diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,22675 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 2.9952,
+  "eval_steps": 156,
+  "global_step": 3744,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "learning_rate": 0.0001,
+      "loss": 2.5411,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 0.0002,
+      "loss": 2.5059,
+      "step": 2
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 0.00019994663820704374,
+      "loss": 2.3172,
+      "step": 3
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 0.00019989327641408753,
+      "loss": 2.1123,
+      "step": 4
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 0.00019983991462113127,
+      "loss": 1.9615,
+      "step": 5
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 0.00019978655282817506,
+      "loss": 1.7271,
+      "step": 6
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 0.0001997331910352188,
+      "loss": 1.7111,
+      "step": 7
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 0.00019967982924226253,
+      "loss": 1.7608,
+      "step": 8
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 0.00019962646744930632,
+      "loss": 1.6313,
+      "step": 9
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 0.00019957310565635005,
+      "loss": 1.4983,
+      "step": 10
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 0.0001995197438633938,
+      "loss": 0.7791,
+      "step": 11
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 0.00019946638207043758,
+      "loss": 1.8302,
+      "step": 12
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 0.00019941302027748134,
+      "loss": 1.4667,
+      "step": 13
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 0.0001993596584845251,
+      "loss": 1.4658,
+      "step": 14
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 0.00019930629669156883,
+      "loss": 1.4719,
+      "step": 15
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 0.0001992529348986126,
+      "loss": 1.5187,
+      "step": 16
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 0.00019919957310565636,
+      "loss": 1.3919,
+      "step": 17
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 0.00019914621131270012,
+      "loss": 0.7413,
+      "step": 18
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.00019909284951974388,
+      "loss": 0.6697,
+      "step": 19
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.00019903948772678762,
+      "loss": 1.3548,
+      "step": 20
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.00019898612593383138,
+      "loss": 1.3856,
+      "step": 21
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.00019893276414087514,
+      "loss": 1.5602,
+      "step": 22
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.0001988794023479189,
+      "loss": 1.2768,
+      "step": 23
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.00019882604055496266,
+      "loss": 0.6351,
+      "step": 24
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.0001987726787620064,
+      "loss": 1.4422,
+      "step": 25
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.00019871931696905016,
+      "loss": 1.4349,
+      "step": 26
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.00019866595517609392,
+      "loss": 1.6123,
+      "step": 27
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.00019861259338313769,
+      "loss": 1.3682,
+      "step": 28
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.00019855923159018145,
+      "loss": 1.4204,
+      "step": 29
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.00019850586979722518,
+      "loss": 1.2458,
+      "step": 30
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.00019845250800426894,
+      "loss": 1.3934,
+      "step": 31
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.0001983991462113127,
+      "loss": 1.2083,
+      "step": 32
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.00019834578441835647,
+      "loss": 1.2454,
+      "step": 33
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.00019829242262540023,
+      "loss": 0.7013,
+      "step": 34
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.000198239060832444,
+      "loss": 1.4089,
+      "step": 35
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.00019818569903948773,
+      "loss": 1.3357,
+      "step": 36
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.0001981323372465315,
+      "loss": 1.4583,
+      "step": 37
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.00019807897545357525,
+      "loss": 1.3116,
+      "step": 38
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.000198025613660619,
+      "loss": 1.384,
+      "step": 39
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.00019797225186766277,
+      "loss": 1.1254,
+      "step": 40
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.0001979188900747065,
+      "loss": 1.4479,
+      "step": 41
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.00019786552828175027,
+      "loss": 1.3518,
+      "step": 42
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.00019781216648879403,
+      "loss": 1.3947,
+      "step": 43
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.0001977588046958378,
+      "loss": 1.1684,
+      "step": 44
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.00019770544290288156,
+      "loss": 0.6793,
+      "step": 45
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.0001976520811099253,
+      "loss": 1.3871,
+      "step": 46
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.00019759871931696905,
+      "loss": 1.188,
+      "step": 47
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.00019754535752401282,
+      "loss": 1.3686,
+      "step": 48
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.00019749199573105658,
+      "loss": 1.2913,
+      "step": 49
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.00019743863393810034,
+      "loss": 0.5762,
+      "step": 50
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.00019738527214514407,
+      "loss": 1.2729,
+      "step": 51
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.00019733191035218784,
+      "loss": 1.2666,
+      "step": 52
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.0001972785485592316,
+      "loss": 1.2657,
+      "step": 53
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.00019722518676627536,
+      "loss": 1.2121,
+      "step": 54
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.00019717182497331912,
+      "loss": 1.3209,
+      "step": 55
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.00019711846318036286,
+      "loss": 1.1682,
+      "step": 56
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.00019706510138740665,
+      "loss": 1.2771,
+      "step": 57
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.00019701173959445038,
+      "loss": 1.3008,
+      "step": 58
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.00019695837780149414,
+      "loss": 0.4822,
+      "step": 59
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.0001969050160085379,
+      "loss": 1.2384,
+      "step": 60
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.00019685165421558164,
+      "loss": 1.1817,
+      "step": 61
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.00019679829242262543,
+      "loss": 1.2594,
+      "step": 62
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.00019674493062966916,
+      "loss": 1.2142,
+      "step": 63
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.00019669156883671293,
+      "loss": 0.6465,
+      "step": 64
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.0001966382070437567,
+      "loss": 1.1747,
+      "step": 65
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.00019658484525080042,
+      "loss": 1.295,
+      "step": 66
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.0001965314834578442,
+      "loss": 1.2297,
+      "step": 67
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.00019647812166488795,
+      "loss": 1.3304,
+      "step": 68
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.0001964247598719317,
+      "loss": 0.5761,
+      "step": 69
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.00019637139807897547,
+      "loss": 1.3126,
+      "step": 70
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.0001963180362860192,
+      "loss": 1.1947,
+      "step": 71
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.000196264674493063,
+      "loss": 0.5905,
+      "step": 72
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.00019621131270010673,
+      "loss": 1.1473,
+      "step": 73
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.0001961579509071505,
+      "loss": 1.2785,
+      "step": 74
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.00019610458911419425,
+      "loss": 0.9917,
+      "step": 75
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.000196051227321238,
+      "loss": 1.0241,
+      "step": 76
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.00019599786552828178,
+      "loss": 1.208,
+      "step": 77
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.0001959445037353255,
+      "loss": 1.0015,
+      "step": 78
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.00019589114194236927,
+      "loss": 1.2541,
+      "step": 79
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.00019583778014941304,
+      "loss": 1.1985,
+      "step": 80
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.00019578441835645677,
+      "loss": 1.1642,
+      "step": 81
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.00019573105656350056,
+      "loss": 0.4965,
+      "step": 82
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.0001956776947705443,
+      "loss": 1.1031,
+      "step": 83
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.00019562433297758806,
+      "loss": 0.9948,
+      "step": 84
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.00019557097118463182,
+      "loss": 1.115,
+      "step": 85
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.00019551760939167555,
+      "loss": 0.5886,
+      "step": 86
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.00019546424759871934,
+      "loss": 1.2426,
+      "step": 87
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.00019541088580576308,
+      "loss": 0.5557,
+      "step": 88
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.00019535752401280684,
+      "loss": 1.1919,
+      "step": 89
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.0001953041622198506,
+      "loss": 1.0153,
+      "step": 90
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.00019525080042689434,
+      "loss": 1.1473,
+      "step": 91
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.0001951974386339381,
+      "loss": 1.1842,
+      "step": 92
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.00019514407684098186,
+      "loss": 1.1105,
+      "step": 93
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.00019509071504802562,
+      "loss": 1.0817,
+      "step": 94
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.00019503735325506938,
+      "loss": 1.1232,
+      "step": 95
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.00019498399146211315,
+      "loss": 1.2912,
+      "step": 96
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.00019493062966915688,
+      "loss": 1.1392,
+      "step": 97
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.00019487726787620064,
+      "loss": 0.4942,
+      "step": 98
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.0001948239060832444,
+      "loss": 1.1358,
+      "step": 99
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.00019477054429028817,
+      "loss": 1.2049,
+      "step": 100
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.00019471718249733193,
+      "loss": 1.1129,
+      "step": 101
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.00019466382070437566,
+      "loss": 1.1219,
+      "step": 102
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.00019461045891141943,
+      "loss": 1.3714,
+      "step": 103
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.0001945570971184632,
+      "loss": 1.2126,
+      "step": 104
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.00019450373532550695,
+      "loss": 1.0974,
+      "step": 105
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.0001944503735325507,
+      "loss": 1.0281,
+      "step": 106
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.00019439701173959445,
+      "loss": 1.1251,
+      "step": 107
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.0001943436499466382,
+      "loss": 1.1154,
+      "step": 108
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.00019429028815368197,
+      "loss": 1.1429,
+      "step": 109
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.00019423692636072573,
+      "loss": 1.0763,
+      "step": 110
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.0001941835645677695,
+      "loss": 1.0683,
+      "step": 111
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.00019413020277481323,
+      "loss": 1.133,
+      "step": 112
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.000194076840981857,
+      "loss": 1.0823,
+      "step": 113
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.00019402347918890075,
+      "loss": 0.9965,
+      "step": 114
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.00019397011739594452,
+      "loss": 1.1846,
+      "step": 115
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.00019391675560298828,
+      "loss": 1.1744,
+      "step": 116
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.000193863393810032,
+      "loss": 0.5205,
+      "step": 117
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.0001938100320170758,
+      "loss": 1.0192,
+      "step": 118
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.00019375667022411954,
+      "loss": 1.2701,
+      "step": 119
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.0001937033084311633,
+      "loss": 1.2086,
+      "step": 120
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.00019364994663820706,
+      "loss": 1.0203,
+      "step": 121
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.0001935965848452508,
+      "loss": 1.1033,
+      "step": 122
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.00019354322305229458,
+      "loss": 1.2379,
+      "step": 123
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.00019348986125933832,
+      "loss": 1.1425,
+      "step": 124
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.00019343649946638208,
+      "loss": 0.6124,
+      "step": 125
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.00019338313767342584,
+      "loss": 0.4575,
+      "step": 126
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.00019332977588046958,
+      "loss": 1.083,
+      "step": 127
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.00019327641408751337,
+      "loss": 1.1269,
+      "step": 128
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.0001932230522945571,
+      "loss": 1.0218,
+      "step": 129
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.00019316969050160086,
+      "loss": 1.1164,
+      "step": 130
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.00019311632870864463,
+      "loss": 1.1076,
+      "step": 131
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.00019306296691568836,
+      "loss": 1.2285,
+      "step": 132
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.00019300960512273215,
+      "loss": 1.1005,
+      "step": 133
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.00019295624332977588,
+      "loss": 1.1447,
+      "step": 134
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.00019290288153681965,
+      "loss": 1.1363,
+      "step": 135
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.0001928495197438634,
+      "loss": 0.9458,
+      "step": 136
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.00019279615795090714,
+      "loss": 1.0994,
+      "step": 137
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.00019274279615795093,
+      "loss": 1.2193,
+      "step": 138
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.00019268943436499467,
+      "loss": 1.1445,
+      "step": 139
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.00019263607257203843,
+      "loss": 1.0343,
+      "step": 140
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.0001925827107790822,
+      "loss": 0.5387,
+      "step": 141
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.00019252934898612593,
+      "loss": 1.1307,
+      "step": 142
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.00019247598719316972,
+      "loss": 0.9785,
+      "step": 143
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.00019242262540021345,
+      "loss": 1.0125,
+      "step": 144
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.0001923692636072572,
+      "loss": 1.3369,
+      "step": 145
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.00019231590181430097,
+      "loss": 1.0654,
+      "step": 146
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.0001922625400213447,
+      "loss": 1.1377,
+      "step": 147
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.0001922091782283885,
+      "loss": 0.6606,
+      "step": 148
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.00019215581643543223,
+      "loss": 1.0887,
+      "step": 149
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.000192102454642476,
+      "loss": 1.1166,
+      "step": 150
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.00019204909284951976,
+      "loss": 0.4397,
+      "step": 151
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.0001919957310565635,
+      "loss": 0.9683,
+      "step": 152
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.00019194236926360728,
+      "loss": 1.1446,
+      "step": 153
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.00019188900747065102,
+      "loss": 1.1153,
+      "step": 154
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.00019183564567769478,
+      "loss": 0.5556,
+      "step": 155
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.00019178228388473854,
+      "loss": 0.6059,
+      "step": 156
+    },
+    {
+      "epoch": 0.12,
+      "eval_loss": 1.1082383394241333,
+      "eval_runtime": 7808.506,
+      "eval_samples_per_second": 1.281,
+      "eval_steps_per_second": 0.08,
+      "step": 156
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.0001917289220917823,
+      "loss": 0.9494,
+      "step": 157
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.00019167556029882606,
+      "loss": 1.1537,
+      "step": 158
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.0001916221985058698,
+      "loss": 1.0104,
+      "step": 159
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.00019156883671291356,
+      "loss": 1.0434,
+      "step": 160
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.00019151547491995732,
+      "loss": 1.0498,
+      "step": 161
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.00019146211312700108,
+      "loss": 1.0603,
+      "step": 162
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.00019140875133404485,
+      "loss": 0.5679,
+      "step": 163
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.00019135538954108858,
+      "loss": 0.9749,
+      "step": 164
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.00019130202774813234,
+      "loss": 1.122,
+      "step": 165
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.0001912486659551761,
+      "loss": 1.2661,
+      "step": 166
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.00019119530416221987,
+      "loss": 1.1415,
+      "step": 167
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.00019114194236926363,
+      "loss": 1.1226,
+      "step": 168
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.00019108858057630736,
+      "loss": 1.0334,
+      "step": 169
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.00019103521878335113,
+      "loss": 1.0155,
+      "step": 170
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.0001909818569903949,
+      "loss": 1.2142,
+      "step": 171
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.00019092849519743865,
+      "loss": 1.2334,
+      "step": 172
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.00019087513340448238,
+      "loss": 1.0776,
+      "step": 173
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.00019082177161152615,
+      "loss": 1.0256,
+      "step": 174
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.0001907684098185699,
+      "loss": 1.1546,
+      "step": 175
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.00019071504802561367,
+      "loss": 1.077,
+      "step": 176
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.00019066168623265743,
+      "loss": 1.2244,
+      "step": 177
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.00019060832443970117,
+      "loss": 1.1666,
+      "step": 178
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.00019055496264674496,
+      "loss": 1.2009,
+      "step": 179
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.0001905016008537887,
+      "loss": 1.0807,
+      "step": 180
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.00019044823906083245,
+      "loss": 1.1708,
+      "step": 181
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.00019039487726787622,
+      "loss": 1.1276,
+      "step": 182
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.00019034151547491995,
+      "loss": 1.1783,
+      "step": 183
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.00019028815368196374,
+      "loss": 0.5062,
+      "step": 184
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.00019023479188900747,
+      "loss": 1.1359,
+      "step": 185
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.00019018143009605124,
+      "loss": 0.9623,
+      "step": 186
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.000190128068303095,
+      "loss": 1.1012,
+      "step": 187
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.00019007470651013873,
+      "loss": 1.1046,
+      "step": 188
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.00019002134471718252,
+      "loss": 1.1456,
+      "step": 189
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.00018996798292422626,
+      "loss": 1.2257,
+      "step": 190
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.00018991462113127002,
+      "loss": 1.0405,
+      "step": 191
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.00018986125933831378,
+      "loss": 0.9709,
+      "step": 192
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.00018980789754535752,
+      "loss": 1.1415,
+      "step": 193
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.0001897545357524013,
+      "loss": 1.0335,
+      "step": 194
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.00018970117395944504,
+      "loss": 1.1355,
+      "step": 195
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.0001896478121664888,
+      "loss": 1.08,
+      "step": 196
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.00018959445037353256,
+      "loss": 0.974,
+      "step": 197
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.0001895410885805763,
+      "loss": 1.1194,
+      "step": 198
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.0001894877267876201,
+      "loss": 1.0174,
+      "step": 199
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.00018943436499466382,
+      "loss": 1.2535,
+      "step": 200
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.00018938100320170758,
+      "loss": 1.26,
+      "step": 201
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.00018932764140875135,
+      "loss": 1.0057,
+      "step": 202
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.00018927427961579508,
+      "loss": 1.1845,
+      "step": 203
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.00018922091782283887,
+      "loss": 0.9873,
+      "step": 204
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.0001891675560298826,
+      "loss": 1.0804,
+      "step": 205
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.00018911419423692637,
+      "loss": 1.2365,
+      "step": 206
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.00018906083244397013,
+      "loss": 1.2553,
+      "step": 207
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.00018900747065101386,
+      "loss": 1.0746,
+      "step": 208
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.00018895410885805765,
+      "loss": 0.9582,
+      "step": 209
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.0001889007470651014,
+      "loss": 0.3971,
+      "step": 210
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.00018884738527214515,
+      "loss": 1.1823,
+      "step": 211
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.0001887940234791889,
+      "loss": 1.1189,
+      "step": 212
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.00018874066168623265,
+      "loss": 1.0236,
+      "step": 213
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.00018868729989327644,
+      "loss": 1.0577,
+      "step": 214
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.00018863393810032017,
+      "loss": 1.0348,
+      "step": 215
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.00018858057630736393,
+      "loss": 0.6006,
+      "step": 216
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.0001885272145144077,
+      "loss": 1.0172,
+      "step": 217
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.00018847385272145146,
+      "loss": 1.1215,
+      "step": 218
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.00018842049092849522,
+      "loss": 0.9849,
+      "step": 219
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.00018836712913553895,
+      "loss": 1.009,
+      "step": 220
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.00018831376734258272,
+      "loss": 1.1488,
+      "step": 221
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.00018826040554962648,
+      "loss": 0.5358,
+      "step": 222
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.00018820704375667024,
+      "loss": 1.0341,
+      "step": 223
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.000188153681963714,
+      "loss": 1.1925,
+      "step": 224
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.00018810032017075774,
+      "loss": 0.9815,
+      "step": 225
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.0001880469583778015,
+      "loss": 1.121,
+      "step": 226
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.00018799359658484526,
+      "loss": 1.0266,
+      "step": 227
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.00018794023479188902,
+      "loss": 1.2245,
+      "step": 228
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.00018788687299893278,
+      "loss": 0.538,
+      "step": 229
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.00018783351120597652,
+      "loss": 1.0804,
+      "step": 230
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.00018778014941302028,
+      "loss": 1.0971,
+      "step": 231
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.00018772678762006404,
+      "loss": 1.0585,
+      "step": 232
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.0001876734258271078,
+      "loss": 0.9903,
+      "step": 233
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.00018762006403415157,
+      "loss": 0.58,
+      "step": 234
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.0001875667022411953,
+      "loss": 0.542,
+      "step": 235
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.00018751334044823906,
+      "loss": 1.1035,
+      "step": 236
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.00018745997865528283,
+      "loss": 0.4969,
+      "step": 237
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.0001874066168623266,
+      "loss": 1.0733,
+      "step": 238
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.00018735325506937035,
+      "loss": 1.029,
+      "step": 239
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.0001872998932764141,
+      "loss": 1.0789,
+      "step": 240
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.00018724653148345785,
+      "loss": 1.0139,
+      "step": 241
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.0001871931696905016,
+      "loss": 0.8493,
+      "step": 242
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.00018713980789754537,
+      "loss": 1.0557,
+      "step": 243
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.00018708644610458913,
+      "loss": 1.1204,
+      "step": 244
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.0001870330843116329,
+      "loss": 1.0769,
+      "step": 245
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.00018697972251867663,
+      "loss": 0.5322,
+      "step": 246
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.0001869263607257204,
+      "loss": 1.1474,
+      "step": 247
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.00018687299893276415,
+      "loss": 0.9072,
+      "step": 248
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.00018681963713980791,
+      "loss": 1.0436,
+      "step": 249
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.00018676627534685168,
+      "loss": 0.4977,
+      "step": 250
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.0001867129135538954,
+      "loss": 1.1886,
+      "step": 251
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.00018665955176093917,
+      "loss": 0.9256,
+      "step": 252
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.00018660618996798294,
+      "loss": 1.0087,
+      "step": 253
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.00018655282817502667,
+      "loss": 0.9193,
+      "step": 254
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.00018649946638207046,
+      "loss": 1.131,
+      "step": 255
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.0001864461045891142,
+      "loss": 0.9315,
+      "step": 256
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.00018639274279615798,
+      "loss": 1.1075,
+      "step": 257
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.00018633938100320172,
+      "loss": 0.9938,
+      "step": 258
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.00018628601921024545,
+      "loss": 1.1761,
+      "step": 259
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.00018623265741728924,
+      "loss": 0.6541,
+      "step": 260
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.00018617929562433298,
+      "loss": 0.5358,
+      "step": 261
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.00018612593383137677,
+      "loss": 1.0229,
+      "step": 262
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.0001860725720384205,
+      "loss": 1.1881,
+      "step": 263
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.00018601921024546424,
+      "loss": 1.1074,
+      "step": 264
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.00018596584845250802,
+      "loss": 0.6958,
+      "step": 265
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.00018591248665955176,
+      "loss": 0.497,
+      "step": 266
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.00018585912486659552,
+      "loss": 1.069,
+      "step": 267
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.00018580576307363928,
+      "loss": 1.036,
+      "step": 268
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.00018575240128068302,
+      "loss": 1.0469,
+      "step": 269
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.0001856990394877268,
+      "loss": 1.0638,
+      "step": 270
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.00018564567769477054,
+      "loss": 1.1851,
+      "step": 271
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.0001855923159018143,
+      "loss": 1.0605,
+      "step": 272
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.00018553895410885807,
+      "loss": 0.4723,
+      "step": 273
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.0001854855923159018,
+      "loss": 1.0018,
+      "step": 274
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.0001854322305229456,
+      "loss": 1.1802,
+      "step": 275
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.00018537886872998933,
+      "loss": 0.9077,
+      "step": 276
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.0001853255069370331,
+      "loss": 1.2894,
+      "step": 277
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.00018527214514407685,
+      "loss": 0.9846,
+      "step": 278
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.0001852187833511206,
+      "loss": 1.012,
+      "step": 279
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.00018516542155816437,
+      "loss": 1.2537,
+      "step": 280
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.0001851120597652081,
+      "loss": 1.1025,
+      "step": 281
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.00018505869797225187,
+      "loss": 0.5568,
+      "step": 282
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.00018500533617929563,
+      "loss": 1.1068,
+      "step": 283
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.0001849519743863394,
+      "loss": 1.1539,
+      "step": 284
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.00018489861259338316,
+      "loss": 0.3933,
+      "step": 285
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.0001848452508004269,
+      "loss": 0.9782,
+      "step": 286
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.00018479188900747065,
+      "loss": 0.9916,
+      "step": 287
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.00018473852721451441,
+      "loss": 1.1263,
+      "step": 288
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.00018468516542155818,
+      "loss": 0.9692,
+      "step": 289
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.00018463180362860194,
+      "loss": 1.0155,
+      "step": 290
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.00018457844183564567,
+      "loss": 1.1557,
+      "step": 291
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.00018452508004268944,
+      "loss": 0.9329,
+      "step": 292
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.0001844717182497332,
+      "loss": 1.1454,
+      "step": 293
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.00018441835645677696,
+      "loss": 1.0124,
+      "step": 294
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.00018436499466382072,
+      "loss": 1.0529,
+      "step": 295
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.00018431163287086446,
+      "loss": 1.0127,
+      "step": 296
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.00018425827107790822,
+      "loss": 1.1314,
+      "step": 297
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.00018420490928495198,
+      "loss": 0.9605,
+      "step": 298
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.00018415154749199574,
+      "loss": 1.0364,
+      "step": 299
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.0001840981856990395,
+      "loss": 0.4641,
+      "step": 300
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.00018404482390608327,
+      "loss": 0.5049,
+      "step": 301
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.000183991462113127,
+      "loss": 1.1168,
+      "step": 302
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.00018393810032017076,
+      "loss": 0.9143,
+      "step": 303
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.00018388473852721452,
+      "loss": 1.0675,
+      "step": 304
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.0001838313767342583,
+      "loss": 1.0851,
+      "step": 305
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.00018377801494130205,
+      "loss": 1.037,
+      "step": 306
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.00018372465314834578,
+      "loss": 1.0835,
+      "step": 307
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.00018367129135538955,
+      "loss": 0.9686,
+      "step": 308
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.0001836179295624333,
+      "loss": 1.1968,
+      "step": 309
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.00018356456776947707,
+      "loss": 1.0021,
+      "step": 310
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.00018351120597652083,
+      "loss": 1.0176,
+      "step": 311
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.00018345784418356457,
+      "loss": 1.0027,
+      "step": 312
+    },
+    {
+      "epoch": 0.25,
+      "eval_loss": 1.061377763748169,
+      "eval_runtime": 7585.2316,
+      "eval_samples_per_second": 1.318,
+      "eval_steps_per_second": 0.082,
+      "step": 312
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.00018340448239060833,
+      "loss": 1.0882,
+      "step": 313
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.0001833511205976521,
+      "loss": 1.1581,
+      "step": 314
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.00018329775880469585,
+      "loss": 0.6106,
+      "step": 315
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.00018324439701173961,
+      "loss": 1.0043,
+      "step": 316
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.00018319103521878335,
+      "loss": 0.7147,
+      "step": 317
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.00018313767342582714,
+      "loss": 0.9992,
+      "step": 318
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.00018308431163287087,
+      "loss": 0.8509,
+      "step": 319
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.00018303094983991464,
+      "loss": 1.0959,
+      "step": 320
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.0001829775880469584,
+      "loss": 1.0642,
+      "step": 321
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.00018292422625400213,
+      "loss": 1.2037,
+      "step": 322
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.00018287086446104592,
+      "loss": 1.062,
+      "step": 323
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.00018281750266808966,
+      "loss": 0.8497,
+      "step": 324
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.00018276414087513342,
+      "loss": 1.1112,
+      "step": 325
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.00018271077908217718,
+      "loss": 1.2825,
+      "step": 326
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.00018265741728922091,
+      "loss": 0.9675,
+      "step": 327
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.0001826040554962647,
+      "loss": 0.9088,
+      "step": 328
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.00018255069370330844,
+      "loss": 1.1744,
+      "step": 329
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.0001824973319103522,
+      "loss": 1.1445,
+      "step": 330
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.00018244397011739596,
+      "loss": 1.1527,
+      "step": 331
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.0001823906083244397,
+      "loss": 0.9322,
+      "step": 332
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.0001823372465314835,
+      "loss": 1.1001,
+      "step": 333
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.00018228388473852722,
+      "loss": 0.5199,
+      "step": 334
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.00018223052294557096,
+      "loss": 0.9137,
+      "step": 335
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.00018217716115261475,
+      "loss": 1.0749,
+      "step": 336
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.00018212379935965848,
+      "loss": 1.1274,
+      "step": 337
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.00018207043756670227,
+      "loss": 1.1026,
+      "step": 338
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.000182017075773746,
+      "loss": 0.5273,
+      "step": 339
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.00018196371398078977,
+      "loss": 0.0,
+      "step": 340
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.00018191035218783353,
+      "loss": 1.1024,
+      "step": 341
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.00018185699039487726,
+      "loss": 0.5081,
+      "step": 342
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.00018180362860192105,
+      "loss": 1.2847,
+      "step": 343
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.0001817502668089648,
+      "loss": 0.596,
+      "step": 344
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.00018169690501600855,
+      "loss": 1.0607,
+      "step": 345
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.0001816435432230523,
+      "loss": 1.0758,
+      "step": 346
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.00018159018143009605,
+      "loss": 1.0138,
+      "step": 347
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.0001815368196371398,
+      "loss": 1.1433,
+      "step": 348
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.00018148345784418357,
+      "loss": 1.1391,
+      "step": 349
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.00018143009605122733,
+      "loss": 0.9667,
+      "step": 350
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.0001813767342582711,
+      "loss": 1.0281,
+      "step": 351
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.00018132337246531483,
+      "loss": 0.4442,
+      "step": 352
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.0001812700106723586,
+      "loss": 1.0739,
+      "step": 353
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.00018121664887940235,
+      "loss": 0.949,
+      "step": 354
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.00018116328708644611,
+      "loss": 1.0872,
+      "step": 355
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.00018110992529348988,
+      "loss": 0.9446,
+      "step": 356
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.0001810565635005336,
+      "loss": 0.4872,
+      "step": 357
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.00018100320170757737,
+      "loss": 0.6241,
+      "step": 358
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.00018094983991462114,
+      "loss": 0.9916,
+      "step": 359
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.0001808964781216649,
+      "loss": 0.5361,
+      "step": 360
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.00018084311632870866,
+      "loss": 1.1123,
+      "step": 361
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.00018078975453575242,
+      "loss": 1.2221,
+      "step": 362
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.00018073639274279616,
+      "loss": 0.4579,
+      "step": 363
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.00018068303094983992,
+      "loss": 1.1237,
+      "step": 364
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.00018062966915688368,
+      "loss": 1.0995,
+      "step": 365
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.00018057630736392744,
+      "loss": 0.5792,
+      "step": 366
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.0001805229455709712,
+      "loss": 1.0448,
+      "step": 367
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.00018046958377801494,
+      "loss": 1.1681,
+      "step": 368
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.0001804162219850587,
+      "loss": 1.162,
+      "step": 369
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.00018036286019210246,
+      "loss": 1.1036,
+      "step": 370
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.00018030949839914622,
+      "loss": 1.1347,
+      "step": 371
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.00018025613660618999,
+      "loss": 1.017,
+      "step": 372
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.00018020277481323372,
+      "loss": 1.0582,
+      "step": 373
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.00018014941302027748,
+      "loss": 0.5043,
+      "step": 374
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.00018009605122732125,
+      "loss": 1.0658,
+      "step": 375
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.000180042689434365,
+      "loss": 1.1595,
+      "step": 376
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.00017998932764140877,
+      "loss": 1.015,
+      "step": 377
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.0001799359658484525,
+      "loss": 0.5704,
+      "step": 378
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.0001798826040554963,
+      "loss": 1.0974,
+      "step": 379
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.00017982924226254003,
+      "loss": 1.1115,
+      "step": 380
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.0001797758804695838,
+      "loss": 1.0787,
+      "step": 381
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.00017972251867662755,
+      "loss": 1.0071,
+      "step": 382
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.0001796691568836713,
+      "loss": 1.1451,
+      "step": 383
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.00017961579509071508,
+      "loss": 1.0778,
+      "step": 384
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.0001795624332977588,
+      "loss": 1.2174,
+      "step": 385
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.00017950907150480257,
+      "loss": 0.3768,
+      "step": 386
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.00017945570971184633,
+      "loss": 1.2107,
+      "step": 387
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.00017940234791889007,
+      "loss": 1.1465,
+      "step": 388
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.00017934898612593386,
+      "loss": 0.45,
+      "step": 389
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.0001792956243329776,
+      "loss": 0.4428,
+      "step": 390
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.00017924226254002136,
+      "loss": 1.1696,
+      "step": 391
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.00017918890074706512,
+      "loss": 1.0768,
+      "step": 392
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.00017913553895410885,
+      "loss": 1.1521,
+      "step": 393
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.00017908217716115264,
+      "loss": 0.9681,
+      "step": 394
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.00017902881536819638,
+      "loss": 0.9624,
+      "step": 395
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.00017897545357524014,
+      "loss": 1.164,
+      "step": 396
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.0001789220917822839,
+      "loss": 1.0969,
+      "step": 397
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.00017886872998932763,
+      "loss": 0.5316,
+      "step": 398
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.00017881536819637142,
+      "loss": 0.6081,
+      "step": 399
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.00017876200640341516,
+      "loss": 0.3939,
+      "step": 400
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.00017870864461045892,
+      "loss": 0.5383,
+      "step": 401
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.00017865528281750268,
+      "loss": 0.9493,
+      "step": 402
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.00017860192102454642,
+      "loss": 1.1332,
+      "step": 403
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.0001785485592315902,
+      "loss": 0.4675,
+      "step": 404
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.00017849519743863394,
+      "loss": 1.0587,
+      "step": 405
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.0001784418356456777,
+      "loss": 0.5324,
+      "step": 406
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.00017838847385272147,
+      "loss": 1.005,
+      "step": 407
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.0001783351120597652,
+      "loss": 1.0999,
+      "step": 408
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.000178281750266809,
+      "loss": 1.1207,
+      "step": 409
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.00017822838847385272,
+      "loss": 0.901,
+      "step": 410
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.00017817502668089649,
+      "loss": 0.9547,
+      "step": 411
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.00017812166488794025,
+      "loss": 1.1402,
+      "step": 412
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.00017806830309498398,
+      "loss": 0.4569,
+      "step": 413
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.00017801494130202777,
+      "loss": 1.1237,
+      "step": 414
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.0001779615795090715,
+      "loss": 0.9173,
+      "step": 415
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.00017790821771611527,
+      "loss": 1.0195,
+      "step": 416
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.00017785485592315903,
+      "loss": 0.9325,
+      "step": 417
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.00017780149413020277,
+      "loss": 1.0993,
+      "step": 418
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.00017774813233724655,
+      "loss": 0.4352,
+      "step": 419
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.0001776947705442903,
+      "loss": 1.0424,
+      "step": 420
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.00017764140875133405,
+      "loss": 0.9037,
+      "step": 421
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.00017758804695837781,
+      "loss": 1.0729,
+      "step": 422
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.00017753468516542158,
+      "loss": 0.4921,
+      "step": 423
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.00017748132337246534,
+      "loss": 0.471,
+      "step": 424
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.00017742796157950907,
+      "loss": 1.0933,
+      "step": 425
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.00017737459978655283,
+      "loss": 0.9881,
+      "step": 426
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.0001773212379935966,
+      "loss": 1.1144,
+      "step": 427
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.00017726787620064036,
+      "loss": 0.4404,
+      "step": 428
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.0001772145144076841,
+      "loss": 0.9607,
+      "step": 429
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.00017716115261472786,
+      "loss": 0.8846,
+      "step": 430
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.00017710779082177162,
+      "loss": 0.4773,
+      "step": 431
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.00017705442902881538,
+      "loss": 0.4873,
+      "step": 432
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.00017700106723585914,
+      "loss": 0.9813,
+      "step": 433
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.00017694770544290288,
+      "loss": 1.1127,
+      "step": 434
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.00017689434364994664,
+      "loss": 1.0873,
+      "step": 435
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.0001768409818569904,
+      "loss": 1.1668,
+      "step": 436
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.00017678762006403416,
+      "loss": 0.9938,
+      "step": 437
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.00017673425827107792,
+      "loss": 0.4594,
+      "step": 438
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.00017668089647812166,
+      "loss": 1.1199,
+      "step": 439
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.00017662753468516545,
+      "loss": 0.484,
+      "step": 440
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.00017657417289220918,
+      "loss": 0.8651,
+      "step": 441
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.00017652081109925294,
+      "loss": 1.0724,
+      "step": 442
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.0001764674493062967,
+      "loss": 1.0886,
+      "step": 443
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.00017641408751334044,
+      "loss": 0.5544,
+      "step": 444
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.00017636072572038423,
+      "loss": 0.5309,
+      "step": 445
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.00017630736392742797,
+      "loss": 1.0713,
+      "step": 446
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.00017625400213447173,
+      "loss": 1.0705,
+      "step": 447
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.0001762006403415155,
+      "loss": 0.9475,
+      "step": 448
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.00017614727854855922,
+      "loss": 0.9265,
+      "step": 449
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.000176093916755603,
+      "loss": 0.9493,
+      "step": 450
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.00017604055496264675,
+      "loss": 1.118,
+      "step": 451
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.0001759871931696905,
+      "loss": 0.4692,
+      "step": 452
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.00017593383137673427,
+      "loss": 1.075,
+      "step": 453
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.000175880469583778,
+      "loss": 1.0884,
+      "step": 454
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.0001758271077908218,
+      "loss": 0.4346,
+      "step": 455
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.00017577374599786553,
+      "loss": 0.4745,
+      "step": 456
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.0001757203842049093,
+      "loss": 1.1329,
+      "step": 457
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.00017566702241195305,
+      "loss": 1.126,
+      "step": 458
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.0001756136606189968,
+      "loss": 0.5075,
+      "step": 459
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.00017556029882604058,
+      "loss": 0.9378,
+      "step": 460
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.00017550693703308431,
+      "loss": 1.0904,
+      "step": 461
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.00017545357524012808,
+      "loss": 1.0923,
+      "step": 462
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.00017540021344717184,
+      "loss": 0.877,
+      "step": 463
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.00017534685165421557,
+      "loss": 1.0494,
+      "step": 464
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.00017529348986125936,
+      "loss": 0.4913,
+      "step": 465
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.0001752401280683031,
+      "loss": 1.0446,
+      "step": 466
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.00017518676627534686,
+      "loss": 0.9659,
+      "step": 467
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.00017513340448239062,
+      "loss": 1.1297,
+      "step": 468
+    },
+    {
+      "epoch": 0.37,
+      "eval_loss": 1.0416920185089111,
+      "eval_runtime": 7645.782,
+      "eval_samples_per_second": 1.308,
+      "eval_steps_per_second": 0.082,
+      "step": 468
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.00017508004268943436,
+      "loss": 1.0977,
+      "step": 469
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.00017502668089647814,
+      "loss": 1.0705,
+      "step": 470
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.00017497331910352188,
+      "loss": 0.0,
+      "step": 471
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.00017491995731056564,
+      "loss": 0.9084,
+      "step": 472
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.0001748665955176094,
+      "loss": 1.0456,
+      "step": 473
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.00017481323372465314,
+      "loss": 1.0242,
+      "step": 474
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.00017475987193169693,
+      "loss": 0.958,
+      "step": 475
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.00017470651013874066,
+      "loss": 1.0935,
+      "step": 476
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.00017465314834578442,
+      "loss": 0.4378,
+      "step": 477
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.00017459978655282819,
+      "loss": 1.0872,
+      "step": 478
