803232998885173,
+      "grad_norm": 0.4614886939525604,
+      "learning_rate": 8e-05,
+      "loss": 1.6074,
+      "step": 5029
+    },
+    {
+      "epoch": 0.28037904124860646,
+      "grad_norm": 0.5836182236671448,
+      "learning_rate": 8e-05,
+      "loss": 1.9167,
+      "step": 5030
+    },
+    {
+      "epoch": 0.28043478260869564,
+      "grad_norm": 0.5373634099960327,
+      "learning_rate": 8e-05,
+      "loss": 1.8484,
+      "step": 5031
+    },
+    {
+      "epoch": 0.2804905239687848,
+      "grad_norm": 0.47761693596839905,
+      "learning_rate": 8e-05,
+      "loss": 1.7066,
+      "step": 5032
+    },
+    {
+      "epoch": 0.280546265328874,
+      "grad_norm": 0.47047945857048035,
+      "learning_rate": 8e-05,
+      "loss": 1.5976,
+      "step": 5033
+    },
+    {
+      "epoch": 0.28060200668896323,
+      "grad_norm": 0.47744011878967285,
+      "learning_rate": 8e-05,
+      "loss": 1.6201,
+      "step": 5034
+    },
+    {
+      "epoch": 0.2806577480490524,
+      "grad_norm": 0.5402758121490479,
+      "learning_rate": 8e-05,
+      "loss": 1.836,
+      "step": 5035
+    },
+    {
+      "epoch": 0.2807134894091416,
+      "grad_norm": 0.5157755017280579,
+      "learning_rate": 8e-05,
+      "loss": 1.7456,
+      "step": 5036
+    },
+    {
+      "epoch": 0.28076923076923077,
+      "grad_norm": 0.5129708051681519,
+      "learning_rate": 8e-05,
+      "loss": 1.5757,
+      "step": 5037
+    },
+    {
+      "epoch": 0.28082497212931995,
+      "grad_norm": 0.5077796578407288,
+      "learning_rate": 8e-05,
+      "loss": 1.7965,
+      "step": 5038
+    },
+    {
+      "epoch": 0.28088071348940913,
+      "grad_norm": 0.4889647364616394,
+      "learning_rate": 8e-05,
+      "loss": 1.6926,
+      "step": 5039
+    },
+    {
+      "epoch": 0.2809364548494983,
+      "grad_norm": 0.48703962564468384,
+      "learning_rate": 8e-05,
+      "loss": 1.6108,
+      "step": 5040
+    },
+    {
+      "epoch": 0.28099219620958754,
+      "grad_norm": 0.5404532551765442,
+      "learning_rate": 8e-05,
+      "loss": 1.8826,
+      "step": 5041
+    },
+    {
+      "epoch": 0.2810479375696767,
+      "grad_norm": 0.5641488432884216,
+      "learning_rate": 8e-05,
+      "loss": 1.7598,
+      "step": 5042
+    },
+    {
+      "epoch": 0.2811036789297659,
+      "grad_norm": 0.4878641366958618,
+      "learning_rate": 8e-05,
+      "loss": 1.724,
+      "step": 5043
+    },
+    {
+      "epoch": 0.2811594202898551,
+      "grad_norm": 0.5139477849006653,
+      "learning_rate": 8e-05,
+      "loss": 1.6586,
+      "step": 5044
+    },
+    {
+      "epoch": 0.28121516164994426,
+      "grad_norm": 0.5409367084503174,
+      "learning_rate": 8e-05,
+      "loss": 1.7619,
+      "step": 5045
+    },
+    {
+      "epoch": 0.28127090301003344,
+      "grad_norm": 0.5175203084945679,
+      "learning_rate": 8e-05,
+      "loss": 1.7523,
+      "step": 5046
+    },
+    {
+      "epoch": 0.2813266443701226,
+      "grad_norm": 0.5151575207710266,
+      "learning_rate": 8e-05,
+      "loss": 1.6643,
+      "step": 5047
+    },
+    {
+      "epoch": 0.2813823857302118,
+      "grad_norm": 0.5394100546836853,
+      "learning_rate": 8e-05,
+      "loss": 1.5896,
+      "step": 5048
+    },
+    {
+      "epoch": 0.281438127090301,
+      "grad_norm": 0.5807749032974243,
+      "learning_rate": 8e-05,
+      "loss": 2.002,
+      "step": 5049
+    },
+    {
+      "epoch": 0.2814938684503902,
+      "grad_norm": 0.4935115575790405,
+      "learning_rate": 8e-05,
+      "loss": 1.5606,
+      "step": 5050
+    },
+    {
+      "epoch": 0.2815496098104794,
+      "grad_norm": 0.4910953640937805,
+      "learning_rate": 8e-05,
+      "loss": 1.6825,
+      "step": 5051
+    },
+    {
+      "epoch": 0.28160535117056856,
+      "grad_norm": 0.5092360377311707,
+      "learning_rate": 8e-05,
+      "loss": 1.7917,
+      "step": 5052
+    },
+    {
+      "epoch": 0.28166109253065774,
+      "grad_norm": 0.5052596926689148,
+      "learning_rate": 8e-05,
+      "loss": 1.6808,
+      "step": 5053
+    },
+    {
+      "epoch": 0.2817168338907469,
+      "grad_norm": 0.5162422060966492,
+      "learning_rate": 8e-05,
+      "loss": 1.8994,
+      "step": 5054
+    },
+    {
+      "epoch": 0.2817725752508361,
+      "grad_norm": 0.48943495750427246,
+      "learning_rate": 8e-05,
+      "loss": 1.6462,
+      "step": 5055
+    },
+    {
+      "epoch": 0.28182831661092533,
+      "grad_norm": 0.5406565070152283,
+      "learning_rate": 8e-05,
+      "loss": 1.8143,
+      "step": 5056
+    },
+    {
+      "epoch": 0.2818840579710145,
+      "grad_norm": 0.502768874168396,
+      "learning_rate": 8e-05,
+      "loss": 1.7451,
+      "step": 5057
+    },
+    {
+      "epoch": 0.2819397993311037,
+      "grad_norm": 0.5879219770431519,
+      "learning_rate": 8e-05,
+      "loss": 2.0077,
+      "step": 5058
+    },
+    {
+      "epoch": 0.28199554069119287,
+      "grad_norm": 0.5409952998161316,
+      "learning_rate": 8e-05,
+      "loss": 1.7412,
+      "step": 5059
+    },
+    {
+      "epoch": 0.28205128205128205,
+      "grad_norm": 0.566804051399231,
+      "learning_rate": 8e-05,
+      "loss": 1.8172,
+      "step": 5060
+    },
+    {
+      "epoch": 0.2821070234113712,
+      "grad_norm": 0.5219933390617371,
+      "learning_rate": 8e-05,
+      "loss": 1.7815,
+      "step": 5061
+    },
+    {
+      "epoch": 0.2821627647714604,
+      "grad_norm": 0.5377077460289001,
+      "learning_rate": 8e-05,
+      "loss": 1.5289,
+      "step": 5062
+    },
+    {
+      "epoch": 0.2822185061315496,
+      "grad_norm": 0.49284616112709045,
+      "learning_rate": 8e-05,
+      "loss": 1.8427,
+      "step": 5063
+    },
+    {
+      "epoch": 0.2822742474916388,
+      "grad_norm": 0.4860459864139557,
+      "learning_rate": 8e-05,
+      "loss": 1.6642,
+      "step": 5064
+    },
+    {
+      "epoch": 0.282329988851728,
+      "grad_norm": 0.49022603034973145,
+      "learning_rate": 8e-05,
+      "loss": 1.5648,
+      "step": 5065
+    },
+    {
+      "epoch": 0.2823857302118172,
+      "grad_norm": 0.5152029395103455,
+      "learning_rate": 8e-05,
+      "loss": 1.778,
+      "step": 5066
+    },
+    {
+      "epoch": 0.28244147157190636,
+      "grad_norm": 0.47602224349975586,
+      "learning_rate": 8e-05,
+      "loss": 1.7115,
+      "step": 5067
+    },
+    {
+      "epoch": 0.28249721293199553,
+      "grad_norm": 0.4774375259876251,
+      "learning_rate": 8e-05,
+      "loss": 1.6216,
+      "step": 5068
+    },
+    {
+      "epoch": 0.2825529542920847,
+      "grad_norm": 0.585257887840271,
+      "learning_rate": 8e-05,
+      "loss": 1.8704,
+      "step": 5069
+    },
+    {
+      "epoch": 0.2826086956521739,
+      "grad_norm": 0.5467697978019714,
+      "learning_rate": 8e-05,
+      "loss": 1.69,
+      "step": 5070
+    },
+    {
+      "epoch": 0.2826644370122631,
+      "grad_norm": 0.5306287407875061,
+      "learning_rate": 8e-05,
+      "loss": 1.7611,
+      "step": 5071
+    },
+    {
+      "epoch": 0.2827201783723523,
+      "grad_norm": 0.54197758436203,
+      "learning_rate": 8e-05,
+      "loss": 1.8773,
+      "step": 5072
+    },
+    {
+      "epoch": 0.2827759197324415,
+      "grad_norm": 0.5408424139022827,
+      "learning_rate": 8e-05,
+      "loss": 1.86,
+      "step": 5073
+    },
+    {
+      "epoch": 0.28283166109253066,
+      "grad_norm": 0.5417904257774353,
+      "learning_rate": 8e-05,
+      "loss": 1.8728,
+      "step": 5074
+    },
+    {
+      "epoch": 0.28288740245261984,
+      "grad_norm": 0.47730568051338196,
+      "learning_rate": 8e-05,
+      "loss": 1.4756,
+      "step": 5075
+    },
+    {
+      "epoch": 0.282943143812709,
+      "grad_norm": 0.5218319892883301,
+      "learning_rate": 8e-05,
+      "loss": 1.7237,
+      "step": 5076
+    },
+    {
+      "epoch": 0.2829988851727982,
+      "grad_norm": 0.5022181272506714,
+      "learning_rate": 8e-05,
+      "loss": 1.7086,
+      "step": 5077
+    },
+    {
+      "epoch": 0.2830546265328874,
+      "grad_norm": 0.49561530351638794,
+      "learning_rate": 8e-05,
+      "loss": 1.7091,
+      "step": 5078
+    },
+    {
+      "epoch": 0.2831103678929766,
+      "grad_norm": 0.494386225938797,
+      "learning_rate": 8e-05,
+      "loss": 1.3795,
+      "step": 5079
+    },
+    {
+      "epoch": 0.2831661092530658,
+      "grad_norm": 0.49516165256500244,
+      "learning_rate": 8e-05,
+      "loss": 1.5951,
+      "step": 5080
+    },
+    {
+      "epoch": 0.28322185061315497,
+      "grad_norm": 0.501606822013855,
+      "learning_rate": 8e-05,
+      "loss": 1.8065,
+      "step": 5081
+    },
+    {
+      "epoch": 0.28327759197324415,
+      "grad_norm": 0.527359127998352,
+      "learning_rate": 8e-05,
+      "loss": 1.8715,
+      "step": 5082
+    },
+    {
+      "epoch": 0.2833333333333333,
+      "grad_norm": 0.5170018076896667,
+      "learning_rate": 8e-05,
+      "loss": 1.7879,
+      "step": 5083
+    },
+    {
+      "epoch": 0.2833890746934225,
+      "grad_norm": 0.49888837337493896,
+      "learning_rate": 8e-05,
+      "loss": 1.4441,
+      "step": 5084
+    },
+    {
+      "epoch": 0.2834448160535117,
+      "grad_norm": 0.521267294883728,
+      "learning_rate": 8e-05,
+      "loss": 1.7673,
+      "step": 5085
+    },
+    {
+      "epoch": 0.2835005574136009,
+      "grad_norm": 0.49815288186073303,
+      "learning_rate": 8e-05,
+      "loss": 1.6354,
+      "step": 5086
+    },
+    {
+      "epoch": 0.2835562987736901,
+      "grad_norm": 0.5356400609016418,
+      "learning_rate": 8e-05,
+      "loss": 1.7102,
+      "step": 5087
+    },
+    {
+      "epoch": 0.2836120401337793,
+      "grad_norm": 0.5145084857940674,
+      "learning_rate": 8e-05,
+      "loss": 1.7456,
+      "step": 5088
+    },
+    {
+      "epoch": 0.28366778149386845,
+      "grad_norm": 0.4927878677845001,
+      "learning_rate": 8e-05,
+      "loss": 1.7089,
+      "step": 5089
+    },
+    {
+      "epoch": 0.28372352285395763,
+      "grad_norm": 0.503649115562439,
+      "learning_rate": 8e-05,
+      "loss": 1.73,
+      "step": 5090
+    },
+    {
+      "epoch": 0.2837792642140468,
+      "grad_norm": 0.5390079021453857,
+      "learning_rate": 8e-05,
+      "loss": 1.8051,
+      "step": 5091
+    },
+    {
+      "epoch": 0.283835005574136,
+      "grad_norm": 0.48613032698631287,
+      "learning_rate": 8e-05,
+      "loss": 1.6163,
+      "step": 5092
+    },
+    {
+      "epoch": 0.28389074693422517,
+      "grad_norm": 0.4882640838623047,
+      "learning_rate": 8e-05,
+      "loss": 1.5683,
+      "step": 5093
+    },
+    {
+      "epoch": 0.2839464882943144,
+      "grad_norm": 0.5540668368339539,
+      "learning_rate": 8e-05,
+      "loss": 1.7487,
+      "step": 5094
+    },
+    {
+      "epoch": 0.2840022296544036,
+      "grad_norm": 0.5007288455963135,
+      "learning_rate": 8e-05,
+      "loss": 1.5905,
+      "step": 5095
+    },
+    {
+      "epoch": 0.28405797101449276,
+      "grad_norm": 0.5038129687309265,
+      "learning_rate": 8e-05,
+      "loss": 1.716,
+      "step": 5096
+    },
+    {
+      "epoch": 0.28411371237458194,
+      "grad_norm": 0.6086083054542542,
+      "learning_rate": 8e-05,
+      "loss": 1.9201,
+      "step": 5097
+    },
+    {
+      "epoch": 0.2841694537346711,
+      "grad_norm": 0.5607197880744934,
+      "learning_rate": 8e-05,
+      "loss": 1.7734,
+      "step": 5098
+    },
+    {
+      "epoch": 0.2842251950947603,
+      "grad_norm": 0.5821718573570251,
+      "learning_rate": 8e-05,
+      "loss": 1.7251,
+      "step": 5099
+    },
+    {
+      "epoch": 0.2842809364548495,
+      "grad_norm": 0.4806770384311676,
+      "learning_rate": 8e-05,
+      "loss": 1.5513,
+      "step": 5100
+    },
+    {
+      "epoch": 0.2843366778149387,
+      "grad_norm": 0.4750522971153259,
+      "learning_rate": 8e-05,
+      "loss": 1.5404,
+      "step": 5101
+    },
+    {
+      "epoch": 0.2843924191750279,
+      "grad_norm": 0.5024062991142273,
+      "learning_rate": 8e-05,
+      "loss": 1.691,
+      "step": 5102
+    },
+    {
+      "epoch": 0.28444816053511707,
+      "grad_norm": 0.498268187046051,
+      "learning_rate": 8e-05,
+      "loss": 1.835,
+      "step": 5103
+    },
+    {
+      "epoch": 0.28450390189520625,
+      "grad_norm": 0.5400916337966919,
+      "learning_rate": 8e-05,
+      "loss": 1.9,
+      "step": 5104
+    },
+    {
+      "epoch": 0.2845596432552954,
+      "grad_norm": 0.4609983563423157,
+      "learning_rate": 8e-05,
+      "loss": 1.6109,
+      "step": 5105
+    },
+    {
+      "epoch": 0.2846153846153846,
+      "grad_norm": 0.532023549079895,
+      "learning_rate": 8e-05,
+      "loss": 1.7207,
+      "step": 5106
+    },
+    {
+      "epoch": 0.2846711259754738,
+      "grad_norm": 0.49450594186782837,
+      "learning_rate": 8e-05,
+      "loss": 1.8646,
+      "step": 5107
+    },
+    {
+      "epoch": 0.28472686733556296,
+      "grad_norm": 0.4583391845226288,
+      "learning_rate": 8e-05,
+      "loss": 1.4679,
+      "step": 5108
+    },
+    {
+      "epoch": 0.2847826086956522,
+      "grad_norm": 0.5723952651023865,
+      "learning_rate": 8e-05,
+      "loss": 1.7621,
+      "step": 5109
+    },
+    {
+      "epoch": 0.2848383500557414,
+      "grad_norm": 0.48113951086997986,
+      "learning_rate": 8e-05,
+      "loss": 1.6413,
+      "step": 5110
+    },
+    {
+      "epoch": 0.28489409141583055,
+      "grad_norm": 0.509999692440033,
+      "learning_rate": 8e-05,
+      "loss": 1.7916,
+      "step": 5111
+    },
+    {
+      "epoch": 0.28494983277591973,
+      "grad_norm": 0.6294270753860474,
+      "learning_rate": 8e-05,
+      "loss": 2.0127,
+      "step": 5112
+    },
+    {
+      "epoch": 0.2850055741360089,
+      "grad_norm": 0.5116695165634155,
+      "learning_rate": 8e-05,
+      "loss": 1.6236,
+      "step": 5113
+    },
+    {
+      "epoch": 0.2850613154960981,
+      "grad_norm": 0.49915406107902527,
+      "learning_rate": 8e-05,
+      "loss": 1.7948,
+      "step": 5114
+    },
+    {
+      "epoch": 0.28511705685618727,
+      "grad_norm": 0.5811174511909485,
+      "learning_rate": 8e-05,
+      "loss": 1.6464,
+      "step": 5115
+    },
+    {
+      "epoch": 0.2851727982162765,
+      "grad_norm": 0.5079129934310913,
+      "learning_rate": 8e-05,
+      "loss": 1.784,
+      "step": 5116
+    },
+    {
+      "epoch": 0.2852285395763657,
+      "grad_norm": 0.5594106912612915,
+      "learning_rate": 8e-05,
+      "loss": 2.0082,
+      "step": 5117
+    },
+    {
+      "epoch": 0.28528428093645486,
+      "grad_norm": 0.5111921429634094,
+      "learning_rate": 8e-05,
+      "loss": 1.7059,
+      "step": 5118
+    },
+    {
+      "epoch": 0.28534002229654404,
+      "grad_norm": 0.5258628129959106,
+      "learning_rate": 8e-05,
+      "loss": 1.8113,
+      "step": 5119
+    },
+    {
+      "epoch": 0.2853957636566332,
+      "grad_norm": 0.49711883068084717,
+      "learning_rate": 8e-05,
+      "loss": 1.525,
+      "step": 5120
+    },
+    {
+      "epoch": 0.2854515050167224,
+      "grad_norm": 0.46852102875709534,
+      "learning_rate": 8e-05,
+      "loss": 1.4837,
+      "step": 5121
+    },
+    {
+      "epoch": 0.2855072463768116,
+      "grad_norm": 0.5145292282104492,
+      "learning_rate": 8e-05,
+      "loss": 1.7547,
+      "step": 5122
+    },
+    {
+      "epoch": 0.28556298773690075,
+      "grad_norm": 0.5454033613204956,
+      "learning_rate": 8e-05,
+      "loss": 1.699,
+      "step": 5123
+    },
+    {
+      "epoch": 0.28561872909699,
+      "grad_norm": 0.5002652406692505,
+      "learning_rate": 8e-05,
+      "loss": 1.7178,
+      "step": 5124
+    },
+    {
+      "epoch": 0.28567447045707917,
+      "grad_norm": 0.4545373320579529,
+      "learning_rate": 8e-05,
+      "loss": 1.4958,
+      "step": 5125
+    },
+    {
+      "epoch": 0.28573021181716834,
+      "grad_norm": 0.5078669786453247,
+      "learning_rate": 8e-05,
+      "loss": 1.7654,
+      "step": 5126
+    },
+    {
+      "epoch": 0.2857859531772575,
+      "grad_norm": 0.49822527170181274,
+      "learning_rate": 8e-05,
+      "loss": 1.7321,
+      "step": 5127
+    },
+    {
+      "epoch": 0.2858416945373467,
+      "grad_norm": 0.4882858097553253,
+      "learning_rate": 8e-05,
+      "loss": 1.6351,
+      "step": 5128
+    },
+    {
+      "epoch": 0.2858974358974359,
+      "grad_norm": 0.5043916702270508,
+      "learning_rate": 8e-05,
+      "loss": 1.7462,
+      "step": 5129
+    },
+    {
+      "epoch": 0.28595317725752506,
+      "grad_norm": 0.4957789480686188,
+      "learning_rate": 8e-05,
+      "loss": 1.7014,
+      "step": 5130
+    },
+    {
+      "epoch": 0.2860089186176143,
+      "grad_norm": 0.47981375455856323,
+      "learning_rate": 8e-05,
+      "loss": 1.5465,
+      "step": 5131
+    },
+    {
+      "epoch": 0.2860646599777035,
+      "grad_norm": 0.46363550424575806,
+      "learning_rate": 8e-05,
+      "loss": 1.6081,
+      "step": 5132
+    },
+    {
+      "epoch": 0.28612040133779265,
+      "grad_norm": 0.5549232959747314,
+      "learning_rate": 8e-05,
+      "loss": 1.7287,
+      "step": 5133
+    },
+    {
+      "epoch": 0.28617614269788183,
+      "grad_norm": 0.4919549822807312,
+      "learning_rate": 8e-05,
+      "loss": 1.7172,
+      "step": 5134
+    },
+    {
+      "epoch": 0.286231884057971,
+      "grad_norm": 0.5521557927131653,
+      "learning_rate": 8e-05,
+      "loss": 2.0098,
+      "step": 5135
+    },
+    {
+      "epoch": 0.2862876254180602,
+      "grad_norm": 0.49092942476272583,
+      "learning_rate": 8e-05,
+      "loss": 1.8048,
+      "step": 5136
+    },
+    {
+      "epoch": 0.28634336677814937,
+      "grad_norm": 0.4958489239215851,
+      "learning_rate": 8e-05,
+      "loss": 1.8119,
+      "step": 5137
+    },
+    {
+      "epoch": 0.2863991081382386,
+      "grad_norm": 0.4862135052680969,
+      "learning_rate": 8e-05,
+      "loss": 1.7403,
+      "step": 5138
+    },
+    {
+      "epoch": 0.2864548494983278,
+      "grad_norm": 0.50542151927948,
+      "learning_rate": 8e-05,
+      "loss": 1.8351,
+      "step": 5139
+    },
+    {
+      "epoch": 0.28651059085841696,
+      "grad_norm": 0.48798254132270813,
+      "learning_rate": 8e-05,
+      "loss": 1.5942,
+      "step": 5140
+    },
+    {
+      "epoch": 0.28656633221850614,
+      "grad_norm": 0.5196491479873657,
+      "learning_rate": 8e-05,
+      "loss": 1.558,
+      "step": 5141
+    },
+    {
+      "epoch": 0.2866220735785953,
+      "grad_norm": 0.4845980703830719,
+      "learning_rate": 8e-05,
+      "loss": 1.4992,
+      "step": 5142
+    },
+    {
+      "epoch": 0.2866778149386845,
+      "grad_norm": 0.5071967244148254,
+      "learning_rate": 8e-05,
+      "loss": 1.69,
+      "step": 5143
+    },
+    {
+      "epoch": 0.2867335562987737,
+      "grad_norm": 0.4898008704185486,
+      "learning_rate": 8e-05,
+      "loss": 1.6055,
+      "step": 5144
+    },
+    {
+      "epoch": 0.28678929765886285,
+      "grad_norm": 0.5290776491165161,
+      "learning_rate": 8e-05,
+      "loss": 1.9193,
+      "step": 5145
+    },
+    {
+      "epoch": 0.2868450390189521,
+      "grad_norm": 0.5390506982803345,
+      "learning_rate": 8e-05,
+      "loss": 1.654,
+      "step": 5146
+    },
+    {
+      "epoch": 0.28690078037904126,
+      "grad_norm": 0.5038630366325378,
+      "learning_rate": 8e-05,
+      "loss": 1.6403,
+      "step": 5147
+    },
+    {
+      "epoch": 0.28695652173913044,
+      "grad_norm": 0.5432470440864563,
+      "learning_rate": 8e-05,
+      "loss": 1.7249,
+      "step": 5148
+    },
+    {
+      "epoch": 0.2870122630992196,
+      "grad_norm": 0.48758211731910706,
+      "learning_rate": 8e-05,
+      "loss": 1.5862,
+      "step": 5149
+    },
+    {
+      "epoch": 0.2870680044593088,
+      "grad_norm": 0.541468620300293,
+      "learning_rate": 8e-05,
+      "loss": 1.971,
+      "step": 5150
+    },
+    {
+      "epoch": 0.287123745819398,
+      "grad_norm": 0.5260459780693054,
+      "learning_rate": 8e-05,
+      "loss": 1.8077,
+      "step": 5151
+    },
+    {
+      "epoch": 0.28717948717948716,
+      "grad_norm": 0.5259945392608643,
+      "learning_rate": 8e-05,
+      "loss": 1.6812,
+      "step": 5152
+    },
+    {
+      "epoch": 0.2872352285395764,
+      "grad_norm": 0.5329565405845642,
+      "learning_rate": 8e-05,
+      "loss": 1.8033,
+      "step": 5153
+    },
+    {
+      "epoch": 0.28729096989966557,
+      "grad_norm": 0.4936564266681671,
+      "learning_rate": 8e-05,
+      "loss": 1.6521,
+      "step": 5154
+    },
+    {
+      "epoch": 0.28734671125975475,
+      "grad_norm": 0.4688653349876404,
+      "learning_rate": 8e-05,
+      "loss": 1.5242,
+      "step": 5155
+    },
+    {
+      "epoch": 0.28740245261984393,
+      "grad_norm": 0.46789103746414185,
+      "learning_rate": 8e-05,
+      "loss": 1.5959,
+      "step": 5156
+    },
+    {
+      "epoch": 0.2874581939799331,
+      "grad_norm": 0.4493112862110138,
+      "learning_rate": 8e-05,
+      "loss": 1.3907,
+      "step": 5157
+    },
+    {
+      "epoch": 0.2875139353400223,
+      "grad_norm": 0.5266214609146118,
+      "learning_rate": 8e-05,
+      "loss": 1.7261,
+      "step": 5158
+    },
+    {
+      "epoch": 0.28756967670011147,
+      "grad_norm": 0.5526112914085388,
+      "learning_rate": 8e-05,
+      "loss": 1.9001,
+      "step": 5159
+    },
+    {
+      "epoch": 0.28762541806020064,
+      "grad_norm": 0.4800586402416229,
+      "learning_rate": 8e-05,
+      "loss": 1.5474,
+      "step": 5160
+    },
+    {
+      "epoch": 0.2876811594202899,
+      "grad_norm": 0.5277012586593628,
+      "learning_rate": 8e-05,
+      "loss": 1.8067,
+      "step": 5161
+    },
+    {
+      "epoch": 0.28773690078037906,
+      "grad_norm": 0.5427259206771851,
+      "learning_rate": 8e-05,
+      "loss": 1.9604,
+      "step": 5162
+    },
+    {
+      "epoch": 0.28779264214046824,
+      "grad_norm": 0.5284680128097534,
+      "learning_rate": 8e-05,
+      "loss": 1.7382,
+      "step": 5163
+    },
+    {
+      "epoch": 0.2878483835005574,
+      "grad_norm": 0.5306553244590759,
+      "learning_rate": 8e-05,
+      "loss": 1.6126,
+      "step": 5164
+    },
+    {
+      "epoch": 0.2879041248606466,
+      "grad_norm": 0.5320262312889099,
+      "learning_rate": 8e-05,
+      "loss": 1.805,
+      "step": 5165
+    },
+    {
+      "epoch": 0.28795986622073577,
+      "grad_norm": 0.4699755609035492,
+      "learning_rate": 8e-05,
+      "loss": 1.4751,
+      "step": 5166
+    },
+    {
+      "epoch": 0.28801560758082495,
+      "grad_norm": 0.5151172876358032,
+      "learning_rate": 8e-05,
+      "loss": 1.7173,
+      "step": 5167
+    },
+    {
+      "epoch": 0.2880713489409142,
+      "grad_norm": 0.480150431394577,
+      "learning_rate": 8e-05,
+      "loss": 1.6986,
+      "step": 5168
+    },
+    {
+      "epoch": 0.28812709030100336,
+      "grad_norm": 0.5028404593467712,
+      "learning_rate": 8e-05,
+      "loss": 1.6821,
+      "step": 5169
+    },
+    {
+      "epoch": 0.28818283166109254,
+      "grad_norm": 0.49127790331840515,
+      "learning_rate": 8e-05,
+      "loss": 1.7527,
+      "step": 5170
+    },
+    {
+      "epoch": 0.2882385730211817,
+      "grad_norm": 0.5041865110397339,
+      "learning_rate": 8e-05,
+      "loss": 1.6777,
+      "step": 5171
+    },
+    {
+      "epoch": 0.2882943143812709,
+      "grad_norm": 0.47198405861854553,
+      "learning_rate": 8e-05,
+      "loss": 1.3965,
+      "step": 5172
+    },
+    {
+      "epoch": 0.2883500557413601,
+      "grad_norm": 0.48658695816993713,
+      "learning_rate": 8e-05,
+      "loss": 1.587,
+      "step": 5173
+    },
+    {
+      "epoch": 0.28840579710144926,
+      "grad_norm": 0.5005254149436951,
+      "learning_rate": 8e-05,
+      "loss": 1.5085,
+      "step": 5174
+    },
+    {
+      "epoch": 0.28846153846153844,
+      "grad_norm": 0.5224685072898865,
+      "learning_rate": 8e-05,
+      "loss": 1.6863,
+      "step": 5175
+    },
+    {
+      "epoch": 0.28851727982162767,
+      "grad_norm": 0.526444673538208,
+      "learning_rate": 8e-05,
+      "loss": 1.715,
+      "step": 5176
+    },
+    {
+      "epoch": 0.28857302118171685,
+      "grad_norm": 0.5164334177970886,
+      "learning_rate": 8e-05,
+      "loss": 1.6896,
+      "step": 5177
+    },
+    {
+      "epoch": 0.288628762541806,
+      "grad_norm": 0.5021805763244629,
+      "learning_rate": 8e-05,
+      "loss": 1.744,
+      "step": 5178
+    },
+    {
+      "epoch": 0.2886845039018952,
+      "grad_norm": 0.5422660708427429,
+      "learning_rate": 8e-05,
+      "loss": 1.7615,
+      "step": 5179
+    },
+    {
+      "epoch": 0.2887402452619844,
+      "grad_norm": 0.5039775967597961,
+      "learning_rate": 8e-05,
+      "loss": 1.8386,
+      "step": 5180
+    },
+    {
+      "epoch": 0.28879598662207356,
+      "grad_norm": 0.482729971408844,
+      "learning_rate": 8e-05,
+      "loss": 1.5795,
+      "step": 5181
+    },
+    {
+      "epoch": 0.28885172798216274,
+      "grad_norm": 0.5006514191627502,
+      "learning_rate": 8e-05,
+      "loss": 1.6781,
+      "step": 5182
+    },
+    {
+      "epoch": 0.288907469342252,
+      "grad_norm": 0.49958372116088867,
+      "learning_rate": 8e-05,
+      "loss": 1.714,
+      "step": 5183
+    },
+    {
+      "epoch": 0.28896321070234116,
+      "grad_norm": 0.4862821400165558,
+      "learning_rate": 8e-05,
+      "loss": 1.7902,
+      "step": 5184
+    },
+    {
+      "epoch": 0.28901895206243033,
+      "grad_norm": 0.5306202173233032,
+      "learning_rate": 8e-05,
+      "loss": 1.6501,
+      "step": 5185
+    },
+    {
+      "epoch": 0.2890746934225195,
+      "grad_norm": 0.49687203764915466,
+      "learning_rate": 8e-05,
+      "loss": 1.7856,
+      "step": 5186
+    },
+    {
+      "epoch": 0.2891304347826087,
+      "grad_norm": 0.4888293147087097,
+      "learning_rate": 8e-05,
+      "loss": 1.6805,
+      "step": 5187
+    },
+    {
+      "epoch": 0.28918617614269787,
+      "grad_norm": 0.5002862215042114,
+      "learning_rate": 8e-05,
+      "loss": 1.6364,
+      "step": 5188
+    },
+    {
+      "epoch": 0.28924191750278705,
+      "grad_norm": 0.5225688815116882,
+      "learning_rate": 8e-05,
+      "loss": 1.6908,
+      "step": 5189
+    },
+    {
+      "epoch": 0.28929765886287623,
+      "grad_norm": 0.5296038389205933,
+      "learning_rate": 8e-05,
+      "loss": 1.7575,
+      "step": 5190
+    },
+    {
+      "epoch": 0.28935340022296546,
+      "grad_norm": 0.5127953290939331,
+      "learning_rate": 8e-05,
+      "loss": 1.6997,
+      "step": 5191
+    },
+    {
+      "epoch": 0.28940914158305464,
+      "grad_norm": 0.525801420211792,
+      "learning_rate": 8e-05,
+      "loss": 1.8364,
+      "step": 5192
+    },
+    {
+      "epoch": 0.2894648829431438,
+      "grad_norm": 0.5201865434646606,
+      "learning_rate": 8e-05,
+      "loss": 1.6582,
+      "step": 5193
+    },
+    {
+      "epoch": 0.289520624303233,
+      "grad_norm": 0.5215128064155579,
+      "learning_rate": 8e-05,
+      "loss": 1.802,
+      "step": 5194
+    },
+    {
+      "epoch": 0.2895763656633222,
+      "grad_norm": 0.5403210520744324,
+      "learning_rate": 8e-05,
+      "loss": 1.7328,
+      "step": 5195
+    },
+    {
+      "epoch": 0.28963210702341136,
+      "grad_norm": 0.5021224021911621,
+      "learning_rate": 8e-05,
+      "loss": 1.6957,
+      "step": 5196
+    },
+    {
+      "epoch": 0.28968784838350053,
+      "grad_norm": 0.4970734715461731,
+      "learning_rate": 8e-05,
+      "loss": 1.8673,
+      "step": 5197
+    },
+    {
+      "epoch": 0.28974358974358977,
+      "grad_norm": 0.498276025056839,
+      "learning_rate": 8e-05,
+      "loss": 1.7437,
+      "step": 5198
+    },
+    {
+      "epoch": 0.28979933110367895,
+      "grad_norm": 0.5261366367340088,
+      "learning_rate": 8e-05,
+      "loss": 1.6561,
+      "step": 5199
+    },
+    {
+      "epoch": 0.2898550724637681,
+      "grad_norm": 0.5094021558761597,
+      "learning_rate": 8e-05,
+      "loss": 1.7651,
+      "step": 5200
+    },
+    {
+      "epoch": 0.2899108138238573,
+      "grad_norm": 0.5043343305587769,
+      "learning_rate": 8e-05,
+      "loss": 1.6313,
+      "step": 5201
+    },
+    {
+      "epoch": 0.2899665551839465,
+      "grad_norm": 0.5139153003692627,
+      "learning_rate": 8e-05,
+      "loss": 1.74,
+      "step": 5202
+    },
+    {
+      "epoch": 0.29002229654403566,
+      "grad_norm": 0.5476909279823303,
+      "learning_rate": 8e-05,
+      "loss": 2.0281,
+      "step": 5203
+    },
+    {
+      "epoch": 0.29007803790412484,
+      "grad_norm": 0.5021553635597229,
+      "learning_rate": 8e-05,
+      "loss": 1.6699,
+      "step": 5204
+    },
+    {
+      "epoch": 0.290133779264214,
+      "grad_norm": 0.5864331722259521,
+      "learning_rate": 8e-05,
+      "loss": 1.8308,
+      "step": 5205
+    },
+    {
+      "epoch": 0.29018952062430325,
+      "grad_norm": 0.5637518167495728,
+      "learning_rate": 8e-05,
+      "loss": 1.9774,
+      "step": 5206
+    },
+    {
+      "epoch": 0.29024526198439243,
+      "grad_norm": 0.4940563142299652,
+      "learning_rate": 8e-05,
+      "loss": 1.7939,
+      "step": 5207
+    },
+    {
+      "epoch": 0.2903010033444816,
+      "grad_norm": 0.4949653744697571,
+      "learning_rate": 8e-05,
+      "loss": 1.8159,
+      "step": 5208
+    },
+    {
+      "epoch": 0.2903567447045708,
+      "grad_norm": 0.5256124138832092,
+      "learning_rate": 8e-05,
+      "loss": 1.4621,
+      "step": 5209
+    },
+    {
+      "epoch": 0.29041248606465997,
+      "grad_norm": 0.4592047333717346,
+      "learning_rate": 8e-05,
+      "loss": 1.5091,
+      "step": 5210
+    },
+    {
+      "epoch": 0.29046822742474915,
+      "grad_norm": 0.5461311936378479,
+      "learning_rate": 8e-05,
+      "loss": 1.8522,
+      "step": 5211
+    },
+    {
+      "epoch": 0.2905239687848383,
+      "grad_norm": 0.48445531725883484,
+      "learning_rate": 8e-05,
+      "loss": 1.6062,
+      "step": 5212
+    },
+    {
+      "epoch": 0.29057971014492756,
+      "grad_norm": 0.5428815484046936,
+      "learning_rate": 8e-05,
+      "loss": 1.991,
+      "step": 5213
+    },
+    {
+      "epoch": 0.29063545150501674,
+      "grad_norm": 0.52653568983078,
+      "learning_rate": 8e-05,
+      "loss": 1.7512,
+      "step": 5214
+    },
+    {
+      "epoch": 0.2906911928651059,
+      "grad_norm": 0.5246649384498596,
+      "learning_rate": 8e-05,
+      "loss": 1.5794,
+      "step": 5215
+    },
+    {
+      "epoch": 0.2907469342251951,
+      "grad_norm": 0.5046067833900452,
+      "learning_rate": 8e-05,
+      "loss": 1.5843,
+      "step": 5216
+    },
+    {
+      "epoch": 0.2908026755852843,
+      "grad_norm": 0.4852803647518158,
+      "learning_rate": 8e-05,
+      "loss": 1.4408,
+      "step": 5217
+    },
+    {
+      "epoch": 0.29085841694537345,
+      "grad_norm": 0.506061851978302,
+      "learning_rate": 8e-05,
+      "loss": 1.6776,
+      "step": 5218
+    },
+    {
+      "epoch": 0.29091415830546263,
+      "grad_norm": 0.5446941256523132,
+      "learning_rate": 8e-05,
+      "loss": 1.7467,
+      "step": 5219
+    },
+    {
+      "epoch": 0.2909698996655518,
+      "grad_norm": 0.5437446236610413,
+      "learning_rate": 8e-05,
+      "loss": 1.5669,
+      "step": 5220
+    },
+    {
+      "epoch": 0.29102564102564105,
+      "grad_norm": 0.5718626976013184,
+      "learning_rate": 8e-05,
+      "loss": 1.7853,
+      "step": 5221
+    },
+    {
+      "epoch": 0.2910813823857302,
+      "grad_norm": 0.5383369326591492,
+      "learning_rate": 8e-05,
+      "loss": 1.8888,
+      "step": 5222
+    },
+    {
+      "epoch": 0.2911371237458194,
+      "grad_norm": 0.5096738338470459,
+      "learning_rate": 8e-05,
+      "loss": 1.6494,
+      "step": 5223
+    },
+    {
+      "epoch": 0.2911928651059086,
+      "grad_norm": 0.4973105788230896,
+      "learning_rate": 8e-05,
+      "loss": 1.7911,
+      "step": 5224
+    },
+    {
+      "epoch": 0.29124860646599776,
+      "grad_norm": 0.5213186144828796,
+      "learning_rate": 8e-05,
+      "loss": 1.6393,
+      "step": 5225
+    },
+    {
+      "epoch": 0.29130434782608694,
+      "grad_norm": 0.5377399325370789,
+      "learning_rate": 8e-05,
+      "loss": 1.8521,
+      "step": 5226
+    },
+    {
+      "epoch": 0.2913600891861761,
+      "grad_norm": 0.47744929790496826,
+      "learning_rate": 8e-05,
+      "loss": 1.5941,
+      "step": 5227
+    },
+    {
+      "epoch": 0.29141583054626535,
+      "grad_norm": 0.48108357191085815,
+      "learning_rate": 8e-05,
+      "loss": 1.544,
+      "step": 5228
+    },
+    {
+      "epoch": 0.29147157190635453,
+      "grad_norm": 0.5099393725395203,
+      "learning_rate": 8e-05,
+      "loss": 1.6005,
+      "step": 5229
+    },
+    {
+      "epoch": 0.2915273132664437,
+      "grad_norm": 0.5258042216300964,
+      "learning_rate": 8e-05,
+      "loss": 1.8084,
+      "step": 5230
+    },
+    {
+      "epoch": 0.2915830546265329,
+      "grad_norm": 0.5149073004722595,
+      "learning_rate": 8e-05,
+      "loss": 1.6328,
+      "step": 5231
+    },
+    {
+      "epoch": 0.29163879598662207,
+      "grad_norm": 0.5075093507766724,
+      "learning_rate": 8e-05,
+      "loss": 1.5967,
+      "step": 5232
+    },
+    {
+      "epoch": 0.29169453734671125,
+      "grad_norm": 0.5436050891876221,
+      "learning_rate": 8e-05,
+      "loss": 1.6569,
+      "step": 5233
+    },
+    {
+      "epoch": 0.2917502787068004,
+      "grad_norm": 0.6043053269386292,
+      "learning_rate": 8e-05,
+      "loss": 2.0332,
+      "step": 5234
+    },
+    {
+      "epoch": 0.29180602006688966,
+      "grad_norm": 0.5168697834014893,
+      "learning_rate": 8e-05,
+      "loss": 1.7984,
+      "step": 5235
+    },
+    {
+      "epoch": 0.29186176142697884,
+      "grad_norm": 0.5074122548103333,
+      "learning_rate": 8e-05,
+      "loss": 1.5661,
+      "step": 5236
+    },
+    {
+      "epoch": 0.291917502787068,
+      "grad_norm": 0.4924065172672272,
+      "learning_rate": 8e-05,
+      "loss": 1.7552,
+      "step": 5237
+    },
+    {
+      "epoch": 0.2919732441471572,
+      "grad_norm": 0.5427802205085754,
+      "learning_rate": 8e-05,
+      "loss": 1.7798,
+      "step": 5238
+    },
+    {
+      "epoch": 0.2920289855072464,
+      "grad_norm": 0.4995916187763214,
+      "learning_rate": 8e-05,
+      "loss": 1.67,
+      "step": 5239
+    },
+    {
+      "epoch": 0.29208472686733555,
+      "grad_norm": 0.49779626727104187,
+      "learning_rate": 8e-05,
+      "loss": 1.4751,
+      "step": 5240
+    },
+    {
+      "epoch": 0.29214046822742473,
+      "grad_norm": 0.46107199788093567,
+      "learning_rate": 8e-05,
+      "loss": 1.5479,
+      "step": 5241
+    },
+    {
+      "epoch": 0.2921962095875139,
+      "grad_norm": 0.534938633441925,
+      "learning_rate": 8e-05,
+      "loss": 1.969,
+      "step": 5242
+    },
+    {
+      "epoch": 0.29225195094760315,
+      "grad_norm": 0.5011855363845825,
+      "learning_rate": 8e-05,
+      "loss": 1.6897,
+      "step": 5243
+    },
+    {
+      "epoch": 0.2923076923076923,
+      "grad_norm": 0.4805475175380707,
+      "learning_rate": 8e-05,
+      "loss": 1.431,
+      "step": 5244
+    },
+    {
+      "epoch": 0.2923634336677815,
+      "grad_norm": 0.5417031645774841,
+      "learning_rate": 8e-05,
+      "loss": 2.0226,
+      "step": 5245
+    },
+    {
+      "epoch": 0.2924191750278707,
+      "grad_norm": 0.48827773332595825,
+      "learning_rate": 8e-05,
+      "loss": 1.6413,
+      "step": 5246
+    },
+    {
+      "epoch": 0.29247491638795986,
+      "grad_norm": 0.5585089921951294,
+      "learning_rate": 8e-05,
+      "loss": 1.7636,
+      "step": 5247
+    },
+    {
+      "epoch": 0.29253065774804904,
+      "grad_norm": 0.5031596422195435,
+      "learning_rate": 8e-05,
+      "loss": 1.9364,
+      "step": 5248
+    },
+    {
+      "epoch": 0.2925863991081382,
+      "grad_norm": 0.50601726770401,
+      "learning_rate": 8e-05,
+      "loss": 1.7698,
+      "step": 5249
+    },
+    {
+      "epoch": 0.29264214046822745,
+      "grad_norm": 0.5573868155479431,
+      "learning_rate": 8e-05,
+      "loss": 1.9684,
+      "step": 5250
+    },
+    {
+      "epoch": 0.29269788182831663,
+      "grad_norm": 0.599646270275116,
+      "learning_rate": 8e-05,
+      "loss": 1.7193,
+      "step": 5251
+    },
+    {
+      "epoch": 0.2927536231884058,
+      "grad_norm": 0.48568132519721985,
+      "learning_rate": 8e-05,
+      "loss": 1.5653,
+      "step": 5252
+    },
+    {
+      "epoch": 0.292809364548495,
+      "grad_norm": 0.47863563895225525,
+      "learning_rate": 8e-05,
+      "loss": 1.6626,
+      "step": 5253
+    },
+    {
+      "epoch": 0.29286510590858417,
+      "grad_norm": 0.4812254011631012,
+      "learning_rate": 8e-05,
+      "loss": 1.5981,
+      "step": 5254
+    },
+    {
+      "epoch": 0.29292084726867335,
+      "grad_norm": 0.5033372640609741,
+      "learning_rate": 8e-05,
+      "loss": 1.6784,
+      "step": 5255
+    },
+    {
+      "epoch": 0.2929765886287625,
+      "grad_norm": 0.5056328177452087,
+      "learning_rate": 8e-05,
+      "loss": 1.5998,
+      "step": 5256
+    },
+    {
+      "epoch": 0.2930323299888517,
+      "grad_norm": 0.4394751191139221,
+      "learning_rate": 8e-05,
+      "loss": 1.4819,
+      "step": 5257
+    },
+    {
+      "epoch": 0.29308807134894094,
+      "grad_norm": 0.5667262673377991,
+      "learning_rate": 8e-05,
+      "loss": 1.7092,
+      "step": 5258
+    },
+    {
+      "epoch": 0.2931438127090301,
+      "grad_norm": 0.5049179792404175,
+      "learning_rate": 8e-05,
+      "loss": 1.7625,
+      "step": 5259
+    },
+    {
+      "epoch": 0.2931995540691193,
+      "grad_norm": 0.5217584371566772,
+      "learning_rate": 8e-05,
+      "loss": 1.8257,
+      "step": 5260
+    },
+    {
+      "epoch": 0.2932552954292085,
+      "grad_norm": 0.5195755362510681,
+      "learning_rate": 8e-05,
+      "loss": 1.9529,
+      "step": 5261
+    },
+    {
+      "epoch": 0.29331103678929765,
+      "grad_norm": 0.49125170707702637,
+      "learning_rate": 8e-05,
+      "loss": 1.5536,
+      "step": 5262
+    },
+    {
+      "epoch": 0.29336677814938683,
+      "grad_norm": 0.49988195300102234,
+      "learning_rate": 8e-05,
+      "loss": 1.5557,
+      "step": 5263
+    },
+    {
+      "epoch": 0.293422519509476,
+      "grad_norm": 0.5084020495414734,
+      "learning_rate": 8e-05,
+      "loss": 1.4499,
+      "step": 5264
+    },
+    {
+      "epoch": 0.29347826086956524,
+      "grad_norm": 0.5017939805984497,
+      "learning_rate": 8e-05,
+      "loss": 1.7718,
+      "step": 5265
+    },
+    {
+      "epoch": 0.2935340022296544,
+      "grad_norm": 0.5145910978317261,
+      "learning_rate": 8e-05,
+      "loss": 1.838,
+      "step": 5266
+    },
+    {
+      "epoch": 0.2935897435897436,
+      "grad_norm": 0.516250729560852,
+      "learning_rate": 8e-05,
+      "loss": 1.7568,
+      "step": 5267
+    },
+    {
+      "epoch": 0.2936454849498328,
+      "grad_norm": 0.4948877990245819,
+      "learning_rate": 8e-05,
+      "loss": 1.7616,
+      "step": 5268
+    },
+    {
+      "epoch": 0.29370122630992196,
+      "grad_norm": 0.5350788831710815,
+      "learning_rate": 8e-05,
+      "loss": 1.9806,
+      "step": 5269
+    },
+    {
+      "epoch": 0.29375696767001114,
+      "grad_norm": 0.5112669467926025,
+      "learning_rate": 8e-05,
+      "loss": 1.7857,
+      "step": 5270
+    },
+    {
+      "epoch": 0.2938127090301003,
+      "grad_norm": 0.4432719945907593,
+      "learning_rate": 8e-05,
+      "loss": 1.4863,
+      "step": 5271
+    },
+    {
+      "epoch": 0.2938684503901895,
+      "grad_norm": 0.5258412957191467,
+      "learning_rate": 8e-05,
+      "loss": 1.7405,
+      "step": 5272
+    },
+    {
+      "epoch": 0.29392419175027873,
+      "grad_norm": 0.5164126753807068,
+      "learning_rate": 8e-05,
+      "loss": 1.7195,
+      "step": 5273
+    },
+    {
+      "epoch": 0.2939799331103679,
+      "grad_norm": 0.5118614435195923,
+      "learning_rate": 8e-05,
+      "loss": 1.744,
+      "step": 5274
+    },
+    {
+      "epoch": 0.2940356744704571,
+      "grad_norm": 0.5361714363098145,
+      "learning_rate": 8e-05,
+      "loss": 1.9204,
+      "step": 5275
+    },
+    {
+      "epoch": 0.29409141583054627,
+      "grad_norm": 0.5519838929176331,
+      "learning_rate": 8e-05,
+      "loss": 1.9456,
+      "step": 5276
+    },
+    {
+      "epoch": 0.29414715719063544,
+      "grad_norm": 0.5510969758033752,
+      "learning_rate": 8e-05,
+      "loss": 1.8676,
+      "step": 5277
+    },
+    {
+      "epoch": 0.2942028985507246,
+      "grad_norm": 0.502593994140625,
+      "learning_rate": 8e-05,
+      "loss": 1.6703,
+      "step": 5278
+    },
+    {
+      "epoch": 0.2942586399108138,
+      "grad_norm": 0.5204991698265076,
+      "learning_rate": 8e-05,
+      "loss": 1.6156,
+      "step": 5279
+    },
+    {
+      "epoch": 0.29431438127090304,
+      "grad_norm": 0.5098587870597839,
+      "learning_rate": 8e-05,
+      "loss": 1.7501,
+      "step": 5280
+    },
+    {
+      "epoch": 0.2943701226309922,
+      "grad_norm": 0.47406288981437683,
+      "learning_rate": 8e-05,
+      "loss": 1.5785,
+      "step": 5281
+    },
+    {
+      "epoch": 0.2944258639910814,
+      "grad_norm": 0.48030877113342285,
+      "learning_rate": 8e-05,
+      "loss": 1.4928,
+      "step": 5282
+    },
+    {
+      "epoch": 0.29448160535117057,
+      "grad_norm": 0.4848138689994812,
+      "learning_rate": 8e-05,
+      "loss": 1.718,
+      "step": 5283
+    },
+    {
+      "epoch": 0.29453734671125975,
+      "grad_norm": 0.503160297870636,
+      "learning_rate": 8e-05,
+      "loss": 1.7247,
+      "step": 5284
+    },
+    {
+      "epoch": 0.29459308807134893,
+      "grad_norm": 0.536435604095459,
+      "learning_rate": 8e-05,
+      "loss": 1.7329,
+      "step": 5285
+    },
+    {
+      "epoch": 0.2946488294314381,
+      "grad_norm": 0.5351728200912476,
+      "learning_rate": 8e-05,
+      "loss": 1.7665,
+      "step": 5286
+    },
+    {
+      "epoch": 0.2947045707915273,
+      "grad_norm": 0.47767600417137146,
+      "learning_rate": 8e-05,
+      "loss": 1.5381,
+      "step": 5287
+    },
+    {
+      "epoch": 0.2947603121516165,
+      "grad_norm": 0.5618817806243896,
+      "learning_rate": 8e-05,
+      "loss": 1.7063,
+      "step": 5288
+    },
+    {
+      "epoch": 0.2948160535117057,
+      "grad_norm": 0.5656747221946716,
+      "learning_rate": 8e-05,
+      "loss": 1.9849,
+      "step": 5289
+    },
+    {
+      "epoch": 0.2948717948717949,
+      "grad_norm": 0.5192245841026306,
+      "learning_rate": 8e-05,
+      "loss": 1.7184,
+      "step": 5290
+    },
+    {
+      "epoch": 0.29492753623188406,
+      "grad_norm": 0.5602400898933411,
+      "learning_rate": 8e-05,
+      "loss": 1.8306,
+      "step": 5291
+    },
+    {
+      "epoch": 0.29498327759197324,
+      "grad_norm": 0.504332423210144,
+      "learning_rate": 8e-05,
+      "loss": 1.7853,
+      "step": 5292
+    },
+    {
+      "epoch": 0.2950390189520624,
+      "grad_norm": 0.5190404653549194,
+      "learning_rate": 8e-05,
+      "loss": 1.9234,
+      "step": 5293
+    },
+    {
+      "epoch": 0.2950947603121516,
+      "grad_norm": 0.4828835427761078,
+      "learning_rate": 8e-05,
+      "loss": 1.7414,
+      "step": 5294
+    },
+    {
+      "epoch": 0.29515050167224083,
+      "grad_norm": 0.4870109260082245,
+      "learning_rate": 8e-05,
+      "loss": 1.5827,
+      "step": 5295
+    },
+    {
+      "epoch": 0.29520624303233,
+      "grad_norm": 0.49624398350715637,
+      "learning_rate": 8e-05,
+      "loss": 1.5151,
+      "step": 5296
+    },
+    {
+      "epoch": 0.2952619843924192,
+      "grad_norm": 0.531680703163147,
+      "learning_rate": 8e-05,
+      "loss": 1.8093,
+      "step": 5297
+    },
+    {
+      "epoch": 0.29531772575250836,
+      "grad_norm": 0.49582523107528687,
+      "learning_rate": 8e-05,
+      "loss": 1.649,
+      "step": 5298
+    },
+    {
+      "epoch": 0.29537346711259754,
+      "grad_norm": 0.48187878727912903,
+      "learning_rate": 8e-05,
+      "loss": 1.4792,
+      "step": 5299
+    },
+    {
+      "epoch": 0.2954292084726867,
+      "grad_norm": 0.5246399641036987,
+      "learning_rate": 8e-05,
+      "loss": 1.6916,
+      "step": 5300
+    },
+    {
+      "epoch": 0.2954849498327759,
+      "grad_norm": 0.5118151307106018,
+      "learning_rate": 8e-05,
+      "loss": 1.6442,
+      "step": 5301
+    },
+    {
+      "epoch": 0.2955406911928651,
+      "grad_norm": 0.5700403451919556,
+      "learning_rate": 8e-05,
+      "loss": 1.7347,
+      "step": 5302
+    },
+    {
+      "epoch": 0.2955964325529543,
+      "grad_norm": 0.4879356026649475,
+      "learning_rate": 8e-05,
+      "loss": 1.4308,
+      "step": 5303
+    },
+    {
+      "epoch": 0.2956521739130435,
+      "grad_norm": 0.4735146164894104,
+      "learning_rate": 8e-05,
+      "loss": 1.4481,
+      "step": 5304
+    },
+    {
+      "epoch": 0.29570791527313267,
+      "grad_norm": 0.5084589123725891,
+      "learning_rate": 8e-05,
+      "loss": 1.6398,
+      "step": 5305
+    },
+    {
+      "epoch": 0.29576365663322185,
+      "grad_norm": 0.5700600147247314,
+      "learning_rate": 8e-05,
+      "loss": 1.8626,
+      "step": 5306
+    },
+    {
+      "epoch": 0.29581939799331103,
+      "grad_norm": 0.5779445171356201,
+      "learning_rate": 8e-05,
+      "loss": 1.7356,
+      "step": 5307
+    },
+    {
+      "epoch": 0.2958751393534002,
+      "grad_norm": 0.5010429620742798,
+      "learning_rate": 8e-05,
+      "loss": 1.6907,
+      "step": 5308
+    },
+    {
+      "epoch": 0.2959308807134894,
+      "grad_norm": 0.5222609639167786,
+      "learning_rate": 8e-05,
+      "loss": 1.8282,
+      "step": 5309
+    },
+    {
+      "epoch": 0.2959866220735786,
+      "grad_norm": 0.5360494256019592,
+      "learning_rate": 8e-05,
+      "loss": 1.6219,
+      "step": 5310
+    },
+    {
+      "epoch": 0.2960423634336678,
+      "grad_norm": 0.5279106497764587,
+      "learning_rate": 8e-05,
+      "loss": 1.8183,
+      "step": 5311
+    },
+    {
+      "epoch": 0.296098104793757,
+      "grad_norm": 0.5564893484115601,
+      "learning_rate": 8e-05,
+      "loss": 1.8302,
+      "step": 5312
+    },
+    {
+      "epoch": 0.29615384615384616,
+      "grad_norm": 0.5432298183441162,
+      "learning_rate": 8e-05,
+      "loss": 1.9451,
+      "step": 5313
+    },
+    {
+      "epoch": 0.29620958751393534,
+      "grad_norm": 0.5052030086517334,
+      "learning_rate": 8e-05,
+      "loss": 1.7607,
+      "step": 5314
+    },
+    {
+      "epoch": 0.2962653288740245,
+      "grad_norm": 0.49412575364112854,
+      "learning_rate": 8e-05,
+      "loss": 1.6384,
+      "step": 5315
+    },
+    {
+      "epoch": 0.2963210702341137,
+      "grad_norm": 0.5110658407211304,
+      "learning_rate": 8e-05,
+      "loss": 1.9352,
+      "step": 5316
+    },
+    {
+      "epoch": 0.29637681159420287,
+      "grad_norm": 0.5185591578483582,
+      "learning_rate": 8e-05,
+      "loss": 1.6942,
+      "step": 5317
+    },
+    {
+      "epoch": 0.2964325529542921,
+      "grad_norm": 0.5007186532020569,
+      "learning_rate": 8e-05,
+      "loss": 1.6965,
+      "step": 5318
+    },
+    {
+      "epoch": 0.2964882943143813,
+      "grad_norm": 0.5010005831718445,
+      "learning_rate": 8e-05,
+      "loss": 1.6339,
+      "step": 5319
+    },
+    {
+      "epoch": 0.29654403567447046,
+      "grad_norm": 0.5402158498764038,
+      "learning_rate": 8e-05,
+      "loss": 1.8016,
+      "step": 5320
+    },
+    {
+      "epoch": 0.29659977703455964,
+      "grad_norm": 0.4774928689002991,
+      "learning_rate": 8e-05,
+      "loss": 1.7116,
+      "step": 5321
+    },
+    {
+      "epoch": 0.2966555183946488,
+      "grad_norm": 0.7964354157447815,
+      "learning_rate": 8e-05,
+      "loss": 1.6258,
+      "step": 5322
+    },
+    {
+      "epoch": 0.296711259754738,
+      "grad_norm": 0.5286622643470764,
+      "learning_rate": 8e-05,
+      "loss": 1.7428,
+      "step": 5323
+    },
+    {
+      "epoch": 0.2967670011148272,
+      "grad_norm": 0.5173184871673584,
+      "learning_rate": 8e-05,
+      "loss": 1.7258,
+      "step": 5324
+    },
+    {
+      "epoch": 0.2968227424749164,
+      "grad_norm": 0.5637500286102295,
+      "learning_rate": 8e-05,
+      "loss": 2.0714,
+      "step": 5325
+    },
+    {
+      "epoch": 0.2968784838350056,
+      "grad_norm": 0.5035580992698669,
+      "learning_rate": 8e-05,
+      "loss": 1.7906,
+      "step": 5326
+    },
+    {
+      "epoch": 0.29693422519509477,
+      "grad_norm": 0.5470683574676514,
+      "learning_rate": 8e-05,
+      "loss": 1.736,
+      "step": 5327
+    },
+    {
+      "epoch": 0.29698996655518395,
+      "grad_norm": 0.5390167832374573,
+      "learning_rate": 8e-05,
+      "loss": 1.8918,
+      "step": 5328
+    },
+    {
+      "epoch": 0.2970457079152731,
+      "grad_norm": 0.501208484172821,
+      "learning_rate": 8e-05,
+      "loss": 1.733,
+      "step": 5329
+    },
+    {
+      "epoch": 0.2971014492753623,
+      "grad_norm": 0.47988513112068176,
+      "learning_rate": 8e-05,
+      "loss": 1.5519,
+      "step": 5330
+    },
+    {
+      "epoch": 0.2971571906354515,
+      "grad_norm": 0.48085033893585205,
+      "learning_rate": 8e-05,
+      "loss": 1.6031,
+      "step": 5331
+    },
+    {
+      "epoch": 0.29721293199554066,
+      "grad_norm": 0.4778374433517456,
+      "learning_rate": 8e-05,
+      "loss": 1.5619,
+      "step": 5332
+    },
+    {
+      "epoch": 0.2972686733556299,
+      "grad_norm": 0.49800926446914673,
+      "learning_rate": 8e-05,
+      "loss": 1.6148,
+      "step": 5333
+    },
+    {
+      "epoch": 0.2973244147157191,
+      "grad_norm": 0.5318037271499634,
+      "learning_rate": 8e-05,
+      "loss": 1.8083,
+      "step": 5334
+    },
+    {
+      "epoch": 0.29738015607580826,
+      "grad_norm": 0.5206844210624695,
+      "learning_rate": 8e-05,
+      "loss": 1.7107,
+      "step": 5335
+    },
+    {
+      "epoch": 0.29743589743589743,
+      "grad_norm": 0.5456350445747375,
+      "learning_rate": 8e-05,
+      "loss": 1.7222,
+      "step": 5336
+    },
+    {
+      "epoch": 0.2974916387959866,
+      "grad_norm": 0.5354796648025513,
+      "learning_rate": 8e-05,
+      "loss": 1.511,
+      "step": 5337
+    },
+    {
+      "epoch": 0.2975473801560758,
+      "grad_norm": 0.5405011773109436,
+      "learning_rate": 8e-05,
+      "loss": 1.7553,
+      "step": 5338
+    },
+    {
+      "epoch": 0.29760312151616497,
+      "grad_norm": 0.5345763564109802,
+      "learning_rate": 8e-05,
+      "loss": 1.7713,
+      "step": 5339
+    },
+    {
+      "epoch": 0.2976588628762542,
+      "grad_norm": 0.4754033088684082,
+      "learning_rate": 8e-05,
+      "loss": 1.5024,
+      "step": 5340
+    },
+    {
+      "epoch": 0.2977146042363434,
+      "grad_norm": 0.45507925748825073,
+      "learning_rate": 8e-05,
+      "loss": 1.4949,
+      "step": 5341
+    },
+    {
+      "epoch": 0.29777034559643256,
+      "grad_norm": 0.49338576197624207,
+      "learning_rate": 8e-05,
+      "loss": 1.6169,
+      "step": 5342
+    },
+    {
+      "epoch": 0.29782608695652174,
+      "grad_norm": 0.48822399973869324,
+      "learning_rate": 8e-05,
+      "loss": 1.6804,
+      "step": 5343
+    },
+    {
+      "epoch": 0.2978818283166109,
+      "grad_norm": 0.4632672965526581,
+      "learning_rate": 8e-05,
+      "loss": 1.1824,
+      "step": 5344
+    },
+    {
+      "epoch": 0.2979375696767001,
+      "grad_norm": 0.5066871047019958,
+      "learning_rate": 8e-05,
+      "loss": 1.6845,
+      "step": 5345
+    },
+    {
+      "epoch": 0.2979933110367893,
+      "grad_norm": 0.4745085537433624,
+      "learning_rate": 8e-05,
+      "loss": 1.7372,
+      "step": 5346
+    },
+    {
+      "epoch": 0.2980490523968785,
+      "grad_norm": 0.47879093885421753,
+      "learning_rate": 8e-05,
+      "loss": 1.5042,
+      "step": 5347
+    },
+    {
+      "epoch": 0.2981047937569677,
+      "grad_norm": 0.4783836305141449,
+      "learning_rate": 8e-05,
+      "loss": 1.7095,
+      "step": 5348
+    },
+    {
+      "epoch": 0.29816053511705687,
+      "grad_norm": 0.4835946559906006,
+      "learning_rate": 8e-05,
+      "loss": 1.6553,
+      "step": 5349
+    },
+    {
+      "epoch": 0.29821627647714605,
+      "grad_norm": 0.5328168272972107,
+      "learning_rate": 8e-05,
+      "loss": 1.7335,
+      "step": 5350
+    },
+    {
+      "epoch": 0.2982720178372352,
+      "grad_norm": 0.5327500700950623,
+      "learning_rate": 8e-05,
+      "loss": 1.7231,
+      "step": 5351
+    },
+    {
+      "epoch": 0.2983277591973244,
+      "grad_norm": 0.6044172644615173,
+      "learning_rate": 8e-05,
+      "loss": 1.8734,
+      "step": 5352
+    },
+    {
+      "epoch": 0.2983835005574136,
+      "grad_norm": 0.5115371346473694,
+      "learning_rate": 8e-05,
+      "loss": 1.6742,
+      "step": 5353
+    },
+    {
+      "epoch": 0.29843924191750276,
+      "grad_norm": 0.570870041847229,
+      "learning_rate": 8e-05,
+      "loss": 1.798,
+      "step": 5354
+    },
+    {
+      "epoch": 0.298494983277592,
+      "grad_norm": 0.5445014834403992,
+      "learning_rate": 8e-05,
+      "loss": 1.6347,
+      "step": 5355
+    },
+    {
+      "epoch": 0.2985507246376812,
+      "grad_norm": 0.5148465037345886,
+      "learning_rate": 8e-05,
+      "loss": 1.7464,
+      "step": 5356
+    },
+    {
+      "epoch": 0.29860646599777035,
+      "grad_norm": 0.5145999789237976,
+      "learning_rate": 8e-05,
+      "loss": 1.6643,
+      "step": 5357
+    },
+    {
+      "epoch": 0.29866220735785953,
+      "grad_norm": 0.49865972995758057,
+      "learning_rate": 8e-05,
+      "loss": 1.6532,
+      "step": 5358
+    },
+    {
+      "epoch": 0.2987179487179487,
+      "grad_norm": 0.495136559009552,
+      "learning_rate": 8e-05,
+      "loss": 1.7584,
+      "step": 5359
+    },
+    {
+      "epoch": 0.2987736900780379,
+      "grad_norm": 0.5089300870895386,
+      "learning_rate": 8e-05,
+      "loss": 1.7906,
+      "step": 5360
+    },
+    {
+      "epoch": 0.29882943143812707,
+      "grad_norm": 0.485957533121109,
+      "learning_rate": 8e-05,
+      "loss": 1.5954,
+      "step": 5361
+    },
+    {
+      "epoch": 0.2988851727982163,
+      "grad_norm": 0.550250232219696,
+      "learning_rate": 8e-05,
+      "loss": 1.814,
+      "step": 5362
+    },
+    {
+      "epoch": 0.2989409141583055,
+      "grad_norm": 0.5336512327194214,
+      "learning_rate": 8e-05,
+      "loss": 1.8362,
+      "step": 5363
+    },
+    {
+      "epoch": 0.29899665551839466,
+      "grad_norm": 0.516875147819519,
+      "learning_rate": 8e-05,
+      "loss": 1.7615,
+      "step": 5364
+    },
+    {
+      "epoch": 0.29905239687848384,
+      "grad_norm": 0.5087434649467468,
+      "learning_rate": 8e-05,
+      "loss": 1.608,
+      "step": 5365
+    },
+    {
+      "epoch": 0.299108138238573,
+      "grad_norm": 0.5574606657028198,
+      "learning_rate": 8e-05,
+      "loss": 1.5601,
+      "step": 5366
+    },
+    {
+      "epoch": 0.2991638795986622,
+      "grad_norm": 0.5068855881690979,
+      "learning_rate": 8e-05,
+      "loss": 1.7151,
+      "step": 5367
+    },
+    {
+      "epoch": 0.2992196209587514,
+      "grad_norm": 0.4880894124507904,
+      "learning_rate": 8e-05,
+      "loss": 1.46,
+      "step": 5368
+    },
+    {
+      "epoch": 0.29927536231884055,
+      "grad_norm": 0.4971500635147095,
+      "learning_rate": 8e-05,
+      "loss": 1.6906,
+      "step": 5369
+    },
+    {
+      "epoch": 0.2993311036789298,
+      "grad_norm": 0.5621704459190369,
+      "learning_rate": 8e-05,
+      "loss": 1.9147,
+      "step": 5370
+    },
+    {
+      "epoch": 0.29938684503901897,
+      "grad_norm": 0.5409892201423645,
+      "learning_rate": 8e-05,
+      "loss": 1.8615,
+      "step": 5371
+    },
+    {
+      "epoch": 0.29944258639910815,
+      "grad_norm": 0.4971315562725067,
+      "learning_rate": 8e-05,
+      "loss": 1.6824,
+      "step": 5372
+    },
+    {
+      "epoch": 0.2994983277591973,
+      "grad_norm": 0.49553629755973816,
+      "learning_rate": 8e-05,
+      "loss": 1.7716,
+      "step": 5373
+    },
+    {
+      "epoch": 0.2995540691192865,
+      "grad_norm": 0.4982132017612457,
+      "learning_rate": 8e-05,
+      "loss": 1.8004,
+      "step": 5374
+    },
+    {
+      "epoch": 0.2996098104793757,
+      "grad_norm": 0.49931418895721436,
+      "learning_rate": 8e-05,
+      "loss": 1.6361,
+      "step": 5375
+    },
+    {
+      "epoch": 0.29966555183946486,
+      "grad_norm": 0.4859783947467804,
+      "learning_rate": 8e-05,
+      "loss": 1.5661,
+      "step": 5376
+    },
+    {
+      "epoch": 0.2997212931995541,
+      "grad_norm": 0.5062422156333923,
+      "learning_rate": 8e-05,
+      "loss": 1.4738,
+      "step": 5377
+    },
+    {
+      "epoch": 0.2997770345596433,
+      "grad_norm": 0.5458669066429138,
+      "learning_rate": 8e-05,
+      "loss": 1.9687,
+      "step": 5378
+    },
+    {
+      "epoch": 0.29983277591973245,
+      "grad_norm": 0.5416478514671326,
+      "learning_rate": 8e-05,
+      "loss": 2.02,
+      "step": 5379
+    },
+    {
+      "epoch": 0.29988851727982163,
+      "grad_norm": 0.5097154974937439,
+      "learning_rate": 8e-05,
+      "loss": 1.5451,
+      "step": 5380
+    },
+    {
+      "epoch": 0.2999442586399108,
+      "grad_norm": 0.4724648594856262,
+      "learning_rate": 8e-05,
+      "loss": 1.6004,
+      "step": 5381
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5238843560218811,
+      "learning_rate": 8e-05,
+      "loss": 1.8858,
+      "step": 5382
+    },
+    {
+      "epoch": 0.30005574136008917,
+      "grad_norm": 0.4457351863384247,
+      "learning_rate": 8e-05,
+      "loss": 1.4063,
+      "step": 5383
+    },
+    {
+      "epoch": 0.30011148272017835,
+      "grad_norm": 0.5066267848014832,
+      "learning_rate": 8e-05,
+      "loss": 1.9073,
+      "step": 5384
+    },
+    {
+      "epoch": 0.3001672240802676,
+      "grad_norm": 0.5180755853652954,
+      "learning_rate": 8e-05,
+      "loss": 1.7075,
+      "step": 5385
+    },
+    {
+      "epoch": 0.30022296544035676,
+      "grad_norm": 0.5159972906112671,
+      "learning_rate": 8e-05,
+      "loss": 1.7433,
+      "step": 5386
+    },
+    {
+      "epoch": 0.30027870680044594,
+      "grad_norm": 0.5109407305717468,
+      "learning_rate": 8e-05,
+      "loss": 1.5583,
+      "step": 5387
+    },
+    {
+      "epoch": 0.3003344481605351,
+      "grad_norm": 0.5248592495918274,
+      "learning_rate": 8e-05,
+      "loss": 1.7224,
+      "step": 5388
+    },
+    {
+      "epoch": 0.3003901895206243,
+      "grad_norm": 0.5111548900604248,
+      "learning_rate": 8e-05,
+      "loss": 1.7059,
+      "step": 5389
+    },
+    {
+      "epoch": 0.3004459308807135,
+      "grad_norm": 0.462700754404068,
+      "learning_rate": 8e-05,
+      "loss": 1.578,
+      "step": 5390
+    },
+    {
+      "epoch": 0.30050167224080265,
+      "grad_norm": 0.5779263377189636,
+      "learning_rate": 8e-05,
+      "loss": 1.8235,
+      "step": 5391
+    },
+    {
+      "epoch": 0.3005574136008919,
+      "grad_norm": 1.2829216718673706,
+      "learning_rate": 8e-05,
+      "loss": 1.575,
+      "step": 5392
+    },
+    {
+      "epoch": 0.30061315496098107,
+      "grad_norm": 0.5034168362617493,
+      "learning_rate": 8e-05,
+      "loss": 1.5587,
+      "step": 5393
+    },
+    {
+      "epoch": 0.30066889632107024,
+      "grad_norm": 0.49088746309280396,
+      "learning_rate": 8e-05,
+      "loss": 1.4346,
+      "step": 5394
+    },
+    {
+      "epoch": 0.3007246376811594,
+      "grad_norm": 0.5063590407371521,
+      "learning_rate": 8e-05,
+      "loss": 1.8368,
+      "step": 5395
+    },
+    {
+      "epoch": 0.3007803790412486,
+      "grad_norm": 0.5310071110725403,
+      "learning_rate": 8e-05,
+      "loss": 1.6953,
+      "step": 5396
+    },
+    {
+      "epoch": 0.3008361204013378,
+      "grad_norm": 0.5646356344223022,
+      "learning_rate": 8e-05,
+      "loss": 1.7004,
+      "step": 5397
+    },
+    {
+      "epoch": 0.30089186176142696,
+      "grad_norm": 0.4882517457008362,
+      "learning_rate": 8e-05,
+      "loss": 1.4631,
+      "step": 5398
+    },
+    {
+      "epoch": 0.30094760312151614,
+      "grad_norm": 0.5490998029708862,
+      "learning_rate": 8e-05,
+      "loss": 1.6273,
+      "step": 5399
+    },
+    {
+      "epoch": 0.3010033444816054,
+      "grad_norm": 0.4808351397514343,
+      "learning_rate": 8e-05,
+      "loss": 1.6429,
+      "step": 5400
+    },
+    {
+      "epoch": 0.30105908584169455,
+      "grad_norm": 0.49833354353904724,
+      "learning_rate": 8e-05,
+      "loss": 1.5635,
+      "step": 5401
+    },
+    {
+      "epoch": 0.30111482720178373,
+      "grad_norm": 0.5647911429405212,
+      "learning_rate": 8e-05,
+      "loss": 1.7225,
+      "step": 5402
+    },
+    {
+      "epoch": 0.3011705685618729,
+      "grad_norm": 0.5790539979934692,
+      "learning_rate": 8e-05,
+      "loss": 1.8063,
+      "step": 5403
+    },
+    {
+      "epoch": 0.3012263099219621,
+      "grad_norm": 0.44960105419158936,
+      "learning_rate": 8e-05,
+      "loss": 1.4754,
+      "step": 5404
+    },
+    {
+      "epoch": 0.30128205128205127,
+      "grad_norm": 0.5029364228248596,
+      "learning_rate": 8e-05,
+      "loss": 1.7869,
+      "step": 5405
+    },
+    {
+      "epoch": 0.30133779264214045,
+      "grad_norm": 0.5034712553024292,
+      "learning_rate": 8e-05,
+      "loss": 1.6554,
+      "step": 5406
+    },
+    {
+      "epoch": 0.3013935340022297,
+      "grad_norm": 0.5302921533584595,
+      "learning_rate": 8e-05,
+      "loss": 1.8542,
+      "step": 5407
+    },
+    {
+      "epoch": 0.30144927536231886,
+      "grad_norm": 0.5185369253158569,
+      "learning_rate": 8e-05,
+      "loss": 1.6084,
+      "step": 5408
+    },
+    {
+      "epoch": 0.30150501672240804,
+      "grad_norm": 0.48464417457580566,
+      "learning_rate": 8e-05,
+      "loss": 1.5505,
+      "step": 5409
+    },
+    {
+      "epoch": 0.3015607580824972,
+      "grad_norm": 0.5956443548202515,
+      "learning_rate": 8e-05,
+      "loss": 1.6358,
+      "step": 5410
+    },
+    {
+      "epoch": 0.3016164994425864,
+      "grad_norm": 0.5080055594444275,
+      "learning_rate": 8e-05,
+      "loss": 1.7066,
+      "step": 5411
+    },
+    {
+      "epoch": 0.3016722408026756,
+      "grad_norm": 0.5118548274040222,
+      "learning_rate": 8e-05,
+      "loss": 1.7763,
+      "step": 5412
+    },
+    {
+      "epoch": 0.30172798216276475,
+      "grad_norm": 0.5360879898071289,
+      "learning_rate": 8e-05,
+      "loss": 1.8087,
+      "step": 5413
+    },
+    {
+      "epoch": 0.30178372352285393,
+      "grad_norm": 0.4963992238044739,
+      "learning_rate": 8e-05,
+      "loss": 1.7252,
+      "step": 5414
+    },
+    {
+      "epoch": 0.30183946488294316,
+      "grad_norm": 0.5266253352165222,
+      "learning_rate": 8e-05,
+      "loss": 1.7846,
+      "step": 5415
+    },
+    {
+      "epoch": 0.30189520624303234,
+      "grad_norm": 0.5456589460372925,
+      "learning_rate": 8e-05,
+      "loss": 1.8456,
+      "step": 5416
+    },
+    {
+      "epoch": 0.3019509476031215,
+      "grad_norm": 0.5212188959121704,
+      "learning_rate": 8e-05,
+      "loss": 1.6749,
+      "step": 5417
+    },
+    {
+      "epoch": 0.3020066889632107,
+      "grad_norm": 0.550622284412384,
+      "learning_rate": 8e-05,
+      "loss": 1.6791,
+      "step": 5418
+    },
+    {
+      "epoch": 0.3020624303232999,
+      "grad_norm": 0.4872855246067047,
+      "learning_rate": 8e-05,
+      "loss": 1.6679,
+      "step": 5419
+    },
+    {
+      "epoch": 0.30211817168338906,
+      "grad_norm": 0.5363277196884155,
+      "learning_rate": 8e-05,
+      "loss": 1.8039,
+      "step": 5420
+    },
+    {
+      "epoch": 0.30217391304347824,
+      "grad_norm": 0.48044562339782715,
+      "learning_rate": 8e-05,
+      "loss": 1.5689,
+      "step": 5421
+    },
+    {
+      "epoch": 0.30222965440356747,
+      "grad_norm": 0.5340180993080139,
+      "learning_rate": 8e-05,
+      "loss": 1.6537,
+      "step": 5422
+    },
+    {
+      "epoch": 0.30228539576365665,
+      "grad_norm": 0.4993794560432434,
+      "learning_rate": 8e-05,
+      "loss": 1.7971,
+      "step": 5423
+    },
+    {
+      "epoch": 0.30234113712374583,
+      "grad_norm": 0.5311400294303894,
+      "learning_rate": 8e-05,
+      "loss": 1.7135,
+      "step": 5424
+    },
+    {
+      "epoch": 0.302396878483835,
+      "grad_norm": 0.5154658555984497,
+      "learning_rate": 8e-05,
+      "loss": 1.7769,
+      "step": 5425
+    },
+    {
+      "epoch": 0.3024526198439242,
+      "grad_norm": 0.5265001058578491,
+      "learning_rate": 8e-05,
+      "loss": 1.7175,
+      "step": 5426
+    },
+    {
+      "epoch": 0.30250836120401337,
+      "grad_norm": 0.540492594242096,
+      "learning_rate": 8e-05,
+      "loss": 1.8148,
+      "step": 5427
+    },
+    {
+      "epoch": 0.30256410256410254,
+      "grad_norm": 0.5124557614326477,
+      "learning_rate": 8e-05,
+      "loss": 1.7311,
+      "step": 5428
+    },
+    {
+      "epoch": 0.3026198439241917,
+      "grad_norm": 0.5288158059120178,
+      "learning_rate": 8e-05,
+      "loss": 1.8849,
+      "step": 5429
+    },
+    {
+      "epoch": 0.30267558528428096,
+      "grad_norm": 0.4883910119533539,
+      "learning_rate": 8e-05,
+      "loss": 1.6892,
+      "step": 5430
+    },
+    {
+      "epoch": 0.30273132664437014,
+      "grad_norm": 0.5443893074989319,
+      "learning_rate": 8e-05,
+      "loss": 2.0222,
+      "step": 5431
+    },
+    {
+      "epoch": 0.3027870680044593,
+      "grad_norm": 0.5112580060958862,
+      "learning_rate": 8e-05,
+      "loss": 1.7861,
+      "step": 5432
+    },
+    {
+      "epoch": 0.3028428093645485,
+      "grad_norm": 0.4806899428367615,
+      "learning_rate": 8e-05,
+      "loss": 1.7085,
+      "step": 5433
+    },
+    {
+      "epoch": 0.30289855072463767,
+      "grad_norm": 0.4477013647556305,
+      "learning_rate": 8e-05,
+      "loss": 1.2138,
+      "step": 5434
+    },
+    {
+      "epoch": 0.30295429208472685,
+      "grad_norm": 0.5485127568244934,
+      "learning_rate": 8e-05,
+      "loss": 1.8087,
+      "step": 5435
+    },
+    {
+      "epoch": 0.30301003344481603,
+      "grad_norm": 0.5672134757041931,
+      "learning_rate": 8e-05,
+      "loss": 1.8478,
+      "step": 5436
+    },
+    {
+      "epoch": 0.30306577480490526,
+      "grad_norm": 0.4998778998851776,
+      "learning_rate": 8e-05,
+      "loss": 1.6146,
+      "step": 5437
+    },
+    {
+      "epoch": 0.30312151616499444,
+      "grad_norm": 0.5240135788917542,
+      "learning_rate": 8e-05,
+      "loss": 1.6686,
+      "step": 5438
+    },
+    {
+      "epoch": 0.3031772575250836,
+      "grad_norm": 0.5832547545433044,
+      "learning_rate": 8e-05,
+      "loss": 1.9084,
+      "step": 5439
+    },
+    {
+      "epoch": 0.3032329988851728,
+      "grad_norm": 0.4926720857620239,
+      "learning_rate": 8e-05,
+      "loss": 1.675,
+      "step": 5440
+    },
+    {
+      "epoch": 0.303288740245262,
+      "grad_norm": 0.5205129981040955,
+      "learning_rate": 8e-05,
+      "loss": 1.7816,
+      "step": 5441
+    },
+    {
+      "epoch": 0.30334448160535116,
+      "grad_norm": 0.4628882110118866,
+      "learning_rate": 8e-05,
+      "loss": 1.4962,
+      "step": 5442
+    },
+    {
+      "epoch": 0.30340022296544034,
+      "grad_norm": 0.5137123465538025,
+      "learning_rate": 8e-05,
+      "loss": 1.651,
+      "step": 5443
+    },
+    {
+      "epoch": 0.30345596432552957,
+      "grad_norm": 0.5469315648078918,
+      "learning_rate": 8e-05,
+      "loss": 1.553,
+      "step": 5444
+    },
+    {
+      "epoch": 0.30351170568561875,
+      "grad_norm": 0.5205637812614441,
+      "learning_rate": 8e-05,
+      "loss": 1.7943,
+      "step": 5445
+    },
+    {
+      "epoch": 0.3035674470457079,
+      "grad_norm": 0.54112309217453,
+      "learning_rate": 8e-05,
+      "loss": 1.7074,
+      "step": 5446
+    },
+    {
+      "epoch": 0.3036231884057971,
+      "grad_norm": 0.4951598644256592,
+      "learning_rate": 8e-05,
+      "loss": 1.792,
+      "step": 5447
+    },
+    {
+      "epoch": 0.3036789297658863,
+      "grad_norm": 0.5104148983955383,
+      "learning_rate": 8e-05,
+      "loss": 1.807,
+      "step": 5448
+    },
+    {
+      "epoch": 0.30373467112597546,
+      "grad_norm": 0.5905737280845642,
+      "learning_rate": 8e-05,
+      "loss": 1.6313,
+      "step": 5449
+    },
+    {
+      "epoch": 0.30379041248606464,
+      "grad_norm": 0.4556120038032532,
+      "learning_rate": 8e-05,
+      "loss": 1.2949,
+      "step": 5450
+    },
+    {
+      "epoch": 0.3038461538461538,
+      "grad_norm": 0.5705082416534424,
+      "learning_rate": 8e-05,
+      "loss": 1.8666,
+      "step": 5451
+    },
+    {
+      "epoch": 0.30390189520624306,
+      "grad_norm": 0.5135024785995483,
+      "learning_rate": 8e-05,
+      "loss": 1.858,
+      "step": 5452
+    },
+    {
+      "epoch": 0.30395763656633223,
+      "grad_norm": 0.6063351631164551,
+      "learning_rate": 8e-05,
+      "loss": 1.831,
+      "step": 5453
+    },
+    {
+      "epoch": 0.3040133779264214,
+      "grad_norm": 0.5353469252586365,
+      "learning_rate": 8e-05,
+      "loss": 1.7517,
+      "step": 5454
+    },
+    {
+      "epoch": 0.3040691192865106,
+      "grad_norm": 0.5837903618812561,
+      "learning_rate": 8e-05,
+      "loss": 2.0732,
+      "step": 5455
+    },
+    {
+      "epoch": 0.30412486064659977,
+      "grad_norm": 0.46746090054512024,
+      "learning_rate": 8e-05,
+      "loss": 1.5466,
+      "step": 5456
+    },
+    {
+      "epoch": 0.30418060200668895,
+      "grad_norm": 0.5374937653541565,
+      "learning_rate": 8e-05,
+      "loss": 1.8763,
+      "step": 5457
+    },
+    {
+      "epoch": 0.30423634336677813,
+      "grad_norm": 0.5043895244598389,
+      "learning_rate": 8e-05,
+      "loss": 1.8551,
+      "step": 5458
+    },
+    {
+      "epoch": 0.30429208472686736,
+      "grad_norm": 0.522622287273407,
+      "learning_rate": 8e-05,
+      "loss": 1.7695,
+      "step": 5459
+    },
+    {
+      "epoch": 0.30434782608695654,
+      "grad_norm": 0.5085619688034058,
+      "learning_rate": 8e-05,
+      "loss": 1.7313,
+      "step": 5460
+    },
+    {
+      "epoch": 0.3044035674470457,
+      "grad_norm": 0.5232932567596436,
+      "learning_rate": 8e-05,
+      "loss": 1.5676,
+      "step": 5461
+    },
+    {
+      "epoch": 0.3044593088071349,
+      "grad_norm": 0.499271422624588,
+      "learning_rate": 8e-05,
+      "loss": 1.8006,
+      "step": 5462
+    },
+    {
+      "epoch": 0.3045150501672241,
+      "grad_norm": 0.5241374969482422,
+      "learning_rate": 8e-05,
+      "loss": 1.7208,
+      "step": 5463
+    },
+    {
+      "epoch": 0.30457079152731326,
+      "grad_norm": 0.5245078206062317,
+      "learning_rate": 8e-05,
+      "loss": 1.9892,
+      "step": 5464
+    },
+    {
+      "epoch": 0.30462653288740243,
+      "grad_norm": 0.49973732233047485,
+      "learning_rate": 8e-05,
+      "loss": 1.6267,
+      "step": 5465
+    },
+    {
+      "epoch": 0.3046822742474916,
+      "grad_norm": 0.5292554497718811,
+      "learning_rate": 8e-05,
+      "loss": 1.3804,
+      "step": 5466
+    },
+    {
+      "epoch": 0.30473801560758085,
+      "grad_norm": 0.5208456516265869,
+      "learning_rate": 8e-05,
+      "loss": 1.6966,
+      "step": 5467
+    },
+    {
+      "epoch": 0.30479375696767,
+      "grad_norm": 0.49586382508277893,
+      "learning_rate": 8e-05,
+      "loss": 1.7829,
+      "step": 5468
+    },
+    {
+      "epoch": 0.3048494983277592,
+      "grad_norm": 0.5234887599945068,
+      "learning_rate": 8e-05,
+      "loss": 1.6937,
+      "step": 5469
+    },
+    {
+      "epoch": 0.3049052396878484,
+      "grad_norm": 0.49152565002441406,
+      "learning_rate": 8e-05,
+      "loss": 1.7413,
+      "step": 5470
+    },
+    {
+      "epoch": 0.30496098104793756,
+      "grad_norm": 0.4878113269805908,
+      "learning_rate": 8e-05,
+      "loss": 1.528,
+      "step": 5471
+    },
+    {
+      "epoch": 0.30501672240802674,
+      "grad_norm": 0.6157654523849487,
+      "learning_rate": 8e-05,
+      "loss": 1.8596,
+      "step": 5472
+    },
+    {
+      "epoch": 0.3050724637681159,
+      "grad_norm": 0.5470446944236755,
+      "learning_rate": 8e-05,
+      "loss": 1.7643,
+      "step": 5473
+    },
+    {
+      "epoch": 0.30512820512820515,
+      "grad_norm": 0.4969521164894104,
+      "learning_rate": 8e-05,
+      "loss": 1.5607,
+      "step": 5474
+    },
+    {
+      "epoch": 0.30518394648829433,
+      "grad_norm": 0.4928933084011078,
+      "learning_rate": 8e-05,
+      "loss": 1.7484,
+      "step": 5475
+    },
+    {
+      "epoch": 0.3052396878483835,
+      "grad_norm": 0.5175816416740417,
+      "learning_rate": 8e-05,
+      "loss": 1.6989,
+      "step": 5476
+    },
+    {
+      "epoch": 0.3052954292084727,
+      "grad_norm": 0.6144953370094299,
+      "learning_rate": 8e-05,
+      "loss": 1.7375,
+      "step": 5477
+    },
+    {
+      "epoch": 0.30535117056856187,
+      "grad_norm": 0.5331835150718689,
+      "learning_rate": 8e-05,
+      "loss": 1.7825,
+      "step": 5478
+    },
+    {
+      "epoch": 0.30540691192865105,
+      "grad_norm": 0.49090802669525146,
+      "learning_rate": 8e-05,
+      "loss": 1.6912,
+      "step": 5479
+    },
+    {
+      "epoch": 0.3054626532887402,
+      "grad_norm": 0.5607184767723083,
+      "learning_rate": 8e-05,
+      "loss": 1.8406,
+      "step": 5480
+    },
+    {
+      "epoch": 0.3055183946488294,
+      "grad_norm": 0.48209232091903687,
+      "learning_rate": 8e-05,
+      "loss": 1.7219,
+      "step": 5481
+    },
+    {
+      "epoch": 0.30557413600891864,
+      "grad_norm": 0.5262097120285034,
+      "learning_rate": 8e-05,
+      "loss": 1.7903,
+      "step": 5482
+    },
+    {
+      "epoch": 0.3056298773690078,
+      "grad_norm": 0.5120763182640076,
+      "learning_rate": 8e-05,
+      "loss": 1.5862,
+      "step": 5483
+    },
+    {
+      "epoch": 0.305685618729097,
+      "grad_norm": 0.5171751379966736,
+      "learning_rate": 8e-05,
+      "loss": 1.6802,
+      "step": 5484
+    },
+    {
+      "epoch": 0.3057413600891862,
+      "grad_norm": 0.4827384352684021,
+      "learning_rate": 8e-05,
+      "loss": 1.6011,
+      "step": 5485
+    },
+    {
+      "epoch": 0.30579710144927535,
+      "grad_norm": 0.5480244159698486,
+      "learning_rate": 8e-05,
+      "loss": 1.7993,
+      "step": 5486
+    },
+    {
+      "epoch": 0.30585284280936453,
+      "grad_norm": 0.506036639213562,
+      "learning_rate": 8e-05,
+      "loss": 1.7115,
+      "step": 5487
+    },
+    {
+      "epoch": 0.3059085841694537,
+      "grad_norm": 0.4862762689590454,
+      "learning_rate": 8e-05,
+      "loss": 1.5576,
+      "step": 5488
+    },
+    {
+      "epoch": 0.30596432552954295,
+      "grad_norm": 0.5329150557518005,
+      "learning_rate": 8e-05,
+      "loss": 1.7333,
+      "step": 5489
+    },
+    {
+      "epoch": 0.3060200668896321,
+      "grad_norm": 0.5178322196006775,
+      "learning_rate": 8e-05,
+      "loss": 1.5498,
+      "step": 5490
+    },
+    {
+      "epoch": 0.3060758082497213,
+      "grad_norm": 0.4974115490913391,
+      "learning_rate": 8e-05,
+      "loss": 1.7897,
+      "step": 5491
+    },
+    {
+      "epoch": 0.3061315496098105,
+      "grad_norm": 0.494586318731308,
+      "learning_rate": 8e-05,
+      "loss": 1.6622,
+      "step": 5492
+    },
+    {
+      "epoch": 0.30618729096989966,
+      "grad_norm": 0.48024600744247437,
+      "learning_rate": 8e-05,
+      "loss": 1.6855,
+      "step": 5493
+    },
+    {
+      "epoch": 0.30624303232998884,
+      "grad_norm": 0.5140489339828491,
+      "learning_rate": 8e-05,
+      "loss": 1.706,
+      "step": 5494
+    },
+    {
+      "epoch": 0.306298773690078,
+      "grad_norm": 0.5151708722114563,
+      "learning_rate": 8e-05,
+      "loss": 1.6387,
+      "step": 5495
+    },
+    {
+      "epoch": 0.3063545150501672,
+      "grad_norm": 0.4907633364200592,
+      "learning_rate": 8e-05,
+      "loss": 1.6687,
+      "step": 5496
+    },
+    {
+      "epoch": 0.30641025641025643,
+      "grad_norm": 0.5302457809448242,
+      "learning_rate": 8e-05,
+      "loss": 1.7727,
+      "step": 5497
+    },
+    {
+      "epoch": 0.3064659977703456,
+      "grad_norm": 0.47479885816574097,
+      "learning_rate": 8e-05,
+      "loss": 1.6267,
+      "step": 5498
+    },
+    {
+      "epoch": 0.3065217391304348,
+      "grad_norm": 0.5059468746185303,
+      "learning_rate": 8e-05,
+      "loss": 1.642,
+      "step": 5499
+    },
+    {
+      "epoch": 0.30657748049052397,
+      "grad_norm": 0.46230581402778625,
+      "learning_rate": 8e-05,
+      "loss": 1.5666,
+      "step": 5500
+    },
+    {
+      "epoch": 0.30663322185061315,
+      "grad_norm": 0.49751943349838257,
+      "learning_rate": 8e-05,
+      "loss": 1.6127,
+      "step": 5501
+    },
+    {
+      "epoch": 0.3066889632107023,
+      "grad_norm": 0.5246805548667908,
+      "learning_rate": 8e-05,
+      "loss": 1.9026,
+      "step": 5502
+    },
+    {
+      "epoch": 0.3067447045707915,
+      "grad_norm": 0.5045078992843628,
+      "learning_rate": 8e-05,
+      "loss": 1.5361,
+      "step": 5503
+    },
+    {
+      "epoch": 0.30680044593088074,
+      "grad_norm": 0.5169194340705872,
+      "learning_rate": 8e-05,
+      "loss": 1.7405,
+      "step": 5504
+    },
+    {
+      "epoch": 0.3068561872909699,
+      "grad_norm": 0.521348774433136,
+      "learning_rate": 8e-05,
+      "loss": 1.7625,
+      "step": 5505
+    },
+    {
+      "epoch": 0.3069119286510591,
+      "grad_norm": 0.5427587032318115,
+      "learning_rate": 8e-05,
+      "loss": 1.8069,
+      "step": 5506
+    },
+    {
+      "epoch": 0.3069676700111483,
+      "grad_norm": 0.5333295464515686,
+      "learning_rate": 8e-05,
+      "loss": 1.707,
+      "step": 5507
+    },
+    {
+      "epoch": 0.30702341137123745,
+      "grad_norm": 0.5025301575660706,
+      "learning_rate": 8e-05,
+      "loss": 1.4767,
+      "step": 5508
+    },
+    {
+      "epoch": 0.30707915273132663,
+      "grad_norm": 0.5663952827453613,
+      "learning_rate": 8e-05,
+      "loss": 1.8625,
+      "step": 5509
+    },
+    {
+      "epoch": 0.3071348940914158,
+      "grad_norm": 0.4802990257740021,
+      "learning_rate": 8e-05,
+      "loss": 1.5655,
+      "step": 5510
+    },
+    {
+      "epoch": 0.307190635451505,
+      "grad_norm": 0.5298011898994446,
+      "learning_rate": 8e-05,
+      "loss": 1.8292,
+      "step": 5511
+    },
+    {
+      "epoch": 0.3072463768115942,
+      "grad_norm": 0.5581608414649963,
+      "learning_rate": 8e-05,
+      "loss": 1.7637,
+      "step": 5512
+    },
+    {
+      "epoch": 0.3073021181716834,
+      "grad_norm": 0.5245633125305176,
+      "learning_rate": 8e-05,
+      "loss": 1.8399,
+      "step": 5513
+    },
+    {
+      "epoch": 0.3073578595317726,
+      "grad_norm": 0.5069871544837952,
+      "learning_rate": 8e-05,
+      "loss": 1.669,
+      "step": 5514
+    },
+    {
+      "epoch": 0.30741360089186176,
+      "grad_norm": 0.5340516567230225,
+      "learning_rate": 8e-05,
+      "loss": 1.802,
+      "step": 5515
+    },
+    {
+      "epoch": 0.30746934225195094,
+      "grad_norm": 0.506650447845459,
+      "learning_rate": 8e-05,
+      "loss": 1.7352,
+      "step": 5516
+    },
+    {
+      "epoch": 0.3075250836120401,
+      "grad_norm": 0.5162363052368164,
+      "learning_rate": 8e-05,
+      "loss": 1.8099,
+      "step": 5517
+    },
+    {
+      "epoch": 0.3075808249721293,
+      "grad_norm": 0.4924274682998657,
+      "learning_rate": 8e-05,
+      "loss": 1.6239,
+      "step": 5518
+    },
+    {
+      "epoch": 0.30763656633221853,
+      "grad_norm": 0.47701767086982727,
+      "learning_rate": 8e-05,
+      "loss": 1.6575,
+      "step": 5519
+    },
+    {
+      "epoch": 0.3076923076923077,
+      "grad_norm": 0.49266427755355835,
+      "learning_rate": 8e-05,
+      "loss": 1.8003,
+      "step": 5520
+    },
+    {
+      "epoch": 0.3077480490523969,
+      "grad_norm": 0.5872482061386108,
+      "learning_rate": 8e-05,
+      "loss": 1.6614,
+      "step": 5521
+    },
+    {
+      "epoch": 0.30780379041248607,
+      "grad_norm": 0.5668823719024658,
+      "learning_rate": 8e-05,
+      "loss": 1.5553,
+      "step": 5522
+    },
+    {
+      "epoch": 0.30785953177257525,
+      "grad_norm": 0.4845235347747803,
+      "learning_rate": 8e-05,
+      "loss": 1.7007,
+      "step": 5523
+    },
+    {
+      "epoch": 0.3079152731326644,
+      "grad_norm": 0.48626527190208435,
+      "learning_rate": 8e-05,
+      "loss": 1.7046,
+      "step": 5524
+    },
+    {
+      "epoch": 0.3079710144927536,
+      "grad_norm": 0.5815119743347168,
+      "learning_rate": 8e-05,
