diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,77028 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.9660032712383503,
+  "eval_steps": 1000,
+  "global_step": 22000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 4.39092396017432e-05,
+      "grad_norm": 350.0,
+      "learning_rate": 7.246376811594203e-07,
+      "loss": 6.5493,
+      "step": 1
+    },
+    {
+      "epoch": 8.78184792034864e-05,
+      "grad_norm": 318.0,
+      "learning_rate": 1.4492753623188406e-06,
+      "loss": 6.505,
+      "step": 2
+    },
+    {
+      "epoch": 0.0001756369584069728,
+      "grad_norm": 366.0,
+      "learning_rate": 2.898550724637681e-06,
+      "loss": 6.4657,
+      "step": 4
+    },
+    {
+      "epoch": 0.00026345543761045916,
+      "grad_norm": 390.0,
+      "learning_rate": 4.347826086956522e-06,
+      "loss": 6.1079,
+      "step": 6
+    },
+    {
+      "epoch": 0.0003512739168139456,
+      "grad_norm": 300.0,
+      "learning_rate": 5.797101449275362e-06,
+      "loss": 5.4335,
+      "step": 8
+    },
+    {
+      "epoch": 0.000439092396017432,
+      "grad_norm": 151.0,
+      "learning_rate": 7.246376811594203e-06,
+      "loss": 4.4914,
+      "step": 10
+    },
+    {
+      "epoch": 0.0005269108752209183,
+      "grad_norm": 85.0,
+      "learning_rate": 8.695652173913044e-06,
+      "loss": 3.9792,
+      "step": 12
+    },
+    {
+      "epoch": 0.0006147293544244047,
+      "grad_norm": 15.0,
+      "learning_rate": 1.0144927536231885e-05,
+      "loss": 3.7765,
+      "step": 14
+    },
+    {
+      "epoch": 0.0007025478336278912,
+      "grad_norm": 9.5625,
+      "learning_rate": 1.1594202898550725e-05,
+      "loss": 3.7259,
+      "step": 16
+    },
+    {
+      "epoch": 0.0007903663128313776,
+      "grad_norm": 7.25,
+      "learning_rate": 1.3043478260869566e-05,
+      "loss": 3.657,
+      "step": 18
+    },
+    {
+      "epoch": 0.000878184792034864,
+      "grad_norm": 16.25,
+      "learning_rate": 1.4492753623188407e-05,
+      "loss": 3.6366,
+      "step": 20
+    },
+    {
+      "epoch": 0.0009660032712383504,
+      "grad_norm": 8.0,
+      "learning_rate": 1.5942028985507246e-05,
+      "loss": 3.5582,
+      "step": 22
+    },
+    {
+      "epoch": 0.0010538217504418366,
+      "grad_norm": 10.8125,
+      "learning_rate": 1.739130434782609e-05,
+      "loss": 3.4616,
+      "step": 24
+    },
+    {
+      "epoch": 0.0011416402296453231,
+      "grad_norm": 6.65625,
+      "learning_rate": 1.8840579710144928e-05,
+      "loss": 3.3355,
+      "step": 26
+    },
+    {
+      "epoch": 0.0012294587088488094,
+      "grad_norm": 6.21875,
+      "learning_rate": 2.028985507246377e-05,
+      "loss": 3.2562,
+      "step": 28
+    },
+    {
+      "epoch": 0.001317277188052296,
+      "grad_norm": 5.75,
+      "learning_rate": 2.173913043478261e-05,
+      "loss": 3.1432,
+      "step": 30
+    },
+    {
+      "epoch": 0.0014050956672557824,
+      "grad_norm": 7.15625,
+      "learning_rate": 2.318840579710145e-05,
+      "loss": 3.0361,
+      "step": 32
+    },
+    {
+      "epoch": 0.0014929141464592687,
+      "grad_norm": 8.25,
+      "learning_rate": 2.4637681159420292e-05,
+      "loss": 2.9816,
+      "step": 34
+    },
+    {
+      "epoch": 0.0015807326256627552,
+      "grad_norm": 7.4375,
+      "learning_rate": 2.608695652173913e-05,
+      "loss": 2.8625,
+      "step": 36
+    },
+    {
+      "epoch": 0.0016685511048662414,
+      "grad_norm": 6.03125,
+      "learning_rate": 2.753623188405797e-05,
+      "loss": 2.712,
+      "step": 38
+    },
+    {
+      "epoch": 0.001756369584069728,
+      "grad_norm": 9.375,
+      "learning_rate": 2.8985507246376814e-05,
+      "loss": 2.5506,
+      "step": 40
+    },
+    {
+      "epoch": 0.0018441880632732142,
+      "grad_norm": 8.8125,
+      "learning_rate": 3.0434782608695656e-05,
+      "loss": 2.4374,
+      "step": 42
+    },
+    {
+      "epoch": 0.0019320065424767007,
+      "grad_norm": 16.625,
+      "learning_rate": 3.188405797101449e-05,
+      "loss": 2.3639,
+      "step": 44
+    },
+    {
+      "epoch": 0.002019825021680187,
+      "grad_norm": 15.1875,
+      "learning_rate": 3.3333333333333335e-05,
+      "loss": 2.2934,
+      "step": 46
+    },
+    {
+      "epoch": 0.0021076435008836733,
+      "grad_norm": 9.375,
+      "learning_rate": 3.478260869565218e-05,
+      "loss": 2.2077,
+      "step": 48
+    },
+    {
+      "epoch": 0.00219546198008716,
+      "grad_norm": 9.875,
+      "learning_rate": 3.6231884057971014e-05,
+      "loss": 2.1561,
+      "step": 50
+    },
+    {
+      "epoch": 0.0022832804592906463,
+      "grad_norm": 13.0,
+      "learning_rate": 3.7681159420289856e-05,
+      "loss": 2.1002,
+      "step": 52
+    },
+    {
+      "epoch": 0.0023710989384941325,
+      "grad_norm": 9.3125,
+      "learning_rate": 3.91304347826087e-05,
+      "loss": 2.0411,
+      "step": 54
+    },
+    {
+      "epoch": 0.002458917417697619,
+      "grad_norm": 9.0625,
+      "learning_rate": 4.057971014492754e-05,
+      "loss": 2.0039,
+      "step": 56
+    },
+    {
+      "epoch": 0.0025467358969011055,
+      "grad_norm": 8.0,
+      "learning_rate": 4.202898550724638e-05,
+      "loss": 2.0052,
+      "step": 58
+    },
+    {
+      "epoch": 0.002634554376104592,
+      "grad_norm": 8.625,
+      "learning_rate": 4.347826086956522e-05,
+      "loss": 1.9758,
+      "step": 60
+    },
+    {
+      "epoch": 0.002722372855308078,
+      "grad_norm": 13.125,
+      "learning_rate": 4.492753623188406e-05,
+      "loss": 1.9008,
+      "step": 62
+    },
+    {
+      "epoch": 0.002810191334511565,
+      "grad_norm": 6.6875,
+      "learning_rate": 4.63768115942029e-05,
+      "loss": 1.8888,
+      "step": 64
+    },
+    {
+      "epoch": 0.002898009813715051,
+      "grad_norm": 11.1875,
+      "learning_rate": 4.782608695652174e-05,
+      "loss": 1.8357,
+      "step": 66
+    },
+    {
+      "epoch": 0.0029858282929185373,
+      "grad_norm": 29.625,
+      "learning_rate": 4.9275362318840584e-05,
+      "loss": 1.9061,
+      "step": 68
+    },
+    {
+      "epoch": 0.0030736467721220236,
+      "grad_norm": 31.125,
+      "learning_rate": 4.9999999760686754e-05,
+      "loss": 1.8563,
+      "step": 70
+    },
+    {
+      "epoch": 0.0031614652513255103,
+      "grad_norm": 13.0625,
+      "learning_rate": 4.999999784618078e-05,
+      "loss": 1.8569,
+      "step": 72
+    },
+    {
+      "epoch": 0.0032492837305289966,
+      "grad_norm": 8.5625,
+      "learning_rate": 4.999999401716897e-05,
+      "loss": 1.8224,
+      "step": 74
+    },
+    {
+      "epoch": 0.003337102209732483,
+      "grad_norm": 8.6875,
+      "learning_rate": 4.9999988273651635e-05,
+      "loss": 1.8069,
+      "step": 76
+    },
+    {
+      "epoch": 0.003424920688935969,
+      "grad_norm": 10.0625,
+      "learning_rate": 4.9999980615629205e-05,
+      "loss": 1.7401,
+      "step": 78
+    },
+    {
+      "epoch": 0.003512739168139456,
+      "grad_norm": 6.75,
+      "learning_rate": 4.9999971043102264e-05,
+      "loss": 1.7421,
+      "step": 80
+    },
+    {
+      "epoch": 0.003600557647342942,
+      "grad_norm": 6.65625,
+      "learning_rate": 4.999995955607155e-05,
+      "loss": 1.7189,
+      "step": 82
+    },
+    {
+      "epoch": 0.0036883761265464284,
+      "grad_norm": 7.125,
+      "learning_rate": 4.999994615453794e-05,
+      "loss": 1.714,
+      "step": 84
+    },
+    {
+      "epoch": 0.0037761946057499147,
+      "grad_norm": 11.25,
+      "learning_rate": 4.9999930838502456e-05,
+      "loss": 1.6762,
+      "step": 86
+    },
+    {
+      "epoch": 0.0038640130849534014,
+      "grad_norm": 12.25,
+      "learning_rate": 4.999991360796628e-05,
+      "loss": 1.6994,
+      "step": 88
+    },
+    {
+      "epoch": 0.003951831564156888,
+      "grad_norm": 7.78125,
+      "learning_rate": 4.9999894462930725e-05,
+      "loss": 1.6493,
+      "step": 90
+    },
+    {
+      "epoch": 0.004039650043360374,
+      "grad_norm": 6.9375,
+      "learning_rate": 4.9999873403397254e-05,
+      "loss": 1.6475,
+      "step": 92
+    },
+    {
+      "epoch": 0.004127468522563861,
+      "grad_norm": 12.8125,
+      "learning_rate": 4.99998504293675e-05,
+      "loss": 1.6591,
+      "step": 94
+    },
+    {
+      "epoch": 0.0042152870017673465,
+      "grad_norm": 11.25,
+      "learning_rate": 4.999982554084319e-05,
+      "loss": 1.6258,
+      "step": 96
+    },
+    {
+      "epoch": 0.004303105480970833,
+      "grad_norm": 12.125,
+      "learning_rate": 4.999979873782625e-05,
+      "loss": 1.6365,
+      "step": 98
+    },
+    {
+      "epoch": 0.00439092396017432,
+      "grad_norm": 10.9375,
+      "learning_rate": 4.999977002031874e-05,
+      "loss": 1.6345,
+      "step": 100
+    },
+    {
+      "epoch": 0.004478742439377806,
+      "grad_norm": 8.9375,
+      "learning_rate": 4.999973938832284e-05,
+      "loss": 1.6242,
+      "step": 102
+    },
+    {
+      "epoch": 0.0045665609185812925,
+      "grad_norm": 9.9375,
+      "learning_rate": 4.999970684184091e-05,
+      "loss": 1.6097,
+      "step": 104
+    },
+    {
+      "epoch": 0.004654379397784779,
+      "grad_norm": 13.6875,
+      "learning_rate": 4.999967238087544e-05,
+      "loss": 1.5788,
+      "step": 106
+    },
+    {
+      "epoch": 0.004742197876988265,
+      "grad_norm": 9.125,
+      "learning_rate": 4.999963600542906e-05,
+      "loss": 1.5573,
+      "step": 108
+    },
+    {
+      "epoch": 0.004830016356191752,
+      "grad_norm": 7.46875,
+      "learning_rate": 4.999959771550457e-05,
+      "loss": 1.574,
+      "step": 110
+    },
+    {
+      "epoch": 0.004917834835395238,
+      "grad_norm": 8.9375,
+      "learning_rate": 4.999955751110488e-05,
+      "loss": 1.5836,
+      "step": 112
+    },
+    {
+      "epoch": 0.005005653314598724,
+      "grad_norm": 14.625,
+      "learning_rate": 4.999951539223309e-05,
+      "loss": 1.5808,
+      "step": 114
+    },
+    {
+      "epoch": 0.005093471793802211,
+      "grad_norm": 10.625,
+      "learning_rate": 4.999947135889242e-05,
+      "loss": 1.569,
+      "step": 116
+    },
+    {
+      "epoch": 0.005181290273005697,
+      "grad_norm": 8.1875,
+      "learning_rate": 4.999942541108624e-05,
+      "loss": 1.5499,
+      "step": 118
+    },
+    {
+      "epoch": 0.005269108752209184,
+      "grad_norm": 8.375,
+      "learning_rate": 4.999937754881807e-05,
+      "loss": 1.5262,
+      "step": 120
+    },
+    {
+      "epoch": 0.00535692723141267,
+      "grad_norm": 8.625,
+      "learning_rate": 4.999932777209157e-05,
+      "loss": 1.5405,
+      "step": 122
+    },
+    {
+      "epoch": 0.005444745710616156,
+      "grad_norm": 9.5625,
+      "learning_rate": 4.999927608091056e-05,
+      "loss": 1.5341,
+      "step": 124
+    },
+    {
+      "epoch": 0.005532564189819643,
+      "grad_norm": 9.9375,
+      "learning_rate": 4.9999222475279003e-05,
+      "loss": 1.5037,
+      "step": 126
+    },
+    {
+      "epoch": 0.00562038266902313,
+      "grad_norm": 7.75,
+      "learning_rate": 4.999916695520098e-05,
+      "loss": 1.5074,
+      "step": 128
+    },
+    {
+      "epoch": 0.005708201148226615,
+      "grad_norm": 9.25,
+      "learning_rate": 4.999910952068077e-05,
+      "loss": 1.5116,
+      "step": 130
+    },
+    {
+      "epoch": 0.005796019627430102,
+      "grad_norm": 6.71875,
+      "learning_rate": 4.999905017172276e-05,
+      "loss": 1.5003,
+      "step": 132
+    },
+    {
+      "epoch": 0.005883838106633588,
+      "grad_norm": 10.5,
+      "learning_rate": 4.999898890833149e-05,
+      "loss": 1.5161,
+      "step": 134
+    },
+    {
+      "epoch": 0.005971656585837075,
+      "grad_norm": 10.125,
+      "learning_rate": 4.999892573051166e-05,
+      "loss": 1.477,
+      "step": 136
+    },
+    {
+      "epoch": 0.006059475065040561,
+      "grad_norm": 10.5,
+      "learning_rate": 4.999886063826811e-05,
+      "loss": 1.4996,
+      "step": 138
+    },
+    {
+      "epoch": 0.006147293544244047,
+      "grad_norm": 10.625,
+      "learning_rate": 4.999879363160581e-05,
+      "loss": 1.4417,
+      "step": 140
+    },
+    {
+      "epoch": 0.006235112023447534,
+      "grad_norm": 13.375,
+      "learning_rate": 4.999872471052991e-05,
+      "loss": 1.457,
+      "step": 142
+    },
+    {
+      "epoch": 0.006322930502651021,
+      "grad_norm": 10.3125,
+      "learning_rate": 4.9998653875045666e-05,
+      "loss": 1.4707,
+      "step": 144
+    },
+    {
+      "epoch": 0.0064107489818545065,
+      "grad_norm": 10.0625,
+      "learning_rate": 4.999858112515853e-05,
+      "loss": 1.4844,
+      "step": 146
+    },
+    {
+      "epoch": 0.006498567461057993,
+      "grad_norm": 11.8125,
+      "learning_rate": 4.999850646087405e-05,
+      "loss": 1.4749,
+      "step": 148
+    },
+    {
+      "epoch": 0.00658638594026148,
+      "grad_norm": 9.125,
+      "learning_rate": 4.999842988219795e-05,
+      "loss": 1.4431,
+      "step": 150
+    },
+    {
+      "epoch": 0.006674204419464966,
+      "grad_norm": 8.1875,
+      "learning_rate": 4.99983513891361e-05,
+      "loss": 1.4359,
+      "step": 152
+    },
+    {
+      "epoch": 0.0067620228986684525,
+      "grad_norm": 6.53125,
+      "learning_rate": 4.999827098169452e-05,
+      "loss": 1.4179,
+      "step": 154
+    },
+    {
+      "epoch": 0.006849841377871938,
+      "grad_norm": 7.9375,
+      "learning_rate": 4.999818865987934e-05,
+      "loss": 1.4185,
+      "step": 156
+    },
+    {
+      "epoch": 0.006937659857075425,
+      "grad_norm": 9.1875,
+      "learning_rate": 4.999810442369689e-05,
+      "loss": 1.449,
+      "step": 158
+    },
+    {
+      "epoch": 0.007025478336278912,
+      "grad_norm": 11.5625,
+      "learning_rate": 4.9998018273153605e-05,
+      "loss": 1.416,
+      "step": 160
+    },
+    {
+      "epoch": 0.007113296815482398,
+      "grad_norm": 9.375,
+      "learning_rate": 4.999793020825609e-05,
+      "loss": 1.3884,
+      "step": 162
+    },
+    {
+      "epoch": 0.007201115294685884,
+      "grad_norm": 5.8125,
+      "learning_rate": 4.9997840229011085e-05,
+      "loss": 1.3853,
+      "step": 164
+    },
+    {
+      "epoch": 0.007288933773889371,
+      "grad_norm": 7.1875,
+      "learning_rate": 4.999774833542549e-05,
+      "loss": 1.4316,
+      "step": 166
+    },
+    {
+      "epoch": 0.007376752253092857,
+      "grad_norm": 9.5,
+      "learning_rate": 4.999765452750633e-05,
+      "loss": 1.3818,
+      "step": 168
+    },
+    {
+      "epoch": 0.007464570732296344,
+      "grad_norm": 7.28125,
+      "learning_rate": 4.999755880526079e-05,
+      "loss": 1.3969,
+      "step": 170
+    },
+    {
+      "epoch": 0.007552389211499829,
+      "grad_norm": 7.9375,
+      "learning_rate": 4.999746116869621e-05,
+      "loss": 1.3647,
+      "step": 172
+    },
+    {
+      "epoch": 0.007640207690703316,
+      "grad_norm": 6.9375,
+      "learning_rate": 4.999736161782006e-05,
+      "loss": 1.3707,
+      "step": 174
+    },
+    {
+      "epoch": 0.007728026169906803,
+      "grad_norm": 7.53125,
+      "learning_rate": 4.9997260152639966e-05,
+      "loss": 1.3951,
+      "step": 176
+    },
+    {
+      "epoch": 0.007815844649110289,
+      "grad_norm": 5.875,
+      "learning_rate": 4.9997156773163694e-05,
+      "loss": 1.3626,
+      "step": 178
+    },
+    {
+      "epoch": 0.007903663128313776,
+      "grad_norm": 7.3125,
+      "learning_rate": 4.999705147939917e-05,
+      "loss": 1.3481,
+      "step": 180
+    },
+    {
+      "epoch": 0.007991481607517262,
+      "grad_norm": 7.03125,
+      "learning_rate": 4.999694427135445e-05,
+      "loss": 1.3754,
+      "step": 182
+    },
+    {
+      "epoch": 0.008079300086720748,
+      "grad_norm": 6.46875,
+      "learning_rate": 4.9996835149037735e-05,
+      "loss": 1.3436,
+      "step": 184
+    },
+    {
+      "epoch": 0.008167118565924234,
+      "grad_norm": 8.625,
+      "learning_rate": 4.999672411245741e-05,
+      "loss": 1.3398,
+      "step": 186
+    },
+    {
+      "epoch": 0.008254937045127721,
+      "grad_norm": 8.125,
+      "learning_rate": 4.999661116162194e-05,
+      "loss": 1.3604,
+      "step": 188
+    },
+    {
+      "epoch": 0.008342755524331207,
+      "grad_norm": 11.875,
+      "learning_rate": 4.9996496296540005e-05,
+      "loss": 1.3354,
+      "step": 190
+    },
+    {
+      "epoch": 0.008430574003534693,
+      "grad_norm": 7.53125,
+      "learning_rate": 4.99963795172204e-05,
+      "loss": 1.3341,
+      "step": 192
+    },
+    {
+      "epoch": 0.00851839248273818,
+      "grad_norm": 6.84375,
+      "learning_rate": 4.999626082367205e-05,
+      "loss": 1.3466,
+      "step": 194
+    },
+    {
+      "epoch": 0.008606210961941667,
+      "grad_norm": 7.8125,
+      "learning_rate": 4.999614021590405e-05,
+      "loss": 1.3208,
+      "step": 196
+    },
+    {
+      "epoch": 0.008694029441145152,
+      "grad_norm": 6.21875,
+      "learning_rate": 4.999601769392565e-05,
+      "loss": 1.3211,
+      "step": 198
+    },
+    {
+      "epoch": 0.00878184792034864,
+      "grad_norm": 6.3125,
+      "learning_rate": 4.999589325774622e-05,
+      "loss": 1.3199,
+      "step": 200
+    },
+    {
+      "epoch": 0.008869666399552126,
+      "grad_norm": 8.875,
+      "learning_rate": 4.99957669073753e-05,
+      "loss": 1.2907,
+      "step": 202
+    },
+    {
+      "epoch": 0.008957484878755612,
+      "grad_norm": 6.03125,
+      "learning_rate": 4.9995638642822536e-05,
+      "loss": 1.3227,
+      "step": 204
+    },
+    {
+      "epoch": 0.0090453033579591,
+      "grad_norm": 10.25,
+      "learning_rate": 4.9995508464097796e-05,
+      "loss": 1.3157,
+      "step": 206
+    },
+    {
+      "epoch": 0.009133121837162585,
+      "grad_norm": 7.625,
+      "learning_rate": 4.999537637121101e-05,
+      "loss": 1.2974,
+      "step": 208
+    },
+    {
+      "epoch": 0.009220940316366071,
+      "grad_norm": 6.4375,
+      "learning_rate": 4.999524236417232e-05,
+      "loss": 1.3103,
+      "step": 210
+    },
+    {
+      "epoch": 0.009308758795569558,
+      "grad_norm": 10.1875,
+      "learning_rate": 4.999510644299198e-05,
+      "loss": 1.3,
+      "step": 212
+    },
+    {
+      "epoch": 0.009396577274773044,
+      "grad_norm": 7.53125,
+      "learning_rate": 4.999496860768039e-05,
+      "loss": 1.2689,
+      "step": 214
+    },
+    {
+      "epoch": 0.00948439575397653,
+      "grad_norm": 6.0625,
+      "learning_rate": 4.999482885824811e-05,
+      "loss": 1.2809,
+      "step": 216
+    },
+    {
+      "epoch": 0.009572214233180018,
+      "grad_norm": 8.125,
+      "learning_rate": 4.9994687194705846e-05,
+      "loss": 1.2937,
+      "step": 218
+    },
+    {
+      "epoch": 0.009660032712383504,
+      "grad_norm": 13.4375,
+      "learning_rate": 4.9994543617064445e-05,
+      "loss": 1.3081,
+      "step": 220
+    },
+    {
+      "epoch": 0.00974785119158699,
+      "grad_norm": 10.5,
+      "learning_rate": 4.999439812533491e-05,
+      "loss": 1.3092,
+      "step": 222
+    },
+    {
+      "epoch": 0.009835669670790475,
+      "grad_norm": 8.375,
+      "learning_rate": 4.9994250719528366e-05,
+      "loss": 1.304,
+      "step": 224
+    },
+    {
+      "epoch": 0.009923488149993963,
+      "grad_norm": 6.53125,
+      "learning_rate": 4.999410139965612e-05,
+      "loss": 1.2551,
+      "step": 226
+    },
+    {
+      "epoch": 0.010011306629197449,
+      "grad_norm": 8.4375,
+      "learning_rate": 4.9993950165729585e-05,
+      "loss": 1.2683,
+      "step": 228
+    },
+    {
+      "epoch": 0.010099125108400935,
+      "grad_norm": 10.5,
+      "learning_rate": 4.9993797017760364e-05,
+      "loss": 1.3054,
+      "step": 230
+    },
+    {
+      "epoch": 0.010186943587604422,
+      "grad_norm": 10.8125,
+      "learning_rate": 4.999364195576017e-05,
+      "loss": 1.2796,
+      "step": 232
+    },
+    {
+      "epoch": 0.010274762066807908,
+      "grad_norm": 13.1875,
+      "learning_rate": 4.999348497974089e-05,
+      "loss": 1.2596,
+      "step": 234
+    },
+    {
+      "epoch": 0.010362580546011394,
+      "grad_norm": 8.9375,
+      "learning_rate": 4.999332608971453e-05,
+      "loss": 1.2413,
+      "step": 236
+    },
+    {
+      "epoch": 0.010450399025214881,
+      "grad_norm": 6.78125,
+      "learning_rate": 4.9993165285693275e-05,
+      "loss": 1.2754,
+      "step": 238
+    },
+    {
+      "epoch": 0.010538217504418367,
+      "grad_norm": 6.75,
+      "learning_rate": 4.999300256768943e-05,
+      "loss": 1.2392,
+      "step": 240
+    },
+    {
+      "epoch": 0.010626035983621853,
+      "grad_norm": 7.4375,
+      "learning_rate": 4.9992837935715455e-05,
+      "loss": 1.2867,
+      "step": 242
+    },
+    {
+      "epoch": 0.01071385446282534,
+      "grad_norm": 8.5625,
+      "learning_rate": 4.999267138978396e-05,
+      "loss": 1.2473,
+      "step": 244
+    },
+    {
+      "epoch": 0.010801672942028826,
+      "grad_norm": 7.75,
+      "learning_rate": 4.9992502929907706e-05,
+      "loss": 1.2448,
+      "step": 246
+    },
+    {
+      "epoch": 0.010889491421232312,
+      "grad_norm": 7.3125,
+      "learning_rate": 4.999233255609957e-05,
+      "loss": 1.2255,
+      "step": 248
+    },
+    {
+      "epoch": 0.0109773099004358,
+      "grad_norm": 8.4375,
+      "learning_rate": 4.999216026837263e-05,
+      "loss": 1.2549,
+      "step": 250
+    },
+    {
+      "epoch": 0.011065128379639286,
+      "grad_norm": 7.71875,
+      "learning_rate": 4.999198606674006e-05,
+      "loss": 1.2665,
+      "step": 252
+    },
+    {
+      "epoch": 0.011152946858842772,
+      "grad_norm": 7.71875,
+      "learning_rate": 4.9991809951215204e-05,
+      "loss": 1.2329,
+      "step": 254
+    },
+    {
+      "epoch": 0.01124076533804626,
+      "grad_norm": 11.0625,
+      "learning_rate": 4.999163192181155e-05,
+      "loss": 1.2516,
+      "step": 256
+    },
+    {
+      "epoch": 0.011328583817249745,
+      "grad_norm": 8.375,
+      "learning_rate": 4.9991451978542744e-05,
+      "loss": 1.2441,
+      "step": 258
+    },
+    {
+      "epoch": 0.01141640229645323,
+      "grad_norm": 7.9375,
+      "learning_rate": 4.999127012142255e-05,
+      "loss": 1.2242,
+      "step": 260
+    },
+    {
+      "epoch": 0.011504220775656717,
+      "grad_norm": 6.34375,
+      "learning_rate": 4.999108635046489e-05,
+      "loss": 1.2447,
+      "step": 262
+    },
+    {
+      "epoch": 0.011592039254860204,
+      "grad_norm": 7.875,
+      "learning_rate": 4.999090066568385e-05,
+      "loss": 1.2263,
+      "step": 264
+    },
+    {
+      "epoch": 0.01167985773406369,
+      "grad_norm": 7.8125,
+      "learning_rate": 4.9990713067093654e-05,
+      "loss": 1.2216,
+      "step": 266
+    },
+    {
+      "epoch": 0.011767676213267176,
+      "grad_norm": 8.5625,
+      "learning_rate": 4.9990523554708655e-05,
+      "loss": 1.2352,
+      "step": 268
+    },
+    {
+      "epoch": 0.011855494692470664,
+      "grad_norm": 6.53125,
+      "learning_rate": 4.999033212854337e-05,
+      "loss": 1.2291,
+      "step": 270
+    },
+    {
+      "epoch": 0.01194331317167415,
+      "grad_norm": 7.8125,
+      "learning_rate": 4.999013878861246e-05,
+      "loss": 1.223,
+      "step": 272
+    },
+    {
+      "epoch": 0.012031131650877635,
+      "grad_norm": 9.5625,
+      "learning_rate": 4.998994353493074e-05,
+      "loss": 1.21,
+      "step": 274
+    },
+    {
+      "epoch": 0.012118950130081123,
+      "grad_norm": 11.5,
+      "learning_rate": 4.998974636751315e-05,
+      "loss": 1.227,
+      "step": 276
+    },
+    {
+      "epoch": 0.012206768609284609,
+      "grad_norm": 9.75,
+      "learning_rate": 4.998954728637478e-05,
+      "loss": 1.2142,
+      "step": 278
+    },
+    {
+      "epoch": 0.012294587088488095,
+      "grad_norm": 7.46875,
+      "learning_rate": 4.9989346291530904e-05,
+      "loss": 1.2299,
+      "step": 280
+    },
+    {
+      "epoch": 0.012382405567691582,
+      "grad_norm": 6.34375,
+      "learning_rate": 4.998914338299689e-05,
+      "loss": 1.2197,
+      "step": 282
+    },
+    {
+      "epoch": 0.012470224046895068,
+      "grad_norm": 7.96875,
+      "learning_rate": 4.998893856078829e-05,
+      "loss": 1.227,
+      "step": 284
+    },
+    {
+      "epoch": 0.012558042526098554,
+      "grad_norm": 11.1875,
+      "learning_rate": 4.998873182492078e-05,
+      "loss": 1.2082,
+      "step": 286
+    },
+    {
+      "epoch": 0.012645861005302041,
+      "grad_norm": 9.9375,
+      "learning_rate": 4.9988523175410204e-05,
+      "loss": 1.2347,
+      "step": 288
+    },
+    {
+      "epoch": 0.012733679484505527,
+      "grad_norm": 8.0625,
+      "learning_rate": 4.9988312612272524e-05,
+      "loss": 1.2427,
+      "step": 290
+    },
+    {
+      "epoch": 0.012821497963709013,
+      "grad_norm": 8.5,
+      "learning_rate": 4.998810013552388e-05,
+      "loss": 1.1799,
+      "step": 292
+    },
+    {
+      "epoch": 0.0129093164429125,
+      "grad_norm": 8.875,
+      "learning_rate": 4.998788574518054e-05,
+      "loss": 1.1828,
+      "step": 294
+    },
+    {
+      "epoch": 0.012997134922115986,
+      "grad_norm": 9.375,
+      "learning_rate": 4.998766944125891e-05,
+      "loss": 1.1921,
+      "step": 296
+    },
+    {
+      "epoch": 0.013084953401319472,
+      "grad_norm": 6.3125,
+      "learning_rate": 4.998745122377557e-05,
+      "loss": 1.1933,
+      "step": 298
+    },
+    {
+      "epoch": 0.01317277188052296,
+      "grad_norm": 6.90625,
+      "learning_rate": 4.998723109274722e-05,
+      "loss": 1.1597,
+      "step": 300
+    },
+    {
+      "epoch": 0.013260590359726446,
+      "grad_norm": 6.71875,
+      "learning_rate": 4.998700904819074e-05,
+      "loss": 1.1891,
+      "step": 302
+    },
+    {
+      "epoch": 0.013348408838929932,
+      "grad_norm": 5.46875,
+      "learning_rate": 4.99867850901231e-05,
+      "loss": 1.1977,
+      "step": 304
+    },
+    {
+      "epoch": 0.013436227318133417,
+      "grad_norm": 7.875,
+      "learning_rate": 4.998655921856147e-05,
+      "loss": 1.1838,
+      "step": 306
+    },
+    {
+      "epoch": 0.013524045797336905,
+      "grad_norm": 5.46875,
+      "learning_rate": 4.9986331433523156e-05,
+      "loss": 1.1977,
+      "step": 308
+    },
+    {
+      "epoch": 0.01361186427654039,
+      "grad_norm": 7.3125,
+      "learning_rate": 4.998610173502558e-05,
+      "loss": 1.1807,
+      "step": 310
+    },
+    {
+      "epoch": 0.013699682755743877,
+      "grad_norm": 5.40625,
+      "learning_rate": 4.9985870123086345e-05,
+      "loss": 1.161,
+      "step": 312
+    },
+    {
+      "epoch": 0.013787501234947364,
+      "grad_norm": 7.8125,
+      "learning_rate": 4.9985636597723195e-05,
+      "loss": 1.1476,
+      "step": 314
+    },
+    {
+      "epoch": 0.01387531971415085,
+      "grad_norm": 5.625,
+      "learning_rate": 4.9985401158954e-05,
+      "loss": 1.1881,
+      "step": 316
+    },
+    {
+      "epoch": 0.013963138193354336,
+      "grad_norm": 7.125,
+      "learning_rate": 4.998516380679679e-05,
+      "loss": 1.1831,
+      "step": 318
+    },
+    {
+      "epoch": 0.014050956672557824,
+      "grad_norm": 9.3125,
+      "learning_rate": 4.998492454126975e-05,
+      "loss": 1.1747,
+      "step": 320
+    },
+    {
+      "epoch": 0.01413877515176131,
+      "grad_norm": 5.5625,
+      "learning_rate": 4.9984683362391204e-05,
+      "loss": 1.1628,
+      "step": 322
+    },
+    {
+      "epoch": 0.014226593630964795,
+      "grad_norm": 7.21875,
+      "learning_rate": 4.998444027017961e-05,
+      "loss": 1.1477,
+      "step": 324
+    },
+    {
+      "epoch": 0.014314412110168283,
+      "grad_norm": 8.625,
+      "learning_rate": 4.9984195264653596e-05,
+      "loss": 1.1537,
+      "step": 326
+    },
+    {
+      "epoch": 0.014402230589371769,
+      "grad_norm": 6.75,
+      "learning_rate": 4.998394834583191e-05,
+      "loss": 1.1832,
+      "step": 328
+    },
+    {
+      "epoch": 0.014490049068575254,
+      "grad_norm": 6.34375,
+      "learning_rate": 4.998369951373348e-05,
+      "loss": 1.1519,
+      "step": 330
+    },
+    {
+      "epoch": 0.014577867547778742,
+      "grad_norm": 6.5,
+      "learning_rate": 4.998344876837735e-05,
+      "loss": 1.1786,
+      "step": 332
+    },
+    {
+      "epoch": 0.014665686026982228,
+      "grad_norm": 7.09375,
+      "learning_rate": 4.998319610978273e-05,
+      "loss": 1.1379,
+      "step": 334
+    },
+    {
+      "epoch": 0.014753504506185714,
+      "grad_norm": 6.5,
+      "learning_rate": 4.998294153796895e-05,
+      "loss": 1.1754,
+      "step": 336
+    },
+    {
+      "epoch": 0.014841322985389201,
+      "grad_norm": 6.5,
+      "learning_rate": 4.998268505295552e-05,
+      "loss": 1.1211,
+      "step": 338
+    },
+    {
+      "epoch": 0.014929141464592687,
+      "grad_norm": 8.5625,
+      "learning_rate": 4.9982426654762095e-05,
+      "loss": 1.1669,
+      "step": 340
+    },
+    {
+      "epoch": 0.015016959943796173,
+      "grad_norm": 7.4375,
+      "learning_rate": 4.998216634340844e-05,
+      "loss": 1.1609,
+      "step": 342
+    },
+    {
+      "epoch": 0.015104778422999659,
+      "grad_norm": 5.625,
+      "learning_rate": 4.998190411891449e-05,
+      "loss": 1.1381,
+      "step": 344
+    },
+    {
+      "epoch": 0.015192596902203146,
+      "grad_norm": 7.09375,
+      "learning_rate": 4.9981639981300344e-05,
+      "loss": 1.1383,
+      "step": 346
+    },
+    {
+      "epoch": 0.015280415381406632,
+      "grad_norm": 8.5625,
+      "learning_rate": 4.998137393058622e-05,
+      "loss": 1.1462,
+      "step": 348
+    },
+    {
+      "epoch": 0.015368233860610118,
+      "grad_norm": 7.0,
+      "learning_rate": 4.9981105966792485e-05,
+      "loss": 1.1386,
+      "step": 350
+    },
+    {
+      "epoch": 0.015456052339813606,
+      "grad_norm": 8.0,
+      "learning_rate": 4.9980836089939665e-05,
+      "loss": 1.1212,
+      "step": 352
+    },
+    {
+      "epoch": 0.015543870819017092,
+      "grad_norm": 6.3125,
+      "learning_rate": 4.998056430004844e-05,
+      "loss": 1.1352,
+      "step": 354
+    },
+    {
+      "epoch": 0.015631689298220577,
+      "grad_norm": 8.8125,
+      "learning_rate": 4.99802905971396e-05,
+      "loss": 1.1435,
+      "step": 356
+    },
+    {
+      "epoch": 0.015719507777424063,
+      "grad_norm": 7.3125,
+      "learning_rate": 4.998001498123413e-05,
+      "loss": 1.1223,
+      "step": 358
+    },
+    {
+      "epoch": 0.015807326256627553,
+      "grad_norm": 5.84375,
+      "learning_rate": 4.9979737452353114e-05,
+      "loss": 1.1079,
+      "step": 360
+    },
+    {
+      "epoch": 0.01589514473583104,
+      "grad_norm": 16.5,
+      "learning_rate": 4.997945801051782e-05,
+      "loss": 1.1403,
+      "step": 362
+    },
+    {
+      "epoch": 0.015982963215034524,
+      "grad_norm": 9.6875,
+      "learning_rate": 4.997917665574964e-05,
+      "loss": 1.1463,
+      "step": 364
+    },
+    {
+      "epoch": 0.01607078169423801,
+      "grad_norm": 7.78125,
+      "learning_rate": 4.9978893388070135e-05,
+      "loss": 1.1436,
+      "step": 366
+    },
+    {
+      "epoch": 0.016158600173441496,
+      "grad_norm": 6.125,
+      "learning_rate": 4.997860820750098e-05,
+      "loss": 1.1402,
+      "step": 368
+    },
+    {
+      "epoch": 0.016246418652644982,
+      "grad_norm": 7.8125,
+      "learning_rate": 4.997832111406402e-05,
+      "loss": 1.098,
+      "step": 370
+    },
+    {
+      "epoch": 0.016334237131848468,
+      "grad_norm": 6.28125,
+      "learning_rate": 4.997803210778124e-05,
+      "loss": 1.1279,
+      "step": 372
+    },
+    {
+      "epoch": 0.016422055611051957,
+      "grad_norm": 5.5625,
+      "learning_rate": 4.997774118867477e-05,
+      "loss": 1.1178,
+      "step": 374
+    },
+    {
+      "epoch": 0.016509874090255443,
+      "grad_norm": 8.8125,
+      "learning_rate": 4.99774483567669e-05,
+      "loss": 1.1477,
+      "step": 376
+    },
+    {
+      "epoch": 0.01659769256945893,
+      "grad_norm": 6.9375,
+      "learning_rate": 4.997715361208004e-05,
+      "loss": 1.1008,
+      "step": 378
+    },
+    {
+      "epoch": 0.016685511048662414,
+      "grad_norm": 5.15625,
+      "learning_rate": 4.997685695463677e-05,
+      "loss": 1.1023,
+      "step": 380
+    },
+    {
+      "epoch": 0.0167733295278659,
+      "grad_norm": 9.375,
+      "learning_rate": 4.9976558384459815e-05,
+      "loss": 1.1327,
+      "step": 382
+    },
+    {
+      "epoch": 0.016861148007069386,
+      "grad_norm": 11.0625,
+      "learning_rate": 4.997625790157203e-05,
+      "loss": 1.1455,
+      "step": 384
+    },
+    {
+      "epoch": 0.016948966486272875,
+      "grad_norm": 8.0625,
+      "learning_rate": 4.997595550599642e-05,
+      "loss": 1.1056,
+      "step": 386
+    },
+    {
+      "epoch": 0.01703678496547636,
+      "grad_norm": 8.6875,
+      "learning_rate": 4.9975651197756155e-05,
+      "loss": 1.1077,
+      "step": 388
+    },
+    {
+      "epoch": 0.017124603444679847,
+      "grad_norm": 7.03125,
+      "learning_rate": 4.9975344976874536e-05,
+      "loss": 1.1418,
+      "step": 390
+    },
+    {
+      "epoch": 0.017212421923883333,
+      "grad_norm": 9.875,
+      "learning_rate": 4.997503684337501e-05,
+      "loss": 1.0905,
+      "step": 392
+    },
+    {
+      "epoch": 0.01730024040308682,
+      "grad_norm": 8.375,
+      "learning_rate": 4.997472679728118e-05,
+      "loss": 1.1365,
+      "step": 394
+    },
+    {
+      "epoch": 0.017388058882290305,
+      "grad_norm": 6.875,
+      "learning_rate": 4.997441483861678e-05,
+      "loss": 1.1076,
+      "step": 396
+    },
+    {
+      "epoch": 0.017475877361493794,
+      "grad_norm": 6.21875,
+      "learning_rate": 4.99741009674057e-05,
+      "loss": 1.0944,
+      "step": 398
+    },
+    {
+      "epoch": 0.01756369584069728,
+      "grad_norm": 9.3125,
+      "learning_rate": 4.997378518367199e-05,
+      "loss": 1.1178,
+      "step": 400
+    },
+    {
+      "epoch": 0.017651514319900766,
+      "grad_norm": 11.0625,
+      "learning_rate": 4.9973467487439816e-05,
+      "loss": 1.1157,
+      "step": 402
+    },
+    {
+      "epoch": 0.01773933279910425,
+      "grad_norm": 10.6875,
+      "learning_rate": 4.997314787873352e-05,
+      "loss": 1.1005,
+      "step": 404
+    },
+    {
+      "epoch": 0.017827151278307737,
+      "grad_norm": 6.9375,
+      "learning_rate": 4.9972826357577576e-05,
+      "loss": 1.0849,
+      "step": 406
+    },
+    {
+      "epoch": 0.017914969757511223,
+      "grad_norm": 5.46875,
+      "learning_rate": 4.99725029239966e-05,
+      "loss": 1.1028,
+      "step": 408
+    },
+    {
+      "epoch": 0.01800278823671471,
+      "grad_norm": 7.34375,
+      "learning_rate": 4.9972177578015364e-05,
+      "loss": 1.1027,
+      "step": 410
+    },
+    {
+      "epoch": 0.0180906067159182,
+      "grad_norm": 6.90625,
+      "learning_rate": 4.997185031965878e-05,
+      "loss": 1.0742,
+      "step": 412
+    },
+    {
+      "epoch": 0.018178425195121684,
+      "grad_norm": 6.9375,
+      "learning_rate": 4.9971521148951914e-05,
+      "loss": 1.0998,
+      "step": 414
+    },
+    {
+      "epoch": 0.01826624367432517,
+      "grad_norm": 6.3125,
+      "learning_rate": 4.997119006591997e-05,
+      "loss": 1.0511,
+      "step": 416
+    },
+    {
+      "epoch": 0.018354062153528656,
+      "grad_norm": 8.875,
+      "learning_rate": 4.997085707058832e-05,
+      "loss": 1.0712,
+      "step": 418
+    },
+    {
+      "epoch": 0.018441880632732142,
+      "grad_norm": 7.625,
+      "learning_rate": 4.997052216298243e-05,
+      "loss": 1.1015,
+      "step": 420
+    },
+    {
+      "epoch": 0.018529699111935628,
+      "grad_norm": 7.5625,
+      "learning_rate": 4.9970185343127975e-05,
+      "loss": 1.071,
+      "step": 422
+    },
+    {
+      "epoch": 0.018617517591139117,
+      "grad_norm": 5.78125,
+      "learning_rate": 4.9969846611050744e-05,
+      "loss": 1.0926,
+      "step": 424
+    },
+    {
+      "epoch": 0.018705336070342603,
+      "grad_norm": 6.34375,
+      "learning_rate": 4.9969505966776664e-05,
+      "loss": 1.0808,
+      "step": 426
+    },
+    {
+      "epoch": 0.01879315454954609,
+      "grad_norm": 6.53125,
+      "learning_rate": 4.9969163410331845e-05,
+      "loss": 1.0853,
+      "step": 428
+    },
+    {
+      "epoch": 0.018880973028749574,
+      "grad_norm": 5.65625,
+      "learning_rate": 4.996881894174249e-05,
+      "loss": 1.0674,
+      "step": 430
+    },
+    {
+      "epoch": 0.01896879150795306,
+      "grad_norm": 9.8125,
+      "learning_rate": 4.996847256103501e-05,
+      "loss": 1.0528,
+      "step": 432
+    },
+    {
+      "epoch": 0.019056609987156546,
+      "grad_norm": 9.1875,
+      "learning_rate": 4.9968124268235906e-05,
+      "loss": 1.1181,
+      "step": 434
+    },
+    {
+      "epoch": 0.019144428466360035,
+      "grad_norm": 7.0625,
+      "learning_rate": 4.996777406337186e-05,
+      "loss": 1.0888,
+      "step": 436
+    },
+    {
+      "epoch": 0.01923224694556352,
+      "grad_norm": 7.0,
+      "learning_rate": 4.99674219464697e-05,
+      "loss": 1.043,
+      "step": 438
+    },
+    {
+      "epoch": 0.019320065424767007,
+      "grad_norm": 5.96875,
+      "learning_rate": 4.9967067917556376e-05,
+      "loss": 1.0879,
+      "step": 440
+    },
+    {
+      "epoch": 0.019407883903970493,
+      "grad_norm": 5.65625,
+      "learning_rate": 4.996671197665901e-05,
+      "loss": 1.072,
+      "step": 442
+    },
+    {
+      "epoch": 0.01949570238317398,
+      "grad_norm": 5.9375,
+      "learning_rate": 4.996635412380485e-05,
+      "loss": 1.0603,
+      "step": 444
+    },
+    {
+      "epoch": 0.019583520862377465,
+      "grad_norm": 10.125,
+      "learning_rate": 4.996599435902131e-05,
+      "loss": 1.0686,
+      "step": 446
+    },
+    {
+      "epoch": 0.01967133934158095,
+      "grad_norm": 13.3125,
+      "learning_rate": 4.996563268233594e-05,
+      "loss": 1.075,
+      "step": 448
+    },
+    {
+      "epoch": 0.01975915782078444,
+      "grad_norm": 12.25,
+      "learning_rate": 4.996526909377643e-05,
+      "loss": 1.0792,
+      "step": 450
+    },
+    {
+      "epoch": 0.019846976299987926,
+      "grad_norm": 7.8125,
+      "learning_rate": 4.996490359337062e-05,
+      "loss": 1.0456,
+      "step": 452
+    },
+    {
+      "epoch": 0.01993479477919141,
+      "grad_norm": 6.59375,
+      "learning_rate": 4.9964536181146525e-05,
+      "loss": 1.0603,
+      "step": 454
+    },
+    {
+      "epoch": 0.020022613258394897,
+      "grad_norm": 9.125,
+      "learning_rate": 4.9964166857132255e-05,
+      "loss": 1.0536,
+      "step": 456
+    },
+    {
+      "epoch": 0.020110431737598383,
+      "grad_norm": 11.0,
+      "learning_rate": 4.996379562135611e-05,
+      "loss": 1.0881,
+      "step": 458
+    },
+    {
+      "epoch": 0.02019825021680187,
+      "grad_norm": 5.4375,
+      "learning_rate": 4.9963422473846504e-05,
+      "loss": 1.0952,
+      "step": 460
+    },
+    {
+      "epoch": 0.02028606869600536,
+      "grad_norm": 5.9375,
+      "learning_rate": 4.9963047414632024e-05,
+      "loss": 1.0648,
+      "step": 462
+    },
+    {
+      "epoch": 0.020373887175208844,
+      "grad_norm": 6.53125,
+      "learning_rate": 4.996267044374139e-05,
+      "loss": 1.0648,
+      "step": 464
+    },
+    {
+      "epoch": 0.02046170565441233,
+      "grad_norm": 8.0625,
+      "learning_rate": 4.9962291561203464e-05,
+      "loss": 1.0814,
+      "step": 466
+    },
+    {
+      "epoch": 0.020549524133615816,
+      "grad_norm": 9.8125,
+      "learning_rate": 4.9961910767047275e-05,
+      "loss": 1.0556,
+      "step": 468
+    },
+    {
+      "epoch": 0.020637342612819302,
+      "grad_norm": 8.375,
+      "learning_rate": 4.996152806130198e-05,
+      "loss": 1.0354,
+      "step": 470
+    },
+    {
+      "epoch": 0.020725161092022788,
+      "grad_norm": 8.375,
+      "learning_rate": 4.996114344399687e-05,
+      "loss": 1.0932,
+      "step": 472
+    },
+    {
+      "epoch": 0.020812979571226277,
+      "grad_norm": 7.90625,
+      "learning_rate": 4.996075691516141e-05,
+      "loss": 1.0409,
+      "step": 474
+    },
+    {
+      "epoch": 0.020900798050429763,
+      "grad_norm": 7.4375,
+      "learning_rate": 4.996036847482521e-05,
+      "loss": 1.0539,
+      "step": 476
+    },
+    {
+      "epoch": 0.02098861652963325,
+      "grad_norm": 7.6875,
+      "learning_rate": 4.9959978123018006e-05,
+      "loss": 1.0252,
+      "step": 478
+    },
+    {
+      "epoch": 0.021076435008836734,
+      "grad_norm": 6.875,
+      "learning_rate": 4.9959585859769694e-05,
+      "loss": 1.0492,
+      "step": 480
+    },
+    {
+      "epoch": 0.02116425348804022,
+      "grad_norm": 6.625,
+      "learning_rate": 4.995919168511032e-05,
+      "loss": 1.0662,
+      "step": 482
+    },
+    {
+      "epoch": 0.021252071967243706,
+      "grad_norm": 6.5,
+      "learning_rate": 4.995879559907005e-05,
+      "loss": 1.035,
+      "step": 484
+    },
+    {
+      "epoch": 0.021339890446447192,
+      "grad_norm": 6.59375,
+      "learning_rate": 4.995839760167924e-05,
+      "loss": 1.0408,
+      "step": 486
+    },
+    {
+      "epoch": 0.02142770892565068,
+      "grad_norm": 6.875,
+      "learning_rate": 4.995799769296836e-05,
+      "loss": 1.0301,
+      "step": 488
+    },
+    {
+      "epoch": 0.021515527404854167,
+      "grad_norm": 5.90625,
+      "learning_rate": 4.995759587296803e-05,
+      "loss": 1.0198,
+      "step": 490
+    },
+    {
+      "epoch": 0.021603345884057653,
+      "grad_norm": 5.53125,
+      "learning_rate": 4.995719214170902e-05,
+      "loss": 1.0354,
+      "step": 492
+    },
+    {
+      "epoch": 0.02169116436326114,
+      "grad_norm": 7.5625,
+      "learning_rate": 4.9956786499222263e-05,
+      "loss": 1.044,
+      "step": 494
+    },
+    {
+      "epoch": 0.021778982842464625,
+      "grad_norm": 6.375,
+      "learning_rate": 4.995637894553881e-05,
+      "loss": 1.0413,
+      "step": 496
+    },
+    {
+      "epoch": 0.02186680132166811,
+      "grad_norm": 6.40625,
+      "learning_rate": 4.9955969480689865e-05,
+      "loss": 1.0098,
+      "step": 498
+    },
+    {
+      "epoch": 0.0219546198008716,
+      "grad_norm": 6.1875,
+      "learning_rate": 4.995555810470681e-05,
+      "loss": 1.0262,
+      "step": 500
+    },
+    {
+      "epoch": 0.022042438280075086,
+      "grad_norm": 4.96875,
+      "learning_rate": 4.995514481762112e-05,
+      "loss": 1.0375,
+      "step": 502
+    },
+    {
+      "epoch": 0.02213025675927857,
+      "grad_norm": 5.8125,
+      "learning_rate": 4.995472961946447e-05,
+      "loss": 1.035,
+      "step": 504
+    },
+    {
+      "epoch": 0.022218075238482057,
+      "grad_norm": 5.5,
+      "learning_rate": 4.9954312510268636e-05,
+      "loss": 1.0307,
+      "step": 506
+    },
+    {
+      "epoch": 0.022305893717685543,
+      "grad_norm": 5.0625,
+      "learning_rate": 4.995389349006557e-05,
+      "loss": 0.9841,
+      "step": 508
+    },
+    {
+      "epoch": 0.02239371219688903,
+      "grad_norm": 6.53125,
+      "learning_rate": 4.995347255888736e-05,
+      "loss": 1.0169,
+      "step": 510
+    },
+    {
+      "epoch": 0.02248153067609252,
+      "grad_norm": 6.125,
+      "learning_rate": 4.9953049716766234e-05,
+      "loss": 1.0283,
+      "step": 512
+    },
+    {
+      "epoch": 0.022569349155296004,
+      "grad_norm": 6.4375,
+      "learning_rate": 4.9952624963734584e-05,
+      "loss": 1.0303,
+      "step": 514
+    },
+    {
+      "epoch": 0.02265716763449949,
+      "grad_norm": 5.84375,
+      "learning_rate": 4.995219829982494e-05,
+      "loss": 1.0268,
+      "step": 516
+    },
+    {
+      "epoch": 0.022744986113702976,
+      "grad_norm": 6.59375,
+      "learning_rate": 4.9951769725069954e-05,
+      "loss": 1.0065,
+      "step": 518
+    },
+    {
+      "epoch": 0.02283280459290646,
+      "grad_norm": 5.5625,
+      "learning_rate": 4.995133923950247e-05,
+      "loss": 1.0092,
+      "step": 520
+    },
+    {
+      "epoch": 0.022920623072109948,
+      "grad_norm": 11.5625,
+      "learning_rate": 4.995090684315544e-05,
+      "loss": 1.0142,
+      "step": 522
+    },
+    {
+      "epoch": 0.023008441551313433,
+      "grad_norm": 9.5625,
+      "learning_rate": 4.9950472536061984e-05,
+      "loss": 0.9864,
+      "step": 524
+    },
+    {
+      "epoch": 0.023096260030516923,
+      "grad_norm": 7.09375,
+      "learning_rate": 4.9950036318255364e-05,
+      "loss": 1.0048,
+      "step": 526
+    },
+    {
+      "epoch": 0.02318407850972041,
+      "grad_norm": 6.8125,
+      "learning_rate": 4.9949598189768985e-05,
+      "loss": 1.0157,
+      "step": 528
+    },
+    {
+      "epoch": 0.023271896988923894,
+      "grad_norm": 5.5,
+      "learning_rate": 4.994915815063639e-05,
+      "loss": 1.0379,
+      "step": 530
+    },
+    {
+      "epoch": 0.02335971546812738,
+      "grad_norm": 8.25,
+      "learning_rate": 4.994871620089129e-05,
+      "loss": 1.0017,
+      "step": 532
+    },
+    {
+      "epoch": 0.023447533947330866,
+      "grad_norm": 7.8125,
+      "learning_rate": 4.9948272340567514e-05,
+      "loss": 1.0021,
+      "step": 534
+    },
+    {
+      "epoch": 0.023535352426534352,
+      "grad_norm": 6.0625,
+      "learning_rate": 4.994782656969906e-05,
+      "loss": 0.9881,
+      "step": 536
+    },
+    {
+      "epoch": 0.02362317090573784,
+      "grad_norm": 6.84375,
+      "learning_rate": 4.994737888832007e-05,
+      "loss": 0.9894,
+      "step": 538
+    },
+    {
+      "epoch": 0.023710989384941327,
+      "grad_norm": 7.46875,
+      "learning_rate": 4.9946929296464825e-05,
+      "loss": 1.0131,
+      "step": 540
+    },
+    {
+      "epoch": 0.023798807864144813,
+      "grad_norm": 8.6875,
+      "learning_rate": 4.994647779416776e-05,
+      "loss": 1.017,
+      "step": 542
+    },
+    {
+      "epoch": 0.0238866263433483,
+      "grad_norm": 10.0625,
+      "learning_rate": 4.994602438146344e-05,
+      "loss": 0.982,
+      "step": 544
+    },
+    {
+      "epoch": 0.023974444822551785,
+      "grad_norm": 8.625,
+      "learning_rate": 4.99455690583866e-05,
+      "loss": 1.0208,
+      "step": 546
+    },
+    {
+      "epoch": 0.02406226330175527,
+      "grad_norm": 11.0625,
+      "learning_rate": 4.994511182497209e-05,
+      "loss": 1.0111,
+      "step": 548
+    },
+    {
+      "epoch": 0.02415008178095876,
+      "grad_norm": 11.875,
+      "learning_rate": 4.9944652681254944e-05,
+      "loss": 0.9969,
+      "step": 550
+    },
+    {
+      "epoch": 0.024237900260162246,
+      "grad_norm": 8.375,
+      "learning_rate": 4.9944191627270314e-05,
+      "loss": 1.0336,
+      "step": 552
+    },
+    {
+      "epoch": 0.02432571873936573,
+      "grad_norm": 7.1875,
+      "learning_rate": 4.994372866305351e-05,
+      "loss": 1.0003,
+      "step": 554
+    },
+    {
+      "epoch": 0.024413537218569217,
+      "grad_norm": 8.6875,
+      "learning_rate": 4.9943263788639985e-05,
+      "loss": 1.0197,
+      "step": 556
+    },
+    {
+      "epoch": 0.024501355697772703,
+      "grad_norm": 6.3125,
+      "learning_rate": 4.994279700406534e-05,
+      "loss": 0.9861,
+      "step": 558
+    },
+    {
+      "epoch": 0.02458917417697619,
+      "grad_norm": 7.125,
+      "learning_rate": 4.994232830936532e-05,
+      "loss": 1.0242,
+      "step": 560
+    },
+    {
+      "epoch": 0.02467699265617968,
+      "grad_norm": 6.09375,
+      "learning_rate": 4.994185770457582e-05,
+      "loss": 0.999,
+      "step": 562
+    },
+    {
+      "epoch": 0.024764811135383164,
+      "grad_norm": 7.5,
+      "learning_rate": 4.994138518973288e-05,
+      "loss": 0.9646,
+      "step": 564
+    },
+    {
+      "epoch": 0.02485262961458665,
+      "grad_norm": 5.59375,
+      "learning_rate": 4.9940910764872685e-05,
+      "loss": 0.9966,
+      "step": 566
+    },
+    {
+      "epoch": 0.024940448093790136,
+      "grad_norm": 5.28125,
+      "learning_rate": 4.994043443003156e-05,
+      "loss": 0.9982,
+      "step": 568
+    },
+    {
+      "epoch": 0.02502826657299362,
+      "grad_norm": 6.15625,
+      "learning_rate": 4.993995618524598e-05,
+      "loss": 1.006,
+      "step": 570
+    },
+    {
+      "epoch": 0.025116085052197108,
+      "grad_norm": 6.40625,
+      "learning_rate": 4.993947603055259e-05,
+      "loss": 0.9813,
+      "step": 572
+    },
+    {
+      "epoch": 0.025203903531400593,
+      "grad_norm": 7.625,
+      "learning_rate": 4.9938993965988145e-05,
+      "loss": 0.9576,
+      "step": 574
+    },
+    {
+      "epoch": 0.025291722010604083,
+      "grad_norm": 5.65625,
+      "learning_rate": 4.993850999158956e-05,
+      "loss": 1.0088,
+      "step": 576
+    },
+    {
+      "epoch": 0.02537954048980757,
+      "grad_norm": 5.6875,
+      "learning_rate": 4.993802410739391e-05,
+      "loss": 0.9918,
+      "step": 578
+    },
+    {
+      "epoch": 0.025467358969011054,
+      "grad_norm": 6.125,
+      "learning_rate": 4.993753631343838e-05,
+      "loss": 0.9659,
+      "step": 580
+    },
+    {
+      "epoch": 0.02555517744821454,
+      "grad_norm": 8.5,
+      "learning_rate": 4.9937046609760356e-05,
+      "loss": 0.9646,
+      "step": 582
+    },
+    {
+      "epoch": 0.025642995927418026,
+      "grad_norm": 7.0625,
+      "learning_rate": 4.993655499639732e-05,
+      "loss": 0.9554,
+      "step": 584
+    },
+    {
+      "epoch": 0.025730814406621512,
+      "grad_norm": 5.0,
+      "learning_rate": 4.9936061473386925e-05,
+      "loss": 0.994,
+      "step": 586
+    },
+    {
+      "epoch": 0.025818632885825,
+      "grad_norm": 5.28125,
+      "learning_rate": 4.993556604076696e-05,
+      "loss": 0.9731,
+      "step": 588
+    },
+    {
+      "epoch": 0.025906451365028487,
+      "grad_norm": 7.0625,
+      "learning_rate": 4.9935068698575375e-05,
+      "loss": 0.9583,
+      "step": 590
+    },
+    {
+      "epoch": 0.025994269844231973,
+      "grad_norm": 5.5,
+      "learning_rate": 4.9934569446850256e-05,
+      "loss": 0.9676,
+      "step": 592
+    },
+    {
+      "epoch": 0.02608208832343546,
+      "grad_norm": 5.3125,
+      "learning_rate": 4.993406828562982e-05,
+      "loss": 0.9957,
+      "step": 594
+    },
+    {
+      "epoch": 0.026169906802638945,
+      "grad_norm": 6.5,
+      "learning_rate": 4.993356521495246e-05,
+      "loss": 0.9717,
+      "step": 596
+    },
+    {
+      "epoch": 0.02625772528184243,
+      "grad_norm": 10.5,
+      "learning_rate": 4.993306023485671e-05,
+      "loss": 0.9933,
+      "step": 598
+    },
+    {
+      "epoch": 0.02634554376104592,
+      "grad_norm": 10.75,
+      "learning_rate": 4.993255334538122e-05,
+      "loss": 0.9773,
+      "step": 600
+    },
+    {
+      "epoch": 0.026433362240249406,
+      "grad_norm": 11.1875,
+      "learning_rate": 4.9932044546564824e-05,
+      "loss": 0.9854,
+      "step": 602
+    },
+    {
+      "epoch": 0.02652118071945289,
+      "grad_norm": 7.625,
+      "learning_rate": 4.993153383844648e-05,
+      "loss": 0.995,
+      "step": 604
+    },
+    {
+      "epoch": 0.026608999198656377,
+      "grad_norm": 5.25,
+      "learning_rate": 4.993102122106529e-05,
+      "loss": 0.9621,
+      "step": 606
+    },
+    {
+      "epoch": 0.026696817677859863,
+      "grad_norm": 6.71875,
+      "learning_rate": 4.993050669446053e-05,
+      "loss": 0.9785,
+      "step": 608
+    },
+    {
+      "epoch": 0.02678463615706335,
+      "grad_norm": 6.1875,
+      "learning_rate": 4.9929990258671574e-05,
+      "loss": 0.9875,
+      "step": 610
+    },
+    {
+      "epoch": 0.026872454636266835,
+      "grad_norm": 6.46875,
+      "learning_rate": 4.9929471913738e-05,
+      "loss": 0.9624,
+      "step": 612
+    },
+    {
+      "epoch": 0.026960273115470324,
+      "grad_norm": 5.46875,
+      "learning_rate": 4.99289516596995e-05,
+      "loss": 0.9641,
+      "step": 614
+    },
+    {
+      "epoch": 0.02704809159467381,
+      "grad_norm": 5.5625,
+      "learning_rate": 4.992842949659589e-05,
+      "loss": 0.9695,
+      "step": 616
+    },
+    {
+      "epoch": 0.027135910073877296,
+      "grad_norm": 5.1875,
+      "learning_rate": 4.9927905424467184e-05,
+      "loss": 0.9731,
+      "step": 618
+    },
+    {
+      "epoch": 0.02722372855308078,
+      "grad_norm": 6.78125,
+      "learning_rate": 4.992737944335349e-05,
+      "loss": 0.9354,
+      "step": 620
+    },
+    {
+      "epoch": 0.027311547032284268,
+      "grad_norm": 8.25,
+      "learning_rate": 4.992685155329512e-05,
+      "loss": 0.944,
+      "step": 622
+    },
+    {
+      "epoch": 0.027399365511487753,
+      "grad_norm": 8.4375,
+      "learning_rate": 4.992632175433247e-05,
+      "loss": 0.961,
+      "step": 624
+    },
+    {
+      "epoch": 0.027487183990691243,
+      "grad_norm": 8.4375,
+      "learning_rate": 4.9925790046506136e-05,
+      "loss": 0.9396,
+      "step": 626
+    },
+    {
+      "epoch": 0.02757500246989473,
+      "grad_norm": 7.65625,
+      "learning_rate": 4.9925256429856814e-05,
+      "loss": 0.9557,
+      "step": 628
+    },
+    {
+      "epoch": 0.027662820949098214,
+      "grad_norm": 7.1875,
+      "learning_rate": 4.992472090442539e-05,
+      "loss": 0.9546,
+      "step": 630
+    },
+    {
+      "epoch": 0.0277506394283017,
+      "grad_norm": 8.0625,
+      "learning_rate": 4.992418347025286e-05,
+      "loss": 0.9652,
+      "step": 632
+    },
+    {
+      "epoch": 0.027838457907505186,
+      "grad_norm": 8.0625,
+      "learning_rate": 4.9923644127380384e-05,
+      "loss": 0.9419,
+      "step": 634
+    },
+    {
+      "epoch": 0.027926276386708672,
+      "grad_norm": 5.75,
+      "learning_rate": 4.992310287584926e-05,
+      "loss": 0.9576,
+      "step": 636
+    },
+    {
+      "epoch": 0.02801409486591216,
+      "grad_norm": 5.5,
+      "learning_rate": 4.992255971570095e-05,
+      "loss": 0.9492,
+      "step": 638
+    },
+    {
+      "epoch": 0.028101913345115647,
+      "grad_norm": 6.0625,
+      "learning_rate": 4.9922014646977046e-05,
+      "loss": 0.9552,
+      "step": 640
+    },
+    {
+      "epoch": 0.028189731824319133,
+      "grad_norm": 6.84375,
+      "learning_rate": 4.992146766971928e-05,
+      "loss": 0.9551,
+      "step": 642
+    },
+    {
+      "epoch": 0.02827755030352262,
+      "grad_norm": 6.03125,
+      "learning_rate": 4.992091878396955e-05,
+      "loss": 0.9738,
+      "step": 644
+    },
+    {
+      "epoch": 0.028365368782726105,
+      "grad_norm": 7.46875,
+      "learning_rate": 4.9920367989769885e-05,
+      "loss": 0.9445,
+      "step": 646
+    },
+    {
+      "epoch": 0.02845318726192959,
+      "grad_norm": 8.4375,
+      "learning_rate": 4.991981528716246e-05,
+      "loss": 0.9412,
+      "step": 648
+    },
+    {
+      "epoch": 0.028541005741133076,
+      "grad_norm": 6.875,
+      "learning_rate": 4.991926067618962e-05,
+      "loss": 0.9425,
+      "step": 650
+    },
+    {
+      "epoch": 0.028628824220336566,
+      "grad_norm": 5.59375,
+      "learning_rate": 4.991870415689381e-05,
+      "loss": 0.9306,
+      "step": 652
+    },
+    {
+      "epoch": 0.02871664269954005,
+      "grad_norm": 5.53125,
+      "learning_rate": 4.9918145729317664e-05,
+      "loss": 0.9298,
+      "step": 654
+    },
+    {
+      "epoch": 0.028804461178743537,
+      "grad_norm": 4.84375,
+      "learning_rate": 4.991758539350395e-05,
+      "loss": 0.9475,
+      "step": 656
+    },
+    {
+      "epoch": 0.028892279657947023,
+      "grad_norm": 6.65625,
+      "learning_rate": 4.991702314949557e-05,
+      "loss": 0.9702,
+      "step": 658
+    },
+    {
+      "epoch": 0.02898009813715051,
+      "grad_norm": 7.5625,
+      "learning_rate": 4.9916458997335583e-05,
+      "loss": 0.9722,
+      "step": 660
+    },
+    {
+      "epoch": 0.029067916616353995,
+      "grad_norm": 7.8125,
+      "learning_rate": 4.99158929370672e-05,
+      "loss": 0.9374,
+      "step": 662
+    },
+    {
+      "epoch": 0.029155735095557484,
+      "grad_norm": 7.96875,
+      "learning_rate": 4.991532496873376e-05,
+      "loss": 0.9239,
+      "step": 664
+    },
+    {
+      "epoch": 0.02924355357476097,
+      "grad_norm": 6.75,
+      "learning_rate": 4.991475509237876e-05,
+      "loss": 0.9186,
+      "step": 666
+    },
+    {
+      "epoch": 0.029331372053964456,
+      "grad_norm": 6.03125,
+      "learning_rate": 4.9914183308045836e-05,
+      "loss": 0.9669,
+      "step": 668
+    },
+    {
+      "epoch": 0.02941919053316794,
+      "grad_norm": 6.28125,
+      "learning_rate": 4.991360961577879e-05,
+      "loss": 0.9288,
+      "step": 670
+    },
+    {
+      "epoch": 0.029507009012371428,
+      "grad_norm": 6.625,
+      "learning_rate": 4.9913034015621545e-05,
+      "loss": 0.9278,
+      "step": 672
+    },
+    {
+      "epoch": 0.029594827491574913,
+      "grad_norm": 6.6875,
+      "learning_rate": 4.9912456507618185e-05,
+      "loss": 0.907,
+      "step": 674
+    },
+    {
+      "epoch": 0.029682645970778403,
+      "grad_norm": 5.78125,
+      "learning_rate": 4.991187709181293e-05,
+      "loss": 0.908,
+      "step": 676
+    },
+    {
+      "epoch": 0.02977046444998189,
+      "grad_norm": 6.5625,
+      "learning_rate": 4.991129576825016e-05,
+      "loss": 0.9143,
+      "step": 678
+    },
+    {
+      "epoch": 0.029858282929185374,
+      "grad_norm": 5.875,
+      "learning_rate": 4.991071253697439e-05,
+      "loss": 0.9022,
+      "step": 680
+    },
+    {
+      "epoch": 0.02994610140838886,
+      "grad_norm": 5.46875,
+      "learning_rate": 4.991012739803028e-05,
+      "loss": 0.9249,
+      "step": 682
+    },
+    {
+      "epoch": 0.030033919887592346,
+      "grad_norm": 4.8125,
+      "learning_rate": 4.990954035146264e-05,
+      "loss": 0.9483,
+      "step": 684
+    },
+    {
+      "epoch": 0.030121738366795832,
+      "grad_norm": 4.75,
+      "learning_rate": 4.990895139731643e-05,
+      "loss": 0.9301,
+      "step": 686
+    },
+    {
+      "epoch": 0.030209556845999318,
+      "grad_norm": 5.0,
+      "learning_rate": 4.990836053563674e-05,
+      "loss": 0.9256,
+      "step": 688
+    },
+    {
+      "epoch": 0.030297375325202807,
+      "grad_norm": 4.84375,
+      "learning_rate": 4.990776776646885e-05,
+      "loss": 0.9601,
+      "step": 690
+    },
+    {
+      "epoch": 0.030385193804406293,
+      "grad_norm": 4.65625,
+      "learning_rate": 4.990717308985812e-05,
+      "loss": 0.9244,
+      "step": 692
+    },
+    {
+      "epoch": 0.03047301228360978,
+      "grad_norm": 5.0,
+      "learning_rate": 4.990657650585011e-05,
+      "loss": 0.9357,
+      "step": 694
+    },
+    {
+      "epoch": 0.030560830762813265,
+      "grad_norm": 6.28125,
+      "learning_rate": 4.99059780144905e-05,
+      "loss": 0.9379,
+      "step": 696
+    },
+    {
+      "epoch": 0.03064864924201675,
+      "grad_norm": 5.90625,
+      "learning_rate": 4.990537761582512e-05,
+      "loss": 0.9476,
+      "step": 698
+    },
+    {
+      "epoch": 0.030736467721220236,
+      "grad_norm": 5.34375,
+      "learning_rate": 4.9904775309899955e-05,
+      "loss": 0.9505,
+      "step": 700
+    },
+    {
+      "epoch": 0.030824286200423726,
+      "grad_norm": 5.65625,
+      "learning_rate": 4.9904171096761124e-05,
+      "loss": 0.9479,
+      "step": 702
+    },
+    {
+      "epoch": 0.03091210467962721,
+      "grad_norm": 4.78125,
+      "learning_rate": 4.99035649764549e-05,
+      "loss": 0.9047,
+      "step": 704
+    },
+    {
+      "epoch": 0.030999923158830697,
+      "grad_norm": 5.46875,
+      "learning_rate": 4.99029569490277e-05,
+      "loss": 0.9234,
+      "step": 706
+    },
+    {
+      "epoch": 0.031087741638034183,
+      "grad_norm": 7.96875,
+      "learning_rate": 4.9902347014526093e-05,
+      "loss": 0.9298,
+      "step": 708
+    },
+    {
+      "epoch": 0.03117556011723767,
+      "grad_norm": 8.8125,
+      "learning_rate": 4.9901735172996775e-05,
+      "loss": 0.9372,
+      "step": 710
+    },
+    {
+      "epoch": 0.031263378596441155,
+      "grad_norm": 7.3125,
+      "learning_rate": 4.990112142448662e-05,
+      "loss": 0.9224,
+      "step": 712
+    },
+    {
+      "epoch": 0.03135119707564464,
+      "grad_norm": 9.75,
+      "learning_rate": 4.990050576904261e-05,
+      "loss": 0.9216,
+      "step": 714
+    },
+    {
+      "epoch": 0.031439015554848126,
+      "grad_norm": 9.625,
+      "learning_rate": 4.989988820671189e-05,
+      "loss": 0.9117,
+      "step": 716
+    },
+    {
+      "epoch": 0.03152683403405161,
+      "grad_norm": 8.25,
+      "learning_rate": 4.989926873754177e-05,
+      "loss": 0.9419,
+      "step": 718
+    },
+    {
+      "epoch": 0.031614652513255105,
+      "grad_norm": 6.375,
+      "learning_rate": 4.989864736157969e-05,
+      "loss": 0.8973,
+      "step": 720
+    },
+    {
+      "epoch": 0.03170247099245859,
+      "grad_norm": 8.1875,
+      "learning_rate": 4.989802407887321e-05,
+      "loss": 0.9172,
+      "step": 722
+    },
+    {
+      "epoch": 0.03179028947166208,
+      "grad_norm": 9.9375,
+      "learning_rate": 4.9897398889470086e-05,
+      "loss": 0.9366,
+      "step": 724
+    },
+    {
+      "epoch": 0.03187810795086556,
+      "grad_norm": 5.46875,
+      "learning_rate": 4.989677179341818e-05,
+      "loss": 0.9221,
+      "step": 726
+    },
+    {
+      "epoch": 0.03196592643006905,
+      "grad_norm": 5.46875,
+      "learning_rate": 4.989614279076553e-05,
+      "loss": 0.8898,
+      "step": 728
+    },
+    {
+      "epoch": 0.032053744909272534,
+      "grad_norm": 7.15625,
+      "learning_rate": 4.9895511881560294e-05,
+      "loss": 0.9261,
+      "step": 730
+    },
+    {
+      "epoch": 0.03214156338847602,
+      "grad_norm": 7.78125,
+      "learning_rate": 4.98948790658508e-05,
+      "loss": 0.8904,
+      "step": 732
+    },
+    {
+      "epoch": 0.032229381867679506,
+      "grad_norm": 7.71875,
+      "learning_rate": 4.989424434368549e-05,
+      "loss": 0.9143,
+      "step": 734
+    },
+    {
+      "epoch": 0.03231720034688299,
+      "grad_norm": 7.875,
+      "learning_rate": 4.989360771511298e-05,
+      "loss": 0.9082,
+      "step": 736
+    },
+    {
+      "epoch": 0.03240501882608648,
+      "grad_norm": 6.28125,
+      "learning_rate": 4.989296918018203e-05,
+      "loss": 0.8653,
+      "step": 738
+    },
+    {
+      "epoch": 0.032492837305289964,
+      "grad_norm": 5.46875,
+      "learning_rate": 4.989232873894152e-05,
+      "loss": 0.8899,
+      "step": 740
+    },
+    {
+      "epoch": 0.03258065578449345,
+      "grad_norm": 5.625,
+      "learning_rate": 4.989168639144052e-05,
+      "loss": 0.8836,
+      "step": 742
+    },
+    {
+      "epoch": 0.032668474263696935,
+      "grad_norm": 5.4375,
+      "learning_rate": 4.98910421377282e-05,
+      "loss": 0.8981,
+      "step": 744
+    },
+    {
+      "epoch": 0.03275629274290043,
+      "grad_norm": 6.03125,
+      "learning_rate": 4.989039597785392e-05,
+      "loss": 0.903,
+      "step": 746
+    },
+    {
+      "epoch": 0.032844111222103914,
+      "grad_norm": 7.59375,
+      "learning_rate": 4.988974791186713e-05,
+      "loss": 0.885,
+      "step": 748
+    },
+    {
+      "epoch": 0.0329319297013074,
+      "grad_norm": 8.3125,
+      "learning_rate": 4.98890979398175e-05,
+      "loss": 0.9071,
+      "step": 750
+    },
+    {
+      "epoch": 0.033019748180510886,
+      "grad_norm": 6.75,
+      "learning_rate": 4.988844606175477e-05,
+      "loss": 0.9012,
+      "step": 752
+    },
+    {
+      "epoch": 0.03310756665971437,
+      "grad_norm": 6.0625,
+      "learning_rate": 4.988779227772888e-05,
+      "loss": 0.8822,
+      "step": 754
+    },
+    {
+      "epoch": 0.03319538513891786,
+      "grad_norm": 7.09375,
+      "learning_rate": 4.9887136587789885e-05,
+      "loss": 0.8973,
+      "step": 756
+    },
+    {
+      "epoch": 0.03328320361812134,
+      "grad_norm": 9.125,
+      "learning_rate": 4.9886478991988004e-05,
+      "loss": 0.911,
+      "step": 758
+    },
+    {
+      "epoch": 0.03337102209732483,
+      "grad_norm": 8.25,
+      "learning_rate": 4.9885819490373605e-05,
+      "loss": 0.8973,
+      "step": 760
+    },
+    {
+      "epoch": 0.033458840576528315,
+      "grad_norm": 6.5,
+      "learning_rate": 4.988515808299718e-05,
+      "loss": 0.9305,
+      "step": 762
+    },
+    {
+      "epoch": 0.0335466590557318,
+      "grad_norm": 6.46875,
+      "learning_rate": 4.988449476990938e-05,
+      "loss": 0.8778,
+      "step": 764
+    },
+    {
+      "epoch": 0.033634477534935286,
+      "grad_norm": 6.21875,
+      "learning_rate": 4.988382955116101e-05,
+      "loss": 0.8987,
+      "step": 766
+    },
+    {
+      "epoch": 0.03372229601413877,
+      "grad_norm": 5.84375,
+      "learning_rate": 4.988316242680301e-05,
+      "loss": 0.8735,
+      "step": 768
+    },
+    {
+      "epoch": 0.03381011449334226,
+      "grad_norm": 5.28125,
+      "learning_rate": 4.988249339688646e-05,
+      "loss": 0.901,
+      "step": 770
+    },
+    {
+      "epoch": 0.03389793297254575,
+      "grad_norm": 5.21875,
+      "learning_rate": 4.98818224614626e-05,
+      "loss": 0.8986,
+      "step": 772
+    },
+    {
+      "epoch": 0.03398575145174924,
+      "grad_norm": 4.90625,
+      "learning_rate": 4.9881149620582815e-05,
+      "loss": 0.895,
+      "step": 774
+    },
+    {
+      "epoch": 0.03407356993095272,
+      "grad_norm": 5.09375,
+      "learning_rate": 4.9880474874298626e-05,
+      "loss": 0.8893,
+      "step": 776
+    },
+    {
+      "epoch": 0.03416138841015621,
+      "grad_norm": 5.1875,
+      "learning_rate": 4.987979822266171e-05,
+      "loss": 0.8624,
+      "step": 778
+    },
+    {
+      "epoch": 0.034249206889359694,
+      "grad_norm": 6.0625,
+      "learning_rate": 4.9879119665723885e-05,
+      "loss": 0.8925,
+      "step": 780
+    },
+    {
+      "epoch": 0.03433702536856318,
+      "grad_norm": 5.53125,
+      "learning_rate": 4.9878439203537104e-05,
+      "loss": 0.8488,
+      "step": 782
+    },
+    {
+      "epoch": 0.034424843847766666,
+      "grad_norm": 6.46875,
+      "learning_rate": 4.987775683615349e-05,
+      "loss": 0.8915,
+      "step": 784
+    },
+    {
+      "epoch": 0.03451266232697015,
+      "grad_norm": 5.09375,
+      "learning_rate": 4.9877072563625285e-05,
+      "loss": 0.8532,
+      "step": 786
+    },
+    {
+      "epoch": 0.03460048080617364,
+      "grad_norm": 4.4375,
+      "learning_rate": 4.9876386386004905e-05,
+      "loss": 0.879,
+      "step": 788
+    },
+    {
+      "epoch": 0.034688299285377124,
+      "grad_norm": 4.75,
+      "learning_rate": 4.9875698303344896e-05,
+      "loss": 0.8712,
+      "step": 790
+    },
+    {
+      "epoch": 0.03477611776458061,
+      "grad_norm": 4.78125,
+      "learning_rate": 4.987500831569795e-05,
+      "loss": 0.8756,
+      "step": 792
+    },
+    {
+      "epoch": 0.034863936243784095,
+      "grad_norm": 5.78125,
+      "learning_rate": 4.987431642311689e-05,
+      "loss": 0.8567,
+      "step": 794
+    },
+    {
+      "epoch": 0.03495175472298759,
+      "grad_norm": 5.9375,
+      "learning_rate": 4.987362262565474e-05,
+      "loss": 0.887,
+      "step": 796
+    },
+    {
+      "epoch": 0.035039573202191074,
+      "grad_norm": 4.9375,
+      "learning_rate": 4.987292692336458e-05,
+      "loss": 0.859,
+      "step": 798
+    },
+    {
+      "epoch": 0.03512739168139456,
+      "grad_norm": 6.09375,
+      "learning_rate": 4.9872229316299734e-05,
+      "loss": 0.8759,
+      "step": 800
+    },
+    {
+      "epoch": 0.035215210160598046,
+      "grad_norm": 4.90625,
+      "learning_rate": 4.987152980451359e-05,
+      "loss": 0.8624,
+      "step": 802
+    },
+    {
+      "epoch": 0.03530302863980153,
+      "grad_norm": 5.90625,
+      "learning_rate": 4.987082838805974e-05,
+      "loss": 0.8837,
+      "step": 804
+    },
+    {
+      "epoch": 0.03539084711900502,
+      "grad_norm": 5.0625,
+      "learning_rate": 4.9870125066991894e-05,
+      "loss": 0.8992,
+      "step": 806
+    },
+    {
+      "epoch": 0.0354786655982085,
+      "grad_norm": 5.1875,
+      "learning_rate": 4.98694198413639e-05,
+      "loss": 0.849,
+      "step": 808
+    },
+    {
+      "epoch": 0.03556648407741199,
+      "grad_norm": 4.875,
+      "learning_rate": 4.986871271122977e-05,
+      "loss": 0.8616,
+      "step": 810
+    },
+    {
+      "epoch": 0.035654302556615475,
+      "grad_norm": 6.1875,
+      "learning_rate": 4.986800367664367e-05,
+      "loss": 0.9031,
+      "step": 812
+    },
+    {
+      "epoch": 0.03574212103581896,
+      "grad_norm": 4.8125,
+      "learning_rate": 4.986729273765988e-05,
+      "loss": 0.8518,
+      "step": 814
+    },
+    {
+      "epoch": 0.035829939515022446,
+      "grad_norm": 4.59375,
+      "learning_rate": 4.9866579894332857e-05,
+      "loss": 0.8672,
+      "step": 816
+    },
+    {
+      "epoch": 0.03591775799422593,
+      "grad_norm": 5.25,
+      "learning_rate": 4.9865865146717176e-05,
+      "loss": 0.8661,
+      "step": 818
+    },
+    {
+      "epoch": 0.03600557647342942,
+      "grad_norm": 6.84375,
+      "learning_rate": 4.9865148494867584e-05,
+      "loss": 0.8709,
+      "step": 820
+    },
+    {
+      "epoch": 0.03609339495263291,
+      "grad_norm": 4.875,
+      "learning_rate": 4.986442993883896e-05,
+      "loss": 0.8915,
+      "step": 822
+    },
+    {
+      "epoch": 0.0361812134318364,
+      "grad_norm": 4.53125,
+      "learning_rate": 4.986370947868634e-05,
+      "loss": 0.8497,
+      "step": 824
+    },
+    {
+      "epoch": 0.03626903191103988,
+      "grad_norm": 5.46875,
+      "learning_rate": 4.986298711446488e-05,
+      "loss": 0.8722,
+      "step": 826
+    },
+    {
+      "epoch": 0.03635685039024337,
+      "grad_norm": 8.0625,
+      "learning_rate": 4.986226284622991e-05,
+      "loss": 0.8664,
+      "step": 828
+    },
+    {
+      "epoch": 0.036444668869446854,
+      "grad_norm": 9.0,
+      "learning_rate": 4.9861536674036885e-05,
+      "loss": 0.8541,
+      "step": 830
+    },
+    {
+      "epoch": 0.03653248734865034,
+      "grad_norm": 8.3125,
+      "learning_rate": 4.986080859794142e-05,
+      "loss": 0.8752,
+      "step": 832
+    },
+    {
+      "epoch": 0.036620305827853826,
+      "grad_norm": 8.8125,
+      "learning_rate": 4.9860078617999284e-05,
+      "loss": 0.8758,
+      "step": 834
+    },
+    {
+      "epoch": 0.03670812430705731,
+      "grad_norm": 8.25,
+      "learning_rate": 4.9859346734266365e-05,
+      "loss": 0.8715,
+      "step": 836
+    },
+    {
+      "epoch": 0.0367959427862608,
+      "grad_norm": 5.84375,
+      "learning_rate": 4.9858612946798714e-05,
+      "loss": 0.8564,
+      "step": 838
+    },
+    {
+      "epoch": 0.036883761265464284,
+      "grad_norm": 5.71875,
+      "learning_rate": 4.985787725565252e-05,
+      "loss": 0.8502,
+      "step": 840
+    },
+    {
+      "epoch": 0.03697157974466777,
+      "grad_norm": 5.71875,
+      "learning_rate": 4.985713966088412e-05,
+      "loss": 0.8582,
+      "step": 842
+    },
+    {
+      "epoch": 0.037059398223871255,
+      "grad_norm": 6.59375,
+      "learning_rate": 4.985640016255002e-05,
+      "loss": 0.8623,
+      "step": 844
+    },
+    {
+      "epoch": 0.03714721670307474,
+      "grad_norm": 5.375,
+      "learning_rate": 4.985565876070683e-05,
+      "loss": 0.8719,
+      "step": 846
+    },
+    {
+      "epoch": 0.037235035182278234,
+      "grad_norm": 5.0,
+      "learning_rate": 4.9854915455411334e-05,
+      "loss": 0.9025,
+      "step": 848
+    },
+    {
+      "epoch": 0.03732285366148172,
+      "grad_norm": 5.75,
+      "learning_rate": 4.9854170246720456e-05,
+      "loss": 0.8393,
+      "step": 850
+    },
+    {
+      "epoch": 0.037410672140685206,
+      "grad_norm": 4.6875,
+      "learning_rate": 4.9853423134691265e-05,
+      "loss": 0.8611,
+      "step": 852
+    },
+    {
+      "epoch": 0.03749849061988869,
+      "grad_norm": 6.375,
+      "learning_rate": 4.985267411938097e-05,
+      "loss": 0.8633,
+      "step": 854
+    },
+    {
+      "epoch": 0.03758630909909218,
+      "grad_norm": 5.78125,
+      "learning_rate": 4.9851923200846934e-05,
+      "loss": 0.8516,
+      "step": 856
+    },
+    {
+      "epoch": 0.03767412757829566,
+      "grad_norm": 6.15625,
+      "learning_rate": 4.985117037914666e-05,
+      "loss": 0.872,
+      "step": 858
+    },
+    {
+      "epoch": 0.03776194605749915,
+      "grad_norm": 5.5625,
+      "learning_rate": 4.9850415654337804e-05,
+      "loss": 0.8479,
+      "step": 860
+    },
+    {
+      "epoch": 0.037849764536702635,
+      "grad_norm": 6.125,
+      "learning_rate": 4.9849659026478154e-05,
+      "loss": 0.8318,
+      "step": 862
+    },
+    {
+      "epoch": 0.03793758301590612,
+      "grad_norm": 5.125,
+      "learning_rate": 4.9848900495625665e-05,
+      "loss": 0.8792,
+      "step": 864
+    },
+    {
+      "epoch": 0.038025401495109606,
+      "grad_norm": 5.09375,
+      "learning_rate": 4.9848140061838424e-05,
+      "loss": 0.855,
+      "step": 866
+    },
+    {
+      "epoch": 0.03811321997431309,
+      "grad_norm": 4.53125,
+      "learning_rate": 4.984737772517465e-05,
+      "loss": 0.8471,
+      "step": 868
+    },
+    {
+      "epoch": 0.03820103845351658,
+      "grad_norm": 6.375,
+      "learning_rate": 4.984661348569274e-05,
+      "loss": 0.8514,
+      "step": 870
+    },
+    {
+      "epoch": 0.03828885693272007,
+      "grad_norm": 5.96875,
+      "learning_rate": 4.984584734345121e-05,
+      "loss": 0.842,
+      "step": 872
+    },
+    {
+      "epoch": 0.03837667541192356,
+      "grad_norm": 8.4375,
+      "learning_rate": 4.984507929850873e-05,
+      "loss": 0.872,
+      "step": 874
+    },
+    {
+      "epoch": 0.03846449389112704,
+      "grad_norm": 8.375,
+      "learning_rate": 4.9844309350924135e-05,
+      "loss": 0.8892,
+      "step": 876
+    },
+    {
+      "epoch": 0.03855231237033053,
+      "grad_norm": 6.6875,
+      "learning_rate": 4.9843537500756364e-05,
+      "loss": 0.8866,
+      "step": 878
+    },
+    {
+      "epoch": 0.038640130849534014,
+      "grad_norm": 5.03125,
+      "learning_rate": 4.9842763748064536e-05,
+      "loss": 0.8492,
+      "step": 880
+    },
+    {
+      "epoch": 0.0387279493287375,
+      "grad_norm": 5.21875,
+      "learning_rate": 4.984198809290791e-05,
+      "loss": 0.8236,
+      "step": 882
+    },
+    {
+      "epoch": 0.038815767807940986,
+      "grad_norm": 5.875,
+      "learning_rate": 4.984121053534588e-05,
+      "loss": 0.8429,
+      "step": 884
+    },
+    {
+      "epoch": 0.03890358628714447,
+      "grad_norm": 6.03125,
+      "learning_rate": 4.984043107543799e-05,
+      "loss": 0.864,
+      "step": 886
+    },
+    {
+      "epoch": 0.03899140476634796,
+      "grad_norm": 5.65625,
+      "learning_rate": 4.983964971324393e-05,
+      "loss": 0.8531,
+      "step": 888
+    },
+    {
+      "epoch": 0.039079223245551443,
+      "grad_norm": 6.34375,
+      "learning_rate": 4.983886644882354e-05,
+      "loss": 0.8143,
+      "step": 890
+    },
+    {
+      "epoch": 0.03916704172475493,
+      "grad_norm": 6.53125,
+      "learning_rate": 4.9838081282236814e-05,
+      "loss": 0.8478,
+      "step": 892
+    },
+    {
+      "epoch": 0.039254860203958415,
+      "grad_norm": 6.96875,
+      "learning_rate": 4.983729421354386e-05,
+      "loss": 0.8864,
+      "step": 894
+    },
+    {
+      "epoch": 0.0393426786831619,
+      "grad_norm": 8.5,
+      "learning_rate": 4.9836505242804966e-05,
+      "loss": 0.8228,
+      "step": 896
+    },
+    {
+      "epoch": 0.039430497162365394,
+      "grad_norm": 7.21875,
+      "learning_rate": 4.9835714370080546e-05,
+      "loss": 0.8414,
+      "step": 898
+    },
+    {
+      "epoch": 0.03951831564156888,
+      "grad_norm": 4.96875,
+      "learning_rate": 4.983492159543116e-05,
+      "loss": 0.8727,
+      "step": 900
+    },
+    {
+      "epoch": 0.039606134120772366,
+      "grad_norm": 5.65625,
+      "learning_rate": 4.983412691891753e-05,
+      "loss": 0.8376,
+      "step": 902
+    },
+    {
+      "epoch": 0.03969395259997585,
+      "grad_norm": 5.25,
+      "learning_rate": 4.983333034060051e-05,
+      "loss": 0.8527,
+      "step": 904
+    },
+    {
+      "epoch": 0.03978177107917934,
+      "grad_norm": 6.1875,
+      "learning_rate": 4.9832531860541096e-05,
+      "loss": 0.8227,
+      "step": 906
+    },
+    {
+      "epoch": 0.03986958955838282,
+      "grad_norm": 4.96875,
+      "learning_rate": 4.9831731478800434e-05,
+      "loss": 0.862,
+      "step": 908
+    },
+    {
+      "epoch": 0.03995740803758631,
+      "grad_norm": 5.125,
+      "learning_rate": 4.983092919543983e-05,
+      "loss": 0.8069,
+      "step": 910
+    },
+    {
+      "epoch": 0.040045226516789795,
+      "grad_norm": 4.03125,
+      "learning_rate": 4.983012501052072e-05,
+      "loss": 0.8215,
+      "step": 912
+    },
+    {
+      "epoch": 0.04013304499599328,
+      "grad_norm": 4.0625,
+      "learning_rate": 4.982931892410468e-05,
+      "loss": 0.8391,
+      "step": 914
+    },
+    {
+      "epoch": 0.040220863475196766,
+      "grad_norm": 5.5625,
+      "learning_rate": 4.982851093625344e-05,
+      "loss": 0.8268,
+      "step": 916
+    },
+    {
+      "epoch": 0.04030868195440025,
+      "grad_norm": 5.25,
+      "learning_rate": 4.982770104702888e-05,
+      "loss": 0.8418,
+      "step": 918
+    },
+    {
+      "epoch": 0.04039650043360374,
+      "grad_norm": 4.71875,
+      "learning_rate": 4.9826889256493034e-05,
+      "loss": 0.839,
+      "step": 920
+    },
+    {
+      "epoch": 0.04048431891280723,
+      "grad_norm": 4.5625,
+      "learning_rate": 4.9826075564708056e-05,
+      "loss": 0.8414,
+      "step": 922
+    },
+    {
+      "epoch": 0.04057213739201072,
+      "grad_norm": 5.4375,
+      "learning_rate": 4.982525997173625e-05,
+      "loss": 0.8294,
+      "step": 924
+    },
+    {
+      "epoch": 0.0406599558712142,
+      "grad_norm": 5.75,
+      "learning_rate": 4.982444247764009e-05,
+      "loss": 0.8287,
+      "step": 926
+    },
+    {
+      "epoch": 0.04074777435041769,
+      "grad_norm": 5.21875,
+      "learning_rate": 4.982362308248217e-05,
+      "loss": 0.8332,
+      "step": 928
+    },
+    {
+      "epoch": 0.040835592829621174,
+      "grad_norm": 3.59375,
+      "learning_rate": 4.9822801786325245e-05,
+      "loss": 0.8231,
+      "step": 930
+    },
+    {
+      "epoch": 0.04092341130882466,
+      "grad_norm": 4.34375,
+      "learning_rate": 4.982197858923221e-05,
+      "loss": 0.8413,
+      "step": 932
+    },
+    {
+      "epoch": 0.041011229788028146,
+      "grad_norm": 5.4375,
+      "learning_rate": 4.98211534912661e-05,
+      "loss": 0.8083,
+      "step": 934
+    },
+    {
+      "epoch": 0.04109904826723163,
+      "grad_norm": 6.21875,
+      "learning_rate": 4.982032649249011e-05,
+      "loss": 0.8205,
+      "step": 936
+    },
+    {
+      "epoch": 0.04118686674643512,
+      "grad_norm": 6.15625,
+      "learning_rate": 4.981949759296757e-05,
+      "loss": 0.8354,
+      "step": 938
+    },
+    {
+      "epoch": 0.041274685225638603,
+      "grad_norm": 5.90625,
+      "learning_rate": 4.981866679276195e-05,
+      "loss": 0.8274,
+      "step": 940
+    },
+    {
+      "epoch": 0.04136250370484209,
+      "grad_norm": 4.9375,
+      "learning_rate": 4.981783409193689e-05,
+      "loss": 0.8363,
+      "step": 942
+    },
+    {
+      "epoch": 0.041450322184045575,
+      "grad_norm": 4.65625,
+      "learning_rate": 4.981699949055613e-05,
+      "loss": 0.8658,
+      "step": 944
+    },
+    {
+      "epoch": 0.04153814066324906,
+      "grad_norm": 4.4375,
+      "learning_rate": 4.9816162988683604e-05,
+      "loss": 0.8392,
+      "step": 946
+    },
+    {
+      "epoch": 0.041625959142452554,
+      "grad_norm": 4.59375,
+      "learning_rate": 4.981532458638337e-05,
+      "loss": 0.8418,
+      "step": 948
+    },
+    {
+      "epoch": 0.04171377762165604,
+      "grad_norm": 4.75,
+      "learning_rate": 4.9814484283719634e-05,
+      "loss": 0.8229,
+      "step": 950
+    },
+    {
+      "epoch": 0.041801596100859525,
+      "grad_norm": 5.5,
+      "learning_rate": 4.981364208075673e-05,
+      "loss": 0.8107,
+      "step": 952
+    },
+    {
+      "epoch": 0.04188941458006301,
+      "grad_norm": 4.90625,
+      "learning_rate": 4.9812797977559176e-05,
+      "loss": 0.8279,
+      "step": 954
+    },
+    {
+      "epoch": 0.0419772330592665,
+      "grad_norm": 5.0,
+      "learning_rate": 4.981195197419161e-05,
+      "loss": 0.8215,
+      "step": 956
+    },
+    {
+      "epoch": 0.04206505153846998,
+      "grad_norm": 4.46875,
+      "learning_rate": 4.981110407071881e-05,
+      "loss": 0.8302,
+      "step": 958
+    },
+    {
+      "epoch": 0.04215287001767347,
+      "grad_norm": 3.78125,
+      "learning_rate": 4.981025426720571e-05,
+      "loss": 0.8162,
+      "step": 960
+    },
+    {
+      "epoch": 0.042240688496876955,
+      "grad_norm": 4.15625,
+      "learning_rate": 4.980940256371739e-05,
+      "loss": 0.8318,
+      "step": 962
+    },
+    {
+      "epoch": 0.04232850697608044,
+      "grad_norm": 3.78125,
+      "learning_rate": 4.980854896031908e-05,
+      "loss": 0.8177,
+      "step": 964
+    },
+    {
+      "epoch": 0.042416325455283926,
+      "grad_norm": 4.125,
+      "learning_rate": 4.9807693457076144e-05,
+      "loss": 0.844,
+      "step": 966
+    },
+    {
+      "epoch": 0.04250414393448741,
+      "grad_norm": 4.65625,
+      "learning_rate": 4.980683605405408e-05,
+      "loss": 0.8244,
+      "step": 968
+    },
+    {
+      "epoch": 0.0425919624136909,
+      "grad_norm": 4.46875,
+      "learning_rate": 4.980597675131858e-05,
+      "loss": 0.7891,
+      "step": 970
+    },
+    {
+      "epoch": 0.042679780892894384,
+      "grad_norm": 3.90625,
+      "learning_rate": 4.980511554893543e-05,
+      "loss": 0.8173,
+      "step": 972
+    },
+    {
+      "epoch": 0.04276759937209788,
+      "grad_norm": 4.75,
+      "learning_rate": 4.980425244697059e-05,
+      "loss": 0.8342,
+      "step": 974
+    },
+    {
+      "epoch": 0.04285541785130136,
+      "grad_norm": 4.375,
+      "learning_rate": 4.9803387445490144e-05,
+      "loss": 0.8508,
+      "step": 976
+    },
+    {
+      "epoch": 0.04294323633050485,
+      "grad_norm": 4.8125,
+      "learning_rate": 4.980252054456035e-05,
+      "loss": 0.8379,
+      "step": 978
+    },
+    {
+      "epoch": 0.043031054809708334,
+      "grad_norm": 5.65625,
+      "learning_rate": 4.980165174424759e-05,
+      "loss": 0.8501,
+      "step": 980
+    },
+    {
+      "epoch": 0.04311887328891182,
+      "grad_norm": 5.3125,
+      "learning_rate": 4.980078104461838e-05,
+      "loss": 0.8109,
+      "step": 982
+    },
+    {
+      "epoch": 0.043206691768115306,
+      "grad_norm": 4.375,
+      "learning_rate": 4.979990844573942e-05,
+      "loss": 0.8135,
+      "step": 984
+    },
+    {
+      "epoch": 0.04329451024731879,
+      "grad_norm": 4.4375,
+      "learning_rate": 4.979903394767752e-05,
+      "loss": 0.7713,
+      "step": 986
+    },
+    {
+      "epoch": 0.04338232872652228,
+      "grad_norm": 6.03125,
+      "learning_rate": 4.979815755049967e-05,
+      "loss": 0.8369,
+      "step": 988
+    },
+    {
+      "epoch": 0.04347014720572576,
+      "grad_norm": 5.09375,
+      "learning_rate": 4.9797279254272956e-05,
+      "loss": 0.8063,
+      "step": 990
+    },
+    {
+      "epoch": 0.04355796568492925,
+      "grad_norm": 4.875,
+      "learning_rate": 4.979639905906466e-05,
+      "loss": 0.8432,
+      "step": 992
+    },
+    {
+      "epoch": 0.043645784164132735,
+      "grad_norm": 4.0625,
+      "learning_rate": 4.9795516964942175e-05,
+      "loss": 0.8128,
+      "step": 994
+    },
+    {
+      "epoch": 0.04373360264333622,
+      "grad_norm": 5.6875,
+      "learning_rate": 4.979463297197306e-05,
+      "loss": 0.7988,
+      "step": 996
+    },
+    {
+      "epoch": 0.043821421122539714,
+      "grad_norm": 5.21875,
+      "learning_rate": 4.9793747080225004e-05,
+      "loss": 0.7873,
+      "step": 998
+    },
+    {
+      "epoch": 0.0439092396017432,
+      "grad_norm": 4.0,
+      "learning_rate": 4.979285928976586e-05,
+      "loss": 0.8221,
+      "step": 1000
+    },
+    {
+      "epoch": 0.043997058080946685,
+      "grad_norm": 4.5,
+      "learning_rate": 4.9791969600663605e-05,
+      "loss": 0.8012,
+      "step": 1002
+    },
+    {
+      "epoch": 0.04408487656015017,
+      "grad_norm": 4.5625,
+      "learning_rate": 4.9791078012986375e-05,
+      "loss": 0.8453,
+      "step": 1004
+    },
+    {
+      "epoch": 0.04417269503935366,
+      "grad_norm": 4.84375,
+      "learning_rate": 4.9790184526802444e-05,
+      "loss": 0.8158,
+      "step": 1006
+    },
+    {
+      "epoch": 0.04426051351855714,
+      "grad_norm": 3.984375,
+      "learning_rate": 4.978928914218025e-05,
+      "loss": 0.8095,
+      "step": 1008
+    },
+    {
+      "epoch": 0.04434833199776063,
+      "grad_norm": 4.40625,
+      "learning_rate": 4.9788391859188346e-05,
+      "loss": 0.8196,
+      "step": 1010
+    },
+    {
+      "epoch": 0.044436150476964115,
+      "grad_norm": 3.90625,
+      "learning_rate": 4.9787492677895445e-05,
+      "loss": 0.8302,
+      "step": 1012
+    },
+    {
+      "epoch": 0.0445239689561676,
+      "grad_norm": 4.40625,
+      "learning_rate": 4.978659159837041e-05,
+      "loss": 0.7852,
+      "step": 1014
+    },
+    {
+      "epoch": 0.044611787435371086,
+      "grad_norm": 4.0625,
+      "learning_rate": 4.9785688620682265e-05,
+      "loss": 0.8035,
+      "step": 1016
+    },
+    {
+      "epoch": 0.04469960591457457,
+      "grad_norm": 5.0,
+      "learning_rate": 4.978478374490013e-05,
+      "loss": 0.7945,
+      "step": 1018
+    },
+    {
+      "epoch": 0.04478742439377806,
+      "grad_norm": 4.21875,
+      "learning_rate": 4.978387697109333e-05,
+      "loss": 0.802,
+      "step": 1020
+    },
+    {
+      "epoch": 0.044875242872981544,
+      "grad_norm": 4.875,
+      "learning_rate": 4.978296829933127e-05,
+      "loss": 0.8112,
+      "step": 1022
+    },
+    {
+      "epoch": 0.04496306135218504,
+      "grad_norm": 4.84375,
+      "learning_rate": 4.978205772968357e-05,
+      "loss": 0.8123,
+      "step": 1024
+    },
+    {
+      "epoch": 0.04505087983138852,
+      "grad_norm": 4.84375,
+      "learning_rate": 4.978114526221994e-05,
+      "loss": 0.8189,
+      "step": 1026
+    },
+    {
+      "epoch": 0.04513869831059201,
+      "grad_norm": 4.59375,
+      "learning_rate": 4.978023089701027e-05,
+      "loss": 0.7864,
+      "step": 1028
+    },
+    {
+      "epoch": 0.045226516789795494,
+      "grad_norm": 5.0,
+      "learning_rate": 4.977931463412459e-05,
+      "loss": 0.784,
+      "step": 1030
+    },
+    {
+      "epoch": 0.04531433526899898,
+      "grad_norm": 4.875,
+      "learning_rate": 4.9778396473633035e-05,
+      "loss": 0.8088,
+      "step": 1032
+    },
+    {
+      "epoch": 0.045402153748202466,
+      "grad_norm": 4.40625,
+      "learning_rate": 4.977747641560595e-05,
+      "loss": 0.7992,
+      "step": 1034
+    },
+    {
+      "epoch": 0.04548997222740595,
+      "grad_norm": 4.21875,
+      "learning_rate": 4.977655446011378e-05,
+      "loss": 0.7899,
+      "step": 1036
+    },
+    {
+      "epoch": 0.04557779070660944,
+      "grad_norm": 4.0625,
+      "learning_rate": 4.9775630607227126e-05,
+      "loss": 0.7699,
+      "step": 1038
+    },
+    {
+      "epoch": 0.04566560918581292,
+      "grad_norm": 4.59375,
+      "learning_rate": 4.977470485701674e-05,
+      "loss": 0.8143,
+      "step": 1040
+    },
+    {
+      "epoch": 0.04575342766501641,
+      "grad_norm": 4.0625,
+      "learning_rate": 4.9773777209553517e-05,
+      "loss": 0.8162,
+      "step": 1042
+    },
+    {
+      "epoch": 0.045841246144219895,
+      "grad_norm": 3.90625,
+      "learning_rate": 4.9772847664908505e-05,
+      "loss": 0.7954,
+      "step": 1044
+    },
+    {
+      "epoch": 0.04592906462342338,
+      "grad_norm": 4.125,
+      "learning_rate": 4.977191622315288e-05,
+      "loss": 0.7932,
+      "step": 1046
+    },
+    {
+      "epoch": 0.04601688310262687,
+      "grad_norm": 5.375,
+      "learning_rate": 4.977098288435796e-05,
+      "loss": 0.791,
+      "step": 1048
+    },
+    {
+      "epoch": 0.04610470158183036,
+      "grad_norm": 4.28125,
+      "learning_rate": 4.977004764859524e-05,
+      "loss": 0.7752,
+      "step": 1050
+    },
+    {
+      "epoch": 0.046192520061033845,
+      "grad_norm": 5.34375,
+      "learning_rate": 4.976911051593633e-05,
+      "loss": 0.7977,
+      "step": 1052
+    },
+    {
+      "epoch": 0.04628033854023733,
+      "grad_norm": 4.09375,
+      "learning_rate": 4.976817148645301e-05,
+      "loss": 0.7915,
+      "step": 1054
+    },
+    {
+      "epoch": 0.04636815701944082,
+      "grad_norm": 5.0625,
+      "learning_rate": 4.976723056021717e-05,
+      "loss": 0.8112,
+      "step": 1056
+    },
+    {
+      "epoch": 0.0464559754986443,
+      "grad_norm": 6.65625,
+      "learning_rate": 4.976628773730088e-05,
+      "loss": 0.818,
+      "step": 1058
+    },
+    {
+      "epoch": 0.04654379397784779,
+      "grad_norm": 5.25,
+      "learning_rate": 4.976534301777634e-05,
+      "loss": 0.8005,
+      "step": 1060
+    },
+    {
+      "epoch": 0.046631612457051275,
+      "grad_norm": 4.28125,
+      "learning_rate": 4.9764396401715895e-05,
+      "loss": 0.8215,
+      "step": 1062
+    },
+    {
+      "epoch": 0.04671943093625476,
+      "grad_norm": 4.25,
+      "learning_rate": 4.9763447889192034e-05,
+      "loss": 0.785,
+      "step": 1064
+    },
+    {
+      "epoch": 0.046807249415458246,
+      "grad_norm": 4.4375,
+      "learning_rate": 4.97624974802774e-05,
+      "loss": 0.8075,
+      "step": 1066
+    },
+    {
+      "epoch": 0.04689506789466173,
+      "grad_norm": 4.84375,
+      "learning_rate": 4.9761545175044764e-05,
+      "loss": 0.8031,
+      "step": 1068
+    },
+    {
+      "epoch": 0.04698288637386522,
+      "grad_norm": 3.765625,
+      "learning_rate": 4.976059097356708e-05,
+      "loss": 0.7689,
+      "step": 1070
+    },
+    {
+      "epoch": 0.047070704853068704,
+      "grad_norm": 5.125,
+      "learning_rate": 4.975963487591739e-05,
+      "loss": 0.7636,
+      "step": 1072
+    },
+    {
+      "epoch": 0.0471585233322722,
+      "grad_norm": 3.921875,
+      "learning_rate": 4.9758676882168934e-05,
+      "loss": 0.7856,
+      "step": 1074
+    },
+    {
+      "epoch": 0.04724634181147568,
+      "grad_norm": 5.25,
+      "learning_rate": 4.975771699239505e-05,
+      "loss": 0.7742,
+      "step": 1076
+    },
+    {
+      "epoch": 0.04733416029067917,
+      "grad_norm": 5.84375,
+      "learning_rate": 4.975675520666928e-05,
+      "loss": 0.8068,
+      "step": 1078
+    },
+    {
+      "epoch": 0.047421978769882654,
+      "grad_norm": 6.96875,
+      "learning_rate": 4.9755791525065266e-05,
+      "loss": 0.8261,
+      "step": 1080
+    },
+    {
+      "epoch": 0.04750979724908614,
+      "grad_norm": 7.625,
+      "learning_rate": 4.975482594765679e-05,
+      "loss": 0.8396,
+      "step": 1082
+    },
+    {
+      "epoch": 0.047597615728289626,
+      "grad_norm": 5.46875,
+      "learning_rate": 4.9753858474517815e-05,
+      "loss": 0.7966,
+      "step": 1084
+    },
+    {
+      "epoch": 0.04768543420749311,
+      "grad_norm": 4.28125,
+      "learning_rate": 4.975288910572242e-05,
+      "loss": 0.7745,
+      "step": 1086
+    },
+    {
+      "epoch": 0.0477732526866966,
+      "grad_norm": 4.59375,
+      "learning_rate": 4.975191784134485e-05,
+      "loss": 0.7827,
+      "step": 1088
+    },
+    {
+      "epoch": 0.04786107116590008,
+      "grad_norm": 4.4375,
+      "learning_rate": 4.975094468145948e-05,
+      "loss": 0.7728,
+      "step": 1090
+    },
+    {
+      "epoch": 0.04794888964510357,
+      "grad_norm": 5.0,
+      "learning_rate": 4.974996962614083e-05,
+      "loss": 0.7792,
+      "step": 1092
+    },
+    {
+      "epoch": 0.048036708124307055,
+      "grad_norm": 4.125,
+      "learning_rate": 4.974899267546357e-05,
+      "loss": 0.7991,
+      "step": 1094
+    },
+    {
+      "epoch": 0.04812452660351054,
+      "grad_norm": 7.0625,
+      "learning_rate": 4.974801382950252e-05,
+      "loss": 0.7764,
+      "step": 1096
+    },
+    {
+      "epoch": 0.04821234508271403,
+      "grad_norm": 8.0,
+      "learning_rate": 4.9747033088332635e-05,
+      "loss": 0.8007,
+      "step": 1098
+    },
+    {
+      "epoch": 0.04830016356191752,
+      "grad_norm": 5.84375,
+      "learning_rate": 4.9746050452029023e-05,
+      "loss": 0.8096,
+      "step": 1100
+    },
+    {
+      "epoch": 0.048387982041121005,
+      "grad_norm": 6.5625,
+      "learning_rate": 4.974506592066695e-05,
+      "loss": 0.8101,
+      "step": 1102
+    },
+    {
+      "epoch": 0.04847580052032449,
+      "grad_norm": 8.8125,
+      "learning_rate": 4.974407949432178e-05,
+      "loss": 0.7943,
+      "step": 1104
+    },
+    {
+      "epoch": 0.04856361899952798,
+      "grad_norm": 5.25,
+      "learning_rate": 4.9743091173069075e-05,
+      "loss": 0.8159,
+      "step": 1106
+    },
+    {
+      "epoch": 0.04865143747873146,
+      "grad_norm": 4.03125,
+      "learning_rate": 4.974210095698452e-05,
+      "loss": 0.8159,
+      "step": 1108
+    },
+    {
+      "epoch": 0.04873925595793495,
+      "grad_norm": 4.9375,
+      "learning_rate": 4.9741108846143934e-05,
+      "loss": 0.7488,
+      "step": 1110
+    },
+    {
+      "epoch": 0.048827074437138435,
+      "grad_norm": 6.25,
+      "learning_rate": 4.974011484062331e-05,
+      "loss": 0.7729,
+      "step": 1112
+    },
+    {
+      "epoch": 0.04891489291634192,
+      "grad_norm": 5.0,
+      "learning_rate": 4.9739118940498766e-05,
+      "loss": 0.7887,
+      "step": 1114
+    },
+    {
+      "epoch": 0.049002711395545406,
+      "grad_norm": 4.40625,
+      "learning_rate": 4.973812114584655e-05,
+      "loss": 0.8019,
+      "step": 1116
+    },
+    {
+      "epoch": 0.04909052987474889,
+      "grad_norm": 4.4375,
+      "learning_rate": 4.9737121456743095e-05,
+      "loss": 0.7422,
+      "step": 1118
+    },
+    {
+      "epoch": 0.04917834835395238,
+      "grad_norm": 4.59375,
+      "learning_rate": 4.9736119873264946e-05,
+      "loss": 0.7503,
+      "step": 1120
+    },
+    {
+      "epoch": 0.049266166833155864,
+      "grad_norm": 4.625,
+      "learning_rate": 4.973511639548881e-05,
+      "loss": 0.7732,
+      "step": 1122
+    },
+    {
+      "epoch": 0.04935398531235936,
+      "grad_norm": 46.25,
+      "learning_rate": 4.973411102349153e-05,
+      "loss": 0.8023,
+      "step": 1124
+    },
+    {
+      "epoch": 0.04944180379156284,
+      "grad_norm": 4.375,
+      "learning_rate": 4.9733103757350096e-05,
+      "loss": 0.7675,
+      "step": 1126
+    },
+    {
+      "epoch": 0.04952962227076633,
+      "grad_norm": 4.65625,
+      "learning_rate": 4.9732094597141654e-05,
+      "loss": 0.8107,
+      "step": 1128
+    },
+    {
+      "epoch": 0.049617440749969814,
+      "grad_norm": 4.46875,
+      "learning_rate": 4.973108354294347e-05,
+      "loss": 0.8025,
+      "step": 1130
+    },
+    {
+      "epoch": 0.0497052592291733,
+      "grad_norm": 5.625,
+      "learning_rate": 4.973007059483299e-05,
+      "loss": 0.7767,
+      "step": 1132
+    },
+    {
+      "epoch": 0.049793077708376786,
+      "grad_norm": 6.5,
+      "learning_rate": 4.9729055752887764e-05,
+      "loss": 0.8024,
+      "step": 1134
+    },
+    {
+      "epoch": 0.04988089618758027,
+      "grad_norm": 6.28125,
+      "learning_rate": 4.9728039017185535e-05,
+      "loss": 0.7989,
+      "step": 1136
+    },
+    {
+      "epoch": 0.04996871466678376,
+      "grad_norm": 5.71875,
+      "learning_rate": 4.9727020387804136e-05,
+      "loss": 0.7659,
+      "step": 1138
+    },
+    {
+      "epoch": 0.05005653314598724,
+      "grad_norm": 5.46875,
+      "learning_rate": 4.972599986482159e-05,
+      "loss": 0.7758,
+      "step": 1140
+    },
+    {
+      "epoch": 0.05014435162519073,
+      "grad_norm": 4.65625,
+      "learning_rate": 4.972497744831606e-05,
+      "loss": 0.7691,
+      "step": 1142
+    },
+    {
+      "epoch": 0.050232170104394215,
+      "grad_norm": 4.84375,
+      "learning_rate": 4.972395313836582e-05,
+      "loss": 0.7584,
+      "step": 1144
+    },
+    {
+      "epoch": 0.0503199885835977,
+      "grad_norm": 3.90625,
+      "learning_rate": 4.9722926935049316e-05,
+      "loss": 0.7792,
+      "step": 1146
+    },
+    {
+      "epoch": 0.05040780706280119,
+      "grad_norm": 4.9375,
+      "learning_rate": 4.9721898838445155e-05,
+      "loss": 0.7649,
+      "step": 1148
+    },
+    {
+      "epoch": 0.05049562554200468,
+      "grad_norm": 5.1875,
+      "learning_rate": 4.972086884863204e-05,
+      "loss": 0.7919,
+      "step": 1150
+    },
+    {
+      "epoch": 0.050583444021208165,
+      "grad_norm": 5.34375,
+      "learning_rate": 4.971983696568888e-05,
+      "loss": 0.7349,
+      "step": 1152
+    },
+    {
+      "epoch": 0.05067126250041165,
+      "grad_norm": 4.375,
+      "learning_rate": 4.9718803189694666e-05,
+      "loss": 0.7765,
+      "step": 1154
+    },
+    {
+      "epoch": 0.05075908097961514,
+      "grad_norm": 3.703125,
+      "learning_rate": 4.9717767520728585e-05,
+      "loss": 0.7909,
+      "step": 1156
+    },
+    {
+      "epoch": 0.05084689945881862,
+      "grad_norm": 5.1875,
+      "learning_rate": 4.971672995886994e-05,
+      "loss": 0.7727,
+      "step": 1158
+    },
+    {
+      "epoch": 0.05093471793802211,
+      "grad_norm": 4.53125,
+      "learning_rate": 4.9715690504198186e-05,
+      "loss": 0.7504,
+      "step": 1160
+    },
+    {
+      "epoch": 0.051022536417225595,
+      "grad_norm": 5.125,
+      "learning_rate": 4.971464915679293e-05,
+      "loss": 0.7518,
+      "step": 1162
+    },
+    {
+      "epoch": 0.05111035489642908,
+      "grad_norm": 4.65625,
+      "learning_rate": 4.971360591673392e-05,
+      "loss": 0.764,
+      "step": 1164
+    },
+    {
+      "epoch": 0.051198173375632566,
+      "grad_norm": 5.4375,
+      "learning_rate": 4.971256078410104e-05,
+      "loss": 0.7693,
+      "step": 1166
+    },
+    {
+      "epoch": 0.05128599185483605,
+      "grad_norm": 4.53125,
+      "learning_rate": 4.971151375897434e-05,
+      "loss": 0.7753,
+      "step": 1168
+    },
+    {
+      "epoch": 0.05137381033403954,
+      "grad_norm": 3.640625,
+      "learning_rate": 4.9710464841433984e-05,
+      "loss": 0.7606,
+      "step": 1170
+    },
+    {
+      "epoch": 0.051461628813243024,
+      "grad_norm": 4.0,
+      "learning_rate": 4.9709414031560306e-05,
+      "loss": 0.7781,
+      "step": 1172
+    },
+    {
+      "epoch": 0.05154944729244651,
+      "grad_norm": 4.25,
+      "learning_rate": 4.9708361329433787e-05,
+      "loss": 0.7369,
+      "step": 1174
+    },
+    {
+      "epoch": 0.05163726577165,
+      "grad_norm": 3.90625,
+      "learning_rate": 4.970730673513503e-05,
+      "loss": 0.7699,
+      "step": 1176
+    },
+    {
+      "epoch": 0.05172508425085349,
+      "grad_norm": 5.25,
+      "learning_rate": 4.97062502487448e-05,
+      "loss": 0.778,
+      "step": 1178
+    },
+    {
+      "epoch": 0.051812902730056974,
+      "grad_norm": 6.21875,
+      "learning_rate": 4.9705191870344e-05,
+      "loss": 0.7993,
+      "step": 1180
+    },
+    {
+      "epoch": 0.05190072120926046,
+      "grad_norm": 6.625,
+      "learning_rate": 4.9704131600013686e-05,
+      "loss": 0.7635,
+      "step": 1182
+    },
+    {
+      "epoch": 0.051988539688463946,
+      "grad_norm": 6.40625,
+      "learning_rate": 4.970306943783506e-05,
+      "loss": 0.7772,
+      "step": 1184
+    },
+    {
+      "epoch": 0.05207635816766743,
+      "grad_norm": 6.4375,
+      "learning_rate": 4.9702005383889446e-05,
+      "loss": 0.765,
+      "step": 1186
+    },
+    {
+      "epoch": 0.05216417664687092,
+      "grad_norm": 7.3125,
+      "learning_rate": 4.9700939438258334e-05,
+      "loss": 0.7375,
+      "step": 1188
+    },
+    {
+      "epoch": 0.0522519951260744,
+      "grad_norm": 6.71875,
+      "learning_rate": 4.969987160102337e-05,
+      "loss": 0.7828,
+      "step": 1190
+    },
+    {
+      "epoch": 0.05233981360527789,
+      "grad_norm": 6.28125,
+      "learning_rate": 4.969880187226631e-05,
+      "loss": 0.7674,
+      "step": 1192
+    },
+    {
+      "epoch": 0.052427632084481375,
+      "grad_norm": 7.34375,
+      "learning_rate": 4.969773025206908e-05,
+      "loss": 0.7823,
+      "step": 1194
+    },
+    {
+      "epoch": 0.05251545056368486,
+      "grad_norm": 6.375,
+      "learning_rate": 4.969665674051376e-05,
+      "loss": 0.7578,
+      "step": 1196
+    },
+    {
+      "epoch": 0.05260326904288835,
+      "grad_norm": 4.625,
+      "learning_rate": 4.969558133768254e-05,
+      "loss": 0.738,
+      "step": 1198
+    },
+    {
+      "epoch": 0.05269108752209184,
+      "grad_norm": 5.09375,
+      "learning_rate": 4.969450404365777e-05,
+      "loss": 0.7709,
+      "step": 1200
+    },
+    {
+      "epoch": 0.052778906001295325,
+      "grad_norm": 6.9375,
+      "learning_rate": 4.969342485852197e-05,
+      "loss": 0.7349,
+      "step": 1202
+    },
+    {
+      "epoch": 0.05286672448049881,
+      "grad_norm": 3.828125,
+      "learning_rate": 4.969234378235778e-05,
+      "loss": 0.7782,
+      "step": 1204
+    },
+    {
+      "epoch": 0.0529545429597023,
+      "grad_norm": 4.0,
+      "learning_rate": 4.969126081524798e-05,
+      "loss": 0.7804,
+      "step": 1206
+    },
+    {
+      "epoch": 0.05304236143890578,
+      "grad_norm": 4.375,
+      "learning_rate": 4.969017595727551e-05,
+      "loss": 0.7332,
+      "step": 1208
+    },
+    {
+      "epoch": 0.05313017991810927,
+      "grad_norm": 4.65625,
+      "learning_rate": 4.968908920852344e-05,
+      "loss": 0.7742,
+      "step": 1210
+    },
+    {
+      "epoch": 0.053217998397312755,
+      "grad_norm": 4.40625,
+      "learning_rate": 4.9688000569075e-05,
+      "loss": 0.7555,
+      "step": 1212
+    },
+    {
+      "epoch": 0.05330581687651624,
+      "grad_norm": 3.609375,
+      "learning_rate": 4.9686910039013566e-05,
+      "loss": 0.7543,
+      "step": 1214
+    },
+    {
+      "epoch": 0.053393635355719726,
+      "grad_norm": 4.15625,
+      "learning_rate": 4.9685817618422635e-05,
+      "loss": 0.7569,
+      "step": 1216
+    },
+    {
+      "epoch": 0.05348145383492321,
+      "grad_norm": 4.25,
+      "learning_rate": 4.968472330738588e-05,
+      "loss": 0.7303,
+      "step": 1218
+    },
+    {
+      "epoch": 0.0535692723141267,
+      "grad_norm": 4.0625,
+      "learning_rate": 4.96836271059871e-05,
+      "loss": 0.7597,
+      "step": 1220
+    },
+    {
+      "epoch": 0.053657090793330184,
+      "grad_norm": 4.6875,
+      "learning_rate": 4.968252901431023e-05,
+      "loss": 0.7583,
+      "step": 1222
+    },
+    {
+      "epoch": 0.05374490927253367,
+      "grad_norm": 4.8125,
+      "learning_rate": 4.968142903243938e-05,
+      "loss": 0.754,
+      "step": 1224
+    },
+    {
+      "epoch": 0.05383272775173716,
+      "grad_norm": 4.21875,
+      "learning_rate": 4.968032716045877e-05,
+      "loss": 0.7428,
+      "step": 1226
+    },
+    {
+      "epoch": 0.05392054623094065,
+      "grad_norm": 3.46875,
+      "learning_rate": 4.967922339845279e-05,
+      "loss": 0.7675,
+      "step": 1228
+    },
+    {
+      "epoch": 0.054008364710144134,
+      "grad_norm": 4.125,
+      "learning_rate": 4.967811774650597e-05,
+      "loss": 0.7386,
+      "step": 1230
+    },
+    {
+      "epoch": 0.05409618318934762,
+      "grad_norm": 3.84375,
+      "learning_rate": 4.967701020470298e-05,
+      "loss": 0.7233,
+      "step": 1232
+    },
+    {
+      "epoch": 0.054184001668551106,
+      "grad_norm": 3.90625,
+      "learning_rate": 4.967590077312863e-05,
+      "loss": 0.7297,
+      "step": 1234
+    },
+    {
+      "epoch": 0.05427182014775459,
+      "grad_norm": 3.84375,
+      "learning_rate": 4.967478945186788e-05,
+      "loss": 0.7155,
+      "step": 1236
+    },
+    {
+      "epoch": 0.05435963862695808,
+      "grad_norm": 4.09375,
+      "learning_rate": 4.967367624100584e-05,
+      "loss": 0.7515,
+      "step": 1238
+    },
+    {
+      "epoch": 0.05444745710616156,
+      "grad_norm": 3.578125,
+      "learning_rate": 4.967256114062776e-05,
+      "loss": 0.7572,
+      "step": 1240
+    },
+    {
+      "epoch": 0.05453527558536505,
+      "grad_norm": 4.53125,
+      "learning_rate": 4.967144415081903e-05,
+      "loss": 0.7531,
+      "step": 1242
+    },
+    {
+      "epoch": 0.054623094064568535,
+      "grad_norm": 4.3125,
+      "learning_rate": 4.96703252716652e-05,
+      "loss": 0.7622,
+      "step": 1244
+    },
+    {
+      "epoch": 0.05471091254377202,
+      "grad_norm": 3.53125,
+      "learning_rate": 4.966920450325194e-05,
+      "loss": 0.7755,
+      "step": 1246
+    },
+    {
+      "epoch": 0.05479873102297551,
+      "grad_norm": 3.9375,
+      "learning_rate": 4.9668081845665085e-05,
+      "loss": 0.726,
+      "step": 1248
+    },
+    {
+      "epoch": 0.05488654950217899,
+      "grad_norm": 4.09375,
+      "learning_rate": 4.9666957298990616e-05,
+      "loss": 0.7504,
+      "step": 1250
+    },
+    {
+      "epoch": 0.054974367981382485,
+      "grad_norm": 4.375,
+      "learning_rate": 4.9665830863314645e-05,
+      "loss": 0.7783,
+      "step": 1252
+    },
+    {
+      "epoch": 0.05506218646058597,
+      "grad_norm": 4.15625,
+      "learning_rate": 4.966470253872343e-05,
+      "loss": 0.7782,
+      "step": 1254
+    },
+    {
+      "epoch": 0.05515000493978946,
+      "grad_norm": 3.65625,
+      "learning_rate": 4.9663572325303376e-05,
+      "loss": 0.7509,
+      "step": 1256
+    },
+    {
+      "epoch": 0.05523782341899294,
+      "grad_norm": 3.984375,
+      "learning_rate": 4.966244022314105e-05,
+      "loss": 0.7475,
+      "step": 1258
+    },
+    {
+      "epoch": 0.05532564189819643,
+      "grad_norm": 4.375,
+      "learning_rate": 4.9661306232323134e-05,
+      "loss": 0.7448,
+      "step": 1260
+    },
+    {
+      "epoch": 0.055413460377399915,
+      "grad_norm": 4.375,
+      "learning_rate": 4.966017035293648e-05,
+      "loss": 0.7464,
+      "step": 1262
+    },
+    {
+      "epoch": 0.0555012788566034,
+      "grad_norm": 4.125,
+      "learning_rate": 4.965903258506806e-05,
+      "loss": 0.771,
+      "step": 1264
+    },
+    {
+      "epoch": 0.055589097335806886,
+      "grad_norm": 4.25,
+      "learning_rate": 4.965789292880502e-05,
+      "loss": 0.7344,
+      "step": 1266
+    },
+    {
+      "epoch": 0.05567691581501037,
+      "grad_norm": 3.828125,
+      "learning_rate": 4.965675138423463e-05,
+      "loss": 0.7208,
+      "step": 1268
+    },
+    {
+      "epoch": 0.05576473429421386,
+      "grad_norm": 3.625,
+      "learning_rate": 4.9655607951444305e-05,
+      "loss": 0.7481,
+      "step": 1270
+    },
+    {
+      "epoch": 0.055852552773417344,
+      "grad_norm": 3.59375,
+      "learning_rate": 4.9654462630521615e-05,
+      "loss": 0.7729,
+      "step": 1272
+    },
+    {
+      "epoch": 0.05594037125262083,
+      "grad_norm": 3.609375,
+      "learning_rate": 4.9653315421554266e-05,
+      "loss": 0.7597,
+      "step": 1274
+    },
+    {
+      "epoch": 0.05602818973182432,
+      "grad_norm": 4.625,
+      "learning_rate": 4.965216632463011e-05,
+      "loss": 0.7347,
+      "step": 1276
+    },
+    {
+      "epoch": 0.05611600821102781,
+      "grad_norm": 4.25,
+      "learning_rate": 4.965101533983715e-05,
+      "loss": 0.7614,
+      "step": 1278
+    },
+    {
+      "epoch": 0.056203826690231294,
+      "grad_norm": 4.53125,
+      "learning_rate": 4.9649862467263526e-05,
+      "loss": 0.7808,
+      "step": 1280
+    },
+    {
+      "epoch": 0.05629164516943478,
+      "grad_norm": 3.65625,
+      "learning_rate": 4.964870770699752e-05,
+      "loss": 0.7464,
+      "step": 1282
+    },
+    {
+      "epoch": 0.056379463648638266,
+      "grad_norm": 4.5625,
+      "learning_rate": 4.964755105912758e-05,
+      "loss": 0.7517,
+      "step": 1284
+    },
+    {
+      "epoch": 0.05646728212784175,
+      "grad_norm": 3.96875,
+      "learning_rate": 4.964639252374226e-05,
+      "loss": 0.7742,
+      "step": 1286
+    },
+    {
+      "epoch": 0.05655510060704524,
+      "grad_norm": 4.75,
+      "learning_rate": 4.96452321009303e-05,
+      "loss": 0.7392,
+      "step": 1288
+    },
+    {
+      "epoch": 0.05664291908624872,
+      "grad_norm": 3.390625,
+      "learning_rate": 4.964406979078056e-05,
+      "loss": 0.7295,
+      "step": 1290
+    },
+    {
+      "epoch": 0.05673073756545221,
+      "grad_norm": 3.71875,
+      "learning_rate": 4.964290559338204e-05,
+      "loss": 0.722,
+      "step": 1292
+    },
+    {
+      "epoch": 0.056818556044655695,
+      "grad_norm": 4.03125,
+      "learning_rate": 4.96417395088239e-05,
+      "loss": 0.7417,
+      "step": 1294
+    },
+    {
+      "epoch": 0.05690637452385918,
+      "grad_norm": 3.28125,
+      "learning_rate": 4.964057153719545e-05,
+      "loss": 0.7169,
+      "step": 1296
+    },
+    {
+      "epoch": 0.05699419300306267,
+      "grad_norm": 3.46875,
+      "learning_rate": 4.963940167858613e-05,
+      "loss": 0.7288,
+      "step": 1298
+    },
+    {
+      "epoch": 0.05708201148226615,
+      "grad_norm": 4.03125,
+      "learning_rate": 4.963822993308551e-05,
+      "loss": 0.7436,
+      "step": 1300
+    },
+    {
+      "epoch": 0.057169829961469645,
+      "grad_norm": 4.5,
+      "learning_rate": 4.9637056300783343e-05,
+      "loss": 0.7619,
+      "step": 1302
+    },
+    {
+      "epoch": 0.05725764844067313,
+      "grad_norm": 4.1875,
+      "learning_rate": 4.9635880781769495e-05,
+      "loss": 0.7431,
+      "step": 1304
+    },
+    {
+      "epoch": 0.05734546691987662,
+      "grad_norm": 3.875,
+      "learning_rate": 4.963470337613399e-05,
+      "loss": 0.7413,
+      "step": 1306
+    },
+    {
+      "epoch": 0.0574332853990801,
+      "grad_norm": 4.3125,
+      "learning_rate": 4.9633524083967e-05,
+      "loss": 0.7325,
+      "step": 1308
+    },
+    {
+      "epoch": 0.05752110387828359,
+      "grad_norm": 4.0625,
+      "learning_rate": 4.963234290535883e-05,
+      "loss": 0.7389,
+      "step": 1310
+    },
+    {
+      "epoch": 0.057608922357487075,
+      "grad_norm": 4.21875,
+      "learning_rate": 4.9631159840399935e-05,
+      "loss": 0.7306,
+      "step": 1312
+    },
+    {
+      "epoch": 0.05769674083669056,
+      "grad_norm": 3.953125,
+      "learning_rate": 4.962997488918091e-05,
+      "loss": 0.7374,
+      "step": 1314
+    },
+    {
+      "epoch": 0.057784559315894046,
+      "grad_norm": 4.375,
+      "learning_rate": 4.962878805179251e-05,
+      "loss": 0.7699,
+      "step": 1316
+    },
+    {
+      "epoch": 0.05787237779509753,
+      "grad_norm": 5.5625,
+      "learning_rate": 4.9627599328325606e-05,
+      "loss": 0.7359,
+      "step": 1318
+    },
+    {
+      "epoch": 0.05796019627430102,
+      "grad_norm": 4.375,
+      "learning_rate": 4.962640871887126e-05,
+      "loss": 0.7165,
+      "step": 1320
+    },
+    {
+      "epoch": 0.058048014753504504,
+      "grad_norm": 3.90625,
+      "learning_rate": 4.962521622352061e-05,
+      "loss": 0.7183,
+      "step": 1322
+    },
+    {
+      "epoch": 0.05813583323270799,
+      "grad_norm": 4.09375,
+      "learning_rate": 4.962402184236501e-05,
+      "loss": 0.7297,
+      "step": 1324
+    },
+    {
+      "epoch": 0.05822365171191148,
+      "grad_norm": 3.328125,
+      "learning_rate": 4.962282557549591e-05,
+      "loss": 0.7111,
+      "step": 1326
+    },
+    {
+      "epoch": 0.05831147019111497,
+      "grad_norm": 3.796875,
+      "learning_rate": 4.9621627423004933e-05,
+      "loss": 0.7451,
+      "step": 1328
+    },
+    {
+      "epoch": 0.058399288670318454,
+      "grad_norm": 4.03125,
+      "learning_rate": 4.9620427384983824e-05,
+      "loss": 0.7387,
+      "step": 1330
+    },
+    {
+      "epoch": 0.05848710714952194,
+      "grad_norm": 4.40625,
+      "learning_rate": 4.9619225461524484e-05,
+      "loss": 0.7362,
+      "step": 1332
+    },
+    {
+      "epoch": 0.058574925628725426,
+      "grad_norm": 3.8125,
+      "learning_rate": 4.961802165271895e-05,
+      "loss": 0.7149,
+      "step": 1334
+    },
+    {
+      "epoch": 0.05866274410792891,
+      "grad_norm": 4.09375,
+      "learning_rate": 4.9616815958659425e-05,
+      "loss": 0.715,
+      "step": 1336
+    },
+    {
+      "epoch": 0.0587505625871324,
+      "grad_norm": 3.75,
+      "learning_rate": 4.961560837943823e-05,
+      "loss": 0.7248,
+      "step": 1338
+    },
+    {
+      "epoch": 0.05883838106633588,
+      "grad_norm": 4.09375,
+      "learning_rate": 4.961439891514784e-05,
+      "loss": 0.7568,
+      "step": 1340
+    },
+    {
+      "epoch": 0.05892619954553937,
+      "grad_norm": 5.875,
+      "learning_rate": 4.961318756588088e-05,
+      "loss": 0.7527,
+      "step": 1342
+    },
+    {
+      "epoch": 0.059014018024742855,
+      "grad_norm": 8.0625,
+      "learning_rate": 4.961197433173012e-05,
+      "loss": 0.7473,
+      "step": 1344
+    },
+    {
+      "epoch": 0.05910183650394634,
+      "grad_norm": 9.0,
+      "learning_rate": 4.961075921278846e-05,
+      "loss": 0.7226,
+      "step": 1346
+    },
+    {
+      "epoch": 0.05918965498314983,
+      "grad_norm": 5.8125,
+      "learning_rate": 4.960954220914897e-05,
+      "loss": 0.7375,
+      "step": 1348
+    },
+    {
+      "epoch": 0.05927747346235331,
+      "grad_norm": 5.1875,
+      "learning_rate": 4.9608323320904836e-05,
+      "loss": 0.7501,
+      "step": 1350
+    },
+    {
+      "epoch": 0.059365291941556805,
+      "grad_norm": 5.8125,
+      "learning_rate": 4.9607102548149396e-05,
+      "loss": 0.7003,
+      "step": 1352
+    },
+    {
+      "epoch": 0.05945311042076029,
+      "grad_norm": 6.96875,
+      "learning_rate": 4.960587989097615e-05,
+      "loss": 0.7404,
+      "step": 1354
+    },
+    {
+      "epoch": 0.05954092889996378,
+      "grad_norm": 5.1875,
+      "learning_rate": 4.9604655349478726e-05,
+      "loss": 0.7491,
+      "step": 1356
+    },
+    {
+      "epoch": 0.05962874737916726,
+      "grad_norm": 5.25,
+      "learning_rate": 4.960342892375089e-05,
+      "loss": 0.7291,
+      "step": 1358
+    },
+    {
+      "epoch": 0.05971656585837075,
+      "grad_norm": 5.53125,
+      "learning_rate": 4.960220061388657e-05,
+      "loss": 0.7123,
+      "step": 1360
+    },
+    {
+      "epoch": 0.059804384337574235,
+      "grad_norm": 5.9375,
+      "learning_rate": 4.960097041997984e-05,
+      "loss": 0.7116,
+      "step": 1362
+    },
+    {
+      "epoch": 0.05989220281677772,
+      "grad_norm": 7.65625,
+      "learning_rate": 4.9599738342124884e-05,
+      "loss": 0.7631,
+      "step": 1364
+    },
+    {
+      "epoch": 0.059980021295981206,
+      "grad_norm": 3.625,
+      "learning_rate": 4.959850438041608e-05,
+      "loss": 0.7352,
+      "step": 1366
+    },
+    {
+      "epoch": 0.06006783977518469,
+      "grad_norm": 3.84375,
+      "learning_rate": 4.9597268534947906e-05,
+      "loss": 0.7008,
+      "step": 1368
+    },
+    {
+      "epoch": 0.06015565825438818,
+      "grad_norm": 3.546875,
+      "learning_rate": 4.9596030805815016e-05,
+      "loss": 0.718,
+      "step": 1370
+    },
+    {
+      "epoch": 0.060243476733591664,
+      "grad_norm": 4.09375,
+      "learning_rate": 4.9594791193112186e-05,
+      "loss": 0.7109,
+      "step": 1372
+    },
+    {
+      "epoch": 0.06033129521279515,
+      "grad_norm": 4.125,
+      "learning_rate": 4.959354969693436e-05,
+      "loss": 0.6931,
+      "step": 1374
+    },
+    {
+      "epoch": 0.060419113691998635,
+      "grad_norm": 4.375,
+      "learning_rate": 4.959230631737659e-05,
+      "loss": 0.7214,
+      "step": 1376
+    },
+    {
+      "epoch": 0.06050693217120213,
+      "grad_norm": 4.90625,
+      "learning_rate": 4.9591061054534116e-05,
+      "loss": 0.7364,
+      "step": 1378
+    },
+    {
+      "epoch": 0.060594750650405614,
+      "grad_norm": 5.96875,
+      "learning_rate": 4.9589813908502284e-05,
+      "loss": 0.7526,
+      "step": 1380
+    },
+    {
+      "epoch": 0.0606825691296091,
+      "grad_norm": 6.78125,
+      "learning_rate": 4.958856487937661e-05,
+      "loss": 0.7243,
+      "step": 1382
+    },
+    {
+      "epoch": 0.060770387608812586,
+      "grad_norm": 6.59375,
+      "learning_rate": 4.9587313967252755e-05,
+      "loss": 0.7135,
+      "step": 1384
+    },
+    {
+      "epoch": 0.06085820608801607,
+      "grad_norm": 8.5,
+      "learning_rate": 4.958606117222649e-05,
+      "loss": 0.7682,
+      "step": 1386
+    },
+    {
+      "epoch": 0.06094602456721956,
+      "grad_norm": 4.59375,
+      "learning_rate": 4.958480649439377e-05,
+      "loss": 0.6967,
+      "step": 1388
+    },
+    {
+      "epoch": 0.06103384304642304,
+      "grad_norm": 6.46875,
+      "learning_rate": 4.958354993385068e-05,
+      "loss": 0.737,
+      "step": 1390
+    },
+    {
+      "epoch": 0.06112166152562653,
+      "grad_norm": 6.875,
+      "learning_rate": 4.9582291490693434e-05,
+      "loss": 0.751,
+      "step": 1392
+    },
+    {
+      "epoch": 0.061209480004830015,
+      "grad_norm": 8.0,
+      "learning_rate": 4.958103116501842e-05,
+      "loss": 0.715,
+      "step": 1394
+    },
+    {
+      "epoch": 0.0612972984840335,
+      "grad_norm": 4.8125,
+      "learning_rate": 4.9579768956922145e-05,
+      "loss": 0.7409,
+      "step": 1396
+    },
+    {
+      "epoch": 0.06138511696323699,
+      "grad_norm": 4.3125,
+      "learning_rate": 4.957850486650127e-05,
+      "loss": 0.7097,
+      "step": 1398
+    },
+    {
+      "epoch": 0.06147293544244047,
+      "grad_norm": 3.609375,
+      "learning_rate": 4.957723889385259e-05,
+      "loss": 0.733,
+      "step": 1400
+    },
+    {
+      "epoch": 0.061560753921643965,
+      "grad_norm": 4.5,
+      "learning_rate": 4.957597103907309e-05,
+      "loss": 0.7219,
+      "step": 1402
+    },
+    {
+      "epoch": 0.06164857240084745,
+      "grad_norm": 4.96875,
+      "learning_rate": 4.957470130225982e-05,
+      "loss": 0.7023,
+      "step": 1404
+    },
+    {
+      "epoch": 0.06173639088005094,
+      "grad_norm": 6.0,
+      "learning_rate": 4.957342968351003e-05,
+      "loss": 0.6959,
+      "step": 1406
+    },
+    {
+      "epoch": 0.06182420935925442,
+      "grad_norm": 5.875,
+      "learning_rate": 4.957215618292111e-05,
+      "loss": 0.7014,
+      "step": 1408
+    },
+    {
+      "epoch": 0.06191202783845791,
+      "grad_norm": 7.25,
+      "learning_rate": 4.957088080059058e-05,
+      "loss": 0.7056,
+      "step": 1410
+    },
+    {
+      "epoch": 0.061999846317661395,
+      "grad_norm": 5.8125,
+      "learning_rate": 4.95696035366161e-05,
+      "loss": 0.7327,
+      "step": 1412
+    },
+    {
+      "epoch": 0.06208766479686488,
+      "grad_norm": 3.734375,
+      "learning_rate": 4.95683243910955e-05,
+      "loss": 0.6793,
+      "step": 1414
+    },
+    {
+      "epoch": 0.062175483276068366,
+      "grad_norm": 3.3125,
+      "learning_rate": 4.956704336412673e-05,
+      "loss": 0.6629,
+      "step": 1416
+    },
+    {
+      "epoch": 0.06226330175527185,
+      "grad_norm": 4.15625,
+      "learning_rate": 4.9565760455807887e-05,
+      "loss": 0.7449,
+      "step": 1418
+    },
+    {
+      "epoch": 0.06235112023447534,
+      "grad_norm": 4.28125,
+      "learning_rate": 4.956447566623722e-05,
+      "loss": 0.723,
+      "step": 1420
+    },
+    {
+      "epoch": 0.062438938713678824,
+      "grad_norm": 4.03125,
+      "learning_rate": 4.956318899551311e-05,
+      "loss": 0.695,
+      "step": 1422
+    },
+    {
+      "epoch": 0.06252675719288231,
+      "grad_norm": 4.96875,
+      "learning_rate": 4.956190044373411e-05,
+      "loss": 0.7091,
+      "step": 1424
+    },
+    {
+      "epoch": 0.0626145756720858,
+      "grad_norm": 5.46875,
+      "learning_rate": 4.956061001099888e-05,
+      "loss": 0.716,
+      "step": 1426
+    },
+    {
+      "epoch": 0.06270239415128928,
+      "grad_norm": 4.6875,
+      "learning_rate": 4.955931769740625e-05,
+      "loss": 0.7345,
+      "step": 1428
+    },
+    {
+      "epoch": 0.06279021263049277,
+      "grad_norm": 4.4375,
+      "learning_rate": 4.955802350305518e-05,
+      "loss": 0.7461,
+      "step": 1430
+    },
+    {
+      "epoch": 0.06287803110969625,
+      "grad_norm": 5.15625,
+      "learning_rate": 4.955672742804479e-05,
+      "loss": 0.7486,
+      "step": 1432
+    },
+    {
+      "epoch": 0.06296584958889974,
+      "grad_norm": 4.53125,
+      "learning_rate": 4.955542947247432e-05,
+      "loss": 0.7167,
+      "step": 1434
+    },
+    {
+      "epoch": 0.06305366806810322,
+      "grad_norm": 4.4375,
+      "learning_rate": 4.955412963644318e-05,
+      "loss": 0.6938,
+      "step": 1436
+    },
+    {
+      "epoch": 0.06314148654730671,
+      "grad_norm": 4.5,
+      "learning_rate": 4.9552827920050906e-05,
+      "loss": 0.7398,
+      "step": 1438
+    },
+    {
+      "epoch": 0.06322930502651021,
+      "grad_norm": 4.3125,
+      "learning_rate": 4.955152432339718e-05,
+      "loss": 0.6898,
+      "step": 1440
+    },
+    {
+      "epoch": 0.0633171235057137,
+      "grad_norm": 4.375,
+      "learning_rate": 4.955021884658184e-05,
+      "loss": 0.7295,
+      "step": 1442
+    },
+    {
+      "epoch": 0.06340494198491718,
+      "grad_norm": 3.765625,
+      "learning_rate": 4.9548911489704854e-05,
+      "loss": 0.7075,
+      "step": 1444
+    },
+    {
+      "epoch": 0.06349276046412067,
+      "grad_norm": 4.75,
+      "learning_rate": 4.9547602252866343e-05,
+      "loss": 0.7071,
+      "step": 1446
+    },
+    {
+      "epoch": 0.06358057894332415,
+      "grad_norm": 4.53125,
+      "learning_rate": 4.954629113616656e-05,
+      "loss": 0.6837,
+      "step": 1448
+    },
+    {
+      "epoch": 0.06366839742252764,
+      "grad_norm": 4.5,
+      "learning_rate": 4.954497813970592e-05,
+      "loss": 0.7418,
+      "step": 1450
+    },
+    {
+      "epoch": 0.06375621590173113,
+      "grad_norm": 3.671875,
+      "learning_rate": 4.9543663263584974e-05,
+      "loss": 0.7308,
+      "step": 1452
+    },
+    {
+      "epoch": 0.06384403438093461,
+      "grad_norm": 4.375,
+      "learning_rate": 4.9542346507904415e-05,
+      "loss": 0.7102,
+      "step": 1454
+    },
+    {
+      "epoch": 0.0639318528601381,
+      "grad_norm": 3.640625,
+      "learning_rate": 4.954102787276507e-05,
+      "loss": 0.7463,
+      "step": 1456
+    },
+    {
+      "epoch": 0.06401967133934158,
+      "grad_norm": 4.6875,
+      "learning_rate": 4.9539707358267935e-05,
+      "loss": 0.7124,
+      "step": 1458
+    },
+    {
+      "epoch": 0.06410748981854507,
+      "grad_norm": 4.3125,
+      "learning_rate": 4.9538384964514116e-05,
+      "loss": 0.7378,
+      "step": 1460
+    },
+    {
+      "epoch": 0.06419530829774855,
+      "grad_norm": 4.625,
+      "learning_rate": 4.953706069160491e-05,
+      "loss": 0.7251,
+      "step": 1462
+    },
+    {
+      "epoch": 0.06428312677695204,
+      "grad_norm": 4.6875,
+      "learning_rate": 4.95357345396417e-05,
+      "loss": 0.7246,
+      "step": 1464
+    },
+    {
+      "epoch": 0.06437094525615553,
+      "grad_norm": 4.71875,
+      "learning_rate": 4.9534406508726065e-05,
+      "loss": 0.722,
+      "step": 1466
+    },
+    {
+      "epoch": 0.06445876373535901,
+      "grad_norm": 4.65625,
+      "learning_rate": 4.95330765989597e-05,
+      "loss": 0.7198,
+      "step": 1468
+    },
+    {
+      "epoch": 0.0645465822145625,
+      "grad_norm": 3.6875,
+      "learning_rate": 4.9531744810444443e-05,
+      "loss": 0.691,
+      "step": 1470
+    },
+    {
+      "epoch": 0.06463440069376598,
+      "grad_norm": 4.125,
+      "learning_rate": 4.9530411143282283e-05,
+      "loss": 0.7077,
+      "step": 1472
+    },
+    {
+      "epoch": 0.06472221917296947,
+      "grad_norm": 3.796875,
+      "learning_rate": 4.952907559757537e-05,
+      "loss": 0.7558,
+      "step": 1474
+    },
+    {
+      "epoch": 0.06481003765217296,
+      "grad_norm": 5.65625,
+      "learning_rate": 4.9527738173425965e-05,
+      "loss": 0.7045,
+      "step": 1476
+    },
+    {
+      "epoch": 0.06489785613137644,
+      "grad_norm": 5.96875,
+      "learning_rate": 4.952639887093648e-05,
+      "loss": 0.7126,
+      "step": 1478
+    },
+    {
+      "epoch": 0.06498567461057993,
+      "grad_norm": 5.03125,
+      "learning_rate": 4.95250576902095e-05,
+      "loss": 0.6971,
+      "step": 1480
+    },
+    {
+      "epoch": 0.06507349308978341,
+      "grad_norm": 4.5,
+      "learning_rate": 4.9523714631347716e-05,
+      "loss": 0.7151,
+      "step": 1482
+    },
+    {
+      "epoch": 0.0651613115689869,
+      "grad_norm": 4.0625,
+      "learning_rate": 4.9522369694453996e-05,
+      "loss": 0.7222,
+      "step": 1484
+    },
+    {
+      "epoch": 0.06524913004819038,
+      "grad_norm": 4.96875,
+      "learning_rate": 4.9521022879631325e-05,
+      "loss": 0.6965,
+      "step": 1486
+    },
+    {
+      "epoch": 0.06533694852739387,
+      "grad_norm": 3.71875,
+      "learning_rate": 4.951967418698284e-05,
+      "loss": 0.6897,
+      "step": 1488
+    },
+    {
+      "epoch": 0.06542476700659736,
+      "grad_norm": 3.765625,
+      "learning_rate": 4.951832361661183e-05,
+      "loss": 0.7026,
+      "step": 1490
+    },
+    {
+      "epoch": 0.06551258548580086,
+      "grad_norm": 4.90625,
+      "learning_rate": 4.9516971168621716e-05,
+      "loss": 0.702,
+      "step": 1492
+    },
+    {
+      "epoch": 0.06560040396500434,
+      "grad_norm": 4.96875,
+      "learning_rate": 4.951561684311608e-05,
+      "loss": 0.698,
+      "step": 1494
+    },
+    {
+      "epoch": 0.06568822244420783,
+      "grad_norm": 6.3125,
+      "learning_rate": 4.951426064019862e-05,
+      "loss": 0.7013,
+      "step": 1496
+    },
+    {
+      "epoch": 0.06577604092341131,
+      "grad_norm": 4.53125,
+      "learning_rate": 4.951290255997321e-05,
+      "loss": 0.6693,
+      "step": 1498
+    },
+    {
+      "epoch": 0.0658638594026148,
+      "grad_norm": 4.375,
+      "learning_rate": 4.9511542602543836e-05,
+      "loss": 0.6918,
+      "step": 1500
+    },
+    {
+      "epoch": 0.06595167788181829,
+      "grad_norm": 4.34375,
+      "learning_rate": 4.951018076801467e-05,
+      "loss": 0.6929,
+      "step": 1502
+    },
+    {
+      "epoch": 0.06603949636102177,
+      "grad_norm": 3.890625,
+      "learning_rate": 4.950881705648998e-05,
+      "loss": 0.7204,
+      "step": 1504
+    },
+    {
+      "epoch": 0.06612731484022526,
+      "grad_norm": 4.15625,
+      "learning_rate": 4.9507451468074194e-05,
+      "loss": 0.7144,
+      "step": 1506
+    },
+    {
+      "epoch": 0.06621513331942874,
+      "grad_norm": 3.578125,
+      "learning_rate": 4.950608400287191e-05,
+      "loss": 0.7081,
+      "step": 1508
+    },
+    {
+      "epoch": 0.06630295179863223,
+      "grad_norm": 4.03125,
+      "learning_rate": 4.950471466098784e-05,
+      "loss": 0.7034,
+      "step": 1510
+    },
+    {
+      "epoch": 0.06639077027783571,
+      "grad_norm": 3.421875,
+      "learning_rate": 4.950334344252684e-05,
+      "loss": 0.6886,
+      "step": 1512
+    },
+    {
+      "epoch": 0.0664785887570392,
+      "grad_norm": 3.265625,
+      "learning_rate": 4.950197034759393e-05,
+      "loss": 0.7148,
+      "step": 1514
+    },
+    {
+      "epoch": 0.06656640723624269,
+      "grad_norm": 3.421875,
+      "learning_rate": 4.950059537629425e-05,
+      "loss": 0.7137,
+      "step": 1516
+    },
+    {
+      "epoch": 0.06665422571544617,
+      "grad_norm": 4.625,
+      "learning_rate": 4.949921852873311e-05,
+      "loss": 0.7077,
+      "step": 1518
+    },
+    {
+      "epoch": 0.06674204419464966,
+      "grad_norm": 3.375,
+      "learning_rate": 4.9497839805015945e-05,
+      "loss": 0.7082,
+      "step": 1520
+    },
+    {
+      "epoch": 0.06682986267385314,
+      "grad_norm": 3.671875,
+      "learning_rate": 4.9496459205248325e-05,
+      "loss": 0.6821,
+      "step": 1522
+    },
+    {
+      "epoch": 0.06691768115305663,
+      "grad_norm": 3.53125,
+      "learning_rate": 4.9495076729535994e-05,
+      "loss": 0.697,
+      "step": 1524
+    },
+    {
+      "epoch": 0.06700549963226012,
+      "grad_norm": 4.1875,
+      "learning_rate": 4.9493692377984815e-05,
+      "loss": 0.6875,
+      "step": 1526
+    },
+    {
+      "epoch": 0.0670933181114636,
+      "grad_norm": 4.21875,
+      "learning_rate": 4.94923061507008e-05,
+      "loss": 0.6914,
+      "step": 1528
+    },
+    {
+      "epoch": 0.06718113659066709,
+      "grad_norm": 4.21875,
+      "learning_rate": 4.9490918047790114e-05,
+      "loss": 0.7216,
+      "step": 1530
+    },
+    {
+      "epoch": 0.06726895506987057,
+      "grad_norm": 5.0625,
+      "learning_rate": 4.9489528069359047e-05,
+      "loss": 0.7153,
+      "step": 1532
+    },
+    {
+      "epoch": 0.06735677354907406,
+      "grad_norm": 4.875,
+      "learning_rate": 4.9488136215514045e-05,
+      "loss": 0.7029,
+      "step": 1534
+    },
+    {
+      "epoch": 0.06744459202827754,
+      "grad_norm": 3.984375,
+      "learning_rate": 4.9486742486361714e-05,
+      "loss": 0.6867,
+      "step": 1536
+    },
+    {
+      "epoch": 0.06753241050748103,
+      "grad_norm": 3.5,
+      "learning_rate": 4.9485346882008765e-05,
+      "loss": 0.6959,
+      "step": 1538
+    },
+    {
+      "epoch": 0.06762022898668452,
+      "grad_norm": 6.3125,
+      "learning_rate": 4.948394940256209e-05,
+      "loss": 0.7248,
+      "step": 1540
+    },
+    {
+      "epoch": 0.06770804746588802,
+      "grad_norm": 5.96875,
+      "learning_rate": 4.948255004812869e-05,
+      "loss": 0.7015,
+      "step": 1542
+    },
+    {
+      "epoch": 0.0677958659450915,
+      "grad_norm": 5.5625,
+      "learning_rate": 4.9481148818815746e-05,
+      "loss": 0.7161,
+      "step": 1544
+    },
+    {
+      "epoch": 0.06788368442429499,
+      "grad_norm": 5.0,
+      "learning_rate": 4.947974571473055e-05,
+      "loss": 0.7092,
+      "step": 1546
+    },
+    {
+      "epoch": 0.06797150290349847,
+      "grad_norm": 3.65625,
+      "learning_rate": 4.9478340735980565e-05,
+      "loss": 0.7008,
+      "step": 1548
+    },
+    {
+      "epoch": 0.06805932138270196,
+      "grad_norm": 4.5,
+      "learning_rate": 4.947693388267338e-05,
+      "loss": 0.7142,
+      "step": 1550
+    },
+    {
+      "epoch": 0.06814713986190545,
+      "grad_norm": 4.5,
+      "learning_rate": 4.947552515491673e-05,
+      "loss": 0.6736,
+      "step": 1552
+    },
+    {
+      "epoch": 0.06823495834110893,
+      "grad_norm": 5.3125,
+      "learning_rate": 4.94741145528185e-05,
+      "loss": 0.664,
+      "step": 1554
+    },
+    {
+      "epoch": 0.06832277682031242,
+      "grad_norm": 4.71875,
+      "learning_rate": 4.94727020764867e-05,
+      "loss": 0.7062,
+      "step": 1556
+    },
+    {
+      "epoch": 0.0684105952995159,
+      "grad_norm": 3.78125,
+      "learning_rate": 4.947128772602951e-05,
+      "loss": 0.6652,
+      "step": 1558
+    },
+    {
+      "epoch": 0.06849841377871939,
+      "grad_norm": 4.15625,
+      "learning_rate": 4.946987150155525e-05,
+      "loss": 0.7125,
+      "step": 1560
+    },
+    {
+      "epoch": 0.06858623225792287,
+      "grad_norm": 5.3125,
+      "learning_rate": 4.9468453403172356e-05,
+      "loss": 0.6811,
+      "step": 1562
+    },
+    {
+      "epoch": 0.06867405073712636,
+      "grad_norm": 4.28125,
+      "learning_rate": 4.946703343098944e-05,
+      "loss": 0.7141,
+      "step": 1564
+    },
+    {
+      "epoch": 0.06876186921632985,
+      "grad_norm": 3.578125,
+      "learning_rate": 4.9465611585115235e-05,
+      "loss": 0.7154,
+      "step": 1566
+    },
+    {
+      "epoch": 0.06884968769553333,
+      "grad_norm": 4.0,
+      "learning_rate": 4.946418786565863e-05,
+      "loss": 0.7087,
+      "step": 1568
+    },
+    {
+      "epoch": 0.06893750617473682,
+      "grad_norm": 3.640625,
+      "learning_rate": 4.946276227272865e-05,
+      "loss": 0.6681,
+      "step": 1570
+    },
+    {
+      "epoch": 0.0690253246539403,
+      "grad_norm": 4.71875,
+      "learning_rate": 4.9461334806434475e-05,
+      "loss": 0.6754,
+      "step": 1572
+    },
+    {
+      "epoch": 0.06911314313314379,
+      "grad_norm": 5.8125,
+      "learning_rate": 4.945990546688542e-05,
+      "loss": 0.6888,
+      "step": 1574
+    },
+    {
+      "epoch": 0.06920096161234728,
+      "grad_norm": 5.21875,
+      "learning_rate": 4.945847425419094e-05,
+      "loss": 0.6945,
+      "step": 1576
+    },
+    {
+      "epoch": 0.06928878009155076,
+      "grad_norm": 4.5625,
+      "learning_rate": 4.945704116846064e-05,
+      "loss": 0.6836,
+      "step": 1578
+    },
+    {
+      "epoch": 0.06937659857075425,
+      "grad_norm": 3.828125,
+      "learning_rate": 4.945560620980426e-05,
+      "loss": 0.6874,
+      "step": 1580
+    },
+    {
+      "epoch": 0.06946441704995773,
+      "grad_norm": 3.65625,
+      "learning_rate": 4.945416937833169e-05,
+      "loss": 0.6591,
+      "step": 1582
+    },
+    {
+      "epoch": 0.06955223552916122,
+      "grad_norm": 3.125,
+      "learning_rate": 4.945273067415298e-05,
+      "loss": 0.68,
+      "step": 1584
+    },
+    {
+      "epoch": 0.0696400540083647,
+      "grad_norm": 3.078125,
+      "learning_rate": 4.945129009737828e-05,
+      "loss": 0.6588,
+      "step": 1586
+    },
+    {
+      "epoch": 0.06972787248756819,
+      "grad_norm": 3.515625,
+      "learning_rate": 4.944984764811793e-05,
+      "loss": 0.667,
+      "step": 1588
+    },
+    {
+      "epoch": 0.06981569096677168,
+      "grad_norm": 4.625,
+      "learning_rate": 4.9448403326482386e-05,
+      "loss": 0.6798,
+      "step": 1590
+    },
+    {
+      "epoch": 0.06990350944597518,
+      "grad_norm": 5.28125,
+      "learning_rate": 4.944695713258225e-05,
+      "loss": 0.6867,
+      "step": 1592
+    },
+    {
+      "epoch": 0.06999132792517866,
+      "grad_norm": 4.84375,
+      "learning_rate": 4.944550906652828e-05,
+      "loss": 0.6764,
+      "step": 1594
+    },
+    {
+      "epoch": 0.07007914640438215,
+      "grad_norm": 6.84375,
+      "learning_rate": 4.944405912843136e-05,
+      "loss": 0.6841,
+      "step": 1596
+    },
+    {
+      "epoch": 0.07016696488358563,
+      "grad_norm": 8.1875,
+      "learning_rate": 4.9442607318402543e-05,
+      "loss": 0.7033,
+      "step": 1598
+    },
+    {
+      "epoch": 0.07025478336278912,
+      "grad_norm": 4.46875,
+      "learning_rate": 4.944115363655299e-05,
+      "loss": 0.6732,
+      "step": 1600
+    },
+    {
+      "epoch": 0.0703426018419926,
+      "grad_norm": 3.78125,
+      "learning_rate": 4.943969808299404e-05,
+      "loss": 0.6968,
+      "step": 1602
+    },
+    {
+      "epoch": 0.07043042032119609,
+      "grad_norm": 3.71875,
+      "learning_rate": 4.943824065783714e-05,
+      "loss": 0.6775,
+      "step": 1604
+    },
+    {
+      "epoch": 0.07051823880039958,
+      "grad_norm": 4.125,
+      "learning_rate": 4.9436781361193926e-05,
+      "loss": 0.646,
+      "step": 1606
+    },
+    {
+      "epoch": 0.07060605727960306,
+      "grad_norm": 6.15625,
+      "learning_rate": 4.943532019317613e-05,
+      "loss": 0.7055,
+      "step": 1608
+    },
+    {
+      "epoch": 0.07069387575880655,
+      "grad_norm": 4.8125,
+      "learning_rate": 4.943385715389566e-05,
+      "loss": 0.6727,
+      "step": 1610
+    },
+    {
+      "epoch": 0.07078169423801003,
+      "grad_norm": 4.65625,
+      "learning_rate": 4.9432392243464546e-05,
+      "loss": 0.6792,
+      "step": 1612
+    },
+    {
+      "epoch": 0.07086951271721352,
+      "grad_norm": 3.921875,
+      "learning_rate": 4.943092546199498e-05,
+      "loss": 0.651,
+      "step": 1614
+    },
+    {
+      "epoch": 0.070957331196417,
+      "grad_norm": 4.0625,
+      "learning_rate": 4.9429456809599286e-05,
+      "loss": 0.6823,
+      "step": 1616
+    },
+    {
+      "epoch": 0.07104514967562049,
+      "grad_norm": 3.828125,
+      "learning_rate": 4.942798628638994e-05,
+      "loss": 0.6846,
+      "step": 1618
+    },
+    {
+      "epoch": 0.07113296815482398,
+      "grad_norm": 3.90625,
+      "learning_rate": 4.942651389247954e-05,
+      "loss": 0.6784,
+      "step": 1620
+    },
+    {
+      "epoch": 0.07122078663402746,
+      "grad_norm": 3.28125,
+      "learning_rate": 4.942503962798085e-05,
+      "loss": 0.6826,
+      "step": 1622
+    },
+    {
+      "epoch": 0.07130860511323095,
+      "grad_norm": 3.453125,
+      "learning_rate": 4.9423563493006776e-05,
+      "loss": 0.7006,
+      "step": 1624
+    },
+    {
+      "epoch": 0.07139642359243444,
+      "grad_norm": 4.1875,
+      "learning_rate": 4.9422085487670344e-05,
+      "loss": 0.6968,
+      "step": 1626
+    },
+    {
+      "epoch": 0.07148424207163792,
+      "grad_norm": 4.34375,
+      "learning_rate": 4.942060561208476e-05,
+      "loss": 0.7066,
+      "step": 1628
+    },
+    {
+      "epoch": 0.07157206055084141,
+      "grad_norm": 4.8125,
+      "learning_rate": 4.941912386636335e-05,
+      "loss": 0.6673,
+      "step": 1630
+    },
+    {
+      "epoch": 0.07165987903004489,
+      "grad_norm": 4.40625,
+      "learning_rate": 4.941764025061957e-05,
+      "loss": 0.6775,
+      "step": 1632
+    },
+    {
+      "epoch": 0.07174769750924838,
+      "grad_norm": 4.65625,
+      "learning_rate": 4.9416154764967046e-05,
+      "loss": 0.6848,
+      "step": 1634
+    },
+    {
+      "epoch": 0.07183551598845186,
+      "grad_norm": 3.90625,
+      "learning_rate": 4.941466740951954e-05,
+      "loss": 0.6707,
+      "step": 1636
+    },
+    {
+      "epoch": 0.07192333446765535,
+      "grad_norm": 3.5,
+      "learning_rate": 4.941317818439095e-05,
+      "loss": 0.6565,
+      "step": 1638
+    },
+    {
+      "epoch": 0.07201115294685884,
+      "grad_norm": 3.734375,
+      "learning_rate": 4.941168708969533e-05,
+      "loss": 0.6783,
+      "step": 1640
+    },
+    {
+      "epoch": 0.07209897142606234,
+      "grad_norm": 3.71875,
+      "learning_rate": 4.941019412554686e-05,
+      "loss": 0.6587,
+      "step": 1642
+    },
+    {
+      "epoch": 0.07218678990526582,
+      "grad_norm": 3.59375,
+      "learning_rate": 4.9408699292059865e-05,
+      "loss": 0.6873,
+      "step": 1644
+    },
+    {
+      "epoch": 0.07227460838446931,
+      "grad_norm": 4.25,
+      "learning_rate": 4.940720258934883e-05,
+      "loss": 0.6964,
+      "step": 1646
+    },
+    {
+      "epoch": 0.0723624268636728,
+      "grad_norm": 4.4375,
+      "learning_rate": 4.940570401752836e-05,
+      "loss": 0.6669,
+      "step": 1648
+    },
+    {
+      "epoch": 0.07245024534287628,
+      "grad_norm": 4.25,
+      "learning_rate": 4.940420357671324e-05,
+      "loss": 0.7105,
+      "step": 1650
+    },
+    {
+      "epoch": 0.07253806382207977,
+      "grad_norm": 3.4375,
+      "learning_rate": 4.940270126701836e-05,
+      "loss": 0.6843,
+      "step": 1652
+    },
+    {
+      "epoch": 0.07262588230128325,
+      "grad_norm": 3.6875,
+      "learning_rate": 4.940119708855876e-05,
+      "loss": 0.6931,
+      "step": 1654
+    },
+    {
+      "epoch": 0.07271370078048674,
+      "grad_norm": 3.96875,
+      "learning_rate": 4.939969104144964e-05,
+      "loss": 0.6722,
+      "step": 1656
+    },
+    {
+      "epoch": 0.07280151925969022,
+      "grad_norm": 5.46875,
+      "learning_rate": 4.939818312580633e-05,
+      "loss": 0.708,
+      "step": 1658
+    },
+    {
+      "epoch": 0.07288933773889371,
+      "grad_norm": 5.3125,
+      "learning_rate": 4.939667334174431e-05,
+      "loss": 0.7094,
+      "step": 1660
+    },
+    {
+      "epoch": 0.0729771562180972,
+      "grad_norm": 4.28125,
+      "learning_rate": 4.93951616893792e-05,
+      "loss": 0.7017,
+      "step": 1662
+    },
+    {
+      "epoch": 0.07306497469730068,
+      "grad_norm": 4.5,
+      "learning_rate": 4.939364816882676e-05,
+      "loss": 0.6851,
+      "step": 1664
+    },
+    {
+      "epoch": 0.07315279317650417,
+      "grad_norm": 3.640625,
+      "learning_rate": 4.939213278020288e-05,
+      "loss": 0.6595,
+      "step": 1666
+    },
+    {
+      "epoch": 0.07324061165570765,
+      "grad_norm": 3.359375,
+      "learning_rate": 4.939061552362364e-05,
+      "loss": 0.6878,
+      "step": 1668
+    },
+    {
+      "epoch": 0.07332843013491114,
+      "grad_norm": 3.515625,
+      "learning_rate": 4.938909639920521e-05,
+      "loss": 0.6921,
+      "step": 1670
+    },
+    {
+      "epoch": 0.07341624861411462,
+      "grad_norm": 3.59375,
+      "learning_rate": 4.9387575407063936e-05,
+      "loss": 0.6692,
+      "step": 1672
+    },
+    {
+      "epoch": 0.07350406709331811,
+      "grad_norm": 3.796875,
+      "learning_rate": 4.938605254731629e-05,
+      "loss": 0.6822,
+      "step": 1674
+    },
+    {
+      "epoch": 0.0735918855725216,
+      "grad_norm": 3.984375,
+      "learning_rate": 4.93845278200789e-05,
+      "loss": 0.6804,
+      "step": 1676
+    },
+    {
+      "epoch": 0.07367970405172508,
+      "grad_norm": 4.25,
+      "learning_rate": 4.938300122546851e-05,
+      "loss": 0.6923,
+      "step": 1678
+    },
+    {
+      "epoch": 0.07376752253092857,
+      "grad_norm": 4.8125,
+      "learning_rate": 4.938147276360205e-05,
+      "loss": 0.6968,
+      "step": 1680
+    },
+    {
+      "epoch": 0.07385534101013205,
+      "grad_norm": 4.71875,
+      "learning_rate": 4.937994243459656e-05,
+      "loss": 0.6804,
+      "step": 1682
+    },
+    {
+      "epoch": 0.07394315948933554,
+      "grad_norm": 4.21875,
+      "learning_rate": 4.937841023856923e-05,
+      "loss": 0.6759,
+      "step": 1684
+    },
+    {
+      "epoch": 0.07403097796853902,
+      "grad_norm": 4.90625,
+      "learning_rate": 4.937687617563741e-05,
+      "loss": 0.6791,
+      "step": 1686
+    },
+    {
+      "epoch": 0.07411879644774251,
+      "grad_norm": 4.53125,
+      "learning_rate": 4.937534024591856e-05,
+      "loss": 0.6773,
+      "step": 1688
+    },
+    {
+      "epoch": 0.074206614926946,
+      "grad_norm": 5.78125,
+      "learning_rate": 4.9373802449530316e-05,
+      "loss": 0.7157,
+      "step": 1690
+    },
+    {
+      "epoch": 0.07429443340614948,
+      "grad_norm": 5.34375,
+      "learning_rate": 4.9372262786590436e-05,
+      "loss": 0.6767,
+      "step": 1692
+    },
+    {
+      "epoch": 0.07438225188535298,
+      "grad_norm": 4.15625,
+      "learning_rate": 4.9370721257216824e-05,
+      "loss": 0.6637,
+      "step": 1694
+    },
+    {
+      "epoch": 0.07447007036455647,
+      "grad_norm": 4.21875,
+      "learning_rate": 4.936917786152754e-05,
+      "loss": 0.6797,
+      "step": 1696
+    },
+    {
+      "epoch": 0.07455788884375995,
+      "grad_norm": 4.21875,
+      "learning_rate": 4.936763259964078e-05,
+      "loss": 0.7028,
+      "step": 1698
+    },
+    {
+      "epoch": 0.07464570732296344,
+      "grad_norm": 4.15625,
+      "learning_rate": 4.9366085471674864e-05,
+      "loss": 0.6994,
+      "step": 1700
+    },
+    {
+      "epoch": 0.07473352580216693,
+      "grad_norm": 4.1875,
+      "learning_rate": 4.936453647774829e-05,
+      "loss": 0.693,
+      "step": 1702
+    },
+    {
+      "epoch": 0.07482134428137041,
+      "grad_norm": 3.9375,
+      "learning_rate": 4.936298561797966e-05,
+      "loss": 0.6734,
+      "step": 1704
+    },
+    {
+      "epoch": 0.0749091627605739,
+      "grad_norm": 3.609375,
+      "learning_rate": 4.936143289248776e-05,
+      "loss": 0.681,
+      "step": 1706
+    },
+    {
+      "epoch": 0.07499698123977738,
+      "grad_norm": 4.15625,
+      "learning_rate": 4.9359878301391495e-05,
+      "loss": 0.6764,
+      "step": 1708
+    },
+    {
+      "epoch": 0.07508479971898087,
+      "grad_norm": 4.625,
+      "learning_rate": 4.93583218448099e-05,
+      "loss": 0.6641,
+      "step": 1710
+    },
+    {
+      "epoch": 0.07517261819818435,
+      "grad_norm": 4.0,
+      "learning_rate": 4.935676352286218e-05,
+      "loss": 0.6791,
+      "step": 1712
+    },
+    {
+      "epoch": 0.07526043667738784,
+      "grad_norm": 5.03125,
+      "learning_rate": 4.935520333566768e-05,
+      "loss": 0.6714,
+      "step": 1714
+    },
+    {
+      "epoch": 0.07534825515659133,
+      "grad_norm": 4.90625,
+      "learning_rate": 4.9353641283345863e-05,
+      "loss": 0.7237,
+      "step": 1716
+    },
+    {
+      "epoch": 0.07543607363579481,
+      "grad_norm": 4.46875,
+      "learning_rate": 4.9352077366016355e-05,
+      "loss": 0.6888,
+      "step": 1718
+    },
+    {
+      "epoch": 0.0755238921149983,
+      "grad_norm": 4.0,
+      "learning_rate": 4.935051158379893e-05,
+      "loss": 0.6826,
+      "step": 1720
+    },
+    {
+      "epoch": 0.07561171059420178,
+      "grad_norm": 3.25,
+      "learning_rate": 4.934894393681349e-05,
+      "loss": 0.666,
+      "step": 1722
+    },
+    {
+      "epoch": 0.07569952907340527,
+      "grad_norm": 4.0625,
+      "learning_rate": 4.934737442518009e-05,
+      "loss": 0.6854,
+      "step": 1724
+    },
+    {
+      "epoch": 0.07578734755260876,
+      "grad_norm": 3.21875,
+      "learning_rate": 4.9345803049018914e-05,
+      "loss": 0.692,
+      "step": 1726
+    },
+    {
+      "epoch": 0.07587516603181224,
+      "grad_norm": 3.59375,
+      "learning_rate": 4.9344229808450305e-05,
+      "loss": 0.6561,
+      "step": 1728
+    },
+    {
+      "epoch": 0.07596298451101573,
+      "grad_norm": 4.125,
+      "learning_rate": 4.934265470359474e-05,
+      "loss": 0.6502,
+      "step": 1730
+    },
+    {
+      "epoch": 0.07605080299021921,
+      "grad_norm": 3.859375,
+      "learning_rate": 4.934107773457285e-05,
+      "loss": 0.6824,
+      "step": 1732
+    },
+    {
+      "epoch": 0.0761386214694227,
+      "grad_norm": 3.625,
+      "learning_rate": 4.933949890150539e-05,
+      "loss": 0.6821,
+      "step": 1734
+    },
+    {
+      "epoch": 0.07622643994862618,
+      "grad_norm": 3.90625,
+      "learning_rate": 4.933791820451327e-05,
+      "loss": 0.6656,
+      "step": 1736
+    },
+    {
+      "epoch": 0.07631425842782967,
+      "grad_norm": 3.65625,
+      "learning_rate": 4.933633564371753e-05,
+      "loss": 0.6663,
+      "step": 1738
+    },
+    {
+      "epoch": 0.07640207690703316,
+      "grad_norm": 3.625,
+      "learning_rate": 4.933475121923938e-05,
+      "loss": 0.6612,
+      "step": 1740
+    },
+    {
+      "epoch": 0.07648989538623664,
+      "grad_norm": 3.46875,
+      "learning_rate": 4.933316493120015e-05,
+      "loss": 0.662,
+      "step": 1742
+    },
+    {
+      "epoch": 0.07657771386544014,
+      "grad_norm": 4.21875,
+      "learning_rate": 4.9331576779721314e-05,
+      "loss": 0.6569,
+      "step": 1744
+    },
+    {
+      "epoch": 0.07666553234464363,
+      "grad_norm": 3.921875,
+      "learning_rate": 4.93299867649245e-05,
+      "loss": 0.6772,
+      "step": 1746
+    },
+    {
+      "epoch": 0.07675335082384711,
+      "grad_norm": 3.078125,
+      "learning_rate": 4.9328394886931456e-05,
+      "loss": 0.6619,
+      "step": 1748
+    },
+    {
+      "epoch": 0.0768411693030506,
+      "grad_norm": 3.609375,
+      "learning_rate": 4.932680114586411e-05,
+      "loss": 0.6713,
+      "step": 1750
+    },
+    {
+      "epoch": 0.07692898778225409,
+      "grad_norm": 3.25,
+      "learning_rate": 4.9325205541844497e-05,
+      "loss": 0.6795,
+      "step": 1752
+    },
+    {
+      "epoch": 0.07701680626145757,
+      "grad_norm": 3.359375,
+      "learning_rate": 4.932360807499481e-05,
+      "loss": 0.6586,
+      "step": 1754
+    },
+    {
+      "epoch": 0.07710462474066106,
+      "grad_norm": 3.515625,
+      "learning_rate": 4.9322008745437385e-05,
+      "loss": 0.675,
+      "step": 1756
+    },
+    {
+      "epoch": 0.07719244321986454,
+      "grad_norm": 3.40625,
+      "learning_rate": 4.932040755329471e-05,
+      "loss": 0.6854,
+      "step": 1758
+    },
+    {
+      "epoch": 0.07728026169906803,
+      "grad_norm": 3.28125,
+      "learning_rate": 4.9318804498689384e-05,
+      "loss": 0.694,
+      "step": 1760
+    },
+    {
+      "epoch": 0.07736808017827151,
+      "grad_norm": 3.296875,
+      "learning_rate": 4.9317199581744187e-05,
+      "loss": 0.6677,
+      "step": 1762
+    },
+    {
+      "epoch": 0.077455898657475,
+      "grad_norm": 3.4375,
+      "learning_rate": 4.931559280258201e-05,
+      "loss": 0.6727,
+      "step": 1764
+    },
+    {
+      "epoch": 0.07754371713667849,
+      "grad_norm": 3.953125,
+      "learning_rate": 4.931398416132591e-05,
+      "loss": 0.6514,
+      "step": 1766
+    },
+    {
+      "epoch": 0.07763153561588197,
+      "grad_norm": 3.640625,
+      "learning_rate": 4.9312373658099076e-05,
+      "loss": 0.6559,
+      "step": 1768
+    },
+    {
+      "epoch": 0.07771935409508546,
+      "grad_norm": 4.21875,
+      "learning_rate": 4.931076129302484e-05,
+      "loss": 0.633,
+      "step": 1770
+    },
+    {
+      "epoch": 0.07780717257428894,
+      "grad_norm": 3.625,
+      "learning_rate": 4.930914706622668e-05,
+      "loss": 0.6615,
+      "step": 1772
+    },
+    {
+      "epoch": 0.07789499105349243,
+      "grad_norm": 3.75,
+      "learning_rate": 4.93075309778282e-05,
+      "loss": 0.6607,
+      "step": 1774
+    },
+    {
+      "epoch": 0.07798280953269592,
+      "grad_norm": 3.40625,
+      "learning_rate": 4.930591302795318e-05,
+      "loss": 0.652,
+      "step": 1776
+    },
+    {
+      "epoch": 0.0780706280118994,
+      "grad_norm": 3.40625,
+      "learning_rate": 4.9304293216725505e-05,
+      "loss": 0.7057,
+      "step": 1778
+    },
+    {
+      "epoch": 0.07815844649110289,
+      "grad_norm": 3.453125,
+      "learning_rate": 4.930267154426924e-05,
+      "loss": 0.6711,
+      "step": 1780
+    },
+    {
+      "epoch": 0.07824626497030637,
+      "grad_norm": 3.796875,
+      "learning_rate": 4.9301048010708556e-05,
+      "loss": 0.6659,
+      "step": 1782
+    },
+    {
+      "epoch": 0.07833408344950986,
+      "grad_norm": 3.859375,
+      "learning_rate": 4.929942261616779e-05,
+      "loss": 0.6763,
+      "step": 1784
+    },
+    {
+      "epoch": 0.07842190192871334,
+      "grad_norm": 3.328125,
+      "learning_rate": 4.929779536077142e-05,
+      "loss": 0.6799,
+      "step": 1786
+    },
+    {
+      "epoch": 0.07850972040791683,
+      "grad_norm": 3.328125,
+      "learning_rate": 4.929616624464405e-05,
+      "loss": 0.6641,
+      "step": 1788
+    },
+    {
+      "epoch": 0.07859753888712032,
+      "grad_norm": 3.65625,
+      "learning_rate": 4.9294535267910446e-05,
+      "loss": 0.6557,
+      "step": 1790
+    },
+    {
+      "epoch": 0.0786853573663238,
+      "grad_norm": 5.125,
+      "learning_rate": 4.929290243069551e-05,
+      "loss": 0.6742,
+      "step": 1792
+    },
+    {
+      "epoch": 0.0787731758455273,
+      "grad_norm": 3.484375,
+      "learning_rate": 4.929126773312428e-05,
+      "loss": 0.6501,
+      "step": 1794
+    },
+    {
+      "epoch": 0.07886099432473079,
+      "grad_norm": 3.90625,
+      "learning_rate": 4.928963117532195e-05,
+      "loss": 0.671,
+      "step": 1796
+    },
+    {
+      "epoch": 0.07894881280393427,
+      "grad_norm": 3.25,
+      "learning_rate": 4.928799275741384e-05,
+      "loss": 0.6796,
+      "step": 1798
+    },
+    {
+      "epoch": 0.07903663128313776,
+      "grad_norm": 3.40625,
+      "learning_rate": 4.928635247952541e-05,
+      "loss": 0.6689,
+      "step": 1800
+    },
+    {
+      "epoch": 0.07912444976234125,
+      "grad_norm": 4.71875,
+      "learning_rate": 4.92847103417823e-05,
+      "loss": 0.6507,
+      "step": 1802
+    },
+    {
+      "epoch": 0.07921226824154473,
+      "grad_norm": 4.71875,
+      "learning_rate": 4.928306634431025e-05,
+      "loss": 0.6544,
+      "step": 1804
+    },
+    {
+      "epoch": 0.07930008672074822,
+      "grad_norm": 4.34375,
+      "learning_rate": 4.9281420487235144e-05,
+      "loss": 0.6738,
+      "step": 1806
+    },
+    {
+      "epoch": 0.0793879051999517,
+      "grad_norm": 3.984375,
+      "learning_rate": 4.927977277068305e-05,
+      "loss": 0.6526,
+      "step": 1808
+    },
+    {
+      "epoch": 0.07947572367915519,
+      "grad_norm": 3.890625,
+      "learning_rate": 4.9278123194780134e-05,
+      "loss": 0.7137,
+      "step": 1810
+    },
+    {
+      "epoch": 0.07956354215835867,
+      "grad_norm": 3.234375,
+      "learning_rate": 4.927647175965272e-05,
+      "loss": 0.6777,
+      "step": 1812
+    },
+    {
+      "epoch": 0.07965136063756216,
+      "grad_norm": 3.765625,
+      "learning_rate": 4.9274818465427285e-05,
+      "loss": 0.6403,
+      "step": 1814
+    },
+    {
+      "epoch": 0.07973917911676565,
+      "grad_norm": 3.828125,
+      "learning_rate": 4.927316331223043e-05,
+      "loss": 0.6577,
+      "step": 1816
+    },
+    {
+      "epoch": 0.07982699759596913,
+      "grad_norm": 4.25,
+      "learning_rate": 4.927150630018891e-05,
+      "loss": 0.646,
+      "step": 1818
+    },
+    {
+      "epoch": 0.07991481607517262,
+      "grad_norm": 3.375,
+      "learning_rate": 4.926984742942961e-05,
+      "loss": 0.6679,
+      "step": 1820
+    },
+    {
+      "epoch": 0.0800026345543761,
+      "grad_norm": 3.5625,
+      "learning_rate": 4.9268186700079594e-05,
+      "loss": 0.6644,
+      "step": 1822
+    },
+    {
+      "epoch": 0.08009045303357959,
+      "grad_norm": 3.53125,
+      "learning_rate": 4.926652411226601e-05,
+      "loss": 0.6556,
+      "step": 1824
+    },
+    {
+      "epoch": 0.08017827151278308,
+      "grad_norm": 3.578125,
+      "learning_rate": 4.92648596661162e-05,
+      "loss": 0.6853,
+      "step": 1826
+    },
+    {
+      "epoch": 0.08026608999198656,
+      "grad_norm": 4.40625,
+      "learning_rate": 4.926319336175762e-05,
+      "loss": 0.6583,
+      "step": 1828
+    },
+    {
+      "epoch": 0.08035390847119005,
+      "grad_norm": 4.21875,
+      "learning_rate": 4.926152519931787e-05,
+      "loss": 0.6808,
+      "step": 1830
+    },
+    {
+      "epoch": 0.08044172695039353,
+      "grad_norm": 4.28125,
+      "learning_rate": 4.925985517892471e-05,
+      "loss": 0.7052,
+      "step": 1832
+    },
+    {
+      "epoch": 0.08052954542959702,
+      "grad_norm": 3.328125,
+      "learning_rate": 4.9258183300706016e-05,
+      "loss": 0.6652,
+      "step": 1834
+    },
+    {
+      "epoch": 0.0806173639088005,
+      "grad_norm": 3.3125,
+      "learning_rate": 4.9256509564789836e-05,
+      "loss": 0.6473,
+      "step": 1836
+    },
+    {
+      "epoch": 0.08070518238800399,
+      "grad_norm": 3.265625,
+      "learning_rate": 4.9254833971304334e-05,
+      "loss": 0.6697,
+      "step": 1838
+    },
+    {
+      "epoch": 0.08079300086720748,
+      "grad_norm": 4.375,
+      "learning_rate": 4.925315652037784e-05,
+      "loss": 0.6887,
+      "step": 1840
+    },
+    {
+      "epoch": 0.08088081934641096,
+      "grad_norm": 3.890625,
+      "learning_rate": 4.925147721213881e-05,
+      "loss": 0.652,
+      "step": 1842
+    },
+    {
+      "epoch": 0.08096863782561446,
+      "grad_norm": 4.125,
+      "learning_rate": 4.924979604671583e-05,
+      "loss": 0.6581,
+      "step": 1844
+    },
+    {
+      "epoch": 0.08105645630481795,
+      "grad_norm": 3.5,
+      "learning_rate": 4.924811302423766e-05,
+      "loss": 0.6593,
+      "step": 1846
+    },
+    {
+      "epoch": 0.08114427478402143,
+      "grad_norm": 5.65625,
+      "learning_rate": 4.924642814483318e-05,
+      "loss": 0.6676,
+      "step": 1848
+    },
+    {
+      "epoch": 0.08123209326322492,
+      "grad_norm": 4.8125,
+      "learning_rate": 4.924474140863142e-05,
+      "loss": 0.6476,
+      "step": 1850
+    },
+    {
+      "epoch": 0.0813199117424284,
+      "grad_norm": 3.953125,
+      "learning_rate": 4.924305281576156e-05,
+      "loss": 0.6434,
+      "step": 1852
+    },
+    {
+      "epoch": 0.08140773022163189,
+      "grad_norm": 4.09375,
+      "learning_rate": 4.924136236635289e-05,
+      "loss": 0.647,
+      "step": 1854
+    },
+    {
+      "epoch": 0.08149554870083538,
+      "grad_norm": 4.75,
+      "learning_rate": 4.923967006053489e-05,
+      "loss": 0.6646,
+      "step": 1856
+    },
+    {
+      "epoch": 0.08158336718003886,
+      "grad_norm": 4.34375,
+      "learning_rate": 4.9237975898437144e-05,
+      "loss": 0.6349,
+      "step": 1858
+    },
+    {
+      "epoch": 0.08167118565924235,
+      "grad_norm": 4.09375,
+      "learning_rate": 4.923627988018939e-05,
+      "loss": 0.6667,
+      "step": 1860
+    },
+    {
+      "epoch": 0.08175900413844583,
+      "grad_norm": 4.0,
+      "learning_rate": 4.9234582005921514e-05,
+      "loss": 0.6273,
+      "step": 1862
+    },
+    {
+      "epoch": 0.08184682261764932,
+      "grad_norm": 3.359375,
+      "learning_rate": 4.923288227576354e-05,
+      "loss": 0.6516,
+      "step": 1864
+    },
+    {
+      "epoch": 0.0819346410968528,
+      "grad_norm": 3.265625,
+      "learning_rate": 4.923118068984564e-05,
+      "loss": 0.6475,
+      "step": 1866
+    },
+    {
+      "epoch": 0.08202245957605629,
+      "grad_norm": 3.171875,
+      "learning_rate": 4.92294772482981e-05,
+      "loss": 0.6575,
+      "step": 1868
+    },
+    {
+      "epoch": 0.08211027805525978,
+      "grad_norm": 3.375,
+      "learning_rate": 4.922777195125139e-05,
+      "loss": 0.6461,
+      "step": 1870
+    },
+    {
+      "epoch": 0.08219809653446326,
+      "grad_norm": 3.46875,
+      "learning_rate": 4.922606479883609e-05,
+      "loss": 0.664,
+      "step": 1872
+    },
+    {
+      "epoch": 0.08228591501366675,
+      "grad_norm": 4.125,
+      "learning_rate": 4.9224355791182955e-05,
+      "loss": 0.6572,
+      "step": 1874
+    },
+    {
+      "epoch": 0.08237373349287024,
+      "grad_norm": 4.28125,
+      "learning_rate": 4.922264492842283e-05,
+      "loss": 0.6558,
+      "step": 1876
+    },
+    {
+      "epoch": 0.08246155197207372,
+      "grad_norm": 3.65625,
+      "learning_rate": 4.922093221068676e-05,
+      "loss": 0.6248,
+      "step": 1878
+    },
+    {
+      "epoch": 0.08254937045127721,
+      "grad_norm": 3.34375,
+      "learning_rate": 4.92192176381059e-05,
+      "loss": 0.6539,
+      "step": 1880
+    },
+    {
+      "epoch": 0.08263718893048069,
+      "grad_norm": 3.03125,
+      "learning_rate": 4.9217501210811536e-05,
+      "loss": 0.662,
+      "step": 1882
+    },
+    {
+      "epoch": 0.08272500740968418,
+      "grad_norm": 3.734375,
+      "learning_rate": 4.9215782928935126e-05,
+      "loss": 0.6361,
+      "step": 1884
+    },
+    {
+      "epoch": 0.08281282588888766,
+      "grad_norm": 3.875,
+      "learning_rate": 4.921406279260826e-05,
+      "loss": 0.6498,
+      "step": 1886
+    },
+    {
+      "epoch": 0.08290064436809115,
+      "grad_norm": 3.078125,
+      "learning_rate": 4.9212340801962655e-05,
+      "loss": 0.6539,
+      "step": 1888
+    },
+    {
+      "epoch": 0.08298846284729464,
+      "grad_norm": 3.75,
+      "learning_rate": 4.9210616957130185e-05,
+      "loss": 0.6717,
+      "step": 1890
+    },
+    {
+      "epoch": 0.08307628132649812,
+      "grad_norm": 3.3125,
+      "learning_rate": 4.9208891258242874e-05,
+      "loss": 0.6808,
+      "step": 1892
+    },
+    {
+      "epoch": 0.08316409980570161,
+      "grad_norm": 3.59375,
+      "learning_rate": 4.9207163705432855e-05,
+      "loss": 0.6205,
+      "step": 1894
+    },
+    {
+      "epoch": 0.08325191828490511,
+      "grad_norm": 3.28125,
+      "learning_rate": 4.920543429883245e-05,
+      "loss": 0.6618,
+      "step": 1896
+    },
+    {
+      "epoch": 0.0833397367641086,
+      "grad_norm": 3.34375,
+      "learning_rate": 4.9203703038574076e-05,
+      "loss": 0.6543,
+      "step": 1898
+    },
+    {
+      "epoch": 0.08342755524331208,
+      "grad_norm": 3.40625,
+      "learning_rate": 4.9201969924790324e-05,
+      "loss": 0.6634,
+      "step": 1900
+    },
+    {
+      "epoch": 0.08351537372251557,
+      "grad_norm": 4.28125,
+      "learning_rate": 4.9200234957613915e-05,
+      "loss": 0.6681,
+      "step": 1902
+    },
+    {
+      "epoch": 0.08360319220171905,
+      "grad_norm": 5.09375,
+      "learning_rate": 4.9198498137177705e-05,
+      "loss": 0.6415,
+      "step": 1904
+    },
+    {
+      "epoch": 0.08369101068092254,
+      "grad_norm": 5.5625,
+      "learning_rate": 4.919675946361472e-05,
+      "loss": 0.6455,
+      "step": 1906
+    },
+    {
+      "epoch": 0.08377882916012602,
+      "grad_norm": 7.09375,
+      "learning_rate": 4.919501893705808e-05,
+      "loss": 0.6539,
+      "step": 1908
+    },
+    {
+      "epoch": 0.08386664763932951,
+      "grad_norm": 6.65625,
+      "learning_rate": 4.91932765576411e-05,
+      "loss": 0.669,
+      "step": 1910
+    },
+    {
+      "epoch": 0.083954466118533,
+      "grad_norm": 5.625,
+      "learning_rate": 4.91915323254972e-05,
+      "loss": 0.6528,
+      "step": 1912
+    },
+    {
+      "epoch": 0.08404228459773648,
+      "grad_norm": 3.5625,
+      "learning_rate": 4.918978624075995e-05,
+      "loss": 0.6532,
+      "step": 1914
+    },
+    {
+      "epoch": 0.08413010307693997,
+      "grad_norm": 3.0625,
+      "learning_rate": 4.918803830356308e-05,
+      "loss": 0.6647,
+      "step": 1916
+    },
+    {
+      "epoch": 0.08421792155614345,
+      "grad_norm": 3.8125,
+      "learning_rate": 4.918628851404043e-05,
+      "loss": 0.7015,
+      "step": 1918
+    },
+    {
+      "epoch": 0.08430574003534694,
+      "grad_norm": 3.9375,
+      "learning_rate": 4.918453687232601e-05,
+      "loss": 0.6318,
+      "step": 1920
+    },
+    {
+      "epoch": 0.08439355851455042,
+      "grad_norm": 3.9375,
+      "learning_rate": 4.918278337855396e-05,
+      "loss": 0.6575,
+      "step": 1922
+    },
+    {
+      "epoch": 0.08448137699375391,
+      "grad_norm": 5.21875,
+      "learning_rate": 4.918102803285856e-05,
+      "loss": 0.6639,
+      "step": 1924
+    },
+    {
+      "epoch": 0.0845691954729574,
+      "grad_norm": 3.875,
+      "learning_rate": 4.917927083537423e-05,
+      "loss": 0.6426,
+      "step": 1926
+    },
+    {
+      "epoch": 0.08465701395216088,
+      "grad_norm": 3.265625,
+      "learning_rate": 4.9177511786235556e-05,
+      "loss": 0.6318,
+      "step": 1928
+    },
+    {
+      "epoch": 0.08474483243136437,
+      "grad_norm": 3.09375,
+      "learning_rate": 4.917575088557723e-05,
+      "loss": 0.6358,
+      "step": 1930
+    },
+    {
+      "epoch": 0.08483265091056785,
+      "grad_norm": 3.5,
+      "learning_rate": 4.917398813353411e-05,
+      "loss": 0.6354,
+      "step": 1932
+    },
+    {
+      "epoch": 0.08492046938977134,
+      "grad_norm": 4.5625,
+      "learning_rate": 4.917222353024118e-05,
+      "loss": 0.6469,
+      "step": 1934
+    },
+    {
+      "epoch": 0.08500828786897482,
+      "grad_norm": 4.28125,
+      "learning_rate": 4.9170457075833574e-05,
+      "loss": 0.6359,
+      "step": 1936
+    },
+    {
+      "epoch": 0.08509610634817831,
+      "grad_norm": 4.15625,
+      "learning_rate": 4.916868877044657e-05,
+      "loss": 0.6663,
+      "step": 1938
+    },
+    {
+      "epoch": 0.0851839248273818,
+      "grad_norm": 5.78125,
+      "learning_rate": 4.916691861421559e-05,
+      "loss": 0.6841,
+      "step": 1940
+    },
+    {
+      "epoch": 0.08527174330658528,
+      "grad_norm": 4.625,
+      "learning_rate": 4.916514660727619e-05,
+      "loss": 0.6488,
+      "step": 1942
+    },
+    {
+      "epoch": 0.08535956178578877,
+      "grad_norm": 4.3125,
+      "learning_rate": 4.916337274976407e-05,
+      "loss": 0.6324,
+      "step": 1944
+    },
+    {
+      "epoch": 0.08544738026499227,
+      "grad_norm": 3.84375,
+      "learning_rate": 4.9161597041815075e-05,
+      "loss": 0.6443,
+      "step": 1946
+    },
+    {
+      "epoch": 0.08553519874419575,
+      "grad_norm": 4.03125,
+      "learning_rate": 4.9159819483565175e-05,
+      "loss": 0.6394,
+      "step": 1948
+    },
+    {
+      "epoch": 0.08562301722339924,
+      "grad_norm": 3.96875,
+      "learning_rate": 4.915804007515052e-05,
+      "loss": 0.6319,
+      "step": 1950
+    },
+    {
+      "epoch": 0.08571083570260273,
+      "grad_norm": 3.984375,
+      "learning_rate": 4.915625881670736e-05,
+      "loss": 0.6318,
+      "step": 1952
+    },
+    {
+      "epoch": 0.08579865418180621,
+      "grad_norm": 3.4375,
+      "learning_rate": 4.915447570837211e-05,
+      "loss": 0.6436,
+      "step": 1954
+    },
+    {
+      "epoch": 0.0858864726610097,
+      "grad_norm": 3.53125,
+      "learning_rate": 4.9152690750281314e-05,
+      "loss": 0.6554,
+      "step": 1956
+    },
+    {
+      "epoch": 0.08597429114021318,
+      "grad_norm": 3.4375,
+      "learning_rate": 4.915090394257168e-05,
+      "loss": 0.6466,
+      "step": 1958
+    },
+    {
+      "epoch": 0.08606210961941667,
+      "grad_norm": 3.046875,
+      "learning_rate": 4.914911528538003e-05,
+      "loss": 0.6518,
+      "step": 1960
+    },
+    {
+      "epoch": 0.08614992809862015,
+      "grad_norm": 3.46875,
+      "learning_rate": 4.914732477884334e-05,
+      "loss": 0.6549,
+      "step": 1962
+    },
+    {
+      "epoch": 0.08623774657782364,
+      "grad_norm": 4.65625,
+      "learning_rate": 4.914553242309873e-05,
+      "loss": 0.6223,
+      "step": 1964
+    },
+    {
+      "epoch": 0.08632556505702713,
+      "grad_norm": 3.953125,
+      "learning_rate": 4.9143738218283466e-05,
+      "loss": 0.6593,
+      "step": 1966
+    },
+    {
+      "epoch": 0.08641338353623061,
+      "grad_norm": 3.25,
+      "learning_rate": 4.9141942164534936e-05,
+      "loss": 0.6879,
+      "step": 1968
+    },
+    {
+      "epoch": 0.0865012020154341,
+      "grad_norm": 4.0625,
+      "learning_rate": 4.9140144261990687e-05,
+      "loss": 0.6215,
+      "step": 1970
+    },
+    {
+      "epoch": 0.08658902049463758,
+      "grad_norm": 3.5625,
+      "learning_rate": 4.91383445107884e-05,
+      "loss": 0.6731,
+      "step": 1972
+    },
+    {
+      "epoch": 0.08667683897384107,
+      "grad_norm": 3.125,
+      "learning_rate": 4.913654291106591e-05,
+      "loss": 0.6251,
+      "step": 1974
+    },
+    {
+      "epoch": 0.08676465745304456,
+      "grad_norm": 3.25,
+      "learning_rate": 4.9134739462961174e-05,
+      "loss": 0.6461,
+      "step": 1976
+    },
+    {
+      "epoch": 0.08685247593224804,
+      "grad_norm": 3.5,
+      "learning_rate": 4.913293416661231e-05,
+      "loss": 0.6691,
+      "step": 1978
+    },
+    {
+      "epoch": 0.08694029441145153,
+      "grad_norm": 3.125,
+      "learning_rate": 4.913112702215756e-05,
+      "loss": 0.6742,
+      "step": 1980
+    },
+    {
+      "epoch": 0.08702811289065501,
+      "grad_norm": 4.4375,
+      "learning_rate": 4.9129318029735315e-05,
+      "loss": 0.6441,
+      "step": 1982
+    },
+    {
+      "epoch": 0.0871159313698585,
+      "grad_norm": 4.0625,
+      "learning_rate": 4.912750718948411e-05,
+      "loss": 0.6627,
+      "step": 1984
+    },
+    {
+      "epoch": 0.08720374984906198,
+      "grad_norm": 4.96875,
+      "learning_rate": 4.912569450154263e-05,
+      "loss": 0.667,
+      "step": 1986
+    },
+    {
+      "epoch": 0.08729156832826547,
+      "grad_norm": 4.75,
+      "learning_rate": 4.912387996604968e-05,
+      "loss": 0.6616,
+      "step": 1988
+    },
+    {
+      "epoch": 0.08737938680746896,
+      "grad_norm": 4.90625,
+      "learning_rate": 4.9122063583144204e-05,
+      "loss": 0.6596,
+      "step": 1990
+    },
+    {
+      "epoch": 0.08746720528667244,
+      "grad_norm": 4.9375,
+      "learning_rate": 4.912024535296533e-05,
+      "loss": 0.6549,
+      "step": 1992
+    },
+    {
+      "epoch": 0.08755502376587593,
+      "grad_norm": 3.3125,
+      "learning_rate": 4.9118425275652286e-05,
+      "loss": 0.6287,
+      "step": 1994
+    },
+    {
+      "epoch": 0.08764284224507943,
+      "grad_norm": 3.78125,
+      "learning_rate": 4.911660335134445e-05,
+      "loss": 0.6587,
+      "step": 1996
+    },
+    {
+      "epoch": 0.08773066072428291,
+      "grad_norm": 3.09375,
+      "learning_rate": 4.9114779580181345e-05,
+      "loss": 0.6227,
+      "step": 1998
+    },
+    {
+      "epoch": 0.0878184792034864,
+      "grad_norm": 3.921875,
+      "learning_rate": 4.9112953962302646e-05,
+      "loss": 0.6558,
+      "step": 2000
+    },
+    {
+      "epoch": 0.08790629768268989,
+      "grad_norm": 4.65625,
+      "learning_rate": 4.9111126497848144e-05,
+      "loss": 0.6681,
+      "step": 2002
+    },
+    {
+      "epoch": 0.08799411616189337,
+      "grad_norm": 5.34375,
+      "learning_rate": 4.9109297186957796e-05,
+      "loss": 0.6741,
+      "step": 2004
+    },
+    {
+      "epoch": 0.08808193464109686,
+      "grad_norm": 3.90625,
+      "learning_rate": 4.91074660297717e-05,
+      "loss": 0.6307,
+      "step": 2006
+    },
+    {
+      "epoch": 0.08816975312030034,
+      "grad_norm": 4.3125,
+      "learning_rate": 4.910563302643007e-05,
+      "loss": 0.6463,
+      "step": 2008
+    },
+    {
+      "epoch": 0.08825757159950383,
+      "grad_norm": 3.984375,
+      "learning_rate": 4.910379817707328e-05,
+      "loss": 0.6463,
+      "step": 2010
+    },
+    {
+      "epoch": 0.08834539007870731,
+      "grad_norm": 3.59375,
+      "learning_rate": 4.910196148184185e-05,
+      "loss": 0.6553,
+      "step": 2012
+    },
+    {
+      "epoch": 0.0884332085579108,
+      "grad_norm": 3.0625,
+      "learning_rate": 4.9100122940876433e-05,
+      "loss": 0.6324,
+      "step": 2014
+    },
+    {
+      "epoch": 0.08852102703711429,
+      "grad_norm": 3.265625,
+      "learning_rate": 4.9098282554317823e-05,
+      "loss": 0.6297,
+      "step": 2016
+    },
+    {
+      "epoch": 0.08860884551631777,
+      "grad_norm": 3.765625,
+      "learning_rate": 4.9096440322306956e-05,
+      "loss": 0.6239,
+      "step": 2018
+    },
+    {
+      "epoch": 0.08869666399552126,
+      "grad_norm": 3.140625,
+      "learning_rate": 4.909459624498491e-05,
+      "loss": 0.6589,
+      "step": 2020
+    },
+    {
+      "epoch": 0.08878448247472474,
+      "grad_norm": 3.734375,
+      "learning_rate": 4.909275032249292e-05,
+      "loss": 0.6233,
+      "step": 2022
+    },
+    {
+      "epoch": 0.08887230095392823,
+      "grad_norm": 3.484375,
+      "learning_rate": 4.909090255497233e-05,
+      "loss": 0.6599,
+      "step": 2024
+    },
+    {
+      "epoch": 0.08896011943313172,
+      "grad_norm": 3.28125,
+      "learning_rate": 4.908905294256464e-05,
+      "loss": 0.652,
+      "step": 2026
+    },
+    {
+      "epoch": 0.0890479379123352,
+      "grad_norm": 3.515625,
+      "learning_rate": 4.908720148541152e-05,
+      "loss": 0.632,
+      "step": 2028
+    },
+    {
+      "epoch": 0.08913575639153869,
+      "grad_norm": 3.484375,
+      "learning_rate": 4.9085348183654714e-05,
+      "loss": 0.6419,
+      "step": 2030
+    },
+    {
+      "epoch": 0.08922357487074217,
+      "grad_norm": 3.15625,
+      "learning_rate": 4.908349303743618e-05,
+      "loss": 0.6383,
+      "step": 2032
+    },
+    {
+      "epoch": 0.08931139334994566,
+      "grad_norm": 3.203125,
+      "learning_rate": 4.908163604689798e-05,
+      "loss": 0.6406,
+      "step": 2034
+    },
+    {
+      "epoch": 0.08939921182914914,
+      "grad_norm": 3.09375,
+      "learning_rate": 4.907977721218231e-05,
+      "loss": 0.6367,
+      "step": 2036
+    },
+    {
+      "epoch": 0.08948703030835263,
+      "grad_norm": 3.765625,
+      "learning_rate": 4.907791653343153e-05,
+      "loss": 0.6365,
+      "step": 2038
+    },
+    {
+      "epoch": 0.08957484878755612,
+      "grad_norm": 3.0,
+      "learning_rate": 4.907605401078814e-05,
+      "loss": 0.6033,
+      "step": 2040
+    },
+    {
+      "epoch": 0.0896626672667596,
+      "grad_norm": 3.28125,
+      "learning_rate": 4.907418964439475e-05,
+      "loss": 0.6359,
+      "step": 2042
+    },
+    {
+      "epoch": 0.08975048574596309,
+      "grad_norm": 3.296875,
+      "learning_rate": 4.907232343439415e-05,
+      "loss": 0.6408,
+      "step": 2044
+    },
+    {
+      "epoch": 0.08983830422516659,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.907045538092926e-05,
+      "loss": 0.6232,
+      "step": 2046
+    },
+    {
+      "epoch": 0.08992612270437007,
+      "grad_norm": 3.171875,
+      "learning_rate": 4.906858548414311e-05,
+      "loss": 0.6624,
+      "step": 2048
+    },
+    {
+      "epoch": 0.09001394118357356,
+      "grad_norm": 3.5625,
+      "learning_rate": 4.906671374417893e-05,
+      "loss": 0.6527,
+      "step": 2050
+    },
+    {
+      "epoch": 0.09010175966277705,
+      "grad_norm": 4.875,
+      "learning_rate": 4.906484016118004e-05,
+      "loss": 0.6559,
+      "step": 2052
+    },
+    {
+      "epoch": 0.09018957814198053,
+      "grad_norm": 3.453125,
+      "learning_rate": 4.906296473528991e-05,
+      "loss": 0.6431,
+      "step": 2054
+    },
+    {
+      "epoch": 0.09027739662118402,
+      "grad_norm": 3.609375,
+      "learning_rate": 4.9061087466652183e-05,
+      "loss": 0.6458,
+      "step": 2056
+    },
+    {
+      "epoch": 0.0903652151003875,
+      "grad_norm": 3.015625,
+      "learning_rate": 4.905920835541061e-05,
+      "loss": 0.6394,
+      "step": 2058
+    },
+    {
+      "epoch": 0.09045303357959099,
+      "grad_norm": 3.90625,
+      "learning_rate": 4.9057327401709084e-05,
+      "loss": 0.6526,
+      "step": 2060
+    },
+    {
+      "epoch": 0.09054085205879447,
+      "grad_norm": 4.71875,
+      "learning_rate": 4.905544460569167e-05,
+      "loss": 0.668,
+      "step": 2062
+    },
+    {
+      "epoch": 0.09062867053799796,
+      "grad_norm": 3.546875,
+      "learning_rate": 4.9053559967502535e-05,
+      "loss": 0.6496,
+      "step": 2064
+    },
+    {
+      "epoch": 0.09071648901720145,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.905167348728601e-05,
+      "loss": 0.6582,
+      "step": 2066
+    },
+    {
+      "epoch": 0.09080430749640493,
+      "grad_norm": 3.3125,
+      "learning_rate": 4.904978516518657e-05,
+      "loss": 0.6153,
+      "step": 2068
+    },
+    {
+      "epoch": 0.09089212597560842,
+      "grad_norm": 4.65625,
+      "learning_rate": 4.904789500134881e-05,
+      "loss": 0.6161,
+      "step": 2070
+    },
+    {
+      "epoch": 0.0909799444548119,
+      "grad_norm": 6.375,
+      "learning_rate": 4.90460029959175e-05,
+      "loss": 0.6207,
+      "step": 2072
+    },
+    {
+      "epoch": 0.09106776293401539,
+      "grad_norm": 5.5,
+      "learning_rate": 4.90441091490375e-05,
+      "loss": 0.6453,
+      "step": 2074
+    },
+    {
+      "epoch": 0.09115558141321888,
+      "grad_norm": 5.8125,
+      "learning_rate": 4.904221346085387e-05,
+      "loss": 0.6337,
+      "step": 2076
+    },
+    {
+      "epoch": 0.09124339989242236,
+      "grad_norm": 4.8125,
+      "learning_rate": 4.904031593151176e-05,
+      "loss": 0.6836,
+      "step": 2078
+    },
+    {
+      "epoch": 0.09133121837162585,
+      "grad_norm": 3.25,
+      "learning_rate": 4.90384165611565e-05,
+      "loss": 0.6126,
+      "step": 2080
+    },
+    {
+      "epoch": 0.09141903685082933,
+      "grad_norm": 3.546875,
+      "learning_rate": 4.9036515349933534e-05,
+      "loss": 0.6287,
+      "step": 2082
+    },
+    {
+      "epoch": 0.09150685533003282,
+      "grad_norm": 3.015625,
+      "learning_rate": 4.903461229798846e-05,
+      "loss": 0.6347,
+      "step": 2084
+    },
+    {
+      "epoch": 0.0915946738092363,
+      "grad_norm": 3.640625,
+      "learning_rate": 4.903270740546701e-05,
+      "loss": 0.6548,
+      "step": 2086
+    },
+    {
+      "epoch": 0.09168249228843979,
+      "grad_norm": 3.796875,
+      "learning_rate": 4.9030800672515075e-05,
+      "loss": 0.639,
+      "step": 2088
+    },
+    {
+      "epoch": 0.09177031076764328,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.902889209927866e-05,
+      "loss": 0.6299,
+      "step": 2090
+    },
+    {
+      "epoch": 0.09185812924684676,
+      "grad_norm": 3.34375,
+      "learning_rate": 4.902698168590393e-05,
+      "loss": 0.6132,
+      "step": 2092
+    },
+    {
+      "epoch": 0.09194594772605025,
+      "grad_norm": 3.234375,
+      "learning_rate": 4.902506943253717e-05,
+      "loss": 0.6099,
+      "step": 2094
+    },
+    {
+      "epoch": 0.09203376620525373,
+      "grad_norm": 3.4375,
+      "learning_rate": 4.902315533932485e-05,
+      "loss": 0.6253,
+      "step": 2096
+    },
+    {
+      "epoch": 0.09212158468445723,
+      "grad_norm": 3.0625,
+      "learning_rate": 4.9021239406413534e-05,
+      "loss": 0.6292,
+      "step": 2098
+    },
+    {
+      "epoch": 0.09220940316366072,
+      "grad_norm": 4.09375,
+      "learning_rate": 4.901932163394994e-05,
+      "loss": 0.6073,
+      "step": 2100
+    },
+    {
+      "epoch": 0.0922972216428642,
+      "grad_norm": 3.84375,
+      "learning_rate": 4.901740202208094e-05,
+      "loss": 0.6577,
+      "step": 2102
+    },
+    {
+      "epoch": 0.09238504012206769,
+      "grad_norm": 3.3125,
+      "learning_rate": 4.901548057095353e-05,
+      "loss": 0.6495,
+      "step": 2104
+    },
+    {
+      "epoch": 0.09247285860127118,
+      "grad_norm": 3.40625,
+      "learning_rate": 4.9013557280714874e-05,
+      "loss": 0.6435,
+      "step": 2106
+    },
+    {
+      "epoch": 0.09256067708047466,
+      "grad_norm": 3.484375,
+      "learning_rate": 4.901163215151223e-05,
+      "loss": 0.6131,
+      "step": 2108
+    },
+    {
+      "epoch": 0.09264849555967815,
+      "grad_norm": 3.34375,
+      "learning_rate": 4.900970518349305e-05,
+      "loss": 0.6703,
+      "step": 2110
+    },
+    {
+      "epoch": 0.09273631403888163,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.900777637680489e-05,
+      "loss": 0.6473,
+      "step": 2112
+    },
+    {
+      "epoch": 0.09282413251808512,
+      "grad_norm": 3.015625,
+      "learning_rate": 4.9005845731595456e-05,
+      "loss": 0.6411,
+      "step": 2114
+    },
+    {
+      "epoch": 0.0929119509972886,
+      "grad_norm": 3.34375,
+      "learning_rate": 4.9003913248012605e-05,
+      "loss": 0.6409,
+      "step": 2116
+    },
+    {
+      "epoch": 0.09299976947649209,
+      "grad_norm": 3.03125,
+      "learning_rate": 4.900197892620432e-05,
+      "loss": 0.6326,
+      "step": 2118
+    },
+    {
+      "epoch": 0.09308758795569558,
+      "grad_norm": 3.0625,
+      "learning_rate": 4.9000042766318744e-05,
+      "loss": 0.617,
+      "step": 2120
+    },
+    {
+      "epoch": 0.09317540643489906,
+      "grad_norm": 4.28125,
+      "learning_rate": 4.899810476850413e-05,
+      "loss": 0.627,
+      "step": 2122
+    },
+    {
+      "epoch": 0.09326322491410255,
+      "grad_norm": 3.171875,
+      "learning_rate": 4.899616493290891e-05,
+      "loss": 0.6311,
+      "step": 2124
+    },
+    {
+      "epoch": 0.09335104339330604,
+      "grad_norm": 3.0625,
+      "learning_rate": 4.8994223259681615e-05,
+      "loss": 0.6436,
+      "step": 2126
+    },
+    {
+      "epoch": 0.09343886187250952,
+      "grad_norm": 3.4375,
+      "learning_rate": 4.899227974897095e-05,
+      "loss": 0.6159,
+      "step": 2128
+    },
+    {
+      "epoch": 0.093526680351713,
+      "grad_norm": 3.515625,
+      "learning_rate": 4.899033440092576e-05,
+      "loss": 0.6209,
+      "step": 2130
+    },
+    {
+      "epoch": 0.09361449883091649,
+      "grad_norm": 3.84375,
+      "learning_rate": 4.8988387215695007e-05,
+      "loss": 0.643,
+      "step": 2132
+    },
+    {
+      "epoch": 0.09370231731011998,
+      "grad_norm": 3.859375,
+      "learning_rate": 4.898643819342781e-05,
+      "loss": 0.6131,
+      "step": 2134
+    },
+    {
+      "epoch": 0.09379013578932346,
+      "grad_norm": 3.53125,
+      "learning_rate": 4.898448733427343e-05,
+      "loss": 0.6001,
+      "step": 2136
+    },
+    {
+      "epoch": 0.09387795426852695,
+      "grad_norm": 3.328125,
+      "learning_rate": 4.898253463838126e-05,
+      "loss": 0.6278,
+      "step": 2138
+    },
+    {
+      "epoch": 0.09396577274773044,
+      "grad_norm": 3.75,
+      "learning_rate": 4.898058010590083e-05,
+      "loss": 0.6501,
+      "step": 2140
+    },
+    {
+      "epoch": 0.09405359122693392,
+      "grad_norm": 3.234375,
+      "learning_rate": 4.897862373698184e-05,
+      "loss": 0.6412,
+      "step": 2142
+    },
+    {
+      "epoch": 0.09414140970613741,
+      "grad_norm": 3.796875,
+      "learning_rate": 4.8976665531774094e-05,
+      "loss": 0.6422,
+      "step": 2144
+    },
+    {
+      "epoch": 0.0942292281853409,
+      "grad_norm": 6.4375,
+      "learning_rate": 4.897470549042754e-05,
+      "loss": 0.6464,
+      "step": 2146
+    },
+    {
+      "epoch": 0.0943170466645444,
+      "grad_norm": 4.4375,
+      "learning_rate": 4.8972743613092304e-05,
+      "loss": 0.615,
+      "step": 2148
+    },
+    {
+      "epoch": 0.09440486514374788,
+      "grad_norm": 4.375,
+      "learning_rate": 4.897077989991862e-05,
+      "loss": 0.6224,
+      "step": 2150
+    },
+    {
+      "epoch": 0.09449268362295137,
+      "grad_norm": 5.0,
+      "learning_rate": 4.896881435105685e-05,
+      "loss": 0.6222,
+      "step": 2152
+    },
+    {
+      "epoch": 0.09458050210215485,
+      "grad_norm": 4.625,
+      "learning_rate": 4.896684696665754e-05,
+      "loss": 0.632,
+      "step": 2154
+    },
+    {
+      "epoch": 0.09466832058135834,
+      "grad_norm": 4.21875,
+      "learning_rate": 4.896487774687135e-05,
+      "loss": 0.6198,
+      "step": 2156
+    },
+    {
+      "epoch": 0.09475613906056182,
+      "grad_norm": 4.28125,
+      "learning_rate": 4.8962906691849066e-05,
+      "loss": 0.6298,
+      "step": 2158
+    },
+    {
+      "epoch": 0.09484395753976531,
+      "grad_norm": 3.4375,
+      "learning_rate": 4.8960933801741646e-05,
+      "loss": 0.6407,
+      "step": 2160
+    },
+    {
+      "epoch": 0.0949317760189688,
+      "grad_norm": 3.328125,
+      "learning_rate": 4.895895907670017e-05,
+      "loss": 0.6368,
+      "step": 2162
+    },
+    {
+      "epoch": 0.09501959449817228,
+      "grad_norm": 3.640625,
+      "learning_rate": 4.895698251687587e-05,
+      "loss": 0.6023,
+      "step": 2164
+    },
+    {
+      "epoch": 0.09510741297737577,
+      "grad_norm": 3.359375,
+      "learning_rate": 4.895500412242011e-05,
+      "loss": 0.611,
+      "step": 2166
+    },
+    {
+      "epoch": 0.09519523145657925,
+      "grad_norm": 3.859375,
+      "learning_rate": 4.895302389348438e-05,
+      "loss": 0.6478,
+      "step": 2168
+    },
+    {
+      "epoch": 0.09528304993578274,
+      "grad_norm": 4.34375,
+      "learning_rate": 4.8951041830220344e-05,
+      "loss": 0.6434,
+      "step": 2170
+    },
+    {
+      "epoch": 0.09537086841498622,
+      "grad_norm": 3.765625,
+      "learning_rate": 4.8949057932779784e-05,
+      "loss": 0.6129,
+      "step": 2172
+    },
+    {
+      "epoch": 0.09545868689418971,
+      "grad_norm": 5.375,
+      "learning_rate": 4.894707220131463e-05,
+      "loss": 0.6264,
+      "step": 2174
+    },
+    {
+      "epoch": 0.0955465053733932,
+      "grad_norm": 4.78125,
+      "learning_rate": 4.8945084635976944e-05,
+      "loss": 0.64,
+      "step": 2176
+    },
+    {
+      "epoch": 0.09563432385259668,
+      "grad_norm": 3.09375,
+      "learning_rate": 4.894309523691893e-05,
+      "loss": 0.6259,
+      "step": 2178
+    },
+    {
+      "epoch": 0.09572214233180017,
+      "grad_norm": 4.03125,
+      "learning_rate": 4.8941104004292955e-05,
+      "loss": 0.6351,
+      "step": 2180
+    },
+    {
+      "epoch": 0.09580996081100365,
+      "grad_norm": 4.09375,
+      "learning_rate": 4.8939110938251485e-05,
+      "loss": 0.6288,
+      "step": 2182
+    },
+    {
+      "epoch": 0.09589777929020714,
+      "grad_norm": 4.34375,
+      "learning_rate": 4.8937116038947164e-05,
+      "loss": 0.612,
+      "step": 2184
+    },
+    {
+      "epoch": 0.09598559776941062,
+      "grad_norm": 3.46875,
+      "learning_rate": 4.8935119306532764e-05,
+      "loss": 0.6268,
+      "step": 2186
+    },
+    {
+      "epoch": 0.09607341624861411,
+      "grad_norm": 3.703125,
+      "learning_rate": 4.893312074116119e-05,
+      "loss": 0.609,
+      "step": 2188
+    },
+    {
+      "epoch": 0.0961612347278176,
+      "grad_norm": 3.625,
+      "learning_rate": 4.893112034298548e-05,
+      "loss": 0.6281,
+      "step": 2190
+    },
+    {
+      "epoch": 0.09624905320702108,
+      "grad_norm": 3.265625,
+      "learning_rate": 4.892911811215885e-05,
+      "loss": 0.6052,
+      "step": 2192
+    },
+    {
+      "epoch": 0.09633687168622457,
+      "grad_norm": 3.484375,
+      "learning_rate": 4.8927114048834613e-05,
+      "loss": 0.6247,
+      "step": 2194
+    },
+    {
+      "epoch": 0.09642469016542805,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.892510815316625e-05,
+      "loss": 0.6106,
+      "step": 2196
+    },
+    {
+      "epoch": 0.09651250864463155,
+      "grad_norm": 3.328125,
+      "learning_rate": 4.8923100425307365e-05,
+      "loss": 0.6224,
+      "step": 2198
+    },
+    {
+      "epoch": 0.09660032712383504,
+      "grad_norm": 3.078125,
+      "learning_rate": 4.892109086541172e-05,
+      "loss": 0.6136,
+      "step": 2200
+    },
+    {
+      "epoch": 0.09668814560303853,
+      "grad_norm": 3.0625,
+      "learning_rate": 4.89190794736332e-05,
+      "loss": 0.6062,
+      "step": 2202
+    },
+    {
+      "epoch": 0.09677596408224201,
+      "grad_norm": 3.265625,
+      "learning_rate": 4.8917066250125834e-05,
+      "loss": 0.65,
+      "step": 2204
+    },
+    {
+      "epoch": 0.0968637825614455,
+      "grad_norm": 3.40625,
+      "learning_rate": 4.891505119504381e-05,
+      "loss": 0.5928,
+      "step": 2206
+    },
+    {
+      "epoch": 0.09695160104064898,
+      "grad_norm": 3.84375,
+      "learning_rate": 4.8913034308541425e-05,
+      "loss": 0.6516,
+      "step": 2208
+    },
+    {
+      "epoch": 0.09703941951985247,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.8911015590773145e-05,
+      "loss": 0.6375,
+      "step": 2210
+    },
+    {
+      "epoch": 0.09712723799905595,
+      "grad_norm": 3.28125,
+      "learning_rate": 4.890899504189356e-05,
+      "loss": 0.6432,
+      "step": 2212
+    },
+    {
+      "epoch": 0.09721505647825944,
+      "grad_norm": 4.1875,
+      "learning_rate": 4.8906972662057406e-05,
+      "loss": 0.6211,
+      "step": 2214
+    },
+    {
+      "epoch": 0.09730287495746293,
+      "grad_norm": 3.46875,
+      "learning_rate": 4.890494845141955e-05,
+      "loss": 0.6126,
+      "step": 2216
+    },
+    {
+      "epoch": 0.09739069343666641,
+      "grad_norm": 3.265625,
+      "learning_rate": 4.890292241013501e-05,
+      "loss": 0.6094,
+      "step": 2218
+    },
+    {
+      "epoch": 0.0974785119158699,
+      "grad_norm": 3.328125,
+      "learning_rate": 4.8900894538358944e-05,
+      "loss": 0.6204,
+      "step": 2220
+    },
+    {
+      "epoch": 0.09756633039507338,
+      "grad_norm": 2.953125,
+      "learning_rate": 4.889886483624664e-05,
+      "loss": 0.6227,
+      "step": 2222
+    },
+    {
+      "epoch": 0.09765414887427687,
+      "grad_norm": 4.09375,
+      "learning_rate": 4.889683330395355e-05,
+      "loss": 0.6102,
+      "step": 2224
+    },
+    {
+      "epoch": 0.09774196735348036,
+      "grad_norm": 3.625,
+      "learning_rate": 4.889479994163523e-05,
+      "loss": 0.6213,
+      "step": 2226
+    },
+    {
+      "epoch": 0.09782978583268384,
+      "grad_norm": 3.125,
+      "learning_rate": 4.8892764749447395e-05,
+      "loss": 0.61,
+      "step": 2228
+    },
+    {
+      "epoch": 0.09791760431188733,
+      "grad_norm": 3.0,
+      "learning_rate": 4.8890727727545916e-05,
+      "loss": 0.6275,
+      "step": 2230
+    },
+    {
+      "epoch": 0.09800542279109081,
+      "grad_norm": 3.0,
+      "learning_rate": 4.8888688876086786e-05,
+      "loss": 0.6525,
+      "step": 2232
+    },
+    {
+      "epoch": 0.0980932412702943,
+      "grad_norm": 3.671875,
+      "learning_rate": 4.8886648195226124e-05,
+      "loss": 0.6094,
+      "step": 2234
+    },
+    {
+      "epoch": 0.09818105974949778,
+      "grad_norm": 4.5625,
+      "learning_rate": 4.8884605685120224e-05,
+      "loss": 0.6006,
+      "step": 2236
+    },
+    {
+      "epoch": 0.09826887822870127,
+      "grad_norm": 5.5,
+      "learning_rate": 4.888256134592549e-05,
+      "loss": 0.6413,
+      "step": 2238
+    },
+    {
+      "epoch": 0.09835669670790476,
+      "grad_norm": 6.125,
+      "learning_rate": 4.888051517779849e-05,
+      "loss": 0.6073,
+      "step": 2240
+    },
+    {
+      "epoch": 0.09844451518710824,
+      "grad_norm": 4.875,
+      "learning_rate": 4.8878467180895906e-05,
+      "loss": 0.5941,
+      "step": 2242
+    },
+    {
+      "epoch": 0.09853233366631173,
+      "grad_norm": 4.53125,
+      "learning_rate": 4.887641735537459e-05,
+      "loss": 0.6117,
+      "step": 2244
+    },
+    {
+      "epoch": 0.09862015214551521,
+      "grad_norm": 5.46875,
+      "learning_rate": 4.88743657013915e-05,
+      "loss": 0.6266,
+      "step": 2246
+    },
+    {
+      "epoch": 0.09870797062471871,
+      "grad_norm": 4.84375,
+      "learning_rate": 4.887231221910376e-05,
+      "loss": 0.6203,
+      "step": 2248
+    },
+    {
+      "epoch": 0.0987957891039222,
+      "grad_norm": 3.109375,
+      "learning_rate": 4.8870256908668646e-05,
+      "loss": 0.618,
+      "step": 2250
+    },
+    {
+      "epoch": 0.09888360758312568,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.886819977024352e-05,
+      "loss": 0.6034,
+      "step": 2252
+    },
+    {
+      "epoch": 0.09897142606232917,
+      "grad_norm": 3.25,
+      "learning_rate": 4.886614080398594e-05,
+      "loss": 0.6134,
+      "step": 2254
+    },
+    {
+      "epoch": 0.09905924454153266,
+      "grad_norm": 3.09375,
+      "learning_rate": 4.886408001005357e-05,
+      "loss": 0.6086,
+      "step": 2256
+    },
+    {
+      "epoch": 0.09914706302073614,
+      "grad_norm": 3.953125,
+      "learning_rate": 4.886201738860423e-05,
+      "loss": 0.6463,
+      "step": 2258
+    },
+    {
+      "epoch": 0.09923488149993963,
+      "grad_norm": 4.3125,
+      "learning_rate": 4.885995293979589e-05,
+      "loss": 0.6026,
+      "step": 2260
+    },
+    {
+      "epoch": 0.09932269997914311,
+      "grad_norm": 4.78125,
+      "learning_rate": 4.8857886663786626e-05,
+      "loss": 0.6457,
+      "step": 2262
+    },
+    {
+      "epoch": 0.0994105184583466,
+      "grad_norm": 4.375,
+      "learning_rate": 4.885581856073468e-05,
+      "loss": 0.5953,
+      "step": 2264
+    },
+    {
+      "epoch": 0.09949833693755009,
+      "grad_norm": 5.03125,
+      "learning_rate": 4.8853748630798434e-05,
+      "loss": 0.606,
+      "step": 2266
+    },
+    {
+      "epoch": 0.09958615541675357,
+      "grad_norm": 6.4375,
+      "learning_rate": 4.88516768741364e-05,
+      "loss": 0.6011,
+      "step": 2268
+    },
+    {
+      "epoch": 0.09967397389595706,
+      "grad_norm": 5.1875,
+      "learning_rate": 4.8849603290907234e-05,
+      "loss": 0.6245,
+      "step": 2270
+    },
+    {
+      "epoch": 0.09976179237516054,
+      "grad_norm": 3.234375,
+      "learning_rate": 4.884752788126973e-05,
+      "loss": 0.6454,
+      "step": 2272
+    },
+    {
+      "epoch": 0.09984961085436403,
+      "grad_norm": 3.15625,
+      "learning_rate": 4.884545064538283e-05,
+      "loss": 0.6246,
+      "step": 2274
+    },
+    {
+      "epoch": 0.09993742933356752,
+      "grad_norm": 2.96875,
+      "learning_rate": 4.884337158340559e-05,
+      "loss": 0.6095,
+      "step": 2276
+    },
+    {
+      "epoch": 0.100025247812771,
+      "grad_norm": 3.6875,
+      "learning_rate": 4.884129069549726e-05,
+      "loss": 0.6126,
+      "step": 2278
+    },
+    {
+      "epoch": 0.10011306629197449,
+      "grad_norm": 3.796875,
+      "learning_rate": 4.883920798181715e-05,
+      "loss": 0.621,
+      "step": 2280
+    },
+    {
+      "epoch": 0.10020088477117797,
+      "grad_norm": 3.9375,
+      "learning_rate": 4.883712344252479e-05,
+      "loss": 0.604,
+      "step": 2282
+    },
+    {
+      "epoch": 0.10028870325038146,
+      "grad_norm": 3.6875,
+      "learning_rate": 4.88350370777798e-05,
+      "loss": 0.607,
+      "step": 2284
+    },
+    {
+      "epoch": 0.10037652172958494,
+      "grad_norm": 2.984375,
+      "learning_rate": 4.8832948887741956e-05,
+      "loss": 0.6061,
+      "step": 2286
+    },
+    {
+      "epoch": 0.10046434020878843,
+      "grad_norm": 3.296875,
+      "learning_rate": 4.883085887257117e-05,
+      "loss": 0.6153,
+      "step": 2288
+    },
+    {
+      "epoch": 0.10055215868799192,
+      "grad_norm": 3.890625,
+      "learning_rate": 4.882876703242751e-05,
+      "loss": 0.6025,
+      "step": 2290
+    },
+    {
+      "epoch": 0.1006399771671954,
+      "grad_norm": 4.03125,
+      "learning_rate": 4.882667336747115e-05,
+      "loss": 0.5971,
+      "step": 2292
+    },
+    {
+      "epoch": 0.10072779564639889,
+      "grad_norm": 3.015625,
+      "learning_rate": 4.882457787786243e-05,
+      "loss": 0.5883,
+      "step": 2294
+    },
+    {
+      "epoch": 0.10081561412560237,
+      "grad_norm": 3.09375,
+      "learning_rate": 4.882248056376183e-05,
+      "loss": 0.6166,
+      "step": 2296
+    },
+    {
+      "epoch": 0.10090343260480586,
+      "grad_norm": 3.5625,
+      "learning_rate": 4.882038142532995e-05,
+      "loss": 0.6313,
+      "step": 2298
+    },
+    {
+      "epoch": 0.10099125108400936,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.881828046272756e-05,
+      "loss": 0.5937,
+      "step": 2300
+    },
+    {
+      "epoch": 0.10107906956321284,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.881617767611554e-05,
+      "loss": 0.6125,
+      "step": 2302
+    },
+    {
+      "epoch": 0.10116688804241633,
+      "grad_norm": 3.390625,
+      "learning_rate": 4.881407306565492e-05,
+      "loss": 0.596,
+      "step": 2304
+    },
+    {
+      "epoch": 0.10125470652161982,
+      "grad_norm": 3.296875,
+      "learning_rate": 4.881196663150689e-05,
+      "loss": 0.6083,
+      "step": 2306
+    },
+    {
+      "epoch": 0.1013425250008233,
+      "grad_norm": 3.484375,
+      "learning_rate": 4.8809858373832726e-05,
+      "loss": 0.597,
+      "step": 2308
+    },
+    {
+      "epoch": 0.10143034348002679,
+      "grad_norm": 3.25,
+      "learning_rate": 4.880774829279392e-05,
+      "loss": 0.6188,
+      "step": 2310
+    },
+    {
+      "epoch": 0.10151816195923027,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.8805636388552035e-05,
+      "loss": 0.6203,
+      "step": 2312
+    },
+    {
+      "epoch": 0.10160598043843376,
+      "grad_norm": 3.078125,
+      "learning_rate": 4.8803522661268805e-05,
+      "loss": 0.6083,
+      "step": 2314
+    },
+    {
+      "epoch": 0.10169379891763725,
+      "grad_norm": 3.140625,
+      "learning_rate": 4.88014071111061e-05,
+      "loss": 0.6002,
+      "step": 2316
+    },
+    {
+      "epoch": 0.10178161739684073,
+      "grad_norm": 3.65625,
+      "learning_rate": 4.8799289738225936e-05,
+      "loss": 0.6068,
+      "step": 2318
+    },
+    {
+      "epoch": 0.10186943587604422,
+      "grad_norm": 3.28125,
+      "learning_rate": 4.879717054279047e-05,
+      "loss": 0.6208,
+      "step": 2320
+    },
+    {
+      "epoch": 0.1019572543552477,
+      "grad_norm": 2.953125,
+      "learning_rate": 4.879504952496197e-05,
+      "loss": 0.6341,
+      "step": 2322
+    },
+    {
+      "epoch": 0.10204507283445119,
+      "grad_norm": 3.4375,
+      "learning_rate": 4.8792926684902875e-05,
+      "loss": 0.6172,
+      "step": 2324
+    },
+    {
+      "epoch": 0.10213289131365468,
+      "grad_norm": 3.34375,
+      "learning_rate": 4.879080202277575e-05,
+      "loss": 0.5982,
+      "step": 2326
+    },
+    {
+      "epoch": 0.10222070979285816,
+      "grad_norm": 4.09375,
+      "learning_rate": 4.87886755387433e-05,
+      "loss": 0.6081,
+      "step": 2328
+    },
+    {
+      "epoch": 0.10230852827206165,
+      "grad_norm": 3.265625,
+      "learning_rate": 4.878654723296838e-05,
+      "loss": 0.6505,
+      "step": 2330
+    },
+    {
+      "epoch": 0.10239634675126513,
+      "grad_norm": 3.453125,
+      "learning_rate": 4.878441710561397e-05,
+      "loss": 0.6074,
+      "step": 2332
+    },
+    {
+      "epoch": 0.10248416523046862,
+      "grad_norm": 3.125,
+      "learning_rate": 4.878228515684319e-05,
+      "loss": 0.6145,
+      "step": 2334
+    },
+    {
+      "epoch": 0.1025719837096721,
+      "grad_norm": 3.734375,
+      "learning_rate": 4.878015138681932e-05,
+      "loss": 0.6188,
+      "step": 2336
+    },
+    {
+      "epoch": 0.10265980218887559,
+      "grad_norm": 4.1875,
+      "learning_rate": 4.877801579570575e-05,
+      "loss": 0.6209,
+      "step": 2338
+    },
+    {
+      "epoch": 0.10274762066807908,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.8775878383666035e-05,
+      "loss": 0.6161,
+      "step": 2340
+    },
+    {
+      "epoch": 0.10283543914728256,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.877373915086385e-05,
+      "loss": 0.6258,
+      "step": 2342
+    },
+    {
+      "epoch": 0.10292325762648605,
+      "grad_norm": 3.59375,
+      "learning_rate": 4.8771598097463026e-05,
+      "loss": 0.6049,
+      "step": 2344
+    },
+    {
+      "epoch": 0.10301107610568953,
+      "grad_norm": 3.546875,
+      "learning_rate": 4.876945522362752e-05,
+      "loss": 0.6479,
+      "step": 2346
+    },
+    {
+      "epoch": 0.10309889458489302,
+      "grad_norm": 3.6875,
+      "learning_rate": 4.876731052952144e-05,
+      "loss": 0.6041,
+      "step": 2348
+    },
+    {
+      "epoch": 0.10318671306409652,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.876516401530901e-05,
+      "loss": 0.6092,
+      "step": 2350
+    },
+    {
+      "epoch": 0.1032745315433,
+      "grad_norm": 3.515625,
+      "learning_rate": 4.876301568115463e-05,
+      "loss": 0.5955,
+      "step": 2352
+    },
+    {
+      "epoch": 0.10336235002250349,
+      "grad_norm": 3.546875,
+      "learning_rate": 4.876086552722281e-05,
+      "loss": 0.6191,
+      "step": 2354
+    },
+    {
+      "epoch": 0.10345016850170698,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.875871355367822e-05,
+      "loss": 0.6051,
+      "step": 2356
+    },
+    {
+      "epoch": 0.10353798698091046,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.8756559760685644e-05,
+      "loss": 0.601,
+      "step": 2358
+    },
+    {
+      "epoch": 0.10362580546011395,
+      "grad_norm": 3.125,
+      "learning_rate": 4.8754404148410025e-05,
+      "loss": 0.6016,
+      "step": 2360
+    },
+    {
+      "epoch": 0.10371362393931743,
+      "grad_norm": 5.03125,
+      "learning_rate": 4.875224671701645e-05,
+      "loss": 0.633,
+      "step": 2362
+    },
+    {
+      "epoch": 0.10380144241852092,
+      "grad_norm": 4.875,
+      "learning_rate": 4.8750087466670116e-05,
+      "loss": 0.6384,
+      "step": 2364
+    },
+    {
+      "epoch": 0.1038892608977244,
+      "grad_norm": 4.3125,
+      "learning_rate": 4.87479263975364e-05,
+      "loss": 0.6113,
+      "step": 2366
+    },
+    {
+      "epoch": 0.10397707937692789,
+      "grad_norm": 3.78125,
+      "learning_rate": 4.8745763509780785e-05,
+      "loss": 0.5765,
+      "step": 2368
+    },
+    {
+      "epoch": 0.10406489785613138,
+      "grad_norm": 3.65625,
+      "learning_rate": 4.874359880356891e-05,
+      "loss": 0.6533,
+      "step": 2370
+    },
+    {
+      "epoch": 0.10415271633533486,
+      "grad_norm": 3.671875,
+      "learning_rate": 4.874143227906654e-05,
+      "loss": 0.6002,
+      "step": 2372
+    },
+    {
+      "epoch": 0.10424053481453835,
+      "grad_norm": 4.4375,
+      "learning_rate": 4.87392639364396e-05,
+      "loss": 0.5987,
+      "step": 2374
+    },
+    {
+      "epoch": 0.10432835329374184,
+      "grad_norm": 4.0625,
+      "learning_rate": 4.873709377585414e-05,
+      "loss": 0.6137,
+      "step": 2376
+    },
+    {
+      "epoch": 0.10441617177294532,
+      "grad_norm": 4.5625,
+      "learning_rate": 4.873492179747634e-05,
+      "loss": 0.6091,
+      "step": 2378
+    },
+    {
+      "epoch": 0.1045039902521488,
+      "grad_norm": 4.75,
+      "learning_rate": 4.873274800147255e-05,
+      "loss": 0.6199,
+      "step": 2380
+    },
+    {
+      "epoch": 0.10459180873135229,
+      "grad_norm": 4.5625,
+      "learning_rate": 4.873057238800922e-05,
+      "loss": 0.6132,
+      "step": 2382
+    },
+    {
+      "epoch": 0.10467962721055578,
+      "grad_norm": 6.375,
+      "learning_rate": 4.872839495725297e-05,
+      "loss": 0.6083,
+      "step": 2384
+    },
+    {
+      "epoch": 0.10476744568975926,
+      "grad_norm": 4.46875,
+      "learning_rate": 4.8726215709370546e-05,
+      "loss": 0.6209,
+      "step": 2386
+    },
+    {
+      "epoch": 0.10485526416896275,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.872403464452884e-05,
+      "loss": 0.6004,
+      "step": 2388
+    },
+    {
+      "epoch": 0.10494308264816624,
+      "grad_norm": 2.96875,
+      "learning_rate": 4.8721851762894865e-05,
+      "loss": 0.5997,
+      "step": 2390
+    },
+    {
+      "epoch": 0.10503090112736972,
+      "grad_norm": 4.0625,
+      "learning_rate": 4.871966706463581e-05,
+      "loss": 0.6023,
+      "step": 2392
+    },
+    {
+      "epoch": 0.10511871960657321,
+      "grad_norm": 3.25,
+      "learning_rate": 4.871748054991895e-05,
+      "loss": 0.5691,
+      "step": 2394
+    },
+    {
+      "epoch": 0.1052065380857767,
+      "grad_norm": 3.59375,
+      "learning_rate": 4.871529221891175e-05,
+      "loss": 0.6085,
+      "step": 2396
+    },
+    {
+      "epoch": 0.10529435656498018,
+      "grad_norm": 4.46875,
+      "learning_rate": 4.871310207178179e-05,
+      "loss": 0.6212,
+      "step": 2398
+    },
+    {
+      "epoch": 0.10538217504418368,
+      "grad_norm": 4.03125,
+      "learning_rate": 4.8710910108696786e-05,
+      "loss": 0.6219,
+      "step": 2400
+    },
+    {
+      "epoch": 0.10546999352338716,
+      "grad_norm": 3.953125,
+      "learning_rate": 4.87087163298246e-05,
+      "loss": 0.6105,
+      "step": 2402
+    },
+    {
+      "epoch": 0.10555781200259065,
+      "grad_norm": 3.96875,
+      "learning_rate": 4.870652073533324e-05,
+      "loss": 0.621,
+      "step": 2404
+    },
+    {
+      "epoch": 0.10564563048179414,
+      "grad_norm": 3.828125,
+      "learning_rate": 4.8704323325390834e-05,
+      "loss": 0.6141,
+      "step": 2406
+    },
+    {
+      "epoch": 0.10573344896099762,
+      "grad_norm": 3.8125,
+      "learning_rate": 4.8702124100165666e-05,
+      "loss": 0.6142,
+      "step": 2408
+    },
+    {
+      "epoch": 0.10582126744020111,
+      "grad_norm": 3.46875,
+      "learning_rate": 4.869992305982615e-05,
+      "loss": 0.5919,
+      "step": 2410
+    },
+    {
+      "epoch": 0.1059090859194046,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.8697720204540846e-05,
+      "loss": 0.5876,
+      "step": 2412
+    },
+    {
+      "epoch": 0.10599690439860808,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.8695515534478456e-05,
+      "loss": 0.6464,
+      "step": 2414
+    },
+    {
+      "epoch": 0.10608472287781157,
+      "grad_norm": 3.0625,
+      "learning_rate": 4.8693309049807795e-05,
+      "loss": 0.5842,
+      "step": 2416
+    },
+    {
+      "epoch": 0.10617254135701505,
+      "grad_norm": 3.40625,
+      "learning_rate": 4.8691100750697856e-05,
+      "loss": 0.5864,
+      "step": 2418
+    },
+    {
+      "epoch": 0.10626035983621854,
+      "grad_norm": 3.296875,
+      "learning_rate": 4.8688890637317734e-05,
+      "loss": 0.6038,
+      "step": 2420
+    },
+    {
+      "epoch": 0.10634817831542202,
+      "grad_norm": 3.375,
+      "learning_rate": 4.86866787098367e-05,
+      "loss": 0.6135,
+      "step": 2422
+    },
+    {
+      "epoch": 0.10643599679462551,
+      "grad_norm": 3.5,
+      "learning_rate": 4.868446496842412e-05,
+      "loss": 0.6135,
+      "step": 2424
+    },
+    {
+      "epoch": 0.106523815273829,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.868224941324954e-05,
+      "loss": 0.6238,
+      "step": 2426
+    },
+    {
+      "epoch": 0.10661163375303248,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.868003204448263e-05,
+      "loss": 0.5728,
+      "step": 2428
+    },
+    {
+      "epoch": 0.10669945223223597,
+      "grad_norm": 3.0,
+      "learning_rate": 4.8677812862293184e-05,
+      "loss": 0.5991,
+      "step": 2430
+    },
+    {
+      "epoch": 0.10678727071143945,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.867559186685115e-05,
+      "loss": 0.6117,
+      "step": 2432
+    },
+    {
+      "epoch": 0.10687508919064294,
+      "grad_norm": 3.3125,
+      "learning_rate": 4.867336905832661e-05,
+      "loss": 0.6006,
+      "step": 2434
+    },
+    {
+      "epoch": 0.10696290766984642,
+      "grad_norm": 3.375,
+      "learning_rate": 4.8671144436889805e-05,
+      "loss": 0.5793,
+      "step": 2436
+    },
+    {
+      "epoch": 0.10705072614904991,
+      "grad_norm": 3.140625,
+      "learning_rate": 4.866891800271108e-05,
+      "loss": 0.5901,
+      "step": 2438
+    },
+    {
+      "epoch": 0.1071385446282534,
+      "grad_norm": 3.0,
+      "learning_rate": 4.8666689755960936e-05,
+      "loss": 0.6089,
+      "step": 2440
+    },
+    {
+      "epoch": 0.10722636310745688,
+      "grad_norm": 3.046875,
+      "learning_rate": 4.866445969681003e-05,
+      "loss": 0.6008,
+      "step": 2442
+    },
+    {
+      "epoch": 0.10731418158666037,
+      "grad_norm": 3.328125,
+      "learning_rate": 4.866222782542912e-05,
+      "loss": 0.6111,
+      "step": 2444
+    },
+    {
+      "epoch": 0.10740200006586385,
+      "grad_norm": 3.078125,
+      "learning_rate": 4.865999414198913e-05,
+      "loss": 0.6166,
+      "step": 2446
+    },
+    {
+      "epoch": 0.10748981854506734,
+      "grad_norm": 3.03125,
+      "learning_rate": 4.865775864666111e-05,
+      "loss": 0.5883,
+      "step": 2448
+    },
+    {
+      "epoch": 0.10757763702427084,
+      "grad_norm": 3.046875,
+      "learning_rate": 4.8655521339616274e-05,
+      "loss": 0.5968,
+      "step": 2450
+    },
+    {
+      "epoch": 0.10766545550347432,
+      "grad_norm": 3.734375,
+      "learning_rate": 4.865328222102594e-05,
+      "loss": 0.61,
+      "step": 2452
+    },
+    {
+      "epoch": 0.10775327398267781,
+      "grad_norm": 3.734375,
+      "learning_rate": 4.865104129106158e-05,
+      "loss": 0.5919,
+      "step": 2454
+    },
+    {
+      "epoch": 0.1078410924618813,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.86487985498948e-05,
+      "loss": 0.6421,
+      "step": 2456
+    },
+    {
+      "epoch": 0.10792891094108478,
+      "grad_norm": 3.3125,
+      "learning_rate": 4.8646553997697375e-05,
+      "loss": 0.6171,
+      "step": 2458
+    },
+    {
+      "epoch": 0.10801672942028827,
+      "grad_norm": 2.875,
+      "learning_rate": 4.864430763464117e-05,
+      "loss": 0.6143,
+      "step": 2460
+    },
+    {
+      "epoch": 0.10810454789949175,
+      "grad_norm": 3.390625,
+      "learning_rate": 4.8642059460898214e-05,
+      "loss": 0.616,
+      "step": 2462
+    },
+    {
+      "epoch": 0.10819236637869524,
+      "grad_norm": 2.875,
+      "learning_rate": 4.8639809476640685e-05,
+      "loss": 0.586,
+      "step": 2464
+    },
+    {
+      "epoch": 0.10828018485789873,
+      "grad_norm": 3.8125,
+      "learning_rate": 4.8637557682040876e-05,
+      "loss": 0.581,
+      "step": 2466
+    },
+    {
+      "epoch": 0.10836800333710221,
+      "grad_norm": 3.53125,
+      "learning_rate": 4.863530407727123e-05,
+      "loss": 0.6136,
+      "step": 2468
+    },
+    {
+      "epoch": 0.1084558218163057,
+      "grad_norm": 3.75,
+      "learning_rate": 4.863304866250433e-05,
+      "loss": 0.6392,
+      "step": 2470
+    },
+    {
+      "epoch": 0.10854364029550918,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.86307914379129e-05,
+      "loss": 0.5975,
+      "step": 2472
+    },
+    {
+      "epoch": 0.10863145877471267,
+      "grad_norm": 3.5,
+      "learning_rate": 4.8628532403669805e-05,
+      "loss": 0.6307,
+      "step": 2474
+    },
+    {
+      "epoch": 0.10871927725391616,
+      "grad_norm": 3.0625,
+      "learning_rate": 4.8626271559948036e-05,
+      "loss": 0.6062,
+      "step": 2476
+    },
+    {
+      "epoch": 0.10880709573311964,
+      "grad_norm": 3.34375,
+      "learning_rate": 4.8624008906920714e-05,
+      "loss": 0.6003,
+      "step": 2478
+    },
+    {
+      "epoch": 0.10889491421232313,
+      "grad_norm": 3.0625,
+      "learning_rate": 4.862174444476113e-05,
+      "loss": 0.6069,
+      "step": 2480
+    },
+    {
+      "epoch": 0.10898273269152661,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.86194781736427e-05,
+      "loss": 0.5926,
+      "step": 2482
+    },
+    {
+      "epoch": 0.1090705511707301,
+      "grad_norm": 3.5625,
+      "learning_rate": 4.861721009373897e-05,
+      "loss": 0.6071,
+      "step": 2484
+    },
+    {
+      "epoch": 0.10915836964993358,
+      "grad_norm": 3.046875,
+      "learning_rate": 4.8614940205223625e-05,
+      "loss": 0.585,
+      "step": 2486
+    },
+    {
+      "epoch": 0.10924618812913707,
+      "grad_norm": 3.6875,
+      "learning_rate": 4.86126685082705e-05,
+      "loss": 0.5705,
+      "step": 2488
+    },
+    {
+      "epoch": 0.10933400660834056,
+      "grad_norm": 2.953125,
+      "learning_rate": 4.861039500305356e-05,
+      "loss": 0.6318,
+      "step": 2490
+    },
+    {
+      "epoch": 0.10942182508754404,
+      "grad_norm": 3.203125,
+      "learning_rate": 4.860811968974691e-05,
+      "loss": 0.5763,
+      "step": 2492
+    },
+    {
+      "epoch": 0.10950964356674753,
+      "grad_norm": 3.46875,
+      "learning_rate": 4.86058425685248e-05,
+      "loss": 0.619,
+      "step": 2494
+    },
+    {
+      "epoch": 0.10959746204595101,
+      "grad_norm": 3.640625,
+      "learning_rate": 4.86035636395616e-05,
+      "loss": 0.6244,
+      "step": 2496
+    },
+    {
+      "epoch": 0.1096852805251545,
+      "grad_norm": 3.265625,
+      "learning_rate": 4.860128290303184e-05,
+      "loss": 0.5996,
+      "step": 2498
+    },
+    {
+      "epoch": 0.10977309900435799,
+      "grad_norm": 3.203125,
+      "learning_rate": 4.859900035911018e-05,
+      "loss": 0.5985,
+      "step": 2500
+    },
+    {
+      "epoch": 0.10986091748356148,
+      "grad_norm": 3.671875,
+      "learning_rate": 4.859671600797141e-05,
+      "loss": 0.6311,
+      "step": 2502
+    },
+    {
+      "epoch": 0.10994873596276497,
+      "grad_norm": 4.46875,
+      "learning_rate": 4.8594429849790476e-05,
+      "loss": 0.6137,
+      "step": 2504
+    },
+    {
+      "epoch": 0.11003655444196846,
+      "grad_norm": 4.53125,
+      "learning_rate": 4.8592141884742445e-05,
+      "loss": 0.5877,
+      "step": 2506
+    },
+    {
+      "epoch": 0.11012437292117194,
+      "grad_norm": 3.265625,
+      "learning_rate": 4.8589852113002546e-05,
+      "loss": 0.5774,
+      "step": 2508
+    },
+    {
+      "epoch": 0.11021219140037543,
+      "grad_norm": 2.984375,
+      "learning_rate": 4.85875605347461e-05,
+      "loss": 0.6018,
+      "step": 2510
+    },
+    {
+      "epoch": 0.11030000987957891,
+      "grad_norm": 3.25,
+      "learning_rate": 4.8585267150148625e-05,
+      "loss": 0.5877,
+      "step": 2512
+    },
+    {
+      "epoch": 0.1103878283587824,
+      "grad_norm": 4.59375,
+      "learning_rate": 4.8582971959385735e-05,
+      "loss": 0.5886,
+      "step": 2514
+    },
+    {
+      "epoch": 0.11047564683798589,
+      "grad_norm": 3.6875,
+      "learning_rate": 4.85806749626332e-05,
+      "loss": 0.6101,
+      "step": 2516
+    },
+    {
+      "epoch": 0.11056346531718937,
+      "grad_norm": 3.421875,
+      "learning_rate": 4.8578376160066916e-05,
+      "loss": 0.5614,
+      "step": 2518
+    },
+    {
+      "epoch": 0.11065128379639286,
+      "grad_norm": 4.53125,
+      "learning_rate": 4.857607555186294e-05,
+      "loss": 0.596,
+      "step": 2520
+    },
+    {
+      "epoch": 0.11073910227559634,
+      "grad_norm": 3.65625,
+      "learning_rate": 4.857377313819745e-05,
+      "loss": 0.6134,
+      "step": 2522
+    },
+    {
+      "epoch": 0.11082692075479983,
+      "grad_norm": 2.96875,
+      "learning_rate": 4.8571468919246755e-05,
+      "loss": 0.5675,
+      "step": 2524
+    },
+    {
+      "epoch": 0.11091473923400332,
+      "grad_norm": 3.484375,
+      "learning_rate": 4.8569162895187324e-05,
+      "loss": 0.5876,
+      "step": 2526
+    },
+    {
+      "epoch": 0.1110025577132068,
+      "grad_norm": 3.953125,
+      "learning_rate": 4.856685506619575e-05,
+      "loss": 0.6046,
+      "step": 2528
+    },
+    {
+      "epoch": 0.11109037619241029,
+      "grad_norm": 4.125,
+      "learning_rate": 4.8564545432448763e-05,
+      "loss": 0.5904,
+      "step": 2530
+    },
+    {
+      "epoch": 0.11117819467161377,
+      "grad_norm": 3.515625,
+      "learning_rate": 4.856223399412324e-05,
+      "loss": 0.5889,
+      "step": 2532
+    },
+    {
+      "epoch": 0.11126601315081726,
+      "grad_norm": 3.203125,
+      "learning_rate": 4.855992075139618e-05,
+      "loss": 0.5859,
+      "step": 2534
+    },
+    {
+      "epoch": 0.11135383163002074,
+      "grad_norm": 3.578125,
+      "learning_rate": 4.8557605704444754e-05,
+      "loss": 0.6318,
+      "step": 2536
+    },
+    {
+      "epoch": 0.11144165010922423,
+      "grad_norm": 3.234375,
+      "learning_rate": 4.8555288853446226e-05,
+      "loss": 0.6019,
+      "step": 2538
+    },
+    {
+      "epoch": 0.11152946858842772,
+      "grad_norm": 3.46875,
+      "learning_rate": 4.8552970198578044e-05,
+      "loss": 0.5788,
+      "step": 2540
+    },
+    {
+      "epoch": 0.1116172870676312,
+      "grad_norm": 3.015625,
+      "learning_rate": 4.8550649740017744e-05,
+      "loss": 0.6114,
+      "step": 2542
+    },
+    {
+      "epoch": 0.11170510554683469,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.854832747794305e-05,
+      "loss": 0.5551,
+      "step": 2544
+    },
+    {
+      "epoch": 0.11179292402603817,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.8546003412531785e-05,
+      "loss": 0.5562,
+      "step": 2546
+    },
+    {
+      "epoch": 0.11188074250524166,
+      "grad_norm": 3.078125,
+      "learning_rate": 4.854367754396194e-05,
+      "loss": 0.6005,
+      "step": 2548
+    },
+    {
+      "epoch": 0.11196856098444515,
+      "grad_norm": 3.65625,
+      "learning_rate": 4.854134987241162e-05,
+      "loss": 0.5951,
+      "step": 2550
+    },
+    {
+      "epoch": 0.11205637946364864,
+      "grad_norm": 4.4375,
+      "learning_rate": 4.85390203980591e-05,
+      "loss": 0.5995,
+      "step": 2552
+    },
+    {
+      "epoch": 0.11214419794285213,
+      "grad_norm": 4.9375,
+      "learning_rate": 4.853668912108273e-05,
+      "loss": 0.628,
+      "step": 2554
+    },
+    {
+      "epoch": 0.11223201642205562,
+      "grad_norm": 3.75,
+      "learning_rate": 4.8534356041661085e-05,
+      "loss": 0.6037,
+      "step": 2556
+    },
+    {
+      "epoch": 0.1123198349012591,
+      "grad_norm": 3.46875,
+      "learning_rate": 4.8532021159972804e-05,
+      "loss": 0.5707,
+      "step": 2558
+    },
+    {
+      "epoch": 0.11240765338046259,
+      "grad_norm": 3.4375,
+      "learning_rate": 4.8529684476196705e-05,
+      "loss": 0.6165,
+      "step": 2560
+    },
+    {
+      "epoch": 0.11249547185966607,
+      "grad_norm": 3.234375,
+      "learning_rate": 4.852734599051173e-05,
+      "loss": 0.6,
+      "step": 2562
+    },
+    {
+      "epoch": 0.11258329033886956,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.852500570309695e-05,
+      "loss": 0.5954,
+      "step": 2564
+    },
+    {
+      "epoch": 0.11267110881807305,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.8522663614131603e-05,
+      "loss": 0.6065,
+      "step": 2566
+    },
+    {
+      "epoch": 0.11275892729727653,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.8520319723795036e-05,
+      "loss": 0.6159,
+      "step": 2568
+    },
+    {
+      "epoch": 0.11284674577648002,
+      "grad_norm": 3.40625,
+      "learning_rate": 4.8517974032266745e-05,
+      "loss": 0.5932,
+      "step": 2570
+    },
+    {
+      "epoch": 0.1129345642556835,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.851562653972637e-05,
+      "loss": 0.5993,
+      "step": 2572
+    },
+    {
+      "epoch": 0.11302238273488699,
+      "grad_norm": 3.375,
+      "learning_rate": 4.851327724635366e-05,
+      "loss": 0.5716,
+      "step": 2574
+    },
+    {
+      "epoch": 0.11311020121409048,
+      "grad_norm": 3.328125,
+      "learning_rate": 4.851092615232856e-05,
+      "loss": 0.5877,
+      "step": 2576
+    },
+    {
+      "epoch": 0.11319801969329396,
+      "grad_norm": 3.40625,
+      "learning_rate": 4.8508573257831094e-05,
+      "loss": 0.5821,
+      "step": 2578
+    },
+    {
+      "epoch": 0.11328583817249745,
+      "grad_norm": 3.03125,
+      "learning_rate": 4.850621856304145e-05,
+      "loss": 0.5923,
+      "step": 2580
+    },
+    {
+      "epoch": 0.11337365665170093,
+      "grad_norm": 3.359375,
+      "learning_rate": 4.850386206813996e-05,
+      "loss": 0.6087,
+      "step": 2582
+    },
+    {
+      "epoch": 0.11346147513090442,
+      "grad_norm": 3.40625,
+      "learning_rate": 4.8501503773307075e-05,
+      "loss": 0.6313,
+      "step": 2584
+    },
+    {
+      "epoch": 0.1135492936101079,
+      "grad_norm": 3.40625,
+      "learning_rate": 4.849914367872339e-05,
+      "loss": 0.5768,
+      "step": 2586
+    },
+    {
+      "epoch": 0.11363711208931139,
+      "grad_norm": 3.40625,
+      "learning_rate": 4.849678178456966e-05,
+      "loss": 0.5633,
+      "step": 2588
+    },
+    {
+      "epoch": 0.11372493056851488,
+      "grad_norm": 3.015625,
+      "learning_rate": 4.8494418091026745e-05,
+      "loss": 0.5882,
+      "step": 2590
+    },
+    {
+      "epoch": 0.11381274904771836,
+      "grad_norm": 3.015625,
+      "learning_rate": 4.849205259827566e-05,
+      "loss": 0.5957,
+      "step": 2592
+    },
+    {
+      "epoch": 0.11390056752692185,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.8489685306497554e-05,
+      "loss": 0.5899,
+      "step": 2594
+    },
+    {
+      "epoch": 0.11398838600612533,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.8487316215873715e-05,
+      "loss": 0.5853,
+      "step": 2596
+    },
+    {
+      "epoch": 0.11407620448532882,
+      "grad_norm": 3.078125,
+      "learning_rate": 4.848494532658557e-05,
+      "loss": 0.5899,
+      "step": 2598
+    },
+    {
+      "epoch": 0.1141640229645323,
+      "grad_norm": 3.21875,
+      "learning_rate": 4.848257263881469e-05,
+      "loss": 0.5872,
+      "step": 2600
+    },
+    {
+      "epoch": 0.1142518414437358,
+      "grad_norm": 3.046875,
+      "learning_rate": 4.848019815274276e-05,
+      "loss": 0.5806,
+      "step": 2602
+    },
+    {
+      "epoch": 0.11433965992293929,
+      "grad_norm": 3.046875,
+      "learning_rate": 4.847782186855163e-05,
+      "loss": 0.5754,
+      "step": 2604
+    },
+    {
+      "epoch": 0.11442747840214278,
+      "grad_norm": 3.0625,
+      "learning_rate": 4.847544378642327e-05,
+      "loss": 0.6083,
+      "step": 2606
+    },
+    {
+      "epoch": 0.11451529688134626,
+      "grad_norm": 3.046875,
+      "learning_rate": 4.8473063906539804e-05,
+      "loss": 0.5894,
+      "step": 2608
+    },
+    {
+      "epoch": 0.11460311536054975,
+      "grad_norm": 3.015625,
+      "learning_rate": 4.8470682229083477e-05,
+      "loss": 0.6113,
+      "step": 2610
+    },
+    {
+      "epoch": 0.11469093383975323,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.846829875423667e-05,
+      "loss": 0.5863,
+      "step": 2612
+    },
+    {
+      "epoch": 0.11477875231895672,
+      "grad_norm": 3.390625,
+      "learning_rate": 4.846591348218192e-05,
+      "loss": 0.5958,
+      "step": 2614
+    },
+    {
+      "epoch": 0.1148665707981602,
+      "grad_norm": 3.421875,
+      "learning_rate": 4.84635264131019e-05,
+      "loss": 0.5882,
+      "step": 2616
+    },
+    {
+      "epoch": 0.11495438927736369,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.84611375471794e-05,
+      "loss": 0.6003,
+      "step": 2618
+    },
+    {
+      "epoch": 0.11504220775656718,
+      "grad_norm": 3.5625,
+      "learning_rate": 4.845874688459736e-05,
+      "loss": 0.6034,
+      "step": 2620
+    },
+    {
+      "epoch": 0.11513002623577066,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.845635442553885e-05,
+      "loss": 0.5967,
+      "step": 2622
+    },
+    {
+      "epoch": 0.11521784471497415,
+      "grad_norm": 3.046875,
+      "learning_rate": 4.8453960170187104e-05,
+      "loss": 0.593,
+      "step": 2624
+    },
+    {
+      "epoch": 0.11530566319417763,
+      "grad_norm": 3.234375,
+      "learning_rate": 4.8451564118725474e-05,
+      "loss": 0.5784,
+      "step": 2626
+    },
+    {
+      "epoch": 0.11539348167338112,
+      "grad_norm": 3.0625,
+      "learning_rate": 4.8449166271337434e-05,
+      "loss": 0.5907,
+      "step": 2628
+    },
+    {
+      "epoch": 0.1154813001525846,
+      "grad_norm": 3.265625,
+      "learning_rate": 4.844676662820662e-05,
+      "loss": 0.5818,
+      "step": 2630
+    },
+    {
+      "epoch": 0.11556911863178809,
+      "grad_norm": 2.875,
+      "learning_rate": 4.8444365189516796e-05,
+      "loss": 0.6006,
+      "step": 2632
+    },
+    {
+      "epoch": 0.11565693711099158,
+      "grad_norm": 3.296875,
+      "learning_rate": 4.8441961955451865e-05,
+      "loss": 0.5846,
+      "step": 2634
+    },
+    {
+      "epoch": 0.11574475559019506,
+      "grad_norm": 2.953125,
+      "learning_rate": 4.843955692619587e-05,
+      "loss": 0.5827,
+      "step": 2636
+    },
+    {
+      "epoch": 0.11583257406939855,
+      "grad_norm": 2.375,
+      "learning_rate": 4.8437150101932996e-05,
+      "loss": 0.6059,
+      "step": 2638
+    },
+    {
+      "epoch": 0.11592039254860204,
+      "grad_norm": 3.078125,
+      "learning_rate": 4.843474148284753e-05,
+      "loss": 0.5945,
+      "step": 2640
+    },
+    {
+      "epoch": 0.11600821102780552,
+      "grad_norm": 3.421875,
+      "learning_rate": 4.843233106912396e-05,
+      "loss": 0.5843,
+      "step": 2642
+    },
+    {
+      "epoch": 0.11609602950700901,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.842991886094686e-05,
+      "loss": 0.5994,
+      "step": 2644
+    },
+    {
+      "epoch": 0.1161838479862125,
+      "grad_norm": 3.625,
+      "learning_rate": 4.842750485850094e-05,
+      "loss": 0.5767,
+      "step": 2646
+    },
+    {
+      "epoch": 0.11627166646541598,
+      "grad_norm": 3.625,
+      "learning_rate": 4.8425089061971094e-05,
+      "loss": 0.5543,
+      "step": 2648
+    },
+    {
+      "epoch": 0.11635948494461947,
+      "grad_norm": 3.140625,
+      "learning_rate": 4.8422671471542314e-05,
+      "loss": 0.6315,
+      "step": 2650
+    },
+    {
+      "epoch": 0.11644730342382296,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.842025208739973e-05,
+      "loss": 0.5896,
+      "step": 2652
+    },
+    {
+      "epoch": 0.11653512190302645,
+      "grad_norm": 3.0,
+      "learning_rate": 4.8417830909728637e-05,
+      "loss": 0.5782,
+      "step": 2654
+    },
+    {
+      "epoch": 0.11662294038222994,
+      "grad_norm": 3.046875,
+      "learning_rate": 4.841540793871443e-05,
+      "loss": 0.58,
+      "step": 2656
+    },
+    {
+      "epoch": 0.11671075886143342,
+      "grad_norm": 3.125,
+      "learning_rate": 4.841298317454267e-05,
+      "loss": 0.5825,
+      "step": 2658
+    },
+    {
+      "epoch": 0.11679857734063691,
+      "grad_norm": 3.28125,
+      "learning_rate": 4.841055661739905e-05,
+      "loss": 0.6037,
+      "step": 2660
+    },
+    {
+      "epoch": 0.1168863958198404,
+      "grad_norm": 3.28125,
+      "learning_rate": 4.8408128267469394e-05,
+      "loss": 0.6085,
+      "step": 2662
+    },
+    {
+      "epoch": 0.11697421429904388,
+      "grad_norm": 2.75,
+      "learning_rate": 4.840569812493966e-05,
+      "loss": 0.6025,
+      "step": 2664
+    },
+    {
+      "epoch": 0.11706203277824737,
+      "grad_norm": 3.0625,
+      "learning_rate": 4.840326618999595e-05,
+      "loss": 0.5656,
+      "step": 2666
+    },
+    {
+      "epoch": 0.11714985125745085,
+      "grad_norm": 3.125,
+      "learning_rate": 4.840083246282452e-05,
+      "loss": 0.5908,
+      "step": 2668
+    },
+    {
+      "epoch": 0.11723766973665434,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.8398396943611715e-05,
+      "loss": 0.5884,
+      "step": 2670
+    },
+    {
+      "epoch": 0.11732548821585782,
+      "grad_norm": 3.328125,
+      "learning_rate": 4.839595963254407e-05,
+      "loss": 0.6311,
+      "step": 2672
+    },
+    {
+      "epoch": 0.11741330669506131,
+      "grad_norm": 3.21875,
+      "learning_rate": 4.8393520529808224e-05,
+      "loss": 0.603,
+      "step": 2674
+    },
+    {
+      "epoch": 0.1175011251742648,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.839107963559097e-05,
+      "loss": 0.5678,
+      "step": 2676
+    },
+    {
+      "epoch": 0.11758894365346828,
+      "grad_norm": 3.25,
+      "learning_rate": 4.838863695007923e-05,
+      "loss": 0.6127,
+      "step": 2678
+    },
+    {
+      "epoch": 0.11767676213267177,
+      "grad_norm": 3.234375,
+      "learning_rate": 4.838619247346007e-05,
+      "loss": 0.6211,
+      "step": 2680
+    },
+    {
+      "epoch": 0.11776458061187525,
+      "grad_norm": 3.046875,
+      "learning_rate": 4.838374620592068e-05,
+      "loss": 0.6083,
+      "step": 2682
+    },
+    {
+      "epoch": 0.11785239909107874,
+      "grad_norm": 3.234375,
+      "learning_rate": 4.83812981476484e-05,
+      "loss": 0.5768,
+      "step": 2684
+    },
+    {
+      "epoch": 0.11794021757028222,
+      "grad_norm": 3.265625,
+      "learning_rate": 4.8378848298830706e-05,
+      "loss": 0.5962,
+      "step": 2686
+    },
+    {
+      "epoch": 0.11802803604948571,
+      "grad_norm": 5.21875,
+      "learning_rate": 4.83763966596552e-05,
+      "loss": 0.5941,
+      "step": 2688
+    },
+    {
+      "epoch": 0.1181158545286892,
+      "grad_norm": 5.71875,
+      "learning_rate": 4.837394323030964e-05,
+      "loss": 0.6159,
+      "step": 2690
+    },
+    {
+      "epoch": 0.11820367300789268,
+      "grad_norm": 4.65625,
+      "learning_rate": 4.8371488010981894e-05,
+      "loss": 0.5818,
+      "step": 2692
+    },
+    {
+      "epoch": 0.11829149148709617,
+      "grad_norm": 4.09375,
+      "learning_rate": 4.8369031001860005e-05,
+      "loss": 0.5715,
+      "step": 2694
+    },
+    {
+      "epoch": 0.11837930996629965,
+      "grad_norm": 4.375,
+      "learning_rate": 4.836657220313211e-05,
+      "loss": 0.5782,
+      "step": 2696
+    },
+    {
+      "epoch": 0.11846712844550314,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.8364111614986527e-05,
+      "loss": 0.5572,
+      "step": 2698
+    },
+    {
+      "epoch": 0.11855494692470663,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.836164923761166e-05,
+      "loss": 0.603,
+      "step": 2700
+    },
+    {
+      "epoch": 0.11864276540391011,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.835918507119611e-05,
+      "loss": 0.5851,
+      "step": 2702
+    },
+    {
+      "epoch": 0.11873058388311361,
+      "grad_norm": 3.203125,
+      "learning_rate": 4.8356719115928564e-05,
+      "loss": 0.602,
+      "step": 2704
+    },
+    {
+      "epoch": 0.1188184023623171,
+      "grad_norm": 2.984375,
+      "learning_rate": 4.835425137199786e-05,
+      "loss": 0.5614,
+      "step": 2706
+    },
+    {
+      "epoch": 0.11890622084152058,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.835178183959299e-05,
+      "loss": 0.6055,
+      "step": 2708
+    },
+    {