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.00017454642475987192,
+      "loss": 0.9914,
+      "step": 479
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.0001744930629669157,
+      "loss": 1.0239,
+      "step": 480
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.00017443970117395944,
+      "loss": 1.1007,
+      "step": 481
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.0001743863393810032,
+      "loss": 1.0081,
+      "step": 482
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.00017433297758804697,
+      "loss": 0.6599,
+      "step": 483
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.00017427961579509073,
+      "loss": 0.976,
+      "step": 484
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.0001742262540021345,
+      "loss": 0.5315,
+      "step": 485
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.00017417289220917823,
+      "loss": 1.1714,
+      "step": 486
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.000174119530416222,
+      "loss": 1.0608,
+      "step": 487
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.00017406616862326575,
+      "loss": 0.9823,
+      "step": 488
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.0001740128068303095,
+      "loss": 1.2254,
+      "step": 489
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.00017395944503735328,
+      "loss": 1.0277,
+      "step": 490
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.000173906083244397,
+      "loss": 1.0276,
+      "step": 491
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.00017385272145144077,
+      "loss": 0.5229,
+      "step": 492
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.00017379935965848453,
+      "loss": 1.0602,
+      "step": 493
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.0001737459978655283,
+      "loss": 1.1166,
+      "step": 494
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.00017369263607257206,
+      "loss": 1.1538,
+      "step": 495
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.0001736392742796158,
+      "loss": 0.8856,
+      "step": 496
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.00017358591248665955,
+      "loss": 0.6337,
+      "step": 497
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.00017353255069370332,
+      "loss": 1.0339,
+      "step": 498
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.00017347918890074708,
+      "loss": 0.4579,
+      "step": 499
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.00017342582710779084,
+      "loss": 0.9007,
+      "step": 500
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.00017337246531483458,
+      "loss": 0.4134,
+      "step": 501
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.00017331910352187834,
+      "loss": 1.1064,
+      "step": 502
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.0001732657417289221,
+      "loss": 1.0298,
+      "step": 503
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.00017321237993596586,
+      "loss": 0.4635,
+      "step": 504
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.00017315901814300962,
+      "loss": 0.999,
+      "step": 505
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.00017310565635005339,
+      "loss": 1.0121,
+      "step": 506
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.00017305229455709712,
+      "loss": 1.0084,
+      "step": 507
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.00017299893276414088,
+      "loss": 0.9686,
+      "step": 508
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.00017294557097118464,
+      "loss": 1.1039,
+      "step": 509
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.00017289220917822838,
+      "loss": 1.0879,
+      "step": 510
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.00017283884738527217,
+      "loss": 0.5787,
+      "step": 511
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.0001727854855923159,
+      "loss": 0.9498,
+      "step": 512
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.00017273212379935967,
+      "loss": 1.1038,
+      "step": 513
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.00017267876200640343,
+      "loss": 1.097,
+      "step": 514
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.00017262540021344716,
+      "loss": 1.0153,
+      "step": 515
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.00017257203842049095,
+      "loss": 1.0373,
+      "step": 516
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.00017251867662753469,
+      "loss": 1.1495,
+      "step": 517
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.00017246531483457845,
+      "loss": 1.1499,
+      "step": 518
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.0001724119530416222,
+      "loss": 0.4953,
+      "step": 519
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.00017235859124866594,
+      "loss": 0.0,
+      "step": 520
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.00017230522945570973,
+      "loss": 1.0972,
+      "step": 521
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.00017225186766275347,
+      "loss": 0.5496,
+      "step": 522
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.00017219850586979726,
+      "loss": 0.9975,
+      "step": 523
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.000172145144076841,
+      "loss": 0.8849,
+      "step": 524
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.00017209178228388473,
+      "loss": 1.073,
+      "step": 525
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.00017203842049092852,
+      "loss": 0.4959,
+      "step": 526
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.00017198505869797225,
+      "loss": 0.9048,
+      "step": 527
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.000171931696905016,
+      "loss": 0.9914,
+      "step": 528
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.00017187833511205978,
+      "loss": 1.0294,
+      "step": 529
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.0001718249733191035,
+      "loss": 0.5432,
+      "step": 530
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.0001717716115261473,
+      "loss": 1.0735,
+      "step": 531
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.00017171824973319103,
+      "loss": 1.1844,
+      "step": 532
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.0001716648879402348,
+      "loss": 0.9519,
+      "step": 533
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.00017161152614727856,
+      "loss": 0.6172,
+      "step": 534
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.0001715581643543223,
+      "loss": 1.0706,
+      "step": 535
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.00017150480256136608,
+      "loss": 0.5517,
+      "step": 536
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.00017145144076840982,
+      "loss": 1.0026,
+      "step": 537
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.00017139807897545358,
+      "loss": 1.0712,
+      "step": 538
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.00017134471718249734,
+      "loss": 0.639,
+      "step": 539
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.00017129135538954108,
+      "loss": 1.0147,
+      "step": 540
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.00017123799359658486,
+      "loss": 0.9226,
+      "step": 541
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.0001711846318036286,
+      "loss": 0.8838,
+      "step": 542
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.00017113127001067236,
+      "loss": 1.1968,
+      "step": 543
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.00017107790821771612,
+      "loss": 1.2515,
+      "step": 544
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.00017102454642475989,
+      "loss": 0.5444,
+      "step": 545
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.00017097118463180365,
+      "loss": 0.9173,
+      "step": 546
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.00017091782283884738,
+      "loss": 0.9666,
+      "step": 547
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.00017086446104589114,
+      "loss": 1.1568,
+      "step": 548
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.0001708110992529349,
+      "loss": 1.1426,
+      "step": 549
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.00017075773745997867,
+      "loss": 1.0291,
+      "step": 550
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.00017070437566702243,
+      "loss": 0.5584,
+      "step": 551
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.00017065101387406616,
+      "loss": 1.1671,
+      "step": 552
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.00017059765208110993,
+      "loss": 1.1641,
+      "step": 553
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.0001705442902881537,
+      "loss": 0.9845,
+      "step": 554
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.00017049092849519745,
+      "loss": 0.9024,
+      "step": 555
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.0001704375667022412,
+      "loss": 1.1306,
+      "step": 556
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.00017038420490928495,
+      "loss": 1.1239,
+      "step": 557
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.0001703308431163287,
+      "loss": 1.0562,
+      "step": 558
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.00017027748132337247,
+      "loss": 1.0007,
+      "step": 559
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.00017022411953041623,
+      "loss": 1.1607,
+      "step": 560
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.00017017075773746,
+      "loss": 1.1117,
+      "step": 561
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.00017011739594450373,
+      "loss": 0.4774,
+      "step": 562
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.0001700640341515475,
+      "loss": 1.0862,
+      "step": 563
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.00017001067235859125,
+      "loss": 0.4884,
+      "step": 564
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.00016995731056563502,
+      "loss": 0.9976,
+      "step": 565
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.00016990394877267878,
+      "loss": 1.0913,
+      "step": 566
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.00016985058697972254,
+      "loss": 1.0932,
+      "step": 567
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.00016979722518676628,
+      "loss": 1.09,
+      "step": 568
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.00016974386339381004,
+      "loss": 1.0882,
+      "step": 569
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.0001696905016008538,
+      "loss": 1.1009,
+      "step": 570
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.00016963713980789756,
+      "loss": 1.1179,
+      "step": 571
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.00016958377801494132,
+      "loss": 0.9515,
+      "step": 572
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.00016953041622198506,
+      "loss": 1.091,
+      "step": 573
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.00016947705442902882,
+      "loss": 1.0886,
+      "step": 574
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.00016942369263607258,
+      "loss": 1.1075,
+      "step": 575
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.00016937033084311634,
+      "loss": 1.0977,
+      "step": 576
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.0001693169690501601,
+      "loss": 0.5217,
+      "step": 577
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.00016926360725720384,
+      "loss": 0.5712,
+      "step": 578
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.0001692102454642476,
+      "loss": 0.978,
+      "step": 579
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.00016915688367129136,
+      "loss": 0.575,
+      "step": 580
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.00016910352187833513,
+      "loss": 1.1473,
+      "step": 581
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.0001690501600853789,
+      "loss": 0.5728,
+      "step": 582
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.00016899679829242262,
+      "loss": 1.1,
+      "step": 583
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.0001689434364994664,
+      "loss": 0.9962,
+      "step": 584
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.00016889007470651015,
+      "loss": 0.5251,
+      "step": 585
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.0001688367129135539,
+      "loss": 1.0941,
+      "step": 586
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.00016878335112059767,
+      "loss": 0.8959,
+      "step": 587
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.0001687299893276414,
+      "loss": 1.0659,
+      "step": 588
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.0001686766275346852,
+      "loss": 1.0254,
+      "step": 589
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.00016862326574172893,
+      "loss": 1.0832,
+      "step": 590
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.00016856990394877266,
+      "loss": 0.9456,
+      "step": 591
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.00016851654215581645,
+      "loss": 0.4577,
+      "step": 592
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.0001684631803628602,
+      "loss": 0.8165,
+      "step": 593
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.00016840981856990398,
+      "loss": 1.0459,
+      "step": 594
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.0001683564567769477,
+      "loss": 1.1722,
+      "step": 595
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.00016830309498399145,
+      "loss": 0.4605,
+      "step": 596
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.00016824973319103524,
+      "loss": 1.0344,
+      "step": 597
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.00016819637139807897,
+      "loss": 1.0925,
+      "step": 598
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.00016814300960512276,
+      "loss": 0.9867,
+      "step": 599
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.0001680896478121665,
+      "loss": 0.5835,
+      "step": 600
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.00016803628601921023,
+      "loss": 0.9548,
+      "step": 601
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.00016798292422625402,
+      "loss": 0.5163,
+      "step": 602
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.00016792956243329775,
+      "loss": 0.8802,
+      "step": 603
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.00016787620064034154,
+      "loss": 1.1112,
+      "step": 604
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.00016782283884738528,
+      "loss": 1.0443,
+      "step": 605
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.00016776947705442904,
+      "loss": 1.0097,
+      "step": 606
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.0001677161152614728,
+      "loss": 0.5051,
+      "step": 607
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.00016766275346851654,
+      "loss": 1.0562,
+      "step": 608
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.0001676093916755603,
+      "loss": 1.021,
+      "step": 609
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.00016755602988260406,
+      "loss": 0.8668,
+      "step": 610
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.00016750266808964782,
+      "loss": 1.0367,
+      "step": 611
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.00016744930629669158,
+      "loss": 1.0454,
+      "step": 612
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.00016739594450373532,
+      "loss": 0.4903,
+      "step": 613
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.00016734258271077908,
+      "loss": 1.0882,
+      "step": 614
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.00016728922091782284,
+      "loss": 0.9924,
+      "step": 615
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.0001672358591248666,
+      "loss": 1.0083,
+      "step": 616
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.00016718249733191037,
+      "loss": 1.0707,
+      "step": 617
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.0001671291355389541,
+      "loss": 0.9974,
+      "step": 618
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.00016707577374599786,
+      "loss": 1.0326,
+      "step": 619
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.00016702241195304163,
+      "loss": 0.5022,
+      "step": 620
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.0001669690501600854,
+      "loss": 0.4394,
+      "step": 621
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.00016691568836712915,
+      "loss": 0.9108,
+      "step": 622
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.00016686232657417289,
+      "loss": 0.5307,
+      "step": 623
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.00016680896478121665,
+      "loss": 1.139,
+      "step": 624
+    },
+    {
+      "epoch": 0.5,
+      "eval_loss": 1.0324418544769287,
+      "eval_runtime": 7693.2976,
+      "eval_samples_per_second": 1.3,
+      "eval_steps_per_second": 0.081,
+      "step": 624
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.0001667556029882604,
+      "loss": 0.4763,
+      "step": 625
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.00016670224119530417,
+      "loss": 0.9759,
+      "step": 626
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.00016664887940234793,
+      "loss": 0.5212,
+      "step": 627
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.0001665955176093917,
+      "loss": 0.9865,
+      "step": 628
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.00016654215581643543,
+      "loss": 0.9897,
+      "step": 629
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.0001664887940234792,
+      "loss": 0.5332,
+      "step": 630
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.00016643543223052295,
+      "loss": 0.9782,
+      "step": 631
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.00016638207043756672,
+      "loss": 0.8704,
+      "step": 632
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.00016632870864461048,
+      "loss": 0.8564,
+      "step": 633
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.0001662753468516542,
+      "loss": 1.0263,
+      "step": 634
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.00016622198505869797,
+      "loss": 1.0711,
+      "step": 635
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.00016616862326574174,
+      "loss": 1.0019,
+      "step": 636
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.0001661152614727855,
+      "loss": 0.5154,
+      "step": 637
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.00016606189967982926,
+      "loss": 1.0962,
+      "step": 638
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.000166008537886873,
+      "loss": 1.2123,
+      "step": 639
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.00016595517609391676,
+      "loss": 1.1048,
+      "step": 640
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.00016590181430096052,
+      "loss": 0.5041,
+      "step": 641
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.00016584845250800428,
+      "loss": 0.9918,
+      "step": 642
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.00016579509071504804,
+      "loss": 1.0451,
+      "step": 643
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 0.00016574172892209178,
+      "loss": 0.9378,
+      "step": 644
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 0.00016568836712913557,
+      "loss": 1.0238,
+      "step": 645
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 0.0001656350053361793,
+      "loss": 0.9408,
+      "step": 646
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 0.00016558164354322306,
+      "loss": 1.0064,
+      "step": 647
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 0.00016552828175026683,
+      "loss": 0.9393,
+      "step": 648
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 0.00016547491995731056,
+      "loss": 0.0,
+      "step": 649
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 0.00016542155816435435,
+      "loss": 1.1206,
+      "step": 650
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 0.00016536819637139808,
+      "loss": 0.9842,
+      "step": 651
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 0.00016531483457844185,
+      "loss": 0.9113,
+      "step": 652
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 0.0001652614727854856,
+      "loss": 0.8976,
+      "step": 653
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 0.00016520811099252934,
+      "loss": 1.1653,
+      "step": 654
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 0.00016515474919957313,
+      "loss": 1.0802,
+      "step": 655
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 0.00016510138740661687,
+      "loss": 1.0728,
+      "step": 656
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 0.00016504802561366063,
+      "loss": 0.9816,
+      "step": 657
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 0.0001649946638207044,
+      "loss": 1.0463,
+      "step": 658
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 0.00016494130202774813,
+      "loss": 1.161,
+      "step": 659
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 0.00016488794023479192,
+      "loss": 1.0802,
+      "step": 660
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 0.00016483457844183565,
+      "loss": 0.9969,
+      "step": 661
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 0.0001647812166488794,
+      "loss": 0.519,
+      "step": 662
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 0.00016472785485592317,
+      "loss": 0.9435,
+      "step": 663
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 0.0001646744930629669,
+      "loss": 1.0836,
+      "step": 664
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 0.0001646211312700107,
+      "loss": 0.9339,
+      "step": 665
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 0.00016456776947705443,
+      "loss": 0.5543,
+      "step": 666
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 0.0001645144076840982,
+      "loss": 1.0607,
+      "step": 667
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 0.00016446104589114196,
+      "loss": 0.9302,
+      "step": 668
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 0.0001644076840981857,
+      "loss": 0.8631,
+      "step": 669
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 0.00016435432230522948,
+      "loss": 1.1412,
+      "step": 670
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 0.00016430096051227322,
+      "loss": 1.1048,
+      "step": 671
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 0.00016424759871931698,
+      "loss": 0.9878,
+      "step": 672
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 0.00016419423692636074,
+      "loss": 1.0459,
+      "step": 673
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 0.00016414087513340447,
+      "loss": 1.0843,
+      "step": 674
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 0.00016408751334044826,
+      "loss": 1.1555,
+      "step": 675
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 0.000164034151547492,
+      "loss": 0.6055,
+      "step": 676
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 0.00016398078975453576,
+      "loss": 1.0291,
+      "step": 677
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 0.00016392742796157952,
+      "loss": 0.972,
+      "step": 678
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 0.00016387406616862326,
+      "loss": 1.0229,
+      "step": 679
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 0.00016382070437566705,
+      "loss": 0.9884,
+      "step": 680
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 0.00016376734258271078,
+      "loss": 0.5586,
+      "step": 681
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 0.00016371398078975454,
+      "loss": 0.4514,
+      "step": 682
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 0.0001636606189967983,
+      "loss": 0.9777,
+      "step": 683
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 0.00016360725720384204,
+      "loss": 1.1091,
+      "step": 684
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 0.00016355389541088583,
+      "loss": 1.1015,
+      "step": 685
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 0.00016350053361792956,
+      "loss": 1.1266,
+      "step": 686
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 0.00016344717182497333,
+      "loss": 1.0593,
+      "step": 687
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 0.0001633938100320171,
+      "loss": 1.0013,
+      "step": 688
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 0.00016334044823906085,
+      "loss": 1.0621,
+      "step": 689
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 0.00016328708644610458,
+      "loss": 1.0731,
+      "step": 690
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 0.00016323372465314835,
+      "loss": 1.0191,
+      "step": 691
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 0.0001631803628601921,
+      "loss": 0.5885,
+      "step": 692
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 0.00016312700106723587,
+      "loss": 1.0535,
+      "step": 693
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 0.00016307363927427963,
+      "loss": 0.995,
+      "step": 694
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 0.00016302027748132337,
+      "loss": 1.0865,
+      "step": 695
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 0.00016296691568836713,
+      "loss": 1.0961,
+      "step": 696
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 0.0001629135538954109,
+      "loss": 1.1594,
+      "step": 697
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 0.00016286019210245465,
+      "loss": 0.5644,
+      "step": 698
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 0.00016280683030949842,
+      "loss": 1.1907,
+      "step": 699
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 0.00016275346851654215,
+      "loss": 0.9697,
+      "step": 700
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 0.0001627001067235859,
+      "loss": 0.9149,
+      "step": 701
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 0.00016264674493062967,
+      "loss": 0.9315,
+      "step": 702
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 0.00016259338313767344,
+      "loss": 0.4553,
+      "step": 703
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 0.0001625400213447172,
+      "loss": 0.9045,
+      "step": 704
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 0.00016248665955176093,
+      "loss": 1.0346,
+      "step": 705
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 0.00016243329775880472,
+      "loss": 1.1302,
+      "step": 706
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 0.00016237993596584846,
+      "loss": 0.5481,
+      "step": 707
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 0.00016232657417289222,
+      "loss": 1.0604,
+      "step": 708
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 0.00016227321237993598,
+      "loss": 1.2022,
+      "step": 709
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 0.00016221985058697972,
+      "loss": 1.052,
+      "step": 710
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 0.0001621664887940235,
+      "loss": 0.4294,
+      "step": 711
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 0.00016211312700106724,
+      "loss": 0.4681,
+      "step": 712
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 0.000162059765208111,
+      "loss": 1.0656,
+      "step": 713
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 0.00016200640341515476,
+      "loss": 1.0664,
+      "step": 714
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 0.0001619530416221985,
+      "loss": 1.0269,
+      "step": 715
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 0.0001618996798292423,
+      "loss": 0.5399,
+      "step": 716
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 0.00016184631803628602,
+      "loss": 0.9102,
+      "step": 717
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 0.00016179295624332978,
+      "loss": 1.0963,
+      "step": 718
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 0.00016173959445037355,
+      "loss": 0.9108,
+      "step": 719
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 0.00016168623265741728,
+      "loss": 0.5802,
+      "step": 720
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 0.00016163287086446107,
+      "loss": 0.9575,
+      "step": 721
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 0.0001615795090715048,
+      "loss": 1.2637,
+      "step": 722
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 0.00016152614727854857,
+      "loss": 0.9947,
+      "step": 723
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 0.00016147278548559233,
+      "loss": 0.9765,
+      "step": 724
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 0.00016141942369263606,
+      "loss": 1.2414,
+      "step": 725
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 0.00016136606189967985,
+      "loss": 0.6237,
+      "step": 726
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 0.0001613127001067236,
+      "loss": 0.6022,
+      "step": 727
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 0.00016125933831376735,
+      "loss": 1.0344,
+      "step": 728
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 0.0001612059765208111,
+      "loss": 0.5557,
+      "step": 729
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 0.00016115261472785485,
+      "loss": 1.0574,
+      "step": 730
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 0.00016109925293489864,
+      "loss": 1.2399,
+      "step": 731
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 0.00016104589114194237,
+      "loss": 1.0246,
+      "step": 732
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 0.00016099252934898613,
+      "loss": 0.5686,
+      "step": 733
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 0.0001609391675560299,
+      "loss": 0.9374,
+      "step": 734
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 0.00016088580576307363,
+      "loss": 0.4838,
+      "step": 735
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 0.00016083244397011742,
+      "loss": 1.0791,
+      "step": 736
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 0.00016077908217716115,
+      "loss": 1.0243,
+      "step": 737
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 0.00016072572038420492,
+      "loss": 0.9554,
+      "step": 738
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 0.00016067235859124868,
+      "loss": 0.849,
+      "step": 739
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 0.0001606189967982924,
+      "loss": 1.1682,
+      "step": 740
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 0.0001605656350053362,
+      "loss": 1.0378,
+      "step": 741
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 0.00016051227321237994,
+      "loss": 0.4219,
+      "step": 742
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 0.0001604589114194237,
+      "loss": 0.5372,
+      "step": 743
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 0.00016040554962646746,
+      "loss": 1.0954,
+      "step": 744
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 0.0001603521878335112,
+      "loss": 1.1553,
+      "step": 745
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 0.00016029882604055498,
+      "loss": 0.8858,
+      "step": 746
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 0.00016024546424759872,
+      "loss": 0.9074,
+      "step": 747
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 0.00016019210245464248,
+      "loss": 0.9555,
+      "step": 748
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 0.00016013874066168624,
+      "loss": 0.5119,
+      "step": 749
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 0.00016008537886873,
+      "loss": 0.8394,
+      "step": 750
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 0.00016003201707577377,
+      "loss": 1.1788,
+      "step": 751
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 0.0001599786552828175,
+      "loss": 1.0103,
+      "step": 752
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 0.00015992529348986126,
+      "loss": 1.0731,
+      "step": 753
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 0.00015987193169690503,
+      "loss": 1.0132,
+      "step": 754
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 0.0001598185699039488,
+      "loss": 1.0449,
+      "step": 755
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 0.00015976520811099255,
+      "loss": 1.0034,
+      "step": 756
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 0.00015971184631803628,
+      "loss": 1.0551,
+      "step": 757
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 0.00015965848452508005,
+      "loss": 1.0514,
+      "step": 758
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 0.0001596051227321238,
+      "loss": 0.8679,
+      "step": 759
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 0.00015955176093916757,
+      "loss": 1.0695,
+      "step": 760
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 0.00015949839914621133,
+      "loss": 0.9836,
+      "step": 761
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 0.00015944503735325507,
+      "loss": 0.9552,
+      "step": 762
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 0.00015939167556029883,
+      "loss": 1.0349,
+      "step": 763
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 0.0001593383137673426,
+      "loss": 1.1175,
+      "step": 764
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 0.00015928495197438635,
+      "loss": 1.0847,
+      "step": 765
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 0.00015923159018143011,
+      "loss": 1.1216,
+      "step": 766
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 0.00015917822838847388,
+      "loss": 0.9448,
+      "step": 767
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 0.0001591248665955176,
+      "loss": 0.4849,
+      "step": 768
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 0.00015907150480256137,
+      "loss": 0.9747,
+      "step": 769
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 0.00015901814300960514,
+      "loss": 0.969,
+      "step": 770
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 0.00015896478121664887,
+      "loss": 1.0239,
+      "step": 771
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 0.00015891141942369266,
+      "loss": 0.0,
+      "step": 772
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 0.0001588580576307364,
+      "loss": 0.4925,
+      "step": 773
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 0.00015880469583778016,
+      "loss": 1.0909,
+      "step": 774
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 0.00015875133404482392,
+      "loss": 1.049,
+      "step": 775
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 0.00015869797225186765,
+      "loss": 0.5879,
+      "step": 776
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 0.00015864461045891144,
+      "loss": 0.9503,
+      "step": 777
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 0.00015859124866595518,
+      "loss": 1.0921,
+      "step": 778
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 0.00015853788687299894,
+      "loss": 1.0678,
+      "step": 779
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 0.0001584845250800427,
+      "loss": 1.0691,
+      "step": 780
+    },
+    {
+      "epoch": 0.62,
+      "eval_loss": 1.0126371383666992,
+      "eval_runtime": 7411.4661,
+      "eval_samples_per_second": 1.349,
+      "eval_steps_per_second": 0.084,
+      "step": 780
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 0.00015843116328708644,
+      "loss": 1.0329,
+      "step": 781
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 0.00015837780149413023,
+      "loss": 0.6579,
+      "step": 782
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 0.00015832443970117396,
+      "loss": 0.8682,
+      "step": 783
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 0.00015827107790821772,
+      "loss": 0.9793,
+      "step": 784
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 0.00015821771611526148,
+      "loss": 0.901,
+      "step": 785
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 0.00015816435432230522,
+      "loss": 1.0282,
+      "step": 786
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 0.000158110992529349,
+      "loss": 1.211,
+      "step": 787
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 0.00015805763073639274,
+      "loss": 1.0884,
+      "step": 788
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 0.0001580042689434365,
+      "loss": 0.9297,
+      "step": 789
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 0.00015795090715048027,
+      "loss": 1.0366,
+      "step": 790
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 0.000157897545357524,
+      "loss": 1.0916,
+      "step": 791
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 0.0001578441835645678,
+      "loss": 0.9871,
+      "step": 792
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 0.00015779082177161153,
+      "loss": 0.9455,
+      "step": 793
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 0.0001577374599786553,
+      "loss": 0.757,
+      "step": 794
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 0.00015768409818569905,
+      "loss": 1.0785,
+      "step": 795
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 0.00015763073639274278,
+      "loss": 1.1182,
+      "step": 796
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 0.00015757737459978657,
+      "loss": 1.0009,
+      "step": 797
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 0.0001575240128068303,
+      "loss": 1.0297,
+      "step": 798
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 0.00015747065101387407,
+      "loss": 0.9969,
+      "step": 799
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 0.00015741728922091783,
+      "loss": 0.6021,
+      "step": 800
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 0.00015736392742796157,
+      "loss": 1.053,
+      "step": 801
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 0.00015731056563500536,
+      "loss": 0.4809,
+      "step": 802
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 0.0001572572038420491,
+      "loss": 0.4817,
+      "step": 803
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 0.00015720384204909285,
+      "loss": 1.1404,
+      "step": 804
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 0.00015715048025613661,
+      "loss": 0.9509,
+      "step": 805
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 0.00015709711846318035,
+      "loss": 1.0033,
+      "step": 806
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 0.00015704375667022414,
+      "loss": 0.4916,
+      "step": 807
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 0.00015699039487726787,
+      "loss": 0.878,
+      "step": 808
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 0.00015693703308431164,
+      "loss": 0.4637,
+      "step": 809
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 0.0001568836712913554,
+      "loss": 1.0392,
+      "step": 810
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 0.00015683030949839916,
+      "loss": 1.0927,
+      "step": 811
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 0.00015677694770544292,
+      "loss": 1.1097,
+      "step": 812
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 0.00015672358591248666,
+      "loss": 1.0494,
+      "step": 813
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 0.00015667022411953042,
+      "loss": 1.0615,
+      "step": 814
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 0.00015661686232657418,
+      "loss": 0.9662,
+      "step": 815
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 0.00015656350053361794,
+      "loss": 0.9882,
+      "step": 816
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 0.0001565101387406617,
+      "loss": 0.9934,
+      "step": 817
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 0.00015645677694770544,
+      "loss": 0.5429,
+      "step": 818
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 0.0001564034151547492,
+      "loss": 1.0816,
+      "step": 819
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 0.00015635005336179296,
+      "loss": 1.1042,
+      "step": 820
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 0.00015629669156883673,
+      "loss": 0.0,
+      "step": 821
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 0.0001562433297758805,
+      "loss": 1.0767,
+      "step": 822
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 0.00015618996798292422,
+      "loss": 0.9884,
+      "step": 823
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 0.00015613660618996798,
+      "loss": 1.0118,
+      "step": 824
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 0.00015608324439701175,
+      "loss": 0.4776,
+      "step": 825
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 0.0001560298826040555,
+      "loss": 1.081,
+      "step": 826
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 0.00015597652081109927,
+      "loss": 0.9692,
+      "step": 827
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 0.00015592315901814303,
+      "loss": 0.9165,
+      "step": 828
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 0.00015586979722518677,
+      "loss": 1.0548,
+      "step": 829
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 0.00015581643543223053,
+      "loss": 0.9511,
+      "step": 830
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 0.0001557630736392743,
+      "loss": 1.0149,
+      "step": 831
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 0.00015570971184631805,
+      "loss": 0.0,
+      "step": 832
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 0.00015565635005336181,
+      "loss": 0.9583,
+      "step": 833
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 0.00015560298826040555,
+      "loss": 0.5001,
+      "step": 834
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 0.0001555496264674493,
+      "loss": 0.5705,
+      "step": 835
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 0.00015549626467449307,
+      "loss": 1.0148,
+      "step": 836
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 0.00015544290288153684,
+      "loss": 1.0967,
+      "step": 837
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 0.0001553895410885806,
+      "loss": 1.1149,
+      "step": 838
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 0.00015533617929562433,
+      "loss": 1.1094,
+      "step": 839
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 0.0001552828175026681,
+      "loss": 0.5016,
+      "step": 840
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 0.00015522945570971186,
+      "loss": 1.0922,
+      "step": 841
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 0.00015517609391675562,
+      "loss": 1.0357,
+      "step": 842
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 0.00015512273212379938,
+      "loss": 1.0708,
+      "step": 843
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 0.00015506937033084311,
+      "loss": 1.0069,
+      "step": 844
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 0.00015501600853788688,
+      "loss": 1.0105,
+      "step": 845
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 0.00015496264674493064,
+      "loss": 0.8718,
+      "step": 846
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 0.00015490928495197437,
+      "loss": 0.9742,
+      "step": 847
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 0.00015485592315901816,
+      "loss": 0.9387,
+      "step": 848
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 0.0001548025613660619,
+      "loss": 1.006,
+      "step": 849
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 0.0001547491995731057,
+      "loss": 1.0243,
+      "step": 850
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 0.00015469583778014942,
+      "loss": 0.5631,
+      "step": 851
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 0.00015464247598719316,
+      "loss": 1.0628,
+      "step": 852
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 0.00015458911419423695,
+      "loss": 0.9543,
+      "step": 853
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 0.00015453575240128068,
+      "loss": 1.1145,
+      "step": 854
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 0.00015448239060832447,
+      "loss": 0.5453,
+      "step": 855
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 0.0001544290288153682,
+      "loss": 1.0097,
+      "step": 856
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 0.00015437566702241194,
+      "loss": 1.1266,
+      "step": 857
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 0.00015432230522945573,
+      "loss": 1.1492,
+      "step": 858
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 0.00015426894343649946,
+      "loss": 1.0456,
+      "step": 859
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 0.00015421558164354325,
+      "loss": 0.9157,
+      "step": 860
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 0.000154162219850587,
+      "loss": 1.0461,
+      "step": 861
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 0.00015410885805763072,
+      "loss": 1.0926,
+      "step": 862
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 0.0001540554962646745,
+      "loss": 0.9818,
+      "step": 863
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 0.00015400213447171825,
+      "loss": 1.0394,
+      "step": 864
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 0.000153948772678762,
+      "loss": 0.3557,
+      "step": 865
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 0.00015389541088580577,
+      "loss": 1.2159,
+      "step": 866
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 0.0001538420490928495,
+      "loss": 1.2171,
+      "step": 867
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 0.0001537886872998933,
+      "loss": 1.0515,
+      "step": 868
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 0.00015373532550693703,
+      "loss": 1.0764,
+      "step": 869
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 0.0001536819637139808,
+      "loss": 0.5037,
+      "step": 870
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 0.00015362860192102455,
+      "loss": 0.9322,
+      "step": 871
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 0.00015357524012806831,
+      "loss": 1.0585,
+      "step": 872
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 0.00015352187833511208,
+      "loss": 1.0687,
+      "step": 873
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 0.0001534685165421558,
+      "loss": 1.0021,
+      "step": 874
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 0.00015341515474919957,
+      "loss": 1.0395,
+      "step": 875
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 0.00015336179295624334,
+      "loss": 1.0081,
+      "step": 876
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 0.0001533084311632871,
+      "loss": 0.9581,
+      "step": 877
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 0.00015325506937033086,
+      "loss": 1.072,
+      "step": 878
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 0.0001532017075773746,
+      "loss": 1.0238,
+      "step": 879
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 0.00015314834578441836,
+      "loss": 1.1553,
+      "step": 880
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 0.00015309498399146212,
+      "loss": 1.1387,
+      "step": 881
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 0.00015304162219850588,
+      "loss": 0.9974,
+      "step": 882
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 0.00015298826040554964,
+      "loss": 0.9595,
+      "step": 883
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 0.00015293489861259338,
+      "loss": 0.6062,
+      "step": 884
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 0.00015288153681963714,
+      "loss": 1.1244,
+      "step": 885
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 0.0001528281750266809,
+      "loss": 0.5634,
+      "step": 886
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 0.00015277481323372466,
+      "loss": 0.9954,
+      "step": 887
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 0.00015272145144076842,
+      "loss": 0.5323,
+      "step": 888
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 0.0001526680896478122,
+      "loss": 0.9408,
+      "step": 889
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 0.00015261472785485592,
+      "loss": 0.5052,
+      "step": 890
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 0.00015256136606189968,
+      "loss": 1.0724,
+      "step": 891
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 0.00015250800426894345,
+      "loss": 1.0296,
+      "step": 892
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 0.0001524546424759872,
+      "loss": 0.8167,
+      "step": 893
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 0.00015240128068303097,
+      "loss": 0.4784,
+      "step": 894
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 0.0001523479188900747,
+      "loss": 0.8929,
+      "step": 895
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 0.00015229455709711847,
+      "loss": 0.9722,
+      "step": 896
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 0.00015224119530416223,
+      "loss": 0.9025,
+      "step": 897
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 0.000152187833511206,
+      "loss": 1.2236,
+      "step": 898
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 0.00015213447171824975,
+      "loss": 0.8301,
+      "step": 899
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 0.0001520811099252935,
+      "loss": 0.942,
+      "step": 900
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 0.00015202774813233725,
+      "loss": 0.858,
+      "step": 901
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 0.000151974386339381,
+      "loss": 1.0273,
+      "step": 902
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 0.00015192102454642477,
+      "loss": 0.0,
+      "step": 903
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 0.00015186766275346853,
+      "loss": 1.0231,
+      "step": 904
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 0.00015181430096051227,
+      "loss": 0.8854,
+      "step": 905
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 0.00015176093916755603,
+      "loss": 0.9622,
+      "step": 906
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 0.0001517075773745998,
+      "loss": 0.988,
+      "step": 907
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 0.00015165421558164356,
+      "loss": 0.8837,
+      "step": 908
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 0.00015160085378868732,
+      "loss": 1.0911,
+      "step": 909
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 0.00015154749199573105,
+      "loss": 0.9309,
+      "step": 910
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 0.00015149413020277484,
+      "loss": 1.0552,
+      "step": 911
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 0.00015144076840981858,
+      "loss": 1.1842,
+      "step": 912
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 0.00015138740661686234,
+      "loss": 1.0977,
+      "step": 913
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 0.0001513340448239061,
+      "loss": 1.0087,
+      "step": 914
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 0.00015128068303094984,
+      "loss": 0.5042,
+      "step": 915
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 0.00015122732123799362,
+      "loss": 0.9916,
+      "step": 916
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 0.00015117395944503736,
+      "loss": 1.0364,
+      "step": 917
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 0.00015112059765208112,
+      "loss": 1.0322,
+      "step": 918
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 0.00015106723585912488,
+      "loss": 0.9738,
+      "step": 919
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 0.00015101387406616862,
+      "loss": 1.0488,
+      "step": 920
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 0.0001509605122732124,
+      "loss": 1.0533,
+      "step": 921
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 0.00015090715048025614,
+      "loss": 1.262,
+      "step": 922
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 0.0001508537886872999,
+      "loss": 0.9225,
+      "step": 923
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 0.00015080042689434367,
+      "loss": 0.9962,
+      "step": 924
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 0.0001507470651013874,
+      "loss": 0.4091,
+      "step": 925
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 0.0001506937033084312,
+      "loss": 1.1501,
+      "step": 926
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 0.00015064034151547492,
+      "loss": 0.8534,
+      "step": 927
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 0.00015058697972251866,
+      "loss": 1.0476,
+      "step": 928
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 0.00015053361792956245,
+      "loss": 1.1795,
+      "step": 929
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 0.00015048025613660618,
+      "loss": 1.1305,
+      "step": 930
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 0.00015042689434364997,
+      "loss": 0.4651,
+      "step": 931
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 0.0001503735325506937,
+      "loss": 0.9084,
+      "step": 932
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 0.00015032017075773747,
+      "loss": 0.9786,
+      "step": 933
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 0.00015026680896478123,
+      "loss": 1.0144,
+      "step": 934
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 0.00015021344717182497,
+      "loss": 0.9939,
+      "step": 935
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 0.00015016008537886876,
+      "loss": 1.1301,
+      "step": 936
+    },
+    {
+      "epoch": 0.75,
+      "eval_loss": 1.0134271383285522,
+      "eval_runtime": 7747.5,
+      "eval_samples_per_second": 1.291,
+      "eval_steps_per_second": 0.081,
+      "step": 936
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 0.0001501067235859125,
+      "loss": 0.9582,
+      "step": 937
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 0.00015005336179295625,
+      "loss": 0.5606,
+      "step": 938
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 0.00015000000000000001,
+      "loss": 0.9534,
+      "step": 939
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 0.00014994663820704375,
+      "loss": 1.0762,
+      "step": 940
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 0.00014989327641408754,
+      "loss": 1.0264,
+      "step": 941
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 0.00014983991462113127,
+      "loss": 1.1293,
+      "step": 942
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 0.00014978655282817503,
+      "loss": 0.5043,
+      "step": 943
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 0.0001497331910352188,
+      "loss": 0.9935,
+      "step": 944
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 0.00014967982924226253,
+      "loss": 0.8615,
+      "step": 945
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 0.0001496264674493063,
+      "loss": 1.0159,
+      "step": 946
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 0.00014957310565635006,
+      "loss": 0.8983,
+      "step": 947
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 0.00014951974386339382,
+      "loss": 1.1834,
+      "step": 948
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 0.00014946638207043758,
+      "loss": 1.0165,
+      "step": 949
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 0.00014941302027748134,
+      "loss": 0.945,
+      "step": 950
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 0.00014935965848452508,
+      "loss": 0.445,
+      "step": 951