+      "loss": 2.0158,
+      "step": 5525
+    },
+    {
+      "epoch": 0.3080267558528428,
+      "grad_norm": 0.4801939129829407,
+      "learning_rate": 8e-05,
+      "loss": 1.5951,
+      "step": 5526
+    },
+    {
+      "epoch": 0.308082497212932,
+      "grad_norm": 0.5311259627342224,
+      "learning_rate": 8e-05,
+      "loss": 2.0277,
+      "step": 5527
+    },
+    {
+      "epoch": 0.3081382385730212,
+      "grad_norm": 0.5235937237739563,
+      "learning_rate": 8e-05,
+      "loss": 1.9141,
+      "step": 5528
+    },
+    {
+      "epoch": 0.3081939799331104,
+      "grad_norm": 0.5488254427909851,
+      "learning_rate": 8e-05,
+      "loss": 1.8013,
+      "step": 5529
+    },
+    {
+      "epoch": 0.30824972129319955,
+      "grad_norm": 0.4970884919166565,
+      "learning_rate": 8e-05,
+      "loss": 1.7506,
+      "step": 5530
+    },
+    {
+      "epoch": 0.30830546265328873,
+      "grad_norm": 0.5055081844329834,
+      "learning_rate": 8e-05,
+      "loss": 1.8335,
+      "step": 5531
+    },
+    {
+      "epoch": 0.3083612040133779,
+      "grad_norm": 0.4809627830982208,
+      "learning_rate": 8e-05,
+      "loss": 1.6935,
+      "step": 5532
+    },
+    {
+      "epoch": 0.3084169453734671,
+      "grad_norm": 0.47954025864601135,
+      "learning_rate": 8e-05,
+      "loss": 1.6094,
+      "step": 5533
+    },
+    {
+      "epoch": 0.3084726867335563,
+      "grad_norm": 0.5082470178604126,
+      "learning_rate": 8e-05,
+      "loss": 1.5929,
+      "step": 5534
+    },
+    {
+      "epoch": 0.3085284280936455,
+      "grad_norm": 0.5128086805343628,
+      "learning_rate": 8e-05,
+      "loss": 1.6322,
+      "step": 5535
+    },
+    {
+      "epoch": 0.3085841694537347,
+      "grad_norm": 0.4984266459941864,
+      "learning_rate": 8e-05,
+      "loss": 1.7852,
+      "step": 5536
+    },
+    {
+      "epoch": 0.30863991081382386,
+      "grad_norm": 0.49187591671943665,
+      "learning_rate": 8e-05,
+      "loss": 1.5455,
+      "step": 5537
+    },
+    {
+      "epoch": 0.30869565217391304,
+      "grad_norm": 0.5065392851829529,
+      "learning_rate": 8e-05,
+      "loss": 1.6408,
+      "step": 5538
+    },
+    {
+      "epoch": 0.3087513935340022,
+      "grad_norm": 0.511028528213501,
+      "learning_rate": 8e-05,
+      "loss": 1.566,
+      "step": 5539
+    },
+    {
+      "epoch": 0.3088071348940914,
+      "grad_norm": 0.5834546685218811,
+      "learning_rate": 8e-05,
+      "loss": 1.6334,
+      "step": 5540
+    },
+    {
+      "epoch": 0.30886287625418063,
+      "grad_norm": 0.5535149574279785,
+      "learning_rate": 8e-05,
+      "loss": 1.776,
+      "step": 5541
+    },
+    {
+      "epoch": 0.3089186176142698,
+      "grad_norm": 0.5388762354850769,
+      "learning_rate": 8e-05,
+      "loss": 1.2867,
+      "step": 5542
+    },
+    {
+      "epoch": 0.308974358974359,
+      "grad_norm": 0.5357641577720642,
+      "learning_rate": 8e-05,
+      "loss": 1.8904,
+      "step": 5543
+    },
+    {
+      "epoch": 0.30903010033444817,
+      "grad_norm": 0.5241168141365051,
+      "learning_rate": 8e-05,
+      "loss": 1.7835,
+      "step": 5544
+    },
+    {
+      "epoch": 0.30908584169453734,
+      "grad_norm": 0.5548775792121887,
+      "learning_rate": 8e-05,
+      "loss": 1.8013,
+      "step": 5545
+    },
+    {
+      "epoch": 0.3091415830546265,
+      "grad_norm": 0.4995080828666687,
+      "learning_rate": 8e-05,
+      "loss": 1.6661,
+      "step": 5546
+    },
+    {
+      "epoch": 0.3091973244147157,
+      "grad_norm": 0.5398724675178528,
+      "learning_rate": 8e-05,
+      "loss": 1.8868,
+      "step": 5547
+    },
+    {
+      "epoch": 0.3092530657748049,
+      "grad_norm": 0.4948381185531616,
+      "learning_rate": 8e-05,
+      "loss": 1.8113,
+      "step": 5548
+    },
+    {
+      "epoch": 0.3093088071348941,
+      "grad_norm": 0.5208935737609863,
+      "learning_rate": 8e-05,
+      "loss": 1.7515,
+      "step": 5549
+    },
+    {
+      "epoch": 0.3093645484949833,
+      "grad_norm": 0.5310214757919312,
+      "learning_rate": 8e-05,
+      "loss": 1.6944,
+      "step": 5550
+    },
+    {
+      "epoch": 0.3094202898550725,
+      "grad_norm": 0.5287116765975952,
+      "learning_rate": 8e-05,
+      "loss": 1.797,
+      "step": 5551
+    },
+    {
+      "epoch": 0.30947603121516165,
+      "grad_norm": 0.5271665453910828,
+      "learning_rate": 8e-05,
+      "loss": 1.677,
+      "step": 5552
+    },
+    {
+      "epoch": 0.30953177257525083,
+      "grad_norm": 0.46743589639663696,
+      "learning_rate": 8e-05,
+      "loss": 1.5052,
+      "step": 5553
+    },
+    {
+      "epoch": 0.30958751393534,
+      "grad_norm": 0.4670599400997162,
+      "learning_rate": 8e-05,
+      "loss": 1.6102,
+      "step": 5554
+    },
+    {
+      "epoch": 0.3096432552954292,
+      "grad_norm": 0.5504299402236938,
+      "learning_rate": 8e-05,
+      "loss": 1.943,
+      "step": 5555
+    },
+    {
+      "epoch": 0.3096989966555184,
+      "grad_norm": 0.516667902469635,
+      "learning_rate": 8e-05,
+      "loss": 1.8104,
+      "step": 5556
+    },
+    {
+      "epoch": 0.3097547380156076,
+      "grad_norm": 0.5134239196777344,
+      "learning_rate": 8e-05,
+      "loss": 1.6219,
+      "step": 5557
+    },
+    {
+      "epoch": 0.3098104793756968,
+      "grad_norm": 0.5604968667030334,
+      "learning_rate": 8e-05,
+      "loss": 1.5912,
+      "step": 5558
+    },
+    {
+      "epoch": 0.30986622073578596,
+      "grad_norm": 0.4922977685928345,
+      "learning_rate": 8e-05,
+      "loss": 1.5991,
+      "step": 5559
+    },
+    {
+      "epoch": 0.30992196209587514,
+      "grad_norm": 0.49453943967819214,
+      "learning_rate": 8e-05,
+      "loss": 1.6388,
+      "step": 5560
+    },
+    {
+      "epoch": 0.3099777034559643,
+      "grad_norm": 0.48977014422416687,
+      "learning_rate": 8e-05,
+      "loss": 1.5058,
+      "step": 5561
+    },
+    {
+      "epoch": 0.3100334448160535,
+      "grad_norm": 0.5432201623916626,
+      "learning_rate": 8e-05,
+      "loss": 1.6409,
+      "step": 5562
+    },
+    {
+      "epoch": 0.3100891861761427,
+      "grad_norm": 0.560623288154602,
+      "learning_rate": 8e-05,
+      "loss": 1.8318,
+      "step": 5563
+    },
+    {
+      "epoch": 0.3101449275362319,
+      "grad_norm": 0.509880542755127,
+      "learning_rate": 8e-05,
+      "loss": 1.855,
+      "step": 5564
+    },
+    {
+      "epoch": 0.3102006688963211,
+      "grad_norm": 0.5146164298057556,
+      "learning_rate": 8e-05,
+      "loss": 1.5432,
+      "step": 5565
+    },
+    {
+      "epoch": 0.31025641025641026,
+      "grad_norm": 0.5598461031913757,
+      "learning_rate": 8e-05,
+      "loss": 1.8373,
+      "step": 5566
+    },
+    {
+      "epoch": 0.31031215161649944,
+      "grad_norm": 0.4729394018650055,
+      "learning_rate": 8e-05,
+      "loss": 1.5871,
+      "step": 5567
+    },
+    {
+      "epoch": 0.3103678929765886,
+      "grad_norm": 0.47193753719329834,
+      "learning_rate": 8e-05,
+      "loss": 1.5819,
+      "step": 5568
+    },
+    {
+      "epoch": 0.3104236343366778,
+      "grad_norm": 0.5307363867759705,
+      "learning_rate": 8e-05,
+      "loss": 1.6914,
+      "step": 5569
+    },
+    {
+      "epoch": 0.310479375696767,
+      "grad_norm": 0.5264828205108643,
+      "learning_rate": 8e-05,
+      "loss": 1.6782,
+      "step": 5570
+    },
+    {
+      "epoch": 0.3105351170568562,
+      "grad_norm": 0.5106006860733032,
+      "learning_rate": 8e-05,
+      "loss": 1.6445,
+      "step": 5571
+    },
+    {
+      "epoch": 0.3105908584169454,
+      "grad_norm": 0.49598777294158936,
+      "learning_rate": 8e-05,
+      "loss": 1.5761,
+      "step": 5572
+    },
+    {
+      "epoch": 0.31064659977703457,
+      "grad_norm": 0.48438408970832825,
+      "learning_rate": 8e-05,
+      "loss": 1.6204,
+      "step": 5573
+    },
+    {
+      "epoch": 0.31070234113712375,
+      "grad_norm": 0.5019887089729309,
+      "learning_rate": 8e-05,
+      "loss": 1.5752,
+      "step": 5574
+    },
+    {
+      "epoch": 0.31075808249721293,
+      "grad_norm": 0.49323540925979614,
+      "learning_rate": 8e-05,
+      "loss": 1.5624,
+      "step": 5575
+    },
+    {
+      "epoch": 0.3108138238573021,
+      "grad_norm": 0.4469727575778961,
+      "learning_rate": 8e-05,
+      "loss": 1.3725,
+      "step": 5576
+    },
+    {
+      "epoch": 0.3108695652173913,
+      "grad_norm": 0.6091485619544983,
+      "learning_rate": 8e-05,
+      "loss": 1.6382,
+      "step": 5577
+    },
+    {
+      "epoch": 0.31092530657748046,
+      "grad_norm": 0.5549405813217163,
+      "learning_rate": 8e-05,
+      "loss": 1.6422,
+      "step": 5578
+    },
+    {
+      "epoch": 0.3109810479375697,
+      "grad_norm": 0.5730343461036682,
+      "learning_rate": 8e-05,
+      "loss": 1.7795,
+      "step": 5579
+    },
+    {
+      "epoch": 0.3110367892976589,
+      "grad_norm": 0.5435800552368164,
+      "learning_rate": 8e-05,
+      "loss": 1.7752,
+      "step": 5580
+    },
+    {
+      "epoch": 0.31109253065774806,
+      "grad_norm": 0.6630704998970032,
+      "learning_rate": 8e-05,
+      "loss": 2.1515,
+      "step": 5581
+    },
+    {
+      "epoch": 0.31114827201783724,
+      "grad_norm": 0.588741660118103,
+      "learning_rate": 8e-05,
+      "loss": 1.8684,
+      "step": 5582
+    },
+    {
+      "epoch": 0.3112040133779264,
+      "grad_norm": 0.5837110877037048,
+      "learning_rate": 8e-05,
+      "loss": 1.9009,
+      "step": 5583
+    },
+    {
+      "epoch": 0.3112597547380156,
+      "grad_norm": 0.5275765657424927,
+      "learning_rate": 8e-05,
+      "loss": 1.7688,
+      "step": 5584
+    },
+    {
+      "epoch": 0.31131549609810477,
+      "grad_norm": 0.5006588697433472,
+      "learning_rate": 8e-05,
+      "loss": 1.6435,
+      "step": 5585
+    },
+    {
+      "epoch": 0.311371237458194,
+      "grad_norm": 0.5062305331230164,
+      "learning_rate": 8e-05,
+      "loss": 1.5436,
+      "step": 5586
+    },
+    {
+      "epoch": 0.3114269788182832,
+      "grad_norm": 0.5304542779922485,
+      "learning_rate": 8e-05,
+      "loss": 1.8129,
+      "step": 5587
+    },
+    {
+      "epoch": 0.31148272017837236,
+      "grad_norm": 0.5773730874061584,
+      "learning_rate": 8e-05,
+      "loss": 2.1254,
+      "step": 5588
+    },
+    {
+      "epoch": 0.31153846153846154,
+      "grad_norm": 0.4913802742958069,
+      "learning_rate": 8e-05,
+      "loss": 1.7781,
+      "step": 5589
+    },
+    {
+      "epoch": 0.3115942028985507,
+      "grad_norm": 0.5441120266914368,
+      "learning_rate": 8e-05,
+      "loss": 1.878,
+      "step": 5590
+    },
+    {
+      "epoch": 0.3116499442586399,
+      "grad_norm": 0.5085309147834778,
+      "learning_rate": 8e-05,
+      "loss": 1.7295,
+      "step": 5591
+    },
+    {
+      "epoch": 0.3117056856187291,
+      "grad_norm": 0.49739134311676025,
+      "learning_rate": 8e-05,
+      "loss": 1.6436,
+      "step": 5592
+    },
+    {
+      "epoch": 0.31176142697881826,
+      "grad_norm": 0.5027386546134949,
+      "learning_rate": 8e-05,
+      "loss": 1.6793,
+      "step": 5593
+    },
+    {
+      "epoch": 0.3118171683389075,
+      "grad_norm": 0.4946807026863098,
+      "learning_rate": 8e-05,
+      "loss": 1.7849,
+      "step": 5594
+    },
+    {
+      "epoch": 0.31187290969899667,
+      "grad_norm": 0.5363293290138245,
+      "learning_rate": 8e-05,
+      "loss": 1.568,
+      "step": 5595
+    },
+    {
+      "epoch": 0.31192865105908585,
+      "grad_norm": 0.501739501953125,
+      "learning_rate": 8e-05,
+      "loss": 1.7039,
+      "step": 5596
+    },
+    {
+      "epoch": 0.311984392419175,
+      "grad_norm": 0.5119023323059082,
+      "learning_rate": 8e-05,
+      "loss": 1.5821,
+      "step": 5597
+    },
+    {
+      "epoch": 0.3120401337792642,
+      "grad_norm": 0.5860930681228638,
+      "learning_rate": 8e-05,
+      "loss": 1.8778,
+      "step": 5598
+    },
+    {
+      "epoch": 0.3120958751393534,
+      "grad_norm": 0.5110373497009277,
+      "learning_rate": 8e-05,
+      "loss": 1.6558,
+      "step": 5599
+    },
+    {
+      "epoch": 0.31215161649944256,
+      "grad_norm": 0.45514994859695435,
+      "learning_rate": 8e-05,
+      "loss": 1.5136,
+      "step": 5600
+    },
+    {
+      "epoch": 0.3122073578595318,
+      "grad_norm": 0.5342680215835571,
+      "learning_rate": 8e-05,
+      "loss": 1.6569,
+      "step": 5601
+    },
+    {
+      "epoch": 0.312263099219621,
+      "grad_norm": 0.5177780389785767,
+      "learning_rate": 8e-05,
+      "loss": 1.7188,
+      "step": 5602
+    },
+    {
+      "epoch": 0.31231884057971016,
+      "grad_norm": 0.5124519467353821,
+      "learning_rate": 8e-05,
+      "loss": 1.6296,
+      "step": 5603
+    },
+    {
+      "epoch": 0.31237458193979933,
+      "grad_norm": 0.49985089898109436,
+      "learning_rate": 8e-05,
+      "loss": 1.5149,
+      "step": 5604
+    },
+    {
+      "epoch": 0.3124303232998885,
+      "grad_norm": 0.4940367639064789,
+      "learning_rate": 8e-05,
+      "loss": 1.6693,
+      "step": 5605
+    },
+    {
+      "epoch": 0.3124860646599777,
+      "grad_norm": 0.4849971532821655,
+      "learning_rate": 8e-05,
+      "loss": 1.4553,
+      "step": 5606
+    },
+    {
+      "epoch": 0.31254180602006687,
+      "grad_norm": 0.47547417879104614,
+      "learning_rate": 8e-05,
+      "loss": 1.8421,
+      "step": 5607
+    },
+    {
+      "epoch": 0.31259754738015605,
+      "grad_norm": 0.5307517647743225,
+      "learning_rate": 8e-05,
+      "loss": 1.7207,
+      "step": 5608
+    },
+    {
+      "epoch": 0.3126532887402453,
+      "grad_norm": 0.493900328874588,
+      "learning_rate": 8e-05,
+      "loss": 1.7417,
+      "step": 5609
+    },
+    {
+      "epoch": 0.31270903010033446,
+      "grad_norm": 0.4873904883861542,
+      "learning_rate": 8e-05,
+      "loss": 1.4415,
+      "step": 5610
+    },
+    {
+      "epoch": 0.31276477146042364,
+      "grad_norm": 0.5056470036506653,
+      "learning_rate": 8e-05,
+      "loss": 1.6919,
+      "step": 5611
+    },
+    {
+      "epoch": 0.3128205128205128,
+      "grad_norm": 0.47142380475997925,
+      "learning_rate": 8e-05,
+      "loss": 1.6281,
+      "step": 5612
+    },
+    {
+      "epoch": 0.312876254180602,
+      "grad_norm": 0.5202730894088745,
+      "learning_rate": 8e-05,
+      "loss": 1.9738,
+      "step": 5613
+    },
+    {
+      "epoch": 0.3129319955406912,
+      "grad_norm": 0.5567691326141357,
+      "learning_rate": 8e-05,
+      "loss": 1.6902,
+      "step": 5614
+    },
+    {
+      "epoch": 0.31298773690078036,
+      "grad_norm": 0.46796953678131104,
+      "learning_rate": 8e-05,
+      "loss": 1.5595,
+      "step": 5615
+    },
+    {
+      "epoch": 0.3130434782608696,
+      "grad_norm": 0.5604295134544373,
+      "learning_rate": 8e-05,
+      "loss": 2.0457,
+      "step": 5616
+    },
+    {
+      "epoch": 0.31309921962095877,
+      "grad_norm": 0.5760122537612915,
+      "learning_rate": 8e-05,
+      "loss": 2.0031,
+      "step": 5617
+    },
+    {
+      "epoch": 0.31315496098104795,
+      "grad_norm": 0.5523648858070374,
+      "learning_rate": 8e-05,
+      "loss": 1.7164,
+      "step": 5618
+    },
+    {
+      "epoch": 0.3132107023411371,
+      "grad_norm": 0.5255864262580872,
+      "learning_rate": 8e-05,
+      "loss": 1.6715,
+      "step": 5619
+    },
+    {
+      "epoch": 0.3132664437012263,
+      "grad_norm": 0.5574448704719543,
+      "learning_rate": 8e-05,
+      "loss": 1.6039,
+      "step": 5620
+    },
+    {
+      "epoch": 0.3133221850613155,
+      "grad_norm": 0.4774945378303528,
+      "learning_rate": 8e-05,
+      "loss": 1.8164,
+      "step": 5621
+    },
+    {
+      "epoch": 0.31337792642140466,
+      "grad_norm": 0.5023643970489502,
+      "learning_rate": 8e-05,
+      "loss": 1.5432,
+      "step": 5622
+    },
+    {
+      "epoch": 0.31343366778149384,
+      "grad_norm": 0.5021160244941711,
+      "learning_rate": 8e-05,
+      "loss": 1.8182,
+      "step": 5623
+    },
+    {
+      "epoch": 0.3134894091415831,
+      "grad_norm": 0.5544974207878113,
+      "learning_rate": 8e-05,
+      "loss": 1.6318,
+      "step": 5624
+    },
+    {
+      "epoch": 0.31354515050167225,
+      "grad_norm": 0.513565719127655,
+      "learning_rate": 8e-05,
+      "loss": 1.6485,
+      "step": 5625
+    },
+    {
+      "epoch": 0.31360089186176143,
+      "grad_norm": 0.5158968567848206,
+      "learning_rate": 8e-05,
+      "loss": 1.546,
+      "step": 5626
+    },
+    {
+      "epoch": 0.3136566332218506,
+      "grad_norm": 0.51337730884552,
+      "learning_rate": 8e-05,
+      "loss": 1.8192,
+      "step": 5627
+    },
+    {
+      "epoch": 0.3137123745819398,
+      "grad_norm": 0.5285170078277588,
+      "learning_rate": 8e-05,
+      "loss": 1.6804,
+      "step": 5628
+    },
+    {
+      "epoch": 0.31376811594202897,
+      "grad_norm": 0.5417566299438477,
+      "learning_rate": 8e-05,
+      "loss": 1.6337,
+      "step": 5629
+    },
+    {
+      "epoch": 0.31382385730211815,
+      "grad_norm": 0.5149359703063965,
+      "learning_rate": 8e-05,
+      "loss": 1.688,
+      "step": 5630
+    },
+    {
+      "epoch": 0.3138795986622074,
+      "grad_norm": 0.5370135307312012,
+      "learning_rate": 8e-05,
+      "loss": 1.7254,
+      "step": 5631
+    },
+    {
+      "epoch": 0.31393534002229656,
+      "grad_norm": 0.49705618619918823,
+      "learning_rate": 8e-05,
+      "loss": 1.6279,
+      "step": 5632
+    },
+    {
+      "epoch": 0.31399108138238574,
+      "grad_norm": 0.536984384059906,
+      "learning_rate": 8e-05,
+      "loss": 1.604,
+      "step": 5633
+    },
+    {
+      "epoch": 0.3140468227424749,
+      "grad_norm": 0.5114649534225464,
+      "learning_rate": 8e-05,
+      "loss": 1.8026,
+      "step": 5634
+    },
+    {
+      "epoch": 0.3141025641025641,
+      "grad_norm": 0.5303798317909241,
+      "learning_rate": 8e-05,
+      "loss": 1.5936,
+      "step": 5635
+    },
+    {
+      "epoch": 0.3141583054626533,
+      "grad_norm": 0.5324503183364868,
+      "learning_rate": 8e-05,
+      "loss": 1.7611,
+      "step": 5636
+    },
+    {
+      "epoch": 0.31421404682274245,
+      "grad_norm": 0.4875989258289337,
+      "learning_rate": 8e-05,
+      "loss": 1.5046,
+      "step": 5637
+    },
+    {
+      "epoch": 0.3142697881828317,
+      "grad_norm": 0.48831331729888916,
+      "learning_rate": 8e-05,
+      "loss": 1.6284,
+      "step": 5638
+    },
+    {
+      "epoch": 0.31432552954292087,
+      "grad_norm": 0.5204409956932068,
+      "learning_rate": 8e-05,
+      "loss": 1.7106,
+      "step": 5639
+    },
+    {
+      "epoch": 0.31438127090301005,
+      "grad_norm": 0.516324520111084,
+      "learning_rate": 8e-05,
+      "loss": 1.7483,
+      "step": 5640
+    },
+    {
+      "epoch": 0.3144370122630992,
+      "grad_norm": 0.5173261165618896,
+      "learning_rate": 8e-05,
+      "loss": 1.6939,
+      "step": 5641
+    },
+    {
+      "epoch": 0.3144927536231884,
+      "grad_norm": 0.5068714618682861,
+      "learning_rate": 8e-05,
+      "loss": 1.7087,
+      "step": 5642
+    },
+    {
+      "epoch": 0.3145484949832776,
+      "grad_norm": 0.4229615032672882,
+      "learning_rate": 8e-05,
+      "loss": 1.4433,
+      "step": 5643
+    },
+    {
+      "epoch": 0.31460423634336676,
+      "grad_norm": 0.5388889908790588,
+      "learning_rate": 8e-05,
+      "loss": 1.9121,
+      "step": 5644
+    },
+    {
+      "epoch": 0.31465997770345594,
+      "grad_norm": 0.4938584566116333,
+      "learning_rate": 8e-05,
+      "loss": 1.6384,
+      "step": 5645
+    },
+    {
+      "epoch": 0.3147157190635452,
+      "grad_norm": 0.5267650485038757,
+      "learning_rate": 8e-05,
+      "loss": 1.9155,
+      "step": 5646
+    },
+    {
+      "epoch": 0.31477146042363435,
+      "grad_norm": 0.5120892524719238,
+      "learning_rate": 8e-05,
+      "loss": 1.5734,
+      "step": 5647
+    },
+    {
+      "epoch": 0.31482720178372353,
+      "grad_norm": 0.5487409830093384,
+      "learning_rate": 8e-05,
+      "loss": 1.4687,
+      "step": 5648
+    },
+    {
+      "epoch": 0.3148829431438127,
+      "grad_norm": 0.5335332751274109,
+      "learning_rate": 8e-05,
+      "loss": 1.9635,
+      "step": 5649
+    },
+    {
+      "epoch": 0.3149386845039019,
+      "grad_norm": 0.4990962743759155,
+      "learning_rate": 8e-05,
+      "loss": 1.5531,
+      "step": 5650
+    },
+    {
+      "epoch": 0.31499442586399107,
+      "grad_norm": 0.5237401127815247,
+      "learning_rate": 8e-05,
+      "loss": 1.8537,
+      "step": 5651
+    },
+    {
+      "epoch": 0.31505016722408025,
+      "grad_norm": 0.5074672698974609,
+      "learning_rate": 8e-05,
+      "loss": 1.6768,
+      "step": 5652
+    },
+    {
+      "epoch": 0.3151059085841695,
+      "grad_norm": 0.5036709308624268,
+      "learning_rate": 8e-05,
+      "loss": 1.6335,
+      "step": 5653
+    },
+    {
+      "epoch": 0.31516164994425866,
+      "grad_norm": 0.5381163358688354,
+      "learning_rate": 8e-05,
+      "loss": 1.701,
+      "step": 5654
+    },
+    {
+      "epoch": 0.31521739130434784,
+      "grad_norm": 0.5201897025108337,
+      "learning_rate": 8e-05,
+      "loss": 1.702,
+      "step": 5655
+    },
+    {
+      "epoch": 0.315273132664437,
+      "grad_norm": 0.5196989178657532,
+      "learning_rate": 8e-05,
+      "loss": 1.5655,
+      "step": 5656
+    },
+    {
+      "epoch": 0.3153288740245262,
+      "grad_norm": 0.5282043218612671,
+      "learning_rate": 8e-05,
+      "loss": 1.8807,
+      "step": 5657
+    },
+    {
+      "epoch": 0.3153846153846154,
+      "grad_norm": 0.5148075819015503,
+      "learning_rate": 8e-05,
+      "loss": 1.6046,
+      "step": 5658
+    },
+    {
+      "epoch": 0.31544035674470455,
+      "grad_norm": 0.5050842761993408,
+      "learning_rate": 8e-05,
+      "loss": 1.6702,
+      "step": 5659
+    },
+    {
+      "epoch": 0.31549609810479373,
+      "grad_norm": 0.5463430285453796,
+      "learning_rate": 8e-05,
+      "loss": 1.9318,
+      "step": 5660
+    },
+    {
+      "epoch": 0.31555183946488297,
+      "grad_norm": 0.5092810392379761,
+      "learning_rate": 8e-05,
+      "loss": 1.7894,
+      "step": 5661
+    },
+    {
+      "epoch": 0.31560758082497214,
+      "grad_norm": 0.5707821249961853,
+      "learning_rate": 8e-05,
+      "loss": 1.6085,
+      "step": 5662
+    },
+    {
+      "epoch": 0.3156633221850613,
+      "grad_norm": 0.5087369680404663,
+      "learning_rate": 8e-05,
+      "loss": 1.6965,
+      "step": 5663
+    },
+    {
+      "epoch": 0.3157190635451505,
+      "grad_norm": 0.5232574343681335,
+      "learning_rate": 8e-05,
+      "loss": 1.8687,
+      "step": 5664
+    },
+    {
+      "epoch": 0.3157748049052397,
+      "grad_norm": 0.5404393076896667,
+      "learning_rate": 8e-05,
+      "loss": 1.9312,
+      "step": 5665
+    },
+    {
+      "epoch": 0.31583054626532886,
+      "grad_norm": 0.4950511157512665,
+      "learning_rate": 8e-05,
+      "loss": 1.6128,
+      "step": 5666
+    },
+    {
+      "epoch": 0.31588628762541804,
+      "grad_norm": 0.4811279773712158,
+      "learning_rate": 8e-05,
+      "loss": 1.7009,
+      "step": 5667
+    },
+    {
+      "epoch": 0.3159420289855073,
+      "grad_norm": 0.48683613538742065,
+      "learning_rate": 8e-05,
+      "loss": 1.5999,
+      "step": 5668
+    },
+    {
+      "epoch": 0.31599777034559645,
+      "grad_norm": 0.4522567391395569,
+      "learning_rate": 8e-05,
+      "loss": 1.4242,
+      "step": 5669
+    },
+    {
+      "epoch": 0.31605351170568563,
+      "grad_norm": 0.5745465159416199,
+      "learning_rate": 8e-05,
+      "loss": 1.3288,
+      "step": 5670
+    },
+    {
+      "epoch": 0.3161092530657748,
+      "grad_norm": 0.577358067035675,
+      "learning_rate": 8e-05,
+      "loss": 1.9754,
+      "step": 5671
+    },
+    {
+      "epoch": 0.316164994425864,
+      "grad_norm": 0.5135446190834045,
+      "learning_rate": 8e-05,
+      "loss": 1.8025,
+      "step": 5672
+    },
+    {
+      "epoch": 0.31622073578595317,
+      "grad_norm": 0.4923860728740692,
+      "learning_rate": 8e-05,
+      "loss": 1.6862,
+      "step": 5673
+    },
+    {
+      "epoch": 0.31627647714604235,
+      "grad_norm": 0.5172598958015442,
+      "learning_rate": 8e-05,
+      "loss": 1.7144,
+      "step": 5674
+    },
+    {
+      "epoch": 0.3163322185061315,
+      "grad_norm": 0.5160543918609619,
+      "learning_rate": 8e-05,
+      "loss": 1.6935,
+      "step": 5675
+    },
+    {
+      "epoch": 0.31638795986622076,
+      "grad_norm": 0.44993856549263,
+      "learning_rate": 8e-05,
+      "loss": 1.4565,
+      "step": 5676
+    },
+    {
+      "epoch": 0.31644370122630994,
+      "grad_norm": 0.5209502577781677,
+      "learning_rate": 8e-05,
+      "loss": 1.6713,
+      "step": 5677
+    },
+    {
+      "epoch": 0.3164994425863991,
+      "grad_norm": 0.5097101330757141,
+      "learning_rate": 8e-05,
+      "loss": 1.8201,
+      "step": 5678
+    },
+    {
+      "epoch": 0.3165551839464883,
+      "grad_norm": 0.51374351978302,
+      "learning_rate": 8e-05,
+      "loss": 1.7345,
+      "step": 5679
+    },
+    {
+      "epoch": 0.3166109253065775,
+      "grad_norm": 0.6368524432182312,
+      "learning_rate": 8e-05,
+      "loss": 2.1954,
+      "step": 5680
+    },
+    {
+      "epoch": 0.31666666666666665,
+      "grad_norm": 0.5003951191902161,
+      "learning_rate": 8e-05,
+      "loss": 1.5184,
+      "step": 5681
+    },
+    {
+      "epoch": 0.31672240802675583,
+      "grad_norm": 0.5287581086158752,
+      "learning_rate": 8e-05,
+      "loss": 1.8125,
+      "step": 5682
+    },
+    {
+      "epoch": 0.31677814938684506,
+      "grad_norm": 0.5246060490608215,
+      "learning_rate": 8e-05,
+      "loss": 1.7134,
+      "step": 5683
+    },
+    {
+      "epoch": 0.31683389074693424,
+      "grad_norm": 0.5368450284004211,
+      "learning_rate": 8e-05,
+      "loss": 1.9415,
+      "step": 5684
+    },
+    {
+      "epoch": 0.3168896321070234,
+      "grad_norm": 0.5108862519264221,
+      "learning_rate": 8e-05,
+      "loss": 1.7992,
+      "step": 5685
+    },
+    {
+      "epoch": 0.3169453734671126,
+      "grad_norm": 0.5072665214538574,
+      "learning_rate": 8e-05,
+      "loss": 1.6403,
+      "step": 5686
+    },
+    {
+      "epoch": 0.3170011148272018,
+      "grad_norm": 0.5059048533439636,
+      "learning_rate": 8e-05,
+      "loss": 1.6965,
+      "step": 5687
+    },
+    {
+      "epoch": 0.31705685618729096,
+      "grad_norm": 0.4943862557411194,
+      "learning_rate": 8e-05,
+      "loss": 1.7319,
+      "step": 5688
+    },
+    {
+      "epoch": 0.31711259754738014,
+      "grad_norm": 0.5123859643936157,
+      "learning_rate": 8e-05,
+      "loss": 1.715,
+      "step": 5689
+    },
+    {
+      "epoch": 0.3171683389074693,
+      "grad_norm": 0.5278445482254028,
+      "learning_rate": 8e-05,
+      "loss": 1.7291,
+      "step": 5690
+    },
+    {
+      "epoch": 0.31722408026755855,
+      "grad_norm": 0.4850660562515259,
+      "learning_rate": 8e-05,
+      "loss": 1.6429,
+      "step": 5691
+    },
+    {
+      "epoch": 0.31727982162764773,
+      "grad_norm": 0.4988071322441101,
+      "learning_rate": 8e-05,
+      "loss": 1.7774,
+      "step": 5692
+    },
+    {
+      "epoch": 0.3173355629877369,
+      "grad_norm": 0.5240743160247803,
+      "learning_rate": 8e-05,
+      "loss": 1.8149,
+      "step": 5693
+    },
+    {
+      "epoch": 0.3173913043478261,
+      "grad_norm": 0.5346617698669434,
+      "learning_rate": 8e-05,
+      "loss": 1.9376,
+      "step": 5694
+    },
+    {
+      "epoch": 0.31744704570791527,
+      "grad_norm": 0.4753016531467438,
+      "learning_rate": 8e-05,
+      "loss": 1.5924,
+      "step": 5695
+    },
+    {
+      "epoch": 0.31750278706800444,
+      "grad_norm": 0.48297837376594543,
+      "learning_rate": 8e-05,
+      "loss": 1.5537,
+      "step": 5696
+    },
+    {
+      "epoch": 0.3175585284280936,
+      "grad_norm": 0.5097260475158691,
+      "learning_rate": 8e-05,
+      "loss": 1.7266,
+      "step": 5697
+    },
+    {
+      "epoch": 0.31761426978818286,
+      "grad_norm": 0.4937364459037781,
+      "learning_rate": 8e-05,
+      "loss": 1.6826,
+      "step": 5698
+    },
+    {
+      "epoch": 0.31767001114827204,
+      "grad_norm": 0.4757471978664398,
+      "learning_rate": 8e-05,
+      "loss": 1.6335,
+      "step": 5699
+    },
+    {
+      "epoch": 0.3177257525083612,
+      "grad_norm": 0.5071625709533691,
+      "learning_rate": 8e-05,
+      "loss": 1.6613,
+      "step": 5700
+    },
+    {
+      "epoch": 0.3177814938684504,
+      "grad_norm": 0.5106151700019836,
+      "learning_rate": 8e-05,
+      "loss": 1.6416,
+      "step": 5701
+    },
+    {
+      "epoch": 0.31783723522853957,
+      "grad_norm": 0.5215732455253601,
+      "learning_rate": 8e-05,
+      "loss": 1.5861,
+      "step": 5702
+    },
+    {
+      "epoch": 0.31789297658862875,
+      "grad_norm": 0.5385761260986328,
+      "learning_rate": 8e-05,
+      "loss": 1.7751,
+      "step": 5703
+    },
+    {
+      "epoch": 0.31794871794871793,
+      "grad_norm": 0.5467097163200378,
+      "learning_rate": 8e-05,
+      "loss": 1.6679,
+      "step": 5704
+    },
+    {
+      "epoch": 0.3180044593088071,
+      "grad_norm": 0.5180816650390625,
+      "learning_rate": 8e-05,
+      "loss": 1.6077,
+      "step": 5705
+    },
+    {
+      "epoch": 0.31806020066889634,
+      "grad_norm": 0.5191653966903687,
+      "learning_rate": 8e-05,
+      "loss": 1.6352,
+      "step": 5706
+    },
+    {
+      "epoch": 0.3181159420289855,
+      "grad_norm": 0.4900323152542114,
+      "learning_rate": 8e-05,
+      "loss": 1.6899,
+      "step": 5707
+    },
+    {
+      "epoch": 0.3181716833890747,
+      "grad_norm": 0.5084742307662964,
+      "learning_rate": 8e-05,
+      "loss": 1.5656,
+      "step": 5708
+    },
+    {
+      "epoch": 0.3182274247491639,
+      "grad_norm": 0.5018293857574463,
+      "learning_rate": 8e-05,
+      "loss": 1.5945,
+      "step": 5709
+    },
+    {
+      "epoch": 0.31828316610925306,
+      "grad_norm": 0.5255138278007507,
+      "learning_rate": 8e-05,
+      "loss": 1.8835,
+      "step": 5710
+    },
+    {
+      "epoch": 0.31833890746934224,
+      "grad_norm": 0.5178418755531311,
+      "learning_rate": 8e-05,
+      "loss": 1.6581,
+      "step": 5711
+    },
+    {
+      "epoch": 0.3183946488294314,
+      "grad_norm": 0.496112197637558,
+      "learning_rate": 8e-05,
+      "loss": 1.605,
+      "step": 5712
+    },
+    {
+      "epoch": 0.31845039018952065,
+      "grad_norm": 0.519536018371582,
+      "learning_rate": 8e-05,
+      "loss": 1.8334,
+      "step": 5713
+    },
+    {
+      "epoch": 0.31850613154960983,
+      "grad_norm": 0.47761979699134827,
+      "learning_rate": 8e-05,
+      "loss": 1.6399,
+      "step": 5714
+    },
+    {
+      "epoch": 0.318561872909699,
+      "grad_norm": 0.5270336866378784,
+      "learning_rate": 8e-05,
+      "loss": 1.6663,
+      "step": 5715
+    },
+    {
+      "epoch": 0.3186176142697882,
+      "grad_norm": 0.5025304555892944,
+      "learning_rate": 8e-05,
+      "loss": 1.649,
+      "step": 5716
+    },
+    {
+      "epoch": 0.31867335562987736,
+      "grad_norm": 0.5432649850845337,
+      "learning_rate": 8e-05,
+      "loss": 1.895,
+      "step": 5717
+    },
+    {
+      "epoch": 0.31872909698996654,
+      "grad_norm": 0.553049623966217,
+      "learning_rate": 8e-05,
+      "loss": 1.8611,
+      "step": 5718
+    },
+    {
+      "epoch": 0.3187848383500557,
+      "grad_norm": 0.4806937873363495,
+      "learning_rate": 8e-05,
+      "loss": 1.6396,
+      "step": 5719
+    },
+    {
+      "epoch": 0.3188405797101449,
+      "grad_norm": 0.478455126285553,
+      "learning_rate": 8e-05,
+      "loss": 1.5705,
+      "step": 5720
+    },
+    {
+      "epoch": 0.31889632107023413,
+      "grad_norm": 0.5210627913475037,
+      "learning_rate": 8e-05,
+      "loss": 1.4042,
+      "step": 5721
+    },
+    {
+      "epoch": 0.3189520624303233,
+      "grad_norm": 0.5083332657814026,
+      "learning_rate": 8e-05,
+      "loss": 1.671,
+      "step": 5722
+    },
+    {
+      "epoch": 0.3190078037904125,
+      "grad_norm": 0.49854928255081177,
+      "learning_rate": 8e-05,
+      "loss": 1.8224,
+      "step": 5723
+    },
+    {
+      "epoch": 0.31906354515050167,
+      "grad_norm": 0.5092363953590393,
+      "learning_rate": 8e-05,
+      "loss": 1.8449,
+      "step": 5724
+    },
+    {
+      "epoch": 0.31911928651059085,
+      "grad_norm": 0.5526291728019714,
+      "learning_rate": 8e-05,
+      "loss": 1.6813,
+      "step": 5725
+    },
+    {
+      "epoch": 0.31917502787068003,
+      "grad_norm": 0.4988034665584564,
+      "learning_rate": 8e-05,
+      "loss": 1.6409,
+      "step": 5726
+    },
+    {
+      "epoch": 0.3192307692307692,
+      "grad_norm": 0.49596595764160156,
+      "learning_rate": 8e-05,
+      "loss": 1.6347,
+      "step": 5727
+    },
+    {
+      "epoch": 0.31928651059085844,
+      "grad_norm": 0.5038948059082031,
+      "learning_rate": 8e-05,
+      "loss": 1.7396,
+      "step": 5728
+    },
+    {
+      "epoch": 0.3193422519509476,
+      "grad_norm": 0.5159317851066589,
+      "learning_rate": 8e-05,
+      "loss": 1.5135,
+      "step": 5729
+    },
+    {
+      "epoch": 0.3193979933110368,
+      "grad_norm": 0.536002516746521,
+      "learning_rate": 8e-05,
+      "loss": 1.6583,
+      "step": 5730
+    },
+    {
+      "epoch": 0.319453734671126,
+      "grad_norm": 0.5048378705978394,
+      "learning_rate": 8e-05,
+      "loss": 1.7345,
+      "step": 5731
+    },
+    {
+      "epoch": 0.31950947603121516,
+      "grad_norm": 0.4891044497489929,
+      "learning_rate": 8e-05,
+      "loss": 1.6499,
+      "step": 5732
+    },
+    {
+      "epoch": 0.31956521739130433,
+      "grad_norm": 0.5046207308769226,
+      "learning_rate": 8e-05,
+      "loss": 1.6562,
+      "step": 5733
+    },
+    {
+      "epoch": 0.3196209587513935,
+      "grad_norm": 0.4974361062049866,
+      "learning_rate": 8e-05,
+      "loss": 1.6007,
+      "step": 5734
+    },
+    {
+      "epoch": 0.31967670011148275,
+      "grad_norm": 0.506041407585144,
+      "learning_rate": 8e-05,
+      "loss": 1.5283,
+      "step": 5735
+    },
+    {
+      "epoch": 0.3197324414715719,
+      "grad_norm": 0.5237359404563904,
+      "learning_rate": 8e-05,
+      "loss": 1.7842,
+      "step": 5736
+    },
+    {
+      "epoch": 0.3197881828316611,
+      "grad_norm": 0.5203297138214111,
+      "learning_rate": 8e-05,
+      "loss": 1.787,
+      "step": 5737
+    },
+    {
+      "epoch": 0.3198439241917503,
+      "grad_norm": 0.5403977036476135,
+      "learning_rate": 8e-05,
+      "loss": 1.8318,
+      "step": 5738
+    },
+    {
+      "epoch": 0.31989966555183946,
+      "grad_norm": 0.5268109440803528,
+      "learning_rate": 8e-05,
+      "loss": 1.8816,
+      "step": 5739
+    },
+    {
+      "epoch": 0.31995540691192864,
+      "grad_norm": 0.5357729196548462,
+      "learning_rate": 8e-05,
+      "loss": 1.689,
+      "step": 5740
+    },
+    {
+      "epoch": 0.3200111482720178,
+      "grad_norm": 0.5156355500221252,
+      "learning_rate": 8e-05,
+      "loss": 1.6791,
+      "step": 5741
+    },
+    {
+      "epoch": 0.320066889632107,
+      "grad_norm": 0.493498831987381,
+      "learning_rate": 8e-05,
+      "loss": 1.6286,
+      "step": 5742
+    },
+    {
+      "epoch": 0.32012263099219623,
+      "grad_norm": 0.5426554083824158,
+      "learning_rate": 8e-05,
+      "loss": 1.9256,
+      "step": 5743
+    },
+    {
+      "epoch": 0.3201783723522854,
+      "grad_norm": 0.5164318680763245,
+      "learning_rate": 8e-05,
+      "loss": 1.6362,
+      "step": 5744
+    },
+    {
+      "epoch": 0.3202341137123746,
+      "grad_norm": 0.5163710713386536,
+      "learning_rate": 8e-05,
+      "loss": 1.4276,
+      "step": 5745
+    },
+    {
+      "epoch": 0.32028985507246377,
+      "grad_norm": 0.529309093952179,
+      "learning_rate": 8e-05,
+      "loss": 1.7972,
+      "step": 5746
+    },
+    {
+      "epoch": 0.32034559643255295,
+      "grad_norm": 0.4801371097564697,
+      "learning_rate": 8e-05,
+      "loss": 1.6195,
+      "step": 5747
+    },
+    {
+      "epoch": 0.3204013377926421,
+      "grad_norm": 0.5078192949295044,
+      "learning_rate": 8e-05,
+      "loss": 1.6717,
+      "step": 5748
+    },
+    {
+      "epoch": 0.3204570791527313,
+      "grad_norm": 0.5345617532730103,
+      "learning_rate": 8e-05,
+      "loss": 1.629,
+      "step": 5749
+    },
+    {
+      "epoch": 0.32051282051282054,
+      "grad_norm": 0.4875049591064453,
+      "learning_rate": 8e-05,
+      "loss": 1.6211,
+      "step": 5750
+    },
+    {
+      "epoch": 0.3205685618729097,
+      "grad_norm": 0.48048990964889526,
+      "learning_rate": 8e-05,
+      "loss": 1.6774,
+      "step": 5751
+    },
+    {
+      "epoch": 0.3206243032329989,
+      "grad_norm": 0.5032520890235901,
+      "learning_rate": 8e-05,
+      "loss": 1.8064,
+      "step": 5752
+    },
+    {
+      "epoch": 0.3206800445930881,
+      "grad_norm": 0.4913550615310669,
+      "learning_rate": 8e-05,
+      "loss": 1.6187,
+      "step": 5753
+    },
+    {
+      "epoch": 0.32073578595317725,
+      "grad_norm": 0.5281867980957031,
+      "learning_rate": 8e-05,
+      "loss": 1.8087,
+      "step": 5754
+    },
+    {
+      "epoch": 0.32079152731326643,
+      "grad_norm": 0.49618878960609436,
+      "learning_rate": 8e-05,
+      "loss": 1.6328,
+      "step": 5755
+    },
+    {
+      "epoch": 0.3208472686733556,
+      "grad_norm": 0.5247597694396973,
+      "learning_rate": 8e-05,
+      "loss": 1.8147,
+      "step": 5756
+    },
+    {
+      "epoch": 0.3209030100334448,
+      "grad_norm": 0.5163070559501648,
+      "learning_rate": 8e-05,
+      "loss": 1.6118,
+      "step": 5757
+    },
+    {
+      "epoch": 0.320958751393534,
+      "grad_norm": 0.49757692217826843,
+      "learning_rate": 8e-05,
+      "loss": 1.5848,
+      "step": 5758
+    },
+    {
+      "epoch": 0.3210144927536232,
+      "grad_norm": 0.5048686265945435,
+      "learning_rate": 8e-05,
+      "loss": 1.7403,
+      "step": 5759
+    },
+    {
+      "epoch": 0.3210702341137124,
+      "grad_norm": 0.5307775139808655,
+      "learning_rate": 8e-05,
+      "loss": 1.8107,
+      "step": 5760
+    },
+    {
+      "epoch": 0.32112597547380156,
+      "grad_norm": 0.4905966520309448,
+      "learning_rate": 8e-05,
+      "loss": 1.5565,
+      "step": 5761
+    },
+    {
+      "epoch": 0.32118171683389074,
+      "grad_norm": 0.5203377604484558,
+      "learning_rate": 8e-05,
+      "loss": 1.6875,
+      "step": 5762
+    },
+    {
+      "epoch": 0.3212374581939799,
+      "grad_norm": 0.48834821581840515,
+      "learning_rate": 8e-05,
+      "loss": 1.6393,
+      "step": 5763
+    },
+    {
+      "epoch": 0.3212931995540691,
+      "grad_norm": 0.5044954419136047,
+      "learning_rate": 8e-05,
+      "loss": 1.8203,
+      "step": 5764
+    },
+    {
+      "epoch": 0.32134894091415833,
+      "grad_norm": 0.5049839615821838,
+      "learning_rate": 8e-05,
+      "loss": 1.6837,
+      "step": 5765
+    },
+    {
+      "epoch": 0.3214046822742475,
+      "grad_norm": 0.4517059922218323,
+      "learning_rate": 8e-05,
+      "loss": 1.5599,
+      "step": 5766
+    },
+    {
+      "epoch": 0.3214604236343367,
+      "grad_norm": 0.48859703540802,
+      "learning_rate": 8e-05,
+      "loss": 1.4708,
+      "step": 5767
+    },
+    {
+      "epoch": 0.32151616499442587,
+      "grad_norm": 0.5075569152832031,
+      "learning_rate": 8e-05,
+      "loss": 1.6787,
+      "step": 5768
+    },
+    {
+      "epoch": 0.32157190635451505,
+      "grad_norm": 0.4926925599575043,
+      "learning_rate": 8e-05,
+      "loss": 1.6868,
+      "step": 5769
+    },
+    {
+      "epoch": 0.3216276477146042,
+      "grad_norm": 0.5494069457054138,
+      "learning_rate": 8e-05,
+      "loss": 1.6931,
+      "step": 5770
+    },
+    {
+      "epoch": 0.3216833890746934,
+      "grad_norm": 0.5246031284332275,
+      "learning_rate": 8e-05,
+      "loss": 1.6851,
+      "step": 5771
+    },
+    {
+      "epoch": 0.3217391304347826,
+      "grad_norm": 0.5542581677436829,
+      "learning_rate": 8e-05,
+      "loss": 1.8545,
+      "step": 5772
+    },
+    {
+      "epoch": 0.3217948717948718,
+      "grad_norm": 0.514141321182251,
+      "learning_rate": 8e-05,
+      "loss": 1.5756,
+      "step": 5773
+    },
+    {
+      "epoch": 0.321850613154961,
+      "grad_norm": 0.48406264185905457,
+      "learning_rate": 8e-05,
+      "loss": 1.703,
+      "step": 5774
+    },
+    {
+      "epoch": 0.3219063545150502,
+      "grad_norm": 0.5315650105476379,
+      "learning_rate": 8e-05,
+      "loss": 1.681,
+      "step": 5775
+    },
+    {
+      "epoch": 0.32196209587513935,
+      "grad_norm": 0.5040196776390076,
+      "learning_rate": 8e-05,
+      "loss": 1.8194,
+      "step": 5776
+    },
+    {
+      "epoch": 0.32201783723522853,
+      "grad_norm": 0.5190441608428955,
+      "learning_rate": 8e-05,
+      "loss": 1.5079,
+      "step": 5777
+    },
+    {
+      "epoch": 0.3220735785953177,
+      "grad_norm": 0.5137271285057068,
+      "learning_rate": 8e-05,
+      "loss": 1.7097,
+      "step": 5778
+    },
+    {
+      "epoch": 0.3221293199554069,
+      "grad_norm": 0.5241565108299255,
+      "learning_rate": 8e-05,
+      "loss": 1.7308,
+      "step": 5779
+    },
+    {
+      "epoch": 0.3221850613154961,
+      "grad_norm": 0.5239174365997314,
+      "learning_rate": 8e-05,
+      "loss": 1.7569,
+      "step": 5780
+    },
+    {
+      "epoch": 0.3222408026755853,
+      "grad_norm": 0.5030615925788879,
+      "learning_rate": 8e-05,
+      "loss": 1.6763,
+      "step": 5781
+    },
+    {
+      "epoch": 0.3222965440356745,
+      "grad_norm": 0.5678612589836121,
+      "learning_rate": 8e-05,
+      "loss": 1.9383,
+      "step": 5782
+    },
+    {
+      "epoch": 0.32235228539576366,
+      "grad_norm": 0.5458984375,
+      "learning_rate": 8e-05,
+      "loss": 1.7486,
+      "step": 5783
+    },
+    {
+      "epoch": 0.32240802675585284,
+      "grad_norm": 0.48285484313964844,
+      "learning_rate": 8e-05,
+      "loss": 1.6632,
+      "step": 5784
+    },
+    {
+      "epoch": 0.322463768115942,
+      "grad_norm": 0.5187762975692749,
+      "learning_rate": 8e-05,
+      "loss": 1.7322,
+      "step": 5785
+    },
+    {
+      "epoch": 0.3225195094760312,
+      "grad_norm": 0.5410240292549133,
+      "learning_rate": 8e-05,
+      "loss": 1.9998,
+      "step": 5786
+    },
+    {
+      "epoch": 0.3225752508361204,
+      "grad_norm": 0.4980333149433136,
+      "learning_rate": 8e-05,
+      "loss": 1.8457,
+      "step": 5787
+    },
+    {
+      "epoch": 0.3226309921962096,
+      "grad_norm": 0.475260853767395,
+      "learning_rate": 8e-05,
+      "loss": 1.6318,
+      "step": 5788
+    },
+    {
+      "epoch": 0.3226867335562988,
+      "grad_norm": 0.575297474861145,
+      "learning_rate": 8e-05,
+      "loss": 1.9413,
+      "step": 5789
+    },
+    {
+      "epoch": 0.32274247491638797,
+      "grad_norm": 0.5367628335952759,
+      "learning_rate": 8e-05,
+      "loss": 1.8158,
+      "step": 5790
+    },
+    {
+      "epoch": 0.32279821627647715,
+      "grad_norm": 0.5099639296531677,
+      "learning_rate": 8e-05,
+      "loss": 1.8162,
+      "step": 5791
+    },
+    {
+      "epoch": 0.3228539576365663,
+      "grad_norm": 0.5249901413917542,
+      "learning_rate": 8e-05,
+      "loss": 1.6299,
+      "step": 5792
+    },
+    {
+      "epoch": 0.3229096989966555,
+      "grad_norm": 0.5240976214408875,
+      "learning_rate": 8e-05,
+      "loss": 1.7886,
+      "step": 5793
+    },
+    {
+      "epoch": 0.3229654403567447,
+      "grad_norm": 0.49148687720298767,
+      "learning_rate": 8e-05,
+      "loss": 1.6332,
+      "step": 5794
+    },
+    {
+      "epoch": 0.3230211817168339,
+      "grad_norm": 0.5600979924201965,
+      "learning_rate": 8e-05,
+      "loss": 1.6854,
+      "step": 5795
+    },
+    {
+      "epoch": 0.3230769230769231,
+      "grad_norm": 0.4522090554237366,
+      "learning_rate": 8e-05,
+      "loss": 1.6729,
+      "step": 5796
+    },
+    {
+      "epoch": 0.3231326644370123,
+      "grad_norm": 0.5222184658050537,
+      "learning_rate": 8e-05,
+      "loss": 1.8499,
+      "step": 5797
+    },
+    {
+      "epoch": 0.32318840579710145,
+      "grad_norm": 0.5561509132385254,
+      "learning_rate": 8e-05,
+      "loss": 1.8946,
+      "step": 5798
+    },
+    {
+      "epoch": 0.32324414715719063,
+      "grad_norm": 0.49291518330574036,
+      "learning_rate": 8e-05,
+      "loss": 1.8201,
+      "step": 5799
+    },
+    {
+      "epoch": 0.3232998885172798,
+      "grad_norm": 0.4637990891933441,
+      "learning_rate": 8e-05,
+      "loss": 1.6406,
+      "step": 5800
+    },
+    {
+      "epoch": 0.323355629877369,
+      "grad_norm": 0.6826379895210266,
+      "learning_rate": 8e-05,
+      "loss": 2.0064,
+      "step": 5801
+    },
+    {
+      "epoch": 0.32341137123745817,
+      "grad_norm": 0.49902430176734924,
+      "learning_rate": 8e-05,
+      "loss": 1.5408,
+      "step": 5802
+    },
+    {
+      "epoch": 0.3234671125975474,
+      "grad_norm": 0.6668875813484192,
+      "learning_rate": 8e-05,
+      "loss": 2.1603,
+      "step": 5803
+    },
+    {
+      "epoch": 0.3235228539576366,
+      "grad_norm": 0.498045951128006,
+      "learning_rate": 8e-05,
+      "loss": 1.5568,
+      "step": 5804
+    },
+    {
+      "epoch": 0.32357859531772576,
+      "grad_norm": 0.5072228908538818,
+      "learning_rate": 8e-05,
+      "loss": 1.6002,
+      "step": 5805
+    },
+    {
+      "epoch": 0.32363433667781494,
+      "grad_norm": 0.5403440594673157,
+      "learning_rate": 8e-05,
+      "loss": 1.5875,
+      "step": 5806
+    },
+    {
+      "epoch": 0.3236900780379041,
+      "grad_norm": 0.47508928179740906,
+      "learning_rate": 8e-05,
+      "loss": 1.5958,
+      "step": 5807
+    },
+    {
+      "epoch": 0.3237458193979933,
+      "grad_norm": 0.5487821698188782,
+      "learning_rate": 8e-05,
+      "loss": 1.6753,
+      "step": 5808
+    },
+    {
+      "epoch": 0.3238015607580825,
+      "grad_norm": 0.4984150826931,
+      "learning_rate": 8e-05,
+      "loss": 1.6836,
+      "step": 5809
+    },
+    {
+      "epoch": 0.3238573021181717,
+      "grad_norm": 0.5286592245101929,
+      "learning_rate": 8e-05,
+      "loss": 1.8199,
+      "step": 5810
+    },
+    {
+      "epoch": 0.3239130434782609,
+      "grad_norm": 0.4898530840873718,
+      "learning_rate": 8e-05,
+      "loss": 1.584,
+      "step": 5811
+    },
+    {
+      "epoch": 0.32396878483835007,
+      "grad_norm": 0.48954761028289795,
+      "learning_rate": 8e-05,
+      "loss": 1.6911,
+      "step": 5812
+    },
+    {
+      "epoch": 0.32402452619843924,
+      "grad_norm": 0.5214674472808838,
+      "learning_rate": 8e-05,
+      "loss": 1.7365,
+      "step": 5813
+    },
+    {
+      "epoch": 0.3240802675585284,
+      "grad_norm": 0.47776153683662415,
+      "learning_rate": 8e-05,
+      "loss": 1.5675,
+      "step": 5814
+    },
+    {
+      "epoch": 0.3241360089186176,
+      "grad_norm": 0.4698529839515686,
+      "learning_rate": 8e-05,
+      "loss": 1.4761,
+      "step": 5815
+    },
+    {
+      "epoch": 0.3241917502787068,
+      "grad_norm": 0.4994838833808899,
+      "learning_rate": 8e-05,
+      "loss": 1.6804,
+      "step": 5816
+    },
+    {
+      "epoch": 0.32424749163879596,
+      "grad_norm": 0.5514016151428223,
+      "learning_rate": 8e-05,
+      "loss": 1.701,
+      "step": 5817
+    },
+    {
+      "epoch": 0.3243032329988852,
+      "grad_norm": 0.5442790389060974,
+      "learning_rate": 8e-05,
+      "loss": 1.7503,
+      "step": 5818
+    },
+    {
+      "epoch": 0.3243589743589744,
+      "grad_norm": 0.5837408304214478,
+      "learning_rate": 8e-05,
+      "loss": 1.6954,
+      "step": 5819
+    },
+    {
+      "epoch": 0.32441471571906355,
+      "grad_norm": 0.531805694103241,
+      "learning_rate": 8e-05,
+      "loss": 1.7151,
+      "step": 5820
+    },
+    {
+      "epoch": 0.32447045707915273,
+      "grad_norm": 0.49297383427619934,
+      "learning_rate": 8e-05,
+      "loss": 1.6382,
+      "step": 5821
+    },
+    {
+      "epoch": 0.3245261984392419,
+      "grad_norm": 0.5175424218177795,
+      "learning_rate": 8e-05,
+      "loss": 1.7722,
+      "step": 5822
+    },
+    {
+      "epoch": 0.3245819397993311,
+      "grad_norm": 0.5085750222206116,
+      "learning_rate": 8e-05,
+      "loss": 1.7484,
+      "step": 5823
+    },
+    {
+      "epoch": 0.32463768115942027,
+      "grad_norm": 0.45293039083480835,
+      "learning_rate": 8e-05,
+      "loss": 1.4519,
+      "step": 5824
+    },
+    {
+      "epoch": 0.3246934225195095,
+      "grad_norm": 0.5808523893356323,
+      "learning_rate": 8e-05,
+      "loss": 1.7628,
+      "step": 5825
+    },
+    {
+      "epoch": 0.3247491638795987,
+      "grad_norm": 0.5092936754226685,
+      "learning_rate": 8e-05,
+      "loss": 1.5918,
+      "step": 5826
+    },
+    {
+      "epoch": 0.32480490523968786,
+      "grad_norm": 0.5005234479904175,
+      "learning_rate": 8e-05,
+      "loss": 1.5775,
+      "step": 5827
+    },
+    {
+      "epoch": 0.32486064659977704,
+      "grad_norm": 0.4908946454524994,
+      "learning_rate": 8e-05,
+      "loss": 1.6302,
+      "step": 5828
+    },
+    {
+      "epoch": 0.3249163879598662,
+      "grad_norm": 0.5337138175964355,
+      "learning_rate": 8e-05,
+      "loss": 1.7442,
+      "step": 5829
+    },
+    {
+      "epoch": 0.3249721293199554,
+      "grad_norm": 0.5431962013244629,
+      "learning_rate": 8e-05,
+      "loss": 1.8264,
+      "step": 5830
+    },
+    {
+      "epoch": 0.3250278706800446,
+      "grad_norm": 0.489956259727478,
+      "learning_rate": 8e-05,
+      "loss": 1.4519,
+      "step": 5831
+    },
+    {
+      "epoch": 0.3250836120401338,
+      "grad_norm": 0.4591715931892395,
+      "learning_rate": 8e-05,
+      "loss": 1.3764,
+      "step": 5832
+    },
+    {
+      "epoch": 0.325139353400223,
+      "grad_norm": 0.5103399753570557,
+      "learning_rate": 8e-05,
+      "loss": 1.8932,
+      "step": 5833
+    },
+    {
+      "epoch": 0.32519509476031216,
+      "grad_norm": 0.5174875855445862,
+      "learning_rate": 8e-05,
+      "loss": 1.6451,
+      "step": 5834
+    },
+    {
+      "epoch": 0.32525083612040134,
+      "grad_norm": 0.5021142959594727,
+      "learning_rate": 8e-05,
+      "loss": 1.7556,
+      "step": 5835
+    },
+    {
+      "epoch": 0.3253065774804905,
+      "grad_norm": 0.517684817314148,
+      "learning_rate": 8e-05,
+      "loss": 1.8316,
+      "step": 5836
+    },
+    {
+      "epoch": 0.3253623188405797,
+      "grad_norm": 0.500555694103241,
+      "learning_rate": 8e-05,
+      "loss": 1.7339,
+      "step": 5837
+    },
+    {
+      "epoch": 0.3254180602006689,
+      "grad_norm": 0.5088232159614563,
+      "learning_rate": 8e-05,
+      "loss": 1.771,
+      "step": 5838
+    },
+    {
+      "epoch": 0.32547380156075806,
+      "grad_norm": 0.4987097680568695,
+      "learning_rate": 8e-05,
+      "loss": 1.737,
+      "step": 5839
+    },
+    {
+      "epoch": 0.3255295429208473,
+      "grad_norm": 0.5131197571754456,
+      "learning_rate": 8e-05,
+      "loss": 1.846,
+      "step": 5840
+    },
+    {
+      "epoch": 0.32558528428093647,
+      "grad_norm": 0.5244218707084656,
+      "learning_rate": 8e-05,
+      "loss": 1.8961,
+      "step": 5841
+    },
+    {
+      "epoch": 0.32564102564102565,
+      "grad_norm": 0.4499856233596802,
+      "learning_rate": 8e-05,
+      "loss": 1.5501,
+      "step": 5842
+    },
+    {
+      "epoch": 0.32569676700111483,
+      "grad_norm": 0.4926725924015045,
+      "learning_rate": 8e-05,
+      "loss": 1.5282,
+      "step": 5843
+    },
+    {
+      "epoch": 0.325752508361204,
+      "grad_norm": 0.49711015820503235,
+      "learning_rate": 8e-05,
+      "loss": 1.6335,
+      "step": 5844
+    },
+    {
+      "epoch": 0.3258082497212932,
+      "grad_norm": 0.48810988664627075,
+      "learning_rate": 8e-05,
+      "loss": 1.7216,
+      "step": 5845
+    },
+    {
+      "epoch": 0.32586399108138236,
+      "grad_norm": 0.49561434984207153,
+      "learning_rate": 8e-05,
+      "loss": 1.6098,
+      "step": 5846
+    },
+    {
+      "epoch": 0.3259197324414716,
+      "grad_norm": 0.5036212205886841,
+      "learning_rate": 8e-05,
+      "loss": 1.749,
+      "step": 5847
+    },
+    {
+      "epoch": 0.3259754738015608,
+      "grad_norm": 0.5274983644485474,
+      "learning_rate": 8e-05,
+      "loss": 1.7112,
+      "step": 5848
+    },
+    {
+      "epoch": 0.32603121516164996,
+      "grad_norm": 0.45901793241500854,
+      "learning_rate": 8e-05,
+      "loss": 1.5345,
+      "step": 5849
+    },
+    {
+      "epoch": 0.32608695652173914,
+      "grad_norm": 0.4732411801815033,
+      "learning_rate": 8e-05,
+      "loss": 1.6603,
+      "step": 5850
+    },
+    {
+      "epoch": 0.3261426978818283,
+      "grad_norm": 0.5366707444190979,
+      "learning_rate": 8e-05,
+      "loss": 1.6819,
+      "step": 5851
+    },
+    {
+      "epoch": 0.3261984392419175,
+      "grad_norm": 0.48991480469703674,
+      "learning_rate": 8e-05,
+      "loss": 1.667,
+      "step": 5852
+    },
+    {
+      "epoch": 0.32625418060200667,
+      "grad_norm": 0.5174232721328735,
+      "learning_rate": 8e-05,
+      "loss": 1.7938,
+      "step": 5853
+    },
+    {
+      "epoch": 0.32630992196209585,
+      "grad_norm": 0.6045074462890625,
+      "learning_rate": 8e-05,
+      "loss": 1.9182,
+      "step": 5854
+    },
+    {
+      "epoch": 0.3263656633221851,
+      "grad_norm": 0.5650275945663452,
+      "learning_rate": 8e-05,
+      "loss": 1.8663,
+      "step": 5855
+    },
+    {
+      "epoch": 0.32642140468227426,
+      "grad_norm": 0.5077049732208252,
+      "learning_rate": 8e-05,
+      "loss": 1.3586,
+      "step": 5856
+    },
+    {
+      "epoch": 0.32647714604236344,
+      "grad_norm": 0.47277364134788513,
+      "learning_rate": 8e-05,
+      "loss": 1.5135,
+      "step": 5857
+    },
+    {
+      "epoch": 0.3265328874024526,
+      "grad_norm": 0.49664178490638733,
+      "learning_rate": 8e-05,
+      "loss": 1.7544,
+      "step": 5858
+    },
+    {
+      "epoch": 0.3265886287625418,
+      "grad_norm": 0.47362759709358215,
+      "learning_rate": 8e-05,
+      "loss": 1.588,
+      "step": 5859
+    },
+    {
+      "epoch": 0.326644370122631,
+      "grad_norm": 0.5836063623428345,
+      "learning_rate": 8e-05,
+      "loss": 1.8409,
+      "step": 5860
+    },
+    {
+      "epoch": 0.32670011148272016,
+      "grad_norm": 0.524232029914856,
+      "learning_rate": 8e-05,
+      "loss": 1.7189,
+      "step": 5861
+    },
+    {
+      "epoch": 0.3267558528428094,
+      "grad_norm": 0.4531126618385315,
+      "learning_rate": 8e-05,
+      "loss": 1.5504,
+      "step": 5862
+    },
+    {
+      "epoch": 0.32681159420289857,
+      "grad_norm": 0.5362951755523682,
+      "learning_rate": 8e-05,
+      "loss": 1.8944,
+      "step": 5863
+    },
+    {
+      "epoch": 0.32686733556298775,
+      "grad_norm": 0.542114794254303,
+      "learning_rate": 8e-05,
+      "loss": 1.7909,
+      "step": 5864
+    },
+    {
+      "epoch": 0.3269230769230769,
+      "grad_norm": 0.5658450126647949,
+      "learning_rate": 8e-05,
+      "loss": 1.9196,
+      "step": 5865
+    },
+    {
+      "epoch": 0.3269788182831661,
+      "grad_norm": 0.495639830827713,
+      "learning_rate": 8e-05,
+      "loss": 1.5991,
+      "step": 5866
+    },
+    {
+      "epoch": 0.3270345596432553,
+      "grad_norm": 0.5281422734260559,
+      "learning_rate": 8e-05,
+      "loss": 1.5525,
+      "step": 5867
+    },
+    {
+      "epoch": 0.32709030100334446,
+      "grad_norm": 0.5133866667747498,
+      "learning_rate": 8e-05,
+      "loss": 1.8848,
+      "step": 5868
+    },
+    {
+      "epoch": 0.32714604236343364,
+      "grad_norm": 0.5450278520584106,
+      "learning_rate": 8e-05,
+      "loss": 1.7997,
+      "step": 5869
+    },
+    {
+      "epoch": 0.3272017837235229,
+      "grad_norm": 0.5412588119506836,
+      "learning_rate": 8e-05,
+      "loss": 1.8669,
+      "step": 5870
+    },
+    {
+      "epoch": 0.32725752508361206,
+      "grad_norm": 0.5211524963378906,
+      "learning_rate": 8e-05,
+      "loss": 1.6767,
+      "step": 5871
+    },
+    {
+      "epoch": 0.32731326644370123,
+      "grad_norm": 0.505767285823822,
+      "learning_rate": 8e-05,
+      "loss": 1.7258,
+      "step": 5872
+    },
+    {
+      "epoch": 0.3273690078037904,
+      "grad_norm": 0.5113710761070251,
+      "learning_rate": 8e-05,
+      "loss": 1.8295,
+      "step": 5873
+    },
+    {
+      "epoch": 0.3274247491638796,
+      "grad_norm": 0.5271311402320862,
+      "learning_rate": 8e-05,
+      "loss": 2.0436,
+      "step": 5874
+    },
+    {
+      "epoch": 0.32748049052396877,
+      "grad_norm": 0.4804200530052185,
+      "learning_rate": 8e-05,
+      "loss": 1.5799,
+      "step": 5875
+    },
+    {
+      "epoch": 0.32753623188405795,
+      "grad_norm": 0.5401303768157959,
+      "learning_rate": 8e-05,
+      "loss": 1.887,
+      "step": 5876
+    },
+    {
+      "epoch": 0.3275919732441472,
+      "grad_norm": 0.5247963666915894,
+      "learning_rate": 8e-05,
+      "loss": 1.9989,
+      "step": 5877
+    },
+    {
+      "epoch": 0.32764771460423636,
+      "grad_norm": 0.5222291350364685,
+      "learning_rate": 8e-05,
+      "loss": 1.9649,
+      "step": 5878
+    },
+    {
+      "epoch": 0.32770345596432554,
+      "grad_norm": 0.5129970908164978,
+      "learning_rate": 8e-05,
+      "loss": 1.861,
+      "step": 5879
+    },
+    {
+      "epoch": 0.3277591973244147,
+      "grad_norm": 0.48588502407073975,
+      "learning_rate": 8e-05,
+      "loss": 1.5694,
+      "step": 5880
+    },
+    {
+      "epoch": 0.3278149386845039,
+      "grad_norm": 0.4927371144294739,
+      "learning_rate": 8e-05,
+      "loss": 1.7139,
+      "step": 5881
+    },
+    {
+      "epoch": 0.3278706800445931,
+      "grad_norm": 0.4911452829837799,
+      "learning_rate": 8e-05,
+      "loss": 1.5714,
+      "step": 5882
+    },
+    {
+      "epoch": 0.32792642140468226,
+      "grad_norm": 0.5015363097190857,
+      "learning_rate": 8e-05,
+      "loss": 1.5972,
+      "step": 5883
+    },
+    {
+      "epoch": 0.32798216276477143,
+      "grad_norm": 0.5393746495246887,
+      "learning_rate": 8e-05,
+      "loss": 1.7549,
+      "step": 5884
+    },
+    {
+      "epoch": 0.32803790412486067,
+      "grad_norm": 0.4892618656158447,
+      "learning_rate": 8e-05,
+      "loss": 1.7331,
+      "step": 5885
+    },
+    {
+      "epoch": 0.32809364548494985,
+      "grad_norm": 0.47445163130760193,
+      "learning_rate": 8e-05,
+      "loss": 1.3907,
+      "step": 5886
+    },
+    {
+      "epoch": 0.328149386845039,
+      "grad_norm": 0.5472539663314819,
+      "learning_rate": 8e-05,
+      "loss": 1.8212,
+      "step": 5887
+    },
+    {
+      "epoch": 0.3282051282051282,
+      "grad_norm": 0.51388019323349,
+      "learning_rate": 8e-05,
+      "loss": 1.7439,
+      "step": 5888
+    },
+    {
+      "epoch": 0.3282608695652174,
+      "grad_norm": 0.4770544767379761,
+      "learning_rate": 8e-05,
+      "loss": 1.5598,
+      "step": 5889
+    },
+    {
+      "epoch": 0.32831661092530656,
+      "grad_norm": 0.5853254199028015,
+      "learning_rate": 8e-05,
+      "loss": 1.9501,
+      "step": 5890
+    },
+    {
+      "epoch": 0.32837235228539574,
+      "grad_norm": 0.559180736541748,
+      "learning_rate": 8e-05,
+      "loss": 1.9934,
+      "step": 5891
+    },
+    {
+      "epoch": 0.328428093645485,
+      "grad_norm": 0.5387535095214844,
+      "learning_rate": 8e-05,
+      "loss": 1.648,
+      "step": 5892
+    },
+    {
+      "epoch": 0.32848383500557415,
+      "grad_norm": 0.5571669936180115,
+      "learning_rate": 8e-05,
+      "loss": 1.9088,
+      "step": 5893
+    },
+    {
+      "epoch": 0.32853957636566333,
+      "grad_norm": 0.5683614611625671,
+      "learning_rate": 8e-05,
+      "loss": 1.9428,
+      "step": 5894
+    },
+    {
+      "epoch": 0.3285953177257525,
+      "grad_norm": 0.4627940058708191,
+      "learning_rate": 8e-05,
+      "loss": 1.43,
+      "step": 5895
+    },
+    {
+      "epoch": 0.3286510590858417,
+      "grad_norm": 0.5188383460044861,
+      "learning_rate": 8e-05,
+      "loss": 1.5124,
+      "step": 5896
+    },
+    {
+      "epoch": 0.32870680044593087,
+      "grad_norm": 0.5177381634712219,
+      "learning_rate": 8e-05,
+      "loss": 1.7765,
+      "step": 5897
+    },
+    {
+      "epoch": 0.32876254180602005,
+      "grad_norm": 0.5070524215698242,
+      "learning_rate": 8e-05,
+      "loss": 1.6511,
+      "step": 5898
+    },
+    {
+      "epoch": 0.3288182831661092,
+      "grad_norm": 0.4960431158542633,
+      "learning_rate": 8e-05,
+      "loss": 1.7412,
+      "step": 5899
+    },
+    {
+      "epoch": 0.32887402452619846,
+      "grad_norm": 0.5233199596405029,
+      "learning_rate": 8e-05,
+      "loss": 1.647,
+      "step": 5900
+    },
+    {
+      "epoch": 0.32892976588628764,
+      "grad_norm": 0.535688579082489,
+      "learning_rate": 8e-05,
+      "loss": 1.7367,
+      "step": 5901
+    },
+    {
+      "epoch": 0.3289855072463768,
+      "grad_norm": 0.5019519925117493,
+      "learning_rate": 8e-05,
+      "loss": 2.0115,
+      "step": 5902
+    },
+    {
+      "epoch": 0.329041248606466,
+      "grad_norm": 0.5208526253700256,
+      "learning_rate": 8e-05,
+      "loss": 1.6949,
+      "step": 5903
+    },
+    {
+      "epoch": 0.3290969899665552,
+      "grad_norm": 0.5380985140800476,
+      "learning_rate": 8e-05,
+      "loss": 1.8018,
+      "step": 5904
+    },
+    {
+      "epoch": 0.32915273132664435,
+      "grad_norm": 0.5154558420181274,
+      "learning_rate": 8e-05,
+      "loss": 1.8893,
+      "step": 5905
+    },
+    {
+      "epoch": 0.32920847268673353,
+      "grad_norm": 0.5352508425712585,
+      "learning_rate": 8e-05,
+      "loss": 1.7274,
+      "step": 5906
+    },
+    {
+      "epoch": 0.32926421404682277,
+      "grad_norm": 0.4826267957687378,
+      "learning_rate": 8e-05,
+      "loss": 1.342,
+      "step": 5907
+    },
+    {
+      "epoch": 0.32931995540691195,
+      "grad_norm": 0.5051599144935608,
+      "learning_rate": 8e-05,
+      "loss": 1.5785,
+      "step": 5908
+    },
+    {
+      "epoch": 0.3293756967670011,
+      "grad_norm": 0.5084307193756104,
+      "learning_rate": 8e-05,
+      "loss": 1.6622,
+      "step": 5909
+    },
+    {
+      "epoch": 0.3294314381270903,
+      "grad_norm": 0.5318410992622375,
+      "learning_rate": 8e-05,
+      "loss": 1.7986,
+      "step": 5910
+    },
+    {
+      "epoch": 0.3294871794871795,
+      "grad_norm": 0.4830995202064514,
+      "learning_rate": 8e-05,
+      "loss": 1.4463,
+      "step": 5911
+    },
+    {
+      "epoch": 0.32954292084726866,
+      "grad_norm": 0.552952229976654,
+      "learning_rate": 8e-05,
+      "loss": 1.8705,
+      "step": 5912
+    },
+    {
+      "epoch": 0.32959866220735784,
+      "grad_norm": 0.5508795976638794,
+      "learning_rate": 8e-05,
+      "loss": 1.7107,
+      "step": 5913
+    },
+    {
+      "epoch": 0.329654403567447,
+      "grad_norm": 0.49108216166496277,
+      "learning_rate": 8e-05,
+      "loss": 1.6975,
+      "step": 5914
+    },
+    {
+      "epoch": 0.32971014492753625,
+      "grad_norm": 0.5228166580200195,
+      "learning_rate": 8e-05,
+      "loss": 1.6342,
+      "step": 5915
+    },
+    {
+      "epoch": 0.32976588628762543,
+      "grad_norm": 0.5140056610107422,
+      "learning_rate": 8e-05,
+      "loss": 1.738,
+      "step": 5916
+    },
+    {
+      "epoch": 0.3298216276477146,
+      "grad_norm": 0.5653532147407532,
+      "learning_rate": 8e-05,
+      "loss": 1.8411,
+      "step": 5917
+    },
+    {
+      "epoch": 0.3298773690078038,
+      "grad_norm": 0.5260170102119446,
+      "learning_rate": 8e-05,
+      "loss": 1.7909,
+      "step": 5918
+    },
+    {
+      "epoch": 0.32993311036789297,
+      "grad_norm": 0.5184810757637024,
+      "learning_rate": 8e-05,
+      "loss": 1.5628,
+      "step": 5919
+    },
+    {
+      "epoch": 0.32998885172798215,
+      "grad_norm": 0.507621705532074,
+      "learning_rate": 8e-05,
+      "loss": 1.7064,
+      "step": 5920
+    },
+    {
+      "epoch": 0.3300445930880713,
+      "grad_norm": 0.4716615378856659,
+      "learning_rate": 8e-05,
+      "loss": 1.5359,
+      "step": 5921
+    },
+    {
+      "epoch": 0.33010033444816056,
+      "grad_norm": 0.5044958591461182,
+      "learning_rate": 8e-05,
+      "loss": 1.5808,
+      "step": 5922
+    },
+    {
+      "epoch": 0.33015607580824974,
+      "grad_norm": 0.5160906910896301,
+      "learning_rate": 8e-05,
+      "loss": 1.5896,
+      "step": 5923
+    },
+    {
+      "epoch": 0.3302118171683389,
+      "grad_norm": 0.5261350274085999,
+      "learning_rate": 8e-05,
+      "loss": 1.7253,
+      "step": 5924
+    },
+    {
+      "epoch": 0.3302675585284281,
+      "grad_norm": 0.5316385626792908,
+      "learning_rate": 8e-05,
+      "loss": 1.6345,
+      "step": 5925
+    },
+    {
+      "epoch": 0.3303232998885173,
+      "grad_norm": 0.5023976564407349,
+      "learning_rate": 8e-05,
+      "loss": 1.6139,
+      "step": 5926
+    },
+    {
+      "epoch": 0.33037904124860645,
+      "grad_norm": 0.5029090046882629,
+      "learning_rate": 8e-05,
+      "loss": 1.5976,
+      "step": 5927
+    },
+    {
+      "epoch": 0.33043478260869563,
+      "grad_norm": 0.4609101116657257,
+      "learning_rate": 8e-05,
+      "loss": 1.2439,
+      "step": 5928
+    },
+    {
+      "epoch": 0.3304905239687848,
+      "grad_norm": 0.5388715267181396,
+      "learning_rate": 8e-05,
+      "loss": 1.5517,
+      "step": 5929
+    },
+    {
+      "epoch": 0.33054626532887404,
+      "grad_norm": 0.5388485193252563,
+      "learning_rate": 8e-05,
+      "loss": 1.6922,
+      "step": 5930
+    },
+    {
+      "epoch": 0.3306020066889632,
+      "grad_norm": 0.46442028880119324,
+      "learning_rate": 8e-05,
+      "loss": 1.5227,
+      "step": 5931
+    },
+    {
+      "epoch": 0.3306577480490524,
+      "grad_norm": 0.47519493103027344,
+      "learning_rate": 8e-05,
+      "loss": 1.7977,
+      "step": 5932
+    },
+    {
+      "epoch": 0.3307134894091416,
+      "grad_norm": 0.5591146945953369,
+      "learning_rate": 8e-05,
+      "loss": 1.8752,
+      "step": 5933
+    },
+    {
+      "epoch": 0.33076923076923076,
+      "grad_norm": 0.5389036536216736,
+      "learning_rate": 8e-05,
+      "loss": 1.7068,
+      "step": 5934
+    },
+    {
+      "epoch": 0.33082497212931994,
+      "grad_norm": 0.53550124168396,
+      "learning_rate": 8e-05,
+      "loss": 1.7691,
+      "step": 5935
+    },
+    {
+      "epoch": 0.3308807134894091,
+      "grad_norm": 0.49706998467445374,
+      "learning_rate": 8e-05,
+      "loss": 1.6261,
+      "step": 5936
+    },
+    {
+      "epoch": 0.33093645484949835,
+      "grad_norm": 0.5716555118560791,
+      "learning_rate": 8e-05,
+      "loss": 1.671,
+      "step": 5937
+    },
+    {
+      "epoch": 0.33099219620958753,
+      "grad_norm": 0.5462481379508972,
+      "learning_rate": 8e-05,
+      "loss": 1.6752,
+      "step": 5938
+    },
+    {
+      "epoch": 0.3310479375696767,
+      "grad_norm": 0.5143697261810303,
+      "learning_rate": 8e-05,
+      "loss": 1.6878,
+      "step": 5939
+    },
+    {
+      "epoch": 0.3311036789297659,
+      "grad_norm": 0.4737571179866791,
+      "learning_rate": 8e-05,
+      "loss": 1.5534,
+      "step": 5940
+    },
+    {
+      "epoch": 0.33115942028985507,
+      "grad_norm": 0.519430935382843,
+      "learning_rate": 8e-05,
+      "loss": 1.6366,
+      "step": 5941
+    },
+    {
+      "epoch": 0.33121516164994425,
+      "grad_norm": 0.5122805833816528,
+      "learning_rate": 8e-05,
+      "loss": 1.7942,
+      "step": 5942
+    },
+    {
+      "epoch": 0.3312709030100334,
+      "grad_norm": 0.5048823952674866,
+      "learning_rate": 8e-05,
+      "loss": 1.6511,
+      "step": 5943
+    },
+    {
+      "epoch": 0.33132664437012266,
+      "grad_norm": 0.4657517075538635,
+      "learning_rate": 8e-05,
+      "loss": 1.6489,
+      "step": 5944
+    },
+    {
+      "epoch": 0.33138238573021184,
+      "grad_norm": 0.515855610370636,
+      "learning_rate": 8e-05,
+      "loss": 1.6975,
+      "step": 5945
+    },
+    {
+      "epoch": 0.331438127090301,
+      "grad_norm": 0.532136082649231,
+      "learning_rate": 8e-05,
+      "loss": 1.6791,
+      "step": 5946
+    },
+    {
+      "epoch": 0.3314938684503902,
+      "grad_norm": 0.541500449180603,
+      "learning_rate": 8e-05,
+      "loss": 1.7849,
+      "step": 5947
+    },
+    {
+      "epoch": 0.3315496098104794,
+      "grad_norm": 0.584726870059967,
+      "learning_rate": 8e-05,
+      "loss": 1.7246,
+      "step": 5948
+    },
+    {
+      "epoch": 0.33160535117056855,
+      "grad_norm": 0.5696508884429932,
+      "learning_rate": 8e-05,
+      "loss": 1.9362,
+      "step": 5949
+    },
+    {
+      "epoch": 0.33166109253065773,
+      "grad_norm": 0.5092194080352783,
+      "learning_rate": 8e-05,
+      "loss": 1.5644,
+      "step": 5950
+    },
+    {
+      "epoch": 0.3317168338907469,
+      "grad_norm": 0.5051195621490479,
+      "learning_rate": 8e-05,
+      "loss": 1.3563,
+      "step": 5951
+    },
+    {
+      "epoch": 0.33177257525083614,
+      "grad_norm": 0.5067978501319885,
+      "learning_rate": 8e-05,
+      "loss": 1.9072,
+      "step": 5952
+    },
+    {
+      "epoch": 0.3318283166109253,
+      "grad_norm": 0.5300750136375427,
+      "learning_rate": 8e-05,
+      "loss": 1.7816,
+      "step": 5953
+    },
+    {
+      "epoch": 0.3318840579710145,
+      "grad_norm": 0.5331507325172424,
+      "learning_rate": 8e-05,
+      "loss": 1.8968,
+      "step": 5954
+    },
+    {
+      "epoch": 0.3319397993311037,
+      "grad_norm": 0.480129212141037,
+      "learning_rate": 8e-05,
+      "loss": 1.338,
+      "step": 5955
+    },
+    {
+      "epoch": 0.33199554069119286,
+      "grad_norm": 0.5224987864494324,
+      "learning_rate": 8e-05,
+      "loss": 1.7565,
+      "step": 5956
+    },
+    {
+      "epoch": 0.33205128205128204,
+      "grad_norm": 0.5832299590110779,
+      "learning_rate": 8e-05,
+      "loss": 1.8018,
+      "step": 5957
+    },
+    {
+      "epoch": 0.3321070234113712,
+      "grad_norm": 0.5152959227561951,
+      "learning_rate": 8e-05,
+      "loss": 1.7769,
+      "step": 5958
+    },
+    {
+      "epoch": 0.33216276477146045,
+      "grad_norm": 0.5299351215362549,
+      "learning_rate": 8e-05,
+      "loss": 1.7476,
+      "step": 5959
+    },
+    {
+      "epoch": 0.33221850613154963,
+      "grad_norm": 0.5079212784767151,
+      "learning_rate": 8e-05,
+      "loss": 1.5925,
+      "step": 5960
+    },
+    {
+      "epoch": 0.3322742474916388,
+      "grad_norm": 0.5432602167129517,
+      "learning_rate": 8e-05,
+      "loss": 1.7368,
+      "step": 5961
+    },
+    {
+      "epoch": 0.332329988851728,
+      "grad_norm": 0.5113498568534851,
+      "learning_rate": 8e-05,
+      "loss": 1.9163,
+      "step": 5962
+    },
+    {
+      "epoch": 0.33238573021181717,
+      "grad_norm": 0.5326887965202332,
+      "learning_rate": 8e-05,
+      "loss": 1.6331,
+      "step": 5963
+    },
+    {
+      "epoch": 0.33244147157190634,
+      "grad_norm": 0.5543447136878967,
+      "learning_rate": 8e-05,
+      "loss": 1.8223,
+      "step": 5964
+    },
+    {
+      "epoch": 0.3324972129319955,
+      "grad_norm": 0.4961104691028595,
+      "learning_rate": 8e-05,
+      "loss": 1.6367,
+      "step": 5965
+    },
+    {
+      "epoch": 0.3325529542920847,
+      "grad_norm": 0.4903685748577118,
+      "learning_rate": 8e-05,
+      "loss": 1.5351,
+      "step": 5966
+    },
+    {
+      "epoch": 0.33260869565217394,
+      "grad_norm": 0.5134567618370056,
+      "learning_rate": 8e-05,
+      "loss": 1.9116,
+      "step": 5967
+    },
+    {
+      "epoch": 0.3326644370122631,
+      "grad_norm": 0.5132040977478027,
+      "learning_rate": 8e-05,
+      "loss": 1.4611,
+      "step": 5968
+    },
+    {
+      "epoch": 0.3327201783723523,
+      "grad_norm": 0.5018925070762634,
+      "learning_rate": 8e-05,
+      "loss": 1.8217,
+      "step": 5969
+    },
+    {
+      "epoch": 0.33277591973244147,
+      "grad_norm": 0.6516698002815247,
+      "learning_rate": 8e-05,
+      "loss": 1.4352,
+      "step": 5970
+    },
+    {
+      "epoch": 0.33283166109253065,
+      "grad_norm": 0.5061227679252625,
+      "learning_rate": 8e-05,
+      "loss": 1.6212,
+      "step": 5971
+    },
+    {
+      "epoch": 0.33288740245261983,
+      "grad_norm": 0.4874931573867798,
+      "learning_rate": 8e-05,
+      "loss": 1.5874,
+      "step": 5972
+    },
+    {
+      "epoch": 0.332943143812709,
+      "grad_norm": 0.5170412063598633,
+      "learning_rate": 8e-05,
+      "loss": 1.8376,
+      "step": 5973
+    },
+    {
+      "epoch": 0.33299888517279824,
+      "grad_norm": 0.48148781061172485,
+      "learning_rate": 8e-05,
+      "loss": 1.6972,
+      "step": 5974
+    },
+    {
+      "epoch": 0.3330546265328874,
+      "grad_norm": 0.5671499967575073,
+      "learning_rate": 8e-05,
+      "loss": 1.8067,
+      "step": 5975
+    },
+    {
+      "epoch": 0.3331103678929766,
+      "grad_norm": 0.5110177397727966,
+      "learning_rate": 8e-05,
+      "loss": 1.7419,
+      "step": 5976
+    },
+    {
+      "epoch": 0.3331661092530658,
+      "grad_norm": 0.49894124269485474,
+      "learning_rate": 8e-05,
+      "loss": 1.847,
+      "step": 5977
+    },
+    {
+      "epoch": 0.33322185061315496,
+      "grad_norm": 0.5159071087837219,
+      "learning_rate": 8e-05,
+      "loss": 1.759,
+      "step": 5978
+    },
+    {
+      "epoch": 0.33327759197324414,
+      "grad_norm": 0.5592484474182129,
+      "learning_rate": 8e-05,
+      "loss": 1.9145,
+      "step": 5979
+    },
+    {
+      "epoch": 0.3333333333333333,
+      "grad_norm": 0.5127020478248596,
+      "learning_rate": 8e-05,
+      "loss": 1.7149,
+      "step": 5980
+    },
+    {
+      "epoch": 0.3333890746934225,
+      "grad_norm": 0.5237054228782654,
+      "learning_rate": 8e-05,
+      "loss": 1.721,
+      "step": 5981
+    },
+    {
+      "epoch": 0.33344481605351173,
+      "grad_norm": 0.47445914149284363,
+      "learning_rate": 8e-05,
+      "loss": 1.5014,
+      "step": 5982
+    },
+    {
+      "epoch": 0.3335005574136009,
+      "grad_norm": 0.4995151162147522,
+      "learning_rate": 8e-05,
+      "loss": 1.7636,
+      "step": 5983
+    },
+    {
+      "epoch": 0.3335562987736901,
+      "grad_norm": 0.5311259031295776,
+      "learning_rate": 8e-05,
+      "loss": 1.8416,
+      "step": 5984
+    },
+    {
+      "epoch": 0.33361204013377926,
+      "grad_norm": 0.5449182391166687,
+      "learning_rate": 8e-05,
+      "loss": 1.9423,
+      "step": 5985
+    },
+    {
+      "epoch": 0.33366778149386844,
+      "grad_norm": 0.5676257014274597,
+      "learning_rate": 8e-05,
+      "loss": 1.8433,
+      "step": 5986
+    },
+    {
+      "epoch": 0.3337235228539576,
+      "grad_norm": 0.5525195002555847,
+      "learning_rate": 8e-05,
+      "loss": 1.7521,
+      "step": 5987
+    },
+    {
+      "epoch": 0.3337792642140468,
+      "grad_norm": 0.5231057405471802,
+      "learning_rate": 8e-05,
+      "loss": 1.7974,
+      "step": 5988
+    },
+    {
+      "epoch": 0.33383500557413603,
+      "grad_norm": 0.5249040126800537,
+      "learning_rate": 8e-05,
+      "loss": 1.5266,
+      "step": 5989
+    },
+    {
+      "epoch": 0.3338907469342252,
+      "grad_norm": 0.6000924706459045,
+      "learning_rate": 8e-05,
+      "loss": 1.9797,
+      "step": 5990
+    },
+    {
+      "epoch": 0.3339464882943144,
+      "grad_norm": 0.5164068937301636,
+      "learning_rate": 8e-05,
+      "loss": 1.6044,
+      "step": 5991
+    },
+    {
+      "epoch": 0.33400222965440357,
+      "grad_norm": 0.4820178747177124,
+      "learning_rate": 8e-05,
+      "loss": 1.6805,
+      "step": 5992
+    },
+    {
+      "epoch": 0.33405797101449275,
+      "grad_norm": 0.4954069256782532,
+      "learning_rate": 8e-05,
+      "loss": 1.6487,
+      "step": 5993
+    },
+    {
+      "epoch": 0.33411371237458193,
+      "grad_norm": 0.5077088475227356,
+      "learning_rate": 8e-05,
+      "loss": 1.7436,
+      "step": 5994
+    },
+    {
+      "epoch": 0.3341694537346711,
+      "grad_norm": 0.5106683373451233,
+      "learning_rate": 8e-05,
+      "loss": 1.7808,
+      "step": 5995
+    },
+    {
+      "epoch": 0.3342251950947603,
+      "grad_norm": 0.5065857768058777,
+      "learning_rate": 8e-05,
+      "loss": 1.5292,
+      "step": 5996
+    },
+    {
+      "epoch": 0.3342809364548495,
+      "grad_norm": 0.5739662647247314,
+      "learning_rate": 8e-05,
+      "loss": 1.9316,
+      "step": 5997
+    },
+    {
+      "epoch": 0.3343366778149387,
+      "grad_norm": 0.5219335556030273,
+      "learning_rate": 8e-05,
+      "loss": 1.5769,
+      "step": 5998
+    },
+    {
+      "epoch": 0.3343924191750279,
+      "grad_norm": 0.5704739689826965,
+      "learning_rate": 8e-05,
+      "loss": 2.1513,
+      "step": 5999
+    },
+    {
+      "epoch": 0.33444816053511706,
+      "grad_norm": 0.5052824020385742,
+      "learning_rate": 8e-05,
+      "loss": 1.6403,
+      "step": 6000
+    },
+    {
+      "epoch": 0.33450390189520623,
+      "grad_norm": 0.5386701226234436,
+      "learning_rate": 8e-05,
+      "loss": 1.881,
+      "step": 6001
+    },
+    {
+      "epoch": 0.3345596432552954,
+      "grad_norm": 0.5315998792648315,
+      "learning_rate": 8e-05,
+      "loss": 1.5746,
+      "step": 6002
+    },
+    {
+      "epoch": 0.3346153846153846,
+      "grad_norm": 0.5564172863960266,
+      "learning_rate": 8e-05,
+      "loss": 1.9253,
+      "step": 6003
+    },
+    {
+      "epoch": 0.3346711259754738,
+      "grad_norm": 0.500681459903717,
+      "learning_rate": 8e-05,
+      "loss": 1.9605,
+      "step": 6004
+    },
+    {
+      "epoch": 0.334726867335563,
+      "grad_norm": 0.5126969218254089,
+      "learning_rate": 8e-05,
+      "loss": 1.6541,
+      "step": 6005
+    },
+    {
+      "epoch": 0.3347826086956522,
+      "grad_norm": 0.5156999826431274,
+      "learning_rate": 8e-05,
+      "loss": 1.6307,
+      "step": 6006
+    },
+    {
+      "epoch": 0.33483835005574136,
+      "grad_norm": 0.5102741122245789,
+      "learning_rate": 8e-05,
+      "loss": 1.6506,
+      "step": 6007
+    },
+    {
+      "epoch": 0.33489409141583054,
+      "grad_norm": 0.5284451246261597,
+      "learning_rate": 8e-05,
+      "loss": 1.8545,
+      "step": 6008
+    },
+    {
+      "epoch": 0.3349498327759197,
+      "grad_norm": 0.5557467341423035,
+      "learning_rate": 8e-05,
+      "loss": 1.8182,
+      "step": 6009
+    },
+    {
+      "epoch": 0.3350055741360089,
+      "grad_norm": 0.4997308552265167,
+      "learning_rate": 8e-05,
+      "loss": 1.7201,
+      "step": 6010
+    },
+    {
+      "epoch": 0.3350613154960981,
+      "grad_norm": 0.5237441062927246,
+      "learning_rate": 8e-05,
+      "loss": 1.7175,
+      "step": 6011
+    },
+    {
+      "epoch": 0.3351170568561873,
+      "grad_norm": 0.5137330889701843,
+      "learning_rate": 8e-05,
+      "loss": 1.8051,
+      "step": 6012
+    },
+    {
+      "epoch": 0.3351727982162765,
+      "grad_norm": 0.5211109519004822,
+      "learning_rate": 8e-05,
+      "loss": 1.7609,
+      "step": 6013
+    },
+    {
+      "epoch": 0.33522853957636567,
+      "grad_norm": 0.5370551943778992,
+      "learning_rate": 8e-05,
+      "loss": 1.8979,
+      "step": 6014
+    },
+    {
+      "epoch": 0.33528428093645485,
+      "grad_norm": 0.5110623240470886,
+      "learning_rate": 8e-05,
+      "loss": 1.7271,
+      "step": 6015
+    },
+    {
+      "epoch": 0.335340022296544,
+      "grad_norm": 0.5268702507019043,
+      "learning_rate": 8e-05,
+      "loss": 1.8703,
+      "step": 6016
+    },
+    {
+      "epoch": 0.3353957636566332,
+      "grad_norm": 0.5116859078407288,
+      "learning_rate": 8e-05,
+      "loss": 1.6818,
+      "step": 6017
+    },
+    {
+      "epoch": 0.3354515050167224,
+      "grad_norm": 0.669144868850708,
+      "learning_rate": 8e-05,
+      "loss": 1.7151,
+      "step": 6018
+    },
+    {
+      "epoch": 0.3355072463768116,
+      "grad_norm": 0.5689597129821777,
+      "learning_rate": 8e-05,
+      "loss": 1.8494,
+      "step": 6019
+    },
+    {
+      "epoch": 0.3355629877369008,
+      "grad_norm": 0.557144820690155,
+      "learning_rate": 8e-05,
+      "loss": 1.7948,
+      "step": 6020
+    },
+    {
+      "epoch": 0.33561872909699,
+      "grad_norm": 0.5054897665977478,
+      "learning_rate": 8e-05,
+      "loss": 1.8311,
+      "step": 6021
+    },
+    {