+      "epoch": 0.11899403932072407,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.834931051890308e-05,
+      "loss": 0.5567,
+      "step": 2710
+    },
+    {
+      "epoch": 0.11908185779992755,
+      "grad_norm": 3.375,
+      "learning_rate": 4.8346837410117365e-05,
+      "loss": 0.5798,
+      "step": 2712
+    },
+    {
+      "epoch": 0.11916967627913104,
+      "grad_norm": 4.0625,
+      "learning_rate": 4.834436251342524e-05,
+      "loss": 0.5719,
+      "step": 2714
+    },
+    {
+      "epoch": 0.11925749475833453,
+      "grad_norm": 4.125,
+      "learning_rate": 4.834188582901624e-05,
+      "loss": 0.6048,
+      "step": 2716
+    },
+    {
+      "epoch": 0.11934531323753801,
+      "grad_norm": 3.296875,
+      "learning_rate": 4.833940735708003e-05,
+      "loss": 0.5629,
+      "step": 2718
+    },
+    {
+      "epoch": 0.1194331317167415,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.8336927097806415e-05,
+      "loss": 0.5777,
+      "step": 2720
+    },
+    {
+      "epoch": 0.11952095019594498,
+      "grad_norm": 3.171875,
+      "learning_rate": 4.833444505138533e-05,
+      "loss": 0.5959,
+      "step": 2722
+    },
+    {
+      "epoch": 0.11960876867514847,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.833196121800684e-05,
+      "loss": 0.5682,
+      "step": 2724
+    },
+    {
+      "epoch": 0.11969658715435195,
+      "grad_norm": 3.03125,
+      "learning_rate": 4.832947559786116e-05,
+      "loss": 0.5562,
+      "step": 2726
+    },
+    {
+      "epoch": 0.11978440563355544,
+      "grad_norm": 4.28125,
+      "learning_rate": 4.8326988191138664e-05,
+      "loss": 0.5755,
+      "step": 2728
+    },
+    {
+      "epoch": 0.11987222411275893,
+      "grad_norm": 4.375,
+      "learning_rate": 4.832449899802981e-05,
+      "loss": 0.5914,
+      "step": 2730
+    },
+    {
+      "epoch": 0.11996004259196241,
+      "grad_norm": 5.53125,
+      "learning_rate": 4.832200801872523e-05,
+      "loss": 0.5999,
+      "step": 2732
+    },
+    {
+      "epoch": 0.1200478610711659,
+      "grad_norm": 6.25,
+      "learning_rate": 4.8319515253415695e-05,
+      "loss": 0.5996,
+      "step": 2734
+    },
+    {
+      "epoch": 0.12013567955036938,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.831702070229208e-05,
+      "loss": 0.6215,
+      "step": 2736
+    },
+    {
+      "epoch": 0.12022349802957287,
+      "grad_norm": 2.75,
+      "learning_rate": 4.831452436554544e-05,
+      "loss": 0.6128,
+      "step": 2738
+    },
+    {
+      "epoch": 0.12031131650877636,
+      "grad_norm": 2.96875,
+      "learning_rate": 4.831202624336693e-05,
+      "loss": 0.5602,
+      "step": 2740
+    },
+    {
+      "epoch": 0.12039913498797984,
+      "grad_norm": 3.109375,
+      "learning_rate": 4.830952633594786e-05,
+      "loss": 0.5663,
+      "step": 2742
+    },
+    {
+      "epoch": 0.12048695346718333,
+      "grad_norm": 3.359375,
+      "learning_rate": 4.8307024643479684e-05,
+      "loss": 0.5638,
+      "step": 2744
+    },
+    {
+      "epoch": 0.12057477194638681,
+      "grad_norm": 4.5625,
+      "learning_rate": 4.830452116615397e-05,
+      "loss": 0.5799,
+      "step": 2746
+    },
+    {
+      "epoch": 0.1206625904255903,
+      "grad_norm": 3.953125,
+      "learning_rate": 4.830201590416243e-05,
+      "loss": 0.5996,
+      "step": 2748
+    },
+    {
+      "epoch": 0.12075040890479379,
+      "grad_norm": 3.25,
+      "learning_rate": 4.829950885769694e-05,
+      "loss": 0.5995,
+      "step": 2750
+    },
+    {
+      "epoch": 0.12083822738399727,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.8297000026949466e-05,
+      "loss": 0.5927,
+      "step": 2752
+    },
+    {
+      "epoch": 0.12092604586320077,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.829448941211215e-05,
+      "loss": 0.5874,
+      "step": 2754
+    },
+    {
+      "epoch": 0.12101386434240426,
+      "grad_norm": 3.21875,
+      "learning_rate": 4.8291977013377254e-05,
+      "loss": 0.5724,
+      "step": 2756
+    },
+    {
+      "epoch": 0.12110168282160774,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.828946283093717e-05,
+      "loss": 0.6041,
+      "step": 2758
+    },
+    {
+      "epoch": 0.12118950130081123,
+      "grad_norm": 3.09375,
+      "learning_rate": 4.828694686498444e-05,
+      "loss": 0.5648,
+      "step": 2760
+    },
+    {
+      "epoch": 0.12127731978001471,
+      "grad_norm": 3.265625,
+      "learning_rate": 4.8284429115711735e-05,
+      "loss": 0.5902,
+      "step": 2762
+    },
+    {
+      "epoch": 0.1213651382592182,
+      "grad_norm": 2.75,
+      "learning_rate": 4.828190958331187e-05,
+      "loss": 0.5663,
+      "step": 2764
+    },
+    {
+      "epoch": 0.12145295673842169,
+      "grad_norm": 2.953125,
+      "learning_rate": 4.827938826797779e-05,
+      "loss": 0.5614,
+      "step": 2766
+    },
+    {
+      "epoch": 0.12154077521762517,
+      "grad_norm": 3.78125,
+      "learning_rate": 4.827686516990257e-05,
+      "loss": 0.581,
+      "step": 2768
+    },
+    {
+      "epoch": 0.12162859369682866,
+      "grad_norm": 4.375,
+      "learning_rate": 4.827434028927944e-05,
+      "loss": 0.5845,
+      "step": 2770
+    },
+    {
+      "epoch": 0.12171641217603214,
+      "grad_norm": 3.65625,
+      "learning_rate": 4.827181362630175e-05,
+      "loss": 0.5912,
+      "step": 2772
+    },
+    {
+      "epoch": 0.12180423065523563,
+      "grad_norm": 3.015625,
+      "learning_rate": 4.8269285181162995e-05,
+      "loss": 0.5885,
+      "step": 2774
+    },
+    {
+      "epoch": 0.12189204913443911,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.82667549540568e-05,
+      "loss": 0.5797,
+      "step": 2776
+    },
+    {
+      "epoch": 0.1219798676136426,
+      "grad_norm": 3.25,
+      "learning_rate": 4.826422294517693e-05,
+      "loss": 0.5669,
+      "step": 2778
+    },
+    {
+      "epoch": 0.12206768609284609,
+      "grad_norm": 3.125,
+      "learning_rate": 4.826168915471729e-05,
+      "loss": 0.6038,
+      "step": 2780
+    },
+    {
+      "epoch": 0.12215550457204957,
+      "grad_norm": 3.609375,
+      "learning_rate": 4.825915358287193e-05,
+      "loss": 0.578,
+      "step": 2782
+    },
+    {
+      "epoch": 0.12224332305125306,
+      "grad_norm": 3.140625,
+      "learning_rate": 4.8256616229835e-05,
+      "loss": 0.5762,
+      "step": 2784
+    },
+    {
+      "epoch": 0.12233114153045654,
+      "grad_norm": 3.96875,
+      "learning_rate": 4.825407709580083e-05,
+      "loss": 0.612,
+      "step": 2786
+    },
+    {
+      "epoch": 0.12241896000966003,
+      "grad_norm": 3.28125,
+      "learning_rate": 4.825153618096386e-05,
+      "loss": 0.6211,
+      "step": 2788
+    },
+    {
+      "epoch": 0.12250677848886352,
+      "grad_norm": 3.6875,
+      "learning_rate": 4.824899348551868e-05,
+      "loss": 0.5634,
+      "step": 2790
+    },
+    {
+      "epoch": 0.122594596968067,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.8246449009660004e-05,
+      "loss": 0.5689,
+      "step": 2792
+    },
+    {
+      "epoch": 0.12268241544727049,
+      "grad_norm": 3.25,
+      "learning_rate": 4.8243902753582695e-05,
+      "loss": 0.6071,
+      "step": 2794
+    },
+    {
+      "epoch": 0.12277023392647397,
+      "grad_norm": 3.15625,
+      "learning_rate": 4.8241354717481734e-05,
+      "loss": 0.5847,
+      "step": 2796
+    },
+    {
+      "epoch": 0.12285805240567746,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.823880490155226e-05,
+      "loss": 0.5787,
+      "step": 2798
+    },
+    {
+      "epoch": 0.12294587088488095,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.823625330598953e-05,
+      "loss": 0.5927,
+      "step": 2800
+    },
+    {
+      "epoch": 0.12303368936408443,
+      "grad_norm": 3.4375,
+      "learning_rate": 4.823369993098896e-05,
+      "loss": 0.593,
+      "step": 2802
+    },
+    {
+      "epoch": 0.12312150784328793,
+      "grad_norm": 3.203125,
+      "learning_rate": 4.823114477674607e-05,
+      "loss": 0.5603,
+      "step": 2804
+    },
+    {
+      "epoch": 0.12320932632249142,
+      "grad_norm": 2.96875,
+      "learning_rate": 4.822858784345655e-05,
+      "loss": 0.5736,
+      "step": 2806
+    },
+    {
+      "epoch": 0.1232971448016949,
+      "grad_norm": 3.015625,
+      "learning_rate": 4.822602913131621e-05,
+      "loss": 0.594,
+      "step": 2808
+    },
+    {
+      "epoch": 0.12338496328089839,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.8223468640520984e-05,
+      "loss": 0.5968,
+      "step": 2810
+    },
+    {
+      "epoch": 0.12347278176010187,
+      "grad_norm": 3.0625,
+      "learning_rate": 4.822090637126697e-05,
+      "loss": 0.5796,
+      "step": 2812
+    },
+    {
+      "epoch": 0.12356060023930536,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.821834232375037e-05,
+      "loss": 0.5728,
+      "step": 2814
+    },
+    {
+      "epoch": 0.12364841871850885,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.8215776498167555e-05,
+      "loss": 0.5684,
+      "step": 2816
+    },
+    {
+      "epoch": 0.12373623719771233,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.8213208894715e-05,
+      "loss": 0.5754,
+      "step": 2818
+    },
+    {
+      "epoch": 0.12382405567691582,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.821063951358936e-05,
+      "loss": 0.579,
+      "step": 2820
+    },
+    {
+      "epoch": 0.1239118741561193,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.820806835498737e-05,
+      "loss": 0.6119,
+      "step": 2822
+    },
+    {
+      "epoch": 0.12399969263532279,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.820549541910595e-05,
+      "loss": 0.5824,
+      "step": 2824
+    },
+    {
+      "epoch": 0.12408751111452627,
+      "grad_norm": 3.28125,
+      "learning_rate": 4.820292070614212e-05,
+      "loss": 0.5643,
+      "step": 2826
+    },
+    {
+      "epoch": 0.12417532959372976,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.820034421629307e-05,
+      "loss": 0.5791,
+      "step": 2828
+    },
+    {
+      "epoch": 0.12426314807293325,
+      "grad_norm": 3.71875,
+      "learning_rate": 4.819776594975609e-05,
+      "loss": 0.5905,
+      "step": 2830
+    },
+    {
+      "epoch": 0.12435096655213673,
+      "grad_norm": 3.21875,
+      "learning_rate": 4.819518590672863e-05,
+      "loss": 0.6018,
+      "step": 2832
+    },
+    {
+      "epoch": 0.12443878503134022,
+      "grad_norm": 3.515625,
+      "learning_rate": 4.8192604087408285e-05,
+      "loss": 0.5928,
+      "step": 2834
+    },
+    {
+      "epoch": 0.1245266035105437,
+      "grad_norm": 4.15625,
+      "learning_rate": 4.819002049199276e-05,
+      "loss": 0.5817,
+      "step": 2836
+    },
+    {
+      "epoch": 0.12461442198974719,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.818743512067989e-05,
+      "loss": 0.6143,
+      "step": 2838
+    },
+    {
+      "epoch": 0.12470224046895068,
+      "grad_norm": 3.21875,
+      "learning_rate": 4.8184847973667695e-05,
+      "loss": 0.5789,
+      "step": 2840
+    },
+    {
+      "epoch": 0.12479005894815416,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.818225905115428e-05,
+      "loss": 0.5805,
+      "step": 2842
+    },
+    {
+      "epoch": 0.12487787742735765,
+      "grad_norm": 3.265625,
+      "learning_rate": 4.817966835333791e-05,
+      "loss": 0.5753,
+      "step": 2844
+    },
+    {
+      "epoch": 0.12496569590656113,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.817707588041698e-05,
+      "loss": 0.5686,
+      "step": 2846
+    },
+    {
+      "epoch": 0.12505351438576462,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.817448163259002e-05,
+      "loss": 0.5567,
+      "step": 2848
+    },
+    {
+      "epoch": 0.12514133286496812,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.8171885610055705e-05,
+      "loss": 0.5747,
+      "step": 2850
+    },
+    {
+      "epoch": 0.1252291513441716,
+      "grad_norm": 3.0,
+      "learning_rate": 4.816928781301283e-05,
+      "loss": 0.5622,
+      "step": 2852
+    },
+    {
+      "epoch": 0.1253169698233751,
+      "grad_norm": 3.03125,
+      "learning_rate": 4.816668824166035e-05,
+      "loss": 0.5603,
+      "step": 2854
+    },
+    {
+      "epoch": 0.12540478830257856,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.8164086896197325e-05,
+      "loss": 0.5682,
+      "step": 2856
+    },
+    {
+      "epoch": 0.12549260678178206,
+      "grad_norm": 3.09375,
+      "learning_rate": 4.816148377682297e-05,
+      "loss": 0.6089,
+      "step": 2858
+    },
+    {
+      "epoch": 0.12558042526098553,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.8158878883736644e-05,
+      "loss": 0.5962,
+      "step": 2860
+    },
+    {
+      "epoch": 0.12566824374018903,
+      "grad_norm": 3.375,
+      "learning_rate": 4.815627221713781e-05,
+      "loss": 0.5576,
+      "step": 2862
+    },
+    {
+      "epoch": 0.1257560622193925,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.815366377722611e-05,
+      "loss": 0.6009,
+      "step": 2864
+    },
+    {
+      "epoch": 0.125843880698596,
+      "grad_norm": 3.109375,
+      "learning_rate": 4.815105356420128e-05,
+      "loss": 0.5888,
+      "step": 2866
+    },
+    {
+      "epoch": 0.12593169917779948,
+      "grad_norm": 3.03125,
+      "learning_rate": 4.814844157826323e-05,
+      "loss": 0.5765,
+      "step": 2868
+    },
+    {
+      "epoch": 0.12601951765700298,
+      "grad_norm": 3.25,
+      "learning_rate": 4.814582781961195e-05,
+      "loss": 0.5808,
+      "step": 2870
+    },
+    {
+      "epoch": 0.12610733613620645,
+      "grad_norm": 3.515625,
+      "learning_rate": 4.814321228844765e-05,
+      "loss": 0.5683,
+      "step": 2872
+    },
+    {
+      "epoch": 0.12619515461540995,
+      "grad_norm": 3.25,
+      "learning_rate": 4.81405949849706e-05,
+      "loss": 0.5604,
+      "step": 2874
+    },
+    {
+      "epoch": 0.12628297309461342,
+      "grad_norm": 3.15625,
+      "learning_rate": 4.813797590938124e-05,
+      "loss": 0.5842,
+      "step": 2876
+    },
+    {
+      "epoch": 0.12637079157381692,
+      "grad_norm": 2.875,
+      "learning_rate": 4.8135355061880124e-05,
+      "loss": 0.577,
+      "step": 2878
+    },
+    {
+      "epoch": 0.12645861005302042,
+      "grad_norm": 4.0625,
+      "learning_rate": 4.813273244266799e-05,
+      "loss": 0.5972,
+      "step": 2880
+    },
+    {
+      "epoch": 0.1265464285322239,
+      "grad_norm": 4.6875,
+      "learning_rate": 4.8130108051945655e-05,
+      "loss": 0.5887,
+      "step": 2882
+    },
+    {
+      "epoch": 0.1266342470114274,
+      "grad_norm": 4.5625,
+      "learning_rate": 4.8127481889914096e-05,
+      "loss": 0.6029,
+      "step": 2884
+    },
+    {
+      "epoch": 0.12672206549063086,
+      "grad_norm": 3.609375,
+      "learning_rate": 4.812485395677443e-05,
+      "loss": 0.589,
+      "step": 2886
+    },
+    {
+      "epoch": 0.12680988396983436,
+      "grad_norm": 3.34375,
+      "learning_rate": 4.812222425272791e-05,
+      "loss": 0.5527,
+      "step": 2888
+    },
+    {
+      "epoch": 0.12689770244903784,
+      "grad_norm": 3.015625,
+      "learning_rate": 4.811959277797591e-05,
+      "loss": 0.6044,
+      "step": 2890
+    },
+    {
+      "epoch": 0.12698552092824134,
+      "grad_norm": 3.3125,
+      "learning_rate": 4.8116959532719954e-05,
+      "loss": 0.5792,
+      "step": 2892
+    },
+    {
+      "epoch": 0.1270733394074448,
+      "grad_norm": 3.734375,
+      "learning_rate": 4.81143245171617e-05,
+      "loss": 0.5781,
+      "step": 2894
+    },
+    {
+      "epoch": 0.1271611578866483,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.8111687731502925e-05,
+      "loss": 0.5827,
+      "step": 2896
+    },
+    {
+      "epoch": 0.12724897636585178,
+      "grad_norm": 2.984375,
+      "learning_rate": 4.8109049175945566e-05,
+      "loss": 0.5314,
+      "step": 2898
+    },
+    {
+      "epoch": 0.12733679484505528,
+      "grad_norm": 3.140625,
+      "learning_rate": 4.810640885069169e-05,
+      "loss": 0.5687,
+      "step": 2900
+    },
+    {
+      "epoch": 0.12742461332425875,
+      "grad_norm": 3.0,
+      "learning_rate": 4.810376675594347e-05,
+      "loss": 0.5552,
+      "step": 2902
+    },
+    {
+      "epoch": 0.12751243180346225,
+      "grad_norm": 3.375,
+      "learning_rate": 4.8101122891903263e-05,
+      "loss": 0.5637,
+      "step": 2904
+    },
+    {
+      "epoch": 0.12760025028266572,
+      "grad_norm": 3.265625,
+      "learning_rate": 4.809847725877352e-05,
+      "loss": 0.5685,
+      "step": 2906
+    },
+    {
+      "epoch": 0.12768806876186922,
+      "grad_norm": 3.109375,
+      "learning_rate": 4.8095829856756866e-05,
+      "loss": 0.581,
+      "step": 2908
+    },
+    {
+      "epoch": 0.1277758872410727,
+      "grad_norm": 3.421875,
+      "learning_rate": 4.809318068605602e-05,
+      "loss": 0.5857,
+      "step": 2910
+    },
+    {
+      "epoch": 0.1278637057202762,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.8090529746873845e-05,
+      "loss": 0.5687,
+      "step": 2912
+    },
+    {
+      "epoch": 0.12795152419947967,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.8087877039413386e-05,
+      "loss": 0.5868,
+      "step": 2914
+    },
+    {
+      "epoch": 0.12803934267868317,
+      "grad_norm": 3.03125,
+      "learning_rate": 4.8085222563877766e-05,
+      "loss": 0.5768,
+      "step": 2916
+    },
+    {
+      "epoch": 0.12812716115788664,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.808256632047026e-05,
+      "loss": 0.5833,
+      "step": 2918
+    },
+    {
+      "epoch": 0.12821497963709014,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.80799083093943e-05,
+      "loss": 0.5633,
+      "step": 2920
+    },
+    {
+      "epoch": 0.1283027981162936,
+      "grad_norm": 3.296875,
+      "learning_rate": 4.8077248530853416e-05,
+      "loss": 0.5658,
+      "step": 2922
+    },
+    {
+      "epoch": 0.1283906165954971,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.807458698505132e-05,
+      "loss": 0.6031,
+      "step": 2924
+    },
+    {
+      "epoch": 0.12847843507470058,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.807192367219182e-05,
+      "loss": 0.5631,
+      "step": 2926
+    },
+    {
+      "epoch": 0.12856625355390408,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.8069258592478875e-05,
+      "loss": 0.5809,
+      "step": 2928
+    },
+    {
+      "epoch": 0.12865407203310758,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.8066591746116575e-05,
+      "loss": 0.5332,
+      "step": 2930
+    },
+    {
+      "epoch": 0.12874189051231105,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.8063923133309144e-05,
+      "loss": 0.5767,
+      "step": 2932
+    },
+    {
+      "epoch": 0.12882970899151455,
+      "grad_norm": 3.5625,
+      "learning_rate": 4.8061252754260954e-05,
+      "loss": 0.5972,
+      "step": 2934
+    },
+    {
+      "epoch": 0.12891752747071802,
+      "grad_norm": 3.734375,
+      "learning_rate": 4.805858060917651e-05,
+      "loss": 0.5573,
+      "step": 2936
+    },
+    {
+      "epoch": 0.12900534594992152,
+      "grad_norm": 4.4375,
+      "learning_rate": 4.8055906698260424e-05,
+      "loss": 0.582,
+      "step": 2938
+    },
+    {
+      "epoch": 0.129093164429125,
+      "grad_norm": 4.09375,
+      "learning_rate": 4.805323102171748e-05,
+      "loss": 0.5729,
+      "step": 2940
+    },
+    {
+      "epoch": 0.1291809829083285,
+      "grad_norm": 4.40625,
+      "learning_rate": 4.805055357975257e-05,
+      "loss": 0.5866,
+      "step": 2942
+    },
+    {
+      "epoch": 0.12926880138753197,
+      "grad_norm": 4.59375,
+      "learning_rate": 4.804787437257075e-05,
+      "loss": 0.6013,
+      "step": 2944
+    },
+    {
+      "epoch": 0.12935661986673547,
+      "grad_norm": 4.59375,
+      "learning_rate": 4.8045193400377186e-05,
+      "loss": 0.5556,
+      "step": 2946
+    },
+    {
+      "epoch": 0.12944443834593894,
+      "grad_norm": 5.9375,
+      "learning_rate": 4.8042510663377184e-05,
+      "loss": 0.5914,
+      "step": 2948
+    },
+    {
+      "epoch": 0.12953225682514244,
+      "grad_norm": 4.375,
+      "learning_rate": 4.803982616177619e-05,
+      "loss": 0.5841,
+      "step": 2950
+    },
+    {
+      "epoch": 0.1296200753043459,
+      "grad_norm": 2.34375,
+      "learning_rate": 4.803713989577979e-05,
+      "loss": 0.5732,
+      "step": 2952
+    },
+    {
+      "epoch": 0.1297078937835494,
+      "grad_norm": 3.015625,
+      "learning_rate": 4.803445186559369e-05,
+      "loss": 0.5683,
+      "step": 2954
+    },
+    {
+      "epoch": 0.12979571226275288,
+      "grad_norm": 3.59375,
+      "learning_rate": 4.803176207142375e-05,
+      "loss": 0.5492,
+      "step": 2956
+    },
+    {
+      "epoch": 0.12988353074195638,
+      "grad_norm": 4.0625,
+      "learning_rate": 4.802907051347595e-05,
+      "loss": 0.5642,
+      "step": 2958
+    },
+    {
+      "epoch": 0.12997134922115985,
+      "grad_norm": 3.421875,
+      "learning_rate": 4.8026377191956404e-05,
+      "loss": 0.5762,
+      "step": 2960
+    },
+    {
+      "epoch": 0.13005916770036335,
+      "grad_norm": 3.625,
+      "learning_rate": 4.802368210707138e-05,
+      "loss": 0.5616,
+      "step": 2962
+    },
+    {
+      "epoch": 0.13014698617956683,
+      "grad_norm": 3.203125,
+      "learning_rate": 4.802098525902725e-05,
+      "loss": 0.557,
+      "step": 2964
+    },
+    {
+      "epoch": 0.13023480465877033,
+      "grad_norm": 2.875,
+      "learning_rate": 4.801828664803056e-05,
+      "loss": 0.5589,
+      "step": 2966
+    },
+    {
+      "epoch": 0.1303226231379738,
+      "grad_norm": 2.5,
+      "learning_rate": 4.8015586274287954e-05,
+      "loss": 0.5587,
+      "step": 2968
+    },
+    {
+      "epoch": 0.1304104416171773,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.8012884138006236e-05,
+      "loss": 0.5536,
+      "step": 2970
+    },
+    {
+      "epoch": 0.13049826009638077,
+      "grad_norm": 2.53125,
+      "learning_rate": 4.8010180239392336e-05,
+      "loss": 0.5832,
+      "step": 2972
+    },
+    {
+      "epoch": 0.13058607857558427,
+      "grad_norm": 2.75,
+      "learning_rate": 4.8007474578653315e-05,
+      "loss": 0.5525,
+      "step": 2974
+    },
+    {
+      "epoch": 0.13067389705478774,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.800476715599638e-05,
+      "loss": 0.554,
+      "step": 2976
+    },
+    {
+      "epoch": 0.13076171553399124,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.800205797162885e-05,
+      "loss": 0.5811,
+      "step": 2978
+    },
+    {
+      "epoch": 0.1308495340131947,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.7999347025758226e-05,
+      "loss": 0.5601,
+      "step": 2980
+    },
+    {
+      "epoch": 0.1309373524923982,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.7996634318592084e-05,
+      "loss": 0.5816,
+      "step": 2982
+    },
+    {
+      "epoch": 0.1310251709716017,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.7993919850338165e-05,
+      "loss": 0.5796,
+      "step": 2984
+    },
+    {
+      "epoch": 0.13111298945080518,
+      "grad_norm": 3.15625,
+      "learning_rate": 4.799120362120436e-05,
+      "loss": 0.5404,
+      "step": 2986
+    },
+    {
+      "epoch": 0.13120080793000868,
+      "grad_norm": 3.359375,
+      "learning_rate": 4.798848563139867e-05,
+      "loss": 0.5436,
+      "step": 2988
+    },
+    {
+      "epoch": 0.13128862640921216,
+      "grad_norm": 3.0,
+      "learning_rate": 4.7985765881129244e-05,
+      "loss": 0.5516,
+      "step": 2990
+    },
+    {
+      "epoch": 0.13137644488841566,
+      "grad_norm": 2.546875,
+      "learning_rate": 4.798304437060435e-05,
+      "loss": 0.5522,
+      "step": 2992
+    },
+    {
+      "epoch": 0.13146426336761913,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.798032110003241e-05,
+      "loss": 0.5484,
+      "step": 2994
+    },
+    {
+      "epoch": 0.13155208184682263,
+      "grad_norm": 3.078125,
+      "learning_rate": 4.797759606962196e-05,
+      "loss": 0.5625,
+      "step": 2996
+    },
+    {
+      "epoch": 0.1316399003260261,
+      "grad_norm": 3.984375,
+      "learning_rate": 4.797486927958171e-05,
+      "loss": 0.5799,
+      "step": 2998
+    },
+    {
+      "epoch": 0.1317277188052296,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.797214073012046e-05,
+      "loss": 0.5912,
+      "step": 3000
+    },
+    {
+      "epoch": 0.13181553728443307,
+      "grad_norm": 2.484375,
+      "learning_rate": 4.796941042144717e-05,
+      "loss": 0.5771,
+      "step": 3002
+    },
+    {
+      "epoch": 0.13190335576363657,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.796667835377092e-05,
+      "loss": 0.5842,
+      "step": 3004
+    },
+    {
+      "epoch": 0.13199117424284004,
+      "grad_norm": 3.328125,
+      "learning_rate": 4.796394452730094e-05,
+      "loss": 0.5651,
+      "step": 3006
+    },
+    {
+      "epoch": 0.13207899272204354,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.796120894224657e-05,
+      "loss": 0.5555,
+      "step": 3008
+    },
+    {
+      "epoch": 0.13216681120124701,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.795847159881733e-05,
+      "loss": 0.5811,
+      "step": 3010
+    },
+    {
+      "epoch": 0.13225462968045051,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.795573249722282e-05,
+      "loss": 0.5715,
+      "step": 3012
+    },
+    {
+      "epoch": 0.13234244815965399,
+      "grad_norm": 2.5,
+      "learning_rate": 4.795299163767282e-05,
+      "loss": 0.5762,
+      "step": 3014
+    },
+    {
+      "epoch": 0.13243026663885749,
+      "grad_norm": 2.984375,
+      "learning_rate": 4.7950249020377215e-05,
+      "loss": 0.5867,
+      "step": 3016
+    },
+    {
+      "epoch": 0.13251808511806096,
+      "grad_norm": 3.34375,
+      "learning_rate": 4.7947504645546034e-05,
+      "loss": 0.5801,
+      "step": 3018
+    },
+    {
+      "epoch": 0.13260590359726446,
+      "grad_norm": 3.203125,
+      "learning_rate": 4.794475851338946e-05,
+      "loss": 0.5964,
+      "step": 3020
+    },
+    {
+      "epoch": 0.13269372207646793,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.794201062411777e-05,
+      "loss": 0.54,
+      "step": 3022
+    },
+    {
+      "epoch": 0.13278154055567143,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.79392609779414e-05,
+      "loss": 0.5766,
+      "step": 3024
+    },
+    {
+      "epoch": 0.1328693590348749,
+      "grad_norm": 3.0,
+      "learning_rate": 4.7936509575070945e-05,
+      "loss": 0.6018,
+      "step": 3026
+    },
+    {
+      "epoch": 0.1329571775140784,
+      "grad_norm": 3.0,
+      "learning_rate": 4.793375641571707e-05,
+      "loss": 0.5358,
+      "step": 3028
+    },
+    {
+      "epoch": 0.13304499599328187,
+      "grad_norm": 3.109375,
+      "learning_rate": 4.793100150009064e-05,
+      "loss": 0.5788,
+      "step": 3030
+    },
+    {
+      "epoch": 0.13313281447248537,
+      "grad_norm": 3.015625,
+      "learning_rate": 4.7928244828402613e-05,
+      "loss": 0.5617,
+      "step": 3032
+    },
+    {
+      "epoch": 0.13322063295168887,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.7925486400864104e-05,
+      "loss": 0.5712,
+      "step": 3034
+    },
+    {
+      "epoch": 0.13330845143089234,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.7922726217686355e-05,
+      "loss": 0.5818,
+      "step": 3036
+    },
+    {
+      "epoch": 0.13339626991009584,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.7919964279080724e-05,
+      "loss": 0.58,
+      "step": 3038
+    },
+    {
+      "epoch": 0.13348408838929932,
+      "grad_norm": 3.015625,
+      "learning_rate": 4.7917200585258746e-05,
+      "loss": 0.563,
+      "step": 3040
+    },
+    {
+      "epoch": 0.13357190686850282,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.791443513643205e-05,
+      "loss": 0.5491,
+      "step": 3042
+    },
+    {
+      "epoch": 0.1336597253477063,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.791166793281242e-05,
+      "loss": 0.5647,
+      "step": 3044
+    },
+    {
+      "epoch": 0.1337475438269098,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.790889897461176e-05,
+      "loss": 0.5556,
+      "step": 3046
+    },
+    {
+      "epoch": 0.13383536230611326,
+      "grad_norm": 3.1875,
+      "learning_rate": 4.790612826204214e-05,
+      "loss": 0.5618,
+      "step": 3048
+    },
+    {
+      "epoch": 0.13392318078531676,
+      "grad_norm": 3.25,
+      "learning_rate": 4.7903355795315714e-05,
+      "loss": 0.5705,
+      "step": 3050
+    },
+    {
+      "epoch": 0.13401099926452023,
+      "grad_norm": 3.125,
+      "learning_rate": 4.790058157464481e-05,
+      "loss": 0.5471,
+      "step": 3052
+    },
+    {
+      "epoch": 0.13409881774372373,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.789780560024188e-05,
+      "loss": 0.6094,
+      "step": 3054
+    },
+    {
+      "epoch": 0.1341866362229272,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.789502787231952e-05,
+      "loss": 0.5525,
+      "step": 3056
+    },
+    {
+      "epoch": 0.1342744547021307,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.7892248391090426e-05,
+      "loss": 0.5455,
+      "step": 3058
+    },
+    {
+      "epoch": 0.13436227318133417,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.788946715676747e-05,
+      "loss": 0.5991,
+      "step": 3060
+    },
+    {
+      "epoch": 0.13445009166053767,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.788668416956362e-05,
+      "loss": 0.5645,
+      "step": 3062
+    },
+    {
+      "epoch": 0.13453791013974115,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.788389942969202e-05,
+      "loss": 0.5732,
+      "step": 3064
+    },
+    {
+      "epoch": 0.13462572861894465,
+      "grad_norm": 3.3125,
+      "learning_rate": 4.788111293736591e-05,
+      "loss": 0.5562,
+      "step": 3066
+    },
+    {
+      "epoch": 0.13471354709814812,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.7878324692798694e-05,
+      "loss": 0.5768,
+      "step": 3068
+    },
+    {
+      "epoch": 0.13480136557735162,
+      "grad_norm": 2.5,
+      "learning_rate": 4.787553469620388e-05,
+      "loss": 0.5782,
+      "step": 3070
+    },
+    {
+      "epoch": 0.1348891840565551,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.787274294779515e-05,
+      "loss": 0.5552,
+      "step": 3072
+    },
+    {
+      "epoch": 0.1349770025357586,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.7869949447786266e-05,
+      "loss": 0.5638,
+      "step": 3074
+    },
+    {
+      "epoch": 0.13506482101496206,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.7867154196391184e-05,
+      "loss": 0.5941,
+      "step": 3076
+    },
+    {
+      "epoch": 0.13515263949416556,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.786435719382394e-05,
+      "loss": 0.5407,
+      "step": 3078
+    },
+    {
+      "epoch": 0.13524045797336903,
+      "grad_norm": 3.109375,
+      "learning_rate": 4.7861558440298745e-05,
+      "loss": 0.5807,
+      "step": 3080
+    },
+    {
+      "epoch": 0.13532827645257253,
+      "grad_norm": 3.296875,
+      "learning_rate": 4.785875793602993e-05,
+      "loss": 0.5439,
+      "step": 3082
+    },
+    {
+      "epoch": 0.13541609493177603,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.785595568123195e-05,
+      "loss": 0.5342,
+      "step": 3084
+    },
+    {
+      "epoch": 0.1355039134109795,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.78531516761194e-05,
+      "loss": 0.5854,
+      "step": 3086
+    },
+    {
+      "epoch": 0.135591731890183,
+      "grad_norm": 2.96875,
+      "learning_rate": 4.785034592090702e-05,
+      "loss": 0.5676,
+      "step": 3088
+    },
+    {
+      "epoch": 0.13567955036938648,
+      "grad_norm": 3.203125,
+      "learning_rate": 4.784753841580967e-05,
+      "loss": 0.5573,
+      "step": 3090
+    },
+    {
+      "epoch": 0.13576736884858998,
+      "grad_norm": 3.375,
+      "learning_rate": 4.7844729161042355e-05,
+      "loss": 0.567,
+      "step": 3092
+    },
+    {
+      "epoch": 0.13585518732779345,
+      "grad_norm": 3.046875,
+      "learning_rate": 4.78419181568202e-05,
+      "loss": 0.5619,
+      "step": 3094
+    },
+    {
+      "epoch": 0.13594300580699695,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.783910540335848e-05,
+      "loss": 0.5454,
+      "step": 3096
+    },
+    {
+      "epoch": 0.13603082428620042,
+      "grad_norm": 3.4375,
+      "learning_rate": 4.783629090087259e-05,
+      "loss": 0.5401,
+      "step": 3098
+    },
+    {
+      "epoch": 0.13611864276540392,
+      "grad_norm": 4.375,
+      "learning_rate": 4.783347464957807e-05,
+      "loss": 0.5515,
+      "step": 3100
+    },
+    {
+      "epoch": 0.1362064612446074,
+      "grad_norm": 3.9375,
+      "learning_rate": 4.783065664969059e-05,
+      "loss": 0.5766,
+      "step": 3102
+    },
+    {
+      "epoch": 0.1362942797238109,
+      "grad_norm": 4.3125,
+      "learning_rate": 4.782783690142595e-05,
+      "loss": 0.5804,
+      "step": 3104
+    },
+    {
+      "epoch": 0.13638209820301436,
+      "grad_norm": 3.96875,
+      "learning_rate": 4.782501540500009e-05,
+      "loss": 0.5908,
+      "step": 3106
+    },
+    {
+      "epoch": 0.13646991668221786,
+      "grad_norm": 3.15625,
+      "learning_rate": 4.7822192160629074e-05,
+      "loss": 0.5553,
+      "step": 3108
+    },
+    {
+      "epoch": 0.13655773516142133,
+      "grad_norm": 3.796875,
+      "learning_rate": 4.781936716852912e-05,
+      "loss": 0.5416,
+      "step": 3110
+    },
+    {
+      "epoch": 0.13664555364062483,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.781654042891655e-05,
+      "loss": 0.5252,
+      "step": 3112
+    },
+    {
+      "epoch": 0.1367333721198283,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.781371194200784e-05,
+      "loss": 0.5582,
+      "step": 3114
+    },
+    {
+      "epoch": 0.1368211905990318,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.781088170801961e-05,
+      "loss": 0.5422,
+      "step": 3116
+    },
+    {
+      "epoch": 0.13690900907823528,
+      "grad_norm": 3.0625,
+      "learning_rate": 4.780804972716859e-05,
+      "loss": 0.5477,
+      "step": 3118
+    },
+    {
+      "epoch": 0.13699682755743878,
+      "grad_norm": 2.484375,
+      "learning_rate": 4.780521599967165e-05,
+      "loss": 0.567,
+      "step": 3120
+    },
+    {
+      "epoch": 0.13708464603664225,
+      "grad_norm": 3.0625,
+      "learning_rate": 4.78023805257458e-05,
+      "loss": 0.5809,
+      "step": 3122
+    },
+    {
+      "epoch": 0.13717246451584575,
+      "grad_norm": 2.75,
+      "learning_rate": 4.77995433056082e-05,
+      "loss": 0.5676,
+      "step": 3124
+    },
+    {
+      "epoch": 0.13726028299504922,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.779670433947608e-05,
+      "loss": 0.5617,
+      "step": 3126
+    },
+    {
+      "epoch": 0.13734810147425272,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.77938636275669e-05,
+      "loss": 0.5592,
+      "step": 3128
+    },
+    {
+      "epoch": 0.1374359199534562,
+      "grad_norm": 2.984375,
+      "learning_rate": 4.779102117009817e-05,
+      "loss": 0.5645,
+      "step": 3130
+    },
+    {
+      "epoch": 0.1375237384326597,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.778817696728758e-05,
+      "loss": 0.5756,
+      "step": 3132
+    },
+    {
+      "epoch": 0.1376115569118632,
+      "grad_norm": 3.09375,
+      "learning_rate": 4.778533101935293e-05,
+      "loss": 0.5531,
+      "step": 3134
+    },
+    {
+      "epoch": 0.13769937539106666,
+      "grad_norm": 2.953125,
+      "learning_rate": 4.778248332651217e-05,
+      "loss": 0.588,
+      "step": 3136
+    },
+    {
+      "epoch": 0.13778719387027016,
+      "grad_norm": 2.96875,
+      "learning_rate": 4.7779633888983375e-05,
+      "loss": 0.5559,
+      "step": 3138
+    },
+    {
+      "epoch": 0.13787501234947364,
+      "grad_norm": 3.015625,
+      "learning_rate": 4.7776782706984754e-05,
+      "loss": 0.5385,
+      "step": 3140
+    },
+    {
+      "epoch": 0.13796283082867714,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.777392978073466e-05,
+      "loss": 0.5593,
+      "step": 3142
+    },
+    {
+      "epoch": 0.1380506493078806,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.777107511045157e-05,
+      "loss": 0.5674,
+      "step": 3144
+    },
+    {
+      "epoch": 0.1381384677870841,
+      "grad_norm": 3.34375,
+      "learning_rate": 4.776821869635407e-05,
+      "loss": 0.5298,
+      "step": 3146
+    },
+    {
+      "epoch": 0.13822628626628758,
+      "grad_norm": 3.34375,
+      "learning_rate": 4.776536053866094e-05,
+      "loss": 0.5588,
+      "step": 3148
+    },
+    {
+      "epoch": 0.13831410474549108,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.7762500637591036e-05,
+      "loss": 0.5656,
+      "step": 3150
+    },
+    {
+      "epoch": 0.13840192322469455,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.775963899336338e-05,
+      "loss": 0.5465,
+      "step": 3152
+    },
+    {
+      "epoch": 0.13848974170389805,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.7756775606197114e-05,
+      "loss": 0.556,
+      "step": 3154
+    },
+    {
+      "epoch": 0.13857756018310152,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.775391047631151e-05,
+      "loss": 0.5616,
+      "step": 3156
+    },
+    {
+      "epoch": 0.13866537866230502,
+      "grad_norm": 3.09375,
+      "learning_rate": 4.7751043603925996e-05,
+      "loss": 0.5721,
+      "step": 3158
+    },
+    {
+      "epoch": 0.1387531971415085,
+      "grad_norm": 3.1875,
+      "learning_rate": 4.77481749892601e-05,
+      "loss": 0.5668,
+      "step": 3160
+    },
+    {
+      "epoch": 0.138841015620712,
+      "grad_norm": 3.0625,
+      "learning_rate": 4.774530463253352e-05,
+      "loss": 0.5333,
+      "step": 3162
+    },
+    {
+      "epoch": 0.13892883409991547,
+      "grad_norm": 2.984375,
+      "learning_rate": 4.774243253396605e-05,
+      "loss": 0.5342,
+      "step": 3164
+    },
+    {
+      "epoch": 0.13901665257911897,
+      "grad_norm": 3.171875,
+      "learning_rate": 4.7739558693777654e-05,
+      "loss": 0.5521,
+      "step": 3166
+    },
+    {
+      "epoch": 0.13910447105832244,
+      "grad_norm": 3.796875,
+      "learning_rate": 4.7736683112188396e-05,
+      "loss": 0.5404,
+      "step": 3168
+    },
+    {
+      "epoch": 0.13919228953752594,
+      "grad_norm": 3.921875,
+      "learning_rate": 4.77338057894185e-05,
+      "loss": 0.5795,
+      "step": 3170
+    },
+    {
+      "epoch": 0.1392801080167294,
+      "grad_norm": 4.75,
+      "learning_rate": 4.773092672568829e-05,
+      "loss": 0.5364,
+      "step": 3172
+    },
+    {
+      "epoch": 0.1393679264959329,
+      "grad_norm": 5.125,
+      "learning_rate": 4.7728045921218286e-05,
+      "loss": 0.5539,
+      "step": 3174
+    },
+    {
+      "epoch": 0.13945574497513638,
+      "grad_norm": 4.59375,
+      "learning_rate": 4.7725163376229064e-05,
+      "loss": 0.5781,
+      "step": 3176
+    },
+    {
+      "epoch": 0.13954356345433988,
+      "grad_norm": 3.5,
+      "learning_rate": 4.772227909094139e-05,
+      "loss": 0.572,
+      "step": 3178
+    },
+    {
+      "epoch": 0.13963138193354335,
+      "grad_norm": 3.59375,
+      "learning_rate": 4.771939306557613e-05,
+      "loss": 0.5433,
+      "step": 3180
+    },
+    {
+      "epoch": 0.13971920041274685,
+      "grad_norm": 2.625,
+      "learning_rate": 4.77165053003543e-05,
+      "loss": 0.5494,
+      "step": 3182
+    },
+    {
+      "epoch": 0.13980701889195035,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.7713615795497055e-05,
+      "loss": 0.5664,
+      "step": 3184
+    },
+    {
+      "epoch": 0.13989483737115382,
+      "grad_norm": 2.984375,
+      "learning_rate": 4.771072455122567e-05,
+      "loss": 0.5408,
+      "step": 3186
+    },
+    {
+      "epoch": 0.13998265585035732,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.770783156776155e-05,
+      "loss": 0.5764,
+      "step": 3188
+    },
+    {
+      "epoch": 0.1400704743295608,
+      "grad_norm": 3.109375,
+      "learning_rate": 4.770493684532624e-05,
+      "loss": 0.5377,
+      "step": 3190
+    },
+    {
+      "epoch": 0.1401582928087643,
+      "grad_norm": 3.078125,
+      "learning_rate": 4.770204038414143e-05,
+      "loss": 0.5654,
+      "step": 3192
+    },
+    {
+      "epoch": 0.14024611128796777,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.769914218442892e-05,
+      "loss": 0.5729,
+      "step": 3194
+    },
+    {
+      "epoch": 0.14033392976717127,
+      "grad_norm": 2.53125,
+      "learning_rate": 4.7696242246410674e-05,
+      "loss": 0.5623,
+      "step": 3196
+    },
+    {
+      "epoch": 0.14042174824637474,
+      "grad_norm": 3.03125,
+      "learning_rate": 4.769334057030874e-05,
+      "loss": 0.5485,
+      "step": 3198
+    },
+    {
+      "epoch": 0.14050956672557824,
+      "grad_norm": 3.140625,
+      "learning_rate": 4.7690437156345356e-05,
+      "loss": 0.5654,
+      "step": 3200
+    },
+    {
+      "epoch": 0.1405973852047817,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.768753200474285e-05,
+      "loss": 0.5322,
+      "step": 3202
+    },
+    {
+      "epoch": 0.1406852036839852,
+      "grad_norm": 2.46875,
+      "learning_rate": 4.768462511572371e-05,
+      "loss": 0.5448,
+      "step": 3204
+    },
+    {
+      "epoch": 0.14077302216318868,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.768171648951054e-05,
+      "loss": 0.5783,
+      "step": 3206
+    },
+    {
+      "epoch": 0.14086084064239218,
+      "grad_norm": 3.25,
+      "learning_rate": 4.767880612632608e-05,
+      "loss": 0.5502,
+      "step": 3208
+    },
+    {
+      "epoch": 0.14094865912159565,
+      "grad_norm": 3.28125,
+      "learning_rate": 4.767589402639321e-05,
+      "loss": 0.5776,
+      "step": 3210
+    },
+    {
+      "epoch": 0.14103647760079915,
+      "grad_norm": 3.09375,
+      "learning_rate": 4.7672980189934935e-05,
+      "loss": 0.5423,
+      "step": 3212
+    },
+    {
+      "epoch": 0.14112429608000263,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.7670064617174414e-05,
+      "loss": 0.5671,
+      "step": 3214
+    },
+    {
+      "epoch": 0.14121211455920613,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.7667147308334906e-05,
+      "loss": 0.5757,
+      "step": 3216
+    },
+    {
+      "epoch": 0.1412999330384096,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.766422826363982e-05,
+      "loss": 0.5452,
+      "step": 3218
+    },
+    {
+      "epoch": 0.1413877515176131,
+      "grad_norm": 2.515625,
+      "learning_rate": 4.76613074833127e-05,
+      "loss": 0.5573,
+      "step": 3220
+    },
+    {
+      "epoch": 0.14147556999681657,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.765838496757722e-05,
+      "loss": 0.5512,
+      "step": 3222
+    },
+    {
+      "epoch": 0.14156338847602007,
+      "grad_norm": 3.390625,
+      "learning_rate": 4.765546071665719e-05,
+      "loss": 0.5689,
+      "step": 3224
+    },
+    {
+      "epoch": 0.14165120695522354,
+      "grad_norm": 3.34375,
+      "learning_rate": 4.765253473077655e-05,
+      "loss": 0.5606,
+      "step": 3226
+    },
+    {
+      "epoch": 0.14173902543442704,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.764960701015937e-05,
+      "loss": 0.5573,
+      "step": 3228
+    },
+    {
+      "epoch": 0.1418268439136305,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.764667755502985e-05,
+      "loss": 0.5643,
+      "step": 3230
+    },
+    {
+      "epoch": 0.141914662392834,
+      "grad_norm": 3.34375,
+      "learning_rate": 4.764374636561234e-05,
+      "loss": 0.5225,
+      "step": 3232
+    },
+    {
+      "epoch": 0.1420024808720375,
+      "grad_norm": 2.953125,
+      "learning_rate": 4.76408134421313e-05,
+      "loss": 0.5291,
+      "step": 3234
+    },
+    {
+      "epoch": 0.14209029935124098,
+      "grad_norm": 2.984375,
+      "learning_rate": 4.7637878784811343e-05,
+      "loss": 0.5567,
+      "step": 3236
+    },
+    {
+      "epoch": 0.14217811783044448,
+      "grad_norm": 2.453125,
+      "learning_rate": 4.76349423938772e-05,
+      "loss": 0.5848,
+      "step": 3238
+    },
+    {
+      "epoch": 0.14226593630964796,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.7632004269553746e-05,
+      "loss": 0.5553,
+      "step": 3240
+    },
+    {
+      "epoch": 0.14235375478885146,
+      "grad_norm": 3.65625,
+      "learning_rate": 4.762906441206597e-05,
+      "loss": 0.5725,
+      "step": 3242
+    },
+    {
+      "epoch": 0.14244157326805493,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.762612282163903e-05,
+      "loss": 0.5571,
+      "step": 3244
+    },
+    {
+      "epoch": 0.14252939174725843,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.762317949849817e-05,
+      "loss": 0.5418,
+      "step": 3246
+    },
+    {
+      "epoch": 0.1426172102264619,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.7620234442868806e-05,
+      "loss": 0.5296,
+      "step": 3248
+    },
+    {
+      "epoch": 0.1427050287056654,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.7617287654976466e-05,
+      "loss": 0.5702,
+      "step": 3250
+    },
+    {
+      "epoch": 0.14279284718486887,
+      "grad_norm": 2.96875,
+      "learning_rate": 4.7614339135046816e-05,
+      "loss": 0.5605,
+      "step": 3252
+    },
+    {
+      "epoch": 0.14288066566407237,
+      "grad_norm": 2.96875,
+      "learning_rate": 4.761138888330565e-05,
+      "loss": 0.546,
+      "step": 3254
+    },
+    {
+      "epoch": 0.14296848414327584,
+      "grad_norm": 2.953125,
+      "learning_rate": 4.760843689997891e-05,
+      "loss": 0.5669,
+      "step": 3256
+    },
+    {
+      "epoch": 0.14305630262247934,
+      "grad_norm": 2.875,
+      "learning_rate": 4.760548318529265e-05,
+      "loss": 0.5651,
+      "step": 3258
+    },
+    {
+      "epoch": 0.14314412110168281,
+      "grad_norm": 2.984375,
+      "learning_rate": 4.760252773947307e-05,
+      "loss": 0.5565,
+      "step": 3260
+    },
+    {
+      "epoch": 0.1432319395808863,
+      "grad_norm": 2.546875,
+      "learning_rate": 4.7599570562746486e-05,
+      "loss": 0.5608,
+      "step": 3262
+    },
+    {
+      "epoch": 0.14331975806008979,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.7596611655339384e-05,
+      "loss": 0.5184,
+      "step": 3264
+    },
+    {
+      "epoch": 0.14340757653929329,
+      "grad_norm": 3.328125,
+      "learning_rate": 4.759365101747833e-05,
+      "loss": 0.5226,
+      "step": 3266
+    },
+    {
+      "epoch": 0.14349539501849676,
+      "grad_norm": 3.046875,
+      "learning_rate": 4.759068864939008e-05,
+      "loss": 0.5637,
+      "step": 3268
+    },
+    {
+      "epoch": 0.14358321349770026,
+      "grad_norm": 3.734375,
+      "learning_rate": 4.7587724551301474e-05,
+      "loss": 0.5445,
+      "step": 3270
+    },
+    {
+      "epoch": 0.14367103197690373,
+      "grad_norm": 3.203125,
+      "learning_rate": 4.758475872343951e-05,
+      "loss": 0.5533,
+      "step": 3272
+    },
+    {
+      "epoch": 0.14375885045610723,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.7581791166031307e-05,
+      "loss": 0.5308,
+      "step": 3274
+    },
+    {
+      "epoch": 0.1438466689353107,
+      "grad_norm": 2.484375,
+      "learning_rate": 4.757882187930412e-05,
+      "loss": 0.5278,
+      "step": 3276
+    },
+    {
+      "epoch": 0.1439344874145142,
+      "grad_norm": 3.15625,
+      "learning_rate": 4.7575850863485345e-05,
+      "loss": 0.5135,
+      "step": 3278
+    },
+    {
+      "epoch": 0.14402230589371767,
+      "grad_norm": 2.953125,
+      "learning_rate": 4.7572878118802496e-05,
+      "loss": 0.5093,
+      "step": 3280
+    },
+    {
+      "epoch": 0.14411012437292117,
+      "grad_norm": 3.484375,
+      "learning_rate": 4.756990364548323e-05,
+      "loss": 0.5549,
+      "step": 3282
+    },
+    {
+      "epoch": 0.14419794285212467,
+      "grad_norm": 3.390625,
+      "learning_rate": 4.7566927443755324e-05,
+      "loss": 0.5789,
+      "step": 3284
+    },
+    {
+      "epoch": 0.14428576133132814,
+      "grad_norm": 2.328125,
+      "learning_rate": 4.756394951384672e-05,
+      "loss": 0.533,
+      "step": 3286
+    },
+    {
+      "epoch": 0.14437357981053164,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.756096985598545e-05,
+      "loss": 0.5552,
+      "step": 3288
+    },
+    {
+      "epoch": 0.14446139828973512,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.7557988470399695e-05,
+      "loss": 0.5436,
+      "step": 3290
+    },
+    {
+      "epoch": 0.14454921676893862,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.7555005357317774e-05,
+      "loss": 0.5551,
+      "step": 3292
+    },
+    {
+      "epoch": 0.1446370352481421,
+      "grad_norm": 3.0,
+      "learning_rate": 4.7552020516968144e-05,
+      "loss": 0.5224,
+      "step": 3294
+    },
+    {
+      "epoch": 0.1447248537273456,
+      "grad_norm": 2.75,
+      "learning_rate": 4.754903394957937e-05,
+      "loss": 0.5368,
+      "step": 3296
+    },
+    {
+      "epoch": 0.14481267220654906,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.7546045655380174e-05,
+      "loss": 0.5495,
+      "step": 3298
+    },
+    {
+      "epoch": 0.14490049068575256,
+      "grad_norm": 2.375,
+      "learning_rate": 4.7543055634599394e-05,
+      "loss": 0.5518,
+      "step": 3300
+    },
+    {
+      "epoch": 0.14498830916495603,
+      "grad_norm": 2.484375,
+      "learning_rate": 4.754006388746601e-05,
+      "loss": 0.5558,
+      "step": 3302
+    },
+    {
+      "epoch": 0.14507612764415953,
+      "grad_norm": 2.40625,
+      "learning_rate": 4.7537070414209134e-05,
+      "loss": 0.5431,
+      "step": 3304
+    },
+    {
+      "epoch": 0.145163946123363,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.7534075215058e-05,
+      "loss": 0.5392,
+      "step": 3306
+    },
+    {
+      "epoch": 0.1452517646025665,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.753107829024198e-05,
+      "loss": 0.5357,
+      "step": 3308
+    },
+    {
+      "epoch": 0.14533958308176997,
+      "grad_norm": 2.46875,
+      "learning_rate": 4.7528079639990596e-05,
+      "loss": 0.5398,
+      "step": 3310
+    },
+    {
+      "epoch": 0.14542740156097347,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.7525079264533464e-05,
+      "loss": 0.5326,
+      "step": 3312
+    },
+    {
+      "epoch": 0.14551522004017695,
+      "grad_norm": 3.125,
+      "learning_rate": 4.752207716410036e-05,
+      "loss": 0.5444,
+      "step": 3314
+    },
+    {
+      "epoch": 0.14560303851938045,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.7519073338921196e-05,
+      "loss": 0.5697,
+      "step": 3316
+    },
+    {
+      "epoch": 0.14569085699858392,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.751606778922599e-05,
+      "loss": 0.551,
+      "step": 3318
+    },
+    {
+      "epoch": 0.14577867547778742,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.751306051524492e-05,
+      "loss": 0.5416,
+      "step": 3320
+    },
+    {
+      "epoch": 0.1458664939569909,
+      "grad_norm": 3.359375,
+      "learning_rate": 4.7510051517208276e-05,
+      "loss": 0.5223,
+      "step": 3322
+    },
+    {
+      "epoch": 0.1459543124361944,
+      "grad_norm": 4.34375,
+      "learning_rate": 4.750704079534649e-05,
+      "loss": 0.5606,
+      "step": 3324
+    },
+    {
+      "epoch": 0.14604213091539786,
+      "grad_norm": 3.921875,
+      "learning_rate": 4.750402834989013e-05,
+      "loss": 0.5694,
+      "step": 3326
+    },
+    {
+      "epoch": 0.14612994939460136,
+      "grad_norm": 3.578125,
+      "learning_rate": 4.7501014181069884e-05,
+      "loss": 0.5552,
+      "step": 3328
+    },
+    {
+      "epoch": 0.14621776787380483,
+      "grad_norm": 3.640625,
+      "learning_rate": 4.749799828911657e-05,
+      "loss": 0.5606,
+      "step": 3330
+    },
+    {
+      "epoch": 0.14630558635300833,
+      "grad_norm": 3.0625,
+      "learning_rate": 4.749498067426116e-05,
+      "loss": 0.5545,
+      "step": 3332
+    },
+    {
+      "epoch": 0.14639340483221183,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.7491961336734735e-05,
+      "loss": 0.5326,
+      "step": 3334
+    },
+    {
+      "epoch": 0.1464812233114153,
+      "grad_norm": 2.546875,
+      "learning_rate": 4.7488940276768525e-05,
+      "loss": 0.5556,
+      "step": 3336
+    },
+    {
+      "epoch": 0.1465690417906188,
+      "grad_norm": 3.328125,
+      "learning_rate": 4.7485917494593866e-05,
+      "loss": 0.5552,
+      "step": 3338
+    },
+    {
+      "epoch": 0.14665686026982228,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.748289299044226e-05,
+      "loss": 0.5554,
+      "step": 3340
+    },
+    {
+      "epoch": 0.14674467874902578,
+      "grad_norm": 3.09375,
+      "learning_rate": 4.747986676454533e-05,
+      "loss": 0.5636,
+      "step": 3342
+    },
+    {
+      "epoch": 0.14683249722822925,
+      "grad_norm": 3.109375,
+      "learning_rate": 4.747683881713481e-05,
+      "loss": 0.5333,
+      "step": 3344
+    },
+    {
+      "epoch": 0.14692031570743275,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.747380914844257e-05,
+      "loss": 0.5331,
+      "step": 3346
+    },
+    {
+      "epoch": 0.14700813418663622,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.7470777758700655e-05,
+      "loss": 0.5181,
+      "step": 3348
+    },
+    {
+      "epoch": 0.14709595266583972,
+      "grad_norm": 3.078125,
+      "learning_rate": 4.746774464814119e-05,
+      "loss": 0.5348,
+      "step": 3350
+    },
+    {
+      "epoch": 0.1471837711450432,
+      "grad_norm": 3.03125,
+      "learning_rate": 4.7464709816996445e-05,
+      "loss": 0.5384,
+      "step": 3352
+    },
+    {
+      "epoch": 0.1472715896242467,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.746167326549884e-05,
+      "loss": 0.5614,
+      "step": 3354
+    },
+    {
+      "epoch": 0.14735940810345016,
+      "grad_norm": 3.078125,
+      "learning_rate": 4.745863499388092e-05,
+      "loss": 0.5327,
+      "step": 3356
+    },
+    {
+      "epoch": 0.14744722658265366,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.7455595002375344e-05,
+      "loss": 0.5381,
+      "step": 3358
+    },
+    {
+      "epoch": 0.14753504506185713,
+      "grad_norm": 3.6875,
+      "learning_rate": 4.745255329121492e-05,
+      "loss": 0.5582,
+      "step": 3360
+    },
+    {
+      "epoch": 0.14762286354106063,
+      "grad_norm": 4.28125,
+      "learning_rate": 4.744950986063258e-05,
+      "loss": 0.5844,
+      "step": 3362
+    },
+    {
+      "epoch": 0.1477106820202641,
+      "grad_norm": 3.390625,
+      "learning_rate": 4.744646471086139e-05,
+      "loss": 0.525,
+      "step": 3364
+    },
+    {
+      "epoch": 0.1477985004994676,
+      "grad_norm": 3.1875,
+      "learning_rate": 4.744341784213456e-05,
+      "loss": 0.5444,
+      "step": 3366
+    },
+    {
+      "epoch": 0.14788631897867108,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.744036925468541e-05,
+      "loss": 0.5375,
+      "step": 3368
+    },
+    {
+      "epoch": 0.14797413745787458,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.74373189487474e-05,
+      "loss": 0.5399,
+      "step": 3370
+    },
+    {
+      "epoch": 0.14806195593707805,
+      "grad_norm": 3.4375,
+      "learning_rate": 4.743426692455413e-05,
+      "loss": 0.5581,
+      "step": 3372
+    },
+    {
+      "epoch": 0.14814977441628155,
+      "grad_norm": 3.71875,
+      "learning_rate": 4.7431213182339315e-05,
+      "loss": 0.5768,
+      "step": 3374
+    },
+    {
+      "epoch": 0.14823759289548502,
+      "grad_norm": 3.375,
+      "learning_rate": 4.742815772233682e-05,
+      "loss": 0.5311,
+      "step": 3376
+    },
+    {
+      "epoch": 0.14832541137468852,
+      "grad_norm": 3.25,
+      "learning_rate": 4.742510054478063e-05,
+      "loss": 0.539,
+      "step": 3378
+    },
+    {
+      "epoch": 0.148413229853892,
+      "grad_norm": 3.375,
+      "learning_rate": 4.7422041649904867e-05,
+      "loss": 0.5256,
+      "step": 3380
+    },
+    {
+      "epoch": 0.1485010483330955,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.7418981037943785e-05,
+      "loss": 0.5492,
+      "step": 3382
+    },
+    {
+      "epoch": 0.14858886681229896,
+      "grad_norm": 3.34375,
+      "learning_rate": 4.741591870913175e-05,
+      "loss": 0.5769,
+      "step": 3384
+    },
+    {
+      "epoch": 0.14867668529150246,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.741285466370329e-05,
+      "loss": 0.5478,
+      "step": 3386
+    },
+    {
+      "epoch": 0.14876450377070596,
+      "grad_norm": 3.046875,
+      "learning_rate": 4.740978890189305e-05,
+      "loss": 0.5265,
+      "step": 3388
+    },
+    {
+      "epoch": 0.14885232224990944,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.740672142393581e-05,
+      "loss": 0.5097,
+      "step": 3390
+    },
+    {
+      "epoch": 0.14894014072911294,
+      "grad_norm": 2.46875,
+      "learning_rate": 4.740365223006646e-05,
+      "loss": 0.5272,
+      "step": 3392
+    },
+    {
+      "epoch": 0.1490279592083164,
+      "grad_norm": 3.0,
+      "learning_rate": 4.7400581320520055e-05,
+      "loss": 0.5462,
+      "step": 3394
+    },
+    {
+      "epoch": 0.1491157776875199,
+      "grad_norm": 2.421875,
+      "learning_rate": 4.7397508695531764e-05,
+      "loss": 0.5324,
+      "step": 3396
+    },
+    {
+      "epoch": 0.14920359616672338,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.739443435533689e-05,
+      "loss": 0.5603,
+      "step": 3398
+    },
+    {
+      "epoch": 0.14929141464592688,
+      "grad_norm": 2.515625,
+      "learning_rate": 4.7391358300170865e-05,
+      "loss": 0.5247,
+      "step": 3400
+    },
+    {
+      "epoch": 0.14937923312513035,
+      "grad_norm": 3.046875,
+      "learning_rate": 4.738828053026925e-05,
+      "loss": 0.5094,
+      "step": 3402
+    },
+    {
+      "epoch": 0.14946705160433385,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.7385201045867747e-05,
+      "loss": 0.5529,
+      "step": 3404
+    },
+    {
+      "epoch": 0.14955487008353732,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.738211984720218e-05,
+      "loss": 0.5386,
+      "step": 3406
+    },
+    {
+      "epoch": 0.14964268856274082,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.7379036934508506e-05,
+      "loss": 0.5707,
+      "step": 3408
+    },
+    {
+      "epoch": 0.1497305070419443,
+      "grad_norm": 2.75,
+      "learning_rate": 4.7375952308022824e-05,
+      "loss": 0.5598,
+      "step": 3410
+    },
+    {
+      "epoch": 0.1498183255211478,
+      "grad_norm": 3.65625,
+      "learning_rate": 4.737286596798135e-05,
+      "loss": 0.5739,
+      "step": 3412
+    },
+    {
+      "epoch": 0.14990614400035127,
+      "grad_norm": 3.734375,
+      "learning_rate": 4.7369777914620436e-05,
+      "loss": 0.5426,
+      "step": 3414
+    },
+    {
+      "epoch": 0.14999396247955477,
+      "grad_norm": 2.625,
+      "learning_rate": 4.736668814817657e-05,
+      "loss": 0.5247,
+      "step": 3416
+    },
+    {
+      "epoch": 0.15008178095875824,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.7363596668886364e-05,
+      "loss": 0.5513,
+      "step": 3418
+    },
+    {
+      "epoch": 0.15016959943796174,
+      "grad_norm": 2.96875,
+      "learning_rate": 4.736050347698656e-05,
+      "loss": 0.5728,
+      "step": 3420
+    },
+    {
+      "epoch": 0.1502574179171652,
+      "grad_norm": 3.0625,
+      "learning_rate": 4.7357408572714046e-05,
+      "loss": 0.554,
+      "step": 3422
+    },
+    {
+      "epoch": 0.1503452363963687,
+      "grad_norm": 2.96875,
+      "learning_rate": 4.735431195630582e-05,
+      "loss": 0.5098,
+      "step": 3424
+    },
+    {
+      "epoch": 0.15043305487557218,
+      "grad_norm": 2.390625,
+      "learning_rate": 4.7351213627999027e-05,
+      "loss": 0.5609,
+      "step": 3426
+    },
+    {
+      "epoch": 0.15052087335477568,
+      "grad_norm": 3.0,
+      "learning_rate": 4.734811358803093e-05,
+      "loss": 0.5503,
+      "step": 3428
+    },
+    {
+      "epoch": 0.15060869183397915,
+      "grad_norm": 3.046875,
+      "learning_rate": 4.734501183663894e-05,
+      "loss": 0.5682,
+      "step": 3430
+    },
+    {
+      "epoch": 0.15069651031318265,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.7341908374060595e-05,
+      "loss": 0.5374,
+      "step": 3432
+    },
+    {
+      "epoch": 0.15078432879238612,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.733880320053354e-05,
+      "loss": 0.5598,
+      "step": 3434
+    },
+    {
+      "epoch": 0.15087214727158962,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.733569631629559e-05,
+      "loss": 0.5408,
+      "step": 3436
+    },
+    {
+      "epoch": 0.15095996575079312,
+      "grad_norm": 2.875,
+      "learning_rate": 4.7332587721584656e-05,
+      "loss": 0.5214,
+      "step": 3438
+    },
+    {
+      "epoch": 0.1510477842299966,
+      "grad_norm": 2.625,
+      "learning_rate": 4.732947741663881e-05,
+      "loss": 0.5384,
+      "step": 3440
+    },
+    {
+      "epoch": 0.1511356027092001,
+      "grad_norm": 2.515625,
+      "learning_rate": 4.732636540169621e-05,
+      "loss": 0.5455,
+      "step": 3442
+    },
+    {
+      "epoch": 0.15122342118840357,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.732325167699522e-05,
+      "loss": 0.5517,
+      "step": 3444
+    },
+    {
+      "epoch": 0.15131123966760707,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.732013624277425e-05,
+      "loss": 0.5336,
+      "step": 3446
+    },
+    {
+      "epoch": 0.15139905814681054,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.73170190992719e-05,
+      "loss": 0.5347,
+      "step": 3448
+    },
+    {
+      "epoch": 0.15148687662601404,
+      "grad_norm": 2.5,
+      "learning_rate": 4.731390024672688e-05,
+      "loss": 0.5212,
+      "step": 3450
+    },
+    {
+      "epoch": 0.1515746951052175,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.731077968537803e-05,
+      "loss": 0.5469,
+      "step": 3452
+    },
+    {
+      "epoch": 0.151662513584421,
+      "grad_norm": 3.234375,
+      "learning_rate": 4.7307657415464324e-05,
+      "loss": 0.559,
+      "step": 3454
+    },
+    {
+      "epoch": 0.15175033206362448,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.7304533437224866e-05,
+      "loss": 0.5247,
+      "step": 3456
+    },
+    {
+      "epoch": 0.15183815054282798,
+      "grad_norm": 3.765625,
+      "learning_rate": 4.730140775089888e-05,
+      "loss": 0.5356,
+      "step": 3458
+    },
+    {
+      "epoch": 0.15192596902203145,
+      "grad_norm": 3.59375,
+      "learning_rate": 4.729828035672576e-05,
+      "loss": 0.5345,
+      "step": 3460
+    },
+    {
+      "epoch": 0.15201378750123495,
+      "grad_norm": 3.578125,
+      "learning_rate": 4.729515125494497e-05,
+      "loss": 0.5386,
+      "step": 3462
+    },
+    {
+      "epoch": 0.15210160598043843,
+      "grad_norm": 5.0625,
+      "learning_rate": 4.729202044579616e-05,
+      "loss": 0.531,
+      "step": 3464
+    },
+    {
+      "epoch": 0.15218942445964193,
+      "grad_norm": 7.71875,
+      "learning_rate": 4.7288887929519074e-05,
+      "loss": 0.5581,
+      "step": 3466
+    },
+    {
+      "epoch": 0.1522772429388454,
+      "grad_norm": 3.484375,
+      "learning_rate": 4.7285753706353614e-05,
+      "loss": 0.5721,
+      "step": 3468
+    },
+    {
+      "epoch": 0.1523650614180489,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.728261777653979e-05,
+      "loss": 0.5701,
+      "step": 3470
+    },
+    {
+      "epoch": 0.15245287989725237,
+      "grad_norm": 4.1875,
+      "learning_rate": 4.7279480140317756e-05,
+      "loss": 0.5465,
+      "step": 3472
+    },
+    {
+      "epoch": 0.15254069837645587,
+      "grad_norm": 3.90625,
+      "learning_rate": 4.727634079792779e-05,
+      "loss": 0.5398,
+      "step": 3474
+    },
+    {
+      "epoch": 0.15262851685565934,
+      "grad_norm": 3.3125,
+      "learning_rate": 4.727319974961031e-05,
+      "loss": 0.5536,
+      "step": 3476
+    },
+    {
+      "epoch": 0.15271633533486284,
+      "grad_norm": 3.71875,
+      "learning_rate": 4.7270056995605846e-05,
+      "loss": 0.5496,
+      "step": 3478
+    },
+    {
+      "epoch": 0.1528041538140663,
+      "grad_norm": 3.65625,
+      "learning_rate": 4.726691253615509e-05,
+      "loss": 0.5222,
+      "step": 3480
+    },
+    {
+      "epoch": 0.1528919722932698,
+      "grad_norm": 3.25,
+      "learning_rate": 4.726376637149883e-05,
+      "loss": 0.5456,
+      "step": 3482
+    },
+    {
+      "epoch": 0.15297979077247328,
+      "grad_norm": 3.078125,
+      "learning_rate": 4.7260618501877994e-05,
+      "loss": 0.5076,
+      "step": 3484
+    },
+    {
+      "epoch": 0.15306760925167678,
+      "grad_norm": 3.3125,
+      "learning_rate": 4.725746892753367e-05,
+      "loss": 0.5468,
+      "step": 3486
+    },
+    {
+      "epoch": 0.15315542773088028,
+      "grad_norm": 3.171875,
+      "learning_rate": 4.725431764870704e-05,
+      "loss": 0.5179,
+      "step": 3488
+    },
+    {
+      "epoch": 0.15324324621008376,
+      "grad_norm": 3.1875,
+      "learning_rate": 4.7251164665639426e-05,
+      "loss": 0.5522,
+      "step": 3490
+    },
+    {
+      "epoch": 0.15333106468928726,
+      "grad_norm": 3.953125,
+      "learning_rate": 4.724800997857228e-05,
+      "loss": 0.5371,
+      "step": 3492
+    },
+    {
+      "epoch": 0.15341888316849073,
+      "grad_norm": 3.796875,
+      "learning_rate": 4.724485358774721e-05,
+      "loss": 0.5303,
+      "step": 3494
+    },
+    {
+      "epoch": 0.15350670164769423,
+      "grad_norm": 3.59375,
+      "learning_rate": 4.724169549340591e-05,
+      "loss": 0.5051,
+      "step": 3496
+    },
+    {
+      "epoch": 0.1535945201268977,
+      "grad_norm": 3.59375,
+      "learning_rate": 4.723853569579024e-05,
+      "loss": 0.5499,
+      "step": 3498
+    },
+    {
+      "epoch": 0.1536823386061012,
+      "grad_norm": 3.265625,
+      "learning_rate": 4.723537419514218e-05,
+      "loss": 0.5448,
+      "step": 3500
+    },
+    {
+      "epoch": 0.15377015708530467,
+      "grad_norm": 3.953125,
+      "learning_rate": 4.723221099170383e-05,
+      "loss": 0.5605,
+      "step": 3502
+    },
+    {
+      "epoch": 0.15385797556450817,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.7229046085717434e-05,
+      "loss": 0.5353,
+      "step": 3504
+    },
+    {
+      "epoch": 0.15394579404371164,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.7225879477425364e-05,
+      "loss": 0.5276,
+      "step": 3506
+    },
+    {
+      "epoch": 0.15403361252291514,
+      "grad_norm": 2.546875,
+      "learning_rate": 4.722271116707011e-05,
+      "loss": 0.5488,
+      "step": 3508
+    },
+    {
+      "epoch": 0.15412143100211861,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.72195411548943e-05,
+      "loss": 0.5281,
+      "step": 3510
+    },
+    {
+      "epoch": 0.1542092494813221,
+      "grad_norm": 3.203125,
+      "learning_rate": 4.7216369441140715e-05,
+      "loss": 0.5644,
+      "step": 3512
+    },
+    {
+      "epoch": 0.15429706796052559,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.721319602605223e-05,
+      "loss": 0.503,
+      "step": 3514
+    },
+    {
+      "epoch": 0.15438488643972909,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.721002090987187e-05,
+      "loss": 0.5472,
+      "step": 3516
+    },
+    {
+      "epoch": 0.15447270491893256,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.720684409284277e-05,
+      "loss": 0.5533,
+      "step": 3518
+    },
+    {
+      "epoch": 0.15456052339813606,
+      "grad_norm": 3.125,
+      "learning_rate": 4.7203665575208244e-05,
+      "loss": 0.5354,
+      "step": 3520
+    },
+    {
+      "epoch": 0.15464834187733953,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.720048535721168e-05,
+      "loss": 0.5225,
+      "step": 3522
+    },
+    {
+      "epoch": 0.15473616035654303,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.7197303439096626e-05,
+      "loss": 0.5436,
+      "step": 3524
+    },
+    {
+      "epoch": 0.1548239788357465,
+      "grad_norm": 3.21875,
+      "learning_rate": 4.7194119821106754e-05,
+      "loss": 0.5509,
+      "step": 3526
+    },
+    {
+      "epoch": 0.15491179731495,
+      "grad_norm": 3.28125,
+      "learning_rate": 4.719093450348586e-05,
+      "loss": 0.5602,
+      "step": 3528
+    },
+    {
+      "epoch": 0.15499961579415347,
+      "grad_norm": 3.09375,
+      "learning_rate": 4.718774748647789e-05,
+      "loss": 0.5344,
+      "step": 3530
+    },
+    {
+      "epoch": 0.15508743427335697,
+      "grad_norm": 3.1875,
+      "learning_rate": 4.718455877032689e-05,
+      "loss": 0.5435,
+      "step": 3532
+    },
+    {
+      "epoch": 0.15517525275256044,
+      "grad_norm": 3.0625,
+      "learning_rate": 4.718136835527707e-05,
+      "loss": 0.5291,
+      "step": 3534
+    },
+    {
+      "epoch": 0.15526307123176394,
+      "grad_norm": 2.953125,
+      "learning_rate": 4.7178176241572735e-05,
+      "loss": 0.5308,
+      "step": 3536
+    },
+    {
+      "epoch": 0.15535088971096744,
+      "grad_norm": 3.0,
+      "learning_rate": 4.717498242945836e-05,
+      "loss": 0.5524,
+      "step": 3538
+    },
+    {
+      "epoch": 0.15543870819017092,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.717178691917851e-05,
+      "loss": 0.5144,
+      "step": 3540
+    },
+    {
+      "epoch": 0.15552652666937442,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.71685897109779e-05,
+      "loss": 0.566,
+      "step": 3542
+    },
+    {
+      "epoch": 0.1556143451485779,
+      "grad_norm": 3.578125,
+      "learning_rate": 4.716539080510137e-05,
+      "loss": 0.5421,
+      "step": 3544
+    },
+    {
+      "epoch": 0.1557021636277814,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.7162190201793904e-05,
+      "loss": 0.5453,
+      "step": 3546
+    },
+    {
+      "epoch": 0.15578998210698486,
+      "grad_norm": 3.09375,
+      "learning_rate": 4.71589879013006e-05,
+      "loss": 0.5369,
+      "step": 3548
+    },
+    {
+      "epoch": 0.15587780058618836,
+      "grad_norm": 3.0,
+      "learning_rate": 4.715578390386669e-05,
+      "loss": 0.5164,
+      "step": 3550
+    },
+    {
+      "epoch": 0.15596561906539183,
+      "grad_norm": 2.46875,
+      "learning_rate": 4.715257820973754e-05,
+      "loss": 0.5128,
+      "step": 3552
+    },
+    {
+      "epoch": 0.15605343754459533,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.7149370819158635e-05,
+      "loss": 0.5106,
+      "step": 3554
+    },
+    {
+      "epoch": 0.1561412560237988,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.714616173237561e-05,
+      "loss": 0.5405,
+      "step": 3556
+    },
+    {
+      "epoch": 0.1562290745030023,
+      "grad_norm": 2.515625,
+      "learning_rate": 4.7142950949634206e-05,
+      "loss": 0.5533,
+      "step": 3558
+    },
+    {
+      "epoch": 0.15631689298220577,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.7139738471180314e-05,
+      "loss": 0.5214,
+      "step": 3560
+    },
+    {
+      "epoch": 0.15640471146140927,
+      "grad_norm": 2.875,
+      "learning_rate": 4.713652429725994e-05,
+      "loss": 0.5371,
+      "step": 3562
+    },
+    {
+      "epoch": 0.15649252994061275,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.713330842811923e-05,
+      "loss": 0.5322,
+      "step": 3564
+    },
+    {
+      "epoch": 0.15658034841981625,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.713009086400445e-05,
+      "loss": 0.5006,
+      "step": 3566
+    },
+    {
+      "epoch": 0.15666816689901972,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.712687160516202e-05,
+      "loss": 0.5391,
+      "step": 3568
+    },
+    {
+      "epoch": 0.15675598537822322,
+      "grad_norm": 2.515625,
+      "learning_rate": 4.712365065183844e-05,
+      "loss": 0.5302,
+      "step": 3570
+    },
+    {
+      "epoch": 0.1568438038574267,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.71204280042804e-05,
+      "loss": 0.5188,
+      "step": 3572
+    },
+    {
+      "epoch": 0.1569316223366302,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.711720366273468e-05,
+      "loss": 0.5314,
+      "step": 3574
+    },
+    {
+      "epoch": 0.15701944081583366,
+      "grad_norm": 2.46875,
+      "learning_rate": 4.71139776274482e-05,
+      "loss": 0.5141,
+      "step": 3576
+    },
+    {
+      "epoch": 0.15710725929503716,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.711074989866802e-05,
+      "loss": 0.5583,
+      "step": 3578
+    },
+    {
+      "epoch": 0.15719507777424063,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.71075204766413e-05,
+      "loss": 0.5613,
+      "step": 3580
+    },
+    {
+      "epoch": 0.15728289625344413,
+      "grad_norm": 2.53125,
+      "learning_rate": 4.710428936161537e-05,
+      "loss": 0.5293,
+      "step": 3582
+    },
+    {
+      "epoch": 0.1573707147326476,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.7101056553837665e-05,
+      "loss": 0.5258,
+      "step": 3584
+    },
+    {
+      "epoch": 0.1574585332118511,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.709782205355574e-05,
+      "loss": 0.5555,
+      "step": 3586
+    },
+    {
+      "epoch": 0.1575463516910546,
+      "grad_norm": 4.46875,
+      "learning_rate": 4.709458586101731e-05,
+      "loss": 0.5505,
+      "step": 3588
+    },
+    {
+      "epoch": 0.15763417017025808,
+      "grad_norm": 4.09375,
+      "learning_rate": 4.70913479764702e-05,
+      "loss": 0.5502,
+      "step": 3590
+    },
+    {
+      "epoch": 0.15772198864946158,
+      "grad_norm": 3.28125,
+      "learning_rate": 4.708810840016237e-05,
+      "loss": 0.5308,
+      "step": 3592
+    },
+    {
+      "epoch": 0.15780980712866505,
+      "grad_norm": 3.25,
+      "learning_rate": 4.7084867132341895e-05,
+      "loss": 0.5257,
+      "step": 3594
+    },
+    {
+      "epoch": 0.15789762560786855,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.708162417325701e-05,
+      "loss": 0.5202,
+      "step": 3596
+    },
+    {
+      "epoch": 0.15798544408707202,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.7078379523156045e-05,
+      "loss": 0.5356,
+      "step": 3598
+    },
+    {
+      "epoch": 0.15807326256627552,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.707513318228749e-05,
+      "loss": 0.5395,
+      "step": 3600
+    },
+    {
+      "epoch": 0.158161081045479,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.707188515089994e-05,
+      "loss": 0.5471,
+      "step": 3602
+    },
+    {
+      "epoch": 0.1582488995246825,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.706863542924213e-05,
+      "loss": 0.544,
+      "step": 3604
+    },
+    {
+      "epoch": 0.15833671800388596,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.706538401756294e-05,
+      "loss": 0.5097,
+      "step": 3606
+    },
+    {
+      "epoch": 0.15842453648308946,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.7062130916111344e-05,
+      "loss": 0.5489,
+      "step": 3608
+    },
+    {
+      "epoch": 0.15851235496229293,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.705887612513647e-05,
+      "loss": 0.5344,
+      "step": 3610
+    },
+    {
+      "epoch": 0.15860017344149643,
+      "grad_norm": 3.359375,
+      "learning_rate": 4.705561964488758e-05,
+      "loss": 0.5395,
+      "step": 3612
+    },
+    {
+      "epoch": 0.1586879919206999,
+      "grad_norm": 2.75,
+      "learning_rate": 4.705236147561405e-05,
+      "loss": 0.5617,
+      "step": 3614
+    },
+    {
+      "epoch": 0.1587758103999034,
+      "grad_norm": 2.40625,
+      "learning_rate": 4.70491016175654e-05,
+      "loss": 0.5352,
+      "step": 3616
+    },
+    {
+      "epoch": 0.15886362887910688,
+      "grad_norm": 2.421875,
+      "learning_rate": 4.704584007099125e-05,
+      "loss": 0.5261,
+      "step": 3618
+    },
+    {
+      "epoch": 0.15895144735831038,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.7042576836141395e-05,
+      "loss": 0.5221,
+      "step": 3620
+    },
+    {
+      "epoch": 0.15903926583751385,
+      "grad_norm": 2.4375,
+      "learning_rate": 4.703931191326572e-05,
+      "loss": 0.5078,
+      "step": 3622
+    },
+    {
+      "epoch": 0.15912708431671735,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.703604530261424e-05,
+      "loss": 0.5281,
+      "step": 3624
+    },
+    {
+      "epoch": 0.15921490279592082,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.7032777004437136e-05,
+      "loss": 0.532,
+      "step": 3626
+    },
+    {
+      "epoch": 0.15930272127512432,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.70295070189847e-05,
+      "loss": 0.5323,
+      "step": 3628
+    },
+    {
+      "epoch": 0.1593905397543278,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.702623534650732e-05,
+      "loss": 0.5521,
+      "step": 3630
+    },
+    {
+      "epoch": 0.1594783582335313,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.7022961987255566e-05,
+      "loss": 0.5377,
+      "step": 3632
+    },
+    {
+      "epoch": 0.15956617671273476,
+      "grad_norm": 3.515625,
+      "learning_rate": 4.70196869414801e-05,
+      "loss": 0.5471,
+      "step": 3634
+    },
+    {
+      "epoch": 0.15965399519193826,
+      "grad_norm": 3.609375,
+      "learning_rate": 4.701641020943173e-05,
+      "loss": 0.5355,
+      "step": 3636
+    },
+    {
+      "epoch": 0.15974181367114176,
+      "grad_norm": 3.546875,
+      "learning_rate": 4.7013131791361385e-05,
+      "loss": 0.5412,
+      "step": 3638
+    },
+    {
+      "epoch": 0.15982963215034524,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.7009851687520134e-05,
+      "loss": 0.5286,
+      "step": 3640
+    },
+    {
+      "epoch": 0.15991745062954874,
+      "grad_norm": 3.234375,
+      "learning_rate": 4.7006569898159165e-05,
+      "loss": 0.5662,
+      "step": 3642
+    },
+    {
+      "epoch": 0.1600052691087522,
+      "grad_norm": 3.0,
+      "learning_rate": 4.7003286423529795e-05,
+      "loss": 0.5374,
+      "step": 3644
+    },
+    {
+      "epoch": 0.1600930875879557,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.700000126388348e-05,
+      "loss": 0.535,
+      "step": 3646
+    },
+    {
+      "epoch": 0.16018090606715918,
+      "grad_norm": 2.515625,
+      "learning_rate": 4.69967144194718e-05,
+      "loss": 0.5094,
+      "step": 3648
+    },
+    {
+      "epoch": 0.16026872454636268,
+      "grad_norm": 2.453125,
+      "learning_rate": 4.6993425890546444e-05,
+      "loss": 0.5443,
+      "step": 3650
+    },
+    {
+      "epoch": 0.16035654302556615,
+      "grad_norm": 2.21875,
+      "learning_rate": 4.699013567735927e-05,
+      "loss": 0.5244,
+      "step": 3652
+    },
+    {
+      "epoch": 0.16044436150476965,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.698684378016222e-05,
+      "loss": 0.5368,
+      "step": 3654
+    },
+    {
+      "epoch": 0.16053217998397312,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.698355019920743e-05,
+      "loss": 0.5424,
+      "step": 3656
+    },
+    {
+      "epoch": 0.16061999846317662,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.698025493474707e-05,
+      "loss": 0.5187,
+      "step": 3658
+    },
+    {
+      "epoch": 0.1607078169423801,
+      "grad_norm": 3.609375,
+      "learning_rate": 4.697695798703353e-05,
+      "loss": 0.5546,
+      "step": 3660
+    },
+    {
+      "epoch": 0.1607956354215836,
+      "grad_norm": 3.09375,
+      "learning_rate": 4.697365935631928e-05,
+      "loss": 0.5323,
+      "step": 3662
+    },
+    {
+      "epoch": 0.16088345390078707,
+      "grad_norm": 2.3125,
+      "learning_rate": 4.697035904285693e-05,
+      "loss": 0.5484,
+      "step": 3664
+    },
+    {
+      "epoch": 0.16097127237999057,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.696705704689921e-05,
+      "loss": 0.524,
+      "step": 3666
+    },
+    {
+      "epoch": 0.16105909085919404,
+      "grad_norm": 3.34375,
+      "learning_rate": 4.6963753368699e-05,
+      "loss": 0.5624,
+      "step": 3668
+    },
+    {
+      "epoch": 0.16114690933839754,
+      "grad_norm": 2.53125,
+      "learning_rate": 4.6960448008509296e-05,
+      "loss": 0.5487,
+      "step": 3670
+    },
+    {
+      "epoch": 0.161234727817601,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.695714096658321e-05,
+      "loss": 0.5272,
+      "step": 3672
+    },
+    {
+      "epoch": 0.1613225462968045,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.695383224317401e-05,
+      "loss": 0.5231,
+      "step": 3674
+    },
+    {
+      "epoch": 0.16141036477600798,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.695052183853508e-05,
+      "loss": 0.5264,
+      "step": 3676
+    },
+    {
+      "epoch": 0.16149818325521148,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.694720975291991e-05,
+      "loss": 0.5296,
+      "step": 3678
+    },
+    {
+      "epoch": 0.16158600173441495,
+      "grad_norm": 3.140625,
+      "learning_rate": 4.694389598658217e-05,
+      "loss": 0.533,
+      "step": 3680
+    },
+    {
+      "epoch": 0.16167382021361845,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.6940580539775616e-05,
+      "loss": 0.4755,
+      "step": 3682
+    },
+    {
+      "epoch": 0.16176163869282192,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.6937263412754135e-05,
+      "loss": 0.5346,
+      "step": 3684
+    },
+    {
+      "epoch": 0.16184945717202542,
+      "grad_norm": 3.09375,
+      "learning_rate": 4.693394460577177e-05,
+      "loss": 0.5129,
+      "step": 3686
+    },
+    {
+      "epoch": 0.16193727565122892,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.693062411908267e-05,
+      "loss": 0.5131,
+      "step": 3688
+    },
+    {
+      "epoch": 0.1620250941304324,
+      "grad_norm": 2.5,
+      "learning_rate": 4.6927301952941105e-05,
+      "loss": 0.5093,
+      "step": 3690
+    },
+    {
+      "epoch": 0.1621129126096359,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.6923978107601516e-05,
+      "loss": 0.5425,
+      "step": 3692
+    },
+    {
+      "epoch": 0.16220073108883937,
+      "grad_norm": 2.5,
+      "learning_rate": 4.692065258331842e-05,
+      "loss": 0.5489,
+      "step": 3694
+    },
+    {
+      "epoch": 0.16228854956804287,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.69173253803465e-05,
+      "loss": 0.5584,
+      "step": 3696
+    },
+    {
+      "epoch": 0.16237636804724634,
+      "grad_norm": 3.34375,
+      "learning_rate": 4.691399649894054e-05,
+      "loss": 0.5217,
+      "step": 3698
+    },
+    {
+      "epoch": 0.16246418652644984,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.691066593935548e-05,
+      "loss": 0.5334,
+      "step": 3700
+    },
+    {
+      "epoch": 0.1625520050056533,
+      "grad_norm": 3.140625,
+      "learning_rate": 4.6907333701846365e-05,
+      "loss": 0.5613,
+      "step": 3702
+    },
+    {
+      "epoch": 0.1626398234848568,
+      "grad_norm": 4.25,
+      "learning_rate": 4.690399978666839e-05,
+      "loss": 0.5262,
+      "step": 3704
+    },
+    {
+      "epoch": 0.16272764196406028,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.690066419407686e-05,
+      "loss": 0.5465,
+      "step": 3706
+    },
+    {
+      "epoch": 0.16281546044326378,
+      "grad_norm": 3.25,
+      "learning_rate": 4.689732692432722e-05,
+      "loss": 0.511,
+      "step": 3708
+    },
+    {
+      "epoch": 0.16290327892246725,
+      "grad_norm": 3.109375,
+      "learning_rate": 4.6893987977675026e-05,
+      "loss": 0.5364,
+      "step": 3710
+    },
+    {
+      "epoch": 0.16299109740167075,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.689064735437599e-05,
+      "loss": 0.5153,
+      "step": 3712
+    },
+    {
+      "epoch": 0.16307891588087423,
+      "grad_norm": 2.515625,
+      "learning_rate": 4.688730505468593e-05,
+      "loss": 0.5287,
+      "step": 3714
+    },
+    {
+      "epoch": 0.16316673436007773,
+      "grad_norm": 2.625,
+      "learning_rate": 4.688396107886081e-05,
+      "loss": 0.5298,
+      "step": 3716
+    },
+    {
+      "epoch": 0.1632545528392812,
+      "grad_norm": 2.75,
+      "learning_rate": 4.688061542715669e-05,
+      "loss": 0.5106,
+      "step": 3718
+    },
+    {
+      "epoch": 0.1633423713184847,
+      "grad_norm": 2.375,
+      "learning_rate": 4.6877268099829804e-05,
+      "loss": 0.555,
+      "step": 3720
+    },
+    {
+      "epoch": 0.16343018979768817,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.687391909713648e-05,
+      "loss": 0.5382,
+      "step": 3722
+    },
+    {
+      "epoch": 0.16351800827689167,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.6870568419333185e-05,
+      "loss": 0.5163,
+      "step": 3724
+    },
+    {
+      "epoch": 0.16360582675609514,
+      "grad_norm": 2.4375,
+      "learning_rate": 4.6867216066676524e-05,
+      "loss": 0.5185,
+      "step": 3726
+    },
+    {
+      "epoch": 0.16369364523529864,
+      "grad_norm": 2.515625,
+      "learning_rate": 4.686386203942321e-05,
+      "loss": 0.5344,
+      "step": 3728
+    },
+    {
+      "epoch": 0.1637814637145021,
+      "grad_norm": 2.953125,
+      "learning_rate": 4.6860506337830105e-05,
+      "loss": 0.5251,
+      "step": 3730
+    },
+    {
+      "epoch": 0.1638692821937056,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.6857148962154185e-05,
+      "loss": 0.4998,
+      "step": 3732
+    },
+    {
+      "epoch": 0.16395710067290908,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.6853789912652554e-05,
+      "loss": 0.5155,
+      "step": 3734
+    },
+    {
+      "epoch": 0.16404491915211258,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.6850429189582454e-05,
+      "loss": 0.5498,
+      "step": 3736
+    },
+    {
+      "epoch": 0.16413273763131608,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.684706679320125e-05,
+      "loss": 0.4993,
+      "step": 3738
+    },
+    {
+      "epoch": 0.16422055611051956,
+      "grad_norm": 2.46875,
+      "learning_rate": 4.684370272376643e-05,
+      "loss": 0.543,
+      "step": 3740
+    },
+    {
+      "epoch": 0.16430837458972306,
+      "grad_norm": 2.40625,
+      "learning_rate": 4.684033698153562e-05,
+      "loss": 0.5161,
+      "step": 3742
+    },
+    {
+      "epoch": 0.16439619306892653,
+      "grad_norm": 3.140625,
+      "learning_rate": 4.683696956676657e-05,
+      "loss": 0.5639,
+      "step": 3744
+    },
+    {
+      "epoch": 0.16448401154813003,
+      "grad_norm": 2.53125,
+      "learning_rate": 4.6833600479717155e-05,
+      "loss": 0.5518,
+      "step": 3746
+    },
+    {
+      "epoch": 0.1645718300273335,
+      "grad_norm": 2.625,
+      "learning_rate": 4.683022972064538e-05,
+      "loss": 0.5062,
+      "step": 3748
+    },
+    {
+      "epoch": 0.164659648506537,
+      "grad_norm": 2.203125,
+      "learning_rate": 4.682685728980939e-05,
+      "loss": 0.537,
+      "step": 3750
+    },
+    {
+      "epoch": 0.16474746698574047,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.682348318746742e-05,
+      "loss": 0.5347,
+      "step": 3752
+    },
+    {
+      "epoch": 0.16483528546494397,
+      "grad_norm": 2.75,
+      "learning_rate": 4.682010741387789e-05,
+      "loss": 0.514,
+      "step": 3754
+    },
+    {
+      "epoch": 0.16492310394414744,
+      "grad_norm": 3.3125,
+      "learning_rate": 4.6816729969299295e-05,
+      "loss": 0.5405,
+      "step": 3756
+    },
+    {
+      "epoch": 0.16501092242335094,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.681335085399029e-05,
+      "loss": 0.5304,
+      "step": 3758
+    },
+    {
+      "epoch": 0.16509874090255441,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.680997006820965e-05,
+      "loss": 0.5137,
+      "step": 3760
+    },
+    {
+      "epoch": 0.1651865593817579,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.680658761221628e-05,
+      "loss": 0.53,
+      "step": 3762
+    },
+    {
+      "epoch": 0.16527437786096139,
+      "grad_norm": 2.34375,
+      "learning_rate": 4.6803203486269195e-05,
+      "loss": 0.552,
+      "step": 3764
+    },
+    {
+      "epoch": 0.16536219634016489,
+      "grad_norm": 2.421875,
+      "learning_rate": 4.679981769062756e-05,
+      "loss": 0.527,
+      "step": 3766
+    },
+    {
+      "epoch": 0.16545001481936836,
+      "grad_norm": 2.46875,
+      "learning_rate": 4.6796430225550664e-05,
+      "loss": 0.5067,
+      "step": 3768
+    },
+    {
+      "epoch": 0.16553783329857186,
+      "grad_norm": 2.625,
+      "learning_rate": 4.679304109129792e-05,
+      "loss": 0.5127,
+      "step": 3770
+    },
+    {
+      "epoch": 0.16562565177777533,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.6789650288128855e-05,
+      "loss": 0.5087,
+      "step": 3772
+    },
+    {
+      "epoch": 0.16571347025697883,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.678625781630315e-05,
+      "loss": 0.5192,
+      "step": 3774
+    },
+    {
+      "epoch": 0.1658012887361823,
+      "grad_norm": 3.46875,
+      "learning_rate": 4.6782863676080605e-05,
+      "loss": 0.5258,
+      "step": 3776
+    },
+    {
+      "epoch": 0.1658891072153858,
+      "grad_norm": 3.296875,
+      "learning_rate": 4.6779467867721135e-05,
+      "loss": 0.5352,
+      "step": 3778
+    },
+    {
+      "epoch": 0.16597692569458927,
+      "grad_norm": 3.28125,
+      "learning_rate": 4.67760703914848e-05,
+      "loss": 0.5331,
+      "step": 3780
+    },
+    {
+      "epoch": 0.16606474417379277,
+      "grad_norm": 2.75,
+      "learning_rate": 4.6772671247631764e-05,
+      "loss": 0.5256,
+      "step": 3782
+    },
+    {
+      "epoch": 0.16615256265299624,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.676927043642235e-05,
+      "loss": 0.5047,
+      "step": 3784
+    },
+    {
+      "epoch": 0.16624038113219974,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.676586795811699e-05,
+      "loss": 0.5068,
+      "step": 3786
+    },
+    {
+      "epoch": 0.16632819961140322,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.676246381297624e-05,
+      "loss": 0.5406,
+      "step": 3788
+    },
+    {
+      "epoch": 0.16641601809060672,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.6759058001260794e-05,
+      "loss": 0.4963,
+      "step": 3790
+    },
+    {
+      "epoch": 0.16650383656981022,
+      "grad_norm": 2.75,
+      "learning_rate": 4.6755650523231476e-05,
+      "loss": 0.5411,
+      "step": 3792
+    },
+    {
+      "epoch": 0.1665916550490137,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.675224137914922e-05,
+      "loss": 0.5449,
+      "step": 3794
+    },
+    {
+      "epoch": 0.1666794735282172,
+      "grad_norm": 3.0625,
+      "learning_rate": 4.674883056927511e-05,
+      "loss": 0.5374,
+      "step": 3796
+    },
+    {
+      "epoch": 0.16676729200742066,
+      "grad_norm": 3.40625,
+      "learning_rate": 4.674541809387033e-05,
+      "loss": 0.531,
+      "step": 3798
+    },
+    {
+      "epoch": 0.16685511048662416,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.674200395319623e-05,
+      "loss": 0.5496,
+      "step": 3800
+    },
+    {
+      "epoch": 0.16694292896582763,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.673858814751425e-05,
+      "loss": 0.5026,
+      "step": 3802
+    },
+    {
+      "epoch": 0.16703074744503113,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.673517067708598e-05,
+      "loss": 0.5376,
+      "step": 3804
+    },
+    {
+      "epoch": 0.1671185659242346,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.6731751542173136e-05,
+      "loss": 0.4992,
+      "step": 3806
+    },
+    {
+      "epoch": 0.1672063844034381,
+      "grad_norm": 2.515625,
+      "learning_rate": 4.672833074303754e-05,
+      "loss": 0.5246,
+      "step": 3808
+    },
+    {
+      "epoch": 0.16729420288264157,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.672490827994117e-05,
+      "loss": 0.5128,
+      "step": 3810
+    },
+    {
+      "epoch": 0.16738202136184507,
+      "grad_norm": 3.078125,
+      "learning_rate": 4.6721484153146124e-05,
+      "loss": 0.5459,
+      "step": 3812
+    },
+    {
+      "epoch": 0.16746983984104855,
+      "grad_norm": 3.265625,
+      "learning_rate": 4.671805836291461e-05,
+      "loss": 0.5388,
+      "step": 3814
+    },
+    {
+      "epoch": 0.16755765832025205,
+      "grad_norm": 2.75,
+      "learning_rate": 4.671463090950897e-05,
+      "loss": 0.5167,
+      "step": 3816
+    },
+    {
+      "epoch": 0.16764547679945552,
+      "grad_norm": 3.125,
+      "learning_rate": 4.67112017931917e-05,
+      "loss": 0.5059,
+      "step": 3818
+    },
+    {
+      "epoch": 0.16773329527865902,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.670777101422539e-05,
+      "loss": 0.5303,
+      "step": 3820
+    },
+    {
+      "epoch": 0.1678211137578625,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.6704338572872773e-05,
+      "loss": 0.5233,
+      "step": 3822
+    },
+    {
+      "epoch": 0.167908932237066,
+      "grad_norm": 3.21875,
+      "learning_rate": 4.67009044693967e-05,
+      "loss": 0.5091,
+      "step": 3824
+    },
+    {
+      "epoch": 0.16799675071626946,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.6697468704060166e-05,
+      "loss": 0.524,
+      "step": 3826
+    },
+    {
+      "epoch": 0.16808456919547296,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.6694031277126285e-05,
+      "loss": 0.5386,
+      "step": 3828
+    },
+    {
+      "epoch": 0.16817238767467643,
+      "grad_norm": 3.1875,
+      "learning_rate": 4.6690592188858275e-05,
+      "loss": 0.5196,
+      "step": 3830
+    },
+    {
+      "epoch": 0.16826020615387993,
+      "grad_norm": 3.4375,
+      "learning_rate": 4.6687151439519516e-05,
+      "loss": 0.5395,
+      "step": 3832
+    },
+    {
+      "epoch": 0.1683480246330834,
+      "grad_norm": 3.34375,
+      "learning_rate": 4.668370902937351e-05,
+      "loss": 0.5279,
+      "step": 3834
+    },
+    {
+      "epoch": 0.1684358431122869,
+      "grad_norm": 3.46875,
+      "learning_rate": 4.668026495868386e-05,
+      "loss": 0.5156,
+      "step": 3836
+    },
+    {
+      "epoch": 0.16852366159149038,
+      "grad_norm": 2.625,
+      "learning_rate": 4.667681922771433e-05,
+      "loss": 0.4916,
+      "step": 3838
+    },
+    {
+      "epoch": 0.16861148007069388,
+      "grad_norm": 2.546875,
+      "learning_rate": 4.6673371836728785e-05,
+      "loss": 0.5422,
+      "step": 3840
+    },
+    {
+      "epoch": 0.16869929854989738,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.6669922785991225e-05,
+      "loss": 0.5224,
+      "step": 3842
+    },
+    {
+      "epoch": 0.16878711702910085,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.666647207576579e-05,
+      "loss": 0.5403,
+      "step": 3844
+    },
+    {
+      "epoch": 0.16887493550830435,
+      "grad_norm": 2.515625,
+      "learning_rate": 4.666301970631672e-05,
+      "loss": 0.5061,
+      "step": 3846
+    },
+    {
+      "epoch": 0.16896275398750782,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.6659565677908414e-05,
+      "loss": 0.504,
+      "step": 3848
+    },
+    {
+      "epoch": 0.16905057246671132,
+      "grad_norm": 3.953125,
+      "learning_rate": 4.665610999080537e-05,
+      "loss": 0.5388,
+      "step": 3850
+    },
+    {
+      "epoch": 0.1691383909459148,
+      "grad_norm": 3.0625,
+      "learning_rate": 4.6652652645272244e-05,
+      "loss": 0.5559,
+      "step": 3852
+    },
+    {
+      "epoch": 0.1692262094251183,
+      "grad_norm": 3.046875,
+      "learning_rate": 4.6649193641573784e-05,
+      "loss": 0.5189,
+      "step": 3854
+    },
+    {
+      "epoch": 0.16931402790432176,
+      "grad_norm": 3.453125,
+      "learning_rate": 4.6645732979974884e-05,
+      "loss": 0.5244,
+      "step": 3856
+    },
+    {
+      "epoch": 0.16940184638352526,
+      "grad_norm": 3.0,
+      "learning_rate": 4.664227066074056e-05,
+      "loss": 0.5008,
+      "step": 3858
+    },
+    {
+      "epoch": 0.16948966486272873,
+      "grad_norm": 2.953125,
+      "learning_rate": 4.663880668413596e-05,
+      "loss": 0.5339,
+      "step": 3860
+    },
+    {
+      "epoch": 0.16957748334193223,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.663534105042636e-05,
+      "loss": 0.5334,
+      "step": 3862
+    },
+    {
+      "epoch": 0.1696653018211357,
+      "grad_norm": 3.296875,
+      "learning_rate": 4.6631873759877156e-05,
+      "loss": 0.5465,
+      "step": 3864
+    },
+    {
+      "epoch": 0.1697531203003392,
+      "grad_norm": 3.9375,
+      "learning_rate": 4.6628404812753876e-05,
+      "loss": 0.523,
+      "step": 3866
+    },
+    {
+      "epoch": 0.16984093877954268,
+      "grad_norm": 3.453125,
+      "learning_rate": 4.662493420932217e-05,
+      "loss": 0.5374,
+      "step": 3868
+    },
+    {
+      "epoch": 0.16992875725874618,
+      "grad_norm": 3.234375,
+      "learning_rate": 4.662146194984782e-05,
+      "loss": 0.5576,
+      "step": 3870
+    },
+    {
+      "epoch": 0.17001657573794965,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.661798803459673e-05,
+      "loss": 0.521,
+      "step": 3872
+    },
+    {
+      "epoch": 0.17010439421715315,
+      "grad_norm": 3.328125,
+      "learning_rate": 4.661451246383492e-05,
+      "loss": 0.5187,
+      "step": 3874
+    },
+    {
+      "epoch": 0.17019221269635662,
+      "grad_norm": 3.46875,
+      "learning_rate": 4.661103523782858e-05,
+      "loss": 0.4808,
+      "step": 3876
+    },
+    {
+      "epoch": 0.17028003117556012,
+      "grad_norm": 3.703125,
+      "learning_rate": 4.660755635684398e-05,
+      "loss": 0.5104,
+      "step": 3878
+    },
+    {
+      "epoch": 0.1703678496547636,
+      "grad_norm": 3.734375,
+      "learning_rate": 4.660407582114753e-05,
+      "loss": 0.5527,
+      "step": 3880
+    },
+    {
+      "epoch": 0.1704556681339671,
+      "grad_norm": 3.328125,
+      "learning_rate": 4.6600593631005776e-05,
+      "loss": 0.5165,
+      "step": 3882
+    },
+    {
+      "epoch": 0.17054348661317056,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.6597109786685386e-05,
+      "loss": 0.5142,
+      "step": 3884
+    },
+    {
+      "epoch": 0.17063130509237406,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.6593624288453146e-05,
+      "loss": 0.5223,
+      "step": 3886
+    },
+    {
+      "epoch": 0.17071912357157754,
+      "grad_norm": 2.4375,
+      "learning_rate": 4.6590137136575986e-05,
+      "loss": 0.54,
+      "step": 3888
+    },
+    {
+      "epoch": 0.17080694205078104,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.6586648331320946e-05,
+      "loss": 0.4989,
+      "step": 3890
+    },
+    {
+      "epoch": 0.17089476052998454,
+      "grad_norm": 2.53125,
+      "learning_rate": 4.6583157872955205e-05,
+      "loss": 0.5059,
+      "step": 3892
+    },
+    {
+      "epoch": 0.170982579009188,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.657966576174606e-05,
+      "loss": 0.5045,
+      "step": 3894
+    },
+    {
+      "epoch": 0.1710703974883915,
+      "grad_norm": 2.75,
+      "learning_rate": 4.657617199796094e-05,
+      "loss": 0.5024,
+      "step": 3896
+    },
+    {
+      "epoch": 0.17115821596759498,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.657267658186739e-05,
+      "loss": 0.5587,
+      "step": 3898
+    },
+    {
+      "epoch": 0.17124603444679848,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.65691795137331e-05,
+      "loss": 0.4818,
+      "step": 3900
+    },
+    {
+      "epoch": 0.17133385292600195,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.6565680793825874e-05,
+      "loss": 0.5296,
+      "step": 3902
+    },
+    {
+      "epoch": 0.17142167140520545,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.6562180422413635e-05,
+      "loss": 0.5506,
+      "step": 3904
+    },
+    {
+      "epoch": 0.17150948988440892,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.655867839976447e-05,
+      "loss": 0.5211,
+      "step": 3906
+    },
+    {
+      "epoch": 0.17159730836361242,
+      "grad_norm": 2.484375,
+      "learning_rate": 4.655517472614652e-05,
+      "loss": 0.5108,
+      "step": 3908
+    },
+    {
+      "epoch": 0.1716851268428159,
+      "grad_norm": 2.390625,
+      "learning_rate": 4.655166940182813e-05,
+      "loss": 0.5492,
+      "step": 3910
+    },
+    {
+      "epoch": 0.1717729453220194,
+      "grad_norm": 2.75,
+      "learning_rate": 4.654816242707774e-05,
+      "loss": 0.5077,
+      "step": 3912
+    },
+    {
+      "epoch": 0.17186076380122287,
+      "grad_norm": 2.53125,
+      "learning_rate": 4.6544653802163904e-05,
+      "loss": 0.5369,
+      "step": 3914
+    },
+    {
+      "epoch": 0.17194858228042637,
+      "grad_norm": 2.328125,
+      "learning_rate": 4.654114352735531e-05,
+      "loss": 0.512,
+      "step": 3916
+    },
+    {
+      "epoch": 0.17203640075962984,
+      "grad_norm": 2.28125,
+      "learning_rate": 4.653763160292078e-05,
+      "loss": 0.5409,
+      "step": 3918
+    },
+    {
+      "epoch": 0.17212421923883334,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.653411802912926e-05,
+      "loss": 0.5141,
+      "step": 3920
+    },
+    {
+      "epoch": 0.1722120377180368,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.6530602806249825e-05,
+      "loss": 0.5354,
+      "step": 3922
+    },
+    {
+      "epoch": 0.1722998561972403,
+      "grad_norm": 2.390625,
+      "learning_rate": 4.652708593455166e-05,
+      "loss": 0.5425,
+      "step": 3924
+    },
+    {
+      "epoch": 0.17238767467644378,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.6523567414304095e-05,
+      "loss": 0.5343,
+      "step": 3926
+    },
+    {
+      "epoch": 0.17247549315564728,
+      "grad_norm": 2.953125,
+      "learning_rate": 4.6520047245776575e-05,
+      "loss": 0.5373,
+      "step": 3928
+    },
+    {
+      "epoch": 0.17256331163485075,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.651652542923869e-05,
+      "loss": 0.5034,
+      "step": 3930
+    },
+    {
+      "epoch": 0.17265113011405425,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.6513001964960116e-05,
+      "loss": 0.514,
+      "step": 3932
+    },
+    {
+      "epoch": 0.17273894859325772,
+      "grad_norm": 3.515625,
+      "learning_rate": 4.65094768532107e-05,
+      "loss": 0.5189,
+      "step": 3934
+    },
+    {
+      "epoch": 0.17282676707246122,
+      "grad_norm": 3.25,
+      "learning_rate": 4.650595009426039e-05,
+      "loss": 0.5327,
+      "step": 3936
+    },
+    {
+      "epoch": 0.1729145855516647,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.650242168837927e-05,
+      "loss": 0.5272,
+      "step": 3938
+    },
+    {
+      "epoch": 0.1730024040308682,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.649889163583754e-05,
+      "loss": 0.5732,
+      "step": 3940
+    },
+    {
+      "epoch": 0.1730902225100717,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.649535993690554e-05,
+      "loss": 0.5236,
+      "step": 3942
+    },
+    {
+      "epoch": 0.17317804098927517,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.649182659185371e-05,
+      "loss": 0.4678,
+      "step": 3944
+    },
+    {
+      "epoch": 0.17326585946847867,
+      "grad_norm": 2.40625,
+      "learning_rate": 4.648829160095265e-05,
+      "loss": 0.5066,
+      "step": 3946
+    },
+    {
+      "epoch": 0.17335367794768214,
+      "grad_norm": 2.421875,
+      "learning_rate": 4.6484754964473076e-05,
+      "loss": 0.4917,
+      "step": 3948
+    },
+    {
+      "epoch": 0.17344149642688564,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.6481216682685816e-05,
+      "loss": 0.5121,
+      "step": 3950
+    },
+    {
+      "epoch": 0.1735293149060891,
+      "grad_norm": 2.359375,
+      "learning_rate": 4.647767675586183e-05,
+      "loss": 0.5145,
+      "step": 3952
+    },
+    {
+      "epoch": 0.1736171333852926,
+      "grad_norm": 2.875,
+      "learning_rate": 4.647413518427221e-05,
+      "loss": 0.5074,
+      "step": 3954
+    },
+    {
+      "epoch": 0.17370495186449608,
+      "grad_norm": 2.546875,
+      "learning_rate": 4.647059196818817e-05,
+      "loss": 0.5168,
+      "step": 3956
+    },
+    {
+      "epoch": 0.17379277034369958,
+      "grad_norm": 2.421875,
+      "learning_rate": 4.646704710788105e-05,
+      "loss": 0.5364,
+      "step": 3958
+    },
+    {
+      "epoch": 0.17388058882290305,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.6463500603622316e-05,
+      "loss": 0.5028,
+      "step": 3960
+    },
+    {
+      "epoch": 0.17396840730210655,
+      "grad_norm": 3.140625,
+      "learning_rate": 4.645995245568357e-05,
+      "loss": 0.5176,
+      "step": 3962
+    },
+    {
+      "epoch": 0.17405622578131003,
+      "grad_norm": 3.203125,
+      "learning_rate": 4.645640266433651e-05,
+      "loss": 0.5466,
+      "step": 3964
+    },
+    {
+      "epoch": 0.17414404426051353,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.645285122985299e-05,
+      "loss": 0.5072,
+      "step": 3966
+    },
+    {
+      "epoch": 0.174231862739717,
+      "grad_norm": 3.40625,
+      "learning_rate": 4.6449298152504986e-05,
+      "loss": 0.5595,
+      "step": 3968
+    },
+    {
+      "epoch": 0.1743196812189205,
+      "grad_norm": 3.546875,
+      "learning_rate": 4.644574343256458e-05,
+      "loss": 0.5338,
+      "step": 3970
+    },
+    {
+      "epoch": 0.17440749969812397,
+      "grad_norm": 3.46875,
+      "learning_rate": 4.6442187070304004e-05,
+      "loss": 0.5155,
+      "step": 3972
+    },
+    {
+      "epoch": 0.17449531817732747,
+      "grad_norm": 3.421875,
+      "learning_rate": 4.643862906599561e-05,
+      "loss": 0.513,
+      "step": 3974
+    },
+    {
+      "epoch": 0.17458313665653094,
+      "grad_norm": 3.5625,
+      "learning_rate": 4.6435069419911854e-05,
+      "loss": 0.5436,
+      "step": 3976
+    },
+    {
+      "epoch": 0.17467095513573444,
+      "grad_norm": 3.09375,
+      "learning_rate": 4.643150813232534e-05,
+      "loss": 0.5065,
+      "step": 3978
+    },
+    {
+      "epoch": 0.1747587736149379,
+      "grad_norm": 3.90625,
+      "learning_rate": 4.6427945203508806e-05,
+      "loss": 0.5314,
+      "step": 3980
+    },
+    {
+      "epoch": 0.1748465920941414,
+      "grad_norm": 4.03125,
+      "learning_rate": 4.642438063373509e-05,
+      "loss": 0.5111,
+      "step": 3982
+    },
+    {
+      "epoch": 0.17493441057334488,
+      "grad_norm": 3.125,
+      "learning_rate": 4.642081442327716e-05,
+      "loss": 0.5112,
+      "step": 3984
+    },
+    {
+      "epoch": 0.17502222905254838,
+      "grad_norm": 3.015625,
+      "learning_rate": 4.6417246572408134e-05,
+      "loss": 0.5393,
+      "step": 3986
+    },
+    {
+      "epoch": 0.17511004753175186,
+      "grad_norm": 2.546875,
+      "learning_rate": 4.641367708140124e-05,
+      "loss": 0.5214,
+      "step": 3988
+    },
+    {
+      "epoch": 0.17519786601095536,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.641010595052981e-05,
+      "loss": 0.5073,
+      "step": 3990
+    },
+    {
+      "epoch": 0.17528568449015886,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.640653318006733e-05,
+      "loss": 0.5078,
+      "step": 3992
+    },
+    {
+      "epoch": 0.17537350296936233,
+      "grad_norm": 2.625,
+      "learning_rate": 4.640295877028742e-05,
+      "loss": 0.5006,
+      "step": 3994
+    },
+    {
+      "epoch": 0.17546132144856583,
+      "grad_norm": 2.390625,
+      "learning_rate": 4.639938272146379e-05,
+      "loss": 0.5217,
+      "step": 3996
+    },
+    {
+      "epoch": 0.1755491399277693,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.639580503387031e-05,
+      "loss": 0.5335,
+      "step": 3998
+    },
+    {
+      "epoch": 0.1756369584069728,
+      "grad_norm": 2.875,
+      "learning_rate": 4.6392225707780935e-05,
+      "loss": 0.5138,
+      "step": 4000
+    },
+    {
+      "epoch": 0.17572477688617627,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.63886447434698e-05,
+      "loss": 0.5127,
+      "step": 4002
+    },
+    {
+      "epoch": 0.17581259536537977,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.638506214121112e-05,
+      "loss": 0.5268,
+      "step": 4004
+    },
+    {
+      "epoch": 0.17590041384458324,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.6381477901279255e-05,
+      "loss": 0.5176,
+      "step": 4006
+    },
+    {
+      "epoch": 0.17598823232378674,
+      "grad_norm": 2.53125,
+      "learning_rate": 4.6377892023948684e-05,
+      "loss": 0.5373,
+      "step": 4008
+    },
+    {
+      "epoch": 0.1760760508029902,
+      "grad_norm": 2.390625,
+      "learning_rate": 4.6374304509494016e-05,
+      "loss": 0.5187,
+      "step": 4010
+    },
+    {
+      "epoch": 0.1761638692821937,
+      "grad_norm": 3.1875,
+      "learning_rate": 4.637071535818999e-05,
+      "loss": 0.5387,
+      "step": 4012
+    },
+    {
+      "epoch": 0.17625168776139719,
+      "grad_norm": 2.53125,
+      "learning_rate": 4.636712457031146e-05,
+      "loss": 0.5008,
+      "step": 4014
+    },
+    {
+      "epoch": 0.17633950624060069,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.63635321461334e-05,
+      "loss": 0.5233,
+      "step": 4016
+    },
+    {
+      "epoch": 0.17642732471980416,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.635993808593093e-05,
+      "loss": 0.5024,
+      "step": 4018
+    },
+    {
+      "epoch": 0.17651514319900766,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.635634238997929e-05,
+      "loss": 0.5411,
+      "step": 4020
+    },
+    {
+      "epoch": 0.17660296167821113,
+      "grad_norm": 2.28125,
+      "learning_rate": 4.6352745058553826e-05,
+      "loss": 0.5314,
+      "step": 4022
+    },
+    {
+      "epoch": 0.17669078015741463,
+      "grad_norm": 2.484375,
+      "learning_rate": 4.634914609193002e-05,
+      "loss": 0.5405,
+      "step": 4024
+    },
+    {
+      "epoch": 0.1767785986366181,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.63455454903835e-05,
+      "loss": 0.5601,
+      "step": 4026
+    },
+    {
+      "epoch": 0.1768664171158216,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.634194325418998e-05,
+      "loss": 0.5059,
+      "step": 4028
+    },
+    {
+      "epoch": 0.17695423559502507,
+      "grad_norm": 2.5,
+      "learning_rate": 4.633833938362533e-05,
+      "loss": 0.5236,
+      "step": 4030
+    },
+    {
+      "epoch": 0.17704205407422857,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.633473387896554e-05,
+      "loss": 0.5242,
+      "step": 4032
+    },
+    {
+      "epoch": 0.17712987255343204,
+      "grad_norm": 2.625,
+      "learning_rate": 4.6331126740486704e-05,
+      "loss": 0.5302,
+      "step": 4034
+    },
+    {
+      "epoch": 0.17721769103263554,
+      "grad_norm": 2.21875,
+      "learning_rate": 4.632751796846508e-05,
+      "loss": 0.4761,
+      "step": 4036
+    },
+    {
+      "epoch": 0.17730550951183902,
+      "grad_norm": 2.5,
+      "learning_rate": 4.632390756317702e-05,
+      "loss": 0.5416,
+      "step": 4038
+    },
+    {
+      "epoch": 0.17739332799104252,
+      "grad_norm": 2.34375,
+      "learning_rate": 4.6320295524898996e-05,
+      "loss": 0.5317,
+      "step": 4040
+    },
+    {
+      "epoch": 0.17748114647024602,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.631668185390764e-05,
+      "loss": 0.5271,
+      "step": 4042
+    },
+    {
+      "epoch": 0.1775689649494495,
+      "grad_norm": 2.3125,
+      "learning_rate": 4.631306655047968e-05,
+      "loss": 0.5316,
+      "step": 4044
+    },
+    {
+      "epoch": 0.177656783428653,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.630944961489196e-05,
+      "loss": 0.548,
+      "step": 4046
+    },
+    {
+      "epoch": 0.17774460190785646,
+      "grad_norm": 2.484375,
+      "learning_rate": 4.630583104742149e-05,
+      "loss": 0.5027,
+      "step": 4048
+    },
+    {
+      "epoch": 0.17783242038705996,
+      "grad_norm": 2.421875,
+      "learning_rate": 4.6302210848345367e-05,
+      "loss": 0.5404,
+      "step": 4050
+    },
+    {
+      "epoch": 0.17792023886626343,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.6298589017940836e-05,
+      "loss": 0.5267,
+      "step": 4052
+    },
+    {
+      "epoch": 0.17800805734546693,
+      "grad_norm": 2.28125,
+      "learning_rate": 4.629496555648525e-05,
+      "loss": 0.5045,
+      "step": 4054
+    },
+    {
+      "epoch": 0.1780958758246704,
+      "grad_norm": 2.25,
+      "learning_rate": 4.629134046425609e-05,
+      "loss": 0.4785,
+      "step": 4056
+    },
+    {
+      "epoch": 0.1781836943038739,
+      "grad_norm": 2.34375,
+      "learning_rate": 4.628771374153098e-05,
+      "loss": 0.5289,
+      "step": 4058
+    },
+    {
+      "epoch": 0.17827151278307737,
+      "grad_norm": 2.4375,
+      "learning_rate": 4.628408538858765e-05,
+      "loss": 0.5295,
+      "step": 4060
+    },
+    {
+      "epoch": 0.17835933126228087,
+      "grad_norm": 2.421875,
+      "learning_rate": 4.628045540570396e-05,
+      "loss": 0.4988,
+      "step": 4062
+    },
+    {
+      "epoch": 0.17844714974148435,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.6276823793157884e-05,
+      "loss": 0.5448,
+      "step": 4064
+    },
+    {
+      "epoch": 0.17853496822068785,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.627319055122755e-05,
+      "loss": 0.5266,
+      "step": 4066
+    },
+    {
+      "epoch": 0.17862278669989132,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.626955568019119e-05,
+      "loss": 0.5096,
+      "step": 4068
+    },
+    {
+      "epoch": 0.17871060517909482,
+      "grad_norm": 3.21875,
+      "learning_rate": 4.626591918032715e-05,
+      "loss": 0.5105,
+      "step": 4070
+    },
+    {
+      "epoch": 0.1787984236582983,
+      "grad_norm": 3.265625,
+      "learning_rate": 4.626228105191392e-05,
+      "loss": 0.4999,
+      "step": 4072
+    },
+    {
+      "epoch": 0.1788862421375018,
+      "grad_norm": 3.0,
+      "learning_rate": 4.625864129523011e-05,
+      "loss": 0.5009,
+      "step": 4074
+    },
+    {
+      "epoch": 0.17897406061670526,
+      "grad_norm": 3.265625,
+      "learning_rate": 4.6254999910554456e-05,
+      "loss": 0.5187,
+      "step": 4076
+    },
+    {
+      "epoch": 0.17906187909590876,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.625135689816581e-05,
+      "loss": 0.5063,
+      "step": 4078
+    },
+    {
+      "epoch": 0.17914969757511223,
+      "grad_norm": 3.015625,
+      "learning_rate": 4.624771225834316e-05,
+      "loss": 0.5143,
+      "step": 4080
+    },
+    {
+      "epoch": 0.17923751605431573,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.624406599136562e-05,
+      "loss": 0.519,
+      "step": 4082
+    },
+    {
+      "epoch": 0.1793253345335192,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.62404180975124e-05,
+      "loss": 0.5254,
+      "step": 4084
+    },
+    {
+      "epoch": 0.1794131530127227,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.6236768577062876e-05,
+      "loss": 0.5358,
+      "step": 4086
+    },
+    {
+      "epoch": 0.17950097149192618,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.623311743029652e-05,
+      "loss": 0.4824,
+      "step": 4088
+    },
+    {
+      "epoch": 0.17958878997112968,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.622946465749295e-05,
+      "loss": 0.5151,
+      "step": 4090
+    },
+    {
+      "epoch": 0.17967660845033318,
+      "grad_norm": 3.1875,
+      "learning_rate": 4.622581025893189e-05,
+      "loss": 0.5062,
+      "step": 4092
+    },
+    {
+      "epoch": 0.17976442692953665,
+      "grad_norm": 3.234375,
+      "learning_rate": 4.622215423489318e-05,
+      "loss": 0.5197,
+      "step": 4094
+    },
+    {
+      "epoch": 0.17985224540874015,
+      "grad_norm": 3.09375,
+      "learning_rate": 4.6218496585656815e-05,
+      "loss": 0.5057,
+      "step": 4096
+    },
+    {
+      "epoch": 0.17994006388794362,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.62148373115029e-05,
+      "loss": 0.516,
+      "step": 4098
+    },
+    {
+      "epoch": 0.18002788236714712,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.621117641271166e-05,
+      "loss": 0.5342,
+      "step": 4100
+    },
+    {
+      "epoch": 0.1801157008463506,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.620751388956344e-05,
+      "loss": 0.52,
+      "step": 4102
+    },
+    {
+      "epoch": 0.1802035193255541,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.620384974233872e-05,
+      "loss": 0.5188,
+      "step": 4104
+    },
+    {
+      "epoch": 0.18029133780475756,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.6200183971318104e-05,
+      "loss": 0.5444,
+      "step": 4106
+    },
+    {
+      "epoch": 0.18037915628396106,
+      "grad_norm": 3.171875,
+      "learning_rate": 4.619651657678232e-05,
+      "loss": 0.526,
+      "step": 4108
+    },
+    {
+      "epoch": 0.18046697476316453,
+      "grad_norm": 2.34375,
+      "learning_rate": 4.619284755901222e-05,
+      "loss": 0.476,
+      "step": 4110
+    },
+    {
+      "epoch": 0.18055479324236803,
+      "grad_norm": 3.078125,
+      "learning_rate": 4.618917691828877e-05,
+      "loss": 0.5217,
+      "step": 4112
+    },
+    {
+      "epoch": 0.1806426117215715,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.618550465489307e-05,
+      "loss": 0.5009,
+      "step": 4114
+    },
+    {
+      "epoch": 0.180730430200775,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.618183076910635e-05,
+      "loss": 0.5349,
+      "step": 4116
+    },
+    {
+      "epoch": 0.18081824867997848,
+      "grad_norm": 3.109375,
+      "learning_rate": 4.6178155261209946e-05,
+      "loss": 0.5149,
+      "step": 4118
+    },
+    {
+      "epoch": 0.18090606715918198,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.6174478131485334e-05,
+      "loss": 0.513,
+      "step": 4120
+    },
+    {
+      "epoch": 0.18099388563838545,
+      "grad_norm": 3.359375,
+      "learning_rate": 4.617079938021412e-05,
+      "loss": 0.5095,
+      "step": 4122
+    },
+    {
+      "epoch": 0.18108170411758895,
+      "grad_norm": 3.5625,
+      "learning_rate": 4.616711900767801e-05,
+      "loss": 0.4875,
+      "step": 4124
+    },
+    {
+      "epoch": 0.18116952259679242,
+      "grad_norm": 2.953125,
+      "learning_rate": 4.616343701415885e-05,
+      "loss": 0.5149,
+      "step": 4126
+    },
+    {
+      "epoch": 0.18125734107599592,
+      "grad_norm": 3.125,
+      "learning_rate": 4.615975339993861e-05,
+      "loss": 0.5326,
+      "step": 4128
+    },
+    {
+      "epoch": 0.1813451595551994,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.615606816529939e-05,
+      "loss": 0.4981,
+      "step": 4130
+    },
+    {
+      "epoch": 0.1814329780344029,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.6152381310523387e-05,
+      "loss": 0.5039,
+      "step": 4132
+    },
+    {
+      "epoch": 0.18152079651360636,
+      "grad_norm": 2.359375,
+      "learning_rate": 4.614869283589296e-05,
+      "loss": 0.5351,
+      "step": 4134
+    },
+    {
+      "epoch": 0.18160861499280986,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.614500274169057e-05,
+      "loss": 0.5039,
+      "step": 4136
+    },
+    {
+      "epoch": 0.18169643347201334,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.6141311028198786e-05,
+      "loss": 0.5379,
+      "step": 4138
+    },
+    {
+      "epoch": 0.18178425195121684,
+      "grad_norm": 2.4375,
+      "learning_rate": 4.613761769570035e-05,
+      "loss": 0.5038,
+      "step": 4140
+    },
+    {
+      "epoch": 0.18187207043042034,
+      "grad_norm": 3.3125,
+      "learning_rate": 4.6133922744478076e-05,
+      "loss": 0.509,
+      "step": 4142
+    },
+    {
+      "epoch": 0.1819598889096238,
+      "grad_norm": 3.109375,
+      "learning_rate": 4.613022617481494e-05,
+      "loss": 0.4996,
+      "step": 4144
+    },
+    {
+      "epoch": 0.1820477073888273,
+      "grad_norm": 2.546875,
+      "learning_rate": 4.612652798699401e-05,
+      "loss": 0.5164,
+      "step": 4146
+    },
+    {
+      "epoch": 0.18213552586803078,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.612282818129851e-05,
+      "loss": 0.4928,
+      "step": 4148
+    },
+    {
+      "epoch": 0.18222334434723428,
+      "grad_norm": 2.5,
+      "learning_rate": 4.611912675801176e-05,
+      "loss": 0.499,
+      "step": 4150
+    },
+    {
+      "epoch": 0.18231116282643775,
+      "grad_norm": 2.359375,
+      "learning_rate": 4.611542371741722e-05,
+      "loss": 0.4836,
+      "step": 4152
+    },
+    {
+      "epoch": 0.18239898130564125,
+      "grad_norm": 2.453125,
+      "learning_rate": 4.6111719059798466e-05,
+      "loss": 0.4947,
+      "step": 4154
+    },
+    {
+      "epoch": 0.18248679978484472,
+      "grad_norm": 2.5,
+      "learning_rate": 4.610801278543921e-05,
+      "loss": 0.5091,
+      "step": 4156
+    },
+    {
+      "epoch": 0.18257461826404822,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.610430489462327e-05,
+      "loss": 0.5287,
+      "step": 4158
+    },
+    {
+      "epoch": 0.1826624367432517,
+      "grad_norm": 3.015625,
+      "learning_rate": 4.6100595387634616e-05,
+      "loss": 0.4934,
+      "step": 4160
+    },
+    {
+      "epoch": 0.1827502552224552,
+      "grad_norm": 2.875,
+      "learning_rate": 4.60968842647573e-05,
+      "loss": 0.5256,
+      "step": 4162
+    },
+    {
+      "epoch": 0.18283807370165867,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.6093171526275524e-05,
+      "loss": 0.5091,
+      "step": 4164
+    },
+    {
+      "epoch": 0.18292589218086217,
+      "grad_norm": 2.421875,
+      "learning_rate": 4.608945717247363e-05,
+      "loss": 0.5385,
+      "step": 4166
+    },
+    {
+      "epoch": 0.18301371066006564,
+      "grad_norm": 2.546875,
+      "learning_rate": 4.6085741203636035e-05,
+      "loss": 0.5128,
+      "step": 4168
+    },
+    {
+      "epoch": 0.18310152913926914,
+      "grad_norm": 2.53125,
+      "learning_rate": 4.6082023620047335e-05,
+      "loss": 0.5135,
+      "step": 4170
+    },
+    {
+      "epoch": 0.1831893476184726,
+      "grad_norm": 2.453125,
+      "learning_rate": 4.60783044219922e-05,
+      "loss": 0.5336,
+      "step": 4172
+    },
+    {
+      "epoch": 0.1832771660976761,
+      "grad_norm": 2.46875,
+      "learning_rate": 4.607458360975547e-05,
+      "loss": 0.5043,
+      "step": 4174
+    },
+    {
+      "epoch": 0.18336498457687958,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.607086118362208e-05,
+      "loss": 0.4748,
+      "step": 4176
+    },
+    {
+      "epoch": 0.18345280305608308,
+      "grad_norm": 3.125,
+      "learning_rate": 4.6067137143877084e-05,
+      "loss": 0.5004,
+      "step": 4178
+    },
+    {
+      "epoch": 0.18354062153528655,
+      "grad_norm": 3.6875,
+      "learning_rate": 4.606341149080567e-05,
+      "loss": 0.4805,
+      "step": 4180
+    },
+    {
+      "epoch": 0.18362844001449005,
+      "grad_norm": 3.71875,
+      "learning_rate": 4.605968422469316e-05,
+      "loss": 0.4993,
+      "step": 4182
+    },
+    {
+      "epoch": 0.18371625849369352,
+      "grad_norm": 4.28125,
+      "learning_rate": 4.605595534582498e-05,
+      "loss": 0.539,
+      "step": 4184
+    },
+    {
+      "epoch": 0.18380407697289702,
+      "grad_norm": 3.953125,
+      "learning_rate": 4.60522248544867e-05,
+      "loss": 0.5105,
+      "step": 4186
+    },
+    {
+      "epoch": 0.1838918954521005,
+      "grad_norm": 3.15625,
+      "learning_rate": 4.604849275096399e-05,
+      "loss": 0.5194,
+      "step": 4188
+    },
+    {
+      "epoch": 0.183979713931304,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.604475903554266e-05,
+      "loss": 0.5028,
+      "step": 4190
+    },
+    {
+      "epoch": 0.18406753241050747,
+      "grad_norm": 2.53125,
+      "learning_rate": 4.6041023708508635e-05,
+      "loss": 0.49,
+      "step": 4192
+    },
+    {
+      "epoch": 0.18415535088971097,
+      "grad_norm": 2.296875,
+      "learning_rate": 4.603728677014797e-05,
+      "loss": 0.4791,
+      "step": 4194
+    },
+    {
+      "epoch": 0.18424316936891447,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.6033548220746846e-05,
+      "loss": 0.5261,
+      "step": 4196
+    },
+    {
+      "epoch": 0.18433098784811794,
+      "grad_norm": 2.359375,
+      "learning_rate": 4.602980806059155e-05,
+      "loss": 0.5155,
+      "step": 4198
+    },
+    {
+      "epoch": 0.18441880632732144,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.602606628996852e-05,
+      "loss": 0.4972,
+      "step": 4200
+    },
+    {
+      "epoch": 0.1845066248065249,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.6022322909164286e-05,
+      "loss": 0.5362,
+      "step": 4202
+    },
+    {
+      "epoch": 0.1845944432857284,
+      "grad_norm": 3.5625,
+      "learning_rate": 4.601857791846553e-05,
+      "loss": 0.4848,
+      "step": 4204
+    },
+    {
+      "epoch": 0.18468226176493188,
+      "grad_norm": 3.59375,
+      "learning_rate": 4.6014831318159036e-05,
+      "loss": 0.5352,
+      "step": 4206
+    },
+    {
+      "epoch": 0.18477008024413538,
+      "grad_norm": 3.4375,
+      "learning_rate": 4.6011083108531725e-05,
+      "loss": 0.5178,
+      "step": 4208
+    },
+    {
+      "epoch": 0.18485789872333885,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.6007333289870625e-05,
+      "loss": 0.4915,
+      "step": 4210
+    },
+    {
+      "epoch": 0.18494571720254235,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.600358186246291e-05,
+      "loss": 0.4989,
+      "step": 4212
+    },
+    {
+      "epoch": 0.18503353568174583,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.599982882659586e-05,
+      "loss": 0.5105,
+      "step": 4214
+    },
+    {
+      "epoch": 0.18512135416094933,
+      "grad_norm": 2.4375,
+      "learning_rate": 4.599607418255689e-05,
+      "loss": 0.5102,
+      "step": 4216
+    },
+    {
+      "epoch": 0.1852091726401528,
+      "grad_norm": 2.421875,
+      "learning_rate": 4.5992317930633524e-05,
+      "loss": 0.5115,
+      "step": 4218
+    },
+    {
+      "epoch": 0.1852969911193563,
+      "grad_norm": 2.46875,
+      "learning_rate": 4.5988560071113416e-05,
+      "loss": 0.5128,
+      "step": 4220
+    },
+    {
+      "epoch": 0.18538480959855977,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.598480060428435e-05,
+      "loss": 0.5167,
+      "step": 4222
+    },
+    {
+      "epoch": 0.18547262807776327,
+      "grad_norm": 3.25,
+      "learning_rate": 4.598103953043422e-05,
+      "loss": 0.525,
+      "step": 4224
+    },
+    {
+      "epoch": 0.18556044655696674,
+      "grad_norm": 3.640625,
+      "learning_rate": 4.597727684985105e-05,
+      "loss": 0.5016,
+      "step": 4226
+    },
+    {
+      "epoch": 0.18564826503617024,
+      "grad_norm": 4.9375,
+      "learning_rate": 4.5973512562823e-05,
+      "loss": 0.5174,
+      "step": 4228
+    },
+    {
+      "epoch": 0.1857360835153737,
+      "grad_norm": 3.78125,
+      "learning_rate": 4.596974666963832e-05,
+      "loss": 0.5166,
+      "step": 4230
+    },
+    {
+      "epoch": 0.1858239019945772,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.596597917058542e-05,
+      "loss": 0.5226,
+      "step": 4232
+    },
+    {
+      "epoch": 0.18591172047378068,
+      "grad_norm": 2.203125,
+      "learning_rate": 4.59622100659528e-05,
+      "loss": 0.5479,
+      "step": 4234
+    },
+    {
+      "epoch": 0.18599953895298418,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.595843935602913e-05,
+      "loss": 0.5321,
+      "step": 4236
+    },
+    {
+      "epoch": 0.18608735743218766,
+      "grad_norm": 3.25,
+      "learning_rate": 4.5954667041103126e-05,
+      "loss": 0.5096,
+      "step": 4238
+    },
+    {
+      "epoch": 0.18617517591139116,
+      "grad_norm": 5.15625,
+      "learning_rate": 4.595089312146371e-05,
+      "loss": 0.4967,
+      "step": 4240
+    },
+    {
+      "epoch": 0.18626299439059463,
+      "grad_norm": 4.5625,
+      "learning_rate": 4.594711759739987e-05,
+      "loss": 0.5105,
+      "step": 4242
+    },
+    {
+      "epoch": 0.18635081286979813,
+      "grad_norm": 3.453125,
+      "learning_rate": 4.594334046920075e-05,
+      "loss": 0.5182,
+      "step": 4244
+    },
+    {
+      "epoch": 0.18643863134900163,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.593956173715559e-05,
+      "loss": 0.5176,
+      "step": 4246
+    },
+    {
+      "epoch": 0.1865264498282051,
+      "grad_norm": 3.25,
+      "learning_rate": 4.593578140155377e-05,
+      "loss": 0.4972,
+      "step": 4248
+    },
+    {
+      "epoch": 0.1866142683074086,
+      "grad_norm": 3.3125,
+      "learning_rate": 4.593199946268479e-05,
+      "loss": 0.4917,
+      "step": 4250
+    },
+    {
+      "epoch": 0.18670208678661207,
+      "grad_norm": 2.984375,
+      "learning_rate": 4.592821592083829e-05,
+      "loss": 0.5161,
+      "step": 4252
+    },
+    {
+      "epoch": 0.18678990526581557,
+      "grad_norm": 3.0625,
+      "learning_rate": 4.592443077630398e-05,
+      "loss": 0.4934,
+      "step": 4254
+    },
+    {
+      "epoch": 0.18687772374501904,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.5920644029371764e-05,
+      "loss": 0.4934,
+      "step": 4256
+    },
+    {
+      "epoch": 0.18696554222422254,
+      "grad_norm": 2.625,
+      "learning_rate": 4.5916855680331596e-05,
+      "loss": 0.493,
+      "step": 4258
+    },
+    {
+      "epoch": 0.187053360703426,
+      "grad_norm": 2.5,
+      "learning_rate": 4.5913065729473614e-05,
+      "loss": 0.5358,
+      "step": 4260
+    },
+    {
+      "epoch": 0.1871411791826295,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.590927417708804e-05,
+      "loss": 0.4723,
+      "step": 4262
+    },
+    {
+      "epoch": 0.18722899766183299,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.590548102346525e-05,
+      "loss": 0.5063,
+      "step": 4264
+    },
+    {
+      "epoch": 0.18731681614103649,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.5901686268895694e-05,
+      "loss": 0.5412,
+      "step": 4266
+    },
+    {
+      "epoch": 0.18740463462023996,
+      "grad_norm": 2.953125,
+      "learning_rate": 4.589788991367e-05,
+      "loss": 0.496,
+      "step": 4268
+    },
+    {
+      "epoch": 0.18749245309944346,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.5894091958078886e-05,
+      "loss": 0.4843,
+      "step": 4270
+    },
+    {
+      "epoch": 0.18758027157864693,
+      "grad_norm": 2.34375,
+      "learning_rate": 4.5890292402413205e-05,
+      "loss": 0.4969,
+      "step": 4272
+    },
+    {
+      "epoch": 0.18766809005785043,
+      "grad_norm": 2.359375,
+      "learning_rate": 4.5886491246963925e-05,
+      "loss": 0.4921,
+      "step": 4274
+    },
+    {
+      "epoch": 0.1877559085370539,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.5882688492022134e-05,
+      "loss": 0.5024,
+      "step": 4276
+    },
+    {
+      "epoch": 0.1878437270162574,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.587888413787905e-05,
+      "loss": 0.4909,
+      "step": 4278
+    },
+    {
+      "epoch": 0.18793154549546087,
+      "grad_norm": 2.5,
+      "learning_rate": 4.587507818482603e-05,
+      "loss": 0.5371,
+      "step": 4280
+    },
+    {
+      "epoch": 0.18801936397466437,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.58712706331545e-05,
+      "loss": 0.5265,
+      "step": 4282
+    },
+    {
+      "epoch": 0.18810718245386784,
+      "grad_norm": 2.265625,
+      "learning_rate": 4.586746148315607e-05,
+      "loss": 0.5251,
+      "step": 4284
+    },
+    {
+      "epoch": 0.18819500093307134,
+      "grad_norm": 2.546875,
+      "learning_rate": 4.5863650735122435e-05,
+      "loss": 0.4889,
+      "step": 4286
+    },
+    {
+      "epoch": 0.18828281941227482,
+      "grad_norm": 2.484375,
+      "learning_rate": 4.5859838389345425e-05,
+      "loss": 0.4883,
+      "step": 4288
+    },
+    {
+      "epoch": 0.18837063789147832,
+      "grad_norm": 2.359375,
+      "learning_rate": 4.585602444611699e-05,
+      "loss": 0.4715,
+      "step": 4290
+    },
+    {
+      "epoch": 0.1884584563706818,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.58522089057292e-05,
+      "loss": 0.4844,
+      "step": 4292
+    },
+    {
+      "epoch": 0.1885462748498853,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.584839176847426e-05,
+      "loss": 0.516,
+      "step": 4294
+    },
+    {
+      "epoch": 0.1886340933290888,
+      "grad_norm": 2.25,
+      "learning_rate": 4.5844573034644485e-05,
+      "loss": 0.5009,
+      "step": 4296
+    },
+    {
+      "epoch": 0.18872191180829226,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.5840752704532304e-05,
+      "loss": 0.5116,
+      "step": 4298
+    },
+    {
+      "epoch": 0.18880973028749576,
+      "grad_norm": 3.015625,
+      "learning_rate": 4.583693077843028e-05,
+      "loss": 0.5064,
+      "step": 4300
+    },
+    {
+      "epoch": 0.18889754876669923,
+      "grad_norm": 3.078125,
+      "learning_rate": 4.5833107256631114e-05,
+      "loss": 0.513,
+      "step": 4302
+    },
+    {
+      "epoch": 0.18898536724590273,
+      "grad_norm": 2.484375,
+      "learning_rate": 4.582928213942759e-05,
+      "loss": 0.5207,
+      "step": 4304
+    },
+    {
+      "epoch": 0.1890731857251062,
+      "grad_norm": 2.3125,
+      "learning_rate": 4.582545542711265e-05,
+      "loss": 0.5153,
+      "step": 4306
+    },
+    {
+      "epoch": 0.1891610042043097,
+      "grad_norm": 2.40625,
+      "learning_rate": 4.582162711997934e-05,
+      "loss": 0.4788,
+      "step": 4308
+    },
+    {
+      "epoch": 0.18924882268351317,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.581779721832084e-05,
+      "loss": 0.4913,
+      "step": 4310
+    },
+    {
+      "epoch": 0.18933664116271667,
+      "grad_norm": 2.296875,
+      "learning_rate": 4.581396572243043e-05,
+      "loss": 0.507,
+      "step": 4312
+    },
+    {
+      "epoch": 0.18942445964192015,
+      "grad_norm": 2.375,
+      "learning_rate": 4.581013263260154e-05,
+      "loss": 0.4794,
+      "step": 4314
+    },
+    {
+      "epoch": 0.18951227812112365,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.58062979491277e-05,
+      "loss": 0.4966,
+      "step": 4316
+    },
+    {
+      "epoch": 0.18960009660032712,
+      "grad_norm": 2.625,
+      "learning_rate": 4.580246167230258e-05,
+      "loss": 0.5002,
+      "step": 4318
+    },
+    {
+      "epoch": 0.18968791507953062,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.579862380241996e-05,
+      "loss": 0.4809,
+      "step": 4320
+    },
+    {
+      "epoch": 0.1897757335587341,
+      "grad_norm": 3.046875,
+      "learning_rate": 4.579478433977374e-05,
+      "loss": 0.4919,
+      "step": 4322
+    },
+    {
+      "epoch": 0.1898635520379376,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.579094328465795e-05,
+      "loss": 0.5036,
+      "step": 4324
+    },
+    {
+      "epoch": 0.18995137051714106,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.578710063736674e-05,
+      "loss": 0.5281,
+      "step": 4326
+    },
+    {
+      "epoch": 0.19003918899634456,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.578325639819438e-05,
+      "loss": 0.5063,
+      "step": 4328
+    },
+    {
+      "epoch": 0.19012700747554803,
+      "grad_norm": 3.046875,
+      "learning_rate": 4.5779410567435264e-05,
+      "loss": 0.5193,
+      "step": 4330
+    },
+    {
+      "epoch": 0.19021482595475153,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.577556314538391e-05,
+      "loss": 0.5139,
+      "step": 4332
+    },
+    {
+      "epoch": 0.190302644433955,
+      "grad_norm": 2.28125,
+      "learning_rate": 4.5771714132334945e-05,
+      "loss": 0.5132,
+      "step": 4334
+    },
+    {
+      "epoch": 0.1903904629131585,
+      "grad_norm": 2.46875,
+      "learning_rate": 4.576786352858313e-05,
+      "loss": 0.5337,
+      "step": 4336
+    },
+    {
+      "epoch": 0.19047828139236198,
+      "grad_norm": 2.234375,
+      "learning_rate": 4.5764011334423354e-05,
+      "loss": 0.5202,
+      "step": 4338
+    },
+    {
+      "epoch": 0.19056609987156548,
+      "grad_norm": 2.484375,
+      "learning_rate": 4.576015755015061e-05,
+      "loss": 0.5288,
+      "step": 4340
+    },
+    {
+      "epoch": 0.19065391835076895,
+      "grad_norm": 2.625,
+      "learning_rate": 4.575630217606002e-05,
+      "loss": 0.5126,
+      "step": 4342
+    },
+    {
+      "epoch": 0.19074173682997245,
+      "grad_norm": 2.359375,
+      "learning_rate": 4.5752445212446836e-05,
+      "loss": 0.5032,
+      "step": 4344
+    },
+    {
+      "epoch": 0.19082955530917595,
+      "grad_norm": 2.390625,
+      "learning_rate": 4.574858665960643e-05,
+      "loss": 0.5159,
+      "step": 4346
+    },
+    {
+      "epoch": 0.19091737378837942,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.574472651783428e-05,
+      "loss": 0.5548,
+      "step": 4348
+    },
+    {
+      "epoch": 0.19100519226758292,
+      "grad_norm": 2.96875,
+      "learning_rate": 4.5740864787425995e-05,
+      "loss": 0.4725,
+      "step": 4350
+    },
+    {
+      "epoch": 0.1910930107467864,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.573700146867732e-05,
+      "loss": 0.4993,
+      "step": 4352
+    },
+    {
+      "epoch": 0.1911808292259899,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.573313656188409e-05,
+      "loss": 0.5261,
+      "step": 4354
+    },
+    {
+      "epoch": 0.19126864770519336,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.57292700673423e-05,
+      "loss": 0.5127,
+      "step": 4356
+    },
+    {
+      "epoch": 0.19135646618439686,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.572540198534804e-05,
+      "loss": 0.5066,
+      "step": 4358
+    },
+    {
+      "epoch": 0.19144428466360033,
+      "grad_norm": 2.40625,
+      "learning_rate": 4.572153231619752e-05,
+      "loss": 0.5019,
+      "step": 4360
+    },
+    {
+      "epoch": 0.19153210314280383,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.57176610601871e-05,
+      "loss": 0.4771,
+      "step": 4362
+    },
+    {
+      "epoch": 0.1916199216220073,
+      "grad_norm": 2.1875,
+      "learning_rate": 4.571378821761322e-05,
+      "loss": 0.525,
+      "step": 4364
+    },
+    {
+      "epoch": 0.1917077401012108,
+      "grad_norm": 2.296875,
+      "learning_rate": 4.5709913788772486e-05,
+      "loss": 0.5299,
+      "step": 4366
+    },
+    {
+      "epoch": 0.19179555858041428,
+      "grad_norm": 2.953125,
+      "learning_rate": 4.570603777396158e-05,
+      "loss": 0.5113,
+      "step": 4368
+    },
+    {
+      "epoch": 0.19188337705961778,
+      "grad_norm": 4.09375,
+      "learning_rate": 4.5702160173477336e-05,
+      "loss": 0.4981,
+      "step": 4370
+    },
+    {
+      "epoch": 0.19197119553882125,
+      "grad_norm": 4.59375,
+      "learning_rate": 4.56982809876167e-05,
+      "loss": 0.4895,
+      "step": 4372
+    },
+    {
+      "epoch": 0.19205901401802475,
+      "grad_norm": 4.03125,
+      "learning_rate": 4.5694400216676755e-05,
+      "loss": 0.4936,
+      "step": 4374
+    },
+    {
+      "epoch": 0.19214683249722822,
+      "grad_norm": 3.234375,
+      "learning_rate": 4.569051786095467e-05,
+      "loss": 0.5052,
+      "step": 4376
+    },
+    {
+      "epoch": 0.19223465097643172,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.568663392074778e-05,
+      "loss": 0.4914,
+      "step": 4378
+    },
+    {
+      "epoch": 0.1923224694556352,
+      "grad_norm": 2.4375,
+      "learning_rate": 4.568274839635349e-05,
+      "loss": 0.4874,
+      "step": 4380
+    },
+    {
+      "epoch": 0.1924102879348387,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.567886128806939e-05,
+      "loss": 0.5129,
+      "step": 4382
+    },
+    {
+      "epoch": 0.19249810641404216,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.5674972596193114e-05,
+      "loss": 0.5016,
+      "step": 4384
+    },
+    {
+      "epoch": 0.19258592489324566,
+      "grad_norm": 3.0625,
+      "learning_rate": 4.567108232102249e-05,
+      "loss": 0.5084,
+      "step": 4386
+    },
+    {
+      "epoch": 0.19267374337244914,
+      "grad_norm": 2.96875,
+      "learning_rate": 4.566719046285542e-05,
+      "loss": 0.5344,
+      "step": 4388
+    },
+    {
+      "epoch": 0.19276156185165264,
+      "grad_norm": 2.625,
+      "learning_rate": 4.5663297021989963e-05,
+      "loss": 0.4685,
+      "step": 4390
+    },
+    {
+      "epoch": 0.1928493803308561,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.565940199872426e-05,
+      "loss": 0.5278,
+      "step": 4392
+    },
+    {
+      "epoch": 0.1929371988100596,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.56555053933566e-05,
+      "loss": 0.4925,
+      "step": 4394
+    },
+    {
+      "epoch": 0.1930250172892631,
+      "grad_norm": 3.078125,
+      "learning_rate": 4.5651607206185384e-05,
+      "loss": 0.5183,
+      "step": 4396
+    },
+    {
+      "epoch": 0.19311283576846658,
+      "grad_norm": 2.984375,
+      "learning_rate": 4.564770743750914e-05,
+      "loss": 0.4626,
+      "step": 4398
+    },
+    {
+      "epoch": 0.19320065424767008,
+      "grad_norm": 3.453125,
+      "learning_rate": 4.564380608762651e-05,
+      "loss": 0.4903,
+      "step": 4400
+    },
+    {
+      "epoch": 0.19328847272687355,
+      "grad_norm": 3.984375,
+      "learning_rate": 4.563990315683626e-05,
+      "loss": 0.4885,
+      "step": 4402
+    },
+    {
+      "epoch": 0.19337629120607705,
+      "grad_norm": 3.25,
+      "learning_rate": 4.5635998645437286e-05,
+      "loss": 0.5027,
+      "step": 4404
+    },
+    {
+      "epoch": 0.19346410968528052,
+      "grad_norm": 3.140625,
+      "learning_rate": 4.5632092553728576e-05,
+      "loss": 0.4832,
+      "step": 4406
+    },
+    {
+      "epoch": 0.19355192816448402,
+      "grad_norm": 3.34375,
+      "learning_rate": 4.5628184882009276e-05,
+      "loss": 0.514,
+      "step": 4408
+    },
+    {
+      "epoch": 0.1936397466436875,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.562427563057864e-05,
+      "loss": 0.5072,
+      "step": 4410
+    },
+    {
+      "epoch": 0.193727565122891,
+      "grad_norm": 2.515625,
+      "learning_rate": 4.5620364799736026e-05,
+      "loss": 0.4741,
+      "step": 4412
+    },
+    {
+      "epoch": 0.19381538360209447,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.561645238978094e-05,
+      "loss": 0.4841,
+      "step": 4414
+    },
+    {
+      "epoch": 0.19390320208129797,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.561253840101298e-05,
+      "loss": 0.5125,
+      "step": 4416
+    },
+    {
+      "epoch": 0.19399102056050144,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.560862283373188e-05,
+      "loss": 0.4877,
+      "step": 4418
+    },
+    {
+      "epoch": 0.19407883903970494,
+      "grad_norm": 3.21875,
+      "learning_rate": 4.560470568823753e-05,
+      "loss": 0.5164,
+      "step": 4420
+    },
+    {
+      "epoch": 0.1941666575189084,
+      "grad_norm": 3.328125,
+      "learning_rate": 4.5600786964829854e-05,
+      "loss": 0.481,
+      "step": 4422
+    },
+    {
+      "epoch": 0.1942544759981119,
+      "grad_norm": 3.15625,
+      "learning_rate": 4.559686666380898e-05,
+      "loss": 0.4857,
+      "step": 4424
+    },
+    {
+      "epoch": 0.19434229447731538,
+      "grad_norm": 3.546875,
+      "learning_rate": 4.559294478547512e-05,
+      "loss": 0.4905,
+      "step": 4426
+    },
+    {
+      "epoch": 0.19443011295651888,
+      "grad_norm": 3.96875,
+      "learning_rate": 4.5589021330128615e-05,
+      "loss": 0.512,
+      "step": 4428
+    },
+    {
+      "epoch": 0.19451793143572235,
+      "grad_norm": 3.140625,
+      "learning_rate": 4.558509629806992e-05,
+      "loss": 0.5058,
+      "step": 4430
+    },
+    {
+      "epoch": 0.19460574991492585,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.558116968959962e-05,
+      "loss": 0.4877,
+      "step": 4432
+    },
+    {
+      "epoch": 0.19469356839412932,
+      "grad_norm": 2.5,
+      "learning_rate": 4.5577241505018404e-05,
+      "loss": 0.4911,
+      "step": 4434
+    },
+    {
+      "epoch": 0.19478138687333282,
+      "grad_norm": 2.546875,
+      "learning_rate": 4.557331174462711e-05,
+      "loss": 0.5044,
+      "step": 4436
+    },
+    {
+      "epoch": 0.1948692053525363,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.556938040872666e-05,
+      "loss": 0.5064,
+      "step": 4438
+    },
+    {
+      "epoch": 0.1949570238317398,
+      "grad_norm": 2.4375,
+      "learning_rate": 4.556544749761813e-05,
+      "loss": 0.4767,
+      "step": 4440
+    },
+    {
+      "epoch": 0.19504484231094327,
+      "grad_norm": 2.265625,
+      "learning_rate": 4.556151301160271e-05,
+      "loss": 0.5161,
+      "step": 4442
+    },
+    {
+      "epoch": 0.19513266079014677,
+      "grad_norm": 2.28125,
+      "learning_rate": 4.5557576950981684e-05,
+      "loss": 0.5014,
+      "step": 4444
+    },
+    {
+      "epoch": 0.19522047926935027,
+      "grad_norm": 2.46875,
+      "learning_rate": 4.555363931605649e-05,
+      "loss": 0.4857,
+      "step": 4446
+    },
+    {
+      "epoch": 0.19530829774855374,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.554970010712867e-05,
+      "loss": 0.5094,
+      "step": 4448
+    },
+    {
+      "epoch": 0.19539611622775724,
+      "grad_norm": 2.265625,
+      "learning_rate": 4.5545759324499896e-05,
+      "loss": 0.5082,
+      "step": 4450
+    },
+    {
+      "epoch": 0.1954839347069607,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.554181696847194e-05,
+      "loss": 0.515,
+      "step": 4452
+    },
+    {
+      "epoch": 0.1955717531861642,
+      "grad_norm": 2.3125,
+      "learning_rate": 4.5537873039346716e-05,
+      "loss": 0.5094,
+      "step": 4454
+    },
+    {
+      "epoch": 0.19565957166536768,
+      "grad_norm": 2.390625,
+      "learning_rate": 4.553392753742626e-05,
+      "loss": 0.5027,
+      "step": 4456
+    },
+    {
+      "epoch": 0.19574739014457118,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.55299804630127e-05,
+      "loss": 0.5025,
+      "step": 4458
+    },
+    {
+      "epoch": 0.19583520862377465,
+      "grad_norm": 2.546875,
+      "learning_rate": 4.5526031816408324e-05,
+      "loss": 0.4915,
+      "step": 4460
+    },
+    {
+      "epoch": 0.19592302710297815,
+      "grad_norm": 2.75,
+      "learning_rate": 4.55220815979155e-05,
+      "loss": 0.5053,
+      "step": 4462
+    },
+    {
+      "epoch": 0.19601084558218163,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.5518129807836755e-05,
+      "loss": 0.5056,
+      "step": 4464
+    },
+    {
+      "epoch": 0.19609866406138513,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.551417644647471e-05,
+      "loss": 0.476,
+      "step": 4466
+    },
+    {
+      "epoch": 0.1961864825405886,
+      "grad_norm": 2.3125,
+      "learning_rate": 4.551022151413212e-05,
+      "loss": 0.5206,
+      "step": 4468
+    },
+    {
+      "epoch": 0.1962743010197921,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.5506265011111845e-05,
+      "loss": 0.4979,
+      "step": 4470
+    },
+    {
+      "epoch": 0.19636211949899557,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.5502306937716873e-05,
+      "loss": 0.5007,
+      "step": 4472
+    },
+    {
+      "epoch": 0.19644993797819907,
+      "grad_norm": 2.96875,
+      "learning_rate": 4.5498347294250335e-05,
+      "loss": 0.4995,
+      "step": 4474
+    },
+    {
+      "epoch": 0.19653775645740254,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.549438608101544e-05,
+      "loss": 0.5034,
+      "step": 4476
+    },
+    {
+      "epoch": 0.19662557493660604,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.549042329831554e-05,
+      "loss": 0.5002,
+      "step": 4478
+    },
+    {
+      "epoch": 0.1967133934158095,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.548645894645413e-05,
+      "loss": 0.5025,
+      "step": 4480
+    },
+    {
+      "epoch": 0.196801211895013,
+      "grad_norm": 2.96875,
+      "learning_rate": 4.548249302573477e-05,
+      "loss": 0.5033,
+      "step": 4482
+    },
+    {
+      "epoch": 0.19688903037421648,
+      "grad_norm": 2.75,
+      "learning_rate": 4.547852553646118e-05,
+      "loss": 0.4976,
+      "step": 4484
+    },
+    {
+      "epoch": 0.19697684885341998,
+      "grad_norm": 2.375,
+      "learning_rate": 4.54745564789372e-05,
+      "loss": 0.4862,
+      "step": 4486
+    },
+    {
+      "epoch": 0.19706466733262346,
+      "grad_norm": 2.375,
+      "learning_rate": 4.547058585346678e-05,
+      "loss": 0.4548,
+      "step": 4488
+    },
+    {
+      "epoch": 0.19715248581182696,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.5466613660353985e-05,
+      "loss": 0.5254,
+      "step": 4490
+    },
+    {
+      "epoch": 0.19724030429103043,
+      "grad_norm": 2.234375,
+      "learning_rate": 4.546263989990302e-05,
+      "loss": 0.4946,
+      "step": 4492
+    },
+    {
+      "epoch": 0.19732812277023393,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.545866457241817e-05,
+      "loss": 0.4971,
+      "step": 4494
+    },
+    {
+      "epoch": 0.19741594124943743,
+      "grad_norm": 2.625,
+      "learning_rate": 4.545468767820389e-05,
+      "loss": 0.4876,
+      "step": 4496
+    },
+    {
+      "epoch": 0.1975037597286409,
+      "grad_norm": 3.015625,
+      "learning_rate": 4.545070921756472e-05,
+      "loss": 0.5235,
+      "step": 4498
+    },
+    {
+      "epoch": 0.1975915782078444,
+      "grad_norm": 3.0625,
+      "learning_rate": 4.5446729190805345e-05,
+      "loss": 0.501,
+      "step": 4500
+    },
+    {
+      "epoch": 0.19767939668704787,
+      "grad_norm": 3.015625,
+      "learning_rate": 4.544274759823054e-05,
+      "loss": 0.5097,
+      "step": 4502
+    },
+    {
+      "epoch": 0.19776721516625137,
+      "grad_norm": 2.375,
+      "learning_rate": 4.5438764440145225e-05,
+      "loss": 0.4899,
+      "step": 4504
+    },
+    {
+      "epoch": 0.19785503364545484,
+      "grad_norm": 2.296875,
+      "learning_rate": 4.543477971685443e-05,
+      "loss": 0.4941,
+      "step": 4506
+    },
+    {
+      "epoch": 0.19794285212465834,
+      "grad_norm": 2.625,
+      "learning_rate": 4.543079342866331e-05,
+      "loss": 0.5089,
+      "step": 4508
+    },
+    {
+      "epoch": 0.1980306706038618,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.542680557587712e-05,
+      "loss": 0.4812,
+      "step": 4510
+    },
+    {
+      "epoch": 0.1981184890830653,
+      "grad_norm": 2.390625,
+      "learning_rate": 4.5422816158801275e-05,
+      "loss": 0.4979,
+      "step": 4512
+    },
+    {
+      "epoch": 0.19820630756226879,
+      "grad_norm": 2.4375,
+      "learning_rate": 4.5418825177741266e-05,
+      "loss": 0.4836,
+      "step": 4514
+    },
+    {
+      "epoch": 0.19829412604147229,
+      "grad_norm": 2.53125,
+      "learning_rate": 4.541483263300273e-05,
+      "loss": 0.5024,
+      "step": 4516
+    },
+    {
+      "epoch": 0.19838194452067576,
+      "grad_norm": 2.28125,
+      "learning_rate": 4.5410838524891416e-05,
+      "loss": 0.4997,
+      "step": 4518
+    },
+    {
+      "epoch": 0.19846976299987926,
+      "grad_norm": 2.4375,
+      "learning_rate": 4.540684285371321e-05,
+      "loss": 0.5085,
+      "step": 4520
+    },
+    {
+      "epoch": 0.19855758147908273,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.540284561977406e-05,
+      "loss": 0.4939,
+      "step": 4522
+    },
+    {
+      "epoch": 0.19864539995828623,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.5398846823380124e-05,
+      "loss": 0.5064,
+      "step": 4524
+    },
+    {
+      "epoch": 0.1987332184374897,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.53948464648376e-05,
+      "loss": 0.462,
+      "step": 4526
+    },
+    {
+      "epoch": 0.1988210369166932,
+      "grad_norm": 2.375,
+      "learning_rate": 4.539084454445284e-05,
+      "loss": 0.492,
+      "step": 4528
+    },
+    {
+      "epoch": 0.19890885539589667,
+      "grad_norm": 2.359375,
+      "learning_rate": 4.5386841062532325e-05,
+      "loss": 0.4829,
+      "step": 4530
+    },
+    {
+      "epoch": 0.19899667387510017,
+      "grad_norm": 2.421875,
+      "learning_rate": 4.538283601938263e-05,
+      "loss": 0.4771,
+      "step": 4532
+    },
+    {
+      "epoch": 0.19908449235430364,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.5378829415310465e-05,
+      "loss": 0.5275,
+      "step": 4534
+    },
+    {
+      "epoch": 0.19917231083350714,
+      "grad_norm": 2.5,
+      "learning_rate": 4.5374821250622665e-05,
+      "loss": 0.5158,
+      "step": 4536
+    },
+    {
+      "epoch": 0.19926012931271062,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.5370811525626166e-05,
+      "loss": 0.5223,
+      "step": 4538
+    },
+    {
+      "epoch": 0.19934794779191412,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.5366800240628034e-05,
+      "loss": 0.5189,
+      "step": 4540
+    },
+    {
+      "epoch": 0.1994357662711176,
+      "grad_norm": 2.40625,
+      "learning_rate": 4.536278739593546e-05,
+      "loss": 0.5121,
+      "step": 4542
+    },
+    {
+      "epoch": 0.1995235847503211,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.535877299185575e-05,
+      "loss": 0.5178,
+      "step": 4544
+    },
+    {
+      "epoch": 0.19961140322952456,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.5354757028696315e-05,
+      "loss": 0.5079,
+      "step": 4546
+    },
+    {
+      "epoch": 0.19969922170872806,
+      "grad_norm": 2.3125,
+      "learning_rate": 4.5350739506764714e-05,
+      "loss": 0.4885,
+      "step": 4548
+    },
+    {
+      "epoch": 0.19978704018793156,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.53467204263686e-05,
+      "loss": 0.5099,
+      "step": 4550
+    },
+    {
+      "epoch": 0.19987485866713503,
+      "grad_norm": 2.46875,
+      "learning_rate": 4.5342699787815755e-05,
+      "loss": 0.4823,
+      "step": 4552
+    },
+    {
+      "epoch": 0.19996267714633853,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.5338677591414095e-05,
+      "loss": 0.4902,
+      "step": 4554
+    },
+    {
+      "epoch": 0.200050495625542,
+      "grad_norm": 2.3125,
+      "learning_rate": 4.533465383747162e-05,
+      "loss": 0.5072,
+      "step": 4556
+    },
+    {
+      "epoch": 0.2001383141047455,
+      "grad_norm": 2.40625,
+      "learning_rate": 4.533062852629648e-05,
+      "loss": 0.5244,
+      "step": 4558
+    },
+    {
+      "epoch": 0.20022613258394897,
+      "grad_norm": 2.265625,
+      "learning_rate": 4.5326601658196937e-05,
+      "loss": 0.5189,
+      "step": 4560
+    },
+    {
+      "epoch": 0.20031395106315247,
+      "grad_norm": 2.53125,
+      "learning_rate": 4.532257323348137e-05,
+      "loss": 0.5431,
+      "step": 4562
+    },
+    {
+      "epoch": 0.20040176954235595,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.5318543252458266e-05,
+      "loss": 0.4636,
+      "step": 4564
+    },
+    {
+      "epoch": 0.20048958802155945,
+      "grad_norm": 2.484375,
+      "learning_rate": 4.5314511715436246e-05,
+      "loss": 0.4813,
+      "step": 4566
+    },
+    {
+      "epoch": 0.20057740650076292,
+      "grad_norm": 3.140625,
+      "learning_rate": 4.531047862272406e-05,
+      "loss": 0.5049,
+      "step": 4568
+    },
+    {
+      "epoch": 0.20066522497996642,
+      "grad_norm": 2.296875,
+      "learning_rate": 4.530644397463055e-05,
+      "loss": 0.5004,
+      "step": 4570
+    },
+    {
+      "epoch": 0.2007530434591699,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.530240777146468e-05,
+      "loss": 0.4946,
+      "step": 4572
+    },
+    {
+      "epoch": 0.2008408619383734,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.529837001353557e-05,
+      "loss": 0.4981,
+      "step": 4574
+    },
+    {
+      "epoch": 0.20092868041757686,
+      "grad_norm": 2.296875,
+      "learning_rate": 4.529433070115241e-05,
+      "loss": 0.5206,
+      "step": 4576
+    },
+    {
+      "epoch": 0.20101649889678036,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.529028983462455e-05,
+      "loss": 0.5008,
+      "step": 4578
+    },
+    {
+      "epoch": 0.20110431737598383,
+      "grad_norm": 2.40625,
+      "learning_rate": 4.528624741426143e-05,
+      "loss": 0.4744,
+      "step": 4580
+    },
+    {
+      "epoch": 0.20119213585518733,
+      "grad_norm": 2.515625,
+      "learning_rate": 4.528220344037261e-05,
+      "loss": 0.5024,
+      "step": 4582
+    },
+    {
+      "epoch": 0.2012799543343908,
+      "grad_norm": 3.0625,
+      "learning_rate": 4.527815791326779e-05,
+      "loss": 0.5053,
+      "step": 4584
+    },
+    {
+      "epoch": 0.2013677728135943,
+      "grad_norm": 2.625,
+      "learning_rate": 4.5274110833256785e-05,
+      "loss": 0.5126,
+      "step": 4586
+    },
+    {
+      "epoch": 0.20145559129279778,
+      "grad_norm": 2.5,
+      "learning_rate": 4.52700622006495e-05,
+      "loss": 0.4648,
+      "step": 4588
+    },
+    {
+      "epoch": 0.20154340977200128,
+      "grad_norm": 2.53125,
+      "learning_rate": 4.5266012015756e-05,
+      "loss": 0.5069,
+      "step": 4590
+    },
+    {
+      "epoch": 0.20163122825120475,
+      "grad_norm": 2.53125,
+      "learning_rate": 4.5261960278886443e-05,
+      "loss": 0.5303,
+      "step": 4592
+    },
+    {
+      "epoch": 0.20171904673040825,
+      "grad_norm": 3.015625,
+      "learning_rate": 4.5257906990351114e-05,
+      "loss": 0.5173,
+      "step": 4594
+    },
+    {
+      "epoch": 0.20180686520961172,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.5253852150460406e-05,
+      "loss": 0.4768,
+      "step": 4596
+    },
+    {
+      "epoch": 0.20189468368881522,
+      "grad_norm": 2.234375,
+      "learning_rate": 4.524979575952484e-05,
+      "loss": 0.4905,
+      "step": 4598
+    },
+    {
+      "epoch": 0.20198250216801872,
+      "grad_norm": 2.375,
+      "learning_rate": 4.524573781785507e-05,
+      "loss": 0.4856,
+      "step": 4600
+    },
+    {
+      "epoch": 0.2020703206472222,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.5241678325761835e-05,
+      "loss": 0.5001,
+      "step": 4602
+    },
+    {
+      "epoch": 0.2021581391264257,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.523761728355603e-05,
+      "loss": 0.5012,
+      "step": 4604
+    },
+    {
+      "epoch": 0.20224595760562916,
+      "grad_norm": 2.5,
+      "learning_rate": 4.523355469154864e-05,
+      "loss": 0.497,
+      "step": 4606
+    },
+    {
+      "epoch": 0.20233377608483266,
+      "grad_norm": 2.484375,
+      "learning_rate": 4.522949055005078e-05,
+      "loss": 0.5138,
+      "step": 4608
+    },
+    {
+      "epoch": 0.20242159456403613,
+      "grad_norm": 2.515625,
+      "learning_rate": 4.522542485937369e-05,
+      "loss": 0.506,
+      "step": 4610
+    },
+    {
+      "epoch": 0.20250941304323963,
+      "grad_norm": 2.3125,
+      "learning_rate": 4.522135761982871e-05,
+      "loss": 0.5185,
+      "step": 4612
+    },
+    {
+      "epoch": 0.2025972315224431,
+      "grad_norm": 2.515625,
+      "learning_rate": 4.521728883172732e-05,
+      "loss": 0.4962,
+      "step": 4614
+    },
+    {
+      "epoch": 0.2026850500016466,
+      "grad_norm": 2.390625,
+      "learning_rate": 4.52132184953811e-05,
+      "loss": 0.4924,
+      "step": 4616
+    },
+    {
+      "epoch": 0.20277286848085008,
+      "grad_norm": 2.453125,
+      "learning_rate": 4.520914661110176e-05,
+      "loss": 0.4647,
+      "step": 4618
+    },
+    {
+      "epoch": 0.20286068696005358,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.5205073179201134e-05,
+      "loss": 0.4827,
+      "step": 4620
+    },
+    {
+      "epoch": 0.20294850543925705,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.5200998199991165e-05,
+      "loss": 0.4757,
+      "step": 4622
+    },
+    {
+      "epoch": 0.20303632391846055,
+      "grad_norm": 2.5,
+      "learning_rate": 4.5196921673783896e-05,
+      "loss": 0.5093,
+      "step": 4624
+    },
+    {
+      "epoch": 0.20312414239766402,
+      "grad_norm": 2.3125,
+      "learning_rate": 4.519284360089153e-05,
+      "loss": 0.5032,
+      "step": 4626
+    },
+    {
+      "epoch": 0.20321196087686752,
+      "grad_norm": 2.484375,
+      "learning_rate": 4.518876398162637e-05,
+      "loss": 0.5029,
+      "step": 4628
+    },
+    {
+      "epoch": 0.203299779356071,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.5184682816300815e-05,
+      "loss": 0.4898,
+      "step": 4630
+    },
+    {
+      "epoch": 0.2033875978352745,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.518060010522742e-05,
+      "loss": 0.4841,
+      "step": 4632
+    },
+    {
+      "epoch": 0.20347541631447796,
+      "grad_norm": 2.484375,
+      "learning_rate": 4.5176515848718825e-05,
+      "loss": 0.4545,
+      "step": 4634
+    },
+    {
+      "epoch": 0.20356323479368146,
+      "grad_norm": 2.21875,
+      "learning_rate": 4.517243004708781e-05,
+      "loss": 0.4862,
+      "step": 4636
+    },
+    {
+      "epoch": 0.20365105327288494,
+      "grad_norm": 2.484375,
+      "learning_rate": 4.5168342700647264e-05,
+      "loss": 0.4902,
+      "step": 4638
+    },
+    {
+      "epoch": 0.20373887175208844,
+      "grad_norm": 2.515625,
+      "learning_rate": 4.51642538097102e-05,
+      "loss": 0.478,
+      "step": 4640
+    },
+    {
+      "epoch": 0.2038266902312919,
+      "grad_norm": 2.234375,
+      "learning_rate": 4.516016337458975e-05,
+      "loss": 0.5129,
+      "step": 4642
+    },
+    {
+      "epoch": 0.2039145087104954,
+      "grad_norm": 2.5,
+      "learning_rate": 4.5156071395599156e-05,
+      "loss": 0.4711,
+      "step": 4644
+    },
+    {
+      "epoch": 0.20400232718969888,
+      "grad_norm": 2.453125,
+      "learning_rate": 4.5151977873051774e-05,
+      "loss": 0.4948,
+      "step": 4646
+    },
+    {
+      "epoch": 0.20409014566890238,
+      "grad_norm": 2.390625,
+      "learning_rate": 4.5147882807261104e-05,
+      "loss": 0.4951,
+      "step": 4648
+    },
+    {
+      "epoch": 0.20417796414810588,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.5143786198540736e-05,
+      "loss": 0.4779,
+      "step": 4650
+    },
+    {
+      "epoch": 0.20426578262730935,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.51396880472044e-05,
+      "loss": 0.518,
+      "step": 4652
+    },
+    {
+      "epoch": 0.20435360110651285,
+      "grad_norm": 2.96875,
+      "learning_rate": 4.5135588353565914e-05,
+      "loss": 0.491,
+      "step": 4654
+    },
+    {
+      "epoch": 0.20444141958571632,
+      "grad_norm": 3.453125,
+      "learning_rate": 4.5131487117939246e-05,
+      "loss": 0.5192,
+      "step": 4656
+    },
+    {
+      "epoch": 0.20452923806491982,
+      "grad_norm": 2.75,
+      "learning_rate": 4.5127384340638474e-05,
+      "loss": 0.4967,
+      "step": 4658
+    },
+    {
+      "epoch": 0.2046170565441233,
+      "grad_norm": 2.359375,
+      "learning_rate": 4.512328002197779e-05,
+      "loss": 0.5179,
+      "step": 4660
+    },
+    {
+      "epoch": 0.2047048750233268,
+      "grad_norm": 2.34375,
+      "learning_rate": 4.5119174162271484e-05,
+      "loss": 0.4966,
+      "step": 4662
+    },
+    {
+      "epoch": 0.20479269350253027,
+      "grad_norm": 2.421875,
+      "learning_rate": 4.5115066761834e-05,
+      "loss": 0.5001,
+      "step": 4664
+    },
+    {
+      "epoch": 0.20488051198173376,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.511095782097989e-05,
+      "loss": 0.5273,
+      "step": 4666
+    },
+    {
+      "epoch": 0.20496833046093724,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.510684734002381e-05,
+      "loss": 0.4918,
+      "step": 4668
+    },
+    {
+      "epoch": 0.20505614894014074,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.510273531928053e-05,
+      "loss": 0.4928,
+      "step": 4670
+    },
+    {
+      "epoch": 0.2051439674193442,
+      "grad_norm": 3.265625,
+      "learning_rate": 4.5098621759064964e-05,
+      "loss": 0.4911,
+      "step": 4672
+    },
+    {
+      "epoch": 0.2052317858985477,
+      "grad_norm": 3.109375,
+      "learning_rate": 4.5094506659692126e-05,
+      "loss": 0.5041,
+      "step": 4674
+    },
+    {
+      "epoch": 0.20531960437775118,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.509039002147716e-05,
+      "loss": 0.4829,
+      "step": 4676
+    },
+    {
+      "epoch": 0.20540742285695468,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.50862718447353e-05,
+      "loss": 0.4726,
+      "step": 4678
+    },
+    {
+      "epoch": 0.20549524133615815,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.508215212978193e-05,
+      "loss": 0.4512,
+      "step": 4680
+    },
+    {
+      "epoch": 0.20558305981536165,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.507803087693253e-05,
+      "loss": 0.497,
+      "step": 4682
+    },
+    {
+      "epoch": 0.20567087829456512,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.507390808650272e-05,
+      "loss": 0.4954,
+      "step": 4684
+    },
+    {
+      "epoch": 0.20575869677376862,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.506978375880822e-05,
+      "loss": 0.4796,
+      "step": 4686
+    },
+    {
+      "epoch": 0.2058465152529721,
+      "grad_norm": 2.5,
+      "learning_rate": 4.5065657894164856e-05,
+      "loss": 0.478,
+      "step": 4688
+    },
+    {
+      "epoch": 0.2059343337321756,
+      "grad_norm": 2.28125,
+      "learning_rate": 4.506153049288861e-05,
+      "loss": 0.5176,
+      "step": 4690
+    },
+    {
+      "epoch": 0.20602215221137907,
+      "grad_norm": 2.625,
+      "learning_rate": 4.505740155529555e-05,
+      "loss": 0.5044,
+      "step": 4692
+    },
+    {
+      "epoch": 0.20610997069058257,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.505327108170187e-05,
+      "loss": 0.5094,
+      "step": 4694
+    },
+    {
+      "epoch": 0.20619778916978604,
+      "grad_norm": 2.359375,
+      "learning_rate": 4.504913907242389e-05,
+      "loss": 0.4767,
+      "step": 4696
+    },
+    {
+      "epoch": 0.20628560764898954,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.504500552777802e-05,
+      "loss": 0.4734,
+      "step": 4698
+    },
+    {
+      "epoch": 0.20637342612819304,
+      "grad_norm": 2.234375,
+      "learning_rate": 4.5040870448080824e-05,
+      "loss": 0.4629,
+      "step": 4700
+    },
+    {
+      "epoch": 0.2064612446073965,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.503673383364897e-05,
+      "loss": 0.499,
+      "step": 4702
+    },
+    {
+      "epoch": 0.2065490630866,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.5032595684799236e-05,
+      "loss": 0.5027,
+      "step": 4704
+    },
+    {
+      "epoch": 0.20663688156580348,
+      "grad_norm": 2.359375,
+      "learning_rate": 4.502845600184852e-05,
+      "loss": 0.5164,
+      "step": 4706
+    },
+    {
+      "epoch": 0.20672470004500698,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.502431478511384e-05,
+      "loss": 0.4971,
+      "step": 4708
+    },
+    {
+      "epoch": 0.20681251852421045,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.5020172034912336e-05,
+      "loss": 0.448,
+      "step": 4710
+    },
+    {
+      "epoch": 0.20690033700341395,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.501602775156126e-05,
+      "loss": 0.4835,
+      "step": 4712
+    },
+    {
+      "epoch": 0.20698815548261743,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.501188193537798e-05,
+      "loss": 0.4712,
+      "step": 4714
+    },
+    {
+      "epoch": 0.20707597396182092,
+      "grad_norm": 3.03125,
+      "learning_rate": 4.500773458667998e-05,
+      "loss": 0.492,
+      "step": 4716
+    },
+    {
+      "epoch": 0.2071637924410244,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.500358570578488e-05,
+      "loss": 0.4948,
+      "step": 4718
+    },
+    {
+      "epoch": 0.2072516109202279,
+      "grad_norm": 2.359375,
+      "learning_rate": 4.4999435293010384e-05,
+      "loss": 0.458,
+      "step": 4720
+    },
+    {
+      "epoch": 0.20733942939943137,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.4995283348674336e-05,
+      "loss": 0.4802,
+      "step": 4722
+    },
+    {
+      "epoch": 0.20742724787863487,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.49911298730947e-05,
+      "loss": 0.4884,
+      "step": 4724
+    },
+    {
+      "epoch": 0.20751506635783834,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.4986974866589546e-05,
+      "loss": 0.476,
+      "step": 4726
+    },
+    {
+      "epoch": 0.20760288483704184,
+      "grad_norm": 2.46875,
+      "learning_rate": 4.498281832947706e-05,
+      "loss": 0.5139,
+      "step": 4728
+    },
+    {
+      "epoch": 0.2076907033162453,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.4978660262075566e-05,
+      "loss": 0.4709,
+      "step": 4730
+    },
+    {
+      "epoch": 0.2077785217954488,
+      "grad_norm": 3.0625,
+      "learning_rate": 4.4974500664703475e-05,
+      "loss": 0.5018,
+      "step": 4732
+    },
+    {
+      "epoch": 0.20786634027465228,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.4970339537679336e-05,
+      "loss": 0.4881,
+      "step": 4734
+    },
+    {
+      "epoch": 0.20795415875385578,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.49661768813218e-05,
+      "loss": 0.5178,
+      "step": 4736
+    },
+    {
+      "epoch": 0.20804197723305926,
+      "grad_norm": 2.375,
+      "learning_rate": 4.496201269594966e-05,
+      "loss": 0.5246,
+      "step": 4738
+    },
+    {
+      "epoch": 0.20812979571226276,
+      "grad_norm": 2.390625,
+      "learning_rate": 4.49578469818818e-05,
+      "loss": 0.4945,
+      "step": 4740
+    },
+    {
+      "epoch": 0.20821761419146623,
+      "grad_norm": 2.453125,
+      "learning_rate": 4.495367973943724e-05,
+      "loss": 0.5048,
+      "step": 4742
+    },
+    {
+      "epoch": 0.20830543267066973,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.494951096893509e-05,
+      "loss": 0.4675,
+      "step": 4744
+    },
+    {
+      "epoch": 0.2083932511498732,
+      "grad_norm": 3.234375,
+      "learning_rate": 4.494534067069463e-05,
+      "loss": 0.5013,
+      "step": 4746
+    },
+    {
+      "epoch": 0.2084810696290767,
+      "grad_norm": 3.171875,
+      "learning_rate": 4.494116884503518e-05,
+      "loss": 0.4897,
+      "step": 4748
+    },
+    {
+      "epoch": 0.2085688881082802,
+      "grad_norm": 2.296875,
+      "learning_rate": 4.493699549227624e-05,
+      "loss": 0.5059,
+      "step": 4750
+    },
+    {
+      "epoch": 0.20865670658748367,
+      "grad_norm": 2.515625,
+      "learning_rate": 4.493282061273741e-05,
+      "loss": 0.5089,
+      "step": 4752
+    },
+    {
+      "epoch": 0.20874452506668717,
+      "grad_norm": 2.125,
+      "learning_rate": 4.4928644206738404e-05,
+      "loss": 0.4795,
+      "step": 4754
+    },
+    {
+      "epoch": 0.20883234354589064,
+      "grad_norm": 2.3125,
+      "learning_rate": 4.492446627459904e-05,
+      "loss": 0.4664,
+      "step": 4756
+    },
+    {
+      "epoch": 0.20892016202509414,
+      "grad_norm": 2.515625,
+      "learning_rate": 4.4920286816639277e-05,
+      "loss": 0.5075,
+      "step": 4758
+    },
+    {
+      "epoch": 0.2090079805042976,
+      "grad_norm": 2.109375,
+      "learning_rate": 4.491610583317918e-05,
+      "loss": 0.4821,
+      "step": 4760
+    },
+    {
+      "epoch": 0.2090957989835011,
+      "grad_norm": 2.09375,
+      "learning_rate": 4.491192332453892e-05,
+      "loss": 0.5116,
+      "step": 4762
+    },
+    {
+      "epoch": 0.20918361746270459,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.490773929103879e-05,
+      "loss": 0.4873,
+      "step": 4764
+    },
+    {
+      "epoch": 0.20927143594190808,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.4903553732999225e-05,
+      "loss": 0.5058,
+      "step": 4766
+    },
+    {
+      "epoch": 0.20935925442111156,
+      "grad_norm": 2.484375,
+      "learning_rate": 4.489936665074074e-05,
+      "loss": 0.4896,
+      "step": 4768
+    },
+    {
+      "epoch": 0.20944707290031506,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.489517804458398e-05,
+      "loss": 0.4906,
+      "step": 4770
+    },
+    {
+      "epoch": 0.20953489137951853,
+      "grad_norm": 3.3125,
+      "learning_rate": 4.4890987914849724e-05,
+      "loss": 0.4861,
+      "step": 4772
+    },
+    {
+      "epoch": 0.20962270985872203,
+      "grad_norm": 3.125,
+      "learning_rate": 4.488679626185884e-05,
+      "loss": 0.4878,
+      "step": 4774
+    },
+    {
+      "epoch": 0.2097105283379255,
+      "grad_norm": 3.15625,
+      "learning_rate": 4.488260308593233e-05,
+      "loss": 0.4785,
+      "step": 4776
+    },
+    {
+      "epoch": 0.209798346817129,
+      "grad_norm": 3.84375,
+      "learning_rate": 4.4878408387391314e-05,
+      "loss": 0.4722,
+      "step": 4778
+    },
+    {
+      "epoch": 0.20988616529633247,
+      "grad_norm": 3.046875,
+      "learning_rate": 4.487421216655702e-05,
+      "loss": 0.4687,
+      "step": 4780
+    },
+    {
+      "epoch": 0.20997398377553597,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.487001442375079e-05,
+      "loss": 0.4903,
+      "step": 4782
+    },
+    {
+      "epoch": 0.21006180225473944,
+      "grad_norm": 2.53125,
+      "learning_rate": 4.4865815159294094e-05,
+      "loss": 0.4738,
+      "step": 4784
+    },
+    {
+      "epoch": 0.21014962073394294,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.486161437350851e-05,
+      "loss": 0.4617,
+      "step": 4786
+    },
+    {
+      "epoch": 0.21023743921314642,
+      "grad_norm": 2.515625,
+      "learning_rate": 4.485741206671574e-05,
+      "loss": 0.4792,
+      "step": 4788
+    },
+    {
+      "epoch": 0.21032525769234992,
+      "grad_norm": 2.5,
+      "learning_rate": 4.4853208239237586e-05,
+      "loss": 0.4983,
+      "step": 4790
+    },
+    {
+      "epoch": 0.2104130761715534,
+      "grad_norm": 3.078125,
+      "learning_rate": 4.4849002891395995e-05,
+      "loss": 0.5093,
+      "step": 4792
+    },
+    {
+      "epoch": 0.2105008946507569,
+      "grad_norm": 2.546875,
+      "learning_rate": 4.4844796023512995e-05,
+      "loss": 0.4946,
+      "step": 4794
+    },
+    {
+      "epoch": 0.21058871312996036,
+      "grad_norm": 2.390625,
+      "learning_rate": 4.484058763591077e-05,
+      "loss": 0.5118,
+      "step": 4796
+    },
+    {
+      "epoch": 0.21067653160916386,
+      "grad_norm": 2.515625,
+      "learning_rate": 4.4836377728911574e-05,
+      "loss": 0.5033,
+      "step": 4798
+    },
+    {
+      "epoch": 0.21076435008836736,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.483216630283783e-05,
+      "loss": 0.4579,
+      "step": 4800
+    },
+    {
+      "epoch": 0.21085216856757083,
+      "grad_norm": 2.21875,
+      "learning_rate": 4.4827953358012024e-05,
+      "loss": 0.5025,
+      "step": 4802
+    },
+    {
+      "epoch": 0.21093998704677433,
+      "grad_norm": 2.515625,
+      "learning_rate": 4.482373889475681e-05,
+      "loss": 0.474,
+      "step": 4804
+    },
+    {
+      "epoch": 0.2110278055259778,
+      "grad_norm": 2.453125,
+      "learning_rate": 4.481952291339491e-05,
+      "loss": 0.5077,
+      "step": 4806
+    },
+    {
+      "epoch": 0.2111156240051813,
+      "grad_norm": 2.4375,
+      "learning_rate": 4.48153054142492e-05,
+      "loss": 0.4601,
+      "step": 4808
+    },
+    {
+      "epoch": 0.21120344248438477,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.481108639764264e-05,
+      "loss": 0.4722,
+      "step": 4810
+    },
+    {
+      "epoch": 0.21129126096358827,
+      "grad_norm": 2.421875,
+      "learning_rate": 4.4806865863898346e-05,
+      "loss": 0.4735,
+      "step": 4812
+    },
+    {
+      "epoch": 0.21137907944279175,
+      "grad_norm": 2.484375,
+      "learning_rate": 4.480264381333951e-05,
+      "loss": 0.459,
+      "step": 4814
+    },
+    {
+      "epoch": 0.21146689792199524,
+      "grad_norm": 2.625,
+      "learning_rate": 4.479842024628946e-05,
+      "loss": 0.4863,
+      "step": 4816
+    },
+    {
+      "epoch": 0.21155471640119872,
+      "grad_norm": 2.234375,
+      "learning_rate": 4.4794195163071656e-05,
+      "loss": 0.4931,
+      "step": 4818
+    },
+    {
+      "epoch": 0.21164253488040222,
+      "grad_norm": 2.15625,
+      "learning_rate": 4.478996856400963e-05,
+      "loss": 0.4681,
+      "step": 4820
+    },
+    {
+      "epoch": 0.2117303533596057,
+      "grad_norm": 2.40625,
+      "learning_rate": 4.478574044942707e-05,
+      "loss": 0.5064,
+      "step": 4822
+    },
+    {
+      "epoch": 0.2118181718388092,
+      "grad_norm": 2.40625,
+      "learning_rate": 4.478151081964777e-05,
+      "loss": 0.4992,
+      "step": 4824
+    },
+    {
+      "epoch": 0.21190599031801266,
+      "grad_norm": 2.75,
+      "learning_rate": 4.477727967499562e-05,
+      "loss": 0.4752,
+      "step": 4826
+    },
+    {
+      "epoch": 0.21199380879721616,
+      "grad_norm": 2.4375,
+      "learning_rate": 4.477304701579466e-05,
+      "loss": 0.4972,
+      "step": 4828
+    },
+    {
+      "epoch": 0.21208162727641963,
+      "grad_norm": 2.203125,
+      "learning_rate": 4.476881284236901e-05,
+      "loss": 0.4839,
+      "step": 4830
+    },
+    {
+      "epoch": 0.21216944575562313,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.4764577155042935e-05,
+      "loss": 0.4915,
+      "step": 4832
+    },
+    {
+      "epoch": 0.2122572642348266,
+      "grad_norm": 2.421875,
+      "learning_rate": 4.4760339954140805e-05,
+      "loss": 0.516,
+      "step": 4834
+    },
+    {
+      "epoch": 0.2123450827140301,
+      "grad_norm": 2.265625,
+      "learning_rate": 4.475610123998711e-05,
+      "loss": 0.4683,
+      "step": 4836
+    },
+    {
+      "epoch": 0.21243290119323358,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.4751861012906445e-05,
+      "loss": 0.4993,
+      "step": 4838
+    },
+    {
+      "epoch": 0.21252071967243708,
+      "grad_norm": 2.21875,
+      "learning_rate": 4.4747619273223525e-05,
+      "loss": 0.483,
+      "step": 4840
+    },
+    {
+      "epoch": 0.21260853815164055,
+      "grad_norm": 2.203125,
+      "learning_rate": 4.474337602126319e-05,
+      "loss": 0.4766,
+      "step": 4842
+    },
+    {
+      "epoch": 0.21269635663084405,
+      "grad_norm": 2.53125,
+      "learning_rate": 4.473913125735038e-05,
+      "loss": 0.4787,
+      "step": 4844
+    },
+    {
+      "epoch": 0.21278417511004752,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.4734884981810174e-05,
+      "loss": 0.4699,
+      "step": 4846
+    },
+    {
+      "epoch": 0.21287199358925102,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.473063719496774e-05,
+      "loss": 0.4876,
+      "step": 4848
+    },
+    {
+      "epoch": 0.21295981206845452,
+      "grad_norm": 2.453125,
+      "learning_rate": 4.472638789714838e-05,
+      "loss": 0.483,
+      "step": 4850
+    },
+    {
+      "epoch": 0.213047630547658,
+      "grad_norm": 2.34375,
+      "learning_rate": 4.472213708867751e-05,
+      "loss": 0.5009,
+      "step": 4852
+    },
+    {
+      "epoch": 0.2131354490268615,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.471788476988066e-05,
+      "loss": 0.4802,
+      "step": 4854
+    },
+    {
+      "epoch": 0.21322326750606496,
+      "grad_norm": 2.453125,
+      "learning_rate": 4.4713630941083454e-05,
+      "loss": 0.4822,
+      "step": 4856
+    },
+    {
+      "epoch": 0.21331108598526846,
+      "grad_norm": 2.25,
+      "learning_rate": 4.470937560261167e-05,
+      "loss": 0.4849,
+      "step": 4858
+    },
+    {
+      "epoch": 0.21339890446447193,
+      "grad_norm": 2.46875,
+      "learning_rate": 4.4705118754791184e-05,
+      "loss": 0.4776,
+      "step": 4860
+    },
+    {
+      "epoch": 0.21348672294367543,
+      "grad_norm": 2.265625,
+      "learning_rate": 4.470086039794797e-05,
+      "loss": 0.4665,
+      "step": 4862
+    },
+    {
+      "epoch": 0.2135745414228789,
+      "grad_norm": 2.953125,
+      "learning_rate": 4.469660053240815e-05,
+      "loss": 0.4877,
+      "step": 4864
+    },
+    {
+      "epoch": 0.2136623599020824,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.469233915849794e-05,
+      "loss": 0.5137,
+      "step": 4866
+    },
+    {
+      "epoch": 0.21375017838128588,
+      "grad_norm": 2.46875,
+      "learning_rate": 4.468807627654368e-05,
+      "loss": 0.498,
+      "step": 4868
+    },
+    {
+      "epoch": 0.21383799686048938,
+      "grad_norm": 2.984375,
+      "learning_rate": 4.4683811886871804e-05,
+      "loss": 0.495,
+      "step": 4870
+    },
+    {
+      "epoch": 0.21392581533969285,
+      "grad_norm": 2.453125,
+      "learning_rate": 4.467954598980891e-05,
+      "loss": 0.4879,
+      "step": 4872
+    },
+    {
+      "epoch": 0.21401363381889635,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.4675278585681665e-05,
+      "loss": 0.4437,
+      "step": 4874
+    },
+    {
+      "epoch": 0.21410145229809982,
+      "grad_norm": 2.453125,
+      "learning_rate": 4.467100967481687e-05,
+      "loss": 0.4846,
+      "step": 4876
+    },
+    {
+      "epoch": 0.21418927077730332,
+      "grad_norm": 2.484375,
+      "learning_rate": 4.466673925754143e-05,
+      "loss": 0.4913,
+      "step": 4878
+    },
+    {
+      "epoch": 0.2142770892565068,
+      "grad_norm": 2.109375,
+      "learning_rate": 4.4662467334182387e-05,
+      "loss": 0.4757,
+      "step": 4880
+    },
+    {
+      "epoch": 0.2143649077357103,
+      "grad_norm": 2.40625,
+      "learning_rate": 4.465819390506689e-05,
+      "loss": 0.4906,
+      "step": 4882
+    },
+    {
+      "epoch": 0.21445272621491376,
+      "grad_norm": 2.3125,
+      "learning_rate": 4.465391897052218e-05,
+      "loss": 0.4691,
+      "step": 4884
+    },
+    {
+      "epoch": 0.21454054469411726,
+      "grad_norm": 2.328125,
+      "learning_rate": 4.4649642530875645e-05,
+      "loss": 0.497,
+      "step": 4886
+    },
+    {
+      "epoch": 0.21462836317332074,
+      "grad_norm": 2.515625,
+      "learning_rate": 4.464536458645479e-05,
+      "loss": 0.4595,
+      "step": 4888
+    },
+    {
+      "epoch": 0.21471618165252424,
+      "grad_norm": 2.1875,
+      "learning_rate": 4.464108513758719e-05,
+      "loss": 0.4769,
+      "step": 4890
+    },
+    {
+      "epoch": 0.2148040001317277,
+      "grad_norm": 2.046875,
+      "learning_rate": 4.4636804184600575e-05,
+      "loss": 0.4988,
+      "step": 4892
+    },
+    {
+      "epoch": 0.2148918186109312,
+      "grad_norm": 2.171875,
+      "learning_rate": 4.4632521727822805e-05,
+      "loss": 0.4872,
+      "step": 4894
+    },
+    {
+      "epoch": 0.21497963709013468,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.4628237767581814e-05,
+      "loss": 0.4902,
+      "step": 4896
+    },
+    {
+      "epoch": 0.21506745556933818,
+      "grad_norm": 2.46875,
+      "learning_rate": 4.462395230420566e-05,
+      "loss": 0.5361,
+      "step": 4898
+    },
+    {
+      "epoch": 0.21515527404854168,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.4619665338022545e-05,
+      "loss": 0.5288,
+      "step": 4900
+    },
+    {
+      "epoch": 0.21524309252774515,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.461537686936075e-05,
+      "loss": 0.481,
+      "step": 4902
+    },
+    {
+      "epoch": 0.21533091100694865,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.461108689854869e-05,
+      "loss": 0.4923,
+      "step": 4904
+    },
+    {
+      "epoch": 0.21541872948615212,
+      "grad_norm": 2.4375,
+      "learning_rate": 4.460679542591489e-05,
+      "loss": 0.4724,
+      "step": 4906
+    },
+    {
+      "epoch": 0.21550654796535562,
+      "grad_norm": 2.421875,
+      "learning_rate": 4.4602502451788005e-05,
+      "loss": 0.4788,
+      "step": 4908
+    },
+    {
+      "epoch": 0.2155943664445591,
+      "grad_norm": 2.4375,
+      "learning_rate": 4.459820797649678e-05,
+      "loss": 0.4707,
+      "step": 4910
+    },
+    {
+      "epoch": 0.2156821849237626,
+      "grad_norm": 2.28125,
+      "learning_rate": 4.4593912000370085e-05,
+      "loss": 0.4789,
+      "step": 4912
+    },
+    {
+      "epoch": 0.21577000340296607,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.458961452373692e-05,
+      "loss": 0.4999,
+      "step": 4914
+    },
+    {
+      "epoch": 0.21585782188216956,
+      "grad_norm": 2.484375,
+      "learning_rate": 4.458531554692638e-05,
+      "loss": 0.4717,
+      "step": 4916
+    },
+    {
+      "epoch": 0.21594564036137304,
+      "grad_norm": 2.4375,
+      "learning_rate": 4.458101507026767e-05,
+      "loss": 0.4894,
+      "step": 4918
+    },
+    {
+      "epoch": 0.21603345884057654,
+      "grad_norm": 2.09375,
+      "learning_rate": 4.4576713094090146e-05,
+      "loss": 0.4699,
+      "step": 4920
+    },
+    {
+      "epoch": 0.21612127731978,
+      "grad_norm": 2.28125,
+      "learning_rate": 4.457240961872323e-05,
+      "loss": 0.4841,
+      "step": 4922
+    },
+    {
+      "epoch": 0.2162090957989835,
+      "grad_norm": 2.296875,
+      "learning_rate": 4.45681046444965e-05,
+      "loss": 0.4954,
+      "step": 4924
+    },
+    {
+      "epoch": 0.21629691427818698,
+      "grad_norm": 2.984375,
+      "learning_rate": 4.4563798171739626e-05,
+      "loss": 0.4853,
+      "step": 4926
+    },
+    {
+      "epoch": 0.21638473275739048,
+      "grad_norm": 2.3125,
+      "learning_rate": 4.455949020078239e-05,
+      "loss": 0.5074,
+      "step": 4928
+    },
+    {
+      "epoch": 0.21647255123659395,
+      "grad_norm": 2.328125,
+      "learning_rate": 4.455518073195471e-05,
+      "loss": 0.5,
+      "step": 4930
+    },
+    {
+      "epoch": 0.21656036971579745,
+      "grad_norm": 2.546875,
+      "learning_rate": 4.45508697655866e-05,
+      "loss": 0.473,
+      "step": 4932
+    },
+    {
+      "epoch": 0.21664818819500092,
+      "grad_norm": 2.421875,
+      "learning_rate": 4.4546557302008195e-05,
+      "loss": 0.4746,
+      "step": 4934
+    },
+    {
+      "epoch": 0.21673600667420442,
+      "grad_norm": 2.234375,
+      "learning_rate": 4.454224334154975e-05,
+      "loss": 0.4836,
+      "step": 4936
+    },
+    {
+      "epoch": 0.2168238251534079,
+      "grad_norm": 2.34375,
+      "learning_rate": 4.453792788454163e-05,
+      "loss": 0.4828,
+      "step": 4938
+    },
+    {
+      "epoch": 0.2169116436326114,
+      "grad_norm": 2.3125,
+      "learning_rate": 4.45336109313143e-05,
+      "loss": 0.5062,
+      "step": 4940
+    },
+    {
+      "epoch": 0.21699946211181487,
+      "grad_norm": 2.546875,
+      "learning_rate": 4.452929248219837e-05,
+      "loss": 0.4902,
+      "step": 4942
+    },
+    {
+      "epoch": 0.21708728059101837,
+      "grad_norm": 2.515625,
+      "learning_rate": 4.4524972537524535e-05,
+      "loss": 0.4679,
+      "step": 4944
+    },
+    {
+      "epoch": 0.21717509907022184,
+      "grad_norm": 2.40625,
+      "learning_rate": 4.4520651097623625e-05,
+      "loss": 0.4979,
+      "step": 4946
+    },
+    {
+      "epoch": 0.21726291754942534,
+      "grad_norm": 2.21875,
+      "learning_rate": 4.451632816282657e-05,
+      "loss": 0.4708,
+      "step": 4948
+    },
+    {
+      "epoch": 0.2173507360286288,
+      "grad_norm": 2.40625,
+      "learning_rate": 4.4512003733464435e-05,
+      "loss": 0.4937,
+      "step": 4950
+    },
+    {
+      "epoch": 0.2174385545078323,
+      "grad_norm": 2.359375,
+      "learning_rate": 4.450767780986837e-05,
+      "loss": 0.47,
+      "step": 4952
+    },
+    {
+      "epoch": 0.2175263729870358,
+      "grad_norm": 2.46875,
+      "learning_rate": 4.4503350392369664e-05,
+      "loss": 0.4776,
+      "step": 4954
+    },
+    {
+      "epoch": 0.21761419146623928,
+      "grad_norm": 2.40625,
+      "learning_rate": 4.4499021481299705e-05,
+      "loss": 0.4668,
+      "step": 4956
+    },
+    {
+      "epoch": 0.21770200994544278,
+      "grad_norm": 2.15625,
+      "learning_rate": 4.449469107699001e-05,
+      "loss": 0.4934,
+      "step": 4958
+    },
+    {
+      "epoch": 0.21778982842464625,
+      "grad_norm": 2.421875,
+      "learning_rate": 4.4490359179772204e-05,
+      "loss": 0.4434,
+      "step": 4960
+    },
+    {
+      "epoch": 0.21787764690384975,
+      "grad_norm": 2.484375,
+      "learning_rate": 4.4486025789978016e-05,
+      "loss": 0.4831,
+      "step": 4962
+    },
+    {
+      "epoch": 0.21796546538305323,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.44816909079393e-05,
+      "loss": 0.4856,
+      "step": 4964
+    },
+    {
+      "epoch": 0.21805328386225672,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.4477354533988025e-05,
+      "loss": 0.4619,
+      "step": 4966
+    },
+    {
+      "epoch": 0.2181411023414602,
+      "grad_norm": 2.53125,
+      "learning_rate": 4.447301666845628e-05,
+      "loss": 0.5101,
+      "step": 4968
+    },
+    {
+      "epoch": 0.2182289208206637,
+      "grad_norm": 2.953125,
+      "learning_rate": 4.4468677311676236e-05,
+      "loss": 0.4775,
+      "step": 4970
+    },
+    {
+      "epoch": 0.21831673929986717,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.4464336463980226e-05,
+      "loss": 0.4981,
+      "step": 4972
+    },
+    {
+      "epoch": 0.21840455777907067,
+      "grad_norm": 2.140625,
+      "learning_rate": 4.445999412570065e-05,
+      "loss": 0.496,
+      "step": 4974
+    },
+    {
+      "epoch": 0.21849237625827414,
+      "grad_norm": 2.125,
+      "learning_rate": 4.445565029717008e-05,
+      "loss": 0.499,
+      "step": 4976
+    },
+    {
+      "epoch": 0.21858019473747764,
+      "grad_norm": 2.1875,
+      "learning_rate": 4.445130497872113e-05,
+      "loss": 0.4771,
+      "step": 4978
+    },
+    {
+      "epoch": 0.2186680132166811,
+      "grad_norm": 2.375,
+      "learning_rate": 4.4446958170686593e-05,
+      "loss": 0.5265,
+      "step": 4980
+    },
+    {
+      "epoch": 0.2187558316958846,
+      "grad_norm": 2.40625,
+      "learning_rate": 4.444260987339933e-05,
+      "loss": 0.4932,
+      "step": 4982
+    },
+    {
+      "epoch": 0.21884365017508808,
+      "grad_norm": 2.5,
+      "learning_rate": 4.443826008719235e-05,
+      "loss": 0.4904,
+      "step": 4984
+    },
+    {
+      "epoch": 0.21893146865429158,
+      "grad_norm": 2.34375,
+      "learning_rate": 4.4433908812398736e-05,
+      "loss": 0.4713,
+      "step": 4986
+    },
+    {
+      "epoch": 0.21901928713349506,
+      "grad_norm": 2.484375,
+      "learning_rate": 4.442955604935174e-05,
+      "loss": 0.5122,
+      "step": 4988
+    },
+    {
+      "epoch": 0.21910710561269856,
+      "grad_norm": 2.34375,
+      "learning_rate": 4.4425201798384686e-05,
+      "loss": 0.4549,
+      "step": 4990
+    },
+    {
+      "epoch": 0.21919492409190203,
+      "grad_norm": 2.25,
+      "learning_rate": 4.442084605983102e-05,
+      "loss": 0.4765,
+      "step": 4992
+    },
+    {
+      "epoch": 0.21928274257110553,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.441648883402431e-05,
+      "loss": 0.5142,
+      "step": 4994
+    },
+    {
+      "epoch": 0.219370561050309,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.441213012129822e-05,
+      "loss": 0.4851,
+      "step": 4996
+    },
+    {
+      "epoch": 0.2194583795295125,
+      "grad_norm": 2.265625,
+      "learning_rate": 4.4407769921986554e-05,
+      "loss": 0.4862,
+      "step": 4998
+    },
+    {
+      "epoch": 0.21954619800871597,
+      "grad_norm": 2.34375,
+      "learning_rate": 4.4403408236423224e-05,
+      "loss": 0.4953,
+      "step": 5000
+    },
+    {
+      "epoch": 0.21963401648791947,
+      "grad_norm": 2.453125,
+      "learning_rate": 4.4399045064942236e-05,
+      "loss": 0.478,
+      "step": 5002
+    },
+    {
+      "epoch": 0.21972183496712297,
+      "grad_norm": 2.171875,
+      "learning_rate": 4.439468040787772e-05,
+      "loss": 0.5044,
+      "step": 5004
+    },
+    {
+      "epoch": 0.21980965344632644,
+      "grad_norm": 2.375,
+      "learning_rate": 4.439031426556394e-05,
+      "loss": 0.5156,
+      "step": 5006
+    },
+    {
+      "epoch": 0.21989747192552994,
+      "grad_norm": 2.296875,
+      "learning_rate": 4.438594663833523e-05,
+      "loss": 0.4612,
+      "step": 5008
+    },
+    {
+      "epoch": 0.2199852904047334,
+      "grad_norm": 2.203125,
+      "learning_rate": 4.4381577526526094e-05,
+      "loss": 0.4944,
+      "step": 5010
+    },
+    {
+      "epoch": 0.2200731088839369,
+      "grad_norm": 2.34375,
+      "learning_rate": 4.43772069304711e-05,
+      "loss": 0.511,
+      "step": 5012
+    },
+    {
+      "epoch": 0.22016092736314039,
+      "grad_norm": 2.375,
+      "learning_rate": 4.437283485050495e-05,
+      "loss": 0.4886,
+      "step": 5014
+    },
+    {
+      "epoch": 0.22024874584234388,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.436846128696247e-05,
+      "loss": 0.5139,
+      "step": 5016
+    },
+    {
+      "epoch": 0.22033656432154736,
+      "grad_norm": 2.28125,
+      "learning_rate": 4.4364086240178584e-05,
+      "loss": 0.4997,
+      "step": 5018
+    },
+    {
+      "epoch": 0.22042438280075086,
+      "grad_norm": 2.40625,
+      "learning_rate": 4.435970971048832e-05,
+      "loss": 0.4923,
+      "step": 5020
+    },
+    {
+      "epoch": 0.22051220127995433,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.435533169822685e-05,
+      "loss": 0.5019,
+      "step": 5022
+    },
+    {
+      "epoch": 0.22060001975915783,
+      "grad_norm": 2.25,
+      "learning_rate": 4.435095220372945e-05,
+      "loss": 0.4697,
+      "step": 5024
+    },
+    {
+      "epoch": 0.2206878382383613,
+      "grad_norm": 2.109375,
+      "learning_rate": 4.434657122733148e-05,
+      "loss": 0.4798,
+      "step": 5026
+    },
+    {
+      "epoch": 0.2207756567175648,
+      "grad_norm": 2.4375,
+      "learning_rate": 4.4342188769368446e-05,
+      "loss": 0.4827,
+      "step": 5028
+    },
+    {
+      "epoch": 0.22086347519676827,
+      "grad_norm": 2.3125,
+      "learning_rate": 4.433780483017597e-05,
+      "loss": 0.4766,
+      "step": 5030
+    },
+    {
+      "epoch": 0.22095129367597177,
+      "grad_norm": 2.25,
+      "learning_rate": 4.433341941008975e-05,
+      "loss": 0.502,
+      "step": 5032
+    },
+    {
+      "epoch": 0.22103911215517524,
+      "grad_norm": 2.40625,
+      "learning_rate": 4.432903250944565e-05,
+      "loss": 0.4846,
+      "step": 5034
+    },
+    {
+      "epoch": 0.22112693063437874,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.43246441285796e-05,
+      "loss": 0.4871,
+      "step": 5036
+    },
+    {
+      "epoch": 0.22121474911358222,
+      "grad_norm": 2.0625,
+      "learning_rate": 4.4320254267827675e-05,
+      "loss": 0.4614,
+      "step": 5038
+    },
+    {
+      "epoch": 0.22130256759278571,
+      "grad_norm": 2.296875,
+      "learning_rate": 4.4315862927526044e-05,
+      "loss": 0.525,
+      "step": 5040
+    },
+    {
+      "epoch": 0.2213903860719892,
+      "grad_norm": 2.296875,
+      "learning_rate": 4.4311470108011e-05,
+      "loss": 0.4742,
+      "step": 5042
+    },
+    {
+      "epoch": 0.2214782045511927,
+      "grad_norm": 2.078125,
+      "learning_rate": 4.4307075809618946e-05,
+      "loss": 0.4935,
+      "step": 5044
+    },
+    {
+      "epoch": 0.22156602303039616,
+      "grad_norm": 2.484375,
+      "learning_rate": 4.4302680032686395e-05,
+      "loss": 0.4644,
+      "step": 5046
+    },
+    {
+      "epoch": 0.22165384150959966,
+      "grad_norm": 2.234375,
+      "learning_rate": 4.429828277754998e-05,
+      "loss": 0.4973,
+      "step": 5048
+    },
+    {
+      "epoch": 0.22174165998880313,
+      "grad_norm": 2.140625,
+      "learning_rate": 4.4293884044546455e-05,
+      "loss": 0.4719,
+      "step": 5050
+    },
+    {
+      "epoch": 0.22182947846800663,
+      "grad_norm": 2.484375,
+      "learning_rate": 4.428948383401265e-05,
+      "loss": 0.4938,
+      "step": 5052
+    },
+    {
+      "epoch": 0.22191729694721013,
+      "grad_norm": 2.375,
+      "learning_rate": 4.4285082146285556e-05,
+      "loss": 0.4923,
+      "step": 5054
+    },
+    {
+      "epoch": 0.2220051154264136,
+      "grad_norm": 2.28125,
+      "learning_rate": 4.428067898170225e-05,
+      "loss": 0.4724,
+      "step": 5056
+    },
+    {
+      "epoch": 0.2220929339056171,
+      "grad_norm": 2.171875,
+      "learning_rate": 4.427627434059992e-05,
+      "loss": 0.5209,
+      "step": 5058
+    },
+    {
+      "epoch": 0.22218075238482057,
+      "grad_norm": 2.421875,
+      "learning_rate": 4.4271868223315884e-05,
+      "loss": 0.4846,
+      "step": 5060
+    },
+    {
+      "epoch": 0.22226857086402407,
+      "grad_norm": 2.125,
+      "learning_rate": 4.4267460630187566e-05,
+      "loss": 0.4762,
+      "step": 5062
+    },
+    {
+      "epoch": 0.22235638934322755,
+      "grad_norm": 2.328125,
+      "learning_rate": 4.4263051561552485e-05,
+      "loss": 0.4835,
+      "step": 5064
+    },
+    {
+      "epoch": 0.22244420782243104,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.42586410177483e-05,
+      "loss": 0.4987,
+      "step": 5066
+    },
+    {
+      "epoch": 0.22253202630163452,
+      "grad_norm": 2.3125,
+      "learning_rate": 4.425422899911277e-05,
+      "loss": 0.4749,
+      "step": 5068
+    },
+    {
+      "epoch": 0.22261984478083802,
+      "grad_norm": 2.421875,
+      "learning_rate": 4.424981550598376e-05,
+      "loss": 0.5063,
+      "step": 5070
+    },
+    {
+      "epoch": 0.2227076632600415,
+      "grad_norm": 2.484375,
+      "learning_rate": 4.4245400538699275e-05,
+      "loss": 0.4963,
+      "step": 5072
+    },
+    {
+      "epoch": 0.222795481739245,
+      "grad_norm": 2.359375,
+      "learning_rate": 4.42409840975974e-05,
+      "loss": 0.5134,
+      "step": 5074
+    },
+    {
+      "epoch": 0.22288330021844846,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.4236566183016345e-05,
+      "loss": 0.4914,
+      "step": 5076
+    },
+    {
+      "epoch": 0.22297111869765196,
+      "grad_norm": 2.328125,
+      "learning_rate": 4.423214679529445e-05,
+      "loss": 0.5302,
+      "step": 5078
+    },
+    {
+      "epoch": 0.22305893717685543,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.422772593477014e-05,
+      "loss": 0.4669,
+      "step": 5080
+    },
+    {
+      "epoch": 0.22314675565605893,
+      "grad_norm": 2.625,
+      "learning_rate": 4.4223303601781966e-05,
+      "loss": 0.4804,
+      "step": 5082
+    },
+    {
+      "epoch": 0.2232345741352624,
+      "grad_norm": 2.3125,
+      "learning_rate": 4.421887979666859e-05,
+      "loss": 0.4832,
+      "step": 5084
+    },
+    {
+      "epoch": 0.2233223926144659,
+      "grad_norm": 2.53125,
+      "learning_rate": 4.42144545197688e-05,
+      "loss": 0.4871,
+      "step": 5086
+    },
+    {
+      "epoch": 0.22341021109366938,
+      "grad_norm": 2.3125,
+      "learning_rate": 4.421002777142148e-05,
+      "loss": 0.4789,
+      "step": 5088
+    },
+    {
+      "epoch": 0.22349802957287287,
+      "grad_norm": 2.3125,
+      "learning_rate": 4.420559955196562e-05,
+      "loss": 0.4875,
+      "step": 5090
+    },
+    {
+      "epoch": 0.22358584805207635,
+      "grad_norm": 2.4375,
+      "learning_rate": 4.420116986174034e-05,
+      "loss": 0.48,
+      "step": 5092
+    },
+    {
+      "epoch": 0.22367366653127985,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.419673870108488e-05,
+      "loss": 0.4575,
+      "step": 5094
+    },
+    {
+      "epoch": 0.22376148501048332,
+      "grad_norm": 3.109375,
+      "learning_rate": 4.419230607033856e-05,
+      "loss": 0.4913,
+      "step": 5096
+    },
+    {
+      "epoch": 0.22384930348968682,
+      "grad_norm": 2.328125,
+      "learning_rate": 4.4187871969840844e-05,
+      "loss": 0.4545,
+      "step": 5098
+    },
+    {
+      "epoch": 0.2239371219688903,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.418343639993129e-05,
+      "loss": 0.4741,
+      "step": 5100
+    },
+    {
+      "epoch": 0.2240249404480938,
+      "grad_norm": 2.296875,
+      "learning_rate": 4.417899936094958e-05,
+      "loss": 0.4738,
+      "step": 5102
+    },
+    {
+      "epoch": 0.2241127589272973,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.4174560853235505e-05,
+      "loss": 0.5039,
+      "step": 5104
+    },
+    {
+      "epoch": 0.22420057740650076,
+      "grad_norm": 2.359375,
+      "learning_rate": 4.4170120877128964e-05,
+      "loss": 0.4679,
+      "step": 5106
+    },
+    {
+      "epoch": 0.22428839588570426,
+      "grad_norm": 2.25,
+      "learning_rate": 4.4165679432969956e-05,
+      "loss": 0.4549,
+      "step": 5108
+    },
+    {
+      "epoch": 0.22437621436490773,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.416123652109864e-05,
+      "loss": 0.5007,
+      "step": 5110
+    },
+    {
+      "epoch": 0.22446403284411123,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.415679214185523e-05,
+      "loss": 0.4696,
+      "step": 5112
+    },
+    {
+      "epoch": 0.2245518513233147,
+      "grad_norm": 2.234375,
+      "learning_rate": 4.415234629558008e-05,
+      "loss": 0.4881,
+      "step": 5114
+    },
+    {
+      "epoch": 0.2246396698025182,
+      "grad_norm": 2.359375,
+      "learning_rate": 4.4147898982613675e-05,
+      "loss": 0.4651,
+      "step": 5116
+    },
+    {
+      "epoch": 0.22472748828172168,
+      "grad_norm": 2.5,
+      "learning_rate": 4.4143450203296566e-05,
+      "loss": 0.4887,
+      "step": 5118
+    },
+    {
+      "epoch": 0.22481530676092518,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.413899995796945e-05,
+      "loss": 0.4802,
+      "step": 5120
+    },
+    {
+      "epoch": 0.22490312524012865,
+      "grad_norm": 2.3125,
+      "learning_rate": 4.413454824697313e-05,
+      "loss": 0.4907,
+      "step": 5122
+    },
+    {
+      "epoch": 0.22499094371933215,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.4130095070648524e-05,
+      "loss": 0.4741,
+      "step": 5124
+    },
+    {
+      "epoch": 0.22507876219853562,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.4125640429336646e-05,
+      "loss": 0.4801,
+      "step": 5126
+    },
+    {
+      "epoch": 0.22516658067773912,
+      "grad_norm": 2.3125,
+      "learning_rate": 4.4121184323378636e-05,
+      "loss": 0.4828,
+      "step": 5128
+    },
+    {
+      "epoch": 0.2252543991569426,
+      "grad_norm": 2.265625,
+      "learning_rate": 4.411672675311576e-05,
+      "loss": 0.468,
+      "step": 5130
+    },
+    {
+      "epoch": 0.2253422176361461,
+      "grad_norm": 2.453125,
+      "learning_rate": 4.4112267718889355e-05,
+      "loss": 0.4707,
+      "step": 5132
+    },
+    {
+      "epoch": 0.22543003611534956,
+      "grad_norm": 3.015625,
+      "learning_rate": 4.410780722104091e-05,
+      "loss": 0.5055,
+      "step": 5134
+    },
+    {
+      "epoch": 0.22551785459455306,
+      "grad_norm": 2.546875,
+      "learning_rate": 4.410334525991201e-05,
+      "loss": 0.4714,
+      "step": 5136
+    },
+    {
+      "epoch": 0.22560567307375654,
+      "grad_norm": 2.953125,
+      "learning_rate": 4.409888183584435e-05,
+      "loss": 0.4737,
+      "step": 5138
+    },
+    {
+      "epoch": 0.22569349155296003,
+      "grad_norm": 3.421875,
+      "learning_rate": 4.409441694917973e-05,
+      "loss": 0.4804,
+      "step": 5140
+    },
+    {
+      "epoch": 0.2257813100321635,
+      "grad_norm": 2.5,
+      "learning_rate": 4.40899506002601e-05,
+      "loss": 0.4848,
+      "step": 5142
+    },
+    {
+      "epoch": 0.225869128511367,
+      "grad_norm": 2.09375,
+      "learning_rate": 4.408548278942747e-05,
+      "loss": 0.4637,
+      "step": 5144
+    },
+    {
+      "epoch": 0.22595694699057048,
+      "grad_norm": 2.03125,
+      "learning_rate": 4.408101351702398e-05,
+      "loss": 0.4872,
+      "step": 5146
+    },
+    {
+      "epoch": 0.22604476546977398,
+      "grad_norm": 2.296875,
+      "learning_rate": 4.4076542783391925e-05,
+      "loss": 0.4784,
+      "step": 5148
+    },
+    {
+      "epoch": 0.22613258394897745,
+      "grad_norm": 2.328125,
+      "learning_rate": 4.4072070588873635e-05,
+      "loss": 0.5011,
+      "step": 5150
+    },
+    {
+      "epoch": 0.22622040242818095,
+      "grad_norm": 2.21875,
+      "learning_rate": 4.406759693381161e-05,
+      "loss": 0.4579,
+      "step": 5152
+    },
+    {
+      "epoch": 0.22630822090738445,
+      "grad_norm": 2.96875,
+      "learning_rate": 4.4063121818548435e-05,
+      "loss": 0.4964,
+      "step": 5154
+    },
+    {
+      "epoch": 0.22639603938658792,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.4058645243426835e-05,
+      "loss": 0.4828,
+      "step": 5156
+    },
+    {
+      "epoch": 0.22648385786579142,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.40541672087896e-05,
+      "loss": 0.4821,
+      "step": 5158
+    },
+    {
+      "epoch": 0.2265716763449949,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.404968771497968e-05,
+      "loss": 0.4809,
+      "step": 5160
+    },
+    {
+      "epoch": 0.2266594948241984,
+      "grad_norm": 2.515625,
+      "learning_rate": 4.404520676234011e-05,
+      "loss": 0.4946,
+      "step": 5162
+    },
+    {
+      "epoch": 0.22674731330340187,
+      "grad_norm": 1.9921875,
+      "learning_rate": 4.404072435121404e-05,
+      "loss": 0.46,
+      "step": 5164
+    },
+    {
+      "epoch": 0.22683513178260536,
+      "grad_norm": 2.25,
+      "learning_rate": 4.403624048194474e-05,
+      "loss": 0.4687,
+      "step": 5166
+    },
+    {
+      "epoch": 0.22692295026180884,
+      "grad_norm": 2.390625,
+      "learning_rate": 4.403175515487557e-05,
+      "loss": 0.494,
+      "step": 5168
+    },
+    {
+      "epoch": 0.22701076874101234,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.402726837035002e-05,
+      "loss": 0.4823,
+      "step": 5170
+    },
+    {
+      "epoch": 0.2270985872202158,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.402278012871172e-05,
+      "loss": 0.4808,
+      "step": 5172
+    },
+    {
+      "epoch": 0.2271864056994193,
+      "grad_norm": 2.546875,
+      "learning_rate": 4.401829043030434e-05,
+      "loss": 0.4776,
+      "step": 5174
+    },
+    {
+      "epoch": 0.22727422417862278,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.401379927547172e-05,
+      "loss": 0.4646,
+      "step": 5176
+    },
+    {
+      "epoch": 0.22736204265782628,
+      "grad_norm": 2.3125,
+      "learning_rate": 4.40093066645578e-05,
+      "loss": 0.4778,
+      "step": 5178
+    },
+    {
+      "epoch": 0.22744986113702975,
+      "grad_norm": 2.265625,
+      "learning_rate": 4.400481259790662e-05,
+      "loss": 0.4713,
+      "step": 5180
+    },
+    {
+      "epoch": 0.22753767961623325,
+      "grad_norm": 2.28125,
+      "learning_rate": 4.400031707586234e-05,
+      "loss": 0.4591,
+      "step": 5182
+    },
+    {
+      "epoch": 0.22762549809543672,
+      "grad_norm": 2.75,
+      "learning_rate": 4.3995820098769217e-05,
+      "loss": 0.4826,
+      "step": 5184
+    },
+    {
+      "epoch": 0.22771331657464022,
+      "grad_norm": 2.53125,
+      "learning_rate": 4.3991321666971636e-05,
+      "loss": 0.4527,
+      "step": 5186
+    },
+    {
+      "epoch": 0.2278011350538437,
+      "grad_norm": 2.359375,
+      "learning_rate": 4.3986821780814095e-05,
+      "loss": 0.4742,
+      "step": 5188
+    },
+    {
+      "epoch": 0.2278889535330472,
+      "grad_norm": 2.390625,
+      "learning_rate": 4.398232044064118e-05,
+      "loss": 0.464,
+      "step": 5190
+    },
+    {
+      "epoch": 0.22797677201225067,
+      "grad_norm": 2.359375,
+      "learning_rate": 4.397781764679762e-05,
+      "loss": 0.4821,
+      "step": 5192
+    },
+    {
+      "epoch": 0.22806459049145417,
+      "grad_norm": 2.125,
+      "learning_rate": 4.397331339962824e-05,
+      "loss": 0.5096,
+      "step": 5194
+    },
+    {
+      "epoch": 0.22815240897065764,
+      "grad_norm": 2.34375,
+      "learning_rate": 4.396880769947796e-05,
+      "loss": 0.5087,
+      "step": 5196
+    },
+    {
+      "epoch": 0.22824022744986114,
+      "grad_norm": 2.125,
+      "learning_rate": 4.396430054669186e-05,
+      "loss": 0.4886,
+      "step": 5198
+    },
+    {
+      "epoch": 0.2283280459290646,
+      "grad_norm": 2.5,
+      "learning_rate": 4.395979194161506e-05,
+      "loss": 0.4747,
+      "step": 5200
+    },
+    {
+      "epoch": 0.2284158644082681,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.395528188459286e-05,
+      "loss": 0.4934,
+      "step": 5202
+    },
+    {
+      "epoch": 0.2285036828874716,
+      "grad_norm": 2.234375,
+      "learning_rate": 4.395077037597062e-05,
+      "loss": 0.4631,
+      "step": 5204
+    },
+    {
+      "epoch": 0.22859150136667508,
+      "grad_norm": 2.453125,
+      "learning_rate": 4.394625741609384e-05,
+      "loss": 0.4963,
+      "step": 5206
+    },
+    {
+      "epoch": 0.22867931984587858,
+      "grad_norm": 2.15625,
+      "learning_rate": 4.3941743005308136e-05,
+      "loss": 0.4941,
+      "step": 5208
+    },
+    {
+      "epoch": 0.22876713832508205,
+      "grad_norm": 2.40625,
+      "learning_rate": 4.39372271439592e-05,
+      "loss": 0.4873,
+      "step": 5210
+    },
+    {
+      "epoch": 0.22885495680428555,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.393270983239288e-05,
+      "loss": 0.4807,
+      "step": 5212
+    },
+    {
+      "epoch": 0.22894277528348903,
+      "grad_norm": 2.453125,
+      "learning_rate": 4.39281910709551e-05,
+      "loss": 0.5022,
+      "step": 5214
+    },
+    {
+      "epoch": 0.22903059376269252,
+      "grad_norm": 2.421875,
+      "learning_rate": 4.3923670859991906e-05,
+      "loss": 0.4848,
+      "step": 5216
+    },
+    {
+      "epoch": 0.229118412241896,
+      "grad_norm": 2.28125,
+      "learning_rate": 4.391914919984947e-05,
+      "loss": 0.4556,
+      "step": 5218
+    },
+    {
+      "epoch": 0.2292062307210995,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.3914626090874044e-05,
+      "loss": 0.4881,
+      "step": 5220
+    },
+    {
+      "epoch": 0.22929404920030297,
+      "grad_norm": 2.46875,
+      "learning_rate": 4.391010153341203e-05,
+      "loss": 0.478,
+      "step": 5222
+    },
+    {
+      "epoch": 0.22938186767950647,
+      "grad_norm": 2.328125,
+      "learning_rate": 4.390557552780989e-05,
+      "loss": 0.485,
+      "step": 5224
+    },
+    {
+      "epoch": 0.22946968615870994,
+      "grad_norm": 2.3125,
+      "learning_rate": 4.390104807441425e-05,
+      "loss": 0.4779,
+      "step": 5226
+    },
+    {
+      "epoch": 0.22955750463791344,
+      "grad_norm": 2.0,
+      "learning_rate": 4.3896519173571824e-05,
+      "loss": 0.4636,
+      "step": 5228
+    },
+    {
+      "epoch": 0.2296453231171169,
+      "grad_norm": 2.265625,
+      "learning_rate": 4.389198882562943e-05,
+      "loss": 0.4515,
+      "step": 5230
+    },
+    {
+      "epoch": 0.2297331415963204,
+      "grad_norm": 2.546875,
+      "learning_rate": 4.3887457030934e-05,
+      "loss": 0.4524,
+      "step": 5232
+    },
+    {
+      "epoch": 0.22982096007552388,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.388292378983258e-05,
+      "loss": 0.4781,
+      "step": 5234
+    },
+    {
+      "epoch": 0.22990877855472738,
+      "grad_norm": 2.53125,
+      "learning_rate": 4.387838910267233e-05,
+      "loss": 0.48,
+      "step": 5236
+    },
+    {
+      "epoch": 0.22999659703393086,
+      "grad_norm": 2.375,
+      "learning_rate": 4.387385296980052e-05,
+      "loss": 0.4608,
+      "step": 5238
+    },
+    {
+      "epoch": 0.23008441551313435,
+      "grad_norm": 2.515625,
+      "learning_rate": 4.3869315391564525e-05,
+      "loss": 0.4516,
+      "step": 5240
+    },
+    {
+      "epoch": 0.23017223399233783,
+      "grad_norm": 2.53125,
+      "learning_rate": 4.3864776368311835e-05,
+      "loss": 0.4758,
+      "step": 5242
+    },
+    {
+      "epoch": 0.23026005247154133,
+      "grad_norm": 2.28125,
+      "learning_rate": 4.3860235900390046e-05,
+      "loss": 0.474,
+      "step": 5244
+    },
+    {
+      "epoch": 0.2303478709507448,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.3855693988146876e-05,
+      "loss": 0.4854,
+      "step": 5246
+    },
+    {
+      "epoch": 0.2304356894299483,
+      "grad_norm": 2.390625,
+      "learning_rate": 4.3851150631930124e-05,
+      "loss": 0.4763,
+      "step": 5248
+    },
+    {
+      "epoch": 0.23052350790915177,
+      "grad_norm": 2.25,
+      "learning_rate": 4.384660583208776e-05,
+      "loss": 0.4844,
+      "step": 5250
+    },
+    {
+      "epoch": 0.23061132638835527,
+      "grad_norm": 2.4375,
+      "learning_rate": 4.3842059588967785e-05,
+      "loss": 0.485,
+      "step": 5252
+    },
+    {
+      "epoch": 0.23069914486755877,
+      "grad_norm": 2.546875,
+      "learning_rate": 4.3837511902918384e-05,
+      "loss": 0.4942,
+      "step": 5254
+    },
+    {
+      "epoch": 0.23078696334676224,
+      "grad_norm": 2.625,
+      "learning_rate": 4.38329627742878e-05,
+      "loss": 0.4815,
+      "step": 5256
+    },
+    {
+      "epoch": 0.23087478182596574,
+      "grad_norm": 2.15625,
+      "learning_rate": 4.382841220342441e-05,
+      "loss": 0.4634,
+      "step": 5258
+    },
+    {
+      "epoch": 0.2309626003051692,
+      "grad_norm": 2.390625,
+      "learning_rate": 4.38238601906767e-05,
+      "loss": 0.4528,
+      "step": 5260
+    },
+    {
+      "epoch": 0.2310504187843727,
+      "grad_norm": 2.4375,
+      "learning_rate": 4.3819306736393265e-05,
+      "loss": 0.4836,
+      "step": 5262
+    },
+    {
+      "epoch": 0.23113823726357619,
+      "grad_norm": 2.390625,
+      "learning_rate": 4.3814751840922816e-05,
+      "loss": 0.4351,
+      "step": 5264
+    },
+    {
+      "epoch": 0.23122605574277968,
+      "grad_norm": 2.296875,
+      "learning_rate": 4.381019550461415e-05,
+      "loss": 0.4912,
+      "step": 5266
+    },
+    {
+      "epoch": 0.23131387422198316,
+      "grad_norm": 2.21875,
+      "learning_rate": 4.3805637727816205e-05,
+      "loss": 0.4542,
+      "step": 5268
+    },
+    {
+      "epoch": 0.23140169270118666,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.3801078510878025e-05,
+      "loss": 0.4752,
+      "step": 5270
+    },
+    {
+      "epoch": 0.23148951118039013,
+      "grad_norm": 2.453125,
+      "learning_rate": 4.3796517854148735e-05,
+      "loss": 0.4868,
+      "step": 5272
+    },
+    {
+      "epoch": 0.23157732965959363,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.3791955757977604e-05,
+      "loss": 0.4772,
+      "step": 5274
+    },
+    {
+      "epoch": 0.2316651481387971,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.3787392222713996e-05,
+      "loss": 0.5021,
+      "step": 5276
+    },
+    {
+      "epoch": 0.2317529666180006,
+      "grad_norm": 2.96875,
+      "learning_rate": 4.378282724870739e-05,
+      "loss": 0.4864,
+      "step": 5278
+    },
+    {
+      "epoch": 0.23184078509720407,
+      "grad_norm": 3.765625,
+      "learning_rate": 4.3778260836307373e-05,
+      "loss": 0.4819,
+      "step": 5280
+    },
+    {
+      "epoch": 0.23192860357640757,
+      "grad_norm": 3.9375,
+      "learning_rate": 4.3773692985863635e-05,
+      "loss": 0.4849,
+      "step": 5282
+    },
+    {
+      "epoch": 0.23201642205561104,
+      "grad_norm": 3.921875,
+      "learning_rate": 4.3769123697725986e-05,
+      "loss": 0.4509,
+      "step": 5284
+    },
+    {
+      "epoch": 0.23210424053481454,
+      "grad_norm": 2.484375,
+      "learning_rate": 4.376455297224435e-05,
+      "loss": 0.4961,
+      "step": 5286
+    },
+    {
+      "epoch": 0.23219205901401802,
+      "grad_norm": 2.359375,
+      "learning_rate": 4.3759980809768756e-05,
+      "loss": 0.4676,
+      "step": 5288
+    },
+    {
+      "epoch": 0.23227987749322151,
+      "grad_norm": 3.875,
+      "learning_rate": 4.3755407210649325e-05,
+      "loss": 0.5097,
+      "step": 5290
+    },
+    {
+      "epoch": 0.232367695972425,
+      "grad_norm": 3.609375,
+      "learning_rate": 4.375083217523631e-05,
+      "loss": 0.4754,
+      "step": 5292
+    },
+    {
+      "epoch": 0.2324555144516285,
+      "grad_norm": 3.453125,
+      "learning_rate": 4.374625570388008e-05,
+      "loss": 0.466,
+      "step": 5294
+    },
+    {
+      "epoch": 0.23254333293083196,
+      "grad_norm": 3.28125,
+      "learning_rate": 4.374167779693109e-05,
+      "loss": 0.4691,
+      "step": 5296
+    },
+    {
+      "epoch": 0.23263115141003546,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.3737098454739924e-05,
+      "loss": 0.4651,
+      "step": 5298
+    },
+    {
+      "epoch": 0.23271896988923893,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.373251767765727e-05,
+      "loss": 0.4742,
+      "step": 5300
+    },
+    {
+      "epoch": 0.23280678836844243,
+      "grad_norm": 3.625,
+      "learning_rate": 4.372793546603392e-05,
+      "loss": 0.4641,
+      "step": 5302
+    },
+    {
+      "epoch": 0.23289460684764593,
+      "grad_norm": 2.234375,
+      "learning_rate": 4.372335182022078e-05,
+      "loss": 0.4854,
+      "step": 5304
+    },
+    {
+      "epoch": 0.2329824253268494,
+      "grad_norm": 2.359375,
+      "learning_rate": 4.371876674056886e-05,
+      "loss": 0.4809,
+      "step": 5306
+    },
+    {
+      "epoch": 0.2330702438060529,
+      "grad_norm": 2.078125,
+      "learning_rate": 4.3714180227429316e-05,
+      "loss": 0.4776,
+      "step": 5308
+    },
+    {
+      "epoch": 0.23315806228525637,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.370959228115335e-05,
+      "loss": 0.4895,
+      "step": 5310
+    },
+    {
+      "epoch": 0.23324588076445987,
+      "grad_norm": 2.515625,
+      "learning_rate": 4.3705002902092326e-05,
+      "loss": 0.4962,
+      "step": 5312
+    },
+    {
+      "epoch": 0.23333369924366335,
+      "grad_norm": 2.328125,
+      "learning_rate": 4.37004120905977e-05,
+      "loss": 0.4616,
+      "step": 5314
+    },
+    {
+      "epoch": 0.23342151772286684,
+      "grad_norm": 2.296875,
+      "learning_rate": 4.369581984702102e-05,
+      "loss": 0.4799,
+      "step": 5316
+    },
+    {
+      "epoch": 0.23350933620207032,
+      "grad_norm": 2.328125,
+      "learning_rate": 4.3691226171713986e-05,
+      "loss": 0.4715,
+      "step": 5318
+    },
+    {
+      "epoch": 0.23359715468127382,
+      "grad_norm": 2.5,
+      "learning_rate": 4.3686631065028374e-05,
+      "loss": 0.4639,
+      "step": 5320
+    },
+    {
+      "epoch": 0.2336849731604773,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.3682034527316064e-05,
+      "loss": 0.4588,
+      "step": 5322
+    },
+    {
+      "epoch": 0.2337727916396808,
+      "grad_norm": 3.328125,
+      "learning_rate": 4.367743655892908e-05,
+      "loss": 0.4905,
+      "step": 5324
+    },
+    {
+      "epoch": 0.23386061011888426,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.367283716021953e-05,
+      "loss": 0.4552,
+      "step": 5326
+    },
+    {
+      "epoch": 0.23394842859808776,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.366823633153963e-05,
+      "loss": 0.5003,
+      "step": 5328
+    },
+    {
+      "epoch": 0.23403624707729123,
+      "grad_norm": 3.078125,
+      "learning_rate": 4.366363407324171e-05,
+      "loss": 0.4557,
+      "step": 5330
+    },
+    {
+      "epoch": 0.23412406555649473,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.365903038567822e-05,
+      "loss": 0.4511,
+      "step": 5332
+    },
+    {
+      "epoch": 0.2342118840356982,
+      "grad_norm": 2.875,
+      "learning_rate": 4.3654425269201716e-05,
+      "loss": 0.4675,
+      "step": 5334
+    },
+    {
+      "epoch": 0.2342997025149017,
+      "grad_norm": 3.203125,
+      "learning_rate": 4.364981872416485e-05,
+      "loss": 0.4511,
+      "step": 5336
+    },
+    {
+      "epoch": 0.23438752099410518,
+      "grad_norm": 3.21875,
+      "learning_rate": 4.36452107509204e-05,
+      "loss": 0.4933,
+      "step": 5338
+    },
+    {
+      "epoch": 0.23447533947330867,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.364060134982124e-05,
+      "loss": 0.4911,
+      "step": 5340
+    },
+    {
+      "epoch": 0.23456315795251215,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.3635990521220355e-05,
+      "loss": 0.5039,
+      "step": 5342
+    },
+    {
+      "epoch": 0.23465097643171565,
+      "grad_norm": 3.390625,
+      "learning_rate": 4.363137826547085e-05,
+      "loss": 0.4649,
+      "step": 5344
+    },
+    {
+      "epoch": 0.23473879491091912,
+      "grad_norm": 2.203125,
+      "learning_rate": 4.362676458292594e-05,
+      "loss": 0.4797,
+      "step": 5346
+    },
+    {
+      "epoch": 0.23482661339012262,
+      "grad_norm": 2.125,
+      "learning_rate": 4.362214947393892e-05,
+      "loss": 0.4704,
+      "step": 5348
+    },
+    {
+      "epoch": 0.2349144318693261,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.361753293886324e-05,
+      "loss": 0.4758,
+      "step": 5350
+    },
+    {
+      "epoch": 0.2350022503485296,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.361291497805242e-05,
+      "loss": 0.4899,
+      "step": 5352
+    },
+    {
+      "epoch": 0.23509006882773306,
+      "grad_norm": 3.03125,
+      "learning_rate": 4.3608295591860105e-05,
+      "loss": 0.4964,
+      "step": 5354
+    },
+    {
+      "epoch": 0.23517788730693656,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.360367478064006e-05,
+      "loss": 0.4743,
+      "step": 5356
+    },
+    {
+      "epoch": 0.23526570578614006,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.3599052544746136e-05,
+      "loss": 0.4564,
+      "step": 5358
+    },
+    {
+      "epoch": 0.23535352426534353,
+      "grad_norm": 2.21875,
+      "learning_rate": 4.359442888453231e-05,
+      "loss": 0.4697,
+      "step": 5360
+    },
+    {
+      "epoch": 0.23544134274454703,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.3589803800352666e-05,
+      "loss": 0.4618,
+      "step": 5362
+    },
+    {
+      "epoch": 0.2355291612237505,
+      "grad_norm": 2.34375,
+      "learning_rate": 4.3585177292561386e-05,
+      "loss": 0.4454,
+      "step": 5364
+    },
+    {
+      "epoch": 0.235616979702954,
+      "grad_norm": 2.4375,
+      "learning_rate": 4.358054936151278e-05,
+      "loss": 0.4874,
+      "step": 5366
+    },
+    {
+      "epoch": 0.23570479818215748,
+      "grad_norm": 2.328125,
+      "learning_rate": 4.3575920007561245e-05,
+      "loss": 0.4826,
+      "step": 5368
+    },
+    {
+      "epoch": 0.23579261666136098,
+      "grad_norm": 2.421875,
+      "learning_rate": 4.35712892310613e-05,
+      "loss": 0.4856,
+      "step": 5370
+    },
+    {
+      "epoch": 0.23588043514056445,
+      "grad_norm": 2.140625,
+      "learning_rate": 4.356665703236758e-05,
+      "loss": 0.4558,
+      "step": 5372
+    },
+    {
+      "epoch": 0.23596825361976795,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.356202341183481e-05,
+      "loss": 0.4791,
+      "step": 5374
+    },
+    {
+      "epoch": 0.23605607209897142,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.355738836981784e-05,
+      "loss": 0.4504,
+      "step": 5376
+    },
+    {
+      "epoch": 0.23614389057817492,
+      "grad_norm": 3.296875,
+      "learning_rate": 4.3552751906671616e-05,
+      "loss": 0.4624,
+      "step": 5378
+    },
+    {
+      "epoch": 0.2362317090573784,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.3548114022751206e-05,
+      "loss": 0.4574,
+      "step": 5380
+    },
+    {
+      "epoch": 0.2363195275365819,
+      "grad_norm": 2.203125,
+      "learning_rate": 4.354347471841178e-05,
+      "loss": 0.476,
+      "step": 5382
+    },
+    {
+      "epoch": 0.23640734601578536,
+      "grad_norm": 2.3125,
+      "learning_rate": 4.3538833994008614e-05,
+      "loss": 0.4668,
+      "step": 5384
+    },
+    {
+      "epoch": 0.23649516449498886,
+      "grad_norm": 2.265625,
+      "learning_rate": 4.35341918498971e-05,
+      "loss": 0.4589,
+      "step": 5386
+    },
+    {
+      "epoch": 0.23658298297419234,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.3529548286432724e-05,
+      "loss": 0.456,
+      "step": 5388
+    },
+    {
+      "epoch": 0.23667080145339583,
+      "grad_norm": 2.140625,
+      "learning_rate": 4.3524903303971104e-05,
+      "loss": 0.4643,
+      "step": 5390
+    },
+    {
+      "epoch": 0.2367586199325993,
+      "grad_norm": 2.140625,
+      "learning_rate": 4.352025690286795e-05,
+      "loss": 0.4801,
+      "step": 5392
+    },
+    {
+      "epoch": 0.2368464384118028,
+      "grad_norm": 2.1875,
+      "learning_rate": 4.3515609083479066e-05,
+      "loss": 0.4824,
+      "step": 5394
+    },
+    {
+      "epoch": 0.23693425689100628,
+      "grad_norm": 2.359375,
+      "learning_rate": 4.351095984616042e-05,
+      "loss": 0.4977,
+      "step": 5396
+    },
+    {
+      "epoch": 0.23702207537020978,
+      "grad_norm": 2.4375,
+      "learning_rate": 4.350630919126803e-05,
+      "loss": 0.4508,
+      "step": 5398
+    },
+    {
+      "epoch": 0.23710989384941325,
+      "grad_norm": 2.140625,
+      "learning_rate": 4.350165711915803e-05,
+      "loss": 0.453,
+      "step": 5400
+    },
+    {
+      "epoch": 0.23719771232861675,
+      "grad_norm": 2.25,
+      "learning_rate": 4.349700363018671e-05,
+      "loss": 0.4878,
+      "step": 5402
+    },
+    {
+      "epoch": 0.23728553080782022,
+      "grad_norm": 2.25,
+      "learning_rate": 4.349234872471041e-05,
+      "loss": 0.454,
+      "step": 5404
+    },
+    {
+      "epoch": 0.23737334928702372,
+      "grad_norm": 2.546875,
+      "learning_rate": 4.348769240308561e-05,
+      "loss": 0.4635,
+      "step": 5406
+    },
+    {
+      "epoch": 0.23746116776622722,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.34830346656689e-05,
+      "loss": 0.4546,
+      "step": 5408
+    },
+    {
+      "epoch": 0.2375489862454307,
+      "grad_norm": 3.5,
+      "learning_rate": 4.347837551281696e-05,
+      "loss": 0.4531,
+      "step": 5410
+    },
+    {
+      "epoch": 0.2376368047246342,
+      "grad_norm": 3.0,
+      "learning_rate": 4.347371494488659e-05,
+      "loss": 0.4688,
+      "step": 5412
+    },
+    {
+      "epoch": 0.23772462320383766,
+      "grad_norm": 2.375,
+      "learning_rate": 4.346905296223471e-05,
+      "loss": 0.4568,
+      "step": 5414
+    },
+    {
+      "epoch": 0.23781244168304116,
+      "grad_norm": 3.078125,
+      "learning_rate": 4.346438956521832e-05,
+      "loss": 0.4738,
+      "step": 5416
+    },
+    {
+      "epoch": 0.23790026016224464,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.345972475419455e-05,
+      "loss": 0.4641,
+      "step": 5418
+    },
+    {
+      "epoch": 0.23798807864144814,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.345505852952064e-05,
+      "loss": 0.4585,
+      "step": 5420
+    },
+    {
+      "epoch": 0.2380758971206516,
+      "grad_norm": 2.375,
+      "learning_rate": 4.345039089155392e-05,
+      "loss": 0.4776,
+      "step": 5422
+    },
+    {
+      "epoch": 0.2381637155998551,
+      "grad_norm": 2.21875,
+      "learning_rate": 4.344572184065184e-05,
+      "loss": 0.4306,
+      "step": 5424