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 0.00014930629669156884,
+      "loss": 0.9647,
+      "step": 952
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 0.0001492529348986126,
+      "loss": 0.8839,
+      "step": 953
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 0.00014919957310565636,
+      "loss": 0.5329,
+      "step": 954
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 0.00014914621131270012,
+      "loss": 0.5116,
+      "step": 955
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 0.00014909284951974386,
+      "loss": 1.145,
+      "step": 956
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 0.00014903948772678762,
+      "loss": 1.0478,
+      "step": 957
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 0.00014898612593383138,
+      "loss": 1.1366,
+      "step": 958
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 0.00014893276414087514,
+      "loss": 1.0773,
+      "step": 959
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 0.0001488794023479189,
+      "loss": 1.0069,
+      "step": 960
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 0.00014882604055496264,
+      "loss": 0.9631,
+      "step": 961
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 0.0001487726787620064,
+      "loss": 0.9943,
+      "step": 962
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 0.00014871931696905017,
+      "loss": 1.1106,
+      "step": 963
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 0.00014866595517609393,
+      "loss": 1.0695,
+      "step": 964
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 0.0001486125933831377,
+      "loss": 0.4745,
+      "step": 965
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 0.00014855923159018142,
+      "loss": 1.0845,
+      "step": 966
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 0.0001485058697972252,
+      "loss": 1.0355,
+      "step": 967
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 0.00014845250800426895,
+      "loss": 1.0491,
+      "step": 968
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 0.0001483991462113127,
+      "loss": 0.928,
+      "step": 969
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 0.00014834578441835647,
+      "loss": 0.5162,
+      "step": 970
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 0.0001482924226254002,
+      "loss": 1.0268,
+      "step": 971
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 0.000148239060832444,
+      "loss": 1.0318,
+      "step": 972
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 0.00014818569903948773,
+      "loss": 0.98,
+      "step": 973
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 0.0001481323372465315,
+      "loss": 0.9238,
+      "step": 974
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 0.00014807897545357526,
+      "loss": 0.9857,
+      "step": 975
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 0.000148025613660619,
+      "loss": 1.1403,
+      "step": 976
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 0.00014797225186766278,
+      "loss": 0.8808,
+      "step": 977
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 0.00014791889007470651,
+      "loss": 1.0693,
+      "step": 978
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 0.00014786552828175028,
+      "loss": 1.0816,
+      "step": 979
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 0.00014781216648879404,
+      "loss": 0.8874,
+      "step": 980
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 0.00014775880469583777,
+      "loss": 1.001,
+      "step": 981
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 0.00014770544290288156,
+      "loss": 0.8989,
+      "step": 982
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 0.0001476520811099253,
+      "loss": 1.0204,
+      "step": 983
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 0.00014759871931696906,
+      "loss": 1.0636,
+      "step": 984
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 0.00014754535752401282,
+      "loss": 1.0742,
+      "step": 985
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 0.00014749199573105656,
+      "loss": 1.0278,
+      "step": 986
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 0.00014743863393810034,
+      "loss": 0.9625,
+      "step": 987
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 0.00014738527214514408,
+      "loss": 1.1469,
+      "step": 988
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 0.00014733191035218784,
+      "loss": 1.0142,
+      "step": 989
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 0.0001472785485592316,
+      "loss": 1.0899,
+      "step": 990
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 0.00014722518676627534,
+      "loss": 1.0757,
+      "step": 991
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 0.00014717182497331913,
+      "loss": 1.0005,
+      "step": 992
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 0.00014711846318036286,
+      "loss": 0.9755,
+      "step": 993
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 0.00014706510138740662,
+      "loss": 1.0767,
+      "step": 994
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 0.00014701173959445039,
+      "loss": 0.9095,
+      "step": 995
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 0.00014695837780149412,
+      "loss": 0.4031,
+      "step": 996
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 0.0001469050160085379,
+      "loss": 0.9096,
+      "step": 997
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 0.00014685165421558164,
+      "loss": 1.0495,
+      "step": 998
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 0.0001467982924226254,
+      "loss": 1.0648,
+      "step": 999
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 0.00014674493062966917,
+      "loss": 1.0925,
+      "step": 1000
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 0.0001466915688367129,
+      "loss": 1.0427,
+      "step": 1001
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 0.0001466382070437567,
+      "loss": 1.1475,
+      "step": 1002
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 0.00014658484525080043,
+      "loss": 1.0127,
+      "step": 1003
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 0.0001465314834578442,
+      "loss": 0.0,
+      "step": 1004
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 0.00014647812166488795,
+      "loss": 1.019,
+      "step": 1005
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 0.00014642475987193169,
+      "loss": 0.9903,
+      "step": 1006
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 0.00014637139807897548,
+      "loss": 0.9035,
+      "step": 1007
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 0.0001463180362860192,
+      "loss": 0.9837,
+      "step": 1008
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 0.00014626467449306297,
+      "loss": 0.4798,
+      "step": 1009
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 0.00014621131270010673,
+      "loss": 0.5839,
+      "step": 1010
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 0.0001461579509071505,
+      "loss": 1.0004,
+      "step": 1011
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 0.00014610458911419426,
+      "loss": 1.047,
+      "step": 1012
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 0.000146051227321238,
+      "loss": 0.9386,
+      "step": 1013
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 0.00014599786552828176,
+      "loss": 1.0565,
+      "step": 1014
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 0.00014594450373532552,
+      "loss": 1.0494,
+      "step": 1015
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 0.00014589114194236928,
+      "loss": 0.5341,
+      "step": 1016
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 0.00014583778014941304,
+      "loss": 0.6025,
+      "step": 1017
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 0.00014578441835645678,
+      "loss": 0.5132,
+      "step": 1018
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 0.00014573105656350054,
+      "loss": 0.8616,
+      "step": 1019
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 0.0001456776947705443,
+      "loss": 1.2084,
+      "step": 1020
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 0.00014562433297758806,
+      "loss": 0.5997,
+      "step": 1021
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 0.00014557097118463182,
+      "loss": 1.1256,
+      "step": 1022
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 0.00014551760939167556,
+      "loss": 1.0581,
+      "step": 1023
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 0.00014546424759871932,
+      "loss": 0.9828,
+      "step": 1024
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 0.00014541088580576308,
+      "loss": 0.9353,
+      "step": 1025
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 0.00014535752401280684,
+      "loss": 0.4527,
+      "step": 1026
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 0.00014530416221985058,
+      "loss": 0.9478,
+      "step": 1027
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 0.00014525080042689434,
+      "loss": 0.9671,
+      "step": 1028
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 0.0001451974386339381,
+      "loss": 1.0632,
+      "step": 1029
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 0.00014514407684098187,
+      "loss": 0.9497,
+      "step": 1030
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 0.00014509071504802563,
+      "loss": 0.9879,
+      "step": 1031
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 0.00014503735325506936,
+      "loss": 0.6012,
+      "step": 1032
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 0.00014498399146211315,
+      "loss": 1.0932,
+      "step": 1033
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 0.00014493062966915689,
+      "loss": 1.1288,
+      "step": 1034
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 0.00014487726787620065,
+      "loss": 1.0945,
+      "step": 1035
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 0.0001448239060832444,
+      "loss": 1.0313,
+      "step": 1036
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 0.00014477054429028814,
+      "loss": 0.9115,
+      "step": 1037
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 0.00014471718249733193,
+      "loss": 1.0269,
+      "step": 1038
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 0.00014466382070437567,
+      "loss": 0.9252,
+      "step": 1039
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 0.00014461045891141943,
+      "loss": 1.0894,
+      "step": 1040
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 0.0001445570971184632,
+      "loss": 1.0725,
+      "step": 1041
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 0.00014450373532550693,
+      "loss": 1.1451,
+      "step": 1042
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 0.00014445037353255072,
+      "loss": 1.0423,
+      "step": 1043
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 0.00014439701173959445,
+      "loss": 0.9527,
+      "step": 1044
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 0.0001443436499466382,
+      "loss": 1.1023,
+      "step": 1045
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 0.00014429028815368198,
+      "loss": 0.9763,
+      "step": 1046
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 0.0001442369263607257,
+      "loss": 1.0052,
+      "step": 1047
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 0.0001441835645677695,
+      "loss": 1.1,
+      "step": 1048
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 0.00014413020277481323,
+      "loss": 1.0828,
+      "step": 1049
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 0.000144076840981857,
+      "loss": 1.0546,
+      "step": 1050
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 0.00014402347918890076,
+      "loss": 0.5143,
+      "step": 1051
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 0.0001439701173959445,
+      "loss": 0.9333,
+      "step": 1052
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 0.00014391675560298828,
+      "loss": 1.0661,
+      "step": 1053
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 0.00014386339381003202,
+      "loss": 0.9487,
+      "step": 1054
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 0.00014381003201707578,
+      "loss": 1.0821,
+      "step": 1055
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 0.00014375667022411954,
+      "loss": 1.0582,
+      "step": 1056
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 0.00014370330843116328,
+      "loss": 0.9907,
+      "step": 1057
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 0.00014364994663820706,
+      "loss": 0.9768,
+      "step": 1058
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 0.0001435965848452508,
+      "loss": 0.9033,
+      "step": 1059
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 0.00014354322305229456,
+      "loss": 1.0131,
+      "step": 1060
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 0.00014348986125933832,
+      "loss": 0.8933,
+      "step": 1061
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 0.00014343649946638206,
+      "loss": 1.0551,
+      "step": 1062
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 0.00014338313767342585,
+      "loss": 0.9622,
+      "step": 1063
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 0.00014332977588046958,
+      "loss": 1.1067,
+      "step": 1064
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 0.00014327641408751334,
+      "loss": 1.0019,
+      "step": 1065
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 0.0001432230522945571,
+      "loss": 1.0692,
+      "step": 1066
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 0.00014316969050160084,
+      "loss": 1.1178,
+      "step": 1067
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 0.00014311632870864463,
+      "loss": 0.9691,
+      "step": 1068
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 0.00014306296691568837,
+      "loss": 1.1678,
+      "step": 1069
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 0.00014300960512273213,
+      "loss": 1.1715,
+      "step": 1070
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 0.0001429562433297759,
+      "loss": 1.1242,
+      "step": 1071
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 0.00014290288153681965,
+      "loss": 0.4185,
+      "step": 1072
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 0.0001428495197438634,
+      "loss": 0.9074,
+      "step": 1073
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 0.00014279615795090715,
+      "loss": 0.8861,
+      "step": 1074
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 0.0001427427961579509,
+      "loss": 1.0168,
+      "step": 1075
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 0.00014268943436499467,
+      "loss": 0.988,
+      "step": 1076
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 0.00014263607257203843,
+      "loss": 0.9337,
+      "step": 1077
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 0.0001425827107790822,
+      "loss": 0.9532,
+      "step": 1078
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 0.00014252934898612593,
+      "loss": 0.9963,
+      "step": 1079
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 0.0001424759871931697,
+      "loss": 0.9427,
+      "step": 1080
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 0.00014242262540021345,
+      "loss": 1.0086,
+      "step": 1081
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 0.00014236926360725722,
+      "loss": 0.547,
+      "step": 1082
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 0.00014231590181430098,
+      "loss": 1.0772,
+      "step": 1083
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 0.0001422625400213447,
+      "loss": 0.9701,
+      "step": 1084
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 0.00014220917822838848,
+      "loss": 1.0098,
+      "step": 1085
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 0.00014215581643543224,
+      "loss": 0.8448,
+      "step": 1086
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 0.000142102454642476,
+      "loss": 0.8402,
+      "step": 1087
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 0.00014204909284951976,
+      "loss": 1.0383,
+      "step": 1088
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 0.0001419957310565635,
+      "loss": 1.1086,
+      "step": 1089
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 0.00014194236926360726,
+      "loss": 0.4384,
+      "step": 1090
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 0.00014188900747065102,
+      "loss": 0.966,
+      "step": 1091
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 0.00014183564567769478,
+      "loss": 1.0663,
+      "step": 1092
+    },
+    {
+      "epoch": 0.87,
+      "eval_loss": 0.9992431402206421,
+      "eval_runtime": 7470.9578,
+      "eval_samples_per_second": 1.339,
+      "eval_steps_per_second": 0.084,
+      "step": 1092
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 0.00014178228388473854,
+      "loss": 0.4867,
+      "step": 1093
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 0.0001417289220917823,
+      "loss": 0.8186,
+      "step": 1094
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 0.00014167556029882604,
+      "loss": 0.4386,
+      "step": 1095
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 0.0001416221985058698,
+      "loss": 1.0082,
+      "step": 1096
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 0.00014156883671291356,
+      "loss": 1.085,
+      "step": 1097
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 0.00014151547491995733,
+      "loss": 0.9514,
+      "step": 1098
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 0.0001414621131270011,
+      "loss": 0.8705,
+      "step": 1099
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 0.00014140875133404482,
+      "loss": 0.9373,
+      "step": 1100
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 0.00014135538954108859,
+      "loss": 0.9153,
+      "step": 1101
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 0.00014130202774813235,
+      "loss": 0.9372,
+      "step": 1102
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 0.0001412486659551761,
+      "loss": 1.0411,
+      "step": 1103
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 0.00014119530416221987,
+      "loss": 0.5953,
+      "step": 1104
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 0.0001411419423692636,
+      "loss": 1.0412,
+      "step": 1105
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 0.00014108858057630737,
+      "loss": 1.0046,
+      "step": 1106
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 0.00014103521878335113,
+      "loss": 0.9752,
+      "step": 1107
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 0.00014098185699039487,
+      "loss": 1.1188,
+      "step": 1108
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 0.00014092849519743865,
+      "loss": 1.1191,
+      "step": 1109
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 0.0001408751334044824,
+      "loss": 1.1046,
+      "step": 1110
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 0.00014082177161152615,
+      "loss": 1.0752,
+      "step": 1111
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 0.0001407684098185699,
+      "loss": 0.8701,
+      "step": 1112
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 0.00014071504802561365,
+      "loss": 0.9714,
+      "step": 1113
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 0.00014066168623265744,
+      "loss": 1.0512,
+      "step": 1114
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 0.00014060832443970117,
+      "loss": 0.9965,
+      "step": 1115
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 0.00014055496264674496,
+      "loss": 1.1963,
+      "step": 1116
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 0.0001405016008537887,
+      "loss": 1.3336,
+      "step": 1117
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 0.00014044823906083243,
+      "loss": 1.0086,
+      "step": 1118
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 0.00014039487726787622,
+      "loss": 0.9733,
+      "step": 1119
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 0.00014034151547491995,
+      "loss": 0.8798,
+      "step": 1120
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 0.00014028815368196372,
+      "loss": 1.1468,
+      "step": 1121
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 0.00014023479188900748,
+      "loss": 0.9594,
+      "step": 1122
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 0.0001401814300960512,
+      "loss": 1.1004,
+      "step": 1123
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 0.000140128068303095,
+      "loss": 1.1131,
+      "step": 1124
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 0.00014007470651013874,
+      "loss": 0.947,
+      "step": 1125
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 0.0001400213447171825,
+      "loss": 0.8907,
+      "step": 1126
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 0.00013996798292422626,
+      "loss": 1.1192,
+      "step": 1127
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 0.00013991462113127,
+      "loss": 1.0567,
+      "step": 1128
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 0.00013986125933831379,
+      "loss": 0.4861,
+      "step": 1129
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 0.00013980789754535752,
+      "loss": 0.9884,
+      "step": 1130
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 0.00013975453575240128,
+      "loss": 1.0053,
+      "step": 1131
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 0.00013970117395944504,
+      "loss": 1.1022,
+      "step": 1132
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 0.0001396478121664888,
+      "loss": 1.0057,
+      "step": 1133
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 0.00013959445037353257,
+      "loss": 0.996,
+      "step": 1134
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 0.0001395410885805763,
+      "loss": 0.5079,
+      "step": 1135
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 0.00013948772678762006,
+      "loss": 1.1166,
+      "step": 1136
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 0.00013943436499466383,
+      "loss": 1.0032,
+      "step": 1137
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 0.0001393810032017076,
+      "loss": 0.9624,
+      "step": 1138
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 0.00013932764140875135,
+      "loss": 0.8619,
+      "step": 1139
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 0.00013927427961579509,
+      "loss": 1.0412,
+      "step": 1140
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 0.00013922091782283885,
+      "loss": 0.9122,
+      "step": 1141
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 0.0001391675560298826,
+      "loss": 1.048,
+      "step": 1142
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 0.00013911419423692637,
+      "loss": 1.0417,
+      "step": 1143
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 0.00013906083244397013,
+      "loss": 0.9233,
+      "step": 1144
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 0.00013900747065101387,
+      "loss": 1.0966,
+      "step": 1145
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 0.00013895410885805763,
+      "loss": 1.0836,
+      "step": 1146
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 0.0001389007470651014,
+      "loss": 1.0188,
+      "step": 1147
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 0.00013884738527214515,
+      "loss": 1.0257,
+      "step": 1148
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 0.00013879402347918892,
+      "loss": 0.8841,
+      "step": 1149
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 0.00013874066168623265,
+      "loss": 0.5585,
+      "step": 1150
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 0.0001386872998932764,
+      "loss": 0.9663,
+      "step": 1151
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 0.00013863393810032017,
+      "loss": 1.1117,
+      "step": 1152
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 0.00013858057630736394,
+      "loss": 1.1031,
+      "step": 1153
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 0.0001385272145144077,
+      "loss": 0.9946,
+      "step": 1154
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 0.00013847385272145146,
+      "loss": 1.0987,
+      "step": 1155
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 0.0001384204909284952,
+      "loss": 1.1321,
+      "step": 1156
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 0.00013836712913553896,
+      "loss": 1.0098,
+      "step": 1157
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 0.00013831376734258272,
+      "loss": 1.1628,
+      "step": 1158
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 0.00013826040554962648,
+      "loss": 1.1084,
+      "step": 1159
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 0.00013820704375667024,
+      "loss": 0.4941,
+      "step": 1160
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 0.00013815368196371398,
+      "loss": 1.0579,
+      "step": 1161
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 0.00013810032017075774,
+      "loss": 1.0302,
+      "step": 1162
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 0.0001380469583778015,
+      "loss": 0.8658,
+      "step": 1163
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 0.00013799359658484526,
+      "loss": 1.0229,
+      "step": 1164
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 0.00013794023479188903,
+      "loss": 0.9376,
+      "step": 1165
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 0.00013788687299893276,
+      "loss": 1.053,
+      "step": 1166
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 0.00013783351120597652,
+      "loss": 0.8967,
+      "step": 1167
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 0.00013778014941302029,
+      "loss": 1.0042,
+      "step": 1168
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 0.00013772678762006405,
+      "loss": 0.8408,
+      "step": 1169
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 0.0001376734258271078,
+      "loss": 0.603,
+      "step": 1170
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 0.00013762006403415154,
+      "loss": 1.0225,
+      "step": 1171
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 0.0001375667022411953,
+      "loss": 0.9679,
+      "step": 1172
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 0.00013751334044823907,
+      "loss": 1.0626,
+      "step": 1173
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 0.00013745997865528283,
+      "loss": 0.9494,
+      "step": 1174
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 0.0001374066168623266,
+      "loss": 0.895,
+      "step": 1175
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 0.00013735325506937033,
+      "loss": 1.1,
+      "step": 1176
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 0.00013729989327641412,
+      "loss": 0.5616,
+      "step": 1177
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 0.00013724653148345785,
+      "loss": 0.9887,
+      "step": 1178
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 0.0001371931696905016,
+      "loss": 0.4414,
+      "step": 1179
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 0.00013713980789754537,
+      "loss": 0.5463,
+      "step": 1180
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 0.0001370864461045891,
+      "loss": 0.9723,
+      "step": 1181
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 0.0001370330843116329,
+      "loss": 0.813,
+      "step": 1182
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 0.00013697972251867663,
+      "loss": 0.4609,
+      "step": 1183
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 0.0001369263607257204,
+      "loss": 0.8537,
+      "step": 1184
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 0.00013687299893276416,
+      "loss": 0.8776,
+      "step": 1185
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 0.0001368196371398079,
+      "loss": 0.9736,
+      "step": 1186
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 0.00013676627534685168,
+      "loss": 1.0127,
+      "step": 1187
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 0.00013671291355389542,
+      "loss": 0.8555,
+      "step": 1188
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 0.00013665955176093915,
+      "loss": 0.9424,
+      "step": 1189
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 0.00013660618996798294,
+      "loss": 1.1072,
+      "step": 1190
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 0.00013655282817502667,
+      "loss": 0.586,
+      "step": 1191
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 0.00013649946638207046,
+      "loss": 0.9215,
+      "step": 1192
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 0.0001364461045891142,
+      "loss": 0.5003,
+      "step": 1193
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 0.00013639274279615796,
+      "loss": 1.0378,
+      "step": 1194
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 0.00013633938100320172,
+      "loss": 1.0935,
+      "step": 1195
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 0.00013628601921024546,
+      "loss": 0.4671,
+      "step": 1196
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 0.00013623265741728925,
+      "loss": 0.9028,
+      "step": 1197
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 0.00013617929562433298,
+      "loss": 0.9088,
+      "step": 1198
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 0.00013612593383137674,
+      "loss": 0.481,
+      "step": 1199
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 0.0001360725720384205,
+      "loss": 0.4536,
+      "step": 1200
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 0.00013601921024546424,
+      "loss": 0.9669,
+      "step": 1201
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 0.000135965848452508,
+      "loss": 0.873,
+      "step": 1202
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 0.00013591248665955176,
+      "loss": 0.9855,
+      "step": 1203
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 0.00013585912486659553,
+      "loss": 0.9402,
+      "step": 1204
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 0.0001358057630736393,
+      "loss": 0.9524,
+      "step": 1205
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 0.00013575240128068302,
+      "loss": 0.9984,
+      "step": 1206
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 0.00013569903948772678,
+      "loss": 0.5185,
+      "step": 1207
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 0.00013564567769477055,
+      "loss": 0.9278,
+      "step": 1208
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 0.0001355923159018143,
+      "loss": 1.0713,
+      "step": 1209
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 0.00013553895410885807,
+      "loss": 1.0537,
+      "step": 1210
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 0.0001354855923159018,
+      "loss": 1.17,
+      "step": 1211
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 0.00013543223052294557,
+      "loss": 1.0292,
+      "step": 1212
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 0.00013537886872998933,
+      "loss": 0.9722,
+      "step": 1213
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 0.0001353255069370331,
+      "loss": 0.8511,
+      "step": 1214
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 0.00013527214514407685,
+      "loss": 0.9515,
+      "step": 1215
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 0.00013521878335112062,
+      "loss": 0.8991,
+      "step": 1216
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 0.00013516542155816435,
+      "loss": 1.0182,
+      "step": 1217
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 0.0001351120597652081,
+      "loss": 0.5212,
+      "step": 1218
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 0.00013505869797225187,
+      "loss": 0.9403,
+      "step": 1219
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 0.00013500533617929564,
+      "loss": 0.8946,
+      "step": 1220
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 0.0001349519743863394,
+      "loss": 0.98,
+      "step": 1221
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 0.00013489861259338313,
+      "loss": 0.8032,
+      "step": 1222
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 0.0001348452508004269,
+      "loss": 1.0669,
+      "step": 1223
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 0.00013479188900747066,
+      "loss": 0.9896,
+      "step": 1224
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 0.00013473852721451442,
+      "loss": 0.9414,
+      "step": 1225
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 0.00013468516542155818,
+      "loss": 1.0967,
+      "step": 1226
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 0.00013463180362860192,
+      "loss": 0.0,
+      "step": 1227
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 0.00013457844183564568,
+      "loss": 0.8779,
+      "step": 1228
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 0.00013452508004268944,
+      "loss": 0.8589,
+      "step": 1229
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 0.0001344717182497332,
+      "loss": 1.1506,
+      "step": 1230
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 0.00013441835645677696,
+      "loss": 0.8691,
+      "step": 1231
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 0.0001343649946638207,
+      "loss": 1.033,
+      "step": 1232
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 0.00013431163287086446,
+      "loss": 1.062,
+      "step": 1233
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 0.00013425827107790822,
+      "loss": 0.518,
+      "step": 1234
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 0.00013420490928495198,
+      "loss": 1.1051,
+      "step": 1235
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 0.00013415154749199575,
+      "loss": 0.6192,
+      "step": 1236
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 0.00013409818569903948,
+      "loss": 0.8531,
+      "step": 1237
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 0.00013404482390608327,
+      "loss": 0.9213,
+      "step": 1238
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 0.000133991462113127,
+      "loss": 1.1007,
+      "step": 1239
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 0.00013393810032017077,
+      "loss": 0.8937,
+      "step": 1240
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 0.00013388473852721453,
+      "loss": 1.0594,
+      "step": 1241
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 0.00013383137673425826,
+      "loss": 0.9308,
+      "step": 1242
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 0.00013377801494130205,
+      "loss": 0.95,
+      "step": 1243
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 0.0001337246531483458,
+      "loss": 0.7238,
+      "step": 1244
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 0.00013367129135538955,
+      "loss": 1.0651,
+      "step": 1245
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 0.0001336179295624333,
+      "loss": 0.5158,
+      "step": 1246
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 0.00013356456776947705,
+      "loss": 0.9512,
+      "step": 1247
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 0.00013351120597652084,
+      "loss": 0.4826,
+      "step": 1248
+    },
+    {
+      "epoch": 1.0,
+      "eval_loss": 0.9929017424583435,
+      "eval_runtime": 7461.0267,
+      "eval_samples_per_second": 1.34,
+      "eval_steps_per_second": 0.084,
+      "step": 1248
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 0.00013345784418356457,
+      "loss": 1.0137,
+      "step": 1249
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 0.00013340448239060833,
+      "loss": 0.8201,
+      "step": 1250
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 0.0001333511205976521,
+      "loss": 0.9085,
+      "step": 1251
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 0.00013329775880469583,
+      "loss": 0.8492,
+      "step": 1252
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 0.00013324439701173962,
+      "loss": 0.8746,
+      "step": 1253
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 0.00013319103521878335,
+      "loss": 0.8142,
+      "step": 1254
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 0.00013313767342582712,
+      "loss": 0.8823,
+      "step": 1255
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 0.00013308431163287088,
+      "loss": 0.7729,
+      "step": 1256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 0.0001330309498399146,
+      "loss": 0.9025,
+      "step": 1257
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 0.0001329775880469584,
+      "loss": 0.8882,
+      "step": 1258
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 0.00013292422625400214,
+      "loss": 0.4586,
+      "step": 1259
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 0.0001328708644610459,
+      "loss": 0.8345,
+      "step": 1260
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 0.00013281750266808966,
+      "loss": 0.8789,
+      "step": 1261
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 0.0001327641408751334,
+      "loss": 0.8282,
+      "step": 1262
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 0.00013271077908217718,
+      "loss": 0.832,
+      "step": 1263
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 0.00013265741728922092,
+      "loss": 0.8707,
+      "step": 1264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 0.00013260405549626468,
+      "loss": 0.8917,
+      "step": 1265
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 0.00013255069370330844,
+      "loss": 0.3882,
+      "step": 1266
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 0.00013249733191035218,
+      "loss": 0.8699,
+      "step": 1267
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 0.00013244397011739597,
+      "loss": 1.0002,
+      "step": 1268
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 0.0001323906083244397,
+      "loss": 0.7834,
+      "step": 1269
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 0.00013233724653148346,
+      "loss": 0.8876,
+      "step": 1270
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 0.00013228388473852723,
+      "loss": 0.8235,
+      "step": 1271
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 0.00013223052294557096,
+      "loss": 0.7933,
+      "step": 1272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 0.00013217716115261475,
+      "loss": 0.8027,
+      "step": 1273
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 0.00013212379935965848,
+      "loss": 0.9239,
+      "step": 1274
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 0.00013207043756670225,
+      "loss": 0.8636,
+      "step": 1275
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 0.000132017075773746,
+      "loss": 0.8196,
+      "step": 1276
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 0.00013196371398078977,
+      "loss": 0.4623,
+      "step": 1277
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 0.00013191035218783353,
+      "loss": 0.3703,
+      "step": 1278
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 0.00013185699039487727,
+      "loss": 0.0,
+      "step": 1279
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 0.00013180362860192103,
+      "loss": 0.7787,
+      "step": 1280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 0.0001317502668089648,
+      "loss": 0.957,
+      "step": 1281
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 0.00013169690501600855,
+      "loss": 0.9059,
+      "step": 1282
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 0.0001316435432230523,
+      "loss": 0.7953,
+      "step": 1283
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 0.00013159018143009605,
+      "loss": 0.9349,
+      "step": 1284
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 0.0001315368196371398,
+      "loss": 0.8744,
+      "step": 1285
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 0.00013148345784418357,
+      "loss": 0.8202,
+      "step": 1286
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 0.00013143009605122734,
+      "loss": 0.8653,
+      "step": 1287
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 0.00013137673425827107,
+      "loss": 0.9884,
+      "step": 1288
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 0.00013132337246531483,
+      "loss": 0.8592,
+      "step": 1289
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 0.0001312700106723586,
+      "loss": 0.8642,
+      "step": 1290
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 0.00013121664887940236,
+      "loss": 0.8757,
+      "step": 1291
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 0.00013116328708644612,
+      "loss": 0.8991,
+      "step": 1292
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 0.00013110992529348985,
+      "loss": 0.9611,
+      "step": 1293
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 0.00013105656350053362,
+      "loss": 0.938,
+      "step": 1294
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 0.00013100320170757738,
+      "loss": 0.9546,
+      "step": 1295
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 0.00013094983991462114,
+      "loss": 0.8866,
+      "step": 1296
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 0.0001308964781216649,
+      "loss": 0.8269,
+      "step": 1297
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 0.00013084311632870864,
+      "loss": 0.0,
+      "step": 1298
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 0.00013078975453575243,
+      "loss": 0.689,
+      "step": 1299
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 0.00013073639274279616,
+      "loss": 0.6058,
+      "step": 1300
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 0.00013068303094983992,
+      "loss": 0.8578,
+      "step": 1301
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 0.00013062966915688368,
+      "loss": 0.8422,
+      "step": 1302
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 0.00013057630736392742,
+      "loss": 0.8456,
+      "step": 1303
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 0.0001305229455709712,
+      "loss": 1.0007,
+      "step": 1304
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 0.00013046958377801494,
+      "loss": 0.8144,
+      "step": 1305
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 0.0001304162219850587,
+      "loss": 0.85,
+      "step": 1306
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 0.00013036286019210247,
+      "loss": 0.9267,
+      "step": 1307
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 0.0001303094983991462,
+      "loss": 1.0948,
+      "step": 1308
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 0.00013025613660619,
+      "loss": 0.4134,
+      "step": 1309
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 0.00013020277481323373,
+      "loss": 0.9371,
+      "step": 1310
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 0.0001301494130202775,
+      "loss": 0.8692,
+      "step": 1311
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 0.00013009605122732125,
+      "loss": 1.0209,
+      "step": 1312
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 0.00013004268943436498,
+      "loss": 0.9861,
+      "step": 1313
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 0.00012998932764140877,
+      "loss": 0.926,
+      "step": 1314
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 0.0001299359658484525,
+      "loss": 0.4701,
+      "step": 1315
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 0.00012988260405549627,
+      "loss": 0.3699,
+      "step": 1316
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 0.00012982924226254003,
+      "loss": 0.9223,
+      "step": 1317
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 0.00012977588046958377,
+      "loss": 0.4722,
+      "step": 1318
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 0.00012972251867662756,
+      "loss": 0.9206,
+      "step": 1319
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 0.0001296691568836713,
+      "loss": 0.4964,
+      "step": 1320
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 0.00012961579509071505,
+      "loss": 0.7793,
+      "step": 1321
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 0.00012956243329775882,
+      "loss": 0.9424,
+      "step": 1322
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 0.00012950907150480255,
+      "loss": 1.0181,
+      "step": 1323
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 0.00012945570971184634,
+      "loss": 0.4783,
+      "step": 1324
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 0.00012940234791889007,
+      "loss": 0.749,
+      "step": 1325
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 0.00012934898612593384,
+      "loss": 0.9465,
+      "step": 1326
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 0.0001292956243329776,
+      "loss": 0.9187,
+      "step": 1327
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 0.00012924226254002133,
+      "loss": 0.9587,
+      "step": 1328
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 0.00012918890074706512,
+      "loss": 0.8469,
+      "step": 1329
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 0.00012913553895410886,
+      "loss": 0.8871,
+      "step": 1330
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 0.00012908217716115262,
+      "loss": 0.8263,
+      "step": 1331
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 0.00012902881536819638,
+      "loss": 0.8875,
+      "step": 1332
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 0.00012897545357524012,
+      "loss": 0.8185,
+      "step": 1333
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 0.0001289220917822839,
+      "loss": 0.9249,
+      "step": 1334
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 0.00012886872998932764,
+      "loss": 1.0406,
+      "step": 1335
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 0.0001288153681963714,
+      "loss": 0.4823,
+      "step": 1336
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 0.00012876200640341516,
+      "loss": 0.4568,
+      "step": 1337
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 0.00012870864461045893,
+      "loss": 0.7613,
+      "step": 1338
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 0.0001286552828175027,
+      "loss": 0.8353,
+      "step": 1339
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 0.00012860192102454642,
+      "loss": 0.508,
+      "step": 1340
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 0.00012854855923159018,
+      "loss": 0.87,
+      "step": 1341
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 0.00012849519743863395,
+      "loss": 0.7962,
+      "step": 1342
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 0.0001284418356456777,
+      "loss": 0.5013,
+      "step": 1343
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 0.00012838847385272147,
+      "loss": 0.4757,
+      "step": 1344
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 0.0001283351120597652,
+      "loss": 0.453,
+      "step": 1345
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 0.00012828175026680897,
+      "loss": 0.488,
+      "step": 1346
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 0.00012822838847385273,
+      "loss": 0.9087,
+      "step": 1347
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 0.0001281750266808965,
+      "loss": 0.9718,
+      "step": 1348
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 0.00012812166488794025,
+      "loss": 1.0564,
+      "step": 1349
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 0.000128068303094984,
+      "loss": 0.4675,
+      "step": 1350
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 0.00012801494130202775,
+      "loss": 0.9077,
+      "step": 1351
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 0.0001279615795090715,
+      "loss": 0.7227,
+      "step": 1352
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 0.00012790821771611527,
+      "loss": 0.8843,
+      "step": 1353
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 0.00012785485592315904,
+      "loss": 0.4639,
+      "step": 1354
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 0.00012780149413020277,
+      "loss": 0.8108,
+      "step": 1355
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 0.00012774813233724653,
+      "loss": 0.3159,
+      "step": 1356
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 0.0001276947705442903,
+      "loss": 0.888,
+      "step": 1357
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 0.00012764140875133406,
+      "loss": 1.0539,
+      "step": 1358
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 0.00012758804695837782,
+      "loss": 0.8193,
+      "step": 1359
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 0.00012753468516542158,
+      "loss": 0.9161,
+      "step": 1360
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 0.00012748132337246532,
+      "loss": 0.9063,
+      "step": 1361
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 0.00012742796157950908,
+      "loss": 0.9463,
+      "step": 1362
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 0.00012737459978655284,
+      "loss": 1.0409,
+      "step": 1363
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 0.00012732123799359657,
+      "loss": 0.8532,
+      "step": 1364
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 0.00012726787620064036,
+      "loss": 0.8823,
+      "step": 1365
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 0.0001272145144076841,
+      "loss": 0.956,
+      "step": 1366
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 0.00012716115261472786,
+      "loss": 0.8265,
+      "step": 1367
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 0.00012710779082177162,
+      "loss": 0.722,
+      "step": 1368
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 0.00012705442902881536,
+      "loss": 0.8585,
+      "step": 1369
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 0.00012700106723585915,
+      "loss": 0.9162,
+      "step": 1370
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 0.00012694770544290288,
+      "loss": 0.8274,
+      "step": 1371
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 0.00012689434364994664,
+      "loss": 0.8017,
+      "step": 1372
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 0.0001268409818569904,
+      "loss": 0.8936,
+      "step": 1373
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 0.00012678762006403414,
+      "loss": 0.4042,
+      "step": 1374
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 0.00012673425827107793,
+      "loss": 0.94,
+      "step": 1375
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 0.00012668089647812166,
+      "loss": 0.9356,
+      "step": 1376
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 0.00012662753468516543,
+      "loss": 0.8979,
+      "step": 1377
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 0.0001265741728922092,
+      "loss": 0.884,
+      "step": 1378
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 0.00012652081109925292,
+      "loss": 0.9782,
+      "step": 1379
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 0.0001264674493062967,
+      "loss": 0.843,
+      "step": 1380
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 0.00012641408751334045,
+      "loss": 0.8253,
+      "step": 1381
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 0.0001263607257203842,
+      "loss": 0.8375,
+      "step": 1382
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 0.00012630736392742797,
+      "loss": 0.8646,
+      "step": 1383
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 0.0001262540021344717,
+      "loss": 0.8128,
+      "step": 1384
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 0.0001262006403415155,
+      "loss": 0.9228,
+      "step": 1385
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 0.00012614727854855923,
+      "loss": 0.4619,
+      "step": 1386
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 0.000126093916755603,
+      "loss": 0.8552,
+      "step": 1387
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 0.00012604055496264675,
+      "loss": 0.7954,
+      "step": 1388
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 0.0001259871931696905,
+      "loss": 0.9201,
+      "step": 1389
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 0.00012593383137673428,
+      "loss": 0.7734,
+      "step": 1390
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 0.000125880469583778,
+      "loss": 0.8724,
+      "step": 1391
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 0.00012582710779082177,
+      "loss": 0.8402,
+      "step": 1392
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 0.00012577374599786554,
+      "loss": 0.8721,
+      "step": 1393
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 0.00012572038420490927,
+      "loss": 0.9225,
+      "step": 1394
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 0.00012566702241195306,
+      "loss": 0.8941,
+      "step": 1395
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 0.0001256136606189968,
+      "loss": 0.4001,
+      "step": 1396
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 0.00012556029882604056,
+      "loss": 0.7869,
+      "step": 1397
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 0.00012550693703308432,
+      "loss": 0.7927,
+      "step": 1398
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 0.00012545357524012808,
+      "loss": 0.9212,
+      "step": 1399
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 0.00012540021344717184,
+      "loss": 0.7395,
+      "step": 1400
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 0.00012534685165421558,
+      "loss": 1.0579,
+      "step": 1401
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 0.00012529348986125934,
+      "loss": 0.7112,
+      "step": 1402
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 0.0001252401280683031,
+      "loss": 0.8591,
+      "step": 1403
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 0.00012518676627534686,
+      "loss": 0.7781,
+      "step": 1404
+    },
+    {
+      "epoch": 1.12,
+      "eval_loss": 1.0063356161117554,
+      "eval_runtime": 7586.6912,
+      "eval_samples_per_second": 1.318,
+      "eval_steps_per_second": 0.082,
+      "step": 1404
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 0.00012513340448239062,
+      "loss": 0.7893,
+      "step": 1405
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 0.00012508004268943436,
+      "loss": 0.7654,
+      "step": 1406
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 0.00012502668089647812,
+      "loss": 0.8606,
+      "step": 1407
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 0.00012497331910352188,
+      "loss": 1.0433,
+      "step": 1408
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 0.00012491995731056565,
+      "loss": 0.8695,
+      "step": 1409
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 0.0001248665955176094,
+      "loss": 0.954,
+      "step": 1410
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 0.00012481323372465314,
+      "loss": 0.4456,
+      "step": 1411
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 0.0001247598719316969,
+      "loss": 0.8777,
+      "step": 1412
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 0.00012470651013874067,
+      "loss": 0.8512,
+      "step": 1413
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 0.00012465314834578443,
+      "loss": 0.8447,
+      "step": 1414
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 0.0001245997865528282,
+      "loss": 0.989,
+      "step": 1415
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 0.00012454642475987193,
+      "loss": 1.011,
+      "step": 1416
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 0.0001244930629669157,
+      "loss": 0.8999,
+      "step": 1417
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 0.00012443970117395945,
+      "loss": 0.4918,
+      "step": 1418
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 0.0001243863393810032,