+      "epoch": 0.33567447045707915,
+      "grad_norm": 0.516497790813446,
+      "learning_rate": 8e-05,
+      "loss": 1.6477,
+      "step": 6022
+    },
+    {
+      "epoch": 0.33573021181716833,
+      "grad_norm": 0.534079372882843,
+      "learning_rate": 8e-05,
+      "loss": 1.5736,
+      "step": 6023
+    },
+    {
+      "epoch": 0.3357859531772575,
+      "grad_norm": 0.5206681489944458,
+      "learning_rate": 8e-05,
+      "loss": 1.7455,
+      "step": 6024
+    },
+    {
+      "epoch": 0.3358416945373467,
+      "grad_norm": 0.5690346360206604,
+      "learning_rate": 8e-05,
+      "loss": 1.7757,
+      "step": 6025
+    },
+    {
+      "epoch": 0.33589743589743587,
+      "grad_norm": 0.5441909432411194,
+      "learning_rate": 8e-05,
+      "loss": 1.8619,
+      "step": 6026
+    },
+    {
+      "epoch": 0.3359531772575251,
+      "grad_norm": 0.5242137908935547,
+      "learning_rate": 8e-05,
+      "loss": 1.618,
+      "step": 6027
+    },
+    {
+      "epoch": 0.3360089186176143,
+      "grad_norm": 0.5588452219963074,
+      "learning_rate": 8e-05,
+      "loss": 1.8945,
+      "step": 6028
+    },
+    {
+      "epoch": 0.33606465997770346,
+      "grad_norm": 0.4702637791633606,
+      "learning_rate": 8e-05,
+      "loss": 1.548,
+      "step": 6029
+    },
+    {
+      "epoch": 0.33612040133779264,
+      "grad_norm": 0.5237563848495483,
+      "learning_rate": 8e-05,
+      "loss": 1.6127,
+      "step": 6030
+    },
+    {
+      "epoch": 0.3361761426978818,
+      "grad_norm": 0.48560264706611633,
+      "learning_rate": 8e-05,
+      "loss": 1.6201,
+      "step": 6031
+    },
+    {
+      "epoch": 0.336231884057971,
+      "grad_norm": 0.47915685176849365,
+      "learning_rate": 8e-05,
+      "loss": 1.6187,
+      "step": 6032
+    },
+    {
+      "epoch": 0.3362876254180602,
+      "grad_norm": 0.5250439047813416,
+      "learning_rate": 8e-05,
+      "loss": 1.5716,
+      "step": 6033
+    },
+    {
+      "epoch": 0.3363433667781494,
+      "grad_norm": 0.4620858430862427,
+      "learning_rate": 8e-05,
+      "loss": 1.4799,
+      "step": 6034
+    },
+    {
+      "epoch": 0.3363991081382386,
+      "grad_norm": 0.5148380994796753,
+      "learning_rate": 8e-05,
+      "loss": 1.7257,
+      "step": 6035
+    },
+    {
+      "epoch": 0.33645484949832777,
+      "grad_norm": 0.5669644474983215,
+      "learning_rate": 8e-05,
+      "loss": 2.0105,
+      "step": 6036
+    },
+    {
+      "epoch": 0.33651059085841695,
+      "grad_norm": 0.4910716414451599,
+      "learning_rate": 8e-05,
+      "loss": 1.7302,
+      "step": 6037
+    },
+    {
+      "epoch": 0.3365663322185061,
+      "grad_norm": 0.506493866443634,
+      "learning_rate": 8e-05,
+      "loss": 1.3614,
+      "step": 6038
+    },
+    {
+      "epoch": 0.3366220735785953,
+      "grad_norm": 0.5282632112503052,
+      "learning_rate": 8e-05,
+      "loss": 1.9953,
+      "step": 6039
+    },
+    {
+      "epoch": 0.3366778149386845,
+      "grad_norm": 0.49607253074645996,
+      "learning_rate": 8e-05,
+      "loss": 1.6552,
+      "step": 6040
+    },
+    {
+      "epoch": 0.3367335562987737,
+      "grad_norm": 0.5147976875305176,
+      "learning_rate": 8e-05,
+      "loss": 1.806,
+      "step": 6041
+    },
+    {
+      "epoch": 0.3367892976588629,
+      "grad_norm": 0.49952930212020874,
+      "learning_rate": 8e-05,
+      "loss": 1.5324,
+      "step": 6042
+    },
+    {
+      "epoch": 0.3368450390189521,
+      "grad_norm": 0.5874413251876831,
+      "learning_rate": 8e-05,
+      "loss": 2.2245,
+      "step": 6043
+    },
+    {
+      "epoch": 0.33690078037904125,
+      "grad_norm": 0.5175328850746155,
+      "learning_rate": 8e-05,
+      "loss": 1.762,
+      "step": 6044
+    },
+    {
+      "epoch": 0.33695652173913043,
+      "grad_norm": 0.5300764441490173,
+      "learning_rate": 8e-05,
+      "loss": 1.756,
+      "step": 6045
+    },
+    {
+      "epoch": 0.3370122630992196,
+      "grad_norm": 0.524803102016449,
+      "learning_rate": 8e-05,
+      "loss": 1.85,
+      "step": 6046
+    },
+    {
+      "epoch": 0.3370680044593088,
+      "grad_norm": 0.5062090158462524,
+      "learning_rate": 8e-05,
+      "loss": 1.4843,
+      "step": 6047
+    },
+    {
+      "epoch": 0.33712374581939797,
+      "grad_norm": 0.5162714123725891,
+      "learning_rate": 8e-05,
+      "loss": 1.9312,
+      "step": 6048
+    },
+    {
+      "epoch": 0.3371794871794872,
+      "grad_norm": 0.4670993685722351,
+      "learning_rate": 8e-05,
+      "loss": 1.5798,
+      "step": 6049
+    },
+    {
+      "epoch": 0.3372352285395764,
+      "grad_norm": 0.5087465643882751,
+      "learning_rate": 8e-05,
+      "loss": 1.7965,
+      "step": 6050
+    },
+    {
+      "epoch": 0.33729096989966556,
+      "grad_norm": 0.5819141864776611,
+      "learning_rate": 8e-05,
+      "loss": 1.9564,
+      "step": 6051
+    },
+    {
+      "epoch": 0.33734671125975474,
+      "grad_norm": 0.5214961171150208,
+      "learning_rate": 8e-05,
+      "loss": 1.6992,
+      "step": 6052
+    },
+    {
+      "epoch": 0.3374024526198439,
+      "grad_norm": 0.5660673379898071,
+      "learning_rate": 8e-05,
+      "loss": 1.8737,
+      "step": 6053
+    },
+    {
+      "epoch": 0.3374581939799331,
+      "grad_norm": 0.5306689739227295,
+      "learning_rate": 8e-05,
+      "loss": 1.3794,
+      "step": 6054
+    },
+    {
+      "epoch": 0.3375139353400223,
+      "grad_norm": 0.5800758004188538,
+      "learning_rate": 8e-05,
+      "loss": 1.6946,
+      "step": 6055
+    },
+    {
+      "epoch": 0.3375696767001115,
+      "grad_norm": 0.5740100741386414,
+      "learning_rate": 8e-05,
+      "loss": 2.1177,
+      "step": 6056
+    },
+    {
+      "epoch": 0.3376254180602007,
+      "grad_norm": 0.571855366230011,
+      "learning_rate": 8e-05,
+      "loss": 2.0269,
+      "step": 6057
+    },
+    {
+      "epoch": 0.33768115942028987,
+      "grad_norm": 0.5222740769386292,
+      "learning_rate": 8e-05,
+      "loss": 1.575,
+      "step": 6058
+    },
+    {
+      "epoch": 0.33773690078037905,
+      "grad_norm": 0.46818023920059204,
+      "learning_rate": 8e-05,
+      "loss": 1.3433,
+      "step": 6059
+    },
+    {
+      "epoch": 0.3377926421404682,
+      "grad_norm": 0.5272983312606812,
+      "learning_rate": 8e-05,
+      "loss": 1.4582,
+      "step": 6060
+    },
+    {
+      "epoch": 0.3378483835005574,
+      "grad_norm": 0.5301730632781982,
+      "learning_rate": 8e-05,
+      "loss": 1.7741,
+      "step": 6061
+    },
+    {
+      "epoch": 0.3379041248606466,
+      "grad_norm": 0.5356550812721252,
+      "learning_rate": 8e-05,
+      "loss": 1.7668,
+      "step": 6062
+    },
+    {
+      "epoch": 0.33795986622073576,
+      "grad_norm": 0.5093060731887817,
+      "learning_rate": 8e-05,
+      "loss": 1.7492,
+      "step": 6063
+    },
+    {
+      "epoch": 0.338015607580825,
+      "grad_norm": 0.4938412308692932,
+      "learning_rate": 8e-05,
+      "loss": 1.621,
+      "step": 6064
+    },
+    {
+      "epoch": 0.3380713489409142,
+      "grad_norm": 0.5446529388427734,
+      "learning_rate": 8e-05,
+      "loss": 1.7603,
+      "step": 6065
+    },
+    {
+      "epoch": 0.33812709030100335,
+      "grad_norm": 0.5054848194122314,
+      "learning_rate": 8e-05,
+      "loss": 1.4763,
+      "step": 6066
+    },
+    {
+      "epoch": 0.33818283166109253,
+      "grad_norm": 0.48166799545288086,
+      "learning_rate": 8e-05,
+      "loss": 1.4409,
+      "step": 6067
+    },
+    {
+      "epoch": 0.3382385730211817,
+      "grad_norm": 0.5645819306373596,
+      "learning_rate": 8e-05,
+      "loss": 1.5985,
+      "step": 6068
+    },
+    {
+      "epoch": 0.3382943143812709,
+      "grad_norm": 0.5451626181602478,
+      "learning_rate": 8e-05,
+      "loss": 1.7397,
+      "step": 6069
+    },
+    {
+      "epoch": 0.33835005574136007,
+      "grad_norm": 0.4711497128009796,
+      "learning_rate": 8e-05,
+      "loss": 1.577,
+      "step": 6070
+    },
+    {
+      "epoch": 0.3384057971014493,
+      "grad_norm": 0.47993114590644836,
+      "learning_rate": 8e-05,
+      "loss": 1.5361,
+      "step": 6071
+    },
+    {
+      "epoch": 0.3384615384615385,
+      "grad_norm": 0.5799311995506287,
+      "learning_rate": 8e-05,
+      "loss": 1.7867,
+      "step": 6072
+    },
+    {
+      "epoch": 0.33851727982162766,
+      "grad_norm": 0.542795717716217,
+      "learning_rate": 8e-05,
+      "loss": 1.7606,
+      "step": 6073
+    },
+    {
+      "epoch": 0.33857302118171684,
+      "grad_norm": 0.5645315051078796,
+      "learning_rate": 8e-05,
+      "loss": 1.9333,
+      "step": 6074
+    },
+    {
+      "epoch": 0.338628762541806,
+      "grad_norm": 0.50261390209198,
+      "learning_rate": 8e-05,
+      "loss": 1.5687,
+      "step": 6075
+    },
+    {
+      "epoch": 0.3386845039018952,
+      "grad_norm": 0.5475636124610901,
+      "learning_rate": 8e-05,
+      "loss": 1.8117,
+      "step": 6076
+    },
+    {
+      "epoch": 0.3387402452619844,
+      "grad_norm": 0.5432072877883911,
+      "learning_rate": 8e-05,
+      "loss": 1.6912,
+      "step": 6077
+    },
+    {
+      "epoch": 0.33879598662207355,
+      "grad_norm": 0.48782506585121155,
+      "learning_rate": 8e-05,
+      "loss": 1.6705,
+      "step": 6078
+    },
+    {
+      "epoch": 0.3388517279821628,
+      "grad_norm": 0.552790641784668,
+      "learning_rate": 8e-05,
+      "loss": 1.941,
+      "step": 6079
+    },
+    {
+      "epoch": 0.33890746934225197,
+      "grad_norm": 0.45601341128349304,
+      "learning_rate": 8e-05,
+      "loss": 1.4809,
+      "step": 6080
+    },
+    {
+      "epoch": 0.33896321070234114,
+      "grad_norm": 0.5363773703575134,
+      "learning_rate": 8e-05,
+      "loss": 1.6767,
+      "step": 6081
+    },
+    {
+      "epoch": 0.3390189520624303,
+      "grad_norm": 0.5174465179443359,
+      "learning_rate": 8e-05,
+      "loss": 1.6438,
+      "step": 6082
+    },
+    {
+      "epoch": 0.3390746934225195,
+      "grad_norm": 0.5154566764831543,
+      "learning_rate": 8e-05,
+      "loss": 1.637,
+      "step": 6083
+    },
+    {
+      "epoch": 0.3391304347826087,
+      "grad_norm": 0.4398244619369507,
+      "learning_rate": 8e-05,
+      "loss": 1.3843,
+      "step": 6084
+    },
+    {
+      "epoch": 0.33918617614269786,
+      "grad_norm": 0.5024697780609131,
+      "learning_rate": 8e-05,
+      "loss": 1.6403,
+      "step": 6085
+    },
+    {
+      "epoch": 0.3392419175027871,
+      "grad_norm": 0.5018582344055176,
+      "learning_rate": 8e-05,
+      "loss": 1.5614,
+      "step": 6086
+    },
+    {
+      "epoch": 0.3392976588628763,
+      "grad_norm": 0.5023205280303955,
+      "learning_rate": 8e-05,
+      "loss": 1.819,
+      "step": 6087
+    },
+    {
+      "epoch": 0.33935340022296545,
+      "grad_norm": 0.5069661140441895,
+      "learning_rate": 8e-05,
+      "loss": 1.6936,
+      "step": 6088
+    },
+    {
+      "epoch": 0.33940914158305463,
+      "grad_norm": 0.49317628145217896,
+      "learning_rate": 8e-05,
+      "loss": 1.4198,
+      "step": 6089
+    },
+    {
+      "epoch": 0.3394648829431438,
+      "grad_norm": 0.4833192825317383,
+      "learning_rate": 8e-05,
+      "loss": 1.6803,
+      "step": 6090
+    },
+    {
+      "epoch": 0.339520624303233,
+      "grad_norm": 0.6212770938873291,
+      "learning_rate": 8e-05,
+      "loss": 2.2949,
+      "step": 6091
+    },
+    {
+      "epoch": 0.33957636566332217,
+      "grad_norm": 0.486361026763916,
+      "learning_rate": 8e-05,
+      "loss": 1.7127,
+      "step": 6092
+    },
+    {
+      "epoch": 0.33963210702341134,
+      "grad_norm": 0.5241421461105347,
+      "learning_rate": 8e-05,
+      "loss": 1.8728,
+      "step": 6093
+    },
+    {
+      "epoch": 0.3396878483835006,
+      "grad_norm": 0.5817832946777344,
+      "learning_rate": 8e-05,
+      "loss": 1.8213,
+      "step": 6094
+    },
+    {
+      "epoch": 0.33974358974358976,
+      "grad_norm": 0.5687096118927002,
+      "learning_rate": 8e-05,
+      "loss": 1.8387,
+      "step": 6095
+    },
+    {
+      "epoch": 0.33979933110367894,
+      "grad_norm": 0.5000899434089661,
+      "learning_rate": 8e-05,
+      "loss": 1.4907,
+      "step": 6096
+    },
+    {
+      "epoch": 0.3398550724637681,
+      "grad_norm": 0.5093439221382141,
+      "learning_rate": 8e-05,
+      "loss": 1.7178,
+      "step": 6097
+    },
+    {
+      "epoch": 0.3399108138238573,
+      "grad_norm": 0.487914115190506,
+      "learning_rate": 8e-05,
+      "loss": 1.4353,
+      "step": 6098
+    },
+    {
+      "epoch": 0.3399665551839465,
+      "grad_norm": 0.5182102918624878,
+      "learning_rate": 8e-05,
+      "loss": 1.833,
+      "step": 6099
+    },
+    {
+      "epoch": 0.34002229654403565,
+      "grad_norm": 0.45939892530441284,
+      "learning_rate": 8e-05,
+      "loss": 1.4572,
+      "step": 6100
+    },
+    {
+      "epoch": 0.3400780379041249,
+      "grad_norm": 0.4695865213871002,
+      "learning_rate": 8e-05,
+      "loss": 1.5965,
+      "step": 6101
+    },
+    {
+      "epoch": 0.34013377926421406,
+      "grad_norm": 0.54413241147995,
+      "learning_rate": 8e-05,
+      "loss": 1.8549,
+      "step": 6102
+    },
+    {
+      "epoch": 0.34018952062430324,
+      "grad_norm": 0.5373137593269348,
+      "learning_rate": 8e-05,
+      "loss": 1.6635,
+      "step": 6103
+    },
+    {
+      "epoch": 0.3402452619843924,
+      "grad_norm": 0.5292257070541382,
+      "learning_rate": 8e-05,
+      "loss": 1.6941,
+      "step": 6104
+    },
+    {
+      "epoch": 0.3403010033444816,
+      "grad_norm": 0.5034290552139282,
+      "learning_rate": 8e-05,
+      "loss": 1.8359,
+      "step": 6105
+    },
+    {
+      "epoch": 0.3403567447045708,
+      "grad_norm": 0.5438308119773865,
+      "learning_rate": 8e-05,
+      "loss": 1.6713,
+      "step": 6106
+    },
+    {
+      "epoch": 0.34041248606465996,
+      "grad_norm": 0.5302205085754395,
+      "learning_rate": 8e-05,
+      "loss": 1.761,
+      "step": 6107
+    },
+    {
+      "epoch": 0.34046822742474914,
+      "grad_norm": 0.5735040903091431,
+      "learning_rate": 8e-05,
+      "loss": 1.8512,
+      "step": 6108
+    },
+    {
+      "epoch": 0.34052396878483837,
+      "grad_norm": 0.542824923992157,
+      "learning_rate": 8e-05,
+      "loss": 1.6337,
+      "step": 6109
+    },
+    {
+      "epoch": 0.34057971014492755,
+      "grad_norm": 0.532397449016571,
+      "learning_rate": 8e-05,
+      "loss": 1.6446,
+      "step": 6110
+    },
+    {
+      "epoch": 0.34063545150501673,
+      "grad_norm": 0.5326183438301086,
+      "learning_rate": 8e-05,
+      "loss": 1.6518,
+      "step": 6111
+    },
+    {
+      "epoch": 0.3406911928651059,
+      "grad_norm": 0.560077965259552,
+      "learning_rate": 8e-05,
+      "loss": 1.736,
+      "step": 6112
+    },
+    {
+      "epoch": 0.3407469342251951,
+      "grad_norm": 0.56466144323349,
+      "learning_rate": 8e-05,
+      "loss": 1.7615,
+      "step": 6113
+    },
+    {
+      "epoch": 0.34080267558528426,
+      "grad_norm": 0.5226140022277832,
+      "learning_rate": 8e-05,
+      "loss": 1.9275,
+      "step": 6114
+    },
+    {
+      "epoch": 0.34085841694537344,
+      "grad_norm": 0.48684361577033997,
+      "learning_rate": 8e-05,
+      "loss": 1.164,
+      "step": 6115
+    },
+    {
+      "epoch": 0.3409141583054627,
+      "grad_norm": 0.5470103621482849,
+      "learning_rate": 8e-05,
+      "loss": 1.9557,
+      "step": 6116
+    },
+    {
+      "epoch": 0.34096989966555186,
+      "grad_norm": 0.4782428741455078,
+      "learning_rate": 8e-05,
+      "loss": 1.5178,
+      "step": 6117
+    },
+    {
+      "epoch": 0.34102564102564104,
+      "grad_norm": 0.49948710203170776,
+      "learning_rate": 8e-05,
+      "loss": 1.6443,
+      "step": 6118
+    },
+    {
+      "epoch": 0.3410813823857302,
+      "grad_norm": 0.5285187363624573,
+      "learning_rate": 8e-05,
+      "loss": 1.8012,
+      "step": 6119
+    },
+    {
+      "epoch": 0.3411371237458194,
+      "grad_norm": 0.5237113237380981,
+      "learning_rate": 8e-05,
+      "loss": 1.6156,
+      "step": 6120
+    },
+    {
+      "epoch": 0.34119286510590857,
+      "grad_norm": 0.5836105346679688,
+      "learning_rate": 8e-05,
+      "loss": 1.7364,
+      "step": 6121
+    },
+    {
+      "epoch": 0.34124860646599775,
+      "grad_norm": 0.5258899927139282,
+      "learning_rate": 8e-05,
+      "loss": 1.6938,
+      "step": 6122
+    },
+    {
+      "epoch": 0.34130434782608693,
+      "grad_norm": 0.5039440393447876,
+      "learning_rate": 8e-05,
+      "loss": 1.6582,
+      "step": 6123
+    },
+    {
+      "epoch": 0.34136008918617616,
+      "grad_norm": 0.5295379161834717,
+      "learning_rate": 8e-05,
+      "loss": 1.7974,
+      "step": 6124
+    },
+    {
+      "epoch": 0.34141583054626534,
+      "grad_norm": 0.5055199861526489,
+      "learning_rate": 8e-05,
+      "loss": 1.6236,
+      "step": 6125
+    },
+    {
+      "epoch": 0.3414715719063545,
+      "grad_norm": 0.516697108745575,
+      "learning_rate": 8e-05,
+      "loss": 1.5589,
+      "step": 6126
+    },
+    {
+      "epoch": 0.3415273132664437,
+      "grad_norm": 0.5174638628959656,
+      "learning_rate": 8e-05,
+      "loss": 1.4963,
+      "step": 6127
+    },
+    {
+      "epoch": 0.3415830546265329,
+      "grad_norm": 0.4991219937801361,
+      "learning_rate": 8e-05,
+      "loss": 1.4234,
+      "step": 6128
+    },
+    {
+      "epoch": 0.34163879598662206,
+      "grad_norm": 0.5199254155158997,
+      "learning_rate": 8e-05,
+      "loss": 1.6689,
+      "step": 6129
+    },
+    {
+      "epoch": 0.34169453734671124,
+      "grad_norm": 0.5277366042137146,
+      "learning_rate": 8e-05,
+      "loss": 1.8916,
+      "step": 6130
+    },
+    {
+      "epoch": 0.34175027870680047,
+      "grad_norm": 0.5037539601325989,
+      "learning_rate": 8e-05,
+      "loss": 1.7343,
+      "step": 6131
+    },
+    {
+      "epoch": 0.34180602006688965,
+      "grad_norm": 0.5122715830802917,
+      "learning_rate": 8e-05,
+      "loss": 1.673,
+      "step": 6132
+    },
+    {
+      "epoch": 0.3418617614269788,
+      "grad_norm": 0.4962614178657532,
+      "learning_rate": 8e-05,
+      "loss": 1.8025,
+      "step": 6133
+    },
+    {
+      "epoch": 0.341917502787068,
+      "grad_norm": 0.48662999272346497,
+      "learning_rate": 8e-05,
+      "loss": 1.5983,
+      "step": 6134
+    },
+    {
+      "epoch": 0.3419732441471572,
+      "grad_norm": 0.5290865302085876,
+      "learning_rate": 8e-05,
+      "loss": 1.6749,
+      "step": 6135
+    },
+    {
+      "epoch": 0.34202898550724636,
+      "grad_norm": 0.4744897782802582,
+      "learning_rate": 8e-05,
+      "loss": 1.452,
+      "step": 6136
+    },
+    {
+      "epoch": 0.34208472686733554,
+      "grad_norm": 0.5886744856834412,
+      "learning_rate": 8e-05,
+      "loss": 1.9132,
+      "step": 6137
+    },
+    {
+      "epoch": 0.3421404682274248,
+      "grad_norm": 0.5426172614097595,
+      "learning_rate": 8e-05,
+      "loss": 1.8144,
+      "step": 6138
+    },
+    {
+      "epoch": 0.34219620958751396,
+      "grad_norm": 0.5077410936355591,
+      "learning_rate": 8e-05,
+      "loss": 1.592,
+      "step": 6139
+    },
+    {
+      "epoch": 0.34225195094760313,
+      "grad_norm": 0.4840519428253174,
+      "learning_rate": 8e-05,
+      "loss": 1.581,
+      "step": 6140
+    },
+    {
+      "epoch": 0.3423076923076923,
+      "grad_norm": 0.5276888608932495,
+      "learning_rate": 8e-05,
+      "loss": 1.8234,
+      "step": 6141
+    },
+    {
+      "epoch": 0.3423634336677815,
+      "grad_norm": 0.5416687726974487,
+      "learning_rate": 8e-05,
+      "loss": 1.8545,
+      "step": 6142
+    },
+    {
+      "epoch": 0.34241917502787067,
+      "grad_norm": 0.48407217860221863,
+      "learning_rate": 8e-05,
+      "loss": 1.7246,
+      "step": 6143
+    },
+    {
+      "epoch": 0.34247491638795985,
+      "grad_norm": 0.5333887934684753,
+      "learning_rate": 8e-05,
+      "loss": 1.7349,
+      "step": 6144
+    },
+    {
+      "epoch": 0.34253065774804903,
+      "grad_norm": 0.49881041049957275,
+      "learning_rate": 8e-05,
+      "loss": 1.7543,
+      "step": 6145
+    },
+    {
+      "epoch": 0.34258639910813826,
+      "grad_norm": 0.5575197339057922,
+      "learning_rate": 8e-05,
+      "loss": 1.8022,
+      "step": 6146
+    },
+    {
+      "epoch": 0.34264214046822744,
+      "grad_norm": 0.5086601972579956,
+      "learning_rate": 8e-05,
+      "loss": 1.7278,
+      "step": 6147
+    },
+    {
+      "epoch": 0.3426978818283166,
+      "grad_norm": 0.5459162592887878,
+      "learning_rate": 8e-05,
+      "loss": 1.7454,
+      "step": 6148
+    },
+    {
+      "epoch": 0.3427536231884058,
+      "grad_norm": 0.5540486574172974,
+      "learning_rate": 8e-05,
+      "loss": 1.9281,
+      "step": 6149
+    },
+    {
+      "epoch": 0.342809364548495,
+      "grad_norm": 0.5860450267791748,
+      "learning_rate": 8e-05,
+      "loss": 1.8497,
+      "step": 6150
+    },
+    {
+      "epoch": 0.34286510590858416,
+      "grad_norm": 0.48789045214653015,
+      "learning_rate": 8e-05,
+      "loss": 1.5751,
+      "step": 6151
+    },
+    {
+      "epoch": 0.34292084726867333,
+      "grad_norm": 0.5314328670501709,
+      "learning_rate": 8e-05,
+      "loss": 1.7379,
+      "step": 6152
+    },
+    {
+      "epoch": 0.34297658862876257,
+      "grad_norm": 0.5092215538024902,
+      "learning_rate": 8e-05,
+      "loss": 1.6208,
+      "step": 6153
+    },
+    {
+      "epoch": 0.34303232998885175,
+      "grad_norm": 0.5262496471405029,
+      "learning_rate": 8e-05,
+      "loss": 1.6981,
+      "step": 6154
+    },
+    {
+      "epoch": 0.3430880713489409,
+      "grad_norm": 0.5319823622703552,
+      "learning_rate": 8e-05,
+      "loss": 1.6881,
+      "step": 6155
+    },
+    {
+      "epoch": 0.3431438127090301,
+      "grad_norm": 0.5669171214103699,
+      "learning_rate": 8e-05,
+      "loss": 1.9915,
+      "step": 6156
+    },
+    {
+      "epoch": 0.3431995540691193,
+      "grad_norm": 0.5114725232124329,
+      "learning_rate": 8e-05,
+      "loss": 1.7512,
+      "step": 6157
+    },
+    {
+      "epoch": 0.34325529542920846,
+      "grad_norm": 0.48004892468452454,
+      "learning_rate": 8e-05,
+      "loss": 1.3144,
+      "step": 6158
+    },
+    {
+      "epoch": 0.34331103678929764,
+      "grad_norm": 0.523747444152832,
+      "learning_rate": 8e-05,
+      "loss": 1.684,
+      "step": 6159
+    },
+    {
+      "epoch": 0.3433667781493868,
+      "grad_norm": 0.522097647190094,
+      "learning_rate": 8e-05,
+      "loss": 1.7017,
+      "step": 6160
+    },
+    {
+      "epoch": 0.34342251950947605,
+      "grad_norm": 0.5229143500328064,
+      "learning_rate": 8e-05,
+      "loss": 1.8869,
+      "step": 6161
+    },
+    {
+      "epoch": 0.34347826086956523,
+      "grad_norm": 0.5871850252151489,
+      "learning_rate": 8e-05,
+      "loss": 1.9267,
+      "step": 6162
+    },
+    {
+      "epoch": 0.3435340022296544,
+      "grad_norm": 0.4945095479488373,
+      "learning_rate": 8e-05,
+      "loss": 1.5654,
+      "step": 6163
+    },
+    {
+      "epoch": 0.3435897435897436,
+      "grad_norm": 0.5154891610145569,
+      "learning_rate": 8e-05,
+      "loss": 1.4761,
+      "step": 6164
+    },
+    {
+      "epoch": 0.34364548494983277,
+      "grad_norm": 0.5434823632240295,
+      "learning_rate": 8e-05,
+      "loss": 1.6896,
+      "step": 6165
+    },
+    {
+      "epoch": 0.34370122630992195,
+      "grad_norm": 0.5289078950881958,
+      "learning_rate": 8e-05,
+      "loss": 1.3661,
+      "step": 6166
+    },
+    {
+      "epoch": 0.3437569676700111,
+      "grad_norm": 0.507692813873291,
+      "learning_rate": 8e-05,
+      "loss": 1.7057,
+      "step": 6167
+    },
+    {
+      "epoch": 0.34381270903010036,
+      "grad_norm": 0.5627875924110413,
+      "learning_rate": 8e-05,
+      "loss": 1.7958,
+      "step": 6168
+    },
+    {
+      "epoch": 0.34386845039018954,
+      "grad_norm": 0.5589924454689026,
+      "learning_rate": 8e-05,
+      "loss": 1.8935,
+      "step": 6169
+    },
+    {
+      "epoch": 0.3439241917502787,
+      "grad_norm": 0.5078762769699097,
+      "learning_rate": 8e-05,
+      "loss": 1.7232,
+      "step": 6170
+    },
+    {
+      "epoch": 0.3439799331103679,
+      "grad_norm": 0.4899272918701172,
+      "learning_rate": 8e-05,
+      "loss": 1.5772,
+      "step": 6171
+    },
+    {
+      "epoch": 0.3440356744704571,
+      "grad_norm": 0.5079790353775024,
+      "learning_rate": 8e-05,
+      "loss": 1.4953,
+      "step": 6172
+    },
+    {
+      "epoch": 0.34409141583054625,
+      "grad_norm": 0.5527381896972656,
+      "learning_rate": 8e-05,
+      "loss": 1.6653,
+      "step": 6173
+    },
+    {
+      "epoch": 0.34414715719063543,
+      "grad_norm": 0.5279654264450073,
+      "learning_rate": 8e-05,
+      "loss": 1.8035,
+      "step": 6174
+    },
+    {
+      "epoch": 0.3442028985507246,
+      "grad_norm": 0.572637677192688,
+      "learning_rate": 8e-05,
+      "loss": 1.7902,
+      "step": 6175
+    },
+    {
+      "epoch": 0.34425863991081385,
+      "grad_norm": 0.5261125564575195,
+      "learning_rate": 8e-05,
+      "loss": 1.8534,
+      "step": 6176
+    },
+    {
+      "epoch": 0.344314381270903,
+      "grad_norm": 0.521367073059082,
+      "learning_rate": 8e-05,
+      "loss": 1.5969,
+      "step": 6177
+    },
+    {
+      "epoch": 0.3443701226309922,
+      "grad_norm": 0.4918879568576813,
+      "learning_rate": 8e-05,
+      "loss": 1.7713,
+      "step": 6178
+    },
+    {
+      "epoch": 0.3444258639910814,
+      "grad_norm": 0.5039586424827576,
+      "learning_rate": 8e-05,
+      "loss": 1.6867,
+      "step": 6179
+    },
+    {
+      "epoch": 0.34448160535117056,
+      "grad_norm": 0.5196630954742432,
+      "learning_rate": 8e-05,
+      "loss": 1.8582,
+      "step": 6180
+    },
+    {
+      "epoch": 0.34453734671125974,
+      "grad_norm": 0.5284832715988159,
+      "learning_rate": 8e-05,
+      "loss": 1.7363,
+      "step": 6181
+    },
+    {
+      "epoch": 0.3445930880713489,
+      "grad_norm": 0.5373867750167847,
+      "learning_rate": 8e-05,
+      "loss": 1.5278,
+      "step": 6182
+    },
+    {
+      "epoch": 0.34464882943143815,
+      "grad_norm": 0.5544673800468445,
+      "learning_rate": 8e-05,
+      "loss": 1.7299,
+      "step": 6183
+    },
+    {
+      "epoch": 0.34470457079152733,
+      "grad_norm": 0.5601715445518494,
+      "learning_rate": 8e-05,
+      "loss": 1.7765,
+      "step": 6184
+    },
+    {
+      "epoch": 0.3447603121516165,
+      "grad_norm": 0.564301073551178,
+      "learning_rate": 8e-05,
+      "loss": 1.649,
+      "step": 6185
+    },
+    {
+      "epoch": 0.3448160535117057,
+      "grad_norm": 0.5370534062385559,
+      "learning_rate": 8e-05,
+      "loss": 2.023,
+      "step": 6186
+    },
+    {
+      "epoch": 0.34487179487179487,
+      "grad_norm": 0.5042305588722229,
+      "learning_rate": 8e-05,
+      "loss": 1.7116,
+      "step": 6187
+    },
+    {
+      "epoch": 0.34492753623188405,
+      "grad_norm": 0.4756479859352112,
+      "learning_rate": 8e-05,
+      "loss": 1.6318,
+      "step": 6188
+    },
+    {
+      "epoch": 0.3449832775919732,
+      "grad_norm": 0.573596715927124,
+      "learning_rate": 8e-05,
+      "loss": 1.8775,
+      "step": 6189
+    },
+    {
+      "epoch": 0.3450390189520624,
+      "grad_norm": 0.5731935501098633,
+      "learning_rate": 8e-05,
+      "loss": 1.5896,
+      "step": 6190
+    },
+    {
+      "epoch": 0.34509476031215164,
+      "grad_norm": 0.48869970440864563,
+      "learning_rate": 8e-05,
+      "loss": 1.7313,
+      "step": 6191
+    },
+    {
+      "epoch": 0.3451505016722408,
+      "grad_norm": 0.5444765686988831,
+      "learning_rate": 8e-05,
+      "loss": 1.7901,
+      "step": 6192
+    },
+    {
+      "epoch": 0.34520624303233,
+      "grad_norm": 0.5213355422019958,
+      "learning_rate": 8e-05,
+      "loss": 1.7404,
+      "step": 6193
+    },
+    {
+      "epoch": 0.3452619843924192,
+      "grad_norm": 0.4946276545524597,
+      "learning_rate": 8e-05,
+      "loss": 1.6513,
+      "step": 6194
+    },
+    {
+      "epoch": 0.34531772575250835,
+      "grad_norm": 0.505746603012085,
+      "learning_rate": 8e-05,
+      "loss": 1.5607,
+      "step": 6195
+    },
+    {
+      "epoch": 0.34537346711259753,
+      "grad_norm": 0.48780500888824463,
+      "learning_rate": 8e-05,
+      "loss": 1.6017,
+      "step": 6196
+    },
+    {
+      "epoch": 0.3454292084726867,
+      "grad_norm": 0.527519941329956,
+      "learning_rate": 8e-05,
+      "loss": 1.5679,
+      "step": 6197
+    },
+    {
+      "epoch": 0.34548494983277594,
+      "grad_norm": 0.5447553992271423,
+      "learning_rate": 8e-05,
+      "loss": 1.8504,
+      "step": 6198
+    },
+    {
+      "epoch": 0.3455406911928651,
+      "grad_norm": 0.505783200263977,
+      "learning_rate": 8e-05,
+      "loss": 1.7171,
+      "step": 6199
+    },
+    {
+      "epoch": 0.3455964325529543,
+      "grad_norm": 0.5543640851974487,
+      "learning_rate": 8e-05,
+      "loss": 1.7813,
+      "step": 6200
+    },
+    {
+      "epoch": 0.3456521739130435,
+      "grad_norm": 0.6018317341804504,
+      "learning_rate": 8e-05,
+      "loss": 1.8848,
+      "step": 6201
+    },
+    {
+      "epoch": 0.34570791527313266,
+      "grad_norm": 0.5265443921089172,
+      "learning_rate": 8e-05,
+      "loss": 1.7686,
+      "step": 6202
+    },
+    {
+      "epoch": 0.34576365663322184,
+      "grad_norm": 0.48967650532722473,
+      "learning_rate": 8e-05,
+      "loss": 1.5091,
+      "step": 6203
+    },
+    {
+      "epoch": 0.345819397993311,
+      "grad_norm": 0.5279107093811035,
+      "learning_rate": 8e-05,
+      "loss": 1.6745,
+      "step": 6204
+    },
+    {
+      "epoch": 0.3458751393534002,
+      "grad_norm": 0.5227549076080322,
+      "learning_rate": 8e-05,
+      "loss": 1.7417,
+      "step": 6205
+    },
+    {
+      "epoch": 0.34593088071348943,
+      "grad_norm": 0.5156877040863037,
+      "learning_rate": 8e-05,
+      "loss": 1.6155,
+      "step": 6206
+    },
+    {
+      "epoch": 0.3459866220735786,
+      "grad_norm": 0.48437830805778503,
+      "learning_rate": 8e-05,
+      "loss": 1.4722,
+      "step": 6207
+    },
+    {
+      "epoch": 0.3460423634336678,
+      "grad_norm": 0.4888782203197479,
+      "learning_rate": 8e-05,
+      "loss": 1.7159,
+      "step": 6208
+    },
+    {
+      "epoch": 0.34609810479375697,
+      "grad_norm": 0.5110427141189575,
+      "learning_rate": 8e-05,
+      "loss": 1.6829,
+      "step": 6209
+    },
+    {
+      "epoch": 0.34615384615384615,
+      "grad_norm": 0.47628217935562134,
+      "learning_rate": 8e-05,
+      "loss": 1.6921,
+      "step": 6210
+    },
+    {
+      "epoch": 0.3462095875139353,
+      "grad_norm": 0.5243502259254456,
+      "learning_rate": 8e-05,
+      "loss": 1.7033,
+      "step": 6211
+    },
+    {
+      "epoch": 0.3462653288740245,
+      "grad_norm": 0.481799453496933,
+      "learning_rate": 8e-05,
+      "loss": 1.6188,
+      "step": 6212
+    },
+    {
+      "epoch": 0.34632107023411374,
+      "grad_norm": 0.4980103075504303,
+      "learning_rate": 8e-05,
+      "loss": 1.6574,
+      "step": 6213
+    },
+    {
+      "epoch": 0.3463768115942029,
+      "grad_norm": 0.5031689405441284,
+      "learning_rate": 8e-05,
+      "loss": 1.632,
+      "step": 6214
+    },
+    {
+      "epoch": 0.3464325529542921,
+      "grad_norm": 0.5591900944709778,
+      "learning_rate": 8e-05,
+      "loss": 1.8869,
+      "step": 6215
+    },
+    {
+      "epoch": 0.3464882943143813,
+      "grad_norm": 0.5446702837944031,
+      "learning_rate": 8e-05,
+      "loss": 1.7415,
+      "step": 6216
+    },
+    {
+      "epoch": 0.34654403567447045,
+      "grad_norm": 0.5175807476043701,
+      "learning_rate": 8e-05,
+      "loss": 1.6912,
+      "step": 6217
+    },
+    {
+      "epoch": 0.34659977703455963,
+      "grad_norm": 0.5340683460235596,
+      "learning_rate": 8e-05,
+      "loss": 1.7797,
+      "step": 6218
+    },
+    {
+      "epoch": 0.3466555183946488,
+      "grad_norm": 0.5005671977996826,
+      "learning_rate": 8e-05,
+      "loss": 1.6165,
+      "step": 6219
+    },
+    {
+      "epoch": 0.346711259754738,
+      "grad_norm": 0.5165536403656006,
+      "learning_rate": 8e-05,
+      "loss": 1.6956,
+      "step": 6220
+    },
+    {
+      "epoch": 0.3467670011148272,
+      "grad_norm": 0.5185880661010742,
+      "learning_rate": 8e-05,
+      "loss": 1.6665,
+      "step": 6221
+    },
+    {
+      "epoch": 0.3468227424749164,
+      "grad_norm": 0.5302398800849915,
+      "learning_rate": 8e-05,
+      "loss": 1.8853,
+      "step": 6222
+    },
+    {
+      "epoch": 0.3468784838350056,
+      "grad_norm": 0.4764886200428009,
+      "learning_rate": 8e-05,
+      "loss": 1.5683,
+      "step": 6223
+    },
+    {
+      "epoch": 0.34693422519509476,
+      "grad_norm": 0.515983521938324,
+      "learning_rate": 8e-05,
+      "loss": 1.6717,
+      "step": 6224
+    },
+    {
+      "epoch": 0.34698996655518394,
+      "grad_norm": 0.5122112035751343,
+      "learning_rate": 8e-05,
+      "loss": 1.7558,
+      "step": 6225
+    },
+    {
+      "epoch": 0.3470457079152731,
+      "grad_norm": 0.5676699876785278,
+      "learning_rate": 8e-05,
+      "loss": 2.0176,
+      "step": 6226
+    },
+    {
+      "epoch": 0.3471014492753623,
+      "grad_norm": 0.5953008532524109,
+      "learning_rate": 8e-05,
+      "loss": 1.8701,
+      "step": 6227
+    },
+    {
+      "epoch": 0.34715719063545153,
+      "grad_norm": 0.5327503681182861,
+      "learning_rate": 8e-05,
+      "loss": 1.7862,
+      "step": 6228
+    },
+    {
+      "epoch": 0.3472129319955407,
+      "grad_norm": 0.514793872833252,
+      "learning_rate": 8e-05,
+      "loss": 1.7198,
+      "step": 6229
+    },
+    {
+      "epoch": 0.3472686733556299,
+      "grad_norm": 0.531711757183075,
+      "learning_rate": 8e-05,
+      "loss": 1.7455,
+      "step": 6230
+    },
+    {
+      "epoch": 0.34732441471571907,
+      "grad_norm": 0.5069992542266846,
+      "learning_rate": 8e-05,
+      "loss": 1.6869,
+      "step": 6231
+    },
+    {
+      "epoch": 0.34738015607580824,
+      "grad_norm": 0.5153529644012451,
+      "learning_rate": 8e-05,
+      "loss": 1.8834,
+      "step": 6232
+    },
+    {
+      "epoch": 0.3474358974358974,
+      "grad_norm": 0.4857705235481262,
+      "learning_rate": 8e-05,
+      "loss": 1.6374,
+      "step": 6233
+    },
+    {
+      "epoch": 0.3474916387959866,
+      "grad_norm": 0.5472427606582642,
+      "learning_rate": 8e-05,
+      "loss": 1.9739,
+      "step": 6234
+    },
+    {
+      "epoch": 0.34754738015607584,
+      "grad_norm": 0.5616702437400818,
+      "learning_rate": 8e-05,
+      "loss": 1.6326,
+      "step": 6235
+    },
+    {
+      "epoch": 0.347603121516165,
+      "grad_norm": 0.5345883369445801,
+      "learning_rate": 8e-05,
+      "loss": 1.7221,
+      "step": 6236
+    },
+    {
+      "epoch": 0.3476588628762542,
+      "grad_norm": 0.48029625415802,
+      "learning_rate": 8e-05,
+      "loss": 1.5286,
+      "step": 6237
+    },
+    {
+      "epoch": 0.34771460423634337,
+      "grad_norm": 0.5803802013397217,
+      "learning_rate": 8e-05,
+      "loss": 1.7394,
+      "step": 6238
+    },
+    {
+      "epoch": 0.34777034559643255,
+      "grad_norm": 0.523873507976532,
+      "learning_rate": 8e-05,
+      "loss": 1.7965,
+      "step": 6239
+    },
+    {
+      "epoch": 0.34782608695652173,
+      "grad_norm": 0.48509761691093445,
+      "learning_rate": 8e-05,
+      "loss": 1.5798,
+      "step": 6240
+    },
+    {
+      "epoch": 0.3478818283166109,
+      "grad_norm": 0.5138911008834839,
+      "learning_rate": 8e-05,
+      "loss": 1.8276,
+      "step": 6241
+    },
+    {
+      "epoch": 0.3479375696767001,
+      "grad_norm": 0.5050959587097168,
+      "learning_rate": 8e-05,
+      "loss": 1.6347,
+      "step": 6242
+    },
+    {
+      "epoch": 0.3479933110367893,
+      "grad_norm": 0.4868180751800537,
+      "learning_rate": 8e-05,
+      "loss": 1.561,
+      "step": 6243
+    },
+    {
+      "epoch": 0.3480490523968785,
+      "grad_norm": 0.48366305232048035,
+      "learning_rate": 8e-05,
+      "loss": 1.8501,
+      "step": 6244
+    },
+    {
+      "epoch": 0.3481047937569677,
+      "grad_norm": 0.5158886313438416,
+      "learning_rate": 8e-05,
+      "loss": 1.7387,
+      "step": 6245
+    },
+    {
+      "epoch": 0.34816053511705686,
+      "grad_norm": 0.516280472278595,
+      "learning_rate": 8e-05,
+      "loss": 1.72,
+      "step": 6246
+    },
+    {
+      "epoch": 0.34821627647714604,
+      "grad_norm": 0.5243182182312012,
+      "learning_rate": 8e-05,
+      "loss": 1.8727,
+      "step": 6247
+    },
+    {
+      "epoch": 0.3482720178372352,
+      "grad_norm": 0.5397647619247437,
+      "learning_rate": 8e-05,
+      "loss": 1.5581,
+      "step": 6248
+    },
+    {
+      "epoch": 0.3483277591973244,
+      "grad_norm": 0.5544797778129578,
+      "learning_rate": 8e-05,
+      "loss": 1.5943,
+      "step": 6249
+    },
+    {
+      "epoch": 0.34838350055741363,
+      "grad_norm": 0.5458196997642517,
+      "learning_rate": 8e-05,
+      "loss": 1.6336,
+      "step": 6250
+    },
+    {
+      "epoch": 0.3484392419175028,
+      "grad_norm": 0.5505799055099487,
+      "learning_rate": 8e-05,
+      "loss": 1.8118,
+      "step": 6251
+    },
+    {
+      "epoch": 0.348494983277592,
+      "grad_norm": 0.5446839928627014,
+      "learning_rate": 8e-05,
+      "loss": 1.8311,
+      "step": 6252
+    },
+    {
+      "epoch": 0.34855072463768116,
+      "grad_norm": 0.5084923505783081,
+      "learning_rate": 8e-05,
+      "loss": 1.6369,
+      "step": 6253
+    },
+    {
+      "epoch": 0.34860646599777034,
+      "grad_norm": 0.5445770025253296,
+      "learning_rate": 8e-05,
+      "loss": 1.7987,
+      "step": 6254
+    },
+    {
+      "epoch": 0.3486622073578595,
+      "grad_norm": 0.5154903531074524,
+      "learning_rate": 8e-05,
+      "loss": 1.6847,
+      "step": 6255
+    },
+    {
+      "epoch": 0.3487179487179487,
+      "grad_norm": 0.5390969514846802,
+      "learning_rate": 8e-05,
+      "loss": 1.8958,
+      "step": 6256
+    },
+    {
+      "epoch": 0.3487736900780379,
+      "grad_norm": 0.496901273727417,
+      "learning_rate": 8e-05,
+      "loss": 1.7611,
+      "step": 6257
+    },
+    {
+      "epoch": 0.3488294314381271,
+      "grad_norm": 0.4763997495174408,
+      "learning_rate": 8e-05,
+      "loss": 1.46,
+      "step": 6258
+    },
+    {
+      "epoch": 0.3488851727982163,
+      "grad_norm": 0.5262258052825928,
+      "learning_rate": 8e-05,
+      "loss": 1.8513,
+      "step": 6259
+    },
+    {
+      "epoch": 0.34894091415830547,
+      "grad_norm": 0.5122326016426086,
+      "learning_rate": 8e-05,
+      "loss": 1.6108,
+      "step": 6260
+    },
+    {
+      "epoch": 0.34899665551839465,
+      "grad_norm": 0.555101752281189,
+      "learning_rate": 8e-05,
+      "loss": 1.8278,
+      "step": 6261
+    },
+    {
+      "epoch": 0.34905239687848383,
+      "grad_norm": 0.5074917078018188,
+      "learning_rate": 8e-05,
+      "loss": 1.7252,
+      "step": 6262
+    },
+    {
+      "epoch": 0.349108138238573,
+      "grad_norm": 0.5135613083839417,
+      "learning_rate": 8e-05,
+      "loss": 1.6203,
+      "step": 6263
+    },
+    {
+      "epoch": 0.3491638795986622,
+      "grad_norm": 0.5522807836532593,
+      "learning_rate": 8e-05,
+      "loss": 1.8089,
+      "step": 6264
+    },
+    {
+      "epoch": 0.3492196209587514,
+      "grad_norm": 0.5222398042678833,
+      "learning_rate": 8e-05,
+      "loss": 1.788,
+      "step": 6265
+    },
+    {
+      "epoch": 0.3492753623188406,
+      "grad_norm": 0.5117469429969788,
+      "learning_rate": 8e-05,
+      "loss": 1.6157,
+      "step": 6266
+    },
+    {
+      "epoch": 0.3493311036789298,
+      "grad_norm": 0.5744069218635559,
+      "learning_rate": 8e-05,
+      "loss": 1.8723,
+      "step": 6267
+    },
+    {
+      "epoch": 0.34938684503901896,
+      "grad_norm": 0.5354539155960083,
+      "learning_rate": 8e-05,
+      "loss": 1.6102,
+      "step": 6268
+    },
+    {
+      "epoch": 0.34944258639910813,
+      "grad_norm": 0.6402024626731873,
+      "learning_rate": 8e-05,
+      "loss": 1.7213,
+      "step": 6269
+    },
+    {
+      "epoch": 0.3494983277591973,
+      "grad_norm": 0.4755125641822815,
+      "learning_rate": 8e-05,
+      "loss": 1.6028,
+      "step": 6270
+    },
+    {
+      "epoch": 0.3495540691192865,
+      "grad_norm": 0.5351192951202393,
+      "learning_rate": 8e-05,
+      "loss": 1.6167,
+      "step": 6271
+    },
+    {
+      "epoch": 0.34960981047937567,
+      "grad_norm": 0.5391672253608704,
+      "learning_rate": 8e-05,
+      "loss": 1.7226,
+      "step": 6272
+    },
+    {
+      "epoch": 0.3496655518394649,
+      "grad_norm": 0.5133790969848633,
+      "learning_rate": 8e-05,
+      "loss": 1.7819,
+      "step": 6273
+    },
+    {
+      "epoch": 0.3497212931995541,
+      "grad_norm": 0.5002404451370239,
+      "learning_rate": 8e-05,
+      "loss": 1.5547,
+      "step": 6274
+    },
+    {
+      "epoch": 0.34977703455964326,
+      "grad_norm": 0.4975956380367279,
+      "learning_rate": 8e-05,
+      "loss": 1.3753,
+      "step": 6275
+    },
+    {
+      "epoch": 0.34983277591973244,
+      "grad_norm": 0.5085654854774475,
+      "learning_rate": 8e-05,
+      "loss": 1.5977,
+      "step": 6276
+    },
+    {
+      "epoch": 0.3498885172798216,
+      "grad_norm": 0.5318102836608887,
+      "learning_rate": 8e-05,
+      "loss": 1.6468,
+      "step": 6277
+    },
+    {
+      "epoch": 0.3499442586399108,
+      "grad_norm": 0.4818606376647949,
+      "learning_rate": 8e-05,
+      "loss": 1.5751,
+      "step": 6278
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5304341316223145,
+      "learning_rate": 8e-05,
+      "loss": 1.6784,
+      "step": 6279
+    },
+    {
+      "epoch": 0.3500557413600892,
+      "grad_norm": 0.5689952969551086,
+      "learning_rate": 8e-05,
+      "loss": 1.9105,
+      "step": 6280
+    },
+    {
+      "epoch": 0.3501114827201784,
+      "grad_norm": 0.5430139303207397,
+      "learning_rate": 8e-05,
+      "loss": 1.8026,
+      "step": 6281
+    },
+    {
+      "epoch": 0.35016722408026757,
+      "grad_norm": 0.49770841002464294,
+      "learning_rate": 8e-05,
+      "loss": 1.5024,
+      "step": 6282
+    },
+    {
+      "epoch": 0.35022296544035675,
+      "grad_norm": 0.5617845058441162,
+      "learning_rate": 8e-05,
+      "loss": 1.8061,
+      "step": 6283
+    },
+    {
+      "epoch": 0.3502787068004459,
+      "grad_norm": 0.5018987655639648,
+      "learning_rate": 8e-05,
+      "loss": 1.6195,
+      "step": 6284
+    },
+    {
+      "epoch": 0.3503344481605351,
+      "grad_norm": 0.5285188555717468,
+      "learning_rate": 8e-05,
+      "loss": 1.6761,
+      "step": 6285
+    },
+    {
+      "epoch": 0.3503901895206243,
+      "grad_norm": 0.5185275077819824,
+      "learning_rate": 8e-05,
+      "loss": 1.6678,
+      "step": 6286
+    },
+    {
+      "epoch": 0.35044593088071346,
+      "grad_norm": 0.5226163864135742,
+      "learning_rate": 8e-05,
+      "loss": 1.709,
+      "step": 6287
+    },
+    {
+      "epoch": 0.3505016722408027,
+      "grad_norm": 0.520292341709137,
+      "learning_rate": 8e-05,
+      "loss": 1.7416,
+      "step": 6288
+    },
+    {
+      "epoch": 0.3505574136008919,
+      "grad_norm": 0.5003277063369751,
+      "learning_rate": 8e-05,
+      "loss": 1.6716,
+      "step": 6289
+    },
+    {
+      "epoch": 0.35061315496098105,
+      "grad_norm": 0.5214517116546631,
+      "learning_rate": 8e-05,
+      "loss": 1.6144,
+      "step": 6290
+    },
+    {
+      "epoch": 0.35066889632107023,
+      "grad_norm": 0.4936576783657074,
+      "learning_rate": 8e-05,
+      "loss": 1.5691,
+      "step": 6291
+    },
+    {
+      "epoch": 0.3507246376811594,
+      "grad_norm": 0.5251075029373169,
+      "learning_rate": 8e-05,
+      "loss": 1.7951,
+      "step": 6292
+    },
+    {
+      "epoch": 0.3507803790412486,
+      "grad_norm": 0.4800378382205963,
+      "learning_rate": 8e-05,
+      "loss": 1.613,
+      "step": 6293
+    },
+    {
+      "epoch": 0.35083612040133777,
+      "grad_norm": 0.4930437207221985,
+      "learning_rate": 8e-05,
+      "loss": 1.5607,
+      "step": 6294
+    },
+    {
+      "epoch": 0.350891861761427,
+      "grad_norm": 0.5461827516555786,
+      "learning_rate": 8e-05,
+      "loss": 1.7602,
+      "step": 6295
+    },
+    {
+      "epoch": 0.3509476031215162,
+      "grad_norm": 0.4616333544254303,
+      "learning_rate": 8e-05,
+      "loss": 1.4898,
+      "step": 6296
+    },
+    {
+      "epoch": 0.35100334448160536,
+      "grad_norm": 0.5303559899330139,
+      "learning_rate": 8e-05,
+      "loss": 1.7979,
+      "step": 6297
+    },
+    {
+      "epoch": 0.35105908584169454,
+      "grad_norm": 0.5381184220314026,
+      "learning_rate": 8e-05,
+      "loss": 1.7486,
+      "step": 6298
+    },
+    {
+      "epoch": 0.3511148272017837,
+      "grad_norm": 0.5684160590171814,
+      "learning_rate": 8e-05,
+      "loss": 1.8218,
+      "step": 6299
+    },
+    {
+      "epoch": 0.3511705685618729,
+      "grad_norm": 0.5133114457130432,
+      "learning_rate": 8e-05,
+      "loss": 1.6521,
+      "step": 6300
+    },
+    {
+      "epoch": 0.3512263099219621,
+      "grad_norm": 0.6434251070022583,
+      "learning_rate": 8e-05,
+      "loss": 2.0408,
+      "step": 6301
+    },
+    {
+      "epoch": 0.35128205128205126,
+      "grad_norm": 0.45654892921447754,
+      "learning_rate": 8e-05,
+      "loss": 1.5327,
+      "step": 6302
+    },
+    {
+      "epoch": 0.3513377926421405,
+      "grad_norm": 0.5252170562744141,
+      "learning_rate": 8e-05,
+      "loss": 1.7286,
+      "step": 6303
+    },
+    {
+      "epoch": 0.35139353400222967,
+      "grad_norm": 0.5745198726654053,
+      "learning_rate": 8e-05,
+      "loss": 1.8548,
+      "step": 6304
+    },
+    {
+      "epoch": 0.35144927536231885,
+      "grad_norm": 0.5161747336387634,
+      "learning_rate": 8e-05,
+      "loss": 1.7415,
+      "step": 6305
+    },
+    {
+      "epoch": 0.351505016722408,
+      "grad_norm": 0.47877398133277893,
+      "learning_rate": 8e-05,
+      "loss": 1.6255,
+      "step": 6306
+    },
+    {
+      "epoch": 0.3515607580824972,
+      "grad_norm": 0.5062041878700256,
+      "learning_rate": 8e-05,
+      "loss": 1.5955,
+      "step": 6307
+    },
+    {
+      "epoch": 0.3516164994425864,
+      "grad_norm": 0.5496783256530762,
+      "learning_rate": 8e-05,
+      "loss": 1.9515,
+      "step": 6308
+    },
+    {
+      "epoch": 0.35167224080267556,
+      "grad_norm": 0.4971919655799866,
+      "learning_rate": 8e-05,
+      "loss": 1.6802,
+      "step": 6309
+    },
+    {
+      "epoch": 0.3517279821627648,
+      "grad_norm": 0.5056297779083252,
+      "learning_rate": 8e-05,
+      "loss": 1.6415,
+      "step": 6310
+    },
+    {
+      "epoch": 0.351783723522854,
+      "grad_norm": 0.5026938319206238,
+      "learning_rate": 8e-05,
+      "loss": 1.6872,
+      "step": 6311
+    },
+    {
+      "epoch": 0.35183946488294315,
+      "grad_norm": 0.5117358565330505,
+      "learning_rate": 8e-05,
+      "loss": 1.4986,
+      "step": 6312
+    },
+    {
+      "epoch": 0.35189520624303233,
+      "grad_norm": 0.49614956974983215,
+      "learning_rate": 8e-05,
+      "loss": 1.6072,
+      "step": 6313
+    },
+    {
+      "epoch": 0.3519509476031215,
+      "grad_norm": 0.5110148191452026,
+      "learning_rate": 8e-05,
+      "loss": 1.718,
+      "step": 6314
+    },
+    {
+      "epoch": 0.3520066889632107,
+      "grad_norm": 0.5317930579185486,
+      "learning_rate": 8e-05,
+      "loss": 1.631,
+      "step": 6315
+    },
+    {
+      "epoch": 0.35206243032329987,
+      "grad_norm": 0.553537130355835,
+      "learning_rate": 8e-05,
+      "loss": 1.7386,
+      "step": 6316
+    },
+    {
+      "epoch": 0.35211817168338905,
+      "grad_norm": 0.5011419653892517,
+      "learning_rate": 8e-05,
+      "loss": 1.572,
+      "step": 6317
+    },
+    {
+      "epoch": 0.3521739130434783,
+      "grad_norm": 0.5681989192962646,
+      "learning_rate": 8e-05,
+      "loss": 2.0663,
+      "step": 6318
+    },
+    {
+      "epoch": 0.35222965440356746,
+      "grad_norm": 0.5317137241363525,
+      "learning_rate": 8e-05,
+      "loss": 1.5384,
+      "step": 6319
+    },
+    {
+      "epoch": 0.35228539576365664,
+      "grad_norm": 0.4983222484588623,
+      "learning_rate": 8e-05,
+      "loss": 1.6649,
+      "step": 6320
+    },
+    {
+      "epoch": 0.3523411371237458,
+      "grad_norm": 0.5296564102172852,
+      "learning_rate": 8e-05,
+      "loss": 1.7091,
+      "step": 6321
+    },
+    {
+      "epoch": 0.352396878483835,
+      "grad_norm": 0.49374499917030334,
+      "learning_rate": 8e-05,
+      "loss": 1.4785,
+      "step": 6322
+    },
+    {
+      "epoch": 0.3524526198439242,
+      "grad_norm": 0.490217924118042,
+      "learning_rate": 8e-05,
+      "loss": 1.6052,
+      "step": 6323
+    },
+    {
+      "epoch": 0.35250836120401335,
+      "grad_norm": 0.5159461498260498,
+      "learning_rate": 8e-05,
+      "loss": 1.61,
+      "step": 6324
+    },
+    {
+      "epoch": 0.3525641025641026,
+      "grad_norm": 0.5397430658340454,
+      "learning_rate": 8e-05,
+      "loss": 1.8227,
+      "step": 6325
+    },
+    {
+      "epoch": 0.35261984392419177,
+      "grad_norm": 0.5083862543106079,
+      "learning_rate": 8e-05,
+      "loss": 1.6645,
+      "step": 6326
+    },
+    {
+      "epoch": 0.35267558528428095,
+      "grad_norm": 0.5055984854698181,
+      "learning_rate": 8e-05,
+      "loss": 1.8023,
+      "step": 6327
+    },
+    {
+      "epoch": 0.3527313266443701,
+      "grad_norm": 0.5376937389373779,
+      "learning_rate": 8e-05,
+      "loss": 1.8579,
+      "step": 6328
+    },
+    {
+      "epoch": 0.3527870680044593,
+      "grad_norm": 0.49673518538475037,
+      "learning_rate": 8e-05,
+      "loss": 1.6094,
+      "step": 6329
+    },
+    {
+      "epoch": 0.3528428093645485,
+      "grad_norm": 0.47112712264060974,
+      "learning_rate": 8e-05,
+      "loss": 1.6966,
+      "step": 6330
+    },
+    {
+      "epoch": 0.35289855072463766,
+      "grad_norm": 0.5200251340866089,
+      "learning_rate": 8e-05,
+      "loss": 1.8557,
+      "step": 6331
+    },
+    {
+      "epoch": 0.3529542920847269,
+      "grad_norm": 0.5550200939178467,
+      "learning_rate": 8e-05,
+      "loss": 1.9868,
+      "step": 6332
+    },
+    {
+      "epoch": 0.3530100334448161,
+      "grad_norm": 0.4754127860069275,
+      "learning_rate": 8e-05,
+      "loss": 1.441,
+      "step": 6333
+    },
+    {
+      "epoch": 0.35306577480490525,
+      "grad_norm": 0.5317299962043762,
+      "learning_rate": 8e-05,
+      "loss": 1.7727,
+      "step": 6334
+    },
+    {
+      "epoch": 0.35312151616499443,
+      "grad_norm": 0.5096808075904846,
+      "learning_rate": 8e-05,
+      "loss": 1.7812,
+      "step": 6335
+    },
+    {
+      "epoch": 0.3531772575250836,
+      "grad_norm": 0.5273366570472717,
+      "learning_rate": 8e-05,
+      "loss": 1.6957,
+      "step": 6336
+    },
+    {
+      "epoch": 0.3532329988851728,
+      "grad_norm": 0.5220344662666321,
+      "learning_rate": 8e-05,
+      "loss": 1.6494,
+      "step": 6337
+    },
+    {
+      "epoch": 0.35328874024526197,
+      "grad_norm": 0.5042638778686523,
+      "learning_rate": 8e-05,
+      "loss": 1.6709,
+      "step": 6338
+    },
+    {
+      "epoch": 0.35334448160535115,
+      "grad_norm": 0.5413642525672913,
+      "learning_rate": 8e-05,
+      "loss": 1.7465,
+      "step": 6339
+    },
+    {
+      "epoch": 0.3534002229654404,
+      "grad_norm": 0.4990393817424774,
+      "learning_rate": 8e-05,
+      "loss": 1.6118,
+      "step": 6340
+    },
+    {
+      "epoch": 0.35345596432552956,
+      "grad_norm": 0.5102233290672302,
+      "learning_rate": 8e-05,
+      "loss": 1.6453,
+      "step": 6341
+    },
+    {
+      "epoch": 0.35351170568561874,
+      "grad_norm": 0.4874460995197296,
+      "learning_rate": 8e-05,
+      "loss": 1.4917,
+      "step": 6342
+    },
+    {
+      "epoch": 0.3535674470457079,
+      "grad_norm": 0.562700092792511,
+      "learning_rate": 8e-05,
+      "loss": 1.9172,
+      "step": 6343
+    },
+    {
+      "epoch": 0.3536231884057971,
+      "grad_norm": 0.5399877429008484,
+      "learning_rate": 8e-05,
+      "loss": 1.5813,
+      "step": 6344
+    },
+    {
+      "epoch": 0.3536789297658863,
+      "grad_norm": 0.45111870765686035,
+      "learning_rate": 8e-05,
+      "loss": 1.0555,
+      "step": 6345
+    },
+    {
+      "epoch": 0.35373467112597545,
+      "grad_norm": 0.49734896421432495,
+      "learning_rate": 8e-05,
+      "loss": 1.7715,
+      "step": 6346
+    },
+    {
+      "epoch": 0.3537904124860647,
+      "grad_norm": 0.5130085945129395,
+      "learning_rate": 8e-05,
+      "loss": 1.6804,
+      "step": 6347
+    },
+    {
+      "epoch": 0.35384615384615387,
+      "grad_norm": 0.5188106894493103,
+      "learning_rate": 8e-05,
+      "loss": 1.8666,
+      "step": 6348
+    },
+    {
+      "epoch": 0.35390189520624304,
+      "grad_norm": 0.5795222520828247,
+      "learning_rate": 8e-05,
+      "loss": 1.6269,
+      "step": 6349
+    },
+    {
+      "epoch": 0.3539576365663322,
+      "grad_norm": 0.48987892270088196,
+      "learning_rate": 8e-05,
+      "loss": 1.4014,
+      "step": 6350
+    },
+    {
+      "epoch": 0.3540133779264214,
+      "grad_norm": 0.5103864669799805,
+      "learning_rate": 8e-05,
+      "loss": 1.5583,
+      "step": 6351
+    },
+    {
+      "epoch": 0.3540691192865106,
+      "grad_norm": 0.5661724805831909,
+      "learning_rate": 8e-05,
+      "loss": 1.877,
+      "step": 6352
+    },
+    {
+      "epoch": 0.35412486064659976,
+      "grad_norm": 0.5406143665313721,
+      "learning_rate": 8e-05,
+      "loss": 1.8725,
+      "step": 6353
+    },
+    {
+      "epoch": 0.35418060200668894,
+      "grad_norm": 0.5083308219909668,
+      "learning_rate": 8e-05,
+      "loss": 1.5352,
+      "step": 6354
+    },
+    {
+      "epoch": 0.3542363433667782,
+      "grad_norm": 0.48224666714668274,
+      "learning_rate": 8e-05,
+      "loss": 1.7073,
+      "step": 6355
+    },
+    {
+      "epoch": 0.35429208472686735,
+      "grad_norm": 0.5201107263565063,
+      "learning_rate": 8e-05,
+      "loss": 1.7611,
+      "step": 6356
+    },
+    {
+      "epoch": 0.35434782608695653,
+      "grad_norm": 0.5390413999557495,
+      "learning_rate": 8e-05,
+      "loss": 1.8358,
+      "step": 6357
+    },
+    {
+      "epoch": 0.3544035674470457,
+      "grad_norm": 0.49942028522491455,
+      "learning_rate": 8e-05,
+      "loss": 1.6606,
+      "step": 6358
+    },
+    {
+      "epoch": 0.3544593088071349,
+      "grad_norm": 0.5088225603103638,
+      "learning_rate": 8e-05,
+      "loss": 1.6044,
+      "step": 6359
+    },
+    {
+      "epoch": 0.35451505016722407,
+      "grad_norm": 0.5360525250434875,
+      "learning_rate": 8e-05,
+      "loss": 1.756,
+      "step": 6360
+    },
+    {
+      "epoch": 0.35457079152731324,
+      "grad_norm": 0.5383181571960449,
+      "learning_rate": 8e-05,
+      "loss": 1.726,
+      "step": 6361
+    },
+    {
+      "epoch": 0.3546265328874025,
+      "grad_norm": 0.5377655029296875,
+      "learning_rate": 8e-05,
+      "loss": 1.8818,
+      "step": 6362
+    },
+    {
+      "epoch": 0.35468227424749166,
+      "grad_norm": 0.48970893025398254,
+      "learning_rate": 8e-05,
+      "loss": 1.7495,
+      "step": 6363
+    },
+    {
+      "epoch": 0.35473801560758084,
+      "grad_norm": 0.4807761609554291,
+      "learning_rate": 8e-05,
+      "loss": 1.6249,
+      "step": 6364
+    },
+    {
+      "epoch": 0.35479375696767,
+      "grad_norm": 0.4989910125732422,
+      "learning_rate": 8e-05,
+      "loss": 1.6189,
+      "step": 6365
+    },
+    {
+      "epoch": 0.3548494983277592,
+      "grad_norm": 0.4965146780014038,
+      "learning_rate": 8e-05,
+      "loss": 1.5913,
+      "step": 6366
+    },
+    {
+      "epoch": 0.3549052396878484,
+      "grad_norm": 0.5162652134895325,
+      "learning_rate": 8e-05,
+      "loss": 1.8145,
+      "step": 6367
+    },
+    {
+      "epoch": 0.35496098104793755,
+      "grad_norm": 0.5167195200920105,
+      "learning_rate": 8e-05,
+      "loss": 1.7582,
+      "step": 6368
+    },
+    {
+      "epoch": 0.35501672240802673,
+      "grad_norm": 0.5981922149658203,
+      "learning_rate": 8e-05,
+      "loss": 1.7429,
+      "step": 6369
+    },
+    {
+      "epoch": 0.35507246376811596,
+      "grad_norm": 0.49923574924468994,
+      "learning_rate": 8e-05,
+      "loss": 1.6594,
+      "step": 6370
+    },
+    {
+      "epoch": 0.35512820512820514,
+      "grad_norm": 0.5337264537811279,
+      "learning_rate": 8e-05,
+      "loss": 1.7549,
+      "step": 6371
+    },
+    {
+      "epoch": 0.3551839464882943,
+      "grad_norm": 0.5021322965621948,
+      "learning_rate": 8e-05,
+      "loss": 1.7241,
+      "step": 6372
+    },
+    {
+      "epoch": 0.3552396878483835,
+      "grad_norm": 0.48789548873901367,
+      "learning_rate": 8e-05,
+      "loss": 1.4845,
+      "step": 6373
+    },
+    {
+      "epoch": 0.3552954292084727,
+      "grad_norm": 0.49966785311698914,
+      "learning_rate": 8e-05,
+      "loss": 1.809,
+      "step": 6374
+    },
+    {
+      "epoch": 0.35535117056856186,
+      "grad_norm": 0.5539280772209167,
+      "learning_rate": 8e-05,
+      "loss": 1.9406,
+      "step": 6375
+    },
+    {
+      "epoch": 0.35540691192865104,
+      "grad_norm": 0.5110356211662292,
+      "learning_rate": 8e-05,
+      "loss": 1.7554,
+      "step": 6376
+    },
+    {
+      "epoch": 0.35546265328874027,
+      "grad_norm": 0.5013741850852966,
+      "learning_rate": 8e-05,
+      "loss": 1.6376,
+      "step": 6377
+    },
+    {
+      "epoch": 0.35551839464882945,
+      "grad_norm": 0.5059554576873779,
+      "learning_rate": 8e-05,
+      "loss": 1.5147,
+      "step": 6378
+    },
+    {
+      "epoch": 0.35557413600891863,
+      "grad_norm": 0.5380630493164062,
+      "learning_rate": 8e-05,
+      "loss": 1.6972,
+      "step": 6379
+    },
+    {
+      "epoch": 0.3556298773690078,
+      "grad_norm": 0.5238441824913025,
+      "learning_rate": 8e-05,
+      "loss": 2.0016,
+      "step": 6380
+    },
+    {
+      "epoch": 0.355685618729097,
+      "grad_norm": 0.5427252650260925,
+      "learning_rate": 8e-05,
+      "loss": 1.8643,
+      "step": 6381
+    },
+    {
+      "epoch": 0.35574136008918616,
+      "grad_norm": 0.5313915610313416,
+      "learning_rate": 8e-05,
+      "loss": 1.9013,
+      "step": 6382
+    },
+    {
+      "epoch": 0.35579710144927534,
+      "grad_norm": 0.4783153235912323,
+      "learning_rate": 8e-05,
+      "loss": 1.6119,
+      "step": 6383
+    },
+    {
+      "epoch": 0.3558528428093645,
+      "grad_norm": 0.506710410118103,
+      "learning_rate": 8e-05,
+      "loss": 1.7333,
+      "step": 6384
+    },
+    {
+      "epoch": 0.35590858416945376,
+      "grad_norm": 0.552340567111969,
+      "learning_rate": 8e-05,
+      "loss": 1.8149,
+      "step": 6385
+    },
+    {
+      "epoch": 0.35596432552954294,
+      "grad_norm": 0.49930626153945923,
+      "learning_rate": 8e-05,
+      "loss": 1.4079,
+      "step": 6386
+    },
+    {
+      "epoch": 0.3560200668896321,
+      "grad_norm": 0.4994630515575409,
+      "learning_rate": 8e-05,
+      "loss": 1.6608,
+      "step": 6387
+    },
+    {
+      "epoch": 0.3560758082497213,
+      "grad_norm": 0.5089203119277954,
+      "learning_rate": 8e-05,
+      "loss": 1.5017,
+      "step": 6388
+    },
+    {
+      "epoch": 0.35613154960981047,
+      "grad_norm": 0.5555450320243835,
+      "learning_rate": 8e-05,
+      "loss": 1.7565,
+      "step": 6389
+    },
+    {
+      "epoch": 0.35618729096989965,
+      "grad_norm": 0.4765130281448364,
+      "learning_rate": 8e-05,
+      "loss": 1.7225,
+      "step": 6390
+    },
+    {
+      "epoch": 0.35624303232998883,
+      "grad_norm": 0.5181832313537598,
+      "learning_rate": 8e-05,
+      "loss": 1.6676,
+      "step": 6391
+    },
+    {
+      "epoch": 0.35629877369007806,
+      "grad_norm": 0.553586483001709,
+      "learning_rate": 8e-05,
+      "loss": 1.7953,
+      "step": 6392
+    },
+    {
+      "epoch": 0.35635451505016724,
+      "grad_norm": 0.5278187394142151,
+      "learning_rate": 8e-05,
+      "loss": 1.7932,
+      "step": 6393
+    },
+    {
+      "epoch": 0.3564102564102564,
+      "grad_norm": 0.5500335693359375,
+      "learning_rate": 8e-05,
+      "loss": 1.6132,
+      "step": 6394
+    },
+    {
+      "epoch": 0.3564659977703456,
+      "grad_norm": 0.5639261603355408,
+      "learning_rate": 8e-05,
+      "loss": 1.7654,
+      "step": 6395
+    },
+    {
+      "epoch": 0.3565217391304348,
+      "grad_norm": 0.5675365924835205,
+      "learning_rate": 8e-05,
+      "loss": 1.897,
+      "step": 6396
+    },
+    {
+      "epoch": 0.35657748049052396,
+      "grad_norm": 0.5094690322875977,
+      "learning_rate": 8e-05,
+      "loss": 1.5559,
+      "step": 6397
+    },
+    {
+      "epoch": 0.35663322185061314,
+      "grad_norm": 0.5150576829910278,
+      "learning_rate": 8e-05,
+      "loss": 1.6622,
+      "step": 6398
+    },
+    {
+      "epoch": 0.3566889632107023,
+      "grad_norm": 0.5280166864395142,
+      "learning_rate": 8e-05,
+      "loss": 1.6834,
+      "step": 6399
+    },
+    {
+      "epoch": 0.35674470457079155,
+      "grad_norm": 0.5540995001792908,
+      "learning_rate": 8e-05,
+      "loss": 1.9026,
+      "step": 6400
+    },
+    {
+      "epoch": 0.3568004459308807,
+      "grad_norm": 0.5353748798370361,
+      "learning_rate": 8e-05,
+      "loss": 1.7235,
+      "step": 6401
+    },
+    {
+      "epoch": 0.3568561872909699,
+      "grad_norm": 0.5143972635269165,
+      "learning_rate": 8e-05,
+      "loss": 1.6317,
+      "step": 6402
+    },
+    {
+      "epoch": 0.3569119286510591,
+      "grad_norm": 0.4793776273727417,
+      "learning_rate": 8e-05,
+      "loss": 1.4453,
+      "step": 6403
+    },
+    {
+      "epoch": 0.35696767001114826,
+      "grad_norm": 0.5332164168357849,
+      "learning_rate": 8e-05,
+      "loss": 1.6527,
+      "step": 6404
+    },
+    {
+      "epoch": 0.35702341137123744,
+      "grad_norm": 0.5050964951515198,
+      "learning_rate": 8e-05,
+      "loss": 1.627,
+      "step": 6405
+    },
+    {
+      "epoch": 0.3570791527313266,
+      "grad_norm": 0.4644363224506378,
+      "learning_rate": 8e-05,
+      "loss": 1.451,
+      "step": 6406
+    },
+    {
+      "epoch": 0.35713489409141586,
+      "grad_norm": 0.5282057523727417,
+      "learning_rate": 8e-05,
+      "loss": 1.5854,
+      "step": 6407
+    },
+    {
+      "epoch": 0.35719063545150503,
+      "grad_norm": 0.5058812499046326,
+      "learning_rate": 8e-05,
+      "loss": 1.8473,
+      "step": 6408
+    },
+    {
+      "epoch": 0.3572463768115942,
+      "grad_norm": 0.49288514256477356,
+      "learning_rate": 8e-05,
+      "loss": 1.6421,
+      "step": 6409
+    },
+    {
+      "epoch": 0.3573021181716834,
+      "grad_norm": 0.5028522610664368,
+      "learning_rate": 8e-05,
+      "loss": 1.7249,
+      "step": 6410
+    },
+    {
+      "epoch": 0.35735785953177257,
+      "grad_norm": 0.4753098487854004,
+      "learning_rate": 8e-05,
+      "loss": 1.5235,
+      "step": 6411
+    },
+    {
+      "epoch": 0.35741360089186175,
+      "grad_norm": 0.49716320633888245,
+      "learning_rate": 8e-05,
+      "loss": 1.582,
+      "step": 6412
+    },
+    {
+      "epoch": 0.35746934225195093,
+      "grad_norm": 0.5260248184204102,
+      "learning_rate": 8e-05,
+      "loss": 1.5284,
+      "step": 6413
+    },
+    {
+      "epoch": 0.3575250836120401,
+      "grad_norm": 0.5542808175086975,
+      "learning_rate": 8e-05,
+      "loss": 1.7856,
+      "step": 6414
+    },
+    {
+      "epoch": 0.35758082497212934,
+      "grad_norm": 0.5370000600814819,
+      "learning_rate": 8e-05,
+      "loss": 1.7328,
+      "step": 6415
+    },
+    {
+      "epoch": 0.3576365663322185,
+      "grad_norm": 0.5088168382644653,
+      "learning_rate": 8e-05,
+      "loss": 1.5591,
+      "step": 6416
+    },
+    {
+      "epoch": 0.3576923076923077,
+      "grad_norm": 0.5215413570404053,
+      "learning_rate": 8e-05,
+      "loss": 1.2636,
+      "step": 6417
+    },
+    {
+      "epoch": 0.3577480490523969,
+      "grad_norm": 0.528469979763031,
+      "learning_rate": 8e-05,
+      "loss": 1.6741,
+      "step": 6418
+    },
+    {
+      "epoch": 0.35780379041248606,
+      "grad_norm": 0.49583855271339417,
+      "learning_rate": 8e-05,
+      "loss": 1.6579,
+      "step": 6419
+    },
+    {
+      "epoch": 0.35785953177257523,
+      "grad_norm": 0.5597677826881409,
+      "learning_rate": 8e-05,
+      "loss": 1.7846,
+      "step": 6420
+    },
+    {
+      "epoch": 0.3579152731326644,
+      "grad_norm": 0.5637001395225525,
+      "learning_rate": 8e-05,
+      "loss": 1.8262,
+      "step": 6421
+    },
+    {
+      "epoch": 0.35797101449275365,
+      "grad_norm": 0.5613079071044922,
+      "learning_rate": 8e-05,
+      "loss": 1.7265,
+      "step": 6422
+    },
+    {
+      "epoch": 0.3580267558528428,
+      "grad_norm": 0.5505542159080505,
+      "learning_rate": 8e-05,
+      "loss": 1.6039,
+      "step": 6423
+    },
+    {
+      "epoch": 0.358082497212932,
+      "grad_norm": 0.5501434206962585,
+      "learning_rate": 8e-05,
+      "loss": 1.7493,
+      "step": 6424
+    },
+    {
+      "epoch": 0.3581382385730212,
+      "grad_norm": 0.5145764946937561,
+      "learning_rate": 8e-05,
+      "loss": 1.7458,
+      "step": 6425
+    },
+    {
+      "epoch": 0.35819397993311036,
+      "grad_norm": 0.5651512742042542,
+      "learning_rate": 8e-05,
+      "loss": 1.8958,
+      "step": 6426
+    },
+    {
+      "epoch": 0.35824972129319954,
+      "grad_norm": 0.5379459857940674,
+      "learning_rate": 8e-05,
+      "loss": 1.9652,
+      "step": 6427
+    },
+    {
+      "epoch": 0.3583054626532887,
+      "grad_norm": 0.46046844124794006,
+      "learning_rate": 8e-05,
+      "loss": 1.4552,
+      "step": 6428
+    },
+    {
+      "epoch": 0.35836120401337795,
+      "grad_norm": 0.5376448035240173,
+      "learning_rate": 8e-05,
+      "loss": 1.6586,
+      "step": 6429
+    },
+    {
+      "epoch": 0.35841694537346713,
+      "grad_norm": 0.4953959584236145,
+      "learning_rate": 8e-05,
+      "loss": 1.3764,
+      "step": 6430
+    },
+    {
+      "epoch": 0.3584726867335563,
+      "grad_norm": 0.5916145443916321,
+      "learning_rate": 8e-05,
+      "loss": 2.108,
+      "step": 6431
+    },
+    {
+      "epoch": 0.3585284280936455,
+      "grad_norm": 0.5416443943977356,
+      "learning_rate": 8e-05,
+      "loss": 1.7952,
+      "step": 6432
+    },
+    {
+      "epoch": 0.35858416945373467,
+      "grad_norm": 0.516238272190094,
+      "learning_rate": 8e-05,
+      "loss": 1.6125,
+      "step": 6433
+    },
+    {
+      "epoch": 0.35863991081382385,
+      "grad_norm": 0.4888264536857605,
+      "learning_rate": 8e-05,
+      "loss": 1.5603,
+      "step": 6434
+    },
+    {
+      "epoch": 0.358695652173913,
+      "grad_norm": 0.5799896717071533,
+      "learning_rate": 8e-05,
+      "loss": 1.9252,
+      "step": 6435
+    },
+    {
+      "epoch": 0.3587513935340022,
+      "grad_norm": 0.4998963177204132,
+      "learning_rate": 8e-05,
+      "loss": 1.6797,
+      "step": 6436
+    },
+    {
+      "epoch": 0.35880713489409144,
+      "grad_norm": 0.5072805881500244,
+      "learning_rate": 8e-05,
+      "loss": 1.557,
+      "step": 6437
+    },
+    {
+      "epoch": 0.3588628762541806,
+      "grad_norm": 0.5456826686859131,
+      "learning_rate": 8e-05,
+      "loss": 1.5075,
+      "step": 6438
+    },
+    {
+      "epoch": 0.3589186176142698,
+      "grad_norm": 0.5327960252761841,
+      "learning_rate": 8e-05,
+      "loss": 1.6091,
+      "step": 6439
+    },
+    {
+      "epoch": 0.358974358974359,
+      "grad_norm": 0.5547055602073669,
+      "learning_rate": 8e-05,
+      "loss": 1.6923,
+      "step": 6440
+    },
+    {
+      "epoch": 0.35903010033444815,
+      "grad_norm": 0.5358332395553589,
+      "learning_rate": 8e-05,
+      "loss": 1.7189,
+      "step": 6441
+    },
+    {
+      "epoch": 0.35908584169453733,
+      "grad_norm": 0.5085732340812683,
+      "learning_rate": 8e-05,
+      "loss": 1.5005,
+      "step": 6442
+    },
+    {
+      "epoch": 0.3591415830546265,
+      "grad_norm": 0.5534481406211853,
+      "learning_rate": 8e-05,
+      "loss": 2.0119,
+      "step": 6443
+    },
+    {
+      "epoch": 0.35919732441471575,
+      "grad_norm": 0.5384095907211304,
+      "learning_rate": 8e-05,
+      "loss": 1.792,
+      "step": 6444
+    },
+    {
+      "epoch": 0.3592530657748049,
+      "grad_norm": 0.4997360110282898,
+      "learning_rate": 8e-05,
+      "loss": 1.6842,
+      "step": 6445
+    },
+    {
+      "epoch": 0.3593088071348941,
+      "grad_norm": 0.5220462679862976,
+      "learning_rate": 8e-05,
+      "loss": 1.7234,
+      "step": 6446
+    },
+    {
+      "epoch": 0.3593645484949833,
+      "grad_norm": 0.5385878682136536,
+      "learning_rate": 8e-05,
+      "loss": 1.7578,
+      "step": 6447
+    },
+    {
+      "epoch": 0.35942028985507246,
+      "grad_norm": 0.5040372014045715,
+      "learning_rate": 8e-05,
+      "loss": 1.7622,
+      "step": 6448
+    },
+    {
+      "epoch": 0.35947603121516164,
+      "grad_norm": 0.512959897518158,
+      "learning_rate": 8e-05,
+      "loss": 1.8103,
+      "step": 6449
+    },
+    {
+      "epoch": 0.3595317725752508,
+      "grad_norm": 0.4947265088558197,
+      "learning_rate": 8e-05,
+      "loss": 1.711,
+      "step": 6450
+    },
+    {
+      "epoch": 0.35958751393534,
+      "grad_norm": 0.5018380284309387,
+      "learning_rate": 8e-05,
+      "loss": 1.695,
+      "step": 6451
+    },
+    {
+      "epoch": 0.35964325529542923,
+      "grad_norm": 0.5547189116477966,
+      "learning_rate": 8e-05,
+      "loss": 1.865,
+      "step": 6452
+    },
+    {
+      "epoch": 0.3596989966555184,
+      "grad_norm": 0.5413358807563782,
+      "learning_rate": 8e-05,
+      "loss": 1.8049,
+      "step": 6453
+    },
+    {
+      "epoch": 0.3597547380156076,
+      "grad_norm": 0.5174626708030701,
+      "learning_rate": 8e-05,
+      "loss": 1.5548,
+      "step": 6454
+    },
+    {
+      "epoch": 0.35981047937569677,
+      "grad_norm": 0.5683441758155823,
+      "learning_rate": 8e-05,
+      "loss": 1.9717,
+      "step": 6455
+    },
+    {
+      "epoch": 0.35986622073578595,
+      "grad_norm": 0.5123081207275391,
+      "learning_rate": 8e-05,
+      "loss": 1.625,
+      "step": 6456
+    },
+    {
+      "epoch": 0.3599219620958751,
+      "grad_norm": 0.5193832516670227,
+      "learning_rate": 8e-05,
+      "loss": 1.6638,
+      "step": 6457
+    },
+    {
+      "epoch": 0.3599777034559643,
+      "grad_norm": 0.5199078321456909,
+      "learning_rate": 8e-05,
+      "loss": 1.6831,
+      "step": 6458
+    },
+    {
+      "epoch": 0.36003344481605354,
+      "grad_norm": 0.5766400098800659,
+      "learning_rate": 8e-05,
+      "loss": 1.7626,
+      "step": 6459
+    },
+    {
+      "epoch": 0.3600891861761427,
+      "grad_norm": 0.523861289024353,
+      "learning_rate": 8e-05,
+      "loss": 1.7019,
+      "step": 6460
+    },
+    {
+      "epoch": 0.3601449275362319,
+      "grad_norm": 0.5427674651145935,
+      "learning_rate": 8e-05,
+      "loss": 1.821,
+      "step": 6461
+    },
+    {
+      "epoch": 0.3602006688963211,
+      "grad_norm": 0.5394901633262634,
+      "learning_rate": 8e-05,
+      "loss": 1.7174,
+      "step": 6462
+    },
+    {
+      "epoch": 0.36025641025641025,
+      "grad_norm": 0.7988752126693726,
+      "learning_rate": 8e-05,
+      "loss": 1.9678,
+      "step": 6463
+    },
+    {
+      "epoch": 0.36031215161649943,
+      "grad_norm": 0.5276942253112793,
+      "learning_rate": 8e-05,
+      "loss": 1.8143,
+      "step": 6464
+    },
+    {
+      "epoch": 0.3603678929765886,
+      "grad_norm": 0.5382676720619202,
+      "learning_rate": 8e-05,
+      "loss": 1.9482,
+      "step": 6465
+    },
+    {
+      "epoch": 0.3604236343366778,
+      "grad_norm": 0.5459386110305786,
+      "learning_rate": 8e-05,
+      "loss": 1.9035,
+      "step": 6466
+    },
+    {
+      "epoch": 0.360479375696767,
+      "grad_norm": 0.4714954197406769,
+      "learning_rate": 8e-05,
+      "loss": 1.4949,
+      "step": 6467
+    },
+    {
+      "epoch": 0.3605351170568562,
+      "grad_norm": 0.5136751532554626,
+      "learning_rate": 8e-05,
+      "loss": 1.7965,
+      "step": 6468
+    },
+    {
+      "epoch": 0.3605908584169454,
+      "grad_norm": 0.524118185043335,
+      "learning_rate": 8e-05,
+      "loss": 1.6161,
+      "step": 6469
+    },
+    {
+      "epoch": 0.36064659977703456,
+      "grad_norm": 0.4913913309574127,
+      "learning_rate": 8e-05,
+      "loss": 1.523,
+      "step": 6470
+    },
+    {
+      "epoch": 0.36070234113712374,
+      "grad_norm": 0.5972789525985718,
+      "learning_rate": 8e-05,
+      "loss": 1.9685,
+      "step": 6471
+    },
+    {
+      "epoch": 0.3607580824972129,
+      "grad_norm": 0.6274723410606384,
+      "learning_rate": 8e-05,
+      "loss": 1.7745,
+      "step": 6472
+    },
+    {
+      "epoch": 0.3608138238573021,
+      "grad_norm": 0.5555365681648254,
+      "learning_rate": 8e-05,
+      "loss": 1.67,
+      "step": 6473
+    },
+    {
+      "epoch": 0.36086956521739133,
+      "grad_norm": 0.5084021687507629,
+      "learning_rate": 8e-05,
+      "loss": 1.6604,
+      "step": 6474
+    },
+    {
+      "epoch": 0.3609253065774805,
+      "grad_norm": 0.5389943718910217,
+      "learning_rate": 8e-05,
+      "loss": 1.6026,
+      "step": 6475
+    },
+    {
+      "epoch": 0.3609810479375697,
+      "grad_norm": 0.5192217826843262,
+      "learning_rate": 8e-05,
+      "loss": 1.5066,
+      "step": 6476
+    },
+    {
+      "epoch": 0.36103678929765887,
+      "grad_norm": 0.5360554456710815,
+      "learning_rate": 8e-05,
+      "loss": 1.7987,
+      "step": 6477
+    },
+    {
+      "epoch": 0.36109253065774805,
+      "grad_norm": 0.5452953577041626,
+      "learning_rate": 8e-05,
+      "loss": 1.7616,
+      "step": 6478
+    },
+    {
+      "epoch": 0.3611482720178372,
+      "grad_norm": 0.5165794491767883,
+      "learning_rate": 8e-05,
+      "loss": 1.7592,
+      "step": 6479
+    },
+    {
+      "epoch": 0.3612040133779264,
+      "grad_norm": 0.48414236307144165,
+      "learning_rate": 8e-05,
+      "loss": 1.5535,
+      "step": 6480
+    },
+    {
+      "epoch": 0.3612597547380156,
+      "grad_norm": 0.5539426207542419,
+      "learning_rate": 8e-05,
+      "loss": 1.7877,
+      "step": 6481
+    },
+    {
+      "epoch": 0.3613154960981048,
+      "grad_norm": 0.5607216358184814,
+      "learning_rate": 8e-05,
+      "loss": 1.708,
+      "step": 6482
+    },
+    {
+      "epoch": 0.361371237458194,
+      "grad_norm": 0.5305792689323425,
+      "learning_rate": 8e-05,
+      "loss": 1.861,
+      "step": 6483
+    },
+    {
+      "epoch": 0.3614269788182832,
+      "grad_norm": 0.4897555112838745,
+      "learning_rate": 8e-05,
+      "loss": 1.5322,
+      "step": 6484
+    },
+    {
+      "epoch": 0.36148272017837235,
+      "grad_norm": 0.4937852919101715,
+      "learning_rate": 8e-05,
+      "loss": 1.5797,
+      "step": 6485
+    },
+    {
+      "epoch": 0.36153846153846153,
+      "grad_norm": 0.49905410408973694,
+      "learning_rate": 8e-05,
+      "loss": 1.6547,
+      "step": 6486
+    },
+    {
+      "epoch": 0.3615942028985507,
+      "grad_norm": 0.49061068892478943,
+      "learning_rate": 8e-05,
+      "loss": 1.4706,
+      "step": 6487
+    },
+    {
+      "epoch": 0.3616499442586399,
+      "grad_norm": 0.4852028787136078,
+      "learning_rate": 8e-05,
+      "loss": 1.4891,
+      "step": 6488
+    },
+    {
+      "epoch": 0.3617056856187291,
+      "grad_norm": 0.5745864510536194,
+      "learning_rate": 8e-05,
+      "loss": 1.9354,
+      "step": 6489
+    },
+    {
+      "epoch": 0.3617614269788183,
+      "grad_norm": 0.5412963628768921,
+      "learning_rate": 8e-05,
+      "loss": 1.7391,
+      "step": 6490
+    },
+    {
+      "epoch": 0.3618171683389075,
+      "grad_norm": 0.4556156396865845,
+      "learning_rate": 8e-05,
+      "loss": 1.5082,
+      "step": 6491
+    },
+    {
+      "epoch": 0.36187290969899666,
+      "grad_norm": 0.575387179851532,
+      "learning_rate": 8e-05,
+      "loss": 1.8843,
+      "step": 6492
+    },
+    {
+      "epoch": 0.36192865105908584,
+      "grad_norm": 0.5384889245033264,
+      "learning_rate": 8e-05,
+      "loss": 1.8429,
+      "step": 6493
+    },
+    {
+      "epoch": 0.361984392419175,
+      "grad_norm": 0.6044797301292419,
+      "learning_rate": 8e-05,
+      "loss": 1.736,
+      "step": 6494
+    },
+    {
+      "epoch": 0.3620401337792642,
+      "grad_norm": 0.5215487480163574,
+      "learning_rate": 8e-05,
+      "loss": 1.5594,
+      "step": 6495
+    },
+    {
+      "epoch": 0.3620958751393534,
+      "grad_norm": 0.5874164700508118,
+      "learning_rate": 8e-05,
+      "loss": 1.8676,
+      "step": 6496
+    },
+    {
+      "epoch": 0.3621516164994426,
+      "grad_norm": 0.5641337037086487,
+      "learning_rate": 8e-05,
+      "loss": 1.6421,
+      "step": 6497
+    },
+    {
+      "epoch": 0.3622073578595318,
+      "grad_norm": 0.5324305295944214,
+      "learning_rate": 8e-05,
+      "loss": 1.5296,
+      "step": 6498
+    },
+    {
+      "epoch": 0.36226309921962097,
+      "grad_norm": 0.4652421772480011,
+      "learning_rate": 8e-05,
+      "loss": 1.4351,
+      "step": 6499
+    },
+    {
+      "epoch": 0.36231884057971014,
+      "grad_norm": 0.5650059580802917,
+      "learning_rate": 8e-05,
+      "loss": 1.9025,
+      "step": 6500
+    },
+    {
+      "epoch": 0.3623745819397993,
+      "grad_norm": 0.509933590888977,
+      "learning_rate": 8e-05,
+      "loss": 1.6018,
+      "step": 6501
+    },
+    {
+      "epoch": 0.3624303232998885,
+      "grad_norm": 0.6044166088104248,
+      "learning_rate": 8e-05,
+      "loss": 1.8909,
+      "step": 6502
+    },
+    {
+      "epoch": 0.3624860646599777,
+      "grad_norm": 0.5065827965736389,
+      "learning_rate": 8e-05,
+      "loss": 1.7081,
+      "step": 6503
+    },
+    {
+      "epoch": 0.3625418060200669,
+      "grad_norm": 0.5435416102409363,
+      "learning_rate": 8e-05,
+      "loss": 1.9141,
+      "step": 6504
+    },
+    {
+      "epoch": 0.3625975473801561,
+      "grad_norm": 0.538236677646637,
+      "learning_rate": 8e-05,
+      "loss": 1.6744,
+      "step": 6505
+    },
+    {
+      "epoch": 0.36265328874024527,
+      "grad_norm": 0.49007493257522583,
+      "learning_rate": 8e-05,
+      "loss": 1.5196,
+      "step": 6506
+    },
+    {
+      "epoch": 0.36270903010033445,
+      "grad_norm": 0.5888476967811584,
+      "learning_rate": 8e-05,
+      "loss": 1.9067,
+      "step": 6507
+    },
+    {
+      "epoch": 0.36276477146042363,
+      "grad_norm": 0.5109419226646423,
+      "learning_rate": 8e-05,
+      "loss": 1.9334,
+      "step": 6508
+    },
+    {
+      "epoch": 0.3628205128205128,
+      "grad_norm": 0.537640392780304,
+      "learning_rate": 8e-05,
+      "loss": 1.8233,
+      "step": 6509
+    },
+    {
+      "epoch": 0.362876254180602,
+      "grad_norm": 0.48754364252090454,
+      "learning_rate": 8e-05,
+      "loss": 1.5459,
+      "step": 6510
+    },
+    {
+      "epoch": 0.36293199554069117,
+      "grad_norm": 0.5003510117530823,
+      "learning_rate": 8e-05,
+      "loss": 1.6937,
+      "step": 6511
+    },
+    {
+      "epoch": 0.3629877369007804,
+      "grad_norm": 0.46658939123153687,
+      "learning_rate": 8e-05,
+      "loss": 1.4756,
+      "step": 6512
+    },
+    {
+      "epoch": 0.3630434782608696,
+      "grad_norm": 0.5610101819038391,
+      "learning_rate": 8e-05,
+      "loss": 1.8659,
+      "step": 6513
+    },
+    {
+      "epoch": 0.36309921962095876,
+      "grad_norm": 0.49573513865470886,
+      "learning_rate": 8e-05,
+      "loss": 1.6819,
+      "step": 6514
+    },
+    {
+      "epoch": 0.36315496098104794,
+      "grad_norm": 0.49012234807014465,
+      "learning_rate": 8e-05,
+      "loss": 1.5958,
+      "step": 6515
+    },
+    {
+      "epoch": 0.3632107023411371,
+      "grad_norm": 0.5030548572540283,
+      "learning_rate": 8e-05,
+      "loss": 1.7471,
+      "step": 6516
+    },
+    {
+      "epoch": 0.3632664437012263,
+      "grad_norm": 0.4662206470966339,
+      "learning_rate": 8e-05,
+      "loss": 1.3778,
+      "step": 6517
+    },
+    {
+      "epoch": 0.3633221850613155,
+      "grad_norm": 0.5224818587303162,
+      "learning_rate": 8e-05,