+    },
+    {
+      "epoch": 0.23825153407905858,
+      "grad_norm": 2.15625,
+      "learning_rate": 4.344105137717197e-05,
+      "loss": 0.455,
+      "step": 5426
+    },
+    {
+      "epoch": 0.23833935255826208,
+      "grad_norm": 2.28125,
+      "learning_rate": 4.343637950147196e-05,
+      "loss": 0.4614,
+      "step": 5428
+    },
+    {
+      "epoch": 0.23842717103746555,
+      "grad_norm": 2.4375,
+      "learning_rate": 4.343170621390958e-05,
+      "loss": 0.453,
+      "step": 5430
+    },
+    {
+      "epoch": 0.23851498951666905,
+      "grad_norm": 2.5,
+      "learning_rate": 4.3427031514842733e-05,
+      "loss": 0.4712,
+      "step": 5432
+    },
+    {
+      "epoch": 0.23860280799587252,
+      "grad_norm": 2.203125,
+      "learning_rate": 4.3422355404629384e-05,
+      "loss": 0.452,
+      "step": 5434
+    },
+    {
+      "epoch": 0.23869062647507602,
+      "grad_norm": 2.25,
+      "learning_rate": 4.3417677883627644e-05,
+      "loss": 0.455,
+      "step": 5436
+    },
+    {
+      "epoch": 0.2387784449542795,
+      "grad_norm": 2.3125,
+      "learning_rate": 4.341299895219572e-05,
+      "loss": 0.4595,
+      "step": 5438
+    },
+    {
+      "epoch": 0.238866263433483,
+      "grad_norm": 2.0625,
+      "learning_rate": 4.340831861069192e-05,
+      "loss": 0.4738,
+      "step": 5440
+    },
+    {
+      "epoch": 0.23895408191268647,
+      "grad_norm": 2.453125,
+      "learning_rate": 4.340363685947467e-05,
+      "loss": 0.4465,
+      "step": 5442
+    },
+    {
+      "epoch": 0.23904190039188997,
+      "grad_norm": 2.421875,
+      "learning_rate": 4.33989536989025e-05,
+      "loss": 0.4713,
+      "step": 5444
+    },
+    {
+      "epoch": 0.23912971887109344,
+      "grad_norm": 2.15625,
+      "learning_rate": 4.3394269129334044e-05,
+      "loss": 0.4876,
+      "step": 5446
+    },
+    {
+      "epoch": 0.23921753735029694,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.338958315112804e-05,
+      "loss": 0.484,
+      "step": 5448
+    },
+    {
+      "epoch": 0.2393053558295004,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.338489576464336e-05,
+      "loss": 0.4481,
+      "step": 5450
+    },
+    {
+      "epoch": 0.2393931743087039,
+      "grad_norm": 2.46875,
+      "learning_rate": 4.338020697023895e-05,
+      "loss": 0.465,
+      "step": 5452
+    },
+    {
+      "epoch": 0.23948099278790738,
+      "grad_norm": 2.234375,
+      "learning_rate": 4.337551676827389e-05,
+      "loss": 0.4872,
+      "step": 5454
+    },
+    {
+      "epoch": 0.23956881126711088,
+      "grad_norm": 2.171875,
+      "learning_rate": 4.337082515910734e-05,
+      "loss": 0.488,
+      "step": 5456
+    },
+    {
+      "epoch": 0.23965662974631438,
+      "grad_norm": 2.546875,
+      "learning_rate": 4.3366132143098606e-05,
+      "loss": 0.4662,
+      "step": 5458
+    },
+    {
+      "epoch": 0.23974444822551785,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.336143772060707e-05,
+      "loss": 0.4747,
+      "step": 5460
+    },
+    {
+      "epoch": 0.23983226670472135,
+      "grad_norm": 3.0,
+      "learning_rate": 4.3356741891992226e-05,
+      "loss": 0.4739,
+      "step": 5462
+    },
+    {
+      "epoch": 0.23992008518392482,
+      "grad_norm": 2.3125,
+      "learning_rate": 4.335204465761369e-05,
+      "loss": 0.427,
+      "step": 5464
+    },
+    {
+      "epoch": 0.24000790366312832,
+      "grad_norm": 2.1875,
+      "learning_rate": 4.334734601783117e-05,
+      "loss": 0.4881,
+      "step": 5466
+    },
+    {
+      "epoch": 0.2400957221423318,
+      "grad_norm": 2.046875,
+      "learning_rate": 4.3342645973004504e-05,
+      "loss": 0.4672,
+      "step": 5468
+    },
+    {
+      "epoch": 0.2401835406215353,
+      "grad_norm": 2.75,
+      "learning_rate": 4.33379445234936e-05,
+      "loss": 0.468,
+      "step": 5470
+    },
+    {
+      "epoch": 0.24027135910073877,
+      "grad_norm": 2.53125,
+      "learning_rate": 4.333324166965852e-05,
+      "loss": 0.491,
+      "step": 5472
+    },
+    {
+      "epoch": 0.24035917757994227,
+      "grad_norm": 2.484375,
+      "learning_rate": 4.3328537411859394e-05,
+      "loss": 0.4633,
+      "step": 5474
+    },
+    {
+      "epoch": 0.24044699605914574,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.332383175045648e-05,
+      "loss": 0.4597,
+      "step": 5476
+    },
+    {
+      "epoch": 0.24053481453834924,
+      "grad_norm": 2.390625,
+      "learning_rate": 4.331912468581013e-05,
+      "loss": 0.453,
+      "step": 5478
+    },
+    {
+      "epoch": 0.2406226330175527,
+      "grad_norm": 2.171875,
+      "learning_rate": 4.331441621828083e-05,
+      "loss": 0.4825,
+      "step": 5480
+    },
+    {
+      "epoch": 0.2407104514967562,
+      "grad_norm": 2.265625,
+      "learning_rate": 4.330970634822914e-05,
+      "loss": 0.5258,
+      "step": 5482
+    },
+    {
+      "epoch": 0.24079826997595968,
+      "grad_norm": 2.46875,
+      "learning_rate": 4.330499507601575e-05,
+      "loss": 0.4577,
+      "step": 5484
+    },
+    {
+      "epoch": 0.24088608845516318,
+      "grad_norm": 2.171875,
+      "learning_rate": 4.330028240200146e-05,
+      "loss": 0.4376,
+      "step": 5486
+    },
+    {
+      "epoch": 0.24097390693436666,
+      "grad_norm": 2.453125,
+      "learning_rate": 4.3295568326547144e-05,
+      "loss": 0.4581,
+      "step": 5488
+    },
+    {
+      "epoch": 0.24106172541357015,
+      "grad_norm": 2.25,
+      "learning_rate": 4.329085285001382e-05,
+      "loss": 0.4796,
+      "step": 5490
+    },
+    {
+      "epoch": 0.24114954389277363,
+      "grad_norm": 2.1875,
+      "learning_rate": 4.3286135972762596e-05,
+      "loss": 0.4707,
+      "step": 5492
+    },
+    {
+      "epoch": 0.24123736237197713,
+      "grad_norm": 2.296875,
+      "learning_rate": 4.328141769515471e-05,
+      "loss": 0.4675,
+      "step": 5494
+    },
+    {
+      "epoch": 0.2413251808511806,
+      "grad_norm": 2.109375,
+      "learning_rate": 4.3276698017551464e-05,
+      "loss": 0.4809,
+      "step": 5496
+    },
+    {
+      "epoch": 0.2414129993303841,
+      "grad_norm": 3.109375,
+      "learning_rate": 4.327197694031431e-05,
+      "loss": 0.4557,
+      "step": 5498
+    },
+    {
+      "epoch": 0.24150081780958757,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.3267254463804775e-05,
+      "loss": 0.4709,
+      "step": 5500
+    },
+    {
+      "epoch": 0.24158863628879107,
+      "grad_norm": 2.375,
+      "learning_rate": 4.326253058838452e-05,
+      "loss": 0.4782,
+      "step": 5502
+    },
+    {
+      "epoch": 0.24167645476799454,
+      "grad_norm": 2.171875,
+      "learning_rate": 4.325780531441529e-05,
+      "loss": 0.4655,
+      "step": 5504
+    },
+    {
+      "epoch": 0.24176427324719804,
+      "grad_norm": 2.28125,
+      "learning_rate": 4.325307864225895e-05,
+      "loss": 0.4288,
+      "step": 5506
+    },
+    {
+      "epoch": 0.24185209172640154,
+      "grad_norm": 2.265625,
+      "learning_rate": 4.324835057227748e-05,
+      "loss": 0.4439,
+      "step": 5508
+    },
+    {
+      "epoch": 0.241939910205605,
+      "grad_norm": 2.203125,
+      "learning_rate": 4.324362110483294e-05,
+      "loss": 0.4501,
+      "step": 5510
+    },
+    {
+      "epoch": 0.2420277286848085,
+      "grad_norm": 2.390625,
+      "learning_rate": 4.3238890240287536e-05,
+      "loss": 0.4746,
+      "step": 5512
+    },
+    {
+      "epoch": 0.24211554716401198,
+      "grad_norm": 2.390625,
+      "learning_rate": 4.323415797900353e-05,
+      "loss": 0.4878,
+      "step": 5514
+    },
+    {
+      "epoch": 0.24220336564321548,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.322942432134335e-05,
+      "loss": 0.4404,
+      "step": 5516
+    },
+    {
+      "epoch": 0.24229118412241896,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.322468926766947e-05,
+      "loss": 0.4871,
+      "step": 5518
+    },
+    {
+      "epoch": 0.24237900260162246,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.321995281834452e-05,
+      "loss": 0.4665,
+      "step": 5520
+    },
+    {
+      "epoch": 0.24246682108082593,
+      "grad_norm": 2.5,
+      "learning_rate": 4.3215214973731225e-05,
+      "loss": 0.4825,
+      "step": 5522
+    },
+    {
+      "epoch": 0.24255463956002943,
+      "grad_norm": 2.3125,
+      "learning_rate": 4.32104757341924e-05,
+      "loss": 0.4422,
+      "step": 5524
+    },
+    {
+      "epoch": 0.2426424580392329,
+      "grad_norm": 2.25,
+      "learning_rate": 4.320573510009097e-05,
+      "loss": 0.4534,
+      "step": 5526
+    },
+    {
+      "epoch": 0.2427302765184364,
+      "grad_norm": 2.15625,
+      "learning_rate": 4.320099307178999e-05,
+      "loss": 0.4488,
+      "step": 5528
+    },
+    {
+      "epoch": 0.24281809499763987,
+      "grad_norm": 2.203125,
+      "learning_rate": 4.3196249649652585e-05,
+      "loss": 0.4662,
+      "step": 5530
+    },
+    {
+      "epoch": 0.24290591347684337,
+      "grad_norm": 2.3125,
+      "learning_rate": 4.319150483404203e-05,
+      "loss": 0.4582,
+      "step": 5532
+    },
+    {
+      "epoch": 0.24299373195604684,
+      "grad_norm": 2.1875,
+      "learning_rate": 4.318675862532167e-05,
+      "loss": 0.4754,
+      "step": 5534
+    },
+    {
+      "epoch": 0.24308155043525034,
+      "grad_norm": 2.375,
+      "learning_rate": 4.318201102385497e-05,
+      "loss": 0.4612,
+      "step": 5536
+    },
+    {
+      "epoch": 0.24316936891445382,
+      "grad_norm": 2.359375,
+      "learning_rate": 4.317726203000552e-05,
+      "loss": 0.4556,
+      "step": 5538
+    },
+    {
+      "epoch": 0.24325718739365731,
+      "grad_norm": 2.328125,
+      "learning_rate": 4.317251164413698e-05,
+      "loss": 0.5085,
+      "step": 5540
+    },
+    {
+      "epoch": 0.2433450058728608,
+      "grad_norm": 2.171875,
+      "learning_rate": 4.316775986661314e-05,
+      "loss": 0.4796,
+      "step": 5542
+    },
+    {
+      "epoch": 0.2434328243520643,
+      "grad_norm": 2.421875,
+      "learning_rate": 4.31630066977979e-05,
+      "loss": 0.4599,
+      "step": 5544
+    },
+    {
+      "epoch": 0.24352064283126776,
+      "grad_norm": 2.046875,
+      "learning_rate": 4.315825213805525e-05,
+      "loss": 0.4443,
+      "step": 5546
+    },
+    {
+      "epoch": 0.24360846131047126,
+      "grad_norm": 2.1875,
+      "learning_rate": 4.31534961877493e-05,
+      "loss": 0.4585,
+      "step": 5548
+    },
+    {
+      "epoch": 0.24369627978967473,
+      "grad_norm": 2.359375,
+      "learning_rate": 4.314873884724425e-05,
+      "loss": 0.4578,
+      "step": 5550
+    },
+    {
+      "epoch": 0.24378409826887823,
+      "grad_norm": 2.40625,
+      "learning_rate": 4.3143980116904436e-05,
+      "loss": 0.4922,
+      "step": 5552
+    },
+    {
+      "epoch": 0.2438719167480817,
+      "grad_norm": 2.25,
+      "learning_rate": 4.313921999709428e-05,
+      "loss": 0.4907,
+      "step": 5554
+    },
+    {
+      "epoch": 0.2439597352272852,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.313445848817831e-05,
+      "loss": 0.4842,
+      "step": 5556
+    },
+    {
+      "epoch": 0.2440475537064887,
+      "grad_norm": 2.3125,
+      "learning_rate": 4.312969559052115e-05,
+      "loss": 0.4853,
+      "step": 5558
+    },
+    {
+      "epoch": 0.24413537218569217,
+      "grad_norm": 2.4375,
+      "learning_rate": 4.312493130448756e-05,
+      "loss": 0.4377,
+      "step": 5560
+    },
+    {
+      "epoch": 0.24422319066489567,
+      "grad_norm": 2.375,
+      "learning_rate": 4.312016563044239e-05,
+      "loss": 0.4781,
+      "step": 5562
+    },
+    {
+      "epoch": 0.24431100914409914,
+      "grad_norm": 2.5,
+      "learning_rate": 4.311539856875059e-05,
+      "loss": 0.4625,
+      "step": 5564
+    },
+    {
+      "epoch": 0.24439882762330264,
+      "grad_norm": 2.34375,
+      "learning_rate": 4.311063011977723e-05,
+      "loss": 0.4681,
+      "step": 5566
+    },
+    {
+      "epoch": 0.24448664610250612,
+      "grad_norm": 2.421875,
+      "learning_rate": 4.3105860283887464e-05,
+      "loss": 0.4618,
+      "step": 5568
+    },
+    {
+      "epoch": 0.24457446458170962,
+      "grad_norm": 2.75,
+      "learning_rate": 4.3101089061446585e-05,
+      "loss": 0.4472,
+      "step": 5570
+    },
+    {
+      "epoch": 0.2446622830609131,
+      "grad_norm": 2.296875,
+      "learning_rate": 4.3096316452819964e-05,
+      "loss": 0.4859,
+      "step": 5572
+    },
+    {
+      "epoch": 0.2447501015401166,
+      "grad_norm": 2.265625,
+      "learning_rate": 4.309154245837309e-05,
+      "loss": 0.4876,
+      "step": 5574
+    },
+    {
+      "epoch": 0.24483792001932006,
+      "grad_norm": 2.515625,
+      "learning_rate": 4.308676707847156e-05,
+      "loss": 0.4829,
+      "step": 5576
+    },
+    {
+      "epoch": 0.24492573849852356,
+      "grad_norm": 2.203125,
+      "learning_rate": 4.308199031348107e-05,
+      "loss": 0.4657,
+      "step": 5578
+    },
+    {
+      "epoch": 0.24501355697772703,
+      "grad_norm": 2.390625,
+      "learning_rate": 4.3077212163767425e-05,
+      "loss": 0.4184,
+      "step": 5580
+    },
+    {
+      "epoch": 0.24510137545693053,
+      "grad_norm": 2.515625,
+      "learning_rate": 4.307243262969654e-05,
+      "loss": 0.471,
+      "step": 5582
+    },
+    {
+      "epoch": 0.245189193936134,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.306765171163443e-05,
+      "loss": 0.4686,
+      "step": 5584
+    },
+    {
+      "epoch": 0.2452770124153375,
+      "grad_norm": 1.9765625,
+      "learning_rate": 4.306286940994723e-05,
+      "loss": 0.48,
+      "step": 5586
+    },
+    {
+      "epoch": 0.24536483089454098,
+      "grad_norm": 2.421875,
+      "learning_rate": 4.3058085725001154e-05,
+      "loss": 0.4761,
+      "step": 5588
+    },
+    {
+      "epoch": 0.24545264937374447,
+      "grad_norm": 2.03125,
+      "learning_rate": 4.305330065716254e-05,
+      "loss": 0.492,
+      "step": 5590
+    },
+    {
+      "epoch": 0.24554046785294795,
+      "grad_norm": 2.28125,
+      "learning_rate": 4.304851420679784e-05,
+      "loss": 0.466,
+      "step": 5592
+    },
+    {
+      "epoch": 0.24562828633215145,
+      "grad_norm": 2.15625,
+      "learning_rate": 4.30437263742736e-05,
+      "loss": 0.4475,
+      "step": 5594
+    },
+    {
+      "epoch": 0.24571610481135492,
+      "grad_norm": 2.328125,
+      "learning_rate": 4.303893715995646e-05,
+      "loss": 0.459,
+      "step": 5596
+    },
+    {
+      "epoch": 0.24580392329055842,
+      "grad_norm": 2.296875,
+      "learning_rate": 4.303414656421319e-05,
+      "loss": 0.4527,
+      "step": 5598
+    },
+    {
+      "epoch": 0.2458917417697619,
+      "grad_norm": 2.265625,
+      "learning_rate": 4.302935458741066e-05,
+      "loss": 0.4525,
+      "step": 5600
+    },
+    {
+      "epoch": 0.2459795602489654,
+      "grad_norm": 2.53125,
+      "learning_rate": 4.3024561229915826e-05,
+      "loss": 0.4693,
+      "step": 5602
+    },
+    {
+      "epoch": 0.24606737872816886,
+      "grad_norm": 2.375,
+      "learning_rate": 4.301976649209577e-05,
+      "loss": 0.4656,
+      "step": 5604
+    },
+    {
+      "epoch": 0.24615519720737236,
+      "grad_norm": 2.1875,
+      "learning_rate": 4.301497037431769e-05,
+      "loss": 0.4751,
+      "step": 5606
+    },
+    {
+      "epoch": 0.24624301568657586,
+      "grad_norm": 2.5,
+      "learning_rate": 4.3010172876948844e-05,
+      "loss": 0.463,
+      "step": 5608
+    },
+    {
+      "epoch": 0.24633083416577933,
+      "grad_norm": 2.203125,
+      "learning_rate": 4.300537400035665e-05,
+      "loss": 0.4701,
+      "step": 5610
+    },
+    {
+      "epoch": 0.24641865264498283,
+      "grad_norm": 2.1875,
+      "learning_rate": 4.30005737449086e-05,
+      "loss": 0.4647,
+      "step": 5612
+    },
+    {
+      "epoch": 0.2465064711241863,
+      "grad_norm": 2.203125,
+      "learning_rate": 4.2995772110972296e-05,
+      "loss": 0.4869,
+      "step": 5614
+    },
+    {
+      "epoch": 0.2465942896033898,
+      "grad_norm": 2.328125,
+      "learning_rate": 4.299096909891545e-05,
+      "loss": 0.445,
+      "step": 5616
+    },
+    {
+      "epoch": 0.24668210808259328,
+      "grad_norm": 2.296875,
+      "learning_rate": 4.2986164709105877e-05,
+      "loss": 0.475,
+      "step": 5618
+    },
+    {
+      "epoch": 0.24676992656179678,
+      "grad_norm": 2.1875,
+      "learning_rate": 4.29813589419115e-05,
+      "loss": 0.4494,
+      "step": 5620
+    },
+    {
+      "epoch": 0.24685774504100025,
+      "grad_norm": 2.265625,
+      "learning_rate": 4.2976551797700336e-05,
+      "loss": 0.4499,
+      "step": 5622
+    },
+    {
+      "epoch": 0.24694556352020375,
+      "grad_norm": 2.5,
+      "learning_rate": 4.297174327684054e-05,
+      "loss": 0.4804,
+      "step": 5624
+    },
+    {
+      "epoch": 0.24703338199940722,
+      "grad_norm": 2.421875,
+      "learning_rate": 4.296693337970033e-05,
+      "loss": 0.4644,
+      "step": 5626
+    },
+    {
+      "epoch": 0.24712120047861072,
+      "grad_norm": 2.625,
+      "learning_rate": 4.296212210664805e-05,
+      "loss": 0.4614,
+      "step": 5628
+    },
+    {
+      "epoch": 0.2472090189578142,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.2957309458052156e-05,
+      "loss": 0.4825,
+      "step": 5630
+    },
+    {
+      "epoch": 0.2472968374370177,
+      "grad_norm": 2.390625,
+      "learning_rate": 4.2952495434281204e-05,
+      "loss": 0.4424,
+      "step": 5632
+    },
+    {
+      "epoch": 0.24738465591622116,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.294768003570384e-05,
+      "loss": 0.486,
+      "step": 5634
+    },
+    {
+      "epoch": 0.24747247439542466,
+      "grad_norm": 2.078125,
+      "learning_rate": 4.294286326268885e-05,
+      "loss": 0.4528,
+      "step": 5636
+    },
+    {
+      "epoch": 0.24756029287462814,
+      "grad_norm": 2.46875,
+      "learning_rate": 4.2938045115605074e-05,
+      "loss": 0.4554,
+      "step": 5638
+    },
+    {
+      "epoch": 0.24764811135383163,
+      "grad_norm": 2.234375,
+      "learning_rate": 4.2933225594821514e-05,
+      "loss": 0.4572,
+      "step": 5640
+    },
+    {
+      "epoch": 0.2477359298330351,
+      "grad_norm": 2.15625,
+      "learning_rate": 4.292840470070724e-05,
+      "loss": 0.4692,
+      "step": 5642
+    },
+    {
+      "epoch": 0.2478237483122386,
+      "grad_norm": 2.21875,
+      "learning_rate": 4.2923582433631424e-05,
+      "loss": 0.4767,
+      "step": 5644
+    },
+    {
+      "epoch": 0.24791156679144208,
+      "grad_norm": 2.484375,
+      "learning_rate": 4.291875879396338e-05,
+      "loss": 0.4665,
+      "step": 5646
+    },
+    {
+      "epoch": 0.24799938527064558,
+      "grad_norm": 2.53125,
+      "learning_rate": 4.291393378207249e-05,
+      "loss": 0.4811,
+      "step": 5648
+    },
+    {
+      "epoch": 0.24808720374984905,
+      "grad_norm": 2.09375,
+      "learning_rate": 4.290910739832825e-05,
+      "loss": 0.4653,
+      "step": 5650
+    },
+    {
+      "epoch": 0.24817502222905255,
+      "grad_norm": 2.234375,
+      "learning_rate": 4.2904279643100276e-05,
+      "loss": 0.4598,
+      "step": 5652
+    },
+    {
+      "epoch": 0.24826284070825602,
+      "grad_norm": 2.390625,
+      "learning_rate": 4.2899450516758275e-05,
+      "loss": 0.4546,
+      "step": 5654
+    },
+    {
+      "epoch": 0.24835065918745952,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.289462001967207e-05,
+      "loss": 0.4407,
+      "step": 5656
+    },
+    {
+      "epoch": 0.24843847766666302,
+      "grad_norm": 3.125,
+      "learning_rate": 4.288978815221157e-05,
+      "loss": 0.466,
+      "step": 5658
+    },
+    {
+      "epoch": 0.2485262961458665,
+      "grad_norm": 3.328125,
+      "learning_rate": 4.28849549147468e-05,
+      "loss": 0.4913,
+      "step": 5660
+    },
+    {
+      "epoch": 0.24861411462507,
+      "grad_norm": 3.484375,
+      "learning_rate": 4.28801203076479e-05,
+      "loss": 0.4755,
+      "step": 5662
+    },
+    {
+      "epoch": 0.24870193310427346,
+      "grad_norm": 3.09375,
+      "learning_rate": 4.2875284331285105e-05,
+      "loss": 0.4563,
+      "step": 5664
+    },
+    {
+      "epoch": 0.24878975158347696,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.287044698602874e-05,
+      "loss": 0.4635,
+      "step": 5666
+    },
+    {
+      "epoch": 0.24887757006268044,
+      "grad_norm": 2.203125,
+      "learning_rate": 4.286560827224927e-05,
+      "loss": 0.4683,
+      "step": 5668
+    },
+    {
+      "epoch": 0.24896538854188394,
+      "grad_norm": 2.15625,
+      "learning_rate": 4.286076819031723e-05,
+      "loss": 0.4425,
+      "step": 5670
+    },
+    {
+      "epoch": 0.2490532070210874,
+      "grad_norm": 2.390625,
+      "learning_rate": 4.285592674060328e-05,
+      "loss": 0.4542,
+      "step": 5672
+    },
+    {
+      "epoch": 0.2491410255002909,
+      "grad_norm": 2.53125,
+      "learning_rate": 4.2851083923478186e-05,
+      "loss": 0.4687,
+      "step": 5674
+    },
+    {
+      "epoch": 0.24922884397949438,
+      "grad_norm": 2.140625,
+      "learning_rate": 4.2846239739312796e-05,
+      "loss": 0.4361,
+      "step": 5676
+    },
+    {
+      "epoch": 0.24931666245869788,
+      "grad_norm": 2.21875,
+      "learning_rate": 4.28413941884781e-05,
+      "loss": 0.4518,
+      "step": 5678
+    },
+    {
+      "epoch": 0.24940448093790135,
+      "grad_norm": 2.53125,
+      "learning_rate": 4.283654727134515e-05,
+      "loss": 0.47,
+      "step": 5680
+    },
+    {
+      "epoch": 0.24949229941710485,
+      "grad_norm": 3.09375,
+      "learning_rate": 4.2831698988285144e-05,
+      "loss": 0.4626,
+      "step": 5682
+    },
+    {
+      "epoch": 0.24958011789630832,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.282684933966935e-05,
+      "loss": 0.4662,
+      "step": 5684
+    },
+    {
+      "epoch": 0.24966793637551182,
+      "grad_norm": 2.296875,
+      "learning_rate": 4.282199832586916e-05,
+      "loss": 0.4362,
+      "step": 5686
+    },
+    {
+      "epoch": 0.2497557548547153,
+      "grad_norm": 2.328125,
+      "learning_rate": 4.2817145947256064e-05,
+      "loss": 0.4661,
+      "step": 5688
+    },
+    {
+      "epoch": 0.2498435733339188,
+      "grad_norm": 2.40625,
+      "learning_rate": 4.281229220420167e-05,
+      "loss": 0.4632,
+      "step": 5690
+    },
+    {
+      "epoch": 0.24993139181312227,
+      "grad_norm": 2.40625,
+      "learning_rate": 4.2807437097077654e-05,
+      "loss": 0.4642,
+      "step": 5692
+    },
+    {
+      "epoch": 0.25001921029232577,
+      "grad_norm": 2.28125,
+      "learning_rate": 4.280258062625585e-05,
+      "loss": 0.4546,
+      "step": 5694
+    },
+    {
+      "epoch": 0.25010702877152924,
+      "grad_norm": 2.265625,
+      "learning_rate": 4.279772279210814e-05,
+      "loss": 0.441,
+      "step": 5696
+    },
+    {
+      "epoch": 0.2501948472507327,
+      "grad_norm": 2.1875,
+      "learning_rate": 4.279286359500657e-05,
+      "loss": 0.4551,
+      "step": 5698
+    },
+    {
+      "epoch": 0.25028266572993624,
+      "grad_norm": 2.359375,
+      "learning_rate": 4.2788003035323225e-05,
+      "loss": 0.4417,
+      "step": 5700
+    },
+    {
+      "epoch": 0.2503704842091397,
+      "grad_norm": 2.078125,
+      "learning_rate": 4.278314111343035e-05,
+      "loss": 0.4275,
+      "step": 5702
+    },
+    {
+      "epoch": 0.2504583026883432,
+      "grad_norm": 2.15625,
+      "learning_rate": 4.277827782970026e-05,
+      "loss": 0.4649,
+      "step": 5704
+    },
+    {
+      "epoch": 0.25054612116754665,
+      "grad_norm": 2.453125,
+      "learning_rate": 4.277341318450541e-05,
+      "loss": 0.4618,
+      "step": 5706
+    },
+    {
+      "epoch": 0.2506339396467502,
+      "grad_norm": 2.4375,
+      "learning_rate": 4.27685471782183e-05,
+      "loss": 0.4681,
+      "step": 5708
+    },
+    {
+      "epoch": 0.25072175812595365,
+      "grad_norm": 2.21875,
+      "learning_rate": 4.276367981121159e-05,
+      "loss": 0.4521,
+      "step": 5710
+    },
+    {
+      "epoch": 0.2508095766051571,
+      "grad_norm": 2.4375,
+      "learning_rate": 4.275881108385802e-05,
+      "loss": 0.455,
+      "step": 5712
+    },
+    {
+      "epoch": 0.25089739508436065,
+      "grad_norm": 2.109375,
+      "learning_rate": 4.275394099653045e-05,
+      "loss": 0.4738,
+      "step": 5714
+    },
+    {
+      "epoch": 0.2509852135635641,
+      "grad_norm": 2.109375,
+      "learning_rate": 4.2749069549601816e-05,
+      "loss": 0.4609,
+      "step": 5716
+    },
+    {
+      "epoch": 0.2510730320427676,
+      "grad_norm": 2.171875,
+      "learning_rate": 4.274419674344519e-05,
+      "loss": 0.4693,
+      "step": 5718
+    },
+    {
+      "epoch": 0.25116085052197107,
+      "grad_norm": 2.40625,
+      "learning_rate": 4.273932257843371e-05,
+      "loss": 0.4611,
+      "step": 5720
+    },
+    {
+      "epoch": 0.2512486690011746,
+      "grad_norm": 2.1875,
+      "learning_rate": 4.273444705494066e-05,
+      "loss": 0.4809,
+      "step": 5722
+    },
+    {
+      "epoch": 0.25133648748037807,
+      "grad_norm": 2.390625,
+      "learning_rate": 4.272957017333941e-05,
+      "loss": 0.446,
+      "step": 5724
+    },
+    {
+      "epoch": 0.25142430595958154,
+      "grad_norm": 2.265625,
+      "learning_rate": 4.2724691934003414e-05,
+      "loss": 0.463,
+      "step": 5726
+    },
+    {
+      "epoch": 0.251512124438785,
+      "grad_norm": 2.203125,
+      "learning_rate": 4.271981233730626e-05,
+      "loss": 0.4614,
+      "step": 5728
+    },
+    {
+      "epoch": 0.25159994291798854,
+      "grad_norm": 2.15625,
+      "learning_rate": 4.271493138362165e-05,
+      "loss": 0.4615,
+      "step": 5730
+    },
+    {
+      "epoch": 0.251687761397192,
+      "grad_norm": 2.40625,
+      "learning_rate": 4.271004907332333e-05,
+      "loss": 0.4866,
+      "step": 5732
+    },
+    {
+      "epoch": 0.2517755798763955,
+      "grad_norm": 2.453125,
+      "learning_rate": 4.270516540678521e-05,
+      "loss": 0.473,
+      "step": 5734
+    },
+    {
+      "epoch": 0.25186339835559896,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.2700280384381275e-05,
+      "loss": 0.4526,
+      "step": 5736
+    },
+    {
+      "epoch": 0.2519512168348025,
+      "grad_norm": 2.5,
+      "learning_rate": 4.269539400648563e-05,
+      "loss": 0.4582,
+      "step": 5738
+    },
+    {
+      "epoch": 0.25203903531400595,
+      "grad_norm": 2.328125,
+      "learning_rate": 4.269050627347247e-05,
+      "loss": 0.4638,
+      "step": 5740
+    },
+    {
+      "epoch": 0.2521268537932094,
+      "grad_norm": 1.9609375,
+      "learning_rate": 4.2685617185716104e-05,
+      "loss": 0.4635,
+      "step": 5742
+    },
+    {
+      "epoch": 0.2522146722724129,
+      "grad_norm": 2.15625,
+      "learning_rate": 4.268072674359093e-05,
+      "loss": 0.4475,
+      "step": 5744
+    },
+    {
+      "epoch": 0.2523024907516164,
+      "grad_norm": 2.375,
+      "learning_rate": 4.267583494747146e-05,
+      "loss": 0.4297,
+      "step": 5746
+    },
+    {
+      "epoch": 0.2523903092308199,
+      "grad_norm": 2.421875,
+      "learning_rate": 4.267094179773232e-05,
+      "loss": 0.4615,
+      "step": 5748
+    },
+    {
+      "epoch": 0.25247812771002337,
+      "grad_norm": 2.515625,
+      "learning_rate": 4.266604729474821e-05,
+      "loss": 0.4567,
+      "step": 5750
+    },
+    {
+      "epoch": 0.25256594618922684,
+      "grad_norm": 2.390625,
+      "learning_rate": 4.2661151438893974e-05,
+      "loss": 0.4569,
+      "step": 5752
+    },
+    {
+      "epoch": 0.25265376466843037,
+      "grad_norm": 2.390625,
+      "learning_rate": 4.265625423054452e-05,
+      "loss": 0.4475,
+      "step": 5754
+    },
+    {
+      "epoch": 0.25274158314763384,
+      "grad_norm": 2.4375,
+      "learning_rate": 4.265135567007489e-05,
+      "loss": 0.4421,
+      "step": 5756
+    },
+    {
+      "epoch": 0.2528294016268373,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.264645575786021e-05,
+      "loss": 0.4542,
+      "step": 5758
+    },
+    {
+      "epoch": 0.25291722010604084,
+      "grad_norm": 2.109375,
+      "learning_rate": 4.264155449427572e-05,
+      "loss": 0.4345,
+      "step": 5760
+    },
+    {
+      "epoch": 0.2530050385852443,
+      "grad_norm": 2.21875,
+      "learning_rate": 4.263665187969675e-05,
+      "loss": 0.4468,
+      "step": 5762
+    },
+    {
+      "epoch": 0.2530928570644478,
+      "grad_norm": 2.25,
+      "learning_rate": 4.263174791449875e-05,
+      "loss": 0.4696,
+      "step": 5764
+    },
+    {
+      "epoch": 0.25318067554365126,
+      "grad_norm": 2.21875,
+      "learning_rate": 4.262684259905728e-05,
+      "loss": 0.4805,
+      "step": 5766
+    },
+    {
+      "epoch": 0.2532684940228548,
+      "grad_norm": 2.359375,
+      "learning_rate": 4.2621935933747974e-05,
+      "loss": 0.4899,
+      "step": 5768
+    },
+    {
+      "epoch": 0.25335631250205826,
+      "grad_norm": 2.203125,
+      "learning_rate": 4.261702791894659e-05,
+      "loss": 0.4855,
+      "step": 5770
+    },
+    {
+      "epoch": 0.25344413098126173,
+      "grad_norm": 2.203125,
+      "learning_rate": 4.261211855502898e-05,
+      "loss": 0.4947,
+      "step": 5772
+    },
+    {
+      "epoch": 0.2535319494604652,
+      "grad_norm": 2.140625,
+      "learning_rate": 4.260720784237111e-05,
+      "loss": 0.457,
+      "step": 5774
+    },
+    {
+      "epoch": 0.2536197679396687,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.260229578134904e-05,
+      "loss": 0.4763,
+      "step": 5776
+    },
+    {
+      "epoch": 0.2537075864188722,
+      "grad_norm": 2.28125,
+      "learning_rate": 4.259738237233896e-05,
+      "loss": 0.4496,
+      "step": 5778
+    },
+    {
+      "epoch": 0.25379540489807567,
+      "grad_norm": 2.140625,
+      "learning_rate": 4.2592467615717105e-05,
+      "loss": 0.4484,
+      "step": 5780
+    },
+    {
+      "epoch": 0.25388322337727914,
+      "grad_norm": 2.109375,
+      "learning_rate": 4.258755151185986e-05,
+      "loss": 0.4883,
+      "step": 5782
+    },
+    {
+      "epoch": 0.25397104185648267,
+      "grad_norm": 2.53125,
+      "learning_rate": 4.25826340611437e-05,
+      "loss": 0.4727,
+      "step": 5784
+    },
+    {
+      "epoch": 0.25405886033568614,
+      "grad_norm": 2.46875,
+      "learning_rate": 4.2577715263945216e-05,
+      "loss": 0.4786,
+      "step": 5786
+    },
+    {
+      "epoch": 0.2541466788148896,
+      "grad_norm": 2.53125,
+      "learning_rate": 4.2572795120641086e-05,
+      "loss": 0.4564,
+      "step": 5788
+    },
+    {
+      "epoch": 0.2542344972940931,
+      "grad_norm": 2.34375,
+      "learning_rate": 4.256787363160809e-05,
+      "loss": 0.477,
+      "step": 5790
+    },
+    {
+      "epoch": 0.2543223157732966,
+      "grad_norm": 2.234375,
+      "learning_rate": 4.256295079722311e-05,
+      "loss": 0.4591,
+      "step": 5792
+    },
+    {
+      "epoch": 0.2544101342525001,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.255802661786316e-05,
+      "loss": 0.4847,
+      "step": 5794
+    },
+    {
+      "epoch": 0.25449795273170356,
+      "grad_norm": 2.96875,
+      "learning_rate": 4.2553101093905325e-05,
+      "loss": 0.4582,
+      "step": 5796
+    },
+    {
+      "epoch": 0.25458577121090703,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.2548174225726795e-05,
+      "loss": 0.4743,
+      "step": 5798
+    },
+    {
+      "epoch": 0.25467358969011056,
+      "grad_norm": 2.3125,
+      "learning_rate": 4.254324601370487e-05,
+      "loss": 0.4626,
+      "step": 5800
+    },
+    {
+      "epoch": 0.25476140816931403,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.253831645821697e-05,
+      "loss": 0.453,
+      "step": 5802
+    },
+    {
+      "epoch": 0.2548492266485175,
+      "grad_norm": 2.3125,
+      "learning_rate": 4.253338555964059e-05,
+      "loss": 0.455,
+      "step": 5804
+    },
+    {
+      "epoch": 0.254937045127721,
+      "grad_norm": 2.265625,
+      "learning_rate": 4.252845331835333e-05,
+      "loss": 0.447,
+      "step": 5806
+    },
+    {
+      "epoch": 0.2550248636069245,
+      "grad_norm": 2.265625,
+      "learning_rate": 4.252351973473293e-05,
+      "loss": 0.4533,
+      "step": 5808
+    },
+    {
+      "epoch": 0.255112682086128,
+      "grad_norm": 2.234375,
+      "learning_rate": 4.251858480915718e-05,
+      "loss": 0.4477,
+      "step": 5810
+    },
+    {
+      "epoch": 0.25520050056533145,
+      "grad_norm": 2.359375,
+      "learning_rate": 4.2513648542004006e-05,
+      "loss": 0.4466,
+      "step": 5812
+    },
+    {
+      "epoch": 0.255288319044535,
+      "grad_norm": 2.296875,
+      "learning_rate": 4.250871093365143e-05,
+      "loss": 0.4656,
+      "step": 5814
+    },
+    {
+      "epoch": 0.25537613752373844,
+      "grad_norm": 2.390625,
+      "learning_rate": 4.250377198447757e-05,
+      "loss": 0.4657,
+      "step": 5816
+    },
+    {
+      "epoch": 0.2554639560029419,
+      "grad_norm": 2.1875,
+      "learning_rate": 4.249883169486066e-05,
+      "loss": 0.4741,
+      "step": 5818
+    },
+    {
+      "epoch": 0.2555517744821454,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.249389006517902e-05,
+      "loss": 0.4608,
+      "step": 5820
+    },
+    {
+      "epoch": 0.2556395929613489,
+      "grad_norm": 2.125,
+      "learning_rate": 4.248894709581108e-05,
+      "loss": 0.4406,
+      "step": 5822
+    },
+    {
+      "epoch": 0.2557274114405524,
+      "grad_norm": 2.421875,
+      "learning_rate": 4.248400278713539e-05,
+      "loss": 0.4559,
+      "step": 5824
+    },
+    {
+      "epoch": 0.25581522991975586,
+      "grad_norm": 2.984375,
+      "learning_rate": 4.247905713953057e-05,
+      "loss": 0.441,
+      "step": 5826
+    },
+    {
+      "epoch": 0.25590304839895933,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.247411015337537e-05,
+      "loss": 0.464,
+      "step": 5828
+    },
+    {
+      "epoch": 0.25599086687816286,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.246916182904862e-05,
+      "loss": 0.4813,
+      "step": 5830
+    },
+    {
+      "epoch": 0.25607868535736633,
+      "grad_norm": 2.515625,
+      "learning_rate": 4.246421216692926e-05,
+      "loss": 0.4666,
+      "step": 5832
+    },
+    {
+      "epoch": 0.2561665038365698,
+      "grad_norm": 2.34375,
+      "learning_rate": 4.245926116739636e-05,
+      "loss": 0.4726,
+      "step": 5834
+    },
+    {
+      "epoch": 0.2562543223157733,
+      "grad_norm": 2.28125,
+      "learning_rate": 4.245430883082905e-05,
+      "loss": 0.4512,
+      "step": 5836
+    },
+    {
+      "epoch": 0.2563421407949768,
+      "grad_norm": 3.1875,
+      "learning_rate": 4.2449355157606584e-05,
+      "loss": 0.4321,
+      "step": 5838
+    },
+    {
+      "epoch": 0.2564299592741803,
+      "grad_norm": 2.875,
+      "learning_rate": 4.244440014810832e-05,
+      "loss": 0.4447,
+      "step": 5840
+    },
+    {
+      "epoch": 0.25651777775338375,
+      "grad_norm": 3.125,
+      "learning_rate": 4.243944380271372e-05,
+      "loss": 0.4621,
+      "step": 5842
+    },
+    {
+      "epoch": 0.2566055962325872,
+      "grad_norm": 2.625,
+      "learning_rate": 4.243448612180232e-05,
+      "loss": 0.445,
+      "step": 5844
+    },
+    {
+      "epoch": 0.25669341471179075,
+      "grad_norm": 2.875,
+      "learning_rate": 4.242952710575379e-05,
+      "loss": 0.4493,
+      "step": 5846
+    },
+    {
+      "epoch": 0.2567812331909942,
+      "grad_norm": 3.1875,
+      "learning_rate": 4.24245667549479e-05,
+      "loss": 0.472,
+      "step": 5848
+    },
+    {
+      "epoch": 0.2568690516701977,
+      "grad_norm": 3.671875,
+      "learning_rate": 4.241960506976452e-05,
+      "loss": 0.4749,
+      "step": 5850
+    },
+    {
+      "epoch": 0.25695687014940116,
+      "grad_norm": 3.140625,
+      "learning_rate": 4.2414642050583595e-05,
+      "loss": 0.4881,
+      "step": 5852
+    },
+    {
+      "epoch": 0.2570446886286047,
+      "grad_norm": 3.03125,
+      "learning_rate": 4.240967769778522e-05,
+      "loss": 0.4714,
+      "step": 5854
+    },
+    {
+      "epoch": 0.25713250710780816,
+      "grad_norm": 2.234375,
+      "learning_rate": 4.240471201174955e-05,
+      "loss": 0.424,
+      "step": 5856
+    },
+    {
+      "epoch": 0.25722032558701163,
+      "grad_norm": 2.28125,
+      "learning_rate": 4.239974499285686e-05,
+      "loss": 0.4458,
+      "step": 5858
+    },
+    {
+      "epoch": 0.25730814406621516,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.2394776641487525e-05,
+      "loss": 0.4459,
+      "step": 5860
+    },
+    {
+      "epoch": 0.25739596254541863,
+      "grad_norm": 2.515625,
+      "learning_rate": 4.238980695802203e-05,
+      "loss": 0.4279,
+      "step": 5862
+    },
+    {
+      "epoch": 0.2574837810246221,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.238483594284094e-05,
+      "loss": 0.471,
+      "step": 5864
+    },
+    {
+      "epoch": 0.2575715995038256,
+      "grad_norm": 3.03125,
+      "learning_rate": 4.2379863596324953e-05,
+      "loss": 0.4962,
+      "step": 5866
+    },
+    {
+      "epoch": 0.2576594179830291,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.2374889918854846e-05,
+      "loss": 0.4755,
+      "step": 5868
+    },
+    {
+      "epoch": 0.2577472364622326,
+      "grad_norm": 3.265625,
+      "learning_rate": 4.236991491081151e-05,
+      "loss": 0.4539,
+      "step": 5870
+    },
+    {
+      "epoch": 0.25783505494143605,
+      "grad_norm": 2.546875,
+      "learning_rate": 4.236493857257591e-05,
+      "loss": 0.4678,
+      "step": 5872
+    },
+    {
+      "epoch": 0.2579228734206395,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.235996090452916e-05,
+      "loss": 0.4569,
+      "step": 5874
+    },
+    {
+      "epoch": 0.25801069189984305,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.2354981907052446e-05,
+      "loss": 0.4773,
+      "step": 5876
+    },
+    {
+      "epoch": 0.2580985103790465,
+      "grad_norm": 2.46875,
+      "learning_rate": 4.2350001580527057e-05,
+      "loss": 0.4491,
+      "step": 5878
+    },
+    {
+      "epoch": 0.25818632885825,
+      "grad_norm": 1.9765625,
+      "learning_rate": 4.234501992533438e-05,
+      "loss": 0.4411,
+      "step": 5880
+    },
+    {
+      "epoch": 0.25827414733745346,
+      "grad_norm": 2.0,
+      "learning_rate": 4.2340036941855924e-05,
+      "loss": 0.4516,
+      "step": 5882
+    },
+    {
+      "epoch": 0.258361965816657,
+      "grad_norm": 2.15625,
+      "learning_rate": 4.233505263047328e-05,
+      "loss": 0.4443,
+      "step": 5884
+    },
+    {
+      "epoch": 0.25844978429586046,
+      "grad_norm": 2.15625,
+      "learning_rate": 4.233006699156816e-05,
+      "loss": 0.4445,
+      "step": 5886
+    },
+    {
+      "epoch": 0.25853760277506393,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.232508002552235e-05,
+      "loss": 0.458,
+      "step": 5888
+    },
+    {
+      "epoch": 0.2586254212542674,
+      "grad_norm": 2.234375,
+      "learning_rate": 4.232009173271776e-05,
+      "loss": 0.4336,
+      "step": 5890
+    },
+    {
+      "epoch": 0.25871323973347093,
+      "grad_norm": 2.015625,
+      "learning_rate": 4.231510211353639e-05,
+      "loss": 0.4569,
+      "step": 5892
+    },
+    {
+      "epoch": 0.2588010582126744,
+      "grad_norm": 2.046875,
+      "learning_rate": 4.2310111168360345e-05,
+      "loss": 0.4377,
+      "step": 5894
+    },
+    {
+      "epoch": 0.2588888766918779,
+      "grad_norm": 2.078125,
+      "learning_rate": 4.230511889757184e-05,
+      "loss": 0.4822,
+      "step": 5896
+    },
+    {
+      "epoch": 0.25897669517108135,
+      "grad_norm": 2.0,
+      "learning_rate": 4.230012530155318e-05,
+      "loss": 0.4582,
+      "step": 5898
+    },
+    {
+      "epoch": 0.2590645136502849,
+      "grad_norm": 2.078125,
+      "learning_rate": 4.229513038068678e-05,
+      "loss": 0.4536,
+      "step": 5900
+    },
+    {
+      "epoch": 0.25915233212948835,
+      "grad_norm": 2.203125,
+      "learning_rate": 4.229013413535515e-05,
+      "loss": 0.4522,
+      "step": 5902
+    },
+    {
+      "epoch": 0.2592401506086918,
+      "grad_norm": 2.328125,
+      "learning_rate": 4.228513656594091e-05,
+      "loss": 0.4709,
+      "step": 5904
+    },
+    {
+      "epoch": 0.2593279690878953,
+      "grad_norm": 2.171875,
+      "learning_rate": 4.228013767282676e-05,
+      "loss": 0.4666,
+      "step": 5906
+    },
+    {
+      "epoch": 0.2594157875670988,
+      "grad_norm": 2.484375,
+      "learning_rate": 4.227513745639553e-05,
+      "loss": 0.4448,
+      "step": 5908
+    },
+    {
+      "epoch": 0.2595036060463023,
+      "grad_norm": 2.359375,
+      "learning_rate": 4.227013591703012e-05,
+      "loss": 0.4491,
+      "step": 5910
+    },
+    {
+      "epoch": 0.25959142452550577,
+      "grad_norm": 2.328125,
+      "learning_rate": 4.226513305511357e-05,
+      "loss": 0.4451,
+      "step": 5912
+    },
+    {
+      "epoch": 0.2596792430047093,
+      "grad_norm": 2.09375,
+      "learning_rate": 4.226012887102899e-05,
+      "loss": 0.4603,
+      "step": 5914
+    },
+    {
+      "epoch": 0.25976706148391276,
+      "grad_norm": 2.328125,
+      "learning_rate": 4.225512336515961e-05,
+      "loss": 0.4343,
+      "step": 5916
+    },
+    {
+      "epoch": 0.25985487996311624,
+      "grad_norm": 2.109375,
+      "learning_rate": 4.225011653788874e-05,
+      "loss": 0.465,
+      "step": 5918
+    },
+    {
+      "epoch": 0.2599426984423197,
+      "grad_norm": 2.3125,
+      "learning_rate": 4.224510838959981e-05,
+      "loss": 0.4542,
+      "step": 5920
+    },
+    {
+      "epoch": 0.26003051692152324,
+      "grad_norm": 2.25,
+      "learning_rate": 4.2240098920676353e-05,
+      "loss": 0.4268,
+      "step": 5922
+    },
+    {
+      "epoch": 0.2601183354007267,
+      "grad_norm": 2.125,
+      "learning_rate": 4.223508813150198e-05,
+      "loss": 0.476,
+      "step": 5924
+    },
+    {
+      "epoch": 0.2602061538799302,
+      "grad_norm": 2.265625,
+      "learning_rate": 4.223007602246043e-05,
+      "loss": 0.4375,
+      "step": 5926
+    },
+    {
+      "epoch": 0.26029397235913365,
+      "grad_norm": 2.25,
+      "learning_rate": 4.2225062593935524e-05,
+      "loss": 0.4468,
+      "step": 5928
+    },
+    {
+      "epoch": 0.2603817908383372,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.22200478463112e-05,
+      "loss": 0.4386,
+      "step": 5930
+    },
+    {
+      "epoch": 0.26046960931754065,
+      "grad_norm": 2.359375,
+      "learning_rate": 4.221503177997148e-05,
+      "loss": 0.4495,
+      "step": 5932
+    },
+    {
+      "epoch": 0.2605574277967441,
+      "grad_norm": 2.34375,
+      "learning_rate": 4.221001439530051e-05,
+      "loss": 0.4712,
+      "step": 5934
+    },
+    {
+      "epoch": 0.2606452462759476,
+      "grad_norm": 2.15625,
+      "learning_rate": 4.2204995692682504e-05,
+      "loss": 0.4401,
+      "step": 5936
+    },
+    {
+      "epoch": 0.2607330647551511,
+      "grad_norm": 2.203125,
+      "learning_rate": 4.21999756725018e-05,
+      "loss": 0.4546,
+      "step": 5938
+    },
+    {
+      "epoch": 0.2608208832343546,
+      "grad_norm": 2.359375,
+      "learning_rate": 4.219495433514284e-05,
+      "loss": 0.4559,
+      "step": 5940
+    },
+    {
+      "epoch": 0.26090870171355807,
+      "grad_norm": 3.0,
+      "learning_rate": 4.2189931680990155e-05,
+      "loss": 0.444,
+      "step": 5942
+    },
+    {
+      "epoch": 0.26099652019276154,
+      "grad_norm": 3.0,
+      "learning_rate": 4.2184907710428375e-05,
+      "loss": 0.4885,
+      "step": 5944
+    },
+    {
+      "epoch": 0.26108433867196507,
+      "grad_norm": 2.421875,
+      "learning_rate": 4.217988242384225e-05,
+      "loss": 0.468,
+      "step": 5946
+    },
+    {
+      "epoch": 0.26117215715116854,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.217485582161661e-05,
+      "loss": 0.4635,
+      "step": 5948
+    },
+    {
+      "epoch": 0.261259975630372,
+      "grad_norm": 2.625,
+      "learning_rate": 4.2169827904136396e-05,
+      "loss": 0.4728,
+      "step": 5950
+    },
+    {
+      "epoch": 0.2613477941095755,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.216479867178664e-05,
+      "loss": 0.4572,
+      "step": 5952
+    },
+    {
+      "epoch": 0.261435612588779,
+      "grad_norm": 2.5,
+      "learning_rate": 4.215976812495249e-05,
+      "loss": 0.4628,
+      "step": 5954
+    },
+    {
+      "epoch": 0.2615234310679825,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.2154736264019184e-05,
+      "loss": 0.444,
+      "step": 5956
+    },
+    {
+      "epoch": 0.26161124954718595,
+      "grad_norm": 2.53125,
+      "learning_rate": 4.214970308937206e-05,
+      "loss": 0.4545,
+      "step": 5958
+    },
+    {
+      "epoch": 0.2616990680263894,
+      "grad_norm": 2.25,
+      "learning_rate": 4.2144668601396566e-05,
+      "loss": 0.4274,
+      "step": 5960
+    },
+    {
+      "epoch": 0.26178688650559295,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.2139632800478234e-05,
+      "loss": 0.4839,
+      "step": 5962
+    },
+    {
+      "epoch": 0.2618747049847964,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.213459568700273e-05,
+      "loss": 0.4527,
+      "step": 5964
+    },
+    {
+      "epoch": 0.2619625234639999,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.212955726135577e-05,
+      "loss": 0.453,
+      "step": 5966
+    },
+    {
+      "epoch": 0.2620503419432034,
+      "grad_norm": 2.09375,
+      "learning_rate": 4.21245175239232e-05,
+      "loss": 0.4747,
+      "step": 5968
+    },
+    {
+      "epoch": 0.2621381604224069,
+      "grad_norm": 2.25,
+      "learning_rate": 4.211947647509098e-05,
+      "loss": 0.4459,
+      "step": 5970
+    },
+    {
+      "epoch": 0.26222597890161037,
+      "grad_norm": 2.171875,
+      "learning_rate": 4.211443411524515e-05,
+      "loss": 0.4452,
+      "step": 5972
+    },
+    {
+      "epoch": 0.26231379738081384,
+      "grad_norm": 2.296875,
+      "learning_rate": 4.210939044477185e-05,
+      "loss": 0.4412,
+      "step": 5974
+    },
+    {
+      "epoch": 0.26240161586001737,
+      "grad_norm": 2.421875,
+      "learning_rate": 4.210434546405733e-05,
+      "loss": 0.4675,
+      "step": 5976
+    },
+    {
+      "epoch": 0.26248943433922084,
+      "grad_norm": 2.171875,
+      "learning_rate": 4.2099299173487936e-05,
+      "loss": 0.4373,
+      "step": 5978
+    },
+    {
+      "epoch": 0.2625772528184243,
+      "grad_norm": 2.1875,
+      "learning_rate": 4.209425157345011e-05,
+      "loss": 0.4514,
+      "step": 5980
+    },
+    {
+      "epoch": 0.2626650712976278,
+      "grad_norm": 2.421875,
+      "learning_rate": 4.20892026643304e-05,
+      "loss": 0.4374,
+      "step": 5982
+    },
+    {
+      "epoch": 0.2627528897768313,
+      "grad_norm": 2.5,
+      "learning_rate": 4.208415244651546e-05,
+      "loss": 0.4501,
+      "step": 5984
+    },
+    {
+      "epoch": 0.2628407082560348,
+      "grad_norm": 2.28125,
+      "learning_rate": 4.207910092039202e-05,
+      "loss": 0.4409,
+      "step": 5986
+    },
+    {
+      "epoch": 0.26292852673523825,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.207404808634694e-05,
+      "loss": 0.4469,
+      "step": 5988
+    },
+    {
+      "epoch": 0.2630163452144417,
+      "grad_norm": 2.109375,
+      "learning_rate": 4.206899394476717e-05,
+      "loss": 0.4775,
+      "step": 5990
+    },
+    {
+      "epoch": 0.26310416369364525,
+      "grad_norm": 2.234375,
+      "learning_rate": 4.2063938496039746e-05,
+      "loss": 0.4512,
+      "step": 5992
+    },
+    {
+      "epoch": 0.2631919821728487,
+      "grad_norm": 2.40625,
+      "learning_rate": 4.2058881740551825e-05,
+      "loss": 0.4476,
+      "step": 5994
+    },
+    {
+      "epoch": 0.2632798006520522,
+      "grad_norm": 2.421875,
+      "learning_rate": 4.2053823678690655e-05,
+      "loss": 0.4599,
+      "step": 5996
+    },
+    {
+      "epoch": 0.26336761913125567,
+      "grad_norm": 2.09375,
+      "learning_rate": 4.2048764310843566e-05,
+      "loss": 0.4945,
+      "step": 5998
+    },
+    {
+      "epoch": 0.2634554376104592,
+      "grad_norm": 2.15625,
+      "learning_rate": 4.204370363739803e-05,
+      "loss": 0.4648,
+      "step": 6000
+    },
+    {
+      "epoch": 0.26354325608966267,
+      "grad_norm": 2.46875,
+      "learning_rate": 4.203864165874158e-05,
+      "loss": 0.4385,
+      "step": 6002
+    },
+    {
+      "epoch": 0.26363107456886614,
+      "grad_norm": 2.125,
+      "learning_rate": 4.203357837526187e-05,
+      "loss": 0.4218,
+      "step": 6004
+    },
+    {
+      "epoch": 0.2637188930480696,
+      "grad_norm": 2.09375,
+      "learning_rate": 4.202851378734664e-05,
+      "loss": 0.4294,
+      "step": 6006
+    },
+    {
+      "epoch": 0.26380671152727314,
+      "grad_norm": 2.078125,
+      "learning_rate": 4.2023447895383746e-05,
+      "loss": 0.4554,
+      "step": 6008
+    },
+    {
+      "epoch": 0.2638945300064766,
+      "grad_norm": 2.375,
+      "learning_rate": 4.201838069976114e-05,
+      "loss": 0.4748,
+      "step": 6010
+    },
+    {
+      "epoch": 0.2639823484856801,
+      "grad_norm": 2.046875,
+      "learning_rate": 4.201331220086685e-05,
+      "loss": 0.4362,
+      "step": 6012
+    },
+    {
+      "epoch": 0.2640701669648836,
+      "grad_norm": 2.265625,
+      "learning_rate": 4.2008242399089036e-05,
+      "loss": 0.4517,
+      "step": 6014
+    },
+    {
+      "epoch": 0.2641579854440871,
+      "grad_norm": 2.375,
+      "learning_rate": 4.200317129481594e-05,
+      "loss": 0.4676,
+      "step": 6016
+    },
+    {
+      "epoch": 0.26424580392329056,
+      "grad_norm": 2.125,
+      "learning_rate": 4.199809888843591e-05,
+      "loss": 0.4595,
+      "step": 6018
+    },
+    {
+      "epoch": 0.26433362240249403,
+      "grad_norm": 2.421875,
+      "learning_rate": 4.19930251803374e-05,
+      "loss": 0.434,
+      "step": 6020
+    },
+    {
+      "epoch": 0.26442144088169756,
+      "grad_norm": 2.28125,
+      "learning_rate": 4.198795017090894e-05,
+      "loss": 0.4706,
+      "step": 6022
+    },
+    {
+      "epoch": 0.26450925936090103,
+      "grad_norm": 2.21875,
+      "learning_rate": 4.1982873860539186e-05,
+      "loss": 0.4541,
+      "step": 6024
+    },
+    {
+      "epoch": 0.2645970778401045,
+      "grad_norm": 2.234375,
+      "learning_rate": 4.197779624961688e-05,
+      "loss": 0.4787,
+      "step": 6026
+    },
+    {
+      "epoch": 0.26468489631930797,
+      "grad_norm": 2.203125,
+      "learning_rate": 4.1972717338530865e-05,
+      "loss": 0.4714,
+      "step": 6028
+    },
+    {
+      "epoch": 0.2647727147985115,
+      "grad_norm": 2.03125,
+      "learning_rate": 4.196763712767009e-05,
+      "loss": 0.4667,
+      "step": 6030
+    },
+    {
+      "epoch": 0.26486053327771497,
+      "grad_norm": 2.40625,
+      "learning_rate": 4.19625556174236e-05,
+      "loss": 0.4655,
+      "step": 6032
+    },
+    {
+      "epoch": 0.26494835175691844,
+      "grad_norm": 2.234375,
+      "learning_rate": 4.195747280818053e-05,
+      "loss": 0.4729,
+      "step": 6034
+    },
+    {
+      "epoch": 0.2650361702361219,
+      "grad_norm": 2.078125,
+      "learning_rate": 4.195238870033012e-05,
+      "loss": 0.4532,
+      "step": 6036
+    },
+    {
+      "epoch": 0.26512398871532544,
+      "grad_norm": 2.09375,
+      "learning_rate": 4.194730329426173e-05,
+      "loss": 0.4614,
+      "step": 6038
+    },
+    {
+      "epoch": 0.2652118071945289,
+      "grad_norm": 2.046875,
+      "learning_rate": 4.194221659036479e-05,
+      "loss": 0.4355,
+      "step": 6040
+    },
+    {
+      "epoch": 0.2652996256737324,
+      "grad_norm": 2.109375,
+      "learning_rate": 4.1937128589028845e-05,
+      "loss": 0.4523,
+      "step": 6042
+    },
+    {
+      "epoch": 0.26538744415293586,
+      "grad_norm": 2.265625,
+      "learning_rate": 4.193203929064353e-05,
+      "loss": 0.4551,
+      "step": 6044
+    },
+    {
+      "epoch": 0.2654752626321394,
+      "grad_norm": 2.140625,
+      "learning_rate": 4.192694869559859e-05,
+      "loss": 0.4766,
+      "step": 6046
+    },
+    {
+      "epoch": 0.26556308111134286,
+      "grad_norm": 2.328125,
+      "learning_rate": 4.1921856804283854e-05,
+      "loss": 0.4742,
+      "step": 6048
+    },
+    {
+      "epoch": 0.26565089959054633,
+      "grad_norm": 2.5,
+      "learning_rate": 4.191676361708927e-05,
+      "loss": 0.4476,
+      "step": 6050
+    },
+    {
+      "epoch": 0.2657387180697498,
+      "grad_norm": 2.5,
+      "learning_rate": 4.191166913440487e-05,
+      "loss": 0.4644,
+      "step": 6052
+    },
+    {
+      "epoch": 0.26582653654895333,
+      "grad_norm": 2.3125,
+      "learning_rate": 4.1906573356620795e-05,
+      "loss": 0.4646,
+      "step": 6054
+    },
+    {
+      "epoch": 0.2659143550281568,
+      "grad_norm": 2.46875,
+      "learning_rate": 4.190147628412729e-05,
+      "loss": 0.4577,
+      "step": 6056
+    },
+    {
+      "epoch": 0.2660021735073603,
+      "grad_norm": 2.28125,
+      "learning_rate": 4.189637791731467e-05,
+      "loss": 0.4707,
+      "step": 6058
+    },
+    {
+      "epoch": 0.26608999198656375,
+      "grad_norm": 2.109375,
+      "learning_rate": 4.1891278256573384e-05,
+      "loss": 0.4336,
+      "step": 6060
+    },
+    {
+      "epoch": 0.2661778104657673,
+      "grad_norm": 2.75,
+      "learning_rate": 4.188617730229395e-05,
+      "loss": 0.4459,
+      "step": 6062
+    },
+    {
+      "epoch": 0.26626562894497074,
+      "grad_norm": 2.171875,
+      "learning_rate": 4.188107505486702e-05,
+      "loss": 0.4513,
+      "step": 6064
+    },
+    {
+      "epoch": 0.2663534474241742,
+      "grad_norm": 2.15625,
+      "learning_rate": 4.187597151468331e-05,
+      "loss": 0.4408,
+      "step": 6066
+    },
+    {
+      "epoch": 0.26644126590337774,
+      "grad_norm": 2.421875,
+      "learning_rate": 4.187086668213366e-05,
+      "loss": 0.454,
+      "step": 6068
+    },
+    {
+      "epoch": 0.2665290843825812,
+      "grad_norm": 2.328125,
+      "learning_rate": 4.186576055760899e-05,
+      "loss": 0.4469,
+      "step": 6070
+    },
+    {
+      "epoch": 0.2666169028617847,
+      "grad_norm": 2.0625,
+      "learning_rate": 4.186065314150034e-05,
+      "loss": 0.4321,
+      "step": 6072
+    },
+    {
+      "epoch": 0.26670472134098816,
+      "grad_norm": 2.0625,
+      "learning_rate": 4.1855544434198826e-05,
+      "loss": 0.4323,
+      "step": 6074
+    },
+    {
+      "epoch": 0.2667925398201917,
+      "grad_norm": 2.3125,
+      "learning_rate": 4.185043443609569e-05,
+      "loss": 0.4487,
+      "step": 6076
+    },
+    {
+      "epoch": 0.26688035829939516,
+      "grad_norm": 2.171875,
+      "learning_rate": 4.184532314758223e-05,
+      "loss": 0.4653,
+      "step": 6078
+    },
+    {
+      "epoch": 0.26696817677859863,
+      "grad_norm": 2.1875,
+      "learning_rate": 4.184021056904989e-05,
+      "loss": 0.489,
+      "step": 6080
+    },
+    {
+      "epoch": 0.2670559952578021,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.183509670089018e-05,
+      "loss": 0.4378,
+      "step": 6082
+    },
+    {
+      "epoch": 0.26714381373700563,
+      "grad_norm": 3.28125,
+      "learning_rate": 4.1829981543494746e-05,
+      "loss": 0.4667,
+      "step": 6084
+    },
+    {
+      "epoch": 0.2672316322162091,
+      "grad_norm": 3.875,
+      "learning_rate": 4.1824865097255284e-05,
+      "loss": 0.4541,
+      "step": 6086
+    },
+    {
+      "epoch": 0.2673194506954126,
+      "grad_norm": 2.15625,
+      "learning_rate": 4.181974736256362e-05,
+      "loss": 0.4202,
+      "step": 6088
+    },
+    {
+      "epoch": 0.26740726917461605,
+      "grad_norm": 2.515625,
+      "learning_rate": 4.181462833981167e-05,
+      "loss": 0.474,
+      "step": 6090
+    },
+    {
+      "epoch": 0.2674950876538196,
+      "grad_norm": 2.359375,
+      "learning_rate": 4.180950802939145e-05,
+      "loss": 0.4343,
+      "step": 6092
+    },
+    {
+      "epoch": 0.26758290613302305,
+      "grad_norm": 1.9921875,
+      "learning_rate": 4.1804386431695076e-05,
+      "loss": 0.4172,
+      "step": 6094
+    },
+    {
+      "epoch": 0.2676707246122265,
+      "grad_norm": 2.296875,
+      "learning_rate": 4.179926354711476e-05,
+      "loss": 0.442,
+      "step": 6096
+    },
+    {
+      "epoch": 0.26775854309143,
+      "grad_norm": 2.015625,
+      "learning_rate": 4.179413937604282e-05,
+      "loss": 0.4464,
+      "step": 6098
+    },
+    {
+      "epoch": 0.2678463615706335,
+      "grad_norm": 2.21875,
+      "learning_rate": 4.178901391887165e-05,
+      "loss": 0.4462,
+      "step": 6100
+    },
+    {
+      "epoch": 0.267934180049837,
+      "grad_norm": 2.296875,
+      "learning_rate": 4.178388717599378e-05,
+      "loss": 0.4389,
+      "step": 6102
+    },
+    {
+      "epoch": 0.26802199852904046,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.17787591478018e-05,
+      "loss": 0.4595,
+      "step": 6104
+    },
+    {
+      "epoch": 0.26810981700824393,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.177362983468843e-05,
+      "loss": 0.4419,
+      "step": 6106
+    },
+    {
+      "epoch": 0.26819763548744746,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.1768499237046455e-05,
+      "loss": 0.4286,
+      "step": 6108
+    },
+    {
+      "epoch": 0.26828545396665093,
+      "grad_norm": 2.203125,
+      "learning_rate": 4.17633673552688e-05,
+      "loss": 0.4302,
+      "step": 6110
+    },
+    {
+      "epoch": 0.2683732724458544,
+      "grad_norm": 2.1875,
+      "learning_rate": 4.175823418974845e-05,
+      "loss": 0.4421,
+      "step": 6112
+    },
+    {
+      "epoch": 0.26846109092505793,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.17530997408785e-05,
+      "loss": 0.471,
+      "step": 6114
+    },
+    {
+      "epoch": 0.2685489094042614,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.174796400905216e-05,
+      "loss": 0.4329,
+      "step": 6116
+    },
+    {
+      "epoch": 0.2686367278834649,
+      "grad_norm": 2.421875,
+      "learning_rate": 4.1742826994662734e-05,
+      "loss": 0.4473,
+      "step": 6118
+    },
+    {
+      "epoch": 0.26872454636266835,
+      "grad_norm": 2.296875,
+      "learning_rate": 4.1737688698103595e-05,
+      "loss": 0.5077,
+      "step": 6120
+    },
+    {
+      "epoch": 0.2688123648418719,
+      "grad_norm": 2.296875,
+      "learning_rate": 4.173254911976824e-05,
+      "loss": 0.4531,
+      "step": 6122
+    },
+    {
+      "epoch": 0.26890018332107535,
+      "grad_norm": 2.453125,
+      "learning_rate": 4.172740826005027e-05,
+      "loss": 0.4318,
+      "step": 6124
+    },
+    {
+      "epoch": 0.2689880018002788,
+      "grad_norm": 2.40625,
+      "learning_rate": 4.1722266119343357e-05,
+      "loss": 0.4643,
+      "step": 6126
+    },
+    {
+      "epoch": 0.2690758202794823,
+      "grad_norm": 2.34375,
+      "learning_rate": 4.1717122698041296e-05,
+      "loss": 0.4479,
+      "step": 6128
+    },
+    {
+      "epoch": 0.2691636387586858,
+      "grad_norm": 2.015625,
+      "learning_rate": 4.1711977996537976e-05,
+      "loss": 0.4423,
+      "step": 6130
+    },
+    {
+      "epoch": 0.2692514572378893,
+      "grad_norm": 2.0,
+      "learning_rate": 4.170683201522737e-05,
+      "loss": 0.4564,
+      "step": 6132
+    },
+    {
+      "epoch": 0.26933927571709276,
+      "grad_norm": 2.0625,
+      "learning_rate": 4.170168475450357e-05,
+      "loss": 0.4654,
+      "step": 6134
+    },
+    {
+      "epoch": 0.26942709419629624,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.1696536214760746e-05,
+      "loss": 0.4642,
+      "step": 6136
+    },
+    {
+      "epoch": 0.26951491267549976,
+      "grad_norm": 2.390625,
+      "learning_rate": 4.169138639639317e-05,
+      "loss": 0.452,
+      "step": 6138
+    },
+    {
+      "epoch": 0.26960273115470323,
+      "grad_norm": 2.28125,
+      "learning_rate": 4.1686235299795226e-05,
+      "loss": 0.4645,
+      "step": 6140
+    },
+    {
+      "epoch": 0.2696905496339067,
+      "grad_norm": 2.15625,
+      "learning_rate": 4.168108292536139e-05,
+      "loss": 0.4592,
+      "step": 6142
+    },
+    {
+      "epoch": 0.2697783681131102,
+      "grad_norm": 2.34375,
+      "learning_rate": 4.167592927348622e-05,
+      "loss": 0.4458,
+      "step": 6144
+    },
+    {
+      "epoch": 0.2698661865923137,
+      "grad_norm": 2.203125,
+      "learning_rate": 4.167077434456439e-05,
+      "loss": 0.4366,
+      "step": 6146
+    },
+    {
+      "epoch": 0.2699540050715172,
+      "grad_norm": 2.1875,
+      "learning_rate": 4.166561813899066e-05,
+      "loss": 0.4384,
+      "step": 6148
+    },
+    {
+      "epoch": 0.27004182355072065,
+      "grad_norm": 2.265625,
+      "learning_rate": 4.16604606571599e-05,
+      "loss": 0.4309,
+      "step": 6150
+    },
+    {
+      "epoch": 0.2701296420299241,
+      "grad_norm": 2.234375,
+      "learning_rate": 4.165530189946707e-05,
+      "loss": 0.4639,
+      "step": 6152
+    },
+    {
+      "epoch": 0.27021746050912765,
+      "grad_norm": 2.203125,
+      "learning_rate": 4.1650141866307224e-05,
+      "loss": 0.4496,
+      "step": 6154
+    },
+    {
+      "epoch": 0.2703052789883311,
+      "grad_norm": 2.46875,
+      "learning_rate": 4.164498055807553e-05,
+      "loss": 0.4343,
+      "step": 6156
+    },
+    {
+      "epoch": 0.2703930974675346,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.163981797516723e-05,
+      "loss": 0.4438,
+      "step": 6158
+    },
+    {
+      "epoch": 0.27048091594673807,
+      "grad_norm": 3.28125,
+      "learning_rate": 4.163465411797768e-05,
+      "loss": 0.4517,
+      "step": 6160
+    },
+    {
+      "epoch": 0.2705687344259416,
+      "grad_norm": 3.484375,
+      "learning_rate": 4.162948898690233e-05,
+      "loss": 0.4539,
+      "step": 6162
+    },
+    {
+      "epoch": 0.27065655290514506,
+      "grad_norm": 2.5,
+      "learning_rate": 4.162432258233673e-05,
+      "loss": 0.4514,
+      "step": 6164
+    },
+    {
+      "epoch": 0.27074437138434854,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.1619154904676525e-05,
+      "loss": 0.4465,
+      "step": 6166
+    },
+    {
+      "epoch": 0.27083218986355206,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.1613985954317446e-05,
+      "loss": 0.44,
+      "step": 6168
+    },
+    {
+      "epoch": 0.27092000834275554,
+      "grad_norm": 2.0625,
+      "learning_rate": 4.1608815731655345e-05,
+      "loss": 0.4482,
+      "step": 6170
+    },
+    {
+      "epoch": 0.271007826821959,
+      "grad_norm": 2.078125,
+      "learning_rate": 4.160364423708615e-05,
+      "loss": 0.4504,
+      "step": 6172
+    },
+    {
+      "epoch": 0.2710956453011625,
+      "grad_norm": 2.421875,
+      "learning_rate": 4.15984714710059e-05,
+      "loss": 0.4325,
+      "step": 6174
+    },
+    {
+      "epoch": 0.271183463780366,
+      "grad_norm": 2.34375,
+      "learning_rate": 4.159329743381072e-05,
+      "loss": 0.4344,
+      "step": 6176
+    },
+    {
+      "epoch": 0.2712712822595695,
+      "grad_norm": 2.1875,
+      "learning_rate": 4.1588122125896854e-05,
+      "loss": 0.4401,
+      "step": 6178
+    },
+    {
+      "epoch": 0.27135910073877295,
+      "grad_norm": 2.171875,
+      "learning_rate": 4.1582945547660625e-05,
+      "loss": 0.4931,
+      "step": 6180
+    },
+    {
+      "epoch": 0.2714469192179764,
+      "grad_norm": 2.078125,
+      "learning_rate": 4.157776769949844e-05,
+      "loss": 0.4262,
+      "step": 6182
+    },
+    {
+      "epoch": 0.27153473769717995,
+      "grad_norm": 2.140625,
+      "learning_rate": 4.157258858180683e-05,
+      "loss": 0.4675,
+      "step": 6184
+    },
+    {
+      "epoch": 0.2716225561763834,
+      "grad_norm": 2.3125,
+      "learning_rate": 4.156740819498242e-05,
+      "loss": 0.4665,
+      "step": 6186
+    },
+    {
+      "epoch": 0.2717103746555869,
+      "grad_norm": 2.0625,
+      "learning_rate": 4.156222653942191e-05,
+      "loss": 0.4537,
+      "step": 6188
+    },
+    {
+      "epoch": 0.27179819313479037,
+      "grad_norm": 2.234375,
+      "learning_rate": 4.1557043615522125e-05,
+      "loss": 0.4607,
+      "step": 6190
+    },
+    {
+      "epoch": 0.2718860116139939,
+      "grad_norm": 2.21875,
+      "learning_rate": 4.155185942367997e-05,
+      "loss": 0.4709,
+      "step": 6192
+    },
+    {
+      "epoch": 0.27197383009319737,
+      "grad_norm": 2.25,
+      "learning_rate": 4.154667396429246e-05,
+      "loss": 0.4343,
+      "step": 6194
+    },
+    {
+      "epoch": 0.27206164857240084,
+      "grad_norm": 2.296875,
+      "learning_rate": 4.1541487237756686e-05,
+      "loss": 0.4345,
+      "step": 6196
+    },
+    {
+      "epoch": 0.2721494670516043,
+      "grad_norm": 2.171875,
+      "learning_rate": 4.153629924446986e-05,
+      "loss": 0.457,
+      "step": 6198
+    },
+    {
+      "epoch": 0.27223728553080784,
+      "grad_norm": 2.109375,
+      "learning_rate": 4.153110998482926e-05,
+      "loss": 0.4675,
+      "step": 6200
+    },
+    {
+      "epoch": 0.2723251040100113,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.152591945923231e-05,
+      "loss": 0.4952,
+      "step": 6202
+    },
+    {
+      "epoch": 0.2724129224892148,
+      "grad_norm": 2.125,
+      "learning_rate": 4.152072766807648e-05,
+      "loss": 0.46,
+      "step": 6204
+    },
+    {
+      "epoch": 0.27250074096841825,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.151553461175936e-05,
+      "loss": 0.4656,
+      "step": 6206
+    },
+    {
+      "epoch": 0.2725885594476218,
+      "grad_norm": 2.421875,
+      "learning_rate": 4.151034029067864e-05,
+      "loss": 0.4346,
+      "step": 6208
+    },
+    {
+      "epoch": 0.27267637792682525,
+      "grad_norm": 2.359375,
+      "learning_rate": 4.1505144705232114e-05,
+      "loss": 0.4237,
+      "step": 6210
+    },
+    {
+      "epoch": 0.2727641964060287,
+      "grad_norm": 1.9296875,
+      "learning_rate": 4.149994785581764e-05,
+      "loss": 0.4196,
+      "step": 6212
+    },
+    {
+      "epoch": 0.27285201488523225,
+      "grad_norm": 2.03125,
+      "learning_rate": 4.149474974283321e-05,
+      "loss": 0.4359,
+      "step": 6214
+    },
+    {
+      "epoch": 0.2729398333644357,
+      "grad_norm": 2.265625,
+      "learning_rate": 4.148955036667689e-05,
+      "loss": 0.4525,
+      "step": 6216
+    },
+    {
+      "epoch": 0.2730276518436392,
+      "grad_norm": 2.15625,
+      "learning_rate": 4.148434972774685e-05,
+      "loss": 0.4554,
+      "step": 6218
+    },
+    {
+      "epoch": 0.27311547032284267,
+      "grad_norm": 2.09375,
+      "learning_rate": 4.147914782644134e-05,
+      "loss": 0.4339,
+      "step": 6220
+    },
+    {
+      "epoch": 0.2732032888020462,
+      "grad_norm": 2.265625,
+      "learning_rate": 4.147394466315876e-05,
+      "loss": 0.4376,
+      "step": 6222
+    },
+    {
+      "epoch": 0.27329110728124967,
+      "grad_norm": 2.484375,
+      "learning_rate": 4.146874023829754e-05,
+      "loss": 0.4694,
+      "step": 6224
+    },
+    {
+      "epoch": 0.27337892576045314,
+      "grad_norm": 2.125,
+      "learning_rate": 4.146353455225625e-05,
+      "loss": 0.4433,
+      "step": 6226
+    },
+    {
+      "epoch": 0.2734667442396566,
+      "grad_norm": 2.125,
+      "learning_rate": 4.145832760543353e-05,
+      "loss": 0.4423,
+      "step": 6228
+    },
+    {
+      "epoch": 0.27355456271886014,
+      "grad_norm": 2.15625,
+      "learning_rate": 4.1453119398228146e-05,
+      "loss": 0.4558,
+      "step": 6230
+    },
+    {
+      "epoch": 0.2736423811980636,
+      "grad_norm": 2.140625,
+      "learning_rate": 4.144790993103893e-05,
+      "loss": 0.4371,
+      "step": 6232
+    },
+    {
+      "epoch": 0.2737301996772671,
+      "grad_norm": 2.359375,
+      "learning_rate": 4.144269920426482e-05,
+      "loss": 0.4518,
+      "step": 6234
+    },
+    {
+      "epoch": 0.27381801815647056,
+      "grad_norm": 2.40625,
+      "learning_rate": 4.1437487218304875e-05,
+      "loss": 0.4369,
+      "step": 6236
+    },
+    {
+      "epoch": 0.2739058366356741,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.1432273973558215e-05,
+      "loss": 0.4403,
+      "step": 6238
+    },
+    {
+      "epoch": 0.27399365511487755,
+      "grad_norm": 2.28125,
+      "learning_rate": 4.142705947042408e-05,
+      "loss": 0.4483,
+      "step": 6240
+    },
+    {
+      "epoch": 0.274081473594081,
+      "grad_norm": 2.1875,
+      "learning_rate": 4.142184370930178e-05,
+      "loss": 0.4426,
+      "step": 6242
+    },
+    {
+      "epoch": 0.2741692920732845,
+      "grad_norm": 1.890625,
+      "learning_rate": 4.141662669059076e-05,
+      "loss": 0.4282,
+      "step": 6244
+    },
+    {
+      "epoch": 0.274257110552488,
+      "grad_norm": 2.09375,
+      "learning_rate": 4.1411408414690536e-05,
+      "loss": 0.4406,
+      "step": 6246
+    },
+    {
+      "epoch": 0.2743449290316915,
+      "grad_norm": 2.328125,
+      "learning_rate": 4.140618888200072e-05,
+      "loss": 0.4505,
+      "step": 6248
+    },
+    {
+      "epoch": 0.27443274751089497,
+      "grad_norm": 2.25,
+      "learning_rate": 4.140096809292102e-05,
+      "loss": 0.4422,
+      "step": 6250
+    },
+    {
+      "epoch": 0.27452056599009844,
+      "grad_norm": 2.15625,
+      "learning_rate": 4.1395746047851256e-05,
+      "loss": 0.4614,
+      "step": 6252
+    },
+    {
+      "epoch": 0.27460838446930197,
+      "grad_norm": 2.3125,
+      "learning_rate": 4.139052274719133e-05,
+      "loss": 0.4416,
+      "step": 6254
+    },
+    {
+      "epoch": 0.27469620294850544,
+      "grad_norm": 2.21875,
+      "learning_rate": 4.1385298191341246e-05,
+      "loss": 0.468,
+      "step": 6256
+    },
+    {
+      "epoch": 0.2747840214277089,
+      "grad_norm": 3.890625,
+      "learning_rate": 4.1380072380701097e-05,
+      "loss": 0.453,
+      "step": 6258
+    },
+    {
+      "epoch": 0.2748718399069124,
+      "grad_norm": 2.03125,
+      "learning_rate": 4.137484531567107e-05,
+      "loss": 0.4224,
+      "step": 6260
+    },
+    {
+      "epoch": 0.2749596583861159,
+      "grad_norm": 2.234375,
+      "learning_rate": 4.136961699665147e-05,
+      "loss": 0.4824,
+      "step": 6262
+    },
+    {
+      "epoch": 0.2750474768653194,
+      "grad_norm": 2.203125,
+      "learning_rate": 4.136438742404268e-05,
+      "loss": 0.4251,
+      "step": 6264
+    },
+    {
+      "epoch": 0.27513529534452286,
+      "grad_norm": 2.265625,
+      "learning_rate": 4.1359156598245176e-05,
+      "loss": 0.4623,
+      "step": 6266
+    },
+    {
+      "epoch": 0.2752231138237264,
+      "grad_norm": 2.3125,
+      "learning_rate": 4.1353924519659534e-05,
+      "loss": 0.4438,
+      "step": 6268
+    },
+    {
+      "epoch": 0.27531093230292986,
+      "grad_norm": 2.359375,
+      "learning_rate": 4.1348691188686436e-05,
+      "loss": 0.466,
+      "step": 6270
+    },
+    {
+      "epoch": 0.27539875078213333,
+      "grad_norm": 2.1875,
+      "learning_rate": 4.134345660572665e-05,
+      "loss": 0.4888,
+      "step": 6272
+    },
+    {
+      "epoch": 0.2754865692613368,
+      "grad_norm": 2.28125,
+      "learning_rate": 4.1338220771181036e-05,
+      "loss": 0.4234,
+      "step": 6274
+    },
+    {
+      "epoch": 0.2755743877405403,
+      "grad_norm": 2.5,
+      "learning_rate": 4.1332983685450556e-05,
+      "loss": 0.4546,
+      "step": 6276
+    },
+    {
+      "epoch": 0.2756622062197438,
+      "grad_norm": 2.171875,
+      "learning_rate": 4.132774534893628e-05,
+      "loss": 0.4614,
+      "step": 6278
+    },
+    {
+      "epoch": 0.27575002469894727,
+      "grad_norm": 2.15625,
+      "learning_rate": 4.132250576203934e-05,
+      "loss": 0.4509,
+      "step": 6280
+    },
+    {
+      "epoch": 0.27583784317815074,
+      "grad_norm": 2.21875,
+      "learning_rate": 4.131726492516099e-05,
+      "loss": 0.4766,
+      "step": 6282
+    },
+    {
+      "epoch": 0.27592566165735427,
+      "grad_norm": 2.203125,
+      "learning_rate": 4.1312022838702595e-05,
+      "loss": 0.4546,
+      "step": 6284
+    },
+    {
+      "epoch": 0.27601348013655774,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.1306779503065585e-05,
+      "loss": 0.4589,
+      "step": 6286
+    },
+    {
+      "epoch": 0.2761012986157612,
+      "grad_norm": 2.296875,
+      "learning_rate": 4.130153491865148e-05,
+      "loss": 0.4231,
+      "step": 6288
+    },
+    {
+      "epoch": 0.2761891170949647,
+      "grad_norm": 2.0625,
+      "learning_rate": 4.1296289085861924e-05,
+      "loss": 0.4614,
+      "step": 6290
+    },
+    {
+      "epoch": 0.2762769355741682,
+      "grad_norm": 2.34375,
+      "learning_rate": 4.129104200509865e-05,
+      "loss": 0.4485,
+      "step": 6292
+    },
+    {
+      "epoch": 0.2763647540533717,
+      "grad_norm": 2.390625,
+      "learning_rate": 4.128579367676346e-05,
+      "loss": 0.4608,
+      "step": 6294
+    },
+    {
+      "epoch": 0.27645257253257516,
+      "grad_norm": 2.625,
+      "learning_rate": 4.12805441012583e-05,
+      "loss": 0.4491,
+      "step": 6296
+    },
+    {
+      "epoch": 0.27654039101177863,
+      "grad_norm": 2.203125,
+      "learning_rate": 4.1275293278985163e-05,
+      "loss": 0.4294,
+      "step": 6298
+    },
+    {
+      "epoch": 0.27662820949098216,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.127004121034617e-05,
+      "loss": 0.4214,
+      "step": 6300
+    },
+    {
+      "epoch": 0.27671602797018563,
+      "grad_norm": 2.453125,
+      "learning_rate": 4.126478789574352e-05,
+      "loss": 0.4354,
+      "step": 6302
+    },
+    {
+      "epoch": 0.2768038464493891,
+      "grad_norm": 2.25,
+      "learning_rate": 4.1259533335579516e-05,
+      "loss": 0.4204,
+      "step": 6304
+    },
+    {
+      "epoch": 0.2768916649285926,
+      "grad_norm": 2.140625,
+      "learning_rate": 4.125427753025655e-05,
+      "loss": 0.439,
+      "step": 6306
+    },
+    {
+      "epoch": 0.2769794834077961,
+      "grad_norm": 2.421875,
+      "learning_rate": 4.12490204801771e-05,
+      "loss": 0.4297,
+      "step": 6308
+    },
+    {
+      "epoch": 0.2770673018869996,
+      "grad_norm": 2.234375,
+      "learning_rate": 4.1243762185743784e-05,
+      "loss": 0.4689,
+      "step": 6310
+    },
+    {
+      "epoch": 0.27715512036620304,
+      "grad_norm": 2.21875,
+      "learning_rate": 4.123850264735926e-05,
+      "loss": 0.4831,
+      "step": 6312
+    },
+    {
+      "epoch": 0.2772429388454066,
+      "grad_norm": 2.234375,
+      "learning_rate": 4.123324186542631e-05,
+      "loss": 0.4377,
+      "step": 6314
+    },
+    {
+      "epoch": 0.27733075732461004,
+      "grad_norm": 2.4375,
+      "learning_rate": 4.1227979840347806e-05,
+      "loss": 0.4299,
+      "step": 6316
+    },
+    {
+      "epoch": 0.2774185758038135,
+      "grad_norm": 2.203125,
+      "learning_rate": 4.1222716572526725e-05,
+      "loss": 0.4539,
+      "step": 6318
+    },
+    {
+      "epoch": 0.277506394283017,
+      "grad_norm": 2.296875,
+      "learning_rate": 4.121745206236611e-05,
+      "loss": 0.4649,
+      "step": 6320
+    },
+    {
+      "epoch": 0.2775942127622205,
+      "grad_norm": 2.0,
+      "learning_rate": 4.121218631026913e-05,
+      "loss": 0.4221,
+      "step": 6322
+    },
+    {
+      "epoch": 0.277682031241424,
+      "grad_norm": 2.390625,
+      "learning_rate": 4.120691931663904e-05,
+      "loss": 0.4379,
+      "step": 6324
+    },
+    {
+      "epoch": 0.27776984972062746,
+      "grad_norm": 2.21875,
+      "learning_rate": 4.120165108187918e-05,
+      "loss": 0.4545,
+      "step": 6326
+    },
+    {
+      "epoch": 0.27785766819983093,
+      "grad_norm": 2.203125,
+      "learning_rate": 4.1196381606393e-05,
+      "loss": 0.4437,
+      "step": 6328
+    },
+    {
+      "epoch": 0.27794548667903446,
+      "grad_norm": 2.3125,
+      "learning_rate": 4.119111089058403e-05,
+      "loss": 0.4657,
+      "step": 6330
+    },
+    {
+      "epoch": 0.27803330515823793,
+      "grad_norm": 2.265625,
+      "learning_rate": 4.118583893485592e-05,
+      "loss": 0.4183,
+      "step": 6332
+    },
+    {
+      "epoch": 0.2781211236374414,
+      "grad_norm": 2.34375,
+      "learning_rate": 4.1180565739612365e-05,
+      "loss": 0.4495,
+      "step": 6334
+    },
+    {
+      "epoch": 0.2782089421166449,
+      "grad_norm": 2.1875,
+      "learning_rate": 4.117529130525721e-05,
+      "loss": 0.4459,
+      "step": 6336
+    },
+    {
+      "epoch": 0.2782967605958484,
+      "grad_norm": 2.46875,
+      "learning_rate": 4.117001563219438e-05,
+      "loss": 0.4392,
+      "step": 6338
+    },
+    {
+      "epoch": 0.2783845790750519,
+      "grad_norm": 2.078125,
+      "learning_rate": 4.1164738720827864e-05,
+      "loss": 0.4609,
+      "step": 6340
+    },
+    {
+      "epoch": 0.27847239755425535,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.1159460571561795e-05,
+      "loss": 0.451,
+      "step": 6342
+    },
+    {
+      "epoch": 0.2785602160334588,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.1154181184800344e-05,
+      "loss": 0.4578,
+      "step": 6344
+    },
+    {
+      "epoch": 0.27864803451266235,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.114890056094784e-05,
+      "loss": 0.4411,
+      "step": 6346
+    },
+    {
+      "epoch": 0.2787358529918658,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.114361870040866e-05,
+      "loss": 0.4315,
+      "step": 6348
+    },
+    {
+      "epoch": 0.2788236714710693,
+      "grad_norm": 2.53125,
+      "learning_rate": 4.1138335603587284e-05,
+      "loss": 0.4445,
+      "step": 6350
+    },
+    {
+      "epoch": 0.27891148995027276,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.11330512708883e-05,
+      "loss": 0.4583,
+      "step": 6352
+    },
+    {
+      "epoch": 0.2789993084294763,
+      "grad_norm": 2.375,
+      "learning_rate": 4.112776570271639e-05,
+      "loss": 0.4679,
+      "step": 6354
+    },
+    {
+      "epoch": 0.27908712690867976,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.11224788994763e-05,
+      "loss": 0.4387,
+      "step": 6356
+    },
+    {
+      "epoch": 0.27917494538788323,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.111719086157293e-05,
+      "loss": 0.4503,
+      "step": 6358
+    },
+    {
+      "epoch": 0.2792627638670867,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.111190158941121e-05,
+      "loss": 0.4793,
+      "step": 6360
+    },
+    {
+      "epoch": 0.27935058234629023,
+      "grad_norm": 3.171875,
+      "learning_rate": 4.11066110833962e-05,
+      "loss": 0.4142,
+      "step": 6362
+    },
+    {
+      "epoch": 0.2794384008254937,
+      "grad_norm": 3.265625,
+      "learning_rate": 4.1101319343933064e-05,
+      "loss": 0.4495,
+      "step": 6364
+    },
+    {
+      "epoch": 0.2795262193046972,
+      "grad_norm": 3.15625,
+      "learning_rate": 4.109602637142703e-05,
+      "loss": 0.4436,
+      "step": 6366
+    },
+    {
+      "epoch": 0.2796140377839007,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.109073216628343e-05,
+      "loss": 0.4644,
+      "step": 6368
+    },
+    {
+      "epoch": 0.2797018562631042,
+      "grad_norm": 2.34375,
+      "learning_rate": 4.108543672890771e-05,
+      "loss": 0.461,
+      "step": 6370
+    },
+    {
+      "epoch": 0.27978967474230765,
+      "grad_norm": 2.390625,
+      "learning_rate": 4.108014005970538e-05,
+      "loss": 0.4372,
+      "step": 6372
+    },
+    {
+      "epoch": 0.2798774932215111,
+      "grad_norm": 2.28125,
+      "learning_rate": 4.107484215908208e-05,
+      "loss": 0.4156,
+      "step": 6374
+    },
+    {
+      "epoch": 0.27996531170071465,
+      "grad_norm": 2.515625,
+      "learning_rate": 4.106954302744351e-05,
+      "loss": 0.4507,
+      "step": 6376
+    },
+    {
+      "epoch": 0.2800531301799181,
+      "grad_norm": 2.265625,
+      "learning_rate": 4.1064242665195486e-05,
+      "loss": 0.4574,
+      "step": 6378
+    },
+    {
+      "epoch": 0.2801409486591216,
+      "grad_norm": 3.015625,
+      "learning_rate": 4.105894107274391e-05,
+      "loss": 0.4322,
+      "step": 6380
+    },
+    {
+      "epoch": 0.28022876713832506,
+      "grad_norm": 2.203125,
+      "learning_rate": 4.105363825049476e-05,
+      "loss": 0.4504,
+      "step": 6382
+    },
+    {
+      "epoch": 0.2803165856175286,
+      "grad_norm": 2.109375,
+      "learning_rate": 4.104833419885417e-05,
+      "loss": 0.4319,
+      "step": 6384
+    },
+    {
+      "epoch": 0.28040440409673206,
+      "grad_norm": 2.265625,
+      "learning_rate": 4.104302891822828e-05,
+      "loss": 0.4637,
+      "step": 6386
+    },
+    {
+      "epoch": 0.28049222257593553,
+      "grad_norm": 2.125,
+      "learning_rate": 4.1037722409023396e-05,
+      "loss": 0.4497,
+      "step": 6388
+    },
+    {
+      "epoch": 0.280580041055139,
+      "grad_norm": 2.296875,
+      "learning_rate": 4.1032414671645894e-05,
+      "loss": 0.4482,
+      "step": 6390
+    },
+    {
+      "epoch": 0.28066785953434253,
+      "grad_norm": 2.75,
+      "learning_rate": 4.102710570650222e-05,
+      "loss": 0.457,
+      "step": 6392
+    },
+    {
+      "epoch": 0.280755678013546,
+      "grad_norm": 2.96875,
+      "learning_rate": 4.102179551399895e-05,
+      "loss": 0.4158,
+      "step": 6394
+    },
+    {
+      "epoch": 0.2808434964927495,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.1016484094542754e-05,
+      "loss": 0.4455,
+      "step": 6396
+    },
+    {
+      "epoch": 0.28093131497195295,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.101117144854035e-05,
+      "loss": 0.433,
+      "step": 6398
+    },
+    {
+      "epoch": 0.2810191334511565,
+      "grad_norm": 2.515625,
+      "learning_rate": 4.10058575763986e-05,
+      "loss": 0.4576,
+      "step": 6400
+    },
+    {
+      "epoch": 0.28110695193035995,
+      "grad_norm": 2.375,
+      "learning_rate": 4.100054247852445e-05,
+      "loss": 0.4346,
+      "step": 6402
+    },
+    {
+      "epoch": 0.2811947704095634,
+      "grad_norm": 2.390625,
+      "learning_rate": 4.099522615532491e-05,
+      "loss": 0.4536,
+      "step": 6404
+    },
+    {
+      "epoch": 0.2812825888887669,
+      "grad_norm": 2.453125,
+      "learning_rate": 4.098990860720712e-05,
+      "loss": 0.4584,
+      "step": 6406
+    },
+    {
+      "epoch": 0.2813704073679704,
+      "grad_norm": 2.25,
+      "learning_rate": 4.09845898345783e-05,
+      "loss": 0.4555,
+      "step": 6408
+    },
+    {
+      "epoch": 0.2814582258471739,
+      "grad_norm": 2.34375,
+      "learning_rate": 4.0979269837845754e-05,
+      "loss": 0.4722,
+      "step": 6410
+    },
+    {
+      "epoch": 0.28154604432637736,
+      "grad_norm": 2.328125,
+      "learning_rate": 4.09739486174169e-05,
+      "loss": 0.4457,
+      "step": 6412
+    },
+    {
+      "epoch": 0.28163386280558084,
+      "grad_norm": 2.125,
+      "learning_rate": 4.0968626173699234e-05,
+      "loss": 0.4123,
+      "step": 6414
+    },
+    {
+      "epoch": 0.28172168128478436,
+      "grad_norm": 2.0625,
+      "learning_rate": 4.0963302507100336e-05,
+      "loss": 0.4087,
+      "step": 6416
+    },
+    {
+      "epoch": 0.28180949976398784,
+      "grad_norm": 2.34375,
+      "learning_rate": 4.095797761802791e-05,
+      "loss": 0.4396,
+      "step": 6418
+    },
+    {
+      "epoch": 0.2818973182431913,
+      "grad_norm": 2.03125,
+      "learning_rate": 4.0952651506889735e-05,
+      "loss": 0.4509,
+      "step": 6420
+    },
+    {
+      "epoch": 0.28198513672239484,
+      "grad_norm": 2.140625,
+      "learning_rate": 4.094732417409368e-05,
+      "loss": 0.4242,
+      "step": 6422
+    },
+    {
+      "epoch": 0.2820729552015983,
+      "grad_norm": 2.0,
+      "learning_rate": 4.094199562004772e-05,
+      "loss": 0.428,
+      "step": 6424
+    },
+    {
+      "epoch": 0.2821607736808018,
+      "grad_norm": 2.328125,
+      "learning_rate": 4.0936665845159915e-05,
+      "loss": 0.4399,
+      "step": 6426
+    },
+    {
+      "epoch": 0.28224859216000525,
+      "grad_norm": 2.15625,
+      "learning_rate": 4.0931334849838414e-05,
+      "loss": 0.4446,
+      "step": 6428
+    },
+    {
+      "epoch": 0.2823364106392088,
+      "grad_norm": 2.296875,
+      "learning_rate": 4.0926002634491476e-05,
+      "loss": 0.4622,
+      "step": 6430
+    },
+    {
+      "epoch": 0.28242422911841225,
+      "grad_norm": 2.140625,
+      "learning_rate": 4.092066919952743e-05,
+      "loss": 0.4531,
+      "step": 6432
+    },
+    {
+      "epoch": 0.2825120475976157,
+      "grad_norm": 2.140625,
+      "learning_rate": 4.0915334545354734e-05,
+      "loss": 0.4387,
+      "step": 6434
+    },
+    {
+      "epoch": 0.2825998660768192,
+      "grad_norm": 2.09375,
+      "learning_rate": 4.0909998672381897e-05,
+      "loss": 0.4392,
+      "step": 6436
+    },
+    {
+      "epoch": 0.2826876845560227,
+      "grad_norm": 2.359375,
+      "learning_rate": 4.090466158101754e-05,
+      "loss": 0.4385,
+      "step": 6438
+    },
+    {
+      "epoch": 0.2827755030352262,
+      "grad_norm": 2.109375,
+      "learning_rate": 4.0899323271670395e-05,
+      "loss": 0.4336,
+      "step": 6440
+    },
+    {
+      "epoch": 0.28286332151442967,
+      "grad_norm": 2.390625,
+      "learning_rate": 4.0893983744749265e-05,
+      "loss": 0.4414,
+      "step": 6442
+    },
+    {
+      "epoch": 0.28295113999363314,
+      "grad_norm": 2.15625,
+      "learning_rate": 4.088864300066304e-05,
+      "loss": 0.4467,
+      "step": 6444
+    },
+    {
+      "epoch": 0.28303895847283667,
+      "grad_norm": 2.09375,
+      "learning_rate": 4.088330103982074e-05,
+      "loss": 0.4311,
+      "step": 6446
+    },
+    {
+      "epoch": 0.28312677695204014,
+      "grad_norm": 2.546875,
+      "learning_rate": 4.0877957862631425e-05,
+      "loss": 0.4265,
+      "step": 6448
+    },
+    {
+      "epoch": 0.2832145954312436,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.087261346950429e-05,
+      "loss": 0.4405,
+      "step": 6450
+    },
+    {
+      "epoch": 0.2833024139104471,
+      "grad_norm": 2.1875,
+      "learning_rate": 4.086726786084862e-05,
+      "loss": 0.42,
+      "step": 6452
+    },
+    {
+      "epoch": 0.2833902323896506,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.086192103707377e-05,
+      "loss": 0.4197,
+      "step": 6454
+    },
+    {
+      "epoch": 0.2834780508688541,
+      "grad_norm": 2.546875,
+      "learning_rate": 4.0856572998589206e-05,
+      "loss": 0.4347,
+      "step": 6456
+    },
+    {
+      "epoch": 0.28356586934805755,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.0851223745804476e-05,
+      "loss": 0.4318,
+      "step": 6458
+    },
+    {
+      "epoch": 0.283653687827261,
+      "grad_norm": 2.25,
+      "learning_rate": 4.0845873279129246e-05,
+      "loss": 0.4353,
+      "step": 6460
+    },
+    {
+      "epoch": 0.28374150630646455,
+      "grad_norm": 2.109375,
+      "learning_rate": 4.0840521598973223e-05,
+      "loss": 0.4395,
+      "step": 6462
+    },
+    {
+      "epoch": 0.283829324785668,
+      "grad_norm": 2.203125,
+      "learning_rate": 4.083516870574626e-05,
+      "loss": 0.4345,
+      "step": 6464
+    },
+    {
+      "epoch": 0.2839171432648715,
+      "grad_norm": 2.328125,
+      "learning_rate": 4.0829814599858296e-05,
+      "loss": 0.4437,
+      "step": 6466
+    },
+    {
+      "epoch": 0.284004961744075,
+      "grad_norm": 2.140625,
+      "learning_rate": 4.0824459281719326e-05,
+      "loss": 0.4519,
+      "step": 6468
+    },
+    {
+      "epoch": 0.2840927802232785,
+      "grad_norm": 2.015625,
+      "learning_rate": 4.0819102751739466e-05,
+      "loss": 0.4521,
+      "step": 6470
+    },
+    {
+      "epoch": 0.28418059870248197,
+      "grad_norm": 2.171875,
+      "learning_rate": 4.081374501032894e-05,
+      "loss": 0.462,
+      "step": 6472
+    },
+    {
+      "epoch": 0.28426841718168544,
+      "grad_norm": 2.171875,
+      "learning_rate": 4.080838605789802e-05,
+      "loss": 0.4368,
+      "step": 6474
+    },
+    {
+      "epoch": 0.28435623566088897,
+      "grad_norm": 2.0625,
+      "learning_rate": 4.08030258948571e-05,
+      "loss": 0.4292,
+      "step": 6476
+    },
+    {
+      "epoch": 0.28444405414009244,
+      "grad_norm": 2.078125,
+      "learning_rate": 4.0797664521616684e-05,
+      "loss": 0.4388,
+      "step": 6478
+    },
+    {
+      "epoch": 0.2845318726192959,
+      "grad_norm": 2.171875,
+      "learning_rate": 4.079230193858732e-05,
+      "loss": 0.4426,
+      "step": 6480
+    },
+    {
+      "epoch": 0.2846196910984994,
+      "grad_norm": 2.28125,
+      "learning_rate": 4.078693814617969e-05,
+      "loss": 0.4409,
+      "step": 6482
+    },
+    {
+      "epoch": 0.2847075095777029,
+      "grad_norm": 2.015625,
+      "learning_rate": 4.078157314480456e-05,
+      "loss": 0.4673,
+      "step": 6484
+    },
+    {
+      "epoch": 0.2847953280569064,
+      "grad_norm": 2.15625,
+      "learning_rate": 4.077620693487277e-05,
+      "loss": 0.4272,
+      "step": 6486
+    },
+    {
+      "epoch": 0.28488314653610985,
+      "grad_norm": 2.21875,
+      "learning_rate": 4.0770839516795265e-05,
+      "loss": 0.4592,
+      "step": 6488
+    },
+    {
+      "epoch": 0.2849709650153133,
+      "grad_norm": 2.3125,
+      "learning_rate": 4.07654708909831e-05,
+      "loss": 0.4513,
+      "step": 6490
+    },
+    {
+      "epoch": 0.28505878349451685,
+      "grad_norm": 2.28125,
+      "learning_rate": 4.076010105784739e-05,
+      "loss": 0.4398,
+      "step": 6492
+    },
+    {
+      "epoch": 0.2851466019737203,
+      "grad_norm": 2.109375,
+      "learning_rate": 4.075473001779936e-05,
+      "loss": 0.4455,
+      "step": 6494
+    },
+    {
+      "epoch": 0.2852344204529238,
+      "grad_norm": 2.484375,
+      "learning_rate": 4.0749357771250335e-05,
+      "loss": 0.4208,
+      "step": 6496
+    },
+    {
+      "epoch": 0.28532223893212727,
+      "grad_norm": 2.28125,
+      "learning_rate": 4.074398431861171e-05,
+      "loss": 0.4249,
+      "step": 6498
+    },
+    {
+      "epoch": 0.2854100574113308,
+      "grad_norm": 2.453125,
+      "learning_rate": 4.0738609660295e-05,
+      "loss": 0.4631,
+      "step": 6500
+    },
+    {
+      "epoch": 0.28549787589053427,
+      "grad_norm": 2.265625,
+      "learning_rate": 4.073323379671179e-05,
+      "loss": 0.4618,
+      "step": 6502
+    },
+    {
+      "epoch": 0.28558569436973774,
+      "grad_norm": 2.203125,
+      "learning_rate": 4.072785672827375e-05,
+      "loss": 0.4415,
+      "step": 6504
+    },
+    {
+      "epoch": 0.2856735128489412,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.072247845539268e-05,
+      "loss": 0.4629,
+      "step": 6506
+    },
+    {
+      "epoch": 0.28576133132814474,
+      "grad_norm": 2.28125,
+      "learning_rate": 4.0717098978480444e-05,
+      "loss": 0.45,
+      "step": 6508
+    },
+    {
+      "epoch": 0.2858491498073482,
+      "grad_norm": 2.03125,
+      "learning_rate": 4.0711718297949e-05,
+      "loss": 0.4641,
+      "step": 6510
+    },
+    {
+      "epoch": 0.2859369682865517,
+      "grad_norm": 2.203125,
+      "learning_rate": 4.07063364142104e-05,
+      "loss": 0.429,
+      "step": 6512
+    },
+    {
+      "epoch": 0.28602478676575516,
+      "grad_norm": 2.34375,
+      "learning_rate": 4.0700953327676797e-05,
+      "loss": 0.4409,
+      "step": 6514
+    },
+    {
+      "epoch": 0.2861126052449587,
+      "grad_norm": 2.40625,
+      "learning_rate": 4.0695569038760416e-05,
+      "loss": 0.4768,
+      "step": 6516
+    },
+    {
+      "epoch": 0.28620042372416216,
+      "grad_norm": 2.234375,
+      "learning_rate": 4.0690183547873594e-05,
+      "loss": 0.453,
+      "step": 6518
+    },
+    {
+      "epoch": 0.28628824220336563,
+      "grad_norm": 2.125,
+      "learning_rate": 4.068479685542876e-05,
+      "loss": 0.4708,
+      "step": 6520
+    },
+    {
+      "epoch": 0.28637606068256916,
+      "grad_norm": 2.28125,
+      "learning_rate": 4.067940896183843e-05,
+      "loss": 0.4502,
+      "step": 6522
+    },
+    {
+      "epoch": 0.2864638791617726,
+      "grad_norm": 2.140625,
+      "learning_rate": 4.067401986751519e-05,
+      "loss": 0.4225,
+      "step": 6524
+    },
+    {
+      "epoch": 0.2865516976409761,
+      "grad_norm": 2.21875,
+      "learning_rate": 4.0668629572871765e-05,
+      "loss": 0.4193,
+      "step": 6526
+    },
+    {
+      "epoch": 0.28663951612017957,
+      "grad_norm": 2.484375,
+      "learning_rate": 4.066323807832092e-05,
+      "loss": 0.4282,
+      "step": 6528
+    },
+    {
+      "epoch": 0.2867273345993831,
+      "grad_norm": 2.296875,
+      "learning_rate": 4.065784538427555e-05,
+      "loss": 0.4534,
+      "step": 6530
+    },
+    {
+      "epoch": 0.28681515307858657,
+      "grad_norm": 2.328125,
+      "learning_rate": 4.0652451491148636e-05,
+      "loss": 0.4383,
+      "step": 6532
+    },
+    {
+      "epoch": 0.28690297155779004,
+      "grad_norm": 2.21875,
+      "learning_rate": 4.0647056399353225e-05,
+      "loss": 0.4463,
+      "step": 6534
+    },
+    {
+      "epoch": 0.2869907900369935,
+      "grad_norm": 2.328125,
+      "learning_rate": 4.0641660109302485e-05,
+      "loss": 0.4432,
+      "step": 6536
+    },
+    {
+      "epoch": 0.28707860851619704,
+      "grad_norm": 2.015625,
+      "learning_rate": 4.063626262140967e-05,
+      "loss": 0.4369,
+      "step": 6538
+    },
+    {
+      "epoch": 0.2871664269954005,
+      "grad_norm": 2.28125,
+      "learning_rate": 4.0630863936088104e-05,
+      "loss": 0.462,
+      "step": 6540
+    },
+    {
+      "epoch": 0.287254245474604,
+      "grad_norm": 2.015625,
+      "learning_rate": 4.062546405375124e-05,
+      "loss": 0.451,
+      "step": 6542
+    },
+    {
+      "epoch": 0.28734206395380746,
+      "grad_norm": 2.203125,
+      "learning_rate": 4.062006297481259e-05,
+      "loss": 0.4816,
+      "step": 6544
+    },
+    {
+      "epoch": 0.287429882433011,
+      "grad_norm": 2.21875,
+      "learning_rate": 4.061466069968577e-05,
+      "loss": 0.4531,
+      "step": 6546
+    },
+    {
+      "epoch": 0.28751770091221446,
+      "grad_norm": 2.078125,
+      "learning_rate": 4.06092572287845e-05,
+      "loss": 0.4612,
+      "step": 6548
+    },
+    {
+      "epoch": 0.28760551939141793,
+      "grad_norm": 2.3125,
+      "learning_rate": 4.0603852562522564e-05,
+      "loss": 0.4173,
+      "step": 6550
+    },
+    {
+      "epoch": 0.2876933378706214,
+      "grad_norm": 2.234375,
+      "learning_rate": 4.0598446701313865e-05,
+      "loss": 0.457,
+      "step": 6552
+    },
+    {
+      "epoch": 0.28778115634982493,
+      "grad_norm": 2.421875,
+      "learning_rate": 4.059303964557237e-05,
+      "loss": 0.4227,
+      "step": 6554
+    },
+    {
+      "epoch": 0.2878689748290284,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.058763139571216e-05,
+      "loss": 0.4605,
+      "step": 6556
+    },
+    {
+      "epoch": 0.2879567933082319,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.05822219521474e-05,
+      "loss": 0.438,
+      "step": 6558
+    },
+    {
+      "epoch": 0.28804461178743535,
+      "grad_norm": 2.453125,
+      "learning_rate": 4.057681131529235e-05,
+      "loss": 0.4462,
+      "step": 6560
+    },
+    {
+      "epoch": 0.2881324302666389,
+      "grad_norm": 2.328125,
+      "learning_rate": 4.0571399485561366e-05,
+      "loss": 0.4389,
+      "step": 6562
+    },
+    {
+      "epoch": 0.28822024874584234,
+      "grad_norm": 2.34375,
+      "learning_rate": 4.0565986463368865e-05,
+      "loss": 0.4411,
+      "step": 6564
+    },
+    {
+      "epoch": 0.2883080672250458,
+      "grad_norm": 2.296875,
+      "learning_rate": 4.0560572249129394e-05,
+      "loss": 0.4564,
+      "step": 6566
+    },
+    {
+      "epoch": 0.28839588570424934,
+      "grad_norm": 2.0625,
+      "learning_rate": 4.0555156843257566e-05,
+      "loss": 0.4626,
+      "step": 6568
+    },
+    {
+      "epoch": 0.2884837041834528,
+      "grad_norm": 2.109375,
+      "learning_rate": 4.05497402461681e-05,
+      "loss": 0.4507,
+      "step": 6570
+    },
+    {
+      "epoch": 0.2885715226626563,
+      "grad_norm": 1.9453125,
+      "learning_rate": 4.0544322458275796e-05,
+      "loss": 0.4225,
+      "step": 6572
+    },
+    {
+      "epoch": 0.28865934114185976,
+      "grad_norm": 2.34375,
+      "learning_rate": 4.0538903479995554e-05,
+      "loss": 0.4069,
+      "step": 6574
+    },
+    {
+      "epoch": 0.2887471596210633,
+      "grad_norm": 2.421875,
+      "learning_rate": 4.053348331174236e-05,
+      "loss": 0.4189,
+      "step": 6576
+    },
+    {
+      "epoch": 0.28883497810026676,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.052806195393129e-05,
+      "loss": 0.4453,
+      "step": 6578
+    },
+    {
+      "epoch": 0.28892279657947023,
+      "grad_norm": 2.984375,
+      "learning_rate": 4.0522639406977516e-05,
+      "loss": 0.4598,
+      "step": 6580
+    },
+    {
+      "epoch": 0.2890106150586737,
+      "grad_norm": 3.484375,
+      "learning_rate": 4.051721567129629e-05,
+      "loss": 0.4437,
+      "step": 6582
+    },
+    {
+      "epoch": 0.28909843353787723,
+      "grad_norm": 2.625,
+      "learning_rate": 4.051179074730297e-05,
+      "loss": 0.446,
+      "step": 6584
+    },
+    {
+      "epoch": 0.2891862520170807,
+      "grad_norm": 2.5,
+      "learning_rate": 4.0506364635413e-05,
+      "loss": 0.4417,
+      "step": 6586
+    },
+    {
+      "epoch": 0.2892740704962842,
+      "grad_norm": 2.046875,
+      "learning_rate": 4.05009373360419e-05,
+      "loss": 0.4179,
+      "step": 6588
+    },
+    {
+      "epoch": 0.28936188897548765,
+      "grad_norm": 2.375,
+      "learning_rate": 4.049550884960531e-05,
+      "loss": 0.4153,
+      "step": 6590
+    },
+    {
+      "epoch": 0.2894497074546912,
+      "grad_norm": 2.40625,
+      "learning_rate": 4.049007917651894e-05,
+      "loss": 0.457,
+      "step": 6592
+    },
+    {
+      "epoch": 0.28953752593389465,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.0484648317198585e-05,
+      "loss": 0.4651,
+      "step": 6594
+    },
+    {
+      "epoch": 0.2896253444130981,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.047921627206015e-05,
+      "loss": 0.4265,
+      "step": 6596
+    },
+    {
+      "epoch": 0.2897131628923016,
+      "grad_norm": 2.265625,
+      "learning_rate": 4.047378304151963e-05,
+      "loss": 0.437,
+      "step": 6598
+    },
+    {
+      "epoch": 0.2898009813715051,
+      "grad_norm": 2.296875,
+      "learning_rate": 4.046834862599309e-05,
+      "loss": 0.4362,
+      "step": 6600
+    },
+    {
+      "epoch": 0.2898887998507086,
+      "grad_norm": 2.015625,
+      "learning_rate": 4.046291302589671e-05,
+      "loss": 0.4426,
+      "step": 6602
+    },
+    {
+      "epoch": 0.28997661832991206,
+      "grad_norm": 2.328125,
+      "learning_rate": 4.045747624164674e-05,
+      "loss": 0.4191,
+      "step": 6604
+    },
+    {
+      "epoch": 0.29006443680911553,
+      "grad_norm": 2.046875,
+      "learning_rate": 4.045203827365953e-05,
+      "loss": 0.3972,
+      "step": 6606
+    },
+    {
+      "epoch": 0.29015225528831906,
+      "grad_norm": 2.09375,
+      "learning_rate": 4.0446599122351535e-05,
+      "loss": 0.4173,
+      "step": 6608
+    },
+    {
+      "epoch": 0.29024007376752253,
+      "grad_norm": 1.9765625,
+      "learning_rate": 4.044115878813927e-05,
+      "loss": 0.4359,
+      "step": 6610
+    },
+    {
+      "epoch": 0.290327892246726,
+      "grad_norm": 2.15625,
+      "learning_rate": 4.043571727143936e-05,
+      "loss": 0.4373,
+      "step": 6612
+    },
+    {
+      "epoch": 0.2904157107259295,
+      "grad_norm": 2.46875,
+      "learning_rate": 4.043027457266853e-05,
+      "loss": 0.4602,
+      "step": 6614
+    },
+    {
+      "epoch": 0.290503529205133,
+      "grad_norm": 2.359375,
+      "learning_rate": 4.0424830692243566e-05,
+      "loss": 0.4517,
+      "step": 6616
+    },
+    {
+      "epoch": 0.2905913476843365,
+      "grad_norm": 2.40625,
+      "learning_rate": 4.0419385630581376e-05,
+      "loss": 0.4215,
+      "step": 6618
+    },
+    {
+      "epoch": 0.29067916616353995,
+      "grad_norm": 2.328125,
+      "learning_rate": 4.041393938809893e-05,
+      "loss": 0.4377,
+      "step": 6620
+    },
+    {
+      "epoch": 0.2907669846427435,
+      "grad_norm": 1.9765625,
+      "learning_rate": 4.0408491965213315e-05,
+      "loss": 0.4292,
+      "step": 6622
+    },
+    {
+      "epoch": 0.29085480312194695,
+      "grad_norm": 2.109375,
+      "learning_rate": 4.04030433623417e-05,
+      "loss": 0.4647,
+      "step": 6624
+    },
+    {
+      "epoch": 0.2909426216011504,
+      "grad_norm": 2.234375,
+      "learning_rate": 4.039759357990133e-05,
+      "loss": 0.416,
+      "step": 6626
+    },
+    {
+      "epoch": 0.2910304400803539,
+      "grad_norm": 2.234375,
+      "learning_rate": 4.039214261830954e-05,
+      "loss": 0.4596,
+      "step": 6628
+    },
+    {
+      "epoch": 0.2911182585595574,
+      "grad_norm": 2.28125,
+      "learning_rate": 4.0386690477983786e-05,
+      "loss": 0.4676,
+      "step": 6630
+    },
+    {
+      "epoch": 0.2912060770387609,
+      "grad_norm": 2.1875,
+      "learning_rate": 4.038123715934158e-05,
+      "loss": 0.4404,
+      "step": 6632
+    },
+    {
+      "epoch": 0.29129389551796436,
+      "grad_norm": 2.21875,
+      "learning_rate": 4.0375782662800555e-05,
+      "loss": 0.4359,
+      "step": 6634
+    },
+    {
+      "epoch": 0.29138171399716783,
+      "grad_norm": 2.28125,
+      "learning_rate": 4.037032698877841e-05,
+      "loss": 0.4296,
+      "step": 6636
+    },
+    {
+      "epoch": 0.29146953247637136,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.0364870137692925e-05,
+      "loss": 0.4485,
+      "step": 6638
+    },
+    {
+      "epoch": 0.29155735095557483,
+      "grad_norm": 1.984375,
+      "learning_rate": 4.035941210996202e-05,
+      "loss": 0.4222,
+      "step": 6640
+    },
+    {
+      "epoch": 0.2916451694347783,
+      "grad_norm": 2.203125,
+      "learning_rate": 4.035395290600365e-05,
+      "loss": 0.4314,
+      "step": 6642
+    },
+    {
+      "epoch": 0.2917329879139818,
+      "grad_norm": 2.359375,
+      "learning_rate": 4.034849252623587e-05,
+      "loss": 0.455,
+      "step": 6644
+    },
+    {
+      "epoch": 0.2918208063931853,
+      "grad_norm": 1.875,
+      "learning_rate": 4.034303097107687e-05,
+      "loss": 0.4175,
+      "step": 6646
+    },
+    {
+      "epoch": 0.2919086248723888,
+      "grad_norm": 2.09375,
+      "learning_rate": 4.033756824094487e-05,
+      "loss": 0.4194,
+      "step": 6648
+    },
+    {
+      "epoch": 0.29199644335159225,
+      "grad_norm": 2.140625,
+      "learning_rate": 4.033210433625822e-05,
+      "loss": 0.4332,
+      "step": 6650
+    },
+    {
+      "epoch": 0.2920842618307957,
+      "grad_norm": 2.25,
+      "learning_rate": 4.0326639257435343e-05,
+      "loss": 0.4202,
+      "step": 6652
+    },
+    {
+      "epoch": 0.29217208030999925,
+      "grad_norm": 2.3125,
+      "learning_rate": 4.032117300489476e-05,
+      "loss": 0.4369,
+      "step": 6654
+    },
+    {
+      "epoch": 0.2922598987892027,
+      "grad_norm": 2.359375,
+      "learning_rate": 4.031570557905508e-05,
+      "loss": 0.4293,
+      "step": 6656
+    },
+    {
+      "epoch": 0.2923477172684062,
+      "grad_norm": 2.234375,
+      "learning_rate": 4.031023698033499e-05,
+      "loss": 0.4306,
+      "step": 6658
+    },
+    {
+      "epoch": 0.29243553574760967,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.030476720915328e-05,
+      "loss": 0.4439,
+      "step": 6660
+    },
+    {
+      "epoch": 0.2925233542268132,
+      "grad_norm": 2.265625,
+      "learning_rate": 4.029929626592884e-05,
+      "loss": 0.4316,
+      "step": 6662
+    },
+    {
+      "epoch": 0.29261117270601666,
+      "grad_norm": 2.28125,
+      "learning_rate": 4.0293824151080614e-05,
+      "loss": 0.4389,
+      "step": 6664
+    },
+    {
+      "epoch": 0.29269899118522014,
+      "grad_norm": 2.203125,
+      "learning_rate": 4.028835086502767e-05,
+      "loss": 0.423,
+      "step": 6666
+    },
+    {
+      "epoch": 0.29278680966442366,
+      "grad_norm": 2.0625,
+      "learning_rate": 4.028287640818915e-05,
+      "loss": 0.4596,
+      "step": 6668
+    },
+    {
+      "epoch": 0.29287462814362714,
+      "grad_norm": 2.265625,
+      "learning_rate": 4.02774007809843e-05,
+      "loss": 0.4145,
+      "step": 6670
+    },
+    {
+      "epoch": 0.2929624466228306,
+      "grad_norm": 2.21875,
+      "learning_rate": 4.027192398383243e-05,
+      "loss": 0.4104,
+      "step": 6672
+    },
+    {
+      "epoch": 0.2930502651020341,
+      "grad_norm": 2.234375,
+      "learning_rate": 4.0266446017152956e-05,
+      "loss": 0.4175,
+      "step": 6674
+    },
+    {
+      "epoch": 0.2931380835812376,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.0260966881365395e-05,
+      "loss": 0.4496,
+      "step": 6676
+    },
+    {
+      "epoch": 0.2932259020604411,
+      "grad_norm": 2.25,
+      "learning_rate": 4.0255486576889315e-05,
+      "loss": 0.4373,
+      "step": 6678
+    },
+    {
+      "epoch": 0.29331372053964455,
+      "grad_norm": 2.359375,
+      "learning_rate": 4.0250005104144425e-05,
+      "loss": 0.4386,
+      "step": 6680
+    },
+    {
+      "epoch": 0.293401539018848,
+      "grad_norm": 2.140625,
+      "learning_rate": 4.0244522463550494e-05,
+      "loss": 0.4338,
+      "step": 6682
+    },
+    {
+      "epoch": 0.29348935749805155,
+      "grad_norm": 2.046875,
+      "learning_rate": 4.023903865552738e-05,
+      "loss": 0.4828,
+      "step": 6684
+    },
+    {
+      "epoch": 0.293577175977255,
+      "grad_norm": 1.9296875,
+      "learning_rate": 4.0233553680495027e-05,
+      "loss": 0.4387,
+      "step": 6686
+    },
+    {
+      "epoch": 0.2936649944564585,
+      "grad_norm": 2.296875,
+      "learning_rate": 4.022806753887349e-05,
+      "loss": 0.4357,
+      "step": 6688
+    },
+    {
+      "epoch": 0.29375281293566197,
+      "grad_norm": 2.203125,
+      "learning_rate": 4.022258023108288e-05,
+      "loss": 0.4067,
+      "step": 6690
+    },
+    {
+      "epoch": 0.2938406314148655,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.021709175754342e-05,
+      "loss": 0.4142,
+      "step": 6692
+    },
+    {
+      "epoch": 0.29392844989406897,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.021160211867544e-05,
+      "loss": 0.4647,
+      "step": 6694
+    },
+    {
+      "epoch": 0.29401626837327244,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.020611131489932e-05,
+      "loss": 0.4448,
+      "step": 6696
+    },
+    {
+      "epoch": 0.2941040868524759,
+      "grad_norm": 2.25,
+      "learning_rate": 4.020061934663555e-05,
+      "loss": 0.4183,
+      "step": 6698
+    },
+    {
+      "epoch": 0.29419190533167944,
+      "grad_norm": 2.375,
+      "learning_rate": 4.0195126214304704e-05,
+      "loss": 0.4467,
+      "step": 6700
+    },
+    {
+      "epoch": 0.2942797238108829,
+      "grad_norm": 2.140625,
+      "learning_rate": 4.018963191832746e-05,
+      "loss": 0.4432,
+      "step": 6702
+    },
+    {
+      "epoch": 0.2943675422900864,
+      "grad_norm": 2.203125,
+      "learning_rate": 4.018413645912455e-05,
+      "loss": 0.4159,
+      "step": 6704
+    },
+    {
+      "epoch": 0.29445536076928985,
+      "grad_norm": 2.203125,
+      "learning_rate": 4.0178639837116836e-05,
+      "loss": 0.412,
+      "step": 6706
+    },
+    {
+      "epoch": 0.2945431792484934,
+      "grad_norm": 2.328125,
+      "learning_rate": 4.0173142052725244e-05,
+      "loss": 0.4499,
+      "step": 6708
+    },
+    {
+      "epoch": 0.29463099772769685,
+      "grad_norm": 2.328125,
+      "learning_rate": 4.0167643106370786e-05,
+      "loss": 0.4282,
+      "step": 6710
+    },
+    {
+      "epoch": 0.2947188162069003,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.016214299847459e-05,
+      "loss": 0.4674,
+      "step": 6712
+    },
+    {
+      "epoch": 0.2948066346861038,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.0156641729457855e-05,
+      "loss": 0.4387,
+      "step": 6714
+    },
+    {
+      "epoch": 0.2948944531653073,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.015113929974187e-05,
+      "loss": 0.4541,
+      "step": 6716
+    },
+    {
+      "epoch": 0.2949822716445108,
+      "grad_norm": 2.546875,
+      "learning_rate": 4.014563570974799e-05,
+      "loss": 0.4202,
+      "step": 6718
+    },
+    {
+      "epoch": 0.29507009012371427,
+      "grad_norm": 3.296875,
+      "learning_rate": 4.01401309598977e-05,
+      "loss": 0.4555,
+      "step": 6720
+    },
+    {
+      "epoch": 0.2951579086029178,
+      "grad_norm": 2.484375,
+      "learning_rate": 4.013462505061255e-05,
+      "loss": 0.4327,
+      "step": 6722
+    },
+    {
+      "epoch": 0.29524572708212127,
+      "grad_norm": 2.3125,
+      "learning_rate": 4.0129117982314194e-05,
+      "loss": 0.4166,
+      "step": 6724
+    },
+    {
+      "epoch": 0.29533354556132474,
+      "grad_norm": 2.390625,
+      "learning_rate": 4.012360975542434e-05,
+      "loss": 0.3892,
+      "step": 6726
+    },
+    {
+      "epoch": 0.2954213640405282,
+      "grad_norm": 2.34375,
+      "learning_rate": 4.011810037036484e-05,
+      "loss": 0.4572,
+      "step": 6728
+    },
+    {
+      "epoch": 0.29550918251973174,
+      "grad_norm": 2.390625,
+      "learning_rate": 4.011258982755759e-05,
+      "loss": 0.4488,
+      "step": 6730
+    },
+    {
+      "epoch": 0.2955970009989352,
+      "grad_norm": 2.171875,
+      "learning_rate": 4.010707812742459e-05,
+      "loss": 0.4369,
+      "step": 6732
+    },
+    {
+      "epoch": 0.2956848194781387,
+      "grad_norm": 2.296875,
+      "learning_rate": 4.010156527038791e-05,
+      "loss": 0.433,
+      "step": 6734
+    },
+    {
+      "epoch": 0.29577263795734215,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.009605125686975e-05,
+      "loss": 0.4481,
+      "step": 6736
+    },
+    {
+      "epoch": 0.2958604564365457,
+      "grad_norm": 2.421875,
+      "learning_rate": 4.009053608729237e-05,
+      "loss": 0.4412,
+      "step": 6738
+    },
+    {
+      "epoch": 0.29594827491574915,
+      "grad_norm": 2.359375,
+      "learning_rate": 4.0085019762078116e-05,
+      "loss": 0.4396,
+      "step": 6740
+    },
+    {
+      "epoch": 0.2960360933949526,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.007950228164943e-05,
+      "loss": 0.4157,
+      "step": 6742
+    },
+    {
+      "epoch": 0.2961239118741561,
+      "grad_norm": 2.328125,
+      "learning_rate": 4.007398364642885e-05,
+      "loss": 0.4439,
+      "step": 6744
+    },
+    {
+      "epoch": 0.2962117303533596,
+      "grad_norm": 2.46875,
+      "learning_rate": 4.006846385683899e-05,
+      "loss": 0.4452,
+      "step": 6746
+    },
+    {
+      "epoch": 0.2962995488325631,
+      "grad_norm": 2.875,
+      "learning_rate": 4.006294291330255e-05,
+      "loss": 0.4207,
+      "step": 6748
+    },
+    {
+      "epoch": 0.29638736731176657,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.005742081624233e-05,
+      "loss": 0.4459,
+      "step": 6750
+    },
+    {
+      "epoch": 0.29647518579097004,
+      "grad_norm": 2.53125,
+      "learning_rate": 4.005189756608122e-05,
+      "loss": 0.433,
+      "step": 6752
+    },
+    {
+      "epoch": 0.29656300427017357,
+      "grad_norm": 2.359375,
+      "learning_rate": 4.004637316324218e-05,
+      "loss": 0.4451,
+      "step": 6754
+    },
+    {
+      "epoch": 0.29665082274937704,
+      "grad_norm": 2.265625,
+      "learning_rate": 4.004084760814828e-05,
+      "loss": 0.4489,
+      "step": 6756
+    },
+    {
+      "epoch": 0.2967386412285805,
+      "grad_norm": 2.390625,
+      "learning_rate": 4.003532090122266e-05,
+      "loss": 0.4177,
+      "step": 6758
+    },
+    {
+      "epoch": 0.296826459707784,
+      "grad_norm": 2.25,
+      "learning_rate": 4.002979304288857e-05,
+      "loss": 0.4251,
+      "step": 6760
+    },
+    {
+      "epoch": 0.2969142781869875,
+      "grad_norm": 2.375,
+      "learning_rate": 4.002426403356932e-05,
+      "loss": 0.4323,
+      "step": 6762
+    },
+    {
+      "epoch": 0.297002096666191,
+      "grad_norm": 2.125,
+      "learning_rate": 4.001873387368833e-05,
+      "loss": 0.4477,
+      "step": 6764
+    },
+    {
+      "epoch": 0.29708991514539446,
+      "grad_norm": 2.296875,
+      "learning_rate": 4.0013202563669104e-05,
+      "loss": 0.4452,
+      "step": 6766
+    },
+    {
+      "epoch": 0.29717773362459793,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.000767010393522e-05,
+      "loss": 0.4423,
+      "step": 6768
+    },
+    {
+      "epoch": 0.29726555210380146,
+      "grad_norm": 2.265625,
+      "learning_rate": 4.0002136494910366e-05,
+      "loss": 0.4301,
+      "step": 6770
+    },
+    {
+      "epoch": 0.29735337058300493,
+      "grad_norm": 2.5,
+      "learning_rate": 3.9996601737018294e-05,
+      "loss": 0.4507,
+      "step": 6772
+    },
+    {
+      "epoch": 0.2974411890622084,
+      "grad_norm": 2.234375,
+      "learning_rate": 3.9991065830682875e-05,
+      "loss": 0.4243,
+      "step": 6774
+    },
+    {
+      "epoch": 0.2975290075414119,
+      "grad_norm": 2.203125,
+      "learning_rate": 3.998552877632804e-05,
+      "loss": 0.4369,
+      "step": 6776
+    },
+    {
+      "epoch": 0.2976168260206154,
+      "grad_norm": 2.0,
+      "learning_rate": 3.9979990574377825e-05,
+      "loss": 0.4455,
+      "step": 6778
+    },
+    {
+      "epoch": 0.29770464449981887,
+      "grad_norm": 2.265625,
+      "learning_rate": 3.997445122525633e-05,
+      "loss": 0.413,
+      "step": 6780
+    },
+    {
+      "epoch": 0.29779246297902234,
+      "grad_norm": 2.046875,
+      "learning_rate": 3.996891072938778e-05,
+      "loss": 0.4478,
+      "step": 6782
+    },
+    {
+      "epoch": 0.29788028145822587,
+      "grad_norm": 2.359375,
+      "learning_rate": 3.9963369087196444e-05,
+      "loss": 0.4367,
+      "step": 6784
+    },
+    {
+      "epoch": 0.29796809993742934,
+      "grad_norm": 2.328125,
+      "learning_rate": 3.9957826299106724e-05,
+      "loss": 0.4002,
+      "step": 6786
+    },
+    {
+      "epoch": 0.2980559184166328,
+      "grad_norm": 2.28125,
+      "learning_rate": 3.9952282365543083e-05,
+      "loss": 0.4108,
+      "step": 6788
+    },
+    {
+      "epoch": 0.2981437368958363,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.9946737286930076e-05,
+      "loss": 0.4468,
+      "step": 6790
+    },
+    {
+      "epoch": 0.2982315553750398,
+      "grad_norm": 2.125,
+      "learning_rate": 3.994119106369234e-05,
+      "loss": 0.4116,
+      "step": 6792
+    },
+    {
+      "epoch": 0.2983193738542433,
+      "grad_norm": 2.203125,
+      "learning_rate": 3.993564369625461e-05,
+      "loss": 0.4294,
+      "step": 6794
+    },
+    {
+      "epoch": 0.29840719233344676,
+      "grad_norm": 2.296875,
+      "learning_rate": 3.993009518504171e-05,
+      "loss": 0.4148,
+      "step": 6796
+    },
+    {
+      "epoch": 0.29849501081265023,
+      "grad_norm": 2.546875,
+      "learning_rate": 3.9924545530478544e-05,
+      "loss": 0.4224,
+      "step": 6798
+    },
+    {
+      "epoch": 0.29858282929185376,
+      "grad_norm": 2.0,
+      "learning_rate": 3.99189947329901e-05,
+      "loss": 0.3931,
+      "step": 6800
+    },
+    {
+      "epoch": 0.29867064777105723,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.991344279300145e-05,
+      "loss": 0.445,
+      "step": 6802
+    },
+    {
+      "epoch": 0.2987584662502607,
+      "grad_norm": 2.390625,
+      "learning_rate": 3.990788971093779e-05,
+      "loss": 0.4219,
+      "step": 6804
+    },
+    {
+      "epoch": 0.2988462847294642,
+      "grad_norm": 2.28125,
+      "learning_rate": 3.9902335487224364e-05,
+      "loss": 0.4343,
+      "step": 6806
+    },
+    {
+      "epoch": 0.2989341032086677,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.989678012228651e-05,
+      "loss": 0.4439,
+      "step": 6808
+    },
+    {
+      "epoch": 0.2990219216878712,
+      "grad_norm": 2.421875,
+      "learning_rate": 3.989122361654967e-05,
+      "loss": 0.452,
+      "step": 6810
+    },
+    {
+      "epoch": 0.29910974016707464,
+      "grad_norm": 2.03125,
+      "learning_rate": 3.988566597043935e-05,
+      "loss": 0.4407,
+      "step": 6812
+    },
+    {
+      "epoch": 0.2991975586462781,
+      "grad_norm": 2.484375,
+      "learning_rate": 3.988010718438115e-05,
+      "loss": 0.4095,
+      "step": 6814
+    },
+    {
+      "epoch": 0.29928537712548164,
+      "grad_norm": 2.6875,
+      "learning_rate": 3.987454725880079e-05,
+      "loss": 0.4345,
+      "step": 6816
+    },
+    {
+      "epoch": 0.2993731956046851,
+      "grad_norm": 2.0,
+      "learning_rate": 3.986898619412402e-05,
+      "loss": 0.4371,
+      "step": 6818
+    },
+    {
+      "epoch": 0.2994610140838886,
+      "grad_norm": 2.203125,
+      "learning_rate": 3.986342399077674e-05,
+      "loss": 0.4375,
+      "step": 6820
+    },
+    {
+      "epoch": 0.2995488325630921,
+      "grad_norm": 1.9765625,
+      "learning_rate": 3.985786064918489e-05,
+      "loss": 0.4311,
+      "step": 6822
+    },
+    {
+      "epoch": 0.2996366510422956,
+      "grad_norm": 2.359375,
+      "learning_rate": 3.9852296169774493e-05,
+      "loss": 0.4073,
+      "step": 6824
+    },
+    {
+      "epoch": 0.29972446952149906,
+      "grad_norm": 2.171875,
+      "learning_rate": 3.9846730552971705e-05,
+      "loss": 0.4334,
+      "step": 6826
+    },
+    {
+      "epoch": 0.29981228800070253,
+      "grad_norm": 2.328125,
+      "learning_rate": 3.984116379920273e-05,
+      "loss": 0.4197,
+      "step": 6828
+    },
+    {
+      "epoch": 0.29990010647990606,
+      "grad_norm": 2.375,
+      "learning_rate": 3.983559590889387e-05,
+      "loss": 0.453,
+      "step": 6830
+    },
+    {
+      "epoch": 0.29998792495910953,
+      "grad_norm": 2.6875,
+      "learning_rate": 3.9830026882471526e-05,
+      "loss": 0.4483,
+      "step": 6832
+    },
+    {
+      "epoch": 0.300075743438313,
+      "grad_norm": 2.828125,
+      "learning_rate": 3.982445672036216e-05,
+      "loss": 0.4343,
+      "step": 6834
+    },
+    {
+      "epoch": 0.3001635619175165,
+      "grad_norm": 2.5625,
+      "learning_rate": 3.9818885422992355e-05,
+      "loss": 0.4299,
+      "step": 6836
+    },
+    {
+      "epoch": 0.30025138039672,
+      "grad_norm": 2.34375,
+      "learning_rate": 3.9813312990788757e-05,
+      "loss": 0.4304,
+      "step": 6838
+    },
+    {
+      "epoch": 0.3003391988759235,
+      "grad_norm": 2.5,
+      "learning_rate": 3.980773942417808e-05,
+      "loss": 0.4219,
+      "step": 6840
+    },
+    {
+      "epoch": 0.30042701735512695,
+      "grad_norm": 2.265625,
+      "learning_rate": 3.980216472358718e-05,
+      "loss": 0.4112,
+      "step": 6842
+    },
+    {
+      "epoch": 0.3005148358343304,
+      "grad_norm": 1.921875,
+      "learning_rate": 3.979658888944296e-05,
+      "loss": 0.4221,
+      "step": 6844
+    },
+    {
+      "epoch": 0.30060265431353395,
+      "grad_norm": 2.0,
+      "learning_rate": 3.97910119221724e-05,
+      "loss": 0.4513,
+      "step": 6846
+    },
+    {
+      "epoch": 0.3006904727927374,
+      "grad_norm": 2.34375,
+      "learning_rate": 3.9785433822202614e-05,
+      "loss": 0.4373,
+      "step": 6848
+    },
+    {
+      "epoch": 0.3007782912719409,
+      "grad_norm": 2.25,
+      "learning_rate": 3.977985458996076e-05,
+      "loss": 0.43,
+      "step": 6850
+    },
+    {
+      "epoch": 0.30086610975114436,
+      "grad_norm": 1.9765625,
+      "learning_rate": 3.97742742258741e-05,
+      "loss": 0.3945,
+      "step": 6852
+    },
+    {
+      "epoch": 0.3009539282303479,
+      "grad_norm": 2.203125,
+      "learning_rate": 3.976869273036997e-05,
+      "loss": 0.4025,
+      "step": 6854
+    },
+    {
+      "epoch": 0.30104174670955136,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.9763110103875824e-05,
+      "loss": 0.4207,
+      "step": 6856
+    },
+    {
+      "epoch": 0.30112956518875483,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.975752634681915e-05,
+      "loss": 0.4163,
+      "step": 6858
+    },
+    {
+      "epoch": 0.3012173836679583,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.975194145962758e-05,
+      "loss": 0.4299,
+      "step": 6860
+    },
+    {
+      "epoch": 0.30130520214716183,
+      "grad_norm": 1.8984375,
+      "learning_rate": 3.97463554427288e-05,
+      "loss": 0.434,
+      "step": 6862
+    },
+    {
+      "epoch": 0.3013930206263653,
+      "grad_norm": 2.40625,
+      "learning_rate": 3.974076829655058e-05,
+      "loss": 0.4472,
+      "step": 6864
+    },
+    {
+      "epoch": 0.3014808391055688,
+      "grad_norm": 2.28125,
+      "learning_rate": 3.97351800215208e-05,
+      "loss": 0.417,
+      "step": 6866
+    },
+    {
+      "epoch": 0.30156865758477225,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.972959061806739e-05,
+      "loss": 0.425,
+      "step": 6868
+    },
+    {
+      "epoch": 0.3016564760639758,
+      "grad_norm": 2.125,
+      "learning_rate": 3.972400008661841e-05,
+      "loss": 0.4394,
+      "step": 6870
+    },
+    {
+      "epoch": 0.30174429454317925,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.971840842760196e-05,
+      "loss": 0.4763,
+      "step": 6872
+    },
+    {
+      "epoch": 0.3018321130223827,
+      "grad_norm": 2.3125,
+      "learning_rate": 3.971281564144628e-05,
+      "loss": 0.4281,
+      "step": 6874
+    },
+    {
+      "epoch": 0.30191993150158625,
+      "grad_norm": 1.953125,
+      "learning_rate": 3.9707221728579634e-05,
+      "loss": 0.459,
+      "step": 6876
+    },
+    {
+      "epoch": 0.3020077499807897,
+      "grad_norm": 2.5625,
+      "learning_rate": 3.970162668943044e-05,
+      "loss": 0.4476,
+      "step": 6878
+    },
+    {
+      "epoch": 0.3020955684599932,
+      "grad_norm": 2.375,
+      "learning_rate": 3.9696030524427144e-05,
+      "loss": 0.4341,
+      "step": 6880
+    },
+    {
+      "epoch": 0.30218338693919666,
+      "grad_norm": 2.25,
+      "learning_rate": 3.96904332339983e-05,
+      "loss": 0.4166,
+      "step": 6882
+    },
+    {
+      "epoch": 0.3022712054184002,
+      "grad_norm": 2.171875,
+      "learning_rate": 3.968483481857256e-05,
+      "loss": 0.4051,
+      "step": 6884
+    },
+    {
+      "epoch": 0.30235902389760366,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.9679235278578654e-05,
+      "loss": 0.4281,
+      "step": 6886
+    },
+    {
+      "epoch": 0.30244684237680713,
+      "grad_norm": 2.078125,
+      "learning_rate": 3.967363461444539e-05,
+      "loss": 0.431,
+      "step": 6888
+    },
+    {
+      "epoch": 0.3025346608560106,
+      "grad_norm": 2.03125,
+      "learning_rate": 3.966803282660167e-05,
+      "loss": 0.4171,
+      "step": 6890
+    },
+    {
+      "epoch": 0.30262247933521413,
+      "grad_norm": 2.171875,
+      "learning_rate": 3.9662429915476476e-05,
+      "loss": 0.4436,
+      "step": 6892
+    },
+    {
+      "epoch": 0.3027102978144176,
+      "grad_norm": 2.34375,
+      "learning_rate": 3.9656825881498885e-05,
+      "loss": 0.4615,
+      "step": 6894
+    },
+    {
+      "epoch": 0.3027981162936211,
+      "grad_norm": 1.984375,
+      "learning_rate": 3.965122072509806e-05,
+      "loss": 0.446,
+      "step": 6896
+    },
+    {
+      "epoch": 0.30288593477282455,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.964561444670324e-05,
+      "loss": 0.4475,
+      "step": 6898
+    },
+    {
+      "epoch": 0.3029737532520281,
+      "grad_norm": 2.34375,
+      "learning_rate": 3.9640007046743756e-05,
+      "loss": 0.3973,
+      "step": 6900
+    },
+    {
+      "epoch": 0.30306157173123155,
+      "grad_norm": 2.203125,
+      "learning_rate": 3.963439852564901e-05,
+      "loss": 0.427,
+      "step": 6902
+    },
+    {
+      "epoch": 0.303149390210435,
+      "grad_norm": 2.171875,
+      "learning_rate": 3.962878888384853e-05,
+      "loss": 0.427,
+      "step": 6904
+    },
+    {
+      "epoch": 0.3032372086896385,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.9623178121771886e-05,
+      "loss": 0.4341,
+      "step": 6906
+    },
+    {
+      "epoch": 0.303325027168842,
+      "grad_norm": 2.40625,
+      "learning_rate": 3.9617566239848755e-05,
+      "loss": 0.4268,
+      "step": 6908
+    },
+    {
+      "epoch": 0.3034128456480455,
+      "grad_norm": 2.4375,
+      "learning_rate": 3.961195323850889e-05,
+      "loss": 0.4002,
+      "step": 6910
+    },
+    {
+      "epoch": 0.30350066412724896,
+      "grad_norm": 3.109375,
+      "learning_rate": 3.960633911818216e-05,
+      "loss": 0.4214,
+      "step": 6912
+    },
+    {
+      "epoch": 0.30358848260645244,
+      "grad_norm": 2.65625,
+      "learning_rate": 3.960072387929847e-05,
+      "loss": 0.4517,
+      "step": 6914
+    },
+    {
+      "epoch": 0.30367630108565596,
+      "grad_norm": 2.34375,
+      "learning_rate": 3.959510752228784e-05,
+      "loss": 0.4415,
+      "step": 6916
+    },
+    {
+      "epoch": 0.30376411956485944,
+      "grad_norm": 2.375,
+      "learning_rate": 3.958949004758039e-05,
+      "loss": 0.4185,
+      "step": 6918
+    },
+    {
+      "epoch": 0.3038519380440629,
+      "grad_norm": 2.90625,
+      "learning_rate": 3.958387145560628e-05,
+      "loss": 0.431,
+      "step": 6920
+    },
+    {
+      "epoch": 0.30393975652326644,
+      "grad_norm": 2.296875,
+      "learning_rate": 3.957825174679581e-05,
+      "loss": 0.4448,
+      "step": 6922
+    },
+    {
+      "epoch": 0.3040275750024699,
+      "grad_norm": 2.296875,
+      "learning_rate": 3.9572630921579324e-05,
+      "loss": 0.4327,
+      "step": 6924
+    },
+    {
+      "epoch": 0.3041153934816734,
+      "grad_norm": 2.34375,
+      "learning_rate": 3.9567008980387264e-05,
+      "loss": 0.4188,
+      "step": 6926
+    },
+    {
+      "epoch": 0.30420321196087685,
+      "grad_norm": 2.875,
+      "learning_rate": 3.956138592365017e-05,
+      "loss": 0.4284,
+      "step": 6928
+    },
+    {
+      "epoch": 0.3042910304400804,
+      "grad_norm": 3.46875,
+      "learning_rate": 3.9555761751798646e-05,
+      "loss": 0.4069,
+      "step": 6930
+    },
+    {
+      "epoch": 0.30437884891928385,
+      "grad_norm": 2.453125,
+      "learning_rate": 3.95501364652634e-05,
+      "loss": 0.4025,
+      "step": 6932
+    },
+    {
+      "epoch": 0.3044666673984873,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.9544510064475214e-05,
+      "loss": 0.4497,
+      "step": 6934
+    },
+    {
+      "epoch": 0.3045544858776908,
+      "grad_norm": 1.921875,
+      "learning_rate": 3.953888254986496e-05,
+      "loss": 0.4263,
+      "step": 6936
+    },
+    {
+      "epoch": 0.3046423043568943,
+      "grad_norm": 2.375,
+      "learning_rate": 3.95332539218636e-05,
+      "loss": 0.4268,
+      "step": 6938
+    },
+    {
+      "epoch": 0.3047301228360978,
+      "grad_norm": 2.65625,
+      "learning_rate": 3.952762418090217e-05,
+      "loss": 0.4395,
+      "step": 6940
+    },
+    {
+      "epoch": 0.30481794131530127,
+      "grad_norm": 2.3125,
+      "learning_rate": 3.9521993327411797e-05,
+      "loss": 0.4417,
+      "step": 6942
+    },
+    {
+      "epoch": 0.30490575979450474,
+      "grad_norm": 2.03125,
+      "learning_rate": 3.9516361361823696e-05,
+      "loss": 0.4297,
+      "step": 6944
+    },
+    {
+      "epoch": 0.30499357827370827,
+      "grad_norm": 2.28125,
+      "learning_rate": 3.951072828456916e-05,
+      "loss": 0.4091,
+      "step": 6946
+    },
+    {
+      "epoch": 0.30508139675291174,
+      "grad_norm": 2.40625,
+      "learning_rate": 3.9505094096079577e-05,
+      "loss": 0.4255,
+      "step": 6948
+    },
+    {
+      "epoch": 0.3051692152321152,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.9499458796786406e-05,
+      "loss": 0.433,
+      "step": 6950
+    },
+    {
+      "epoch": 0.3052570337113187,
+      "grad_norm": 2.359375,
+      "learning_rate": 3.94938223871212e-05,
+      "loss": 0.4533,
+      "step": 6952
+    },
+    {
+      "epoch": 0.3053448521905222,
+      "grad_norm": 2.4375,
+      "learning_rate": 3.948818486751561e-05,
+      "loss": 0.4138,
+      "step": 6954
+    },
+    {
+      "epoch": 0.3054326706697257,
+      "grad_norm": 2.203125,
+      "learning_rate": 3.948254623840134e-05,
+      "loss": 0.4251,
+      "step": 6956
+    },
+    {
+      "epoch": 0.30552048914892915,
+      "grad_norm": 2.390625,
+      "learning_rate": 3.947690650021022e-05,
+      "loss": 0.4489,
+      "step": 6958
+    },
+    {
+      "epoch": 0.3056083076281326,
+      "grad_norm": 2.28125,
+      "learning_rate": 3.947126565337412e-05,
+      "loss": 0.4479,
+      "step": 6960
+    },
+    {
+      "epoch": 0.30569612610733615,
+      "grad_norm": 1.890625,
+      "learning_rate": 3.946562369832503e-05,
+      "loss": 0.4553,
+      "step": 6962
+    },
+    {
+      "epoch": 0.3057839445865396,
+      "grad_norm": 1.96875,
+      "learning_rate": 3.945998063549501e-05,
+      "loss": 0.4209,
+      "step": 6964
+    },
+    {
+      "epoch": 0.3058717630657431,
+      "grad_norm": 1.9453125,
+      "learning_rate": 3.945433646531621e-05,
+      "loss": 0.3993,
+      "step": 6966
+    },
+    {
+      "epoch": 0.30595958154494657,
+      "grad_norm": 2.125,
+      "learning_rate": 3.9448691188220854e-05,
+      "loss": 0.4497,
+      "step": 6968
+    },
+    {
+      "epoch": 0.3060474000241501,
+      "grad_norm": 2.171875,
+      "learning_rate": 3.944304480464126e-05,
+      "loss": 0.4177,
+      "step": 6970
+    },
+    {
+      "epoch": 0.30613521850335357,
+      "grad_norm": 2.140625,
+      "learning_rate": 3.943739731500984e-05,
+      "loss": 0.4476,
+      "step": 6972
+    },
+    {
+      "epoch": 0.30622303698255704,
+      "grad_norm": 1.96875,
+      "learning_rate": 3.943174871975907e-05,
+      "loss": 0.4581,
+      "step": 6974
+    },
+    {
+      "epoch": 0.30631085546176057,
+      "grad_norm": 1.96875,
+      "learning_rate": 3.942609901932153e-05,
+      "loss": 0.448,
+      "step": 6976
+    },
+    {
+      "epoch": 0.30639867394096404,
+      "grad_norm": 2.109375,
+      "learning_rate": 3.942044821412986e-05,
+      "loss": 0.4786,
+      "step": 6978
+    },
+    {
+      "epoch": 0.3064864924201675,
+      "grad_norm": 2.3125,
+      "learning_rate": 3.941479630461681e-05,
+      "loss": 0.4589,
+      "step": 6980
+    },
+    {
+      "epoch": 0.306574310899371,
+      "grad_norm": 2.4375,
+      "learning_rate": 3.94091432912152e-05,
+      "loss": 0.4237,
+      "step": 6982
+    },
+    {
+      "epoch": 0.3066621293785745,
+      "grad_norm": 2.578125,
+      "learning_rate": 3.940348917435796e-05,
+      "loss": 0.4291,
+      "step": 6984
+    },
+    {
+      "epoch": 0.306749947857778,
+      "grad_norm": 2.296875,
+      "learning_rate": 3.939783395447805e-05,
+      "loss": 0.4183,
+      "step": 6986
+    },
+    {
+      "epoch": 0.30683776633698145,
+      "grad_norm": 2.25,
+      "learning_rate": 3.939217763200857e-05,
+      "loss": 0.4078,
+      "step": 6988
+    },
+    {
+      "epoch": 0.3069255848161849,
+      "grad_norm": 2.328125,
+      "learning_rate": 3.938652020738267e-05,
+      "loss": 0.4128,
+      "step": 6990
+    },
+    {
+      "epoch": 0.30701340329538845,
+      "grad_norm": 2.0,
+      "learning_rate": 3.9380861681033606e-05,
+      "loss": 0.4092,
+      "step": 6992
+    },
+    {
+      "epoch": 0.3071012217745919,
+      "grad_norm": 2.25,
+      "learning_rate": 3.937520205339471e-05,
+      "loss": 0.4189,
+      "step": 6994
+    },
+    {
+      "epoch": 0.3071890402537954,
+      "grad_norm": 2.265625,
+      "learning_rate": 3.93695413248994e-05,
+      "loss": 0.4114,
+      "step": 6996
+    },
+    {
+      "epoch": 0.30727685873299887,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.9363879495981166e-05,
+      "loss": 0.4252,
+      "step": 6998
+    },
+    {
+      "epoch": 0.3073646772122024,
+      "grad_norm": 2.296875,
+      "learning_rate": 3.935821656707359e-05,
+      "loss": 0.4006,
+      "step": 7000
+    },
+    {
+      "epoch": 0.30745249569140587,
+      "grad_norm": 2.203125,
+      "learning_rate": 3.935255253861036e-05,
+      "loss": 0.434,
+      "step": 7002
+    },
+    {
+      "epoch": 0.30754031417060934,
+      "grad_norm": 1.9375,
+      "learning_rate": 3.934688741102521e-05,
+      "loss": 0.4251,
+      "step": 7004
+    },
+    {
+      "epoch": 0.3076281326498128,
+      "grad_norm": 2.421875,
+      "learning_rate": 3.934122118475197e-05,
+      "loss": 0.4504,
+      "step": 7006
+    },
+    {
+      "epoch": 0.30771595112901634,
+      "grad_norm": 2.640625,
+      "learning_rate": 3.93355538602246e-05,
+      "loss": 0.426,
+      "step": 7008
+    },
+    {
+      "epoch": 0.3078037696082198,
+      "grad_norm": 2.75,
+      "learning_rate": 3.932988543787707e-05,
+      "loss": 0.398,
+      "step": 7010
+    },
+    {
+      "epoch": 0.3078915880874233,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.932421591814347e-05,
+      "loss": 0.4376,
+      "step": 7012
+    },
+    {
+      "epoch": 0.30797940656662676,
+      "grad_norm": 2.625,
+      "learning_rate": 3.9318545301457985e-05,
+      "loss": 0.4604,
+      "step": 7014
+    },
+    {
+      "epoch": 0.3080672250458303,
+      "grad_norm": 1.9765625,
+      "learning_rate": 3.931287358825486e-05,
+      "loss": 0.4587,
+      "step": 7016
+    },
+    {
+      "epoch": 0.30815504352503376,
+      "grad_norm": 2.453125,
+      "learning_rate": 3.930720077896846e-05,
+      "loss": 0.4166,
+      "step": 7018
+    },
+    {
+      "epoch": 0.30824286200423723,
+      "grad_norm": 2.140625,
+      "learning_rate": 3.930152687403319e-05,
+      "loss": 0.4667,
+      "step": 7020
+    },
+    {
+      "epoch": 0.30833068048344076,
+      "grad_norm": 2.125,
+      "learning_rate": 3.929585187388356e-05,
+      "loss": 0.4017,
+      "step": 7022
+    },
+    {
+      "epoch": 0.3084184989626442,
+      "grad_norm": 2.09375,
+      "learning_rate": 3.929017577895416e-05,
+      "loss": 0.426,
+      "step": 7024
+    },
+    {
+      "epoch": 0.3085063174418477,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.928449858967969e-05,
+      "loss": 0.4435,
+      "step": 7026
+    },
+    {
+      "epoch": 0.30859413592105117,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.9278820306494876e-05,
+      "loss": 0.4275,
+      "step": 7028
+    },
+    {
+      "epoch": 0.3086819544002547,
+      "grad_norm": 2.109375,
+      "learning_rate": 3.92731409298346e-05,
+      "loss": 0.4406,
+      "step": 7030
+    },
+    {
+      "epoch": 0.30876977287945817,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.9267460460133756e-05,
+      "loss": 0.4393,
+      "step": 7032
+    },
+    {
+      "epoch": 0.30885759135866164,
+      "grad_norm": 2.328125,
+      "learning_rate": 3.9261778897827376e-05,
+      "loss": 0.443,
+      "step": 7034
+    },
+    {
+      "epoch": 0.3089454098378651,
+      "grad_norm": 2.0,
+      "learning_rate": 3.925609624335054e-05,
+      "loss": 0.4156,
+      "step": 7036
+    },
+    {
+      "epoch": 0.30903322831706864,
+      "grad_norm": 2.203125,
+      "learning_rate": 3.925041249713844e-05,
+      "loss": 0.4142,
+      "step": 7038
+    },
+    {
+      "epoch": 0.3091210467962721,
+      "grad_norm": 2.046875,
+      "learning_rate": 3.9244727659626346e-05,
+      "loss": 0.421,
+      "step": 7040
+    },
+    {
+      "epoch": 0.3092088652754756,
+      "grad_norm": 2.265625,
+      "learning_rate": 3.923904173124958e-05,
+      "loss": 0.4122,
+      "step": 7042
+    },
+    {
+      "epoch": 0.30929668375467906,
+      "grad_norm": 2.390625,
+      "learning_rate": 3.923335471244359e-05,
+      "loss": 0.4319,
+      "step": 7044
+    },
+    {
+      "epoch": 0.3093845022338826,
+      "grad_norm": 1.9609375,
+      "learning_rate": 3.922766660364388e-05,
+      "loss": 0.4363,
+      "step": 7046
+    },
+    {
+      "epoch": 0.30947232071308606,
+      "grad_norm": 2.375,
+      "learning_rate": 3.9221977405286057e-05,
+      "loss": 0.4407,
+      "step": 7048
+    },
+    {
+      "epoch": 0.30956013919228953,
+      "grad_norm": 2.078125,
+      "learning_rate": 3.9216287117805787e-05,
+      "loss": 0.4583,
+      "step": 7050
+    },
+    {
+      "epoch": 0.309647957671493,
+      "grad_norm": 2.296875,
+      "learning_rate": 3.9210595741638853e-05,
+      "loss": 0.4364,
+      "step": 7052
+    },
+    {
+      "epoch": 0.30973577615069653,
+      "grad_norm": 1.9140625,
+      "learning_rate": 3.9204903277221086e-05,
+      "loss": 0.4248,
+      "step": 7054
+    },
+    {
+      "epoch": 0.3098235946299,
+      "grad_norm": 1.953125,
+      "learning_rate": 3.9199209724988415e-05,
+      "loss": 0.4312,
+      "step": 7056
+    },
+    {
+      "epoch": 0.3099114131091035,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.9193515085376867e-05,
+      "loss": 0.43,
+      "step": 7058
+    },
+    {
+      "epoch": 0.30999923158830694,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.918781935882253e-05,
+      "loss": 0.4222,
+      "step": 7060
+    },
+    {
+      "epoch": 0.3100870500675105,
+      "grad_norm": 2.515625,
+      "learning_rate": 3.918212254576158e-05,
+      "loss": 0.4161,
+      "step": 7062
+    },
+    {
+      "epoch": 0.31017486854671394,
+      "grad_norm": 2.03125,
+      "learning_rate": 3.91764246466303e-05,
+      "loss": 0.4363,
+      "step": 7064
+    },
+    {
+      "epoch": 0.3102626870259174,
+      "grad_norm": 2.421875,
+      "learning_rate": 3.9170725661865e-05,
+      "loss": 0.3983,
+      "step": 7066
+    },
+    {
+      "epoch": 0.3103505055051209,
+      "grad_norm": 2.171875,
+      "learning_rate": 3.916502559190215e-05,
+      "loss": 0.4453,
+      "step": 7068
+    },
+    {
+      "epoch": 0.3104383239843244,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.9159324437178236e-05,
+      "loss": 0.4246,
+      "step": 7070
+    },
+    {
+      "epoch": 0.3105261424635279,
+      "grad_norm": 2.0625,
+      "learning_rate": 3.915362219812986e-05,
+      "loss": 0.4217,
+      "step": 7072
+    },
+    {
+      "epoch": 0.31061396094273136,
+      "grad_norm": 2.296875,
+      "learning_rate": 3.914791887519371e-05,
+      "loss": 0.4183,
+      "step": 7074
+    },
+    {
+      "epoch": 0.3107017794219349,
+      "grad_norm": 2.25,
+      "learning_rate": 3.914221446880654e-05,
+      "loss": 0.4117,
+      "step": 7076
+    },
+    {
+      "epoch": 0.31078959790113836,
+      "grad_norm": 2.328125,
+      "learning_rate": 3.91365089794052e-05,
+      "loss": 0.4156,
+      "step": 7078
+    },
+    {
+      "epoch": 0.31087741638034183,
+      "grad_norm": 2.375,
+      "learning_rate": 3.91308024074266e-05,
+      "loss": 0.4281,
+      "step": 7080
+    },
+    {
+      "epoch": 0.3109652348595453,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.9125094753307775e-05,
+      "loss": 0.428,
+      "step": 7082
+    },
+    {
+      "epoch": 0.31105305333874883,
+      "grad_norm": 2.296875,
+      "learning_rate": 3.91193860174858e-05,
+      "loss": 0.4356,
+      "step": 7084
+    },
+    {
+      "epoch": 0.3111408718179523,
+      "grad_norm": 2.375,
+      "learning_rate": 3.911367620039787e-05,
+      "loss": 0.4351,
+      "step": 7086
+    },
+    {
+      "epoch": 0.3112286902971558,
+      "grad_norm": 2.140625,
+      "learning_rate": 3.910796530248123e-05,
+      "loss": 0.455,
+      "step": 7088
+    },
+    {
+      "epoch": 0.31131650877635925,
+      "grad_norm": 2.703125,
+      "learning_rate": 3.910225332417322e-05,
+      "loss": 0.3949,
+      "step": 7090
+    },
+    {
+      "epoch": 0.3114043272555628,
+      "grad_norm": 2.546875,
+      "learning_rate": 3.909654026591127e-05,
+      "loss": 0.4086,
+      "step": 7092
+    },
+    {
+      "epoch": 0.31149214573476625,
+      "grad_norm": 2.265625,
+      "learning_rate": 3.9090826128132896e-05,
+      "loss": 0.4271,
+      "step": 7094
+    },
+    {
+      "epoch": 0.3115799642139697,
+      "grad_norm": 2.421875,
+      "learning_rate": 3.908511091127567e-05,
+      "loss": 0.4195,
+      "step": 7096
+    },
+    {
+      "epoch": 0.3116677826931732,
+      "grad_norm": 2.328125,
+      "learning_rate": 3.907939461577727e-05,
+      "loss": 0.4229,
+      "step": 7098
+    },
+    {
+      "epoch": 0.3117556011723767,
+      "grad_norm": 2.0625,
+      "learning_rate": 3.9073677242075466e-05,
+      "loss": 0.4651,
+      "step": 7100
+    },
+    {
+      "epoch": 0.3118434196515802,
+      "grad_norm": 2.296875,
+      "learning_rate": 3.906795879060809e-05,
+      "loss": 0.4029,
+      "step": 7102
+    },
+    {
+      "epoch": 0.31193123813078366,
+      "grad_norm": 2.140625,
+      "learning_rate": 3.906223926181305e-05,
+      "loss": 0.4254,
+      "step": 7104
+    },
+    {
+      "epoch": 0.31201905660998713,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.905651865612835e-05,
+      "loss": 0.4422,
+      "step": 7106
+    },
+    {
+      "epoch": 0.31210687508919066,
+      "grad_norm": 2.03125,
+      "learning_rate": 3.9050796973992084e-05,
+      "loss": 0.432,
+      "step": 7108
+    },
+    {
+      "epoch": 0.31219469356839413,
+      "grad_norm": 2.234375,
+      "learning_rate": 3.9045074215842425e-05,
+      "loss": 0.4265,
+      "step": 7110
+    },
+    {
+      "epoch": 0.3122825120475976,
+      "grad_norm": 2.328125,
+      "learning_rate": 3.903935038211761e-05,
+      "loss": 0.4098,
+      "step": 7112
+    },
+    {
+      "epoch": 0.3123703305268011,
+      "grad_norm": 1.8828125,
+      "learning_rate": 3.903362547325598e-05,
+      "loss": 0.407,
+      "step": 7114
+    },
+    {
+      "epoch": 0.3124581490060046,
+      "grad_norm": 1.8984375,
+      "learning_rate": 3.9027899489695954e-05,
+      "loss": 0.4245,
+      "step": 7116
+    },
+    {
+      "epoch": 0.3125459674852081,
+      "grad_norm": 2.328125,
+      "learning_rate": 3.902217243187601e-05,
+      "loss": 0.4124,
+      "step": 7118
+    },
+    {
+      "epoch": 0.31263378596441155,
+      "grad_norm": 2.4375,
+      "learning_rate": 3.9016444300234747e-05,
+      "loss": 0.4477,
+      "step": 7120
+    },
+    {
+      "epoch": 0.312721604443615,
+      "grad_norm": 2.015625,
+      "learning_rate": 3.9010715095210816e-05,
+      "loss": 0.403,
+      "step": 7122
+    },
+    {
+      "epoch": 0.31280942292281855,
+      "grad_norm": 2.09375,
+      "learning_rate": 3.900498481724296e-05,
+      "loss": 0.4407,
+      "step": 7124
+    },
+    {
+      "epoch": 0.312897241402022,
+      "grad_norm": 2.171875,
+      "learning_rate": 3.899925346677002e-05,
+      "loss": 0.4491,
+      "step": 7126
+    },
+    {
+      "epoch": 0.3129850598812255,
+      "grad_norm": 2.015625,
+      "learning_rate": 3.8993521044230884e-05,
+      "loss": 0.4311,
+      "step": 7128
+    },
+    {
+      "epoch": 0.313072878360429,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.8987787550064555e-05,
+      "loss": 0.4356,
+      "step": 7130
+    },
+    {
+      "epoch": 0.3131606968396325,
+      "grad_norm": 2.546875,
+      "learning_rate": 3.8982052984710105e-05,
+      "loss": 0.4152,
+      "step": 7132
+    },
+    {
+      "epoch": 0.31324851531883596,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.8976317348606684e-05,
+      "loss": 0.411,
+      "step": 7134
+    },
+    {
+      "epoch": 0.31333633379803943,
+      "grad_norm": 2.40625,
+      "learning_rate": 3.8970580642193534e-05,
+      "loss": 0.4311,
+      "step": 7136
+    },
+    {
+      "epoch": 0.31342415227724296,
+      "grad_norm": 2.34375,
+      "learning_rate": 3.896484286590997e-05,
+      "loss": 0.4259,
+      "step": 7138
+    },
+    {
+      "epoch": 0.31351197075644643,
+      "grad_norm": 2.078125,
+      "learning_rate": 3.895910402019538e-05,
+      "loss": 0.4198,
+      "step": 7140
+    },
+    {
+      "epoch": 0.3135997892356499,
+      "grad_norm": 2.0625,
+      "learning_rate": 3.895336410548927e-05,
+      "loss": 0.4099,
+      "step": 7142
+    },
+    {
+      "epoch": 0.3136876077148534,
+      "grad_norm": 2.140625,
+      "learning_rate": 3.894762312223118e-05,
+      "loss": 0.4274,
+      "step": 7144
+    },
+    {
+      "epoch": 0.3137754261940569,
+      "grad_norm": 1.9921875,
+      "learning_rate": 3.894188107086078e-05,
+      "loss": 0.436,
+      "step": 7146
+    },
+    {
+      "epoch": 0.3138632446732604,
+      "grad_norm": 2.046875,
+      "learning_rate": 3.893613795181778e-05,
+      "loss": 0.437,
+      "step": 7148
+    },
+    {
+      "epoch": 0.31395106315246385,
+      "grad_norm": 2.328125,
+      "learning_rate": 3.8930393765542e-05,
+      "loss": 0.4435,
+      "step": 7150
+    },
+    {
+      "epoch": 0.3140388816316673,
+      "grad_norm": 2.65625,
+      "learning_rate": 3.892464851247332e-05,
+      "loss": 0.412,
+      "step": 7152
+    },
+    {
+      "epoch": 0.31412670011087085,
+      "grad_norm": 2.453125,
+      "learning_rate": 3.891890219305172e-05,
+      "loss": 0.4144,
+      "step": 7154
+    },
+    {
+      "epoch": 0.3142145185900743,
+      "grad_norm": 2.046875,
+      "learning_rate": 3.8913154807717255e-05,
+      "loss": 0.4099,
+      "step": 7156
+    },
+    {
+      "epoch": 0.3143023370692778,
+      "grad_norm": 1.9609375,
+      "learning_rate": 3.8907406356910054e-05,
+      "loss": 0.4196,
+      "step": 7158
+    },
+    {
+      "epoch": 0.31439015554848126,
+      "grad_norm": 2.078125,
+      "learning_rate": 3.890165684107035e-05,
+      "loss": 0.4157,
+      "step": 7160
+    },
+    {
+      "epoch": 0.3144779740276848,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.8895906260638426e-05,
+      "loss": 0.4065,
+      "step": 7162
+    },
+    {
+      "epoch": 0.31456579250688826,
+      "grad_norm": 2.328125,
+      "learning_rate": 3.8890154616054676e-05,
+      "loss": 0.4396,
+      "step": 7164
+    },
+    {
+      "epoch": 0.31465361098609174,
+      "grad_norm": 1.9921875,
+      "learning_rate": 3.888440190775955e-05,
+      "loss": 0.4087,
+      "step": 7166
+    },
+    {
+      "epoch": 0.3147414294652952,
+      "grad_norm": 1.859375,
+      "learning_rate": 3.8878648136193596e-05,
+      "loss": 0.4512,
+      "step": 7168
+    },
+    {
+      "epoch": 0.31482924794449874,
+      "grad_norm": 1.921875,
+      "learning_rate": 3.887289330179744e-05,
+      "loss": 0.4095,
+      "step": 7170
+    },
+    {
+      "epoch": 0.3149170664237022,
+      "grad_norm": 1.96875,
+      "learning_rate": 3.8867137405011786e-05,
+      "loss": 0.4463,
+      "step": 7172
+    },
+    {
+      "epoch": 0.3150048849029057,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.886138044627744e-05,
+      "loss": 0.4024,
+      "step": 7174
+    },
+    {
+      "epoch": 0.3150927033821092,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.885562242603525e-05,
+      "loss": 0.4472,
+      "step": 7176
+    },
+    {
+      "epoch": 0.3151805218613127,
+      "grad_norm": 1.9765625,
+      "learning_rate": 3.8849863344726175e-05,
+      "loss": 0.4177,
+      "step": 7178
+    },
+    {
+      "epoch": 0.31526834034051615,
+      "grad_norm": 2.0625,
+      "learning_rate": 3.884410320279124e-05,
+      "loss": 0.4592,
+      "step": 7180
+    },
+    {
+      "epoch": 0.3153561588197196,
+      "grad_norm": 2.140625,
+      "learning_rate": 3.883834200067157e-05,
+      "loss": 0.4344,
+      "step": 7182
+    },
+    {
+      "epoch": 0.31544397729892315,
+      "grad_norm": 2.0625,
+      "learning_rate": 3.883257973880834e-05,
+      "loss": 0.4343,
+      "step": 7184
+    },
+    {
+      "epoch": 0.3155317957781266,
+      "grad_norm": 1.9453125,
+      "learning_rate": 3.8826816417642845e-05,
+      "loss": 0.4252,
+      "step": 7186
+    },
+    {
+      "epoch": 0.3156196142573301,
+      "grad_norm": 2.03125,
+      "learning_rate": 3.882105203761644e-05,
+      "loss": 0.4246,
+      "step": 7188
+    },
+    {
+      "epoch": 0.31570743273653357,
+      "grad_norm": 2.09375,
+      "learning_rate": 3.8815286599170544e-05,
+      "loss": 0.4142,
+      "step": 7190
+    },
+    {
+      "epoch": 0.3157952512157371,
+      "grad_norm": 2.28125,
+      "learning_rate": 3.88095201027467e-05,
+      "loss": 0.4259,
+      "step": 7192
+    },
+    {
+      "epoch": 0.31588306969494057,
+      "grad_norm": 2.046875,
+      "learning_rate": 3.880375254878649e-05,
+      "loss": 0.4249,
+      "step": 7194
+    },
+    {
+      "epoch": 0.31597088817414404,
+      "grad_norm": 2.5,
+      "learning_rate": 3.87979839377316e-05,
+      "loss": 0.4373,
+      "step": 7196
+    },
+    {
+      "epoch": 0.3160587066533475,
+      "grad_norm": 2.03125,
+      "learning_rate": 3.87922142700238e-05,
+      "loss": 0.4233,
+      "step": 7198
+    },
+    {
+      "epoch": 0.31614652513255104,
+      "grad_norm": 2.359375,
+      "learning_rate": 3.878644354610492e-05,
+      "loss": 0.4137,
+      "step": 7200
+    },
+    {
+      "epoch": 0.3162343436117545,
+      "grad_norm": 2.140625,
+      "learning_rate": 3.8780671766416885e-05,
+      "loss": 0.4349,
+      "step": 7202
+    },
+    {
+      "epoch": 0.316322162090958,
+      "grad_norm": 2.765625,
+      "learning_rate": 3.8774898931401706e-05,
+      "loss": 0.425,
+      "step": 7204
+    },
+    {
+      "epoch": 0.31640998057016145,
+      "grad_norm": 2.328125,
+      "learning_rate": 3.8769125041501466e-05,
+      "loss": 0.466,
+      "step": 7206
+    },
+    {
+      "epoch": 0.316497799049365,
+      "grad_norm": 1.9453125,
+      "learning_rate": 3.876335009715833e-05,
+      "loss": 0.4166,
+      "step": 7208
+    },
+    {
+      "epoch": 0.31658561752856845,
+      "grad_norm": 2.140625,
+      "learning_rate": 3.8757574098814544e-05,
+      "loss": 0.4177,
+      "step": 7210
+    },
+    {
+      "epoch": 0.3166734360077719,
+      "grad_norm": 2.65625,
+      "learning_rate": 3.875179704691243e-05,
+      "loss": 0.4321,
+      "step": 7212
+    },
+    {
+      "epoch": 0.3167612544869754,
+      "grad_norm": 2.46875,
+      "learning_rate": 3.874601894189441e-05,
+      "loss": 0.4159,
+      "step": 7214
+    },
+    {
+      "epoch": 0.3168490729661789,
+      "grad_norm": 2.625,
+      "learning_rate": 3.8740239784202956e-05,
+      "loss": 0.4265,
+      "step": 7216
+    },
+    {
+      "epoch": 0.3169368914453824,
+      "grad_norm": 2.75,
+      "learning_rate": 3.873445957428065e-05,
+      "loss": 0.4415,
+      "step": 7218
+    },
+    {
+      "epoch": 0.31702470992458587,
+      "grad_norm": 2.140625,
+      "learning_rate": 3.872867831257014e-05,
+      "loss": 0.4174,
+      "step": 7220
+    },
+    {
+      "epoch": 0.31711252840378934,
+      "grad_norm": 2.375,
+      "learning_rate": 3.872289599951415e-05,
+      "loss": 0.4364,
+      "step": 7222
+    },
+    {
+      "epoch": 0.31720034688299287,
+      "grad_norm": 2.328125,
+      "learning_rate": 3.8717112635555494e-05,
+      "loss": 0.4321,
+      "step": 7224
+    },
+    {
+      "epoch": 0.31728816536219634,
+      "grad_norm": 2.109375,
+      "learning_rate": 3.8711328221137066e-05,
+      "loss": 0.4507,
+      "step": 7226
+    },
+    {
+      "epoch": 0.3173759838413998,
+      "grad_norm": 2.078125,
+      "learning_rate": 3.870554275670184e-05,
+      "loss": 0.4347,
+      "step": 7228
+    },
+    {
+      "epoch": 0.31746380232060334,
+      "grad_norm": 1.9765625,
+      "learning_rate": 3.8699756242692854e-05,
+      "loss": 0.3905,
+      "step": 7230
+    },
+    {
+      "epoch": 0.3175516207998068,
+      "grad_norm": 2.046875,
+      "learning_rate": 3.869396867955326e-05,
+      "loss": 0.4089,
+      "step": 7232
+    },
+    {
+      "epoch": 0.3176394392790103,
+      "grad_norm": 2.109375,
+      "learning_rate": 3.868818006772626e-05,
+      "loss": 0.4114,
+      "step": 7234
+    },
+    {
+      "epoch": 0.31772725775821375,
+      "grad_norm": 2.203125,
+      "learning_rate": 3.8682390407655145e-05,
+      "loss": 0.39,
+      "step": 7236
+    },
+    {
+      "epoch": 0.3178150762374173,
+      "grad_norm": 2.078125,
+      "learning_rate": 3.86765996997833e-05,
+      "loss": 0.4383,
+      "step": 7238
+    },
+    {
+      "epoch": 0.31790289471662075,
+      "grad_norm": 2.046875,
+      "learning_rate": 3.867080794455416e-05,
+      "loss": 0.415,
+      "step": 7240
+    },
+    {
+      "epoch": 0.3179907131958242,
+      "grad_norm": 2.109375,
+      "learning_rate": 3.866501514241129e-05,
+      "loss": 0.4171,
+      "step": 7242
+    },
+    {
+      "epoch": 0.3180785316750277,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.8659221293798265e-05,
+      "loss": 0.4294,
+      "step": 7244
+    },
+    {
+      "epoch": 0.3181663501542312,
+      "grad_norm": 1.984375,
+      "learning_rate": 3.865342639915881e-05,
+      "loss": 0.4058,
+      "step": 7246
+    },
+    {
+      "epoch": 0.3182541686334347,
+      "grad_norm": 2.078125,
+      "learning_rate": 3.864763045893668e-05,
+      "loss": 0.4064,
+      "step": 7248
+    },
+    {
+      "epoch": 0.31834198711263817,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.8641833473575745e-05,
+      "loss": 0.4323,
+      "step": 7250
+    },
+    {
+      "epoch": 0.31842980559184164,
+      "grad_norm": 2.359375,
+      "learning_rate": 3.863603544351993e-05,
+      "loss": 0.4051,
+      "step": 7252
+    },
+    {
+      "epoch": 0.31851762407104517,
+      "grad_norm": 2.484375,
+      "learning_rate": 3.863023636921326e-05,
+      "loss": 0.4387,
+      "step": 7254
+    },
+    {
+      "epoch": 0.31860544255024864,
+      "grad_norm": 2.421875,
+      "learning_rate": 3.862443625109981e-05,
+      "loss": 0.4106,
+      "step": 7256
+    },
+    {
+      "epoch": 0.3186932610294521,
+      "grad_norm": 2.578125,
+      "learning_rate": 3.861863508962377e-05,
+      "loss": 0.4498,
+      "step": 7258
+    },
+    {
+      "epoch": 0.3187810795086556,
+      "grad_norm": 2.8125,
+      "learning_rate": 3.861283288522939e-05,
+      "loss": 0.4025,
+      "step": 7260
+    },
+    {
+      "epoch": 0.3188688979878591,
+      "grad_norm": 2.390625,
+      "learning_rate": 3.8607029638361005e-05,
+      "loss": 0.4204,
+      "step": 7262
+    },
+    {
+      "epoch": 0.3189567164670626,
+      "grad_norm": 2.34375,
+      "learning_rate": 3.860122534946302e-05,
+      "loss": 0.4304,
+      "step": 7264
+    },
+    {
+      "epoch": 0.31904453494626606,
+      "grad_norm": 1.96875,
+      "learning_rate": 3.859542001897994e-05,
+      "loss": 0.4281,
+      "step": 7266
+    },
+    {
+      "epoch": 0.31913235342546953,
+      "grad_norm": 1.953125,
+      "learning_rate": 3.858961364735635e-05,
+      "loss": 0.4151,
+      "step": 7268
+    },
+    {
+      "epoch": 0.31922017190467306,
+      "grad_norm": 2.0625,
+      "learning_rate": 3.858380623503688e-05,
+      "loss": 0.421,
+      "step": 7270
+    },
+    {
+      "epoch": 0.3193079903838765,
+      "grad_norm": 1.890625,
+      "learning_rate": 3.857799778246627e-05,
+      "loss": 0.429,
+      "step": 7272
+    },
+    {
+      "epoch": 0.31939580886308,
+      "grad_norm": 2.015625,
+      "learning_rate": 3.857218829008934e-05,
+      "loss": 0.442,
+      "step": 7274
+    },
+    {
+      "epoch": 0.3194836273422835,
+      "grad_norm": 2.0625,
+      "learning_rate": 3.856637775835097e-05,
+      "loss": 0.4132,
+      "step": 7276
+    },
+    {
+      "epoch": 0.319571445821487,
+      "grad_norm": 2.203125,
+      "learning_rate": 3.856056618769614e-05,
+      "loss": 0.4476,
+      "step": 7278
+    },
+    {
+      "epoch": 0.31965926430069047,
+      "grad_norm": 1.9765625,
+      "learning_rate": 3.855475357856991e-05,
+      "loss": 0.4238,
+      "step": 7280
+    },
+    {
+      "epoch": 0.31974708277989394,
+      "grad_norm": 2.046875,
+      "learning_rate": 3.854893993141739e-05,
+      "loss": 0.3769,
+      "step": 7282
+    },
+    {
+      "epoch": 0.31983490125909747,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.854312524668381e-05,
+      "loss": 0.4558,
+      "step": 7284
+    },
+    {
+      "epoch": 0.31992271973830094,
+      "grad_norm": 2.359375,
+      "learning_rate": 3.853730952481446e-05,
+      "loss": 0.4316,
+      "step": 7286
+    },
+    {
+      "epoch": 0.3200105382175044,
+      "grad_norm": 2.125,
+      "learning_rate": 3.853149276625468e-05,
+      "loss": 0.4224,
+      "step": 7288
+    },
+    {
+      "epoch": 0.3200983566967079,
+      "grad_norm": 2.296875,
+      "learning_rate": 3.8525674971449956e-05,
+      "loss": 0.4253,
+      "step": 7290
+    },
+    {
+      "epoch": 0.3201861751759114,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.85198561408458e-05,
+      "loss": 0.435,
+      "step": 7292
+    },
+    {
+      "epoch": 0.3202739936551149,
+      "grad_norm": 2.296875,
+      "learning_rate": 3.851403627488781e-05,
+      "loss": 0.4294,
+      "step": 7294
+    },
+    {
+      "epoch": 0.32036181213431836,
+      "grad_norm": 2.640625,
+      "learning_rate": 3.8508215374021695e-05,
+      "loss": 0.4705,
+      "step": 7296
+    },
+    {
+      "epoch": 0.32044963061352183,
+      "grad_norm": 2.140625,
+      "learning_rate": 3.8502393438693203e-05,
+      "loss": 0.4311,
+      "step": 7298
+    },
+    {
+      "epoch": 0.32053744909272536,
+      "grad_norm": 2.109375,
+      "learning_rate": 3.8496570469348184e-05,
+      "loss": 0.436,
+      "step": 7300
+    },
+    {
+      "epoch": 0.32062526757192883,
+      "grad_norm": 2.0625,
+      "learning_rate": 3.8490746466432556e-05,
+      "loss": 0.4277,
+      "step": 7302
+    },
+    {
+      "epoch": 0.3207130860511323,
+      "grad_norm": 2.25,
+      "learning_rate": 3.848492143039234e-05,
+      "loss": 0.4082,
+      "step": 7304
+    },
+    {
+      "epoch": 0.3208009045303358,
+      "grad_norm": 2.0625,
+      "learning_rate": 3.8479095361673604e-05,
+      "loss": 0.4182,
+      "step": 7306
+    },
+    {
+      "epoch": 0.3208887230095393,
+      "grad_norm": 1.9609375,
+      "learning_rate": 3.847326826072252e-05,
+      "loss": 0.4079,
+      "step": 7308
+    },
+    {
+      "epoch": 0.3209765414887428,
+      "grad_norm": 2.046875,
+      "learning_rate": 3.846744012798531e-05,
+      "loss": 0.4158,
+      "step": 7310
+    },
+    {
+      "epoch": 0.32106435996794624,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.8461610963908314e-05,
+      "loss": 0.4115,
+      "step": 7312
+    },
+    {
+      "epoch": 0.3211521784471497,
+      "grad_norm": 1.875,
+      "learning_rate": 3.845578076893793e-05,
+      "loss": 0.4487,
+      "step": 7314
+    },
+    {
+      "epoch": 0.32123999692635324,
+      "grad_norm": 2.171875,
+      "learning_rate": 3.8449949543520625e-05,
+      "loss": 0.4127,
+      "step": 7316
+    },
+    {
+      "epoch": 0.3213278154055567,
+      "grad_norm": 2.125,
+      "learning_rate": 3.8444117288102956e-05,
+      "loss": 0.434,
+      "step": 7318
+    },
+    {
+      "epoch": 0.3214156338847602,
+      "grad_norm": 1.9765625,
+      "learning_rate": 3.8438284003131566e-05,
+      "loss": 0.4443,
+      "step": 7320
+    },
+    {
+      "epoch": 0.32150345236396366,
+      "grad_norm": 2.125,
+      "learning_rate": 3.843244968905316e-05,
+      "loss": 0.4392,
+      "step": 7322
+    },
+    {
+      "epoch": 0.3215912708431672,
+      "grad_norm": 2.078125,
+      "learning_rate": 3.842661434631454e-05,
+      "loss": 0.434,
+      "step": 7324
+    },
+    {
+      "epoch": 0.32167908932237066,
+      "grad_norm": 2.109375,
+      "learning_rate": 3.842077797536258e-05,
+      "loss": 0.4356,
+      "step": 7326
+    },
+    {
+      "epoch": 0.32176690780157413,
+      "grad_norm": 2.5,
+      "learning_rate": 3.8414940576644215e-05,
+      "loss": 0.4315,
+      "step": 7328
+    },
+    {
+      "epoch": 0.32185472628077766,
+      "grad_norm": 2.3125,
+      "learning_rate": 3.840910215060649e-05,
+      "loss": 0.4577,
+      "step": 7330
+    },
+    {
+      "epoch": 0.32194254475998113,
+      "grad_norm": 3.03125,
+      "learning_rate": 3.8403262697696514e-05,
+      "loss": 0.4243,
+      "step": 7332
+    },
+    {
+      "epoch": 0.3220303632391846,
+      "grad_norm": 2.71875,
+      "learning_rate": 3.8397422218361454e-05,
+      "loss": 0.4241,
+      "step": 7334
+    },
+    {
+      "epoch": 0.3221181817183881,
+      "grad_norm": 3.078125,
+      "learning_rate": 3.8391580713048604e-05,
+      "loss": 0.3999,
+      "step": 7336
+    },
+    {
+      "epoch": 0.3222060001975916,
+      "grad_norm": 3.734375,
+      "learning_rate": 3.8385738182205276e-05,
+      "loss": 0.4357,
+      "step": 7338
+    },
+    {
+      "epoch": 0.3222938186767951,
+      "grad_norm": 2.34375,
+      "learning_rate": 3.837989462627893e-05,
+      "loss": 0.4123,
+      "step": 7340
+    },
+    {
+      "epoch": 0.32238163715599855,
+      "grad_norm": 2.09375,
+      "learning_rate": 3.837405004571703e-05,
+      "loss": 0.4154,
+      "step": 7342
+    },
+    {
+      "epoch": 0.322469455635202,
+      "grad_norm": 2.234375,
+      "learning_rate": 3.836820444096718e-05,
+      "loss": 0.3876,
+      "step": 7344
+    },
+    {
+      "epoch": 0.32255727411440555,
+      "grad_norm": 2.671875,
+      "learning_rate": 3.8362357812477025e-05,
+      "loss": 0.4418,
+      "step": 7346
+    },
+    {
+      "epoch": 0.322645092593609,
+      "grad_norm": 2.671875,
+      "learning_rate": 3.8356510160694305e-05,
+      "loss": 0.4353,
+      "step": 7348
+    },
+    {
+      "epoch": 0.3227329110728125,
+      "grad_norm": 2.65625,
+      "learning_rate": 3.835066148606683e-05,
+      "loss": 0.4217,
+      "step": 7350
+    },
+    {
+      "epoch": 0.32282072955201596,
+      "grad_norm": 2.640625,
+      "learning_rate": 3.834481178904251e-05,
+      "loss": 0.4379,
+      "step": 7352
+    },
+    {
+      "epoch": 0.3229085480312195,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.83389610700693e-05,
+      "loss": 0.4549,
+      "step": 7354
+    },
+    {
+      "epoch": 0.32299636651042296,
+      "grad_norm": 2.046875,
+      "learning_rate": 3.833310932959525e-05,
+      "loss": 0.4456,
+      "step": 7356
+    },
+    {
+      "epoch": 0.32308418498962643,
+      "grad_norm": 2.46875,
+      "learning_rate": 3.8327256568068495e-05,
+      "loss": 0.4234,
+      "step": 7358
+    },
+    {
+      "epoch": 0.3231720034688299,
+      "grad_norm": 2.453125,
+      "learning_rate": 3.832140278593724e-05,
+      "loss": 0.42,
+      "step": 7360
+    },
+    {
+      "epoch": 0.32325982194803343,
+      "grad_norm": 2.796875,
+      "learning_rate": 3.8315547983649764e-05,
+      "loss": 0.4267,
+      "step": 7362
+    },
+    {
+      "epoch": 0.3233476404272369,
+      "grad_norm": 2.359375,
+      "learning_rate": 3.830969216165443e-05,
+      "loss": 0.4332,
+      "step": 7364
+    },
+    {
+      "epoch": 0.3234354589064404,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.8303835320399685e-05,
+      "loss": 0.4232,
+      "step": 7366
+    },
+    {
+      "epoch": 0.32352327738564385,
+      "grad_norm": 2.28125,
+      "learning_rate": 3.829797746033404e-05,
+      "loss": 0.4553,
+      "step": 7368
+    },
+    {
+      "epoch": 0.3236110958648474,
+      "grad_norm": 2.4375,
+      "learning_rate": 3.829211858190608e-05,
+      "loss": 0.4008,
+      "step": 7370
+    },
+    {
+      "epoch": 0.32369891434405085,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.8286258685564505e-05,
+      "loss": 0.4607,
+      "step": 7372
+    },
+    {
+      "epoch": 0.3237867328232543,
+      "grad_norm": 2.515625,
+      "learning_rate": 3.828039777175805e-05,
+      "loss": 0.4038,
+      "step": 7374
+    },
+    {
+      "epoch": 0.32387455130245785,
+      "grad_norm": 2.671875,
+      "learning_rate": 3.8274535840935553e-05,
+      "loss": 0.4249,
+      "step": 7376
+    },
+    {
+      "epoch": 0.3239623697816613,
+      "grad_norm": 2.140625,
+      "learning_rate": 3.8268672893545924e-05,
+      "loss": 0.4222,
+      "step": 7378
+    },
+    {
+      "epoch": 0.3240501882608648,
+      "grad_norm": 2.140625,
+      "learning_rate": 3.826280893003814e-05,
+      "loss": 0.4074,
+      "step": 7380
+    },
+    {
+      "epoch": 0.32413800674006826,
+      "grad_norm": 2.03125,
+      "learning_rate": 3.8256943950861264e-05,
+      "loss": 0.4298,
+      "step": 7382
+    },
+    {
+      "epoch": 0.3242258252192718,
+      "grad_norm": 2.203125,
+      "learning_rate": 3.825107795646444e-05,
+      "loss": 0.4323,
+      "step": 7384
+    },
+    {
+      "epoch": 0.32431364369847526,
+      "grad_norm": 2.265625,
+      "learning_rate": 3.82452109472969e-05,
+      "loss": 0.4272,
+      "step": 7386
+    },
+    {
+      "epoch": 0.32440146217767873,
+      "grad_norm": 2.125,
+      "learning_rate": 3.823934292380793e-05,
+      "loss": 0.4031,
+      "step": 7388
+    },
+    {
+      "epoch": 0.3244892806568822,
+      "grad_norm": 2.203125,
+      "learning_rate": 3.82334738864469e-05,
+      "loss": 0.4268,
+      "step": 7390
+    },
+    {
+      "epoch": 0.32457709913608573,
+      "grad_norm": 2.203125,
+      "learning_rate": 3.822760383566327e-05,
+      "loss": 0.413,
+      "step": 7392
+    },
+    {
+      "epoch": 0.3246649176152892,
+      "grad_norm": 2.0625,
+      "learning_rate": 3.822173277190657e-05,
+      "loss": 0.4249,
+      "step": 7394
+    },
+    {
+      "epoch": 0.3247527360944927,
+      "grad_norm": 2.109375,
+      "learning_rate": 3.8215860695626396e-05,
+      "loss": 0.4091,
+      "step": 7396
+    },
+    {
+      "epoch": 0.32484055457369615,
+      "grad_norm": 1.9375,
+      "learning_rate": 3.8209987607272444e-05,
+      "loss": 0.4415,
+      "step": 7398
+    },
+    {
+      "epoch": 0.3249283730528997,
+      "grad_norm": 2.34375,
+      "learning_rate": 3.820411350729448e-05,
+      "loss": 0.4045,
+      "step": 7400
+    },
+    {
+      "epoch": 0.32501619153210315,
+      "grad_norm": 2.046875,
+      "learning_rate": 3.819823839614234e-05,
+      "loss": 0.4163,
+      "step": 7402
+    },
+    {
+      "epoch": 0.3251040100113066,
+      "grad_norm": 2.375,
+      "learning_rate": 3.8192362274265934e-05,
+      "loss": 0.3869,
+      "step": 7404
+    },
+    {
+      "epoch": 0.3251918284905101,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.8186485142115266e-05,
+      "loss": 0.4192,
+      "step": 7406
+    },
+    {
+      "epoch": 0.3252796469697136,
+      "grad_norm": 1.9609375,
+      "learning_rate": 3.81806070001404e-05,
+      "loss": 0.4053,
+      "step": 7408
+    },
+    {
+      "epoch": 0.3253674654489171,
+      "grad_norm": 1.96875,
+      "learning_rate": 3.817472784879149e-05,
+      "loss": 0.4142,
+      "step": 7410
+    },
+    {
+      "epoch": 0.32545528392812056,
+      "grad_norm": 1.984375,
+      "learning_rate": 3.816884768851877e-05,
+      "loss": 0.4375,
+      "step": 7412
+    },
+    {
+      "epoch": 0.32554310240732404,
+      "grad_norm": 1.921875,
+      "learning_rate": 3.816296651977254e-05,
+      "loss": 0.4372,
+      "step": 7414
+    },
+    {
+      "epoch": 0.32563092088652756,
+      "grad_norm": 2.171875,
+      "learning_rate": 3.815708434300317e-05,
+      "loss": 0.4311,
+      "step": 7416
+    },
+    {
+      "epoch": 0.32571873936573104,
+      "grad_norm": 2.375,
+      "learning_rate": 3.815120115866113e-05,
+      "loss": 0.4264,
+      "step": 7418
+    },
+    {
+      "epoch": 0.3258065578449345,
+      "grad_norm": 2.171875,
+      "learning_rate": 3.814531696719695e-05,
+      "loss": 0.4248,
+      "step": 7420
+    },
+    {
+      "epoch": 0.325894376324138,
+      "grad_norm": 2.359375,
+      "learning_rate": 3.813943176906125e-05,
+      "loss": 0.4473,
+      "step": 7422
+    },
+    {
+      "epoch": 0.3259821948033415,
+      "grad_norm": 2.3125,
+      "learning_rate": 3.813354556470471e-05,
+      "loss": 0.4678,
+      "step": 7424
+    },
+    {
+      "epoch": 0.326070013282545,
+      "grad_norm": 2.46875,
+      "learning_rate": 3.812765835457811e-05,
+      "loss": 0.432,
+      "step": 7426
+    },
+    {
+      "epoch": 0.32615783176174845,
+      "grad_norm": 2.0625,
+      "learning_rate": 3.812177013913228e-05,
+      "loss": 0.4186,
+      "step": 7428
+    },
+    {
+      "epoch": 0.326245650240952,
+      "grad_norm": 2.03125,
+      "learning_rate": 3.8115880918818147e-05,
+      "loss": 0.4128,
+      "step": 7430
+    },
+    {
+      "epoch": 0.32633346872015545,
+      "grad_norm": 2.109375,
+      "learning_rate": 3.810999069408671e-05,
+      "loss": 0.4101,
+      "step": 7432
+    },
+    {
+      "epoch": 0.3264212871993589,
+      "grad_norm": 2.03125,
+      "learning_rate": 3.810409946538904e-05,
+      "loss": 0.3973,
+      "step": 7434
+    },
+    {
+      "epoch": 0.3265091056785624,
+      "grad_norm": 1.921875,
+      "learning_rate": 3.80982072331763e-05,
+      "loss": 0.417,
+      "step": 7436
+    },
+    {
+      "epoch": 0.3265969241577659,
+      "grad_norm": 2.015625,
+      "learning_rate": 3.809231399789971e-05,
+      "loss": 0.4263,
+      "step": 7438
+    },
+    {
+      "epoch": 0.3266847426369694,
+      "grad_norm": 2.203125,
+      "learning_rate": 3.808641976001057e-05,
+      "loss": 0.4417,
+      "step": 7440
+    },
+    {
+      "epoch": 0.32677256111617287,
+      "grad_norm": 2.234375,
+      "learning_rate": 3.808052451996027e-05,
+      "loss": 0.4295,
+      "step": 7442
+    },
+    {
+      "epoch": 0.32686037959537634,
+      "grad_norm": 2.28125,
+      "learning_rate": 3.8074628278200266e-05,
+      "loss": 0.405,
+      "step": 7444
+    },
+    {
+      "epoch": 0.32694819807457987,
+      "grad_norm": 2.265625,
+      "learning_rate": 3.806873103518209e-05,
+      "loss": 0.4161,
+      "step": 7446
+    },
+    {
+      "epoch": 0.32703601655378334,
+      "grad_norm": 2.125,
+      "learning_rate": 3.806283279135736e-05,
+      "loss": 0.4151,
+      "step": 7448
+    },
+    {
+      "epoch": 0.3271238350329868,
+      "grad_norm": 2.125,
+      "learning_rate": 3.805693354717777e-05,
+      "loss": 0.4141,
+      "step": 7450
+    },
+    {
+      "epoch": 0.3272116535121903,
+      "grad_norm": 2.140625,
+      "learning_rate": 3.805103330309508e-05,
+      "loss": 0.4028,
+      "step": 7452
+    },
+    {
+      "epoch": 0.3272994719913938,
+      "grad_norm": 2.203125,
+      "learning_rate": 3.804513205956113e-05,
+      "loss": 0.4393,
+      "step": 7454
+    },
+    {
+      "epoch": 0.3273872904705973,
+      "grad_norm": 2.171875,
+      "learning_rate": 3.8039229817027834e-05,
+      "loss": 0.4222,
+      "step": 7456
+    },
+    {
+      "epoch": 0.32747510894980075,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.803332657594719e-05,
+      "loss": 0.4171,
+      "step": 7458
+    },
+    {
+      "epoch": 0.3275629274290042,
+      "grad_norm": 2.03125,
+      "learning_rate": 3.8027422336771275e-05,
+      "loss": 0.3707,
+      "step": 7460
+    },
+    {
+      "epoch": 0.32765074590820775,
+      "grad_norm": 2.03125,
+      "learning_rate": 3.802151709995224e-05,
+      "loss": 0.39,
+      "step": 7462
+    },
+    {
+      "epoch": 0.3277385643874112,
+      "grad_norm": 2.5,
+      "learning_rate": 3.80156108659423e-05,
+      "loss": 0.44,
+      "step": 7464
+    },
+    {
+      "epoch": 0.3278263828666147,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.800970363519376e-05,
+      "loss": 0.3907,
+      "step": 7466
+    },
+    {
+      "epoch": 0.32791420134581817,
+      "grad_norm": 2.421875,
+      "learning_rate": 3.8003795408159004e-05,
+      "loss": 0.412,
+      "step": 7468
+    },
+    {
+      "epoch": 0.3280020198250217,
+      "grad_norm": 2.109375,
+      "learning_rate": 3.799788618529046e-05,
+      "loss": 0.4174,
+      "step": 7470
+    },
+    {
+      "epoch": 0.32808983830422517,
+      "grad_norm": 2.25,
+      "learning_rate": 3.7991975967040694e-05,
+      "loss": 0.4239,
+      "step": 7472
+    },
+    {
+      "epoch": 0.32817765678342864,
+      "grad_norm": 2.453125,
+      "learning_rate": 3.798606475386229e-05,
+      "loss": 0.4289,
+      "step": 7474
+    },
+    {
+      "epoch": 0.32826547526263217,
+      "grad_norm": 2.234375,
+      "learning_rate": 3.798015254620794e-05,
+      "loss": 0.4332,
+      "step": 7476
+    },
+    {
+      "epoch": 0.32835329374183564,
+      "grad_norm": 1.8984375,
+      "learning_rate": 3.797423934453038e-05,
+      "loss": 0.4123,
+      "step": 7478
+    },
+    {
+      "epoch": 0.3284411122210391,
+      "grad_norm": 2.0625,
+      "learning_rate": 3.796832514928247e-05,
+      "loss": 0.4131,
+      "step": 7480
+    },
+    {
+      "epoch": 0.3285289307002426,
+      "grad_norm": 2.0625,
+      "learning_rate": 3.796240996091711e-05,
+      "loss": 0.4259,
+      "step": 7482
+    },
+    {
+      "epoch": 0.3286167491794461,
+      "grad_norm": 2.34375,
+      "learning_rate": 3.795649377988729e-05,
+      "loss": 0.3994,
+      "step": 7484
+    },
+    {
+      "epoch": 0.3287045676586496,
+      "grad_norm": 1.8046875,
+      "learning_rate": 3.795057660664607e-05,
+      "loss": 0.4358,
+      "step": 7486
+    },
+    {
+      "epoch": 0.32879238613785305,
+      "grad_norm": 2.03125,
+      "learning_rate": 3.794465844164659e-05,
+      "loss": 0.4326,
+      "step": 7488
+    },
+    {
+      "epoch": 0.3288802046170565,
+      "grad_norm": 2.234375,
+      "learning_rate": 3.793873928534206e-05,
+      "loss": 0.4255,
+      "step": 7490
+    },
+    {
+      "epoch": 0.32896802309626005,
+      "grad_norm": 1.9765625,
+      "learning_rate": 3.793281913818578e-05,
+      "loss": 0.4384,
+      "step": 7492
+    },
+    {
+      "epoch": 0.3290558415754635,
+      "grad_norm": 1.90625,
+      "learning_rate": 3.7926898000631106e-05,
+      "loss": 0.4206,
+      "step": 7494
+    },
+    {
+      "epoch": 0.329143660054667,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.792097587313148e-05,
+      "loss": 0.4219,
+      "step": 7496
+    },
+    {
+      "epoch": 0.32923147853387047,
+      "grad_norm": 2.140625,
+      "learning_rate": 3.791505275614043e-05,
+      "loss": 0.4579,
+      "step": 7498
+    },
+    {
+      "epoch": 0.329319297013074,
+      "grad_norm": 2.0,
+      "learning_rate": 3.790912865011154e-05,
+      "loss": 0.4063,
+      "step": 7500
+    },
+    {
+      "epoch": 0.32940711549227747,
+      "grad_norm": 2.015625,
+      "learning_rate": 3.790320355549849e-05,
+      "loss": 0.4278,
+      "step": 7502
+    },
+    {
+      "epoch": 0.32949493397148094,
+      "grad_norm": 2.03125,
+      "learning_rate": 3.789727747275502e-05,
+      "loss": 0.4088,
+      "step": 7504
+    },
+    {
+      "epoch": 0.3295827524506844,
+      "grad_norm": 2.28125,
+      "learning_rate": 3.7891350402334935e-05,
+      "loss": 0.3918,
+      "step": 7506
+    },
+    {
+      "epoch": 0.32967057092988794,
+      "grad_norm": 2.0,
+      "learning_rate": 3.788542234469216e-05,
+      "loss": 0.4272,
+      "step": 7508
+    },
+    {
+      "epoch": 0.3297583894090914,
+      "grad_norm": 2.53125,
+      "learning_rate": 3.7879493300280643e-05,
+      "loss": 0.4163,
+      "step": 7510
+    },
+    {
+      "epoch": 0.3298462078882949,
+      "grad_norm": 2.3125,
+      "learning_rate": 3.7873563269554454e-05,
+      "loss": 0.436,
+      "step": 7512
+    },
+    {
+      "epoch": 0.32993402636749836,
+      "grad_norm": 2.53125,
+      "learning_rate": 3.78676322529677e-05,
+      "loss": 0.4093,
+      "step": 7514
+    },
+    {
+      "epoch": 0.3300218448467019,
+      "grad_norm": 2.625,
+      "learning_rate": 3.786170025097457e-05,
+      "loss": 0.4155,
+      "step": 7516
+    },
+    {
+      "epoch": 0.33010966332590536,
+      "grad_norm": 2.046875,
+      "learning_rate": 3.7855767264029366e-05,
+      "loss": 0.401,
+      "step": 7518
+    },
+    {
+      "epoch": 0.33019748180510883,
+      "grad_norm": 2.296875,
+      "learning_rate": 3.784983329258642e-05,
+      "loss": 0.4521,
+      "step": 7520
+    },
+    {
+      "epoch": 0.3302853002843123,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.784389833710016e-05,
+      "loss": 0.3999,
+      "step": 7522
+    },
+    {
+      "epoch": 0.3303731187635158,
+      "grad_norm": 1.9375,
+      "learning_rate": 3.783796239802509e-05,
+      "loss": 0.44,
+      "step": 7524
+    },
+    {
+      "epoch": 0.3304609372427193,
+      "grad_norm": 1.9609375,
+      "learning_rate": 3.783202547581577e-05,
+      "loss": 0.4255,
+      "step": 7526
+    },
+    {
+      "epoch": 0.33054875572192277,
+      "grad_norm": 2.296875,
+      "learning_rate": 3.782608757092687e-05,
+      "loss": 0.4362,
+      "step": 7528
+    },
+    {
+      "epoch": 0.3306365742011263,
+      "grad_norm": 2.03125,
+      "learning_rate": 3.782014868381312e-05,
+      "loss": 0.4049,
+      "step": 7530
+    },
+    {
+      "epoch": 0.33072439268032977,
+      "grad_norm": 2.171875,
+      "learning_rate": 3.781420881492929e-05,
+      "loss": 0.4214,
+      "step": 7532
+    },
+    {
+      "epoch": 0.33081221115953324,
+      "grad_norm": 2.109375,
+      "learning_rate": 3.780826796473029e-05,
+      "loss": 0.4206,
+      "step": 7534
+    },
+    {
+      "epoch": 0.3309000296387367,
+      "grad_norm": 2.140625,
+      "learning_rate": 3.780232613367105e-05,
+      "loss": 0.4223,
+      "step": 7536
+    },
+    {
+      "epoch": 0.33098784811794024,
+      "grad_norm": 1.8359375,
+      "learning_rate": 3.779638332220662e-05,
+      "loss": 0.3965,
+      "step": 7538
+    },
+    {
+      "epoch": 0.3310756665971437,
+      "grad_norm": 2.0,
+      "learning_rate": 3.7790439530792075e-05,
+      "loss": 0.386,
+      "step": 7540
+    },
+    {
+      "epoch": 0.3311634850763472,
+      "grad_norm": 2.109375,
+      "learning_rate": 3.778449475988261e-05,
+      "loss": 0.4401,
+      "step": 7542
+    },
+    {
+      "epoch": 0.33125130355555066,
+      "grad_norm": 1.921875,
+      "learning_rate": 3.777854900993347e-05,
+      "loss": 0.4216,
+      "step": 7544
+    },
+    {
+      "epoch": 0.3313391220347542,
+      "grad_norm": 2.015625,
+      "learning_rate": 3.777260228139999e-05,
+      "loss": 0.3927,
+      "step": 7546
+    },
+    {
+      "epoch": 0.33142694051395766,
+      "grad_norm": 2.234375,
+      "learning_rate": 3.776665457473756e-05,
+      "loss": 0.4544,
+      "step": 7548
+    },
+    {
+      "epoch": 0.33151475899316113,
+      "grad_norm": 1.9375,
+      "learning_rate": 3.776070589040166e-05,
+      "loss": 0.401,
+      "step": 7550
+    },
+    {
+      "epoch": 0.3316025774723646,
+      "grad_norm": 2.09375,
+      "learning_rate": 3.775475622884785e-05,
+      "loss": 0.4037,
+      "step": 7552
+    },
+    {
+      "epoch": 0.33169039595156813,
+      "grad_norm": 2.078125,
+      "learning_rate": 3.774880559053175e-05,
+      "loss": 0.4327,
+      "step": 7554
+    },
+    {
+      "epoch": 0.3317782144307716,
+      "grad_norm": 2.0,
+      "learning_rate": 3.7742853975909056e-05,
+      "loss": 0.4202,
+      "step": 7556
+    },
+    {
+      "epoch": 0.3318660329099751,
+      "grad_norm": 2.03125,
+      "learning_rate": 3.773690138543555e-05,
+      "loss": 0.3931,
+      "step": 7558
+    },
+    {
+      "epoch": 0.33195385138917854,
+      "grad_norm": 2.09375,
+      "learning_rate": 3.773094781956709e-05,
+      "loss": 0.4,
+      "step": 7560
+    },
+    {
+      "epoch": 0.33204166986838207,
+      "grad_norm": 2.125,
+      "learning_rate": 3.772499327875959e-05,
+      "loss": 0.4021,
+      "step": 7562
+    },
+    {
+      "epoch": 0.33212948834758554,
+      "grad_norm": 2.4375,
+      "learning_rate": 3.771903776346905e-05,
+      "loss": 0.4401,
+      "step": 7564
+    },
+    {
+      "epoch": 0.332217306826789,
+      "grad_norm": 1.9140625,
+      "learning_rate": 3.771308127415155e-05,
+      "loss": 0.4201,
+      "step": 7566
+    },
+    {
+      "epoch": 0.3323051253059925,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.770712381126325e-05,
+      "loss": 0.4519,
+      "step": 7568
+    },
+    {
+      "epoch": 0.332392943785196,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.7701165375260344e-05,
+      "loss": 0.4011,
+      "step": 7570
+    },
+    {
+      "epoch": 0.3324807622643995,
+      "grad_norm": 2.171875,
+      "learning_rate": 3.7695205966599154e-05,
+      "loss": 0.3996,
+      "step": 7572
+    },
+    {
+      "epoch": 0.33256858074360296,
+      "grad_norm": 2.171875,
+      "learning_rate": 3.768924558573606e-05,
+      "loss": 0.406,
+      "step": 7574
+    },
+    {
+      "epoch": 0.33265639922280643,
+      "grad_norm": 2.09375,
+      "learning_rate": 3.768328423312749e-05,
+      "loss": 0.4221,
+      "step": 7576
+    },
+    {
+      "epoch": 0.33274421770200996,
+      "grad_norm": 2.03125,
+      "learning_rate": 3.767732190922997e-05,
+      "loss": 0.4178,
+      "step": 7578
+    },
+    {
+      "epoch": 0.33283203618121343,
+      "grad_norm": 2.09375,
+      "learning_rate": 3.767135861450011e-05,
+      "loss": 0.4026,
+      "step": 7580
+    },
+    {
+      "epoch": 0.3329198546604169,
+      "grad_norm": 1.921875,
+      "learning_rate": 3.7665394349394556e-05,
+      "loss": 0.4005,
+      "step": 7582
+    },
+    {
+      "epoch": 0.33300767313962043,
+      "grad_norm": 2.171875,
+      "learning_rate": 3.765942911437007e-05,
+      "loss": 0.4165,
+      "step": 7584
+    },
+    {
+      "epoch": 0.3330954916188239,
+      "grad_norm": 1.9140625,
+      "learning_rate": 3.7653462909883474e-05,
+      "loss": 0.3877,
+      "step": 7586
+    },
+    {
+      "epoch": 0.3331833100980274,
+      "grad_norm": 1.953125,
+      "learning_rate": 3.764749573639165e-05,
+      "loss": 0.4211,
+      "step": 7588
+    },
+    {
+      "epoch": 0.33327112857723085,
+      "grad_norm": 2.453125,
+      "learning_rate": 3.7641527594351577e-05,
+      "loss": 0.4552,
+      "step": 7590
+    },
+    {
+      "epoch": 0.3333589470564344,
+      "grad_norm": 2.203125,
+      "learning_rate": 3.763555848422028e-05,
+      "loss": 0.4462,
+      "step": 7592
+    },
+    {
+      "epoch": 0.33344676553563785,
+      "grad_norm": 2.0625,
+      "learning_rate": 3.762958840645489e-05,
+      "loss": 0.4193,
+      "step": 7594
+    },
+    {
+      "epoch": 0.3335345840148413,
+      "grad_norm": 2.0625,
+      "learning_rate": 3.7623617361512595e-05,
+      "loss": 0.4079,
+      "step": 7596
+    },
+    {
+      "epoch": 0.3336224024940448,
+      "grad_norm": 2.3125,
+      "learning_rate": 3.7617645349850655e-05,
+      "loss": 0.4191,
+      "step": 7598
+    },
+    {
+      "epoch": 0.3337102209732483,
+      "grad_norm": 2.109375,
+      "learning_rate": 3.761167237192641e-05,
+      "loss": 0.4295,
+      "step": 7600
+    },
+    {
+      "epoch": 0.3337980394524518,
+      "grad_norm": 2.203125,
+      "learning_rate": 3.7605698428197265e-05,
+      "loss": 0.4517,
+      "step": 7602
+    },
+    {
+      "epoch": 0.33388585793165526,
+      "grad_norm": 2.078125,
+      "learning_rate": 3.7599723519120725e-05,
+      "loss": 0.4201,
+      "step": 7604
+    },
+    {
+      "epoch": 0.33397367641085873,
+      "grad_norm": 1.90625,
+      "learning_rate": 3.759374764515433e-05,
+      "loss": 0.4051,
+      "step": 7606
+    },
+    {
+      "epoch": 0.33406149489006226,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.7587770806755715e-05,
+      "loss": 0.4344,
+      "step": 7608
+    },
+    {
+      "epoch": 0.33414931336926573,
+      "grad_norm": 2.0625,
+      "learning_rate": 3.7581793004382603e-05,
+      "loss": 0.4551,
+      "step": 7610
+    },
+    {
+      "epoch": 0.3342371318484692,
+      "grad_norm": 2.296875,
+      "learning_rate": 3.757581423849277e-05,
+      "loss": 0.4337,
+      "step": 7612
+    },
+    {
+      "epoch": 0.3343249503276727,
+      "grad_norm": 2.140625,
+      "learning_rate": 3.7569834509544054e-05,
+      "loss": 0.4249,
+      "step": 7614
+    },
+    {
+      "epoch": 0.3344127688068762,
+      "grad_norm": 2.28125,
+      "learning_rate": 3.756385381799441e-05,
+      "loss": 0.4328,
+      "step": 7616
+    },
+    {
+      "epoch": 0.3345005872860797,
+      "grad_norm": 1.984375,
+      "learning_rate": 3.755787216430182e-05,
+      "loss": 0.3575,
+      "step": 7618
+    },
+    {
+      "epoch": 0.33458840576528315,
+      "grad_norm": 2.46875,
+      "learning_rate": 3.755188954892438e-05,
+      "loss": 0.4209,
+      "step": 7620
+    },
+    {
+      "epoch": 0.3346762242444866,
+      "grad_norm": 2.40625,
+      "learning_rate": 3.754590597232023e-05,
+      "loss": 0.4171,
+      "step": 7622
+    },
+    {
+      "epoch": 0.33476404272369015,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.753992143494759e-05,
+      "loss": 0.4345,
+      "step": 7624
+    },
+    {
+      "epoch": 0.3348518612028936,
+      "grad_norm": 2.109375,
+      "learning_rate": 3.753393593726475e-05,
+      "loss": 0.4031,
+      "step": 7626
+    },
+    {
+      "epoch": 0.3349396796820971,
+      "grad_norm": 2.203125,
+      "learning_rate": 3.7527949479730104e-05,
+      "loss": 0.4153,
+      "step": 7628
+    },
+    {
+      "epoch": 0.3350274981613006,
+      "grad_norm": 2.078125,
+      "learning_rate": 3.752196206280207e-05,
+      "loss": 0.4425,
+      "step": 7630
+    },
+    {
+      "epoch": 0.3351153166405041,
+      "grad_norm": 1.9609375,
+      "learning_rate": 3.7515973686939184e-05,
+      "loss": 0.3951,
+      "step": 7632
+    },
+    {
+      "epoch": 0.33520313511970756,
+      "grad_norm": 1.984375,
+      "learning_rate": 3.7509984352600044e-05,
+      "loss": 0.4232,
+      "step": 7634
+    },
+    {
+      "epoch": 0.33529095359891103,
+      "grad_norm": 2.609375,
+      "learning_rate": 3.7503994060243296e-05,
+      "loss": 0.3971,
+      "step": 7636
+    },
+    {
+      "epoch": 0.33537877207811456,
+      "grad_norm": 2.234375,
+      "learning_rate": 3.7498002810327694e-05,
+      "loss": 0.451,
+      "step": 7638
+    },
+    {
+      "epoch": 0.33546659055731803,
+      "grad_norm": 2.234375,
+      "learning_rate": 3.749201060331203e-05,
+      "loss": 0.3823,
+      "step": 7640
+    },
+    {
+      "epoch": 0.3355544090365215,
+      "grad_norm": 2.328125,
+      "learning_rate": 3.74860174396552e-05,
+      "loss": 0.4138,
+      "step": 7642
+    },
+    {
+      "epoch": 0.335642227515725,
+      "grad_norm": 1.7890625,
+      "learning_rate": 3.7480023319816164e-05,
+      "loss": 0.4207,
+      "step": 7644
+    },
+    {
+      "epoch": 0.3357300459949285,
+      "grad_norm": 1.8828125,
+      "learning_rate": 3.747402824425395e-05,
+      "loss": 0.3933,
+      "step": 7646
+    },
+    {
+      "epoch": 0.335817864474132,
+      "grad_norm": 2.359375,
+      "learning_rate": 3.7468032213427666e-05,
+      "loss": 0.4109,
+      "step": 7648
+    },
+    {
+      "epoch": 0.33590568295333545,
+      "grad_norm": 1.8046875,
+      "learning_rate": 3.7462035227796484e-05,
+      "loss": 0.447,
+      "step": 7650
+    },
+    {
+      "epoch": 0.3359935014325389,
+      "grad_norm": 2.171875,
+      "learning_rate": 3.745603728781966e-05,
+      "loss": 0.3935,
+      "step": 7652
+    },
+    {
+      "epoch": 0.33608131991174245,
+      "grad_norm": 2.03125,
+      "learning_rate": 3.745003839395651e-05,
+      "loss": 0.4265,
+      "step": 7654
+    },
+    {
+      "epoch": 0.3361691383909459,
+      "grad_norm": 1.9453125,
+      "learning_rate": 3.744403854666643e-05,
+      "loss": 0.4304,
+      "step": 7656
+    },
+    {
+      "epoch": 0.3362569568701494,
+      "grad_norm": 1.953125,
+      "learning_rate": 3.743803774640891e-05,
+      "loss": 0.4293,
+      "step": 7658
+    },
+    {
+      "epoch": 0.33634477534935286,
+      "grad_norm": 2.125,
+      "learning_rate": 3.743203599364347e-05,
+      "loss": 0.3955,
+      "step": 7660
+    },
+    {
+      "epoch": 0.3364325938285564,
+      "grad_norm": 2.0625,
+      "learning_rate": 3.7426033288829725e-05,
+      "loss": 0.4054,
+      "step": 7662
+    },
+    {
+      "epoch": 0.33652041230775986,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.7420029632427375e-05,
+      "loss": 0.4184,
+      "step": 7664
+    },
+    {
+      "epoch": 0.33660823078696334,
+      "grad_norm": 2.28125,
+      "learning_rate": 3.7414025024896195e-05,
+      "loss": 0.4242,
+      "step": 7666
+    },
+    {
+      "epoch": 0.3366960492661668,
+      "grad_norm": 2.34375,
+      "learning_rate": 3.740801946669599e-05,
+      "loss": 0.4264,
+      "step": 7668
+    },
+    {
+      "epoch": 0.33678386774537034,
+      "grad_norm": 2.34375,
+      "learning_rate": 3.740201295828668e-05,
+      "loss": 0.4062,
+      "step": 7670
+    },
+    {
+      "epoch": 0.3368716862245738,
+      "grad_norm": 2.453125,
+      "learning_rate": 3.739600550012824e-05,
+      "loss": 0.4151,
+      "step": 7672
+    },
+    {
+      "epoch": 0.3369595047037773,
+      "grad_norm": 2.046875,
+      "learning_rate": 3.738999709268074e-05,
+      "loss": 0.3959,
+      "step": 7674
+    },
+    {
+      "epoch": 0.33704732318298075,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.738398773640428e-05,
+      "loss": 0.4408,
+      "step": 7676
+    },
+    {
+      "epoch": 0.3371351416621843,
+      "grad_norm": 1.921875,
+      "learning_rate": 3.737797743175907e-05,
+      "loss": 0.4428,
+      "step": 7678
+    },
+    {
+      "epoch": 0.33722296014138775,
+      "grad_norm": 2.078125,
+      "learning_rate": 3.7371966179205386e-05,
+      "loss": 0.4136,
+      "step": 7680
+    },
+    {
+      "epoch": 0.3373107786205912,
+      "grad_norm": 1.8359375,
+      "learning_rate": 3.7365953979203574e-05,
+      "loss": 0.3965,
+      "step": 7682
+    },
+    {
+      "epoch": 0.33739859709979475,
+      "grad_norm": 1.890625,
+      "learning_rate": 3.735994083221403e-05,
+      "loss": 0.4064,
+      "step": 7684
+    },
+    {
+      "epoch": 0.3374864155789982,
+      "grad_norm": 1.984375,
+      "learning_rate": 3.735392673869726e-05,
+      "loss": 0.4215,
+      "step": 7686
+    },
+    {
+      "epoch": 0.3375742340582017,
+      "grad_norm": 2.25,
+      "learning_rate": 3.734791169911382e-05,
+      "loss": 0.4152,
+      "step": 7688
+    },
+    {
+      "epoch": 0.33766205253740517,
+      "grad_norm": 2.328125,
+      "learning_rate": 3.734189571392434e-05,
+      "loss": 0.402,
+      "step": 7690
+    },
+    {
+      "epoch": 0.3377498710166087,
+      "grad_norm": 2.0,
+      "learning_rate": 3.7335878783589525e-05,
+      "loss": 0.4335,
+      "step": 7692
+    },
+    {
+      "epoch": 0.33783768949581217,
+      "grad_norm": 1.984375,
+      "learning_rate": 3.7329860908570156e-05,
+      "loss": 0.4104,
+      "step": 7694
+    },
+    {
+      "epoch": 0.33792550797501564,
+      "grad_norm": 1.921875,
+      "learning_rate": 3.732384208932709e-05,
+      "loss": 0.4066,
+      "step": 7696
+    },
+    {
+      "epoch": 0.3380133264542191,
+      "grad_norm": 1.9609375,
+      "learning_rate": 3.7317822326321236e-05,
+      "loss": 0.4157,
+      "step": 7698
+    },
+    {
+      "epoch": 0.33810114493342264,
+      "grad_norm": 2.296875,
+      "learning_rate": 3.7311801620013596e-05,
+      "loss": 0.4167,
+      "step": 7700
+    },
+    {
+      "epoch": 0.3381889634126261,
+      "grad_norm": 2.0,
+      "learning_rate": 3.730577997086524e-05,
+      "loss": 0.4178,
+      "step": 7702
+    },
+    {
+      "epoch": 0.3382767818918296,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.72997573793373e-05,
+      "loss": 0.4049,
+      "step": 7704
+    },
+    {
+      "epoch": 0.33836460037103305,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.729373384589099e-05,
+      "loss": 0.428,
+      "step": 7706
+    },
+    {
+      "epoch": 0.3384524188502366,
+      "grad_norm": 2.171875,
+      "learning_rate": 3.72877093709876e-05,
+      "loss": 0.4219,
+      "step": 7708
+    },
+    {
+      "epoch": 0.33854023732944005,
+      "grad_norm": 2.328125,
+      "learning_rate": 3.728168395508848e-05,
+      "loss": 0.415,
+      "step": 7710
+    },
+    {
+      "epoch": 0.3386280558086435,
+      "grad_norm": 2.109375,
+      "learning_rate": 3.7275657598655066e-05,
+      "loss": 0.4394,
+      "step": 7712
+    },
+    {
+      "epoch": 0.338715874287847,
+      "grad_norm": 2.171875,
+      "learning_rate": 3.726963030214884e-05,
+      "loss": 0.425,
+      "step": 7714
+    },
+    {
+      "epoch": 0.3388036927670505,
+      "grad_norm": 2.125,
+      "learning_rate": 3.726360206603138e-05,
+      "loss": 0.4231,
+      "step": 7716
+    },
+    {
+      "epoch": 0.338891511246254,
+      "grad_norm": 2.03125,
+      "learning_rate": 3.725757289076434e-05,
+      "loss": 0.4206,
+      "step": 7718
+    },
+    {
+      "epoch": 0.33897932972545747,
+      "grad_norm": 2.203125,
+      "learning_rate": 3.725154277680943e-05,
+      "loss": 0.4093,
+      "step": 7720
+    },
+    {
+      "epoch": 0.33906714820466094,
+      "grad_norm": 1.90625,
+      "learning_rate": 3.7245511724628444e-05,
+      "loss": 0.4074,
+      "step": 7722
+    },
+    {
+      "epoch": 0.33915496668386447,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.7239479734683226e-05,
+      "loss": 0.4005,
+      "step": 7724
+    },
+    {
+      "epoch": 0.33924278516306794,
+      "grad_norm": 2.484375,
+      "learning_rate": 3.723344680743571e-05,
+      "loss": 0.436,
+      "step": 7726
+    },
+    {
+      "epoch": 0.3393306036422714,
+      "grad_norm": 1.96875,
+      "learning_rate": 3.722741294334792e-05,
+      "loss": 0.4087,
+      "step": 7728
+    },
+    {
+      "epoch": 0.33941842212147494,
+      "grad_norm": 2.265625,
+      "learning_rate": 3.722137814288191e-05,
+      "loss": 0.4382,
+      "step": 7730
+    },
+    {
+      "epoch": 0.3395062406006784,
+      "grad_norm": 2.140625,
+      "learning_rate": 3.721534240649983e-05,
+      "loss": 0.3787,
+      "step": 7732
+    },
+    {
+      "epoch": 0.3395940590798819,
+      "grad_norm": 2.28125,
+      "learning_rate": 3.720930573466389e-05,
+      "loss": 0.401,
+      "step": 7734
+    },
+    {
+      "epoch": 0.33968187755908535,
+      "grad_norm": 2.125,
+      "learning_rate": 3.720326812783641e-05,
+      "loss": 0.4417,
+      "step": 7736
+    },
+    {
+      "epoch": 0.3397696960382889,
+      "grad_norm": 2.3125,
+      "learning_rate": 3.719722958647972e-05,
+      "loss": 0.4282,
+      "step": 7738
+    },
+    {
+      "epoch": 0.33985751451749235,
+      "grad_norm": 2.125,
+      "learning_rate": 3.719119011105627e-05,
+      "loss": 0.4352,
+      "step": 7740
+    },
+    {
+      "epoch": 0.3399453329966958,
+      "grad_norm": 2.09375,
+      "learning_rate": 3.7185149702028555e-05,
+      "loss": 0.4154,
+      "step": 7742
+    },
+    {
+      "epoch": 0.3400331514758993,
+      "grad_norm": 2.234375,
+      "learning_rate": 3.717910835985916e-05,
+      "loss": 0.4041,
+      "step": 7744
+    },
+    {
+      "epoch": 0.3401209699551028,
+      "grad_norm": 2.09375,
+      "learning_rate": 3.717306608501072e-05,
+      "loss": 0.3961,
+      "step": 7746
+    },
+    {
+      "epoch": 0.3402087884343063,
+      "grad_norm": 2.140625,
+      "learning_rate": 3.716702287794597e-05,
+      "loss": 0.4197,
+      "step": 7748
+    },
+    {
+      "epoch": 0.34029660691350977,
+      "grad_norm": 2.046875,
+      "learning_rate": 3.7160978739127684e-05,
+      "loss": 0.4037,
+      "step": 7750
+    },
+    {
+      "epoch": 0.34038442539271324,
+      "grad_norm": 2.234375,
+      "learning_rate": 3.7154933669018724e-05,
+      "loss": 0.4529,
+      "step": 7752
+    },
+    {
+      "epoch": 0.34047224387191677,
+      "grad_norm": 2.03125,
+      "learning_rate": 3.714888766808204e-05,
+      "loss": 0.414,
+      "step": 7754
+    },
+    {
+      "epoch": 0.34056006235112024,
+      "grad_norm": 2.265625,
+      "learning_rate": 3.714284073678063e-05,
+      "loss": 0.4048,
+      "step": 7756
+    },
+    {
+      "epoch": 0.3406478808303237,
+      "grad_norm": 2.0625,
+      "learning_rate": 3.7136792875577556e-05,
+      "loss": 0.4016,
+      "step": 7758
+    },
+    {
+      "epoch": 0.3407356993095272,
+      "grad_norm": 2.46875,
+      "learning_rate": 3.713074408493598e-05,
+      "loss": 0.4241,
+      "step": 7760
+    },
+    {
+      "epoch": 0.3408235177887307,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.7124694365319114e-05,
+      "loss": 0.4218,
+      "step": 7762
+    },
+    {
+      "epoch": 0.3409113362679342,
+      "grad_norm": 1.90625,
+      "learning_rate": 3.711864371719024e-05,
+      "loss": 0.4143,
+      "step": 7764
+    },
+    {
+      "epoch": 0.34099915474713766,
+      "grad_norm": 2.0,
+      "learning_rate": 3.711259214101273e-05,
+      "loss": 0.4018,
+      "step": 7766
+    },
+    {
+      "epoch": 0.34108697322634113,
+      "grad_norm": 2.171875,
+      "learning_rate": 3.710653963725001e-05,
+      "loss": 0.4196,
+      "step": 7768
+    },
+    {
+      "epoch": 0.34117479170554466,
+      "grad_norm": 2.109375,
+      "learning_rate": 3.710048620636558e-05,
+      "loss": 0.404,
+      "step": 7770
+    },
+    {
+      "epoch": 0.3412626101847481,
+      "grad_norm": 1.984375,
+      "learning_rate": 3.7094431848823026e-05,
+      "loss": 0.3881,
+      "step": 7772
+    },
+    {
+      "epoch": 0.3413504286639516,
+      "grad_norm": 2.34375,
+      "learning_rate": 3.708837656508597e-05,
+      "loss": 0.3954,
+      "step": 7774
+    },
+    {
+      "epoch": 0.34143824714315507,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.708232035561815e-05,
+      "loss": 0.3994,
+      "step": 7776
+    },
+    {
+      "epoch": 0.3415260656223586,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.707626322088333e-05,
+      "loss": 0.4353,
+      "step": 7778
+    },
+    {
+      "epoch": 0.34161388410156207,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.707020516134539e-05,
+      "loss": 0.4419,
+      "step": 7780
+    },
+    {
+      "epoch": 0.34170170258076554,
+      "grad_norm": 2.203125,
+      "learning_rate": 3.706414617746823e-05,
+      "loss": 0.4037,
+      "step": 7782
+    },
+    {
+      "epoch": 0.34178952105996907,
+      "grad_norm": 1.9453125,
+      "learning_rate": 3.705808626971587e-05,
+      "loss": 0.4231,
+      "step": 7784
+    },
+    {
+      "epoch": 0.34187733953917254,
+      "grad_norm": 2.09375,
+      "learning_rate": 3.705202543855237e-05,
+      "loss": 0.4201,
+      "step": 7786
+    },
+    {
+      "epoch": 0.341965158018376,
+      "grad_norm": 2.125,
+      "learning_rate": 3.704596368444187e-05,
+      "loss": 0.4198,
+      "step": 7788
+    },
+    {
+      "epoch": 0.3420529764975795,
+      "grad_norm": 1.8671875,
+      "learning_rate": 3.70399010078486e-05,
+      "loss": 0.4011,
+      "step": 7790
+    },
+    {
+      "epoch": 0.342140794976783,
+      "grad_norm": 2.015625,
+      "learning_rate": 3.703383740923682e-05,
+      "loss": 0.4002,
+      "step": 7792
+    },
+    {
+      "epoch": 0.3422286134559865,
+      "grad_norm": 2.171875,
+      "learning_rate": 3.7027772889070875e-05,
+      "loss": 0.4169,
+      "step": 7794
+    },
+    {
+      "epoch": 0.34231643193518996,
+      "grad_norm": 1.859375,
+      "learning_rate": 3.702170744781521e-05,
+      "loss": 0.3867,
+      "step": 7796
+    },
+    {
+      "epoch": 0.34240425041439343,
+      "grad_norm": 2.03125,
+      "learning_rate": 3.70156410859343e-05,
+      "loss": 0.4391,
+      "step": 7798
+    },
+    {
+      "epoch": 0.34249206889359696,
+      "grad_norm": 2.296875,
+      "learning_rate": 3.700957380389272e-05,
+      "loss": 0.4103,
+      "step": 7800
+    },
+    {
+      "epoch": 0.34257988737280043,
+      "grad_norm": 2.28125,
+      "learning_rate": 3.7003505602155085e-05,
+      "loss": 0.4221,
+      "step": 7802
+    },
+    {
+      "epoch": 0.3426677058520039,
+      "grad_norm": 2.0,
+      "learning_rate": 3.6997436481186134e-05,
+      "loss": 0.4157,
+      "step": 7804
+    },
+    {
+      "epoch": 0.3427555243312074,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.699136644145061e-05,
+      "loss": 0.4063,
+      "step": 7806
+    },
+    {
+      "epoch": 0.3428433428104109,
+      "grad_norm": 2.125,
+      "learning_rate": 3.698529548341337e-05,
+      "loss": 0.4183,
+      "step": 7808
+    },
+    {
+      "epoch": 0.3429311612896144,
+      "grad_norm": 1.875,
+      "learning_rate": 3.697922360753933e-05,
+      "loss": 0.4057,
+      "step": 7810
+    },
+    {
+      "epoch": 0.34301897976881784,
+      "grad_norm": 1.828125,
+      "learning_rate": 3.697315081429348e-05,
+      "loss": 0.3984,
+      "step": 7812
+    },
+    {
+      "epoch": 0.3431067982480213,
+      "grad_norm": 1.9609375,
+      "learning_rate": 3.696707710414086e-05,
+      "loss": 0.4262,
+      "step": 7814
+    },
+    {
+      "epoch": 0.34319461672722484,
+      "grad_norm": 2.0625,
+      "learning_rate": 3.696100247754661e-05,
+      "loss": 0.4487,
+      "step": 7816
+    },
+    {
+      "epoch": 0.3432824352064283,
+      "grad_norm": 2.25,
+      "learning_rate": 3.6954926934975925e-05,
+      "loss": 0.4345,
+      "step": 7818
+    },
+    {
+      "epoch": 0.3433702536856318,
+      "grad_norm": 2.3125,
+      "learning_rate": 3.694885047689407e-05,
+      "loss": 0.3907,
+      "step": 7820
+    },
+    {
+      "epoch": 0.34345807216483526,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.694277310376639e-05,
+      "loss": 0.4209,
+      "step": 7822
+    },
+    {
+      "epoch": 0.3435458906440388,
+      "grad_norm": 2.390625,
+      "learning_rate": 3.693669481605827e-05,
+      "loss": 0.4242,
+      "step": 7824
+    },
+    {
+      "epoch": 0.34363370912324226,
+      "grad_norm": 2.640625,
+      "learning_rate": 3.69306156142352e-05,
+      "loss": 0.3965,
+      "step": 7826
+    },
+    {
+      "epoch": 0.34372152760244573,
+      "grad_norm": 2.671875,
+      "learning_rate": 3.692453549876273e-05,
+      "loss": 0.4348,
+      "step": 7828
+    },
+    {
+      "epoch": 0.34380934608164926,
+      "grad_norm": 2.375,
+      "learning_rate": 3.691845447010647e-05,
+      "loss": 0.3989,
+      "step": 7830
+    },
+    {
+      "epoch": 0.34389716456085273,
+      "grad_norm": 2.40625,
+      "learning_rate": 3.691237252873211e-05,
+      "loss": 0.3943,
+      "step": 7832
+    },
+    {
+      "epoch": 0.3439849830400562,
+      "grad_norm": 2.234375,
+      "learning_rate": 3.690628967510541e-05,
+      "loss": 0.4105,
+      "step": 7834
+    },
+    {
+      "epoch": 0.3440728015192597,
+      "grad_norm": 2.46875,
+      "learning_rate": 3.69002059096922e-05,
+      "loss": 0.4227,
+      "step": 7836
+    },
+    {
+      "epoch": 0.3441606199984632,
+      "grad_norm": 2.265625,
+      "learning_rate": 3.6894121232958354e-05,
+      "loss": 0.4361,
+      "step": 7838
+    },
+    {
+      "epoch": 0.3442484384776667,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.688803564536986e-05,
+      "loss": 0.4565,
+      "step": 7840
+    },
+    {
+      "epoch": 0.34433625695687015,
+      "grad_norm": 2.09375,
+      "learning_rate": 3.688194914739274e-05,
+      "loss": 0.4395,
+      "step": 7842
+    },
+    {
+      "epoch": 0.3444240754360736,
+      "grad_norm": 2.484375,
+      "learning_rate": 3.687586173949311e-05,
+      "loss": 0.4276,
+      "step": 7844
+    },
+    {
+      "epoch": 0.34451189391527715,
+      "grad_norm": 2.3125,
+      "learning_rate": 3.686977342213714e-05,
+      "loss": 0.4263,
+      "step": 7846
+    },
+    {
+      "epoch": 0.3445997123944806,
+      "grad_norm": 2.765625,
+      "learning_rate": 3.686368419579108e-05,
+      "loss": 0.3967,
+      "step": 7848
+    },
+    {
+      "epoch": 0.3446875308736841,
+      "grad_norm": 2.328125,
+      "learning_rate": 3.685759406092124e-05,
+      "loss": 0.4122,
+      "step": 7850
+    },
+    {
+      "epoch": 0.34477534935288756,
+      "grad_norm": 1.8984375,
+      "learning_rate": 3.6851503017994e-05,
+      "loss": 0.4182,
+      "step": 7852
+    },
+    {
+      "epoch": 0.3448631678320911,
+      "grad_norm": 2.0,
+      "learning_rate": 3.6845411067475825e-05,
+      "loss": 0.4255,
+      "step": 7854
+    },
+    {
+      "epoch": 0.34495098631129456,
+      "grad_norm": 2.140625,
+      "learning_rate": 3.683931820983322e-05,
+      "loss": 0.3841,
+      "step": 7856
+    },
+    {
+      "epoch": 0.34503880479049803,
+      "grad_norm": 2.09375,
+      "learning_rate": 3.68332244455328e-05,
+      "loss": 0.4518,
+      "step": 7858
+    },
+    {
+      "epoch": 0.3451266232697015,
+      "grad_norm": 2.625,
+      "learning_rate": 3.6827129775041216e-05,
+      "loss": 0.3772,
+      "step": 7860
+    },
+    {
+      "epoch": 0.34521444174890503,
+      "grad_norm": 2.0625,
+      "learning_rate": 3.682103419882519e-05,
+      "loss": 0.4353,
+      "step": 7862
+    },
+    {
+      "epoch": 0.3453022602281085,
+      "grad_norm": 2.078125,
+      "learning_rate": 3.6814937717351525e-05,
+      "loss": 0.3897,
+      "step": 7864
+    },
+    {
+      "epoch": 0.345390078707312,
+      "grad_norm": 2.296875,
+      "learning_rate": 3.6808840331087115e-05,
+      "loss": 0.4197,
+      "step": 7866
+    },
+    {
+      "epoch": 0.34547789718651545,
+      "grad_norm": 1.953125,
+      "learning_rate": 3.6802742040498875e-05,
+      "loss": 0.4177,
+      "step": 7868
+    },
+    {
+      "epoch": 0.345565715665719,
+      "grad_norm": 2.203125,
+      "learning_rate": 3.679664284605381e-05,
+      "loss": 0.4177,
+      "step": 7870
+    },
+    {
+      "epoch": 0.34565353414492245,
+      "grad_norm": 2.390625,
+      "learning_rate": 3.679054274821903e-05,
+      "loss": 0.4031,
+      "step": 7872
+    },
+    {
+      "epoch": 0.3457413526241259,
+      "grad_norm": 2.6875,
+      "learning_rate": 3.678444174746164e-05,
+      "loss": 0.4048,
+      "step": 7874
+    },
+    {
+      "epoch": 0.3458291711033294,
+      "grad_norm": 2.265625,
+      "learning_rate": 3.677833984424888e-05,
+      "loss": 0.4164,
+      "step": 7876
+    },
+    {
+      "epoch": 0.3459169895825329,
+      "grad_norm": 2.140625,
+      "learning_rate": 3.677223703904803e-05,
+      "loss": 0.4107,
+      "step": 7878
+    },
+    {
+      "epoch": 0.3460048080617364,
+      "grad_norm": 2.65625,
+      "learning_rate": 3.6766133332326455e-05,
+      "loss": 0.4107,
+      "step": 7880
+    },
+    {
+      "epoch": 0.34609262654093986,
+      "grad_norm": 2.375,
+      "learning_rate": 3.676002872455157e-05,
+      "loss": 0.4097,
+      "step": 7882
+    },
+    {
+      "epoch": 0.3461804450201434,
+      "grad_norm": 2.421875,
+      "learning_rate": 3.675392321619086e-05,
+      "loss": 0.4145,
+      "step": 7884
+    },
+    {
+      "epoch": 0.34626826349934686,
+      "grad_norm": 1.921875,
+      "learning_rate": 3.674781680771189e-05,
+      "loss": 0.4026,
+      "step": 7886
+    },
+    {
+      "epoch": 0.34635608197855033,
+      "grad_norm": 1.921875,
+      "learning_rate": 3.674170949958229e-05,
+      "loss": 0.4055,
+      "step": 7888
+    },
+    {
+      "epoch": 0.3464439004577538,
+      "grad_norm": 1.8125,
+      "learning_rate": 3.673560129226976e-05,
+      "loss": 0.4262,
+      "step": 7890
+    },
+    {
+      "epoch": 0.34653171893695733,
+      "grad_norm": 2.359375,
+      "learning_rate": 3.6729492186242073e-05,
+      "loss": 0.4097,
+      "step": 7892
+    },
+    {
+      "epoch": 0.3466195374161608,
+      "grad_norm": 2.140625,
+      "learning_rate": 3.672338218196708e-05,
+      "loss": 0.4043,
+      "step": 7894
+    },
+    {
+      "epoch": 0.3467073558953643,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.6717271279912645e-05,
+      "loss": 0.4346,
+      "step": 7896
+    },
+    {
+      "epoch": 0.34679517437456775,
+      "grad_norm": 2.296875,
+      "learning_rate": 3.6711159480546785e-05,
+      "loss": 0.4076,
+      "step": 7898
+    },
+    {
+      "epoch": 0.3468829928537713,
+      "grad_norm": 2.515625,
+      "learning_rate": 3.6705046784337514e-05,
+      "loss": 0.4282,
+      "step": 7900
+    },
+    {
+      "epoch": 0.34697081133297475,
+      "grad_norm": 2.734375,
+      "learning_rate": 3.669893319175296e-05,
+      "loss": 0.4104,
+      "step": 7902
+    },
+    {
+      "epoch": 0.3470586298121782,
+      "grad_norm": 2.5,
+      "learning_rate": 3.6692818703261286e-05,
+      "loss": 0.4188,
+      "step": 7904
+    },
+    {
+      "epoch": 0.3471464482913817,
+      "grad_norm": 1.8671875,
+      "learning_rate": 3.668670331933076e-05,
+      "loss": 0.4156,
+      "step": 7906
+    },
+    {
+      "epoch": 0.3472342667705852,
+      "grad_norm": 2.078125,
+      "learning_rate": 3.6680587040429696e-05,
+      "loss": 0.4276,
+      "step": 7908
+    },
+    {
+      "epoch": 0.3473220852497887,
+      "grad_norm": 1.796875,
+      "learning_rate": 3.667446986702647e-05,
+      "loss": 0.4195,
+      "step": 7910
+    },
+    {
+      "epoch": 0.34740990372899216,
+      "grad_norm": 2.0,
+      "learning_rate": 3.6668351799589557e-05,
+      "loss": 0.4033,
+      "step": 7912
+    },
+    {
+      "epoch": 0.34749772220819564,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.666223283858745e-05,
+      "loss": 0.4044,
+      "step": 7914
+    },
+    {
+      "epoch": 0.34758554068739916,
+      "grad_norm": 1.9921875,
+      "learning_rate": 3.6656112984488765e-05,
+      "loss": 0.4108,
+      "step": 7916
+    },
+    {
+      "epoch": 0.34767335916660264,
+      "grad_norm": 2.28125,
+      "learning_rate": 3.664999223776215e-05,
+      "loss": 0.4093,
+      "step": 7918
+    },
+    {
+      "epoch": 0.3477611776458061,
+      "grad_norm": 2.109375,
+      "learning_rate": 3.664387059887634e-05,
+      "loss": 0.3886,
+      "step": 7920
+    },
+    {
+      "epoch": 0.3478489961250096,
+      "grad_norm": 2.0625,
+      "learning_rate": 3.6637748068300123e-05,
+      "loss": 0.4283,
+      "step": 7922
+    },
+    {
+      "epoch": 0.3479368146042131,
+      "grad_norm": 2.015625,
+      "learning_rate": 3.663162464650237e-05,
+      "loss": 0.398,
+      "step": 7924
+    },
+    {
+      "epoch": 0.3480246330834166,
+      "grad_norm": 1.7578125,
+      "learning_rate": 3.662550033395202e-05,
+      "loss": 0.4176,
+      "step": 7926
+    },
+    {
+      "epoch": 0.34811245156262005,
+      "grad_norm": 2.0,
+      "learning_rate": 3.661937513111806e-05,
+      "loss": 0.3965,
+      "step": 7928
+    },
+    {
+      "epoch": 0.3482002700418235,
+      "grad_norm": 2.34375,
+      "learning_rate": 3.661324903846957e-05,
+      "loss": 0.4322,
+      "step": 7930
+    },
+    {
+      "epoch": 0.34828808852102705,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.660712205647568e-05,
+      "loss": 0.3999,
+      "step": 7932
+    },
+    {
+      "epoch": 0.3483759070002305,
+      "grad_norm": 2.234375,
+      "learning_rate": 3.6600994185605614e-05,
+      "loss": 0.4251,
+      "step": 7934
+    },
+    {
+      "epoch": 0.348463725479434,
+      "grad_norm": 2.4375,
+      "learning_rate": 3.6594865426328625e-05,
+      "loss": 0.4191,
+      "step": 7936
+    },
+    {
+      "epoch": 0.3485515439586375,
+      "grad_norm": 2.484375,
+      "learning_rate": 3.658873577911406e-05,
+      "loss": 0.3965,
+      "step": 7938
+    },
+    {
+      "epoch": 0.348639362437841,
+      "grad_norm": 1.9140625,
+      "learning_rate": 3.658260524443133e-05,
+      "loss": 0.4126,
+      "step": 7940
+    },
+    {
+      "epoch": 0.34872718091704447,
+      "grad_norm": 2.046875,
+      "learning_rate": 3.657647382274992e-05,
+      "loss": 0.4249,
+      "step": 7942
+    },
+    {
+      "epoch": 0.34881499939624794,
+      "grad_norm": 2.203125,
+      "learning_rate": 3.657034151453936e-05,
+      "loss": 0.39,
+      "step": 7944
+    },
+    {
+      "epoch": 0.34890281787545147,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.656420832026928e-05,
+      "loss": 0.4209,
+      "step": 7946
+    },
+    {
+      "epoch": 0.34899063635465494,
+      "grad_norm": 1.890625,
+      "learning_rate": 3.655807424040936e-05,
+      "loss": 0.4177,
+      "step": 7948
+    },
+    {
+      "epoch": 0.3490784548338584,
+      "grad_norm": 1.921875,
+      "learning_rate": 3.655193927542933e-05,
+      "loss": 0.4071,
+      "step": 7950
+    },
+    {
+      "epoch": 0.3491662733130619,
+      "grad_norm": 1.921875,
+      "learning_rate": 3.654580342579903e-05,
+      "loss": 0.4208,
+      "step": 7952
+    },
+    {
+      "epoch": 0.3492540917922654,
+      "grad_norm": 2.015625,
+      "learning_rate": 3.6539666691988336e-05,
+      "loss": 0.3849,
+      "step": 7954
+    },
+    {
+      "epoch": 0.3493419102714689,
+      "grad_norm": 1.8828125,
+      "learning_rate": 3.65335290744672e-05,
+      "loss": 0.4318,
+      "step": 7956
+    },
+    {
+      "epoch": 0.34942972875067235,
+      "grad_norm": 1.890625,
+      "learning_rate": 3.6527390573705645e-05,
+      "loss": 0.3816,
+      "step": 7958
+    },
+    {
+      "epoch": 0.3495175472298758,
+      "grad_norm": 2.015625,
+      "learning_rate": 3.652125119017375e-05,
+      "loss": 0.4051,
+      "step": 7960
+    },
+    {
+      "epoch": 0.34960536570907935,
+      "grad_norm": 2.109375,
+      "learning_rate": 3.651511092434168e-05,
+      "loss": 0.386,
+      "step": 7962
+    },
+    {
+      "epoch": 0.3496931841882828,
+      "grad_norm": 1.84375,
+      "learning_rate": 3.650896977667965e-05,
+      "loss": 0.438,
+      "step": 7964
+    },
+    {
+      "epoch": 0.3497810026674863,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.6502827747657964e-05,
+      "loss": 0.3965,
+      "step": 7966
+    },
+    {
+      "epoch": 0.34986882114668977,
+      "grad_norm": 1.9375,
+      "learning_rate": 3.649668483774696e-05,
+      "loss": 0.419,
+      "step": 7968
+    },
+    {
+      "epoch": 0.3499566396258933,
+      "grad_norm": 1.9140625,
+      "learning_rate": 3.649054104741709e-05,
+      "loss": 0.3912,
+      "step": 7970
+    },
+    {
+      "epoch": 0.35004445810509677,
+      "grad_norm": 2.109375,
+      "learning_rate": 3.648439637713883e-05,
+      "loss": 0.4051,
+      "step": 7972
+    },
+    {
+      "epoch": 0.35013227658430024,
+      "grad_norm": 1.9453125,
+      "learning_rate": 3.6478250827382734e-05,
+      "loss": 0.4089,
+      "step": 7974
+    },
+    {
+      "epoch": 0.3502200950635037,
+      "grad_norm": 1.96875,
+      "learning_rate": 3.647210439861944e-05,
+      "loss": 0.4123,
+      "step": 7976
+    },
+    {
+      "epoch": 0.35030791354270724,
+      "grad_norm": 2.125,
+      "learning_rate": 3.646595709131965e-05,
+      "loss": 0.4088,
+      "step": 7978
+    },
+    {
+      "epoch": 0.3503957320219107,
+      "grad_norm": 1.8671875,
+      "learning_rate": 3.6459808905954105e-05,
+      "loss": 0.396,
+      "step": 7980
+    },
+    {
+      "epoch": 0.3504835505011142,
+      "grad_norm": 1.8203125,
+      "learning_rate": 3.645365984299366e-05,
+      "loss": 0.4095,
+      "step": 7982
+    },
+    {
+      "epoch": 0.3505713689803177,
+      "grad_norm": 2.015625,
+      "learning_rate": 3.644750990290919e-05,
+      "loss": 0.4179,
+      "step": 7984
+    },
+    {
+      "epoch": 0.3506591874595212,
+      "grad_norm": 2.171875,
+      "learning_rate": 3.6441359086171665e-05,
+      "loss": 0.4492,
+      "step": 7986
+    },
+    {
+      "epoch": 0.35074700593872465,
+      "grad_norm": 2.015625,
+      "learning_rate": 3.643520739325213e-05,
+      "loss": 0.4249,
+      "step": 7988
+    },
+    {
+      "epoch": 0.3508348244179281,
+      "grad_norm": 2.3125,
+      "learning_rate": 3.6429054824621656e-05,
+      "loss": 0.4198,
+      "step": 7990
+    },
+    {
+      "epoch": 0.35092264289713165,
+      "grad_norm": 1.9296875,
+      "learning_rate": 3.6422901380751436e-05,
+      "loss": 0.4131,
+      "step": 7992
+    },
+    {
+      "epoch": 0.3510104613763351,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.641674706211269e-05,
+      "loss": 0.4094,
+      "step": 7994
+    },
+    {
+      "epoch": 0.3510982798555386,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.641059186917671e-05,
+      "loss": 0.418,
+      "step": 7996
+    },
+    {
+      "epoch": 0.35118609833474207,
+      "grad_norm": 2.34375,
+      "learning_rate": 3.6404435802414866e-05,
+      "loss": 0.4279,
+      "step": 7998
+    },
+    {
+      "epoch": 0.3512739168139456,
+      "grad_norm": 1.8984375,
+      "learning_rate": 3.63982788622986e-05,
+      "loss": 0.4254,
+      "step": 8000
+    },
+    {
+      "epoch": 0.35136173529314907,
+      "grad_norm": 2.125,
+      "learning_rate": 3.63921210492994e-05,
+      "loss": 0.3997,
+      "step": 8002
+    },
+    {
+      "epoch": 0.35144955377235254,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.638596236388886e-05,
+      "loss": 0.4476,
+      "step": 8004
+    },
+    {
+      "epoch": 0.351537372251556,
+      "grad_norm": 2.09375,
+      "learning_rate": 3.637980280653858e-05,
+      "loss": 0.4054,
+      "step": 8006
+    },
+    {
+      "epoch": 0.35162519073075954,
+      "grad_norm": 1.984375,
+      "learning_rate": 3.637364237772027e-05,
+      "loss": 0.4005,
+      "step": 8008
+    },
+    {
+      "epoch": 0.351713009209963,
+      "grad_norm": 2.140625,
+      "learning_rate": 3.63674810779057e-05,
+      "loss": 0.4214,
+      "step": 8010
+    },
+    {
+      "epoch": 0.3518008276891665,
+      "grad_norm": 2.015625,
+      "learning_rate": 3.63613189075667e-05,
+      "loss": 0.4369,
+      "step": 8012
+    },
+    {
+      "epoch": 0.35188864616836996,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.6355155867175185e-05,
+      "loss": 0.4079,
+      "step": 8014
+    },
+    {
+      "epoch": 0.3519764646475735,
+      "grad_norm": 2.328125,
+      "learning_rate": 3.63489919572031e-05,
+      "loss": 0.3932,
+      "step": 8016
+    },
+    {
+      "epoch": 0.35206428312677696,
+      "grad_norm": 2.6875,
+      "learning_rate": 3.6342827178122505e-05,
+      "loss": 0.4338,
+      "step": 8018
+    },
+    {
+      "epoch": 0.3521521016059804,
+      "grad_norm": 2.234375,
+      "learning_rate": 3.6336661530405486e-05,
+      "loss": 0.4306,
+      "step": 8020
+    },
+    {
+      "epoch": 0.3522399200851839,
+      "grad_norm": 2.125,
+      "learning_rate": 3.63304950145242e-05,
+      "loss": 0.3814,
+      "step": 8022
+    },
+    {
+      "epoch": 0.3523277385643874,
+      "grad_norm": 2.375,
+      "learning_rate": 3.6324327630950886e-05,
+      "loss": 0.4214,
+      "step": 8024
+    },
+    {
+      "epoch": 0.3524155570435909,
+      "grad_norm": 1.9921875,
+      "learning_rate": 3.6318159380157855e-05,
+      "loss": 0.4044,
+      "step": 8026
+    },
+    {
+      "epoch": 0.35250337552279437,
+      "grad_norm": 1.953125,
+      "learning_rate": 3.631199026261746e-05,
+      "loss": 0.402,
+      "step": 8028
+    },
+    {
+      "epoch": 0.35259119400199784,
+      "grad_norm": 1.78125,
+      "learning_rate": 3.6305820278802143e-05,
+      "loss": 0.4106,
+      "step": 8030
+    },
+    {
+      "epoch": 0.35267901248120137,
+      "grad_norm": 1.90625,
+      "learning_rate": 3.62996494291844e-05,
+      "loss": 0.3897,
+      "step": 8032
+    },
+    {
+      "epoch": 0.35276683096040484,
+      "grad_norm": 1.9765625,
+      "learning_rate": 3.62934777142368e-05,
+      "loss": 0.4002,
+      "step": 8034
+    },
+    {
+      "epoch": 0.3528546494396083,
+      "grad_norm": 1.9609375,
+      "learning_rate": 3.6287305134431956e-05,
+      "loss": 0.4293,
+      "step": 8036
+    },
+    {
+      "epoch": 0.35294246791881184,
+      "grad_norm": 2.078125,
+      "learning_rate": 3.6281131690242594e-05,
+      "loss": 0.4017,
+      "step": 8038
+    },
+    {
+      "epoch": 0.3530302863980153,
+      "grad_norm": 1.9609375,
+      "learning_rate": 3.627495738214145e-05,
+      "loss": 0.3893,
+      "step": 8040
+    },
+    {
+      "epoch": 0.3531181048772188,
+      "grad_norm": 1.9453125,
+      "learning_rate": 3.626878221060137e-05,
+      "loss": 0.4135,
+      "step": 8042
+    },
+    {
+      "epoch": 0.35320592335642226,
+      "grad_norm": 2.296875,
+      "learning_rate": 3.626260617609525e-05,
+      "loss": 0.4264,
+      "step": 8044
+    },
+    {
+      "epoch": 0.3532937418356258,
+      "grad_norm": 1.96875,
+      "learning_rate": 3.625642927909605e-05,
+      "loss": 0.4158,
+      "step": 8046
+    },
+    {
+      "epoch": 0.35338156031482926,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.625025152007679e-05,
+      "loss": 0.4414,
+      "step": 8048
+    },
+    {
+      "epoch": 0.35346937879403273,
+      "grad_norm": 2.09375,
+      "learning_rate": 3.6244072899510586e-05,
+      "loss": 0.4108,
+      "step": 8050
+    },
+    {
+      "epoch": 0.3535571972732362,
+      "grad_norm": 1.8671875,
+      "learning_rate": 3.6237893417870574e-05,
+      "loss": 0.4202,
+      "step": 8052
+    },
+    {
+      "epoch": 0.35364501575243973,
+      "grad_norm": 2.171875,
+      "learning_rate": 3.623171307562999e-05,
+      "loss": 0.4186,
+      "step": 8054
+    },
+    {
+      "epoch": 0.3537328342316432,
+      "grad_norm": 2.046875,
+      "learning_rate": 3.622553187326213e-05,
+      "loss": 0.3793,
+      "step": 8056
+    },
+    {
+      "epoch": 0.3538206527108467,
+      "grad_norm": 2.078125,
+      "learning_rate": 3.6219349811240345e-05,
+      "loss": 0.4024,
+      "step": 8058
+    },
+    {
+      "epoch": 0.35390847119005014,
+      "grad_norm": 1.96875,
+      "learning_rate": 3.621316689003806e-05,
+      "loss": 0.4224,
+      "step": 8060
+    },
+    {
+      "epoch": 0.35399628966925367,
+      "grad_norm": 2.109375,
+      "learning_rate": 3.6206983110128765e-05,
+      "loss": 0.4361,
+      "step": 8062
+    },
+    {
+      "epoch": 0.35408410814845714,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.620079847198602e-05,
+      "loss": 0.4311,
+      "step": 8064
+    },
+    {
+      "epoch": 0.3541719266276606,
+      "grad_norm": 2.09375,
+      "learning_rate": 3.619461297608345e-05,
+      "loss": 0.3927,
+      "step": 8066
+    },
+    {
+      "epoch": 0.3542597451068641,
+      "grad_norm": 2.03125,
+      "learning_rate": 3.6188426622894726e-05,
+      "loss": 0.429,
+      "step": 8068
+    },
+    {
+      "epoch": 0.3543475635860676,
+      "grad_norm": 2.140625,
+      "learning_rate": 3.618223941289362e-05,
+      "loss": 0.3921,
+      "step": 8070
+    },
+    {
+      "epoch": 0.3544353820652711,
+      "grad_norm": 2.046875,
+      "learning_rate": 3.617605134655393e-05,
+      "loss": 0.4144,
+      "step": 8072
+    },
+    {
+      "epoch": 0.35452320054447456,
+      "grad_norm": 2.5625,
+      "learning_rate": 3.6169862424349544e-05,
+      "loss": 0.3993,
+      "step": 8074
+    },
+    {
+      "epoch": 0.35461101902367803,
+      "grad_norm": 2.34375,
+      "learning_rate": 3.6163672646754423e-05,
+      "loss": 0.3845,
+      "step": 8076
+    },
+    {
+      "epoch": 0.35469883750288156,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.615748201424257e-05,
+      "loss": 0.4187,
+      "step": 8078
+    },
+    {
+      "epoch": 0.35478665598208503,
+      "grad_norm": 2.421875,
+      "learning_rate": 3.615129052728808e-05,
+      "loss": 0.3998,
+      "step": 8080
+    },
+    {
+      "epoch": 0.3548744744612885,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.6145098186365085e-05,
+      "loss": 0.437,
+      "step": 8082
+    },
+    {
+      "epoch": 0.35496229294049203,
+      "grad_norm": 2.53125,
+      "learning_rate": 3.6138904991947794e-05,
+      "loss": 0.4331,
+      "step": 8084
+    },
+    {
+      "epoch": 0.3550501114196955,
+      "grad_norm": 1.9765625,
+      "learning_rate": 3.613271094451049e-05,
+      "loss": 0.392,
+      "step": 8086
+    },
+    {
+      "epoch": 0.355137929898899,
+      "grad_norm": 2.046875,
+      "learning_rate": 3.612651604452752e-05,
+      "loss": 0.407,
+      "step": 8088
+    },
+    {
+      "epoch": 0.35522574837810245,
+      "grad_norm": 2.140625,
+      "learning_rate": 3.612032029247326e-05,
+      "loss": 0.4184,
+      "step": 8090
+    },
+    {
+      "epoch": 0.355313566857306,
+      "grad_norm": 2.078125,
+      "learning_rate": 3.611412368882223e-05,
+      "loss": 0.4275,
+      "step": 8092
+    },
+    {
+      "epoch": 0.35540138533650945,
+      "grad_norm": 1.9921875,
+      "learning_rate": 3.610792623404894e-05,
+      "loss": 0.3976,
+      "step": 8094
+    },
+    {
+      "epoch": 0.3554892038157129,
+      "grad_norm": 2.25,
+      "learning_rate": 3.6101727928628e-05,
+      "loss": 0.3745,
+      "step": 8096
+    },
+    {
+      "epoch": 0.3555770222949164,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.6095528773034065e-05,
+      "loss": 0.4196,
+      "step": 8098
+    },
+    {
+      "epoch": 0.3556648407741199,
+      "grad_norm": 2.25,
+      "learning_rate": 3.608932876774188e-05,
+      "loss": 0.3838,
+      "step": 8100
+    },
+    {
+      "epoch": 0.3557526592533234,
+      "grad_norm": 2.359375,
+      "learning_rate": 3.6083127913226235e-05,
+      "loss": 0.4328,
+      "step": 8102
+    },
+    {
+      "epoch": 0.35584047773252686,
+      "grad_norm": 2.6875,
+      "learning_rate": 3.6076926209962e-05,
+      "loss": 0.41,
+      "step": 8104
+    },
+    {
+      "epoch": 0.35592829621173033,
+      "grad_norm": 2.640625,
+      "learning_rate": 3.607072365842411e-05,
+      "loss": 0.3981,
+      "step": 8106
+    },
+    {
+      "epoch": 0.35601611469093386,
+      "grad_norm": 2.171875,
+      "learning_rate": 3.606452025908754e-05,
+      "loss": 0.4053,
+      "step": 8108
+    },
+    {
+      "epoch": 0.35610393317013733,
+      "grad_norm": 2.109375,
+      "learning_rate": 3.605831601242735e-05,
+      "loss": 0.4144,
+      "step": 8110
+    },
+    {
+      "epoch": 0.3561917516493408,
+      "grad_norm": 1.8515625,
+      "learning_rate": 3.605211091891868e-05,
+      "loss": 0.3831,
+      "step": 8112
+    },
+    {
+      "epoch": 0.3562795701285443,
+      "grad_norm": 2.109375,
+      "learning_rate": 3.604590497903671e-05,
+      "loss": 0.404,
+      "step": 8114
+    },
+    {
+      "epoch": 0.3563673886077478,
+      "grad_norm": 2.6875,
+      "learning_rate": 3.603969819325668e-05,
+      "loss": 0.4024,
+      "step": 8116
+    },
+    {
+      "epoch": 0.3564552070869513,
+      "grad_norm": 2.375,
+      "learning_rate": 3.6033490562053915e-05,
+      "loss": 0.4221,
+      "step": 8118
+    },
+    {
+      "epoch": 0.35654302556615475,
+      "grad_norm": 2.09375,
+      "learning_rate": 3.6027282085903795e-05,
+      "loss": 0.3815,
+      "step": 8120
+    },
+    {
+      "epoch": 0.3566308440453582,
+      "grad_norm": 2.265625,
+      "learning_rate": 3.6021072765281776e-05,
+      "loss": 0.3893,
+      "step": 8122
+    },
+    {
+      "epoch": 0.35671866252456175,
+      "grad_norm": 2.421875,
+      "learning_rate": 3.6014862600663354e-05,
+      "loss": 0.4161,
+      "step": 8124
+    },
+    {
+      "epoch": 0.3568064810037652,
+      "grad_norm": 2.3125,
+      "learning_rate": 3.600865159252413e-05,
+      "loss": 0.4167,
+      "step": 8126
+    },
+    {
+      "epoch": 0.3568942994829687,
+      "grad_norm": 2.09375,
+      "learning_rate": 3.6002439741339715e-05,
+      "loss": 0.4056,
+      "step": 8128
+    },
+    {
+      "epoch": 0.35698211796217216,
+      "grad_norm": 1.90625,
+      "learning_rate": 3.5996227047585837e-05,
+      "loss": 0.416,
+      "step": 8130
+    },
+    {
+      "epoch": 0.3570699364413757,
+      "grad_norm": 2.515625,
+      "learning_rate": 3.599001351173825e-05,
+      "loss": 0.386,
+      "step": 8132
+    },
+    {
+      "epoch": 0.35715775492057916,
+      "grad_norm": 2.25,
+      "learning_rate": 3.598379913427279e-05,