+      "loss": 0.5677,
+      "step": 1419
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 0.00012433297758804697,
+      "loss": 0.8701,
+      "step": 1420
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 0.00012427961579509073,
+      "loss": 0.9587,
+      "step": 1421
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 0.00012422625400213447,
+      "loss": 0.9178,
+      "step": 1422
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 0.00012417289220917823,
+      "loss": 0.9091,
+      "step": 1423
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 0.000124119530416222,
+      "loss": 1.0468,
+      "step": 1424
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 0.00012406616862326576,
+      "loss": 0.842,
+      "step": 1425
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 0.00012401280683030952,
+      "loss": 0.877,
+      "step": 1426
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 0.00012395944503735325,
+      "loss": 0.8408,
+      "step": 1427
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 0.00012390608324439701,
+      "loss": 0.4075,
+      "step": 1428
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 0.00012385272145144078,
+      "loss": 1.0811,
+      "step": 1429
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 0.00012379935965848454,
+      "loss": 0.8761,
+      "step": 1430
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 0.0001237459978655283,
+      "loss": 0.9878,
+      "step": 1431
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 0.00012369263607257204,
+      "loss": 0.355,
+      "step": 1432
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 0.0001236392742796158,
+      "loss": 0.8794,
+      "step": 1433
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 0.00012358591248665956,
+      "loss": 0.3886,
+      "step": 1434
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 0.00012353255069370332,
+      "loss": 0.8672,
+      "step": 1435
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 0.00012347918890074708,
+      "loss": 0.7406,
+      "step": 1436
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 0.00012342582710779082,
+      "loss": 0.7086,
+      "step": 1437
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 0.00012337246531483458,
+      "loss": 1.0002,
+      "step": 1438
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 0.00012331910352187834,
+      "loss": 0.8608,
+      "step": 1439
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 0.0001232657417289221,
+      "loss": 0.7091,
+      "step": 1440
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 0.00012321237993596587,
+      "loss": 0.756,
+      "step": 1441
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 0.0001231590181430096,
+      "loss": 0.8388,
+      "step": 1442
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 0.0001231056563500534,
+      "loss": 0.4718,
+      "step": 1443
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 0.00012305229455709712,
+      "loss": 0.9323,
+      "step": 1444
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 0.00012299893276414086,
+      "loss": 0.9571,
+      "step": 1445
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 0.00012294557097118465,
+      "loss": 1.0927,
+      "step": 1446
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 0.00012289220917822838,
+      "loss": 0.8512,
+      "step": 1447
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 0.00012283884738527217,
+      "loss": 0.7472,
+      "step": 1448
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 0.0001227854855923159,
+      "loss": 0.9354,
+      "step": 1449
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 0.00012273212379935964,
+      "loss": 0.8479,
+      "step": 1450
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 0.00012267876200640343,
+      "loss": 0.8855,
+      "step": 1451
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 0.00012262540021344717,
+      "loss": 0.8796,
+      "step": 1452
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 0.00012257203842049096,
+      "loss": 0.8209,
+      "step": 1453
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 0.0001225186766275347,
+      "loss": 0.8339,
+      "step": 1454
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 0.00012246531483457843,
+      "loss": 0.9091,
+      "step": 1455
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 0.00012241195304162221,
+      "loss": 0.8095,
+      "step": 1456
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 0.00012235859124866595,
+      "loss": 0.9496,
+      "step": 1457
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 0.0001223052294557097,
+      "loss": 0.4381,
+      "step": 1458
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 0.00012225186766275347,
+      "loss": 0.8629,
+      "step": 1459
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 0.00012219850586979723,
+      "loss": 0.8725,
+      "step": 1460
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 0.000122145144076841,
+      "loss": 0.9059,
+      "step": 1461
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 0.00012209178228388473,
+      "loss": 0.8456,
+      "step": 1462
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 0.00012203842049092851,
+      "loss": 0.9624,
+      "step": 1463
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 0.00012198505869797226,
+      "loss": 0.3843,
+      "step": 1464
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 0.00012193169690501602,
+      "loss": 0.8961,
+      "step": 1465
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 0.00012187833511205977,
+      "loss": 0.9863,
+      "step": 1466
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 0.00012182497331910351,
+      "loss": 0.8795,
+      "step": 1467
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 0.00012177161152614729,
+      "loss": 1.104,
+      "step": 1468
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 0.00012171824973319104,
+      "loss": 0.8854,
+      "step": 1469
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 0.0001216648879402348,
+      "loss": 0.7585,
+      "step": 1470
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 0.00012161152614727855,
+      "loss": 0.9674,
+      "step": 1471
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 0.0001215581643543223,
+      "loss": 0.4067,
+      "step": 1472
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 0.00012150480256136607,
+      "loss": 0.4891,
+      "step": 1473
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 0.00012145144076840982,
+      "loss": 0.907,
+      "step": 1474
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 0.00012139807897545358,
+      "loss": 0.8781,
+      "step": 1475
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 0.00012134471718249733,
+      "loss": 0.4291,
+      "step": 1476
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 0.00012129135538954108,
+      "loss": 0.9259,
+      "step": 1477
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 0.00012123799359658486,
+      "loss": 0.8346,
+      "step": 1478
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 0.0001211846318036286,
+      "loss": 0.3869,
+      "step": 1479
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 0.00012113127001067237,
+      "loss": 0.888,
+      "step": 1480
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 0.00012107790821771611,
+      "loss": 0.4784,
+      "step": 1481
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 0.00012102454642475989,
+      "loss": 0.4682,
+      "step": 1482
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 0.00012097118463180364,
+      "loss": 0.8026,
+      "step": 1483
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 0.00012091782283884739,
+      "loss": 0.8931,
+      "step": 1484
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 0.00012086446104589115,
+      "loss": 0.4369,
+      "step": 1485
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 0.0001208110992529349,
+      "loss": 0.9108,
+      "step": 1486
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 0.00012075773745997867,
+      "loss": 0.8989,
+      "step": 1487
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 0.00012070437566702242,
+      "loss": 0.8637,
+      "step": 1488
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 0.00012065101387406617,
+      "loss": 0.8155,
+      "step": 1489
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 0.00012059765208110993,
+      "loss": 0.4708,
+      "step": 1490
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 0.00012054429028815368,
+      "loss": 0.8556,
+      "step": 1491
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 0.00012049092849519746,
+      "loss": 0.7542,
+      "step": 1492
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 0.0001204375667022412,
+      "loss": 0.9391,
+      "step": 1493
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 0.00012038420490928495,
+      "loss": 0.9102,
+      "step": 1494
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 0.00012033084311632871,
+      "loss": 0.752,
+      "step": 1495
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 0.00012027748132337246,
+      "loss": 0.8637,
+      "step": 1496
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 0.00012022411953041624,
+      "loss": 0.8059,
+      "step": 1497
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 0.00012017075773745999,
+      "loss": 0.7951,
+      "step": 1498
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 0.00012011739594450373,
+      "loss": 0.7915,
+      "step": 1499
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 0.0001200640341515475,
+      "loss": 0.7485,
+      "step": 1500
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 0.00012001067235859125,
+      "loss": 0.925,
+      "step": 1501
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 0.00011995731056563502,
+      "loss": 0.9402,
+      "step": 1502
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 0.00011990394877267877,
+      "loss": 0.8412,
+      "step": 1503
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 0.00011985058697972253,
+      "loss": 0.8805,
+      "step": 1504
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 0.00011979722518676628,
+      "loss": 0.3785,
+      "step": 1505
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 0.00011974386339381003,
+      "loss": 0.9524,
+      "step": 1506
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 0.0001196905016008538,
+      "loss": 0.8691,
+      "step": 1507
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 0.00011963713980789755,
+      "loss": 0.7829,
+      "step": 1508
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 0.00011958377801494131,
+      "loss": 0.9056,
+      "step": 1509
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 0.00011953041622198506,
+      "loss": 0.573,
+      "step": 1510
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 0.00011947705442902881,
+      "loss": 0.722,
+      "step": 1511
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 0.00011942369263607259,
+      "loss": 0.8298,
+      "step": 1512
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 0.00011937033084311633,
+      "loss": 0.887,
+      "step": 1513
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 0.0001193169690501601,
+      "loss": 1.0447,
+      "step": 1514
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 0.00011926360725720385,
+      "loss": 0.9466,
+      "step": 1515
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 0.0001192102454642476,
+      "loss": 0.8716,
+      "step": 1516
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 0.00011915688367129137,
+      "loss": 0.8747,
+      "step": 1517
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 0.00011910352187833512,
+      "loss": 0.8173,
+      "step": 1518
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 0.00011905016008537888,
+      "loss": 0.9705,
+      "step": 1519
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 0.00011899679829242263,
+      "loss": 0.859,
+      "step": 1520
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 0.0001189434364994664,
+      "loss": 0.7728,
+      "step": 1521
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 0.00011889007470651015,
+      "loss": 0.9839,
+      "step": 1522
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 0.0001188367129135539,
+      "loss": 0.441,
+      "step": 1523
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 0.00011878335112059766,
+      "loss": 0.7104,
+      "step": 1524
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 0.00011872998932764141,
+      "loss": 0.8919,
+      "step": 1525
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 0.00011867662753468519,
+      "loss": 0.515,
+      "step": 1526
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 0.00011862326574172893,
+      "loss": 0.8686,
+      "step": 1527
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 0.00011856990394877267,
+      "loss": 0.8811,
+      "step": 1528
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 0.00011851654215581644,
+      "loss": 0.8584,
+      "step": 1529
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 0.0001184631803628602,
+      "loss": 0.8976,
+      "step": 1530
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 0.00011840981856990397,
+      "loss": 0.8456,
+      "step": 1531
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 0.00011835645677694772,
+      "loss": 0.4022,
+      "step": 1532
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 0.00011830309498399145,
+      "loss": 0.8688,
+      "step": 1533
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 0.00011824973319103523,
+      "loss": 0.371,
+      "step": 1534
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 0.00011819637139807898,
+      "loss": 0.8851,
+      "step": 1535
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 0.00011814300960512275,
+      "loss": 0.7682,
+      "step": 1536
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 0.00011808964781216649,
+      "loss": 0.8643,
+      "step": 1537
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 0.00011803628601921023,
+      "loss": 0.7461,
+      "step": 1538
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 0.00011798292422625401,
+      "loss": 0.5474,
+      "step": 1539
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 0.00011792956243329776,
+      "loss": 0.9887,
+      "step": 1540
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 0.00011787620064034153,
+      "loss": 0.9513,
+      "step": 1541
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 0.00011782283884738527,
+      "loss": 0.7677,
+      "step": 1542
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 0.00011776947705442904,
+      "loss": 0.9572,
+      "step": 1543
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 0.00011771611526147279,
+      "loss": 0.9447,
+      "step": 1544
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 0.00011766275346851654,
+      "loss": 0.9586,
+      "step": 1545
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 0.0001176093916755603,
+      "loss": 0.3773,
+      "step": 1546
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 0.00011755602988260405,
+      "loss": 0.7902,
+      "step": 1547
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 0.00011750266808964783,
+      "loss": 1.0833,
+      "step": 1548
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 0.00011744930629669158,
+      "loss": 0.8154,
+      "step": 1549
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 0.00011739594450373532,
+      "loss": 0.8196,
+      "step": 1550
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 0.00011734258271077909,
+      "loss": 0.9038,
+      "step": 1551
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 0.00011728922091782283,
+      "loss": 0.8647,
+      "step": 1552
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 0.00011723585912486661,
+      "loss": 0.9396,
+      "step": 1553
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 0.00011718249733191036,
+      "loss": 0.3881,
+      "step": 1554
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 0.00011712913553895411,
+      "loss": 0.8572,
+      "step": 1555
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 0.00011707577374599787,
+      "loss": 0.8467,
+      "step": 1556
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 0.00011702241195304162,
+      "loss": 0.7851,
+      "step": 1557
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 0.00011696905016008539,
+      "loss": 0.8379,
+      "step": 1558
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 0.00011691568836712914,
+      "loss": 0.8585,
+      "step": 1559
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 0.00011686232657417289,
+      "loss": 0.7545,
+      "step": 1560
+    },
+    {
+      "epoch": 1.25,
+      "eval_loss": 1.0079351663589478,
+      "eval_runtime": 7638.6414,
+      "eval_samples_per_second": 1.309,
+      "eval_steps_per_second": 0.082,
+      "step": 1560
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 0.00011680896478121665,
+      "loss": 0.8977,
+      "step": 1561
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 0.0001167556029882604,
+      "loss": 0.8243,
+      "step": 1562
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 0.00011670224119530418,
+      "loss": 0.8968,
+      "step": 1563
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 0.00011664887940234792,
+      "loss": 0.4638,
+      "step": 1564
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 0.00011659551760939169,
+      "loss": 0.9026,
+      "step": 1565
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 0.00011654215581643543,
+      "loss": 0.744,
+      "step": 1566
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 0.00011648879402347918,
+      "loss": 0.949,
+      "step": 1567
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 0.00011643543223052296,
+      "loss": 0.8913,
+      "step": 1568
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 0.0001163820704375667,
+      "loss": 0.9044,
+      "step": 1569
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 0.00011632870864461047,
+      "loss": 0.4823,
+      "step": 1570
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 0.00011627534685165422,
+      "loss": 0.9303,
+      "step": 1571
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 0.00011622198505869797,
+      "loss": 0.7867,
+      "step": 1572
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 0.00011616862326574174,
+      "loss": 0.7273,
+      "step": 1573
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 0.00011611526147278549,
+      "loss": 0.996,
+      "step": 1574
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 0.00011606189967982925,
+      "loss": 0.8927,
+      "step": 1575
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 0.000116008537886873,
+      "loss": 0.8964,
+      "step": 1576
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 0.00011595517609391675,
+      "loss": 0.8129,
+      "step": 1577
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 0.00011590181430096052,
+      "loss": 0.8344,
+      "step": 1578
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 0.00011584845250800427,
+      "loss": 0.9367,
+      "step": 1579
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 0.00011579509071504803,
+      "loss": 0.8866,
+      "step": 1580
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 0.00011574172892209178,
+      "loss": 0.8083,
+      "step": 1581
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 0.00011568836712913556,
+      "loss": 1.0552,
+      "step": 1582
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 0.0001156350053361793,
+      "loss": 0.9335,
+      "step": 1583
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 0.00011558164354322305,
+      "loss": 0.8835,
+      "step": 1584
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 0.00011552828175026682,
+      "loss": 0.8223,
+      "step": 1585
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 0.00011547491995731057,
+      "loss": 0.418,
+      "step": 1586
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 0.00011542155816435434,
+      "loss": 0.4517,
+      "step": 1587
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 0.00011536819637139809,
+      "loss": 0.4115,
+      "step": 1588
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 0.00011531483457844184,
+      "loss": 0.8058,
+      "step": 1589
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 0.0001152614727854856,
+      "loss": 0.9118,
+      "step": 1590
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 0.00011520811099252935,
+      "loss": 0.8723,
+      "step": 1591
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 0.00011515474919957312,
+      "loss": 0.8345,
+      "step": 1592
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 0.00011510138740661687,
+      "loss": 0.8923,
+      "step": 1593
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 0.00011504802561366062,
+      "loss": 0.9109,
+      "step": 1594
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 0.00011499466382070438,
+      "loss": 0.9467,
+      "step": 1595
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 0.00011494130202774813,
+      "loss": 0.8848,
+      "step": 1596
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 0.0001148879402347919,
+      "loss": 0.0,
+      "step": 1597
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 0.00011483457844183565,
+      "loss": 0.8814,
+      "step": 1598
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 0.0001147812166488794,
+      "loss": 0.8503,
+      "step": 1599
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 0.00011472785485592317,
+      "loss": 0.8714,
+      "step": 1600
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 0.00011467449306296691,
+      "loss": 0.0,
+      "step": 1601
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 0.00011462113127001069,
+      "loss": 0.8605,
+      "step": 1602
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 0.00011456776947705444,
+      "loss": 0.9107,
+      "step": 1603
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 0.0001145144076840982,
+      "loss": 0.8477,
+      "step": 1604
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 0.00011446104589114195,
+      "loss": 1.023,
+      "step": 1605
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 0.0001144076840981857,
+      "loss": 0.9452,
+      "step": 1606
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 0.00011435432230522947,
+      "loss": 0.6586,
+      "step": 1607
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 0.00011430096051227322,
+      "loss": 0.9294,
+      "step": 1608
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 0.00011424759871931698,
+      "loss": 1.0008,
+      "step": 1609
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 0.00011419423692636073,
+      "loss": 0.4311,
+      "step": 1610
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 0.00011414087513340448,
+      "loss": 0.5121,
+      "step": 1611
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 0.00011408751334044825,
+      "loss": 1.0271,
+      "step": 1612
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 0.000114034151547492,
+      "loss": 0.7657,
+      "step": 1613
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 0.00011398078975453576,
+      "loss": 0.8402,
+      "step": 1614
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 0.00011392742796157951,
+      "loss": 0.8389,
+      "step": 1615
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 0.00011387406616862326,
+      "loss": 0.8655,
+      "step": 1616
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 0.00011382070437566704,
+      "loss": 0.8399,
+      "step": 1617
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 0.00011376734258271077,
+      "loss": 0.8673,
+      "step": 1618
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 0.00011371398078975455,
+      "loss": 0.8553,
+      "step": 1619
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 0.0001136606189967983,
+      "loss": 0.8876,
+      "step": 1620
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 0.00011360725720384204,
+      "loss": 0.9605,
+      "step": 1621
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 0.00011355389541088582,
+      "loss": 0.4542,
+      "step": 1622
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 0.00011350053361792955,
+      "loss": 0.7955,
+      "step": 1623
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 0.00011344717182497333,
+      "loss": 0.9756,
+      "step": 1624
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 0.00011339381003201708,
+      "loss": 0.9457,
+      "step": 1625
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 0.00011334044823906085,
+      "loss": 1.033,
+      "step": 1626
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 0.00011328708644610459,
+      "loss": 0.8776,
+      "step": 1627
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 0.00011323372465314834,
+      "loss": 0.9193,
+      "step": 1628
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 0.00011318036286019211,
+      "loss": 1.0516,
+      "step": 1629
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 0.00011312700106723586,
+      "loss": 0.9505,
+      "step": 1630
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 0.00011307363927427962,
+      "loss": 0.4932,
+      "step": 1631
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 0.00011302027748132337,
+      "loss": 0.9201,
+      "step": 1632
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 0.00011296691568836712,
+      "loss": 0.9391,
+      "step": 1633
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 0.0001129135538954109,
+      "loss": 0.843,
+      "step": 1634
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 0.00011286019210245464,
+      "loss": 0.8774,
+      "step": 1635
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 0.0001128068303094984,
+      "loss": 0.5023,
+      "step": 1636
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 0.00011275346851654215,
+      "loss": 0.8064,
+      "step": 1637
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 0.0001127001067235859,
+      "loss": 0.5676,
+      "step": 1638
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 0.00011264674493062968,
+      "loss": 0.8896,
+      "step": 1639
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 0.00011259338313767343,
+      "loss": 0.486,
+      "step": 1640
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 0.00011254002134471719,
+      "loss": 0.7887,
+      "step": 1641
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 0.00011248665955176094,
+      "loss": 0.5275,
+      "step": 1642
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 0.00011243329775880471,
+      "loss": 0.487,
+      "step": 1643
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 0.00011237993596584846,
+      "loss": 0.9553,
+      "step": 1644
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 0.00011232657417289221,
+      "loss": 0.8834,
+      "step": 1645
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 0.00011227321237993597,
+      "loss": 0.3591,
+      "step": 1646
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 0.00011221985058697972,
+      "loss": 0.4045,
+      "step": 1647
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 0.0001121664887940235,
+      "loss": 0.7576,
+      "step": 1648
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 0.00011211312700106724,
+      "loss": 0.8562,
+      "step": 1649
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 0.00011205976520811099,
+      "loss": 0.8917,
+      "step": 1650
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 0.00011200640341515475,
+      "loss": 0.86,
+      "step": 1651
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 0.0001119530416221985,
+      "loss": 0.8254,
+      "step": 1652
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 0.00011189967982924228,
+      "loss": 0.7794,
+      "step": 1653
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 0.00011184631803628603,
+      "loss": 0.4457,
+      "step": 1654
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 0.00011179295624332978,
+      "loss": 0.9286,
+      "step": 1655
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 0.00011173959445037354,
+      "loss": 0.8559,
+      "step": 1656
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 0.00011168623265741729,
+      "loss": 0.8212,
+      "step": 1657
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 0.00011163287086446106,
+      "loss": 0.9286,
+      "step": 1658
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 0.00011157950907150481,
+      "loss": 0.8288,
+      "step": 1659
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 0.00011152614727854856,
+      "loss": 0.9041,
+      "step": 1660
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 0.00011147278548559232,
+      "loss": 0.9381,
+      "step": 1661
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 0.00011141942369263607,
+      "loss": 0.9555,
+      "step": 1662
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 0.00011136606189967984,
+      "loss": 0.8625,
+      "step": 1663
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 0.00011131270010672359,
+      "loss": 0.964,
+      "step": 1664
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 0.00011125933831376735,
+      "loss": 0.8381,
+      "step": 1665
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 0.0001112059765208111,
+      "loss": 0.4031,
+      "step": 1666
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 0.00011115261472785485,
+      "loss": 0.4396,
+      "step": 1667
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 0.00011109925293489863,
+      "loss": 0.4524,
+      "step": 1668
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 0.00011104589114194238,
+      "loss": 0.9978,
+      "step": 1669
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 0.00011099252934898614,
+      "loss": 0.9049,
+      "step": 1670
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 0.00011093916755602989,
+      "loss": 0.4195,
+      "step": 1671
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 0.00011088580576307363,
+      "loss": 1.0154,
+      "step": 1672
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 0.00011083244397011741,
+      "loss": 1.0343,
+      "step": 1673
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 0.00011077908217716116,
+      "loss": 0.7798,
+      "step": 1674
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 0.00011072572038420492,
+      "loss": 0.4205,
+      "step": 1675
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 0.00011067235859124867,
+      "loss": 0.4534,
+      "step": 1676
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 0.00011061899679829242,
+      "loss": 0.859,
+      "step": 1677
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 0.00011056563500533619,
+      "loss": 0.842,
+      "step": 1678
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 0.00011051227321237994,
+      "loss": 0.8335,
+      "step": 1679
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 0.0001104589114194237,
+      "loss": 0.9805,
+      "step": 1680
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 0.00011040554962646745,
+      "loss": 0.9027,
+      "step": 1681
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 0.0001103521878335112,
+      "loss": 1.0538,
+      "step": 1682
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 0.00011029882604055497,
+      "loss": 0.9097,
+      "step": 1683
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 0.00011024546424759872,
+      "loss": 1.0097,
+      "step": 1684
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 0.00011019210245464249,
+      "loss": 0.9058,
+      "step": 1685
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 0.00011013874066168623,
+      "loss": 0.4284,
+      "step": 1686
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 0.00011008537886873001,
+      "loss": 0.9716,
+      "step": 1687
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 0.00011003201707577376,
+      "loss": 0.7554,
+      "step": 1688
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 0.0001099786552828175,
+      "loss": 0.8629,
+      "step": 1689
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 0.00010992529348986127,
+      "loss": 0.7938,
+      "step": 1690
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 0.00010987193169690502,
+      "loss": 0.8187,
+      "step": 1691
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 0.00010981856990394879,
+      "loss": 0.8699,
+      "step": 1692
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 0.00010976520811099254,
+      "loss": 0.7733,
+      "step": 1693
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 0.00010971184631803629,
+      "loss": 1.0697,
+      "step": 1694
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 0.00010965848452508005,
+      "loss": 0.8532,
+      "step": 1695
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 0.0001096051227321238,
+      "loss": 0.8201,
+      "step": 1696
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 0.00010955176093916757,
+      "loss": 0.9239,
+      "step": 1697
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 0.00010949839914621132,
+      "loss": 0.4898,
+      "step": 1698
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 0.00010944503735325506,
+      "loss": 0.4312,
+      "step": 1699
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 0.00010939167556029883,
+      "loss": 0.4369,
+      "step": 1700
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 0.00010933831376734258,
+      "loss": 0.9137,
+      "step": 1701
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 0.00010928495197438636,
+      "loss": 0.8271,
+      "step": 1702
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 0.0001092315901814301,
+      "loss": 0.9434,
+      "step": 1703
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 0.00010917822838847387,
+      "loss": 0.8782,
+      "step": 1704
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 0.00010912486659551762,
+      "loss": 0.8865,
+      "step": 1705
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 0.00010907150480256136,
+      "loss": 0.8469,
+      "step": 1706
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 0.00010901814300960514,
+      "loss": 0.3995,
+      "step": 1707
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 0.00010896478121664887,
+      "loss": 0.8929,
+      "step": 1708
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 0.00010891141942369265,
+      "loss": 0.898,
+      "step": 1709
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 0.0001088580576307364,
+      "loss": 0.8161,
+      "step": 1710
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 0.00010880469583778015,
+      "loss": 0.9052,
+      "step": 1711
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 0.00010875133404482391,
+      "loss": 0.7814,
+      "step": 1712
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 0.00010869797225186766,
+      "loss": 0.0,
+      "step": 1713
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 0.00010864461045891143,
+      "loss": 0.9118,
+      "step": 1714
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 0.00010859124866595518,
+      "loss": 0.9156,
+      "step": 1715
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 0.00010853788687299893,
+      "loss": 0.8593,
+      "step": 1716
+    },
+    {
+      "epoch": 1.37,
+      "eval_loss": 0.9985359311103821,
+      "eval_runtime": 7469.135,
+      "eval_samples_per_second": 1.339,
+      "eval_steps_per_second": 0.084,
+      "step": 1716
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 0.00010848452508004269,
+      "loss": 0.3792,
+      "step": 1717
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 0.00010843116328708644,
+      "loss": 0.9552,
+      "step": 1718
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 0.00010837780149413022,
+      "loss": 0.9399,
+      "step": 1719
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 0.00010832443970117396,
+      "loss": 0.4532,
+      "step": 1720
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 0.00010827107790821771,
+      "loss": 0.7878,
+      "step": 1721
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 0.00010821771611526147,
+      "loss": 0.9158,
+      "step": 1722
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 0.00010816435432230522,
+      "loss": 0.9541,
+      "step": 1723
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 0.000108110992529349,
+      "loss": 0.9115,
+      "step": 1724
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 0.00010805763073639275,
+      "loss": 0.8496,
+      "step": 1725
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 0.00010800426894343651,
+      "loss": 0.8989,
+      "step": 1726
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 0.00010795090715048026,
+      "loss": 0.4742,
+      "step": 1727
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 0.000107897545357524,
+      "loss": 0.7981,
+      "step": 1728
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 0.00010784418356456778,
+      "loss": 0.8631,
+      "step": 1729
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 0.00010779082177161153,
+      "loss": 0.9764,
+      "step": 1730
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 0.00010773745997865529,
+      "loss": 0.8675,
+      "step": 1731
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 0.00010768409818569904,
+      "loss": 1.0222,
+      "step": 1732
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 0.00010763073639274279,
+      "loss": 0.9954,
+      "step": 1733
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 0.00010757737459978656,
+      "loss": 0.8184,
+      "step": 1734
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 0.00010752401280683031,
+      "loss": 0.8747,
+      "step": 1735
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 0.00010747065101387407,
+      "loss": 0.8522,
+      "step": 1736
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 0.00010741728922091782,
+      "loss": 0.9638,
+      "step": 1737
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 0.00010736392742796157,
+      "loss": 0.9802,
+      "step": 1738
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 0.00010731056563500535,
+      "loss": 1.011,
+      "step": 1739
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 0.0001072572038420491,
+      "loss": 0.6797,
+      "step": 1740
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 0.00010720384204909286,
+      "loss": 0.9086,
+      "step": 1741
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 0.0001071504802561366,
+      "loss": 0.9508,
+      "step": 1742
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 0.00010709711846318035,
+      "loss": 0.8772,
+      "step": 1743
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 0.00010704375667022413,
+      "loss": 0.7428,
+      "step": 1744
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 0.00010699039487726788,
+      "loss": 0.8686,
+      "step": 1745
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 0.00010693703308431164,
+      "loss": 0.8978,
+      "step": 1746
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 0.00010688367129135539,
+      "loss": 0.782,
+      "step": 1747
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 0.00010683030949839916,
+      "loss": 0.9041,
+      "step": 1748
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 0.00010677694770544291,
+      "loss": 0.9276,
+      "step": 1749
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 0.00010672358591248666,
+      "loss": 0.4647,
+      "step": 1750
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 0.00010667022411953042,
+      "loss": 0.8631,
+      "step": 1751
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 0.00010661686232657417,
+      "loss": 0.8799,
+      "step": 1752
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 0.00010656350053361795,
+      "loss": 0.4648,
+      "step": 1753
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 0.0001065101387406617,
+      "loss": 1.0756,
+      "step": 1754
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 0.00010645677694770544,
+      "loss": 1.1229,
+      "step": 1755
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 0.0001064034151547492,
+      "loss": 0.7679,
+      "step": 1756
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 0.00010635005336179295,
+      "loss": 1.0028,
+      "step": 1757
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 0.00010629669156883673,
+      "loss": 0.9621,
+      "step": 1758
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 0.00010624332977588048,
+      "loss": 0.6736,
+      "step": 1759
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 0.00010618996798292423,
+      "loss": 0.7874,
+      "step": 1760
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 0.00010613660618996799,
+      "loss": 0.7373,
+      "step": 1761
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 0.00010608324439701174,
+      "loss": 0.9847,
+      "step": 1762
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 0.00010602988260405551,
+      "loss": 0.9269,
+      "step": 1763
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 0.00010597652081109926,
+      "loss": 1.0382,
+      "step": 1764
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 0.00010592315901814302,
+      "loss": 0.7673,
+      "step": 1765
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 0.00010586979722518677,
+      "loss": 0.8027,
+      "step": 1766
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 0.00010581643543223052,
+      "loss": 0.9417,
+      "step": 1767
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 0.0001057630736392743,
+      "loss": 1.0463,
+      "step": 1768
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 0.00010570971184631804,
+      "loss": 0.6785,
+      "step": 1769
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 0.0001056563500533618,
+      "loss": 0.784,
+      "step": 1770
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 0.00010560298826040555,
+      "loss": 0.7935,
+      "step": 1771
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 0.0001055496264674493,
+      "loss": 1.0542,
+      "step": 1772
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 0.00010549626467449308,
+      "loss": 0.9118,
+      "step": 1773
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 0.00010544290288153683,
+      "loss": 0.8866,
+      "step": 1774
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 0.00010538954108858059,
+      "loss": 0.5435,
+      "step": 1775
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 0.00010533617929562434,
+      "loss": 0.9209,
+      "step": 1776
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 0.00010528281750266808,
+      "loss": 0.9333,
+      "step": 1777
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 0.00010522945570971186,
+      "loss": 0.9183,
+      "step": 1778
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 0.00010517609391675561,
+      "loss": 0.9315,
+      "step": 1779
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 0.00010512273212379937,
+      "loss": 0.5751,
+      "step": 1780
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 0.00010506937033084312,
+      "loss": 0.6947,
+      "step": 1781
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 0.00010501600853788687,
+      "loss": 0.8276,
+      "step": 1782
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 0.00010496264674493064,
+      "loss": 0.9229,
+      "step": 1783
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 0.00010490928495197438,
+      "loss": 0.8911,
+      "step": 1784
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 0.00010485592315901815,
+      "loss": 1.046,
+      "step": 1785
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 0.0001048025613660619,
+      "loss": 1.04,
+      "step": 1786
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 0.00010474919957310568,
+      "loss": 0.4773,
+      "step": 1787
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 0.00010469583778014943,
+      "loss": 0.7709,
+      "step": 1788
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 0.00010464247598719316,
+      "loss": 0.8272,
+      "step": 1789
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 0.00010458911419423694,
+      "loss": 0.9234,
+      "step": 1790
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 0.00010453575240128068,
+      "loss": 0.9089,
+      "step": 1791
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 0.00010448239060832446,
+      "loss": 0.8311,
+      "step": 1792
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 0.0001044290288153682,
+      "loss": 0.9424,
+      "step": 1793
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 0.00010437566702241194,
+      "loss": 0.9383,
+      "step": 1794
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 0.00010432230522945572,
+      "loss": 0.9969,
+      "step": 1795
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 0.00010426894343649947,
+      "loss": 0.8116,
+      "step": 1796
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 0.00010421558164354324,
+      "loss": 0.8201,
+      "step": 1797
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 0.00010416221985058698,
+      "loss": 0.8734,
+      "step": 1798
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 0.00010410885805763073,
+      "loss": 0.8989,
+      "step": 1799
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 0.0001040554962646745,
+      "loss": 0.9238,
+      "step": 1800
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 0.00010400213447171825,
+      "loss": 0.4867,
+      "step": 1801
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 0.00010394877267876201,
+      "loss": 0.8783,
+      "step": 1802
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 0.00010389541088580576,
+      "loss": 0.8879,
+      "step": 1803
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 0.00010384204909284951,
+      "loss": 0.8567,
+      "step": 1804
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 0.00010378868729989328,
+      "loss": 0.8253,
+      "step": 1805
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 0.00010373532550693703,
+      "loss": 0.9467,
+      "step": 1806
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 0.0001036819637139808,
+      "loss": 0.4182,
+      "step": 1807
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 0.00010362860192102454,
+      "loss": 0.7999,
+      "step": 1808
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 0.00010357524012806832,
+      "loss": 0.8709,
+      "step": 1809
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 0.00010352187833511207,
+      "loss": 0.9995,
+      "step": 1810
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 0.00010346851654215582,
+      "loss": 0.9845,
+      "step": 1811
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 0.00010341515474919958,
+      "loss": 0.8148,
+      "step": 1812
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 0.00010336179295624333,
+      "loss": 0.7489,
+      "step": 1813
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 0.0001033084311632871,
+      "loss": 0.9689,
+      "step": 1814
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 0.00010325506937033085,
+      "loss": 0.8776,
+      "step": 1815
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 0.0001032017075773746,
+      "loss": 0.9055,
+      "step": 1816
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 0.00010314834578441836,
+      "loss": 1.0643,
+      "step": 1817
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 0.00010309498399146211,
+      "loss": 0.9014,
+      "step": 1818
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 0.00010304162219850588,
+      "loss": 0.7694,
+      "step": 1819
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 0.00010298826040554963,
+      "loss": 0.8488,
+      "step": 1820
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 0.00010293489861259338,
+      "loss": 0.7804,
+      "step": 1821
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 0.00010288153681963714,
+      "loss": 0.7808,
+      "step": 1822
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 0.00010282817502668089,
+      "loss": 0.8632,
+      "step": 1823
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 0.00010277481323372467,
+      "loss": 0.7149,
+      "step": 1824
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 0.00010272145144076842,
+      "loss": 0.796,
+      "step": 1825
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 0.00010266808964781218,
+      "loss": 0.4705,
+      "step": 1826
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 0.00010261472785485593,
+      "loss": 0.8115,
+      "step": 1827
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 0.00010256136606189967,
+      "loss": 0.8614,
+      "step": 1828
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 0.00010250800426894345,
+      "loss": 0.4436,
+      "step": 1829
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 0.0001024546424759872,
+      "loss": 0.8367,
+      "step": 1830
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 0.00010240128068303096,
+      "loss": 0.4671,
+      "step": 1831
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 0.00010234791889007471,
+      "loss": 0.8819,
+      "step": 1832
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 0.00010229455709711846,
+      "loss": 0.8543,
+      "step": 1833
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 0.00010224119530416223,
+      "loss": 0.8971,
+      "step": 1834
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 0.00010218783351120598,
+      "loss": 1.0111,
+      "step": 1835
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 0.00010213447171824974,
+      "loss": 1.0373,
+      "step": 1836
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 0.00010208110992529349,
+      "loss": 1.0013,
+      "step": 1837
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 0.00010202774813233724,
+      "loss": 0.7675,
+      "step": 1838
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 0.00010197438633938102,
+      "loss": 0.9012,
+      "step": 1839
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 0.00010192102454642476,
+      "loss": 0.7214,
+      "step": 1840
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 0.00010186766275346853,
+      "loss": 0.9429,
+      "step": 1841
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 0.00010181430096051227,
+      "loss": 0.8371,
+      "step": 1842
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 0.00010176093916755602,
+      "loss": 0.8738,
+      "step": 1843
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 0.0001017075773745998,
+      "loss": 0.4561,
+      "step": 1844
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 0.00010165421558164355,
+      "loss": 0.811,
+      "step": 1845
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 0.00010160085378868731,
+      "loss": 0.9679,
+      "step": 1846
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 0.00010154749199573106,
+      "loss": 0.9866,
+      "step": 1847
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 0.00010149413020277483,
+      "loss": 0.8314,
+      "step": 1848
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 0.00010144076840981858,
+      "loss": 0.5767,
+      "step": 1849
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 0.00010138740661686233,
+      "loss": 0.8133,
+      "step": 1850
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 0.00010133404482390609,
+      "loss": 0.5525,
+      "step": 1851
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 0.00010128068303094984,
+      "loss": 0.8997,
+      "step": 1852
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 0.00010122732123799362,
+      "loss": 0.9946,
+      "step": 1853
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 0.00010117395944503736,
+      "loss": 0.3709,
+      "step": 1854
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 0.00010112059765208111,
+      "loss": 0.9388,
+      "step": 1855
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 0.00010106723585912487,
+      "loss": 0.8165,
+      "step": 1856
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 0.00010101387406616862,
+      "loss": 0.896,
+      "step": 1857
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 0.0001009605122732124,
+      "loss": 0.9617,
+      "step": 1858
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 0.00010090715048025615,
+      "loss": 0.9868,
+      "step": 1859
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 0.0001008537886872999,
+      "loss": 0.8389,
+      "step": 1860
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 0.00010080042689434366,
+      "loss": 0.9469,
+      "step": 1861
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 0.0001007470651013874,
+      "loss": 0.8176,
+      "step": 1862
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 0.00010069370330843118,
+      "loss": 0.469,
+      "step": 1863
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 0.00010064034151547493,
+      "loss": 0.8996,
+      "step": 1864
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 0.00010058697972251866,
+      "loss": 0.9687,
+      "step": 1865
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 0.00010053361792956244,
+      "loss": 0.3448,
+      "step": 1866
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 0.00010048025613660619,
+      "loss": 0.9627,
+      "step": 1867
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 0.00010042689434364996,
+      "loss": 0.9318,
+      "step": 1868
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 0.00010037353255069371,
+      "loss": 0.7098,
+      "step": 1869
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 0.00010032017075773747,
+      "loss": 0.8405,
+      "step": 1870
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 0.00010026680896478122,
+      "loss": 0.8631,
+      "step": 1871
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 0.00010021344717182497,
+      "loss": 0.9688,
+      "step": 1872
+    },
+    {
+      "epoch": 1.5,
+      "eval_loss": 1.0002734661102295,
+      "eval_runtime": 7631.7851,
+      "eval_samples_per_second": 1.31,
+      "eval_steps_per_second": 0.082,
+      "step": 1872
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 0.00010016008537886875,
+      "loss": 0.7903,
+      "step": 1873
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 0.00010010672358591248,
+      "loss": 1.0131,
+      "step": 1874
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 0.00010005336179295626,
+      "loss": 0.9893,
+      "step": 1875
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 0.0001,
+      "loss": 0.9663,
+      "step": 1876
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 9.994663820704377e-05,
+      "loss": 0.456,
+      "step": 1877
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 9.989327641408753e-05,
+      "loss": 0.7286,
+      "step": 1878
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 9.983991462113126e-05,
+      "loss": 0.9732,
+      "step": 1879
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 9.978655282817503e-05,
+      "loss": 0.854,
+      "step": 1880
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 9.973319103521879e-05,
+      "loss": 0.9321,
+      "step": 1881
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 9.967982924226255e-05,
+      "loss": 0.9728,
+      "step": 1882
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 9.96264674493063e-05,
+      "loss": 0.8229,
+      "step": 1883
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 9.957310565635006e-05,
+      "loss": 0.9016,
+      "step": 1884
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 9.951974386339381e-05,