+      "loss": 1.7238,
+      "step": 6518
+    },
+    {
+      "epoch": 0.3633779264214047,
+      "grad_norm": 0.4647224247455597,
+      "learning_rate": 8e-05,
+      "loss": 1.2414,
+      "step": 6519
+    },
+    {
+      "epoch": 0.3634336677814939,
+      "grad_norm": 0.5260183811187744,
+      "learning_rate": 8e-05,
+      "loss": 1.7534,
+      "step": 6520
+    },
+    {
+      "epoch": 0.36348940914158306,
+      "grad_norm": 0.4867696166038513,
+      "learning_rate": 8e-05,
+      "loss": 1.5077,
+      "step": 6521
+    },
+    {
+      "epoch": 0.36354515050167224,
+      "grad_norm": 0.5324907898902893,
+      "learning_rate": 8e-05,
+      "loss": 1.6597,
+      "step": 6522
+    },
+    {
+      "epoch": 0.3636008918617614,
+      "grad_norm": 0.5028617978096008,
+      "learning_rate": 8e-05,
+      "loss": 1.5938,
+      "step": 6523
+    },
+    {
+      "epoch": 0.3636566332218506,
+      "grad_norm": 0.556318998336792,
+      "learning_rate": 8e-05,
+      "loss": 1.899,
+      "step": 6524
+    },
+    {
+      "epoch": 0.3637123745819398,
+      "grad_norm": 0.5142127275466919,
+      "learning_rate": 8e-05,
+      "loss": 1.6227,
+      "step": 6525
+    },
+    {
+      "epoch": 0.36376811594202896,
+      "grad_norm": 0.5022664070129395,
+      "learning_rate": 8e-05,
+      "loss": 1.677,
+      "step": 6526
+    },
+    {
+      "epoch": 0.3638238573021182,
+      "grad_norm": 0.4740968942642212,
+      "learning_rate": 8e-05,
+      "loss": 1.4775,
+      "step": 6527
+    },
+    {
+      "epoch": 0.36387959866220737,
+      "grad_norm": 0.6123348474502563,
+      "learning_rate": 8e-05,
+      "loss": 1.6875,
+      "step": 6528
+    },
+    {
+      "epoch": 0.36393534002229655,
+      "grad_norm": 0.5256551504135132,
+      "learning_rate": 8e-05,
+      "loss": 1.7736,
+      "step": 6529
+    },
+    {
+      "epoch": 0.36399108138238573,
+      "grad_norm": 0.5395634174346924,
+      "learning_rate": 8e-05,
+      "loss": 1.8819,
+      "step": 6530
+    },
+    {
+      "epoch": 0.3640468227424749,
+      "grad_norm": 0.5305240154266357,
+      "learning_rate": 8e-05,
+      "loss": 1.5619,
+      "step": 6531
+    },
+    {
+      "epoch": 0.3641025641025641,
+      "grad_norm": 0.4834372401237488,
+      "learning_rate": 8e-05,
+      "loss": 1.4774,
+      "step": 6532
+    },
+    {
+      "epoch": 0.36415830546265326,
+      "grad_norm": 0.5600225925445557,
+      "learning_rate": 8e-05,
+      "loss": 1.9042,
+      "step": 6533
+    },
+    {
+      "epoch": 0.3642140468227425,
+      "grad_norm": 0.5327090620994568,
+      "learning_rate": 8e-05,
+      "loss": 1.7592,
+      "step": 6534
+    },
+    {
+      "epoch": 0.3642697881828317,
+      "grad_norm": 0.48773810267448425,
+      "learning_rate": 8e-05,
+      "loss": 1.6035,
+      "step": 6535
+    },
+    {
+      "epoch": 0.36432552954292086,
+      "grad_norm": 0.5857282876968384,
+      "learning_rate": 8e-05,
+      "loss": 1.8036,
+      "step": 6536
+    },
+    {
+      "epoch": 0.36438127090301003,
+      "grad_norm": 0.5390805602073669,
+      "learning_rate": 8e-05,
+      "loss": 1.6281,
+      "step": 6537
+    },
+    {
+      "epoch": 0.3644370122630992,
+      "grad_norm": 0.4638005495071411,
+      "learning_rate": 8e-05,
+      "loss": 1.4809,
+      "step": 6538
+    },
+    {
+      "epoch": 0.3644927536231884,
+      "grad_norm": 0.4900088906288147,
+      "learning_rate": 8e-05,
+      "loss": 1.4965,
+      "step": 6539
+    },
+    {
+      "epoch": 0.36454849498327757,
+      "grad_norm": 0.5323425531387329,
+      "learning_rate": 8e-05,
+      "loss": 1.8889,
+      "step": 6540
+    },
+    {
+      "epoch": 0.3646042363433668,
+      "grad_norm": 0.5697327256202698,
+      "learning_rate": 8e-05,
+      "loss": 1.8247,
+      "step": 6541
+    },
+    {
+      "epoch": 0.364659977703456,
+      "grad_norm": 0.5275499224662781,
+      "learning_rate": 8e-05,
+      "loss": 1.6131,
+      "step": 6542
+    },
+    {
+      "epoch": 0.36471571906354516,
+      "grad_norm": 0.5853115320205688,
+      "learning_rate": 8e-05,
+      "loss": 1.4321,
+      "step": 6543
+    },
+    {
+      "epoch": 0.36477146042363434,
+      "grad_norm": 0.5601681470870972,
+      "learning_rate": 8e-05,
+      "loss": 1.7446,
+      "step": 6544
+    },
+    {
+      "epoch": 0.3648272017837235,
+      "grad_norm": 0.5006238222122192,
+      "learning_rate": 8e-05,
+      "loss": 1.5411,
+      "step": 6545
+    },
+    {
+      "epoch": 0.3648829431438127,
+      "grad_norm": 0.5464562177658081,
+      "learning_rate": 8e-05,
+      "loss": 1.7087,
+      "step": 6546
+    },
+    {
+      "epoch": 0.3649386845039019,
+      "grad_norm": 0.5300664901733398,
+      "learning_rate": 8e-05,
+      "loss": 1.7823,
+      "step": 6547
+    },
+    {
+      "epoch": 0.36499442586399106,
+      "grad_norm": 0.5564868450164795,
+      "learning_rate": 8e-05,
+      "loss": 2.0165,
+      "step": 6548
+    },
+    {
+      "epoch": 0.3650501672240803,
+      "grad_norm": 0.574718177318573,
+      "learning_rate": 8e-05,
+      "loss": 1.9779,
+      "step": 6549
+    },
+    {
+      "epoch": 0.36510590858416947,
+      "grad_norm": 0.4962438941001892,
+      "learning_rate": 8e-05,
+      "loss": 1.7449,
+      "step": 6550
+    },
+    {
+      "epoch": 0.36516164994425865,
+      "grad_norm": 0.5202713012695312,
+      "learning_rate": 8e-05,
+      "loss": 1.6674,
+      "step": 6551
+    },
+    {
+      "epoch": 0.3652173913043478,
+      "grad_norm": 0.5091273188591003,
+      "learning_rate": 8e-05,
+      "loss": 1.7102,
+      "step": 6552
+    },
+    {
+      "epoch": 0.365273132664437,
+      "grad_norm": 0.4962243139743805,
+      "learning_rate": 8e-05,
+      "loss": 1.7279,
+      "step": 6553
+    },
+    {
+      "epoch": 0.3653288740245262,
+      "grad_norm": 0.5044227242469788,
+      "learning_rate": 8e-05,
+      "loss": 1.8611,
+      "step": 6554
+    },
+    {
+      "epoch": 0.36538461538461536,
+      "grad_norm": 0.5568208694458008,
+      "learning_rate": 8e-05,
+      "loss": 1.5701,
+      "step": 6555
+    },
+    {
+      "epoch": 0.3654403567447046,
+      "grad_norm": 0.4962807297706604,
+      "learning_rate": 8e-05,
+      "loss": 1.5269,
+      "step": 6556
+    },
+    {
+      "epoch": 0.3654960981047938,
+      "grad_norm": 0.48871636390686035,
+      "learning_rate": 8e-05,
+      "loss": 1.6167,
+      "step": 6557
+    },
+    {
+      "epoch": 0.36555183946488295,
+      "grad_norm": 0.5097386837005615,
+      "learning_rate": 8e-05,
+      "loss": 1.7949,
+      "step": 6558
+    },
+    {
+      "epoch": 0.36560758082497213,
+      "grad_norm": 0.5374859571456909,
+      "learning_rate": 8e-05,
+      "loss": 1.64,
+      "step": 6559
+    },
+    {
+      "epoch": 0.3656633221850613,
+      "grad_norm": 0.5586405396461487,
+      "learning_rate": 8e-05,
+      "loss": 1.4684,
+      "step": 6560
+    },
+    {
+      "epoch": 0.3657190635451505,
+      "grad_norm": 0.5146247148513794,
+      "learning_rate": 8e-05,
+      "loss": 1.5647,
+      "step": 6561
+    },
+    {
+      "epoch": 0.36577480490523967,
+      "grad_norm": 0.5160596966743469,
+      "learning_rate": 8e-05,
+      "loss": 1.5378,
+      "step": 6562
+    },
+    {
+      "epoch": 0.36583054626532885,
+      "grad_norm": 0.5124092102050781,
+      "learning_rate": 8e-05,
+      "loss": 1.5913,
+      "step": 6563
+    },
+    {
+      "epoch": 0.3658862876254181,
+      "grad_norm": 0.5166916251182556,
+      "learning_rate": 8e-05,
+      "loss": 1.6796,
+      "step": 6564
+    },
+    {
+      "epoch": 0.36594202898550726,
+      "grad_norm": 0.512089729309082,
+      "learning_rate": 8e-05,
+      "loss": 1.5568,
+      "step": 6565
+    },
+    {
+      "epoch": 0.36599777034559644,
+      "grad_norm": 0.48678961396217346,
+      "learning_rate": 8e-05,
+      "loss": 1.5374,
+      "step": 6566
+    },
+    {
+      "epoch": 0.3660535117056856,
+      "grad_norm": 0.5077835321426392,
+      "learning_rate": 8e-05,
+      "loss": 1.695,
+      "step": 6567
+    },
+    {
+      "epoch": 0.3661092530657748,
+      "grad_norm": 0.5288443565368652,
+      "learning_rate": 8e-05,
+      "loss": 1.7829,
+      "step": 6568
+    },
+    {
+      "epoch": 0.366164994425864,
+      "grad_norm": 0.5403035283088684,
+      "learning_rate": 8e-05,
+      "loss": 1.7271,
+      "step": 6569
+    },
+    {
+      "epoch": 0.36622073578595316,
+      "grad_norm": 0.5623093247413635,
+      "learning_rate": 8e-05,
+      "loss": 1.862,
+      "step": 6570
+    },
+    {
+      "epoch": 0.3662764771460424,
+      "grad_norm": 0.5390522480010986,
+      "learning_rate": 8e-05,
+      "loss": 1.693,
+      "step": 6571
+    },
+    {
+      "epoch": 0.36633221850613157,
+      "grad_norm": 0.5267409086227417,
+      "learning_rate": 8e-05,
+      "loss": 1.5246,
+      "step": 6572
+    },
+    {
+      "epoch": 0.36638795986622075,
+      "grad_norm": 0.507342517375946,
+      "learning_rate": 8e-05,
+      "loss": 1.6128,
+      "step": 6573
+    },
+    {
+      "epoch": 0.3664437012263099,
+      "grad_norm": 0.5371213555335999,
+      "learning_rate": 8e-05,
+      "loss": 1.6659,
+      "step": 6574
+    },
+    {
+      "epoch": 0.3664994425863991,
+      "grad_norm": 0.5376846194267273,
+      "learning_rate": 8e-05,
+      "loss": 1.7991,
+      "step": 6575
+    },
+    {
+      "epoch": 0.3665551839464883,
+      "grad_norm": 0.5879801511764526,
+      "learning_rate": 8e-05,
+      "loss": 1.9445,
+      "step": 6576
+    },
+    {
+      "epoch": 0.36661092530657746,
+      "grad_norm": 0.5080875754356384,
+      "learning_rate": 8e-05,
+      "loss": 1.4541,
+      "step": 6577
+    },
+    {
+      "epoch": 0.36666666666666664,
+      "grad_norm": 0.49109145998954773,
+      "learning_rate": 8e-05,
+      "loss": 1.5978,
+      "step": 6578
+    },
+    {
+      "epoch": 0.3667224080267559,
+      "grad_norm": 0.5372000932693481,
+      "learning_rate": 8e-05,
+      "loss": 1.7174,
+      "step": 6579
+    },
+    {
+      "epoch": 0.36677814938684505,
+      "grad_norm": 0.5064409375190735,
+      "learning_rate": 8e-05,
+      "loss": 1.6052,
+      "step": 6580
+    },
+    {
+      "epoch": 0.36683389074693423,
+      "grad_norm": 0.4575348496437073,
+      "learning_rate": 8e-05,
+      "loss": 1.4182,
+      "step": 6581
+    },
+    {
+      "epoch": 0.3668896321070234,
+      "grad_norm": 0.5588882565498352,
+      "learning_rate": 8e-05,
+      "loss": 1.8103,
+      "step": 6582
+    },
+    {
+      "epoch": 0.3669453734671126,
+      "grad_norm": 0.5252819657325745,
+      "learning_rate": 8e-05,
+      "loss": 1.7079,
+      "step": 6583
+    },
+    {
+      "epoch": 0.36700111482720177,
+      "grad_norm": 0.5528586506843567,
+      "learning_rate": 8e-05,
+      "loss": 1.864,
+      "step": 6584
+    },
+    {
+      "epoch": 0.36705685618729095,
+      "grad_norm": 0.5005204081535339,
+      "learning_rate": 8e-05,
+      "loss": 1.5407,
+      "step": 6585
+    },
+    {
+      "epoch": 0.3671125975473802,
+      "grad_norm": 0.5155864953994751,
+      "learning_rate": 8e-05,
+      "loss": 1.7737,
+      "step": 6586
+    },
+    {
+      "epoch": 0.36716833890746936,
+      "grad_norm": 0.5482629537582397,
+      "learning_rate": 8e-05,
+      "loss": 1.7748,
+      "step": 6587
+    },
+    {
+      "epoch": 0.36722408026755854,
+      "grad_norm": 0.5585765838623047,
+      "learning_rate": 8e-05,
+      "loss": 1.7237,
+      "step": 6588
+    },
+    {
+      "epoch": 0.3672798216276477,
+      "grad_norm": 0.49062028527259827,
+      "learning_rate": 8e-05,
+      "loss": 1.5748,
+      "step": 6589
+    },
+    {
+      "epoch": 0.3673355629877369,
+      "grad_norm": 0.5338550209999084,
+      "learning_rate": 8e-05,
+      "loss": 1.7891,
+      "step": 6590
+    },
+    {
+      "epoch": 0.3673913043478261,
+      "grad_norm": 0.5884079337120056,
+      "learning_rate": 8e-05,
+      "loss": 1.6608,
+      "step": 6591
+    },
+    {
+      "epoch": 0.36744704570791525,
+      "grad_norm": 0.5195145606994629,
+      "learning_rate": 8e-05,
+      "loss": 1.7409,
+      "step": 6592
+    },
+    {
+      "epoch": 0.36750278706800443,
+      "grad_norm": 0.5180572271347046,
+      "learning_rate": 8e-05,
+      "loss": 1.6582,
+      "step": 6593
+    },
+    {
+      "epoch": 0.36755852842809367,
+      "grad_norm": 0.5171594619750977,
+      "learning_rate": 8e-05,
+      "loss": 1.7377,
+      "step": 6594
+    },
+    {
+      "epoch": 0.36761426978818285,
+      "grad_norm": 0.5494788289070129,
+      "learning_rate": 8e-05,
+      "loss": 2.0252,
+      "step": 6595
+    },
+    {
+      "epoch": 0.367670011148272,
+      "grad_norm": 0.5018215179443359,
+      "learning_rate": 8e-05,
+      "loss": 1.7187,
+      "step": 6596
+    },
+    {
+      "epoch": 0.3677257525083612,
+      "grad_norm": 0.5216608047485352,
+      "learning_rate": 8e-05,
+      "loss": 1.6828,
+      "step": 6597
+    },
+    {
+      "epoch": 0.3677814938684504,
+      "grad_norm": 0.4886898100376129,
+      "learning_rate": 8e-05,
+      "loss": 1.695,
+      "step": 6598
+    },
+    {
+      "epoch": 0.36783723522853956,
+      "grad_norm": 0.5049815773963928,
+      "learning_rate": 8e-05,
+      "loss": 1.6745,
+      "step": 6599
+    },
+    {
+      "epoch": 0.36789297658862874,
+      "grad_norm": 0.51988685131073,
+      "learning_rate": 8e-05,
+      "loss": 1.6152,
+      "step": 6600
+    },
+    {
+      "epoch": 0.367948717948718,
+      "grad_norm": 0.4952639043331146,
+      "learning_rate": 8e-05,
+      "loss": 1.5671,
+      "step": 6601
+    },
+    {
+      "epoch": 0.36800445930880715,
+      "grad_norm": 0.5094792246818542,
+      "learning_rate": 8e-05,
+      "loss": 1.6354,
+      "step": 6602
+    },
+    {
+      "epoch": 0.36806020066889633,
+      "grad_norm": 0.5375295877456665,
+      "learning_rate": 8e-05,
+      "loss": 1.5954,
+      "step": 6603
+    },
+    {
+      "epoch": 0.3681159420289855,
+      "grad_norm": 0.5196813941001892,
+      "learning_rate": 8e-05,
+      "loss": 1.5596,
+      "step": 6604
+    },
+    {
+      "epoch": 0.3681716833890747,
+      "grad_norm": 0.524641752243042,
+      "learning_rate": 8e-05,
+      "loss": 1.6315,
+      "step": 6605
+    },
+    {
+      "epoch": 0.36822742474916387,
+      "grad_norm": 0.5774500370025635,
+      "learning_rate": 8e-05,
+      "loss": 1.7054,
+      "step": 6606
+    },
+    {
+      "epoch": 0.36828316610925305,
+      "grad_norm": 0.5841911435127258,
+      "learning_rate": 8e-05,
+      "loss": 1.9424,
+      "step": 6607
+    },
+    {
+      "epoch": 0.3683389074693422,
+      "grad_norm": 0.5551748871803284,
+      "learning_rate": 8e-05,
+      "loss": 1.7258,
+      "step": 6608
+    },
+    {
+      "epoch": 0.36839464882943146,
+      "grad_norm": 0.5234334468841553,
+      "learning_rate": 8e-05,
+      "loss": 1.866,
+      "step": 6609
+    },
+    {
+      "epoch": 0.36845039018952064,
+      "grad_norm": 0.48736587166786194,
+      "learning_rate": 8e-05,
+      "loss": 1.3496,
+      "step": 6610
+    },
+    {
+      "epoch": 0.3685061315496098,
+      "grad_norm": 0.5116720795631409,
+      "learning_rate": 8e-05,
+      "loss": 1.8454,
+      "step": 6611
+    },
+    {
+      "epoch": 0.368561872909699,
+      "grad_norm": 0.4924512803554535,
+      "learning_rate": 8e-05,
+      "loss": 1.6009,
+      "step": 6612
+    },
+    {
+      "epoch": 0.3686176142697882,
+      "grad_norm": 0.49612775444984436,
+      "learning_rate": 8e-05,
+      "loss": 1.5169,
+      "step": 6613
+    },
+    {
+      "epoch": 0.36867335562987735,
+      "grad_norm": 0.5605433583259583,
+      "learning_rate": 8e-05,
+      "loss": 1.6678,
+      "step": 6614
+    },
+    {
+      "epoch": 0.36872909698996653,
+      "grad_norm": 0.5453359484672546,
+      "learning_rate": 8e-05,
+      "loss": 1.7332,
+      "step": 6615
+    },
+    {
+      "epoch": 0.36878483835005577,
+      "grad_norm": 0.507093608379364,
+      "learning_rate": 8e-05,
+      "loss": 1.7661,
+      "step": 6616
+    },
+    {
+      "epoch": 0.36884057971014494,
+      "grad_norm": 0.5282663106918335,
+      "learning_rate": 8e-05,
+      "loss": 1.7891,
+      "step": 6617
+    },
+    {
+      "epoch": 0.3688963210702341,
+      "grad_norm": 0.5238633155822754,
+      "learning_rate": 8e-05,
+      "loss": 1.7113,
+      "step": 6618
+    },
+    {
+      "epoch": 0.3689520624303233,
+      "grad_norm": 0.5179184079170227,
+      "learning_rate": 8e-05,
+      "loss": 1.7173,
+      "step": 6619
+    },
+    {
+      "epoch": 0.3690078037904125,
+      "grad_norm": 0.5235730409622192,
+      "learning_rate": 8e-05,
+      "loss": 1.6982,
+      "step": 6620
+    },
+    {
+      "epoch": 0.36906354515050166,
+      "grad_norm": 0.49404484033584595,
+      "learning_rate": 8e-05,
+      "loss": 1.5118,
+      "step": 6621
+    },
+    {
+      "epoch": 0.36911928651059084,
+      "grad_norm": 0.5173282623291016,
+      "learning_rate": 8e-05,
+      "loss": 1.6177,
+      "step": 6622
+    },
+    {
+      "epoch": 0.36917502787068,
+      "grad_norm": 0.5297023057937622,
+      "learning_rate": 8e-05,
+      "loss": 1.7344,
+      "step": 6623
+    },
+    {
+      "epoch": 0.36923076923076925,
+      "grad_norm": 0.512212872505188,
+      "learning_rate": 8e-05,
+      "loss": 1.5957,
+      "step": 6624
+    },
+    {
+      "epoch": 0.36928651059085843,
+      "grad_norm": 0.504366934299469,
+      "learning_rate": 8e-05,
+      "loss": 1.672,
+      "step": 6625
+    },
+    {
+      "epoch": 0.3693422519509476,
+      "grad_norm": 0.5258657932281494,
+      "learning_rate": 8e-05,
+      "loss": 1.5942,
+      "step": 6626
+    },
+    {
+      "epoch": 0.3693979933110368,
+      "grad_norm": 0.5346935987472534,
+      "learning_rate": 8e-05,
+      "loss": 1.6521,
+      "step": 6627
+    },
+    {
+      "epoch": 0.36945373467112597,
+      "grad_norm": 0.4969017505645752,
+      "learning_rate": 8e-05,
+      "loss": 1.6157,
+      "step": 6628
+    },
+    {
+      "epoch": 0.36950947603121514,
+      "grad_norm": 0.5460304021835327,
+      "learning_rate": 8e-05,
+      "loss": 1.6787,
+      "step": 6629
+    },
+    {
+      "epoch": 0.3695652173913043,
+      "grad_norm": 0.4940931797027588,
+      "learning_rate": 8e-05,
+      "loss": 1.6568,
+      "step": 6630
+    },
+    {
+      "epoch": 0.36962095875139356,
+      "grad_norm": 0.5654184818267822,
+      "learning_rate": 8e-05,
+      "loss": 1.7727,
+      "step": 6631
+    },
+    {
+      "epoch": 0.36967670011148274,
+      "grad_norm": 0.48522743582725525,
+      "learning_rate": 8e-05,
+      "loss": 1.4585,
+      "step": 6632
+    },
+    {
+      "epoch": 0.3697324414715719,
+      "grad_norm": 0.5106379985809326,
+      "learning_rate": 8e-05,
+      "loss": 1.5965,
+      "step": 6633
+    },
+    {
+      "epoch": 0.3697881828316611,
+      "grad_norm": 0.5287076830863953,
+      "learning_rate": 8e-05,
+      "loss": 1.6599,
+      "step": 6634
+    },
+    {
+      "epoch": 0.3698439241917503,
+      "grad_norm": 0.4883654713630676,
+      "learning_rate": 8e-05,
+      "loss": 1.6157,
+      "step": 6635
+    },
+    {
+      "epoch": 0.36989966555183945,
+      "grad_norm": 0.459191232919693,
+      "learning_rate": 8e-05,
+      "loss": 1.3488,
+      "step": 6636
+    },
+    {
+      "epoch": 0.36995540691192863,
+      "grad_norm": 0.5547154545783997,
+      "learning_rate": 8e-05,
+      "loss": 1.8267,
+      "step": 6637
+    },
+    {
+      "epoch": 0.37001114827201786,
+      "grad_norm": 0.5515210628509521,
+      "learning_rate": 8e-05,
+      "loss": 1.7662,
+      "step": 6638
+    },
+    {
+      "epoch": 0.37006688963210704,
+      "grad_norm": 0.571776270866394,
+      "learning_rate": 8e-05,
+      "loss": 1.8165,
+      "step": 6639
+    },
+    {
+      "epoch": 0.3701226309921962,
+      "grad_norm": 0.4707722067832947,
+      "learning_rate": 8e-05,
+      "loss": 1.6566,
+      "step": 6640
+    },
+    {
+      "epoch": 0.3701783723522854,
+      "grad_norm": 0.5285392999649048,
+      "learning_rate": 8e-05,
+      "loss": 1.6576,
+      "step": 6641
+    },
+    {
+      "epoch": 0.3702341137123746,
+      "grad_norm": 0.49318474531173706,
+      "learning_rate": 8e-05,
+      "loss": 1.6507,
+      "step": 6642
+    },
+    {
+      "epoch": 0.37028985507246376,
+      "grad_norm": 0.5350597500801086,
+      "learning_rate": 8e-05,
+      "loss": 1.7359,
+      "step": 6643
+    },
+    {
+      "epoch": 0.37034559643255294,
+      "grad_norm": 0.561744213104248,
+      "learning_rate": 8e-05,
+      "loss": 1.7681,
+      "step": 6644
+    },
+    {
+      "epoch": 0.3704013377926421,
+      "grad_norm": 0.529123842716217,
+      "learning_rate": 8e-05,
+      "loss": 1.6686,
+      "step": 6645
+    },
+    {
+      "epoch": 0.37045707915273135,
+      "grad_norm": 0.5006415247917175,
+      "learning_rate": 8e-05,
+      "loss": 1.4867,
+      "step": 6646
+    },
+    {
+      "epoch": 0.37051282051282053,
+      "grad_norm": 0.47411513328552246,
+      "learning_rate": 8e-05,
+      "loss": 1.6221,
+      "step": 6647
+    },
+    {
+      "epoch": 0.3705685618729097,
+      "grad_norm": 0.535148561000824,
+      "learning_rate": 8e-05,
+      "loss": 1.4756,
+      "step": 6648
+    },
+    {
+      "epoch": 0.3706243032329989,
+      "grad_norm": 0.483823299407959,
+      "learning_rate": 8e-05,
+      "loss": 1.6542,
+      "step": 6649
+    },
+    {
+      "epoch": 0.37068004459308806,
+      "grad_norm": 0.5823395252227783,
+      "learning_rate": 8e-05,
+      "loss": 1.8614,
+      "step": 6650
+    },
+    {
+      "epoch": 0.37073578595317724,
+      "grad_norm": 0.4946765899658203,
+      "learning_rate": 8e-05,
+      "loss": 1.5182,
+      "step": 6651
+    },
+    {
+      "epoch": 0.3707915273132664,
+      "grad_norm": 0.5327061414718628,
+      "learning_rate": 8e-05,
+      "loss": 1.6711,
+      "step": 6652
+    },
+    {
+      "epoch": 0.37084726867335566,
+      "grad_norm": 0.49041929841041565,
+      "learning_rate": 8e-05,
+      "loss": 1.6707,
+      "step": 6653
+    },
+    {
+      "epoch": 0.37090301003344484,
+      "grad_norm": 0.5231907367706299,
+      "learning_rate": 8e-05,
+      "loss": 1.5303,
+      "step": 6654
+    },
+    {
+      "epoch": 0.370958751393534,
+      "grad_norm": 0.47933900356292725,
+      "learning_rate": 8e-05,
+      "loss": 1.4115,
+      "step": 6655
+    },
+    {
+      "epoch": 0.3710144927536232,
+      "grad_norm": 0.5437414646148682,
+      "learning_rate": 8e-05,
+      "loss": 1.7237,
+      "step": 6656
+    },
+    {
+      "epoch": 0.37107023411371237,
+      "grad_norm": 0.5546405911445618,
+      "learning_rate": 8e-05,
+      "loss": 1.8704,
+      "step": 6657
+    },
+    {
+      "epoch": 0.37112597547380155,
+      "grad_norm": 0.4913389980792999,
+      "learning_rate": 8e-05,
+      "loss": 1.6487,
+      "step": 6658
+    },
+    {
+      "epoch": 0.37118171683389073,
+      "grad_norm": 0.501367449760437,
+      "learning_rate": 8e-05,
+      "loss": 1.6942,
+      "step": 6659
+    },
+    {
+      "epoch": 0.3712374581939799,
+      "grad_norm": 0.5529482960700989,
+      "learning_rate": 8e-05,
+      "loss": 1.336,
+      "step": 6660
+    },
+    {
+      "epoch": 0.37129319955406914,
+      "grad_norm": 0.5055440664291382,
+      "learning_rate": 8e-05,
+      "loss": 1.3953,
+      "step": 6661
+    },
+    {
+      "epoch": 0.3713489409141583,
+      "grad_norm": 0.497901052236557,
+      "learning_rate": 8e-05,
+      "loss": 1.5928,
+      "step": 6662
+    },
+    {
+      "epoch": 0.3714046822742475,
+      "grad_norm": 0.5057888031005859,
+      "learning_rate": 8e-05,
+      "loss": 1.7644,
+      "step": 6663
+    },
+    {
+      "epoch": 0.3714604236343367,
+      "grad_norm": 0.5101109147071838,
+      "learning_rate": 8e-05,
+      "loss": 1.5919,
+      "step": 6664
+    },
+    {
+      "epoch": 0.37151616499442586,
+      "grad_norm": 0.4964488744735718,
+      "learning_rate": 8e-05,
+      "loss": 1.6232,
+      "step": 6665
+    },
+    {
+      "epoch": 0.37157190635451504,
+      "grad_norm": 0.5438522696495056,
+      "learning_rate": 8e-05,
+      "loss": 1.8482,
+      "step": 6666
+    },
+    {
+      "epoch": 0.3716276477146042,
+      "grad_norm": 0.5171390175819397,
+      "learning_rate": 8e-05,
+      "loss": 1.5641,
+      "step": 6667
+    },
+    {
+      "epoch": 0.37168338907469345,
+      "grad_norm": 0.6002838611602783,
+      "learning_rate": 8e-05,
+      "loss": 2.0131,
+      "step": 6668
+    },
+    {
+      "epoch": 0.3717391304347826,
+      "grad_norm": 0.5407944321632385,
+      "learning_rate": 8e-05,
+      "loss": 1.7647,
+      "step": 6669
+    },
+    {
+      "epoch": 0.3717948717948718,
+      "grad_norm": 0.5388062596321106,
+      "learning_rate": 8e-05,
+      "loss": 1.7898,
+      "step": 6670
+    },
+    {
+      "epoch": 0.371850613154961,
+      "grad_norm": 0.49192193150520325,
+      "learning_rate": 8e-05,
+      "loss": 1.6744,
+      "step": 6671
+    },
+    {
+      "epoch": 0.37190635451505016,
+      "grad_norm": 0.5632032752037048,
+      "learning_rate": 8e-05,
+      "loss": 1.946,
+      "step": 6672
+    },
+    {
+      "epoch": 0.37196209587513934,
+      "grad_norm": 0.5415680408477783,
+      "learning_rate": 8e-05,
+      "loss": 1.6364,
+      "step": 6673
+    },
+    {
+      "epoch": 0.3720178372352285,
+      "grad_norm": 0.5188789367675781,
+      "learning_rate": 8e-05,
+      "loss": 1.4743,
+      "step": 6674
+    },
+    {
+      "epoch": 0.3720735785953177,
+      "grad_norm": 0.5327577590942383,
+      "learning_rate": 8e-05,
+      "loss": 1.7789,
+      "step": 6675
+    },
+    {
+      "epoch": 0.37212931995540693,
+      "grad_norm": 0.5899271368980408,
+      "learning_rate": 8e-05,
+      "loss": 1.9106,
+      "step": 6676
+    },
+    {
+      "epoch": 0.3721850613154961,
+      "grad_norm": 0.5051175355911255,
+      "learning_rate": 8e-05,
+      "loss": 1.6377,
+      "step": 6677
+    },
+    {
+      "epoch": 0.3722408026755853,
+      "grad_norm": 0.5093581676483154,
+      "learning_rate": 8e-05,
+      "loss": 1.6424,
+      "step": 6678
+    },
+    {
+      "epoch": 0.37229654403567447,
+      "grad_norm": 0.5512845516204834,
+      "learning_rate": 8e-05,
+      "loss": 1.6879,
+      "step": 6679
+    },
+    {
+      "epoch": 0.37235228539576365,
+      "grad_norm": 0.5061362385749817,
+      "learning_rate": 8e-05,
+      "loss": 1.9053,
+      "step": 6680
+    },
+    {
+      "epoch": 0.37240802675585283,
+      "grad_norm": 0.48424118757247925,
+      "learning_rate": 8e-05,
+      "loss": 1.3633,
+      "step": 6681
+    },
+    {
+      "epoch": 0.372463768115942,
+      "grad_norm": 0.5293522477149963,
+      "learning_rate": 8e-05,
+      "loss": 1.6503,
+      "step": 6682
+    },
+    {
+      "epoch": 0.37251950947603124,
+      "grad_norm": 0.4804575443267822,
+      "learning_rate": 8e-05,
+      "loss": 1.3826,
+      "step": 6683
+    },
+    {
+      "epoch": 0.3725752508361204,
+      "grad_norm": 0.5213553309440613,
+      "learning_rate": 8e-05,
+      "loss": 1.9151,
+      "step": 6684
+    },
+    {
+      "epoch": 0.3726309921962096,
+      "grad_norm": 0.5248057842254639,
+      "learning_rate": 8e-05,
+      "loss": 1.8208,
+      "step": 6685
+    },
+    {
+      "epoch": 0.3726867335562988,
+      "grad_norm": 0.6494954824447632,
+      "learning_rate": 8e-05,
+      "loss": 1.5739,
+      "step": 6686
+    },
+    {
+      "epoch": 0.37274247491638796,
+      "grad_norm": 0.5033650994300842,
+      "learning_rate": 8e-05,
+      "loss": 1.5725,
+      "step": 6687
+    },
+    {
+      "epoch": 0.37279821627647713,
+      "grad_norm": 0.5093197226524353,
+      "learning_rate": 8e-05,
+      "loss": 1.7493,
+      "step": 6688
+    },
+    {
+      "epoch": 0.3728539576365663,
+      "grad_norm": 0.485580712556839,
+      "learning_rate": 8e-05,
+      "loss": 1.5371,
+      "step": 6689
+    },
+    {
+      "epoch": 0.3729096989966555,
+      "grad_norm": 0.5913149118423462,
+      "learning_rate": 8e-05,
+      "loss": 1.7906,
+      "step": 6690
+    },
+    {
+      "epoch": 0.3729654403567447,
+      "grad_norm": 0.5249662399291992,
+      "learning_rate": 8e-05,
+      "loss": 1.6659,
+      "step": 6691
+    },
+    {
+      "epoch": 0.3730211817168339,
+      "grad_norm": 0.5243616700172424,
+      "learning_rate": 8e-05,
+      "loss": 1.6433,
+      "step": 6692
+    },
+    {
+      "epoch": 0.3730769230769231,
+      "grad_norm": 0.5441550612449646,
+      "learning_rate": 8e-05,
+      "loss": 1.7376,
+      "step": 6693
+    },
+    {
+      "epoch": 0.37313266443701226,
+      "grad_norm": 0.5584784746170044,
+      "learning_rate": 8e-05,
+      "loss": 1.834,
+      "step": 6694
+    },
+    {
+      "epoch": 0.37318840579710144,
+      "grad_norm": 0.5328128933906555,
+      "learning_rate": 8e-05,
+      "loss": 1.7491,
+      "step": 6695
+    },
+    {
+      "epoch": 0.3732441471571906,
+      "grad_norm": 0.5425513982772827,
+      "learning_rate": 8e-05,
+      "loss": 1.8401,
+      "step": 6696
+    },
+    {
+      "epoch": 0.3732998885172798,
+      "grad_norm": 0.5354945063591003,
+      "learning_rate": 8e-05,
+      "loss": 1.626,
+      "step": 6697
+    },
+    {
+      "epoch": 0.37335562987736903,
+      "grad_norm": 0.5174336433410645,
+      "learning_rate": 8e-05,
+      "loss": 1.6046,
+      "step": 6698
+    },
+    {
+      "epoch": 0.3734113712374582,
+      "grad_norm": 0.6389894485473633,
+      "learning_rate": 8e-05,
+      "loss": 1.9224,
+      "step": 6699
+    },
+    {
+      "epoch": 0.3734671125975474,
+      "grad_norm": 0.4668770134449005,
+      "learning_rate": 8e-05,
+      "loss": 1.4228,
+      "step": 6700
+    },
+    {
+      "epoch": 0.37352285395763657,
+      "grad_norm": 0.46430280804634094,
+      "learning_rate": 8e-05,
+      "loss": 1.585,
+      "step": 6701
+    },
+    {
+      "epoch": 0.37357859531772575,
+      "grad_norm": 0.5275490880012512,
+      "learning_rate": 8e-05,
+      "loss": 1.7373,
+      "step": 6702
+    },
+    {
+      "epoch": 0.3736343366778149,
+      "grad_norm": 0.5056618452072144,
+      "learning_rate": 8e-05,
+      "loss": 1.6823,
+      "step": 6703
+    },
+    {
+      "epoch": 0.3736900780379041,
+      "grad_norm": 0.5731016993522644,
+      "learning_rate": 8e-05,
+      "loss": 1.8575,
+      "step": 6704
+    },
+    {
+      "epoch": 0.3737458193979933,
+      "grad_norm": 0.5207464098930359,
+      "learning_rate": 8e-05,
+      "loss": 1.5405,
+      "step": 6705
+    },
+    {
+      "epoch": 0.3738015607580825,
+      "grad_norm": 0.49173352122306824,
+      "learning_rate": 8e-05,
+      "loss": 1.4675,
+      "step": 6706
+    },
+    {
+      "epoch": 0.3738573021181717,
+      "grad_norm": 0.5099393129348755,
+      "learning_rate": 8e-05,
+      "loss": 1.582,
+      "step": 6707
+    },
+    {
+      "epoch": 0.3739130434782609,
+      "grad_norm": 0.5217099785804749,
+      "learning_rate": 8e-05,
+      "loss": 1.6141,
+      "step": 6708
+    },
+    {
+      "epoch": 0.37396878483835005,
+      "grad_norm": 0.5156376957893372,
+      "learning_rate": 8e-05,
+      "loss": 1.8233,
+      "step": 6709
+    },
+    {
+      "epoch": 0.37402452619843923,
+      "grad_norm": 0.4843328893184662,
+      "learning_rate": 8e-05,
+      "loss": 1.3098,
+      "step": 6710
+    },
+    {
+      "epoch": 0.3740802675585284,
+      "grad_norm": 0.5049893856048584,
+      "learning_rate": 8e-05,
+      "loss": 1.5087,
+      "step": 6711
+    },
+    {
+      "epoch": 0.3741360089186176,
+      "grad_norm": 0.5323360562324524,
+      "learning_rate": 8e-05,
+      "loss": 1.8198,
+      "step": 6712
+    },
+    {
+      "epoch": 0.3741917502787068,
+      "grad_norm": 0.5711098313331604,
+      "learning_rate": 8e-05,
+      "loss": 1.5649,
+      "step": 6713
+    },
+    {
+      "epoch": 0.374247491638796,
+      "grad_norm": 0.538690984249115,
+      "learning_rate": 8e-05,
+      "loss": 1.4958,
+      "step": 6714
+    },
+    {
+      "epoch": 0.3743032329988852,
+      "grad_norm": 0.5207656621932983,
+      "learning_rate": 8e-05,
+      "loss": 1.5492,
+      "step": 6715
+    },
+    {
+      "epoch": 0.37435897435897436,
+      "grad_norm": 0.5162418484687805,
+      "learning_rate": 8e-05,
+      "loss": 1.695,
+      "step": 6716
+    },
+    {
+      "epoch": 0.37441471571906354,
+      "grad_norm": 0.5555497407913208,
+      "learning_rate": 8e-05,
+      "loss": 1.7426,
+      "step": 6717
+    },
+    {
+      "epoch": 0.3744704570791527,
+      "grad_norm": 0.5439112782478333,
+      "learning_rate": 8e-05,
+      "loss": 1.6575,
+      "step": 6718
+    },
+    {
+      "epoch": 0.3745261984392419,
+      "grad_norm": 0.5380793809890747,
+      "learning_rate": 8e-05,
+      "loss": 1.7351,
+      "step": 6719
+    },
+    {
+      "epoch": 0.3745819397993311,
+      "grad_norm": 0.5121132731437683,
+      "learning_rate": 8e-05,
+      "loss": 1.544,
+      "step": 6720
+    },
+    {
+      "epoch": 0.3746376811594203,
+      "grad_norm": 0.5033418536186218,
+      "learning_rate": 8e-05,
+      "loss": 1.544,
+      "step": 6721
+    },
+    {
+      "epoch": 0.3746934225195095,
+      "grad_norm": 0.5126864314079285,
+      "learning_rate": 8e-05,
+      "loss": 1.6579,
+      "step": 6722
+    },
+    {
+      "epoch": 0.37474916387959867,
+      "grad_norm": 0.5681543350219727,
+      "learning_rate": 8e-05,
+      "loss": 1.7232,
+      "step": 6723
+    },
+    {
+      "epoch": 0.37480490523968785,
+      "grad_norm": 0.5577385425567627,
+      "learning_rate": 8e-05,
+      "loss": 1.5906,
+      "step": 6724
+    },
+    {
+      "epoch": 0.374860646599777,
+      "grad_norm": 0.5277745723724365,
+      "learning_rate": 8e-05,
+      "loss": 1.6748,
+      "step": 6725
+    },
+    {
+      "epoch": 0.3749163879598662,
+      "grad_norm": 0.5073549747467041,
+      "learning_rate": 8e-05,
+      "loss": 1.3841,
+      "step": 6726
+    },
+    {
+      "epoch": 0.3749721293199554,
+      "grad_norm": 0.5121777653694153,
+      "learning_rate": 8e-05,
+      "loss": 1.6002,
+      "step": 6727
+    },
+    {
+      "epoch": 0.3750278706800446,
+      "grad_norm": 0.5338273644447327,
+      "learning_rate": 8e-05,
+      "loss": 1.6037,
+      "step": 6728
+    },
+    {
+      "epoch": 0.3750836120401338,
+      "grad_norm": 0.550675630569458,
+      "learning_rate": 8e-05,
+      "loss": 1.7317,
+      "step": 6729
+    },
+    {
+      "epoch": 0.375139353400223,
+      "grad_norm": 0.5400722026824951,
+      "learning_rate": 8e-05,
+      "loss": 1.8084,
+      "step": 6730
+    },
+    {
+      "epoch": 0.37519509476031215,
+      "grad_norm": 0.5145625472068787,
+      "learning_rate": 8e-05,
+      "loss": 1.4093,
+      "step": 6731
+    },
+    {
+      "epoch": 0.37525083612040133,
+      "grad_norm": 0.6066479086875916,
+      "learning_rate": 8e-05,
+      "loss": 1.9785,
+      "step": 6732
+    },
+    {
+      "epoch": 0.3753065774804905,
+      "grad_norm": 0.5051720142364502,
+      "learning_rate": 8e-05,
+      "loss": 1.7888,
+      "step": 6733
+    },
+    {
+      "epoch": 0.3753623188405797,
+      "grad_norm": 0.5142165422439575,
+      "learning_rate": 8e-05,
+      "loss": 1.6872,
+      "step": 6734
+    },
+    {
+      "epoch": 0.3754180602006689,
+      "grad_norm": 0.5311247706413269,
+      "learning_rate": 8e-05,
+      "loss": 1.6527,
+      "step": 6735
+    },
+    {
+      "epoch": 0.3754738015607581,
+      "grad_norm": 0.5546833872795105,
+      "learning_rate": 8e-05,
+      "loss": 1.8813,
+      "step": 6736
+    },
+    {
+      "epoch": 0.3755295429208473,
+      "grad_norm": 0.5479986071586609,
+      "learning_rate": 8e-05,
+      "loss": 1.6774,
+      "step": 6737
+    },
+    {
+      "epoch": 0.37558528428093646,
+      "grad_norm": 0.5656196475028992,
+      "learning_rate": 8e-05,
+      "loss": 1.9011,
+      "step": 6738
+    },
+    {
+      "epoch": 0.37564102564102564,
+      "grad_norm": 0.5137097239494324,
+      "learning_rate": 8e-05,
+      "loss": 1.7263,
+      "step": 6739
+    },
+    {
+      "epoch": 0.3756967670011148,
+      "grad_norm": 0.5318892598152161,
+      "learning_rate": 8e-05,
+      "loss": 1.75,
+      "step": 6740
+    },
+    {
+      "epoch": 0.375752508361204,
+      "grad_norm": 0.5124412178993225,
+      "learning_rate": 8e-05,
+      "loss": 1.6559,
+      "step": 6741
+    },
+    {
+      "epoch": 0.3758082497212932,
+      "grad_norm": 0.5239470601081848,
+      "learning_rate": 8e-05,
+      "loss": 1.7523,
+      "step": 6742
+    },
+    {
+      "epoch": 0.3758639910813824,
+      "grad_norm": 0.5155774354934692,
+      "learning_rate": 8e-05,
+      "loss": 1.7701,
+      "step": 6743
+    },
+    {
+      "epoch": 0.3759197324414716,
+      "grad_norm": 0.4978020191192627,
+      "learning_rate": 8e-05,
+      "loss": 1.5988,
+      "step": 6744
+    },
+    {
+      "epoch": 0.37597547380156077,
+      "grad_norm": 0.4940713047981262,
+      "learning_rate": 8e-05,
+      "loss": 1.78,
+      "step": 6745
+    },
+    {
+      "epoch": 0.37603121516164995,
+      "grad_norm": 0.4915592670440674,
+      "learning_rate": 8e-05,
+      "loss": 1.6901,
+      "step": 6746
+    },
+    {
+      "epoch": 0.3760869565217391,
+      "grad_norm": 0.5367061495780945,
+      "learning_rate": 8e-05,
+      "loss": 1.5299,
+      "step": 6747
+    },
+    {
+      "epoch": 0.3761426978818283,
+      "grad_norm": 0.5281707644462585,
+      "learning_rate": 8e-05,
+      "loss": 1.6676,
+      "step": 6748
+    },
+    {
+      "epoch": 0.3761984392419175,
+      "grad_norm": 0.5502864122390747,
+      "learning_rate": 8e-05,
+      "loss": 1.7959,
+      "step": 6749
+    },
+    {
+      "epoch": 0.3762541806020067,
+      "grad_norm": 0.48569023609161377,
+      "learning_rate": 8e-05,
+      "loss": 1.5379,
+      "step": 6750
+    },
+    {
+      "epoch": 0.3763099219620959,
+      "grad_norm": 0.5276660919189453,
+      "learning_rate": 8e-05,
+      "loss": 1.5918,
+      "step": 6751
+    },
+    {
+      "epoch": 0.3763656633221851,
+      "grad_norm": 0.5254923701286316,
+      "learning_rate": 8e-05,
+      "loss": 1.6608,
+      "step": 6752
+    },
+    {
+      "epoch": 0.37642140468227425,
+      "grad_norm": 0.5019527077674866,
+      "learning_rate": 8e-05,
+      "loss": 1.4723,
+      "step": 6753
+    },
+    {
+      "epoch": 0.37647714604236343,
+      "grad_norm": 0.5919958353042603,
+      "learning_rate": 8e-05,
+      "loss": 1.8617,
+      "step": 6754
+    },
+    {
+      "epoch": 0.3765328874024526,
+      "grad_norm": 0.5578399300575256,
+      "learning_rate": 8e-05,
+      "loss": 1.8262,
+      "step": 6755
+    },
+    {
+      "epoch": 0.3765886287625418,
+      "grad_norm": 0.5366002917289734,
+      "learning_rate": 8e-05,
+      "loss": 1.7505,
+      "step": 6756
+    },
+    {
+      "epoch": 0.37664437012263097,
+      "grad_norm": 0.5503619909286499,
+      "learning_rate": 8e-05,
+      "loss": 1.7405,
+      "step": 6757
+    },
+    {
+      "epoch": 0.3767001114827202,
+      "grad_norm": 0.5201141238212585,
+      "learning_rate": 8e-05,
+      "loss": 1.8242,
+      "step": 6758
+    },
+    {
+      "epoch": 0.3767558528428094,
+      "grad_norm": 0.5590203404426575,
+      "learning_rate": 8e-05,
+      "loss": 1.7174,
+      "step": 6759
+    },
+    {
+      "epoch": 0.37681159420289856,
+      "grad_norm": 0.5271986722946167,
+      "learning_rate": 8e-05,
+      "loss": 1.7366,
+      "step": 6760
+    },
+    {
+      "epoch": 0.37686733556298774,
+      "grad_norm": 0.49543046951293945,
+      "learning_rate": 8e-05,
+      "loss": 1.5968,
+      "step": 6761
+    },
+    {
+      "epoch": 0.3769230769230769,
+      "grad_norm": 0.4976285398006439,
+      "learning_rate": 8e-05,
+      "loss": 1.7351,
+      "step": 6762
+    },
+    {
+      "epoch": 0.3769788182831661,
+      "grad_norm": 0.49661460518836975,
+      "learning_rate": 8e-05,
+      "loss": 1.4948,
+      "step": 6763
+    },
+    {
+      "epoch": 0.3770345596432553,
+      "grad_norm": 0.5244497656822205,
+      "learning_rate": 8e-05,
+      "loss": 1.8646,
+      "step": 6764
+    },
+    {
+      "epoch": 0.3770903010033445,
+      "grad_norm": 0.5240281224250793,
+      "learning_rate": 8e-05,
+      "loss": 1.8795,
+      "step": 6765
+    },
+    {
+      "epoch": 0.3771460423634337,
+      "grad_norm": 0.5917136073112488,
+      "learning_rate": 8e-05,
+      "loss": 2.0025,
+      "step": 6766
+    },
+    {
+      "epoch": 0.37720178372352287,
+      "grad_norm": 0.5583435893058777,
+      "learning_rate": 8e-05,
+      "loss": 1.9267,
+      "step": 6767
+    },
+    {
+      "epoch": 0.37725752508361204,
+      "grad_norm": 0.5000171065330505,
+      "learning_rate": 8e-05,
+      "loss": 1.6172,
+      "step": 6768
+    },
+    {
+      "epoch": 0.3773132664437012,
+      "grad_norm": 0.5241192579269409,
+      "learning_rate": 8e-05,
+      "loss": 1.5843,
+      "step": 6769
+    },
+    {
+      "epoch": 0.3773690078037904,
+      "grad_norm": 0.5357404947280884,
+      "learning_rate": 8e-05,
+      "loss": 1.8782,
+      "step": 6770
+    },
+    {
+      "epoch": 0.3774247491638796,
+      "grad_norm": 0.5429965257644653,
+      "learning_rate": 8e-05,
+      "loss": 1.7346,
+      "step": 6771
+    },
+    {
+      "epoch": 0.37748049052396876,
+      "grad_norm": 0.5275600552558899,
+      "learning_rate": 8e-05,
+      "loss": 1.603,
+      "step": 6772
+    },
+    {
+      "epoch": 0.377536231884058,
+      "grad_norm": 0.49835145473480225,
+      "learning_rate": 8e-05,
+      "loss": 1.8675,
+      "step": 6773
+    },
+    {
+      "epoch": 0.37759197324414717,
+      "grad_norm": 0.5513197183609009,
+      "learning_rate": 8e-05,
+      "loss": 1.905,
+      "step": 6774
+    },
+    {
+      "epoch": 0.37764771460423635,
+      "grad_norm": 0.48942214250564575,
+      "learning_rate": 8e-05,
+      "loss": 1.6214,
+      "step": 6775
+    },
+    {
+      "epoch": 0.37770345596432553,
+      "grad_norm": 0.45932549238204956,
+      "learning_rate": 8e-05,
+      "loss": 1.369,
+      "step": 6776
+    },
+    {
+      "epoch": 0.3777591973244147,
+      "grad_norm": 0.7865234613418579,
+      "learning_rate": 8e-05,
+      "loss": 1.7867,
+      "step": 6777
+    },
+    {
+      "epoch": 0.3778149386845039,
+      "grad_norm": 0.5142460465431213,
+      "learning_rate": 8e-05,
+      "loss": 1.7239,
+      "step": 6778
+    },
+    {
+      "epoch": 0.37787068004459307,
+      "grad_norm": 0.5079149007797241,
+      "learning_rate": 8e-05,
+      "loss": 1.6643,
+      "step": 6779
+    },
+    {
+      "epoch": 0.3779264214046823,
+      "grad_norm": 0.518399715423584,
+      "learning_rate": 8e-05,
+      "loss": 1.57,
+      "step": 6780
+    },
+    {
+      "epoch": 0.3779821627647715,
+      "grad_norm": 0.5325887799263,
+      "learning_rate": 8e-05,
+      "loss": 1.7435,
+      "step": 6781
+    },
+    {
+      "epoch": 0.37803790412486066,
+      "grad_norm": 0.5779329538345337,
+      "learning_rate": 8e-05,
+      "loss": 1.7795,
+      "step": 6782
+    },
+    {
+      "epoch": 0.37809364548494984,
+      "grad_norm": 0.5537606477737427,
+      "learning_rate": 8e-05,
+      "loss": 1.6176,
+      "step": 6783
+    },
+    {
+      "epoch": 0.378149386845039,
+      "grad_norm": 0.5910248160362244,
+      "learning_rate": 8e-05,
+      "loss": 1.7853,
+      "step": 6784
+    },
+    {
+      "epoch": 0.3782051282051282,
+      "grad_norm": 0.5300013422966003,
+      "learning_rate": 8e-05,
+      "loss": 1.5137,
+      "step": 6785
+    },
+    {
+      "epoch": 0.3782608695652174,
+      "grad_norm": 0.5802608132362366,
+      "learning_rate": 8e-05,
+      "loss": 1.7391,
+      "step": 6786
+    },
+    {
+      "epoch": 0.37831661092530655,
+      "grad_norm": 0.520869255065918,
+      "learning_rate": 8e-05,
+      "loss": 1.6134,
+      "step": 6787
+    },
+    {
+      "epoch": 0.3783723522853958,
+      "grad_norm": 0.5778328776359558,
+      "learning_rate": 8e-05,
+      "loss": 1.8975,
+      "step": 6788
+    },
+    {
+      "epoch": 0.37842809364548496,
+      "grad_norm": 0.5080073475837708,
+      "learning_rate": 8e-05,
+      "loss": 1.5634,
+      "step": 6789
+    },
+    {
+      "epoch": 0.37848383500557414,
+      "grad_norm": 0.5058872699737549,
+      "learning_rate": 8e-05,
+      "loss": 1.8211,
+      "step": 6790
+    },
+    {
+      "epoch": 0.3785395763656633,
+      "grad_norm": 0.5163450837135315,
+      "learning_rate": 8e-05,
+      "loss": 1.7062,
+      "step": 6791
+    },
+    {
+      "epoch": 0.3785953177257525,
+      "grad_norm": 0.5297355651855469,
+      "learning_rate": 8e-05,
+      "loss": 1.715,
+      "step": 6792
+    },
+    {
+      "epoch": 0.3786510590858417,
+      "grad_norm": 0.5093514919281006,
+      "learning_rate": 8e-05,
+      "loss": 1.7468,
+      "step": 6793
+    },
+    {
+      "epoch": 0.37870680044593086,
+      "grad_norm": 0.5215211510658264,
+      "learning_rate": 8e-05,
+      "loss": 1.504,
+      "step": 6794
+    },
+    {
+      "epoch": 0.3787625418060201,
+      "grad_norm": 0.5218589901924133,
+      "learning_rate": 8e-05,
+      "loss": 1.7833,
+      "step": 6795
+    },
+    {
+      "epoch": 0.37881828316610927,
+      "grad_norm": 0.5648636817932129,
+      "learning_rate": 8e-05,
+      "loss": 1.7638,
+      "step": 6796
+    },
+    {
+      "epoch": 0.37887402452619845,
+      "grad_norm": 0.5256296992301941,
+      "learning_rate": 8e-05,
+      "loss": 1.7571,
+      "step": 6797
+    },
+    {
+      "epoch": 0.37892976588628763,
+      "grad_norm": 0.4967195391654968,
+      "learning_rate": 8e-05,
+      "loss": 1.7674,
+      "step": 6798
+    },
+    {
+      "epoch": 0.3789855072463768,
+      "grad_norm": 0.5004242658615112,
+      "learning_rate": 8e-05,
+      "loss": 1.6802,
+      "step": 6799
+    },
+    {
+      "epoch": 0.379041248606466,
+      "grad_norm": 0.5254338979721069,
+      "learning_rate": 8e-05,
+      "loss": 1.7531,
+      "step": 6800
+    },
+    {
+      "epoch": 0.37909698996655516,
+      "grad_norm": 0.5258433222770691,
+      "learning_rate": 8e-05,
+      "loss": 1.6567,
+      "step": 6801
+    },
+    {
+      "epoch": 0.37915273132664434,
+      "grad_norm": 0.5207537412643433,
+      "learning_rate": 8e-05,
+      "loss": 1.6734,
+      "step": 6802
+    },
+    {
+      "epoch": 0.3792084726867336,
+      "grad_norm": 0.5069604516029358,
+      "learning_rate": 8e-05,
+      "loss": 1.5255,
+      "step": 6803
+    },
+    {
+      "epoch": 0.37926421404682276,
+      "grad_norm": 0.5184929966926575,
+      "learning_rate": 8e-05,
+      "loss": 1.6458,
+      "step": 6804
+    },
+    {
+      "epoch": 0.37931995540691194,
+      "grad_norm": 0.5532615184783936,
+      "learning_rate": 8e-05,
+      "loss": 1.6805,
+      "step": 6805
+    },
+    {
+      "epoch": 0.3793756967670011,
+      "grad_norm": 0.49667033553123474,
+      "learning_rate": 8e-05,
+      "loss": 1.512,
+      "step": 6806
+    },
+    {
+      "epoch": 0.3794314381270903,
+      "grad_norm": 0.5174059271812439,
+      "learning_rate": 8e-05,
+      "loss": 1.6957,
+      "step": 6807
+    },
+    {
+      "epoch": 0.37948717948717947,
+      "grad_norm": 0.48986855149269104,
+      "learning_rate": 8e-05,
+      "loss": 1.5954,
+      "step": 6808
+    },
+    {
+      "epoch": 0.37954292084726865,
+      "grad_norm": 0.529532790184021,
+      "learning_rate": 8e-05,
+      "loss": 1.6124,
+      "step": 6809
+    },
+    {
+      "epoch": 0.3795986622073579,
+      "grad_norm": 0.564482569694519,
+      "learning_rate": 8e-05,
+      "loss": 1.8692,
+      "step": 6810
+    },
+    {
+      "epoch": 0.37965440356744706,
+      "grad_norm": 0.5190215706825256,
+      "learning_rate": 8e-05,
+      "loss": 1.6718,
+      "step": 6811
+    },
+    {
+      "epoch": 0.37971014492753624,
+      "grad_norm": 0.5358736515045166,
+      "learning_rate": 8e-05,
+      "loss": 1.7521,
+      "step": 6812
+    },
+    {
+      "epoch": 0.3797658862876254,
+      "grad_norm": 0.5234962701797485,
+      "learning_rate": 8e-05,
+      "loss": 1.948,
+      "step": 6813
+    },
+    {
+      "epoch": 0.3798216276477146,
+      "grad_norm": 0.5307388305664062,
+      "learning_rate": 8e-05,
+      "loss": 1.6149,
+      "step": 6814
+    },
+    {
+      "epoch": 0.3798773690078038,
+      "grad_norm": 0.5353745222091675,
+      "learning_rate": 8e-05,
+      "loss": 1.8958,
+      "step": 6815
+    },
+    {
+      "epoch": 0.37993311036789296,
+      "grad_norm": 0.5111658573150635,
+      "learning_rate": 8e-05,
+      "loss": 1.5686,
+      "step": 6816
+    },
+    {
+      "epoch": 0.37998885172798214,
+      "grad_norm": 0.5570499300956726,
+      "learning_rate": 8e-05,
+      "loss": 1.7011,
+      "step": 6817
+    },
+    {
+      "epoch": 0.38004459308807137,
+      "grad_norm": 0.5320193767547607,
+      "learning_rate": 8e-05,
+      "loss": 1.641,
+      "step": 6818
+    },
+    {
+      "epoch": 0.38010033444816055,
+      "grad_norm": 0.498035192489624,
+      "learning_rate": 8e-05,
+      "loss": 1.6247,
+      "step": 6819
+    },
+    {
+      "epoch": 0.3801560758082497,
+      "grad_norm": 0.5418937802314758,
+      "learning_rate": 8e-05,
+      "loss": 1.8077,
+      "step": 6820
+    },
+    {
+      "epoch": 0.3802118171683389,
+      "grad_norm": 0.5447816252708435,
+      "learning_rate": 8e-05,
+      "loss": 1.7445,
+      "step": 6821
+    },
+    {
+      "epoch": 0.3802675585284281,
+      "grad_norm": 0.535697340965271,
+      "learning_rate": 8e-05,
+      "loss": 1.7219,
+      "step": 6822
+    },
+    {
+      "epoch": 0.38032329988851726,
+      "grad_norm": 0.543074369430542,
+      "learning_rate": 8e-05,
+      "loss": 1.6885,
+      "step": 6823
+    },
+    {
+      "epoch": 0.38037904124860644,
+      "grad_norm": 0.4806436002254486,
+      "learning_rate": 8e-05,
+      "loss": 1.5289,
+      "step": 6824
+    },
+    {
+      "epoch": 0.3804347826086957,
+      "grad_norm": 0.4711361825466156,
+      "learning_rate": 8e-05,
+      "loss": 1.4881,
+      "step": 6825
+    },
+    {
+      "epoch": 0.38049052396878486,
+      "grad_norm": 0.5814215540885925,
+      "learning_rate": 8e-05,
+      "loss": 1.6463,
+      "step": 6826
+    },
+    {
+      "epoch": 0.38054626532887403,
+      "grad_norm": 0.5161845684051514,
+      "learning_rate": 8e-05,
+      "loss": 1.7158,
+      "step": 6827
+    },
+    {
+      "epoch": 0.3806020066889632,
+      "grad_norm": 0.5395570993423462,
+      "learning_rate": 8e-05,
+      "loss": 1.6931,
+      "step": 6828
+    },
+    {
+      "epoch": 0.3806577480490524,
+      "grad_norm": 0.47877243161201477,
+      "learning_rate": 8e-05,
+      "loss": 1.4561,
+      "step": 6829
+    },
+    {
+      "epoch": 0.38071348940914157,
+      "grad_norm": 0.5157475471496582,
+      "learning_rate": 8e-05,
+      "loss": 1.6233,
+      "step": 6830
+    },
+    {
+      "epoch": 0.38076923076923075,
+      "grad_norm": 0.47996920347213745,
+      "learning_rate": 8e-05,
+      "loss": 1.747,
+      "step": 6831
+    },
+    {
+      "epoch": 0.38082497212932,
+      "grad_norm": 0.5504299402236938,
+      "learning_rate": 8e-05,
+      "loss": 1.8097,
+      "step": 6832
+    },
+    {
+      "epoch": 0.38088071348940916,
+      "grad_norm": 0.5216795802116394,
+      "learning_rate": 8e-05,
+      "loss": 1.6967,
+      "step": 6833
+    },
+    {
+      "epoch": 0.38093645484949834,
+      "grad_norm": 0.5156157612800598,
+      "learning_rate": 8e-05,
+      "loss": 1.6106,
+      "step": 6834
+    },
+    {
+      "epoch": 0.3809921962095875,
+      "grad_norm": 0.5302389860153198,
+      "learning_rate": 8e-05,
+      "loss": 1.6393,
+      "step": 6835
+    },
+    {
+      "epoch": 0.3810479375696767,
+      "grad_norm": 0.6114341020584106,
+      "learning_rate": 8e-05,
+      "loss": 1.7414,
+      "step": 6836
+    },
+    {
+      "epoch": 0.3811036789297659,
+      "grad_norm": 0.589827835559845,
+      "learning_rate": 8e-05,
+      "loss": 1.8041,
+      "step": 6837
+    },
+    {
+      "epoch": 0.38115942028985506,
+      "grad_norm": 0.5321907997131348,
+      "learning_rate": 8e-05,
+      "loss": 1.871,
+      "step": 6838
+    },
+    {
+      "epoch": 0.38121516164994423,
+      "grad_norm": 0.5287553071975708,
+      "learning_rate": 8e-05,
+      "loss": 1.6608,
+      "step": 6839
+    },
+    {
+      "epoch": 0.38127090301003347,
+      "grad_norm": 0.5082924962043762,
+      "learning_rate": 8e-05,
+      "loss": 1.6053,
+      "step": 6840
+    },
+    {
+      "epoch": 0.38132664437012265,
+      "grad_norm": 0.5522058010101318,
+      "learning_rate": 8e-05,
+      "loss": 2.0149,
+      "step": 6841
+    },
+    {
+      "epoch": 0.3813823857302118,
+      "grad_norm": 0.5427312850952148,
+      "learning_rate": 8e-05,
+      "loss": 1.6895,
+      "step": 6842
+    },
+    {
+      "epoch": 0.381438127090301,
+      "grad_norm": 0.49162036180496216,
+      "learning_rate": 8e-05,
+      "loss": 1.5884,
+      "step": 6843
+    },
+    {
+      "epoch": 0.3814938684503902,
+      "grad_norm": 0.5044377446174622,
+      "learning_rate": 8e-05,
+      "loss": 1.8749,
+      "step": 6844
+    },
+    {
+      "epoch": 0.38154960981047936,
+      "grad_norm": 0.5048327445983887,
+      "learning_rate": 8e-05,
+      "loss": 1.7494,
+      "step": 6845
+    },
+    {
+      "epoch": 0.38160535117056854,
+      "grad_norm": 0.5744604468345642,
+      "learning_rate": 8e-05,
+      "loss": 2.0678,
+      "step": 6846
+    },
+    {
+      "epoch": 0.3816610925306578,
+      "grad_norm": 0.5219566226005554,
+      "learning_rate": 8e-05,
+      "loss": 1.685,
+      "step": 6847
+    },
+    {
+      "epoch": 0.38171683389074695,
+      "grad_norm": 0.5850833654403687,
+      "learning_rate": 8e-05,
+      "loss": 1.8884,
+      "step": 6848
+    },
+    {
+      "epoch": 0.38177257525083613,
+      "grad_norm": 0.510691225528717,
+      "learning_rate": 8e-05,
+      "loss": 1.6582,
+      "step": 6849
+    },
+    {
+      "epoch": 0.3818283166109253,
+      "grad_norm": 0.5360484719276428,
+      "learning_rate": 8e-05,
+      "loss": 1.8495,
+      "step": 6850
+    },
+    {
+      "epoch": 0.3818840579710145,
+      "grad_norm": 0.5596208572387695,
+      "learning_rate": 8e-05,
+      "loss": 1.8184,
+      "step": 6851
+    },
+    {
+      "epoch": 0.38193979933110367,
+      "grad_norm": 0.5178501009941101,
+      "learning_rate": 8e-05,
+      "loss": 1.6066,
+      "step": 6852
+    },
+    {
+      "epoch": 0.38199554069119285,
+      "grad_norm": 0.535118043422699,
+      "learning_rate": 8e-05,
+      "loss": 1.6646,
+      "step": 6853
+    },
+    {
+      "epoch": 0.382051282051282,
+      "grad_norm": 0.5289649963378906,
+      "learning_rate": 8e-05,
+      "loss": 1.8509,
+      "step": 6854
+    },
+    {
+      "epoch": 0.38210702341137126,
+      "grad_norm": 0.5231328010559082,
+      "learning_rate": 8e-05,
+      "loss": 1.6752,
+      "step": 6855
+    },
+    {
+      "epoch": 0.38216276477146044,
+      "grad_norm": 0.5241426825523376,
+      "learning_rate": 8e-05,
+      "loss": 1.6721,
+      "step": 6856
+    },
+    {
+      "epoch": 0.3822185061315496,
+      "grad_norm": 0.5819863677024841,
+      "learning_rate": 8e-05,
+      "loss": 2.0239,
+      "step": 6857
+    },
+    {
+      "epoch": 0.3822742474916388,
+      "grad_norm": 0.5138490200042725,
+      "learning_rate": 8e-05,
+      "loss": 1.6158,
+      "step": 6858
+    },
+    {
+      "epoch": 0.382329988851728,
+      "grad_norm": 0.5236282348632812,
+      "learning_rate": 8e-05,
+      "loss": 1.7532,
+      "step": 6859
+    },
+    {
+      "epoch": 0.38238573021181715,
+      "grad_norm": 0.503381609916687,
+      "learning_rate": 8e-05,
+      "loss": 1.454,
+      "step": 6860
+    },
+    {
+      "epoch": 0.38244147157190633,
+      "grad_norm": 0.5174334049224854,
+      "learning_rate": 8e-05,
+      "loss": 1.5932,
+      "step": 6861
+    },
+    {
+      "epoch": 0.38249721293199557,
+      "grad_norm": 0.5292040705680847,
+      "learning_rate": 8e-05,
+      "loss": 1.7081,
+      "step": 6862
+    },
+    {
+      "epoch": 0.38255295429208475,
+      "grad_norm": 0.5251057147979736,
+      "learning_rate": 8e-05,
+      "loss": 1.8752,
+      "step": 6863
+    },
+    {
+      "epoch": 0.3826086956521739,
+      "grad_norm": 0.5487580895423889,
+      "learning_rate": 8e-05,
+      "loss": 1.8481,
+      "step": 6864
+    },
+    {
+      "epoch": 0.3826644370122631,
+      "grad_norm": 0.5405415892601013,
+      "learning_rate": 8e-05,
+      "loss": 1.9514,
+      "step": 6865
+    },
+    {
+      "epoch": 0.3827201783723523,
+      "grad_norm": 0.517487645149231,
+      "learning_rate": 8e-05,
+      "loss": 1.5871,
+      "step": 6866
+    },
+    {
+      "epoch": 0.38277591973244146,
+      "grad_norm": 0.5205005407333374,
+      "learning_rate": 8e-05,
+      "loss": 1.8664,
+      "step": 6867
+    },
+    {
+      "epoch": 0.38283166109253064,
+      "grad_norm": 0.5625116229057312,
+      "learning_rate": 8e-05,
+      "loss": 1.6272,
+      "step": 6868
+    },
+    {
+      "epoch": 0.3828874024526198,
+      "grad_norm": 0.5461683869361877,
+      "learning_rate": 8e-05,
+      "loss": 1.7246,
+      "step": 6869
+    },
+    {
+      "epoch": 0.38294314381270905,
+      "grad_norm": 0.5413379073143005,
+      "learning_rate": 8e-05,
+      "loss": 1.7691,
+      "step": 6870
+    },
+    {
+      "epoch": 0.38299888517279823,
+      "grad_norm": 0.498203843832016,
+      "learning_rate": 8e-05,
+      "loss": 1.3738,
+      "step": 6871
+    },
+    {
+      "epoch": 0.3830546265328874,
+      "grad_norm": 0.5851048827171326,
+      "learning_rate": 8e-05,
+      "loss": 1.8351,
+      "step": 6872
+    },
+    {
+      "epoch": 0.3831103678929766,
+      "grad_norm": 0.693454921245575,
+      "learning_rate": 8e-05,
+      "loss": 1.7745,
+      "step": 6873
+    },
+    {
+      "epoch": 0.38316610925306577,
+      "grad_norm": 0.5207720994949341,
+      "learning_rate": 8e-05,
+      "loss": 1.6717,
+      "step": 6874
+    },
+    {
+      "epoch": 0.38322185061315495,
+      "grad_norm": 0.5102624297142029,
+      "learning_rate": 8e-05,
+      "loss": 1.7611,
+      "step": 6875
+    },
+    {
+      "epoch": 0.3832775919732441,
+      "grad_norm": 0.4984310567378998,
+      "learning_rate": 8e-05,
+      "loss": 1.6392,
+      "step": 6876
+    },
+    {
+      "epoch": 0.38333333333333336,
+      "grad_norm": 0.529915988445282,
+      "learning_rate": 8e-05,
+      "loss": 1.8102,
+      "step": 6877
+    },
+    {
+      "epoch": 0.38338907469342254,
+      "grad_norm": 0.50824373960495,
+      "learning_rate": 8e-05,
+      "loss": 1.553,
+      "step": 6878
+    },
+    {
+      "epoch": 0.3834448160535117,
+      "grad_norm": 0.5229014754295349,
+      "learning_rate": 8e-05,
+      "loss": 1.67,
+      "step": 6879
+    },
+    {
+      "epoch": 0.3835005574136009,
+      "grad_norm": 0.5327182412147522,
+      "learning_rate": 8e-05,
+      "loss": 1.8009,
+      "step": 6880
+    },
+    {
+      "epoch": 0.3835562987736901,
+      "grad_norm": 0.5739976167678833,
+      "learning_rate": 8e-05,
+      "loss": 1.5915,
+      "step": 6881
+    },
+    {
+      "epoch": 0.38361204013377925,
+      "grad_norm": 0.5743573307991028,
+      "learning_rate": 8e-05,
+      "loss": 1.8215,
+      "step": 6882
+    },
+    {
+      "epoch": 0.38366778149386843,
+      "grad_norm": 0.5695149302482605,
+      "learning_rate": 8e-05,
+      "loss": 1.8339,
+      "step": 6883
+    },
+    {
+      "epoch": 0.3837235228539576,
+      "grad_norm": 0.5400159955024719,
+      "learning_rate": 8e-05,
+      "loss": 1.6151,
+      "step": 6884
+    },
+    {
+      "epoch": 0.38377926421404684,
+      "grad_norm": 0.5263397693634033,
+      "learning_rate": 8e-05,
+      "loss": 1.7196,
+      "step": 6885
+    },
+    {
+      "epoch": 0.383835005574136,
+      "grad_norm": 0.4755968451499939,
+      "learning_rate": 8e-05,
+      "loss": 1.6788,
+      "step": 6886
+    },
+    {
+      "epoch": 0.3838907469342252,
+      "grad_norm": 0.542471706867218,
+      "learning_rate": 8e-05,
+      "loss": 1.716,
+      "step": 6887
+    },
+    {
+      "epoch": 0.3839464882943144,
+      "grad_norm": 0.5046743154525757,
+      "learning_rate": 8e-05,
+      "loss": 1.6081,
+      "step": 6888
+    },
+    {
+      "epoch": 0.38400222965440356,
+      "grad_norm": 0.5250272750854492,
+      "learning_rate": 8e-05,
+      "loss": 1.8675,
+      "step": 6889
+    },
+    {
+      "epoch": 0.38405797101449274,
+      "grad_norm": 0.5886393785476685,
+      "learning_rate": 8e-05,
+      "loss": 1.9787,
+      "step": 6890
+    },
+    {
+      "epoch": 0.3841137123745819,
+      "grad_norm": 0.5748692154884338,
+      "learning_rate": 8e-05,
+      "loss": 1.7817,
+      "step": 6891
+    },
+    {
+      "epoch": 0.38416945373467115,
+      "grad_norm": 0.5253850817680359,
+      "learning_rate": 8e-05,
+      "loss": 1.535,
+      "step": 6892
+    },
+    {
+      "epoch": 0.38422519509476033,
+      "grad_norm": 0.5223969221115112,
+      "learning_rate": 8e-05,
+      "loss": 1.6191,
+      "step": 6893
+    },
+    {
+      "epoch": 0.3842809364548495,
+      "grad_norm": 0.5565198659896851,
+      "learning_rate": 8e-05,
+      "loss": 1.7,
+      "step": 6894
+    },
+    {
+      "epoch": 0.3843366778149387,
+      "grad_norm": 0.5230225324630737,
+      "learning_rate": 8e-05,
+      "loss": 1.6912,
+      "step": 6895
+    },
+    {
+      "epoch": 0.38439241917502787,
+      "grad_norm": 0.4867858290672302,
+      "learning_rate": 8e-05,
+      "loss": 1.4106,
+      "step": 6896
+    },
+    {
+      "epoch": 0.38444816053511705,
+      "grad_norm": 0.5560249090194702,
+      "learning_rate": 8e-05,
+      "loss": 1.78,
+      "step": 6897
+    },
+    {
+      "epoch": 0.3845039018952062,
+      "grad_norm": 0.5637099742889404,
+      "learning_rate": 8e-05,
+      "loss": 1.8057,
+      "step": 6898
+    },
+    {
+      "epoch": 0.3845596432552954,
+      "grad_norm": 0.5312997102737427,
+      "learning_rate": 8e-05,
+      "loss": 1.646,
+      "step": 6899
+    },
+    {
+      "epoch": 0.38461538461538464,
+      "grad_norm": 0.516556441783905,
+      "learning_rate": 8e-05,
+      "loss": 1.815,
+      "step": 6900
+    },
+    {
+      "epoch": 0.3846711259754738,
+      "grad_norm": 0.49218010902404785,
+      "learning_rate": 8e-05,
+      "loss": 1.5516,
+      "step": 6901
+    },
+    {
+      "epoch": 0.384726867335563,
+      "grad_norm": 0.5333677530288696,
+      "learning_rate": 8e-05,
+      "loss": 1.7512,
+      "step": 6902
+    },
+    {
+      "epoch": 0.3847826086956522,
+      "grad_norm": 0.5032493472099304,
+      "learning_rate": 8e-05,
+      "loss": 1.5828,
+      "step": 6903
+    },
+    {
+      "epoch": 0.38483835005574135,
+      "grad_norm": 0.5718285441398621,
+      "learning_rate": 8e-05,
+      "loss": 1.9966,
+      "step": 6904
+    },
+    {
+      "epoch": 0.38489409141583053,
+      "grad_norm": 0.5370532274246216,
+      "learning_rate": 8e-05,
+      "loss": 1.7849,
+      "step": 6905
+    },
+    {
+      "epoch": 0.3849498327759197,
+      "grad_norm": 0.5484476089477539,
+      "learning_rate": 8e-05,
+      "loss": 1.6092,
+      "step": 6906
+    },
+    {
+      "epoch": 0.38500557413600894,
+      "grad_norm": 0.5581532120704651,
+      "learning_rate": 8e-05,
+      "loss": 1.4118,
+      "step": 6907
+    },
+    {
+      "epoch": 0.3850613154960981,
+      "grad_norm": 0.47904351353645325,
+      "learning_rate": 8e-05,
+      "loss": 1.31,
+      "step": 6908
+    },
+    {
+      "epoch": 0.3851170568561873,
+      "grad_norm": 0.5527108311653137,
+      "learning_rate": 8e-05,
+      "loss": 1.7555,
+      "step": 6909
+    },
+    {
+      "epoch": 0.3851727982162765,
+      "grad_norm": 0.5678485631942749,
+      "learning_rate": 8e-05,
+      "loss": 1.6078,
+      "step": 6910
+    },
+    {
+      "epoch": 0.38522853957636566,
+      "grad_norm": 0.5740782022476196,
+      "learning_rate": 8e-05,
+      "loss": 1.9674,
+      "step": 6911
+    },
+    {
+      "epoch": 0.38528428093645484,
+      "grad_norm": 0.5345740914344788,
+      "learning_rate": 8e-05,
+      "loss": 1.6925,
+      "step": 6912
+    },
+    {
+      "epoch": 0.385340022296544,
+      "grad_norm": 0.5236703753471375,
+      "learning_rate": 8e-05,
+      "loss": 1.6034,
+      "step": 6913
+    },
+    {
+      "epoch": 0.3853957636566332,
+      "grad_norm": 0.6058480739593506,
+      "learning_rate": 8e-05,
+      "loss": 1.6309,
+      "step": 6914
+    },
+    {
+      "epoch": 0.38545150501672243,
+      "grad_norm": 0.5399019122123718,
+      "learning_rate": 8e-05,
+      "loss": 1.6592,
+      "step": 6915
+    },
+    {
+      "epoch": 0.3855072463768116,
+      "grad_norm": 0.5349549651145935,
+      "learning_rate": 8e-05,
+      "loss": 1.4853,
+      "step": 6916
+    },
+    {
+      "epoch": 0.3855629877369008,
+      "grad_norm": 0.5857842564582825,
+      "learning_rate": 8e-05,
+      "loss": 1.749,
+      "step": 6917
+    },
+    {
+      "epoch": 0.38561872909698997,
+      "grad_norm": 0.5716098546981812,
+      "learning_rate": 8e-05,
+      "loss": 1.7058,
+      "step": 6918
+    },
+    {
+      "epoch": 0.38567447045707914,
+      "grad_norm": 0.5078105330467224,
+      "learning_rate": 8e-05,
+      "loss": 1.6227,
+      "step": 6919
+    },
+    {
+      "epoch": 0.3857302118171683,
+      "grad_norm": 0.5407280921936035,
+      "learning_rate": 8e-05,
+      "loss": 1.7624,
+      "step": 6920
+    },
+    {
+      "epoch": 0.3857859531772575,
+      "grad_norm": 0.47967904806137085,
+      "learning_rate": 8e-05,
+      "loss": 1.6142,
+      "step": 6921
+    },
+    {
+      "epoch": 0.38584169453734674,
+      "grad_norm": 0.5285566449165344,
+      "learning_rate": 8e-05,
+      "loss": 1.7721,
+      "step": 6922
+    },
+    {
+      "epoch": 0.3858974358974359,
+      "grad_norm": 0.5036161541938782,
+      "learning_rate": 8e-05,
+      "loss": 1.7378,
+      "step": 6923
+    },
+    {
+      "epoch": 0.3859531772575251,
+      "grad_norm": 0.4978451132774353,
+      "learning_rate": 8e-05,
+      "loss": 1.8718,
+      "step": 6924
+    },
+    {
+      "epoch": 0.38600891861761427,
+      "grad_norm": 0.5122378468513489,
+      "learning_rate": 8e-05,
+      "loss": 1.658,
+      "step": 6925
+    },
+    {
+      "epoch": 0.38606465997770345,
+      "grad_norm": 0.5669604539871216,
+      "learning_rate": 8e-05,
+      "loss": 1.9589,
+      "step": 6926
+    },
+    {
+      "epoch": 0.38612040133779263,
+      "grad_norm": 0.5110654234886169,
+      "learning_rate": 8e-05,
+      "loss": 1.6009,
+      "step": 6927
+    },
+    {
+      "epoch": 0.3861761426978818,
+      "grad_norm": 0.5333113670349121,
+      "learning_rate": 8e-05,
+      "loss": 1.6887,
+      "step": 6928
+    },
+    {
+      "epoch": 0.38623188405797104,
+      "grad_norm": 0.5038242340087891,
+      "learning_rate": 8e-05,
+      "loss": 1.5244,
+      "step": 6929
+    },
+    {
+      "epoch": 0.3862876254180602,
+      "grad_norm": 0.5251333713531494,
+      "learning_rate": 8e-05,
+      "loss": 1.7132,
+      "step": 6930
+    },
+    {
+      "epoch": 0.3863433667781494,
+      "grad_norm": 0.5106619000434875,
+      "learning_rate": 8e-05,
+      "loss": 1.6036,
+      "step": 6931
+    },
+    {
+      "epoch": 0.3863991081382386,
+      "grad_norm": 0.5227919816970825,
+      "learning_rate": 8e-05,
+      "loss": 1.8909,
+      "step": 6932
+    },
+    {
+      "epoch": 0.38645484949832776,
+      "grad_norm": 0.5325038433074951,
+      "learning_rate": 8e-05,
+      "loss": 1.6279,
+      "step": 6933
+    },
+    {
+      "epoch": 0.38651059085841694,
+      "grad_norm": 0.5435842275619507,
+      "learning_rate": 8e-05,
+      "loss": 1.6574,
+      "step": 6934
+    },
+    {
+      "epoch": 0.3865663322185061,
+      "grad_norm": 0.5161607265472412,
+      "learning_rate": 8e-05,
+      "loss": 1.6595,
+      "step": 6935
+    },
+    {
+      "epoch": 0.3866220735785953,
+      "grad_norm": 0.5756670236587524,
+      "learning_rate": 8e-05,
+      "loss": 1.7544,
+      "step": 6936
+    },
+    {
+      "epoch": 0.3866778149386845,
+      "grad_norm": 0.5451154112815857,
+      "learning_rate": 8e-05,
+      "loss": 1.761,
+      "step": 6937
+    },
+    {
+      "epoch": 0.3867335562987737,
+      "grad_norm": 0.5060601830482483,
+      "learning_rate": 8e-05,
+      "loss": 1.8706,
+      "step": 6938
+    },
+    {
+      "epoch": 0.3867892976588629,
+      "grad_norm": 0.5163840055465698,
+      "learning_rate": 8e-05,
+      "loss": 1.7337,
+      "step": 6939
+    },
+    {
+      "epoch": 0.38684503901895206,
+      "grad_norm": 0.5261103510856628,
+      "learning_rate": 8e-05,
+      "loss": 1.6595,
+      "step": 6940
+    },
+    {
+      "epoch": 0.38690078037904124,
+      "grad_norm": 0.5704602003097534,
+      "learning_rate": 8e-05,
+      "loss": 1.738,
+      "step": 6941
+    },
+    {
+      "epoch": 0.3869565217391304,
+      "grad_norm": 0.5094172954559326,
+      "learning_rate": 8e-05,
+      "loss": 1.7202,
+      "step": 6942
+    },
+    {
+      "epoch": 0.3870122630992196,
+      "grad_norm": 0.5927731394767761,
+      "learning_rate": 8e-05,
+      "loss": 1.9517,
+      "step": 6943
+    },
+    {
+      "epoch": 0.38706800445930883,
+      "grad_norm": 0.5634392499923706,
+      "learning_rate": 8e-05,
+      "loss": 1.6567,
+      "step": 6944
+    },
+    {
+      "epoch": 0.387123745819398,
+      "grad_norm": 0.5229656100273132,
+      "learning_rate": 8e-05,
+      "loss": 1.6196,
+      "step": 6945
+    },
+    {
+      "epoch": 0.3871794871794872,
+      "grad_norm": 0.5949238538742065,
+      "learning_rate": 8e-05,
+      "loss": 1.9731,
+      "step": 6946
+    },
+    {
+      "epoch": 0.38723522853957637,
+      "grad_norm": 0.4674675166606903,
+      "learning_rate": 8e-05,
+      "loss": 1.5122,
+      "step": 6947
+    },
+    {
+      "epoch": 0.38729096989966555,
+      "grad_norm": 0.5145415663719177,
+      "learning_rate": 8e-05,
+      "loss": 1.5898,
+      "step": 6948
+    },
+    {
+      "epoch": 0.38734671125975473,
+      "grad_norm": 0.5329121351242065,
+      "learning_rate": 8e-05,
+      "loss": 1.6453,
+      "step": 6949
+    },
+    {
+      "epoch": 0.3874024526198439,
+      "grad_norm": 0.5534684658050537,
+      "learning_rate": 8e-05,
+      "loss": 1.9438,
+      "step": 6950
+    },
+    {
+      "epoch": 0.3874581939799331,
+      "grad_norm": 0.5507635474205017,
+      "learning_rate": 8e-05,
+      "loss": 1.7876,
+      "step": 6951
+    },
+    {
+      "epoch": 0.3875139353400223,
+      "grad_norm": 0.5376663208007812,
+      "learning_rate": 8e-05,
+      "loss": 1.7594,
+      "step": 6952
+    },
+    {
+      "epoch": 0.3875696767001115,
+      "grad_norm": 0.5363755822181702,
+      "learning_rate": 8e-05,
+      "loss": 1.7429,
+      "step": 6953
+    },
+    {
+      "epoch": 0.3876254180602007,
+      "grad_norm": 0.5718921422958374,
+      "learning_rate": 8e-05,
+      "loss": 1.7017,
+      "step": 6954
+    },
+    {
+      "epoch": 0.38768115942028986,
+      "grad_norm": 0.555400013923645,
+      "learning_rate": 8e-05,
+      "loss": 1.7949,
+      "step": 6955
+    },
+    {
+      "epoch": 0.38773690078037903,
+      "grad_norm": 0.49279558658599854,
+      "learning_rate": 8e-05,
+      "loss": 1.5243,
+      "step": 6956
+    },
+    {
+      "epoch": 0.3877926421404682,
+      "grad_norm": 0.5423111319541931,
+      "learning_rate": 8e-05,
+      "loss": 1.48,
+      "step": 6957
+    },
+    {
+      "epoch": 0.3878483835005574,
+      "grad_norm": 0.4846994876861572,
+      "learning_rate": 8e-05,
+      "loss": 1.574,
+      "step": 6958
+    },
+    {
+      "epoch": 0.3879041248606466,
+      "grad_norm": 0.5224431157112122,
+      "learning_rate": 8e-05,
+      "loss": 1.6236,
+      "step": 6959
+    },
+    {
+      "epoch": 0.3879598662207358,
+      "grad_norm": 0.5335124135017395,
+      "learning_rate": 8e-05,
+      "loss": 1.7444,
+      "step": 6960
+    },
+    {
+      "epoch": 0.388015607580825,
+      "grad_norm": 0.5206796526908875,
+      "learning_rate": 8e-05,
+      "loss": 1.8457,
+      "step": 6961
+    },
+    {
+      "epoch": 0.38807134894091416,
+      "grad_norm": 0.5847195386886597,
+      "learning_rate": 8e-05,
+      "loss": 1.9442,
+      "step": 6962
+    },
+    {
+      "epoch": 0.38812709030100334,
+      "grad_norm": 0.4814377725124359,
+      "learning_rate": 8e-05,
+      "loss": 1.4659,
+      "step": 6963
+    },
+    {
+      "epoch": 0.3881828316610925,
+      "grad_norm": 0.4975985884666443,
+      "learning_rate": 8e-05,
+      "loss": 1.5948,
+      "step": 6964
+    },
+    {
+      "epoch": 0.3882385730211817,
+      "grad_norm": 0.5406726002693176,
+      "learning_rate": 8e-05,
+      "loss": 1.7501,
+      "step": 6965
+    },
+    {
+      "epoch": 0.3882943143812709,
+      "grad_norm": 0.5482088923454285,
+      "learning_rate": 8e-05,
+      "loss": 1.8163,
+      "step": 6966
+    },
+    {
+      "epoch": 0.3883500557413601,
+      "grad_norm": 0.5207962989807129,
+      "learning_rate": 8e-05,
+      "loss": 1.6247,
+      "step": 6967
+    },
+    {
+      "epoch": 0.3884057971014493,
+      "grad_norm": 0.5449426770210266,
+      "learning_rate": 8e-05,
+      "loss": 1.8502,
+      "step": 6968
+    },
+    {
+      "epoch": 0.38846153846153847,
+      "grad_norm": 0.6138060688972473,
+      "learning_rate": 8e-05,
+      "loss": 2.0023,
+      "step": 6969
+    },
+    {
+      "epoch": 0.38851727982162765,
+      "grad_norm": 0.5436367988586426,
+      "learning_rate": 8e-05,
+      "loss": 1.6418,
+      "step": 6970
+    },
+    {
+      "epoch": 0.3885730211817168,
+      "grad_norm": 0.5104245543479919,
+      "learning_rate": 8e-05,
+      "loss": 1.7074,
+      "step": 6971
+    },
+    {
+      "epoch": 0.388628762541806,
+      "grad_norm": 0.5240477919578552,
+      "learning_rate": 8e-05,
+      "loss": 1.7272,
+      "step": 6972
+    },
+    {
+      "epoch": 0.3886845039018952,
+      "grad_norm": 0.5188844203948975,
+      "learning_rate": 8e-05,
+      "loss": 1.6639,
+      "step": 6973
+    },
+    {
+      "epoch": 0.3887402452619844,
+      "grad_norm": 0.5346798300743103,
+      "learning_rate": 8e-05,
+      "loss": 1.88,
+      "step": 6974
+    },
+    {
+      "epoch": 0.3887959866220736,
+      "grad_norm": 0.5370174050331116,
+      "learning_rate": 8e-05,
+      "loss": 1.6109,
+      "step": 6975
+    },
+    {
+      "epoch": 0.3888517279821628,
+      "grad_norm": 0.497453510761261,
+      "learning_rate": 8e-05,
+      "loss": 1.4677,
+      "step": 6976
+    },
+    {
+      "epoch": 0.38890746934225195,
+      "grad_norm": 0.5281513333320618,
+      "learning_rate": 8e-05,
+      "loss": 1.7995,
+      "step": 6977
+    },
+    {
+      "epoch": 0.38896321070234113,
+      "grad_norm": 0.5266953706741333,
+      "learning_rate": 8e-05,
+      "loss": 1.7261,
+      "step": 6978
+    },
+    {
+      "epoch": 0.3890189520624303,
+      "grad_norm": 0.4925101399421692,
+      "learning_rate": 8e-05,
+      "loss": 1.6437,
+      "step": 6979
+    },
+    {
+      "epoch": 0.3890746934225195,
+      "grad_norm": 0.5393596291542053,
+      "learning_rate": 8e-05,
+      "loss": 1.7864,
+      "step": 6980
+    },
+    {
+      "epoch": 0.38913043478260867,
+      "grad_norm": 0.537490963935852,
+      "learning_rate": 8e-05,
+      "loss": 1.6231,
+      "step": 6981
+    },
+    {
+      "epoch": 0.3891861761426979,
+      "grad_norm": 0.5116598606109619,
+      "learning_rate": 8e-05,
+      "loss": 1.605,
+      "step": 6982
+    },
+    {
+      "epoch": 0.3892419175027871,
+      "grad_norm": 0.559950590133667,
+      "learning_rate": 8e-05,
+      "loss": 1.541,
+      "step": 6983
+    },
+    {
+      "epoch": 0.38929765886287626,
+      "grad_norm": 0.4906415343284607,
+      "learning_rate": 8e-05,
+      "loss": 1.3907,
+      "step": 6984
+    },
+    {
+      "epoch": 0.38935340022296544,
+      "grad_norm": 0.5318133234977722,
+      "learning_rate": 8e-05,
+      "loss": 1.9337,
+      "step": 6985
+    },
+    {
+      "epoch": 0.3894091415830546,
+      "grad_norm": 0.5461350083351135,
+      "learning_rate": 8e-05,
+      "loss": 1.7324,
+      "step": 6986
+    },
+    {
+      "epoch": 0.3894648829431438,
+      "grad_norm": 0.5531414151191711,
+      "learning_rate": 8e-05,
+      "loss": 1.7923,
+      "step": 6987
+    },
+    {
+      "epoch": 0.389520624303233,
+      "grad_norm": 0.49649012088775635,
+      "learning_rate": 8e-05,
+      "loss": 1.5552,
+      "step": 6988
+    },
+    {
+      "epoch": 0.3895763656633222,
+      "grad_norm": 0.48604315519332886,
+      "learning_rate": 8e-05,
+      "loss": 1.6701,
+      "step": 6989
+    },
+    {
+      "epoch": 0.3896321070234114,
+      "grad_norm": 0.5059698820114136,
+      "learning_rate": 8e-05,
+      "loss": 1.5653,
+      "step": 6990
+    },
+    {
+      "epoch": 0.38968784838350057,
+      "grad_norm": 0.5029332637786865,
+      "learning_rate": 8e-05,
+      "loss": 1.5919,
+      "step": 6991
+    },
+    {
+      "epoch": 0.38974358974358975,
+      "grad_norm": 0.5438098907470703,
+      "learning_rate": 8e-05,
+      "loss": 1.7432,
+      "step": 6992
+    },
+    {
+      "epoch": 0.3897993311036789,
+      "grad_norm": 0.4913785755634308,
+      "learning_rate": 8e-05,
+      "loss": 1.5916,
+      "step": 6993
+    },
+    {
+      "epoch": 0.3898550724637681,
+      "grad_norm": 0.47481903433799744,
+      "learning_rate": 8e-05,
+      "loss": 1.5417,
+      "step": 6994
+    },
+    {
+      "epoch": 0.3899108138238573,
+      "grad_norm": 0.5361449122428894,
+      "learning_rate": 8e-05,
+      "loss": 1.9055,
+      "step": 6995
+    },
+    {
+      "epoch": 0.38996655518394646,
+      "grad_norm": 0.5341494083404541,
+      "learning_rate": 8e-05,
+      "loss": 1.7104,
+      "step": 6996
+    },
+    {
+      "epoch": 0.3900222965440357,
+      "grad_norm": 0.5641982555389404,
+      "learning_rate": 8e-05,
+      "loss": 1.9479,
+      "step": 6997
+    },
+    {
+      "epoch": 0.3900780379041249,
+      "grad_norm": 0.5358514189720154,
+      "learning_rate": 8e-05,
+      "loss": 1.7432,
+      "step": 6998
+    },
+    {
+      "epoch": 0.39013377926421405,
+      "grad_norm": 0.5224436521530151,
+      "learning_rate": 8e-05,
+      "loss": 1.8123,
+      "step": 6999
+    },
+    {
+      "epoch": 0.39018952062430323,
+      "grad_norm": 0.4992210268974304,
+      "learning_rate": 8e-05,
+      "loss": 1.8403,
+      "step": 7000
+    },
+    {
+      "epoch": 0.3902452619843924,
+      "grad_norm": 0.5072396397590637,
+      "learning_rate": 8e-05,
+      "loss": 1.6148,
+      "step": 7001
+    },
+    {
+      "epoch": 0.3903010033444816,
+      "grad_norm": 0.5484432578086853,
+      "learning_rate": 8e-05,
+      "loss": 1.7956,
+      "step": 7002
+    },
+    {
+      "epoch": 0.39035674470457077,
+      "grad_norm": 0.502047598361969,
+      "learning_rate": 8e-05,
+      "loss": 1.6104,
+      "step": 7003
+    },
+    {
+      "epoch": 0.39041248606466,
+      "grad_norm": 0.5068725943565369,
+      "learning_rate": 8e-05,
+      "loss": 1.6485,
+      "step": 7004
+    },
+    {
+      "epoch": 0.3904682274247492,
+      "grad_norm": 0.49112656712532043,
+      "learning_rate": 8e-05,
+      "loss": 1.557,
+      "step": 7005
+    },
+    {
+      "epoch": 0.39052396878483836,
+      "grad_norm": 0.5446252226829529,
+      "learning_rate": 8e-05,
+      "loss": 1.6979,
+      "step": 7006
+    },
+    {
+      "epoch": 0.39057971014492754,
+      "grad_norm": 0.47817152738571167,
+      "learning_rate": 8e-05,
+      "loss": 1.5078,
+      "step": 7007
+    },
+    {
+      "epoch": 0.3906354515050167,
+      "grad_norm": 0.4783325493335724,
+      "learning_rate": 8e-05,
+      "loss": 1.5962,
+      "step": 7008
+    },
+    {
+      "epoch": 0.3906911928651059,
+      "grad_norm": 0.5405673980712891,
+      "learning_rate": 8e-05,
+      "loss": 1.7521,
+      "step": 7009
+    },
+    {
+      "epoch": 0.3907469342251951,
+      "grad_norm": 0.5000050067901611,
+      "learning_rate": 8e-05,
+      "loss": 1.613,
+      "step": 7010
+    },
+    {
+      "epoch": 0.39080267558528425,
+      "grad_norm": 0.5314992666244507,
+      "learning_rate": 8e-05,
+      "loss": 1.6482,
+      "step": 7011
+    },
+    {
+      "epoch": 0.3908584169453735,
+      "grad_norm": 0.49853751063346863,
+      "learning_rate": 8e-05,
+      "loss": 1.6888,
+      "step": 7012
+    },
+    {
+      "epoch": 0.39091415830546267,
+      "grad_norm": 0.6656997203826904,
+      "learning_rate": 8e-05,
+      "loss": 2.1214,
+      "step": 7013
+    },
+    {
+      "epoch": 0.39096989966555185,
+      "grad_norm": 0.5526180863380432,
+      "learning_rate": 8e-05,
+      "loss": 1.9484,
+      "step": 7014
+    },
+    {
+      "epoch": 0.391025641025641,