+      "loss": 0.4333,
+      "step": 8134
+    },
+    {
+      "epoch": 0.35724557339978263,
+      "grad_norm": 1.9921875,
+      "learning_rate": 3.597758391566536e-05,
+      "loss": 0.396,
+      "step": 8136
+    },
+    {
+      "epoch": 0.35733339187898616,
+      "grad_norm": 2.140625,
+      "learning_rate": 3.5971367856391925e-05,
+      "loss": 0.3873,
+      "step": 8138
+    },
+    {
+      "epoch": 0.35742121035818963,
+      "grad_norm": 2.25,
+      "learning_rate": 3.596515095692851e-05,
+      "loss": 0.4016,
+      "step": 8140
+    },
+    {
+      "epoch": 0.3575090288373931,
+      "grad_norm": 2.28125,
+      "learning_rate": 3.5958933217751214e-05,
+      "loss": 0.3966,
+      "step": 8142
+    },
+    {
+      "epoch": 0.3575968473165966,
+      "grad_norm": 2.46875,
+      "learning_rate": 3.595271463933617e-05,
+      "loss": 0.3895,
+      "step": 8144
+    },
+    {
+      "epoch": 0.3576846657958001,
+      "grad_norm": 2.28125,
+      "learning_rate": 3.5946495222159624e-05,
+      "loss": 0.4021,
+      "step": 8146
+    },
+    {
+      "epoch": 0.3577724842750036,
+      "grad_norm": 2.09375,
+      "learning_rate": 3.5940274966697846e-05,
+      "loss": 0.4019,
+      "step": 8148
+    },
+    {
+      "epoch": 0.35786030275420705,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.593405387342719e-05,
+      "loss": 0.4004,
+      "step": 8150
+    },
+    {
+      "epoch": 0.3579481212334105,
+      "grad_norm": 1.84375,
+      "learning_rate": 3.5927831942824056e-05,
+      "loss": 0.3691,
+      "step": 8152
+    },
+    {
+      "epoch": 0.35803593971261405,
+      "grad_norm": 2.0,
+      "learning_rate": 3.592160917536495e-05,
+      "loss": 0.386,
+      "step": 8154
+    },
+    {
+      "epoch": 0.3581237581918175,
+      "grad_norm": 1.8828125,
+      "learning_rate": 3.5915385571526385e-05,
+      "loss": 0.394,
+      "step": 8156
+    },
+    {
+      "epoch": 0.358211576671021,
+      "grad_norm": 2.03125,
+      "learning_rate": 3.590916113178498e-05,
+      "loss": 0.398,
+      "step": 8158
+    },
+    {
+      "epoch": 0.35829939515022446,
+      "grad_norm": 1.9609375,
+      "learning_rate": 3.5902935856617403e-05,
+      "loss": 0.411,
+      "step": 8160
+    },
+    {
+      "epoch": 0.358387213629428,
+      "grad_norm": 2.203125,
+      "learning_rate": 3.589670974650038e-05,
+      "loss": 0.4071,
+      "step": 8162
+    },
+    {
+      "epoch": 0.35847503210863146,
+      "grad_norm": 2.390625,
+      "learning_rate": 3.5890482801910705e-05,
+      "loss": 0.4023,
+      "step": 8164
+    },
+    {
+      "epoch": 0.35856285058783494,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.5884255023325256e-05,
+      "loss": 0.3921,
+      "step": 8166
+    },
+    {
+      "epoch": 0.3586506690670384,
+      "grad_norm": 2.203125,
+      "learning_rate": 3.587802641122095e-05,
+      "loss": 0.449,
+      "step": 8168
+    },
+    {
+      "epoch": 0.35873848754624194,
+      "grad_norm": 2.390625,
+      "learning_rate": 3.5871796966074775e-05,
+      "loss": 0.398,
+      "step": 8170
+    },
+    {
+      "epoch": 0.3588263060254454,
+      "grad_norm": 2.25,
+      "learning_rate": 3.586556668836378e-05,
+      "loss": 0.4156,
+      "step": 8172
+    },
+    {
+      "epoch": 0.3589141245046489,
+      "grad_norm": 2.0625,
+      "learning_rate": 3.585933557856508e-05,
+      "loss": 0.4224,
+      "step": 8174
+    },
+    {
+      "epoch": 0.35900194298385235,
+      "grad_norm": 2.296875,
+      "learning_rate": 3.5853103637155854e-05,
+      "loss": 0.4249,
+      "step": 8176
+    },
+    {
+      "epoch": 0.3590897614630559,
+      "grad_norm": 1.8046875,
+      "learning_rate": 3.5846870864613355e-05,
+      "loss": 0.4294,
+      "step": 8178
+    },
+    {
+      "epoch": 0.35917757994225935,
+      "grad_norm": 2.203125,
+      "learning_rate": 3.584063726141489e-05,
+      "loss": 0.4183,
+      "step": 8180
+    },
+    {
+      "epoch": 0.3592653984214628,
+      "grad_norm": 1.9140625,
+      "learning_rate": 3.5834402828037816e-05,
+      "loss": 0.4175,
+      "step": 8182
+    },
+    {
+      "epoch": 0.35935321690066635,
+      "grad_norm": 1.8984375,
+      "learning_rate": 3.582816756495958e-05,
+      "loss": 0.4311,
+      "step": 8184
+    },
+    {
+      "epoch": 0.3594410353798698,
+      "grad_norm": 2.0625,
+      "learning_rate": 3.5821931472657674e-05,
+      "loss": 0.38,
+      "step": 8186
+    },
+    {
+      "epoch": 0.3595288538590733,
+      "grad_norm": 2.375,
+      "learning_rate": 3.581569455160967e-05,
+      "loss": 0.4245,
+      "step": 8188
+    },
+    {
+      "epoch": 0.35961667233827677,
+      "grad_norm": 1.984375,
+      "learning_rate": 3.580945680229317e-05,
+      "loss": 0.3865,
+      "step": 8190
+    },
+    {
+      "epoch": 0.3597044908174803,
+      "grad_norm": 1.8125,
+      "learning_rate": 3.580321822518588e-05,
+      "loss": 0.4175,
+      "step": 8192
+    },
+    {
+      "epoch": 0.35979230929668377,
+      "grad_norm": 1.8671875,
+      "learning_rate": 3.579697882076557e-05,
+      "loss": 0.3859,
+      "step": 8194
+    },
+    {
+      "epoch": 0.35988012777588724,
+      "grad_norm": 2.25,
+      "learning_rate": 3.5790738589510015e-05,
+      "loss": 0.3937,
+      "step": 8196
+    },
+    {
+      "epoch": 0.3599679462550907,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.578449753189711e-05,
+      "loss": 0.397,
+      "step": 8198
+    },
+    {
+      "epoch": 0.36005576473429424,
+      "grad_norm": 2.125,
+      "learning_rate": 3.5778255648404805e-05,
+      "loss": 0.409,
+      "step": 8200
+    },
+    {
+      "epoch": 0.3601435832134977,
+      "grad_norm": 2.140625,
+      "learning_rate": 3.57720129395111e-05,
+      "loss": 0.4007,
+      "step": 8202
+    },
+    {
+      "epoch": 0.3602314016927012,
+      "grad_norm": 2.03125,
+      "learning_rate": 3.576576940569406e-05,
+      "loss": 0.4047,
+      "step": 8204
+    },
+    {
+      "epoch": 0.36031922017190465,
+      "grad_norm": 1.8515625,
+      "learning_rate": 3.5759525047431816e-05,
+      "loss": 0.4141,
+      "step": 8206
+    },
+    {
+      "epoch": 0.3604070386511082,
+      "grad_norm": 1.9375,
+      "learning_rate": 3.575327986520257e-05,
+      "loss": 0.3848,
+      "step": 8208
+    },
+    {
+      "epoch": 0.36049485713031165,
+      "grad_norm": 2.0625,
+      "learning_rate": 3.574703385948457e-05,
+      "loss": 0.4296,
+      "step": 8210
+    },
+    {
+      "epoch": 0.3605826756095151,
+      "grad_norm": 2.03125,
+      "learning_rate": 3.574078703075613e-05,
+      "loss": 0.4021,
+      "step": 8212
+    },
+    {
+      "epoch": 0.3606704940887186,
+      "grad_norm": 2.046875,
+      "learning_rate": 3.573453937949566e-05,
+      "loss": 0.4013,
+      "step": 8214
+    },
+    {
+      "epoch": 0.3607583125679221,
+      "grad_norm": 2.015625,
+      "learning_rate": 3.572829090618159e-05,
+      "loss": 0.376,
+      "step": 8216
+    },
+    {
+      "epoch": 0.3608461310471256,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.572204161129243e-05,
+      "loss": 0.4091,
+      "step": 8218
+    },
+    {
+      "epoch": 0.36093394952632907,
+      "grad_norm": 1.953125,
+      "learning_rate": 3.571579149530675e-05,
+      "loss": 0.4112,
+      "step": 8220
+    },
+    {
+      "epoch": 0.36102176800553254,
+      "grad_norm": 2.046875,
+      "learning_rate": 3.5709540558703186e-05,
+      "loss": 0.4014,
+      "step": 8222
+    },
+    {
+      "epoch": 0.36110958648473607,
+      "grad_norm": 1.921875,
+      "learning_rate": 3.570328880196044e-05,
+      "loss": 0.3932,
+      "step": 8224
+    },
+    {
+      "epoch": 0.36119740496393954,
+      "grad_norm": 2.125,
+      "learning_rate": 3.569703622555727e-05,
+      "loss": 0.4222,
+      "step": 8226
+    },
+    {
+      "epoch": 0.361285223443143,
+      "grad_norm": 2.359375,
+      "learning_rate": 3.569078282997251e-05,
+      "loss": 0.4291,
+      "step": 8228
+    },
+    {
+      "epoch": 0.3613730419223465,
+      "grad_norm": 1.9609375,
+      "learning_rate": 3.568452861568503e-05,
+      "loss": 0.3827,
+      "step": 8230
+    },
+    {
+      "epoch": 0.36146086040155,
+      "grad_norm": 1.9140625,
+      "learning_rate": 3.5678273583173795e-05,
+      "loss": 0.4059,
+      "step": 8232
+    },
+    {
+      "epoch": 0.3615486788807535,
+      "grad_norm": 2.125,
+      "learning_rate": 3.567201773291781e-05,
+      "loss": 0.4157,
+      "step": 8234
+    },
+    {
+      "epoch": 0.36163649735995695,
+      "grad_norm": 2.375,
+      "learning_rate": 3.5665761065396134e-05,
+      "loss": 0.3946,
+      "step": 8236
+    },
+    {
+      "epoch": 0.3617243158391605,
+      "grad_norm": 2.171875,
+      "learning_rate": 3.565950358108793e-05,
+      "loss": 0.3954,
+      "step": 8238
+    },
+    {
+      "epoch": 0.36181213431836395,
+      "grad_norm": 2.40625,
+      "learning_rate": 3.565324528047238e-05,
+      "loss": 0.4223,
+      "step": 8240
+    },
+    {
+      "epoch": 0.3618999527975674,
+      "grad_norm": 2.09375,
+      "learning_rate": 3.5646986164028765e-05,
+      "loss": 0.4139,
+      "step": 8242
+    },
+    {
+      "epoch": 0.3619877712767709,
+      "grad_norm": 2.515625,
+      "learning_rate": 3.564072623223639e-05,
+      "loss": 0.4005,
+      "step": 8244
+    },
+    {
+      "epoch": 0.3620755897559744,
+      "grad_norm": 2.328125,
+      "learning_rate": 3.5634465485574644e-05,
+      "loss": 0.3918,
+      "step": 8246
+    },
+    {
+      "epoch": 0.3621634082351779,
+      "grad_norm": 2.84375,
+      "learning_rate": 3.5628203924522984e-05,
+      "loss": 0.3848,
+      "step": 8248
+    },
+    {
+      "epoch": 0.36225122671438137,
+      "grad_norm": 2.78125,
+      "learning_rate": 3.562194154956093e-05,
+      "loss": 0.391,
+      "step": 8250
+    },
+    {
+      "epoch": 0.36233904519358484,
+      "grad_norm": 2.171875,
+      "learning_rate": 3.561567836116804e-05,
+      "loss": 0.4067,
+      "step": 8252
+    },
+    {
+      "epoch": 0.36242686367278837,
+      "grad_norm": 1.9921875,
+      "learning_rate": 3.5609414359823956e-05,
+      "loss": 0.4135,
+      "step": 8254
+    },
+    {
+      "epoch": 0.36251468215199184,
+      "grad_norm": 2.0,
+      "learning_rate": 3.5603149546008373e-05,
+      "loss": 0.4034,
+      "step": 8256
+    },
+    {
+      "epoch": 0.3626025006311953,
+      "grad_norm": 1.875,
+      "learning_rate": 3.5596883920201054e-05,
+      "loss": 0.4041,
+      "step": 8258
+    },
+    {
+      "epoch": 0.3626903191103988,
+      "grad_norm": 1.984375,
+      "learning_rate": 3.559061748288183e-05,
+      "loss": 0.4162,
+      "step": 8260
+    },
+    {
+      "epoch": 0.3627781375896023,
+      "grad_norm": 2.078125,
+      "learning_rate": 3.558435023453058e-05,
+      "loss": 0.3935,
+      "step": 8262
+    },
+    {
+      "epoch": 0.3628659560688058,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.557808217562726e-05,
+      "loss": 0.4009,
+      "step": 8264
+    },
+    {
+      "epoch": 0.36295377454800926,
+      "grad_norm": 1.8828125,
+      "learning_rate": 3.557181330665186e-05,
+      "loss": 0.369,
+      "step": 8266
+    },
+    {
+      "epoch": 0.36304159302721273,
+      "grad_norm": 2.046875,
+      "learning_rate": 3.5565543628084466e-05,
+      "loss": 0.3915,
+      "step": 8268
+    },
+    {
+      "epoch": 0.36312941150641626,
+      "grad_norm": 1.9453125,
+      "learning_rate": 3.55592731404052e-05,
+      "loss": 0.3905,
+      "step": 8270
+    },
+    {
+      "epoch": 0.3632172299856197,
+      "grad_norm": 2.078125,
+      "learning_rate": 3.555300184409428e-05,
+      "loss": 0.4033,
+      "step": 8272
+    },
+    {
+      "epoch": 0.3633050484648232,
+      "grad_norm": 2.0625,
+      "learning_rate": 3.554672973963194e-05,
+      "loss": 0.3804,
+      "step": 8274
+    },
+    {
+      "epoch": 0.36339286694402667,
+      "grad_norm": 2.03125,
+      "learning_rate": 3.554045682749851e-05,
+      "loss": 0.388,
+      "step": 8276
+    },
+    {
+      "epoch": 0.3634806854232302,
+      "grad_norm": 2.40625,
+      "learning_rate": 3.553418310817437e-05,
+      "loss": 0.4108,
+      "step": 8278
+    },
+    {
+      "epoch": 0.36356850390243367,
+      "grad_norm": 1.90625,
+      "learning_rate": 3.5527908582139965e-05,
+      "loss": 0.3755,
+      "step": 8280
+    },
+    {
+      "epoch": 0.36365632238163714,
+      "grad_norm": 1.8984375,
+      "learning_rate": 3.5521633249875796e-05,
+      "loss": 0.3783,
+      "step": 8282
+    },
+    {
+      "epoch": 0.36374414086084067,
+      "grad_norm": 2.09375,
+      "learning_rate": 3.551535711186243e-05,
+      "loss": 0.405,
+      "step": 8284
+    },
+    {
+      "epoch": 0.36383195934004414,
+      "grad_norm": 2.171875,
+      "learning_rate": 3.550908016858049e-05,
+      "loss": 0.3899,
+      "step": 8286
+    },
+    {
+      "epoch": 0.3639197778192476,
+      "grad_norm": 2.234375,
+      "learning_rate": 3.550280242051067e-05,
+      "loss": 0.4381,
+      "step": 8288
+    },
+    {
+      "epoch": 0.3640075962984511,
+      "grad_norm": 2.140625,
+      "learning_rate": 3.5496523868133735e-05,
+      "loss": 0.3999,
+      "step": 8290
+    },
+    {
+      "epoch": 0.3640954147776546,
+      "grad_norm": 2.015625,
+      "learning_rate": 3.549024451193048e-05,
+      "loss": 0.4372,
+      "step": 8292
+    },
+    {
+      "epoch": 0.3641832332568581,
+      "grad_norm": 2.0625,
+      "learning_rate": 3.548396435238179e-05,
+      "loss": 0.3942,
+      "step": 8294
+    },
+    {
+      "epoch": 0.36427105173606156,
+      "grad_norm": 2.203125,
+      "learning_rate": 3.547768338996859e-05,
+      "loss": 0.3907,
+      "step": 8296
+    },
+    {
+      "epoch": 0.36435887021526503,
+      "grad_norm": 2.71875,
+      "learning_rate": 3.547140162517189e-05,
+      "loss": 0.3835,
+      "step": 8298
+    },
+    {
+      "epoch": 0.36444668869446856,
+      "grad_norm": 1.84375,
+      "learning_rate": 3.5465119058472736e-05,
+      "loss": 0.4294,
+      "step": 8300
+    },
+    {
+      "epoch": 0.36453450717367203,
+      "grad_norm": 2.015625,
+      "learning_rate": 3.545883569035226e-05,
+      "loss": 0.4001,
+      "step": 8302
+    },
+    {
+      "epoch": 0.3646223256528755,
+      "grad_norm": 1.890625,
+      "learning_rate": 3.545255152129164e-05,
+      "loss": 0.3965,
+      "step": 8304
+    },
+    {
+      "epoch": 0.364710144132079,
+      "grad_norm": 1.859375,
+      "learning_rate": 3.544626655177212e-05,
+      "loss": 0.411,
+      "step": 8306
+    },
+    {
+      "epoch": 0.3647979626112825,
+      "grad_norm": 1.9140625,
+      "learning_rate": 3.5439980782275e-05,
+      "loss": 0.4039,
+      "step": 8308
+    },
+    {
+      "epoch": 0.36488578109048597,
+      "grad_norm": 2.140625,
+      "learning_rate": 3.543369421328165e-05,
+      "loss": 0.4358,
+      "step": 8310
+    },
+    {
+      "epoch": 0.36497359956968944,
+      "grad_norm": 2.171875,
+      "learning_rate": 3.5427406845273506e-05,
+      "loss": 0.4205,
+      "step": 8312
+    },
+    {
+      "epoch": 0.3650614180488929,
+      "grad_norm": 1.9609375,
+      "learning_rate": 3.542111867873203e-05,
+      "loss": 0.391,
+      "step": 8314
+    },
+    {
+      "epoch": 0.36514923652809644,
+      "grad_norm": 1.8984375,
+      "learning_rate": 3.5414829714138795e-05,
+      "loss": 0.4033,
+      "step": 8316
+    },
+    {
+      "epoch": 0.3652370550072999,
+      "grad_norm": 2.046875,
+      "learning_rate": 3.540853995197541e-05,
+      "loss": 0.3773,
+      "step": 8318
+    },
+    {
+      "epoch": 0.3653248734865034,
+      "grad_norm": 2.0625,
+      "learning_rate": 3.540224939272353e-05,
+      "loss": 0.396,
+      "step": 8320
+    },
+    {
+      "epoch": 0.36541269196570686,
+      "grad_norm": 2.0625,
+      "learning_rate": 3.5395958036864896e-05,
+      "loss": 0.3724,
+      "step": 8322
+    },
+    {
+      "epoch": 0.3655005104449104,
+      "grad_norm": 2.046875,
+      "learning_rate": 3.538966588488131e-05,
+      "loss": 0.398,
+      "step": 8324
+    },
+    {
+      "epoch": 0.36558832892411386,
+      "grad_norm": 1.84375,
+      "learning_rate": 3.538337293725462e-05,
+      "loss": 0.3854,
+      "step": 8326
+    },
+    {
+      "epoch": 0.36567614740331733,
+      "grad_norm": 1.9609375,
+      "learning_rate": 3.5377079194466737e-05,
+      "loss": 0.4364,
+      "step": 8328
+    },
+    {
+      "epoch": 0.3657639658825208,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.5370784656999655e-05,
+      "loss": 0.414,
+      "step": 8330
+    },
+    {
+      "epoch": 0.36585178436172433,
+      "grad_norm": 2.34375,
+      "learning_rate": 3.536448932533538e-05,
+      "loss": 0.3896,
+      "step": 8332
+    },
+    {
+      "epoch": 0.3659396028409278,
+      "grad_norm": 1.921875,
+      "learning_rate": 3.5358193199956036e-05,
+      "loss": 0.4054,
+      "step": 8334
+    },
+    {
+      "epoch": 0.3660274213201313,
+      "grad_norm": 2.34375,
+      "learning_rate": 3.5351896281343774e-05,
+      "loss": 0.3983,
+      "step": 8336
+    },
+    {
+      "epoch": 0.3661152397993348,
+      "grad_norm": 1.9453125,
+      "learning_rate": 3.5345598569980814e-05,
+      "loss": 0.4626,
+      "step": 8338
+    },
+    {
+      "epoch": 0.3662030582785383,
+      "grad_norm": 2.375,
+      "learning_rate": 3.5339300066349435e-05,
+      "loss": 0.4138,
+      "step": 8340
+    },
+    {
+      "epoch": 0.36629087675774175,
+      "grad_norm": 1.7890625,
+      "learning_rate": 3.5333000770931986e-05,
+      "loss": 0.4116,
+      "step": 8342
+    },
+    {
+      "epoch": 0.3663786952369452,
+      "grad_norm": 1.9765625,
+      "learning_rate": 3.532670068421085e-05,
+      "loss": 0.4138,
+      "step": 8344
+    },
+    {
+      "epoch": 0.36646651371614875,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.5320399806668506e-05,
+      "loss": 0.4108,
+      "step": 8346
+    },
+    {
+      "epoch": 0.3665543321953522,
+      "grad_norm": 2.328125,
+      "learning_rate": 3.531409813878746e-05,
+      "loss": 0.3859,
+      "step": 8348
+    },
+    {
+      "epoch": 0.3666421506745557,
+      "grad_norm": 2.34375,
+      "learning_rate": 3.5307795681050316e-05,
+      "loss": 0.4094,
+      "step": 8350
+    },
+    {
+      "epoch": 0.36672996915375916,
+      "grad_norm": 2.09375,
+      "learning_rate": 3.530149243393971e-05,
+      "loss": 0.3775,
+      "step": 8352
+    },
+    {
+      "epoch": 0.3668177876329627,
+      "grad_norm": 2.0,
+      "learning_rate": 3.5295188397938336e-05,
+      "loss": 0.4058,
+      "step": 8354
+    },
+    {
+      "epoch": 0.36690560611216616,
+      "grad_norm": 2.078125,
+      "learning_rate": 3.528888357352898e-05,
+      "loss": 0.374,
+      "step": 8356
+    },
+    {
+      "epoch": 0.36699342459136963,
+      "grad_norm": 2.046875,
+      "learning_rate": 3.528257796119443e-05,
+      "loss": 0.3964,
+      "step": 8358
+    },
+    {
+      "epoch": 0.3670812430705731,
+      "grad_norm": 2.140625,
+      "learning_rate": 3.527627156141761e-05,
+      "loss": 0.4265,
+      "step": 8360
+    },
+    {
+      "epoch": 0.36716906154977663,
+      "grad_norm": 1.8671875,
+      "learning_rate": 3.5269964374681454e-05,
+      "loss": 0.3935,
+      "step": 8362
+    },
+    {
+      "epoch": 0.3672568800289801,
+      "grad_norm": 2.0,
+      "learning_rate": 3.526365640146896e-05,
+      "loss": 0.4064,
+      "step": 8364
+    },
+    {
+      "epoch": 0.3673446985081836,
+      "grad_norm": 2.03125,
+      "learning_rate": 3.5257347642263194e-05,
+      "loss": 0.4018,
+      "step": 8366
+    },
+    {
+      "epoch": 0.36743251698738705,
+      "grad_norm": 2.046875,
+      "learning_rate": 3.52510380975473e-05,
+      "loss": 0.405,
+      "step": 8368
+    },
+    {
+      "epoch": 0.3675203354665906,
+      "grad_norm": 2.171875,
+      "learning_rate": 3.524472776780443e-05,
+      "loss": 0.4101,
+      "step": 8370
+    },
+    {
+      "epoch": 0.36760815394579405,
+      "grad_norm": 2.0625,
+      "learning_rate": 3.523841665351787e-05,
+      "loss": 0.4013,
+      "step": 8372
+    },
+    {
+      "epoch": 0.3676959724249975,
+      "grad_norm": 1.90625,
+      "learning_rate": 3.5232104755170905e-05,
+      "loss": 0.4238,
+      "step": 8374
+    },
+    {
+      "epoch": 0.367783790904201,
+      "grad_norm": 2.765625,
+      "learning_rate": 3.522579207324689e-05,
+      "loss": 0.3948,
+      "step": 8376
+    },
+    {
+      "epoch": 0.3678716093834045,
+      "grad_norm": 1.9921875,
+      "learning_rate": 3.5219478608229284e-05,
+      "loss": 0.4009,
+      "step": 8378
+    },
+    {
+      "epoch": 0.367959427862608,
+      "grad_norm": 2.0625,
+      "learning_rate": 3.5213164360601555e-05,
+      "loss": 0.4123,
+      "step": 8380
+    },
+    {
+      "epoch": 0.36804724634181146,
+      "grad_norm": 2.015625,
+      "learning_rate": 3.5206849330847244e-05,
+      "loss": 0.3838,
+      "step": 8382
+    },
+    {
+      "epoch": 0.36813506482101493,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.520053351944996e-05,
+      "loss": 0.4118,
+      "step": 8384
+    },
+    {
+      "epoch": 0.36822288330021846,
+      "grad_norm": 2.34375,
+      "learning_rate": 3.5194216926893395e-05,
+      "loss": 0.4044,
+      "step": 8386
+    },
+    {
+      "epoch": 0.36831070177942193,
+      "grad_norm": 2.03125,
+      "learning_rate": 3.5187899553661236e-05,
+      "loss": 0.4166,
+      "step": 8388
+    },
+    {
+      "epoch": 0.3683985202586254,
+      "grad_norm": 2.03125,
+      "learning_rate": 3.518158140023729e-05,
+      "loss": 0.4072,
+      "step": 8390
+    },
+    {
+      "epoch": 0.36848633873782893,
+      "grad_norm": 2.015625,
+      "learning_rate": 3.5175262467105404e-05,
+      "loss": 0.4168,
+      "step": 8392
+    },
+    {
+      "epoch": 0.3685741572170324,
+      "grad_norm": 2.28125,
+      "learning_rate": 3.5168942754749476e-05,
+      "loss": 0.3971,
+      "step": 8394
+    },
+    {
+      "epoch": 0.3686619756962359,
+      "grad_norm": 2.203125,
+      "learning_rate": 3.516262226365347e-05,
+      "loss": 0.3804,
+      "step": 8396
+    },
+    {
+      "epoch": 0.36874979417543935,
+      "grad_norm": 2.03125,
+      "learning_rate": 3.515630099430142e-05,
+      "loss": 0.4292,
+      "step": 8398
+    },
+    {
+      "epoch": 0.3688376126546429,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.5149978947177396e-05,
+      "loss": 0.4162,
+      "step": 8400
+    },
+    {
+      "epoch": 0.36892543113384635,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.514365612276557e-05,
+      "loss": 0.4324,
+      "step": 8402
+    },
+    {
+      "epoch": 0.3690132496130498,
+      "grad_norm": 2.03125,
+      "learning_rate": 3.5137332521550116e-05,
+      "loss": 0.4021,
+      "step": 8404
+    },
+    {
+      "epoch": 0.3691010680922533,
+      "grad_norm": 1.8671875,
+      "learning_rate": 3.513100814401531e-05,
+      "loss": 0.4144,
+      "step": 8406
+    },
+    {
+      "epoch": 0.3691888865714568,
+      "grad_norm": 1.9765625,
+      "learning_rate": 3.512468299064546e-05,
+      "loss": 0.3933,
+      "step": 8408
+    },
+    {
+      "epoch": 0.3692767050506603,
+      "grad_norm": 2.0,
+      "learning_rate": 3.5118357061924974e-05,
+      "loss": 0.3806,
+      "step": 8410
+    },
+    {
+      "epoch": 0.36936452352986376,
+      "grad_norm": 2.4375,
+      "learning_rate": 3.511203035833827e-05,
+      "loss": 0.3926,
+      "step": 8412
+    },
+    {
+      "epoch": 0.36945234200906724,
+      "grad_norm": 2.3125,
+      "learning_rate": 3.5105702880369864e-05,
+      "loss": 0.4062,
+      "step": 8414
+    },
+    {
+      "epoch": 0.36954016048827076,
+      "grad_norm": 2.578125,
+      "learning_rate": 3.509937462850431e-05,
+      "loss": 0.4053,
+      "step": 8416
+    },
+    {
+      "epoch": 0.36962797896747424,
+      "grad_norm": 2.390625,
+      "learning_rate": 3.509304560322622e-05,
+      "loss": 0.4269,
+      "step": 8418
+    },
+    {
+      "epoch": 0.3697157974466777,
+      "grad_norm": 2.25,
+      "learning_rate": 3.5086715805020274e-05,
+      "loss": 0.38,
+      "step": 8420
+    },
+    {
+      "epoch": 0.3698036159258812,
+      "grad_norm": 2.09375,
+      "learning_rate": 3.508038523437122e-05,
+      "loss": 0.3872,
+      "step": 8422
+    },
+    {
+      "epoch": 0.3698914344050847,
+      "grad_norm": 2.171875,
+      "learning_rate": 3.5074053891763844e-05,
+      "loss": 0.3803,
+      "step": 8424
+    },
+    {
+      "epoch": 0.3699792528842882,
+      "grad_norm": 2.203125,
+      "learning_rate": 3.506772177768301e-05,
+      "loss": 0.4023,
+      "step": 8426
+    },
+    {
+      "epoch": 0.37006707136349165,
+      "grad_norm": 2.046875,
+      "learning_rate": 3.506138889261364e-05,
+      "loss": 0.4042,
+      "step": 8428
+    },
+    {
+      "epoch": 0.3701548898426951,
+      "grad_norm": 1.8984375,
+      "learning_rate": 3.505505523704068e-05,
+      "loss": 0.3977,
+      "step": 8430
+    },
+    {
+      "epoch": 0.37024270832189865,
+      "grad_norm": 2.046875,
+      "learning_rate": 3.5048720811449185e-05,
+      "loss": 0.3787,
+      "step": 8432
+    },
+    {
+      "epoch": 0.3703305268011021,
+      "grad_norm": 2.03125,
+      "learning_rate": 3.504238561632424e-05,
+      "loss": 0.3843,
+      "step": 8434
+    },
+    {
+      "epoch": 0.3704183452803056,
+      "grad_norm": 2.0,
+      "learning_rate": 3.5036049652151e-05,
+      "loss": 0.3983,
+      "step": 8436
+    },
+    {
+      "epoch": 0.3705061637595091,
+      "grad_norm": 1.984375,
+      "learning_rate": 3.5029712919414664e-05,
+      "loss": 0.4091,
+      "step": 8438
+    },
+    {
+      "epoch": 0.3705939822387126,
+      "grad_norm": 1.9296875,
+      "learning_rate": 3.5023375418600524e-05,
+      "loss": 0.3897,
+      "step": 8440
+    },
+    {
+      "epoch": 0.37068180071791607,
+      "grad_norm": 1.9921875,
+      "learning_rate": 3.501703715019388e-05,
+      "loss": 0.4017,
+      "step": 8442
+    },
+    {
+      "epoch": 0.37076961919711954,
+      "grad_norm": 1.953125,
+      "learning_rate": 3.501069811468013e-05,
+      "loss": 0.4036,
+      "step": 8444
+    },
+    {
+      "epoch": 0.37085743767632307,
+      "grad_norm": 2.0625,
+      "learning_rate": 3.5004358312544714e-05,
+      "loss": 0.3832,
+      "step": 8446
+    },
+    {
+      "epoch": 0.37094525615552654,
+      "grad_norm": 1.8203125,
+      "learning_rate": 3.499801774427315e-05,
+      "loss": 0.4105,
+      "step": 8448
+    },
+    {
+      "epoch": 0.37103307463473,
+      "grad_norm": 2.3125,
+      "learning_rate": 3.499167641035099e-05,
+      "loss": 0.4005,
+      "step": 8450
+    },
+    {
+      "epoch": 0.3711208931139335,
+      "grad_norm": 2.09375,
+      "learning_rate": 3.498533431126386e-05,
+      "loss": 0.3815,
+      "step": 8452
+    },
+    {
+      "epoch": 0.371208711593137,
+      "grad_norm": 2.203125,
+      "learning_rate": 3.497899144749742e-05,
+      "loss": 0.4332,
+      "step": 8454
+    },
+    {
+      "epoch": 0.3712965300723405,
+      "grad_norm": 2.390625,
+      "learning_rate": 3.497264781953743e-05,
+      "loss": 0.4117,
+      "step": 8456
+    },
+    {
+      "epoch": 0.37138434855154395,
+      "grad_norm": 2.3125,
+      "learning_rate": 3.496630342786968e-05,
+      "loss": 0.3979,
+      "step": 8458
+    },
+    {
+      "epoch": 0.3714721670307474,
+      "grad_norm": 2.3125,
+      "learning_rate": 3.495995827298002e-05,
+      "loss": 0.4247,
+      "step": 8460
+    },
+    {
+      "epoch": 0.37155998550995095,
+      "grad_norm": 1.9453125,
+      "learning_rate": 3.4953612355354373e-05,
+      "loss": 0.4288,
+      "step": 8462
+    },
+    {
+      "epoch": 0.3716478039891544,
+      "grad_norm": 1.890625,
+      "learning_rate": 3.494726567547871e-05,
+      "loss": 0.4028,
+      "step": 8464
+    },
+    {
+      "epoch": 0.3717356224683579,
+      "grad_norm": 2.265625,
+      "learning_rate": 3.4940918233839056e-05,
+      "loss": 0.379,
+      "step": 8466
+    },
+    {
+      "epoch": 0.37182344094756137,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.4934570030921494e-05,
+      "loss": 0.4114,
+      "step": 8468
+    },
+    {
+      "epoch": 0.3719112594267649,
+      "grad_norm": 2.078125,
+      "learning_rate": 3.492822106721217e-05,
+      "loss": 0.3756,
+      "step": 8470
+    },
+    {
+      "epoch": 0.37199907790596837,
+      "grad_norm": 2.296875,
+      "learning_rate": 3.492187134319731e-05,
+      "loss": 0.4375,
+      "step": 8472
+    },
+    {
+      "epoch": 0.37208689638517184,
+      "grad_norm": 2.0625,
+      "learning_rate": 3.491552085936316e-05,
+      "loss": 0.4181,
+      "step": 8474
+    },
+    {
+      "epoch": 0.3721747148643753,
+      "grad_norm": 2.53125,
+      "learning_rate": 3.4909169616196055e-05,
+      "loss": 0.3895,
+      "step": 8476
+    },
+    {
+      "epoch": 0.37226253334357884,
+      "grad_norm": 2.109375,
+      "learning_rate": 3.490281761418236e-05,
+      "loss": 0.3932,
+      "step": 8478
+    },
+    {
+      "epoch": 0.3723503518227823,
+      "grad_norm": 2.046875,
+      "learning_rate": 3.489646485380851e-05,
+      "loss": 0.4005,
+      "step": 8480
+    },
+    {
+      "epoch": 0.3724381703019858,
+      "grad_norm": 1.78125,
+      "learning_rate": 3.4890111335561016e-05,
+      "loss": 0.3688,
+      "step": 8482
+    },
+    {
+      "epoch": 0.37252598878118925,
+      "grad_norm": 2.0625,
+      "learning_rate": 3.488375705992642e-05,
+      "loss": 0.3811,
+      "step": 8484
+    },
+    {
+      "epoch": 0.3726138072603928,
+      "grad_norm": 2.109375,
+      "learning_rate": 3.487740202739134e-05,
+      "loss": 0.4228,
+      "step": 8486
+    },
+    {
+      "epoch": 0.37270162573959625,
+      "grad_norm": 2.015625,
+      "learning_rate": 3.487104623844245e-05,
+      "loss": 0.3903,
+      "step": 8488
+    },
+    {
+      "epoch": 0.3727894442187997,
+      "grad_norm": 2.0625,
+      "learning_rate": 3.486468969356647e-05,
+      "loss": 0.3835,
+      "step": 8490
+    },
+    {
+      "epoch": 0.37287726269800325,
+      "grad_norm": 1.984375,
+      "learning_rate": 3.485833239325019e-05,
+      "loss": 0.4037,
+      "step": 8492
+    },
+    {
+      "epoch": 0.3729650811772067,
+      "grad_norm": 2.375,
+      "learning_rate": 3.485197433798045e-05,
+      "loss": 0.372,
+      "step": 8494
+    },
+    {
+      "epoch": 0.3730528996564102,
+      "grad_norm": 2.25,
+      "learning_rate": 3.484561552824416e-05,
+      "loss": 0.3698,
+      "step": 8496
+    },
+    {
+      "epoch": 0.37314071813561367,
+      "grad_norm": 1.984375,
+      "learning_rate": 3.483925596452826e-05,
+      "loss": 0.4107,
+      "step": 8498
+    },
+    {
+      "epoch": 0.3732285366148172,
+      "grad_norm": 2.03125,
+      "learning_rate": 3.4832895647319786e-05,
+      "loss": 0.4162,
+      "step": 8500
+    },
+    {
+      "epoch": 0.37331635509402067,
+      "grad_norm": 2.171875,
+      "learning_rate": 3.482653457710581e-05,
+      "loss": 0.3997,
+      "step": 8502
+    },
+    {
+      "epoch": 0.37340417357322414,
+      "grad_norm": 2.140625,
+      "learning_rate": 3.482017275437346e-05,
+      "loss": 0.4082,
+      "step": 8504
+    },
+    {
+      "epoch": 0.3734919920524276,
+      "grad_norm": 2.203125,
+      "learning_rate": 3.481381017960992e-05,
+      "loss": 0.3791,
+      "step": 8506
+    },
+    {
+      "epoch": 0.37357981053163114,
+      "grad_norm": 2.09375,
+      "learning_rate": 3.480744685330244e-05,
+      "loss": 0.3819,
+      "step": 8508
+    },
+    {
+      "epoch": 0.3736676290108346,
+      "grad_norm": 2.296875,
+      "learning_rate": 3.480108277593834e-05,
+      "loss": 0.4309,
+      "step": 8510
+    },
+    {
+      "epoch": 0.3737554474900381,
+      "grad_norm": 1.953125,
+      "learning_rate": 3.4794717948004974e-05,
+      "loss": 0.3988,
+      "step": 8512
+    },
+    {
+      "epoch": 0.37384326596924156,
+      "grad_norm": 1.828125,
+      "learning_rate": 3.478835236998976e-05,
+      "loss": 0.4027,
+      "step": 8514
+    },
+    {
+      "epoch": 0.3739310844484451,
+      "grad_norm": 1.8046875,
+      "learning_rate": 3.4781986042380164e-05,
+      "loss": 0.3995,
+      "step": 8516
+    },
+    {
+      "epoch": 0.37401890292764856,
+      "grad_norm": 1.9609375,
+      "learning_rate": 3.4775618965663736e-05,
+      "loss": 0.4285,
+      "step": 8518
+    },
+    {
+      "epoch": 0.374106721406852,
+      "grad_norm": 2.015625,
+      "learning_rate": 3.476925114032806e-05,
+      "loss": 0.3945,
+      "step": 8520
+    },
+    {
+      "epoch": 0.3741945398860555,
+      "grad_norm": 1.921875,
+      "learning_rate": 3.47628825668608e-05,
+      "loss": 0.3842,
+      "step": 8522
+    },
+    {
+      "epoch": 0.374282358365259,
+      "grad_norm": 1.9765625,
+      "learning_rate": 3.475651324574965e-05,
+      "loss": 0.4019,
+      "step": 8524
+    },
+    {
+      "epoch": 0.3743701768444625,
+      "grad_norm": 2.0,
+      "learning_rate": 3.4750143177482366e-05,
+      "loss": 0.4039,
+      "step": 8526
+    },
+    {
+      "epoch": 0.37445799532366597,
+      "grad_norm": 2.25,
+      "learning_rate": 3.47437723625468e-05,
+      "loss": 0.3881,
+      "step": 8528
+    },
+    {
+      "epoch": 0.37454581380286944,
+      "grad_norm": 2.296875,
+      "learning_rate": 3.47374008014308e-05,
+      "loss": 0.4122,
+      "step": 8530
+    },
+    {
+      "epoch": 0.37463363228207297,
+      "grad_norm": 1.8671875,
+      "learning_rate": 3.473102849462231e-05,
+      "loss": 0.4023,
+      "step": 8532
+    },
+    {
+      "epoch": 0.37472145076127644,
+      "grad_norm": 2.390625,
+      "learning_rate": 3.472465544260932e-05,
+      "loss": 0.406,
+      "step": 8534
+    },
+    {
+      "epoch": 0.3748092692404799,
+      "grad_norm": 2.328125,
+      "learning_rate": 3.471828164587989e-05,
+      "loss": 0.3963,
+      "step": 8536
+    },
+    {
+      "epoch": 0.37489708771968344,
+      "grad_norm": 1.9296875,
+      "learning_rate": 3.471190710492213e-05,
+      "loss": 0.3803,
+      "step": 8538
+    },
+    {
+      "epoch": 0.3749849061988869,
+      "grad_norm": 1.953125,
+      "learning_rate": 3.470553182022419e-05,
+      "loss": 0.4057,
+      "step": 8540
+    },
+    {
+      "epoch": 0.3750727246780904,
+      "grad_norm": 2.125,
+      "learning_rate": 3.46991557922743e-05,
+      "loss": 0.384,
+      "step": 8542
+    },
+    {
+      "epoch": 0.37516054315729386,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.4692779021560726e-05,
+      "loss": 0.3787,
+      "step": 8544
+    },
+    {
+      "epoch": 0.3752483616364974,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.468640150857181e-05,
+      "loss": 0.3673,
+      "step": 8546
+    },
+    {
+      "epoch": 0.37533618011570086,
+      "grad_norm": 1.8359375,
+      "learning_rate": 3.4680023253795956e-05,
+      "loss": 0.3796,
+      "step": 8548
+    },
+    {
+      "epoch": 0.37542399859490433,
+      "grad_norm": 1.8125,
+      "learning_rate": 3.46736442577216e-05,
+      "loss": 0.3703,
+      "step": 8550
+    },
+    {
+      "epoch": 0.3755118170741078,
+      "grad_norm": 1.8125,
+      "learning_rate": 3.466726452083724e-05,
+      "loss": 0.4013,
+      "step": 8552
+    },
+    {
+      "epoch": 0.37559963555331133,
+      "grad_norm": 2.28125,
+      "learning_rate": 3.466088404363145e-05,
+      "loss": 0.3854,
+      "step": 8554
+    },
+    {
+      "epoch": 0.3756874540325148,
+      "grad_norm": 2.09375,
+      "learning_rate": 3.465450282659285e-05,
+      "loss": 0.4204,
+      "step": 8556
+    },
+    {
+      "epoch": 0.3757752725117183,
+      "grad_norm": 2.078125,
+      "learning_rate": 3.464812087021009e-05,
+      "loss": 0.412,
+      "step": 8558
+    },
+    {
+      "epoch": 0.37586309099092174,
+      "grad_norm": 1.984375,
+      "learning_rate": 3.4641738174971936e-05,
+      "loss": 0.4043,
+      "step": 8560
+    },
+    {
+      "epoch": 0.37595090947012527,
+      "grad_norm": 2.046875,
+      "learning_rate": 3.463535474136716e-05,
+      "loss": 0.3919,
+      "step": 8562
+    },
+    {
+      "epoch": 0.37603872794932874,
+      "grad_norm": 2.25,
+      "learning_rate": 3.462897056988461e-05,
+      "loss": 0.389,
+      "step": 8564
+    },
+    {
+      "epoch": 0.3761265464285322,
+      "grad_norm": 1.859375,
+      "learning_rate": 3.462258566101318e-05,
+      "loss": 0.3905,
+      "step": 8566
+    },
+    {
+      "epoch": 0.3762143649077357,
+      "grad_norm": 2.6875,
+      "learning_rate": 3.461620001524183e-05,
+      "loss": 0.4025,
+      "step": 8568
+    },
+    {
+      "epoch": 0.3763021833869392,
+      "grad_norm": 2.046875,
+      "learning_rate": 3.460981363305959e-05,
+      "loss": 0.4231,
+      "step": 8570
+    },
+    {
+      "epoch": 0.3763900018661427,
+      "grad_norm": 2.28125,
+      "learning_rate": 3.460342651495551e-05,
+      "loss": 0.423,
+      "step": 8572
+    },
+    {
+      "epoch": 0.37647782034534616,
+      "grad_norm": 1.984375,
+      "learning_rate": 3.459703866141872e-05,
+      "loss": 0.396,
+      "step": 8574
+    },
+    {
+      "epoch": 0.37656563882454963,
+      "grad_norm": 2.046875,
+      "learning_rate": 3.459065007293842e-05,
+      "loss": 0.4062,
+      "step": 8576
+    },
+    {
+      "epoch": 0.37665345730375316,
+      "grad_norm": 1.9296875,
+      "learning_rate": 3.458426075000383e-05,
+      "loss": 0.409,
+      "step": 8578
+    },
+    {
+      "epoch": 0.37674127578295663,
+      "grad_norm": 1.875,
+      "learning_rate": 3.4577870693104256e-05,
+      "loss": 0.4096,
+      "step": 8580
+    },
+    {
+      "epoch": 0.3768290942621601,
+      "grad_norm": 1.96875,
+      "learning_rate": 3.4571479902729045e-05,
+      "loss": 0.4013,
+      "step": 8582
+    },
+    {
+      "epoch": 0.3769169127413636,
+      "grad_norm": 2.078125,
+      "learning_rate": 3.456508837936762e-05,
+      "loss": 0.4027,
+      "step": 8584
+    },
+    {
+      "epoch": 0.3770047312205671,
+      "grad_norm": 1.9140625,
+      "learning_rate": 3.4558696123509426e-05,
+      "loss": 0.4047,
+      "step": 8586
+    },
+    {
+      "epoch": 0.3770925496997706,
+      "grad_norm": 1.8984375,
+      "learning_rate": 3.455230313564399e-05,
+      "loss": 0.3918,
+      "step": 8588
+    },
+    {
+      "epoch": 0.37718036817897405,
+      "grad_norm": 2.078125,
+      "learning_rate": 3.45459094162609e-05,
+      "loss": 0.3935,
+      "step": 8590
+    },
+    {
+      "epoch": 0.3772681866581776,
+      "grad_norm": 1.9765625,
+      "learning_rate": 3.453951496584977e-05,
+      "loss": 0.4381,
+      "step": 8592
+    },
+    {
+      "epoch": 0.37735600513738105,
+      "grad_norm": 1.9296875,
+      "learning_rate": 3.45331197849003e-05,
+      "loss": 0.3799,
+      "step": 8594
+    },
+    {
+      "epoch": 0.3774438236165845,
+      "grad_norm": 2.125,
+      "learning_rate": 3.452672387390223e-05,
+      "loss": 0.3954,
+      "step": 8596
+    },
+    {
+      "epoch": 0.377531642095788,
+      "grad_norm": 2.203125,
+      "learning_rate": 3.452032723334536e-05,
+      "loss": 0.4193,
+      "step": 8598
+    },
+    {
+      "epoch": 0.3776194605749915,
+      "grad_norm": 2.234375,
+      "learning_rate": 3.451392986371955e-05,
+      "loss": 0.3764,
+      "step": 8600
+    },
+    {
+      "epoch": 0.377707279054195,
+      "grad_norm": 1.90625,
+      "learning_rate": 3.450753176551472e-05,
+      "loss": 0.4466,
+      "step": 8602
+    },
+    {
+      "epoch": 0.37779509753339846,
+      "grad_norm": 2.0625,
+      "learning_rate": 3.4501132939220816e-05,
+      "loss": 0.3867,
+      "step": 8604
+    },
+    {
+      "epoch": 0.37788291601260193,
+      "grad_norm": 1.7890625,
+      "learning_rate": 3.4494733385327875e-05,
+      "loss": 0.4123,
+      "step": 8606
+    },
+    {
+      "epoch": 0.37797073449180546,
+      "grad_norm": 1.9765625,
+      "learning_rate": 3.4488333104325975e-05,
+      "loss": 0.3987,
+      "step": 8608
+    },
+    {
+      "epoch": 0.37805855297100893,
+      "grad_norm": 2.03125,
+      "learning_rate": 3.448193209670526e-05,
+      "loss": 0.3798,
+      "step": 8610
+    },
+    {
+      "epoch": 0.3781463714502124,
+      "grad_norm": 2.03125,
+      "learning_rate": 3.44755303629559e-05,
+      "loss": 0.3834,
+      "step": 8612
+    },
+    {
+      "epoch": 0.3782341899294159,
+      "grad_norm": 1.9453125,
+      "learning_rate": 3.446912790356817e-05,
+      "loss": 0.3721,
+      "step": 8614
+    },
+    {
+      "epoch": 0.3783220084086194,
+      "grad_norm": 2.109375,
+      "learning_rate": 3.446272471903235e-05,
+      "loss": 0.3949,
+      "step": 8616
+    },
+    {
+      "epoch": 0.3784098268878229,
+      "grad_norm": 2.296875,
+      "learning_rate": 3.445632080983879e-05,
+      "loss": 0.389,
+      "step": 8618
+    },
+    {
+      "epoch": 0.37849764536702635,
+      "grad_norm": 2.5625,
+      "learning_rate": 3.444991617647792e-05,
+      "loss": 0.3948,
+      "step": 8620
+    },
+    {
+      "epoch": 0.3785854638462298,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.44435108194402e-05,
+      "loss": 0.4126,
+      "step": 8622
+    },
+    {
+      "epoch": 0.37867328232543335,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.443710473921617e-05,
+      "loss": 0.4307,
+      "step": 8624
+    },
+    {
+      "epoch": 0.3787611008046368,
+      "grad_norm": 2.265625,
+      "learning_rate": 3.44306979362964e-05,
+      "loss": 0.3778,
+      "step": 8626
+    },
+    {
+      "epoch": 0.3788489192838403,
+      "grad_norm": 2.34375,
+      "learning_rate": 3.4424290411171505e-05,
+      "loss": 0.4118,
+      "step": 8628
+    },
+    {
+      "epoch": 0.37893673776304376,
+      "grad_norm": 2.359375,
+      "learning_rate": 3.44178821643322e-05,
+      "loss": 0.3992,
+      "step": 8630
+    },
+    {
+      "epoch": 0.3790245562422473,
+      "grad_norm": 2.53125,
+      "learning_rate": 3.441147319626922e-05,
+      "loss": 0.4063,
+      "step": 8632
+    },
+    {
+      "epoch": 0.37911237472145076,
+      "grad_norm": 1.890625,
+      "learning_rate": 3.440506350747337e-05,
+      "loss": 0.3811,
+      "step": 8634
+    },
+    {
+      "epoch": 0.37920019320065423,
+      "grad_norm": 2.078125,
+      "learning_rate": 3.43986530984355e-05,
+      "loss": 0.3884,
+      "step": 8636
+    },
+    {
+      "epoch": 0.37928801167985776,
+      "grad_norm": 2.078125,
+      "learning_rate": 3.439224196964652e-05,
+      "loss": 0.3733,
+      "step": 8638
+    },
+    {
+      "epoch": 0.37937583015906123,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.43858301215974e-05,
+      "loss": 0.3773,
+      "step": 8640
+    },
+    {
+      "epoch": 0.3794636486382647,
+      "grad_norm": 2.046875,
+      "learning_rate": 3.437941755477916e-05,
+      "loss": 0.4134,
+      "step": 8642
+    },
+    {
+      "epoch": 0.3795514671174682,
+      "grad_norm": 1.890625,
+      "learning_rate": 3.437300426968287e-05,
+      "loss": 0.3893,
+      "step": 8644
+    },
+    {
+      "epoch": 0.3796392855966717,
+      "grad_norm": 1.9921875,
+      "learning_rate": 3.436659026679967e-05,
+      "loss": 0.4045,
+      "step": 8646
+    },
+    {
+      "epoch": 0.3797271040758752,
+      "grad_norm": 1.9921875,
+      "learning_rate": 3.436017554662074e-05,
+      "loss": 0.4129,
+      "step": 8648
+    },
+    {
+      "epoch": 0.37981492255507865,
+      "grad_norm": 2.140625,
+      "learning_rate": 3.4353760109637336e-05,
+      "loss": 0.4034,
+      "step": 8650
+    },
+    {
+      "epoch": 0.3799027410342821,
+      "grad_norm": 2.125,
+      "learning_rate": 3.4347343956340726e-05,
+      "loss": 0.3959,
+      "step": 8652
+    },
+    {
+      "epoch": 0.37999055951348565,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.434092708722228e-05,
+      "loss": 0.4124,
+      "step": 8654
+    },
+    {
+      "epoch": 0.3800783779926891,
+      "grad_norm": 1.9375,
+      "learning_rate": 3.43345095027734e-05,
+      "loss": 0.4073,
+      "step": 8656
+    },
+    {
+      "epoch": 0.3801661964718926,
+      "grad_norm": 2.140625,
+      "learning_rate": 3.432809120348553e-05,
+      "loss": 0.3992,
+      "step": 8658
+    },
+    {
+      "epoch": 0.38025401495109606,
+      "grad_norm": 1.9296875,
+      "learning_rate": 3.432167218985022e-05,
+      "loss": 0.4143,
+      "step": 8660
+    },
+    {
+      "epoch": 0.3803418334302996,
+      "grad_norm": 1.890625,
+      "learning_rate": 3.4315252462359015e-05,
+      "loss": 0.4135,
+      "step": 8662
+    },
+    {
+      "epoch": 0.38042965190950306,
+      "grad_norm": 2.109375,
+      "learning_rate": 3.4308832021503544e-05,
+      "loss": 0.3911,
+      "step": 8664
+    },
+    {
+      "epoch": 0.38051747038870654,
+      "grad_norm": 2.078125,
+      "learning_rate": 3.430241086777548e-05,
+      "loss": 0.4136,
+      "step": 8666
+    },
+    {
+      "epoch": 0.38060528886791,
+      "grad_norm": 2.25,
+      "learning_rate": 3.429598900166656e-05,
+      "loss": 0.3959,
+      "step": 8668
+    },
+    {
+      "epoch": 0.38069310734711354,
+      "grad_norm": 1.7890625,
+      "learning_rate": 3.428956642366857e-05,
+      "loss": 0.3828,
+      "step": 8670
+    },
+    {
+      "epoch": 0.380780925826317,
+      "grad_norm": 1.9921875,
+      "learning_rate": 3.4283143134273365e-05,
+      "loss": 0.414,
+      "step": 8672
+    },
+    {
+      "epoch": 0.3808687443055205,
+      "grad_norm": 1.921875,
+      "learning_rate": 3.427671913397283e-05,
+      "loss": 0.4011,
+      "step": 8674
+    },
+    {
+      "epoch": 0.38095656278472395,
+      "grad_norm": 2.0625,
+      "learning_rate": 3.427029442325893e-05,
+      "loss": 0.37,
+      "step": 8676
+    },
+    {
+      "epoch": 0.3810443812639275,
+      "grad_norm": 2.078125,
+      "learning_rate": 3.426386900262365e-05,
+      "loss": 0.4073,
+      "step": 8678
+    },
+    {
+      "epoch": 0.38113219974313095,
+      "grad_norm": 2.5625,
+      "learning_rate": 3.425744287255907e-05,
+      "loss": 0.4084,
+      "step": 8680
+    },
+    {
+      "epoch": 0.3812200182223344,
+      "grad_norm": 1.9609375,
+      "learning_rate": 3.425101603355728e-05,
+      "loss": 0.4015,
+      "step": 8682
+    },
+    {
+      "epoch": 0.3813078367015379,
+      "grad_norm": 1.8515625,
+      "learning_rate": 3.4244588486110475e-05,
+      "loss": 0.3831,
+      "step": 8684
+    },
+    {
+      "epoch": 0.3813956551807414,
+      "grad_norm": 2.03125,
+      "learning_rate": 3.423816023071087e-05,
+      "loss": 0.4229,
+      "step": 8686
+    },
+    {
+      "epoch": 0.3814834736599449,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.423173126785073e-05,
+      "loss": 0.4307,
+      "step": 8688
+    },
+    {
+      "epoch": 0.38157129213914837,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.42253015980224e-05,
+      "loss": 0.4027,
+      "step": 8690
+    },
+    {
+      "epoch": 0.3816591106183519,
+      "grad_norm": 1.9765625,
+      "learning_rate": 3.4218871221718266e-05,
+      "loss": 0.4117,
+      "step": 8692
+    },
+    {
+      "epoch": 0.38174692909755537,
+      "grad_norm": 2.03125,
+      "learning_rate": 3.4212440139430765e-05,
+      "loss": 0.362,
+      "step": 8694
+    },
+    {
+      "epoch": 0.38183474757675884,
+      "grad_norm": 1.9296875,
+      "learning_rate": 3.420600835165239e-05,
+      "loss": 0.396,
+      "step": 8696
+    },
+    {
+      "epoch": 0.3819225660559623,
+      "grad_norm": 2.0625,
+      "learning_rate": 3.419957585887568e-05,
+      "loss": 0.3867,
+      "step": 8698
+    },
+    {
+      "epoch": 0.38201038453516584,
+      "grad_norm": 1.9140625,
+      "learning_rate": 3.4193142661593255e-05,
+      "loss": 0.4201,
+      "step": 8700
+    },
+    {
+      "epoch": 0.3820982030143693,
+      "grad_norm": 2.34375,
+      "learning_rate": 3.418670876029776e-05,
+      "loss": 0.3989,
+      "step": 8702
+    },
+    {
+      "epoch": 0.3821860214935728,
+      "grad_norm": 1.765625,
+      "learning_rate": 3.41802741554819e-05,
+      "loss": 0.3829,
+      "step": 8704
+    },
+    {
+      "epoch": 0.38227383997277625,
+      "grad_norm": 1.8671875,
+      "learning_rate": 3.417383884763845e-05,
+      "loss": 0.4079,
+      "step": 8706
+    },
+    {
+      "epoch": 0.3823616584519798,
+      "grad_norm": 2.453125,
+      "learning_rate": 3.416740283726022e-05,
+      "loss": 0.396,
+      "step": 8708
+    },
+    {
+      "epoch": 0.38244947693118325,
+      "grad_norm": 1.9609375,
+      "learning_rate": 3.416096612484008e-05,
+      "loss": 0.3813,
+      "step": 8710
+    },
+    {
+      "epoch": 0.3825372954103867,
+      "grad_norm": 2.140625,
+      "learning_rate": 3.415452871087097e-05,
+      "loss": 0.4056,
+      "step": 8712
+    },
+    {
+      "epoch": 0.3826251138895902,
+      "grad_norm": 2.03125,
+      "learning_rate": 3.414809059584585e-05,
+      "loss": 0.3915,
+      "step": 8714
+    },
+    {
+      "epoch": 0.3827129323687937,
+      "grad_norm": 1.96875,
+      "learning_rate": 3.414165178025775e-05,
+      "loss": 0.3685,
+      "step": 8716
+    },
+    {
+      "epoch": 0.3828007508479972,
+      "grad_norm": 2.65625,
+      "learning_rate": 3.413521226459977e-05,
+      "loss": 0.4022,
+      "step": 8718
+    },
+    {
+      "epoch": 0.38288856932720067,
+      "grad_norm": 2.640625,
+      "learning_rate": 3.412877204936505e-05,
+      "loss": 0.4032,
+      "step": 8720
+    },
+    {
+      "epoch": 0.38297638780640414,
+      "grad_norm": 2.328125,
+      "learning_rate": 3.412233113504677e-05,
+      "loss": 0.3842,
+      "step": 8722
+    },
+    {
+      "epoch": 0.38306420628560767,
+      "grad_norm": 2.125,
+      "learning_rate": 3.41158895221382e-05,
+      "loss": 0.3474,
+      "step": 8724
+    },
+    {
+      "epoch": 0.38315202476481114,
+      "grad_norm": 2.015625,
+      "learning_rate": 3.4109447211132616e-05,
+      "loss": 0.38,
+      "step": 8726
+    },
+    {
+      "epoch": 0.3832398432440146,
+      "grad_norm": 1.9765625,
+      "learning_rate": 3.410300420252338e-05,
+      "loss": 0.3728,
+      "step": 8728
+    },
+    {
+      "epoch": 0.3833276617232181,
+      "grad_norm": 2.015625,
+      "learning_rate": 3.40965604968039e-05,
+      "loss": 0.3944,
+      "step": 8730
+    },
+    {
+      "epoch": 0.3834154802024216,
+      "grad_norm": 2.515625,
+      "learning_rate": 3.409011609446763e-05,
+      "loss": 0.4302,
+      "step": 8732
+    },
+    {
+      "epoch": 0.3835032986816251,
+      "grad_norm": 2.140625,
+      "learning_rate": 3.40836709960081e-05,
+      "loss": 0.4424,
+      "step": 8734
+    },
+    {
+      "epoch": 0.38359111716082855,
+      "grad_norm": 2.140625,
+      "learning_rate": 3.407722520191887e-05,
+      "loss": 0.4114,
+      "step": 8736
+    },
+    {
+      "epoch": 0.383678935640032,
+      "grad_norm": 2.0625,
+      "learning_rate": 3.4070778712693555e-05,
+      "loss": 0.388,
+      "step": 8738
+    },
+    {
+      "epoch": 0.38376675411923555,
+      "grad_norm": 2.109375,
+      "learning_rate": 3.4064331528825834e-05,
+      "loss": 0.3817,
+      "step": 8740
+    },
+    {
+      "epoch": 0.383854572598439,
+      "grad_norm": 2.046875,
+      "learning_rate": 3.405788365080942e-05,
+      "loss": 0.4326,
+      "step": 8742
+    },
+    {
+      "epoch": 0.3839423910776425,
+      "grad_norm": 1.7890625,
+      "learning_rate": 3.405143507913812e-05,
+      "loss": 0.4267,
+      "step": 8744
+    },
+    {
+      "epoch": 0.384030209556846,
+      "grad_norm": 1.984375,
+      "learning_rate": 3.404498581430574e-05,
+      "loss": 0.4047,
+      "step": 8746
+    },
+    {
+      "epoch": 0.3841180280360495,
+      "grad_norm": 1.9140625,
+      "learning_rate": 3.403853585680619e-05,
+      "loss": 0.3583,
+      "step": 8748
+    },
+    {
+      "epoch": 0.38420584651525297,
+      "grad_norm": 1.9921875,
+      "learning_rate": 3.403208520713338e-05,
+      "loss": 0.3668,
+      "step": 8750
+    },
+    {
+      "epoch": 0.38429366499445644,
+      "grad_norm": 2.125,
+      "learning_rate": 3.402563386578133e-05,
+      "loss": 0.4008,
+      "step": 8752
+    },
+    {
+      "epoch": 0.38438148347365997,
+      "grad_norm": 1.9140625,
+      "learning_rate": 3.401918183324408e-05,
+      "loss": 0.3952,
+      "step": 8754
+    },
+    {
+      "epoch": 0.38446930195286344,
+      "grad_norm": 2.09375,
+      "learning_rate": 3.4012729110015715e-05,
+      "loss": 0.3952,
+      "step": 8756
+    },
+    {
+      "epoch": 0.3845571204320669,
+      "grad_norm": 1.9296875,
+      "learning_rate": 3.4006275696590394e-05,
+      "loss": 0.4012,
+      "step": 8758
+    },
+    {
+      "epoch": 0.3846449389112704,
+      "grad_norm": 1.8203125,
+      "learning_rate": 3.399982159346232e-05,
+      "loss": 0.3867,
+      "step": 8760
+    },
+    {
+      "epoch": 0.3847327573904739,
+      "grad_norm": 2.046875,
+      "learning_rate": 3.3993366801125766e-05,
+      "loss": 0.3622,
+      "step": 8762
+    },
+    {
+      "epoch": 0.3848205758696774,
+      "grad_norm": 2.015625,
+      "learning_rate": 3.398691132007501e-05,
+      "loss": 0.4084,
+      "step": 8764
+    },
+    {
+      "epoch": 0.38490839434888086,
+      "grad_norm": 1.921875,
+      "learning_rate": 3.398045515080443e-05,
+      "loss": 0.4114,
+      "step": 8766
+    },
+    {
+      "epoch": 0.3849962128280843,
+      "grad_norm": 2.171875,
+      "learning_rate": 3.397399829380845e-05,
+      "loss": 0.3966,
+      "step": 8768
+    },
+    {
+      "epoch": 0.38508403130728786,
+      "grad_norm": 2.203125,
+      "learning_rate": 3.3967540749581535e-05,
+      "loss": 0.4238,
+      "step": 8770
+    },
+    {
+      "epoch": 0.3851718497864913,
+      "grad_norm": 2.5625,
+      "learning_rate": 3.3961082518618195e-05,
+      "loss": 0.3761,
+      "step": 8772
+    },
+    {
+      "epoch": 0.3852596682656948,
+      "grad_norm": 2.03125,
+      "learning_rate": 3.395462360141301e-05,
+      "loss": 0.368,
+      "step": 8774
+    },
+    {
+      "epoch": 0.38534748674489827,
+      "grad_norm": 1.90625,
+      "learning_rate": 3.394816399846059e-05,
+      "loss": 0.4066,
+      "step": 8776
+    },
+    {
+      "epoch": 0.3854353052241018,
+      "grad_norm": 1.8828125,
+      "learning_rate": 3.3941703710255634e-05,
+      "loss": 0.4185,
+      "step": 8778
+    },
+    {
+      "epoch": 0.38552312370330527,
+      "grad_norm": 1.9140625,
+      "learning_rate": 3.393524273729286e-05,
+      "loss": 0.3871,
+      "step": 8780
+    },
+    {
+      "epoch": 0.38561094218250874,
+      "grad_norm": 2.5,
+      "learning_rate": 3.3928781080067064e-05,
+      "loss": 0.4065,
+      "step": 8782
+    },
+    {
+      "epoch": 0.3856987606617122,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.392231873907307e-05,
+      "loss": 0.4016,
+      "step": 8784
+    },
+    {
+      "epoch": 0.38578657914091574,
+      "grad_norm": 2.109375,
+      "learning_rate": 3.3915855714805766e-05,
+      "loss": 0.3787,
+      "step": 8786
+    },
+    {
+      "epoch": 0.3858743976201192,
+      "grad_norm": 2.109375,
+      "learning_rate": 3.39093920077601e-05,
+      "loss": 0.3677,
+      "step": 8788
+    },
+    {
+      "epoch": 0.3859622160993227,
+      "grad_norm": 1.9609375,
+      "learning_rate": 3.3902927618431044e-05,
+      "loss": 0.3667,
+      "step": 8790
+    },
+    {
+      "epoch": 0.3860500345785262,
+      "grad_norm": 1.875,
+      "learning_rate": 3.3896462547313665e-05,
+      "loss": 0.3921,
+      "step": 8792
+    },
+    {
+      "epoch": 0.3861378530577297,
+      "grad_norm": 2.421875,
+      "learning_rate": 3.3889996794903055e-05,
+      "loss": 0.4078,
+      "step": 8794
+    },
+    {
+      "epoch": 0.38622567153693316,
+      "grad_norm": 2.328125,
+      "learning_rate": 3.3883530361694355e-05,
+      "loss": 0.388,
+      "step": 8796
+    },
+    {
+      "epoch": 0.38631349001613663,
+      "grad_norm": 2.4375,
+      "learning_rate": 3.3877063248182775e-05,
+      "loss": 0.386,
+      "step": 8798
+    },
+    {
+      "epoch": 0.38640130849534016,
+      "grad_norm": 1.9453125,
+      "learning_rate": 3.3870595454863564e-05,
+      "loss": 0.4035,
+      "step": 8800
+    },
+    {
+      "epoch": 0.38648912697454363,
+      "grad_norm": 2.4375,
+      "learning_rate": 3.386412698223202e-05,
+      "loss": 0.3703,
+      "step": 8802
+    },
+    {
+      "epoch": 0.3865769454537471,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.385765783078351e-05,
+      "loss": 0.4085,
+      "step": 8804
+    },
+    {
+      "epoch": 0.3866647639329506,
+      "grad_norm": 2.25,
+      "learning_rate": 3.385118800101344e-05,
+      "loss": 0.3968,
+      "step": 8806
+    },
+    {
+      "epoch": 0.3867525824121541,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.384471749341727e-05,
+      "loss": 0.3996,
+      "step": 8808
+    },
+    {
+      "epoch": 0.38684040089135757,
+      "grad_norm": 1.890625,
+      "learning_rate": 3.383824630849052e-05,
+      "loss": 0.4039,
+      "step": 8810
+    },
+    {
+      "epoch": 0.38692821937056104,
+      "grad_norm": 1.9453125,
+      "learning_rate": 3.383177444672874e-05,
+      "loss": 0.3871,
+      "step": 8812
+    },
+    {
+      "epoch": 0.3870160378497645,
+      "grad_norm": 2.078125,
+      "learning_rate": 3.3825301908627556e-05,
+      "loss": 0.4155,
+      "step": 8814
+    },
+    {
+      "epoch": 0.38710385632896804,
+      "grad_norm": 1.8515625,
+      "learning_rate": 3.381882869468264e-05,
+      "loss": 0.374,
+      "step": 8816
+    },
+    {
+      "epoch": 0.3871916748081715,
+      "grad_norm": 1.8046875,
+      "learning_rate": 3.3812354805389713e-05,
+      "loss": 0.4077,
+      "step": 8818
+    },
+    {
+      "epoch": 0.387279493287375,
+      "grad_norm": 1.8359375,
+      "learning_rate": 3.380588024124454e-05,
+      "loss": 0.3662,
+      "step": 8820
+    },
+    {
+      "epoch": 0.38736731176657846,
+      "grad_norm": 2.078125,
+      "learning_rate": 3.379940500274294e-05,
+      "loss": 0.4216,
+      "step": 8822
+    },
+    {
+      "epoch": 0.387455130245782,
+      "grad_norm": 2.015625,
+      "learning_rate": 3.3792929090380806e-05,
+      "loss": 0.4041,
+      "step": 8824
+    },
+    {
+      "epoch": 0.38754294872498546,
+      "grad_norm": 1.9609375,
+      "learning_rate": 3.3786452504654045e-05,
+      "loss": 0.3596,
+      "step": 8826
+    },
+    {
+      "epoch": 0.38763076720418893,
+      "grad_norm": 1.75,
+      "learning_rate": 3.377997524605865e-05,
+      "loss": 0.3894,
+      "step": 8828
+    },
+    {
+      "epoch": 0.3877185856833924,
+      "grad_norm": 1.9453125,
+      "learning_rate": 3.377349731509064e-05,
+      "loss": 0.3847,
+      "step": 8830
+    },
+    {
+      "epoch": 0.38780640416259593,
+      "grad_norm": 2.265625,
+      "learning_rate": 3.376701871224611e-05,
+      "loss": 0.39,
+      "step": 8832
+    },
+    {
+      "epoch": 0.3878942226417994,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.3760539438021184e-05,
+      "loss": 0.4149,
+      "step": 8834
+    },
+    {
+      "epoch": 0.3879820411210029,
+      "grad_norm": 2.09375,
+      "learning_rate": 3.375405949291205e-05,
+      "loss": 0.3674,
+      "step": 8836
+    },
+    {
+      "epoch": 0.38806985960020635,
+      "grad_norm": 1.90625,
+      "learning_rate": 3.374757887741494e-05,
+      "loss": 0.3604,
+      "step": 8838
+    },
+    {
+      "epoch": 0.3881576780794099,
+      "grad_norm": 1.921875,
+      "learning_rate": 3.3741097592026136e-05,
+      "loss": 0.392,
+      "step": 8840
+    },
+    {
+      "epoch": 0.38824549655861335,
+      "grad_norm": 2.03125,
+      "learning_rate": 3.373461563724198e-05,
+      "loss": 0.411,
+      "step": 8842
+    },
+    {
+      "epoch": 0.3883333150378168,
+      "grad_norm": 2.109375,
+      "learning_rate": 3.372813301355888e-05,
+      "loss": 0.4219,
+      "step": 8844
+    },
+    {
+      "epoch": 0.38842113351702034,
+      "grad_norm": 2.3125,
+      "learning_rate": 3.3721649721473255e-05,
+      "loss": 0.4035,
+      "step": 8846
+    },
+    {
+      "epoch": 0.3885089519962238,
+      "grad_norm": 1.9140625,
+      "learning_rate": 3.3715165761481606e-05,
+      "loss": 0.3774,
+      "step": 8848
+    },
+    {
+      "epoch": 0.3885967704754273,
+      "grad_norm": 1.9453125,
+      "learning_rate": 3.370868113408047e-05,
+      "loss": 0.3749,
+      "step": 8850
+    },
+    {
+      "epoch": 0.38868458895463076,
+      "grad_norm": 1.8984375,
+      "learning_rate": 3.3702195839766445e-05,
+      "loss": 0.4078,
+      "step": 8852
+    },
+    {
+      "epoch": 0.3887724074338343,
+      "grad_norm": 1.9609375,
+      "learning_rate": 3.369570987903618e-05,
+      "loss": 0.3886,
+      "step": 8854
+    },
+    {
+      "epoch": 0.38886022591303776,
+      "grad_norm": 2.421875,
+      "learning_rate": 3.368922325238636e-05,
+      "loss": 0.3892,
+      "step": 8856
+    },
+    {
+      "epoch": 0.38894804439224123,
+      "grad_norm": 1.9140625,
+      "learning_rate": 3.368273596031374e-05,
+      "loss": 0.388,
+      "step": 8858
+    },
+    {
+      "epoch": 0.3890358628714447,
+      "grad_norm": 2.140625,
+      "learning_rate": 3.367624800331513e-05,
+      "loss": 0.3847,
+      "step": 8860
+    },
+    {
+      "epoch": 0.38912368135064823,
+      "grad_norm": 2.40625,
+      "learning_rate": 3.366975938188737e-05,
+      "loss": 0.3817,
+      "step": 8862
+    },
+    {
+      "epoch": 0.3892114998298517,
+      "grad_norm": 2.109375,
+      "learning_rate": 3.3663270096527344e-05,
+      "loss": 0.3949,
+      "step": 8864
+    },
+    {
+      "epoch": 0.3892993183090552,
+      "grad_norm": 2.078125,
+      "learning_rate": 3.3656780147732024e-05,
+      "loss": 0.3954,
+      "step": 8866
+    },
+    {
+      "epoch": 0.38938713678825865,
+      "grad_norm": 2.078125,
+      "learning_rate": 3.3650289535998406e-05,
+      "loss": 0.3788,
+      "step": 8868
+    },
+    {
+      "epoch": 0.3894749552674622,
+      "grad_norm": 1.90625,
+      "learning_rate": 3.364379826182354e-05,
+      "loss": 0.3814,
+      "step": 8870
+    },
+    {
+      "epoch": 0.38956277374666565,
+      "grad_norm": 2.125,
+      "learning_rate": 3.363730632570453e-05,
+      "loss": 0.4086,
+      "step": 8872
+    },
+    {
+      "epoch": 0.3896505922258691,
+      "grad_norm": 1.7578125,
+      "learning_rate": 3.363081372813853e-05,
+      "loss": 0.3798,
+      "step": 8874
+    },
+    {
+      "epoch": 0.3897384107050726,
+      "grad_norm": 1.9296875,
+      "learning_rate": 3.362432046962275e-05,
+      "loss": 0.4,
+      "step": 8876
+    },
+    {
+      "epoch": 0.3898262291842761,
+      "grad_norm": 1.8515625,
+      "learning_rate": 3.3617826550654445e-05,
+      "loss": 0.4174,
+      "step": 8878
+    },
+    {
+      "epoch": 0.3899140476634796,
+      "grad_norm": 2.0,
+      "learning_rate": 3.361133197173091e-05,
+      "loss": 0.388,
+      "step": 8880
+    },
+    {
+      "epoch": 0.39000186614268306,
+      "grad_norm": 2.03125,
+      "learning_rate": 3.360483673334951e-05,
+      "loss": 0.3948,
+      "step": 8882
+    },
+    {
+      "epoch": 0.39008968462188653,
+      "grad_norm": 2.078125,
+      "learning_rate": 3.359834083600765e-05,
+      "loss": 0.3934,
+      "step": 8884
+    },
+    {
+      "epoch": 0.39017750310109006,
+      "grad_norm": 1.7421875,
+      "learning_rate": 3.359184428020279e-05,
+      "loss": 0.3894,
+      "step": 8886
+    },
+    {
+      "epoch": 0.39026532158029353,
+      "grad_norm": 1.8515625,
+      "learning_rate": 3.3585347066432435e-05,
+      "loss": 0.4051,
+      "step": 8888
+    },
+    {
+      "epoch": 0.390353140059497,
+      "grad_norm": 1.828125,
+      "learning_rate": 3.357884919519414e-05,
+      "loss": 0.3849,
+      "step": 8890
+    },
+    {
+      "epoch": 0.39044095853870053,
+      "grad_norm": 1.78125,
+      "learning_rate": 3.357235066698552e-05,
+      "loss": 0.4023,
+      "step": 8892
+    },
+    {
+      "epoch": 0.390528777017904,
+      "grad_norm": 1.90625,
+      "learning_rate": 3.356585148230423e-05,
+      "loss": 0.3851,
+      "step": 8894
+    },
+    {
+      "epoch": 0.3906165954971075,
+      "grad_norm": 1.890625,
+      "learning_rate": 3.355935164164798e-05,
+      "loss": 0.4059,
+      "step": 8896
+    },
+    {
+      "epoch": 0.39070441397631095,
+      "grad_norm": 1.8046875,
+      "learning_rate": 3.355285114551453e-05,
+      "loss": 0.3912,
+      "step": 8898
+    },
+    {
+      "epoch": 0.3907922324555145,
+      "grad_norm": 2.015625,
+      "learning_rate": 3.3546349994401686e-05,
+      "loss": 0.4149,
+      "step": 8900
+    },
+    {
+      "epoch": 0.39088005093471795,
+      "grad_norm": 2.046875,
+      "learning_rate": 3.3539848188807315e-05,
+      "loss": 0.3998,
+      "step": 8902
+    },
+    {
+      "epoch": 0.3909678694139214,
+      "grad_norm": 1.890625,
+      "learning_rate": 3.3533345729229315e-05,
+      "loss": 0.3907,
+      "step": 8904
+    },
+    {
+      "epoch": 0.3910556878931249,
+      "grad_norm": 2.109375,
+      "learning_rate": 3.352684261616566e-05,
+      "loss": 0.4005,
+      "step": 8906
+    },
+    {
+      "epoch": 0.3911435063723284,
+      "grad_norm": 2.0,
+      "learning_rate": 3.352033885011436e-05,
+      "loss": 0.3899,
+      "step": 8908
+    },
+    {
+      "epoch": 0.3912313248515319,
+      "grad_norm": 2.078125,
+      "learning_rate": 3.351383443157347e-05,
+      "loss": 0.3959,
+      "step": 8910
+    },
+    {
+      "epoch": 0.39131914333073536,
+      "grad_norm": 2.40625,
+      "learning_rate": 3.350732936104108e-05,
+      "loss": 0.3978,
+      "step": 8912
+    },
+    {
+      "epoch": 0.39140696180993884,
+      "grad_norm": 2.078125,
+      "learning_rate": 3.3500823639015376e-05,
+      "loss": 0.3955,
+      "step": 8914
+    },
+    {
+      "epoch": 0.39149478028914236,
+      "grad_norm": 2.484375,
+      "learning_rate": 3.3494317265994565e-05,
+      "loss": 0.4163,
+      "step": 8916
+    },
+    {
+      "epoch": 0.39158259876834584,
+      "grad_norm": 2.34375,
+      "learning_rate": 3.348781024247689e-05,
+      "loss": 0.3811,
+      "step": 8918
+    },
+    {
+      "epoch": 0.3916704172475493,
+      "grad_norm": 2.09375,
+      "learning_rate": 3.3481302568960684e-05,
+      "loss": 0.3828,
+      "step": 8920
+    },
+    {
+      "epoch": 0.3917582357267528,
+      "grad_norm": 1.8125,
+      "learning_rate": 3.3474794245944294e-05,
+      "loss": 0.3912,
+      "step": 8922
+    },
+    {
+      "epoch": 0.3918460542059563,
+      "grad_norm": 1.9765625,
+      "learning_rate": 3.3468285273926123e-05,
+      "loss": 0.3943,
+      "step": 8924
+    },
+    {
+      "epoch": 0.3919338726851598,
+      "grad_norm": 1.7578125,
+      "learning_rate": 3.346177565340464e-05,
+      "loss": 0.3705,
+      "step": 8926
+    },
+    {
+      "epoch": 0.39202169116436325,
+      "grad_norm": 2.078125,
+      "learning_rate": 3.345526538487834e-05,
+      "loss": 0.4072,
+      "step": 8928
+    },
+    {
+      "epoch": 0.3921095096435667,
+      "grad_norm": 2.0625,
+      "learning_rate": 3.34487544688458e-05,
+      "loss": 0.4101,
+      "step": 8930
+    },
+    {
+      "epoch": 0.39219732812277025,
+      "grad_norm": 1.875,
+      "learning_rate": 3.3442242905805614e-05,
+      "loss": 0.3849,
+      "step": 8932
+    },
+    {
+      "epoch": 0.3922851466019737,
+      "grad_norm": 2.046875,
+      "learning_rate": 3.343573069625645e-05,
+      "loss": 0.3763,
+      "step": 8934
+    },
+    {
+      "epoch": 0.3923729650811772,
+      "grad_norm": 2.0625,
+      "learning_rate": 3.3429217840697e-05,
+      "loss": 0.3715,
+      "step": 8936
+    },
+    {
+      "epoch": 0.39246078356038067,
+      "grad_norm": 1.953125,
+      "learning_rate": 3.3422704339626024e-05,
+      "loss": 0.4216,
+      "step": 8938
+    },
+    {
+      "epoch": 0.3925486020395842,
+      "grad_norm": 1.875,
+      "learning_rate": 3.341619019354233e-05,
+      "loss": 0.3712,
+      "step": 8940
+    },
+    {
+      "epoch": 0.39263642051878767,
+      "grad_norm": 1.8359375,
+      "learning_rate": 3.3409675402944786e-05,
+      "loss": 0.3723,
+      "step": 8942
+    },
+    {
+      "epoch": 0.39272423899799114,
+      "grad_norm": 1.9921875,
+      "learning_rate": 3.3403159968332273e-05,
+      "loss": 0.3759,
+      "step": 8944
+    },
+    {
+      "epoch": 0.39281205747719466,
+      "grad_norm": 2.3125,
+      "learning_rate": 3.339664389020376e-05,
+      "loss": 0.3894,
+      "step": 8946
+    },
+    {
+      "epoch": 0.39289987595639814,
+      "grad_norm": 1.9375,
+      "learning_rate": 3.339012716905824e-05,
+      "loss": 0.3792,
+      "step": 8948
+    },
+    {
+      "epoch": 0.3929876944356016,
+      "grad_norm": 2.109375,
+      "learning_rate": 3.338360980539477e-05,
+      "loss": 0.4002,
+      "step": 8950
+    },
+    {
+      "epoch": 0.3930755129148051,
+      "grad_norm": 2.140625,
+      "learning_rate": 3.3377091799712454e-05,
+      "loss": 0.3979,
+      "step": 8952
+    },
+    {
+      "epoch": 0.3931633313940086,
+      "grad_norm": 1.8125,
+      "learning_rate": 3.3370573152510445e-05,
+      "loss": 0.4109,
+      "step": 8954
+    },
+    {
+      "epoch": 0.3932511498732121,
+      "grad_norm": 1.96875,
+      "learning_rate": 3.336405386428792e-05,
+      "loss": 0.4168,
+      "step": 8956
+    },
+    {
+      "epoch": 0.39333896835241555,
+      "grad_norm": 1.78125,
+      "learning_rate": 3.335753393554416e-05,
+      "loss": 0.3975,
+      "step": 8958
+    },
+    {
+      "epoch": 0.393426786831619,
+      "grad_norm": 1.984375,
+      "learning_rate": 3.335101336677845e-05,
+      "loss": 0.388,
+      "step": 8960
+    },
+    {
+      "epoch": 0.39351460531082255,
+      "grad_norm": 1.9296875,
+      "learning_rate": 3.3344492158490134e-05,
+      "loss": 0.4136,
+      "step": 8962
+    },
+    {
+      "epoch": 0.393602423790026,
+      "grad_norm": 2.171875,
+      "learning_rate": 3.33379703111786e-05,
+      "loss": 0.4321,
+      "step": 8964
+    },
+    {
+      "epoch": 0.3936902422692295,
+      "grad_norm": 2.015625,
+      "learning_rate": 3.3331447825343306e-05,
+      "loss": 0.3847,
+      "step": 8966
+    },
+    {
+      "epoch": 0.39377806074843297,
+      "grad_norm": 1.984375,
+      "learning_rate": 3.3324924701483734e-05,
+      "loss": 0.3922,
+      "step": 8968
+    },
+    {
+      "epoch": 0.3938658792276365,
+      "grad_norm": 1.890625,
+      "learning_rate": 3.331840094009944e-05,
+      "loss": 0.3798,
+      "step": 8970
+    },
+    {
+      "epoch": 0.39395369770683997,
+      "grad_norm": 1.921875,
+      "learning_rate": 3.3311876541690014e-05,
+      "loss": 0.3968,
+      "step": 8972
+    },
+    {
+      "epoch": 0.39404151618604344,
+      "grad_norm": 2.046875,
+      "learning_rate": 3.330535150675508e-05,
+      "loss": 0.4071,
+      "step": 8974
+    },
+    {
+      "epoch": 0.3941293346652469,
+      "grad_norm": 2.140625,
+      "learning_rate": 3.329882583579433e-05,
+      "loss": 0.381,
+      "step": 8976
+    },
+    {
+      "epoch": 0.39421715314445044,
+      "grad_norm": 2.421875,
+      "learning_rate": 3.329229952930752e-05,
+      "loss": 0.3968,
+      "step": 8978
+    },
+    {
+      "epoch": 0.3943049716236539,
+      "grad_norm": 2.296875,
+      "learning_rate": 3.3285772587794417e-05,
+      "loss": 0.3864,
+      "step": 8980
+    },
+    {
+      "epoch": 0.3943927901028574,
+      "grad_norm": 2.03125,
+      "learning_rate": 3.3279245011754874e-05,
+      "loss": 0.4029,
+      "step": 8982
+    },
+    {
+      "epoch": 0.39448060858206085,
+      "grad_norm": 2.046875,
+      "learning_rate": 3.3272716801688754e-05,
+      "loss": 0.3987,
+      "step": 8984
+    },
+    {
+      "epoch": 0.3945684270612644,
+      "grad_norm": 1.859375,
+      "learning_rate": 3.3266187958096e-05,
+      "loss": 0.4145,
+      "step": 8986
+    },
+    {
+      "epoch": 0.39465624554046785,
+      "grad_norm": 1.96875,
+      "learning_rate": 3.325965848147659e-05,
+      "loss": 0.3853,
+      "step": 8988
+    },
+    {
+      "epoch": 0.3947440640196713,
+      "grad_norm": 2.015625,
+      "learning_rate": 3.3253128372330556e-05,
+      "loss": 0.3784,
+      "step": 8990
+    },
+    {
+      "epoch": 0.39483188249887485,
+      "grad_norm": 1.7734375,
+      "learning_rate": 3.324659763115797e-05,
+      "loss": 0.3869,
+      "step": 8992
+    },
+    {
+      "epoch": 0.3949197009780783,
+      "grad_norm": 1.78125,
+      "learning_rate": 3.324006625845896e-05,
+      "loss": 0.4024,
+      "step": 8994
+    },
+    {
+      "epoch": 0.3950075194572818,
+      "grad_norm": 2.046875,
+      "learning_rate": 3.3233534254733706e-05,
+      "loss": 0.4002,
+      "step": 8996
+    },
+    {
+      "epoch": 0.39509533793648527,
+      "grad_norm": 1.875,
+      "learning_rate": 3.322700162048242e-05,
+      "loss": 0.3972,
+      "step": 8998
+    },
+    {
+      "epoch": 0.3951831564156888,
+      "grad_norm": 1.84375,
+      "learning_rate": 3.322046835620538e-05,
+      "loss": 0.3898,
+      "step": 9000
+    },
+    {
+      "epoch": 0.39527097489489227,
+      "grad_norm": 2.046875,
+      "learning_rate": 3.32139344624029e-05,
+      "loss": 0.3801,
+      "step": 9002
+    },
+    {
+      "epoch": 0.39535879337409574,
+      "grad_norm": 1.9453125,
+      "learning_rate": 3.320739993957535e-05,
+      "loss": 0.4005,
+      "step": 9004
+    },
+    {
+      "epoch": 0.3954466118532992,
+      "grad_norm": 2.078125,
+      "learning_rate": 3.320086478822315e-05,
+      "loss": 0.3623,
+      "step": 9006
+    },
+    {
+      "epoch": 0.39553443033250274,
+      "grad_norm": 1.953125,
+      "learning_rate": 3.319432900884676e-05,
+      "loss": 0.379,
+      "step": 9008
+    },
+    {
+      "epoch": 0.3956222488117062,
+      "grad_norm": 1.8515625,
+      "learning_rate": 3.318779260194668e-05,
+      "loss": 0.3844,
+      "step": 9010
+    },
+    {
+      "epoch": 0.3957100672909097,
+      "grad_norm": 1.8359375,
+      "learning_rate": 3.318125556802348e-05,
+      "loss": 0.3694,
+      "step": 9012
+    },
+    {
+      "epoch": 0.39579788577011316,
+      "grad_norm": 2.078125,
+      "learning_rate": 3.317471790757778e-05,
+      "loss": 0.3808,
+      "step": 9014
+    },
+    {
+      "epoch": 0.3958857042493167,
+      "grad_norm": 1.8984375,
+      "learning_rate": 3.316817962111022e-05,
+      "loss": 0.4207,
+      "step": 9016
+    },
+    {
+      "epoch": 0.39597352272852016,
+      "grad_norm": 1.9296875,
+      "learning_rate": 3.31616407091215e-05,
+      "loss": 0.3916,
+      "step": 9018
+    },
+    {
+      "epoch": 0.3960613412077236,
+      "grad_norm": 1.96875,
+      "learning_rate": 3.315510117211238e-05,
+      "loss": 0.379,
+      "step": 9020
+    },
+    {
+      "epoch": 0.3961491596869271,
+      "grad_norm": 1.7734375,
+      "learning_rate": 3.314856101058366e-05,
+      "loss": 0.3991,
+      "step": 9022
+    },
+    {
+      "epoch": 0.3962369781661306,
+      "grad_norm": 2.03125,
+      "learning_rate": 3.314202022503618e-05,
+      "loss": 0.3913,
+      "step": 9024
+    },
+    {
+      "epoch": 0.3963247966453341,
+      "grad_norm": 1.90625,
+      "learning_rate": 3.313547881597084e-05,
+      "loss": 0.3706,
+      "step": 9026
+    },
+    {
+      "epoch": 0.39641261512453757,
+      "grad_norm": 1.9375,
+      "learning_rate": 3.312893678388858e-05,
+      "loss": 0.3919,
+      "step": 9028
+    },
+    {
+      "epoch": 0.39650043360374104,
+      "grad_norm": 1.8671875,
+      "learning_rate": 3.31223941292904e-05,
+      "loss": 0.3523,
+      "step": 9030
+    },
+    {
+      "epoch": 0.39658825208294457,
+      "grad_norm": 2.125,
+      "learning_rate": 3.3115850852677327e-05,
+      "loss": 0.3827,
+      "step": 9032
+    },
+    {
+      "epoch": 0.39667607056214804,
+      "grad_norm": 2.0,
+      "learning_rate": 3.310930695455046e-05,
+      "loss": 0.3588,
+      "step": 9034
+    },
+    {
+      "epoch": 0.3967638890413515,
+      "grad_norm": 2.03125,
+      "learning_rate": 3.3102762435410904e-05,
+      "loss": 0.4117,
+      "step": 9036
+    },
+    {
+      "epoch": 0.396851707520555,
+      "grad_norm": 1.9921875,
+      "learning_rate": 3.3096217295759866e-05,
+      "loss": 0.3824,
+      "step": 9038
+    },
+    {
+      "epoch": 0.3969395259997585,
+      "grad_norm": 2.09375,
+      "learning_rate": 3.308967153609857e-05,
+      "loss": 0.3642,
+      "step": 9040
+    },
+    {
+      "epoch": 0.397027344478962,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.308312515692828e-05,
+      "loss": 0.3674,
+      "step": 9042
+    },
+    {
+      "epoch": 0.39711516295816546,
+      "grad_norm": 1.8671875,
+      "learning_rate": 3.307657815875034e-05,
+      "loss": 0.4016,
+      "step": 9044
+    },
+    {
+      "epoch": 0.397202981437369,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.30700305420661e-05,
+      "loss": 0.372,
+      "step": 9046
+    },
+    {
+      "epoch": 0.39729079991657246,
+      "grad_norm": 2.203125,
+      "learning_rate": 3.306348230737699e-05,
+      "loss": 0.426,
+      "step": 9048
+    },
+    {
+      "epoch": 0.39737861839577593,
+      "grad_norm": 1.859375,
+      "learning_rate": 3.305693345518447e-05,
+      "loss": 0.3717,
+      "step": 9050
+    },
+    {
+      "epoch": 0.3974664368749794,
+      "grad_norm": 2.171875,
+      "learning_rate": 3.3050383985990056e-05,
+      "loss": 0.4166,
+      "step": 9052
+    },
+    {
+      "epoch": 0.39755425535418293,
+      "grad_norm": 2.328125,
+      "learning_rate": 3.304383390029531e-05,
+      "loss": 0.4124,
+      "step": 9054
+    },
+    {
+      "epoch": 0.3976420738333864,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.3037283198601826e-05,
+      "loss": 0.4108,
+      "step": 9056
+    },
+    {
+      "epoch": 0.39772989231258987,
+      "grad_norm": 1.8515625,
+      "learning_rate": 3.303073188141128e-05,
+      "loss": 0.3818,
+      "step": 9058
+    },
+    {
+      "epoch": 0.39781771079179334,
+      "grad_norm": 2.046875,
+      "learning_rate": 3.3024179949225355e-05,
+      "loss": 0.3845,
+      "step": 9060
+    },
+    {
+      "epoch": 0.39790552927099687,
+      "grad_norm": 2.078125,
+      "learning_rate": 3.30176274025458e-05,
+      "loss": 0.4132,
+      "step": 9062
+    },
+    {
+      "epoch": 0.39799334775020034,
+      "grad_norm": 2.3125,
+      "learning_rate": 3.301107424187443e-05,
+      "loss": 0.3977,
+      "step": 9064
+    },
+    {
+      "epoch": 0.3980811662294038,
+      "grad_norm": 2.265625,
+      "learning_rate": 3.300452046771306e-05,
+      "loss": 0.3447,
+      "step": 9066
+    },
+    {
+      "epoch": 0.3981689847086073,
+      "grad_norm": 1.953125,
+      "learning_rate": 3.29979660805636e-05,
+      "loss": 0.3903,
+      "step": 9068
+    },
+    {
+      "epoch": 0.3982568031878108,
+      "grad_norm": 1.984375,
+      "learning_rate": 3.299141108092799e-05,
+      "loss": 0.3841,
+      "step": 9070
+    },
+    {
+      "epoch": 0.3983446216670143,
+      "grad_norm": 2.125,
+      "learning_rate": 3.298485546930819e-05,
+      "loss": 0.3854,
+      "step": 9072
+    },
+    {
+      "epoch": 0.39843244014621776,
+      "grad_norm": 1.8125,
+      "learning_rate": 3.2978299246206246e-05,
+      "loss": 0.3778,
+      "step": 9074
+    },
+    {
+      "epoch": 0.39852025862542123,
+      "grad_norm": 2.125,
+      "learning_rate": 3.297174241212424e-05,
+      "loss": 0.4199,
+      "step": 9076
+    },
+    {
+      "epoch": 0.39860807710462476,
+      "grad_norm": 2.09375,
+      "learning_rate": 3.296518496756428e-05,
+      "loss": 0.3954,
+      "step": 9078
+    },
+    {
+      "epoch": 0.39869589558382823,
+      "grad_norm": 1.984375,
+      "learning_rate": 3.295862691302855e-05,
+      "loss": 0.3856,
+      "step": 9080
+    },
+    {
+      "epoch": 0.3987837140630317,
+      "grad_norm": 1.84375,
+      "learning_rate": 3.295206824901926e-05,
+      "loss": 0.3974,
+      "step": 9082
+    },
+    {
+      "epoch": 0.3988715325422352,
+      "grad_norm": 2.046875,
+      "learning_rate": 3.2945508976038694e-05,
+      "loss": 0.4178,
+      "step": 9084
+    },
+    {
+      "epoch": 0.3989593510214387,
+      "grad_norm": 1.9609375,
+      "learning_rate": 3.293894909458913e-05,
+      "loss": 0.3664,
+      "step": 9086
+    },
+    {
+      "epoch": 0.3990471695006422,
+      "grad_norm": 1.984375,
+      "learning_rate": 3.2932388605172946e-05,
+      "loss": 0.3838,
+      "step": 9088
+    },
+    {
+      "epoch": 0.39913498797984565,
+      "grad_norm": 2.625,
+      "learning_rate": 3.2925827508292535e-05,
+      "loss": 0.3785,
+      "step": 9090
+    },
+    {
+      "epoch": 0.3992228064590491,
+      "grad_norm": 1.96875,
+      "learning_rate": 3.2919265804450364e-05,
+      "loss": 0.3911,
+      "step": 9092
+    },
+    {
+      "epoch": 0.39931062493825265,
+      "grad_norm": 2.109375,
+      "learning_rate": 3.291270349414891e-05,
+      "loss": 0.3595,
+      "step": 9094
+    },
+    {
+      "epoch": 0.3993984434174561,
+      "grad_norm": 1.921875,
+      "learning_rate": 3.290614057789073e-05,
+      "loss": 0.3703,
+      "step": 9096
+    },
+    {
+      "epoch": 0.3994862618966596,
+      "grad_norm": 1.9453125,
+      "learning_rate": 3.289957705617841e-05,
+      "loss": 0.3909,
+      "step": 9098
+    },
+    {
+      "epoch": 0.3995740803758631,
+      "grad_norm": 2.015625,
+      "learning_rate": 3.2893012929514574e-05,
+      "loss": 0.391,
+      "step": 9100
+    },
+    {
+      "epoch": 0.3996618988550666,
+      "grad_norm": 2.53125,
+      "learning_rate": 3.288644819840193e-05,
+      "loss": 0.3775,
+      "step": 9102
+    },
+    {
+      "epoch": 0.39974971733427006,
+      "grad_norm": 2.0,
+      "learning_rate": 3.2879882863343184e-05,
+      "loss": 0.4158,
+      "step": 9104
+    },
+    {
+      "epoch": 0.39983753581347353,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.287331692484113e-05,
+      "loss": 0.37,
+      "step": 9106
+    },
+    {
+      "epoch": 0.39992535429267706,
+      "grad_norm": 2.109375,
+      "learning_rate": 3.286675038339857e-05,
+      "loss": 0.3767,
+      "step": 9108
+    },
+    {
+      "epoch": 0.40001317277188053,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.286018323951838e-05,
+      "loss": 0.3919,
+      "step": 9110
+    },
+    {
+      "epoch": 0.400100991251084,
+      "grad_norm": 1.9375,
+      "learning_rate": 3.2853615493703475e-05,
+      "loss": 0.4056,
+      "step": 9112
+    },
+    {
+      "epoch": 0.4001888097302875,
+      "grad_norm": 1.8046875,
+      "learning_rate": 3.284704714645681e-05,
+      "loss": 0.4012,
+      "step": 9114
+    },
+    {
+      "epoch": 0.400276628209491,
+      "grad_norm": 1.9453125,
+      "learning_rate": 3.28404781982814e-05,
+      "loss": 0.4101,
+      "step": 9116
+    },
+    {
+      "epoch": 0.4003644466886945,
+      "grad_norm": 1.8359375,
+      "learning_rate": 3.283390864968029e-05,
+      "loss": 0.3904,
+      "step": 9118
+    },
+    {
+      "epoch": 0.40045226516789795,
+      "grad_norm": 1.8359375,
+      "learning_rate": 3.282733850115657e-05,
+      "loss": 0.4022,
+      "step": 9120
+    },
+    {
+      "epoch": 0.4005400836471014,
+      "grad_norm": 1.796875,
+      "learning_rate": 3.28207677532134e-05,
+      "loss": 0.3885,
+      "step": 9122
+    },
+    {
+      "epoch": 0.40062790212630495,
+      "grad_norm": 1.8046875,
+      "learning_rate": 3.281419640635395e-05,
+      "loss": 0.4061,
+      "step": 9124
+    },
+    {
+      "epoch": 0.4007157206055084,
+      "grad_norm": 1.875,
+      "learning_rate": 3.2807624461081477e-05,
+      "loss": 0.3681,
+      "step": 9126
+    },
+    {
+      "epoch": 0.4008035390847119,
+      "grad_norm": 1.8515625,
+      "learning_rate": 3.280105191789925e-05,
+      "loss": 0.3851,
+      "step": 9128
+    },
+    {
+      "epoch": 0.40089135756391536,
+      "grad_norm": 1.8125,
+      "learning_rate": 3.279447877731058e-05,
+      "loss": 0.3873,
+      "step": 9130
+    },
+    {
+      "epoch": 0.4009791760431189,
+      "grad_norm": 1.8828125,
+      "learning_rate": 3.2787905039818875e-05,
+      "loss": 0.3851,
+      "step": 9132
+    },
+    {
+      "epoch": 0.40106699452232236,
+      "grad_norm": 2.125,
+      "learning_rate": 3.278133070592753e-05,
+      "loss": 0.3864,
+      "step": 9134
+    },
+    {
+      "epoch": 0.40115481300152583,
+      "grad_norm": 2.0625,
+      "learning_rate": 3.277475577614002e-05,
+      "loss": 0.3968,
+      "step": 9136
+    },
+    {
+      "epoch": 0.4012426314807293,
+      "grad_norm": 1.8671875,
+      "learning_rate": 3.276818025095984e-05,
+      "loss": 0.3765,
+      "step": 9138
+    },
+    {
+      "epoch": 0.40133044995993283,
+      "grad_norm": 1.859375,
+      "learning_rate": 3.276160413089056e-05,
+      "loss": 0.3751,
+      "step": 9140
+    },
+    {
+      "epoch": 0.4014182684391363,
+      "grad_norm": 1.7890625,
+      "learning_rate": 3.275502741643577e-05,
+      "loss": 0.4117,
+      "step": 9142
+    },
+    {
+      "epoch": 0.4015060869183398,
+      "grad_norm": 1.8046875,
+      "learning_rate": 3.274845010809913e-05,
+      "loss": 0.3599,
+      "step": 9144
+    },
+    {
+      "epoch": 0.4015939053975433,
+      "grad_norm": 1.96875,
+      "learning_rate": 3.274187220638431e-05,
+      "loss": 0.4034,
+      "step": 9146
+    },
+    {
+      "epoch": 0.4016817238767468,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.273529371179507e-05,
+      "loss": 0.3909,
+      "step": 9148
+    },
+    {
+      "epoch": 0.40176954235595025,
+      "grad_norm": 2.171875,
+      "learning_rate": 3.2728714624835174e-05,
+      "loss": 0.3765,
+      "step": 9150
+    },
+    {
+      "epoch": 0.4018573608351537,
+      "grad_norm": 2.359375,
+      "learning_rate": 3.272213494600847e-05,
+      "loss": 0.3848,
+      "step": 9152
+    },
+    {
+      "epoch": 0.40194517931435725,
+      "grad_norm": 1.984375,
+      "learning_rate": 3.271555467581882e-05,
+      "loss": 0.3683,
+      "step": 9154
+    },
+    {
+      "epoch": 0.4020329977935607,
+      "grad_norm": 2.078125,
+      "learning_rate": 3.270897381477014e-05,
+      "loss": 0.376,
+      "step": 9156
+    },
+    {
+      "epoch": 0.4021208162727642,
+      "grad_norm": 2.265625,
+      "learning_rate": 3.27023923633664e-05,
+      "loss": 0.4018,
+      "step": 9158
+    },
+    {
+      "epoch": 0.40220863475196766,
+      "grad_norm": 1.9609375,
+      "learning_rate": 3.26958103221116e-05,
+      "loss": 0.4136,
+      "step": 9160
+    },
+    {
+      "epoch": 0.4022964532311712,
+      "grad_norm": 1.875,
+      "learning_rate": 3.26892276915098e-05,
+      "loss": 0.3917,
+      "step": 9162
+    },
+    {
+      "epoch": 0.40238427171037466,
+      "grad_norm": 1.8359375,
+      "learning_rate": 3.268264447206511e-05,
+      "loss": 0.3996,
+      "step": 9164
+    },
+    {
+      "epoch": 0.40247209018957814,
+      "grad_norm": 2.0,
+      "learning_rate": 3.267606066428166e-05,
+      "loss": 0.3714,
+      "step": 9166
+    },
+    {
+      "epoch": 0.4025599086687816,
+      "grad_norm": 2.25,
+      "learning_rate": 3.266947626866365e-05,
+      "loss": 0.3664,
+      "step": 9168
+    },
+    {
+      "epoch": 0.40264772714798513,
+      "grad_norm": 1.984375,
+      "learning_rate": 3.26628912857153e-05,
+      "loss": 0.3979,
+      "step": 9170
+    },
+    {
+      "epoch": 0.4027355456271886,
+      "grad_norm": 1.796875,
+      "learning_rate": 3.2656305715940905e-05,
+      "loss": 0.3641,
+      "step": 9172
+    },
+    {
+      "epoch": 0.4028233641063921,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.264971955984478e-05,
+      "loss": 0.3947,
+      "step": 9174
+    },
+    {
+      "epoch": 0.40291118258559555,
+      "grad_norm": 1.9140625,
+      "learning_rate": 3.2643132817931294e-05,
+      "loss": 0.3627,
+      "step": 9176
+    },
+    {
+      "epoch": 0.4029990010647991,
+      "grad_norm": 1.96875,
+      "learning_rate": 3.263654549070486e-05,
+      "loss": 0.3983,
+      "step": 9178
+    },
+    {
+      "epoch": 0.40308681954400255,
+      "grad_norm": 1.8671875,
+      "learning_rate": 3.262995757866996e-05,
+      "loss": 0.4208,
+      "step": 9180
+    },
+    {
+      "epoch": 0.403174638023206,
+      "grad_norm": 2.09375,
+      "learning_rate": 3.262336908233106e-05,
+      "loss": 0.4071,
+      "step": 9182
+    },
+    {
+      "epoch": 0.4032624565024095,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.2616780002192746e-05,
+      "loss": 0.4131,
+      "step": 9184
+    },
+    {
+      "epoch": 0.403350274981613,
+      "grad_norm": 1.9375,
+      "learning_rate": 3.2610190338759586e-05,
+      "loss": 0.4001,
+      "step": 9186
+    },
+    {
+      "epoch": 0.4034380934608165,
+      "grad_norm": 1.875,
+      "learning_rate": 3.2603600092536216e-05,
+      "loss": 0.4055,
+      "step": 9188
+    },
+    {
+      "epoch": 0.40352591194001997,
+      "grad_norm": 1.8046875,
+      "learning_rate": 3.259700926402734e-05,
+      "loss": 0.3863,
+      "step": 9190
+    },
+    {
+      "epoch": 0.40361373041922344,
+      "grad_norm": 1.8828125,
+      "learning_rate": 3.2590417853737666e-05,
+      "loss": 0.4105,
+      "step": 9192
+    },
+    {
+      "epoch": 0.40370154889842697,
+      "grad_norm": 1.8828125,
+      "learning_rate": 3.258382586217198e-05,
+      "loss": 0.4068,
+      "step": 9194
+    },
+    {
+      "epoch": 0.40378936737763044,
+      "grad_norm": 1.9453125,
+      "learning_rate": 3.2577233289835085e-05,
+      "loss": 0.3828,
+      "step": 9196
+    },
+    {
+      "epoch": 0.4038771858568339,
+      "grad_norm": 1.9609375,
+      "learning_rate": 3.257064013723185e-05,
+      "loss": 0.3971,
+      "step": 9198
+    },
+    {
+      "epoch": 0.40396500433603744,
+      "grad_norm": 1.984375,
+      "learning_rate": 3.256404640486719e-05,
+      "loss": 0.4,
+      "step": 9200
+    },
+    {
+      "epoch": 0.4040528228152409,
+      "grad_norm": 1.875,
+      "learning_rate": 3.2557452093246025e-05,
+      "loss": 0.3995,
+      "step": 9202
+    },
+    {
+      "epoch": 0.4041406412944444,
+      "grad_norm": 1.828125,
+      "learning_rate": 3.255085720287337e-05,
+      "loss": 0.3784,
+      "step": 9204
+    },
+    {
+      "epoch": 0.40422845977364785,
+      "grad_norm": 2.140625,
+      "learning_rate": 3.254426173425428e-05,
+      "loss": 0.4098,
+      "step": 9206
+    },
+    {
+      "epoch": 0.4043162782528514,
+      "grad_norm": 2.234375,
+      "learning_rate": 3.2537665687893796e-05,
+      "loss": 0.4037,
+      "step": 9208
+    },
+    {
+      "epoch": 0.40440409673205485,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.253106906429707e-05,
+      "loss": 0.3656,
+      "step": 9210
+    },
+    {
+      "epoch": 0.4044919152112583,
+      "grad_norm": 2.09375,
+      "learning_rate": 3.2524471863969274e-05,
+      "loss": 0.382,
+      "step": 9212
+    },
+    {
+      "epoch": 0.4045797336904618,
+      "grad_norm": 1.828125,
+      "learning_rate": 3.251787408741562e-05,
+      "loss": 0.3949,
+      "step": 9214
+    },
+    {
+      "epoch": 0.4046675521696653,
+      "grad_norm": 2.40625,
+      "learning_rate": 3.2511275735141365e-05,
+      "loss": 0.4174,
+      "step": 9216
+    },
+    {
+      "epoch": 0.4047553706488688,
+      "grad_norm": 1.9609375,
+      "learning_rate": 3.250467680765181e-05,
+      "loss": 0.4042,
+      "step": 9218
+    },
+    {
+      "epoch": 0.40484318912807227,
+      "grad_norm": 1.9140625,
+      "learning_rate": 3.2498077305452316e-05,
+      "loss": 0.3811,
+      "step": 9220
+    },
+    {
+      "epoch": 0.40493100760727574,
+      "grad_norm": 1.796875,
+      "learning_rate": 3.249147722904826e-05,
+      "loss": 0.3887,
+      "step": 9222
+    },
+    {
+      "epoch": 0.40501882608647927,
+      "grad_norm": 1.8828125,
+      "learning_rate": 3.248487657894508e-05,
+      "loss": 0.3687,
+      "step": 9224
+    },
+    {
+      "epoch": 0.40510664456568274,
+      "grad_norm": 1.90625,
+      "learning_rate": 3.247827535564826e-05,
+      "loss": 0.377,
+      "step": 9226
+    },
+    {
+      "epoch": 0.4051944630448862,
+      "grad_norm": 2.046875,
+      "learning_rate": 3.2471673559663314e-05,
+      "loss": 0.382,
+      "step": 9228
+    },
+    {
+      "epoch": 0.4052822815240897,
+      "grad_norm": 1.828125,
+      "learning_rate": 3.246507119149582e-05,
+      "loss": 0.3965,
+      "step": 9230
+    },
+    {
+      "epoch": 0.4053701000032932,
+      "grad_norm": 1.859375,
+      "learning_rate": 3.245846825165139e-05,
+      "loss": 0.3912,
+      "step": 9232
+    },
+    {
+      "epoch": 0.4054579184824967,
+      "grad_norm": 2.203125,
+      "learning_rate": 3.245186474063566e-05,
+      "loss": 0.3826,
+      "step": 9234
+    },
+    {
+      "epoch": 0.40554573696170015,
+      "grad_norm": 2.3125,
+      "learning_rate": 3.244526065895436e-05,
+      "loss": 0.3925,
+      "step": 9236
+    },
+    {
+      "epoch": 0.4056335554409036,
+      "grad_norm": 2.0625,
+      "learning_rate": 3.24386560071132e-05,
+      "loss": 0.3818,
+      "step": 9238
+    },
+    {
+      "epoch": 0.40572137392010715,
+      "grad_norm": 1.7890625,
+      "learning_rate": 3.243205078561798e-05,
+      "loss": 0.3815,
+      "step": 9240
+    },
+    {
+      "epoch": 0.4058091923993106,
+      "grad_norm": 1.875,
+      "learning_rate": 3.242544499497453e-05,
+      "loss": 0.3767,
+      "step": 9242
+    },
+    {
+      "epoch": 0.4058970108785141,
+      "grad_norm": 2.125,
+      "learning_rate": 3.241883863568873e-05,
+      "loss": 0.3857,
+      "step": 9244
+    },
+    {
+      "epoch": 0.4059848293577176,
+      "grad_norm": 1.9453125,
+      "learning_rate": 3.241223170826648e-05,
+      "loss": 0.393,
+      "step": 9246
+    },
+    {
+      "epoch": 0.4060726478369211,
+      "grad_norm": 2.234375,
+      "learning_rate": 3.240562421321376e-05,
+      "loss": 0.3945,
+      "step": 9248
+    },
+    {
+      "epoch": 0.40616046631612457,
+      "grad_norm": 1.8671875,
+      "learning_rate": 3.2399016151036555e-05,
+      "loss": 0.3851,
+      "step": 9250
+    },
+    {
+      "epoch": 0.40624828479532804,
+      "grad_norm": 2.4375,
+      "learning_rate": 3.239240752224091e-05,
+      "loss": 0.377,
+      "step": 9252
+    },
+    {
+      "epoch": 0.40633610327453157,
+      "grad_norm": 2.03125,
+      "learning_rate": 3.238579832733294e-05,
+      "loss": 0.3738,
+      "step": 9254
+    },
+    {
+      "epoch": 0.40642392175373504,
+      "grad_norm": 2.125,
+      "learning_rate": 3.2379188566818765e-05,
+      "loss": 0.3562,
+      "step": 9256
+    },
+    {
+      "epoch": 0.4065117402329385,
+      "grad_norm": 1.9765625,
+      "learning_rate": 3.237257824120455e-05,
+      "loss": 0.3817,
+      "step": 9258
+    },
+    {
+      "epoch": 0.406599558712142,
+      "grad_norm": 2.078125,
+      "learning_rate": 3.2365967350996526e-05,
+      "loss": 0.3536,
+      "step": 9260
+    },
+    {
+      "epoch": 0.4066873771913455,
+      "grad_norm": 2.4375,
+      "learning_rate": 3.2359355896700964e-05,
+      "loss": 0.3519,
+      "step": 9262
+    },
+    {
+      "epoch": 0.406775195670549,
+      "grad_norm": 1.859375,
+      "learning_rate": 3.235274387882416e-05,
+      "loss": 0.3692,
+      "step": 9264
+    },
+    {
+      "epoch": 0.40686301414975246,
+      "grad_norm": 1.9375,
+      "learning_rate": 3.234613129787246e-05,
+      "loss": 0.3724,
+      "step": 9266
+    },
+    {
+      "epoch": 0.4069508326289559,
+      "grad_norm": 1.75,
+      "learning_rate": 3.233951815435228e-05,
+      "loss": 0.3574,
+      "step": 9268
+    },
+    {
+      "epoch": 0.40703865110815945,
+      "grad_norm": 2.25,
+      "learning_rate": 3.233290444877003e-05,
+      "loss": 0.3948,
+      "step": 9270
+    },
+    {
+      "epoch": 0.4071264695873629,
+      "grad_norm": 2.0625,
+      "learning_rate": 3.2326290181632204e-05,
+      "loss": 0.4047,
+      "step": 9272
+    },
+    {
+      "epoch": 0.4072142880665664,
+      "grad_norm": 1.953125,
+      "learning_rate": 3.2319675353445314e-05,
+      "loss": 0.3698,
+      "step": 9274
+    },
+    {
+      "epoch": 0.40730210654576987,
+      "grad_norm": 1.875,
+      "learning_rate": 3.231305996471593e-05,
+      "loss": 0.3717,
+      "step": 9276
+    },
+    {
+      "epoch": 0.4073899250249734,
+      "grad_norm": 2.015625,
+      "learning_rate": 3.230644401595067e-05,
+      "loss": 0.4248,
+      "step": 9278
+    },
+    {
+      "epoch": 0.40747774350417687,
+      "grad_norm": 2.140625,
+      "learning_rate": 3.2299827507656165e-05,
+      "loss": 0.3944,
+      "step": 9280
+    },
+    {
+      "epoch": 0.40756556198338034,
+      "grad_norm": 2.46875,
+      "learning_rate": 3.229321044033913e-05,
+      "loss": 0.3982,
+      "step": 9282
+    },
+    {
+      "epoch": 0.4076533804625838,
+      "grad_norm": 3.28125,
+      "learning_rate": 3.228659281450628e-05,
+      "loss": 0.3924,
+      "step": 9284
+    },
+    {
+      "epoch": 0.40774119894178734,
+      "grad_norm": 2.125,
+      "learning_rate": 3.227997463066441e-05,
+      "loss": 0.3693,
+      "step": 9286
+    },
+    {
+      "epoch": 0.4078290174209908,
+      "grad_norm": 2.3125,
+      "learning_rate": 3.227335588932034e-05,
+      "loss": 0.3468,
+      "step": 9288
+    },
+    {
+      "epoch": 0.4079168359001943,
+      "grad_norm": 1.6875,
+      "learning_rate": 3.226673659098093e-05,
+      "loss": 0.4128,
+      "step": 9290
+    },
+    {
+      "epoch": 0.40800465437939776,
+      "grad_norm": 1.828125,
+      "learning_rate": 3.226011673615309e-05,
+      "loss": 0.3843,
+      "step": 9292
+    },
+    {
+      "epoch": 0.4080924728586013,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.225349632534378e-05,
+      "loss": 0.4072,
+      "step": 9294
+    },
+    {
+      "epoch": 0.40818029133780476,
+      "grad_norm": 1.765625,
+      "learning_rate": 3.2246875359059966e-05,
+      "loss": 0.3747,
+      "step": 9296
+    },
+    {
+      "epoch": 0.40826810981700823,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.2240253837808706e-05,
+      "loss": 0.3809,
+      "step": 9298
+    },
+    {
+      "epoch": 0.40835592829621176,
+      "grad_norm": 2.234375,
+      "learning_rate": 3.223363176209708e-05,
+      "loss": 0.3934,
+      "step": 9300
+    },
+    {
+      "epoch": 0.40844374677541523,
+      "grad_norm": 1.8828125,
+      "learning_rate": 3.222700913243219e-05,
+      "loss": 0.4021,
+      "step": 9302
+    },
+    {
+      "epoch": 0.4085315652546187,
+      "grad_norm": 1.796875,
+      "learning_rate": 3.2220385949321215e-05,
+      "loss": 0.379,
+      "step": 9304
+    },
+    {
+      "epoch": 0.4086193837338222,
+      "grad_norm": 1.859375,
+      "learning_rate": 3.221376221327135e-05,
+      "loss": 0.4074,
+      "step": 9306
+    },
+    {
+      "epoch": 0.4087072022130257,
+      "grad_norm": 1.9765625,
+      "learning_rate": 3.220713792478984e-05,
+      "loss": 0.4007,
+      "step": 9308
+    },
+    {
+      "epoch": 0.40879502069222917,
+      "grad_norm": 1.8515625,
+      "learning_rate": 3.220051308438399e-05,
+      "loss": 0.3983,
+      "step": 9310
+    },
+    {
+      "epoch": 0.40888283917143264,
+      "grad_norm": 1.8828125,
+      "learning_rate": 3.2193887692561115e-05,
+      "loss": 0.3699,
+      "step": 9312
+    },
+    {
+      "epoch": 0.4089706576506361,
+      "grad_norm": 2.0,
+      "learning_rate": 3.2187261749828594e-05,
+      "loss": 0.3806,
+      "step": 9314
+    },
+    {
+      "epoch": 0.40905847612983964,
+      "grad_norm": 2.296875,
+      "learning_rate": 3.218063525669385e-05,
+      "loss": 0.3909,
+      "step": 9316
+    },
+    {
+      "epoch": 0.4091462946090431,
+      "grad_norm": 1.8984375,
+      "learning_rate": 3.2174008213664335e-05,
+      "loss": 0.3955,
+      "step": 9318
+    },
+    {
+      "epoch": 0.4092341130882466,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.216738062124756e-05,
+      "loss": 0.3532,
+      "step": 9320
+    },
+    {
+      "epoch": 0.40932193156745006,
+      "grad_norm": 1.9375,
+      "learning_rate": 3.216075247995105e-05,
+      "loss": 0.375,
+      "step": 9322
+    },
+    {
+      "epoch": 0.4094097500466536,
+      "grad_norm": 1.9140625,
+      "learning_rate": 3.21541237902824e-05,
+      "loss": 0.378,
+      "step": 9324
+    },
+    {
+      "epoch": 0.40949756852585706,
+      "grad_norm": 1.9296875,
+      "learning_rate": 3.214749455274923e-05,
+      "loss": 0.3989,
+      "step": 9326
+    },
+    {
+      "epoch": 0.40958538700506053,
+      "grad_norm": 2.28125,
+      "learning_rate": 3.2140864767859216e-05,
+      "loss": 0.3949,
+      "step": 9328
+    },
+    {
+      "epoch": 0.409673205484264,
+      "grad_norm": 1.9453125,
+      "learning_rate": 3.213423443612007e-05,
+      "loss": 0.4091,
+      "step": 9330
+    },
+    {
+      "epoch": 0.40976102396346753,
+      "grad_norm": 1.9296875,
+      "learning_rate": 3.2127603558039545e-05,
+      "loss": 0.4058,
+      "step": 9332
+    },
+    {
+      "epoch": 0.409848842442671,
+      "grad_norm": 1.890625,
+      "learning_rate": 3.212097213412542e-05,
+      "loss": 0.3779,
+      "step": 9334
+    },
+    {
+      "epoch": 0.4099366609218745,
+      "grad_norm": 1.9921875,
+      "learning_rate": 3.211434016488555e-05,
+      "loss": 0.3734,
+      "step": 9336
+    },
+    {
+      "epoch": 0.41002447940107795,
+      "grad_norm": 2.0625,
+      "learning_rate": 3.2107707650827804e-05,
+      "loss": 0.3795,
+      "step": 9338
+    },
+    {
+      "epoch": 0.4101122978802815,
+      "grad_norm": 1.921875,
+      "learning_rate": 3.2101074592460094e-05,
+      "loss": 0.3807,
+      "step": 9340
+    },
+    {
+      "epoch": 0.41020011635948495,
+      "grad_norm": 1.8359375,
+      "learning_rate": 3.2094440990290395e-05,
+      "loss": 0.3797,
+      "step": 9342
+    },
+    {
+      "epoch": 0.4102879348386884,
+      "grad_norm": 1.8515625,
+      "learning_rate": 3.208780684482671e-05,
+      "loss": 0.3895,
+      "step": 9344
+    },
+    {
+      "epoch": 0.41037575331789194,
+      "grad_norm": 2.015625,
+      "learning_rate": 3.208117215657707e-05,
+      "loss": 0.3704,
+      "step": 9346
+    },
+    {
+      "epoch": 0.4104635717970954,
+      "grad_norm": 1.9453125,
+      "learning_rate": 3.207453692604957e-05,
+      "loss": 0.3739,
+      "step": 9348
+    },
+    {
+      "epoch": 0.4105513902762989,
+      "grad_norm": 2.03125,
+      "learning_rate": 3.206790115375234e-05,
+      "loss": 0.3953,
+      "step": 9350
+    },
+    {
+      "epoch": 0.41063920875550236,
+      "grad_norm": 1.9453125,
+      "learning_rate": 3.206126484019354e-05,
+      "loss": 0.3564,
+      "step": 9352
+    },
+    {
+      "epoch": 0.4107270272347059,
+      "grad_norm": 2.015625,
+      "learning_rate": 3.205462798588139e-05,
+      "loss": 0.3534,
+      "step": 9354
+    },
+    {
+      "epoch": 0.41081484571390936,
+      "grad_norm": 1.8828125,
+      "learning_rate": 3.204799059132414e-05,
+      "loss": 0.4022,
+      "step": 9356
+    },
+    {
+      "epoch": 0.41090266419311283,
+      "grad_norm": 2.28125,
+      "learning_rate": 3.204135265703008e-05,
+      "loss": 0.385,
+      "step": 9358
+    },
+    {
+      "epoch": 0.4109904826723163,
+      "grad_norm": 2.203125,
+      "learning_rate": 3.203471418350754e-05,
+      "loss": 0.3867,
+      "step": 9360
+    },
+    {
+      "epoch": 0.41107830115151983,
+      "grad_norm": 2.0625,
+      "learning_rate": 3.20280751712649e-05,
+      "loss": 0.4418,
+      "step": 9362
+    },
+    {
+      "epoch": 0.4111661196307233,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.2021435620810587e-05,
+      "loss": 0.3827,
+      "step": 9364
+    },
+    {
+      "epoch": 0.4112539381099268,
+      "grad_norm": 1.796875,
+      "learning_rate": 3.2014795532653054e-05,
+      "loss": 0.4018,
+      "step": 9366
+    },
+    {
+      "epoch": 0.41134175658913025,
+      "grad_norm": 1.9921875,
+      "learning_rate": 3.200815490730079e-05,
+      "loss": 0.4129,
+      "step": 9368
+    },
+    {
+      "epoch": 0.4114295750683338,
+      "grad_norm": 2.046875,
+      "learning_rate": 3.200151374526234e-05,
+      "loss": 0.3971,
+      "step": 9370
+    },
+    {
+      "epoch": 0.41151739354753725,
+      "grad_norm": 1.984375,
+      "learning_rate": 3.19948720470463e-05,
+      "loss": 0.3694,
+      "step": 9372
+    },
+    {
+      "epoch": 0.4116052120267407,
+      "grad_norm": 1.9921875,
+      "learning_rate": 3.198822981316127e-05,
+      "loss": 0.3975,
+      "step": 9374
+    },
+    {
+      "epoch": 0.4116930305059442,
+      "grad_norm": 1.984375,
+      "learning_rate": 3.198158704411593e-05,
+      "loss": 0.3705,
+      "step": 9376
+    },
+    {
+      "epoch": 0.4117808489851477,
+      "grad_norm": 2.0,
+      "learning_rate": 3.1974943740418986e-05,
+      "loss": 0.4057,
+      "step": 9378
+    },
+    {
+      "epoch": 0.4118686674643512,
+      "grad_norm": 1.953125,
+      "learning_rate": 3.1968299902579166e-05,
+      "loss": 0.3794,
+      "step": 9380
+    },
+    {
+      "epoch": 0.41195648594355466,
+      "grad_norm": 2.265625,
+      "learning_rate": 3.196165553110528e-05,
+      "loss": 0.3638,
+      "step": 9382
+    },
+    {
+      "epoch": 0.41204430442275813,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.1955010626506146e-05,
+      "loss": 0.3815,
+      "step": 9384
+    },
+    {
+      "epoch": 0.41213212290196166,
+      "grad_norm": 2.109375,
+      "learning_rate": 3.1948365189290625e-05,
+      "loss": 0.379,
+      "step": 9386
+    },
+    {
+      "epoch": 0.41221994138116513,
+      "grad_norm": 1.8828125,
+      "learning_rate": 3.194171921996763e-05,
+      "loss": 0.3453,
+      "step": 9388
+    },
+    {
+      "epoch": 0.4123077598603686,
+      "grad_norm": 1.65625,
+      "learning_rate": 3.1935072719046115e-05,
+      "loss": 0.3958,
+      "step": 9390
+    },
+    {
+      "epoch": 0.4123955783395721,
+      "grad_norm": 1.9375,
+      "learning_rate": 3.192842568703508e-05,
+      "loss": 0.4017,
+      "step": 9392
+    },
+    {
+      "epoch": 0.4124833968187756,
+      "grad_norm": 2.140625,
+      "learning_rate": 3.192177812444353e-05,
+      "loss": 0.4008,
+      "step": 9394
+    },
+    {
+      "epoch": 0.4125712152979791,
+      "grad_norm": 1.8359375,
+      "learning_rate": 3.191513003178055e-05,
+      "loss": 0.3786,
+      "step": 9396
+    },
+    {
+      "epoch": 0.41265903377718255,
+      "grad_norm": 2.015625,
+      "learning_rate": 3.1908481409555266e-05,
+      "loss": 0.3702,
+      "step": 9398
+    },
+    {
+      "epoch": 0.4127468522563861,
+      "grad_norm": 1.8671875,
+      "learning_rate": 3.190183225827682e-05,
+      "loss": 0.3878,
+      "step": 9400
+    },
+    {
+      "epoch": 0.41283467073558955,
+      "grad_norm": 2.171875,
+      "learning_rate": 3.1895182578454395e-05,
+      "loss": 0.3722,
+      "step": 9402
+    },
+    {
+      "epoch": 0.412922489214793,
+      "grad_norm": 1.8671875,
+      "learning_rate": 3.188853237059725e-05,
+      "loss": 0.3754,
+      "step": 9404
+    },
+    {
+      "epoch": 0.4130103076939965,
+      "grad_norm": 2.25,
+      "learning_rate": 3.188188163521463e-05,
+      "loss": 0.4122,
+      "step": 9406
+    },
+    {
+      "epoch": 0.4130981261732,
+      "grad_norm": 2.03125,
+      "learning_rate": 3.1875230372815864e-05,
+      "loss": 0.3988,
+      "step": 9408
+    },
+    {
+      "epoch": 0.4131859446524035,
+      "grad_norm": 1.9296875,
+      "learning_rate": 3.1868578583910316e-05,
+      "loss": 0.364,
+      "step": 9410
+    },
+    {
+      "epoch": 0.41327376313160696,
+      "grad_norm": 2.0625,
+      "learning_rate": 3.186192626900737e-05,
+      "loss": 0.379,
+      "step": 9412
+    },
+    {
+      "epoch": 0.41336158161081044,
+      "grad_norm": 2.0,
+      "learning_rate": 3.185527342861647e-05,
+      "loss": 0.4191,
+      "step": 9414
+    },
+    {
+      "epoch": 0.41344940009001396,
+      "grad_norm": 2.25,
+      "learning_rate": 3.184862006324709e-05,
+      "loss": 0.414,
+      "step": 9416
+    },
+    {
+      "epoch": 0.41353721856921744,
+      "grad_norm": 1.984375,
+      "learning_rate": 3.184196617340874e-05,
+      "loss": 0.4011,
+      "step": 9418
+    },
+    {
+      "epoch": 0.4136250370484209,
+      "grad_norm": 1.7421875,
+      "learning_rate": 3.1835311759610975e-05,
+      "loss": 0.3796,
+      "step": 9420
+    },
+    {
+      "epoch": 0.4137128555276244,
+      "grad_norm": 1.9765625,
+      "learning_rate": 3.18286568223634e-05,
+      "loss": 0.3948,
+      "step": 9422
+    },
+    {
+      "epoch": 0.4138006740068279,
+      "grad_norm": 1.9140625,
+      "learning_rate": 3.1822001362175646e-05,
+      "loss": 0.3655,
+      "step": 9424
+    },
+    {
+      "epoch": 0.4138884924860314,
+      "grad_norm": 2.109375,
+      "learning_rate": 3.18153453795574e-05,
+      "loss": 0.3793,
+      "step": 9426
+    },
+    {
+      "epoch": 0.41397631096523485,
+      "grad_norm": 1.8125,
+      "learning_rate": 3.180868887501837e-05,
+      "loss": 0.3982,
+      "step": 9428
+    },
+    {
+      "epoch": 0.4140641294444383,
+      "grad_norm": 1.96875,
+      "learning_rate": 3.1802031849068316e-05,
+      "loss": 0.3917,
+      "step": 9430
+    },
+    {
+      "epoch": 0.41415194792364185,
+      "grad_norm": 1.9453125,
+      "learning_rate": 3.1795374302217025e-05,
+      "loss": 0.3937,
+      "step": 9432
+    },
+    {
+      "epoch": 0.4142397664028453,
+      "grad_norm": 2.0,
+      "learning_rate": 3.178871623497434e-05,
+      "loss": 0.3886,
+      "step": 9434
+    },
+    {
+      "epoch": 0.4143275848820488,
+      "grad_norm": 2.046875,
+      "learning_rate": 3.178205764785014e-05,
+      "loss": 0.3583,
+      "step": 9436
+    },
+    {
+      "epoch": 0.41441540336125227,
+      "grad_norm": 1.8359375,
+      "learning_rate": 3.177539854135434e-05,
+      "loss": 0.3792,
+      "step": 9438
+    },
+    {
+      "epoch": 0.4145032218404558,
+      "grad_norm": 1.8828125,
+      "learning_rate": 3.1768738915996896e-05,
+      "loss": 0.3909,
+      "step": 9440
+    },
+    {
+      "epoch": 0.41459104031965927,
+      "grad_norm": 2.015625,
+      "learning_rate": 3.1762078772287804e-05,
+      "loss": 0.3741,
+      "step": 9442
+    },
+    {
+      "epoch": 0.41467885879886274,
+      "grad_norm": 2.3125,
+      "learning_rate": 3.17554181107371e-05,
+      "loss": 0.3985,
+      "step": 9444
+    },
+    {
+      "epoch": 0.41476667727806626,
+      "grad_norm": 1.921875,
+      "learning_rate": 3.174875693185486e-05,
+      "loss": 0.3948,
+      "step": 9446
+    },
+    {
+      "epoch": 0.41485449575726974,
+      "grad_norm": 2.078125,
+      "learning_rate": 3.174209523615119e-05,
+      "loss": 0.3797,
+      "step": 9448
+    },
+    {
+      "epoch": 0.4149423142364732,
+      "grad_norm": 2.234375,
+      "learning_rate": 3.173543302413625e-05,
+      "loss": 0.3699,
+      "step": 9450
+    },
+    {
+      "epoch": 0.4150301327156767,
+      "grad_norm": 1.953125,
+      "learning_rate": 3.172877029632023e-05,
+      "loss": 0.3681,
+      "step": 9452
+    },
+    {
+      "epoch": 0.4151179511948802,
+      "grad_norm": 1.90625,
+      "learning_rate": 3.1722107053213386e-05,
+      "loss": 0.361,
+      "step": 9454
+    },
+    {
+      "epoch": 0.4152057696740837,
+      "grad_norm": 1.9921875,
+      "learning_rate": 3.171544329532596e-05,
+      "loss": 0.3652,
+      "step": 9456
+    },
+    {
+      "epoch": 0.41529358815328715,
+      "grad_norm": 2.265625,
+      "learning_rate": 3.1708779023168275e-05,
+      "loss": 0.3872,
+      "step": 9458
+    },
+    {
+      "epoch": 0.4153814066324906,
+      "grad_norm": 1.984375,
+      "learning_rate": 3.170211423725069e-05,
+      "loss": 0.3984,
+      "step": 9460
+    },
+    {
+      "epoch": 0.41546922511169415,
+      "grad_norm": 2.09375,
+      "learning_rate": 3.169544893808359e-05,
+      "loss": 0.4018,
+      "step": 9462
+    },
+    {
+      "epoch": 0.4155570435908976,
+      "grad_norm": 1.9296875,
+      "learning_rate": 3.16887831261774e-05,
+      "loss": 0.4096,
+      "step": 9464
+    },
+    {
+      "epoch": 0.4156448620701011,
+      "grad_norm": 1.859375,
+      "learning_rate": 3.16821168020426e-05,
+      "loss": 0.3844,
+      "step": 9466
+    },
+    {
+      "epoch": 0.41573268054930457,
+      "grad_norm": 1.9921875,
+      "learning_rate": 3.167544996618969e-05,
+      "loss": 0.4051,
+      "step": 9468
+    },
+    {
+      "epoch": 0.4158204990285081,
+      "grad_norm": 2.09375,
+      "learning_rate": 3.166878261912922e-05,
+      "loss": 0.3821,
+      "step": 9470
+    },
+    {
+      "epoch": 0.41590831750771157,
+      "grad_norm": 1.9140625,
+      "learning_rate": 3.166211476137178e-05,
+      "loss": 0.4056,
+      "step": 9472
+    },
+    {
+      "epoch": 0.41599613598691504,
+      "grad_norm": 1.9765625,
+      "learning_rate": 3.1655446393427994e-05,
+      "loss": 0.3796,
+      "step": 9474
+    },
+    {
+      "epoch": 0.4160839544661185,
+      "grad_norm": 2.484375,
+      "learning_rate": 3.164877751580853e-05,
+      "loss": 0.3845,
+      "step": 9476
+    },
+    {
+      "epoch": 0.41617177294532204,
+      "grad_norm": 2.078125,
+      "learning_rate": 3.164210812902409e-05,
+      "loss": 0.3913,
+      "step": 9478
+    },
+    {
+      "epoch": 0.4162595914245255,
+      "grad_norm": 2.09375,
+      "learning_rate": 3.1635438233585425e-05,
+      "loss": 0.401,
+      "step": 9480
+    },
+    {
+      "epoch": 0.416347409903729,
+      "grad_norm": 1.9140625,
+      "learning_rate": 3.162876783000329e-05,
+      "loss": 0.3596,
+      "step": 9482
+    },
+    {
+      "epoch": 0.41643522838293245,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.162209691878854e-05,
+      "loss": 0.3905,
+      "step": 9484
+    },
+    {
+      "epoch": 0.416523046862136,
+      "grad_norm": 1.828125,
+      "learning_rate": 3.161542550045202e-05,
+      "loss": 0.3869,
+      "step": 9486
+    },
+    {
+      "epoch": 0.41661086534133945,
+      "grad_norm": 2.015625,
+      "learning_rate": 3.160875357550462e-05,
+      "loss": 0.3911,
+      "step": 9488
+    },
+    {
+      "epoch": 0.4166986838205429,
+      "grad_norm": 1.7890625,
+      "learning_rate": 3.1602081144457297e-05,
+      "loss": 0.3836,
+      "step": 9490
+    },
+    {
+      "epoch": 0.4167865022997464,
+      "grad_norm": 1.9296875,
+      "learning_rate": 3.159540820782102e-05,
+      "loss": 0.3814,
+      "step": 9492
+    },
+    {
+      "epoch": 0.4168743207789499,
+      "grad_norm": 1.8828125,
+      "learning_rate": 3.1588734766106794e-05,
+      "loss": 0.3724,
+      "step": 9494
+    },
+    {
+      "epoch": 0.4169621392581534,
+      "grad_norm": 1.7421875,
+      "learning_rate": 3.1582060819825674e-05,
+      "loss": 0.387,
+      "step": 9496
+    },
+    {
+      "epoch": 0.41704995773735687,
+      "grad_norm": 1.7890625,
+      "learning_rate": 3.1575386369488766e-05,
+      "loss": 0.3574,
+      "step": 9498
+    },
+    {
+      "epoch": 0.4171377762165604,
+      "grad_norm": 1.9609375,
+      "learning_rate": 3.15687114156072e-05,
+      "loss": 0.3769,
+      "step": 9500
+    },
+    {
+      "epoch": 0.41722559469576387,
+      "grad_norm": 1.7734375,
+      "learning_rate": 3.156203595869213e-05,
+      "loss": 0.4157,
+      "step": 9502
+    },
+    {
+      "epoch": 0.41731341317496734,
+      "grad_norm": 2.015625,
+      "learning_rate": 3.155535999925478e-05,
+      "loss": 0.401,
+      "step": 9504
+    },
+    {
+      "epoch": 0.4174012316541708,
+      "grad_norm": 1.8203125,
+      "learning_rate": 3.1548683537806384e-05,
+      "loss": 0.3857,
+      "step": 9506
+    },
+    {
+      "epoch": 0.41748905013337434,
+      "grad_norm": 1.90625,
+      "learning_rate": 3.1542006574858236e-05,
+      "loss": 0.365,
+      "step": 9508
+    },
+    {
+      "epoch": 0.4175768686125778,
+      "grad_norm": 2.140625,
+      "learning_rate": 3.153532911092165e-05,
+      "loss": 0.3915,
+      "step": 9510
+    },
+    {
+      "epoch": 0.4176646870917813,
+      "grad_norm": 2.125,
+      "learning_rate": 3.1528651146508e-05,
+      "loss": 0.3779,
+      "step": 9512
+    },
+    {
+      "epoch": 0.41775250557098476,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.1521972682128674e-05,
+      "loss": 0.3807,
+      "step": 9514
+    },
+    {
+      "epoch": 0.4178403240501883,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.151529371829513e-05,
+      "loss": 0.3603,
+      "step": 9516
+    },
+    {
+      "epoch": 0.41792814252939176,
+      "grad_norm": 1.8828125,
+      "learning_rate": 3.150861425551882e-05,
+      "loss": 0.3942,
+      "step": 9518
+    },
+    {
+      "epoch": 0.4180159610085952,
+      "grad_norm": 1.890625,
+      "learning_rate": 3.150193429431127e-05,
+      "loss": 0.3824,
+      "step": 9520
+    },
+    {
+      "epoch": 0.4181037794877987,
+      "grad_norm": 1.984375,
+      "learning_rate": 3.149525383518404e-05,
+      "loss": 0.4087,
+      "step": 9522
+    },
+    {
+      "epoch": 0.4181915979670022,
+      "grad_norm": 1.953125,
+      "learning_rate": 3.148857287864871e-05,
+      "loss": 0.3705,
+      "step": 9524
+    },
+    {
+      "epoch": 0.4182794164462057,
+      "grad_norm": 1.8984375,
+      "learning_rate": 3.148189142521691e-05,
+      "loss": 0.3858,
+      "step": 9526
+    },
+    {
+      "epoch": 0.41836723492540917,
+      "grad_norm": 1.984375,
+      "learning_rate": 3.1475209475400316e-05,
+      "loss": 0.3966,
+      "step": 9528
+    },
+    {
+      "epoch": 0.41845505340461264,
+      "grad_norm": 1.8828125,
+      "learning_rate": 3.146852702971063e-05,
+      "loss": 0.3847,
+      "step": 9530
+    },
+    {
+      "epoch": 0.41854287188381617,
+      "grad_norm": 1.9296875,
+      "learning_rate": 3.146184408865959e-05,
+      "loss": 0.4065,
+      "step": 9532
+    },
+    {
+      "epoch": 0.41863069036301964,
+      "grad_norm": 1.8515625,
+      "learning_rate": 3.1455160652758975e-05,
+      "loss": 0.3839,
+      "step": 9534
+    },
+    {
+      "epoch": 0.4187185088422231,
+      "grad_norm": 1.8359375,
+      "learning_rate": 3.1448476722520625e-05,
+      "loss": 0.3895,
+      "step": 9536
+    },
+    {
+      "epoch": 0.4188063273214266,
+      "grad_norm": 2.125,
+      "learning_rate": 3.144179229845637e-05,
+      "loss": 0.4202,
+      "step": 9538
+    },
+    {
+      "epoch": 0.4188941458006301,
+      "grad_norm": 1.953125,
+      "learning_rate": 3.143510738107812e-05,
+      "loss": 0.4095,
+      "step": 9540
+    },
+    {
+      "epoch": 0.4189819642798336,
+      "grad_norm": 1.6484375,
+      "learning_rate": 3.1428421970897804e-05,
+      "loss": 0.4062,
+      "step": 9542
+    },
+    {
+      "epoch": 0.41906978275903706,
+      "grad_norm": 1.9609375,
+      "learning_rate": 3.142173606842739e-05,
+      "loss": 0.3639,
+      "step": 9544
+    },
+    {
+      "epoch": 0.41915760123824053,
+      "grad_norm": 1.9140625,
+      "learning_rate": 3.1415049674178884e-05,
+      "loss": 0.3899,
+      "step": 9546
+    },
+    {
+      "epoch": 0.41924541971744406,
+      "grad_norm": 2.03125,
+      "learning_rate": 3.1408362788664346e-05,
+      "loss": 0.3779,
+      "step": 9548
+    },
+    {
+      "epoch": 0.41933323819664753,
+      "grad_norm": 2.0,
+      "learning_rate": 3.1401675412395845e-05,
+      "loss": 0.4029,
+      "step": 9550
+    },
+    {
+      "epoch": 0.419421056675851,
+      "grad_norm": 1.9140625,
+      "learning_rate": 3.139498754588551e-05,
+      "loss": 0.3936,
+      "step": 9552
+    },
+    {
+      "epoch": 0.41950887515505453,
+      "grad_norm": 1.7578125,
+      "learning_rate": 3.1388299189645496e-05,
+      "loss": 0.407,
+      "step": 9554
+    },
+    {
+      "epoch": 0.419596693634258,
+      "grad_norm": 1.9140625,
+      "learning_rate": 3.1381610344187995e-05,
+      "loss": 0.3474,
+      "step": 9556
+    },
+    {
+      "epoch": 0.41968451211346147,
+      "grad_norm": 2.296875,
+      "learning_rate": 3.1374921010025246e-05,
+      "loss": 0.4169,
+      "step": 9558
+    },
+    {
+      "epoch": 0.41977233059266494,
+      "grad_norm": 1.9921875,
+      "learning_rate": 3.136823118766951e-05,
+      "loss": 0.3908,
+      "step": 9560
+    },
+    {
+      "epoch": 0.41986014907186847,
+      "grad_norm": 1.859375,
+      "learning_rate": 3.1361540877633114e-05,
+      "loss": 0.3901,
+      "step": 9562
+    },
+    {
+      "epoch": 0.41994796755107194,
+      "grad_norm": 2.09375,
+      "learning_rate": 3.135485008042839e-05,
+      "loss": 0.3783,
+      "step": 9564
+    },
+    {
+      "epoch": 0.4200357860302754,
+      "grad_norm": 1.8359375,
+      "learning_rate": 3.134815879656772e-05,
+      "loss": 0.3723,
+      "step": 9566
+    },
+    {
+      "epoch": 0.4201236045094789,
+      "grad_norm": 1.7734375,
+      "learning_rate": 3.134146702656353e-05,
+      "loss": 0.391,
+      "step": 9568
+    },
+    {
+      "epoch": 0.4202114229886824,
+      "grad_norm": 1.828125,
+      "learning_rate": 3.1334774770928276e-05,
+      "loss": 0.3982,
+      "step": 9570
+    },
+    {
+      "epoch": 0.4202992414678859,
+      "grad_norm": 1.9375,
+      "learning_rate": 3.132808203017445e-05,
+      "loss": 0.3727,
+      "step": 9572
+    },
+    {
+      "epoch": 0.42038705994708936,
+      "grad_norm": 1.8984375,
+      "learning_rate": 3.1321388804814584e-05,
+      "loss": 0.365,
+      "step": 9574
+    },
+    {
+      "epoch": 0.42047487842629283,
+      "grad_norm": 1.703125,
+      "learning_rate": 3.131469509536125e-05,
+      "loss": 0.4185,
+      "step": 9576
+    },
+    {
+      "epoch": 0.42056269690549636,
+      "grad_norm": 1.953125,
+      "learning_rate": 3.130800090232705e-05,
+      "loss": 0.3825,
+      "step": 9578
+    },
+    {
+      "epoch": 0.42065051538469983,
+      "grad_norm": 1.8515625,
+      "learning_rate": 3.1301306226224625e-05,
+      "loss": 0.3819,
+      "step": 9580
+    },
+    {
+      "epoch": 0.4207383338639033,
+      "grad_norm": 1.828125,
+      "learning_rate": 3.129461106756667e-05,
+      "loss": 0.3885,
+      "step": 9582
+    },
+    {
+      "epoch": 0.4208261523431068,
+      "grad_norm": 1.9140625,
+      "learning_rate": 3.128791542686588e-05,
+      "loss": 0.3542,
+      "step": 9584
+    },
+    {
+      "epoch": 0.4209139708223103,
+      "grad_norm": 2.140625,
+      "learning_rate": 3.128121930463502e-05,
+      "loss": 0.3729,
+      "step": 9586
+    },
+    {
+      "epoch": 0.4210017893015138,
+      "grad_norm": 1.96875,
+      "learning_rate": 3.1274522701386886e-05,
+      "loss": 0.4038,
+      "step": 9588
+    },
+    {
+      "epoch": 0.42108960778071725,
+      "grad_norm": 2.40625,
+      "learning_rate": 3.12678256176343e-05,
+      "loss": 0.3794,
+      "step": 9590
+    },
+    {
+      "epoch": 0.4211774262599207,
+      "grad_norm": 1.9140625,
+      "learning_rate": 3.126112805389012e-05,
+      "loss": 0.3976,
+      "step": 9592
+    },
+    {
+      "epoch": 0.42126524473912424,
+      "grad_norm": 1.7890625,
+      "learning_rate": 3.1254430010667254e-05,
+      "loss": 0.3973,
+      "step": 9594
+    },
+    {
+      "epoch": 0.4213530632183277,
+      "grad_norm": 1.96875,
+      "learning_rate": 3.1247731488478636e-05,
+      "loss": 0.3854,
+      "step": 9596
+    },
+    {
+      "epoch": 0.4214408816975312,
+      "grad_norm": 2.140625,
+      "learning_rate": 3.124103248783725e-05,
+      "loss": 0.4122,
+      "step": 9598
+    },
+    {
+      "epoch": 0.4215287001767347,
+      "grad_norm": 1.9453125,
+      "learning_rate": 3.1234333009256095e-05,
+      "loss": 0.3758,
+      "step": 9600
+    },
+    {
+      "epoch": 0.4216165186559382,
+      "grad_norm": 2.09375,
+      "learning_rate": 3.122763305324823e-05,
+      "loss": 0.374,
+      "step": 9602
+    },
+    {
+      "epoch": 0.42170433713514166,
+      "grad_norm": 1.9453125,
+      "learning_rate": 3.1220932620326726e-05,
+      "loss": 0.3778,
+      "step": 9604
+    },
+    {
+      "epoch": 0.42179215561434513,
+      "grad_norm": 2.09375,
+      "learning_rate": 3.1214231711004716e-05,
+      "loss": 0.3878,
+      "step": 9606
+    },
+    {
+      "epoch": 0.42187997409354866,
+      "grad_norm": 1.90625,
+      "learning_rate": 3.1207530325795344e-05,
+      "loss": 0.3833,
+      "step": 9608
+    },
+    {
+      "epoch": 0.42196779257275213,
+      "grad_norm": 1.7890625,
+      "learning_rate": 3.120082846521181e-05,
+      "loss": 0.3614,
+      "step": 9610
+    },
+    {
+      "epoch": 0.4220556110519556,
+      "grad_norm": 1.90625,
+      "learning_rate": 3.1194126129767356e-05,
+      "loss": 0.3733,
+      "step": 9612
+    },
+    {
+      "epoch": 0.4221434295311591,
+      "grad_norm": 1.875,
+      "learning_rate": 3.1187423319975236e-05,
+      "loss": 0.3895,
+      "step": 9614
+    },
+    {
+      "epoch": 0.4222312480103626,
+      "grad_norm": 1.8515625,
+      "learning_rate": 3.118072003634876e-05,
+      "loss": 0.3825,
+      "step": 9616
+    },
+    {
+      "epoch": 0.4223190664895661,
+      "grad_norm": 1.9453125,
+      "learning_rate": 3.117401627940125e-05,
+      "loss": 0.3708,
+      "step": 9618
+    },
+    {
+      "epoch": 0.42240688496876955,
+      "grad_norm": 1.890625,
+      "learning_rate": 3.1167312049646094e-05,
+      "loss": 0.3634,
+      "step": 9620
+    },
+    {
+      "epoch": 0.422494703447973,
+      "grad_norm": 1.7734375,
+      "learning_rate": 3.116060734759671e-05,
+      "loss": 0.3742,
+      "step": 9622
+    },
+    {
+      "epoch": 0.42258252192717655,
+      "grad_norm": 1.96875,
+      "learning_rate": 3.115390217376654e-05,
+      "loss": 0.3826,
+      "step": 9624
+    },
+    {
+      "epoch": 0.42267034040638,
+      "grad_norm": 1.8984375,
+      "learning_rate": 3.1147196528669056e-05,
+      "loss": 0.3936,
+      "step": 9626
+    },
+    {
+      "epoch": 0.4227581588855835,
+      "grad_norm": 1.9453125,
+      "learning_rate": 3.11404904128178e-05,
+      "loss": 0.3715,
+      "step": 9628
+    },
+    {
+      "epoch": 0.42284597736478696,
+      "grad_norm": 1.7578125,
+      "learning_rate": 3.113378382672631e-05,
+      "loss": 0.3611,
+      "step": 9630
+    },
+    {
+      "epoch": 0.4229337958439905,
+      "grad_norm": 1.71875,
+      "learning_rate": 3.112707677090818e-05,
+      "loss": 0.3716,
+      "step": 9632
+    },
+    {
+      "epoch": 0.42302161432319396,
+      "grad_norm": 1.7890625,
+      "learning_rate": 3.1120369245877045e-05,
+      "loss": 0.3768,
+      "step": 9634
+    },
+    {
+      "epoch": 0.42310943280239743,
+      "grad_norm": 1.890625,
+      "learning_rate": 3.111366125214657e-05,
+      "loss": 0.356,
+      "step": 9636
+    },
+    {
+      "epoch": 0.4231972512816009,
+      "grad_norm": 1.8671875,
+      "learning_rate": 3.110695279023045e-05,
+      "loss": 0.3886,
+      "step": 9638
+    },
+    {
+      "epoch": 0.42328506976080443,
+      "grad_norm": 2.46875,
+      "learning_rate": 3.110024386064242e-05,
+      "loss": 0.3937,
+      "step": 9640
+    },
+    {
+      "epoch": 0.4233728882400079,
+      "grad_norm": 1.9609375,
+      "learning_rate": 3.109353446389625e-05,
+      "loss": 0.3737,
+      "step": 9642
+    },
+    {
+      "epoch": 0.4234607067192114,
+      "grad_norm": 1.8984375,
+      "learning_rate": 3.108682460050576e-05,
+      "loss": 0.3738,
+      "step": 9644
+    },
+    {
+      "epoch": 0.42354852519841485,
+      "grad_norm": 2.125,
+      "learning_rate": 3.108011427098477e-05,
+      "loss": 0.38,
+      "step": 9646
+    },
+    {
+      "epoch": 0.4236363436776184,
+      "grad_norm": 2.171875,
+      "learning_rate": 3.107340347584718e-05,
+      "loss": 0.3693,
+      "step": 9648
+    },
+    {
+      "epoch": 0.42372416215682185,
+      "grad_norm": 2.0625,
+      "learning_rate": 3.1066692215606894e-05,
+      "loss": 0.3689,
+      "step": 9650
+    },
+    {
+      "epoch": 0.4238119806360253,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.1059980490777865e-05,
+      "loss": 0.3757,
+      "step": 9652
+    },
+    {
+      "epoch": 0.42389979911522885,
+      "grad_norm": 1.8046875,
+      "learning_rate": 3.1053268301874074e-05,
+      "loss": 0.3936,
+      "step": 9654
+    },
+    {
+      "epoch": 0.4239876175944323,
+      "grad_norm": 2.25,
+      "learning_rate": 3.104655564940954e-05,
+      "loss": 0.3597,
+      "step": 9656
+    },
+    {
+      "epoch": 0.4240754360736358,
+      "grad_norm": 2.046875,
+      "learning_rate": 3.103984253389833e-05,
+      "loss": 0.3453,
+      "step": 9658
+    },
+    {
+      "epoch": 0.42416325455283926,
+      "grad_norm": 2.390625,
+      "learning_rate": 3.103312895585454e-05,
+      "loss": 0.3657,
+      "step": 9660
+    },
+    {
+      "epoch": 0.4242510730320428,
+      "grad_norm": 2.3125,
+      "learning_rate": 3.102641491579228e-05,
+      "loss": 0.3713,
+      "step": 9662
+    },
+    {
+      "epoch": 0.42433889151124626,
+      "grad_norm": 1.859375,
+      "learning_rate": 3.101970041422572e-05,
+      "loss": 0.3867,
+      "step": 9664
+    },
+    {
+      "epoch": 0.42442670999044974,
+      "grad_norm": 1.875,
+      "learning_rate": 3.1012985451669065e-05,
+      "loss": 0.4136,
+      "step": 9666
+    },
+    {
+      "epoch": 0.4245145284696532,
+      "grad_norm": 1.890625,
+      "learning_rate": 3.100627002863654e-05,
+      "loss": 0.4042,
+      "step": 9668
+    },
+    {
+      "epoch": 0.42460234694885673,
+      "grad_norm": 1.953125,
+      "learning_rate": 3.099955414564241e-05,
+      "loss": 0.3814,
+      "step": 9670
+    },
+    {
+      "epoch": 0.4246901654280602,
+      "grad_norm": 1.84375,
+      "learning_rate": 3.099283780320099e-05,
+      "loss": 0.3804,
+      "step": 9672
+    },
+    {
+      "epoch": 0.4247779839072637,
+      "grad_norm": 2.171875,
+      "learning_rate": 3.098612100182662e-05,
+      "loss": 0.3602,
+      "step": 9674
+    },
+    {
+      "epoch": 0.42486580238646715,
+      "grad_norm": 2.0625,
+      "learning_rate": 3.0979403742033664e-05,
+      "loss": 0.3581,
+      "step": 9676
+    },
+    {
+      "epoch": 0.4249536208656707,
+      "grad_norm": 1.8125,
+      "learning_rate": 3.097268602433654e-05,
+      "loss": 0.3673,
+      "step": 9678
+    },
+    {
+      "epoch": 0.42504143934487415,
+      "grad_norm": 1.8515625,
+      "learning_rate": 3.096596784924968e-05,
+      "loss": 0.3822,
+      "step": 9680
+    },
+    {
+      "epoch": 0.4251292578240776,
+      "grad_norm": 1.8984375,
+      "learning_rate": 3.0959249217287574e-05,
+      "loss": 0.3619,
+      "step": 9682
+    },
+    {
+      "epoch": 0.4252170763032811,
+      "grad_norm": 1.8359375,
+      "learning_rate": 3.095253012896474e-05,
+      "loss": 0.4173,
+      "step": 9684
+    },
+    {
+      "epoch": 0.4253048947824846,
+      "grad_norm": 1.984375,
+      "learning_rate": 3.094581058479571e-05,
+      "loss": 0.3643,
+      "step": 9686
+    },
+    {
+      "epoch": 0.4253927132616881,
+      "grad_norm": 2.046875,
+      "learning_rate": 3.0939090585295094e-05,
+      "loss": 0.3801,
+      "step": 9688
+    },
+    {
+      "epoch": 0.42548053174089157,
+      "grad_norm": 1.78125,
+      "learning_rate": 3.09323701309775e-05,
+      "loss": 0.4019,
+      "step": 9690
+    },
+    {
+      "epoch": 0.42556835022009504,
+      "grad_norm": 1.78125,
+      "learning_rate": 3.092564922235757e-05,
+      "loss": 0.4079,
+      "step": 9692
+    },
+    {
+      "epoch": 0.42565616869929856,
+      "grad_norm": 1.703125,
+      "learning_rate": 3.091892785995e-05,
+      "loss": 0.3587,
+      "step": 9694
+    },
+    {
+      "epoch": 0.42574398717850204,
+      "grad_norm": 1.8125,
+      "learning_rate": 3.0912206044269515e-05,
+      "loss": 0.4073,
+      "step": 9696
+    },
+    {
+      "epoch": 0.4258318056577055,
+      "grad_norm": 2.203125,
+      "learning_rate": 3.090548377583088e-05,
+      "loss": 0.3923,
+      "step": 9698
+    },
+    {
+      "epoch": 0.42591962413690904,
+      "grad_norm": 1.984375,
+      "learning_rate": 3.089876105514888e-05,
+      "loss": 0.4088,
+      "step": 9700
+    },
+    {
+      "epoch": 0.4260074426161125,
+      "grad_norm": 1.921875,
+      "learning_rate": 3.089203788273834e-05,
+      "loss": 0.3799,
+      "step": 9702
+    },
+    {
+      "epoch": 0.426095261095316,
+      "grad_norm": 2.28125,
+      "learning_rate": 3.088531425911413e-05,
+      "loss": 0.3668,
+      "step": 9704
+    },
+    {
+      "epoch": 0.42618307957451945,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.087859018479115e-05,
+      "loss": 0.3907,
+      "step": 9706
+    },
+    {
+      "epoch": 0.426270898053723,
+      "grad_norm": 1.8046875,
+      "learning_rate": 3.0871865660284316e-05,
+      "loss": 0.4238,
+      "step": 9708
+    },
+    {
+      "epoch": 0.42635871653292645,
+      "grad_norm": 2.71875,
+      "learning_rate": 3.0865140686108606e-05,
+      "loss": 0.3756,
+      "step": 9710
+    },
+    {
+      "epoch": 0.4264465350121299,
+      "grad_norm": 2.078125,
+      "learning_rate": 3.085841526277901e-05,
+      "loss": 0.3871,
+      "step": 9712
+    },
+    {
+      "epoch": 0.4265343534913334,
+      "grad_norm": 1.875,
+      "learning_rate": 3.085168939081058e-05,
+      "loss": 0.3876,
+      "step": 9714
+    },
+    {
+      "epoch": 0.4266221719705369,
+      "grad_norm": 2.03125,
+      "learning_rate": 3.0844963070718366e-05,
+      "loss": 0.3558,
+      "step": 9716
+    },
+    {
+      "epoch": 0.4267099904497404,
+      "grad_norm": 2.0625,
+      "learning_rate": 3.0838236303017476e-05,
+      "loss": 0.3686,
+      "step": 9718
+    },
+    {
+      "epoch": 0.42679780892894387,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.083150908822306e-05,
+      "loss": 0.3684,
+      "step": 9720
+    },
+    {
+      "epoch": 0.42688562740814734,
+      "grad_norm": 1.9765625,
+      "learning_rate": 3.082478142685027e-05,
+      "loss": 0.3866,
+      "step": 9722
+    },
+    {
+      "epoch": 0.42697344588735087,
+      "grad_norm": 1.9453125,
+      "learning_rate": 3.081805331941433e-05,
+      "loss": 0.3644,
+      "step": 9724
+    },
+    {
+      "epoch": 0.42706126436655434,
+      "grad_norm": 1.8515625,
+      "learning_rate": 3.081132476643047e-05,
+      "loss": 0.4389,
+      "step": 9726
+    },
+    {
+      "epoch": 0.4271490828457578,
+      "grad_norm": 1.8125,
+      "learning_rate": 3.0804595768413964e-05,
+      "loss": 0.3857,
+      "step": 9728
+    },
+    {
+      "epoch": 0.4272369013249613,
+      "grad_norm": 1.890625,
+      "learning_rate": 3.079786632588012e-05,
+      "loss": 0.3536,
+      "step": 9730
+    },
+    {
+      "epoch": 0.4273247198041648,
+      "grad_norm": 1.9609375,
+      "learning_rate": 3.079113643934429e-05,
+      "loss": 0.4163,
+      "step": 9732
+    },
+    {
+      "epoch": 0.4274125382833683,
+      "grad_norm": 1.9765625,
+      "learning_rate": 3.078440610932184e-05,
+      "loss": 0.3987,
+      "step": 9734
+    },
+    {
+      "epoch": 0.42750035676257175,
+      "grad_norm": 2.015625,
+      "learning_rate": 3.077767533632818e-05,
+      "loss": 0.3931,
+      "step": 9736
+    },
+    {
+      "epoch": 0.4275881752417752,
+      "grad_norm": 1.8046875,
+      "learning_rate": 3.077094412087877e-05,
+      "loss": 0.3957,
+      "step": 9738
+    },
+    {
+      "epoch": 0.42767599372097875,
+      "grad_norm": 1.921875,
+      "learning_rate": 3.076421246348906e-05,
+      "loss": 0.342,
+      "step": 9740
+    },
+    {
+      "epoch": 0.4277638122001822,
+      "grad_norm": 1.9140625,
+      "learning_rate": 3.075748036467458e-05,
+      "loss": 0.3825,
+      "step": 9742
+    },
+    {
+      "epoch": 0.4278516306793857,
+      "grad_norm": 1.8828125,
+      "learning_rate": 3.0750747824950885e-05,
+      "loss": 0.3787,
+      "step": 9744
+    },
+    {
+      "epoch": 0.42793944915858917,
+      "grad_norm": 1.8671875,
+      "learning_rate": 3.0744014844833535e-05,
+      "loss": 0.3855,
+      "step": 9746
+    },
+    {
+      "epoch": 0.4280272676377927,
+      "grad_norm": 1.703125,
+      "learning_rate": 3.0737281424838146e-05,
+      "loss": 0.3554,
+      "step": 9748
+    },
+    {
+      "epoch": 0.42811508611699617,
+      "grad_norm": 1.9609375,
+      "learning_rate": 3.073054756548038e-05,
+      "loss": 0.4223,
+      "step": 9750
+    },
+    {
+      "epoch": 0.42820290459619964,
+      "grad_norm": 1.9375,
+      "learning_rate": 3.0723813267275915e-05,
+      "loss": 0.3837,
+      "step": 9752
+    },
+    {
+      "epoch": 0.42829072307540317,
+      "grad_norm": 1.9609375,
+      "learning_rate": 3.071707853074045e-05,
+      "loss": 0.4027,
+      "step": 9754
+    },
+    {
+      "epoch": 0.42837854155460664,
+      "grad_norm": 1.875,
+      "learning_rate": 3.071034335638973e-05,
+      "loss": 0.3697,
+      "step": 9756
+    },
+    {
+      "epoch": 0.4284663600338101,
+      "grad_norm": 2.234375,
+      "learning_rate": 3.070360774473956e-05,
+      "loss": 0.3743,
+      "step": 9758
+    },
+    {
+      "epoch": 0.4285541785130136,
+      "grad_norm": 2.03125,
+      "learning_rate": 3.0696871696305726e-05,
+      "loss": 0.3974,
+      "step": 9760
+    },
+    {
+      "epoch": 0.4286419969922171,
+      "grad_norm": 1.8984375,
+      "learning_rate": 3.069013521160411e-05,
+      "loss": 0.3907,
+      "step": 9762
+    },
+    {
+      "epoch": 0.4287298154714206,
+      "grad_norm": 1.859375,
+      "learning_rate": 3.068339829115057e-05,
+      "loss": 0.427,
+      "step": 9764
+    },
+    {
+      "epoch": 0.42881763395062406,
+      "grad_norm": 1.875,
+      "learning_rate": 3.067666093546102e-05,
+      "loss": 0.3888,
+      "step": 9766
+    },
+    {
+      "epoch": 0.4289054524298275,
+      "grad_norm": 2.0,
+      "learning_rate": 3.066992314505142e-05,
+      "loss": 0.3796,
+      "step": 9768
+    },
+    {
+      "epoch": 0.42899327090903105,
+      "grad_norm": 2.03125,
+      "learning_rate": 3.066318492043774e-05,
+      "loss": 0.357,
+      "step": 9770
+    },
+    {
+      "epoch": 0.4290810893882345,
+      "grad_norm": 1.78125,
+      "learning_rate": 3.065644626213601e-05,
+      "loss": 0.3919,
+      "step": 9772
+    },
+    {
+      "epoch": 0.429168907867438,
+      "grad_norm": 1.890625,
+      "learning_rate": 3.064970717066227e-05,
+      "loss": 0.3633,
+      "step": 9774
+    },
+    {
+      "epoch": 0.42925672634664147,
+      "grad_norm": 1.8359375,
+      "learning_rate": 3.064296764653259e-05,
+      "loss": 0.3975,
+      "step": 9776
+    },
+    {
+      "epoch": 0.429344544825845,
+      "grad_norm": 2.015625,
+      "learning_rate": 3.0636227690263104e-05,
+      "loss": 0.3941,
+      "step": 9778
+    },
+    {
+      "epoch": 0.42943236330504847,
+      "grad_norm": 1.984375,
+      "learning_rate": 3.0629487302369945e-05,
+      "loss": 0.3861,
+      "step": 9780
+    },
+    {
+      "epoch": 0.42952018178425194,
+      "grad_norm": 1.9609375,
+      "learning_rate": 3.0622746483369306e-05,
+      "loss": 0.3714,
+      "step": 9782
+    },
+    {
+      "epoch": 0.4296080002634554,
+      "grad_norm": 1.9609375,
+      "learning_rate": 3.061600523377739e-05,
+      "loss": 0.3706,
+      "step": 9784
+    },
+    {
+      "epoch": 0.42969581874265894,
+      "grad_norm": 2.046875,
+      "learning_rate": 3.0609263554110445e-05,
+      "loss": 0.3677,
+      "step": 9786
+    },
+    {
+      "epoch": 0.4297836372218624,
+      "grad_norm": 1.875,
+      "learning_rate": 3.060252144488476e-05,
+      "loss": 0.3784,
+      "step": 9788
+    },
+    {
+      "epoch": 0.4298714557010659,
+      "grad_norm": 2.203125,
+      "learning_rate": 3.059577890661663e-05,
+      "loss": 0.3611,
+      "step": 9790
+    },
+    {
+      "epoch": 0.42995927418026936,
+      "grad_norm": 2.328125,
+      "learning_rate": 3.058903593982241e-05,
+      "loss": 0.41,
+      "step": 9792
+    },
+    {
+      "epoch": 0.4300470926594729,
+      "grad_norm": 2.0625,
+      "learning_rate": 3.058229254501848e-05,
+      "loss": 0.371,
+      "step": 9794
+    },
+    {
+      "epoch": 0.43013491113867636,
+      "grad_norm": 1.6953125,
+      "learning_rate": 3.057554872272125e-05,
+      "loss": 0.3496,
+      "step": 9796
+    },
+    {
+      "epoch": 0.43022272961787983,
+      "grad_norm": 2.0,
+      "learning_rate": 3.0568804473447164e-05,
+      "loss": 0.3618,
+      "step": 9798
+    },
+    {
+      "epoch": 0.43031054809708336,
+      "grad_norm": 1.9140625,
+      "learning_rate": 3.05620597977127e-05,
+      "loss": 0.3626,
+      "step": 9800
+    },
+    {
+      "epoch": 0.43039836657628683,
+      "grad_norm": 2.078125,
+      "learning_rate": 3.0555314696034356e-05,
+      "loss": 0.3567,
+      "step": 9802
+    },
+    {
+      "epoch": 0.4304861850554903,
+      "grad_norm": 1.875,
+      "learning_rate": 3.054856916892868e-05,
+      "loss": 0.3825,
+      "step": 9804
+    },
+    {
+      "epoch": 0.43057400353469377,
+      "grad_norm": 1.7265625,
+      "learning_rate": 3.0541823216912245e-05,
+      "loss": 0.3721,
+      "step": 9806
+    },
+    {
+      "epoch": 0.4306618220138973,
+      "grad_norm": 1.796875,
+      "learning_rate": 3.0535076840501665e-05,
+      "loss": 0.3852,
+      "step": 9808
+    },
+    {
+      "epoch": 0.43074964049310077,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.052833004021357e-05,
+      "loss": 0.3748,
+      "step": 9810
+    },
+    {
+      "epoch": 0.43083745897230424,
+      "grad_norm": 1.90625,
+      "learning_rate": 3.052158281656465e-05,
+      "loss": 0.4204,
+      "step": 9812
+    },
+    {
+      "epoch": 0.4309252774515077,
+      "grad_norm": 1.7578125,
+      "learning_rate": 3.0514835170071582e-05,
+      "loss": 0.3771,
+      "step": 9814
+    },
+    {
+      "epoch": 0.43101309593071124,
+      "grad_norm": 1.9375,
+      "learning_rate": 3.0508087101251115e-05,
+      "loss": 0.3823,
+      "step": 9816
+    },
+    {
+      "epoch": 0.4311009144099147,
+      "grad_norm": 2.234375,
+      "learning_rate": 3.0501338610620017e-05,
+      "loss": 0.377,
+      "step": 9818
+    },
+    {
+      "epoch": 0.4311887328891182,
+      "grad_norm": 1.8984375,
+      "learning_rate": 3.0494589698695087e-05,
+      "loss": 0.405,
+      "step": 9820
+    },
+    {
+      "epoch": 0.43127655136832166,
+      "grad_norm": 1.96875,
+      "learning_rate": 3.0487840365993164e-05,
+      "loss": 0.3876,
+      "step": 9822
+    },
+    {
+      "epoch": 0.4313643698475252,
+      "grad_norm": 1.9453125,
+      "learning_rate": 3.0481090613031115e-05,
+      "loss": 0.3827,
+      "step": 9824
+    },
+    {
+      "epoch": 0.43145218832672866,
+      "grad_norm": 2.015625,
+      "learning_rate": 3.0474340440325822e-05,
+      "loss": 0.4024,
+      "step": 9826
+    },
+    {
+      "epoch": 0.43154000680593213,
+      "grad_norm": 1.859375,
+      "learning_rate": 3.046758984839424e-05,
+      "loss": 0.3622,
+      "step": 9828
+    },
+    {
+      "epoch": 0.4316278252851356,
+      "grad_norm": 1.765625,
+      "learning_rate": 3.0460838837753304e-05,
+      "loss": 0.4005,
+      "step": 9830
+    },
+    {
+      "epoch": 0.43171564376433913,
+      "grad_norm": 1.7890625,
+      "learning_rate": 3.0454087408920024e-05,
+      "loss": 0.3643,
+      "step": 9832
+    },
+    {
+      "epoch": 0.4318034622435426,
+      "grad_norm": 2.03125,
+      "learning_rate": 3.0447335562411423e-05,
+      "loss": 0.3626,
+      "step": 9834
+    },
+    {
+      "epoch": 0.4318912807227461,
+      "grad_norm": 1.953125,
+      "learning_rate": 3.044058329874456e-05,
+      "loss": 0.3631,
+      "step": 9836
+    },
+    {
+      "epoch": 0.43197909920194955,
+      "grad_norm": 2.03125,
+      "learning_rate": 3.0433830618436528e-05,
+      "loss": 0.413,
+      "step": 9838
+    },
+    {
+      "epoch": 0.4320669176811531,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.042707752200444e-05,
+      "loss": 0.3812,
+      "step": 9840
+    },
+    {
+      "epoch": 0.43215473616035655,
+      "grad_norm": 1.65625,
+      "learning_rate": 3.042032400996545e-05,
+      "loss": 0.3471,
+      "step": 9842
+    },
+    {
+      "epoch": 0.43224255463956,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.0413570082836757e-05,
+      "loss": 0.3952,
+      "step": 9844
+    },
+    {
+      "epoch": 0.4323303731187635,
+      "grad_norm": 1.8046875,
+      "learning_rate": 3.0406815741135563e-05,
+      "loss": 0.3993,
+      "step": 9846
+    },
+    {
+      "epoch": 0.432418191597967,
+      "grad_norm": 2.171875,
+      "learning_rate": 3.0400060985379124e-05,
+      "loss": 0.3543,
+      "step": 9848
+    },
+    {
+      "epoch": 0.4325060100771705,
+      "grad_norm": 1.8359375,
+      "learning_rate": 3.0393305816084728e-05,
+      "loss": 0.4002,
+      "step": 9850
+    },
+    {
+      "epoch": 0.43259382855637396,
+      "grad_norm": 2.171875,
+      "learning_rate": 3.0386550233769673e-05,
+      "loss": 0.3702,
+      "step": 9852
+    },
+    {
+      "epoch": 0.4326816470355775,
+      "grad_norm": 1.8515625,
+      "learning_rate": 3.037979423895131e-05,
+      "loss": 0.402,
+      "step": 9854
+    },
+    {
+      "epoch": 0.43276946551478096,
+      "grad_norm": 1.875,
+      "learning_rate": 3.037303783214701e-05,
+      "loss": 0.3626,
+      "step": 9856
+    },
+    {
+      "epoch": 0.43285728399398443,
+      "grad_norm": 2.03125,
+      "learning_rate": 3.03662810138742e-05,
+      "loss": 0.3723,
+      "step": 9858
+    },
+    {
+      "epoch": 0.4329451024731879,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.035952378465029e-05,
+      "loss": 0.3783,
+      "step": 9860
+    },
+    {
+      "epoch": 0.43303292095239143,
+      "grad_norm": 2.03125,
+      "learning_rate": 3.0352766144992768e-05,
+      "loss": 0.3801,
+      "step": 9862
+    },
+    {
+      "epoch": 0.4331207394315949,
+      "grad_norm": 1.9296875,
+      "learning_rate": 3.034600809541913e-05,
+      "loss": 0.3734,
+      "step": 9864
+    },
+    {
+      "epoch": 0.4332085579107984,
+      "grad_norm": 1.859375,
+      "learning_rate": 3.033924963644691e-05,
+      "loss": 0.3632,
+      "step": 9866
+    },
+    {
+      "epoch": 0.43329637639000185,
+      "grad_norm": 1.8359375,
+      "learning_rate": 3.0332490768593675e-05,
+      "loss": 0.3899,
+      "step": 9868
+    },
+    {
+      "epoch": 0.4333841948692054,
+      "grad_norm": 1.921875,
+      "learning_rate": 3.0325731492377015e-05,
+      "loss": 0.3708,
+      "step": 9870
+    },
+    {
+      "epoch": 0.43347201334840885,
+      "grad_norm": 2.109375,
+      "learning_rate": 3.0318971808314566e-05,
+      "loss": 0.3751,
+      "step": 9872
+    },
+    {
+      "epoch": 0.4335598318276123,
+      "grad_norm": 1.8515625,
+      "learning_rate": 3.0312211716923978e-05,
+      "loss": 0.3686,
+      "step": 9874
+    },
+    {
+      "epoch": 0.4336476503068158,
+      "grad_norm": 2.265625,
+      "learning_rate": 3.0305451218722947e-05,
+      "loss": 0.3846,
+      "step": 9876
+    },
+    {
+      "epoch": 0.4337354687860193,
+      "grad_norm": 1.9453125,
+      "learning_rate": 3.0298690314229184e-05,
+      "loss": 0.3559,
+      "step": 9878
+    },
+    {
+      "epoch": 0.4338232872652228,
+      "grad_norm": 2.28125,
+      "learning_rate": 3.029192900396045e-05,
+      "loss": 0.4058,
+      "step": 9880
+    },
+    {
+      "epoch": 0.43391110574442626,
+      "grad_norm": 1.9609375,
+      "learning_rate": 3.0285167288434518e-05,
+      "loss": 0.3496,
+      "step": 9882
+    },
+    {
+      "epoch": 0.43399892422362973,
+      "grad_norm": 1.8515625,
+      "learning_rate": 3.0278405168169215e-05,
+      "loss": 0.3949,
+      "step": 9884
+    },
+    {
+      "epoch": 0.43408674270283326,
+      "grad_norm": 1.828125,
+      "learning_rate": 3.0271642643682378e-05,
+      "loss": 0.3738,
+      "step": 9886
+    },
+    {
+      "epoch": 0.43417456118203673,
+      "grad_norm": 1.75,
+      "learning_rate": 3.0264879715491883e-05,
+      "loss": 0.3839,
+      "step": 9888
+    },
+    {
+      "epoch": 0.4342623796612402,
+      "grad_norm": 2.015625,
+      "learning_rate": 3.0258116384115643e-05,
+      "loss": 0.3591,
+      "step": 9890
+    },
+    {
+      "epoch": 0.4343501981404437,
+      "grad_norm": 1.8359375,
+      "learning_rate": 3.0251352650071578e-05,
+      "loss": 0.3743,
+      "step": 9892
+    },
+    {
+      "epoch": 0.4344380166196472,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.0244588513877676e-05,
+      "loss": 0.3767,
+      "step": 9894
+    },
+    {
+      "epoch": 0.4345258350988507,
+      "grad_norm": 1.890625,
+      "learning_rate": 3.0237823976051925e-05,
+      "loss": 0.3498,
+      "step": 9896
+    },
+    {
+      "epoch": 0.43461365357805415,
+      "grad_norm": 1.78125,
+      "learning_rate": 3.0231059037112363e-05,
+      "loss": 0.3935,
+      "step": 9898
+    },
+    {
+      "epoch": 0.4347014720572576,
+      "grad_norm": 1.734375,
+      "learning_rate": 3.0224293697577045e-05,
+      "loss": 0.3661,
+      "step": 9900
+    },
+    {
+      "epoch": 0.43478929053646115,
+      "grad_norm": 2.046875,
+      "learning_rate": 3.021752795796406e-05,
+      "loss": 0.4047,
+      "step": 9902
+    },
+    {
+      "epoch": 0.4348771090156646,
+      "grad_norm": 1.984375,
+      "learning_rate": 3.021076181879154e-05,
+      "loss": 0.3855,
+      "step": 9904
+    },
+    {
+      "epoch": 0.4349649274948681,
+      "grad_norm": 1.984375,
+      "learning_rate": 3.0203995280577618e-05,
+      "loss": 0.3811,
+      "step": 9906
+    },
+    {
+      "epoch": 0.4350527459740716,
+      "grad_norm": 1.7890625,
+      "learning_rate": 3.0197228343840502e-05,
+      "loss": 0.3624,
+      "step": 9908
+    },
+    {
+      "epoch": 0.4351405644532751,
+      "grad_norm": 1.953125,
+      "learning_rate": 3.0190461009098382e-05,
+      "loss": 0.3668,
+      "step": 9910
+    },
+    {
+      "epoch": 0.43522838293247856,
+      "grad_norm": 1.859375,
+      "learning_rate": 3.018369327686953e-05,
+      "loss": 0.3465,
+      "step": 9912
+    },
+    {
+      "epoch": 0.43531620141168204,
+      "grad_norm": 1.984375,
+      "learning_rate": 3.0176925147672192e-05,
+      "loss": 0.4023,
+      "step": 9914
+    },
+    {
+      "epoch": 0.43540401989088556,
+      "grad_norm": 1.921875,
+      "learning_rate": 3.017015662202468e-05,
+      "loss": 0.3568,
+      "step": 9916
+    },
+    {
+      "epoch": 0.43549183837008903,
+      "grad_norm": 1.7265625,
+      "learning_rate": 3.0163387700445345e-05,
+      "loss": 0.3598,
+      "step": 9918
+    },
+    {
+      "epoch": 0.4355796568492925,
+      "grad_norm": 1.796875,
+      "learning_rate": 3.0156618383452545e-05,
+      "loss": 0.3737,
+      "step": 9920
+    },
+    {
+      "epoch": 0.435667475328496,
+      "grad_norm": 1.8203125,
+      "learning_rate": 3.0149848671564663e-05,
+      "loss": 0.4093,
+      "step": 9922
+    },
+    {
+      "epoch": 0.4357552938076995,
+      "grad_norm": 1.734375,
+      "learning_rate": 3.014307856530015e-05,
+      "loss": 0.3829,
+      "step": 9924
+    },
+    {
+      "epoch": 0.435843112286903,
+      "grad_norm": 1.8828125,
+      "learning_rate": 3.0136308065177434e-05,
+      "loss": 0.3661,
+      "step": 9926
+    },
+    {
+      "epoch": 0.43593093076610645,
+      "grad_norm": 2.078125,
+      "learning_rate": 3.0129537171715016e-05,
+      "loss": 0.389,
+      "step": 9928
+    },
+    {
+      "epoch": 0.4360187492453099,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.0122765885431414e-05,
+      "loss": 0.4172,
+      "step": 9930
+    },
+    {
+      "epoch": 0.43610656772451345,
+      "grad_norm": 1.84375,
+      "learning_rate": 3.0115994206845173e-05,
+      "loss": 0.3846,
+      "step": 9932
+    },
+    {
+      "epoch": 0.4361943862037169,
+      "grad_norm": 1.8828125,
+      "learning_rate": 3.010922213647487e-05,
+      "loss": 0.3701,
+      "step": 9934
+    },
+    {
+      "epoch": 0.4362822046829204,
+      "grad_norm": 2.09375,
+      "learning_rate": 3.0102449674839117e-05,
+      "loss": 0.3837,
+      "step": 9936
+    },
+    {
+      "epoch": 0.43637002316212387,
+      "grad_norm": 2.0625,
+      "learning_rate": 3.0095676822456532e-05,
+      "loss": 0.394,
+      "step": 9938
+    },
+    {
+      "epoch": 0.4364578416413274,
+      "grad_norm": 1.890625,
+      "learning_rate": 3.00889035798458e-05,
+      "loss": 0.3849,
+      "step": 9940
+    },
+    {
+      "epoch": 0.43654566012053087,
+      "grad_norm": 1.796875,
+      "learning_rate": 3.008212994752561e-05,
+      "loss": 0.3773,
+      "step": 9942
+    },
+    {
+      "epoch": 0.43663347859973434,
+      "grad_norm": 1.9375,
+      "learning_rate": 3.007535592601469e-05,
+      "loss": 0.3691,
+      "step": 9944
+    },
+    {
+      "epoch": 0.4367212970789378,
+      "grad_norm": 1.8125,
+      "learning_rate": 3.00685815158318e-05,
+      "loss": 0.3364,
+      "step": 9946
+    },
+    {
+      "epoch": 0.43680911555814134,
+      "grad_norm": 2.09375,
+      "learning_rate": 3.0061806717495728e-05,
+      "loss": 0.3628,
+      "step": 9948
+    },
+    {
+      "epoch": 0.4368969340373448,
+      "grad_norm": 1.828125,
+      "learning_rate": 3.005503153152528e-05,
+      "loss": 0.3841,
+      "step": 9950
+    },
+    {
+      "epoch": 0.4369847525165483,
+      "grad_norm": 2.296875,
+      "learning_rate": 3.0048255958439303e-05,
+      "loss": 0.347,
+      "step": 9952
+    },
+    {
+      "epoch": 0.4370725709957518,
+      "grad_norm": 1.8828125,
+      "learning_rate": 3.0041479998756673e-05,
+      "loss": 0.386,
+      "step": 9954
+    },
+    {
+      "epoch": 0.4371603894749553,
+      "grad_norm": 1.84375,
+      "learning_rate": 3.00347036529963e-05,
+      "loss": 0.3891,
+      "step": 9956
+    },
+    {
+      "epoch": 0.43724820795415875,
+      "grad_norm": 1.8671875,
+      "learning_rate": 3.0027926921677108e-05,
+      "loss": 0.4011,
+      "step": 9958
+    },
+    {
+      "epoch": 0.4373360264333622,
+      "grad_norm": 1.9921875,
+      "learning_rate": 3.0021149805318072e-05,
+      "loss": 0.4107,
+      "step": 9960
+    },
+    {
+      "epoch": 0.43742384491256575,
+      "grad_norm": 2.078125,
+      "learning_rate": 3.001437230443818e-05,
+      "loss": 0.3708,
+      "step": 9962
+    },
+    {
+      "epoch": 0.4375116633917692,
+      "grad_norm": 1.9140625,
+      "learning_rate": 3.0007594419556456e-05,
+      "loss": 0.3526,
+      "step": 9964
+    },
+    {
+      "epoch": 0.4375994818709727,
+      "grad_norm": 2.3125,
+      "learning_rate": 3.0000816151191952e-05,
+      "loss": 0.3973,
+      "step": 9966
+    },
+    {
+      "epoch": 0.43768730035017617,
+      "grad_norm": 1.953125,
+      "learning_rate": 2.9994037499863747e-05,
+      "loss": 0.3634,
+      "step": 9968
+    },
+    {
+      "epoch": 0.4377751188293797,
+      "grad_norm": 1.875,
+      "learning_rate": 2.998725846609095e-05,
+      "loss": 0.3703,
+      "step": 9970
+    },
+    {
+      "epoch": 0.43786293730858317,
+      "grad_norm": 1.9296875,
+      "learning_rate": 2.9980479050392702e-05,
+      "loss": 0.382,
+      "step": 9972
+    },
+    {
+      "epoch": 0.43795075578778664,
+      "grad_norm": 2.234375,
+      "learning_rate": 2.9973699253288186e-05,
+      "loss": 0.3804,
+      "step": 9974
+    },
+    {
+      "epoch": 0.4380385742669901,
+      "grad_norm": 2.109375,
+      "learning_rate": 2.996691907529658e-05,
+      "loss": 0.3785,
+      "step": 9976
+    },
+    {
+      "epoch": 0.43812639274619364,
+      "grad_norm": 1.7578125,
+      "learning_rate": 2.996013851693712e-05,
+      "loss": 0.3478,
+      "step": 9978
+    },
+    {
+      "epoch": 0.4382142112253971,
+      "grad_norm": 1.7890625,
+      "learning_rate": 2.9953357578729064e-05,
+      "loss": 0.3744,
+      "step": 9980
+    },
+    {
+      "epoch": 0.4383020297046006,
+      "grad_norm": 1.9140625,
+      "learning_rate": 2.99465762611917e-05,
+      "loss": 0.3685,
+      "step": 9982
+    },
+    {
+      "epoch": 0.43838984818380405,
+      "grad_norm": 2.125,
+      "learning_rate": 2.9939794564844335e-05,
+      "loss": 0.3512,
+      "step": 9984
+    },
+    {
+      "epoch": 0.4384776666630076,
+      "grad_norm": 2.546875,
+      "learning_rate": 2.993301249020633e-05,
+      "loss": 0.3636,
+      "step": 9986
+    },
+    {
+      "epoch": 0.43856548514221105,
+      "grad_norm": 2.125,
+      "learning_rate": 2.9926230037797036e-05,
+      "loss": 0.3955,
+      "step": 9988
+    },
+    {
+      "epoch": 0.4386533036214145,
+      "grad_norm": 1.875,
+      "learning_rate": 2.9919447208135865e-05,
+      "loss": 0.383,
+      "step": 9990
+    },
+    {
+      "epoch": 0.438741122100618,
+      "grad_norm": 2.84375,
+      "learning_rate": 2.9912664001742246e-05,
+      "loss": 0.3825,
+      "step": 9992
+    },
+    {
+      "epoch": 0.4388289405798215,
+      "grad_norm": 2.390625,
+      "learning_rate": 2.9905880419135646e-05,
+      "loss": 0.3879,
+      "step": 9994
+    },
+    {
+      "epoch": 0.438916759059025,
+      "grad_norm": 2.109375,
+      "learning_rate": 2.9899096460835545e-05,
+      "loss": 0.3831,
+      "step": 9996
+    },
+    {
+      "epoch": 0.43900457753822847,
+      "grad_norm": 1.6796875,
+      "learning_rate": 2.9892312127361464e-05,
+      "loss": 0.3886,
+      "step": 9998
+    },
+    {
+      "epoch": 0.43909239601743194,
+      "grad_norm": 2.953125,
+      "learning_rate": 2.988552741923295e-05,
+      "loss": 0.3905,
+      "step": 10000
+    },
+    {
+      "epoch": 0.43918021449663547,
+      "grad_norm": 2.46875,
+      "learning_rate": 2.9878742336969568e-05,
+      "loss": 0.367,
+      "step": 10002
+    },
+    {
+      "epoch": 0.43926803297583894,
+      "grad_norm": 2.15625,
+      "learning_rate": 2.987195688109093e-05,
+      "loss": 0.4148,
+      "step": 10004
+    },
+    {
+      "epoch": 0.4393558514550424,
+      "grad_norm": 1.9453125,
+      "learning_rate": 2.9865171052116664e-05,
+      "loss": 0.3781,
+      "step": 10006
+    },
+    {
+      "epoch": 0.43944366993424594,
+      "grad_norm": 2.359375,
+      "learning_rate": 2.9858384850566435e-05,
+      "loss": 0.3622,
+      "step": 10008
+    },
+    {
+      "epoch": 0.4395314884134494,
+      "grad_norm": 2.25,
+      "learning_rate": 2.9851598276959935e-05,
+      "loss": 0.3743,
+      "step": 10010
+    },
+    {
+      "epoch": 0.4396193068926529,
+      "grad_norm": 1.765625,
+      "learning_rate": 2.984481133181688e-05,
+      "loss": 0.4079,
+      "step": 10012
+    },
+    {
+      "epoch": 0.43970712537185636,
+      "grad_norm": 1.9765625,
+      "learning_rate": 2.9838024015657e-05,
+      "loss": 0.3881,
+      "step": 10014
+    },
+    {
+      "epoch": 0.4397949438510599,
+      "grad_norm": 2.0625,
+      "learning_rate": 2.9831236329000084e-05,
+      "loss": 0.3975,
+      "step": 10016
+    },
+    {
+      "epoch": 0.43988276233026335,
+      "grad_norm": 2.140625,
+      "learning_rate": 2.9824448272365928e-05,
+      "loss": 0.3801,
+      "step": 10018
+    },
+    {
+      "epoch": 0.4399705808094668,
+      "grad_norm": 2.046875,
+      "learning_rate": 2.981765984627437e-05,
+      "loss": 0.3606,
+      "step": 10020
+    },
+    {
+      "epoch": 0.4400583992886703,
+      "grad_norm": 2.046875,
+      "learning_rate": 2.981087105124527e-05,
+      "loss": 0.3692,
+      "step": 10022
+    },
+    {
+      "epoch": 0.4401462177678738,
+      "grad_norm": 1.8828125,
+      "learning_rate": 2.9804081887798508e-05,
+      "loss": 0.3693,
+      "step": 10024
+    },
+    {
+      "epoch": 0.4402340362470773,
+      "grad_norm": 1.6875,
+      "learning_rate": 2.9797292356454004e-05,
+      "loss": 0.3681,
+      "step": 10026
+    },
+    {
+      "epoch": 0.44032185472628077,
+      "grad_norm": 1.671875,
+      "learning_rate": 2.9790502457731706e-05,
+      "loss": 0.3738,
+      "step": 10028
+    },
+    {
+      "epoch": 0.44040967320548424,
+      "grad_norm": 1.8671875,
+      "learning_rate": 2.9783712192151576e-05,
+      "loss": 0.3856,
+      "step": 10030
+    },
+    {
+      "epoch": 0.44049749168468777,
+      "grad_norm": 2.03125,
+      "learning_rate": 2.9776921560233616e-05,
+      "loss": 0.3701,
+      "step": 10032
+    },
+    {
+      "epoch": 0.44058531016389124,
+      "grad_norm": 2.0,
+      "learning_rate": 2.9770130562497867e-05,
+      "loss": 0.3902,
+      "step": 10034
+    },
+    {
+      "epoch": 0.4406731286430947,
+      "grad_norm": 1.7734375,
+      "learning_rate": 2.9763339199464374e-05,
+      "loss": 0.3585,
+      "step": 10036
+    },
+    {
+      "epoch": 0.4407609471222982,
+      "grad_norm": 1.8828125,
+      "learning_rate": 2.9756547471653218e-05,
+      "loss": 0.3656,
+      "step": 10038
+    },
+    {
+      "epoch": 0.4408487656015017,
+      "grad_norm": 1.9140625,
+      "learning_rate": 2.9749755379584515e-05,
+      "loss": 0.3603,
+      "step": 10040
+    },
+    {
+      "epoch": 0.4409365840807052,
+      "grad_norm": 1.828125,
+      "learning_rate": 2.9742962923778417e-05,
+      "loss": 0.4048,
+      "step": 10042
+    },
+    {
+      "epoch": 0.44102440255990866,
+      "grad_norm": 2.09375,
+      "learning_rate": 2.9736170104755075e-05,
+      "loss": 0.4039,
+      "step": 10044
+    },
+    {
+      "epoch": 0.44111222103911213,
+      "grad_norm": 1.921875,
+      "learning_rate": 2.9729376923034684e-05,
+      "loss": 0.3617,
+      "step": 10046
+    },
+    {
+      "epoch": 0.44120003951831566,
+      "grad_norm": 2.1875,
+      "learning_rate": 2.9722583379137493e-05,
+      "loss": 0.4059,
+      "step": 10048
+    },
+    {
+      "epoch": 0.44128785799751913,
+      "grad_norm": 1.9609375,
+      "learning_rate": 2.9715789473583715e-05,
+      "loss": 0.376,
+      "step": 10050
+    },
+    {
+      "epoch": 0.4413756764767226,
+      "grad_norm": 1.859375,
+      "learning_rate": 2.9708995206893658e-05,
+      "loss": 0.351,
+      "step": 10052
+    },
+    {
+      "epoch": 0.44146349495592613,
+      "grad_norm": 1.875,
+      "learning_rate": 2.970220057958762e-05,
+      "loss": 0.3672,
+      "step": 10054
+    },
+    {
+      "epoch": 0.4415513134351296,
+      "grad_norm": 1.9375,
+      "learning_rate": 2.9695405592185925e-05,
+      "loss": 0.36,
+      "step": 10056
+    },
+    {
+      "epoch": 0.44163913191433307,
+      "grad_norm": 2.1875,
+      "learning_rate": 2.968861024520896e-05,
+      "loss": 0.3663,
+      "step": 10058
+    },
+    {
+      "epoch": 0.44172695039353654,
+      "grad_norm": 1.9609375,
+      "learning_rate": 2.9681814539177094e-05,
+      "loss": 0.3859,
+      "step": 10060
+    },
+    {
+      "epoch": 0.44181476887274007,
+      "grad_norm": 2.25,
+      "learning_rate": 2.9675018474610743e-05,
+      "loss": 0.3617,
+      "step": 10062
+    },
+    {
+      "epoch": 0.44190258735194354,
+      "grad_norm": 1.75,
+      "learning_rate": 2.9668222052030353e-05,
+      "loss": 0.364,
+      "step": 10064
+    },
+    {
+      "epoch": 0.441990405831147,
+      "grad_norm": 1.9453125,
+      "learning_rate": 2.9661425271956406e-05,
+      "loss": 0.3872,
+      "step": 10066
+    },
+    {
+      "epoch": 0.4420782243103505,
+      "grad_norm": 1.859375,
+      "learning_rate": 2.965462813490939e-05,
+      "loss": 0.3916,
+      "step": 10068
+    },
+    {
+      "epoch": 0.442166042789554,
+      "grad_norm": 1.8046875,
+      "learning_rate": 2.964783064140984e-05,
+      "loss": 0.3791,
+      "step": 10070
+    },
+    {
+      "epoch": 0.4422538612687575,
+      "grad_norm": 2.171875,
+      "learning_rate": 2.9641032791978307e-05,
+      "loss": 0.3698,
+      "step": 10072
+    },
+    {
+      "epoch": 0.44234167974796096,
+      "grad_norm": 1.75,
+      "learning_rate": 2.9634234587135366e-05,
+      "loss": 0.3751,
+      "step": 10074
+    },
+    {
+      "epoch": 0.44242949822716443,
+      "grad_norm": 1.921875,
+      "learning_rate": 2.9627436027401633e-05,
+      "loss": 0.3814,
+      "step": 10076
+    },
+    {
+      "epoch": 0.44251731670636796,
+      "grad_norm": 1.828125,
+      "learning_rate": 2.9620637113297735e-05,
+      "loss": 0.3538,
+      "step": 10078
+    },
+    {
+      "epoch": 0.44260513518557143,
+      "grad_norm": 2.40625,
+      "learning_rate": 2.961383784534434e-05,
+      "loss": 0.3811,
+      "step": 10080
+    },
+    {
+      "epoch": 0.4426929536647749,
+      "grad_norm": 1.7890625,
+      "learning_rate": 2.9607038224062133e-05,
+      "loss": 0.3739,
+      "step": 10082
+    },
+    {
+      "epoch": 0.4427807721439784,
+      "grad_norm": 1.984375,
+      "learning_rate": 2.9600238249971846e-05,
+      "loss": 0.3835,
+      "step": 10084
+    },
+    {
+      "epoch": 0.4428685906231819,
+      "grad_norm": 1.984375,
+      "learning_rate": 2.9593437923594204e-05,
+      "loss": 0.3694,
+      "step": 10086
+    },
+    {
+      "epoch": 0.4429564091023854,
+      "grad_norm": 1.890625,
+      "learning_rate": 2.958663724544999e-05,
+      "loss": 0.3621,
+      "step": 10088
+    },
+    {
+      "epoch": 0.44304422758158885,
+      "grad_norm": 1.734375,
+      "learning_rate": 2.9579836216059988e-05,
+      "loss": 0.3912,
+      "step": 10090
+    },
+    {
+      "epoch": 0.4431320460607923,
+      "grad_norm": 2.28125,
+      "learning_rate": 2.9573034835945028e-05,
+      "loss": 0.3912,
+      "step": 10092
+    },
+    {
+      "epoch": 0.44321986453999584,
+      "grad_norm": 1.8984375,
+      "learning_rate": 2.9566233105625973e-05,
+      "loss": 0.3883,
+      "step": 10094
+    },
+    {
+      "epoch": 0.4433076830191993,
+      "grad_norm": 1.9140625,
+      "learning_rate": 2.955943102562369e-05,
+      "loss": 0.361,
+      "step": 10096
+    },
+    {
+      "epoch": 0.4433955014984028,
+      "grad_norm": 1.9140625,
+      "learning_rate": 2.9552628596459086e-05,
+      "loss": 0.3964,
+      "step": 10098
+    },
+    {
+      "epoch": 0.44348331997760626,
+      "grad_norm": 1.9453125,
+      "learning_rate": 2.9545825818653087e-05,
+      "loss": 0.3786,
+      "step": 10100
+    },
+    {
+      "epoch": 0.4435711384568098,
+      "grad_norm": 1.859375,
+      "learning_rate": 2.9539022692726665e-05,
+      "loss": 0.3958,
+      "step": 10102
+    },
+    {
+      "epoch": 0.44365895693601326,
+      "grad_norm": 1.90625,
+      "learning_rate": 2.9532219219200797e-05,
+      "loss": 0.3772,
+      "step": 10104
+    },
+    {
+      "epoch": 0.44374677541521673,
+      "grad_norm": 1.921875,
+      "learning_rate": 2.952541539859649e-05,
+      "loss": 0.3753,
+      "step": 10106
+    },
+    {
+      "epoch": 0.44383459389442026,
+      "grad_norm": 1.7890625,
+      "learning_rate": 2.951861123143479e-05,
+      "loss": 0.373,
+      "step": 10108
+    },
+    {
+      "epoch": 0.44392241237362373,
+      "grad_norm": 2.03125,
+      "learning_rate": 2.9511806718236764e-05,
+      "loss": 0.3747,
+      "step": 10110
+    },
+    {
+      "epoch": 0.4440102308528272,
+      "grad_norm": 1.8203125,
+      "learning_rate": 2.9505001859523484e-05,
+      "loss": 0.357,
+      "step": 10112
+    },
+    {
+      "epoch": 0.4440980493320307,
+      "grad_norm": 1.84375,
+      "learning_rate": 2.949819665581609e-05,
+      "loss": 0.3729,
+      "step": 10114
+    },
+    {
+      "epoch": 0.4441858678112342,
+      "grad_norm": 2.03125,
+      "learning_rate": 2.9491391107635715e-05,
+      "loss": 0.3629,
+      "step": 10116
+    },
+    {
+      "epoch": 0.4442736862904377,
+      "grad_norm": 1.953125,
+      "learning_rate": 2.9484585215503537e-05,
+      "loss": 0.3556,
+      "step": 10118
+    },
+    {
+      "epoch": 0.44436150476964115,
+      "grad_norm": 2.046875,
+      "learning_rate": 2.9477778979940745e-05,
+      "loss": 0.3578,
+      "step": 10120
+    },
+    {
+      "epoch": 0.4444493232488446,
+      "grad_norm": 1.734375,
+      "learning_rate": 2.947097240146857e-05,
+      "loss": 0.3788,
+      "step": 10122
+    },
+    {
+      "epoch": 0.44453714172804815,
+      "grad_norm": 1.9296875,
+      "learning_rate": 2.9464165480608252e-05,
+      "loss": 0.3716,
+      "step": 10124
+    },
+    {
+      "epoch": 0.4446249602072516,
+      "grad_norm": 2.078125,
+      "learning_rate": 2.945735821788107e-05,
+      "loss": 0.3521,
+      "step": 10126
+    },
+    {
+      "epoch": 0.4447127786864551,
+      "grad_norm": 1.9453125,
+      "learning_rate": 2.945055061380833e-05,
+      "loss": 0.3624,
+      "step": 10128
+    },
+    {
+      "epoch": 0.44480059716565856,
+      "grad_norm": 1.9296875,
+      "learning_rate": 2.9443742668911357e-05,
+      "loss": 0.3788,
+      "step": 10130
+    },
+    {
+      "epoch": 0.4448884156448621,
+      "grad_norm": 2.03125,
+      "learning_rate": 2.9436934383711508e-05,
+      "loss": 0.3681,
+      "step": 10132
+    },
+    {
+      "epoch": 0.44497623412406556,
+      "grad_norm": 1.9140625,
+      "learning_rate": 2.943012575873016e-05,
+      "loss": 0.3704,
+      "step": 10134
+    },
+    {
+      "epoch": 0.44506405260326903,
+      "grad_norm": 1.8828125,
+      "learning_rate": 2.9423316794488716e-05,
+      "loss": 0.4046,
+      "step": 10136
+    },
+    {
+      "epoch": 0.4451518710824725,
+      "grad_norm": 1.6640625,
+      "learning_rate": 2.9416507491508606e-05,
+      "loss": 0.3611,
+      "step": 10138
+    },
+    {
+      "epoch": 0.44523968956167603,
+      "grad_norm": 1.9296875,
+      "learning_rate": 2.9409697850311296e-05,
+      "loss": 0.3797,
+      "step": 10140
+    },
+    {
+      "epoch": 0.4453275080408795,
+      "grad_norm": 2.171875,
+      "learning_rate": 2.940288787141827e-05,
+      "loss": 0.4003,
+      "step": 10142
+    },
+    {
+      "epoch": 0.445415326520083,
+      "grad_norm": 2.1875,
+      "learning_rate": 2.9396077555351038e-05,
+      "loss": 0.3505,
+      "step": 10144
+    },
+    {
+      "epoch": 0.44550314499928645,
+      "grad_norm": 2.265625,
+      "learning_rate": 2.9389266902631137e-05,
+      "loss": 0.3681,
+      "step": 10146
+    },
+    {
+      "epoch": 0.44559096347849,
+      "grad_norm": 2.140625,
+      "learning_rate": 2.9382455913780115e-05,
+      "loss": 0.3881,
+      "step": 10148
+    },
+    {
+      "epoch": 0.44567878195769345,
+      "grad_norm": 1.765625,
+      "learning_rate": 2.9375644589319572e-05,
+      "loss": 0.3798,
+      "step": 10150
+    },
+    {
+      "epoch": 0.4457666004368969,
+      "grad_norm": 2.15625,
+      "learning_rate": 2.936883292977112e-05,
+      "loss": 0.39,
+      "step": 10152
+    },
+    {
+      "epoch": 0.44585441891610045,
+      "grad_norm": 2.09375,
+      "learning_rate": 2.936202093565639e-05,
+      "loss": 0.3701,
+      "step": 10154
+    },
+    {
+      "epoch": 0.4459422373953039,
+      "grad_norm": 2.0625,
+      "learning_rate": 2.9355208607497053e-05,
+      "loss": 0.3647,
+      "step": 10156
+    },
+    {
+      "epoch": 0.4460300558745074,
+      "grad_norm": 1.9609375,
+      "learning_rate": 2.93483959458148e-05,
+      "loss": 0.3879,
+      "step": 10158
+    },
+    {
+      "epoch": 0.44611787435371086,
+      "grad_norm": 1.875,
+      "learning_rate": 2.9341582951131343e-05,
+      "loss": 0.3899,
+      "step": 10160
+    },
+    {
+      "epoch": 0.4462056928329144,
+      "grad_norm": 2.015625,
+      "learning_rate": 2.9334769623968417e-05,
+      "loss": 0.3776,
+      "step": 10162
+    },
+    {
+      "epoch": 0.44629351131211786,
+      "grad_norm": 2.0625,
+      "learning_rate": 2.9327955964847798e-05,
+      "loss": 0.3546,
+      "step": 10164
+    },
+    {
+      "epoch": 0.44638132979132134,
+      "grad_norm": 2.15625,
+      "learning_rate": 2.9321141974291277e-05,
+      "loss": 0.3971,
+      "step": 10166
+    },
+    {
+      "epoch": 0.4464691482705248,
+      "grad_norm": 2.1875,
+      "learning_rate": 2.931432765282066e-05,
+      "loss": 0.3852,
+      "step": 10168
+    },
+    {
+      "epoch": 0.44655696674972833,
+      "grad_norm": 1.9921875,
+      "learning_rate": 2.9307513000957797e-05,
+      "loss": 0.3891,
+      "step": 10170
+    },
+    {
+      "epoch": 0.4466447852289318,
+      "grad_norm": 1.9296875,
+      "learning_rate": 2.930069801922457e-05,
+      "loss": 0.3529,
+      "step": 10172
+    },
+    {
+      "epoch": 0.4467326037081353,
+      "grad_norm": 1.9453125,
+      "learning_rate": 2.9293882708142846e-05,
+      "loss": 0.3754,
+      "step": 10174
+    },
+    {
+      "epoch": 0.44682042218733875,
+      "grad_norm": 1.71875,
+      "learning_rate": 2.9287067068234554e-05,
+      "loss": 0.3734,
+      "step": 10176
+    },
+    {
+      "epoch": 0.4469082406665423,
+      "grad_norm": 1.859375,
+      "learning_rate": 2.928025110002164e-05,
+      "loss": 0.378,
+      "step": 10178
+    },
+    {
+      "epoch": 0.44699605914574575,
+      "grad_norm": 1.7578125,
+      "learning_rate": 2.9273434804026072e-05,
+      "loss": 0.3614,
+      "step": 10180
+    },
+    {
+      "epoch": 0.4470838776249492,
+      "grad_norm": 1.75,
+      "learning_rate": 2.9266618180769846e-05,
+      "loss": 0.3388,
+      "step": 10182
+    },
+    {
+      "epoch": 0.4471716961041527,
+      "grad_norm": 1.875,
+      "learning_rate": 2.9259801230774974e-05,
+      "loss": 0.3852,
+      "step": 10184
+    },
+    {
+      "epoch": 0.4472595145833562,
+      "grad_norm": 1.7890625,
+      "learning_rate": 2.92529839545635e-05,
+      "loss": 0.3585,
+      "step": 10186
+    },
+    {
+      "epoch": 0.4473473330625597,
+      "grad_norm": 1.8984375,
+      "learning_rate": 2.9246166352657494e-05,
+      "loss": 0.3929,
+      "step": 10188
+    },
+    {
+      "epoch": 0.44743515154176317,
+      "grad_norm": 2.1875,
+      "learning_rate": 2.923934842557905e-05,
+      "loss": 0.3593,
+      "step": 10190
+    },
+    {
+      "epoch": 0.44752297002096664,
+      "grad_norm": 1.703125,
+      "learning_rate": 2.923253017385029e-05,
+      "loss": 0.4071,
+      "step": 10192
+    },
+    {
+      "epoch": 0.44761078850017016,
+      "grad_norm": 1.984375,
+      "learning_rate": 2.9225711597993362e-05,
+      "loss": 0.34,
+      "step": 10194
+    },
+    {
+      "epoch": 0.44769860697937364,
+      "grad_norm": 2.140625,
+      "learning_rate": 2.9218892698530427e-05,
+      "loss": 0.3652,
+      "step": 10196
+    },
+    {
+      "epoch": 0.4477864254585771,
+      "grad_norm": 1.9140625,
+      "learning_rate": 2.9212073475983663e-05,
+      "loss": 0.398,
+      "step": 10198
+    },
+    {
+      "epoch": 0.4478742439377806,
+      "grad_norm": 2.140625,
+      "learning_rate": 2.9205253930875315e-05,
+      "loss": 0.3866,
+      "step": 10200
+    },
+    {
+      "epoch": 0.4479620624169841,
+      "grad_norm": 1.8828125,
+      "learning_rate": 2.9198434063727602e-05,
+      "loss": 0.4063,
+      "step": 10202
+    },
+    {
+      "epoch": 0.4480498808961876,
+      "grad_norm": 2.25,
+      "learning_rate": 2.919161387506281e-05,
+      "loss": 0.4008,
+      "step": 10204
+    },
+    {
+      "epoch": 0.44813769937539105,
+      "grad_norm": 1.671875,
+      "learning_rate": 2.9184793365403217e-05,
+      "loss": 0.3405,
+      "step": 10206
+    },
+    {
+      "epoch": 0.4482255178545946,
+      "grad_norm": 2.125,
+      "learning_rate": 2.917797253527116e-05,
+      "loss": 0.4057,
+      "step": 10208
+    },
+    {
+      "epoch": 0.44831333633379805,
+      "grad_norm": 1.671875,
+      "learning_rate": 2.917115138518895e-05,
+      "loss": 0.3616,
+      "step": 10210
+    },
+    {
+      "epoch": 0.4484011548130015,
+      "grad_norm": 1.6640625,
+      "learning_rate": 2.916432991567897e-05,
+      "loss": 0.3644,
+      "step": 10212
+    },
+    {
+      "epoch": 0.448488973292205,
+      "grad_norm": 1.90625,
+      "learning_rate": 2.9157508127263612e-05,
+      "loss": 0.3851,
+      "step": 10214
+    },
+    {
+      "epoch": 0.4485767917714085,
+      "grad_norm": 1.859375,
+      "learning_rate": 2.915068602046528e-05,
+      "loss": 0.3828,
+      "step": 10216
+    },
+    {
+      "epoch": 0.448664610250612,
+      "grad_norm": 1.8515625,
+      "learning_rate": 2.9143863595806413e-05,
+      "loss": 0.382,
+      "step": 10218
+    },
+    {
+      "epoch": 0.44875242872981547,
+      "grad_norm": 2.03125,
+      "learning_rate": 2.9137040853809487e-05,
+      "loss": 0.3827,
+      "step": 10220
+    },
+    {
+      "epoch": 0.44884024720901894,
+      "grad_norm": 1.90625,
+      "learning_rate": 2.9130217794996977e-05,
+      "loss": 0.3805,
+      "step": 10222
+    },
+    {
+      "epoch": 0.44892806568822247,
+      "grad_norm": 1.734375,
+      "learning_rate": 2.9123394419891396e-05,
+      "loss": 0.3908,
+      "step": 10224
+    },
+    {
+      "epoch": 0.44901588416742594,
+      "grad_norm": 2.03125,
+      "learning_rate": 2.911657072901529e-05,
+      "loss": 0.3937,
+      "step": 10226
+    },
+    {
+      "epoch": 0.4491037026466294,
+      "grad_norm": 1.8984375,
+      "learning_rate": 2.91097467228912e-05,
+      "loss": 0.3825,
+      "step": 10228
+    },
+    {
+      "epoch": 0.4491915211258329,
+      "grad_norm": 2.078125,
+      "learning_rate": 2.9102922402041728e-05,
+      "loss": 0.388,
+      "step": 10230
+    },
+    {
+      "epoch": 0.4492793396050364,
+      "grad_norm": 1.859375,
+      "learning_rate": 2.9096097766989478e-05,
+      "loss": 0.3475,
+      "step": 10232
+    },
+    {
+      "epoch": 0.4493671580842399,
+      "grad_norm": 1.7734375,
+      "learning_rate": 2.9089272818257073e-05,
+      "loss": 0.4138,
+      "step": 10234
+    },
+    {
+      "epoch": 0.44945497656344335,
+      "grad_norm": 2.0,
+      "learning_rate": 2.908244755636717e-05,
+      "loss": 0.3681,
+      "step": 10236
+    },
+    {
+      "epoch": 0.4495427950426468,
+      "grad_norm": 1.75,
+      "learning_rate": 2.907562198184246e-05,
+      "loss": 0.3713,
+      "step": 10238
+    },
+    {
+      "epoch": 0.44963061352185035,
+      "grad_norm": 1.765625,
+      "learning_rate": 2.906879609520564e-05,
+      "loss": 0.3717,
+      "step": 10240
+    },
+    {
+      "epoch": 0.4497184320010538,
+      "grad_norm": 1.859375,
+      "learning_rate": 2.9061969896979447e-05,
+      "loss": 0.3774,
+      "step": 10242
+    },
+    {
+      "epoch": 0.4498062504802573,
+      "grad_norm": 1.84375,
+      "learning_rate": 2.9055143387686624e-05,
+      "loss": 0.3603,
+      "step": 10244
+    },
+    {
+      "epoch": 0.44989406895946077,
+      "grad_norm": 1.796875,
+      "learning_rate": 2.9048316567849947e-05,
+      "loss": 0.3547,
+      "step": 10246
+    },
+    {
+      "epoch": 0.4499818874386643,
+      "grad_norm": 1.765625,
+      "learning_rate": 2.9041489437992215e-05,
+      "loss": 0.3901,
+      "step": 10248
+    },
+    {
+      "epoch": 0.45006970591786777,
+      "grad_norm": 2.09375,
+      "learning_rate": 2.9034661998636248e-05,
+      "loss": 0.3638,
+      "step": 10250
+    },
+    {
+      "epoch": 0.45015752439707124,
+      "grad_norm": 1.8671875,
+      "learning_rate": 2.9027834250304904e-05,
+      "loss": 0.3831,
+      "step": 10252
+    },
+    {
+      "epoch": 0.45024534287627477,
+      "grad_norm": 1.8203125,
+      "learning_rate": 2.9021006193521043e-05,
+      "loss": 0.3749,
+      "step": 10254
+    },
+    {
+      "epoch": 0.45033316135547824,
+      "grad_norm": 1.890625,
+      "learning_rate": 2.901417782880757e-05,
+      "loss": 0.3908,
+      "step": 10256
+    },
+    {
+      "epoch": 0.4504209798346817,
+      "grad_norm": 1.8125,
+      "learning_rate": 2.9007349156687404e-05,
+      "loss": 0.3754,
+      "step": 10258
+    },
+    {
+      "epoch": 0.4505087983138852,
+      "grad_norm": 1.8203125,
+      "learning_rate": 2.900052017768346e-05,
+      "loss": 0.375,
+      "step": 10260
+    },
+    {
+      "epoch": 0.4505966167930887,
+      "grad_norm": 2.015625,
+      "learning_rate": 2.899369089231873e-05,
+      "loss": 0.3856,
+      "step": 10262
+    },
+    {
+      "epoch": 0.4506844352722922,
+      "grad_norm": 1.8515625,
+      "learning_rate": 2.8986861301116196e-05,
+      "loss": 0.3618,
+      "step": 10264
+    },
+    {
+      "epoch": 0.45077225375149566,
+      "grad_norm": 2.171875,
+      "learning_rate": 2.8980031404598862e-05,
+      "loss": 0.414,
+      "step": 10266
+    },
+    {
+      "epoch": 0.4508600722306991,
+      "grad_norm": 1.8203125,
+      "learning_rate": 2.897320120328978e-05,
+      "loss": 0.4064,
+      "step": 10268
+    },
+    {
+      "epoch": 0.45094789070990265,
+      "grad_norm": 1.828125,
+      "learning_rate": 2.8966370697711988e-05,
+      "loss": 0.3708,
+      "step": 10270
+    },
+    {
+      "epoch": 0.4510357091891061,
+      "grad_norm": 2.09375,
+      "learning_rate": 2.895953988838859e-05,
+      "loss": 0.3726,
+      "step": 10272
+    },
+    {
+      "epoch": 0.4511235276683096,
+      "grad_norm": 1.703125,
+      "learning_rate": 2.8952708775842664e-05,
+      "loss": 0.3504,
+      "step": 10274
+    },
+    {
+      "epoch": 0.45121134614751307,
+      "grad_norm": 1.8203125,
+      "learning_rate": 2.8945877360597352e-05,
+      "loss": 0.3642,
+      "step": 10276
+    },
+    {
+      "epoch": 0.4512991646267166,
+      "grad_norm": 2.15625,
+      "learning_rate": 2.8939045643175812e-05,
+      "loss": 0.3775,
+      "step": 10278
+    },
+    {
+      "epoch": 0.45138698310592007,
+      "grad_norm": 1.828125,
+      "learning_rate": 2.8932213624101207e-05,
+      "loss": 0.3676,
+      "step": 10280
+    },
+    {
+      "epoch": 0.45147480158512354,
+      "grad_norm": 2.921875,
+      "learning_rate": 2.8925381303896747e-05,
+      "loss": 0.3688,
+      "step": 10282
+    },
+    {
+      "epoch": 0.451562620064327,
+      "grad_norm": 2.0625,
+      "learning_rate": 2.8918548683085643e-05,
+      "loss": 0.3705,
+      "step": 10284
+    },
+    {
+      "epoch": 0.45165043854353054,
+      "grad_norm": 2.015625,
+      "learning_rate": 2.891171576219114e-05,
+      "loss": 0.3961,
+      "step": 10286
+    },
+    {
+      "epoch": 0.451738257022734,
+      "grad_norm": 1.7890625,
+      "learning_rate": 2.8904882541736512e-05,
+      "loss": 0.3567,
+      "step": 10288
+    },
+    {
+      "epoch": 0.4518260755019375,
+      "grad_norm": 2.484375,
+      "learning_rate": 2.8898049022245034e-05,
+      "loss": 0.3565,
+      "step": 10290
+    },
+    {
+      "epoch": 0.45191389398114096,
+      "grad_norm": 1.8671875,
+      "learning_rate": 2.889121520424003e-05,
+      "loss": 0.3552,
+      "step": 10292
+    },
+    {
+      "epoch": 0.4520017124603445,
+      "grad_norm": 1.8984375,
+      "learning_rate": 2.888438108824484e-05,
+      "loss": 0.3441,
+      "step": 10294
+    },
+    {
+      "epoch": 0.45208953093954796,
+      "grad_norm": 1.75,
+      "learning_rate": 2.8877546674782806e-05,
+      "loss": 0.3561,
+      "step": 10296
+    },
+    {
+      "epoch": 0.45217734941875143,
+      "grad_norm": 1.734375,
+      "learning_rate": 2.8870711964377322e-05,
+      "loss": 0.365,
+      "step": 10298
+    },
+    {
+      "epoch": 0.4522651678979549,
+      "grad_norm": 2.078125,
+      "learning_rate": 2.8863876957551784e-05,
+      "loss": 0.381,
+      "step": 10300
+    },
+    {
+      "epoch": 0.45235298637715843,
+      "grad_norm": 2.203125,
+      "learning_rate": 2.8857041654829625e-05,
+      "loss": 0.3885,
+      "step": 10302
+    },
+    {
+      "epoch": 0.4524408048563619,
+      "grad_norm": 2.390625,
+      "learning_rate": 2.8850206056734297e-05,
+      "loss": 0.3848,
+      "step": 10304
+    },
+    {
+      "epoch": 0.45252862333556537,
+      "grad_norm": 1.9140625,
+      "learning_rate": 2.8843370163789264e-05,
+      "loss": 0.3426,
+      "step": 10306
+    },
+    {
+      "epoch": 0.4526164418147689,
+      "grad_norm": 1.6328125,
+      "learning_rate": 2.883653397651802e-05,
+      "loss": 0.3648,
+      "step": 10308
+    },
+    {
+      "epoch": 0.45270426029397237,
+      "grad_norm": 1.7265625,
+      "learning_rate": 2.8829697495444087e-05,
+      "loss": 0.365,
+      "step": 10310
+    },
+    {
+      "epoch": 0.45279207877317584,
+      "grad_norm": 1.8984375,
+      "learning_rate": 2.8822860721090995e-05,
+      "loss": 0.3502,
+      "step": 10312
+    },
+    {
+      "epoch": 0.4528798972523793,
+      "grad_norm": 1.828125,
+      "learning_rate": 2.8816023653982317e-05,
+      "loss": 0.3714,
+      "step": 10314
+    },
+    {
+      "epoch": 0.45296771573158284,
+      "grad_norm": 1.8046875,
+      "learning_rate": 2.8809186294641634e-05,
+      "loss": 0.3823,
+      "step": 10316
+    },
+    {
+      "epoch": 0.4530555342107863,
+      "grad_norm": 1.9765625,
+      "learning_rate": 2.8802348643592552e-05,
+      "loss": 0.3837,
+      "step": 10318
+    },
+    {
+      "epoch": 0.4531433526899898,
+      "grad_norm": 1.921875,
+      "learning_rate": 2.8795510701358703e-05,
+      "loss": 0.3675,
+      "step": 10320
+    },
+    {
+      "epoch": 0.45323117116919326,
+      "grad_norm": 1.921875,
+      "learning_rate": 2.878867246846373e-05,
+      "loss": 0.373,
+      "step": 10322
+    },
+    {
+      "epoch": 0.4533189896483968,
+      "grad_norm": 1.8125,
+      "learning_rate": 2.87818339454313e-05,
+      "loss": 0.3807,
+      "step": 10324
+    },
+    {
+      "epoch": 0.45340680812760026,
+      "grad_norm": 1.9375,
+      "learning_rate": 2.877499513278513e-05,
+      "loss": 0.3547,
+      "step": 10326
+    },
+    {
+      "epoch": 0.45349462660680373,
+      "grad_norm": 1.9609375,
+      "learning_rate": 2.876815603104893e-05,
+      "loss": 0.3703,
+      "step": 10328
+    },
+    {
+      "epoch": 0.4535824450860072,
+      "grad_norm": 1.796875,
+      "learning_rate": 2.8761316640746437e-05,
+      "loss": 0.3889,
+      "step": 10330
+    },
+    {
+      "epoch": 0.45367026356521073,
+      "grad_norm": 1.796875,
+      "learning_rate": 2.8754476962401418e-05,
+      "loss": 0.384,
+      "step": 10332
+    },
+    {
+      "epoch": 0.4537580820444142,
+      "grad_norm": 1.8046875,
+      "learning_rate": 2.874763699653765e-05,
+      "loss": 0.3686,
+      "step": 10334
+    },
+    {
+      "epoch": 0.4538459005236177,
+      "grad_norm": 1.796875,
+      "learning_rate": 2.874079674367894e-05,
+      "loss": 0.3623,
+      "step": 10336
+    },
+    {
+      "epoch": 0.45393371900282115,
+      "grad_norm": 1.7734375,
+      "learning_rate": 2.8733956204349117e-05,
+      "loss": 0.3689,
+      "step": 10338
+    },
+    {
+      "epoch": 0.4540215374820247,
+      "grad_norm": 1.828125,
+      "learning_rate": 2.8727115379072034e-05,
+      "loss": 0.3677,
+      "step": 10340
+    },
+    {
+      "epoch": 0.45410935596122814,
+      "grad_norm": 1.8359375,
+      "learning_rate": 2.872027426837156e-05,
+      "loss": 0.3815,
+      "step": 10342
+    },
+    {
+      "epoch": 0.4541971744404316,
+      "grad_norm": 1.984375,
+      "learning_rate": 2.87134328727716e-05,
+      "loss": 0.3619,
+      "step": 10344
+    },
+    {
+      "epoch": 0.4542849929196351,
+      "grad_norm": 1.8203125,
+      "learning_rate": 2.870659119279605e-05,
+      "loss": 0.3892,
+      "step": 10346
+    },
+    {
+      "epoch": 0.4543728113988386,
+      "grad_norm": 1.8203125,
+      "learning_rate": 2.8699749228968865e-05,
+      "loss": 0.4045,
+      "step": 10348
+    },
+    {
+      "epoch": 0.4544606298780421,
+      "grad_norm": 2.0625,
+      "learning_rate": 2.8692906981813993e-05,
+      "loss": 0.4142,
+      "step": 10350
+    },
+    {
+      "epoch": 0.45454844835724556,
+      "grad_norm": 1.65625,
+      "learning_rate": 2.8686064451855422e-05,
+      "loss": 0.3741,
+      "step": 10352
+    },
+    {
+      "epoch": 0.45463626683644903,
+      "grad_norm": 1.9921875,
+      "learning_rate": 2.867922163961715e-05,
+      "loss": 0.3552,
+      "step": 10354
+    },
+    {
+      "epoch": 0.45472408531565256,
+      "grad_norm": 2.109375,
+      "learning_rate": 2.86723785456232e-05,
+      "loss": 0.398,
+      "step": 10356
+    },
+    {
+      "epoch": 0.45481190379485603,
+      "grad_norm": 1.8203125,
+      "learning_rate": 2.866553517039763e-05,
+      "loss": 0.3416,
+      "step": 10358
+    },
+    {
+      "epoch": 0.4548997222740595,
+      "grad_norm": 1.9375,
+      "learning_rate": 2.8658691514464488e-05,
+      "loss": 0.3898,
+      "step": 10360
+    },
+    {
+      "epoch": 0.45498754075326303,
+      "grad_norm": 1.7890625,
+      "learning_rate": 2.8651847578347873e-05,
+      "loss": 0.3765,
+      "step": 10362
+    },
+    {
+      "epoch": 0.4550753592324665,
+      "grad_norm": 2.0,
+      "learning_rate": 2.8645003362571897e-05,
+      "loss": 0.3381,
+      "step": 10364
+    },
+    {
+      "epoch": 0.45516317771167,
+      "grad_norm": 1.8671875,
+      "learning_rate": 2.863815886766069e-05,
+      "loss": 0.3649,
+      "step": 10366
+    },
+    {
+      "epoch": 0.45525099619087345,
+      "grad_norm": 1.9296875,
+      "learning_rate": 2.8631314094138405e-05,
+      "loss": 0.3858,
+      "step": 10368
+    },
+    {
+      "epoch": 0.455338814670077,
+      "grad_norm": 1.9921875,
+      "learning_rate": 2.862446904252922e-05,
+      "loss": 0.3454,
+      "step": 10370
+    },
+    {
+      "epoch": 0.45542663314928045,
+      "grad_norm": 1.953125,
+      "learning_rate": 2.8617623713357326e-05,
+      "loss": 0.3886,
+      "step": 10372
+    },
+    {
+      "epoch": 0.4555144516284839,
+      "grad_norm": 1.7109375,
+      "learning_rate": 2.8610778107146934e-05,
+      "loss": 0.3711,
+      "step": 10374
+    },
+    {
+      "epoch": 0.4556022701076874,
+      "grad_norm": 1.890625,
+      "learning_rate": 2.8603932224422297e-05,
+      "loss": 0.3676,
+      "step": 10376
+    },
+    {
+      "epoch": 0.4556900885868909,
+      "grad_norm": 1.8671875,
+      "learning_rate": 2.8597086065707655e-05,
+      "loss": 0.3693,
+      "step": 10378
+    },
+    {
+      "epoch": 0.4557779070660944,
+      "grad_norm": 2.03125,
+      "learning_rate": 2.8590239631527314e-05,
+      "loss": 0.34,
+      "step": 10380
+    },
+    {
+      "epoch": 0.45586572554529786,
+      "grad_norm": 1.7578125,
+      "learning_rate": 2.858339292240556e-05,
+      "loss": 0.3826,
+      "step": 10382
+    },
+    {
+      "epoch": 0.45595354402450133,
+      "grad_norm": 1.875,
+      "learning_rate": 2.857654593886671e-05,
+      "loss": 0.3651,
+      "step": 10384
+    },
+    {
+      "epoch": 0.45604136250370486,
+      "grad_norm": 1.9765625,
+      "learning_rate": 2.856969868143512e-05,
+      "loss": 0.3667,
+      "step": 10386
+    },
+    {
+      "epoch": 0.45612918098290833,
+      "grad_norm": 2.03125,
+      "learning_rate": 2.856285115063514e-05,
+      "loss": 0.3768,
+      "step": 10388
+    },
+    {
+      "epoch": 0.4562169994621118,
+      "grad_norm": 1.9765625,
+      "learning_rate": 2.8556003346991174e-05,
+      "loss": 0.3444,
+      "step": 10390
+    },
+    {
+      "epoch": 0.4563048179413153,
+      "grad_norm": 1.84375,
+      "learning_rate": 2.8549155271027617e-05,
+      "loss": 0.3787,
+      "step": 10392
+    },
+    {
+      "epoch": 0.4563926364205188,
+      "grad_norm": 2.0,
+      "learning_rate": 2.8542306923268897e-05,
+      "loss": 0.379,
+      "step": 10394
+    },
+    {
+      "epoch": 0.4564804548997223,
+      "grad_norm": 1.90625,
+      "learning_rate": 2.853545830423947e-05,
+      "loss": 0.373,
+      "step": 10396
+    },
+    {
+      "epoch": 0.45656827337892575,
+      "grad_norm": 1.9375,
+      "learning_rate": 2.8528609414463793e-05,
+      "loss": 0.3874,
+      "step": 10398
+    },
+    {
+      "epoch": 0.4566560918581292,
+      "grad_norm": 1.765625,
+      "learning_rate": 2.8521760254466355e-05,
+      "loss": 0.3398,
+      "step": 10400
+    },
+    {
+      "epoch": 0.45674391033733275,
+      "grad_norm": 1.8359375,
+      "learning_rate": 2.851491082477168e-05,
+      "loss": 0.3801,
+      "step": 10402
+    },
+    {
+      "epoch": 0.4568317288165362,
+      "grad_norm": 1.8515625,
+      "learning_rate": 2.8508061125904284e-05,
+      "loss": 0.3614,
+      "step": 10404
+    },
+    {
+      "epoch": 0.4569195472957397,
+      "grad_norm": 1.9765625,
+      "learning_rate": 2.850121115838874e-05,
+      "loss": 0.3851,
+      "step": 10406
+    },
+    {
+      "epoch": 0.4570073657749432,
+      "grad_norm": 2.046875,
+      "learning_rate": 2.8494360922749595e-05,
+      "loss": 0.3495,
+      "step": 10408
+    },
+    {
+      "epoch": 0.4570951842541467,
+      "grad_norm": 1.7734375,
+      "learning_rate": 2.848751041951146e-05,
+      "loss": 0.3773,
+      "step": 10410
+    },
+    {
+      "epoch": 0.45718300273335016,
+      "grad_norm": 1.8828125,
+      "learning_rate": 2.8480659649198937e-05,
+      "loss": 0.3699,
+      "step": 10412
+    },
+    {
+      "epoch": 0.45727082121255364,
+      "grad_norm": 2.0,
+      "learning_rate": 2.8473808612336662e-05,
+      "loss": 0.3567,
+      "step": 10414
+    },
+    {
+      "epoch": 0.45735863969175716,
+      "grad_norm": 1.84375,
+      "learning_rate": 2.8466957309449287e-05,
+      "loss": 0.3647,
+      "step": 10416
+    },
+    {
+      "epoch": 0.45744645817096063,
+      "grad_norm": 1.9453125,
+      "learning_rate": 2.8460105741061505e-05,
+      "loss": 0.3707,
+      "step": 10418
+    },
+    {
+      "epoch": 0.4575342766501641,
+      "grad_norm": 2.25,
+      "learning_rate": 2.8453253907697985e-05,
+      "loss": 0.3728,
+      "step": 10420
+    },
+    {
+      "epoch": 0.4576220951293676,
+      "grad_norm": 2.0,
+      "learning_rate": 2.844640180988345e-05,
+      "loss": 0.3649,
+      "step": 10422
+    },
+    {
+      "epoch": 0.4577099136085711,
+      "grad_norm": 1.9296875,
+      "learning_rate": 2.8439549448142644e-05,
+      "loss": 0.3704,
+      "step": 10424
+    },
+    {
+      "epoch": 0.4577977320877746,
+      "grad_norm": 2.09375,
+      "learning_rate": 2.8432696823000314e-05,
+      "loss": 0.3709,
+      "step": 10426
+    },
+    {
+      "epoch": 0.45788555056697805,
+      "grad_norm": 2.0625,
+      "learning_rate": 2.8425843934981245e-05,
+      "loss": 0.3849,
+      "step": 10428
+    },
+    {
+      "epoch": 0.4579733690461815,
+      "grad_norm": 2.046875,
+      "learning_rate": 2.8418990784610223e-05,
+      "loss": 0.388,
+      "step": 10430
+    },
+    {
+      "epoch": 0.45806118752538505,
+      "grad_norm": 2.125,
+      "learning_rate": 2.8412137372412062e-05,
+      "loss": 0.3925,
+      "step": 10432
+    },
+    {
+      "epoch": 0.4581490060045885,
+      "grad_norm": 1.890625,
+      "learning_rate": 2.8405283698911605e-05,
+      "loss": 0.3642,
+      "step": 10434
+    },
+    {
+      "epoch": 0.458236824483792,
+      "grad_norm": 1.921875,
+      "learning_rate": 2.8398429764633706e-05,
+      "loss": 0.3798,
+      "step": 10436
+    },
+    {
+      "epoch": 0.45832464296299547,
+      "grad_norm": 1.8046875,
+      "learning_rate": 2.839157557010324e-05,
+      "loss": 0.3691,
+      "step": 10438
+    },
+    {
+      "epoch": 0.458412461442199,
+      "grad_norm": 1.71875,
+      "learning_rate": 2.838472111584511e-05,
+      "loss": 0.3631,
+      "step": 10440
+    },
+    {
+      "epoch": 0.45850027992140246,
+      "grad_norm": 2.0625,
+      "learning_rate": 2.8377866402384223e-05,
+      "loss": 0.4049,
+      "step": 10442
+    },
+    {
+      "epoch": 0.45858809840060594,
+      "grad_norm": 1.8984375,
+      "learning_rate": 2.837101143024552e-05,
+      "loss": 0.374,
+      "step": 10444
+    },
+    {
+      "epoch": 0.4586759168798094,
+      "grad_norm": 1.8828125,
+      "learning_rate": 2.836415619995395e-05,
+      "loss": 0.3938,
+      "step": 10446
+    },
+    {
+      "epoch": 0.45876373535901294,
+      "grad_norm": 1.7421875,
+      "learning_rate": 2.835730071203449e-05,
+      "loss": 0.3711,
+      "step": 10448
+    },
+    {
+      "epoch": 0.4588515538382164,
+      "grad_norm": 2.1875,
+      "learning_rate": 2.8350444967012134e-05,
+      "loss": 0.3739,
+      "step": 10450
+    },
+    {
+      "epoch": 0.4589393723174199,
+      "grad_norm": 1.859375,
+      "learning_rate": 2.8343588965411905e-05,
+      "loss": 0.3799,
+      "step": 10452
+    },
+    {
+      "epoch": 0.45902719079662335,
+      "grad_norm": 2.015625,
+      "learning_rate": 2.833673270775883e-05,
+      "loss": 0.3414,
+      "step": 10454
+    },
+    {
+      "epoch": 0.4591150092758269,
+      "grad_norm": 1.7578125,
+      "learning_rate": 2.832987619457797e-05,
+      "loss": 0.3798,
+      "step": 10456
+    },
+    {
+      "epoch": 0.45920282775503035,
+      "grad_norm": 2.390625,
+      "learning_rate": 2.832301942639439e-05,
+      "loss": 0.3828,
+      "step": 10458
+    },
+    {
+      "epoch": 0.4592906462342338,
+      "grad_norm": 1.984375,
+      "learning_rate": 2.8316162403733177e-05,
+      "loss": 0.3562,
+      "step": 10460
+    },
+    {
+      "epoch": 0.45937846471343735,
+      "grad_norm": 2.140625,
+      "learning_rate": 2.8309305127119456e-05,
+      "loss": 0.3902,
+      "step": 10462
+    },
+    {
+      "epoch": 0.4594662831926408,
+      "grad_norm": 2.203125,
+      "learning_rate": 2.8302447597078353e-05,
+      "loss": 0.3925,
+      "step": 10464
+    },
+    {
+      "epoch": 0.4595541016718443,
+      "grad_norm": 2.09375,
+      "learning_rate": 2.8295589814135032e-05,
+      "loss": 0.3432,
+      "step": 10466
+    },
+    {
+      "epoch": 0.45964192015104777,
+      "grad_norm": 2.078125,
+      "learning_rate": 2.8288731778814642e-05,
+      "loss": 0.3987,
+      "step": 10468
+    },
+    {
+      "epoch": 0.4597297386302513,
+      "grad_norm": 1.9140625,
+      "learning_rate": 2.8281873491642392e-05,
+      "loss": 0.3686,
+      "step": 10470
+    },
+    {
+      "epoch": 0.45981755710945477,
+      "grad_norm": 1.9765625,
+      "learning_rate": 2.827501495314348e-05,
+      "loss": 0.3607,
+      "step": 10472
+    },
+    {
+      "epoch": 0.45990537558865824,
+      "grad_norm": 2.109375,
+      "learning_rate": 2.8268156163843136e-05,
+      "loss": 0.3818,
+      "step": 10474
+    },
+    {
+      "epoch": 0.4599931940678617,
+      "grad_norm": 2.1875,
+      "learning_rate": 2.8261297124266613e-05,
+      "loss": 0.3941,
+      "step": 10476
+    },
+    {
+      "epoch": 0.46008101254706524,
+      "grad_norm": 1.84375,
+      "learning_rate": 2.825443783493917e-05,
+      "loss": 0.3418,
+      "step": 10478
+    },
+    {
+      "epoch": 0.4601688310262687,
+      "grad_norm": 1.859375,
+      "learning_rate": 2.8247578296386102e-05,
+      "loss": 0.3479,
+      "step": 10480
+    },
+    {
+      "epoch": 0.4602566495054722,
+      "grad_norm": 1.7265625,
+      "learning_rate": 2.824071850913271e-05,
+      "loss": 0.3828,
+      "step": 10482
+    },
+    {
+      "epoch": 0.46034446798467565,
+      "grad_norm": 1.78125,
+      "learning_rate": 2.823385847370431e-05,
+      "loss": 0.3475,
+      "step": 10484
+    },
+    {
+      "epoch": 0.4604322864638792,
+      "grad_norm": 2.015625,
+      "learning_rate": 2.822699819062626e-05,
+      "loss": 0.3772,
+      "step": 10486
+    },
+    {
+      "epoch": 0.46052010494308265,
+      "grad_norm": 1.7578125,
+      "learning_rate": 2.822013766042391e-05,
+      "loss": 0.3601,
+      "step": 10488
+    },
+    {
+      "epoch": 0.4606079234222861,
+      "grad_norm": 1.8828125,
+      "learning_rate": 2.8213276883622654e-05,
+      "loss": 0.383,
+      "step": 10490
+    },
+    {
+      "epoch": 0.4606957419014896,
+      "grad_norm": 1.8984375,
+      "learning_rate": 2.820641586074788e-05,
+      "loss": 0.3687,
+      "step": 10492
+    },
+    {
+      "epoch": 0.4607835603806931,
+      "grad_norm": 1.8125,
+      "learning_rate": 2.8199554592325005e-05,
+      "loss": 0.3863,
+      "step": 10494
+    },
+    {
+      "epoch": 0.4608713788598966,
+      "grad_norm": 2.015625,
+      "learning_rate": 2.819269307887948e-05,
+      "loss": 0.3222,
+      "step": 10496
+    },
+    {
+      "epoch": 0.46095919733910007,
+      "grad_norm": 1.7421875,
+      "learning_rate": 2.818583132093675e-05,
+      "loss": 0.3738,
+      "step": 10498
+    },
+    {
+      "epoch": 0.46104701581830354,
+      "grad_norm": 2.0625,
+      "learning_rate": 2.8178969319022292e-05,
+      "loss": 0.3666,
+      "step": 10500
+    },
+    {
+      "epoch": 0.46113483429750707,
+      "grad_norm": 1.859375,
+      "learning_rate": 2.8172107073661607e-05,
+      "loss": 0.3504,
+      "step": 10502
+    },
+    {
+      "epoch": 0.46122265277671054,
+      "grad_norm": 1.7578125,
+      "learning_rate": 2.8165244585380197e-05,
+      "loss": 0.3886,
+      "step": 10504
+    },
+    {
+      "epoch": 0.461310471255914,
+      "grad_norm": 1.8515625,
+      "learning_rate": 2.815838185470361e-05,
+      "loss": 0.3935,
+      "step": 10506
+    },
+    {
+      "epoch": 0.46139828973511754,
+      "grad_norm": 1.8125,
+      "learning_rate": 2.815151888215737e-05,
+      "loss": 0.3804,
+      "step": 10508
+    },
+    {
+      "epoch": 0.461486108214321,
+      "grad_norm": 1.6640625,
+      "learning_rate": 2.8144655668267056e-05,
+      "loss": 0.3321,
+      "step": 10510
+    },
+    {
+      "epoch": 0.4615739266935245,
+      "grad_norm": 2.015625,
+      "learning_rate": 2.813779221355826e-05,
+      "loss": 0.3835,
+      "step": 10512
+    },
+    {
+      "epoch": 0.46166174517272796,
+      "grad_norm": 1.828125,
+      "learning_rate": 2.8130928518556588e-05,
+      "loss": 0.3922,
+      "step": 10514
+    },
+    {
+      "epoch": 0.4617495636519315,
+      "grad_norm": 1.734375,
+      "learning_rate": 2.8124064583787662e-05,
+      "loss": 0.3696,
+      "step": 10516
+    },
+    {
+      "epoch": 0.46183738213113495,
+      "grad_norm": 2.15625,
+      "learning_rate": 2.8117200409777124e-05,
+      "loss": 0.3481,
+      "step": 10518
+    },
+    {
+      "epoch": 0.4619252006103384,
+      "grad_norm": 1.7890625,
+      "learning_rate": 2.8110335997050624e-05,
+      "loss": 0.3764,
+      "step": 10520
+    },
+    {
+      "epoch": 0.4620130190895419,
+      "grad_norm": 1.859375,
+      "learning_rate": 2.8103471346133848e-05,
+      "loss": 0.3645,
+      "step": 10522
+    },
+    {
+      "epoch": 0.4621008375687454,
+      "grad_norm": 1.9375,
+      "learning_rate": 2.8096606457552488e-05,
+      "loss": 0.3537,
+      "step": 10524
+    },
+    {
+      "epoch": 0.4621886560479489,
+      "grad_norm": 1.703125,
+      "learning_rate": 2.808974133183227e-05,
+      "loss": 0.3692,
+      "step": 10526
+    },
+    {
+      "epoch": 0.46227647452715237,
+      "grad_norm": 1.90625,
+      "learning_rate": 2.8082875969498922e-05,
+      "loss": 0.3552,
+      "step": 10528
+    },
+    {
+      "epoch": 0.46236429300635584,
+      "grad_norm": 1.8671875,
+      "learning_rate": 2.8076010371078186e-05,
+      "loss": 0.3821,
+      "step": 10530
+    },
+    {
+      "epoch": 0.46245211148555937,
+      "grad_norm": 2.109375,
+      "learning_rate": 2.8069144537095842e-05,
+      "loss": 0.3728,
+      "step": 10532
+    },
+    {
+      "epoch": 0.46253992996476284,
+      "grad_norm": 1.859375,
+      "learning_rate": 2.8062278468077678e-05,
+      "loss": 0.3553,
+      "step": 10534
+    },
+    {
+      "epoch": 0.4626277484439663,
+      "grad_norm": 1.875,
+      "learning_rate": 2.8055412164549488e-05,
+      "loss": 0.36,
+      "step": 10536
+    },
+    {
+      "epoch": 0.4627155669231698,
+      "grad_norm": 1.8984375,
+      "learning_rate": 2.8048545627037102e-05,
+      "loss": 0.3602,
+      "step": 10538
+    },
+    {
+      "epoch": 0.4628033854023733,
+      "grad_norm": 2.140625,
+      "learning_rate": 2.8041678856066367e-05,
+      "loss": 0.3827,
+      "step": 10540
+    },
+    {
+      "epoch": 0.4628912038815768,
+      "grad_norm": 2.015625,
+      "learning_rate": 2.8034811852163136e-05,
+      "loss": 0.3562,
+      "step": 10542
+    },
+    {
+      "epoch": 0.46297902236078026,
+      "grad_norm": 1.828125,
+      "learning_rate": 2.802794461585328e-05,
+      "loss": 0.3726,
+      "step": 10544
+    },
+    {
+      "epoch": 0.46306684083998373,
+      "grad_norm": 2.140625,
+      "learning_rate": 2.80210771476627e-05,
+      "loss": 0.3806,
+      "step": 10546
+    },
+    {
+      "epoch": 0.46315465931918726,
+      "grad_norm": 1.9140625,
+      "learning_rate": 2.8014209448117317e-05,
+      "loss": 0.3816,
+      "step": 10548
+    },
+    {
+      "epoch": 0.46324247779839073,
+      "grad_norm": 1.921875,
+      "learning_rate": 2.8007341517743042e-05,
+      "loss": 0.3968,
+      "step": 10550
+    },
+    {
+      "epoch": 0.4633302962775942,
+      "grad_norm": 1.8515625,
+      "learning_rate": 2.8000473357065838e-05,
+      "loss": 0.3709,
+      "step": 10552
+    },
+    {
+      "epoch": 0.46341811475679767,
+      "grad_norm": 2.203125,
+      "learning_rate": 2.7993604966611665e-05,
+      "loss": 0.3425,
+      "step": 10554
+    },
+    {
+      "epoch": 0.4635059332360012,
+      "grad_norm": 1.75,
+      "learning_rate": 2.79867363469065e-05,
+      "loss": 0.3793,
+      "step": 10556
+    },
+    {
+      "epoch": 0.46359375171520467,
+      "grad_norm": 2.3125,
+      "learning_rate": 2.7979867498476354e-05,
+      "loss": 0.3728,
+      "step": 10558
+    },
+    {
+      "epoch": 0.46368157019440814,
+      "grad_norm": 1.828125,
+      "learning_rate": 2.7972998421847235e-05,
+      "loss": 0.382,
+      "step": 10560
+    },
+    {
+      "epoch": 0.46376938867361167,
+      "grad_norm": 2.484375,
+      "learning_rate": 2.7966129117545194e-05,
+      "loss": 0.3628,
+      "step": 10562
+    },
+    {
+      "epoch": 0.46385720715281514,
+      "grad_norm": 1.828125,
+      "learning_rate": 2.7959259586096272e-05,
+      "loss": 0.372,
+      "step": 10564
+    },
+    {
+      "epoch": 0.4639450256320186,
+      "grad_norm": 2.28125,
+      "learning_rate": 2.7952389828026538e-05,
+      "loss": 0.3911,
+      "step": 10566
+    },
+    {
+      "epoch": 0.4640328441112221,
+      "grad_norm": 1.734375,
+      "learning_rate": 2.7945519843862083e-05,
+      "loss": 0.3848,
+      "step": 10568
+    },
+    {
+      "epoch": 0.4641206625904256,
+      "grad_norm": 1.890625,
+      "learning_rate": 2.7938649634129015e-05,
+      "loss": 0.3828,
+      "step": 10570
+    },
+    {
+      "epoch": 0.4642084810696291,
+      "grad_norm": 1.8359375,
+      "learning_rate": 2.7931779199353448e-05,
+      "loss": 0.3414,
+      "step": 10572
+    },
+    {
+      "epoch": 0.46429629954883256,
+      "grad_norm": 1.953125,
+      "learning_rate": 2.792490854006153e-05,
+      "loss": 0.3428,
+      "step": 10574
+    },
+    {
+      "epoch": 0.46438411802803603,
+      "grad_norm": 1.84375,
+      "learning_rate": 2.7918037656779417e-05,
+      "loss": 0.3689,
+      "step": 10576
+    },
+    {
+      "epoch": 0.46447193650723956,
+      "grad_norm": 2.09375,
+      "learning_rate": 2.791116655003328e-05,
+      "loss": 0.3652,
+      "step": 10578
+    },
+    {
+      "epoch": 0.46455975498644303,
+      "grad_norm": 1.984375,
+      "learning_rate": 2.790429522034932e-05,
+      "loss": 0.3702,
+      "step": 10580
+    },
+    {
+      "epoch": 0.4646475734656465,
+      "grad_norm": 2.40625,
+      "learning_rate": 2.789742366825372e-05,
+      "loss": 0.3559,
+      "step": 10582
+    },
+    {
+      "epoch": 0.46473539194485,
+      "grad_norm": 2.25,
+      "learning_rate": 2.7890551894272726e-05,
+      "loss": 0.3607,
+      "step": 10584
+    },
+    {
+      "epoch": 0.4648232104240535,
+      "grad_norm": 1.9453125,
+      "learning_rate": 2.7883679898932575e-05,
+      "loss": 0.3528,
+      "step": 10586
+    },
+    {
+      "epoch": 0.464911028903257,
+      "grad_norm": 2.1875,
+      "learning_rate": 2.7876807682759526e-05,
+      "loss": 0.3599,
+      "step": 10588
+    },
+    {
+      "epoch": 0.46499884738246045,
+      "grad_norm": 2.203125,
+      "learning_rate": 2.786993524627986e-05,
+      "loss": 0.3848,
+      "step": 10590
+    },
+    {
+      "epoch": 0.4650866658616639,
+      "grad_norm": 2.046875,
+      "learning_rate": 2.786306259001986e-05,
+      "loss": 0.3684,
+      "step": 10592
+    },
+    {
+      "epoch": 0.46517448434086744,
+      "grad_norm": 2.09375,
+      "learning_rate": 2.7856189714505843e-05,
+      "loss": 0.38,
+      "step": 10594
+    },
+    {
+      "epoch": 0.4652623028200709,
+      "grad_norm": 1.9140625,
+      "learning_rate": 2.7849316620264133e-05,
+      "loss": 0.3607,
+      "step": 10596
+    },
+    {
+      "epoch": 0.4653501212992744,
+      "grad_norm": 1.8359375,
+      "learning_rate": 2.784244330782107e-05,
+      "loss": 0.3588,
+      "step": 10598
+    },
+    {
+      "epoch": 0.46543793977847786,
+      "grad_norm": 2.015625,
+      "learning_rate": 2.783556977770302e-05,
+      "loss": 0.3734,
+      "step": 10600
+    },
+    {
+      "epoch": 0.4655257582576814,
+      "grad_norm": 1.984375,
+      "learning_rate": 2.7828696030436353e-05,
+      "loss": 0.3406,
+      "step": 10602
+    },
+    {
+      "epoch": 0.46561357673688486,
+      "grad_norm": 2.171875,
+      "learning_rate": 2.782182206654747e-05,
+      "loss": 0.3711,
+      "step": 10604
+    },
+    {
+      "epoch": 0.46570139521608833,
+      "grad_norm": 1.8515625,
+      "learning_rate": 2.7814947886562774e-05,
+      "loss": 0.3544,
+      "step": 10606
+    },
+    {
+      "epoch": 0.46578921369529186,
+      "grad_norm": 1.6484375,
+      "learning_rate": 2.7808073491008698e-05,
+      "loss": 0.3632,
+      "step": 10608
+    },
+    {
+      "epoch": 0.46587703217449533,
+      "grad_norm": 1.8828125,
+      "learning_rate": 2.780119888041168e-05,
+      "loss": 0.3707,
+      "step": 10610
+    },
+    {
+      "epoch": 0.4659648506536988,
+      "grad_norm": 2.296875,
+      "learning_rate": 2.7794324055298178e-05,
+      "loss": 0.3833,
+      "step": 10612
+    },
+    {
+      "epoch": 0.4660526691329023,
+      "grad_norm": 2.203125,
+      "learning_rate": 2.7787449016194665e-05,
+      "loss": 0.3892,
+      "step": 10614
+    },
+    {
+      "epoch": 0.4661404876121058,
+      "grad_norm": 1.78125,
+      "learning_rate": 2.7780573763627648e-05,
+      "loss": 0.3694,
+      "step": 10616
+    },
+    {
+      "epoch": 0.4662283060913093,
+      "grad_norm": 1.8671875,
+      "learning_rate": 2.777369829812362e-05,
+      "loss": 0.3798,
+      "step": 10618
+    },
+    {
+      "epoch": 0.46631612457051275,
+      "grad_norm": 1.9140625,
+      "learning_rate": 2.776682262020911e-05,
+      "loss": 0.388,
+      "step": 10620
+    },
+    {
+      "epoch": 0.4664039430497162,
+      "grad_norm": 2.015625,
+      "learning_rate": 2.7759946730410663e-05,
+      "loss": 0.3559,
+      "step": 10622
+    },
+    {
+      "epoch": 0.46649176152891975,
+      "grad_norm": 2.09375,
+      "learning_rate": 2.7753070629254835e-05,
+      "loss": 0.3877,
+      "step": 10624
+    },
+    {
+      "epoch": 0.4665795800081232,
+      "grad_norm": 1.7734375,
+      "learning_rate": 2.7746194317268197e-05,
+      "loss": 0.3885,
+      "step": 10626
+    },
+    {
+      "epoch": 0.4666673984873267,
+      "grad_norm": 1.7578125,
+      "learning_rate": 2.773931779497735e-05,
+      "loss": 0.3809,
+      "step": 10628
+    },
+    {
+      "epoch": 0.46675521696653016,
+      "grad_norm": 1.734375,
+      "learning_rate": 2.7732441062908877e-05,
+      "loss": 0.3759,
+      "step": 10630
+    },
+    {
+      "epoch": 0.4668430354457337,
+      "grad_norm": 2.3125,
+      "learning_rate": 2.772556412158941e-05,
+      "loss": 0.3576,
+      "step": 10632
+    },
+    {
+      "epoch": 0.46693085392493716,
+      "grad_norm": 1.6953125,
+      "learning_rate": 2.7718686971545592e-05,
+      "loss": 0.3965,
+      "step": 10634
+    },
+    {
+      "epoch": 0.46701867240414063,
+      "grad_norm": 1.78125,
+      "learning_rate": 2.771180961330408e-05,
+      "loss": 0.3518,
+      "step": 10636
+    },
+    {
+      "epoch": 0.4671064908833441,
+      "grad_norm": 1.7421875,
+      "learning_rate": 2.770493204739153e-05,
+      "loss": 0.3663,
+      "step": 10638
+    },
+    {
+      "epoch": 0.46719430936254763,
+      "grad_norm": 1.8203125,
+      "learning_rate": 2.7698054274334645e-05,
+      "loss": 0.3468,
+      "step": 10640
+    },
+    {
+      "epoch": 0.4672821278417511,
+      "grad_norm": 1.859375,
+      "learning_rate": 2.7691176294660114e-05,
+      "loss": 0.3605,
+      "step": 10642
+    },
+    {
+      "epoch": 0.4673699463209546,
+      "grad_norm": 1.7734375,
+      "learning_rate": 2.768429810889465e-05,
+      "loss": 0.3526,
+      "step": 10644
+    },
+    {
+      "epoch": 0.46745776480015805,
+      "grad_norm": 2.078125,
+      "learning_rate": 2.7677419717565e-05,
+      "loss": 0.3834,
+      "step": 10646
+    },
+    {
+      "epoch": 0.4675455832793616,
+      "grad_norm": 2.140625,
+      "learning_rate": 2.76705411211979e-05,
+      "loss": 0.3985,
+      "step": 10648
+    },
+    {
+      "epoch": 0.46763340175856505,
+      "grad_norm": 1.78125,
+      "learning_rate": 2.766366232032013e-05,
+      "loss": 0.3632,
+      "step": 10650
+    },
+    {
+      "epoch": 0.4677212202377685,
+      "grad_norm": 1.7109375,
+      "learning_rate": 2.7656783315458457e-05,
+      "loss": 0.341,
+      "step": 10652
+    },
+    {
+      "epoch": 0.467809038716972,
+      "grad_norm": 2.546875,
+      "learning_rate": 2.7649904107139675e-05,
+      "loss": 0.3662,
+      "step": 10654
+    },
+    {
+      "epoch": 0.4678968571961755,
+      "grad_norm": 2.3125,
+      "learning_rate": 2.7643024695890613e-05,
+      "loss": 0.3479,
+      "step": 10656
+    },
+    {
+      "epoch": 0.467984675675379,
+      "grad_norm": 1.7265625,
+      "learning_rate": 2.7636145082238074e-05,
+      "loss": 0.3481,
+      "step": 10658
+    },
+    {
+      "epoch": 0.46807249415458246,
+      "grad_norm": 1.9140625,
+      "learning_rate": 2.762926526670892e-05,
+      "loss": 0.3795,
+      "step": 10660
+    },
+    {
+      "epoch": 0.468160312633786,
+      "grad_norm": 1.9296875,
+      "learning_rate": 2.7622385249829997e-05,
+      "loss": 0.3054,
+      "step": 10662
+    },
+    {
+      "epoch": 0.46824813111298946,
+      "grad_norm": 2.453125,
+      "learning_rate": 2.761550503212818e-05,
+      "loss": 0.3707,
+      "step": 10664
+    },
+    {
+      "epoch": 0.46833594959219293,
+      "grad_norm": 1.8984375,
+      "learning_rate": 2.7608624614130374e-05,
+      "loss": 0.3639,
+      "step": 10666
+    },
+    {
+      "epoch": 0.4684237680713964,
+      "grad_norm": 2.265625,
+      "learning_rate": 2.760174399636346e-05,
+      "loss": 0.3831,
+      "step": 10668
+    },
+    {
+      "epoch": 0.46851158655059993,
+      "grad_norm": 2.03125,
+      "learning_rate": 2.759486317935437e-05,
+      "loss": 0.3721,
+      "step": 10670
+    },
+    {
+      "epoch": 0.4685994050298034,
+      "grad_norm": 1.8984375,
+      "learning_rate": 2.7587982163630033e-05,
+      "loss": 0.365,
+      "step": 10672
+    },
+    {
+      "epoch": 0.4686872235090069,
+      "grad_norm": 1.9765625,
+      "learning_rate": 2.7581100949717402e-05,
+      "loss": 0.3823,
+      "step": 10674
+    },
+    {
+      "epoch": 0.46877504198821035,
+      "grad_norm": 1.796875,
+      "learning_rate": 2.7574219538143443e-05,
+      "loss": 0.3744,
+      "step": 10676
+    },
+    {
+      "epoch": 0.4688628604674139,
+      "grad_norm": 1.8984375,
+      "learning_rate": 2.756733792943514e-05,
+      "loss": 0.3758,
+      "step": 10678
+    },
+    {
+      "epoch": 0.46895067894661735,
+      "grad_norm": 1.890625,
+      "learning_rate": 2.7560456124119473e-05,
+      "loss": 0.3831,
+      "step": 10680
+    },
+    {
+      "epoch": 0.4690384974258208,
+      "grad_norm": 1.7109375,
+      "learning_rate": 2.7553574122723462e-05,
+      "loss": 0.3569,
+      "step": 10682
+    },
+    {
+      "epoch": 0.4691263159050243,
+      "grad_norm": 1.9609375,
+      "learning_rate": 2.7546691925774137e-05,
+      "loss": 0.3486,
+      "step": 10684
+    },
+    {
+      "epoch": 0.4692141343842278,
+      "grad_norm": 1.78125,
+      "learning_rate": 2.753980953379854e-05,
+      "loss": 0.3897,
+      "step": 10686
+    },
+    {
+      "epoch": 0.4693019528634313,
+      "grad_norm": 1.875,
+      "learning_rate": 2.753292694732371e-05,
+      "loss": 0.3905,
+      "step": 10688
+    },
+    {
+      "epoch": 0.46938977134263477,
+      "grad_norm": 1.84375,
+      "learning_rate": 2.7526044166876737e-05,
+      "loss": 0.3352,
+      "step": 10690
+    },
+    {
+      "epoch": 0.46947758982183824,
+      "grad_norm": 1.875,
+      "learning_rate": 2.751916119298469e-05,
+      "loss": 0.3845,
+      "step": 10692
+    },
+    {
+      "epoch": 0.46956540830104176,
+      "grad_norm": 1.6640625,
+      "learning_rate": 2.751227802617467e-05,
+      "loss": 0.3655,
+      "step": 10694
+    },
+    {
+      "epoch": 0.46965322678024524,
+      "grad_norm": 1.7265625,
+      "learning_rate": 2.7505394666973803e-05,
+      "loss": 0.3813,
+      "step": 10696
+    },
+    {
+      "epoch": 0.4697410452594487,
+      "grad_norm": 1.9609375,
+      "learning_rate": 2.749851111590921e-05,
+      "loss": 0.4094,
+      "step": 10698
+    },
+    {
+      "epoch": 0.4698288637386522,
+      "grad_norm": 1.7890625,
+      "learning_rate": 2.7491627373508035e-05,
+      "loss": 0.3623,
+      "step": 10700
+    },
+    {
+      "epoch": 0.4699166822178557,
+      "grad_norm": 1.921875,
+      "learning_rate": 2.7484743440297445e-05,
+      "loss": 0.3805,
+      "step": 10702
+    },
+    {
+      "epoch": 0.4700045006970592,
+      "grad_norm": 1.875,
+      "learning_rate": 2.7477859316804604e-05,
+      "loss": 0.3795,
+      "step": 10704
+    },
+    {
+      "epoch": 0.47009231917626265,
+      "grad_norm": 1.953125,
+      "learning_rate": 2.7470975003556705e-05,
+      "loss": 0.3744,
+      "step": 10706
+    },
+    {
+      "epoch": 0.4701801376554661,
+      "grad_norm": 1.671875,
+      "learning_rate": 2.746409050108094e-05,
+      "loss": 0.3915,
+      "step": 10708
+    },
+    {
+      "epoch": 0.47026795613466965,
+      "grad_norm": 1.8828125,
+      "learning_rate": 2.7457205809904534e-05,
+      "loss": 0.3823,
+      "step": 10710
+    },
+    {
+      "epoch": 0.4703557746138731,
+      "grad_norm": 1.890625,
+      "learning_rate": 2.7450320930554724e-05,
+      "loss": 0.3686,
+      "step": 10712
+    },
+    {
+      "epoch": 0.4704435930930766,
+      "grad_norm": 1.9921875,
+      "learning_rate": 2.7443435863558748e-05,
+      "loss": 0.3657,
+      "step": 10714
+    },
+    {
+      "epoch": 0.4705314115722801,
+      "grad_norm": 1.8125,
+      "learning_rate": 2.743655060944387e-05,
+      "loss": 0.3807,
+      "step": 10716
+    },
+    {
+      "epoch": 0.4706192300514836,
+      "grad_norm": 2.015625,
+      "learning_rate": 2.7429665168737363e-05,
+      "loss": 0.3675,
+      "step": 10718
+    },
+    {
+      "epoch": 0.47070704853068707,
+      "grad_norm": 1.6875,
+      "learning_rate": 2.742277954196651e-05,
+      "loss": 0.3704,
+      "step": 10720
+    },
+    {
+      "epoch": 0.47079486700989054,
+      "grad_norm": 2.0,
+      "learning_rate": 2.7415893729658616e-05,
+      "loss": 0.3794,
+      "step": 10722
+    },
+    {
+      "epoch": 0.47088268548909407,
+      "grad_norm": 1.90625,
+      "learning_rate": 2.7409007732341004e-05,
+      "loss": 0.3534,
+      "step": 10724
+    },
+    {
+      "epoch": 0.47097050396829754,
+      "grad_norm": 2.046875,
+      "learning_rate": 2.740212155054101e-05,
+      "loss": 0.3631,
+      "step": 10726
+    },
+    {
+      "epoch": 0.471058322447501,
+      "grad_norm": 1.8515625,
+      "learning_rate": 2.739523518478596e-05,
+      "loss": 0.3783,
+      "step": 10728
+    },
+    {
+      "epoch": 0.4711461409267045,
+      "grad_norm": 2.03125,
+      "learning_rate": 2.7388348635603235e-05,
+      "loss": 0.3625,
+      "step": 10730
+    },
+    {
+      "epoch": 0.471233959405908,
+      "grad_norm": 1.8671875,
+      "learning_rate": 2.7381461903520194e-05,
+      "loss": 0.3539,
+      "step": 10732
+    },
+    {
+      "epoch": 0.4713217778851115,
+      "grad_norm": 1.9296875,
+      "learning_rate": 2.7374574989064234e-05,
+      "loss": 0.3576,
+      "step": 10734
+    },
+    {
+      "epoch": 0.47140959636431495,
+      "grad_norm": 1.8125,
+      "learning_rate": 2.7367687892762745e-05,
+      "loss": 0.3955,
+      "step": 10736
+    },
+    {
+      "epoch": 0.4714974148435184,
+      "grad_norm": 1.8125,
+      "learning_rate": 2.736080061514315e-05,
+      "loss": 0.3655,
+      "step": 10738
+    },
+    {
+      "epoch": 0.47158523332272195,
+      "grad_norm": 1.859375,
+      "learning_rate": 2.7353913156732884e-05,
+      "loss": 0.3624,
+      "step": 10740
+    },
+    {
+      "epoch": 0.4716730518019254,
+      "grad_norm": 1.984375,
+      "learning_rate": 2.7347025518059383e-05,
+      "loss": 0.3643,
+      "step": 10742
+    },
+    {
+      "epoch": 0.4717608702811289,
+      "grad_norm": 1.7890625,
+      "learning_rate": 2.7340137699650103e-05,
+      "loss": 0.353,
+      "step": 10744
+    },
+    {
+      "epoch": 0.47184868876033237,
+      "grad_norm": 2.03125,
+      "learning_rate": 2.7333249702032516e-05,
+      "loss": 0.3639,
+      "step": 10746
+    },
+    {
+      "epoch": 0.4719365072395359,
+      "grad_norm": 1.875,
+      "learning_rate": 2.7326361525734112e-05,
+      "loss": 0.3855,
+      "step": 10748
+    },
+    {
+      "epoch": 0.47202432571873937,
+      "grad_norm": 1.8125,
+      "learning_rate": 2.731947317128238e-05,
+      "loss": 0.3408,
+      "step": 10750
+    },
+    {
+      "epoch": 0.47211214419794284,
+      "grad_norm": 1.7578125,
+      "learning_rate": 2.731258463920484e-05,
+      "loss": 0.3931,
+      "step": 10752
+    },
+    {
+      "epoch": 0.4721999626771463,
+      "grad_norm": 1.890625,
+      "learning_rate": 2.730569593002901e-05,
+      "loss": 0.3563,
+      "step": 10754
+    },
+    {
+      "epoch": 0.47228778115634984,
+      "grad_norm": 2.0625,
+      "learning_rate": 2.7298807044282438e-05,
+      "loss": 0.3742,
+      "step": 10756
+    },
+    {
+      "epoch": 0.4723755996355533,
+      "grad_norm": 2.140625,
+      "learning_rate": 2.7291917982492664e-05,
+      "loss": 0.4072,
+      "step": 10758
+    },
+    {
+      "epoch": 0.4724634181147568,
+      "grad_norm": 2.0625,
+      "learning_rate": 2.7285028745187268e-05,
+      "loss": 0.3598,
+      "step": 10760
+    },
+    {
+      "epoch": 0.4725512365939603,
+      "grad_norm": 1.8828125,
+      "learning_rate": 2.7278139332893827e-05,
+      "loss": 0.3398,
+      "step": 10762
+    },
+    {
+      "epoch": 0.4726390550731638,
+      "grad_norm": 2.0,
+      "learning_rate": 2.727124974613992e-05,
+      "loss": 0.3882,
+      "step": 10764
+    },
+    {
+      "epoch": 0.47272687355236725,
+      "grad_norm": 1.75,
+      "learning_rate": 2.7264359985453168e-05,
+      "loss": 0.391,
+      "step": 10766
+    },
+    {
+      "epoch": 0.4728146920315707,
+      "grad_norm": 1.7578125,
+      "learning_rate": 2.7257470051361184e-05,
+      "loss": 0.3629,
+      "step": 10768
+    },
+    {
+      "epoch": 0.47290251051077425,
+      "grad_norm": 2.0,
+      "learning_rate": 2.72505799443916e-05,
+      "loss": 0.3444,
+      "step": 10770
+    },
+    {
+      "epoch": 0.4729903289899777,
+      "grad_norm": 2.390625,
+      "learning_rate": 2.724368966507207e-05,
+      "loss": 0.3814,
+      "step": 10772
+    },
+    {
+      "epoch": 0.4730781474691812,
+      "grad_norm": 1.8984375,
+      "learning_rate": 2.7236799213930243e-05,
+      "loss": 0.3891,
+      "step": 10774
+    },
+    {
+      "epoch": 0.47316596594838467,
+      "grad_norm": 1.8984375,
+      "learning_rate": 2.7229908591493804e-05,
+      "loss": 0.3767,
+      "step": 10776
+    },
+    {
+      "epoch": 0.4732537844275882,
+      "grad_norm": 1.890625,
+      "learning_rate": 2.7223017798290423e-05,
+      "loss": 0.3763,
+      "step": 10778
+    },
+    {
+      "epoch": 0.47334160290679167,
+      "grad_norm": 1.8125,
+      "learning_rate": 2.7216126834847805e-05,
+      "loss": 0.3715,
+      "step": 10780
+    },
+    {
+      "epoch": 0.47342942138599514,
+      "grad_norm": 1.8203125,
+      "learning_rate": 2.7209235701693663e-05,
+      "loss": 0.3626,
+      "step": 10782
+    },
+    {
+      "epoch": 0.4735172398651986,
+      "grad_norm": 1.796875,
+      "learning_rate": 2.7202344399355722e-05,
+      "loss": 0.3342,
+      "step": 10784
+    },
+    {
+      "epoch": 0.47360505834440214,
+      "grad_norm": 1.84375,
+      "learning_rate": 2.719545292836172e-05,
+      "loss": 0.3692,
+      "step": 10786
+    },
+    {
+      "epoch": 0.4736928768236056,
+      "grad_norm": 1.7421875,
+      "learning_rate": 2.7188561289239413e-05,
+      "loss": 0.3625,
+      "step": 10788
+    },
+    {
+      "epoch": 0.4737806953028091,
+      "grad_norm": 1.859375,
+      "learning_rate": 2.7181669482516546e-05,
+      "loss": 0.3626,
+      "step": 10790
+    },
+    {
+      "epoch": 0.47386851378201256,
+      "grad_norm": 1.890625,
+      "learning_rate": 2.7174777508720917e-05,
+      "loss": 0.3878,
+      "step": 10792
+    },
+    {
+      "epoch": 0.4739563322612161,
+      "grad_norm": 1.8125,
+      "learning_rate": 2.7167885368380304e-05,
+      "loss": 0.347,
+      "step": 10794
+    },
+    {
+      "epoch": 0.47404415074041956,
+      "grad_norm": 1.921875,
+      "learning_rate": 2.7160993062022504e-05,
+      "loss": 0.3518,
+      "step": 10796
+    },
+    {
+      "epoch": 0.47413196921962303,
+      "grad_norm": 1.953125,
+      "learning_rate": 2.715410059017534e-05,
+      "loss": 0.3369,
+      "step": 10798
+    },
+    {
+      "epoch": 0.4742197876988265,
+      "grad_norm": 1.890625,
+      "learning_rate": 2.7147207953366633e-05,
+      "loss": 0.3438,
+      "step": 10800
+    },
+    {
+      "epoch": 0.47430760617803003,
+      "grad_norm": 1.859375,
+      "learning_rate": 2.7140315152124233e-05,
+      "loss": 0.3813,
+      "step": 10802
+    },
+    {
+      "epoch": 0.4743954246572335,
+      "grad_norm": 1.8828125,
+      "learning_rate": 2.713342218697598e-05,
+      "loss": 0.3613,
+      "step": 10804
+    },
+    {
+      "epoch": 0.47448324313643697,
+      "grad_norm": 1.625,
+      "learning_rate": 2.7126529058449747e-05,
+      "loss": 0.3814,
+      "step": 10806
+    },
+    {
+      "epoch": 0.47457106161564044,
+      "grad_norm": 2.078125,
+      "learning_rate": 2.711963576707341e-05,
+      "loss": 0.3403,
+      "step": 10808
+    },
+    {
+      "epoch": 0.47465888009484397,
+      "grad_norm": 2.0,
+      "learning_rate": 2.7112742313374855e-05,
+      "loss": 0.3498,
+      "step": 10810
+    },
+    {
+      "epoch": 0.47474669857404744,
+      "grad_norm": 1.84375,
+      "learning_rate": 2.710584869788199e-05,
+      "loss": 0.3991,
+      "step": 10812
+    },
+    {
+      "epoch": 0.4748345170532509,
+      "grad_norm": 1.734375,
+      "learning_rate": 2.7098954921122726e-05,
+      "loss": 0.3418,
+      "step": 10814
+    },
+    {
+      "epoch": 0.47492233553245444,
+      "grad_norm": 1.75,
+      "learning_rate": 2.709206098362499e-05,
+      "loss": 0.3639,
+      "step": 10816
+    },
+    {
+      "epoch": 0.4750101540116579,
+      "grad_norm": 1.9609375,
+      "learning_rate": 2.7085166885916723e-05,
+      "loss": 0.3521,
+      "step": 10818
+    },
+    {
+      "epoch": 0.4750979724908614,
+      "grad_norm": 1.90625,
+      "learning_rate": 2.7078272628525875e-05,
+      "loss": 0.3859,
+      "step": 10820
+    },
+    {
+      "epoch": 0.47518579097006486,
+      "grad_norm": 1.78125,
+      "learning_rate": 2.707137821198041e-05,
+      "loss": 0.3664,
+      "step": 10822
+    },
+    {
+      "epoch": 0.4752736094492684,
+      "grad_norm": 1.890625,
+      "learning_rate": 2.7064483636808313e-05,
+      "loss": 0.3766,
+      "step": 10824
+    },
+    {
+      "epoch": 0.47536142792847186,
+      "grad_norm": 1.7265625,
+      "learning_rate": 2.705758890353756e-05,