+      "loss": 0.8312,
+      "step": 1885
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 9.946638207043757e-05,
+      "loss": 0.7903,
+      "step": 1886
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 9.941302027748133e-05,
+      "loss": 0.4309,
+      "step": 1887
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 9.935965848452508e-05,
+      "loss": 0.7926,
+      "step": 1888
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 9.930629669156884e-05,
+      "loss": 0.3717,
+      "step": 1889
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 9.925293489861259e-05,
+      "loss": 0.7696,
+      "step": 1890
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 9.919957310565635e-05,
+      "loss": 0.8862,
+      "step": 1891
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 9.914621131270011e-05,
+      "loss": 1.0347,
+      "step": 1892
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 9.909284951974386e-05,
+      "loss": 0.9563,
+      "step": 1893
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 9.903948772678763e-05,
+      "loss": 0.8225,
+      "step": 1894
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 9.898612593383139e-05,
+      "loss": 0.8492,
+      "step": 1895
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 9.893276414087514e-05,
+      "loss": 0.9232,
+      "step": 1896
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 9.88794023479189e-05,
+      "loss": 0.9468,
+      "step": 1897
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 9.882604055496265e-05,
+      "loss": 0.8264,
+      "step": 1898
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 9.877267876200641e-05,
+      "loss": 0.8868,
+      "step": 1899
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 9.871931696905017e-05,
+      "loss": 0.389,
+      "step": 1900
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 9.866595517609392e-05,
+      "loss": 0.9643,
+      "step": 1901
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 9.861259338313768e-05,
+      "loss": 0.9184,
+      "step": 1902
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 9.855923159018143e-05,
+      "loss": 0.8675,
+      "step": 1903
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 9.850586979722519e-05,
+      "loss": 0.4143,
+      "step": 1904
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 9.845250800426895e-05,
+      "loss": 0.8021,
+      "step": 1905
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 9.839914621131271e-05,
+      "loss": 0.3934,
+      "step": 1906
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 9.834578441835646e-05,
+      "loss": 0.8215,
+      "step": 1907
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 9.829242262540021e-05,
+      "loss": 0.7566,
+      "step": 1908
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 9.823906083244397e-05,
+      "loss": 0.9592,
+      "step": 1909
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 9.818569903948774e-05,
+      "loss": 0.9011,
+      "step": 1910
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 9.81323372465315e-05,
+      "loss": 0.9497,
+      "step": 1911
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 9.807897545357525e-05,
+      "loss": 1.0322,
+      "step": 1912
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 9.8025613660619e-05,
+      "loss": 0.8216,
+      "step": 1913
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 9.797225186766276e-05,
+      "loss": 0.8817,
+      "step": 1914
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 9.791889007470652e-05,
+      "loss": 0.859,
+      "step": 1915
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 9.786552828175028e-05,
+      "loss": 0.8018,
+      "step": 1916
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 9.781216648879403e-05,
+      "loss": 0.9676,
+      "step": 1917
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 9.775880469583778e-05,
+      "loss": 0.7943,
+      "step": 1918
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 9.770544290288154e-05,
+      "loss": 0.7896,
+      "step": 1919
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 9.76520811099253e-05,
+      "loss": 0.9429,
+      "step": 1920
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 9.759871931696905e-05,
+      "loss": 0.737,
+      "step": 1921
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 9.754535752401281e-05,
+      "loss": 0.3472,
+      "step": 1922
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 9.749199573105657e-05,
+      "loss": 0.9751,
+      "step": 1923
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 9.743863393810032e-05,
+      "loss": 0.7492,
+      "step": 1924
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 9.738527214514408e-05,
+      "loss": 0.9019,
+      "step": 1925
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 9.733191035218783e-05,
+      "loss": 0.7601,
+      "step": 1926
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 9.72785485592316e-05,
+      "loss": 0.911,
+      "step": 1927
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 9.722518676627536e-05,
+      "loss": 0.9312,
+      "step": 1928
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 9.71718249733191e-05,
+      "loss": 1.0699,
+      "step": 1929
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 9.711846318036287e-05,
+      "loss": 0.8478,
+      "step": 1930
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 9.706510138740661e-05,
+      "loss": 0.9209,
+      "step": 1931
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 9.701173959445038e-05,
+      "loss": 0.8971,
+      "step": 1932
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 9.695837780149414e-05,
+      "loss": 0.9643,
+      "step": 1933
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 9.69050160085379e-05,
+      "loss": 0.9387,
+      "step": 1934
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 9.685165421558165e-05,
+      "loss": 0.9916,
+      "step": 1935
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 9.67982924226254e-05,
+      "loss": 0.4154,
+      "step": 1936
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 9.674493062966916e-05,
+      "loss": 0.3716,
+      "step": 1937
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 9.669156883671292e-05,
+      "loss": 0.7018,
+      "step": 1938
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 9.663820704375668e-05,
+      "loss": 0.931,
+      "step": 1939
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 9.658484525080043e-05,
+      "loss": 0.8068,
+      "step": 1940
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 9.653148345784418e-05,
+      "loss": 0.9394,
+      "step": 1941
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 9.647812166488794e-05,
+      "loss": 0.7985,
+      "step": 1942
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 9.64247598719317e-05,
+      "loss": 0.4686,
+      "step": 1943
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 9.637139807897547e-05,
+      "loss": 0.8829,
+      "step": 1944
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 9.631803628601921e-05,
+      "loss": 0.874,
+      "step": 1945
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 9.626467449306296e-05,
+      "loss": 0.8847,
+      "step": 1946
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 9.621131270010673e-05,
+      "loss": 0.8267,
+      "step": 1947
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 9.615795090715049e-05,
+      "loss": 0.8218,
+      "step": 1948
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 9.610458911419425e-05,
+      "loss": 0.3615,
+      "step": 1949
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 9.6051227321238e-05,
+      "loss": 0.4364,
+      "step": 1950
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 9.599786552828175e-05,
+      "loss": 0.9089,
+      "step": 1951
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 9.594450373532551e-05,
+      "loss": 0.7676,
+      "step": 1952
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 9.589114194236927e-05,
+      "loss": 0.8273,
+      "step": 1953
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 9.583778014941303e-05,
+      "loss": 0.8046,
+      "step": 1954
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 9.578441835645678e-05,
+      "loss": 0.4514,
+      "step": 1955
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 9.573105656350054e-05,
+      "loss": 0.9613,
+      "step": 1956
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 9.567769477054429e-05,
+      "loss": 0.5192,
+      "step": 1957
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 9.562433297758805e-05,
+      "loss": 0.9744,
+      "step": 1958
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 9.557097118463181e-05,
+      "loss": 0.8119,
+      "step": 1959
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 9.551760939167556e-05,
+      "loss": 0.897,
+      "step": 1960
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 9.546424759871932e-05,
+      "loss": 0.8033,
+      "step": 1961
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 9.541088580576307e-05,
+      "loss": 0.8948,
+      "step": 1962
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 9.535752401280684e-05,
+      "loss": 0.8916,
+      "step": 1963
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 9.530416221985058e-05,
+      "loss": 0.4438,
+      "step": 1964
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 9.525080042689435e-05,
+      "loss": 0.8337,
+      "step": 1965
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 9.519743863393811e-05,
+      "loss": 0.8546,
+      "step": 1966
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 9.514407684098187e-05,
+      "loss": 0.8343,
+      "step": 1967
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 9.509071504802562e-05,
+      "loss": 0.7763,
+      "step": 1968
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 9.503735325506937e-05,
+      "loss": 0.8583,
+      "step": 1969
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 9.498399146211313e-05,
+      "loss": 1.1261,
+      "step": 1970
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 9.493062966915689e-05,
+      "loss": 1.0732,
+      "step": 1971
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 9.487726787620065e-05,
+      "loss": 0.9576,
+      "step": 1972
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 9.48239060832444e-05,
+      "loss": 0.8152,
+      "step": 1973
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 9.477054429028815e-05,
+      "loss": 0.8144,
+      "step": 1974
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 9.471718249733191e-05,
+      "loss": 0.8321,
+      "step": 1975
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 9.466382070437567e-05,
+      "loss": 1.1534,
+      "step": 1976
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 9.461045891141944e-05,
+      "loss": 1.1164,
+      "step": 1977
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 9.455709711846318e-05,
+      "loss": 0.8367,
+      "step": 1978
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 9.450373532550693e-05,
+      "loss": 0.7238,
+      "step": 1979
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 9.44503735325507e-05,
+      "loss": 0.9117,
+      "step": 1980
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 9.439701173959446e-05,
+      "loss": 0.7405,
+      "step": 1981
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 9.434364994663822e-05,
+      "loss": 0.9113,
+      "step": 1982
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 9.429028815368197e-05,
+      "loss": 1.0566,
+      "step": 1983
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 9.423692636072573e-05,
+      "loss": 0.8302,
+      "step": 1984
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 9.418356456776948e-05,
+      "loss": 0.8672,
+      "step": 1985
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 9.413020277481324e-05,
+      "loss": 0.9032,
+      "step": 1986
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 9.4076840981857e-05,
+      "loss": 0.9857,
+      "step": 1987
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 9.402347918890075e-05,
+      "loss": 0.9556,
+      "step": 1988
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 9.397011739594451e-05,
+      "loss": 0.8109,
+      "step": 1989
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 9.391675560298826e-05,
+      "loss": 0.7776,
+      "step": 1990
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 9.386339381003202e-05,
+      "loss": 1.0226,
+      "step": 1991
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 9.381003201707578e-05,
+      "loss": 0.8941,
+      "step": 1992
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 9.375667022411953e-05,
+      "loss": 0.3553,
+      "step": 1993
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 9.37033084311633e-05,
+      "loss": 0.7412,
+      "step": 1994
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 9.364994663820706e-05,
+      "loss": 0.4657,
+      "step": 1995
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 9.35965848452508e-05,
+      "loss": 0.5139,
+      "step": 1996
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 9.354322305229457e-05,
+      "loss": 0.8632,
+      "step": 1997
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 9.348986125933831e-05,
+      "loss": 0.8369,
+      "step": 1998
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 9.343649946638208e-05,
+      "loss": 0.6952,
+      "step": 1999
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 9.338313767342584e-05,
+      "loss": 0.8944,
+      "step": 2000
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 9.332977588046959e-05,
+      "loss": 0.8329,
+      "step": 2001
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 9.327641408751334e-05,
+      "loss": 0.7575,
+      "step": 2002
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 9.32230522945571e-05,
+      "loss": 0.8136,
+      "step": 2003
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 9.316969050160086e-05,
+      "loss": 0.8281,
+      "step": 2004
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 9.311632870864462e-05,
+      "loss": 0.7539,
+      "step": 2005
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 9.306296691568838e-05,
+      "loss": 0.8367,
+      "step": 2006
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 9.300960512273212e-05,
+      "loss": 0.7879,
+      "step": 2007
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 9.295624332977588e-05,
+      "loss": 0.0,
+      "step": 2008
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 9.290288153681964e-05,
+      "loss": 0.9878,
+      "step": 2009
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 9.28495197438634e-05,
+      "loss": 0.4569,
+      "step": 2010
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 9.279615795090715e-05,
+      "loss": 0.9631,
+      "step": 2011
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 9.27427961579509e-05,
+      "loss": 0.8711,
+      "step": 2012
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 9.268943436499466e-05,
+      "loss": 0.8437,
+      "step": 2013
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 9.263607257203842e-05,
+      "loss": 0.9054,
+      "step": 2014
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 9.258271077908219e-05,
+      "loss": 0.9486,
+      "step": 2015
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 9.252934898612594e-05,
+      "loss": 0.8262,
+      "step": 2016
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 9.24759871931697e-05,
+      "loss": 0.9252,
+      "step": 2017
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 9.242262540021345e-05,
+      "loss": 0.8257,
+      "step": 2018
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 9.236926360725721e-05,
+      "loss": 0.7877,
+      "step": 2019
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 9.231590181430097e-05,
+      "loss": 0.902,
+      "step": 2020
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 9.226254002134472e-05,
+      "loss": 0.9815,
+      "step": 2021
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 9.220917822838848e-05,
+      "loss": 0.9223,
+      "step": 2022
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 9.215581643543223e-05,
+      "loss": 0.8726,
+      "step": 2023
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 9.210245464247599e-05,
+      "loss": 0.9455,
+      "step": 2024
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 9.204909284951975e-05,
+      "loss": 0.4656,
+      "step": 2025
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 9.19957310565635e-05,
+      "loss": 0.8762,
+      "step": 2026
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 9.194236926360726e-05,
+      "loss": 0.8315,
+      "step": 2027
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 9.188900747065102e-05,
+      "loss": 0.8505,
+      "step": 2028
+    },
+    {
+      "epoch": 1.62,
+      "eval_loss": 0.9914842844009399,
+      "eval_runtime": 7418.4995,
+      "eval_samples_per_second": 1.348,
+      "eval_steps_per_second": 0.084,
+      "step": 2028
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 9.183564567769477e-05,
+      "loss": 0.8211,
+      "step": 2029
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 9.178228388473853e-05,
+      "loss": 0.815,
+      "step": 2030
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 9.172892209178228e-05,
+      "loss": 0.8878,
+      "step": 2031
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 9.167556029882605e-05,
+      "loss": 0.8859,
+      "step": 2032
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 9.162219850586981e-05,
+      "loss": 0.895,
+      "step": 2033
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 9.156883671291357e-05,
+      "loss": 0.8897,
+      "step": 2034
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 9.151547491995732e-05,
+      "loss": 0.8406,
+      "step": 2035
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 9.146211312700107e-05,
+      "loss": 0.8957,
+      "step": 2036
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 9.140875133404483e-05,
+      "loss": 0.4778,
+      "step": 2037
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 9.135538954108859e-05,
+      "loss": 0.4697,
+      "step": 2038
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 9.130202774813235e-05,
+      "loss": 0.8975,
+      "step": 2039
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 9.12486659551761e-05,
+      "loss": 0.5611,
+      "step": 2040
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 9.119530416221985e-05,
+      "loss": 0.7951,
+      "step": 2041
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 9.114194236926361e-05,
+      "loss": 0.5006,
+      "step": 2042
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 9.108858057630737e-05,
+      "loss": 0.7717,
+      "step": 2043
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 9.103521878335113e-05,
+      "loss": 0.4862,
+      "step": 2044
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 9.098185699039488e-05,
+      "loss": 0.4301,
+      "step": 2045
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 9.092849519743863e-05,
+      "loss": 0.9462,
+      "step": 2046
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 9.08751334044824e-05,
+      "loss": 0.9189,
+      "step": 2047
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 9.082177161152616e-05,
+      "loss": 0.4859,
+      "step": 2048
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 9.07684098185699e-05,
+      "loss": 1.0724,
+      "step": 2049
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 9.071504802561367e-05,
+      "loss": 0.9027,
+      "step": 2050
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 9.066168623265741e-05,
+      "loss": 0.4385,
+      "step": 2051
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 9.060832443970118e-05,
+      "loss": 0.7411,
+      "step": 2052
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 9.055496264674494e-05,
+      "loss": 0.8346,
+      "step": 2053
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 9.050160085378869e-05,
+      "loss": 0.9097,
+      "step": 2054
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 9.044823906083245e-05,
+      "loss": 0.8561,
+      "step": 2055
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 9.039487726787621e-05,
+      "loss": 0.8306,
+      "step": 2056
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 9.034151547491996e-05,
+      "loss": 0.7182,
+      "step": 2057
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 9.028815368196372e-05,
+      "loss": 0.4933,
+      "step": 2058
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 9.023479188900747e-05,
+      "loss": 0.7787,
+      "step": 2059
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 9.018143009605123e-05,
+      "loss": 0.8648,
+      "step": 2060
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 9.012806830309499e-05,
+      "loss": 0.9109,
+      "step": 2061
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 9.007470651013874e-05,
+      "loss": 0.9827,
+      "step": 2062
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 9.00213447171825e-05,
+      "loss": 0.9503,
+      "step": 2063
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 8.996798292422625e-05,
+      "loss": 0.9847,
+      "step": 2064
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 8.991462113127001e-05,
+      "loss": 0.9742,
+      "step": 2065
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 8.986125933831378e-05,
+      "loss": 0.459,
+      "step": 2066
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 8.980789754535754e-05,
+      "loss": 0.7713,
+      "step": 2067
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 8.975453575240129e-05,
+      "loss": 1.0052,
+      "step": 2068
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 8.970117395944503e-05,
+      "loss": 0.3872,
+      "step": 2069
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 8.96478121664888e-05,
+      "loss": 0.4987,
+      "step": 2070
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 8.959445037353256e-05,
+      "loss": 0.7851,
+      "step": 2071
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 8.954108858057632e-05,
+      "loss": 0.8191,
+      "step": 2072
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 8.948772678762007e-05,
+      "loss": 0.9953,
+      "step": 2073
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 8.943436499466382e-05,
+      "loss": 1.1162,
+      "step": 2074
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 8.938100320170758e-05,
+      "loss": 0.8391,
+      "step": 2075
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 8.932764140875134e-05,
+      "loss": 0.8809,
+      "step": 2076
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 8.92742796157951e-05,
+      "loss": 0.4724,
+      "step": 2077
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 8.922091782283885e-05,
+      "loss": 1.0439,
+      "step": 2078
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 8.91675560298826e-05,
+      "loss": 0.5988,
+      "step": 2079
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 8.911419423692636e-05,
+      "loss": 0.9659,
+      "step": 2080
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 8.906083244397012e-05,
+      "loss": 0.9394,
+      "step": 2081
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 8.900747065101389e-05,
+      "loss": 1.0201,
+      "step": 2082
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 8.895410885805763e-05,
+      "loss": 0.8498,
+      "step": 2083
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 8.890074706510138e-05,
+      "loss": 0.9357,
+      "step": 2084
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 8.884738527214514e-05,
+      "loss": 0.4722,
+      "step": 2085
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 8.879402347918891e-05,
+      "loss": 0.7826,
+      "step": 2086
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 8.874066168623267e-05,
+      "loss": 0.9063,
+      "step": 2087
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 8.868729989327642e-05,
+      "loss": 0.5063,
+      "step": 2088
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 8.863393810032018e-05,
+      "loss": 0.8806,
+      "step": 2089
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 8.858057630736393e-05,
+      "loss": 0.7826,
+      "step": 2090
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 8.852721451440769e-05,
+      "loss": 0.844,
+      "step": 2091
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 8.847385272145144e-05,
+      "loss": 1.0269,
+      "step": 2092
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 8.84204909284952e-05,
+      "loss": 0.7621,
+      "step": 2093
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 8.836712913553896e-05,
+      "loss": 0.8076,
+      "step": 2094
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 8.831376734258272e-05,
+      "loss": 0.4652,
+      "step": 2095
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 8.826040554962647e-05,
+      "loss": 0.7672,
+      "step": 2096
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 8.820704375667022e-05,
+      "loss": 0.8658,
+      "step": 2097
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 8.815368196371398e-05,
+      "loss": 0.4702,
+      "step": 2098
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 8.810032017075774e-05,
+      "loss": 0.8075,
+      "step": 2099
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 8.80469583778015e-05,
+      "loss": 0.7958,
+      "step": 2100
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 8.799359658484526e-05,
+      "loss": 0.4253,
+      "step": 2101
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 8.7940234791889e-05,
+      "loss": 0.7844,
+      "step": 2102
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 8.788687299893277e-05,
+      "loss": 0.8714,
+      "step": 2103
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 8.783351120597653e-05,
+      "loss": 0.9125,
+      "step": 2104
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 8.778014941302029e-05,
+      "loss": 0.9085,
+      "step": 2105
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 8.772678762006404e-05,
+      "loss": 0.4605,
+      "step": 2106
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 8.767342582710779e-05,
+      "loss": 0.4791,
+      "step": 2107
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 8.762006403415155e-05,
+      "loss": 0.4594,
+      "step": 2108
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 8.756670224119531e-05,
+      "loss": 0.8305,
+      "step": 2109
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 8.751334044823907e-05,
+      "loss": 0.7878,
+      "step": 2110
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 8.745997865528282e-05,
+      "loss": 0.4442,
+      "step": 2111
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 8.740661686232657e-05,
+      "loss": 0.9078,
+      "step": 2112
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 8.735325506937033e-05,
+      "loss": 0.9267,
+      "step": 2113
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 8.729989327641409e-05,
+      "loss": 0.732,
+      "step": 2114
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 8.724653148345785e-05,
+      "loss": 0.8866,
+      "step": 2115
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 8.71931696905016e-05,
+      "loss": 1.0318,
+      "step": 2116
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 8.713980789754537e-05,
+      "loss": 1.0498,
+      "step": 2117
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 8.708644610458911e-05,
+      "loss": 0.466,
+      "step": 2118
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 8.703308431163288e-05,
+      "loss": 0.8791,
+      "step": 2119
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 8.697972251867664e-05,
+      "loss": 0.7693,
+      "step": 2120
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 8.692636072572039e-05,
+      "loss": 0.9791,
+      "step": 2121
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 8.687299893276415e-05,
+      "loss": 0.7924,
+      "step": 2122
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 8.68196371398079e-05,
+      "loss": 0.9065,
+      "step": 2123
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 8.676627534685166e-05,
+      "loss": 0.9161,
+      "step": 2124
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 8.671291355389542e-05,
+      "loss": 0.3404,
+      "step": 2125
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 8.665955176093917e-05,
+      "loss": 1.0227,
+      "step": 2126
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 8.660618996798293e-05,
+      "loss": 0.9122,
+      "step": 2127
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 8.655282817502669e-05,
+      "loss": 0.7313,
+      "step": 2128
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 8.649946638207044e-05,
+      "loss": 0.9096,
+      "step": 2129
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 8.644610458911419e-05,
+      "loss": 0.6846,
+      "step": 2130
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 8.639274279615795e-05,
+      "loss": 0.76,
+      "step": 2131
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 8.633938100320171e-05,
+      "loss": 0.7201,
+      "step": 2132
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 8.628601921024548e-05,
+      "loss": 0.8132,
+      "step": 2133
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 8.623265741728922e-05,
+      "loss": 0.856,
+      "step": 2134
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 8.617929562433297e-05,
+      "loss": 0.7248,
+      "step": 2135
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 8.612593383137673e-05,
+      "loss": 0.8232,
+      "step": 2136
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 8.60725720384205e-05,
+      "loss": 0.9782,
+      "step": 2137
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 8.601921024546426e-05,
+      "loss": 0.378,
+      "step": 2138
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 8.5965848452508e-05,
+      "loss": 0.4053,
+      "step": 2139
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 8.591248665955176e-05,
+      "loss": 0.9102,
+      "step": 2140
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 8.585912486659552e-05,
+      "loss": 0.8575,
+      "step": 2141
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 8.580576307363928e-05,
+      "loss": 0.0,
+      "step": 2142
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 8.575240128068304e-05,
+      "loss": 0.99,
+      "step": 2143
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 8.569903948772679e-05,
+      "loss": 1.011,
+      "step": 2144
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 8.564567769477054e-05,
+      "loss": 0.7783,
+      "step": 2145
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 8.55923159018143e-05,
+      "loss": 0.9033,
+      "step": 2146
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 8.553895410885806e-05,
+      "loss": 0.9682,
+      "step": 2147
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 8.548559231590182e-05,
+      "loss": 0.5177,
+      "step": 2148
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 8.543223052294557e-05,
+      "loss": 0.4291,
+      "step": 2149
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 8.537886872998933e-05,
+      "loss": 0.7913,
+      "step": 2150
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 8.532550693703308e-05,
+      "loss": 0.9911,
+      "step": 2151
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 8.527214514407684e-05,
+      "loss": 0.8606,
+      "step": 2152
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 8.52187833511206e-05,
+      "loss": 0.3508,
+      "step": 2153
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 8.516542155816435e-05,
+      "loss": 0.8515,
+      "step": 2154
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 8.511205976520812e-05,
+      "loss": 1.041,
+      "step": 2155
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 8.505869797225187e-05,
+      "loss": 0.9513,
+      "step": 2156
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 8.500533617929563e-05,
+      "loss": 0.9887,
+      "step": 2157
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 8.495197438633939e-05,
+      "loss": 0.808,
+      "step": 2158
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 8.489861259338314e-05,
+      "loss": 0.8529,
+      "step": 2159
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 8.48452508004269e-05,
+      "loss": 0.4566,
+      "step": 2160
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 8.479188900747066e-05,
+      "loss": 0.9727,
+      "step": 2161
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 8.473852721451441e-05,
+      "loss": 0.9473,
+      "step": 2162
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 8.468516542155817e-05,
+      "loss": 0.9478,
+      "step": 2163
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 8.463180362860192e-05,
+      "loss": 0.795,
+      "step": 2164
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 8.457844183564568e-05,
+      "loss": 0.9441,
+      "step": 2165
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 8.452508004268944e-05,
+      "loss": 1.0208,
+      "step": 2166
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 8.44717182497332e-05,
+      "loss": 0.8765,
+      "step": 2167
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 8.441835645677695e-05,
+      "loss": 0.0,
+      "step": 2168
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 8.43649946638207e-05,
+      "loss": 0.8708,
+      "step": 2169
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 8.431163287086447e-05,
+      "loss": 0.7676,
+      "step": 2170
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 8.425827107790823e-05,
+      "loss": 0.9473,
+      "step": 2171
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 8.420490928495199e-05,
+      "loss": 0.9434,
+      "step": 2172
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 8.415154749199572e-05,
+      "loss": 0.9175,
+      "step": 2173
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 8.409818569903949e-05,
+      "loss": 0.8325,
+      "step": 2174
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 8.404482390608325e-05,
+      "loss": 0.4898,
+      "step": 2175
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 8.399146211312701e-05,
+      "loss": 0.8591,
+      "step": 2176
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 8.393810032017077e-05,
+      "loss": 0.8806,
+      "step": 2177
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 8.388473852721452e-05,
+      "loss": 0.8283,
+      "step": 2178
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 8.383137673425827e-05,
+      "loss": 0.7642,
+      "step": 2179
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 8.377801494130203e-05,
+      "loss": 0.7774,
+      "step": 2180
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 8.372465314834579e-05,
+      "loss": 0.4386,
+      "step": 2181
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 8.367129135538954e-05,
+      "loss": 0.3094,
+      "step": 2182
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 8.36179295624333e-05,
+      "loss": 0.3854,
+      "step": 2183
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 8.356456776947705e-05,
+      "loss": 0.704,
+      "step": 2184
+    },
+    {
+      "epoch": 1.75,
+      "eval_loss": 0.9970366358757019,
+      "eval_runtime": 7699.2981,
+      "eval_samples_per_second": 1.299,
+      "eval_steps_per_second": 0.081,
+      "step": 2184
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 8.351120597652081e-05,
+      "loss": 0.8424,
+      "step": 2185
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 8.345784418356458e-05,
+      "loss": 0.8751,
+      "step": 2186
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 8.340448239060832e-05,
+      "loss": 0.9549,
+      "step": 2187
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 8.335112059765209e-05,
+      "loss": 0.3615,
+      "step": 2188
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 8.329775880469585e-05,
+      "loss": 0.8535,
+      "step": 2189
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 8.32443970117396e-05,
+      "loss": 0.8692,
+      "step": 2190
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 8.319103521878336e-05,
+      "loss": 0.8593,
+      "step": 2191
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 8.31376734258271e-05,
+      "loss": 0.799,
+      "step": 2192
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 8.308431163287087e-05,
+      "loss": 0.9762,
+      "step": 2193
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 8.303094983991463e-05,
+      "loss": 0.7901,
+      "step": 2194
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 8.297758804695838e-05,
+      "loss": 0.9837,
+      "step": 2195
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 8.292422625400214e-05,
+      "loss": 0.4507,
+      "step": 2196
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 8.287086446104589e-05,
+      "loss": 0.889,
+      "step": 2197
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 8.281750266808965e-05,
+      "loss": 0.8046,
+      "step": 2198
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 8.276414087513341e-05,
+      "loss": 0.9359,
+      "step": 2199
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 8.271077908217718e-05,
+      "loss": 0.7655,
+      "step": 2200
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 8.265741728922092e-05,
+      "loss": 0.8919,
+      "step": 2201
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 8.260405549626467e-05,
+      "loss": 0.7915,
+      "step": 2202
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 8.255069370330843e-05,
+      "loss": 0.7936,
+      "step": 2203
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 8.24973319103522e-05,
+      "loss": 0.8924,
+      "step": 2204
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 8.244397011739596e-05,
+      "loss": 0.8213,
+      "step": 2205
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 8.23906083244397e-05,
+      "loss": 0.8548,
+      "step": 2206
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 8.233724653148345e-05,
+      "loss": 0.8916,
+      "step": 2207
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 8.228388473852722e-05,
+      "loss": 0.8047,
+      "step": 2208
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 8.223052294557098e-05,
+      "loss": 0.9559,
+      "step": 2209
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 8.217716115261474e-05,
+      "loss": 0.9222,
+      "step": 2210
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 8.212379935965849e-05,
+      "loss": 0.4107,
+      "step": 2211
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 8.207043756670224e-05,
+      "loss": 0.8491,
+      "step": 2212
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 8.2017075773746e-05,
+      "loss": 0.8177,
+      "step": 2213
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 8.196371398078976e-05,
+      "loss": 0.9608,
+      "step": 2214
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 8.191035218783352e-05,
+      "loss": 0.8279,
+      "step": 2215
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 8.185699039487727e-05,
+      "loss": 0.7583,
+      "step": 2216
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 8.180362860192102e-05,
+      "loss": 0.8961,
+      "step": 2217
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 8.175026680896478e-05,
+      "loss": 0.4439,
+      "step": 2218
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 8.169690501600854e-05,
+      "loss": 0.4085,
+      "step": 2219
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 8.164354322305229e-05,
+      "loss": 0.4273,
+      "step": 2220
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 8.159018143009605e-05,
+      "loss": 0.8467,
+      "step": 2221
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 8.153681963713982e-05,
+      "loss": 0.9339,
+      "step": 2222
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 8.148345784418356e-05,
+      "loss": 0.4235,
+      "step": 2223
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 8.143009605122733e-05,
+      "loss": 0.8787,
+      "step": 2224
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 8.137673425827108e-05,
+      "loss": 0.8559,
+      "step": 2225
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 8.132337246531484e-05,
+      "loss": 0.9365,
+      "step": 2226
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 8.12700106723586e-05,
+      "loss": 0.8825,
+      "step": 2227
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 8.121664887940236e-05,
+      "loss": 0.8362,
+      "step": 2228
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 8.116328708644611e-05,
+      "loss": 1.0341,
+      "step": 2229
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 8.110992529348986e-05,
+      "loss": 0.4825,
+      "step": 2230
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 8.105656350053362e-05,
+      "loss": 0.9559,
+      "step": 2231
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 8.100320170757738e-05,
+      "loss": 0.7901,
+      "step": 2232
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 8.094983991462114e-05,
+      "loss": 0.8766,
+      "step": 2233
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 8.089647812166489e-05,
+      "loss": 0.8474,
+      "step": 2234
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 8.084311632870864e-05,
+      "loss": 0.7598,
+      "step": 2235
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 8.07897545357524e-05,
+      "loss": 0.9119,
+      "step": 2236
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 8.073639274279616e-05,
+      "loss": 0.4252,
+      "step": 2237
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 8.068303094983993e-05,
+      "loss": 0.8269,
+      "step": 2238
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 8.062966915688367e-05,
+      "loss": 0.9821,
+      "step": 2239
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 8.057630736392742e-05,
+      "loss": 0.7845,
+      "step": 2240
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 8.052294557097119e-05,
+      "loss": 0.8883,
+      "step": 2241
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 8.046958377801495e-05,
+      "loss": 0.8534,
+      "step": 2242
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 8.041622198505871e-05,
+      "loss": 0.9714,
+      "step": 2243
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 8.036286019210246e-05,
+      "loss": 0.9922,
+      "step": 2244
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 8.03094983991462e-05,
+      "loss": 0.8855,
+      "step": 2245
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 8.025613660618997e-05,
+      "loss": 0.7454,
+      "step": 2246
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 8.020277481323373e-05,
+      "loss": 1.1087,
+      "step": 2247
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 8.014941302027749e-05,
+      "loss": 0.7769,
+      "step": 2248
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 8.009605122732124e-05,
+      "loss": 0.8608,
+      "step": 2249
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 8.0042689434365e-05,
+      "loss": 0.4287,
+      "step": 2250
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 7.998932764140875e-05,
+      "loss": 0.7658,
+      "step": 2251
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 7.993596584845251e-05,
+      "loss": 0.9512,
+      "step": 2252
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 7.988260405549627e-05,
+      "loss": 0.9311,
+      "step": 2253
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 7.982924226254002e-05,
+      "loss": 0.9727,
+      "step": 2254
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 7.977588046958379e-05,
+      "loss": 0.8801,
+      "step": 2255
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 7.972251867662753e-05,
+      "loss": 0.9117,
+      "step": 2256
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 7.96691568836713e-05,
+      "loss": 0.7347,
+      "step": 2257
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 7.961579509071506e-05,
+      "loss": 0.8344,
+      "step": 2258
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 7.95624332977588e-05,
+      "loss": 0.9308,
+      "step": 2259
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 7.950907150480257e-05,
+      "loss": 0.8247,
+      "step": 2260
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 7.945570971184633e-05,
+      "loss": 0.8866,
+      "step": 2261
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 7.940234791889008e-05,
+      "loss": 1.0452,
+      "step": 2262
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 7.934898612593383e-05,
+      "loss": 0.7961,
+      "step": 2263
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 7.929562433297759e-05,
+      "loss": 0.3126,
+      "step": 2264
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 7.924226254002135e-05,
+      "loss": 0.7347,
+      "step": 2265
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 7.918890074706511e-05,
+      "loss": 0.3327,
+      "step": 2266
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 7.913553895410886e-05,
+      "loss": 0.7647,
+      "step": 2267
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 7.908217716115261e-05,
+      "loss": 0.8463,
+      "step": 2268
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 7.902881536819637e-05,
+      "loss": 0.8754,
+      "step": 2269
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 7.897545357524013e-05,
+      "loss": 0.7137,
+      "step": 2270
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 7.89220917822839e-05,
+      "loss": 0.8547,
+      "step": 2271
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 7.886872998932764e-05,
+      "loss": 0.9191,
+      "step": 2272
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 7.881536819637139e-05,
+      "loss": 0.8798,
+      "step": 2273
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 7.876200640341515e-05,
+      "loss": 0.3671,
+      "step": 2274
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 7.870864461045892e-05,
+      "loss": 0.7629,
+      "step": 2275
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 7.865528281750268e-05,
+      "loss": 0.8156,
+      "step": 2276
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 7.860192102454643e-05,
+      "loss": 0.8222,
+      "step": 2277
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 7.854855923159017e-05,
+      "loss": 0.4997,
+      "step": 2278
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 7.849519743863394e-05,
+      "loss": 0.7666,
+      "step": 2279
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 7.84418356456777e-05,
+      "loss": 1.0471,
+      "step": 2280
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 7.838847385272146e-05,
+      "loss": 0.8073,
+      "step": 2281
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 7.833511205976521e-05,
+      "loss": 0.9088,
+      "step": 2282
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 7.828175026680897e-05,
+      "loss": 0.9403,
+      "step": 2283
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 7.822838847385272e-05,
+      "loss": 0.4867,
+      "step": 2284
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 7.817502668089648e-05,
+      "loss": 0.9833,
+      "step": 2285
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 7.812166488794024e-05,
+      "loss": 0.871,
+      "step": 2286
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 7.806830309498399e-05,
+      "loss": 0.7711,
+      "step": 2287
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 7.801494130202775e-05,
+      "loss": 0.8225,
+      "step": 2288
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 7.796157950907152e-05,
+      "loss": 0.491,
+      "step": 2289
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 7.790821771611526e-05,
+      "loss": 0.7884,
+      "step": 2290
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 7.785485592315903e-05,
+      "loss": 0.4709,
+      "step": 2291
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 7.780149413020277e-05,
+      "loss": 0.4694,
+      "step": 2292
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 7.774813233724654e-05,
+      "loss": 0.8271,
+      "step": 2293
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 7.76947705442903e-05,
+      "loss": 0.4556,
+      "step": 2294
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 7.764140875133405e-05,
+      "loss": 0.7932,
+      "step": 2295
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 7.758804695837781e-05,
+      "loss": 0.9199,
+      "step": 2296
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 7.753468516542156e-05,
+      "loss": 0.4909,
+      "step": 2297
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 7.748132337246532e-05,
+      "loss": 0.9329,
+      "step": 2298
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 7.742796157950908e-05,
+      "loss": 0.8202,
+      "step": 2299
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 7.737459978655284e-05,
+      "loss": 0.7967,
+      "step": 2300
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 7.732123799359658e-05,
+      "loss": 0.8342,
+      "step": 2301
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 7.726787620064034e-05,
+      "loss": 0.7708,
+      "step": 2302
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 7.72145144076841e-05,
+      "loss": 0.0,
+      "step": 2303
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 7.716115261472786e-05,
+      "loss": 0.7461,
+      "step": 2304
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 7.710779082177163e-05,
+      "loss": 0.9379,
+      "step": 2305
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 7.705442902881536e-05,
+      "loss": 0.888,
+      "step": 2306
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 7.700106723585912e-05,
+      "loss": 0.5415,
+      "step": 2307
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 7.694770544290288e-05,
+      "loss": 0.877,
+      "step": 2308
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 7.689434364994665e-05,
+      "loss": 0.8804,
+      "step": 2309
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 7.68409818569904e-05,
+      "loss": 0.9484,
+      "step": 2310
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 7.678762006403416e-05,
+      "loss": 1.0547,
+      "step": 2311
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 7.67342582710779e-05,
+      "loss": 0.0,
+      "step": 2312
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 7.668089647812167e-05,
+      "loss": 0.7955,
+      "step": 2313
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 7.662753468516543e-05,
+      "loss": 0.9886,
+      "step": 2314
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 7.657417289220918e-05,
+      "loss": 0.4376,
+      "step": 2315
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 7.652081109925294e-05,
+      "loss": 0.9336,
+      "step": 2316
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 7.646744930629669e-05,
+      "loss": 0.8459,
+      "step": 2317
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 7.641408751334045e-05,
+      "loss": 0.8343,
+      "step": 2318
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 7.636072572038421e-05,
+      "loss": 0.8851,
+      "step": 2319
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 7.630736392742796e-05,
+      "loss": 0.4835,
+      "step": 2320
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 7.625400213447172e-05,
+      "loss": 0.9941,
+      "step": 2321
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 7.620064034151548e-05,
+      "loss": 0.8886,
+      "step": 2322
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 7.614727854855923e-05,
+      "loss": 0.7274,
+      "step": 2323
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 7.6093916755603e-05,
+      "loss": 0.9607,
+      "step": 2324
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 7.604055496264674e-05,
+      "loss": 1.0546,
+      "step": 2325
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 7.59871931696905e-05,
+      "loss": 0.8916,
+      "step": 2326
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 7.593383137673427e-05,
+      "loss": 0.8132,
+      "step": 2327
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 7.588046958377802e-05,
+      "loss": 0.8675,
+      "step": 2328
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 7.582710779082178e-05,
+      "loss": 0.7941,
+      "step": 2329
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 7.577374599786553e-05,
+      "loss": 0.8091,
+      "step": 2330
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 7.572038420490929e-05,
+      "loss": 0.8856,
+      "step": 2331
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 7.566702241195305e-05,
+      "loss": 0.707,
+      "step": 2332
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 7.561366061899681e-05,
+      "loss": 0.4155,
+      "step": 2333
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 7.556029882604056e-05,
+      "loss": 0.9851,
+      "step": 2334
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 7.550693703308431e-05,
+      "loss": 0.8714,
+      "step": 2335
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 7.545357524012807e-05,
+      "loss": 1.0882,
+      "step": 2336
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 7.540021344717183e-05,
+      "loss": 1.2311,
+      "step": 2337
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 7.53468516542156e-05,
+      "loss": 0.9767,
+      "step": 2338
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 7.529348986125933e-05,
+      "loss": 0.8202,
+      "step": 2339
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 7.524012806830309e-05,
+      "loss": 0.8,
+      "step": 2340
+    },
+    {
+      "epoch": 1.87,
+      "eval_loss": 0.9943591952323914,
+      "eval_runtime": 7639.1444,
+      "eval_samples_per_second": 1.309,
+      "eval_steps_per_second": 0.082,
+      "step": 2340
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 7.518676627534685e-05,
+      "loss": 0.8826,
+      "step": 2341
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 7.513340448239062e-05,
+      "loss": 0.8318,
+      "step": 2342
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 7.508004268943438e-05,
+      "loss": 0.3752,
+      "step": 2343
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 7.502668089647813e-05,
+      "loss": 0.779,
+      "step": 2344
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 7.497331910352187e-05,
+      "loss": 0.8601,
+      "step": 2345
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 7.491995731056564e-05,
+      "loss": 0.9762,
+      "step": 2346
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 7.48665955176094e-05,
+      "loss": 0.4314,
+      "step": 2347
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 7.481323372465315e-05,
+      "loss": 0.9164,
+      "step": 2348
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 7.475987193169691e-05,
+      "loss": 0.8565,
+      "step": 2349
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 7.470651013874067e-05,
+      "loss": 0.0,
+      "step": 2350
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 7.465314834578442e-05,
+      "loss": 0.9432,
+      "step": 2351
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 7.459978655282818e-05,
+      "loss": 0.4035,
+      "step": 2352
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 7.454642475987193e-05,
+      "loss": 0.4378,
+      "step": 2353
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 7.449306296691569e-05,
+      "loss": 0.3672,
+      "step": 2354
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 7.443970117395945e-05,