+      "grad_norm": 0.5586705803871155,
+      "learning_rate": 8e-05,
+      "loss": 1.8734,
+      "step": 7015
+    },
+    {
+      "epoch": 0.3910813823857302,
+      "grad_norm": 0.5152546763420105,
+      "learning_rate": 8e-05,
+      "loss": 1.7155,
+      "step": 7016
+    },
+    {
+      "epoch": 0.3911371237458194,
+      "grad_norm": 0.5178968906402588,
+      "learning_rate": 8e-05,
+      "loss": 1.6688,
+      "step": 7017
+    },
+    {
+      "epoch": 0.39119286510590856,
+      "grad_norm": 0.5010287165641785,
+      "learning_rate": 8e-05,
+      "loss": 1.6412,
+      "step": 7018
+    },
+    {
+      "epoch": 0.3912486064659978,
+      "grad_norm": 0.49992191791534424,
+      "learning_rate": 8e-05,
+      "loss": 1.7122,
+      "step": 7019
+    },
+    {
+      "epoch": 0.391304347826087,
+      "grad_norm": 0.5425409078598022,
+      "learning_rate": 8e-05,
+      "loss": 1.846,
+      "step": 7020
+    },
+    {
+      "epoch": 0.39136008918617615,
+      "grad_norm": 0.4892907738685608,
+      "learning_rate": 8e-05,
+      "loss": 1.6649,
+      "step": 7021
+    },
+    {
+      "epoch": 0.39141583054626533,
+      "grad_norm": 0.5175873637199402,
+      "learning_rate": 8e-05,
+      "loss": 1.6877,
+      "step": 7022
+    },
+    {
+      "epoch": 0.3914715719063545,
+      "grad_norm": 0.5006701946258545,
+      "learning_rate": 8e-05,
+      "loss": 1.6376,
+      "step": 7023
+    },
+    {
+      "epoch": 0.3915273132664437,
+      "grad_norm": 0.5196325778961182,
+      "learning_rate": 8e-05,
+      "loss": 1.616,
+      "step": 7024
+    },
+    {
+      "epoch": 0.39158305462653287,
+      "grad_norm": 0.5506092309951782,
+      "learning_rate": 8e-05,
+      "loss": 1.8984,
+      "step": 7025
+    },
+    {
+      "epoch": 0.39163879598662205,
+      "grad_norm": 0.4763076603412628,
+      "learning_rate": 8e-05,
+      "loss": 1.5839,
+      "step": 7026
+    },
+    {
+      "epoch": 0.3916945373467113,
+      "grad_norm": 0.5732759833335876,
+      "learning_rate": 8e-05,
+      "loss": 1.6241,
+      "step": 7027
+    },
+    {
+      "epoch": 0.39175027870680046,
+      "grad_norm": 0.4679046869277954,
+      "learning_rate": 8e-05,
+      "loss": 1.4369,
+      "step": 7028
+    },
+    {
+      "epoch": 0.39180602006688964,
+      "grad_norm": 0.48116782307624817,
+      "learning_rate": 8e-05,
+      "loss": 1.4916,
+      "step": 7029
+    },
+    {
+      "epoch": 0.3918617614269788,
+      "grad_norm": 0.5586615800857544,
+      "learning_rate": 8e-05,
+      "loss": 1.8211,
+      "step": 7030
+    },
+    {
+      "epoch": 0.391917502787068,
+      "grad_norm": 0.49537360668182373,
+      "learning_rate": 8e-05,
+      "loss": 1.6,
+      "step": 7031
+    },
+    {
+      "epoch": 0.3919732441471572,
+      "grad_norm": 0.5766502618789673,
+      "learning_rate": 8e-05,
+      "loss": 1.7938,
+      "step": 7032
+    },
+    {
+      "epoch": 0.39202898550724635,
+      "grad_norm": 0.5225504636764526,
+      "learning_rate": 8e-05,
+      "loss": 1.7192,
+      "step": 7033
+    },
+    {
+      "epoch": 0.3920847268673356,
+      "grad_norm": 0.5227795243263245,
+      "learning_rate": 8e-05,
+      "loss": 1.5299,
+      "step": 7034
+    },
+    {
+      "epoch": 0.39214046822742477,
+      "grad_norm": 0.5982847213745117,
+      "learning_rate": 8e-05,
+      "loss": 2.0238,
+      "step": 7035
+    },
+    {
+      "epoch": 0.39219620958751394,
+      "grad_norm": 0.5800369381904602,
+      "learning_rate": 8e-05,
+      "loss": 1.6763,
+      "step": 7036
+    },
+    {
+      "epoch": 0.3922519509476031,
+      "grad_norm": 0.567676305770874,
+      "learning_rate": 8e-05,
+      "loss": 1.3901,
+      "step": 7037
+    },
+    {
+      "epoch": 0.3923076923076923,
+      "grad_norm": 0.5485690236091614,
+      "learning_rate": 8e-05,
+      "loss": 1.8169,
+      "step": 7038
+    },
+    {
+      "epoch": 0.3923634336677815,
+      "grad_norm": 0.5342855453491211,
+      "learning_rate": 8e-05,
+      "loss": 1.6875,
+      "step": 7039
+    },
+    {
+      "epoch": 0.39241917502787066,
+      "grad_norm": 0.4871984124183655,
+      "learning_rate": 8e-05,
+      "loss": 1.5819,
+      "step": 7040
+    },
+    {
+      "epoch": 0.3924749163879599,
+      "grad_norm": 0.5237048864364624,
+      "learning_rate": 8e-05,
+      "loss": 1.7995,
+      "step": 7041
+    },
+    {
+      "epoch": 0.39253065774804907,
+      "grad_norm": 0.5661356449127197,
+      "learning_rate": 8e-05,
+      "loss": 2.0471,
+      "step": 7042
+    },
+    {
+      "epoch": 0.39258639910813825,
+      "grad_norm": 0.49327969551086426,
+      "learning_rate": 8e-05,
+      "loss": 1.6038,
+      "step": 7043
+    },
+    {
+      "epoch": 0.39264214046822743,
+      "grad_norm": 0.5772015452384949,
+      "learning_rate": 8e-05,
+      "loss": 1.7543,
+      "step": 7044
+    },
+    {
+      "epoch": 0.3926978818283166,
+      "grad_norm": 0.5665246844291687,
+      "learning_rate": 8e-05,
+      "loss": 1.806,
+      "step": 7045
+    },
+    {
+      "epoch": 0.3927536231884058,
+      "grad_norm": 0.49184301495552063,
+      "learning_rate": 8e-05,
+      "loss": 1.5393,
+      "step": 7046
+    },
+    {
+      "epoch": 0.39280936454849497,
+      "grad_norm": 0.5374985337257385,
+      "learning_rate": 8e-05,
+      "loss": 1.7513,
+      "step": 7047
+    },
+    {
+      "epoch": 0.39286510590858414,
+      "grad_norm": 0.5255217552185059,
+      "learning_rate": 8e-05,
+      "loss": 1.7094,
+      "step": 7048
+    },
+    {
+      "epoch": 0.3929208472686734,
+      "grad_norm": 0.5010694861412048,
+      "learning_rate": 8e-05,
+      "loss": 1.5043,
+      "step": 7049
+    },
+    {
+      "epoch": 0.39297658862876256,
+      "grad_norm": 0.5062034130096436,
+      "learning_rate": 8e-05,
+      "loss": 1.4959,
+      "step": 7050
+    },
+    {
+      "epoch": 0.39303232998885174,
+      "grad_norm": 0.5208863019943237,
+      "learning_rate": 8e-05,
+      "loss": 1.4794,
+      "step": 7051
+    },
+    {
+      "epoch": 0.3930880713489409,
+      "grad_norm": 0.7044099569320679,
+      "learning_rate": 8e-05,
+      "loss": 1.6254,
+      "step": 7052
+    },
+    {
+      "epoch": 0.3931438127090301,
+      "grad_norm": 0.5580699443817139,
+      "learning_rate": 8e-05,
+      "loss": 1.6458,
+      "step": 7053
+    },
+    {
+      "epoch": 0.3931995540691193,
+      "grad_norm": 0.5844243168830872,
+      "learning_rate": 8e-05,
+      "loss": 1.8115,
+      "step": 7054
+    },
+    {
+      "epoch": 0.39325529542920845,
+      "grad_norm": 0.5109538435935974,
+      "learning_rate": 8e-05,
+      "loss": 1.7011,
+      "step": 7055
+    },
+    {
+      "epoch": 0.3933110367892977,
+      "grad_norm": 0.5302479267120361,
+      "learning_rate": 8e-05,
+      "loss": 1.6663,
+      "step": 7056
+    },
+    {
+      "epoch": 0.39336677814938686,
+      "grad_norm": 0.5831972360610962,
+      "learning_rate": 8e-05,
+      "loss": 2.0732,
+      "step": 7057
+    },
+    {
+      "epoch": 0.39342251950947604,
+      "grad_norm": 0.5115814805030823,
+      "learning_rate": 8e-05,
+      "loss": 1.5539,
+      "step": 7058
+    },
+    {
+      "epoch": 0.3934782608695652,
+      "grad_norm": 0.5479879379272461,
+      "learning_rate": 8e-05,
+      "loss": 1.6709,
+      "step": 7059
+    },
+    {
+      "epoch": 0.3935340022296544,
+      "grad_norm": 0.5461530089378357,
+      "learning_rate": 8e-05,
+      "loss": 1.6039,
+      "step": 7060
+    },
+    {
+      "epoch": 0.3935897435897436,
+      "grad_norm": 0.5113889575004578,
+      "learning_rate": 8e-05,
+      "loss": 1.6394,
+      "step": 7061
+    },
+    {
+      "epoch": 0.39364548494983276,
+      "grad_norm": 0.5324178338050842,
+      "learning_rate": 8e-05,
+      "loss": 1.7084,
+      "step": 7062
+    },
+    {
+      "epoch": 0.39370122630992194,
+      "grad_norm": 0.5632378458976746,
+      "learning_rate": 8e-05,
+      "loss": 1.7095,
+      "step": 7063
+    },
+    {
+      "epoch": 0.39375696767001117,
+      "grad_norm": 0.5247626304626465,
+      "learning_rate": 8e-05,
+      "loss": 1.9485,
+      "step": 7064
+    },
+    {
+      "epoch": 0.39381270903010035,
+      "grad_norm": 0.5283797383308411,
+      "learning_rate": 8e-05,
+      "loss": 1.8232,
+      "step": 7065
+    },
+    {
+      "epoch": 0.39386845039018953,
+      "grad_norm": 0.5559601187705994,
+      "learning_rate": 8e-05,
+      "loss": 1.9815,
+      "step": 7066
+    },
+    {
+      "epoch": 0.3939241917502787,
+      "grad_norm": 0.5030553340911865,
+      "learning_rate": 8e-05,
+      "loss": 1.6524,
+      "step": 7067
+    },
+    {
+      "epoch": 0.3939799331103679,
+      "grad_norm": 0.4583421051502228,
+      "learning_rate": 8e-05,
+      "loss": 1.0681,
+      "step": 7068
+    },
+    {
+      "epoch": 0.39403567447045706,
+      "grad_norm": 0.4787176847457886,
+      "learning_rate": 8e-05,
+      "loss": 1.5423,
+      "step": 7069
+    },
+    {
+      "epoch": 0.39409141583054624,
+      "grad_norm": 0.5295685529708862,
+      "learning_rate": 8e-05,
+      "loss": 1.667,
+      "step": 7070
+    },
+    {
+      "epoch": 0.3941471571906355,
+      "grad_norm": 0.5449760556221008,
+      "learning_rate": 8e-05,
+      "loss": 1.7835,
+      "step": 7071
+    },
+    {
+      "epoch": 0.39420289855072466,
+      "grad_norm": 0.5539211630821228,
+      "learning_rate": 8e-05,
+      "loss": 1.7981,
+      "step": 7072
+    },
+    {
+      "epoch": 0.39425863991081384,
+      "grad_norm": 0.5042213201522827,
+      "learning_rate": 8e-05,
+      "loss": 1.304,
+      "step": 7073
+    },
+    {
+      "epoch": 0.394314381270903,
+      "grad_norm": 0.544682502746582,
+      "learning_rate": 8e-05,
+      "loss": 1.7827,
+      "step": 7074
+    },
+    {
+      "epoch": 0.3943701226309922,
+      "grad_norm": 0.5186235904693604,
+      "learning_rate": 8e-05,
+      "loss": 1.6544,
+      "step": 7075
+    },
+    {
+      "epoch": 0.39442586399108137,
+      "grad_norm": 0.5746743679046631,
+      "learning_rate": 8e-05,
+      "loss": 1.7657,
+      "step": 7076
+    },
+    {
+      "epoch": 0.39448160535117055,
+      "grad_norm": 0.5321332216262817,
+      "learning_rate": 8e-05,
+      "loss": 1.7641,
+      "step": 7077
+    },
+    {
+      "epoch": 0.39453734671125973,
+      "grad_norm": 0.5100351572036743,
+      "learning_rate": 8e-05,
+      "loss": 1.3928,
+      "step": 7078
+    },
+    {
+      "epoch": 0.39459308807134896,
+      "grad_norm": 0.555130124092102,
+      "learning_rate": 8e-05,
+      "loss": 1.7825,
+      "step": 7079
+    },
+    {
+      "epoch": 0.39464882943143814,
+      "grad_norm": 0.5125172138214111,
+      "learning_rate": 8e-05,
+      "loss": 1.6196,
+      "step": 7080
+    },
+    {
+      "epoch": 0.3947045707915273,
+      "grad_norm": 0.5487686991691589,
+      "learning_rate": 8e-05,
+      "loss": 1.8139,
+      "step": 7081
+    },
+    {
+      "epoch": 0.3947603121516165,
+      "grad_norm": 0.554560124874115,
+      "learning_rate": 8e-05,
+      "loss": 1.8361,
+      "step": 7082
+    },
+    {
+      "epoch": 0.3948160535117057,
+      "grad_norm": 0.4659266173839569,
+      "learning_rate": 8e-05,
+      "loss": 1.4233,
+      "step": 7083
+    },
+    {
+      "epoch": 0.39487179487179486,
+      "grad_norm": 0.5364368557929993,
+      "learning_rate": 8e-05,
+      "loss": 1.6832,
+      "step": 7084
+    },
+    {
+      "epoch": 0.39492753623188404,
+      "grad_norm": 0.5149415731430054,
+      "learning_rate": 8e-05,
+      "loss": 1.5629,
+      "step": 7085
+    },
+    {
+      "epoch": 0.39498327759197327,
+      "grad_norm": 0.5535032749176025,
+      "learning_rate": 8e-05,
+      "loss": 1.6994,
+      "step": 7086
+    },
+    {
+      "epoch": 0.39503901895206245,
+      "grad_norm": 0.5155476927757263,
+      "learning_rate": 8e-05,
+      "loss": 1.4715,
+      "step": 7087
+    },
+    {
+      "epoch": 0.3950947603121516,
+      "grad_norm": 0.5397098064422607,
+      "learning_rate": 8e-05,
+      "loss": 1.6871,
+      "step": 7088
+    },
+    {
+      "epoch": 0.3951505016722408,
+      "grad_norm": 0.5228619575500488,
+      "learning_rate": 8e-05,
+      "loss": 1.6322,
+      "step": 7089
+    },
+    {
+      "epoch": 0.39520624303233,
+      "grad_norm": 0.5426560044288635,
+      "learning_rate": 8e-05,
+      "loss": 1.8848,
+      "step": 7090
+    },
+    {
+      "epoch": 0.39526198439241916,
+      "grad_norm": 0.6139596700668335,
+      "learning_rate": 8e-05,
+      "loss": 1.7845,
+      "step": 7091
+    },
+    {
+      "epoch": 0.39531772575250834,
+      "grad_norm": 0.5236718058586121,
+      "learning_rate": 8e-05,
+      "loss": 1.6079,
+      "step": 7092
+    },
+    {
+      "epoch": 0.3953734671125975,
+      "grad_norm": 0.5317121148109436,
+      "learning_rate": 8e-05,
+      "loss": 1.8536,
+      "step": 7093
+    },
+    {
+      "epoch": 0.39542920847268676,
+      "grad_norm": 0.5438995361328125,
+      "learning_rate": 8e-05,
+      "loss": 1.5735,
+      "step": 7094
+    },
+    {
+      "epoch": 0.39548494983277593,
+      "grad_norm": 0.5763745307922363,
+      "learning_rate": 8e-05,
+      "loss": 1.7013,
+      "step": 7095
+    },
+    {
+      "epoch": 0.3955406911928651,
+      "grad_norm": 0.52718585729599,
+      "learning_rate": 8e-05,
+      "loss": 1.5212,
+      "step": 7096
+    },
+    {
+      "epoch": 0.3955964325529543,
+      "grad_norm": 0.5630643367767334,
+      "learning_rate": 8e-05,
+      "loss": 1.8717,
+      "step": 7097
+    },
+    {
+      "epoch": 0.39565217391304347,
+      "grad_norm": 0.5001387596130371,
+      "learning_rate": 8e-05,
+      "loss": 1.6107,
+      "step": 7098
+    },
+    {
+      "epoch": 0.39570791527313265,
+      "grad_norm": 0.5191920399665833,
+      "learning_rate": 8e-05,
+      "loss": 1.6995,
+      "step": 7099
+    },
+    {
+      "epoch": 0.3957636566332218,
+      "grad_norm": 0.5507417917251587,
+      "learning_rate": 8e-05,
+      "loss": 1.983,
+      "step": 7100
+    },
+    {
+      "epoch": 0.39581939799331106,
+      "grad_norm": 0.540391206741333,
+      "learning_rate": 8e-05,
+      "loss": 1.8635,
+      "step": 7101
+    },
+    {
+      "epoch": 0.39587513935340024,
+      "grad_norm": 0.534968912601471,
+      "learning_rate": 8e-05,
+      "loss": 1.6873,
+      "step": 7102
+    },
+    {
+      "epoch": 0.3959308807134894,
+      "grad_norm": 0.5403704047203064,
+      "learning_rate": 8e-05,
+      "loss": 1.8149,
+      "step": 7103
+    },
+    {
+      "epoch": 0.3959866220735786,
+      "grad_norm": 0.5288872122764587,
+      "learning_rate": 8e-05,
+      "loss": 1.6222,
+      "step": 7104
+    },
+    {
+      "epoch": 0.3960423634336678,
+      "grad_norm": 0.4927409589290619,
+      "learning_rate": 8e-05,
+      "loss": 1.7369,
+      "step": 7105
+    },
+    {
+      "epoch": 0.39609810479375696,
+      "grad_norm": 0.49208763241767883,
+      "learning_rate": 8e-05,
+      "loss": 1.5327,
+      "step": 7106
+    },
+    {
+      "epoch": 0.39615384615384613,
+      "grad_norm": 0.4779423177242279,
+      "learning_rate": 8e-05,
+      "loss": 1.428,
+      "step": 7107
+    },
+    {
+      "epoch": 0.3962095875139353,
+      "grad_norm": 0.52891606092453,
+      "learning_rate": 8e-05,
+      "loss": 1.792,
+      "step": 7108
+    },
+    {
+      "epoch": 0.39626532887402455,
+      "grad_norm": 0.5548779368400574,
+      "learning_rate": 8e-05,
+      "loss": 1.7652,
+      "step": 7109
+    },
+    {
+      "epoch": 0.3963210702341137,
+      "grad_norm": 0.5618845820426941,
+      "learning_rate": 8e-05,
+      "loss": 1.8268,
+      "step": 7110
+    },
+    {
+      "epoch": 0.3963768115942029,
+      "grad_norm": 0.5512269139289856,
+      "learning_rate": 8e-05,
+      "loss": 1.7835,
+      "step": 7111
+    },
+    {
+      "epoch": 0.3964325529542921,
+      "grad_norm": 0.5383572578430176,
+      "learning_rate": 8e-05,
+      "loss": 1.6328,
+      "step": 7112
+    },
+    {
+      "epoch": 0.39648829431438126,
+      "grad_norm": 0.5181984305381775,
+      "learning_rate": 8e-05,
+      "loss": 1.4108,
+      "step": 7113
+    },
+    {
+      "epoch": 0.39654403567447044,
+      "grad_norm": 0.6135736107826233,
+      "learning_rate": 8e-05,
+      "loss": 2.1073,
+      "step": 7114
+    },
+    {
+      "epoch": 0.3965997770345596,
+      "grad_norm": 0.5372074246406555,
+      "learning_rate": 8e-05,
+      "loss": 1.7575,
+      "step": 7115
+    },
+    {
+      "epoch": 0.39665551839464885,
+      "grad_norm": 0.5308025479316711,
+      "learning_rate": 8e-05,
+      "loss": 1.6895,
+      "step": 7116
+    },
+    {
+      "epoch": 0.39671125975473803,
+      "grad_norm": 0.5664640069007874,
+      "learning_rate": 8e-05,
+      "loss": 1.7114,
+      "step": 7117
+    },
+    {
+      "epoch": 0.3967670011148272,
+      "grad_norm": 0.5988225936889648,
+      "learning_rate": 8e-05,
+      "loss": 1.9322,
+      "step": 7118
+    },
+    {
+      "epoch": 0.3968227424749164,
+      "grad_norm": 0.5092202425003052,
+      "learning_rate": 8e-05,
+      "loss": 1.7817,
+      "step": 7119
+    },
+    {
+      "epoch": 0.39687848383500557,
+      "grad_norm": 0.5076021552085876,
+      "learning_rate": 8e-05,
+      "loss": 1.6029,
+      "step": 7120
+    },
+    {
+      "epoch": 0.39693422519509475,
+      "grad_norm": 0.5385578870773315,
+      "learning_rate": 8e-05,
+      "loss": 1.5666,
+      "step": 7121
+    },
+    {
+      "epoch": 0.3969899665551839,
+      "grad_norm": 0.52571040391922,
+      "learning_rate": 8e-05,
+      "loss": 1.6277,
+      "step": 7122
+    },
+    {
+      "epoch": 0.3970457079152731,
+      "grad_norm": 0.5378811359405518,
+      "learning_rate": 8e-05,
+      "loss": 1.7184,
+      "step": 7123
+    },
+    {
+      "epoch": 0.39710144927536234,
+      "grad_norm": 0.4979332983493805,
+      "learning_rate": 8e-05,
+      "loss": 1.6225,
+      "step": 7124
+    },
+    {
+      "epoch": 0.3971571906354515,
+      "grad_norm": 0.5347927212715149,
+      "learning_rate": 8e-05,
+      "loss": 1.6643,
+      "step": 7125
+    },
+    {
+      "epoch": 0.3972129319955407,
+      "grad_norm": 0.557803750038147,
+      "learning_rate": 8e-05,
+      "loss": 1.7504,
+      "step": 7126
+    },
+    {
+      "epoch": 0.3972686733556299,
+      "grad_norm": 0.5096932053565979,
+      "learning_rate": 8e-05,
+      "loss": 1.7177,
+      "step": 7127
+    },
+    {
+      "epoch": 0.39732441471571905,
+      "grad_norm": 0.5407583117485046,
+      "learning_rate": 8e-05,
+      "loss": 1.7523,
+      "step": 7128
+    },
+    {
+      "epoch": 0.39738015607580823,
+      "grad_norm": 0.46805867552757263,
+      "learning_rate": 8e-05,
+      "loss": 1.549,
+      "step": 7129
+    },
+    {
+      "epoch": 0.3974358974358974,
+      "grad_norm": 0.5960009098052979,
+      "learning_rate": 8e-05,
+      "loss": 1.941,
+      "step": 7130
+    },
+    {
+      "epoch": 0.39749163879598665,
+      "grad_norm": 0.5466600060462952,
+      "learning_rate": 8e-05,
+      "loss": 1.5596,
+      "step": 7131
+    },
+    {
+      "epoch": 0.3975473801560758,
+      "grad_norm": 0.5576518177986145,
+      "learning_rate": 8e-05,
+      "loss": 1.7435,
+      "step": 7132
+    },
+    {
+      "epoch": 0.397603121516165,
+      "grad_norm": 0.56622713804245,
+      "learning_rate": 8e-05,
+      "loss": 1.7719,
+      "step": 7133
+    },
+    {
+      "epoch": 0.3976588628762542,
+      "grad_norm": 0.515771746635437,
+      "learning_rate": 8e-05,
+      "loss": 1.5704,
+      "step": 7134
+    },
+    {
+      "epoch": 0.39771460423634336,
+      "grad_norm": 0.5629921555519104,
+      "learning_rate": 8e-05,
+      "loss": 1.7414,
+      "step": 7135
+    },
+    {
+      "epoch": 0.39777034559643254,
+      "grad_norm": 0.5997464656829834,
+      "learning_rate": 8e-05,
+      "loss": 1.8351,
+      "step": 7136
+    },
+    {
+      "epoch": 0.3978260869565217,
+      "grad_norm": 0.5276950001716614,
+      "learning_rate": 8e-05,
+      "loss": 1.5901,
+      "step": 7137
+    },
+    {
+      "epoch": 0.39788182831661095,
+      "grad_norm": 0.562311589717865,
+      "learning_rate": 8e-05,
+      "loss": 1.7781,
+      "step": 7138
+    },
+    {
+      "epoch": 0.39793756967670013,
+      "grad_norm": 0.5696790218353271,
+      "learning_rate": 8e-05,
+      "loss": 1.8847,
+      "step": 7139
+    },
+    {
+      "epoch": 0.3979933110367893,
+      "grad_norm": 0.5362411141395569,
+      "learning_rate": 8e-05,
+      "loss": 1.7381,
+      "step": 7140
+    },
+    {
+      "epoch": 0.3980490523968785,
+      "grad_norm": 0.5526409149169922,
+      "learning_rate": 8e-05,
+      "loss": 1.7339,
+      "step": 7141
+    },
+    {
+      "epoch": 0.39810479375696767,
+      "grad_norm": 0.5474283695220947,
+      "learning_rate": 8e-05,
+      "loss": 1.575,
+      "step": 7142
+    },
+    {
+      "epoch": 0.39816053511705685,
+      "grad_norm": 0.5464802980422974,
+      "learning_rate": 8e-05,
+      "loss": 1.9082,
+      "step": 7143
+    },
+    {
+      "epoch": 0.398216276477146,
+      "grad_norm": 0.5198264718055725,
+      "learning_rate": 8e-05,
+      "loss": 1.7022,
+      "step": 7144
+    },
+    {
+      "epoch": 0.3982720178372352,
+      "grad_norm": 0.5149979591369629,
+      "learning_rate": 8e-05,
+      "loss": 1.7212,
+      "step": 7145
+    },
+    {
+      "epoch": 0.39832775919732444,
+      "grad_norm": 0.5334619283676147,
+      "learning_rate": 8e-05,
+      "loss": 1.6983,
+      "step": 7146
+    },
+    {
+      "epoch": 0.3983835005574136,
+      "grad_norm": 0.5151000022888184,
+      "learning_rate": 8e-05,
+      "loss": 1.6459,
+      "step": 7147
+    },
+    {
+      "epoch": 0.3984392419175028,
+      "grad_norm": 0.557898998260498,
+      "learning_rate": 8e-05,
+      "loss": 1.8314,
+      "step": 7148
+    },
+    {
+      "epoch": 0.398494983277592,
+      "grad_norm": 0.5185675621032715,
+      "learning_rate": 8e-05,
+      "loss": 1.7921,
+      "step": 7149
+    },
+    {
+      "epoch": 0.39855072463768115,
+      "grad_norm": 0.5042035579681396,
+      "learning_rate": 8e-05,
+      "loss": 1.7054,
+      "step": 7150
+    },
+    {
+      "epoch": 0.39860646599777033,
+      "grad_norm": 0.5184841156005859,
+      "learning_rate": 8e-05,
+      "loss": 1.7933,
+      "step": 7151
+    },
+    {
+      "epoch": 0.3986622073578595,
+      "grad_norm": 0.505743682384491,
+      "learning_rate": 8e-05,
+      "loss": 1.7687,
+      "step": 7152
+    },
+    {
+      "epoch": 0.39871794871794874,
+      "grad_norm": 0.47850340604782104,
+      "learning_rate": 8e-05,
+      "loss": 1.6095,
+      "step": 7153
+    },
+    {
+      "epoch": 0.3987736900780379,
+      "grad_norm": 0.5684821009635925,
+      "learning_rate": 8e-05,
+      "loss": 1.7756,
+      "step": 7154
+    },
+    {
+      "epoch": 0.3988294314381271,
+      "grad_norm": 0.5087734460830688,
+      "learning_rate": 8e-05,
+      "loss": 1.6656,
+      "step": 7155
+    },
+    {
+      "epoch": 0.3988851727982163,
+      "grad_norm": 0.5265142321586609,
+      "learning_rate": 8e-05,
+      "loss": 1.7659,
+      "step": 7156
+    },
+    {
+      "epoch": 0.39894091415830546,
+      "grad_norm": 0.4864795506000519,
+      "learning_rate": 8e-05,
+      "loss": 1.5684,
+      "step": 7157
+    },
+    {
+      "epoch": 0.39899665551839464,
+      "grad_norm": 0.4981810748577118,
+      "learning_rate": 8e-05,
+      "loss": 1.6102,
+      "step": 7158
+    },
+    {
+      "epoch": 0.3990523968784838,
+      "grad_norm": 0.5119472742080688,
+      "learning_rate": 8e-05,
+      "loss": 1.5057,
+      "step": 7159
+    },
+    {
+      "epoch": 0.399108138238573,
+      "grad_norm": 0.507240355014801,
+      "learning_rate": 8e-05,
+      "loss": 1.7858,
+      "step": 7160
+    },
+    {
+      "epoch": 0.39916387959866223,
+      "grad_norm": 0.5334985256195068,
+      "learning_rate": 8e-05,
+      "loss": 1.6685,
+      "step": 7161
+    },
+    {
+      "epoch": 0.3992196209587514,
+      "grad_norm": 0.5807051658630371,
+      "learning_rate": 8e-05,
+      "loss": 1.9291,
+      "step": 7162
+    },
+    {
+      "epoch": 0.3992753623188406,
+      "grad_norm": 0.5425968766212463,
+      "learning_rate": 8e-05,
+      "loss": 1.7255,
+      "step": 7163
+    },
+    {
+      "epoch": 0.39933110367892977,
+      "grad_norm": 0.5361913442611694,
+      "learning_rate": 8e-05,
+      "loss": 1.8213,
+      "step": 7164
+    },
+    {
+      "epoch": 0.39938684503901895,
+      "grad_norm": 0.5138218998908997,
+      "learning_rate": 8e-05,
+      "loss": 1.7108,
+      "step": 7165
+    },
+    {
+      "epoch": 0.3994425863991081,
+      "grad_norm": 0.4844566881656647,
+      "learning_rate": 8e-05,
+      "loss": 1.4822,
+      "step": 7166
+    },
+    {
+      "epoch": 0.3994983277591973,
+      "grad_norm": 0.5321471691131592,
+      "learning_rate": 8e-05,
+      "loss": 1.7647,
+      "step": 7167
+    },
+    {
+      "epoch": 0.39955406911928654,
+      "grad_norm": 0.5588659048080444,
+      "learning_rate": 8e-05,
+      "loss": 1.8536,
+      "step": 7168
+    },
+    {
+      "epoch": 0.3996098104793757,
+      "grad_norm": 0.5184299945831299,
+      "learning_rate": 8e-05,
+      "loss": 1.4886,
+      "step": 7169
+    },
+    {
+      "epoch": 0.3996655518394649,
+      "grad_norm": 0.5337185263633728,
+      "learning_rate": 8e-05,
+      "loss": 1.6478,
+      "step": 7170
+    },
+    {
+      "epoch": 0.3997212931995541,
+      "grad_norm": 0.5601459741592407,
+      "learning_rate": 8e-05,
+      "loss": 1.7805,
+      "step": 7171
+    },
+    {
+      "epoch": 0.39977703455964325,
+      "grad_norm": 0.5493620038032532,
+      "learning_rate": 8e-05,
+      "loss": 1.9224,
+      "step": 7172
+    },
+    {
+      "epoch": 0.39983277591973243,
+      "grad_norm": 0.49881303310394287,
+      "learning_rate": 8e-05,
+      "loss": 1.5991,
+      "step": 7173
+    },
+    {
+      "epoch": 0.3998885172798216,
+      "grad_norm": 0.6255406141281128,
+      "learning_rate": 8e-05,
+      "loss": 1.8032,
+      "step": 7174
+    },
+    {
+      "epoch": 0.3999442586399108,
+      "grad_norm": 0.46618714928627014,
+      "learning_rate": 8e-05,
+      "loss": 1.4614,
+      "step": 7175
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5251069664955139,
+      "learning_rate": 8e-05,
+      "loss": 1.695,
+      "step": 7176
+    },
+    {
+      "epoch": 0.4000557413600892,
+      "grad_norm": 0.5375117063522339,
+      "learning_rate": 8e-05,
+      "loss": 1.7528,
+      "step": 7177
+    },
+    {
+      "epoch": 0.4001114827201784,
+      "grad_norm": 0.5366592407226562,
+      "learning_rate": 8e-05,
+      "loss": 1.8064,
+      "step": 7178
+    },
+    {
+      "epoch": 0.40016722408026756,
+      "grad_norm": 0.5342915654182434,
+      "learning_rate": 8e-05,
+      "loss": 1.7025,
+      "step": 7179
+    },
+    {
+      "epoch": 0.40022296544035674,
+      "grad_norm": 0.510113000869751,
+      "learning_rate": 8e-05,
+      "loss": 1.6142,
+      "step": 7180
+    },
+    {
+      "epoch": 0.4002787068004459,
+      "grad_norm": 0.5079729557037354,
+      "learning_rate": 8e-05,
+      "loss": 1.6545,
+      "step": 7181
+    },
+    {
+      "epoch": 0.4003344481605351,
+      "grad_norm": 0.5121543407440186,
+      "learning_rate": 8e-05,
+      "loss": 1.5406,
+      "step": 7182
+    },
+    {
+      "epoch": 0.40039018952062433,
+      "grad_norm": 0.5430922508239746,
+      "learning_rate": 8e-05,
+      "loss": 1.6463,
+      "step": 7183
+    },
+    {
+      "epoch": 0.4004459308807135,
+      "grad_norm": 0.5592001080513,
+      "learning_rate": 8e-05,
+      "loss": 1.7763,
+      "step": 7184
+    },
+    {
+      "epoch": 0.4005016722408027,
+      "grad_norm": 0.5771913528442383,
+      "learning_rate": 8e-05,
+      "loss": 2.056,
+      "step": 7185
+    },
+    {
+      "epoch": 0.40055741360089187,
+      "grad_norm": 0.5104514360427856,
+      "learning_rate": 8e-05,
+      "loss": 1.5396,
+      "step": 7186
+    },
+    {
+      "epoch": 0.40061315496098104,
+      "grad_norm": 0.543381929397583,
+      "learning_rate": 8e-05,
+      "loss": 1.6641,
+      "step": 7187
+    },
+    {
+      "epoch": 0.4006688963210702,
+      "grad_norm": 0.5523921251296997,
+      "learning_rate": 8e-05,
+      "loss": 1.5779,
+      "step": 7188
+    },
+    {
+      "epoch": 0.4007246376811594,
+      "grad_norm": 0.4858691394329071,
+      "learning_rate": 8e-05,
+      "loss": 1.4625,
+      "step": 7189
+    },
+    {
+      "epoch": 0.4007803790412486,
+      "grad_norm": 0.5553864240646362,
+      "learning_rate": 8e-05,
+      "loss": 1.6574,
+      "step": 7190
+    },
+    {
+      "epoch": 0.4008361204013378,
+      "grad_norm": 0.6046305298805237,
+      "learning_rate": 8e-05,
+      "loss": 1.8414,
+      "step": 7191
+    },
+    {
+      "epoch": 0.400891861761427,
+      "grad_norm": 0.48796287178993225,
+      "learning_rate": 8e-05,
+      "loss": 1.4957,
+      "step": 7192
+    },
+    {
+      "epoch": 0.40094760312151617,
+      "grad_norm": 0.5404077768325806,
+      "learning_rate": 8e-05,
+      "loss": 1.4942,
+      "step": 7193
+    },
+    {
+      "epoch": 0.40100334448160535,
+      "grad_norm": 0.5653712153434753,
+      "learning_rate": 8e-05,
+      "loss": 1.8882,
+      "step": 7194
+    },
+    {
+      "epoch": 0.40105908584169453,
+      "grad_norm": 0.5747902393341064,
+      "learning_rate": 8e-05,
+      "loss": 1.8261,
+      "step": 7195
+    },
+    {
+      "epoch": 0.4011148272017837,
+      "grad_norm": 0.537380039691925,
+      "learning_rate": 8e-05,
+      "loss": 1.7867,
+      "step": 7196
+    },
+    {
+      "epoch": 0.4011705685618729,
+      "grad_norm": 0.5171547532081604,
+      "learning_rate": 8e-05,
+      "loss": 1.6586,
+      "step": 7197
+    },
+    {
+      "epoch": 0.4012263099219621,
+      "grad_norm": 0.5561977624893188,
+      "learning_rate": 8e-05,
+      "loss": 1.6361,
+      "step": 7198
+    },
+    {
+      "epoch": 0.4012820512820513,
+      "grad_norm": 0.5058720707893372,
+      "learning_rate": 8e-05,
+      "loss": 1.5449,
+      "step": 7199
+    },
+    {
+      "epoch": 0.4013377926421405,
+      "grad_norm": 0.562718391418457,
+      "learning_rate": 8e-05,
+      "loss": 1.7685,
+      "step": 7200
+    },
+    {
+      "epoch": 0.40139353400222966,
+      "grad_norm": 0.5348118543624878,
+      "learning_rate": 8e-05,
+      "loss": 1.7649,
+      "step": 7201
+    },
+    {
+      "epoch": 0.40144927536231884,
+      "grad_norm": 0.5601673126220703,
+      "learning_rate": 8e-05,
+      "loss": 1.796,
+      "step": 7202
+    },
+    {
+      "epoch": 0.401505016722408,
+      "grad_norm": 0.492948979139328,
+      "learning_rate": 8e-05,
+      "loss": 1.5824,
+      "step": 7203
+    },
+    {
+      "epoch": 0.4015607580824972,
+      "grad_norm": 0.5347782969474792,
+      "learning_rate": 8e-05,
+      "loss": 1.559,
+      "step": 7204
+    },
+    {
+      "epoch": 0.40161649944258637,
+      "grad_norm": 0.5501527786254883,
+      "learning_rate": 8e-05,
+      "loss": 1.5969,
+      "step": 7205
+    },
+    {
+      "epoch": 0.4016722408026756,
+      "grad_norm": 0.5141997337341309,
+      "learning_rate": 8e-05,
+      "loss": 1.5827,
+      "step": 7206
+    },
+    {
+      "epoch": 0.4017279821627648,
+      "grad_norm": 0.49190807342529297,
+      "learning_rate": 8e-05,
+      "loss": 1.5513,
+      "step": 7207
+    },
+    {
+      "epoch": 0.40178372352285396,
+      "grad_norm": 0.5163851976394653,
+      "learning_rate": 8e-05,
+      "loss": 1.4754,
+      "step": 7208
+    },
+    {
+      "epoch": 0.40183946488294314,
+      "grad_norm": 0.5286582708358765,
+      "learning_rate": 8e-05,
+      "loss": 1.6241,
+      "step": 7209
+    },
+    {
+      "epoch": 0.4018952062430323,
+      "grad_norm": 0.5705527663230896,
+      "learning_rate": 8e-05,
+      "loss": 1.8793,
+      "step": 7210
+    },
+    {
+      "epoch": 0.4019509476031215,
+      "grad_norm": 0.505829393863678,
+      "learning_rate": 8e-05,
+      "loss": 1.6368,
+      "step": 7211
+    },
+    {
+      "epoch": 0.4020066889632107,
+      "grad_norm": 0.6229416728019714,
+      "learning_rate": 8e-05,
+      "loss": 2.0446,
+      "step": 7212
+    },
+    {
+      "epoch": 0.4020624303232999,
+      "grad_norm": 0.5273987054824829,
+      "learning_rate": 8e-05,
+      "loss": 1.6026,
+      "step": 7213
+    },
+    {
+      "epoch": 0.4021181716833891,
+      "grad_norm": 0.5581182837486267,
+      "learning_rate": 8e-05,
+      "loss": 1.8189,
+      "step": 7214
+    },
+    {
+      "epoch": 0.40217391304347827,
+      "grad_norm": 0.5493568778038025,
+      "learning_rate": 8e-05,
+      "loss": 1.7455,
+      "step": 7215
+    },
+    {
+      "epoch": 0.40222965440356745,
+      "grad_norm": 0.5482844710350037,
+      "learning_rate": 8e-05,
+      "loss": 1.8096,
+      "step": 7216
+    },
+    {
+      "epoch": 0.40228539576365663,
+      "grad_norm": 0.5216826796531677,
+      "learning_rate": 8e-05,
+      "loss": 1.59,
+      "step": 7217
+    },
+    {
+      "epoch": 0.4023411371237458,
+      "grad_norm": 0.5004620552062988,
+      "learning_rate": 8e-05,
+      "loss": 1.5273,
+      "step": 7218
+    },
+    {
+      "epoch": 0.402396878483835,
+      "grad_norm": 0.5231639742851257,
+      "learning_rate": 8e-05,
+      "loss": 1.6749,
+      "step": 7219
+    },
+    {
+      "epoch": 0.40245261984392416,
+      "grad_norm": 0.4889169931411743,
+      "learning_rate": 8e-05,
+      "loss": 1.6555,
+      "step": 7220
+    },
+    {
+      "epoch": 0.4025083612040134,
+      "grad_norm": 0.49795204401016235,
+      "learning_rate": 8e-05,
+      "loss": 1.583,
+      "step": 7221
+    },
+    {
+      "epoch": 0.4025641025641026,
+      "grad_norm": 0.5205188989639282,
+      "learning_rate": 8e-05,
+      "loss": 1.6421,
+      "step": 7222
+    },
+    {
+      "epoch": 0.40261984392419176,
+      "grad_norm": 0.5181834101676941,
+      "learning_rate": 8e-05,
+      "loss": 1.6711,
+      "step": 7223
+    },
+    {
+      "epoch": 0.40267558528428093,
+      "grad_norm": 0.5610384941101074,
+      "learning_rate": 8e-05,
+      "loss": 1.5956,
+      "step": 7224
+    },
+    {
+      "epoch": 0.4027313266443701,
+      "grad_norm": 0.5183853507041931,
+      "learning_rate": 8e-05,
+      "loss": 1.5579,
+      "step": 7225
+    },
+    {
+      "epoch": 0.4027870680044593,
+      "grad_norm": 0.5921036005020142,
+      "learning_rate": 8e-05,
+      "loss": 1.8994,
+      "step": 7226
+    },
+    {
+      "epoch": 0.40284280936454847,
+      "grad_norm": 0.54100501537323,
+      "learning_rate": 8e-05,
+      "loss": 1.6054,
+      "step": 7227
+    },
+    {
+      "epoch": 0.4028985507246377,
+      "grad_norm": 0.5243507027626038,
+      "learning_rate": 8e-05,
+      "loss": 1.5235,
+      "step": 7228
+    },
+    {
+      "epoch": 0.4029542920847269,
+      "grad_norm": 0.5220246315002441,
+      "learning_rate": 8e-05,
+      "loss": 1.5431,
+      "step": 7229
+    },
+    {
+      "epoch": 0.40301003344481606,
+      "grad_norm": 0.48561525344848633,
+      "learning_rate": 8e-05,
+      "loss": 1.4805,
+      "step": 7230
+    },
+    {
+      "epoch": 0.40306577480490524,
+      "grad_norm": 0.548069953918457,
+      "learning_rate": 8e-05,
+      "loss": 1.567,
+      "step": 7231
+    },
+    {
+      "epoch": 0.4031215161649944,
+      "grad_norm": 0.5422705411911011,
+      "learning_rate": 8e-05,
+      "loss": 1.7502,
+      "step": 7232
+    },
+    {
+      "epoch": 0.4031772575250836,
+      "grad_norm": 0.508958101272583,
+      "learning_rate": 8e-05,
+      "loss": 1.4607,
+      "step": 7233
+    },
+    {
+      "epoch": 0.4032329988851728,
+      "grad_norm": 0.5446996092796326,
+      "learning_rate": 8e-05,
+      "loss": 1.7624,
+      "step": 7234
+    },
+    {
+      "epoch": 0.403288740245262,
+      "grad_norm": 0.5247542858123779,
+      "learning_rate": 8e-05,
+      "loss": 1.5096,
+      "step": 7235
+    },
+    {
+      "epoch": 0.4033444816053512,
+      "grad_norm": 0.5961942076683044,
+      "learning_rate": 8e-05,
+      "loss": 1.9071,
+      "step": 7236
+    },
+    {
+      "epoch": 0.40340022296544037,
+      "grad_norm": 0.5433081984519958,
+      "learning_rate": 8e-05,
+      "loss": 1.6365,
+      "step": 7237
+    },
+    {
+      "epoch": 0.40345596432552955,
+      "grad_norm": 0.5088001489639282,
+      "learning_rate": 8e-05,
+      "loss": 1.7051,
+      "step": 7238
+    },
+    {
+      "epoch": 0.4035117056856187,
+      "grad_norm": 0.510620653629303,
+      "learning_rate": 8e-05,
+      "loss": 1.592,
+      "step": 7239
+    },
+    {
+      "epoch": 0.4035674470457079,
+      "grad_norm": 0.5772462487220764,
+      "learning_rate": 8e-05,
+      "loss": 1.7582,
+      "step": 7240
+    },
+    {
+      "epoch": 0.4036231884057971,
+      "grad_norm": 0.5558195114135742,
+      "learning_rate": 8e-05,
+      "loss": 1.7582,
+      "step": 7241
+    },
+    {
+      "epoch": 0.40367892976588626,
+      "grad_norm": 0.5695681571960449,
+      "learning_rate": 8e-05,
+      "loss": 1.819,
+      "step": 7242
+    },
+    {
+      "epoch": 0.4037346711259755,
+      "grad_norm": 0.4986192584037781,
+      "learning_rate": 8e-05,
+      "loss": 1.6319,
+      "step": 7243
+    },
+    {
+      "epoch": 0.4037904124860647,
+      "grad_norm": 0.5238422155380249,
+      "learning_rate": 8e-05,
+      "loss": 1.7019,
+      "step": 7244
+    },
+    {
+      "epoch": 0.40384615384615385,
+      "grad_norm": 0.517815351486206,
+      "learning_rate": 8e-05,
+      "loss": 1.6851,
+      "step": 7245
+    },
+    {
+      "epoch": 0.40390189520624303,
+      "grad_norm": 0.5185407400131226,
+      "learning_rate": 8e-05,
+      "loss": 1.6868,
+      "step": 7246
+    },
+    {
+      "epoch": 0.4039576365663322,
+      "grad_norm": 0.5118069648742676,
+      "learning_rate": 8e-05,
+      "loss": 1.8089,
+      "step": 7247
+    },
+    {
+      "epoch": 0.4040133779264214,
+      "grad_norm": 0.5143292546272278,
+      "learning_rate": 8e-05,
+      "loss": 1.4618,
+      "step": 7248
+    },
+    {
+      "epoch": 0.40406911928651057,
+      "grad_norm": 0.5812262892723083,
+      "learning_rate": 8e-05,
+      "loss": 1.8726,
+      "step": 7249
+    },
+    {
+      "epoch": 0.4041248606465998,
+      "grad_norm": 0.5687528848648071,
+      "learning_rate": 8e-05,
+      "loss": 1.9506,
+      "step": 7250
+    },
+    {
+      "epoch": 0.404180602006689,
+      "grad_norm": 0.5691890716552734,
+      "learning_rate": 8e-05,
+      "loss": 1.9047,
+      "step": 7251
+    },
+    {
+      "epoch": 0.40423634336677816,
+      "grad_norm": 0.5662111043930054,
+      "learning_rate": 8e-05,
+      "loss": 1.6631,
+      "step": 7252
+    },
+    {
+      "epoch": 0.40429208472686734,
+      "grad_norm": 0.5995144844055176,
+      "learning_rate": 8e-05,
+      "loss": 1.8842,
+      "step": 7253
+    },
+    {
+      "epoch": 0.4043478260869565,
+      "grad_norm": 0.5363901853561401,
+      "learning_rate": 8e-05,
+      "loss": 1.7278,
+      "step": 7254
+    },
+    {
+      "epoch": 0.4044035674470457,
+      "grad_norm": 0.5433478355407715,
+      "learning_rate": 8e-05,
+      "loss": 1.7402,
+      "step": 7255
+    },
+    {
+      "epoch": 0.4044593088071349,
+      "grad_norm": 0.5265305042266846,
+      "learning_rate": 8e-05,
+      "loss": 1.7695,
+      "step": 7256
+    },
+    {
+      "epoch": 0.40451505016722406,
+      "grad_norm": 0.4668007493019104,
+      "learning_rate": 8e-05,
+      "loss": 1.4825,
+      "step": 7257
+    },
+    {
+      "epoch": 0.4045707915273133,
+      "grad_norm": 0.5050549507141113,
+      "learning_rate": 8e-05,
+      "loss": 1.6319,
+      "step": 7258
+    },
+    {
+      "epoch": 0.40462653288740247,
+      "grad_norm": 0.5017398595809937,
+      "learning_rate": 8e-05,
+      "loss": 1.6405,
+      "step": 7259
+    },
+    {
+      "epoch": 0.40468227424749165,
+      "grad_norm": 0.5323314666748047,
+      "learning_rate": 8e-05,
+      "loss": 1.7371,
+      "step": 7260
+    },
+    {
+      "epoch": 0.4047380156075808,
+      "grad_norm": 0.5021803975105286,
+      "learning_rate": 8e-05,
+      "loss": 1.6651,
+      "step": 7261
+    },
+    {
+      "epoch": 0.40479375696767,
+      "grad_norm": 0.4910186529159546,
+      "learning_rate": 8e-05,
+      "loss": 1.5458,
+      "step": 7262
+    },
+    {
+      "epoch": 0.4048494983277592,
+      "grad_norm": 0.6216732263565063,
+      "learning_rate": 8e-05,
+      "loss": 2.0608,
+      "step": 7263
+    },
+    {
+      "epoch": 0.40490523968784836,
+      "grad_norm": 0.553127110004425,
+      "learning_rate": 8e-05,
+      "loss": 1.6896,
+      "step": 7264
+    },
+    {
+      "epoch": 0.4049609810479376,
+      "grad_norm": 0.5868520736694336,
+      "learning_rate": 8e-05,
+      "loss": 1.7069,
+      "step": 7265
+    },
+    {
+      "epoch": 0.4050167224080268,
+      "grad_norm": 0.46473756432533264,
+      "learning_rate": 8e-05,
+      "loss": 1.5142,
+      "step": 7266
+    },
+    {
+      "epoch": 0.40507246376811595,
+      "grad_norm": 0.5369247198104858,
+      "learning_rate": 8e-05,
+      "loss": 1.6509,
+      "step": 7267
+    },
+    {
+      "epoch": 0.40512820512820513,
+      "grad_norm": 0.515741765499115,
+      "learning_rate": 8e-05,
+      "loss": 1.7297,
+      "step": 7268
+    },
+    {
+      "epoch": 0.4051839464882943,
+      "grad_norm": 0.5385709404945374,
+      "learning_rate": 8e-05,
+      "loss": 1.7021,
+      "step": 7269
+    },
+    {
+      "epoch": 0.4052396878483835,
+      "grad_norm": 0.5441043972969055,
+      "learning_rate": 8e-05,
+      "loss": 1.6534,
+      "step": 7270
+    },
+    {
+      "epoch": 0.40529542920847267,
+      "grad_norm": 0.5380592346191406,
+      "learning_rate": 8e-05,
+      "loss": 1.6558,
+      "step": 7271
+    },
+    {
+      "epoch": 0.40535117056856185,
+      "grad_norm": 0.5165267586708069,
+      "learning_rate": 8e-05,
+      "loss": 1.6123,
+      "step": 7272
+    },
+    {
+      "epoch": 0.4054069119286511,
+      "grad_norm": 0.527855396270752,
+      "learning_rate": 8e-05,
+      "loss": 1.5234,
+      "step": 7273
+    },
+    {
+      "epoch": 0.40546265328874026,
+      "grad_norm": 0.5355624556541443,
+      "learning_rate": 8e-05,
+      "loss": 1.7963,
+      "step": 7274
+    },
+    {
+      "epoch": 0.40551839464882944,
+      "grad_norm": 0.5941885113716125,
+      "learning_rate": 8e-05,
+      "loss": 1.6391,
+      "step": 7275
+    },
+    {
+      "epoch": 0.4055741360089186,
+      "grad_norm": 0.525371253490448,
+      "learning_rate": 8e-05,
+      "loss": 1.7042,
+      "step": 7276
+    },
+    {
+      "epoch": 0.4056298773690078,
+      "grad_norm": 0.5089641809463501,
+      "learning_rate": 8e-05,
+      "loss": 1.5949,
+      "step": 7277
+    },
+    {
+      "epoch": 0.405685618729097,
+      "grad_norm": 0.47336864471435547,
+      "learning_rate": 8e-05,
+      "loss": 1.4321,
+      "step": 7278
+    },
+    {
+      "epoch": 0.40574136008918615,
+      "grad_norm": 0.520519495010376,
+      "learning_rate": 8e-05,
+      "loss": 1.6113,
+      "step": 7279
+    },
+    {
+      "epoch": 0.4057971014492754,
+      "grad_norm": 0.521248996257782,
+      "learning_rate": 8e-05,
+      "loss": 1.4845,
+      "step": 7280
+    },
+    {
+      "epoch": 0.40585284280936457,
+      "grad_norm": 0.51210618019104,
+      "learning_rate": 8e-05,
+      "loss": 1.6622,
+      "step": 7281
+    },
+    {
+      "epoch": 0.40590858416945375,
+      "grad_norm": 0.5206047892570496,
+      "learning_rate": 8e-05,
+      "loss": 1.8067,
+      "step": 7282
+    },
+    {
+      "epoch": 0.4059643255295429,
+      "grad_norm": 0.5226845145225525,
+      "learning_rate": 8e-05,
+      "loss": 1.644,
+      "step": 7283
+    },
+    {
+      "epoch": 0.4060200668896321,
+      "grad_norm": 0.5129978656768799,
+      "learning_rate": 8e-05,
+      "loss": 1.5691,
+      "step": 7284
+    },
+    {
+      "epoch": 0.4060758082497213,
+      "grad_norm": 0.5226308107376099,
+      "learning_rate": 8e-05,
+      "loss": 1.4112,
+      "step": 7285
+    },
+    {
+      "epoch": 0.40613154960981046,
+      "grad_norm": 0.5508782267570496,
+      "learning_rate": 8e-05,
+      "loss": 1.6141,
+      "step": 7286
+    },
+    {
+      "epoch": 0.40618729096989964,
+      "grad_norm": 0.54766446352005,
+      "learning_rate": 8e-05,
+      "loss": 1.7305,
+      "step": 7287
+    },
+    {
+      "epoch": 0.4062430323299889,
+      "grad_norm": 0.5656951665878296,
+      "learning_rate": 8e-05,
+      "loss": 1.8182,
+      "step": 7288
+    },
+    {
+      "epoch": 0.40629877369007805,
+      "grad_norm": 0.6196290850639343,
+      "learning_rate": 8e-05,
+      "loss": 1.7658,
+      "step": 7289
+    },
+    {
+      "epoch": 0.40635451505016723,
+      "grad_norm": 0.5594131350517273,
+      "learning_rate": 8e-05,
+      "loss": 1.6559,
+      "step": 7290
+    },
+    {
+      "epoch": 0.4064102564102564,
+      "grad_norm": 0.7854731678962708,
+      "learning_rate": 8e-05,
+      "loss": 1.3569,
+      "step": 7291
+    },
+    {
+      "epoch": 0.4064659977703456,
+      "grad_norm": 0.49075648188591003,
+      "learning_rate": 8e-05,
+      "loss": 1.6244,
+      "step": 7292
+    },
+    {
+      "epoch": 0.40652173913043477,
+      "grad_norm": 0.5666341781616211,
+      "learning_rate": 8e-05,
+      "loss": 1.7787,
+      "step": 7293
+    },
+    {
+      "epoch": 0.40657748049052395,
+      "grad_norm": 0.5504280924797058,
+      "learning_rate": 8e-05,
+      "loss": 1.7641,
+      "step": 7294
+    },
+    {
+      "epoch": 0.4066332218506132,
+      "grad_norm": 0.5278729200363159,
+      "learning_rate": 8e-05,
+      "loss": 1.7668,
+      "step": 7295
+    },
+    {
+      "epoch": 0.40668896321070236,
+      "grad_norm": 0.5682119131088257,
+      "learning_rate": 8e-05,
+      "loss": 1.6458,
+      "step": 7296
+    },
+    {
+      "epoch": 0.40674470457079154,
+      "grad_norm": 0.6224889755249023,
+      "learning_rate": 8e-05,
+      "loss": 1.8824,
+      "step": 7297
+    },
+    {
+      "epoch": 0.4068004459308807,
+      "grad_norm": 0.5042921304702759,
+      "learning_rate": 8e-05,
+      "loss": 1.614,
+      "step": 7298
+    },
+    {
+      "epoch": 0.4068561872909699,
+      "grad_norm": 0.5012341141700745,
+      "learning_rate": 8e-05,
+      "loss": 1.5258,
+      "step": 7299
+    },
+    {
+      "epoch": 0.4069119286510591,
+      "grad_norm": 0.5259014964103699,
+      "learning_rate": 8e-05,
+      "loss": 1.6493,
+      "step": 7300
+    },
+    {
+      "epoch": 0.40696767001114825,
+      "grad_norm": 0.493158221244812,
+      "learning_rate": 8e-05,
+      "loss": 1.4587,
+      "step": 7301
+    },
+    {
+      "epoch": 0.40702341137123743,
+      "grad_norm": 0.5391311049461365,
+      "learning_rate": 8e-05,
+      "loss": 1.8054,
+      "step": 7302
+    },
+    {
+      "epoch": 0.40707915273132667,
+      "grad_norm": 0.5361586213111877,
+      "learning_rate": 8e-05,
+      "loss": 1.4909,
+      "step": 7303
+    },
+    {
+      "epoch": 0.40713489409141584,
+      "grad_norm": 0.5893044471740723,
+      "learning_rate": 8e-05,
+      "loss": 1.8674,
+      "step": 7304
+    },
+    {
+      "epoch": 0.407190635451505,
+      "grad_norm": 0.540810763835907,
+      "learning_rate": 8e-05,
+      "loss": 1.6365,
+      "step": 7305
+    },
+    {
+      "epoch": 0.4072463768115942,
+      "grad_norm": 0.5277107357978821,
+      "learning_rate": 8e-05,
+      "loss": 1.6131,
+      "step": 7306
+    },
+    {
+      "epoch": 0.4073021181716834,
+      "grad_norm": 0.47652947902679443,
+      "learning_rate": 8e-05,
+      "loss": 1.3556,
+      "step": 7307
+    },
+    {
+      "epoch": 0.40735785953177256,
+      "grad_norm": 0.5251367688179016,
+      "learning_rate": 8e-05,
+      "loss": 1.795,
+      "step": 7308
+    },
+    {
+      "epoch": 0.40741360089186174,
+      "grad_norm": 0.5432030558586121,
+      "learning_rate": 8e-05,
+      "loss": 2.0776,
+      "step": 7309
+    },
+    {
+      "epoch": 0.40746934225195097,
+      "grad_norm": 0.5877689719200134,
+      "learning_rate": 8e-05,
+      "loss": 1.88,
+      "step": 7310
+    },
+    {
+      "epoch": 0.40752508361204015,
+      "grad_norm": 0.5788148641586304,
+      "learning_rate": 8e-05,
+      "loss": 1.8192,
+      "step": 7311
+    },
+    {
+      "epoch": 0.40758082497212933,
+      "grad_norm": 0.551967978477478,
+      "learning_rate": 8e-05,
+      "loss": 1.7107,
+      "step": 7312
+    },
+    {
+      "epoch": 0.4076365663322185,
+      "grad_norm": 0.49300578236579895,
+      "learning_rate": 8e-05,
+      "loss": 1.599,
+      "step": 7313
+    },
+    {
+      "epoch": 0.4076923076923077,
+      "grad_norm": 0.505588173866272,
+      "learning_rate": 8e-05,
+      "loss": 1.4808,
+      "step": 7314
+    },
+    {
+      "epoch": 0.40774804905239687,
+      "grad_norm": 0.6424862742424011,
+      "learning_rate": 8e-05,
+      "loss": 1.3891,
+      "step": 7315
+    },
+    {
+      "epoch": 0.40780379041248604,
+      "grad_norm": 0.5149382948875427,
+      "learning_rate": 8e-05,
+      "loss": 1.5358,
+      "step": 7316
+    },
+    {
+      "epoch": 0.4078595317725752,
+      "grad_norm": 0.5695319771766663,
+      "learning_rate": 8e-05,
+      "loss": 1.7848,
+      "step": 7317
+    },
+    {
+      "epoch": 0.40791527313266446,
+      "grad_norm": 0.562187910079956,
+      "learning_rate": 8e-05,
+      "loss": 2.0304,
+      "step": 7318
+    },
+    {
+      "epoch": 0.40797101449275364,
+      "grad_norm": 0.5698602795600891,
+      "learning_rate": 8e-05,
+      "loss": 1.7808,
+      "step": 7319
+    },
+    {
+      "epoch": 0.4080267558528428,
+      "grad_norm": 0.4977051615715027,
+      "learning_rate": 8e-05,
+      "loss": 1.528,
+      "step": 7320
+    },
+    {
+      "epoch": 0.408082497212932,
+      "grad_norm": 0.48804786801338196,
+      "learning_rate": 8e-05,
+      "loss": 1.8455,
+      "step": 7321
+    },
+    {
+      "epoch": 0.4081382385730212,
+      "grad_norm": 0.4927138090133667,
+      "learning_rate": 8e-05,
+      "loss": 1.5468,
+      "step": 7322
+    },
+    {
+      "epoch": 0.40819397993311035,
+      "grad_norm": 0.5586680173873901,
+      "learning_rate": 8e-05,
+      "loss": 1.5726,
+      "step": 7323
+    },
+    {
+      "epoch": 0.40824972129319953,
+      "grad_norm": 0.5101229548454285,
+      "learning_rate": 8e-05,
+      "loss": 1.704,
+      "step": 7324
+    },
+    {
+      "epoch": 0.40830546265328876,
+      "grad_norm": 0.5254568457603455,
+      "learning_rate": 8e-05,
+      "loss": 1.7521,
+      "step": 7325
+    },
+    {
+      "epoch": 0.40836120401337794,
+      "grad_norm": 0.490192174911499,
+      "learning_rate": 8e-05,
+      "loss": 1.5068,
+      "step": 7326
+    },
+    {
+      "epoch": 0.4084169453734671,
+      "grad_norm": 0.5417366623878479,
+      "learning_rate": 8e-05,
+      "loss": 1.7213,
+      "step": 7327
+    },
+    {
+      "epoch": 0.4084726867335563,
+      "grad_norm": 0.5311920642852783,
+      "learning_rate": 8e-05,
+      "loss": 1.7244,
+      "step": 7328
+    },
+    {
+      "epoch": 0.4085284280936455,
+      "grad_norm": 0.5254058837890625,
+      "learning_rate": 8e-05,
+      "loss": 1.4401,
+      "step": 7329
+    },
+    {
+      "epoch": 0.40858416945373466,
+      "grad_norm": 0.5589635968208313,
+      "learning_rate": 8e-05,
+      "loss": 1.7735,
+      "step": 7330
+    },
+    {
+      "epoch": 0.40863991081382384,
+      "grad_norm": 0.5018014907836914,
+      "learning_rate": 8e-05,
+      "loss": 1.6477,
+      "step": 7331
+    },
+    {
+      "epoch": 0.40869565217391307,
+      "grad_norm": 0.525607168674469,
+      "learning_rate": 8e-05,
+      "loss": 1.5617,
+      "step": 7332
+    },
+    {
+      "epoch": 0.40875139353400225,
+      "grad_norm": 0.5388331413269043,
+      "learning_rate": 8e-05,
+      "loss": 1.6982,
+      "step": 7333
+    },
+    {
+      "epoch": 0.40880713489409143,
+      "grad_norm": 0.5607791543006897,
+      "learning_rate": 8e-05,
+      "loss": 1.7448,
+      "step": 7334
+    },
+    {
+      "epoch": 0.4088628762541806,
+      "grad_norm": 0.5692734122276306,
+      "learning_rate": 8e-05,
+      "loss": 1.5466,
+      "step": 7335
+    },
+    {
+      "epoch": 0.4089186176142698,
+      "grad_norm": 0.5394876003265381,
+      "learning_rate": 8e-05,
+      "loss": 1.7048,
+      "step": 7336
+    },
+    {
+      "epoch": 0.40897435897435896,
+      "grad_norm": 0.5117727518081665,
+      "learning_rate": 8e-05,
+      "loss": 1.5873,
+      "step": 7337
+    },
+    {
+      "epoch": 0.40903010033444814,
+      "grad_norm": 0.5213082432746887,
+      "learning_rate": 8e-05,
+      "loss": 1.6449,
+      "step": 7338
+    },
+    {
+      "epoch": 0.4090858416945373,
+      "grad_norm": 0.5686284303665161,
+      "learning_rate": 8e-05,
+      "loss": 2.0047,
+      "step": 7339
+    },
+    {
+      "epoch": 0.40914158305462656,
+      "grad_norm": 0.5361032485961914,
+      "learning_rate": 8e-05,
+      "loss": 1.9062,
+      "step": 7340
+    },
+    {
+      "epoch": 0.40919732441471574,
+      "grad_norm": 0.5334948897361755,
+      "learning_rate": 8e-05,
+      "loss": 1.812,
+      "step": 7341
+    },
+    {
+      "epoch": 0.4092530657748049,
+      "grad_norm": 0.5158889293670654,
+      "learning_rate": 8e-05,
+      "loss": 1.649,
+      "step": 7342
+    },
+    {
+      "epoch": 0.4093088071348941,
+      "grad_norm": 0.4964984357357025,
+      "learning_rate": 8e-05,
+      "loss": 1.4696,
+      "step": 7343
+    },
+    {
+      "epoch": 0.40936454849498327,
+      "grad_norm": NaN,
+      "learning_rate": 8e-05,
+      "loss": 1.7212,
+      "step": 7344
+    },
+    {
+      "epoch": 0.40942028985507245,
+      "grad_norm": 0.5129902362823486,
+      "learning_rate": 8e-05,
+      "loss": 1.6362,
+      "step": 7345
+    },
+    {
+      "epoch": 0.40947603121516163,
+      "grad_norm": 0.5674779415130615,
+      "learning_rate": 8e-05,
+      "loss": 1.634,
+      "step": 7346
+    },
+    {
+      "epoch": 0.40953177257525086,
+      "grad_norm": 0.4873337149620056,
+      "learning_rate": 8e-05,
+      "loss": 1.7257,
+      "step": 7347
+    },
+    {
+      "epoch": 0.40958751393534004,
+      "grad_norm": 0.5569888353347778,
+      "learning_rate": 8e-05,
+      "loss": 1.6974,
+      "step": 7348
+    },
+    {
+      "epoch": 0.4096432552954292,
+      "grad_norm": 0.5687206387519836,
+      "learning_rate": 8e-05,
+      "loss": 1.8742,
+      "step": 7349
+    },
+    {
+      "epoch": 0.4096989966555184,
+      "grad_norm": 0.7041868567466736,
+      "learning_rate": 8e-05,
+      "loss": 1.7045,
+      "step": 7350
+    },
+    {
+      "epoch": 0.4097547380156076,
+      "grad_norm": 0.5319523811340332,
+      "learning_rate": 8e-05,
+      "loss": 1.759,
+      "step": 7351
+    },
+    {
+      "epoch": 0.40981047937569676,
+      "grad_norm": 0.523441731929779,
+      "learning_rate": 8e-05,
+      "loss": 1.8373,
+      "step": 7352
+    },
+    {
+      "epoch": 0.40986622073578594,
+      "grad_norm": 0.5501686930656433,
+      "learning_rate": 8e-05,
+      "loss": 1.6488,
+      "step": 7353
+    },
+    {
+      "epoch": 0.4099219620958751,
+      "grad_norm": 0.4980147182941437,
+      "learning_rate": 8e-05,
+      "loss": 1.6566,
+      "step": 7354
+    },
+    {
+      "epoch": 0.40997770345596435,
+      "grad_norm": 0.5074655413627625,
+      "learning_rate": 8e-05,
+      "loss": 1.5941,
+      "step": 7355
+    },
+    {
+      "epoch": 0.4100334448160535,
+      "grad_norm": 0.5464773178100586,
+      "learning_rate": 8e-05,
+      "loss": 1.7869,
+      "step": 7356
+    },
+    {
+      "epoch": 0.4100891861761427,
+      "grad_norm": 0.6277008652687073,
+      "learning_rate": 8e-05,
+      "loss": 2.2296,
+      "step": 7357
+    },
+    {
+      "epoch": 0.4101449275362319,
+      "grad_norm": 0.5693956613540649,
+      "learning_rate": 8e-05,
+      "loss": 1.7652,
+      "step": 7358
+    },
+    {
+      "epoch": 0.41020066889632106,
+      "grad_norm": 0.5990443229675293,
+      "learning_rate": 8e-05,
+      "loss": 2.0915,
+      "step": 7359
+    },
+    {
+      "epoch": 0.41025641025641024,
+      "grad_norm": 0.5059307813644409,
+      "learning_rate": 8e-05,
+      "loss": 1.6951,
+      "step": 7360
+    },
+    {
+      "epoch": 0.4103121516164994,
+      "grad_norm": 0.5810990333557129,
+      "learning_rate": 8e-05,
+      "loss": 2.0471,
+      "step": 7361
+    },
+    {
+      "epoch": 0.41036789297658866,
+      "grad_norm": 0.5133877992630005,
+      "learning_rate": 8e-05,
+      "loss": 1.4961,
+      "step": 7362
+    },
+    {
+      "epoch": 0.41042363433667783,
+      "grad_norm": 0.5552747845649719,
+      "learning_rate": 8e-05,
+      "loss": 1.8481,
+      "step": 7363
+    },
+    {
+      "epoch": 0.410479375696767,
+      "grad_norm": 0.5545029044151306,
+      "learning_rate": 8e-05,
+      "loss": 1.5988,
+      "step": 7364
+    },
+    {
+      "epoch": 0.4105351170568562,
+      "grad_norm": 0.522911787033081,
+      "learning_rate": 8e-05,
+      "loss": 1.6247,
+      "step": 7365
+    },
+    {
+      "epoch": 0.41059085841694537,
+      "grad_norm": 0.5258331894874573,
+      "learning_rate": 8e-05,
+      "loss": 1.7336,
+      "step": 7366
+    },
+    {
+      "epoch": 0.41064659977703455,
+      "grad_norm": 0.5287671685218811,
+      "learning_rate": 8e-05,
+      "loss": 1.5069,
+      "step": 7367
+    },
+    {
+      "epoch": 0.4107023411371237,
+      "grad_norm": 0.5577136278152466,
+      "learning_rate": 8e-05,
+      "loss": 1.5476,
+      "step": 7368
+    },
+    {
+      "epoch": 0.4107580824972129,
+      "grad_norm": 0.5114900469779968,
+      "learning_rate": 8e-05,
+      "loss": 1.5314,
+      "step": 7369
+    },
+    {
+      "epoch": 0.41081382385730214,
+      "grad_norm": 0.5116535425186157,
+      "learning_rate": 8e-05,
+      "loss": 1.6569,
+      "step": 7370
+    },
+    {
+      "epoch": 0.4108695652173913,
+      "grad_norm": 0.5354006290435791,
+      "learning_rate": 8e-05,
+      "loss": 1.628,
+      "step": 7371
+    },
+    {
+      "epoch": 0.4109253065774805,
+      "grad_norm": 0.4987451434135437,
+      "learning_rate": 8e-05,
+      "loss": 1.6674,
+      "step": 7372
+    },
+    {
+      "epoch": 0.4109810479375697,
+      "grad_norm": 0.5362124443054199,
+      "learning_rate": 8e-05,
+      "loss": 1.7775,
+      "step": 7373
+    },
+    {
+      "epoch": 0.41103678929765886,
+      "grad_norm": 0.5331243872642517,
+      "learning_rate": 8e-05,
+      "loss": 1.4266,
+      "step": 7374
+    },
+    {
+      "epoch": 0.41109253065774803,
+      "grad_norm": 0.552800178527832,
+      "learning_rate": 8e-05,
+      "loss": 1.5609,
+      "step": 7375
+    },
+    {
+      "epoch": 0.4111482720178372,
+      "grad_norm": 0.5207287669181824,
+      "learning_rate": 8e-05,
+      "loss": 1.7087,
+      "step": 7376
+    },
+    {
+      "epoch": 0.41120401337792645,
+      "grad_norm": 0.5250851511955261,
+      "learning_rate": 8e-05,
+      "loss": 1.6695,
+      "step": 7377
+    },
+    {
+      "epoch": 0.4112597547380156,
+      "grad_norm": 0.552610456943512,
+      "learning_rate": 8e-05,
+      "loss": 1.8226,
+      "step": 7378
+    },
+    {
+      "epoch": 0.4113154960981048,
+      "grad_norm": 0.5105003714561462,
+      "learning_rate": 8e-05,
+      "loss": 1.7584,
+      "step": 7379
+    },
+    {
+      "epoch": 0.411371237458194,
+      "grad_norm": 0.5159006714820862,
+      "learning_rate": 8e-05,
+      "loss": 1.627,
+      "step": 7380
+    },
+    {
+      "epoch": 0.41142697881828316,
+      "grad_norm": 0.5639457106590271,
+      "learning_rate": 8e-05,
+      "loss": 1.8453,
+      "step": 7381
+    },
+    {
+      "epoch": 0.41148272017837234,
+      "grad_norm": 0.4808430075645447,
+      "learning_rate": 8e-05,
+      "loss": 1.7023,
+      "step": 7382
+    },
+    {
+      "epoch": 0.4115384615384615,
+      "grad_norm": 0.5087721347808838,
+      "learning_rate": 8e-05,
+      "loss": 1.7405,
+      "step": 7383
+    },
+    {
+      "epoch": 0.4115942028985507,
+      "grad_norm": 0.5479757189750671,
+      "learning_rate": 8e-05,
+      "loss": 1.7867,
+      "step": 7384
+    },
+    {
+      "epoch": 0.41164994425863993,
+      "grad_norm": 0.5063915252685547,
+      "learning_rate": 8e-05,
+      "loss": 1.7109,
+      "step": 7385
+    },
+    {
+      "epoch": 0.4117056856187291,
+      "grad_norm": 0.5211130380630493,
+      "learning_rate": 8e-05,
+      "loss": 1.6334,
+      "step": 7386
+    },
+    {
+      "epoch": 0.4117614269788183,
+      "grad_norm": 0.5052521824836731,
+      "learning_rate": 8e-05,
+      "loss": 1.8544,
+      "step": 7387
+    },
+    {
+      "epoch": 0.41181716833890747,
+      "grad_norm": 0.5772587060928345,
+      "learning_rate": 8e-05,
+      "loss": 1.9451,
+      "step": 7388
+    },
+    {
+      "epoch": 0.41187290969899665,
+      "grad_norm": 0.5105565786361694,
+      "learning_rate": 8e-05,
+      "loss": 1.5803,
+      "step": 7389
+    },
+    {
+      "epoch": 0.4119286510590858,
+      "grad_norm": 0.5484120845794678,
+      "learning_rate": 8e-05,
+      "loss": 1.7197,
+      "step": 7390
+    },
+    {
+      "epoch": 0.411984392419175,
+      "grad_norm": 0.5253351926803589,
+      "learning_rate": 8e-05,
+      "loss": 1.5121,
+      "step": 7391
+    },
+    {
+      "epoch": 0.41204013377926424,
+      "grad_norm": 0.523414134979248,
+      "learning_rate": 8e-05,
+      "loss": 1.6663,
+      "step": 7392
+    },
+    {
+      "epoch": 0.4120958751393534,
+      "grad_norm": 0.5172882676124573,
+      "learning_rate": 8e-05,
+      "loss": 1.5974,
+      "step": 7393
+    },
+    {
+      "epoch": 0.4121516164994426,
+      "grad_norm": 0.5120169520378113,
+      "learning_rate": 8e-05,
+      "loss": 1.6205,
+      "step": 7394
+    },
+    {
+      "epoch": 0.4122073578595318,
+      "grad_norm": 0.5658069849014282,
+      "learning_rate": 8e-05,
+      "loss": 1.6488,
+      "step": 7395
+    },
+    {
+      "epoch": 0.41226309921962095,
+      "grad_norm": 0.5338782668113708,
+      "learning_rate": 8e-05,
+      "loss": 1.8355,
+      "step": 7396
+    },
+    {
+      "epoch": 0.41231884057971013,
+      "grad_norm": 0.504389226436615,
+      "learning_rate": 8e-05,
+      "loss": 1.7015,
+      "step": 7397
+    },
+    {
+      "epoch": 0.4123745819397993,
+      "grad_norm": 0.535618007183075,
+      "learning_rate": 8e-05,
+      "loss": 1.5951,
+      "step": 7398
+    },
+    {
+      "epoch": 0.4124303232998885,
+      "grad_norm": 0.501131534576416,
+      "learning_rate": 8e-05,
+      "loss": 1.733,
+      "step": 7399
+    },
+    {
+      "epoch": 0.4124860646599777,
+      "grad_norm": 0.5304338932037354,
+      "learning_rate": 8e-05,
+      "loss": 1.674,
+      "step": 7400
+    },
+    {
+      "epoch": 0.4125418060200669,
+      "grad_norm": 0.6905393004417419,
+      "learning_rate": 8e-05,
+      "loss": 1.4084,
+      "step": 7401
+    },
+    {
+      "epoch": 0.4125975473801561,
+      "grad_norm": 0.5419348478317261,
+      "learning_rate": 8e-05,
+      "loss": 1.7082,
+      "step": 7402
+    },
+    {
+      "epoch": 0.41265328874024526,
+      "grad_norm": 0.4964982867240906,
+      "learning_rate": 8e-05,
+      "loss": 1.5791,
+      "step": 7403
+    },
+    {
+      "epoch": 0.41270903010033444,
+      "grad_norm": 0.5790861248970032,
+      "learning_rate": 8e-05,
+      "loss": 1.8076,
+      "step": 7404
+    },
+    {
+      "epoch": 0.4127647714604236,
+      "grad_norm": 0.5126116871833801,
+      "learning_rate": 8e-05,
+      "loss": 1.7128,
+      "step": 7405
+    },
+    {
+      "epoch": 0.4128205128205128,
+      "grad_norm": 0.5443001985549927,
+      "learning_rate": 8e-05,
+      "loss": 1.7598,
+      "step": 7406
+    },
+    {
+      "epoch": 0.41287625418060203,
+      "grad_norm": 0.4740257263183594,
+      "learning_rate": 8e-05,
+      "loss": 1.4271,
+      "step": 7407
+    },
+    {
+      "epoch": 0.4129319955406912,
+      "grad_norm": 0.5965816974639893,
+      "learning_rate": 8e-05,
+      "loss": 1.8464,
+      "step": 7408
+    },
+    {
+      "epoch": 0.4129877369007804,
+      "grad_norm": 0.5371137261390686,
+      "learning_rate": 8e-05,
+      "loss": 1.561,
+      "step": 7409
+    },
+    {
+      "epoch": 0.41304347826086957,
+      "grad_norm": 0.5329623222351074,
+      "learning_rate": 8e-05,
+      "loss": 1.7707,
+      "step": 7410
+    },
+    {
+      "epoch": 0.41309921962095875,
+      "grad_norm": 0.5217059254646301,
+      "learning_rate": 8e-05,
+      "loss": 1.5821,
+      "step": 7411
+    },
+    {
+      "epoch": 0.4131549609810479,
+      "grad_norm": 0.5564699172973633,
+      "learning_rate": 8e-05,
+      "loss": 1.7106,
+      "step": 7412
+    },
+    {
+      "epoch": 0.4132107023411371,
+      "grad_norm": 0.5382006168365479,
+      "learning_rate": 8e-05,
+      "loss": 1.7405,
+      "step": 7413
+    },
+    {
+      "epoch": 0.4132664437012263,
+      "grad_norm": 0.5441609025001526,
+      "learning_rate": 8e-05,
+      "loss": 1.6926,
+      "step": 7414
+    },
+    {
+      "epoch": 0.4133221850613155,
+      "grad_norm": 0.578251302242279,
+      "learning_rate": 8e-05,
+      "loss": 1.51,
+      "step": 7415
+    },
+    {
+      "epoch": 0.4133779264214047,
+      "grad_norm": 0.5474327802658081,
+      "learning_rate": 8e-05,
+      "loss": 1.7875,
+      "step": 7416
+    },
+    {
+      "epoch": 0.4134336677814939,
+      "grad_norm": 0.5703834891319275,
+      "learning_rate": 8e-05,
+      "loss": 1.8034,
+      "step": 7417
+    },
+    {
+      "epoch": 0.41348940914158305,
+      "grad_norm": 0.5354348421096802,
+      "learning_rate": 8e-05,
+      "loss": 1.746,
+      "step": 7418
+    },
+    {
+      "epoch": 0.41354515050167223,
+      "grad_norm": 0.5187190771102905,
+      "learning_rate": 8e-05,
+      "loss": 1.6389,
+      "step": 7419
+    },
+    {
+      "epoch": 0.4136008918617614,
+      "grad_norm": 0.5314764380455017,
+      "learning_rate": 8e-05,
+      "loss": 1.7257,
+      "step": 7420
+    },
+    {
+      "epoch": 0.4136566332218506,
+      "grad_norm": 0.5247149467468262,
+      "learning_rate": 8e-05,
+      "loss": 1.568,
+      "step": 7421
+    },
+    {
+      "epoch": 0.4137123745819398,
+      "grad_norm": 0.4993450343608856,
+      "learning_rate": 8e-05,
+      "loss": 1.5826,
+      "step": 7422
+    },
+    {
+      "epoch": 0.413768115942029,
+      "grad_norm": 0.5261116623878479,
+      "learning_rate": 8e-05,
+      "loss": 1.6117,
+      "step": 7423
+    },
+    {
+      "epoch": 0.4138238573021182,
+      "grad_norm": 0.5468164086341858,
+      "learning_rate": 8e-05,
+      "loss": 1.5543,
+      "step": 7424
+    },
+    {
+      "epoch": 0.41387959866220736,
+      "grad_norm": 0.5935494303703308,
+      "learning_rate": 8e-05,
+      "loss": 1.7585,
+      "step": 7425
+    },
+    {
+      "epoch": 0.41393534002229654,
+      "grad_norm": 0.5638031363487244,
+      "learning_rate": 8e-05,
+      "loss": 1.9934,
+      "step": 7426
+    },
+    {
+      "epoch": 0.4139910813823857,
+      "grad_norm": 0.47739219665527344,
+      "learning_rate": 8e-05,
+      "loss": 1.5811,
+      "step": 7427
+    },
+    {
+      "epoch": 0.4140468227424749,
+      "grad_norm": 0.5171627998352051,
+      "learning_rate": 8e-05,
+      "loss": 1.7371,
+      "step": 7428
+    },
+    {
+      "epoch": 0.41410256410256413,
+      "grad_norm": 0.5365909337997437,
+      "learning_rate": 8e-05,
+      "loss": 1.7186,
+      "step": 7429
+    },
+    {
+      "epoch": 0.4141583054626533,
+      "grad_norm": 0.5287392735481262,
+      "learning_rate": 8e-05,
+      "loss": 1.5597,
+      "step": 7430
+    },
+    {
+      "epoch": 0.4142140468227425,
+      "grad_norm": 0.5019273161888123,
+      "learning_rate": 8e-05,
+      "loss": 1.4832,
+      "step": 7431
+    },
+    {
+      "epoch": 0.41426978818283167,
+      "grad_norm": 0.535295844078064,
+      "learning_rate": 8e-05,
+      "loss": 1.6458,
+      "step": 7432
+    },
+    {
+      "epoch": 0.41432552954292085,
+      "grad_norm": 0.5648143291473389,
+      "learning_rate": 8e-05,
+      "loss": 1.7733,
+      "step": 7433
+    },
+    {
+      "epoch": 0.41438127090301,
+      "grad_norm": 0.5082452893257141,
+      "learning_rate": 8e-05,
+      "loss": 1.7593,
+      "step": 7434
+    },
+    {
+      "epoch": 0.4144370122630992,
+      "grad_norm": 0.515472412109375,
+      "learning_rate": 8e-05,
+      "loss": 1.5832,
+      "step": 7435
+    },
+    {
+      "epoch": 0.4144927536231884,
+      "grad_norm": 0.540582001209259,
+      "learning_rate": 8e-05,
+      "loss": 1.6328,
+      "step": 7436
+    },
+    {
+      "epoch": 0.4145484949832776,
+      "grad_norm": 0.5583634972572327,
+      "learning_rate": 8e-05,
+      "loss": 1.8097,
+      "step": 7437
+    },
+    {
+      "epoch": 0.4146042363433668,
+      "grad_norm": 0.5404437184333801,
+      "learning_rate": 8e-05,
+      "loss": 1.7001,
+      "step": 7438
+    },
+    {
+      "epoch": 0.414659977703456,
+      "grad_norm": 0.5578569769859314,
+      "learning_rate": 8e-05,
+      "loss": 1.6861,
+      "step": 7439
+    },
+    {
+      "epoch": 0.41471571906354515,
+      "grad_norm": 0.5309574007987976,
+      "learning_rate": 8e-05,
+      "loss": 1.5784,
+      "step": 7440
+    },
+    {
+      "epoch": 0.41477146042363433,
+      "grad_norm": 0.5452367663383484,
+      "learning_rate": 8e-05,
+      "loss": 1.687,
+      "step": 7441
+    },
+    {
+      "epoch": 0.4148272017837235,
+      "grad_norm": 0.5491836071014404,
+      "learning_rate": 8e-05,
+      "loss": 1.6507,
+      "step": 7442
+    },
+    {
+      "epoch": 0.4148829431438127,
+      "grad_norm": 0.5284445285797119,
+      "learning_rate": 8e-05,
+      "loss": 1.5607,
+      "step": 7443
+    },
+    {
+      "epoch": 0.4149386845039019,
+      "grad_norm": 0.5563288927078247,
+      "learning_rate": 8e-05,
+      "loss": 1.666,
+      "step": 7444
+    },
+    {
+      "epoch": 0.4149944258639911,
+      "grad_norm": 0.5210450887680054,
+      "learning_rate": 8e-05,
+      "loss": 1.5811,
+      "step": 7445
+    },
+    {
+      "epoch": 0.4150501672240803,
+      "grad_norm": 0.5133852958679199,
+      "learning_rate": 8e-05,
+      "loss": 1.7379,
+      "step": 7446
+    },
+    {
+      "epoch": 0.41510590858416946,
+      "grad_norm": 0.5505521297454834,
+      "learning_rate": 8e-05,
+      "loss": 1.7915,
+      "step": 7447
+    },
+    {
+      "epoch": 0.41516164994425864,
+      "grad_norm": 0.5652416348457336,
+      "learning_rate": 8e-05,
+      "loss": 1.6962,
+      "step": 7448
+    },
+    {
+      "epoch": 0.4152173913043478,
+      "grad_norm": 0.4884105622768402,
+      "learning_rate": 8e-05,
+      "loss": 1.6399,
+      "step": 7449
+    },
+    {
+      "epoch": 0.415273132664437,
+      "grad_norm": 0.5113058686256409,
+      "learning_rate": 8e-05,
+      "loss": 1.7623,
+      "step": 7450
+    },
+    {
+      "epoch": 0.4153288740245262,
+      "grad_norm": 0.533737063407898,
+      "learning_rate": 8e-05,
+      "loss": 1.5663,
+      "step": 7451
+    },
+    {
+      "epoch": 0.4153846153846154,
+      "grad_norm": 0.5064999461174011,
+      "learning_rate": 8e-05,
+      "loss": 1.639,
+      "step": 7452
+    },
+    {
+      "epoch": 0.4154403567447046,
+      "grad_norm": 0.600654661655426,
+      "learning_rate": 8e-05,
+      "loss": 1.62,
+      "step": 7453
+    },
+    {
+      "epoch": 0.41549609810479377,
+      "grad_norm": 0.5428963899612427,
+      "learning_rate": 8e-05,
+      "loss": 1.7833,
+      "step": 7454
+    },
+    {
+      "epoch": 0.41555183946488294,
+      "grad_norm": 0.5852295756340027,
+      "learning_rate": 8e-05,
+      "loss": 1.7379,
+      "step": 7455
+    },
+    {
+      "epoch": 0.4156075808249721,
+      "grad_norm": 0.49108296632766724,
+      "learning_rate": 8e-05,
+      "loss": 1.4555,
+      "step": 7456
+    },
+    {
+      "epoch": 0.4156633221850613,
+      "grad_norm": 0.5304026007652283,
+      "learning_rate": 8e-05,
+      "loss": 1.4129,
+      "step": 7457
+    },
+    {
+      "epoch": 0.4157190635451505,
+      "grad_norm": 0.5354292392730713,
+      "learning_rate": 8e-05,
+      "loss": 1.5815,
+      "step": 7458
+    },
+    {
+      "epoch": 0.4157748049052397,
+      "grad_norm": 0.49475952982902527,
+      "learning_rate": 8e-05,
+      "loss": 1.6129,
+      "step": 7459
+    },
+    {
+      "epoch": 0.4158305462653289,
+      "grad_norm": 0.5129939317703247,
+      "learning_rate": 8e-05,
+      "loss": 1.6688,
+      "step": 7460
+    },
+    {
+      "epoch": 0.41588628762541807,
+      "grad_norm": 0.5419784784317017,
+      "learning_rate": 8e-05,
+      "loss": 1.7965,
+      "step": 7461
+    },
+    {
+      "epoch": 0.41594202898550725,
+      "grad_norm": 0.521435022354126,
+      "learning_rate": 8e-05,
+      "loss": 1.6395,
+      "step": 7462
+    },
+    {
+      "epoch": 0.41599777034559643,
+      "grad_norm": 0.5104933977127075,
+      "learning_rate": 8e-05,
+      "loss": 1.6178,
+      "step": 7463
+    },
+    {
+      "epoch": 0.4160535117056856,
+      "grad_norm": 0.5301722288131714,
+      "learning_rate": 8e-05,
+      "loss": 1.7177,
+      "step": 7464
+    },
+    {
+      "epoch": 0.4161092530657748,
+      "grad_norm": 0.5151179432868958,
+      "learning_rate": 8e-05,
+      "loss": 1.667,
+      "step": 7465
+    },
+    {
+      "epoch": 0.41616499442586397,
+      "grad_norm": 0.4956851303577423,
+      "learning_rate": 8e-05,
+      "loss": 1.6084,
+      "step": 7466
+    },
+    {
+      "epoch": 0.4162207357859532,
+      "grad_norm": 0.5060463547706604,
+      "learning_rate": 8e-05,
+      "loss": 1.4551,
+      "step": 7467
+    },
+    {
+      "epoch": 0.4162764771460424,
+      "grad_norm": 0.5542873740196228,
+      "learning_rate": 8e-05,
+      "loss": 1.4436,
+      "step": 7468
+    },
+    {
+      "epoch": 0.41633221850613156,
+      "grad_norm": 0.5257938504219055,
+      "learning_rate": 8e-05,
+      "loss": 1.5713,
+      "step": 7469
+    },
+    {
+      "epoch": 0.41638795986622074,
+      "grad_norm": 0.5059783458709717,
+      "learning_rate": 8e-05,
+      "loss": 1.5782,
+      "step": 7470
+    },
+    {
+      "epoch": 0.4164437012263099,
+      "grad_norm": 0.5536264181137085,
+      "learning_rate": 8e-05,
+      "loss": 1.8385,
+      "step": 7471
+    },
+    {
+      "epoch": 0.4164994425863991,
+      "grad_norm": 0.5066181421279907,
+      "learning_rate": 8e-05,
+      "loss": 1.5314,
+      "step": 7472
+    },
+    {
+      "epoch": 0.41655518394648827,
+      "grad_norm": 0.5219850540161133,
+      "learning_rate": 8e-05,
+      "loss": 1.7979,
+      "step": 7473
+    },
+    {
+      "epoch": 0.4166109253065775,
+      "grad_norm": 0.5205042362213135,
+      "learning_rate": 8e-05,
+      "loss": 1.6734,
+      "step": 7474
+    },
+    {
+      "epoch": 0.4166666666666667,
+      "grad_norm": 0.49642428755760193,
+      "learning_rate": 8e-05,
+      "loss": 1.697,
+      "step": 7475
+    },
+    {
+      "epoch": 0.41672240802675586,
+      "grad_norm": 0.4865456223487854,
+      "learning_rate": 8e-05,
+      "loss": 1.5118,
+      "step": 7476
+    },
+    {
+      "epoch": 0.41677814938684504,
+      "grad_norm": 0.5142179131507874,
+      "learning_rate": 8e-05,
+      "loss": 1.5317,
+      "step": 7477
+    },
+    {
+      "epoch": 0.4168338907469342,
+      "grad_norm": 0.5317469835281372,
+      "learning_rate": 8e-05,
+      "loss": 1.601,
+      "step": 7478
+    },
+    {
+      "epoch": 0.4168896321070234,
+      "grad_norm": 0.501403272151947,
+      "learning_rate": 8e-05,
+      "loss": 1.6438,
+      "step": 7479
+    },
+    {
+      "epoch": 0.4169453734671126,
+      "grad_norm": 0.5238836407661438,
+      "learning_rate": 8e-05,
+      "loss": 1.7402,
+      "step": 7480
+    },
+    {
+      "epoch": 0.41700111482720176,
+      "grad_norm": 0.6039742231369019,
+      "learning_rate": 8e-05,
+      "loss": 1.8983,
+      "step": 7481
+    },
+    {
+      "epoch": 0.417056856187291,
+      "grad_norm": 0.593430757522583,
+      "learning_rate": 8e-05,
+      "loss": 1.9938,
+      "step": 7482
+    },
+    {
+      "epoch": 0.41711259754738017,
+      "grad_norm": 0.5287467837333679,
+      "learning_rate": 8e-05,
+      "loss": 1.7788,
+      "step": 7483
+    },
+    {
+      "epoch": 0.41716833890746935,
+      "grad_norm": 0.5970814228057861,
+      "learning_rate": 8e-05,
+      "loss": 1.8968,
+      "step": 7484
+    },
+    {
+      "epoch": 0.41722408026755853,
+      "grad_norm": 0.49905112385749817,
+      "learning_rate": 8e-05,
+      "loss": 1.5382,
+      "step": 7485
+    },
+    {
+      "epoch": 0.4172798216276477,
+      "grad_norm": 0.5791721343994141,
+      "learning_rate": 8e-05,
+      "loss": 1.87,
+      "step": 7486
+    },
+    {
+      "epoch": 0.4173355629877369,
+      "grad_norm": 0.534599244594574,
+      "learning_rate": 8e-05,
+      "loss": 1.7648,
+      "step": 7487
+    },
+    {
+      "epoch": 0.41739130434782606,
+      "grad_norm": 0.5286064147949219,
+      "learning_rate": 8e-05,
+      "loss": 1.4987,
+      "step": 7488
+    },
+    {
+      "epoch": 0.4174470457079153,
+      "grad_norm": 0.47725871205329895,
+      "learning_rate": 8e-05,
+      "loss": 1.4845,
+      "step": 7489
+    },
+    {
+      "epoch": 0.4175027870680045,
+      "grad_norm": 0.5819158554077148,
+      "learning_rate": 8e-05,
+      "loss": 1.8628,
+      "step": 7490
+    },
+    {
+      "epoch": 0.41755852842809366,
+      "grad_norm": 0.5254594683647156,
+      "learning_rate": 8e-05,
+      "loss": 1.5828,
+      "step": 7491
+    },
+    {
+      "epoch": 0.41761426978818283,
+      "grad_norm": 0.5000957250595093,
+      "learning_rate": 8e-05,
+      "loss": 1.5273,
+      "step": 7492
+    },
+    {
+      "epoch": 0.417670011148272,
+      "grad_norm": 0.5213155746459961,
+      "learning_rate": 8e-05,
+      "loss": 1.8125,
+      "step": 7493
+    },
+    {
+      "epoch": 0.4177257525083612,
+      "grad_norm": 0.4522757828235626,
+      "learning_rate": 8e-05,
+      "loss": 1.3973,
+      "step": 7494
+    },
+    {
+      "epoch": 0.41778149386845037,
+      "grad_norm": 0.5139923691749573,
+      "learning_rate": 8e-05,
+      "loss": 1.7558,
+      "step": 7495
+    },
+    {
+      "epoch": 0.41783723522853955,
+      "grad_norm": 0.5346956849098206,
+      "learning_rate": 8e-05,
+      "loss": 1.3351,
+      "step": 7496
+    },
+    {
+      "epoch": 0.4178929765886288,
+      "grad_norm": 0.48926034569740295,
+      "learning_rate": 8e-05,
+      "loss": 1.4097,
+      "step": 7497
+    },
+    {
+      "epoch": 0.41794871794871796,
+      "grad_norm": 0.5781483054161072,
+      "learning_rate": 8e-05,
+      "loss": 1.7858,
+      "step": 7498
+    },
+    {
+      "epoch": 0.41800445930880714,
+      "grad_norm": 0.5323367118835449,
+      "learning_rate": 8e-05,
+      "loss": 1.4932,
+      "step": 7499
+    },
+    {
+      "epoch": 0.4180602006688963,
+      "grad_norm": 0.6245522499084473,
+      "learning_rate": 8e-05,
+      "loss": 1.6934,
+      "step": 7500
+    },
+    {
+      "epoch": 0.4181159420289855,
+      "grad_norm": 0.5793091654777527,
+      "learning_rate": 8e-05,
+      "loss": 1.8131,
+      "step": 7501
+    },
+    {
+      "epoch": 0.4181716833890747,
+      "grad_norm": 0.5998666286468506,
+      "learning_rate": 8e-05,
+      "loss": 1.5865,
+      "step": 7502
+    },
+    {
+      "epoch": 0.41822742474916386,
+      "grad_norm": 0.513708770275116,
+      "learning_rate": 8e-05,
+      "loss": 1.4626,
+      "step": 7503
+    },
+    {
+      "epoch": 0.4182831661092531,
+      "grad_norm": 0.5088089108467102,
+      "learning_rate": 8e-05,
+      "loss": 1.6985,
+      "step": 7504
+    },
+    {
+      "epoch": 0.41833890746934227,
+      "grad_norm": 0.5325367450714111,
+      "learning_rate": 8e-05,
+      "loss": 1.7215,
+      "step": 7505
+    },
+    {
+      "epoch": 0.41839464882943145,
+      "grad_norm": 0.5224018692970276,
+      "learning_rate": 8e-05,
+      "loss": 1.5529,
+      "step": 7506
+    },
+    {
+      "epoch": 0.4184503901895206,
+      "grad_norm": 0.5467714667320251,
+      "learning_rate": 8e-05,
+      "loss": 1.8319,
+      "step": 7507
+    },
+    {
+      "epoch": 0.4185061315496098,
+      "grad_norm": 0.5274876356124878,
+      "learning_rate": 8e-05,
+      "loss": 1.7315,
+      "step": 7508
+    },
+    {
+      "epoch": 0.418561872909699,
+      "grad_norm": 0.5074488520622253,
+      "learning_rate": 8e-05,
+      "loss": 1.5532,
+      "step": 7509
+    },
+    {
+      "epoch": 0.41861761426978816,
+      "grad_norm": 0.4909979999065399,
+      "learning_rate": 8e-05,
+      "loss": 1.4936,
+      "step": 7510
+    },
+    {
+      "epoch": 0.41867335562987734,
+      "grad_norm": 0.5296882390975952,
+      "learning_rate": 8e-05,
+      "loss": 1.6747,
+      "step": 7511
+    },
+    {
+      "epoch": 0.4187290969899666,