+      "loss": 0.3534,
+      "step": 10826
+    },
+    {
+      "epoch": 0.47544924640767533,
+      "grad_norm": 1.859375,
+      "learning_rate": 2.7050694012696155e-05,
+      "loss": 0.3678,
+      "step": 10828
+    },
+    {
+      "epoch": 0.4755370648868788,
+      "grad_norm": 1.8359375,
+      "learning_rate": 2.7043798964812107e-05,
+      "loss": 0.3803,
+      "step": 10830
+    },
+    {
+      "epoch": 0.47562488336608233,
+      "grad_norm": 1.890625,
+      "learning_rate": 2.7036903760413447e-05,
+      "loss": 0.3502,
+      "step": 10832
+    },
+    {
+      "epoch": 0.4757127018452858,
+      "grad_norm": 1.9453125,
+      "learning_rate": 2.7030008400028205e-05,
+      "loss": 0.381,
+      "step": 10834
+    },
+    {
+      "epoch": 0.4758005203244893,
+      "grad_norm": 1.859375,
+      "learning_rate": 2.702311288418443e-05,
+      "loss": 0.3547,
+      "step": 10836
+    },
+    {
+      "epoch": 0.47588833880369275,
+      "grad_norm": 1.734375,
+      "learning_rate": 2.70162172134102e-05,
+      "loss": 0.3324,
+      "step": 10838
+    },
+    {
+      "epoch": 0.4759761572828963,
+      "grad_norm": 2.125,
+      "learning_rate": 2.7009321388233567e-05,
+      "loss": 0.3896,
+      "step": 10840
+    },
+    {
+      "epoch": 0.47606397576209974,
+      "grad_norm": 2.265625,
+      "learning_rate": 2.700242540918262e-05,
+      "loss": 0.3554,
+      "step": 10842
+    },
+    {
+      "epoch": 0.4761517942413032,
+      "grad_norm": 1.9609375,
+      "learning_rate": 2.6995529276785446e-05,
+      "loss": 0.3728,
+      "step": 10844
+    },
+    {
+      "epoch": 0.4762396127205067,
+      "grad_norm": 1.7421875,
+      "learning_rate": 2.6988632991570174e-05,
+      "loss": 0.364,
+      "step": 10846
+    },
+    {
+      "epoch": 0.4763274311997102,
+      "grad_norm": 1.875,
+      "learning_rate": 2.6981736554064903e-05,
+      "loss": 0.4019,
+      "step": 10848
+    },
+    {
+      "epoch": 0.4764152496789137,
+      "grad_norm": 1.9921875,
+      "learning_rate": 2.6974839964797776e-05,
+      "loss": 0.3606,
+      "step": 10850
+    },
+    {
+      "epoch": 0.47650306815811716,
+      "grad_norm": 1.828125,
+      "learning_rate": 2.6967943224296927e-05,
+      "loss": 0.3407,
+      "step": 10852
+    },
+    {
+      "epoch": 0.47659088663732063,
+      "grad_norm": 1.828125,
+      "learning_rate": 2.6961046333090517e-05,
+      "loss": 0.3656,
+      "step": 10854
+    },
+    {
+      "epoch": 0.47667870511652416,
+      "grad_norm": 1.9375,
+      "learning_rate": 2.695414929170671e-05,
+      "loss": 0.3354,
+      "step": 10856
+    },
+    {
+      "epoch": 0.47676652359572763,
+      "grad_norm": 1.8984375,
+      "learning_rate": 2.694725210067368e-05,
+      "loss": 0.3991,
+      "step": 10858
+    },
+    {
+      "epoch": 0.4768543420749311,
+      "grad_norm": 2.15625,
+      "learning_rate": 2.6940354760519616e-05,
+      "loss": 0.3437,
+      "step": 10860
+    },
+    {
+      "epoch": 0.47694216055413463,
+      "grad_norm": 1.9140625,
+      "learning_rate": 2.6933457271772718e-05,
+      "loss": 0.3562,
+      "step": 10862
+    },
+    {
+      "epoch": 0.4770299790333381,
+      "grad_norm": 1.921875,
+      "learning_rate": 2.692655963496121e-05,
+      "loss": 0.4094,
+      "step": 10864
+    },
+    {
+      "epoch": 0.4771177975125416,
+      "grad_norm": 1.859375,
+      "learning_rate": 2.6919661850613287e-05,
+      "loss": 0.3828,
+      "step": 10866
+    },
+    {
+      "epoch": 0.47720561599174505,
+      "grad_norm": 1.8984375,
+      "learning_rate": 2.6912763919257207e-05,
+      "loss": 0.3572,
+      "step": 10868
+    },
+    {
+      "epoch": 0.4772934344709486,
+      "grad_norm": 2.28125,
+      "learning_rate": 2.690586584142121e-05,
+      "loss": 0.3654,
+      "step": 10870
+    },
+    {
+      "epoch": 0.47738125295015205,
+      "grad_norm": 2.375,
+      "learning_rate": 2.6898967617633546e-05,
+      "loss": 0.3713,
+      "step": 10872
+    },
+    {
+      "epoch": 0.4774690714293555,
+      "grad_norm": 1.953125,
+      "learning_rate": 2.6892069248422486e-05,
+      "loss": 0.3445,
+      "step": 10874
+    },
+    {
+      "epoch": 0.477556889908559,
+      "grad_norm": 1.8203125,
+      "learning_rate": 2.6885170734316317e-05,
+      "loss": 0.36,
+      "step": 10876
+    },
+    {
+      "epoch": 0.4776447083877625,
+      "grad_norm": 1.9765625,
+      "learning_rate": 2.6878272075843313e-05,
+      "loss": 0.334,
+      "step": 10878
+    },
+    {
+      "epoch": 0.477732526866966,
+      "grad_norm": 1.875,
+      "learning_rate": 2.6871373273531785e-05,
+      "loss": 0.3442,
+      "step": 10880
+    },
+    {
+      "epoch": 0.47782034534616946,
+      "grad_norm": 1.890625,
+      "learning_rate": 2.6864474327910038e-05,
+      "loss": 0.3635,
+      "step": 10882
+    },
+    {
+      "epoch": 0.47790816382537293,
+      "grad_norm": 1.6796875,
+      "learning_rate": 2.6857575239506406e-05,
+      "loss": 0.3776,
+      "step": 10884
+    },
+    {
+      "epoch": 0.47799598230457646,
+      "grad_norm": 1.6953125,
+      "learning_rate": 2.6850676008849222e-05,
+      "loss": 0.3681,
+      "step": 10886
+    },
+    {
+      "epoch": 0.47808380078377993,
+      "grad_norm": 1.78125,
+      "learning_rate": 2.6843776636466828e-05,
+      "loss": 0.3486,
+      "step": 10888
+    },
+    {
+      "epoch": 0.4781716192629834,
+      "grad_norm": 1.75,
+      "learning_rate": 2.6836877122887565e-05,
+      "loss": 0.3536,
+      "step": 10890
+    },
+    {
+      "epoch": 0.4782594377421869,
+      "grad_norm": 1.7421875,
+      "learning_rate": 2.6829977468639815e-05,
+      "loss": 0.3377,
+      "step": 10892
+    },
+    {
+      "epoch": 0.4783472562213904,
+      "grad_norm": 2.4375,
+      "learning_rate": 2.682307767425195e-05,
+      "loss": 0.357,
+      "step": 10894
+    },
+    {
+      "epoch": 0.4784350747005939,
+      "grad_norm": 1.96875,
+      "learning_rate": 2.6816177740252362e-05,
+      "loss": 0.3781,
+      "step": 10896
+    },
+    {
+      "epoch": 0.47852289317979735,
+      "grad_norm": 1.71875,
+      "learning_rate": 2.6809277667169446e-05,
+      "loss": 0.3761,
+      "step": 10898
+    },
+    {
+      "epoch": 0.4786107116590008,
+      "grad_norm": 1.78125,
+      "learning_rate": 2.680237745553162e-05,
+      "loss": 0.3505,
+      "step": 10900
+    },
+    {
+      "epoch": 0.47869853013820435,
+      "grad_norm": 1.9765625,
+      "learning_rate": 2.6795477105867295e-05,
+      "loss": 0.3985,
+      "step": 10902
+    },
+    {
+      "epoch": 0.4787863486174078,
+      "grad_norm": 2.03125,
+      "learning_rate": 2.6788576618704897e-05,
+      "loss": 0.382,
+      "step": 10904
+    },
+    {
+      "epoch": 0.4788741670966113,
+      "grad_norm": 1.6875,
+      "learning_rate": 2.6781675994572876e-05,
+      "loss": 0.3692,
+      "step": 10906
+    },
+    {
+      "epoch": 0.47896198557581476,
+      "grad_norm": 2.046875,
+      "learning_rate": 2.677477523399967e-05,
+      "loss": 0.3521,
+      "step": 10908
+    },
+    {
+      "epoch": 0.4790498040550183,
+      "grad_norm": 2.328125,
+      "learning_rate": 2.676787433751376e-05,
+      "loss": 0.3662,
+      "step": 10910
+    },
+    {
+      "epoch": 0.47913762253422176,
+      "grad_norm": 2.234375,
+      "learning_rate": 2.6760973305643616e-05,
+      "loss": 0.3562,
+      "step": 10912
+    },
+    {
+      "epoch": 0.47922544101342524,
+      "grad_norm": 1.75,
+      "learning_rate": 2.6754072138917708e-05,
+      "loss": 0.362,
+      "step": 10914
+    },
+    {
+      "epoch": 0.47931325949262876,
+      "grad_norm": 2.265625,
+      "learning_rate": 2.674717083786454e-05,
+      "loss": 0.3507,
+      "step": 10916
+    },
+    {
+      "epoch": 0.47940107797183223,
+      "grad_norm": 3.125,
+      "learning_rate": 2.6740269403012595e-05,
+      "loss": 0.3584,
+      "step": 10918
+    },
+    {
+      "epoch": 0.4794888964510357,
+      "grad_norm": 1.796875,
+      "learning_rate": 2.6733367834890415e-05,
+      "loss": 0.376,
+      "step": 10920
+    },
+    {
+      "epoch": 0.4795767149302392,
+      "grad_norm": 1.75,
+      "learning_rate": 2.6726466134026507e-05,
+      "loss": 0.3462,
+      "step": 10922
+    },
+    {
+      "epoch": 0.4796645334094427,
+      "grad_norm": 1.7265625,
+      "learning_rate": 2.6719564300949413e-05,
+      "loss": 0.3701,
+      "step": 10924
+    },
+    {
+      "epoch": 0.4797523518886462,
+      "grad_norm": 1.9921875,
+      "learning_rate": 2.6712662336187667e-05,
+      "loss": 0.3719,
+      "step": 10926
+    },
+    {
+      "epoch": 0.47984017036784965,
+      "grad_norm": 1.9921875,
+      "learning_rate": 2.6705760240269828e-05,
+      "loss": 0.3555,
+      "step": 10928
+    },
+    {
+      "epoch": 0.4799279888470531,
+      "grad_norm": 1.78125,
+      "learning_rate": 2.6698858013724466e-05,
+      "loss": 0.365,
+      "step": 10930
+    },
+    {
+      "epoch": 0.48001580732625665,
+      "grad_norm": 1.7734375,
+      "learning_rate": 2.669195565708015e-05,
+      "loss": 0.3934,
+      "step": 10932
+    },
+    {
+      "epoch": 0.4801036258054601,
+      "grad_norm": 1.96875,
+      "learning_rate": 2.668505317086546e-05,
+      "loss": 0.3618,
+      "step": 10934
+    },
+    {
+      "epoch": 0.4801914442846636,
+      "grad_norm": 2.171875,
+      "learning_rate": 2.6678150555608993e-05,
+      "loss": 0.3615,
+      "step": 10936
+    },
+    {
+      "epoch": 0.48027926276386707,
+      "grad_norm": 2.25,
+      "learning_rate": 2.667124781183936e-05,
+      "loss": 0.3968,
+      "step": 10938
+    },
+    {
+      "epoch": 0.4803670812430706,
+      "grad_norm": 1.671875,
+      "learning_rate": 2.6664344940085166e-05,
+      "loss": 0.3858,
+      "step": 10940
+    },
+    {
+      "epoch": 0.48045489972227406,
+      "grad_norm": 2.046875,
+      "learning_rate": 2.665744194087503e-05,
+      "loss": 0.3488,
+      "step": 10942
+    },
+    {
+      "epoch": 0.48054271820147754,
+      "grad_norm": 2.40625,
+      "learning_rate": 2.6650538814737603e-05,
+      "loss": 0.3385,
+      "step": 10944
+    },
+    {
+      "epoch": 0.480630536680681,
+      "grad_norm": 1.890625,
+      "learning_rate": 2.6643635562201513e-05,
+      "loss": 0.3834,
+      "step": 10946
+    },
+    {
+      "epoch": 0.48071835515988454,
+      "grad_norm": 1.9375,
+      "learning_rate": 2.6636732183795426e-05,
+      "loss": 0.3588,
+      "step": 10948
+    },
+    {
+      "epoch": 0.480806173639088,
+      "grad_norm": 1.9140625,
+      "learning_rate": 2.6629828680047998e-05,
+      "loss": 0.3857,
+      "step": 10950
+    },
+    {
+      "epoch": 0.4808939921182915,
+      "grad_norm": 2.234375,
+      "learning_rate": 2.6622925051487884e-05,
+      "loss": 0.3765,
+      "step": 10952
+    },
+    {
+      "epoch": 0.48098181059749495,
+      "grad_norm": 2.421875,
+      "learning_rate": 2.6616021298643788e-05,
+      "loss": 0.3642,
+      "step": 10954
+    },
+    {
+      "epoch": 0.4810696290766985,
+      "grad_norm": 1.8046875,
+      "learning_rate": 2.6609117422044394e-05,
+      "loss": 0.3503,
+      "step": 10956
+    },
+    {
+      "epoch": 0.48115744755590195,
+      "grad_norm": 2.125,
+      "learning_rate": 2.66022134222184e-05,
+      "loss": 0.3387,
+      "step": 10958
+    },
+    {
+      "epoch": 0.4812452660351054,
+      "grad_norm": 2.109375,
+      "learning_rate": 2.659530929969452e-05,
+      "loss": 0.3796,
+      "step": 10960
+    },
+    {
+      "epoch": 0.48133308451430895,
+      "grad_norm": 1.9296875,
+      "learning_rate": 2.658840505500147e-05,
+      "loss": 0.3883,
+      "step": 10962
+    },
+    {
+      "epoch": 0.4814209029935124,
+      "grad_norm": 2.4375,
+      "learning_rate": 2.658150068866798e-05,
+      "loss": 0.3505,
+      "step": 10964
+    },
+    {
+      "epoch": 0.4815087214727159,
+      "grad_norm": 1.5859375,
+      "learning_rate": 2.657459620122279e-05,
+      "loss": 0.3476,
+      "step": 10966
+    },
+    {
+      "epoch": 0.48159653995191937,
+      "grad_norm": 2.046875,
+      "learning_rate": 2.6567691593194644e-05,
+      "loss": 0.3604,
+      "step": 10968
+    },
+    {
+      "epoch": 0.4816843584311229,
+      "grad_norm": 1.8828125,
+      "learning_rate": 2.6560786865112296e-05,
+      "loss": 0.3394,
+      "step": 10970
+    },
+    {
+      "epoch": 0.48177217691032637,
+      "grad_norm": 1.890625,
+      "learning_rate": 2.6553882017504523e-05,
+      "loss": 0.3632,
+      "step": 10972
+    },
+    {
+      "epoch": 0.48185999538952984,
+      "grad_norm": 1.765625,
+      "learning_rate": 2.654697705090009e-05,
+      "loss": 0.3815,
+      "step": 10974
+    },
+    {
+      "epoch": 0.4819478138687333,
+      "grad_norm": 1.9296875,
+      "learning_rate": 2.6540071965827784e-05,
+      "loss": 0.3448,
+      "step": 10976
+    },
+    {
+      "epoch": 0.48203563234793684,
+      "grad_norm": 1.8671875,
+      "learning_rate": 2.6533166762816404e-05,
+      "loss": 0.3669,
+      "step": 10978
+    },
+    {
+      "epoch": 0.4821234508271403,
+      "grad_norm": 2.109375,
+      "learning_rate": 2.6526261442394734e-05,
+      "loss": 0.3838,
+      "step": 10980
+    },
+    {
+      "epoch": 0.4822112693063438,
+      "grad_norm": 2.03125,
+      "learning_rate": 2.65193560050916e-05,
+      "loss": 0.3641,
+      "step": 10982
+    },
+    {
+      "epoch": 0.48229908778554725,
+      "grad_norm": 1.7421875,
+      "learning_rate": 2.6512450451435822e-05,
+      "loss": 0.3629,
+      "step": 10984
+    },
+    {
+      "epoch": 0.4823869062647508,
+      "grad_norm": 2.03125,
+      "learning_rate": 2.6505544781956225e-05,
+      "loss": 0.3732,
+      "step": 10986
+    },
+    {
+      "epoch": 0.48247472474395425,
+      "grad_norm": 1.7734375,
+      "learning_rate": 2.6498638997181648e-05,
+      "loss": 0.3918,
+      "step": 10988
+    },
+    {
+      "epoch": 0.4825625432231577,
+      "grad_norm": 1.8515625,
+      "learning_rate": 2.649173309764093e-05,
+      "loss": 0.3696,
+      "step": 10990
+    },
+    {
+      "epoch": 0.4826503617023612,
+      "grad_norm": 1.8515625,
+      "learning_rate": 2.648482708386294e-05,
+      "loss": 0.3544,
+      "step": 10992
+    },
+    {
+      "epoch": 0.4827381801815647,
+      "grad_norm": 1.734375,
+      "learning_rate": 2.647792095637654e-05,
+      "loss": 0.3607,
+      "step": 10994
+    },
+    {
+      "epoch": 0.4828259986607682,
+      "grad_norm": 2.140625,
+      "learning_rate": 2.6471014715710595e-05,
+      "loss": 0.3559,
+      "step": 10996
+    },
+    {
+      "epoch": 0.48291381713997167,
+      "grad_norm": 1.8515625,
+      "learning_rate": 2.646410836239399e-05,
+      "loss": 0.3294,
+      "step": 10998
+    },
+    {
+      "epoch": 0.48300163561917514,
+      "grad_norm": 1.734375,
+      "learning_rate": 2.645720189695562e-05,
+      "loss": 0.3633,
+      "step": 11000
+    },
+    {
+      "epoch": 0.48308945409837867,
+      "grad_norm": 1.9140625,
+      "learning_rate": 2.6450295319924374e-05,
+      "loss": 0.359,
+      "step": 11002
+    },
+    {
+      "epoch": 0.48317727257758214,
+      "grad_norm": 1.953125,
+      "learning_rate": 2.6443388631829162e-05,
+      "loss": 0.3626,
+      "step": 11004
+    },
+    {
+      "epoch": 0.4832650910567856,
+      "grad_norm": 1.84375,
+      "learning_rate": 2.6436481833198905e-05,
+      "loss": 0.3442,
+      "step": 11006
+    },
+    {
+      "epoch": 0.4833529095359891,
+      "grad_norm": 1.8203125,
+      "learning_rate": 2.6429574924562534e-05,
+      "loss": 0.3785,
+      "step": 11008
+    },
+    {
+      "epoch": 0.4834407280151926,
+      "grad_norm": 1.71875,
+      "learning_rate": 2.642266790644896e-05,
+      "loss": 0.3794,
+      "step": 11010
+    },
+    {
+      "epoch": 0.4835285464943961,
+      "grad_norm": 2.015625,
+      "learning_rate": 2.641576077938715e-05,
+      "loss": 0.3482,
+      "step": 11012
+    },
+    {
+      "epoch": 0.48361636497359956,
+      "grad_norm": 1.875,
+      "learning_rate": 2.640885354390603e-05,
+      "loss": 0.3612,
+      "step": 11014
+    },
+    {
+      "epoch": 0.4837041834528031,
+      "grad_norm": 1.875,
+      "learning_rate": 2.6401946200534567e-05,
+      "loss": 0.3577,
+      "step": 11016
+    },
+    {
+      "epoch": 0.48379200193200655,
+      "grad_norm": 1.9296875,
+      "learning_rate": 2.639503874980173e-05,
+      "loss": 0.3491,
+      "step": 11018
+    },
+    {
+      "epoch": 0.48387982041121,
+      "grad_norm": 1.78125,
+      "learning_rate": 2.6388131192236493e-05,
+      "loss": 0.372,
+      "step": 11020
+    },
+    {
+      "epoch": 0.4839676388904135,
+      "grad_norm": 1.7421875,
+      "learning_rate": 2.6381223528367836e-05,
+      "loss": 0.3848,
+      "step": 11022
+    },
+    {
+      "epoch": 0.484055457369617,
+      "grad_norm": 1.8359375,
+      "learning_rate": 2.6374315758724755e-05,
+      "loss": 0.3491,
+      "step": 11024
+    },
+    {
+      "epoch": 0.4841432758488205,
+      "grad_norm": 1.859375,
+      "learning_rate": 2.6367407883836238e-05,
+      "loss": 0.3776,
+      "step": 11026
+    },
+    {
+      "epoch": 0.48423109432802397,
+      "grad_norm": 1.8359375,
+      "learning_rate": 2.6360499904231297e-05,
+      "loss": 0.3584,
+      "step": 11028
+    },
+    {
+      "epoch": 0.48431891280722744,
+      "grad_norm": 1.703125,
+      "learning_rate": 2.6353591820438945e-05,
+      "loss": 0.356,
+      "step": 11030
+    },
+    {
+      "epoch": 0.48440673128643097,
+      "grad_norm": 1.8515625,
+      "learning_rate": 2.6346683632988213e-05,
+      "loss": 0.3673,
+      "step": 11032
+    },
+    {
+      "epoch": 0.48449454976563444,
+      "grad_norm": 1.8671875,
+      "learning_rate": 2.633977534240812e-05,
+      "loss": 0.3659,
+      "step": 11034
+    },
+    {
+      "epoch": 0.4845823682448379,
+      "grad_norm": 1.84375,
+      "learning_rate": 2.6332866949227713e-05,
+      "loss": 0.3441,
+      "step": 11036
+    },
+    {
+      "epoch": 0.4846701867240414,
+      "grad_norm": 1.828125,
+      "learning_rate": 2.6325958453976036e-05,
+      "loss": 0.374,
+      "step": 11038
+    },
+    {
+      "epoch": 0.4847580052032449,
+      "grad_norm": 1.7265625,
+      "learning_rate": 2.6319049857182144e-05,
+      "loss": 0.3385,
+      "step": 11040
+    },
+    {
+      "epoch": 0.4848458236824484,
+      "grad_norm": 2.03125,
+      "learning_rate": 2.631214115937509e-05,
+      "loss": 0.3632,
+      "step": 11042
+    },
+    {
+      "epoch": 0.48493364216165186,
+      "grad_norm": 1.859375,
+      "learning_rate": 2.630523236108396e-05,
+      "loss": 0.3257,
+      "step": 11044
+    },
+    {
+      "epoch": 0.48502146064085533,
+      "grad_norm": 1.7578125,
+      "learning_rate": 2.629832346283782e-05,
+      "loss": 0.3693,
+      "step": 11046
+    },
+    {
+      "epoch": 0.48510927912005886,
+      "grad_norm": 2.15625,
+      "learning_rate": 2.629141446516576e-05,
+      "loss": 0.3815,
+      "step": 11048
+    },
+    {
+      "epoch": 0.48519709759926233,
+      "grad_norm": 1.984375,
+      "learning_rate": 2.628450536859687e-05,
+      "loss": 0.3517,
+      "step": 11050
+    },
+    {
+      "epoch": 0.4852849160784658,
+      "grad_norm": 1.75,
+      "learning_rate": 2.6277596173660246e-05,
+      "loss": 0.3754,
+      "step": 11052
+    },
+    {
+      "epoch": 0.48537273455766927,
+      "grad_norm": 1.75,
+      "learning_rate": 2.627068688088501e-05,
+      "loss": 0.401,
+      "step": 11054
+    },
+    {
+      "epoch": 0.4854605530368728,
+      "grad_norm": 1.7578125,
+      "learning_rate": 2.6263777490800263e-05,
+      "loss": 0.3613,
+      "step": 11056
+    },
+    {
+      "epoch": 0.48554837151607627,
+      "grad_norm": 1.6875,
+      "learning_rate": 2.6256868003935127e-05,
+      "loss": 0.3688,
+      "step": 11058
+    },
+    {
+      "epoch": 0.48563618999527974,
+      "grad_norm": 1.75,
+      "learning_rate": 2.6249958420818744e-05,
+      "loss": 0.3926,
+      "step": 11060
+    },
+    {
+      "epoch": 0.48572400847448327,
+      "grad_norm": 1.6796875,
+      "learning_rate": 2.6243048741980246e-05,
+      "loss": 0.3964,
+      "step": 11062
+    },
+    {
+      "epoch": 0.48581182695368674,
+      "grad_norm": 1.7578125,
+      "learning_rate": 2.6236138967948776e-05,
+      "loss": 0.3771,
+      "step": 11064
+    },
+    {
+      "epoch": 0.4858996454328902,
+      "grad_norm": 1.859375,
+      "learning_rate": 2.622922909925349e-05,
+      "loss": 0.3446,
+      "step": 11066
+    },
+    {
+      "epoch": 0.4859874639120937,
+      "grad_norm": 1.9140625,
+      "learning_rate": 2.6222319136423533e-05,
+      "loss": 0.337,
+      "step": 11068
+    },
+    {
+      "epoch": 0.4860752823912972,
+      "grad_norm": 1.8125,
+      "learning_rate": 2.62154090799881e-05,
+      "loss": 0.3735,
+      "step": 11070
+    },
+    {
+      "epoch": 0.4861631008705007,
+      "grad_norm": 1.7421875,
+      "learning_rate": 2.6208498930476337e-05,
+      "loss": 0.3807,
+      "step": 11072
+    },
+    {
+      "epoch": 0.48625091934970416,
+      "grad_norm": 1.671875,
+      "learning_rate": 2.620158868841745e-05,
+      "loss": 0.3486,
+      "step": 11074
+    },
+    {
+      "epoch": 0.48633873782890763,
+      "grad_norm": 1.6640625,
+      "learning_rate": 2.6194678354340602e-05,
+      "loss": 0.3378,
+      "step": 11076
+    },
+    {
+      "epoch": 0.48642655630811116,
+      "grad_norm": 1.8359375,
+      "learning_rate": 2.6187767928774998e-05,
+      "loss": 0.3508,
+      "step": 11078
+    },
+    {
+      "epoch": 0.48651437478731463,
+      "grad_norm": 1.8359375,
+      "learning_rate": 2.6180857412249842e-05,
+      "loss": 0.3512,
+      "step": 11080
+    },
+    {
+      "epoch": 0.4866021932665181,
+      "grad_norm": 1.75,
+      "learning_rate": 2.6173946805294348e-05,
+      "loss": 0.338,
+      "step": 11082
+    },
+    {
+      "epoch": 0.4866900117457216,
+      "grad_norm": 1.859375,
+      "learning_rate": 2.6167036108437724e-05,
+      "loss": 0.3863,
+      "step": 11084
+    },
+    {
+      "epoch": 0.4867778302249251,
+      "grad_norm": 1.7734375,
+      "learning_rate": 2.6160125322209194e-05,
+      "loss": 0.3513,
+      "step": 11086
+    },
+    {
+      "epoch": 0.4868656487041286,
+      "grad_norm": 1.8515625,
+      "learning_rate": 2.615321444713799e-05,
+      "loss": 0.3806,
+      "step": 11088
+    },
+    {
+      "epoch": 0.48695346718333204,
+      "grad_norm": 2.0,
+      "learning_rate": 2.6146303483753343e-05,
+      "loss": 0.3856,
+      "step": 11090
+    },
+    {
+      "epoch": 0.4870412856625355,
+      "grad_norm": 1.71875,
+      "learning_rate": 2.6139392432584504e-05,
+      "loss": 0.3408,
+      "step": 11092
+    },
+    {
+      "epoch": 0.48712910414173904,
+      "grad_norm": 1.8203125,
+      "learning_rate": 2.6132481294160715e-05,
+      "loss": 0.3723,
+      "step": 11094
+    },
+    {
+      "epoch": 0.4872169226209425,
+      "grad_norm": 2.125,
+      "learning_rate": 2.612557006901124e-05,
+      "loss": 0.3541,
+      "step": 11096
+    },
+    {
+      "epoch": 0.487304741100146,
+      "grad_norm": 1.7109375,
+      "learning_rate": 2.6118658757665343e-05,
+      "loss": 0.37,
+      "step": 11098
+    },
+    {
+      "epoch": 0.48739255957934946,
+      "grad_norm": 1.9453125,
+      "learning_rate": 2.6111747360652295e-05,
+      "loss": 0.3398,
+      "step": 11100
+    },
+    {
+      "epoch": 0.487480378058553,
+      "grad_norm": 1.671875,
+      "learning_rate": 2.6104835878501353e-05,
+      "loss": 0.3591,
+      "step": 11102
+    },
+    {
+      "epoch": 0.48756819653775646,
+      "grad_norm": 1.9453125,
+      "learning_rate": 2.609792431174182e-05,
+      "loss": 0.3655,
+      "step": 11104
+    },
+    {
+      "epoch": 0.48765601501695993,
+      "grad_norm": 1.7109375,
+      "learning_rate": 2.609101266090298e-05,
+      "loss": 0.3481,
+      "step": 11106
+    },
+    {
+      "epoch": 0.4877438334961634,
+      "grad_norm": 1.9765625,
+      "learning_rate": 2.608410092651413e-05,
+      "loss": 0.372,
+      "step": 11108
+    },
+    {
+      "epoch": 0.48783165197536693,
+      "grad_norm": 1.9140625,
+      "learning_rate": 2.6077189109104577e-05,
+      "loss": 0.3566,
+      "step": 11110
+    },
+    {
+      "epoch": 0.4879194704545704,
+      "grad_norm": 2.09375,
+      "learning_rate": 2.6070277209203613e-05,
+      "loss": 0.352,
+      "step": 11112
+    },
+    {
+      "epoch": 0.4880072889337739,
+      "grad_norm": 1.828125,
+      "learning_rate": 2.606336522734057e-05,
+      "loss": 0.3427,
+      "step": 11114
+    },
+    {
+      "epoch": 0.4880951074129774,
+      "grad_norm": 1.8203125,
+      "learning_rate": 2.605645316404477e-05,
+      "loss": 0.3579,
+      "step": 11116
+    },
+    {
+      "epoch": 0.4881829258921809,
+      "grad_norm": 2.015625,
+      "learning_rate": 2.604954101984552e-05,
+      "loss": 0.3444,
+      "step": 11118
+    },
+    {
+      "epoch": 0.48827074437138435,
+      "grad_norm": 1.8671875,
+      "learning_rate": 2.6042628795272174e-05,
+      "loss": 0.345,
+      "step": 11120
+    },
+    {
+      "epoch": 0.4883585628505878,
+      "grad_norm": 1.8046875,
+      "learning_rate": 2.6035716490854067e-05,
+      "loss": 0.3319,
+      "step": 11122
+    },
+    {
+      "epoch": 0.48844638132979135,
+      "grad_norm": 1.75,
+      "learning_rate": 2.6028804107120535e-05,
+      "loss": 0.3964,
+      "step": 11124
+    },
+    {
+      "epoch": 0.4885341998089948,
+      "grad_norm": 1.6875,
+      "learning_rate": 2.6021891644600944e-05,
+      "loss": 0.3471,
+      "step": 11126
+    },
+    {
+      "epoch": 0.4886220182881983,
+      "grad_norm": 2.015625,
+      "learning_rate": 2.6014979103824637e-05,
+      "loss": 0.3365,
+      "step": 11128
+    },
+    {
+      "epoch": 0.48870983676740176,
+      "grad_norm": 1.640625,
+      "learning_rate": 2.6008066485320998e-05,
+      "loss": 0.3383,
+      "step": 11130
+    },
+    {
+      "epoch": 0.4887976552466053,
+      "grad_norm": 2.125,
+      "learning_rate": 2.6001153789619383e-05,
+      "loss": 0.3568,
+      "step": 11132
+    },
+    {
+      "epoch": 0.48888547372580876,
+      "grad_norm": 1.8359375,
+      "learning_rate": 2.5994241017249167e-05,
+      "loss": 0.3439,
+      "step": 11134
+    },
+    {
+      "epoch": 0.48897329220501223,
+      "grad_norm": 2.1875,
+      "learning_rate": 2.5987328168739745e-05,
+      "loss": 0.3543,
+      "step": 11136
+    },
+    {
+      "epoch": 0.4890611106842157,
+      "grad_norm": 1.8984375,
+      "learning_rate": 2.5980415244620486e-05,
+      "loss": 0.3801,
+      "step": 11138
+    },
+    {
+      "epoch": 0.48914892916341923,
+      "grad_norm": 1.8359375,
+      "learning_rate": 2.5973502245420793e-05,
+      "loss": 0.3759,
+      "step": 11140
+    },
+    {
+      "epoch": 0.4892367476426227,
+      "grad_norm": 1.6953125,
+      "learning_rate": 2.5966589171670064e-05,
+      "loss": 0.3677,
+      "step": 11142
+    },
+    {
+      "epoch": 0.4893245661218262,
+      "grad_norm": 1.9453125,
+      "learning_rate": 2.5959676023897706e-05,
+      "loss": 0.3884,
+      "step": 11144
+    },
+    {
+      "epoch": 0.48941238460102965,
+      "grad_norm": 1.75,
+      "learning_rate": 2.5952762802633136e-05,
+      "loss": 0.3424,
+      "step": 11146
+    },
+    {
+      "epoch": 0.4895002030802332,
+      "grad_norm": 1.734375,
+      "learning_rate": 2.5945849508405755e-05,
+      "loss": 0.3576,
+      "step": 11148
+    },
+    {
+      "epoch": 0.48958802155943665,
+      "grad_norm": 1.7421875,
+      "learning_rate": 2.593893614174499e-05,
+      "loss": 0.3672,
+      "step": 11150
+    },
+    {
+      "epoch": 0.4896758400386401,
+      "grad_norm": 1.796875,
+      "learning_rate": 2.5932022703180277e-05,
+      "loss": 0.3371,
+      "step": 11152
+    },
+    {
+      "epoch": 0.4897636585178436,
+      "grad_norm": 1.6953125,
+      "learning_rate": 2.592510919324103e-05,
+      "loss": 0.3762,
+      "step": 11154
+    },
+    {
+      "epoch": 0.4898514769970471,
+      "grad_norm": 1.65625,
+      "learning_rate": 2.591819561245671e-05,
+      "loss": 0.3559,
+      "step": 11156
+    },
+    {
+      "epoch": 0.4899392954762506,
+      "grad_norm": 1.8515625,
+      "learning_rate": 2.591128196135675e-05,
+      "loss": 0.3713,
+      "step": 11158
+    },
+    {
+      "epoch": 0.49002711395545406,
+      "grad_norm": 2.046875,
+      "learning_rate": 2.5904368240470595e-05,
+      "loss": 0.3572,
+      "step": 11160
+    },
+    {
+      "epoch": 0.49011493243465754,
+      "grad_norm": 1.8984375,
+      "learning_rate": 2.5897454450327707e-05,
+      "loss": 0.3401,
+      "step": 11162
+    },
+    {
+      "epoch": 0.49020275091386106,
+      "grad_norm": 2.15625,
+      "learning_rate": 2.5890540591457535e-05,
+      "loss": 0.3697,
+      "step": 11164
+    },
+    {
+      "epoch": 0.49029056939306453,
+      "grad_norm": 1.7265625,
+      "learning_rate": 2.5883626664389556e-05,
+      "loss": 0.3574,
+      "step": 11166
+    },
+    {
+      "epoch": 0.490378387872268,
+      "grad_norm": 1.7578125,
+      "learning_rate": 2.5876712669653232e-05,
+      "loss": 0.3785,
+      "step": 11168
+    },
+    {
+      "epoch": 0.49046620635147153,
+      "grad_norm": 1.7578125,
+      "learning_rate": 2.586979860777804e-05,
+      "loss": 0.3473,
+      "step": 11170
+    },
+    {
+      "epoch": 0.490554024830675,
+      "grad_norm": 1.609375,
+      "learning_rate": 2.586288447929347e-05,
+      "loss": 0.349,
+      "step": 11172
+    },
+    {
+      "epoch": 0.4906418433098785,
+      "grad_norm": 1.8671875,
+      "learning_rate": 2.5855970284728993e-05,
+      "loss": 0.3541,
+      "step": 11174
+    },
+    {
+      "epoch": 0.49072966178908195,
+      "grad_norm": 1.8515625,
+      "learning_rate": 2.584905602461411e-05,
+      "loss": 0.3555,
+      "step": 11176
+    },
+    {
+      "epoch": 0.4908174802682855,
+      "grad_norm": 1.8125,
+      "learning_rate": 2.5842141699478317e-05,
+      "loss": 0.3449,
+      "step": 11178
+    },
+    {
+      "epoch": 0.49090529874748895,
+      "grad_norm": 1.6953125,
+      "learning_rate": 2.58352273098511e-05,
+      "loss": 0.3378,
+      "step": 11180
+    },
+    {
+      "epoch": 0.4909931172266924,
+      "grad_norm": 1.7890625,
+      "learning_rate": 2.5828312856261982e-05,
+      "loss": 0.3507,
+      "step": 11182
+    },
+    {
+      "epoch": 0.4910809357058959,
+      "grad_norm": 1.90625,
+      "learning_rate": 2.582139833924047e-05,
+      "loss": 0.3718,
+      "step": 11184
+    },
+    {
+      "epoch": 0.4911687541850994,
+      "grad_norm": 2.03125,
+      "learning_rate": 2.581448375931607e-05,
+      "loss": 0.3828,
+      "step": 11186
+    },
+    {
+      "epoch": 0.4912565726643029,
+      "grad_norm": 1.8515625,
+      "learning_rate": 2.580756911701831e-05,
+      "loss": 0.3782,
+      "step": 11188
+    },
+    {
+      "epoch": 0.49134439114350636,
+      "grad_norm": 2.0,
+      "learning_rate": 2.5800654412876713e-05,
+      "loss": 0.3224,
+      "step": 11190
+    },
+    {
+      "epoch": 0.49143220962270984,
+      "grad_norm": 1.8671875,
+      "learning_rate": 2.5793739647420817e-05,
+      "loss": 0.3679,
+      "step": 11192
+    },
+    {
+      "epoch": 0.49152002810191336,
+      "grad_norm": 1.765625,
+      "learning_rate": 2.578682482118014e-05,
+      "loss": 0.3788,
+      "step": 11194
+    },
+    {
+      "epoch": 0.49160784658111684,
+      "grad_norm": 1.9609375,
+      "learning_rate": 2.5779909934684227e-05,
+      "loss": 0.3789,
+      "step": 11196
+    },
+    {
+      "epoch": 0.4916956650603203,
+      "grad_norm": 1.9609375,
+      "learning_rate": 2.5772994988462634e-05,
+      "loss": 0.3832,
+      "step": 11198
+    },
+    {
+      "epoch": 0.4917834835395238,
+      "grad_norm": 1.8671875,
+      "learning_rate": 2.5766079983044893e-05,
+      "loss": 0.3572,
+      "step": 11200
+    },
+    {
+      "epoch": 0.4918713020187273,
+      "grad_norm": 1.703125,
+      "learning_rate": 2.5759164918960567e-05,
+      "loss": 0.3766,
+      "step": 11202
+    },
+    {
+      "epoch": 0.4919591204979308,
+      "grad_norm": 1.75,
+      "learning_rate": 2.5752249796739202e-05,
+      "loss": 0.3484,
+      "step": 11204
+    },
+    {
+      "epoch": 0.49204693897713425,
+      "grad_norm": 1.9921875,
+      "learning_rate": 2.574533461691037e-05,
+      "loss": 0.3933,
+      "step": 11206
+    },
+    {
+      "epoch": 0.4921347574563377,
+      "grad_norm": 2.078125,
+      "learning_rate": 2.5738419380003642e-05,
+      "loss": 0.346,
+      "step": 11208
+    },
+    {
+      "epoch": 0.49222257593554125,
+      "grad_norm": 2.28125,
+      "learning_rate": 2.573150408654858e-05,
+      "loss": 0.3779,
+      "step": 11210
+    },
+    {
+      "epoch": 0.4923103944147447,
+      "grad_norm": 1.7734375,
+      "learning_rate": 2.572458873707475e-05,
+      "loss": 0.3431,
+      "step": 11212
+    },
+    {
+      "epoch": 0.4923982128939482,
+      "grad_norm": 1.765625,
+      "learning_rate": 2.5717673332111748e-05,
+      "loss": 0.3638,
+      "step": 11214
+    },
+    {
+      "epoch": 0.4924860313731517,
+      "grad_norm": 1.8515625,
+      "learning_rate": 2.5710757872189145e-05,
+      "loss": 0.3546,
+      "step": 11216
+    },
+    {
+      "epoch": 0.4925738498523552,
+      "grad_norm": 1.5859375,
+      "learning_rate": 2.5703842357836537e-05,
+      "loss": 0.3372,
+      "step": 11218
+    },
+    {
+      "epoch": 0.49266166833155867,
+      "grad_norm": 1.7421875,
+      "learning_rate": 2.5696926789583513e-05,
+      "loss": 0.3499,
+      "step": 11220
+    },
+    {
+      "epoch": 0.49274948681076214,
+      "grad_norm": 2.0,
+      "learning_rate": 2.569001116795967e-05,
+      "loss": 0.3435,
+      "step": 11222
+    },
+    {
+      "epoch": 0.49283730528996567,
+      "grad_norm": 1.875,
+      "learning_rate": 2.5683095493494607e-05,
+      "loss": 0.3628,
+      "step": 11224
+    },
+    {
+      "epoch": 0.49292512376916914,
+      "grad_norm": 1.8203125,
+      "learning_rate": 2.5676179766717923e-05,
+      "loss": 0.3492,
+      "step": 11226
+    },
+    {
+      "epoch": 0.4930129422483726,
+      "grad_norm": 1.6640625,
+      "learning_rate": 2.566926398815923e-05,
+      "loss": 0.3667,
+      "step": 11228
+    },
+    {
+      "epoch": 0.4931007607275761,
+      "grad_norm": 1.6484375,
+      "learning_rate": 2.5662348158348142e-05,
+      "loss": 0.3326,
+      "step": 11230
+    },
+    {
+      "epoch": 0.4931885792067796,
+      "grad_norm": 1.78125,
+      "learning_rate": 2.5655432277814273e-05,
+      "loss": 0.3653,
+      "step": 11232
+    },
+    {
+      "epoch": 0.4932763976859831,
+      "grad_norm": 1.7109375,
+      "learning_rate": 2.5648516347087248e-05,
+      "loss": 0.3682,
+      "step": 11234
+    },
+    {
+      "epoch": 0.49336421616518655,
+      "grad_norm": 1.96875,
+      "learning_rate": 2.564160036669668e-05,
+      "loss": 0.3406,
+      "step": 11236
+    },
+    {
+      "epoch": 0.49345203464439,
+      "grad_norm": 1.90625,
+      "learning_rate": 2.5634684337172205e-05,
+      "loss": 0.3647,
+      "step": 11238
+    },
+    {
+      "epoch": 0.49353985312359355,
+      "grad_norm": 1.8046875,
+      "learning_rate": 2.562776825904345e-05,
+      "loss": 0.3419,
+      "step": 11240
+    },
+    {
+      "epoch": 0.493627671602797,
+      "grad_norm": 1.828125,
+      "learning_rate": 2.562085213284005e-05,
+      "loss": 0.3472,
+      "step": 11242
+    },
+    {
+      "epoch": 0.4937154900820005,
+      "grad_norm": 1.8125,
+      "learning_rate": 2.5613935959091647e-05,
+      "loss": 0.3834,
+      "step": 11244
+    },
+    {
+      "epoch": 0.49380330856120397,
+      "grad_norm": 1.9609375,
+      "learning_rate": 2.5607019738327882e-05,
+      "loss": 0.3647,
+      "step": 11246
+    },
+    {
+      "epoch": 0.4938911270404075,
+      "grad_norm": 1.78125,
+      "learning_rate": 2.5600103471078397e-05,
+      "loss": 0.3764,
+      "step": 11248
+    },
+    {
+      "epoch": 0.49397894551961097,
+      "grad_norm": 1.8203125,
+      "learning_rate": 2.5593187157872845e-05,
+      "loss": 0.364,
+      "step": 11250
+    },
+    {
+      "epoch": 0.49406676399881444,
+      "grad_norm": 1.6953125,
+      "learning_rate": 2.5586270799240876e-05,
+      "loss": 0.3456,
+      "step": 11252
+    },
+    {
+      "epoch": 0.4941545824780179,
+      "grad_norm": 1.6640625,
+      "learning_rate": 2.557935439571215e-05,
+      "loss": 0.3279,
+      "step": 11254
+    },
+    {
+      "epoch": 0.49424240095722144,
+      "grad_norm": 1.6875,
+      "learning_rate": 2.557243794781633e-05,
+      "loss": 0.3656,
+      "step": 11256
+    },
+    {
+      "epoch": 0.4943302194364249,
+      "grad_norm": 1.6953125,
+      "learning_rate": 2.556552145608307e-05,
+      "loss": 0.3393,
+      "step": 11258
+    },
+    {
+      "epoch": 0.4944180379156284,
+      "grad_norm": 1.6796875,
+      "learning_rate": 2.5558604921042045e-05,
+      "loss": 0.3982,
+      "step": 11260
+    },
+    {
+      "epoch": 0.49450585639483186,
+      "grad_norm": 1.703125,
+      "learning_rate": 2.555168834322292e-05,
+      "loss": 0.3508,
+      "step": 11262
+    },
+    {
+      "epoch": 0.4945936748740354,
+      "grad_norm": 1.7890625,
+      "learning_rate": 2.5544771723155365e-05,
+      "loss": 0.3574,
+      "step": 11264
+    },
+    {
+      "epoch": 0.49468149335323885,
+      "grad_norm": 1.703125,
+      "learning_rate": 2.553785506136906e-05,
+      "loss": 0.3464,
+      "step": 11266
+    },
+    {
+      "epoch": 0.4947693118324423,
+      "grad_norm": 1.8125,
+      "learning_rate": 2.553093835839369e-05,
+      "loss": 0.3589,
+      "step": 11268
+    },
+    {
+      "epoch": 0.49485713031164585,
+      "grad_norm": 1.78125,
+      "learning_rate": 2.5524021614758934e-05,
+      "loss": 0.3419,
+      "step": 11270
+    },
+    {
+      "epoch": 0.4949449487908493,
+      "grad_norm": 1.734375,
+      "learning_rate": 2.551710483099448e-05,
+      "loss": 0.3627,
+      "step": 11272
+    },
+    {
+      "epoch": 0.4950327672700528,
+      "grad_norm": 1.6796875,
+      "learning_rate": 2.551018800763001e-05,
+      "loss": 0.3796,
+      "step": 11274
+    },
+    {
+      "epoch": 0.49512058574925627,
+      "grad_norm": 1.625,
+      "learning_rate": 2.5503271145195217e-05,
+      "loss": 0.3389,
+      "step": 11276
+    },
+    {
+      "epoch": 0.4952084042284598,
+      "grad_norm": 1.609375,
+      "learning_rate": 2.5496354244219805e-05,
+      "loss": 0.3406,
+      "step": 11278
+    },
+    {
+      "epoch": 0.49529622270766327,
+      "grad_norm": 1.734375,
+      "learning_rate": 2.548943730523346e-05,
+      "loss": 0.3669,
+      "step": 11280
+    },
+    {
+      "epoch": 0.49538404118686674,
+      "grad_norm": 1.7890625,
+      "learning_rate": 2.5482520328765898e-05,
+      "loss": 0.3812,
+      "step": 11282
+    },
+    {
+      "epoch": 0.4954718596660702,
+      "grad_norm": 1.8125,
+      "learning_rate": 2.5475603315346807e-05,
+      "loss": 0.3608,
+      "step": 11284
+    },
+    {
+      "epoch": 0.49555967814527374,
+      "grad_norm": 1.765625,
+      "learning_rate": 2.5468686265505903e-05,
+      "loss": 0.3618,
+      "step": 11286
+    },
+    {
+      "epoch": 0.4956474966244772,
+      "grad_norm": 1.7265625,
+      "learning_rate": 2.5461769179772886e-05,
+      "loss": 0.3699,
+      "step": 11288
+    },
+    {
+      "epoch": 0.4957353151036807,
+      "grad_norm": 1.84375,
+      "learning_rate": 2.545485205867748e-05,
+      "loss": 0.3482,
+      "step": 11290
+    },
+    {
+      "epoch": 0.49582313358288416,
+      "grad_norm": 1.765625,
+      "learning_rate": 2.5447934902749393e-05,
+      "loss": 0.3649,
+      "step": 11292
+    },
+    {
+      "epoch": 0.4959109520620877,
+      "grad_norm": 2.078125,
+      "learning_rate": 2.5441017712518337e-05,
+      "loss": 0.3537,
+      "step": 11294
+    },
+    {
+      "epoch": 0.49599877054129116,
+      "grad_norm": 2.0,
+      "learning_rate": 2.5434100488514053e-05,
+      "loss": 0.3684,
+      "step": 11296
+    },
+    {
+      "epoch": 0.49608658902049463,
+      "grad_norm": 1.8046875,
+      "learning_rate": 2.542718323126624e-05,
+      "loss": 0.3429,
+      "step": 11298
+    },
+    {
+      "epoch": 0.4961744074996981,
+      "grad_norm": 1.6796875,
+      "learning_rate": 2.542026594130464e-05,
+      "loss": 0.3532,
+      "step": 11300
+    },
+    {
+      "epoch": 0.4962622259789016,
+      "grad_norm": 1.796875,
+      "learning_rate": 2.5413348619158967e-05,
+      "loss": 0.3644,
+      "step": 11302
+    },
+    {
+      "epoch": 0.4963500444581051,
+      "grad_norm": 1.7734375,
+      "learning_rate": 2.5406431265358955e-05,
+      "loss": 0.3644,
+      "step": 11304
+    },
+    {
+      "epoch": 0.49643786293730857,
+      "grad_norm": 1.78125,
+      "learning_rate": 2.5399513880434345e-05,
+      "loss": 0.3695,
+      "step": 11306
+    },
+    {
+      "epoch": 0.49652568141651204,
+      "grad_norm": 1.8203125,
+      "learning_rate": 2.5392596464914864e-05,
+      "loss": 0.3724,
+      "step": 11308
+    },
+    {
+      "epoch": 0.49661349989571557,
+      "grad_norm": 1.8515625,
+      "learning_rate": 2.5385679019330257e-05,
+      "loss": 0.3812,
+      "step": 11310
+    },
+    {
+      "epoch": 0.49670131837491904,
+      "grad_norm": 2.015625,
+      "learning_rate": 2.537876154421025e-05,
+      "loss": 0.3289,
+      "step": 11312
+    },
+    {
+      "epoch": 0.4967891368541225,
+      "grad_norm": 1.828125,
+      "learning_rate": 2.537184404008459e-05,
+      "loss": 0.3673,
+      "step": 11314
+    },
+    {
+      "epoch": 0.49687695533332604,
+      "grad_norm": 1.9609375,
+      "learning_rate": 2.536492650748304e-05,
+      "loss": 0.3545,
+      "step": 11316
+    },
+    {
+      "epoch": 0.4969647738125295,
+      "grad_norm": 2.0,
+      "learning_rate": 2.535800894693532e-05,
+      "loss": 0.3674,
+      "step": 11318
+    },
+    {
+      "epoch": 0.497052592291733,
+      "grad_norm": 1.9765625,
+      "learning_rate": 2.535109135897119e-05,
+      "loss": 0.3536,
+      "step": 11320
+    },
+    {
+      "epoch": 0.49714041077093646,
+      "grad_norm": 1.7265625,
+      "learning_rate": 2.5344173744120402e-05,
+      "loss": 0.3547,
+      "step": 11322
+    },
+    {
+      "epoch": 0.49722822925014,
+      "grad_norm": 1.8359375,
+      "learning_rate": 2.53372561029127e-05,
+      "loss": 0.3864,
+      "step": 11324
+    },
+    {
+      "epoch": 0.49731604772934346,
+      "grad_norm": 1.6796875,
+      "learning_rate": 2.533033843587785e-05,
+      "loss": 0.3391,
+      "step": 11326
+    },
+    {
+      "epoch": 0.49740386620854693,
+      "grad_norm": 1.890625,
+      "learning_rate": 2.53234207435456e-05,
+      "loss": 0.374,
+      "step": 11328
+    },
+    {
+      "epoch": 0.4974916846877504,
+      "grad_norm": 1.796875,
+      "learning_rate": 2.531650302644572e-05,
+      "loss": 0.3793,
+      "step": 11330
+    },
+    {
+      "epoch": 0.49757950316695393,
+      "grad_norm": 1.7890625,
+      "learning_rate": 2.530958528510795e-05,
+      "loss": 0.3879,
+      "step": 11332
+    },
+    {
+      "epoch": 0.4976673216461574,
+      "grad_norm": 1.90625,
+      "learning_rate": 2.530266752006208e-05,
+      "loss": 0.3623,
+      "step": 11334
+    },
+    {
+      "epoch": 0.4977551401253609,
+      "grad_norm": 1.9296875,
+      "learning_rate": 2.529574973183785e-05,
+      "loss": 0.3668,
+      "step": 11336
+    },
+    {
+      "epoch": 0.49784295860456435,
+      "grad_norm": 2.21875,
+      "learning_rate": 2.5288831920965028e-05,
+      "loss": 0.3766,
+      "step": 11338
+    },
+    {
+      "epoch": 0.4979307770837679,
+      "grad_norm": 2.171875,
+      "learning_rate": 2.5281914087973395e-05,
+      "loss": 0.375,
+      "step": 11340
+    },
+    {
+      "epoch": 0.49801859556297134,
+      "grad_norm": 1.6484375,
+      "learning_rate": 2.5274996233392712e-05,
+      "loss": 0.3424,
+      "step": 11342
+    },
+    {
+      "epoch": 0.4981064140421748,
+      "grad_norm": 1.8359375,
+      "learning_rate": 2.5268078357752757e-05,
+      "loss": 0.3336,
+      "step": 11344
+    },
+    {
+      "epoch": 0.4981942325213783,
+      "grad_norm": 1.984375,
+      "learning_rate": 2.52611604615833e-05,
+      "loss": 0.3356,
+      "step": 11346
+    },
+    {
+      "epoch": 0.4982820510005818,
+      "grad_norm": 1.8046875,
+      "learning_rate": 2.5254242545414104e-05,
+      "loss": 0.3336,
+      "step": 11348
+    },
+    {
+      "epoch": 0.4983698694797853,
+      "grad_norm": 2.171875,
+      "learning_rate": 2.5247324609774953e-05,
+      "loss": 0.3769,
+      "step": 11350
+    },
+    {
+      "epoch": 0.49845768795898876,
+      "grad_norm": 1.9375,
+      "learning_rate": 2.524040665519562e-05,
+      "loss": 0.3574,
+      "step": 11352
+    },
+    {
+      "epoch": 0.49854550643819223,
+      "grad_norm": 1.734375,
+      "learning_rate": 2.523348868220589e-05,
+      "loss": 0.3693,
+      "step": 11354
+    },
+    {
+      "epoch": 0.49863332491739576,
+      "grad_norm": 1.8671875,
+      "learning_rate": 2.522657069133555e-05,
+      "loss": 0.3505,
+      "step": 11356
+    },
+    {
+      "epoch": 0.49872114339659923,
+      "grad_norm": 1.9453125,
+      "learning_rate": 2.5219652683114363e-05,
+      "loss": 0.3886,
+      "step": 11358
+    },
+    {
+      "epoch": 0.4988089618758027,
+      "grad_norm": 2.234375,
+      "learning_rate": 2.5212734658072124e-05,
+      "loss": 0.384,
+      "step": 11360
+    },
+    {
+      "epoch": 0.4988967803550062,
+      "grad_norm": 1.7734375,
+      "learning_rate": 2.520581661673862e-05,
+      "loss": 0.3613,
+      "step": 11362
+    },
+    {
+      "epoch": 0.4989845988342097,
+      "grad_norm": 1.8515625,
+      "learning_rate": 2.5198898559643618e-05,
+      "loss": 0.3673,
+      "step": 11364
+    },
+    {
+      "epoch": 0.4990724173134132,
+      "grad_norm": 1.953125,
+      "learning_rate": 2.5191980487316924e-05,
+      "loss": 0.353,
+      "step": 11366
+    },
+    {
+      "epoch": 0.49916023579261665,
+      "grad_norm": 1.9609375,
+      "learning_rate": 2.5185062400288316e-05,
+      "loss": 0.3536,
+      "step": 11368
+    },
+    {
+      "epoch": 0.4992480542718202,
+      "grad_norm": 1.8359375,
+      "learning_rate": 2.517814429908759e-05,
+      "loss": 0.35,
+      "step": 11370
+    },
+    {
+      "epoch": 0.49933587275102365,
+      "grad_norm": 1.8125,
+      "learning_rate": 2.5171226184244525e-05,
+      "loss": 0.3362,
+      "step": 11372
+    },
+    {
+      "epoch": 0.4994236912302271,
+      "grad_norm": 1.7734375,
+      "learning_rate": 2.5164308056288915e-05,
+      "loss": 0.3457,
+      "step": 11374
+    },
+    {
+      "epoch": 0.4995115097094306,
+      "grad_norm": 1.828125,
+      "learning_rate": 2.515738991575057e-05,
+      "loss": 0.3596,
+      "step": 11376
+    },
+    {
+      "epoch": 0.4995993281886341,
+      "grad_norm": 2.078125,
+      "learning_rate": 2.515047176315925e-05,
+      "loss": 0.3743,
+      "step": 11378
+    },
+    {
+      "epoch": 0.4996871466678376,
+      "grad_norm": 2.140625,
+      "learning_rate": 2.5143553599044773e-05,
+      "loss": 0.3283,
+      "step": 11380
+    },
+    {
+      "epoch": 0.49977496514704106,
+      "grad_norm": 1.8203125,
+      "learning_rate": 2.513663542393693e-05,
+      "loss": 0.3711,
+      "step": 11382
+    },
+    {
+      "epoch": 0.49986278362624453,
+      "grad_norm": 2.296875,
+      "learning_rate": 2.5129717238365508e-05,
+      "loss": 0.3386,
+      "step": 11384
+    },
+    {
+      "epoch": 0.49995060210544806,
+      "grad_norm": 1.8125,
+      "learning_rate": 2.512279904286031e-05,
+      "loss": 0.3563,
+      "step": 11386
+    },
+    {
+      "epoch": 0.5000384205846515,
+      "grad_norm": 1.890625,
+      "learning_rate": 2.5115880837951134e-05,
+      "loss": 0.3429,
+      "step": 11388
+    },
+    {
+      "epoch": 0.500126239063855,
+      "grad_norm": 1.84375,
+      "learning_rate": 2.510896262416777e-05,
+      "loss": 0.3711,
+      "step": 11390
+    },
+    {
+      "epoch": 0.5002140575430585,
+      "grad_norm": 1.671875,
+      "learning_rate": 2.5102044402040033e-05,
+      "loss": 0.3793,
+      "step": 11392
+    },
+    {
+      "epoch": 0.500301876022262,
+      "grad_norm": 1.9296875,
+      "learning_rate": 2.5095126172097705e-05,
+      "loss": 0.3527,
+      "step": 11394
+    },
+    {
+      "epoch": 0.5003896945014654,
+      "grad_norm": 1.6953125,
+      "learning_rate": 2.5088207934870593e-05,
+      "loss": 0.3562,
+      "step": 11396
+    },
+    {
+      "epoch": 0.500477512980669,
+      "grad_norm": 1.9609375,
+      "learning_rate": 2.5081289690888495e-05,
+      "loss": 0.3647,
+      "step": 11398
+    },
+    {
+      "epoch": 0.5005653314598725,
+      "grad_norm": 1.734375,
+      "learning_rate": 2.507437144068121e-05,
+      "loss": 0.344,
+      "step": 11400
+    },
+    {
+      "epoch": 0.500653149939076,
+      "grad_norm": 1.8515625,
+      "learning_rate": 2.5067453184778545e-05,
+      "loss": 0.3791,
+      "step": 11402
+    },
+    {
+      "epoch": 0.5007409684182794,
+      "grad_norm": 1.953125,
+      "learning_rate": 2.5060534923710298e-05,
+      "loss": 0.3648,
+      "step": 11404
+    },
+    {
+      "epoch": 0.5008287868974829,
+      "grad_norm": 1.703125,
+      "learning_rate": 2.505361665800628e-05,
+      "loss": 0.3255,
+      "step": 11406
+    },
+    {
+      "epoch": 0.5009166053766864,
+      "grad_norm": 1.75,
+      "learning_rate": 2.5046698388196284e-05,
+      "loss": 0.3838,
+      "step": 11408
+    },
+    {
+      "epoch": 0.5010044238558898,
+      "grad_norm": 1.6796875,
+      "learning_rate": 2.50397801148101e-05,
+      "loss": 0.3467,
+      "step": 11410
+    },
+    {
+      "epoch": 0.5010922423350933,
+      "grad_norm": 1.8984375,
+      "learning_rate": 2.5032861838377557e-05,
+      "loss": 0.3475,
+      "step": 11412
+    },
+    {
+      "epoch": 0.5011800608142969,
+      "grad_norm": 1.734375,
+      "learning_rate": 2.5025943559428444e-05,
+      "loss": 0.3663,
+      "step": 11414
+    },
+    {
+      "epoch": 0.5012678792935004,
+      "grad_norm": 1.7265625,
+      "learning_rate": 2.5019025278492565e-05,
+      "loss": 0.3312,
+      "step": 11416
+    },
+    {
+      "epoch": 0.5013556977727038,
+      "grad_norm": 1.7265625,
+      "learning_rate": 2.501210699609973e-05,
+      "loss": 0.3312,
+      "step": 11418
+    },
+    {
+      "epoch": 0.5014435162519073,
+      "grad_norm": 1.90625,
+      "learning_rate": 2.5005188712779736e-05,
+      "loss": 0.3418,
+      "step": 11420
+    },
+    {
+      "epoch": 0.5015313347311108,
+      "grad_norm": 1.796875,
+      "learning_rate": 2.4998270429062394e-05,
+      "loss": 0.3551,
+      "step": 11422
+    },
+    {
+      "epoch": 0.5016191532103143,
+      "grad_norm": 1.7265625,
+      "learning_rate": 2.4991352145477497e-05,
+      "loss": 0.3624,
+      "step": 11424
+    },
+    {
+      "epoch": 0.5017069716895177,
+      "grad_norm": 1.8671875,
+      "learning_rate": 2.498443386255485e-05,
+      "loss": 0.3742,
+      "step": 11426
+    },
+    {
+      "epoch": 0.5017947901687213,
+      "grad_norm": 2.015625,
+      "learning_rate": 2.497751558082427e-05,
+      "loss": 0.3658,
+      "step": 11428
+    },
+    {
+      "epoch": 0.5018826086479248,
+      "grad_norm": 1.7734375,
+      "learning_rate": 2.4970597300815545e-05,
+      "loss": 0.3503,
+      "step": 11430
+    },
+    {
+      "epoch": 0.5019704271271282,
+      "grad_norm": 1.6875,
+      "learning_rate": 2.4963679023058495e-05,
+      "loss": 0.3415,
+      "step": 11432
+    },
+    {
+      "epoch": 0.5020582456063317,
+      "grad_norm": 1.9140625,
+      "learning_rate": 2.495676074808291e-05,
+      "loss": 0.3646,
+      "step": 11434
+    },
+    {
+      "epoch": 0.5021460640855352,
+      "grad_norm": 2.0625,
+      "learning_rate": 2.4949842476418592e-05,
+      "loss": 0.3604,
+      "step": 11436
+    },
+    {
+      "epoch": 0.5022338825647387,
+      "grad_norm": 1.8359375,
+      "learning_rate": 2.4942924208595352e-05,
+      "loss": 0.371,
+      "step": 11438
+    },
+    {
+      "epoch": 0.5023217010439421,
+      "grad_norm": 1.859375,
+      "learning_rate": 2.493600594514299e-05,
+      "loss": 0.3412,
+      "step": 11440
+    },
+    {
+      "epoch": 0.5024095195231456,
+      "grad_norm": 1.7734375,
+      "learning_rate": 2.4929087686591317e-05,
+      "loss": 0.344,
+      "step": 11442
+    },
+    {
+      "epoch": 0.5024973380023492,
+      "grad_norm": 1.6875,
+      "learning_rate": 2.4922169433470112e-05,
+      "loss": 0.3685,
+      "step": 11444
+    },
+    {
+      "epoch": 0.5025851564815527,
+      "grad_norm": 1.9765625,
+      "learning_rate": 2.491525118630921e-05,
+      "loss": 0.3741,
+      "step": 11446
+    },
+    {
+      "epoch": 0.5026729749607561,
+      "grad_norm": 2.125,
+      "learning_rate": 2.4908332945638378e-05,
+      "loss": 0.3448,
+      "step": 11448
+    },
+    {
+      "epoch": 0.5027607934399596,
+      "grad_norm": 1.8046875,
+      "learning_rate": 2.4901414711987432e-05,
+      "loss": 0.3645,
+      "step": 11450
+    },
+    {
+      "epoch": 0.5028486119191631,
+      "grad_norm": 1.78125,
+      "learning_rate": 2.489449648588617e-05,
+      "loss": 0.3583,
+      "step": 11452
+    },
+    {
+      "epoch": 0.5029364303983666,
+      "grad_norm": 1.625,
+      "learning_rate": 2.4887578267864393e-05,
+      "loss": 0.341,
+      "step": 11454
+    },
+    {
+      "epoch": 0.50302424887757,
+      "grad_norm": 1.703125,
+      "learning_rate": 2.4880660058451895e-05,
+      "loss": 0.3411,
+      "step": 11456
+    },
+    {
+      "epoch": 0.5031120673567735,
+      "grad_norm": 1.828125,
+      "learning_rate": 2.487374185817849e-05,
+      "loss": 0.3633,
+      "step": 11458
+    },
+    {
+      "epoch": 0.5031998858359771,
+      "grad_norm": 1.8203125,
+      "learning_rate": 2.4866823667573954e-05,
+      "loss": 0.3506,
+      "step": 11460
+    },
+    {
+      "epoch": 0.5032877043151806,
+      "grad_norm": 2.0,
+      "learning_rate": 2.4859905487168095e-05,
+      "loss": 0.3882,
+      "step": 11462
+    },
+    {
+      "epoch": 0.503375522794384,
+      "grad_norm": 1.78125,
+      "learning_rate": 2.485298731749071e-05,
+      "loss": 0.3581,
+      "step": 11464
+    },
+    {
+      "epoch": 0.5034633412735875,
+      "grad_norm": 1.671875,
+      "learning_rate": 2.4846069159071586e-05,
+      "loss": 0.3581,
+      "step": 11466
+    },
+    {
+      "epoch": 0.503551159752791,
+      "grad_norm": 2.328125,
+      "learning_rate": 2.4839151012440532e-05,
+      "loss": 0.356,
+      "step": 11468
+    },
+    {
+      "epoch": 0.5036389782319944,
+      "grad_norm": 1.734375,
+      "learning_rate": 2.4832232878127327e-05,
+      "loss": 0.3814,
+      "step": 11470
+    },
+    {
+      "epoch": 0.5037267967111979,
+      "grad_norm": 1.9765625,
+      "learning_rate": 2.4825314756661765e-05,
+      "loss": 0.3626,
+      "step": 11472
+    },
+    {
+      "epoch": 0.5038146151904015,
+      "grad_norm": 1.6796875,
+      "learning_rate": 2.4818396648573645e-05,
+      "loss": 0.3471,
+      "step": 11474
+    },
+    {
+      "epoch": 0.503902433669605,
+      "grad_norm": 1.765625,
+      "learning_rate": 2.4811478554392762e-05,
+      "loss": 0.3433,
+      "step": 11476
+    },
+    {
+      "epoch": 0.5039902521488084,
+      "grad_norm": 1.6171875,
+      "learning_rate": 2.4804560474648885e-05,
+      "loss": 0.3593,
+      "step": 11478
+    },
+    {
+      "epoch": 0.5040780706280119,
+      "grad_norm": 1.6875,
+      "learning_rate": 2.4797642409871816e-05,
+      "loss": 0.3472,
+      "step": 11480
+    },
+    {
+      "epoch": 0.5041658891072154,
+      "grad_norm": 1.921875,
+      "learning_rate": 2.479072436059134e-05,
+      "loss": 0.3718,
+      "step": 11482
+    },
+    {
+      "epoch": 0.5042537075864189,
+      "grad_norm": 1.6953125,
+      "learning_rate": 2.4783806327337242e-05,
+      "loss": 0.3797,
+      "step": 11484
+    },
+    {
+      "epoch": 0.5043415260656223,
+      "grad_norm": 1.78125,
+      "learning_rate": 2.477688831063931e-05,
+      "loss": 0.3475,
+      "step": 11486
+    },
+    {
+      "epoch": 0.5044293445448258,
+      "grad_norm": 1.7265625,
+      "learning_rate": 2.476997031102732e-05,
+      "loss": 0.4019,
+      "step": 11488
+    },
+    {
+      "epoch": 0.5045171630240294,
+      "grad_norm": 1.7265625,
+      "learning_rate": 2.4763052329031055e-05,
+      "loss": 0.3631,
+      "step": 11490
+    },
+    {
+      "epoch": 0.5046049815032329,
+      "grad_norm": 1.796875,
+      "learning_rate": 2.4756134365180306e-05,
+      "loss": 0.3593,
+      "step": 11492
+    },
+    {
+      "epoch": 0.5046927999824363,
+      "grad_norm": 1.7734375,
+      "learning_rate": 2.4749216420004852e-05,
+      "loss": 0.357,
+      "step": 11494
+    },
+    {
+      "epoch": 0.5047806184616398,
+      "grad_norm": 1.7265625,
+      "learning_rate": 2.4742298494034453e-05,
+      "loss": 0.3796,
+      "step": 11496
+    },
+    {
+      "epoch": 0.5048684369408433,
+      "grad_norm": 1.7421875,
+      "learning_rate": 2.4735380587798902e-05,
+      "loss": 0.354,
+      "step": 11498
+    },
+    {
+      "epoch": 0.5049562554200467,
+      "grad_norm": 1.734375,
+      "learning_rate": 2.4728462701827968e-05,
+      "loss": 0.3939,
+      "step": 11500
+    },
+    {
+      "epoch": 0.5050440738992502,
+      "grad_norm": 1.8515625,
+      "learning_rate": 2.472154483665142e-05,
+      "loss": 0.376,
+      "step": 11502
+    },
+    {
+      "epoch": 0.5051318923784537,
+      "grad_norm": 1.78125,
+      "learning_rate": 2.4714626992799047e-05,
+      "loss": 0.3363,
+      "step": 11504
+    },
+    {
+      "epoch": 0.5052197108576573,
+      "grad_norm": 2.28125,
+      "learning_rate": 2.4707709170800608e-05,
+      "loss": 0.3781,
+      "step": 11506
+    },
+    {
+      "epoch": 0.5053075293368607,
+      "grad_norm": 1.75,
+      "learning_rate": 2.4700791371185868e-05,
+      "loss": 0.3709,
+      "step": 11508
+    },
+    {
+      "epoch": 0.5053953478160642,
+      "grad_norm": 1.8125,
+      "learning_rate": 2.469387359448459e-05,
+      "loss": 0.345,
+      "step": 11510
+    },
+    {
+      "epoch": 0.5054831662952677,
+      "grad_norm": 1.9296875,
+      "learning_rate": 2.4686955841226546e-05,
+      "loss": 0.3357,
+      "step": 11512
+    },
+    {
+      "epoch": 0.5055709847744712,
+      "grad_norm": 1.921875,
+      "learning_rate": 2.4680038111941504e-05,
+      "loss": 0.3312,
+      "step": 11514
+    },
+    {
+      "epoch": 0.5056588032536746,
+      "grad_norm": 2.109375,
+      "learning_rate": 2.4673120407159214e-05,
+      "loss": 0.3445,
+      "step": 11516
+    },
+    {
+      "epoch": 0.5057466217328781,
+      "grad_norm": 1.828125,
+      "learning_rate": 2.466620272740945e-05,
+      "loss": 0.3678,
+      "step": 11518
+    },
+    {
+      "epoch": 0.5058344402120817,
+      "grad_norm": 1.6796875,
+      "learning_rate": 2.4659285073221955e-05,
+      "loss": 0.3549,
+      "step": 11520
+    },
+    {
+      "epoch": 0.5059222586912852,
+      "grad_norm": 2.21875,
+      "learning_rate": 2.465236744512649e-05,
+      "loss": 0.3821,
+      "step": 11522
+    },
+    {
+      "epoch": 0.5060100771704886,
+      "grad_norm": 2.15625,
+      "learning_rate": 2.464544984365281e-05,
+      "loss": 0.358,
+      "step": 11524
+    },
+    {
+      "epoch": 0.5060978956496921,
+      "grad_norm": 1.953125,
+      "learning_rate": 2.4638532269330665e-05,
+      "loss": 0.3459,
+      "step": 11526
+    },
+    {
+      "epoch": 0.5061857141288956,
+      "grad_norm": 1.8671875,
+      "learning_rate": 2.4631614722689806e-05,
+      "loss": 0.3657,
+      "step": 11528
+    },
+    {
+      "epoch": 0.506273532608099,
+      "grad_norm": 1.7265625,
+      "learning_rate": 2.4624697204259986e-05,
+      "loss": 0.3525,
+      "step": 11530
+    },
+    {
+      "epoch": 0.5063613510873025,
+      "grad_norm": 1.90625,
+      "learning_rate": 2.461777971457094e-05,
+      "loss": 0.3466,
+      "step": 11532
+    },
+    {
+      "epoch": 0.506449169566506,
+      "grad_norm": 2.0,
+      "learning_rate": 2.461086225415241e-05,
+      "loss": 0.3451,
+      "step": 11534
+    },
+    {
+      "epoch": 0.5065369880457096,
+      "grad_norm": 1.6875,
+      "learning_rate": 2.4603944823534146e-05,
+      "loss": 0.3576,
+      "step": 11536
+    },
+    {
+      "epoch": 0.506624806524913,
+      "grad_norm": 1.6640625,
+      "learning_rate": 2.4597027423245893e-05,
+      "loss": 0.3392,
+      "step": 11538
+    },
+    {
+      "epoch": 0.5067126250041165,
+      "grad_norm": 1.7265625,
+      "learning_rate": 2.4590110053817364e-05,
+      "loss": 0.3546,
+      "step": 11540
+    },
+    {
+      "epoch": 0.50680044348332,
+      "grad_norm": 1.921875,
+      "learning_rate": 2.4583192715778304e-05,
+      "loss": 0.3363,
+      "step": 11542
+    },
+    {
+      "epoch": 0.5068882619625235,
+      "grad_norm": 1.7890625,
+      "learning_rate": 2.4576275409658454e-05,
+      "loss": 0.359,
+      "step": 11544
+    },
+    {
+      "epoch": 0.5069760804417269,
+      "grad_norm": 1.75,
+      "learning_rate": 2.456935813598753e-05,
+      "loss": 0.3529,
+      "step": 11546
+    },
+    {
+      "epoch": 0.5070638989209304,
+      "grad_norm": 1.890625,
+      "learning_rate": 2.4562440895295265e-05,
+      "loss": 0.3435,
+      "step": 11548
+    },
+    {
+      "epoch": 0.5071517174001339,
+      "grad_norm": 1.8515625,
+      "learning_rate": 2.4555523688111377e-05,
+      "loss": 0.3664,
+      "step": 11550
+    },
+    {
+      "epoch": 0.5072395358793375,
+      "grad_norm": 1.9453125,
+      "learning_rate": 2.454860651496559e-05,
+      "loss": 0.3318,
+      "step": 11552
+    },
+    {
+      "epoch": 0.5073273543585409,
+      "grad_norm": 1.71875,
+      "learning_rate": 2.4541689376387627e-05,
+      "loss": 0.3551,
+      "step": 11554
+    },
+    {
+      "epoch": 0.5074151728377444,
+      "grad_norm": 1.796875,
+      "learning_rate": 2.453477227290721e-05,
+      "loss": 0.3396,
+      "step": 11556
+    },
+    {
+      "epoch": 0.5075029913169479,
+      "grad_norm": 1.8125,
+      "learning_rate": 2.452785520505403e-05,
+      "loss": 0.3486,
+      "step": 11558
+    },
+    {
+      "epoch": 0.5075908097961513,
+      "grad_norm": 1.8828125,
+      "learning_rate": 2.452093817335782e-05,
+      "loss": 0.357,
+      "step": 11560
+    },
+    {
+      "epoch": 0.5076786282753548,
+      "grad_norm": 1.6640625,
+      "learning_rate": 2.4514021178348276e-05,
+      "loss": 0.356,
+      "step": 11562
+    },
+    {
+      "epoch": 0.5077664467545583,
+      "grad_norm": 1.7265625,
+      "learning_rate": 2.4507104220555106e-05,
+      "loss": 0.3339,
+      "step": 11564
+    },
+    {
+      "epoch": 0.5078542652337618,
+      "grad_norm": 2.015625,
+      "learning_rate": 2.450018730050802e-05,
+      "loss": 0.3823,
+      "step": 11566
+    },
+    {
+      "epoch": 0.5079420837129653,
+      "grad_norm": 1.59375,
+      "learning_rate": 2.4493270418736707e-05,
+      "loss": 0.3737,
+      "step": 11568
+    },
+    {
+      "epoch": 0.5080299021921688,
+      "grad_norm": 1.953125,
+      "learning_rate": 2.448635357577087e-05,
+      "loss": 0.3471,
+      "step": 11570
+    },
+    {
+      "epoch": 0.5081177206713723,
+      "grad_norm": 2.0,
+      "learning_rate": 2.4479436772140195e-05,
+      "loss": 0.3523,
+      "step": 11572
+    },
+    {
+      "epoch": 0.5082055391505758,
+      "grad_norm": 2.015625,
+      "learning_rate": 2.4472520008374375e-05,
+      "loss": 0.3427,
+      "step": 11574
+    },
+    {
+      "epoch": 0.5082933576297792,
+      "grad_norm": 1.875,
+      "learning_rate": 2.44656032850031e-05,
+      "loss": 0.34,
+      "step": 11576
+    },
+    {
+      "epoch": 0.5083811761089827,
+      "grad_norm": 1.9453125,
+      "learning_rate": 2.4458686602556052e-05,
+      "loss": 0.3527,
+      "step": 11578
+    },
+    {
+      "epoch": 0.5084689945881862,
+      "grad_norm": 1.734375,
+      "learning_rate": 2.445176996156292e-05,
+      "loss": 0.3461,
+      "step": 11580
+    },
+    {
+      "epoch": 0.5085568130673898,
+      "grad_norm": 1.9296875,
+      "learning_rate": 2.444485336255337e-05,
+      "loss": 0.3473,
+      "step": 11582
+    },
+    {
+      "epoch": 0.5086446315465932,
+      "grad_norm": 1.703125,
+      "learning_rate": 2.4437936806057082e-05,
+      "loss": 0.3454,
+      "step": 11584
+    },
+    {
+      "epoch": 0.5087324500257967,
+      "grad_norm": 1.8828125,
+      "learning_rate": 2.443102029260373e-05,
+      "loss": 0.371,
+      "step": 11586
+    },
+    {
+      "epoch": 0.5088202685050002,
+      "grad_norm": 1.7890625,
+      "learning_rate": 2.442410382272298e-05,
+      "loss": 0.3645,
+      "step": 11588
+    },
+    {
+      "epoch": 0.5089080869842036,
+      "grad_norm": 1.859375,
+      "learning_rate": 2.4417187396944496e-05,
+      "loss": 0.3277,
+      "step": 11590
+    },
+    {
+      "epoch": 0.5089959054634071,
+      "grad_norm": 1.8359375,
+      "learning_rate": 2.4410271015797947e-05,
+      "loss": 0.3338,
+      "step": 11592
+    },
+    {
+      "epoch": 0.5090837239426106,
+      "grad_norm": 1.828125,
+      "learning_rate": 2.440335467981298e-05,
+      "loss": 0.3514,
+      "step": 11594
+    },
+    {
+      "epoch": 0.5091715424218141,
+      "grad_norm": 1.8125,
+      "learning_rate": 2.439643838951925e-05,
+      "loss": 0.365,
+      "step": 11596
+    },
+    {
+      "epoch": 0.5092593609010176,
+      "grad_norm": 1.828125,
+      "learning_rate": 2.438952214544643e-05,
+      "loss": 0.3909,
+      "step": 11598
+    },
+    {
+      "epoch": 0.5093471793802211,
+      "grad_norm": 1.9609375,
+      "learning_rate": 2.4382605948124137e-05,
+      "loss": 0.3652,
+      "step": 11600
+    },
+    {
+      "epoch": 0.5094349978594246,
+      "grad_norm": 1.75,
+      "learning_rate": 2.437568979808203e-05,
+      "loss": 0.3538,
+      "step": 11602
+    },
+    {
+      "epoch": 0.5095228163386281,
+      "grad_norm": 1.7578125,
+      "learning_rate": 2.436877369584975e-05,
+      "loss": 0.3569,
+      "step": 11604
+    },
+    {
+      "epoch": 0.5096106348178315,
+      "grad_norm": 1.890625,
+      "learning_rate": 2.436185764195693e-05,
+      "loss": 0.3627,
+      "step": 11606
+    },
+    {
+      "epoch": 0.509698453297035,
+      "grad_norm": 1.7890625,
+      "learning_rate": 2.4354941636933203e-05,
+      "loss": 0.3579,
+      "step": 11608
+    },
+    {
+      "epoch": 0.5097862717762385,
+      "grad_norm": 1.9765625,
+      "learning_rate": 2.43480256813082e-05,
+      "loss": 0.3583,
+      "step": 11610
+    },
+    {
+      "epoch": 0.509874090255442,
+      "grad_norm": 1.7734375,
+      "learning_rate": 2.4341109775611547e-05,
+      "loss": 0.3649,
+      "step": 11612
+    },
+    {
+      "epoch": 0.5099619087346455,
+      "grad_norm": 1.84375,
+      "learning_rate": 2.4334193920372867e-05,
+      "loss": 0.3639,
+      "step": 11614
+    },
+    {
+      "epoch": 0.510049727213849,
+      "grad_norm": 1.6875,
+      "learning_rate": 2.432727811612178e-05,
+      "loss": 0.3859,
+      "step": 11616
+    },
+    {
+      "epoch": 0.5101375456930525,
+      "grad_norm": 1.7578125,
+      "learning_rate": 2.4320362363387893e-05,
+      "loss": 0.3314,
+      "step": 11618
+    },
+    {
+      "epoch": 0.510225364172256,
+      "grad_norm": 1.734375,
+      "learning_rate": 2.4313446662700813e-05,
+      "loss": 0.3727,
+      "step": 11620
+    },
+    {
+      "epoch": 0.5103131826514594,
+      "grad_norm": 1.8125,
+      "learning_rate": 2.4306531014590156e-05,
+      "loss": 0.3434,
+      "step": 11622
+    },
+    {
+      "epoch": 0.5104010011306629,
+      "grad_norm": 1.75,
+      "learning_rate": 2.429961541958552e-05,
+      "loss": 0.3612,
+      "step": 11624
+    },
+    {
+      "epoch": 0.5104888196098664,
+      "grad_norm": 1.65625,
+      "learning_rate": 2.4292699878216505e-05,
+      "loss": 0.3464,
+      "step": 11626
+    },
+    {
+      "epoch": 0.51057663808907,
+      "grad_norm": 1.75,
+      "learning_rate": 2.4285784391012706e-05,
+      "loss": 0.337,
+      "step": 11628
+    },
+    {
+      "epoch": 0.5106644565682734,
+      "grad_norm": 1.6875,
+      "learning_rate": 2.4278868958503708e-05,
+      "loss": 0.3909,
+      "step": 11630
+    },
+    {
+      "epoch": 0.5107522750474769,
+      "grad_norm": 1.828125,
+      "learning_rate": 2.42719535812191e-05,
+      "loss": 0.3323,
+      "step": 11632
+    },
+    {
+      "epoch": 0.5108400935266804,
+      "grad_norm": 1.6875,
+      "learning_rate": 2.4265038259688456e-05,
+      "loss": 0.3545,
+      "step": 11634
+    },
+    {
+      "epoch": 0.5109279120058838,
+      "grad_norm": 1.859375,
+      "learning_rate": 2.425812299444136e-05,
+      "loss": 0.3492,
+      "step": 11636
+    },
+    {
+      "epoch": 0.5110157304850873,
+      "grad_norm": 1.6875,
+      "learning_rate": 2.425120778600738e-05,
+      "loss": 0.3437,
+      "step": 11638
+    },
+    {
+      "epoch": 0.5111035489642908,
+      "grad_norm": 1.8515625,
+      "learning_rate": 2.424429263491609e-05,
+      "loss": 0.3651,
+      "step": 11640
+    },
+    {
+      "epoch": 0.5111913674434942,
+      "grad_norm": 1.6015625,
+      "learning_rate": 2.4237377541697055e-05,
+      "loss": 0.3285,
+      "step": 11642
+    },
+    {
+      "epoch": 0.5112791859226978,
+      "grad_norm": 1.78125,
+      "learning_rate": 2.4230462506879824e-05,
+      "loss": 0.338,
+      "step": 11644
+    },
+    {
+      "epoch": 0.5113670044019013,
+      "grad_norm": 1.9453125,
+      "learning_rate": 2.4223547530993958e-05,
+      "loss": 0.3836,
+      "step": 11646
+    },
+    {
+      "epoch": 0.5114548228811048,
+      "grad_norm": 1.875,
+      "learning_rate": 2.4216632614569012e-05,
+      "loss": 0.382,
+      "step": 11648
+    },
+    {
+      "epoch": 0.5115426413603082,
+      "grad_norm": 1.734375,
+      "learning_rate": 2.4209717758134525e-05,
+      "loss": 0.3605,
+      "step": 11650
+    },
+    {
+      "epoch": 0.5116304598395117,
+      "grad_norm": 1.7265625,
+      "learning_rate": 2.4202802962220047e-05,
+      "loss": 0.3905,
+      "step": 11652
+    },
+    {
+      "epoch": 0.5117182783187152,
+      "grad_norm": 1.9140625,
+      "learning_rate": 2.4195888227355105e-05,
+      "loss": 0.329,
+      "step": 11654
+    },
+    {
+      "epoch": 0.5118060967979187,
+      "grad_norm": 1.75,
+      "learning_rate": 2.4188973554069236e-05,
+      "loss": 0.3697,
+      "step": 11656
+    },
+    {
+      "epoch": 0.5118939152771221,
+      "grad_norm": 1.7421875,
+      "learning_rate": 2.4182058942891966e-05,
+      "loss": 0.3456,
+      "step": 11658
+    },
+    {
+      "epoch": 0.5119817337563257,
+      "grad_norm": 1.828125,
+      "learning_rate": 2.417514439435283e-05,
+      "loss": 0.3294,
+      "step": 11660
+    },
+    {
+      "epoch": 0.5120695522355292,
+      "grad_norm": 1.7109375,
+      "learning_rate": 2.416822990898132e-05,
+      "loss": 0.3599,
+      "step": 11662
+    },
+    {
+      "epoch": 0.5121573707147327,
+      "grad_norm": 1.8671875,
+      "learning_rate": 2.4161315487306965e-05,
+      "loss": 0.3617,
+      "step": 11664
+    },
+    {
+      "epoch": 0.5122451891939361,
+      "grad_norm": 1.765625,
+      "learning_rate": 2.415440112985927e-05,
+      "loss": 0.3578,
+      "step": 11666
+    },
+    {
+      "epoch": 0.5123330076731396,
+      "grad_norm": 1.7734375,
+      "learning_rate": 2.4147486837167748e-05,
+      "loss": 0.3638,
+      "step": 11668
+    },
+    {
+      "epoch": 0.5124208261523431,
+      "grad_norm": 1.9140625,
+      "learning_rate": 2.414057260976188e-05,
+      "loss": 0.3481,
+      "step": 11670
+    },
+    {
+      "epoch": 0.5125086446315466,
+      "grad_norm": 1.734375,
+      "learning_rate": 2.413365844817117e-05,
+      "loss": 0.3752,
+      "step": 11672
+    },
+    {
+      "epoch": 0.5125964631107501,
+      "grad_norm": 1.8203125,
+      "learning_rate": 2.41267443529251e-05,
+      "loss": 0.353,
+      "step": 11674
+    },
+    {
+      "epoch": 0.5126842815899536,
+      "grad_norm": 1.65625,
+      "learning_rate": 2.411983032455316e-05,
+      "loss": 0.3513,
+      "step": 11676
+    },
+    {
+      "epoch": 0.5127721000691571,
+      "grad_norm": 1.8515625,
+      "learning_rate": 2.4112916363584828e-05,
+      "loss": 0.3692,
+      "step": 11678
+    },
+    {
+      "epoch": 0.5128599185483605,
+      "grad_norm": 1.9609375,
+      "learning_rate": 2.410600247054957e-05,
+      "loss": 0.3459,
+      "step": 11680
+    },
+    {
+      "epoch": 0.512947737027564,
+      "grad_norm": 1.7421875,
+      "learning_rate": 2.4099088645976855e-05,
+      "loss": 0.3597,
+      "step": 11682
+    },
+    {
+      "epoch": 0.5130355555067675,
+      "grad_norm": 1.921875,
+      "learning_rate": 2.409217489039615e-05,
+      "loss": 0.3649,
+      "step": 11684
+    },
+    {
+      "epoch": 0.513123373985971,
+      "grad_norm": 1.7890625,
+      "learning_rate": 2.4085261204336905e-05,
+      "loss": 0.3304,
+      "step": 11686
+    },
+    {
+      "epoch": 0.5132111924651744,
+      "grad_norm": 2.0,
+      "learning_rate": 2.407834758832858e-05,
+      "loss": 0.3645,
+      "step": 11688
+    },
+    {
+      "epoch": 0.513299010944378,
+      "grad_norm": 2.109375,
+      "learning_rate": 2.4071434042900627e-05,
+      "loss": 0.4048,
+      "step": 11690
+    },
+    {
+      "epoch": 0.5133868294235815,
+      "grad_norm": 1.7265625,
+      "learning_rate": 2.4064520568582468e-05,
+      "loss": 0.3575,
+      "step": 11692
+    },
+    {
+      "epoch": 0.513474647902785,
+      "grad_norm": 1.84375,
+      "learning_rate": 2.4057607165903557e-05,
+      "loss": 0.3888,
+      "step": 11694
+    },
+    {
+      "epoch": 0.5135624663819884,
+      "grad_norm": 2.03125,
+      "learning_rate": 2.4050693835393303e-05,
+      "loss": 0.3731,
+      "step": 11696
+    },
+    {
+      "epoch": 0.5136502848611919,
+      "grad_norm": 1.8125,
+      "learning_rate": 2.4043780577581145e-05,
+      "loss": 0.3508,
+      "step": 11698
+    },
+    {
+      "epoch": 0.5137381033403954,
+      "grad_norm": 1.734375,
+      "learning_rate": 2.40368673929965e-05,
+      "loss": 0.359,
+      "step": 11700
+    },
+    {
+      "epoch": 0.5138259218195989,
+      "grad_norm": 1.8359375,
+      "learning_rate": 2.4029954282168782e-05,
+      "loss": 0.3378,
+      "step": 11702
+    },
+    {
+      "epoch": 0.5139137402988023,
+      "grad_norm": 1.6953125,
+      "learning_rate": 2.40230412456274e-05,
+      "loss": 0.3509,
+      "step": 11704
+    },
+    {
+      "epoch": 0.5140015587780059,
+      "grad_norm": 1.8984375,
+      "learning_rate": 2.401612828390175e-05,
+      "loss": 0.3493,
+      "step": 11706
+    },
+    {
+      "epoch": 0.5140893772572094,
+      "grad_norm": 1.7734375,
+      "learning_rate": 2.4009215397521234e-05,
+      "loss": 0.3398,
+      "step": 11708
+    },
+    {
+      "epoch": 0.5141771957364129,
+      "grad_norm": 1.8203125,
+      "learning_rate": 2.4002302587015236e-05,
+      "loss": 0.3232,
+      "step": 11710
+    },
+    {
+      "epoch": 0.5142650142156163,
+      "grad_norm": 1.8828125,
+      "learning_rate": 2.399538985291315e-05,
+      "loss": 0.3542,
+      "step": 11712
+    },
+    {
+      "epoch": 0.5143528326948198,
+      "grad_norm": 1.75,
+      "learning_rate": 2.3988477195744353e-05,
+      "loss": 0.353,
+      "step": 11714
+    },
+    {
+      "epoch": 0.5144406511740233,
+      "grad_norm": 1.8828125,
+      "learning_rate": 2.398156461603821e-05,
+      "loss": 0.3577,
+      "step": 11716
+    },
+    {
+      "epoch": 0.5145284696532267,
+      "grad_norm": 1.9296875,
+      "learning_rate": 2.397465211432409e-05,
+      "loss": 0.3709,
+      "step": 11718
+    },
+    {
+      "epoch": 0.5146162881324303,
+      "grad_norm": 2.015625,
+      "learning_rate": 2.396773969113136e-05,
+      "loss": 0.3302,
+      "step": 11720
+    },
+    {
+      "epoch": 0.5147041066116338,
+      "grad_norm": 1.9140625,
+      "learning_rate": 2.396082734698938e-05,
+      "loss": 0.3461,
+      "step": 11722
+    },
+    {
+      "epoch": 0.5147919250908373,
+      "grad_norm": 1.7734375,
+      "learning_rate": 2.3953915082427482e-05,
+      "loss": 0.3456,
+      "step": 11724
+    },
+    {
+      "epoch": 0.5148797435700407,
+      "grad_norm": 1.9375,
+      "learning_rate": 2.3947002897975018e-05,
+      "loss": 0.3699,
+      "step": 11726
+    },
+    {
+      "epoch": 0.5149675620492442,
+      "grad_norm": 1.7421875,
+      "learning_rate": 2.3940090794161324e-05,
+      "loss": 0.3854,
+      "step": 11728
+    },
+    {
+      "epoch": 0.5150553805284477,
+      "grad_norm": 1.796875,
+      "learning_rate": 2.3933178771515735e-05,
+      "loss": 0.3623,
+      "step": 11730
+    },
+    {
+      "epoch": 0.5151431990076512,
+      "grad_norm": 1.8671875,
+      "learning_rate": 2.3926266830567567e-05,
+      "loss": 0.3466,
+      "step": 11732
+    },
+    {
+      "epoch": 0.5152310174868546,
+      "grad_norm": 1.734375,
+      "learning_rate": 2.3919354971846143e-05,
+      "loss": 0.3677,
+      "step": 11734
+    },
+    {
+      "epoch": 0.5153188359660582,
+      "grad_norm": 1.8125,
+      "learning_rate": 2.3912443195880776e-05,
+      "loss": 0.3551,
+      "step": 11736
+    },
+    {
+      "epoch": 0.5154066544452617,
+      "grad_norm": 1.8515625,
+      "learning_rate": 2.3905531503200768e-05,
+      "loss": 0.3481,
+      "step": 11738
+    },
+    {
+      "epoch": 0.5154944729244652,
+      "grad_norm": 2.109375,
+      "learning_rate": 2.3898619894335425e-05,
+      "loss": 0.3936,
+      "step": 11740
+    },
+    {
+      "epoch": 0.5155822914036686,
+      "grad_norm": 1.75,
+      "learning_rate": 2.3891708369814028e-05,
+      "loss": 0.3743,
+      "step": 11742
+    },
+    {
+      "epoch": 0.5156701098828721,
+      "grad_norm": 1.7421875,
+      "learning_rate": 2.3884796930165875e-05,
+      "loss": 0.3537,
+      "step": 11744
+    },
+    {
+      "epoch": 0.5157579283620756,
+      "grad_norm": 1.6796875,
+      "learning_rate": 2.3877885575920235e-05,
+      "loss": 0.3446,
+      "step": 11746
+    },
+    {
+      "epoch": 0.515845746841279,
+      "grad_norm": 1.8359375,
+      "learning_rate": 2.3870974307606386e-05,
+      "loss": 0.3612,
+      "step": 11748
+    },
+    {
+      "epoch": 0.5159335653204825,
+      "grad_norm": 1.90625,
+      "learning_rate": 2.38640631257536e-05,
+      "loss": 0.3155,
+      "step": 11750
+    },
+    {
+      "epoch": 0.5160213837996861,
+      "grad_norm": 2.03125,
+      "learning_rate": 2.385715203089114e-05,
+      "loss": 0.3664,
+      "step": 11752
+    },
+    {
+      "epoch": 0.5161092022788896,
+      "grad_norm": 1.9140625,
+      "learning_rate": 2.385024102354824e-05,
+      "loss": 0.3454,
+      "step": 11754
+    },
+    {
+      "epoch": 0.516197020758093,
+      "grad_norm": 1.6796875,
+      "learning_rate": 2.3843330104254165e-05,
+      "loss": 0.3397,
+      "step": 11756
+    },
+    {
+      "epoch": 0.5162848392372965,
+      "grad_norm": 1.796875,
+      "learning_rate": 2.383641927353814e-05,
+      "loss": 0.3574,
+      "step": 11758
+    },
+    {
+      "epoch": 0.5163726577165,
+      "grad_norm": 1.7734375,
+      "learning_rate": 2.3829508531929408e-05,
+      "loss": 0.3543,
+      "step": 11760
+    },
+    {
+      "epoch": 0.5164604761957035,
+      "grad_norm": 2.0,
+      "learning_rate": 2.3822597879957192e-05,
+      "loss": 0.3621,
+      "step": 11762
+    },
+    {
+      "epoch": 0.5165482946749069,
+      "grad_norm": 1.765625,
+      "learning_rate": 2.3815687318150716e-05,
+      "loss": 0.3541,
+      "step": 11764
+    },
+    {
+      "epoch": 0.5166361131541104,
+      "grad_norm": 1.765625,
+      "learning_rate": 2.3808776847039187e-05,
+      "loss": 0.3748,
+      "step": 11766
+    },
+    {
+      "epoch": 0.516723931633314,
+      "grad_norm": 2.4375,
+      "learning_rate": 2.3801866467151813e-05,
+      "loss": 0.3511,
+      "step": 11768
+    },
+    {
+      "epoch": 0.5168117501125175,
+      "grad_norm": 1.75,
+      "learning_rate": 2.3794956179017792e-05,
+      "loss": 0.3926,
+      "step": 11770
+    },
+    {
+      "epoch": 0.5168995685917209,
+      "grad_norm": 1.6640625,
+      "learning_rate": 2.3788045983166317e-05,
+      "loss": 0.3538,
+      "step": 11772
+    },
+    {
+      "epoch": 0.5169873870709244,
+      "grad_norm": 1.765625,
+      "learning_rate": 2.378113588012657e-05,
+      "loss": 0.334,
+      "step": 11774
+    },
+    {
+      "epoch": 0.5170752055501279,
+      "grad_norm": 1.5859375,
+      "learning_rate": 2.377422587042773e-05,
+      "loss": 0.3454,
+      "step": 11776
+    },
+    {
+      "epoch": 0.5171630240293313,
+      "grad_norm": 1.578125,
+      "learning_rate": 2.376731595459897e-05,
+      "loss": 0.3581,
+      "step": 11778
+    },
+    {
+      "epoch": 0.5172508425085348,
+      "grad_norm": 1.75,
+      "learning_rate": 2.3760406133169443e-05,
+      "loss": 0.3241,
+      "step": 11780
+    },
+    {
+      "epoch": 0.5173386609877384,
+      "grad_norm": 1.8203125,
+      "learning_rate": 2.3753496406668325e-05,
+      "loss": 0.3687,
+      "step": 11782
+    },
+    {
+      "epoch": 0.5174264794669419,
+      "grad_norm": 1.6015625,
+      "learning_rate": 2.3746586775624738e-05,
+      "loss": 0.3487,
+      "step": 11784
+    },
+    {
+      "epoch": 0.5175142979461453,
+      "grad_norm": 1.7265625,
+      "learning_rate": 2.3739677240567836e-05,
+      "loss": 0.349,
+      "step": 11786
+    },
+    {
+      "epoch": 0.5176021164253488,
+      "grad_norm": 1.6328125,
+      "learning_rate": 2.3732767802026757e-05,
+      "loss": 0.3642,
+      "step": 11788
+    },
+    {
+      "epoch": 0.5176899349045523,
+      "grad_norm": 1.734375,
+      "learning_rate": 2.372585846053062e-05,
+      "loss": 0.3601,
+      "step": 11790
+    },
+    {
+      "epoch": 0.5177777533837558,
+      "grad_norm": 2.015625,
+      "learning_rate": 2.3718949216608556e-05,
+      "loss": 0.3388,
+      "step": 11792
+    },
+    {
+      "epoch": 0.5178655718629592,
+      "grad_norm": 1.7265625,
+      "learning_rate": 2.371204007078966e-05,
+      "loss": 0.3462,
+      "step": 11794
+    },
+    {
+      "epoch": 0.5179533903421627,
+      "grad_norm": 1.7421875,
+      "learning_rate": 2.3705131023603043e-05,
+      "loss": 0.3661,
+      "step": 11796
+    },
+    {
+      "epoch": 0.5180412088213663,
+      "grad_norm": 1.78125,
+      "learning_rate": 2.3698222075577805e-05,
+      "loss": 0.3453,
+      "step": 11798
+    },
+    {
+      "epoch": 0.5181290273005698,
+      "grad_norm": 1.90625,
+      "learning_rate": 2.3691313227243033e-05,
+      "loss": 0.3558,
+      "step": 11800
+    },
+    {
+      "epoch": 0.5182168457797732,
+      "grad_norm": 1.9609375,
+      "learning_rate": 2.3684404479127813e-05,
+      "loss": 0.3436,
+      "step": 11802
+    },
+    {
+      "epoch": 0.5183046642589767,
+      "grad_norm": 1.6875,
+      "learning_rate": 2.3677495831761205e-05,
+      "loss": 0.3589,
+      "step": 11804
+    },
+    {
+      "epoch": 0.5183924827381802,
+      "grad_norm": 1.703125,
+      "learning_rate": 2.3670587285672284e-05,
+      "loss": 0.3264,
+      "step": 11806
+    },
+    {
+      "epoch": 0.5184803012173836,
+      "grad_norm": 1.703125,
+      "learning_rate": 2.366367884139011e-05,
+      "loss": 0.3578,
+      "step": 11808
+    },
+    {
+      "epoch": 0.5185681196965871,
+      "grad_norm": 1.6796875,
+      "learning_rate": 2.365677049944373e-05,
+      "loss": 0.3635,
+      "step": 11810
+    },
+    {
+      "epoch": 0.5186559381757906,
+      "grad_norm": 1.6953125,
+      "learning_rate": 2.364986226036219e-05,
+      "loss": 0.3389,
+      "step": 11812
+    },
+    {
+      "epoch": 0.5187437566549942,
+      "grad_norm": 1.7578125,
+      "learning_rate": 2.3642954124674523e-05,
+      "loss": 0.3554,
+      "step": 11814
+    },
+    {
+      "epoch": 0.5188315751341976,
+      "grad_norm": 1.875,
+      "learning_rate": 2.3636046092909754e-05,
+      "loss": 0.3582,
+      "step": 11816
+    },
+    {
+      "epoch": 0.5189193936134011,
+      "grad_norm": 1.9140625,
+      "learning_rate": 2.36291381655969e-05,
+      "loss": 0.3479,
+      "step": 11818
+    },
+    {
+      "epoch": 0.5190072120926046,
+      "grad_norm": 1.8515625,
+      "learning_rate": 2.362223034326497e-05,
+      "loss": 0.3522,
+      "step": 11820
+    },
+    {
+      "epoch": 0.5190950305718081,
+      "grad_norm": 1.9609375,
+      "learning_rate": 2.3615322626442977e-05,
+      "loss": 0.3455,
+      "step": 11822
+    },
+    {
+      "epoch": 0.5191828490510115,
+      "grad_norm": 2.109375,
+      "learning_rate": 2.36084150156599e-05,
+      "loss": 0.3522,
+      "step": 11824
+    },
+    {
+      "epoch": 0.519270667530215,
+      "grad_norm": 1.8984375,
+      "learning_rate": 2.360150751144474e-05,
+      "loss": 0.3604,
+      "step": 11826
+    },
+    {
+      "epoch": 0.5193584860094186,
+      "grad_norm": 1.625,
+      "learning_rate": 2.3594600114326476e-05,
+      "loss": 0.3782,
+      "step": 11828
+    },
+    {
+      "epoch": 0.5194463044886221,
+      "grad_norm": 1.6015625,
+      "learning_rate": 2.3587692824834066e-05,
+      "loss": 0.3688,
+      "step": 11830
+    },
+    {
+      "epoch": 0.5195341229678255,
+      "grad_norm": 1.8984375,
+      "learning_rate": 2.358078564349648e-05,
+      "loss": 0.3603,
+      "step": 11832
+    },
+    {
+      "epoch": 0.519621941447029,
+      "grad_norm": 1.75,
+      "learning_rate": 2.3573878570842664e-05,
+      "loss": 0.3842,
+      "step": 11834
+    },
+    {
+      "epoch": 0.5197097599262325,
+      "grad_norm": 1.7890625,
+      "learning_rate": 2.3566971607401574e-05,
+      "loss": 0.3662,
+      "step": 11836
+    },
+    {
+      "epoch": 0.519797578405436,
+      "grad_norm": 1.8125,
+      "learning_rate": 2.3560064753702144e-05,
+      "loss": 0.3501,
+      "step": 11838
+    },
+    {
+      "epoch": 0.5198853968846394,
+      "grad_norm": 1.625,
+      "learning_rate": 2.3553158010273295e-05,
+      "loss": 0.3431,
+      "step": 11840
+    },
+    {
+      "epoch": 0.5199732153638429,
+      "grad_norm": 1.75,
+      "learning_rate": 2.3546251377643955e-05,
+      "loss": 0.3624,
+      "step": 11842
+    },
+    {
+      "epoch": 0.5200610338430465,
+      "grad_norm": 1.828125,
+      "learning_rate": 2.353934485634304e-05,
+      "loss": 0.3681,
+      "step": 11844
+    },
+    {
+      "epoch": 0.5201488523222499,
+      "grad_norm": 1.8828125,
+      "learning_rate": 2.3532438446899437e-05,
+      "loss": 0.3293,
+      "step": 11846
+    },
+    {
+      "epoch": 0.5202366708014534,
+      "grad_norm": 1.75,
+      "learning_rate": 2.3525532149842054e-05,
+      "loss": 0.351,
+      "step": 11848
+    },
+    {
+      "epoch": 0.5203244892806569,
+      "grad_norm": 1.6328125,
+      "learning_rate": 2.3518625965699767e-05,
+      "loss": 0.3449,
+      "step": 11850
+    },
+    {
+      "epoch": 0.5204123077598604,
+      "grad_norm": 1.8984375,
+      "learning_rate": 2.351171989500146e-05,
+      "loss": 0.3673,
+      "step": 11852
+    },
+    {
+      "epoch": 0.5205001262390638,
+      "grad_norm": 1.6875,
+      "learning_rate": 2.3504813938276005e-05,
+      "loss": 0.3409,
+      "step": 11854
+    },
+    {
+      "epoch": 0.5205879447182673,
+      "grad_norm": 1.8671875,
+      "learning_rate": 2.349790809605225e-05,
+      "loss": 0.3701,
+      "step": 11856
+    },
+    {
+      "epoch": 0.5206757631974708,
+      "grad_norm": 1.859375,
+      "learning_rate": 2.349100236885906e-05,
+      "loss": 0.3485,
+      "step": 11858
+    },
+    {
+      "epoch": 0.5207635816766744,
+      "grad_norm": 1.7265625,
+      "learning_rate": 2.3484096757225263e-05,
+      "loss": 0.3583,
+      "step": 11860
+    },
+    {
+      "epoch": 0.5208514001558778,
+      "grad_norm": 2.0625,
+      "learning_rate": 2.3477191261679704e-05,
+      "loss": 0.35,
+      "step": 11862
+    },
+    {
+      "epoch": 0.5209392186350813,
+      "grad_norm": 1.8125,
+      "learning_rate": 2.3470285882751205e-05,
+      "loss": 0.3593,
+      "step": 11864
+    },
+    {
+      "epoch": 0.5210270371142848,
+      "grad_norm": 1.8359375,
+      "learning_rate": 2.3463380620968576e-05,
+      "loss": 0.3636,
+      "step": 11866
+    },
+    {
+      "epoch": 0.5211148555934882,
+      "grad_norm": 1.75,
+      "learning_rate": 2.345647547686063e-05,
+      "loss": 0.3752,
+      "step": 11868
+    },
+    {
+      "epoch": 0.5212026740726917,
+      "grad_norm": 1.84375,
+      "learning_rate": 2.344957045095616e-05,
+      "loss": 0.3505,
+      "step": 11870
+    },
+    {
+      "epoch": 0.5212904925518952,
+      "grad_norm": 2.140625,
+      "learning_rate": 2.3442665543783957e-05,
+      "loss": 0.3737,
+      "step": 11872
+    },
+    {
+      "epoch": 0.5213783110310988,
+      "grad_norm": 2.09375,
+      "learning_rate": 2.343576075587281e-05,
+      "loss": 0.3533,
+      "step": 11874
+    },
+    {
+      "epoch": 0.5214661295103022,
+      "grad_norm": 1.8359375,
+      "learning_rate": 2.342885608775147e-05,
+      "loss": 0.3613,
+      "step": 11876
+    },
+    {
+      "epoch": 0.5215539479895057,
+      "grad_norm": 1.9140625,
+      "learning_rate": 2.342195153994871e-05,
+      "loss": 0.342,
+      "step": 11878
+    },
+    {
+      "epoch": 0.5216417664687092,
+      "grad_norm": 1.7421875,
+      "learning_rate": 2.3415047112993274e-05,
+      "loss": 0.3746,
+      "step": 11880
+    },
+    {
+      "epoch": 0.5217295849479127,
+      "grad_norm": 1.7734375,
+      "learning_rate": 2.340814280741391e-05,
+      "loss": 0.3362,
+      "step": 11882
+    },
+    {
+      "epoch": 0.5218174034271161,
+      "grad_norm": 1.8671875,
+      "learning_rate": 2.3401238623739354e-05,
+      "loss": 0.3482,
+      "step": 11884
+    },
+    {
+      "epoch": 0.5219052219063196,
+      "grad_norm": 1.6015625,
+      "learning_rate": 2.3394334562498328e-05,
+      "loss": 0.3655,
+      "step": 11886
+    },
+    {
+      "epoch": 0.5219930403855231,
+      "grad_norm": 1.609375,
+      "learning_rate": 2.3387430624219546e-05,
+      "loss": 0.3439,
+      "step": 11888
+    },
+    {
+      "epoch": 0.5220808588647267,
+      "grad_norm": 2.109375,
+      "learning_rate": 2.3380526809431717e-05,
+      "loss": 0.3549,
+      "step": 11890
+    },
+    {
+      "epoch": 0.5221686773439301,
+      "grad_norm": 1.8515625,
+      "learning_rate": 2.3373623118663528e-05,
+      "loss": 0.3444,
+      "step": 11892
+    },
+    {
+      "epoch": 0.5222564958231336,
+      "grad_norm": 1.796875,
+      "learning_rate": 2.336671955244367e-05,
+      "loss": 0.3373,
+      "step": 11894
+    },
+    {
+      "epoch": 0.5223443143023371,
+      "grad_norm": 1.734375,
+      "learning_rate": 2.335981611130082e-05,
+      "loss": 0.3576,
+      "step": 11896
+    },
+    {
+      "epoch": 0.5224321327815405,
+      "grad_norm": 2.046875,
+      "learning_rate": 2.3352912795763646e-05,
+      "loss": 0.3655,
+      "step": 11898
+    },
+    {
+      "epoch": 0.522519951260744,
+      "grad_norm": 2.0,
+      "learning_rate": 2.334600960636081e-05,
+      "loss": 0.376,
+      "step": 11900
+    },
+    {
+      "epoch": 0.5226077697399475,
+      "grad_norm": 1.84375,
+      "learning_rate": 2.333910654362095e-05,
+      "loss": 0.3281,
+      "step": 11902
+    },
+    {
+      "epoch": 0.522695588219151,
+      "grad_norm": 1.8203125,
+      "learning_rate": 2.333220360807271e-05,
+      "loss": 0.3497,
+      "step": 11904
+    },
+    {
+      "epoch": 0.5227834066983545,
+      "grad_norm": 1.828125,
+      "learning_rate": 2.3325300800244726e-05,
+      "loss": 0.3409,
+      "step": 11906
+    },
+    {
+      "epoch": 0.522871225177558,
+      "grad_norm": 1.890625,
+      "learning_rate": 2.3318398120665598e-05,
+      "loss": 0.3553,
+      "step": 11908
+    },
+    {
+      "epoch": 0.5229590436567615,
+      "grad_norm": 1.6875,
+      "learning_rate": 2.3311495569863945e-05,
+      "loss": 0.3339,
+      "step": 11910
+    },
+    {
+      "epoch": 0.523046862135965,
+      "grad_norm": 1.859375,
+      "learning_rate": 2.3304593148368366e-05,
+      "loss": 0.3571,
+      "step": 11912
+    },
+    {
+      "epoch": 0.5231346806151684,
+      "grad_norm": 1.8515625,
+      "learning_rate": 2.3297690856707458e-05,
+      "loss": 0.3476,
+      "step": 11914
+    },
+    {
+      "epoch": 0.5232224990943719,
+      "grad_norm": 1.8203125,
+      "learning_rate": 2.3290788695409785e-05,
+      "loss": 0.3489,
+      "step": 11916
+    },
+    {
+      "epoch": 0.5233103175735754,
+      "grad_norm": 1.6875,
+      "learning_rate": 2.3283886665003924e-05,
+      "loss": 0.3602,
+      "step": 11918
+    },
+    {
+      "epoch": 0.5233981360527789,
+      "grad_norm": 1.8671875,
+      "learning_rate": 2.327698476601843e-05,
+      "loss": 0.3553,
+      "step": 11920
+    },
+    {
+      "epoch": 0.5234859545319824,
+      "grad_norm": 1.7421875,
+      "learning_rate": 2.327008299898186e-05,
+      "loss": 0.3447,
+      "step": 11922
+    },
+    {
+      "epoch": 0.5235737730111859,
+      "grad_norm": 1.7578125,
+      "learning_rate": 2.3263181364422748e-05,
+      "loss": 0.3234,
+      "step": 11924
+    },
+    {
+      "epoch": 0.5236615914903894,
+      "grad_norm": 1.828125,
+      "learning_rate": 2.3256279862869626e-05,
+      "loss": 0.3697,
+      "step": 11926
+    },
+    {
+      "epoch": 0.5237494099695928,
+      "grad_norm": 1.7265625,
+      "learning_rate": 2.3249378494851008e-05,
+      "loss": 0.3523,
+      "step": 11928
+    },
+    {
+      "epoch": 0.5238372284487963,
+      "grad_norm": 1.8203125,
+      "learning_rate": 2.3242477260895404e-05,
+      "loss": 0.356,
+      "step": 11930
+    },
+    {
+      "epoch": 0.5239250469279998,
+      "grad_norm": 1.7109375,
+      "learning_rate": 2.323557616153131e-05,
+      "loss": 0.3551,
+      "step": 11932
+    },
+    {
+      "epoch": 0.5240128654072033,
+      "grad_norm": 1.8359375,
+      "learning_rate": 2.322867519728722e-05,
+      "loss": 0.3631,
+      "step": 11934
+    },
+    {
+      "epoch": 0.5241006838864068,
+      "grad_norm": 1.8046875,
+      "learning_rate": 2.3221774368691616e-05,
+      "loss": 0.34,
+      "step": 11936
+    },
+    {
+      "epoch": 0.5241885023656103,
+      "grad_norm": 1.578125,
+      "learning_rate": 2.3214873676272948e-05,
+      "loss": 0.3139,
+      "step": 11938
+    },
+    {
+      "epoch": 0.5242763208448138,
+      "grad_norm": 1.765625,
+      "learning_rate": 2.320797312055969e-05,
+      "loss": 0.3523,
+      "step": 11940
+    },
+    {
+      "epoch": 0.5243641393240173,
+      "grad_norm": 1.7734375,
+      "learning_rate": 2.3201072702080267e-05,
+      "loss": 0.3513,
+      "step": 11942
+    },
+    {
+      "epoch": 0.5244519578032207,
+      "grad_norm": 1.6015625,
+      "learning_rate": 2.3194172421363132e-05,
+      "loss": 0.3215,
+      "step": 11944
+    },
+    {
+      "epoch": 0.5245397762824242,
+      "grad_norm": 1.75,
+      "learning_rate": 2.3187272278936705e-05,
+      "loss": 0.3471,
+      "step": 11946
+    },
+    {
+      "epoch": 0.5246275947616277,
+      "grad_norm": 1.9921875,
+      "learning_rate": 2.3180372275329404e-05,
+      "loss": 0.3485,
+      "step": 11948
+    },
+    {
+      "epoch": 0.5247154132408312,
+      "grad_norm": 1.9140625,
+      "learning_rate": 2.317347241106963e-05,
+      "loss": 0.3585,
+      "step": 11950
+    },
+    {
+      "epoch": 0.5248032317200347,
+      "grad_norm": 1.65625,
+      "learning_rate": 2.316657268668578e-05,
+      "loss": 0.3414,
+      "step": 11952
+    },
+    {
+      "epoch": 0.5248910501992382,
+      "grad_norm": 1.96875,
+      "learning_rate": 2.315967310270623e-05,
+      "loss": 0.3344,
+      "step": 11954
+    },
+    {
+      "epoch": 0.5249788686784417,
+      "grad_norm": 2.1875,
+      "learning_rate": 2.315277365965935e-05,
+      "loss": 0.3659,
+      "step": 11956
+    },
+    {
+      "epoch": 0.5250666871576452,
+      "grad_norm": 2.046875,
+      "learning_rate": 2.314587435807351e-05,
+      "loss": 0.3661,
+      "step": 11958
+    },
+    {
+      "epoch": 0.5251545056368486,
+      "grad_norm": 2.1875,
+      "learning_rate": 2.3138975198477056e-05,
+      "loss": 0.3642,
+      "step": 11960
+    },
+    {
+      "epoch": 0.5252423241160521,
+      "grad_norm": 1.9140625,
+      "learning_rate": 2.3132076181398332e-05,
+      "loss": 0.3845,
+      "step": 11962
+    },
+    {
+      "epoch": 0.5253301425952556,
+      "grad_norm": 1.703125,
+      "learning_rate": 2.3125177307365658e-05,
+      "loss": 0.3446,
+      "step": 11964
+    },
+    {
+      "epoch": 0.525417961074459,
+      "grad_norm": 1.9296875,
+      "learning_rate": 2.3118278576907366e-05,
+      "loss": 0.3537,
+      "step": 11966
+    },
+    {
+      "epoch": 0.5255057795536626,
+      "grad_norm": 2.0625,
+      "learning_rate": 2.3111379990551736e-05,
+      "loss": 0.3397,
+      "step": 11968
+    },
+    {
+      "epoch": 0.5255935980328661,
+      "grad_norm": 1.6015625,
+      "learning_rate": 2.3104481548827084e-05,
+      "loss": 0.3742,
+      "step": 11970
+    },
+    {
+      "epoch": 0.5256814165120696,
+      "grad_norm": 1.828125,
+      "learning_rate": 2.309758325226169e-05,
+      "loss": 0.3405,
+      "step": 11972
+    },
+    {
+      "epoch": 0.525769234991273,
+      "grad_norm": 1.8828125,
+      "learning_rate": 2.3090685101383824e-05,
+      "loss": 0.333,
+      "step": 11974
+    },
+    {
+      "epoch": 0.5258570534704765,
+      "grad_norm": 1.7421875,
+      "learning_rate": 2.3083787096721756e-05,
+      "loss": 0.3498,
+      "step": 11976
+    },
+    {
+      "epoch": 0.52594487194968,
+      "grad_norm": 1.9765625,
+      "learning_rate": 2.3076889238803727e-05,
+      "loss": 0.3476,
+      "step": 11978
+    },
+    {
+      "epoch": 0.5260326904288835,
+      "grad_norm": 1.671875,
+      "learning_rate": 2.3069991528157982e-05,
+      "loss": 0.3544,
+      "step": 11980
+    },
+    {
+      "epoch": 0.526120508908087,
+      "grad_norm": 1.90625,
+      "learning_rate": 2.3063093965312747e-05,
+      "loss": 0.3416,
+      "step": 11982
+    },
+    {
+      "epoch": 0.5262083273872905,
+      "grad_norm": 1.75,
+      "learning_rate": 2.305619655079624e-05,
+      "loss": 0.3546,
+      "step": 11984
+    },
+    {
+      "epoch": 0.526296145866494,
+      "grad_norm": 1.984375,
+      "learning_rate": 2.3049299285136667e-05,
+      "loss": 0.3546,
+      "step": 11986
+    },
+    {
+      "epoch": 0.5263839643456975,
+      "grad_norm": 1.796875,
+      "learning_rate": 2.304240216886223e-05,
+      "loss": 0.3563,
+      "step": 11988
+    },
+    {
+      "epoch": 0.5264717828249009,
+      "grad_norm": 1.5546875,
+      "learning_rate": 2.3035505202501095e-05,
+      "loss": 0.3371,
+      "step": 11990
+    },
+    {
+      "epoch": 0.5265596013041044,
+      "grad_norm": 1.8046875,
+      "learning_rate": 2.3028608386581446e-05,
+      "loss": 0.3759,
+      "step": 11992
+    },
+    {
+      "epoch": 0.5266474197833079,
+      "grad_norm": 1.609375,
+      "learning_rate": 2.302171172163144e-05,
+      "loss": 0.3462,
+      "step": 11994
+    },
+    {
+      "epoch": 0.5267352382625113,
+      "grad_norm": 1.7890625,
+      "learning_rate": 2.301481520817922e-05,
+      "loss": 0.3657,
+      "step": 11996
+    },
+    {
+      "epoch": 0.5268230567417149,
+      "grad_norm": 1.65625,
+      "learning_rate": 2.300791884675294e-05,
+      "loss": 0.3392,
+      "step": 11998
+    },
+    {
+      "epoch": 0.5269108752209184,
+      "grad_norm": 1.640625,
+      "learning_rate": 2.3001022637880705e-05,
+      "loss": 0.3366,
+      "step": 12000
+    },
+    {
+      "epoch": 0.5269986937001219,
+      "grad_norm": 1.65625,
+      "learning_rate": 2.2994126582090635e-05,
+      "loss": 0.3486,
+      "step": 12002
+    },
+    {
+      "epoch": 0.5270865121793253,
+      "grad_norm": 1.71875,
+      "learning_rate": 2.298723067991083e-05,
+      "loss": 0.3687,
+      "step": 12004
+    },
+    {
+      "epoch": 0.5271743306585288,
+      "grad_norm": 1.7421875,
+      "learning_rate": 2.2980334931869386e-05,
+      "loss": 0.3494,
+      "step": 12006
+    },
+    {
+      "epoch": 0.5272621491377323,
+      "grad_norm": 1.8671875,
+      "learning_rate": 2.2973439338494372e-05,
+      "loss": 0.3455,
+      "step": 12008
+    },
+    {
+      "epoch": 0.5273499676169358,
+      "grad_norm": 1.6484375,
+      "learning_rate": 2.296654390031386e-05,
+      "loss": 0.3771,
+      "step": 12010
+    },
+    {
+      "epoch": 0.5274377860961392,
+      "grad_norm": 1.671875,
+      "learning_rate": 2.295964861785591e-05,
+      "loss": 0.3492,
+      "step": 12012
+    },
+    {
+      "epoch": 0.5275256045753428,
+      "grad_norm": 1.7890625,
+      "learning_rate": 2.295275349164855e-05,
+      "loss": 0.3332,
+      "step": 12014
+    },
+    {
+      "epoch": 0.5276134230545463,
+      "grad_norm": 1.6953125,
+      "learning_rate": 2.2945858522219822e-05,
+      "loss": 0.3673,
+      "step": 12016
+    },
+    {
+      "epoch": 0.5277012415337498,
+      "grad_norm": 1.734375,
+      "learning_rate": 2.293896371009774e-05,
+      "loss": 0.3439,
+      "step": 12018
+    },
+    {
+      "epoch": 0.5277890600129532,
+      "grad_norm": 1.671875,
+      "learning_rate": 2.293206905581031e-05,
+      "loss": 0.3268,
+      "step": 12020
+    },
+    {
+      "epoch": 0.5278768784921567,
+      "grad_norm": 1.7265625,
+      "learning_rate": 2.2925174559885526e-05,
+      "loss": 0.3308,
+      "step": 12022
+    },
+    {
+      "epoch": 0.5279646969713602,
+      "grad_norm": 2.0,
+      "learning_rate": 2.291828022285138e-05,
+      "loss": 0.3547,
+      "step": 12024
+    },
+    {
+      "epoch": 0.5280525154505636,
+      "grad_norm": 1.703125,
+      "learning_rate": 2.2911386045235826e-05,
+      "loss": 0.3416,
+      "step": 12026
+    },
+    {
+      "epoch": 0.5281403339297672,
+      "grad_norm": 2.171875,
+      "learning_rate": 2.290449202756684e-05,
+      "loss": 0.3571,
+      "step": 12028
+    },
+    {
+      "epoch": 0.5282281524089707,
+      "grad_norm": 2.0625,
+      "learning_rate": 2.2897598170372346e-05,
+      "loss": 0.3452,
+      "step": 12030
+    },
+    {
+      "epoch": 0.5283159708881742,
+      "grad_norm": 1.7578125,
+      "learning_rate": 2.289070447418029e-05,
+      "loss": 0.3385,
+      "step": 12032
+    },
+    {
+      "epoch": 0.5284037893673776,
+      "grad_norm": 1.6796875,
+      "learning_rate": 2.288381093951859e-05,
+      "loss": 0.3539,
+      "step": 12034
+    },
+    {
+      "epoch": 0.5284916078465811,
+      "grad_norm": 1.953125,
+      "learning_rate": 2.2876917566915158e-05,
+      "loss": 0.3663,
+      "step": 12036
+    },
+    {
+      "epoch": 0.5285794263257846,
+      "grad_norm": 1.640625,
+      "learning_rate": 2.2870024356897887e-05,
+      "loss": 0.3517,
+      "step": 12038
+    },
+    {
+      "epoch": 0.5286672448049881,
+      "grad_norm": 1.765625,
+      "learning_rate": 2.286313130999466e-05,
+      "loss": 0.321,
+      "step": 12040
+    },
+    {
+      "epoch": 0.5287550632841915,
+      "grad_norm": 2.09375,
+      "learning_rate": 2.285623842673335e-05,
+      "loss": 0.3667,
+      "step": 12042
+    },
+    {
+      "epoch": 0.5288428817633951,
+      "grad_norm": 1.6640625,
+      "learning_rate": 2.2849345707641815e-05,
+      "loss": 0.3561,
+      "step": 12044
+    },
+    {
+      "epoch": 0.5289307002425986,
+      "grad_norm": 1.71875,
+      "learning_rate": 2.28424531532479e-05,
+      "loss": 0.3227,
+      "step": 12046
+    },
+    {
+      "epoch": 0.5290185187218021,
+      "grad_norm": 1.7578125,
+      "learning_rate": 2.2835560764079446e-05,
+      "loss": 0.3416,
+      "step": 12048
+    },
+    {
+      "epoch": 0.5291063372010055,
+      "grad_norm": 1.8828125,
+      "learning_rate": 2.282866854066426e-05,
+      "loss": 0.3529,
+      "step": 12050
+    },
+    {
+      "epoch": 0.529194155680209,
+      "grad_norm": 2.125,
+      "learning_rate": 2.282177648353016e-05,
+      "loss": 0.3454,
+      "step": 12052
+    },
+    {
+      "epoch": 0.5292819741594125,
+      "grad_norm": 1.7578125,
+      "learning_rate": 2.281488459320494e-05,
+      "loss": 0.3768,
+      "step": 12054
+    },
+    {
+      "epoch": 0.5293697926386159,
+      "grad_norm": 1.8828125,
+      "learning_rate": 2.280799287021638e-05,
+      "loss": 0.3357,
+      "step": 12056
+    },
+    {
+      "epoch": 0.5294576111178194,
+      "grad_norm": 1.796875,
+      "learning_rate": 2.280110131509226e-05,
+      "loss": 0.3533,
+      "step": 12058
+    },
+    {
+      "epoch": 0.529545429597023,
+      "grad_norm": 1.671875,
+      "learning_rate": 2.2794209928360322e-05,
+      "loss": 0.3469,
+      "step": 12060
+    },
+    {
+      "epoch": 0.5296332480762265,
+      "grad_norm": 1.875,
+      "learning_rate": 2.2787318710548313e-05,
+      "loss": 0.3259,
+      "step": 12062
+    },
+    {
+      "epoch": 0.5297210665554299,
+      "grad_norm": 1.890625,
+      "learning_rate": 2.2780427662183975e-05,
+      "loss": 0.3613,
+      "step": 12064
+    },
+    {
+      "epoch": 0.5298088850346334,
+      "grad_norm": 1.859375,
+      "learning_rate": 2.2773536783795012e-05,
+      "loss": 0.3602,
+      "step": 12066
+    },
+    {
+      "epoch": 0.5298967035138369,
+      "grad_norm": 1.6015625,
+      "learning_rate": 2.2766646075909137e-05,
+      "loss": 0.3444,
+      "step": 12068
+    },
+    {
+      "epoch": 0.5299845219930404,
+      "grad_norm": 1.8046875,
+      "learning_rate": 2.275975553905404e-05,
+      "loss": 0.3097,
+      "step": 12070
+    },
+    {
+      "epoch": 0.5300723404722438,
+      "grad_norm": 1.734375,
+      "learning_rate": 2.2752865173757403e-05,
+      "loss": 0.3388,
+      "step": 12072
+    },
+    {
+      "epoch": 0.5301601589514474,
+      "grad_norm": 1.78125,
+      "learning_rate": 2.2745974980546893e-05,
+      "loss": 0.3593,
+      "step": 12074
+    },
+    {
+      "epoch": 0.5302479774306509,
+      "grad_norm": 1.6484375,
+      "learning_rate": 2.273908495995015e-05,
+      "loss": 0.3607,
+      "step": 12076
+    },
+    {
+      "epoch": 0.5303357959098544,
+      "grad_norm": 1.75,
+      "learning_rate": 2.273219511249483e-05,
+      "loss": 0.3437,
+      "step": 12078
+    },
+    {
+      "epoch": 0.5304236143890578,
+      "grad_norm": 1.703125,
+      "learning_rate": 2.272530543870855e-05,
+      "loss": 0.3302,
+      "step": 12080
+    },
+    {
+      "epoch": 0.5305114328682613,
+      "grad_norm": 1.84375,
+      "learning_rate": 2.2718415939118924e-05,
+      "loss": 0.3521,
+      "step": 12082
+    },
+    {
+      "epoch": 0.5305992513474648,
+      "grad_norm": 1.859375,
+      "learning_rate": 2.271152661425355e-05,
+      "loss": 0.3512,
+      "step": 12084
+    },
+    {
+      "epoch": 0.5306870698266682,
+      "grad_norm": 1.6640625,
+      "learning_rate": 2.2704637464640026e-05,
+      "loss": 0.3374,
+      "step": 12086
+    },
+    {
+      "epoch": 0.5307748883058717,
+      "grad_norm": 1.859375,
+      "learning_rate": 2.2697748490805905e-05,
+      "loss": 0.3573,
+      "step": 12088
+    },
+    {
+      "epoch": 0.5308627067850753,
+      "grad_norm": 1.5625,
+      "learning_rate": 2.2690859693278772e-05,
+      "loss": 0.3552,
+      "step": 12090
+    },
+    {
+      "epoch": 0.5309505252642788,
+      "grad_norm": 1.8671875,
+      "learning_rate": 2.2683971072586145e-05,
+      "loss": 0.3585,
+      "step": 12092
+    },
+    {
+      "epoch": 0.5310383437434822,
+      "grad_norm": 1.828125,
+      "learning_rate": 2.2677082629255566e-05,
+      "loss": 0.3361,
+      "step": 12094
+    },
+    {
+      "epoch": 0.5311261622226857,
+      "grad_norm": 1.734375,
+      "learning_rate": 2.2670194363814558e-05,
+      "loss": 0.3917,
+      "step": 12096
+    },
+    {
+      "epoch": 0.5312139807018892,
+      "grad_norm": 1.6484375,
+      "learning_rate": 2.2663306276790625e-05,
+      "loss": 0.3448,
+      "step": 12098
+    },
+    {
+      "epoch": 0.5313017991810927,
+      "grad_norm": 1.7109375,
+      "learning_rate": 2.265641836871126e-05,
+      "loss": 0.3332,
+      "step": 12100
+    },
+    {
+      "epoch": 0.5313896176602961,
+      "grad_norm": 1.765625,
+      "learning_rate": 2.2649530640103936e-05,
+      "loss": 0.3468,
+      "step": 12102
+    },
+    {
+      "epoch": 0.5314774361394996,
+      "grad_norm": 1.8671875,
+      "learning_rate": 2.264264309149612e-05,
+      "loss": 0.362,
+      "step": 12104
+    },
+    {
+      "epoch": 0.5315652546187032,
+      "grad_norm": 1.7734375,
+      "learning_rate": 2.2635755723415256e-05,
+      "loss": 0.3369,
+      "step": 12106
+    },
+    {
+      "epoch": 0.5316530730979067,
+      "grad_norm": 1.9375,
+      "learning_rate": 2.262886853638879e-05,
+      "loss": 0.3461,
+      "step": 12108
+    },
+    {
+      "epoch": 0.5317408915771101,
+      "grad_norm": 1.625,
+      "learning_rate": 2.262198153094414e-05,
+      "loss": 0.3503,
+      "step": 12110
+    },
+    {
+      "epoch": 0.5318287100563136,
+      "grad_norm": 1.8359375,
+      "learning_rate": 2.2615094707608715e-05,
+      "loss": 0.3132,
+      "step": 12112
+    },
+    {
+      "epoch": 0.5319165285355171,
+      "grad_norm": 1.8359375,
+      "learning_rate": 2.2608208066909904e-05,
+      "loss": 0.3519,
+      "step": 12114
+    },
+    {
+      "epoch": 0.5320043470147205,
+      "grad_norm": 2.046875,
+      "learning_rate": 2.2601321609375097e-05,
+      "loss": 0.3466,
+      "step": 12116
+    },
+    {
+      "epoch": 0.532092165493924,
+      "grad_norm": 2.078125,
+      "learning_rate": 2.2594435335531655e-05,
+      "loss": 0.3607,
+      "step": 12118
+    },
+    {
+      "epoch": 0.5321799839731275,
+      "grad_norm": 1.5859375,
+      "learning_rate": 2.258754924590694e-05,
+      "loss": 0.3258,
+      "step": 12120
+    },
+    {
+      "epoch": 0.5322678024523311,
+      "grad_norm": 1.859375,
+      "learning_rate": 2.2580663341028273e-05,
+      "loss": 0.3427,
+      "step": 12122
+    },
+    {
+      "epoch": 0.5323556209315345,
+      "grad_norm": 1.7734375,
+      "learning_rate": 2.2573777621422985e-05,
+      "loss": 0.3639,
+      "step": 12124
+    },
+    {
+      "epoch": 0.532443439410738,
+      "grad_norm": 1.9765625,
+      "learning_rate": 2.25668920876184e-05,
+      "loss": 0.3382,
+      "step": 12126
+    },
+    {
+      "epoch": 0.5325312578899415,
+      "grad_norm": 1.734375,
+      "learning_rate": 2.256000674014179e-05,
+      "loss": 0.3362,
+      "step": 12128
+    },
+    {
+      "epoch": 0.532619076369145,
+      "grad_norm": 1.78125,
+      "learning_rate": 2.2553121579520454e-05,
+      "loss": 0.3644,
+      "step": 12130
+    },
+    {
+      "epoch": 0.5327068948483484,
+      "grad_norm": 1.7578125,
+      "learning_rate": 2.254623660628165e-05,
+      "loss": 0.3392,
+      "step": 12132
+    },
+    {
+      "epoch": 0.5327947133275519,
+      "grad_norm": 1.6875,
+      "learning_rate": 2.2539351820952636e-05,
+      "loss": 0.3422,
+      "step": 12134
+    },
+    {
+      "epoch": 0.5328825318067555,
+      "grad_norm": 1.890625,
+      "learning_rate": 2.2532467224060656e-05,
+      "loss": 0.3382,
+      "step": 12136
+    },
+    {
+      "epoch": 0.532970350285959,
+      "grad_norm": 1.7265625,
+      "learning_rate": 2.2525582816132922e-05,
+      "loss": 0.3704,
+      "step": 12138
+    },
+    {
+      "epoch": 0.5330581687651624,
+      "grad_norm": 1.890625,
+      "learning_rate": 2.2518698597696646e-05,
+      "loss": 0.3222,
+      "step": 12140
+    },
+    {
+      "epoch": 0.5331459872443659,
+      "grad_norm": 1.640625,
+      "learning_rate": 2.251181456927903e-05,
+      "loss": 0.3759,
+      "step": 12142
+    },
+    {
+      "epoch": 0.5332338057235694,
+      "grad_norm": 1.8671875,
+      "learning_rate": 2.2504930731407252e-05,
+      "loss": 0.3332,
+      "step": 12144
+    },
+    {
+      "epoch": 0.5333216242027728,
+      "grad_norm": 1.75,
+      "learning_rate": 2.2498047084608477e-05,
+      "loss": 0.343,
+      "step": 12146
+    },
+    {
+      "epoch": 0.5334094426819763,
+      "grad_norm": 1.6953125,
+      "learning_rate": 2.2491163629409852e-05,
+      "loss": 0.3348,
+      "step": 12148
+    },
+    {
+      "epoch": 0.5334972611611798,
+      "grad_norm": 1.6640625,
+      "learning_rate": 2.2484280366338527e-05,
+      "loss": 0.3569,
+      "step": 12150
+    },
+    {
+      "epoch": 0.5335850796403834,
+      "grad_norm": 1.71875,
+      "learning_rate": 2.247739729592161e-05,
+      "loss": 0.357,
+      "step": 12152
+    },
+    {
+      "epoch": 0.5336728981195868,
+      "grad_norm": 1.6875,
+      "learning_rate": 2.247051441868621e-05,
+      "loss": 0.3752,
+      "step": 12154
+    },
+    {
+      "epoch": 0.5337607165987903,
+      "grad_norm": 1.6328125,
+      "learning_rate": 2.2463631735159422e-05,
+      "loss": 0.3608,
+      "step": 12156
+    },
+    {
+      "epoch": 0.5338485350779938,
+      "grad_norm": 1.5546875,
+      "learning_rate": 2.2456749245868323e-05,
+      "loss": 0.3554,
+      "step": 12158
+    },
+    {
+      "epoch": 0.5339363535571973,
+      "grad_norm": 1.6328125,
+      "learning_rate": 2.2449866951339976e-05,
+      "loss": 0.3527,
+      "step": 12160
+    },
+    {
+      "epoch": 0.5340241720364007,
+      "grad_norm": 1.8671875,
+      "learning_rate": 2.2442984852101435e-05,
+      "loss": 0.3513,
+      "step": 12162
+    },
+    {
+      "epoch": 0.5341119905156042,
+      "grad_norm": 1.546875,
+      "learning_rate": 2.2436102948679722e-05,
+      "loss": 0.3651,
+      "step": 12164
+    },
+    {
+      "epoch": 0.5341998089948077,
+      "grad_norm": 1.671875,
+      "learning_rate": 2.242922124160186e-05,
+      "loss": 0.3336,
+      "step": 12166
+    },
+    {
+      "epoch": 0.5342876274740113,
+      "grad_norm": 1.859375,
+      "learning_rate": 2.2422339731394848e-05,
+      "loss": 0.3534,
+      "step": 12168
+    },
+    {
+      "epoch": 0.5343754459532147,
+      "grad_norm": 1.8125,
+      "learning_rate": 2.2415458418585682e-05,
+      "loss": 0.3558,
+      "step": 12170
+    },
+    {
+      "epoch": 0.5344632644324182,
+      "grad_norm": 1.8671875,
+      "learning_rate": 2.2408577303701334e-05,
+      "loss": 0.3285,
+      "step": 12172
+    },
+    {
+      "epoch": 0.5345510829116217,
+      "grad_norm": 1.9453125,
+      "learning_rate": 2.240169638726875e-05,
+      "loss": 0.3693,
+      "step": 12174
+    },
+    {
+      "epoch": 0.5346389013908251,
+      "grad_norm": 1.8515625,
+      "learning_rate": 2.239481566981488e-05,
+      "loss": 0.3854,
+      "step": 12176
+    },
+    {
+      "epoch": 0.5347267198700286,
+      "grad_norm": 1.7578125,
+      "learning_rate": 2.2387935151866653e-05,
+      "loss": 0.3369,
+      "step": 12178
+    },
+    {
+      "epoch": 0.5348145383492321,
+      "grad_norm": 1.9609375,
+      "learning_rate": 2.2381054833950976e-05,
+      "loss": 0.3574,
+      "step": 12180
+    },
+    {
+      "epoch": 0.5349023568284357,
+      "grad_norm": 1.8671875,
+      "learning_rate": 2.2374174716594758e-05,
+      "loss": 0.3278,
+      "step": 12182
+    },
+    {
+      "epoch": 0.5349901753076391,
+      "grad_norm": 1.6484375,
+      "learning_rate": 2.2367294800324862e-05,
+      "loss": 0.3384,
+      "step": 12184
+    },
+    {
+      "epoch": 0.5350779937868426,
+      "grad_norm": 1.8671875,
+      "learning_rate": 2.236041508566816e-05,
+      "loss": 0.3491,
+      "step": 12186
+    },
+    {
+      "epoch": 0.5351658122660461,
+      "grad_norm": 1.734375,
+      "learning_rate": 2.2353535573151506e-05,
+      "loss": 0.3535,
+      "step": 12188
+    },
+    {
+      "epoch": 0.5352536307452496,
+      "grad_norm": 1.84375,
+      "learning_rate": 2.234665626330173e-05,
+      "loss": 0.3765,
+      "step": 12190
+    },
+    {
+      "epoch": 0.535341449224453,
+      "grad_norm": 1.765625,
+      "learning_rate": 2.233977715664565e-05,
+      "loss": 0.3506,
+      "step": 12192
+    },
+    {
+      "epoch": 0.5354292677036565,
+      "grad_norm": 1.578125,
+      "learning_rate": 2.2332898253710077e-05,
+      "loss": 0.3636,
+      "step": 12194
+    },
+    {
+      "epoch": 0.53551708618286,
+      "grad_norm": 1.6796875,
+      "learning_rate": 2.2326019555021793e-05,
+      "loss": 0.331,
+      "step": 12196
+    },
+    {
+      "epoch": 0.5356049046620636,
+      "grad_norm": 1.6484375,
+      "learning_rate": 2.231914106110758e-05,
+      "loss": 0.3456,
+      "step": 12198
+    },
+    {
+      "epoch": 0.535692723141267,
+      "grad_norm": 1.8203125,
+      "learning_rate": 2.231226277249418e-05,
+      "loss": 0.3505,
+      "step": 12200
+    },
+    {
+      "epoch": 0.5357805416204705,
+      "grad_norm": 1.90625,
+      "learning_rate": 2.2305384689708342e-05,
+      "loss": 0.3472,
+      "step": 12202
+    },
+    {
+      "epoch": 0.535868360099674,
+      "grad_norm": 1.84375,
+      "learning_rate": 2.229850681327679e-05,
+      "loss": 0.3306,
+      "step": 12204
+    },
+    {
+      "epoch": 0.5359561785788775,
+      "grad_norm": 1.9375,
+      "learning_rate": 2.2291629143726232e-05,
+      "loss": 0.3345,
+      "step": 12206
+    },
+    {
+      "epoch": 0.5360439970580809,
+      "grad_norm": 1.6328125,
+      "learning_rate": 2.228475168158337e-05,
+      "loss": 0.3393,
+      "step": 12208
+    },
+    {
+      "epoch": 0.5361318155372844,
+      "grad_norm": 1.671875,
+      "learning_rate": 2.227787442737487e-05,
+      "loss": 0.327,
+      "step": 12210
+    },
+    {
+      "epoch": 0.5362196340164879,
+      "grad_norm": 1.8828125,
+      "learning_rate": 2.2270997381627407e-05,
+      "loss": 0.3662,
+      "step": 12212
+    },
+    {
+      "epoch": 0.5363074524956915,
+      "grad_norm": 1.6640625,
+      "learning_rate": 2.2264120544867615e-05,
+      "loss": 0.365,
+      "step": 12214
+    },
+    {
+      "epoch": 0.5363952709748949,
+      "grad_norm": 1.7265625,
+      "learning_rate": 2.2257243917622124e-05,
+      "loss": 0.3556,
+      "step": 12216
+    },
+    {
+      "epoch": 0.5364830894540984,
+      "grad_norm": 1.6875,
+      "learning_rate": 2.2250367500417547e-05,
+      "loss": 0.3615,
+      "step": 12218
+    },
+    {
+      "epoch": 0.5365709079333019,
+      "grad_norm": 2.53125,
+      "learning_rate": 2.2243491293780493e-05,
+      "loss": 0.3571,
+      "step": 12220
+    },
+    {
+      "epoch": 0.5366587264125053,
+      "grad_norm": 1.6875,
+      "learning_rate": 2.223661529823753e-05,
+      "loss": 0.3442,
+      "step": 12222
+    },
+    {
+      "epoch": 0.5367465448917088,
+      "grad_norm": 1.984375,
+      "learning_rate": 2.222973951431524e-05,
+      "loss": 0.3453,
+      "step": 12224
+    },
+    {
+      "epoch": 0.5368343633709123,
+      "grad_norm": 1.796875,
+      "learning_rate": 2.2222863942540153e-05,
+      "loss": 0.3537,
+      "step": 12226
+    },
+    {
+      "epoch": 0.5369221818501159,
+      "grad_norm": 1.8046875,
+      "learning_rate": 2.2215988583438814e-05,
+      "loss": 0.3566,
+      "step": 12228
+    },
+    {
+      "epoch": 0.5370100003293193,
+      "grad_norm": 1.796875,
+      "learning_rate": 2.2209113437537738e-05,
+      "loss": 0.3607,
+      "step": 12230
+    },
+    {
+      "epoch": 0.5370978188085228,
+      "grad_norm": 1.75,
+      "learning_rate": 2.220223850536342e-05,
+      "loss": 0.3418,
+      "step": 12232
+    },
+    {
+      "epoch": 0.5371856372877263,
+      "grad_norm": 1.625,
+      "learning_rate": 2.2195363787442356e-05,
+      "loss": 0.3311,
+      "step": 12234
+    },
+    {
+      "epoch": 0.5372734557669298,
+      "grad_norm": 1.8984375,
+      "learning_rate": 2.2188489284300997e-05,
+      "loss": 0.3293,
+      "step": 12236
+    },
+    {
+      "epoch": 0.5373612742461332,
+      "grad_norm": 1.671875,
+      "learning_rate": 2.2181614996465805e-05,
+      "loss": 0.3569,
+      "step": 12238
+    },
+    {
+      "epoch": 0.5374490927253367,
+      "grad_norm": 2.015625,
+      "learning_rate": 2.217474092446321e-05,
+      "loss": 0.3676,
+      "step": 12240
+    },
+    {
+      "epoch": 0.5375369112045402,
+      "grad_norm": 1.78125,
+      "learning_rate": 2.2167867068819646e-05,
+      "loss": 0.3687,
+      "step": 12242
+    },
+    {
+      "epoch": 0.5376247296837438,
+      "grad_norm": 1.7421875,
+      "learning_rate": 2.2160993430061488e-05,
+      "loss": 0.3539,
+      "step": 12244
+    },
+    {
+      "epoch": 0.5377125481629472,
+      "grad_norm": 1.703125,
+      "learning_rate": 2.2154120008715135e-05,
+      "loss": 0.3229,
+      "step": 12246
+    },
+    {
+      "epoch": 0.5378003666421507,
+      "grad_norm": 1.8203125,
+      "learning_rate": 2.2147246805306955e-05,
+      "loss": 0.3524,
+      "step": 12248
+    },
+    {
+      "epoch": 0.5378881851213542,
+      "grad_norm": 1.921875,
+      "learning_rate": 2.21403738203633e-05,
+      "loss": 0.3487,
+      "step": 12250
+    },
+    {
+      "epoch": 0.5379760036005576,
+      "grad_norm": 1.8125,
+      "learning_rate": 2.2133501054410503e-05,
+      "loss": 0.3325,
+      "step": 12252
+    },
+    {
+      "epoch": 0.5380638220797611,
+      "grad_norm": 1.6875,
+      "learning_rate": 2.2126628507974877e-05,
+      "loss": 0.315,
+      "step": 12254
+    },
+    {
+      "epoch": 0.5381516405589646,
+      "grad_norm": 1.7421875,
+      "learning_rate": 2.2119756181582733e-05,
+      "loss": 0.3404,
+      "step": 12256
+    },
+    {
+      "epoch": 0.538239459038168,
+      "grad_norm": 1.6484375,
+      "learning_rate": 2.2112884075760347e-05,
+      "loss": 0.3195,
+      "step": 12258
+    },
+    {
+      "epoch": 0.5383272775173716,
+      "grad_norm": 1.75,
+      "learning_rate": 2.2106012191033998e-05,
+      "loss": 0.3416,
+      "step": 12260
+    },
+    {
+      "epoch": 0.5384150959965751,
+      "grad_norm": 1.6875,
+      "learning_rate": 2.2099140527929926e-05,
+      "loss": 0.3675,
+      "step": 12262
+    },
+    {
+      "epoch": 0.5385029144757786,
+      "grad_norm": 1.6015625,
+      "learning_rate": 2.2092269086974367e-05,
+      "loss": 0.3592,
+      "step": 12264
+    },
+    {
+      "epoch": 0.538590732954982,
+      "grad_norm": 1.5859375,
+      "learning_rate": 2.2085397868693537e-05,
+      "loss": 0.3565,
+      "step": 12266
+    },
+    {
+      "epoch": 0.5386785514341855,
+      "grad_norm": 1.875,
+      "learning_rate": 2.2078526873613637e-05,
+      "loss": 0.3504,
+      "step": 12268
+    },
+    {
+      "epoch": 0.538766369913389,
+      "grad_norm": 1.859375,
+      "learning_rate": 2.207165610226085e-05,
+      "loss": 0.3347,
+      "step": 12270
+    },
+    {
+      "epoch": 0.5388541883925925,
+      "grad_norm": 1.71875,
+      "learning_rate": 2.2064785555161343e-05,
+      "loss": 0.3376,
+      "step": 12272
+    },
+    {
+      "epoch": 0.5389420068717959,
+      "grad_norm": 2.015625,
+      "learning_rate": 2.205791523284127e-05,
+      "loss": 0.366,
+      "step": 12274
+    },
+    {
+      "epoch": 0.5390298253509995,
+      "grad_norm": 2.3125,
+      "learning_rate": 2.2051045135826743e-05,
+      "loss": 0.3163,
+      "step": 12276
+    },
+    {
+      "epoch": 0.539117643830203,
+      "grad_norm": 1.6953125,
+      "learning_rate": 2.2044175264643884e-05,
+      "loss": 0.3666,
+      "step": 12278
+    },
+    {
+      "epoch": 0.5392054623094065,
+      "grad_norm": 1.8671875,
+      "learning_rate": 2.2037305619818792e-05,
+      "loss": 0.3586,
+      "step": 12280
+    },
+    {
+      "epoch": 0.5392932807886099,
+      "grad_norm": 2.25,
+      "learning_rate": 2.203043620187755e-05,
+      "loss": 0.3081,
+      "step": 12282
+    },
+    {
+      "epoch": 0.5393810992678134,
+      "grad_norm": 1.984375,
+      "learning_rate": 2.2023567011346213e-05,
+      "loss": 0.3113,
+      "step": 12284
+    },
+    {
+      "epoch": 0.5394689177470169,
+      "grad_norm": 2.140625,
+      "learning_rate": 2.2016698048750832e-05,
+      "loss": 0.339,
+      "step": 12286
+    },
+    {
+      "epoch": 0.5395567362262204,
+      "grad_norm": 1.6953125,
+      "learning_rate": 2.200982931461743e-05,
+      "loss": 0.3853,
+      "step": 12288
+    },
+    {
+      "epoch": 0.5396445547054239,
+      "grad_norm": 2.09375,
+      "learning_rate": 2.200296080947201e-05,
+      "loss": 0.3577,
+      "step": 12290
+    },
+    {
+      "epoch": 0.5397323731846274,
+      "grad_norm": 1.6484375,
+      "learning_rate": 2.199609253384057e-05,
+      "loss": 0.3771,
+      "step": 12292
+    },
+    {
+      "epoch": 0.5398201916638309,
+      "grad_norm": 1.71875,
+      "learning_rate": 2.198922448824908e-05,
+      "loss": 0.342,
+      "step": 12294
+    },
+    {
+      "epoch": 0.5399080101430344,
+      "grad_norm": 1.9765625,
+      "learning_rate": 2.1982356673223516e-05,
+      "loss": 0.3599,
+      "step": 12296
+    },
+    {
+      "epoch": 0.5399958286222378,
+      "grad_norm": 1.7890625,
+      "learning_rate": 2.1975489089289788e-05,
+      "loss": 0.3807,
+      "step": 12298
+    },
+    {
+      "epoch": 0.5400836471014413,
+      "grad_norm": 1.75,
+      "learning_rate": 2.1968621736973836e-05,
+      "loss": 0.3404,
+      "step": 12300
+    },
+    {
+      "epoch": 0.5401714655806448,
+      "grad_norm": 1.765625,
+      "learning_rate": 2.1961754616801553e-05,
+      "loss": 0.3492,
+      "step": 12302
+    },
+    {
+      "epoch": 0.5402592840598482,
+      "grad_norm": 1.5703125,
+      "learning_rate": 2.1954887729298844e-05,
+      "loss": 0.3583,
+      "step": 12304
+    },
+    {
+      "epoch": 0.5403471025390518,
+      "grad_norm": 1.65625,
+      "learning_rate": 2.1948021074991552e-05,
+      "loss": 0.3626,
+      "step": 12306
+    },
+    {
+      "epoch": 0.5404349210182553,
+      "grad_norm": 2.015625,
+      "learning_rate": 2.1941154654405536e-05,
+      "loss": 0.3354,
+      "step": 12308
+    },
+    {
+      "epoch": 0.5405227394974588,
+      "grad_norm": 1.6796875,
+      "learning_rate": 2.1934288468066627e-05,
+      "loss": 0.3288,
+      "step": 12310
+    },
+    {
+      "epoch": 0.5406105579766622,
+      "grad_norm": 1.6484375,
+      "learning_rate": 2.192742251650065e-05,
+      "loss": 0.3443,
+      "step": 12312
+    },
+    {
+      "epoch": 0.5406983764558657,
+      "grad_norm": 1.8671875,
+      "learning_rate": 2.1920556800233392e-05,
+      "loss": 0.3248,
+      "step": 12314
+    },
+    {
+      "epoch": 0.5407861949350692,
+      "grad_norm": 1.75,
+      "learning_rate": 2.1913691319790627e-05,
+      "loss": 0.3393,
+      "step": 12316
+    },
+    {
+      "epoch": 0.5408740134142727,
+      "grad_norm": 1.640625,
+      "learning_rate": 2.190682607569812e-05,
+      "loss": 0.3671,
+      "step": 12318
+    },
+    {
+      "epoch": 0.5409618318934761,
+      "grad_norm": 1.6796875,
+      "learning_rate": 2.189996106848162e-05,
+      "loss": 0.3408,
+      "step": 12320
+    },
+    {
+      "epoch": 0.5410496503726797,
+      "grad_norm": 1.7265625,
+      "learning_rate": 2.1893096298666844e-05,
+      "loss": 0.3222,
+      "step": 12322
+    },
+    {
+      "epoch": 0.5411374688518832,
+      "grad_norm": 1.7109375,
+      "learning_rate": 2.1886231766779495e-05,
+      "loss": 0.3502,
+      "step": 12324
+    },
+    {
+      "epoch": 0.5412252873310867,
+      "grad_norm": 1.6875,
+      "learning_rate": 2.1879367473345263e-05,
+      "loss": 0.3585,
+      "step": 12326
+    },
+    {
+      "epoch": 0.5413131058102901,
+      "grad_norm": 1.640625,
+      "learning_rate": 2.187250341888982e-05,
+      "loss": 0.3369,
+      "step": 12328
+    },
+    {
+      "epoch": 0.5414009242894936,
+      "grad_norm": 1.828125,
+      "learning_rate": 2.186563960393881e-05,
+      "loss": 0.3656,
+      "step": 12330
+    },
+    {
+      "epoch": 0.5414887427686971,
+      "grad_norm": 1.78125,
+      "learning_rate": 2.185877602901788e-05,
+      "loss": 0.3492,
+      "step": 12332
+    },
+    {
+      "epoch": 0.5415765612479005,
+      "grad_norm": 1.75,
+      "learning_rate": 2.1851912694652634e-05,
+      "loss": 0.3264,
+      "step": 12334
+    },
+    {
+      "epoch": 0.5416643797271041,
+      "grad_norm": 1.71875,
+      "learning_rate": 2.184504960136867e-05,
+      "loss": 0.3433,
+      "step": 12336
+    },
+    {
+      "epoch": 0.5417521982063076,
+      "grad_norm": 1.859375,
+      "learning_rate": 2.1838186749691557e-05,
+      "loss": 0.3419,
+      "step": 12338
+    },
+    {
+      "epoch": 0.5418400166855111,
+      "grad_norm": 2.03125,
+      "learning_rate": 2.183132414014686e-05,
+      "loss": 0.3602,
+      "step": 12340
+    },
+    {
+      "epoch": 0.5419278351647145,
+      "grad_norm": 1.7890625,
+      "learning_rate": 2.1824461773260122e-05,
+      "loss": 0.3398,
+      "step": 12342
+    },
+    {
+      "epoch": 0.542015653643918,
+      "grad_norm": 1.671875,
+      "learning_rate": 2.1817599649556858e-05,
+      "loss": 0.3386,
+      "step": 12344
+    },
+    {
+      "epoch": 0.5421034721231215,
+      "grad_norm": 1.671875,
+      "learning_rate": 2.181073776956258e-05,
+      "loss": 0.3365,
+      "step": 12346
+    },
+    {
+      "epoch": 0.542191290602325,
+      "grad_norm": 1.9453125,
+      "learning_rate": 2.1803876133802773e-05,
+      "loss": 0.316,
+      "step": 12348
+    },
+    {
+      "epoch": 0.5422791090815284,
+      "grad_norm": 1.9140625,
+      "learning_rate": 2.179701474280289e-05,
+      "loss": 0.357,
+      "step": 12350
+    },
+    {
+      "epoch": 0.542366927560732,
+      "grad_norm": 1.9375,
+      "learning_rate": 2.179015359708839e-05,
+      "loss": 0.3212,
+      "step": 12352
+    },
+    {
+      "epoch": 0.5424547460399355,
+      "grad_norm": 1.6953125,
+      "learning_rate": 2.1783292697184694e-05,
+      "loss": 0.3282,
+      "step": 12354
+    },
+    {
+      "epoch": 0.542542564519139,
+      "grad_norm": 1.890625,
+      "learning_rate": 2.1776432043617214e-05,
+      "loss": 0.3623,
+      "step": 12356
+    },
+    {
+      "epoch": 0.5426303829983424,
+      "grad_norm": 1.7578125,
+      "learning_rate": 2.1769571636911347e-05,
+      "loss": 0.3273,
+      "step": 12358
+    },
+    {
+      "epoch": 0.5427182014775459,
+      "grad_norm": 1.6171875,
+      "learning_rate": 2.1762711477592454e-05,
+      "loss": 0.3147,
+      "step": 12360
+    },
+    {
+      "epoch": 0.5428060199567494,
+      "grad_norm": 1.9140625,
+      "learning_rate": 2.1755851566185888e-05,
+      "loss": 0.3477,
+      "step": 12362
+    },
+    {
+      "epoch": 0.5428938384359528,
+      "grad_norm": 2.125,
+      "learning_rate": 2.174899190321699e-05,
+      "loss": 0.3381,
+      "step": 12364
+    },
+    {
+      "epoch": 0.5429816569151563,
+      "grad_norm": 1.7578125,
+      "learning_rate": 2.1742132489211082e-05,
+      "loss": 0.35,
+      "step": 12366
+    },
+    {
+      "epoch": 0.5430694753943599,
+      "grad_norm": 1.6328125,
+      "learning_rate": 2.173527332469344e-05,
+      "loss": 0.3457,
+      "step": 12368
+    },
+    {
+      "epoch": 0.5431572938735634,
+      "grad_norm": 1.8515625,
+      "learning_rate": 2.1728414410189347e-05,
+      "loss": 0.3632,
+      "step": 12370
+    },
+    {
+      "epoch": 0.5432451123527668,
+      "grad_norm": 1.859375,
+      "learning_rate": 2.172155574622407e-05,
+      "loss": 0.3292,
+      "step": 12372
+    },
+    {
+      "epoch": 0.5433329308319703,
+      "grad_norm": 1.734375,
+      "learning_rate": 2.1714697333322832e-05,
+      "loss": 0.3531,
+      "step": 12374
+    },
+    {
+      "epoch": 0.5434207493111738,
+      "grad_norm": 1.828125,
+      "learning_rate": 2.1707839172010862e-05,
+      "loss": 0.3582,
+      "step": 12376
+    },
+    {
+      "epoch": 0.5435085677903773,
+      "grad_norm": 1.703125,
+      "learning_rate": 2.1700981262813357e-05,
+      "loss": 0.3639,
+      "step": 12378
+    },
+    {
+      "epoch": 0.5435963862695807,
+      "grad_norm": 1.734375,
+      "learning_rate": 2.1694123606255496e-05,
+      "loss": 0.3343,
+      "step": 12380
+    },
+    {
+      "epoch": 0.5436842047487843,
+      "grad_norm": 1.8046875,
+      "learning_rate": 2.1687266202862445e-05,
+      "loss": 0.3665,
+      "step": 12382
+    },
+    {
+      "epoch": 0.5437720232279878,
+      "grad_norm": 1.75,
+      "learning_rate": 2.1680409053159345e-05,
+      "loss": 0.3628,
+      "step": 12384
+    },
+    {
+      "epoch": 0.5438598417071913,
+      "grad_norm": 1.703125,
+      "learning_rate": 2.1673552157671308e-05,
+      "loss": 0.3443,
+      "step": 12386
+    },
+    {
+      "epoch": 0.5439476601863947,
+      "grad_norm": 1.640625,
+      "learning_rate": 2.1666695516923445e-05,
+      "loss": 0.3487,
+      "step": 12388
+    },
+    {
+      "epoch": 0.5440354786655982,
+      "grad_norm": 1.7421875,
+      "learning_rate": 2.1659839131440844e-05,
+      "loss": 0.3419,
+      "step": 12390
+    },
+    {
+      "epoch": 0.5441232971448017,
+      "grad_norm": 1.65625,
+      "learning_rate": 2.1652983001748558e-05,
+      "loss": 0.3457,
+      "step": 12392
+    },
+    {
+      "epoch": 0.5442111156240051,
+      "grad_norm": 1.796875,
+      "learning_rate": 2.1646127128371644e-05,
+      "loss": 0.3612,
+      "step": 12394
+    },
+    {
+      "epoch": 0.5442989341032086,
+      "grad_norm": 1.6484375,
+      "learning_rate": 2.1639271511835117e-05,
+      "loss": 0.3232,
+      "step": 12396
+    },
+    {
+      "epoch": 0.5443867525824122,
+      "grad_norm": 1.8359375,
+      "learning_rate": 2.163241615266398e-05,
+      "loss": 0.3363,
+      "step": 12398
+    },
+    {
+      "epoch": 0.5444745710616157,
+      "grad_norm": 1.765625,
+      "learning_rate": 2.162556105138322e-05,
+      "loss": 0.3479,
+      "step": 12400
+    },
+    {
+      "epoch": 0.5445623895408191,
+      "grad_norm": 1.8203125,
+      "learning_rate": 2.1618706208517804e-05,
+      "loss": 0.3632,
+      "step": 12402
+    },
+    {
+      "epoch": 0.5446502080200226,
+      "grad_norm": 1.78125,
+      "learning_rate": 2.1611851624592678e-05,
+      "loss": 0.3729,
+      "step": 12404
+    },
+    {
+      "epoch": 0.5447380264992261,
+      "grad_norm": 1.6953125,
+      "learning_rate": 2.160499730013276e-05,
+      "loss": 0.3483,
+      "step": 12406
+    },
+    {
+      "epoch": 0.5448258449784296,
+      "grad_norm": 1.6875,
+      "learning_rate": 2.1598143235662977e-05,
+      "loss": 0.3499,
+      "step": 12408
+    },
+    {
+      "epoch": 0.544913663457633,
+      "grad_norm": 1.7734375,
+      "learning_rate": 2.1591289431708188e-05,
+      "loss": 0.3555,
+      "step": 12410
+    },
+    {
+      "epoch": 0.5450014819368365,
+      "grad_norm": 1.765625,
+      "learning_rate": 2.1584435888793276e-05,
+      "loss": 0.3504,
+      "step": 12412
+    },
+    {
+      "epoch": 0.5450893004160401,
+      "grad_norm": 1.9765625,
+      "learning_rate": 2.157758260744308e-05,
+      "loss": 0.34,
+      "step": 12414
+    },
+    {
+      "epoch": 0.5451771188952436,
+      "grad_norm": 1.84375,
+      "learning_rate": 2.157072958818243e-05,
+      "loss": 0.3406,
+      "step": 12416
+    },
+    {
+      "epoch": 0.545264937374447,
+      "grad_norm": 1.84375,
+      "learning_rate": 2.1563876831536127e-05,
+      "loss": 0.3288,
+      "step": 12418
+    },
+    {
+      "epoch": 0.5453527558536505,
+      "grad_norm": 1.8984375,
+      "learning_rate": 2.1557024338028968e-05,
+      "loss": 0.3326,
+      "step": 12420
+    },
+    {
+      "epoch": 0.545440574332854,
+      "grad_norm": 1.78125,
+      "learning_rate": 2.1550172108185703e-05,
+      "loss": 0.3643,
+      "step": 12422
+    },
+    {
+      "epoch": 0.5455283928120574,
+      "grad_norm": 1.7734375,
+      "learning_rate": 2.1543320142531087e-05,
+      "loss": 0.368,
+      "step": 12424
+    },
+    {
+      "epoch": 0.5456162112912609,
+      "grad_norm": 1.6640625,
+      "learning_rate": 2.1536468441589855e-05,
+      "loss": 0.3537,
+      "step": 12426
+    },
+    {
+      "epoch": 0.5457040297704645,
+      "grad_norm": 1.8359375,
+      "learning_rate": 2.1529617005886687e-05,
+      "loss": 0.3417,
+      "step": 12428
+    },
+    {
+      "epoch": 0.545791848249668,
+      "grad_norm": 1.921875,
+      "learning_rate": 2.1522765835946283e-05,
+      "loss": 0.3658,
+      "step": 12430
+    },
+    {
+      "epoch": 0.5458796667288714,
+      "grad_norm": 1.765625,
+      "learning_rate": 2.1515914932293302e-05,
+      "loss": 0.3662,
+      "step": 12432
+    },
+    {
+      "epoch": 0.5459674852080749,
+      "grad_norm": 1.8125,
+      "learning_rate": 2.1509064295452395e-05,
+      "loss": 0.3521,
+      "step": 12434
+    },
+    {
+      "epoch": 0.5460553036872784,
+      "grad_norm": 1.8125,
+      "learning_rate": 2.150221392594818e-05,
+      "loss": 0.3552,
+      "step": 12436
+    },
+    {
+      "epoch": 0.5461431221664819,
+      "grad_norm": 1.8984375,
+      "learning_rate": 2.149536382430526e-05,
+      "loss": 0.333,
+      "step": 12438
+    },
+    {
+      "epoch": 0.5462309406456853,
+      "grad_norm": 1.875,
+      "learning_rate": 2.1488513991048215e-05,
+      "loss": 0.3332,
+      "step": 12440
+    },
+    {
+      "epoch": 0.5463187591248888,
+      "grad_norm": 1.7421875,
+      "learning_rate": 2.1481664426701615e-05,
+      "loss": 0.3674,
+      "step": 12442
+    },
+    {
+      "epoch": 0.5464065776040924,
+      "grad_norm": 1.6875,
+      "learning_rate": 2.1474815131789993e-05,
+      "loss": 0.361,
+      "step": 12444
+    },
+    {
+      "epoch": 0.5464943960832959,
+      "grad_norm": 1.6953125,
+      "learning_rate": 2.1467966106837877e-05,
+      "loss": 0.3478,
+      "step": 12446
+    },
+    {
+      "epoch": 0.5465822145624993,
+      "grad_norm": 1.96875,
+      "learning_rate": 2.1461117352369765e-05,
+      "loss": 0.3433,
+      "step": 12448
+    },
+    {
+      "epoch": 0.5466700330417028,
+      "grad_norm": 1.75,
+      "learning_rate": 2.1454268868910126e-05,
+      "loss": 0.3391,
+      "step": 12450
+    },
+    {
+      "epoch": 0.5467578515209063,
+      "grad_norm": 1.9765625,
+      "learning_rate": 2.1447420656983432e-05,
+      "loss": 0.3514,
+      "step": 12452
+    },
+    {
+      "epoch": 0.5468456700001098,
+      "grad_norm": 1.671875,
+      "learning_rate": 2.1440572717114115e-05,
+      "loss": 0.35,
+      "step": 12454
+    },
+    {
+      "epoch": 0.5469334884793132,
+      "grad_norm": 1.75,
+      "learning_rate": 2.1433725049826597e-05,
+      "loss": 0.3498,
+      "step": 12456
+    },
+    {
+      "epoch": 0.5470213069585167,
+      "grad_norm": 1.625,
+      "learning_rate": 2.1426877655645274e-05,
+      "loss": 0.3254,
+      "step": 12458
+    },
+    {
+      "epoch": 0.5471091254377203,
+      "grad_norm": 1.796875,
+      "learning_rate": 2.142003053509451e-05,
+      "loss": 0.3818,
+      "step": 12460
+    },
+    {
+      "epoch": 0.5471969439169238,
+      "grad_norm": 1.8125,
+      "learning_rate": 2.1413183688698664e-05,
+      "loss": 0.3622,
+      "step": 12462
+    },
+    {
+      "epoch": 0.5472847623961272,
+      "grad_norm": 1.953125,
+      "learning_rate": 2.1406337116982074e-05,
+      "loss": 0.3316,
+      "step": 12464
+    },
+    {
+      "epoch": 0.5473725808753307,
+      "grad_norm": 1.671875,
+      "learning_rate": 2.1399490820469042e-05,
+      "loss": 0.3456,
+      "step": 12466
+    },
+    {
+      "epoch": 0.5474603993545342,
+      "grad_norm": 1.765625,
+      "learning_rate": 2.139264479968387e-05,
+      "loss": 0.3177,
+      "step": 12468
+    },
+    {
+      "epoch": 0.5475482178337376,
+      "grad_norm": 1.6875,
+      "learning_rate": 2.138579905515083e-05,
+      "loss": 0.3627,
+      "step": 12470
+    },
+    {
+      "epoch": 0.5476360363129411,
+      "grad_norm": 1.640625,
+      "learning_rate": 2.137895358739416e-05,
+      "loss": 0.3348,
+      "step": 12472
+    },
+    {
+      "epoch": 0.5477238547921446,
+      "grad_norm": 1.6875,
+      "learning_rate": 2.137210839693809e-05,
+      "loss": 0.3557,
+      "step": 12474
+    },
+    {
+      "epoch": 0.5478116732713482,
+      "grad_norm": 1.8984375,
+      "learning_rate": 2.1365263484306828e-05,
+      "loss": 0.3701,
+      "step": 12476
+    },
+    {
+      "epoch": 0.5478994917505516,
+      "grad_norm": 1.859375,
+      "learning_rate": 2.135841885002456e-05,
+      "loss": 0.3403,
+      "step": 12478
+    },
+    {
+      "epoch": 0.5479873102297551,
+      "grad_norm": 1.7265625,
+      "learning_rate": 2.135157449461545e-05,
+      "loss": 0.3498,
+      "step": 12480
+    },
+    {
+      "epoch": 0.5480751287089586,
+      "grad_norm": 1.6015625,
+      "learning_rate": 2.1344730418603646e-05,
+      "loss": 0.3349,
+      "step": 12482
+    },
+    {
+      "epoch": 0.548162947188162,
+      "grad_norm": 1.8203125,
+      "learning_rate": 2.1337886622513257e-05,
+      "loss": 0.3614,
+      "step": 12484
+    },
+    {
+      "epoch": 0.5482507656673655,
+      "grad_norm": 1.921875,
+      "learning_rate": 2.1331043106868386e-05,
+      "loss": 0.3407,
+      "step": 12486
+    },
+    {
+      "epoch": 0.548338584146569,
+      "grad_norm": 1.703125,
+      "learning_rate": 2.1324199872193128e-05,
+      "loss": 0.3853,
+      "step": 12488
+    },
+    {
+      "epoch": 0.5484264026257726,
+      "grad_norm": 1.6328125,
+      "learning_rate": 2.1317356919011513e-05,
+      "loss": 0.3742,
+      "step": 12490
+    },
+    {
+      "epoch": 0.548514221104976,
+      "grad_norm": 1.6171875,
+      "learning_rate": 2.1310514247847587e-05,
+      "loss": 0.3465,
+      "step": 12492
+    },
+    {
+      "epoch": 0.5486020395841795,
+      "grad_norm": 1.8046875,
+      "learning_rate": 2.1303671859225364e-05,
+      "loss": 0.3242,
+      "step": 12494
+    },
+    {
+      "epoch": 0.548689858063383,
+      "grad_norm": 1.90625,
+      "learning_rate": 2.1296829753668844e-05,
+      "loss": 0.3473,
+      "step": 12496
+    },
+    {
+      "epoch": 0.5487776765425865,
+      "grad_norm": 1.7265625,
+      "learning_rate": 2.1289987931701983e-05,
+      "loss": 0.3608,
+      "step": 12498
+    },
+    {
+      "epoch": 0.5488654950217899,
+      "grad_norm": 1.4921875,
+      "learning_rate": 2.1283146393848733e-05,
+      "loss": 0.3289,
+      "step": 12500
+    },
+    {
+      "epoch": 0.5489533135009934,
+      "grad_norm": 1.6796875,
+      "learning_rate": 2.1276305140633024e-05,
+      "loss": 0.349,
+      "step": 12502
+    },
+    {
+      "epoch": 0.5490411319801969,
+      "grad_norm": 1.640625,
+      "learning_rate": 2.1269464172578764e-05,
+      "loss": 0.3402,
+      "step": 12504
+    },
+    {
+      "epoch": 0.5491289504594005,
+      "grad_norm": 1.890625,
+      "learning_rate": 2.1262623490209834e-05,
+      "loss": 0.3441,
+      "step": 12506
+    },
+    {
+      "epoch": 0.5492167689386039,
+      "grad_norm": 1.734375,
+      "learning_rate": 2.1255783094050087e-05,
+      "loss": 0.3782,
+      "step": 12508
+    },
+    {
+      "epoch": 0.5493045874178074,
+      "grad_norm": 1.6796875,
+      "learning_rate": 2.1248942984623367e-05,
+      "loss": 0.3275,
+      "step": 12510
+    },
+    {
+      "epoch": 0.5493924058970109,
+      "grad_norm": 1.8671875,
+      "learning_rate": 2.1242103162453495e-05,
+      "loss": 0.3401,
+      "step": 12512
+    },
+    {
+      "epoch": 0.5494802243762144,
+      "grad_norm": 1.8359375,
+      "learning_rate": 2.123526362806426e-05,
+      "loss": 0.3422,
+      "step": 12514
+    },
+    {
+      "epoch": 0.5495680428554178,
+      "grad_norm": 1.703125,
+      "learning_rate": 2.1228424381979446e-05,
+      "loss": 0.3726,
+      "step": 12516
+    },
+    {
+      "epoch": 0.5496558613346213,
+      "grad_norm": 1.5625,
+      "learning_rate": 2.12215854247228e-05,
+      "loss": 0.3331,
+      "step": 12518
+    },
+    {
+      "epoch": 0.5497436798138248,
+      "grad_norm": 1.78125,
+      "learning_rate": 2.1214746756818044e-05,
+      "loss": 0.3523,
+      "step": 12520
+    },
+    {
+      "epoch": 0.5498314982930284,
+      "grad_norm": 1.9765625,
+      "learning_rate": 2.1207908378788888e-05,
+      "loss": 0.3616,
+      "step": 12522
+    },
+    {
+      "epoch": 0.5499193167722318,
+      "grad_norm": 1.7265625,
+      "learning_rate": 2.1201070291159014e-05,
+      "loss": 0.3123,
+      "step": 12524
+    },
+    {
+      "epoch": 0.5500071352514353,
+      "grad_norm": 1.8125,
+      "learning_rate": 2.119423249445209e-05,
+      "loss": 0.3212,
+      "step": 12526
+    },
+    {
+      "epoch": 0.5500949537306388,
+      "grad_norm": 1.84375,
+      "learning_rate": 2.1187394989191754e-05,
+      "loss": 0.3603,
+      "step": 12528
+    },
+    {
+      "epoch": 0.5501827722098422,
+      "grad_norm": 1.78125,
+      "learning_rate": 2.118055777590162e-05,
+      "loss": 0.3227,
+      "step": 12530
+    },
+    {
+      "epoch": 0.5502705906890457,
+      "grad_norm": 1.875,
+      "learning_rate": 2.1173720855105295e-05,
+      "loss": 0.3258,
+      "step": 12532
+    },
+    {
+      "epoch": 0.5503584091682492,
+      "grad_norm": 1.875,
+      "learning_rate": 2.1166884227326338e-05,
+      "loss": 0.3497,
+      "step": 12534
+    },
+    {
+      "epoch": 0.5504462276474528,
+      "grad_norm": 1.8203125,
+      "learning_rate": 2.116004789308831e-05,
+      "loss": 0.3479,
+      "step": 12536
+    },
+    {
+      "epoch": 0.5505340461266562,
+      "grad_norm": 1.84375,
+      "learning_rate": 2.115321185291473e-05,
+      "loss": 0.3353,
+      "step": 12538
+    },
+    {
+      "epoch": 0.5506218646058597,
+      "grad_norm": 1.6953125,
+      "learning_rate": 2.114637610732911e-05,
+      "loss": 0.3391,
+      "step": 12540
+    },
+    {
+      "epoch": 0.5507096830850632,
+      "grad_norm": 1.9296875,
+      "learning_rate": 2.1139540656854934e-05,
+      "loss": 0.3559,
+      "step": 12542
+    },
+    {
+      "epoch": 0.5507975015642667,
+      "grad_norm": 1.828125,
+      "learning_rate": 2.1132705502015657e-05,
+      "loss": 0.359,
+      "step": 12544
+    },
+    {
+      "epoch": 0.5508853200434701,
+      "grad_norm": 1.78125,
+      "learning_rate": 2.112587064333472e-05,
+      "loss": 0.3664,
+      "step": 12546
+    },
+    {
+      "epoch": 0.5509731385226736,
+      "grad_norm": 1.7109375,
+      "learning_rate": 2.1119036081335536e-05,
+      "loss": 0.3311,
+      "step": 12548
+    },
+    {
+      "epoch": 0.5510609570018771,
+      "grad_norm": 1.7109375,
+      "learning_rate": 2.111220181654151e-05,
+      "loss": 0.3288,
+      "step": 12550
+    },
+    {
+      "epoch": 0.5511487754810807,
+      "grad_norm": 1.703125,
+      "learning_rate": 2.1105367849475993e-05,
+      "loss": 0.331,
+      "step": 12552
+    },
+    {
+      "epoch": 0.5512365939602841,
+      "grad_norm": 1.6796875,
+      "learning_rate": 2.1098534180662337e-05,
+      "loss": 0.3219,
+      "step": 12554
+    },
+    {
+      "epoch": 0.5513244124394876,
+      "grad_norm": 1.8359375,
+      "learning_rate": 2.109170081062387e-05,
+      "loss": 0.3198,
+      "step": 12556
+    },
+    {
+      "epoch": 0.5514122309186911,
+      "grad_norm": 1.7421875,
+      "learning_rate": 2.1084867739883894e-05,
+      "loss": 0.339,
+      "step": 12558
+    },
+    {
+      "epoch": 0.5515000493978945,
+      "grad_norm": 1.6875,
+      "learning_rate": 2.1078034968965685e-05,
+      "loss": 0.3557,
+      "step": 12560
+    },
+    {
+      "epoch": 0.551587867877098,
+      "grad_norm": 1.84375,
+      "learning_rate": 2.1071202498392493e-05,
+      "loss": 0.3697,
+      "step": 12562
+    },
+    {
+      "epoch": 0.5516756863563015,
+      "grad_norm": 1.640625,
+      "learning_rate": 2.106437032868756e-05,
+      "loss": 0.3533,
+      "step": 12564
+    },
+    {
+      "epoch": 0.551763504835505,
+      "grad_norm": 1.84375,
+      "learning_rate": 2.105753846037409e-05,
+      "loss": 0.3537,
+      "step": 12566
+    },
+    {
+      "epoch": 0.5518513233147085,
+      "grad_norm": 1.7578125,
+      "learning_rate": 2.1050706893975274e-05,
+      "loss": 0.3369,
+      "step": 12568
+    },
+    {
+      "epoch": 0.551939141793912,
+      "grad_norm": 1.640625,
+      "learning_rate": 2.1043875630014265e-05,
+      "loss": 0.3325,
+      "step": 12570
+    },
+    {
+      "epoch": 0.5520269602731155,
+      "grad_norm": 1.609375,
+      "learning_rate": 2.103704466901421e-05,
+      "loss": 0.3235,
+      "step": 12572
+    },
+    {
+      "epoch": 0.552114778752319,
+      "grad_norm": 1.8515625,
+      "learning_rate": 2.1030214011498224e-05,
+      "loss": 0.3459,
+      "step": 12574
+    },
+    {
+      "epoch": 0.5522025972315224,
+      "grad_norm": 1.8671875,
+      "learning_rate": 2.10233836579894e-05,
+      "loss": 0.3483,
+      "step": 12576
+    },
+    {
+      "epoch": 0.5522904157107259,
+      "grad_norm": 1.6875,
+      "learning_rate": 2.1016553609010807e-05,
+      "loss": 0.326,
+      "step": 12578
+    },
+    {
+      "epoch": 0.5523782341899294,
+      "grad_norm": 1.6875,
+      "learning_rate": 2.1009723865085506e-05,
+      "loss": 0.3566,
+      "step": 12580
+    },
+    {
+      "epoch": 0.552466052669133,
+      "grad_norm": 1.7421875,
+      "learning_rate": 2.10028944267365e-05,
+      "loss": 0.3525,
+      "step": 12582
+    },
+    {
+      "epoch": 0.5525538711483364,
+      "grad_norm": 1.703125,
+      "learning_rate": 2.0996065294486798e-05,
+      "loss": 0.3291,
+      "step": 12584
+    },
+    {
+      "epoch": 0.5526416896275399,
+      "grad_norm": 1.53125,
+      "learning_rate": 2.098923646885937e-05,
+      "loss": 0.3655,
+      "step": 12586
+    },
+    {
+      "epoch": 0.5527295081067434,
+      "grad_norm": 1.5625,
+      "learning_rate": 2.0982407950377177e-05,
+      "loss": 0.3266,
+      "step": 12588
+    },
+    {
+      "epoch": 0.5528173265859468,
+      "grad_norm": 1.71875,
+      "learning_rate": 2.0975579739563144e-05,
+      "loss": 0.348,
+      "step": 12590
+    },
+    {
+      "epoch": 0.5529051450651503,
+      "grad_norm": 1.8046875,
+      "learning_rate": 2.096875183694018e-05,
+      "loss": 0.3656,
+      "step": 12592
+    },
+    {
+      "epoch": 0.5529929635443538,
+      "grad_norm": 1.7578125,
+      "learning_rate": 2.0961924243031177e-05,
+      "loss": 0.3601,
+      "step": 12594
+    },
+    {
+      "epoch": 0.5530807820235573,
+      "grad_norm": 1.6328125,
+      "learning_rate": 2.0955096958358973e-05,
+      "loss": 0.365,
+      "step": 12596
+    },
+    {
+      "epoch": 0.5531686005027608,
+      "grad_norm": 1.734375,
+      "learning_rate": 2.0948269983446413e-05,
+      "loss": 0.3289,
+      "step": 12598
+    },
+    {
+      "epoch": 0.5532564189819643,
+      "grad_norm": 1.703125,
+      "learning_rate": 2.0941443318816313e-05,
+      "loss": 0.3671,
+      "step": 12600
+    },
+    {
+      "epoch": 0.5533442374611678,
+      "grad_norm": 1.90625,
+      "learning_rate": 2.093461696499145e-05,
+      "loss": 0.34,
+      "step": 12602
+    },
+    {
+      "epoch": 0.5534320559403713,
+      "grad_norm": 1.9765625,
+      "learning_rate": 2.0927790922494607e-05,
+      "loss": 0.3728,
+      "step": 12604
+    },
+    {
+      "epoch": 0.5535198744195747,
+      "grad_norm": 1.84375,
+      "learning_rate": 2.0920965191848503e-05,
+      "loss": 0.3345,
+      "step": 12606
+    },
+    {
+      "epoch": 0.5536076928987782,
+      "grad_norm": 1.9765625,
+      "learning_rate": 2.091413977357586e-05,
+      "loss": 0.3539,
+      "step": 12608
+    },
+    {
+      "epoch": 0.5536955113779817,
+      "grad_norm": 1.828125,
+      "learning_rate": 2.0907314668199386e-05,
+      "loss": 0.3434,
+      "step": 12610
+    },
+    {
+      "epoch": 0.5537833298571851,
+      "grad_norm": 1.7890625,
+      "learning_rate": 2.0900489876241724e-05,
+      "loss": 0.3751,
+      "step": 12612
+    },
+    {
+      "epoch": 0.5538711483363887,
+      "grad_norm": 1.96875,
+      "learning_rate": 2.089366539822553e-05,
+      "loss": 0.3286,
+      "step": 12614
+    },
+    {
+      "epoch": 0.5539589668155922,
+      "grad_norm": 1.6484375,
+      "learning_rate": 2.0886841234673426e-05,
+      "loss": 0.3511,
+      "step": 12616
+    },
+    {
+      "epoch": 0.5540467852947957,
+      "grad_norm": 1.96875,
+      "learning_rate": 2.0880017386108005e-05,
+      "loss": 0.3467,
+      "step": 12618
+    },
+    {
+      "epoch": 0.5541346037739991,
+      "grad_norm": 1.765625,
+      "learning_rate": 2.0873193853051844e-05,
+      "loss": 0.3602,
+      "step": 12620
+    },
+    {
+      "epoch": 0.5542224222532026,
+      "grad_norm": 1.8046875,
+      "learning_rate": 2.0866370636027484e-05,
+      "loss": 0.3364,
+      "step": 12622
+    },
+    {
+      "epoch": 0.5543102407324061,
+      "grad_norm": 1.8671875,
+      "learning_rate": 2.0859547735557453e-05,
+      "loss": 0.3398,
+      "step": 12624
+    },
+    {
+      "epoch": 0.5543980592116096,
+      "grad_norm": 1.6875,
+      "learning_rate": 2.0852725152164247e-05,
+      "loss": 0.3592,
+      "step": 12626
+    },
+    {
+      "epoch": 0.5544858776908131,
+      "grad_norm": 1.84375,
+      "learning_rate": 2.0845902886370342e-05,
+      "loss": 0.364,
+      "step": 12628
+    },
+    {
+      "epoch": 0.5545736961700166,
+      "grad_norm": 1.6953125,
+      "learning_rate": 2.0839080938698193e-05,
+      "loss": 0.3523,
+      "step": 12630
+    },
+    {
+      "epoch": 0.5546615146492201,
+      "grad_norm": 1.734375,
+      "learning_rate": 2.0832259309670223e-05,
+      "loss": 0.3408,
+      "step": 12632
+    },
+    {
+      "epoch": 0.5547493331284236,
+      "grad_norm": 2.015625,
+      "learning_rate": 2.082543799980883e-05,
+      "loss": 0.3592,
+      "step": 12634
+    },
+    {
+      "epoch": 0.554837151607627,
+      "grad_norm": 1.828125,
+      "learning_rate": 2.08186170096364e-05,
+      "loss": 0.3799,
+      "step": 12636
+    },
+    {
+      "epoch": 0.5549249700868305,
+      "grad_norm": 1.765625,
+      "learning_rate": 2.0811796339675277e-05,
+      "loss": 0.3433,
+      "step": 12638
+    },
+    {
+      "epoch": 0.555012788566034,
+      "grad_norm": 1.765625,
+      "learning_rate": 2.0804975990447802e-05,
+      "loss": 0.3505,
+      "step": 12640
+    },
+    {
+      "epoch": 0.5551006070452374,
+      "grad_norm": 1.984375,
+      "learning_rate": 2.079815596247627e-05,
+      "loss": 0.3466,
+      "step": 12642
+    },
+    {
+      "epoch": 0.555188425524441,
+      "grad_norm": 1.7421875,
+      "learning_rate": 2.0791336256282958e-05,
+      "loss": 0.3317,
+      "step": 12644
+    },
+    {
+      "epoch": 0.5552762440036445,
+      "grad_norm": 1.703125,
+      "learning_rate": 2.078451687239013e-05,
+      "loss": 0.3306,
+      "step": 12646
+    },
+    {
+      "epoch": 0.555364062482848,
+      "grad_norm": 1.921875,
+      "learning_rate": 2.0777697811320003e-05,
+      "loss": 0.3522,
+      "step": 12648
+    },
+    {
+      "epoch": 0.5554518809620514,
+      "grad_norm": 1.671875,
+      "learning_rate": 2.0770879073594794e-05,
+      "loss": 0.3407,
+      "step": 12650
+    },
+    {
+      "epoch": 0.5555396994412549,
+      "grad_norm": 1.765625,
+      "learning_rate": 2.0764060659736674e-05,
+      "loss": 0.3543,
+      "step": 12652
+    },
+    {
+      "epoch": 0.5556275179204584,
+      "grad_norm": 1.6875,
+      "learning_rate": 2.0757242570267808e-05,
+      "loss": 0.3488,
+      "step": 12654
+    },
+    {
+      "epoch": 0.5557153363996619,
+      "grad_norm": 1.640625,
+      "learning_rate": 2.0750424805710326e-05,
+      "loss": 0.344,
+      "step": 12656
+    },
+    {
+      "epoch": 0.5558031548788653,
+      "grad_norm": 1.734375,
+      "learning_rate": 2.0743607366586327e-05,
+      "loss": 0.3543,
+      "step": 12658
+    },
+    {
+      "epoch": 0.5558909733580689,
+      "grad_norm": 1.640625,
+      "learning_rate": 2.0736790253417893e-05,
+      "loss": 0.3518,
+      "step": 12660
+    },
+    {
+      "epoch": 0.5559787918372724,
+      "grad_norm": 1.734375,
+      "learning_rate": 2.0729973466727086e-05,
+      "loss": 0.3597,
+      "step": 12662
+    },
+    {
+      "epoch": 0.5560666103164759,
+      "grad_norm": 1.6484375,
+      "learning_rate": 2.0723157007035938e-05,
+      "loss": 0.3324,
+      "step": 12664
+    },
+    {
+      "epoch": 0.5561544287956793,
+      "grad_norm": 1.6484375,
+      "learning_rate": 2.0716340874866453e-05,
+      "loss": 0.3592,
+      "step": 12666
+    },
+    {
+      "epoch": 0.5562422472748828,
+      "grad_norm": 1.90625,
+      "learning_rate": 2.0709525070740604e-05,
+      "loss": 0.3427,
+      "step": 12668
+    },
+    {
+      "epoch": 0.5563300657540863,
+      "grad_norm": 1.609375,
+      "learning_rate": 2.0702709595180357e-05,
+      "loss": 0.3546,
+      "step": 12670
+    },
+    {
+      "epoch": 0.5564178842332898,
+      "grad_norm": 1.6640625,
+      "learning_rate": 2.0695894448707647e-05,
+      "loss": 0.3577,
+      "step": 12672
+    },
+    {
+      "epoch": 0.5565057027124932,
+      "grad_norm": 1.6953125,
+      "learning_rate": 2.0689079631844363e-05,
+      "loss": 0.3441,
+      "step": 12674
+    },
+    {
+      "epoch": 0.5565935211916968,
+      "grad_norm": 1.8203125,
+      "learning_rate": 2.0682265145112395e-05,
+      "loss": 0.3469,
+      "step": 12676
+    },
+    {
+      "epoch": 0.5566813396709003,
+      "grad_norm": 1.7421875,
+      "learning_rate": 2.0675450989033595e-05,
+      "loss": 0.368,
+      "step": 12678
+    },
+    {
+      "epoch": 0.5567691581501037,
+      "grad_norm": 1.671875,
+      "learning_rate": 2.0668637164129796e-05,
+      "loss": 0.333,
+      "step": 12680
+    },
+    {
+      "epoch": 0.5568569766293072,
+      "grad_norm": 1.6953125,
+      "learning_rate": 2.0661823670922806e-05,
+      "loss": 0.3168,
+      "step": 12682
+    },
+    {
+      "epoch": 0.5569447951085107,
+      "grad_norm": 1.6875,
+      "learning_rate": 2.0655010509934395e-05,
+      "loss": 0.3233,
+      "step": 12684
+    },
+    {
+      "epoch": 0.5570326135877142,
+      "grad_norm": 2.046875,
+      "learning_rate": 2.064819768168632e-05,
+      "loss": 0.3385,
+      "step": 12686
+    },
+    {
+      "epoch": 0.5571204320669176,
+      "grad_norm": 1.71875,
+      "learning_rate": 2.064138518670031e-05,
+      "loss": 0.3491,
+      "step": 12688
+    },
+    {
+      "epoch": 0.5572082505461212,
+      "grad_norm": 1.6875,
+      "learning_rate": 2.0634573025498067e-05,
+      "loss": 0.329,
+      "step": 12690
+    },
+    {
+      "epoch": 0.5572960690253247,
+      "grad_norm": 1.8203125,
+      "learning_rate": 2.0627761198601273e-05,
+      "loss": 0.3355,
+      "step": 12692
+    },
+    {
+      "epoch": 0.5573838875045282,
+      "grad_norm": 1.6328125,
+      "learning_rate": 2.062094970653157e-05,
+      "loss": 0.349,
+      "step": 12694
+    },
+    {
+      "epoch": 0.5574717059837316,
+      "grad_norm": 1.75,
+      "learning_rate": 2.0614138549810587e-05,
+      "loss": 0.3676,
+      "step": 12696
+    },
+    {
+      "epoch": 0.5575595244629351,
+      "grad_norm": 1.71875,
+      "learning_rate": 2.0607327728959928e-05,
+      "loss": 0.3912,
+      "step": 12698
+    },
+    {
+      "epoch": 0.5576473429421386,
+      "grad_norm": 1.9375,
+      "learning_rate": 2.060051724450116e-05,
+      "loss": 0.3439,
+      "step": 12700
+    },
+    {
+      "epoch": 0.557735161421342,
+      "grad_norm": 1.7578125,
+      "learning_rate": 2.059370709695585e-05,
+      "loss": 0.3781,
+      "step": 12702
+    },
+    {
+      "epoch": 0.5578229799005455,
+      "grad_norm": 1.6953125,
+      "learning_rate": 2.0586897286845498e-05,
+      "loss": 0.3375,
+      "step": 12704
+    },
+    {
+      "epoch": 0.5579107983797491,
+      "grad_norm": 1.6796875,
+      "learning_rate": 2.0580087814691607e-05,
+      "loss": 0.3421,
+      "step": 12706
+    },
+    {
+      "epoch": 0.5579986168589526,
+      "grad_norm": 1.875,
+      "learning_rate": 2.0573278681015657e-05,
+      "loss": 0.3377,
+      "step": 12708
+    },
+    {
+      "epoch": 0.558086435338156,
+      "grad_norm": 1.671875,
+      "learning_rate": 2.056646988633908e-05,
+      "loss": 0.3251,
+      "step": 12710
+    },
+    {
+      "epoch": 0.5581742538173595,
+      "grad_norm": 1.640625,
+      "learning_rate": 2.0559661431183303e-05,
+      "loss": 0.3299,
+      "step": 12712
+    },
+    {
+      "epoch": 0.558262072296563,
+      "grad_norm": 1.65625,
+      "learning_rate": 2.0552853316069717e-05,
+      "loss": 0.3172,
+      "step": 12714
+    },
+    {
+      "epoch": 0.5583498907757665,
+      "grad_norm": 1.7109375,
+      "learning_rate": 2.054604554151969e-05,
+      "loss": 0.3287,
+      "step": 12716
+    },
+    {
+      "epoch": 0.5584377092549699,
+      "grad_norm": 1.7890625,
+      "learning_rate": 2.053923810805457e-05,
+      "loss": 0.3309,
+      "step": 12718
+    },
+    {
+      "epoch": 0.5585255277341734,
+      "grad_norm": 1.71875,
+      "learning_rate": 2.0532431016195658e-05,
+      "loss": 0.3412,
+      "step": 12720
+    },
+    {
+      "epoch": 0.558613346213377,
+      "grad_norm": 1.859375,
+      "learning_rate": 2.052562426646425e-05,
+      "loss": 0.3359,
+      "step": 12722
+    },
+    {
+      "epoch": 0.5587011646925805,
+      "grad_norm": 1.734375,
+      "learning_rate": 2.0518817859381607e-05,
+      "loss": 0.351,
+      "step": 12724
+    },
+    {
+      "epoch": 0.5587889831717839,
+      "grad_norm": 2.28125,
+      "learning_rate": 2.0512011795468972e-05,
+      "loss": 0.3586,
+      "step": 12726
+    },
+    {
+      "epoch": 0.5588768016509874,
+      "grad_norm": 1.6640625,
+      "learning_rate": 2.0505206075247552e-05,
+      "loss": 0.3383,
+      "step": 12728
+    },
+    {
+      "epoch": 0.5589646201301909,
+      "grad_norm": 1.625,
+      "learning_rate": 2.0498400699238522e-05,
+      "loss": 0.3388,
+      "step": 12730
+    },
+    {
+      "epoch": 0.5590524386093944,
+      "grad_norm": 1.7109375,
+      "learning_rate": 2.0491595667963042e-05,
+      "loss": 0.3217,
+      "step": 12732
+    },
+    {
+      "epoch": 0.5591402570885978,
+      "grad_norm": 1.9453125,
+      "learning_rate": 2.0484790981942263e-05,
+      "loss": 0.3533,
+      "step": 12734
+    },
+    {
+      "epoch": 0.5592280755678014,
+      "grad_norm": 1.59375,
+      "learning_rate": 2.047798664169726e-05,
+      "loss": 0.3279,
+      "step": 12736
+    },
+    {
+      "epoch": 0.5593158940470049,
+      "grad_norm": 1.8671875,
+      "learning_rate": 2.0471182647749123e-05,
+      "loss": 0.3693,
+      "step": 12738
+    },
+    {
+      "epoch": 0.5594037125262084,
+      "grad_norm": 1.6796875,
+      "learning_rate": 2.0464379000618906e-05,
+      "loss": 0.3364,
+      "step": 12740
+    },
+    {
+      "epoch": 0.5594915310054118,
+      "grad_norm": 1.6875,
+      "learning_rate": 2.0457575700827634e-05,
+      "loss": 0.3385,
+      "step": 12742
+    },
+    {
+      "epoch": 0.5595793494846153,
+      "grad_norm": 1.625,
+      "learning_rate": 2.0450772748896307e-05,
+      "loss": 0.3681,
+      "step": 12744
+    },
+    {
+      "epoch": 0.5596671679638188,
+      "grad_norm": 1.6015625,
+      "learning_rate": 2.0443970145345887e-05,
+      "loss": 0.3346,
+      "step": 12746
+    },
+    {
+      "epoch": 0.5597549864430222,
+      "grad_norm": 1.8359375,
+      "learning_rate": 2.0437167890697324e-05,
+      "loss": 0.328,
+      "step": 12748
+    },
+    {
+      "epoch": 0.5598428049222257,
+      "grad_norm": 1.8046875,
+      "learning_rate": 2.0430365985471542e-05,
+      "loss": 0.3155,
+      "step": 12750
+    },
+    {
+      "epoch": 0.5599306234014293,
+      "grad_norm": 1.6484375,
+      "learning_rate": 2.0423564430189427e-05,
+      "loss": 0.3309,
+      "step": 12752
+    },
+    {
+      "epoch": 0.5600184418806328,
+      "grad_norm": 1.8671875,
+      "learning_rate": 2.0416763225371848e-05,
+      "loss": 0.3572,
+      "step": 12754
+    },
+    {
+      "epoch": 0.5601062603598362,
+      "grad_norm": 1.8203125,
+      "learning_rate": 2.0409962371539637e-05,
+      "loss": 0.341,
+      "step": 12756
+    },
+    {
+      "epoch": 0.5601940788390397,
+      "grad_norm": 1.78125,
+      "learning_rate": 2.0403161869213608e-05,
+      "loss": 0.3409,
+      "step": 12758
+    },
+    {
+      "epoch": 0.5602818973182432,
+      "grad_norm": 1.78125,
+      "learning_rate": 2.0396361718914545e-05,
+      "loss": 0.3501,
+      "step": 12760
+    },
+    {
+      "epoch": 0.5603697157974467,
+      "grad_norm": 1.8671875,
+      "learning_rate": 2.0389561921163203e-05,
+      "loss": 0.3461,
+      "step": 12762
+    },
+    {
+      "epoch": 0.5604575342766501,
+      "grad_norm": 1.8671875,
+      "learning_rate": 2.0382762476480326e-05,
+      "loss": 0.3411,
+      "step": 12764
+    },
+    {
+      "epoch": 0.5605453527558536,
+      "grad_norm": 1.8984375,
+      "learning_rate": 2.03759633853866e-05,
+      "loss": 0.3693,
+      "step": 12766
+    },
+    {
+      "epoch": 0.5606331712350572,
+      "grad_norm": 1.90625,
+      "learning_rate": 2.0369164648402704e-05,
+      "loss": 0.3263,
+      "step": 12768
+    },
+    {
+      "epoch": 0.5607209897142607,
+      "grad_norm": 1.6640625,
+      "learning_rate": 2.0362366266049288e-05,
+      "loss": 0.3549,
+      "step": 12770
+    },
+    {
+      "epoch": 0.5608088081934641,
+      "grad_norm": 1.6953125,
+      "learning_rate": 2.0355568238846972e-05,
+      "loss": 0.3589,
+      "step": 12772
+    },
+    {
+      "epoch": 0.5608966266726676,
+      "grad_norm": 1.6015625,
+      "learning_rate": 2.0348770567316354e-05,
+      "loss": 0.336,
+      "step": 12774
+    },
+    {
+      "epoch": 0.5609844451518711,
+      "grad_norm": 1.7734375,
+      "learning_rate": 2.0341973251978003e-05,
+      "loss": 0.3383,
+      "step": 12776
+    },
+    {
+      "epoch": 0.5610722636310745,
+      "grad_norm": 1.6171875,
+      "learning_rate": 2.0335176293352456e-05,
+      "loss": 0.331,
+      "step": 12778
+    },
+    {
+      "epoch": 0.561160082110278,
+      "grad_norm": 1.875,
+      "learning_rate": 2.0328379691960235e-05,
+      "loss": 0.3607,
+      "step": 12780
+    },
+    {
+      "epoch": 0.5612479005894816,
+      "grad_norm": 1.7109375,
+      "learning_rate": 2.032158344832181e-05,
+      "loss": 0.3026,
+      "step": 12782
+    },
+    {
+      "epoch": 0.5613357190686851,
+      "grad_norm": 1.90625,
+      "learning_rate": 2.0314787562957645e-05,
+      "loss": 0.3605,
+      "step": 12784
+    },
+    {
+      "epoch": 0.5614235375478885,
+      "grad_norm": 2.015625,
+      "learning_rate": 2.0307992036388167e-05,
+      "loss": 0.3378,
+      "step": 12786
+    },
+    {
+      "epoch": 0.561511356027092,
+      "grad_norm": 1.796875,
+      "learning_rate": 2.030119686913379e-05,
+      "loss": 0.3292,
+      "step": 12788
+    },
+    {
+      "epoch": 0.5615991745062955,
+      "grad_norm": 1.734375,
+      "learning_rate": 2.0294402061714887e-05,
+      "loss": 0.3457,
+      "step": 12790
+    },
+    {
+      "epoch": 0.561686992985499,
+      "grad_norm": 1.6640625,
+      "learning_rate": 2.0287607614651797e-05,
+      "loss": 0.3479,
+      "step": 12792
+    },
+    {
+      "epoch": 0.5617748114647024,
+      "grad_norm": 1.65625,
+      "learning_rate": 2.0280813528464855e-05,
+      "loss": 0.3728,
+      "step": 12794
+    },
+    {
+      "epoch": 0.5618626299439059,
+      "grad_norm": 1.84375,
+      "learning_rate": 2.0274019803674335e-05,
+      "loss": 0.3711,
+      "step": 12796
+    },
+    {
+      "epoch": 0.5619504484231095,
+      "grad_norm": 1.8671875,
+      "learning_rate": 2.026722644080051e-05,
+      "loss": 0.3665,
+      "step": 12798
+    },
+    {
+      "epoch": 0.562038266902313,
+      "grad_norm": 1.6953125,
+      "learning_rate": 2.0260433440363617e-05,
+      "loss": 0.3475,
+      "step": 12800
+    },
+    {
+      "epoch": 0.5621260853815164,
+      "grad_norm": 1.6328125,
+      "learning_rate": 2.025364080288387e-05,
+      "loss": 0.3542,
+      "step": 12802
+    },
+    {
+      "epoch": 0.5622139038607199,
+      "grad_norm": 1.9609375,
+      "learning_rate": 2.0246848528881455e-05,
+      "loss": 0.3597,
+      "step": 12804
+    },
+    {
+      "epoch": 0.5623017223399234,
+      "grad_norm": 1.9453125,
+      "learning_rate": 2.0240056618876514e-05,
+      "loss": 0.3296,
+      "step": 12806
+    },
+    {
+      "epoch": 0.5623895408191268,
+      "grad_norm": 1.671875,
+      "learning_rate": 2.0233265073389173e-05,
+      "loss": 0.3503,
+      "step": 12808
+    },
+    {
+      "epoch": 0.5624773592983303,
+      "grad_norm": 1.609375,
+      "learning_rate": 2.022647389293954e-05,
+      "loss": 0.3532,
+      "step": 12810
+    },
+    {
+      "epoch": 0.5625651777775338,
+      "grad_norm": 1.640625,
+      "learning_rate": 2.0219683078047678e-05,
+      "loss": 0.3695,
+      "step": 12812
+    },
+    {
+      "epoch": 0.5626529962567374,
+      "grad_norm": 1.6875,
+      "learning_rate": 2.0212892629233633e-05,
+      "loss": 0.3369,
+      "step": 12814
+    },
+    {
+      "epoch": 0.5627408147359408,
+      "grad_norm": 1.734375,
+      "learning_rate": 2.0206102547017426e-05,
+      "loss": 0.3601,
+      "step": 12816
+    },
+    {
+      "epoch": 0.5628286332151443,
+      "grad_norm": 1.671875,
+      "learning_rate": 2.0199312831919024e-05,
+      "loss": 0.3639,
+      "step": 12818
+    },
+    {
+      "epoch": 0.5629164516943478,
+      "grad_norm": 1.796875,
+      "learning_rate": 2.0192523484458397e-05,
+      "loss": 0.3868,
+      "step": 12820
+    },
+    {
+      "epoch": 0.5630042701735513,
+      "grad_norm": 1.7578125,
+      "learning_rate": 2.018573450515548e-05,
+      "loss": 0.3354,
+      "step": 12822
+    },
+    {
+      "epoch": 0.5630920886527547,
+      "grad_norm": 1.734375,
+      "learning_rate": 2.0178945894530165e-05,
+      "loss": 0.3583,
+      "step": 12824
+    },
+    {
+      "epoch": 0.5631799071319582,
+      "grad_norm": 1.71875,
+      "learning_rate": 2.0172157653102337e-05,
+      "loss": 0.3593,
+      "step": 12826
+    },
+    {
+      "epoch": 0.5632677256111617,
+      "grad_norm": 1.703125,
+      "learning_rate": 2.0165369781391824e-05,
+      "loss": 0.3227,
+      "step": 12828
+    },
+    {
+      "epoch": 0.5633555440903653,
+      "grad_norm": 1.9453125,
+      "learning_rate": 2.0158582279918458e-05,
+      "loss": 0.3631,
+      "step": 12830
+    },
+    {
+      "epoch": 0.5634433625695687,
+      "grad_norm": 1.78125,
+      "learning_rate": 2.015179514920202e-05,
+      "loss": 0.3561,
+      "step": 12832
+    },
+    {
+      "epoch": 0.5635311810487722,
+      "grad_norm": 1.7109375,
+      "learning_rate": 2.0145008389762265e-05,
+      "loss": 0.3265,
+      "step": 12834
+    },
+    {
+      "epoch": 0.5636189995279757,
+      "grad_norm": 1.6953125,
+      "learning_rate": 2.0138222002118934e-05,
+      "loss": 0.3476,
+      "step": 12836
+    },
+    {
+      "epoch": 0.5637068180071791,
+      "grad_norm": 1.859375,
+      "learning_rate": 2.013143598679173e-05,
+      "loss": 0.3589,
+      "step": 12838
+    },
+    {
+      "epoch": 0.5637946364863826,
+      "grad_norm": 1.703125,
+      "learning_rate": 2.0124650344300323e-05,
+      "loss": 0.3578,
+      "step": 12840
+    },
+    {
+      "epoch": 0.5638824549655861,
+      "grad_norm": 1.71875,
+      "learning_rate": 2.0117865075164366e-05,
+      "loss": 0.3233,
+      "step": 12842
+    },
+    {
+      "epoch": 0.5639702734447897,
+      "grad_norm": 1.609375,
+      "learning_rate": 2.011108017990347e-05,
+      "loss": 0.3266,
+      "step": 12844
+    },
+    {
+      "epoch": 0.5640580919239931,
+      "grad_norm": 1.703125,
+      "learning_rate": 2.010429565903722e-05,
+      "loss": 0.3269,
+      "step": 12846
+    },
+    {
+      "epoch": 0.5641459104031966,
+      "grad_norm": 1.921875,
+      "learning_rate": 2.009751151308519e-05,
+      "loss": 0.3282,
+      "step": 12848
+    },
+    {
+      "epoch": 0.5642337288824001,
+      "grad_norm": 1.828125,
+      "learning_rate": 2.0090727742566894e-05,
+      "loss": 0.3494,
+      "step": 12850
+    },
+    {
+      "epoch": 0.5643215473616036,
+      "grad_norm": 1.875,
+      "learning_rate": 2.0083944348001856e-05,
+      "loss": 0.3248,
+      "step": 12852
+    },
+    {
+      "epoch": 0.564409365840807,
+      "grad_norm": 1.9296875,
+      "learning_rate": 2.0077161329909533e-05,
+      "loss": 0.3379,
+      "step": 12854
+    },
+    {
+      "epoch": 0.5644971843200105,
+      "grad_norm": 1.5859375,
+      "learning_rate": 2.007037868880938e-05,
+      "loss": 0.3335,
+      "step": 12856
+    },
+    {
+      "epoch": 0.564585002799214,
+      "grad_norm": 1.734375,
+      "learning_rate": 2.0063596425220803e-05,
+      "loss": 0.3479,
+      "step": 12858
+    },
+    {
+      "epoch": 0.5646728212784176,
+      "grad_norm": 1.8828125,
+      "learning_rate": 2.0056814539663196e-05,
+      "loss": 0.3678,
+      "step": 12860
+    },
+    {
+      "epoch": 0.564760639757621,
+      "grad_norm": 1.53125,
+      "learning_rate": 2.0050033032655915e-05,
+      "loss": 0.3086,
+      "step": 12862
+    },
+    {
+      "epoch": 0.5648484582368245,
+      "grad_norm": 1.65625,
+      "learning_rate": 2.004325190471829e-05,
+      "loss": 0.354,
+      "step": 12864
+    },
+    {
+      "epoch": 0.564936276716028,
+      "grad_norm": 1.6015625,
+      "learning_rate": 2.0036471156369632e-05,
+      "loss": 0.3582,
+      "step": 12866
+    },
+    {
+      "epoch": 0.5650240951952314,
+      "grad_norm": 1.6953125,
+      "learning_rate": 2.0029690788129194e-05,
+      "loss": 0.3552,
+      "step": 12868
+    },
+    {
+      "epoch": 0.5651119136744349,
+      "grad_norm": 1.6484375,
+      "learning_rate": 2.0022910800516227e-05,
+      "loss": 0.3399,
+      "step": 12870
+    },
+    {
+      "epoch": 0.5651997321536384,
+      "grad_norm": 1.6796875,
+      "learning_rate": 2.001613119404994e-05,
+      "loss": 0.3803,
+      "step": 12872
+    },
+    {
+      "epoch": 0.5652875506328419,
+      "grad_norm": 1.609375,
+      "learning_rate": 2.000935196924953e-05,
+      "loss": 0.3496,
+      "step": 12874
+    },
+    {
+      "epoch": 0.5653753691120454,
+      "grad_norm": 1.640625,
+      "learning_rate": 2.0002573126634136e-05,
+      "loss": 0.3368,
+      "step": 12876
+    },
+    {
+      "epoch": 0.5654631875912489,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.99957946667229e-05,
+      "loss": 0.3569,
+      "step": 12878
+    },
+    {
+      "epoch": 0.5655510060704524,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.9989016590034896e-05,
+      "loss": 0.3643,
+      "step": 12880
+    },
+    {
+      "epoch": 0.5656388245496559,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.9982238897089207e-05,
+      "loss": 0.3609,
+      "step": 12882
+    },
+    {
+      "epoch": 0.5657266430288593,
+      "grad_norm": 1.8671875,
+      "learning_rate": 1.9975461588404866e-05,
+      "loss": 0.3383,
+      "step": 12884
+    },
+    {
+      "epoch": 0.5658144615080628,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.9968684664500888e-05,
+      "loss": 0.3505,
+      "step": 12886
+    },
+    {
+      "epoch": 0.5659022799872663,
+      "grad_norm": 1.7578125,
+      "learning_rate": 1.996190812589624e-05,
+      "loss": 0.3292,
+      "step": 12888
+    },
+    {
+      "epoch": 0.5659900984664699,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.995513197310987e-05,
+      "loss": 0.3501,
+      "step": 12890
+    },
+    {
+      "epoch": 0.5660779169456733,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.994835620666071e-05,
+      "loss": 0.3541,
+      "step": 12892
+    },
+    {
+      "epoch": 0.5661657354248768,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.994158082706764e-05,
+      "loss": 0.3566,
+      "step": 12894
+    },
+    {
+      "epoch": 0.5662535539040803,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.9934805834849518e-05,
+      "loss": 0.3237,
+      "step": 12896
+    },
+    {
+      "epoch": 0.5663413723832837,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.992803123052518e-05,
+      "loss": 0.3658,
+      "step": 12898
+    },
+    {
+      "epoch": 0.5664291908624872,
+      "grad_norm": 1.8515625,
+      "learning_rate": 1.992125701461343e-05,
+      "loss": 0.3631,
+      "step": 12900
+    },
+    {
+      "epoch": 0.5665170093416907,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.9914483187633042e-05,
+      "loss": 0.3414,
+      "step": 12902
+    },
+    {
+      "epoch": 0.5666048278208942,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.990770975010274e-05,
+      "loss": 0.34,
+      "step": 12904
+    },
+    {
+      "epoch": 0.5666926463000977,
+      "grad_norm": 1.7578125,
+      "learning_rate": 1.990093670254125e-05,
+      "loss": 0.3301,
+      "step": 12906
+    },
+    {
+      "epoch": 0.5667804647793012,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.9894164045467246e-05,
+      "loss": 0.3471,
+      "step": 12908
+    },
+    {
+      "epoch": 0.5668682832585047,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.9887391779399383e-05,
+      "loss": 0.3316,
+      "step": 12910
+    },
+    {
+      "epoch": 0.5669561017377082,
+      "grad_norm": 1.75,
+      "learning_rate": 1.9880619904856288e-05,
+      "loss": 0.3578,
+      "step": 12912
+    },
+    {
+      "epoch": 0.5670439202169116,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.987384842235655e-05,
+      "loss": 0.339,
+      "step": 12914
+    },
+    {
+      "epoch": 0.5671317386961151,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.9867077332418723e-05,
+      "loss": 0.3202,
+      "step": 12916
+    },
+    {
+      "epoch": 0.5672195571753186,
+      "grad_norm": 1.8203125,
+      "learning_rate": 1.9860306635561353e-05,
+      "loss": 0.3416,
+      "step": 12918
+    },
+    {
+      "epoch": 0.567307375654522,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.985353633230293e-05,
+      "loss": 0.3111,
+      "step": 12920
+    },
+    {
+      "epoch": 0.5673951941337256,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.984676642316192e-05,
+      "loss": 0.3586,
+      "step": 12922
+    },
+    {
+      "epoch": 0.5674830126129291,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.983999690865678e-05,
+      "loss": 0.3161,
+      "step": 12924
+    },
+    {
+      "epoch": 0.5675708310921326,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.9833227789305906e-05,
+      "loss": 0.3488,
+      "step": 12926
+    },
+    {
+      "epoch": 0.567658649571336,
+      "grad_norm": 1.8203125,
+      "learning_rate": 1.9826459065627694e-05,
+      "loss": 0.3508,
+      "step": 12928
+    },
+    {
+      "epoch": 0.5677464680505395,
+      "grad_norm": 2.03125,
+      "learning_rate": 1.9819690738140484e-05,
+      "loss": 0.332,
+      "step": 12930
+    },
+    {
+      "epoch": 0.567834286529743,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.9812922807362598e-05,
+      "loss": 0.3561,
+      "step": 12932
+    },
+    {
+      "epoch": 0.5679221050089465,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.980615527381233e-05,
+      "loss": 0.3248,
+      "step": 12934
+    },
+    {
+      "epoch": 0.56800992348815,
+      "grad_norm": 1.9765625,
+      "learning_rate": 1.9799388138007928e-05,
+      "loss": 0.3775,
+      "step": 12936
+    },
+    {
+      "epoch": 0.5680977419673535,
+      "grad_norm": 1.75,
+      "learning_rate": 1.979262140046764e-05,
+      "loss": 0.3259,
+      "step": 12938
+    },
+    {
+      "epoch": 0.568185560446557,
+      "grad_norm": 1.84375,
+      "learning_rate": 1.978585506170965e-05,
+      "loss": 0.3159,
+      "step": 12940
+    },
+    {
+      "epoch": 0.5682733789257605,
+      "grad_norm": 1.8984375,
+      "learning_rate": 1.9779089122252125e-05,
+      "loss": 0.3324,
+      "step": 12942
+    },
+    {
+      "epoch": 0.5683611974049639,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.977232358261321e-05,
+      "loss": 0.3292,
+      "step": 12944
+    },
+    {
+      "epoch": 0.5684490158841674,
+      "grad_norm": 1.8125,
+      "learning_rate": 1.9765558443311004e-05,
+      "loss": 0.3388,
+      "step": 12946
+    },
+    {
+      "epoch": 0.5685368343633709,
+      "grad_norm": 1.9375,
+      "learning_rate": 1.9758793704863606e-05,
+      "loss": 0.334,
+      "step": 12948
+    },
+    {
+      "epoch": 0.5686246528425744,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.9752029367789023e-05,
+      "loss": 0.3225,
+      "step": 12950
+    },
+    {
+      "epoch": 0.5687124713217779,
+      "grad_norm": 1.75,
+      "learning_rate": 1.9745265432605293e-05,
+      "loss": 0.3441,
+      "step": 12952
+    },
+    {
+      "epoch": 0.5688002898009814,
+      "grad_norm": 1.8984375,
+      "learning_rate": 1.97385018998304e-05,
+      "loss": 0.3379,
+      "step": 12954
+    },
+    {
+      "epoch": 0.5688881082801849,
+      "grad_norm": 1.75,
+      "learning_rate": 1.9731738769982287e-05,
+      "loss": 0.3578,
+      "step": 12956
+    },
+    {
+      "epoch": 0.5689759267593884,
+      "grad_norm": 2.15625,
+      "learning_rate": 1.9724976043578882e-05,
+      "loss": 0.3437,
+      "step": 12958
+    },
+    {
+      "epoch": 0.5690637452385918,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.971821372113808e-05,
+      "loss": 0.3526,
+      "step": 12960
+    },
+    {
+      "epoch": 0.5691515637177953,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.9711451803177734e-05,
+      "loss": 0.3622,
+      "step": 12962
+    },
+    {
+      "epoch": 0.5692393821969988,
+      "grad_norm": 2.140625,
+      "learning_rate": 1.970469029021568e-05,
+      "loss": 0.3646,
+      "step": 12964
+    },
+    {
+      "epoch": 0.5693272006762022,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.9697929182769712e-05,
+      "loss": 0.3192,
+      "step": 12966
+    },
+    {
+      "epoch": 0.5694150191554058,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.9691168481357594e-05,
+      "loss": 0.3408,
+      "step": 12968
+    },
+    {
+      "epoch": 0.5695028376346093,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.9684408186497068e-05,
+      "loss": 0.3438,
+      "step": 12970
+    },
+    {
+      "epoch": 0.5695906561138128,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.9677648298705838e-05,
+      "loss": 0.3413,
+      "step": 12972
+    },
+    {
+      "epoch": 0.5696784745930162,
+      "grad_norm": 1.9453125,
+      "learning_rate": 1.9670888818501576e-05,
+      "loss": 0.35,
+      "step": 12974
+    },
+    {
+      "epoch": 0.5697662930722197,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.9664129746401933e-05,
+      "loss": 0.327,
+      "step": 12976
+    },
+    {
+      "epoch": 0.5698541115514232,
+      "grad_norm": 1.8125,
+      "learning_rate": 1.9657371082924515e-05,
+      "loss": 0.3242,
+      "step": 12978
+    },
+    {
+      "epoch": 0.5699419300306267,
+      "grad_norm": 1.859375,
+      "learning_rate": 1.9650612828586895e-05,
+      "loss": 0.3106,
+      "step": 12980
+    },
+    {
+      "epoch": 0.5700297485098302,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.964385498390662e-05,
+      "loss": 0.3208,
+      "step": 12982
+    },
+    {
+      "epoch": 0.5701175669890337,
+      "grad_norm": 1.8046875,
+      "learning_rate": 1.9637097549401223e-05,
+      "loss": 0.3237,
+      "step": 12984
+    },
+    {
+      "epoch": 0.5702053854682372,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.9630340525588176e-05,
+      "loss": 0.3198,
+      "step": 12986
+    },
+    {
+      "epoch": 0.5702932039474407,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.9623583912984938e-05,
+      "loss": 0.3426,
+      "step": 12988
+    },
+    {
+      "epoch": 0.5703810224266441,
+      "grad_norm": 1.7890625,
+      "learning_rate": 1.961682771210894e-05,
+      "loss": 0.3656,
+      "step": 12990
+    },
+    {
+      "epoch": 0.5704688409058476,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.961007192347756e-05,
+      "loss": 0.3708,
+      "step": 12992
+    },
+    {
+      "epoch": 0.5705566593850511,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.9603316547608165e-05,
+      "loss": 0.3506,
+      "step": 12994
+    },
+    {
+      "epoch": 0.5706444778642545,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.9596561585018082e-05,
+      "loss": 0.3227,
+      "step": 12996
+    },
+    {
+      "epoch": 0.5707322963434581,
+      "grad_norm": 1.8984375,
+      "learning_rate": 1.9589807036224607e-05,
+      "loss": 0.3619,
+      "step": 12998
+    },
+    {
+      "epoch": 0.5708201148226616,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.9583052901745012e-05,
+      "loss": 0.3424,
+      "step": 13000
+    },
+    {
+      "epoch": 0.5709079333018651,
+      "grad_norm": 1.7890625,
+      "learning_rate": 1.957629918209652e-05,
+      "loss": 0.3276,
+      "step": 13002
+    },
+    {
+      "epoch": 0.5709957517810685,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.9569545877796336e-05,
+      "loss": 0.3318,
+      "step": 13004
+    },
+    {
+      "epoch": 0.571083570260272,
+      "grad_norm": 1.7734375,
+      "learning_rate": 1.9562792989361628e-05,
+      "loss": 0.3371,
+      "step": 13006
+    },
+    {
+      "epoch": 0.5711713887394755,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.955604051730954e-05,
+      "loss": 0.3476,
+      "step": 13008
+    },
+    {
+      "epoch": 0.571259207218679,
+      "grad_norm": 1.8671875,
+      "learning_rate": 1.9549288462157185e-05,
+      "loss": 0.3263,
+      "step": 13010
+    },
+    {
+      "epoch": 0.5713470256978824,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.9542536824421613e-05,
+      "loss": 0.3142,
+      "step": 13012
+    },
+    {
+      "epoch": 0.571434844177086,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.953578560461988e-05,
+      "loss": 0.356,
+      "step": 13014
+    },
+    {
+      "epoch": 0.5715226626562895,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.9529034803269e-05,
+      "loss": 0.3482,
+      "step": 13016
+    },
+    {
+      "epoch": 0.571610481135493,
+      "grad_norm": 1.9296875,
+      "learning_rate": 1.9522284420885942e-05,
+      "loss": 0.3243,
+      "step": 13018
+    },
+    {
+      "epoch": 0.5716982996146964,
+      "grad_norm": 1.9140625,
+      "learning_rate": 1.9515534457987655e-05,
+      "loss": 0.3365,
+      "step": 13020
+    },
+    {
+      "epoch": 0.5717861180938999,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.9508784915091057e-05,
+      "loss": 0.3138,
+      "step": 13022
+    },
+    {
+      "epoch": 0.5718739365731034,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.9502035792713026e-05,
+      "loss": 0.3385,
+      "step": 13024
+    },
+    {
+      "epoch": 0.5719617550523068,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.9495287091370414e-05,
+      "loss": 0.3446,
+      "step": 13026
+    },
+    {
+      "epoch": 0.5720495735315103,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.9488538811580033e-05,
+      "loss": 0.3483,
+      "step": 13028
+    },
+    {
+      "epoch": 0.5721373920107139,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.948179095385867e-05,
+      "loss": 0.3579,
+      "step": 13030
+    },
+    {
+      "epoch": 0.5722252104899174,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.947504351872308e-05,
+      "loss": 0.3803,
+      "step": 13032
+    },
+    {
+      "epoch": 0.5723130289691208,
+      "grad_norm": 1.75,
+      "learning_rate": 1.946829650668998e-05,
+      "loss": 0.3694,
+      "step": 13034
+    },
+    {
+      "epoch": 0.5724008474483243,
+      "grad_norm": 1.90625,
+      "learning_rate": 1.9461549918276066e-05,
+      "loss": 0.3297,
+      "step": 13036
+    },
+    {
+      "epoch": 0.5724886659275278,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.945480375399798e-05,
+      "loss": 0.3655,
+      "step": 13038
+    },
+    {
+      "epoch": 0.5725764844067313,
+      "grad_norm": 2.03125,
+      "learning_rate": 1.944805801437236e-05,
+      "loss": 0.3315,
+      "step": 13040
+    },
+    {
+      "epoch": 0.5726643028859347,
+      "grad_norm": 1.9609375,
+      "learning_rate": 1.944131269991579e-05,
+      "loss": 0.3349,
+      "step": 13042
+    },
+    {
+      "epoch": 0.5727521213651383,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.9434567811144813e-05,
+      "loss": 0.3429,
+      "step": 13044
+    },
+    {
+      "epoch": 0.5728399398443418,
+      "grad_norm": 1.828125,
+      "learning_rate": 1.9427823348575972e-05,
+      "loss": 0.3571,
+      "step": 13046
+    },
+    {
+      "epoch": 0.5729277583235453,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.9421079312725755e-05,
+      "loss": 0.3535,
+      "step": 13048
+    },
+    {
+      "epoch": 0.5730155768027487,
+      "grad_norm": 1.9140625,
+      "learning_rate": 1.941433570411062e-05,
+      "loss": 0.298,
+      "step": 13050
+    },
+    {
+      "epoch": 0.5731033952819522,
+      "grad_norm": 1.75,
+      "learning_rate": 1.9407592523247002e-05,
+      "loss": 0.3675,
+      "step": 13052
+    },
+    {
+      "epoch": 0.5731912137611557,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.940084977065128e-05,
+      "loss": 0.3698,
+      "step": 13054
+    },
+    {
+      "epoch": 0.5732790322403591,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.9394107446839828e-05,
+      "loss": 0.3116,
+      "step": 13056
+    },
+    {
+      "epoch": 0.5733668507195626,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.9387365552328973e-05,
+      "loss": 0.3263,
+      "step": 13058
+    },
+    {
+      "epoch": 0.5734546691987662,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.938062408763501e-05,
+      "loss": 0.3317,
+      "step": 13060
+    },
+    {
+      "epoch": 0.5735424876779697,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.9373883053274206e-05,
+      "loss": 0.346,
+      "step": 13062
+    },
+    {
+      "epoch": 0.5736303061571731,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.936714244976278e-05,
+      "loss": 0.3404,
+      "step": 13064
+    },
+    {
+      "epoch": 0.5737181246363766,
+      "grad_norm": 1.7578125,
+      "learning_rate": 1.9360402277616936e-05,
+      "loss": 0.3712,
+      "step": 13066
+    },
+    {
+      "epoch": 0.5738059431155801,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.9353662537352833e-05,
+      "loss": 0.3626,
+      "step": 13068
+    },
+    {
+      "epoch": 0.5738937615947836,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.9346923229486625e-05,
+      "loss": 0.3185,
+      "step": 13070
+    },
+    {
+      "epoch": 0.573981580073987,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.9340184354534376e-05,
+      "loss": 0.4062,
+      "step": 13072
+    },
+    {
+      "epoch": 0.5740693985531905,
+      "grad_norm": 1.8203125,
+      "learning_rate": 1.9333445913012165e-05,
+      "loss": 0.3457,
+      "step": 13074
+    },
+    {
+      "epoch": 0.5741572170323941,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.9326707905436026e-05,
+      "loss": 0.3193,
+      "step": 13076
+    },
+    {
+      "epoch": 0.5742450355115976,
+      "grad_norm": 1.8046875,
+      "learning_rate": 1.931997033232196e-05,
+      "loss": 0.3622,
+      "step": 13078
+    },
+    {
+      "epoch": 0.574332853990801,
+      "grad_norm": 1.765625,
+      "learning_rate": 1.9313233194185925e-05,
+      "loss": 0.336,
+      "step": 13080
+    },
+    {
+      "epoch": 0.5744206724700045,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.9306496491543853e-05,
+      "loss": 0.3208,
+      "step": 13082
+    },
+    {
+      "epoch": 0.574508490949208,
+      "grad_norm": 1.7578125,
+      "learning_rate": 1.9299760224911644e-05,
+      "loss": 0.3113,
+      "step": 13084
+    },
+    {
+      "epoch": 0.5745963094284114,
+      "grad_norm": 1.8125,
+      "learning_rate": 1.9293024394805166e-05,
+      "loss": 0.3527,
+      "step": 13086
+    },
+    {
+      "epoch": 0.5746841279076149,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.928628900174025e-05,
+      "loss": 0.3387,
+      "step": 13088
+    },
+    {
+      "epoch": 0.5747719463868185,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.927955404623269e-05,
+      "loss": 0.3336,
+      "step": 13090
+    },
+    {
+      "epoch": 0.574859764866022,
+      "grad_norm": 1.765625,
+      "learning_rate": 1.927281952879825e-05,
+      "loss": 0.3422,
+      "step": 13092
+    },
+    {
+      "epoch": 0.5749475833452254,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.9266085449952664e-05,
+      "loss": 0.3387,
+      "step": 13094
+    },
+    {
+      "epoch": 0.5750354018244289,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.925935181021163e-05,
+      "loss": 0.3232,
+      "step": 13096
+    },
+    {
+      "epoch": 0.5751232203036324,
+      "grad_norm": 1.8203125,
+      "learning_rate": 1.925261861009081e-05,
+      "loss": 0.3756,
+      "step": 13098
+    },
+    {
+      "epoch": 0.5752110387828359,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.9245885850105834e-05,
+      "loss": 0.3472,
+      "step": 13100
+    },
+    {
+      "epoch": 0.5752988572620393,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.9239153530772307e-05,
+      "loss": 0.3247,
+      "step": 13102
+    },
+    {
+      "epoch": 0.5753866757412428,
+      "grad_norm": 1.7578125,
+      "learning_rate": 1.923242165260578e-05,
+      "loss": 0.3371,
+      "step": 13104
+    },
+    {
+      "epoch": 0.5754744942204464,
+      "grad_norm": 1.8515625,
+      "learning_rate": 1.9225690216121782e-05,
+      "loss": 0.3503,
+      "step": 13106
+    },
+    {
+      "epoch": 0.5755623126996499,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.921895922183581e-05,
+      "loss": 0.3884,
+      "step": 13108
+    },
+    {
+      "epoch": 0.5756501311788533,
+      "grad_norm": 1.875,
+      "learning_rate": 1.9212228670263326e-05,
+      "loss": 0.322,
+      "step": 13110
+    },
+    {
+      "epoch": 0.5757379496580568,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.920549856191976e-05,
+      "loss": 0.3361,
+      "step": 13112
+    },
+    {
+      "epoch": 0.5758257681372603,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.9198768897320503e-05,
+      "loss": 0.3715,
+      "step": 13114
+    },
+    {
+      "epoch": 0.5759135866164637,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.9192039676980917e-05,
+      "loss": 0.3504,
+      "step": 13116
+    },
+    {
+      "epoch": 0.5760014050956672,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.918531090141632e-05,
+      "loss": 0.3315,
+      "step": 13118
+    },
+    {
+      "epoch": 0.5760892235748707,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.9178582571142008e-05,
+      "loss": 0.3459,
+      "step": 13120
+    },
+    {
+      "epoch": 0.5761770420540743,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.917185468667324e-05,
+      "loss": 0.3454,
+      "step": 13122
+    },
+    {
+      "epoch": 0.5762648605332777,
+      "grad_norm": 1.78125,
+      "learning_rate": 1.916512724852524e-05,
+      "loss": 0.3098,
+      "step": 13124
+    },
+    {
+      "epoch": 0.5763526790124812,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.9158400257213195e-05,
+      "loss": 0.321,
+      "step": 13126
+    },
+    {
+      "epoch": 0.5764404974916847,
+      "grad_norm": 1.84375,
+      "learning_rate": 1.9151673713252253e-05,
+      "loss": 0.3499,
+      "step": 13128
+    },
+    {
+      "epoch": 0.5765283159708882,
+      "grad_norm": 1.84375,
+      "learning_rate": 1.914494761715754e-05,
+      "loss": 0.3356,
+      "step": 13130
+    },
+    {
+      "epoch": 0.5766161344500916,
+      "grad_norm": 1.859375,
+      "learning_rate": 1.9138221969444153e-05,
+      "loss": 0.3424,
+      "step": 13132
+    },
+    {
+      "epoch": 0.5767039529292951,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.9131496770627123e-05,
+      "loss": 0.3522,
+      "step": 13134
+    },
+    {
+      "epoch": 0.5767917714084987,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.9124772021221476e-05,
+      "loss": 0.3515,
+      "step": 13136
+    },
+    {
+      "epoch": 0.5768795898877022,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.91180477217422e-05,
+      "loss": 0.3192,
+      "step": 13138
+    },
+    {
+      "epoch": 0.5769674083669056,
+      "grad_norm": 1.75,
+      "learning_rate": 1.911132387270424e-05,
+      "loss": 0.353,
+      "step": 13140
+    },
+    {
+      "epoch": 0.5770552268461091,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.9104600474622505e-05,
+      "loss": 0.3523,
+      "step": 13142
+    },
+    {
+      "epoch": 0.5771430453253126,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.9097877528011882e-05,
+      "loss": 0.3311,
+      "step": 13144
+    },
+    {
+      "epoch": 0.577230863804516,
+      "grad_norm": 1.7890625,
+      "learning_rate": 1.9091155033387208e-05,
+      "loss": 0.3442,
+      "step": 13146
+    },
+    {
+      "epoch": 0.5773186822837195,
+      "grad_norm": 1.8125,
+      "learning_rate": 1.9084432991263302e-05,
+      "loss": 0.3278,
+      "step": 13148
+    },
+    {
+      "epoch": 0.577406500762923,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.9077711402154942e-05,
+      "loss": 0.3689,
+      "step": 13150
+    },
+    {
+      "epoch": 0.5774943192421266,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.907099026657685e-05,
+      "loss": 0.3226,
+      "step": 13152
+    },
+    {
+      "epoch": 0.57758213772133,
+      "grad_norm": 1.7734375,
+      "learning_rate": 1.906426958504375e-05,
+      "loss": 0.3529,
+      "step": 13154
+    },
+    {
+      "epoch": 0.5776699562005335,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.905754935807031e-05,
+      "loss": 0.3441,
+      "step": 13156
+    },
+    {
+      "epoch": 0.577757774679737,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.905082958617116e-05,
+      "loss": 0.3536,
+      "step": 13158
+    },
+    {
+      "epoch": 0.5778455931589405,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.9044110269860916e-05,
+      "loss": 0.3286,
+      "step": 13160
+    },
+    {
+      "epoch": 0.5779334116381439,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.9037391409654133e-05,
+      "loss": 0.3429,
+      "step": 13162
+    },
+    {
+      "epoch": 0.5780212301173474,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.9030673006065346e-05,
+      "loss": 0.3406,
+      "step": 13164
+    },
+    {
+      "epoch": 0.5781090485965509,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.9023955059609042e-05,
+      "loss": 0.3143,
+      "step": 13166
+    },
+    {
+      "epoch": 0.5781968670757545,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.9017237570799697e-05,
+      "loss": 0.3395,
+      "step": 13168
+    },
+    {
+      "epoch": 0.5782846855549579,
+      "grad_norm": 1.8671875,
+      "learning_rate": 1.901052054015173e-05,
+      "loss": 0.3471,
+      "step": 13170
+    },
+    {
+      "epoch": 0.5783725040341614,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.9003803968179537e-05,
+      "loss": 0.3708,
+      "step": 13172
+    },
+    {
+      "epoch": 0.5784603225133649,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.899708785539747e-05,
+      "loss": 0.3453,
+      "step": 13174
+    },
+    {
+      "epoch": 0.5785481409925683,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.8990372202319864e-05,
+      "loss": 0.3442,
+      "step": 13176
+    },
+    {
+      "epoch": 0.5786359594717718,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.898365700946099e-05,
+      "loss": 0.3297,
+      "step": 13178
+    },
+    {
+      "epoch": 0.5787237779509753,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.8976942277335103e-05,
+      "loss": 0.3381,
+      "step": 13180
+    },
+    {
+      "epoch": 0.5788115964301788,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.897022800645642e-05,
+      "loss": 0.335,
+      "step": 13182
+    },
+    {
+      "epoch": 0.5788994149093823,
+      "grad_norm": 1.8671875,
+      "learning_rate": 1.896351419733912e-05,
+      "loss": 0.3333,
+      "step": 13184
+    },
+    {
+      "epoch": 0.5789872333885858,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.895680085049736e-05,
+      "loss": 0.3333,
+      "step": 13186
+    },
+    {
+      "epoch": 0.5790750518677893,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.895008796644523e-05,
+      "loss": 0.3366,
+      "step": 13188
+    },
+    {
+      "epoch": 0.5791628703469928,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.8943375545696816e-05,
+      "loss": 0.3414,
+      "step": 13190
+    },
+    {
+      "epoch": 0.5792506888261962,
+      "grad_norm": 1.7890625,
+      "learning_rate": 1.8936663588766154e-05,
+      "loss": 0.3546,
+      "step": 13192
+    },
+    {
+      "epoch": 0.5793385073053997,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.8929952096167258e-05,
+      "loss": 0.3291,
+      "step": 13194
+    },
+    {
+      "epoch": 0.5794263257846032,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.8923241068414076e-05,
+      "loss": 0.3341,
+      "step": 13196
+    },
+    {
+      "epoch": 0.5795141442638068,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.891653050602055e-05,
+      "loss": 0.3349,
+      "step": 13198
+    },
+    {
+      "epoch": 0.5796019627430102,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.8909820409500573e-05,
+      "loss": 0.3762,
+      "step": 13200
+    },
+    {
+      "epoch": 0.5796897812222137,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.8903110779368018e-05,
+      "loss": 0.3333,
+      "step": 13202
+    },
+    {
+      "epoch": 0.5797775997014172,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.8896401616136692e-05,
+      "loss": 0.3159,
+      "step": 13204
+    },
+    {
+      "epoch": 0.5798654181806207,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.8889692920320397e-05,
+      "loss": 0.3489,
+      "step": 13206
+    },
+    {
+      "epoch": 0.5799532366598241,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.888298469243288e-05,
+      "loss": 0.3647,
+      "step": 13208
+    },
+    {
+      "epoch": 0.5800410551390276,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.8876276932987864e-05,
+      "loss": 0.3274,
+      "step": 13210
+    },
+    {
+      "epoch": 0.5801288736182311,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.8869569642499032e-05,
+      "loss": 0.328,
+      "step": 13212
+    },
+    {
+      "epoch": 0.5802166920974346,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.8862862821480025e-05,
+      "loss": 0.3536,
+      "step": 13214
+    },
+    {
+      "epoch": 0.5803045105766381,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.885615647044445e-05,
+      "loss": 0.331,
+      "step": 13216
+    },
+    {
+      "epoch": 0.5803923290558416,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.8849450589905887e-05,
+      "loss": 0.3362,
+      "step": 13218
+    },
+    {
+      "epoch": 0.5804801475350451,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.884274518037787e-05,
+      "loss": 0.3125,
+      "step": 13220
+    },
+    {
+      "epoch": 0.5805679660142485,
+      "grad_norm": 1.8828125,
+      "learning_rate": 1.883604024237391e-05,
+      "loss": 0.3284,
+      "step": 13222
+    },
+    {
+      "epoch": 0.580655784493452,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.8829335776407464e-05,
+      "loss": 0.3651,
+      "step": 13224
+    },
+    {
+      "epoch": 0.5807436029726555,
+      "grad_norm": 1.8671875,
+      "learning_rate": 1.8822631782991967e-05,
+      "loss": 0.337,
+      "step": 13226
+    },
+    {
+      "epoch": 0.580831421451859,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.8815928262640804e-05,
+      "loss": 0.3373,
+      "step": 13228
+    },
+    {
+      "epoch": 0.5809192399310625,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.8809225215867334e-05,
+      "loss": 0.3681,
+      "step": 13230
+    },
+    {
+      "epoch": 0.581007058410266,
+      "grad_norm": 1.8046875,
+      "learning_rate": 1.8802522643184882e-05,
+      "loss": 0.3509,
+      "step": 13232
+    },
+    {
+      "epoch": 0.5810948768894695,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.879582054510673e-05,
+      "loss": 0.3459,
+      "step": 13234
+    },
+    {
+      "epoch": 0.581182695368673,
+      "grad_norm": 1.8046875,
+      "learning_rate": 1.878911892214613e-05,
+      "loss": 0.3446,
+      "step": 13236
+    },
+    {
+      "epoch": 0.5812705138478764,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.8782417774816295e-05,
+      "loss": 0.3414,
+      "step": 13238
+    },
+    {
+      "epoch": 0.5813583323270799,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.8775717103630395e-05,
+      "loss": 0.3279,
+      "step": 13240
+    },
+    {
+      "epoch": 0.5814461508062834,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.876901690910157e-05,
+      "loss": 0.3356,
+      "step": 13242
+    },
+    {
+      "epoch": 0.581533969285487,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.8762317191742924e-05,
+      "loss": 0.3271,
+      "step": 13244
+    },
+    {
+      "epoch": 0.5816217877646904,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.875561795206752e-05,
+      "loss": 0.3237,
+      "step": 13246
+    },
+    {
+      "epoch": 0.5817096062438939,
+      "grad_norm": 1.7578125,
+      "learning_rate": 1.8748919190588398e-05,
+      "loss": 0.3382,
+      "step": 13248
+    },
+    {
+      "epoch": 0.5817974247230974,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.8742220907818535e-05,
+      "loss": 0.3319,
+      "step": 13250
+    },
+    {
+      "epoch": 0.5818852432023008,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.87355231042709e-05,
+      "loss": 0.3463,
+      "step": 13252
+    },
+    {
+      "epoch": 0.5819730616815043,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.8728825780458415e-05,
+      "loss": 0.3481,
+      "step": 13254
+    },
+    {
+      "epoch": 0.5820608801607078,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.8722128936893946e-05,
+      "loss": 0.3337,
+      "step": 13256
+    },
+    {
+      "epoch": 0.5821486986399113,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.8715432574090344e-05,
+      "loss": 0.3141,
+      "step": 13258
+    },
+    {
+      "epoch": 0.5822365171191148,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.8708736692560424e-05,
+      "loss": 0.3838,
+      "step": 13260
+    },
+    {
+      "epoch": 0.5823243355983183,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.8702041292816963e-05,
+      "loss": 0.3102,
+      "step": 13262
+    },
+    {
+      "epoch": 0.5824121540775218,
+      "grad_norm": 1.796875,
+      "learning_rate": 1.8695346375372686e-05,
+      "loss": 0.3297,
+      "step": 13264
+    },
+    {
+      "epoch": 0.5824999725567253,
+      "grad_norm": 1.7734375,
+      "learning_rate": 1.868865194074029e-05,
+      "loss": 0.3576,
+      "step": 13266
+    },
+    {
+      "epoch": 0.5825877910359287,
+      "grad_norm": 1.953125,
+      "learning_rate": 1.868195798943244e-05,
+      "loss": 0.3449,
+      "step": 13268
+    },
+    {
+      "epoch": 0.5826756095151322,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.8675264521961765e-05,
+      "loss": 0.3412,
+      "step": 13270
+    },
+    {
+      "epoch": 0.5827634279943357,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.866857153884085e-05,
+      "loss": 0.3659,
+      "step": 13272
+    },
+    {
+      "epoch": 0.5828512464735391,
+      "grad_norm": 1.7734375,
+      "learning_rate": 1.8661879040582254e-05,
+      "loss": 0.3646,
+      "step": 13274
+    },
+    {
+      "epoch": 0.5829390649527427,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.8655187027698467e-05,
+      "loss": 0.3396,
+      "step": 13276
+    },
+    {
+      "epoch": 0.5830268834319462,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.864849550070198e-05,
+      "loss": 0.3503,
+      "step": 13278
+    },
+    {
+      "epoch": 0.5831147019111497,
+      "grad_norm": 1.8125,
+      "learning_rate": 1.8641804460105233e-05,
+      "loss": 0.3492,
+      "step": 13280
+    },
+    {
+      "epoch": 0.5832025203903531,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.8635113906420622e-05,
+      "loss": 0.3658,
+      "step": 13282
+    },
+    {
+      "epoch": 0.5832903388695566,
+      "grad_norm": 1.796875,
+      "learning_rate": 1.862842384016052e-05,
+      "loss": 0.3581,
+      "step": 13284
+    },
+    {
+      "epoch": 0.5833781573487601,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.862173426183725e-05,
+      "loss": 0.3505,
+      "step": 13286
+    },
+    {
+      "epoch": 0.5834659758279636,
+      "grad_norm": 1.9765625,
+      "learning_rate": 1.8615045171963098e-05,
+      "loss": 0.3254,
+      "step": 13288
+    },
+    {
+      "epoch": 0.5835537943071671,
+      "grad_norm": 1.8515625,
+      "learning_rate": 1.860835657105031e-05,
+      "loss": 0.318,
+      "step": 13290
+    },
+    {
+      "epoch": 0.5836416127863706,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.8601668459611106e-05,
+      "loss": 0.3535,
+      "step": 13292
+    },
+    {
+      "epoch": 0.5837294312655741,
+      "grad_norm": 1.859375,
+      "learning_rate": 1.859498083815767e-05,
+      "loss": 0.3671,
+      "step": 13294
+    },
+    {
+      "epoch": 0.5838172497447776,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.8588293707202133e-05,
+      "loss": 0.3476,
+      "step": 13296
+    },
+    {
+      "epoch": 0.583905068223981,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.858160706725661e-05,
+      "loss": 0.3515,
+      "step": 13298
+    },
+    {
+      "epoch": 0.5839928867031845,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.857492091883315e-05,
+      "loss": 0.3621,
+      "step": 13300
+    },
+    {
+      "epoch": 0.584080705182388,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.8568235262443782e-05,
+      "loss": 0.3171,
+      "step": 13302
+    },
+    {
+      "epoch": 0.5841685236615914,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.8561550098600505e-05,
+      "loss": 0.351,
+      "step": 13304
+    },
+    {
+      "epoch": 0.584256342140795,
+      "grad_norm": 2.046875,
+      "learning_rate": 1.8554865427815262e-05,
+      "loss": 0.3387,
+      "step": 13306
+    },
+    {
+      "epoch": 0.5843441606199985,
+      "grad_norm": 1.765625,
+      "learning_rate": 1.854818125059997e-05,
+      "loss": 0.3562,
+      "step": 13308
+    },
+    {
+      "epoch": 0.584431979099202,
+      "grad_norm": 1.8984375,
+      "learning_rate": 1.8541497567466508e-05,
+      "loss": 0.3394,
+      "step": 13310
+    },
+    {
+      "epoch": 0.5845197975784054,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.853481437892671e-05,
+      "loss": 0.3271,
+      "step": 13312
+    },
+    {
+      "epoch": 0.5846076160576089,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.852813168549237e-05,
+      "loss": 0.3523,
+      "step": 13314
+    },
+    {
+      "epoch": 0.5846954345368124,
+      "grad_norm": 1.75,
+      "learning_rate": 1.8521449487675275e-05,
+      "loss": 0.3484,
+      "step": 13316
+    },
+    {
+      "epoch": 0.5847832530160159,
+      "grad_norm": 1.7734375,
+      "learning_rate": 1.851476778598712e-05,
+      "loss": 0.3407,
+      "step": 13318
+    },
+    {
+      "epoch": 0.5848710714952193,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.8508086580939598e-05,
+      "loss": 0.3632,
+      "step": 13320
+    },
+    {
+      "epoch": 0.5849588899744229,
+      "grad_norm": 1.859375,
+      "learning_rate": 1.8501405873044363e-05,
+      "loss": 0.3391,
+      "step": 13322
+    },
+    {
+      "epoch": 0.5850467084536264,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.8494725662813028e-05,
+      "loss": 0.3268,
+      "step": 13324
+    },
+    {
+      "epoch": 0.5851345269328299,
+      "grad_norm": 1.9609375,
+      "learning_rate": 1.848804595075716e-05,
+      "loss": 0.3522,
+      "step": 13326
+    },
+    {
+      "epoch": 0.5852223454120333,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.848136673738829e-05,
+      "loss": 0.316,
+      "step": 13328
+    },
+    {
+      "epoch": 0.5853101638912368,
+      "grad_norm": 1.8515625,
+      "learning_rate": 1.847468802321792e-05,
+      "loss": 0.3361,
+      "step": 13330
+    },
+    {
+      "epoch": 0.5853979823704403,
+      "grad_norm": 1.7734375,
+      "learning_rate": 1.8468009808757505e-05,
+      "loss": 0.3726,
+      "step": 13332
+    },
+    {
+      "epoch": 0.5854858008496437,
+      "grad_norm": 1.828125,
+      "learning_rate": 1.8461332094518465e-05,
+      "loss": 0.345,
+      "step": 13334
+    },
+    {
+      "epoch": 0.5855736193288473,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.8454654881012184e-05,
+      "loss": 0.3274,
+      "step": 13336
+    },
+    {
+      "epoch": 0.5856614378080508,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.8447978168750003e-05,
+      "loss": 0.3359,
+      "step": 13338
+    },
+    {
+      "epoch": 0.5857492562872543,
+      "grad_norm": 1.90625,
+      "learning_rate": 1.844130195824322e-05,
+      "loss": 0.387,
+      "step": 13340
+    },
+    {
+      "epoch": 0.5858370747664577,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.84346262500031e-05,
+      "loss": 0.3119,
+      "step": 13342
+    },
+    {
+      "epoch": 0.5859248932456612,
+      "grad_norm": 1.90625,
+      "learning_rate": 1.842795104454088e-05,
+      "loss": 0.3082,
+      "step": 13344
+    },
+    {
+      "epoch": 0.5860127117248647,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.8421276342367756e-05,
+      "loss": 0.3793,
+      "step": 13346
+    },
+    {
+      "epoch": 0.5861005302040682,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.8414602143994856e-05,
+      "loss": 0.3599,
+      "step": 13348
+    },
+    {
+      "epoch": 0.5861883486832716,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.8407928449933304e-05,
+      "loss": 0.3041,
+      "step": 13350
+    },
+    {
+      "epoch": 0.5862761671624752,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.8401255260694173e-05,
+      "loss": 0.315,
+      "step": 13352
+    },
+    {
+      "epoch": 0.5863639856416787,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.8394582576788487e-05,
+      "loss": 0.3373,
+      "step": 13354
+    },
+    {
+      "epoch": 0.5864518041208822,
+      "grad_norm": 1.7578125,
+      "learning_rate": 1.838791039872726e-05,
+      "loss": 0.3263,
+      "step": 13356
+    },
+    {
+      "epoch": 0.5865396226000856,
+      "grad_norm": 1.90625,
+      "learning_rate": 1.8381238727021433e-05,
+      "loss": 0.3309,
+      "step": 13358
+    },
+    {
+      "epoch": 0.5866274410792891,
+      "grad_norm": 1.8515625,
+      "learning_rate": 1.837456756218194e-05,
+      "loss": 0.3174,
+      "step": 13360
+    },
+    {
+      "epoch": 0.5867152595584926,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.8367896904719645e-05,
+      "loss": 0.3269,
+      "step": 13362
+    },
+    {
+      "epoch": 0.586803078037696,
+      "grad_norm": 1.8515625,
+      "learning_rate": 1.8361226755145394e-05,
+      "loss": 0.3741,
+      "step": 13364
+    },
+    {
+      "epoch": 0.5868908965168995,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.835455711396999e-05,
+      "loss": 0.3335,
+      "step": 13366
+    },
+    {
+      "epoch": 0.5869787149961031,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.8347887981704195e-05,
+      "loss": 0.3346,
+      "step": 13368
+    },
+    {
+      "epoch": 0.5870665334753066,
+      "grad_norm": 1.7890625,
+      "learning_rate": 1.8341219358858734e-05,
+      "loss": 0.3419,
+      "step": 13370
+    },
+    {
+      "epoch": 0.58715435195451,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.833455124594429e-05,
+      "loss": 0.3505,
+      "step": 13372
+    },
+    {
+      "epoch": 0.5872421704337135,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.8327883643471514e-05,
+      "loss": 0.3448,
+      "step": 13374
+    },
+    {
+      "epoch": 0.587329988912917,
+      "grad_norm": 1.765625,
+      "learning_rate": 1.8321216551951002e-05,
+      "loss": 0.3481,
+      "step": 13376
+    },
+    {
+      "epoch": 0.5874178073921205,
+      "grad_norm": 1.828125,
+      "learning_rate": 1.8314549971893342e-05,
+      "loss": 0.3309,
+      "step": 13378
+    },
+    {
+      "epoch": 0.5875056258713239,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.8307883903809035e-05,
+      "loss": 0.3391,
+      "step": 13380
+    },
+    {
+      "epoch": 0.5875934443505274,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.8301218348208583e-05,
+      "loss": 0.3251,
+      "step": 13382
+    },
+    {
+      "epoch": 0.587681262829731,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.8294553305602434e-05,
+      "loss": 0.3132,
+      "step": 13384
+    },
+    {
+      "epoch": 0.5877690813089345,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.8287888776501007e-05,
+      "loss": 0.3206,
+      "step": 13386
+    },
+    {
+      "epoch": 0.5878568997881379,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.828122476141466e-05,
+      "loss": 0.3317,
+      "step": 13388
+    },
+    {
+      "epoch": 0.5879447182673414,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.8274561260853733e-05,
+      "loss": 0.3096,
+      "step": 13390
+    },
+    {
+      "epoch": 0.5880325367465449,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.8267898275328516e-05,
+      "loss": 0.3219,
+      "step": 13392
+    },
+    {
+      "epoch": 0.5881203552257483,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.8261235805349263e-05,
+      "loss": 0.322,
+      "step": 13394
+    },
+    {
+      "epoch": 0.5882081737049518,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.8254573851426195e-05,
+      "loss": 0.354,
+      "step": 13396
+    },
+    {
+      "epoch": 0.5882959921841554,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.824791241406947e-05,
+      "loss": 0.3446,
+      "step": 13398
+    },
+    {
+      "epoch": 0.5883838106633589,
+      "grad_norm": 1.75,
+      "learning_rate": 1.824125149378923e-05,
+      "loss": 0.3736,
+      "step": 13400
+    },
+    {
+      "epoch": 0.5884716291425623,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.823459109109557e-05,
+      "loss": 0.3377,
+      "step": 13402
+    },
+    {
+      "epoch": 0.5885594476217658,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.822793120649855e-05,
+      "loss": 0.3308,
+      "step": 13404
+    },
+    {
+      "epoch": 0.5886472661009693,
+      "grad_norm": 1.828125,
+      "learning_rate": 1.822127184050818e-05,
+      "loss": 0.3304,
+      "step": 13406
+    },
+    {
+      "epoch": 0.5887350845801728,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.821461299363445e-05,
+      "loss": 0.3367,
+      "step": 13408
+    },
+    {
+      "epoch": 0.5888229030593762,
+      "grad_norm": 1.8046875,
+      "learning_rate": 1.820795466638727e-05,
+      "loss": 0.3298,
+      "step": 13410
+    },
+    {
+      "epoch": 0.5889107215385797,
+      "grad_norm": 1.7734375,
+      "learning_rate": 1.8201296859276555e-05,
+      "loss": 0.3305,
+      "step": 13412
+    },
+    {
+      "epoch": 0.5889985400177833,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.8194639572812157e-05,
+      "loss": 0.3244,
+      "step": 13414
+    },
+    {
+      "epoch": 0.5890863584969868,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.8187982807503885e-05,
+      "loss": 0.3538,
+      "step": 13416
+    },
+    {
+      "epoch": 0.5891741769761902,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.818132656386152e-05,
+      "loss": 0.3192,
+      "step": 13418
+    },
+    {
+      "epoch": 0.5892619954553937,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.817467084239481e-05,
+      "loss": 0.3559,
+      "step": 13420
+    },
+    {
+      "epoch": 0.5893498139345972,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.8168015643613445e-05,
+      "loss": 0.333,
+      "step": 13422
+    },
+    {
+      "epoch": 0.5894376324138006,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.8161360968027073e-05,
+      "loss": 0.324,
+      "step": 13424
+    },
+    {
+      "epoch": 0.5895254508930041,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.8154706816145315e-05,
+      "loss": 0.3465,
+      "step": 13426
+    },
+    {
+      "epoch": 0.5896132693722076,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.8148053188477754e-05,
+      "loss": 0.3364,
+      "step": 13428
+    },
+    {
+      "epoch": 0.5897010878514112,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.814140008553392e-05,
+      "loss": 0.3523,
+      "step": 13430
+    },
+    {
+      "epoch": 0.5897889063306146,
+      "grad_norm": 1.9375,
+      "learning_rate": 1.8134747507823323e-05,
+      "loss": 0.3177,
+      "step": 13432
+    },
+    {
+      "epoch": 0.5898767248098181,
+      "grad_norm": 1.7734375,
+      "learning_rate": 1.8128095455855395e-05,
+      "loss": 0.3485,
+      "step": 13434
+    },
+    {
+      "epoch": 0.5899645432890216,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.8121443930139568e-05,
+      "loss": 0.3294,
+      "step": 13436
+    },
+    {
+      "epoch": 0.5900523617682251,
+      "grad_norm": 1.8046875,
+      "learning_rate": 1.8114792931185223e-05,
+      "loss": 0.3434,
+      "step": 13438
+    },
+    {
+      "epoch": 0.5901401802474285,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.8108142459501674e-05,
+      "loss": 0.3251,
+      "step": 13440
+    },
+    {
+      "epoch": 0.590227998726632,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.810149251559823e-05,
+      "loss": 0.3167,
+      "step": 13442
+    },
+    {
+      "epoch": 0.5903158172058356,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.8094843099984144e-05,
+      "loss": 0.3115,
+      "step": 13444
+    },
+    {
+      "epoch": 0.5904036356850391,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.8088194213168626e-05,
+      "loss": 0.34,
+      "step": 13446
+    },
+    {
+      "epoch": 0.5904914541642425,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.8081545855660858e-05,
+      "loss": 0.3425,
+      "step": 13448
+    },
+    {
+      "epoch": 0.590579272643446,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.807489802796996e-05,
+      "loss": 0.3633,
+      "step": 13450
+    },
+    {
+      "epoch": 0.5906670911226495,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.8068250730605034e-05,
+      "loss": 0.332,
+      "step": 13452
+    },
+    {
+      "epoch": 0.590754909601853,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.8061603964075125e-05,
+      "loss": 0.3062,
+      "step": 13454
+    },
+    {
+      "epoch": 0.5908427280810564,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.805495772888925e-05,
+      "loss": 0.3278,
+      "step": 13456
+    },
+    {
+      "epoch": 0.5909305465602599,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.8048312025556384e-05,
+      "loss": 0.3581,
+      "step": 13458
+    },
+    {
+      "epoch": 0.5910183650394635,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.8041666854585443e-05,
+      "loss": 0.3449,
+      "step": 13460
+    },
+    {
+      "epoch": 0.591106183518667,
+      "grad_norm": 1.8203125,
+      "learning_rate": 1.803502221648532e-05,
+      "loss": 0.3533,
+      "step": 13462
+    },
+    {
+      "epoch": 0.5911940019978704,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.8028378111764864e-05,
+      "loss": 0.2833,
+      "step": 13464
+    },
+    {
+      "epoch": 0.5912818204770739,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.802173454093289e-05,
+      "loss": 0.338,
+      "step": 13466
+    },
+    {
+      "epoch": 0.5913696389562774,
+      "grad_norm": 1.8828125,
+      "learning_rate": 1.801509150449815e-05,
+      "loss": 0.3585,
+      "step": 13468
+    },
+    {
+      "epoch": 0.5914574574354808,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.8008449002969395e-05,
+      "loss": 0.3445,
+      "step": 13470
+    },
+    {
+      "epoch": 0.5915452759146843,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.8001807036855278e-05,
+      "loss": 0.3491,
+      "step": 13472
+    },
+    {
+      "epoch": 0.5916330943938878,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.799516560666446e-05,
+      "loss": 0.3108,
+      "step": 13474
+    },
+    {
+      "epoch": 0.5917209128730914,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.7988524712905532e-05,
+      "loss": 0.3521,
+      "step": 13476
+    },
+    {
+      "epoch": 0.5918087313522948,
+      "grad_norm": 1.7890625,
+      "learning_rate": 1.7981884356087065e-05,
+      "loss": 0.3642,
+      "step": 13478
+    },
+    {
+      "epoch": 0.5918965498314983,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.7975244536717576e-05,
+      "loss": 0.3198,
+      "step": 13480
+    },
+    {
+      "epoch": 0.5919843683107018,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.7968605255305542e-05,
+      "loss": 0.344,
+      "step": 13482
+    },
+    {
+      "epoch": 0.5920721867899053,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.7961966512359414e-05,
+      "loss": 0.3449,
+      "step": 13484
+    },
+    {
+      "epoch": 0.5921600052691087,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.7955328308387563e-05,
+      "loss": 0.3294,
+      "step": 13486
+    },
+    {
+      "epoch": 0.5922478237483122,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.794869064389837e-05,
+      "loss": 0.3317,
+      "step": 13488
+    },
+    {
+      "epoch": 0.5923356422275158,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.7942053519400133e-05,
+      "loss": 0.3468,
+      "step": 13490
+    },
+    {
+      "epoch": 0.5924234607067193,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.7935416935401128e-05,
+      "loss": 0.3326,
+      "step": 13492
+    },
+    {
+      "epoch": 0.5925112791859227,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.7928780892409592e-05,
+      "loss": 0.32,
+      "step": 13494
+    },
+    {
+      "epoch": 0.5925990976651262,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.7922145390933707e-05,
+      "loss": 0.3368,
+      "step": 13496
+    },
+    {
+      "epoch": 0.5926869161443297,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.7915510431481625e-05,
+      "loss": 0.3298,
+      "step": 13498
+    },
+    {
+      "epoch": 0.5927747346235331,
+      "grad_norm": 1.796875,
+      "learning_rate": 1.790887601456146e-05,
+      "loss": 0.3706,
+      "step": 13500
+    },
+    {
+      "epoch": 0.5928625531027366,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.7902242140681264e-05,
+      "loss": 0.3379,
+      "step": 13502
+    },
+    {
+      "epoch": 0.5929503715819401,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.7895608810349062e-05,
+      "loss": 0.3249,
+      "step": 13504
+    },
+    {
+      "epoch": 0.5930381900611437,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.788897602407284e-05,
+      "loss": 0.3453,
+      "step": 13506
+    },
+    {
+      "epoch": 0.5931260085403471,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.7882343782360546e-05,
+      "loss": 0.3356,
+      "step": 13508
+    },
+    {
+      "epoch": 0.5932138270195506,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.787571208572007e-05,
+      "loss": 0.3301,
+      "step": 13510
+    },
+    {
+      "epoch": 0.5933016454987541,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.7869080934659265e-05,