+      "loss": 0.4772,
+      "step": 2355
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 7.43863393810032e-05,
+      "loss": 0.9104,
+      "step": 2356
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 7.433297758804696e-05,
+      "loss": 0.8218,
+      "step": 2357
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 7.427961579509071e-05,
+      "loss": 0.9344,
+      "step": 2358
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 7.422625400213447e-05,
+      "loss": 0.8979,
+      "step": 2359
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 7.417289220917824e-05,
+      "loss": 0.8159,
+      "step": 2360
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 7.4119530416222e-05,
+      "loss": 0.782,
+      "step": 2361
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 7.406616862326575e-05,
+      "loss": 0.8834,
+      "step": 2362
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 7.40128068303095e-05,
+      "loss": 0.405,
+      "step": 2363
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 7.395944503735326e-05,
+      "loss": 0.5173,
+      "step": 2364
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 7.390608324439702e-05,
+      "loss": 0.962,
+      "step": 2365
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 7.385272145144078e-05,
+      "loss": 0.9101,
+      "step": 2366
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 7.379935965848453e-05,
+      "loss": 0.7896,
+      "step": 2367
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 7.374599786552828e-05,
+      "loss": 0.8628,
+      "step": 2368
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 7.369263607257204e-05,
+      "loss": 0.7721,
+      "step": 2369
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 7.36392742796158e-05,
+      "loss": 0.9357,
+      "step": 2370
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 7.358591248665956e-05,
+      "loss": 0.9285,
+      "step": 2371
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 7.353255069370331e-05,
+      "loss": 0.8037,
+      "step": 2372
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 7.347918890074706e-05,
+      "loss": 1.0331,
+      "step": 2373
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 7.342582710779082e-05,
+      "loss": 0.4626,
+      "step": 2374
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 7.337246531483458e-05,
+      "loss": 0.8717,
+      "step": 2375
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 7.331910352187835e-05,
+      "loss": 0.9286,
+      "step": 2376
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 7.32657417289221e-05,
+      "loss": 0.8101,
+      "step": 2377
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 7.321237993596584e-05,
+      "loss": 0.4788,
+      "step": 2378
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 7.31590181430096e-05,
+      "loss": 0.9047,
+      "step": 2379
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 7.310565635005337e-05,
+      "loss": 0.8193,
+      "step": 2380
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 7.305229455709713e-05,
+      "loss": 0.9454,
+      "step": 2381
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 7.299893276414088e-05,
+      "loss": 0.7493,
+      "step": 2382
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 7.294557097118464e-05,
+      "loss": 0.9644,
+      "step": 2383
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 7.289220917822839e-05,
+      "loss": 0.9774,
+      "step": 2384
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 7.283884738527215e-05,
+      "loss": 0.7951,
+      "step": 2385
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 7.278548559231591e-05,
+      "loss": 0.8423,
+      "step": 2386
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 7.273212379935966e-05,
+      "loss": 0.9462,
+      "step": 2387
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 7.267876200640342e-05,
+      "loss": 0.8601,
+      "step": 2388
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 7.262540021344717e-05,
+      "loss": 0.8656,
+      "step": 2389
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 7.257203842049093e-05,
+      "loss": 0.869,
+      "step": 2390
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 7.251867662753468e-05,
+      "loss": 0.8255,
+      "step": 2391
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 7.246531483457844e-05,
+      "loss": 0.8714,
+      "step": 2392
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 7.24119530416222e-05,
+      "loss": 0.4373,
+      "step": 2393
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 7.235859124866597e-05,
+      "loss": 0.4292,
+      "step": 2394
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 7.230522945570972e-05,
+      "loss": 0.848,
+      "step": 2395
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 7.225186766275346e-05,
+      "loss": 0.3574,
+      "step": 2396
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 7.219850586979723e-05,
+      "loss": 0.8886,
+      "step": 2397
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 7.214514407684099e-05,
+      "loss": 0.8979,
+      "step": 2398
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 7.209178228388475e-05,
+      "loss": 0.8428,
+      "step": 2399
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 7.20384204909285e-05,
+      "loss": 0.806,
+      "step": 2400
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 7.198505869797225e-05,
+      "loss": 0.7955,
+      "step": 2401
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 7.193169690501601e-05,
+      "loss": 0.7544,
+      "step": 2402
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 7.187833511205977e-05,
+      "loss": 0.8038,
+      "step": 2403
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 7.182497331910353e-05,
+      "loss": 0.51,
+      "step": 2404
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 7.177161152614728e-05,
+      "loss": 0.795,
+      "step": 2405
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 7.171824973319103e-05,
+      "loss": 0.9405,
+      "step": 2406
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 7.166488794023479e-05,
+      "loss": 0.8525,
+      "step": 2407
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 7.161152614727855e-05,
+      "loss": 0.8731,
+      "step": 2408
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 7.155816435432232e-05,
+      "loss": 0.9391,
+      "step": 2409
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 7.150480256136606e-05,
+      "loss": 0.5151,
+      "step": 2410
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 7.145144076840983e-05,
+      "loss": 0.4415,
+      "step": 2411
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 7.139807897545357e-05,
+      "loss": 0.7966,
+      "step": 2412
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 7.134471718249734e-05,
+      "loss": 0.9679,
+      "step": 2413
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 7.12913553895411e-05,
+      "loss": 0.7924,
+      "step": 2414
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 7.123799359658485e-05,
+      "loss": 0.4305,
+      "step": 2415
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 7.118463180362861e-05,
+      "loss": 0.4907,
+      "step": 2416
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 7.113127001067236e-05,
+      "loss": 0.3336,
+      "step": 2417
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 7.107790821771612e-05,
+      "loss": 0.8789,
+      "step": 2418
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 7.102454642475988e-05,
+      "loss": 0.7333,
+      "step": 2419
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 7.097118463180363e-05,
+      "loss": 0.8186,
+      "step": 2420
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 7.091782283884739e-05,
+      "loss": 0.9501,
+      "step": 2421
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 7.086446104589115e-05,
+      "loss": 0.9132,
+      "step": 2422
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 7.08110992529349e-05,
+      "loss": 0.9138,
+      "step": 2423
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 7.075773745997866e-05,
+      "loss": 0.8264,
+      "step": 2424
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 7.070437566702241e-05,
+      "loss": 0.5217,
+      "step": 2425
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 7.065101387406617e-05,
+      "loss": 0.3591,
+      "step": 2426
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 7.059765208110994e-05,
+      "loss": 0.8504,
+      "step": 2427
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 7.054429028815368e-05,
+      "loss": 0.9017,
+      "step": 2428
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 7.049092849519743e-05,
+      "loss": 0.397,
+      "step": 2429
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 7.04375667022412e-05,
+      "loss": 0.7841,
+      "step": 2430
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 7.038420490928496e-05,
+      "loss": 0.3144,
+      "step": 2431
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 7.033084311632872e-05,
+      "loss": 0.951,
+      "step": 2432
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 7.027748132337248e-05,
+      "loss": 0.8897,
+      "step": 2433
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 7.022411953041622e-05,
+      "loss": 0.9301,
+      "step": 2434
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 7.017075773745998e-05,
+      "loss": 0.7439,
+      "step": 2435
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 7.011739594450374e-05,
+      "loss": 0.5264,
+      "step": 2436
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 7.00640341515475e-05,
+      "loss": 0.8636,
+      "step": 2437
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 7.001067235859125e-05,
+      "loss": 0.9111,
+      "step": 2438
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 6.9957310565635e-05,
+      "loss": 0.8404,
+      "step": 2439
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 6.990394877267876e-05,
+      "loss": 0.8439,
+      "step": 2440
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 6.985058697972252e-05,
+      "loss": 0.4241,
+      "step": 2441
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 6.979722518676628e-05,
+      "loss": 0.3794,
+      "step": 2442
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 6.974386339381003e-05,
+      "loss": 0.9248,
+      "step": 2443
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 6.96905016008538e-05,
+      "loss": 0.8678,
+      "step": 2444
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 6.963713980789754e-05,
+      "loss": 0.4997,
+      "step": 2445
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 6.95837780149413e-05,
+      "loss": 0.9508,
+      "step": 2446
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 6.953041622198507e-05,
+      "loss": 0.4647,
+      "step": 2447
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 6.947705442902882e-05,
+      "loss": 0.6667,
+      "step": 2448
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 6.942369263607258e-05,
+      "loss": 0.9389,
+      "step": 2449
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 6.937033084311633e-05,
+      "loss": 0.9093,
+      "step": 2450
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 6.931696905016009e-05,
+      "loss": 0.9188,
+      "step": 2451
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 6.926360725720385e-05,
+      "loss": 0.9977,
+      "step": 2452
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 6.92102454642476e-05,
+      "loss": 0.7926,
+      "step": 2453
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 6.915688367129136e-05,
+      "loss": 0.7421,
+      "step": 2454
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 6.910352187833512e-05,
+      "loss": 0.887,
+      "step": 2455
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 6.905016008537887e-05,
+      "loss": 0.9402,
+      "step": 2456
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 6.899679829242263e-05,
+      "loss": 0.8065,
+      "step": 2457
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 6.894343649946638e-05,
+      "loss": 0.6928,
+      "step": 2458
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 6.889007470651014e-05,
+      "loss": 0.846,
+      "step": 2459
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 6.88367129135539e-05,
+      "loss": 0.894,
+      "step": 2460
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 6.878335112059765e-05,
+      "loss": 0.8956,
+      "step": 2461
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 6.872998932764141e-05,
+      "loss": 0.3987,
+      "step": 2462
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 6.867662753468516e-05,
+      "loss": 0.9568,
+      "step": 2463
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 6.862326574172893e-05,
+      "loss": 0.414,
+      "step": 2464
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 6.856990394877269e-05,
+      "loss": 0.8134,
+      "step": 2465
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 6.851654215581645e-05,
+      "loss": 0.9064,
+      "step": 2466
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 6.84631803628602e-05,
+      "loss": 0.5324,
+      "step": 2467
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 6.840981856990395e-05,
+      "loss": 0.817,
+      "step": 2468
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 6.835645677694771e-05,
+      "loss": 0.5052,
+      "step": 2469
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 6.830309498399147e-05,
+      "loss": 0.4366,
+      "step": 2470
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 6.824973319103523e-05,
+      "loss": 0.933,
+      "step": 2471
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 6.819637139807898e-05,
+      "loss": 0.8377,
+      "step": 2472
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 6.814300960512273e-05,
+      "loss": 0.9906,
+      "step": 2473
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 6.808964781216649e-05,
+      "loss": 0.7894,
+      "step": 2474
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 6.803628601921025e-05,
+      "loss": 0.7458,
+      "step": 2475
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 6.7982924226254e-05,
+      "loss": 0.8296,
+      "step": 2476
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 6.792956243329776e-05,
+      "loss": 0.8754,
+      "step": 2477
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 6.787620064034151e-05,
+      "loss": 0.8495,
+      "step": 2478
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 6.782283884738527e-05,
+      "loss": 0.4544,
+      "step": 2479
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 6.776947705442904e-05,
+      "loss": 0.7526,
+      "step": 2480
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 6.771611526147278e-05,
+      "loss": 0.9622,
+      "step": 2481
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 6.766275346851655e-05,
+      "loss": 0.9733,
+      "step": 2482
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 6.760939167556031e-05,
+      "loss": 0.9543,
+      "step": 2483
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 6.755602988260406e-05,
+      "loss": 0.8135,
+      "step": 2484
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 6.750266808964782e-05,
+      "loss": 0.8332,
+      "step": 2485
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 6.744930629669157e-05,
+      "loss": 0.8497,
+      "step": 2486
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 6.739594450373533e-05,
+      "loss": 0.8181,
+      "step": 2487
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 6.734258271077909e-05,
+      "loss": 0.4249,
+      "step": 2488
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 6.728922091782284e-05,
+      "loss": 0.8563,
+      "step": 2489
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 6.72358591248666e-05,
+      "loss": 0.9138,
+      "step": 2490
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 6.718249733191035e-05,
+      "loss": 0.855,
+      "step": 2491
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 6.712913553895411e-05,
+      "loss": 0.7979,
+      "step": 2492
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 6.707577374599787e-05,
+      "loss": 0.3848,
+      "step": 2493
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 6.702241195304164e-05,
+      "loss": 0.8677,
+      "step": 2494
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 6.696905016008538e-05,
+      "loss": 0.9003,
+      "step": 2495
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 6.691568836712913e-05,
+      "loss": 0.8405,
+      "step": 2496
+    },
+    {
+      "epoch": 2.0,
+      "eval_loss": 0.98443204164505,
+      "eval_runtime": 7418.4679,
+      "eval_samples_per_second": 1.348,
+      "eval_steps_per_second": 0.084,
+      "step": 2496
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 6.68623265741729e-05,
+      "loss": 0.9147,
+      "step": 2497
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 6.680896478121666e-05,
+      "loss": 0.7995,
+      "step": 2498
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 6.675560298826042e-05,
+      "loss": 0.7943,
+      "step": 2499
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 6.670224119530417e-05,
+      "loss": 0.8097,
+      "step": 2500
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 6.664887940234791e-05,
+      "loss": 0.8224,
+      "step": 2501
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 6.659551760939168e-05,
+      "loss": 0.5647,
+      "step": 2502
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 6.654215581643544e-05,
+      "loss": 0.7268,
+      "step": 2503
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 6.64887940234792e-05,
+      "loss": 0.6974,
+      "step": 2504
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 6.643543223052295e-05,
+      "loss": 0.2986,
+      "step": 2505
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 6.63820704375667e-05,
+      "loss": 0.6696,
+      "step": 2506
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 6.632870864461046e-05,
+      "loss": 0.3596,
+      "step": 2507
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 6.627534685165422e-05,
+      "loss": 0.6946,
+      "step": 2508
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 6.622198505869798e-05,
+      "loss": 0.4893,
+      "step": 2509
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 6.616862326574173e-05,
+      "loss": 0.751,
+      "step": 2510
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 6.611526147278548e-05,
+      "loss": 0.7372,
+      "step": 2511
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 6.606189967982924e-05,
+      "loss": 0.7397,
+      "step": 2512
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 6.6008537886873e-05,
+      "loss": 0.6961,
+      "step": 2513
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 6.595517609391677e-05,
+      "loss": 0.666,
+      "step": 2514
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 6.590181430096051e-05,
+      "loss": 0.33,
+      "step": 2515
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 6.584845250800428e-05,
+      "loss": 0.3031,
+      "step": 2516
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 6.579509071504802e-05,
+      "loss": 0.6226,
+      "step": 2517
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 6.574172892209179e-05,
+      "loss": 0.7875,
+      "step": 2518
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 6.568836712913554e-05,
+      "loss": 0.4156,
+      "step": 2519
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 6.56350053361793e-05,
+      "loss": 0.7381,
+      "step": 2520
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 6.558164354322306e-05,
+      "loss": 0.7007,
+      "step": 2521
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 6.552828175026681e-05,
+      "loss": 0.7706,
+      "step": 2522
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 6.547491995731057e-05,
+      "loss": 0.3384,
+      "step": 2523
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 6.542155816435432e-05,
+      "loss": 0.6606,
+      "step": 2524
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 6.536819637139808e-05,
+      "loss": 0.3656,
+      "step": 2525
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 6.531483457844184e-05,
+      "loss": 0.5949,
+      "step": 2526
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 6.52614727854856e-05,
+      "loss": 0.8104,
+      "step": 2527
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 6.520811099252935e-05,
+      "loss": 0.7178,
+      "step": 2528
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 6.51547491995731e-05,
+      "loss": 0.7954,
+      "step": 2529
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 6.510138740661686e-05,
+      "loss": 0.6876,
+      "step": 2530
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 6.504802561366062e-05,
+      "loss": 0.6398,
+      "step": 2531
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 6.499466382070439e-05,
+      "loss": 0.6938,
+      "step": 2532
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 6.494130202774814e-05,
+      "loss": 0.6903,
+      "step": 2533
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 6.488794023479188e-05,
+      "loss": 0.6446,
+      "step": 2534
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 6.483457844183565e-05,
+      "loss": 0.8096,
+      "step": 2535
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 6.478121664887941e-05,
+      "loss": 0.8013,
+      "step": 2536
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 6.472785485592317e-05,
+      "loss": 0.3523,
+      "step": 2537
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 6.467449306296692e-05,
+      "loss": 0.784,
+      "step": 2538
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 6.462113127001067e-05,
+      "loss": 0.5807,
+      "step": 2539
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 6.456776947705443e-05,
+      "loss": 0.6419,
+      "step": 2540
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 6.451440768409819e-05,
+      "loss": 0.8019,
+      "step": 2541
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 6.446104589114195e-05,
+      "loss": 0.7003,
+      "step": 2542
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 6.44076840981857e-05,
+      "loss": 0.6614,
+      "step": 2543
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 6.435432230522946e-05,
+      "loss": 1.0036,
+      "step": 2544
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 6.430096051227321e-05,
+      "loss": 0.7434,
+      "step": 2545
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 6.424759871931697e-05,
+      "loss": 0.7949,
+      "step": 2546
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 6.419423692636073e-05,
+      "loss": 0.7279,
+      "step": 2547
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 6.414087513340448e-05,
+      "loss": 0.7272,
+      "step": 2548
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 6.408751334044825e-05,
+      "loss": 0.7495,
+      "step": 2549
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 6.4034151547492e-05,
+      "loss": 0.7475,
+      "step": 2550
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 6.398078975453576e-05,
+      "loss": 0.7115,
+      "step": 2551
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 6.392742796157952e-05,
+      "loss": 0.5902,
+      "step": 2552
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 6.387406616862327e-05,
+      "loss": 0.8086,
+      "step": 2553
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 6.382070437566703e-05,
+      "loss": 0.6273,
+      "step": 2554
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 6.376734258271079e-05,
+      "loss": 0.8093,
+      "step": 2555
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 6.371398078975454e-05,
+      "loss": 0.7256,
+      "step": 2556
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 6.366061899679829e-05,
+      "loss": 0.3326,
+      "step": 2557
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 6.360725720384205e-05,
+      "loss": 0.7453,
+      "step": 2558
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 6.355389541088581e-05,
+      "loss": 0.8717,
+      "step": 2559
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 6.350053361792957e-05,
+      "loss": 0.7742,
+      "step": 2560
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 6.344717182497332e-05,
+      "loss": 0.67,
+      "step": 2561
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 6.339381003201707e-05,
+      "loss": 0.7667,
+      "step": 2562
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 6.334044823906083e-05,
+      "loss": 0.6732,
+      "step": 2563
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 6.32870864461046e-05,
+      "loss": 0.734,
+      "step": 2564
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 6.323372465314836e-05,
+      "loss": 0.7376,
+      "step": 2565
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 6.31803628601921e-05,
+      "loss": 0.7902,
+      "step": 2566
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 6.312700106723585e-05,
+      "loss": 0.7415,
+      "step": 2567
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 6.307363927427961e-05,
+      "loss": 0.3339,
+      "step": 2568
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 6.302027748132338e-05,
+      "loss": 0.6647,
+      "step": 2569
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 6.296691568836714e-05,
+      "loss": 0.5982,
+      "step": 2570
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 6.291355389541089e-05,
+      "loss": 0.6098,
+      "step": 2571
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 6.286019210245464e-05,
+      "loss": 0.7196,
+      "step": 2572
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 6.28068303094984e-05,
+      "loss": 0.6818,
+      "step": 2573
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 6.275346851654216e-05,
+      "loss": 0.704,
+      "step": 2574
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 6.270010672358592e-05,
+      "loss": 0.772,
+      "step": 2575
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 6.264674493062967e-05,
+      "loss": 0.6989,
+      "step": 2576
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 6.259338313767343e-05,
+      "loss": 0.8629,
+      "step": 2577
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 6.254002134471718e-05,
+      "loss": 0.6965,
+      "step": 2578
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 6.248665955176094e-05,
+      "loss": 0.7148,
+      "step": 2579
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 6.24332977588047e-05,
+      "loss": 0.3413,
+      "step": 2580
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 6.237993596584845e-05,
+      "loss": 0.6789,
+      "step": 2581
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 6.232657417289221e-05,
+      "loss": 0.6628,
+      "step": 2582
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 6.227321237993596e-05,
+      "loss": 0.6455,
+      "step": 2583
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 6.221985058697972e-05,
+      "loss": 0.6035,
+      "step": 2584
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 6.216648879402349e-05,
+      "loss": 0.6688,
+      "step": 2585
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 6.211312700106723e-05,
+      "loss": 0.649,
+      "step": 2586
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 6.2059765208111e-05,
+      "loss": 0.7667,
+      "step": 2587
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 6.200640341515476e-05,
+      "loss": 0.6972,
+      "step": 2588
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 6.195304162219851e-05,
+      "loss": 0.5825,
+      "step": 2589
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 6.189967982924227e-05,
+      "loss": 0.7156,
+      "step": 2590
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 6.184631803628602e-05,
+      "loss": 0.665,
+      "step": 2591
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 6.179295624332978e-05,
+      "loss": 0.7665,
+      "step": 2592
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 6.173959445037354e-05,
+      "loss": 0.8238,
+      "step": 2593
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 6.168623265741729e-05,
+      "loss": 0.6379,
+      "step": 2594
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 6.163287086446105e-05,
+      "loss": 0.8164,
+      "step": 2595
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 6.15795090715048e-05,
+      "loss": 0.5701,
+      "step": 2596
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 6.152614727854856e-05,
+      "loss": 0.7154,
+      "step": 2597
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 6.147278548559232e-05,
+      "loss": 0.7495,
+      "step": 2598
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 6.141942369263609e-05,
+      "loss": 0.3599,
+      "step": 2599
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 6.136606189967982e-05,
+      "loss": 0.3024,
+      "step": 2600
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 6.131270010672358e-05,
+      "loss": 0.6361,
+      "step": 2601
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 6.125933831376735e-05,
+      "loss": 0.6687,
+      "step": 2602
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 6.120597652081111e-05,
+      "loss": 0.6596,
+      "step": 2603
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 6.115261472785486e-05,
+      "loss": 0.6214,
+      "step": 2604
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 6.109925293489862e-05,
+      "loss": 0.7901,
+      "step": 2605
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 6.104589114194237e-05,
+      "loss": 0.7863,
+      "step": 2606
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 6.099252934898613e-05,
+      "loss": 0.7971,
+      "step": 2607
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 6.093916755602988e-05,
+      "loss": 0.3232,
+      "step": 2608
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 6.0885805763073645e-05,
+      "loss": 0.6116,
+      "step": 2609
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 6.08324439701174e-05,
+      "loss": 0.6641,
+      "step": 2610
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 6.077908217716115e-05,
+      "loss": 0.8583,
+      "step": 2611
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 6.072572038420491e-05,
+      "loss": 0.637,
+      "step": 2612
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 6.0672358591248666e-05,
+      "loss": 0.742,
+      "step": 2613
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 6.061899679829243e-05,
+      "loss": 0.6112,
+      "step": 2614
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 6.056563500533618e-05,
+      "loss": 0.5487,
+      "step": 2615
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 6.0512273212379945e-05,
+      "loss": 0.7378,
+      "step": 2616
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 6.045891141942369e-05,
+      "loss": 0.3308,
+      "step": 2617
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 6.040554962646745e-05,
+      "loss": 0.4606,
+      "step": 2618
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 6.035218783351121e-05,
+      "loss": 0.6825,
+      "step": 2619
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 6.0298826040554966e-05,
+      "loss": 0.3105,
+      "step": 2620
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 6.024546424759873e-05,
+      "loss": 0.4129,
+      "step": 2621
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 6.0192102454642476e-05,
+      "loss": 0.7579,
+      "step": 2622
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 6.013874066168623e-05,
+      "loss": 0.7539,
+      "step": 2623
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 6.008537886872999e-05,
+      "loss": 0.6856,
+      "step": 2624
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 6.003201707577375e-05,
+      "loss": 0.8288,
+      "step": 2625
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 5.997865528281751e-05,
+      "loss": 0.7713,
+      "step": 2626
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 5.9925293489861266e-05,
+      "loss": 0.5961,
+      "step": 2627
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 5.9871931696905014e-05,
+      "loss": 0.6973,
+      "step": 2628
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 5.9818569903948776e-05,
+      "loss": 0.7862,
+      "step": 2629
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 5.976520811099253e-05,
+      "loss": 0.7008,
+      "step": 2630
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 5.971184631803629e-05,
+      "loss": 0.3675,
+      "step": 2631
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 5.965848452508005e-05,
+      "loss": 0.6427,
+      "step": 2632
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 5.96051227321238e-05,
+      "loss": 0.6241,
+      "step": 2633
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 5.955176093916756e-05,
+      "loss": 0.6737,
+      "step": 2634
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 5.9498399146211314e-05,
+      "loss": 0.7907,
+      "step": 2635
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 5.9445037353255076e-05,
+      "loss": 0.7699,
+      "step": 2636
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 5.939167556029883e-05,
+      "loss": 0.8032,
+      "step": 2637
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 5.933831376734259e-05,
+      "loss": 0.7918,
+      "step": 2638
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 5.9284951974386335e-05,
+      "loss": 0.6912,
+      "step": 2639
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 5.92315901814301e-05,
+      "loss": 0.7133,
+      "step": 2640
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 5.917822838847386e-05,
+      "loss": 0.2728,
+      "step": 2641
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 5.9124866595517614e-05,
+      "loss": 0.7552,
+      "step": 2642
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 5.9071504802561376e-05,
+      "loss": 0.3421,
+      "step": 2643
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 5.901814300960512e-05,
+      "loss": 0.7596,
+      "step": 2644
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 5.896478121664888e-05,
+      "loss": 0.5969,
+      "step": 2645
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 5.8911419423692635e-05,
+      "loss": 0.7908,
+      "step": 2646
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 5.8858057630736397e-05,
+      "loss": 0.8913,
+      "step": 2647
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 5.880469583778015e-05,
+      "loss": 0.7513,
+      "step": 2648
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 5.8751334044823914e-05,
+      "loss": 0.6387,
+      "step": 2649
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 5.869797225186766e-05,
+      "loss": 0.6462,
+      "step": 2650
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 5.864461045891142e-05,
+      "loss": 0.3905,
+      "step": 2651
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 5.859124866595518e-05,
+      "loss": 0.7633,
+      "step": 2652
+    },
+    {
+      "epoch": 2.12,
+      "eval_loss": 1.0448030233383179,
+      "eval_runtime": 7634.0996,
+      "eval_samples_per_second": 1.31,
+      "eval_steps_per_second": 0.082,
+      "step": 2652
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 5.8537886872998934e-05,
+      "loss": 0.6811,
+      "step": 2653
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 5.8484525080042696e-05,
+      "loss": 0.3842,
+      "step": 2654
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 5.8431163287086445e-05,
+      "loss": 0.6698,
+      "step": 2655
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 5.83778014941302e-05,
+      "loss": 0.7143,
+      "step": 2656
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 5.832443970117396e-05,
+      "loss": 0.4214,
+      "step": 2657
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 5.827107790821772e-05,
+      "loss": 0.8207,
+      "step": 2658
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 5.821771611526148e-05,
+      "loss": 0.7071,
+      "step": 2659
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 5.8164354322305234e-05,
+      "loss": 0.6817,
+      "step": 2660
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 5.811099252934898e-05,
+      "loss": 0.3712,
+      "step": 2661
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 5.8057630736392745e-05,
+      "loss": 0.8466,
+      "step": 2662
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 5.80042689434365e-05,
+      "loss": 0.7719,
+      "step": 2663
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 5.795090715048026e-05,
+      "loss": 0.6462,
+      "step": 2664
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 5.789754535752402e-05,
+      "loss": 0.7187,
+      "step": 2665
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 5.784418356456778e-05,
+      "loss": 0.7026,
+      "step": 2666
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 5.779082177161153e-05,
+      "loss": 0.7027,
+      "step": 2667
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 5.773745997865528e-05,
+      "loss": 0.3115,
+      "step": 2668
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 5.7684098185699045e-05,
+      "loss": 0.6877,
+      "step": 2669
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 5.76307363927428e-05,
+      "loss": 0.6525,
+      "step": 2670
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 5.757737459978656e-05,
+      "loss": 0.7568,
+      "step": 2671
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 5.752401280683031e-05,
+      "loss": 0.6503,
+      "step": 2672
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 5.7470651013874065e-05,
+      "loss": 0.6575,
+      "step": 2673
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 5.741728922091783e-05,
+      "loss": 0.7704,
+      "step": 2674
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 5.736392742796158e-05,
+      "loss": 0.7523,
+      "step": 2675
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 5.7310565635005345e-05,
+      "loss": 0.6738,
+      "step": 2676
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 5.72572038420491e-05,
+      "loss": 0.8537,
+      "step": 2677
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 5.720384204909285e-05,
+      "loss": 0.6891,
+      "step": 2678
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 5.715048025613661e-05,
+      "loss": 0.7803,
+      "step": 2679
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 5.7097118463180365e-05,
+      "loss": 0.7379,
+      "step": 2680
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 5.704375667022413e-05,
+      "loss": 0.3222,
+      "step": 2681
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 5.699039487726788e-05,
+      "loss": 0.7507,
+      "step": 2682
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 5.693703308431163e-05,
+      "loss": 0.7468,
+      "step": 2683
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 5.6883671291355386e-05,
+      "loss": 0.8258,
+      "step": 2684
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 5.683030949839915e-05,
+      "loss": 0.359,
+      "step": 2685
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 5.677694770544291e-05,
+      "loss": 0.6944,
+      "step": 2686
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 5.6723585912486665e-05,
+      "loss": 0.314,
+      "step": 2687
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 5.667022411953043e-05,
+      "loss": 0.6539,
+      "step": 2688
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 5.661686232657417e-05,
+      "loss": 0.6013,
+      "step": 2689
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 5.656350053361793e-05,
+      "loss": 0.6861,
+      "step": 2690
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 5.6510138740661686e-05,
+      "loss": 0.3179,
+      "step": 2691
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 5.645677694770545e-05,
+      "loss": 0.7027,
+      "step": 2692
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 5.64034151547492e-05,
+      "loss": 0.3793,
+      "step": 2693
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 5.635005336179295e-05,
+      "loss": 0.6823,
+      "step": 2694
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 5.6296691568836714e-05,
+      "loss": 0.716,
+      "step": 2695
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 5.624332977588047e-05,
+      "loss": 0.6971,
+      "step": 2696
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 5.618996798292423e-05,
+      "loss": 0.6235,
+      "step": 2697
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 5.6136606189967986e-05,
+      "loss": 0.7483,
+      "step": 2698
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 5.608324439701175e-05,
+      "loss": 0.6631,
+      "step": 2699
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 5.6029882604055496e-05,
+      "loss": 0.7947,
+      "step": 2700
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 5.597652081109925e-05,
+      "loss": 0.2721,
+      "step": 2701
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 5.5923159018143013e-05,
+      "loss": 0.0,
+      "step": 2702
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 5.586979722518677e-05,
+      "loss": 0.7122,
+      "step": 2703
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 5.581643543223053e-05,
+      "loss": 0.8608,
+      "step": 2704
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 5.576307363927428e-05,
+      "loss": 0.7835,
+      "step": 2705
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 5.5709711846318034e-05,
+      "loss": 0.6312,
+      "step": 2706
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 5.5656350053361796e-05,
+      "loss": 0.7897,
+      "step": 2707
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 5.560298826040555e-05,
+      "loss": 0.4181,
+      "step": 2708
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 5.554962646744931e-05,
+      "loss": 0.7735,
+      "step": 2709
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 5.549626467449307e-05,
+      "loss": 0.6815,
+      "step": 2710
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 5.544290288153682e-05,
+      "loss": 0.8068,
+      "step": 2711
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 5.538954108858058e-05,
+      "loss": 0.6069,
+      "step": 2712
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 5.5336179295624334e-05,
+      "loss": 0.8352,
+      "step": 2713
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 5.5282817502668096e-05,
+      "loss": 0.5886,
+      "step": 2714
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 5.522945570971185e-05,
+      "loss": 0.762,
+      "step": 2715
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 5.51760939167556e-05,
+      "loss": 0.6404,
+      "step": 2716
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 5.512273212379936e-05,
+      "loss": 0.7635,
+      "step": 2717
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 5.506937033084312e-05,
+      "loss": 0.8292,
+      "step": 2718
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 5.501600853788688e-05,
+      "loss": 0.5925,
+      "step": 2719
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 5.4962646744930634e-05,
+      "loss": 0.8053,
+      "step": 2720
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 5.4909284951974396e-05,
+      "loss": 0.7777,
+      "step": 2721
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 5.4855923159018144e-05,
+      "loss": 0.3901,
+      "step": 2722
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 5.48025613660619e-05,
+      "loss": 0.6993,
+      "step": 2723
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 5.474919957310566e-05,
+      "loss": 0.5119,
+      "step": 2724
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 5.469583778014942e-05,
+      "loss": 0.7539,
+      "step": 2725
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 5.464247598719318e-05,
+      "loss": 0.6342,
+      "step": 2726
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 5.4589114194236934e-05,
+      "loss": 0.8731,
+      "step": 2727
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 5.453575240128068e-05,
+      "loss": 0.7249,
+      "step": 2728
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 5.448239060832444e-05,
+      "loss": 0.7306,
+      "step": 2729
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 5.44290288153682e-05,
+      "loss": 0.7591,
+      "step": 2730
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 5.4375667022411955e-05,
+      "loss": 0.6461,
+      "step": 2731
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 5.432230522945572e-05,
+      "loss": 0.6698,
+      "step": 2732
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 5.4268943436499465e-05,
+      "loss": 0.7248,
+      "step": 2733
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 5.421558164354322e-05,
+      "loss": 0.3429,
+      "step": 2734
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 5.416221985058698e-05,
+      "loss": 0.6979,
+      "step": 2735
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 5.410885805763074e-05,
+      "loss": 0.7332,
+      "step": 2736
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 5.40554962646745e-05,
+      "loss": 0.3716,
+      "step": 2737
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 5.4002134471718255e-05,
+      "loss": 0.7527,
+      "step": 2738
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 5.3948772678762e-05,
+      "loss": 0.8209,
+      "step": 2739
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 5.3895410885805765e-05,
+      "loss": 0.7745,
+      "step": 2740
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 5.384204909284952e-05,
+      "loss": 0.7303,
+      "step": 2741
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 5.378868729989328e-05,
+      "loss": 0.7138,
+      "step": 2742
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 5.373532550693704e-05,
+      "loss": 0.3966,
+      "step": 2743
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 5.3681963713980786e-05,
+      "loss": 0.4295,
+      "step": 2744
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 5.362860192102455e-05,
+      "loss": 0.3635,
+      "step": 2745
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 5.35752401280683e-05,
+      "loss": 0.623,
+      "step": 2746
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 5.3521878335112065e-05,
+      "loss": 0.7299,
+      "step": 2747
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 5.346851654215582e-05,
+      "loss": 0.6206,
+      "step": 2748
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 5.341515474919958e-05,
+      "loss": 0.2775,
+      "step": 2749
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 5.336179295624333e-05,
+      "loss": 0.7344,
+      "step": 2750
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 5.3308431163287086e-05,
+      "loss": 0.734,
+      "step": 2751
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 5.325506937033085e-05,
+      "loss": 0.3783,
+      "step": 2752
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 5.32017075773746e-05,
+      "loss": 0.3447,
+      "step": 2753
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 5.3148345784418365e-05,
+      "loss": 0.3263,
+      "step": 2754
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 5.309498399146211e-05,
+      "loss": 0.4336,
+      "step": 2755
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 5.304162219850587e-05,
+      "loss": 0.6649,
+      "step": 2756
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 5.298826040554963e-05,
+      "loss": 0.6814,
+      "step": 2757
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 5.2934898612593386e-05,
+      "loss": 0.3721,
+      "step": 2758
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 5.288153681963715e-05,
+      "loss": 0.2821,
+      "step": 2759
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 5.28281750266809e-05,
+      "loss": 0.7883,
+      "step": 2760
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 5.277481323372465e-05,
+      "loss": 0.6036,
+      "step": 2761
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 5.272145144076841e-05,
+      "loss": 0.7523,
+      "step": 2762
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 5.266808964781217e-05,
+      "loss": 0.6207,
+      "step": 2763
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 5.261472785485593e-05,
+      "loss": 0.7795,
+      "step": 2764
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 5.2561366061899685e-05,
+      "loss": 0.3925,
+      "step": 2765
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 5.2508004268943434e-05,
+      "loss": 0.3165,
+      "step": 2766
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 5.245464247598719e-05,
+      "loss": 0.4065,
+      "step": 2767
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 5.240128068303095e-05,
+      "loss": 0.7459,
+      "step": 2768
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 5.234791889007471e-05,
+      "loss": 0.7709,
+      "step": 2769
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 5.229455709711847e-05,
+      "loss": 0.6922,
+      "step": 2770
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 5.224119530416223e-05,
+      "loss": 0.6852,
+      "step": 2771
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 5.218783351120597e-05,
+      "loss": 0.6856,
+      "step": 2772
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 5.2134471718249734e-05,
+      "loss": 0.8285,
+      "step": 2773
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 5.208110992529349e-05,
+      "loss": 0.8501,
+      "step": 2774
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 5.202774813233725e-05,
+      "loss": 0.6971,
+      "step": 2775
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 5.1974386339381006e-05,
+      "loss": 0.7158,
+      "step": 2776
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 5.1921024546424754e-05,
+      "loss": 0.7485,
+      "step": 2777
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 5.1867662753468516e-05,
+      "loss": 0.7647,
+      "step": 2778
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 5.181430096051227e-05,
+      "loss": 0.7758,
+      "step": 2779
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 5.1760939167556034e-05,
+      "loss": 0.8332,
+      "step": 2780
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 5.170757737459979e-05,
+      "loss": 0.7001,
+      "step": 2781
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 5.165421558164355e-05,
+      "loss": 0.0,
+      "step": 2782
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 5.16008537886873e-05,
+      "loss": 0.6908,
+      "step": 2783
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 5.1547491995731054e-05,
+      "loss": 0.7112,
+      "step": 2784
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 5.1494130202774816e-05,
+      "loss": 0.737,
+      "step": 2785
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 5.144076840981857e-05,
+      "loss": 0.6293,
+      "step": 2786
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 5.1387406616862334e-05,
+      "loss": 0.3223,
+      "step": 2787
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 5.133404482390609e-05,
+      "loss": 0.8353,
+      "step": 2788
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 5.128068303094984e-05,
+      "loss": 0.634,
+      "step": 2789
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 5.12273212379936e-05,
+      "loss": 0.3204,
+      "step": 2790
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 5.1173959445037354e-05,
+      "loss": 0.2451,
+      "step": 2791
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 5.1120597652081116e-05,
+      "loss": 0.7066,
+      "step": 2792
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 5.106723585912487e-05,
+      "loss": 0.6909,
+      "step": 2793
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 5.101387406616862e-05,
+      "loss": 0.6627,
+      "step": 2794
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 5.096051227321238e-05,
+      "loss": 0.6134,
+      "step": 2795
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 5.090715048025614e-05,
+      "loss": 0.5597,
+      "step": 2796
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 5.08537886872999e-05,
+      "loss": 0.7434,
+      "step": 2797
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 5.0800426894343654e-05,
+      "loss": 0.7816,
+      "step": 2798
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 5.0747065101387416e-05,
+      "loss": 0.6875,
+      "step": 2799
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 5.0693703308431165e-05,
+      "loss": 0.6686,
+      "step": 2800
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 5.064034151547492e-05,
+      "loss": 0.3226,
+      "step": 2801
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 5.058697972251868e-05,
+      "loss": 0.7105,
+      "step": 2802
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 5.053361792956244e-05,
+      "loss": 0.6806,
+      "step": 2803
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 5.04802561366062e-05,
+      "loss": 0.7617,
+      "step": 2804
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 5.042689434364995e-05,
+      "loss": 0.8228,
+      "step": 2805
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 5.03735325506937e-05,
+      "loss": 0.3095,
+      "step": 2806
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 5.0320170757737464e-05,
+      "loss": 0.6831,
+      "step": 2807
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 5.026680896478122e-05,
+      "loss": 0.7197,
+      "step": 2808
+    },
+    {
+      "epoch": 2.25,
+      "eval_loss": 1.0402891635894775,
+      "eval_runtime": 7473.8062,
+      "eval_samples_per_second": 1.338,
+      "eval_steps_per_second": 0.084,
+      "step": 2808
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 5.021344717182498e-05,
+      "loss": 0.338,
+      "step": 2809
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 5.016008537886874e-05,
+      "loss": 0.5579,
+      "step": 2810
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 5.0106723585912485e-05,
+      "loss": 0.5297,
+      "step": 2811
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 5.005336179295624e-05,
+      "loss": 0.7387,
+      "step": 2812
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 5e-05,
+      "loss": 0.3931,
+      "step": 2813
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 4.9946638207043764e-05,
+      "loss": 0.3221,
+      "step": 2814
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 4.989327641408751e-05,
+      "loss": 0.3504,
+      "step": 2815
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 4.9839914621131275e-05,
+      "loss": 0.416,
+      "step": 2816
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 4.978655282817503e-05,
+      "loss": 0.6261,
+      "step": 2817
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 4.9733191035218785e-05,
+      "loss": 0.6962,
+      "step": 2818
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 4.967982924226254e-05,
+      "loss": 0.7801,
+      "step": 2819
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 4.9626467449306296e-05,
+      "loss": 0.3704,
+      "step": 2820
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 4.957310565635006e-05,
+      "loss": 0.8298,
+      "step": 2821
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 4.951974386339381e-05,
+      "loss": 0.7485,
+      "step": 2822
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 4.946638207043757e-05,
+      "loss": 0.3208,
+      "step": 2823
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 4.941302027748132e-05,
+      "loss": 0.7395,