+      "grad_norm": 0.5459797382354736,
+      "learning_rate": 8e-05,
+      "loss": 1.7246,
+      "step": 7512
+    },
+    {
+      "epoch": 0.41878483835005575,
+      "grad_norm": 0.4935992360115051,
+      "learning_rate": 8e-05,
+      "loss": 1.5092,
+      "step": 7513
+    },
+    {
+      "epoch": 0.41884057971014493,
+      "grad_norm": 0.5477376580238342,
+      "learning_rate": 8e-05,
+      "loss": 1.7749,
+      "step": 7514
+    },
+    {
+      "epoch": 0.4188963210702341,
+      "grad_norm": 0.5266591906547546,
+      "learning_rate": 8e-05,
+      "loss": 1.7828,
+      "step": 7515
+    },
+    {
+      "epoch": 0.4189520624303233,
+      "grad_norm": 0.5456550717353821,
+      "learning_rate": 8e-05,
+      "loss": 1.6885,
+      "step": 7516
+    },
+    {
+      "epoch": 0.41900780379041247,
+      "grad_norm": 0.56923508644104,
+      "learning_rate": 8e-05,
+      "loss": 1.8137,
+      "step": 7517
+    },
+    {
+      "epoch": 0.41906354515050165,
+      "grad_norm": 0.5321353673934937,
+      "learning_rate": 8e-05,
+      "loss": 1.718,
+      "step": 7518
+    },
+    {
+      "epoch": 0.4191192865105909,
+      "grad_norm": 0.5207672715187073,
+      "learning_rate": 8e-05,
+      "loss": 1.4387,
+      "step": 7519
+    },
+    {
+      "epoch": 0.41917502787068006,
+      "grad_norm": 0.5619326233863831,
+      "learning_rate": 8e-05,
+      "loss": 1.8107,
+      "step": 7520
+    },
+    {
+      "epoch": 0.41923076923076924,
+      "grad_norm": 0.5416442155838013,
+      "learning_rate": 8e-05,
+      "loss": 1.7936,
+      "step": 7521
+    },
+    {
+      "epoch": 0.4192865105908584,
+      "grad_norm": 0.5835678577423096,
+      "learning_rate": 8e-05,
+      "loss": 1.7206,
+      "step": 7522
+    },
+    {
+      "epoch": 0.4193422519509476,
+      "grad_norm": 0.5465813279151917,
+      "learning_rate": 8e-05,
+      "loss": 1.6876,
+      "step": 7523
+    },
+    {
+      "epoch": 0.4193979933110368,
+      "grad_norm": 0.5443020462989807,
+      "learning_rate": 8e-05,
+      "loss": 1.6822,
+      "step": 7524
+    },
+    {
+      "epoch": 0.41945373467112596,
+      "grad_norm": 0.4806618392467499,
+      "learning_rate": 8e-05,
+      "loss": 1.5436,
+      "step": 7525
+    },
+    {
+      "epoch": 0.4195094760312152,
+      "grad_norm": 0.5466196537017822,
+      "learning_rate": 8e-05,
+      "loss": 1.6569,
+      "step": 7526
+    },
+    {
+      "epoch": 0.41956521739130437,
+      "grad_norm": 0.520419716835022,
+      "learning_rate": 8e-05,
+      "loss": 1.6902,
+      "step": 7527
+    },
+    {
+      "epoch": 0.41962095875139355,
+      "grad_norm": 0.512060821056366,
+      "learning_rate": 8e-05,
+      "loss": 1.7119,
+      "step": 7528
+    },
+    {
+      "epoch": 0.4196767001114827,
+      "grad_norm": 0.5478929281234741,
+      "learning_rate": 8e-05,
+      "loss": 1.5737,
+      "step": 7529
+    },
+    {
+      "epoch": 0.4197324414715719,
+      "grad_norm": 0.5517770051956177,
+      "learning_rate": 8e-05,
+      "loss": 1.8199,
+      "step": 7530
+    },
+    {
+      "epoch": 0.4197881828316611,
+      "grad_norm": 0.5207688212394714,
+      "learning_rate": 8e-05,
+      "loss": 1.3724,
+      "step": 7531
+    },
+    {
+      "epoch": 0.41984392419175026,
+      "grad_norm": 0.5325443148612976,
+      "learning_rate": 8e-05,
+      "loss": 1.7362,
+      "step": 7532
+    },
+    {
+      "epoch": 0.41989966555183944,
+      "grad_norm": 0.5486525893211365,
+      "learning_rate": 8e-05,
+      "loss": 1.6209,
+      "step": 7533
+    },
+    {
+      "epoch": 0.4199554069119287,
+      "grad_norm": 0.5634220838546753,
+      "learning_rate": 8e-05,
+      "loss": 1.6899,
+      "step": 7534
+    },
+    {
+      "epoch": 0.42001114827201785,
+      "grad_norm": 0.603767454624176,
+      "learning_rate": 8e-05,
+      "loss": 1.9529,
+      "step": 7535
+    },
+    {
+      "epoch": 0.42006688963210703,
+      "grad_norm": 0.5488311052322388,
+      "learning_rate": 8e-05,
+      "loss": 1.6988,
+      "step": 7536
+    },
+    {
+      "epoch": 0.4201226309921962,
+      "grad_norm": 0.5185443758964539,
+      "learning_rate": 8e-05,
+      "loss": 1.7136,
+      "step": 7537
+    },
+    {
+      "epoch": 0.4201783723522854,
+      "grad_norm": 0.5392804741859436,
+      "learning_rate": 8e-05,
+      "loss": 1.9436,
+      "step": 7538
+    },
+    {
+      "epoch": 0.42023411371237457,
+      "grad_norm": 0.5228851437568665,
+      "learning_rate": 8e-05,
+      "loss": 1.4759,
+      "step": 7539
+    },
+    {
+      "epoch": 0.42028985507246375,
+      "grad_norm": 0.48584380745887756,
+      "learning_rate": 8e-05,
+      "loss": 1.5736,
+      "step": 7540
+    },
+    {
+      "epoch": 0.420345596432553,
+      "grad_norm": 0.5334780812263489,
+      "learning_rate": 8e-05,
+      "loss": 1.5748,
+      "step": 7541
+    },
+    {
+      "epoch": 0.42040133779264216,
+      "grad_norm": 0.5429795384407043,
+      "learning_rate": 8e-05,
+      "loss": 1.627,
+      "step": 7542
+    },
+    {
+      "epoch": 0.42045707915273134,
+      "grad_norm": 0.5226675271987915,
+      "learning_rate": 8e-05,
+      "loss": 1.7006,
+      "step": 7543
+    },
+    {
+      "epoch": 0.4205128205128205,
+      "grad_norm": 0.5478998422622681,
+      "learning_rate": 8e-05,
+      "loss": 1.6005,
+      "step": 7544
+    },
+    {
+      "epoch": 0.4205685618729097,
+      "grad_norm": 0.5705276131629944,
+      "learning_rate": 8e-05,
+      "loss": 1.7977,
+      "step": 7545
+    },
+    {
+      "epoch": 0.4206243032329989,
+      "grad_norm": 0.5187798738479614,
+      "learning_rate": 8e-05,
+      "loss": 1.6782,
+      "step": 7546
+    },
+    {
+      "epoch": 0.42068004459308805,
+      "grad_norm": 0.5031110644340515,
+      "learning_rate": 8e-05,
+      "loss": 1.5819,
+      "step": 7547
+    },
+    {
+      "epoch": 0.42073578595317723,
+      "grad_norm": 0.5376731157302856,
+      "learning_rate": 8e-05,
+      "loss": 1.6474,
+      "step": 7548
+    },
+    {
+      "epoch": 0.42079152731326647,
+      "grad_norm": 0.5520094037055969,
+      "learning_rate": 8e-05,
+      "loss": 1.6007,
+      "step": 7549
+    },
+    {
+      "epoch": 0.42084726867335565,
+      "grad_norm": 0.51263827085495,
+      "learning_rate": 8e-05,
+      "loss": 1.5592,
+      "step": 7550
+    },
+    {
+      "epoch": 0.4209030100334448,
+      "grad_norm": 0.5615968108177185,
+      "learning_rate": 8e-05,
+      "loss": 1.6022,
+      "step": 7551
+    },
+    {
+      "epoch": 0.420958751393534,
+      "grad_norm": 0.5113683938980103,
+      "learning_rate": 8e-05,
+      "loss": 1.6102,
+      "step": 7552
+    },
+    {
+      "epoch": 0.4210144927536232,
+      "grad_norm": 0.5493238568305969,
+      "learning_rate": 8e-05,
+      "loss": 1.6307,
+      "step": 7553
+    },
+    {
+      "epoch": 0.42107023411371236,
+      "grad_norm": 0.47366848587989807,
+      "learning_rate": 8e-05,
+      "loss": 1.4715,
+      "step": 7554
+    },
+    {
+      "epoch": 0.42112597547380154,
+      "grad_norm": 0.5336489677429199,
+      "learning_rate": 8e-05,
+      "loss": 1.7972,
+      "step": 7555
+    },
+    {
+      "epoch": 0.4211817168338908,
+      "grad_norm": 0.5131495594978333,
+      "learning_rate": 8e-05,
+      "loss": 1.6308,
+      "step": 7556
+    },
+    {
+      "epoch": 0.42123745819397995,
+      "grad_norm": 0.5199673771858215,
+      "learning_rate": 8e-05,
+      "loss": 1.6602,
+      "step": 7557
+    },
+    {
+      "epoch": 0.42129319955406913,
+      "grad_norm": 0.5832521915435791,
+      "learning_rate": 8e-05,
+      "loss": 1.7775,
+      "step": 7558
+    },
+    {
+      "epoch": 0.4213489409141583,
+      "grad_norm": 0.5680006146430969,
+      "learning_rate": 8e-05,
+      "loss": 1.7212,
+      "step": 7559
+    },
+    {
+      "epoch": 0.4214046822742475,
+      "grad_norm": 0.5315585136413574,
+      "learning_rate": 8e-05,
+      "loss": 1.672,
+      "step": 7560
+    },
+    {
+      "epoch": 0.42146042363433667,
+      "grad_norm": 0.5046219229698181,
+      "learning_rate": 8e-05,
+      "loss": 1.5325,
+      "step": 7561
+    },
+    {
+      "epoch": 0.42151616499442585,
+      "grad_norm": 0.5093092322349548,
+      "learning_rate": 8e-05,
+      "loss": 1.4543,
+      "step": 7562
+    },
+    {
+      "epoch": 0.421571906354515,
+      "grad_norm": 0.49265122413635254,
+      "learning_rate": 8e-05,
+      "loss": 1.4613,
+      "step": 7563
+    },
+    {
+      "epoch": 0.42162764771460426,
+      "grad_norm": 0.5940898656845093,
+      "learning_rate": 8e-05,
+      "loss": 1.7415,
+      "step": 7564
+    },
+    {
+      "epoch": 0.42168338907469344,
+      "grad_norm": 0.4995904862880707,
+      "learning_rate": 8e-05,
+      "loss": 1.6841,
+      "step": 7565
+    },
+    {
+      "epoch": 0.4217391304347826,
+      "grad_norm": 0.5423473715782166,
+      "learning_rate": 8e-05,
+      "loss": 1.6731,
+      "step": 7566
+    },
+    {
+      "epoch": 0.4217948717948718,
+      "grad_norm": 0.5287518501281738,
+      "learning_rate": 8e-05,
+      "loss": 1.5821,
+      "step": 7567
+    },
+    {
+      "epoch": 0.421850613154961,
+      "grad_norm": 0.6142107248306274,
+      "learning_rate": 8e-05,
+      "loss": 1.7342,
+      "step": 7568
+    },
+    {
+      "epoch": 0.42190635451505015,
+      "grad_norm": 0.5168833136558533,
+      "learning_rate": 8e-05,
+      "loss": 1.6403,
+      "step": 7569
+    },
+    {
+      "epoch": 0.42196209587513933,
+      "grad_norm": 0.49429258704185486,
+      "learning_rate": 8e-05,
+      "loss": 1.4016,
+      "step": 7570
+    },
+    {
+      "epoch": 0.42201783723522857,
+      "grad_norm": 0.5727221369743347,
+      "learning_rate": 8e-05,
+      "loss": 1.722,
+      "step": 7571
+    },
+    {
+      "epoch": 0.42207357859531774,
+      "grad_norm": 0.5960108041763306,
+      "learning_rate": 8e-05,
+      "loss": 1.8959,
+      "step": 7572
+    },
+    {
+      "epoch": 0.4221293199554069,
+      "grad_norm": 0.5104274153709412,
+      "learning_rate": 8e-05,
+      "loss": 1.6805,
+      "step": 7573
+    },
+    {
+      "epoch": 0.4221850613154961,
+      "grad_norm": 0.5847427248954773,
+      "learning_rate": 8e-05,
+      "loss": 1.7601,
+      "step": 7574
+    },
+    {
+      "epoch": 0.4222408026755853,
+      "grad_norm": 0.6041691899299622,
+      "learning_rate": 8e-05,
+      "loss": 1.6447,
+      "step": 7575
+    },
+    {
+      "epoch": 0.42229654403567446,
+      "grad_norm": 0.5371209383010864,
+      "learning_rate": 8e-05,
+      "loss": 1.6438,
+      "step": 7576
+    },
+    {
+      "epoch": 0.42235228539576364,
+      "grad_norm": 0.540814220905304,
+      "learning_rate": 8e-05,
+      "loss": 1.6554,
+      "step": 7577
+    },
+    {
+      "epoch": 0.4224080267558528,
+      "grad_norm": 0.5552213788032532,
+      "learning_rate": 8e-05,
+      "loss": 1.9482,
+      "step": 7578
+    },
+    {
+      "epoch": 0.42246376811594205,
+      "grad_norm": 0.565314769744873,
+      "learning_rate": 8e-05,
+      "loss": 1.9007,
+      "step": 7579
+    },
+    {
+      "epoch": 0.42251950947603123,
+      "grad_norm": 0.538690447807312,
+      "learning_rate": 8e-05,
+      "loss": 1.6039,
+      "step": 7580
+    },
+    {
+      "epoch": 0.4225752508361204,
+      "grad_norm": 0.5818107724189758,
+      "learning_rate": 8e-05,
+      "loss": 1.29,
+      "step": 7581
+    },
+    {
+      "epoch": 0.4226309921962096,
+      "grad_norm": 0.5533572435379028,
+      "learning_rate": 8e-05,
+      "loss": 1.7616,
+      "step": 7582
+    },
+    {
+      "epoch": 0.42268673355629877,
+      "grad_norm": 0.5512927174568176,
+      "learning_rate": 8e-05,
+      "loss": 1.651,
+      "step": 7583
+    },
+    {
+      "epoch": 0.42274247491638794,
+      "grad_norm": 0.5334979295730591,
+      "learning_rate": 8e-05,
+      "loss": 1.4996,
+      "step": 7584
+    },
+    {
+      "epoch": 0.4227982162764771,
+      "grad_norm": 0.48545053601264954,
+      "learning_rate": 8e-05,
+      "loss": 1.6261,
+      "step": 7585
+    },
+    {
+      "epoch": 0.42285395763656636,
+      "grad_norm": 0.47293198108673096,
+      "learning_rate": 8e-05,
+      "loss": 1.4688,
+      "step": 7586
+    },
+    {
+      "epoch": 0.42290969899665554,
+      "grad_norm": 0.5610591173171997,
+      "learning_rate": 8e-05,
+      "loss": 1.524,
+      "step": 7587
+    },
+    {
+      "epoch": 0.4229654403567447,
+      "grad_norm": 0.5938591957092285,
+      "learning_rate": 8e-05,
+      "loss": 2.0942,
+      "step": 7588
+    },
+    {
+      "epoch": 0.4230211817168339,
+      "grad_norm": 0.49670886993408203,
+      "learning_rate": 8e-05,
+      "loss": 1.7034,
+      "step": 7589
+    },
+    {
+      "epoch": 0.4230769230769231,
+      "grad_norm": 0.5407470464706421,
+      "learning_rate": 8e-05,
+      "loss": 1.7156,
+      "step": 7590
+    },
+    {
+      "epoch": 0.42313266443701225,
+      "grad_norm": 0.5480191111564636,
+      "learning_rate": 8e-05,
+      "loss": 1.4605,
+      "step": 7591
+    },
+    {
+      "epoch": 0.42318840579710143,
+      "grad_norm": 0.4956515431404114,
+      "learning_rate": 8e-05,
+      "loss": 1.5794,
+      "step": 7592
+    },
+    {
+      "epoch": 0.4232441471571906,
+      "grad_norm": 0.6380559802055359,
+      "learning_rate": 8e-05,
+      "loss": 1.9092,
+      "step": 7593
+    },
+    {
+      "epoch": 0.42329988851727984,
+      "grad_norm": 0.5132410526275635,
+      "learning_rate": 8e-05,
+      "loss": 1.5389,
+      "step": 7594
+    },
+    {
+      "epoch": 0.423355629877369,
+      "grad_norm": 0.5750256776809692,
+      "learning_rate": 8e-05,
+      "loss": 1.7148,
+      "step": 7595
+    },
+    {
+      "epoch": 0.4234113712374582,
+      "grad_norm": 0.5305300354957581,
+      "learning_rate": 8e-05,
+      "loss": 1.8729,
+      "step": 7596
+    },
+    {
+      "epoch": 0.4234671125975474,
+      "grad_norm": 0.5310443043708801,
+      "learning_rate": 8e-05,
+      "loss": 1.5722,
+      "step": 7597
+    },
+    {
+      "epoch": 0.42352285395763656,
+      "grad_norm": 0.5453668832778931,
+      "learning_rate": 8e-05,
+      "loss": 1.5986,
+      "step": 7598
+    },
+    {
+      "epoch": 0.42357859531772574,
+      "grad_norm": 0.4966369867324829,
+      "learning_rate": 8e-05,
+      "loss": 1.5603,
+      "step": 7599
+    },
+    {
+      "epoch": 0.4236343366778149,
+      "grad_norm": 0.5009279847145081,
+      "learning_rate": 8e-05,
+      "loss": 1.6332,
+      "step": 7600
+    },
+    {
+      "epoch": 0.42369007803790415,
+      "grad_norm": 0.5626919269561768,
+      "learning_rate": 8e-05,
+      "loss": 1.8475,
+      "step": 7601
+    },
+    {
+      "epoch": 0.42374581939799333,
+      "grad_norm": 0.48185285925865173,
+      "learning_rate": 8e-05,
+      "loss": 1.3049,
+      "step": 7602
+    },
+    {
+      "epoch": 0.4238015607580825,
+      "grad_norm": 0.5500190258026123,
+      "learning_rate": 8e-05,
+      "loss": 1.8254,
+      "step": 7603
+    },
+    {
+      "epoch": 0.4238573021181717,
+      "grad_norm": 0.5302898287773132,
+      "learning_rate": 8e-05,
+      "loss": 1.8387,
+      "step": 7604
+    },
+    {
+      "epoch": 0.42391304347826086,
+      "grad_norm": 0.5317866206169128,
+      "learning_rate": 8e-05,
+      "loss": 1.5846,
+      "step": 7605
+    },
+    {
+      "epoch": 0.42396878483835004,
+      "grad_norm": 0.5135228037834167,
+      "learning_rate": 8e-05,
+      "loss": 1.6452,
+      "step": 7606
+    },
+    {
+      "epoch": 0.4240245261984392,
+      "grad_norm": 0.5610030293464661,
+      "learning_rate": 8e-05,
+      "loss": 1.678,
+      "step": 7607
+    },
+    {
+      "epoch": 0.4240802675585284,
+      "grad_norm": 0.5058049559593201,
+      "learning_rate": 8e-05,
+      "loss": 1.8044,
+      "step": 7608
+    },
+    {
+      "epoch": 0.42413600891861764,
+      "grad_norm": 0.5683199167251587,
+      "learning_rate": 8e-05,
+      "loss": 1.8105,
+      "step": 7609
+    },
+    {
+      "epoch": 0.4241917502787068,
+      "grad_norm": 0.5189591646194458,
+      "learning_rate": 8e-05,
+      "loss": 1.4554,
+      "step": 7610
+    },
+    {
+      "epoch": 0.424247491638796,
+      "grad_norm": 0.5114647746086121,
+      "learning_rate": 8e-05,
+      "loss": 1.6832,
+      "step": 7611
+    },
+    {
+      "epoch": 0.42430323299888517,
+      "grad_norm": 0.5456311702728271,
+      "learning_rate": 8e-05,
+      "loss": 1.7849,
+      "step": 7612
+    },
+    {
+      "epoch": 0.42435897435897435,
+      "grad_norm": 0.5578476190567017,
+      "learning_rate": 8e-05,
+      "loss": 1.8521,
+      "step": 7613
+    },
+    {
+      "epoch": 0.42441471571906353,
+      "grad_norm": 0.5680450201034546,
+      "learning_rate": 8e-05,
+      "loss": 1.8202,
+      "step": 7614
+    },
+    {
+      "epoch": 0.4244704570791527,
+      "grad_norm": 0.5414755344390869,
+      "learning_rate": 8e-05,
+      "loss": 1.7448,
+      "step": 7615
+    },
+    {
+      "epoch": 0.42452619843924194,
+      "grad_norm": 0.5823304653167725,
+      "learning_rate": 8e-05,
+      "loss": 1.8596,
+      "step": 7616
+    },
+    {
+      "epoch": 0.4245819397993311,
+      "grad_norm": 0.5097913146018982,
+      "learning_rate": 8e-05,
+      "loss": 1.4833,
+      "step": 7617
+    },
+    {
+      "epoch": 0.4246376811594203,
+      "grad_norm": 0.5405616164207458,
+      "learning_rate": 8e-05,
+      "loss": 1.764,
+      "step": 7618
+    },
+    {
+      "epoch": 0.4246934225195095,
+      "grad_norm": 0.513064980506897,
+      "learning_rate": 8e-05,
+      "loss": 1.6893,
+      "step": 7619
+    },
+    {
+      "epoch": 0.42474916387959866,
+      "grad_norm": 0.5439816117286682,
+      "learning_rate": 8e-05,
+      "loss": 1.6872,
+      "step": 7620
+    },
+    {
+      "epoch": 0.42480490523968784,
+      "grad_norm": 0.7042144536972046,
+      "learning_rate": 8e-05,
+      "loss": 1.8371,
+      "step": 7621
+    },
+    {
+      "epoch": 0.424860646599777,
+      "grad_norm": 0.4817110598087311,
+      "learning_rate": 8e-05,
+      "loss": 1.6254,
+      "step": 7622
+    },
+    {
+      "epoch": 0.4249163879598662,
+      "grad_norm": 0.5761969685554504,
+      "learning_rate": 8e-05,
+      "loss": 1.7819,
+      "step": 7623
+    },
+    {
+      "epoch": 0.4249721293199554,
+      "grad_norm": 0.6097943186759949,
+      "learning_rate": 8e-05,
+      "loss": 1.8692,
+      "step": 7624
+    },
+    {
+      "epoch": 0.4250278706800446,
+      "grad_norm": 0.5209263563156128,
+      "learning_rate": 8e-05,
+      "loss": 1.6544,
+      "step": 7625
+    },
+    {
+      "epoch": 0.4250836120401338,
+      "grad_norm": 0.5559192299842834,
+      "learning_rate": 8e-05,
+      "loss": 1.7777,
+      "step": 7626
+    },
+    {
+      "epoch": 0.42513935340022296,
+      "grad_norm": 0.5229249000549316,
+      "learning_rate": 8e-05,
+      "loss": 1.749,
+      "step": 7627
+    },
+    {
+      "epoch": 0.42519509476031214,
+      "grad_norm": 0.4869654178619385,
+      "learning_rate": 8e-05,
+      "loss": 1.4507,
+      "step": 7628
+    },
+    {
+      "epoch": 0.4252508361204013,
+      "grad_norm": 0.5297756791114807,
+      "learning_rate": 8e-05,
+      "loss": 1.7302,
+      "step": 7629
+    },
+    {
+      "epoch": 0.4253065774804905,
+      "grad_norm": 0.5154840350151062,
+      "learning_rate": 8e-05,
+      "loss": 1.7402,
+      "step": 7630
+    },
+    {
+      "epoch": 0.42536231884057973,
+      "grad_norm": 0.4986840784549713,
+      "learning_rate": 8e-05,
+      "loss": 1.4875,
+      "step": 7631
+    },
+    {
+      "epoch": 0.4254180602006689,
+      "grad_norm": 0.48561593890190125,
+      "learning_rate": 8e-05,
+      "loss": 1.5017,
+      "step": 7632
+    },
+    {
+      "epoch": 0.4254738015607581,
+      "grad_norm": 0.5005348920822144,
+      "learning_rate": 8e-05,
+      "loss": 1.67,
+      "step": 7633
+    },
+    {
+      "epoch": 0.42552954292084727,
+      "grad_norm": 0.5824643969535828,
+      "learning_rate": 8e-05,
+      "loss": 2.0515,
+      "step": 7634
+    },
+    {
+      "epoch": 0.42558528428093645,
+      "grad_norm": 0.512254536151886,
+      "learning_rate": 8e-05,
+      "loss": 1.8098,
+      "step": 7635
+    },
+    {
+      "epoch": 0.4256410256410256,
+      "grad_norm": 0.5454243421554565,
+      "learning_rate": 8e-05,
+      "loss": 1.9424,
+      "step": 7636
+    },
+    {
+      "epoch": 0.4256967670011148,
+      "grad_norm": 0.4775199890136719,
+      "learning_rate": 8e-05,
+      "loss": 1.3979,
+      "step": 7637
+    },
+    {
+      "epoch": 0.42575250836120404,
+      "grad_norm": 0.4912828207015991,
+      "learning_rate": 8e-05,
+      "loss": 1.4566,
+      "step": 7638
+    },
+    {
+      "epoch": 0.4258082497212932,
+      "grad_norm": 0.5596845149993896,
+      "learning_rate": 8e-05,
+      "loss": 1.7422,
+      "step": 7639
+    },
+    {
+      "epoch": 0.4258639910813824,
+      "grad_norm": 0.5112817287445068,
+      "learning_rate": 8e-05,
+      "loss": 1.5523,
+      "step": 7640
+    },
+    {
+      "epoch": 0.4259197324414716,
+      "grad_norm": 0.5548750758171082,
+      "learning_rate": 8e-05,
+      "loss": 1.7622,
+      "step": 7641
+    },
+    {
+      "epoch": 0.42597547380156076,
+      "grad_norm": 0.5437869429588318,
+      "learning_rate": 8e-05,
+      "loss": 1.6663,
+      "step": 7642
+    },
+    {
+      "epoch": 0.42603121516164993,
+      "grad_norm": 0.5219390392303467,
+      "learning_rate": 8e-05,
+      "loss": 1.6779,
+      "step": 7643
+    },
+    {
+      "epoch": 0.4260869565217391,
+      "grad_norm": 0.5230270624160767,
+      "learning_rate": 8e-05,
+      "loss": 1.6244,
+      "step": 7644
+    },
+    {
+      "epoch": 0.4261426978818283,
+      "grad_norm": 0.5277723670005798,
+      "learning_rate": 8e-05,
+      "loss": 1.5704,
+      "step": 7645
+    },
+    {
+      "epoch": 0.4261984392419175,
+      "grad_norm": 0.5310590863227844,
+      "learning_rate": 8e-05,
+      "loss": 1.7704,
+      "step": 7646
+    },
+    {
+      "epoch": 0.4262541806020067,
+      "grad_norm": 0.5522536039352417,
+      "learning_rate": 8e-05,
+      "loss": 1.6834,
+      "step": 7647
+    },
+    {
+      "epoch": 0.4263099219620959,
+      "grad_norm": 0.5156833529472351,
+      "learning_rate": 8e-05,
+      "loss": 1.5644,
+      "step": 7648
+    },
+    {
+      "epoch": 0.42636566332218506,
+      "grad_norm": 0.5626975893974304,
+      "learning_rate": 8e-05,
+      "loss": 1.7787,
+      "step": 7649
+    },
+    {
+      "epoch": 0.42642140468227424,
+      "grad_norm": 0.5107052326202393,
+      "learning_rate": 8e-05,
+      "loss": 1.5207,
+      "step": 7650
+    },
+    {
+      "epoch": 0.4264771460423634,
+      "grad_norm": 0.5529700517654419,
+      "learning_rate": 8e-05,
+      "loss": 1.9028,
+      "step": 7651
+    },
+    {
+      "epoch": 0.4265328874024526,
+      "grad_norm": 0.5312693119049072,
+      "learning_rate": 8e-05,
+      "loss": 1.7605,
+      "step": 7652
+    },
+    {
+      "epoch": 0.42658862876254183,
+      "grad_norm": 0.5670287013053894,
+      "learning_rate": 8e-05,
+      "loss": 1.8526,
+      "step": 7653
+    },
+    {
+      "epoch": 0.426644370122631,
+      "grad_norm": 0.5672020316123962,
+      "learning_rate": 8e-05,
+      "loss": 1.4338,
+      "step": 7654
+    },
+    {
+      "epoch": 0.4267001114827202,
+      "grad_norm": 0.5201690793037415,
+      "learning_rate": 8e-05,
+      "loss": 1.5336,
+      "step": 7655
+    },
+    {
+      "epoch": 0.42675585284280937,
+      "grad_norm": 0.4957461953163147,
+      "learning_rate": 8e-05,
+      "loss": 1.5938,
+      "step": 7656
+    },
+    {
+      "epoch": 0.42681159420289855,
+      "grad_norm": 0.5092799663543701,
+      "learning_rate": 8e-05,
+      "loss": 1.5157,
+      "step": 7657
+    },
+    {
+      "epoch": 0.4268673355629877,
+      "grad_norm": 0.5307939052581787,
+      "learning_rate": 8e-05,
+      "loss": 1.9267,
+      "step": 7658
+    },
+    {
+      "epoch": 0.4269230769230769,
+      "grad_norm": 0.5566995143890381,
+      "learning_rate": 8e-05,
+      "loss": 1.7914,
+      "step": 7659
+    },
+    {
+      "epoch": 0.4269788182831661,
+      "grad_norm": 0.5886173248291016,
+      "learning_rate": 8e-05,
+      "loss": 2.1544,
+      "step": 7660
+    },
+    {
+      "epoch": 0.4270345596432553,
+      "grad_norm": 0.547525942325592,
+      "learning_rate": 8e-05,
+      "loss": 1.7621,
+      "step": 7661
+    },
+    {
+      "epoch": 0.4270903010033445,
+      "grad_norm": 0.5709888935089111,
+      "learning_rate": 8e-05,
+      "loss": 1.7429,
+      "step": 7662
+    },
+    {
+      "epoch": 0.4271460423634337,
+      "grad_norm": 0.5912901759147644,
+      "learning_rate": 8e-05,
+      "loss": 1.6187,
+      "step": 7663
+    },
+    {
+      "epoch": 0.42720178372352285,
+      "grad_norm": 0.5148410797119141,
+      "learning_rate": 8e-05,
+      "loss": 1.7212,
+      "step": 7664
+    },
+    {
+      "epoch": 0.42725752508361203,
+      "grad_norm": 0.5268368124961853,
+      "learning_rate": 8e-05,
+      "loss": 1.5642,
+      "step": 7665
+    },
+    {
+      "epoch": 0.4273132664437012,
+      "grad_norm": 0.5531495213508606,
+      "learning_rate": 8e-05,
+      "loss": 1.7811,
+      "step": 7666
+    },
+    {
+      "epoch": 0.4273690078037904,
+      "grad_norm": 0.5273078680038452,
+      "learning_rate": 8e-05,
+      "loss": 1.7557,
+      "step": 7667
+    },
+    {
+      "epoch": 0.4274247491638796,
+      "grad_norm": 0.6302218437194824,
+      "learning_rate": 8e-05,
+      "loss": 1.6079,
+      "step": 7668
+    },
+    {
+      "epoch": 0.4274804905239688,
+      "grad_norm": 0.5131771564483643,
+      "learning_rate": 8e-05,
+      "loss": 1.6148,
+      "step": 7669
+    },
+    {
+      "epoch": 0.427536231884058,
+      "grad_norm": 0.4759463965892792,
+      "learning_rate": 8e-05,
+      "loss": 1.4748,
+      "step": 7670
+    },
+    {
+      "epoch": 0.42759197324414716,
+      "grad_norm": 0.6624416708946228,
+      "learning_rate": 8e-05,
+      "loss": 1.8419,
+      "step": 7671
+    },
+    {
+      "epoch": 0.42764771460423634,
+      "grad_norm": 0.5604111552238464,
+      "learning_rate": 8e-05,
+      "loss": 1.819,
+      "step": 7672
+    },
+    {
+      "epoch": 0.4277034559643255,
+      "grad_norm": 0.5581426620483398,
+      "learning_rate": 8e-05,
+      "loss": 1.585,
+      "step": 7673
+    },
+    {
+      "epoch": 0.4277591973244147,
+      "grad_norm": 0.5654292106628418,
+      "learning_rate": 8e-05,
+      "loss": 1.8578,
+      "step": 7674
+    },
+    {
+      "epoch": 0.4278149386845039,
+      "grad_norm": 0.5510108470916748,
+      "learning_rate": 8e-05,
+      "loss": 1.8643,
+      "step": 7675
+    },
+    {
+      "epoch": 0.4278706800445931,
+      "grad_norm": 0.5483967661857605,
+      "learning_rate": 8e-05,
+      "loss": 1.7746,
+      "step": 7676
+    },
+    {
+      "epoch": 0.4279264214046823,
+      "grad_norm": 0.5236601829528809,
+      "learning_rate": 8e-05,
+      "loss": 1.6946,
+      "step": 7677
+    },
+    {
+      "epoch": 0.42798216276477147,
+      "grad_norm": 0.5318403244018555,
+      "learning_rate": 8e-05,
+      "loss": 1.5735,
+      "step": 7678
+    },
+    {
+      "epoch": 0.42803790412486065,
+      "grad_norm": 0.5390852093696594,
+      "learning_rate": 8e-05,
+      "loss": 1.7182,
+      "step": 7679
+    },
+    {
+      "epoch": 0.4280936454849498,
+      "grad_norm": 0.6264533400535583,
+      "learning_rate": 8e-05,
+      "loss": 1.844,
+      "step": 7680
+    },
+    {
+      "epoch": 0.428149386845039,
+      "grad_norm": 0.5331520438194275,
+      "learning_rate": 8e-05,
+      "loss": 1.5223,
+      "step": 7681
+    },
+    {
+      "epoch": 0.4282051282051282,
+      "grad_norm": 0.5553402304649353,
+      "learning_rate": 8e-05,
+      "loss": 1.728,
+      "step": 7682
+    },
+    {
+      "epoch": 0.4282608695652174,
+      "grad_norm": 0.5179986357688904,
+      "learning_rate": 8e-05,
+      "loss": 1.4644,
+      "step": 7683
+    },
+    {
+      "epoch": 0.4283166109253066,
+      "grad_norm": 0.5654995441436768,
+      "learning_rate": 8e-05,
+      "loss": 1.5609,
+      "step": 7684
+    },
+    {
+      "epoch": 0.4283723522853958,
+      "grad_norm": 0.5537652373313904,
+      "learning_rate": 8e-05,
+      "loss": 1.5997,
+      "step": 7685
+    },
+    {
+      "epoch": 0.42842809364548495,
+      "grad_norm": 0.5693234801292419,
+      "learning_rate": 8e-05,
+      "loss": 1.7669,
+      "step": 7686
+    },
+    {
+      "epoch": 0.42848383500557413,
+      "grad_norm": 0.5033535361289978,
+      "learning_rate": 8e-05,
+      "loss": 1.662,
+      "step": 7687
+    },
+    {
+      "epoch": 0.4285395763656633,
+      "grad_norm": 0.5271948575973511,
+      "learning_rate": 8e-05,
+      "loss": 1.5819,
+      "step": 7688
+    },
+    {
+      "epoch": 0.4285953177257525,
+      "grad_norm": 0.5141314268112183,
+      "learning_rate": 8e-05,
+      "loss": 1.6355,
+      "step": 7689
+    },
+    {
+      "epoch": 0.42865105908584167,
+      "grad_norm": 0.5537899732589722,
+      "learning_rate": 8e-05,
+      "loss": 1.7471,
+      "step": 7690
+    },
+    {
+      "epoch": 0.4287068004459309,
+      "grad_norm": 0.5063492059707642,
+      "learning_rate": 8e-05,
+      "loss": 1.3736,
+      "step": 7691
+    },
+    {
+      "epoch": 0.4287625418060201,
+      "grad_norm": 0.47556009888648987,
+      "learning_rate": 8e-05,
+      "loss": 1.5181,
+      "step": 7692
+    },
+    {
+      "epoch": 0.42881828316610926,
+      "grad_norm": 0.5070196986198425,
+      "learning_rate": 8e-05,
+      "loss": 1.7123,
+      "step": 7693
+    },
+    {
+      "epoch": 0.42887402452619844,
+      "grad_norm": 0.5480073094367981,
+      "learning_rate": 8e-05,
+      "loss": 1.6972,
+      "step": 7694
+    },
+    {
+      "epoch": 0.4289297658862876,
+      "grad_norm": 0.5210518836975098,
+      "learning_rate": 8e-05,
+      "loss": 1.7618,
+      "step": 7695
+    },
+    {
+      "epoch": 0.4289855072463768,
+      "grad_norm": 0.5803531408309937,
+      "learning_rate": 8e-05,
+      "loss": 1.8326,
+      "step": 7696
+    },
+    {
+      "epoch": 0.429041248606466,
+      "grad_norm": 0.4972003400325775,
+      "learning_rate": 8e-05,
+      "loss": 1.5183,
+      "step": 7697
+    },
+    {
+      "epoch": 0.4290969899665552,
+      "grad_norm": 0.5382688045501709,
+      "learning_rate": 8e-05,
+      "loss": 1.8117,
+      "step": 7698
+    },
+    {
+      "epoch": 0.4291527313266444,
+      "grad_norm": 0.47422686219215393,
+      "learning_rate": 8e-05,
+      "loss": 1.6818,
+      "step": 7699
+    },
+    {
+      "epoch": 0.42920847268673357,
+      "grad_norm": 0.5724363923072815,
+      "learning_rate": 8e-05,
+      "loss": 1.8173,
+      "step": 7700
+    },
+    {
+      "epoch": 0.42926421404682275,
+      "grad_norm": 0.5626831650733948,
+      "learning_rate": 8e-05,
+      "loss": 1.6295,
+      "step": 7701
+    },
+    {
+      "epoch": 0.4293199554069119,
+      "grad_norm": 0.5423737168312073,
+      "learning_rate": 8e-05,
+      "loss": 1.7521,
+      "step": 7702
+    },
+    {
+      "epoch": 0.4293756967670011,
+      "grad_norm": 0.5201063752174377,
+      "learning_rate": 8e-05,
+      "loss": 1.6914,
+      "step": 7703
+    },
+    {
+      "epoch": 0.4294314381270903,
+      "grad_norm": 0.5397342443466187,
+      "learning_rate": 8e-05,
+      "loss": 1.6416,
+      "step": 7704
+    },
+    {
+      "epoch": 0.42948717948717946,
+      "grad_norm": 0.54704749584198,
+      "learning_rate": 8e-05,
+      "loss": 1.6658,
+      "step": 7705
+    },
+    {
+      "epoch": 0.4295429208472687,
+      "grad_norm": 0.5956423878669739,
+      "learning_rate": 8e-05,
+      "loss": 1.7879,
+      "step": 7706
+    },
+    {
+      "epoch": 0.4295986622073579,
+      "grad_norm": 0.5553403496742249,
+      "learning_rate": 8e-05,
+      "loss": 1.7613,
+      "step": 7707
+    },
+    {
+      "epoch": 0.42965440356744705,
+      "grad_norm": 0.5432010293006897,
+      "learning_rate": 8e-05,
+      "loss": 1.6631,
+      "step": 7708
+    },
+    {
+      "epoch": 0.42971014492753623,
+      "grad_norm": 0.5349817872047424,
+      "learning_rate": 8e-05,
+      "loss": 1.7419,
+      "step": 7709
+    },
+    {
+      "epoch": 0.4297658862876254,
+      "grad_norm": 0.5355441570281982,
+      "learning_rate": 8e-05,
+      "loss": 1.676,
+      "step": 7710
+    },
+    {
+      "epoch": 0.4298216276477146,
+      "grad_norm": 0.5206706523895264,
+      "learning_rate": 8e-05,
+      "loss": 1.499,
+      "step": 7711
+    },
+    {
+      "epoch": 0.42987736900780377,
+      "grad_norm": 0.5900402069091797,
+      "learning_rate": 8e-05,
+      "loss": 1.9597,
+      "step": 7712
+    },
+    {
+      "epoch": 0.429933110367893,
+      "grad_norm": 0.5922344326972961,
+      "learning_rate": 8e-05,
+      "loss": 1.969,
+      "step": 7713
+    },
+    {
+      "epoch": 0.4299888517279822,
+      "grad_norm": 0.5381130576133728,
+      "learning_rate": 8e-05,
+      "loss": 1.5954,
+      "step": 7714
+    },
+    {
+      "epoch": 0.43004459308807136,
+      "grad_norm": 0.5215221047401428,
+      "learning_rate": 8e-05,
+      "loss": 1.513,
+      "step": 7715
+    },
+    {
+      "epoch": 0.43010033444816054,
+      "grad_norm": 0.6906350255012512,
+      "learning_rate": 8e-05,
+      "loss": 1.7071,
+      "step": 7716
+    },
+    {
+      "epoch": 0.4301560758082497,
+      "grad_norm": 0.49984869360923767,
+      "learning_rate": 8e-05,
+      "loss": 1.5308,
+      "step": 7717
+    },
+    {
+      "epoch": 0.4302118171683389,
+      "grad_norm": 0.5267676711082458,
+      "learning_rate": 8e-05,
+      "loss": 1.6652,
+      "step": 7718
+    },
+    {
+      "epoch": 0.4302675585284281,
+      "grad_norm": 0.533749520778656,
+      "learning_rate": 8e-05,
+      "loss": 1.6965,
+      "step": 7719
+    },
+    {
+      "epoch": 0.43032329988851725,
+      "grad_norm": 0.5423862338066101,
+      "learning_rate": 8e-05,
+      "loss": 1.5306,
+      "step": 7720
+    },
+    {
+      "epoch": 0.4303790412486065,
+      "grad_norm": 0.525157630443573,
+      "learning_rate": 8e-05,
+      "loss": 1.5907,
+      "step": 7721
+    },
+    {
+      "epoch": 0.43043478260869567,
+      "grad_norm": 0.5443845391273499,
+      "learning_rate": 8e-05,
+      "loss": 1.6161,
+      "step": 7722
+    },
+    {
+      "epoch": 0.43049052396878484,
+      "grad_norm": 0.5968034267425537,
+      "learning_rate": 8e-05,
+      "loss": 1.7725,
+      "step": 7723
+    },
+    {
+      "epoch": 0.430546265328874,
+      "grad_norm": 0.5567757487297058,
+      "learning_rate": 8e-05,
+      "loss": 1.7576,
+      "step": 7724
+    },
+    {
+      "epoch": 0.4306020066889632,
+      "grad_norm": 0.5223360657691956,
+      "learning_rate": 8e-05,
+      "loss": 1.7543,
+      "step": 7725
+    },
+    {
+      "epoch": 0.4306577480490524,
+      "grad_norm": 0.5313509702682495,
+      "learning_rate": 8e-05,
+      "loss": 1.6799,
+      "step": 7726
+    },
+    {
+      "epoch": 0.43071348940914156,
+      "grad_norm": 0.5988759994506836,
+      "learning_rate": 8e-05,
+      "loss": 1.9165,
+      "step": 7727
+    },
+    {
+      "epoch": 0.4307692307692308,
+      "grad_norm": 0.5237802267074585,
+      "learning_rate": 8e-05,
+      "loss": 1.555,
+      "step": 7728
+    },
+    {
+      "epoch": 0.43082497212931997,
+      "grad_norm": 0.5154265761375427,
+      "learning_rate": 8e-05,
+      "loss": 1.7844,
+      "step": 7729
+    },
+    {
+      "epoch": 0.43088071348940915,
+      "grad_norm": 0.510456919670105,
+      "learning_rate": 8e-05,
+      "loss": 1.7146,
+      "step": 7730
+    },
+    {
+      "epoch": 0.43093645484949833,
+      "grad_norm": 0.47713032364845276,
+      "learning_rate": 8e-05,
+      "loss": 1.4793,
+      "step": 7731
+    },
+    {
+      "epoch": 0.4309921962095875,
+      "grad_norm": 0.5002168416976929,
+      "learning_rate": 8e-05,
+      "loss": 1.7028,
+      "step": 7732
+    },
+    {
+      "epoch": 0.4310479375696767,
+      "grad_norm": 0.5037677884101868,
+      "learning_rate": 8e-05,
+      "loss": 1.7953,
+      "step": 7733
+    },
+    {
+      "epoch": 0.43110367892976587,
+      "grad_norm": 0.5853288173675537,
+      "learning_rate": 8e-05,
+      "loss": 1.8959,
+      "step": 7734
+    },
+    {
+      "epoch": 0.4311594202898551,
+      "grad_norm": 0.5197460651397705,
+      "learning_rate": 8e-05,
+      "loss": 1.6534,
+      "step": 7735
+    },
+    {
+      "epoch": 0.4312151616499443,
+      "grad_norm": 0.5077546834945679,
+      "learning_rate": 8e-05,
+      "loss": 1.6536,
+      "step": 7736
+    },
+    {
+      "epoch": 0.43127090301003346,
+      "grad_norm": 0.5528148412704468,
+      "learning_rate": 8e-05,
+      "loss": 1.8002,
+      "step": 7737
+    },
+    {
+      "epoch": 0.43132664437012264,
+      "grad_norm": 0.5611364245414734,
+      "learning_rate": 8e-05,
+      "loss": 1.6509,
+      "step": 7738
+    },
+    {
+      "epoch": 0.4313823857302118,
+      "grad_norm": 0.5474463701248169,
+      "learning_rate": 8e-05,
+      "loss": 1.8108,
+      "step": 7739
+    },
+    {
+      "epoch": 0.431438127090301,
+      "grad_norm": 0.568392276763916,
+      "learning_rate": 8e-05,
+      "loss": 1.7376,
+      "step": 7740
+    },
+    {
+      "epoch": 0.43149386845039017,
+      "grad_norm": 0.5201650261878967,
+      "learning_rate": 8e-05,
+      "loss": 1.5288,
+      "step": 7741
+    },
+    {
+      "epoch": 0.43154960981047935,
+      "grad_norm": 0.5635396242141724,
+      "learning_rate": 8e-05,
+      "loss": 1.5296,
+      "step": 7742
+    },
+    {
+      "epoch": 0.4316053511705686,
+      "grad_norm": 0.5454962849617004,
+      "learning_rate": 8e-05,
+      "loss": 1.7872,
+      "step": 7743
+    },
+    {
+      "epoch": 0.43166109253065776,
+      "grad_norm": 0.5474128127098083,
+      "learning_rate": 8e-05,
+      "loss": 1.7107,
+      "step": 7744
+    },
+    {
+      "epoch": 0.43171683389074694,
+      "grad_norm": 0.5751205682754517,
+      "learning_rate": 8e-05,
+      "loss": 2.0157,
+      "step": 7745
+    },
+    {
+      "epoch": 0.4317725752508361,
+      "grad_norm": 0.5645526647567749,
+      "learning_rate": 8e-05,
+      "loss": 1.5263,
+      "step": 7746
+    },
+    {
+      "epoch": 0.4318283166109253,
+      "grad_norm": 0.5280218720436096,
+      "learning_rate": 8e-05,
+      "loss": 1.6702,
+      "step": 7747
+    },
+    {
+      "epoch": 0.4318840579710145,
+      "grad_norm": 0.5345996618270874,
+      "learning_rate": 8e-05,
+      "loss": 1.7274,
+      "step": 7748
+    },
+    {
+      "epoch": 0.43193979933110366,
+      "grad_norm": 0.500987708568573,
+      "learning_rate": 8e-05,
+      "loss": 1.5633,
+      "step": 7749
+    },
+    {
+      "epoch": 0.4319955406911929,
+      "grad_norm": 0.5084331631660461,
+      "learning_rate": 8e-05,
+      "loss": 1.493,
+      "step": 7750
+    },
+    {
+      "epoch": 0.43205128205128207,
+      "grad_norm": 0.5245458483695984,
+      "learning_rate": 8e-05,
+      "loss": 1.7815,
+      "step": 7751
+    },
+    {
+      "epoch": 0.43210702341137125,
+      "grad_norm": 0.5459529757499695,
+      "learning_rate": 8e-05,
+      "loss": 1.6608,
+      "step": 7752
+    },
+    {
+      "epoch": 0.43216276477146043,
+      "grad_norm": 0.505776584148407,
+      "learning_rate": 8e-05,
+      "loss": 1.5744,
+      "step": 7753
+    },
+    {
+      "epoch": 0.4322185061315496,
+      "grad_norm": 0.5133358240127563,
+      "learning_rate": 8e-05,
+      "loss": 1.7084,
+      "step": 7754
+    },
+    {
+      "epoch": 0.4322742474916388,
+      "grad_norm": 0.4842239320278168,
+      "learning_rate": 8e-05,
+      "loss": 1.5998,
+      "step": 7755
+    },
+    {
+      "epoch": 0.43232998885172796,
+      "grad_norm": 0.49821797013282776,
+      "learning_rate": 8e-05,
+      "loss": 1.5764,
+      "step": 7756
+    },
+    {
+      "epoch": 0.43238573021181714,
+      "grad_norm": 0.5373196005821228,
+      "learning_rate": 8e-05,
+      "loss": 1.536,
+      "step": 7757
+    },
+    {
+      "epoch": 0.4324414715719064,
+      "grad_norm": 0.5395300388336182,
+      "learning_rate": 8e-05,
+      "loss": 1.6068,
+      "step": 7758
+    },
+    {
+      "epoch": 0.43249721293199556,
+      "grad_norm": 0.5366533994674683,
+      "learning_rate": 8e-05,
+      "loss": 1.6368,
+      "step": 7759
+    },
+    {
+      "epoch": 0.43255295429208473,
+      "grad_norm": 0.5885009169578552,
+      "learning_rate": 8e-05,
+      "loss": 1.7807,
+      "step": 7760
+    },
+    {
+      "epoch": 0.4326086956521739,
+      "grad_norm": 0.5301985740661621,
+      "learning_rate": 8e-05,
+      "loss": 1.7191,
+      "step": 7761
+    },
+    {
+      "epoch": 0.4326644370122631,
+      "grad_norm": 0.5283817648887634,
+      "learning_rate": 8e-05,
+      "loss": 1.6551,
+      "step": 7762
+    },
+    {
+      "epoch": 0.43272017837235227,
+      "grad_norm": 0.5542014837265015,
+      "learning_rate": 8e-05,
+      "loss": 1.7364,
+      "step": 7763
+    },
+    {
+      "epoch": 0.43277591973244145,
+      "grad_norm": 0.5553522706031799,
+      "learning_rate": 8e-05,
+      "loss": 1.5643,
+      "step": 7764
+    },
+    {
+      "epoch": 0.4328316610925307,
+      "grad_norm": 0.5287862420082092,
+      "learning_rate": 8e-05,
+      "loss": 1.763,
+      "step": 7765
+    },
+    {
+      "epoch": 0.43288740245261986,
+      "grad_norm": 0.5084366798400879,
+      "learning_rate": 8e-05,
+      "loss": 1.6354,
+      "step": 7766
+    },
+    {
+      "epoch": 0.43294314381270904,
+      "grad_norm": 0.5567695498466492,
+      "learning_rate": 8e-05,
+      "loss": 1.7928,
+      "step": 7767
+    },
+    {
+      "epoch": 0.4329988851727982,
+      "grad_norm": 0.5375629663467407,
+      "learning_rate": 8e-05,
+      "loss": 1.6925,
+      "step": 7768
+    },
+    {
+      "epoch": 0.4330546265328874,
+      "grad_norm": 0.5264116525650024,
+      "learning_rate": 8e-05,
+      "loss": 1.775,
+      "step": 7769
+    },
+    {
+      "epoch": 0.4331103678929766,
+      "grad_norm": 0.5006895661354065,
+      "learning_rate": 8e-05,
+      "loss": 1.626,
+      "step": 7770
+    },
+    {
+      "epoch": 0.43316610925306576,
+      "grad_norm": 0.5429582595825195,
+      "learning_rate": 8e-05,
+      "loss": 1.9145,
+      "step": 7771
+    },
+    {
+      "epoch": 0.43322185061315494,
+      "grad_norm": 0.46108904480934143,
+      "learning_rate": 8e-05,
+      "loss": 1.2811,
+      "step": 7772
+    },
+    {
+      "epoch": 0.43327759197324417,
+      "grad_norm": 0.5806688070297241,
+      "learning_rate": 8e-05,
+      "loss": 1.8541,
+      "step": 7773
+    },
+    {
+      "epoch": 0.43333333333333335,
+      "grad_norm": 0.5763636231422424,
+      "learning_rate": 8e-05,
+      "loss": 1.8307,
+      "step": 7774
+    },
+    {
+      "epoch": 0.4333890746934225,
+      "grad_norm": 0.5186631083488464,
+      "learning_rate": 8e-05,
+      "loss": 1.6144,
+      "step": 7775
+    },
+    {
+      "epoch": 0.4334448160535117,
+      "grad_norm": 0.5161386132240295,
+      "learning_rate": 8e-05,
+      "loss": 1.7043,
+      "step": 7776
+    },
+    {
+      "epoch": 0.4335005574136009,
+      "grad_norm": 0.5380999445915222,
+      "learning_rate": 8e-05,
+      "loss": 1.6233,
+      "step": 7777
+    },
+    {
+      "epoch": 0.43355629877369006,
+      "grad_norm": 0.5027433037757874,
+      "learning_rate": 8e-05,
+      "loss": 1.5026,
+      "step": 7778
+    },
+    {
+      "epoch": 0.43361204013377924,
+      "grad_norm": 0.4948747158050537,
+      "learning_rate": 8e-05,
+      "loss": 1.5456,
+      "step": 7779
+    },
+    {
+      "epoch": 0.4336677814938685,
+      "grad_norm": 0.5553415417671204,
+      "learning_rate": 8e-05,
+      "loss": 1.6105,
+      "step": 7780
+    },
+    {
+      "epoch": 0.43372352285395765,
+      "grad_norm": 0.5105648040771484,
+      "learning_rate": 8e-05,
+      "loss": 1.5264,
+      "step": 7781
+    },
+    {
+      "epoch": 0.43377926421404683,
+      "grad_norm": 0.5923818349838257,
+      "learning_rate": 8e-05,
+      "loss": 1.672,
+      "step": 7782
+    },
+    {
+      "epoch": 0.433835005574136,
+      "grad_norm": 0.5075348019599915,
+      "learning_rate": 8e-05,
+      "loss": 1.4665,
+      "step": 7783
+    },
+    {
+      "epoch": 0.4338907469342252,
+      "grad_norm": 0.5268019437789917,
+      "learning_rate": 8e-05,
+      "loss": 1.7234,
+      "step": 7784
+    },
+    {
+      "epoch": 0.43394648829431437,
+      "grad_norm": 0.48290082812309265,
+      "learning_rate": 8e-05,
+      "loss": 1.3133,
+      "step": 7785
+    },
+    {
+      "epoch": 0.43400222965440355,
+      "grad_norm": 0.5417959690093994,
+      "learning_rate": 8e-05,
+      "loss": 1.6926,
+      "step": 7786
+    },
+    {
+      "epoch": 0.4340579710144927,
+      "grad_norm": 0.7672746181488037,
+      "learning_rate": 8e-05,
+      "loss": 1.8159,
+      "step": 7787
+    },
+    {
+      "epoch": 0.43411371237458196,
+      "grad_norm": 0.5545976758003235,
+      "learning_rate": 8e-05,
+      "loss": 1.9052,
+      "step": 7788
+    },
+    {
+      "epoch": 0.43416945373467114,
+      "grad_norm": 0.5124754905700684,
+      "learning_rate": 8e-05,
+      "loss": 1.6812,
+      "step": 7789
+    },
+    {
+      "epoch": 0.4342251950947603,
+      "grad_norm": 0.5179421901702881,
+      "learning_rate": 8e-05,
+      "loss": 1.7352,
+      "step": 7790
+    },
+    {
+      "epoch": 0.4342809364548495,
+      "grad_norm": 0.5736626386642456,
+      "learning_rate": 8e-05,
+      "loss": 1.5641,
+      "step": 7791
+    },
+    {
+      "epoch": 0.4343366778149387,
+      "grad_norm": 0.5768195986747742,
+      "learning_rate": 8e-05,
+      "loss": 1.7259,
+      "step": 7792
+    },
+    {
+      "epoch": 0.43439241917502786,
+      "grad_norm": 0.5163231492042542,
+      "learning_rate": 8e-05,
+      "loss": 1.7629,
+      "step": 7793
+    },
+    {
+      "epoch": 0.43444816053511703,
+      "grad_norm": 0.5233259797096252,
+      "learning_rate": 8e-05,
+      "loss": 1.5611,
+      "step": 7794
+    },
+    {
+      "epoch": 0.43450390189520627,
+      "grad_norm": 0.4899364411830902,
+      "learning_rate": 8e-05,
+      "loss": 1.2202,
+      "step": 7795
+    },
+    {
+      "epoch": 0.43455964325529545,
+      "grad_norm": 0.5041932463645935,
+      "learning_rate": 8e-05,
+      "loss": 1.6126,
+      "step": 7796
+    },
+    {
+      "epoch": 0.4346153846153846,
+      "grad_norm": 0.5223237872123718,
+      "learning_rate": 8e-05,
+      "loss": 1.7619,
+      "step": 7797
+    },
+    {
+      "epoch": 0.4346711259754738,
+      "grad_norm": 0.5277586579322815,
+      "learning_rate": 8e-05,
+      "loss": 1.6513,
+      "step": 7798
+    },
+    {
+      "epoch": 0.434726867335563,
+      "grad_norm": 0.5886757373809814,
+      "learning_rate": 8e-05,
+      "loss": 1.9693,
+      "step": 7799
+    },
+    {
+      "epoch": 0.43478260869565216,
+      "grad_norm": 0.503447413444519,
+      "learning_rate": 8e-05,
+      "loss": 1.6416,
+      "step": 7800
+    },
+    {
+      "epoch": 0.43483835005574134,
+      "grad_norm": 0.5232421159744263,
+      "learning_rate": 8e-05,
+      "loss": 1.5618,
+      "step": 7801
+    },
+    {
+      "epoch": 0.4348940914158305,
+      "grad_norm": 0.5342657566070557,
+      "learning_rate": 8e-05,
+      "loss": 1.7106,
+      "step": 7802
+    },
+    {
+      "epoch": 0.43494983277591975,
+      "grad_norm": 0.5697629451751709,
+      "learning_rate": 8e-05,
+      "loss": 1.6205,
+      "step": 7803
+    },
+    {
+      "epoch": 0.43500557413600893,
+      "grad_norm": 0.54024338722229,
+      "learning_rate": 8e-05,
+      "loss": 1.5715,
+      "step": 7804
+    },
+    {
+      "epoch": 0.4350613154960981,
+      "grad_norm": 0.4912549555301666,
+      "learning_rate": 8e-05,
+      "loss": 1.4373,
+      "step": 7805
+    },
+    {
+      "epoch": 0.4351170568561873,
+      "grad_norm": 0.5158436894416809,
+      "learning_rate": 8e-05,
+      "loss": 1.7721,
+      "step": 7806
+    },
+    {
+      "epoch": 0.43517279821627647,
+      "grad_norm": 0.5614274740219116,
+      "learning_rate": 8e-05,
+      "loss": 1.7082,
+      "step": 7807
+    },
+    {
+      "epoch": 0.43522853957636565,
+      "grad_norm": 0.5338239669799805,
+      "learning_rate": 8e-05,
+      "loss": 1.679,
+      "step": 7808
+    },
+    {
+      "epoch": 0.4352842809364548,
+      "grad_norm": 0.5141422748565674,
+      "learning_rate": 8e-05,
+      "loss": 1.481,
+      "step": 7809
+    },
+    {
+      "epoch": 0.43534002229654406,
+      "grad_norm": 0.5167403221130371,
+      "learning_rate": 8e-05,
+      "loss": 1.5169,
+      "step": 7810
+    },
+    {
+      "epoch": 0.43539576365663324,
+      "grad_norm": 0.5393883585929871,
+      "learning_rate": 8e-05,
+      "loss": 1.7736,
+      "step": 7811
+    },
+    {
+      "epoch": 0.4354515050167224,
+      "grad_norm": 0.5399210453033447,
+      "learning_rate": 8e-05,
+      "loss": 1.8407,
+      "step": 7812
+    },
+    {
+      "epoch": 0.4355072463768116,
+      "grad_norm": 0.5002278685569763,
+      "learning_rate": 8e-05,
+      "loss": 1.6819,
+      "step": 7813
+    },
+    {
+      "epoch": 0.4355629877369008,
+      "grad_norm": 0.5279465317726135,
+      "learning_rate": 8e-05,
+      "loss": 1.6198,
+      "step": 7814
+    },
+    {
+      "epoch": 0.43561872909698995,
+      "grad_norm": 0.5590936541557312,
+      "learning_rate": 8e-05,
+      "loss": 1.7535,
+      "step": 7815
+    },
+    {
+      "epoch": 0.43567447045707913,
+      "grad_norm": 0.5703123211860657,
+      "learning_rate": 8e-05,
+      "loss": 1.8356,
+      "step": 7816
+    },
+    {
+      "epoch": 0.4357302118171683,
+      "grad_norm": 0.5183179378509521,
+      "learning_rate": 8e-05,
+      "loss": 1.6322,
+      "step": 7817
+    },
+    {
+      "epoch": 0.43578595317725755,
+      "grad_norm": 0.5476742386817932,
+      "learning_rate": 8e-05,
+      "loss": 1.6851,
+      "step": 7818
+    },
+    {
+      "epoch": 0.4358416945373467,
+      "grad_norm": 0.5798848867416382,
+      "learning_rate": 8e-05,
+      "loss": 1.7472,
+      "step": 7819
+    },
+    {
+      "epoch": 0.4358974358974359,
+      "grad_norm": 0.4962586760520935,
+      "learning_rate": 8e-05,
+      "loss": 1.6727,
+      "step": 7820
+    },
+    {
+      "epoch": 0.4359531772575251,
+      "grad_norm": 0.5649489760398865,
+      "learning_rate": 8e-05,
+      "loss": 1.8386,
+      "step": 7821
+    },
+    {
+      "epoch": 0.43600891861761426,
+      "grad_norm": 0.5474053621292114,
+      "learning_rate": 8e-05,
+      "loss": 1.9251,
+      "step": 7822
+    },
+    {
+      "epoch": 0.43606465997770344,
+      "grad_norm": 0.5361340641975403,
+      "learning_rate": 8e-05,
+      "loss": 1.8305,
+      "step": 7823
+    },
+    {
+      "epoch": 0.4361204013377926,
+      "grad_norm": 0.5183270573616028,
+      "learning_rate": 8e-05,
+      "loss": 1.3515,
+      "step": 7824
+    },
+    {
+      "epoch": 0.43617614269788185,
+      "grad_norm": 0.5586016774177551,
+      "learning_rate": 8e-05,
+      "loss": 1.6641,
+      "step": 7825
+    },
+    {
+      "epoch": 0.43623188405797103,
+      "grad_norm": 0.5187327861785889,
+      "learning_rate": 8e-05,
+      "loss": 1.7939,
+      "step": 7826
+    },
+    {
+      "epoch": 0.4362876254180602,
+      "grad_norm": 0.49453333020210266,
+      "learning_rate": 8e-05,
+      "loss": 1.6929,
+      "step": 7827
+    },
+    {
+      "epoch": 0.4363433667781494,
+      "grad_norm": 0.485514372587204,
+      "learning_rate": 8e-05,
+      "loss": 1.352,
+      "step": 7828
+    },
+    {
+      "epoch": 0.43639910813823857,
+      "grad_norm": 0.6950998306274414,
+      "learning_rate": 8e-05,
+      "loss": 2.265,
+      "step": 7829
+    },
+    {
+      "epoch": 0.43645484949832775,
+      "grad_norm": 0.5308253765106201,
+      "learning_rate": 8e-05,
+      "loss": 1.8375,
+      "step": 7830
+    },
+    {
+      "epoch": 0.4365105908584169,
+      "grad_norm": 0.4920623302459717,
+      "learning_rate": 8e-05,
+      "loss": 1.5364,
+      "step": 7831
+    },
+    {
+      "epoch": 0.43656633221850616,
+      "grad_norm": 0.48871949315071106,
+      "learning_rate": 8e-05,
+      "loss": 1.6355,
+      "step": 7832
+    },
+    {
+      "epoch": 0.43662207357859534,
+      "grad_norm": 0.5611724257469177,
+      "learning_rate": 8e-05,
+      "loss": 1.7873,
+      "step": 7833
+    },
+    {
+      "epoch": 0.4366778149386845,
+      "grad_norm": 0.5402656197547913,
+      "learning_rate": 8e-05,
+      "loss": 1.5936,
+      "step": 7834
+    },
+    {
+      "epoch": 0.4367335562987737,
+      "grad_norm": 0.565010666847229,
+      "learning_rate": 8e-05,
+      "loss": 1.8024,
+      "step": 7835
+    },
+    {
+      "epoch": 0.4367892976588629,
+      "grad_norm": 0.5193888545036316,
+      "learning_rate": 8e-05,
+      "loss": 1.6301,
+      "step": 7836
+    },
+    {
+      "epoch": 0.43684503901895205,
+      "grad_norm": 0.5342622995376587,
+      "learning_rate": 8e-05,
+      "loss": 1.5462,
+      "step": 7837
+    },
+    {
+      "epoch": 0.43690078037904123,
+      "grad_norm": 0.5685501098632812,
+      "learning_rate": 8e-05,
+      "loss": 1.8972,
+      "step": 7838
+    },
+    {
+      "epoch": 0.4369565217391304,
+      "grad_norm": 0.5733136534690857,
+      "learning_rate": 8e-05,
+      "loss": 1.7854,
+      "step": 7839
+    },
+    {
+      "epoch": 0.43701226309921964,
+      "grad_norm": 0.5411170125007629,
+      "learning_rate": 8e-05,
+      "loss": 1.8297,
+      "step": 7840
+    },
+    {
+      "epoch": 0.4370680044593088,
+      "grad_norm": 0.5439412593841553,
+      "learning_rate": 8e-05,
+      "loss": 1.8925,
+      "step": 7841
+    },
+    {
+      "epoch": 0.437123745819398,
+      "grad_norm": 0.5242007374763489,
+      "learning_rate": 8e-05,
+      "loss": 1.6811,
+      "step": 7842
+    },
+    {
+      "epoch": 0.4371794871794872,
+      "grad_norm": 0.5315377116203308,
+      "learning_rate": 8e-05,
+      "loss": 1.5802,
+      "step": 7843
+    },
+    {
+      "epoch": 0.43723522853957636,
+      "grad_norm": 0.5164995193481445,
+      "learning_rate": 8e-05,
+      "loss": 1.6785,
+      "step": 7844
+    },
+    {
+      "epoch": 0.43729096989966554,
+      "grad_norm": 0.5350561141967773,
+      "learning_rate": 8e-05,
+      "loss": 1.7579,
+      "step": 7845
+    },
+    {
+      "epoch": 0.4373467112597547,
+      "grad_norm": 0.5074156522750854,
+      "learning_rate": 8e-05,
+      "loss": 1.4924,
+      "step": 7846
+    },
+    {
+      "epoch": 0.43740245261984395,
+      "grad_norm": 0.5803791880607605,
+      "learning_rate": 8e-05,
+      "loss": 1.9426,
+      "step": 7847
+    },
+    {
+      "epoch": 0.43745819397993313,
+      "grad_norm": 0.4957139492034912,
+      "learning_rate": 8e-05,
+      "loss": 1.6834,
+      "step": 7848
+    },
+    {
+      "epoch": 0.4375139353400223,
+      "grad_norm": 0.5319690108299255,
+      "learning_rate": 8e-05,
+      "loss": 1.6855,
+      "step": 7849
+    },
+    {
+      "epoch": 0.4375696767001115,
+      "grad_norm": 0.5462393760681152,
+      "learning_rate": 8e-05,
+      "loss": 1.5779,
+      "step": 7850
+    },
+    {
+      "epoch": 0.43762541806020067,
+      "grad_norm": 0.5485098958015442,
+      "learning_rate": 8e-05,
+      "loss": 1.6278,
+      "step": 7851
+    },
+    {
+      "epoch": 0.43768115942028984,
+      "grad_norm": 0.5800246000289917,
+      "learning_rate": 8e-05,
+      "loss": 1.8766,
+      "step": 7852
+    },
+    {
+      "epoch": 0.437736900780379,
+      "grad_norm": 0.5502092838287354,
+      "learning_rate": 8e-05,
+      "loss": 1.807,
+      "step": 7853
+    },
+    {
+      "epoch": 0.4377926421404682,
+      "grad_norm": 0.5767754912376404,
+      "learning_rate": 8e-05,
+      "loss": 1.8275,
+      "step": 7854
+    },
+    {
+      "epoch": 0.43784838350055744,
+      "grad_norm": 0.541012704372406,
+      "learning_rate": 8e-05,
+      "loss": 1.6968,
+      "step": 7855
+    },
+    {
+      "epoch": 0.4379041248606466,
+      "grad_norm": 0.4932102859020233,
+      "learning_rate": 8e-05,
+      "loss": 1.5194,
+      "step": 7856
+    },
+    {
+      "epoch": 0.4379598662207358,
+      "grad_norm": 0.5406380891799927,
+      "learning_rate": 8e-05,
+      "loss": 1.6381,
+      "step": 7857
+    },
+    {
+      "epoch": 0.438015607580825,
+      "grad_norm": 0.5157513618469238,
+      "learning_rate": 8e-05,
+      "loss": 1.7066,
+      "step": 7858
+    },
+    {
+      "epoch": 0.43807134894091415,
+      "grad_norm": 0.5263302326202393,
+      "learning_rate": 8e-05,
+      "loss": 1.5629,
+      "step": 7859
+    },
+    {
+      "epoch": 0.43812709030100333,
+      "grad_norm": 0.49297571182250977,
+      "learning_rate": 8e-05,
+      "loss": 1.4853,
+      "step": 7860
+    },
+    {
+      "epoch": 0.4381828316610925,
+      "grad_norm": 0.5538367629051208,
+      "learning_rate": 8e-05,
+      "loss": 1.76,
+      "step": 7861
+    },
+    {
+      "epoch": 0.43823857302118174,
+      "grad_norm": 0.5233355760574341,
+      "learning_rate": 8e-05,
+      "loss": 1.5433,
+      "step": 7862
+    },
+    {
+      "epoch": 0.4382943143812709,
+      "grad_norm": 0.5237699747085571,
+      "learning_rate": 8e-05,
+      "loss": 1.4886,
+      "step": 7863
+    },
+    {
+      "epoch": 0.4383500557413601,
+      "grad_norm": 0.5608038306236267,
+      "learning_rate": 8e-05,
+      "loss": 1.8291,
+      "step": 7864
+    },
+    {
+      "epoch": 0.4384057971014493,
+      "grad_norm": 0.5415871739387512,
+      "learning_rate": 8e-05,
+      "loss": 1.6691,
+      "step": 7865
+    },
+    {
+      "epoch": 0.43846153846153846,
+      "grad_norm": 0.5716668963432312,
+      "learning_rate": 8e-05,
+      "loss": 1.7024,
+      "step": 7866
+    },
+    {
+      "epoch": 0.43851727982162764,
+      "grad_norm": 0.5166226625442505,
+      "learning_rate": 8e-05,
+      "loss": 1.6799,
+      "step": 7867
+    },
+    {
+      "epoch": 0.4385730211817168,
+      "grad_norm": 0.5312373638153076,
+      "learning_rate": 8e-05,
+      "loss": 1.6249,
+      "step": 7868
+    },
+    {
+      "epoch": 0.438628762541806,
+      "grad_norm": 0.5225555896759033,
+      "learning_rate": 8e-05,
+      "loss": 1.6405,
+      "step": 7869
+    },
+    {
+      "epoch": 0.43868450390189523,
+      "grad_norm": 0.5243179202079773,
+      "learning_rate": 8e-05,
+      "loss": 1.5884,
+      "step": 7870
+    },
+    {
+      "epoch": 0.4387402452619844,
+      "grad_norm": 0.5011211037635803,
+      "learning_rate": 8e-05,
+      "loss": 1.6761,
+      "step": 7871
+    },
+    {
+      "epoch": 0.4387959866220736,
+      "grad_norm": 0.5171734690666199,
+      "learning_rate": 8e-05,
+      "loss": 1.1844,
+      "step": 7872
+    },
+    {
+      "epoch": 0.43885172798216276,
+      "grad_norm": 0.5413135290145874,
+      "learning_rate": 8e-05,
+      "loss": 1.6946,
+      "step": 7873
+    },
+    {
+      "epoch": 0.43890746934225194,
+      "grad_norm": 0.5040024518966675,
+      "learning_rate": 8e-05,
+      "loss": 1.7646,
+      "step": 7874
+    },
+    {
+      "epoch": 0.4389632107023411,
+      "grad_norm": 0.5325048565864563,
+      "learning_rate": 8e-05,
+      "loss": 1.5024,
+      "step": 7875
+    },
+    {
+      "epoch": 0.4390189520624303,
+      "grad_norm": 0.4997828006744385,
+      "learning_rate": 8e-05,
+      "loss": 1.5721,
+      "step": 7876
+    },
+    {
+      "epoch": 0.43907469342251954,
+      "grad_norm": 0.5768607258796692,
+      "learning_rate": 8e-05,
+      "loss": 1.8523,
+      "step": 7877
+    },
+    {
+      "epoch": 0.4391304347826087,
+      "grad_norm": 0.5463048815727234,
+      "learning_rate": 8e-05,
+      "loss": 1.8202,
+      "step": 7878
+    },
+    {
+      "epoch": 0.4391861761426979,
+      "grad_norm": 0.5254857540130615,
+      "learning_rate": 8e-05,
+      "loss": 1.6109,
+      "step": 7879
+    },
+    {
+      "epoch": 0.43924191750278707,
+      "grad_norm": 0.5765327215194702,
+      "learning_rate": 8e-05,
+      "loss": 1.7585,
+      "step": 7880
+    },
+    {
+      "epoch": 0.43929765886287625,
+      "grad_norm": 0.47521743178367615,
+      "learning_rate": 8e-05,
+      "loss": 1.5456,
+      "step": 7881
+    },
+    {
+      "epoch": 0.43935340022296543,
+      "grad_norm": 0.5405141115188599,
+      "learning_rate": 8e-05,
+      "loss": 1.811,
+      "step": 7882
+    },
+    {
+      "epoch": 0.4394091415830546,
+      "grad_norm": 0.5534927248954773,
+      "learning_rate": 8e-05,
+      "loss": 1.7256,
+      "step": 7883
+    },
+    {
+      "epoch": 0.4394648829431438,
+      "grad_norm": 0.5321674942970276,
+      "learning_rate": 8e-05,
+      "loss": 1.6274,
+      "step": 7884
+    },
+    {
+      "epoch": 0.439520624303233,
+      "grad_norm": 0.5295695066452026,
+      "learning_rate": 8e-05,
+      "loss": 1.5753,
+      "step": 7885
+    },
+    {
+      "epoch": 0.4395763656633222,
+      "grad_norm": 0.574080228805542,
+      "learning_rate": 8e-05,
+      "loss": 1.8412,
+      "step": 7886
+    },
+    {
+      "epoch": 0.4396321070234114,
+      "grad_norm": 0.5514618754386902,
+      "learning_rate": 8e-05,
+      "loss": 1.7035,
+      "step": 7887
+    },
+    {
+      "epoch": 0.43968784838350056,
+      "grad_norm": 0.516113817691803,
+      "learning_rate": 8e-05,
+      "loss": 1.6944,
+      "step": 7888
+    },
+    {
+      "epoch": 0.43974358974358974,
+      "grad_norm": 0.49809619784355164,
+      "learning_rate": 8e-05,
+      "loss": 1.5288,
+      "step": 7889
+    },
+    {
+      "epoch": 0.4397993311036789,
+      "grad_norm": 0.5675715804100037,
+      "learning_rate": 8e-05,
+      "loss": 1.4908,
+      "step": 7890
+    },
+    {
+      "epoch": 0.4398550724637681,
+      "grad_norm": 0.585404098033905,
+      "learning_rate": 8e-05,
+      "loss": 1.7151,
+      "step": 7891
+    },
+    {
+      "epoch": 0.4399108138238573,
+      "grad_norm": 0.5580907464027405,
+      "learning_rate": 8e-05,
+      "loss": 1.8979,
+      "step": 7892
+    },
+    {
+      "epoch": 0.4399665551839465,
+      "grad_norm": 0.5543110966682434,
+      "learning_rate": 8e-05,
+      "loss": 1.6024,
+      "step": 7893
+    },
+    {
+      "epoch": 0.4400222965440357,
+      "grad_norm": 0.4939456582069397,
+      "learning_rate": 8e-05,
+      "loss": 1.4742,
+      "step": 7894
+    },
+    {
+      "epoch": 0.44007803790412486,
+      "grad_norm": 0.5418000817298889,
+      "learning_rate": 8e-05,
+      "loss": 1.6381,
+      "step": 7895
+    },
+    {
+      "epoch": 0.44013377926421404,
+      "grad_norm": 0.5609815716743469,
+      "learning_rate": 8e-05,
+      "loss": 1.7848,
+      "step": 7896
+    },
+    {
+      "epoch": 0.4401895206243032,
+      "grad_norm": 0.5530222654342651,
+      "learning_rate": 8e-05,
+      "loss": 1.7114,
+      "step": 7897
+    },
+    {
+      "epoch": 0.4402452619843924,
+      "grad_norm": 0.5275927782058716,
+      "learning_rate": 8e-05,
+      "loss": 1.5007,
+      "step": 7898
+    },
+    {
+      "epoch": 0.4403010033444816,
+      "grad_norm": 0.5392025709152222,
+      "learning_rate": 8e-05,
+      "loss": 1.5712,
+      "step": 7899
+    },
+    {
+      "epoch": 0.4403567447045708,
+      "grad_norm": 0.5584535598754883,
+      "learning_rate": 8e-05,
+      "loss": 1.5063,
+      "step": 7900
+    },
+    {
+      "epoch": 0.44041248606466,
+      "grad_norm": 0.545294463634491,
+      "learning_rate": 8e-05,
+      "loss": 1.7439,
+      "step": 7901
+    },
+    {
+      "epoch": 0.44046822742474917,
+      "grad_norm": 0.5408179759979248,
+      "learning_rate": 8e-05,
+      "loss": 1.4718,
+      "step": 7902
+    },
+    {
+      "epoch": 0.44052396878483835,
+      "grad_norm": 0.4967776834964752,
+      "learning_rate": 8e-05,
+      "loss": 1.3889,
+      "step": 7903
+    },
+    {
+      "epoch": 0.4405797101449275,
+      "grad_norm": 0.5502292513847351,
+      "learning_rate": 8e-05,
+      "loss": 1.607,
+      "step": 7904
+    },
+    {
+      "epoch": 0.4406354515050167,
+      "grad_norm": 0.5698793530464172,
+      "learning_rate": 8e-05,
+      "loss": 1.8125,
+      "step": 7905
+    },
+    {
+      "epoch": 0.4406911928651059,
+      "grad_norm": 0.5307198762893677,
+      "learning_rate": 8e-05,
+      "loss": 1.5135,
+      "step": 7906
+    },
+    {
+      "epoch": 0.4407469342251951,
+      "grad_norm": 0.5222761034965515,
+      "learning_rate": 8e-05,
+      "loss": 1.7111,
+      "step": 7907
+    },
+    {
+      "epoch": 0.4408026755852843,
+      "grad_norm": 0.5538968443870544,
+      "learning_rate": 8e-05,
+      "loss": 1.7281,
+      "step": 7908
+    },
+    {
+      "epoch": 0.4408584169453735,
+      "grad_norm": 0.653179943561554,
+      "learning_rate": 8e-05,
+      "loss": 1.596,
+      "step": 7909
+    },
+    {
+      "epoch": 0.44091415830546266,
+      "grad_norm": 0.5297260880470276,
+      "learning_rate": 8e-05,
+      "loss": 1.554,
+      "step": 7910
+    },
+    {
+      "epoch": 0.44096989966555183,
+      "grad_norm": 0.5243282914161682,
+      "learning_rate": 8e-05,
+      "loss": 1.6496,
+      "step": 7911
+    },
+    {
+      "epoch": 0.441025641025641,
+      "grad_norm": 0.5334591269493103,
+      "learning_rate": 8e-05,
+      "loss": 1.5722,
+      "step": 7912
+    },
+    {
+      "epoch": 0.4410813823857302,
+      "grad_norm": 0.606288492679596,
+      "learning_rate": 8e-05,
+      "loss": 1.8366,
+      "step": 7913
+    },
+    {
+      "epoch": 0.44113712374581937,
+      "grad_norm": 0.582847535610199,
+      "learning_rate": 8e-05,
+      "loss": 1.7688,
+      "step": 7914
+    },
+    {
+      "epoch": 0.4411928651059086,
+      "grad_norm": 0.5157939195632935,
+      "learning_rate": 8e-05,
+      "loss": 1.6076,
+      "step": 7915
+    },
+    {
+      "epoch": 0.4412486064659978,
+      "grad_norm": 0.49170541763305664,
+      "learning_rate": 8e-05,
+      "loss": 1.5521,
+      "step": 7916
+    },
+    {
+      "epoch": 0.44130434782608696,
+      "grad_norm": 0.6261030435562134,
+      "learning_rate": 8e-05,
+      "loss": 1.823,
+      "step": 7917
+    },
+    {
+      "epoch": 0.44136008918617614,
+      "grad_norm": 0.5169453620910645,
+      "learning_rate": 8e-05,
+      "loss": 1.7102,
+      "step": 7918
+    },
+    {
+      "epoch": 0.4414158305462653,
+      "grad_norm": 0.5462013483047485,
+      "learning_rate": 8e-05,
+      "loss": 1.7069,
+      "step": 7919
+    },
+    {
+      "epoch": 0.4414715719063545,
+      "grad_norm": 0.5447542667388916,
+      "learning_rate": 8e-05,
+      "loss": 1.5881,
+      "step": 7920
+    },
+    {
+      "epoch": 0.4415273132664437,
+      "grad_norm": 0.5971205830574036,
+      "learning_rate": 8e-05,
+      "loss": 1.8295,
+      "step": 7921
+    },
+    {
+      "epoch": 0.4415830546265329,
+      "grad_norm": 0.5625289678573608,
+      "learning_rate": 8e-05,
+      "loss": 1.6345,
+      "step": 7922
+    },
+    {
+      "epoch": 0.4416387959866221,
+      "grad_norm": 0.5436424016952515,
+      "learning_rate": 8e-05,
+      "loss": 1.8182,
+      "step": 7923
+    },
+    {
+      "epoch": 0.44169453734671127,
+      "grad_norm": 0.5405243635177612,
+      "learning_rate": 8e-05,
+      "loss": 1.744,
+      "step": 7924
+    },
+    {
+      "epoch": 0.44175027870680045,
+      "grad_norm": 0.5950716733932495,
+      "learning_rate": 8e-05,
+      "loss": 1.859,
+      "step": 7925
+    },
+    {
+      "epoch": 0.4418060200668896,
+      "grad_norm": 0.531568169593811,
+      "learning_rate": 8e-05,
+      "loss": 1.7732,
+      "step": 7926
+    },
+    {
+      "epoch": 0.4418617614269788,
+      "grad_norm": 0.5521816611289978,
+      "learning_rate": 8e-05,
+      "loss": 1.5794,
+      "step": 7927
+    },
+    {
+      "epoch": 0.441917502787068,
+      "grad_norm": 0.5484359264373779,
+      "learning_rate": 8e-05,
+      "loss": 1.6711,
+      "step": 7928
+    },
+    {
+      "epoch": 0.4419732441471572,
+      "grad_norm": 0.5186398029327393,
+      "learning_rate": 8e-05,
+      "loss": 1.7833,
+      "step": 7929
+    },
+    {
+      "epoch": 0.4420289855072464,
+      "grad_norm": 0.5243989825248718,
+      "learning_rate": 8e-05,
+      "loss": 1.5285,
+      "step": 7930
+    },
+    {
+      "epoch": 0.4420847268673356,
+      "grad_norm": 0.4838677942752838,
+      "learning_rate": 8e-05,
+      "loss": 1.589,
+      "step": 7931
+    },
+    {
+      "epoch": 0.44214046822742475,
+      "grad_norm": 0.510401725769043,
+      "learning_rate": 8e-05,
+      "loss": 1.7556,
+      "step": 7932
+    },
+    {
+      "epoch": 0.44219620958751393,
+      "grad_norm": 0.5910479426383972,
+      "learning_rate": 8e-05,
+      "loss": 1.7874,
+      "step": 7933
+    },
+    {
+      "epoch": 0.4422519509476031,
+      "grad_norm": 0.577433168888092,
+      "learning_rate": 8e-05,
+      "loss": 1.784,
+      "step": 7934
+    },
+    {
+      "epoch": 0.4423076923076923,
+      "grad_norm": 0.49999353289604187,
+      "learning_rate": 8e-05,
+      "loss": 1.6503,
+      "step": 7935
+    },
+    {
+      "epoch": 0.44236343366778147,
+      "grad_norm": 0.48399224877357483,
+      "learning_rate": 8e-05,
+      "loss": 1.478,
+      "step": 7936
+    },
+    {
+      "epoch": 0.4424191750278707,
+      "grad_norm": 0.5222609639167786,
+      "learning_rate": 8e-05,
+      "loss": 1.5898,
+      "step": 7937
+    },
+    {
+      "epoch": 0.4424749163879599,
+      "grad_norm": 0.5327225923538208,
+      "learning_rate": 8e-05,
+      "loss": 1.7029,
+      "step": 7938
+    },
+    {
+      "epoch": 0.44253065774804906,
+      "grad_norm": 0.4968123137950897,
+      "learning_rate": 8e-05,
+      "loss": 1.4246,
+      "step": 7939
+    },
+    {
+      "epoch": 0.44258639910813824,
+      "grad_norm": 0.5522922873497009,
+      "learning_rate": 8e-05,
+      "loss": 1.8884,
+      "step": 7940
+    },
+    {
+      "epoch": 0.4426421404682274,
+      "grad_norm": 0.5389187932014465,
+      "learning_rate": 8e-05,
+      "loss": 1.6626,
+      "step": 7941
+    },
+    {
+      "epoch": 0.4426978818283166,
+      "grad_norm": 0.5206403732299805,
+      "learning_rate": 8e-05,
+      "loss": 1.5838,
+      "step": 7942
+    },
+    {
+      "epoch": 0.4427536231884058,
+      "grad_norm": 0.4918576776981354,
+      "learning_rate": 8e-05,
+      "loss": 1.4641,
+      "step": 7943
+    },
+    {
+      "epoch": 0.442809364548495,
+      "grad_norm": 0.5084859728813171,
+      "learning_rate": 8e-05,
+      "loss": 1.7413,
+      "step": 7944
+    },
+    {
+      "epoch": 0.4428651059085842,
+      "grad_norm": 0.5442695021629333,
+      "learning_rate": 8e-05,
+      "loss": 1.8337,
+      "step": 7945
+    },
+    {
+      "epoch": 0.44292084726867337,
+      "grad_norm": 0.5141339302062988,
+      "learning_rate": 8e-05,
+      "loss": 1.6379,
+      "step": 7946
+    },
+    {
+      "epoch": 0.44297658862876255,
+      "grad_norm": 0.5276790261268616,
+      "learning_rate": 8e-05,
+      "loss": 1.7592,
+      "step": 7947
+    },
+    {
+      "epoch": 0.4430323299888517,
+      "grad_norm": 0.5303980708122253,
+      "learning_rate": 8e-05,
+      "loss": 1.6602,
+      "step": 7948
+    },
+    {
+      "epoch": 0.4430880713489409,
+      "grad_norm": 0.5147736072540283,
+      "learning_rate": 8e-05,
+      "loss": 1.6483,
+      "step": 7949
+    },
+    {
+      "epoch": 0.4431438127090301,
+      "grad_norm": 0.5501843094825745,
+      "learning_rate": 8e-05,
+      "loss": 1.7459,
+      "step": 7950
+    },
+    {
+      "epoch": 0.44319955406911926,
+      "grad_norm": 0.5527022480964661,
+      "learning_rate": 8e-05,
+      "loss": 1.8121,
+      "step": 7951
+    },
+    {
+      "epoch": 0.4432552954292085,
+      "grad_norm": 0.534709095954895,
+      "learning_rate": 8e-05,
+      "loss": 1.6494,
+      "step": 7952
+    },
+    {
+      "epoch": 0.4433110367892977,
+      "grad_norm": 0.5975018739700317,
+      "learning_rate": 8e-05,
+      "loss": 1.69,
+      "step": 7953
+    },
+    {
+      "epoch": 0.44336677814938685,
+      "grad_norm": 0.5425289869308472,
+      "learning_rate": 8e-05,
+      "loss": 1.6613,
+      "step": 7954
+    },
+    {
+      "epoch": 0.44342251950947603,
+      "grad_norm": 0.5638290047645569,
+      "learning_rate": 8e-05,
+      "loss": 1.8224,
+      "step": 7955
+    },
+    {
+      "epoch": 0.4434782608695652,
+      "grad_norm": 0.5615090131759644,
+      "learning_rate": 8e-05,
+      "loss": 1.7671,
+      "step": 7956
+    },
+    {
+      "epoch": 0.4435340022296544,
+      "grad_norm": 0.5214371681213379,
+      "learning_rate": 8e-05,
+      "loss": 1.5474,
+      "step": 7957
+    },
+    {
+      "epoch": 0.44358974358974357,
+      "grad_norm": 0.524421215057373,
+      "learning_rate": 8e-05,
+      "loss": 1.7009,
+      "step": 7958
+    },
+    {
+      "epoch": 0.4436454849498328,
+      "grad_norm": 0.5154446363449097,
+      "learning_rate": 8e-05,
+      "loss": 1.6709,
+      "step": 7959
+    },
+    {
+      "epoch": 0.443701226309922,
+      "grad_norm": 0.5744690299034119,
+      "learning_rate": 8e-05,
+      "loss": 2.029,
+      "step": 7960
+    },
+    {
+      "epoch": 0.44375696767001116,
+      "grad_norm": 0.5760914087295532,
+      "learning_rate": 8e-05,
+      "loss": 1.9362,
+      "step": 7961
+    },
+    {
+      "epoch": 0.44381270903010034,
+      "grad_norm": 0.5083246231079102,
+      "learning_rate": 8e-05,
+      "loss": 1.6227,
+      "step": 7962
+    },
+    {
+      "epoch": 0.4438684503901895,
+      "grad_norm": 0.5764332413673401,
+      "learning_rate": 8e-05,
+      "loss": 1.7809,
+      "step": 7963
+    },
+    {
+      "epoch": 0.4439241917502787,
+      "grad_norm": 0.5300820469856262,
+      "learning_rate": 8e-05,
+      "loss": 1.6536,
+      "step": 7964
+    },
+    {
+      "epoch": 0.4439799331103679,
+      "grad_norm": 0.4995599687099457,
+      "learning_rate": 8e-05,
+      "loss": 1.5449,
+      "step": 7965
+    },
+    {
+      "epoch": 0.44403567447045705,
+      "grad_norm": 0.5778630375862122,
+      "learning_rate": 8e-05,
+      "loss": 1.6878,
+      "step": 7966
+    },
+    {
+      "epoch": 0.4440914158305463,
+      "grad_norm": 0.5369563698768616,
+      "learning_rate": 8e-05,
+      "loss": 1.5915,
+      "step": 7967
+    },
+    {
+      "epoch": 0.44414715719063547,
+      "grad_norm": 0.5145952701568604,
+      "learning_rate": 8e-05,
+      "loss": 1.5626,
+      "step": 7968
+    },
+    {
+      "epoch": 0.44420289855072465,
+      "grad_norm": 0.48712003231048584,
+      "learning_rate": 8e-05,
+      "loss": 1.4918,
+      "step": 7969
+    },
+    {
+      "epoch": 0.4442586399108138,
+      "grad_norm": 0.5495256781578064,
+      "learning_rate": 8e-05,
+      "loss": 1.7003,
+      "step": 7970
+    },
+    {
+      "epoch": 0.444314381270903,
+      "grad_norm": 0.5205605626106262,
+      "learning_rate": 8e-05,
+      "loss": 1.7063,
+      "step": 7971
+    },
+    {
+      "epoch": 0.4443701226309922,
+      "grad_norm": 0.5068941712379456,
+      "learning_rate": 8e-05,
+      "loss": 1.6017,
+      "step": 7972
+    },
+    {
+      "epoch": 0.44442586399108136,
+      "grad_norm": 0.5757874250411987,
+      "learning_rate": 8e-05,
+      "loss": 1.7222,
+      "step": 7973
+    },
+    {
+      "epoch": 0.4444816053511706,
+      "grad_norm": 0.5113906264305115,
+      "learning_rate": 8e-05,
+      "loss": 1.4284,
+      "step": 7974
+    },
+    {
+      "epoch": 0.4445373467112598,
+      "grad_norm": 0.6202803254127502,
+      "learning_rate": 8e-05,
+      "loss": 1.7175,
+      "step": 7975
+    },
+    {
+      "epoch": 0.44459308807134895,
+      "grad_norm": 0.5060017704963684,
+      "learning_rate": 8e-05,
+      "loss": 1.5773,
+      "step": 7976
+    },
+    {
+      "epoch": 0.44464882943143813,
+      "grad_norm": 0.5798066258430481,
+      "learning_rate": 8e-05,
+      "loss": 1.788,
+      "step": 7977
+    },
+    {
+      "epoch": 0.4447045707915273,
+      "grad_norm": 0.5086560249328613,
+      "learning_rate": 8e-05,
+      "loss": 1.5831,
+      "step": 7978
+    },
+    {
+      "epoch": 0.4447603121516165,
+      "grad_norm": 0.5149025917053223,
+      "learning_rate": 8e-05,
+      "loss": 1.5965,
+      "step": 7979
+    },
+    {
+      "epoch": 0.44481605351170567,
+      "grad_norm": 0.5903439521789551,
+      "learning_rate": 8e-05,
+      "loss": 1.8206,
+      "step": 7980
+    },
+    {
+      "epoch": 0.44487179487179485,
+      "grad_norm": 0.5610959529876709,
+      "learning_rate": 8e-05,
+      "loss": 1.6128,
+      "step": 7981
+    },
+    {
+      "epoch": 0.4449275362318841,
+      "grad_norm": 0.5200814008712769,
+      "learning_rate": 8e-05,
+      "loss": 1.5292,
+      "step": 7982
+    },
+    {
+      "epoch": 0.44498327759197326,
+      "grad_norm": 0.5364866256713867,
+      "learning_rate": 8e-05,
+      "loss": 1.7843,
+      "step": 7983
+    },
+    {
+      "epoch": 0.44503901895206244,
+      "grad_norm": 0.5320987105369568,
+      "learning_rate": 8e-05,
+      "loss": 1.662,
+      "step": 7984
+    },
+    {
+      "epoch": 0.4450947603121516,
+      "grad_norm": 0.5191739201545715,
+      "learning_rate": 8e-05,
+      "loss": 1.6594,
+      "step": 7985
+    },
+    {
+      "epoch": 0.4451505016722408,
+      "grad_norm": 0.5547230243682861,
+      "learning_rate": 8e-05,
+      "loss": 1.655,
+      "step": 7986
+    },
+    {
+      "epoch": 0.44520624303233,
+      "grad_norm": 0.5270265340805054,
+      "learning_rate": 8e-05,
+      "loss": 1.5838,
+      "step": 7987
+    },
+    {
+      "epoch": 0.44526198439241915,
+      "grad_norm": 0.550730288028717,
+      "learning_rate": 8e-05,
+      "loss": 1.748,
+      "step": 7988
+    },
+    {
+      "epoch": 0.4453177257525084,
+      "grad_norm": 0.4731638729572296,
+      "learning_rate": 8e-05,
+      "loss": 1.5173,
+      "step": 7989
+    },
+    {
+      "epoch": 0.44537346711259757,
+      "grad_norm": 0.542497992515564,
+      "learning_rate": 8e-05,
+      "loss": 1.6523,
+      "step": 7990
+    },
+    {
+      "epoch": 0.44542920847268674,
+      "grad_norm": 0.6007255911827087,
+      "learning_rate": 8e-05,
+      "loss": 1.8565,
+      "step": 7991
+    },
+    {
+      "epoch": 0.4454849498327759,
+      "grad_norm": 0.5397570133209229,
+      "learning_rate": 8e-05,
+      "loss": 1.5885,
+      "step": 7992
+    },
+    {
+      "epoch": 0.4455406911928651,
+      "grad_norm": 0.5419074892997742,
+      "learning_rate": 8e-05,
+      "loss": 1.7126,
+      "step": 7993
+    },
+    {
+      "epoch": 0.4455964325529543,
+      "grad_norm": 0.5122414827346802,
+      "learning_rate": 8e-05,
+      "loss": 1.7067,
+      "step": 7994
+    },
+    {
+      "epoch": 0.44565217391304346,
+      "grad_norm": 0.5504677295684814,
+      "learning_rate": 8e-05,
+      "loss": 1.7417,
+      "step": 7995
+    },
+    {
+      "epoch": 0.44570791527313264,
+      "grad_norm": 0.536055862903595,
+      "learning_rate": 8e-05,
+      "loss": 1.7303,
+      "step": 7996
+    },
+    {
+      "epoch": 0.44576365663322187,
+      "grad_norm": 0.5506960153579712,
+      "learning_rate": 8e-05,
+      "loss": 1.6504,
+      "step": 7997
+    },
+    {
+      "epoch": 0.44581939799331105,
+      "grad_norm": 0.580141544342041,
+      "learning_rate": 8e-05,
+      "loss": 1.7647,
+      "step": 7998
+    },
+    {
+      "epoch": 0.44587513935340023,
+      "grad_norm": 0.5886245369911194,
+      "learning_rate": 8e-05,
+      "loss": 1.7583,
+      "step": 7999
+    },
+    {
+      "epoch": 0.4459308807134894,
+      "grad_norm": 0.5789358019828796,
+      "learning_rate": 8e-05,
+      "loss": 1.8195,
+      "step": 8000
+    },
+    {
+      "epoch": 0.4459866220735786,
+      "grad_norm": 0.5332310795783997,
+      "learning_rate": 8e-05,
+      "loss": 1.8922,
+      "step": 8001
+    },
+    {
+      "epoch": 0.44604236343366777,
+      "grad_norm": 0.5233482718467712,
+      "learning_rate": 8e-05,
+      "loss": 1.5484,
+      "step": 8002
+    },
+    {
+      "epoch": 0.44609810479375694,
+      "grad_norm": 0.5520485639572144,
+      "learning_rate": 8e-05,
+      "loss": 1.8006,
+      "step": 8003
+    },
+    {
+      "epoch": 0.4461538461538462,
+      "grad_norm": 0.5176988840103149,
+      "learning_rate": 8e-05,
+      "loss": 1.7473,
+      "step": 8004
+    },
+    {
+      "epoch": 0.44620958751393536,
+      "grad_norm": 0.5222984552383423,
+      "learning_rate": 8e-05,
+      "loss": 1.68,
+      "step": 8005
+    },
+    {
+      "epoch": 0.44626532887402454,
+      "grad_norm": 0.5123849511146545,
+      "learning_rate": 8e-05,
+      "loss": 1.5285,
+      "step": 8006
+    },
+    {
+      "epoch": 0.4463210702341137,
+      "grad_norm": 0.6144101619720459,
+      "learning_rate": 8e-05,
+      "loss": 2.081,
+      "step": 8007
+    },
+    {
+      "epoch": 0.4463768115942029,
+      "grad_norm": 0.5395132303237915,
+      "learning_rate": 8e-05,
+      "loss": 1.7676,