+      "loss": 0.3048,
+      "step": 13512
+    },
+    {
+      "epoch": 0.5933894639779576,
+      "grad_norm": 1.7890625,
+      "learning_rate": 1.7862450329685952e-05,
+      "loss": 0.3584,
+      "step": 13514
+    },
+    {
+      "epoch": 0.593477282457161,
+      "grad_norm": 1.625,
+      "learning_rate": 1.7855820271307906e-05,
+      "loss": 0.3314,
+      "step": 13516
+    },
+    {
+      "epoch": 0.5935651009363645,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.7849190760032853e-05,
+      "loss": 0.3615,
+      "step": 13518
+    },
+    {
+      "epoch": 0.593652919415568,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.7842561796368496e-05,
+      "loss": 0.3443,
+      "step": 13520
+    },
+    {
+      "epoch": 0.5937407378947716,
+      "grad_norm": 1.8828125,
+      "learning_rate": 1.7835933380822462e-05,
+      "loss": 0.3515,
+      "step": 13522
+    },
+    {
+      "epoch": 0.593828556373975,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.7829305513902366e-05,
+      "loss": 0.3048,
+      "step": 13524
+    },
+    {
+      "epoch": 0.5939163748531785,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.782267819611578e-05,
+      "loss": 0.3268,
+      "step": 13526
+    },
+    {
+      "epoch": 0.594004193332382,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.7816051427970213e-05,
+      "loss": 0.3424,
+      "step": 13528
+    },
+    {
+      "epoch": 0.5940920118115854,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.780942520997316e-05,
+      "loss": 0.3329,
+      "step": 13530
+    },
+    {
+      "epoch": 0.5941798302907889,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.780279954263204e-05,
+      "loss": 0.3417,
+      "step": 13532
+    },
+    {
+      "epoch": 0.5942676487699924,
+      "grad_norm": 1.859375,
+      "learning_rate": 1.7796174426454255e-05,
+      "loss": 0.3385,
+      "step": 13534
+    },
+    {
+      "epoch": 0.5943554672491959,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.7789549861947165e-05,
+      "loss": 0.3405,
+      "step": 13536
+    },
+    {
+      "epoch": 0.5944432857283994,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.778292584961807e-05,
+      "loss": 0.3766,
+      "step": 13538
+    },
+    {
+      "epoch": 0.5945311042076029,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.7776302389974243e-05,
+      "loss": 0.3446,
+      "step": 13540
+    },
+    {
+      "epoch": 0.5946189226868064,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.776967948352291e-05,
+      "loss": 0.318,
+      "step": 13542
+    },
+    {
+      "epoch": 0.5947067411660099,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.7763057130771254e-05,
+      "loss": 0.353,
+      "step": 13544
+    },
+    {
+      "epoch": 0.5947945596452133,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.7756435332226424e-05,
+      "loss": 0.3147,
+      "step": 13546
+    },
+    {
+      "epoch": 0.5948823781244168,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.774981408839551e-05,
+      "loss": 0.3275,
+      "step": 13548
+    },
+    {
+      "epoch": 0.5949701966036203,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.774319339978557e-05,
+      "loss": 0.3295,
+      "step": 13550
+    },
+    {
+      "epoch": 0.5950580150828239,
+      "grad_norm": 1.75,
+      "learning_rate": 1.7736573266903617e-05,
+      "loss": 0.3415,
+      "step": 13552
+    },
+    {
+      "epoch": 0.5951458335620273,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.7729953690256625e-05,
+      "loss": 0.3136,
+      "step": 13554
+    },
+    {
+      "epoch": 0.5952336520412308,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.772333467035153e-05,
+      "loss": 0.3453,
+      "step": 13556
+    },
+    {
+      "epoch": 0.5953214705204343,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.7716716207695202e-05,
+      "loss": 0.3261,
+      "step": 13558
+    },
+    {
+      "epoch": 0.5954092889996377,
+      "grad_norm": 2.015625,
+      "learning_rate": 1.7710098302794495e-05,
+      "loss": 0.3554,
+      "step": 13560
+    },
+    {
+      "epoch": 0.5954971074788412,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.7703480956156215e-05,
+      "loss": 0.3133,
+      "step": 13562
+    },
+    {
+      "epoch": 0.5955849259580447,
+      "grad_norm": 1.875,
+      "learning_rate": 1.7696864168287105e-05,
+      "loss": 0.3392,
+      "step": 13564
+    },
+    {
+      "epoch": 0.5956727444372482,
+      "grad_norm": 1.7734375,
+      "learning_rate": 1.7690247939693887e-05,
+      "loss": 0.3364,
+      "step": 13566
+    },
+    {
+      "epoch": 0.5957605629164517,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.7683632270883233e-05,
+      "loss": 0.3271,
+      "step": 13568
+    },
+    {
+      "epoch": 0.5958483813956552,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.7677017162361776e-05,
+      "loss": 0.3657,
+      "step": 13570
+    },
+    {
+      "epoch": 0.5959361998748587,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.7670402614636104e-05,
+      "loss": 0.327,
+      "step": 13572
+    },
+    {
+      "epoch": 0.5960240183540622,
+      "grad_norm": 1.625,
+      "learning_rate": 1.7663788628212752e-05,
+      "loss": 0.3258,
+      "step": 13574
+    },
+    {
+      "epoch": 0.5961118368332656,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.7657175203598222e-05,
+      "loss": 0.3347,
+      "step": 13576
+    },
+    {
+      "epoch": 0.5961996553124691,
+      "grad_norm": 1.7578125,
+      "learning_rate": 1.765056234129898e-05,
+      "loss": 0.3133,
+      "step": 13578
+    },
+    {
+      "epoch": 0.5962874737916726,
+      "grad_norm": 1.84375,
+      "learning_rate": 1.7643950041821434e-05,
+      "loss": 0.3587,
+      "step": 13580
+    },
+    {
+      "epoch": 0.596375292270876,
+      "grad_norm": 1.8125,
+      "learning_rate": 1.763733830567196e-05,
+      "loss": 0.3578,
+      "step": 13582
+    },
+    {
+      "epoch": 0.5964631107500796,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.763072713335688e-05,
+      "loss": 0.3336,
+      "step": 13584
+    },
+    {
+      "epoch": 0.5965509292292831,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.7624116525382482e-05,
+      "loss": 0.3035,
+      "step": 13586
+    },
+    {
+      "epoch": 0.5966387477084866,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.761750648225501e-05,
+      "loss": 0.33,
+      "step": 13588
+    },
+    {
+      "epoch": 0.59672656618769,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.7610897004480658e-05,
+      "loss": 0.2962,
+      "step": 13590
+    },
+    {
+      "epoch": 0.5968143846668935,
+      "grad_norm": 1.9453125,
+      "learning_rate": 1.76042880925656e-05,
+      "loss": 0.3426,
+      "step": 13592
+    },
+    {
+      "epoch": 0.596902203146097,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.7597679747015922e-05,
+      "loss": 0.3492,
+      "step": 13594
+    },
+    {
+      "epoch": 0.5969900216253005,
+      "grad_norm": 1.96875,
+      "learning_rate": 1.75910719683377e-05,
+      "loss": 0.3161,
+      "step": 13596
+    },
+    {
+      "epoch": 0.597077840104504,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.758446475703697e-05,
+      "loss": 0.3393,
+      "step": 13598
+    },
+    {
+      "epoch": 0.5971656585837075,
+      "grad_norm": 1.921875,
+      "learning_rate": 1.75778581136197e-05,
+      "loss": 0.3211,
+      "step": 13600
+    },
+    {
+      "epoch": 0.597253477062911,
+      "grad_norm": 1.828125,
+      "learning_rate": 1.7571252038591835e-05,
+      "loss": 0.3378,
+      "step": 13602
+    },
+    {
+      "epoch": 0.5973412955421145,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.7564646532459273e-05,
+      "loss": 0.34,
+      "step": 13604
+    },
+    {
+      "epoch": 0.5974291140213179,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.755804159572786e-05,
+      "loss": 0.3273,
+      "step": 13606
+    },
+    {
+      "epoch": 0.5975169325005214,
+      "grad_norm": 1.921875,
+      "learning_rate": 1.7551437228903407e-05,
+      "loss": 0.3273,
+      "step": 13608
+    },
+    {
+      "epoch": 0.5976047509797249,
+      "grad_norm": 1.796875,
+      "learning_rate": 1.7544833432491674e-05,
+      "loss": 0.3469,
+      "step": 13610
+    },
+    {
+      "epoch": 0.5976925694589283,
+      "grad_norm": 2.015625,
+      "learning_rate": 1.7538230206998386e-05,
+      "loss": 0.3423,
+      "step": 13612
+    },
+    {
+      "epoch": 0.5977803879381319,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.7531627552929214e-05,
+      "loss": 0.3512,
+      "step": 13614
+    },
+    {
+      "epoch": 0.5978682064173354,
+      "grad_norm": 1.8984375,
+      "learning_rate": 1.7525025470789797e-05,
+      "loss": 0.3472,
+      "step": 13616
+    },
+    {
+      "epoch": 0.5979560248965389,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.7518423961085725e-05,
+      "loss": 0.3479,
+      "step": 13618
+    },
+    {
+      "epoch": 0.5980438433757423,
+      "grad_norm": 1.890625,
+      "learning_rate": 1.7511823024322534e-05,
+      "loss": 0.3252,
+      "step": 13620
+    },
+    {
+      "epoch": 0.5981316618549458,
+      "grad_norm": 1.9453125,
+      "learning_rate": 1.7505222661005745e-05,
+      "loss": 0.3326,
+      "step": 13622
+    },
+    {
+      "epoch": 0.5982194803341493,
+      "grad_norm": 1.96875,
+      "learning_rate": 1.7498622871640785e-05,
+      "loss": 0.3116,
+      "step": 13624
+    },
+    {
+      "epoch": 0.5983072988133528,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.7492023656733085e-05,
+      "loss": 0.3341,
+      "step": 13626
+    },
+    {
+      "epoch": 0.5983951172925562,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.7485425016788016e-05,
+      "loss": 0.3458,
+      "step": 13628
+    },
+    {
+      "epoch": 0.5984829357717598,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.74788269523109e-05,
+      "loss": 0.3681,
+      "step": 13630
+    },
+    {
+      "epoch": 0.5985707542509633,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.747222946380702e-05,
+      "loss": 0.3796,
+      "step": 13632
+    },
+    {
+      "epoch": 0.5986585727301668,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.7465632551781614e-05,
+      "loss": 0.3366,
+      "step": 13634
+    },
+    {
+      "epoch": 0.5987463912093702,
+      "grad_norm": 1.765625,
+      "learning_rate": 1.745903621673987e-05,
+      "loss": 0.3106,
+      "step": 13636
+    },
+    {
+      "epoch": 0.5988342096885737,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.745244045918694e-05,
+      "loss": 0.3482,
+      "step": 13638
+    },
+    {
+      "epoch": 0.5989220281677772,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.744584527962793e-05,
+      "loss": 0.3297,
+      "step": 13640
+    },
+    {
+      "epoch": 0.5990098466469806,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.7439250678567897e-05,
+      "loss": 0.3305,
+      "step": 13642
+    },
+    {
+      "epoch": 0.5990976651261842,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.7432656656511866e-05,
+      "loss": 0.3571,
+      "step": 13644
+    },
+    {
+      "epoch": 0.5991854836053877,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.7426063213964796e-05,
+      "loss": 0.3437,
+      "step": 13646
+    },
+    {
+      "epoch": 0.5992733020845912,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.741947035143162e-05,
+      "loss": 0.3357,
+      "step": 13648
+    },
+    {
+      "epoch": 0.5993611205637946,
+      "grad_norm": 1.765625,
+      "learning_rate": 1.7412878069417227e-05,
+      "loss": 0.3361,
+      "step": 13650
+    },
+    {
+      "epoch": 0.5994489390429981,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.7406286368426445e-05,
+      "loss": 0.3687,
+      "step": 13652
+    },
+    {
+      "epoch": 0.5995367575222016,
+      "grad_norm": 1.625,
+      "learning_rate": 1.7399695248964086e-05,
+      "loss": 0.3249,
+      "step": 13654
+    },
+    {
+      "epoch": 0.5996245760014051,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.7393104711534874e-05,
+      "loss": 0.3255,
+      "step": 13656
+    },
+    {
+      "epoch": 0.5997123944806085,
+      "grad_norm": 1.796875,
+      "learning_rate": 1.7386514756643536e-05,
+      "loss": 0.3569,
+      "step": 13658
+    },
+    {
+      "epoch": 0.5998002129598121,
+      "grad_norm": 1.5,
+      "learning_rate": 1.7379925384794716e-05,
+      "loss": 0.3281,
+      "step": 13660
+    },
+    {
+      "epoch": 0.5998880314390156,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.7373336596493033e-05,
+      "loss": 0.3503,
+      "step": 13662
+    },
+    {
+      "epoch": 0.5999758499182191,
+      "grad_norm": 1.75,
+      "learning_rate": 1.7366748392243064e-05,
+      "loss": 0.3302,
+      "step": 13664
+    },
+    {
+      "epoch": 0.6000636683974225,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.7360160772549333e-05,
+      "loss": 0.3629,
+      "step": 13666
+    },
+    {
+      "epoch": 0.600151486876626,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.7353573737916322e-05,
+      "loss": 0.369,
+      "step": 13668
+    },
+    {
+      "epoch": 0.6002393053558295,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.7346987288848473e-05,
+      "loss": 0.3343,
+      "step": 13670
+    },
+    {
+      "epoch": 0.600327123835033,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.7340401425850168e-05,
+      "loss": 0.3406,
+      "step": 13672
+    },
+    {
+      "epoch": 0.6004149423142364,
+      "grad_norm": 1.625,
+      "learning_rate": 1.7333816149425753e-05,
+      "loss": 0.3344,
+      "step": 13674
+    },
+    {
+      "epoch": 0.60050276079344,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.732723146007954e-05,
+      "loss": 0.3399,
+      "step": 13676
+    },
+    {
+      "epoch": 0.6005905792726435,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.7320647358315777e-05,
+      "loss": 0.3329,
+      "step": 13678
+    },
+    {
+      "epoch": 0.600678397751847,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.731406384463869e-05,
+      "loss": 0.345,
+      "step": 13680
+    },
+    {
+      "epoch": 0.6007662162310504,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.7307480919552427e-05,
+      "loss": 0.3567,
+      "step": 13682
+    },
+    {
+      "epoch": 0.6008540347102539,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.730089858356113e-05,
+      "loss": 0.3426,
+      "step": 13684
+    },
+    {
+      "epoch": 0.6009418531894574,
+      "grad_norm": 1.828125,
+      "learning_rate": 1.7294316837168857e-05,
+      "loss": 0.2965,
+      "step": 13686
+    },
+    {
+      "epoch": 0.6010296716686608,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.7287735680879645e-05,
+      "loss": 0.3596,
+      "step": 13688
+    },
+    {
+      "epoch": 0.6011174901478644,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.7281155115197484e-05,
+      "loss": 0.3422,
+      "step": 13690
+    },
+    {
+      "epoch": 0.6012053086270679,
+      "grad_norm": 1.828125,
+      "learning_rate": 1.7274575140626318e-05,
+      "loss": 0.3203,
+      "step": 13692
+    },
+    {
+      "epoch": 0.6012931271062714,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.726799575767004e-05,
+      "loss": 0.3258,
+      "step": 13694
+    },
+    {
+      "epoch": 0.6013809455854748,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.72614169668325e-05,
+      "loss": 0.3454,
+      "step": 13696
+    },
+    {
+      "epoch": 0.6014687640646783,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.7254838768617497e-05,
+      "loss": 0.3279,
+      "step": 13698
+    },
+    {
+      "epoch": 0.6015565825438818,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.7248261163528806e-05,
+      "loss": 0.3692,
+      "step": 13700
+    },
+    {
+      "epoch": 0.6016444010230853,
+      "grad_norm": 1.8671875,
+      "learning_rate": 1.724168415207013e-05,
+      "loss": 0.366,
+      "step": 13702
+    },
+    {
+      "epoch": 0.6017322195022887,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.7235107734745136e-05,
+      "loss": 0.3487,
+      "step": 13704
+    },
+    {
+      "epoch": 0.6018200379814923,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.7228531912057465e-05,
+      "loss": 0.3285,
+      "step": 13706
+    },
+    {
+      "epoch": 0.6019078564606958,
+      "grad_norm": 1.765625,
+      "learning_rate": 1.7221956684510677e-05,
+      "loss": 0.3312,
+      "step": 13708
+    },
+    {
+      "epoch": 0.6019956749398993,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.721538205260831e-05,
+      "loss": 0.3263,
+      "step": 13710
+    },
+    {
+      "epoch": 0.6020834934191027,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.720880801685385e-05,
+      "loss": 0.3295,
+      "step": 13712
+    },
+    {
+      "epoch": 0.6021713118983062,
+      "grad_norm": 1.765625,
+      "learning_rate": 1.720223457775075e-05,
+      "loss": 0.3395,
+      "step": 13714
+    },
+    {
+      "epoch": 0.6022591303775097,
+      "grad_norm": 1.8671875,
+      "learning_rate": 1.719566173580239e-05,
+      "loss": 0.3319,
+      "step": 13716
+    },
+    {
+      "epoch": 0.6023469488567131,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.7189089491512116e-05,
+      "loss": 0.3508,
+      "step": 13718
+    },
+    {
+      "epoch": 0.6024347673359166,
+      "grad_norm": 1.796875,
+      "learning_rate": 1.7182517845383252e-05,
+      "loss": 0.3301,
+      "step": 13720
+    },
+    {
+      "epoch": 0.6025225858151202,
+      "grad_norm": 1.7578125,
+      "learning_rate": 1.717594679791904e-05,
+      "loss": 0.3258,
+      "step": 13722
+    },
+    {
+      "epoch": 0.6026104042943237,
+      "grad_norm": 1.8125,
+      "learning_rate": 1.7169376349622698e-05,
+      "loss": 0.3483,
+      "step": 13724
+    },
+    {
+      "epoch": 0.6026982227735271,
+      "grad_norm": 1.8125,
+      "learning_rate": 1.716280650099739e-05,
+      "loss": 0.3264,
+      "step": 13726
+    },
+    {
+      "epoch": 0.6027860412527306,
+      "grad_norm": 1.84375,
+      "learning_rate": 1.7156237252546242e-05,
+      "loss": 0.3246,
+      "step": 13728
+    },
+    {
+      "epoch": 0.6028738597319341,
+      "grad_norm": 1.8203125,
+      "learning_rate": 1.7149668604772324e-05,
+      "loss": 0.3438,
+      "step": 13730
+    },
+    {
+      "epoch": 0.6029616782111376,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.714310055817867e-05,
+      "loss": 0.3036,
+      "step": 13732
+    },
+    {
+      "epoch": 0.603049496690341,
+      "grad_norm": 1.9296875,
+      "learning_rate": 1.7136533113268256e-05,
+      "loss": 0.35,
+      "step": 13734
+    },
+    {
+      "epoch": 0.6031373151695445,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.712996627054402e-05,
+      "loss": 0.3397,
+      "step": 13736
+    },
+    {
+      "epoch": 0.6032251336487481,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.7123400030508852e-05,
+      "loss": 0.3716,
+      "step": 13738
+    },
+    {
+      "epoch": 0.6033129521279516,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.71168343936656e-05,
+      "loss": 0.3035,
+      "step": 13740
+    },
+    {
+      "epoch": 0.603400770607155,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.7110269360517066e-05,
+      "loss": 0.3052,
+      "step": 13742
+    },
+    {
+      "epoch": 0.6034885890863585,
+      "grad_norm": 1.7890625,
+      "learning_rate": 1.710370493156599e-05,
+      "loss": 0.3412,
+      "step": 13744
+    },
+    {
+      "epoch": 0.603576407565562,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.709714110731509e-05,
+      "loss": 0.3488,
+      "step": 13746
+    },
+    {
+      "epoch": 0.6036642260447654,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.709057788826701e-05,
+      "loss": 0.324,
+      "step": 13748
+    },
+    {
+      "epoch": 0.6037520445239689,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.7084015274924373e-05,
+      "loss": 0.3332,
+      "step": 13750
+    },
+    {
+      "epoch": 0.6038398630031725,
+      "grad_norm": 1.7734375,
+      "learning_rate": 1.7077453267789746e-05,
+      "loss": 0.3286,
+      "step": 13752
+    },
+    {
+      "epoch": 0.603927681482376,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.707089186736564e-05,
+      "loss": 0.3647,
+      "step": 13754
+    },
+    {
+      "epoch": 0.6040154999615794,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.7064331074154543e-05,
+      "loss": 0.33,
+      "step": 13756
+    },
+    {
+      "epoch": 0.6041033184407829,
+      "grad_norm": 1.765625,
+      "learning_rate": 1.7057770888658873e-05,
+      "loss": 0.3588,
+      "step": 13758
+    },
+    {
+      "epoch": 0.6041911369199864,
+      "grad_norm": 1.484375,
+      "learning_rate": 1.705121131138101e-05,
+      "loss": 0.3133,
+      "step": 13760
+    },
+    {
+      "epoch": 0.6042789553991899,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.7044652342823292e-05,
+      "loss": 0.3258,
+      "step": 13762
+    },
+    {
+      "epoch": 0.6043667738783933,
+      "grad_norm": 1.8515625,
+      "learning_rate": 1.7038093983488003e-05,
+      "loss": 0.3074,
+      "step": 13764
+    },
+    {
+      "epoch": 0.6044545923575968,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.703153623387738e-05,
+      "loss": 0.3299,
+      "step": 13766
+    },
+    {
+      "epoch": 0.6045424108368004,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.7024979094493637e-05,
+      "loss": 0.3334,
+      "step": 13768
+    },
+    {
+      "epoch": 0.6046302293160039,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.7018422565838896e-05,
+      "loss": 0.3128,
+      "step": 13770
+    },
+    {
+      "epoch": 0.6047180477952073,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.701186664841527e-05,
+      "loss": 0.3236,
+      "step": 13772
+    },
+    {
+      "epoch": 0.6048058662744108,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.7005311342724812e-05,
+      "loss": 0.347,
+      "step": 13774
+    },
+    {
+      "epoch": 0.6048936847536143,
+      "grad_norm": 1.8046875,
+      "learning_rate": 1.6998756649269535e-05,
+      "loss": 0.359,
+      "step": 13776
+    },
+    {
+      "epoch": 0.6049815032328177,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.6992202568551383e-05,
+      "loss": 0.322,
+      "step": 13778
+    },
+    {
+      "epoch": 0.6050693217120212,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.6985649101072277e-05,
+      "loss": 0.3372,
+      "step": 13780
+    },
+    {
+      "epoch": 0.6051571401912247,
+      "grad_norm": 1.625,
+      "learning_rate": 1.6979096247334092e-05,
+      "loss": 0.3177,
+      "step": 13782
+    },
+    {
+      "epoch": 0.6052449586704283,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.697254400783863e-05,
+      "loss": 0.3389,
+      "step": 13784
+    },
+    {
+      "epoch": 0.6053327771496317,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.6965992383087677e-05,
+      "loss": 0.3674,
+      "step": 13786
+    },
+    {
+      "epoch": 0.6054205956288352,
+      "grad_norm": 1.84375,
+      "learning_rate": 1.6959441373582947e-05,
+      "loss": 0.3298,
+      "step": 13788
+    },
+    {
+      "epoch": 0.6055084141080387,
+      "grad_norm": 1.515625,
+      "learning_rate": 1.6952890979826132e-05,
+      "loss": 0.3318,
+      "step": 13790
+    },
+    {
+      "epoch": 0.6055962325872422,
+      "grad_norm": 1.90625,
+      "learning_rate": 1.6946341202318854e-05,
+      "loss": 0.3524,
+      "step": 13792
+    },
+    {
+      "epoch": 0.6056840510664456,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.6939792041562695e-05,
+      "loss": 0.3478,
+      "step": 13794
+    },
+    {
+      "epoch": 0.6057718695456491,
+      "grad_norm": 1.75,
+      "learning_rate": 1.693324349805919e-05,
+      "loss": 0.3476,
+      "step": 13796
+    },
+    {
+      "epoch": 0.6058596880248527,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.6926695572309837e-05,
+      "loss": 0.3191,
+      "step": 13798
+    },
+    {
+      "epoch": 0.6059475065040562,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.6920148264816064e-05,
+      "loss": 0.3439,
+      "step": 13800
+    },
+    {
+      "epoch": 0.6060353249832596,
+      "grad_norm": 1.8125,
+      "learning_rate": 1.691360157607928e-05,
+      "loss": 0.3311,
+      "step": 13802
+    },
+    {
+      "epoch": 0.6061231434624631,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.6907055506600834e-05,
+      "loss": 0.347,
+      "step": 13804
+    },
+    {
+      "epoch": 0.6062109619416666,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.6900510056882012e-05,
+      "loss": 0.3227,
+      "step": 13806
+    },
+    {
+      "epoch": 0.60629878042087,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.6893965227424073e-05,
+      "loss": 0.3598,
+      "step": 13808
+    },
+    {
+      "epoch": 0.6063865989000735,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.6887421018728215e-05,
+      "loss": 0.3397,
+      "step": 13810
+    },
+    {
+      "epoch": 0.606474417379277,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.68808774312956e-05,
+      "loss": 0.3515,
+      "step": 13812
+    },
+    {
+      "epoch": 0.6065622358584806,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.6874334465627335e-05,
+      "loss": 0.3346,
+      "step": 13814
+    },
+    {
+      "epoch": 0.606650054337684,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.686779212222449e-05,
+      "loss": 0.3284,
+      "step": 13816
+    },
+    {
+      "epoch": 0.6067378728168875,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.6861250401588075e-05,
+      "loss": 0.3169,
+      "step": 13818
+    },
+    {
+      "epoch": 0.606825691296091,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.685470930421905e-05,
+      "loss": 0.3321,
+      "step": 13820
+    },
+    {
+      "epoch": 0.6069135097752945,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.6848168830618338e-05,
+      "loss": 0.3241,
+      "step": 13822
+    },
+    {
+      "epoch": 0.6070013282544979,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.6841628981286814e-05,
+      "loss": 0.3567,
+      "step": 13824
+    },
+    {
+      "epoch": 0.6070891467337014,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.6835089756725297e-05,
+      "loss": 0.3199,
+      "step": 13826
+    },
+    {
+      "epoch": 0.6071769652129049,
+      "grad_norm": 1.8515625,
+      "learning_rate": 1.6828551157434568e-05,
+      "loss": 0.3317,
+      "step": 13828
+    },
+    {
+      "epoch": 0.6072647836921085,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.6822013183915347e-05,
+      "loss": 0.3371,
+      "step": 13830
+    },
+    {
+      "epoch": 0.6073526021713119,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.6815475836668317e-05,
+      "loss": 0.3358,
+      "step": 13832
+    },
+    {
+      "epoch": 0.6074404206505154,
+      "grad_norm": 1.7734375,
+      "learning_rate": 1.6808939116194107e-05,
+      "loss": 0.3335,
+      "step": 13834
+    },
+    {
+      "epoch": 0.6075282391297189,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.6802403022993304e-05,
+      "loss": 0.3321,
+      "step": 13836
+    },
+    {
+      "epoch": 0.6076160576089223,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.6795867557566454e-05,
+      "loss": 0.3463,
+      "step": 13838
+    },
+    {
+      "epoch": 0.6077038760881258,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.6789332720414023e-05,
+      "loss": 0.3516,
+      "step": 13840
+    },
+    {
+      "epoch": 0.6077916945673293,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.6782798512036457e-05,
+      "loss": 0.3261,
+      "step": 13842
+    },
+    {
+      "epoch": 0.6078795130465329,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.677626493293416e-05,
+      "loss": 0.3435,
+      "step": 13844
+    },
+    {
+      "epoch": 0.6079673315257363,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.676973198360746e-05,
+      "loss": 0.3584,
+      "step": 13846
+    },
+    {
+      "epoch": 0.6080551500049398,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.6763199664556656e-05,
+      "loss": 0.3631,
+      "step": 13848
+    },
+    {
+      "epoch": 0.6081429684841433,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.6756667976281997e-05,
+      "loss": 0.3259,
+      "step": 13850
+    },
+    {
+      "epoch": 0.6082307869633468,
+      "grad_norm": 1.953125,
+      "learning_rate": 1.6750136919283678e-05,
+      "loss": 0.3486,
+      "step": 13852
+    },
+    {
+      "epoch": 0.6083186054425502,
+      "grad_norm": 1.625,
+      "learning_rate": 1.674360649406186e-05,
+      "loss": 0.3215,
+      "step": 13854
+    },
+    {
+      "epoch": 0.6084064239217537,
+      "grad_norm": 1.9609375,
+      "learning_rate": 1.673707670111663e-05,
+      "loss": 0.3284,
+      "step": 13856
+    },
+    {
+      "epoch": 0.6084942424009572,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.6730547540948048e-05,
+      "loss": 0.336,
+      "step": 13858
+    },
+    {
+      "epoch": 0.6085820608801608,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.6724019014056115e-05,
+      "loss": 0.3323,
+      "step": 13860
+    },
+    {
+      "epoch": 0.6086698793593642,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.6717491120940793e-05,
+      "loss": 0.358,
+      "step": 13862
+    },
+    {
+      "epoch": 0.6087576978385677,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.671096386210198e-05,
+      "loss": 0.331,
+      "step": 13864
+    },
+    {
+      "epoch": 0.6088455163177712,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.670443723803955e-05,
+      "loss": 0.3196,
+      "step": 13866
+    },
+    {
+      "epoch": 0.6089333347969746,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.669791124925331e-05,
+      "loss": 0.3557,
+      "step": 13868
+    },
+    {
+      "epoch": 0.6090211532761781,
+      "grad_norm": 1.84375,
+      "learning_rate": 1.669138589624301e-05,
+      "loss": 0.3677,
+      "step": 13870
+    },
+    {
+      "epoch": 0.6091089717553816,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.668486117950837e-05,
+      "loss": 0.2952,
+      "step": 13872
+    },
+    {
+      "epoch": 0.6091967902345851,
+      "grad_norm": 1.78125,
+      "learning_rate": 1.6678337099549052e-05,
+      "loss": 0.3623,
+      "step": 13874
+    },
+    {
+      "epoch": 0.6092846087137886,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.667181365686467e-05,
+      "loss": 0.3241,
+      "step": 13876
+    },
+    {
+      "epoch": 0.6093724271929921,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.66652908519548e-05,
+      "loss": 0.3287,
+      "step": 13878
+    },
+    {
+      "epoch": 0.6094602456721956,
+      "grad_norm": 1.765625,
+      "learning_rate": 1.6658768685318955e-05,
+      "loss": 0.337,
+      "step": 13880
+    },
+    {
+      "epoch": 0.6095480641513991,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.6652247157456603e-05,
+      "loss": 0.3112,
+      "step": 13882
+    },
+    {
+      "epoch": 0.6096358826306025,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.6645726268867163e-05,
+      "loss": 0.3299,
+      "step": 13884
+    },
+    {
+      "epoch": 0.609723701109806,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.6639206020050006e-05,
+      "loss": 0.322,
+      "step": 13886
+    },
+    {
+      "epoch": 0.6098115195890095,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.6632686411504455e-05,
+      "loss": 0.3119,
+      "step": 13888
+    },
+    {
+      "epoch": 0.609899338068213,
+      "grad_norm": 1.8984375,
+      "learning_rate": 1.6626167443729797e-05,
+      "loss": 0.3334,
+      "step": 13890
+    },
+    {
+      "epoch": 0.6099871565474165,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.6619649117225233e-05,
+      "loss": 0.354,
+      "step": 13892
+    },
+    {
+      "epoch": 0.61007497502662,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.6613131432489947e-05,
+      "loss": 0.3137,
+      "step": 13894
+    },
+    {
+      "epoch": 0.6101627935058235,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.6606614390023066e-05,
+      "loss": 0.3491,
+      "step": 13896
+    },
+    {
+      "epoch": 0.610250611985027,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.660009799032368e-05,
+      "loss": 0.3024,
+      "step": 13898
+    },
+    {
+      "epoch": 0.6103384304642304,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.65935822338908e-05,
+      "loss": 0.3068,
+      "step": 13900
+    },
+    {
+      "epoch": 0.6104262489434339,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.6587067121223397e-05,
+      "loss": 0.3341,
+      "step": 13902
+    },
+    {
+      "epoch": 0.6105140674226374,
+      "grad_norm": 1.8203125,
+      "learning_rate": 1.6580552652820412e-05,
+      "loss": 0.3257,
+      "step": 13904
+    },
+    {
+      "epoch": 0.610601885901841,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.6574038829180733e-05,
+      "loss": 0.3143,
+      "step": 13906
+    },
+    {
+      "epoch": 0.6106897043810444,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.6567525650803174e-05,
+      "loss": 0.3188,
+      "step": 13908
+    },
+    {
+      "epoch": 0.6107775228602479,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.656101311818652e-05,
+      "loss": 0.3143,
+      "step": 13910
+    },
+    {
+      "epoch": 0.6108653413394514,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.6554501231829504e-05,
+      "loss": 0.3411,
+      "step": 13912
+    },
+    {
+      "epoch": 0.6109531598186548,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.654798999223081e-05,
+      "loss": 0.3352,
+      "step": 13914
+    },
+    {
+      "epoch": 0.6110409782978583,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.654147939988907e-05,
+      "loss": 0.3489,
+      "step": 13916
+    },
+    {
+      "epoch": 0.6111287967770618,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.6534969455302864e-05,
+      "loss": 0.351,
+      "step": 13918
+    },
+    {
+      "epoch": 0.6112166152562652,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.6528460158970727e-05,
+      "loss": 0.3228,
+      "step": 13920
+    },
+    {
+      "epoch": 0.6113044337354688,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.652195151139114e-05,
+      "loss": 0.3586,
+      "step": 13922
+    },
+    {
+      "epoch": 0.6113922522146723,
+      "grad_norm": 1.8125,
+      "learning_rate": 1.651544351306254e-05,
+      "loss": 0.326,
+      "step": 13924
+    },
+    {
+      "epoch": 0.6114800706938758,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.6508936164483314e-05,
+      "loss": 0.3237,
+      "step": 13926
+    },
+    {
+      "epoch": 0.6115678891730792,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.6502429466151788e-05,
+      "loss": 0.3357,
+      "step": 13928
+    },
+    {
+      "epoch": 0.6116557076522827,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.649592341856625e-05,
+      "loss": 0.33,
+      "step": 13930
+    },
+    {
+      "epoch": 0.6117435261314862,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.648941802222494e-05,
+      "loss": 0.3332,
+      "step": 13932
+    },
+    {
+      "epoch": 0.6118313446106897,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.6482913277626033e-05,
+      "loss": 0.3155,
+      "step": 13934
+    },
+    {
+      "epoch": 0.6119191630898931,
+      "grad_norm": 1.7890625,
+      "learning_rate": 1.6476409185267666e-05,
+      "loss": 0.3584,
+      "step": 13936
+    },
+    {
+      "epoch": 0.6120069815690967,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.6469905745647928e-05,
+      "loss": 0.343,
+      "step": 13938
+    },
+    {
+      "epoch": 0.6120948000483002,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.6463402959264858e-05,
+      "loss": 0.3573,
+      "step": 13940
+    },
+    {
+      "epoch": 0.6121826185275037,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.6456900826616433e-05,
+      "loss": 0.3155,
+      "step": 13942
+    },
+    {
+      "epoch": 0.6122704370067071,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.645039934820059e-05,
+      "loss": 0.3426,
+      "step": 13944
+    },
+    {
+      "epoch": 0.6123582554859106,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.644389852451521e-05,
+      "loss": 0.3317,
+      "step": 13946
+    },
+    {
+      "epoch": 0.6124460739651141,
+      "grad_norm": 1.5,
+      "learning_rate": 1.6437398356058137e-05,
+      "loss": 0.3334,
+      "step": 13948
+    },
+    {
+      "epoch": 0.6125338924443176,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.643089884332715e-05,
+      "loss": 0.3079,
+      "step": 13950
+    },
+    {
+      "epoch": 0.6126217109235211,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.642439998681999e-05,
+      "loss": 0.3297,
+      "step": 13952
+    },
+    {
+      "epoch": 0.6127095294027246,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.6417901787034324e-05,
+      "loss": 0.309,
+      "step": 13954
+    },
+    {
+      "epoch": 0.6127973478819281,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.64114042444678e-05,
+      "loss": 0.3131,
+      "step": 13956
+    },
+    {
+      "epoch": 0.6128851663611316,
+      "grad_norm": 1.8203125,
+      "learning_rate": 1.6404907359618e-05,
+      "loss": 0.3473,
+      "step": 13958
+    },
+    {
+      "epoch": 0.612972984840335,
+      "grad_norm": 1.84375,
+      "learning_rate": 1.639841113298246e-05,
+      "loss": 0.3242,
+      "step": 13960
+    },
+    {
+      "epoch": 0.6130608033195385,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.6391915565058653e-05,
+      "loss": 0.3567,
+      "step": 13962
+    },
+    {
+      "epoch": 0.613148621798742,
+      "grad_norm": 1.625,
+      "learning_rate": 1.6385420656344007e-05,
+      "loss": 0.3253,
+      "step": 13964
+    },
+    {
+      "epoch": 0.6132364402779454,
+      "grad_norm": 1.7890625,
+      "learning_rate": 1.637892640733592e-05,
+      "loss": 0.3385,
+      "step": 13966
+    },
+    {
+      "epoch": 0.613324258757149,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.637243281853172e-05,
+      "loss": 0.3117,
+      "step": 13968
+    },
+    {
+      "epoch": 0.6134120772363525,
+      "grad_norm": 1.625,
+      "learning_rate": 1.6365939890428673e-05,
+      "loss": 0.3177,
+      "step": 13970
+    },
+    {
+      "epoch": 0.613499895715556,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.6359447623524022e-05,
+      "loss": 0.3271,
+      "step": 13972
+    },
+    {
+      "epoch": 0.6135877141947594,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.635295601831494e-05,
+      "loss": 0.3415,
+      "step": 13974
+    },
+    {
+      "epoch": 0.6136755326739629,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.6346465075298564e-05,
+      "loss": 0.3051,
+      "step": 13976
+    },
+    {
+      "epoch": 0.6137633511531664,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.633997479497197e-05,
+      "loss": 0.3285,
+      "step": 13978
+    },
+    {
+      "epoch": 0.6138511696323699,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.6333485177832176e-05,
+      "loss": 0.3322,
+      "step": 13980
+    },
+    {
+      "epoch": 0.6139389881115733,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.632699622437617e-05,
+      "loss": 0.3341,
+      "step": 13982
+    },
+    {
+      "epoch": 0.6140268065907769,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.6320507935100863e-05,
+      "loss": 0.3281,
+      "step": 13984
+    },
+    {
+      "epoch": 0.6141146250699804,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.6314020310503144e-05,
+      "loss": 0.3057,
+      "step": 13986
+    },
+    {
+      "epoch": 0.6142024435491839,
+      "grad_norm": 1.8359375,
+      "learning_rate": 1.630753335107984e-05,
+      "loss": 0.3259,
+      "step": 13988
+    },
+    {
+      "epoch": 0.6142902620283873,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.630104705732771e-05,
+      "loss": 0.3332,
+      "step": 13990
+    },
+    {
+      "epoch": 0.6143780805075908,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.6294561429743475e-05,
+      "loss": 0.3268,
+      "step": 13992
+    },
+    {
+      "epoch": 0.6144658989867943,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.6288076468823827e-05,
+      "loss": 0.311,
+      "step": 13994
+    },
+    {
+      "epoch": 0.6145537174659977,
+      "grad_norm": 1.875,
+      "learning_rate": 1.6281592175065357e-05,
+      "loss": 0.3465,
+      "step": 13996
+    },
+    {
+      "epoch": 0.6146415359452013,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.6275108548964653e-05,
+      "loss": 0.3331,
+      "step": 13998
+    },
+    {
+      "epoch": 0.6147293544244048,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.626862559101823e-05,
+      "loss": 0.3438,
+      "step": 14000
+    },
+    {
+      "epoch": 0.6148171729036083,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.6262143301722547e-05,
+      "loss": 0.3152,
+      "step": 14002
+    },
+    {
+      "epoch": 0.6149049913828117,
+      "grad_norm": 1.8125,
+      "learning_rate": 1.625566168157403e-05,
+      "loss": 0.3089,
+      "step": 14004
+    },
+    {
+      "epoch": 0.6149928098620152,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.6249180731069036e-05,
+      "loss": 0.3415,
+      "step": 14006
+    },
+    {
+      "epoch": 0.6150806283412187,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.6242700450703876e-05,
+      "loss": 0.3317,
+      "step": 14008
+    },
+    {
+      "epoch": 0.6151684468204222,
+      "grad_norm": 1.84375,
+      "learning_rate": 1.6236220840974815e-05,
+      "loss": 0.3361,
+      "step": 14010
+    },
+    {
+      "epoch": 0.6152562652996256,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.6229741902378063e-05,
+      "loss": 0.3206,
+      "step": 14012
+    },
+    {
+      "epoch": 0.6153440837788292,
+      "grad_norm": 1.9765625,
+      "learning_rate": 1.6223263635409785e-05,
+      "loss": 0.3096,
+      "step": 14014
+    },
+    {
+      "epoch": 0.6154319022580327,
+      "grad_norm": 1.84375,
+      "learning_rate": 1.621678604056608e-05,
+      "loss": 0.3521,
+      "step": 14016
+    },
+    {
+      "epoch": 0.6155197207372362,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.6210309118343e-05,
+      "loss": 0.3101,
+      "step": 14018
+    },
+    {
+      "epoch": 0.6156075392164396,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.6203832869236557e-05,
+      "loss": 0.3519,
+      "step": 14020
+    },
+    {
+      "epoch": 0.6156953576956431,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.619735729374271e-05,
+      "loss": 0.3418,
+      "step": 14022
+    },
+    {
+      "epoch": 0.6157831761748466,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.6190882392357342e-05,
+      "loss": 0.3315,
+      "step": 14024
+    },
+    {
+      "epoch": 0.61587099465405,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.6184408165576316e-05,
+      "loss": 0.3609,
+      "step": 14026
+    },
+    {
+      "epoch": 0.6159588131332535,
+      "grad_norm": 1.78125,
+      "learning_rate": 1.6177934613895422e-05,
+      "loss": 0.3276,
+      "step": 14028
+    },
+    {
+      "epoch": 0.6160466316124571,
+      "grad_norm": 1.7578125,
+      "learning_rate": 1.6171461737810413e-05,
+      "loss": 0.3124,
+      "step": 14030
+    },
+    {
+      "epoch": 0.6161344500916606,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.616498953781698e-05,
+      "loss": 0.3149,
+      "step": 14032
+    },
+    {
+      "epoch": 0.616222268570864,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.6158518014410762e-05,
+      "loss": 0.3245,
+      "step": 14034
+    },
+    {
+      "epoch": 0.6163100870500675,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.615204716808736e-05,
+      "loss": 0.341,
+      "step": 14036
+    },
+    {
+      "epoch": 0.616397905529271,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.61455769993423e-05,
+      "loss": 0.3343,
+      "step": 14038
+    },
+    {
+      "epoch": 0.6164857240084745,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.6139107508671086e-05,
+      "loss": 0.3379,
+      "step": 14040
+    },
+    {
+      "epoch": 0.6165735424876779,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.6132638696569134e-05,
+      "loss": 0.3453,
+      "step": 14042
+    },
+    {
+      "epoch": 0.6166613609668815,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.612617056353184e-05,
+      "loss": 0.3367,
+      "step": 14044
+    },
+    {
+      "epoch": 0.616749179446085,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.611970311005453e-05,
+      "loss": 0.3481,
+      "step": 14046
+    },
+    {
+      "epoch": 0.6168369979252885,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.611323633663248e-05,
+      "loss": 0.3198,
+      "step": 14048
+    },
+    {
+      "epoch": 0.6169248164044919,
+      "grad_norm": 2.015625,
+      "learning_rate": 1.610677024376093e-05,
+      "loss": 0.3636,
+      "step": 14050
+    },
+    {
+      "epoch": 0.6170126348836954,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.6100304831935052e-05,
+      "loss": 0.316,
+      "step": 14052
+    },
+    {
+      "epoch": 0.6171004533628989,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.609384010164996e-05,
+      "loss": 0.3195,
+      "step": 14054
+    },
+    {
+      "epoch": 0.6171882718421023,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.608737605340072e-05,
+      "loss": 0.3629,
+      "step": 14056
+    },
+    {
+      "epoch": 0.6172760903213058,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.608091268768236e-05,
+      "loss": 0.2921,
+      "step": 14058
+    },
+    {
+      "epoch": 0.6173639088005094,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.6074450004989844e-05,
+      "loss": 0.3225,
+      "step": 14060
+    },
+    {
+      "epoch": 0.6174517272797129,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.606798800581809e-05,
+      "loss": 0.3192,
+      "step": 14062
+    },
+    {
+      "epoch": 0.6175395457589163,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.6061526690661947e-05,
+      "loss": 0.3356,
+      "step": 14064
+    },
+    {
+      "epoch": 0.6176273642381198,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.6055066060016247e-05,
+      "loss": 0.3436,
+      "step": 14066
+    },
+    {
+      "epoch": 0.6177151827173233,
+      "grad_norm": 1.875,
+      "learning_rate": 1.6048606114375723e-05,
+      "loss": 0.3204,
+      "step": 14068
+    },
+    {
+      "epoch": 0.6178030011965268,
+      "grad_norm": 1.5,
+      "learning_rate": 1.604214685423509e-05,
+      "loss": 0.3223,
+      "step": 14070
+    },
+    {
+      "epoch": 0.6178908196757302,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.6035688280088995e-05,
+      "loss": 0.3394,
+      "step": 14072
+    },
+    {
+      "epoch": 0.6179786381549337,
+      "grad_norm": 1.515625,
+      "learning_rate": 1.6029230392432043e-05,
+      "loss": 0.3128,
+      "step": 14074
+    },
+    {
+      "epoch": 0.6180664566341373,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.6022773191758784e-05,
+      "loss": 0.3205,
+      "step": 14076
+    },
+    {
+      "epoch": 0.6181542751133408,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.60163166785637e-05,
+      "loss": 0.343,
+      "step": 14078
+    },
+    {
+      "epoch": 0.6182420935925442,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.6009860853341237e-05,
+      "loss": 0.3274,
+      "step": 14080
+    },
+    {
+      "epoch": 0.6183299120717477,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.6003405716585798e-05,
+      "loss": 0.3696,
+      "step": 14082
+    },
+    {
+      "epoch": 0.6184177305509512,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.5996951268791695e-05,
+      "loss": 0.3001,
+      "step": 14084
+    },
+    {
+      "epoch": 0.6185055490301546,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.599049751045322e-05,
+      "loss": 0.3314,
+      "step": 14086
+    },
+    {
+      "epoch": 0.6185933675093581,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.5984044442064606e-05,
+      "loss": 0.3451,
+      "step": 14088
+    },
+    {
+      "epoch": 0.6186811859885616,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.5977592064120027e-05,
+      "loss": 0.3189,
+      "step": 14090
+    },
+    {
+      "epoch": 0.6187690044677652,
+      "grad_norm": 1.625,
+      "learning_rate": 1.5971140377113623e-05,
+      "loss": 0.3403,
+      "step": 14092
+    },
+    {
+      "epoch": 0.6188568229469686,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.5964689381539445e-05,
+      "loss": 0.3287,
+      "step": 14094
+    },
+    {
+      "epoch": 0.6189446414261721,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.595823907789152e-05,
+      "loss": 0.3246,
+      "step": 14096
+    },
+    {
+      "epoch": 0.6190324599053756,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.595178946666381e-05,
+      "loss": 0.3468,
+      "step": 14098
+    },
+    {
+      "epoch": 0.6191202783845791,
+      "grad_norm": 1.75,
+      "learning_rate": 1.5945340548350235e-05,
+      "loss": 0.329,
+      "step": 14100
+    },
+    {
+      "epoch": 0.6192080968637825,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.5938892323444653e-05,
+      "loss": 0.3288,
+      "step": 14102
+    },
+    {
+      "epoch": 0.619295915342986,
+      "grad_norm": 1.75,
+      "learning_rate": 1.593244479244087e-05,
+      "loss": 0.3357,
+      "step": 14104
+    },
+    {
+      "epoch": 0.6193837338221896,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.5925997955832633e-05,
+      "loss": 0.3161,
+      "step": 14106
+    },
+    {
+      "epoch": 0.6194715523013931,
+      "grad_norm": 1.7578125,
+      "learning_rate": 1.591955181411365e-05,
+      "loss": 0.3194,
+      "step": 14108
+    },
+    {
+      "epoch": 0.6195593707805965,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.5913106367777567e-05,
+      "loss": 0.3085,
+      "step": 14110
+    },
+    {
+      "epoch": 0.6196471892598,
+      "grad_norm": 1.8828125,
+      "learning_rate": 1.590666161731798e-05,
+      "loss": 0.3346,
+      "step": 14112
+    },
+    {
+      "epoch": 0.6197350077390035,
+      "grad_norm": 1.7578125,
+      "learning_rate": 1.5900217563228426e-05,
+      "loss": 0.3722,
+      "step": 14114
+    },
+    {
+      "epoch": 0.619822826218207,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.5893774206002393e-05,
+      "loss": 0.3329,
+      "step": 14116
+    },
+    {
+      "epoch": 0.6199106446974104,
+      "grad_norm": 1.8203125,
+      "learning_rate": 1.588733154613331e-05,
+      "loss": 0.367,
+      "step": 14118
+    },
+    {
+      "epoch": 0.6199984631766139,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.588088958411456e-05,
+      "loss": 0.3305,
+      "step": 14120
+    },
+    {
+      "epoch": 0.6200862816558175,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.5874448320439475e-05,
+      "loss": 0.3372,
+      "step": 14122
+    },
+    {
+      "epoch": 0.620174100135021,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.586800775560132e-05,
+      "loss": 0.3231,
+      "step": 14124
+    },
+    {
+      "epoch": 0.6202619186142244,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.5861567890093328e-05,
+      "loss": 0.3355,
+      "step": 14126
+    },
+    {
+      "epoch": 0.6203497370934279,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.5855128724408655e-05,
+      "loss": 0.3658,
+      "step": 14128
+    },
+    {
+      "epoch": 0.6204375555726314,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.5848690259040414e-05,
+      "loss": 0.3452,
+      "step": 14130
+    },
+    {
+      "epoch": 0.6205253740518348,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.5842252494481664e-05,
+      "loss": 0.3205,
+      "step": 14132
+    },
+    {
+      "epoch": 0.6206131925310383,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.5835815431225418e-05,
+      "loss": 0.3294,
+      "step": 14134
+    },
+    {
+      "epoch": 0.6207010110102418,
+      "grad_norm": 1.921875,
+      "learning_rate": 1.5829379069764622e-05,
+      "loss": 0.3441,
+      "step": 14136
+    },
+    {
+      "epoch": 0.6207888294894454,
+      "grad_norm": 1.765625,
+      "learning_rate": 1.582294341059218e-05,
+      "loss": 0.3576,
+      "step": 14138
+    },
+    {
+      "epoch": 0.6208766479686488,
+      "grad_norm": 1.796875,
+      "learning_rate": 1.5816508454200922e-05,
+      "loss": 0.3253,
+      "step": 14140
+    },
+    {
+      "epoch": 0.6209644664478523,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.581007420108365e-05,
+      "loss": 0.3422,
+      "step": 14142
+    },
+    {
+      "epoch": 0.6210522849270558,
+      "grad_norm": 1.9140625,
+      "learning_rate": 1.5803640651733115e-05,
+      "loss": 0.329,
+      "step": 14144
+    },
+    {
+      "epoch": 0.6211401034062592,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.579720780664197e-05,
+      "loss": 0.3193,
+      "step": 14146
+    },
+    {
+      "epoch": 0.6212279218854627,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.5790775666302855e-05,
+      "loss": 0.3378,
+      "step": 14148
+    },
+    {
+      "epoch": 0.6213157403646662,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.5784344231208347e-05,
+      "loss": 0.3437,
+      "step": 14150
+    },
+    {
+      "epoch": 0.6214035588438698,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.577791350185097e-05,
+      "loss": 0.3464,
+      "step": 14152
+    },
+    {
+      "epoch": 0.6214913773230732,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.5771483478723188e-05,
+      "loss": 0.3521,
+      "step": 14154
+    },
+    {
+      "epoch": 0.6215791958022767,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.576505416231741e-05,
+      "loss": 0.344,
+      "step": 14156
+    },
+    {
+      "epoch": 0.6216670142814802,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.5758625553126e-05,
+      "loss": 0.3501,
+      "step": 14158
+    },
+    {
+      "epoch": 0.6217548327606837,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.575219765164126e-05,
+      "loss": 0.3525,
+      "step": 14160
+    },
+    {
+      "epoch": 0.6218426512398871,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.5745770458355442e-05,
+      "loss": 0.3339,
+      "step": 14162
+    },
+    {
+      "epoch": 0.6219304697190906,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.5739343973760743e-05,
+      "loss": 0.3257,
+      "step": 14164
+    },
+    {
+      "epoch": 0.6220182881982941,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.57329181983493e-05,
+      "loss": 0.3317,
+      "step": 14166
+    },
+    {
+      "epoch": 0.6221061066774977,
+      "grad_norm": 1.796875,
+      "learning_rate": 1.5726493132613203e-05,
+      "loss": 0.3308,
+      "step": 14168
+    },
+    {
+      "epoch": 0.6221939251567011,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.5720068777044476e-05,
+      "loss": 0.3298,
+      "step": 14170
+    },
+    {
+      "epoch": 0.6222817436359046,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.5713645132135118e-05,
+      "loss": 0.3412,
+      "step": 14172
+    },
+    {
+      "epoch": 0.6223695621151081,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.570722219837705e-05,
+      "loss": 0.3044,
+      "step": 14174
+    },
+    {
+      "epoch": 0.6224573805943115,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.570079997626212e-05,
+      "loss": 0.3277,
+      "step": 14176
+    },
+    {
+      "epoch": 0.622545199073515,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.569437846628216e-05,
+      "loss": 0.3342,
+      "step": 14178
+    },
+    {
+      "epoch": 0.6226330175527185,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.5687957668928927e-05,
+      "loss": 0.3118,
+      "step": 14180
+    },
+    {
+      "epoch": 0.622720836031922,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.5681537584694128e-05,
+      "loss": 0.3523,
+      "step": 14182
+    },
+    {
+      "epoch": 0.6228086545111255,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.567511821406941e-05,
+      "loss": 0.3283,
+      "step": 14184
+    },
+    {
+      "epoch": 0.622896472990329,
+      "grad_norm": 1.484375,
+      "learning_rate": 1.566869955754638e-05,
+      "loss": 0.324,
+      "step": 14186
+    },
+    {
+      "epoch": 0.6229842914695325,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.5662281615616582e-05,
+      "loss": 0.3403,
+      "step": 14188
+    },
+    {
+      "epoch": 0.623072109948736,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.5655864388771486e-05,
+      "loss": 0.3209,
+      "step": 14190
+    },
+    {
+      "epoch": 0.6231599284279394,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.5649447877502537e-05,
+      "loss": 0.3108,
+      "step": 14192
+    },
+    {
+      "epoch": 0.6232477469071429,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.5643032082301106e-05,
+      "loss": 0.3284,
+      "step": 14194
+    },
+    {
+      "epoch": 0.6233355653863464,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.5636617003658527e-05,
+      "loss": 0.3531,
+      "step": 14196
+    },
+    {
+      "epoch": 0.62342338386555,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.5630202642066062e-05,
+      "loss": 0.3265,
+      "step": 14198
+    },
+    {
+      "epoch": 0.6235112023447534,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.5623788998014925e-05,
+      "loss": 0.3469,
+      "step": 14200
+    },
+    {
+      "epoch": 0.6235990208239569,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.5617376071996277e-05,
+      "loss": 0.3165,
+      "step": 14202
+    },
+    {
+      "epoch": 0.6236868393031604,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.5610963864501212e-05,
+      "loss": 0.3144,
+      "step": 14204
+    },
+    {
+      "epoch": 0.6237746577823639,
+      "grad_norm": 1.625,
+      "learning_rate": 1.5604552376020797e-05,
+      "loss": 0.3202,
+      "step": 14206
+    },
+    {
+      "epoch": 0.6238624762615673,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.5598141607046004e-05,
+      "loss": 0.3172,
+      "step": 14208
+    },
+    {
+      "epoch": 0.6239502947407708,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.559173155806778e-05,
+      "loss": 0.3439,
+      "step": 14210
+    },
+    {
+      "epoch": 0.6240381132199743,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.558532222957701e-05,
+      "loss": 0.292,
+      "step": 14212
+    },
+    {
+      "epoch": 0.6241259316991778,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.5578913622064523e-05,
+      "loss": 0.3282,
+      "step": 14214
+    },
+    {
+      "epoch": 0.6242137501783813,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.5572505736021088e-05,
+      "loss": 0.3597,
+      "step": 14216
+    },
+    {
+      "epoch": 0.6243015686575848,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.5566098571937416e-05,
+      "loss": 0.3474,
+      "step": 14218
+    },
+    {
+      "epoch": 0.6243893871367883,
+      "grad_norm": 2.1875,
+      "learning_rate": 1.5559692130304185e-05,
+      "loss": 0.3342,
+      "step": 14220
+    },
+    {
+      "epoch": 0.6244772056159917,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.555328641161199e-05,
+      "loss": 0.3256,
+      "step": 14222
+    },
+    {
+      "epoch": 0.6245650240951952,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.5546881416351385e-05,
+      "loss": 0.3209,
+      "step": 14224
+    },
+    {
+      "epoch": 0.6246528425743987,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.5540477145012876e-05,
+      "loss": 0.3217,
+      "step": 14226
+    },
+    {
+      "epoch": 0.6247406610536022,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.5534073598086888e-05,
+      "loss": 0.3278,
+      "step": 14228
+    },
+    {
+      "epoch": 0.6248284795328057,
+      "grad_norm": 1.8046875,
+      "learning_rate": 1.5527670776063812e-05,
+      "loss": 0.3114,
+      "step": 14230
+    },
+    {
+      "epoch": 0.6249162980120092,
+      "grad_norm": 1.765625,
+      "learning_rate": 1.552126867943398e-05,
+      "loss": 0.3408,
+      "step": 14232
+    },
+    {
+      "epoch": 0.6250041164912127,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.5514867308687665e-05,
+      "loss": 0.3077,
+      "step": 14234
+    },
+    {
+      "epoch": 0.6250919349704162,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.5508466664315092e-05,
+      "loss": 0.3432,
+      "step": 14236
+    },
+    {
+      "epoch": 0.6251797534496196,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.550206674680641e-05,
+      "loss": 0.3146,
+      "step": 14238
+    },
+    {
+      "epoch": 0.6252675719288231,
+      "grad_norm": 1.8203125,
+      "learning_rate": 1.5495667556651738e-05,
+      "loss": 0.3549,
+      "step": 14240
+    },
+    {
+      "epoch": 0.6253553904080266,
+      "grad_norm": 1.4765625,
+      "learning_rate": 1.548926909434112e-05,
+      "loss": 0.3519,
+      "step": 14242
+    },
+    {
+      "epoch": 0.62544320888723,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.5482871360364548e-05,
+      "loss": 0.3228,
+      "step": 14244
+    },
+    {
+      "epoch": 0.6255310273664336,
+      "grad_norm": 1.8046875,
+      "learning_rate": 1.5476474355211973e-05,
+      "loss": 0.3409,
+      "step": 14246
+    },
+    {
+      "epoch": 0.6256188458456371,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.5470078079373275e-05,
+      "loss": 0.3254,
+      "step": 14248
+    },
+    {
+      "epoch": 0.6257066643248406,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.5463682533338286e-05,
+      "loss": 0.3055,
+      "step": 14250
+    },
+    {
+      "epoch": 0.625794482804044,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.545728771759677e-05,
+      "loss": 0.3273,
+      "step": 14252
+    },
+    {
+      "epoch": 0.6258823012832475,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.545089363263845e-05,
+      "loss": 0.3481,
+      "step": 14254
+    },
+    {
+      "epoch": 0.625970119762451,
+      "grad_norm": 1.7734375,
+      "learning_rate": 1.5444500278952982e-05,
+      "loss": 0.3302,
+      "step": 14256
+    },
+    {
+      "epoch": 0.6260579382416545,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.5438107657029975e-05,
+      "loss": 0.3343,
+      "step": 14258
+    },
+    {
+      "epoch": 0.626145756720858,
+      "grad_norm": 1.8984375,
+      "learning_rate": 1.543171576735898e-05,
+      "loss": 0.3062,
+      "step": 14260
+    },
+    {
+      "epoch": 0.6262335752000615,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.542532461042948e-05,
+      "loss": 0.3297,
+      "step": 14262
+    },
+    {
+      "epoch": 0.626321393679265,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.5418934186730923e-05,
+      "loss": 0.3328,
+      "step": 14264
+    },
+    {
+      "epoch": 0.6264092121584685,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.5412544496752686e-05,
+      "loss": 0.3082,
+      "step": 14266
+    },
+    {
+      "epoch": 0.6264970306376719,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.540615554098408e-05,
+      "loss": 0.3179,
+      "step": 14268
+    },
+    {
+      "epoch": 0.6265848491168754,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.539976731991438e-05,
+      "loss": 0.3258,
+      "step": 14270
+    },
+    {
+      "epoch": 0.6266726675960789,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.5393379834032804e-05,
+      "loss": 0.3596,
+      "step": 14272
+    },
+    {
+      "epoch": 0.6267604860752823,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.53869930838285e-05,
+      "loss": 0.3398,
+      "step": 14274
+    },
+    {
+      "epoch": 0.6268483045544859,
+      "grad_norm": 1.890625,
+      "learning_rate": 1.5380607069790577e-05,
+      "loss": 0.3567,
+      "step": 14276
+    },
+    {
+      "epoch": 0.6269361230336894,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.5374221792408067e-05,
+      "loss": 0.3124,
+      "step": 14278
+    },
+    {
+      "epoch": 0.6270239415128929,
+      "grad_norm": 1.625,
+      "learning_rate": 1.536783725216996e-05,
+      "loss": 0.3209,
+      "step": 14280
+    },
+    {
+      "epoch": 0.6271117599920963,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.5361453449565183e-05,
+      "loss": 0.3161,
+      "step": 14282
+    },
+    {
+      "epoch": 0.6271995784712998,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.535507038508261e-05,
+      "loss": 0.35,
+      "step": 14284
+    },
+    {
+      "epoch": 0.6272873969505033,
+      "grad_norm": 1.7578125,
+      "learning_rate": 1.5348688059211067e-05,
+      "loss": 0.3225,
+      "step": 14286
+    },
+    {
+      "epoch": 0.6273752154297068,
+      "grad_norm": 1.515625,
+      "learning_rate": 1.53423064724393e-05,
+      "loss": 0.3094,
+      "step": 14288
+    },
+    {
+      "epoch": 0.6274630339089102,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.5335925625256017e-05,
+      "loss": 0.3151,
+      "step": 14290
+    },
+    {
+      "epoch": 0.6275508523881138,
+      "grad_norm": 1.7734375,
+      "learning_rate": 1.5329545518149867e-05,
+      "loss": 0.3424,
+      "step": 14292
+    },
+    {
+      "epoch": 0.6276386708673173,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.532316615160944e-05,
+      "loss": 0.3404,
+      "step": 14294
+    },
+    {
+      "epoch": 0.6277264893465208,
+      "grad_norm": 2.0,
+      "learning_rate": 1.5316787526123273e-05,
+      "loss": 0.3184,
+      "step": 14296
+    },
+    {
+      "epoch": 0.6278143078257242,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.531040964217984e-05,
+      "loss": 0.3307,
+      "step": 14298
+    },
+    {
+      "epoch": 0.6279021263049277,
+      "grad_norm": 1.7890625,
+      "learning_rate": 1.5304032500267557e-05,
+      "loss": 0.3201,
+      "step": 14300
+    },
+    {
+      "epoch": 0.6279899447841312,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.529765610087479e-05,
+      "loss": 0.3219,
+      "step": 14302
+    },
+    {
+      "epoch": 0.6280777632633346,
+      "grad_norm": 1.8515625,
+      "learning_rate": 1.529128044448984e-05,
+      "loss": 0.3523,
+      "step": 14304
+    },
+    {
+      "epoch": 0.6281655817425382,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.528490553160096e-05,
+      "loss": 0.3338,
+      "step": 14306
+    },
+    {
+      "epoch": 0.6282534002217417,
+      "grad_norm": 1.859375,
+      "learning_rate": 1.5278531362696348e-05,
+      "loss": 0.3669,
+      "step": 14308
+    },
+    {
+      "epoch": 0.6283412187009452,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.5272157938264127e-05,
+      "loss": 0.3476,
+      "step": 14310
+    },
+    {
+      "epoch": 0.6284290371801486,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.5265785258792395e-05,
+      "loss": 0.3176,
+      "step": 14312
+    },
+    {
+      "epoch": 0.6285168556593521,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.5259413324769153e-05,
+      "loss": 0.311,
+      "step": 14314
+    },
+    {
+      "epoch": 0.6286046741385556,
+      "grad_norm": 1.921875,
+      "learning_rate": 1.5253042136682374e-05,
+      "loss": 0.338,
+      "step": 14316
+    },
+    {
+      "epoch": 0.6286924926177591,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.5246671695019966e-05,
+      "loss": 0.3778,
+      "step": 14318
+    },
+    {
+      "epoch": 0.6287803110969625,
+      "grad_norm": 1.7734375,
+      "learning_rate": 1.5240302000269774e-05,
+      "loss": 0.3273,
+      "step": 14320
+    },
+    {
+      "epoch": 0.6288681295761661,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.5233933052919602e-05,
+      "loss": 0.3136,
+      "step": 14322
+    },
+    {
+      "epoch": 0.6289559480553696,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.5227564853457173e-05,
+      "loss": 0.335,
+      "step": 14324
+    },
+    {
+      "epoch": 0.6290437665345731,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.5221197402370172e-05,
+      "loss": 0.3299,
+      "step": 14326
+    },
+    {
+      "epoch": 0.6291315850137765,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.5214830700146227e-05,
+      "loss": 0.3594,
+      "step": 14328
+    },
+    {
+      "epoch": 0.62921940349298,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.520846474727288e-05,
+      "loss": 0.3397,
+      "step": 14330
+    },
+    {
+      "epoch": 0.6293072219721835,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.5202099544237653e-05,
+      "loss": 0.3367,
+      "step": 14332
+    },
+    {
+      "epoch": 0.629395040451387,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.519573509152799e-05,
+      "loss": 0.3105,
+      "step": 14334
+    },
+    {
+      "epoch": 0.6294828589305904,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.5189371389631284e-05,
+      "loss": 0.3246,
+      "step": 14336
+    },
+    {
+      "epoch": 0.629570677409794,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.5183008439034873e-05,
+      "loss": 0.3085,
+      "step": 14338
+    },
+    {
+      "epoch": 0.6296584958889975,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.5176646240226025e-05,
+      "loss": 0.3442,
+      "step": 14340
+    },
+    {
+      "epoch": 0.6297463143682009,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.5170284793691963e-05,
+      "loss": 0.3371,
+      "step": 14342
+    },
+    {
+      "epoch": 0.6298341328474044,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.5163924099919846e-05,
+      "loss": 0.3339,
+      "step": 14344
+    },
+    {
+      "epoch": 0.6299219513266079,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.5157564159396781e-05,
+      "loss": 0.3161,
+      "step": 14346
+    },
+    {
+      "epoch": 0.6300097698058114,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.5151204972609818e-05,
+      "loss": 0.3199,
+      "step": 14348
+    },
+    {
+      "epoch": 0.6300975882850148,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.5144846540045932e-05,
+      "loss": 0.3202,
+      "step": 14350
+    },
+    {
+      "epoch": 0.6301854067642184,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.5138488862192063e-05,
+      "loss": 0.3289,
+      "step": 14352
+    },
+    {
+      "epoch": 0.6302732252434219,
+      "grad_norm": 1.796875,
+      "learning_rate": 1.5132131939535076e-05,
+      "loss": 0.3225,
+      "step": 14354
+    },
+    {
+      "epoch": 0.6303610437226254,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.5125775772561795e-05,
+      "loss": 0.3161,
+      "step": 14356
+    },
+    {
+      "epoch": 0.6304488622018288,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.5119420361758982e-05,
+      "loss": 0.3352,
+      "step": 14358
+    },
+    {
+      "epoch": 0.6305366806810323,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.5113065707613317e-05,
+      "loss": 0.3359,
+      "step": 14360
+    },
+    {
+      "epoch": 0.6306244991602358,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.5106711810611446e-05,
+      "loss": 0.3457,
+      "step": 14362
+    },
+    {
+      "epoch": 0.6307123176394392,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.5100358671239964e-05,
+      "loss": 0.3188,
+      "step": 14364
+    },
+    {
+      "epoch": 0.6308001361186427,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.5094006289985385e-05,
+      "loss": 0.3278,
+      "step": 14366
+    },
+    {
+      "epoch": 0.6308879545978463,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.5087654667334174e-05,
+      "loss": 0.3287,
+      "step": 14368
+    },
+    {
+      "epoch": 0.6309757730770498,
+      "grad_norm": 1.890625,
+      "learning_rate": 1.5081303803772751e-05,
+      "loss": 0.3183,
+      "step": 14370
+    },
+    {
+      "epoch": 0.6310635915562532,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.5074953699787452e-05,
+      "loss": 0.3278,
+      "step": 14372
+    },
+    {
+      "epoch": 0.6311514100354567,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.506860435586459e-05,
+      "loss": 0.2985,
+      "step": 14374
+    },
+    {
+      "epoch": 0.6312392285146602,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.506225577249038e-05,
+      "loss": 0.3348,
+      "step": 14376
+    },
+    {
+      "epoch": 0.6313270469938637,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.5055907950151004e-05,
+      "loss": 0.3477,
+      "step": 14378
+    },
+    {
+      "epoch": 0.6314148654730671,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.5049560889332581e-05,
+      "loss": 0.3124,
+      "step": 14380
+    },
+    {
+      "epoch": 0.6315026839522706,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.5043214590521174e-05,
+      "loss": 0.3292,
+      "step": 14382
+    },
+    {
+      "epoch": 0.6315905024314742,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.5036869054202782e-05,
+      "loss": 0.3574,
+      "step": 14384
+    },
+    {
+      "epoch": 0.6316783209106777,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.5030524280863342e-05,
+      "loss": 0.2946,
+      "step": 14386
+    },
+    {
+      "epoch": 0.6317661393898811,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.5024180270988741e-05,
+      "loss": 0.352,
+      "step": 14388
+    },
+    {
+      "epoch": 0.6318539578690846,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.5017837025064818e-05,
+      "loss": 0.3146,
+      "step": 14390
+    },
+    {
+      "epoch": 0.6319417763482881,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.501149454357732e-05,
+      "loss": 0.3527,
+      "step": 14392
+    },
+    {
+      "epoch": 0.6320295948274915,
+      "grad_norm": 1.859375,
+      "learning_rate": 1.5005152827011962e-05,
+      "loss": 0.3064,
+      "step": 14394
+    },
+    {
+      "epoch": 0.632117413306695,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.4998811875854396e-05,
+      "loss": 0.3194,
+      "step": 14396
+    },
+    {
+      "epoch": 0.6322052317858986,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.4992471690590216e-05,
+      "loss": 0.3234,
+      "step": 14398
+    },
+    {
+      "epoch": 0.6322930502651021,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.4986132271704955e-05,
+      "loss": 0.3459,
+      "step": 14400
+    },
+    {
+      "epoch": 0.6323808687443055,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.4979793619684082e-05,
+      "loss": 0.3138,
+      "step": 14402
+    },
+    {
+      "epoch": 0.632468687223509,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.4973455735013015e-05,
+      "loss": 0.327,
+      "step": 14404
+    },
+    {
+      "epoch": 0.6325565057027125,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.4967118618177112e-05,
+      "loss": 0.3255,
+      "step": 14406
+    },
+    {
+      "epoch": 0.632644324181916,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.4960782269661672e-05,
+      "loss": 0.3421,
+      "step": 14408
+    },
+    {
+      "epoch": 0.6327321426611194,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.4954446689951934e-05,
+      "loss": 0.3058,
+      "step": 14410
+    },
+    {
+      "epoch": 0.6328199611403229,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.4948111879533071e-05,
+      "loss": 0.3665,
+      "step": 14412
+    },
+    {
+      "epoch": 0.6329077796195265,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.4941777838890215e-05,
+      "loss": 0.3343,
+      "step": 14414
+    },
+    {
+      "epoch": 0.63299559809873,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.4935444568508419e-05,
+      "loss": 0.3379,
+      "step": 14416
+    },
+    {
+      "epoch": 0.6330834165779334,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.4929112068872691e-05,
+      "loss": 0.3326,
+      "step": 14418
+    },
+    {
+      "epoch": 0.6331712350571369,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.4922780340467984e-05,
+      "loss": 0.3167,
+      "step": 14420
+    },
+    {
+      "epoch": 0.6332590535363404,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.4916449383779169e-05,
+      "loss": 0.3186,
+      "step": 14422
+    },
+    {
+      "epoch": 0.6333468720155438,
+      "grad_norm": 1.8125,
+      "learning_rate": 1.4910119199291072e-05,
+      "loss": 0.3316,
+      "step": 14424
+    },
+    {
+      "epoch": 0.6334346904947473,
+      "grad_norm": 1.8359375,
+      "learning_rate": 1.4903789787488474e-05,
+      "loss": 0.3321,
+      "step": 14426
+    },
+    {
+      "epoch": 0.6335225089739508,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.4897461148856068e-05,
+      "loss": 0.3481,
+      "step": 14428
+    },
+    {
+      "epoch": 0.6336103274531544,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.4891133283878509e-05,
+      "loss": 0.3035,
+      "step": 14430
+    },
+    {
+      "epoch": 0.6336981459323578,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.4884806193040384e-05,
+      "loss": 0.3554,
+      "step": 14432
+    },
+    {
+      "epoch": 0.6337859644115613,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.487847987682623e-05,
+      "loss": 0.3422,
+      "step": 14434
+    },
+    {
+      "epoch": 0.6338737828907648,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.4872154335720518e-05,
+      "loss": 0.3272,
+      "step": 14436
+    },
+    {
+      "epoch": 0.6339616013699683,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.4865829570207645e-05,
+      "loss": 0.3523,
+      "step": 14438
+    },
+    {
+      "epoch": 0.6340494198491717,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.4859505580771977e-05,
+      "loss": 0.3274,
+      "step": 14440
+    },
+    {
+      "epoch": 0.6341372383283752,
+      "grad_norm": 1.625,
+      "learning_rate": 1.48531823678978e-05,
+      "loss": 0.3391,
+      "step": 14442
+    },
+    {
+      "epoch": 0.6342250568075787,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.484685993206935e-05,
+      "loss": 0.3422,
+      "step": 14444
+    },
+    {
+      "epoch": 0.6343128752867823,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.4840538273770807e-05,
+      "loss": 0.3105,
+      "step": 14446
+    },
+    {
+      "epoch": 0.6344006937659857,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.4834217393486272e-05,
+      "loss": 0.3629,
+      "step": 14448
+    },
+    {
+      "epoch": 0.6344885122451892,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.4827897291699816e-05,
+      "loss": 0.3606,
+      "step": 14450
+    },
+    {
+      "epoch": 0.6345763307243927,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.4821577968895414e-05,
+      "loss": 0.3118,
+      "step": 14452
+    },
+    {
+      "epoch": 0.6346641492035962,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.4815259425557013e-05,
+      "loss": 0.3374,
+      "step": 14454
+    },
+    {
+      "epoch": 0.6347519676827996,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.4808941662168485e-05,
+      "loss": 0.3343,
+      "step": 14456
+    },
+    {
+      "epoch": 0.6348397861620031,
+      "grad_norm": 1.625,
+      "learning_rate": 1.4802624679213645e-05,
+      "loss": 0.3287,
+      "step": 14458
+    },
+    {
+      "epoch": 0.6349276046412067,
+      "grad_norm": 1.78125,
+      "learning_rate": 1.4796308477176258e-05,
+      "loss": 0.3231,
+      "step": 14460
+    },
+    {
+      "epoch": 0.6350154231204101,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.4789993056540013e-05,
+      "loss": 0.3116,
+      "step": 14462
+    },
+    {
+      "epoch": 0.6351032415996136,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.4783678417788544e-05,
+      "loss": 0.3213,
+      "step": 14464
+    },
+    {
+      "epoch": 0.6351910600788171,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.477736456140543e-05,
+      "loss": 0.313,
+      "step": 14466
+    },
+    {
+      "epoch": 0.6352788785580206,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.4771051487874189e-05,
+      "loss": 0.321,
+      "step": 14468
+    },
+    {
+      "epoch": 0.635366697037224,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.4764739197678279e-05,
+      "loss": 0.332,
+      "step": 14470
+    },
+    {
+      "epoch": 0.6354545155164275,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.4758427691301096e-05,
+      "loss": 0.3222,
+      "step": 14472
+    },
+    {
+      "epoch": 0.635542333995631,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.475211696922597e-05,
+      "loss": 0.3467,
+      "step": 14474
+    },
+    {
+      "epoch": 0.6356301524748346,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.4745807031936188e-05,
+      "loss": 0.3384,
+      "step": 14476
+    },
+    {
+      "epoch": 0.635717970954038,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.473949787991496e-05,
+      "loss": 0.3454,
+      "step": 14478
+    },
+    {
+      "epoch": 0.6358057894332415,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.4733189513645443e-05,
+      "loss": 0.3301,
+      "step": 14480
+    },
+    {
+      "epoch": 0.635893607912445,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.4726881933610742e-05,
+      "loss": 0.3258,
+      "step": 14482
+    },
+    {
+      "epoch": 0.6359814263916485,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.472057514029388e-05,
+      "loss": 0.3416,
+      "step": 14484
+    },
+    {
+      "epoch": 0.6360692448708519,
+      "grad_norm": 1.5,
+      "learning_rate": 1.4714269134177836e-05,
+      "loss": 0.3131,
+      "step": 14486
+    },
+    {
+      "epoch": 0.6361570633500554,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.4707963915745531e-05,
+      "loss": 0.3115,
+      "step": 14488
+    },
+    {
+      "epoch": 0.6362448818292589,
+      "grad_norm": 1.515625,
+      "learning_rate": 1.4701659485479813e-05,
+      "loss": 0.3337,
+      "step": 14490
+    },
+    {
+      "epoch": 0.6363327003084625,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.4695355843863479e-05,
+      "loss": 0.348,
+      "step": 14492
+    },
+    {
+      "epoch": 0.6364205187876659,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.4689052991379266e-05,
+      "loss": 0.3505,
+      "step": 14494
+    },
+    {
+      "epoch": 0.6365083372668694,
+      "grad_norm": 1.9296875,
+      "learning_rate": 1.4682750928509845e-05,
+      "loss": 0.3427,
+      "step": 14496
+    },
+    {
+      "epoch": 0.6365961557460729,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.4676449655737837e-05,
+      "loss": 0.3425,
+      "step": 14498
+    },
+    {
+      "epoch": 0.6366839742252763,
+      "grad_norm": 1.7890625,
+      "learning_rate": 1.4670149173545783e-05,
+      "loss": 0.3313,
+      "step": 14500
+    },
+    {
+      "epoch": 0.6367717927044798,
+      "grad_norm": 1.625,
+      "learning_rate": 1.4663849482416184e-05,
+      "loss": 0.3294,
+      "step": 14502
+    },
+    {
+      "epoch": 0.6368596111836833,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.4657550582831467e-05,
+      "loss": 0.3425,
+      "step": 14504
+    },
+    {
+      "epoch": 0.6369474296628869,
+      "grad_norm": 1.828125,
+      "learning_rate": 1.4651252475274007e-05,
+      "loss": 0.3109,
+      "step": 14506
+    },
+    {
+      "epoch": 0.6370352481420903,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.4644955160226118e-05,
+      "loss": 0.3427,
+      "step": 14508
+    },
+    {
+      "epoch": 0.6371230666212938,
+      "grad_norm": 1.8125,
+      "learning_rate": 1.4638658638170038e-05,
+      "loss": 0.3422,
+      "step": 14510
+    },
+    {
+      "epoch": 0.6372108851004973,
+      "grad_norm": 1.9453125,
+      "learning_rate": 1.4632362909587977e-05,
+      "loss": 0.3501,
+      "step": 14512
+    },
+    {
+      "epoch": 0.6372987035797008,
+      "grad_norm": 1.8828125,
+      "learning_rate": 1.4626067974962038e-05,
+      "loss": 0.3177,
+      "step": 14514
+    },
+    {
+      "epoch": 0.6373865220589042,
+      "grad_norm": 1.5,
+      "learning_rate": 1.46197738347743e-05,
+      "loss": 0.3215,
+      "step": 14516
+    },
+    {
+      "epoch": 0.6374743405381077,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.461348048950677e-05,
+      "loss": 0.3276,
+      "step": 14518
+    },
+    {
+      "epoch": 0.6375621590173112,
+      "grad_norm": 1.9140625,
+      "learning_rate": 1.4607187939641393e-05,
+      "loss": 0.3418,
+      "step": 14520
+    },
+    {
+      "epoch": 0.6376499774965148,
+      "grad_norm": 1.7578125,
+      "learning_rate": 1.4600896185660057e-05,
+      "loss": 0.3179,
+      "step": 14522
+    },
+    {
+      "epoch": 0.6377377959757182,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.459460522804459e-05,
+      "loss": 0.328,
+      "step": 14524
+    },
+    {
+      "epoch": 0.6378256144549217,
+      "grad_norm": 1.9453125,
+      "learning_rate": 1.4588315067276737e-05,
+      "loss": 0.3368,
+      "step": 14526
+    },
+    {
+      "epoch": 0.6379134329341252,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.458202570383822e-05,
+      "loss": 0.3467,
+      "step": 14528
+    },
+    {
+      "epoch": 0.6380012514133286,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.4575737138210674e-05,
+      "loss": 0.2989,
+      "step": 14530
+    },
+    {
+      "epoch": 0.6380890698925321,
+      "grad_norm": 1.7734375,
+      "learning_rate": 1.4569449370875665e-05,
+      "loss": 0.3633,
+      "step": 14532
+    },
+    {
+      "epoch": 0.6381768883717356,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.4563162402314737e-05,
+      "loss": 0.3455,
+      "step": 14534
+    },
+    {
+      "epoch": 0.6382647068509391,
+      "grad_norm": 1.78125,
+      "learning_rate": 1.4556876233009323e-05,
+      "loss": 0.334,
+      "step": 14536
+    },
+    {
+      "epoch": 0.6383525253301426,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.4550590863440838e-05,
+      "loss": 0.3335,
+      "step": 14538
+    },
+    {
+      "epoch": 0.6384403438093461,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.4544306294090612e-05,
+      "loss": 0.3423,
+      "step": 14540
+    },
+    {
+      "epoch": 0.6385281622885496,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.4538022525439915e-05,
+      "loss": 0.3175,
+      "step": 14542
+    },
+    {
+      "epoch": 0.638615980767753,
+      "grad_norm": 1.4921875,
+      "learning_rate": 1.4531739557969964e-05,
+      "loss": 0.3087,
+      "step": 14544
+    },
+    {
+      "epoch": 0.6387037992469565,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.4525457392161895e-05,
+      "loss": 0.3289,
+      "step": 14546
+    },
+    {
+      "epoch": 0.63879161772616,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.4519176028496817e-05,
+      "loss": 0.3231,
+      "step": 14548
+    },
+    {
+      "epoch": 0.6388794362053635,
+      "grad_norm": 1.796875,
+      "learning_rate": 1.4512895467455745e-05,
+      "loss": 0.3068,
+      "step": 14550
+    },
+    {
+      "epoch": 0.638967254684567,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.4506615709519661e-05,
+      "loss": 0.3506,
+      "step": 14552
+    },
+    {
+      "epoch": 0.6390550731637705,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.4500336755169464e-05,
+      "loss": 0.3405,
+      "step": 14554
+    },
+    {
+      "epoch": 0.639142891642974,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.449405860488598e-05,
+      "loss": 0.3079,
+      "step": 14556
+    },
+    {
+      "epoch": 0.6392307101221775,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.4487781259150018e-05,
+      "loss": 0.3093,
+      "step": 14558
+    },
+    {
+      "epoch": 0.6393185286013809,
+      "grad_norm": 1.9609375,
+      "learning_rate": 1.448150471844228e-05,
+      "loss": 0.3463,
+      "step": 14560
+    },
+    {
+      "epoch": 0.6394063470805844,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.4475228983243438e-05,
+      "loss": 0.3015,
+      "step": 14562
+    },
+    {
+      "epoch": 0.6394941655597879,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.4468954054034087e-05,
+      "loss": 0.3246,
+      "step": 14564
+    },
+    {
+      "epoch": 0.6395819840389914,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.4462679931294749e-05,
+      "loss": 0.3138,
+      "step": 14566
+    },
+    {
+      "epoch": 0.6396698025181949,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.4456406615505916e-05,
+      "loss": 0.3133,
+      "step": 14568
+    },
+    {
+      "epoch": 0.6397576209973984,
+      "grad_norm": 1.9609375,
+      "learning_rate": 1.4450134107147983e-05,
+      "loss": 0.3602,
+      "step": 14570
+    },
+    {
+      "epoch": 0.6398454394766019,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.4443862406701325e-05,
+      "loss": 0.331,
+      "step": 14572
+    },
+    {
+      "epoch": 0.6399332579558054,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.4437591514646198e-05,
+      "loss": 0.3548,
+      "step": 14574
+    },
+    {
+      "epoch": 0.6400210764350088,
+      "grad_norm": 1.8671875,
+      "learning_rate": 1.4431321431462858e-05,
+      "loss": 0.3461,
+      "step": 14576
+    },
+    {
+      "epoch": 0.6401088949142123,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.4425052157631441e-05,
+      "loss": 0.3374,
+      "step": 14578
+    },
+    {
+      "epoch": 0.6401967133934158,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.4418783693632077e-05,
+      "loss": 0.3284,
+      "step": 14580
+    },
+    {
+      "epoch": 0.6402845318726192,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.4412516039944793e-05,
+      "loss": 0.3184,
+      "step": 14582
+    },
+    {
+      "epoch": 0.6403723503518228,
+      "grad_norm": 1.8828125,
+      "learning_rate": 1.4406249197049559e-05,
+      "loss": 0.32,
+      "step": 14584
+    },
+    {
+      "epoch": 0.6404601688310263,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.4399983165426312e-05,
+      "loss": 0.3213,
+      "step": 14586
+    },
+    {
+      "epoch": 0.6405479873102298,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.4393717945554885e-05,
+      "loss": 0.3198,
+      "step": 14588
+    },
+    {
+      "epoch": 0.6406358057894332,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.438745353791509e-05,
+      "loss": 0.3314,
+      "step": 14590
+    },
+    {
+      "epoch": 0.6407236242686367,
+      "grad_norm": 1.484375,
+      "learning_rate": 1.4381189942986644e-05,
+      "loss": 0.3409,
+      "step": 14592
+    },
+    {
+      "epoch": 0.6408114427478402,
+      "grad_norm": 1.5078125,
+      "learning_rate": 1.4374927161249212e-05,
+      "loss": 0.3322,
+      "step": 14594
+    },
+    {
+      "epoch": 0.6408992612270437,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.4368665193182416e-05,
+      "loss": 0.3318,
+      "step": 14596
+    },
+    {
+      "epoch": 0.6409870797062472,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.4362404039265775e-05,
+      "loss": 0.3138,
+      "step": 14598
+    },
+    {
+      "epoch": 0.6410748981854507,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.435614369997879e-05,
+      "loss": 0.3314,
+      "step": 14600
+    },
+    {
+      "epoch": 0.6411627166646542,
+      "grad_norm": 1.5,
+      "learning_rate": 1.4349884175800876e-05,
+      "loss": 0.2977,
+      "step": 14602
+    },
+    {
+      "epoch": 0.6412505351438577,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.4343625467211386e-05,
+      "loss": 0.3434,
+      "step": 14604
+    },
+    {
+      "epoch": 0.6413383536230611,
+      "grad_norm": 1.8125,
+      "learning_rate": 1.4337367574689609e-05,
+      "loss": 0.3048,
+      "step": 14606
+    },
+    {
+      "epoch": 0.6414261721022646,
+      "grad_norm": 1.8359375,
+      "learning_rate": 1.4331110498714773e-05,
+      "loss": 0.3218,
+      "step": 14608
+    },
+    {
+      "epoch": 0.6415139905814681,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.4324854239766059e-05,
+      "loss": 0.3395,
+      "step": 14610
+    },
+    {
+      "epoch": 0.6416018090606715,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.4318598798322557e-05,
+      "loss": 0.3108,
+      "step": 14612
+    },
+    {
+      "epoch": 0.6416896275398751,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.4312344174863329e-05,
+      "loss": 0.3237,
+      "step": 14614
+    },
+    {
+      "epoch": 0.6417774460190786,
+      "grad_norm": 1.625,
+      "learning_rate": 1.4306090369867348e-05,
+      "loss": 0.343,
+      "step": 14616
+    },
+    {
+      "epoch": 0.6418652644982821,
+      "grad_norm": 1.5,
+      "learning_rate": 1.429983738381352e-05,
+      "loss": 0.3108,
+      "step": 14618
+    },
+    {
+      "epoch": 0.6419530829774855,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.4293585217180717e-05,
+      "loss": 0.3313,
+      "step": 14620
+    },
+    {
+      "epoch": 0.642040901456689,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.4287333870447716e-05,
+      "loss": 0.3604,
+      "step": 14622
+    },
+    {
+      "epoch": 0.6421287199358925,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.4281083344093265e-05,
+      "loss": 0.3235,
+      "step": 14624
+    },
+    {
+      "epoch": 0.642216538415096,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.4274833638596024e-05,
+      "loss": 0.3281,
+      "step": 14626
+    },
+    {
+      "epoch": 0.6423043568942994,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.4268584754434583e-05,
+      "loss": 0.309,
+      "step": 14628
+    },
+    {
+      "epoch": 0.642392175373503,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.4262336692087503e-05,
+      "loss": 0.3422,
+      "step": 14630
+    },
+    {
+      "epoch": 0.6424799938527065,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.4256089452033241e-05,
+      "loss": 0.3389,
+      "step": 14632
+    },
+    {
+      "epoch": 0.64256781233191,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.4249843034750246e-05,
+      "loss": 0.3091,
+      "step": 14634
+    },
+    {
+      "epoch": 0.6426556308111134,
+      "grad_norm": 1.625,
+      "learning_rate": 1.4243597440716827e-05,
+      "loss": 0.3495,
+      "step": 14636
+    },
+    {
+      "epoch": 0.6427434492903169,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.4237352670411308e-05,
+      "loss": 0.3063,
+      "step": 14638
+    },
+    {
+      "epoch": 0.6428312677695204,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.423110872431189e-05,
+      "loss": 0.3091,
+      "step": 14640
+    },
+    {
+      "epoch": 0.6429190862487238,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.4224865602896757e-05,
+      "loss": 0.3372,
+      "step": 14642
+    },
+    {
+      "epoch": 0.6430069047279273,
+      "grad_norm": 1.5078125,
+      "learning_rate": 1.4218623306643997e-05,
+      "loss": 0.3061,
+      "step": 14644
+    },
+    {
+      "epoch": 0.6430947232071309,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.421238183603164e-05,
+      "loss": 0.3507,
+      "step": 14646
+    },
+    {
+      "epoch": 0.6431825416863344,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.4206141191537682e-05,
+      "loss": 0.3205,
+      "step": 14648
+    },
+    {
+      "epoch": 0.6432703601655378,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.4199901373640005e-05,
+      "loss": 0.3764,
+      "step": 14650
+    },
+    {
+      "epoch": 0.6433581786447413,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.419366238281648e-05,
+      "loss": 0.3491,
+      "step": 14652
+    },
+    {
+      "epoch": 0.6434459971239448,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.4187424219544882e-05,
+      "loss": 0.3463,
+      "step": 14654
+    },
+    {
+      "epoch": 0.6435338156031483,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.4181186884302916e-05,
+      "loss": 0.343,
+      "step": 14656
+    },
+    {
+      "epoch": 0.6436216340823517,
+      "grad_norm": 1.8984375,
+      "learning_rate": 1.4174950377568264e-05,
+      "loss": 0.373,
+      "step": 14658
+    },
+    {
+      "epoch": 0.6437094525615553,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.4168714699818498e-05,
+      "loss": 0.3412,
+      "step": 14660
+    },
+    {
+      "epoch": 0.6437972710407588,
+      "grad_norm": 1.90625,
+      "learning_rate": 1.4162479851531163e-05,
+      "loss": 0.3298,
+      "step": 14662
+    },
+    {
+      "epoch": 0.6438850895199623,
+      "grad_norm": 1.625,
+      "learning_rate": 1.4156245833183723e-05,
+      "loss": 0.3244,
+      "step": 14664
+    },
+    {
+      "epoch": 0.6439729079991657,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.4150012645253575e-05,
+      "loss": 0.3533,
+      "step": 14666
+    },
+    {
+      "epoch": 0.6440607264783692,
+      "grad_norm": 1.8125,
+      "learning_rate": 1.4143780288218058e-05,
+      "loss": 0.3331,
+      "step": 14668
+    },
+    {
+      "epoch": 0.6441485449575727,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.4137548762554443e-05,
+      "loss": 0.3425,
+      "step": 14670
+    },
+    {
+      "epoch": 0.6442363634367761,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.4131318068739951e-05,
+      "loss": 0.3373,
+      "step": 14672
+    },
+    {
+      "epoch": 0.6443241819159796,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.4125088207251722e-05,
+      "loss": 0.3299,
+      "step": 14674
+    },
+    {
+      "epoch": 0.6444120003951832,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.4118859178566853e-05,
+      "loss": 0.3592,
+      "step": 14676
+    },
+    {
+      "epoch": 0.6444998188743867,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.4112630983162356e-05,
+      "loss": 0.3314,
+      "step": 14678
+    },
+    {
+      "epoch": 0.6445876373535901,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.410640362151518e-05,
+      "loss": 0.3222,
+      "step": 14680
+    },
+    {
+      "epoch": 0.6446754558327936,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.4100177094102235e-05,
+      "loss": 0.3264,
+      "step": 14682
+    },
+    {
+      "epoch": 0.6447632743119971,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.4093951401400335e-05,
+      "loss": 0.3309,
+      "step": 14684
+    },
+    {
+      "epoch": 0.6448510927912006,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.4087726543886254e-05,
+      "loss": 0.3072,
+      "step": 14686
+    },
+    {
+      "epoch": 0.644938911270404,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.4081502522036693e-05,
+      "loss": 0.3241,
+      "step": 14688
+    },
+    {
+      "epoch": 0.6450267297496075,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.4075279336328279e-05,
+      "loss": 0.3302,
+      "step": 14690
+    },
+    {
+      "epoch": 0.6451145482288111,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.40690569872376e-05,
+      "loss": 0.3425,
+      "step": 14692
+    },
+    {
+      "epoch": 0.6452023667080146,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.4062835475241148e-05,
+      "loss": 0.3169,
+      "step": 14694
+    },
+    {
+      "epoch": 0.645290185187218,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.4056614800815396e-05,
+      "loss": 0.3517,
+      "step": 14696
+    },
+    {
+      "epoch": 0.6453780036664215,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.4050394964436686e-05,
+      "loss": 0.3589,
+      "step": 14698
+    },
+    {
+      "epoch": 0.645465822145625,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.4044175966581363e-05,
+      "loss": 0.3343,
+      "step": 14700
+    },
+    {
+      "epoch": 0.6455536406248285,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.4037957807725666e-05,
+      "loss": 0.3384,
+      "step": 14702
+    },
+    {
+      "epoch": 0.6456414591040319,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.403174048834579e-05,
+      "loss": 0.3348,
+      "step": 14704
+    },
+    {
+      "epoch": 0.6457292775832355,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.4025524008917861e-05,
+      "loss": 0.3253,
+      "step": 14706
+    },
+    {
+      "epoch": 0.645817096062439,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.4019308369917928e-05,
+      "loss": 0.3263,
+      "step": 14708
+    },
+    {
+      "epoch": 0.6459049145416424,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.4013093571821994e-05,
+      "loss": 0.3211,
+      "step": 14710
+    },
+    {
+      "epoch": 0.6459927330208459,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.4006879615105984e-05,
+      "loss": 0.3175,
+      "step": 14712
+    },
+    {
+      "epoch": 0.6460805515000494,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.400066650024578e-05,
+      "loss": 0.3165,
+      "step": 14714
+    },
+    {
+      "epoch": 0.6461683699792529,
+      "grad_norm": 1.453125,
+      "learning_rate": 1.3994454227717168e-05,
+      "loss": 0.3272,
+      "step": 14716
+    },
+    {
+      "epoch": 0.6462561884584563,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.398824279799588e-05,
+      "loss": 0.3444,
+      "step": 14718
+    },
+    {
+      "epoch": 0.6463440069376598,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.3982032211557609e-05,
+      "loss": 0.3389,
+      "step": 14720
+    },
+    {
+      "epoch": 0.6464318254168634,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.3975822468877942e-05,
+      "loss": 0.3294,
+      "step": 14722
+    },
+    {
+      "epoch": 0.6465196438960669,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.396961357043244e-05,
+      "loss": 0.3239,
+      "step": 14724
+    },
+    {
+      "epoch": 0.6466074623752703,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.3963405516696579e-05,
+      "loss": 0.3458,
+      "step": 14726
+    },
+    {
+      "epoch": 0.6466952808544738,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.3957198308145769e-05,
+      "loss": 0.3198,
+      "step": 14728
+    },
+    {
+      "epoch": 0.6467830993336773,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.395099194525536e-05,
+      "loss": 0.3392,
+      "step": 14730
+    },
+    {
+      "epoch": 0.6468709178128808,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.3944786428500623e-05,
+      "loss": 0.3456,
+      "step": 14732
+    },
+    {
+      "epoch": 0.6469587362920842,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.3938581758356806e-05,
+      "loss": 0.3485,
+      "step": 14734
+    },
+    {
+      "epoch": 0.6470465547712877,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.3932377935299035e-05,
+      "loss": 0.3474,
+      "step": 14736
+    },
+    {
+      "epoch": 0.6471343732504913,
+      "grad_norm": 1.75,
+      "learning_rate": 1.3926174959802429e-05,
+      "loss": 0.3339,
+      "step": 14738
+    },
+    {
+      "epoch": 0.6472221917296948,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.3919972832341997e-05,
+      "loss": 0.3329,
+      "step": 14740
+    },
+    {
+      "epoch": 0.6473100102088982,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.391377155339269e-05,
+      "loss": 0.3519,
+      "step": 14742
+    },
+    {
+      "epoch": 0.6473978286881017,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.3907571123429427e-05,
+      "loss": 0.3232,
+      "step": 14744
+    },
+    {
+      "epoch": 0.6474856471673052,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.3901371542927016e-05,
+      "loss": 0.3239,
+      "step": 14746
+    },
+    {
+      "epoch": 0.6475734656465086,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.3895172812360244e-05,
+      "loss": 0.3496,
+      "step": 14748
+    },
+    {
+      "epoch": 0.6476612841257121,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.3888974932203797e-05,
+      "loss": 0.3183,
+      "step": 14750
+    },
+    {
+      "epoch": 0.6477491026049157,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.3882777902932306e-05,
+      "loss": 0.3365,
+      "step": 14752
+    },
+    {
+      "epoch": 0.6478369210841192,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.387658172502036e-05,
+      "loss": 0.314,
+      "step": 14754
+    },
+    {
+      "epoch": 0.6479247395633226,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.3870386398942447e-05,
+      "loss": 0.3087,
+      "step": 14756
+    },
+    {
+      "epoch": 0.6480125580425261,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.3864191925173015e-05,
+      "loss": 0.3027,
+      "step": 14758
+    },
+    {
+      "epoch": 0.6481003765217296,
+      "grad_norm": 1.7890625,
+      "learning_rate": 1.3857998304186423e-05,
+      "loss": 0.3307,
+      "step": 14760
+    },
+    {
+      "epoch": 0.648188195000933,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.3851805536457003e-05,
+      "loss": 0.3298,
+      "step": 14762
+    },
+    {
+      "epoch": 0.6482760134801365,
+      "grad_norm": 1.8046875,
+      "learning_rate": 1.3845613622458986e-05,
+      "loss": 0.3379,
+      "step": 14764
+    },
+    {
+      "epoch": 0.64836383195934,
+      "grad_norm": 1.625,
+      "learning_rate": 1.3839422562666543e-05,
+      "loss": 0.3376,
+      "step": 14766
+    },
+    {
+      "epoch": 0.6484516504385436,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.3833232357553804e-05,
+      "loss": 0.3247,
+      "step": 14768
+    },
+    {
+      "epoch": 0.648539468917747,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.3827043007594798e-05,
+      "loss": 0.2991,
+      "step": 14770
+    },
+    {
+      "epoch": 0.6486272873969505,
+      "grad_norm": 1.7578125,
+      "learning_rate": 1.3820854513263532e-05,
+      "loss": 0.3261,
+      "step": 14772
+    },
+    {
+      "epoch": 0.648715105876154,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.381466687503389e-05,
+      "loss": 0.3293,
+      "step": 14774
+    },
+    {
+      "epoch": 0.6488029243553575,
+      "grad_norm": 1.515625,
+      "learning_rate": 1.3808480093379755e-05,
+      "loss": 0.3416,
+      "step": 14776
+    },
+    {
+      "epoch": 0.6488907428345609,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.3802294168774893e-05,
+      "loss": 0.3469,
+      "step": 14778
+    },
+    {
+      "epoch": 0.6489785613137644,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.3796109101693022e-05,
+      "loss": 0.3225,
+      "step": 14780
+    },
+    {
+      "epoch": 0.6490663797929679,
+      "grad_norm": 1.90625,
+      "learning_rate": 1.3789924892607808e-05,
+      "loss": 0.3461,
+      "step": 14782
+    },
+    {
+      "epoch": 0.6491541982721715,
+      "grad_norm": 1.625,
+      "learning_rate": 1.3783741541992826e-05,
+      "loss": 0.332,
+      "step": 14784
+    },
+    {
+      "epoch": 0.6492420167513749,
+      "grad_norm": 1.890625,
+      "learning_rate": 1.3777559050321615e-05,
+      "loss": 0.3151,
+      "step": 14786
+    },
+    {
+      "epoch": 0.6493298352305784,
+      "grad_norm": 1.7734375,
+      "learning_rate": 1.3771377418067621e-05,
+      "loss": 0.2975,
+      "step": 14788
+    },
+    {
+      "epoch": 0.6494176537097819,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.3765196645704236e-05,
+      "loss": 0.3272,
+      "step": 14790
+    },
+    {
+      "epoch": 0.6495054721889854,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.3759016733704783e-05,
+      "loss": 0.3647,
+      "step": 14792
+    },
+    {
+      "epoch": 0.6495932906681888,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.375283768254252e-05,
+      "loss": 0.3253,
+      "step": 14794
+    },
+    {
+      "epoch": 0.6496811091473923,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.3746659492690645e-05,
+      "loss": 0.3017,
+      "step": 14796
+    },
+    {
+      "epoch": 0.6497689276265958,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.3740482164622279e-05,
+      "loss": 0.3483,
+      "step": 14798
+    },
+    {
+      "epoch": 0.6498567461057994,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.3734305698810496e-05,
+      "loss": 0.3406,
+      "step": 14800
+    },
+    {
+      "epoch": 0.6499445645850028,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.3728130095728284e-05,
+      "loss": 0.3054,
+      "step": 14802
+    },
+    {
+      "epoch": 0.6500323830642063,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.3721955355848562e-05,
+      "loss": 0.335,
+      "step": 14804
+    },
+    {
+      "epoch": 0.6501202015434098,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.371578147964421e-05,
+      "loss": 0.3368,
+      "step": 14806
+    },
+    {
+      "epoch": 0.6502080200226132,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.3709608467588008e-05,
+      "loss": 0.3173,
+      "step": 14808
+    },
+    {
+      "epoch": 0.6502958385018167,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.3703436320152708e-05,
+      "loss": 0.3083,
+      "step": 14810
+    },
+    {
+      "epoch": 0.6503836569810202,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.3697265037810964e-05,
+      "loss": 0.3077,
+      "step": 14812
+    },
+    {
+      "epoch": 0.6504714754602238,
+      "grad_norm": 1.515625,
+      "learning_rate": 1.3691094621035358e-05,
+      "loss": 0.3309,
+      "step": 14814
+    },
+    {
+      "epoch": 0.6505592939394272,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.368492507029845e-05,
+      "loss": 0.3468,
+      "step": 14816
+    },
+    {
+      "epoch": 0.6506471124186307,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.367875638607269e-05,
+      "loss": 0.3506,
+      "step": 14818
+    },
+    {
+      "epoch": 0.6507349308978342,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.367258856883048e-05,
+      "loss": 0.3203,
+      "step": 14820
+    },
+    {
+      "epoch": 0.6508227493770377,
+      "grad_norm": 1.765625,
+      "learning_rate": 1.3666421619044146e-05,
+      "loss": 0.3387,
+      "step": 14822
+    },
+    {
+      "epoch": 0.6509105678562411,
+      "grad_norm": 1.625,
+      "learning_rate": 1.366025553718597e-05,
+      "loss": 0.335,
+      "step": 14824
+    },
+    {
+      "epoch": 0.6509983863354446,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.3654090323728142e-05,
+      "loss": 0.3325,
+      "step": 14826
+    },
+    {
+      "epoch": 0.6510862048146481,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.3647925979142789e-05,
+      "loss": 0.326,
+      "step": 14828
+    },
+    {
+      "epoch": 0.6511740232938517,
+      "grad_norm": 1.625,
+      "learning_rate": 1.3641762503901994e-05,
+      "loss": 0.3521,
+      "step": 14830
+    },
+    {
+      "epoch": 0.6512618417730551,
+      "grad_norm": 1.8671875,
+      "learning_rate": 1.3635599898477738e-05,
+      "loss": 0.348,
+      "step": 14832
+    },
+    {
+      "epoch": 0.6513496602522586,
+      "grad_norm": 1.8046875,
+      "learning_rate": 1.3629438163341978e-05,
+      "loss": 0.3248,
+      "step": 14834
+    },
+    {
+      "epoch": 0.6514374787314621,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.3623277298966558e-05,
+      "loss": 0.3101,
+      "step": 14836
+    },
+    {
+      "epoch": 0.6515252972106655,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.36171173058233e-05,
+      "loss": 0.3505,
+      "step": 14838
+    },
+    {
+      "epoch": 0.651613115689869,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.3610958184383928e-05,
+      "loss": 0.2986,
+      "step": 14840
+    },
+    {
+      "epoch": 0.6517009341690725,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.3604799935120099e-05,
+      "loss": 0.307,
+      "step": 14842
+    },
+    {
+      "epoch": 0.651788752648276,
+      "grad_norm": 1.625,
+      "learning_rate": 1.3598642558503432e-05,
+      "loss": 0.3515,
+      "step": 14844
+    },
+    {
+      "epoch": 0.6518765711274795,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.3592486055005441e-05,
+      "loss": 0.3184,
+      "step": 14846
+    },
+    {
+      "epoch": 0.651964389606683,
+      "grad_norm": 1.7890625,
+      "learning_rate": 1.3586330425097621e-05,
+      "loss": 0.3355,
+      "step": 14848
+    },
+    {
+      "epoch": 0.6520522080858865,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.3580175669251336e-05,
+      "loss": 0.3188,
+      "step": 14850
+    },
+    {
+      "epoch": 0.65214002656509,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.3574021787937944e-05,
+      "loss": 0.3224,
+      "step": 14852
+    },
+    {
+      "epoch": 0.6522278450442934,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.35678687816287e-05,
+      "loss": 0.3439,
+      "step": 14854
+    },
+    {
+      "epoch": 0.6523156635234969,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.3561716650794798e-05,
+      "loss": 0.356,
+      "step": 14856
+    },
+    {
+      "epoch": 0.6524034820027004,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.3555565395907388e-05,
+      "loss": 0.3344,
+      "step": 14858
+    },
+    {
+      "epoch": 0.652491300481904,
+      "grad_norm": 1.625,
+      "learning_rate": 1.3549415017437512e-05,
+      "loss": 0.3208,
+      "step": 14860
+    },
+    {
+      "epoch": 0.6525791189611074,
+      "grad_norm": 1.625,
+      "learning_rate": 1.3543265515856191e-05,
+      "loss": 0.3234,
+      "step": 14862
+    },
+    {
+      "epoch": 0.6526669374403109,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.3537116891634338e-05,
+      "loss": 0.3535,
+      "step": 14864
+    },
+    {
+      "epoch": 0.6527547559195144,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.3530969145242816e-05,
+      "loss": 0.3412,
+      "step": 14866
+    },
+    {
+      "epoch": 0.6528425743987178,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.3524822277152433e-05,
+      "loss": 0.328,
+      "step": 14868
+    },
+    {
+      "epoch": 0.6529303928779213,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.3518676287833904e-05,
+      "loss": 0.3258,
+      "step": 14870
+    },
+    {
+      "epoch": 0.6530182113571248,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.3512531177757904e-05,
+      "loss": 0.3029,
+      "step": 14872
+    },
+    {
+      "epoch": 0.6531060298363283,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.3506386947395022e-05,
+      "loss": 0.3254,
+      "step": 14874
+    },
+    {
+      "epoch": 0.6531938483155318,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.3500243597215773e-05,
+      "loss": 0.3469,
+      "step": 14876
+    },
+    {
+      "epoch": 0.6532816667947353,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.3494101127690633e-05,
+      "loss": 0.3451,
+      "step": 14878
+    },
+    {
+      "epoch": 0.6533694852739388,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.348795953928999e-05,
+      "loss": 0.3097,
+      "step": 14880
+    },
+    {
+      "epoch": 0.6534573037531423,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.3481818832484163e-05,
+      "loss": 0.3298,
+      "step": 14882
+    },
+    {
+      "epoch": 0.6535451222323457,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.3475679007743402e-05,
+      "loss": 0.3281,
+      "step": 14884
+    },
+    {
+      "epoch": 0.6536329407115492,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.3469540065537917e-05,
+      "loss": 0.3299,
+      "step": 14886
+    },
+    {
+      "epoch": 0.6537207591907527,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.3463402006337817e-05,
+      "loss": 0.3511,
+      "step": 14888
+    },
+    {
+      "epoch": 0.6538085776699561,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.3457264830613141e-05,
+      "loss": 0.3132,
+      "step": 14890
+    },
+    {
+      "epoch": 0.6538963961491597,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.3451128538833906e-05,
+      "loss": 0.3346,
+      "step": 14892
+    },
+    {
+      "epoch": 0.6539842146283632,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.3444993131470006e-05,
+      "loss": 0.3232,
+      "step": 14894
+    },
+    {
+      "epoch": 0.6540720331075667,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.3438858608991315e-05,
+      "loss": 0.3156,
+      "step": 14896
+    },
+    {
+      "epoch": 0.6541598515867701,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.3432724971867599e-05,
+      "loss": 0.3412,
+      "step": 14898
+    },
+    {
+      "epoch": 0.6542476700659736,
+      "grad_norm": 1.7578125,
+      "learning_rate": 1.3426592220568568e-05,
+      "loss": 0.3014,
+      "step": 14900
+    },
+    {
+      "epoch": 0.6543354885451771,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.3420460355563891e-05,
+      "loss": 0.3055,
+      "step": 14902
+    },
+    {
+      "epoch": 0.6544233070243806,
+      "grad_norm": 1.515625,
+      "learning_rate": 1.3414329377323126e-05,
+      "loss": 0.3393,
+      "step": 14904
+    },
+    {
+      "epoch": 0.6545111255035841,
+      "grad_norm": 1.78125,
+      "learning_rate": 1.3408199286315803e-05,
+      "loss": 0.344,
+      "step": 14906
+    },
+    {
+      "epoch": 0.6545989439827876,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.3402070083011348e-05,
+      "loss": 0.3015,
+      "step": 14908
+    },
+    {
+      "epoch": 0.6546867624619911,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.3395941767879164e-05,
+      "loss": 0.3608,
+      "step": 14910
+    },
+    {
+      "epoch": 0.6547745809411946,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.3389814341388523e-05,
+      "loss": 0.3229,
+      "step": 14912
+    },
+    {
+      "epoch": 0.654862399420398,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.3383687804008693e-05,
+      "loss": 0.3233,
+      "step": 14914
+    },
+    {
+      "epoch": 0.6549502178996015,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.3377562156208833e-05,
+      "loss": 0.3179,
+      "step": 14916
+    },
+    {
+      "epoch": 0.655038036378805,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.3371437398458042e-05,
+      "loss": 0.3171,
+      "step": 14918
+    },
+    {
+      "epoch": 0.6551258548580084,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.3365313531225374e-05,
+      "loss": 0.3321,
+      "step": 14920
+    },
+    {
+      "epoch": 0.655213673337212,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.3359190554979772e-05,
+      "loss": 0.3269,
+      "step": 14922
+    },
+    {
+      "epoch": 0.6553014918164155,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.3353068470190161e-05,
+      "loss": 0.3241,
+      "step": 14924
+    },
+    {
+      "epoch": 0.655389310295619,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.3346947277325356e-05,
+      "loss": 0.2946,
+      "step": 14926
+    },
+    {
+      "epoch": 0.6554771287748224,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.334082697685411e-05,
+      "loss": 0.3364,
+      "step": 14928
+    },
+    {
+      "epoch": 0.6555649472540259,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.3334707569245142e-05,
+      "loss": 0.35,
+      "step": 14930
+    },
+    {
+      "epoch": 0.6556527657332294,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.3328589054967056e-05,
+      "loss": 0.3134,
+      "step": 14932
+    },
+    {
+      "epoch": 0.6557405842124329,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.3322471434488424e-05,
+      "loss": 0.3198,
+      "step": 14934
+    },
+    {
+      "epoch": 0.6558284026916363,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.3316354708277728e-05,
+      "loss": 0.3519,
+      "step": 14936
+    },
+    {
+      "epoch": 0.6559162211708399,
+      "grad_norm": 1.765625,
+      "learning_rate": 1.3310238876803383e-05,
+      "loss": 0.3331,
+      "step": 14938
+    },
+    {
+      "epoch": 0.6560040396500434,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.3304123940533767e-05,
+      "loss": 0.339,
+      "step": 14940
+    },
+    {
+      "epoch": 0.6560918581292469,
+      "grad_norm": 1.625,
+      "learning_rate": 1.3298009899937123e-05,
+      "loss": 0.3369,
+      "step": 14942
+    },
+    {
+      "epoch": 0.6561796766084503,
+      "grad_norm": 1.8046875,
+      "learning_rate": 1.3291896755481694e-05,
+      "loss": 0.3124,
+      "step": 14944
+    },
+    {
+      "epoch": 0.6562674950876538,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.3285784507635609e-05,
+      "loss": 0.3247,
+      "step": 14946
+    },
+    {
+      "epoch": 0.6563553135668573,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.3279673156866967e-05,
+      "loss": 0.3425,
+      "step": 14948
+    },
+    {
+      "epoch": 0.6564431320460608,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.327356270364376e-05,
+      "loss": 0.3239,
+      "step": 14950
+    },
+    {
+      "epoch": 0.6565309505252643,
+      "grad_norm": 1.7734375,
+      "learning_rate": 1.3267453148433926e-05,
+      "loss": 0.322,
+      "step": 14952
+    },
+    {
+      "epoch": 0.6566187690044678,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.326134449170535e-05,
+      "loss": 0.3282,
+      "step": 14954
+    },
+    {
+      "epoch": 0.6567065874836713,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.3255236733925819e-05,
+      "loss": 0.3349,
+      "step": 14956
+    },
+    {
+      "epoch": 0.6567944059628747,
+      "grad_norm": 1.75,
+      "learning_rate": 1.3249129875563083e-05,
+      "loss": 0.3605,
+      "step": 14958
+    },
+    {
+      "epoch": 0.6568822244420782,
+      "grad_norm": 1.484375,
+      "learning_rate": 1.3243023917084796e-05,
+      "loss": 0.3174,
+      "step": 14960
+    },
+    {
+      "epoch": 0.6569700429212817,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.3236918858958547e-05,
+      "loss": 0.3258,
+      "step": 14962
+    },
+    {
+      "epoch": 0.6570578614004852,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.3230814701651884e-05,
+      "loss": 0.2999,
+      "step": 14964
+    },
+    {
+      "epoch": 0.6571456798796886,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.3224711445632237e-05,
+      "loss": 0.2968,
+      "step": 14966
+    },
+    {
+      "epoch": 0.6572334983588922,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.3218609091367024e-05,
+      "loss": 0.3169,
+      "step": 14968
+    },
+    {
+      "epoch": 0.6573213168380957,
+      "grad_norm": 1.625,
+      "learning_rate": 1.321250763932354e-05,
+      "loss": 0.3351,
+      "step": 14970
+    },
+    {
+      "epoch": 0.6574091353172992,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.320640708996907e-05,
+      "loss": 0.3572,
+      "step": 14972
+    },
+    {
+      "epoch": 0.6574969537965026,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.3200307443770748e-05,
+      "loss": 0.3086,
+      "step": 14974
+    },
+    {
+      "epoch": 0.6575847722757061,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.319420870119572e-05,
+      "loss": 0.3494,
+      "step": 14976
+    },
+    {
+      "epoch": 0.6576725907549096,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.3188110862711023e-05,
+      "loss": 0.3088,
+      "step": 14978
+    },
+    {
+      "epoch": 0.657760409234113,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.3182013928783618e-05,
+      "loss": 0.3322,
+      "step": 14980
+    },
+    {
+      "epoch": 0.6578482277133165,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.3175917899880427e-05,
+      "loss": 0.3124,
+      "step": 14982
+    },
+    {
+      "epoch": 0.6579360461925201,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.3169822776468268e-05,
+      "loss": 0.3098,
+      "step": 14984
+    },
+    {
+      "epoch": 0.6580238646717236,
+      "grad_norm": 1.5,
+      "learning_rate": 1.3163728559013928e-05,
+      "loss": 0.2701,
+      "step": 14986
+    },
+    {
+      "epoch": 0.658111683150927,
+      "grad_norm": 1.625,
+      "learning_rate": 1.3157635247984091e-05,
+      "loss": 0.3256,
+      "step": 14988
+    },
+    {
+      "epoch": 0.6581995016301305,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.3151542843845377e-05,
+      "loss": 0.3451,
+      "step": 14990
+    },
+    {
+      "epoch": 0.658287320109334,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.3145451347064358e-05,
+      "loss": 0.3386,
+      "step": 14992
+    },
+    {
+      "epoch": 0.6583751385885375,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.313936075810751e-05,
+      "loss": 0.3218,
+      "step": 14994
+    },
+    {
+      "epoch": 0.6584629570677409,
+      "grad_norm": 1.5078125,
+      "learning_rate": 1.313327107744127e-05,
+      "loss": 0.3355,
+      "step": 14996
+    },
+    {
+      "epoch": 0.6585507755469444,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.3127182305531971e-05,
+      "loss": 0.3392,
+      "step": 14998
+    },
+    {
+      "epoch": 0.658638594026148,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.3121094442845893e-05,
+      "loss": 0.3311,
+      "step": 15000
+    },
+    {
+      "epoch": 0.6587264125053515,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.3115007489849265e-05,
+      "loss": 0.3168,
+      "step": 15002
+    },
+    {
+      "epoch": 0.6588142309845549,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.3108921447008194e-05,
+      "loss": 0.3516,
+      "step": 15004
+    },
+    {
+      "epoch": 0.6589020494637584,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.310283631478878e-05,
+      "loss": 0.3232,
+      "step": 15006
+    },
+    {
+      "epoch": 0.6589898679429619,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.3096752093657002e-05,
+      "loss": 0.3072,
+      "step": 15008
+    },
+    {
+      "epoch": 0.6590776864221654,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.3090668784078813e-05,
+      "loss": 0.3238,
+      "step": 15010
+    },
+    {
+      "epoch": 0.6591655049013688,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.3084586386520062e-05,
+      "loss": 0.3492,
+      "step": 15012
+    },
+    {
+      "epoch": 0.6592533233805724,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.3078504901446533e-05,
+      "loss": 0.3287,
+      "step": 15014
+    },
+    {
+      "epoch": 0.6593411418597759,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.3072424329323968e-05,
+      "loss": 0.3733,
+      "step": 15016
+    },
+    {
+      "epoch": 0.6594289603389794,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.3066344670617991e-05,
+      "loss": 0.3267,
+      "step": 15018
+    },
+    {
+      "epoch": 0.6595167788181828,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.3060265925794218e-05,
+      "loss": 0.326,
+      "step": 15020
+    },
+    {
+      "epoch": 0.6596045972973863,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.3054188095318137e-05,
+      "loss": 0.2889,
+      "step": 15022
+    },
+    {
+      "epoch": 0.6596924157765898,
+      "grad_norm": 1.5,
+      "learning_rate": 1.3048111179655186e-05,
+      "loss": 0.2946,
+      "step": 15024
+    },
+    {
+      "epoch": 0.6597802342557932,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.3042035179270756e-05,
+      "loss": 0.3259,
+      "step": 15026
+    },
+    {
+      "epoch": 0.6598680527349967,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.3035960094630132e-05,
+      "loss": 0.3238,
+      "step": 15028
+    },
+    {
+      "epoch": 0.6599558712142003,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.302988592619856e-05,
+      "loss": 0.3381,
+      "step": 15030
+    },
+    {
+      "epoch": 0.6600436896934038,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.3023812674441189e-05,
+      "loss": 0.3074,
+      "step": 15032
+    },
+    {
+      "epoch": 0.6601315081726072,
+      "grad_norm": 1.5,
+      "learning_rate": 1.301774033982312e-05,
+      "loss": 0.3234,
+      "step": 15034
+    },
+    {
+      "epoch": 0.6602193266518107,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.3011668922809355e-05,
+      "loss": 0.3027,
+      "step": 15036
+    },
+    {
+      "epoch": 0.6603071451310142,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.3005598423864868e-05,
+      "loss": 0.3438,
+      "step": 15038
+    },
+    {
+      "epoch": 0.6603949636102177,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.2999528843454528e-05,
+      "loss": 0.3359,
+      "step": 15040
+    },
+    {
+      "epoch": 0.6604827820894211,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.2993460182043138e-05,
+      "loss": 0.3283,
+      "step": 15042
+    },
+    {
+      "epoch": 0.6605706005686246,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.2987392440095455e-05,
+      "loss": 0.3231,
+      "step": 15044
+    },
+    {
+      "epoch": 0.6606584190478282,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.2981325618076129e-05,
+      "loss": 0.302,
+      "step": 15046
+    },
+    {
+      "epoch": 0.6607462375270317,
+      "grad_norm": 1.625,
+      "learning_rate": 1.2975259716449778e-05,
+      "loss": 0.3357,
+      "step": 15048
+    },
+    {
+      "epoch": 0.6608340560062351,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.2969194735680917e-05,
+      "loss": 0.3215,
+      "step": 15050
+    },
+    {
+      "epoch": 0.6609218744854386,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.2963130676234003e-05,
+      "loss": 0.3265,
+      "step": 15052
+    },
+    {
+      "epoch": 0.6610096929646421,
+      "grad_norm": 1.75,
+      "learning_rate": 1.2957067538573434e-05,
+      "loss": 0.3357,
+      "step": 15054
+    },
+    {
+      "epoch": 0.6610975114438455,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.2951005323163509e-05,
+      "loss": 0.3197,
+      "step": 15056
+    },
+    {
+      "epoch": 0.661185329923049,
+      "grad_norm": 1.4765625,
+      "learning_rate": 1.2944944030468498e-05,
+      "loss": 0.3311,
+      "step": 15058
+    },
+    {
+      "epoch": 0.6612731484022526,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.2938883660952558e-05,
+      "loss": 0.3386,
+      "step": 15060
+    },
+    {
+      "epoch": 0.6613609668814561,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.2932824215079792e-05,
+      "loss": 0.2935,
+      "step": 15062
+    },
+    {
+      "epoch": 0.6614487853606595,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.2926765693314257e-05,
+      "loss": 0.324,
+      "step": 15064
+    },
+    {
+      "epoch": 0.661536603839863,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.2920708096119883e-05,
+      "loss": 0.3579,
+      "step": 15066
+    },
+    {
+      "epoch": 0.6616244223190665,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.291465142396059e-05,
+      "loss": 0.3081,
+      "step": 15068
+    },
+    {
+      "epoch": 0.66171224079827,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.2908595677300172e-05,
+      "loss": 0.3196,
+      "step": 15070
+    },
+    {
+      "epoch": 0.6618000592774734,
+      "grad_norm": 1.625,
+      "learning_rate": 1.2902540856602414e-05,
+      "loss": 0.319,
+      "step": 15072
+    },
+    {
+      "epoch": 0.6618878777566769,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.289648696233097e-05,
+      "loss": 0.3505,
+      "step": 15074
+    },
+    {
+      "epoch": 0.6619756962358805,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.2890433994949447e-05,
+      "loss": 0.3762,
+      "step": 15076
+    },
+    {
+      "epoch": 0.662063514715084,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.2884381954921404e-05,
+      "loss": 0.3002,
+      "step": 15078
+    },
+    {
+      "epoch": 0.6621513331942874,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.2878330842710284e-05,
+      "loss": 0.317,
+      "step": 15080
+    },
+    {
+      "epoch": 0.6622391516734909,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.2872280658779501e-05,
+      "loss": 0.329,
+      "step": 15082
+    },
+    {
+      "epoch": 0.6623269701526944,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.2866231403592377e-05,
+      "loss": 0.2896,
+      "step": 15084
+    },
+    {
+      "epoch": 0.6624147886318978,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.2860183077612148e-05,
+      "loss": 0.3293,
+      "step": 15086
+    },
+    {
+      "epoch": 0.6625026071111013,
+      "grad_norm": 1.8046875,
+      "learning_rate": 1.285413568130202e-05,
+      "loss": 0.3361,
+      "step": 15088
+    },
+    {
+      "epoch": 0.6625904255903048,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.2848089215125084e-05,
+      "loss": 0.3096,
+      "step": 15090
+    },
+    {
+      "epoch": 0.6626782440695084,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.2842043679544397e-05,
+      "loss": 0.3276,
+      "step": 15092
+    },
+    {
+      "epoch": 0.6627660625487118,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.283599907502292e-05,
+      "loss": 0.3178,
+      "step": 15094
+    },
+    {
+      "epoch": 0.6628538810279153,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.2829955402023549e-05,
+      "loss": 0.3348,
+      "step": 15096
+    },
+    {
+      "epoch": 0.6629416995071188,
+      "grad_norm": 1.515625,
+      "learning_rate": 1.2823912661009102e-05,
+      "loss": 0.2979,
+      "step": 15098
+    },
+    {
+      "epoch": 0.6630295179863223,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.2817870852442355e-05,
+      "loss": 0.3386,
+      "step": 15100
+    },
+    {
+      "epoch": 0.6631173364655257,
+      "grad_norm": 1.625,
+      "learning_rate": 1.2811829976785971e-05,
+      "loss": 0.314,
+      "step": 15102
+    },
+    {
+      "epoch": 0.6632051549447292,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.2805790034502565e-05,
+      "loss": 0.3095,
+      "step": 15104
+    },
+    {
+      "epoch": 0.6632929734239328,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.2799751026054691e-05,
+      "loss": 0.3297,
+      "step": 15106
+    },
+    {
+      "epoch": 0.6633807919031363,
+      "grad_norm": 1.4921875,
+      "learning_rate": 1.2793712951904796e-05,
+      "loss": 0.3125,
+      "step": 15108
+    },
+    {
+      "epoch": 0.6634686103823397,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.2787675812515299e-05,
+      "loss": 0.3212,
+      "step": 15110
+    },
+    {
+      "epoch": 0.6635564288615432,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.2781639608348517e-05,
+      "loss": 0.3171,
+      "step": 15112
+    },
+    {
+      "epoch": 0.6636442473407467,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.2775604339866692e-05,
+      "loss": 0.3389,
+      "step": 15114
+    },
+    {
+      "epoch": 0.6637320658199501,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.2769570007532027e-05,
+      "loss": 0.3024,
+      "step": 15116
+    },
+    {
+      "epoch": 0.6638198842991536,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.2763536611806615e-05,
+      "loss": 0.3248,
+      "step": 15118
+    },
+    {
+      "epoch": 0.6639077027783571,
+      "grad_norm": 1.8203125,
+      "learning_rate": 1.275750415315251e-05,
+      "loss": 0.3536,
+      "step": 15120
+    },
+    {
+      "epoch": 0.6639955212575607,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.2751472632031672e-05,
+      "loss": 0.3393,
+      "step": 15122
+    },
+    {
+      "epoch": 0.6640833397367641,
+      "grad_norm": 1.4765625,
+      "learning_rate": 1.2745442048905998e-05,
+      "loss": 0.3293,
+      "step": 15124
+    },
+    {
+      "epoch": 0.6641711582159676,
+      "grad_norm": 1.796875,
+      "learning_rate": 1.2739412404237306e-05,
+      "loss": 0.3239,
+      "step": 15126
+    },
+    {
+      "epoch": 0.6642589766951711,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.2733383698487344e-05,
+      "loss": 0.3344,
+      "step": 15128
+    },
+    {
+      "epoch": 0.6643467951743746,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.2727355932117806e-05,
+      "loss": 0.3109,
+      "step": 15130
+    },
+    {
+      "epoch": 0.664434613653578,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.2721329105590284e-05,
+      "loss": 0.3262,
+      "step": 15132
+    },
+    {
+      "epoch": 0.6645224321327815,
+      "grad_norm": 1.75,
+      "learning_rate": 1.2715303219366337e-05,
+      "loss": 0.3469,
+      "step": 15134
+    },
+    {
+      "epoch": 0.664610250611985,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.2709278273907408e-05,
+      "loss": 0.3071,
+      "step": 15136
+    },
+    {
+      "epoch": 0.6646980690911886,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.2703254269674885e-05,
+      "loss": 0.3076,
+      "step": 15138
+    },
+    {
+      "epoch": 0.664785887570392,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.269723120713011e-05,
+      "loss": 0.3035,
+      "step": 15140
+    },
+    {
+      "epoch": 0.6648737060495955,
+      "grad_norm": 1.7890625,
+      "learning_rate": 1.2691209086734313e-05,
+      "loss": 0.3316,
+      "step": 15142
+    },
+    {
+      "epoch": 0.664961524528799,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.2685187908948678e-05,
+      "loss": 0.3274,
+      "step": 15144
+    },
+    {
+      "epoch": 0.6650493430080024,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.2679167674234308e-05,
+      "loss": 0.305,
+      "step": 15146
+    },
+    {
+      "epoch": 0.6651371614872059,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.267314838305222e-05,
+      "loss": 0.3264,
+      "step": 15148
+    },
+    {
+      "epoch": 0.6652249799664094,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.2667130035863395e-05,
+      "loss": 0.3286,
+      "step": 15150
+    },
+    {
+      "epoch": 0.6653127984456129,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.2661112633128696e-05,
+      "loss": 0.3141,
+      "step": 15152
+    },
+    {
+      "epoch": 0.6654006169248164,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.2655096175308962e-05,
+      "loss": 0.3165,
+      "step": 15154
+    },
+    {
+      "epoch": 0.6654884354040199,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.264908066286492e-05,
+      "loss": 0.3351,
+      "step": 15156
+    },
+    {
+      "epoch": 0.6655762538832234,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.2643066096257244e-05,
+      "loss": 0.3032,
+      "step": 15158
+    },
+    {
+      "epoch": 0.6656640723624269,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.2637052475946526e-05,
+      "loss": 0.303,
+      "step": 15160
+    },
+    {
+      "epoch": 0.6657518908416303,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.2631039802393286e-05,
+      "loss": 0.3241,
+      "step": 15162
+    },
+    {
+      "epoch": 0.6658397093208338,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.2625028076057987e-05,
+      "loss": 0.3363,
+      "step": 15164
+    },
+    {
+      "epoch": 0.6659275278000373,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.2619017297400998e-05,
+      "loss": 0.3022,
+      "step": 15166
+    },
+    {
+      "epoch": 0.6660153462792409,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.2613007466882643e-05,
+      "loss": 0.3297,
+      "step": 15168
+    },
+    {
+      "epoch": 0.6661031647584443,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.2606998584963136e-05,
+      "loss": 0.3167,
+      "step": 15170
+    },
+    {
+      "epoch": 0.6661909832376478,
+      "grad_norm": 1.75,
+      "learning_rate": 1.2600990652102656e-05,
+      "loss": 0.3309,
+      "step": 15172
+    },
+    {
+      "epoch": 0.6662788017168513,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.2594983668761286e-05,
+      "loss": 0.3045,
+      "step": 15174
+    },
+    {
+      "epoch": 0.6663666201960547,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.2588977635399029e-05,
+      "loss": 0.3155,
+      "step": 15176
+    },
+    {
+      "epoch": 0.6664544386752582,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.2582972552475852e-05,
+      "loss": 0.3326,
+      "step": 15178
+    },
+    {
+      "epoch": 0.6665422571544617,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.2576968420451601e-05,
+      "loss": 0.3442,
+      "step": 15180
+    },
+    {
+      "epoch": 0.6666300756336652,
+      "grad_norm": 1.7578125,
+      "learning_rate": 1.2570965239786098e-05,
+      "loss": 0.313,
+      "step": 15182
+    },
+    {
+      "epoch": 0.6667178941128687,
+      "grad_norm": 1.75,
+      "learning_rate": 1.2564963010939057e-05,
+      "loss": 0.3379,
+      "step": 15184
+    },
+    {
+      "epoch": 0.6668057125920722,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.2558961734370128e-05,
+      "loss": 0.3332,
+      "step": 15186
+    },
+    {
+      "epoch": 0.6668935310712757,
+      "grad_norm": 1.7890625,
+      "learning_rate": 1.2552961410538894e-05,
+      "loss": 0.3414,
+      "step": 15188
+    },
+    {
+      "epoch": 0.6669813495504792,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.2546962039904847e-05,
+      "loss": 0.334,
+      "step": 15190
+    },
+    {
+      "epoch": 0.6670691680296826,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.254096362292744e-05,
+      "loss": 0.3496,
+      "step": 15192
+    },
+    {
+      "epoch": 0.6671569865088861,
+      "grad_norm": 1.4921875,
+      "learning_rate": 1.253496616006602e-05,
+      "loss": 0.3553,
+      "step": 15194
+    },
+    {
+      "epoch": 0.6672448049880896,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.2528969651779888e-05,
+      "loss": 0.3289,
+      "step": 15196
+    },
+    {
+      "epoch": 0.667332623467293,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.2522974098528245e-05,
+      "loss": 0.2986,
+      "step": 15198
+    },
+    {
+      "epoch": 0.6674204419464966,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.2516979500770232e-05,
+      "loss": 0.3219,
+      "step": 15200
+    },
+    {
+      "epoch": 0.6675082604257001,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.251098585896493e-05,
+      "loss": 0.3009,
+      "step": 15202
+    },
+    {
+      "epoch": 0.6675960789049036,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.250499317357131e-05,
+      "loss": 0.3442,
+      "step": 15204
+    },
+    {
+      "epoch": 0.667683897384107,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.2499001445048325e-05,
+      "loss": 0.3178,
+      "step": 15206
+    },
+    {
+      "epoch": 0.6677717158633105,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.2493010673854803e-05,
+      "loss": 0.3308,
+      "step": 15208
+    },
+    {
+      "epoch": 0.667859534342514,
+      "grad_norm": 1.421875,
+      "learning_rate": 1.2487020860449511e-05,
+      "loss": 0.3033,
+      "step": 15210
+    },
+    {
+      "epoch": 0.6679473528217175,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.2481032005291173e-05,
+      "loss": 0.3283,
+      "step": 15212
+    },
+    {
+      "epoch": 0.668035171300921,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.2475044108838393e-05,
+      "loss": 0.2983,
+      "step": 15214
+    },
+    {
+      "epoch": 0.6681229897801245,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.2469057171549759e-05,
+      "loss": 0.3213,
+      "step": 15216
+    },
+    {
+      "epoch": 0.668210808259328,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.246307119388371e-05,
+      "loss": 0.3042,
+      "step": 15218
+    },
+    {
+      "epoch": 0.6682986267385315,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.2457086176298685e-05,
+      "loss": 0.3172,
+      "step": 15220
+    },
+    {
+      "epoch": 0.6683864452177349,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.2451102119253009e-05,
+      "loss": 0.3245,
+      "step": 15222
+    },
+    {
+      "epoch": 0.6684742636969384,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.2445119023204926e-05,
+      "loss": 0.32,
+      "step": 15224
+    },
+    {
+      "epoch": 0.6685620821761419,
+      "grad_norm": 1.7734375,
+      "learning_rate": 1.2439136888612652e-05,
+      "loss": 0.3245,
+      "step": 15226
+    },
+    {
+      "epoch": 0.6686499006553454,
+      "grad_norm": 1.625,
+      "learning_rate": 1.2433155715934275e-05,
+      "loss": 0.3037,
+      "step": 15228
+    },
+    {
+      "epoch": 0.6687377191345489,
+      "grad_norm": 1.5078125,
+      "learning_rate": 1.2427175505627856e-05,
+      "loss": 0.3555,
+      "step": 15230
+    },
+    {
+      "epoch": 0.6688255376137524,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.2421196258151337e-05,
+      "loss": 0.3333,
+      "step": 15232
+    },
+    {
+      "epoch": 0.6689133560929559,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.241521797396264e-05,
+      "loss": 0.3125,
+      "step": 15234
+    },
+    {
+      "epoch": 0.6690011745721594,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.2409240653519564e-05,
+      "loss": 0.3263,
+      "step": 15236
+    },
+    {
+      "epoch": 0.6690889930513628,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.2403264297279849e-05,
+      "loss": 0.3191,
+      "step": 15238
+    },
+    {
+      "epoch": 0.6691768115305663,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.239728890570118e-05,
+      "loss": 0.3162,
+      "step": 15240
+    },
+    {
+      "epoch": 0.6692646300097698,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.239131447924114e-05,
+      "loss": 0.3191,
+      "step": 15242
+    },
+    {
+      "epoch": 0.6693524484889732,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.238534101835727e-05,
+      "loss": 0.3004,
+      "step": 15244
+    },
+    {
+      "epoch": 0.6694402669681768,
+      "grad_norm": 1.859375,
+      "learning_rate": 1.2379368523507007e-05,
+      "loss": 0.3416,
+      "step": 15246
+    },
+    {
+      "epoch": 0.6695280854473803,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.2373396995147729e-05,
+      "loss": 0.3377,
+      "step": 15248
+    },
+    {
+      "epoch": 0.6696159039265838,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.2367426433736737e-05,
+      "loss": 0.3255,
+      "step": 15250
+    },
+    {
+      "epoch": 0.6697037224057872,
+      "grad_norm": 1.7890625,
+      "learning_rate": 1.2361456839731245e-05,
+      "loss": 0.3023,
+      "step": 15252
+    },
+    {
+      "epoch": 0.6697915408849907,
+      "grad_norm": 1.75,
+      "learning_rate": 1.2355488213588429e-05,
+      "loss": 0.3236,
+      "step": 15254
+    },
+    {
+      "epoch": 0.6698793593641942,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.2349520555765348e-05,
+      "loss": 0.3063,
+      "step": 15256
+    },
+    {
+      "epoch": 0.6699671778433977,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.2343553866719024e-05,
+      "loss": 0.3225,
+      "step": 15258
+    },
+    {
+      "epoch": 0.6700549963226012,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.2337588146906378e-05,
+      "loss": 0.2899,
+      "step": 15260
+    },
+    {
+      "epoch": 0.6701428148018047,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.2331623396784258e-05,
+      "loss": 0.3267,
+      "step": 15262
+    },
+    {
+      "epoch": 0.6702306332810082,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.2325659616809466e-05,
+      "loss": 0.299,
+      "step": 15264
+    },
+    {
+      "epoch": 0.6703184517602117,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.2319696807438686e-05,
+      "loss": 0.3019,
+      "step": 15266
+    },
+    {
+      "epoch": 0.6704062702394151,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.2313734969128576e-05,
+      "loss": 0.2836,
+      "step": 15268
+    },
+    {
+      "epoch": 0.6704940887186186,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.2307774102335685e-05,
+      "loss": 0.3468,
+      "step": 15270
+    },
+    {
+      "epoch": 0.6705819071978221,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.2301814207516482e-05,
+      "loss": 0.3267,
+      "step": 15272
+    },
+    {
+      "epoch": 0.6706697256770255,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.2295855285127403e-05,
+      "loss": 0.3108,
+      "step": 15274
+    },
+    {
+      "epoch": 0.6707575441562291,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.2289897335624761e-05,
+      "loss": 0.3076,
+      "step": 15276
+    },
+    {
+      "epoch": 0.6708453626354326,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.2283940359464849e-05,
+      "loss": 0.3156,
+      "step": 15278
+    },
+    {
+      "epoch": 0.6709331811146361,
+      "grad_norm": 1.625,
+      "learning_rate": 1.2277984357103811e-05,
+      "loss": 0.3288,
+      "step": 15280
+    },
+    {
+      "epoch": 0.6710209995938395,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.2272029328997791e-05,
+      "loss": 0.3448,
+      "step": 15282
+    },
+    {
+      "epoch": 0.671108818073043,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.2266075275602818e-05,
+      "loss": 0.3192,
+      "step": 15284
+    },
+    {
+      "epoch": 0.6711966365522465,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.226012219737484e-05,
+      "loss": 0.3315,
+      "step": 15286
+    },
+    {
+      "epoch": 0.67128445503145,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.2254170094769771e-05,
+      "loss": 0.3169,
+      "step": 15288
+    },
+    {
+      "epoch": 0.6713722735106534,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.22482189682434e-05,
+      "loss": 0.3263,
+      "step": 15290
+    },
+    {
+      "epoch": 0.671460091989857,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.2242268818251486e-05,
+      "loss": 0.3187,
+      "step": 15292
+    },
+    {
+      "epoch": 0.6715479104690605,
+      "grad_norm": 1.84375,
+      "learning_rate": 1.2236319645249677e-05,
+      "loss": 0.3476,
+      "step": 15294
+    },
+    {
+      "epoch": 0.671635728948264,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.2230371449693578e-05,
+      "loss": 0.3365,
+      "step": 15296
+    },
+    {
+      "epoch": 0.6717235474274674,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.2224424232038692e-05,
+      "loss": 0.299,
+      "step": 15298
+    },
+    {
+      "epoch": 0.6718113659066709,
+      "grad_norm": 1.625,
+      "learning_rate": 1.2218477992740455e-05,
+      "loss": 0.3148,
+      "step": 15300
+    },
+    {
+      "epoch": 0.6718991843858744,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.2212532732254245e-05,
+      "loss": 0.3289,
+      "step": 15302
+    },
+    {
+      "epoch": 0.6719870028650778,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.2206588451035334e-05,
+      "loss": 0.3381,
+      "step": 15304
+    },
+    {
+      "epoch": 0.6720748213442814,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.2200645149538955e-05,
+      "loss": 0.3261,
+      "step": 15306
+    },
+    {
+      "epoch": 0.6721626398234849,
+      "grad_norm": 1.8125,
+      "learning_rate": 1.2194702828220241e-05,
+      "loss": 0.3341,
+      "step": 15308
+    },
+    {
+      "epoch": 0.6722504583026884,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.2188761487534254e-05,
+      "loss": 0.3285,
+      "step": 15310
+    },
+    {
+      "epoch": 0.6723382767818918,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.218282112793598e-05,
+      "loss": 0.3094,
+      "step": 15312
+    },
+    {
+      "epoch": 0.6724260952610953,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.2176881749880328e-05,
+      "loss": 0.3299,
+      "step": 15314
+    },
+    {
+      "epoch": 0.6725139137402988,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.2170943353822155e-05,
+      "loss": 0.3249,
+      "step": 15316
+    },
+    {
+      "epoch": 0.6726017322195023,
+      "grad_norm": 1.625,
+      "learning_rate": 1.2165005940216209e-05,
+      "loss": 0.3287,
+      "step": 15318
+    },
+    {
+      "epoch": 0.6726895506987057,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.2159069509517193e-05,
+      "loss": 0.3378,
+      "step": 15320
+    },
+    {
+      "epoch": 0.6727773691779093,
+      "grad_norm": 1.515625,
+      "learning_rate": 1.2153134062179711e-05,
+      "loss": 0.3203,
+      "step": 15322
+    },
+    {
+      "epoch": 0.6728651876571128,
+      "grad_norm": 1.4921875,
+      "learning_rate": 1.2147199598658293e-05,
+      "loss": 0.3337,
+      "step": 15324
+    },
+    {
+      "epoch": 0.6729530061363163,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.2141266119407421e-05,
+      "loss": 0.3331,
+      "step": 15326
+    },
+    {
+      "epoch": 0.6730408246155197,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.2135333624881463e-05,
+      "loss": 0.3097,
+      "step": 15328
+    },
+    {
+      "epoch": 0.6731286430947232,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.2129402115534747e-05,
+      "loss": 0.3504,
+      "step": 15330
+    },
+    {
+      "epoch": 0.6732164615739267,
+      "grad_norm": 1.75,
+      "learning_rate": 1.2123471591821503e-05,
+      "loss": 0.337,
+      "step": 15332
+    },
+    {
+      "epoch": 0.6733042800531301,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.2117542054195882e-05,
+      "loss": 0.2912,
+      "step": 15334
+    },
+    {
+      "epoch": 0.6733920985323336,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.2111613503111987e-05,
+      "loss": 0.3178,
+      "step": 15336
+    },
+    {
+      "epoch": 0.6734799170115372,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.2105685939023811e-05,
+      "loss": 0.3123,
+      "step": 15338
+    },
+    {
+      "epoch": 0.6735677354907407,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.2099759362385318e-05,
+      "loss": 0.3066,
+      "step": 15340
+    },
+    {
+      "epoch": 0.6736555539699441,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.2093833773650324e-05,
+      "loss": 0.3333,
+      "step": 15342
+    },
+    {
+      "epoch": 0.6737433724491476,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.2087909173272644e-05,
+      "loss": 0.3225,
+      "step": 15344
+    },
+    {
+      "epoch": 0.6738311909283511,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.2081985561705975e-05,
+      "loss": 0.3382,
+      "step": 15346
+    },
+    {
+      "epoch": 0.6739190094075546,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.2076062939403937e-05,
+      "loss": 0.2952,
+      "step": 15348
+    },
+    {
+      "epoch": 0.674006827886758,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.2070141306820106e-05,
+      "loss": 0.3388,
+      "step": 15350
+    },
+    {
+      "epoch": 0.6740946463659615,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.2064220664407946e-05,
+      "loss": 0.3128,
+      "step": 15352
+    },
+    {
+      "epoch": 0.6741824648451651,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.205830101262088e-05,
+      "loss": 0.3194,
+      "step": 15354
+    },
+    {
+      "epoch": 0.6742702833243686,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.2052382351912219e-05,
+      "loss": 0.3389,
+      "step": 15356
+    },
+    {
+      "epoch": 0.674358101803572,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.2046464682735217e-05,
+      "loss": 0.3324,
+      "step": 15358
+    },
+    {
+      "epoch": 0.6744459202827755,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.2040548005543062e-05,
+      "loss": 0.327,
+      "step": 15360
+    },
+    {
+      "epoch": 0.674533738761979,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.2034632320788839e-05,
+      "loss": 0.3616,
+      "step": 15362
+    },
+    {
+      "epoch": 0.6746215572411824,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.2028717628925587e-05,
+      "loss": 0.3157,
+      "step": 15364
+    },
+    {
+      "epoch": 0.6747093757203859,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.2022803930406242e-05,
+      "loss": 0.3295,
+      "step": 15366
+    },
+    {
+      "epoch": 0.6747971941995895,
+      "grad_norm": 1.8046875,
+      "learning_rate": 1.2016891225683691e-05,
+      "loss": 0.3492,
+      "step": 15368
+    },
+    {
+      "epoch": 0.674885012678793,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.2010979515210724e-05,
+      "loss": 0.3435,
+      "step": 15370
+    },
+    {
+      "epoch": 0.6749728311579964,
+      "grad_norm": 1.625,
+      "learning_rate": 1.2005068799440059e-05,
+      "loss": 0.3334,
+      "step": 15372
+    },
+    {
+      "epoch": 0.6750606496371999,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.1999159078824337e-05,
+      "loss": 0.3308,
+      "step": 15374
+    },
+    {
+      "epoch": 0.6751484681164034,
+      "grad_norm": 1.765625,
+      "learning_rate": 1.1993250353816124e-05,
+      "loss": 0.3166,
+      "step": 15376
+    },
+    {
+      "epoch": 0.6752362865956069,
+      "grad_norm": 1.796875,
+      "learning_rate": 1.1987342624867926e-05,
+      "loss": 0.3151,
+      "step": 15378
+    },
+    {
+      "epoch": 0.6753241050748103,
+      "grad_norm": 2.015625,
+      "learning_rate": 1.1981435892432139e-05,
+      "loss": 0.3483,
+      "step": 15380
+    },
+    {
+      "epoch": 0.6754119235540138,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.1975530156961119e-05,
+      "loss": 0.3139,
+      "step": 15382
+    },
+    {
+      "epoch": 0.6754997420332174,
+      "grad_norm": 1.5,
+      "learning_rate": 1.1969625418907123e-05,
+      "loss": 0.3511,
+      "step": 15384
+    },
+    {
+      "epoch": 0.6755875605124209,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.1963721678722328e-05,
+      "loss": 0.332,
+      "step": 15386
+    },
+    {
+      "epoch": 0.6756753789916243,
+      "grad_norm": 1.875,
+      "learning_rate": 1.1957818936858862e-05,
+      "loss": 0.3403,
+      "step": 15388
+    },
+    {
+      "epoch": 0.6757631974708278,
+      "grad_norm": 1.5078125,
+      "learning_rate": 1.1951917193768736e-05,
+      "loss": 0.3234,
+      "step": 15390
+    },
+    {
+      "epoch": 0.6758510159500313,
+      "grad_norm": 1.765625,
+      "learning_rate": 1.194601644990393e-05,
+      "loss": 0.3602,
+      "step": 15392
+    },
+    {
+      "epoch": 0.6759388344292347,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.1940116705716315e-05,
+      "loss": 0.3491,
+      "step": 15394
+    },
+    {
+      "epoch": 0.6760266529084382,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.1934217961657682e-05,
+      "loss": 0.335,
+      "step": 15396
+    },
+    {
+      "epoch": 0.6761144713876417,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.1928320218179779e-05,
+      "loss": 0.3248,
+      "step": 15398
+    },
+    {
+      "epoch": 0.6762022898668453,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.1922423475734248e-05,
+      "loss": 0.3105,
+      "step": 15400
+    },
+    {
+      "epoch": 0.6762901083460487,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.1916527734772661e-05,
+      "loss": 0.3222,
+      "step": 15402
+    },
+    {
+      "epoch": 0.6763779268252522,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.191063299574651e-05,
+      "loss": 0.3442,
+      "step": 15404
+    },
+    {
+      "epoch": 0.6764657453044557,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.1904739259107228e-05,
+      "loss": 0.3104,
+      "step": 15406
+    },
+    {
+      "epoch": 0.6765535637836592,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.1898846525306154e-05,
+      "loss": 0.3163,
+      "step": 15408
+    },
+    {
+      "epoch": 0.6766413822628626,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.1892954794794545e-05,
+      "loss": 0.3274,
+      "step": 15410
+    },
+    {
+      "epoch": 0.6767292007420661,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.1887064068023607e-05,
+      "loss": 0.336,
+      "step": 15412
+    },
+    {
+      "epoch": 0.6768170192212697,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.1881174345444437e-05,
+      "loss": 0.3256,
+      "step": 15414
+    },
+    {
+      "epoch": 0.6769048377004732,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.187528562750809e-05,
+      "loss": 0.3031,
+      "step": 15416
+    },
+    {
+      "epoch": 0.6769926561796766,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.1869397914665516e-05,
+      "loss": 0.355,
+      "step": 15418
+    },
+    {
+      "epoch": 0.6770804746588801,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.186351120736759e-05,
+      "loss": 0.3176,
+      "step": 15420
+    },
+    {
+      "epoch": 0.6771682931380836,
+      "grad_norm": 1.625,
+      "learning_rate": 1.1857625506065135e-05,
+      "loss": 0.3048,
+      "step": 15422
+    },
+    {
+      "epoch": 0.677256111617287,
+      "grad_norm": 1.625,
+      "learning_rate": 1.1851740811208856e-05,
+      "loss": 0.3217,
+      "step": 15424
+    },
+    {
+      "epoch": 0.6773439300964905,
+      "grad_norm": 1.75,
+      "learning_rate": 1.1845857123249427e-05,
+      "loss": 0.3281,
+      "step": 15426
+    },
+    {
+      "epoch": 0.677431748575694,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.1839974442637406e-05,
+      "loss": 0.3514,
+      "step": 15428
+    },
+    {
+      "epoch": 0.6775195670548976,
+      "grad_norm": 1.4921875,
+      "learning_rate": 1.1834092769823304e-05,
+      "loss": 0.3142,
+      "step": 15430
+    },
+    {
+      "epoch": 0.677607385534101,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.1828212105257536e-05,
+      "loss": 0.3072,
+      "step": 15432
+    },
+    {
+      "epoch": 0.6776952040133045,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.1822332449390441e-05,
+      "loss": 0.3315,
+      "step": 15434
+    },
+    {
+      "epoch": 0.677783022492508,
+      "grad_norm": 1.515625,
+      "learning_rate": 1.1816453802672286e-05,
+      "loss": 0.2974,
+      "step": 15436
+    },
+    {
+      "epoch": 0.6778708409717115,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.181057616555325e-05,
+      "loss": 0.3113,
+      "step": 15438
+    },
+    {
+      "epoch": 0.6779586594509149,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.1804699538483462e-05,
+      "loss": 0.3069,
+      "step": 15440
+    },
+    {
+      "epoch": 0.6780464779301184,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.1798823921912937e-05,
+      "loss": 0.3202,
+      "step": 15442
+    },
+    {
+      "epoch": 0.6781342964093219,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.1792949316291651e-05,
+      "loss": 0.3149,
+      "step": 15444
+    },
+    {
+      "epoch": 0.6782221148885255,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.1787075722069471e-05,
+      "loss": 0.3165,
+      "step": 15446
+    },
+    {
+      "epoch": 0.6783099333677289,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.1781203139696192e-05,
+      "loss": 0.348,
+      "step": 15448
+    },
+    {
+      "epoch": 0.6783977518469324,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.1775331569621553e-05,
+      "loss": 0.3292,
+      "step": 15450
+    },
+    {
+      "epoch": 0.6784855703261359,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.1769461012295183e-05,
+      "loss": 0.3128,
+      "step": 15452
+    },
+    {
+      "epoch": 0.6785733888053394,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.1763591468166671e-05,
+      "loss": 0.3219,
+      "step": 15454
+    },
+    {
+      "epoch": 0.6786612072845428,
+      "grad_norm": 1.78125,
+      "learning_rate": 1.1757722937685498e-05,
+      "loss": 0.337,
+      "step": 15456
+    },
+    {
+      "epoch": 0.6787490257637463,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.1751855421301064e-05,
+      "loss": 0.3249,
+      "step": 15458
+    },
+    {
+      "epoch": 0.6788368442429499,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.174598891946273e-05,
+      "loss": 0.31,
+      "step": 15460
+    },
+    {
+      "epoch": 0.6789246627221533,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.1740123432619741e-05,
+      "loss": 0.3139,
+      "step": 15462
+    },
+    {
+      "epoch": 0.6790124812013568,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.1734258961221278e-05,
+      "loss": 0.3365,
+      "step": 15464
+    },
+    {
+      "epoch": 0.6791002996805603,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.1728395505716433e-05,
+      "loss": 0.3293,
+      "step": 15466
+    },
+    {
+      "epoch": 0.6791881181597638,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.1722533066554254e-05,
+      "loss": 0.336,
+      "step": 15468
+    },
+    {
+      "epoch": 0.6792759366389672,
+      "grad_norm": 1.4453125,
+      "learning_rate": 1.1716671644183674e-05,
+      "loss": 0.3193,
+      "step": 15470
+    },
+    {
+      "epoch": 0.6793637551181707,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.1710811239053553e-05,
+      "loss": 0.3072,
+      "step": 15472
+    },
+    {
+      "epoch": 0.6794515735973742,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.1704951851612705e-05,
+      "loss": 0.3524,
+      "step": 15474
+    },
+    {
+      "epoch": 0.6795393920765778,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.169909348230982e-05,
+      "loss": 0.323,
+      "step": 15476
+    },
+    {
+      "epoch": 0.6796272105557812,
+      "grad_norm": 1.4921875,
+      "learning_rate": 1.1693236131593555e-05,
+      "loss": 0.3359,
+      "step": 15478
+    },
+    {
+      "epoch": 0.6797150290349847,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.1687379799912457e-05,
+      "loss": 0.3165,
+      "step": 15480
+    },
+    {
+      "epoch": 0.6798028475141882,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.1681524487714995e-05,
+      "loss": 0.3152,
+      "step": 15482
+    },
+    {
+      "epoch": 0.6798906659933917,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.167567019544959e-05,
+      "loss": 0.3278,
+      "step": 15484
+    },
+    {
+      "epoch": 0.6799784844725951,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.1669816923564544e-05,
+      "loss": 0.3325,
+      "step": 15486
+    },
+    {
+      "epoch": 0.6800663029517986,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.1663964672508126e-05,
+      "loss": 0.327,
+      "step": 15488
+    },
+    {
+      "epoch": 0.6801541214310021,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.1658113442728489e-05,
+      "loss": 0.3238,
+      "step": 15490
+    },
+    {
+      "epoch": 0.6802419399102057,
+      "grad_norm": 1.453125,
+      "learning_rate": 1.1652263234673725e-05,
+      "loss": 0.3403,
+      "step": 15492
+    },
+    {
+      "epoch": 0.6803297583894091,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.164641404879183e-05,
+      "loss": 0.321,
+      "step": 15494
+    },
+    {
+      "epoch": 0.6804175768686126,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.1640565885530758e-05,
+      "loss": 0.3183,
+      "step": 15496
+    },
+    {
+      "epoch": 0.6805053953478161,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.1634718745338353e-05,
+      "loss": 0.3223,
+      "step": 15498
+    },
+    {
+      "epoch": 0.6805932138270195,
+      "grad_norm": 1.625,
+      "learning_rate": 1.1628872628662381e-05,
+      "loss": 0.3443,
+      "step": 15500
+    },
+    {
+      "epoch": 0.680681032306223,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.1623027535950559e-05,
+      "loss": 0.3374,
+      "step": 15502
+    },
+    {
+      "epoch": 0.6807688507854265,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.1617183467650483e-05,
+      "loss": 0.33,
+      "step": 15504
+    },
+    {
+      "epoch": 0.68085666926463,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.1611340424209715e-05,
+      "loss": 0.3205,
+      "step": 15506
+    },
+    {
+      "epoch": 0.6809444877438335,
+      "grad_norm": 1.75,
+      "learning_rate": 1.160549840607571e-05,
+      "loss": 0.3222,
+      "step": 15508
+    },
+    {
+      "epoch": 0.681032306223037,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.1599657413695836e-05,
+      "loss": 0.362,
+      "step": 15510
+    },
+    {
+      "epoch": 0.6811201247022405,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.1593817447517419e-05,
+      "loss": 0.3281,
+      "step": 15512
+    },
+    {
+      "epoch": 0.681207943181444,
+      "grad_norm": 1.4296875,
+      "learning_rate": 1.1587978507987667e-05,
+      "loss": 0.2968,
+      "step": 15514
+    },
+    {
+      "epoch": 0.6812957616606474,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.1582140595553746e-05,
+      "loss": 0.3268,
+      "step": 15516
+    },
+    {
+      "epoch": 0.6813835801398509,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.157630371066271e-05,
+      "loss": 0.312,
+      "step": 15518
+    },
+    {
+      "epoch": 0.6814713986190544,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.1570467853761552e-05,
+      "loss": 0.3154,
+      "step": 15520
+    },
+    {
+      "epoch": 0.681559217098258,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.156463302529719e-05,
+      "loss": 0.2936,
+      "step": 15522
+    },
+    {
+      "epoch": 0.6816470355774614,
+      "grad_norm": 1.4609375,
+      "learning_rate": 1.1558799225716451e-05,
+      "loss": 0.3225,
+      "step": 15524
+    },
+    {
+      "epoch": 0.6817348540566649,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.155296645546609e-05,
+      "loss": 0.3127,
+      "step": 15526
+    },
+    {
+      "epoch": 0.6818226725358684,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.1547134714992772e-05,
+      "loss": 0.3079,
+      "step": 15528
+    },
+    {
+      "epoch": 0.6819104910150718,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.1541304004743112e-05,
+      "loss": 0.3163,
+      "step": 15530
+    },
+    {
+      "epoch": 0.6819983094942753,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.1535474325163618e-05,
+      "loss": 0.2959,
+      "step": 15532
+    },
+    {
+      "epoch": 0.6820861279734788,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.1529645676700717e-05,
+      "loss": 0.3165,
+      "step": 15534
+    },
+    {
+      "epoch": 0.6821739464526823,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.1523818059800793e-05,
+      "loss": 0.3352,
+      "step": 15536
+    },
+    {
+      "epoch": 0.6822617649318858,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.1517991474910097e-05,
+      "loss": 0.321,
+      "step": 15538
+    },
+    {
+      "epoch": 0.6823495834110893,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.1512165922474857e-05,
+      "loss": 0.3393,
+      "step": 15540
+    },
+    {
+      "epoch": 0.6824374018902928,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.1506341402941187e-05,
+      "loss": 0.3403,
+      "step": 15542
+    },
+    {
+      "epoch": 0.6825252203694963,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.1500517916755115e-05,
+      "loss": 0.316,
+      "step": 15544
+    },
+    {
+      "epoch": 0.6826130388486997,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.1494695464362627e-05,
+      "loss": 0.3239,
+      "step": 15546
+    },
+    {
+      "epoch": 0.6827008573279032,
+      "grad_norm": 1.625,
+      "learning_rate": 1.1488874046209588e-05,
+      "loss": 0.3235,
+      "step": 15548
+    },
+    {
+      "epoch": 0.6827886758071067,
+      "grad_norm": 1.5078125,
+      "learning_rate": 1.1483053662741822e-05,
+      "loss": 0.3239,
+      "step": 15550
+    },
+    {
+      "epoch": 0.6828764942863101,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.1477234314405048e-05,
+      "loss": 0.304,
+      "step": 15552
+    },
+    {
+      "epoch": 0.6829643127655137,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.1471416001644911e-05,
+      "loss": 0.3451,
+      "step": 15554
+    },
+    {
+      "epoch": 0.6830521312447172,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.146559872490697e-05,
+      "loss": 0.2889,
+      "step": 15556
+    },
+    {
+      "epoch": 0.6831399497239207,
+      "grad_norm": 1.765625,
+      "learning_rate": 1.1459782484636734e-05,
+      "loss": 0.3077,
+      "step": 15558
+    },
+    {
+      "epoch": 0.6832277682031241,
+      "grad_norm": 1.625,
+      "learning_rate": 1.1453967281279601e-05,
+      "loss": 0.346,
+      "step": 15560
+    },
+    {
+      "epoch": 0.6833155866823276,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.144815311528089e-05,
+      "loss": 0.3292,
+      "step": 15562
+    },
+    {
+      "epoch": 0.6834034051615311,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.1442339987085873e-05,
+      "loss": 0.3335,
+      "step": 15564
+    },
+    {
+      "epoch": 0.6834912236407346,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.1436527897139698e-05,
+      "loss": 0.2855,
+      "step": 15566
+    },
+    {
+      "epoch": 0.6835790421199381,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.1430716845887478e-05,
+      "loss": 0.322,
+      "step": 15568
+    },
+    {
+      "epoch": 0.6836668605991416,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.1424906833774218e-05,
+      "loss": 0.297,
+      "step": 15570
+    },
+    {
+      "epoch": 0.6837546790783451,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.1419097861244834e-05,
+      "loss": 0.299,
+      "step": 15572
+    },
+    {
+      "epoch": 0.6838424975575486,
+      "grad_norm": 1.625,
+      "learning_rate": 1.1413289928744203e-05,
+      "loss": 0.3015,
+      "step": 15574
+    },
+    {
+      "epoch": 0.683930316036752,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.1407483036717076e-05,
+      "loss": 0.3119,
+      "step": 15576
+    },
+    {
+      "epoch": 0.6840181345159555,
+      "grad_norm": 1.4453125,
+      "learning_rate": 1.1401677185608165e-05,
+      "loss": 0.3085,
+      "step": 15578
+    },
+    {
+      "epoch": 0.684105952995159,
+      "grad_norm": 1.625,
+      "learning_rate": 1.1395872375862074e-05,
+      "loss": 0.3267,
+      "step": 15580
+    },
+    {
+      "epoch": 0.6841937714743624,
+      "grad_norm": 1.8671875,
+      "learning_rate": 1.139006860792333e-05,
+      "loss": 0.3639,
+      "step": 15582
+    },
+    {
+      "epoch": 0.684281589953566,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.1384265882236414e-05,
+      "loss": 0.3307,
+      "step": 15584
+    },
+    {
+      "epoch": 0.6843694084327695,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.1378464199245659e-05,
+      "loss": 0.3185,
+      "step": 15586
+    },
+    {
+      "epoch": 0.684457226911973,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.137266355939539e-05,
+      "loss": 0.2997,
+      "step": 15588
+    },
+    {
+      "epoch": 0.6845450453911764,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.1366863963129805e-05,
+      "loss": 0.3336,
+      "step": 15590
+    },
+    {
+      "epoch": 0.6846328638703799,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.1361065410893057e-05,
+      "loss": 0.2917,
+      "step": 15592
+    },
+    {
+      "epoch": 0.6847206823495834,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.1355267903129187e-05,
+      "loss": 0.3479,
+      "step": 15594
+    },
+    {
+      "epoch": 0.6848085008287869,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.1349471440282164e-05,
+      "loss": 0.2962,
+      "step": 15596
+    },
+    {
+      "epoch": 0.6848963193079903,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.1343676022795898e-05,
+      "loss": 0.3089,
+      "step": 15598
+    },
+    {
+      "epoch": 0.6849841377871939,
+      "grad_norm": 1.625,
+      "learning_rate": 1.1337881651114188e-05,
+      "loss": 0.3061,
+      "step": 15600
+    },
+    {
+      "epoch": 0.6850719562663974,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.1332088325680783e-05,
+      "loss": 0.3318,
+      "step": 15602
+    },
+    {
+      "epoch": 0.6851597747456009,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.1326296046939333e-05,
+      "loss": 0.2978,
+      "step": 15604
+    },
+    {
+      "epoch": 0.6852475932248043,
+      "grad_norm": 1.4921875,
+      "learning_rate": 1.1320504815333399e-05,
+      "loss": 0.3091,
+      "step": 15606
+    },
+    {
+      "epoch": 0.6853354117040078,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.1314714631306495e-05,
+      "loss": 0.3161,
+      "step": 15608
+    },
+    {
+      "epoch": 0.6854232301832113,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.1308925495302017e-05,
+      "loss": 0.3394,
+      "step": 15610
+    },
+    {
+      "epoch": 0.6855110486624147,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.1303137407763314e-05,
+      "loss": 0.3152,
+      "step": 15612
+    },
+    {
+      "epoch": 0.6855988671416183,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.1297350369133632e-05,
+      "loss": 0.3281,
+      "step": 15614
+    },
+    {
+      "epoch": 0.6856866856208218,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.1291564379856145e-05,
+      "loss": 0.3239,
+      "step": 15616
+    },
+    {
+      "epoch": 0.6857745041000253,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.1285779440373943e-05,
+      "loss": 0.3456,
+      "step": 15618
+    },
+    {
+      "epoch": 0.6858623225792287,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.1279995551130029e-05,
+      "loss": 0.3397,
+      "step": 15620
+    },
+    {
+      "epoch": 0.6859501410584322,
+      "grad_norm": 1.5,
+      "learning_rate": 1.1274212712567354e-05,
+      "loss": 0.3237,
+      "step": 15622
+    },
+    {
+      "epoch": 0.6860379595376357,
+      "grad_norm": 1.5,
+      "learning_rate": 1.126843092512875e-05,
+      "loss": 0.3207,
+      "step": 15624
+    },
+    {
+      "epoch": 0.6861257780168392,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.126265018925701e-05,
+      "loss": 0.3607,
+      "step": 15626
+    },
+    {
+      "epoch": 0.6862135964960426,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.1256870505394798e-05,
+      "loss": 0.2892,
+      "step": 15628
+    },
+    {
+      "epoch": 0.6863014149752462,
+      "grad_norm": 1.515625,
+      "learning_rate": 1.1251091873984748e-05,
+      "loss": 0.326,
+      "step": 15630
+    },
+    {
+      "epoch": 0.6863892334544497,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.1245314295469379e-05,
+      "loss": 0.3374,
+      "step": 15632
+    },
+    {
+      "epoch": 0.6864770519336532,
+      "grad_norm": 1.5078125,
+      "learning_rate": 1.1239537770291128e-05,
+      "loss": 0.3499,
+      "step": 15634
+    },
+    {
+      "epoch": 0.6865648704128566,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.1233762298892384e-05,
+      "loss": 0.3229,
+      "step": 15636
+    },
+    {
+      "epoch": 0.6866526888920601,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.1227987881715412e-05,
+      "loss": 0.3101,
+      "step": 15638
+    },
+    {
+      "epoch": 0.6867405073712636,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.1222214519202439e-05,
+      "loss": 0.326,
+      "step": 15640
+    },
+    {
+      "epoch": 0.686828325850467,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.1216442211795582e-05,
+      "loss": 0.3178,
+      "step": 15642
+    },
+    {
+      "epoch": 0.6869161443296705,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.1210670959936875e-05,
+      "loss": 0.2925,
+      "step": 15644
+    },
+    {
+      "epoch": 0.6870039628088741,
+      "grad_norm": 1.46875,
+      "learning_rate": 1.120490076406831e-05,
+      "loss": 0.2749,
+      "step": 15646
+    },
+    {
+      "epoch": 0.6870917812880776,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.1199131624631734e-05,
+      "loss": 0.313,
+      "step": 15648
+    },
+    {
+      "epoch": 0.687179599767281,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.1193363542068974e-05,
+      "loss": 0.3055,
+      "step": 15650
+    },
+    {
+      "epoch": 0.6872674182464845,
+      "grad_norm": 1.484375,
+      "learning_rate": 1.1187596516821734e-05,
+      "loss": 0.3377,
+      "step": 15652
+    },
+    {
+      "epoch": 0.687355236725688,
+      "grad_norm": 1.484375,
+      "learning_rate": 1.1181830549331674e-05,
+      "loss": 0.3177,
+      "step": 15654
+    },
+    {
+      "epoch": 0.6874430552048915,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.1176065640040342e-05,
+      "loss": 0.3335,
+      "step": 15656
+    },
+    {
+      "epoch": 0.6875308736840949,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.1170301789389209e-05,
+      "loss": 0.2978,
+      "step": 15658
+    },
+    {
+      "epoch": 0.6876186921632985,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.116453899781969e-05,
+      "loss": 0.3564,
+      "step": 15660
+    },
+    {
+      "epoch": 0.687706510642502,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.115877726577308e-05,
+      "loss": 0.3511,
+      "step": 15662
+    },
+    {
+      "epoch": 0.6877943291217055,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.1153016593690634e-05,
+      "loss": 0.3411,
+      "step": 15664
+    },
+    {
+      "epoch": 0.6878821476009089,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.11472569820135e-05,
+      "loss": 0.333,
+      "step": 15666
+    },
+    {
+      "epoch": 0.6879699660801124,
+      "grad_norm": 2.015625,
+      "learning_rate": 1.1141498431182735e-05,
+      "loss": 0.3437,
+      "step": 15668
+    },
+    {
+      "epoch": 0.6880577845593159,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.1135740941639353e-05,
+      "loss": 0.3365,
+      "step": 15670
+    },
+    {
+      "epoch": 0.6881456030385193,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.1129984513824241e-05,
+      "loss": 0.3471,
+      "step": 15672
+    },
+    {
+      "epoch": 0.6882334215177228,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.1124229148178253e-05,
+      "loss": 0.3298,
+      "step": 15674
+    },
+    {
+      "epoch": 0.6883212399969264,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.1118474845142122e-05,
+      "loss": 0.3135,
+      "step": 15676
+    },
+    {
+      "epoch": 0.6884090584761299,
+      "grad_norm": 1.78125,
+      "learning_rate": 1.1112721605156511e-05,
+      "loss": 0.3343,
+      "step": 15678
+    },
+    {
+      "epoch": 0.6884968769553333,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.1106969428662011e-05,
+      "loss": 0.3215,
+      "step": 15680
+    },
+    {
+      "epoch": 0.6885846954345368,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.1101218316099115e-05,
+      "loss": 0.3083,
+      "step": 15682
+    },
+    {
+      "epoch": 0.6886725139137403,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.109546826790826e-05,
+      "loss": 0.3814,
+      "step": 15684
+    },
+    {
+      "epoch": 0.6887603323929438,
+      "grad_norm": 1.765625,
+      "learning_rate": 1.108971928452977e-05,
+      "loss": 0.3308,
+      "step": 15686
+    },
+    {
+      "epoch": 0.6888481508721472,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.108397136640392e-05,
+      "loss": 0.3188,
+      "step": 15688
+    },
+    {
+      "epoch": 0.6889359693513507,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.107822451397087e-05,
+      "loss": 0.3336,
+      "step": 15690
+    },
+    {
+      "epoch": 0.6890237878305543,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.1072478727670732e-05,
+      "loss": 0.3128,
+      "step": 15692
+    },
+    {
+      "epoch": 0.6891116063097578,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.1066734007943514e-05,
+      "loss": 0.3277,
+      "step": 15694
+    },
+    {
+      "epoch": 0.6891994247889612,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.1060990355229134e-05,
+      "loss": 0.3215,
+      "step": 15696
+    },
+    {
+      "epoch": 0.6892872432681647,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.1055247769967465e-05,
+      "loss": 0.3183,
+      "step": 15698
+    },
+    {
+      "epoch": 0.6893750617473682,
+      "grad_norm": 1.5078125,
+      "learning_rate": 1.1049506252598255e-05,
+      "loss": 0.3155,
+      "step": 15700
+    },
+    {
+      "epoch": 0.6894628802265717,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.1043765803561207e-05,
+      "loss": 0.3152,
+      "step": 15702
+    },
+    {
+      "epoch": 0.6895506987057751,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.1038026423295923e-05,
+      "loss": 0.3361,
+      "step": 15704
+    },
+    {
+      "epoch": 0.6896385171849786,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.103228811224191e-05,
+      "loss": 0.3342,
+      "step": 15706
+    },
+    {
+      "epoch": 0.6897263356641822,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.1026550870838643e-05,
+      "loss": 0.3308,
+      "step": 15708
+    },
+    {
+      "epoch": 0.6898141541433856,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.1020814699525439e-05,
+      "loss": 0.3376,
+      "step": 15710
+    },
+    {
+      "epoch": 0.6899019726225891,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.1015079598741607e-05,
+      "loss": 0.3261,
+      "step": 15712
+    },
+    {
+      "epoch": 0.6899897911017926,
+      "grad_norm": 1.625,
+      "learning_rate": 1.1009345568926321e-05,
+      "loss": 0.3297,
+      "step": 15714
+    },
+    {
+      "epoch": 0.6900776095809961,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.1003612610518718e-05,
+      "loss": 0.3,
+      "step": 15716
+    },
+    {
+      "epoch": 0.6901654280601995,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.0997880723957812e-05,
+      "loss": 0.2955,
+      "step": 15718
+    },
+    {
+      "epoch": 0.690253246539403,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.099214990968255e-05,
+      "loss": 0.3062,
+      "step": 15720
+    },
+    {
+      "epoch": 0.6903410650186066,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.0986420168131817e-05,
+      "loss": 0.3384,
+      "step": 15722
+    },
+    {
+      "epoch": 0.6904288834978101,
+      "grad_norm": 1.5078125,
+      "learning_rate": 1.0980691499744375e-05,
+      "loss": 0.3259,
+      "step": 15724
+    },
+    {
+      "epoch": 0.6905167019770135,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.0974963904958947e-05,
+      "loss": 0.3113,
+      "step": 15726
+    },
+    {
+      "epoch": 0.690604520456217,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.0969237384214146e-05,
+      "loss": 0.3524,
+      "step": 15728
+    },
+    {
+      "epoch": 0.6906923389354205,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.0963511937948501e-05,
+      "loss": 0.2989,
+      "step": 15730
+    },
+    {
+      "epoch": 0.690780157414624,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.0957787566600486e-05,
+      "loss": 0.3107,
+      "step": 15732
+    },
+    {
+      "epoch": 0.6908679758938274,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.0952064270608452e-05,
+      "loss": 0.3216,
+      "step": 15734
+    },
+    {
+      "epoch": 0.6909557943730309,
+      "grad_norm": 1.515625,
+      "learning_rate": 1.0946342050410719e-05,
+      "loss": 0.3078,
+      "step": 15736
+    },
+    {
+      "epoch": 0.6910436128522345,
+      "grad_norm": 1.515625,
+      "learning_rate": 1.0940620906445478e-05,
+      "loss": 0.323,
+      "step": 15738
+    },
+    {
+      "epoch": 0.691131431331438,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.0934900839150858e-05,
+      "loss": 0.2987,
+      "step": 15740
+    },
+    {
+      "epoch": 0.6912192498106414,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.0929181848964904e-05,
+      "loss": 0.3072,
+      "step": 15742
+    },
+    {
+      "epoch": 0.6913070682898449,
+      "grad_norm": 1.515625,
+      "learning_rate": 1.0923463936325568e-05,
+      "loss": 0.2987,
+      "step": 15744
+    },
+    {
+      "epoch": 0.6913948867690484,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.0917747101670744e-05,
+      "loss": 0.2951,
+      "step": 15746
+    },
+    {
+      "epoch": 0.6914827052482518,
+      "grad_norm": 1.484375,
+      "learning_rate": 1.0912031345438218e-05,
+      "loss": 0.3291,
+      "step": 15748
+    },
+    {
+      "epoch": 0.6915705237274553,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.0906316668065717e-05,
+      "loss": 0.3089,
+      "step": 15750
+    },
+    {
+      "epoch": 0.6916583422066588,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.0900603069990861e-05,
+      "loss": 0.3456,
+      "step": 15752
+    },
+    {
+      "epoch": 0.6917461606858624,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.0894890551651197e-05,
+      "loss": 0.337,
+      "step": 15754
+    },
+    {
+      "epoch": 0.6918339791650658,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.0889179113484202e-05,
+      "loss": 0.3076,
+      "step": 15756
+    },
+    {
+      "epoch": 0.6919217976442693,
+      "grad_norm": 1.484375,
+      "learning_rate": 1.0883468755927245e-05,
+      "loss": 0.322,
+      "step": 15758
+    },
+    {
+      "epoch": 0.6920096161234728,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.0877759479417643e-05,
+      "loss": 0.3365,
+      "step": 15760
+    },
+    {
+      "epoch": 0.6920974346026763,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.0872051284392596e-05,
+      "loss": 0.3051,
+      "step": 15762
+    },
+    {
+      "epoch": 0.6921852530818797,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.0866344171289259e-05,
+      "loss": 0.2926,
+      "step": 15764
+    },
+    {
+      "epoch": 0.6922730715610832,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.0860638140544672e-05,
+      "loss": 0.302,
+      "step": 15766
+    },
+    {
+      "epoch": 0.6923608900402868,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.0854933192595806e-05,
+      "loss": 0.3183,
+      "step": 15768
+    },
+    {
+      "epoch": 0.6924487085194903,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.0849229327879548e-05,
+      "loss": 0.3038,
+      "step": 15770
+    },
+    {
+      "epoch": 0.6925365269986937,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.0843526546832688e-05,
+      "loss": 0.3107,
+      "step": 15772
+    },
+    {
+      "epoch": 0.6926243454778972,
+      "grad_norm": 1.4921875,
+      "learning_rate": 1.0837824849891972e-05,
+      "loss": 0.32,
+      "step": 15774
+    },
+    {
+      "epoch": 0.6927121639571007,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.0832124237494013e-05,
+      "loss": 0.3063,
+      "step": 15776
+    },
+    {
+      "epoch": 0.6927999824363041,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.0826424710075383e-05,
+      "loss": 0.3032,
+      "step": 15778
+    },
+    {
+      "epoch": 0.6928878009155076,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.082072626807255e-05,
+      "loss": 0.3558,
+      "step": 15780
+    },
+    {
+      "epoch": 0.6929756193947111,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.0815028911921887e-05,
+      "loss": 0.2991,
+      "step": 15782
+    },
+    {
+      "epoch": 0.6930634378739147,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.0809332642059721e-05,
+      "loss": 0.3309,
+      "step": 15784
+    },
+    {
+      "epoch": 0.6931512563531181,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.0803637458922253e-05,
+      "loss": 0.313,
+      "step": 15786
+    },
+    {
+      "epoch": 0.6932390748323216,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.079794336294564e-05,
+      "loss": 0.2886,
+      "step": 15788
+    },
+    {
+      "epoch": 0.6933268933115251,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.079225035456593e-05,
+      "loss": 0.3351,
+      "step": 15790
+    },
+    {
+      "epoch": 0.6934147117907286,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.0786558434219082e-05,
+      "loss": 0.3275,
+      "step": 15792
+    },
+    {
+      "epoch": 0.693502530269932,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.0780867602341007e-05,
+      "loss": 0.2797,
+      "step": 15794
+    },
+    {
+      "epoch": 0.6935903487491355,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.0775177859367492e-05,
+      "loss": 0.319,
+      "step": 15796
+    },
+    {
+      "epoch": 0.693678167228339,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.0769489205734276e-05,
+      "loss": 0.3419,
+      "step": 15798
+    },
+    {
+      "epoch": 0.6937659857075426,
+      "grad_norm": 1.453125,
+      "learning_rate": 1.0763801641876986e-05,
+      "loss": 0.2933,
+      "step": 15800
+    },
+    {
+      "epoch": 0.693853804186746,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.0758115168231178e-05,
+      "loss": 0.2888,
+      "step": 15802
+    },
+    {
+      "epoch": 0.6939416226659495,
+      "grad_norm": 1.8046875,
+      "learning_rate": 1.0752429785232326e-05,
+      "loss": 0.3036,
+      "step": 15804
+    },
+    {
+      "epoch": 0.694029441145153,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.0746745493315807e-05,
+      "loss": 0.316,
+      "step": 15806
+    },
+    {
+      "epoch": 0.6941172596243564,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.0741062292916943e-05,
+      "loss": 0.3299,
+      "step": 15808
+    },
+    {
+      "epoch": 0.6942050781035599,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.073538018447094e-05,
+      "loss": 0.3164,
+      "step": 15810
+    },
+    {
+      "epoch": 0.6942928965827634,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.072969916841295e-05,
+      "loss": 0.3163,
+      "step": 15812
+    },
+    {
+      "epoch": 0.694380715061967,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.0724019245178016e-05,
+      "loss": 0.3286,
+      "step": 15814
+    },
+    {
+      "epoch": 0.6944685335411704,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.0718340415201104e-05,
+      "loss": 0.3321,
+      "step": 15816
+    },
+    {
+      "epoch": 0.6945563520203739,
+      "grad_norm": 1.5078125,
+      "learning_rate": 1.0712662678917115e-05,
+      "loss": 0.3012,
+      "step": 15818
+    },
+    {
+      "epoch": 0.6946441704995774,
+      "grad_norm": 1.5,
+      "learning_rate": 1.0706986036760833e-05,
+      "loss": 0.3009,
+      "step": 15820
+    },
+    {
+      "epoch": 0.6947319889787809,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.0701310489166997e-05,
+      "loss": 0.3225,
+      "step": 15822
+    },
+    {
+      "epoch": 0.6948198074579843,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.0695636036570222e-05,
+      "loss": 0.3273,
+      "step": 15824
+    },
+    {
+      "epoch": 0.6949076259371878,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.0689962679405077e-05,
+      "loss": 0.3408,
+      "step": 15826
+    },
+    {
+      "epoch": 0.6949954444163913,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.0684290418106022e-05,
+      "loss": 0.2963,
+      "step": 15828
+    },
+    {
+      "epoch": 0.6950832628955949,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.0678619253107436e-05,
+      "loss": 0.3406,
+      "step": 15830
+    },
+    {
+      "epoch": 0.6951710813747983,
+      "grad_norm": 1.515625,
+      "learning_rate": 1.0672949184843622e-05,
+      "loss": 0.3546,
+      "step": 15832
+    },
+    {
+      "epoch": 0.6952588998540018,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.0667280213748784e-05,
+      "loss": 0.3345,
+      "step": 15834
+    },
+    {
+      "epoch": 0.6953467183332053,
+      "grad_norm": 1.625,
+      "learning_rate": 1.0661612340257071e-05,
+      "loss": 0.3073,
+      "step": 15836
+    },
+    {
+      "epoch": 0.6954345368124087,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.0655945564802517e-05,
+      "loss": 0.3455,
+      "step": 15838
+    },
+    {
+      "epoch": 0.6955223552916122,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.0650279887819094e-05,
+      "loss": 0.3232,
+      "step": 15840
+    },
+    {
+      "epoch": 0.6956101737708157,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.0644615309740683e-05,
+      "loss": 0.3087,
+      "step": 15842
+    },
+    {
+      "epoch": 0.6956979922500192,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.063895183100106e-05,
+      "loss": 0.3733,
+      "step": 15844
+    },
+    {
+      "epoch": 0.6957858107292227,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.0633289452033957e-05,
+      "loss": 0.3111,
+      "step": 15846
+    },
+    {
+      "epoch": 0.6958736292084262,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.0627628173272986e-05,
+      "loss": 0.3288,
+      "step": 15848
+    },
+    {
+      "epoch": 0.6959614476876297,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.0621967995151699e-05,
+      "loss": 0.3649,
+      "step": 15850
+    },
+    {
+      "epoch": 0.6960492661668332,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.0616308918103554e-05,
+      "loss": 0.3319,
+      "step": 15852
+    },
+    {
+      "epoch": 0.6961370846460366,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.0610650942561909e-05,
+      "loss": 0.3189,
+      "step": 15854
+    },
+    {
+      "epoch": 0.6962249031252401,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.0604994068960072e-05,
+      "loss": 0.3257,
+      "step": 15856
+    },
+    {
+      "epoch": 0.6963127216044436,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.0599338297731231e-05,
+      "loss": 0.3204,
+      "step": 15858
+    },
+    {
+      "epoch": 0.696400540083647,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.0593683629308537e-05,
+      "loss": 0.3182,
+      "step": 15860
+    },
+    {
+      "epoch": 0.6964883585628506,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.0588030064124981e-05,
+      "loss": 0.3205,
+      "step": 15862
+    },
+    {
+      "epoch": 0.6965761770420541,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.058237760261355e-05,
+      "loss": 0.3209,
+      "step": 15864
+    },
+    {
+      "epoch": 0.6966639955212576,
+      "grad_norm": 1.4921875,
+      "learning_rate": 1.05767262452071e-05,
+      "loss": 0.2977,
+      "step": 15866
+    },
+    {
+      "epoch": 0.696751814000461,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.0571075992338398e-05,
+      "loss": 0.3416,
+      "step": 15868
+    },
+    {
+      "epoch": 0.6968396324796645,
+      "grad_norm": 1.515625,
+      "learning_rate": 1.0565426844440166e-05,
+      "loss": 0.3263,
+      "step": 15870
+    },
+    {
+      "epoch": 0.696927450958868,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.0559778801945e-05,
+      "loss": 0.3241,
+      "step": 15872
+    },
+    {
+      "epoch": 0.6970152694380715,
+      "grad_norm": 1.4296875,
+      "learning_rate": 1.0554131865285441e-05,
+      "loss": 0.332,
+      "step": 15874
+    },
+    {
+      "epoch": 0.697103087917275,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.0548486034893926e-05,
+      "loss": 0.3466,
+      "step": 15876
+    },
+    {
+      "epoch": 0.6971909063964785,
+      "grad_norm": 1.4609375,
+      "learning_rate": 1.0542841311202809e-05,
+      "loss": 0.3141,
+      "step": 15878
+    },
+    {
+      "epoch": 0.697278724875682,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.0537197694644376e-05,
+      "loss": 0.3049,
+      "step": 15880
+    },
+    {
+      "epoch": 0.6973665433548855,
+      "grad_norm": 1.4375,
+      "learning_rate": 1.0531555185650802e-05,
+      "loss": 0.3313,
+      "step": 15882
+    },
+    {
+      "epoch": 0.6974543618340889,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.052591378465421e-05,
+      "loss": 0.3541,
+      "step": 15884
+    },
+    {
+      "epoch": 0.6975421803132924,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.052027349208661e-05,
+      "loss": 0.3362,
+      "step": 15886
+    },
+    {
+      "epoch": 0.6976299987924959,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.0514634308379928e-05,
+      "loss": 0.3457,
+      "step": 15888
+    },
+    {
+      "epoch": 0.6977178172716993,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.050899623396603e-05,
+      "loss": 0.3524,
+      "step": 15890
+    },
+    {
+      "epoch": 0.6978056357509029,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.0503359269276678e-05,
+      "loss": 0.3214,
+      "step": 15892
+    },
+    {
+      "epoch": 0.6978934542301064,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.0497723414743546e-05,
+      "loss": 0.2927,
+      "step": 15894
+    },
+    {
+      "epoch": 0.6979812727093099,
+      "grad_norm": 1.453125,
+      "learning_rate": 1.0492088670798223e-05,
+      "loss": 0.317,
+      "step": 15896
+    },
+    {
+      "epoch": 0.6980690911885133,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.0486455037872236e-05,
+      "loss": 0.3112,
+      "step": 15898
+    },
+    {
+      "epoch": 0.6981569096677168,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.0480822516396994e-05,
+      "loss": 0.3209,
+      "step": 15900
+    },
+    {
+      "epoch": 0.6982447281469203,
+      "grad_norm": 1.4765625,
+      "learning_rate": 1.047519110680385e-05,
+      "loss": 0.3126,
+      "step": 15902
+    },
+    {
+      "epoch": 0.6983325466261238,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.0469560809524056e-05,
+      "loss": 0.2958,
+      "step": 15904
+    },
+    {
+      "epoch": 0.6984203651053272,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.0463931624988769e-05,
+      "loss": 0.308,
+      "step": 15906
+    },
+    {
+      "epoch": 0.6985081835845308,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.045830355362909e-05,
+      "loss": 0.3281,
+      "step": 15908
+    },
+    {
+      "epoch": 0.6985960020637343,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.0452676595876001e-05,
+      "loss": 0.3242,
+      "step": 15910
+    },
+    {
+      "epoch": 0.6986838205429378,
+      "grad_norm": 1.625,
+      "learning_rate": 1.0447050752160436e-05,
+      "loss": 0.3326,
+      "step": 15912
+    },
+    {
+      "epoch": 0.6987716390221412,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.0441426022913211e-05,
+      "loss": 0.3141,
+      "step": 15914
+    },
+    {
+      "epoch": 0.6988594575013447,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.0435802408565065e-05,
+      "loss": 0.3224,
+      "step": 15916
+    },
+    {
+      "epoch": 0.6989472759805482,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.043017990954667e-05,
+      "loss": 0.3246,
+      "step": 15918
+    },
+    {
+      "epoch": 0.6990350944597516,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.042455852628858e-05,
+      "loss": 0.318,
+      "step": 15920
+    },
+    {
+      "epoch": 0.6991229129389552,
+      "grad_norm": 1.7890625,
+      "learning_rate": 1.0418938259221311e-05,
+      "loss": 0.3173,
+      "step": 15922
+    },
+    {
+      "epoch": 0.6992107314181587,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.041331910877523e-05,
+      "loss": 0.2939,
+      "step": 15924
+    },
+    {
+      "epoch": 0.6992985498973622,
+      "grad_norm": 1.5078125,
+      "learning_rate": 1.0407701075380674e-05,
+      "loss": 0.313,
+      "step": 15926
+    },
+    {
+      "epoch": 0.6993863683765656,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.0402084159467867e-05,
+      "loss": 0.3124,
+      "step": 15928
+    },
+    {
+      "epoch": 0.6994741868557691,
+      "grad_norm": 1.625,
+      "learning_rate": 1.0396468361466947e-05,
+      "loss": 0.3338,
+      "step": 15930
+    },
+    {
+      "epoch": 0.6995620053349726,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.0390853681807989e-05,
+      "loss": 0.2974,
+      "step": 15932
+    },
+    {
+      "epoch": 0.6996498238141761,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.038524012092095e-05,
+      "loss": 0.3551,
+      "step": 15934
+    },
+    {
+      "epoch": 0.6997376422933795,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.0379627679235734e-05,
+      "loss": 0.3138,
+      "step": 15936
+    },
+    {
+      "epoch": 0.6998254607725831,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.0374016357182137e-05,
+      "loss": 0.3576,
+      "step": 15938
+    },
+    {
+      "epoch": 0.6999132792517866,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.0368406155189862e-05,
+      "loss": 0.3219,
+      "step": 15940
+    },
+    {
+      "epoch": 0.7000010977309901,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.036279707368856e-05,
+      "loss": 0.3306,
+      "step": 15942
+    },
+    {
+      "epoch": 0.7000889162101935,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.035718911310776e-05,
+      "loss": 0.3197,
+      "step": 15944
+    },
+    {
+      "epoch": 0.700176734689397,
+      "grad_norm": 1.625,
+      "learning_rate": 1.0351582273876936e-05,
+      "loss": 0.2982,
+      "step": 15946
+    },
+    {
+      "epoch": 0.7002645531686005,
+      "grad_norm": 1.5,
+      "learning_rate": 1.0345976556425452e-05,
+      "loss": 0.3038,
+      "step": 15948
+    },
+    {
+      "epoch": 0.700352371647804,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.0340371961182588e-05,
+      "loss": 0.3031,
+      "step": 15950
+    },
+    {
+      "epoch": 0.7004401901270074,
+      "grad_norm": 1.4765625,
+      "learning_rate": 1.0334768488577563e-05,
+      "loss": 0.3268,
+      "step": 15952
+    },
+    {
+      "epoch": 0.700528008606211,
+      "grad_norm": 1.8671875,
+      "learning_rate": 1.032916613903948e-05,
+      "loss": 0.317,
+      "step": 15954
+    },
+    {
+      "epoch": 0.7006158270854145,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.0323564912997371e-05,
+      "loss": 0.3188,
+      "step": 15956
+    },
+    {
+      "epoch": 0.700703645564618,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.0317964810880173e-05,
+      "loss": 0.33,
+      "step": 15958
+    },
+    {
+      "epoch": 0.7007914640438214,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.0312365833116757e-05,
+      "loss": 0.3089,
+      "step": 15960
+    },
+    {
+      "epoch": 0.7008792825230249,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.0306767980135878e-05,
+      "loss": 0.3248,
+      "step": 15962
+    },
+    {
+      "epoch": 0.7009671010022284,
+      "grad_norm": 1.5,
+      "learning_rate": 1.0301171252366238e-05,
+      "loss": 0.3073,
+      "step": 15964
+    },
+    {
+      "epoch": 0.7010549194814318,
+      "grad_norm": 1.4609375,
+      "learning_rate": 1.0295575650236428e-05,
+      "loss": 0.339,
+      "step": 15966
+    },
+    {
+      "epoch": 0.7011427379606354,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.0289981174174947e-05,
+      "loss": 0.3299,
+      "step": 15968
+    },
+    {
+      "epoch": 0.7012305564398389,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.0284387824610247e-05,
+      "loss": 0.3266,
+      "step": 15970
+    },
+    {
+      "epoch": 0.7013183749190424,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.0278795601970646e-05,
+      "loss": 0.3139,
+      "step": 15972
+    },
+    {
+      "epoch": 0.7014061933982458,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.027320450668441e-05,
+      "loss": 0.3182,
+      "step": 15974
+    },
+    {
+      "epoch": 0.7014940118774493,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.026761453917971e-05,
+      "loss": 0.3215,
+      "step": 15976
+    },
+    {
+      "epoch": 0.7015818303566528,
+      "grad_norm": 1.4609375,
+      "learning_rate": 1.026202569988461e-05,
+      "loss": 0.3387,
+      "step": 15978
+    },
+    {
+      "epoch": 0.7016696488358563,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.025643798922712e-05,
+      "loss": 0.3404,
+      "step": 15980
+    },
+    {
+      "epoch": 0.7017574673150597,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.0250851407635137e-05,
+      "loss": 0.3075,
+      "step": 15982
+    },
+    {
+      "epoch": 0.7018452857942633,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.0245265955536503e-05,
+      "loss": 0.3429,
+      "step": 15984
+    },
+    {
+      "epoch": 0.7019331042734668,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.0239681633358924e-05,
+      "loss": 0.3293,
+      "step": 15986
+    },
+    {
+      "epoch": 0.7020209227526703,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.0234098441530075e-05,
+      "loss": 0.3247,
+      "step": 15988
+    },
+    {
+      "epoch": 0.7021087412318737,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.0228516380477504e-05,
+      "loss": 0.3329,
+      "step": 15990
+    },
+    {
+      "epoch": 0.7021965597110772,
+      "grad_norm": 1.4609375,
+      "learning_rate": 1.0222935450628681e-05,
+      "loss": 0.2793,
+      "step": 15992
+    },
+    {
+      "epoch": 0.7022843781902807,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.0217355652411015e-05,
+      "loss": 0.3479,
+      "step": 15994
+    },
+    {
+      "epoch": 0.7023721966694841,
+      "grad_norm": 1.796875,
+      "learning_rate": 1.0211776986251784e-05,
+      "loss": 0.3216,
+      "step": 15996
+    },
+    {
+      "epoch": 0.7024600151486876,
+      "grad_norm": 1.625,
+      "learning_rate": 1.0206199452578228e-05,
+      "loss": 0.3071,
+      "step": 15998
+    },
+    {
+      "epoch": 0.7025478336278912,
+      "grad_norm": 1.625,
+      "learning_rate": 1.0200623051817462e-05,
+      "loss": 0.303,
+      "step": 16000
+    },
+    {
+      "epoch": 0.7026356521070947,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.0195047784396524e-05,
+      "loss": 0.3267,
+      "step": 16002
+    },
+    {
+      "epoch": 0.7027234705862981,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.0189473650742385e-05,
+      "loss": 0.3194,
+      "step": 16004
+    },
+    {
+      "epoch": 0.7028112890655016,
+      "grad_norm": 1.796875,
+      "learning_rate": 1.018390065128189e-05,
+      "loss": 0.3244,
+      "step": 16006
+    },
+    {
+      "epoch": 0.7028991075447051,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.0178328786441848e-05,
+      "loss": 0.3149,
+      "step": 16008
+    },
+    {
+      "epoch": 0.7029869260239086,
+      "grad_norm": 1.4921875,
+      "learning_rate": 1.017275805664894e-05,
+      "loss": 0.3069,
+      "step": 16010
+    },
+    {
+      "epoch": 0.703074744503112,
+      "grad_norm": 1.8046875,
+      "learning_rate": 1.0167188462329767e-05,
+      "loss": 0.3143,
+      "step": 16012
+    },
+    {
+      "epoch": 0.7031625629823156,
+      "grad_norm": 1.765625,
+      "learning_rate": 1.016162000391087e-05,
+      "loss": 0.3319,
+      "step": 16014
+    },
+    {
+      "epoch": 0.7032503814615191,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.0156052681818659e-05,
+      "loss": 0.3057,
+      "step": 16016
+    },
+    {
+      "epoch": 0.7033381999407226,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.0150486496479498e-05,
+      "loss": 0.2966,
+      "step": 16018
+    },
+    {
+      "epoch": 0.703426018419926,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.014492144831963e-05,
+      "loss": 0.3022,
+      "step": 16020
+    },
+    {
+      "epoch": 0.7035138368991295,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.0139357537765249e-05,
+      "loss": 0.2942,
+      "step": 16022
+    },
+    {
+      "epoch": 0.703601655378333,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.013379476524242e-05,
+      "loss": 0.3157,
+      "step": 16024
+    },
+    {
+      "epoch": 0.7036894738575364,
+      "grad_norm": 1.421875,
+      "learning_rate": 1.0128233131177161e-05,
+      "loss": 0.3504,
+      "step": 16026
+    },
+    {
+      "epoch": 0.7037772923367399,
+      "grad_norm": 1.515625,
+      "learning_rate": 1.0122672635995375e-05,
+      "loss": 0.2993,
+      "step": 16028
+    },
+    {
+      "epoch": 0.7038651108159435,
+      "grad_norm": 1.921875,
+      "learning_rate": 1.0117113280122875e-05,
+      "loss": 0.3401,
+      "step": 16030
+    },
+    {
+      "epoch": 0.703952929295147,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.0111555063985418e-05,
+      "loss": 0.3181,
+      "step": 16032
+    },
+    {
+      "epoch": 0.7040407477743504,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.0105997988008631e-05,
+      "loss": 0.3173,
+      "step": 16034
+    },
+    {
+      "epoch": 0.7041285662535539,
+      "grad_norm": 1.4921875,
+      "learning_rate": 1.01004420526181e-05,
+      "loss": 0.3183,
+      "step": 16036
+    },
+    {
+      "epoch": 0.7042163847327574,
+      "grad_norm": 1.4765625,
+      "learning_rate": 1.0094887258239288e-05,
+      "loss": 0.3163,
+      "step": 16038
+    },
+    {
+      "epoch": 0.7043042032119609,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.0089333605297574e-05,
+      "loss": 0.3289,
+      "step": 16040
+    },
+    {
+      "epoch": 0.7043920216911643,
+      "grad_norm": 1.4609375,
+      "learning_rate": 1.0083781094218275e-05,
+      "loss": 0.2794,
+      "step": 16042
+    },
+    {
+      "epoch": 0.7044798401703678,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.0078229725426594e-05,
+      "loss": 0.3269,
+      "step": 16044
+    },
+    {
+      "epoch": 0.7045676586495714,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.0072679499347663e-05,
+      "loss": 0.3037,
+      "step": 16046
+    },
+    {
+      "epoch": 0.7046554771287749,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.00671304164065e-05,
+      "loss": 0.3021,
+      "step": 16048
+    },
+    {
+      "epoch": 0.7047432956079783,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.0061582477028078e-05,
+      "loss": 0.3212,
+      "step": 16050
+    },
+    {
+      "epoch": 0.7048311140871818,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.0056035681637254e-05,
+      "loss": 0.313,
+      "step": 16052
+    },
+    {
+      "epoch": 0.7049189325663853,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.005049003065879e-05,
+      "loss": 0.3348,
+      "step": 16054
+    },
+    {
+      "epoch": 0.7050067510455887,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.0044945524517391e-05,
+      "loss": 0.324,
+      "step": 16056
+    },
+    {
+      "epoch": 0.7050945695247922,
+      "grad_norm": 1.8046875,
+      "learning_rate": 1.003940216363764e-05,
+      "loss": 0.3451,
+      "step": 16058
+    },
+    {
+      "epoch": 0.7051823880039957,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.0033859948444069e-05,
+      "loss": 0.32,
+      "step": 16060
+    },
+    {
+      "epoch": 0.7052702064831993,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.0028318879361087e-05,
+      "loss": 0.3068,
+      "step": 16062
+    },
+    {
+      "epoch": 0.7053580249624027,
+      "grad_norm": 1.515625,
+      "learning_rate": 1.0022778956813028e-05,
+      "loss": 0.3484,
+      "step": 16064
+    },
+    {
+      "epoch": 0.7054458434416062,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.0017240181224155e-05,
+      "loss": 0.325,
+      "step": 16066
+    },
+    {
+      "epoch": 0.7055336619208097,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.0011702553018612e-05,
+      "loss": 0.3345,
+      "step": 16068
+    },
+    {
+      "epoch": 0.7056214804000132,
+      "grad_norm": 1.5,
+      "learning_rate": 1.0006166072620488e-05,
+      "loss": 0.3413,
+      "step": 16070
+    },
+    {
+      "epoch": 0.7057092988792166,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.0000630740453762e-05,
+      "loss": 0.2878,
+      "step": 16072
+    },
+    {
+      "epoch": 0.7057971173584201,
+      "grad_norm": 1.59375,
+      "learning_rate": 9.995096556942319e-06,
+      "loss": 0.3334,
+      "step": 16074
+    },
+    {
+      "epoch": 0.7058849358376237,
+      "grad_norm": 1.5,
+      "learning_rate": 9.989563522509997e-06,
+      "loss": 0.3158,
+      "step": 16076
+    },
+    {
+      "epoch": 0.7059727543168272,
+      "grad_norm": 1.515625,
+      "learning_rate": 9.984031637580483e-06,
+      "loss": 0.2987,
+      "step": 16078
+    },
+    {
+      "epoch": 0.7060605727960306,
+      "grad_norm": 1.5078125,
+      "learning_rate": 9.978500902577432e-06,
+      "loss": 0.3314,
+      "step": 16080
+    },
+    {
+      "epoch": 0.7061483912752341,
+      "grad_norm": 1.6328125,
+      "learning_rate": 9.972971317924374e-06,
+      "loss": 0.3223,
+      "step": 16082
+    },
+    {
+      "epoch": 0.7062362097544376,
+      "grad_norm": 1.6328125,
+      "learning_rate": 9.967442884044784e-06,
+      "loss": 0.3217,
+      "step": 16084
+    },
+    {
+      "epoch": 0.706324028233641,
+      "grad_norm": 1.453125,
+      "learning_rate": 9.961915601362013e-06,
+      "loss": 0.3206,
+      "step": 16086
+    },
+    {
+      "epoch": 0.7064118467128445,
+      "grad_norm": 1.4453125,
+      "learning_rate": 9.95638947029936e-06,
+      "loss": 0.3122,
+      "step": 16088
+    },
+    {
+      "epoch": 0.706499665192048,
+      "grad_norm": 1.4765625,
+      "learning_rate": 9.950864491280004e-06,
+      "loss": 0.2964,
+      "step": 16090
+    },
+    {
+      "epoch": 0.7065874836712516,
+      "grad_norm": 1.6015625,
+      "learning_rate": 9.945340664727048e-06,
+      "loss": 0.3622,
+      "step": 16092
+    },
+    {
+      "epoch": 0.706675302150455,
+      "grad_norm": 1.421875,
+      "learning_rate": 9.939817991063518e-06,
+      "loss": 0.3222,
+      "step": 16094
+    },
+    {
+      "epoch": 0.7067631206296585,
+      "grad_norm": 1.671875,
+      "learning_rate": 9.934296470712331e-06,
+      "loss": 0.307,
+      "step": 16096
+    },
+    {
+      "epoch": 0.706850939108862,
+      "grad_norm": 1.609375,
+      "learning_rate": 9.928776104096338e-06,
+      "loss": 0.298,
+      "step": 16098
+    },
+    {
+      "epoch": 0.7069387575880655,
+      "grad_norm": 1.5,
+      "learning_rate": 9.923256891638285e-06,
+      "loss": 0.3036,
+      "step": 16100
+    },
+    {
+      "epoch": 0.7070265760672689,
+      "grad_norm": 1.5234375,
+      "learning_rate": 9.917738833760826e-06,
+      "loss": 0.3059,
+      "step": 16102
+    },
+    {
+      "epoch": 0.7071143945464724,
+      "grad_norm": 1.4921875,
+      "learning_rate": 9.91222193088655e-06,
+      "loss": 0.325,
+      "step": 16104
+    },
+    {
+      "epoch": 0.7072022130256759,
+      "grad_norm": 1.796875,
+      "learning_rate": 9.906706183437933e-06,
+      "loss": 0.3055,
+      "step": 16106
+    },
+    {
+      "epoch": 0.7072900315048795,
+      "grad_norm": 1.5546875,
+      "learning_rate": 9.901191591837378e-06,
+      "loss": 0.3273,
+      "step": 16108
+    },
+    {
+      "epoch": 0.7073778499840829,
+      "grad_norm": 1.53125,
+      "learning_rate": 9.89567815650718e-06,
+      "loss": 0.2972,
+      "step": 16110
+    },
+    {
+      "epoch": 0.7074656684632864,
+      "grad_norm": 1.5234375,
+      "learning_rate": 9.890165877869578e-06,
+      "loss": 0.3642,
+      "step": 16112
+    },
+    {
+      "epoch": 0.7075534869424899,
+      "grad_norm": 1.4375,
+      "learning_rate": 9.884654756346698e-06,
+      "loss": 0.3163,
+      "step": 16114
+    },
+    {
+      "epoch": 0.7076413054216933,
+      "grad_norm": 1.5859375,
+      "learning_rate": 9.879144792360567e-06,
+      "loss": 0.3013,
+      "step": 16116
+    },
+    {
+      "epoch": 0.7077291239008968,
+      "grad_norm": 1.5234375,
+      "learning_rate": 9.873635986333162e-06,
+      "loss": 0.3158,
+      "step": 16118
+    },
+    {
+      "epoch": 0.7078169423801003,
+      "grad_norm": 1.53125,
+      "learning_rate": 9.868128338686334e-06,
+      "loss": 0.3253,
+      "step": 16120
+    },
+    {
+      "epoch": 0.7079047608593039,
+      "grad_norm": 1.609375,
+      "learning_rate": 9.862621849841871e-06,
+      "loss": 0.3257,
+      "step": 16122
+    },
+    {
+      "epoch": 0.7079925793385073,
+      "grad_norm": 1.4140625,
+      "learning_rate": 9.857116520221457e-06,
+      "loss": 0.3153,
+      "step": 16124
+    },
+    {
+      "epoch": 0.7080803978177108,
+      "grad_norm": 1.5703125,
+      "learning_rate": 9.85161235024668e-06,
+      "loss": 0.3007,
+      "step": 16126
+    },
+    {
+      "epoch": 0.7081682162969143,
+      "grad_norm": 1.6875,
+      "learning_rate": 9.846109340339068e-06,
+      "loss": 0.334,
+      "step": 16128
+    },
+    {
+      "epoch": 0.7082560347761178,
+      "grad_norm": 1.59375,
+      "learning_rate": 9.840607490920031e-06,
+      "loss": 0.3236,
+      "step": 16130
+    },
+    {
+      "epoch": 0.7083438532553212,
+      "grad_norm": 1.578125,
+      "learning_rate": 9.835106802410913e-06,
+      "loss": 0.2867,
+      "step": 16132
+    },
+    {
+      "epoch": 0.7084316717345247,
+      "grad_norm": 1.5390625,
+      "learning_rate": 9.829607275232949e-06,
+      "loss": 0.294,
+      "step": 16134
+    },
+    {
+      "epoch": 0.7085194902137282,
+      "grad_norm": 1.4453125,
+      "learning_rate": 9.824108909807297e-06,
+      "loss": 0.3177,
+      "step": 16136
+    },
+    {
+      "epoch": 0.7086073086929318,
+      "grad_norm": 1.5078125,
+      "learning_rate": 9.818611706555026e-06,
+      "loss": 0.3327,
+      "step": 16138
+    },
+    {
+      "epoch": 0.7086951271721352,
+      "grad_norm": 1.5078125,
+      "learning_rate": 9.813115665897096e-06,
+      "loss": 0.3253,
+      "step": 16140
+    },
+    {
+      "epoch": 0.7087829456513387,
+      "grad_norm": 1.5234375,
+      "learning_rate": 9.807620788254421e-06,
+      "loss": 0.3269,
+      "step": 16142
+    },
+    {
+      "epoch": 0.7088707641305422,
+      "grad_norm": 1.5078125,
+      "learning_rate": 9.802127074047779e-06,
+      "loss": 0.3377,
+      "step": 16144
+    },
+    {
+      "epoch": 0.7089585826097456,
+      "grad_norm": 1.65625,
+      "learning_rate": 9.796634523697898e-06,
+      "loss": 0.3253,
+      "step": 16146
+    },
+    {
+      "epoch": 0.7090464010889491,
+      "grad_norm": 1.5859375,
+      "learning_rate": 9.79114313762539e-06,
+      "loss": 0.3284,
+      "step": 16148
+    },
+    {
+      "epoch": 0.7091342195681526,
+      "grad_norm": 1.6015625,
+      "learning_rate": 9.785652916250773e-06,
+      "loss": 0.3334,
+      "step": 16150
+    },
+    {
+      "epoch": 0.7092220380473561,
+      "grad_norm": 1.5859375,
+      "learning_rate": 9.780163859994515e-06,
+      "loss": 0.2905,
+      "step": 16152
+    },
+    {
+      "epoch": 0.7093098565265596,
+      "grad_norm": 1.5546875,
+      "learning_rate": 9.77467596927695e-06,
+      "loss": 0.3026,
+      "step": 16154
+    },
+    {
+      "epoch": 0.7093976750057631,
+      "grad_norm": 1.578125,
+      "learning_rate": 9.769189244518354e-06,
+      "loss": 0.3614,
+      "step": 16156
+    },
+    {
+      "epoch": 0.7094854934849666,
+      "grad_norm": 1.6484375,
+      "learning_rate": 9.763703686138892e-06,
+      "loss": 0.3132,
+      "step": 16158
+    },
+    {
+      "epoch": 0.7095733119641701,
+      "grad_norm": 1.5625,
+      "learning_rate": 9.75821929455866e-06,
+      "loss": 0.2955,
+      "step": 16160
+    },
+    {
+      "epoch": 0.7096611304433735,
+      "grad_norm": 1.6484375,
+      "learning_rate": 9.75273607019765e-06,
+      "loss": 0.3125,
+      "step": 16162
+    },
+    {
+      "epoch": 0.709748948922577,
+      "grad_norm": 1.5,
+      "learning_rate": 9.747254013475754e-06,
+      "loss": 0.3066,
+      "step": 16164
+    },
+    {
+      "epoch": 0.7098367674017805,
+      "grad_norm": 1.78125,
+      "learning_rate": 9.741773124812814e-06,
+      "loss": 0.3302,
+      "step": 16166
+    },
+    {
+      "epoch": 0.7099245858809841,
+      "grad_norm": 1.5703125,
+      "learning_rate": 9.736293404628546e-06,
+      "loss": 0.3334,
+      "step": 16168
+    },
+    {
+      "epoch": 0.7100124043601875,
+      "grad_norm": 1.4765625,
+      "learning_rate": 9.730814853342587e-06,
+      "loss": 0.3064,
+      "step": 16170
+    },
+    {
+      "epoch": 0.710100222839391,
+      "grad_norm": 1.5546875,
+      "learning_rate": 9.72533747137448e-06,
+      "loss": 0.3152,
+      "step": 16172
+    },
+    {
+      "epoch": 0.7101880413185945,
+      "grad_norm": 1.578125,
+      "learning_rate": 9.719861259143698e-06,
+      "loss": 0.3324,
+      "step": 16174
+    },
+    {
+      "epoch": 0.710275859797798,
+      "grad_norm": 1.7578125,
+      "learning_rate": 9.714386217069604e-06,
+      "loss": 0.347,
+      "step": 16176
+    },
+    {
+      "epoch": 0.7103636782770014,
+      "grad_norm": 1.625,
+      "learning_rate": 9.708912345571469e-06,
+      "loss": 0.3197,
+      "step": 16178
+    },
+    {
+      "epoch": 0.7104514967562049,
+      "grad_norm": 1.515625,
+      "learning_rate": 9.7034396450685e-06,
+      "loss": 0.3114,
+      "step": 16180
+    },
+    {
+      "epoch": 0.7105393152354084,
+      "grad_norm": 1.703125,
+      "learning_rate": 9.69796811597978e-06,
+      "loss": 0.3416,
+      "step": 16182
+    },
+    {
+      "epoch": 0.710627133714612,
+      "grad_norm": 1.8125,
+      "learning_rate": 9.692497758724342e-06,
+      "loss": 0.3315,
+      "step": 16184
+    },
+    {
+      "epoch": 0.7107149521938154,
+      "grad_norm": 1.5,
+      "learning_rate": 9.687028573721094e-06,
+      "loss": 0.3204,
+      "step": 16186
+    },
+    {
+      "epoch": 0.7108027706730189,
+      "grad_norm": 1.5859375,
+      "learning_rate": 9.681560561388858e-06,
+      "loss": 0.3229,
+      "step": 16188
+    },
+    {
+      "epoch": 0.7108905891522224,
+      "grad_norm": 1.5390625,
+      "learning_rate": 9.676093722146399e-06,
+      "loss": 0.3268,
+      "step": 16190
+    },
+    {
+      "epoch": 0.7109784076314258,
+      "grad_norm": 1.5703125,
+      "learning_rate": 9.670628056412342e-06,
+      "loss": 0.2892,
+      "step": 16192
+    },
+    {
+      "epoch": 0.7110662261106293,
+      "grad_norm": 1.5234375,
+      "learning_rate": 9.665163564605275e-06,
+      "loss": 0.3335,
+      "step": 16194
+    },
+    {
+      "epoch": 0.7111540445898328,
+      "grad_norm": 1.71875,
+      "learning_rate": 9.659700247143658e-06,
+      "loss": 0.3094,
+      "step": 16196
+    },
+    {
+      "epoch": 0.7112418630690363,
+      "grad_norm": 1.515625,
+      "learning_rate": 9.654238104445873e-06,
+      "loss": 0.3172,
+      "step": 16198
+    },
+    {
+      "epoch": 0.7113296815482398,
+      "grad_norm": 1.546875,
+      "learning_rate": 9.648777136930215e-06,
+      "loss": 0.3063,
+      "step": 16200
+    },
+    {
+      "epoch": 0.7114175000274433,
+      "grad_norm": 1.53125,
+      "learning_rate": 9.64331734501487e-06,
+      "loss": 0.3165,
+      "step": 16202
+    },
+    {
+      "epoch": 0.7115053185066468,
+      "grad_norm": 1.5,
+      "learning_rate": 9.637858729117977e-06,
+      "loss": 0.3141,
+      "step": 16204
+    },
+    {
+      "epoch": 0.7115931369858502,
+      "grad_norm": 1.5078125,
+      "learning_rate": 9.632401289657537e-06,
+      "loss": 0.3346,
+      "step": 16206
+    },
+    {
+      "epoch": 0.7116809554650537,
+      "grad_norm": 1.59375,
+      "learning_rate": 9.626945027051495e-06,
+      "loss": 0.3514,
+      "step": 16208
+    },
+    {
+      "epoch": 0.7117687739442572,
+      "grad_norm": 1.5625,
+      "learning_rate": 9.621489941717691e-06,
+      "loss": 0.3265,
+      "step": 16210
+    },
+    {
+      "epoch": 0.7118565924234607,
+      "grad_norm": 1.546875,
+      "learning_rate": 9.616036034073863e-06,
+      "loss": 0.3285,
+      "step": 16212
+    },
+    {
+      "epoch": 0.7119444109026641,
+      "grad_norm": 1.5703125,
+      "learning_rate": 9.610583304537693e-06,
+      "loss": 0.3249,
+      "step": 16214
+    },
+    {
+      "epoch": 0.7120322293818677,
+      "grad_norm": 1.4765625,
+      "learning_rate": 9.605131753526733e-06,
+      "loss": 0.3098,
+      "step": 16216
+    },
+    {
+      "epoch": 0.7121200478610712,
+      "grad_norm": 1.546875,
+      "learning_rate": 9.599681381458483e-06,
+      "loss": 0.3362,
+      "step": 16218
+    },
+    {
+      "epoch": 0.7122078663402747,
+      "grad_norm": 1.6328125,
+      "learning_rate": 9.594232188750316e-06,
+      "loss": 0.3121,
+      "step": 16220
+    },
+    {
+      "epoch": 0.7122956848194781,
+      "grad_norm": 1.546875,
+      "learning_rate": 9.58878417581955e-06,
+      "loss": 0.3073,
+      "step": 16222
+    },
+    {
+      "epoch": 0.7123835032986816,
+      "grad_norm": 1.515625,
+      "learning_rate": 9.583337343083387e-06,
+      "loss": 0.3131,
+      "step": 16224
+    },
+    {
+      "epoch": 0.7124713217778851,
+      "grad_norm": 1.484375,