+      "step": 2824
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 4.9359658484525085e-05,
+      "loss": 0.6553,
+      "step": 2825
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 4.930629669156884e-05,
+      "loss": 0.6292,
+      "step": 2826
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 4.9252934898612595e-05,
+      "loss": 0.7014,
+      "step": 2827
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 4.919957310565636e-05,
+      "loss": 0.7689,
+      "step": 2828
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 4.9146211312700106e-05,
+      "loss": 0.8,
+      "step": 2829
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 4.909284951974387e-05,
+      "loss": 0.8588,
+      "step": 2830
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 4.903948772678762e-05,
+      "loss": 0.3236,
+      "step": 2831
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 4.898612593383138e-05,
+      "loss": 0.9383,
+      "step": 2832
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 4.893276414087514e-05,
+      "loss": 0.6979,
+      "step": 2833
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 4.887940234791889e-05,
+      "loss": 0.7886,
+      "step": 2834
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 4.882604055496265e-05,
+      "loss": 0.7111,
+      "step": 2835
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 4.8772678762006406e-05,
+      "loss": 0.7407,
+      "step": 2836
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 4.871931696905016e-05,
+      "loss": 0.6538,
+      "step": 2837
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 4.8665955176093916e-05,
+      "loss": 0.6985,
+      "step": 2838
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 4.861259338313768e-05,
+      "loss": 0.6636,
+      "step": 2839
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 4.855923159018143e-05,
+      "loss": 0.6528,
+      "step": 2840
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 4.850586979722519e-05,
+      "loss": 0.7047,
+      "step": 2841
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 4.845250800426895e-05,
+      "loss": 0.7031,
+      "step": 2842
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 4.83991462113127e-05,
+      "loss": 0.7632,
+      "step": 2843
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 4.834578441835646e-05,
+      "loss": 0.6714,
+      "step": 2844
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 4.8292422625400216e-05,
+      "loss": 0.9021,
+      "step": 2845
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 4.823906083244397e-05,
+      "loss": 0.8299,
+      "step": 2846
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 4.818569903948773e-05,
+      "loss": 0.7984,
+      "step": 2847
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 4.813233724653148e-05,
+      "loss": 0.7445,
+      "step": 2848
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 4.8078975453575244e-05,
+      "loss": 0.6537,
+      "step": 2849
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 4.8025613660619e-05,
+      "loss": 0.6625,
+      "step": 2850
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 4.7972251867662754e-05,
+      "loss": 0.7776,
+      "step": 2851
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 4.7918890074706516e-05,
+      "loss": 0.4115,
+      "step": 2852
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 4.786552828175027e-05,
+      "loss": 0.5406,
+      "step": 2853
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 4.7812166488794026e-05,
+      "loss": 0.3069,
+      "step": 2854
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 4.775880469583778e-05,
+      "loss": 0.4207,
+      "step": 2855
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 4.770544290288154e-05,
+      "loss": 0.4339,
+      "step": 2856
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 4.765208110992529e-05,
+      "loss": 0.3492,
+      "step": 2857
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 4.7598719316969054e-05,
+      "loss": 0.6326,
+      "step": 2858
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 4.754535752401281e-05,
+      "loss": 0.7737,
+      "step": 2859
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 4.7491995731056564e-05,
+      "loss": 0.7161,
+      "step": 2860
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 4.7438633938100326e-05,
+      "loss": 0.7516,
+      "step": 2861
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 4.7385272145144075e-05,
+      "loss": 0.3745,
+      "step": 2862
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 4.7331910352187837e-05,
+      "loss": 0.5986,
+      "step": 2863
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 4.727854855923159e-05,
+      "loss": 0.6629,
+      "step": 2864
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 4.722518676627535e-05,
+      "loss": 0.736,
+      "step": 2865
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 4.717182497331911e-05,
+      "loss": 0.3051,
+      "step": 2866
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 4.7118463180362864e-05,
+      "loss": 0.6888,
+      "step": 2867
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 4.706510138740662e-05,
+      "loss": 0.4206,
+      "step": 2868
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 4.7011739594450374e-05,
+      "loss": 0.7435,
+      "step": 2869
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 4.695837780149413e-05,
+      "loss": 0.5642,
+      "step": 2870
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 4.690501600853789e-05,
+      "loss": 0.7633,
+      "step": 2871
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 4.685165421558165e-05,
+      "loss": 0.8764,
+      "step": 2872
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 4.67982924226254e-05,
+      "loss": 0.7304,
+      "step": 2873
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 4.674493062966916e-05,
+      "loss": 0.3153,
+      "step": 2874
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 4.669156883671292e-05,
+      "loss": 0.6415,
+      "step": 2875
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 4.663820704375667e-05,
+      "loss": 0.626,
+      "step": 2876
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 4.658484525080043e-05,
+      "loss": 0.7553,
+      "step": 2877
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 4.653148345784419e-05,
+      "loss": 0.7157,
+      "step": 2878
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 4.647812166488794e-05,
+      "loss": 0.3016,
+      "step": 2879
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 4.64247598719317e-05,
+      "loss": 0.715,
+      "step": 2880
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 4.637139807897545e-05,
+      "loss": 0.311,
+      "step": 2881
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 4.631803628601921e-05,
+      "loss": 0.7646,
+      "step": 2882
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 4.626467449306297e-05,
+      "loss": 0.6794,
+      "step": 2883
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 4.621131270010672e-05,
+      "loss": 0.6863,
+      "step": 2884
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 4.6157950907150485e-05,
+      "loss": 0.7035,
+      "step": 2885
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 4.610458911419424e-05,
+      "loss": 0.7489,
+      "step": 2886
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 4.6051227321237995e-05,
+      "loss": 0.299,
+      "step": 2887
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 4.599786552828175e-05,
+      "loss": 0.6809,
+      "step": 2888
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 4.594450373532551e-05,
+      "loss": 0.7955,
+      "step": 2889
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 4.589114194236927e-05,
+      "loss": 0.7039,
+      "step": 2890
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 4.583778014941302e-05,
+      "loss": 0.6945,
+      "step": 2891
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 4.5784418356456785e-05,
+      "loss": 0.8682,
+      "step": 2892
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 4.573105656350053e-05,
+      "loss": 0.7333,
+      "step": 2893
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 4.5677694770544295e-05,
+      "loss": 0.7378,
+      "step": 2894
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 4.562433297758805e-05,
+      "loss": 0.3624,
+      "step": 2895
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 4.5570971184631805e-05,
+      "loss": 0.7158,
+      "step": 2896
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 4.551760939167557e-05,
+      "loss": 0.738,
+      "step": 2897
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 4.5464247598719316e-05,
+      "loss": 0.6728,
+      "step": 2898
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 4.541088580576308e-05,
+      "loss": 0.6849,
+      "step": 2899
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 4.535752401280683e-05,
+      "loss": 0.7409,
+      "step": 2900
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 4.530416221985059e-05,
+      "loss": 0.6189,
+      "step": 2901
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 4.525080042689434e-05,
+      "loss": 0.6083,
+      "step": 2902
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 4.5197438633938105e-05,
+      "loss": 0.6771,
+      "step": 2903
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 4.514407684098186e-05,
+      "loss": 0.6542,
+      "step": 2904
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 4.5090715048025616e-05,
+      "loss": 0.6825,
+      "step": 2905
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 4.503735325506937e-05,
+      "loss": 0.3996,
+      "step": 2906
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 4.4983991462113126e-05,
+      "loss": 0.7704,
+      "step": 2907
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 4.493062966915689e-05,
+      "loss": 0.8847,
+      "step": 2908
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 4.487726787620064e-05,
+      "loss": 0.7916,
+      "step": 2909
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 4.48239060832444e-05,
+      "loss": 0.6742,
+      "step": 2910
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 4.477054429028816e-05,
+      "loss": 0.6915,
+      "step": 2911
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 4.471718249733191e-05,
+      "loss": 0.3389,
+      "step": 2912
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 4.466382070437567e-05,
+      "loss": 0.7103,
+      "step": 2913
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 4.4610458911419426e-05,
+      "loss": 0.3754,
+      "step": 2914
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 4.455709711846318e-05,
+      "loss": 0.7145,
+      "step": 2915
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 4.450373532550694e-05,
+      "loss": 0.3026,
+      "step": 2916
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 4.445037353255069e-05,
+      "loss": 0.7063,
+      "step": 2917
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 4.4397011739594453e-05,
+      "loss": 0.6861,
+      "step": 2918
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 4.434364994663821e-05,
+      "loss": 0.6596,
+      "step": 2919
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 4.4290288153681964e-05,
+      "loss": 0.3285,
+      "step": 2920
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 4.423692636072572e-05,
+      "loss": 0.2966,
+      "step": 2921
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 4.418356456776948e-05,
+      "loss": 0.4296,
+      "step": 2922
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 4.4130202774813236e-05,
+      "loss": 0.6378,
+      "step": 2923
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 4.407684098185699e-05,
+      "loss": 0.6738,
+      "step": 2924
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 4.402347918890075e-05,
+      "loss": 0.6204,
+      "step": 2925
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 4.39701173959445e-05,
+      "loss": 0.6662,
+      "step": 2926
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 4.3916755602988264e-05,
+      "loss": 0.694,
+      "step": 2927
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 4.386339381003202e-05,
+      "loss": 0.6491,
+      "step": 2928
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 4.3810032017075774e-05,
+      "loss": 0.6124,
+      "step": 2929
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 4.3756670224119536e-05,
+      "loss": 0.3704,
+      "step": 2930
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 4.3703308431163285e-05,
+      "loss": 0.5942,
+      "step": 2931
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 4.3649946638207046e-05,
+      "loss": 0.7652,
+      "step": 2932
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 4.35965848452508e-05,
+      "loss": 0.7172,
+      "step": 2933
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 4.354322305229456e-05,
+      "loss": 0.2944,
+      "step": 2934
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 4.348986125933832e-05,
+      "loss": 0.7195,
+      "step": 2935
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 4.3436499466382074e-05,
+      "loss": 0.7701,
+      "step": 2936
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 4.338313767342583e-05,
+      "loss": 0.3774,
+      "step": 2937
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 4.3329775880469584e-05,
+      "loss": 0.7203,
+      "step": 2938
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 4.3276414087513346e-05,
+      "loss": 0.7499,
+      "step": 2939
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 4.3223052294557095e-05,
+      "loss": 0.3487,
+      "step": 2940
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 4.316969050160086e-05,
+      "loss": 0.7164,
+      "step": 2941
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 4.311632870864461e-05,
+      "loss": 0.7582,
+      "step": 2942
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 4.306296691568837e-05,
+      "loss": 0.6349,
+      "step": 2943
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 4.300960512273213e-05,
+      "loss": 0.7047,
+      "step": 2944
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 4.295624332977588e-05,
+      "loss": 0.8539,
+      "step": 2945
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 4.290288153681964e-05,
+      "loss": 0.3592,
+      "step": 2946
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 4.2849519743863395e-05,
+      "loss": 0.8845,
+      "step": 2947
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 4.279615795090715e-05,
+      "loss": 0.6823,
+      "step": 2948
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 4.274279615795091e-05,
+      "loss": 0.7028,
+      "step": 2949
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 4.268943436499467e-05,
+      "loss": 0.6444,
+      "step": 2950
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 4.263607257203842e-05,
+      "loss": 0.6822,
+      "step": 2951
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 4.258271077908218e-05,
+      "loss": 0.5153,
+      "step": 2952
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 4.252934898612593e-05,
+      "loss": 0.7033,
+      "step": 2953
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 4.2475987193169695e-05,
+      "loss": 0.3087,
+      "step": 2954
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 4.242262540021345e-05,
+      "loss": 0.7965,
+      "step": 2955
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 4.2369263607257205e-05,
+      "loss": 0.7858,
+      "step": 2956
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 4.231590181430096e-05,
+      "loss": 0.3856,
+      "step": 2957
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 4.226254002134472e-05,
+      "loss": 0.8,
+      "step": 2958
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 4.220917822838848e-05,
+      "loss": 0.64,
+      "step": 2959
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 4.215581643543223e-05,
+      "loss": 0.6325,
+      "step": 2960
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 4.2102454642475994e-05,
+      "loss": 0.6851,
+      "step": 2961
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 4.204909284951974e-05,
+      "loss": 0.4096,
+      "step": 2962
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 4.1995731056563505e-05,
+      "loss": 0.7965,
+      "step": 2963
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 4.194236926360726e-05,
+      "loss": 0.7427,
+      "step": 2964
+    },
+    {
+      "epoch": 2.37,
+      "eval_loss": 1.0400638580322266,
+      "eval_runtime": 7419.9709,
+      "eval_samples_per_second": 1.348,
+      "eval_steps_per_second": 0.084,
+      "step": 2964
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 4.1889007470651015e-05,
+      "loss": 0.7943,
+      "step": 2965
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 4.183564567769477e-05,
+      "loss": 0.5665,
+      "step": 2966
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 4.1782283884738526e-05,
+      "loss": 0.6724,
+      "step": 2967
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 4.172892209178229e-05,
+      "loss": 0.8535,
+      "step": 2968
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 4.167556029882604e-05,
+      "loss": 0.7551,
+      "step": 2969
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 4.16221985058698e-05,
+      "loss": 0.7539,
+      "step": 2970
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 4.156883671291355e-05,
+      "loss": 0.8002,
+      "step": 2971
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 4.1515474919957315e-05,
+      "loss": 0.7543,
+      "step": 2972
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 4.146211312700107e-05,
+      "loss": 0.7202,
+      "step": 2973
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 4.1408751334044826e-05,
+      "loss": 0.722,
+      "step": 2974
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 4.135538954108859e-05,
+      "loss": 0.6433,
+      "step": 2975
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 4.1302027748132336e-05,
+      "loss": 0.7383,
+      "step": 2976
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 4.12486659551761e-05,
+      "loss": 0.6054,
+      "step": 2977
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 4.119530416221985e-05,
+      "loss": 0.8282,
+      "step": 2978
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 4.114194236926361e-05,
+      "loss": 0.6689,
+      "step": 2979
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 4.108858057630737e-05,
+      "loss": 0.7077,
+      "step": 2980
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 4.103521878335112e-05,
+      "loss": 0.7636,
+      "step": 2981
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 4.098185699039488e-05,
+      "loss": 0.7586,
+      "step": 2982
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 4.0928495197438636e-05,
+      "loss": 0.6184,
+      "step": 2983
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 4.087513340448239e-05,
+      "loss": 0.8302,
+      "step": 2984
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 4.0821771611526146e-05,
+      "loss": 0.5703,
+      "step": 2985
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 4.076840981856991e-05,
+      "loss": 0.6709,
+      "step": 2986
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 4.071504802561366e-05,
+      "loss": 0.6082,
+      "step": 2987
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 4.066168623265742e-05,
+      "loss": 0.5998,
+      "step": 2988
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 4.060832443970118e-05,
+      "loss": 0.8156,
+      "step": 2989
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 4.055496264674493e-05,
+      "loss": 0.3311,
+      "step": 2990
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 4.050160085378869e-05,
+      "loss": 0.7383,
+      "step": 2991
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 4.0448239060832446e-05,
+      "loss": 0.8319,
+      "step": 2992
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 4.03948772678762e-05,
+      "loss": 0.5625,
+      "step": 2993
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 4.034151547491996e-05,
+      "loss": 0.6317,
+      "step": 2994
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 4.028815368196371e-05,
+      "loss": 0.7227,
+      "step": 2995
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 4.0234791889007474e-05,
+      "loss": 0.2539,
+      "step": 2996
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 4.018143009605123e-05,
+      "loss": 0.5989,
+      "step": 2997
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 4.0128068303094984e-05,
+      "loss": 0.7708,
+      "step": 2998
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 4.0074706510138746e-05,
+      "loss": 0.6659,
+      "step": 2999
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 4.00213447171825e-05,
+      "loss": 0.7524,
+      "step": 3000
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 3.9967982924226256e-05,
+      "loss": 0.7499,
+      "step": 3001
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 3.991462113127001e-05,
+      "loss": 0.6546,
+      "step": 3002
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 3.986125933831377e-05,
+      "loss": 0.6961,
+      "step": 3003
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 3.980789754535753e-05,
+      "loss": 0.0,
+      "step": 3004
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 3.9754535752401284e-05,
+      "loss": 0.6439,
+      "step": 3005
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 3.970117395944504e-05,
+      "loss": 0.6192,
+      "step": 3006
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 3.9647812166488794e-05,
+      "loss": 0.6668,
+      "step": 3007
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 3.9594450373532556e-05,
+      "loss": 0.613,
+      "step": 3008
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 3.9541088580576305e-05,
+      "loss": 0.6818,
+      "step": 3009
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 3.948772678762007e-05,
+      "loss": 0.4264,
+      "step": 3010
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 3.943436499466382e-05,
+      "loss": 0.6647,
+      "step": 3011
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 3.938100320170758e-05,
+      "loss": 0.6296,
+      "step": 3012
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 3.932764140875134e-05,
+      "loss": 0.6622,
+      "step": 3013
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 3.927427961579509e-05,
+      "loss": 0.7275,
+      "step": 3014
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 3.922091782283885e-05,
+      "loss": 0.701,
+      "step": 3015
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 3.9167556029882605e-05,
+      "loss": 0.0,
+      "step": 3016
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 3.911419423692636e-05,
+      "loss": 0.6942,
+      "step": 3017
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 3.906083244397012e-05,
+      "loss": 0.6335,
+      "step": 3018
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 3.900747065101388e-05,
+      "loss": 0.3818,
+      "step": 3019
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 3.895410885805763e-05,
+      "loss": 0.7013,
+      "step": 3020
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 3.890074706510139e-05,
+      "loss": 0.7377,
+      "step": 3021
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 3.884738527214515e-05,
+      "loss": 0.6648,
+      "step": 3022
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 3.8794023479188905e-05,
+      "loss": 0.6943,
+      "step": 3023
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 3.874066168623266e-05,
+      "loss": 0.7697,
+      "step": 3024
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 3.868729989327642e-05,
+      "loss": 0.7245,
+      "step": 3025
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 3.863393810032017e-05,
+      "loss": 0.4363,
+      "step": 3026
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 3.858057630736393e-05,
+      "loss": 0.4078,
+      "step": 3027
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 3.852721451440768e-05,
+      "loss": 0.7465,
+      "step": 3028
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 3.847385272145144e-05,
+      "loss": 0.7629,
+      "step": 3029
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 3.84204909284952e-05,
+      "loss": 0.7638,
+      "step": 3030
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 3.836712913553895e-05,
+      "loss": 0.6906,
+      "step": 3031
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 3.8313767342582715e-05,
+      "loss": 0.7081,
+      "step": 3032
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 3.826040554962647e-05,
+      "loss": 0.6609,
+      "step": 3033
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 3.8207043756670225e-05,
+      "loss": 0.6913,
+      "step": 3034
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 3.815368196371398e-05,
+      "loss": 0.3176,
+      "step": 3035
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 3.810032017075774e-05,
+      "loss": 0.7011,
+      "step": 3036
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 3.80469583778015e-05,
+      "loss": 0.7123,
+      "step": 3037
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 3.799359658484525e-05,
+      "loss": 0.7807,
+      "step": 3038
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 3.794023479188901e-05,
+      "loss": 0.753,
+      "step": 3039
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 3.788687299893276e-05,
+      "loss": 0.2806,
+      "step": 3040
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 3.7833511205976525e-05,
+      "loss": 0.6558,
+      "step": 3041
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 3.778014941302028e-05,
+      "loss": 0.7746,
+      "step": 3042
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 3.7726787620064035e-05,
+      "loss": 0.7093,
+      "step": 3043
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 3.76734258271078e-05,
+      "loss": 0.7538,
+      "step": 3044
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 3.7620064034151546e-05,
+      "loss": 0.6202,
+      "step": 3045
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 3.756670224119531e-05,
+      "loss": 0.4283,
+      "step": 3046
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 3.751334044823906e-05,
+      "loss": 0.815,
+      "step": 3047
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 3.745997865528282e-05,
+      "loss": 0.3636,
+      "step": 3048
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 3.740661686232657e-05,
+      "loss": 0.2534,
+      "step": 3049
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 3.7353255069370335e-05,
+      "loss": 0.9261,
+      "step": 3050
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 3.729989327641409e-05,
+      "loss": 0.6691,
+      "step": 3051
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 3.7246531483457846e-05,
+      "loss": 0.5998,
+      "step": 3052
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 3.71931696905016e-05,
+      "loss": 0.5889,
+      "step": 3053
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 3.7139807897545356e-05,
+      "loss": 0.6453,
+      "step": 3054
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 3.708644610458912e-05,
+      "loss": 0.7464,
+      "step": 3055
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 3.703308431163287e-05,
+      "loss": 0.6749,
+      "step": 3056
+    },
+    {
+      "epoch": 2.45,
+      "learning_rate": 3.697972251867663e-05,
+      "loss": 0.7842,
+      "step": 3057
+    },
+    {
+      "epoch": 2.45,
+      "learning_rate": 3.692636072572039e-05,
+      "loss": 0.6728,
+      "step": 3058
+    },
+    {
+      "epoch": 2.45,
+      "learning_rate": 3.687299893276414e-05,
+      "loss": 0.7383,
+      "step": 3059
+    },
+    {
+      "epoch": 2.45,
+      "learning_rate": 3.68196371398079e-05,
+      "loss": 0.733,
+      "step": 3060
+    },
+    {
+      "epoch": 2.45,
+      "learning_rate": 3.6766275346851656e-05,
+      "loss": 0.3278,
+      "step": 3061
+    },
+    {
+      "epoch": 2.45,
+      "learning_rate": 3.671291355389541e-05,
+      "loss": 0.6733,
+      "step": 3062
+    },
+    {
+      "epoch": 2.45,
+      "learning_rate": 3.665955176093917e-05,
+      "loss": 0.6965,
+      "step": 3063
+    },
+    {
+      "epoch": 2.45,
+      "learning_rate": 3.660618996798292e-05,
+      "loss": 0.663,
+      "step": 3064
+    },
+    {
+      "epoch": 2.45,
+      "learning_rate": 3.6552828175026684e-05,
+      "loss": 0.7977,
+      "step": 3065
+    },
+    {
+      "epoch": 2.45,
+      "learning_rate": 3.649946638207044e-05,
+      "loss": 0.7323,
+      "step": 3066
+    },
+    {
+      "epoch": 2.45,
+      "learning_rate": 3.6446104589114194e-05,
+      "loss": 0.7038,
+      "step": 3067
+    },
+    {
+      "epoch": 2.45,
+      "learning_rate": 3.6392742796157956e-05,
+      "loss": 0.6776,
+      "step": 3068
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 3.633938100320171e-05,
+      "loss": 0.6112,
+      "step": 3069
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 3.6286019210245466e-05,
+      "loss": 0.664,
+      "step": 3070
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 3.623265741728922e-05,
+      "loss": 0.3705,
+      "step": 3071
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 3.6179295624332983e-05,
+      "loss": 0.709,
+      "step": 3072
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 3.612593383137673e-05,
+      "loss": 0.6527,
+      "step": 3073
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 3.6072572038420494e-05,
+      "loss": 0.5496,
+      "step": 3074
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 3.601921024546425e-05,
+      "loss": 0.7227,
+      "step": 3075
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 3.5965848452508004e-05,
+      "loss": 0.6791,
+      "step": 3076
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 3.5912486659551766e-05,
+      "loss": 0.6924,
+      "step": 3077
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 3.5859124866595515e-05,
+      "loss": 0.6377,
+      "step": 3078
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 3.5805763073639277e-05,
+      "loss": 0.0,
+      "step": 3079
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 3.575240128068303e-05,
+      "loss": 0.6846,
+      "step": 3080
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 3.569903948772679e-05,
+      "loss": 0.6922,
+      "step": 3081
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 3.564567769477055e-05,
+      "loss": 0.683,
+      "step": 3082
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 3.5592315901814304e-05,
+      "loss": 0.8433,
+      "step": 3083
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 3.553895410885806e-05,
+      "loss": 0.6368,
+      "step": 3084
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 3.5485592315901815e-05,
+      "loss": 0.7009,
+      "step": 3085
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 3.5432230522945576e-05,
+      "loss": 0.72,
+      "step": 3086
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 3.537886872998933e-05,
+      "loss": 0.7462,
+      "step": 3087
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 3.532550693703309e-05,
+      "loss": 0.7458,
+      "step": 3088
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 3.527214514407684e-05,
+      "loss": 0.3904,
+      "step": 3089
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 3.52187833511206e-05,
+      "loss": 0.6135,
+      "step": 3090
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 3.516542155816436e-05,
+      "loss": 0.3225,
+      "step": 3091
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 3.511205976520811e-05,
+      "loss": 0.5981,
+      "step": 3092
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 3.505869797225187e-05,
+      "loss": 0.357,
+      "step": 3093
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 3.5005336179295625e-05,
+      "loss": 0.7311,
+      "step": 3094
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 3.495197438633938e-05,
+      "loss": 0.6875,
+      "step": 3095
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 3.489861259338314e-05,
+      "loss": 0.8734,
+      "step": 3096
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 3.48452508004269e-05,
+      "loss": 0.7118,
+      "step": 3097
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 3.479188900747065e-05,
+      "loss": 0.772,
+      "step": 3098
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 3.473852721451441e-05,
+      "loss": 0.8654,
+      "step": 3099
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 3.468516542155816e-05,
+      "loss": 0.7777,
+      "step": 3100
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 3.4631803628601925e-05,
+      "loss": 0.6504,
+      "step": 3101
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 3.457844183564568e-05,
+      "loss": 0.7198,
+      "step": 3102
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 3.4525080042689435e-05,
+      "loss": 0.6608,
+      "step": 3103
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 3.447171824973319e-05,
+      "loss": 0.2869,
+      "step": 3104
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 3.441835645677695e-05,
+      "loss": 0.7128,
+      "step": 3105
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 3.436499466382071e-05,
+      "loss": 0.7077,
+      "step": 3106
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 3.431163287086446e-05,
+      "loss": 0.721,
+      "step": 3107
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 3.4258271077908225e-05,
+      "loss": 0.6795,
+      "step": 3108
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 3.420490928495197e-05,
+      "loss": 0.6943,
+      "step": 3109
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 3.4151547491995735e-05,
+      "loss": 0.7988,
+      "step": 3110
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 3.409818569903949e-05,
+      "loss": 0.6585,
+      "step": 3111
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 3.4044823906083245e-05,
+      "loss": 0.6181,
+      "step": 3112
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 3.3991462113127e-05,
+      "loss": 0.7307,
+      "step": 3113
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 3.3938100320170756e-05,
+      "loss": 0.712,
+      "step": 3114
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 3.388473852721452e-05,
+      "loss": 0.6085,
+      "step": 3115
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 3.383137673425827e-05,
+      "loss": 0.6427,
+      "step": 3116
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 3.377801494130203e-05,
+      "loss": 0.7138,
+      "step": 3117
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 3.372465314834578e-05,
+      "loss": 0.7374,
+      "step": 3118
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 3.3671291355389545e-05,
+      "loss": 0.8266,
+      "step": 3119
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 3.36179295624333e-05,
+      "loss": 0.927,
+      "step": 3120
+    },
+    {
+      "epoch": 2.5,
+      "eval_loss": 1.0407631397247314,
+      "eval_runtime": 7416.4944,
+      "eval_samples_per_second": 1.348,
+      "eval_steps_per_second": 0.084,
+      "step": 3120
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 3.3564567769477056e-05,
+      "loss": 0.7789,
+      "step": 3121
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 3.351120597652082e-05,
+      "loss": 0.7491,
+      "step": 3122
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 3.3457844183564566e-05,
+      "loss": 0.6875,
+      "step": 3123
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 3.340448239060833e-05,
+      "loss": 0.8358,
+      "step": 3124
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 3.335112059765208e-05,
+      "loss": 0.613,
+      "step": 3125
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 3.329775880469584e-05,
+      "loss": 0.4327,
+      "step": 3126
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 3.32443970117396e-05,
+      "loss": 0.7446,
+      "step": 3127
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 3.319103521878335e-05,
+      "loss": 0.7872,
+      "step": 3128
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 3.313767342582711e-05,
+      "loss": 0.6697,
+      "step": 3129
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 3.3084311632870866e-05,
+      "loss": 0.4288,
+      "step": 3130
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 3.303094983991462e-05,
+      "loss": 0.6131,
+      "step": 3131
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 3.297758804695838e-05,
+      "loss": 0.6634,
+      "step": 3132
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 3.292422625400214e-05,
+      "loss": 0.623,
+      "step": 3133
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 3.2870864461045893e-05,
+      "loss": 0.7076,
+      "step": 3134
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 3.281750266808965e-05,
+      "loss": 0.7318,
+      "step": 3135
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 3.2764140875133404e-05,
+      "loss": 0.6438,
+      "step": 3136
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 3.271077908217716e-05,
+      "loss": 0.6637,
+      "step": 3137
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 3.265741728922092e-05,
+      "loss": 0.6189,
+      "step": 3138
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 3.2604055496264676e-05,
+      "loss": 0.6857,
+      "step": 3139
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 3.255069370330843e-05,
+      "loss": 0.8388,
+      "step": 3140
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 3.249733191035219e-05,
+      "loss": 0.3576,
+      "step": 3141
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 3.244397011739594e-05,
+      "loss": 0.7666,
+      "step": 3142
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 3.2390608324439704e-05,
+      "loss": 0.8636,
+      "step": 3143
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 3.233724653148346e-05,
+      "loss": 0.7282,
+      "step": 3144
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 3.2283884738527214e-05,
+      "loss": 0.3244,
+      "step": 3145
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 3.2230522945570976e-05,
+      "loss": 0.7007,
+      "step": 3146
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 3.217716115261473e-05,
+      "loss": 0.7751,
+      "step": 3147
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 3.2123799359658487e-05,
+      "loss": 0.3919,
+      "step": 3148
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 3.207043756670224e-05,
+      "loss": 0.67,
+      "step": 3149
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 3.2017075773746e-05,
+      "loss": 0.6731,
+      "step": 3150
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 3.196371398078976e-05,
+      "loss": 0.6315,
+      "step": 3151
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 3.1910352187833514e-05,
+      "loss": 0.6959,
+      "step": 3152
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 3.185699039487727e-05,
+      "loss": 0.7723,
+      "step": 3153
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 3.1803628601921024e-05,
+      "loss": 0.3677,
+      "step": 3154
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 3.1750266808964786e-05,
+      "loss": 0.7033,
+      "step": 3155
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 3.1696905016008535e-05,
+      "loss": 0.6835,
+      "step": 3156
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 3.16435432230523e-05,
+      "loss": 0.6881,
+      "step": 3157
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 3.159018143009605e-05,
+      "loss": 0.6652,
+      "step": 3158
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 3.153681963713981e-05,
+      "loss": 0.735,
+      "step": 3159
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 3.148345784418357e-05,
+      "loss": 0.717,
+      "step": 3160
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 3.143009605122732e-05,
+      "loss": 0.5997,
+      "step": 3161
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 3.137673425827108e-05,
+      "loss": 0.8306,
+      "step": 3162
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 3.1323372465314835e-05,
+      "loss": 0.7297,
+      "step": 3163
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 3.127001067235859e-05,
+      "loss": 0.7159,
+      "step": 3164
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 3.121664887940235e-05,
+      "loss": 0.5976,
+      "step": 3165
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 3.116328708644611e-05,
+      "loss": 0.8728,
+      "step": 3166
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 3.110992529348986e-05,
+      "loss": 0.7355,
+      "step": 3167
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 3.105656350053362e-05,
+      "loss": 0.7139,
+      "step": 3168
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 3.100320170757738e-05,
+      "loss": 0.3762,
+      "step": 3169
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 3.0949839914621135e-05,
+      "loss": 0.6356,
+      "step": 3170
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 3.089647812166489e-05,
+      "loss": 0.6573,
+      "step": 3171
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 3.0843116328708645e-05,
+      "loss": 0.7328,
+      "step": 3172
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 3.07897545357524e-05,
+      "loss": 0.6547,
+      "step": 3173
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 3.073639274279616e-05,
+      "loss": 0.7944,
+      "step": 3174
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 3.068303094983991e-05,
+      "loss": 0.6708,
+      "step": 3175
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 3.062966915688367e-05,
+      "loss": 0.7251,
+      "step": 3176
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 3.057630736392743e-05,
+      "loss": 0.7959,
+      "step": 3177
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 3.052294557097118e-05,
+      "loss": 0.6188,
+      "step": 3178
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 3.046958377801494e-05,
+      "loss": 0.7932,
+      "step": 3179
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 3.04162219850587e-05,
+      "loss": 0.6698,
+      "step": 3180
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 3.0362860192102455e-05,
+      "loss": 0.8089,
+      "step": 3181
+    },
+    {
+      "epoch": 2.55,
+      "learning_rate": 3.0309498399146214e-05,
+      "loss": 0.6667,
+      "step": 3182
+    },
+    {
+      "epoch": 2.55,
+      "learning_rate": 3.0256136606189972e-05,
+      "loss": 0.6519,
+      "step": 3183
+    },
+    {
+      "epoch": 2.55,
+      "learning_rate": 3.0202774813233724e-05,
+      "loss": 0.7077,
+      "step": 3184
+    },
+    {
+      "epoch": 2.55,
+      "learning_rate": 3.0149413020277483e-05,
+      "loss": 0.7387,
+      "step": 3185
+    },
+    {
+      "epoch": 2.55,
+      "learning_rate": 3.0096051227321238e-05,
+      "loss": 0.7815,
+      "step": 3186
+    },
+    {
+      "epoch": 2.55,
+      "learning_rate": 3.0042689434364997e-05,
+      "loss": 0.7,
+      "step": 3187
+    },
+    {
+      "epoch": 2.55,
+      "learning_rate": 2.9989327641408755e-05,
+      "loss": 0.3845,
+      "step": 3188
+    },
+    {
+      "epoch": 2.55,
+      "learning_rate": 2.9935965848452507e-05,
+      "loss": 0.7395,
+      "step": 3189
+    },
+    {
+      "epoch": 2.55,
+      "learning_rate": 2.9882604055496266e-05,
+      "loss": 0.3958,
+      "step": 3190
+    },
+    {
+      "epoch": 2.55,
+      "learning_rate": 2.9829242262540024e-05,
+      "loss": 0.7855,
+      "step": 3191
+    },
+    {
+      "epoch": 2.55,
+      "learning_rate": 2.977588046958378e-05,
+      "loss": 0.7585,
+      "step": 3192
+    },
+    {
+      "epoch": 2.55,
+      "learning_rate": 2.9722518676627538e-05,
+      "loss": 0.7922,
+      "step": 3193
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 2.9669156883671297e-05,
+      "loss": 0.7581,
+      "step": 3194
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 2.961579509071505e-05,
+      "loss": 0.78,
+      "step": 3195
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 2.9562433297758807e-05,
+      "loss": 0.6757,
+      "step": 3196
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 2.950907150480256e-05,
+      "loss": 0.7598,
+      "step": 3197
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 2.9455709711846317e-05,
+      "loss": 0.8073,
+      "step": 3198
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 2.9402347918890076e-05,
+      "loss": 0.7006,
+      "step": 3199
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 2.934898612593383e-05,
+      "loss": 0.3888,
+      "step": 3200
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 2.929562433297759e-05,
+      "loss": 0.6269,
+      "step": 3201
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 2.9242262540021348e-05,
+      "loss": 0.3432,
+      "step": 3202
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 2.91889007470651e-05,
+      "loss": 0.674,
+      "step": 3203
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 2.913553895410886e-05,
+      "loss": 0.7014,
+      "step": 3204
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 2.9082177161152617e-05,
+      "loss": 0.6713,
+      "step": 3205
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 2.9028815368196372e-05,
+      "loss": 0.6517,
+      "step": 3206
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 2.897545357524013e-05,
+      "loss": 0.3277,
+      "step": 3207
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 2.892209178228389e-05,
+      "loss": 0.6594,
+      "step": 3208
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 2.886872998932764e-05,
+      "loss": 0.6992,
+      "step": 3209
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 2.88153681963714e-05,
+      "loss": 0.4434,
+      "step": 3210
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 2.8762006403415155e-05,
+      "loss": 0.6554,
+      "step": 3211
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 2.8708644610458914e-05,
+      "loss": 0.6649,
+      "step": 3212
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 2.8655282817502672e-05,
+      "loss": 0.5903,
+      "step": 3213
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 2.8601921024546424e-05,
+      "loss": 0.8022,
+      "step": 3214
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 2.8548559231590183e-05,
+      "loss": 0.331,
+      "step": 3215
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 2.849519743863394e-05,
+      "loss": 0.7199,
+      "step": 3216
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 2.8441835645677693e-05,
+      "loss": 0.7952,
+      "step": 3217
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 2.8388473852721455e-05,
+      "loss": 0.7541,
+      "step": 3218
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 2.8335112059765214e-05,
+      "loss": 0.6522,
+      "step": 3219
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 2.8281750266808965e-05,
+      "loss": 0.6825,
+      "step": 3220
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 2.8228388473852724e-05,
+      "loss": 0.3672,
+      "step": 3221
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 2.8175026680896476e-05,
+      "loss": 0.7185,
+      "step": 3222
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 2.8121664887940234e-05,
+      "loss": 0.3796,
+      "step": 3223
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 2.8068303094983993e-05,
+      "loss": 0.6376,
+      "step": 3224
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 2.8014941302027748e-05,
+      "loss": 0.7736,
+      "step": 3225
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 2.7961579509071507e-05,
+      "loss": 0.7553,
+      "step": 3226
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 2.7908217716115265e-05,
+      "loss": 0.4675,
+      "step": 3227
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 2.7854855923159017e-05,
+      "loss": 0.0,
+      "step": 3228
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 2.7801494130202776e-05,
+      "loss": 0.7185,
+      "step": 3229
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 2.7748132337246534e-05,
+      "loss": 0.6868,
+      "step": 3230
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 2.769477054429029e-05,
+      "loss": 0.6509,
+      "step": 3231
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 2.7641408751334048e-05,
+      "loss": 0.5982,
+      "step": 3232
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 2.75880469583778e-05,
+      "loss": 0.6697,
+      "step": 3233
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 2.753468516542156e-05,
+      "loss": 0.61,
+      "step": 3234
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 2.7481323372465317e-05,
+      "loss": 0.3517,
+      "step": 3235
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 2.7427961579509072e-05,
+      "loss": 0.7455,
+      "step": 3236
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 2.737459978655283e-05,
+      "loss": 0.7341,
+      "step": 3237
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 2.732123799359659e-05,
+      "loss": 0.5155,
+      "step": 3238
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 2.726787620064034e-05,
+      "loss": 0.6555,
+      "step": 3239
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 2.72145144076841e-05,
+      "loss": 0.8226,
+      "step": 3240
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 2.716115261472786e-05,
+      "loss": 0.6781,
+      "step": 3241
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 2.710779082177161e-05,
+      "loss": 0.7608,
+      "step": 3242
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 2.705442902881537e-05,
+      "loss": 0.7087,
+      "step": 3243
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 2.7001067235859127e-05,
+      "loss": 0.6656,
+      "step": 3244
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 2.6947705442902882e-05,
+      "loss": 0.8386,
+      "step": 3245
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 2.689434364994664e-05,
+      "loss": 0.3105,
+      "step": 3246
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 2.6840981856990393e-05,
+      "loss": 0.6836,
+      "step": 3247
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 2.678762006403415e-05,
+      "loss": 0.7565,
+      "step": 3248
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 2.673425827107791e-05,
+      "loss": 0.3935,
+      "step": 3249
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 2.6680896478121665e-05,
+      "loss": 0.7279,
+      "step": 3250
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 2.6627534685165424e-05,
+      "loss": 0.6637,
+      "step": 3251
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 2.6574172892209182e-05,
+      "loss": 0.6696,
+      "step": 3252
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 2.6520811099252934e-05,
+      "loss": 0.7764,
+      "step": 3253
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 2.6467449306296693e-05,
+      "loss": 0.645,
+      "step": 3254
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 2.641408751334045e-05,
+      "loss": 0.7352,
+      "step": 3255
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 2.6360725720384207e-05,
+      "loss": 0.6851,
+      "step": 3256
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 2.6307363927427965e-05,
+      "loss": 0.7796,
+      "step": 3257
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 2.6254002134471717e-05,
+      "loss": 0.6355,
+      "step": 3258
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 2.6200640341515475e-05,
+      "loss": 0.6386,
+      "step": 3259
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 2.6147278548559234e-05,
+      "loss": 0.0,
+      "step": 3260
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 2.6093916755602986e-05,
+      "loss": 0.7524,
+      "step": 3261
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 2.6040554962646744e-05,
+      "loss": 0.3635,
+      "step": 3262
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 2.5987193169690503e-05,
+      "loss": 0.743,
+      "step": 3263
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 2.5933831376734258e-05,
+      "loss": 0.37,
+      "step": 3264
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 2.5880469583778017e-05,
+      "loss": 0.7412,
+      "step": 3265
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 2.5827107790821775e-05,
+      "loss": 0.6609,
+      "step": 3266
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 2.5773745997865527e-05,
+      "loss": 0.7458,
+      "step": 3267
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 2.5720384204909286e-05,
+      "loss": 0.49,
+      "step": 3268
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 2.5667022411953044e-05,
+      "loss": 0.801,
+      "step": 3269
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 2.56136606189968e-05,
+      "loss": 0.6224,
+      "step": 3270
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 2.5560298826040558e-05,
+      "loss": 0.7051,
+      "step": 3271
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 2.550693703308431e-05,
+      "loss": 0.8289,
+      "step": 3272
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 2.545357524012807e-05,
+      "loss": 0.6084,
+      "step": 3273
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 2.5400213447171827e-05,
+      "loss": 0.6847,
+      "step": 3274
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 2.5346851654215582e-05,
+      "loss": 0.8357,
+      "step": 3275
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 2.529348986125934e-05,
+      "loss": 0.6796,
+      "step": 3276
+    },
+    {
+      "epoch": 2.62,
+      "eval_loss": 1.0375471115112305,
+      "eval_runtime": 7364.1755,
+      "eval_samples_per_second": 1.358,
+      "eval_steps_per_second": 0.085,
+      "step": 3276
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 2.52401280683031e-05,
+      "loss": 0.682,
+      "step": 3277
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 2.518676627534685e-05,
+      "loss": 0.4051,
+      "step": 3278
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 2.513340448239061e-05,
+      "loss": 0.7056,
+      "step": 3279
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 2.508004268943437e-05,
+      "loss": 0.7224,
+      "step": 3280
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 2.502668089647812e-05,
+      "loss": 0.738,
+      "step": 3281
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 2.4973319103521882e-05,
+      "loss": 0.7322,
+      "step": 3282
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 2.4919957310565637e-05,
+      "loss": 0.7433,
+      "step": 3283
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 2.4866595517609393e-05,
+      "loss": 0.6727,
+      "step": 3284
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 2.4813233724653148e-05,
+      "loss": 0.6851,
+      "step": 3285
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 2.4759871931696906e-05,
+      "loss": 0.6986,
+      "step": 3286
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 2.470651013874066e-05,
+      "loss": 0.3082,
+      "step": 3287
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 2.465314834578442e-05,
+      "loss": 0.6232,
+      "step": 3288
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 2.459978655282818e-05,
+      "loss": 0.6721,
+      "step": 3289
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 2.4546424759871934e-05,
+      "loss": 0.6779,
+      "step": 3290
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 2.449306296691569e-05,