+      "step": 8008
+    },
+    {
+      "epoch": 0.4464325529542921,
+      "grad_norm": 0.6325112581253052,
+      "learning_rate": 8e-05,
+      "loss": 2.0531,
+      "step": 8009
+    },
+    {
+      "epoch": 0.44648829431438125,
+      "grad_norm": 0.5071414709091187,
+      "learning_rate": 8e-05,
+      "loss": 1.551,
+      "step": 8010
+    },
+    {
+      "epoch": 0.44654403567447043,
+      "grad_norm": 0.5221314430236816,
+      "learning_rate": 8e-05,
+      "loss": 1.6285,
+      "step": 8011
+    },
+    {
+      "epoch": 0.44659977703455966,
+      "grad_norm": 0.5296001434326172,
+      "learning_rate": 8e-05,
+      "loss": 1.5697,
+      "step": 8012
+    },
+    {
+      "epoch": 0.44665551839464884,
+      "grad_norm": 0.5084794163703918,
+      "learning_rate": 8e-05,
+      "loss": 1.6643,
+      "step": 8013
+    },
+    {
+      "epoch": 0.446711259754738,
+      "grad_norm": 0.537729024887085,
+      "learning_rate": 8e-05,
+      "loss": 1.8762,
+      "step": 8014
+    },
+    {
+      "epoch": 0.4467670011148272,
+      "grad_norm": 0.5437967777252197,
+      "learning_rate": 8e-05,
+      "loss": 1.5961,
+      "step": 8015
+    },
+    {
+      "epoch": 0.4468227424749164,
+      "grad_norm": 0.5017749071121216,
+      "learning_rate": 8e-05,
+      "loss": 1.6359,
+      "step": 8016
+    },
+    {
+      "epoch": 0.44687848383500556,
+      "grad_norm": 0.5402575731277466,
+      "learning_rate": 8e-05,
+      "loss": 1.7692,
+      "step": 8017
+    },
+    {
+      "epoch": 0.44693422519509474,
+      "grad_norm": 0.6567726135253906,
+      "learning_rate": 8e-05,
+      "loss": 2.1172,
+      "step": 8018
+    },
+    {
+      "epoch": 0.44698996655518397,
+      "grad_norm": 0.5226815342903137,
+      "learning_rate": 8e-05,
+      "loss": 1.5514,
+      "step": 8019
+    },
+    {
+      "epoch": 0.44704570791527315,
+      "grad_norm": 0.5185201168060303,
+      "learning_rate": 8e-05,
+      "loss": 1.6736,
+      "step": 8020
+    },
+    {
+      "epoch": 0.44710144927536233,
+      "grad_norm": 0.5502247214317322,
+      "learning_rate": 8e-05,
+      "loss": 1.712,
+      "step": 8021
+    },
+    {
+      "epoch": 0.4471571906354515,
+      "grad_norm": 0.5351677536964417,
+      "learning_rate": 8e-05,
+      "loss": 1.8209,
+      "step": 8022
+    },
+    {
+      "epoch": 0.4472129319955407,
+      "grad_norm": 0.5364797711372375,
+      "learning_rate": 8e-05,
+      "loss": 1.7622,
+      "step": 8023
+    },
+    {
+      "epoch": 0.44726867335562986,
+      "grad_norm": 0.47554463148117065,
+      "learning_rate": 8e-05,
+      "loss": 1.3493,
+      "step": 8024
+    },
+    {
+      "epoch": 0.44732441471571904,
+      "grad_norm": 0.5028263330459595,
+      "learning_rate": 8e-05,
+      "loss": 1.6095,
+      "step": 8025
+    },
+    {
+      "epoch": 0.4473801560758083,
+      "grad_norm": 0.5260576605796814,
+      "learning_rate": 8e-05,
+      "loss": 1.7044,
+      "step": 8026
+    },
+    {
+      "epoch": 0.44743589743589746,
+      "grad_norm": 0.5328836441040039,
+      "learning_rate": 8e-05,
+      "loss": 1.7109,
+      "step": 8027
+    },
+    {
+      "epoch": 0.44749163879598663,
+      "grad_norm": 0.4906134307384491,
+      "learning_rate": 8e-05,
+      "loss": 1.4654,
+      "step": 8028
+    },
+    {
+      "epoch": 0.4475473801560758,
+      "grad_norm": 0.5100238919258118,
+      "learning_rate": 8e-05,
+      "loss": 1.5336,
+      "step": 8029
+    },
+    {
+      "epoch": 0.447603121516165,
+      "grad_norm": 0.53555828332901,
+      "learning_rate": 8e-05,
+      "loss": 1.6962,
+      "step": 8030
+    },
+    {
+      "epoch": 0.44765886287625417,
+      "grad_norm": 0.5725340247154236,
+      "learning_rate": 8e-05,
+      "loss": 1.9263,
+      "step": 8031
+    },
+    {
+      "epoch": 0.44771460423634335,
+      "grad_norm": 0.5331685543060303,
+      "learning_rate": 8e-05,
+      "loss": 1.7079,
+      "step": 8032
+    },
+    {
+      "epoch": 0.44777034559643253,
+      "grad_norm": 0.5597028136253357,
+      "learning_rate": 8e-05,
+      "loss": 1.5791,
+      "step": 8033
+    },
+    {
+      "epoch": 0.44782608695652176,
+      "grad_norm": 0.5003613829612732,
+      "learning_rate": 8e-05,
+      "loss": 1.6993,
+      "step": 8034
+    },
+    {
+      "epoch": 0.44788182831661094,
+      "grad_norm": 0.5053821206092834,
+      "learning_rate": 8e-05,
+      "loss": 1.5431,
+      "step": 8035
+    },
+    {
+      "epoch": 0.4479375696767001,
+      "grad_norm": 0.6037608981132507,
+      "learning_rate": 8e-05,
+      "loss": 1.6782,
+      "step": 8036
+    },
+    {
+      "epoch": 0.4479933110367893,
+      "grad_norm": 0.5720326900482178,
+      "learning_rate": 8e-05,
+      "loss": 1.6984,
+      "step": 8037
+    },
+    {
+      "epoch": 0.4480490523968785,
+      "grad_norm": 0.5353522300720215,
+      "learning_rate": 8e-05,
+      "loss": 1.615,
+      "step": 8038
+    },
+    {
+      "epoch": 0.44810479375696766,
+      "grad_norm": 0.5460042953491211,
+      "learning_rate": 8e-05,
+      "loss": 1.6436,
+      "step": 8039
+    },
+    {
+      "epoch": 0.44816053511705684,
+      "grad_norm": 0.5569713711738586,
+      "learning_rate": 8e-05,
+      "loss": 1.7649,
+      "step": 8040
+    },
+    {
+      "epoch": 0.44821627647714607,
+      "grad_norm": 0.5759195685386658,
+      "learning_rate": 8e-05,
+      "loss": 1.5646,
+      "step": 8041
+    },
+    {
+      "epoch": 0.44827201783723525,
+      "grad_norm": 0.5390092134475708,
+      "learning_rate": 8e-05,
+      "loss": 1.667,
+      "step": 8042
+    },
+    {
+      "epoch": 0.4483277591973244,
+      "grad_norm": 0.5163521766662598,
+      "learning_rate": 8e-05,
+      "loss": 1.7887,
+      "step": 8043
+    },
+    {
+      "epoch": 0.4483835005574136,
+      "grad_norm": 0.5562042593955994,
+      "learning_rate": 8e-05,
+      "loss": 1.8138,
+      "step": 8044
+    },
+    {
+      "epoch": 0.4484392419175028,
+      "grad_norm": 0.5692409873008728,
+      "learning_rate": 8e-05,
+      "loss": 1.7239,
+      "step": 8045
+    },
+    {
+      "epoch": 0.44849498327759196,
+      "grad_norm": 0.49043577909469604,
+      "learning_rate": 8e-05,
+      "loss": 1.4557,
+      "step": 8046
+    },
+    {
+      "epoch": 0.44855072463768114,
+      "grad_norm": 0.5952504873275757,
+      "learning_rate": 8e-05,
+      "loss": 1.9548,
+      "step": 8047
+    },
+    {
+      "epoch": 0.4486064659977703,
+      "grad_norm": 0.5599199533462524,
+      "learning_rate": 8e-05,
+      "loss": 1.6081,
+      "step": 8048
+    },
+    {
+      "epoch": 0.44866220735785955,
+      "grad_norm": 0.5060661435127258,
+      "learning_rate": 8e-05,
+      "loss": 1.496,
+      "step": 8049
+    },
+    {
+      "epoch": 0.44871794871794873,
+      "grad_norm": 0.5552914142608643,
+      "learning_rate": 8e-05,
+      "loss": 1.6489,
+      "step": 8050
+    },
+    {
+      "epoch": 0.4487736900780379,
+      "grad_norm": 0.5685629844665527,
+      "learning_rate": 8e-05,
+      "loss": 1.8693,
+      "step": 8051
+    },
+    {
+      "epoch": 0.4488294314381271,
+      "grad_norm": 0.45723462104797363,
+      "learning_rate": 8e-05,
+      "loss": 1.3753,
+      "step": 8052
+    },
+    {
+      "epoch": 0.44888517279821627,
+      "grad_norm": 0.4917463958263397,
+      "learning_rate": 8e-05,
+      "loss": 1.4215,
+      "step": 8053
+    },
+    {
+      "epoch": 0.44894091415830545,
+      "grad_norm": 0.5196283459663391,
+      "learning_rate": 8e-05,
+      "loss": 1.7516,
+      "step": 8054
+    },
+    {
+      "epoch": 0.4489966555183946,
+      "grad_norm": 0.538723349571228,
+      "learning_rate": 8e-05,
+      "loss": 1.7441,
+      "step": 8055
+    },
+    {
+      "epoch": 0.44905239687848386,
+      "grad_norm": 0.5024257898330688,
+      "learning_rate": 8e-05,
+      "loss": 1.4925,
+      "step": 8056
+    },
+    {
+      "epoch": 0.44910813823857304,
+      "grad_norm": 0.519081711769104,
+      "learning_rate": 8e-05,
+      "loss": 1.5659,
+      "step": 8057
+    },
+    {
+      "epoch": 0.4491638795986622,
+      "grad_norm": 0.5455350875854492,
+      "learning_rate": 8e-05,
+      "loss": 1.7434,
+      "step": 8058
+    },
+    {
+      "epoch": 0.4492196209587514,
+      "grad_norm": 0.5286114811897278,
+      "learning_rate": 8e-05,
+      "loss": 1.6247,
+      "step": 8059
+    },
+    {
+      "epoch": 0.4492753623188406,
+      "grad_norm": 0.5353128910064697,
+      "learning_rate": 8e-05,
+      "loss": 1.6816,
+      "step": 8060
+    },
+    {
+      "epoch": 0.44933110367892976,
+      "grad_norm": 0.5198448896408081,
+      "learning_rate": 8e-05,
+      "loss": 1.6891,
+      "step": 8061
+    },
+    {
+      "epoch": 0.44938684503901893,
+      "grad_norm": 0.5549296736717224,
+      "learning_rate": 8e-05,
+      "loss": 1.8031,
+      "step": 8062
+    },
+    {
+      "epoch": 0.4494425863991081,
+      "grad_norm": 0.5344325304031372,
+      "learning_rate": 8e-05,
+      "loss": 1.6114,
+      "step": 8063
+    },
+    {
+      "epoch": 0.44949832775919735,
+      "grad_norm": 0.5231785178184509,
+      "learning_rate": 8e-05,
+      "loss": 1.5725,
+      "step": 8064
+    },
+    {
+      "epoch": 0.4495540691192865,
+      "grad_norm": 0.5388935208320618,
+      "learning_rate": 8e-05,
+      "loss": 1.5811,
+      "step": 8065
+    },
+    {
+      "epoch": 0.4496098104793757,
+      "grad_norm": 0.5113973617553711,
+      "learning_rate": 8e-05,
+      "loss": 1.4581,
+      "step": 8066
+    },
+    {
+      "epoch": 0.4496655518394649,
+      "grad_norm": 0.5376384258270264,
+      "learning_rate": 8e-05,
+      "loss": 1.7599,
+      "step": 8067
+    },
+    {
+      "epoch": 0.44972129319955406,
+      "grad_norm": 0.5223584175109863,
+      "learning_rate": 8e-05,
+      "loss": 1.7786,
+      "step": 8068
+    },
+    {
+      "epoch": 0.44977703455964324,
+      "grad_norm": 0.5178060531616211,
+      "learning_rate": 8e-05,
+      "loss": 1.7542,
+      "step": 8069
+    },
+    {
+      "epoch": 0.4498327759197324,
+      "grad_norm": 0.5555859804153442,
+      "learning_rate": 8e-05,
+      "loss": 1.6353,
+      "step": 8070
+    },
+    {
+      "epoch": 0.44988851727982165,
+      "grad_norm": 0.510905921459198,
+      "learning_rate": 8e-05,
+      "loss": 1.6363,
+      "step": 8071
+    },
+    {
+      "epoch": 0.44994425863991083,
+      "grad_norm": 0.5641573667526245,
+      "learning_rate": 8e-05,
+      "loss": 1.8944,
+      "step": 8072
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5425226092338562,
+      "learning_rate": 8e-05,
+      "loss": 1.6644,
+      "step": 8073
+    },
+    {
+      "epoch": 0.4500557413600892,
+      "grad_norm": 0.5517387986183167,
+      "learning_rate": 8e-05,
+      "loss": 1.7881,
+      "step": 8074
+    },
+    {
+      "epoch": 0.45011148272017837,
+      "grad_norm": 0.5022097229957581,
+      "learning_rate": 8e-05,
+      "loss": 1.5671,
+      "step": 8075
+    },
+    {
+      "epoch": 0.45016722408026755,
+      "grad_norm": 0.551947295665741,
+      "learning_rate": 8e-05,
+      "loss": 1.8847,
+      "step": 8076
+    },
+    {
+      "epoch": 0.4502229654403567,
+      "grad_norm": 0.584666907787323,
+      "learning_rate": 8e-05,
+      "loss": 1.7084,
+      "step": 8077
+    },
+    {
+      "epoch": 0.4502787068004459,
+      "grad_norm": 0.536909818649292,
+      "learning_rate": 8e-05,
+      "loss": 1.6183,
+      "step": 8078
+    },
+    {
+      "epoch": 0.45033444816053514,
+      "grad_norm": 0.5274991393089294,
+      "learning_rate": 8e-05,
+      "loss": 1.7355,
+      "step": 8079
+    },
+    {
+      "epoch": 0.4503901895206243,
+      "grad_norm": 0.49505487084388733,
+      "learning_rate": 8e-05,
+      "loss": 1.5412,
+      "step": 8080
+    },
+    {
+      "epoch": 0.4504459308807135,
+      "grad_norm": 0.5947615504264832,
+      "learning_rate": 8e-05,
+      "loss": 1.964,
+      "step": 8081
+    },
+    {
+      "epoch": 0.4505016722408027,
+      "grad_norm": 0.5649814605712891,
+      "learning_rate": 8e-05,
+      "loss": 1.7741,
+      "step": 8082
+    },
+    {
+      "epoch": 0.45055741360089185,
+      "grad_norm": 0.495584100484848,
+      "learning_rate": 8e-05,
+      "loss": 1.7114,
+      "step": 8083
+    },
+    {
+      "epoch": 0.45061315496098103,
+      "grad_norm": 0.5267391204833984,
+      "learning_rate": 8e-05,
+      "loss": 1.8227,
+      "step": 8084
+    },
+    {
+      "epoch": 0.4506688963210702,
+      "grad_norm": 0.5202919840812683,
+      "learning_rate": 8e-05,
+      "loss": 1.598,
+      "step": 8085
+    },
+    {
+      "epoch": 0.45072463768115945,
+      "grad_norm": 0.5348260998725891,
+      "learning_rate": 8e-05,
+      "loss": 1.53,
+      "step": 8086
+    },
+    {
+      "epoch": 0.4507803790412486,
+      "grad_norm": 0.5375006198883057,
+      "learning_rate": 8e-05,
+      "loss": 1.4686,
+      "step": 8087
+    },
+    {
+      "epoch": 0.4508361204013378,
+      "grad_norm": 0.5393670797348022,
+      "learning_rate": 8e-05,
+      "loss": 1.6661,
+      "step": 8088
+    },
+    {
+      "epoch": 0.450891861761427,
+      "grad_norm": 0.5716339349746704,
+      "learning_rate": 8e-05,
+      "loss": 1.8264,
+      "step": 8089
+    },
+    {
+      "epoch": 0.45094760312151616,
+      "grad_norm": 0.5319709777832031,
+      "learning_rate": 8e-05,
+      "loss": 1.6998,
+      "step": 8090
+    },
+    {
+      "epoch": 0.45100334448160534,
+      "grad_norm": 0.5010347366333008,
+      "learning_rate": 8e-05,
+      "loss": 1.5386,
+      "step": 8091
+    },
+    {
+      "epoch": 0.4510590858416945,
+      "grad_norm": 0.5667153596878052,
+      "learning_rate": 8e-05,
+      "loss": 1.9579,
+      "step": 8092
+    },
+    {
+      "epoch": 0.4511148272017837,
+      "grad_norm": 0.500321090221405,
+      "learning_rate": 8e-05,
+      "loss": 1.687,
+      "step": 8093
+    },
+    {
+      "epoch": 0.45117056856187293,
+      "grad_norm": 0.5381765365600586,
+      "learning_rate": 8e-05,
+      "loss": 1.8494,
+      "step": 8094
+    },
+    {
+      "epoch": 0.4512263099219621,
+      "grad_norm": 0.5462884902954102,
+      "learning_rate": 8e-05,
+      "loss": 1.6532,
+      "step": 8095
+    },
+    {
+      "epoch": 0.4512820512820513,
+      "grad_norm": 0.5044798851013184,
+      "learning_rate": 8e-05,
+      "loss": 1.7236,
+      "step": 8096
+    },
+    {
+      "epoch": 0.45133779264214047,
+      "grad_norm": 0.45130854845046997,
+      "learning_rate": 8e-05,
+      "loss": 1.0285,
+      "step": 8097
+    },
+    {
+      "epoch": 0.45139353400222965,
+      "grad_norm": 0.5413728356361389,
+      "learning_rate": 8e-05,
+      "loss": 1.7617,
+      "step": 8098
+    },
+    {
+      "epoch": 0.4514492753623188,
+      "grad_norm": 0.5397976636886597,
+      "learning_rate": 8e-05,
+      "loss": 1.6764,
+      "step": 8099
+    },
+    {
+      "epoch": 0.451505016722408,
+      "grad_norm": 0.562067985534668,
+      "learning_rate": 8e-05,
+      "loss": 1.7465,
+      "step": 8100
+    },
+    {
+      "epoch": 0.45156075808249724,
+      "grad_norm": 0.5796635746955872,
+      "learning_rate": 8e-05,
+      "loss": 1.7751,
+      "step": 8101
+    },
+    {
+      "epoch": 0.4516164994425864,
+      "grad_norm": 0.5075339078903198,
+      "learning_rate": 8e-05,
+      "loss": 1.5174,
+      "step": 8102
+    },
+    {
+      "epoch": 0.4516722408026756,
+      "grad_norm": 0.4336899220943451,
+      "learning_rate": 8e-05,
+      "loss": 1.0754,
+      "step": 8103
+    },
+    {
+      "epoch": 0.4517279821627648,
+      "grad_norm": 0.5077371001243591,
+      "learning_rate": 8e-05,
+      "loss": 1.4769,
+      "step": 8104
+    },
+    {
+      "epoch": 0.45178372352285395,
+      "grad_norm": 0.5493600964546204,
+      "learning_rate": 8e-05,
+      "loss": 1.8034,
+      "step": 8105
+    },
+    {
+      "epoch": 0.45183946488294313,
+      "grad_norm": 0.4964161217212677,
+      "learning_rate": 8e-05,
+      "loss": 1.6415,
+      "step": 8106
+    },
+    {
+      "epoch": 0.4518952062430323,
+      "grad_norm": 0.5899708867073059,
+      "learning_rate": 8e-05,
+      "loss": 1.6904,
+      "step": 8107
+    },
+    {
+      "epoch": 0.4519509476031215,
+      "grad_norm": 0.5874199271202087,
+      "learning_rate": 8e-05,
+      "loss": 1.7159,
+      "step": 8108
+    },
+    {
+      "epoch": 0.4520066889632107,
+      "grad_norm": 0.597144365310669,
+      "learning_rate": 8e-05,
+      "loss": 1.658,
+      "step": 8109
+    },
+    {
+      "epoch": 0.4520624303232999,
+      "grad_norm": 0.5587716102600098,
+      "learning_rate": 8e-05,
+      "loss": 1.4525,
+      "step": 8110
+    },
+    {
+      "epoch": 0.4521181716833891,
+      "grad_norm": 0.5530357360839844,
+      "learning_rate": 8e-05,
+      "loss": 1.6442,
+      "step": 8111
+    },
+    {
+      "epoch": 0.45217391304347826,
+      "grad_norm": 0.5947765111923218,
+      "learning_rate": 8e-05,
+      "loss": 1.8093,
+      "step": 8112
+    },
+    {
+      "epoch": 0.45222965440356744,
+      "grad_norm": 0.5420628190040588,
+      "learning_rate": 8e-05,
+      "loss": 1.4677,
+      "step": 8113
+    },
+    {
+      "epoch": 0.4522853957636566,
+      "grad_norm": 0.5657048225402832,
+      "learning_rate": 8e-05,
+      "loss": 1.796,
+      "step": 8114
+    },
+    {
+      "epoch": 0.4523411371237458,
+      "grad_norm": 0.5704300403594971,
+      "learning_rate": 8e-05,
+      "loss": 1.7856,
+      "step": 8115
+    },
+    {
+      "epoch": 0.45239687848383503,
+      "grad_norm": 0.581071674823761,
+      "learning_rate": 8e-05,
+      "loss": 1.7953,
+      "step": 8116
+    },
+    {
+      "epoch": 0.4524526198439242,
+      "grad_norm": 0.49436646699905396,
+      "learning_rate": 8e-05,
+      "loss": 1.3443,
+      "step": 8117
+    },
+    {
+      "epoch": 0.4525083612040134,
+      "grad_norm": 0.47054535150527954,
+      "learning_rate": 8e-05,
+      "loss": 1.5879,
+      "step": 8118
+    },
+    {
+      "epoch": 0.45256410256410257,
+      "grad_norm": 0.5406829118728638,
+      "learning_rate": 8e-05,
+      "loss": 1.5736,
+      "step": 8119
+    },
+    {
+      "epoch": 0.45261984392419174,
+      "grad_norm": 0.605000913143158,
+      "learning_rate": 8e-05,
+      "loss": 1.9112,
+      "step": 8120
+    },
+    {
+      "epoch": 0.4526755852842809,
+      "grad_norm": 0.56277996301651,
+      "learning_rate": 8e-05,
+      "loss": 1.7241,
+      "step": 8121
+    },
+    {
+      "epoch": 0.4527313266443701,
+      "grad_norm": 0.5486575961112976,
+      "learning_rate": 8e-05,
+      "loss": 1.6242,
+      "step": 8122
+    },
+    {
+      "epoch": 0.45278706800445934,
+      "grad_norm": 0.4798527657985687,
+      "learning_rate": 8e-05,
+      "loss": 1.339,
+      "step": 8123
+    },
+    {
+      "epoch": 0.4528428093645485,
+      "grad_norm": 0.5389805436134338,
+      "learning_rate": 8e-05,
+      "loss": 1.5075,
+      "step": 8124
+    },
+    {
+      "epoch": 0.4528985507246377,
+      "grad_norm": 0.5828685760498047,
+      "learning_rate": 8e-05,
+      "loss": 1.497,
+      "step": 8125
+    },
+    {
+      "epoch": 0.4529542920847269,
+      "grad_norm": 0.5291416645050049,
+      "learning_rate": 8e-05,
+      "loss": 1.5458,
+      "step": 8126
+    },
+    {
+      "epoch": 0.45301003344481605,
+      "grad_norm": 0.5561142563819885,
+      "learning_rate": 8e-05,
+      "loss": 1.7098,
+      "step": 8127
+    },
+    {
+      "epoch": 0.45306577480490523,
+      "grad_norm": 0.5542564392089844,
+      "learning_rate": 8e-05,
+      "loss": 1.9053,
+      "step": 8128
+    },
+    {
+      "epoch": 0.4531215161649944,
+      "grad_norm": 0.5583918690681458,
+      "learning_rate": 8e-05,
+      "loss": 1.7097,
+      "step": 8129
+    },
+    {
+      "epoch": 0.4531772575250836,
+      "grad_norm": 0.6063544750213623,
+      "learning_rate": 8e-05,
+      "loss": 1.6789,
+      "step": 8130
+    },
+    {
+      "epoch": 0.4532329988851728,
+      "grad_norm": 0.5362454056739807,
+      "learning_rate": 8e-05,
+      "loss": 1.5972,
+      "step": 8131
+    },
+    {
+      "epoch": 0.453288740245262,
+      "grad_norm": 0.5259758830070496,
+      "learning_rate": 8e-05,
+      "loss": 1.6535,
+      "step": 8132
+    },
+    {
+      "epoch": 0.4533444816053512,
+      "grad_norm": 0.5174777507781982,
+      "learning_rate": 8e-05,
+      "loss": 1.6159,
+      "step": 8133
+    },
+    {
+      "epoch": 0.45340022296544036,
+      "grad_norm": 0.5773780345916748,
+      "learning_rate": 8e-05,
+      "loss": 1.842,
+      "step": 8134
+    },
+    {
+      "epoch": 0.45345596432552954,
+      "grad_norm": 0.5560287833213806,
+      "learning_rate": 8e-05,
+      "loss": 1.7839,
+      "step": 8135
+    },
+    {
+      "epoch": 0.4535117056856187,
+      "grad_norm": 0.6014232635498047,
+      "learning_rate": 8e-05,
+      "loss": 1.8046,
+      "step": 8136
+    },
+    {
+      "epoch": 0.4535674470457079,
+      "grad_norm": 0.5750653743743896,
+      "learning_rate": 8e-05,
+      "loss": 1.6044,
+      "step": 8137
+    },
+    {
+      "epoch": 0.45362318840579713,
+      "grad_norm": 0.4930271506309509,
+      "learning_rate": 8e-05,
+      "loss": 1.5389,
+      "step": 8138
+    },
+    {
+      "epoch": 0.4536789297658863,
+      "grad_norm": 0.515459418296814,
+      "learning_rate": 8e-05,
+      "loss": 1.5612,
+      "step": 8139
+    },
+    {
+      "epoch": 0.4537346711259755,
+      "grad_norm": 0.5473839640617371,
+      "learning_rate": 8e-05,
+      "loss": 1.645,
+      "step": 8140
+    },
+    {
+      "epoch": 0.45379041248606466,
+      "grad_norm": 0.5208000540733337,
+      "learning_rate": 8e-05,
+      "loss": 1.7808,
+      "step": 8141
+    },
+    {
+      "epoch": 0.45384615384615384,
+      "grad_norm": 0.5229403972625732,
+      "learning_rate": 8e-05,
+      "loss": 1.5968,
+      "step": 8142
+    },
+    {
+      "epoch": 0.453901895206243,
+      "grad_norm": 0.5802630186080933,
+      "learning_rate": 8e-05,
+      "loss": 1.6251,
+      "step": 8143
+    },
+    {
+      "epoch": 0.4539576365663322,
+      "grad_norm": 0.5335087180137634,
+      "learning_rate": 8e-05,
+      "loss": 1.7452,
+      "step": 8144
+    },
+    {
+      "epoch": 0.4540133779264214,
+      "grad_norm": 0.5810218453407288,
+      "learning_rate": 8e-05,
+      "loss": 1.8251,
+      "step": 8145
+    },
+    {
+      "epoch": 0.4540691192865106,
+      "grad_norm": 0.5198366045951843,
+      "learning_rate": 8e-05,
+      "loss": 1.6894,
+      "step": 8146
+    },
+    {
+      "epoch": 0.4541248606465998,
+      "grad_norm": 0.5838040709495544,
+      "learning_rate": 8e-05,
+      "loss": 1.7899,
+      "step": 8147
+    },
+    {
+      "epoch": 0.45418060200668897,
+      "grad_norm": 0.554544985294342,
+      "learning_rate": 8e-05,
+      "loss": 1.7979,
+      "step": 8148
+    },
+    {
+      "epoch": 0.45423634336677815,
+      "grad_norm": 0.5387526750564575,
+      "learning_rate": 8e-05,
+      "loss": 1.8184,
+      "step": 8149
+    },
+    {
+      "epoch": 0.45429208472686733,
+      "grad_norm": 0.5326625108718872,
+      "learning_rate": 8e-05,
+      "loss": 1.7459,
+      "step": 8150
+    },
+    {
+      "epoch": 0.4543478260869565,
+      "grad_norm": 0.5054154396057129,
+      "learning_rate": 8e-05,
+      "loss": 1.4972,
+      "step": 8151
+    },
+    {
+      "epoch": 0.4544035674470457,
+      "grad_norm": 0.5089333653450012,
+      "learning_rate": 8e-05,
+      "loss": 1.5232,
+      "step": 8152
+    },
+    {
+      "epoch": 0.4544593088071349,
+      "grad_norm": 0.5137532353401184,
+      "learning_rate": 8e-05,
+      "loss": 1.709,
+      "step": 8153
+    },
+    {
+      "epoch": 0.4545150501672241,
+      "grad_norm": 0.5468508005142212,
+      "learning_rate": 8e-05,
+      "loss": 1.7988,
+      "step": 8154
+    },
+    {
+      "epoch": 0.4545707915273133,
+      "grad_norm": 0.5106626152992249,
+      "learning_rate": 8e-05,
+      "loss": 1.5714,
+      "step": 8155
+    },
+    {
+      "epoch": 0.45462653288740246,
+      "grad_norm": 0.5454922318458557,
+      "learning_rate": 8e-05,
+      "loss": 1.7014,
+      "step": 8156
+    },
+    {
+      "epoch": 0.45468227424749164,
+      "grad_norm": 0.5221412181854248,
+      "learning_rate": 8e-05,
+      "loss": 1.6882,
+      "step": 8157
+    },
+    {
+      "epoch": 0.4547380156075808,
+      "grad_norm": 0.5512486100196838,
+      "learning_rate": 8e-05,
+      "loss": 1.75,
+      "step": 8158
+    },
+    {
+      "epoch": 0.45479375696767,
+      "grad_norm": 0.49368467926979065,
+      "learning_rate": 8e-05,
+      "loss": 1.5232,
+      "step": 8159
+    },
+    {
+      "epoch": 0.45484949832775917,
+      "grad_norm": 0.5954217910766602,
+      "learning_rate": 8e-05,
+      "loss": 1.8206,
+      "step": 8160
+    },
+    {
+      "epoch": 0.4549052396878484,
+      "grad_norm": 0.5530853271484375,
+      "learning_rate": 8e-05,
+      "loss": 1.7372,
+      "step": 8161
+    },
+    {
+      "epoch": 0.4549609810479376,
+      "grad_norm": 0.5027709007263184,
+      "learning_rate": 8e-05,
+      "loss": 1.3964,
+      "step": 8162
+    },
+    {
+      "epoch": 0.45501672240802676,
+      "grad_norm": 0.5821712017059326,
+      "learning_rate": 8e-05,
+      "loss": 1.8771,
+      "step": 8163
+    },
+    {
+      "epoch": 0.45507246376811594,
+      "grad_norm": 0.5425708293914795,
+      "learning_rate": 8e-05,
+      "loss": 1.6491,
+      "step": 8164
+    },
+    {
+      "epoch": 0.4551282051282051,
+      "grad_norm": 0.5588460564613342,
+      "learning_rate": 8e-05,
+      "loss": 1.7911,
+      "step": 8165
+    },
+    {
+      "epoch": 0.4551839464882943,
+      "grad_norm": 0.5832526683807373,
+      "learning_rate": 8e-05,
+      "loss": 1.5692,
+      "step": 8166
+    },
+    {
+      "epoch": 0.4552396878483835,
+      "grad_norm": 0.5861908197402954,
+      "learning_rate": 8e-05,
+      "loss": 1.8063,
+      "step": 8167
+    },
+    {
+      "epoch": 0.4552954292084727,
+      "grad_norm": 0.5518962144851685,
+      "learning_rate": 8e-05,
+      "loss": 1.693,
+      "step": 8168
+    },
+    {
+      "epoch": 0.4553511705685619,
+      "grad_norm": 0.5097272992134094,
+      "learning_rate": 8e-05,
+      "loss": 1.5919,
+      "step": 8169
+    },
+    {
+      "epoch": 0.45540691192865107,
+      "grad_norm": 0.5612396597862244,
+      "learning_rate": 8e-05,
+      "loss": 1.6945,
+      "step": 8170
+    },
+    {
+      "epoch": 0.45546265328874025,
+      "grad_norm": 0.6262779235839844,
+      "learning_rate": 8e-05,
+      "loss": 1.754,
+      "step": 8171
+    },
+    {
+      "epoch": 0.45551839464882943,
+      "grad_norm": 0.5311568975448608,
+      "learning_rate": 8e-05,
+      "loss": 1.5949,
+      "step": 8172
+    },
+    {
+      "epoch": 0.4555741360089186,
+      "grad_norm": 0.5751360654830933,
+      "learning_rate": 8e-05,
+      "loss": 1.6484,
+      "step": 8173
+    },
+    {
+      "epoch": 0.4556298773690078,
+      "grad_norm": 0.6115161180496216,
+      "learning_rate": 8e-05,
+      "loss": 1.914,
+      "step": 8174
+    },
+    {
+      "epoch": 0.45568561872909696,
+      "grad_norm": 0.580069899559021,
+      "learning_rate": 8e-05,
+      "loss": 2.048,
+      "step": 8175
+    },
+    {
+      "epoch": 0.4557413600891862,
+      "grad_norm": 0.5459027290344238,
+      "learning_rate": 8e-05,
+      "loss": 1.6791,
+      "step": 8176
+    },
+    {
+      "epoch": 0.4557971014492754,
+      "grad_norm": 0.5423399806022644,
+      "learning_rate": 8e-05,
+      "loss": 1.8994,
+      "step": 8177
+    },
+    {
+      "epoch": 0.45585284280936456,
+      "grad_norm": 0.5573829412460327,
+      "learning_rate": 8e-05,
+      "loss": 1.9555,
+      "step": 8178
+    },
+    {
+      "epoch": 0.45590858416945373,
+      "grad_norm": 0.5473054051399231,
+      "learning_rate": 8e-05,
+      "loss": 1.8268,
+      "step": 8179
+    },
+    {
+      "epoch": 0.4559643255295429,
+      "grad_norm": 0.5032323002815247,
+      "learning_rate": 8e-05,
+      "loss": 1.6568,
+      "step": 8180
+    },
+    {
+      "epoch": 0.4560200668896321,
+      "grad_norm": 0.6185225248336792,
+      "learning_rate": 8e-05,
+      "loss": 1.6041,
+      "step": 8181
+    },
+    {
+      "epoch": 0.45607580824972127,
+      "grad_norm": 0.4937954545021057,
+      "learning_rate": 8e-05,
+      "loss": 1.5299,
+      "step": 8182
+    },
+    {
+      "epoch": 0.4561315496098105,
+      "grad_norm": 0.5103335380554199,
+      "learning_rate": 8e-05,
+      "loss": 1.7353,
+      "step": 8183
+    },
+    {
+      "epoch": 0.4561872909698997,
+      "grad_norm": 0.5169482231140137,
+      "learning_rate": 8e-05,
+      "loss": 1.645,
+      "step": 8184
+    },
+    {
+      "epoch": 0.45624303232998886,
+      "grad_norm": 0.5453347563743591,
+      "learning_rate": 8e-05,
+      "loss": 1.6612,
+      "step": 8185
+    },
+    {
+      "epoch": 0.45629877369007804,
+      "grad_norm": 0.5726072192192078,
+      "learning_rate": 8e-05,
+      "loss": 1.7785,
+      "step": 8186
+    },
+    {
+      "epoch": 0.4563545150501672,
+      "grad_norm": 0.5240086317062378,
+      "learning_rate": 8e-05,
+      "loss": 1.4897,
+      "step": 8187
+    },
+    {
+      "epoch": 0.4564102564102564,
+      "grad_norm": 0.6142365336418152,
+      "learning_rate": 8e-05,
+      "loss": 1.6084,
+      "step": 8188
+    },
+    {
+      "epoch": 0.4564659977703456,
+      "grad_norm": 0.5494696497917175,
+      "learning_rate": 8e-05,
+      "loss": 1.7353,
+      "step": 8189
+    },
+    {
+      "epoch": 0.45652173913043476,
+      "grad_norm": 0.5338488817214966,
+      "learning_rate": 8e-05,
+      "loss": 1.7151,
+      "step": 8190
+    },
+    {
+      "epoch": 0.456577480490524,
+      "grad_norm": 0.5470588207244873,
+      "learning_rate": 8e-05,
+      "loss": 1.8233,
+      "step": 8191
+    },
+    {
+      "epoch": 0.45663322185061317,
+      "grad_norm": 0.5292598605155945,
+      "learning_rate": 8e-05,
+      "loss": 1.7907,
+      "step": 8192
+    },
+    {
+      "epoch": 0.45668896321070235,
+      "grad_norm": 0.52900230884552,
+      "learning_rate": 8e-05,
+      "loss": 1.8824,
+      "step": 8193
+    },
+    {
+      "epoch": 0.4567447045707915,
+      "grad_norm": 0.533981204032898,
+      "learning_rate": 8e-05,
+      "loss": 1.734,
+      "step": 8194
+    },
+    {
+      "epoch": 0.4568004459308807,
+      "grad_norm": 0.5497355461120605,
+      "learning_rate": 8e-05,
+      "loss": 1.644,
+      "step": 8195
+    },
+    {
+      "epoch": 0.4568561872909699,
+      "grad_norm": 0.5164412260055542,
+      "learning_rate": 8e-05,
+      "loss": 1.745,
+      "step": 8196
+    },
+    {
+      "epoch": 0.45691192865105906,
+      "grad_norm": 0.5101504921913147,
+      "learning_rate": 8e-05,
+      "loss": 1.7557,
+      "step": 8197
+    },
+    {
+      "epoch": 0.4569676700111483,
+      "grad_norm": 0.5320062637329102,
+      "learning_rate": 8e-05,
+      "loss": 1.7507,
+      "step": 8198
+    },
+    {
+      "epoch": 0.4570234113712375,
+      "grad_norm": 0.5199472308158875,
+      "learning_rate": 8e-05,
+      "loss": 1.6154,
+      "step": 8199
+    },
+    {
+      "epoch": 0.45707915273132665,
+      "grad_norm": 0.54566490650177,
+      "learning_rate": 8e-05,
+      "loss": 1.6001,
+      "step": 8200
+    },
+    {
+      "epoch": 0.45713489409141583,
+      "grad_norm": 0.4972429573535919,
+      "learning_rate": 8e-05,
+      "loss": 1.6148,
+      "step": 8201
+    },
+    {
+      "epoch": 0.457190635451505,
+      "grad_norm": 0.5039562582969666,
+      "learning_rate": 8e-05,
+      "loss": 1.4474,
+      "step": 8202
+    },
+    {
+      "epoch": 0.4572463768115942,
+      "grad_norm": 0.5226442217826843,
+      "learning_rate": 8e-05,
+      "loss": 1.6382,
+      "step": 8203
+    },
+    {
+      "epoch": 0.45730211817168337,
+      "grad_norm": 0.49594616889953613,
+      "learning_rate": 8e-05,
+      "loss": 1.5854,
+      "step": 8204
+    },
+    {
+      "epoch": 0.45735785953177255,
+      "grad_norm": 0.5066409111022949,
+      "learning_rate": 8e-05,
+      "loss": 1.4831,
+      "step": 8205
+    },
+    {
+      "epoch": 0.4574136008918618,
+      "grad_norm": 0.6288407444953918,
+      "learning_rate": 8e-05,
+      "loss": 1.9551,
+      "step": 8206
+    },
+    {
+      "epoch": 0.45746934225195096,
+      "grad_norm": 0.540301501750946,
+      "learning_rate": 8e-05,
+      "loss": 1.7047,
+      "step": 8207
+    },
+    {
+      "epoch": 0.45752508361204014,
+      "grad_norm": 0.5802655220031738,
+      "learning_rate": 8e-05,
+      "loss": 1.9917,
+      "step": 8208
+    },
+    {
+      "epoch": 0.4575808249721293,
+      "grad_norm": 0.5389567017555237,
+      "learning_rate": 8e-05,
+      "loss": 1.6907,
+      "step": 8209
+    },
+    {
+      "epoch": 0.4576365663322185,
+      "grad_norm": 0.49851271510124207,
+      "learning_rate": 8e-05,
+      "loss": 1.695,
+      "step": 8210
+    },
+    {
+      "epoch": 0.4576923076923077,
+      "grad_norm": 0.5140793919563293,
+      "learning_rate": 8e-05,
+      "loss": 1.6777,
+      "step": 8211
+    },
+    {
+      "epoch": 0.45774804905239685,
+      "grad_norm": 0.4872767925262451,
+      "learning_rate": 8e-05,
+      "loss": 1.4623,
+      "step": 8212
+    },
+    {
+      "epoch": 0.4578037904124861,
+      "grad_norm": 0.48600637912750244,
+      "learning_rate": 8e-05,
+      "loss": 1.4297,
+      "step": 8213
+    },
+    {
+      "epoch": 0.45785953177257527,
+      "grad_norm": 0.5143309235572815,
+      "learning_rate": 8e-05,
+      "loss": 1.5519,
+      "step": 8214
+    },
+    {
+      "epoch": 0.45791527313266445,
+      "grad_norm": 0.4996730089187622,
+      "learning_rate": 8e-05,
+      "loss": 1.5866,
+      "step": 8215
+    },
+    {
+      "epoch": 0.4579710144927536,
+      "grad_norm": 0.555542528629303,
+      "learning_rate": 8e-05,
+      "loss": 1.6117,
+      "step": 8216
+    },
+    {
+      "epoch": 0.4580267558528428,
+      "grad_norm": 0.5220987796783447,
+      "learning_rate": 8e-05,
+      "loss": 1.5935,
+      "step": 8217
+    },
+    {
+      "epoch": 0.458082497212932,
+      "grad_norm": 0.5406100749969482,
+      "learning_rate": 8e-05,
+      "loss": 1.6984,
+      "step": 8218
+    },
+    {
+      "epoch": 0.45813823857302116,
+      "grad_norm": 0.5386164784431458,
+      "learning_rate": 8e-05,
+      "loss": 1.7229,
+      "step": 8219
+    },
+    {
+      "epoch": 0.45819397993311034,
+      "grad_norm": 0.5644610524177551,
+      "learning_rate": 8e-05,
+      "loss": 1.7008,
+      "step": 8220
+    },
+    {
+      "epoch": 0.4582497212931996,
+      "grad_norm": 0.5660703182220459,
+      "learning_rate": 8e-05,
+      "loss": 1.7436,
+      "step": 8221
+    },
+    {
+      "epoch": 0.45830546265328875,
+      "grad_norm": 0.5259301662445068,
+      "learning_rate": 8e-05,
+      "loss": 1.7168,
+      "step": 8222
+    },
+    {
+      "epoch": 0.45836120401337793,
+      "grad_norm": 0.5591362714767456,
+      "learning_rate": 8e-05,
+      "loss": 1.968,
+      "step": 8223
+    },
+    {
+      "epoch": 0.4584169453734671,
+      "grad_norm": 0.5394548773765564,
+      "learning_rate": 8e-05,
+      "loss": 1.6454,
+      "step": 8224
+    },
+    {
+      "epoch": 0.4584726867335563,
+      "grad_norm": 0.6189616322517395,
+      "learning_rate": 8e-05,
+      "loss": 2.2171,
+      "step": 8225
+    },
+    {
+      "epoch": 0.45852842809364547,
+      "grad_norm": 0.504211962223053,
+      "learning_rate": 8e-05,
+      "loss": 1.5599,
+      "step": 8226
+    },
+    {
+      "epoch": 0.45858416945373465,
+      "grad_norm": 0.5423411130905151,
+      "learning_rate": 8e-05,
+      "loss": 1.8359,
+      "step": 8227
+    },
+    {
+      "epoch": 0.4586399108138239,
+      "grad_norm": 0.6187911033630371,
+      "learning_rate": 8e-05,
+      "loss": 2.113,
+      "step": 8228
+    },
+    {
+      "epoch": 0.45869565217391306,
+      "grad_norm": 0.5360573530197144,
+      "learning_rate": 8e-05,
+      "loss": 1.6487,
+      "step": 8229
+    },
+    {
+      "epoch": 0.45875139353400224,
+      "grad_norm": 0.5503401160240173,
+      "learning_rate": 8e-05,
+      "loss": 1.6517,
+      "step": 8230
+    },
+    {
+      "epoch": 0.4588071348940914,
+      "grad_norm": 0.49501150846481323,
+      "learning_rate": 8e-05,
+      "loss": 1.5461,
+      "step": 8231
+    },
+    {
+      "epoch": 0.4588628762541806,
+      "grad_norm": 0.5979353785514832,
+      "learning_rate": 8e-05,
+      "loss": 1.9651,
+      "step": 8232
+    },
+    {
+      "epoch": 0.4589186176142698,
+      "grad_norm": 0.5210656523704529,
+      "learning_rate": 8e-05,
+      "loss": 1.5778,
+      "step": 8233
+    },
+    {
+      "epoch": 0.45897435897435895,
+      "grad_norm": 0.5004912614822388,
+      "learning_rate": 8e-05,
+      "loss": 1.5335,
+      "step": 8234
+    },
+    {
+      "epoch": 0.4590301003344482,
+      "grad_norm": 0.5546466708183289,
+      "learning_rate": 8e-05,
+      "loss": 1.8972,
+      "step": 8235
+    },
+    {
+      "epoch": 0.45908584169453737,
+      "grad_norm": 0.5235152840614319,
+      "learning_rate": 8e-05,
+      "loss": 1.6181,
+      "step": 8236
+    },
+    {
+      "epoch": 0.45914158305462655,
+      "grad_norm": 0.5469987392425537,
+      "learning_rate": 8e-05,
+      "loss": 1.7178,
+      "step": 8237
+    },
+    {
+      "epoch": 0.4591973244147157,
+      "grad_norm": 0.5259850025177002,
+      "learning_rate": 8e-05,
+      "loss": 1.5475,
+      "step": 8238
+    },
+    {
+      "epoch": 0.4592530657748049,
+      "grad_norm": 0.5477474927902222,
+      "learning_rate": 8e-05,
+      "loss": 1.6689,
+      "step": 8239
+    },
+    {
+      "epoch": 0.4593088071348941,
+      "grad_norm": 0.5713329315185547,
+      "learning_rate": 8e-05,
+      "loss": 1.8525,
+      "step": 8240
+    },
+    {
+      "epoch": 0.45936454849498326,
+      "grad_norm": 0.5435706973075867,
+      "learning_rate": 8e-05,
+      "loss": 1.7169,
+      "step": 8241
+    },
+    {
+      "epoch": 0.45942028985507244,
+      "grad_norm": 0.5042891502380371,
+      "learning_rate": 8e-05,
+      "loss": 1.5887,
+      "step": 8242
+    },
+    {
+      "epoch": 0.4594760312151617,
+      "grad_norm": 0.5069809556007385,
+      "learning_rate": 8e-05,
+      "loss": 1.58,
+      "step": 8243
+    },
+    {
+      "epoch": 0.45953177257525085,
+      "grad_norm": 0.5203117728233337,
+      "learning_rate": 8e-05,
+      "loss": 1.643,
+      "step": 8244
+    },
+    {
+      "epoch": 0.45958751393534003,
+      "grad_norm": 0.5837569236755371,
+      "learning_rate": 8e-05,
+      "loss": 1.489,
+      "step": 8245
+    },
+    {
+      "epoch": 0.4596432552954292,
+      "grad_norm": 0.544971764087677,
+      "learning_rate": 8e-05,
+      "loss": 1.56,
+      "step": 8246
+    },
+    {
+      "epoch": 0.4596989966555184,
+      "grad_norm": 0.572570264339447,
+      "learning_rate": 8e-05,
+      "loss": 1.6885,
+      "step": 8247
+    },
+    {
+      "epoch": 0.45975473801560757,
+      "grad_norm": 0.5214897990226746,
+      "learning_rate": 8e-05,
+      "loss": 1.6651,
+      "step": 8248
+    },
+    {
+      "epoch": 0.45981047937569675,
+      "grad_norm": 0.5395656228065491,
+      "learning_rate": 8e-05,
+      "loss": 1.6458,
+      "step": 8249
+    },
+    {
+      "epoch": 0.459866220735786,
+      "grad_norm": 0.5213479399681091,
+      "learning_rate": 8e-05,
+      "loss": 1.5345,
+      "step": 8250
+    },
+    {
+      "epoch": 0.45992196209587516,
+      "grad_norm": 0.5405566692352295,
+      "learning_rate": 8e-05,
+      "loss": 1.7846,
+      "step": 8251
+    },
+    {
+      "epoch": 0.45997770345596434,
+      "grad_norm": 0.5674734115600586,
+      "learning_rate": 8e-05,
+      "loss": 1.6661,
+      "step": 8252
+    },
+    {
+      "epoch": 0.4600334448160535,
+      "grad_norm": 0.5487441420555115,
+      "learning_rate": 8e-05,
+      "loss": 1.7767,
+      "step": 8253
+    },
+    {
+      "epoch": 0.4600891861761427,
+      "grad_norm": 0.5241776704788208,
+      "learning_rate": 8e-05,
+      "loss": 1.7186,
+      "step": 8254
+    },
+    {
+      "epoch": 0.4601449275362319,
+      "grad_norm": 0.5231916904449463,
+      "learning_rate": 8e-05,
+      "loss": 1.4834,
+      "step": 8255
+    },
+    {
+      "epoch": 0.46020066889632105,
+      "grad_norm": 0.5078309774398804,
+      "learning_rate": 8e-05,
+      "loss": 1.6286,
+      "step": 8256
+    },
+    {
+      "epoch": 0.46025641025641023,
+      "grad_norm": 0.6028822660446167,
+      "learning_rate": 8e-05,
+      "loss": 1.7446,
+      "step": 8257
+    },
+    {
+      "epoch": 0.46031215161649947,
+      "grad_norm": 0.49096351861953735,
+      "learning_rate": 8e-05,
+      "loss": 1.5099,
+      "step": 8258
+    },
+    {
+      "epoch": 0.46036789297658864,
+      "grad_norm": 0.5947312712669373,
+      "learning_rate": 8e-05,
+      "loss": 1.7774,
+      "step": 8259
+    },
+    {
+      "epoch": 0.4604236343366778,
+      "grad_norm": 0.5490503907203674,
+      "learning_rate": 8e-05,
+      "loss": 1.7369,
+      "step": 8260
+    },
+    {
+      "epoch": 0.460479375696767,
+      "grad_norm": 0.49701809883117676,
+      "learning_rate": 8e-05,
+      "loss": 1.4358,
+      "step": 8261
+    },
+    {
+      "epoch": 0.4605351170568562,
+      "grad_norm": 0.5505596399307251,
+      "learning_rate": 8e-05,
+      "loss": 1.5358,
+      "step": 8262
+    },
+    {
+      "epoch": 0.46059085841694536,
+      "grad_norm": 0.5386398434638977,
+      "learning_rate": 8e-05,
+      "loss": 1.5321,
+      "step": 8263
+    },
+    {
+      "epoch": 0.46064659977703454,
+      "grad_norm": 0.5729582905769348,
+      "learning_rate": 8e-05,
+      "loss": 1.7571,
+      "step": 8264
+    },
+    {
+      "epoch": 0.46070234113712377,
+      "grad_norm": 0.591675877571106,
+      "learning_rate": 8e-05,
+      "loss": 1.8728,
+      "step": 8265
+    },
+    {
+      "epoch": 0.46075808249721295,
+      "grad_norm": 0.5355835556983948,
+      "learning_rate": 8e-05,
+      "loss": 1.6417,
+      "step": 8266
+    },
+    {
+      "epoch": 0.46081382385730213,
+      "grad_norm": 0.5285724997520447,
+      "learning_rate": 8e-05,
+      "loss": 1.7523,
+      "step": 8267
+    },
+    {
+      "epoch": 0.4608695652173913,
+      "grad_norm": 0.49647387862205505,
+      "learning_rate": 8e-05,
+      "loss": 1.5041,
+      "step": 8268
+    },
+    {
+      "epoch": 0.4609253065774805,
+      "grad_norm": 0.5497033596038818,
+      "learning_rate": 8e-05,
+      "loss": 1.5984,
+      "step": 8269
+    },
+    {
+      "epoch": 0.46098104793756967,
+      "grad_norm": 0.5773898959159851,
+      "learning_rate": 8e-05,
+      "loss": 1.7893,
+      "step": 8270
+    },
+    {
+      "epoch": 0.46103678929765884,
+      "grad_norm": 0.5707722902297974,
+      "learning_rate": 8e-05,
+      "loss": 1.6682,
+      "step": 8271
+    },
+    {
+      "epoch": 0.461092530657748,
+      "grad_norm": 0.6010302901268005,
+      "learning_rate": 8e-05,
+      "loss": 1.7398,
+      "step": 8272
+    },
+    {
+      "epoch": 0.46114827201783726,
+      "grad_norm": 0.5731756091117859,
+      "learning_rate": 8e-05,
+      "loss": 1.7073,
+      "step": 8273
+    },
+    {
+      "epoch": 0.46120401337792644,
+      "grad_norm": 0.5730472207069397,
+      "learning_rate": 8e-05,
+      "loss": 1.655,
+      "step": 8274
+    },
+    {
+      "epoch": 0.4612597547380156,
+      "grad_norm": 0.5540065169334412,
+      "learning_rate": 8e-05,
+      "loss": 1.8349,
+      "step": 8275
+    },
+    {
+      "epoch": 0.4613154960981048,
+      "grad_norm": 0.5118550658226013,
+      "learning_rate": 8e-05,
+      "loss": 1.5659,
+      "step": 8276
+    },
+    {
+      "epoch": 0.461371237458194,
+      "grad_norm": 0.5175141096115112,
+      "learning_rate": 8e-05,
+      "loss": 1.7364,
+      "step": 8277
+    },
+    {
+      "epoch": 0.46142697881828315,
+      "grad_norm": 0.5034962296485901,
+      "learning_rate": 8e-05,
+      "loss": 1.5762,
+      "step": 8278
+    },
+    {
+      "epoch": 0.46148272017837233,
+      "grad_norm": 0.5668886303901672,
+      "learning_rate": 8e-05,
+      "loss": 1.7841,
+      "step": 8279
+    },
+    {
+      "epoch": 0.46153846153846156,
+      "grad_norm": 0.4840858280658722,
+      "learning_rate": 8e-05,
+      "loss": 1.6333,
+      "step": 8280
+    },
+    {
+      "epoch": 0.46159420289855074,
+      "grad_norm": 0.5296512246131897,
+      "learning_rate": 8e-05,
+      "loss": 1.7393,
+      "step": 8281
+    },
+    {
+      "epoch": 0.4616499442586399,
+      "grad_norm": 0.5633383393287659,
+      "learning_rate": 8e-05,
+      "loss": 1.9755,
+      "step": 8282
+    },
+    {
+      "epoch": 0.4617056856187291,
+      "grad_norm": 0.5701639652252197,
+      "learning_rate": 8e-05,
+      "loss": 1.5747,
+      "step": 8283
+    },
+    {
+      "epoch": 0.4617614269788183,
+      "grad_norm": 0.5512565970420837,
+      "learning_rate": 8e-05,
+      "loss": 1.5175,
+      "step": 8284
+    },
+    {
+      "epoch": 0.46181716833890746,
+      "grad_norm": 0.5408527255058289,
+      "learning_rate": 8e-05,
+      "loss": 1.7483,
+      "step": 8285
+    },
+    {
+      "epoch": 0.46187290969899664,
+      "grad_norm": 0.5452939867973328,
+      "learning_rate": 8e-05,
+      "loss": 1.5374,
+      "step": 8286
+    },
+    {
+      "epoch": 0.4619286510590858,
+      "grad_norm": 0.5265973806381226,
+      "learning_rate": 8e-05,
+      "loss": 1.6867,
+      "step": 8287
+    },
+    {
+      "epoch": 0.46198439241917505,
+      "grad_norm": 0.5437179803848267,
+      "learning_rate": 8e-05,
+      "loss": 1.7102,
+      "step": 8288
+    },
+    {
+      "epoch": 0.46204013377926423,
+      "grad_norm": 0.5253024101257324,
+      "learning_rate": 8e-05,
+      "loss": 1.6835,
+      "step": 8289
+    },
+    {
+      "epoch": 0.4620958751393534,
+      "grad_norm": 0.524053156375885,
+      "learning_rate": 8e-05,
+      "loss": 1.6644,
+      "step": 8290
+    },
+    {
+      "epoch": 0.4621516164994426,
+      "grad_norm": 0.5442635416984558,
+      "learning_rate": 8e-05,
+      "loss": 1.6942,
+      "step": 8291
+    },
+    {
+      "epoch": 0.46220735785953176,
+      "grad_norm": 0.5474844574928284,
+      "learning_rate": 8e-05,
+      "loss": 1.636,
+      "step": 8292
+    },
+    {
+      "epoch": 0.46226309921962094,
+      "grad_norm": 0.5145108699798584,
+      "learning_rate": 8e-05,
+      "loss": 1.548,
+      "step": 8293
+    },
+    {
+      "epoch": 0.4623188405797101,
+      "grad_norm": 0.4863380789756775,
+      "learning_rate": 8e-05,
+      "loss": 1.5228,
+      "step": 8294
+    },
+    {
+      "epoch": 0.46237458193979936,
+      "grad_norm": 0.5318400263786316,
+      "learning_rate": 8e-05,
+      "loss": 1.3712,
+      "step": 8295
+    },
+    {
+      "epoch": 0.46243032329988853,
+      "grad_norm": 0.551691472530365,
+      "learning_rate": 8e-05,
+      "loss": 1.7128,
+      "step": 8296
+    },
+    {
+      "epoch": 0.4624860646599777,
+      "grad_norm": 0.5184952616691589,
+      "learning_rate": 8e-05,
+      "loss": 1.39,
+      "step": 8297
+    },
+    {
+      "epoch": 0.4625418060200669,
+      "grad_norm": 0.5191618204116821,
+      "learning_rate": 8e-05,
+      "loss": 1.5686,
+      "step": 8298
+    },
+    {
+      "epoch": 0.46259754738015607,
+      "grad_norm": 0.55131596326828,
+      "learning_rate": 8e-05,
+      "loss": 1.5918,
+      "step": 8299
+    },
+    {
+      "epoch": 0.46265328874024525,
+      "grad_norm": 0.5466975569725037,
+      "learning_rate": 8e-05,
+      "loss": 1.7046,
+      "step": 8300
+    },
+    {
+      "epoch": 0.46270903010033443,
+      "grad_norm": 0.5339773297309875,
+      "learning_rate": 8e-05,
+      "loss": 1.6391,
+      "step": 8301
+    },
+    {
+      "epoch": 0.4627647714604236,
+      "grad_norm": 0.5683463215827942,
+      "learning_rate": 8e-05,
+      "loss": 1.7752,
+      "step": 8302
+    },
+    {
+      "epoch": 0.46282051282051284,
+      "grad_norm": 0.5539059042930603,
+      "learning_rate": 8e-05,
+      "loss": 1.7801,
+      "step": 8303
+    },
+    {
+      "epoch": 0.462876254180602,
+      "grad_norm": 0.5928812026977539,
+      "learning_rate": 8e-05,
+      "loss": 1.7587,
+      "step": 8304
+    },
+    {
+      "epoch": 0.4629319955406912,
+      "grad_norm": 0.6081556081771851,
+      "learning_rate": 8e-05,
+      "loss": 1.9025,
+      "step": 8305
+    },
+    {
+      "epoch": 0.4629877369007804,
+      "grad_norm": 0.5395975112915039,
+      "learning_rate": 8e-05,
+      "loss": 1.77,
+      "step": 8306
+    },
+    {
+      "epoch": 0.46304347826086956,
+      "grad_norm": 0.5138643383979797,
+      "learning_rate": 8e-05,
+      "loss": 1.6165,
+      "step": 8307
+    },
+    {
+      "epoch": 0.46309921962095874,
+      "grad_norm": 0.48940345644950867,
+      "learning_rate": 8e-05,
+      "loss": 1.6167,
+      "step": 8308
+    },
+    {
+      "epoch": 0.4631549609810479,
+      "grad_norm": 0.606408417224884,
+      "learning_rate": 8e-05,
+      "loss": 1.4015,
+      "step": 8309
+    },
+    {
+      "epoch": 0.46321070234113715,
+      "grad_norm": 0.5240685939788818,
+      "learning_rate": 8e-05,
+      "loss": 1.5085,
+      "step": 8310
+    },
+    {
+      "epoch": 0.4632664437012263,
+      "grad_norm": 0.5632039308547974,
+      "learning_rate": 8e-05,
+      "loss": 1.7416,
+      "step": 8311
+    },
+    {
+      "epoch": 0.4633221850613155,
+      "grad_norm": 0.574478030204773,
+      "learning_rate": 8e-05,
+      "loss": 1.6456,
+      "step": 8312
+    },
+    {
+      "epoch": 0.4633779264214047,
+      "grad_norm": 0.5275530219078064,
+      "learning_rate": 8e-05,
+      "loss": 1.5196,
+      "step": 8313
+    },
+    {
+      "epoch": 0.46343366778149386,
+      "grad_norm": 0.5936029553413391,
+      "learning_rate": 8e-05,
+      "loss": 1.806,
+      "step": 8314
+    },
+    {
+      "epoch": 0.46348940914158304,
+      "grad_norm": 0.5386390686035156,
+      "learning_rate": 8e-05,
+      "loss": 1.7582,
+      "step": 8315
+    },
+    {
+      "epoch": 0.4635451505016722,
+      "grad_norm": 0.5835428237915039,
+      "learning_rate": 8e-05,
+      "loss": 1.7981,
+      "step": 8316
+    },
+    {
+      "epoch": 0.4636008918617614,
+      "grad_norm": 0.6085085868835449,
+      "learning_rate": 8e-05,
+      "loss": 2.0216,
+      "step": 8317
+    },
+    {
+      "epoch": 0.46365663322185063,
+      "grad_norm": 0.5607771277427673,
+      "learning_rate": 8e-05,
+      "loss": 1.5928,
+      "step": 8318
+    },
+    {
+      "epoch": 0.4637123745819398,
+      "grad_norm": 0.5809671878814697,
+      "learning_rate": 8e-05,
+      "loss": 1.7172,
+      "step": 8319
+    },
+    {
+      "epoch": 0.463768115942029,
+      "grad_norm": 0.5324718952178955,
+      "learning_rate": 8e-05,
+      "loss": 1.6558,
+      "step": 8320
+    },
+    {
+      "epoch": 0.46382385730211817,
+      "grad_norm": 0.5490670800209045,
+      "learning_rate": 8e-05,
+      "loss": 1.755,
+      "step": 8321
+    },
+    {
+      "epoch": 0.46387959866220735,
+      "grad_norm": 0.5422378778457642,
+      "learning_rate": 8e-05,
+      "loss": 1.7185,
+      "step": 8322
+    },
+    {
+      "epoch": 0.4639353400222965,
+      "grad_norm": 0.5793622136116028,
+      "learning_rate": 8e-05,
+      "loss": 1.7017,
+      "step": 8323
+    },
+    {
+      "epoch": 0.4639910813823857,
+      "grad_norm": 0.48198401927948,
+      "learning_rate": 8e-05,
+      "loss": 1.4639,
+      "step": 8324
+    },
+    {
+      "epoch": 0.46404682274247494,
+      "grad_norm": 0.9945459961891174,
+      "learning_rate": 8e-05,
+      "loss": 1.6198,
+      "step": 8325
+    },
+    {
+      "epoch": 0.4641025641025641,
+      "grad_norm": 0.5188176035881042,
+      "learning_rate": 8e-05,
+      "loss": 1.5543,
+      "step": 8326
+    },
+    {
+      "epoch": 0.4641583054626533,
+      "grad_norm": 0.5144892930984497,
+      "learning_rate": 8e-05,
+      "loss": 1.5877,
+      "step": 8327
+    },
+    {
+      "epoch": 0.4642140468227425,
+      "grad_norm": 0.5563486218452454,
+      "learning_rate": 8e-05,
+      "loss": 1.6031,
+      "step": 8328
+    },
+    {
+      "epoch": 0.46426978818283166,
+      "grad_norm": 0.5223346948623657,
+      "learning_rate": 8e-05,
+      "loss": 1.7237,
+      "step": 8329
+    },
+    {
+      "epoch": 0.46432552954292083,
+      "grad_norm": 0.5695260763168335,
+      "learning_rate": 8e-05,
+      "loss": 1.7654,
+      "step": 8330
+    },
+    {
+      "epoch": 0.46438127090301,
+      "grad_norm": 0.5626814365386963,
+      "learning_rate": 8e-05,
+      "loss": 1.881,
+      "step": 8331
+    },
+    {
+      "epoch": 0.46443701226309925,
+      "grad_norm": 0.5374181866645813,
+      "learning_rate": 8e-05,
+      "loss": 1.7529,
+      "step": 8332
+    },
+    {
+      "epoch": 0.4644927536231884,
+      "grad_norm": 0.5513303279876709,
+      "learning_rate": 8e-05,
+      "loss": 1.8025,
+      "step": 8333
+    },
+    {
+      "epoch": 0.4645484949832776,
+      "grad_norm": 0.5162280201911926,
+      "learning_rate": 8e-05,
+      "loss": 1.5826,
+      "step": 8334
+    },
+    {
+      "epoch": 0.4646042363433668,
+      "grad_norm": 0.5183243751525879,
+      "learning_rate": 8e-05,
+      "loss": 1.4256,
+      "step": 8335
+    },
+    {
+      "epoch": 0.46465997770345596,
+      "grad_norm": 0.5621330738067627,
+      "learning_rate": 8e-05,
+      "loss": 1.5479,
+      "step": 8336
+    },
+    {
+      "epoch": 0.46471571906354514,
+      "grad_norm": 0.5799599289894104,
+      "learning_rate": 8e-05,
+      "loss": 1.9382,
+      "step": 8337
+    },
+    {
+      "epoch": 0.4647714604236343,
+      "grad_norm": 0.5067434310913086,
+      "learning_rate": 8e-05,
+      "loss": 1.423,
+      "step": 8338
+    },
+    {
+      "epoch": 0.4648272017837235,
+      "grad_norm": 0.5357816219329834,
+      "learning_rate": 8e-05,
+      "loss": 1.5762,
+      "step": 8339
+    },
+    {
+      "epoch": 0.46488294314381273,
+      "grad_norm": 0.4942900538444519,
+      "learning_rate": 8e-05,
+      "loss": 1.4523,
+      "step": 8340
+    },
+    {
+      "epoch": 0.4649386845039019,
+      "grad_norm": 0.5133934617042542,
+      "learning_rate": 8e-05,
+      "loss": 1.6795,
+      "step": 8341
+    },
+    {
+      "epoch": 0.4649944258639911,
+      "grad_norm": 0.6044417023658752,
+      "learning_rate": 8e-05,
+      "loss": 1.8665,
+      "step": 8342
+    },
+    {
+      "epoch": 0.46505016722408027,
+      "grad_norm": 0.5712794065475464,
+      "learning_rate": 8e-05,
+      "loss": 1.7797,
+      "step": 8343
+    },
+    {
+      "epoch": 0.46510590858416945,
+      "grad_norm": 0.5416268706321716,
+      "learning_rate": 8e-05,
+      "loss": 1.5001,
+      "step": 8344
+    },
+    {
+      "epoch": 0.4651616499442586,
+      "grad_norm": 0.5173828601837158,
+      "learning_rate": 8e-05,
+      "loss": 1.6749,
+      "step": 8345
+    },
+    {
+      "epoch": 0.4652173913043478,
+      "grad_norm": 0.5193343162536621,
+      "learning_rate": 8e-05,
+      "loss": 1.6668,
+      "step": 8346
+    },
+    {
+      "epoch": 0.46527313266443704,
+      "grad_norm": 0.5364198088645935,
+      "learning_rate": 8e-05,
+      "loss": 1.9235,
+      "step": 8347
+    },
+    {
+      "epoch": 0.4653288740245262,
+      "grad_norm": 0.5241788029670715,
+      "learning_rate": 8e-05,
+      "loss": 1.659,
+      "step": 8348
+    },
+    {
+      "epoch": 0.4653846153846154,
+      "grad_norm": 0.5484277606010437,
+      "learning_rate": 8e-05,
+      "loss": 1.6348,
+      "step": 8349
+    },
+    {
+      "epoch": 0.4654403567447046,
+      "grad_norm": 0.5373669266700745,
+      "learning_rate": 8e-05,
+      "loss": 1.5944,
+      "step": 8350
+    },
+    {
+      "epoch": 0.46549609810479375,
+      "grad_norm": 0.5364479422569275,
+      "learning_rate": 8e-05,
+      "loss": 1.977,
+      "step": 8351
+    },
+    {
+      "epoch": 0.46555183946488293,
+      "grad_norm": 0.5501197576522827,
+      "learning_rate": 8e-05,
+      "loss": 1.5936,
+      "step": 8352
+    },
+    {
+      "epoch": 0.4656075808249721,
+      "grad_norm": 0.5254024267196655,
+      "learning_rate": 8e-05,
+      "loss": 1.8216,
+      "step": 8353
+    },
+    {
+      "epoch": 0.4656633221850613,
+      "grad_norm": 0.573980987071991,
+      "learning_rate": 8e-05,
+      "loss": 1.8467,
+      "step": 8354
+    },
+    {
+      "epoch": 0.4657190635451505,
+      "grad_norm": 0.5709452629089355,
+      "learning_rate": 8e-05,
+      "loss": 1.9287,
+      "step": 8355
+    },
+    {
+      "epoch": 0.4657748049052397,
+      "grad_norm": 0.5465955138206482,
+      "learning_rate": 8e-05,
+      "loss": 1.7063,
+      "step": 8356
+    },
+    {
+      "epoch": 0.4658305462653289,
+      "grad_norm": 0.5366584658622742,
+      "learning_rate": 8e-05,
+      "loss": 1.5926,
+      "step": 8357
+    },
+    {
+      "epoch": 0.46588628762541806,
+      "grad_norm": 0.552971363067627,
+      "learning_rate": 8e-05,
+      "loss": 1.7139,
+      "step": 8358
+    },
+    {
+      "epoch": 0.46594202898550724,
+      "grad_norm": 0.46810775995254517,
+      "learning_rate": 8e-05,
+      "loss": 1.4272,
+      "step": 8359
+    },
+    {
+      "epoch": 0.4659977703455964,
+      "grad_norm": 0.536858081817627,
+      "learning_rate": 8e-05,
+      "loss": 1.8526,
+      "step": 8360
+    },
+    {
+      "epoch": 0.4660535117056856,
+      "grad_norm": 0.46676334738731384,
+      "learning_rate": 8e-05,
+      "loss": 1.5068,
+      "step": 8361
+    },
+    {
+      "epoch": 0.46610925306577483,
+      "grad_norm": 0.49330854415893555,
+      "learning_rate": 8e-05,
+      "loss": 1.5198,
+      "step": 8362
+    },
+    {
+      "epoch": 0.466164994425864,
+      "grad_norm": 0.6294517517089844,
+      "learning_rate": 8e-05,
+      "loss": 1.6882,
+      "step": 8363
+    },
+    {
+      "epoch": 0.4662207357859532,
+      "grad_norm": 0.6168404221534729,
+      "learning_rate": 8e-05,
+      "loss": 1.8669,
+      "step": 8364
+    },
+    {
+      "epoch": 0.46627647714604237,
+      "grad_norm": 0.5400106310844421,
+      "learning_rate": 8e-05,
+      "loss": 1.7639,
+      "step": 8365
+    },
+    {
+      "epoch": 0.46633221850613155,
+      "grad_norm": 0.5267928242683411,
+      "learning_rate": 8e-05,
+      "loss": 1.5111,
+      "step": 8366
+    },
+    {
+      "epoch": 0.4663879598662207,
+      "grad_norm": 0.5415551066398621,
+      "learning_rate": 8e-05,
+      "loss": 1.6299,
+      "step": 8367
+    },
+    {
+      "epoch": 0.4664437012263099,
+      "grad_norm": 0.505455493927002,
+      "learning_rate": 8e-05,
+      "loss": 1.5415,
+      "step": 8368
+    },
+    {
+      "epoch": 0.4664994425863991,
+      "grad_norm": 0.5676425695419312,
+      "learning_rate": 8e-05,
+      "loss": 1.7696,
+      "step": 8369
+    },
+    {
+      "epoch": 0.4665551839464883,
+      "grad_norm": 0.48888587951660156,
+      "learning_rate": 8e-05,
+      "loss": 1.4738,
+      "step": 8370
+    },
+    {
+      "epoch": 0.4666109253065775,
+      "grad_norm": 0.5303835272789001,
+      "learning_rate": 8e-05,
+      "loss": 1.6513,
+      "step": 8371
+    },
+    {
+      "epoch": 0.4666666666666667,
+      "grad_norm": 0.5789093971252441,
+      "learning_rate": 8e-05,
+      "loss": 1.8686,
+      "step": 8372
+    },
+    {
+      "epoch": 0.46672240802675585,
+      "grad_norm": 0.5505883693695068,
+      "learning_rate": 8e-05,
+      "loss": 1.6741,
+      "step": 8373
+    },
+    {
+      "epoch": 0.46677814938684503,
+      "grad_norm": 0.6738007664680481,
+      "learning_rate": 8e-05,
+      "loss": 1.6415,
+      "step": 8374
+    },
+    {
+      "epoch": 0.4668338907469342,
+      "grad_norm": 0.6047912836074829,
+      "learning_rate": 8e-05,
+      "loss": 1.9487,
+      "step": 8375
+    },
+    {
+      "epoch": 0.4668896321070234,
+      "grad_norm": 0.5206901431083679,
+      "learning_rate": 8e-05,
+      "loss": 1.4995,
+      "step": 8376
+    },
+    {
+      "epoch": 0.4669453734671126,
+      "grad_norm": 0.5485989451408386,
+      "learning_rate": 8e-05,
+      "loss": 1.6406,
+      "step": 8377
+    },
+    {
+      "epoch": 0.4670011148272018,
+      "grad_norm": 0.5110633373260498,
+      "learning_rate": 8e-05,
+      "loss": 1.5564,
+      "step": 8378
+    },
+    {
+      "epoch": 0.467056856187291,
+      "grad_norm": 0.5560204982757568,
+      "learning_rate": 8e-05,
+      "loss": 1.8834,
+      "step": 8379
+    },
+    {
+      "epoch": 0.46711259754738016,
+      "grad_norm": 0.6657040119171143,
+      "learning_rate": 8e-05,
+      "loss": 1.4447,
+      "step": 8380
+    },
+    {
+      "epoch": 0.46716833890746934,
+      "grad_norm": 0.5403985381126404,
+      "learning_rate": 8e-05,
+      "loss": 1.733,
+      "step": 8381
+    },
+    {
+      "epoch": 0.4672240802675585,
+      "grad_norm": 0.5952733159065247,
+      "learning_rate": 8e-05,
+      "loss": 2.0547,
+      "step": 8382
+    },
+    {
+      "epoch": 0.4672798216276477,
+      "grad_norm": 0.5075384378433228,
+      "learning_rate": 8e-05,
+      "loss": 1.6152,
+      "step": 8383
+    },
+    {
+      "epoch": 0.4673355629877369,
+      "grad_norm": 0.5747597217559814,
+      "learning_rate": 8e-05,
+      "loss": 1.7946,
+      "step": 8384
+    },
+    {
+      "epoch": 0.4673913043478261,
+      "grad_norm": 0.6281068325042725,
+      "learning_rate": 8e-05,
+      "loss": 1.9414,
+      "step": 8385
+    },
+    {
+      "epoch": 0.4674470457079153,
+      "grad_norm": 0.5636025071144104,
+      "learning_rate": 8e-05,
+      "loss": 1.9194,
+      "step": 8386
+    },
+    {
+      "epoch": 0.46750278706800447,
+      "grad_norm": 0.5191909074783325,
+      "learning_rate": 8e-05,
+      "loss": 1.5481,
+      "step": 8387
+    },
+    {
+      "epoch": 0.46755852842809364,
+      "grad_norm": 0.5473641157150269,
+      "learning_rate": 8e-05,
+      "loss": 1.7042,
+      "step": 8388
+    },
+    {
+      "epoch": 0.4676142697881828,
+      "grad_norm": 0.5307797193527222,
+      "learning_rate": 8e-05,
+      "loss": 1.7302,
+      "step": 8389
+    },
+    {
+      "epoch": 0.467670011148272,
+      "grad_norm": 0.539536714553833,
+      "learning_rate": 8e-05,
+      "loss": 1.636,
+      "step": 8390
+    },
+    {
+      "epoch": 0.4677257525083612,
+      "grad_norm": 0.49172070622444153,
+      "learning_rate": 8e-05,
+      "loss": 1.5952,
+      "step": 8391
+    },
+    {
+      "epoch": 0.4677814938684504,
+      "grad_norm": 0.5307680368423462,
+      "learning_rate": 8e-05,
+      "loss": 1.6422,
+      "step": 8392
+    },
+    {
+      "epoch": 0.4678372352285396,
+      "grad_norm": 0.5460218787193298,
+      "learning_rate": 8e-05,
+      "loss": 1.8214,
+      "step": 8393
+    },
+    {
+      "epoch": 0.4678929765886288,
+      "grad_norm": 0.54899662733078,
+      "learning_rate": 8e-05,
+      "loss": 1.5739,
+      "step": 8394
+    },
+    {
+      "epoch": 0.46794871794871795,
+      "grad_norm": 0.5274426937103271,
+      "learning_rate": 8e-05,
+      "loss": 1.7815,
+      "step": 8395
+    },
+    {
+      "epoch": 0.46800445930880713,
+      "grad_norm": 0.49675053358078003,
+      "learning_rate": 8e-05,
+      "loss": 1.5168,
+      "step": 8396
+    },
+    {
+      "epoch": 0.4680602006688963,
+      "grad_norm": 0.5308000445365906,
+      "learning_rate": 8e-05,
+      "loss": 1.6003,
+      "step": 8397
+    },
+    {
+      "epoch": 0.4681159420289855,
+      "grad_norm": 0.5650284886360168,
+      "learning_rate": 8e-05,
+      "loss": 1.6046,
+      "step": 8398
+    },
+    {
+      "epoch": 0.46817168338907467,
+      "grad_norm": 0.5442639589309692,
+      "learning_rate": 8e-05,
+      "loss": 1.6954,
+      "step": 8399
+    },
+    {
+      "epoch": 0.4682274247491639,
+      "grad_norm": 0.6157032251358032,
+      "learning_rate": 8e-05,
+      "loss": 1.9056,
+      "step": 8400
+    },
+    {
+      "epoch": 0.4682831661092531,
+      "grad_norm": 0.545790433883667,
+      "learning_rate": 8e-05,
+      "loss": 1.7362,
+      "step": 8401
+    },
+    {
+      "epoch": 0.46833890746934226,
+      "grad_norm": 0.5099235773086548,
+      "learning_rate": 8e-05,
+      "loss": 1.5272,
+      "step": 8402
+    },
+    {
+      "epoch": 0.46839464882943144,
+      "grad_norm": 0.5452757477760315,
+      "learning_rate": 8e-05,
+      "loss": 1.7499,
+      "step": 8403
+    },
+    {
+      "epoch": 0.4684503901895206,
+      "grad_norm": 0.5631169080734253,
+      "learning_rate": 8e-05,
+      "loss": 1.9252,
+      "step": 8404
+    },
+    {
+      "epoch": 0.4685061315496098,
+      "grad_norm": 0.5493608713150024,
+      "learning_rate": 8e-05,
+      "loss": 1.6625,
+      "step": 8405
+    },
+    {
+      "epoch": 0.468561872909699,
+      "grad_norm": 0.5756227374076843,
+      "learning_rate": 8e-05,
+      "loss": 1.8305,
+      "step": 8406
+    },
+    {
+      "epoch": 0.4686176142697882,
+      "grad_norm": 0.4990144670009613,
+      "learning_rate": 8e-05,
+      "loss": 1.4709,
+      "step": 8407
+    },
+    {
+      "epoch": 0.4686733556298774,
+      "grad_norm": 0.5849319100379944,
+      "learning_rate": 8e-05,
+      "loss": 1.8436,
+      "step": 8408
+    },
+    {
+      "epoch": 0.46872909698996656,
+      "grad_norm": 0.5466557145118713,
+      "learning_rate": 8e-05,
+      "loss": 1.8921,
+      "step": 8409
+    },
+    {
+      "epoch": 0.46878483835005574,
+      "grad_norm": 0.5994129776954651,
+      "learning_rate": 8e-05,
+      "loss": 1.7685,
+      "step": 8410
+    },
+    {
+      "epoch": 0.4688405797101449,
+      "grad_norm": 0.512111485004425,
+      "learning_rate": 8e-05,
+      "loss": 1.6275,
+      "step": 8411
+    },
+    {
+      "epoch": 0.4688963210702341,
+      "grad_norm": 0.536937952041626,
+      "learning_rate": 8e-05,
+      "loss": 1.7214,
+      "step": 8412
+    },
+    {
+      "epoch": 0.4689520624303233,
+      "grad_norm": 0.545914888381958,
+      "learning_rate": 8e-05,
+      "loss": 1.6529,
+      "step": 8413
+    },
+    {
+      "epoch": 0.46900780379041246,
+      "grad_norm": 0.5283071994781494,
+      "learning_rate": 8e-05,
+      "loss": 1.779,
+      "step": 8414
+    },
+    {
+      "epoch": 0.4690635451505017,
+      "grad_norm": 0.4989955723285675,
+      "learning_rate": 8e-05,
+      "loss": 1.3232,
+      "step": 8415
+    },
+    {
+      "epoch": 0.46911928651059087,
+      "grad_norm": 0.5377459526062012,
+      "learning_rate": 8e-05,
+      "loss": 1.6318,
+      "step": 8416
+    },
+    {
+      "epoch": 0.46917502787068005,
+      "grad_norm": 0.5159767270088196,
+      "learning_rate": 8e-05,
+      "loss": 1.6881,
+      "step": 8417
+    },
+    {
+      "epoch": 0.46923076923076923,
+      "grad_norm": 0.49078115820884705,
+      "learning_rate": 8e-05,
+      "loss": 1.7409,
+      "step": 8418
+    },
+    {
+      "epoch": 0.4692865105908584,
+      "grad_norm": 0.5152768492698669,
+      "learning_rate": 8e-05,
+      "loss": 1.5558,
+      "step": 8419
+    },
+    {
+      "epoch": 0.4693422519509476,
+      "grad_norm": 0.545647144317627,
+      "learning_rate": 8e-05,
+      "loss": 1.8672,
+      "step": 8420
+    },
+    {
+      "epoch": 0.46939799331103677,
+      "grad_norm": 0.5529956221580505,
+      "learning_rate": 8e-05,
+      "loss": 1.6215,
+      "step": 8421
+    },
+    {
+      "epoch": 0.469453734671126,
+      "grad_norm": 0.5335527062416077,
+      "learning_rate": 8e-05,
+      "loss": 1.7322,
+      "step": 8422
+    },
+    {
+      "epoch": 0.4695094760312152,
+      "grad_norm": 0.5761918425559998,
+      "learning_rate": 8e-05,
+      "loss": 1.8132,
+      "step": 8423
+    },
+    {
+      "epoch": 0.46956521739130436,
+      "grad_norm": 0.5925495028495789,
+      "learning_rate": 8e-05,
+      "loss": 1.8973,
+      "step": 8424
+    },
+    {
+      "epoch": 0.46962095875139354,
+      "grad_norm": 0.5872844457626343,
+      "learning_rate": 8e-05,
+      "loss": 1.6397,
+      "step": 8425
+    },
+    {
+      "epoch": 0.4696767001114827,
+      "grad_norm": 0.5724424123764038,
+      "learning_rate": 8e-05,
+      "loss": 1.6804,
+      "step": 8426
+    },
+    {
+      "epoch": 0.4697324414715719,
+      "grad_norm": 0.4837213158607483,
+      "learning_rate": 8e-05,
+      "loss": 1.4942,
+      "step": 8427
+    },
+    {
+      "epoch": 0.46978818283166107,
+      "grad_norm": 0.5545235276222229,
+      "learning_rate": 8e-05,
+      "loss": 1.3764,
+      "step": 8428
+    },
+    {
+      "epoch": 0.4698439241917503,
+      "grad_norm": 0.4970060884952545,
+      "learning_rate": 8e-05,
+      "loss": 1.5394,
+      "step": 8429
+    },
+    {
+      "epoch": 0.4698996655518395,
+      "grad_norm": 0.536975085735321,
+      "learning_rate": 8e-05,
+      "loss": 1.4085,
+      "step": 8430
+    },
+    {
+      "epoch": 0.46995540691192866,
+      "grad_norm": 0.5321915745735168,
+      "learning_rate": 8e-05,
+      "loss": 1.4961,
+      "step": 8431
+    },
+    {
+      "epoch": 0.47001114827201784,
+      "grad_norm": 0.5285022258758545,
+      "learning_rate": 8e-05,
+      "loss": 1.4844,
+      "step": 8432
+    },
+    {
+      "epoch": 0.470066889632107,
+      "grad_norm": 0.5677052140235901,
+      "learning_rate": 8e-05,
+      "loss": 1.8176,
+      "step": 8433
+    },
+    {
+      "epoch": 0.4701226309921962,
+      "grad_norm": 0.5485408306121826,
+      "learning_rate": 8e-05,
+      "loss": 1.5995,
+      "step": 8434
+    },
+    {
+      "epoch": 0.4701783723522854,
+      "grad_norm": 0.5371769666671753,
+      "learning_rate": 8e-05,
+      "loss": 1.711,
+      "step": 8435
+    },
+    {
+      "epoch": 0.47023411371237456,
+      "grad_norm": 0.5499669909477234,
+      "learning_rate": 8e-05,
+      "loss": 1.5055,
+      "step": 8436
+    },
+    {
+      "epoch": 0.4702898550724638,
+      "grad_norm": 0.5971784591674805,
+      "learning_rate": 8e-05,
+      "loss": 1.6764,
+      "step": 8437
+    },
+    {
+      "epoch": 0.47034559643255297,
+      "grad_norm": 0.7127771377563477,
+      "learning_rate": 8e-05,
+      "loss": 1.1981,
+      "step": 8438
+    },
+    {
+      "epoch": 0.47040133779264215,
+      "grad_norm": 0.5157849192619324,
+      "learning_rate": 8e-05,
+      "loss": 1.4334,
+      "step": 8439
+    },
+    {
+      "epoch": 0.47045707915273133,
+      "grad_norm": 0.5078108906745911,
+      "learning_rate": 8e-05,
+      "loss": 1.5883,
+      "step": 8440
+    },
+    {
+      "epoch": 0.4705128205128205,
+      "grad_norm": 0.5465660095214844,
+      "learning_rate": 8e-05,
+      "loss": 1.691,
+      "step": 8441
+    },
+    {
+      "epoch": 0.4705685618729097,
+      "grad_norm": 0.5386843681335449,
+      "learning_rate": 8e-05,
+      "loss": 1.7306,
+      "step": 8442
+    },
+    {
+      "epoch": 0.47062430323299886,
+      "grad_norm": 0.5482935905456543,
+      "learning_rate": 8e-05,
+      "loss": 1.8809,
+      "step": 8443
+    },
+    {
+      "epoch": 0.4706800445930881,
+      "grad_norm": 0.5584401488304138,
+      "learning_rate": 8e-05,
+      "loss": 1.668,
+      "step": 8444
+    },
+    {
+      "epoch": 0.4707357859531773,
+      "grad_norm": 0.4966360926628113,
+      "learning_rate": 8e-05,
+      "loss": 1.5818,
+      "step": 8445
+    },
+    {
+      "epoch": 0.47079152731326646,
+      "grad_norm": 0.5326342582702637,
+      "learning_rate": 8e-05,
+      "loss": 1.5713,
+      "step": 8446
+    },
+    {
+      "epoch": 0.47084726867335563,
+      "grad_norm": 0.542566180229187,
+      "learning_rate": 8e-05,
+      "loss": 1.7441,
+      "step": 8447
+    },
+    {
+      "epoch": 0.4709030100334448,
+      "grad_norm": 0.5604332089424133,
+      "learning_rate": 8e-05,
+      "loss": 1.6504,
+      "step": 8448
+    },
+    {
+      "epoch": 0.470958751393534,
+      "grad_norm": 0.5631572008132935,
+      "learning_rate": 8e-05,
+      "loss": 1.8132,
+      "step": 8449
+    },
+    {
+      "epoch": 0.47101449275362317,
+      "grad_norm": 0.5507049560546875,
+      "learning_rate": 8e-05,
+      "loss": 1.6097,
+      "step": 8450
+    },
+    {
+      "epoch": 0.47107023411371235,
+      "grad_norm": 0.5550432205200195,
+      "learning_rate": 8e-05,
+      "loss": 1.6979,
+      "step": 8451
+    },
+    {
+      "epoch": 0.4711259754738016,
+      "grad_norm": 0.5171401500701904,
+      "learning_rate": 8e-05,
+      "loss": 1.4894,
+      "step": 8452
+    },
+    {
+      "epoch": 0.47118171683389076,
+      "grad_norm": 0.5405150055885315,
+      "learning_rate": 8e-05,
+      "loss": 1.6846,
+      "step": 8453
+    },
+    {
+      "epoch": 0.47123745819397994,
+      "grad_norm": 0.536183774471283,
+      "learning_rate": 8e-05,
+      "loss": 1.4982,
+      "step": 8454
+    },
+    {
+      "epoch": 0.4712931995540691,
+      "grad_norm": 0.5685522556304932,
+      "learning_rate": 8e-05,
+      "loss": 1.785,
+      "step": 8455
+    },
+    {
+      "epoch": 0.4713489409141583,
+      "grad_norm": 0.5040280222892761,
+      "learning_rate": 8e-05,
+      "loss": 1.453,
+      "step": 8456
+    },
+    {
+      "epoch": 0.4714046822742475,
+      "grad_norm": 0.5402909517288208,
+      "learning_rate": 8e-05,
+      "loss": 1.7486,
+      "step": 8457
+    },
+    {
+      "epoch": 0.47146042363433666,
+      "grad_norm": 0.5560258030891418,
+      "learning_rate": 8e-05,
+      "loss": 1.636,
+      "step": 8458
+    },
+    {
+      "epoch": 0.4715161649944259,
+      "grad_norm": 0.5390499234199524,
+      "learning_rate": 8e-05,
+      "loss": 1.8613,
+      "step": 8459
+    },
+    {
+      "epoch": 0.47157190635451507,
+      "grad_norm": 0.5030994415283203,
+      "learning_rate": 8e-05,
+      "loss": 1.4811,
+      "step": 8460
+    },
+    {
+      "epoch": 0.47162764771460425,
+      "grad_norm": 0.5098398923873901,
+      "learning_rate": 8e-05,
+      "loss": 1.6269,
+      "step": 8461
+    },
+    {
+      "epoch": 0.4716833890746934,
+      "grad_norm": 0.5928331017494202,
+      "learning_rate": 8e-05,
+      "loss": 1.8651,
+      "step": 8462
+    },
+    {
+      "epoch": 0.4717391304347826,
+      "grad_norm": 0.5353269577026367,
+      "learning_rate": 8e-05,
+      "loss": 1.6416,
+      "step": 8463
+    },
+    {
+      "epoch": 0.4717948717948718,
+      "grad_norm": 0.5925639867782593,
+      "learning_rate": 8e-05,
+      "loss": 1.7972,
+      "step": 8464
+    },
+    {
+      "epoch": 0.47185061315496096,
+      "grad_norm": 0.46216219663619995,
+      "learning_rate": 8e-05,
+      "loss": 1.4756,
+      "step": 8465
+    },
+    {
+      "epoch": 0.47190635451505014,
+      "grad_norm": 0.5451775193214417,
+      "learning_rate": 8e-05,
+      "loss": 1.7737,
+      "step": 8466
+    },
+    {
+      "epoch": 0.4719620958751394,
+      "grad_norm": 0.5439514517784119,
+      "learning_rate": 8e-05,
+      "loss": 1.8168,
+      "step": 8467
+    },
+    {
+      "epoch": 0.47201783723522855,
+      "grad_norm": 0.5746572613716125,
+      "learning_rate": 8e-05,
+      "loss": 1.8156,
+      "step": 8468
+    },
+    {
+      "epoch": 0.47207357859531773,
+      "grad_norm": 0.5776638388633728,
+      "learning_rate": 8e-05,
+      "loss": 1.808,
+      "step": 8469
+    },
+    {
+      "epoch": 0.4721293199554069,
+      "grad_norm": 0.5580638647079468,
+      "learning_rate": 8e-05,
+      "loss": 1.5294,
+      "step": 8470
+    },
+    {
+      "epoch": 0.4721850613154961,
+      "grad_norm": 0.5210429430007935,
+      "learning_rate": 8e-05,
+      "loss": 1.5537,
+      "step": 8471
+    },
+    {
+      "epoch": 0.47224080267558527,
+      "grad_norm": 0.5610958337783813,
+      "learning_rate": 8e-05,
+      "loss": 1.7074,
+      "step": 8472
+    },
+    {
+      "epoch": 0.47229654403567445,
+      "grad_norm": 0.5312358736991882,
+      "learning_rate": 8e-05,
+      "loss": 1.7302,
+      "step": 8473
+    },
+    {
+      "epoch": 0.4723522853957637,
+      "grad_norm": 0.5046020746231079,
+      "learning_rate": 8e-05,
+      "loss": 1.6344,
+      "step": 8474
+    },
+    {
+      "epoch": 0.47240802675585286,
+      "grad_norm": 0.5662330389022827,
+      "learning_rate": 8e-05,
+      "loss": 1.8086,
+      "step": 8475
+    },
+    {
+      "epoch": 0.47246376811594204,
+      "grad_norm": 0.5743188261985779,
+      "learning_rate": 8e-05,
+      "loss": 1.7688,
+      "step": 8476
+    },
+    {
+      "epoch": 0.4725195094760312,
+      "grad_norm": 0.5933166742324829,
+      "learning_rate": 8e-05,
+      "loss": 1.6326,
+      "step": 8477
+    },
+    {
+      "epoch": 0.4725752508361204,
+      "grad_norm": 0.5768831968307495,
+      "learning_rate": 8e-05,
+      "loss": 1.6855,
+      "step": 8478
+    },
+    {
+      "epoch": 0.4726309921962096,
+      "grad_norm": 0.5617559552192688,
+      "learning_rate": 8e-05,
+      "loss": 1.5102,
+      "step": 8479
+    },
+    {
+      "epoch": 0.47268673355629875,
+      "grad_norm": 0.5413428544998169,
+      "learning_rate": 8e-05,
+      "loss": 1.6673,
+      "step": 8480
+    },
+    {
+      "epoch": 0.47274247491638793,
+      "grad_norm": 0.5708975195884705,
+      "learning_rate": 8e-05,
+      "loss": 1.7645,
+      "step": 8481
+    },
+    {
+      "epoch": 0.47279821627647717,
+      "grad_norm": 0.5631591081619263,
+      "learning_rate": 8e-05,
+      "loss": 1.7544,
+      "step": 8482
+    },
+    {
+      "epoch": 0.47285395763656635,
+      "grad_norm": 0.5494202375411987,
+      "learning_rate": 8e-05,
+      "loss": 1.6359,
+      "step": 8483
+    },
+    {
+      "epoch": 0.4729096989966555,
+      "grad_norm": 0.5949761867523193,
+      "learning_rate": 8e-05,
+      "loss": 1.7499,
+      "step": 8484
+    },
+    {
+      "epoch": 0.4729654403567447,
+      "grad_norm": 0.5369747877120972,
+      "learning_rate": 8e-05,
+      "loss": 1.6844,
+      "step": 8485
+    },
+    {
+      "epoch": 0.4730211817168339,
+      "grad_norm": 0.5614414215087891,
+      "learning_rate": 8e-05,
+      "loss": 1.7497,
+      "step": 8486
+    },
+    {
+      "epoch": 0.47307692307692306,
+      "grad_norm": 0.49822041392326355,
+      "learning_rate": 8e-05,
+      "loss": 1.4049,
+      "step": 8487
+    },
+    {
+      "epoch": 0.47313266443701224,
+      "grad_norm": 0.5776722431182861,
+      "learning_rate": 8e-05,
+      "loss": 1.6997,
+      "step": 8488
+    },
+    {
+      "epoch": 0.4731884057971015,
+      "grad_norm": NaN,
+      "learning_rate": 8e-05,
+      "loss": 2.2426,
+      "step": 8489
+    },
+    {
+      "epoch": 0.47324414715719065,
+      "grad_norm": 0.605370044708252,
+      "learning_rate": 8e-05,
+      "loss": 1.8601,
+      "step": 8490
+    },
+    {
+      "epoch": 0.47329988851727983,
+      "grad_norm": 0.5518434643745422,
+      "learning_rate": 8e-05,
+      "loss": 1.5683,
+      "step": 8491
+    },
+    {
+      "epoch": 0.473355629877369,
+      "grad_norm": 0.540189266204834,
+      "learning_rate": 8e-05,
+      "loss": 1.6447,
+      "step": 8492
+    },
+    {
+      "epoch": 0.4734113712374582,
+      "grad_norm": 0.5703498125076294,
+      "learning_rate": 8e-05,
+      "loss": 1.7412,
+      "step": 8493
+    },
+    {
+      "epoch": 0.47346711259754737,
+      "grad_norm": 0.5552210807800293,
+      "learning_rate": 8e-05,
+      "loss": 1.6012,
+      "step": 8494
+    },
+    {
+      "epoch": 0.47352285395763655,
+      "grad_norm": 0.5687077641487122,
+      "learning_rate": 8e-05,
+      "loss": 1.8304,
+      "step": 8495
+    },
+    {
+      "epoch": 0.4735785953177257,
+      "grad_norm": 0.5775901079177856,
+      "learning_rate": 8e-05,
+      "loss": 1.8197,
+      "step": 8496
+    },
+    {
+      "epoch": 0.47363433667781496,
+      "grad_norm": 0.5491219758987427,
+      "learning_rate": 8e-05,
+      "loss": 1.7912,
+      "step": 8497
+    },
+    {
+      "epoch": 0.47369007803790414,
+      "grad_norm": 0.520912230014801,
+      "learning_rate": 8e-05,
+      "loss": 1.7095,
+      "step": 8498
+    },
+    {
+      "epoch": 0.4737458193979933,
+      "grad_norm": 0.4933062791824341,
+      "learning_rate": 8e-05,
+      "loss": 1.4836,
+      "step": 8499
+    },
+    {
+      "epoch": 0.4738015607580825,
+      "grad_norm": 0.5397845506668091,
+      "learning_rate": 8e-05,
+      "loss": 1.8484,
+      "step": 8500
+    },
+    {
+      "epoch": 0.4738573021181717,
+      "grad_norm": 0.5235599279403687,
+      "learning_rate": 8e-05,
+      "loss": 1.823,
+      "step": 8501
+    },
+    {
+      "epoch": 0.47391304347826085,
+      "grad_norm": 0.5046244859695435,
+      "learning_rate": 8e-05,
+      "loss": 1.6333,
+      "step": 8502
+    },
+    {
+      "epoch": 0.47396878483835003,
+      "grad_norm": 0.5144981741905212,
+      "learning_rate": 8e-05,
+      "loss": 1.7431,
+      "step": 8503
+    },
+    {
+      "epoch": 0.47402452619843927,
+      "grad_norm": 0.5081440806388855,
+      "learning_rate": 8e-05,
+      "loss": 1.4979,
+      "step": 8504
+    },
+    {
+      "epoch": 0.47408026755852845,
+      "grad_norm": 0.550340473651886,
+      "learning_rate": 8e-05,