+      "learning_rate": 9.577891690958935e-06,
+      "loss": 0.2908,
+      "step": 16226
+    },
+    {
+      "epoch": 0.7125591402570886,
+      "grad_norm": 1.578125,
+      "learning_rate": 9.572447219863253e-06,
+      "loss": 0.3163,
+      "step": 16228
+    },
+    {
+      "epoch": 0.7126469587362921,
+      "grad_norm": 1.640625,
+      "learning_rate": 9.567003930213241e-06,
+      "loss": 0.3299,
+      "step": 16230
+    },
+    {
+      "epoch": 0.7127347772154956,
+      "grad_norm": 1.484375,
+      "learning_rate": 9.56156182242578e-06,
+      "loss": 0.3156,
+      "step": 16232
+    },
+    {
+      "epoch": 0.7128225956946991,
+      "grad_norm": 1.5859375,
+      "learning_rate": 9.556120896917605e-06,
+      "loss": 0.3299,
+      "step": 16234
+    },
+    {
+      "epoch": 0.7129104141739026,
+      "grad_norm": 1.4921875,
+      "learning_rate": 9.550681154105403e-06,
+      "loss": 0.3117,
+      "step": 16236
+    },
+    {
+      "epoch": 0.712998232653106,
+      "grad_norm": 1.5625,
+      "learning_rate": 9.545242594405743e-06,
+      "loss": 0.295,
+      "step": 16238
+    },
+    {
+      "epoch": 0.7130860511323095,
+      "grad_norm": 1.5625,
+      "learning_rate": 9.539805218235101e-06,
+      "loss": 0.3388,
+      "step": 16240
+    },
+    {
+      "epoch": 0.713173869611513,
+      "grad_norm": 1.546875,
+      "learning_rate": 9.534369026009888e-06,
+      "loss": 0.3074,
+      "step": 16242
+    },
+    {
+      "epoch": 0.7132616880907164,
+      "grad_norm": 1.625,
+      "learning_rate": 9.528934018146396e-06,
+      "loss": 0.3924,
+      "step": 16244
+    },
+    {
+      "epoch": 0.71334950656992,
+      "grad_norm": 1.515625,
+      "learning_rate": 9.523500195060852e-06,
+      "loss": 0.308,
+      "step": 16246
+    },
+    {
+      "epoch": 0.7134373250491235,
+      "grad_norm": 1.5546875,
+      "learning_rate": 9.518067557169375e-06,
+      "loss": 0.3017,
+      "step": 16248
+    },
+    {
+      "epoch": 0.713525143528327,
+      "grad_norm": 1.5078125,
+      "learning_rate": 9.512636104887984e-06,
+      "loss": 0.3188,
+      "step": 16250
+    },
+    {
+      "epoch": 0.7136129620075304,
+      "grad_norm": 1.578125,
+      "learning_rate": 9.507205838632643e-06,
+      "loss": 0.3262,
+      "step": 16252
+    },
+    {
+      "epoch": 0.7137007804867339,
+      "grad_norm": 1.5625,
+      "learning_rate": 9.501776758819186e-06,
+      "loss": 0.3345,
+      "step": 16254
+    },
+    {
+      "epoch": 0.7137885989659374,
+      "grad_norm": 1.5078125,
+      "learning_rate": 9.496348865863386e-06,
+      "loss": 0.3339,
+      "step": 16256
+    },
+    {
+      "epoch": 0.7138764174451409,
+      "grad_norm": 1.5390625,
+      "learning_rate": 9.49092216018091e-06,
+      "loss": 0.3174,
+      "step": 16258
+    },
+    {
+      "epoch": 0.7139642359243443,
+      "grad_norm": 1.6015625,
+      "learning_rate": 9.485496642187328e-06,
+      "loss": 0.3242,
+      "step": 16260
+    },
+    {
+      "epoch": 0.7140520544035479,
+      "grad_norm": 1.5625,
+      "learning_rate": 9.480072312298135e-06,
+      "loss": 0.3445,
+      "step": 16262
+    },
+    {
+      "epoch": 0.7141398728827514,
+      "grad_norm": 1.671875,
+      "learning_rate": 9.474649170928714e-06,
+      "loss": 0.3035,
+      "step": 16264
+    },
+    {
+      "epoch": 0.7142276913619549,
+      "grad_norm": 1.5,
+      "learning_rate": 9.469227218494391e-06,
+      "loss": 0.305,
+      "step": 16266
+    },
+    {
+      "epoch": 0.7143155098411583,
+      "grad_norm": 1.71875,
+      "learning_rate": 9.463806455410365e-06,
+      "loss": 0.3175,
+      "step": 16268
+    },
+    {
+      "epoch": 0.7144033283203618,
+      "grad_norm": 1.5859375,
+      "learning_rate": 9.458386882091769e-06,
+      "loss": 0.3037,
+      "step": 16270
+    },
+    {
+      "epoch": 0.7144911467995653,
+      "grad_norm": 1.53125,
+      "learning_rate": 9.452968498953634e-06,
+      "loss": 0.3173,
+      "step": 16272
+    },
+    {
+      "epoch": 0.7145789652787687,
+      "grad_norm": 1.8125,
+      "learning_rate": 9.447551306410887e-06,
+      "loss": 0.3008,
+      "step": 16274
+    },
+    {
+      "epoch": 0.7146667837579723,
+      "grad_norm": 1.515625,
+      "learning_rate": 9.442135304878403e-06,
+      "loss": 0.31,
+      "step": 16276
+    },
+    {
+      "epoch": 0.7147546022371758,
+      "grad_norm": 1.546875,
+      "learning_rate": 9.436720494770912e-06,
+      "loss": 0.3117,
+      "step": 16278
+    },
+    {
+      "epoch": 0.7148424207163793,
+      "grad_norm": 1.5390625,
+      "learning_rate": 9.431306876503108e-06,
+      "loss": 0.3323,
+      "step": 16280
+    },
+    {
+      "epoch": 0.7149302391955827,
+      "grad_norm": 1.6171875,
+      "learning_rate": 9.425894450489556e-06,
+      "loss": 0.3338,
+      "step": 16282
+    },
+    {
+      "epoch": 0.7150180576747862,
+      "grad_norm": 1.53125,
+      "learning_rate": 9.420483217144729e-06,
+      "loss": 0.3296,
+      "step": 16284
+    },
+    {
+      "epoch": 0.7151058761539897,
+      "grad_norm": 1.5078125,
+      "learning_rate": 9.415073176883043e-06,
+      "loss": 0.3117,
+      "step": 16286
+    },
+    {
+      "epoch": 0.7151936946331932,
+      "grad_norm": 1.5859375,
+      "learning_rate": 9.409664330118778e-06,
+      "loss": 0.3294,
+      "step": 16288
+    },
+    {
+      "epoch": 0.7152815131123966,
+      "grad_norm": 1.65625,
+      "learning_rate": 9.404256677266176e-06,
+      "loss": 0.2995,
+      "step": 16290
+    },
+    {
+      "epoch": 0.7153693315916002,
+      "grad_norm": 1.5859375,
+      "learning_rate": 9.398850218739319e-06,
+      "loss": 0.3314,
+      "step": 16292
+    },
+    {
+      "epoch": 0.7154571500708037,
+      "grad_norm": 1.5078125,
+      "learning_rate": 9.393444954952257e-06,
+      "loss": 0.3079,
+      "step": 16294
+    },
+    {
+      "epoch": 0.7155449685500072,
+      "grad_norm": 1.5703125,
+      "learning_rate": 9.388040886318916e-06,
+      "loss": 0.3207,
+      "step": 16296
+    },
+    {
+      "epoch": 0.7156327870292106,
+      "grad_norm": 1.578125,
+      "learning_rate": 9.382638013253156e-06,
+      "loss": 0.3263,
+      "step": 16298
+    },
+    {
+      "epoch": 0.7157206055084141,
+      "grad_norm": 1.4765625,
+      "learning_rate": 9.377236336168717e-06,
+      "loss": 0.2916,
+      "step": 16300
+    },
+    {
+      "epoch": 0.7158084239876176,
+      "grad_norm": 1.546875,
+      "learning_rate": 9.371835855479258e-06,
+      "loss": 0.3527,
+      "step": 16302
+    },
+    {
+      "epoch": 0.715896242466821,
+      "grad_norm": 1.5703125,
+      "learning_rate": 9.366436571598364e-06,
+      "loss": 0.2921,
+      "step": 16304
+    },
+    {
+      "epoch": 0.7159840609460245,
+      "grad_norm": 1.5703125,
+      "learning_rate": 9.361038484939496e-06,
+      "loss": 0.3471,
+      "step": 16306
+    },
+    {
+      "epoch": 0.7160718794252281,
+      "grad_norm": 1.609375,
+      "learning_rate": 9.355641595916059e-06,
+      "loss": 0.3398,
+      "step": 16308
+    },
+    {
+      "epoch": 0.7161596979044316,
+      "grad_norm": 1.5,
+      "learning_rate": 9.350245904941338e-06,
+      "loss": 0.3452,
+      "step": 16310
+    },
+    {
+      "epoch": 0.716247516383635,
+      "grad_norm": 1.5625,
+      "learning_rate": 9.34485141242853e-06,
+      "loss": 0.3264,
+      "step": 16312
+    },
+    {
+      "epoch": 0.7163353348628385,
+      "grad_norm": 1.5390625,
+      "learning_rate": 9.339458118790761e-06,
+      "loss": 0.2958,
+      "step": 16314
+    },
+    {
+      "epoch": 0.716423153342042,
+      "grad_norm": 1.7109375,
+      "learning_rate": 9.334066024441035e-06,
+      "loss": 0.3358,
+      "step": 16316
+    },
+    {
+      "epoch": 0.7165109718212455,
+      "grad_norm": 1.6875,
+      "learning_rate": 9.328675129792298e-06,
+      "loss": 0.318,
+      "step": 16318
+    },
+    {
+      "epoch": 0.7165987903004489,
+      "grad_norm": 1.703125,
+      "learning_rate": 9.323285435257373e-06,
+      "loss": 0.3243,
+      "step": 16320
+    },
+    {
+      "epoch": 0.7166866087796525,
+      "grad_norm": 1.640625,
+      "learning_rate": 9.31789694124901e-06,
+      "loss": 0.3271,
+      "step": 16322
+    },
+    {
+      "epoch": 0.716774427258856,
+      "grad_norm": 1.53125,
+      "learning_rate": 9.312509648179856e-06,
+      "loss": 0.3216,
+      "step": 16324
+    },
+    {
+      "epoch": 0.7168622457380595,
+      "grad_norm": 1.4765625,
+      "learning_rate": 9.30712355646247e-06,
+      "loss": 0.3215,
+      "step": 16326
+    },
+    {
+      "epoch": 0.7169500642172629,
+      "grad_norm": 1.5390625,
+      "learning_rate": 9.301738666509327e-06,
+      "loss": 0.2933,
+      "step": 16328
+    },
+    {
+      "epoch": 0.7170378826964664,
+      "grad_norm": 1.78125,
+      "learning_rate": 9.296354978732793e-06,
+      "loss": 0.3125,
+      "step": 16330
+    },
+    {
+      "epoch": 0.7171257011756699,
+      "grad_norm": 1.515625,
+      "learning_rate": 9.29097249354517e-06,
+      "loss": 0.3495,
+      "step": 16332
+    },
+    {
+      "epoch": 0.7172135196548733,
+      "grad_norm": 1.6796875,
+      "learning_rate": 9.285591211358637e-06,
+      "loss": 0.3285,
+      "step": 16334
+    },
+    {
+      "epoch": 0.7173013381340768,
+      "grad_norm": 1.4921875,
+      "learning_rate": 9.28021113258529e-06,
+      "loss": 0.3444,
+      "step": 16336
+    },
+    {
+      "epoch": 0.7173891566132804,
+      "grad_norm": 1.6171875,
+      "learning_rate": 9.274832257637148e-06,
+      "loss": 0.3218,
+      "step": 16338
+    },
+    {
+      "epoch": 0.7174769750924839,
+      "grad_norm": 1.5,
+      "learning_rate": 9.269454586926116e-06,
+      "loss": 0.3116,
+      "step": 16340
+    },
+    {
+      "epoch": 0.7175647935716873,
+      "grad_norm": 1.5625,
+      "learning_rate": 9.264078120864029e-06,
+      "loss": 0.3402,
+      "step": 16342
+    },
+    {
+      "epoch": 0.7176526120508908,
+      "grad_norm": 1.546875,
+      "learning_rate": 9.258702859862612e-06,
+      "loss": 0.3325,
+      "step": 16344
+    },
+    {
+      "epoch": 0.7177404305300943,
+      "grad_norm": 1.6484375,
+      "learning_rate": 9.253328804333495e-06,
+      "loss": 0.3402,
+      "step": 16346
+    },
+    {
+      "epoch": 0.7178282490092978,
+      "grad_norm": 1.5234375,
+      "learning_rate": 9.247955954688242e-06,
+      "loss": 0.349,
+      "step": 16348
+    },
+    {
+      "epoch": 0.7179160674885012,
+      "grad_norm": 1.59375,
+      "learning_rate": 9.242584311338288e-06,
+      "loss": 0.3494,
+      "step": 16350
+    },
+    {
+      "epoch": 0.7180038859677047,
+      "grad_norm": 1.5625,
+      "learning_rate": 9.237213874695021e-06,
+      "loss": 0.3199,
+      "step": 16352
+    },
+    {
+      "epoch": 0.7180917044469083,
+      "grad_norm": 1.6484375,
+      "learning_rate": 9.231844645169679e-06,
+      "loss": 0.3132,
+      "step": 16354
+    },
+    {
+      "epoch": 0.7181795229261118,
+      "grad_norm": 1.515625,
+      "learning_rate": 9.226476623173464e-06,
+      "loss": 0.3385,
+      "step": 16356
+    },
+    {
+      "epoch": 0.7182673414053152,
+      "grad_norm": 1.6015625,
+      "learning_rate": 9.22110980911744e-06,
+      "loss": 0.3345,
+      "step": 16358
+    },
+    {
+      "epoch": 0.7183551598845187,
+      "grad_norm": 1.46875,
+      "learning_rate": 9.215744203412619e-06,
+      "loss": 0.3218,
+      "step": 16360
+    },
+    {
+      "epoch": 0.7184429783637222,
+      "grad_norm": 1.5546875,
+      "learning_rate": 9.210379806469888e-06,
+      "loss": 0.317,
+      "step": 16362
+    },
+    {
+      "epoch": 0.7185307968429256,
+      "grad_norm": 1.5234375,
+      "learning_rate": 9.205016618700049e-06,
+      "loss": 0.3086,
+      "step": 16364
+    },
+    {
+      "epoch": 0.7186186153221291,
+      "grad_norm": 1.5,
+      "learning_rate": 9.199654640513833e-06,
+      "loss": 0.3399,
+      "step": 16366
+    },
+    {
+      "epoch": 0.7187064338013327,
+      "grad_norm": 1.484375,
+      "learning_rate": 9.194293872321843e-06,
+      "loss": 0.3437,
+      "step": 16368
+    },
+    {
+      "epoch": 0.7187942522805362,
+      "grad_norm": 1.59375,
+      "learning_rate": 9.188934314534625e-06,
+      "loss": 0.3387,
+      "step": 16370
+    },
+    {
+      "epoch": 0.7188820707597396,
+      "grad_norm": 1.59375,
+      "learning_rate": 9.18357596756261e-06,
+      "loss": 0.2981,
+      "step": 16372
+    },
+    {
+      "epoch": 0.7189698892389431,
+      "grad_norm": 1.5390625,
+      "learning_rate": 9.178218831816126e-06,
+      "loss": 0.315,
+      "step": 16374
+    },
+    {
+      "epoch": 0.7190577077181466,
+      "grad_norm": 1.5859375,
+      "learning_rate": 9.17286290770545e-06,
+      "loss": 0.2909,
+      "step": 16376
+    },
+    {
+      "epoch": 0.7191455261973501,
+      "grad_norm": 1.4921875,
+      "learning_rate": 9.167508195640714e-06,
+      "loss": 0.3028,
+      "step": 16378
+    },
+    {
+      "epoch": 0.7192333446765535,
+      "grad_norm": 1.5078125,
+      "learning_rate": 9.162154696032007e-06,
+      "loss": 0.3327,
+      "step": 16380
+    },
+    {
+      "epoch": 0.719321163155757,
+      "grad_norm": 1.5625,
+      "learning_rate": 9.156802409289289e-06,
+      "loss": 0.3227,
+      "step": 16382
+    },
+    {
+      "epoch": 0.7194089816349606,
+      "grad_norm": 1.5625,
+      "learning_rate": 9.151451335822442e-06,
+      "loss": 0.311,
+      "step": 16384
+    },
+    {
+      "epoch": 0.7194968001141641,
+      "grad_norm": 1.484375,
+      "learning_rate": 9.146101476041249e-06,
+      "loss": 0.3146,
+      "step": 16386
+    },
+    {
+      "epoch": 0.7195846185933675,
+      "grad_norm": 1.65625,
+      "learning_rate": 9.140752830355395e-06,
+      "loss": 0.3515,
+      "step": 16388
+    },
+    {
+      "epoch": 0.719672437072571,
+      "grad_norm": 1.703125,
+      "learning_rate": 9.135405399174504e-06,
+      "loss": 0.3435,
+      "step": 16390
+    },
+    {
+      "epoch": 0.7197602555517745,
+      "grad_norm": 1.4921875,
+      "learning_rate": 9.13005918290806e-06,
+      "loss": 0.3042,
+      "step": 16392
+    },
+    {
+      "epoch": 0.719848074030978,
+      "grad_norm": 1.6328125,
+      "learning_rate": 9.124714181965497e-06,
+      "loss": 0.3166,
+      "step": 16394
+    },
+    {
+      "epoch": 0.7199358925101814,
+      "grad_norm": 1.5859375,
+      "learning_rate": 9.119370396756125e-06,
+      "loss": 0.3493,
+      "step": 16396
+    },
+    {
+      "epoch": 0.7200237109893849,
+      "grad_norm": 1.6328125,
+      "learning_rate": 9.114027827689168e-06,
+      "loss": 0.3101,
+      "step": 16398
+    },
+    {
+      "epoch": 0.7201115294685885,
+      "grad_norm": 1.671875,
+      "learning_rate": 9.108686475173777e-06,
+      "loss": 0.3509,
+      "step": 16400
+    },
+    {
+      "epoch": 0.720199347947792,
+      "grad_norm": 1.5390625,
+      "learning_rate": 9.103346339618975e-06,
+      "loss": 0.3276,
+      "step": 16402
+    },
+    {
+      "epoch": 0.7202871664269954,
+      "grad_norm": 1.515625,
+      "learning_rate": 9.098007421433732e-06,
+      "loss": 0.3511,
+      "step": 16404
+    },
+    {
+      "epoch": 0.7203749849061989,
+      "grad_norm": 1.8125,
+      "learning_rate": 9.092669721026892e-06,
+      "loss": 0.3317,
+      "step": 16406
+    },
+    {
+      "epoch": 0.7204628033854024,
+      "grad_norm": 1.5390625,
+      "learning_rate": 9.087333238807208e-06,
+      "loss": 0.3172,
+      "step": 16408
+    },
+    {
+      "epoch": 0.7205506218646058,
+      "grad_norm": 1.53125,
+      "learning_rate": 9.081997975183368e-06,
+      "loss": 0.3195,
+      "step": 16410
+    },
+    {
+      "epoch": 0.7206384403438093,
+      "grad_norm": 1.6328125,
+      "learning_rate": 9.07666393056394e-06,
+      "loss": 0.3213,
+      "step": 16412
+    },
+    {
+      "epoch": 0.7207262588230128,
+      "grad_norm": 1.6328125,
+      "learning_rate": 9.071331105357406e-06,
+      "loss": 0.3312,
+      "step": 16414
+    },
+    {
+      "epoch": 0.7208140773022164,
+      "grad_norm": 1.515625,
+      "learning_rate": 9.065999499972144e-06,
+      "loss": 0.3171,
+      "step": 16416
+    },
+    {
+      "epoch": 0.7209018957814198,
+      "grad_norm": 1.5,
+      "learning_rate": 9.06066911481647e-06,
+      "loss": 0.3196,
+      "step": 16418
+    },
+    {
+      "epoch": 0.7209897142606233,
+      "grad_norm": 1.625,
+      "learning_rate": 9.055339950298564e-06,
+      "loss": 0.3272,
+      "step": 16420
+    },
+    {
+      "epoch": 0.7210775327398268,
+      "grad_norm": 1.5859375,
+      "learning_rate": 9.050012006826558e-06,
+      "loss": 0.2998,
+      "step": 16422
+    },
+    {
+      "epoch": 0.7211653512190302,
+      "grad_norm": 1.625,
+      "learning_rate": 9.044685284808458e-06,
+      "loss": 0.3191,
+      "step": 16424
+    },
+    {
+      "epoch": 0.7212531696982337,
+      "grad_norm": 1.5546875,
+      "learning_rate": 9.03935978465217e-06,
+      "loss": 0.3121,
+      "step": 16426
+    },
+    {
+      "epoch": 0.7213409881774372,
+      "grad_norm": 1.5859375,
+      "learning_rate": 9.034035506765548e-06,
+      "loss": 0.3239,
+      "step": 16428
+    },
+    {
+      "epoch": 0.7214288066566408,
+      "grad_norm": 1.5703125,
+      "learning_rate": 9.028712451556307e-06,
+      "loss": 0.2768,
+      "step": 16430
+    },
+    {
+      "epoch": 0.7215166251358442,
+      "grad_norm": 1.53125,
+      "learning_rate": 9.023390619432101e-06,
+      "loss": 0.3418,
+      "step": 16432
+    },
+    {
+      "epoch": 0.7216044436150477,
+      "grad_norm": 1.59375,
+      "learning_rate": 9.018070010800472e-06,
+      "loss": 0.3248,
+      "step": 16434
+    },
+    {
+      "epoch": 0.7216922620942512,
+      "grad_norm": 1.53125,
+      "learning_rate": 9.012750626068864e-06,
+      "loss": 0.2989,
+      "step": 16436
+    },
+    {
+      "epoch": 0.7217800805734547,
+      "grad_norm": 1.578125,
+      "learning_rate": 9.007432465644652e-06,
+      "loss": 0.3299,
+      "step": 16438
+    },
+    {
+      "epoch": 0.7218678990526581,
+      "grad_norm": 1.6328125,
+      "learning_rate": 9.00211552993509e-06,
+      "loss": 0.3024,
+      "step": 16440
+    },
+    {
+      "epoch": 0.7219557175318616,
+      "grad_norm": 1.609375,
+      "learning_rate": 8.996799819347363e-06,
+      "loss": 0.3125,
+      "step": 16442
+    },
+    {
+      "epoch": 0.7220435360110651,
+      "grad_norm": 1.6796875,
+      "learning_rate": 8.991485334288542e-06,
+      "loss": 0.3349,
+      "step": 16444
+    },
+    {
+      "epoch": 0.7221313544902687,
+      "grad_norm": 1.578125,
+      "learning_rate": 8.986172075165611e-06,
+      "loss": 0.3068,
+      "step": 16446
+    },
+    {
+      "epoch": 0.7222191729694721,
+      "grad_norm": 1.59375,
+      "learning_rate": 8.98086004238546e-06,
+      "loss": 0.3324,
+      "step": 16448
+    },
+    {
+      "epoch": 0.7223069914486756,
+      "grad_norm": 1.6328125,
+      "learning_rate": 8.975549236354882e-06,
+      "loss": 0.3361,
+      "step": 16450
+    },
+    {
+      "epoch": 0.7223948099278791,
+      "grad_norm": 1.6015625,
+      "learning_rate": 8.970239657480592e-06,
+      "loss": 0.3373,
+      "step": 16452
+    },
+    {
+      "epoch": 0.7224826284070825,
+      "grad_norm": 1.6484375,
+      "learning_rate": 8.964931306169182e-06,
+      "loss": 0.3342,
+      "step": 16454
+    },
+    {
+      "epoch": 0.722570446886286,
+      "grad_norm": 1.6875,
+      "learning_rate": 8.959624182827187e-06,
+      "loss": 0.311,
+      "step": 16456
+    },
+    {
+      "epoch": 0.7226582653654895,
+      "grad_norm": 1.515625,
+      "learning_rate": 8.954318287861016e-06,
+      "loss": 0.3384,
+      "step": 16458
+    },
+    {
+      "epoch": 0.722746083844693,
+      "grad_norm": 1.6015625,
+      "learning_rate": 8.949013621676988e-06,
+      "loss": 0.3401,
+      "step": 16460
+    },
+    {
+      "epoch": 0.7228339023238965,
+      "grad_norm": 1.5859375,
+      "learning_rate": 8.943710184681353e-06,
+      "loss": 0.3354,
+      "step": 16462
+    },
+    {
+      "epoch": 0.7229217208031,
+      "grad_norm": 1.484375,
+      "learning_rate": 8.938407977280233e-06,
+      "loss": 0.3297,
+      "step": 16464
+    },
+    {
+      "epoch": 0.7230095392823035,
+      "grad_norm": 1.53125,
+      "learning_rate": 8.93310699987969e-06,
+      "loss": 0.2986,
+      "step": 16466
+    },
+    {
+      "epoch": 0.723097357761507,
+      "grad_norm": 1.6015625,
+      "learning_rate": 8.927807252885664e-06,
+      "loss": 0.3087,
+      "step": 16468
+    },
+    {
+      "epoch": 0.7231851762407104,
+      "grad_norm": 1.59375,
+      "learning_rate": 8.922508736704002e-06,
+      "loss": 0.3098,
+      "step": 16470
+    },
+    {
+      "epoch": 0.7232729947199139,
+      "grad_norm": 1.59375,
+      "learning_rate": 8.917211451740484e-06,
+      "loss": 0.3198,
+      "step": 16472
+    },
+    {
+      "epoch": 0.7233608131991174,
+      "grad_norm": 1.5078125,
+      "learning_rate": 8.911915398400767e-06,
+      "loss": 0.3407,
+      "step": 16474
+    },
+    {
+      "epoch": 0.723448631678321,
+      "grad_norm": 1.75,
+      "learning_rate": 8.906620577090427e-06,
+      "loss": 0.3181,
+      "step": 16476
+    },
+    {
+      "epoch": 0.7235364501575244,
+      "grad_norm": 1.546875,
+      "learning_rate": 8.90132698821493e-06,
+      "loss": 0.3371,
+      "step": 16478
+    },
+    {
+      "epoch": 0.7236242686367279,
+      "grad_norm": 1.5390625,
+      "learning_rate": 8.896034632179683e-06,
+      "loss": 0.342,
+      "step": 16480
+    },
+    {
+      "epoch": 0.7237120871159314,
+      "grad_norm": 1.640625,
+      "learning_rate": 8.890743509389953e-06,
+      "loss": 0.3339,
+      "step": 16482
+    },
+    {
+      "epoch": 0.7237999055951349,
+      "grad_norm": 1.5703125,
+      "learning_rate": 8.885453620250958e-06,
+      "loss": 0.3264,
+      "step": 16484
+    },
+    {
+      "epoch": 0.7238877240743383,
+      "grad_norm": 1.515625,
+      "learning_rate": 8.880164965167787e-06,
+      "loss": 0.3196,
+      "step": 16486
+    },
+    {
+      "epoch": 0.7239755425535418,
+      "grad_norm": 1.5703125,
+      "learning_rate": 8.874877544545438e-06,
+      "loss": 0.3019,
+      "step": 16488
+    },
+    {
+      "epoch": 0.7240633610327453,
+      "grad_norm": 1.5234375,
+      "learning_rate": 8.86959135878884e-06,
+      "loss": 0.3177,
+      "step": 16490
+    },
+    {
+      "epoch": 0.7241511795119489,
+      "grad_norm": 1.59375,
+      "learning_rate": 8.864306408302795e-06,
+      "loss": 0.3494,
+      "step": 16492
+    },
+    {
+      "epoch": 0.7242389979911523,
+      "grad_norm": 1.6328125,
+      "learning_rate": 8.859022693492042e-06,
+      "loss": 0.3201,
+      "step": 16494
+    },
+    {
+      "epoch": 0.7243268164703558,
+      "grad_norm": 1.59375,
+      "learning_rate": 8.8537402147612e-06,
+      "loss": 0.2976,
+      "step": 16496
+    },
+    {
+      "epoch": 0.7244146349495593,
+      "grad_norm": 1.4765625,
+      "learning_rate": 8.848458972514792e-06,
+      "loss": 0.332,
+      "step": 16498
+    },
+    {
+      "epoch": 0.7245024534287627,
+      "grad_norm": 1.609375,
+      "learning_rate": 8.843178967157278e-06,
+      "loss": 0.312,
+      "step": 16500
+    },
+    {
+      "epoch": 0.7245902719079662,
+      "grad_norm": 1.625,
+      "learning_rate": 8.837900199092986e-06,
+      "loss": 0.3438,
+      "step": 16502
+    },
+    {
+      "epoch": 0.7246780903871697,
+      "grad_norm": 1.484375,
+      "learning_rate": 8.832622668726184e-06,
+      "loss": 0.3145,
+      "step": 16504
+    },
+    {
+      "epoch": 0.7247659088663732,
+      "grad_norm": 1.4765625,
+      "learning_rate": 8.827346376460998e-06,
+      "loss": 0.3059,
+      "step": 16506
+    },
+    {
+      "epoch": 0.7248537273455767,
+      "grad_norm": 1.59375,
+      "learning_rate": 8.822071322701513e-06,
+      "loss": 0.3079,
+      "step": 16508
+    },
+    {
+      "epoch": 0.7249415458247802,
+      "grad_norm": 1.5703125,
+      "learning_rate": 8.816797507851682e-06,
+      "loss": 0.3218,
+      "step": 16510
+    },
+    {
+      "epoch": 0.7250293643039837,
+      "grad_norm": 1.640625,
+      "learning_rate": 8.811524932315371e-06,
+      "loss": 0.31,
+      "step": 16512
+    },
+    {
+      "epoch": 0.7251171827831872,
+      "grad_norm": 1.4609375,
+      "learning_rate": 8.806253596496369e-06,
+      "loss": 0.3252,
+      "step": 16514
+    },
+    {
+      "epoch": 0.7252050012623906,
+      "grad_norm": 1.4765625,
+      "learning_rate": 8.800983500798341e-06,
+      "loss": 0.3262,
+      "step": 16516
+    },
+    {
+      "epoch": 0.7252928197415941,
+      "grad_norm": 1.5625,
+      "learning_rate": 8.795714645624887e-06,
+      "loss": 0.3043,
+      "step": 16518
+    },
+    {
+      "epoch": 0.7253806382207976,
+      "grad_norm": 1.53125,
+      "learning_rate": 8.79044703137949e-06,
+      "loss": 0.3325,
+      "step": 16520
+    },
+    {
+      "epoch": 0.7254684567000012,
+      "grad_norm": 1.46875,
+      "learning_rate": 8.785180658465536e-06,
+      "loss": 0.3342,
+      "step": 16522
+    },
+    {
+      "epoch": 0.7255562751792046,
+      "grad_norm": 1.578125,
+      "learning_rate": 8.779915527286343e-06,
+      "loss": 0.3068,
+      "step": 16524
+    },
+    {
+      "epoch": 0.7256440936584081,
+      "grad_norm": 1.5625,
+      "learning_rate": 8.7746516382451e-06,
+      "loss": 0.3271,
+      "step": 16526
+    },
+    {
+      "epoch": 0.7257319121376116,
+      "grad_norm": 1.6171875,
+      "learning_rate": 8.769388991744928e-06,
+      "loss": 0.2964,
+      "step": 16528
+    },
+    {
+      "epoch": 0.725819730616815,
+      "grad_norm": 1.6953125,
+      "learning_rate": 8.764127588188842e-06,
+      "loss": 0.3511,
+      "step": 16530
+    },
+    {
+      "epoch": 0.7259075490960185,
+      "grad_norm": 1.640625,
+      "learning_rate": 8.758867427979748e-06,
+      "loss": 0.3357,
+      "step": 16532
+    },
+    {
+      "epoch": 0.725995367575222,
+      "grad_norm": 1.5859375,
+      "learning_rate": 8.753608511520489e-06,
+      "loss": 0.348,
+      "step": 16534
+    },
+    {
+      "epoch": 0.7260831860544255,
+      "grad_norm": 1.5546875,
+      "learning_rate": 8.748350839213782e-06,
+      "loss": 0.3523,
+      "step": 16536
+    },
+    {
+      "epoch": 0.726171004533629,
+      "grad_norm": 1.5390625,
+      "learning_rate": 8.743094411462266e-06,
+      "loss": 0.3257,
+      "step": 16538
+    },
+    {
+      "epoch": 0.7262588230128325,
+      "grad_norm": 1.453125,
+      "learning_rate": 8.737839228668468e-06,
+      "loss": 0.307,
+      "step": 16540
+    },
+    {
+      "epoch": 0.726346641492036,
+      "grad_norm": 1.4765625,
+      "learning_rate": 8.73258529123485e-06,
+      "loss": 0.3303,
+      "step": 16542
+    },
+    {
+      "epoch": 0.7264344599712395,
+      "grad_norm": 1.6953125,
+      "learning_rate": 8.727332599563745e-06,
+      "loss": 0.3162,
+      "step": 16544
+    },
+    {
+      "epoch": 0.7265222784504429,
+      "grad_norm": 1.5546875,
+      "learning_rate": 8.722081154057408e-06,
+      "loss": 0.3351,
+      "step": 16546
+    },
+    {
+      "epoch": 0.7266100969296464,
+      "grad_norm": 1.6484375,
+      "learning_rate": 8.716830955118002e-06,
+      "loss": 0.329,
+      "step": 16548
+    },
+    {
+      "epoch": 0.7266979154088499,
+      "grad_norm": 1.6796875,
+      "learning_rate": 8.711582003147578e-06,
+      "loss": 0.3177,
+      "step": 16550
+    },
+    {
+      "epoch": 0.7267857338880533,
+      "grad_norm": 1.5078125,
+      "learning_rate": 8.706334298548119e-06,
+      "loss": 0.3,
+      "step": 16552
+    },
+    {
+      "epoch": 0.7268735523672569,
+      "grad_norm": 1.484375,
+      "learning_rate": 8.701087841721475e-06,
+      "loss": 0.3191,
+      "step": 16554
+    },
+    {
+      "epoch": 0.7269613708464604,
+      "grad_norm": 1.484375,
+      "learning_rate": 8.69584263306944e-06,
+      "loss": 0.3198,
+      "step": 16556
+    },
+    {
+      "epoch": 0.7270491893256639,
+      "grad_norm": 1.5703125,
+      "learning_rate": 8.690598672993683e-06,
+      "loss": 0.3061,
+      "step": 16558
+    },
+    {
+      "epoch": 0.7271370078048673,
+      "grad_norm": 1.4375,
+      "learning_rate": 8.685355961895784e-06,
+      "loss": 0.3073,
+      "step": 16560
+    },
+    {
+      "epoch": 0.7272248262840708,
+      "grad_norm": 1.6328125,
+      "learning_rate": 8.680114500177241e-06,
+      "loss": 0.3,
+      "step": 16562
+    },
+    {
+      "epoch": 0.7273126447632743,
+      "grad_norm": 1.5,
+      "learning_rate": 8.674874288239438e-06,
+      "loss": 0.3184,
+      "step": 16564
+    },
+    {
+      "epoch": 0.7274004632424778,
+      "grad_norm": 1.4453125,
+      "learning_rate": 8.669635326483688e-06,
+      "loss": 0.2938,
+      "step": 16566
+    },
+    {
+      "epoch": 0.7274882817216813,
+      "grad_norm": 1.5703125,
+      "learning_rate": 8.664397615311165e-06,
+      "loss": 0.3619,
+      "step": 16568
+    },
+    {
+      "epoch": 0.7275761002008848,
+      "grad_norm": 1.5859375,
+      "learning_rate": 8.659161155122997e-06,
+      "loss": 0.3037,
+      "step": 16570
+    },
+    {
+      "epoch": 0.7276639186800883,
+      "grad_norm": 1.5546875,
+      "learning_rate": 8.653925946320183e-06,
+      "loss": 0.3034,
+      "step": 16572
+    },
+    {
+      "epoch": 0.7277517371592918,
+      "grad_norm": 1.4375,
+      "learning_rate": 8.648691989303631e-06,
+      "loss": 0.3027,
+      "step": 16574
+    },
+    {
+      "epoch": 0.7278395556384952,
+      "grad_norm": 1.515625,
+      "learning_rate": 8.643459284474176e-06,
+      "loss": 0.3192,
+      "step": 16576
+    },
+    {
+      "epoch": 0.7279273741176987,
+      "grad_norm": 1.53125,
+      "learning_rate": 8.638227832232517e-06,
+      "loss": 0.3333,
+      "step": 16578
+    },
+    {
+      "epoch": 0.7280151925969022,
+      "grad_norm": 1.6015625,
+      "learning_rate": 8.632997632979306e-06,
+      "loss": 0.3472,
+      "step": 16580
+    },
+    {
+      "epoch": 0.7281030110761056,
+      "grad_norm": 1.5859375,
+      "learning_rate": 8.62776868711506e-06,
+      "loss": 0.3078,
+      "step": 16582
+    },
+    {
+      "epoch": 0.7281908295553092,
+      "grad_norm": 1.6484375,
+      "learning_rate": 8.622540995040202e-06,
+      "loss": 0.3002,
+      "step": 16584
+    },
+    {
+      "epoch": 0.7282786480345127,
+      "grad_norm": 1.5546875,
+      "learning_rate": 8.617314557155087e-06,
+      "loss": 0.3192,
+      "step": 16586
+    },
+    {
+      "epoch": 0.7283664665137162,
+      "grad_norm": 1.5234375,
+      "learning_rate": 8.612089373859945e-06,
+      "loss": 0.3664,
+      "step": 16588
+    },
+    {
+      "epoch": 0.7284542849929196,
+      "grad_norm": 1.703125,
+      "learning_rate": 8.606865445554934e-06,
+      "loss": 0.3197,
+      "step": 16590
+    },
+    {
+      "epoch": 0.7285421034721231,
+      "grad_norm": 1.578125,
+      "learning_rate": 8.601642772640097e-06,
+      "loss": 0.3189,
+      "step": 16592
+    },
+    {
+      "epoch": 0.7286299219513266,
+      "grad_norm": 1.578125,
+      "learning_rate": 8.596421355515383e-06,
+      "loss": 0.3084,
+      "step": 16594
+    },
+    {
+      "epoch": 0.7287177404305301,
+      "grad_norm": 1.5859375,
+      "learning_rate": 8.591201194580667e-06,
+      "loss": 0.3358,
+      "step": 16596
+    },
+    {
+      "epoch": 0.7288055589097335,
+      "grad_norm": 1.5859375,
+      "learning_rate": 8.585982290235681e-06,
+      "loss": 0.3179,
+      "step": 16598
+    },
+    {
+      "epoch": 0.7288933773889371,
+      "grad_norm": 1.515625,
+      "learning_rate": 8.580764642880113e-06,
+      "loss": 0.3107,
+      "step": 16600
+    },
+    {
+      "epoch": 0.7289811958681406,
+      "grad_norm": 1.5078125,
+      "learning_rate": 8.575548252913515e-06,
+      "loss": 0.3655,
+      "step": 16602
+    },
+    {
+      "epoch": 0.7290690143473441,
+      "grad_norm": 1.515625,
+      "learning_rate": 8.57033312073538e-06,
+      "loss": 0.3162,
+      "step": 16604
+    },
+    {
+      "epoch": 0.7291568328265475,
+      "grad_norm": 1.5625,
+      "learning_rate": 8.565119246745074e-06,
+      "loss": 0.3325,
+      "step": 16606
+    },
+    {
+      "epoch": 0.729244651305751,
+      "grad_norm": 1.5625,
+      "learning_rate": 8.559906631341866e-06,
+      "loss": 0.3174,
+      "step": 16608
+    },
+    {
+      "epoch": 0.7293324697849545,
+      "grad_norm": 1.4921875,
+      "learning_rate": 8.554695274924956e-06,
+      "loss": 0.3142,
+      "step": 16610
+    },
+    {
+      "epoch": 0.729420288264158,
+      "grad_norm": 1.5625,
+      "learning_rate": 8.549485177893418e-06,
+      "loss": 0.3224,
+      "step": 16612
+    },
+    {
+      "epoch": 0.7295081067433614,
+      "grad_norm": 1.6640625,
+      "learning_rate": 8.544276340646256e-06,
+      "loss": 0.3582,
+      "step": 16614
+    },
+    {
+      "epoch": 0.729595925222565,
+      "grad_norm": 1.53125,
+      "learning_rate": 8.539068763582347e-06,
+      "loss": 0.3463,
+      "step": 16616
+    },
+    {
+      "epoch": 0.7296837437017685,
+      "grad_norm": 1.5078125,
+      "learning_rate": 8.533862447100511e-06,
+      "loss": 0.297,
+      "step": 16618
+    },
+    {
+      "epoch": 0.7297715621809719,
+      "grad_norm": 1.5546875,
+      "learning_rate": 8.528657391599431e-06,
+      "loss": 0.3059,
+      "step": 16620
+    },
+    {
+      "epoch": 0.7298593806601754,
+      "grad_norm": 1.5625,
+      "learning_rate": 8.52345359747771e-06,
+      "loss": 0.3107,
+      "step": 16622
+    },
+    {
+      "epoch": 0.7299471991393789,
+      "grad_norm": 1.5078125,
+      "learning_rate": 8.51825106513387e-06,
+      "loss": 0.3003,
+      "step": 16624
+    },
+    {
+      "epoch": 0.7300350176185824,
+      "grad_norm": 1.515625,
+      "learning_rate": 8.513049794966305e-06,
+      "loss": 0.3071,
+      "step": 16626
+    },
+    {
+      "epoch": 0.7301228360977858,
+      "grad_norm": 1.46875,
+      "learning_rate": 8.507849787373356e-06,
+      "loss": 0.3396,
+      "step": 16628
+    },
+    {
+      "epoch": 0.7302106545769894,
+      "grad_norm": 1.484375,
+      "learning_rate": 8.50265104275321e-06,
+      "loss": 0.3191,
+      "step": 16630
+    },
+    {
+      "epoch": 0.7302984730561929,
+      "grad_norm": 1.5546875,
+      "learning_rate": 8.497453561504007e-06,
+      "loss": 0.2972,
+      "step": 16632
+    },
+    {
+      "epoch": 0.7303862915353964,
+      "grad_norm": 1.71875,
+      "learning_rate": 8.492257344023769e-06,
+      "loss": 0.3168,
+      "step": 16634
+    },
+    {
+      "epoch": 0.7304741100145998,
+      "grad_norm": 1.5078125,
+      "learning_rate": 8.48706239071041e-06,
+      "loss": 0.3114,
+      "step": 16636
+    },
+    {
+      "epoch": 0.7305619284938033,
+      "grad_norm": 1.4375,
+      "learning_rate": 8.481868701961782e-06,
+      "loss": 0.3148,
+      "step": 16638
+    },
+    {
+      "epoch": 0.7306497469730068,
+      "grad_norm": 1.515625,
+      "learning_rate": 8.476676278175597e-06,
+      "loss": 0.3238,
+      "step": 16640
+    },
+    {
+      "epoch": 0.7307375654522102,
+      "grad_norm": 1.6953125,
+      "learning_rate": 8.471485119749514e-06,
+      "loss": 0.3176,
+      "step": 16642
+    },
+    {
+      "epoch": 0.7308253839314137,
+      "grad_norm": 1.53125,
+      "learning_rate": 8.466295227081061e-06,
+      "loss": 0.3225,
+      "step": 16644
+    },
+    {
+      "epoch": 0.7309132024106173,
+      "grad_norm": 1.671875,
+      "learning_rate": 8.461106600567679e-06,
+      "loss": 0.3377,
+      "step": 16646
+    },
+    {
+      "epoch": 0.7310010208898208,
+      "grad_norm": 1.5859375,
+      "learning_rate": 8.455919240606722e-06,
+      "loss": 0.3063,
+      "step": 16648
+    },
+    {
+      "epoch": 0.7310888393690242,
+      "grad_norm": 1.5546875,
+      "learning_rate": 8.450733147595427e-06,
+      "loss": 0.3358,
+      "step": 16650
+    },
+    {
+      "epoch": 0.7311766578482277,
+      "grad_norm": 1.6015625,
+      "learning_rate": 8.445548321930966e-06,
+      "loss": 0.3395,
+      "step": 16652
+    },
+    {
+      "epoch": 0.7312644763274312,
+      "grad_norm": 1.5390625,
+      "learning_rate": 8.44036476401038e-06,
+      "loss": 0.3459,
+      "step": 16654
+    },
+    {
+      "epoch": 0.7313522948066347,
+      "grad_norm": 1.59375,
+      "learning_rate": 8.435182474230624e-06,
+      "loss": 0.3312,
+      "step": 16656
+    },
+    {
+      "epoch": 0.7314401132858381,
+      "grad_norm": 1.6328125,
+      "learning_rate": 8.430001452988582e-06,
+      "loss": 0.3066,
+      "step": 16658
+    },
+    {
+      "epoch": 0.7315279317650416,
+      "grad_norm": 1.546875,
+      "learning_rate": 8.424821700680982e-06,
+      "loss": 0.3105,
+      "step": 16660
+    },
+    {
+      "epoch": 0.7316157502442452,
+      "grad_norm": 1.53125,
+      "learning_rate": 8.419643217704517e-06,
+      "loss": 0.3185,
+      "step": 16662
+    },
+    {
+      "epoch": 0.7317035687234487,
+      "grad_norm": 1.5546875,
+      "learning_rate": 8.414466004455743e-06,
+      "loss": 0.3262,
+      "step": 16664
+    },
+    {
+      "epoch": 0.7317913872026521,
+      "grad_norm": 1.5625,
+      "learning_rate": 8.409290061331145e-06,
+      "loss": 0.3022,
+      "step": 16666
+    },
+    {
+      "epoch": 0.7318792056818556,
+      "grad_norm": 1.5078125,
+      "learning_rate": 8.404115388727093e-06,
+      "loss": 0.3006,
+      "step": 16668
+    },
+    {
+      "epoch": 0.7319670241610591,
+      "grad_norm": 1.5234375,
+      "learning_rate": 8.398941987039854e-06,
+      "loss": 0.3115,
+      "step": 16670
+    },
+    {
+      "epoch": 0.7320548426402625,
+      "grad_norm": 1.5703125,
+      "learning_rate": 8.393769856665626e-06,
+      "loss": 0.3167,
+      "step": 16672
+    },
+    {
+      "epoch": 0.732142661119466,
+      "grad_norm": 1.4453125,
+      "learning_rate": 8.388598998000472e-06,
+      "loss": 0.284,
+      "step": 16674
+    },
+    {
+      "epoch": 0.7322304795986696,
+      "grad_norm": 1.5078125,
+      "learning_rate": 8.383429411440399e-06,
+      "loss": 0.2953,
+      "step": 16676
+    },
+    {
+      "epoch": 0.7323182980778731,
+      "grad_norm": 1.546875,
+      "learning_rate": 8.378261097381285e-06,
+      "loss": 0.3249,
+      "step": 16678
+    },
+    {
+      "epoch": 0.7324061165570765,
+      "grad_norm": 1.6171875,
+      "learning_rate": 8.373094056218913e-06,
+      "loss": 0.2998,
+      "step": 16680
+    },
+    {
+      "epoch": 0.73249393503628,
+      "grad_norm": 1.53125,
+      "learning_rate": 8.367928288348992e-06,
+      "loss": 0.3415,
+      "step": 16682
+    },
+    {
+      "epoch": 0.7325817535154835,
+      "grad_norm": 1.46875,
+      "learning_rate": 8.3627637941671e-06,
+      "loss": 0.2889,
+      "step": 16684
+    },
+    {
+      "epoch": 0.732669571994687,
+      "grad_norm": 1.5859375,
+      "learning_rate": 8.357600574068755e-06,
+      "loss": 0.3368,
+      "step": 16686
+    },
+    {
+      "epoch": 0.7327573904738904,
+      "grad_norm": 1.5546875,
+      "learning_rate": 8.352438628449347e-06,
+      "loss": 0.2997,
+      "step": 16688
+    },
+    {
+      "epoch": 0.7328452089530939,
+      "grad_norm": 1.53125,
+      "learning_rate": 8.347277957704178e-06,
+      "loss": 0.3069,
+      "step": 16690
+    },
+    {
+      "epoch": 0.7329330274322975,
+      "grad_norm": 1.5078125,
+      "learning_rate": 8.34211856222845e-06,
+      "loss": 0.3195,
+      "step": 16692
+    },
+    {
+      "epoch": 0.733020845911501,
+      "grad_norm": 1.59375,
+      "learning_rate": 8.33696044241728e-06,
+      "loss": 0.3095,
+      "step": 16694
+    },
+    {
+      "epoch": 0.7331086643907044,
+      "grad_norm": 1.5078125,
+      "learning_rate": 8.331803598665674e-06,
+      "loss": 0.2801,
+      "step": 16696
+    },
+    {
+      "epoch": 0.7331964828699079,
+      "grad_norm": 1.5,
+      "learning_rate": 8.326648031368536e-06,
+      "loss": 0.3288,
+      "step": 16698
+    },
+    {
+      "epoch": 0.7332843013491114,
+      "grad_norm": 1.5703125,
+      "learning_rate": 8.321493740920699e-06,
+      "loss": 0.3092,
+      "step": 16700
+    },
+    {
+      "epoch": 0.7333721198283149,
+      "grad_norm": 1.5625,
+      "learning_rate": 8.316340727716862e-06,
+      "loss": 0.3149,
+      "step": 16702
+    },
+    {
+      "epoch": 0.7334599383075183,
+      "grad_norm": 1.6484375,
+      "learning_rate": 8.311188992151656e-06,
+      "loss": 0.3189,
+      "step": 16704
+    },
+    {
+      "epoch": 0.7335477567867218,
+      "grad_norm": 1.4296875,
+      "learning_rate": 8.3060385346196e-06,
+      "loss": 0.3034,
+      "step": 16706
+    },
+    {
+      "epoch": 0.7336355752659254,
+      "grad_norm": 1.5078125,
+      "learning_rate": 8.300889355515107e-06,
+      "loss": 0.3473,
+      "step": 16708
+    },
+    {
+      "epoch": 0.7337233937451288,
+      "grad_norm": 1.453125,
+      "learning_rate": 8.295741455232517e-06,
+      "loss": 0.3084,
+      "step": 16710
+    },
+    {
+      "epoch": 0.7338112122243323,
+      "grad_norm": 1.546875,
+      "learning_rate": 8.290594834166043e-06,
+      "loss": 0.332,
+      "step": 16712
+    },
+    {
+      "epoch": 0.7338990307035358,
+      "grad_norm": 1.515625,
+      "learning_rate": 8.285449492709829e-06,
+      "loss": 0.325,
+      "step": 16714
+    },
+    {
+      "epoch": 0.7339868491827393,
+      "grad_norm": 1.5078125,
+      "learning_rate": 8.2803054312579e-06,
+      "loss": 0.3286,
+      "step": 16716
+    },
+    {
+      "epoch": 0.7340746676619427,
+      "grad_norm": 1.5078125,
+      "learning_rate": 8.275162650204182e-06,
+      "loss": 0.3368,
+      "step": 16718
+    },
+    {
+      "epoch": 0.7341624861411462,
+      "grad_norm": 1.78125,
+      "learning_rate": 8.270021149942536e-06,
+      "loss": 0.3279,
+      "step": 16720
+    },
+    {
+      "epoch": 0.7342503046203498,
+      "grad_norm": 1.609375,
+      "learning_rate": 8.26488093086666e-06,
+      "loss": 0.3313,
+      "step": 16722
+    },
+    {
+      "epoch": 0.7343381230995533,
+      "grad_norm": 1.59375,
+      "learning_rate": 8.259741993370227e-06,
+      "loss": 0.3095,
+      "step": 16724
+    },
+    {
+      "epoch": 0.7344259415787567,
+      "grad_norm": 1.5625,
+      "learning_rate": 8.254604337846753e-06,
+      "loss": 0.3407,
+      "step": 16726
+    },
+    {
+      "epoch": 0.7345137600579602,
+      "grad_norm": 1.671875,
+      "learning_rate": 8.2494679646897e-06,
+      "loss": 0.3175,
+      "step": 16728
+    },
+    {
+      "epoch": 0.7346015785371637,
+      "grad_norm": 1.6328125,
+      "learning_rate": 8.24433287429241e-06,
+      "loss": 0.3401,
+      "step": 16730
+    },
+    {
+      "epoch": 0.7346893970163672,
+      "grad_norm": 1.4453125,
+      "learning_rate": 8.239199067048114e-06,
+      "loss": 0.316,
+      "step": 16732
+    },
+    {
+      "epoch": 0.7347772154955706,
+      "grad_norm": 1.484375,
+      "learning_rate": 8.23406654334998e-06,
+      "loss": 0.3004,
+      "step": 16734
+    },
+    {
+      "epoch": 0.7348650339747741,
+      "grad_norm": 1.515625,
+      "learning_rate": 8.22893530359104e-06,
+      "loss": 0.319,
+      "step": 16736
+    },
+    {
+      "epoch": 0.7349528524539777,
+      "grad_norm": 1.546875,
+      "learning_rate": 8.223805348164265e-06,
+      "loss": 0.319,
+      "step": 16738
+    },
+    {
+      "epoch": 0.7350406709331812,
+      "grad_norm": 1.53125,
+      "learning_rate": 8.218676677462497e-06,
+      "loss": 0.3369,
+      "step": 16740
+    },
+    {
+      "epoch": 0.7351284894123846,
+      "grad_norm": 1.5703125,
+      "learning_rate": 8.213549291878483e-06,
+      "loss": 0.3231,
+      "step": 16742
+    },
+    {
+      "epoch": 0.7352163078915881,
+      "grad_norm": 1.4921875,
+      "learning_rate": 8.208423191804899e-06,
+      "loss": 0.3173,
+      "step": 16744
+    },
+    {
+      "epoch": 0.7353041263707916,
+      "grad_norm": 1.3984375,
+      "learning_rate": 8.20329837763428e-06,
+      "loss": 0.3021,
+      "step": 16746
+    },
+    {
+      "epoch": 0.735391944849995,
+      "grad_norm": 1.7578125,
+      "learning_rate": 8.19817484975911e-06,
+      "loss": 0.32,
+      "step": 16748
+    },
+    {
+      "epoch": 0.7354797633291985,
+      "grad_norm": 1.578125,
+      "learning_rate": 8.193052608571736e-06,
+      "loss": 0.309,
+      "step": 16750
+    },
+    {
+      "epoch": 0.735567581808402,
+      "grad_norm": 1.5859375,
+      "learning_rate": 8.18793165446442e-06,
+      "loss": 0.2867,
+      "step": 16752
+    },
+    {
+      "epoch": 0.7356554002876056,
+      "grad_norm": 1.46875,
+      "learning_rate": 8.182811987829323e-06,
+      "loss": 0.3224,
+      "step": 16754
+    },
+    {
+      "epoch": 0.735743218766809,
+      "grad_norm": 1.421875,
+      "learning_rate": 8.177693609058521e-06,
+      "loss": 0.3212,
+      "step": 16756
+    },
+    {
+      "epoch": 0.7358310372460125,
+      "grad_norm": 1.578125,
+      "learning_rate": 8.172576518543976e-06,
+      "loss": 0.3462,
+      "step": 16758
+    },
+    {
+      "epoch": 0.735918855725216,
+      "grad_norm": 1.5390625,
+      "learning_rate": 8.167460716677546e-06,
+      "loss": 0.3368,
+      "step": 16760
+    },
+    {
+      "epoch": 0.7360066742044195,
+      "grad_norm": 1.5234375,
+      "learning_rate": 8.16234620385102e-06,
+      "loss": 0.3412,
+      "step": 16762
+    },
+    {
+      "epoch": 0.7360944926836229,
+      "grad_norm": 1.4921875,
+      "learning_rate": 8.157232980456047e-06,
+      "loss": 0.3209,
+      "step": 16764
+    },
+    {
+      "epoch": 0.7361823111628264,
+      "grad_norm": 1.5234375,
+      "learning_rate": 8.15212104688422e-06,
+      "loss": 0.2946,
+      "step": 16766
+    },
+    {
+      "epoch": 0.7362701296420299,
+      "grad_norm": 1.5234375,
+      "learning_rate": 8.147010403527003e-06,
+      "loss": 0.3094,
+      "step": 16768
+    },
+    {
+      "epoch": 0.7363579481212335,
+      "grad_norm": 1.515625,
+      "learning_rate": 8.141901050775758e-06,
+      "loss": 0.3224,
+      "step": 16770
+    },
+    {
+      "epoch": 0.7364457666004369,
+      "grad_norm": 1.53125,
+      "learning_rate": 8.136792989021783e-06,
+      "loss": 0.3271,
+      "step": 16772
+    },
+    {
+      "epoch": 0.7365335850796404,
+      "grad_norm": 1.7265625,
+      "learning_rate": 8.131686218656231e-06,
+      "loss": 0.3339,
+      "step": 16774
+    },
+    {
+      "epoch": 0.7366214035588439,
+      "grad_norm": 1.5078125,
+      "learning_rate": 8.126580740070202e-06,
+      "loss": 0.311,
+      "step": 16776
+    },
+    {
+      "epoch": 0.7367092220380473,
+      "grad_norm": 1.6328125,
+      "learning_rate": 8.121476553654666e-06,
+      "loss": 0.309,
+      "step": 16778
+    },
+    {
+      "epoch": 0.7367970405172508,
+      "grad_norm": 1.5703125,
+      "learning_rate": 8.116373659800502e-06,
+      "loss": 0.3169,
+      "step": 16780
+    },
+    {
+      "epoch": 0.7368848589964543,
+      "grad_norm": 1.5,
+      "learning_rate": 8.11127205889849e-06,
+      "loss": 0.331,
+      "step": 16782
+    },
+    {
+      "epoch": 0.7369726774756579,
+      "grad_norm": 1.578125,
+      "learning_rate": 8.106171751339303e-06,
+      "loss": 0.3081,
+      "step": 16784
+    },
+    {
+      "epoch": 0.7370604959548613,
+      "grad_norm": 1.5234375,
+      "learning_rate": 8.10107273751354e-06,
+      "loss": 0.3114,
+      "step": 16786
+    },
+    {
+      "epoch": 0.7371483144340648,
+      "grad_norm": 1.546875,
+      "learning_rate": 8.095975017811671e-06,
+      "loss": 0.332,
+      "step": 16788
+    },
+    {
+      "epoch": 0.7372361329132683,
+      "grad_norm": 1.546875,
+      "learning_rate": 8.090878592624097e-06,
+      "loss": 0.3164,
+      "step": 16790
+    },
+    {
+      "epoch": 0.7373239513924718,
+      "grad_norm": 1.4921875,
+      "learning_rate": 8.085783462341093e-06,
+      "loss": 0.3269,
+      "step": 16792
+    },
+    {
+      "epoch": 0.7374117698716752,
+      "grad_norm": 1.5078125,
+      "learning_rate": 8.080689627352837e-06,
+      "loss": 0.3515,
+      "step": 16794
+    },
+    {
+      "epoch": 0.7374995883508787,
+      "grad_norm": 1.5390625,
+      "learning_rate": 8.075597088049434e-06,
+      "loss": 0.2985,
+      "step": 16796
+    },
+    {
+      "epoch": 0.7375874068300822,
+      "grad_norm": 1.59375,
+      "learning_rate": 8.070505844820853e-06,
+      "loss": 0.3259,
+      "step": 16798
+    },
+    {
+      "epoch": 0.7376752253092858,
+      "grad_norm": 1.5078125,
+      "learning_rate": 8.065415898057003e-06,
+      "loss": 0.328,
+      "step": 16800
+    },
+    {
+      "epoch": 0.7377630437884892,
+      "grad_norm": 1.4765625,
+      "learning_rate": 8.060327248147662e-06,
+      "loss": 0.2985,
+      "step": 16802
+    },
+    {
+      "epoch": 0.7378508622676927,
+      "grad_norm": 1.4765625,
+      "learning_rate": 8.055239895482514e-06,
+      "loss": 0.3036,
+      "step": 16804
+    },
+    {
+      "epoch": 0.7379386807468962,
+      "grad_norm": 1.484375,
+      "learning_rate": 8.050153840451163e-06,
+      "loss": 0.3052,
+      "step": 16806
+    },
+    {
+      "epoch": 0.7380264992260996,
+      "grad_norm": 1.5234375,
+      "learning_rate": 8.045069083443088e-06,
+      "loss": 0.339,
+      "step": 16808
+    },
+    {
+      "epoch": 0.7381143177053031,
+      "grad_norm": 1.4609375,
+      "learning_rate": 8.039985624847692e-06,
+      "loss": 0.3002,
+      "step": 16810
+    },
+    {
+      "epoch": 0.7382021361845066,
+      "grad_norm": 1.546875,
+      "learning_rate": 8.034903465054266e-06,
+      "loss": 0.3592,
+      "step": 16812
+    },
+    {
+      "epoch": 0.7382899546637101,
+      "grad_norm": 1.53125,
+      "learning_rate": 8.029822604451997e-06,
+      "loss": 0.3023,
+      "step": 16814
+    },
+    {
+      "epoch": 0.7383777731429136,
+      "grad_norm": 1.5390625,
+      "learning_rate": 8.024743043429975e-06,
+      "loss": 0.3548,
+      "step": 16816
+    },
+    {
+      "epoch": 0.7384655916221171,
+      "grad_norm": 1.6015625,
+      "learning_rate": 8.019664782377207e-06,
+      "loss": 0.3256,
+      "step": 16818
+    },
+    {
+      "epoch": 0.7385534101013206,
+      "grad_norm": 1.59375,
+      "learning_rate": 8.014587821682578e-06,
+      "loss": 0.3362,
+      "step": 16820
+    },
+    {
+      "epoch": 0.7386412285805241,
+      "grad_norm": 1.53125,
+      "learning_rate": 8.009512161734881e-06,
+      "loss": 0.2868,
+      "step": 16822
+    },
+    {
+      "epoch": 0.7387290470597275,
+      "grad_norm": 1.453125,
+      "learning_rate": 8.004437802922823e-06,
+      "loss": 0.3509,
+      "step": 16824
+    },
+    {
+      "epoch": 0.738816865538931,
+      "grad_norm": 1.5234375,
+      "learning_rate": 7.999364745634982e-06,
+      "loss": 0.3372,
+      "step": 16826
+    },
+    {
+      "epoch": 0.7389046840181345,
+      "grad_norm": 1.5546875,
+      "learning_rate": 7.994292990259875e-06,
+      "loss": 0.3092,
+      "step": 16828
+    },
+    {
+      "epoch": 0.738992502497338,
+      "grad_norm": 1.4609375,
+      "learning_rate": 7.989222537185886e-06,
+      "loss": 0.3271,
+      "step": 16830
+    },
+    {
+      "epoch": 0.7390803209765415,
+      "grad_norm": 1.6640625,
+      "learning_rate": 7.984153386801304e-06,
+      "loss": 0.3161,
+      "step": 16832
+    },
+    {
+      "epoch": 0.739168139455745,
+      "grad_norm": 1.546875,
+      "learning_rate": 7.979085539494344e-06,
+      "loss": 0.3168,
+      "step": 16834
+    },
+    {
+      "epoch": 0.7392559579349485,
+      "grad_norm": 1.5078125,
+      "learning_rate": 7.974018995653085e-06,
+      "loss": 0.3135,
+      "step": 16836
+    },
+    {
+      "epoch": 0.7393437764141519,
+      "grad_norm": 1.4609375,
+      "learning_rate": 7.96895375566554e-06,
+      "loss": 0.3193,
+      "step": 16838
+    },
+    {
+      "epoch": 0.7394315948933554,
+      "grad_norm": 1.515625,
+      "learning_rate": 7.963889819919599e-06,
+      "loss": 0.3187,
+      "step": 16840
+    },
+    {
+      "epoch": 0.7395194133725589,
+      "grad_norm": 1.578125,
+      "learning_rate": 7.95882718880306e-06,
+      "loss": 0.3404,
+      "step": 16842
+    },
+    {
+      "epoch": 0.7396072318517624,
+      "grad_norm": 1.53125,
+      "learning_rate": 7.953765862703622e-06,
+      "loss": 0.3411,
+      "step": 16844
+    },
+    {
+      "epoch": 0.7396950503309659,
+      "grad_norm": 1.5078125,
+      "learning_rate": 7.948705842008869e-06,
+      "loss": 0.3225,
+      "step": 16846
+    },
+    {
+      "epoch": 0.7397828688101694,
+      "grad_norm": 1.5234375,
+      "learning_rate": 7.943647127106318e-06,
+      "loss": 0.346,
+      "step": 16848
+    },
+    {
+      "epoch": 0.7398706872893729,
+      "grad_norm": 1.515625,
+      "learning_rate": 7.938589718383354e-06,
+      "loss": 0.2811,
+      "step": 16850
+    },
+    {
+      "epoch": 0.7399585057685764,
+      "grad_norm": 1.59375,
+      "learning_rate": 7.933533616227284e-06,
+      "loss": 0.3548,
+      "step": 16852
+    },
+    {
+      "epoch": 0.7400463242477798,
+      "grad_norm": 1.4453125,
+      "learning_rate": 7.928478821025304e-06,
+      "loss": 0.2942,
+      "step": 16854
+    },
+    {
+      "epoch": 0.7401341427269833,
+      "grad_norm": 1.5078125,
+      "learning_rate": 7.923425333164497e-06,
+      "loss": 0.3173,
+      "step": 16856
+    },
+    {
+      "epoch": 0.7402219612061868,
+      "grad_norm": 1.6484375,
+      "learning_rate": 7.918373153031882e-06,
+      "loss": 0.2976,
+      "step": 16858
+    },
+    {
+      "epoch": 0.7403097796853902,
+      "grad_norm": 1.5390625,
+      "learning_rate": 7.913322281014337e-06,
+      "loss": 0.3343,
+      "step": 16860
+    },
+    {
+      "epoch": 0.7403975981645938,
+      "grad_norm": 1.5546875,
+      "learning_rate": 7.908272717498674e-06,
+      "loss": 0.3248,
+      "step": 16862
+    },
+    {
+      "epoch": 0.7404854166437973,
+      "grad_norm": 1.4921875,
+      "learning_rate": 7.903224462871586e-06,
+      "loss": 0.3092,
+      "step": 16864
+    },
+    {
+      "epoch": 0.7405732351230008,
+      "grad_norm": 1.453125,
+      "learning_rate": 7.898177517519659e-06,
+      "loss": 0.3267,
+      "step": 16866
+    },
+    {
+      "epoch": 0.7406610536022042,
+      "grad_norm": 1.5078125,
+      "learning_rate": 7.893131881829405e-06,
+      "loss": 0.3049,
+      "step": 16868
+    },
+    {
+      "epoch": 0.7407488720814077,
+      "grad_norm": 1.4296875,
+      "learning_rate": 7.888087556187201e-06,
+      "loss": 0.3228,
+      "step": 16870
+    },
+    {
+      "epoch": 0.7408366905606112,
+      "grad_norm": 1.546875,
+      "learning_rate": 7.883044540979373e-06,
+      "loss": 0.2953,
+      "step": 16872
+    },
+    {
+      "epoch": 0.7409245090398147,
+      "grad_norm": 1.625,
+      "learning_rate": 7.878002836592082e-06,
+      "loss": 0.331,
+      "step": 16874
+    },
+    {
+      "epoch": 0.7410123275190182,
+      "grad_norm": 1.640625,
+      "learning_rate": 7.872962443411445e-06,
+      "loss": 0.3062,
+      "step": 16876
+    },
+    {
+      "epoch": 0.7411001459982217,
+      "grad_norm": 1.4921875,
+      "learning_rate": 7.86792336182344e-06,
+      "loss": 0.3085,
+      "step": 16878
+    },
+    {
+      "epoch": 0.7411879644774252,
+      "grad_norm": 1.7265625,
+      "learning_rate": 7.86288559221398e-06,
+      "loss": 0.3122,
+      "step": 16880
+    },
+    {
+      "epoch": 0.7412757829566287,
+      "grad_norm": 1.5859375,
+      "learning_rate": 7.85784913496885e-06,
+      "loss": 0.3141,
+      "step": 16882
+    },
+    {
+      "epoch": 0.7413636014358321,
+      "grad_norm": 1.5703125,
+      "learning_rate": 7.852813990473734e-06,
+      "loss": 0.3276,
+      "step": 16884
+    },
+    {
+      "epoch": 0.7414514199150356,
+      "grad_norm": 1.5,
+      "learning_rate": 7.847780159114243e-06,
+      "loss": 0.3191,
+      "step": 16886
+    },
+    {
+      "epoch": 0.7415392383942391,
+      "grad_norm": 1.7734375,
+      "learning_rate": 7.84274764127585e-06,
+      "loss": 0.3259,
+      "step": 16888
+    },
+    {
+      "epoch": 0.7416270568734425,
+      "grad_norm": 1.640625,
+      "learning_rate": 7.837716437343961e-06,
+      "loss": 0.31,
+      "step": 16890
+    },
+    {
+      "epoch": 0.7417148753526461,
+      "grad_norm": 1.5078125,
+      "learning_rate": 7.832686547703866e-06,
+      "loss": 0.344,
+      "step": 16892
+    },
+    {
+      "epoch": 0.7418026938318496,
+      "grad_norm": 1.4453125,
+      "learning_rate": 7.827657972740738e-06,
+      "loss": 0.3119,
+      "step": 16894
+    },
+    {
+      "epoch": 0.7418905123110531,
+      "grad_norm": 1.6015625,
+      "learning_rate": 7.82263071283969e-06,
+      "loss": 0.3151,
+      "step": 16896
+    },
+    {
+      "epoch": 0.7419783307902565,
+      "grad_norm": 1.5078125,
+      "learning_rate": 7.81760476838569e-06,
+      "loss": 0.3009,
+      "step": 16898
+    },
+    {
+      "epoch": 0.74206614926946,
+      "grad_norm": 1.4453125,
+      "learning_rate": 7.812580139763646e-06,
+      "loss": 0.3434,
+      "step": 16900
+    },
+    {
+      "epoch": 0.7421539677486635,
+      "grad_norm": 1.6171875,
+      "learning_rate": 7.807556827358331e-06,
+      "loss": 0.3139,
+      "step": 16902
+    },
+    {
+      "epoch": 0.742241786227867,
+      "grad_norm": 1.46875,
+      "learning_rate": 7.802534831554437e-06,
+      "loss": 0.3411,
+      "step": 16904
+    },
+    {
+      "epoch": 0.7423296047070704,
+      "grad_norm": 1.4453125,
+      "learning_rate": 7.797514152736548e-06,
+      "loss": 0.3141,
+      "step": 16906
+    },
+    {
+      "epoch": 0.742417423186274,
+      "grad_norm": 1.4921875,
+      "learning_rate": 7.792494791289142e-06,
+      "loss": 0.301,
+      "step": 16908
+    },
+    {
+      "epoch": 0.7425052416654775,
+      "grad_norm": 1.4609375,
+      "learning_rate": 7.787476747596618e-06,
+      "loss": 0.3115,
+      "step": 16910
+    },
+    {
+      "epoch": 0.742593060144681,
+      "grad_norm": 1.5234375,
+      "learning_rate": 7.782460022043242e-06,
+      "loss": 0.3402,
+      "step": 16912
+    },
+    {
+      "epoch": 0.7426808786238844,
+      "grad_norm": 1.53125,
+      "learning_rate": 7.777444615013213e-06,
+      "loss": 0.3368,
+      "step": 16914
+    },
+    {
+      "epoch": 0.7427686971030879,
+      "grad_norm": 1.6015625,
+      "learning_rate": 7.772430526890603e-06,
+      "loss": 0.3026,
+      "step": 16916
+    },
+    {
+      "epoch": 0.7428565155822914,
+      "grad_norm": 1.5078125,
+      "learning_rate": 7.767417758059386e-06,
+      "loss": 0.3252,
+      "step": 16918
+    },
+    {
+      "epoch": 0.7429443340614948,
+      "grad_norm": 1.4921875,
+      "learning_rate": 7.762406308903458e-06,
+      "loss": 0.3343,
+      "step": 16920
+    },
+    {
+      "epoch": 0.7430321525406984,
+      "grad_norm": 1.5625,
+      "learning_rate": 7.757396179806576e-06,
+      "loss": 0.3512,
+      "step": 16922
+    },
+    {
+      "epoch": 0.7431199710199019,
+      "grad_norm": 1.734375,
+      "learning_rate": 7.752387371152436e-06,
+      "loss": 0.3558,
+      "step": 16924
+    },
+    {
+      "epoch": 0.7432077894991054,
+      "grad_norm": 1.4609375,
+      "learning_rate": 7.747379883324606e-06,
+      "loss": 0.3414,
+      "step": 16926
+    },
+    {
+      "epoch": 0.7432956079783088,
+      "grad_norm": 1.5,
+      "learning_rate": 7.742373716706556e-06,
+      "loss": 0.3374,
+      "step": 16928
+    },
+    {
+      "epoch": 0.7433834264575123,
+      "grad_norm": 1.5859375,
+      "learning_rate": 7.737368871681666e-06,
+      "loss": 0.3306,
+      "step": 16930
+    },
+    {
+      "epoch": 0.7434712449367158,
+      "grad_norm": 1.5390625,
+      "learning_rate": 7.732365348633203e-06,
+      "loss": 0.2976,
+      "step": 16932
+    },
+    {
+      "epoch": 0.7435590634159193,
+      "grad_norm": 1.46875,
+      "learning_rate": 7.727363147944352e-06,
+      "loss": 0.3012,
+      "step": 16934
+    },
+    {
+      "epoch": 0.7436468818951227,
+      "grad_norm": 1.46875,
+      "learning_rate": 7.722362269998159e-06,
+      "loss": 0.2872,
+      "step": 16936
+    },
+    {
+      "epoch": 0.7437347003743263,
+      "grad_norm": 1.5,
+      "learning_rate": 7.717362715177611e-06,
+      "loss": 0.3526,
+      "step": 16938
+    },
+    {
+      "epoch": 0.7438225188535298,
+      "grad_norm": 1.671875,
+      "learning_rate": 7.712364483865564e-06,
+      "loss": 0.3291,
+      "step": 16940
+    },
+    {
+      "epoch": 0.7439103373327333,
+      "grad_norm": 1.5546875,
+      "learning_rate": 7.707367576444796e-06,
+      "loss": 0.3077,
+      "step": 16942
+    },
+    {
+      "epoch": 0.7439981558119367,
+      "grad_norm": 1.4609375,
+      "learning_rate": 7.702371993297963e-06,
+      "loss": 0.3227,
+      "step": 16944
+    },
+    {
+      "epoch": 0.7440859742911402,
+      "grad_norm": 1.515625,
+      "learning_rate": 7.697377734807623e-06,
+      "loss": 0.319,
+      "step": 16946
+    },
+    {
+      "epoch": 0.7441737927703437,
+      "grad_norm": 1.53125,
+      "learning_rate": 7.69238480135625e-06,
+      "loss": 0.3222,
+      "step": 16948
+    },
+    {
+      "epoch": 0.7442616112495472,
+      "grad_norm": 1.5390625,
+      "learning_rate": 7.68739319332619e-06,
+      "loss": 0.3384,
+      "step": 16950
+    },
+    {
+      "epoch": 0.7443494297287506,
+      "grad_norm": 1.53125,
+      "learning_rate": 7.682402911099717e-06,
+      "loss": 0.2967,
+      "step": 16952
+    },
+    {
+      "epoch": 0.7444372482079542,
+      "grad_norm": 1.5390625,
+      "learning_rate": 7.677413955058982e-06,
+      "loss": 0.3151,
+      "step": 16954
+    },
+    {
+      "epoch": 0.7445250666871577,
+      "grad_norm": 1.5,
+      "learning_rate": 7.67242632558603e-06,
+      "loss": 0.309,
+      "step": 16956
+    },
+    {
+      "epoch": 0.7446128851663611,
+      "grad_norm": 1.5234375,
+      "learning_rate": 7.667440023062833e-06,
+      "loss": 0.3118,
+      "step": 16958
+    },
+    {
+      "epoch": 0.7447007036455646,
+      "grad_norm": 1.5703125,
+      "learning_rate": 7.662455047871226e-06,
+      "loss": 0.3044,
+      "step": 16960
+    },
+    {
+      "epoch": 0.7447885221247681,
+      "grad_norm": 1.5390625,
+      "learning_rate": 7.657471400392974e-06,
+      "loss": 0.3071,
+      "step": 16962
+    },
+    {
+      "epoch": 0.7448763406039716,
+      "grad_norm": 1.609375,
+      "learning_rate": 7.652489081009718e-06,
+      "loss": 0.3313,
+      "step": 16964
+    },
+    {
+      "epoch": 0.744964159083175,
+      "grad_norm": 1.453125,
+      "learning_rate": 7.647508090103009e-06,
+      "loss": 0.3014,
+      "step": 16966
+    },
+    {
+      "epoch": 0.7450519775623785,
+      "grad_norm": 1.5234375,
+      "learning_rate": 7.642528428054288e-06,
+      "loss": 0.3124,
+      "step": 16968
+    },
+    {
+      "epoch": 0.7451397960415821,
+      "grad_norm": 1.59375,
+      "learning_rate": 7.637550095244894e-06,
+      "loss": 0.2996,
+      "step": 16970
+    },
+    {
+      "epoch": 0.7452276145207856,
+      "grad_norm": 1.6328125,
+      "learning_rate": 7.632573092056086e-06,
+      "loss": 0.3438,
+      "step": 16972
+    },
+    {
+      "epoch": 0.745315432999989,
+      "grad_norm": 1.53125,
+      "learning_rate": 7.627597418868984e-06,
+      "loss": 0.3348,
+      "step": 16974
+    },
+    {
+      "epoch": 0.7454032514791925,
+      "grad_norm": 1.546875,
+      "learning_rate": 7.622623076064645e-06,
+      "loss": 0.3366,
+      "step": 16976
+    },
+    {
+      "epoch": 0.745491069958396,
+      "grad_norm": 1.6640625,
+      "learning_rate": 7.617650064023996e-06,
+      "loss": 0.3047,
+      "step": 16978
+    },
+    {
+      "epoch": 0.7455788884375995,
+      "grad_norm": 1.5546875,
+      "learning_rate": 7.6126783831278605e-06,
+      "loss": 0.3447,
+      "step": 16980
+    },
+    {
+      "epoch": 0.7456667069168029,
+      "grad_norm": 1.5625,
+      "learning_rate": 7.607708033756994e-06,
+      "loss": 0.3101,
+      "step": 16982
+    },
+    {
+      "epoch": 0.7457545253960065,
+      "grad_norm": 1.546875,
+      "learning_rate": 7.602739016292007e-06,
+      "loss": 0.31,
+      "step": 16984
+    },
+    {
+      "epoch": 0.74584234387521,
+      "grad_norm": 1.5625,
+      "learning_rate": 7.5977713311134454e-06,
+      "loss": 0.3373,
+      "step": 16986
+    },
+    {
+      "epoch": 0.7459301623544135,
+      "grad_norm": 1.640625,
+      "learning_rate": 7.592804978601725e-06,
+      "loss": 0.3055,
+      "step": 16988
+    },
+    {
+      "epoch": 0.7460179808336169,
+      "grad_norm": 1.53125,
+      "learning_rate": 7.587839959137166e-06,
+      "loss": 0.3276,
+      "step": 16990
+    },
+    {
+      "epoch": 0.7461057993128204,
+      "grad_norm": 1.6015625,
+      "learning_rate": 7.582876273100004e-06,
+      "loss": 0.3149,
+      "step": 16992
+    },
+    {
+      "epoch": 0.7461936177920239,
+      "grad_norm": 1.5703125,
+      "learning_rate": 7.5779139208703446e-06,
+      "loss": 0.3327,
+      "step": 16994
+    },
+    {
+      "epoch": 0.7462814362712273,
+      "grad_norm": 1.5078125,
+      "learning_rate": 7.572952902828229e-06,
+      "loss": 0.3176,
+      "step": 16996
+    },
+    {
+      "epoch": 0.7463692547504308,
+      "grad_norm": 1.53125,
+      "learning_rate": 7.567993219353542e-06,
+      "loss": 0.3338,
+      "step": 16998
+    },
+    {
+      "epoch": 0.7464570732296344,
+      "grad_norm": 1.5078125,
+      "learning_rate": 7.563034870826121e-06,
+      "loss": 0.3115,
+      "step": 17000
+    },
+    {
+      "epoch": 0.7465448917088379,
+      "grad_norm": 1.578125,
+      "learning_rate": 7.55807785762567e-06,
+      "loss": 0.3558,
+      "step": 17002
+    },
+    {
+      "epoch": 0.7466327101880413,
+      "grad_norm": 1.5234375,
+      "learning_rate": 7.553122180131788e-06,
+      "loss": 0.3067,
+      "step": 17004
+    },
+    {
+      "epoch": 0.7467205286672448,
+      "grad_norm": 1.5390625,
+      "learning_rate": 7.548167838724002e-06,
+      "loss": 0.3201,
+      "step": 17006
+    },
+    {
+      "epoch": 0.7468083471464483,
+      "grad_norm": 1.5390625,
+      "learning_rate": 7.543214833781695e-06,
+      "loss": 0.3435,
+      "step": 17008
+    },
+    {
+      "epoch": 0.7468961656256518,
+      "grad_norm": 1.671875,
+      "learning_rate": 7.538263165684192e-06,
+      "loss": 0.3126,
+      "step": 17010
+    },
+    {
+      "epoch": 0.7469839841048552,
+      "grad_norm": 1.5078125,
+      "learning_rate": 7.533312834810672e-06,
+      "loss": 0.3172,
+      "step": 17012
+    },
+    {
+      "epoch": 0.7470718025840587,
+      "grad_norm": 1.578125,
+      "learning_rate": 7.5283638415402505e-06,
+      "loss": 0.3332,
+      "step": 17014
+    },
+    {
+      "epoch": 0.7471596210632623,
+      "grad_norm": 1.5625,
+      "learning_rate": 7.523416186251917e-06,
+      "loss": 0.3223,
+      "step": 17016
+    },
+    {
+      "epoch": 0.7472474395424658,
+      "grad_norm": 1.484375,
+      "learning_rate": 7.518469869324548e-06,
+      "loss": 0.3223,
+      "step": 17018
+    },
+    {
+      "epoch": 0.7473352580216692,
+      "grad_norm": 1.5546875,
+      "learning_rate": 7.513524891136958e-06,
+      "loss": 0.3237,
+      "step": 17020
+    },
+    {
+      "epoch": 0.7474230765008727,
+      "grad_norm": 1.546875,
+      "learning_rate": 7.5085812520678174e-06,
+      "loss": 0.3332,
+      "step": 17022
+    },
+    {
+      "epoch": 0.7475108949800762,
+      "grad_norm": 1.5390625,
+      "learning_rate": 7.503638952495723e-06,
+      "loss": 0.3227,
+      "step": 17024
+    },
+    {
+      "epoch": 0.7475987134592796,
+      "grad_norm": 1.609375,
+      "learning_rate": 7.498697992799153e-06,
+      "loss": 0.3195,
+      "step": 17026
+    },
+    {
+      "epoch": 0.7476865319384831,
+      "grad_norm": 1.53125,
+      "learning_rate": 7.4937583733564855e-06,
+      "loss": 0.3361,
+      "step": 17028
+    },
+    {
+      "epoch": 0.7477743504176867,
+      "grad_norm": 1.6328125,
+      "learning_rate": 7.488820094545998e-06,
+      "loss": 0.2991,
+      "step": 17030
+    },
+    {
+      "epoch": 0.7478621688968902,
+      "grad_norm": 1.5390625,
+      "learning_rate": 7.483883156745858e-06,
+      "loss": 0.3425,
+      "step": 17032
+    },
+    {
+      "epoch": 0.7479499873760936,
+      "grad_norm": 1.484375,
+      "learning_rate": 7.478947560334151e-06,
+      "loss": 0.3203,
+      "step": 17034
+    },
+    {
+      "epoch": 0.7480378058552971,
+      "grad_norm": 1.5390625,
+      "learning_rate": 7.4740133056888345e-06,
+      "loss": 0.3273,
+      "step": 17036
+    },
+    {
+      "epoch": 0.7481256243345006,
+      "grad_norm": 1.515625,
+      "learning_rate": 7.469080393187786e-06,
+      "loss": 0.3265,
+      "step": 17038
+    },
+    {
+      "epoch": 0.748213442813704,
+      "grad_norm": 1.71875,
+      "learning_rate": 7.464148823208764e-06,
+      "loss": 0.3364,
+      "step": 17040
+    },
+    {
+      "epoch": 0.7483012612929075,
+      "grad_norm": 1.4921875,
+      "learning_rate": 7.459218596129422e-06,
+      "loss": 0.2967,
+      "step": 17042
+    },
+    {
+      "epoch": 0.748389079772111,
+      "grad_norm": 1.5546875,
+      "learning_rate": 7.454289712327333e-06,
+      "loss": 0.3519,
+      "step": 17044
+    },
+    {
+      "epoch": 0.7484768982513146,
+      "grad_norm": 1.5859375,
+      "learning_rate": 7.449362172179936e-06,
+      "loss": 0.3004,
+      "step": 17046
+    },
+    {
+      "epoch": 0.748564716730518,
+      "grad_norm": 1.6171875,
+      "learning_rate": 7.444435976064595e-06,
+      "loss": 0.2956,
+      "step": 17048
+    },
+    {
+      "epoch": 0.7486525352097215,
+      "grad_norm": 1.4609375,
+      "learning_rate": 7.439511124358558e-06,
+      "loss": 0.3085,
+      "step": 17050
+    },
+    {
+      "epoch": 0.748740353688925,
+      "grad_norm": 1.6171875,
+      "learning_rate": 7.434587617438962e-06,
+      "loss": 0.3035,
+      "step": 17052
+    },
+    {
+      "epoch": 0.7488281721681285,
+      "grad_norm": 1.484375,
+      "learning_rate": 7.4296654556828635e-06,
+      "loss": 0.307,
+      "step": 17054
+    },
+    {
+      "epoch": 0.7489159906473319,
+      "grad_norm": 1.53125,
+      "learning_rate": 7.424744639467196e-06,
+      "loss": 0.3258,
+      "step": 17056
+    },
+    {
+      "epoch": 0.7490038091265354,
+      "grad_norm": 1.515625,
+      "learning_rate": 7.419825169168798e-06,
+      "loss": 0.3134,
+      "step": 17058
+    },
+    {
+      "epoch": 0.7490916276057389,
+      "grad_norm": 1.46875,
+      "learning_rate": 7.4149070451643955e-06,
+      "loss": 0.3302,
+      "step": 17060
+    },
+    {
+      "epoch": 0.7491794460849425,
+      "grad_norm": 1.484375,
+      "learning_rate": 7.4099902678306324e-06,
+      "loss": 0.3126,
+      "step": 17062
+    },
+    {
+      "epoch": 0.7492672645641459,
+      "grad_norm": 1.5859375,
+      "learning_rate": 7.405074837544035e-06,
+      "loss": 0.3404,
+      "step": 17064
+    },
+    {
+      "epoch": 0.7493550830433494,
+      "grad_norm": 1.4609375,
+      "learning_rate": 7.400160754681012e-06,
+      "loss": 0.3275,
+      "step": 17066
+    },
+    {
+      "epoch": 0.7494429015225529,
+      "grad_norm": 1.578125,
+      "learning_rate": 7.3952480196179094e-06,
+      "loss": 0.3284,
+      "step": 17068
+    },
+    {
+      "epoch": 0.7495307200017564,
+      "grad_norm": 1.484375,
+      "learning_rate": 7.3903366327309235e-06,
+      "loss": 0.3025,
+      "step": 17070
+    },
+    {
+      "epoch": 0.7496185384809598,
+      "grad_norm": 1.4765625,
+      "learning_rate": 7.385426594396186e-06,
+      "loss": 0.3241,
+      "step": 17072
+    },
+    {
+      "epoch": 0.7497063569601633,
+      "grad_norm": 1.59375,
+      "learning_rate": 7.3805179049896975e-06,
+      "loss": 0.3365,
+      "step": 17074
+    },
+    {
+      "epoch": 0.7497941754393669,
+      "grad_norm": 1.65625,
+      "learning_rate": 7.375610564887378e-06,
+      "loss": 0.3483,
+      "step": 17076
+    },
+    {
+      "epoch": 0.7498819939185704,
+      "grad_norm": 1.46875,
+      "learning_rate": 7.3707045744650265e-06,
+      "loss": 0.31,
+      "step": 17078
+    },
+    {
+      "epoch": 0.7499698123977738,
+      "grad_norm": 1.6640625,
+      "learning_rate": 7.365799934098336e-06,
+      "loss": 0.3239,
+      "step": 17080
+    },
+    {
+      "epoch": 0.7500576308769773,
+      "grad_norm": 1.515625,
+      "learning_rate": 7.360896644162924e-06,
+      "loss": 0.3389,
+      "step": 17082
+    },
+    {
+      "epoch": 0.7501454493561808,
+      "grad_norm": 1.5625,
+      "learning_rate": 7.355994705034267e-06,
+      "loss": 0.3029,
+      "step": 17084
+    },
+    {
+      "epoch": 0.7502332678353842,
+      "grad_norm": 1.40625,
+      "learning_rate": 7.35109411708777e-06,
+      "loss": 0.3065,
+      "step": 17086
+    },
+    {
+      "epoch": 0.7503210863145877,
+      "grad_norm": 1.578125,
+      "learning_rate": 7.346194880698718e-06,
+      "loss": 0.3347,
+      "step": 17088
+    },
+    {
+      "epoch": 0.7504089047937912,
+      "grad_norm": 1.6484375,
+      "learning_rate": 7.341296996242295e-06,
+      "loss": 0.3161,
+      "step": 17090
+    },
+    {
+      "epoch": 0.7504967232729948,
+      "grad_norm": 1.5625,
+      "learning_rate": 7.336400464093579e-06,
+      "loss": 0.318,
+      "step": 17092
+    },
+    {
+      "epoch": 0.7505845417521982,
+      "grad_norm": 1.5078125,
+      "learning_rate": 7.331505284627543e-06,
+      "loss": 0.3086,
+      "step": 17094
+    },
+    {
+      "epoch": 0.7506723602314017,
+      "grad_norm": 1.578125,
+      "learning_rate": 7.326611458219077e-06,
+      "loss": 0.3083,
+      "step": 17096
+    },
+    {
+      "epoch": 0.7507601787106052,
+      "grad_norm": 1.546875,
+      "learning_rate": 7.321718985242931e-06,
+      "loss": 0.3355,
+      "step": 17098
+    },
+    {
+      "epoch": 0.7508479971898087,
+      "grad_norm": 1.828125,
+      "learning_rate": 7.316827866073794e-06,
+      "loss": 0.3235,
+      "step": 17100
+    },
+    {
+      "epoch": 0.7509358156690121,
+      "grad_norm": 1.59375,
+      "learning_rate": 7.3119381010862155e-06,
+      "loss": 0.2957,
+      "step": 17102
+    },
+    {
+      "epoch": 0.7510236341482156,
+      "grad_norm": 1.484375,
+      "learning_rate": 7.307049690654649e-06,
+      "loss": 0.3141,
+      "step": 17104
+    },
+    {
+      "epoch": 0.7511114526274191,
+      "grad_norm": 1.5390625,
+      "learning_rate": 7.30216263515347e-06,
+      "loss": 0.3129,
+      "step": 17106
+    },
+    {
+      "epoch": 0.7511992711066227,
+      "grad_norm": 1.53125,
+      "learning_rate": 7.297276934956909e-06,
+      "loss": 0.3318,
+      "step": 17108
+    },
+    {
+      "epoch": 0.7512870895858261,
+      "grad_norm": 1.6640625,
+      "learning_rate": 7.292392590439132e-06,
+      "loss": 0.3171,
+      "step": 17110
+    },
+    {
+      "epoch": 0.7513749080650296,
+      "grad_norm": 1.625,
+      "learning_rate": 7.287509601974174e-06,
+      "loss": 0.3326,
+      "step": 17112
+    },
+    {
+      "epoch": 0.7514627265442331,
+      "grad_norm": 1.53125,
+      "learning_rate": 7.28262796993597e-06,
+      "loss": 0.3218,
+      "step": 17114
+    },
+    {
+      "epoch": 0.7515505450234365,
+      "grad_norm": 1.5546875,
+      "learning_rate": 7.2777476946983696e-06,
+      "loss": 0.3128,
+      "step": 17116
+    },
+    {
+      "epoch": 0.75163836350264,
+      "grad_norm": 1.5703125,
+      "learning_rate": 7.2728687766351e-06,
+      "loss": 0.3404,
+      "step": 17118
+    },
+    {
+      "epoch": 0.7517261819818435,
+      "grad_norm": 1.6953125,
+      "learning_rate": 7.267991216119791e-06,
+      "loss": 0.3249,
+      "step": 17120
+    },
+    {
+      "epoch": 0.751814000461047,
+      "grad_norm": 1.65625,
+      "learning_rate": 7.263115013525956e-06,
+      "loss": 0.3063,
+      "step": 17122
+    },
+    {
+      "epoch": 0.7519018189402505,
+      "grad_norm": 1.640625,
+      "learning_rate": 7.258240169227032e-06,
+      "loss": 0.3196,
+      "step": 17124
+    },
+    {
+      "epoch": 0.751989637419454,
+      "grad_norm": 1.71875,
+      "learning_rate": 7.25336668359633e-06,
+      "loss": 0.3088,
+      "step": 17126
+    },
+    {
+      "epoch": 0.7520774558986575,
+      "grad_norm": 1.4609375,
+      "learning_rate": 7.248494557007051e-06,
+      "loss": 0.3015,
+      "step": 17128
+    },
+    {
+      "epoch": 0.752165274377861,
+      "grad_norm": 1.65625,
+      "learning_rate": 7.2436237898323236e-06,
+      "loss": 0.3299,
+      "step": 17130
+    },
+    {
+      "epoch": 0.7522530928570644,
+      "grad_norm": 1.5234375,
+      "learning_rate": 7.238754382445137e-06,
+      "loss": 0.3355,
+      "step": 17132
+    },
+    {
+      "epoch": 0.7523409113362679,
+      "grad_norm": 1.4921875,
+      "learning_rate": 7.233886335218404e-06,
+      "loss": 0.3233,
+      "step": 17134
+    },
+    {
+      "epoch": 0.7524287298154714,
+      "grad_norm": 1.5390625,
+      "learning_rate": 7.2290196485249155e-06,
+      "loss": 0.3121,
+      "step": 17136
+    },
+    {
+      "epoch": 0.752516548294675,
+      "grad_norm": 1.453125,
+      "learning_rate": 7.22415432273735e-06,
+      "loss": 0.3085,
+      "step": 17138
+    },
+    {
+      "epoch": 0.7526043667738784,
+      "grad_norm": 1.5390625,
+      "learning_rate": 7.21929035822832e-06,
+      "loss": 0.3517,
+      "step": 17140
+    },
+    {
+      "epoch": 0.7526921852530819,
+      "grad_norm": 1.484375,
+      "learning_rate": 7.214427755370287e-06,
+      "loss": 0.3077,
+      "step": 17142
+    },
+    {
+      "epoch": 0.7527800037322854,
+      "grad_norm": 1.4453125,
+      "learning_rate": 7.209566514535648e-06,
+      "loss": 0.3223,
+      "step": 17144
+    },
+    {
+      "epoch": 0.7528678222114888,
+      "grad_norm": 1.5234375,
+      "learning_rate": 7.204706636096664e-06,
+      "loss": 0.3242,
+      "step": 17146
+    },
+    {
+      "epoch": 0.7529556406906923,
+      "grad_norm": 1.9296875,
+      "learning_rate": 7.199848120425526e-06,
+      "loss": 0.3434,
+      "step": 17148
+    },
+    {
+      "epoch": 0.7530434591698958,
+      "grad_norm": 1.546875,
+      "learning_rate": 7.194990967894269e-06,
+      "loss": 0.298,
+      "step": 17150
+    },
+    {
+      "epoch": 0.7531312776490993,
+      "grad_norm": 1.640625,
+      "learning_rate": 7.1901351788748795e-06,
+      "loss": 0.288,
+      "step": 17152
+    },
+    {
+      "epoch": 0.7532190961283028,
+      "grad_norm": 1.5234375,
+      "learning_rate": 7.1852807537392095e-06,
+      "loss": 0.3107,
+      "step": 17154
+    },
+    {
+      "epoch": 0.7533069146075063,
+      "grad_norm": 1.5234375,
+      "learning_rate": 7.1804276928590015e-06,
+      "loss": 0.2997,
+      "step": 17156
+    },
+    {
+      "epoch": 0.7533947330867098,
+      "grad_norm": 1.6015625,
+      "learning_rate": 7.175575996605918e-06,
+      "loss": 0.2898,
+      "step": 17158
+    },
+    {
+      "epoch": 0.7534825515659133,
+      "grad_norm": 1.4765625,
+      "learning_rate": 7.170725665351493e-06,
+      "loss": 0.3254,
+      "step": 17160
+    },
+    {
+      "epoch": 0.7535703700451167,
+      "grad_norm": 1.5703125,
+      "learning_rate": 7.165876699467175e-06,
+      "loss": 0.3258,
+      "step": 17162
+    },
+    {
+      "epoch": 0.7536581885243202,
+      "grad_norm": 1.4765625,
+      "learning_rate": 7.161029099324299e-06,
+      "loss": 0.3426,
+      "step": 17164
+    },
+    {
+      "epoch": 0.7537460070035237,
+      "grad_norm": 1.4375,
+      "learning_rate": 7.156182865294078e-06,
+      "loss": 0.2992,
+      "step": 17166
+    },
+    {
+      "epoch": 0.7538338254827271,
+      "grad_norm": 1.5078125,
+      "learning_rate": 7.151337997747662e-06,
+      "loss": 0.3336,
+      "step": 17168
+    },
+    {
+      "epoch": 0.7539216439619307,
+      "grad_norm": 1.546875,
+      "learning_rate": 7.14649449705605e-06,
+      "loss": 0.3151,
+      "step": 17170
+    },
+    {
+      "epoch": 0.7540094624411342,
+      "grad_norm": 1.515625,
+      "learning_rate": 7.1416523635901785e-06,
+      "loss": 0.2882,
+      "step": 17172
+    },
+    {
+      "epoch": 0.7540972809203377,
+      "grad_norm": 1.5,
+      "learning_rate": 7.136811597720852e-06,
+      "loss": 0.3184,
+      "step": 17174
+    },
+    {
+      "epoch": 0.7541850993995411,
+      "grad_norm": 1.515625,
+      "learning_rate": 7.131972199818765e-06,
+      "loss": 0.3324,
+      "step": 17176
+    },
+    {
+      "epoch": 0.7542729178787446,
+      "grad_norm": 1.46875,
+      "learning_rate": 7.12713417025454e-06,
+      "loss": 0.323,
+      "step": 17178
+    },
+    {
+      "epoch": 0.7543607363579481,
+      "grad_norm": 1.6328125,
+      "learning_rate": 7.122297509398662e-06,
+      "loss": 0.3348,
+      "step": 17180
+    },
+    {
+      "epoch": 0.7544485548371516,
+      "grad_norm": 1.546875,
+      "learning_rate": 7.117462217621529e-06,
+      "loss": 0.3198,
+      "step": 17182
+    },
+    {
+      "epoch": 0.7545363733163551,
+      "grad_norm": 1.5234375,
+      "learning_rate": 7.112628295293417e-06,
+      "loss": 0.318,
+      "step": 17184
+    },
+    {
+      "epoch": 0.7546241917955586,
+      "grad_norm": 1.4609375,
+      "learning_rate": 7.107795742784526e-06,
+      "loss": 0.2928,
+      "step": 17186
+    },
+    {
+      "epoch": 0.7547120102747621,
+      "grad_norm": 1.53125,
+      "learning_rate": 7.102964560464925e-06,
+      "loss": 0.3126,
+      "step": 17188
+    },
+    {
+      "epoch": 0.7547998287539656,
+      "grad_norm": 1.7421875,
+      "learning_rate": 7.0981347487045825e-06,
+      "loss": 0.3456,
+      "step": 17190
+    },
+    {
+      "epoch": 0.754887647233169,
+      "grad_norm": 1.6484375,
+      "learning_rate": 7.093306307873376e-06,
+      "loss": 0.3137,
+      "step": 17192
+    },
+    {
+      "epoch": 0.7549754657123725,
+      "grad_norm": 1.5859375,
+      "learning_rate": 7.0884792383410615e-06,
+      "loss": 0.3314,
+      "step": 17194
+    },
+    {
+      "epoch": 0.755063284191576,
+      "grad_norm": 1.625,
+      "learning_rate": 7.083653540477306e-06,
+      "loss": 0.3567,
+      "step": 17196
+    },
+    {
+      "epoch": 0.7551511026707795,
+      "grad_norm": 1.484375,
+      "learning_rate": 7.078829214651658e-06,
+      "loss": 0.308,
+      "step": 17198
+    },
+    {
+      "epoch": 0.755238921149983,
+      "grad_norm": 1.4453125,
+      "learning_rate": 7.074006261233559e-06,
+      "loss": 0.3441,
+      "step": 17200
+    },
+    {
+      "epoch": 0.7553267396291865,
+      "grad_norm": 1.5703125,
+      "learning_rate": 7.0691846805923635e-06,
+      "loss": 0.3215,
+      "step": 17202
+    },
+    {
+      "epoch": 0.75541455810839,
+      "grad_norm": 1.5,
+      "learning_rate": 7.064364473097296e-06,
+      "loss": 0.3256,
+      "step": 17204
+    },
+    {
+      "epoch": 0.7555023765875934,
+      "grad_norm": 1.4453125,
+      "learning_rate": 7.059545639117504e-06,
+      "loss": 0.3148,
+      "step": 17206
+    },
+    {
+      "epoch": 0.7555901950667969,
+      "grad_norm": 1.46875,
+      "learning_rate": 7.054728179021999e-06,
+      "loss": 0.3193,
+      "step": 17208
+    },
+    {
+      "epoch": 0.7556780135460004,
+      "grad_norm": 1.5859375,
+      "learning_rate": 7.049912093179728e-06,
+      "loss": 0.2858,
+      "step": 17210
+    },
+    {
+      "epoch": 0.7557658320252039,
+      "grad_norm": 1.515625,
+      "learning_rate": 7.0450973819594785e-06,
+      "loss": 0.3166,
+      "step": 17212
+    },
+    {
+      "epoch": 0.7558536505044073,
+      "grad_norm": 1.609375,
+      "learning_rate": 7.04028404572998e-06,
+      "loss": 0.3183,
+      "step": 17214
+    },
+    {
+      "epoch": 0.7559414689836109,
+      "grad_norm": 1.53125,
+      "learning_rate": 7.035472084859837e-06,
+      "loss": 0.299,
+      "step": 17216
+    },
+    {
+      "epoch": 0.7560292874628144,
+      "grad_norm": 1.5546875,
+      "learning_rate": 7.030661499717539e-06,
+      "loss": 0.3222,
+      "step": 17218
+    },
+    {
+      "epoch": 0.7561171059420179,
+      "grad_norm": 1.484375,
+      "learning_rate": 7.0258522906715e-06,
+      "loss": 0.3416,
+      "step": 17220
+    },
+    {
+      "epoch": 0.7562049244212213,
+      "grad_norm": 1.4921875,
+      "learning_rate": 7.0210444580899925e-06,
+      "loss": 0.3432,
+      "step": 17222
+    },
+    {
+      "epoch": 0.7562927429004248,
+      "grad_norm": 1.5625,
+      "learning_rate": 7.016238002341219e-06,
+      "loss": 0.316,
+      "step": 17224
+    },
+    {
+      "epoch": 0.7563805613796283,
+      "grad_norm": 1.4375,
+      "learning_rate": 7.0114329237932485e-06,
+      "loss": 0.3184,
+      "step": 17226
+    },
+    {
+      "epoch": 0.7564683798588318,
+      "grad_norm": 1.53125,
+      "learning_rate": 7.006629222814048e-06,
+      "loss": 0.3126,
+      "step": 17228
+    },
+    {
+      "epoch": 0.7565561983380353,
+      "grad_norm": 1.4453125,
+      "learning_rate": 7.001826899771505e-06,
+      "loss": 0.3099,
+      "step": 17230
+    },
+    {
+      "epoch": 0.7566440168172388,
+      "grad_norm": 1.4609375,
+      "learning_rate": 6.997025955033365e-06,
+      "loss": 0.3084,
+      "step": 17232
+    },
+    {
+      "epoch": 0.7567318352964423,
+      "grad_norm": 1.6171875,
+      "learning_rate": 6.992226388967302e-06,
+      "loss": 0.3072,
+      "step": 17234
+    },
+    {
+      "epoch": 0.7568196537756458,
+      "grad_norm": 1.5078125,
+      "learning_rate": 6.987428201940854e-06,
+      "loss": 0.3085,
+      "step": 17236
+    },
+    {
+      "epoch": 0.7569074722548492,
+      "grad_norm": 1.5546875,
+      "learning_rate": 6.982631394321468e-06,
+      "loss": 0.3344,
+      "step": 17238
+    },
+    {
+      "epoch": 0.7569952907340527,
+      "grad_norm": 1.6171875,
+      "learning_rate": 6.977835966476503e-06,
+      "loss": 0.3099,
+      "step": 17240
+    },
+    {
+      "epoch": 0.7570831092132562,
+      "grad_norm": 1.5703125,
+      "learning_rate": 6.973041918773168e-06,
+      "loss": 0.3596,
+      "step": 17242
+    },
+    {
+      "epoch": 0.7571709276924596,
+      "grad_norm": 1.4765625,
+      "learning_rate": 6.9682492515786096e-06,
+      "loss": 0.3381,
+      "step": 17244
+    },
+    {
+      "epoch": 0.7572587461716632,
+      "grad_norm": 1.5546875,
+      "learning_rate": 6.963457965259837e-06,
+      "loss": 0.3254,
+      "step": 17246
+    },
+    {
+      "epoch": 0.7573465646508667,
+      "grad_norm": 1.5625,
+      "learning_rate": 6.958668060183785e-06,
+      "loss": 0.3155,
+      "step": 17248
+    },
+    {
+      "epoch": 0.7574343831300702,
+      "grad_norm": 1.515625,
+      "learning_rate": 6.953879536717259e-06,
+      "loss": 0.3348,
+      "step": 17250
+    },
+    {
+      "epoch": 0.7575222016092736,
+      "grad_norm": 1.578125,
+      "learning_rate": 6.949092395226955e-06,
+      "loss": 0.3602,
+      "step": 17252
+    },
+    {
+      "epoch": 0.7576100200884771,
+      "grad_norm": 1.4765625,
+      "learning_rate": 6.944306636079492e-06,
+      "loss": 0.2985,
+      "step": 17254
+    },
+    {
+      "epoch": 0.7576978385676806,
+      "grad_norm": 1.53125,
+      "learning_rate": 6.939522259641346e-06,
+      "loss": 0.2893,
+      "step": 17256
+    },
+    {
+      "epoch": 0.757785657046884,
+      "grad_norm": 1.5390625,
+      "learning_rate": 6.934739266278923e-06,
+      "loss": 0.3334,
+      "step": 17258
+    },
+    {
+      "epoch": 0.7578734755260875,
+      "grad_norm": 1.53125,
+      "learning_rate": 6.929957656358496e-06,
+      "loss": 0.3047,
+      "step": 17260
+    },
+    {
+      "epoch": 0.7579612940052911,
+      "grad_norm": 1.4921875,
+      "learning_rate": 6.925177430246238e-06,
+      "loss": 0.3304,
+      "step": 17262
+    },
+    {
+      "epoch": 0.7580491124844946,
+      "grad_norm": 1.5859375,
+      "learning_rate": 6.920398588308233e-06,
+      "loss": 0.3467,
+      "step": 17264
+    },
+    {
+      "epoch": 0.758136930963698,
+      "grad_norm": 1.515625,
+      "learning_rate": 6.915621130910427e-06,
+      "loss": 0.3235,
+      "step": 17266
+    },
+    {
+      "epoch": 0.7582247494429015,
+      "grad_norm": 1.515625,
+      "learning_rate": 6.9108450584187e-06,
+      "loss": 0.3322,
+      "step": 17268
+    },
+    {
+      "epoch": 0.758312567922105,
+      "grad_norm": 1.546875,
+      "learning_rate": 6.9060703711987944e-06,
+      "loss": 0.3114,
+      "step": 17270
+    },
+    {
+      "epoch": 0.7584003864013085,
+      "grad_norm": 1.5,
+      "learning_rate": 6.9012970696163585e-06,
+      "loss": 0.3164,
+      "step": 17272
+    },
+    {
+      "epoch": 0.7584882048805119,
+      "grad_norm": 1.6328125,
+      "learning_rate": 6.896525154036923e-06,
+      "loss": 0.323,
+      "step": 17274
+    },
+    {
+      "epoch": 0.7585760233597155,
+      "grad_norm": 1.578125,
+      "learning_rate": 6.891754624825939e-06,
+      "loss": 0.3231,
+      "step": 17276
+    },
+    {
+      "epoch": 0.758663841838919,
+      "grad_norm": 1.4921875,
+      "learning_rate": 6.886985482348726e-06,
+      "loss": 0.3322,
+      "step": 17278
+    },
+    {
+      "epoch": 0.7587516603181225,
+      "grad_norm": 1.609375,
+      "learning_rate": 6.8822177269704965e-06,
+      "loss": 0.3385,
+      "step": 17280
+    },
+    {
+      "epoch": 0.7588394787973259,
+      "grad_norm": 1.609375,
+      "learning_rate": 6.877451359056389e-06,
+      "loss": 0.3327,
+      "step": 17282
+    },
+    {
+      "epoch": 0.7589272972765294,
+      "grad_norm": 1.59375,
+      "learning_rate": 6.872686378971391e-06,
+      "loss": 0.3015,
+      "step": 17284
+    },
+    {
+      "epoch": 0.7590151157557329,
+      "grad_norm": 1.640625,
+      "learning_rate": 6.867922787080422e-06,
+      "loss": 0.3224,
+      "step": 17286
+    },
+    {
+      "epoch": 0.7591029342349364,
+      "grad_norm": 1.515625,
+      "learning_rate": 6.863160583748274e-06,
+      "loss": 0.3328,
+      "step": 17288
+    },
+    {
+      "epoch": 0.7591907527141398,
+      "grad_norm": 1.6875,
+      "learning_rate": 6.858399769339627e-06,
+      "loss": 0.3141,
+      "step": 17290
+    },
+    {
+      "epoch": 0.7592785711933434,
+      "grad_norm": 1.5546875,
+      "learning_rate": 6.853640344219084e-06,
+      "loss": 0.3176,
+      "step": 17292
+    },
+    {
+      "epoch": 0.7593663896725469,
+      "grad_norm": 1.4921875,
+      "learning_rate": 6.848882308751106e-06,
+      "loss": 0.2885,
+      "step": 17294
+    },
+    {
+      "epoch": 0.7594542081517504,
+      "grad_norm": 1.484375,
+      "learning_rate": 6.844125663300077e-06,
+      "loss": 0.3246,
+      "step": 17296
+    },
+    {
+      "epoch": 0.7595420266309538,
+      "grad_norm": 1.5859375,
+      "learning_rate": 6.839370408230259e-06,
+      "loss": 0.3065,
+      "step": 17298
+    },
+    {
+      "epoch": 0.7596298451101573,
+      "grad_norm": 1.5078125,
+      "learning_rate": 6.8346165439058e-06,
+      "loss": 0.3019,
+      "step": 17300
+    },
+    {
+      "epoch": 0.7597176635893608,
+      "grad_norm": 1.5859375,
+      "learning_rate": 6.829864070690778e-06,
+      "loss": 0.3003,
+      "step": 17302
+    },
+    {
+      "epoch": 0.7598054820685642,
+      "grad_norm": 1.53125,
+      "learning_rate": 6.825112988949103e-06,
+      "loss": 0.3029,
+      "step": 17304
+    },
+    {
+      "epoch": 0.7598933005477677,
+      "grad_norm": 1.5234375,
+      "learning_rate": 6.820363299044641e-06,
+      "loss": 0.3294,
+      "step": 17306
+    },
+    {
+      "epoch": 0.7599811190269713,
+      "grad_norm": 1.5625,
+      "learning_rate": 6.815615001341108e-06,
+      "loss": 0.2908,
+      "step": 17308
+    },
+    {
+      "epoch": 0.7600689375061748,
+      "grad_norm": 1.53125,
+      "learning_rate": 6.810868096202144e-06,
+      "loss": 0.3225,
+      "step": 17310
+    },
+    {
+      "epoch": 0.7601567559853782,
+      "grad_norm": 1.46875,
+      "learning_rate": 6.806122583991264e-06,
+      "loss": 0.3106,
+      "step": 17312
+    },
+    {
+      "epoch": 0.7602445744645817,
+      "grad_norm": 1.6171875,
+      "learning_rate": 6.801378465071867e-06,
+      "loss": 0.3209,
+      "step": 17314
+    },
+    {
+      "epoch": 0.7603323929437852,
+      "grad_norm": 1.5546875,
+      "learning_rate": 6.7966357398072804e-06,
+      "loss": 0.3088,
+      "step": 17316
+    },
+    {
+      "epoch": 0.7604202114229887,
+      "grad_norm": 1.5,
+      "learning_rate": 6.791894408560681e-06,
+      "loss": 0.3274,
+      "step": 17318
+    },
+    {
+      "epoch": 0.7605080299021921,
+      "grad_norm": 1.5703125,
+      "learning_rate": 6.787154471695184e-06,
+      "loss": 0.3167,
+      "step": 17320
+    },
+    {
+      "epoch": 0.7605958483813956,
+      "grad_norm": 1.4609375,
+      "learning_rate": 6.7824159295737625e-06,
+      "loss": 0.3075,
+      "step": 17322
+    },
+    {
+      "epoch": 0.7606836668605992,
+      "grad_norm": 1.515625,
+      "learning_rate": 6.777678782559288e-06,
+      "loss": 0.3066,
+      "step": 17324
+    },
+    {
+      "epoch": 0.7607714853398027,
+      "grad_norm": 1.453125,
+      "learning_rate": 6.772943031014548e-06,
+      "loss": 0.2794,
+      "step": 17326
+    },
+    {
+      "epoch": 0.7608593038190061,
+      "grad_norm": 1.625,
+      "learning_rate": 6.768208675302193e-06,
+      "loss": 0.3188,
+      "step": 17328
+    },
+    {
+      "epoch": 0.7609471222982096,
+      "grad_norm": 1.4296875,
+      "learning_rate": 6.763475715784795e-06,
+      "loss": 0.2992,
+      "step": 17330
+    },
+    {
+      "epoch": 0.7610349407774131,
+      "grad_norm": 1.625,
+      "learning_rate": 6.758744152824798e-06,
+      "loss": 0.3023,
+      "step": 17332
+    },
+    {
+      "epoch": 0.7611227592566165,
+      "grad_norm": 1.609375,
+      "learning_rate": 6.754013986784546e-06,
+      "loss": 0.3005,
+      "step": 17334
+    },
+    {
+      "epoch": 0.76121057773582,
+      "grad_norm": 1.578125,
+      "learning_rate": 6.749285218026272e-06,
+      "loss": 0.3198,
+      "step": 17336
+    },
+    {
+      "epoch": 0.7612983962150236,
+      "grad_norm": 1.5390625,
+      "learning_rate": 6.744557846912114e-06,
+      "loss": 0.3374,
+      "step": 17338
+    },
+    {
+      "epoch": 0.7613862146942271,
+      "grad_norm": 1.625,
+      "learning_rate": 6.739831873804095e-06,
+      "loss": 0.3041,
+      "step": 17340
+    },
+    {
+      "epoch": 0.7614740331734305,
+      "grad_norm": 1.609375,
+      "learning_rate": 6.7351072990641225e-06,
+      "loss": 0.3155,
+      "step": 17342
+    },
+    {
+      "epoch": 0.761561851652634,
+      "grad_norm": 1.4609375,
+      "learning_rate": 6.730384123054018e-06,
+      "loss": 0.311,
+      "step": 17344
+    },
+    {
+      "epoch": 0.7616496701318375,
+      "grad_norm": 1.59375,
+      "learning_rate": 6.725662346135467e-06,
+      "loss": 0.3149,
+      "step": 17346
+    },
+    {
+      "epoch": 0.761737488611041,
+      "grad_norm": 1.4375,
+      "learning_rate": 6.720941968670083e-06,
+      "loss": 0.309,
+      "step": 17348
+    },
+    {
+      "epoch": 0.7618253070902444,
+      "grad_norm": 1.453125,
+      "learning_rate": 6.716222991019347e-06,
+      "loss": 0.3154,
+      "step": 17350
+    },
+    {
+      "epoch": 0.7619131255694479,
+      "grad_norm": 1.6171875,
+      "learning_rate": 6.711505413544628e-06,
+      "loss": 0.3152,
+      "step": 17352
+    },
+    {
+      "epoch": 0.7620009440486515,
+      "grad_norm": 1.4375,
+      "learning_rate": 6.706789236607214e-06,
+      "loss": 0.298,
+      "step": 17354
+    },
+    {
+      "epoch": 0.762088762527855,
+      "grad_norm": 1.609375,
+      "learning_rate": 6.7020744605682616e-06,
+      "loss": 0.351,
+      "step": 17356
+    },
+    {
+      "epoch": 0.7621765810070584,
+      "grad_norm": 1.5546875,
+      "learning_rate": 6.697361085788839e-06,
+      "loss": 0.3015,
+      "step": 17358
+    },
+    {
+      "epoch": 0.7622643994862619,
+      "grad_norm": 1.5859375,
+      "learning_rate": 6.69264911262989e-06,
+      "loss": 0.3124,
+      "step": 17360
+    },
+    {
+      "epoch": 0.7623522179654654,
+      "grad_norm": 1.390625,
+      "learning_rate": 6.687938541452257e-06,
+      "loss": 0.3441,
+      "step": 17362
+    },
+    {
+      "epoch": 0.7624400364446688,
+      "grad_norm": 1.5625,
+      "learning_rate": 6.6832293726166926e-06,
+      "loss": 0.3319,
+      "step": 17364
+    },
+    {
+      "epoch": 0.7625278549238723,
+      "grad_norm": 1.4765625,
+      "learning_rate": 6.678521606483798e-06,
+      "loss": 0.3132,
+      "step": 17366
+    },
+    {
+      "epoch": 0.7626156734030758,
+      "grad_norm": 1.5078125,
+      "learning_rate": 6.673815243414119e-06,
+      "loss": 0.3261,
+      "step": 17368
+    },
+    {
+      "epoch": 0.7627034918822794,
+      "grad_norm": 1.515625,
+      "learning_rate": 6.669110283768057e-06,
+      "loss": 0.3018,
+      "step": 17370
+    },
+    {
+      "epoch": 0.7627913103614828,
+      "grad_norm": 1.71875,
+      "learning_rate": 6.664406727905928e-06,
+      "loss": 0.3038,
+      "step": 17372
+    },
+    {
+      "epoch": 0.7628791288406863,
+      "grad_norm": 1.5,
+      "learning_rate": 6.659704576187928e-06,
+      "loss": 0.3124,
+      "step": 17374
+    },
+    {
+      "epoch": 0.7629669473198898,
+      "grad_norm": 1.578125,
+      "learning_rate": 6.655003828974141e-06,
+      "loss": 0.3278,
+      "step": 17376
+    },
+    {
+      "epoch": 0.7630547657990933,
+      "grad_norm": 1.4921875,
+      "learning_rate": 6.650304486624565e-06,
+      "loss": 0.3151,
+      "step": 17378
+    },
+    {
+      "epoch": 0.7631425842782967,
+      "grad_norm": 1.5546875,
+      "learning_rate": 6.645606549499062e-06,
+      "loss": 0.3277,
+      "step": 17380
+    },
+    {
+      "epoch": 0.7632304027575002,
+      "grad_norm": 1.453125,
+      "learning_rate": 6.640910017957419e-06,
+      "loss": 0.296,
+      "step": 17382
+    },
+    {
+      "epoch": 0.7633182212367038,
+      "grad_norm": 1.5703125,
+      "learning_rate": 6.6362148923592854e-06,
+      "loss": 0.3152,
+      "step": 17384
+    },
+    {
+      "epoch": 0.7634060397159073,
+      "grad_norm": 1.4453125,
+      "learning_rate": 6.6315211730642114e-06,
+      "loss": 0.3105,
+      "step": 17386
+    },
+    {
+      "epoch": 0.7634938581951107,
+      "grad_norm": 1.625,
+      "learning_rate": 6.626828860431658e-06,
+      "loss": 0.3076,
+      "step": 17388
+    },
+    {
+      "epoch": 0.7635816766743142,
+      "grad_norm": 1.59375,
+      "learning_rate": 6.622137954820945e-06,
+      "loss": 0.3395,
+      "step": 17390
+    },
+    {
+      "epoch": 0.7636694951535177,
+      "grad_norm": 1.5078125,
+      "learning_rate": 6.617448456591321e-06,
+      "loss": 0.3213,
+      "step": 17392
+    },
+    {
+      "epoch": 0.7637573136327211,
+      "grad_norm": 1.59375,
+      "learning_rate": 6.612760366101902e-06,
+      "loss": 0.2909,
+      "step": 17394
+    },
+    {
+      "epoch": 0.7638451321119246,
+      "grad_norm": 1.515625,
+      "learning_rate": 6.608073683711699e-06,
+      "loss": 0.3157,
+      "step": 17396
+    },
+    {
+      "epoch": 0.7639329505911281,
+      "grad_norm": 1.5234375,
+      "learning_rate": 6.603388409779626e-06,
+      "loss": 0.3192,
+      "step": 17398
+    },
+    {
+      "epoch": 0.7640207690703317,
+      "grad_norm": 1.640625,
+      "learning_rate": 6.598704544664469e-06,
+      "loss": 0.3447,
+      "step": 17400
+    },
+    {
+      "epoch": 0.7641085875495351,
+      "grad_norm": 1.53125,
+      "learning_rate": 6.594022088724935e-06,
+      "loss": 0.3035,
+      "step": 17402
+    },
+    {
+      "epoch": 0.7641964060287386,
+      "grad_norm": 1.4921875,
+      "learning_rate": 6.589341042319597e-06,
+      "loss": 0.3009,
+      "step": 17404
+    },
+    {
+      "epoch": 0.7642842245079421,
+      "grad_norm": 1.515625,
+      "learning_rate": 6.5846614058069415e-06,
+      "loss": 0.296,
+      "step": 17406
+    },
+    {
+      "epoch": 0.7643720429871456,
+      "grad_norm": 1.5390625,
+      "learning_rate": 6.579983179545324e-06,
+      "loss": 0.3131,
+      "step": 17408
+    },
+    {
+      "epoch": 0.764459861466349,
+      "grad_norm": 1.578125,
+      "learning_rate": 6.5753063638930165e-06,
+      "loss": 0.3251,
+      "step": 17410
+    },
+    {
+      "epoch": 0.7645476799455525,
+      "grad_norm": 1.765625,
+      "learning_rate": 6.570630959208163e-06,
+      "loss": 0.3128,
+      "step": 17412
+    },
+    {
+      "epoch": 0.764635498424756,
+      "grad_norm": 1.5703125,
+      "learning_rate": 6.5659569658488015e-06,
+      "loss": 0.3027,
+      "step": 17414
+    },
+    {
+      "epoch": 0.7647233169039596,
+      "grad_norm": 1.5546875,
+      "learning_rate": 6.561284384172883e-06,
+      "loss": 0.3251,
+      "step": 17416
+    },
+    {
+      "epoch": 0.764811135383163,
+      "grad_norm": 1.453125,
+      "learning_rate": 6.556613214538218e-06,
+      "loss": 0.3083,
+      "step": 17418
+    },
+    {
+      "epoch": 0.7648989538623665,
+      "grad_norm": 1.484375,
+      "learning_rate": 6.551943457302543e-06,
+      "loss": 0.3032,
+      "step": 17420
+    },
+    {
+      "epoch": 0.76498677234157,
+      "grad_norm": 1.6015625,
+      "learning_rate": 6.547275112823459e-06,
+      "loss": 0.2952,
+      "step": 17422
+    },
+    {
+      "epoch": 0.7650745908207734,
+      "grad_norm": 1.578125,
+      "learning_rate": 6.542608181458471e-06,
+      "loss": 0.3373,
+      "step": 17424
+    },
+    {
+      "epoch": 0.7651624092999769,
+      "grad_norm": 1.46875,
+      "learning_rate": 6.537942663564975e-06,
+      "loss": 0.324,
+      "step": 17426
+    },
+    {
+      "epoch": 0.7652502277791804,
+      "grad_norm": 1.625,
+      "learning_rate": 6.5332785595002446e-06,
+      "loss": 0.3154,
+      "step": 17428
+    },
+    {
+      "epoch": 0.765338046258384,
+      "grad_norm": 1.578125,
+      "learning_rate": 6.528615869621477e-06,
+      "loss": 0.3149,
+      "step": 17430
+    },
+    {
+      "epoch": 0.7654258647375874,
+      "grad_norm": 1.5234375,
+      "learning_rate": 6.523954594285728e-06,
+      "loss": 0.3135,
+      "step": 17432
+    },
+    {
+      "epoch": 0.7655136832167909,
+      "grad_norm": 1.5078125,
+      "learning_rate": 6.51929473384997e-06,
+      "loss": 0.3317,
+      "step": 17434
+    },
+    {
+      "epoch": 0.7656015016959944,
+      "grad_norm": 1.5,
+      "learning_rate": 6.514636288671056e-06,
+      "loss": 0.3129,
+      "step": 17436
+    },
+    {
+      "epoch": 0.7656893201751979,
+      "grad_norm": 1.5234375,
+      "learning_rate": 6.509979259105714e-06,
+      "loss": 0.2847,
+      "step": 17438
+    },
+    {
+      "epoch": 0.7657771386544013,
+      "grad_norm": 1.5859375,
+      "learning_rate": 6.505323645510603e-06,
+      "loss": 0.3047,
+      "step": 17440
+    },
+    {
+      "epoch": 0.7658649571336048,
+      "grad_norm": 1.53125,
+      "learning_rate": 6.500669448242233e-06,
+      "loss": 0.3203,
+      "step": 17442
+    },
+    {
+      "epoch": 0.7659527756128083,
+      "grad_norm": 1.5546875,
+      "learning_rate": 6.496016667657037e-06,
+      "loss": 0.2908,
+      "step": 17444
+    },
+    {
+      "epoch": 0.7660405940920119,
+      "grad_norm": 1.5546875,
+      "learning_rate": 6.491365304111322e-06,
+      "loss": 0.3149,
+      "step": 17446
+    },
+    {
+      "epoch": 0.7661284125712153,
+      "grad_norm": 1.53125,
+      "learning_rate": 6.486715357961281e-06,
+      "loss": 0.3405,
+      "step": 17448
+    },
+    {
+      "epoch": 0.7662162310504188,
+      "grad_norm": 1.4375,
+      "learning_rate": 6.4820668295630245e-06,
+      "loss": 0.2915,
+      "step": 17450
+    },
+    {
+      "epoch": 0.7663040495296223,
+      "grad_norm": 1.4453125,
+      "learning_rate": 6.47741971927252e-06,
+      "loss": 0.3329,
+      "step": 17452
+    },
+    {
+      "epoch": 0.7663918680088257,
+      "grad_norm": 1.5546875,
+      "learning_rate": 6.4727740274456605e-06,
+      "loss": 0.3177,
+      "step": 17454
+    },
+    {
+      "epoch": 0.7664796864880292,
+      "grad_norm": 1.4609375,
+      "learning_rate": 6.46812975443821e-06,
+      "loss": 0.3042,
+      "step": 17456
+    },
+    {
+      "epoch": 0.7665675049672327,
+      "grad_norm": 1.5078125,
+      "learning_rate": 6.463486900605822e-06,
+      "loss": 0.3394,
+      "step": 17458
+    },
+    {
+      "epoch": 0.7666553234464362,
+      "grad_norm": 1.3828125,
+      "learning_rate": 6.458845466304053e-06,
+      "loss": 0.2948,
+      "step": 17460
+    },
+    {
+      "epoch": 0.7667431419256397,
+      "grad_norm": 1.5234375,
+      "learning_rate": 6.454205451888335e-06,
+      "loss": 0.3091,
+      "step": 17462
+    },
+    {
+      "epoch": 0.7668309604048432,
+      "grad_norm": 1.5859375,
+      "learning_rate": 6.449566857714015e-06,
+      "loss": 0.3186,
+      "step": 17464
+    },
+    {
+      "epoch": 0.7669187788840467,
+      "grad_norm": 1.625,
+      "learning_rate": 6.444929684136306e-06,
+      "loss": 0.3018,
+      "step": 17466
+    },
+    {
+      "epoch": 0.7670065973632502,
+      "grad_norm": 1.484375,
+      "learning_rate": 6.440293931510336e-06,
+      "loss": 0.3279,
+      "step": 17468
+    },
+    {
+      "epoch": 0.7670944158424536,
+      "grad_norm": 1.5,
+      "learning_rate": 6.435659600191099e-06,
+      "loss": 0.3197,
+      "step": 17470
+    },
+    {
+      "epoch": 0.7671822343216571,
+      "grad_norm": 1.5078125,
+      "learning_rate": 6.43102669053351e-06,
+      "loss": 0.2848,
+      "step": 17472
+    },
+    {
+      "epoch": 0.7672700528008606,
+      "grad_norm": 1.484375,
+      "learning_rate": 6.4263952028923435e-06,
+      "loss": 0.2845,
+      "step": 17474
+    },
+    {
+      "epoch": 0.767357871280064,
+      "grad_norm": 1.546875,
+      "learning_rate": 6.421765137622282e-06,
+      "loss": 0.3126,
+      "step": 17476
+    },
+    {
+      "epoch": 0.7674456897592676,
+      "grad_norm": 1.5859375,
+      "learning_rate": 6.417136495077905e-06,
+      "loss": 0.3325,
+      "step": 17478
+    },
+    {
+      "epoch": 0.7675335082384711,
+      "grad_norm": 1.4765625,
+      "learning_rate": 6.4125092756136625e-06,
+      "loss": 0.3037,
+      "step": 17480
+    },
+    {
+      "epoch": 0.7676213267176746,
+      "grad_norm": 1.546875,
+      "learning_rate": 6.407883479583921e-06,
+      "loss": 0.3043,
+      "step": 17482
+    },
+    {
+      "epoch": 0.767709145196878,
+      "grad_norm": 1.5078125,
+      "learning_rate": 6.403259107342921e-06,
+      "loss": 0.34,
+      "step": 17484
+    },
+    {
+      "epoch": 0.7677969636760815,
+      "grad_norm": 1.5703125,
+      "learning_rate": 6.398636159244797e-06,
+      "loss": 0.3392,
+      "step": 17486
+    },
+    {
+      "epoch": 0.767884782155285,
+      "grad_norm": 1.5,
+      "learning_rate": 6.394014635643575e-06,
+      "loss": 0.3265,
+      "step": 17488
+    },
+    {
+      "epoch": 0.7679726006344885,
+      "grad_norm": 1.5546875,
+      "learning_rate": 6.389394536893165e-06,
+      "loss": 0.3089,
+      "step": 17490
+    },
+    {
+      "epoch": 0.768060419113692,
+      "grad_norm": 1.4921875,
+      "learning_rate": 6.384775863347389e-06,
+      "loss": 0.2844,
+      "step": 17492
+    },
+    {
+      "epoch": 0.7681482375928955,
+      "grad_norm": 1.7265625,
+      "learning_rate": 6.380158615359932e-06,
+      "loss": 0.3237,
+      "step": 17494
+    },
+    {
+      "epoch": 0.768236056072099,
+      "grad_norm": 1.4375,
+      "learning_rate": 6.3755427932844005e-06,
+      "loss": 0.2955,
+      "step": 17496
+    },
+    {
+      "epoch": 0.7683238745513025,
+      "grad_norm": 1.40625,
+      "learning_rate": 6.3709283974742654e-06,
+      "loss": 0.3182,
+      "step": 17498
+    },
+    {
+      "epoch": 0.7684116930305059,
+      "grad_norm": 1.4921875,
+      "learning_rate": 6.366315428282893e-06,
+      "loss": 0.3128,
+      "step": 17500
+    },
+    {
+      "epoch": 0.7684995115097094,
+      "grad_norm": 1.4375,
+      "learning_rate": 6.361703886063561e-06,
+      "loss": 0.2936,
+      "step": 17502
+    },
+    {
+      "epoch": 0.7685873299889129,
+      "grad_norm": 1.4921875,
+      "learning_rate": 6.357093771169403e-06,
+      "loss": 0.3087,
+      "step": 17504
+    },
+    {
+      "epoch": 0.7686751484681164,
+      "grad_norm": 1.625,
+      "learning_rate": 6.35248508395348e-06,
+      "loss": 0.3173,
+      "step": 17506
+    },
+    {
+      "epoch": 0.7687629669473199,
+      "grad_norm": 1.484375,
+      "learning_rate": 6.347877824768722e-06,
+      "loss": 0.307,
+      "step": 17508
+    },
+    {
+      "epoch": 0.7688507854265234,
+      "grad_norm": 1.5703125,
+      "learning_rate": 6.343271993967942e-06,
+      "loss": 0.3284,
+      "step": 17510
+    },
+    {
+      "epoch": 0.7689386039057269,
+      "grad_norm": 1.546875,
+      "learning_rate": 6.338667591903874e-06,
+      "loss": 0.3296,
+      "step": 17512
+    },
+    {
+      "epoch": 0.7690264223849304,
+      "grad_norm": 1.4921875,
+      "learning_rate": 6.334064618929106e-06,
+      "loss": 0.3028,
+      "step": 17514
+    },
+    {
+      "epoch": 0.7691142408641338,
+      "grad_norm": 1.609375,
+      "learning_rate": 6.329463075396161e-06,
+      "loss": 0.3115,
+      "step": 17516
+    },
+    {
+      "epoch": 0.7692020593433373,
+      "grad_norm": 1.453125,
+      "learning_rate": 6.324862961657393e-06,
+      "loss": 0.3134,
+      "step": 17518
+    },
+    {
+      "epoch": 0.7692898778225408,
+      "grad_norm": 1.5546875,
+      "learning_rate": 6.320264278065103e-06,
+      "loss": 0.3194,
+      "step": 17520
+    },
+    {
+      "epoch": 0.7693776963017442,
+      "grad_norm": 1.5546875,
+      "learning_rate": 6.315667024971453e-06,
+      "loss": 0.2776,
+      "step": 17522
+    },
+    {
+      "epoch": 0.7694655147809478,
+      "grad_norm": 1.421875,
+      "learning_rate": 6.311071202728494e-06,
+      "loss": 0.3067,
+      "step": 17524
+    },
+    {
+      "epoch": 0.7695533332601513,
+      "grad_norm": 1.5,
+      "learning_rate": 6.306476811688189e-06,
+      "loss": 0.3116,
+      "step": 17526
+    },
+    {
+      "epoch": 0.7696411517393548,
+      "grad_norm": 1.546875,
+      "learning_rate": 6.301883852202365e-06,
+      "loss": 0.2875,
+      "step": 17528
+    },
+    {
+      "epoch": 0.7697289702185582,
+      "grad_norm": 1.5,
+      "learning_rate": 6.2972923246227635e-06,
+      "loss": 0.297,
+      "step": 17530
+    },
+    {
+      "epoch": 0.7698167886977617,
+      "grad_norm": 1.546875,
+      "learning_rate": 6.292702229301001e-06,
+      "loss": 0.2835,
+      "step": 17532
+    },
+    {
+      "epoch": 0.7699046071769652,
+      "grad_norm": 1.546875,
+      "learning_rate": 6.288113566588577e-06,
+      "loss": 0.2944,
+      "step": 17534
+    },
+    {
+      "epoch": 0.7699924256561687,
+      "grad_norm": 1.453125,
+      "learning_rate": 6.283526336836912e-06,
+      "loss": 0.3091,
+      "step": 17536
+    },
+    {
+      "epoch": 0.7700802441353722,
+      "grad_norm": 1.53125,
+      "learning_rate": 6.2789405403972765e-06,
+      "loss": 0.3228,
+      "step": 17538
+    },
+    {
+      "epoch": 0.7701680626145757,
+      "grad_norm": 1.5,
+      "learning_rate": 6.274356177620871e-06,
+      "loss": 0.3132,
+      "step": 17540
+    },
+    {
+      "epoch": 0.7702558810937792,
+      "grad_norm": 1.5546875,
+      "learning_rate": 6.269773248858748e-06,
+      "loss": 0.3091,
+      "step": 17542
+    },
+    {
+      "epoch": 0.7703436995729827,
+      "grad_norm": 1.5859375,
+      "learning_rate": 6.265191754461891e-06,
+      "loss": 0.3336,
+      "step": 17544
+    },
+    {
+      "epoch": 0.7704315180521861,
+      "grad_norm": 1.4375,
+      "learning_rate": 6.260611694781138e-06,
+      "loss": 0.2966,
+      "step": 17546
+    },
+    {
+      "epoch": 0.7705193365313896,
+      "grad_norm": 1.5390625,
+      "learning_rate": 6.256033070167236e-06,
+      "loss": 0.2936,
+      "step": 17548
+    },
+    {
+      "epoch": 0.7706071550105931,
+      "grad_norm": 1.5703125,
+      "learning_rate": 6.251455880970811e-06,
+      "loss": 0.3251,
+      "step": 17550
+    },
+    {
+      "epoch": 0.7706949734897965,
+      "grad_norm": 1.5078125,
+      "learning_rate": 6.246880127542385e-06,
+      "loss": 0.3251,
+      "step": 17552
+    },
+    {
+      "epoch": 0.7707827919690001,
+      "grad_norm": 1.5078125,
+      "learning_rate": 6.242305810232379e-06,
+      "loss": 0.3179,
+      "step": 17554
+    },
+    {
+      "epoch": 0.7708706104482036,
+      "grad_norm": 1.53125,
+      "learning_rate": 6.237732929391085e-06,
+      "loss": 0.3123,
+      "step": 17556
+    },
+    {
+      "epoch": 0.7709584289274071,
+      "grad_norm": 1.5390625,
+      "learning_rate": 6.233161485368707e-06,
+      "loss": 0.3289,
+      "step": 17558
+    },
+    {
+      "epoch": 0.7710462474066105,
+      "grad_norm": 1.46875,
+      "learning_rate": 6.228591478515322e-06,
+      "loss": 0.308,
+      "step": 17560
+    },
+    {
+      "epoch": 0.771134065885814,
+      "grad_norm": 1.4296875,
+      "learning_rate": 6.224022909180893e-06,
+      "loss": 0.3368,
+      "step": 17562
+    },
+    {
+      "epoch": 0.7712218843650175,
+      "grad_norm": 1.5,
+      "learning_rate": 6.219455777715299e-06,
+      "loss": 0.3168,
+      "step": 17564
+    },
+    {
+      "epoch": 0.771309702844221,
+      "grad_norm": 1.5390625,
+      "learning_rate": 6.2148900844682775e-06,
+      "loss": 0.3278,
+      "step": 17566
+    },
+    {
+      "epoch": 0.7713975213234244,
+      "grad_norm": 1.4765625,
+      "learning_rate": 6.210325829789481e-06,
+      "loss": 0.3131,
+      "step": 17568
+    },
+    {
+      "epoch": 0.771485339802628,
+      "grad_norm": 1.5625,
+      "learning_rate": 6.205763014028437e-06,
+      "loss": 0.2855,
+      "step": 17570
+    },
+    {
+      "epoch": 0.7715731582818315,
+      "grad_norm": 1.5625,
+      "learning_rate": 6.201201637534562e-06,
+      "loss": 0.3164,
+      "step": 17572
+    },
+    {
+      "epoch": 0.771660976761035,
+      "grad_norm": 1.6484375,
+      "learning_rate": 6.196641700657177e-06,
+      "loss": 0.3106,
+      "step": 17574
+    },
+    {
+      "epoch": 0.7717487952402384,
+      "grad_norm": 1.4609375,
+      "learning_rate": 6.192083203745472e-06,
+      "loss": 0.2999,
+      "step": 17576
+    },
+    {
+      "epoch": 0.7718366137194419,
+      "grad_norm": 1.515625,
+      "learning_rate": 6.187526147148557e-06,
+      "loss": 0.3328,
+      "step": 17578
+    },
+    {
+      "epoch": 0.7719244321986454,
+      "grad_norm": 1.5078125,
+      "learning_rate": 6.182970531215384e-06,
+      "loss": 0.3015,
+      "step": 17580
+    },
+    {
+      "epoch": 0.7720122506778488,
+      "grad_norm": 1.4609375,
+      "learning_rate": 6.1784163562948476e-06,
+      "loss": 0.3401,
+      "step": 17582
+    },
+    {
+      "epoch": 0.7721000691570524,
+      "grad_norm": 1.515625,
+      "learning_rate": 6.173863622735698e-06,
+      "loss": 0.3226,
+      "step": 17584
+    },
+    {
+      "epoch": 0.7721878876362559,
+      "grad_norm": 1.4296875,
+      "learning_rate": 6.169312330886578e-06,
+      "loss": 0.3179,
+      "step": 17586
+    },
+    {
+      "epoch": 0.7722757061154594,
+      "grad_norm": 1.53125,
+      "learning_rate": 6.164762481096042e-06,
+      "loss": 0.3081,
+      "step": 17588
+    },
+    {
+      "epoch": 0.7723635245946628,
+      "grad_norm": 1.484375,
+      "learning_rate": 6.1602140737125e-06,
+      "loss": 0.2958,
+      "step": 17590
+    },
+    {
+      "epoch": 0.7724513430738663,
+      "grad_norm": 1.5234375,
+      "learning_rate": 6.15566710908429e-06,
+      "loss": 0.3202,
+      "step": 17592
+    },
+    {
+      "epoch": 0.7725391615530698,
+      "grad_norm": 1.5625,
+      "learning_rate": 6.151121587559611e-06,
+      "loss": 0.2966,
+      "step": 17594
+    },
+    {
+      "epoch": 0.7726269800322733,
+      "grad_norm": 1.4609375,
+      "learning_rate": 6.146577509486551e-06,
+      "loss": 0.3459,
+      "step": 17596
+    },
+    {
+      "epoch": 0.7727147985114767,
+      "grad_norm": 1.5,
+      "learning_rate": 6.1420348752131095e-06,
+      "loss": 0.3385,
+      "step": 17598
+    },
+    {
+      "epoch": 0.7728026169906803,
+      "grad_norm": 1.46875,
+      "learning_rate": 6.137493685087154e-06,
+      "loss": 0.2862,
+      "step": 17600
+    },
+    {
+      "epoch": 0.7728904354698838,
+      "grad_norm": 1.4140625,
+      "learning_rate": 6.1329539394564596e-06,
+      "loss": 0.2925,
+      "step": 17602
+    },
+    {
+      "epoch": 0.7729782539490873,
+      "grad_norm": 1.65625,
+      "learning_rate": 6.128415638668669e-06,
+      "loss": 0.2776,
+      "step": 17604
+    },
+    {
+      "epoch": 0.7730660724282907,
+      "grad_norm": 1.578125,
+      "learning_rate": 6.123878783071338e-06,
+      "loss": 0.3116,
+      "step": 17606
+    },
+    {
+      "epoch": 0.7731538909074942,
+      "grad_norm": 1.5625,
+      "learning_rate": 6.119343373011896e-06,
+      "loss": 0.3309,
+      "step": 17608
+    },
+    {
+      "epoch": 0.7732417093866977,
+      "grad_norm": 1.484375,
+      "learning_rate": 6.114809408837665e-06,
+      "loss": 0.3367,
+      "step": 17610
+    },
+    {
+      "epoch": 0.7733295278659011,
+      "grad_norm": 1.5234375,
+      "learning_rate": 6.1102768908958555e-06,
+      "loss": 0.326,
+      "step": 17612
+    },
+    {
+      "epoch": 0.7734173463451046,
+      "grad_norm": 1.5,
+      "learning_rate": 6.105745819533562e-06,
+      "loss": 0.3058,
+      "step": 17614
+    },
+    {
+      "epoch": 0.7735051648243082,
+      "grad_norm": 1.3984375,
+      "learning_rate": 6.101216195097792e-06,
+      "loss": 0.3065,
+      "step": 17616
+    },
+    {
+      "epoch": 0.7735929833035117,
+      "grad_norm": 1.609375,
+      "learning_rate": 6.096688017935406e-06,
+      "loss": 0.3105,
+      "step": 17618
+    },
+    {
+      "epoch": 0.7736808017827151,
+      "grad_norm": 1.625,
+      "learning_rate": 6.0921612883931944e-06,
+      "loss": 0.3188,
+      "step": 17620
+    },
+    {
+      "epoch": 0.7737686202619186,
+      "grad_norm": 1.59375,
+      "learning_rate": 6.087636006817801e-06,
+      "loss": 0.3217,
+      "step": 17622
+    },
+    {
+      "epoch": 0.7738564387411221,
+      "grad_norm": 1.5078125,
+      "learning_rate": 6.083112173555769e-06,
+      "loss": 0.3665,
+      "step": 17624
+    },
+    {
+      "epoch": 0.7739442572203256,
+      "grad_norm": 1.5859375,
+      "learning_rate": 6.0785897889535485e-06,
+      "loss": 0.2938,
+      "step": 17626
+    },
+    {
+      "epoch": 0.774032075699529,
+      "grad_norm": 1.546875,
+      "learning_rate": 6.074068853357451e-06,
+      "loss": 0.3167,
+      "step": 17628
+    },
+    {
+      "epoch": 0.7741198941787326,
+      "grad_norm": 1.6171875,
+      "learning_rate": 6.069549367113706e-06,
+      "loss": 0.3021,
+      "step": 17630
+    },
+    {
+      "epoch": 0.7742077126579361,
+      "grad_norm": 1.4296875,
+      "learning_rate": 6.065031330568408e-06,
+      "loss": 0.2957,
+      "step": 17632
+    },
+    {
+      "epoch": 0.7742955311371396,
+      "grad_norm": 1.5078125,
+      "learning_rate": 6.0605147440675415e-06,
+      "loss": 0.3229,
+      "step": 17634
+    },
+    {
+      "epoch": 0.774383349616343,
+      "grad_norm": 1.5234375,
+      "learning_rate": 6.0559996079570025e-06,
+      "loss": 0.2996,
+      "step": 17636
+    },
+    {
+      "epoch": 0.7744711680955465,
+      "grad_norm": 1.4765625,
+      "learning_rate": 6.051485922582548e-06,
+      "loss": 0.315,
+      "step": 17638
+    },
+    {
+      "epoch": 0.77455898657475,
+      "grad_norm": 1.5,
+      "learning_rate": 6.046973688289859e-06,
+      "loss": 0.3164,
+      "step": 17640
+    },
+    {
+      "epoch": 0.7746468050539534,
+      "grad_norm": 1.546875,
+      "learning_rate": 6.042462905424454e-06,
+      "loss": 0.3116,
+      "step": 17642
+    },
+    {
+      "epoch": 0.7747346235331569,
+      "grad_norm": 1.5234375,
+      "learning_rate": 6.0379535743317896e-06,
+      "loss": 0.3056,
+      "step": 17644
+    },
+    {
+      "epoch": 0.7748224420123605,
+      "grad_norm": 1.625,
+      "learning_rate": 6.033445695357187e-06,
+      "loss": 0.2799,
+      "step": 17646
+    },
+    {
+      "epoch": 0.774910260491564,
+      "grad_norm": 1.5390625,
+      "learning_rate": 6.0289392688458544e-06,
+      "loss": 0.3192,
+      "step": 17648
+    },
+    {
+      "epoch": 0.7749980789707674,
+      "grad_norm": 1.5625,
+      "learning_rate": 6.024434295142905e-06,
+      "loss": 0.3216,
+      "step": 17650
+    },
+    {
+      "epoch": 0.7750858974499709,
+      "grad_norm": 1.4296875,
+      "learning_rate": 6.019930774593318e-06,
+      "loss": 0.3274,
+      "step": 17652
+    },
+    {
+      "epoch": 0.7751737159291744,
+      "grad_norm": 1.5234375,
+      "learning_rate": 6.015428707541993e-06,
+      "loss": 0.325,
+      "step": 17654
+    },
+    {
+      "epoch": 0.7752615344083779,
+      "grad_norm": 1.5625,
+      "learning_rate": 6.010928094333684e-06,
+      "loss": 0.336,
+      "step": 17656
+    },
+    {
+      "epoch": 0.7753493528875813,
+      "grad_norm": 1.6875,
+      "learning_rate": 6.006428935313049e-06,
+      "loss": 0.3216,
+      "step": 17658
+    },
+    {
+      "epoch": 0.7754371713667848,
+      "grad_norm": 1.5078125,
+      "learning_rate": 6.001931230824648e-06,
+      "loss": 0.2961,
+      "step": 17660
+    },
+    {
+      "epoch": 0.7755249898459884,
+      "grad_norm": 1.5078125,
+      "learning_rate": 5.997434981212896e-06,
+      "loss": 0.2915,
+      "step": 17662
+    },
+    {
+      "epoch": 0.7756128083251919,
+      "grad_norm": 1.546875,
+      "learning_rate": 5.992940186822138e-06,
+      "loss": 0.2998,
+      "step": 17664
+    },
+    {
+      "epoch": 0.7757006268043953,
+      "grad_norm": 1.5234375,
+      "learning_rate": 5.988446847996579e-06,
+      "loss": 0.3264,
+      "step": 17666
+    },
+    {
+      "epoch": 0.7757884452835988,
+      "grad_norm": 1.484375,
+      "learning_rate": 5.983954965080307e-06,
+      "loss": 0.3098,
+      "step": 17668
+    },
+    {
+      "epoch": 0.7758762637628023,
+      "grad_norm": 1.4453125,
+      "learning_rate": 5.9794645384173314e-06,
+      "loss": 0.2944,
+      "step": 17670
+    },
+    {
+      "epoch": 0.7759640822420057,
+      "grad_norm": 1.6171875,
+      "learning_rate": 5.974975568351521e-06,
+      "loss": 0.3112,
+      "step": 17672
+    },
+    {
+      "epoch": 0.7760519007212092,
+      "grad_norm": 1.46875,
+      "learning_rate": 5.970488055226642e-06,
+      "loss": 0.3,
+      "step": 17674
+    },
+    {
+      "epoch": 0.7761397192004127,
+      "grad_norm": 1.6796875,
+      "learning_rate": 5.966001999386339e-06,
+      "loss": 0.3396,
+      "step": 17676
+    },
+    {
+      "epoch": 0.7762275376796163,
+      "grad_norm": 1.46875,
+      "learning_rate": 5.9615174011741774e-06,
+      "loss": 0.3165,
+      "step": 17678
+    },
+    {
+      "epoch": 0.7763153561588197,
+      "grad_norm": 1.671875,
+      "learning_rate": 5.957034260933567e-06,
+      "loss": 0.3076,
+      "step": 17680
+    },
+    {
+      "epoch": 0.7764031746380232,
+      "grad_norm": 1.65625,
+      "learning_rate": 5.952552579007847e-06,
+      "loss": 0.2645,
+      "step": 17682
+    },
+    {
+      "epoch": 0.7764909931172267,
+      "grad_norm": 1.5859375,
+      "learning_rate": 5.948072355740214e-06,
+      "loss": 0.3043,
+      "step": 17684
+    },
+    {
+      "epoch": 0.7765788115964302,
+      "grad_norm": 1.6328125,
+      "learning_rate": 5.943593591473762e-06,
+      "loss": 0.295,
+      "step": 17686
+    },
+    {
+      "epoch": 0.7766666300756336,
+      "grad_norm": 1.4296875,
+      "learning_rate": 5.939116286551488e-06,
+      "loss": 0.3101,
+      "step": 17688
+    },
+    {
+      "epoch": 0.7767544485548371,
+      "grad_norm": 1.53125,
+      "learning_rate": 5.9346404413162494e-06,
+      "loss": 0.3455,
+      "step": 17690
+    },
+    {
+      "epoch": 0.7768422670340407,
+      "grad_norm": 1.4609375,
+      "learning_rate": 5.930166056110825e-06,
+      "loss": 0.3092,
+      "step": 17692
+    },
+    {
+      "epoch": 0.7769300855132442,
+      "grad_norm": 1.4921875,
+      "learning_rate": 5.925693131277854e-06,
+      "loss": 0.328,
+      "step": 17694
+    },
+    {
+      "epoch": 0.7770179039924476,
+      "grad_norm": 1.5078125,
+      "learning_rate": 5.921221667159868e-06,
+      "loss": 0.3365,
+      "step": 17696
+    },
+    {
+      "epoch": 0.7771057224716511,
+      "grad_norm": 1.4296875,
+      "learning_rate": 5.91675166409931e-06,
+      "loss": 0.2962,
+      "step": 17698
+    },
+    {
+      "epoch": 0.7771935409508546,
+      "grad_norm": 1.5078125,
+      "learning_rate": 5.912283122438481e-06,
+      "loss": 0.3077,
+      "step": 17700
+    },
+    {
+      "epoch": 0.777281359430058,
+      "grad_norm": 1.5234375,
+      "learning_rate": 5.907816042519587e-06,
+      "loss": 0.3033,
+      "step": 17702
+    },
+    {
+      "epoch": 0.7773691779092615,
+      "grad_norm": 1.5,
+      "learning_rate": 5.903350424684712e-06,
+      "loss": 0.3449,
+      "step": 17704
+    },
+    {
+      "epoch": 0.777456996388465,
+      "grad_norm": 1.7578125,
+      "learning_rate": 5.898886269275844e-06,
+      "loss": 0.2861,
+      "step": 17706
+    },
+    {
+      "epoch": 0.7775448148676686,
+      "grad_norm": 1.6875,
+      "learning_rate": 5.894423576634847e-06,
+      "loss": 0.3205,
+      "step": 17708
+    },
+    {
+      "epoch": 0.777632633346872,
+      "grad_norm": 1.46875,
+      "learning_rate": 5.889962347103461e-06,
+      "loss": 0.3092,
+      "step": 17710
+    },
+    {
+      "epoch": 0.7777204518260755,
+      "grad_norm": 1.4453125,
+      "learning_rate": 5.88550258102335e-06,
+      "loss": 0.3223,
+      "step": 17712
+    },
+    {
+      "epoch": 0.777808270305279,
+      "grad_norm": 1.7265625,
+      "learning_rate": 5.881044278736025e-06,
+      "loss": 0.3207,
+      "step": 17714
+    },
+    {
+      "epoch": 0.7778960887844825,
+      "grad_norm": 1.5859375,
+      "learning_rate": 5.87658744058292e-06,
+      "loss": 0.3131,
+      "step": 17716
+    },
+    {
+      "epoch": 0.7779839072636859,
+      "grad_norm": 1.5,
+      "learning_rate": 5.8721320669053335e-06,
+      "loss": 0.3225,
+      "step": 17718
+    },
+    {
+      "epoch": 0.7780717257428894,
+      "grad_norm": 1.515625,
+      "learning_rate": 5.867678158044451e-06,
+      "loss": 0.3102,
+      "step": 17720
+    },
+    {
+      "epoch": 0.7781595442220929,
+      "grad_norm": 1.5,
+      "learning_rate": 5.863225714341367e-06,
+      "loss": 0.3148,
+      "step": 17722
+    },
+    {
+      "epoch": 0.7782473627012965,
+      "grad_norm": 1.484375,
+      "learning_rate": 5.85877473613704e-06,
+      "loss": 0.324,
+      "step": 17724
+    },
+    {
+      "epoch": 0.7783351811804999,
+      "grad_norm": 1.484375,
+      "learning_rate": 5.854325223772339e-06,
+      "loss": 0.3006,
+      "step": 17726
+    },
+    {
+      "epoch": 0.7784229996597034,
+      "grad_norm": 1.5234375,
+      "learning_rate": 5.849877177588e-06,
+      "loss": 0.3068,
+      "step": 17728
+    },
+    {
+      "epoch": 0.7785108181389069,
+      "grad_norm": 1.515625,
+      "learning_rate": 5.845430597924653e-06,
+      "loss": 0.3072,
+      "step": 17730
+    },
+    {
+      "epoch": 0.7785986366181104,
+      "grad_norm": 1.5546875,
+      "learning_rate": 5.840985485122829e-06,
+      "loss": 0.3408,
+      "step": 17732
+    },
+    {
+      "epoch": 0.7786864550973138,
+      "grad_norm": 1.4609375,
+      "learning_rate": 5.836541839522927e-06,
+      "loss": 0.3022,
+      "step": 17734
+    },
+    {
+      "epoch": 0.7787742735765173,
+      "grad_norm": 1.578125,
+      "learning_rate": 5.832099661465248e-06,
+      "loss": 0.3409,
+      "step": 17736
+    },
+    {
+      "epoch": 0.7788620920557209,
+      "grad_norm": 1.5234375,
+      "learning_rate": 5.827658951289963e-06,
+      "loss": 0.3165,
+      "step": 17738
+    },
+    {
+      "epoch": 0.7789499105349244,
+      "grad_norm": 1.515625,
+      "learning_rate": 5.823219709337158e-06,
+      "loss": 0.3617,
+      "step": 17740
+    },
+    {
+      "epoch": 0.7790377290141278,
+      "grad_norm": 1.453125,
+      "learning_rate": 5.818781935946779e-06,
+      "loss": 0.2971,
+      "step": 17742
+    },
+    {
+      "epoch": 0.7791255474933313,
+      "grad_norm": 1.5078125,
+      "learning_rate": 5.814345631458684e-06,
+      "loss": 0.3356,
+      "step": 17744
+    },
+    {
+      "epoch": 0.7792133659725348,
+      "grad_norm": 1.5625,
+      "learning_rate": 5.8099107962125975e-06,
+      "loss": 0.3173,
+      "step": 17746
+    },
+    {
+      "epoch": 0.7793011844517382,
+      "grad_norm": 1.4609375,
+      "learning_rate": 5.80547743054814e-06,
+      "loss": 0.3085,
+      "step": 17748
+    },
+    {
+      "epoch": 0.7793890029309417,
+      "grad_norm": 1.4765625,
+      "learning_rate": 5.801045534804825e-06,
+      "loss": 0.3229,
+      "step": 17750
+    },
+    {
+      "epoch": 0.7794768214101452,
+      "grad_norm": 1.484375,
+      "learning_rate": 5.7966151093220396e-06,
+      "loss": 0.318,
+      "step": 17752
+    },
+    {
+      "epoch": 0.7795646398893488,
+      "grad_norm": 1.4921875,
+      "learning_rate": 5.7921861544390805e-06,
+      "loss": 0.3038,
+      "step": 17754
+    },
+    {
+      "epoch": 0.7796524583685522,
+      "grad_norm": 1.546875,
+      "learning_rate": 5.787758670495108e-06,
+      "loss": 0.334,
+      "step": 17756
+    },
+    {
+      "epoch": 0.7797402768477557,
+      "grad_norm": 1.484375,
+      "learning_rate": 5.783332657829177e-06,
+      "loss": 0.2941,
+      "step": 17758
+    },
+    {
+      "epoch": 0.7798280953269592,
+      "grad_norm": 1.484375,
+      "learning_rate": 5.778908116780244e-06,
+      "loss": 0.3116,
+      "step": 17760
+    },
+    {
+      "epoch": 0.7799159138061627,
+      "grad_norm": 1.59375,
+      "learning_rate": 5.7744850476871335e-06,
+      "loss": 0.3276,
+      "step": 17762
+    },
+    {
+      "epoch": 0.7800037322853661,
+      "grad_norm": 1.5546875,
+      "learning_rate": 5.770063450888569e-06,
+      "loss": 0.3172,
+      "step": 17764
+    },
+    {
+      "epoch": 0.7800915507645696,
+      "grad_norm": 1.3984375,
+      "learning_rate": 5.765643326723147e-06,
+      "loss": 0.3122,
+      "step": 17766
+    },
+    {
+      "epoch": 0.7801793692437731,
+      "grad_norm": 1.6015625,
+      "learning_rate": 5.761224675529375e-06,
+      "loss": 0.3361,
+      "step": 17768
+    },
+    {
+      "epoch": 0.7802671877229767,
+      "grad_norm": 1.4765625,
+      "learning_rate": 5.756807497645633e-06,
+      "loss": 0.3319,
+      "step": 17770
+    },
+    {
+      "epoch": 0.7803550062021801,
+      "grad_norm": 1.4375,
+      "learning_rate": 5.752391793410175e-06,
+      "loss": 0.3019,
+      "step": 17772
+    },
+    {
+      "epoch": 0.7804428246813836,
+      "grad_norm": 1.453125,
+      "learning_rate": 5.7479775631611775e-06,
+      "loss": 0.3042,
+      "step": 17774
+    },
+    {
+      "epoch": 0.7805306431605871,
+      "grad_norm": 1.4375,
+      "learning_rate": 5.743564807236665e-06,
+      "loss": 0.2863,
+      "step": 17776
+    },
+    {
+      "epoch": 0.7806184616397905,
+      "grad_norm": 1.578125,
+      "learning_rate": 5.739153525974583e-06,
+      "loss": 0.3002,
+      "step": 17778
+    },
+    {
+      "epoch": 0.780706280118994,
+      "grad_norm": 1.421875,
+      "learning_rate": 5.73474371971274e-06,
+      "loss": 0.3111,
+      "step": 17780
+    },
+    {
+      "epoch": 0.7807940985981975,
+      "grad_norm": 1.5234375,
+      "learning_rate": 5.730335388788835e-06,
+      "loss": 0.318,
+      "step": 17782
+    },
+    {
+      "epoch": 0.7808819170774011,
+      "grad_norm": 1.453125,
+      "learning_rate": 5.725928533540473e-06,
+      "loss": 0.3323,
+      "step": 17784
+    },
+    {
+      "epoch": 0.7809697355566045,
+      "grad_norm": 1.5390625,
+      "learning_rate": 5.721523154305117e-06,
+      "loss": 0.3277,
+      "step": 17786
+    },
+    {
+      "epoch": 0.781057554035808,
+      "grad_norm": 1.4765625,
+      "learning_rate": 5.717119251420145e-06,
+      "loss": 0.3311,
+      "step": 17788
+    },
+    {
+      "epoch": 0.7811453725150115,
+      "grad_norm": 1.4375,
+      "learning_rate": 5.712716825222803e-06,
+      "loss": 0.319,
+      "step": 17790
+    },
+    {
+      "epoch": 0.781233190994215,
+      "grad_norm": 1.65625,
+      "learning_rate": 5.7083158760502295e-06,
+      "loss": 0.3544,
+      "step": 17792
+    },
+    {
+      "epoch": 0.7813210094734184,
+      "grad_norm": 1.5,
+      "learning_rate": 5.703916404239454e-06,
+      "loss": 0.31,
+      "step": 17794
+    },
+    {
+      "epoch": 0.7814088279526219,
+      "grad_norm": 1.5078125,
+      "learning_rate": 5.699518410127375e-06,
+      "loss": 0.3007,
+      "step": 17796
+    },
+    {
+      "epoch": 0.7814966464318254,
+      "grad_norm": 1.515625,
+      "learning_rate": 5.695121894050812e-06,
+      "loss": 0.3031,
+      "step": 17798
+    },
+    {
+      "epoch": 0.781584464911029,
+      "grad_norm": 1.5078125,
+      "learning_rate": 5.690726856346434e-06,
+      "loss": 0.3103,
+      "step": 17800
+    },
+    {
+      "epoch": 0.7816722833902324,
+      "grad_norm": 1.40625,
+      "learning_rate": 5.686333297350832e-06,
+      "loss": 0.3037,
+      "step": 17802
+    },
+    {
+      "epoch": 0.7817601018694359,
+      "grad_norm": 1.484375,
+      "learning_rate": 5.681941217400446e-06,
+      "loss": 0.3205,
+      "step": 17804
+    },
+    {
+      "epoch": 0.7818479203486394,
+      "grad_norm": 1.5,
+      "learning_rate": 5.677550616831639e-06,
+      "loss": 0.2974,
+      "step": 17806
+    },
+    {
+      "epoch": 0.7819357388278428,
+      "grad_norm": 1.5390625,
+      "learning_rate": 5.673161495980639e-06,
+      "loss": 0.3513,
+      "step": 17808
+    },
+    {
+      "epoch": 0.7820235573070463,
+      "grad_norm": 1.4921875,
+      "learning_rate": 5.668773855183557e-06,
+      "loss": 0.3199,
+      "step": 17810
+    },
+    {
+      "epoch": 0.7821113757862498,
+      "grad_norm": 1.5234375,
+      "learning_rate": 5.664387694776416e-06,
+      "loss": 0.327,
+      "step": 17812
+    },
+    {
+      "epoch": 0.7821991942654533,
+      "grad_norm": 1.4765625,
+      "learning_rate": 5.660003015095092e-06,
+      "loss": 0.3015,
+      "step": 17814
+    },
+    {
+      "epoch": 0.7822870127446568,
+      "grad_norm": 1.59375,
+      "learning_rate": 5.65561981647538e-06,
+      "loss": 0.2809,
+      "step": 17816
+    },
+    {
+      "epoch": 0.7823748312238603,
+      "grad_norm": 1.5546875,
+      "learning_rate": 5.65123809925294e-06,
+      "loss": 0.3464,
+      "step": 17818
+    },
+    {
+      "epoch": 0.7824626497030638,
+      "grad_norm": 1.5390625,
+      "learning_rate": 5.646857863763317e-06,
+      "loss": 0.335,
+      "step": 17820
+    },
+    {
+      "epoch": 0.7825504681822673,
+      "grad_norm": 1.4921875,
+      "learning_rate": 5.642479110341964e-06,
+      "loss": 0.3084,
+      "step": 17822
+    },
+    {
+      "epoch": 0.7826382866614707,
+      "grad_norm": 1.6015625,
+      "learning_rate": 5.638101839324203e-06,
+      "loss": 0.3195,
+      "step": 17824
+    },
+    {
+      "epoch": 0.7827261051406742,
+      "grad_norm": 1.5078125,
+      "learning_rate": 5.633726051045243e-06,
+      "loss": 0.3401,
+      "step": 17826
+    },
+    {
+      "epoch": 0.7828139236198777,
+      "grad_norm": 1.5546875,
+      "learning_rate": 5.629351745840181e-06,
+      "loss": 0.3235,
+      "step": 17828
+    },
+    {
+      "epoch": 0.7829017420990811,
+      "grad_norm": 1.515625,
+      "learning_rate": 5.624978924044008e-06,
+      "loss": 0.3002,
+      "step": 17830
+    },
+    {
+      "epoch": 0.7829895605782847,
+      "grad_norm": 1.46875,
+      "learning_rate": 5.620607585991597e-06,
+      "loss": 0.288,
+      "step": 17832
+    },
+    {
+      "epoch": 0.7830773790574882,
+      "grad_norm": 1.453125,
+      "learning_rate": 5.616237732017693e-06,
+      "loss": 0.3017,
+      "step": 17834
+    },
+    {
+      "epoch": 0.7831651975366917,
+      "grad_norm": 1.46875,
+      "learning_rate": 5.611869362456959e-06,
+      "loss": 0.3192,
+      "step": 17836
+    },
+    {
+      "epoch": 0.7832530160158951,
+      "grad_norm": 1.4609375,
+      "learning_rate": 5.607502477643908e-06,
+      "loss": 0.2933,
+      "step": 17838
+    },
+    {
+      "epoch": 0.7833408344950986,
+      "grad_norm": 1.4296875,
+      "learning_rate": 5.603137077912976e-06,
+      "loss": 0.3554,
+      "step": 17840
+    },
+    {
+      "epoch": 0.7834286529743021,
+      "grad_norm": 1.53125,
+      "learning_rate": 5.598773163598456e-06,
+      "loss": 0.2942,
+      "step": 17842
+    },
+    {
+      "epoch": 0.7835164714535056,
+      "grad_norm": 1.4765625,
+      "learning_rate": 5.5944107350345274e-06,
+      "loss": 0.3343,
+      "step": 17844
+    },
+    {
+      "epoch": 0.7836042899327091,
+      "grad_norm": 1.4609375,
+      "learning_rate": 5.590049792555285e-06,
+      "loss": 0.3123,
+      "step": 17846
+    },
+    {
+      "epoch": 0.7836921084119126,
+      "grad_norm": 1.4453125,
+      "learning_rate": 5.5856903364946785e-06,
+      "loss": 0.3106,
+      "step": 17848
+    },
+    {
+      "epoch": 0.7837799268911161,
+      "grad_norm": 1.5234375,
+      "learning_rate": 5.581332367186562e-06,
+      "loss": 0.3005,
+      "step": 17850
+    },
+    {
+      "epoch": 0.7838677453703196,
+      "grad_norm": 1.546875,
+      "learning_rate": 5.576975884964672e-06,
+      "loss": 0.3089,
+      "step": 17852
+    },
+    {
+      "epoch": 0.783955563849523,
+      "grad_norm": 1.7734375,
+      "learning_rate": 5.572620890162622e-06,
+      "loss": 0.3218,
+      "step": 17854
+    },
+    {
+      "epoch": 0.7840433823287265,
+      "grad_norm": 1.4609375,
+      "learning_rate": 5.568267383113923e-06,
+      "loss": 0.3051,
+      "step": 17856
+    },
+    {
+      "epoch": 0.78413120080793,
+      "grad_norm": 1.546875,
+      "learning_rate": 5.563915364151959e-06,
+      "loss": 0.2974,
+      "step": 17858
+    },
+    {
+      "epoch": 0.7842190192871334,
+      "grad_norm": 1.5546875,
+      "learning_rate": 5.55956483361002e-06,
+      "loss": 0.3184,
+      "step": 17860
+    },
+    {
+      "epoch": 0.784306837766337,
+      "grad_norm": 1.484375,
+      "learning_rate": 5.555215791821261e-06,
+      "loss": 0.2942,
+      "step": 17862
+    },
+    {
+      "epoch": 0.7843946562455405,
+      "grad_norm": 1.578125,
+      "learning_rate": 5.550868239118745e-06,
+      "loss": 0.3245,
+      "step": 17864
+    },
+    {
+      "epoch": 0.784482474724744,
+      "grad_norm": 1.5390625,
+      "learning_rate": 5.5465221758353945e-06,
+      "loss": 0.2906,
+      "step": 17866
+    },
+    {
+      "epoch": 0.7845702932039474,
+      "grad_norm": 1.53125,
+      "learning_rate": 5.542177602304047e-06,
+      "loss": 0.3384,
+      "step": 17868
+    },
+    {
+      "epoch": 0.7846581116831509,
+      "grad_norm": 1.4140625,
+      "learning_rate": 5.537834518857401e-06,
+      "loss": 0.3252,
+      "step": 17870
+    },
+    {
+      "epoch": 0.7847459301623544,
+      "grad_norm": 1.5,
+      "learning_rate": 5.5334929258280485e-06,
+      "loss": 0.3389,
+      "step": 17872
+    },
+    {
+      "epoch": 0.7848337486415579,
+      "grad_norm": 1.4453125,
+      "learning_rate": 5.5291528235484776e-06,
+      "loss": 0.3351,
+      "step": 17874
+    },
+    {
+      "epoch": 0.7849215671207613,
+      "grad_norm": 1.4453125,
+      "learning_rate": 5.524814212351048e-06,
+      "loss": 0.3114,
+      "step": 17876
+    },
+    {
+      "epoch": 0.7850093855999649,
+      "grad_norm": 1.4296875,
+      "learning_rate": 5.520477092568019e-06,
+      "loss": 0.3462,
+      "step": 17878
+    },
+    {
+      "epoch": 0.7850972040791684,
+      "grad_norm": 1.4765625,
+      "learning_rate": 5.516141464531524e-06,
+      "loss": 0.3045,
+      "step": 17880
+    },
+    {
+      "epoch": 0.7851850225583719,
+      "grad_norm": 1.5625,
+      "learning_rate": 5.511807328573579e-06,
+      "loss": 0.3109,
+      "step": 17882
+    },
+    {
+      "epoch": 0.7852728410375753,
+      "grad_norm": 1.4921875,
+      "learning_rate": 5.507474685026115e-06,
+      "loss": 0.3091,
+      "step": 17884
+    },
+    {
+      "epoch": 0.7853606595167788,
+      "grad_norm": 1.484375,
+      "learning_rate": 5.503143534220901e-06,
+      "loss": 0.3276,
+      "step": 17886
+    },
+    {
+      "epoch": 0.7854484779959823,
+      "grad_norm": 1.421875,
+      "learning_rate": 5.4988138764896305e-06,
+      "loss": 0.3302,
+      "step": 17888
+    },
+    {
+      "epoch": 0.7855362964751857,
+      "grad_norm": 1.5390625,
+      "learning_rate": 5.494485712163866e-06,
+      "loss": 0.3134,
+      "step": 17890
+    },
+    {
+      "epoch": 0.7856241149543893,
+      "grad_norm": 1.4375,
+      "learning_rate": 5.490159041575066e-06,
+      "loss": 0.3505,
+      "step": 17892
+    },
+    {
+      "epoch": 0.7857119334335928,
+      "grad_norm": 1.546875,
+      "learning_rate": 5.485833865054563e-06,
+      "loss": 0.3337,
+      "step": 17894
+    },
+    {
+      "epoch": 0.7857997519127963,
+      "grad_norm": 1.46875,
+      "learning_rate": 5.481510182933575e-06,
+      "loss": 0.2966,
+      "step": 17896
+    },
+    {
+      "epoch": 0.7858875703919997,
+      "grad_norm": 1.5625,
+      "learning_rate": 5.4771879955432205e-06,
+      "loss": 0.316,
+      "step": 17898
+    },
+    {
+      "epoch": 0.7859753888712032,
+      "grad_norm": 1.5078125,
+      "learning_rate": 5.472867303214485e-06,
+      "loss": 0.3278,
+      "step": 17900
+    },
+    {
+      "epoch": 0.7860632073504067,
+      "grad_norm": 1.4140625,
+      "learning_rate": 5.46854810627826e-06,
+      "loss": 0.2878,
+      "step": 17902
+    },
+    {
+      "epoch": 0.7861510258296102,
+      "grad_norm": 1.484375,
+      "learning_rate": 5.464230405065301e-06,
+      "loss": 0.3051,
+      "step": 17904
+    },
+    {
+      "epoch": 0.7862388443088136,
+      "grad_norm": 1.46875,
+      "learning_rate": 5.459914199906252e-06,
+      "loss": 0.3092,
+      "step": 17906
+    },
+    {
+      "epoch": 0.7863266627880172,
+      "grad_norm": 1.5546875,
+      "learning_rate": 5.455599491131669e-06,
+      "loss": 0.3125,
+      "step": 17908
+    },
+    {
+      "epoch": 0.7864144812672207,
+      "grad_norm": 1.4921875,
+      "learning_rate": 5.451286279071952e-06,
+      "loss": 0.3291,
+      "step": 17910
+    },
+    {
+      "epoch": 0.7865022997464242,
+      "grad_norm": 1.5625,
+      "learning_rate": 5.446974564057425e-06,
+      "loss": 0.3274,
+      "step": 17912
+    },
+    {
+      "epoch": 0.7865901182256276,
+      "grad_norm": 1.578125,
+      "learning_rate": 5.442664346418275e-06,
+      "loss": 0.3265,
+      "step": 17914
+    },
+    {
+      "epoch": 0.7866779367048311,
+      "grad_norm": 1.453125,
+      "learning_rate": 5.438355626484576e-06,
+      "loss": 0.3356,
+      "step": 17916
+    },
+    {
+      "epoch": 0.7867657551840346,
+      "grad_norm": 1.5,
+      "learning_rate": 5.434048404586292e-06,
+      "loss": 0.3155,
+      "step": 17918
+    },
+    {
+      "epoch": 0.786853573663238,
+      "grad_norm": 1.578125,
+      "learning_rate": 5.429742681053266e-06,
+      "loss": 0.3409,
+      "step": 17920
+    },
+    {
+      "epoch": 0.7869413921424415,
+      "grad_norm": 1.4296875,
+      "learning_rate": 5.4254384562152424e-06,
+      "loss": 0.3067,
+      "step": 17922
+    },
+    {
+      "epoch": 0.7870292106216451,
+      "grad_norm": 1.4921875,
+      "learning_rate": 5.421135730401828e-06,
+      "loss": 0.3168,
+      "step": 17924
+    },
+    {
+      "epoch": 0.7871170291008486,
+      "grad_norm": 1.5625,
+      "learning_rate": 5.416834503942539e-06,
+      "loss": 0.3115,
+      "step": 17926
+    },
+    {
+      "epoch": 0.787204847580052,
+      "grad_norm": 1.609375,
+      "learning_rate": 5.4125347771667585e-06,
+      "loss": 0.3487,
+      "step": 17928
+    },
+    {
+      "epoch": 0.7872926660592555,
+      "grad_norm": 1.546875,
+      "learning_rate": 5.408236550403753e-06,
+      "loss": 0.3071,
+      "step": 17930
+    },
+    {
+      "epoch": 0.787380484538459,
+      "grad_norm": 1.4453125,
+      "learning_rate": 5.403939823982698e-06,
+      "loss": 0.3221,
+      "step": 17932
+    },
+    {
+      "epoch": 0.7874683030176625,
+      "grad_norm": 1.5390625,
+      "learning_rate": 5.39964459823262e-06,
+      "loss": 0.299,
+      "step": 17934
+    },
+    {
+      "epoch": 0.7875561214968659,
+      "grad_norm": 1.421875,
+      "learning_rate": 5.395350873482463e-06,
+      "loss": 0.2888,
+      "step": 17936
+    },
+    {
+      "epoch": 0.7876439399760695,
+      "grad_norm": 1.484375,
+      "learning_rate": 5.391058650061032e-06,
+      "loss": 0.3037,
+      "step": 17938
+    },
+    {
+      "epoch": 0.787731758455273,
+      "grad_norm": 1.484375,
+      "learning_rate": 5.3867679282970345e-06,
+      "loss": 0.3282,
+      "step": 17940
+    },
+    {
+      "epoch": 0.7878195769344765,
+      "grad_norm": 1.46875,
+      "learning_rate": 5.382478708519051e-06,
+      "loss": 0.3471,
+      "step": 17942
+    },
+    {
+      "epoch": 0.7879073954136799,
+      "grad_norm": 1.4609375,
+      "learning_rate": 5.378190991055543e-06,
+      "loss": 0.292,
+      "step": 17944
+    },
+    {
+      "epoch": 0.7879952138928834,
+      "grad_norm": 1.625,
+      "learning_rate": 5.373904776234886e-06,
+      "loss": 0.3057,
+      "step": 17946
+    },
+    {
+      "epoch": 0.7880830323720869,
+      "grad_norm": 1.5078125,
+      "learning_rate": 5.369620064385294e-06,
+      "loss": 0.3136,
+      "step": 17948
+    },
+    {
+      "epoch": 0.7881708508512903,
+      "grad_norm": 1.59375,
+      "learning_rate": 5.36533685583491e-06,
+      "loss": 0.3183,
+      "step": 17950
+    },
+    {
+      "epoch": 0.7882586693304938,
+      "grad_norm": 1.5625,
+      "learning_rate": 5.361055150911729e-06,
+      "loss": 0.3312,
+      "step": 17952
+    },
+    {
+      "epoch": 0.7883464878096974,
+      "grad_norm": 1.4765625,
+      "learning_rate": 5.3567749499436605e-06,
+      "loss": 0.3276,
+      "step": 17954
+    },
+    {
+      "epoch": 0.7884343062889009,
+      "grad_norm": 1.5390625,
+      "learning_rate": 5.352496253258474e-06,
+      "loss": 0.3156,
+      "step": 17956
+    },
+    {
+      "epoch": 0.7885221247681043,
+      "grad_norm": 1.4921875,
+      "learning_rate": 5.348219061183826e-06,
+      "loss": 0.3053,
+      "step": 17958
+    },
+    {
+      "epoch": 0.7886099432473078,
+      "grad_norm": 1.515625,
+      "learning_rate": 5.34394337404728e-06,
+      "loss": 0.3108,
+      "step": 17960
+    },
+    {
+      "epoch": 0.7886977617265113,
+      "grad_norm": 1.484375,
+      "learning_rate": 5.339669192176258e-06,
+      "loss": 0.3091,
+      "step": 17962
+    },
+    {
+      "epoch": 0.7887855802057148,
+      "grad_norm": 1.4453125,
+      "learning_rate": 5.335396515898086e-06,
+      "loss": 0.2981,
+      "step": 17964
+    },
+    {
+      "epoch": 0.7888733986849182,
+      "grad_norm": 1.5078125,
+      "learning_rate": 5.331125345539967e-06,
+      "loss": 0.344,
+      "step": 17966
+    },
+    {
+      "epoch": 0.7889612171641217,
+      "grad_norm": 1.46875,
+      "learning_rate": 5.326855681428974e-06,
+      "loss": 0.3075,
+      "step": 17968
+    },
+    {
+      "epoch": 0.7890490356433253,
+      "grad_norm": 1.4921875,
+      "learning_rate": 5.3225875238920945e-06,
+      "loss": 0.311,
+      "step": 17970
+    },
+    {
+      "epoch": 0.7891368541225288,
+      "grad_norm": 1.5,
+      "learning_rate": 5.318320873256174e-06,
+      "loss": 0.3374,
+      "step": 17972
+    },
+    {
+      "epoch": 0.7892246726017322,
+      "grad_norm": 1.4609375,
+      "learning_rate": 5.314055729847967e-06,
+      "loss": 0.2925,
+      "step": 17974
+    },
+    {
+      "epoch": 0.7893124910809357,
+      "grad_norm": 1.5,
+      "learning_rate": 5.309792093994093e-06,
+      "loss": 0.3213,
+      "step": 17976
+    },
+    {
+      "epoch": 0.7894003095601392,
+      "grad_norm": 1.5546875,
+      "learning_rate": 5.3055299660210555e-06,
+      "loss": 0.3087,
+      "step": 17978
+    },
+    {
+      "epoch": 0.7894881280393427,
+      "grad_norm": 1.46875,
+      "learning_rate": 5.301269346255258e-06,
+      "loss": 0.3155,
+      "step": 17980
+    },
+    {
+      "epoch": 0.7895759465185461,
+      "grad_norm": 1.4375,
+      "learning_rate": 5.2970102350229675e-06,
+      "loss": 0.3026,
+      "step": 17982
+    },
+    {
+      "epoch": 0.7896637649977497,
+      "grad_norm": 1.546875,
+      "learning_rate": 5.292752632650363e-06,
+      "loss": 0.3308,
+      "step": 17984
+    },
+    {
+      "epoch": 0.7897515834769532,
+      "grad_norm": 1.53125,
+      "learning_rate": 5.288496539463481e-06,
+      "loss": 0.3074,
+      "step": 17986
+    },
+    {
+      "epoch": 0.7898394019561567,
+      "grad_norm": 1.578125,
+      "learning_rate": 5.284241955788266e-06,
+      "loss": 0.3303,
+      "step": 17988
+    },
+    {
+      "epoch": 0.7899272204353601,
+      "grad_norm": 1.546875,
+      "learning_rate": 5.279988881950526e-06,
+      "loss": 0.3099,
+      "step": 17990
+    },
+    {
+      "epoch": 0.7900150389145636,
+      "grad_norm": 1.4609375,
+      "learning_rate": 5.2757373182759585e-06,
+      "loss": 0.3013,
+      "step": 17992
+    },
+    {
+      "epoch": 0.7901028573937671,
+      "grad_norm": 1.671875,
+      "learning_rate": 5.271487265090163e-06,
+      "loss": 0.3248,
+      "step": 17994
+    },
+    {
+      "epoch": 0.7901906758729705,
+      "grad_norm": 1.8984375,
+      "learning_rate": 5.2672387227185954e-06,
+      "loss": 0.3134,
+      "step": 17996
+    },
+    {
+      "epoch": 0.790278494352174,
+      "grad_norm": 1.5,
+      "learning_rate": 5.262991691486624e-06,
+      "loss": 0.3233,
+      "step": 17998
+    },
+    {
+      "epoch": 0.7903663128313776,
+      "grad_norm": 1.484375,
+      "learning_rate": 5.25874617171947e-06,
+      "loss": 0.31,
+      "step": 18000
+    },
+    {
+      "epoch": 0.7904541313105811,
+      "grad_norm": 1.5703125,
+      "learning_rate": 5.254502163742275e-06,
+      "loss": 0.2897,
+      "step": 18002
+    },
+    {
+      "epoch": 0.7905419497897845,
+      "grad_norm": 1.4765625,
+      "learning_rate": 5.250259667880039e-06,
+      "loss": 0.317,
+      "step": 18004
+    },
+    {
+      "epoch": 0.790629768268988,
+      "grad_norm": 1.4921875,
+      "learning_rate": 5.246018684457646e-06,
+      "loss": 0.2996,
+      "step": 18006
+    },
+    {
+      "epoch": 0.7907175867481915,
+      "grad_norm": 1.4609375,
+      "learning_rate": 5.241779213799888e-06,
+      "loss": 0.3249,
+      "step": 18008
+    },
+    {
+      "epoch": 0.790805405227395,
+      "grad_norm": 1.515625,
+      "learning_rate": 5.237541256231402e-06,
+      "loss": 0.3213,
+      "step": 18010
+    },
+    {
+      "epoch": 0.7908932237065984,
+      "grad_norm": 1.5546875,
+      "learning_rate": 5.23330481207675e-06,
+      "loss": 0.3026,
+      "step": 18012
+    },
+    {
+      "epoch": 0.7909810421858019,
+      "grad_norm": 1.546875,
+      "learning_rate": 5.22906988166035e-06,
+      "loss": 0.3241,
+      "step": 18014
+    },
+    {
+      "epoch": 0.7910688606650055,
+      "grad_norm": 1.4921875,
+      "learning_rate": 5.224836465306521e-06,
+      "loss": 0.3207,
+      "step": 18016
+    },
+    {
+      "epoch": 0.791156679144209,
+      "grad_norm": 1.4296875,
+      "learning_rate": 5.22060456333946e-06,
+      "loss": 0.3222,
+      "step": 18018
+    },
+    {
+      "epoch": 0.7912444976234124,
+      "grad_norm": 1.453125,
+      "learning_rate": 5.216374176083233e-06,
+      "loss": 0.3284,
+      "step": 18020
+    },
+    {
+      "epoch": 0.7913323161026159,
+      "grad_norm": 1.5625,
+      "learning_rate": 5.212145303861821e-06,
+      "loss": 0.311,
+      "step": 18022
+    },
+    {
+      "epoch": 0.7914201345818194,
+      "grad_norm": 1.5078125,
+      "learning_rate": 5.207917946999058e-06,
+      "loss": 0.3093,
+      "step": 18024
+    },
+    {
+      "epoch": 0.7915079530610228,
+      "grad_norm": 1.5546875,
+      "learning_rate": 5.2036921058186915e-06,
+      "loss": 0.3083,
+      "step": 18026
+    },
+    {
+      "epoch": 0.7915957715402263,
+      "grad_norm": 1.59375,
+      "learning_rate": 5.199467780644329e-06,
+      "loss": 0.3385,
+      "step": 18028
+    },
+    {
+      "epoch": 0.7916835900194298,
+      "grad_norm": 1.5546875,
+      "learning_rate": 5.195244971799462e-06,
+      "loss": 0.3325,
+      "step": 18030
+    },
+    {
+      "epoch": 0.7917714084986334,
+      "grad_norm": 1.4921875,
+      "learning_rate": 5.19102367960749e-06,
+      "loss": 0.3155,
+      "step": 18032
+    },
+    {
+      "epoch": 0.7918592269778368,
+      "grad_norm": 1.5625,
+      "learning_rate": 5.186803904391669e-06,
+      "loss": 0.3371,
+      "step": 18034
+    },
+    {
+      "epoch": 0.7919470454570403,
+      "grad_norm": 1.421875,
+      "learning_rate": 5.1825856464751575e-06,
+      "loss": 0.3148,
+      "step": 18036
+    },
+    {
+      "epoch": 0.7920348639362438,
+      "grad_norm": 1.5,
+      "learning_rate": 5.178368906180989e-06,
+      "loss": 0.2917,
+      "step": 18038
+    },
+    {
+      "epoch": 0.7921226824154473,
+      "grad_norm": 1.484375,
+      "learning_rate": 5.174153683832081e-06,
+      "loss": 0.3068,
+      "step": 18040
+    },
+    {
+      "epoch": 0.7922105008946507,
+      "grad_norm": 1.4453125,
+      "learning_rate": 5.1699399797512375e-06,
+      "loss": 0.3011,
+      "step": 18042
+    },
+    {
+      "epoch": 0.7922983193738542,
+      "grad_norm": 1.59375,
+      "learning_rate": 5.165727794261135e-06,
+      "loss": 0.3002,
+      "step": 18044
+    },
+    {
+      "epoch": 0.7923861378530578,
+      "grad_norm": 1.5234375,
+      "learning_rate": 5.161517127684362e-06,
+      "loss": 0.3056,
+      "step": 18046
+    },
+    {
+      "epoch": 0.7924739563322613,
+      "grad_norm": 1.59375,
+      "learning_rate": 5.157307980343357e-06,
+      "loss": 0.3399,
+      "step": 18048
+    },
+    {
+      "epoch": 0.7925617748114647,
+      "grad_norm": 1.578125,
+      "learning_rate": 5.153100352560467e-06,
+      "loss": 0.2994,
+      "step": 18050
+    },
+    {
+      "epoch": 0.7926495932906682,
+      "grad_norm": 1.4921875,
+      "learning_rate": 5.148894244657912e-06,
+      "loss": 0.3117,
+      "step": 18052
+    },
+    {
+      "epoch": 0.7927374117698717,
+      "grad_norm": 1.4765625,
+      "learning_rate": 5.144689656957785e-06,
+      "loss": 0.3101,
+      "step": 18054
+    },
+    {
+      "epoch": 0.7928252302490751,
+      "grad_norm": 1.4765625,
+      "learning_rate": 5.140486589782092e-06,
+      "loss": 0.3079,
+      "step": 18056
+    },
+    {
+      "epoch": 0.7929130487282786,
+      "grad_norm": 1.6015625,
+      "learning_rate": 5.13628504345269e-06,
+      "loss": 0.3244,
+      "step": 18058
+    },
+    {
+      "epoch": 0.7930008672074821,
+      "grad_norm": 1.4765625,
+      "learning_rate": 5.13208501829135e-06,
+      "loss": 0.3635,
+      "step": 18060
+    },
+    {
+      "epoch": 0.7930886856866857,
+      "grad_norm": 1.4609375,
+      "learning_rate": 5.127886514619698e-06,
+      "loss": 0.3079,
+      "step": 18062
+    },
+    {
+      "epoch": 0.7931765041658891,
+      "grad_norm": 1.4140625,
+      "learning_rate": 5.123689532759254e-06,
+      "loss": 0.2855,
+      "step": 18064
+    },
+    {
+      "epoch": 0.7932643226450926,
+      "grad_norm": 1.5703125,
+      "learning_rate": 5.119494073031439e-06,
+      "loss": 0.3265,
+      "step": 18066
+    },
+    {
+      "epoch": 0.7933521411242961,
+      "grad_norm": 1.4140625,
+      "learning_rate": 5.115300135757534e-06,
+      "loss": 0.3127,
+      "step": 18068
+    },
+    {
+      "epoch": 0.7934399596034996,
+      "grad_norm": 1.5625,
+      "learning_rate": 5.11110772125871e-06,
+      "loss": 0.3004,
+      "step": 18070
+    },
+    {
+      "epoch": 0.793527778082703,
+      "grad_norm": 1.5,
+      "learning_rate": 5.1069168298560176e-06,
+      "loss": 0.3214,
+      "step": 18072
+    },
+    {
+      "epoch": 0.7936155965619065,
+      "grad_norm": 1.453125,
+      "learning_rate": 5.10272746187041e-06,
+      "loss": 0.3196,
+      "step": 18074
+    },
+    {
+      "epoch": 0.79370341504111,
+      "grad_norm": 1.4375,
+      "learning_rate": 5.098539617622697e-06,
+      "loss": 0.3235,
+      "step": 18076
+    },
+    {
+      "epoch": 0.7937912335203136,
+      "grad_norm": 1.484375,
+      "learning_rate": 5.094353297433596e-06,
+      "loss": 0.3163,
+      "step": 18078
+    },
+    {
+      "epoch": 0.793879051999517,
+      "grad_norm": 1.4921875,
+      "learning_rate": 5.090168501623693e-06,
+      "loss": 0.3085,
+      "step": 18080
+    },
+    {
+      "epoch": 0.7939668704787205,
+      "grad_norm": 1.4453125,
+      "learning_rate": 5.085985230513451e-06,
+      "loss": 0.2908,
+      "step": 18082
+    },
+    {
+      "epoch": 0.794054688957924,
+      "grad_norm": 1.53125,
+      "learning_rate": 5.081803484423242e-06,
+      "loss": 0.3026,
+      "step": 18084
+    },
+    {
+      "epoch": 0.7941425074371274,
+      "grad_norm": 1.546875,
+      "learning_rate": 5.077623263673289e-06,
+      "loss": 0.3191,
+      "step": 18086
+    },
+    {
+      "epoch": 0.7942303259163309,
+      "grad_norm": 1.5078125,
+      "learning_rate": 5.07344456858373e-06,
+      "loss": 0.3342,
+      "step": 18088
+    },
+    {
+      "epoch": 0.7943181443955344,
+      "grad_norm": 1.546875,
+      "learning_rate": 5.069267399474559e-06,
+      "loss": 0.3048,
+      "step": 18090
+    },
+    {
+      "epoch": 0.794405962874738,
+      "grad_norm": 1.5625,
+      "learning_rate": 5.0650917566656656e-06,
+      "loss": 0.2963,
+      "step": 18092
+    },
+    {
+      "epoch": 0.7944937813539414,
+      "grad_norm": 1.484375,
+      "learning_rate": 5.0609176404768285e-06,
+      "loss": 0.3213,
+      "step": 18094
+    },
+    {
+      "epoch": 0.7945815998331449,
+      "grad_norm": 1.515625,
+      "learning_rate": 5.056745051227693e-06,
+      "loss": 0.3337,
+      "step": 18096
+    },
+    {
+      "epoch": 0.7946694183123484,
+      "grad_norm": 1.53125,
+      "learning_rate": 5.052573989237808e-06,
+      "loss": 0.3239,
+      "step": 18098
+    },
+    {
+      "epoch": 0.7947572367915519,
+      "grad_norm": 1.53125,
+      "learning_rate": 5.048404454826588e-06,
+      "loss": 0.3279,
+      "step": 18100
+    },
+    {
+      "epoch": 0.7948450552707553,
+      "grad_norm": 1.484375,
+      "learning_rate": 5.044236448313339e-06,
+      "loss": 0.3332,
+      "step": 18102
+    },
+    {
+      "epoch": 0.7949328737499588,
+      "grad_norm": 1.578125,
+      "learning_rate": 5.040069970017247e-06,
+      "loss": 0.3256,
+      "step": 18104
+    },
+    {
+      "epoch": 0.7950206922291623,
+      "grad_norm": 1.4765625,
+      "learning_rate": 5.035905020257373e-06,
+      "loss": 0.3154,
+      "step": 18106
+    },
+    {
+      "epoch": 0.7951085107083659,
+      "grad_norm": 1.5546875,
+      "learning_rate": 5.031741599352685e-06,
+      "loss": 0.3131,
+      "step": 18108
+    },
+    {
+      "epoch": 0.7951963291875693,
+      "grad_norm": 1.4375,
+      "learning_rate": 5.027579707622007e-06,
+      "loss": 0.3013,
+      "step": 18110
+    },
+    {
+      "epoch": 0.7952841476667728,
+      "grad_norm": 1.6171875,
+      "learning_rate": 5.02341934538407e-06,
+      "loss": 0.3147,
+      "step": 18112
+    },
+    {
+      "epoch": 0.7953719661459763,
+      "grad_norm": 1.5546875,
+      "learning_rate": 5.019260512957466e-06,
+      "loss": 0.3319,
+      "step": 18114
+    },
+    {
+      "epoch": 0.7954597846251797,
+      "grad_norm": 1.4921875,
+      "learning_rate": 5.0151032106606764e-06,
+      "loss": 0.3055,
+      "step": 18116
+    },
+    {
+      "epoch": 0.7955476031043832,
+      "grad_norm": 1.484375,
+      "learning_rate": 5.010947438812078e-06,
+      "loss": 0.3211,
+      "step": 18118
+    },
+    {
+      "epoch": 0.7956354215835867,
+      "grad_norm": 1.5859375,
+      "learning_rate": 5.006793197729912e-06,
+      "loss": 0.3466,
+      "step": 18120
+    },
+    {
+      "epoch": 0.7957232400627902,
+      "grad_norm": 1.484375,
+      "learning_rate": 5.002640487732321e-06,
+      "loss": 0.3254,
+      "step": 18122
+    },
+    {
+      "epoch": 0.7958110585419937,
+      "grad_norm": 1.4375,
+      "learning_rate": 4.9984893091373165e-06,
+      "loss": 0.3338,
+      "step": 18124
+    },
+    {
+      "epoch": 0.7958988770211972,
+      "grad_norm": 1.53125,
+      "learning_rate": 4.994339662262787e-06,
+      "loss": 0.3281,
+      "step": 18126
+    },
+    {
+      "epoch": 0.7959866955004007,
+      "grad_norm": 1.453125,
+      "learning_rate": 4.990191547426531e-06,
+      "loss": 0.3051,
+      "step": 18128
+    },
+    {
+      "epoch": 0.7960745139796042,
+      "grad_norm": 1.4453125,
+      "learning_rate": 4.986044964946201e-06,
+      "loss": 0.3234,
+      "step": 18130
+    },
+    {
+      "epoch": 0.7961623324588076,
+      "grad_norm": 1.515625,
+      "learning_rate": 4.981899915139346e-06,
+      "loss": 0.3067,
+      "step": 18132
+    },
+    {
+      "epoch": 0.7962501509380111,
+      "grad_norm": 1.5703125,
+      "learning_rate": 4.977756398323388e-06,
+      "loss": 0.2966,
+      "step": 18134
+    },
+    {
+      "epoch": 0.7963379694172146,
+      "grad_norm": 1.4765625,
+      "learning_rate": 4.97361441481565e-06,
+      "loss": 0.3272,
+      "step": 18136
+    },
+    {
+      "epoch": 0.7964257878964182,
+      "grad_norm": 1.546875,
+      "learning_rate": 4.969473964933313e-06,
+      "loss": 0.2979,
+      "step": 18138
+    },
+    {
+      "epoch": 0.7965136063756216,
+      "grad_norm": 1.484375,
+      "learning_rate": 4.965335048993472e-06,
+      "loss": 0.2876,
+      "step": 18140
+    },
+    {
+      "epoch": 0.7966014248548251,
+      "grad_norm": 1.515625,
+      "learning_rate": 4.961197667313072e-06,
+      "loss": 0.3546,
+      "step": 18142
+    },
+    {
+      "epoch": 0.7966892433340286,
+      "grad_norm": 1.5546875,
+      "learning_rate": 4.957061820208952e-06,
+      "loss": 0.3265,
+      "step": 18144
+    },
+    {
+      "epoch": 0.796777061813232,
+      "grad_norm": 1.453125,
+      "learning_rate": 4.952927507997851e-06,
+      "loss": 0.3119,
+      "step": 18146
+    },
+    {
+      "epoch": 0.7968648802924355,
+      "grad_norm": 1.5,
+      "learning_rate": 4.948794730996359e-06,
+      "loss": 0.3151,
+      "step": 18148
+    },
+    {
+      "epoch": 0.796952698771639,
+      "grad_norm": 1.4375,
+      "learning_rate": 4.9446634895209815e-06,
+      "loss": 0.3152,
+      "step": 18150
+    },
+    {
+      "epoch": 0.7970405172508425,
+      "grad_norm": 1.5234375,
+      "learning_rate": 4.940533783888079e-06,
+      "loss": 0.3161,
+      "step": 18152
+    },
+    {
+      "epoch": 0.797128335730046,
+      "grad_norm": 1.5390625,
+      "learning_rate": 4.936405614413903e-06,
+      "loss": 0.3029,
+      "step": 18154
+    },
+    {
+      "epoch": 0.7972161542092495,
+      "grad_norm": 1.5,
+      "learning_rate": 4.932278981414601e-06,
+      "loss": 0.3507,
+      "step": 18156
+    },
+    {
+      "epoch": 0.797303972688453,
+      "grad_norm": 1.4921875,
+      "learning_rate": 4.92815388520618e-06,
+      "loss": 0.3181,
+      "step": 18158
+    },
+    {
+      "epoch": 0.7973917911676565,
+      "grad_norm": 1.453125,
+      "learning_rate": 4.924030326104556e-06,
+      "loss": 0.303,
+      "step": 18160
+    },
+    {
+      "epoch": 0.7974796096468599,
+      "grad_norm": 1.5625,
+      "learning_rate": 4.9199083044254915e-06,
+      "loss": 0.3098,
+      "step": 18162
+    },
+    {
+      "epoch": 0.7975674281260634,
+      "grad_norm": 1.453125,
+      "learning_rate": 4.915787820484669e-06,
+      "loss": 0.3075,
+      "step": 18164
+    },
+    {
+      "epoch": 0.7976552466052669,
+      "grad_norm": 1.4921875,
+      "learning_rate": 4.911668874597628e-06,
+      "loss": 0.326,
+      "step": 18166
+    },
+    {
+      "epoch": 0.7977430650844703,
+      "grad_norm": 1.5234375,
+      "learning_rate": 4.9075514670797935e-06,
+      "loss": 0.3283,
+      "step": 18168
+    },
+    {
+      "epoch": 0.7978308835636739,
+      "grad_norm": 1.484375,
+      "learning_rate": 4.903435598246492e-06,
+      "loss": 0.3154,
+      "step": 18170
+    },
+    {
+      "epoch": 0.7979187020428774,
+      "grad_norm": 1.484375,
+      "learning_rate": 4.899321268412904e-06,
+      "loss": 0.3441,
+      "step": 18172
+    },
+    {
+      "epoch": 0.7980065205220809,
+      "grad_norm": 1.6015625,
+      "learning_rate": 4.895208477894117e-06,
+      "loss": 0.3069,
+      "step": 18174
+    },
+    {
+      "epoch": 0.7980943390012843,
+      "grad_norm": 1.5234375,
+      "learning_rate": 4.891097227005085e-06,
+      "loss": 0.307,
+      "step": 18176
+    },
+    {
+      "epoch": 0.7981821574804878,
+      "grad_norm": 1.46875,
+      "learning_rate": 4.88698751606064e-06,
+      "loss": 0.3102,
+      "step": 18178
+    },
+    {
+      "epoch": 0.7982699759596913,
+      "grad_norm": 1.5,
+      "learning_rate": 4.882879345375521e-06,
+      "loss": 0.3094,
+      "step": 18180
+    },
+    {
+      "epoch": 0.7983577944388948,
+      "grad_norm": 1.515625,
+      "learning_rate": 4.878772715264315e-06,
+      "loss": 0.3016,
+      "step": 18182
+    },
+    {
+      "epoch": 0.7984456129180982,
+      "grad_norm": 1.4453125,
+      "learning_rate": 4.874667626041526e-06,
+      "loss": 0.3108,
+      "step": 18184
+    },
+    {
+      "epoch": 0.7985334313973018,
+      "grad_norm": 1.5703125,
+      "learning_rate": 4.870564078021514e-06,
+      "loss": 0.3263,
+      "step": 18186
+    },
+    {
+      "epoch": 0.7986212498765053,
+      "grad_norm": 1.46875,
+      "learning_rate": 4.866462071518524e-06,
+      "loss": 0.2997,
+      "step": 18188
+    },
+    {
+      "epoch": 0.7987090683557088,
+      "grad_norm": 1.59375,
+      "learning_rate": 4.862361606846702e-06,
+      "loss": 0.3249,
+      "step": 18190
+    },
+    {
+      "epoch": 0.7987968868349122,
+      "grad_norm": 1.671875,
+      "learning_rate": 4.858262684320056e-06,
+      "loss": 0.3326,
+      "step": 18192
+    },
+    {
+      "epoch": 0.7988847053141157,
+      "grad_norm": 1.5546875,
+      "learning_rate": 4.854165304252481e-06,
+      "loss": 0.3095,
+      "step": 18194
+    },
+    {
+      "epoch": 0.7989725237933192,
+      "grad_norm": 1.46875,
+      "learning_rate": 4.850069466957749e-06,
+      "loss": 0.3536,
+      "step": 18196
+    },
+    {
+      "epoch": 0.7990603422725227,
+      "grad_norm": 1.5078125,
+      "learning_rate": 4.8459751727495335e-06,
+      "loss": 0.3048,
+      "step": 18198
+    },
+    {
+      "epoch": 0.7991481607517262,
+      "grad_norm": 1.5390625,
+      "learning_rate": 4.841882421941365e-06,
+      "loss": 0.3148,
+      "step": 18200
+    },
+    {
+      "epoch": 0.7992359792309297,
+      "grad_norm": 1.53125,
+      "learning_rate": 4.837791214846679e-06,
+      "loss": 0.3154,
+      "step": 18202
+    },
+    {
+      "epoch": 0.7993237977101332,
+      "grad_norm": 1.515625,
+      "learning_rate": 4.833701551778777e-06,
+      "loss": 0.3204,
+      "step": 18204
+    },
+    {
+      "epoch": 0.7994116161893366,
+      "grad_norm": 1.4375,
+      "learning_rate": 4.829613433050837e-06,
+      "loss": 0.3105,
+      "step": 18206
+    },
+    {
+      "epoch": 0.7994994346685401,
+      "grad_norm": 1.4453125,
+      "learning_rate": 4.82552685897594e-06,
+      "loss": 0.2983,
+      "step": 18208
+    },
+    {
+      "epoch": 0.7995872531477436,
+      "grad_norm": 1.4921875,
+      "learning_rate": 4.8214418298670264e-06,
+      "loss": 0.3086,
+      "step": 18210
+    },
+    {
+      "epoch": 0.7996750716269471,
+      "grad_norm": 1.5,
+      "learning_rate": 4.8173583460369435e-06,
+      "loss": 0.3359,
+      "step": 18212
+    },
+    {
+      "epoch": 0.7997628901061505,
+      "grad_norm": 1.484375,
+      "learning_rate": 4.813276407798395e-06,
+      "loss": 0.2954,
+      "step": 18214
+    },
+    {
+      "epoch": 0.7998507085853541,
+      "grad_norm": 1.4921875,
+      "learning_rate": 4.809196015463971e-06,
+      "loss": 0.3317,
+      "step": 18216
+    },
+    {
+      "epoch": 0.7999385270645576,
+      "grad_norm": 1.515625,
+      "learning_rate": 4.805117169346163e-06,
+      "loss": 0.3251,
+      "step": 18218
+    },
+    {
+      "epoch": 0.8000263455437611,
+      "grad_norm": 1.4609375,
+      "learning_rate": 4.801039869757318e-06,
+      "loss": 0.3396,
+      "step": 18220
+    },
+    {
+      "epoch": 0.8001141640229645,
+      "grad_norm": 1.609375,
+      "learning_rate": 4.7969641170096944e-06,
+      "loss": 0.3243,
+      "step": 18222
+    },
+    {
+      "epoch": 0.800201982502168,
+      "grad_norm": 1.5,
+      "learning_rate": 4.792889911415388e-06,
+      "loss": 0.2899,
+      "step": 18224
+    },
+    {
+      "epoch": 0.8002898009813715,
+      "grad_norm": 1.4296875,
+      "learning_rate": 4.788817253286424e-06,
+      "loss": 0.3137,
+      "step": 18226
+    },
+    {
+      "epoch": 0.800377619460575,
+      "grad_norm": 1.4296875,
+      "learning_rate": 4.784746142934676e-06,
+      "loss": 0.2939,
+      "step": 18228
+    },
+    {
+      "epoch": 0.8004654379397784,
+      "grad_norm": 1.5234375,
+      "learning_rate": 4.780676580671911e-06,
+      "loss": 0.3131,
+      "step": 18230
+    },
+    {
+      "epoch": 0.800553256418982,
+      "grad_norm": 1.5078125,
+      "learning_rate": 4.776608566809787e-06,
+      "loss": 0.3141,
+      "step": 18232
+    },
+    {
+      "epoch": 0.8006410748981855,
+      "grad_norm": 1.4609375,
+      "learning_rate": 4.772542101659819e-06,
+      "loss": 0.3322,
+      "step": 18234
+    },
+    {
+      "epoch": 0.800728893377389,
+      "grad_norm": 1.4609375,
+      "learning_rate": 4.7684771855334324e-06,
+      "loss": 0.3079,
+      "step": 18236
+    },
+    {
+      "epoch": 0.8008167118565924,
+      "grad_norm": 1.4453125,
+      "learning_rate": 4.764413818741914e-06,
+      "loss": 0.3413,
+      "step": 18238
+    },
+    {
+      "epoch": 0.8009045303357959,
+      "grad_norm": 1.53125,
+      "learning_rate": 4.760352001596427e-06,
+      "loss": 0.3347,
+      "step": 18240
+    },
+    {
+      "epoch": 0.8009923488149994,
+      "grad_norm": 1.5,
+      "learning_rate": 4.756291734408044e-06,
+      "loss": 0.3092,
+      "step": 18242
+    },
+    {
+      "epoch": 0.8010801672942028,
+      "grad_norm": 1.421875,
+      "learning_rate": 4.752233017487687e-06,
+      "loss": 0.3148,
+      "step": 18244
+    },
+    {
+      "epoch": 0.8011679857734064,
+      "grad_norm": 1.53125,
+      "learning_rate": 4.748175851146186e-06,
+      "loss": 0.3175,
+      "step": 18246
+    },
+    {
+      "epoch": 0.8012558042526099,
+      "grad_norm": 1.453125,
+      "learning_rate": 4.74412023569423e-06,
+      "loss": 0.2983,
+      "step": 18248
+    },
+    {
+      "epoch": 0.8013436227318134,
+      "grad_norm": 1.4296875,
+      "learning_rate": 4.740066171442398e-06,
+      "loss": 0.3073,
+      "step": 18250
+    },
+    {
+      "epoch": 0.8014314412110168,
+      "grad_norm": 1.6171875,
+      "learning_rate": 4.7360136587011665e-06,
+      "loss": 0.321,
+      "step": 18252
+    },
+    {
+      "epoch": 0.8015192596902203,
+      "grad_norm": 1.4921875,
+      "learning_rate": 4.731962697780856e-06,
+      "loss": 0.3364,
+      "step": 18254
+    },
+    {
+      "epoch": 0.8016070781694238,
+      "grad_norm": 1.421875,
+      "learning_rate": 4.727913288991706e-06,
+      "loss": 0.2831,
+      "step": 18256
+    },
+    {
+      "epoch": 0.8016948966486273,
+      "grad_norm": 1.5546875,
+      "learning_rate": 4.723865432643809e-06,
+      "loss": 0.3107,
+      "step": 18258
+    },
+    {
+      "epoch": 0.8017827151278307,
+      "grad_norm": 1.59375,
+      "learning_rate": 4.719819129047165e-06,
+      "loss": 0.322,
+      "step": 18260
+    },
+    {
+      "epoch": 0.8018705336070343,
+      "grad_norm": 1.6640625,
+      "learning_rate": 4.7157743785116255e-06,
+      "loss": 0.2842,
+      "step": 18262
+    },
+    {
+      "epoch": 0.8019583520862378,
+      "grad_norm": 1.546875,
+      "learning_rate": 4.711731181346954e-06,
+      "loss": 0.3266,
+      "step": 18264
+    },
+    {
+      "epoch": 0.8020461705654413,
+      "grad_norm": 1.4296875,
+      "learning_rate": 4.707689537862772e-06,
+      "loss": 0.2794,
+      "step": 18266
+    },
+    {
+      "epoch": 0.8021339890446447,
+      "grad_norm": 1.484375,
+      "learning_rate": 4.703649448368583e-06,
+      "loss": 0.3098,
+      "step": 18268
+    },
+    {
+      "epoch": 0.8022218075238482,
+      "grad_norm": 1.5859375,
+      "learning_rate": 4.699610913173791e-06,
+      "loss": 0.3132,
+      "step": 18270
+    },
+    {
+      "epoch": 0.8023096260030517,
+      "grad_norm": 1.5078125,
+      "learning_rate": 4.695573932587657e-06,
+      "loss": 0.3035,
+      "step": 18272
+    },
+    {
+      "epoch": 0.8023974444822551,
+      "grad_norm": 1.4140625,
+      "learning_rate": 4.691538506919344e-06,
+      "loss": 0.3148,
+      "step": 18274
+    },
+    {
+      "epoch": 0.8024852629614586,
+      "grad_norm": 1.4921875,
+      "learning_rate": 4.68750463647788e-06,
+      "loss": 0.3232,
+      "step": 18276
+    },
+    {
+      "epoch": 0.8025730814406622,
+      "grad_norm": 1.4765625,
+      "learning_rate": 4.683472321572172e-06,
+      "loss": 0.3216,
+      "step": 18278
+    },
+    {
+      "epoch": 0.8026608999198657,
+      "grad_norm": 1.40625,
+      "learning_rate": 4.679441562511033e-06,
+      "loss": 0.3349,
+      "step": 18280
+    },
+    {
+      "epoch": 0.8027487183990691,
+      "grad_norm": 1.59375,
+      "learning_rate": 4.675412359603121e-06,
+      "loss": 0.2902,
+      "step": 18282
+    },
+    {
+      "epoch": 0.8028365368782726,
+      "grad_norm": 1.6015625,
+      "learning_rate": 4.671384713157018e-06,
+      "loss": 0.3061,
+      "step": 18284
+    },
+    {
+      "epoch": 0.8029243553574761,
+      "grad_norm": 1.5078125,
+      "learning_rate": 4.667358623481132e-06,
+      "loss": 0.3198,
+      "step": 18286
+    },
+    {
+      "epoch": 0.8030121738366796,
+      "grad_norm": 1.4921875,
+      "learning_rate": 4.663334090883806e-06,
+      "loss": 0.2927,
+      "step": 18288
+    },
+    {
+      "epoch": 0.803099992315883,
+      "grad_norm": 1.4921875,
+      "learning_rate": 4.659311115673229e-06,
+      "loss": 0.3303,
+      "step": 18290
+    },
+    {
+      "epoch": 0.8031878107950866,
+      "grad_norm": 1.4921875,
+      "learning_rate": 4.655289698157475e-06,
+      "loss": 0.3044,
+      "step": 18292
+    },
+    {
+      "epoch": 0.8032756292742901,
+      "grad_norm": 1.53125,
+      "learning_rate": 4.65126983864452e-06,
+      "loss": 0.3317,
+      "step": 18294
+    },
+    {
+      "epoch": 0.8033634477534936,
+      "grad_norm": 1.484375,
+      "learning_rate": 4.647251537442193e-06,
+      "loss": 0.2918,
+      "step": 18296
+    },
+    {
+      "epoch": 0.803451266232697,
+      "grad_norm": 1.46875,
+      "learning_rate": 4.643234794858229e-06,
+      "loss": 0.3144,
+      "step": 18298
+    },
+    {
+      "epoch": 0.8035390847119005,
+      "grad_norm": 1.5,
+      "learning_rate": 4.639219611200221e-06,
+      "loss": 0.3063,
+      "step": 18300
+    },
+    {
+      "epoch": 0.803626903191104,
+      "grad_norm": 1.53125,
+      "learning_rate": 4.635205986775654e-06,
+      "loss": 0.3148,
+      "step": 18302
+    },
+    {
+      "epoch": 0.8037147216703074,
+      "grad_norm": 1.453125,
+      "learning_rate": 4.6311939218918995e-06,
+      "loss": 0.3087,
+      "step": 18304
+    },
+    {
+      "epoch": 0.8038025401495109,
+      "grad_norm": 1.546875,
+      "learning_rate": 4.627183416856187e-06,
+      "loss": 0.2936,
+      "step": 18306
+    },
+    {
+      "epoch": 0.8038903586287145,
+      "grad_norm": 1.53125,
+      "learning_rate": 4.623174471975664e-06,
+      "loss": 0.2981,
+      "step": 18308
+    },
+    {
+      "epoch": 0.803978177107918,
+      "grad_norm": 1.578125,
+      "learning_rate": 4.619167087557322e-06,
+      "loss": 0.2947,
+      "step": 18310
+    },
+    {
+      "epoch": 0.8040659955871214,
+      "grad_norm": 1.5390625,
+      "learning_rate": 4.615161263908044e-06,
+      "loss": 0.3003,
+      "step": 18312
+    },
+    {
+      "epoch": 0.8041538140663249,
+      "grad_norm": 1.5078125,
+      "learning_rate": 4.611157001334615e-06,
+      "loss": 0.3225,
+      "step": 18314
+    },
+    {
+      "epoch": 0.8042416325455284,
+      "grad_norm": 1.5078125,
+      "learning_rate": 4.607154300143657e-06,
+      "loss": 0.2862,
+      "step": 18316
+    },
+    {
+      "epoch": 0.8043294510247319,
+      "grad_norm": 1.4765625,
+      "learning_rate": 4.603153160641719e-06,
+      "loss": 0.29,
+      "step": 18318
+    },
+    {
+      "epoch": 0.8044172695039353,
+      "grad_norm": 1.5234375,
+      "learning_rate": 4.5991535831351964e-06,
+      "loss": 0.3264,
+      "step": 18320
+    },
+    {
+      "epoch": 0.8045050879831388,
+      "grad_norm": 1.484375,
+      "learning_rate": 4.595155567930387e-06,
+      "loss": 0.2995,
+      "step": 18322
+    },
+    {
+      "epoch": 0.8045929064623424,
+      "grad_norm": 1.4765625,
+      "learning_rate": 4.591159115333454e-06,
+      "loss": 0.3019,
+      "step": 18324
+    },
+    {
+      "epoch": 0.8046807249415459,
+      "grad_norm": 1.4140625,
+      "learning_rate": 4.587164225650445e-06,
+      "loss": 0.3189,
+      "step": 18326
+    },
+    {
+      "epoch": 0.8047685434207493,
+      "grad_norm": 1.53125,
+      "learning_rate": 4.583170899187298e-06,
+      "loss": 0.3148,
+      "step": 18328
+    },
+    {
+      "epoch": 0.8048563618999528,
+      "grad_norm": 1.4609375,
+      "learning_rate": 4.579179136249812e-06,
+      "loss": 0.318,
+      "step": 18330
+    },
+    {
+      "epoch": 0.8049441803791563,
+      "grad_norm": 1.5859375,
+      "learning_rate": 4.5751889371436905e-06,
+      "loss": 0.3232,
+      "step": 18332
+    },
+    {
+      "epoch": 0.8050319988583597,
+      "grad_norm": 1.4296875,
+      "learning_rate": 4.571200302174489e-06,
+      "loss": 0.3035,
+      "step": 18334
+    },
+    {
+      "epoch": 0.8051198173375632,
+      "grad_norm": 1.625,
+      "learning_rate": 4.567213231647669e-06,
+      "loss": 0.3367,
+      "step": 18336
+    },
+    {
+      "epoch": 0.8052076358167668,
+      "grad_norm": 1.4921875,
+      "learning_rate": 4.563227725868561e-06,
+      "loss": 0.3073,
+      "step": 18338
+    },
+    {
+      "epoch": 0.8052954542959703,
+      "grad_norm": 1.4921875,
+      "learning_rate": 4.559243785142367e-06,
+      "loss": 0.3291,
+      "step": 18340
+    },
+    {
+      "epoch": 0.8053832727751737,
+      "grad_norm": 1.4453125,
+      "learning_rate": 4.555261409774187e-06,
+      "loss": 0.3237,
+      "step": 18342
+    },
+    {
+      "epoch": 0.8054710912543772,
+      "grad_norm": 1.4609375,
+      "learning_rate": 4.5512806000689916e-06,
+      "loss": 0.2984,
+      "step": 18344
+    },
+    {
+      "epoch": 0.8055589097335807,
+      "grad_norm": 1.3828125,
+      "learning_rate": 4.547301356331629e-06,
+      "loss": 0.3249,
+      "step": 18346
+    },
+    {
+      "epoch": 0.8056467282127842,
+      "grad_norm": 1.4765625,
+      "learning_rate": 4.543323678866826e-06,
+      "loss": 0.3269,
+      "step": 18348
+    },
+    {
+      "epoch": 0.8057345466919876,
+      "grad_norm": 1.453125,
+      "learning_rate": 4.539347567979205e-06,
+      "loss": 0.3146,
+      "step": 18350
+    },
+    {
+      "epoch": 0.8058223651711911,
+      "grad_norm": 1.59375,
+      "learning_rate": 4.535373023973253e-06,
+      "loss": 0.3335,
+      "step": 18352
+    },
+    {
+      "epoch": 0.8059101836503947,
+      "grad_norm": 1.53125,
+      "learning_rate": 4.531400047153331e-06,
+      "loss": 0.2805,
+      "step": 18354
+    },
+    {
+      "epoch": 0.8059980021295982,
+      "grad_norm": 1.53125,
+      "learning_rate": 4.52742863782371e-06,
+      "loss": 0.3067,
+      "step": 18356
+    },
+    {
+      "epoch": 0.8060858206088016,
+      "grad_norm": 1.5234375,
+      "learning_rate": 4.5234587962885045e-06,
+      "loss": 0.2892,
+      "step": 18358
+    },
+    {
+      "epoch": 0.8061736390880051,
+      "grad_norm": 1.5546875,
+      "learning_rate": 4.519490522851738e-06,
+      "loss": 0.3257,
+      "step": 18360
+    },
+    {
+      "epoch": 0.8062614575672086,
+      "grad_norm": 1.5234375,
+      "learning_rate": 4.515523817817297e-06,
+      "loss": 0.3211,
+      "step": 18362
+    },
+    {
+      "epoch": 0.806349276046412,
+      "grad_norm": 1.5078125,
+      "learning_rate": 4.511558681488945e-06,
+      "loss": 0.2904,
+      "step": 18364
+    },
+    {
+      "epoch": 0.8064370945256155,
+      "grad_norm": 1.4765625,
+      "learning_rate": 4.507595114170349e-06,
+      "loss": 0.3069,
+      "step": 18366
+    },
+    {
+      "epoch": 0.806524913004819,
+      "grad_norm": 1.515625,
+      "learning_rate": 4.503633116165026e-06,
+      "loss": 0.323,
+      "step": 18368
+    },
+    {
+      "epoch": 0.8066127314840226,
+      "grad_norm": 1.484375,
+      "learning_rate": 4.499672687776396e-06,
+      "loss": 0.3047,
+      "step": 18370
+    },
+    {
+      "epoch": 0.806700549963226,
+      "grad_norm": 1.5546875,
+      "learning_rate": 4.495713829307749e-06,
+      "loss": 0.3107,
+      "step": 18372
+    },
+    {
+      "epoch": 0.8067883684424295,
+      "grad_norm": 1.546875,
+      "learning_rate": 4.491756541062242e-06,
+      "loss": 0.3108,
+      "step": 18374
+    },
+    {
+      "epoch": 0.806876186921633,
+      "grad_norm": 1.4765625,
+      "learning_rate": 4.48780082334295e-06,
+      "loss": 0.3186,
+      "step": 18376
+    },
+    {
+      "epoch": 0.8069640054008365,
+      "grad_norm": 1.546875,
+      "learning_rate": 4.483846676452777e-06,
+      "loss": 0.3069,
+      "step": 18378
+    },
+    {
+      "epoch": 0.8070518238800399,
+      "grad_norm": 1.4765625,
+      "learning_rate": 4.479894100694545e-06,
+      "loss": 0.3001,
+      "step": 18380
+    },
+    {
+      "epoch": 0.8071396423592434,
+      "grad_norm": 1.5078125,
+      "learning_rate": 4.4759430963709406e-06,
+      "loss": 0.3085,
+      "step": 18382
+    },
+    {
+      "epoch": 0.8072274608384469,
+      "grad_norm": 1.7578125,
+      "learning_rate": 4.471993663784538e-06,
+      "loss": 0.3099,
+      "step": 18384
+    },
+    {
+      "epoch": 0.8073152793176505,
+      "grad_norm": 1.5625,
+      "learning_rate": 4.468045803237783e-06,
+      "loss": 0.2982,
+      "step": 18386
+    },
+    {
+      "epoch": 0.8074030977968539,
+      "grad_norm": 1.5,
+      "learning_rate": 4.464099515032993e-06,
+      "loss": 0.3134,
+      "step": 18388
+    },
+    {
+      "epoch": 0.8074909162760574,
+      "grad_norm": 1.46875,
+      "learning_rate": 4.460154799472394e-06,
+      "loss": 0.3161,
+      "step": 18390
+    },
+    {
+      "epoch": 0.8075787347552609,
+      "grad_norm": 1.4375,
+      "learning_rate": 4.456211656858056e-06,
+      "loss": 0.3432,
+      "step": 18392
+    },
+    {
+      "epoch": 0.8076665532344643,
+      "grad_norm": 1.5703125,
+      "learning_rate": 4.452270087491961e-06,
+      "loss": 0.317,
+      "step": 18394
+    },
+    {
+      "epoch": 0.8077543717136678,
+      "grad_norm": 1.4609375,
+      "learning_rate": 4.448330091675943e-06,
+      "loss": 0.3255,
+      "step": 18396
+    },
+    {
+      "epoch": 0.8078421901928713,
+      "grad_norm": 1.515625,
+      "learning_rate": 4.444391669711737e-06,
+      "loss": 0.2994,
+      "step": 18398
+    },
+    {
+      "epoch": 0.8079300086720749,
+      "grad_norm": 1.4609375,
+      "learning_rate": 4.440454821900947e-06,
+      "loss": 0.3137,
+      "step": 18400
+    },
+    {
+      "epoch": 0.8080178271512783,
+      "grad_norm": 1.46875,
+      "learning_rate": 4.436519548545049e-06,
+      "loss": 0.3011,
+      "step": 18402
+    },
+    {
+      "epoch": 0.8081056456304818,
+      "grad_norm": 1.4609375,
+      "learning_rate": 4.432585849945417e-06,
+      "loss": 0.3021,
+      "step": 18404
+    },
+    {
+      "epoch": 0.8081934641096853,
+      "grad_norm": 1.4453125,
+      "learning_rate": 4.428653726403292e-06,
+      "loss": 0.321,
+      "step": 18406
+    },
+    {
+      "epoch": 0.8082812825888888,
+      "grad_norm": 1.421875,
+      "learning_rate": 4.424723178219798e-06,
+      "loss": 0.304,
+      "step": 18408
+    },
+    {
+      "epoch": 0.8083691010680922,
+      "grad_norm": 1.5078125,
+      "learning_rate": 4.4207942056959275e-06,
+      "loss": 0.3304,
+      "step": 18410
+    },
+    {
+      "epoch": 0.8084569195472957,
+      "grad_norm": 1.4453125,
+      "learning_rate": 4.416866809132575e-06,
+      "loss": 0.2906,
+      "step": 18412
+    },
+    {
+      "epoch": 0.8085447380264992,
+      "grad_norm": 1.53125,
+      "learning_rate": 4.412940988830497e-06,
+      "loss": 0.3207,
+      "step": 18414
+    },
+    {
+      "epoch": 0.8086325565057028,
+      "grad_norm": 1.515625,
+      "learning_rate": 4.409016745090327e-06,
+      "loss": 0.3424,
+      "step": 18416
+    },
+    {
+      "epoch": 0.8087203749849062,
+      "grad_norm": 1.453125,
+      "learning_rate": 4.405094078212599e-06,
+      "loss": 0.3357,
+      "step": 18418
+    },
+    {
+      "epoch": 0.8088081934641097,
+      "grad_norm": 1.53125,
+      "learning_rate": 4.4011729884976955e-06,
+      "loss": 0.3276,
+      "step": 18420
+    },
+    {
+      "epoch": 0.8088960119433132,
+      "grad_norm": 1.46875,
+      "learning_rate": 4.397253476245908e-06,
+      "loss": 0.3036,
+      "step": 18422
+    },
+    {
+      "epoch": 0.8089838304225166,
+      "grad_norm": 1.484375,
+      "learning_rate": 4.393335541757387e-06,
+      "loss": 0.3055,
+      "step": 18424
+    },
+    {
+      "epoch": 0.8090716489017201,
+      "grad_norm": 1.515625,
+      "learning_rate": 4.389419185332167e-06,
+      "loss": 0.3363,
+      "step": 18426
+    },
+    {
+      "epoch": 0.8091594673809236,
+      "grad_norm": 1.5546875,
+      "learning_rate": 4.3855044072701715e-06,
+      "loss": 0.3146,
+      "step": 18428
+    },
+    {
+      "epoch": 0.8092472858601271,
+      "grad_norm": 1.375,
+      "learning_rate": 4.381591207871183e-06,
+      "loss": 0.311,
+      "step": 18430
+    },
+    {
+      "epoch": 0.8093351043393306,
+      "grad_norm": 1.53125,
+      "learning_rate": 4.377679587434888e-06,
+      "loss": 0.3498,
+      "step": 18432
+    },
+    {
+      "epoch": 0.8094229228185341,
+      "grad_norm": 1.5,
+      "learning_rate": 4.373769546260836e-06,
+      "loss": 0.3082,
+      "step": 18434
+    },
+    {
+      "epoch": 0.8095107412977376,
+      "grad_norm": 1.640625,
+      "learning_rate": 4.369861084648455e-06,
+      "loss": 0.3477,
+      "step": 18436
+    },
+    {
+      "epoch": 0.8095985597769411,
+      "grad_norm": 1.5390625,
+      "learning_rate": 4.365954202897058e-06,
+      "loss": 0.3237,
+      "step": 18438
+    },
+    {
+      "epoch": 0.8096863782561445,
+      "grad_norm": 1.5546875,
+      "learning_rate": 4.362048901305829e-06,
+      "loss": 0.3204,
+      "step": 18440
+    },
+    {
+      "epoch": 0.809774196735348,
+      "grad_norm": 1.4609375,
+      "learning_rate": 4.358145180173847e-06,
+      "loss": 0.3125,
+      "step": 18442
+    },
+    {
+      "epoch": 0.8098620152145515,
+      "grad_norm": 1.5,
+      "learning_rate": 4.354243039800049e-06,
+      "loss": 0.2819,
+      "step": 18444
+    },
+    {
+      "epoch": 0.8099498336937551,
+      "grad_norm": 1.4375,
+      "learning_rate": 4.350342480483277e-06,
+      "loss": 0.3345,
+      "step": 18446
+    },
+    {
+      "epoch": 0.8100376521729585,
+      "grad_norm": 1.515625,
+      "learning_rate": 4.346443502522226e-06,
+      "loss": 0.3361,
+      "step": 18448
+    },
+    {
+      "epoch": 0.810125470652162,
+      "grad_norm": 1.4609375,
+      "learning_rate": 4.3425461062154755e-06,
+      "loss": 0.3344,
+      "step": 18450
+    },
+    {
+      "epoch": 0.8102132891313655,
+      "grad_norm": 1.4453125,
+      "learning_rate": 4.338650291861504e-06,
+      "loss": 0.2927,
+      "step": 18452
+    },
+    {
+      "epoch": 0.810301107610569,
+      "grad_norm": 1.46875,
+      "learning_rate": 4.334756059758638e-06,
+      "loss": 0.3279,
+      "step": 18454
+    },
+    {
+      "epoch": 0.8103889260897724,
+      "grad_norm": 1.5234375,
+      "learning_rate": 4.330863410205116e-06,
+      "loss": 0.327,
+      "step": 18456
+    },
+    {
+      "epoch": 0.8104767445689759,
+      "grad_norm": 1.484375,
+      "learning_rate": 4.326972343499025e-06,
+      "loss": 0.3323,
+      "step": 18458
+    },
+    {
+      "epoch": 0.8105645630481794,
+      "grad_norm": 1.5,
+      "learning_rate": 4.323082859938343e-06,
+      "loss": 0.3349,
+      "step": 18460
+    },
+    {
+      "epoch": 0.810652381527383,
+      "grad_norm": 1.5078125,
+      "learning_rate": 4.319194959820941e-06,
+      "loss": 0.312,
+      "step": 18462
+    },
+    {
+      "epoch": 0.8107402000065864,
+      "grad_norm": 1.5625,
+      "learning_rate": 4.315308643444537e-06,
+      "loss": 0.3173,
+      "step": 18464
+    },
+    {
+      "epoch": 0.8108280184857899,
+      "grad_norm": 1.5,
+      "learning_rate": 4.3114239111067625e-06,
+      "loss": 0.3198,
+      "step": 18466
+    },
+    {
+      "epoch": 0.8109158369649934,
+      "grad_norm": 1.46875,
+      "learning_rate": 4.307540763105103e-06,
+      "loss": 0.3422,
+      "step": 18468
+    },
+    {
+      "epoch": 0.8110036554441968,
+      "grad_norm": 1.4609375,
+      "learning_rate": 4.303659199736934e-06,
+      "loss": 0.3171,
+      "step": 18470
+    },
+    {
+      "epoch": 0.8110914739234003,
+      "grad_norm": 1.5078125,
+      "learning_rate": 4.299779221299499e-06,
+      "loss": 0.3135,
+      "step": 18472
+    },
+    {
+      "epoch": 0.8111792924026038,
+      "grad_norm": 1.546875,
+      "learning_rate": 4.295900828089938e-06,
+      "loss": 0.3031,
+      "step": 18474
+    },
+    {
+      "epoch": 0.8112671108818073,
+      "grad_norm": 1.625,
+      "learning_rate": 4.292024020405255e-06,
+      "loss": 0.3092,
+      "step": 18476
+    },
+    {
+      "epoch": 0.8113549293610108,
+      "grad_norm": 1.59375,
+      "learning_rate": 4.288148798542332e-06,
+      "loss": 0.3243,
+      "step": 18478
+    },
+    {
+      "epoch": 0.8114427478402143,
+      "grad_norm": 1.59375,
+      "learning_rate": 4.284275162797943e-06,
+      "loss": 0.3145,
+      "step": 18480
+    },
+    {
+      "epoch": 0.8115305663194178,
+      "grad_norm": 1.484375,
+      "learning_rate": 4.2804031134687255e-06,
+      "loss": 0.2968,
+      "step": 18482
+    },
+    {
+      "epoch": 0.8116183847986213,
+      "grad_norm": 1.4609375,
+      "learning_rate": 4.276532650851206e-06,
+      "loss": 0.3296,
+      "step": 18484
+    },
+    {
+      "epoch": 0.8117062032778247,
+      "grad_norm": 1.484375,
+      "learning_rate": 4.272663775241787e-06,
+      "loss": 0.3132,
+      "step": 18486
+    },
+    {
+      "epoch": 0.8117940217570282,
+      "grad_norm": 1.5,
+      "learning_rate": 4.268796486936738e-06,
+      "loss": 0.3123,
+      "step": 18488
+    },
+    {
+      "epoch": 0.8118818402362317,
+      "grad_norm": 1.4921875,
+      "learning_rate": 4.264930786232227e-06,
+      "loss": 0.3005,
+      "step": 18490
+    },
+    {
+      "epoch": 0.8119696587154352,
+      "grad_norm": 1.5,
+      "learning_rate": 4.2610666734242825e-06,
+      "loss": 0.3299,
+      "step": 18492
+    },
+    {
+      "epoch": 0.8120574771946387,
+      "grad_norm": 1.5,
+      "learning_rate": 4.2572041488088325e-06,
+      "loss": 0.2917,
+      "step": 18494
+    },
+    {
+      "epoch": 0.8121452956738422,
+      "grad_norm": 1.5546875,
+      "learning_rate": 4.253343212681657e-06,
+      "loss": 0.335,
+      "step": 18496
+    },
+    {
+      "epoch": 0.8122331141530457,
+      "grad_norm": 1.5078125,
+      "learning_rate": 4.249483865338435e-06,
+      "loss": 0.324,
+      "step": 18498
+    },
+    {
+      "epoch": 0.8123209326322491,
+      "grad_norm": 1.546875,
+      "learning_rate": 4.24562610707471e-06,
+      "loss": 0.3325,
+      "step": 18500
+    },
+    {
+      "epoch": 0.8124087511114526,
+      "grad_norm": 1.4375,
+      "learning_rate": 4.241769938185907e-06,
+      "loss": 0.3297,
+      "step": 18502
+    },
+    {
+      "epoch": 0.8124965695906561,
+      "grad_norm": 1.4296875,
+      "learning_rate": 4.237915358967348e-06,
+      "loss": 0.3245,
+      "step": 18504
+    },
+    {
+      "epoch": 0.8125843880698596,
+      "grad_norm": 1.453125,
+      "learning_rate": 4.234062369714198e-06,
+      "loss": 0.2839,
+      "step": 18506
+    },
+    {
+      "epoch": 0.8126722065490631,
+      "grad_norm": 1.4609375,
+      "learning_rate": 4.230210970721538e-06,
+      "loss": 0.3228,
+      "step": 18508
+    },
+    {
+      "epoch": 0.8127600250282666,
+      "grad_norm": 1.4375,
+      "learning_rate": 4.226361162284298e-06,
+      "loss": 0.3114,
+      "step": 18510
+    },
+    {
+      "epoch": 0.8128478435074701,
+      "grad_norm": 1.5859375,
+      "learning_rate": 4.222512944697296e-06,
+      "loss": 0.3034,
+      "step": 18512
+    },
+    {
+      "epoch": 0.8129356619866736,
+      "grad_norm": 1.546875,
+      "learning_rate": 4.218666318255238e-06,
+      "loss": 0.2847,
+      "step": 18514
+    },
+    {
+      "epoch": 0.813023480465877,
+      "grad_norm": 1.453125,
+      "learning_rate": 4.21482128325269e-06,
+      "loss": 0.3017,
+      "step": 18516
+    },
+    {
+      "epoch": 0.8131112989450805,
+      "grad_norm": 1.6171875,
+      "learning_rate": 4.210977839984117e-06,
+      "loss": 0.3169,
+      "step": 18518
+    },
+    {
+      "epoch": 0.813199117424284,
+      "grad_norm": 1.5625,
+      "learning_rate": 4.207135988743844e-06,
+      "loss": 0.3233,
+      "step": 18520
+    },
+    {
+      "epoch": 0.8132869359034874,
+      "grad_norm": 1.5625,
+      "learning_rate": 4.203295729826076e-06,
+      "loss": 0.3397,
+      "step": 18522
+    },
+    {
+      "epoch": 0.813374754382691,
+      "grad_norm": 1.453125,
+      "learning_rate": 4.199457063524911e-06,
+      "loss": 0.3195,
+      "step": 18524
+    },
+    {
+      "epoch": 0.8134625728618945,
+      "grad_norm": 1.53125,
+      "learning_rate": 4.1956199901343055e-06,
+      "loss": 0.3229,
+      "step": 18526
+    },
+    {
+      "epoch": 0.813550391341098,
+      "grad_norm": 1.53125,
+      "learning_rate": 4.191784509948121e-06,
+      "loss": 0.2953,
+      "step": 18528
+    },
+    {
+      "epoch": 0.8136382098203014,
+      "grad_norm": 1.515625,
+      "learning_rate": 4.187950623260053e-06,
+      "loss": 0.3143,
+      "step": 18530
+    },
+    {
+      "epoch": 0.8137260282995049,
+      "grad_norm": 1.4921875,
+      "learning_rate": 4.184118330363721e-06,
+      "loss": 0.3555,
+      "step": 18532
+    },
+    {
+      "epoch": 0.8138138467787084,
+      "grad_norm": 1.4296875,
+      "learning_rate": 4.180287631552593e-06,
+      "loss": 0.3137,
+      "step": 18534
+    },
+    {
+      "epoch": 0.8139016652579119,
+      "grad_norm": 1.46875,
+      "learning_rate": 4.176458527120034e-06,
+      "loss": 0.3324,
+      "step": 18536
+    },
+    {
+      "epoch": 0.8139894837371154,
+      "grad_norm": 1.5546875,
+      "learning_rate": 4.172631017359274e-06,
+      "loss": 0.3093,
+      "step": 18538
+    },
+    {
+      "epoch": 0.8140773022163189,
+      "grad_norm": 1.5390625,
+      "learning_rate": 4.168805102563414e-06,
+      "loss": 0.3012,
+      "step": 18540
+    },
+    {
+      "epoch": 0.8141651206955224,
+      "grad_norm": 1.484375,
+      "learning_rate": 4.164980783025463e-06,
+      "loss": 0.2968,
+      "step": 18542
+    },
+    {
+      "epoch": 0.8142529391747259,
+      "grad_norm": 1.5234375,
+      "learning_rate": 4.1611580590382695e-06,
+      "loss": 0.3131,
+      "step": 18544
+    },
+    {
+      "epoch": 0.8143407576539293,
+      "grad_norm": 1.4453125,
+      "learning_rate": 4.157336930894593e-06,
+      "loss": 0.3175,
+      "step": 18546
+    },
+    {
+      "epoch": 0.8144285761331328,
+      "grad_norm": 1.484375,
+      "learning_rate": 4.153517398887053e-06,
+      "loss": 0.2955,
+      "step": 18548
+    },
+    {
+      "epoch": 0.8145163946123363,
+      "grad_norm": 1.5,
+      "learning_rate": 4.14969946330814e-06,
+      "loss": 0.3392,
+      "step": 18550
+    },
+    {
+      "epoch": 0.8146042130915397,
+      "grad_norm": 1.5703125,
+      "learning_rate": 4.145883124450245e-06,
+      "loss": 0.3102,
+      "step": 18552
+    },
+    {
+      "epoch": 0.8146920315707433,
+      "grad_norm": 1.4453125,
+      "learning_rate": 4.142068382605615e-06,
+      "loss": 0.3169,
+      "step": 18554
+    },
+    {
+      "epoch": 0.8147798500499468,
+      "grad_norm": 1.5,
+      "learning_rate": 4.138255238066397e-06,
+      "loss": 0.3488,
+      "step": 18556
+    },
+    {
+      "epoch": 0.8148676685291503,
+      "grad_norm": 1.6015625,
+      "learning_rate": 4.13444369112459e-06,
+      "loss": 0.3287,
+      "step": 18558
+    },
+    {
+      "epoch": 0.8149554870083537,
+      "grad_norm": 1.40625,
+      "learning_rate": 4.130633742072087e-06,
+      "loss": 0.3307,
+      "step": 18560
+    },
+    {
+      "epoch": 0.8150433054875572,
+      "grad_norm": 1.5703125,
+      "learning_rate": 4.126825391200656e-06,
+      "loss": 0.3352,
+      "step": 18562
+    },
+    {
+      "epoch": 0.8151311239667607,
+      "grad_norm": 1.5234375,
+      "learning_rate": 4.123018638801935e-06,
+      "loss": 0.3291,
+      "step": 18564
+    },
+    {
+      "epoch": 0.8152189424459642,
+      "grad_norm": 1.4609375,
+      "learning_rate": 4.119213485167456e-06,
+      "loss": 0.3104,
+      "step": 18566
+    },
+    {
+      "epoch": 0.8153067609251676,
+      "grad_norm": 1.4765625,
+      "learning_rate": 4.115409930588606e-06,
+      "loss": 0.312,
+      "step": 18568
+    },
+    {
+      "epoch": 0.8153945794043712,
+      "grad_norm": 1.5390625,
+      "learning_rate": 4.111607975356679e-06,
+      "loss": 0.3169,
+      "step": 18570
+    },
+    {
+      "epoch": 0.8154823978835747,
+      "grad_norm": 1.5625,
+      "learning_rate": 4.1078076197628214e-06,
+      "loss": 0.3016,
+      "step": 18572
+    },
+    {
+      "epoch": 0.8155702163627782,
+      "grad_norm": 1.4609375,
+      "learning_rate": 4.104008864098055e-06,
+      "loss": 0.3218,
+      "step": 18574
+    },
+    {
+      "epoch": 0.8156580348419816,
+      "grad_norm": 1.4765625,
+      "learning_rate": 4.100211708653306e-06,
+      "loss": 0.305,
+      "step": 18576
+    },
+    {
+      "epoch": 0.8157458533211851,
+      "grad_norm": 1.5,
+      "learning_rate": 4.0964161537193486e-06,
+      "loss": 0.3193,
+      "step": 18578
+    },
+    {
+      "epoch": 0.8158336718003886,
+      "grad_norm": 1.4609375,
+      "learning_rate": 4.092622199586859e-06,
+      "loss": 0.3265,
+      "step": 18580
+    },
+    {
+      "epoch": 0.815921490279592,
+      "grad_norm": 1.5390625,
+      "learning_rate": 4.088829846546374e-06,
+      "loss": 0.2984,
+      "step": 18582
+    },
+    {
+      "epoch": 0.8160093087587955,
+      "grad_norm": 1.6015625,
+      "learning_rate": 4.085039094888307e-06,
+      "loss": 0.3128,
+      "step": 18584
+    },
+    {
+      "epoch": 0.8160971272379991,
+      "grad_norm": 1.5078125,
+      "learning_rate": 4.0812499449029624e-06,
+      "loss": 0.3125,
+      "step": 18586
+    },
+    {
+      "epoch": 0.8161849457172026,
+      "grad_norm": 1.46875,
+      "learning_rate": 4.077462396880508e-06,
+      "loss": 0.3213,
+      "step": 18588
+    },
+    {
+      "epoch": 0.816272764196406,
+      "grad_norm": 1.4375,
+      "learning_rate": 4.073676451111011e-06,
+      "loss": 0.2875,
+      "step": 18590
+    },
+    {
+      "epoch": 0.8163605826756095,
+      "grad_norm": 1.4609375,
+      "learning_rate": 4.069892107884374e-06,
+      "loss": 0.3203,
+      "step": 18592
+    },
+    {
+      "epoch": 0.816448401154813,
+      "grad_norm": 1.4921875,
+      "learning_rate": 4.066109367490426e-06,
+      "loss": 0.3497,
+      "step": 18594
+    },
+    {
+      "epoch": 0.8165362196340165,
+      "grad_norm": 1.4921875,
+      "learning_rate": 4.062328230218831e-06,
+      "loss": 0.2914,
+      "step": 18596
+    },
+    {
+      "epoch": 0.8166240381132199,
+      "grad_norm": 1.640625,
+      "learning_rate": 4.0585486963591655e-06,
+      "loss": 0.3261,
+      "step": 18598
+    },
+    {
+      "epoch": 0.8167118565924235,
+      "grad_norm": 1.46875,
+      "learning_rate": 4.0547707662008634e-06,
+      "loss": 0.3163,
+      "step": 18600
+    },
+    {
+      "epoch": 0.816799675071627,
+      "grad_norm": 1.5,
+      "learning_rate": 4.050994440033229e-06,
+      "loss": 0.3222,
+      "step": 18602
+    },
+    {
+      "epoch": 0.8168874935508305,
+      "grad_norm": 1.5234375,
+      "learning_rate": 4.04721971814547e-06,
+      "loss": 0.3136,
+      "step": 18604
+    },
+    {
+      "epoch": 0.8169753120300339,
+      "grad_norm": 1.4921875,
+      "learning_rate": 4.0434466008266395e-06,
+      "loss": 0.3103,
+      "step": 18606
+    },
+    {
+      "epoch": 0.8170631305092374,
+      "grad_norm": 1.4296875,
+      "learning_rate": 4.0396750883657e-06,
+      "loss": 0.3185,
+      "step": 18608
+    },
+    {
+      "epoch": 0.8171509489884409,
+      "grad_norm": 1.5,
+      "learning_rate": 4.035905181051464e-06,
+      "loss": 0.3172,
+      "step": 18610
+    },
+    {
+      "epoch": 0.8172387674676443,
+      "grad_norm": 1.515625,
+      "learning_rate": 4.0321368791726325e-06,
+      "loss": 0.3128,
+      "step": 18612
+    },
+    {
+      "epoch": 0.8173265859468478,
+      "grad_norm": 1.3984375,
+      "learning_rate": 4.028370183017788e-06,
+      "loss": 0.3079,
+      "step": 18614
+    },
+    {
+      "epoch": 0.8174144044260514,
+      "grad_norm": 1.578125,
+      "learning_rate": 4.024605092875378e-06,
+      "loss": 0.3365,
+      "step": 18616
+    },
+    {
+      "epoch": 0.8175022229052549,
+      "grad_norm": 1.46875,
+      "learning_rate": 4.020841609033743e-06,
+      "loss": 0.3228,
+      "step": 18618
+    },
+    {
+      "epoch": 0.8175900413844583,
+      "grad_norm": 1.4296875,
+      "learning_rate": 4.0170797317810875e-06,
+      "loss": 0.3432,
+      "step": 18620
+    },
+    {
+      "epoch": 0.8176778598636618,
+      "grad_norm": 1.46875,
+      "learning_rate": 4.013319461405493e-06,
+      "loss": 0.3294,
+      "step": 18622
+    },
+    {
+      "epoch": 0.8177656783428653,
+      "grad_norm": 1.4296875,
+      "learning_rate": 4.009560798194928e-06,
+      "loss": 0.3218,
+      "step": 18624
+    },
+    {
+      "epoch": 0.8178534968220688,
+      "grad_norm": 1.4375,
+      "learning_rate": 4.005803742437222e-06,
+      "loss": 0.3066,
+      "step": 18626
+    },
+    {
+      "epoch": 0.8179413153012722,
+      "grad_norm": 1.5234375,
+      "learning_rate": 4.002048294420105e-06,
+      "loss": 0.3282,
+      "step": 18628
+    },
+    {
+      "epoch": 0.8180291337804757,
+      "grad_norm": 1.4453125,
+      "learning_rate": 3.998294454431157e-06,
+      "loss": 0.3375,
+      "step": 18630
+    },
+    {
+      "epoch": 0.8181169522596793,
+      "grad_norm": 1.5625,
+      "learning_rate": 3.99454222275786e-06,
+      "loss": 0.3129,
+      "step": 18632
+    },
+    {
+      "epoch": 0.8182047707388828,
+      "grad_norm": 1.3828125,
+      "learning_rate": 3.990791599687554e-06,
+      "loss": 0.332,
+      "step": 18634
+    },
+    {
+      "epoch": 0.8182925892180862,
+      "grad_norm": 1.578125,
+      "learning_rate": 3.987042585507458e-06,
+      "loss": 0.325,
+      "step": 18636
+    },
+    {
+      "epoch": 0.8183804076972897,
+      "grad_norm": 1.5625,
+      "learning_rate": 3.983295180504685e-06,
+      "loss": 0.2998,
+      "step": 18638
+    },
+    {
+      "epoch": 0.8184682261764932,
+      "grad_norm": 1.4921875,
+      "learning_rate": 3.979549384966197e-06,
+      "loss": 0.2963,
+      "step": 18640
+    },
+    {
+      "epoch": 0.8185560446556966,
+      "grad_norm": 1.4765625,
+      "learning_rate": 3.975805199178865e-06,
+      "loss": 0.3278,
+      "step": 18642
+    },
+    {
+      "epoch": 0.8186438631349001,
+      "grad_norm": 1.4765625,
+      "learning_rate": 3.972062623429409e-06,
+      "loss": 0.3575,
+      "step": 18644
+    },
+    {
+      "epoch": 0.8187316816141037,
+      "grad_norm": 1.453125,
+      "learning_rate": 3.968321658004431e-06,
+      "loss": 0.3186,
+      "step": 18646
+    },
+    {
+      "epoch": 0.8188195000933072,
+      "grad_norm": 1.5859375,
+      "learning_rate": 3.964582303190428e-06,
+      "loss": 0.3378,
+      "step": 18648
+    },
+    {
+      "epoch": 0.8189073185725106,
+      "grad_norm": 1.578125,
+      "learning_rate": 3.9608445592737575e-06,
+      "loss": 0.311,
+      "step": 18650
+    },
+    {
+      "epoch": 0.8189951370517141,
+      "grad_norm": 1.515625,
+      "learning_rate": 3.957108426540654e-06,
+      "loss": 0.2803,
+      "step": 18652
+    },
+    {
+      "epoch": 0.8190829555309176,
+      "grad_norm": 1.4453125,
+      "learning_rate": 3.953373905277222e-06,
+      "loss": 0.2866,
+      "step": 18654
+    },
+    {
+      "epoch": 0.8191707740101211,
+      "grad_norm": 1.5,
+      "learning_rate": 3.949640995769471e-06,
+      "loss": 0.3107,
+      "step": 18656
+    },
+    {
+      "epoch": 0.8192585924893245,
+      "grad_norm": 1.4375,
+      "learning_rate": 3.945909698303249e-06,
+      "loss": 0.321,
+      "step": 18658
+    },
+    {
+      "epoch": 0.819346410968528,
+      "grad_norm": 1.5625,
+      "learning_rate": 3.942180013164318e-06,
+      "loss": 0.305,
+      "step": 18660
+    },
+    {
+      "epoch": 0.8194342294477316,
+      "grad_norm": 1.4453125,
+      "learning_rate": 3.938451940638291e-06,
+      "loss": 0.3157,
+      "step": 18662
+    },
+    {
+      "epoch": 0.8195220479269351,
+      "grad_norm": 1.5,
+      "learning_rate": 3.934725481010653e-06,
+      "loss": 0.3003,
+      "step": 18664
+    },
+    {
+      "epoch": 0.8196098664061385,
+      "grad_norm": 1.4921875,
+      "learning_rate": 3.931000634566798e-06,
+      "loss": 0.343,
+      "step": 18666
+    },
+    {
+      "epoch": 0.819697684885342,
+      "grad_norm": 1.4921875,
+      "learning_rate": 3.927277401591956e-06,
+      "loss": 0.3026,
+      "step": 18668
+    },
+    {
+      "epoch": 0.8197855033645455,
+      "grad_norm": 1.5,
+      "learning_rate": 3.923555782371269e-06,
+      "loss": 0.3104,
+      "step": 18670
+    },
+    {
+      "epoch": 0.819873321843749,
+      "grad_norm": 1.5703125,
+      "learning_rate": 3.919835777189732e-06,
+      "loss": 0.3062,
+      "step": 18672
+    },
+    {
+      "epoch": 0.8199611403229524,
+      "grad_norm": 1.4921875,
+      "learning_rate": 3.916117386332219e-06,
+      "loss": 0.31,
+      "step": 18674
+    },
+    {
+      "epoch": 0.8200489588021559,
+      "grad_norm": 1.5390625,
+      "learning_rate": 3.912400610083494e-06,
+      "loss": 0.2994,
+      "step": 18676
+    },
+    {
+      "epoch": 0.8201367772813595,
+      "grad_norm": 1.5703125,
+      "learning_rate": 3.908685448728183e-06,
+      "loss": 0.3164,
+      "step": 18678
+    },
+    {
+      "epoch": 0.820224595760563,
+      "grad_norm": 1.5,
+      "learning_rate": 3.9049719025508e-06,
+      "loss": 0.3125,
+      "step": 18680
+    },
+    {
+      "epoch": 0.8203124142397664,
+      "grad_norm": 1.5234375,
+      "learning_rate": 3.901259971835728e-06,
+      "loss": 0.3175,
+      "step": 18682
+    },
+    {
+      "epoch": 0.8204002327189699,
+      "grad_norm": 1.5234375,
+      "learning_rate": 3.897549656867222e-06,
+      "loss": 0.3205,
+      "step": 18684
+    },
+    {
+      "epoch": 0.8204880511981734,
+      "grad_norm": 1.484375,
+      "learning_rate": 3.893840957929423e-06,
+      "loss": 0.3137,
+      "step": 18686
+    },
+    {
+      "epoch": 0.8205758696773768,
+      "grad_norm": 1.578125,
+      "learning_rate": 3.8901338753063375e-06,
+      "loss": 0.3512,
+      "step": 18688
+    },
+    {
+      "epoch": 0.8206636881565803,
+      "grad_norm": 1.4609375,
+      "learning_rate": 3.886428409281867e-06,
+      "loss": 0.3016,
+      "step": 18690
+    },
+    {
+      "epoch": 0.8207515066357839,
+      "grad_norm": 1.578125,
+      "learning_rate": 3.882724560139764e-06,
+      "loss": 0.3361,
+      "step": 18692
+    },
+    {
+      "epoch": 0.8208393251149874,
+      "grad_norm": 1.5,
+      "learning_rate": 3.879022328163681e-06,
+      "loss": 0.3345,
+      "step": 18694
+    },
+    {
+      "epoch": 0.8209271435941908,
+      "grad_norm": 1.6328125,
+      "learning_rate": 3.875321713637131e-06,
+      "loss": 0.3463,
+      "step": 18696
+    },
+    {
+      "epoch": 0.8210149620733943,
+      "grad_norm": 1.5703125,
+      "learning_rate": 3.8716227168435035e-06,
+      "loss": 0.3222,
+      "step": 18698
+    },
+    {
+      "epoch": 0.8211027805525978,
+      "grad_norm": 1.5,
+      "learning_rate": 3.867925338066078e-06,
+      "loss": 0.3165,
+      "step": 18700
+    },
+    {
+      "epoch": 0.8211905990318012,
+      "grad_norm": 1.53125,
+      "learning_rate": 3.864229577587991e-06,
+      "loss": 0.2992,
+      "step": 18702
+    },
+    {
+      "epoch": 0.8212784175110047,
+      "grad_norm": 1.5234375,
+      "learning_rate": 3.860535435692275e-06,
+      "loss": 0.3192,
+      "step": 18704
+    },
+    {
+      "epoch": 0.8213662359902082,
+      "grad_norm": 1.5078125,
+      "learning_rate": 3.856842912661823e-06,
+      "loss": 0.3102,
+      "step": 18706
+    },
+    {
+      "epoch": 0.8214540544694118,
+      "grad_norm": 1.4609375,
+      "learning_rate": 3.853152008779401e-06,
+      "loss": 0.3308,
+      "step": 18708
+    },
+    {
+      "epoch": 0.8215418729486152,
+      "grad_norm": 1.453125,
+      "learning_rate": 3.8494627243276764e-06,
+      "loss": 0.3053,
+      "step": 18710
+    },
+    {
+      "epoch": 0.8216296914278187,
+      "grad_norm": 1.4453125,
+      "learning_rate": 3.8457750595891656e-06,
+      "loss": 0.3065,
+      "step": 18712
+    },
+    {
+      "epoch": 0.8217175099070222,
+      "grad_norm": 1.4609375,
+      "learning_rate": 3.84208901484627e-06,
+      "loss": 0.3132,
+      "step": 18714
+    },
+    {
+      "epoch": 0.8218053283862257,
+      "grad_norm": 1.53125,
+      "learning_rate": 3.838404590381267e-06,
+      "loss": 0.3347,
+      "step": 18716
+    },
+    {
+      "epoch": 0.8218931468654291,
+      "grad_norm": 1.46875,
+      "learning_rate": 3.834721786476317e-06,
+      "loss": 0.2712,
+      "step": 18718
+    },
+    {
+      "epoch": 0.8219809653446326,
+      "grad_norm": 1.4921875,
+      "learning_rate": 3.831040603413441e-06,
+      "loss": 0.3004,
+      "step": 18720
+    },
+    {
+      "epoch": 0.8220687838238361,
+      "grad_norm": 1.5078125,
+      "learning_rate": 3.827361041474556e-06,
+      "loss": 0.3115,
+      "step": 18722
+    },
+    {
+      "epoch": 0.8221566023030397,
+      "grad_norm": 1.46875,
+      "learning_rate": 3.823683100941436e-06,
+      "loss": 0.3081,
+      "step": 18724
+    },
+    {
+      "epoch": 0.8222444207822431,
+      "grad_norm": 1.5,
+      "learning_rate": 3.820006782095736e-06,
+      "loss": 0.3414,
+      "step": 18726
+    },
+    {
+      "epoch": 0.8223322392614466,
+      "grad_norm": 1.4921875,
+      "learning_rate": 3.816332085218999e-06,
+      "loss": 0.3236,
+      "step": 18728
+    },
+    {
+      "epoch": 0.8224200577406501,
+      "grad_norm": 1.4921875,
+      "learning_rate": 3.812659010592626e-06,
+      "loss": 0.3281,
+      "step": 18730
+    },
+    {
+      "epoch": 0.8225078762198536,
+      "grad_norm": 1.4453125,
+      "learning_rate": 3.808987558497906e-06,
+      "loss": 0.302,
+      "step": 18732
+    },
+    {
+      "epoch": 0.822595694699057,
+      "grad_norm": 1.4765625,
+      "learning_rate": 3.8053177292160015e-06,
+      "loss": 0.3059,
+      "step": 18734
+    },
+    {
+      "epoch": 0.8226835131782605,
+      "grad_norm": 1.4765625,
+      "learning_rate": 3.801649523027942e-06,
+      "loss": 0.325,
+      "step": 18736
+    },
+    {
+      "epoch": 0.822771331657464,
+      "grad_norm": 1.5703125,
+      "learning_rate": 3.7979829402146477e-06,
+      "loss": 0.3035,
+      "step": 18738
+    },
+    {
+      "epoch": 0.8228591501366675,
+      "grad_norm": 1.46875,
+      "learning_rate": 3.794317981056894e-06,
+      "loss": 0.3197,
+      "step": 18740
+    },
+    {
+      "epoch": 0.822946968615871,
+      "grad_norm": 1.4765625,
+      "learning_rate": 3.7906546458353677e-06,
+      "loss": 0.3018,
+      "step": 18742
+    },
+    {
+      "epoch": 0.8230347870950745,
+      "grad_norm": 1.4375,
+      "learning_rate": 3.78699293483058e-06,
+      "loss": 0.3275,
+      "step": 18744
+    },
+    {
+      "epoch": 0.823122605574278,
+      "grad_norm": 1.484375,
+      "learning_rate": 3.783332848322965e-06,
+      "loss": 0.3183,
+      "step": 18746
+    },
+    {
+      "epoch": 0.8232104240534814,
+      "grad_norm": 1.484375,
+      "learning_rate": 3.7796743865928045e-06,
+      "loss": 0.3437,
+      "step": 18748
+    },
+    {
+      "epoch": 0.8232982425326849,
+      "grad_norm": 1.3984375,
+      "learning_rate": 3.776017549920263e-06,
+      "loss": 0.3303,
+      "step": 18750
+    },
+    {
+      "epoch": 0.8233860610118884,
+      "grad_norm": 1.453125,
+      "learning_rate": 3.772362338585389e-06,
+      "loss": 0.3062,
+      "step": 18752
+    },
+    {
+      "epoch": 0.823473879491092,
+      "grad_norm": 1.4609375,
+      "learning_rate": 3.7687087528680915e-06,
+      "loss": 0.3194,
+      "step": 18754
+    },
+    {
+      "epoch": 0.8235616979702954,
+      "grad_norm": 1.546875,
+      "learning_rate": 3.7650567930481716e-06,
+      "loss": 0.3231,
+      "step": 18756
+    },
+    {
+      "epoch": 0.8236495164494989,
+      "grad_norm": 1.5,
+      "learning_rate": 3.761406459405292e-06,
+      "loss": 0.3186,
+      "step": 18758
+    },
+    {
+      "epoch": 0.8237373349287024,
+      "grad_norm": 1.453125,
+      "learning_rate": 3.7577577522189935e-06,
+      "loss": 0.3101,
+      "step": 18760
+    },
+    {
+      "epoch": 0.8238251534079059,
+      "grad_norm": 1.5703125,
+      "learning_rate": 3.754110671768704e-06,
+      "loss": 0.313,
+      "step": 18762
+    },
+    {
+      "epoch": 0.8239129718871093,
+      "grad_norm": 1.4140625,
+      "learning_rate": 3.750465218333704e-06,
+      "loss": 0.3161,
+      "step": 18764
+    },
+    {
+      "epoch": 0.8240007903663128,
+      "grad_norm": 1.546875,
+      "learning_rate": 3.746821392193181e-06,
+      "loss": 0.3347,
+      "step": 18766
+    },
+    {
+      "epoch": 0.8240886088455163,
+      "grad_norm": 1.484375,
+      "learning_rate": 3.7431791936261672e-06,
+      "loss": 0.3277,
+      "step": 18768
+    },
+    {
+      "epoch": 0.8241764273247199,
+      "grad_norm": 1.5546875,
+      "learning_rate": 3.739538622911584e-06,
+      "loss": 0.3013,
+      "step": 18770
+    },
+    {
+      "epoch": 0.8242642458039233,
+      "grad_norm": 1.5,
+      "learning_rate": 3.7358996803282335e-06,
+      "loss": 0.3254,
+      "step": 18772
+    },
+    {
+      "epoch": 0.8243520642831268,
+      "grad_norm": 1.4296875,
+      "learning_rate": 3.7322623661547816e-06,
+      "loss": 0.3235,
+      "step": 18774
+    },
+    {
+      "epoch": 0.8244398827623303,
+      "grad_norm": 1.4375,
+      "learning_rate": 3.7286266806697777e-06,
+      "loss": 0.3175,
+      "step": 18776
+    },
+    {
+      "epoch": 0.8245277012415337,
+      "grad_norm": 1.4375,
+      "learning_rate": 3.7249926241516353e-06,
+      "loss": 0.2844,
+      "step": 18778
+    },
+    {
+      "epoch": 0.8246155197207372,
+      "grad_norm": 1.5234375,
+      "learning_rate": 3.721360196878662e-06,
+      "loss": 0.3042,
+      "step": 18780
+    },
+    {
+      "epoch": 0.8247033381999407,
+      "grad_norm": 1.4765625,
+      "learning_rate": 3.7177293991290273e-06,
+      "loss": 0.3065,
+      "step": 18782
+    },
+    {
+      "epoch": 0.8247911566791442,
+      "grad_norm": 1.5078125,
+      "learning_rate": 3.7141002311807698e-06,
+      "loss": 0.2974,
+      "step": 18784
+    },
+    {
+      "epoch": 0.8248789751583477,
+      "grad_norm": 1.4375,
+      "learning_rate": 3.7104726933118282e-06,
+      "loss": 0.3056,
+      "step": 18786
+    },
+    {
+      "epoch": 0.8249667936375512,
+      "grad_norm": 1.421875,
+      "learning_rate": 3.7068467857999832e-06,
+      "loss": 0.3465,
+      "step": 18788
+    },
+    {
+      "epoch": 0.8250546121167547,
+      "grad_norm": 1.515625,
+      "learning_rate": 3.703222508922921e-06,
+      "loss": 0.3242,
+      "step": 18790
+    },
+    {
+      "epoch": 0.8251424305959582,
+      "grad_norm": 1.5625,
+      "learning_rate": 3.699599862958178e-06,
+      "loss": 0.3172,
+      "step": 18792
+    },
+    {
+      "epoch": 0.8252302490751616,
+      "grad_norm": 1.6484375,
+      "learning_rate": 3.69597884818319e-06,
+      "loss": 0.3146,
+      "step": 18794
+    },
+    {
+      "epoch": 0.8253180675543651,
+      "grad_norm": 1.4296875,
+      "learning_rate": 3.69235946487525e-06,
+      "loss": 0.3353,
+      "step": 18796
+    },
+    {
+      "epoch": 0.8254058860335686,
+      "grad_norm": 1.4609375,
+      "learning_rate": 3.688741713311522e-06,
+      "loss": 0.3312,
+      "step": 18798
+    },
+    {
+      "epoch": 0.8254937045127722,
+      "grad_norm": 1.453125,
+      "learning_rate": 3.6851255937690704e-06,
+      "loss": 0.3205,
+      "step": 18800
+    },
+    {
+      "epoch": 0.8255815229919756,
+      "grad_norm": 1.4375,
+      "learning_rate": 3.6815111065248043e-06,
+      "loss": 0.2822,
+      "step": 18802
+    },
+    {
+      "epoch": 0.8256693414711791,
+      "grad_norm": 1.671875,
+      "learning_rate": 3.677898251855538e-06,
+      "loss": 0.3132,
+      "step": 18804
+    },
+    {
+      "epoch": 0.8257571599503826,
+      "grad_norm": 1.53125,
+      "learning_rate": 3.6742870300379232e-06,
+      "loss": 0.3054,
+      "step": 18806
+    },
+    {
+      "epoch": 0.825844978429586,
+      "grad_norm": 1.4453125,
+      "learning_rate": 3.6706774413485275e-06,
+      "loss": 0.2958,
+      "step": 18808
+    },
+    {
+      "epoch": 0.8259327969087895,
+      "grad_norm": 1.5078125,
+      "learning_rate": 3.667069486063765e-06,
+      "loss": 0.3208,
+      "step": 18810
+    },
+    {
+      "epoch": 0.826020615387993,
+      "grad_norm": 1.515625,
+      "learning_rate": 3.6634631644599295e-06,
+      "loss": 0.3191,
+      "step": 18812
+    },
+    {
+      "epoch": 0.8261084338671965,
+      "grad_norm": 1.5,
+      "learning_rate": 3.659858476813205e-06,
+      "loss": 0.3186,
+      "step": 18814
+    },
+    {
+      "epoch": 0.8261962523464,
+      "grad_norm": 1.46875,
+      "learning_rate": 3.6562554233996295e-06,
+      "loss": 0.3111,
+      "step": 18816
+    },
+    {
+      "epoch": 0.8262840708256035,
+      "grad_norm": 1.4453125,
+      "learning_rate": 3.6526540044951346e-06,
+      "loss": 0.3027,
+      "step": 18818
+    },
+    {
+      "epoch": 0.826371889304807,
+      "grad_norm": 1.4140625,
+      "learning_rate": 3.649054220375514e-06,
+      "loss": 0.2807,
+      "step": 18820
+    },
+    {
+      "epoch": 0.8264597077840105,
+      "grad_norm": 1.5,
+      "learning_rate": 3.6454560713164334e-06,
+      "loss": 0.3191,
+      "step": 18822
+    },
+    {
+      "epoch": 0.8265475262632139,
+      "grad_norm": 1.484375,
+      "learning_rate": 3.6418595575934524e-06,
+      "loss": 0.3282,
+      "step": 18824
+    },
+    {
+      "epoch": 0.8266353447424174,
+      "grad_norm": 1.4609375,
+      "learning_rate": 3.6382646794819785e-06,
+      "loss": 0.3097,
+      "step": 18826
+    },
+    {
+      "epoch": 0.8267231632216209,
+      "grad_norm": 1.578125,
+      "learning_rate": 3.6346714372573224e-06,
+      "loss": 0.3229,
+      "step": 18828
+    },
+    {
+      "epoch": 0.8268109817008243,
+      "grad_norm": 1.46875,
+      "learning_rate": 3.6310798311946503e-06,
+      "loss": 0.3392,
+      "step": 18830
+    },
+    {
+      "epoch": 0.8268988001800279,
+      "grad_norm": 1.421875,
+      "learning_rate": 3.627489861569003e-06,
+      "loss": 0.2793,
+      "step": 18832
+    },
+    {
+      "epoch": 0.8269866186592314,
+      "grad_norm": 1.484375,
+      "learning_rate": 3.623901528655313e-06,
+      "loss": 0.3158,
+      "step": 18834
+    },
+    {
+      "epoch": 0.8270744371384349,
+      "grad_norm": 1.4375,
+      "learning_rate": 3.620314832728361e-06,
+      "loss": 0.3022,
+      "step": 18836
+    },
+    {
+      "epoch": 0.8271622556176383,
+      "grad_norm": 1.515625,
+      "learning_rate": 3.616729774062827e-06,
+      "loss": 0.2894,
+      "step": 18838
+    },
+    {
+      "epoch": 0.8272500740968418,
+      "grad_norm": 1.53125,
+      "learning_rate": 3.613146352933247e-06,
+      "loss": 0.2916,
+      "step": 18840
+    },
+    {
+      "epoch": 0.8273378925760453,
+      "grad_norm": 1.3828125,
+      "learning_rate": 3.6095645696140547e-06,
+      "loss": 0.3082,
+      "step": 18842
+    },
+    {
+      "epoch": 0.8274257110552488,
+      "grad_norm": 1.3984375,
+      "learning_rate": 3.6059844243795327e-06,
+      "loss": 0.3201,
+      "step": 18844
+    },
+    {
+      "epoch": 0.8275135295344523,
+      "grad_norm": 1.765625,
+      "learning_rate": 3.6024059175038455e-06,
+      "loss": 0.293,
+      "step": 18846
+    },
+    {
+      "epoch": 0.8276013480136558,
+      "grad_norm": 1.4453125,
+      "learning_rate": 3.5988290492610488e-06,
+      "loss": 0.3128,
+      "step": 18848
+    },
+    {
+      "epoch": 0.8276891664928593,
+      "grad_norm": 1.40625,
+      "learning_rate": 3.5952538199250515e-06,
+      "loss": 0.2913,
+      "step": 18850
+    },
+    {
+      "epoch": 0.8277769849720628,
+      "grad_norm": 1.46875,
+      "learning_rate": 3.5916802297696506e-06,
+      "loss": 0.3064,
+      "step": 18852
+    },
+    {
+      "epoch": 0.8278648034512662,
+      "grad_norm": 1.546875,
+      "learning_rate": 3.5881082790685026e-06,
+      "loss": 0.322,
+      "step": 18854
+    },
+    {
+      "epoch": 0.8279526219304697,
+      "grad_norm": 1.4765625,
+      "learning_rate": 3.5845379680951614e-06,
+      "loss": 0.3225,
+      "step": 18856
+    },
+    {
+      "epoch": 0.8280404404096732,
+      "grad_norm": 1.5390625,
+      "learning_rate": 3.580969297123038e-06,
+      "loss": 0.3068,
+      "step": 18858
+    },
+    {
+      "epoch": 0.8281282588888766,
+      "grad_norm": 1.7578125,
+      "learning_rate": 3.577402266425414e-06,
+      "loss": 0.3302,
+      "step": 18860
+    },
+    {
+      "epoch": 0.8282160773680802,
+      "grad_norm": 1.5078125,
+      "learning_rate": 3.573836876275466e-06,
+      "loss": 0.3012,
+      "step": 18862
+    },
+    {
+      "epoch": 0.8283038958472837,
+      "grad_norm": 1.5,
+      "learning_rate": 3.5702731269462193e-06,
+      "loss": 0.338,
+      "step": 18864
+    },
+    {
+      "epoch": 0.8283917143264872,
+      "grad_norm": 1.4609375,
+      "learning_rate": 3.5667110187106056e-06,
+      "loss": 0.3062,
+      "step": 18866
+    },
+    {
+      "epoch": 0.8284795328056906,
+      "grad_norm": 1.4140625,
+      "learning_rate": 3.5631505518413904e-06,
+      "loss": 0.2983,
+      "step": 18868
+    },
+    {
+      "epoch": 0.8285673512848941,
+      "grad_norm": 1.5234375,
+      "learning_rate": 3.5595917266112474e-06,
+      "loss": 0.309,
+      "step": 18870
+    },
+    {
+      "epoch": 0.8286551697640976,
+      "grad_norm": 1.546875,
+      "learning_rate": 3.5560345432927104e-06,
+      "loss": 0.3137,
+      "step": 18872
+    },
+    {
+      "epoch": 0.8287429882433011,
+      "grad_norm": 1.4453125,
+      "learning_rate": 3.552479002158185e-06,
+      "loss": 0.3057,
+      "step": 18874
+    },
+    {
+      "epoch": 0.8288308067225045,
+      "grad_norm": 1.4765625,
+      "learning_rate": 3.5489251034799608e-06,
+      "loss": 0.328,
+      "step": 18876
+    },
+    {
+      "epoch": 0.8289186252017081,
+      "grad_norm": 1.46875,
+      "learning_rate": 3.545372847530193e-06,
+      "loss": 0.3488,
+      "step": 18878
+    },
+    {
+      "epoch": 0.8290064436809116,
+      "grad_norm": 1.5390625,
+      "learning_rate": 3.5418222345809186e-06,
+      "loss": 0.3083,
+      "step": 18880
+    },
+    {
+      "epoch": 0.8290942621601151,
+      "grad_norm": 1.6640625,
+      "learning_rate": 3.5382732649040405e-06,
+      "loss": 0.3395,
+      "step": 18882
+    },
+    {
+      "epoch": 0.8291820806393185,
+      "grad_norm": 1.53125,
+      "learning_rate": 3.5347259387713354e-06,
+      "loss": 0.314,
+      "step": 18884
+    },
+    {
+      "epoch": 0.829269899118522,
+      "grad_norm": 1.5078125,
+      "learning_rate": 3.5311802564544723e-06,
+      "loss": 0.2924,
+      "step": 18886
+    },
+    {
+      "epoch": 0.8293577175977255,
+      "grad_norm": 1.46875,
+      "learning_rate": 3.527636218224961e-06,
+      "loss": 0.3303,
+      "step": 18888
+    },
+    {
+      "epoch": 0.829445536076929,
+      "grad_norm": 1.578125,
+      "learning_rate": 3.5240938243542244e-06,
+      "loss": 0.3289,
+      "step": 18890
+    },
+    {
+      "epoch": 0.8295333545561325,
+      "grad_norm": 1.4375,
+      "learning_rate": 3.5205530751135307e-06,
+      "loss": 0.3181,
+      "step": 18892
+    },
+    {
+      "epoch": 0.829621173035336,
+      "grad_norm": 1.6015625,
+      "learning_rate": 3.5170139707740247e-06,
+      "loss": 0.3109,
+      "step": 18894
+    },
+    {
+      "epoch": 0.8297089915145395,
+      "grad_norm": 1.5625,
+      "learning_rate": 3.5134765116067504e-06,
+      "loss": 0.3041,
+      "step": 18896
+    },
+    {
+      "epoch": 0.829796809993743,
+      "grad_norm": 1.453125,
+      "learning_rate": 3.5099406978825857e-06,
+      "loss": 0.3023,
+      "step": 18898
+    },
+    {
+      "epoch": 0.8298846284729464,
+      "grad_norm": 1.578125,
+      "learning_rate": 3.5064065298723163e-06,
+      "loss": 0.3104,
+      "step": 18900
+    },
+    {
+      "epoch": 0.8299724469521499,
+      "grad_norm": 1.46875,
+      "learning_rate": 3.502874007846585e-06,
+      "loss": 0.3267,
+      "step": 18902
+    },
+    {
+      "epoch": 0.8300602654313534,
+      "grad_norm": 1.4609375,
+      "learning_rate": 3.499343132075919e-06,
+      "loss": 0.2919,
+      "step": 18904
+    },
+    {
+      "epoch": 0.8301480839105568,
+      "grad_norm": 1.59375,
+      "learning_rate": 3.495813902830711e-06,
+      "loss": 0.3234,
+      "step": 18906
+    },
+    {
+      "epoch": 0.8302359023897604,
+      "grad_norm": 1.4609375,
+      "learning_rate": 3.492286320381222e-06,
+      "loss": 0.3001,
+      "step": 18908
+    },
+    {
+      "epoch": 0.8303237208689639,
+      "grad_norm": 1.5859375,
+      "learning_rate": 3.488760384997608e-06,
+      "loss": 0.2751,
+      "step": 18910
+    },
+    {
+      "epoch": 0.8304115393481674,
+      "grad_norm": 1.53125,
+      "learning_rate": 3.485236096949876e-06,
+      "loss": 0.2929,
+      "step": 18912
+    },
+    {
+      "epoch": 0.8304993578273708,
+      "grad_norm": 1.4453125,
+      "learning_rate": 3.4817134565079264e-06,
+      "loss": 0.3345,
+      "step": 18914