+      "loss": 0.6745,
+      "step": 3291
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 2.4439701173959444e-05,
+      "loss": 0.7071,
+      "step": 3292
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 2.4386339381003203e-05,
+      "loss": 0.2983,
+      "step": 3293
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 2.4332977588046958e-05,
+      "loss": 0.299,
+      "step": 3294
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 2.4279615795090717e-05,
+      "loss": 0.6591,
+      "step": 3295
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 2.4226254002134475e-05,
+      "loss": 0.6608,
+      "step": 3296
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 2.417289220917823e-05,
+      "loss": 0.726,
+      "step": 3297
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 2.4119530416221986e-05,
+      "loss": 0.7285,
+      "step": 3298
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 2.406616862326574e-05,
+      "loss": 0.7858,
+      "step": 3299
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 2.40128068303095e-05,
+      "loss": 0.8687,
+      "step": 3300
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 2.3959445037353258e-05,
+      "loss": 0.6627,
+      "step": 3301
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 2.3906083244397013e-05,
+      "loss": 0.6653,
+      "step": 3302
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 2.385272145144077e-05,
+      "loss": 0.6675,
+      "step": 3303
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 2.3799359658484527e-05,
+      "loss": 0.7501,
+      "step": 3304
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 2.3745997865528282e-05,
+      "loss": 0.7658,
+      "step": 3305
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 2.3692636072572037e-05,
+      "loss": 0.6623,
+      "step": 3306
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 2.3639274279615796e-05,
+      "loss": 0.7281,
+      "step": 3307
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 2.3585912486659554e-05,
+      "loss": 0.6921,
+      "step": 3308
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 2.353255069370331e-05,
+      "loss": 0.7671,
+      "step": 3309
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 2.3479188900747065e-05,
+      "loss": 0.5807,
+      "step": 3310
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 2.3425827107790823e-05,
+      "loss": 0.7643,
+      "step": 3311
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 2.337246531483458e-05,
+      "loss": 0.6427,
+      "step": 3312
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 2.3319103521878334e-05,
+      "loss": 0.7599,
+      "step": 3313
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 2.3265741728922096e-05,
+      "loss": 0.6222,
+      "step": 3314
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 2.321237993596585e-05,
+      "loss": 0.7113,
+      "step": 3315
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 2.3159018143009606e-05,
+      "loss": 0.7476,
+      "step": 3316
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 2.310565635005336e-05,
+      "loss": 0.5102,
+      "step": 3317
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 2.305229455709712e-05,
+      "loss": 0.6741,
+      "step": 3318
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 2.2998932764140875e-05,
+      "loss": 0.5914,
+      "step": 3319
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 2.2945570971184634e-05,
+      "loss": 0.6069,
+      "step": 3320
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 2.2892209178228392e-05,
+      "loss": 0.75,
+      "step": 3321
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 2.2838847385272147e-05,
+      "loss": 0.7045,
+      "step": 3322
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 2.2785485592315903e-05,
+      "loss": 0.6035,
+      "step": 3323
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 2.2732123799359658e-05,
+      "loss": 0.6778,
+      "step": 3324
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 2.2678762006403416e-05,
+      "loss": 0.7057,
+      "step": 3325
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 2.262540021344717e-05,
+      "loss": 0.3298,
+      "step": 3326
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 2.257203842049093e-05,
+      "loss": 0.7455,
+      "step": 3327
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 2.2518676627534685e-05,
+      "loss": 0.8572,
+      "step": 3328
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 2.2465314834578444e-05,
+      "loss": 0.6581,
+      "step": 3329
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 2.24119530416222e-05,
+      "loss": 0.7966,
+      "step": 3330
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 2.2358591248665954e-05,
+      "loss": 0.7499,
+      "step": 3331
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 2.2305229455709713e-05,
+      "loss": 0.6752,
+      "step": 3332
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 2.225186766275347e-05,
+      "loss": 0.5669,
+      "step": 3333
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 2.2198505869797227e-05,
+      "loss": 0.7264,
+      "step": 3334
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 2.2145144076840982e-05,
+      "loss": 0.7512,
+      "step": 3335
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 2.209178228388474e-05,
+      "loss": 0.3318,
+      "step": 3336
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 2.2038420490928496e-05,
+      "loss": 0.7175,
+      "step": 3337
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 2.198505869797225e-05,
+      "loss": 0.7041,
+      "step": 3338
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 2.193169690501601e-05,
+      "loss": 0.7695,
+      "step": 3339
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 2.1878335112059768e-05,
+      "loss": 0.5884,
+      "step": 3340
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 2.1824973319103523e-05,
+      "loss": 0.7674,
+      "step": 3341
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 2.177161152614728e-05,
+      "loss": 0.6876,
+      "step": 3342
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 2.1718249733191037e-05,
+      "loss": 0.6364,
+      "step": 3343
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 2.1664887940234792e-05,
+      "loss": 0.738,
+      "step": 3344
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 2.1611526147278547e-05,
+      "loss": 0.5922,
+      "step": 3345
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 2.1558164354322306e-05,
+      "loss": 0.7955,
+      "step": 3346
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 2.1504802561366065e-05,
+      "loss": 0.7606,
+      "step": 3347
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 2.145144076840982e-05,
+      "loss": 0.6343,
+      "step": 3348
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 2.1398078975453575e-05,
+      "loss": 0.6374,
+      "step": 3349
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 2.1344717182497334e-05,
+      "loss": 0.7153,
+      "step": 3350
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 2.129135538954109e-05,
+      "loss": 0.7485,
+      "step": 3351
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 2.1237993596584847e-05,
+      "loss": 0.3711,
+      "step": 3352
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 2.1184631803628602e-05,
+      "loss": 0.3615,
+      "step": 3353
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 2.113127001067236e-05,
+      "loss": 0.6385,
+      "step": 3354
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 2.1077908217716116e-05,
+      "loss": 0.7621,
+      "step": 3355
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 2.102454642475987e-05,
+      "loss": 0.5661,
+      "step": 3356
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 2.097118463180363e-05,
+      "loss": 0.3224,
+      "step": 3357
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 2.0917822838847385e-05,
+      "loss": 0.7084,
+      "step": 3358
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 2.0864461045891144e-05,
+      "loss": 0.8134,
+      "step": 3359
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 2.08110992529349e-05,
+      "loss": 0.7268,
+      "step": 3360
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 2.0757737459978658e-05,
+      "loss": 0.5934,
+      "step": 3361
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 2.0704375667022413e-05,
+      "loss": 0.7653,
+      "step": 3362
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 2.0651013874066168e-05,
+      "loss": 0.7028,
+      "step": 3363
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 2.0597652081109927e-05,
+      "loss": 0.6611,
+      "step": 3364
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 2.0544290288153685e-05,
+      "loss": 0.7216,
+      "step": 3365
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 2.049092849519744e-05,
+      "loss": 0.3516,
+      "step": 3366
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 2.0437566702241196e-05,
+      "loss": 0.3808,
+      "step": 3367
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 2.0384204909284954e-05,
+      "loss": 0.7189,
+      "step": 3368
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 2.033084311632871e-05,
+      "loss": 0.6527,
+      "step": 3369
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 2.0277481323372464e-05,
+      "loss": 0.7503,
+      "step": 3370
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 2.0224119530416223e-05,
+      "loss": 0.6563,
+      "step": 3371
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 2.017075773745998e-05,
+      "loss": 0.7958,
+      "step": 3372
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 2.0117395944503737e-05,
+      "loss": 0.8148,
+      "step": 3373
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 2.0064034151547492e-05,
+      "loss": 0.6707,
+      "step": 3374
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 2.001067235859125e-05,
+      "loss": 0.5433,
+      "step": 3375
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 1.9957310565635006e-05,
+      "loss": 0.6539,
+      "step": 3376
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 1.9903948772678764e-05,
+      "loss": 0.679,
+      "step": 3377
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 1.985058697972252e-05,
+      "loss": 0.7149,
+      "step": 3378
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 1.9797225186766278e-05,
+      "loss": 0.3527,
+      "step": 3379
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 1.9743863393810033e-05,
+      "loss": 0.7149,
+      "step": 3380
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 1.969050160085379e-05,
+      "loss": 0.6993,
+      "step": 3381
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 1.9637139807897544e-05,
+      "loss": 0.5834,
+      "step": 3382
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 1.9583778014941302e-05,
+      "loss": 0.7522,
+      "step": 3383
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 1.953041622198506e-05,
+      "loss": 0.3132,
+      "step": 3384
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 1.9477054429028816e-05,
+      "loss": 0.5427,
+      "step": 3385
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 1.9423692636072575e-05,
+      "loss": 0.69,
+      "step": 3386
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 1.937033084311633e-05,
+      "loss": 0.7013,
+      "step": 3387
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 1.9316969050160085e-05,
+      "loss": 0.822,
+      "step": 3388
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 1.926360725720384e-05,
+      "loss": 0.789,
+      "step": 3389
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 1.92102454642476e-05,
+      "loss": 0.6778,
+      "step": 3390
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 1.9156883671291357e-05,
+      "loss": 0.9233,
+      "step": 3391
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 1.9103521878335113e-05,
+      "loss": 0.5563,
+      "step": 3392
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 1.905016008537887e-05,
+      "loss": 0.7744,
+      "step": 3393
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 1.8996798292422626e-05,
+      "loss": 0.6949,
+      "step": 3394
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 1.894343649946638e-05,
+      "loss": 0.7536,
+      "step": 3395
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 1.889007470651014e-05,
+      "loss": 0.7802,
+      "step": 3396
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 1.88367129135539e-05,
+      "loss": 0.6807,
+      "step": 3397
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 1.8783351120597654e-05,
+      "loss": 0.7635,
+      "step": 3398
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 1.872998932764141e-05,
+      "loss": 0.7023,
+      "step": 3399
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 1.8676627534685168e-05,
+      "loss": 0.649,
+      "step": 3400
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 1.8623265741728923e-05,
+      "loss": 0.6834,
+      "step": 3401
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 1.8569903948772678e-05,
+      "loss": 0.3656,
+      "step": 3402
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 1.8516542155816437e-05,
+      "loss": 0.7536,
+      "step": 3403
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 1.8463180362860195e-05,
+      "loss": 0.7242,
+      "step": 3404
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 1.840981856990395e-05,
+      "loss": 0.3831,
+      "step": 3405
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 1.8356456776947706e-05,
+      "loss": 0.5478,
+      "step": 3406
+    },
+    {
+      "epoch": 2.73,
+      "learning_rate": 1.830309498399146e-05,
+      "loss": 1.0333,
+      "step": 3407
+    },
+    {
+      "epoch": 2.73,
+      "learning_rate": 1.824973319103522e-05,
+      "loss": 0.5791,
+      "step": 3408
+    },
+    {
+      "epoch": 2.73,
+      "learning_rate": 1.8196371398078978e-05,
+      "loss": 0.7341,
+      "step": 3409
+    },
+    {
+      "epoch": 2.73,
+      "learning_rate": 1.8143009605122733e-05,
+      "loss": 0.7205,
+      "step": 3410
+    },
+    {
+      "epoch": 2.73,
+      "learning_rate": 1.8089647812166492e-05,
+      "loss": 0.6394,
+      "step": 3411
+    },
+    {
+      "epoch": 2.73,
+      "learning_rate": 1.8036286019210247e-05,
+      "loss": 0.3593,
+      "step": 3412
+    },
+    {
+      "epoch": 2.73,
+      "learning_rate": 1.7982924226254002e-05,
+      "loss": 0.6519,
+      "step": 3413
+    },
+    {
+      "epoch": 2.73,
+      "learning_rate": 1.7929562433297757e-05,
+      "loss": 0.6972,
+      "step": 3414
+    },
+    {
+      "epoch": 2.73,
+      "learning_rate": 1.7876200640341516e-05,
+      "loss": 0.7077,
+      "step": 3415
+    },
+    {
+      "epoch": 2.73,
+      "learning_rate": 1.7822838847385274e-05,
+      "loss": 0.7698,
+      "step": 3416
+    },
+    {
+      "epoch": 2.73,
+      "learning_rate": 1.776947705442903e-05,
+      "loss": 0.7196,
+      "step": 3417
+    },
+    {
+      "epoch": 2.73,
+      "learning_rate": 1.7716115261472788e-05,
+      "loss": 0.6608,
+      "step": 3418
+    },
+    {
+      "epoch": 2.74,
+      "learning_rate": 1.7662753468516543e-05,
+      "loss": 0.6732,
+      "step": 3419
+    },
+    {
+      "epoch": 2.74,
+      "learning_rate": 1.76093916755603e-05,
+      "loss": 0.7088,
+      "step": 3420
+    },
+    {
+      "epoch": 2.74,
+      "learning_rate": 1.7556029882604054e-05,
+      "loss": 0.627,
+      "step": 3421
+    },
+    {
+      "epoch": 2.74,
+      "learning_rate": 1.7502668089647812e-05,
+      "loss": 0.7934,
+      "step": 3422
+    },
+    {
+      "epoch": 2.74,
+      "learning_rate": 1.744930629669157e-05,
+      "loss": 0.6117,
+      "step": 3423
+    },
+    {
+      "epoch": 2.74,
+      "learning_rate": 1.7395944503735326e-05,
+      "loss": 0.3573,
+      "step": 3424
+    },
+    {
+      "epoch": 2.74,
+      "learning_rate": 1.734258271077908e-05,
+      "loss": 0.7919,
+      "step": 3425
+    },
+    {
+      "epoch": 2.74,
+      "learning_rate": 1.728922091782284e-05,
+      "loss": 0.76,
+      "step": 3426
+    },
+    {
+      "epoch": 2.74,
+      "learning_rate": 1.7235859124866595e-05,
+      "loss": 0.6371,
+      "step": 3427
+    },
+    {
+      "epoch": 2.74,
+      "learning_rate": 1.7182497331910354e-05,
+      "loss": 0.5717,
+      "step": 3428
+    },
+    {
+      "epoch": 2.74,
+      "learning_rate": 1.7129135538954112e-05,
+      "loss": 0.7123,
+      "step": 3429
+    },
+    {
+      "epoch": 2.74,
+      "learning_rate": 1.7075773745997867e-05,
+      "loss": 0.6805,
+      "step": 3430
+    },
+    {
+      "epoch": 2.74,
+      "learning_rate": 1.7022411953041623e-05,
+      "loss": 0.6616,
+      "step": 3431
+    },
+    {
+      "epoch": 2.75,
+      "learning_rate": 1.6969050160085378e-05,
+      "loss": 0.7887,
+      "step": 3432
+    },
+    {
+      "epoch": 2.75,
+      "eval_loss": 1.0496493577957153,
+      "eval_runtime": 7804.2565,
+      "eval_samples_per_second": 1.281,
+      "eval_steps_per_second": 0.08,
+      "step": 3432
+    },
+    {
+      "epoch": 2.75,
+      "learning_rate": 1.6915688367129136e-05,
+      "loss": 0.6921,
+      "step": 3433
+    },
+    {
+      "epoch": 2.75,
+      "learning_rate": 1.686232657417289e-05,
+      "loss": 0.7124,
+      "step": 3434
+    },
+    {
+      "epoch": 2.75,
+      "learning_rate": 1.680896478121665e-05,
+      "loss": 0.7992,
+      "step": 3435
+    },
+    {
+      "epoch": 2.75,
+      "learning_rate": 1.675560298826041e-05,
+      "loss": 0.6729,
+      "step": 3436
+    },
+    {
+      "epoch": 2.75,
+      "learning_rate": 1.6702241195304164e-05,
+      "loss": 0.3349,
+      "step": 3437
+    },
+    {
+      "epoch": 2.75,
+      "learning_rate": 1.664887940234792e-05,
+      "loss": 0.6453,
+      "step": 3438
+    },
+    {
+      "epoch": 2.75,
+      "learning_rate": 1.6595517609391674e-05,
+      "loss": 0.6588,
+      "step": 3439
+    },
+    {
+      "epoch": 2.75,
+      "learning_rate": 1.6542155816435433e-05,
+      "loss": 0.804,
+      "step": 3440
+    },
+    {
+      "epoch": 2.75,
+      "learning_rate": 1.648879402347919e-05,
+      "loss": 0.7064,
+      "step": 3441
+    },
+    {
+      "epoch": 2.75,
+      "learning_rate": 1.6435432230522947e-05,
+      "loss": 0.7628,
+      "step": 3442
+    },
+    {
+      "epoch": 2.75,
+      "learning_rate": 1.6382070437566702e-05,
+      "loss": 0.7086,
+      "step": 3443
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 1.632870864461046e-05,
+      "loss": 0.7577,
+      "step": 3444
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 1.6275346851654216e-05,
+      "loss": 0.3628,
+      "step": 3445
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 1.622198505869797e-05,
+      "loss": 0.8589,
+      "step": 3446
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 1.616862326574173e-05,
+      "loss": 0.7951,
+      "step": 3447
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 1.6115261472785488e-05,
+      "loss": 0.601,
+      "step": 3448
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 1.6061899679829243e-05,
+      "loss": 0.697,
+      "step": 3449
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 1.6008537886873e-05,
+      "loss": 0.7719,
+      "step": 3450
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 1.5955176093916757e-05,
+      "loss": 0.7125,
+      "step": 3451
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 1.5901814300960512e-05,
+      "loss": 0.761,
+      "step": 3452
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 1.5848452508004267e-05,
+      "loss": 0.6888,
+      "step": 3453
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 1.5795090715048026e-05,
+      "loss": 0.7615,
+      "step": 3454
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 1.5741728922091785e-05,
+      "loss": 0.3961,
+      "step": 3455
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 1.568836712913554e-05,
+      "loss": 0.5947,
+      "step": 3456
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 1.5635005336179295e-05,
+      "loss": 0.7455,
+      "step": 3457
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 1.5581643543223054e-05,
+      "loss": 0.3332,
+      "step": 3458
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 1.552828175026681e-05,
+      "loss": 0.6635,
+      "step": 3459
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 1.5474919957310567e-05,
+      "loss": 0.6725,
+      "step": 3460
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 1.5421558164354323e-05,
+      "loss": 0.7804,
+      "step": 3461
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 1.536819637139808e-05,
+      "loss": 0.7755,
+      "step": 3462
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 1.5314834578441836e-05,
+      "loss": 0.4521,
+      "step": 3463
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 1.526147278548559e-05,
+      "loss": 0.0,
+      "step": 3464
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 1.520811099252935e-05,
+      "loss": 0.7226,
+      "step": 3465
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 1.5154749199573107e-05,
+      "loss": 0.8451,
+      "step": 3466
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 1.5101387406616862e-05,
+      "loss": 0.6809,
+      "step": 3467
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 1.5048025613660619e-05,
+      "loss": 0.748,
+      "step": 3468
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 1.4994663820704378e-05,
+      "loss": 0.3078,
+      "step": 3469
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 1.4941302027748133e-05,
+      "loss": 0.681,
+      "step": 3470
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 1.488794023479189e-05,
+      "loss": 0.7216,
+      "step": 3471
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 1.4834578441835648e-05,
+      "loss": 0.3227,
+      "step": 3472
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 1.4781216648879403e-05,
+      "loss": 0.6377,
+      "step": 3473
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 1.4727854855923159e-05,
+      "loss": 0.371,
+      "step": 3474
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 1.4674493062966916e-05,
+      "loss": 0.6315,
+      "step": 3475
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 1.4621131270010674e-05,
+      "loss": 0.8492,
+      "step": 3476
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 1.456776947705443e-05,
+      "loss": 0.7227,
+      "step": 3477
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 1.4514407684098186e-05,
+      "loss": 0.6784,
+      "step": 3478
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 1.4461045891141945e-05,
+      "loss": 0.7877,
+      "step": 3479
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 1.44076840981857e-05,
+      "loss": 0.6999,
+      "step": 3480
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 1.4354322305229457e-05,
+      "loss": 0.4247,
+      "step": 3481
+    },
+    {
+      "epoch": 2.79,
+      "learning_rate": 1.4300960512273212e-05,
+      "loss": 0.0,
+      "step": 3482
+    },
+    {
+      "epoch": 2.79,
+      "learning_rate": 1.424759871931697e-05,
+      "loss": 0.6634,
+      "step": 3483
+    },
+    {
+      "epoch": 2.79,
+      "learning_rate": 1.4194236926360728e-05,
+      "loss": 0.6452,
+      "step": 3484
+    },
+    {
+      "epoch": 2.79,
+      "learning_rate": 1.4140875133404483e-05,
+      "loss": 0.7205,
+      "step": 3485
+    },
+    {
+      "epoch": 2.79,
+      "learning_rate": 1.4087513340448238e-05,
+      "loss": 0.6753,
+      "step": 3486
+    },
+    {
+      "epoch": 2.79,
+      "learning_rate": 1.4034151547491996e-05,
+      "loss": 0.7782,
+      "step": 3487
+    },
+    {
+      "epoch": 2.79,
+      "learning_rate": 1.3980789754535753e-05,
+      "loss": 0.6635,
+      "step": 3488
+    },
+    {
+      "epoch": 2.79,
+      "learning_rate": 1.3927427961579509e-05,
+      "loss": 0.8089,
+      "step": 3489
+    },
+    {
+      "epoch": 2.79,
+      "learning_rate": 1.3874066168623267e-05,
+      "loss": 0.8115,
+      "step": 3490
+    },
+    {
+      "epoch": 2.79,
+      "learning_rate": 1.3820704375667024e-05,
+      "loss": 0.7017,
+      "step": 3491
+    },
+    {
+      "epoch": 2.79,
+      "learning_rate": 1.376734258271078e-05,
+      "loss": 0.7772,
+      "step": 3492
+    },
+    {
+      "epoch": 2.79,
+      "learning_rate": 1.3713980789754536e-05,
+      "loss": 0.4034,
+      "step": 3493
+    },
+    {
+      "epoch": 2.8,
+      "learning_rate": 1.3660618996798295e-05,
+      "loss": 0.6448,
+      "step": 3494
+    },
+    {
+      "epoch": 2.8,
+      "learning_rate": 1.360725720384205e-05,
+      "loss": 0.7187,
+      "step": 3495
+    },
+    {
+      "epoch": 2.8,
+      "learning_rate": 1.3553895410885805e-05,
+      "loss": 0.8293,
+      "step": 3496
+    },
+    {
+      "epoch": 2.8,
+      "learning_rate": 1.3500533617929564e-05,
+      "loss": 0.7064,
+      "step": 3497
+    },
+    {
+      "epoch": 2.8,
+      "learning_rate": 1.344717182497332e-05,
+      "loss": 0.628,
+      "step": 3498
+    },
+    {
+      "epoch": 2.8,
+      "learning_rate": 1.3393810032017076e-05,
+      "loss": 0.4576,
+      "step": 3499
+    },
+    {
+      "epoch": 2.8,
+      "learning_rate": 1.3340448239060833e-05,
+      "loss": 0.369,
+      "step": 3500
+    },
+    {
+      "epoch": 2.8,
+      "learning_rate": 1.3287086446104591e-05,
+      "loss": 0.4021,
+      "step": 3501
+    },
+    {
+      "epoch": 2.8,
+      "learning_rate": 1.3233724653148346e-05,
+      "loss": 0.6722,
+      "step": 3502
+    },
+    {
+      "epoch": 2.8,
+      "learning_rate": 1.3180362860192103e-05,
+      "loss": 0.5988,
+      "step": 3503
+    },
+    {
+      "epoch": 2.8,
+      "learning_rate": 1.3127001067235858e-05,
+      "loss": 0.7722,
+      "step": 3504
+    },
+    {
+      "epoch": 2.8,
+      "learning_rate": 1.3073639274279617e-05,
+      "loss": 0.3343,
+      "step": 3505
+    },
+    {
+      "epoch": 2.8,
+      "learning_rate": 1.3020277481323372e-05,
+      "loss": 0.6193,
+      "step": 3506
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 1.2966915688367129e-05,
+      "loss": 0.446,
+      "step": 3507
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 1.2913553895410888e-05,
+      "loss": 0.3418,
+      "step": 3508
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 1.2860192102454643e-05,
+      "loss": 0.736,
+      "step": 3509
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 1.28068303094984e-05,
+      "loss": 0.8539,
+      "step": 3510
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 1.2753468516542155e-05,
+      "loss": 0.647,
+      "step": 3511
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 1.2700106723585914e-05,
+      "loss": 0.3861,
+      "step": 3512
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 1.264674493062967e-05,
+      "loss": 0.69,
+      "step": 3513
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 1.2593383137673426e-05,
+      "loss": 0.7055,
+      "step": 3514
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 1.2540021344717184e-05,
+      "loss": 0.5856,
+      "step": 3515
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 1.2486659551760941e-05,
+      "loss": 0.3484,
+      "step": 3516
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 1.2433297758804696e-05,
+      "loss": 0.3915,
+      "step": 3517
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 1.2379935965848453e-05,
+      "loss": 0.3436,
+      "step": 3518
+    },
+    {
+      "epoch": 2.82,
+      "learning_rate": 1.232657417289221e-05,
+      "loss": 0.7104,
+      "step": 3519
+    },
+    {
+      "epoch": 2.82,
+      "learning_rate": 1.2273212379935967e-05,
+      "loss": 0.7734,
+      "step": 3520
+    },
+    {
+      "epoch": 2.82,
+      "learning_rate": 1.2219850586979722e-05,
+      "loss": 0.7549,
+      "step": 3521
+    },
+    {
+      "epoch": 2.82,
+      "learning_rate": 1.2166488794023479e-05,
+      "loss": 0.7751,
+      "step": 3522
+    },
+    {
+      "epoch": 2.82,
+      "learning_rate": 1.2113127001067238e-05,
+      "loss": 0.7033,
+      "step": 3523
+    },
+    {
+      "epoch": 2.82,
+      "learning_rate": 1.2059765208110993e-05,
+      "loss": 0.6302,
+      "step": 3524
+    },
+    {
+      "epoch": 2.82,
+      "learning_rate": 1.200640341515475e-05,
+      "loss": 0.2911,
+      "step": 3525
+    },
+    {
+      "epoch": 2.82,
+      "learning_rate": 1.1953041622198507e-05,
+      "loss": 0.7583,
+      "step": 3526
+    },
+    {
+      "epoch": 2.82,
+      "learning_rate": 1.1899679829242263e-05,
+      "loss": 0.6482,
+      "step": 3527
+    },
+    {
+      "epoch": 2.82,
+      "learning_rate": 1.1846318036286019e-05,
+      "loss": 0.7104,
+      "step": 3528
+    },
+    {
+      "epoch": 2.82,
+      "learning_rate": 1.1792956243329777e-05,
+      "loss": 0.7544,
+      "step": 3529
+    },
+    {
+      "epoch": 2.82,
+      "learning_rate": 1.1739594450373532e-05,
+      "loss": 0.7259,
+      "step": 3530
+    },
+    {
+      "epoch": 2.82,
+      "learning_rate": 1.168623265741729e-05,
+      "loss": 0.7478,
+      "step": 3531
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 1.1632870864461048e-05,
+      "loss": 0.7086,
+      "step": 3532
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 1.1579509071504803e-05,
+      "loss": 0.6978,
+      "step": 3533
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 1.152614727854856e-05,
+      "loss": 0.7481,
+      "step": 3534
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 1.1472785485592317e-05,
+      "loss": 0.6511,
+      "step": 3535
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 1.1419423692636074e-05,
+      "loss": 0.6131,
+      "step": 3536
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 1.1366061899679829e-05,
+      "loss": 0.7275,
+      "step": 3537
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 1.1312700106723586e-05,
+      "loss": 0.5695,
+      "step": 3538
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 1.1259338313767343e-05,
+      "loss": 0.604,
+      "step": 3539
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 1.12059765208111e-05,
+      "loss": 0.7313,
+      "step": 3540
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 1.1152614727854856e-05,
+      "loss": 0.6382,
+      "step": 3541
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 1.1099252934898613e-05,
+      "loss": 0.3701,
+      "step": 3542
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 1.104589114194237e-05,
+      "loss": 0.793,
+      "step": 3543
+    },
+    {
+      "epoch": 2.84,
+      "learning_rate": 1.0992529348986125e-05,
+      "loss": 0.6825,
+      "step": 3544
+    },
+    {
+      "epoch": 2.84,
+      "learning_rate": 1.0939167556029884e-05,
+      "loss": 0.0,
+      "step": 3545
+    },
+    {
+      "epoch": 2.84,
+      "learning_rate": 1.088580576307364e-05,
+      "loss": 0.7319,
+      "step": 3546
+    },
+    {
+      "epoch": 2.84,
+      "learning_rate": 1.0832443970117396e-05,
+      "loss": 0.7155,
+      "step": 3547
+    },
+    {
+      "epoch": 2.84,
+      "learning_rate": 1.0779082177161153e-05,
+      "loss": 0.7002,
+      "step": 3548
+    },
+    {
+      "epoch": 2.84,
+      "learning_rate": 1.072572038420491e-05,
+      "loss": 0.6217,
+      "step": 3549
+    },
+    {
+      "epoch": 2.84,
+      "learning_rate": 1.0672358591248667e-05,
+      "loss": 0.6755,
+      "step": 3550
+    },
+    {
+      "epoch": 2.84,
+      "learning_rate": 1.0618996798292424e-05,
+      "loss": 0.824,
+      "step": 3551
+    },
+    {
+      "epoch": 2.84,
+      "learning_rate": 1.056563500533618e-05,
+      "loss": 0.6632,
+      "step": 3552
+    },
+    {
+      "epoch": 2.84,
+      "learning_rate": 1.0512273212379936e-05,
+      "loss": 0.7786,
+      "step": 3553
+    },
+    {
+      "epoch": 2.84,
+      "learning_rate": 1.0458911419423693e-05,
+      "loss": 0.7546,
+      "step": 3554
+    },
+    {
+      "epoch": 2.84,
+      "learning_rate": 1.040554962646745e-05,
+      "loss": 0.6937,
+      "step": 3555
+    },
+    {
+      "epoch": 2.84,
+      "learning_rate": 1.0352187833511206e-05,
+      "loss": 0.7221,
+      "step": 3556
+    },
+    {
+      "epoch": 2.85,
+      "learning_rate": 1.0298826040554963e-05,
+      "loss": 0.6721,
+      "step": 3557
+    },
+    {
+      "epoch": 2.85,
+      "learning_rate": 1.024546424759872e-05,
+      "loss": 0.7037,
+      "step": 3558
+    },
+    {
+      "epoch": 2.85,
+      "learning_rate": 1.0192102454642477e-05,
+      "loss": 0.6672,
+      "step": 3559
+    },
+    {
+      "epoch": 2.85,
+      "learning_rate": 1.0138740661686232e-05,
+      "loss": 0.3768,
+      "step": 3560
+    },
+    {
+      "epoch": 2.85,
+      "learning_rate": 1.008537886872999e-05,
+      "loss": 0.9035,
+      "step": 3561
+    },
+    {
+      "epoch": 2.85,
+      "learning_rate": 1.0032017075773746e-05,
+      "loss": 0.673,
+      "step": 3562
+    },
+    {
+      "epoch": 2.85,
+      "learning_rate": 9.978655282817503e-06,
+      "loss": 0.6694,
+      "step": 3563
+    },
+    {
+      "epoch": 2.85,
+      "learning_rate": 9.92529348986126e-06,
+      "loss": 0.359,
+      "step": 3564
+    },
+    {
+      "epoch": 2.85,
+      "learning_rate": 9.871931696905017e-06,
+      "loss": 0.7806,
+      "step": 3565
+    },
+    {
+      "epoch": 2.85,
+      "learning_rate": 9.818569903948772e-06,
+      "loss": 0.7206,
+      "step": 3566
+    },
+    {
+      "epoch": 2.85,
+      "learning_rate": 9.76520811099253e-06,
+      "loss": 0.3425,
+      "step": 3567
+    },
+    {
+      "epoch": 2.85,
+      "learning_rate": 9.711846318036287e-06,
+      "loss": 0.7072,
+      "step": 3568
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 9.658484525080043e-06,
+      "loss": 0.7729,
+      "step": 3569
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 9.6051227321238e-06,
+      "loss": 0.659,
+      "step": 3570
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 9.551760939167556e-06,
+      "loss": 0.7876,
+      "step": 3571
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 9.498399146211313e-06,
+      "loss": 0.3176,
+      "step": 3572
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 9.44503735325507e-06,
+      "loss": 0.716,
+      "step": 3573
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 9.391675560298827e-06,
+      "loss": 0.3338,
+      "step": 3574
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 9.338313767342584e-06,
+      "loss": 0.4006,
+      "step": 3575
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 9.284951974386339e-06,
+      "loss": 0.6812,
+      "step": 3576
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 9.231590181430098e-06,
+      "loss": 0.3738,
+      "step": 3577
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 9.178228388473853e-06,
+      "loss": 0.578,
+      "step": 3578
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 9.12486659551761e-06,
+      "loss": 0.8029,
+      "step": 3579
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 9.071504802561367e-06,
+      "loss": 0.6229,
+      "step": 3580
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 9.018143009605123e-06,
+      "loss": 0.8424,
+      "step": 3581
+    },
+    {
+      "epoch": 2.87,
+      "learning_rate": 8.964781216648879e-06,
+      "loss": 0.6148,
+      "step": 3582
+    },
+    {
+      "epoch": 2.87,
+      "learning_rate": 8.911419423692637e-06,
+      "loss": 0.2913,
+      "step": 3583
+    },
+    {
+      "epoch": 2.87,
+      "learning_rate": 8.858057630736394e-06,
+      "loss": 0.7003,
+      "step": 3584
+    },
+    {
+      "epoch": 2.87,
+      "learning_rate": 8.80469583778015e-06,
+      "loss": 0.6602,
+      "step": 3585
+    },
+    {
+      "epoch": 2.87,
+      "learning_rate": 8.751334044823906e-06,
+      "loss": 0.7676,
+      "step": 3586
+    },
+    {
+      "epoch": 2.87,
+      "learning_rate": 8.697972251867663e-06,
+      "loss": 0.6701,
+      "step": 3587
+    },
+    {
+      "epoch": 2.87,
+      "learning_rate": 8.64461045891142e-06,
+      "loss": 0.7076,
+      "step": 3588
+    },
+    {
+      "epoch": 2.87,
+      "eval_loss": 1.036924958229065,
+      "eval_runtime": 7470.467,
+      "eval_samples_per_second": 1.339,
+      "eval_steps_per_second": 0.084,
+      "step": 3588
+    },
+    {
+      "epoch": 2.87,
+      "learning_rate": 8.591248665955177e-06,
+      "loss": 0.6236,
+      "step": 3589
+    },
+    {
+      "epoch": 2.87,
+      "learning_rate": 8.537886872998934e-06,
+      "loss": 0.6494,
+      "step": 3590
+    },
+    {
+      "epoch": 2.87,
+      "learning_rate": 8.484525080042689e-06,
+      "loss": 0.7958,
+      "step": 3591
+    },
+    {
+      "epoch": 2.87,
+      "learning_rate": 8.431163287086446e-06,
+      "loss": 0.7114,
+      "step": 3592
+    },
+    {
+      "epoch": 2.87,
+      "learning_rate": 8.377801494130204e-06,
+      "loss": 0.5759,
+      "step": 3593
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 8.32443970117396e-06,
+      "loss": 0.7708,
+      "step": 3594
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 8.271077908217716e-06,
+      "loss": 0.7467,
+      "step": 3595
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 8.217716115261473e-06,
+      "loss": 0.7081,
+      "step": 3596
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 8.16435432230523e-06,
+      "loss": 0.3816,
+      "step": 3597
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 8.110992529348985e-06,
+      "loss": 0.8583,
+      "step": 3598
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 8.057630736392744e-06,
+      "loss": 0.7053,
+      "step": 3599
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 8.0042689434365e-06,
+      "loss": 0.6158,
+      "step": 3600
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 7.950907150480256e-06,
+      "loss": 0.7084,
+      "step": 3601
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 7.897545357524013e-06,
+      "loss": 0.681,
+      "step": 3602
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 7.84418356456777e-06,
+      "loss": 0.8547,
+      "step": 3603
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 7.790821771611527e-06,
+      "loss": 0.3895,
+      "step": 3604
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 7.737459978655284e-06,
+      "loss": 0.6753,
+      "step": 3605
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 7.68409818569904e-06,
+      "loss": 0.628,
+      "step": 3606
+    },
+    {
+      "epoch": 2.89,
+      "learning_rate": 7.630736392742796e-06,
+      "loss": 0.7251,
+      "step": 3607
+    },
+    {
+      "epoch": 2.89,
+      "learning_rate": 7.5773745997865535e-06,
+      "loss": 0.3177,
+      "step": 3608
+    },
+    {
+      "epoch": 2.89,
+      "learning_rate": 7.5240128068303095e-06,
+      "loss": 0.6644,
+      "step": 3609
+    },
+    {
+      "epoch": 2.89,
+      "learning_rate": 7.470651013874066e-06,
+      "loss": 0.6351,
+      "step": 3610
+    },
+    {
+      "epoch": 2.89,
+      "learning_rate": 7.417289220917824e-06,
+      "loss": 0.5619,
+      "step": 3611
+    },
+    {
+      "epoch": 2.89,
+      "learning_rate": 7.363927427961579e-06,
+      "loss": 0.6573,
+      "step": 3612
+    },
+    {
+      "epoch": 2.89,
+      "learning_rate": 7.310565635005337e-06,
+      "loss": 0.9775,
+      "step": 3613
+    },
+    {
+      "epoch": 2.89,
+      "learning_rate": 7.257203842049093e-06,
+      "loss": 0.3173,
+      "step": 3614
+    },
+    {
+      "epoch": 2.89,
+      "learning_rate": 7.20384204909285e-06,
+      "loss": 0.6497,
+      "step": 3615
+    },
+    {
+      "epoch": 2.89,
+      "learning_rate": 7.150480256136606e-06,
+      "loss": 0.7494,
+      "step": 3616
+    },
+    {
+      "epoch": 2.89,
+      "learning_rate": 7.097118463180364e-06,
+      "loss": 0.6546,
+      "step": 3617
+    },
+    {
+      "epoch": 2.89,
+      "learning_rate": 7.043756670224119e-06,
+      "loss": 0.8081,
+      "step": 3618
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 6.990394877267877e-06,
+      "loss": 0.7733,
+      "step": 3619
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 6.9370330843116336e-06,
+      "loss": 0.7214,
+      "step": 3620
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 6.88367129135539e-06,
+      "loss": 0.7072,
+      "step": 3621
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 6.830309498399147e-06,
+      "loss": 0.3598,
+      "step": 3622
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 6.7769477054429025e-06,
+      "loss": 0.6518,
+      "step": 3623
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 6.72358591248666e-06,
+      "loss": 0.675,
+      "step": 3624
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 6.670224119530416e-06,
+      "loss": 0.7755,
+      "step": 3625
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 6.616862326574173e-06,
+      "loss": 0.6919,
+      "step": 3626
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 6.563500533617929e-06,
+      "loss": 0.6448,
+      "step": 3627
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 6.510138740661686e-06,
+      "loss": 0.6181,
+      "step": 3628
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 6.456776947705444e-06,
+      "loss": 0.721,
+      "step": 3629
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 6.4034151547492e-06,
+      "loss": 0.7536,
+      "step": 3630
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 6.350053361792957e-06,
+      "loss": 0.7881,
+      "step": 3631
+    },
+    {
+      "epoch": 2.91,
+      "learning_rate": 6.296691568836713e-06,
+      "loss": 0.6321,
+      "step": 3632
+    },
+    {
+      "epoch": 2.91,
+      "learning_rate": 6.2433297758804705e-06,
+      "loss": 0.7289,
+      "step": 3633
+    },
+    {
+      "epoch": 2.91,
+      "learning_rate": 6.189967982924227e-06,
+      "loss": 0.7685,
+      "step": 3634
+    },
+    {
+      "epoch": 2.91,
+      "learning_rate": 6.1366061899679835e-06,
+      "loss": 0.681,
+      "step": 3635
+    },
+    {
+      "epoch": 2.91,
+      "learning_rate": 6.0832443970117395e-06,
+      "loss": 0.6585,
+      "step": 3636
+    },
+    {
+      "epoch": 2.91,
+      "learning_rate": 6.029882604055496e-06,
+      "loss": 0.3611,
+      "step": 3637
+    },
+    {
+      "epoch": 2.91,
+      "learning_rate": 5.976520811099253e-06,
+      "loss": 0.625,
+      "step": 3638
+    },
+    {
+      "epoch": 2.91,
+      "learning_rate": 5.923159018143009e-06,
+      "loss": 0.8343,
+      "step": 3639
+    },
+    {
+      "epoch": 2.91,
+      "learning_rate": 5.869797225186766e-06,
+      "loss": 0.9094,
+      "step": 3640
+    },
+    {
+      "epoch": 2.91,
+      "learning_rate": 5.816435432230524e-06,
+      "loss": 0.309,
+      "step": 3641
+    },
+    {
+      "epoch": 2.91,
+      "learning_rate": 5.76307363927428e-06,
+      "loss": 0.592,
+      "step": 3642
+    },
+    {
+      "epoch": 2.91,
+      "learning_rate": 5.709711846318037e-06,
+      "loss": 0.7147,
+      "step": 3643
+    },
+    {
+      "epoch": 2.92,
+      "learning_rate": 5.656350053361793e-06,
+      "loss": 0.6359,
+      "step": 3644
+    },
+    {
+      "epoch": 2.92,
+      "learning_rate": 5.60298826040555e-06,
+      "loss": 0.6968,
+      "step": 3645
+    },
+    {
+      "epoch": 2.92,
+      "learning_rate": 5.549626467449307e-06,
+      "loss": 0.7086,
+      "step": 3646
+    },
+    {
+      "epoch": 2.92,
+      "learning_rate": 5.496264674493063e-06,
+      "loss": 0.7549,
+      "step": 3647
+    },
+    {
+      "epoch": 2.92,
+      "learning_rate": 5.44290288153682e-06,
+      "loss": 0.5961,
+      "step": 3648
+    },
+    {
+      "epoch": 2.92,
+      "learning_rate": 5.3895410885805765e-06,
+      "loss": 0.6684,
+      "step": 3649
+    },
+    {
+      "epoch": 2.92,
+      "learning_rate": 5.336179295624333e-06,
+      "loss": 0.6797,
+      "step": 3650
+    },
+    {
+      "epoch": 2.92,
+      "learning_rate": 5.28281750266809e-06,
+      "loss": 0.7711,
+      "step": 3651
+    },
+    {
+      "epoch": 2.92,
+      "learning_rate": 5.229455709711846e-06,
+      "loss": 0.3264,
+      "step": 3652
+    },
+    {
+      "epoch": 2.92,
+      "learning_rate": 5.176093916755603e-06,
+      "loss": 0.7506,
+      "step": 3653
+    },
+    {
+      "epoch": 2.92,
+      "learning_rate": 5.12273212379936e-06,
+      "loss": 0.5946,
+      "step": 3654
+    },
+    {
+      "epoch": 2.92,
+      "learning_rate": 5.069370330843116e-06,
+      "loss": 0.7703,
+      "step": 3655
+    },
+    {
+      "epoch": 2.92,
+      "learning_rate": 5.016008537886873e-06,
+      "loss": 0.6622,
+      "step": 3656
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 4.96264674493063e-06,
+      "loss": 0.6557,
+      "step": 3657
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 4.909284951974386e-06,
+      "loss": 0.4071,
+      "step": 3658
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 4.855923159018144e-06,
+      "loss": 0.6558,
+      "step": 3659
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 4.8025613660619e-06,
+      "loss": 0.6523,
+      "step": 3660
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 4.749199573105657e-06,
+      "loss": 0.3385,
+      "step": 3661
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 4.6958377801494135e-06,
+      "loss": 0.6412,
+      "step": 3662
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 4.6424759871931695e-06,
+      "loss": 0.7749,
+      "step": 3663
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 4.589114194236926e-06,
+      "loss": 0.7188,
+      "step": 3664
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 4.535752401280683e-06,
+      "loss": 0.6646,
+      "step": 3665
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 4.482390608324439e-06,
+      "loss": 0.2917,
+      "step": 3666
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 4.429028815368197e-06,
+      "loss": 0.5845,
+      "step": 3667
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 4.375667022411953e-06,
+      "loss": 0.7293,
+      "step": 3668
+    },
+    {
+      "epoch": 2.94,
+      "learning_rate": 4.32230522945571e-06,
+      "loss": 0.662,
+      "step": 3669
+    },
+    {
+      "epoch": 2.94,
+      "learning_rate": 4.268943436499467e-06,
+      "loss": 0.7531,
+      "step": 3670
+    },
+    {
+      "epoch": 2.94,
+      "learning_rate": 4.215581643543223e-06,
+      "loss": 0.6364,
+      "step": 3671
+    },
+    {
+      "epoch": 2.94,
+      "learning_rate": 4.16221985058698e-06,
+      "loss": 0.3857,
+      "step": 3672
+    },
+    {
+      "epoch": 2.94,
+      "learning_rate": 4.108858057630737e-06,
+      "loss": 0.6462,
+      "step": 3673
+    },
+    {
+      "epoch": 2.94,
+      "learning_rate": 4.055496264674493e-06,
+      "loss": 0.5848,
+      "step": 3674
+    },
+    {
+      "epoch": 2.94,
+      "learning_rate": 4.00213447171825e-06,
+      "loss": 0.5229,
+      "step": 3675
+    },
+    {
+      "epoch": 2.94,
+      "learning_rate": 3.9487726787620065e-06,
+      "loss": 0.7418,
+      "step": 3676
+    },
+    {
+      "epoch": 2.94,
+      "learning_rate": 3.895410885805763e-06,
+      "loss": 0.6992,
+      "step": 3677
+    },
+    {
+      "epoch": 2.94,
+      "learning_rate": 3.84204909284952e-06,
+      "loss": 0.9014,
+      "step": 3678
+    },
+    {
+      "epoch": 2.94,
+      "learning_rate": 3.7886872998932767e-06,
+      "loss": 0.7162,
+      "step": 3679
+    },
+    {
+      "epoch": 2.94,
+      "learning_rate": 3.735325506937033e-06,
+      "loss": 0.7478,
+      "step": 3680
+    },
+    {
+      "epoch": 2.94,
+      "learning_rate": 3.6819637139807897e-06,
+      "loss": 0.6323,
+      "step": 3681
+    },
+    {
+      "epoch": 2.95,
+      "learning_rate": 3.6286019210245465e-06,
+      "loss": 0.6609,
+      "step": 3682
+    },
+    {
+      "epoch": 2.95,
+      "learning_rate": 3.575240128068303e-06,
+      "loss": 0.5739,
+      "step": 3683
+    },
+    {
+      "epoch": 2.95,
+      "learning_rate": 3.5218783351120595e-06,
+      "loss": 0.6305,
+      "step": 3684
+    },
+    {
+      "epoch": 2.95,
+      "learning_rate": 3.4685165421558168e-06,
+      "loss": 0.3154,
+      "step": 3685
+    },
+    {
+      "epoch": 2.95,
+      "learning_rate": 3.4151547491995737e-06,
+      "loss": 0.7406,
+      "step": 3686
+    },
+    {
+      "epoch": 2.95,
+      "learning_rate": 3.36179295624333e-06,
+      "loss": 0.598,
+      "step": 3687
+    },
+    {
+      "epoch": 2.95,
+      "learning_rate": 3.3084311632870866e-06,
+      "loss": 0.6714,
+      "step": 3688
+    },
+    {
+      "epoch": 2.95,
+      "learning_rate": 3.255069370330843e-06,
+      "loss": 0.7479,
+      "step": 3689
+    },
+    {
+      "epoch": 2.95,
+      "learning_rate": 3.2017075773746e-06,
+      "loss": 0.7551,
+      "step": 3690
+    },
+    {
+      "epoch": 2.95,
+      "learning_rate": 3.1483457844183564e-06,
+      "loss": 0.8139,
+      "step": 3691
+    },
+    {
+      "epoch": 2.95,
+      "learning_rate": 3.0949839914621133e-06,
+      "loss": 0.7379,
+      "step": 3692
+    },
+    {
+      "epoch": 2.95,
+      "learning_rate": 3.0416221985058698e-06,
+      "loss": 0.6276,
+      "step": 3693
+    },
+    {
+      "epoch": 2.96,
+      "learning_rate": 2.9882604055496266e-06,
+      "loss": 0.0,
+      "step": 3694
+    },
+    {
+      "epoch": 2.96,
+      "learning_rate": 2.934898612593383e-06,
+      "loss": 0.7383,
+      "step": 3695
+    },
+    {
+      "epoch": 2.96,
+      "learning_rate": 2.88153681963714e-06,
+      "loss": 0.6868,
+      "step": 3696
+    },
+    {
+      "epoch": 2.96,
+      "learning_rate": 2.8281750266808965e-06,
+      "loss": 0.801,
+      "step": 3697
+    },
+    {
+      "epoch": 2.96,
+      "learning_rate": 2.7748132337246533e-06,
+      "loss": 0.7992,
+      "step": 3698
+    },
+    {
+      "epoch": 2.96,
+      "learning_rate": 2.72145144076841e-06,
+      "loss": 0.636,
+      "step": 3699
+    },
+    {
+      "epoch": 2.96,
+      "learning_rate": 2.6680896478121667e-06,
+      "loss": 0.3689,
+      "step": 3700
+    },
+    {
+      "epoch": 2.96,
+      "learning_rate": 2.614727854855923e-06,
+      "loss": 0.7806,
+      "step": 3701
+    },
+    {
+      "epoch": 2.96,
+      "learning_rate": 2.56136606189968e-06,
+      "loss": 0.6775,
+      "step": 3702
+    },
+    {
+      "epoch": 2.96,
+      "learning_rate": 2.5080042689434365e-06,
+      "loss": 0.7561,
+      "step": 3703
+    },
+    {
+      "epoch": 2.96,
+      "learning_rate": 2.454642475987193e-06,
+      "loss": 0.8101,
+      "step": 3704
+    },
+    {
+      "epoch": 2.96,
+      "learning_rate": 2.40128068303095e-06,
+      "loss": 0.7429,
+      "step": 3705
+    },
+    {
+      "epoch": 2.96,
+      "learning_rate": 2.3479188900747067e-06,
+      "loss": 0.6413,
+      "step": 3706
+    },
+    {
+      "epoch": 2.97,
+      "learning_rate": 2.294557097118463e-06,
+      "loss": 0.7469,
+      "step": 3707
+    },
+    {
+      "epoch": 2.97,
+      "learning_rate": 2.2411953041622197e-06,
+      "loss": 0.778,
+      "step": 3708
+    },
+    {
+      "epoch": 2.97,
+      "learning_rate": 2.1878335112059766e-06,
+      "loss": 0.7112,
+      "step": 3709
+    },
+    {
+      "epoch": 2.97,
+      "learning_rate": 2.1344717182497334e-06,
+      "loss": 0.6584,
+      "step": 3710
+    },
+    {
+      "epoch": 2.97,
+      "learning_rate": 2.08110992529349e-06,
+      "loss": 0.7254,
+      "step": 3711
+    },
+    {
+      "epoch": 2.97,
+      "learning_rate": 2.0277481323372464e-06,
+      "loss": 0.321,
+      "step": 3712
+    },
+    {
+      "epoch": 2.97,
+      "learning_rate": 1.9743863393810032e-06,
+      "loss": 0.4061,
+      "step": 3713
+    },
+    {
+      "epoch": 2.97,
+      "learning_rate": 1.92102454642476e-06,
+      "loss": 0.2926,
+      "step": 3714
+    },
+    {
+      "epoch": 2.97,
+      "learning_rate": 1.8676627534685166e-06,
+      "loss": 0.7069,
+      "step": 3715
+    },
+    {
+      "epoch": 2.97,
+      "learning_rate": 1.8143009605122733e-06,
+      "loss": 0.7413,
+      "step": 3716
+    },
+    {
+      "epoch": 2.97,
+      "learning_rate": 1.7609391675560297e-06,
+      "loss": 0.6506,
+      "step": 3717
+    },
+    {
+      "epoch": 2.97,
+      "learning_rate": 1.7075773745997868e-06,
+      "loss": 0.7362,
+      "step": 3718
+    },
+    {
+      "epoch": 2.98,
+      "learning_rate": 1.6542155816435433e-06,
+      "loss": 0.697,
+      "step": 3719
+    },
+    {
+      "epoch": 2.98,
+      "learning_rate": 1.6008537886873e-06,
+      "loss": 0.595,
+      "step": 3720
+    },
+    {
+      "epoch": 2.98,
+      "learning_rate": 1.5474919957310566e-06,
+      "loss": 0.7717,
+      "step": 3721
+    },
+    {
+      "epoch": 2.98,
+      "learning_rate": 1.4941302027748133e-06,
+      "loss": 0.7109,
+      "step": 3722
+    },
+    {
+      "epoch": 2.98,
+      "learning_rate": 1.44076840981857e-06,
+      "loss": 0.6937,
+      "step": 3723
+    },
+    {
+      "epoch": 2.98,
+      "learning_rate": 1.3874066168623267e-06,
+      "loss": 0.7424,
+      "step": 3724
+    },
+    {
+      "epoch": 2.98,
+      "learning_rate": 1.3340448239060833e-06,
+      "loss": 0.7066,
+      "step": 3725
+    },
+    {
+      "epoch": 2.98,
+      "learning_rate": 1.28068303094984e-06,
+      "loss": 0.6957,
+      "step": 3726
+    },
+    {
+      "epoch": 2.98,
+      "learning_rate": 1.2273212379935965e-06,
+      "loss": 0.7353,
+      "step": 3727
+    },
+    {
+      "epoch": 2.98,
+      "learning_rate": 1.1739594450373534e-06,
+      "loss": 0.6795,
+      "step": 3728
+    },
+    {
+      "epoch": 2.98,
+      "learning_rate": 1.1205976520811098e-06,
+      "loss": 0.7333,
+      "step": 3729
+    },
+    {
+      "epoch": 2.98,
+      "learning_rate": 1.0672358591248667e-06,
+      "loss": 0.6962,
+      "step": 3730
+    },
+    {
+      "epoch": 2.98,
+      "learning_rate": 1.0138740661686232e-06,
+      "loss": 0.8381,
+      "step": 3731
+    },
+    {
+      "epoch": 2.99,
+      "learning_rate": 9.6051227321238e-07,
+      "loss": 0.8629,
+      "step": 3732
+    },
+    {
+      "epoch": 2.99,
+      "learning_rate": 9.071504802561366e-07,
+      "loss": 0.7228,
+      "step": 3733
+    },
+    {
+      "epoch": 2.99,
+      "learning_rate": 8.537886872998934e-07,
+      "loss": 0.4289,
+      "step": 3734
+    },
+    {
+      "epoch": 2.99,
+      "learning_rate": 8.0042689434365e-07,
+      "loss": 0.649,
+      "step": 3735
+    },
+    {
+      "epoch": 2.99,
+      "learning_rate": 7.470651013874067e-07,
+      "loss": 0.3162,
+      "step": 3736
+    },
+    {
+      "epoch": 2.99,
+      "learning_rate": 6.937033084311633e-07,
+      "loss": 0.2608,
+      "step": 3737
+    },
+    {
+      "epoch": 2.99,
+      "learning_rate": 6.4034151547492e-07,
+      "loss": 0.6931,
+      "step": 3738
+    },
+    {
+      "epoch": 2.99,
+      "learning_rate": 5.869797225186767e-07,
+      "loss": 0.6458,
+      "step": 3739
+    },
+    {
+      "epoch": 2.99,
+      "learning_rate": 5.336179295624334e-07,
+      "loss": 0.559,
+      "step": 3740
+    },
+    {
+      "epoch": 2.99,
+      "learning_rate": 4.8025613660619e-07,
+      "loss": 0.7785,
+      "step": 3741
+    },
+    {
+      "epoch": 2.99,
+      "learning_rate": 4.268943436499467e-07,
+      "loss": 0.7793,
+      "step": 3742
+    },
+    {
+      "epoch": 2.99,
+      "learning_rate": 3.7353255069370333e-07,
+      "loss": 0.6703,
+      "step": 3743
+    },
+    {
+      "epoch": 3.0,
+      "learning_rate": 3.2017075773746e-07,
+      "loss": 0.6959,
+      "step": 3744
+    },
+    {
+      "epoch": 3.0,
+      "eval_loss": 1.0379732847213745,
+      "eval_runtime": 7471.6029,
+      "eval_samples_per_second": 1.338,
+      "eval_steps_per_second": 0.084,
+      "step": 3744
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 3750,
+  "num_train_epochs": 3,
+  "save_steps": 156,
+  "total_flos": 2.5272438366142464e+18,
+  "trial_name": null,
+  "trial_params": null
+}