+      "loss": 1.6302,
+      "step": 8505
+    },
+    {
+      "epoch": 0.4741360089186176,
+      "grad_norm": 0.5056584477424622,
+      "learning_rate": 8e-05,
+      "loss": 1.4834,
+      "step": 8506
+    },
+    {
+      "epoch": 0.4741917502787068,
+      "grad_norm": 0.49061113595962524,
+      "learning_rate": 8e-05,
+      "loss": 1.4609,
+      "step": 8507
+    },
+    {
+      "epoch": 0.474247491638796,
+      "grad_norm": 0.6651285290718079,
+      "learning_rate": 8e-05,
+      "loss": 1.7197,
+      "step": 8508
+    },
+    {
+      "epoch": 0.47430323299888516,
+      "grad_norm": 0.5291346311569214,
+      "learning_rate": 8e-05,
+      "loss": 1.6432,
+      "step": 8509
+    },
+    {
+      "epoch": 0.47435897435897434,
+      "grad_norm": 0.5617138743400574,
+      "learning_rate": 8e-05,
+      "loss": 1.7347,
+      "step": 8510
+    },
+    {
+      "epoch": 0.4744147157190635,
+      "grad_norm": 0.498995304107666,
+      "learning_rate": 8e-05,
+      "loss": 1.5052,
+      "step": 8511
+    },
+    {
+      "epoch": 0.47447045707915275,
+      "grad_norm": 0.5444831252098083,
+      "learning_rate": 8e-05,
+      "loss": 1.7847,
+      "step": 8512
+    },
+    {
+      "epoch": 0.47452619843924193,
+      "grad_norm": 0.54232257604599,
+      "learning_rate": 8e-05,
+      "loss": 1.7109,
+      "step": 8513
+    },
+    {
+      "epoch": 0.4745819397993311,
+      "grad_norm": 0.5574086308479309,
+      "learning_rate": 8e-05,
+      "loss": 1.7192,
+      "step": 8514
+    },
+    {
+      "epoch": 0.4746376811594203,
+      "grad_norm": 0.5762770771980286,
+      "learning_rate": 8e-05,
+      "loss": 1.8808,
+      "step": 8515
+    },
+    {
+      "epoch": 0.47469342251950947,
+      "grad_norm": 0.5595988631248474,
+      "learning_rate": 8e-05,
+      "loss": 1.8657,
+      "step": 8516
+    },
+    {
+      "epoch": 0.47474916387959865,
+      "grad_norm": 0.5351122617721558,
+      "learning_rate": 8e-05,
+      "loss": 1.8992,
+      "step": 8517
+    },
+    {
+      "epoch": 0.4748049052396878,
+      "grad_norm": 0.5130336284637451,
+      "learning_rate": 8e-05,
+      "loss": 1.6199,
+      "step": 8518
+    },
+    {
+      "epoch": 0.47486064659977706,
+      "grad_norm": 0.5582343339920044,
+      "learning_rate": 8e-05,
+      "loss": 1.6847,
+      "step": 8519
+    },
+    {
+      "epoch": 0.47491638795986624,
+      "grad_norm": 0.5283982157707214,
+      "learning_rate": 8e-05,
+      "loss": 1.6256,
+      "step": 8520
+    },
+    {
+      "epoch": 0.4749721293199554,
+      "grad_norm": 0.501308262348175,
+      "learning_rate": 8e-05,
+      "loss": 1.4266,
+      "step": 8521
+    },
+    {
+      "epoch": 0.4750278706800446,
+      "grad_norm": 0.5760260820388794,
+      "learning_rate": 8e-05,
+      "loss": 1.7956,
+      "step": 8522
+    },
+    {
+      "epoch": 0.4750836120401338,
+      "grad_norm": 0.5730380415916443,
+      "learning_rate": 8e-05,
+      "loss": 1.564,
+      "step": 8523
+    },
+    {
+      "epoch": 0.47513935340022295,
+      "grad_norm": 0.5436396598815918,
+      "learning_rate": 8e-05,
+      "loss": 1.7967,
+      "step": 8524
+    },
+    {
+      "epoch": 0.47519509476031213,
+      "grad_norm": 0.5383810997009277,
+      "learning_rate": 8e-05,
+      "loss": 1.6471,
+      "step": 8525
+    },
+    {
+      "epoch": 0.47525083612040137,
+      "grad_norm": 0.548120379447937,
+      "learning_rate": 8e-05,
+      "loss": 1.6541,
+      "step": 8526
+    },
+    {
+      "epoch": 0.47530657748049054,
+      "grad_norm": 0.5204375982284546,
+      "learning_rate": 8e-05,
+      "loss": 1.4784,
+      "step": 8527
+    },
+    {
+      "epoch": 0.4753623188405797,
+      "grad_norm": 0.562260091304779,
+      "learning_rate": 8e-05,
+      "loss": 1.6466,
+      "step": 8528
+    },
+    {
+      "epoch": 0.4754180602006689,
+      "grad_norm": 0.523824155330658,
+      "learning_rate": 8e-05,
+      "loss": 1.5682,
+      "step": 8529
+    },
+    {
+      "epoch": 0.4754738015607581,
+      "grad_norm": 0.5674720406532288,
+      "learning_rate": 8e-05,
+      "loss": 1.6642,
+      "step": 8530
+    },
+    {
+      "epoch": 0.47552954292084726,
+      "grad_norm": 0.49034515023231506,
+      "learning_rate": 8e-05,
+      "loss": 1.2825,
+      "step": 8531
+    },
+    {
+      "epoch": 0.47558528428093644,
+      "grad_norm": 0.5165945291519165,
+      "learning_rate": 8e-05,
+      "loss": 1.6967,
+      "step": 8532
+    },
+    {
+      "epoch": 0.4756410256410256,
+      "grad_norm": 0.5310879349708557,
+      "learning_rate": 8e-05,
+      "loss": 1.448,
+      "step": 8533
+    },
+    {
+      "epoch": 0.47569676700111485,
+      "grad_norm": 0.5824185609817505,
+      "learning_rate": 8e-05,
+      "loss": 1.836,
+      "step": 8534
+    },
+    {
+      "epoch": 0.47575250836120403,
+      "grad_norm": 0.4889031946659088,
+      "learning_rate": 8e-05,
+      "loss": 1.3147,
+      "step": 8535
+    },
+    {
+      "epoch": 0.4758082497212932,
+      "grad_norm": 0.6152478456497192,
+      "learning_rate": 8e-05,
+      "loss": 1.8151,
+      "step": 8536
+    },
+    {
+      "epoch": 0.4758639910813824,
+      "grad_norm": 0.5502965450286865,
+      "learning_rate": 8e-05,
+      "loss": 1.6876,
+      "step": 8537
+    },
+    {
+      "epoch": 0.47591973244147157,
+      "grad_norm": 0.5226942896842957,
+      "learning_rate": 8e-05,
+      "loss": 1.6196,
+      "step": 8538
+    },
+    {
+      "epoch": 0.47597547380156074,
+      "grad_norm": 0.5234166383743286,
+      "learning_rate": 8e-05,
+      "loss": 1.7366,
+      "step": 8539
+    },
+    {
+      "epoch": 0.4760312151616499,
+      "grad_norm": 0.5769938230514526,
+      "learning_rate": 8e-05,
+      "loss": 1.8112,
+      "step": 8540
+    },
+    {
+      "epoch": 0.47608695652173916,
+      "grad_norm": 0.5301511287689209,
+      "learning_rate": 8e-05,
+      "loss": 1.805,
+      "step": 8541
+    },
+    {
+      "epoch": 0.47614269788182834,
+      "grad_norm": 0.5410844087600708,
+      "learning_rate": 8e-05,
+      "loss": 1.7966,
+      "step": 8542
+    },
+    {
+      "epoch": 0.4761984392419175,
+      "grad_norm": 0.5708724856376648,
+      "learning_rate": 8e-05,
+      "loss": 1.7287,
+      "step": 8543
+    },
+    {
+      "epoch": 0.4762541806020067,
+      "grad_norm": 0.516877293586731,
+      "learning_rate": 8e-05,
+      "loss": 1.6566,
+      "step": 8544
+    },
+    {
+      "epoch": 0.4763099219620959,
+      "grad_norm": 0.5604516267776489,
+      "learning_rate": 8e-05,
+      "loss": 1.7456,
+      "step": 8545
+    },
+    {
+      "epoch": 0.47636566332218505,
+      "grad_norm": 0.5355717539787292,
+      "learning_rate": 8e-05,
+      "loss": 1.6313,
+      "step": 8546
+    },
+    {
+      "epoch": 0.47642140468227423,
+      "grad_norm": 0.5061933994293213,
+      "learning_rate": 8e-05,
+      "loss": 1.5322,
+      "step": 8547
+    },
+    {
+      "epoch": 0.4764771460423634,
+      "grad_norm": 0.5413419604301453,
+      "learning_rate": 8e-05,
+      "loss": 1.7215,
+      "step": 8548
+    },
+    {
+      "epoch": 0.47653288740245264,
+      "grad_norm": 0.5458749532699585,
+      "learning_rate": 8e-05,
+      "loss": 1.8465,
+      "step": 8549
+    },
+    {
+      "epoch": 0.4765886287625418,
+      "grad_norm": 0.5370244979858398,
+      "learning_rate": 8e-05,
+      "loss": 1.5644,
+      "step": 8550
+    },
+    {
+      "epoch": 0.476644370122631,
+      "grad_norm": 0.5661664605140686,
+      "learning_rate": 8e-05,
+      "loss": 1.7357,
+      "step": 8551
+    },
+    {
+      "epoch": 0.4767001114827202,
+      "grad_norm": 0.5611180663108826,
+      "learning_rate": 8e-05,
+      "loss": 1.8036,
+      "step": 8552
+    },
+    {
+      "epoch": 0.47675585284280936,
+      "grad_norm": 0.5248943567276001,
+      "learning_rate": 8e-05,
+      "loss": 1.515,
+      "step": 8553
+    },
+    {
+      "epoch": 0.47681159420289854,
+      "grad_norm": 0.5423473715782166,
+      "learning_rate": 8e-05,
+      "loss": 1.6425,
+      "step": 8554
+    },
+    {
+      "epoch": 0.4768673355629877,
+      "grad_norm": 0.6139373183250427,
+      "learning_rate": 8e-05,
+      "loss": 1.7463,
+      "step": 8555
+    },
+    {
+      "epoch": 0.47692307692307695,
+      "grad_norm": 0.5869265794754028,
+      "learning_rate": 8e-05,
+      "loss": 1.7739,
+      "step": 8556
+    },
+    {
+      "epoch": 0.47697881828316613,
+      "grad_norm": 0.6074341535568237,
+      "learning_rate": 8e-05,
+      "loss": 1.9197,
+      "step": 8557
+    },
+    {
+      "epoch": 0.4770345596432553,
+      "grad_norm": 0.5665176510810852,
+      "learning_rate": 8e-05,
+      "loss": 1.6443,
+      "step": 8558
+    },
+    {
+      "epoch": 0.4770903010033445,
+      "grad_norm": 0.5026342272758484,
+      "learning_rate": 8e-05,
+      "loss": 1.4091,
+      "step": 8559
+    },
+    {
+      "epoch": 0.47714604236343366,
+      "grad_norm": 0.5059593915939331,
+      "learning_rate": 8e-05,
+      "loss": 1.5653,
+      "step": 8560
+    },
+    {
+      "epoch": 0.47720178372352284,
+      "grad_norm": 0.5121914744377136,
+      "learning_rate": 8e-05,
+      "loss": 1.3864,
+      "step": 8561
+    },
+    {
+      "epoch": 0.477257525083612,
+      "grad_norm": 0.5353608727455139,
+      "learning_rate": 8e-05,
+      "loss": 1.611,
+      "step": 8562
+    },
+    {
+      "epoch": 0.4773132664437012,
+      "grad_norm": 0.49897637963294983,
+      "learning_rate": 8e-05,
+      "loss": 1.6633,
+      "step": 8563
+    },
+    {
+      "epoch": 0.47736900780379043,
+      "grad_norm": 0.5345289707183838,
+      "learning_rate": 8e-05,
+      "loss": 1.7425,
+      "step": 8564
+    },
+    {
+      "epoch": 0.4774247491638796,
+      "grad_norm": 0.5157424807548523,
+      "learning_rate": 8e-05,
+      "loss": 1.4158,
+      "step": 8565
+    },
+    {
+      "epoch": 0.4774804905239688,
+      "grad_norm": 0.49930742383003235,
+      "learning_rate": 8e-05,
+      "loss": 1.6463,
+      "step": 8566
+    },
+    {
+      "epoch": 0.47753623188405797,
+      "grad_norm": 0.5590001344680786,
+      "learning_rate": 8e-05,
+      "loss": 1.7468,
+      "step": 8567
+    },
+    {
+      "epoch": 0.47759197324414715,
+      "grad_norm": 0.603706419467926,
+      "learning_rate": 8e-05,
+      "loss": 1.8045,
+      "step": 8568
+    },
+    {
+      "epoch": 0.47764771460423633,
+      "grad_norm": 0.5573416948318481,
+      "learning_rate": 8e-05,
+      "loss": 1.7055,
+      "step": 8569
+    },
+    {
+      "epoch": 0.4777034559643255,
+      "grad_norm": 0.5743922591209412,
+      "learning_rate": 8e-05,
+      "loss": 1.8453,
+      "step": 8570
+    },
+    {
+      "epoch": 0.47775919732441474,
+      "grad_norm": 0.5786899328231812,
+      "learning_rate": 8e-05,
+      "loss": 1.8062,
+      "step": 8571
+    },
+    {
+      "epoch": 0.4778149386845039,
+      "grad_norm": 0.511255145072937,
+      "learning_rate": 8e-05,
+      "loss": 1.6474,
+      "step": 8572
+    },
+    {
+      "epoch": 0.4778706800445931,
+      "grad_norm": 0.51045823097229,
+      "learning_rate": 8e-05,
+      "loss": 1.4694,
+      "step": 8573
+    },
+    {
+      "epoch": 0.4779264214046823,
+      "grad_norm": 0.5715101361274719,
+      "learning_rate": 8e-05,
+      "loss": 1.7614,
+      "step": 8574
+    },
+    {
+      "epoch": 0.47798216276477146,
+      "grad_norm": 0.5859196186065674,
+      "learning_rate": 8e-05,
+      "loss": 1.6408,
+      "step": 8575
+    },
+    {
+      "epoch": 0.47803790412486064,
+      "grad_norm": 0.5789876580238342,
+      "learning_rate": 8e-05,
+      "loss": 1.8372,
+      "step": 8576
+    },
+    {
+      "epoch": 0.4780936454849498,
+      "grad_norm": 0.5685731768608093,
+      "learning_rate": 8e-05,
+      "loss": 1.5789,
+      "step": 8577
+    },
+    {
+      "epoch": 0.478149386845039,
+      "grad_norm": 0.5826317071914673,
+      "learning_rate": 8e-05,
+      "loss": 1.6124,
+      "step": 8578
+    },
+    {
+      "epoch": 0.4782051282051282,
+      "grad_norm": 0.5494159460067749,
+      "learning_rate": 8e-05,
+      "loss": 1.6284,
+      "step": 8579
+    },
+    {
+      "epoch": 0.4782608695652174,
+      "grad_norm": 0.4774226248264313,
+      "learning_rate": 8e-05,
+      "loss": 1.3509,
+      "step": 8580
+    },
+    {
+      "epoch": 0.4783166109253066,
+      "grad_norm": 0.5364505648612976,
+      "learning_rate": 8e-05,
+      "loss": 1.6744,
+      "step": 8581
+    },
+    {
+      "epoch": 0.47837235228539576,
+      "grad_norm": 0.5289744734764099,
+      "learning_rate": 8e-05,
+      "loss": 1.5394,
+      "step": 8582
+    },
+    {
+      "epoch": 0.47842809364548494,
+      "grad_norm": 0.579842746257782,
+      "learning_rate": 8e-05,
+      "loss": 1.7983,
+      "step": 8583
+    },
+    {
+      "epoch": 0.4784838350055741,
+      "grad_norm": 0.5630995035171509,
+      "learning_rate": 8e-05,
+      "loss": 1.6644,
+      "step": 8584
+    },
+    {
+      "epoch": 0.4785395763656633,
+      "grad_norm": 0.5356509685516357,
+      "learning_rate": 8e-05,
+      "loss": 1.6247,
+      "step": 8585
+    },
+    {
+      "epoch": 0.47859531772575253,
+      "grad_norm": 0.5014398097991943,
+      "learning_rate": 8e-05,
+      "loss": 1.5935,
+      "step": 8586
+    },
+    {
+      "epoch": 0.4786510590858417,
+      "grad_norm": 0.5287692546844482,
+      "learning_rate": 8e-05,
+      "loss": 1.7297,
+      "step": 8587
+    },
+    {
+      "epoch": 0.4787068004459309,
+      "grad_norm": 0.5526402592658997,
+      "learning_rate": 8e-05,
+      "loss": 1.5543,
+      "step": 8588
+    },
+    {
+      "epoch": 0.47876254180602007,
+      "grad_norm": 0.547342836856842,
+      "learning_rate": 8e-05,
+      "loss": 1.5193,
+      "step": 8589
+    },
+    {
+      "epoch": 0.47881828316610925,
+      "grad_norm": 0.5509900450706482,
+      "learning_rate": 8e-05,
+      "loss": 1.7922,
+      "step": 8590
+    },
+    {
+      "epoch": 0.4788740245261984,
+      "grad_norm": 0.5440711975097656,
+      "learning_rate": 8e-05,
+      "loss": 1.5476,
+      "step": 8591
+    },
+    {
+      "epoch": 0.4789297658862876,
+      "grad_norm": 0.5216376185417175,
+      "learning_rate": 8e-05,
+      "loss": 1.4191,
+      "step": 8592
+    },
+    {
+      "epoch": 0.4789855072463768,
+      "grad_norm": 0.5691701769828796,
+      "learning_rate": 8e-05,
+      "loss": 1.6273,
+      "step": 8593
+    },
+    {
+      "epoch": 0.479041248606466,
+      "grad_norm": 0.5230170488357544,
+      "learning_rate": 8e-05,
+      "loss": 1.6368,
+      "step": 8594
+    },
+    {
+      "epoch": 0.4790969899665552,
+      "grad_norm": 0.5238328576087952,
+      "learning_rate": 8e-05,
+      "loss": 1.6142,
+      "step": 8595
+    },
+    {
+      "epoch": 0.4791527313266444,
+      "grad_norm": 0.5560547709465027,
+      "learning_rate": 8e-05,
+      "loss": 1.6971,
+      "step": 8596
+    },
+    {
+      "epoch": 0.47920847268673356,
+      "grad_norm": 0.5416411757469177,
+      "learning_rate": 8e-05,
+      "loss": 1.7525,
+      "step": 8597
+    },
+    {
+      "epoch": 0.47926421404682273,
+      "grad_norm": 0.5670372843742371,
+      "learning_rate": 8e-05,
+      "loss": 1.7209,
+      "step": 8598
+    },
+    {
+      "epoch": 0.4793199554069119,
+      "grad_norm": 0.559459924697876,
+      "learning_rate": 8e-05,
+      "loss": 1.7211,
+      "step": 8599
+    },
+    {
+      "epoch": 0.4793756967670011,
+      "grad_norm": 0.5402252078056335,
+      "learning_rate": 8e-05,
+      "loss": 1.713,
+      "step": 8600
+    },
+    {
+      "epoch": 0.4794314381270903,
+      "grad_norm": 0.5224754214286804,
+      "learning_rate": 8e-05,
+      "loss": 1.5075,
+      "step": 8601
+    },
+    {
+      "epoch": 0.4794871794871795,
+      "grad_norm": 0.5436088442802429,
+      "learning_rate": 8e-05,
+      "loss": 1.4942,
+      "step": 8602
+    },
+    {
+      "epoch": 0.4795429208472687,
+      "grad_norm": 0.7566401362419128,
+      "learning_rate": 8e-05,
+      "loss": 1.8953,
+      "step": 8603
+    },
+    {
+      "epoch": 0.47959866220735786,
+      "grad_norm": 0.5927697420120239,
+      "learning_rate": 8e-05,
+      "loss": 1.8098,
+      "step": 8604
+    },
+    {
+      "epoch": 0.47965440356744704,
+      "grad_norm": 0.5384635925292969,
+      "learning_rate": 8e-05,
+      "loss": 1.7361,
+      "step": 8605
+    },
+    {
+      "epoch": 0.4797101449275362,
+      "grad_norm": 0.527228832244873,
+      "learning_rate": 8e-05,
+      "loss": 1.5468,
+      "step": 8606
+    },
+    {
+      "epoch": 0.4797658862876254,
+      "grad_norm": 0.5674744248390198,
+      "learning_rate": 8e-05,
+      "loss": 1.5369,
+      "step": 8607
+    },
+    {
+      "epoch": 0.4798216276477146,
+      "grad_norm": 0.5231080055236816,
+      "learning_rate": 8e-05,
+      "loss": 1.622,
+      "step": 8608
+    },
+    {
+      "epoch": 0.4798773690078038,
+      "grad_norm": 0.7882029414176941,
+      "learning_rate": 8e-05,
+      "loss": 1.726,
+      "step": 8609
+    },
+    {
+      "epoch": 0.479933110367893,
+      "grad_norm": 0.5434726476669312,
+      "learning_rate": 8e-05,
+      "loss": 1.8462,
+      "step": 8610
+    },
+    {
+      "epoch": 0.47998885172798217,
+      "grad_norm": 0.5238932371139526,
+      "learning_rate": 8e-05,
+      "loss": 1.6484,
+      "step": 8611
+    },
+    {
+      "epoch": 0.48004459308807135,
+      "grad_norm": 0.5556615591049194,
+      "learning_rate": 8e-05,
+      "loss": 1.7166,
+      "step": 8612
+    },
+    {
+      "epoch": 0.4801003344481605,
+      "grad_norm": 0.5250589847564697,
+      "learning_rate": 8e-05,
+      "loss": 1.758,
+      "step": 8613
+    },
+    {
+      "epoch": 0.4801560758082497,
+      "grad_norm": 0.5367085337638855,
+      "learning_rate": 8e-05,
+      "loss": 1.6922,
+      "step": 8614
+    },
+    {
+      "epoch": 0.4802118171683389,
+      "grad_norm": 0.572113037109375,
+      "learning_rate": 8e-05,
+      "loss": 1.6577,
+      "step": 8615
+    },
+    {
+      "epoch": 0.4802675585284281,
+      "grad_norm": 0.5322007536888123,
+      "learning_rate": 8e-05,
+      "loss": 1.7393,
+      "step": 8616
+    },
+    {
+      "epoch": 0.4803232998885173,
+      "grad_norm": 0.5749253630638123,
+      "learning_rate": 8e-05,
+      "loss": 1.4501,
+      "step": 8617
+    },
+    {
+      "epoch": 0.4803790412486065,
+      "grad_norm": 0.5245643854141235,
+      "learning_rate": 8e-05,
+      "loss": 1.5195,
+      "step": 8618
+    },
+    {
+      "epoch": 0.48043478260869565,
+      "grad_norm": 0.5711345672607422,
+      "learning_rate": 8e-05,
+      "loss": 1.8074,
+      "step": 8619
+    },
+    {
+      "epoch": 0.48049052396878483,
+      "grad_norm": 0.5806153416633606,
+      "learning_rate": 8e-05,
+      "loss": 1.6689,
+      "step": 8620
+    },
+    {
+      "epoch": 0.480546265328874,
+      "grad_norm": 0.5262450575828552,
+      "learning_rate": 8e-05,
+      "loss": 1.621,
+      "step": 8621
+    },
+    {
+      "epoch": 0.4806020066889632,
+      "grad_norm": 0.5317466259002686,
+      "learning_rate": 8e-05,
+      "loss": 1.7104,
+      "step": 8622
+    },
+    {
+      "epoch": 0.4806577480490524,
+      "grad_norm": 0.5976824164390564,
+      "learning_rate": 8e-05,
+      "loss": 1.6694,
+      "step": 8623
+    },
+    {
+      "epoch": 0.4807134894091416,
+      "grad_norm": 0.5373111963272095,
+      "learning_rate": 8e-05,
+      "loss": 1.6995,
+      "step": 8624
+    },
+    {
+      "epoch": 0.4807692307692308,
+      "grad_norm": 0.5650762915611267,
+      "learning_rate": 8e-05,
+      "loss": 1.7193,
+      "step": 8625
+    },
+    {
+      "epoch": 0.48082497212931996,
+      "grad_norm": 0.4892181158065796,
+      "learning_rate": 8e-05,
+      "loss": 1.5967,
+      "step": 8626
+    },
+    {
+      "epoch": 0.48088071348940914,
+      "grad_norm": 0.4929823577404022,
+      "learning_rate": 8e-05,
+      "loss": 1.5012,
+      "step": 8627
+    },
+    {
+      "epoch": 0.4809364548494983,
+      "grad_norm": 0.5495728850364685,
+      "learning_rate": 8e-05,
+      "loss": 1.7687,
+      "step": 8628
+    },
+    {
+      "epoch": 0.4809921962095875,
+      "grad_norm": 0.5166272521018982,
+      "learning_rate": 8e-05,
+      "loss": 1.6092,
+      "step": 8629
+    },
+    {
+      "epoch": 0.4810479375696767,
+      "grad_norm": 0.5548287630081177,
+      "learning_rate": 8e-05,
+      "loss": 1.5799,
+      "step": 8630
+    },
+    {
+      "epoch": 0.4811036789297659,
+      "grad_norm": 0.5389736890792847,
+      "learning_rate": 8e-05,
+      "loss": 1.5249,
+      "step": 8631
+    },
+    {
+      "epoch": 0.4811594202898551,
+      "grad_norm": 0.5570147633552551,
+      "learning_rate": 8e-05,
+      "loss": 1.7399,
+      "step": 8632
+    },
+    {
+      "epoch": 0.48121516164994427,
+      "grad_norm": 0.528630793094635,
+      "learning_rate": 8e-05,
+      "loss": 1.5725,
+      "step": 8633
+    },
+    {
+      "epoch": 0.48127090301003345,
+      "grad_norm": 0.5272868275642395,
+      "learning_rate": 8e-05,
+      "loss": 1.6709,
+      "step": 8634
+    },
+    {
+      "epoch": 0.4813266443701226,
+      "grad_norm": 0.5915762782096863,
+      "learning_rate": 8e-05,
+      "loss": 1.9041,
+      "step": 8635
+    },
+    {
+      "epoch": 0.4813823857302118,
+      "grad_norm": 0.5859066247940063,
+      "learning_rate": 8e-05,
+      "loss": 1.942,
+      "step": 8636
+    },
+    {
+      "epoch": 0.481438127090301,
+      "grad_norm": 0.5061094760894775,
+      "learning_rate": 8e-05,
+      "loss": 1.5464,
+      "step": 8637
+    },
+    {
+      "epoch": 0.4814938684503902,
+      "grad_norm": 0.5594993233680725,
+      "learning_rate": 8e-05,
+      "loss": 1.6924,
+      "step": 8638
+    },
+    {
+      "epoch": 0.4815496098104794,
+      "grad_norm": 0.5258274674415588,
+      "learning_rate": 8e-05,
+      "loss": 1.5865,
+      "step": 8639
+    },
+    {
+      "epoch": 0.4816053511705686,
+      "grad_norm": 0.5139769315719604,
+      "learning_rate": 8e-05,
+      "loss": 1.6557,
+      "step": 8640
+    },
+    {
+      "epoch": 0.48166109253065775,
+      "grad_norm": 0.5273594856262207,
+      "learning_rate": 8e-05,
+      "loss": 1.6087,
+      "step": 8641
+    },
+    {
+      "epoch": 0.48171683389074693,
+      "grad_norm": 0.6034706830978394,
+      "learning_rate": 8e-05,
+      "loss": 1.7294,
+      "step": 8642
+    },
+    {
+      "epoch": 0.4817725752508361,
+      "grad_norm": 0.5553348660469055,
+      "learning_rate": 8e-05,
+      "loss": 1.4567,
+      "step": 8643
+    },
+    {
+      "epoch": 0.4818283166109253,
+      "grad_norm": 0.5267349481582642,
+      "learning_rate": 8e-05,
+      "loss": 1.517,
+      "step": 8644
+    },
+    {
+      "epoch": 0.48188405797101447,
+      "grad_norm": 0.5282378196716309,
+      "learning_rate": 8e-05,
+      "loss": 1.5921,
+      "step": 8645
+    },
+    {
+      "epoch": 0.4819397993311037,
+      "grad_norm": 0.526695728302002,
+      "learning_rate": 8e-05,
+      "loss": 1.7406,
+      "step": 8646
+    },
+    {
+      "epoch": 0.4819955406911929,
+      "grad_norm": 0.5395731925964355,
+      "learning_rate": 8e-05,
+      "loss": 1.4693,
+      "step": 8647
+    },
+    {
+      "epoch": 0.48205128205128206,
+      "grad_norm": 0.5188526511192322,
+      "learning_rate": 8e-05,
+      "loss": 1.4089,
+      "step": 8648
+    },
+    {
+      "epoch": 0.48210702341137124,
+      "grad_norm": 0.5890273451805115,
+      "learning_rate": 8e-05,
+      "loss": 1.765,
+      "step": 8649
+    },
+    {
+      "epoch": 0.4821627647714604,
+      "grad_norm": 0.5989070534706116,
+      "learning_rate": 8e-05,
+      "loss": 1.8159,
+      "step": 8650
+    },
+    {
+      "epoch": 0.4822185061315496,
+      "grad_norm": 0.5372903943061829,
+      "learning_rate": 8e-05,
+      "loss": 1.7536,
+      "step": 8651
+    },
+    {
+      "epoch": 0.4822742474916388,
+      "grad_norm": 0.5370186567306519,
+      "learning_rate": 8e-05,
+      "loss": 1.5296,
+      "step": 8652
+    },
+    {
+      "epoch": 0.482329988851728,
+      "grad_norm": 0.5499247908592224,
+      "learning_rate": 8e-05,
+      "loss": 1.5902,
+      "step": 8653
+    },
+    {
+      "epoch": 0.4823857302118172,
+      "grad_norm": 0.5799142122268677,
+      "learning_rate": 8e-05,
+      "loss": 1.9175,
+      "step": 8654
+    },
+    {
+      "epoch": 0.48244147157190637,
+      "grad_norm": 0.562670886516571,
+      "learning_rate": 8e-05,
+      "loss": 1.7646,
+      "step": 8655
+    },
+    {
+      "epoch": 0.48249721293199554,
+      "grad_norm": 0.5562814474105835,
+      "learning_rate": 8e-05,
+      "loss": 1.7472,
+      "step": 8656
+    },
+    {
+      "epoch": 0.4825529542920847,
+      "grad_norm": 0.5497145652770996,
+      "learning_rate": 8e-05,
+      "loss": 1.7556,
+      "step": 8657
+    },
+    {
+      "epoch": 0.4826086956521739,
+      "grad_norm": 0.5814297795295715,
+      "learning_rate": 8e-05,
+      "loss": 1.6638,
+      "step": 8658
+    },
+    {
+      "epoch": 0.4826644370122631,
+      "grad_norm": 0.5385329723358154,
+      "learning_rate": 8e-05,
+      "loss": 1.6465,
+      "step": 8659
+    },
+    {
+      "epoch": 0.48272017837235226,
+      "grad_norm": 0.533889651298523,
+      "learning_rate": 8e-05,
+      "loss": 1.6298,
+      "step": 8660
+    },
+    {
+      "epoch": 0.4827759197324415,
+      "grad_norm": 0.5587199330329895,
+      "learning_rate": 8e-05,
+      "loss": 1.7957,
+      "step": 8661
+    },
+    {
+      "epoch": 0.4828316610925307,
+      "grad_norm": 0.5244656205177307,
+      "learning_rate": 8e-05,
+      "loss": 1.4407,
+      "step": 8662
+    },
+    {
+      "epoch": 0.48288740245261985,
+      "grad_norm": 0.5379255414009094,
+      "learning_rate": 8e-05,
+      "loss": 1.5525,
+      "step": 8663
+    },
+    {
+      "epoch": 0.48294314381270903,
+      "grad_norm": 0.5391064882278442,
+      "learning_rate": 8e-05,
+      "loss": 1.6002,
+      "step": 8664
+    },
+    {
+      "epoch": 0.4829988851727982,
+      "grad_norm": 0.5106524229049683,
+      "learning_rate": 8e-05,
+      "loss": 1.6711,
+      "step": 8665
+    },
+    {
+      "epoch": 0.4830546265328874,
+      "grad_norm": 0.5733168721199036,
+      "learning_rate": 8e-05,
+      "loss": 1.6508,
+      "step": 8666
+    },
+    {
+      "epoch": 0.48311036789297657,
+      "grad_norm": 0.5390850901603699,
+      "learning_rate": 8e-05,
+      "loss": 1.5811,
+      "step": 8667
+    },
+    {
+      "epoch": 0.4831661092530658,
+      "grad_norm": 0.5093480944633484,
+      "learning_rate": 8e-05,
+      "loss": 1.424,
+      "step": 8668
+    },
+    {
+      "epoch": 0.483221850613155,
+      "grad_norm": 0.5804044008255005,
+      "learning_rate": 8e-05,
+      "loss": 2.0014,
+      "step": 8669
+    },
+    {
+      "epoch": 0.48327759197324416,
+      "grad_norm": 0.6597045063972473,
+      "learning_rate": 8e-05,
+      "loss": 2.1137,
+      "step": 8670
+    },
+    {
+      "epoch": 0.48333333333333334,
+      "grad_norm": 0.5206351280212402,
+      "learning_rate": 8e-05,
+      "loss": 1.5059,
+      "step": 8671
+    },
+    {
+      "epoch": 0.4833890746934225,
+      "grad_norm": 0.5995643138885498,
+      "learning_rate": 8e-05,
+      "loss": 1.505,
+      "step": 8672
+    },
+    {
+      "epoch": 0.4834448160535117,
+      "grad_norm": 0.5529085993766785,
+      "learning_rate": 8e-05,
+      "loss": 1.5522,
+      "step": 8673
+    },
+    {
+      "epoch": 0.4835005574136009,
+      "grad_norm": 0.5057050585746765,
+      "learning_rate": 8e-05,
+      "loss": 1.5109,
+      "step": 8674
+    },
+    {
+      "epoch": 0.48355629877369005,
+      "grad_norm": 0.5466620922088623,
+      "learning_rate": 8e-05,
+      "loss": 1.7737,
+      "step": 8675
+    },
+    {
+      "epoch": 0.4836120401337793,
+      "grad_norm": 0.5643208026885986,
+      "learning_rate": 8e-05,
+      "loss": 1.5089,
+      "step": 8676
+    },
+    {
+      "epoch": 0.48366778149386846,
+      "grad_norm": 0.5501327514648438,
+      "learning_rate": 8e-05,
+      "loss": 1.6649,
+      "step": 8677
+    },
+    {
+      "epoch": 0.48372352285395764,
+      "grad_norm": 0.5182400941848755,
+      "learning_rate": 8e-05,
+      "loss": 1.6111,
+      "step": 8678
+    },
+    {
+      "epoch": 0.4837792642140468,
+      "grad_norm": 0.5956326723098755,
+      "learning_rate": 8e-05,
+      "loss": 1.6881,
+      "step": 8679
+    },
+    {
+      "epoch": 0.483835005574136,
+      "grad_norm": 0.5279495716094971,
+      "learning_rate": 8e-05,
+      "loss": 1.542,
+      "step": 8680
+    },
+    {
+      "epoch": 0.4838907469342252,
+      "grad_norm": 0.537891149520874,
+      "learning_rate": 8e-05,
+      "loss": 1.651,
+      "step": 8681
+    },
+    {
+      "epoch": 0.48394648829431436,
+      "grad_norm": 0.5734157562255859,
+      "learning_rate": 8e-05,
+      "loss": 1.6414,
+      "step": 8682
+    },
+    {
+      "epoch": 0.4840022296544036,
+      "grad_norm": 0.5524048209190369,
+      "learning_rate": 8e-05,
+      "loss": 1.5915,
+      "step": 8683
+    },
+    {
+      "epoch": 0.48405797101449277,
+      "grad_norm": 0.5250129103660583,
+      "learning_rate": 8e-05,
+      "loss": 1.7873,
+      "step": 8684
+    },
+    {
+      "epoch": 0.48411371237458195,
+      "grad_norm": 0.4993424415588379,
+      "learning_rate": 8e-05,
+      "loss": 1.5449,
+      "step": 8685
+    },
+    {
+      "epoch": 0.48416945373467113,
+      "grad_norm": 0.5354702472686768,
+      "learning_rate": 8e-05,
+      "loss": 1.6186,
+      "step": 8686
+    },
+    {
+      "epoch": 0.4842251950947603,
+      "grad_norm": 0.7202904224395752,
+      "learning_rate": 8e-05,
+      "loss": 1.5289,
+      "step": 8687
+    },
+    {
+      "epoch": 0.4842809364548495,
+      "grad_norm": 0.5833171010017395,
+      "learning_rate": 8e-05,
+      "loss": 1.7669,
+      "step": 8688
+    },
+    {
+      "epoch": 0.48433667781493867,
+      "grad_norm": 0.5454303026199341,
+      "learning_rate": 8e-05,
+      "loss": 1.5781,
+      "step": 8689
+    },
+    {
+      "epoch": 0.48439241917502784,
+      "grad_norm": 0.5747190117835999,
+      "learning_rate": 8e-05,
+      "loss": 1.8008,
+      "step": 8690
+    },
+    {
+      "epoch": 0.4844481605351171,
+      "grad_norm": 0.5227949023246765,
+      "learning_rate": 8e-05,
+      "loss": 1.7343,
+      "step": 8691
+    },
+    {
+      "epoch": 0.48450390189520626,
+      "grad_norm": 0.5211941599845886,
+      "learning_rate": 8e-05,
+      "loss": 1.773,
+      "step": 8692
+    },
+    {
+      "epoch": 0.48455964325529544,
+      "grad_norm": 0.511934220790863,
+      "learning_rate": 8e-05,
+      "loss": 1.6041,
+      "step": 8693
+    },
+    {
+      "epoch": 0.4846153846153846,
+      "grad_norm": 0.5146392583847046,
+      "learning_rate": 8e-05,
+      "loss": 1.5809,
+      "step": 8694
+    },
+    {
+      "epoch": 0.4846711259754738,
+      "grad_norm": 0.5503633618354797,
+      "learning_rate": 8e-05,
+      "loss": 1.6425,
+      "step": 8695
+    },
+    {
+      "epoch": 0.48472686733556297,
+      "grad_norm": 0.6414245963096619,
+      "learning_rate": 8e-05,
+      "loss": 1.6811,
+      "step": 8696
+    },
+    {
+      "epoch": 0.48478260869565215,
+      "grad_norm": 0.5404537916183472,
+      "learning_rate": 8e-05,
+      "loss": 1.7578,
+      "step": 8697
+    },
+    {
+      "epoch": 0.4848383500557414,
+      "grad_norm": 0.5513060688972473,
+      "learning_rate": 8e-05,
+      "loss": 1.7655,
+      "step": 8698
+    },
+    {
+      "epoch": 0.48489409141583056,
+      "grad_norm": 0.6144512891769409,
+      "learning_rate": 8e-05,
+      "loss": 1.7569,
+      "step": 8699
+    },
+    {
+      "epoch": 0.48494983277591974,
+      "grad_norm": 0.5189093947410583,
+      "learning_rate": 8e-05,
+      "loss": 1.7075,
+      "step": 8700
+    },
+    {
+      "epoch": 0.4850055741360089,
+      "grad_norm": 0.5028544664382935,
+      "learning_rate": 8e-05,
+      "loss": 1.441,
+      "step": 8701
+    },
+    {
+      "epoch": 0.4850613154960981,
+      "grad_norm": 0.5841237902641296,
+      "learning_rate": 8e-05,
+      "loss": 1.7716,
+      "step": 8702
+    },
+    {
+      "epoch": 0.4851170568561873,
+      "grad_norm": 0.5114294290542603,
+      "learning_rate": 8e-05,
+      "loss": 1.4676,
+      "step": 8703
+    },
+    {
+      "epoch": 0.48517279821627646,
+      "grad_norm": 0.597727358341217,
+      "learning_rate": 8e-05,
+      "loss": 1.5788,
+      "step": 8704
+    },
+    {
+      "epoch": 0.48522853957636564,
+      "grad_norm": 0.5356034636497498,
+      "learning_rate": 8e-05,
+      "loss": 1.6905,
+      "step": 8705
+    },
+    {
+      "epoch": 0.48528428093645487,
+      "grad_norm": 0.5260893702507019,
+      "learning_rate": 8e-05,
+      "loss": 1.5722,
+      "step": 8706
+    },
+    {
+      "epoch": 0.48534002229654405,
+      "grad_norm": 0.5867304801940918,
+      "learning_rate": 8e-05,
+      "loss": 1.8526,
+      "step": 8707
+    },
+    {
+      "epoch": 0.48539576365663323,
+      "grad_norm": 0.5377045273780823,
+      "learning_rate": 8e-05,
+      "loss": 1.8251,
+      "step": 8708
+    },
+    {
+      "epoch": 0.4854515050167224,
+      "grad_norm": 0.5542226433753967,
+      "learning_rate": 8e-05,
+      "loss": 1.9328,
+      "step": 8709
+    },
+    {
+      "epoch": 0.4855072463768116,
+      "grad_norm": 0.6472811102867126,
+      "learning_rate": 8e-05,
+      "loss": 1.7699,
+      "step": 8710
+    },
+    {
+      "epoch": 0.48556298773690076,
+      "grad_norm": 0.5780431032180786,
+      "learning_rate": 8e-05,
+      "loss": 1.5492,
+      "step": 8711
+    },
+    {
+      "epoch": 0.48561872909698994,
+      "grad_norm": 0.6103213429450989,
+      "learning_rate": 8e-05,
+      "loss": 1.7291,
+      "step": 8712
+    },
+    {
+      "epoch": 0.4856744704570792,
+      "grad_norm": 0.4791732430458069,
+      "learning_rate": 8e-05,
+      "loss": 1.4316,
+      "step": 8713
+    },
+    {
+      "epoch": 0.48573021181716836,
+      "grad_norm": 0.5246248245239258,
+      "learning_rate": 8e-05,
+      "loss": 1.5101,
+      "step": 8714
+    },
+    {
+      "epoch": 0.48578595317725753,
+      "grad_norm": 0.5121028423309326,
+      "learning_rate": 8e-05,
+      "loss": 1.6613,
+      "step": 8715
+    },
+    {
+      "epoch": 0.4858416945373467,
+      "grad_norm": 0.522429347038269,
+      "learning_rate": 8e-05,
+      "loss": 1.4679,
+      "step": 8716
+    },
+    {
+      "epoch": 0.4858974358974359,
+      "grad_norm": 0.5425698757171631,
+      "learning_rate": 8e-05,
+      "loss": 1.6178,
+      "step": 8717
+    },
+    {
+      "epoch": 0.48595317725752507,
+      "grad_norm": 0.5327525734901428,
+      "learning_rate": 8e-05,
+      "loss": 1.6725,
+      "step": 8718
+    },
+    {
+      "epoch": 0.48600891861761425,
+      "grad_norm": 0.5315846800804138,
+      "learning_rate": 8e-05,
+      "loss": 1.4996,
+      "step": 8719
+    },
+    {
+      "epoch": 0.48606465997770343,
+      "grad_norm": 0.6056599617004395,
+      "learning_rate": 8e-05,
+      "loss": 1.5545,
+      "step": 8720
+    },
+    {
+      "epoch": 0.48612040133779266,
+      "grad_norm": 0.535976767539978,
+      "learning_rate": 8e-05,
+      "loss": 1.6458,
+      "step": 8721
+    },
+    {
+      "epoch": 0.48617614269788184,
+      "grad_norm": 0.5951147079467773,
+      "learning_rate": 8e-05,
+      "loss": 1.8229,
+      "step": 8722
+    },
+    {
+      "epoch": 0.486231884057971,
+      "grad_norm": 0.5291656255722046,
+      "learning_rate": 8e-05,
+      "loss": 1.6449,
+      "step": 8723
+    },
+    {
+      "epoch": 0.4862876254180602,
+      "grad_norm": 0.4991705119609833,
+      "learning_rate": 8e-05,
+      "loss": 1.3861,
+      "step": 8724
+    },
+    {
+      "epoch": 0.4863433667781494,
+      "grad_norm": 0.538762629032135,
+      "learning_rate": 8e-05,
+      "loss": 1.802,
+      "step": 8725
+    },
+    {
+      "epoch": 0.48639910813823856,
+      "grad_norm": 0.5288955569267273,
+      "learning_rate": 8e-05,
+      "loss": 1.5841,
+      "step": 8726
+    },
+    {
+      "epoch": 0.48645484949832773,
+      "grad_norm": 0.5434188842773438,
+      "learning_rate": 8e-05,
+      "loss": 1.6843,
+      "step": 8727
+    },
+    {
+      "epoch": 0.48651059085841697,
+      "grad_norm": 0.5369993448257446,
+      "learning_rate": 8e-05,
+      "loss": 1.7506,
+      "step": 8728
+    },
+    {
+      "epoch": 0.48656633221850615,
+      "grad_norm": 0.5895953178405762,
+      "learning_rate": 8e-05,
+      "loss": 1.6276,
+      "step": 8729
+    },
+    {
+      "epoch": 0.4866220735785953,
+      "grad_norm": 0.5380129814147949,
+      "learning_rate": 8e-05,
+      "loss": 1.7047,
+      "step": 8730
+    },
+    {
+      "epoch": 0.4866778149386845,
+      "grad_norm": 0.5475728511810303,
+      "learning_rate": 8e-05,
+      "loss": 1.9109,
+      "step": 8731
+    },
+    {
+      "epoch": 0.4867335562987737,
+      "grad_norm": 0.5191141963005066,
+      "learning_rate": 8e-05,
+      "loss": 1.5978,
+      "step": 8732
+    },
+    {
+      "epoch": 0.48678929765886286,
+      "grad_norm": 0.5969076752662659,
+      "learning_rate": 8e-05,
+      "loss": 1.8285,
+      "step": 8733
+    },
+    {
+      "epoch": 0.48684503901895204,
+      "grad_norm": 0.49988871812820435,
+      "learning_rate": 8e-05,
+      "loss": 1.4702,
+      "step": 8734
+    },
+    {
+      "epoch": 0.4869007803790413,
+      "grad_norm": 0.5074156522750854,
+      "learning_rate": 8e-05,
+      "loss": 1.5778,
+      "step": 8735
+    },
+    {
+      "epoch": 0.48695652173913045,
+      "grad_norm": 0.5323285460472107,
+      "learning_rate": 8e-05,
+      "loss": 1.8674,
+      "step": 8736
+    },
+    {
+      "epoch": 0.48701226309921963,
+      "grad_norm": 0.5003831386566162,
+      "learning_rate": 8e-05,
+      "loss": 1.3662,
+      "step": 8737
+    },
+    {
+      "epoch": 0.4870680044593088,
+      "grad_norm": 0.5432720184326172,
+      "learning_rate": 8e-05,
+      "loss": 1.6234,
+      "step": 8738
+    },
+    {
+      "epoch": 0.487123745819398,
+      "grad_norm": 0.5797169804573059,
+      "learning_rate": 8e-05,
+      "loss": 1.8015,
+      "step": 8739
+    },
+    {
+      "epoch": 0.48717948717948717,
+      "grad_norm": 0.5347649455070496,
+      "learning_rate": 8e-05,
+      "loss": 1.7881,
+      "step": 8740
+    },
+    {
+      "epoch": 0.48723522853957635,
+      "grad_norm": 0.5669839978218079,
+      "learning_rate": 8e-05,
+      "loss": 1.6679,
+      "step": 8741
+    },
+    {
+      "epoch": 0.4872909698996655,
+      "grad_norm": 0.54695063829422,
+      "learning_rate": 8e-05,
+      "loss": 1.7627,
+      "step": 8742
+    },
+    {
+      "epoch": 0.48734671125975476,
+      "grad_norm": 0.6649922132492065,
+      "learning_rate": 8e-05,
+      "loss": 2.0592,
+      "step": 8743
+    },
+    {
+      "epoch": 0.48740245261984394,
+      "grad_norm": 0.5175009369850159,
+      "learning_rate": 8e-05,
+      "loss": 1.4855,
+      "step": 8744
+    },
+    {
+      "epoch": 0.4874581939799331,
+      "grad_norm": 0.5442166924476624,
+      "learning_rate": 8e-05,
+      "loss": 1.4434,
+      "step": 8745
+    },
+    {
+      "epoch": 0.4875139353400223,
+      "grad_norm": 0.5374280214309692,
+      "learning_rate": 8e-05,
+      "loss": 1.4264,
+      "step": 8746
+    },
+    {
+      "epoch": 0.4875696767001115,
+      "grad_norm": 0.5635949373245239,
+      "learning_rate": 8e-05,
+      "loss": 1.7391,
+      "step": 8747
+    },
+    {
+      "epoch": 0.48762541806020065,
+      "grad_norm": 0.5555373430252075,
+      "learning_rate": 8e-05,
+      "loss": 1.5628,
+      "step": 8748
+    },
+    {
+      "epoch": 0.48768115942028983,
+      "grad_norm": 0.5785765051841736,
+      "learning_rate": 8e-05,
+      "loss": 1.6793,
+      "step": 8749
+    },
+    {
+      "epoch": 0.48773690078037907,
+      "grad_norm": 0.4377272427082062,
+      "learning_rate": 8e-05,
+      "loss": 0.8688,
+      "step": 8750
+    },
+    {
+      "epoch": 0.48779264214046825,
+      "grad_norm": 0.49414724111557007,
+      "learning_rate": 8e-05,
+      "loss": 1.4305,
+      "step": 8751
+    },
+    {
+      "epoch": 0.4878483835005574,
+      "grad_norm": 0.5328716039657593,
+      "learning_rate": 8e-05,
+      "loss": 1.458,
+      "step": 8752
+    },
+    {
+      "epoch": 0.4879041248606466,
+      "grad_norm": 0.5761605501174927,
+      "learning_rate": 8e-05,
+      "loss": 1.8068,
+      "step": 8753
+    },
+    {
+      "epoch": 0.4879598662207358,
+      "grad_norm": 0.47436168789863586,
+      "learning_rate": 8e-05,
+      "loss": 1.1729,
+      "step": 8754
+    },
+    {
+      "epoch": 0.48801560758082496,
+      "grad_norm": 0.5351918339729309,
+      "learning_rate": 8e-05,
+      "loss": 1.7398,
+      "step": 8755
+    },
+    {
+      "epoch": 0.48807134894091414,
+      "grad_norm": 0.6001772880554199,
+      "learning_rate": 8e-05,
+      "loss": 1.7006,
+      "step": 8756
+    },
+    {
+      "epoch": 0.4881270903010033,
+      "grad_norm": 0.5331566333770752,
+      "learning_rate": 8e-05,
+      "loss": 1.3804,
+      "step": 8757
+    },
+    {
+      "epoch": 0.48818283166109255,
+      "grad_norm": 0.5804986953735352,
+      "learning_rate": 8e-05,
+      "loss": 1.7321,
+      "step": 8758
+    },
+    {
+      "epoch": 0.48823857302118173,
+      "grad_norm": 0.6281266212463379,
+      "learning_rate": 8e-05,
+      "loss": 1.5619,
+      "step": 8759
+    },
+    {
+      "epoch": 0.4882943143812709,
+      "grad_norm": 0.5269529819488525,
+      "learning_rate": 8e-05,
+      "loss": 1.5927,
+      "step": 8760
+    },
+    {
+      "epoch": 0.4883500557413601,
+      "grad_norm": 0.628817081451416,
+      "learning_rate": 8e-05,
+      "loss": 1.8212,
+      "step": 8761
+    },
+    {
+      "epoch": 0.48840579710144927,
+      "grad_norm": 0.5161125659942627,
+      "learning_rate": 8e-05,
+      "loss": 1.5617,
+      "step": 8762
+    },
+    {
+      "epoch": 0.48846153846153845,
+      "grad_norm": 0.5620384812355042,
+      "learning_rate": 8e-05,
+      "loss": 1.794,
+      "step": 8763
+    },
+    {
+      "epoch": 0.4885172798216276,
+      "grad_norm": 0.5161763429641724,
+      "learning_rate": 8e-05,
+      "loss": 1.4774,
+      "step": 8764
+    },
+    {
+      "epoch": 0.48857302118171686,
+      "grad_norm": 0.5568657517433167,
+      "learning_rate": 8e-05,
+      "loss": 1.7074,
+      "step": 8765
+    },
+    {
+      "epoch": 0.48862876254180604,
+      "grad_norm": 0.5740475058555603,
+      "learning_rate": 8e-05,
+      "loss": 1.6313,
+      "step": 8766
+    },
+    {
+      "epoch": 0.4886845039018952,
+      "grad_norm": 0.5516563653945923,
+      "learning_rate": 8e-05,
+      "loss": 1.6862,
+      "step": 8767
+    },
+    {
+      "epoch": 0.4887402452619844,
+      "grad_norm": 0.5552970170974731,
+      "learning_rate": 8e-05,
+      "loss": 1.7847,
+      "step": 8768
+    },
+    {
+      "epoch": 0.4887959866220736,
+      "grad_norm": 0.6541142463684082,
+      "learning_rate": 8e-05,
+      "loss": 1.9267,
+      "step": 8769
+    },
+    {
+      "epoch": 0.48885172798216275,
+      "grad_norm": 0.5554009675979614,
+      "learning_rate": 8e-05,
+      "loss": 1.8234,
+      "step": 8770
+    },
+    {
+      "epoch": 0.48890746934225193,
+      "grad_norm": 0.5801352858543396,
+      "learning_rate": 8e-05,
+      "loss": 1.7553,
+      "step": 8771
+    },
+    {
+      "epoch": 0.4889632107023411,
+      "grad_norm": 0.5918492674827576,
+      "learning_rate": 8e-05,
+      "loss": 1.5996,
+      "step": 8772
+    },
+    {
+      "epoch": 0.48901895206243035,
+      "grad_norm": 0.5583415627479553,
+      "learning_rate": 8e-05,
+      "loss": 1.6549,
+      "step": 8773
+    },
+    {
+      "epoch": 0.4890746934225195,
+      "grad_norm": 0.5346972346305847,
+      "learning_rate": 8e-05,
+      "loss": 1.867,
+      "step": 8774
+    },
+    {
+      "epoch": 0.4891304347826087,
+      "grad_norm": 0.5730275511741638,
+      "learning_rate": 8e-05,
+      "loss": 1.8839,
+      "step": 8775
+    },
+    {
+      "epoch": 0.4891861761426979,
+      "grad_norm": 0.5031627416610718,
+      "learning_rate": 8e-05,
+      "loss": 1.5669,
+      "step": 8776
+    },
+    {
+      "epoch": 0.48924191750278706,
+      "grad_norm": 0.5561438798904419,
+      "learning_rate": 8e-05,
+      "loss": 1.804,
+      "step": 8777
+    },
+    {
+      "epoch": 0.48929765886287624,
+      "grad_norm": 0.5484711527824402,
+      "learning_rate": 8e-05,
+      "loss": 1.7645,
+      "step": 8778
+    },
+    {
+      "epoch": 0.4893534002229654,
+      "grad_norm": 0.5624884963035583,
+      "learning_rate": 8e-05,
+      "loss": 1.551,
+      "step": 8779
+    },
+    {
+      "epoch": 0.48940914158305465,
+      "grad_norm": 0.5877845883369446,
+      "learning_rate": 8e-05,
+      "loss": 1.9774,
+      "step": 8780
+    },
+    {
+      "epoch": 0.48946488294314383,
+      "grad_norm": 0.557813823223114,
+      "learning_rate": 8e-05,
+      "loss": 1.7737,
+      "step": 8781
+    },
+    {
+      "epoch": 0.489520624303233,
+      "grad_norm": 0.5099141597747803,
+      "learning_rate": 8e-05,
+      "loss": 1.5899,
+      "step": 8782
+    },
+    {
+      "epoch": 0.4895763656633222,
+      "grad_norm": 0.5315130949020386,
+      "learning_rate": 8e-05,
+      "loss": 1.5977,
+      "step": 8783
+    },
+    {
+      "epoch": 0.48963210702341137,
+      "grad_norm": 0.5222629904747009,
+      "learning_rate": 8e-05,
+      "loss": 1.7737,
+      "step": 8784
+    },
+    {
+      "epoch": 0.48968784838350055,
+      "grad_norm": 0.5800378322601318,
+      "learning_rate": 8e-05,
+      "loss": 1.8103,
+      "step": 8785
+    },
+    {
+      "epoch": 0.4897435897435897,
+      "grad_norm": 0.5630063414573669,
+      "learning_rate": 8e-05,
+      "loss": 1.6663,
+      "step": 8786
+    },
+    {
+      "epoch": 0.4897993311036789,
+      "grad_norm": 0.5052672028541565,
+      "learning_rate": 8e-05,
+      "loss": 1.412,
+      "step": 8787
+    },
+    {
+      "epoch": 0.48985507246376814,
+      "grad_norm": 0.6610358953475952,
+      "learning_rate": 8e-05,
+      "loss": 2.2178,
+      "step": 8788
+    },
+    {
+      "epoch": 0.4899108138238573,
+      "grad_norm": 0.5247467756271362,
+      "learning_rate": 8e-05,
+      "loss": 1.6431,
+      "step": 8789
+    },
+    {
+      "epoch": 0.4899665551839465,
+      "grad_norm": 0.5510267019271851,
+      "learning_rate": 8e-05,
+      "loss": 1.7086,
+      "step": 8790
+    },
+    {
+      "epoch": 0.4900222965440357,
+      "grad_norm": 0.5938252210617065,
+      "learning_rate": 8e-05,
+      "loss": 1.7299,
+      "step": 8791
+    },
+    {
+      "epoch": 0.49007803790412485,
+      "grad_norm": 0.6057933568954468,
+      "learning_rate": 8e-05,
+      "loss": 1.9044,
+      "step": 8792
+    },
+    {
+      "epoch": 0.49013377926421403,
+      "grad_norm": 0.5210702419281006,
+      "learning_rate": 8e-05,
+      "loss": 1.6161,
+      "step": 8793
+    },
+    {
+      "epoch": 0.4901895206243032,
+      "grad_norm": 0.5469611883163452,
+      "learning_rate": 8e-05,
+      "loss": 1.5888,
+      "step": 8794
+    },
+    {
+      "epoch": 0.49024526198439244,
+      "grad_norm": 0.5292215347290039,
+      "learning_rate": 8e-05,
+      "loss": 1.6371,
+      "step": 8795
+    },
+    {
+      "epoch": 0.4903010033444816,
+      "grad_norm": 0.5211275219917297,
+      "learning_rate": 8e-05,
+      "loss": 1.5327,
+      "step": 8796
+    },
+    {
+      "epoch": 0.4903567447045708,
+      "grad_norm": 0.5240673422813416,
+      "learning_rate": 8e-05,
+      "loss": 1.6971,
+      "step": 8797
+    },
+    {
+      "epoch": 0.49041248606466,
+      "grad_norm": 0.4981025755405426,
+      "learning_rate": 8e-05,
+      "loss": 1.4021,
+      "step": 8798
+    },
+    {
+      "epoch": 0.49046822742474916,
+      "grad_norm": 0.5216448307037354,
+      "learning_rate": 8e-05,
+      "loss": 1.4958,
+      "step": 8799
+    },
+    {
+      "epoch": 0.49052396878483834,
+      "grad_norm": 0.531890332698822,
+      "learning_rate": 8e-05,
+      "loss": 1.6907,
+      "step": 8800
+    },
+    {
+      "epoch": 0.4905797101449275,
+      "grad_norm": 0.5103880763053894,
+      "learning_rate": 8e-05,
+      "loss": 1.3802,
+      "step": 8801
+    },
+    {
+      "epoch": 0.4906354515050167,
+      "grad_norm": 0.5482351183891296,
+      "learning_rate": 8e-05,
+      "loss": 1.5857,
+      "step": 8802
+    },
+    {
+      "epoch": 0.49069119286510593,
+      "grad_norm": 0.6087172031402588,
+      "learning_rate": 8e-05,
+      "loss": 1.4996,
+      "step": 8803
+    },
+    {
+      "epoch": 0.4907469342251951,
+      "grad_norm": 0.5360320210456848,
+      "learning_rate": 8e-05,
+      "loss": 1.6422,
+      "step": 8804
+    },
+    {
+      "epoch": 0.4908026755852843,
+      "grad_norm": 0.557442843914032,
+      "learning_rate": 8e-05,
+      "loss": 1.7739,
+      "step": 8805
+    },
+    {
+      "epoch": 0.49085841694537347,
+      "grad_norm": 0.5476164817810059,
+      "learning_rate": 8e-05,
+      "loss": 1.6352,
+      "step": 8806
+    },
+    {
+      "epoch": 0.49091415830546264,
+      "grad_norm": 0.5663220286369324,
+      "learning_rate": 8e-05,
+      "loss": 1.7057,
+      "step": 8807
+    },
+    {
+      "epoch": 0.4909698996655518,
+      "grad_norm": 0.5578323602676392,
+      "learning_rate": 8e-05,
+      "loss": 1.4855,
+      "step": 8808
+    },
+    {
+      "epoch": 0.491025641025641,
+      "grad_norm": 0.7541027069091797,
+      "learning_rate": 8e-05,
+      "loss": 1.7219,
+      "step": 8809
+    },
+    {
+      "epoch": 0.49108138238573024,
+      "grad_norm": 0.5288164615631104,
+      "learning_rate": 8e-05,
+      "loss": 1.4917,
+      "step": 8810
+    },
+    {
+      "epoch": 0.4911371237458194,
+      "grad_norm": 0.5490350127220154,
+      "learning_rate": 8e-05,
+      "loss": 1.7237,
+      "step": 8811
+    },
+    {
+      "epoch": 0.4911928651059086,
+      "grad_norm": 0.5328288078308105,
+      "learning_rate": 8e-05,
+      "loss": 1.6545,
+      "step": 8812
+    },
+    {
+      "epoch": 0.4912486064659978,
+      "grad_norm": 0.6055459380149841,
+      "learning_rate": 8e-05,
+      "loss": 1.5455,
+      "step": 8813
+    },
+    {
+      "epoch": 0.49130434782608695,
+      "grad_norm": 0.5313544273376465,
+      "learning_rate": 8e-05,
+      "loss": 1.5263,
+      "step": 8814
+    },
+    {
+      "epoch": 0.49136008918617613,
+      "grad_norm": 0.5419933199882507,
+      "learning_rate": 8e-05,
+      "loss": 1.7685,
+      "step": 8815
+    },
+    {
+      "epoch": 0.4914158305462653,
+      "grad_norm": 0.5350378155708313,
+      "learning_rate": 8e-05,
+      "loss": 1.5742,
+      "step": 8816
+    },
+    {
+      "epoch": 0.4914715719063545,
+      "grad_norm": 0.5508986711502075,
+      "learning_rate": 8e-05,
+      "loss": 1.7074,
+      "step": 8817
+    },
+    {
+      "epoch": 0.4915273132664437,
+      "grad_norm": 0.5655425786972046,
+      "learning_rate": 8e-05,
+      "loss": 1.7255,
+      "step": 8818
+    },
+    {
+      "epoch": 0.4915830546265329,
+      "grad_norm": 0.5334153175354004,
+      "learning_rate": 8e-05,
+      "loss": 1.534,
+      "step": 8819
+    },
+    {
+      "epoch": 0.4916387959866221,
+      "grad_norm": 0.5287074446678162,
+      "learning_rate": 8e-05,
+      "loss": 1.5271,
+      "step": 8820
+    },
+    {
+      "epoch": 0.49169453734671126,
+      "grad_norm": 0.5889909863471985,
+      "learning_rate": 8e-05,
+      "loss": 1.7964,
+      "step": 8821
+    },
+    {
+      "epoch": 0.49175027870680044,
+      "grad_norm": 0.5158841013908386,
+      "learning_rate": 8e-05,
+      "loss": 1.5585,
+      "step": 8822
+    },
+    {
+      "epoch": 0.4918060200668896,
+      "grad_norm": 0.5091415643692017,
+      "learning_rate": 8e-05,
+      "loss": 1.5794,
+      "step": 8823
+    },
+    {
+      "epoch": 0.4918617614269788,
+      "grad_norm": 0.5212059617042542,
+      "learning_rate": 8e-05,
+      "loss": 1.62,
+      "step": 8824
+    },
+    {
+      "epoch": 0.49191750278706803,
+      "grad_norm": 0.5812239050865173,
+      "learning_rate": 8e-05,
+      "loss": 1.8132,
+      "step": 8825
+    },
+    {
+      "epoch": 0.4919732441471572,
+      "grad_norm": 0.5414086580276489,
+      "learning_rate": 8e-05,
+      "loss": 1.5705,
+      "step": 8826
+    },
+    {
+      "epoch": 0.4920289855072464,
+      "grad_norm": 0.5411535501480103,
+      "learning_rate": 8e-05,
+      "loss": 1.6643,
+      "step": 8827
+    },
+    {
+      "epoch": 0.49208472686733556,
+      "grad_norm": 0.5157851576805115,
+      "learning_rate": 8e-05,
+      "loss": 1.7066,
+      "step": 8828
+    },
+    {
+      "epoch": 0.49214046822742474,
+      "grad_norm": 0.5341320633888245,
+      "learning_rate": 8e-05,
+      "loss": 1.772,
+      "step": 8829
+    },
+    {
+      "epoch": 0.4921962095875139,
+      "grad_norm": 0.5654478669166565,
+      "learning_rate": 8e-05,
+      "loss": 1.7265,
+      "step": 8830
+    },
+    {
+      "epoch": 0.4922519509476031,
+      "grad_norm": 0.5648899078369141,
+      "learning_rate": 8e-05,
+      "loss": 1.6301,
+      "step": 8831
+    },
+    {
+      "epoch": 0.49230769230769234,
+      "grad_norm": 0.5131524801254272,
+      "learning_rate": 8e-05,
+      "loss": 1.648,
+      "step": 8832
+    },
+    {
+      "epoch": 0.4923634336677815,
+      "grad_norm": 0.49470341205596924,
+      "learning_rate": 8e-05,
+      "loss": 1.583,
+      "step": 8833
+    },
+    {
+      "epoch": 0.4924191750278707,
+      "grad_norm": 0.5371274352073669,
+      "learning_rate": 8e-05,
+      "loss": 1.743,
+      "step": 8834
+    },
+    {
+      "epoch": 0.49247491638795987,
+      "grad_norm": 0.5869753956794739,
+      "learning_rate": 8e-05,
+      "loss": 1.6407,
+      "step": 8835
+    },
+    {
+      "epoch": 0.49253065774804905,
+      "grad_norm": 0.49453967809677124,
+      "learning_rate": 8e-05,
+      "loss": 1.4267,
+      "step": 8836
+    },
+    {
+      "epoch": 0.49258639910813823,
+      "grad_norm": 0.5081074237823486,
+      "learning_rate": 8e-05,
+      "loss": 1.5216,
+      "step": 8837
+    },
+    {
+      "epoch": 0.4926421404682274,
+      "grad_norm": 0.5137898325920105,
+      "learning_rate": 8e-05,
+      "loss": 1.5087,
+      "step": 8838
+    },
+    {
+      "epoch": 0.4926978818283166,
+      "grad_norm": 0.5767127275466919,
+      "learning_rate": 8e-05,
+      "loss": 1.6743,
+      "step": 8839
+    },
+    {
+      "epoch": 0.4927536231884058,
+      "grad_norm": 0.5548331141471863,
+      "learning_rate": 8e-05,
+      "loss": 1.7111,
+      "step": 8840
+    },
+    {
+      "epoch": 0.492809364548495,
+      "grad_norm": 0.5236859321594238,
+      "learning_rate": 8e-05,
+      "loss": 1.6453,
+      "step": 8841
+    },
+    {
+      "epoch": 0.4928651059085842,
+      "grad_norm": 0.5255964994430542,
+      "learning_rate": 8e-05,
+      "loss": 1.7073,
+      "step": 8842
+    },
+    {
+      "epoch": 0.49292084726867336,
+      "grad_norm": 0.5545988082885742,
+      "learning_rate": 8e-05,
+      "loss": 1.6276,
+      "step": 8843
+    },
+    {
+      "epoch": 0.49297658862876254,
+      "grad_norm": 0.5834181308746338,
+      "learning_rate": 8e-05,
+      "loss": 1.727,
+      "step": 8844
+    },
+    {
+      "epoch": 0.4930323299888517,
+      "grad_norm": 0.5687046647071838,
+      "learning_rate": 8e-05,
+      "loss": 1.7085,
+      "step": 8845
+    },
+    {
+      "epoch": 0.4930880713489409,
+      "grad_norm": 0.5452062487602234,
+      "learning_rate": 8e-05,
+      "loss": 1.6304,
+      "step": 8846
+    },
+    {
+      "epoch": 0.4931438127090301,
+      "grad_norm": 0.5048710107803345,
+      "learning_rate": 8e-05,
+      "loss": 1.3044,
+      "step": 8847
+    },
+    {
+      "epoch": 0.4931995540691193,
+      "grad_norm": 0.534052848815918,
+      "learning_rate": 8e-05,
+      "loss": 1.7413,
+      "step": 8848
+    },
+    {
+      "epoch": 0.4932552954292085,
+      "grad_norm": 0.6282777190208435,
+      "learning_rate": 8e-05,
+      "loss": 2.0405,
+      "step": 8849
+    },
+    {
+      "epoch": 0.49331103678929766,
+      "grad_norm": 0.5243322253227234,
+      "learning_rate": 8e-05,
+      "loss": 1.4643,
+      "step": 8850
+    },
+    {
+      "epoch": 0.49336677814938684,
+      "grad_norm": 0.5178560018539429,
+      "learning_rate": 8e-05,
+      "loss": 1.5601,
+      "step": 8851
+    },
+    {
+      "epoch": 0.493422519509476,
+      "grad_norm": 0.5159735679626465,
+      "learning_rate": 8e-05,
+      "loss": 1.6775,
+      "step": 8852
+    },
+    {
+      "epoch": 0.4934782608695652,
+      "grad_norm": 0.5154753923416138,
+      "learning_rate": 8e-05,
+      "loss": 1.4601,
+      "step": 8853
+    },
+    {
+      "epoch": 0.4935340022296544,
+      "grad_norm": 0.5437743067741394,
+      "learning_rate": 8e-05,
+      "loss": 1.4441,
+      "step": 8854
+    },
+    {
+      "epoch": 0.4935897435897436,
+      "grad_norm": 0.5819849371910095,
+      "learning_rate": 8e-05,
+      "loss": 1.7588,
+      "step": 8855
+    },
+    {
+      "epoch": 0.4936454849498328,
+      "grad_norm": 0.5482854247093201,
+      "learning_rate": 8e-05,
+      "loss": 1.6496,
+      "step": 8856
+    },
+    {
+      "epoch": 0.49370122630992197,
+      "grad_norm": 0.4933509826660156,
+      "learning_rate": 8e-05,
+      "loss": 1.6415,
+      "step": 8857
+    },
+    {
+      "epoch": 0.49375696767001115,
+      "grad_norm": 0.5568447113037109,
+      "learning_rate": 8e-05,
+      "loss": 1.7922,
+      "step": 8858
+    },
+    {
+      "epoch": 0.4938127090301003,
+      "grad_norm": 0.548936665058136,
+      "learning_rate": 8e-05,
+      "loss": 1.6704,
+      "step": 8859
+    },
+    {
+      "epoch": 0.4938684503901895,
+      "grad_norm": 0.542045533657074,
+      "learning_rate": 8e-05,
+      "loss": 1.6765,
+      "step": 8860
+    },
+    {
+      "epoch": 0.4939241917502787,
+      "grad_norm": 0.5615548491477966,
+      "learning_rate": 8e-05,
+      "loss": 1.6885,
+      "step": 8861
+    },
+    {
+      "epoch": 0.4939799331103679,
+      "grad_norm": 0.5786969661712646,
+      "learning_rate": 8e-05,
+      "loss": 1.6795,
+      "step": 8862
+    },
+    {
+      "epoch": 0.4940356744704571,
+      "grad_norm": 0.5402855277061462,
+      "learning_rate": 8e-05,
+      "loss": 1.6715,
+      "step": 8863
+    },
+    {
+      "epoch": 0.4940914158305463,
+      "grad_norm": 0.5332042574882507,
+      "learning_rate": 8e-05,
+      "loss": 1.5561,
+      "step": 8864
+    },
+    {
+      "epoch": 0.49414715719063546,
+      "grad_norm": 0.5454310178756714,
+      "learning_rate": 8e-05,
+      "loss": 1.6758,
+      "step": 8865
+    },
+    {
+      "epoch": 0.49420289855072463,
+      "grad_norm": 0.5213438868522644,
+      "learning_rate": 8e-05,
+      "loss": 1.5993,
+      "step": 8866
+    },
+    {
+      "epoch": 0.4942586399108138,
+      "grad_norm": 0.550634503364563,
+      "learning_rate": 8e-05,
+      "loss": 1.7791,
+      "step": 8867
+    },
+    {
+      "epoch": 0.494314381270903,
+      "grad_norm": 0.4944699704647064,
+      "learning_rate": 8e-05,
+      "loss": 1.4123,
+      "step": 8868
+    },
+    {
+      "epoch": 0.49437012263099217,
+      "grad_norm": 0.5702661275863647,
+      "learning_rate": 8e-05,
+      "loss": 1.9546,
+      "step": 8869
+    },
+    {
+      "epoch": 0.4944258639910814,
+      "grad_norm": 0.5053265690803528,
+      "learning_rate": 8e-05,
+      "loss": 1.5537,
+      "step": 8870
+    },
+    {
+      "epoch": 0.4944816053511706,
+      "grad_norm": 0.5766624808311462,
+      "learning_rate": 8e-05,
+      "loss": 1.957,
+      "step": 8871
+    },
+    {
+      "epoch": 0.49453734671125976,
+      "grad_norm": 0.5471301078796387,
+      "learning_rate": 8e-05,
+      "loss": 1.695,
+      "step": 8872
+    },
+    {
+      "epoch": 0.49459308807134894,
+      "grad_norm": 0.5663468837738037,
+      "learning_rate": 8e-05,
+      "loss": 1.5985,
+      "step": 8873
+    },
+    {
+      "epoch": 0.4946488294314381,
+      "grad_norm": 0.513062596321106,
+      "learning_rate": 8e-05,
+      "loss": 1.5668,
+      "step": 8874
+    },
+    {
+      "epoch": 0.4947045707915273,
+      "grad_norm": 0.5351914763450623,
+      "learning_rate": 8e-05,
+      "loss": 1.8147,
+      "step": 8875
+    },
+    {
+      "epoch": 0.4947603121516165,
+      "grad_norm": 0.5259719491004944,
+      "learning_rate": 8e-05,
+      "loss": 1.5631,
+      "step": 8876
+    },
+    {
+      "epoch": 0.4948160535117057,
+      "grad_norm": 0.527617871761322,
+      "learning_rate": 8e-05,
+      "loss": 1.484,
+      "step": 8877
+    },
+    {
+      "epoch": 0.4948717948717949,
+      "grad_norm": 0.529051661491394,
+      "learning_rate": 8e-05,
+      "loss": 1.6024,
+      "step": 8878
+    },
+    {
+      "epoch": 0.49492753623188407,
+      "grad_norm": 0.5139374136924744,
+      "learning_rate": 8e-05,
+      "loss": 1.4492,
+      "step": 8879
+    },
+    {
+      "epoch": 0.49498327759197325,
+      "grad_norm": 0.523256778717041,
+      "learning_rate": 8e-05,
+      "loss": 1.6188,
+      "step": 8880
+    },
+    {
+      "epoch": 0.4950390189520624,
+      "grad_norm": 0.5505326390266418,
+      "learning_rate": 8e-05,
+      "loss": 1.675,
+      "step": 8881
+    },
+    {
+      "epoch": 0.4950947603121516,
+      "grad_norm": 0.5377850532531738,
+      "learning_rate": 8e-05,
+      "loss": 1.6074,
+      "step": 8882
+    },
+    {
+      "epoch": 0.4951505016722408,
+      "grad_norm": 0.5315950512886047,
+      "learning_rate": 8e-05,
+      "loss": 1.5816,
+      "step": 8883
+    },
+    {
+      "epoch": 0.49520624303232996,
+      "grad_norm": 0.6018045544624329,
+      "learning_rate": 8e-05,
+      "loss": 1.7899,
+      "step": 8884
+    },
+    {
+      "epoch": 0.4952619843924192,
+      "grad_norm": 0.579967200756073,
+      "learning_rate": 8e-05,
+      "loss": 1.6885,
+      "step": 8885
+    },
+    {
+      "epoch": 0.4953177257525084,
+      "grad_norm": 0.593023955821991,
+      "learning_rate": 8e-05,
+      "loss": 1.7914,
+      "step": 8886
+    },
+    {
+      "epoch": 0.49537346711259755,
+      "grad_norm": 0.5540221333503723,
+      "learning_rate": 8e-05,
+      "loss": 1.7083,
+      "step": 8887
+    },
+    {
+      "epoch": 0.49542920847268673,
+      "grad_norm": 0.5043448805809021,
+      "learning_rate": 8e-05,
+      "loss": 1.5477,
+      "step": 8888
+    },
+    {
+      "epoch": 0.4954849498327759,
+      "grad_norm": 0.6106610894203186,
+      "learning_rate": 8e-05,
+      "loss": 1.9431,
+      "step": 8889
+    },
+    {
+      "epoch": 0.4955406911928651,
+      "grad_norm": 0.5356171131134033,
+      "learning_rate": 8e-05,
+      "loss": 1.6505,
+      "step": 8890
+    },
+    {
+      "epoch": 0.49559643255295427,
+      "grad_norm": 0.5981549620628357,
+      "learning_rate": 8e-05,
+      "loss": 1.8112,
+      "step": 8891
+    },
+    {
+      "epoch": 0.4956521739130435,
+      "grad_norm": 0.5713699460029602,
+      "learning_rate": 8e-05,
+      "loss": 1.4266,
+      "step": 8892
+    },
+    {
+      "epoch": 0.4957079152731327,
+      "grad_norm": 0.5626436471939087,
+      "learning_rate": 8e-05,
+      "loss": 1.8815,
+      "step": 8893
+    },
+    {
+      "epoch": 0.49576365663322186,
+      "grad_norm": 0.5311765670776367,
+      "learning_rate": 8e-05,
+      "loss": 1.6903,
+      "step": 8894
+    },
+    {
+      "epoch": 0.49581939799331104,
+      "grad_norm": 0.5413386225700378,
+      "learning_rate": 8e-05,
+      "loss": 1.5235,
+      "step": 8895
+    },
+    {
+      "epoch": 0.4958751393534002,
+      "grad_norm": 0.4782915413379669,
+      "learning_rate": 8e-05,
+      "loss": 1.2485,
+      "step": 8896
+    },
+    {
+      "epoch": 0.4959308807134894,
+      "grad_norm": 0.5347909331321716,
+      "learning_rate": 8e-05,
+      "loss": 1.6659,
+      "step": 8897
+    },
+    {
+      "epoch": 0.4959866220735786,
+      "grad_norm": 0.5323268175125122,
+      "learning_rate": 8e-05,
+      "loss": 1.7437,
+      "step": 8898
+    },
+    {
+      "epoch": 0.49604236343366775,
+      "grad_norm": 0.5715739727020264,
+      "learning_rate": 8e-05,
+      "loss": 1.7099,
+      "step": 8899
+    },
+    {
+      "epoch": 0.496098104793757,
+      "grad_norm": 0.7310459613800049,
+      "learning_rate": 8e-05,
+      "loss": 1.7719,
+      "step": 8900
+    },
+    {
+      "epoch": 0.49615384615384617,
+      "grad_norm": 0.598199188709259,
+      "learning_rate": 8e-05,
+      "loss": 1.6053,
+      "step": 8901
+    },
+    {
+      "epoch": 0.49620958751393535,
+      "grad_norm": 0.5783191919326782,
+      "learning_rate": 8e-05,
+      "loss": 1.6202,
+      "step": 8902
+    },
+    {
+      "epoch": 0.4962653288740245,
+      "grad_norm": 0.533246636390686,
+      "learning_rate": 8e-05,
+      "loss": 1.4843,
+      "step": 8903
+    },
+    {
+      "epoch": 0.4963210702341137,
+      "grad_norm": 0.615612804889679,
+      "learning_rate": 8e-05,
+      "loss": 1.9378,
+      "step": 8904
+    },
+    {
+      "epoch": 0.4963768115942029,
+      "grad_norm": 0.4893304407596588,
+      "learning_rate": 8e-05,
+      "loss": 1.4438,
+      "step": 8905
+    },
+    {
+      "epoch": 0.49643255295429206,
+      "grad_norm": 0.5676378607749939,
+      "learning_rate": 8e-05,
+      "loss": 1.7779,
+      "step": 8906
+    },
+    {
+      "epoch": 0.4964882943143813,
+      "grad_norm": 0.5795214772224426,
+      "learning_rate": 8e-05,
+      "loss": 1.7449,
+      "step": 8907
+    },
+    {
+      "epoch": 0.4965440356744705,
+      "grad_norm": 0.5832824110984802,
+      "learning_rate": 8e-05,
+      "loss": 1.8669,
+      "step": 8908
+    },
+    {
+      "epoch": 0.49659977703455965,
+      "grad_norm": 0.7791319489479065,
+      "learning_rate": 8e-05,
+      "loss": 1.6592,
+      "step": 8909
+    },
+    {
+      "epoch": 0.49665551839464883,
+      "grad_norm": 0.5205925703048706,
+      "learning_rate": 8e-05,
+      "loss": 1.6496,
+      "step": 8910
+    },
+    {
+      "epoch": 0.496711259754738,
+      "grad_norm": 0.5032536387443542,
+      "learning_rate": 8e-05,
+      "loss": 1.4186,
+      "step": 8911
+    },
+    {
+      "epoch": 0.4967670011148272,
+      "grad_norm": 0.5437055826187134,
+      "learning_rate": 8e-05,
+      "loss": 1.7029,
+      "step": 8912
+    },
+    {
+      "epoch": 0.49682274247491637,
+      "grad_norm": 0.5419133305549622,
+      "learning_rate": 8e-05,
+      "loss": 1.7818,
+      "step": 8913
+    },
+    {
+      "epoch": 0.49687848383500555,
+      "grad_norm": 0.5332748889923096,
+      "learning_rate": 8e-05,
+      "loss": 1.4844,
+      "step": 8914
+    },
+    {
+      "epoch": 0.4969342251950948,
+      "grad_norm": 0.5628235936164856,
+      "learning_rate": 8e-05,
+      "loss": 1.7583,
+      "step": 8915
+    },
+    {
+      "epoch": 0.49698996655518396,
+      "grad_norm": 0.5357251763343811,
+      "learning_rate": 8e-05,
+      "loss": 1.9001,
+      "step": 8916
+    },
+    {
+      "epoch": 0.49704570791527314,
+      "grad_norm": 0.5667791366577148,
+      "learning_rate": 8e-05,
+      "loss": 1.7927,
+      "step": 8917
+    },
+    {
+      "epoch": 0.4971014492753623,
+      "grad_norm": 0.4990066587924957,
+      "learning_rate": 8e-05,
+      "loss": 1.4913,
+      "step": 8918
+    },
+    {
+      "epoch": 0.4971571906354515,
+      "grad_norm": 0.5187414288520813,
+      "learning_rate": 8e-05,
+      "loss": 1.4935,
+      "step": 8919
+    },
+    {
+      "epoch": 0.4972129319955407,
+      "grad_norm": 0.5602273344993591,
+      "learning_rate": 8e-05,
+      "loss": 1.6549,
+      "step": 8920
+    },
+    {
+      "epoch": 0.49726867335562985,
+      "grad_norm": 0.5536525249481201,
+      "learning_rate": 8e-05,
+      "loss": 1.5933,
+      "step": 8921
+    },
+    {
+      "epoch": 0.4973244147157191,
+      "grad_norm": 0.5444284677505493,
+      "learning_rate": 8e-05,
+      "loss": 1.6299,
+      "step": 8922
+    },
+    {
+      "epoch": 0.49738015607580827,
+      "grad_norm": 0.5906103849411011,
+      "learning_rate": 8e-05,
+      "loss": 1.7832,
+      "step": 8923
+    },
+    {
+      "epoch": 0.49743589743589745,
+      "grad_norm": 0.5382599234580994,
+      "learning_rate": 8e-05,
+      "loss": 1.6533,
+      "step": 8924
+    },
+    {
+      "epoch": 0.4974916387959866,
+      "grad_norm": 0.5361124873161316,
+      "learning_rate": 8e-05,
+      "loss": 1.6416,
+      "step": 8925
+    },
+    {
+      "epoch": 0.4975473801560758,
+      "grad_norm": 0.5967959761619568,
+      "learning_rate": 8e-05,
+      "loss": 1.9175,
+      "step": 8926
+    },
+    {
+      "epoch": 0.497603121516165,
+      "grad_norm": 0.5135853886604309,
+      "learning_rate": 8e-05,
+      "loss": 1.545,
+      "step": 8927
+    },
+    {
+      "epoch": 0.49765886287625416,
+      "grad_norm": 0.5927615761756897,
+      "learning_rate": 8e-05,
+      "loss": 1.9414,
+      "step": 8928
+    },
+    {
+      "epoch": 0.4977146042363434,
+      "grad_norm": 0.5337189435958862,
+      "learning_rate": 8e-05,
+      "loss": 1.7404,
+      "step": 8929
+    },
+    {
+      "epoch": 0.4977703455964326,
+      "grad_norm": 0.5466311573982239,
+      "learning_rate": 8e-05,
+      "loss": 1.7012,
+      "step": 8930
+    },
+    {
+      "epoch": 0.49782608695652175,
+      "grad_norm": 0.4856451153755188,
+      "learning_rate": 8e-05,
+      "loss": 1.572,
+      "step": 8931
+    },
+    {
+      "epoch": 0.49788182831661093,
+      "grad_norm": 0.5938416123390198,
+      "learning_rate": 8e-05,
+      "loss": 1.831,
+      "step": 8932
+    },
+    {
+      "epoch": 0.4979375696767001,
+      "grad_norm": 0.46626678109169006,
+      "learning_rate": 8e-05,
+      "loss": 1.2963,
+      "step": 8933
+    },
+    {
+      "epoch": 0.4979933110367893,
+      "grad_norm": 0.6333751678466797,
+      "learning_rate": 8e-05,
+      "loss": 1.4346,
+      "step": 8934
+    },
+    {
+      "epoch": 0.49804905239687847,
+      "grad_norm": 0.5655824542045593,
+      "learning_rate": 8e-05,
+      "loss": 1.8434,
+      "step": 8935
+    },
+    {
+      "epoch": 0.49810479375696765,
+      "grad_norm": 0.5203345417976379,
+      "learning_rate": 8e-05,
+      "loss": 1.7327,
+      "step": 8936
+    },
+    {
+      "epoch": 0.4981605351170569,
+      "grad_norm": 0.5423237085342407,
+      "learning_rate": 8e-05,
+      "loss": 1.6376,
+      "step": 8937
+    },
+    {
+      "epoch": 0.49821627647714606,
+      "grad_norm": 0.5489369630813599,
+      "learning_rate": 8e-05,
+      "loss": 1.7333,
+      "step": 8938
+    },
+    {
+      "epoch": 0.49827201783723524,
+      "grad_norm": 0.5339183807373047,
+      "learning_rate": 8e-05,
+      "loss": 1.6599,
+      "step": 8939
+    },
+    {
+      "epoch": 0.4983277591973244,
+      "grad_norm": 0.5917924642562866,
+      "learning_rate": 8e-05,
+      "loss": 2.0178,
+      "step": 8940
+    },
+    {
+      "epoch": 0.4983835005574136,
+      "grad_norm": 0.5788326263427734,
+      "learning_rate": 8e-05,
+      "loss": 1.607,
+      "step": 8941
+    },
+    {
+      "epoch": 0.4984392419175028,
+      "grad_norm": 0.49358728528022766,
+      "learning_rate": 8e-05,
+      "loss": 1.4111,
+      "step": 8942
+    },
+    {
+      "epoch": 0.49849498327759195,
+      "grad_norm": 0.5892625451087952,
+      "learning_rate": 8e-05,
+      "loss": 1.8303,
+      "step": 8943
+    },
+    {
+      "epoch": 0.4985507246376812,
+      "grad_norm": 0.6209052205085754,
+      "learning_rate": 8e-05,
+      "loss": 1.8335,
+      "step": 8944
+    },
+    {
+      "epoch": 0.49860646599777037,
+      "grad_norm": 0.5494616031646729,
+      "learning_rate": 8e-05,
+      "loss": 1.8625,
+      "step": 8945
+    },
+    {
+      "epoch": 0.49866220735785954,
+      "grad_norm": 0.5134598612785339,
+      "learning_rate": 8e-05,
+      "loss": 1.5633,
+      "step": 8946
+    },
+    {
+      "epoch": 0.4987179487179487,
+      "grad_norm": 0.5650904178619385,
+      "learning_rate": 8e-05,
+      "loss": 1.6744,
+      "step": 8947
+    },
+    {
+      "epoch": 0.4987736900780379,
+      "grad_norm": 0.545432984828949,
+      "learning_rate": 8e-05,
+      "loss": 1.9893,
+      "step": 8948
+    },
+    {
+      "epoch": 0.4988294314381271,
+      "grad_norm": 0.5759139657020569,
+      "learning_rate": 8e-05,
+      "loss": 1.624,
+      "step": 8949
+    },
+    {
+      "epoch": 0.49888517279821626,
+      "grad_norm": 0.5109109878540039,
+      "learning_rate": 8e-05,
+      "loss": 1.4989,
+      "step": 8950
+    },
+    {
+      "epoch": 0.49894091415830544,
+      "grad_norm": 0.48945099115371704,
+      "learning_rate": 8e-05,
+      "loss": 1.3881,
+      "step": 8951
+    },
+    {
+      "epoch": 0.49899665551839467,
+      "grad_norm": 0.532191812992096,
+      "learning_rate": 8e-05,
+      "loss": 1.5542,
+      "step": 8952
+    },
+    {
+      "epoch": 0.49905239687848385,
+      "grad_norm": 0.9853920340538025,
+      "learning_rate": 8e-05,
+      "loss": 1.6086,
+      "step": 8953
+    },
+    {
+      "epoch": 0.49910813823857303,
+      "grad_norm": 0.5455578565597534,
+      "learning_rate": 8e-05,
+      "loss": 1.3766,
+      "step": 8954
+    },
+    {
+      "epoch": 0.4991638795986622,
+      "grad_norm": 0.5894210338592529,
+      "learning_rate": 8e-05,
+      "loss": 1.7001,
+      "step": 8955
+    },
+    {
+      "epoch": 0.4992196209587514,
+      "grad_norm": 0.5502829551696777,
+      "learning_rate": 8e-05,
+      "loss": 1.6782,
+      "step": 8956
+    },
+    {
+      "epoch": 0.49927536231884057,
+      "grad_norm": 0.5422848463058472,
+      "learning_rate": 8e-05,
+      "loss": 1.5955,
+      "step": 8957
+    },
+    {
+      "epoch": 0.49933110367892974,
+      "grad_norm": 0.577444851398468,
+      "learning_rate": 8e-05,
+      "loss": 1.8286,
+      "step": 8958
+    },
+    {
+      "epoch": 0.499386845039019,
+      "grad_norm": 0.5444875359535217,
+      "learning_rate": 8e-05,
+      "loss": 1.5021,
+      "step": 8959
+    },
+    {
+      "epoch": 0.49944258639910816,
+      "grad_norm": 0.5610191822052002,
+      "learning_rate": 8e-05,
+      "loss": 1.7752,
+      "step": 8960
+    },
+    {
+      "epoch": 0.49949832775919734,
+      "grad_norm": 0.5572718977928162,
+      "learning_rate": 8e-05,
+      "loss": 1.7961,
+      "step": 8961
+    },
+    {
+      "epoch": 0.4995540691192865,
+      "grad_norm": 0.5246376991271973,
+      "learning_rate": 8e-05,
+      "loss": 1.2681,
+      "step": 8962
+    },
+    {
+      "epoch": 0.4996098104793757,
+      "grad_norm": 0.5913357138633728,
+      "learning_rate": 8e-05,
+      "loss": 1.7517,
+      "step": 8963
+    },
+    {
+      "epoch": 0.49966555183946487,
+      "grad_norm": 0.520278811454773,
+      "learning_rate": 8e-05,
+      "loss": 1.5818,
+      "step": 8964
+    },
+    {
+      "epoch": 0.49972129319955405,
+      "grad_norm": 0.5451558232307434,
+      "learning_rate": 8e-05,
+      "loss": 1.6493,
+      "step": 8965
+    },
+    {
+      "epoch": 0.49977703455964323,
+      "grad_norm": 0.5383256077766418,
+      "learning_rate": 8e-05,
+      "loss": 1.9797,
+      "step": 8966
+    },
+    {
+      "epoch": 0.49983277591973246,
+      "grad_norm": 0.5744273066520691,
+      "learning_rate": 8e-05,
+      "loss": 1.8506,
+      "step": 8967
+    },
+    {
+      "epoch": 0.49988851727982164,
+      "grad_norm": 0.6074442267417908,
+      "learning_rate": 8e-05,
+      "loss": 1.9342,
+      "step": 8968
+    },
+    {
+      "epoch": 0.4999442586399108,
+      "grad_norm": 0.5620627999305725,
+      "learning_rate": 8e-05,
+      "loss": 1.7157,
+      "step": 8969
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5319579243659973,
+      "learning_rate": 8e-05,
+      "loss": 1.5882,
+      "step": 8970
+    },
+    {
+      "epoch": 0.5000557413600892,
+      "grad_norm": 0.5307610034942627,
+      "learning_rate": 8e-05,
+      "loss": 1.6502,
+      "step": 8971
+    },
+    {
+      "epoch": 0.5001114827201784,
+      "grad_norm": 0.5093584060668945,
+      "learning_rate": 8e-05,
+      "loss": 1.4701,
+      "step": 8972
+    },
+    {
+      "epoch": 0.5001672240802676,
+      "grad_norm": 0.5648424029350281,
+      "learning_rate": 8e-05,
+      "loss": 1.6485,
+      "step": 8973
+    },
+    {
+      "epoch": 0.5002229654403567,
+      "grad_norm": 0.5174546241760254,
+      "learning_rate": 8e-05,
+      "loss": 1.6417,
+      "step": 8974
+    },
+    {
+      "epoch": 0.500278706800446,
+      "grad_norm": 0.5817108154296875,
+      "learning_rate": 8e-05,
+      "loss": 1.5185,
+      "step": 8975
+    },
+    {
+      "epoch": 0.5003344481605351,
+      "grad_norm": 0.5302509665489197,
+      "learning_rate": 8e-05,
+      "loss": 1.7699,
+      "step": 8976
+    },
+    {
+      "epoch": 0.5003901895206243,
+      "grad_norm": 0.5956892967224121,
+      "learning_rate": 8e-05,
+      "loss": 1.6252,
+      "step": 8977
+    },
+    {
+      "epoch": 0.5004459308807135,
+      "grad_norm": 0.564408540725708,
+      "learning_rate": 8e-05,
+      "loss": 1.4983,
+      "step": 8978
+    },
+    {
+      "epoch": 0.5005016722408027,
+      "grad_norm": 0.5849278569221497,
+      "learning_rate": 8e-05,
+      "loss": 1.7398,
+      "step": 8979
+    },
+    {
+      "epoch": 0.5005574136008919,
+      "grad_norm": 0.5324124097824097,
+      "learning_rate": 8e-05,
+      "loss": 1.6415,
+      "step": 8980
+    },
+    {
+      "epoch": 0.500613154960981,
+      "grad_norm": 0.5535297393798828,
+      "learning_rate": 8e-05,
+      "loss": 1.6123,
+      "step": 8981
+    },
+    {
+      "epoch": 0.5006688963210703,
+      "grad_norm": 0.5630205273628235,
+      "learning_rate": 8e-05,
+      "loss": 1.7372,
+      "step": 8982
+    },
+    {
+      "epoch": 0.5007246376811594,
+      "grad_norm": 0.5356989502906799,
+      "learning_rate": 8e-05,
+      "loss": 1.7165,
+      "step": 8983
+    },
+    {
+      "epoch": 0.5007803790412486,
+      "grad_norm": 0.6177685856819153,
+      "learning_rate": 8e-05,
+      "loss": 1.8803,
+      "step": 8984
+    },
+    {
+      "epoch": 0.5008361204013378,
+      "grad_norm": 0.5300348401069641,
+      "learning_rate": 8e-05,
+      "loss": 1.5568,
+      "step": 8985
+    },
+    {
+      "epoch": 0.500891861761427,
+      "grad_norm": 0.5672892332077026,
+      "learning_rate": 8e-05,
+      "loss": 1.7536,
+      "step": 8986
+    },
+    {
+      "epoch": 0.5009476031215162,
+      "grad_norm": 0.5484430193901062,
+      "learning_rate": 8e-05,
+      "loss": 1.6806,
+      "step": 8987
+    },
+    {
+      "epoch": 0.5010033444816053,
+      "grad_norm": 0.49196168780326843,
+      "learning_rate": 8e-05,
+      "loss": 1.4351,
+      "step": 8988
+    },
+    {
+      "epoch": 0.5010590858416946,
+      "grad_norm": 0.49675580859184265,
+      "learning_rate": 8e-05,
+      "loss": 1.5384,
+      "step": 8989
+    },
+    {
+      "epoch": 0.5011148272017837,
+      "grad_norm": 0.5410286784172058,
+      "learning_rate": 8e-05,
+      "loss": 1.7398,
+      "step": 8990
+    },
+    {
+      "epoch": 0.5011705685618729,
+      "grad_norm": 0.5007629990577698,
+      "learning_rate": 8e-05,
+      "loss": 1.6711,
+      "step": 8991
+    },
+    {
+      "epoch": 0.501226309921962,
+      "grad_norm": 0.5875619649887085,
+      "learning_rate": 8e-05,
+      "loss": 1.5316,
+      "step": 8992
+    },
+    {
+      "epoch": 0.5012820512820513,
+      "grad_norm": 0.5303008556365967,
+      "learning_rate": 8e-05,
+      "loss": 1.5361,
+      "step": 8993
+    },
+    {
+      "epoch": 0.5013377926421405,
+      "grad_norm": 0.6002159714698792,
+      "learning_rate": 8e-05,
+      "loss": 1.9255,
+      "step": 8994
+    },
+    {
+      "epoch": 0.5013935340022296,
+      "grad_norm": 0.5164722204208374,
+      "learning_rate": 8e-05,
+      "loss": 1.5173,
+      "step": 8995
+    },
+    {
+      "epoch": 0.5014492753623189,
+      "grad_norm": 0.49862802028656006,
+      "learning_rate": 8e-05,
+      "loss": 1.6109,
+      "step": 8996
+    },
+    {
+      "epoch": 0.501505016722408,
+      "grad_norm": 0.594048798084259,
+      "learning_rate": 8e-05,
+      "loss": 1.8636,
+      "step": 8997
+    },
+    {
+      "epoch": 0.5015607580824972,
+      "grad_norm": 0.5271819233894348,
+      "learning_rate": 8e-05,
+      "loss": 1.6044,
+      "step": 8998
+    },
+    {
+      "epoch": 0.5016164994425863,
+      "grad_norm": 0.5142638683319092,
+      "learning_rate": 8e-05,
+      "loss": 1.6739,
+      "step": 8999
+    },
+    {
+      "epoch": 0.5016722408026756,
+      "grad_norm": 0.5490266680717468,
+      "learning_rate": 8e-05,
+      "loss": 1.5733,
+      "step": 9000
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 17940,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 3000,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 6.788311104356352e+18,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}