+    },
+    {
+      "epoch": 0.8305871763065743,
+      "grad_norm": 1.53125,
+      "learning_rate": 3.478192463941518e-06,
+      "loss": 0.3347,
+      "step": 18916
+    },
+    {
+      "epoch": 0.8306749947857778,
+      "grad_norm": 1.546875,
+      "learning_rate": 3.4746731195202857e-06,
+      "loss": 0.3297,
+      "step": 18918
+    },
+    {
+      "epoch": 0.8307628132649812,
+      "grad_norm": 1.484375,
+      "learning_rate": 3.47115542351375e-06,
+      "loss": 0.2782,
+      "step": 18920
+    },
+    {
+      "epoch": 0.8308506317441847,
+      "grad_norm": 1.375,
+      "learning_rate": 3.467639376191287e-06,
+      "loss": 0.3137,
+      "step": 18922
+    },
+    {
+      "epoch": 0.8309384502233883,
+      "grad_norm": 1.4140625,
+      "learning_rate": 3.4641249778221694e-06,
+      "loss": 0.3205,
+      "step": 18924
+    },
+    {
+      "epoch": 0.8310262687025918,
+      "grad_norm": 1.4609375,
+      "learning_rate": 3.460612228675522e-06,
+      "loss": 0.3171,
+      "step": 18926
+    },
+    {
+      "epoch": 0.8311140871817952,
+      "grad_norm": 1.484375,
+      "learning_rate": 3.4571011290203543e-06,
+      "loss": 0.2997,
+      "step": 18928
+    },
+    {
+      "epoch": 0.8312019056609987,
+      "grad_norm": 1.46875,
+      "learning_rate": 3.4535916791255424e-06,
+      "loss": 0.3184,
+      "step": 18930
+    },
+    {
+      "epoch": 0.8312897241402022,
+      "grad_norm": 1.53125,
+      "learning_rate": 3.4500838792598467e-06,
+      "loss": 0.3275,
+      "step": 18932
+    },
+    {
+      "epoch": 0.8313775426194057,
+      "grad_norm": 1.5,
+      "learning_rate": 3.4465777296918965e-06,
+      "loss": 0.3294,
+      "step": 18934
+    },
+    {
+      "epoch": 0.8314653610986091,
+      "grad_norm": 1.4765625,
+      "learning_rate": 3.443073230690183e-06,
+      "loss": 0.3117,
+      "step": 18936
+    },
+    {
+      "epoch": 0.8315531795778126,
+      "grad_norm": 1.484375,
+      "learning_rate": 3.439570382523094e-06,
+      "loss": 0.2968,
+      "step": 18938
+    },
+    {
+      "epoch": 0.8316409980570162,
+      "grad_norm": 1.484375,
+      "learning_rate": 3.4360691854588707e-06,
+      "loss": 0.3137,
+      "step": 18940
+    },
+    {
+      "epoch": 0.8317288165362197,
+      "grad_norm": 1.4765625,
+      "learning_rate": 3.4325696397656425e-06,
+      "loss": 0.3005,
+      "step": 18942
+    },
+    {
+      "epoch": 0.8318166350154231,
+      "grad_norm": 1.4765625,
+      "learning_rate": 3.429071745711401e-06,
+      "loss": 0.3195,
+      "step": 18944
+    },
+    {
+      "epoch": 0.8319044534946266,
+      "grad_norm": 1.5078125,
+      "learning_rate": 3.42557550356401e-06,
+      "loss": 0.3318,
+      "step": 18946
+    },
+    {
+      "epoch": 0.8319922719738301,
+      "grad_norm": 1.453125,
+      "learning_rate": 3.4220809135912247e-06,
+      "loss": 0.3077,
+      "step": 18948
+    },
+    {
+      "epoch": 0.8320800904530335,
+      "grad_norm": 1.390625,
+      "learning_rate": 3.418587976060653e-06,
+      "loss": 0.2822,
+      "step": 18950
+    },
+    {
+      "epoch": 0.832167908932237,
+      "grad_norm": 1.5234375,
+      "learning_rate": 3.4150966912397888e-06,
+      "loss": 0.329,
+      "step": 18952
+    },
+    {
+      "epoch": 0.8322557274114406,
+      "grad_norm": 1.4296875,
+      "learning_rate": 3.4116070593959963e-06,
+      "loss": 0.3254,
+      "step": 18954
+    },
+    {
+      "epoch": 0.8323435458906441,
+      "grad_norm": 1.5234375,
+      "learning_rate": 3.4081190807965043e-06,
+      "loss": 0.3085,
+      "step": 18956
+    },
+    {
+      "epoch": 0.8324313643698475,
+      "grad_norm": 1.46875,
+      "learning_rate": 3.4046327557084396e-06,
+      "loss": 0.3283,
+      "step": 18958
+    },
+    {
+      "epoch": 0.832519182849051,
+      "grad_norm": 1.40625,
+      "learning_rate": 3.4011480843987643e-06,
+      "loss": 0.3184,
+      "step": 18960
+    },
+    {
+      "epoch": 0.8326070013282545,
+      "grad_norm": 1.4453125,
+      "learning_rate": 3.3976650671343535e-06,
+      "loss": 0.3031,
+      "step": 18962
+    },
+    {
+      "epoch": 0.832694819807458,
+      "grad_norm": 1.609375,
+      "learning_rate": 3.3941837041819247e-06,
+      "loss": 0.3436,
+      "step": 18964
+    },
+    {
+      "epoch": 0.8327826382866614,
+      "grad_norm": 1.4921875,
+      "learning_rate": 3.3907039958080895e-06,
+      "loss": 0.3419,
+      "step": 18966
+    },
+    {
+      "epoch": 0.8328704567658649,
+      "grad_norm": 1.5078125,
+      "learning_rate": 3.3872259422793263e-06,
+      "loss": 0.3346,
+      "step": 18968
+    },
+    {
+      "epoch": 0.8329582752450685,
+      "grad_norm": 1.5234375,
+      "learning_rate": 3.383749543861972e-06,
+      "loss": 0.3143,
+      "step": 18970
+    },
+    {
+      "epoch": 0.833046093724272,
+      "grad_norm": 1.4765625,
+      "learning_rate": 3.3802748008222667e-06,
+      "loss": 0.3023,
+      "step": 18972
+    },
+    {
+      "epoch": 0.8331339122034754,
+      "grad_norm": 1.6015625,
+      "learning_rate": 3.3768017134262945e-06,
+      "loss": 0.3151,
+      "step": 18974
+    },
+    {
+      "epoch": 0.8332217306826789,
+      "grad_norm": 1.5390625,
+      "learning_rate": 3.3733302819400376e-06,
+      "loss": 0.3068,
+      "step": 18976
+    },
+    {
+      "epoch": 0.8333095491618824,
+      "grad_norm": 1.3984375,
+      "learning_rate": 3.36986050662933e-06,
+      "loss": 0.3282,
+      "step": 18978
+    },
+    {
+      "epoch": 0.8333973676410859,
+      "grad_norm": 1.3984375,
+      "learning_rate": 3.366392387759884e-06,
+      "loss": 0.3174,
+      "step": 18980
+    },
+    {
+      "epoch": 0.8334851861202893,
+      "grad_norm": 1.4453125,
+      "learning_rate": 3.3629259255973017e-06,
+      "loss": 0.3048,
+      "step": 18982
+    },
+    {
+      "epoch": 0.8335730045994928,
+      "grad_norm": 1.4140625,
+      "learning_rate": 3.3594611204070313e-06,
+      "loss": 0.3117,
+      "step": 18984
+    },
+    {
+      "epoch": 0.8336608230786964,
+      "grad_norm": 1.46875,
+      "learning_rate": 3.355997972454425e-06,
+      "loss": 0.3045,
+      "step": 18986
+    },
+    {
+      "epoch": 0.8337486415578998,
+      "grad_norm": 1.5390625,
+      "learning_rate": 3.3525364820046782e-06,
+      "loss": 0.3071,
+      "step": 18988
+    },
+    {
+      "epoch": 0.8338364600371033,
+      "grad_norm": 1.46875,
+      "learning_rate": 3.34907664932288e-06,
+      "loss": 0.2948,
+      "step": 18990
+    },
+    {
+      "epoch": 0.8339242785163068,
+      "grad_norm": 1.546875,
+      "learning_rate": 3.3456184746739753e-06,
+      "loss": 0.3179,
+      "step": 18992
+    },
+    {
+      "epoch": 0.8340120969955103,
+      "grad_norm": 1.453125,
+      "learning_rate": 3.3421619583228036e-06,
+      "loss": 0.3317,
+      "step": 18994
+    },
+    {
+      "epoch": 0.8340999154747137,
+      "grad_norm": 1.4296875,
+      "learning_rate": 3.338707100534061e-06,
+      "loss": 0.2953,
+      "step": 18996
+    },
+    {
+      "epoch": 0.8341877339539172,
+      "grad_norm": 1.4140625,
+      "learning_rate": 3.335253901572316e-06,
+      "loss": 0.2883,
+      "step": 18998
+    },
+    {
+      "epoch": 0.8342755524331208,
+      "grad_norm": 1.5625,
+      "learning_rate": 3.3318023617020273e-06,
+      "loss": 0.3376,
+      "step": 19000
+    },
+    {
+      "epoch": 0.8343633709123243,
+      "grad_norm": 1.4375,
+      "learning_rate": 3.3283524811875023e-06,
+      "loss": 0.3111,
+      "step": 19002
+    },
+    {
+      "epoch": 0.8344511893915277,
+      "grad_norm": 1.421875,
+      "learning_rate": 3.3249042602929437e-06,
+      "loss": 0.3274,
+      "step": 19004
+    },
+    {
+      "epoch": 0.8345390078707312,
+      "grad_norm": 1.4921875,
+      "learning_rate": 3.3214576992824125e-06,
+      "loss": 0.336,
+      "step": 19006
+    },
+    {
+      "epoch": 0.8346268263499347,
+      "grad_norm": 1.46875,
+      "learning_rate": 3.3180127984198423e-06,
+      "loss": 0.3123,
+      "step": 19008
+    },
+    {
+      "epoch": 0.8347146448291382,
+      "grad_norm": 1.4921875,
+      "learning_rate": 3.314569557969055e-06,
+      "loss": 0.3108,
+      "step": 19010
+    },
+    {
+      "epoch": 0.8348024633083416,
+      "grad_norm": 1.40625,
+      "learning_rate": 3.311127978193726e-06,
+      "loss": 0.3298,
+      "step": 19012
+    },
+    {
+      "epoch": 0.8348902817875451,
+      "grad_norm": 1.4765625,
+      "learning_rate": 3.3076880593574196e-06,
+      "loss": 0.3208,
+      "step": 19014
+    },
+    {
+      "epoch": 0.8349781002667487,
+      "grad_norm": 1.6328125,
+      "learning_rate": 3.3042498017235607e-06,
+      "loss": 0.342,
+      "step": 19016
+    },
+    {
+      "epoch": 0.8350659187459522,
+      "grad_norm": 1.4375,
+      "learning_rate": 3.300813205555453e-06,
+      "loss": 0.2855,
+      "step": 19018
+    },
+    {
+      "epoch": 0.8351537372251556,
+      "grad_norm": 1.46875,
+      "learning_rate": 3.2973782711162722e-06,
+      "loss": 0.3106,
+      "step": 19020
+    },
+    {
+      "epoch": 0.8352415557043591,
+      "grad_norm": 1.46875,
+      "learning_rate": 3.29394499866906e-06,
+      "loss": 0.3279,
+      "step": 19022
+    },
+    {
+      "epoch": 0.8353293741835626,
+      "grad_norm": 1.4375,
+      "learning_rate": 3.2905133884767486e-06,
+      "loss": 0.3162,
+      "step": 19024
+    },
+    {
+      "epoch": 0.835417192662766,
+      "grad_norm": 1.4609375,
+      "learning_rate": 3.287083440802122e-06,
+      "loss": 0.3103,
+      "step": 19026
+    },
+    {
+      "epoch": 0.8355050111419695,
+      "grad_norm": 1.4453125,
+      "learning_rate": 3.2836551559078525e-06,
+      "loss": 0.2965,
+      "step": 19028
+    },
+    {
+      "epoch": 0.835592829621173,
+      "grad_norm": 1.40625,
+      "learning_rate": 3.280228534056479e-06,
+      "loss": 0.3235,
+      "step": 19030
+    },
+    {
+      "epoch": 0.8356806481003766,
+      "grad_norm": 1.46875,
+      "learning_rate": 3.2768035755104064e-06,
+      "loss": 0.29,
+      "step": 19032
+    },
+    {
+      "epoch": 0.83576846657958,
+      "grad_norm": 1.453125,
+      "learning_rate": 3.273380280531929e-06,
+      "loss": 0.3043,
+      "step": 19034
+    },
+    {
+      "epoch": 0.8358562850587835,
+      "grad_norm": 1.5078125,
+      "learning_rate": 3.2699586493831895e-06,
+      "loss": 0.3144,
+      "step": 19036
+    },
+    {
+      "epoch": 0.835944103537987,
+      "grad_norm": 1.4375,
+      "learning_rate": 3.266538682326234e-06,
+      "loss": 0.3357,
+      "step": 19038
+    },
+    {
+      "epoch": 0.8360319220171905,
+      "grad_norm": 1.4453125,
+      "learning_rate": 3.2631203796229555e-06,
+      "loss": 0.3015,
+      "step": 19040
+    },
+    {
+      "epoch": 0.8361197404963939,
+      "grad_norm": 1.4140625,
+      "learning_rate": 3.259703741535122e-06,
+      "loss": 0.2977,
+      "step": 19042
+    },
+    {
+      "epoch": 0.8362075589755974,
+      "grad_norm": 1.3984375,
+      "learning_rate": 3.256288768324395e-06,
+      "loss": 0.299,
+      "step": 19044
+    },
+    {
+      "epoch": 0.836295377454801,
+      "grad_norm": 1.5234375,
+      "learning_rate": 3.2528754602522804e-06,
+      "loss": 0.3027,
+      "step": 19046
+    },
+    {
+      "epoch": 0.8363831959340045,
+      "grad_norm": 1.6015625,
+      "learning_rate": 3.249463817580181e-06,
+      "loss": 0.3362,
+      "step": 19048
+    },
+    {
+      "epoch": 0.8364710144132079,
+      "grad_norm": 1.5234375,
+      "learning_rate": 3.2460538405693573e-06,
+      "loss": 0.31,
+      "step": 19050
+    },
+    {
+      "epoch": 0.8365588328924114,
+      "grad_norm": 1.46875,
+      "learning_rate": 3.242645529480945e-06,
+      "loss": 0.3407,
+      "step": 19052
+    },
+    {
+      "epoch": 0.8366466513716149,
+      "grad_norm": 1.5390625,
+      "learning_rate": 3.239238884575949e-06,
+      "loss": 0.3327,
+      "step": 19054
+    },
+    {
+      "epoch": 0.8367344698508183,
+      "grad_norm": 1.4609375,
+      "learning_rate": 3.23583390611526e-06,
+      "loss": 0.3321,
+      "step": 19056
+    },
+    {
+      "epoch": 0.8368222883300218,
+      "grad_norm": 1.4765625,
+      "learning_rate": 3.2324305943596256e-06,
+      "loss": 0.2949,
+      "step": 19058
+    },
+    {
+      "epoch": 0.8369101068092253,
+      "grad_norm": 1.5078125,
+      "learning_rate": 3.2290289495696705e-06,
+      "loss": 0.2996,
+      "step": 19060
+    },
+    {
+      "epoch": 0.8369979252884289,
+      "grad_norm": 1.5390625,
+      "learning_rate": 3.2256289720059035e-06,
+      "loss": 0.3195,
+      "step": 19062
+    },
+    {
+      "epoch": 0.8370857437676323,
+      "grad_norm": 1.5390625,
+      "learning_rate": 3.2222306619286852e-06,
+      "loss": 0.2868,
+      "step": 19064
+    },
+    {
+      "epoch": 0.8371735622468358,
+      "grad_norm": 1.4765625,
+      "learning_rate": 3.2188340195982685e-06,
+      "loss": 0.3008,
+      "step": 19066
+    },
+    {
+      "epoch": 0.8372613807260393,
+      "grad_norm": 1.453125,
+      "learning_rate": 3.2154390452747625e-06,
+      "loss": 0.3051,
+      "step": 19068
+    },
+    {
+      "epoch": 0.8373491992052428,
+      "grad_norm": 1.546875,
+      "learning_rate": 3.2120457392181503e-06,
+      "loss": 0.3128,
+      "step": 19070
+    },
+    {
+      "epoch": 0.8374370176844462,
+      "grad_norm": 1.46875,
+      "learning_rate": 3.208654101688305e-06,
+      "loss": 0.3044,
+      "step": 19072
+    },
+    {
+      "epoch": 0.8375248361636497,
+      "grad_norm": 1.3984375,
+      "learning_rate": 3.205264132944946e-06,
+      "loss": 0.3149,
+      "step": 19074
+    },
+    {
+      "epoch": 0.8376126546428532,
+      "grad_norm": 1.4765625,
+      "learning_rate": 3.2018758332476916e-06,
+      "loss": 0.3165,
+      "step": 19076
+    },
+    {
+      "epoch": 0.8377004731220568,
+      "grad_norm": 1.4375,
+      "learning_rate": 3.198489202856009e-06,
+      "loss": 0.2965,
+      "step": 19078
+    },
+    {
+      "epoch": 0.8377882916012602,
+      "grad_norm": 1.5078125,
+      "learning_rate": 3.195104242029251e-06,
+      "loss": 0.3335,
+      "step": 19080
+    },
+    {
+      "epoch": 0.8378761100804637,
+      "grad_norm": 1.6171875,
+      "learning_rate": 3.191720951026636e-06,
+      "loss": 0.3275,
+      "step": 19082
+    },
+    {
+      "epoch": 0.8379639285596672,
+      "grad_norm": 1.515625,
+      "learning_rate": 3.188339330107254e-06,
+      "loss": 0.3249,
+      "step": 19084
+    },
+    {
+      "epoch": 0.8380517470388706,
+      "grad_norm": 1.4609375,
+      "learning_rate": 3.184959379530081e-06,
+      "loss": 0.3148,
+      "step": 19086
+    },
+    {
+      "epoch": 0.8381395655180741,
+      "grad_norm": 1.46875,
+      "learning_rate": 3.1815810995539404e-06,
+      "loss": 0.3339,
+      "step": 19088
+    },
+    {
+      "epoch": 0.8382273839972776,
+      "grad_norm": 1.40625,
+      "learning_rate": 3.178204490437556e-06,
+      "loss": 0.3135,
+      "step": 19090
+    },
+    {
+      "epoch": 0.8383152024764811,
+      "grad_norm": 1.4765625,
+      "learning_rate": 3.174829552439504e-06,
+      "loss": 0.3102,
+      "step": 19092
+    },
+    {
+      "epoch": 0.8384030209556846,
+      "grad_norm": 1.5703125,
+      "learning_rate": 3.1714562858182277e-06,
+      "loss": 0.3405,
+      "step": 19094
+    },
+    {
+      "epoch": 0.8384908394348881,
+      "grad_norm": 1.5546875,
+      "learning_rate": 3.168084690832071e-06,
+      "loss": 0.3078,
+      "step": 19096
+    },
+    {
+      "epoch": 0.8385786579140916,
+      "grad_norm": 1.4765625,
+      "learning_rate": 3.1647147677392157e-06,
+      "loss": 0.3009,
+      "step": 19098
+    },
+    {
+      "epoch": 0.8386664763932951,
+      "grad_norm": 1.515625,
+      "learning_rate": 3.161346516797742e-06,
+      "loss": 0.3162,
+      "step": 19100
+    },
+    {
+      "epoch": 0.8387542948724985,
+      "grad_norm": 1.53125,
+      "learning_rate": 3.157979938265587e-06,
+      "loss": 0.3176,
+      "step": 19102
+    },
+    {
+      "epoch": 0.838842113351702,
+      "grad_norm": 1.5546875,
+      "learning_rate": 3.1546150324005595e-06,
+      "loss": 0.3332,
+      "step": 19104
+    },
+    {
+      "epoch": 0.8389299318309055,
+      "grad_norm": 1.4453125,
+      "learning_rate": 3.1512517994603556e-06,
+      "loss": 0.2953,
+      "step": 19106
+    },
+    {
+      "epoch": 0.8390177503101091,
+      "grad_norm": 1.4140625,
+      "learning_rate": 3.1478902397025196e-06,
+      "loss": 0.3121,
+      "step": 19108
+    },
+    {
+      "epoch": 0.8391055687893125,
+      "grad_norm": 1.4375,
+      "learning_rate": 3.1445303533845005e-06,
+      "loss": 0.303,
+      "step": 19110
+    },
+    {
+      "epoch": 0.839193387268516,
+      "grad_norm": 1.5,
+      "learning_rate": 3.1411721407635738e-06,
+      "loss": 0.3029,
+      "step": 19112
+    },
+    {
+      "epoch": 0.8392812057477195,
+      "grad_norm": 1.5703125,
+      "learning_rate": 3.13781560209693e-06,
+      "loss": 0.3016,
+      "step": 19114
+    },
+    {
+      "epoch": 0.8393690242269229,
+      "grad_norm": 1.4609375,
+      "learning_rate": 3.1344607376416008e-06,
+      "loss": 0.3275,
+      "step": 19116
+    },
+    {
+      "epoch": 0.8394568427061264,
+      "grad_norm": 1.4453125,
+      "learning_rate": 3.131107547654519e-06,
+      "loss": 0.2944,
+      "step": 19118
+    },
+    {
+      "epoch": 0.8395446611853299,
+      "grad_norm": 1.515625,
+      "learning_rate": 3.1277560323924593e-06,
+      "loss": 0.3159,
+      "step": 19120
+    },
+    {
+      "epoch": 0.8396324796645334,
+      "grad_norm": 1.4453125,
+      "learning_rate": 3.1244061921120838e-06,
+      "loss": 0.3307,
+      "step": 19122
+    },
+    {
+      "epoch": 0.8397202981437369,
+      "grad_norm": 1.53125,
+      "learning_rate": 3.1210580270699285e-06,
+      "loss": 0.3018,
+      "step": 19124
+    },
+    {
+      "epoch": 0.8398081166229404,
+      "grad_norm": 1.4453125,
+      "learning_rate": 3.1177115375223915e-06,
+      "loss": 0.3169,
+      "step": 19126
+    },
+    {
+      "epoch": 0.8398959351021439,
+      "grad_norm": 1.5546875,
+      "learning_rate": 3.1143667237257563e-06,
+      "loss": 0.3231,
+      "step": 19128
+    },
+    {
+      "epoch": 0.8399837535813474,
+      "grad_norm": 1.4296875,
+      "learning_rate": 3.1110235859361597e-06,
+      "loss": 0.3246,
+      "step": 19130
+    },
+    {
+      "epoch": 0.8400715720605508,
+      "grad_norm": 1.4375,
+      "learning_rate": 3.107682124409622e-06,
+      "loss": 0.2926,
+      "step": 19132
+    },
+    {
+      "epoch": 0.8401593905397543,
+      "grad_norm": 1.625,
+      "learning_rate": 3.1043423394020417e-06,
+      "loss": 0.3063,
+      "step": 19134
+    },
+    {
+      "epoch": 0.8402472090189578,
+      "grad_norm": 1.4453125,
+      "learning_rate": 3.1010042311691663e-06,
+      "loss": 0.3252,
+      "step": 19136
+    },
+    {
+      "epoch": 0.8403350274981612,
+      "grad_norm": 1.4921875,
+      "learning_rate": 3.0976677999666414e-06,
+      "loss": 0.3084,
+      "step": 19138
+    },
+    {
+      "epoch": 0.8404228459773648,
+      "grad_norm": 1.4765625,
+      "learning_rate": 3.094333046049966e-06,
+      "loss": 0.3037,
+      "step": 19140
+    },
+    {
+      "epoch": 0.8405106644565683,
+      "grad_norm": 1.5859375,
+      "learning_rate": 3.0909999696745185e-06,
+      "loss": 0.3232,
+      "step": 19142
+    },
+    {
+      "epoch": 0.8405984829357718,
+      "grad_norm": 1.4453125,
+      "learning_rate": 3.0876685710955476e-06,
+      "loss": 0.3094,
+      "step": 19144
+    },
+    {
+      "epoch": 0.8406863014149752,
+      "grad_norm": 1.53125,
+      "learning_rate": 3.084338850568161e-06,
+      "loss": 0.2993,
+      "step": 19146
+    },
+    {
+      "epoch": 0.8407741198941787,
+      "grad_norm": 1.546875,
+      "learning_rate": 3.081010808347365e-06,
+      "loss": 0.314,
+      "step": 19148
+    },
+    {
+      "epoch": 0.8408619383733822,
+      "grad_norm": 1.4765625,
+      "learning_rate": 3.0776844446880115e-06,
+      "loss": 0.3055,
+      "step": 19150
+    },
+    {
+      "epoch": 0.8409497568525857,
+      "grad_norm": 1.40625,
+      "learning_rate": 3.074359759844844e-06,
+      "loss": 0.2917,
+      "step": 19152
+    },
+    {
+      "epoch": 0.8410375753317892,
+      "grad_norm": 1.5078125,
+      "learning_rate": 3.071036754072462e-06,
+      "loss": 0.3387,
+      "step": 19154
+    },
+    {
+      "epoch": 0.8411253938109927,
+      "grad_norm": 1.453125,
+      "learning_rate": 3.067715427625334e-06,
+      "loss": 0.3268,
+      "step": 19156
+    },
+    {
+      "epoch": 0.8412132122901962,
+      "grad_norm": 1.4296875,
+      "learning_rate": 3.0643957807578253e-06,
+      "loss": 0.2954,
+      "step": 19158
+    },
+    {
+      "epoch": 0.8413010307693997,
+      "grad_norm": 1.359375,
+      "learning_rate": 3.061077813724139e-06,
+      "loss": 0.3172,
+      "step": 19160
+    },
+    {
+      "epoch": 0.8413888492486031,
+      "grad_norm": 1.453125,
+      "learning_rate": 3.0577615267783773e-06,
+      "loss": 0.2994,
+      "step": 19162
+    },
+    {
+      "epoch": 0.8414766677278066,
+      "grad_norm": 1.453125,
+      "learning_rate": 3.0544469201744976e-06,
+      "loss": 0.3058,
+      "step": 19164
+    },
+    {
+      "epoch": 0.8415644862070101,
+      "grad_norm": 1.4765625,
+      "learning_rate": 3.0511339941663303e-06,
+      "loss": 0.313,
+      "step": 19166
+    },
+    {
+      "epoch": 0.8416523046862135,
+      "grad_norm": 1.4296875,
+      "learning_rate": 3.0478227490075866e-06,
+      "loss": 0.313,
+      "step": 19168
+    },
+    {
+      "epoch": 0.8417401231654171,
+      "grad_norm": 1.4453125,
+      "learning_rate": 3.044513184951833e-06,
+      "loss": 0.3098,
+      "step": 19170
+    },
+    {
+      "epoch": 0.8418279416446206,
+      "grad_norm": 1.515625,
+      "learning_rate": 3.0412053022525367e-06,
+      "loss": 0.2796,
+      "step": 19172
+    },
+    {
+      "epoch": 0.8419157601238241,
+      "grad_norm": 1.4921875,
+      "learning_rate": 3.037899101162989e-06,
+      "loss": 0.3176,
+      "step": 19174
+    },
+    {
+      "epoch": 0.8420035786030275,
+      "grad_norm": 1.46875,
+      "learning_rate": 3.034594581936398e-06,
+      "loss": 0.2792,
+      "step": 19176
+    },
+    {
+      "epoch": 0.842091397082231,
+      "grad_norm": 1.5078125,
+      "learning_rate": 3.0312917448258205e-06,
+      "loss": 0.3012,
+      "step": 19178
+    },
+    {
+      "epoch": 0.8421792155614345,
+      "grad_norm": 1.375,
+      "learning_rate": 3.0279905900841815e-06,
+      "loss": 0.3114,
+      "step": 19180
+    },
+    {
+      "epoch": 0.842267034040638,
+      "grad_norm": 1.4765625,
+      "learning_rate": 3.024691117964298e-06,
+      "loss": 0.3407,
+      "step": 19182
+    },
+    {
+      "epoch": 0.8423548525198414,
+      "grad_norm": 1.5,
+      "learning_rate": 3.0213933287188272e-06,
+      "loss": 0.2972,
+      "step": 19184
+    },
+    {
+      "epoch": 0.842442670999045,
+      "grad_norm": 1.5,
+      "learning_rate": 3.0180972226003327e-06,
+      "loss": 0.2992,
+      "step": 19186
+    },
+    {
+      "epoch": 0.8425304894782485,
+      "grad_norm": 1.4296875,
+      "learning_rate": 3.014802799861216e-06,
+      "loss": 0.2968,
+      "step": 19188
+    },
+    {
+      "epoch": 0.842618307957452,
+      "grad_norm": 1.484375,
+      "learning_rate": 3.011510060753775e-06,
+      "loss": 0.3046,
+      "step": 19190
+    },
+    {
+      "epoch": 0.8427061264366554,
+      "grad_norm": 1.4375,
+      "learning_rate": 3.008219005530166e-06,
+      "loss": 0.3093,
+      "step": 19192
+    },
+    {
+      "epoch": 0.8427939449158589,
+      "grad_norm": 1.5234375,
+      "learning_rate": 3.0049296344424103e-06,
+      "loss": 0.3324,
+      "step": 19194
+    },
+    {
+      "epoch": 0.8428817633950624,
+      "grad_norm": 1.5625,
+      "learning_rate": 3.001641947742423e-06,
+      "loss": 0.3031,
+      "step": 19196
+    },
+    {
+      "epoch": 0.8429695818742658,
+      "grad_norm": 1.53125,
+      "learning_rate": 2.9983559456819633e-06,
+      "loss": 0.3161,
+      "step": 19198
+    },
+    {
+      "epoch": 0.8430574003534694,
+      "grad_norm": 1.4375,
+      "learning_rate": 2.9950716285126827e-06,
+      "loss": 0.3365,
+      "step": 19200
+    },
+    {
+      "epoch": 0.8431452188326729,
+      "grad_norm": 1.390625,
+      "learning_rate": 2.9917889964860917e-06,
+      "loss": 0.3279,
+      "step": 19202
+    },
+    {
+      "epoch": 0.8432330373118764,
+      "grad_norm": 1.5,
+      "learning_rate": 2.9885080498535777e-06,
+      "loss": 0.3001,
+      "step": 19204
+    },
+    {
+      "epoch": 0.8433208557910798,
+      "grad_norm": 1.5,
+      "learning_rate": 2.9852287888663928e-06,
+      "loss": 0.2887,
+      "step": 19206
+    },
+    {
+      "epoch": 0.8434086742702833,
+      "grad_norm": 1.5703125,
+      "learning_rate": 2.9819512137756577e-06,
+      "loss": 0.3053,
+      "step": 19208
+    },
+    {
+      "epoch": 0.8434964927494868,
+      "grad_norm": 1.4453125,
+      "learning_rate": 2.9786753248323833e-06,
+      "loss": 0.3567,
+      "step": 19210
+    },
+    {
+      "epoch": 0.8435843112286903,
+      "grad_norm": 1.5234375,
+      "learning_rate": 2.9754011222874275e-06,
+      "loss": 0.2944,
+      "step": 19212
+    },
+    {
+      "epoch": 0.8436721297078937,
+      "grad_norm": 1.4609375,
+      "learning_rate": 2.972128606391536e-06,
+      "loss": 0.3047,
+      "step": 19214
+    },
+    {
+      "epoch": 0.8437599481870973,
+      "grad_norm": 1.4609375,
+      "learning_rate": 2.9688577773953176e-06,
+      "loss": 0.2938,
+      "step": 19216
+    },
+    {
+      "epoch": 0.8438477666663008,
+      "grad_norm": 1.453125,
+      "learning_rate": 2.965588635549249e-06,
+      "loss": 0.3215,
+      "step": 19218
+    },
+    {
+      "epoch": 0.8439355851455043,
+      "grad_norm": 1.6171875,
+      "learning_rate": 2.9623211811036862e-06,
+      "loss": 0.3035,
+      "step": 19220
+    },
+    {
+      "epoch": 0.8440234036247077,
+      "grad_norm": 1.4609375,
+      "learning_rate": 2.959055414308845e-06,
+      "loss": 0.3194,
+      "step": 19222
+    },
+    {
+      "epoch": 0.8441112221039112,
+      "grad_norm": 1.453125,
+      "learning_rate": 2.9557913354148316e-06,
+      "loss": 0.3443,
+      "step": 19224
+    },
+    {
+      "epoch": 0.8441990405831147,
+      "grad_norm": 1.4765625,
+      "learning_rate": 2.952528944671601e-06,
+      "loss": 0.3636,
+      "step": 19226
+    },
+    {
+      "epoch": 0.8442868590623182,
+      "grad_norm": 1.5703125,
+      "learning_rate": 2.9492682423289843e-06,
+      "loss": 0.2937,
+      "step": 19228
+    },
+    {
+      "epoch": 0.8443746775415216,
+      "grad_norm": 1.421875,
+      "learning_rate": 2.9460092286366955e-06,
+      "loss": 0.3119,
+      "step": 19230
+    },
+    {
+      "epoch": 0.8444624960207252,
+      "grad_norm": 1.4921875,
+      "learning_rate": 2.9427519038443014e-06,
+      "loss": 0.2929,
+      "step": 19232
+    },
+    {
+      "epoch": 0.8445503144999287,
+      "grad_norm": 1.515625,
+      "learning_rate": 2.939496268201264e-06,
+      "loss": 0.3141,
+      "step": 19234
+    },
+    {
+      "epoch": 0.8446381329791322,
+      "grad_norm": 1.4453125,
+      "learning_rate": 2.936242321956881e-06,
+      "loss": 0.3232,
+      "step": 19236
+    },
+    {
+      "epoch": 0.8447259514583356,
+      "grad_norm": 1.4765625,
+      "learning_rate": 2.9329900653603553e-06,
+      "loss": 0.3165,
+      "step": 19238
+    },
+    {
+      "epoch": 0.8448137699375391,
+      "grad_norm": 1.609375,
+      "learning_rate": 2.929739498660741e-06,
+      "loss": 0.3192,
+      "step": 19240
+    },
+    {
+      "epoch": 0.8449015884167426,
+      "grad_norm": 1.4375,
+      "learning_rate": 2.9264906221069586e-06,
+      "loss": 0.3393,
+      "step": 19242
+    },
+    {
+      "epoch": 0.844989406895946,
+      "grad_norm": 1.5234375,
+      "learning_rate": 2.9232434359478256e-06,
+      "loss": 0.3182,
+      "step": 19244
+    },
+    {
+      "epoch": 0.8450772253751496,
+      "grad_norm": 1.4296875,
+      "learning_rate": 2.919997940431993e-06,
+      "loss": 0.2901,
+      "step": 19246
+    },
+    {
+      "epoch": 0.8451650438543531,
+      "grad_norm": 1.5234375,
+      "learning_rate": 2.916754135808017e-06,
+      "loss": 0.2945,
+      "step": 19248
+    },
+    {
+      "epoch": 0.8452528623335566,
+      "grad_norm": 1.4921875,
+      "learning_rate": 2.9135120223242996e-06,
+      "loss": 0.3016,
+      "step": 19250
+    },
+    {
+      "epoch": 0.84534068081276,
+      "grad_norm": 1.4375,
+      "learning_rate": 2.9102716002291307e-06,
+      "loss": 0.2988,
+      "step": 19252
+    },
+    {
+      "epoch": 0.8454284992919635,
+      "grad_norm": 1.7109375,
+      "learning_rate": 2.907032869770657e-06,
+      "loss": 0.3238,
+      "step": 19254
+    },
+    {
+      "epoch": 0.845516317771167,
+      "grad_norm": 1.5078125,
+      "learning_rate": 2.903795831196898e-06,
+      "loss": 0.3231,
+      "step": 19256
+    },
+    {
+      "epoch": 0.8456041362503705,
+      "grad_norm": 1.4375,
+      "learning_rate": 2.900560484755754e-06,
+      "loss": 0.3126,
+      "step": 19258
+    },
+    {
+      "epoch": 0.8456919547295739,
+      "grad_norm": 1.4609375,
+      "learning_rate": 2.8973268306949824e-06,
+      "loss": 0.3204,
+      "step": 19260
+    },
+    {
+      "epoch": 0.8457797732087775,
+      "grad_norm": 1.5703125,
+      "learning_rate": 2.894094869262226e-06,
+      "loss": 0.3058,
+      "step": 19262
+    },
+    {
+      "epoch": 0.845867591687981,
+      "grad_norm": 1.3984375,
+      "learning_rate": 2.8908646007049843e-06,
+      "loss": 0.3002,
+      "step": 19264
+    },
+    {
+      "epoch": 0.8459554101671845,
+      "grad_norm": 1.5234375,
+      "learning_rate": 2.887636025270629e-06,
+      "loss": 0.3175,
+      "step": 19266
+    },
+    {
+      "epoch": 0.8460432286463879,
+      "grad_norm": 1.5234375,
+      "learning_rate": 2.8844091432064097e-06,
+      "loss": 0.3171,
+      "step": 19268
+    },
+    {
+      "epoch": 0.8461310471255914,
+      "grad_norm": 1.4375,
+      "learning_rate": 2.8811839547594337e-06,
+      "loss": 0.3075,
+      "step": 19270
+    },
+    {
+      "epoch": 0.8462188656047949,
+      "grad_norm": 1.4921875,
+      "learning_rate": 2.8779604601766957e-06,
+      "loss": 0.3231,
+      "step": 19272
+    },
+    {
+      "epoch": 0.8463066840839983,
+      "grad_norm": 1.59375,
+      "learning_rate": 2.874738659705048e-06,
+      "loss": 0.3302,
+      "step": 19274
+    },
+    {
+      "epoch": 0.8463945025632018,
+      "grad_norm": 1.484375,
+      "learning_rate": 2.8715185535912182e-06,
+      "loss": 0.3165,
+      "step": 19276
+    },
+    {
+      "epoch": 0.8464823210424054,
+      "grad_norm": 1.5390625,
+      "learning_rate": 2.868300142081806e-06,
+      "loss": 0.3322,
+      "step": 19278
+    },
+    {
+      "epoch": 0.8465701395216089,
+      "grad_norm": 1.46875,
+      "learning_rate": 2.865083425423265e-06,
+      "loss": 0.3214,
+      "step": 19280
+    },
+    {
+      "epoch": 0.8466579580008123,
+      "grad_norm": 1.4765625,
+      "learning_rate": 2.8618684038619504e-06,
+      "loss": 0.327,
+      "step": 19282
+    },
+    {
+      "epoch": 0.8467457764800158,
+      "grad_norm": 1.484375,
+      "learning_rate": 2.858655077644054e-06,
+      "loss": 0.314,
+      "step": 19284
+    },
+    {
+      "epoch": 0.8468335949592193,
+      "grad_norm": 1.609375,
+      "learning_rate": 2.855443447015663e-06,
+      "loss": 0.3064,
+      "step": 19286
+    },
+    {
+      "epoch": 0.8469214134384228,
+      "grad_norm": 1.4765625,
+      "learning_rate": 2.8522335122227214e-06,
+      "loss": 0.29,
+      "step": 19288
+    },
+    {
+      "epoch": 0.8470092319176262,
+      "grad_norm": 1.6953125,
+      "learning_rate": 2.849025273511044e-06,
+      "loss": 0.3206,
+      "step": 19290
+    },
+    {
+      "epoch": 0.8470970503968297,
+      "grad_norm": 1.453125,
+      "learning_rate": 2.8458187311263233e-06,
+      "loss": 0.327,
+      "step": 19292
+    },
+    {
+      "epoch": 0.8471848688760333,
+      "grad_norm": 1.5078125,
+      "learning_rate": 2.842613885314116e-06,
+      "loss": 0.3152,
+      "step": 19294
+    },
+    {
+      "epoch": 0.8472726873552368,
+      "grad_norm": 1.5,
+      "learning_rate": 2.83941073631985e-06,
+      "loss": 0.3103,
+      "step": 19296
+    },
+    {
+      "epoch": 0.8473605058344402,
+      "grad_norm": 1.5234375,
+      "learning_rate": 2.836209284388816e-06,
+      "loss": 0.3559,
+      "step": 19298
+    },
+    {
+      "epoch": 0.8474483243136437,
+      "grad_norm": 1.4453125,
+      "learning_rate": 2.8330095297661925e-06,
+      "loss": 0.3054,
+      "step": 19300
+    },
+    {
+      "epoch": 0.8475361427928472,
+      "grad_norm": 1.453125,
+      "learning_rate": 2.8298114726970138e-06,
+      "loss": 0.2983,
+      "step": 19302
+    },
+    {
+      "epoch": 0.8476239612720506,
+      "grad_norm": 1.421875,
+      "learning_rate": 2.8266151134261815e-06,
+      "loss": 0.2949,
+      "step": 19304
+    },
+    {
+      "epoch": 0.8477117797512541,
+      "grad_norm": 1.546875,
+      "learning_rate": 2.8234204521984857e-06,
+      "loss": 0.3187,
+      "step": 19306
+    },
+    {
+      "epoch": 0.8477995982304577,
+      "grad_norm": 1.4765625,
+      "learning_rate": 2.8202274892585617e-06,
+      "loss": 0.3119,
+      "step": 19308
+    },
+    {
+      "epoch": 0.8478874167096612,
+      "grad_norm": 1.546875,
+      "learning_rate": 2.8170362248509386e-06,
+      "loss": 0.3204,
+      "step": 19310
+    },
+    {
+      "epoch": 0.8479752351888646,
+      "grad_norm": 1.53125,
+      "learning_rate": 2.8138466592199986e-06,
+      "loss": 0.3137,
+      "step": 19312
+    },
+    {
+      "epoch": 0.8480630536680681,
+      "grad_norm": 1.484375,
+      "learning_rate": 2.8106587926099963e-06,
+      "loss": 0.2915,
+      "step": 19314
+    },
+    {
+      "epoch": 0.8481508721472716,
+      "grad_norm": 1.453125,
+      "learning_rate": 2.8074726252650645e-06,
+      "loss": 0.322,
+      "step": 19316
+    },
+    {
+      "epoch": 0.8482386906264751,
+      "grad_norm": 1.484375,
+      "learning_rate": 2.804288157429197e-06,
+      "loss": 0.3051,
+      "step": 19318
+    },
+    {
+      "epoch": 0.8483265091056785,
+      "grad_norm": 1.5234375,
+      "learning_rate": 2.8011053893462676e-06,
+      "loss": 0.3313,
+      "step": 19320
+    },
+    {
+      "epoch": 0.848414327584882,
+      "grad_norm": 1.515625,
+      "learning_rate": 2.7979243212600035e-06,
+      "loss": 0.331,
+      "step": 19322
+    },
+    {
+      "epoch": 0.8485021460640856,
+      "grad_norm": 1.4921875,
+      "learning_rate": 2.7947449534140217e-06,
+      "loss": 0.3045,
+      "step": 19324
+    },
+    {
+      "epoch": 0.848589964543289,
+      "grad_norm": 1.40625,
+      "learning_rate": 2.7915672860517937e-06,
+      "loss": 0.3273,
+      "step": 19326
+    },
+    {
+      "epoch": 0.8486777830224925,
+      "grad_norm": 1.484375,
+      "learning_rate": 2.7883913194166693e-06,
+      "loss": 0.3184,
+      "step": 19328
+    },
+    {
+      "epoch": 0.848765601501696,
+      "grad_norm": 1.4765625,
+      "learning_rate": 2.7852170537518594e-06,
+      "loss": 0.3319,
+      "step": 19330
+    },
+    {
+      "epoch": 0.8488534199808995,
+      "grad_norm": 1.6953125,
+      "learning_rate": 2.782044489300448e-06,
+      "loss": 0.313,
+      "step": 19332
+    },
+    {
+      "epoch": 0.8489412384601029,
+      "grad_norm": 1.546875,
+      "learning_rate": 2.7788736263054033e-06,
+      "loss": 0.3191,
+      "step": 19334
+    },
+    {
+      "epoch": 0.8490290569393064,
+      "grad_norm": 1.5078125,
+      "learning_rate": 2.775704465009535e-06,
+      "loss": 0.3005,
+      "step": 19336
+    },
+    {
+      "epoch": 0.8491168754185099,
+      "grad_norm": 1.40625,
+      "learning_rate": 2.772537005655554e-06,
+      "loss": 0.2972,
+      "step": 19338
+    },
+    {
+      "epoch": 0.8492046938977135,
+      "grad_norm": 1.46875,
+      "learning_rate": 2.7693712484860135e-06,
+      "loss": 0.3037,
+      "step": 19340
+    },
+    {
+      "epoch": 0.8492925123769169,
+      "grad_norm": 1.484375,
+      "learning_rate": 2.7662071937433502e-06,
+      "loss": 0.3017,
+      "step": 19342
+    },
+    {
+      "epoch": 0.8493803308561204,
+      "grad_norm": 1.53125,
+      "learning_rate": 2.7630448416698734e-06,
+      "loss": 0.304,
+      "step": 19344
+    },
+    {
+      "epoch": 0.8494681493353239,
+      "grad_norm": 1.46875,
+      "learning_rate": 2.75988419250775e-06,
+      "loss": 0.2853,
+      "step": 19346
+    },
+    {
+      "epoch": 0.8495559678145274,
+      "grad_norm": 1.53125,
+      "learning_rate": 2.7567252464990283e-06,
+      "loss": 0.3149,
+      "step": 19348
+    },
+    {
+      "epoch": 0.8496437862937308,
+      "grad_norm": 1.4453125,
+      "learning_rate": 2.753568003885623e-06,
+      "loss": 0.2996,
+      "step": 19350
+    },
+    {
+      "epoch": 0.8497316047729343,
+      "grad_norm": 1.421875,
+      "learning_rate": 2.750412464909305e-06,
+      "loss": 0.3227,
+      "step": 19352
+    },
+    {
+      "epoch": 0.8498194232521379,
+      "grad_norm": 1.46875,
+      "learning_rate": 2.7472586298117385e-06,
+      "loss": 0.311,
+      "step": 19354
+    },
+    {
+      "epoch": 0.8499072417313414,
+      "grad_norm": 1.4453125,
+      "learning_rate": 2.7441064988344424e-06,
+      "loss": 0.3119,
+      "step": 19356
+    },
+    {
+      "epoch": 0.8499950602105448,
+      "grad_norm": 1.671875,
+      "learning_rate": 2.740956072218806e-06,
+      "loss": 0.3566,
+      "step": 19358
+    },
+    {
+      "epoch": 0.8500828786897483,
+      "grad_norm": 1.453125,
+      "learning_rate": 2.737807350206084e-06,
+      "loss": 0.3247,
+      "step": 19360
+    },
+    {
+      "epoch": 0.8501706971689518,
+      "grad_norm": 1.4921875,
+      "learning_rate": 2.734660333037417e-06,
+      "loss": 0.3071,
+      "step": 19362
+    },
+    {
+      "epoch": 0.8502585156481552,
+      "grad_norm": 1.4375,
+      "learning_rate": 2.7315150209537984e-06,
+      "loss": 0.3137,
+      "step": 19364
+    },
+    {
+      "epoch": 0.8503463341273587,
+      "grad_norm": 1.5625,
+      "learning_rate": 2.7283714141960905e-06,
+      "loss": 0.3234,
+      "step": 19366
+    },
+    {
+      "epoch": 0.8504341526065622,
+      "grad_norm": 1.515625,
+      "learning_rate": 2.7252295130050483e-06,
+      "loss": 0.2924,
+      "step": 19368
+    },
+    {
+      "epoch": 0.8505219710857658,
+      "grad_norm": 1.4609375,
+      "learning_rate": 2.7220893176212595e-06,
+      "loss": 0.3097,
+      "step": 19370
+    },
+    {
+      "epoch": 0.8506097895649692,
+      "grad_norm": 1.4921875,
+      "learning_rate": 2.718950828285219e-06,
+      "loss": 0.2995,
+      "step": 19372
+    },
+    {
+      "epoch": 0.8506976080441727,
+      "grad_norm": 1.5078125,
+      "learning_rate": 2.7158140452372666e-06,
+      "loss": 0.3103,
+      "step": 19374
+    },
+    {
+      "epoch": 0.8507854265233762,
+      "grad_norm": 1.4765625,
+      "learning_rate": 2.712678968717608e-06,
+      "loss": 0.2902,
+      "step": 19376
+    },
+    {
+      "epoch": 0.8508732450025797,
+      "grad_norm": 1.484375,
+      "learning_rate": 2.7095455989663452e-06,
+      "loss": 0.315,
+      "step": 19378
+    },
+    {
+      "epoch": 0.8509610634817831,
+      "grad_norm": 1.5546875,
+      "learning_rate": 2.7064139362234174e-06,
+      "loss": 0.3209,
+      "step": 19380
+    },
+    {
+      "epoch": 0.8510488819609866,
+      "grad_norm": 1.59375,
+      "learning_rate": 2.7032839807286575e-06,
+      "loss": 0.3109,
+      "step": 19382
+    },
+    {
+      "epoch": 0.8511367004401901,
+      "grad_norm": 1.6640625,
+      "learning_rate": 2.700155732721751e-06,
+      "loss": 0.3309,
+      "step": 19384
+    },
+    {
+      "epoch": 0.8512245189193937,
+      "grad_norm": 1.515625,
+      "learning_rate": 2.6970291924422768e-06,
+      "loss": 0.2905,
+      "step": 19386
+    },
+    {
+      "epoch": 0.8513123373985971,
+      "grad_norm": 1.453125,
+      "learning_rate": 2.6939043601296425e-06,
+      "loss": 0.3297,
+      "step": 19388
+    },
+    {
+      "epoch": 0.8514001558778006,
+      "grad_norm": 1.625,
+      "learning_rate": 2.690781236023163e-06,
+      "loss": 0.347,
+      "step": 19390
+    },
+    {
+      "epoch": 0.8514879743570041,
+      "grad_norm": 1.4296875,
+      "learning_rate": 2.687659820362004e-06,
+      "loss": 0.2977,
+      "step": 19392
+    },
+    {
+      "epoch": 0.8515757928362075,
+      "grad_norm": 1.53125,
+      "learning_rate": 2.6845401133852005e-06,
+      "loss": 0.3306,
+      "step": 19394
+    },
+    {
+      "epoch": 0.851663611315411,
+      "grad_norm": 1.4375,
+      "learning_rate": 2.6814221153316694e-06,
+      "loss": 0.3197,
+      "step": 19396
+    },
+    {
+      "epoch": 0.8517514297946145,
+      "grad_norm": 1.5390625,
+      "learning_rate": 2.6783058264401776e-06,
+      "loss": 0.2946,
+      "step": 19398
+    },
+    {
+      "epoch": 0.8518392482738181,
+      "grad_norm": 1.5078125,
+      "learning_rate": 2.675191246949382e-06,
+      "loss": 0.3442,
+      "step": 19400
+    },
+    {
+      "epoch": 0.8519270667530215,
+      "grad_norm": 1.4296875,
+      "learning_rate": 2.672078377097792e-06,
+      "loss": 0.292,
+      "step": 19402
+    },
+    {
+      "epoch": 0.852014885232225,
+      "grad_norm": 1.4296875,
+      "learning_rate": 2.6689672171237885e-06,
+      "loss": 0.3111,
+      "step": 19404
+    },
+    {
+      "epoch": 0.8521027037114285,
+      "grad_norm": 1.515625,
+      "learning_rate": 2.6658577672656337e-06,
+      "loss": 0.2968,
+      "step": 19406
+    },
+    {
+      "epoch": 0.852190522190632,
+      "grad_norm": 1.4921875,
+      "learning_rate": 2.6627500277614376e-06,
+      "loss": 0.303,
+      "step": 19408
+    },
+    {
+      "epoch": 0.8522783406698354,
+      "grad_norm": 1.46875,
+      "learning_rate": 2.6596439988492065e-06,
+      "loss": 0.2896,
+      "step": 19410
+    },
+    {
+      "epoch": 0.8523661591490389,
+      "grad_norm": 1.546875,
+      "learning_rate": 2.6565396807667895e-06,
+      "loss": 0.3206,
+      "step": 19412
+    },
+    {
+      "epoch": 0.8524539776282424,
+      "grad_norm": 1.546875,
+      "learning_rate": 2.6534370737519188e-06,
+      "loss": 0.3197,
+      "step": 19414
+    },
+    {
+      "epoch": 0.852541796107446,
+      "grad_norm": 1.5,
+      "learning_rate": 2.6503361780421983e-06,
+      "loss": 0.2924,
+      "step": 19416
+    },
+    {
+      "epoch": 0.8526296145866494,
+      "grad_norm": 1.46875,
+      "learning_rate": 2.647236993875088e-06,
+      "loss": 0.3023,
+      "step": 19418
+    },
+    {
+      "epoch": 0.8527174330658529,
+      "grad_norm": 1.421875,
+      "learning_rate": 2.644139521487929e-06,
+      "loss": 0.316,
+      "step": 19420
+    },
+    {
+      "epoch": 0.8528052515450564,
+      "grad_norm": 1.5078125,
+      "learning_rate": 2.64104376111792e-06,
+      "loss": 0.3262,
+      "step": 19422
+    },
+    {
+      "epoch": 0.8528930700242598,
+      "grad_norm": 1.5078125,
+      "learning_rate": 2.6379497130021443e-06,
+      "loss": 0.322,
+      "step": 19424
+    },
+    {
+      "epoch": 0.8529808885034633,
+      "grad_norm": 1.609375,
+      "learning_rate": 2.6348573773775394e-06,
+      "loss": 0.31,
+      "step": 19426
+    },
+    {
+      "epoch": 0.8530687069826668,
+      "grad_norm": 1.4375,
+      "learning_rate": 2.6317667544809134e-06,
+      "loss": 0.3069,
+      "step": 19428
+    },
+    {
+      "epoch": 0.8531565254618703,
+      "grad_norm": 1.5859375,
+      "learning_rate": 2.6286778445489574e-06,
+      "loss": 0.3245,
+      "step": 19430
+    },
+    {
+      "epoch": 0.8532443439410738,
+      "grad_norm": 1.515625,
+      "learning_rate": 2.6255906478182075e-06,
+      "loss": 0.3098,
+      "step": 19432
+    },
+    {
+      "epoch": 0.8533321624202773,
+      "grad_norm": 1.5,
+      "learning_rate": 2.6225051645250965e-06,
+      "loss": 0.2982,
+      "step": 19434
+    },
+    {
+      "epoch": 0.8534199808994808,
+      "grad_norm": 1.484375,
+      "learning_rate": 2.619421394905902e-06,
+      "loss": 0.3077,
+      "step": 19436
+    },
+    {
+      "epoch": 0.8535077993786843,
+      "grad_norm": 1.515625,
+      "learning_rate": 2.61633933919678e-06,
+      "loss": 0.2876,
+      "step": 19438
+    },
+    {
+      "epoch": 0.8535956178578877,
+      "grad_norm": 1.4296875,
+      "learning_rate": 2.613258997633761e-06,
+      "loss": 0.3006,
+      "step": 19440
+    },
+    {
+      "epoch": 0.8536834363370912,
+      "grad_norm": 1.5234375,
+      "learning_rate": 2.610180370452728e-06,
+      "loss": 0.3323,
+      "step": 19442
+    },
+    {
+      "epoch": 0.8537712548162947,
+      "grad_norm": 1.484375,
+      "learning_rate": 2.6071034578894567e-06,
+      "loss": 0.3108,
+      "step": 19444
+    },
+    {
+      "epoch": 0.8538590732954981,
+      "grad_norm": 1.5546875,
+      "learning_rate": 2.6040282601795697e-06,
+      "loss": 0.3092,
+      "step": 19446
+    },
+    {
+      "epoch": 0.8539468917747017,
+      "grad_norm": 1.4609375,
+      "learning_rate": 2.6009547775585648e-06,
+      "loss": 0.3037,
+      "step": 19448
+    },
+    {
+      "epoch": 0.8540347102539052,
+      "grad_norm": 1.4765625,
+      "learning_rate": 2.597883010261809e-06,
+      "loss": 0.3239,
+      "step": 19450
+    },
+    {
+      "epoch": 0.8541225287331087,
+      "grad_norm": 1.453125,
+      "learning_rate": 2.594812958524548e-06,
+      "loss": 0.3089,
+      "step": 19452
+    },
+    {
+      "epoch": 0.8542103472123121,
+      "grad_norm": 1.4765625,
+      "learning_rate": 2.591744622581882e-06,
+      "loss": 0.2894,
+      "step": 19454
+    },
+    {
+      "epoch": 0.8542981656915156,
+      "grad_norm": 1.390625,
+      "learning_rate": 2.588678002668779e-06,
+      "loss": 0.2996,
+      "step": 19456
+    },
+    {
+      "epoch": 0.8543859841707191,
+      "grad_norm": 1.5546875,
+      "learning_rate": 2.585613099020093e-06,
+      "loss": 0.32,
+      "step": 19458
+    },
+    {
+      "epoch": 0.8544738026499226,
+      "grad_norm": 1.46875,
+      "learning_rate": 2.5825499118705213e-06,
+      "loss": 0.2837,
+      "step": 19460
+    },
+    {
+      "epoch": 0.8545616211291261,
+      "grad_norm": 1.4765625,
+      "learning_rate": 2.5794884414546584e-06,
+      "loss": 0.3122,
+      "step": 19462
+    },
+    {
+      "epoch": 0.8546494396083296,
+      "grad_norm": 1.6328125,
+      "learning_rate": 2.5764286880069434e-06,
+      "loss": 0.3227,
+      "step": 19464
+    },
+    {
+      "epoch": 0.8547372580875331,
+      "grad_norm": 1.4296875,
+      "learning_rate": 2.573370651761692e-06,
+      "loss": 0.3458,
+      "step": 19466
+    },
+    {
+      "epoch": 0.8548250765667366,
+      "grad_norm": 1.609375,
+      "learning_rate": 2.5703143329530943e-06,
+      "loss": 0.3311,
+      "step": 19468
+    },
+    {
+      "epoch": 0.85491289504594,
+      "grad_norm": 1.46875,
+      "learning_rate": 2.5672597318152e-06,
+      "loss": 0.2945,
+      "step": 19470
+    },
+    {
+      "epoch": 0.8550007135251435,
+      "grad_norm": 1.4609375,
+      "learning_rate": 2.5642068485819347e-06,
+      "loss": 0.3096,
+      "step": 19472
+    },
+    {
+      "epoch": 0.855088532004347,
+      "grad_norm": 1.5,
+      "learning_rate": 2.5611556834870905e-06,
+      "loss": 0.3302,
+      "step": 19474
+    },
+    {
+      "epoch": 0.8551763504835505,
+      "grad_norm": 1.53125,
+      "learning_rate": 2.5581062367643154e-06,
+      "loss": 0.3125,
+      "step": 19476
+    },
+    {
+      "epoch": 0.855264168962754,
+      "grad_norm": 1.71875,
+      "learning_rate": 2.555058508647154e-06,
+      "loss": 0.2861,
+      "step": 19478
+    },
+    {
+      "epoch": 0.8553519874419575,
+      "grad_norm": 1.4375,
+      "learning_rate": 2.5520124993689858e-06,
+      "loss": 0.3093,
+      "step": 19480
+    },
+    {
+      "epoch": 0.855439805921161,
+      "grad_norm": 1.578125,
+      "learning_rate": 2.548968209163083e-06,
+      "loss": 0.2958,
+      "step": 19482
+    },
+    {
+      "epoch": 0.8555276244003645,
+      "grad_norm": 1.4921875,
+      "learning_rate": 2.5459256382625723e-06,
+      "loss": 0.3501,
+      "step": 19484
+    },
+    {
+      "epoch": 0.8556154428795679,
+      "grad_norm": 1.5625,
+      "learning_rate": 2.542884786900465e-06,
+      "loss": 0.3296,
+      "step": 19486
+    },
+    {
+      "epoch": 0.8557032613587714,
+      "grad_norm": 1.5234375,
+      "learning_rate": 2.539845655309625e-06,
+      "loss": 0.3145,
+      "step": 19488
+    },
+    {
+      "epoch": 0.8557910798379749,
+      "grad_norm": 1.4375,
+      "learning_rate": 2.5368082437227825e-06,
+      "loss": 0.3172,
+      "step": 19490
+    },
+    {
+      "epoch": 0.8558788983171783,
+      "grad_norm": 1.5234375,
+      "learning_rate": 2.533772552372554e-06,
+      "loss": 0.3262,
+      "step": 19492
+    },
+    {
+      "epoch": 0.8559667167963819,
+      "grad_norm": 1.421875,
+      "learning_rate": 2.5307385814914043e-06,
+      "loss": 0.307,
+      "step": 19494
+    },
+    {
+      "epoch": 0.8560545352755854,
+      "grad_norm": 1.484375,
+      "learning_rate": 2.5277063313116854e-06,
+      "loss": 0.3223,
+      "step": 19496
+    },
+    {
+      "epoch": 0.8561423537547889,
+      "grad_norm": 1.4453125,
+      "learning_rate": 2.5246758020656013e-06,
+      "loss": 0.3303,
+      "step": 19498
+    },
+    {
+      "epoch": 0.8562301722339923,
+      "grad_norm": 1.484375,
+      "learning_rate": 2.52164699398523e-06,
+      "loss": 0.3126,
+      "step": 19500
+    },
+    {
+      "epoch": 0.8563179907131958,
+      "grad_norm": 1.5078125,
+      "learning_rate": 2.518619907302522e-06,
+      "loss": 0.2908,
+      "step": 19502
+    },
+    {
+      "epoch": 0.8564058091923993,
+      "grad_norm": 1.3984375,
+      "learning_rate": 2.515594542249289e-06,
+      "loss": 0.2945,
+      "step": 19504
+    },
+    {
+      "epoch": 0.8564936276716028,
+      "grad_norm": 1.5,
+      "learning_rate": 2.512570899057218e-06,
+      "loss": 0.3543,
+      "step": 19506
+    },
+    {
+      "epoch": 0.8565814461508063,
+      "grad_norm": 1.4765625,
+      "learning_rate": 2.5095489779578595e-06,
+      "loss": 0.3082,
+      "step": 19508
+    },
+    {
+      "epoch": 0.8566692646300098,
+      "grad_norm": 1.4609375,
+      "learning_rate": 2.506528779182632e-06,
+      "loss": 0.3024,
+      "step": 19510
+    },
+    {
+      "epoch": 0.8567570831092133,
+      "grad_norm": 1.5546875,
+      "learning_rate": 2.5035103029628166e-06,
+      "loss": 0.3181,
+      "step": 19512
+    },
+    {
+      "epoch": 0.8568449015884168,
+      "grad_norm": 1.4453125,
+      "learning_rate": 2.5004935495295813e-06,
+      "loss": 0.2928,
+      "step": 19514
+    },
+    {
+      "epoch": 0.8569327200676202,
+      "grad_norm": 1.453125,
+      "learning_rate": 2.497478519113941e-06,
+      "loss": 0.295,
+      "step": 19516
+    },
+    {
+      "epoch": 0.8570205385468237,
+      "grad_norm": 1.5234375,
+      "learning_rate": 2.4944652119467864e-06,
+      "loss": 0.3238,
+      "step": 19518
+    },
+    {
+      "epoch": 0.8571083570260272,
+      "grad_norm": 1.5390625,
+      "learning_rate": 2.4914536282588856e-06,
+      "loss": 0.3081,
+      "step": 19520
+    },
+    {
+      "epoch": 0.8571961755052306,
+      "grad_norm": 1.5,
+      "learning_rate": 2.488443768280857e-06,
+      "loss": 0.3123,
+      "step": 19522
+    },
+    {
+      "epoch": 0.8572839939844342,
+      "grad_norm": 1.4375,
+      "learning_rate": 2.4854356322432078e-06,
+      "loss": 0.3266,
+      "step": 19524
+    },
+    {
+      "epoch": 0.8573718124636377,
+      "grad_norm": 1.515625,
+      "learning_rate": 2.482429220376292e-06,
+      "loss": 0.3134,
+      "step": 19526
+    },
+    {
+      "epoch": 0.8574596309428412,
+      "grad_norm": 1.4296875,
+      "learning_rate": 2.4794245329103406e-06,
+      "loss": 0.3128,
+      "step": 19528
+    },
+    {
+      "epoch": 0.8575474494220446,
+      "grad_norm": 1.4765625,
+      "learning_rate": 2.476421570075463e-06,
+      "loss": 0.3205,
+      "step": 19530
+    },
+    {
+      "epoch": 0.8576352679012481,
+      "grad_norm": 1.421875,
+      "learning_rate": 2.4734203321016143e-06,
+      "loss": 0.3162,
+      "step": 19532
+    },
+    {
+      "epoch": 0.8577230863804516,
+      "grad_norm": 1.4609375,
+      "learning_rate": 2.470420819218641e-06,
+      "loss": 0.3417,
+      "step": 19534
+    },
+    {
+      "epoch": 0.857810904859655,
+      "grad_norm": 1.5078125,
+      "learning_rate": 2.4674230316562435e-06,
+      "loss": 0.3085,
+      "step": 19536
+    },
+    {
+      "epoch": 0.8578987233388585,
+      "grad_norm": 1.453125,
+      "learning_rate": 2.4644269696439847e-06,
+      "loss": 0.3028,
+      "step": 19538
+    },
+    {
+      "epoch": 0.8579865418180621,
+      "grad_norm": 1.4296875,
+      "learning_rate": 2.461432633411323e-06,
+      "loss": 0.3199,
+      "step": 19540
+    },
+    {
+      "epoch": 0.8580743602972656,
+      "grad_norm": 1.546875,
+      "learning_rate": 2.4584400231875416e-06,
+      "loss": 0.3034,
+      "step": 19542
+    },
+    {
+      "epoch": 0.858162178776469,
+      "grad_norm": 1.4921875,
+      "learning_rate": 2.4554491392018347e-06,
+      "loss": 0.3213,
+      "step": 19544
+    },
+    {
+      "epoch": 0.8582499972556725,
+      "grad_norm": 1.4765625,
+      "learning_rate": 2.4524599816832276e-06,
+      "loss": 0.3014,
+      "step": 19546
+    },
+    {
+      "epoch": 0.858337815734876,
+      "grad_norm": 1.4453125,
+      "learning_rate": 2.4494725508606488e-06,
+      "loss": 0.2933,
+      "step": 19548
+    },
+    {
+      "epoch": 0.8584256342140795,
+      "grad_norm": 1.4765625,
+      "learning_rate": 2.446486846962867e-06,
+      "loss": 0.3237,
+      "step": 19550
+    },
+    {
+      "epoch": 0.8585134526932829,
+      "grad_norm": 1.484375,
+      "learning_rate": 2.4435028702185227e-06,
+      "loss": 0.3021,
+      "step": 19552
+    },
+    {
+      "epoch": 0.8586012711724865,
+      "grad_norm": 1.4453125,
+      "learning_rate": 2.4405206208561403e-06,
+      "loss": 0.2859,
+      "step": 19554
+    },
+    {
+      "epoch": 0.85868908965169,
+      "grad_norm": 1.4296875,
+      "learning_rate": 2.437540099104094e-06,
+      "loss": 0.3134,
+      "step": 19556
+    },
+    {
+      "epoch": 0.8587769081308935,
+      "grad_norm": 1.5078125,
+      "learning_rate": 2.4345613051906384e-06,
+      "loss": 0.3277,
+      "step": 19558
+    },
+    {
+      "epoch": 0.8588647266100969,
+      "grad_norm": 1.453125,
+      "learning_rate": 2.431584239343887e-06,
+      "loss": 0.2996,
+      "step": 19560
+    },
+    {
+      "epoch": 0.8589525450893004,
+      "grad_norm": 1.4921875,
+      "learning_rate": 2.4286089017918233e-06,
+      "loss": 0.3341,
+      "step": 19562
+    },
+    {
+      "epoch": 0.8590403635685039,
+      "grad_norm": 1.484375,
+      "learning_rate": 2.4256352927623015e-06,
+      "loss": 0.2953,
+      "step": 19564
+    },
+    {
+      "epoch": 0.8591281820477074,
+      "grad_norm": 1.484375,
+      "learning_rate": 2.4226634124830366e-06,
+      "loss": 0.2946,
+      "step": 19566
+    },
+    {
+      "epoch": 0.8592160005269108,
+      "grad_norm": 1.40625,
+      "learning_rate": 2.419693261181627e-06,
+      "loss": 0.3183,
+      "step": 19568
+    },
+    {
+      "epoch": 0.8593038190061144,
+      "grad_norm": 1.4765625,
+      "learning_rate": 2.4167248390855187e-06,
+      "loss": 0.2981,
+      "step": 19570
+    },
+    {
+      "epoch": 0.8593916374853179,
+      "grad_norm": 1.4609375,
+      "learning_rate": 2.413758146422035e-06,
+      "loss": 0.3042,
+      "step": 19572
+    },
+    {
+      "epoch": 0.8594794559645214,
+      "grad_norm": 1.4921875,
+      "learning_rate": 2.4107931834183667e-06,
+      "loss": 0.321,
+      "step": 19574
+    },
+    {
+      "epoch": 0.8595672744437248,
+      "grad_norm": 1.515625,
+      "learning_rate": 2.4078299503015654e-06,
+      "loss": 0.3173,
+      "step": 19576
+    },
+    {
+      "epoch": 0.8596550929229283,
+      "grad_norm": 1.421875,
+      "learning_rate": 2.4048684472985694e-06,
+      "loss": 0.3045,
+      "step": 19578
+    },
+    {
+      "epoch": 0.8597429114021318,
+      "grad_norm": 1.59375,
+      "learning_rate": 2.401908674636158e-06,
+      "loss": 0.3217,
+      "step": 19580
+    },
+    {
+      "epoch": 0.8598307298813352,
+      "grad_norm": 1.484375,
+      "learning_rate": 2.398950632541003e-06,
+      "loss": 0.3452,
+      "step": 19582
+    },
+    {
+      "epoch": 0.8599185483605387,
+      "grad_norm": 1.40625,
+      "learning_rate": 2.3959943212396223e-06,
+      "loss": 0.312,
+      "step": 19584
+    },
+    {
+      "epoch": 0.8600063668397423,
+      "grad_norm": 1.4921875,
+      "learning_rate": 2.3930397409584194e-06,
+      "loss": 0.3037,
+      "step": 19586
+    },
+    {
+      "epoch": 0.8600941853189458,
+      "grad_norm": 1.3984375,
+      "learning_rate": 2.3900868919236515e-06,
+      "loss": 0.2944,
+      "step": 19588
+    },
+    {
+      "epoch": 0.8601820037981492,
+      "grad_norm": 1.515625,
+      "learning_rate": 2.387135774361446e-06,
+      "loss": 0.3224,
+      "step": 19590
+    },
+    {
+      "epoch": 0.8602698222773527,
+      "grad_norm": 1.5,
+      "learning_rate": 2.384186388497808e-06,
+      "loss": 0.3191,
+      "step": 19592
+    },
+    {
+      "epoch": 0.8603576407565562,
+      "grad_norm": 1.4921875,
+      "learning_rate": 2.3812387345585914e-06,
+      "loss": 0.311,
+      "step": 19594
+    },
+    {
+      "epoch": 0.8604454592357597,
+      "grad_norm": 1.515625,
+      "learning_rate": 2.378292812769542e-06,
+      "loss": 0.3169,
+      "step": 19596
+    },
+    {
+      "epoch": 0.8605332777149631,
+      "grad_norm": 1.4765625,
+      "learning_rate": 2.3753486233562496e-06,
+      "loss": 0.2885,
+      "step": 19598
+    },
+    {
+      "epoch": 0.8606210961941667,
+      "grad_norm": 1.515625,
+      "learning_rate": 2.3724061665441777e-06,
+      "loss": 0.33,
+      "step": 19600
+    },
+    {
+      "epoch": 0.8607089146733702,
+      "grad_norm": 1.46875,
+      "learning_rate": 2.3694654425586772e-06,
+      "loss": 0.2909,
+      "step": 19602
+    },
+    {
+      "epoch": 0.8607967331525737,
+      "grad_norm": 1.53125,
+      "learning_rate": 2.3665264516249287e-06,
+      "loss": 0.292,
+      "step": 19604
+    },
+    {
+      "epoch": 0.8608845516317771,
+      "grad_norm": 1.4453125,
+      "learning_rate": 2.363589193968013e-06,
+      "loss": 0.2853,
+      "step": 19606
+    },
+    {
+      "epoch": 0.8609723701109806,
+      "grad_norm": 1.5859375,
+      "learning_rate": 2.3606536698128595e-06,
+      "loss": 0.2869,
+      "step": 19608
+    },
+    {
+      "epoch": 0.8610601885901841,
+      "grad_norm": 1.4765625,
+      "learning_rate": 2.357719879384279e-06,
+      "loss": 0.2986,
+      "step": 19610
+    },
+    {
+      "epoch": 0.8611480070693875,
+      "grad_norm": 1.375,
+      "learning_rate": 2.3547878229069397e-06,
+      "loss": 0.2786,
+      "step": 19612
+    },
+    {
+      "epoch": 0.861235825548591,
+      "grad_norm": 1.484375,
+      "learning_rate": 2.35185750060537e-06,
+      "loss": 0.2973,
+      "step": 19614
+    },
+    {
+      "epoch": 0.8613236440277946,
+      "grad_norm": 1.5,
+      "learning_rate": 2.3489289127039906e-06,
+      "loss": 0.3174,
+      "step": 19616
+    },
+    {
+      "epoch": 0.8614114625069981,
+      "grad_norm": 1.5234375,
+      "learning_rate": 2.3460020594270575e-06,
+      "loss": 0.3406,
+      "step": 19618
+    },
+    {
+      "epoch": 0.8614992809862015,
+      "grad_norm": 1.5390625,
+      "learning_rate": 2.343076940998723e-06,
+      "loss": 0.3065,
+      "step": 19620
+    },
+    {
+      "epoch": 0.861587099465405,
+      "grad_norm": 1.4609375,
+      "learning_rate": 2.340153557642985e-06,
+      "loss": 0.335,
+      "step": 19622
+    },
+    {
+      "epoch": 0.8616749179446085,
+      "grad_norm": 1.5546875,
+      "learning_rate": 2.3372319095837174e-06,
+      "loss": 0.2874,
+      "step": 19624
+    },
+    {
+      "epoch": 0.861762736423812,
+      "grad_norm": 1.4765625,
+      "learning_rate": 2.334311997044666e-06,
+      "loss": 0.2934,
+      "step": 19626
+    },
+    {
+      "epoch": 0.8618505549030154,
+      "grad_norm": 1.453125,
+      "learning_rate": 2.3313938202494307e-06,
+      "loss": 0.3114,
+      "step": 19628
+    },
+    {
+      "epoch": 0.8619383733822189,
+      "grad_norm": 1.53125,
+      "learning_rate": 2.328477379421498e-06,
+      "loss": 0.3105,
+      "step": 19630
+    },
+    {
+      "epoch": 0.8620261918614225,
+      "grad_norm": 1.5234375,
+      "learning_rate": 2.3255626747841985e-06,
+      "loss": 0.319,
+      "step": 19632
+    },
+    {
+      "epoch": 0.862114010340626,
+      "grad_norm": 1.4296875,
+      "learning_rate": 2.3226497065607478e-06,
+      "loss": 0.2853,
+      "step": 19634
+    },
+    {
+      "epoch": 0.8622018288198294,
+      "grad_norm": 1.46875,
+      "learning_rate": 2.3197384749742175e-06,
+      "loss": 0.3208,
+      "step": 19636
+    },
+    {
+      "epoch": 0.8622896472990329,
+      "grad_norm": 1.484375,
+      "learning_rate": 2.316828980247546e-06,
+      "loss": 0.3086,
+      "step": 19638
+    },
+    {
+      "epoch": 0.8623774657782364,
+      "grad_norm": 1.4375,
+      "learning_rate": 2.3139212226035543e-06,
+      "loss": 0.2994,
+      "step": 19640
+    },
+    {
+      "epoch": 0.8624652842574398,
+      "grad_norm": 1.53125,
+      "learning_rate": 2.3110152022649103e-06,
+      "loss": 0.3229,
+      "step": 19642
+    },
+    {
+      "epoch": 0.8625531027366433,
+      "grad_norm": 1.5390625,
+      "learning_rate": 2.308110919454165e-06,
+      "loss": 0.3214,
+      "step": 19644
+    },
+    {
+      "epoch": 0.8626409212158468,
+      "grad_norm": 1.515625,
+      "learning_rate": 2.305208374393719e-06,
+      "loss": 0.3302,
+      "step": 19646
+    },
+    {
+      "epoch": 0.8627287396950504,
+      "grad_norm": 1.4765625,
+      "learning_rate": 2.3023075673058615e-06,
+      "loss": 0.3134,
+      "step": 19648
+    },
+    {
+      "epoch": 0.8628165581742538,
+      "grad_norm": 1.4609375,
+      "learning_rate": 2.299408498412731e-06,
+      "loss": 0.319,
+      "step": 19650
+    },
+    {
+      "epoch": 0.8629043766534573,
+      "grad_norm": 1.5234375,
+      "learning_rate": 2.2965111679363355e-06,
+      "loss": 0.2917,
+      "step": 19652
+    },
+    {
+      "epoch": 0.8629921951326608,
+      "grad_norm": 1.703125,
+      "learning_rate": 2.293615576098562e-06,
+      "loss": 0.3199,
+      "step": 19654
+    },
+    {
+      "epoch": 0.8630800136118643,
+      "grad_norm": 1.4296875,
+      "learning_rate": 2.290721723121145e-06,
+      "loss": 0.3266,
+      "step": 19656
+    },
+    {
+      "epoch": 0.8631678320910677,
+      "grad_norm": 1.5,
+      "learning_rate": 2.2878296092257085e-06,
+      "loss": 0.323,
+      "step": 19658
+    },
+    {
+      "epoch": 0.8632556505702712,
+      "grad_norm": 1.390625,
+      "learning_rate": 2.2849392346337235e-06,
+      "loss": 0.328,
+      "step": 19660
+    },
+    {
+      "epoch": 0.8633434690494748,
+      "grad_norm": 1.4765625,
+      "learning_rate": 2.2820505995665378e-06,
+      "loss": 0.3261,
+      "step": 19662
+    },
+    {
+      "epoch": 0.8634312875286783,
+      "grad_norm": 1.453125,
+      "learning_rate": 2.2791637042453628e-06,
+      "loss": 0.3211,
+      "step": 19664
+    },
+    {
+      "epoch": 0.8635191060078817,
+      "grad_norm": 1.5390625,
+      "learning_rate": 2.2762785488912724e-06,
+      "loss": 0.3188,
+      "step": 19666
+    },
+    {
+      "epoch": 0.8636069244870852,
+      "grad_norm": 1.4609375,
+      "learning_rate": 2.273395133725223e-06,
+      "loss": 0.3157,
+      "step": 19668
+    },
+    {
+      "epoch": 0.8636947429662887,
+      "grad_norm": 1.359375,
+      "learning_rate": 2.270513458968018e-06,
+      "loss": 0.3233,
+      "step": 19670
+    },
+    {
+      "epoch": 0.8637825614454921,
+      "grad_norm": 1.3984375,
+      "learning_rate": 2.267633524840346e-06,
+      "loss": 0.3448,
+      "step": 19672
+    },
+    {
+      "epoch": 0.8638703799246956,
+      "grad_norm": 1.484375,
+      "learning_rate": 2.2647553315627485e-06,
+      "loss": 0.3084,
+      "step": 19674
+    },
+    {
+      "epoch": 0.8639581984038991,
+      "grad_norm": 1.578125,
+      "learning_rate": 2.261878879355633e-06,
+      "loss": 0.297,
+      "step": 19676
+    },
+    {
+      "epoch": 0.8640460168831027,
+      "grad_norm": 1.5625,
+      "learning_rate": 2.2590041684392854e-06,
+      "loss": 0.3149,
+      "step": 19678
+    },
+    {
+      "epoch": 0.8641338353623061,
+      "grad_norm": 1.4609375,
+      "learning_rate": 2.256131199033848e-06,
+      "loss": 0.3068,
+      "step": 19680
+    },
+    {
+      "epoch": 0.8642216538415096,
+      "grad_norm": 1.5390625,
+      "learning_rate": 2.2532599713593396e-06,
+      "loss": 0.3197,
+      "step": 19682
+    },
+    {
+      "epoch": 0.8643094723207131,
+      "grad_norm": 1.4609375,
+      "learning_rate": 2.2503904856356366e-06,
+      "loss": 0.3218,
+      "step": 19684
+    },
+    {
+      "epoch": 0.8643972907999166,
+      "grad_norm": 1.3984375,
+      "learning_rate": 2.247522742082478e-06,
+      "loss": 0.3111,
+      "step": 19686
+    },
+    {
+      "epoch": 0.86448510927912,
+      "grad_norm": 1.421875,
+      "learning_rate": 2.244656740919485e-06,
+      "loss": 0.3329,
+      "step": 19688
+    },
+    {
+      "epoch": 0.8645729277583235,
+      "grad_norm": 1.4921875,
+      "learning_rate": 2.241792482366131e-06,
+      "loss": 0.3249,
+      "step": 19690
+    },
+    {
+      "epoch": 0.864660746237527,
+      "grad_norm": 1.4765625,
+      "learning_rate": 2.238929966641767e-06,
+      "loss": 0.3023,
+      "step": 19692
+    },
+    {
+      "epoch": 0.8647485647167306,
+      "grad_norm": 1.453125,
+      "learning_rate": 2.2360691939656036e-06,
+      "loss": 0.3127,
+      "step": 19694
+    },
+    {
+      "epoch": 0.864836383195934,
+      "grad_norm": 1.4609375,
+      "learning_rate": 2.2332101645567176e-06,
+      "loss": 0.3544,
+      "step": 19696
+    },
+    {
+      "epoch": 0.8649242016751375,
+      "grad_norm": 1.4375,
+      "learning_rate": 2.2303528786340565e-06,
+      "loss": 0.31,
+      "step": 19698
+    },
+    {
+      "epoch": 0.865012020154341,
+      "grad_norm": 1.421875,
+      "learning_rate": 2.227497336416423e-06,
+      "loss": 0.3246,
+      "step": 19700
+    },
+    {
+      "epoch": 0.8650998386335444,
+      "grad_norm": 1.5234375,
+      "learning_rate": 2.224643538122509e-06,
+      "loss": 0.3037,
+      "step": 19702
+    },
+    {
+      "epoch": 0.8651876571127479,
+      "grad_norm": 1.4375,
+      "learning_rate": 2.2217914839708477e-06,
+      "loss": 0.3307,
+      "step": 19704
+    },
+    {
+      "epoch": 0.8652754755919514,
+      "grad_norm": 1.4453125,
+      "learning_rate": 2.2189411741798587e-06,
+      "loss": 0.3167,
+      "step": 19706
+    },
+    {
+      "epoch": 0.865363294071155,
+      "grad_norm": 1.53125,
+      "learning_rate": 2.216092608967818e-06,
+      "loss": 0.3497,
+      "step": 19708
+    },
+    {
+      "epoch": 0.8654511125503584,
+      "grad_norm": 1.46875,
+      "learning_rate": 2.213245788552859e-06,
+      "loss": 0.3256,
+      "step": 19710
+    },
+    {
+      "epoch": 0.8655389310295619,
+      "grad_norm": 1.4375,
+      "learning_rate": 2.2104007131530046e-06,
+      "loss": 0.2998,
+      "step": 19712
+    },
+    {
+      "epoch": 0.8656267495087654,
+      "grad_norm": 1.421875,
+      "learning_rate": 2.207557382986125e-06,
+      "loss": 0.3256,
+      "step": 19714
+    },
+    {
+      "epoch": 0.8657145679879689,
+      "grad_norm": 1.5546875,
+      "learning_rate": 2.2047157982699686e-06,
+      "loss": 0.3113,
+      "step": 19716
+    },
+    {
+      "epoch": 0.8658023864671723,
+      "grad_norm": 1.4453125,
+      "learning_rate": 2.201875959222133e-06,
+      "loss": 0.2938,
+      "step": 19718
+    },
+    {
+      "epoch": 0.8658902049463758,
+      "grad_norm": 1.5546875,
+      "learning_rate": 2.199037866060108e-06,
+      "loss": 0.3194,
+      "step": 19720
+    },
+    {
+      "epoch": 0.8659780234255793,
+      "grad_norm": 1.4765625,
+      "learning_rate": 2.1962015190012315e-06,
+      "loss": 0.2859,
+      "step": 19722
+    },
+    {
+      "epoch": 0.8660658419047829,
+      "grad_norm": 1.4453125,
+      "learning_rate": 2.1933669182627044e-06,
+      "loss": 0.3604,
+      "step": 19724
+    },
+    {
+      "epoch": 0.8661536603839863,
+      "grad_norm": 1.46875,
+      "learning_rate": 2.1905340640616085e-06,
+      "loss": 0.2944,
+      "step": 19726
+    },
+    {
+      "epoch": 0.8662414788631898,
+      "grad_norm": 1.390625,
+      "learning_rate": 2.187702956614879e-06,
+      "loss": 0.2785,
+      "step": 19728
+    },
+    {
+      "epoch": 0.8663292973423933,
+      "grad_norm": 1.515625,
+      "learning_rate": 2.184873596139328e-06,
+      "loss": 0.3243,
+      "step": 19730
+    },
+    {
+      "epoch": 0.8664171158215968,
+      "grad_norm": 1.5,
+      "learning_rate": 2.1820459828516215e-06,
+      "loss": 0.3116,
+      "step": 19732
+    },
+    {
+      "epoch": 0.8665049343008002,
+      "grad_norm": 1.453125,
+      "learning_rate": 2.1792201169683085e-06,
+      "loss": 0.3112,
+      "step": 19734
+    },
+    {
+      "epoch": 0.8665927527800037,
+      "grad_norm": 1.390625,
+      "learning_rate": 2.1763959987057882e-06,
+      "loss": 0.3129,
+      "step": 19736
+    },
+    {
+      "epoch": 0.8666805712592072,
+      "grad_norm": 1.40625,
+      "learning_rate": 2.1735736282803314e-06,
+      "loss": 0.3219,
+      "step": 19738
+    },
+    {
+      "epoch": 0.8667683897384107,
+      "grad_norm": 1.4609375,
+      "learning_rate": 2.17075300590808e-06,
+      "loss": 0.295,
+      "step": 19740
+    },
+    {
+      "epoch": 0.8668562082176142,
+      "grad_norm": 1.4375,
+      "learning_rate": 2.1679341318050327e-06,
+      "loss": 0.2926,
+      "step": 19742
+    },
+    {
+      "epoch": 0.8669440266968177,
+      "grad_norm": 1.515625,
+      "learning_rate": 2.1651170061870675e-06,
+      "loss": 0.3007,
+      "step": 19744
+    },
+    {
+      "epoch": 0.8670318451760212,
+      "grad_norm": 1.46875,
+      "learning_rate": 2.1623016292699137e-06,
+      "loss": 0.3406,
+      "step": 19746
+    },
+    {
+      "epoch": 0.8671196636552246,
+      "grad_norm": 1.4296875,
+      "learning_rate": 2.159488001269175e-06,
+      "loss": 0.3446,
+      "step": 19748
+    },
+    {
+      "epoch": 0.8672074821344281,
+      "grad_norm": 1.515625,
+      "learning_rate": 2.156676122400322e-06,
+      "loss": 0.3072,
+      "step": 19750
+    },
+    {
+      "epoch": 0.8672953006136316,
+      "grad_norm": 1.4765625,
+      "learning_rate": 2.153865992878684e-06,
+      "loss": 0.3402,
+      "step": 19752
+    },
+    {
+      "epoch": 0.8673831190928352,
+      "grad_norm": 1.4375,
+      "learning_rate": 2.151057612919474e-06,
+      "loss": 0.2902,
+      "step": 19754
+    },
+    {
+      "epoch": 0.8674709375720386,
+      "grad_norm": 1.4921875,
+      "learning_rate": 2.14825098273774e-06,
+      "loss": 0.286,
+      "step": 19756
+    },
+    {
+      "epoch": 0.8675587560512421,
+      "grad_norm": 1.5390625,
+      "learning_rate": 2.1454461025484292e-06,
+      "loss": 0.2923,
+      "step": 19758
+    },
+    {
+      "epoch": 0.8676465745304456,
+      "grad_norm": 1.5546875,
+      "learning_rate": 2.1426429725663335e-06,
+      "loss": 0.3102,
+      "step": 19760
+    },
+    {
+      "epoch": 0.867734393009649,
+      "grad_norm": 1.515625,
+      "learning_rate": 2.1398415930061146e-06,
+      "loss": 0.3185,
+      "step": 19762
+    },
+    {
+      "epoch": 0.8678222114888525,
+      "grad_norm": 1.3984375,
+      "learning_rate": 2.1370419640823094e-06,
+      "loss": 0.305,
+      "step": 19764
+    },
+    {
+      "epoch": 0.867910029968056,
+      "grad_norm": 1.4296875,
+      "learning_rate": 2.134244086009307e-06,
+      "loss": 0.3049,
+      "step": 19766
+    },
+    {
+      "epoch": 0.8679978484472595,
+      "grad_norm": 1.5078125,
+      "learning_rate": 2.1314479590013786e-06,
+      "loss": 0.3173,
+      "step": 19768
+    },
+    {
+      "epoch": 0.868085666926463,
+      "grad_norm": 1.46875,
+      "learning_rate": 2.1286535832726467e-06,
+      "loss": 0.3196,
+      "step": 19770
+    },
+    {
+      "epoch": 0.8681734854056665,
+      "grad_norm": 1.5390625,
+      "learning_rate": 2.125860959037104e-06,
+      "loss": 0.3416,
+      "step": 19772
+    },
+    {
+      "epoch": 0.86826130388487,
+      "grad_norm": 1.5234375,
+      "learning_rate": 2.1230700865086134e-06,
+      "loss": 0.3315,
+      "step": 19774
+    },
+    {
+      "epoch": 0.8683491223640735,
+      "grad_norm": 1.4375,
+      "learning_rate": 2.1202809659008976e-06,
+      "loss": 0.317,
+      "step": 19776
+    },
+    {
+      "epoch": 0.8684369408432769,
+      "grad_norm": 1.390625,
+      "learning_rate": 2.1174935974275556e-06,
+      "loss": 0.3034,
+      "step": 19778
+    },
+    {
+      "epoch": 0.8685247593224804,
+      "grad_norm": 1.453125,
+      "learning_rate": 2.114707981302033e-06,
+      "loss": 0.3161,
+      "step": 19780
+    },
+    {
+      "epoch": 0.8686125778016839,
+      "grad_norm": 1.546875,
+      "learning_rate": 2.1119241177376626e-06,
+      "loss": 0.2952,
+      "step": 19782
+    },
+    {
+      "epoch": 0.8687003962808874,
+      "grad_norm": 1.4609375,
+      "learning_rate": 2.1091420069476316e-06,
+      "loss": 0.3344,
+      "step": 19784
+    },
+    {
+      "epoch": 0.8687882147600909,
+      "grad_norm": 1.4609375,
+      "learning_rate": 2.1063616491449917e-06,
+      "loss": 0.3011,
+      "step": 19786
+    },
+    {
+      "epoch": 0.8688760332392944,
+      "grad_norm": 1.4375,
+      "learning_rate": 2.1035830445426648e-06,
+      "loss": 0.3266,
+      "step": 19788
+    },
+    {
+      "epoch": 0.8689638517184979,
+      "grad_norm": 1.4296875,
+      "learning_rate": 2.1008061933534327e-06,
+      "loss": 0.3156,
+      "step": 19790
+    },
+    {
+      "epoch": 0.8690516701977014,
+      "grad_norm": 1.546875,
+      "learning_rate": 2.098031095789957e-06,
+      "loss": 0.3302,
+      "step": 19792
+    },
+    {
+      "epoch": 0.8691394886769048,
+      "grad_norm": 1.5078125,
+      "learning_rate": 2.095257752064747e-06,
+      "loss": 0.309,
+      "step": 19794
+    },
+    {
+      "epoch": 0.8692273071561083,
+      "grad_norm": 1.5703125,
+      "learning_rate": 2.092486162390192e-06,
+      "loss": 0.3126,
+      "step": 19796
+    },
+    {
+      "epoch": 0.8693151256353118,
+      "grad_norm": 1.5390625,
+      "learning_rate": 2.089716326978536e-06,
+      "loss": 0.3339,
+      "step": 19798
+    },
+    {
+      "epoch": 0.8694029441145152,
+      "grad_norm": 1.4765625,
+      "learning_rate": 2.086948246041892e-06,
+      "loss": 0.3164,
+      "step": 19800
+    },
+    {
+      "epoch": 0.8694907625937188,
+      "grad_norm": 1.484375,
+      "learning_rate": 2.0841819197922494e-06,
+      "loss": 0.3027,
+      "step": 19802
+    },
+    {
+      "epoch": 0.8695785810729223,
+      "grad_norm": 1.46875,
+      "learning_rate": 2.0814173484414447e-06,
+      "loss": 0.3284,
+      "step": 19804
+    },
+    {
+      "epoch": 0.8696663995521258,
+      "grad_norm": 1.4453125,
+      "learning_rate": 2.0786545322011965e-06,
+      "loss": 0.3101,
+      "step": 19806
+    },
+    {
+      "epoch": 0.8697542180313292,
+      "grad_norm": 1.5703125,
+      "learning_rate": 2.07589347128308e-06,
+      "loss": 0.3168,
+      "step": 19808
+    },
+    {
+      "epoch": 0.8698420365105327,
+      "grad_norm": 1.5703125,
+      "learning_rate": 2.0731341658985324e-06,
+      "loss": 0.3165,
+      "step": 19810
+    },
+    {
+      "epoch": 0.8699298549897362,
+      "grad_norm": 1.5078125,
+      "learning_rate": 2.0703766162588706e-06,
+      "loss": 0.3176,
+      "step": 19812
+    },
+    {
+      "epoch": 0.8700176734689397,
+      "grad_norm": 1.4765625,
+      "learning_rate": 2.0676208225752585e-06,
+      "loss": 0.3196,
+      "step": 19814
+    },
+    {
+      "epoch": 0.8701054919481432,
+      "grad_norm": 1.515625,
+      "learning_rate": 2.06486678505875e-06,
+      "loss": 0.3524,
+      "step": 19816
+    },
+    {
+      "epoch": 0.8701933104273467,
+      "grad_norm": 1.4609375,
+      "learning_rate": 2.0621145039202343e-06,
+      "loss": 0.2923,
+      "step": 19818
+    },
+    {
+      "epoch": 0.8702811289065502,
+      "grad_norm": 1.4140625,
+      "learning_rate": 2.0593639793704905e-06,
+      "loss": 0.3128,
+      "step": 19820
+    },
+    {
+      "epoch": 0.8703689473857537,
+      "grad_norm": 1.4921875,
+      "learning_rate": 2.0566152116201528e-06,
+      "loss": 0.3216,
+      "step": 19822
+    },
+    {
+      "epoch": 0.8704567658649571,
+      "grad_norm": 1.4375,
+      "learning_rate": 2.053868200879719e-06,
+      "loss": 0.2935,
+      "step": 19824
+    },
+    {
+      "epoch": 0.8705445843441606,
+      "grad_norm": 1.4765625,
+      "learning_rate": 2.0511229473595626e-06,
+      "loss": 0.3261,
+      "step": 19826
+    },
+    {
+      "epoch": 0.8706324028233641,
+      "grad_norm": 1.5546875,
+      "learning_rate": 2.0483794512699077e-06,
+      "loss": 0.305,
+      "step": 19828
+    },
+    {
+      "epoch": 0.8707202213025675,
+      "grad_norm": 1.421875,
+      "learning_rate": 2.045637712820861e-06,
+      "loss": 0.3249,
+      "step": 19830
+    },
+    {
+      "epoch": 0.8708080397817711,
+      "grad_norm": 1.453125,
+      "learning_rate": 2.042897732222379e-06,
+      "loss": 0.3237,
+      "step": 19832
+    },
+    {
+      "epoch": 0.8708958582609746,
+      "grad_norm": 1.4140625,
+      "learning_rate": 2.040159509684289e-06,
+      "loss": 0.3407,
+      "step": 19834
+    },
+    {
+      "epoch": 0.8709836767401781,
+      "grad_norm": 1.4765625,
+      "learning_rate": 2.0374230454162924e-06,
+      "loss": 0.3136,
+      "step": 19836
+    },
+    {
+      "epoch": 0.8710714952193815,
+      "grad_norm": 1.4453125,
+      "learning_rate": 2.034688339627938e-06,
+      "loss": 0.3151,
+      "step": 19838
+    },
+    {
+      "epoch": 0.871159313698585,
+      "grad_norm": 1.390625,
+      "learning_rate": 2.031955392528662e-06,
+      "loss": 0.3123,
+      "step": 19840
+    },
+    {
+      "epoch": 0.8712471321777885,
+      "grad_norm": 1.5234375,
+      "learning_rate": 2.0292242043277455e-06,
+      "loss": 0.3193,
+      "step": 19842
+    },
+    {
+      "epoch": 0.871334950656992,
+      "grad_norm": 1.484375,
+      "learning_rate": 2.0264947752343446e-06,
+      "loss": 0.2988,
+      "step": 19844
+    },
+    {
+      "epoch": 0.8714227691361954,
+      "grad_norm": 1.5,
+      "learning_rate": 2.023767105457483e-06,
+      "loss": 0.3288,
+      "step": 19846
+    },
+    {
+      "epoch": 0.871510587615399,
+      "grad_norm": 1.4453125,
+      "learning_rate": 2.0210411952060444e-06,
+      "loss": 0.2927,
+      "step": 19848
+    },
+    {
+      "epoch": 0.8715984060946025,
+      "grad_norm": 1.53125,
+      "learning_rate": 2.0183170446887807e-06,
+      "loss": 0.2949,
+      "step": 19850
+    },
+    {
+      "epoch": 0.871686224573806,
+      "grad_norm": 1.4609375,
+      "learning_rate": 2.0155946541143e-06,
+      "loss": 0.3076,
+      "step": 19852
+    },
+    {
+      "epoch": 0.8717740430530094,
+      "grad_norm": 1.6796875,
+      "learning_rate": 2.012874023691097e-06,
+      "loss": 0.3349,
+      "step": 19854
+    },
+    {
+      "epoch": 0.8718618615322129,
+      "grad_norm": 1.5703125,
+      "learning_rate": 2.010155153627505e-06,
+      "loss": 0.3102,
+      "step": 19856
+    },
+    {
+      "epoch": 0.8719496800114164,
+      "grad_norm": 1.5,
+      "learning_rate": 2.0074380441317488e-06,
+      "loss": 0.3419,
+      "step": 19858
+    },
+    {
+      "epoch": 0.8720374984906198,
+      "grad_norm": 1.4375,
+      "learning_rate": 2.004722695411898e-06,
+      "loss": 0.3193,
+      "step": 19860
+    },
+    {
+      "epoch": 0.8721253169698234,
+      "grad_norm": 1.484375,
+      "learning_rate": 2.0020091076758922e-06,
+      "loss": 0.3176,
+      "step": 19862
+    },
+    {
+      "epoch": 0.8722131354490269,
+      "grad_norm": 1.421875,
+      "learning_rate": 1.9992972811315458e-06,
+      "loss": 0.2944,
+      "step": 19864
+    },
+    {
+      "epoch": 0.8723009539282304,
+      "grad_norm": 1.421875,
+      "learning_rate": 1.9965872159865227e-06,
+      "loss": 0.3135,
+      "step": 19866
+    },
+    {
+      "epoch": 0.8723887724074338,
+      "grad_norm": 1.4609375,
+      "learning_rate": 1.9938789124483687e-06,
+      "loss": 0.3138,
+      "step": 19868
+    },
+    {
+      "epoch": 0.8724765908866373,
+      "grad_norm": 1.5078125,
+      "learning_rate": 1.991172370724484e-06,
+      "loss": 0.322,
+      "step": 19870
+    },
+    {
+      "epoch": 0.8725644093658408,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.988467591022131e-06,
+      "loss": 0.2898,
+      "step": 19872
+    },
+    {
+      "epoch": 0.8726522278450443,
+      "grad_norm": 1.46875,
+      "learning_rate": 1.9857645735484486e-06,
+      "loss": 0.3098,
+      "step": 19874
+    },
+    {
+      "epoch": 0.8727400463242477,
+      "grad_norm": 1.421875,
+      "learning_rate": 1.9830633185104282e-06,
+      "loss": 0.3381,
+      "step": 19876
+    },
+    {
+      "epoch": 0.8728278648034513,
+      "grad_norm": 1.4609375,
+      "learning_rate": 1.980363826114945e-06,
+      "loss": 0.2921,
+      "step": 19878
+    },
+    {
+      "epoch": 0.8729156832826548,
+      "grad_norm": 1.5078125,
+      "learning_rate": 1.977666096568709e-06,
+      "loss": 0.2974,
+      "step": 19880
+    },
+    {
+      "epoch": 0.8730035017618583,
+      "grad_norm": 1.4921875,
+      "learning_rate": 1.974970130078327e-06,
+      "loss": 0.3526,
+      "step": 19882
+    },
+    {
+      "epoch": 0.8730913202410617,
+      "grad_norm": 1.4453125,
+      "learning_rate": 1.9722759268502507e-06,
+      "loss": 0.3244,
+      "step": 19884
+    },
+    {
+      "epoch": 0.8731791387202652,
+      "grad_norm": 1.4375,
+      "learning_rate": 1.9695834870908007e-06,
+      "loss": 0.3194,
+      "step": 19886
+    },
+    {
+      "epoch": 0.8732669571994687,
+      "grad_norm": 1.5,
+      "learning_rate": 1.966892811006174e-06,
+      "loss": 0.3328,
+      "step": 19888
+    },
+    {
+      "epoch": 0.8733547756786721,
+      "grad_norm": 1.4453125,
+      "learning_rate": 1.964203898802411e-06,
+      "loss": 0.3276,
+      "step": 19890
+    },
+    {
+      "epoch": 0.8734425941578756,
+      "grad_norm": 1.46875,
+      "learning_rate": 1.961516750685441e-06,
+      "loss": 0.2992,
+      "step": 19892
+    },
+    {
+      "epoch": 0.8735304126370792,
+      "grad_norm": 1.421875,
+      "learning_rate": 1.9588313668610385e-06,
+      "loss": 0.3253,
+      "step": 19894
+    },
+    {
+      "epoch": 0.8736182311162827,
+      "grad_norm": 1.4921875,
+      "learning_rate": 1.9561477475348504e-06,
+      "loss": 0.329,
+      "step": 19896
+    },
+    {
+      "epoch": 0.8737060495954861,
+      "grad_norm": 1.46875,
+      "learning_rate": 1.9534658929123984e-06,
+      "loss": 0.2999,
+      "step": 19898
+    },
+    {
+      "epoch": 0.8737938680746896,
+      "grad_norm": 1.4609375,
+      "learning_rate": 1.9507858031990454e-06,
+      "loss": 0.3202,
+      "step": 19900
+    },
+    {
+      "epoch": 0.8738816865538931,
+      "grad_norm": 1.4296875,
+      "learning_rate": 1.9481074786000473e-06,
+      "loss": 0.2992,
+      "step": 19902
+    },
+    {
+      "epoch": 0.8739695050330966,
+      "grad_norm": 1.5078125,
+      "learning_rate": 1.945430919320504e-06,
+      "loss": 0.3101,
+      "step": 19904
+    },
+    {
+      "epoch": 0.8740573235123,
+      "grad_norm": 1.40625,
+      "learning_rate": 1.9427561255653816e-06,
+      "loss": 0.3208,
+      "step": 19906
+    },
+    {
+      "epoch": 0.8741451419915036,
+      "grad_norm": 1.4296875,
+      "learning_rate": 1.9400830975395272e-06,
+      "loss": 0.3327,
+      "step": 19908
+    },
+    {
+      "epoch": 0.8742329604707071,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.937411835447639e-06,
+      "loss": 0.3426,
+      "step": 19910
+    },
+    {
+      "epoch": 0.8743207789499106,
+      "grad_norm": 1.3984375,
+      "learning_rate": 1.9347423394942774e-06,
+      "loss": 0.299,
+      "step": 19912
+    },
+    {
+      "epoch": 0.874408597429114,
+      "grad_norm": 1.4765625,
+      "learning_rate": 1.9320746098838744e-06,
+      "loss": 0.3203,
+      "step": 19914
+    },
+    {
+      "epoch": 0.8744964159083175,
+      "grad_norm": 1.484375,
+      "learning_rate": 1.9294086468207325e-06,
+      "loss": 0.3169,
+      "step": 19916
+    },
+    {
+      "epoch": 0.874584234387521,
+      "grad_norm": 1.5078125,
+      "learning_rate": 1.9267444505090022e-06,
+      "loss": 0.3429,
+      "step": 19918
+    },
+    {
+      "epoch": 0.8746720528667244,
+      "grad_norm": 1.4375,
+      "learning_rate": 1.9240820211527148e-06,
+      "loss": 0.2993,
+      "step": 19920
+    },
+    {
+      "epoch": 0.8747598713459279,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.9214213589557608e-06,
+      "loss": 0.3321,
+      "step": 19922
+    },
+    {
+      "epoch": 0.8748476898251315,
+      "grad_norm": 1.4609375,
+      "learning_rate": 1.918762464121887e-06,
+      "loss": 0.2825,
+      "step": 19924
+    },
+    {
+      "epoch": 0.874935508304335,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.9161053368547207e-06,
+      "loss": 0.3315,
+      "step": 19926
+    },
+    {
+      "epoch": 0.8750233267835384,
+      "grad_norm": 1.5078125,
+      "learning_rate": 1.913449977357737e-06,
+      "loss": 0.3192,
+      "step": 19928
+    },
+    {
+      "epoch": 0.8751111452627419,
+      "grad_norm": 1.4609375,
+      "learning_rate": 1.9107963858342905e-06,
+      "loss": 0.323,
+      "step": 19930
+    },
+    {
+      "epoch": 0.8751989637419454,
+      "grad_norm": 1.4921875,
+      "learning_rate": 1.9081445624875933e-06,
+      "loss": 0.3143,
+      "step": 19932
+    },
+    {
+      "epoch": 0.8752867822211489,
+      "grad_norm": 1.453125,
+      "learning_rate": 1.9054945075207164e-06,
+      "loss": 0.3259,
+      "step": 19934
+    },
+    {
+      "epoch": 0.8753746007003523,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.9028462211366117e-06,
+      "loss": 0.3299,
+      "step": 19936
+    },
+    {
+      "epoch": 0.8754624191795558,
+      "grad_norm": 1.4453125,
+      "learning_rate": 1.9001997035380776e-06,
+      "loss": 0.3065,
+      "step": 19938
+    },
+    {
+      "epoch": 0.8755502376587594,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.897554954927791e-06,
+      "loss": 0.3281,
+      "step": 19940
+    },
+    {
+      "epoch": 0.8756380561379629,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.894911975508276e-06,
+      "loss": 0.3459,
+      "step": 19942
+    },
+    {
+      "epoch": 0.8757258746171663,
+      "grad_norm": 1.515625,
+      "learning_rate": 1.8922707654819488e-06,
+      "loss": 0.3059,
+      "step": 19944
+    },
+    {
+      "epoch": 0.8758136930963698,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.8896313250510612e-06,
+      "loss": 0.2982,
+      "step": 19946
+    },
+    {
+      "epoch": 0.8759015115755733,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.8869936544177458e-06,
+      "loss": 0.3176,
+      "step": 19948
+    },
+    {
+      "epoch": 0.8759893300547767,
+      "grad_norm": 1.4921875,
+      "learning_rate": 1.8843577537839996e-06,
+      "loss": 0.2997,
+      "step": 19950
+    },
+    {
+      "epoch": 0.8760771485339802,
+      "grad_norm": 1.515625,
+      "learning_rate": 1.881723623351675e-06,
+      "loss": 0.3199,
+      "step": 19952
+    },
+    {
+      "epoch": 0.8761649670131838,
+      "grad_norm": 1.4453125,
+      "learning_rate": 1.8790912633224994e-06,
+      "loss": 0.3363,
+      "step": 19954
+    },
+    {
+      "epoch": 0.8762527854923873,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.876460673898059e-06,
+      "loss": 0.3124,
+      "step": 19956
+    },
+    {
+      "epoch": 0.8763406039715907,
+      "grad_norm": 1.421875,
+      "learning_rate": 1.8738318552797978e-06,
+      "loss": 0.2969,
+      "step": 19958
+    },
+    {
+      "epoch": 0.8764284224507942,
+      "grad_norm": 1.5,
+      "learning_rate": 1.8712048076690442e-06,
+      "loss": 0.2909,
+      "step": 19960
+    },
+    {
+      "epoch": 0.8765162409299977,
+      "grad_norm": 1.4453125,
+      "learning_rate": 1.8685795312669619e-06,
+      "loss": 0.3397,
+      "step": 19962
+    },
+    {
+      "epoch": 0.8766040594092012,
+      "grad_norm": 1.4921875,
+      "learning_rate": 1.8659560262746123e-06,
+      "loss": 0.3175,
+      "step": 19964
+    },
+    {
+      "epoch": 0.8766918778884046,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.8633342928928931e-06,
+      "loss": 0.2963,
+      "step": 19966
+    },
+    {
+      "epoch": 0.8767796963676081,
+      "grad_norm": 1.5,
+      "learning_rate": 1.8607143313225773e-06,
+      "loss": 0.296,
+      "step": 19968
+    },
+    {
+      "epoch": 0.8768675148468117,
+      "grad_norm": 1.5078125,
+      "learning_rate": 1.8580961417643123e-06,
+      "loss": 0.3398,
+      "step": 19970
+    },
+    {
+      "epoch": 0.8769553333260152,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.855479724418585e-06,
+      "loss": 0.3342,
+      "step": 19972
+    },
+    {
+      "epoch": 0.8770431518052186,
+      "grad_norm": 1.515625,
+      "learning_rate": 1.8528650794857716e-06,
+      "loss": 0.3249,
+      "step": 19974
+    },
+    {
+      "epoch": 0.8771309702844221,
+      "grad_norm": 1.4453125,
+      "learning_rate": 1.850252207166095e-06,
+      "loss": 0.3134,
+      "step": 19976
+    },
+    {
+      "epoch": 0.8772187887636256,
+      "grad_norm": 1.4921875,
+      "learning_rate": 1.847641107659659e-06,
+      "loss": 0.323,
+      "step": 19978
+    },
+    {
+      "epoch": 0.877306607242829,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.845031781166412e-06,
+      "loss": 0.2874,
+      "step": 19980
+    },
+    {
+      "epoch": 0.8773944257220325,
+      "grad_norm": 1.4140625,
+      "learning_rate": 1.8424242278861858e-06,
+      "loss": 0.3032,
+      "step": 19982
+    },
+    {
+      "epoch": 0.877482244201236,
+      "grad_norm": 1.4609375,
+      "learning_rate": 1.8398184480186654e-06,
+      "loss": 0.3245,
+      "step": 19984
+    },
+    {
+      "epoch": 0.8775700626804396,
+      "grad_norm": 1.4453125,
+      "learning_rate": 1.8372144417633935e-06,
+      "loss": 0.3182,
+      "step": 19986
+    },
+    {
+      "epoch": 0.877657881159643,
+      "grad_norm": 1.5078125,
+      "learning_rate": 1.834612209319797e-06,
+      "loss": 0.3319,
+      "step": 19988
+    },
+    {
+      "epoch": 0.8777456996388465,
+      "grad_norm": 1.453125,
+      "learning_rate": 1.832011750887147e-06,
+      "loss": 0.2969,
+      "step": 19990
+    },
+    {
+      "epoch": 0.87783351811805,
+      "grad_norm": 1.46875,
+      "learning_rate": 1.8294130666645926e-06,
+      "loss": 0.3246,
+      "step": 19992
+    },
+    {
+      "epoch": 0.8779213365972535,
+      "grad_norm": 1.5,
+      "learning_rate": 1.8268161568511438e-06,
+      "loss": 0.3301,
+      "step": 19994
+    },
+    {
+      "epoch": 0.8780091550764569,
+      "grad_norm": 1.5078125,
+      "learning_rate": 1.8242210216456612e-06,
+      "loss": 0.2898,
+      "step": 19996
+    },
+    {
+      "epoch": 0.8780969735556604,
+      "grad_norm": 1.4296875,
+      "learning_rate": 1.8216276612468941e-06,
+      "loss": 0.3018,
+      "step": 19998
+    },
+    {
+      "epoch": 0.8781847920348639,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.8190360758534392e-06,
+      "loss": 0.277,
+      "step": 20000
+    },
+    {
+      "epoch": 0.8782726105140675,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.8164462656637544e-06,
+      "loss": 0.3097,
+      "step": 20002
+    },
+    {
+      "epoch": 0.8783604289932709,
+      "grad_norm": 1.453125,
+      "learning_rate": 1.8138582308761698e-06,
+      "loss": 0.3035,
+      "step": 20004
+    },
+    {
+      "epoch": 0.8784482474724744,
+      "grad_norm": 1.4375,
+      "learning_rate": 1.8112719716888853e-06,
+      "loss": 0.3346,
+      "step": 20006
+    },
+    {
+      "epoch": 0.8785360659516779,
+      "grad_norm": 1.4296875,
+      "learning_rate": 1.808687488299951e-06,
+      "loss": 0.3034,
+      "step": 20008
+    },
+    {
+      "epoch": 0.8786238844308814,
+      "grad_norm": 1.375,
+      "learning_rate": 1.8061047809072834e-06,
+      "loss": 0.3205,
+      "step": 20010
+    },
+    {
+      "epoch": 0.8787117029100848,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.8035238497086743e-06,
+      "loss": 0.3309,
+      "step": 20012
+    },
+    {
+      "epoch": 0.8787995213892883,
+      "grad_norm": 1.4609375,
+      "learning_rate": 1.8009446949017683e-06,
+      "loss": 0.3232,
+      "step": 20014
+    },
+    {
+      "epoch": 0.8788873398684919,
+      "grad_norm": 1.5078125,
+      "learning_rate": 1.7983673166840825e-06,
+      "loss": 0.3017,
+      "step": 20016
+    },
+    {
+      "epoch": 0.8789751583476954,
+      "grad_norm": 1.5078125,
+      "learning_rate": 1.7957917152529869e-06,
+      "loss": 0.3277,
+      "step": 20018
+    },
+    {
+      "epoch": 0.8790629768268988,
+      "grad_norm": 1.5,
+      "learning_rate": 1.7932178908057208e-06,
+      "loss": 0.2818,
+      "step": 20020
+    },
+    {
+      "epoch": 0.8791507953061023,
+      "grad_norm": 1.5078125,
+      "learning_rate": 1.7906458435393986e-06,
+      "loss": 0.3283,
+      "step": 20022
+    },
+    {
+      "epoch": 0.8792386137853058,
+      "grad_norm": 1.3984375,
+      "learning_rate": 1.7880755736509741e-06,
+      "loss": 0.3126,
+      "step": 20024
+    },
+    {
+      "epoch": 0.8793264322645092,
+      "grad_norm": 1.4375,
+      "learning_rate": 1.7855070813372926e-06,
+      "loss": 0.3258,
+      "step": 20026
+    },
+    {
+      "epoch": 0.8794142507437127,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.7829403667950412e-06,
+      "loss": 0.3533,
+      "step": 20028
+    },
+    {
+      "epoch": 0.8795020692229162,
+      "grad_norm": 1.4921875,
+      "learning_rate": 1.7803754302207848e-06,
+      "loss": 0.3357,
+      "step": 20030
+    },
+    {
+      "epoch": 0.8795898877021198,
+      "grad_norm": 1.453125,
+      "learning_rate": 1.7778122718109441e-06,
+      "loss": 0.318,
+      "step": 20032
+    },
+    {
+      "epoch": 0.8796777061813232,
+      "grad_norm": 1.453125,
+      "learning_rate": 1.7752508917618011e-06,
+      "loss": 0.3047,
+      "step": 20034
+    },
+    {
+      "epoch": 0.8797655246605267,
+      "grad_norm": 1.4453125,
+      "learning_rate": 1.7726912902695152e-06,
+      "loss": 0.3174,
+      "step": 20036
+    },
+    {
+      "epoch": 0.8798533431397302,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.770133467530094e-06,
+      "loss": 0.3134,
+      "step": 20038
+    },
+    {
+      "epoch": 0.8799411616189337,
+      "grad_norm": 1.3984375,
+      "learning_rate": 1.7675774237394278e-06,
+      "loss": 0.2914,
+      "step": 20040
+    },
+    {
+      "epoch": 0.8800289800981371,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.765023159093246e-06,
+      "loss": 0.3218,
+      "step": 20042
+    },
+    {
+      "epoch": 0.8801167985773406,
+      "grad_norm": 1.484375,
+      "learning_rate": 1.7624706737871643e-06,
+      "loss": 0.3444,
+      "step": 20044
+    },
+    {
+      "epoch": 0.8802046170565441,
+      "grad_norm": 1.4765625,
+      "learning_rate": 1.7599199680166462e-06,
+      "loss": 0.3358,
+      "step": 20046
+    },
+    {
+      "epoch": 0.8802924355357477,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.7573710419770267e-06,
+      "loss": 0.3213,
+      "step": 20048
+    },
+    {
+      "epoch": 0.8803802540149511,
+      "grad_norm": 1.4921875,
+      "learning_rate": 1.7548238958635082e-06,
+      "loss": 0.2897,
+      "step": 20050
+    },
+    {
+      "epoch": 0.8804680724941546,
+      "grad_norm": 1.453125,
+      "learning_rate": 1.7522785298711457e-06,
+      "loss": 0.3444,
+      "step": 20052
+    },
+    {
+      "epoch": 0.8805558909733581,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.7497349441948668e-06,
+      "loss": 0.2972,
+      "step": 20054
+    },
+    {
+      "epoch": 0.8806437094525615,
+      "grad_norm": 1.46875,
+      "learning_rate": 1.7471931390294627e-06,
+      "loss": 0.3189,
+      "step": 20056
+    },
+    {
+      "epoch": 0.880731527931765,
+      "grad_norm": 1.484375,
+      "learning_rate": 1.744653114569575e-06,
+      "loss": 0.3152,
+      "step": 20058
+    },
+    {
+      "epoch": 0.8808193464109685,
+      "grad_norm": 1.5,
+      "learning_rate": 1.7421148710097312e-06,
+      "loss": 0.3031,
+      "step": 20060
+    },
+    {
+      "epoch": 0.8809071648901721,
+      "grad_norm": 1.4921875,
+      "learning_rate": 1.7395784085443068e-06,
+      "loss": 0.3525,
+      "step": 20062
+    },
+    {
+      "epoch": 0.8809949833693755,
+      "grad_norm": 1.4375,
+      "learning_rate": 1.737043727367546e-06,
+      "loss": 0.3036,
+      "step": 20064
+    },
+    {
+      "epoch": 0.881082801848579,
+      "grad_norm": 1.4921875,
+      "learning_rate": 1.7345108276735467e-06,
+      "loss": 0.3163,
+      "step": 20066
+    },
+    {
+      "epoch": 0.8811706203277825,
+      "grad_norm": 1.4375,
+      "learning_rate": 1.7319797096562867e-06,
+      "loss": 0.3123,
+      "step": 20068
+    },
+    {
+      "epoch": 0.881258438806986,
+      "grad_norm": 1.4765625,
+      "learning_rate": 1.729450373509603e-06,
+      "loss": 0.2987,
+      "step": 20070
+    },
+    {
+      "epoch": 0.8813462572861894,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.726922819427182e-06,
+      "loss": 0.3379,
+      "step": 20072
+    },
+    {
+      "epoch": 0.8814340757653929,
+      "grad_norm": 1.484375,
+      "learning_rate": 1.724397047602594e-06,
+      "loss": 0.3099,
+      "step": 20074
+    },
+    {
+      "epoch": 0.8815218942445964,
+      "grad_norm": 1.4296875,
+      "learning_rate": 1.7218730582292563e-06,
+      "loss": 0.3147,
+      "step": 20076
+    },
+    {
+      "epoch": 0.8816097127238,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.719350851500462e-06,
+      "loss": 0.3137,
+      "step": 20078
+    },
+    {
+      "epoch": 0.8816975312030034,
+      "grad_norm": 1.4375,
+      "learning_rate": 1.7168304276093616e-06,
+      "loss": 0.3025,
+      "step": 20080
+    },
+    {
+      "epoch": 0.8817853496822069,
+      "grad_norm": 1.390625,
+      "learning_rate": 1.714311786748965e-06,
+      "loss": 0.3322,
+      "step": 20082
+    },
+    {
+      "epoch": 0.8818731681614104,
+      "grad_norm": 1.4609375,
+      "learning_rate": 1.7117949291121565e-06,
+      "loss": 0.3313,
+      "step": 20084
+    },
+    {
+      "epoch": 0.8819609866406138,
+      "grad_norm": 1.4453125,
+      "learning_rate": 1.7092798548916738e-06,
+      "loss": 0.3306,
+      "step": 20086
+    },
+    {
+      "epoch": 0.8820488051198173,
+      "grad_norm": 1.5078125,
+      "learning_rate": 1.706766564280124e-06,
+      "loss": 0.3208,
+      "step": 20088
+    },
+    {
+      "epoch": 0.8821366235990208,
+      "grad_norm": 1.4296875,
+      "learning_rate": 1.704255057469975e-06,
+      "loss": 0.3011,
+      "step": 20090
+    },
+    {
+      "epoch": 0.8822244420782243,
+      "grad_norm": 1.4375,
+      "learning_rate": 1.7017453346535595e-06,
+      "loss": 0.3052,
+      "step": 20092
+    },
+    {
+      "epoch": 0.8823122605574278,
+      "grad_norm": 1.453125,
+      "learning_rate": 1.6992373960230707e-06,
+      "loss": 0.3114,
+      "step": 20094
+    },
+    {
+      "epoch": 0.8824000790366313,
+      "grad_norm": 1.484375,
+      "learning_rate": 1.6967312417705634e-06,
+      "loss": 0.2927,
+      "step": 20096
+    },
+    {
+      "epoch": 0.8824878975158348,
+      "grad_norm": 1.4921875,
+      "learning_rate": 1.6942268720879672e-06,
+      "loss": 0.2854,
+      "step": 20098
+    },
+    {
+      "epoch": 0.8825757159950383,
+      "grad_norm": 1.5078125,
+      "learning_rate": 1.6917242871670596e-06,
+      "loss": 0.3352,
+      "step": 20100
+    },
+    {
+      "epoch": 0.8826635344742417,
+      "grad_norm": 1.453125,
+      "learning_rate": 1.6892234871994983e-06,
+      "loss": 0.3241,
+      "step": 20102
+    },
+    {
+      "epoch": 0.8827513529534452,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.686724472376791e-06,
+      "loss": 0.3028,
+      "step": 20104
+    },
+    {
+      "epoch": 0.8828391714326487,
+      "grad_norm": 1.4921875,
+      "learning_rate": 1.6842272428903073e-06,
+      "loss": 0.3198,
+      "step": 20106
+    },
+    {
+      "epoch": 0.8829269899118523,
+      "grad_norm": 1.484375,
+      "learning_rate": 1.6817317989312935e-06,
+      "loss": 0.3084,
+      "step": 20108
+    },
+    {
+      "epoch": 0.8830148083910557,
+      "grad_norm": 1.484375,
+      "learning_rate": 1.6792381406908475e-06,
+      "loss": 0.3003,
+      "step": 20110
+    },
+    {
+      "epoch": 0.8831026268702592,
+      "grad_norm": 1.4765625,
+      "learning_rate": 1.6767462683599355e-06,
+      "loss": 0.3089,
+      "step": 20112
+    },
+    {
+      "epoch": 0.8831904453494627,
+      "grad_norm": 1.4765625,
+      "learning_rate": 1.6742561821293829e-06,
+      "loss": 0.2854,
+      "step": 20114
+    },
+    {
+      "epoch": 0.8832782638286661,
+      "grad_norm": 1.5,
+      "learning_rate": 1.671767882189887e-06,
+      "loss": 0.3033,
+      "step": 20116
+    },
+    {
+      "epoch": 0.8833660823078696,
+      "grad_norm": 1.46875,
+      "learning_rate": 1.6692813687320008e-06,
+      "loss": 0.33,
+      "step": 20118
+    },
+    {
+      "epoch": 0.8834539007870731,
+      "grad_norm": 1.625,
+      "learning_rate": 1.6667966419461333e-06,
+      "loss": 0.3175,
+      "step": 20120
+    },
+    {
+      "epoch": 0.8835417192662766,
+      "grad_norm": 1.46875,
+      "learning_rate": 1.6643137020225824e-06,
+      "loss": 0.3049,
+      "step": 20122
+    },
+    {
+      "epoch": 0.8836295377454801,
+      "grad_norm": 1.3984375,
+      "learning_rate": 1.6618325491514736e-06,
+      "loss": 0.3186,
+      "step": 20124
+    },
+    {
+      "epoch": 0.8837173562246836,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.6593531835228244e-06,
+      "loss": 0.3149,
+      "step": 20126
+    },
+    {
+      "epoch": 0.8838051747038871,
+      "grad_norm": 1.5,
+      "learning_rate": 1.6568756053265023e-06,
+      "loss": 0.3189,
+      "step": 20128
+    },
+    {
+      "epoch": 0.8838929931830906,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.6543998147522444e-06,
+      "loss": 0.3124,
+      "step": 20130
+    },
+    {
+      "epoch": 0.883980811662294,
+      "grad_norm": 1.40625,
+      "learning_rate": 1.6519258119896463e-06,
+      "loss": 0.2992,
+      "step": 20132
+    },
+    {
+      "epoch": 0.8840686301414975,
+      "grad_norm": 1.46875,
+      "learning_rate": 1.6494535972281623e-06,
+      "loss": 0.3163,
+      "step": 20134
+    },
+    {
+      "epoch": 0.884156448620701,
+      "grad_norm": 1.4921875,
+      "learning_rate": 1.6469831706571237e-06,
+      "loss": 0.3111,
+      "step": 20136
+    },
+    {
+      "epoch": 0.8842442670999044,
+      "grad_norm": 1.4453125,
+      "learning_rate": 1.6445145324657075e-06,
+      "loss": 0.3072,
+      "step": 20138
+    },
+    {
+      "epoch": 0.884332085579108,
+      "grad_norm": 1.5,
+      "learning_rate": 1.6420476828429704e-06,
+      "loss": 0.3367,
+      "step": 20140
+    },
+    {
+      "epoch": 0.8844199040583115,
+      "grad_norm": 1.453125,
+      "learning_rate": 1.6395826219778226e-06,
+      "loss": 0.3162,
+      "step": 20142
+    },
+    {
+      "epoch": 0.884507722537515,
+      "grad_norm": 1.4921875,
+      "learning_rate": 1.6371193500590325e-06,
+      "loss": 0.3177,
+      "step": 20144
+    },
+    {
+      "epoch": 0.8845955410167184,
+      "grad_norm": 1.4140625,
+      "learning_rate": 1.6346578672752467e-06,
+      "loss": 0.3382,
+      "step": 20146
+    },
+    {
+      "epoch": 0.8846833594959219,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.6321981738149584e-06,
+      "loss": 0.3199,
+      "step": 20148
+    },
+    {
+      "epoch": 0.8847711779751254,
+      "grad_norm": 1.4921875,
+      "learning_rate": 1.6297402698665392e-06,
+      "loss": 0.3121,
+      "step": 20150
+    },
+    {
+      "epoch": 0.8848589964543289,
+      "grad_norm": 1.4921875,
+      "learning_rate": 1.6272841556182139e-06,
+      "loss": 0.3434,
+      "step": 20152
+    },
+    {
+      "epoch": 0.8849468149335324,
+      "grad_norm": 1.5,
+      "learning_rate": 1.6248298312580706e-06,
+      "loss": 0.2925,
+      "step": 20154
+    },
+    {
+      "epoch": 0.8850346334127359,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.6223772969740592e-06,
+      "loss": 0.3513,
+      "step": 20156
+    },
+    {
+      "epoch": 0.8851224518919394,
+      "grad_norm": 1.4765625,
+      "learning_rate": 1.6199265529539987e-06,
+      "loss": 0.3147,
+      "step": 20158
+    },
+    {
+      "epoch": 0.8852102703711429,
+      "grad_norm": 1.4296875,
+      "learning_rate": 1.6174775993855695e-06,
+      "loss": 0.303,
+      "step": 20160
+    },
+    {
+      "epoch": 0.8852980888503463,
+      "grad_norm": 1.4765625,
+      "learning_rate": 1.615030436456308e-06,
+      "loss": 0.3321,
+      "step": 20162
+    },
+    {
+      "epoch": 0.8853859073295498,
+      "grad_norm": 1.3828125,
+      "learning_rate": 1.6125850643536221e-06,
+      "loss": 0.3239,
+      "step": 20164
+    },
+    {
+      "epoch": 0.8854737258087533,
+      "grad_norm": 1.4765625,
+      "learning_rate": 1.6101414832647822e-06,
+      "loss": 0.3308,
+      "step": 20166
+    },
+    {
+      "epoch": 0.8855615442879567,
+      "grad_norm": 1.46875,
+      "learning_rate": 1.6076996933769073e-06,
+      "loss": 0.3111,
+      "step": 20168
+    },
+    {
+      "epoch": 0.8856493627671603,
+      "grad_norm": 1.5078125,
+      "learning_rate": 1.6052596948770037e-06,
+      "loss": 0.3035,
+      "step": 20170
+    },
+    {
+      "epoch": 0.8857371812463638,
+      "grad_norm": 1.5,
+      "learning_rate": 1.6028214879519166e-06,
+      "loss": 0.3051,
+      "step": 20172
+    },
+    {
+      "epoch": 0.8858249997255673,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.6003850727883713e-06,
+      "loss": 0.2882,
+      "step": 20174
+    },
+    {
+      "epoch": 0.8859128182047707,
+      "grad_norm": 1.4296875,
+      "learning_rate": 1.5979504495729436e-06,
+      "loss": 0.3004,
+      "step": 20176
+    },
+    {
+      "epoch": 0.8860006366839742,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.5955176184920844e-06,
+      "loss": 0.3127,
+      "step": 20178
+    },
+    {
+      "epoch": 0.8860884551631777,
+      "grad_norm": 1.4765625,
+      "learning_rate": 1.5930865797320948e-06,
+      "loss": 0.3208,
+      "step": 20180
+    },
+    {
+      "epoch": 0.8861762736423812,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.5906573334791424e-06,
+      "loss": 0.321,
+      "step": 20182
+    },
+    {
+      "epoch": 0.8862640921215846,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.588229879919273e-06,
+      "loss": 0.3277,
+      "step": 20184
+    },
+    {
+      "epoch": 0.8863519106007882,
+      "grad_norm": 1.46875,
+      "learning_rate": 1.5858042192383626e-06,
+      "loss": 0.3402,
+      "step": 20186
+    },
+    {
+      "epoch": 0.8864397290799917,
+      "grad_norm": 1.5078125,
+      "learning_rate": 1.5833803516221795e-06,
+      "loss": 0.321,
+      "step": 20188
+    },
+    {
+      "epoch": 0.8865275475591952,
+      "grad_norm": 1.3984375,
+      "learning_rate": 1.580958277256342e-06,
+      "loss": 0.3062,
+      "step": 20190
+    },
+    {
+      "epoch": 0.8866153660383986,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.5785379963263347e-06,
+      "loss": 0.3405,
+      "step": 20192
+    },
+    {
+      "epoch": 0.8867031845176021,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.5761195090175041e-06,
+      "loss": 0.3248,
+      "step": 20194
+    },
+    {
+      "epoch": 0.8867910029968056,
+      "grad_norm": 1.453125,
+      "learning_rate": 1.5737028155150517e-06,
+      "loss": 0.3085,
+      "step": 20196
+    },
+    {
+      "epoch": 0.886878821476009,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.5712879160040578e-06,
+      "loss": 0.3068,
+      "step": 20198
+    },
+    {
+      "epoch": 0.8869666399552125,
+      "grad_norm": 1.4609375,
+      "learning_rate": 1.568874810669449e-06,
+      "loss": 0.3134,
+      "step": 20200
+    },
+    {
+      "epoch": 0.8870544584344161,
+      "grad_norm": 1.46875,
+      "learning_rate": 1.5664634996960248e-06,
+      "loss": 0.3131,
+      "step": 20202
+    },
+    {
+      "epoch": 0.8871422769136196,
+      "grad_norm": 1.4609375,
+      "learning_rate": 1.564053983268446e-06,
+      "loss": 0.318,
+      "step": 20204
+    },
+    {
+      "epoch": 0.887230095392823,
+      "grad_norm": 1.515625,
+      "learning_rate": 1.5616462615712257e-06,
+      "loss": 0.2956,
+      "step": 20206
+    },
+    {
+      "epoch": 0.8873179138720265,
+      "grad_norm": 1.421875,
+      "learning_rate": 1.5592403347887558e-06,
+      "loss": 0.2928,
+      "step": 20208
+    },
+    {
+      "epoch": 0.88740573235123,
+      "grad_norm": 1.4296875,
+      "learning_rate": 1.5568362031052803e-06,
+      "loss": 0.3236,
+      "step": 20210
+    },
+    {
+      "epoch": 0.8874935508304335,
+      "grad_norm": 1.4453125,
+      "learning_rate": 1.5544338667049075e-06,
+      "loss": 0.3086,
+      "step": 20212
+    },
+    {
+      "epoch": 0.8875813693096369,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.5520333257716125e-06,
+      "loss": 0.3167,
+      "step": 20214
+    },
+    {
+      "epoch": 0.8876691877888405,
+      "grad_norm": 1.5,
+      "learning_rate": 1.549634580489226e-06,
+      "loss": 0.3088,
+      "step": 20216
+    },
+    {
+      "epoch": 0.887757006268044,
+      "grad_norm": 1.4921875,
+      "learning_rate": 1.5472376310414428e-06,
+      "loss": 0.2893,
+      "step": 20218
+    },
+    {
+      "epoch": 0.8878448247472475,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.5448424776118215e-06,
+      "loss": 0.3047,
+      "step": 20220
+    },
+    {
+      "epoch": 0.8879326432264509,
+      "grad_norm": 1.4609375,
+      "learning_rate": 1.5424491203837903e-06,
+      "loss": 0.2981,
+      "step": 20222
+    },
+    {
+      "epoch": 0.8880204617056544,
+      "grad_norm": 1.46875,
+      "learning_rate": 1.5400575595406226e-06,
+      "loss": 0.3243,
+      "step": 20224
+    },
+    {
+      "epoch": 0.8881082801848579,
+      "grad_norm": 1.4921875,
+      "learning_rate": 1.5376677952654767e-06,
+      "loss": 0.3125,
+      "step": 20226
+    },
+    {
+      "epoch": 0.8881960986640614,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.5352798277413539e-06,
+      "loss": 0.3246,
+      "step": 20228
+    },
+    {
+      "epoch": 0.8882839171432648,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.5328936571511244e-06,
+      "loss": 0.3198,
+      "step": 20230
+    },
+    {
+      "epoch": 0.8883717356224684,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.530509283677528e-06,
+      "loss": 0.3416,
+      "step": 20232
+    },
+    {
+      "epoch": 0.8884595541016719,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.5281267075031497e-06,
+      "loss": 0.313,
+      "step": 20234
+    },
+    {
+      "epoch": 0.8885473725808753,
+      "grad_norm": 1.4765625,
+      "learning_rate": 1.5257459288104625e-06,
+      "loss": 0.3235,
+      "step": 20236
+    },
+    {
+      "epoch": 0.8886351910600788,
+      "grad_norm": 1.4296875,
+      "learning_rate": 1.5233669477817737e-06,
+      "loss": 0.2899,
+      "step": 20238
+    },
+    {
+      "epoch": 0.8887230095392823,
+      "grad_norm": 1.4453125,
+      "learning_rate": 1.520989764599276e-06,
+      "loss": 0.3083,
+      "step": 20240
+    },
+    {
+      "epoch": 0.8888108280184858,
+      "grad_norm": 1.4921875,
+      "learning_rate": 1.5186143794450103e-06,
+      "loss": 0.3121,
+      "step": 20242
+    },
+    {
+      "epoch": 0.8888986464976892,
+      "grad_norm": 1.4140625,
+      "learning_rate": 1.516240792500878e-06,
+      "loss": 0.3316,
+      "step": 20244
+    },
+    {
+      "epoch": 0.8889864649768927,
+      "grad_norm": 1.515625,
+      "learning_rate": 1.5138690039486675e-06,
+      "loss": 0.2773,
+      "step": 20246
+    },
+    {
+      "epoch": 0.8890742834560963,
+      "grad_norm": 1.4609375,
+      "learning_rate": 1.5114990139699886e-06,
+      "loss": 0.3228,
+      "step": 20248
+    },
+    {
+      "epoch": 0.8891621019352998,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.5091308227463492e-06,
+      "loss": 0.3265,
+      "step": 20250
+    },
+    {
+      "epoch": 0.8892499204145032,
+      "grad_norm": 1.4453125,
+      "learning_rate": 1.5067644304590983e-06,
+      "loss": 0.3161,
+      "step": 20252
+    },
+    {
+      "epoch": 0.8893377388937067,
+      "grad_norm": 1.4375,
+      "learning_rate": 1.5043998372894634e-06,
+      "loss": 0.3223,
+      "step": 20254
+    },
+    {
+      "epoch": 0.8894255573729102,
+      "grad_norm": 1.4765625,
+      "learning_rate": 1.5020370434185221e-06,
+      "loss": 0.3077,
+      "step": 20256
+    },
+    {
+      "epoch": 0.8895133758521137,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.4996760490272127e-06,
+      "loss": 0.3164,
+      "step": 20258
+    },
+    {
+      "epoch": 0.8896011943313171,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.4973168542963489e-06,
+      "loss": 0.299,
+      "step": 20260
+    },
+    {
+      "epoch": 0.8896890128105207,
+      "grad_norm": 1.484375,
+      "learning_rate": 1.494959459406589e-06,
+      "loss": 0.3543,
+      "step": 20262
+    },
+    {
+      "epoch": 0.8897768312897242,
+      "grad_norm": 1.515625,
+      "learning_rate": 1.4926038645384749e-06,
+      "loss": 0.3334,
+      "step": 20264
+    },
+    {
+      "epoch": 0.8898646497689277,
+      "grad_norm": 1.46875,
+      "learning_rate": 1.4902500698723897e-06,
+      "loss": 0.3223,
+      "step": 20266
+    },
+    {
+      "epoch": 0.8899524682481311,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.4878980755885868e-06,
+      "loss": 0.3124,
+      "step": 20268
+    },
+    {
+      "epoch": 0.8900402867273346,
+      "grad_norm": 1.46875,
+      "learning_rate": 1.4855478818671887e-06,
+      "loss": 0.3265,
+      "step": 20270
+    },
+    {
+      "epoch": 0.8901281052065381,
+      "grad_norm": 1.40625,
+      "learning_rate": 1.483199488888168e-06,
+      "loss": 0.3368,
+      "step": 20272
+    },
+    {
+      "epoch": 0.8902159236857415,
+      "grad_norm": 1.4453125,
+      "learning_rate": 1.4808528968313695e-06,
+      "loss": 0.3398,
+      "step": 20274
+    },
+    {
+      "epoch": 0.890303742164945,
+      "grad_norm": 1.3828125,
+      "learning_rate": 1.4785081058764972e-06,
+      "loss": 0.3304,
+      "step": 20276
+    },
+    {
+      "epoch": 0.8903915606441486,
+      "grad_norm": 1.4609375,
+      "learning_rate": 1.4761651162031099e-06,
+      "loss": 0.3183,
+      "step": 20278
+    },
+    {
+      "epoch": 0.8904793791233521,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.473823927990639e-06,
+      "loss": 0.3324,
+      "step": 20280
+    },
+    {
+      "epoch": 0.8905671976025555,
+      "grad_norm": 1.4375,
+      "learning_rate": 1.4714845414183665e-06,
+      "loss": 0.3068,
+      "step": 20282
+    },
+    {
+      "epoch": 0.890655016081759,
+      "grad_norm": 1.4453125,
+      "learning_rate": 1.4691469566654514e-06,
+      "loss": 0.3171,
+      "step": 20284
+    },
+    {
+      "epoch": 0.8907428345609625,
+      "grad_norm": 1.5078125,
+      "learning_rate": 1.466811173910898e-06,
+      "loss": 0.3047,
+      "step": 20286
+    },
+    {
+      "epoch": 0.890830653040166,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.4644771933335938e-06,
+      "loss": 0.3207,
+      "step": 20288
+    },
+    {
+      "epoch": 0.8909184715193694,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.4621450151122652e-06,
+      "loss": 0.3226,
+      "step": 20290
+    },
+    {
+      "epoch": 0.8910062899985729,
+      "grad_norm": 1.4140625,
+      "learning_rate": 1.4598146394255109e-06,
+      "loss": 0.3179,
+      "step": 20292
+    },
+    {
+      "epoch": 0.8910941084777765,
+      "grad_norm": 1.46875,
+      "learning_rate": 1.4574860664517964e-06,
+      "loss": 0.2931,
+      "step": 20294
+    },
+    {
+      "epoch": 0.89118192695698,
+      "grad_norm": 1.4921875,
+      "learning_rate": 1.4551592963694404e-06,
+      "loss": 0.305,
+      "step": 20296
+    },
+    {
+      "epoch": 0.8912697454361834,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.4528343293566337e-06,
+      "loss": 0.288,
+      "step": 20298
+    },
+    {
+      "epoch": 0.8913575639153869,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.4505111655914199e-06,
+      "loss": 0.3151,
+      "step": 20300
+    },
+    {
+      "epoch": 0.8914453823945904,
+      "grad_norm": 1.484375,
+      "learning_rate": 1.4481898052517013e-06,
+      "loss": 0.339,
+      "step": 20302
+    },
+    {
+      "epoch": 0.8915332008737938,
+      "grad_norm": 1.4375,
+      "learning_rate": 1.4458702485152549e-06,
+      "loss": 0.2906,
+      "step": 20304
+    },
+    {
+      "epoch": 0.8916210193529973,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.4435524955597134e-06,
+      "loss": 0.3058,
+      "step": 20306
+    },
+    {
+      "epoch": 0.8917088378322009,
+      "grad_norm": 1.4609375,
+      "learning_rate": 1.4412365465625689e-06,
+      "loss": 0.3228,
+      "step": 20308
+    },
+    {
+      "epoch": 0.8917966563114044,
+      "grad_norm": 1.5078125,
+      "learning_rate": 1.4389224017011704e-06,
+      "loss": 0.3381,
+      "step": 20310
+    },
+    {
+      "epoch": 0.8918844747906078,
+      "grad_norm": 1.5,
+      "learning_rate": 1.436610061152749e-06,
+      "loss": 0.323,
+      "step": 20312
+    },
+    {
+      "epoch": 0.8919722932698113,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.4342995250943735e-06,
+      "loss": 0.3117,
+      "step": 20314
+    },
+    {
+      "epoch": 0.8920601117490148,
+      "grad_norm": 1.4375,
+      "learning_rate": 1.431990793702992e-06,
+      "loss": 0.3206,
+      "step": 20316
+    },
+    {
+      "epoch": 0.8921479302282183,
+      "grad_norm": 1.4921875,
+      "learning_rate": 1.429683867155407e-06,
+      "loss": 0.2996,
+      "step": 20318
+    },
+    {
+      "epoch": 0.8922357487074217,
+      "grad_norm": 1.515625,
+      "learning_rate": 1.4273787456282771e-06,
+      "loss": 0.3233,
+      "step": 20320
+    },
+    {
+      "epoch": 0.8923235671866252,
+      "grad_norm": 1.4453125,
+      "learning_rate": 1.4250754292981366e-06,
+      "loss": 0.3058,
+      "step": 20322
+    },
+    {
+      "epoch": 0.8924113856658288,
+      "grad_norm": 1.453125,
+      "learning_rate": 1.4227739183413663e-06,
+      "loss": 0.3134,
+      "step": 20324
+    },
+    {
+      "epoch": 0.8924992041450323,
+      "grad_norm": 1.484375,
+      "learning_rate": 1.420474212934228e-06,
+      "loss": 0.3056,
+      "step": 20326
+    },
+    {
+      "epoch": 0.8925870226242357,
+      "grad_norm": 1.515625,
+      "learning_rate": 1.4181763132528253e-06,
+      "loss": 0.3122,
+      "step": 20328
+    },
+    {
+      "epoch": 0.8926748411034392,
+      "grad_norm": 1.4609375,
+      "learning_rate": 1.4158802194731285e-06,
+      "loss": 0.2905,
+      "step": 20330
+    },
+    {
+      "epoch": 0.8927626595826427,
+      "grad_norm": 1.515625,
+      "learning_rate": 1.4135859317709832e-06,
+      "loss": 0.3222,
+      "step": 20332
+    },
+    {
+      "epoch": 0.8928504780618461,
+      "grad_norm": 1.4140625,
+      "learning_rate": 1.4112934503220766e-06,
+      "loss": 0.2815,
+      "step": 20334
+    },
+    {
+      "epoch": 0.8929382965410496,
+      "grad_norm": 1.484375,
+      "learning_rate": 1.4090027753019763e-06,
+      "loss": 0.3155,
+      "step": 20336
+    },
+    {
+      "epoch": 0.8930261150202531,
+      "grad_norm": 1.453125,
+      "learning_rate": 1.4067139068861007e-06,
+      "loss": 0.3158,
+      "step": 20338
+    },
+    {
+      "epoch": 0.8931139334994567,
+      "grad_norm": 1.625,
+      "learning_rate": 1.404426845249729e-06,
+      "loss": 0.3042,
+      "step": 20340
+    },
+    {
+      "epoch": 0.8932017519786601,
+      "grad_norm": 1.34375,
+      "learning_rate": 1.4021415905680042e-06,
+      "loss": 0.3318,
+      "step": 20342
+    },
+    {
+      "epoch": 0.8932895704578636,
+      "grad_norm": 1.453125,
+      "learning_rate": 1.399858143015928e-06,
+      "loss": 0.3627,
+      "step": 20344
+    },
+    {
+      "epoch": 0.8933773889370671,
+      "grad_norm": 1.4453125,
+      "learning_rate": 1.39757650276838e-06,
+      "loss": 0.3007,
+      "step": 20346
+    },
+    {
+      "epoch": 0.8934652074162706,
+      "grad_norm": 1.5,
+      "learning_rate": 1.3952966700000764e-06,
+      "loss": 0.3236,
+      "step": 20348
+    },
+    {
+      "epoch": 0.893553025895474,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.393018644885613e-06,
+      "loss": 0.3086,
+      "step": 20350
+    },
+    {
+      "epoch": 0.8936408443746775,
+      "grad_norm": 1.5,
+      "learning_rate": 1.3907424275994452e-06,
+      "loss": 0.284,
+      "step": 20352
+    },
+    {
+      "epoch": 0.893728662853881,
+      "grad_norm": 1.421875,
+      "learning_rate": 1.3884680183158748e-06,
+      "loss": 0.2977,
+      "step": 20354
+    },
+    {
+      "epoch": 0.8938164813330846,
+      "grad_norm": 1.4296875,
+      "learning_rate": 1.3861954172090907e-06,
+      "loss": 0.301,
+      "step": 20356
+    },
+    {
+      "epoch": 0.893904299812288,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.3839246244531145e-06,
+      "loss": 0.3359,
+      "step": 20358
+    },
+    {
+      "epoch": 0.8939921182914915,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.3816556402218573e-06,
+      "loss": 0.3188,
+      "step": 20360
+    },
+    {
+      "epoch": 0.894079936770695,
+      "grad_norm": 1.4140625,
+      "learning_rate": 1.3793884646890742e-06,
+      "loss": 0.3213,
+      "step": 20362
+    },
+    {
+      "epoch": 0.8941677552498984,
+      "grad_norm": 1.4921875,
+      "learning_rate": 1.3771230980283795e-06,
+      "loss": 0.3277,
+      "step": 20364
+    },
+    {
+      "epoch": 0.8942555737291019,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.3748595404132648e-06,
+      "loss": 0.3196,
+      "step": 20366
+    },
+    {
+      "epoch": 0.8943433922083054,
+      "grad_norm": 1.453125,
+      "learning_rate": 1.372597792017069e-06,
+      "loss": 0.3485,
+      "step": 20368
+    },
+    {
+      "epoch": 0.894431210687509,
+      "grad_norm": 1.4765625,
+      "learning_rate": 1.3703378530129984e-06,
+      "loss": 0.3237,
+      "step": 20370
+    },
+    {
+      "epoch": 0.8945190291667124,
+      "grad_norm": 1.4296875,
+      "learning_rate": 1.3680797235741171e-06,
+      "loss": 0.3003,
+      "step": 20372
+    },
+    {
+      "epoch": 0.8946068476459159,
+      "grad_norm": 1.4453125,
+      "learning_rate": 1.3658234038733591e-06,
+      "loss": 0.2984,
+      "step": 20374
+    },
+    {
+      "epoch": 0.8946946661251194,
+      "grad_norm": 1.4453125,
+      "learning_rate": 1.3635688940835056e-06,
+      "loss": 0.306,
+      "step": 20376
+    },
+    {
+      "epoch": 0.8947824846043229,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.3613161943772156e-06,
+      "loss": 0.3302,
+      "step": 20378
+    },
+    {
+      "epoch": 0.8948703030835263,
+      "grad_norm": 1.4140625,
+      "learning_rate": 1.3590653049269985e-06,
+      "loss": 0.2974,
+      "step": 20380
+    },
+    {
+      "epoch": 0.8949581215627298,
+      "grad_norm": 1.4765625,
+      "learning_rate": 1.3568162259052247e-06,
+      "loss": 0.3034,
+      "step": 20382
+    },
+    {
+      "epoch": 0.8950459400419333,
+      "grad_norm": 1.4453125,
+      "learning_rate": 1.3545689574841342e-06,
+      "loss": 0.3418,
+      "step": 20384
+    },
+    {
+      "epoch": 0.8951337585211369,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.352323499835817e-06,
+      "loss": 0.3485,
+      "step": 20386
+    },
+    {
+      "epoch": 0.8952215770003403,
+      "grad_norm": 1.484375,
+      "learning_rate": 1.3500798531322412e-06,
+      "loss": 0.3248,
+      "step": 20388
+    },
+    {
+      "epoch": 0.8953093954795438,
+      "grad_norm": 1.484375,
+      "learning_rate": 1.3478380175452165e-06,
+      "loss": 0.3054,
+      "step": 20390
+    },
+    {
+      "epoch": 0.8953972139587473,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.3455979932464253e-06,
+      "loss": 0.3101,
+      "step": 20392
+    },
+    {
+      "epoch": 0.8954850324379507,
+      "grad_norm": 1.4453125,
+      "learning_rate": 1.3433597804074132e-06,
+      "loss": 0.2924,
+      "step": 20394
+    },
+    {
+      "epoch": 0.8955728509171542,
+      "grad_norm": 1.4296875,
+      "learning_rate": 1.3411233791995743e-06,
+      "loss": 0.2853,
+      "step": 20396
+    },
+    {
+      "epoch": 0.8956606693963577,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.3388887897941877e-06,
+      "loss": 0.3011,
+      "step": 20398
+    },
+    {
+      "epoch": 0.8957484878755612,
+      "grad_norm": 1.484375,
+      "learning_rate": 1.336656012362361e-06,
+      "loss": 0.3603,
+      "step": 20400
+    },
+    {
+      "epoch": 0.8958363063547647,
+      "grad_norm": 1.4765625,
+      "learning_rate": 1.3344250470750941e-06,
+      "loss": 0.3254,
+      "step": 20402
+    },
+    {
+      "epoch": 0.8959241248339682,
+      "grad_norm": 1.453125,
+      "learning_rate": 1.3321958941032303e-06,
+      "loss": 0.294,
+      "step": 20404
+    },
+    {
+      "epoch": 0.8960119433131717,
+      "grad_norm": 1.5078125,
+      "learning_rate": 1.3299685536174749e-06,
+      "loss": 0.3111,
+      "step": 20406
+    },
+    {
+      "epoch": 0.8960997617923752,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.3277430257884055e-06,
+      "loss": 0.3373,
+      "step": 20408
+    },
+    {
+      "epoch": 0.8961875802715786,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.3255193107864438e-06,
+      "loss": 0.3106,
+      "step": 20410
+    },
+    {
+      "epoch": 0.8962753987507821,
+      "grad_norm": 1.3828125,
+      "learning_rate": 1.3232974087818955e-06,
+      "loss": 0.328,
+      "step": 20412
+    },
+    {
+      "epoch": 0.8963632172299856,
+      "grad_norm": 1.453125,
+      "learning_rate": 1.321077319944905e-06,
+      "loss": 0.3167,
+      "step": 20414
+    },
+    {
+      "epoch": 0.8964510357091892,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.3188590444454863e-06,
+      "loss": 0.3185,
+      "step": 20416
+    },
+    {
+      "epoch": 0.8965388541883926,
+      "grad_norm": 1.5,
+      "learning_rate": 1.3166425824535227e-06,
+      "loss": 0.2994,
+      "step": 20418
+    },
+    {
+      "epoch": 0.8966266726675961,
+      "grad_norm": 1.46875,
+      "learning_rate": 1.3144279341387427e-06,
+      "loss": 0.3108,
+      "step": 20420
+    },
+    {
+      "epoch": 0.8967144911467996,
+      "grad_norm": 1.4921875,
+      "learning_rate": 1.312215099670755e-06,
+      "loss": 0.3096,
+      "step": 20422
+    },
+    {
+      "epoch": 0.896802309626003,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.3100040792190127e-06,
+      "loss": 0.3016,
+      "step": 20424
+    },
+    {
+      "epoch": 0.8968901281052065,
+      "grad_norm": 1.4140625,
+      "learning_rate": 1.3077948729528333e-06,
+      "loss": 0.3203,
+      "step": 20426
+    },
+    {
+      "epoch": 0.89697794658441,
+      "grad_norm": 1.484375,
+      "learning_rate": 1.3055874810414037e-06,
+      "loss": 0.3113,
+      "step": 20428
+    },
+    {
+      "epoch": 0.8970657650636135,
+      "grad_norm": 1.421875,
+      "learning_rate": 1.3033819036537665e-06,
+      "loss": 0.3138,
+      "step": 20430
+    },
+    {
+      "epoch": 0.897153583542817,
+      "grad_norm": 1.484375,
+      "learning_rate": 1.3011781409588225e-06,
+      "loss": 0.336,
+      "step": 20432
+    },
+    {
+      "epoch": 0.8972414020220205,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.298976193125334e-06,
+      "loss": 0.304,
+      "step": 20434
+    },
+    {
+      "epoch": 0.897329220501224,
+      "grad_norm": 1.453125,
+      "learning_rate": 1.2967760603219358e-06,
+      "loss": 0.3237,
+      "step": 20436
+    },
+    {
+      "epoch": 0.8974170389804275,
+      "grad_norm": 1.453125,
+      "learning_rate": 1.294577742717104e-06,
+      "loss": 0.3243,
+      "step": 20438
+    },
+    {
+      "epoch": 0.8975048574596309,
+      "grad_norm": 1.4609375,
+      "learning_rate": 1.2923812404791958e-06,
+      "loss": 0.2941,
+      "step": 20440
+    },
+    {
+      "epoch": 0.8975926759388344,
+      "grad_norm": 1.4765625,
+      "learning_rate": 1.2901865537764124e-06,
+      "loss": 0.3097,
+      "step": 20442
+    },
+    {
+      "epoch": 0.8976804944180379,
+      "grad_norm": 1.453125,
+      "learning_rate": 1.2879936827768253e-06,
+      "loss": 0.3195,
+      "step": 20444
+    },
+    {
+      "epoch": 0.8977683128972413,
+      "grad_norm": 1.4296875,
+      "learning_rate": 1.2858026276483691e-06,
+      "loss": 0.3152,
+      "step": 20446
+    },
+    {
+      "epoch": 0.8978561313764449,
+      "grad_norm": 1.3828125,
+      "learning_rate": 1.2836133885588297e-06,
+      "loss": 0.3098,
+      "step": 20448
+    },
+    {
+      "epoch": 0.8979439498556484,
+      "grad_norm": 1.484375,
+      "learning_rate": 1.2814259656758643e-06,
+      "loss": 0.322,
+      "step": 20450
+    },
+    {
+      "epoch": 0.8980317683348519,
+      "grad_norm": 1.4765625,
+      "learning_rate": 1.2792403591669832e-06,
+      "loss": 0.2979,
+      "step": 20452
+    },
+    {
+      "epoch": 0.8981195868140553,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.277056569199561e-06,
+      "loss": 0.2918,
+      "step": 20454
+    },
+    {
+      "epoch": 0.8982074052932588,
+      "grad_norm": 1.4765625,
+      "learning_rate": 1.2748745959408365e-06,
+      "loss": 0.3074,
+      "step": 20456
+    },
+    {
+      "epoch": 0.8982952237724623,
+      "grad_norm": 1.40625,
+      "learning_rate": 1.2726944395578978e-06,
+      "loss": 0.2898,
+      "step": 20458
+    },
+    {
+      "epoch": 0.8983830422516658,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.270516100217714e-06,
+      "loss": 0.3297,
+      "step": 20460
+    },
+    {
+      "epoch": 0.8984708607308693,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.2683395780870883e-06,
+      "loss": 0.3182,
+      "step": 20462
+    },
+    {
+      "epoch": 0.8985586792100728,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.266164873332712e-06,
+      "loss": 0.3075,
+      "step": 20464
+    },
+    {
+      "epoch": 0.8986464976892763,
+      "grad_norm": 1.515625,
+      "learning_rate": 1.2639919861211158e-06,
+      "loss": 0.3312,
+      "step": 20466
+    },
+    {
+      "epoch": 0.8987343161684798,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.261820916618703e-06,
+      "loss": 0.3273,
+      "step": 20468
+    },
+    {
+      "epoch": 0.8988221346476832,
+      "grad_norm": 1.4296875,
+      "learning_rate": 1.2596516649917373e-06,
+      "loss": 0.3323,
+      "step": 20470
+    },
+    {
+      "epoch": 0.8989099531268867,
+      "grad_norm": 1.4609375,
+      "learning_rate": 1.2574842314063335e-06,
+      "loss": 0.3316,
+      "step": 20472
+    },
+    {
+      "epoch": 0.8989977716060902,
+      "grad_norm": 1.46875,
+      "learning_rate": 1.2553186160284837e-06,
+      "loss": 0.3054,
+      "step": 20474
+    },
+    {
+      "epoch": 0.8990855900852937,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.2531548190240244e-06,
+      "loss": 0.2856,
+      "step": 20476
+    },
+    {
+      "epoch": 0.8991734085644972,
+      "grad_norm": 1.4140625,
+      "learning_rate": 1.2509928405586596e-06,
+      "loss": 0.2936,
+      "step": 20478
+    },
+    {
+      "epoch": 0.8992612270437007,
+      "grad_norm": 1.4375,
+      "learning_rate": 1.2488326807979594e-06,
+      "loss": 0.3209,
+      "step": 20480
+    },
+    {
+      "epoch": 0.8993490455229042,
+      "grad_norm": 1.4921875,
+      "learning_rate": 1.2466743399073415e-06,
+      "loss": 0.3034,
+      "step": 20482
+    },
+    {
+      "epoch": 0.8994368640021076,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.2445178180521016e-06,
+      "loss": 0.3014,
+      "step": 20484
+    },
+    {
+      "epoch": 0.8995246824813111,
+      "grad_norm": 1.4921875,
+      "learning_rate": 1.2423631153973824e-06,
+      "loss": 0.3369,
+      "step": 20486
+    },
+    {
+      "epoch": 0.8996125009605146,
+      "grad_norm": 1.421875,
+      "learning_rate": 1.2402102321081854e-06,
+      "loss": 0.3253,
+      "step": 20488
+    },
+    {
+      "epoch": 0.8997003194397181,
+      "grad_norm": 1.46875,
+      "learning_rate": 1.2380591683493926e-06,
+      "loss": 0.3104,
+      "step": 20490
+    },
+    {
+      "epoch": 0.8997881379189215,
+      "grad_norm": 1.46875,
+      "learning_rate": 1.2359099242857164e-06,
+      "loss": 0.2955,
+      "step": 20492
+    },
+    {
+      "epoch": 0.8998759563981251,
+      "grad_norm": 1.4296875,
+      "learning_rate": 1.2337625000817616e-06,
+      "loss": 0.3055,
+      "step": 20494
+    },
+    {
+      "epoch": 0.8999637748773286,
+      "grad_norm": 1.5078125,
+      "learning_rate": 1.2316168959019658e-06,
+      "loss": 0.3278,
+      "step": 20496
+    },
+    {
+      "epoch": 0.9000515933565321,
+      "grad_norm": 1.484375,
+      "learning_rate": 1.2294731119106479e-06,
+      "loss": 0.319,
+      "step": 20498
+    },
+    {
+      "epoch": 0.9001394118357355,
+      "grad_norm": 1.5078125,
+      "learning_rate": 1.2273311482719764e-06,
+      "loss": 0.3076,
+      "step": 20500
+    },
+    {
+      "epoch": 0.900227230314939,
+      "grad_norm": 1.40625,
+      "learning_rate": 1.2251910051499865e-06,
+      "loss": 0.2989,
+      "step": 20502
+    },
+    {
+      "epoch": 0.9003150487941425,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.2230526827085698e-06,
+      "loss": 0.3129,
+      "step": 20504
+    },
+    {
+      "epoch": 0.900402867273346,
+      "grad_norm": 1.3984375,
+      "learning_rate": 1.2209161811114727e-06,
+      "loss": 0.3273,
+      "step": 20506
+    },
+    {
+      "epoch": 0.9004906857525495,
+      "grad_norm": 1.75,
+      "learning_rate": 1.2187815005223202e-06,
+      "loss": 0.3165,
+      "step": 20508
+    },
+    {
+      "epoch": 0.900578504231753,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.2166486411045786e-06,
+      "loss": 0.33,
+      "step": 20510
+    },
+    {
+      "epoch": 0.9006663227109565,
+      "grad_norm": 1.484375,
+      "learning_rate": 1.2145176030215866e-06,
+      "loss": 0.3129,
+      "step": 20512
+    },
+    {
+      "epoch": 0.90075414119016,
+      "grad_norm": 1.421875,
+      "learning_rate": 1.2123883864365388e-06,
+      "loss": 0.3177,
+      "step": 20514
+    },
+    {
+      "epoch": 0.9008419596693634,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.210260991512488e-06,
+      "loss": 0.3488,
+      "step": 20516
+    },
+    {
+      "epoch": 0.9009297781485669,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.208135418412354e-06,
+      "loss": 0.3142,
+      "step": 20518
+    },
+    {
+      "epoch": 0.9010175966277704,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.2060116672989118e-06,
+      "loss": 0.3306,
+      "step": 20520
+    },
+    {
+      "epoch": 0.9011054151069738,
+      "grad_norm": 1.453125,
+      "learning_rate": 1.203889738334807e-06,
+      "loss": 0.3044,
+      "step": 20522
+    },
+    {
+      "epoch": 0.9011932335861774,
+      "grad_norm": 1.515625,
+      "learning_rate": 1.2017696316825228e-06,
+      "loss": 0.3305,
+      "step": 20524
+    },
+    {
+      "epoch": 0.9012810520653809,
+      "grad_norm": 1.453125,
+      "learning_rate": 1.1996513475044297e-06,
+      "loss": 0.3275,
+      "step": 20526
+    },
+    {
+      "epoch": 0.9013688705445844,
+      "grad_norm": 1.515625,
+      "learning_rate": 1.1975348859627395e-06,
+      "loss": 0.3127,
+      "step": 20528
+    },
+    {
+      "epoch": 0.9014566890237878,
+      "grad_norm": 1.5078125,
+      "learning_rate": 1.195420247219531e-06,
+      "loss": 0.3118,
+      "step": 20530
+    },
+    {
+      "epoch": 0.9015445075029913,
+      "grad_norm": 1.453125,
+      "learning_rate": 1.193307431436752e-06,
+      "loss": 0.3125,
+      "step": 20532
+    },
+    {
+      "epoch": 0.9016323259821948,
+      "grad_norm": 1.4296875,
+      "learning_rate": 1.1911964387761904e-06,
+      "loss": 0.2804,
+      "step": 20534
+    },
+    {
+      "epoch": 0.9017201444613983,
+      "grad_norm": 1.4765625,
+      "learning_rate": 1.1890872693995165e-06,
+      "loss": 0.3351,
+      "step": 20536
+    },
+    {
+      "epoch": 0.9018079629406017,
+      "grad_norm": 1.4375,
+      "learning_rate": 1.1869799234682482e-06,
+      "loss": 0.3175,
+      "step": 20538
+    },
+    {
+      "epoch": 0.9018957814198053,
+      "grad_norm": 1.5078125,
+      "learning_rate": 1.1848744011437623e-06,
+      "loss": 0.3165,
+      "step": 20540
+    },
+    {
+      "epoch": 0.9019835998990088,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.1827707025873074e-06,
+      "loss": 0.3231,
+      "step": 20542
+    },
+    {
+      "epoch": 0.9020714183782123,
+      "grad_norm": 1.453125,
+      "learning_rate": 1.1806688279599798e-06,
+      "loss": 0.3045,
+      "step": 20544
+    },
+    {
+      "epoch": 0.9021592368574157,
+      "grad_norm": 1.5078125,
+      "learning_rate": 1.1785687774227422e-06,
+      "loss": 0.3261,
+      "step": 20546
+    },
+    {
+      "epoch": 0.9022470553366192,
+      "grad_norm": 1.484375,
+      "learning_rate": 1.1764705511364215e-06,
+      "loss": 0.2926,
+      "step": 20548
+    },
+    {
+      "epoch": 0.9023348738158227,
+      "grad_norm": 1.453125,
+      "learning_rate": 1.1743741492616922e-06,
+      "loss": 0.2985,
+      "step": 20550
+    },
+    {
+      "epoch": 0.9024226922950261,
+      "grad_norm": 1.4296875,
+      "learning_rate": 1.172279571959109e-06,
+      "loss": 0.2882,
+      "step": 20552
+    },
+    {
+      "epoch": 0.9025105107742296,
+      "grad_norm": 1.484375,
+      "learning_rate": 1.170186819389063e-06,
+      "loss": 0.3183,
+      "step": 20554
+    },
+    {
+      "epoch": 0.9025983292534332,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.1680958917118235e-06,
+      "loss": 0.2962,
+      "step": 20556
+    },
+    {
+      "epoch": 0.9026861477326367,
+      "grad_norm": 1.4765625,
+      "learning_rate": 1.1660067890875092e-06,
+      "loss": 0.3091,
+      "step": 20558
+    },
+    {
+      "epoch": 0.9027739662118401,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.1639195116761148e-06,
+      "loss": 0.328,
+      "step": 20560
+    },
+    {
+      "epoch": 0.9028617846910436,
+      "grad_norm": 1.46875,
+      "learning_rate": 1.161834059637476e-06,
+      "loss": 0.2942,
+      "step": 20562
+    },
+    {
+      "epoch": 0.9029496031702471,
+      "grad_norm": 1.5,
+      "learning_rate": 1.1597504331312986e-06,
+      "loss": 0.2917,
+      "step": 20564
+    },
+    {
+      "epoch": 0.9030374216494506,
+      "grad_norm": 1.4296875,
+      "learning_rate": 1.1576686323171493e-06,
+      "loss": 0.297,
+      "step": 20566
+    },
+    {
+      "epoch": 0.903125240128654,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.1555886573544478e-06,
+      "loss": 0.336,
+      "step": 20568
+    },
+    {
+      "epoch": 0.9032130586078576,
+      "grad_norm": 1.625,
+      "learning_rate": 1.1535105084024862e-06,
+      "loss": 0.3405,
+      "step": 20570
+    },
+    {
+      "epoch": 0.9033008770870611,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.1514341856204037e-06,
+      "loss": 0.3324,
+      "step": 20572
+    },
+    {
+      "epoch": 0.9033886955662646,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.149359689167212e-06,
+      "loss": 0.3195,
+      "step": 20574
+    },
+    {
+      "epoch": 0.903476514045468,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.1472870192017732e-06,
+      "loss": 0.319,
+      "step": 20576
+    },
+    {
+      "epoch": 0.9035643325246715,
+      "grad_norm": 1.484375,
+      "learning_rate": 1.1452161758828074e-06,
+      "loss": 0.3324,
+      "step": 20578
+    },
+    {
+      "epoch": 0.903652151003875,
+      "grad_norm": 1.4921875,
+      "learning_rate": 1.1431471593689097e-06,
+      "loss": 0.3128,
+      "step": 20580
+    },
+    {
+      "epoch": 0.9037399694830784,
+      "grad_norm": 1.4765625,
+      "learning_rate": 1.1410799698185204e-06,
+      "loss": 0.3055,
+      "step": 20582
+    },
+    {
+      "epoch": 0.9038277879622819,
+      "grad_norm": 1.46875,
+      "learning_rate": 1.139014607389946e-06,
+      "loss": 0.2876,
+      "step": 20584
+    },
+    {
+      "epoch": 0.9039156064414855,
+      "grad_norm": 1.4140625,
+      "learning_rate": 1.136951072241349e-06,
+      "loss": 0.3179,
+      "step": 20586
+    },
+    {
+      "epoch": 0.904003424920689,
+      "grad_norm": 1.46875,
+      "learning_rate": 1.134889364530764e-06,
+      "loss": 0.3113,
+      "step": 20588
+    },
+    {
+      "epoch": 0.9040912433998924,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.1328294844160732e-06,
+      "loss": 0.3197,
+      "step": 20590
+    },
+    {
+      "epoch": 0.9041790618790959,
+      "grad_norm": 1.4453125,
+      "learning_rate": 1.1307714320550167e-06,
+      "loss": 0.3094,
+      "step": 20592
+    },
+    {
+      "epoch": 0.9042668803582994,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.1287152076052104e-06,
+      "loss": 0.2775,
+      "step": 20594
+    },
+    {
+      "epoch": 0.9043546988375029,
+      "grad_norm": 1.4765625,
+      "learning_rate": 1.1266608112241118e-06,
+      "loss": 0.3146,
+      "step": 20596
+    },
+    {
+      "epoch": 0.9044425173167063,
+      "grad_norm": 1.3984375,
+      "learning_rate": 1.1246082430690558e-06,
+      "loss": 0.2975,
+      "step": 20598
+    },
+    {
+      "epoch": 0.9045303357959098,
+      "grad_norm": 1.453125,
+      "learning_rate": 1.1225575032972223e-06,
+      "loss": 0.306,
+      "step": 20600
+    },
+    {
+      "epoch": 0.9046181542751134,
+      "grad_norm": 1.4609375,
+      "learning_rate": 1.1205085920656556e-06,
+      "loss": 0.3098,
+      "step": 20602
+    },
+    {
+      "epoch": 0.9047059727543169,
+      "grad_norm": 1.453125,
+      "learning_rate": 1.1184615095312684e-06,
+      "loss": 0.3128,
+      "step": 20604
+    },
+    {
+      "epoch": 0.9047937912335203,
+      "grad_norm": 1.5078125,
+      "learning_rate": 1.1164162558508217e-06,
+      "loss": 0.2778,
+      "step": 20606
+    },
+    {
+      "epoch": 0.9048816097127238,
+      "grad_norm": 1.4453125,
+      "learning_rate": 1.114372831180946e-06,
+      "loss": 0.3463,
+      "step": 20608
+    },
+    {
+      "epoch": 0.9049694281919273,
+      "grad_norm": 1.515625,
+      "learning_rate": 1.112331235678124e-06,
+      "loss": 0.3002,
+      "step": 20610
+    },
+    {
+      "epoch": 0.9050572466711307,
+      "grad_norm": 1.4765625,
+      "learning_rate": 1.1102914694987004e-06,
+      "loss": 0.3101,
+      "step": 20612
+    },
+    {
+      "epoch": 0.9051450651503342,
+      "grad_norm": 1.5078125,
+      "learning_rate": 1.1082535327988864e-06,
+      "loss": 0.313,
+      "step": 20614
+    },
+    {
+      "epoch": 0.9052328836295378,
+      "grad_norm": 1.4765625,
+      "learning_rate": 1.1062174257347402e-06,
+      "loss": 0.3067,
+      "step": 20616
+    },
+    {
+      "epoch": 0.9053207021087413,
+      "grad_norm": 1.515625,
+      "learning_rate": 1.1041831484621956e-06,
+      "loss": 0.3436,
+      "step": 20618
+    },
+    {
+      "epoch": 0.9054085205879447,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.102150701137028e-06,
+      "loss": 0.3385,
+      "step": 20620
+    },
+    {
+      "epoch": 0.9054963390671482,
+      "grad_norm": 1.4765625,
+      "learning_rate": 1.1001200839148934e-06,
+      "loss": 0.3233,
+      "step": 20622
+    },
+    {
+      "epoch": 0.9055841575463517,
+      "grad_norm": 1.484375,
+      "learning_rate": 1.0980912969512897e-06,
+      "loss": 0.3154,
+      "step": 20624
+    },
+    {
+      "epoch": 0.9056719760255552,
+      "grad_norm": 1.4921875,
+      "learning_rate": 1.0960643404015813e-06,
+      "loss": 0.3367,
+      "step": 20626
+    },
+    {
+      "epoch": 0.9057597945047586,
+      "grad_norm": 1.515625,
+      "learning_rate": 1.0940392144210027e-06,
+      "loss": 0.3401,
+      "step": 20628
+    },
+    {
+      "epoch": 0.9058476129839621,
+      "grad_norm": 1.4765625,
+      "learning_rate": 1.092015919164624e-06,
+      "loss": 0.3034,
+      "step": 20630
+    },
+    {
+      "epoch": 0.9059354314631657,
+      "grad_norm": 1.46875,
+      "learning_rate": 1.089994454787402e-06,
+      "loss": 0.3058,
+      "step": 20632
+    },
+    {
+      "epoch": 0.9060232499423692,
+      "grad_norm": 1.4765625,
+      "learning_rate": 1.0879748214441348e-06,
+      "loss": 0.3177,
+      "step": 20634
+    },
+    {
+      "epoch": 0.9061110684215726,
+      "grad_norm": 1.515625,
+      "learning_rate": 1.0859570192894908e-06,
+      "loss": 0.32,
+      "step": 20636
+    },
+    {
+      "epoch": 0.9061988869007761,
+      "grad_norm": 1.375,
+      "learning_rate": 1.0839410484779876e-06,
+      "loss": 0.2909,
+      "step": 20638
+    },
+    {
+      "epoch": 0.9062867053799796,
+      "grad_norm": 1.453125,
+      "learning_rate": 1.0819269091640134e-06,
+      "loss": 0.3363,
+      "step": 20640
+    },
+    {
+      "epoch": 0.906374523859183,
+      "grad_norm": 1.421875,
+      "learning_rate": 1.0799146015018109e-06,
+      "loss": 0.3271,
+      "step": 20642
+    },
+    {
+      "epoch": 0.9064623423383865,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.077904125645482e-06,
+      "loss": 0.3047,
+      "step": 20644
+    },
+    {
+      "epoch": 0.90655016081759,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.0758954817489897e-06,
+      "loss": 0.3081,
+      "step": 20646
+    },
+    {
+      "epoch": 0.9066379792967936,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.0738886699661528e-06,
+      "loss": 0.2996,
+      "step": 20648
+    },
+    {
+      "epoch": 0.906725797775997,
+      "grad_norm": 1.4296875,
+      "learning_rate": 1.0718836904506617e-06,
+      "loss": 0.3251,
+      "step": 20650
+    },
+    {
+      "epoch": 0.9068136162552005,
+      "grad_norm": 1.4609375,
+      "learning_rate": 1.0698805433560528e-06,
+      "loss": 0.3055,
+      "step": 20652
+    },
+    {
+      "epoch": 0.906901434734404,
+      "grad_norm": 1.5,
+      "learning_rate": 1.0678792288357249e-06,
+      "loss": 0.3177,
+      "step": 20654
+    },
+    {
+      "epoch": 0.9069892532136075,
+      "grad_norm": 1.484375,
+      "learning_rate": 1.0658797470429443e-06,
+      "loss": 0.3137,
+      "step": 20656
+    },
+    {
+      "epoch": 0.9070770716928109,
+      "grad_norm": 1.4609375,
+      "learning_rate": 1.0638820981308305e-06,
+      "loss": 0.3099,
+      "step": 20658
+    },
+    {
+      "epoch": 0.9071648901720144,
+      "grad_norm": 1.4609375,
+      "learning_rate": 1.0618862822523639e-06,
+      "loss": 0.3337,
+      "step": 20660
+    },
+    {
+      "epoch": 0.907252708651218,
+      "grad_norm": 1.4375,
+      "learning_rate": 1.0598922995603861e-06,
+      "loss": 0.3277,
+      "step": 20662
+    },
+    {
+      "epoch": 0.9073405271304215,
+      "grad_norm": 1.46875,
+      "learning_rate": 1.0579001502075887e-06,
+      "loss": 0.334,
+      "step": 20664
+    },
+    {
+      "epoch": 0.9074283456096249,
+      "grad_norm": 1.46875,
+      "learning_rate": 1.055909834346544e-06,
+      "loss": 0.3086,
+      "step": 20666
+    },
+    {
+      "epoch": 0.9075161640888284,
+      "grad_norm": 1.453125,
+      "learning_rate": 1.0539213521296582e-06,
+      "loss": 0.3132,
+      "step": 20668
+    },
+    {
+      "epoch": 0.9076039825680319,
+      "grad_norm": 1.4296875,
+      "learning_rate": 1.0519347037092175e-06,
+      "loss": 0.3115,
+      "step": 20670
+    },
+    {
+      "epoch": 0.9076918010472353,
+      "grad_norm": 1.4765625,
+      "learning_rate": 1.0499498892373616e-06,
+      "loss": 0.3169,
+      "step": 20672
+    },
+    {
+      "epoch": 0.9077796195264388,
+      "grad_norm": 1.515625,
+      "learning_rate": 1.0479669088660827e-06,
+      "loss": 0.3559,
+      "step": 20674
+    },
+    {
+      "epoch": 0.9078674380056423,
+      "grad_norm": 1.46875,
+      "learning_rate": 1.0459857627472396e-06,
+      "loss": 0.3287,
+      "step": 20676
+    },
+    {
+      "epoch": 0.9079552564848459,
+      "grad_norm": 1.515625,
+      "learning_rate": 1.0440064510325448e-06,
+      "loss": 0.3209,
+      "step": 20678
+    },
+    {
+      "epoch": 0.9080430749640493,
+      "grad_norm": 1.4765625,
+      "learning_rate": 1.0420289738735822e-06,
+      "loss": 0.3123,
+      "step": 20680
+    },
+    {
+      "epoch": 0.9081308934432528,
+      "grad_norm": 1.4375,
+      "learning_rate": 1.0400533314217837e-06,
+      "loss": 0.332,
+      "step": 20682
+    },
+    {
+      "epoch": 0.9082187119224563,
+      "grad_norm": 1.484375,
+      "learning_rate": 1.0380795238284446e-06,
+      "loss": 0.299,
+      "step": 20684
+    },
+    {
+      "epoch": 0.9083065304016598,
+      "grad_norm": 1.4921875,
+      "learning_rate": 1.0361075512447193e-06,
+      "loss": 0.3013,
+      "step": 20686
+    },
+    {
+      "epoch": 0.9083943488808632,
+      "grad_norm": 1.4921875,
+      "learning_rate": 1.0341374138216203e-06,
+      "loss": 0.314,
+      "step": 20688
+    },
+    {
+      "epoch": 0.9084821673600667,
+      "grad_norm": 1.4921875,
+      "learning_rate": 1.0321691117100268e-06,
+      "loss": 0.3044,
+      "step": 20690
+    },
+    {
+      "epoch": 0.9085699858392702,
+      "grad_norm": 1.4296875,
+      "learning_rate": 1.0302026450606656e-06,
+      "loss": 0.3105,
+      "step": 20692
+    },
+    {
+      "epoch": 0.9086578043184738,
+      "grad_norm": 1.484375,
+      "learning_rate": 1.0282380140241326e-06,
+      "loss": 0.3198,
+      "step": 20694
+    },
+    {
+      "epoch": 0.9087456227976772,
+      "grad_norm": 1.421875,
+      "learning_rate": 1.02627521875088e-06,
+      "loss": 0.3752,
+      "step": 20696
+    },
+    {
+      "epoch": 0.9088334412768807,
+      "grad_norm": 1.4453125,
+      "learning_rate": 1.0243142593912153e-06,
+      "loss": 0.3131,
+      "step": 20698
+    },
+    {
+      "epoch": 0.9089212597560842,
+      "grad_norm": 1.421875,
+      "learning_rate": 1.0223551360953154e-06,
+      "loss": 0.2975,
+      "step": 20700
+    },
+    {
+      "epoch": 0.9090090782352876,
+      "grad_norm": 1.421875,
+      "learning_rate": 1.020397849013205e-06,
+      "loss": 0.3036,
+      "step": 20702
+    },
+    {
+      "epoch": 0.9090968967144911,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.018442398294775e-06,
+      "loss": 0.285,
+      "step": 20704
+    },
+    {
+      "epoch": 0.9091847151936946,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.016488784089778e-06,
+      "loss": 0.3174,
+      "step": 20706
+    },
+    {
+      "epoch": 0.9092725336728981,
+      "grad_norm": 1.5078125,
+      "learning_rate": 1.0145370065478194e-06,
+      "loss": 0.3333,
+      "step": 20708
+    },
+    {
+      "epoch": 0.9093603521521016,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.01258706581836e-06,
+      "loss": 0.311,
+      "step": 20710
+    },
+    {
+      "epoch": 0.9094481706313051,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.010638962050739e-06,
+      "loss": 0.3376,
+      "step": 20712
+    },
+    {
+      "epoch": 0.9095359891105086,
+      "grad_norm": 1.46875,
+      "learning_rate": 1.0086926953941368e-06,
+      "loss": 0.3444,
+      "step": 20714
+    },
+    {
+      "epoch": 0.9096238075897121,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.0067482659975953e-06,
+      "loss": 0.2992,
+      "step": 20716
+    },
+    {
+      "epoch": 0.9097116260689155,
+      "grad_norm": 1.4609375,
+      "learning_rate": 1.0048056740100286e-06,
+      "loss": 0.337,
+      "step": 20718
+    },
+    {
+      "epoch": 0.909799444548119,
+      "grad_norm": 1.3984375,
+      "learning_rate": 1.0028649195801903e-06,
+      "loss": 0.3046,
+      "step": 20720
+    },
+    {
+      "epoch": 0.9098872630273225,
+      "grad_norm": 1.4453125,
+      "learning_rate": 1.0009260028567113e-06,
+      "loss": 0.3162,
+      "step": 20722
+    },
+    {
+      "epoch": 0.9099750815065261,
+      "grad_norm": 1.421875,
+      "learning_rate": 9.989889239880729e-07,
+      "loss": 0.301,
+      "step": 20724
+    },
+    {
+      "epoch": 0.9100628999857295,
+      "grad_norm": 1.4453125,
+      "learning_rate": 9.970536831226145e-07,
+      "loss": 0.301,
+      "step": 20726
+    },
+    {
+      "epoch": 0.910150718464933,
+      "grad_norm": 1.484375,
+      "learning_rate": 9.951202804085402e-07,
+      "loss": 0.3301,
+      "step": 20728
+    },
+    {
+      "epoch": 0.9102385369441365,
+      "grad_norm": 1.5546875,
+      "learning_rate": 9.931887159939062e-07,
+      "loss": 0.302,
+      "step": 20730
+    },
+    {
+      "epoch": 0.91032635542334,
+      "grad_norm": 1.5234375,
+      "learning_rate": 9.91258990026639e-07,
+      "loss": 0.292,
+      "step": 20732
+    },
+    {
+      "epoch": 0.9104141739025434,
+      "grad_norm": 1.4921875,
+      "learning_rate": 9.893311026545116e-07,
+      "loss": 0.33,
+      "step": 20734
+    },
+    {
+      "epoch": 0.9105019923817469,
+      "grad_norm": 1.453125,
+      "learning_rate": 9.874050540251672e-07,
+      "loss": 0.2974,
+      "step": 20736
+    },
+    {
+      "epoch": 0.9105898108609504,
+      "grad_norm": 1.4921875,
+      "learning_rate": 9.85480844286099e-07,
+      "loss": 0.3126,
+      "step": 20738
+    },
+    {
+      "epoch": 0.910677629340154,
+      "grad_norm": 1.4609375,
+      "learning_rate": 9.835584735846588e-07,
+      "loss": 0.3131,
+      "step": 20740
+    },
+    {
+      "epoch": 0.9107654478193574,
+      "grad_norm": 1.375,
+      "learning_rate": 9.816379420680727e-07,
+      "loss": 0.3123,
+      "step": 20742
+    },
+    {
+      "epoch": 0.9108532662985609,
+      "grad_norm": 1.4609375,
+      "learning_rate": 9.797192498834096e-07,
+      "loss": 0.3226,
+      "step": 20744
+    },
+    {
+      "epoch": 0.9109410847777644,
+      "grad_norm": 1.453125,
+      "learning_rate": 9.778023971776045e-07,
+      "loss": 0.324,
+      "step": 20746
+    },
+    {
+      "epoch": 0.9110289032569678,
+      "grad_norm": 1.53125,
+      "learning_rate": 9.758873840974514e-07,
+      "loss": 0.3356,
+      "step": 20748
+    },
+    {
+      "epoch": 0.9111167217361713,
+      "grad_norm": 1.53125,
+      "learning_rate": 9.739742107895994e-07,
+      "loss": 0.3209,
+      "step": 20750
+    },
+    {
+      "epoch": 0.9112045402153748,
+      "grad_norm": 1.546875,
+      "learning_rate": 9.720628774005647e-07,
+      "loss": 0.2962,
+      "step": 20752
+    },
+    {
+      "epoch": 0.9112923586945783,
+      "grad_norm": 1.5078125,
+      "learning_rate": 9.701533840767108e-07,
+      "loss": 0.3132,
+      "step": 20754
+    },
+    {
+      "epoch": 0.9113801771737818,
+      "grad_norm": 1.5,
+      "learning_rate": 9.682457309642735e-07,
+      "loss": 0.2998,
+      "step": 20756
+    },
+    {
+      "epoch": 0.9114679956529853,
+      "grad_norm": 1.4140625,
+      "learning_rate": 9.663399182093386e-07,
+      "loss": 0.338,
+      "step": 20758
+    },
+    {
+      "epoch": 0.9115558141321888,
+      "grad_norm": 1.4140625,
+      "learning_rate": 9.644359459578533e-07,
+      "loss": 0.329,
+      "step": 20760
+    },
+    {
+      "epoch": 0.9116436326113923,
+      "grad_norm": 1.4921875,
+      "learning_rate": 9.62533814355626e-07,
+      "loss": 0.2874,
+      "step": 20762
+    },
+    {
+      "epoch": 0.9117314510905957,
+      "grad_norm": 1.46875,
+      "learning_rate": 9.606335235483182e-07,
+      "loss": 0.3186,
+      "step": 20764
+    },
+    {
+      "epoch": 0.9118192695697992,
+      "grad_norm": 1.453125,
+      "learning_rate": 9.58735073681466e-07,
+      "loss": 0.3224,
+      "step": 20766
+    },
+    {
+      "epoch": 0.9119070880490027,
+      "grad_norm": 1.453125,
+      "learning_rate": 9.568384649004363e-07,
+      "loss": 0.337,
+      "step": 20768
+    },
+    {
+      "epoch": 0.9119949065282063,
+      "grad_norm": 1.4375,
+      "learning_rate": 9.549436973504855e-07,
+      "loss": 0.2885,
+      "step": 20770
+    },
+    {
+      "epoch": 0.9120827250074097,
+      "grad_norm": 1.53125,
+      "learning_rate": 9.530507711767056e-07,
+      "loss": 0.3455,
+      "step": 20772
+    },
+    {
+      "epoch": 0.9121705434866132,
+      "grad_norm": 1.4921875,
+      "learning_rate": 9.511596865240669e-07,
+      "loss": 0.3165,
+      "step": 20774
+    },
+    {
+      "epoch": 0.9122583619658167,
+      "grad_norm": 1.4609375,
+      "learning_rate": 9.49270443537384e-07,
+      "loss": 0.294,
+      "step": 20776
+    },
+    {
+      "epoch": 0.9123461804450201,
+      "grad_norm": 1.4296875,
+      "learning_rate": 9.473830423613328e-07,
+      "loss": 0.339,
+      "step": 20778
+    },
+    {
+      "epoch": 0.9124339989242236,
+      "grad_norm": 1.4296875,
+      "learning_rate": 9.454974831404561e-07,
+      "loss": 0.3194,
+      "step": 20780
+    },
+    {
+      "epoch": 0.9125218174034271,
+      "grad_norm": 1.421875,
+      "learning_rate": 9.436137660191469e-07,
+      "loss": 0.3044,
+      "step": 20782
+    },
+    {
+      "epoch": 0.9126096358826306,
+      "grad_norm": 1.53125,
+      "learning_rate": 9.417318911416644e-07,
+      "loss": 0.315,
+      "step": 20784
+    },
+    {
+      "epoch": 0.9126974543618341,
+      "grad_norm": 1.5,
+      "learning_rate": 9.398518586521188e-07,
+      "loss": 0.2958,
+      "step": 20786
+    },
+    {
+      "epoch": 0.9127852728410376,
+      "grad_norm": 1.4296875,
+      "learning_rate": 9.379736686944862e-07,
+      "loss": 0.305,
+      "step": 20788
+    },
+    {
+      "epoch": 0.9128730913202411,
+      "grad_norm": 1.5234375,
+      "learning_rate": 9.36097321412599e-07,
+      "loss": 0.322,
+      "step": 20790
+    },
+    {
+      "epoch": 0.9129609097994446,
+      "grad_norm": 1.4453125,
+      "learning_rate": 9.342228169501449e-07,
+      "loss": 0.3145,
+      "step": 20792
+    },
+    {
+      "epoch": 0.913048728278648,
+      "grad_norm": 1.53125,
+      "learning_rate": 9.323501554506786e-07,
+      "loss": 0.2959,
+      "step": 20794
+    },
+    {
+      "epoch": 0.9131365467578515,
+      "grad_norm": 1.484375,
+      "learning_rate": 9.304793370576076e-07,
+      "loss": 0.311,
+      "step": 20796
+    },
+    {
+      "epoch": 0.913224365237055,
+      "grad_norm": 1.546875,
+      "learning_rate": 9.286103619141978e-07,
+      "loss": 0.3461,
+      "step": 20798
+    },
+    {
+      "epoch": 0.9133121837162584,
+      "grad_norm": 1.4765625,
+      "learning_rate": 9.267432301635792e-07,
+      "loss": 0.2936,
+      "step": 20800
+    },
+    {
+      "epoch": 0.913400002195462,
+      "grad_norm": 1.421875,
+      "learning_rate": 9.248779419487292e-07,
+      "loss": 0.289,
+      "step": 20802
+    },
+    {
+      "epoch": 0.9134878206746655,
+      "grad_norm": 1.453125,
+      "learning_rate": 9.230144974125027e-07,
+      "loss": 0.3029,
+      "step": 20804
+    },
+    {
+      "epoch": 0.913575639153869,
+      "grad_norm": 1.453125,
+      "learning_rate": 9.211528966975941e-07,
+      "loss": 0.3303,
+      "step": 20806
+    },
+    {
+      "epoch": 0.9136634576330724,
+      "grad_norm": 1.4375,
+      "learning_rate": 9.192931399465698e-07,
+      "loss": 0.3011,
+      "step": 20808
+    },
+    {
+      "epoch": 0.9137512761122759,
+      "grad_norm": 1.46875,
+      "learning_rate": 9.174352273018521e-07,
+      "loss": 0.3299,
+      "step": 20810
+    },
+    {
+      "epoch": 0.9138390945914794,
+      "grad_norm": 1.4375,
+      "learning_rate": 9.155791589057133e-07,
+      "loss": 0.3313,
+      "step": 20812
+    },
+    {
+      "epoch": 0.9139269130706829,
+      "grad_norm": 1.3984375,
+      "learning_rate": 9.137249349002979e-07,
+      "loss": 0.3033,
+      "step": 20814
+    },
+    {
+      "epoch": 0.9140147315498864,
+      "grad_norm": 1.4453125,
+      "learning_rate": 9.118725554276008e-07,
+      "loss": 0.3177,
+      "step": 20816
+    },
+    {
+      "epoch": 0.9141025500290899,
+      "grad_norm": 1.546875,
+      "learning_rate": 9.10022020629478e-07,
+      "loss": 0.3122,
+      "step": 20818
+    },
+    {
+      "epoch": 0.9141903685082934,
+      "grad_norm": 1.5234375,
+      "learning_rate": 9.081733306476437e-07,
+      "loss": 0.3234,
+      "step": 20820
+    },
+    {
+      "epoch": 0.9142781869874969,
+      "grad_norm": 1.5,
+      "learning_rate": 9.063264856236708e-07,
+      "loss": 0.3054,
+      "step": 20822
+    },
+    {
+      "epoch": 0.9143660054667003,
+      "grad_norm": 1.4296875,
+      "learning_rate": 9.044814856989908e-07,
+      "loss": 0.3372,
+      "step": 20824
+    },
+    {
+      "epoch": 0.9144538239459038,
+      "grad_norm": 1.4296875,
+      "learning_rate": 9.026383310148933e-07,
+      "loss": 0.3083,
+      "step": 20826
+    },
+    {
+      "epoch": 0.9145416424251073,
+      "grad_norm": 1.4375,
+      "learning_rate": 9.00797021712535e-07,
+      "loss": 0.323,
+      "step": 20828
+    },
+    {
+      "epoch": 0.9146294609043107,
+      "grad_norm": 1.640625,
+      "learning_rate": 8.989575579329113e-07,
+      "loss": 0.3073,
+      "step": 20830
+    },
+    {
+      "epoch": 0.9147172793835143,
+      "grad_norm": 1.484375,
+      "learning_rate": 8.971199398168983e-07,
+      "loss": 0.3294,
+      "step": 20832
+    },
+    {
+      "epoch": 0.9148050978627178,
+      "grad_norm": 1.4921875,
+      "learning_rate": 8.95284167505217e-07,
+      "loss": 0.3088,
+      "step": 20834
+    },
+    {
+      "epoch": 0.9148929163419213,
+      "grad_norm": 1.4765625,
+      "learning_rate": 8.934502411384549e-07,
+      "loss": 0.3042,
+      "step": 20836
+    },
+    {
+      "epoch": 0.9149807348211247,
+      "grad_norm": 1.4921875,
+      "learning_rate": 8.916181608570495e-07,
+      "loss": 0.3142,
+      "step": 20838
+    },
+    {
+      "epoch": 0.9150685533003282,
+      "grad_norm": 1.4296875,
+      "learning_rate": 8.897879268013027e-07,
+      "loss": 0.3137,
+      "step": 20840
+    },
+    {
+      "epoch": 0.9151563717795317,
+      "grad_norm": 1.484375,
+      "learning_rate": 8.879595391113798e-07,
+      "loss": 0.3189,
+      "step": 20842
+    },
+    {
+      "epoch": 0.9152441902587352,
+      "grad_norm": 1.4453125,
+      "learning_rate": 8.861329979272915e-07,
+      "loss": 0.3058,
+      "step": 20844
+    },
+    {
+      "epoch": 0.9153320087379386,
+      "grad_norm": 1.390625,
+      "learning_rate": 8.843083033889227e-07,
+      "loss": 0.3186,
+      "step": 20846
+    },
+    {
+      "epoch": 0.9154198272171422,
+      "grad_norm": 1.46875,
+      "learning_rate": 8.824854556360062e-07,
+      "loss": 0.3289,
+      "step": 20848
+    },
+    {
+      "epoch": 0.9155076456963457,
+      "grad_norm": 1.40625,
+      "learning_rate": 8.806644548081305e-07,
+      "loss": 0.3385,
+      "step": 20850
+    },
+    {
+      "epoch": 0.9155954641755492,
+      "grad_norm": 1.53125,
+      "learning_rate": 8.788453010447534e-07,
+      "loss": 0.3497,
+      "step": 20852
+    },
+    {
+      "epoch": 0.9156832826547526,
+      "grad_norm": 1.4375,
+      "learning_rate": 8.770279944851856e-07,
+      "loss": 0.3319,
+      "step": 20854
+    },
+    {
+      "epoch": 0.9157711011339561,
+      "grad_norm": 1.4765625,
+      "learning_rate": 8.752125352685992e-07,
+      "loss": 0.3535,
+      "step": 20856
+    },
+    {
+      "epoch": 0.9158589196131596,
+      "grad_norm": 1.4921875,
+      "learning_rate": 8.73398923534019e-07,
+      "loss": 0.3008,
+      "step": 20858
+    },
+    {
+      "epoch": 0.915946738092363,
+      "grad_norm": 1.421875,
+      "learning_rate": 8.715871594203312e-07,
+      "loss": 0.3212,
+      "step": 20860
+    },
+    {
+      "epoch": 0.9160345565715666,
+      "grad_norm": 1.53125,
+      "learning_rate": 8.697772430662859e-07,
+      "loss": 0.2935,
+      "step": 20862
+    },
+    {
+      "epoch": 0.9161223750507701,
+      "grad_norm": 1.4140625,
+      "learning_rate": 8.679691746104807e-07,
+      "loss": 0.3038,
+      "step": 20864
+    },
+    {
+      "epoch": 0.9162101935299736,
+      "grad_norm": 1.46875,
+      "learning_rate": 8.661629541913824e-07,
+      "loss": 0.2944,
+      "step": 20866
+    },
+    {
+      "epoch": 0.916298012009177,
+      "grad_norm": 1.5546875,
+      "learning_rate": 8.643585819473055e-07,
+      "loss": 0.3247,
+      "step": 20868
+    },
+    {
+      "epoch": 0.9163858304883805,
+      "grad_norm": 1.4765625,
+      "learning_rate": 8.625560580164394e-07,
+      "loss": 0.3316,
+      "step": 20870
+    },
+    {
+      "epoch": 0.916473648967584,
+      "grad_norm": 1.5390625,
+      "learning_rate": 8.60755382536818e-07,
+      "loss": 0.2913,
+      "step": 20872
+    },
+    {
+      "epoch": 0.9165614674467875,
+      "grad_norm": 1.578125,
+      "learning_rate": 8.589565556463314e-07,
+      "loss": 0.346,
+      "step": 20874
+    },
+    {
+      "epoch": 0.9166492859259909,
+      "grad_norm": 1.53125,
+      "learning_rate": 8.571595774827413e-07,
+      "loss": 0.3234,
+      "step": 20876
+    },
+    {
+      "epoch": 0.9167371044051945,
+      "grad_norm": 1.515625,
+      "learning_rate": 8.553644481836542e-07,
+      "loss": 0.2973,
+      "step": 20878
+    },
+    {
+      "epoch": 0.916824922884398,
+      "grad_norm": 1.609375,
+      "learning_rate": 8.535711678865493e-07,
+      "loss": 0.2843,
+      "step": 20880
+    },
+    {
+      "epoch": 0.9169127413636015,
+      "grad_norm": 1.4921875,
+      "learning_rate": 8.517797367287555e-07,
+      "loss": 0.3122,
+      "step": 20882
+    },
+    {
+      "epoch": 0.9170005598428049,
+      "grad_norm": 1.4296875,
+      "learning_rate": 8.49990154847452e-07,
+      "loss": 0.2951,
+      "step": 20884
+    },
+    {
+      "epoch": 0.9170883783220084,
+      "grad_norm": 1.4453125,
+      "learning_rate": 8.482024223796958e-07,
+      "loss": 0.3027,
+      "step": 20886
+    },
+    {
+      "epoch": 0.9171761968012119,
+      "grad_norm": 1.4375,
+      "learning_rate": 8.464165394623829e-07,
+      "loss": 0.3013,
+      "step": 20888
+    },
+    {
+      "epoch": 0.9172640152804153,
+      "grad_norm": 1.5,
+      "learning_rate": 8.446325062322902e-07,
+      "loss": 0.3374,
+      "step": 20890
+    },
+    {
+      "epoch": 0.9173518337596188,
+      "grad_norm": 1.46875,
+      "learning_rate": 8.428503228260221e-07,
+      "loss": 0.3171,
+      "step": 20892
+    },
+    {
+      "epoch": 0.9174396522388224,
+      "grad_norm": 1.421875,
+      "learning_rate": 8.410699893800722e-07,
+      "loss": 0.3274,
+      "step": 20894
+    },
+    {
+      "epoch": 0.9175274707180259,
+      "grad_norm": 1.4921875,
+      "learning_rate": 8.392915060307704e-07,
+      "loss": 0.2995,
+      "step": 20896
+    },
+    {
+      "epoch": 0.9176152891972293,
+      "grad_norm": 1.453125,
+      "learning_rate": 8.37514872914319e-07,
+      "loss": 0.3186,
+      "step": 20898
+    },
+    {
+      "epoch": 0.9177031076764328,
+      "grad_norm": 1.4765625,
+      "learning_rate": 8.3574009016677e-07,
+      "loss": 0.3063,
+      "step": 20900
+    },
+    {
+      "epoch": 0.9177909261556363,
+      "grad_norm": 1.546875,
+      "learning_rate": 8.339671579240371e-07,
+      "loss": 0.3366,
+      "step": 20902
+    },
+    {
+      "epoch": 0.9178787446348398,
+      "grad_norm": 1.546875,
+      "learning_rate": 8.321960763218922e-07,
+      "loss": 0.3206,
+      "step": 20904
+    },
+    {
+      "epoch": 0.9179665631140432,
+      "grad_norm": 1.484375,
+      "learning_rate": 8.304268454959657e-07,
+      "loss": 0.3006,
+      "step": 20906
+    },
+    {
+      "epoch": 0.9180543815932467,
+      "grad_norm": 1.453125,
+      "learning_rate": 8.286594655817465e-07,
+      "loss": 0.3366,
+      "step": 20908
+    },
+    {
+      "epoch": 0.9181422000724503,
+      "grad_norm": 1.40625,
+      "learning_rate": 8.268939367145789e-07,
+      "loss": 0.3098,
+      "step": 20910
+    },
+    {
+      "epoch": 0.9182300185516538,
+      "grad_norm": 1.4765625,
+      "learning_rate": 8.251302590296661e-07,
+      "loss": 0.3572,
+      "step": 20912
+    },
+    {
+      "epoch": 0.9183178370308572,
+      "grad_norm": 1.4296875,
+      "learning_rate": 8.23368432662075e-07,
+      "loss": 0.271,
+      "step": 20914
+    },
+    {
+      "epoch": 0.9184056555100607,
+      "grad_norm": 1.515625,
+      "learning_rate": 8.216084577467226e-07,
+      "loss": 0.3109,
+      "step": 20916
+    },
+    {
+      "epoch": 0.9184934739892642,
+      "grad_norm": 1.46875,
+      "learning_rate": 8.198503344183955e-07,
+      "loss": 0.3085,
+      "step": 20918
+    },
+    {
+      "epoch": 0.9185812924684676,
+      "grad_norm": 1.4921875,
+      "learning_rate": 8.180940628117223e-07,
+      "loss": 0.3106,
+      "step": 20920
+    },
+    {
+      "epoch": 0.9186691109476711,
+      "grad_norm": 1.484375,
+      "learning_rate": 8.163396430612063e-07,
+      "loss": 0.3375,
+      "step": 20922
+    },
+    {
+      "epoch": 0.9187569294268747,
+      "grad_norm": 1.5,
+      "learning_rate": 8.145870753011958e-07,
+      "loss": 0.3231,
+      "step": 20924
+    },
+    {
+      "epoch": 0.9188447479060782,
+      "grad_norm": 1.5,
+      "learning_rate": 8.128363596659e-07,
+      "loss": 0.2954,
+      "step": 20926
+    },
+    {
+      "epoch": 0.9189325663852816,
+      "grad_norm": 1.46875,
+      "learning_rate": 8.110874962894005e-07,
+      "loss": 0.3181,
+      "step": 20928
+    },
+    {
+      "epoch": 0.9190203848644851,
+      "grad_norm": 1.484375,
+      "learning_rate": 8.093404853056125e-07,
+      "loss": 0.3185,
+      "step": 20930
+    },
+    {
+      "epoch": 0.9191082033436886,
+      "grad_norm": 1.4296875,
+      "learning_rate": 8.075953268483344e-07,
+      "loss": 0.3389,
+      "step": 20932
+    },
+    {
+      "epoch": 0.9191960218228921,
+      "grad_norm": 1.5078125,
+      "learning_rate": 8.058520210512066e-07,
+      "loss": 0.3182,
+      "step": 20934
+    },
+    {
+      "epoch": 0.9192838403020955,
+      "grad_norm": 1.59375,
+      "learning_rate": 8.041105680477251e-07,
+      "loss": 0.3405,
+      "step": 20936
+    },
+    {
+      "epoch": 0.919371658781299,
+      "grad_norm": 1.453125,
+      "learning_rate": 8.02370967971261e-07,
+      "loss": 0.3015,
+      "step": 20938
+    },
+    {
+      "epoch": 0.9194594772605026,
+      "grad_norm": 1.4296875,
+      "learning_rate": 8.006332209550244e-07,
+      "loss": 0.297,
+      "step": 20940
+    },
+    {
+      "epoch": 0.9195472957397061,
+      "grad_norm": 1.4453125,
+      "learning_rate": 7.988973271321004e-07,
+      "loss": 0.3436,
+      "step": 20942
+    },
+    {
+      "epoch": 0.9196351142189095,
+      "grad_norm": 1.421875,
+      "learning_rate": 7.971632866354189e-07,
+      "loss": 0.3226,
+      "step": 20944
+    },
+    {
+      "epoch": 0.919722932698113,
+      "grad_norm": 1.4296875,
+      "learning_rate": 7.954310995977737e-07,
+      "loss": 0.3318,
+      "step": 20946
+    },
+    {
+      "epoch": 0.9198107511773165,
+      "grad_norm": 1.421875,
+      "learning_rate": 7.937007661518198e-07,
+      "loss": 0.3187,
+      "step": 20948
+    },
+    {
+      "epoch": 0.91989856965652,
+      "grad_norm": 1.4609375,
+      "learning_rate": 7.919722864300649e-07,
+      "loss": 0.3513,
+      "step": 20950
+    },
+    {
+      "epoch": 0.9199863881357234,
+      "grad_norm": 1.4609375,
+      "learning_rate": 7.902456605648756e-07,
+      "loss": 0.3281,
+      "step": 20952
+    },
+    {
+      "epoch": 0.9200742066149269,
+      "grad_norm": 1.4140625,
+      "learning_rate": 7.885208886884709e-07,
+      "loss": 0.2876,
+      "step": 20954
+    },
+    {
+      "epoch": 0.9201620250941305,
+      "grad_norm": 1.5,
+      "learning_rate": 7.867979709329454e-07,
+      "loss": 0.2869,
+      "step": 20956
+    },
+    {
+      "epoch": 0.920249843573334,
+      "grad_norm": 1.4609375,
+      "learning_rate": 7.85076907430235e-07,
+      "loss": 0.3085,
+      "step": 20958
+    },
+    {
+      "epoch": 0.9203376620525374,
+      "grad_norm": 1.5625,
+      "learning_rate": 7.833576983121399e-07,
+      "loss": 0.3395,
+      "step": 20960
+    },
+    {
+      "epoch": 0.9204254805317409,
+      "grad_norm": 1.453125,
+      "learning_rate": 7.816403437103159e-07,
+      "loss": 0.3473,
+      "step": 20962
+    },
+    {
+      "epoch": 0.9205132990109444,
+      "grad_norm": 1.4921875,
+      "learning_rate": 7.799248437562801e-07,
+      "loss": 0.2932,
+      "step": 20964
+    },
+    {
+      "epoch": 0.9206011174901478,
+      "grad_norm": 1.40625,
+      "learning_rate": 7.782111985814077e-07,
+      "loss": 0.31,
+      "step": 20966
+    },
+    {
+      "epoch": 0.9206889359693513,
+      "grad_norm": 1.53125,
+      "learning_rate": 7.764994083169247e-07,
+      "loss": 0.3037,
+      "step": 20968
+    },
+    {
+      "epoch": 0.9207767544485549,
+      "grad_norm": 1.4296875,
+      "learning_rate": 7.747894730939259e-07,
+      "loss": 0.3276,
+      "step": 20970
+    },
+    {
+      "epoch": 0.9208645729277584,
+      "grad_norm": 1.40625,
+      "learning_rate": 7.730813930433567e-07,
+      "loss": 0.285,
+      "step": 20972
+    },
+    {
+      "epoch": 0.9209523914069618,
+      "grad_norm": 1.484375,
+      "learning_rate": 7.713751682960207e-07,
+      "loss": 0.3125,
+      "step": 20974
+    },
+    {
+      "epoch": 0.9210402098861653,
+      "grad_norm": 1.4375,
+      "learning_rate": 7.69670798982583e-07,
+      "loss": 0.3246,
+      "step": 20976
+    },
+    {
+      "epoch": 0.9211280283653688,
+      "grad_norm": 1.6015625,
+      "learning_rate": 7.67968285233564e-07,
+      "loss": 0.3272,
+      "step": 20978
+    },
+    {
+      "epoch": 0.9212158468445723,
+      "grad_norm": 1.609375,
+      "learning_rate": 7.662676271793429e-07,
+      "loss": 0.2933,
+      "step": 20980
+    },
+    {
+      "epoch": 0.9213036653237757,
+      "grad_norm": 1.5625,
+      "learning_rate": 7.64568824950157e-07,
+      "loss": 0.3296,
+      "step": 20982
+    },
+    {
+      "epoch": 0.9213914838029792,
+      "grad_norm": 1.46875,
+      "learning_rate": 7.628718786760997e-07,
+      "loss": 0.3073,
+      "step": 20984
+    },
+    {
+      "epoch": 0.9214793022821828,
+      "grad_norm": 1.4375,
+      "learning_rate": 7.611767884871251e-07,
+      "loss": 0.3272,
+      "step": 20986
+    },
+    {
+      "epoch": 0.9215671207613862,
+      "grad_norm": 1.5703125,
+      "learning_rate": 7.59483554513038e-07,
+      "loss": 0.3255,
+      "step": 20988
+    },
+    {
+      "epoch": 0.9216549392405897,
+      "grad_norm": 1.453125,
+      "learning_rate": 7.57792176883515e-07,
+      "loss": 0.3086,
+      "step": 20990
+    },
+    {
+      "epoch": 0.9217427577197932,
+      "grad_norm": 1.453125,
+      "learning_rate": 7.561026557280748e-07,
+      "loss": 0.3066,
+      "step": 20992
+    },
+    {
+      "epoch": 0.9218305761989967,
+      "grad_norm": 1.46875,
+      "learning_rate": 7.544149911761084e-07,
+      "loss": 0.3216,
+      "step": 20994
+    },
+    {
+      "epoch": 0.9219183946782001,
+      "grad_norm": 1.4140625,
+      "learning_rate": 7.527291833568539e-07,
+      "loss": 0.3202,
+      "step": 20996
+    },
+    {
+      "epoch": 0.9220062131574036,
+      "grad_norm": 1.4296875,
+      "learning_rate": 7.510452323994083e-07,
+      "loss": 0.3295,
+      "step": 20998
+    },
+    {
+      "epoch": 0.9220940316366071,
+      "grad_norm": 1.4296875,
+      "learning_rate": 7.493631384327348e-07,
+      "loss": 0.2978,
+      "step": 21000
+    },
+    {
+      "epoch": 0.9221818501158107,
+      "grad_norm": 1.46875,
+      "learning_rate": 7.476829015856446e-07,
+      "loss": 0.3027,
+      "step": 21002
+    },
+    {
+      "epoch": 0.9222696685950141,
+      "grad_norm": 1.4453125,
+      "learning_rate": 7.460045219868095e-07,
+      "loss": 0.3039,
+      "step": 21004
+    },
+    {
+      "epoch": 0.9223574870742176,
+      "grad_norm": 1.59375,
+      "learning_rate": 7.443279997647657e-07,
+      "loss": 0.3202,
+      "step": 21006
+    },
+    {
+      "epoch": 0.9224453055534211,
+      "grad_norm": 1.4453125,
+      "learning_rate": 7.426533350478937e-07,
+      "loss": 0.2813,
+      "step": 21008
+    },
+    {
+      "epoch": 0.9225331240326246,
+      "grad_norm": 1.46875,
+      "learning_rate": 7.409805279644494e-07,
+      "loss": 0.3119,
+      "step": 21010
+    },
+    {
+      "epoch": 0.922620942511828,
+      "grad_norm": 1.453125,
+      "learning_rate": 7.393095786425275e-07,
+      "loss": 0.3038,
+      "step": 21012
+    },
+    {
+      "epoch": 0.9227087609910315,
+      "grad_norm": 1.46875,
+      "learning_rate": 7.376404872100978e-07,
+      "loss": 0.3249,
+      "step": 21014
+    },
+    {
+      "epoch": 0.9227965794702351,
+      "grad_norm": 1.53125,
+      "learning_rate": 7.359732537949693e-07,
+      "loss": 0.3023,
+      "step": 21016
+    },
+    {
+      "epoch": 0.9228843979494386,
+      "grad_norm": 1.3984375,
+      "learning_rate": 7.343078785248315e-07,
+      "loss": 0.3033,
+      "step": 21018
+    },
+    {
+      "epoch": 0.922972216428642,
+      "grad_norm": 1.5,
+      "learning_rate": 7.326443615272099e-07,
+      "loss": 0.3115,
+      "step": 21020
+    },
+    {
+      "epoch": 0.9230600349078455,
+      "grad_norm": 1.546875,
+      "learning_rate": 7.309827029295002e-07,
+      "loss": 0.3266,
+      "step": 21022
+    },
+    {
+      "epoch": 0.923147853387049,
+      "grad_norm": 1.4609375,
+      "learning_rate": 7.29322902858956e-07,
+      "loss": 0.3321,
+      "step": 21024
+    },
+    {
+      "epoch": 0.9232356718662524,
+      "grad_norm": 1.6640625,
+      "learning_rate": 7.276649614426784e-07,
+      "loss": 0.3087,
+      "step": 21026
+    },
+    {
+      "epoch": 0.9233234903454559,
+      "grad_norm": 1.390625,
+      "learning_rate": 7.26008878807638e-07,
+      "loss": 0.3076,
+      "step": 21028
+    },
+    {
+      "epoch": 0.9234113088246594,
+      "grad_norm": 1.4765625,
+      "learning_rate": 7.243546550806557e-07,
+      "loss": 0.3151,
+      "step": 21030
+    },
+    {
+      "epoch": 0.923499127303863,
+      "grad_norm": 1.5078125,
+      "learning_rate": 7.227022903884134e-07,
+      "loss": 0.2999,
+      "step": 21032
+    },
+    {
+      "epoch": 0.9235869457830664,
+      "grad_norm": 1.4453125,
+      "learning_rate": 7.210517848574516e-07,
+      "loss": 0.335,
+      "step": 21034
+    },
+    {
+      "epoch": 0.9236747642622699,
+      "grad_norm": 1.5078125,
+      "learning_rate": 7.194031386141608e-07,
+      "loss": 0.3007,
+      "step": 21036
+    },
+    {
+      "epoch": 0.9237625827414734,
+      "grad_norm": 1.3828125,
+      "learning_rate": 7.177563517848013e-07,
+      "loss": 0.2695,
+      "step": 21038
+    },
+    {
+      "epoch": 0.9238504012206769,
+      "grad_norm": 1.515625,
+      "learning_rate": 7.161114244954775e-07,
+      "loss": 0.3298,
+      "step": 21040
+    },
+    {
+      "epoch": 0.9239382196998803,
+      "grad_norm": 1.4296875,
+      "learning_rate": 7.144683568721694e-07,
+      "loss": 0.3179,
+      "step": 21042
+    },
+    {
+      "epoch": 0.9240260381790838,
+      "grad_norm": 1.5234375,
+      "learning_rate": 7.128271490406873e-07,
+      "loss": 0.3385,
+      "step": 21044
+    },
+    {
+      "epoch": 0.9241138566582873,
+      "grad_norm": 1.4375,
+      "learning_rate": 7.111878011267309e-07,
+      "loss": 0.3113,
+      "step": 21046
+    },
+    {
+      "epoch": 0.9242016751374909,
+      "grad_norm": 1.453125,
+      "learning_rate": 7.095503132558329e-07,
+      "loss": 0.3134,
+      "step": 21048
+    },
+    {
+      "epoch": 0.9242894936166943,
+      "grad_norm": 1.40625,
+      "learning_rate": 7.07914685553393e-07,
+      "loss": 0.2692,
+      "step": 21050
+    },
+    {
+      "epoch": 0.9243773120958978,
+      "grad_norm": 1.5078125,
+      "learning_rate": 7.062809181446695e-07,
+      "loss": 0.3174,
+      "step": 21052
+    },
+    {
+      "epoch": 0.9244651305751013,
+      "grad_norm": 1.5,
+      "learning_rate": 7.046490111547788e-07,
+      "loss": 0.3495,
+      "step": 21054
+    },
+    {
+      "epoch": 0.9245529490543047,
+      "grad_norm": 1.4453125,
+      "learning_rate": 7.030189647086904e-07,
+      "loss": 0.3221,
+      "step": 21056
+    },
+    {
+      "epoch": 0.9246407675335082,
+      "grad_norm": 1.46875,
+      "learning_rate": 7.013907789312352e-07,
+      "loss": 0.3218,
+      "step": 21058
+    },
+    {
+      "epoch": 0.9247285860127117,
+      "grad_norm": 1.390625,
+      "learning_rate": 6.997644539470938e-07,
+      "loss": 0.3023,
+      "step": 21060
+    },
+    {
+      "epoch": 0.9248164044919152,
+      "grad_norm": 1.515625,
+      "learning_rate": 6.981399898808222e-07,
+      "loss": 0.3328,
+      "step": 21062
+    },
+    {
+      "epoch": 0.9249042229711187,
+      "grad_norm": 1.5234375,
+      "learning_rate": 6.965173868568098e-07,
+      "loss": 0.3167,
+      "step": 21064
+    },
+    {
+      "epoch": 0.9249920414503222,
+      "grad_norm": 1.4921875,
+      "learning_rate": 6.948966449993266e-07,
+      "loss": 0.3245,
+      "step": 21066
+    },
+    {
+      "epoch": 0.9250798599295257,
+      "grad_norm": 1.4453125,
+      "learning_rate": 6.932777644324844e-07,
+      "loss": 0.3237,
+      "step": 21068
+    },
+    {
+      "epoch": 0.9251676784087292,
+      "grad_norm": 1.4453125,
+      "learning_rate": 6.916607452802538e-07,
+      "loss": 0.319,
+      "step": 21070
+    },
+    {
+      "epoch": 0.9252554968879326,
+      "grad_norm": 1.5,
+      "learning_rate": 6.90045587666474e-07,
+      "loss": 0.3122,
+      "step": 21072
+    },
+    {
+      "epoch": 0.9253433153671361,
+      "grad_norm": 1.5859375,
+      "learning_rate": 6.884322917148328e-07,
+      "loss": 0.3193,
+      "step": 21074
+    },
+    {
+      "epoch": 0.9254311338463396,
+      "grad_norm": 1.46875,
+      "learning_rate": 6.868208575488699e-07,
+      "loss": 0.3228,
+      "step": 21076
+    },
+    {
+      "epoch": 0.9255189523255432,
+      "grad_norm": 1.4765625,
+      "learning_rate": 6.852112852919951e-07,
+      "loss": 0.3195,
+      "step": 21078
+    },
+    {
+      "epoch": 0.9256067708047466,
+      "grad_norm": 1.5234375,
+      "learning_rate": 6.836035750674708e-07,
+      "loss": 0.3395,
+      "step": 21080
+    },
+    {
+      "epoch": 0.9256945892839501,
+      "grad_norm": 1.4765625,
+      "learning_rate": 6.819977269984123e-07,
+      "loss": 0.3088,
+      "step": 21082
+    },
+    {
+      "epoch": 0.9257824077631536,
+      "grad_norm": 1.5234375,
+      "learning_rate": 6.803937412077965e-07,
+      "loss": 0.3288,
+      "step": 21084
+    },
+    {
+      "epoch": 0.925870226242357,
+      "grad_norm": 1.5,
+      "learning_rate": 6.787916178184583e-07,
+      "loss": 0.3166,
+      "step": 21086
+    },
+    {
+      "epoch": 0.9259580447215605,
+      "grad_norm": 1.4140625,
+      "learning_rate": 6.771913569530857e-07,
+      "loss": 0.2714,
+      "step": 21088
+    },
+    {
+      "epoch": 0.926045863200764,
+      "grad_norm": 1.453125,
+      "learning_rate": 6.755929587342336e-07,
+      "loss": 0.3068,
+      "step": 21090
+    },
+    {
+      "epoch": 0.9261336816799675,
+      "grad_norm": 1.453125,
+      "learning_rate": 6.739964232843038e-07,
+      "loss": 0.3145,
+      "step": 21092
+    },
+    {
+      "epoch": 0.926221500159171,
+      "grad_norm": 1.4921875,
+      "learning_rate": 6.72401750725557e-07,
+      "loss": 0.3062,
+      "step": 21094
+    },
+    {
+      "epoch": 0.9263093186383745,
+      "grad_norm": 1.5078125,
+      "learning_rate": 6.708089411801177e-07,
+      "loss": 0.2989,
+      "step": 21096
+    },
+    {
+      "epoch": 0.926397137117578,
+      "grad_norm": 1.421875,
+      "learning_rate": 6.692179947699579e-07,
+      "loss": 0.2991,
+      "step": 21098
+    },
+    {
+      "epoch": 0.9264849555967815,
+      "grad_norm": 1.5703125,
+      "learning_rate": 6.676289116169188e-07,
+      "loss": 0.342,
+      "step": 21100
+    },
+    {
+      "epoch": 0.9265727740759849,
+      "grad_norm": 1.4765625,
+      "learning_rate": 6.660416918426892e-07,
+      "loss": 0.2854,
+      "step": 21102
+    },
+    {
+      "epoch": 0.9266605925551884,
+      "grad_norm": 1.4296875,
+      "learning_rate": 6.644563355688277e-07,
+      "loss": 0.3511,
+      "step": 21104
+    },
+    {
+      "epoch": 0.9267484110343919,
+      "grad_norm": 1.53125,
+      "learning_rate": 6.628728429167258e-07,
+      "loss": 0.3465,
+      "step": 21106
+    },
+    {
+      "epoch": 0.9268362295135953,
+      "grad_norm": 1.578125,
+      "learning_rate": 6.612912140076588e-07,
+      "loss": 0.3519,
+      "step": 21108
+    },
+    {
+      "epoch": 0.9269240479927989,
+      "grad_norm": 1.515625,
+      "learning_rate": 6.597114489627437e-07,
+      "loss": 0.3272,
+      "step": 21110
+    },
+    {
+      "epoch": 0.9270118664720024,
+      "grad_norm": 1.4453125,
+      "learning_rate": 6.581335479029588e-07,
+      "loss": 0.2801,
+      "step": 21112
+    },
+    {
+      "epoch": 0.9270996849512059,
+      "grad_norm": 1.484375,
+      "learning_rate": 6.565575109491462e-07,
+      "loss": 0.3199,
+      "step": 21114
+    },
+    {
+      "epoch": 0.9271875034304093,
+      "grad_norm": 1.46875,
+      "learning_rate": 6.5498333822199e-07,
+      "loss": 0.2785,
+      "step": 21116
+    },
+    {
+      "epoch": 0.9272753219096128,
+      "grad_norm": 1.40625,
+      "learning_rate": 6.534110298420493e-07,
+      "loss": 0.3323,
+      "step": 21118
+    },
+    {
+      "epoch": 0.9273631403888163,
+      "grad_norm": 1.453125,
+      "learning_rate": 6.518405859297277e-07,
+      "loss": 0.3108,
+      "step": 21120
+    },
+    {
+      "epoch": 0.9274509588680198,
+      "grad_norm": 1.53125,
+      "learning_rate": 6.502720066052903e-07,
+      "loss": 0.315,
+      "step": 21122
+    },
+    {
+      "epoch": 0.9275387773472233,
+      "grad_norm": 1.421875,
+      "learning_rate": 6.487052919888603e-07,
+      "loss": 0.3075,
+      "step": 21124
+    },
+    {
+      "epoch": 0.9276265958264268,
+      "grad_norm": 1.4140625,
+      "learning_rate": 6.47140442200414e-07,
+      "loss": 0.3272,
+      "step": 21126
+    },
+    {
+      "epoch": 0.9277144143056303,
+      "grad_norm": 1.3984375,
+      "learning_rate": 6.455774573597917e-07,
+      "loss": 0.3082,
+      "step": 21128
+    },
+    {
+      "epoch": 0.9278022327848338,
+      "grad_norm": 1.4765625,
+      "learning_rate": 6.440163375866892e-07,
+      "loss": 0.3059,
+      "step": 21130
+    },
+    {
+      "epoch": 0.9278900512640372,
+      "grad_norm": 1.546875,
+      "learning_rate": 6.424570830006498e-07,
+      "loss": 0.3184,
+      "step": 21132
+    },
+    {
+      "epoch": 0.9279778697432407,
+      "grad_norm": 1.46875,
+      "learning_rate": 6.408996937210892e-07,
+      "loss": 0.3014,
+      "step": 21134
+    },
+    {
+      "epoch": 0.9280656882224442,
+      "grad_norm": 1.4453125,
+      "learning_rate": 6.393441698672647e-07,
+      "loss": 0.3257,
+      "step": 21136
+    },
+    {
+      "epoch": 0.9281535067016476,
+      "grad_norm": 1.4296875,
+      "learning_rate": 6.377905115583088e-07,
+      "loss": 0.2957,
+      "step": 21138
+    },
+    {
+      "epoch": 0.9282413251808512,
+      "grad_norm": 1.484375,
+      "learning_rate": 6.362387189131902e-07,
+      "loss": 0.2928,
+      "step": 21140
+    },
+    {
+      "epoch": 0.9283291436600547,
+      "grad_norm": 1.4765625,
+      "learning_rate": 6.346887920507555e-07,
+      "loss": 0.3254,
+      "step": 21142
+    },
+    {
+      "epoch": 0.9284169621392582,
+      "grad_norm": 1.4921875,
+      "learning_rate": 6.33140731089693e-07,
+      "loss": 0.3156,
+      "step": 21144
+    },
+    {
+      "epoch": 0.9285047806184616,
+      "grad_norm": 1.4375,
+      "learning_rate": 6.315945361485498e-07,
+      "loss": 0.3211,
+      "step": 21146
+    },
+    {
+      "epoch": 0.9285925990976651,
+      "grad_norm": 1.5078125,
+      "learning_rate": 6.300502073457448e-07,
+      "loss": 0.3218,
+      "step": 21148
+    },
+    {
+      "epoch": 0.9286804175768686,
+      "grad_norm": 1.515625,
+      "learning_rate": 6.285077447995307e-07,
+      "loss": 0.2848,
+      "step": 21150
+    },
+    {
+      "epoch": 0.9287682360560721,
+      "grad_norm": 1.453125,
+      "learning_rate": 6.26967148628041e-07,
+      "loss": 0.306,
+      "step": 21152
+    },
+    {
+      "epoch": 0.9288560545352755,
+      "grad_norm": 1.515625,
+      "learning_rate": 6.254284189492476e-07,
+      "loss": 0.3131,
+      "step": 21154
+    },
+    {
+      "epoch": 0.9289438730144791,
+      "grad_norm": 1.4609375,
+      "learning_rate": 6.238915558809899e-07,
+      "loss": 0.3059,
+      "step": 21156
+    },
+    {
+      "epoch": 0.9290316914936826,
+      "grad_norm": 1.46875,
+      "learning_rate": 6.223565595409597e-07,
+      "loss": 0.2828,
+      "step": 21158
+    },
+    {
+      "epoch": 0.9291195099728861,
+      "grad_norm": 1.53125,
+      "learning_rate": 6.208234300467048e-07,
+      "loss": 0.3251,
+      "step": 21160
+    },
+    {
+      "epoch": 0.9292073284520895,
+      "grad_norm": 1.46875,
+      "learning_rate": 6.192921675156394e-07,
+      "loss": 0.3106,
+      "step": 21162
+    },
+    {
+      "epoch": 0.929295146931293,
+      "grad_norm": 1.4765625,
+      "learning_rate": 6.177627720650226e-07,
+      "loss": 0.3275,
+      "step": 21164
+    },
+    {
+      "epoch": 0.9293829654104965,
+      "grad_norm": 1.5078125,
+      "learning_rate": 6.1623524381198e-07,
+      "loss": 0.2936,
+      "step": 21166
+    },
+    {
+      "epoch": 0.9294707838897,
+      "grad_norm": 1.453125,
+      "learning_rate": 6.14709582873485e-07,
+      "loss": 0.2886,
+      "step": 21168
+    },
+    {
+      "epoch": 0.9295586023689035,
+      "grad_norm": 1.46875,
+      "learning_rate": 6.131857893663772e-07,
+      "loss": 0.3234,
+      "step": 21170
+    },
+    {
+      "epoch": 0.929646420848107,
+      "grad_norm": 1.515625,
+      "learning_rate": 6.116638634073496e-07,
+      "loss": 0.3274,
+      "step": 21172
+    },
+    {
+      "epoch": 0.9297342393273105,
+      "grad_norm": 1.4921875,
+      "learning_rate": 6.101438051129449e-07,
+      "loss": 0.3255,
+      "step": 21174
+    },
+    {
+      "epoch": 0.929822057806514,
+      "grad_norm": 1.5078125,
+      "learning_rate": 6.086256145995783e-07,
+      "loss": 0.3073,
+      "step": 21176
+    },
+    {
+      "epoch": 0.9299098762857174,
+      "grad_norm": 1.5234375,
+      "learning_rate": 6.071092919835042e-07,
+      "loss": 0.3081,
+      "step": 21178
+    },
+    {
+      "epoch": 0.9299976947649209,
+      "grad_norm": 1.46875,
+      "learning_rate": 6.055948373808517e-07,
+      "loss": 0.3165,
+      "step": 21180
+    },
+    {
+      "epoch": 0.9300855132441244,
+      "grad_norm": 1.46875,
+      "learning_rate": 6.04082250907595e-07,
+      "loss": 0.3216,
+      "step": 21182
+    },
+    {
+      "epoch": 0.9301733317233278,
+      "grad_norm": 1.4375,
+      "learning_rate": 6.025715326795633e-07,
+      "loss": 0.3113,
+      "step": 21184
+    },
+    {
+      "epoch": 0.9302611502025314,
+      "grad_norm": 1.6015625,
+      "learning_rate": 6.01062682812456e-07,
+      "loss": 0.3103,
+      "step": 21186
+    },
+    {
+      "epoch": 0.9303489686817349,
+      "grad_norm": 1.4453125,
+      "learning_rate": 5.995557014218168e-07,
+      "loss": 0.2916,
+      "step": 21188
+    },
+    {
+      "epoch": 0.9304367871609384,
+      "grad_norm": 1.484375,
+      "learning_rate": 5.980505886230503e-07,
+      "loss": 0.2992,
+      "step": 21190
+    },
+    {
+      "epoch": 0.9305246056401418,
+      "grad_norm": 1.5390625,
+      "learning_rate": 5.965473445314201e-07,
+      "loss": 0.2912,
+      "step": 21192
+    },
+    {
+      "epoch": 0.9306124241193453,
+      "grad_norm": 1.515625,
+      "learning_rate": 5.950459692620425e-07,
+      "loss": 0.2946,
+      "step": 21194
+    },
+    {
+      "epoch": 0.9307002425985488,
+      "grad_norm": 1.5546875,
+      "learning_rate": 5.935464629298975e-07,
+      "loss": 0.3201,
+      "step": 21196
+    },
+    {
+      "epoch": 0.9307880610777522,
+      "grad_norm": 1.4765625,
+      "learning_rate": 5.920488256498131e-07,
+      "loss": 0.3221,
+      "step": 21198
+    },
+    {
+      "epoch": 0.9308758795569557,
+      "grad_norm": 1.4609375,
+      "learning_rate": 5.905530575364831e-07,
+      "loss": 0.3377,
+      "step": 21200
+    },
+    {
+      "epoch": 0.9309636980361593,
+      "grad_norm": 1.484375,
+      "learning_rate": 5.89059158704447e-07,
+      "loss": 0.3286,
+      "step": 21202
+    },
+    {
+      "epoch": 0.9310515165153628,
+      "grad_norm": 1.6640625,
+      "learning_rate": 5.875671292681157e-07,
+      "loss": 0.2927,
+      "step": 21204
+    },
+    {
+      "epoch": 0.9311393349945662,
+      "grad_norm": 1.5,
+      "learning_rate": 5.860769693417451e-07,
+      "loss": 0.3109,
+      "step": 21206
+    },
+    {
+      "epoch": 0.9312271534737697,
+      "grad_norm": 1.546875,
+      "learning_rate": 5.845886790394495e-07,
+      "loss": 0.3093,
+      "step": 21208
+    },
+    {
+      "epoch": 0.9313149719529732,
+      "grad_norm": 1.5234375,
+      "learning_rate": 5.8310225847521e-07,
+      "loss": 0.3427,
+      "step": 21210
+    },
+    {
+      "epoch": 0.9314027904321767,
+      "grad_norm": 1.4921875,
+      "learning_rate": 5.816177077628493e-07,
+      "loss": 0.3308,
+      "step": 21212
+    },
+    {
+      "epoch": 0.9314906089113801,
+      "grad_norm": 1.5546875,
+      "learning_rate": 5.801350270160599e-07,
+      "loss": 0.3064,
+      "step": 21214
+    },
+    {
+      "epoch": 0.9315784273905837,
+      "grad_norm": 1.515625,
+      "learning_rate": 5.786542163483843e-07,
+      "loss": 0.3105,
+      "step": 21216
+    },
+    {
+      "epoch": 0.9316662458697872,
+      "grad_norm": 1.5,
+      "learning_rate": 5.771752758732207e-07,
+      "loss": 0.3202,
+      "step": 21218
+    },
+    {
+      "epoch": 0.9317540643489907,
+      "grad_norm": 1.421875,
+      "learning_rate": 5.756982057038312e-07,
+      "loss": 0.3233,
+      "step": 21220
+    },
+    {
+      "epoch": 0.9318418828281941,
+      "grad_norm": 1.453125,
+      "learning_rate": 5.742230059533255e-07,
+      "loss": 0.3054,
+      "step": 21222
+    },
+    {
+      "epoch": 0.9319297013073976,
+      "grad_norm": 1.46875,
+      "learning_rate": 5.727496767346796e-07,
+      "loss": 0.3086,
+      "step": 21224
+    },
+    {
+      "epoch": 0.9320175197866011,
+      "grad_norm": 1.4375,
+      "learning_rate": 5.712782181607202e-07,
+      "loss": 0.3207,
+      "step": 21226
+    },
+    {
+      "epoch": 0.9321053382658046,
+      "grad_norm": 1.5703125,
+      "learning_rate": 5.698086303441292e-07,
+      "loss": 0.3119,
+      "step": 21228
+    },
+    {
+      "epoch": 0.932193156745008,
+      "grad_norm": 1.515625,
+      "learning_rate": 5.683409133974499e-07,
+      "loss": 0.304,
+      "step": 21230
+    },
+    {
+      "epoch": 0.9322809752242116,
+      "grad_norm": 1.546875,
+      "learning_rate": 5.668750674330786e-07,
+      "loss": 0.3022,
+      "step": 21232
+    },
+    {
+      "epoch": 0.9323687937034151,
+      "grad_norm": 1.578125,
+      "learning_rate": 5.654110925632756e-07,
+      "loss": 0.2937,
+      "step": 21234
+    },
+    {
+      "epoch": 0.9324566121826185,
+      "grad_norm": 1.5703125,
+      "learning_rate": 5.639489889001426e-07,
+      "loss": 0.3227,
+      "step": 21236
+    },
+    {
+      "epoch": 0.932544430661822,
+      "grad_norm": 1.4921875,
+      "learning_rate": 5.624887565556596e-07,
+      "loss": 0.3311,
+      "step": 21238
+    },
+    {
+      "epoch": 0.9326322491410255,
+      "grad_norm": 1.4375,
+      "learning_rate": 5.610303956416402e-07,
+      "loss": 0.3161,
+      "step": 21240
+    },
+    {
+      "epoch": 0.932720067620229,
+      "grad_norm": 1.5,
+      "learning_rate": 5.595739062697752e-07,
+      "loss": 0.2872,
+      "step": 21242
+    },
+    {
+      "epoch": 0.9328078860994324,
+      "grad_norm": 1.4609375,
+      "learning_rate": 5.581192885516006e-07,
+      "loss": 0.3168,
+      "step": 21244
+    },
+    {
+      "epoch": 0.9328957045786359,
+      "grad_norm": 1.515625,
+      "learning_rate": 5.566665425985052e-07,
+      "loss": 0.3393,
+      "step": 21246
+    },
+    {
+      "epoch": 0.9329835230578395,
+      "grad_norm": 1.4140625,
+      "learning_rate": 5.552156685217497e-07,
+      "loss": 0.3114,
+      "step": 21248
+    },
+    {
+      "epoch": 0.933071341537043,
+      "grad_norm": 1.46875,
+      "learning_rate": 5.537666664324342e-07,
+      "loss": 0.2953,
+      "step": 21250
+    },
+    {
+      "epoch": 0.9331591600162464,
+      "grad_norm": 1.4296875,
+      "learning_rate": 5.523195364415312e-07,
+      "loss": 0.3083,
+      "step": 21252
+    },
+    {
+      "epoch": 0.9332469784954499,
+      "grad_norm": 1.5,
+      "learning_rate": 5.508742786598575e-07,
+      "loss": 0.3182,
+      "step": 21254
+    },
+    {
+      "epoch": 0.9333347969746534,
+      "grad_norm": 1.484375,
+      "learning_rate": 5.494308931980913e-07,
+      "loss": 0.3156,
+      "step": 21256
+    },
+    {
+      "epoch": 0.9334226154538569,
+      "grad_norm": 1.4375,
+      "learning_rate": 5.479893801667718e-07,
+      "loss": 0.321,
+      "step": 21258
+    },
+    {
+      "epoch": 0.9335104339330603,
+      "grad_norm": 1.4921875,
+      "learning_rate": 5.465497396762831e-07,
+      "loss": 0.3114,
+      "step": 21260
+    },
+    {
+      "epoch": 0.9335982524122638,
+      "grad_norm": 1.484375,
+      "learning_rate": 5.451119718368786e-07,
+      "loss": 0.3205,
+      "step": 21262
+    },
+    {
+      "epoch": 0.9336860708914674,
+      "grad_norm": 1.4921875,
+      "learning_rate": 5.436760767586618e-07,
+      "loss": 0.3173,
+      "step": 21264
+    },
+    {
+      "epoch": 0.9337738893706709,
+      "grad_norm": 1.4921875,
+      "learning_rate": 5.422420545515949e-07,
+      "loss": 0.2962,
+      "step": 21266
+    },
+    {
+      "epoch": 0.9338617078498743,
+      "grad_norm": 1.421875,
+      "learning_rate": 5.408099053254929e-07,
+      "loss": 0.3042,
+      "step": 21268
+    },
+    {
+      "epoch": 0.9339495263290778,
+      "grad_norm": 1.4140625,
+      "learning_rate": 5.393796291900316e-07,
+      "loss": 0.3332,
+      "step": 21270
+    },
+    {
+      "epoch": 0.9340373448082813,
+      "grad_norm": 1.5078125,
+      "learning_rate": 5.379512262547431e-07,
+      "loss": 0.3221,
+      "step": 21272
+    },
+    {
+      "epoch": 0.9341251632874847,
+      "grad_norm": 1.4609375,
+      "learning_rate": 5.365246966290094e-07,
+      "loss": 0.3113,
+      "step": 21274
+    },
+    {
+      "epoch": 0.9342129817666882,
+      "grad_norm": 1.5546875,
+      "learning_rate": 5.351000404220846e-07,
+      "loss": 0.2909,
+      "step": 21276
+    },
+    {
+      "epoch": 0.9343008002458918,
+      "grad_norm": 1.453125,
+      "learning_rate": 5.336772577430593e-07,
+      "loss": 0.2881,
+      "step": 21278
+    },
+    {
+      "epoch": 0.9343886187250953,
+      "grad_norm": 1.4453125,
+      "learning_rate": 5.322563487008964e-07,
+      "loss": 0.3233,
+      "step": 21280
+    },
+    {
+      "epoch": 0.9344764372042987,
+      "grad_norm": 1.40625,
+      "learning_rate": 5.308373134044059e-07,
+      "loss": 0.297,
+      "step": 21282
+    },
+    {
+      "epoch": 0.9345642556835022,
+      "grad_norm": 1.421875,
+      "learning_rate": 5.294201519622594e-07,
+      "loss": 0.3024,
+      "step": 21284
+    },
+    {
+      "epoch": 0.9346520741627057,
+      "grad_norm": 1.4296875,
+      "learning_rate": 5.280048644829866e-07,
+      "loss": 0.2991,
+      "step": 21286
+    },
+    {
+      "epoch": 0.9347398926419092,
+      "grad_norm": 1.453125,
+      "learning_rate": 5.265914510749676e-07,
+      "loss": 0.325,
+      "step": 21288
+    },
+    {
+      "epoch": 0.9348277111211126,
+      "grad_norm": 1.421875,
+      "learning_rate": 5.251799118464407e-07,
+      "loss": 0.311,
+      "step": 21290
+    },
+    {
+      "epoch": 0.9349155296003161,
+      "grad_norm": 1.5,
+      "learning_rate": 5.237702469055028e-07,
+      "loss": 0.3254,
+      "step": 21292
+    },
+    {
+      "epoch": 0.9350033480795197,
+      "grad_norm": 1.4296875,
+      "learning_rate": 5.223624563601065e-07,
+      "loss": 0.3102,
+      "step": 21294
+    },
+    {
+      "epoch": 0.9350911665587232,
+      "grad_norm": 1.5234375,
+      "learning_rate": 5.209565403180627e-07,
+      "loss": 0.3131,
+      "step": 21296
+    },
+    {
+      "epoch": 0.9351789850379266,
+      "grad_norm": 1.46875,
+      "learning_rate": 5.195524988870326e-07,
+      "loss": 0.3322,
+      "step": 21298
+    },
+    {
+      "epoch": 0.9352668035171301,
+      "grad_norm": 1.625,
+      "learning_rate": 5.18150332174544e-07,
+      "loss": 0.3213,
+      "step": 21300
+    },
+    {
+      "epoch": 0.9353546219963336,
+      "grad_norm": 1.4765625,
+      "learning_rate": 5.167500402879665e-07,
+      "loss": 0.2772,
+      "step": 21302
+    },
+    {
+      "epoch": 0.935442440475537,
+      "grad_norm": 1.421875,
+      "learning_rate": 5.153516233345451e-07,
+      "loss": 0.3016,
+      "step": 21304
+    },
+    {
+      "epoch": 0.9355302589547405,
+      "grad_norm": 1.4453125,
+      "learning_rate": 5.139550814213634e-07,
+      "loss": 0.2926,
+      "step": 21306
+    },
+    {
+      "epoch": 0.935618077433944,
+      "grad_norm": 1.5546875,
+      "learning_rate": 5.125604146553692e-07,
+      "loss": 0.3425,
+      "step": 21308
+    },
+    {
+      "epoch": 0.9357058959131476,
+      "grad_norm": 1.4921875,
+      "learning_rate": 5.111676231433715e-07,
+      "loss": 0.3165,
+      "step": 21310
+    },
+    {
+      "epoch": 0.935793714392351,
+      "grad_norm": 1.4140625,
+      "learning_rate": 5.097767069920267e-07,
+      "loss": 0.313,
+      "step": 21312
+    },
+    {
+      "epoch": 0.9358815328715545,
+      "grad_norm": 1.4375,
+      "learning_rate": 5.083876663078523e-07,
+      "loss": 0.3176,
+      "step": 21314
+    },
+    {
+      "epoch": 0.935969351350758,
+      "grad_norm": 1.4140625,
+      "learning_rate": 5.070005011972218e-07,
+      "loss": 0.3126,
+      "step": 21316
+    },
+    {
+      "epoch": 0.9360571698299615,
+      "grad_norm": 1.46875,
+      "learning_rate": 5.056152117663665e-07,
+      "loss": 0.3096,
+      "step": 21318
+    },
+    {
+      "epoch": 0.9361449883091649,
+      "grad_norm": 1.4609375,
+      "learning_rate": 5.042317981213684e-07,
+      "loss": 0.3051,
+      "step": 21320
+    },
+    {
+      "epoch": 0.9362328067883684,
+      "grad_norm": 1.46875,
+      "learning_rate": 5.028502603681678e-07,
+      "loss": 0.3365,
+      "step": 21322
+    },
+    {
+      "epoch": 0.936320625267572,
+      "grad_norm": 1.546875,
+      "learning_rate": 5.01470598612569e-07,
+      "loss": 0.3218,
+      "step": 21324
+    },
+    {
+      "epoch": 0.9364084437467755,
+      "grad_norm": 1.578125,
+      "learning_rate": 5.000928129602234e-07,
+      "loss": 0.2954,
+      "step": 21326
+    },
+    {
+      "epoch": 0.9364962622259789,
+      "grad_norm": 1.6875,
+      "learning_rate": 4.987169035166467e-07,
+      "loss": 0.3334,
+      "step": 21328
+    },
+    {
+      "epoch": 0.9365840807051824,
+      "grad_norm": 1.5,
+      "learning_rate": 4.973428703872018e-07,
+      "loss": 0.3243,
+      "step": 21330
+    },
+    {
+      "epoch": 0.9366718991843859,
+      "grad_norm": 1.453125,
+      "learning_rate": 4.959707136771103e-07,
+      "loss": 0.297,
+      "step": 21332
+    },
+    {
+      "epoch": 0.9367597176635893,
+      "grad_norm": 1.5546875,
+      "learning_rate": 4.9460043349146e-07,
+      "loss": 0.3222,
+      "step": 21334
+    },
+    {
+      "epoch": 0.9368475361427928,
+      "grad_norm": 1.5546875,
+      "learning_rate": 4.932320299351784e-07,
+      "loss": 0.3216,
+      "step": 21336
+    },
+    {
+      "epoch": 0.9369353546219963,
+      "grad_norm": 1.4921875,
+      "learning_rate": 4.918655031130648e-07,
+      "loss": 0.3081,
+      "step": 21338
+    },
+    {
+      "epoch": 0.9370231731011999,
+      "grad_norm": 1.4296875,
+      "learning_rate": 4.905008531297661e-07,
+      "loss": 0.3371,
+      "step": 21340
+    },
+    {
+      "epoch": 0.9371109915804033,
+      "grad_norm": 1.53125,
+      "learning_rate": 4.891380800897877e-07,
+      "loss": 0.3587,
+      "step": 21342
+    },
+    {
+      "epoch": 0.9371988100596068,
+      "grad_norm": 1.3984375,
+      "learning_rate": 4.877771840974904e-07,
+      "loss": 0.2993,
+      "step": 21344
+    },
+    {
+      "epoch": 0.9372866285388103,
+      "grad_norm": 1.546875,
+      "learning_rate": 4.86418165257091e-07,
+      "loss": 0.3142,
+      "step": 21346
+    },
+    {
+      "epoch": 0.9373744470180138,
+      "grad_norm": 1.4140625,
+      "learning_rate": 4.850610236726672e-07,
+      "loss": 0.3118,
+      "step": 21348
+    },
+    {
+      "epoch": 0.9374622654972172,
+      "grad_norm": 1.484375,
+      "learning_rate": 4.83705759448147e-07,
+      "loss": 0.2905,
+      "step": 21350
+    },
+    {
+      "epoch": 0.9375500839764207,
+      "grad_norm": 1.5390625,
+      "learning_rate": 4.823523726873169e-07,
+      "loss": 0.3095,
+      "step": 21352
+    },
+    {
+      "epoch": 0.9376379024556242,
+      "grad_norm": 1.46875,
+      "learning_rate": 4.810008634938163e-07,
+      "loss": 0.3168,
+      "step": 21354
+    },
+    {
+      "epoch": 0.9377257209348278,
+      "grad_norm": 1.4453125,
+      "learning_rate": 4.796512319711482e-07,
+      "loss": 0.3462,
+      "step": 21356
+    },
+    {
+      "epoch": 0.9378135394140312,
+      "grad_norm": 1.515625,
+      "learning_rate": 4.783034782226691e-07,
+      "loss": 0.3285,
+      "step": 21358
+    },
+    {
+      "epoch": 0.9379013578932347,
+      "grad_norm": 1.421875,
+      "learning_rate": 4.769576023515854e-07,
+      "loss": 0.2981,
+      "step": 21360
+    },
+    {
+      "epoch": 0.9379891763724382,
+      "grad_norm": 1.421875,
+      "learning_rate": 4.7561360446096714e-07,
+      "loss": 0.3216,
+      "step": 21362
+    },
+    {
+      "epoch": 0.9380769948516416,
+      "grad_norm": 1.421875,
+      "learning_rate": 4.742714846537377e-07,
+      "loss": 0.298,
+      "step": 21364
+    },
+    {
+      "epoch": 0.9381648133308451,
+      "grad_norm": 1.515625,
+      "learning_rate": 4.729312430326788e-07,
+      "loss": 0.3255,
+      "step": 21366
+    },
+    {
+      "epoch": 0.9382526318100486,
+      "grad_norm": 1.5546875,
+      "learning_rate": 4.7159287970042485e-07,
+      "loss": 0.3263,
+      "step": 21368
+    },
+    {
+      "epoch": 0.9383404502892522,
+      "grad_norm": 1.46875,
+      "learning_rate": 4.702563947594663e-07,
+      "loss": 0.3215,
+      "step": 21370
+    },
+    {
+      "epoch": 0.9384282687684556,
+      "grad_norm": 1.4375,
+      "learning_rate": 4.6892178831215437e-07,
+      "loss": 0.3175,
+      "step": 21372
+    },
+    {
+      "epoch": 0.9385160872476591,
+      "grad_norm": 1.40625,
+      "learning_rate": 4.6758906046069084e-07,
+      "loss": 0.3307,
+      "step": 21374
+    },
+    {
+      "epoch": 0.9386039057268626,
+      "grad_norm": 1.4609375,
+      "learning_rate": 4.662582113071412e-07,
+      "loss": 0.313,
+      "step": 21376
+    },
+    {
+      "epoch": 0.9386917242060661,
+      "grad_norm": 1.46875,
+      "learning_rate": 4.649292409534184e-07,
+      "loss": 0.2962,
+      "step": 21378
+    },
+    {
+      "epoch": 0.9387795426852695,
+      "grad_norm": 1.5,
+      "learning_rate": 4.6360214950129375e-07,
+      "loss": 0.3318,
+      "step": 21380
+    },
+    {
+      "epoch": 0.938867361164473,
+      "grad_norm": 1.4765625,
+      "learning_rate": 4.6227693705239993e-07,
+      "loss": 0.2868,
+      "step": 21382
+    },
+    {
+      "epoch": 0.9389551796436765,
+      "grad_norm": 1.515625,
+      "learning_rate": 4.6095360370821685e-07,
+      "loss": 0.3004,
+      "step": 21384
+    },
+    {
+      "epoch": 0.9390429981228801,
+      "grad_norm": 1.53125,
+      "learning_rate": 4.5963214957009394e-07,
+      "loss": 0.3312,
+      "step": 21386
+    },
+    {
+      "epoch": 0.9391308166020835,
+      "grad_norm": 1.3984375,
+      "learning_rate": 4.5831257473921973e-07,
+      "loss": 0.3139,
+      "step": 21388
+    },
+    {
+      "epoch": 0.939218635081287,
+      "grad_norm": 1.4765625,
+      "learning_rate": 4.569948793166551e-07,
+      "loss": 0.2907,
+      "step": 21390
+    },
+    {
+      "epoch": 0.9393064535604905,
+      "grad_norm": 1.4765625,
+      "learning_rate": 4.5567906340330537e-07,
+      "loss": 0.3269,
+      "step": 21392
+    },
+    {
+      "epoch": 0.939394272039694,
+      "grad_norm": 1.515625,
+      "learning_rate": 4.543651270999344e-07,
+      "loss": 0.3064,
+      "step": 21394
+    },
+    {
+      "epoch": 0.9394820905188974,
+      "grad_norm": 1.4765625,
+      "learning_rate": 4.530530705071673e-07,
+      "loss": 0.3008,
+      "step": 21396
+    },
+    {
+      "epoch": 0.9395699089981009,
+      "grad_norm": 1.375,
+      "learning_rate": 4.5174289372547917e-07,
+      "loss": 0.3103,
+      "step": 21398
+    },
+    {
+      "epoch": 0.9396577274773044,
+      "grad_norm": 1.421875,
+      "learning_rate": 4.504345968552065e-07,
+      "loss": 0.2845,
+      "step": 21400
+    },
+    {
+      "epoch": 0.9397455459565079,
+      "grad_norm": 1.3828125,
+      "learning_rate": 4.4912817999653576e-07,
+      "loss": 0.3474,
+      "step": 21402
+    },
+    {
+      "epoch": 0.9398333644357114,
+      "grad_norm": 1.484375,
+      "learning_rate": 4.4782364324951476e-07,
+      "loss": 0.2995,
+      "step": 21404
+    },
+    {
+      "epoch": 0.9399211829149149,
+      "grad_norm": 1.5234375,
+      "learning_rate": 4.4652098671404685e-07,
+      "loss": 0.2957,
+      "step": 21406
+    },
+    {
+      "epoch": 0.9400090013941184,
+      "grad_norm": 1.46875,
+      "learning_rate": 4.452202104898828e-07,
+      "loss": 0.305,
+      "step": 21408
+    },
+    {
+      "epoch": 0.9400968198733218,
+      "grad_norm": 1.453125,
+      "learning_rate": 4.4392131467664853e-07,
+      "loss": 0.3167,
+      "step": 21410
+    },
+    {
+      "epoch": 0.9401846383525253,
+      "grad_norm": 1.5234375,
+      "learning_rate": 4.4262429937380057e-07,
+      "loss": 0.3332,
+      "step": 21412
+    },
+    {
+      "epoch": 0.9402724568317288,
+      "grad_norm": 1.46875,
+      "learning_rate": 4.4132916468067333e-07,
+      "loss": 0.2977,
+      "step": 21414
+    },
+    {
+      "epoch": 0.9403602753109322,
+      "grad_norm": 1.4609375,
+      "learning_rate": 4.400359106964458e-07,
+      "loss": 0.3122,
+      "step": 21416
+    },
+    {
+      "epoch": 0.9404480937901358,
+      "grad_norm": 1.484375,
+      "learning_rate": 4.387445375201527e-07,
+      "loss": 0.3073,
+      "step": 21418
+    },
+    {
+      "epoch": 0.9405359122693393,
+      "grad_norm": 1.4765625,
+      "learning_rate": 4.374550452506926e-07,
+      "loss": 0.3325,
+      "step": 21420
+    },
+    {
+      "epoch": 0.9406237307485428,
+      "grad_norm": 1.5,
+      "learning_rate": 4.3616743398681157e-07,
+      "loss": 0.283,
+      "step": 21422
+    },
+    {
+      "epoch": 0.9407115492277462,
+      "grad_norm": 1.484375,
+      "learning_rate": 4.348817038271197e-07,
+      "loss": 0.3195,
+      "step": 21424
+    },
+    {
+      "epoch": 0.9407993677069497,
+      "grad_norm": 1.453125,
+      "learning_rate": 4.335978548700742e-07,
+      "loss": 0.3162,
+      "step": 21426
+    },
+    {
+      "epoch": 0.9408871861861532,
+      "grad_norm": 1.4375,
+      "learning_rate": 4.323158872139937e-07,
+      "loss": 0.2864,
+      "step": 21428
+    },
+    {
+      "epoch": 0.9409750046653567,
+      "grad_norm": 1.453125,
+      "learning_rate": 4.3103580095705256e-07,
+      "loss": 0.3208,
+      "step": 21430
+    },
+    {
+      "epoch": 0.9410628231445602,
+      "grad_norm": 1.421875,
+      "learning_rate": 4.2975759619727775e-07,
+      "loss": 0.3209,
+      "step": 21432
+    },
+    {
+      "epoch": 0.9411506416237637,
+      "grad_norm": 1.40625,
+      "learning_rate": 4.284812730325577e-07,
+      "loss": 0.3148,
+      "step": 21434
+    },
+    {
+      "epoch": 0.9412384601029672,
+      "grad_norm": 1.5234375,
+      "learning_rate": 4.272068315606309e-07,
+      "loss": 0.3439,
+      "step": 21436
+    },
+    {
+      "epoch": 0.9413262785821707,
+      "grad_norm": 1.5,
+      "learning_rate": 4.2593427187909705e-07,
+      "loss": 0.2771,
+      "step": 21438
+    },
+    {
+      "epoch": 0.9414140970613741,
+      "grad_norm": 1.46875,
+      "learning_rate": 4.246635940854088e-07,
+      "loss": 0.3518,
+      "step": 21440
+    },
+    {
+      "epoch": 0.9415019155405776,
+      "grad_norm": 1.484375,
+      "learning_rate": 4.233947982768716e-07,
+      "loss": 0.3015,
+      "step": 21442
+    },
+    {
+      "epoch": 0.9415897340197811,
+      "grad_norm": 1.46875,
+      "learning_rate": 4.2212788455065213e-07,
+      "loss": 0.3217,
+      "step": 21444
+    },
+    {
+      "epoch": 0.9416775524989845,
+      "grad_norm": 1.4609375,
+      "learning_rate": 4.2086285300377004e-07,
+      "loss": 0.2966,
+      "step": 21446
+    },
+    {
+      "epoch": 0.9417653709781881,
+      "grad_norm": 1.4375,
+      "learning_rate": 4.195997037331034e-07,
+      "loss": 0.3075,
+      "step": 21448
+    },
+    {
+      "epoch": 0.9418531894573916,
+      "grad_norm": 1.3984375,
+      "learning_rate": 4.183384368353832e-07,
+      "loss": 0.3088,
+      "step": 21450
+    },
+    {
+      "epoch": 0.9419410079365951,
+      "grad_norm": 1.46875,
+      "learning_rate": 4.170790524071988e-07,
+      "loss": 0.3209,
+      "step": 21452
+    },
+    {
+      "epoch": 0.9420288264157985,
+      "grad_norm": 1.4140625,
+      "learning_rate": 4.158215505449953e-07,
+      "loss": 0.2918,
+      "step": 21454
+    },
+    {
+      "epoch": 0.942116644895002,
+      "grad_norm": 1.484375,
+      "learning_rate": 4.145659313450678e-07,
+      "loss": 0.3079,
+      "step": 21456
+    },
+    {
+      "epoch": 0.9422044633742055,
+      "grad_norm": 1.4921875,
+      "learning_rate": 4.1331219490357563e-07,
+      "loss": 0.298,
+      "step": 21458
+    },
+    {
+      "epoch": 0.942292281853409,
+      "grad_norm": 1.5234375,
+      "learning_rate": 4.12060341316528e-07,
+      "loss": 0.31,
+      "step": 21460
+    },
+    {
+      "epoch": 0.9423801003326124,
+      "grad_norm": 1.515625,
+      "learning_rate": 4.1081037067979553e-07,
+      "loss": 0.329,
+      "step": 21462
+    },
+    {
+      "epoch": 0.942467918811816,
+      "grad_norm": 1.5,
+      "learning_rate": 4.095622830891016e-07,
+      "loss": 0.3058,
+      "step": 21464
+    },
+    {
+      "epoch": 0.9425557372910195,
+      "grad_norm": 1.4609375,
+      "learning_rate": 4.08316078640017e-07,
+      "loss": 0.3173,
+      "step": 21466
+    },
+    {
+      "epoch": 0.942643555770223,
+      "grad_norm": 1.5859375,
+      "learning_rate": 4.070717574279875e-07,
+      "loss": 0.339,
+      "step": 21468
+    },
+    {
+      "epoch": 0.9427313742494264,
+      "grad_norm": 1.4375,
+      "learning_rate": 4.058293195482954e-07,
+      "loss": 0.3143,
+      "step": 21470
+    },
+    {
+      "epoch": 0.9428191927286299,
+      "grad_norm": 1.5,
+      "learning_rate": 4.045887650960922e-07,
+      "loss": 0.3114,
+      "step": 21472
+    },
+    {
+      "epoch": 0.9429070112078334,
+      "grad_norm": 1.4609375,
+      "learning_rate": 4.0335009416637426e-07,
+      "loss": 0.3175,
+      "step": 21474
+    },
+    {
+      "epoch": 0.9429948296870369,
+      "grad_norm": 1.484375,
+      "learning_rate": 4.021133068540045e-07,
+      "loss": 0.312,
+      "step": 21476
+    },
+    {
+      "epoch": 0.9430826481662404,
+      "grad_norm": 1.515625,
+      "learning_rate": 4.008784032536933e-07,
+      "loss": 0.3322,
+      "step": 21478
+    },
+    {
+      "epoch": 0.9431704666454439,
+      "grad_norm": 1.4609375,
+      "learning_rate": 3.9964538346000945e-07,
+      "loss": 0.2827,
+      "step": 21480
+    },
+    {
+      "epoch": 0.9432582851246474,
+      "grad_norm": 1.4375,
+      "learning_rate": 3.9841424756738023e-07,
+      "loss": 0.3179,
+      "step": 21482
+    },
+    {
+      "epoch": 0.9433461036038508,
+      "grad_norm": 1.6171875,
+      "learning_rate": 3.9718499567008573e-07,
+      "loss": 0.3429,
+      "step": 21484
+    },
+    {
+      "epoch": 0.9434339220830543,
+      "grad_norm": 1.46875,
+      "learning_rate": 3.959576278622618e-07,
+      "loss": 0.3006,
+      "step": 21486
+    },
+    {
+      "epoch": 0.9435217405622578,
+      "grad_norm": 1.5234375,
+      "learning_rate": 3.9473214423789983e-07,
+      "loss": 0.3245,
+      "step": 21488
+    },
+    {
+      "epoch": 0.9436095590414613,
+      "grad_norm": 1.4765625,
+      "learning_rate": 3.9350854489084985e-07,
+      "loss": 0.3254,
+      "step": 21490
+    },
+    {
+      "epoch": 0.9436973775206647,
+      "grad_norm": 1.4609375,
+      "learning_rate": 3.922868299148119e-07,
+      "loss": 0.3228,
+      "step": 21492
+    },
+    {
+      "epoch": 0.9437851959998683,
+      "grad_norm": 1.4921875,
+      "learning_rate": 3.9106699940335004e-07,
+      "loss": 0.3083,
+      "step": 21494
+    },
+    {
+      "epoch": 0.9438730144790718,
+      "grad_norm": 1.4765625,
+      "learning_rate": 3.898490534498755e-07,
+      "loss": 0.2993,
+      "step": 21496
+    },
+    {
+      "epoch": 0.9439608329582753,
+      "grad_norm": 1.453125,
+      "learning_rate": 3.8863299214765834e-07,
+      "loss": 0.3122,
+      "step": 21498
+    },
+    {
+      "epoch": 0.9440486514374787,
+      "grad_norm": 1.4296875,
+      "learning_rate": 3.874188155898295e-07,
+      "loss": 0.3229,
+      "step": 21500
+    },
+    {
+      "epoch": 0.9441364699166822,
+      "grad_norm": 1.5390625,
+      "learning_rate": 3.862065238693674e-07,
+      "loss": 0.3165,
+      "step": 21502
+    },
+    {
+      "epoch": 0.9442242883958857,
+      "grad_norm": 1.4921875,
+      "learning_rate": 3.8499611707910887e-07,
+      "loss": 0.3188,
+      "step": 21504
+    },
+    {
+      "epoch": 0.9443121068750892,
+      "grad_norm": 1.5234375,
+      "learning_rate": 3.8378759531174924e-07,
+      "loss": 0.2803,
+      "step": 21506
+    },
+    {
+      "epoch": 0.9443999253542926,
+      "grad_norm": 1.421875,
+      "learning_rate": 3.825809586598339e-07,
+      "loss": 0.3319,
+      "step": 21508
+    },
+    {
+      "epoch": 0.9444877438334962,
+      "grad_norm": 1.484375,
+      "learning_rate": 3.8137620721577227e-07,
+      "loss": 0.3344,
+      "step": 21510
+    },
+    {
+      "epoch": 0.9445755623126997,
+      "grad_norm": 1.4609375,
+      "learning_rate": 3.8017334107182113e-07,
+      "loss": 0.2868,
+      "step": 21512
+    },
+    {
+      "epoch": 0.9446633807919032,
+      "grad_norm": 1.4140625,
+      "learning_rate": 3.789723603200984e-07,
+      "loss": 0.2926,
+      "step": 21514
+    },
+    {
+      "epoch": 0.9447511992711066,
+      "grad_norm": 1.40625,
+      "learning_rate": 3.7777326505257503e-07,
+      "loss": 0.2918,
+      "step": 21516
+    },
+    {
+      "epoch": 0.9448390177503101,
+      "grad_norm": 1.546875,
+      "learning_rate": 3.765760553610748e-07,
+      "loss": 0.2923,
+      "step": 21518
+    },
+    {
+      "epoch": 0.9449268362295136,
+      "grad_norm": 1.5390625,
+      "learning_rate": 3.7538073133728256e-07,
+      "loss": 0.3119,
+      "step": 21520
+    },
+    {
+      "epoch": 0.945014654708717,
+      "grad_norm": 1.53125,
+      "learning_rate": 3.7418729307273913e-07,
+      "loss": 0.3021,
+      "step": 21522
+    },
+    {
+      "epoch": 0.9451024731879206,
+      "grad_norm": 1.4921875,
+      "learning_rate": 3.7299574065883527e-07,
+      "loss": 0.3275,
+      "step": 21524
+    },
+    {
+      "epoch": 0.9451902916671241,
+      "grad_norm": 1.4921875,
+      "learning_rate": 3.718060741868229e-07,
+      "loss": 0.3329,
+      "step": 21526
+    },
+    {
+      "epoch": 0.9452781101463276,
+      "grad_norm": 1.5390625,
+      "learning_rate": 3.706182937478014e-07,
+      "loss": 0.3302,
+      "step": 21528
+    },
+    {
+      "epoch": 0.945365928625531,
+      "grad_norm": 1.4765625,
+      "learning_rate": 3.69432399432737e-07,
+      "loss": 0.3176,
+      "step": 21530
+    },
+    {
+      "epoch": 0.9454537471047345,
+      "grad_norm": 1.484375,
+      "learning_rate": 3.6824839133244303e-07,
+      "loss": 0.313,
+      "step": 21532
+    },
+    {
+      "epoch": 0.945541565583938,
+      "grad_norm": 1.484375,
+      "learning_rate": 3.6706626953759427e-07,
+      "loss": 0.3476,
+      "step": 21534
+    },
+    {
+      "epoch": 0.9456293840631415,
+      "grad_norm": 1.4296875,
+      "learning_rate": 3.6588603413871267e-07,
+      "loss": 0.321,
+      "step": 21536
+    },
+    {
+      "epoch": 0.9457172025423449,
+      "grad_norm": 1.5234375,
+      "learning_rate": 3.6470768522618713e-07,
+      "loss": 0.3198,
+      "step": 21538
+    },
+    {
+      "epoch": 0.9458050210215485,
+      "grad_norm": 1.4296875,
+      "learning_rate": 3.6353122289025096e-07,
+      "loss": 0.3369,
+      "step": 21540
+    },
+    {
+      "epoch": 0.945892839500752,
+      "grad_norm": 1.53125,
+      "learning_rate": 3.623566472209988e-07,
+      "loss": 0.304,
+      "step": 21542
+    },
+    {
+      "epoch": 0.9459806579799555,
+      "grad_norm": 1.4453125,
+      "learning_rate": 3.6118395830838095e-07,
+      "loss": 0.3101,
+      "step": 21544
+    },
+    {
+      "epoch": 0.9460684764591589,
+      "grad_norm": 1.46875,
+      "learning_rate": 3.6001315624220046e-07,
+      "loss": 0.3317,
+      "step": 21546
+    },
+    {
+      "epoch": 0.9461562949383624,
+      "grad_norm": 1.5078125,
+      "learning_rate": 3.588442411121218e-07,
+      "loss": 0.2873,
+      "step": 21548
+    },
+    {
+      "epoch": 0.9462441134175659,
+      "grad_norm": 1.4296875,
+      "learning_rate": 3.5767721300765666e-07,
+      "loss": 0.3251,
+      "step": 21550
+    },
+    {
+      "epoch": 0.9463319318967693,
+      "grad_norm": 1.5,
+      "learning_rate": 3.5651207201817527e-07,
+      "loss": 0.3384,
+      "step": 21552
+    },
+    {
+      "epoch": 0.9464197503759728,
+      "grad_norm": 1.4296875,
+      "learning_rate": 3.553488182329118e-07,
+      "loss": 0.3002,
+      "step": 21554
+    },
+    {
+      "epoch": 0.9465075688551764,
+      "grad_norm": 1.4921875,
+      "learning_rate": 3.5418745174093936e-07,
+      "loss": 0.3427,
+      "step": 21556
+    },
+    {
+      "epoch": 0.9465953873343799,
+      "grad_norm": 1.546875,
+      "learning_rate": 3.530279726312008e-07,
+      "loss": 0.3356,
+      "step": 21558
+    },
+    {
+      "epoch": 0.9466832058135833,
+      "grad_norm": 1.4296875,
+      "learning_rate": 3.5187038099248893e-07,
+      "loss": 0.2836,
+      "step": 21560
+    },
+    {
+      "epoch": 0.9467710242927868,
+      "grad_norm": 1.4609375,
+      "learning_rate": 3.5071467691345226e-07,
+      "loss": 0.3062,
+      "step": 21562
+    },
+    {
+      "epoch": 0.9468588427719903,
+      "grad_norm": 1.6328125,
+      "learning_rate": 3.495608604825951e-07,
+      "loss": 0.297,
+      "step": 21564
+    },
+    {
+      "epoch": 0.9469466612511938,
+      "grad_norm": 1.4453125,
+      "learning_rate": 3.4840893178827715e-07,
+      "loss": 0.3281,
+      "step": 21566
+    },
+    {
+      "epoch": 0.9470344797303972,
+      "grad_norm": 1.6015625,
+      "learning_rate": 3.472588909187113e-07,
+      "loss": 0.3245,
+      "step": 21568
+    },
+    {
+      "epoch": 0.9471222982096008,
+      "grad_norm": 1.46875,
+      "learning_rate": 3.461107379619688e-07,
+      "loss": 0.3271,
+      "step": 21570
+    },
+    {
+      "epoch": 0.9472101166888043,
+      "grad_norm": 1.453125,
+      "learning_rate": 3.4496447300597647e-07,
+      "loss": 0.297,
+      "step": 21572
+    },
+    {
+      "epoch": 0.9472979351680078,
+      "grad_norm": 1.5,
+      "learning_rate": 3.438200961385141e-07,
+      "loss": 0.2927,
+      "step": 21574
+    },
+    {
+      "epoch": 0.9473857536472112,
+      "grad_norm": 1.453125,
+      "learning_rate": 3.426776074472199e-07,
+      "loss": 0.3157,
+      "step": 21576
+    },
+    {
+      "epoch": 0.9474735721264147,
+      "grad_norm": 1.5234375,
+      "learning_rate": 3.41537007019585e-07,
+      "loss": 0.3015,
+      "step": 21578
+    },
+    {
+      "epoch": 0.9475613906056182,
+      "grad_norm": 1.5234375,
+      "learning_rate": 3.4039829494295626e-07,
+      "loss": 0.3131,
+      "step": 21580
+    },
+    {
+      "epoch": 0.9476492090848216,
+      "grad_norm": 1.546875,
+      "learning_rate": 3.392614713045389e-07,
+      "loss": 0.3216,
+      "step": 21582
+    },
+    {
+      "epoch": 0.9477370275640251,
+      "grad_norm": 1.4921875,
+      "learning_rate": 3.381265361913882e-07,
+      "loss": 0.3391,
+      "step": 21584
+    },
+    {
+      "epoch": 0.9478248460432287,
+      "grad_norm": 1.5859375,
+      "learning_rate": 3.369934896904209e-07,
+      "loss": 0.2709,
+      "step": 21586
+    },
+    {
+      "epoch": 0.9479126645224322,
+      "grad_norm": 1.5703125,
+      "learning_rate": 3.3586233188840355e-07,
+      "loss": 0.2907,
+      "step": 21588
+    },
+    {
+      "epoch": 0.9480004830016356,
+      "grad_norm": 1.4375,
+      "learning_rate": 3.3473306287196136e-07,
+      "loss": 0.3163,
+      "step": 21590
+    },
+    {
+      "epoch": 0.9480883014808391,
+      "grad_norm": 1.453125,
+      "learning_rate": 3.336056827275752e-07,
+      "loss": 0.2975,
+      "step": 21592
+    },
+    {
+      "epoch": 0.9481761199600426,
+      "grad_norm": 1.6015625,
+      "learning_rate": 3.3248019154157595e-07,
+      "loss": 0.3388,
+      "step": 21594
+    },
+    {
+      "epoch": 0.9482639384392461,
+      "grad_norm": 1.5390625,
+      "learning_rate": 3.3135658940015857e-07,
+      "loss": 0.2906,
+      "step": 21596
+    },
+    {
+      "epoch": 0.9483517569184495,
+      "grad_norm": 1.453125,
+      "learning_rate": 3.302348763893681e-07,
+      "loss": 0.3181,
+      "step": 21598
+    },
+    {
+      "epoch": 0.948439575397653,
+      "grad_norm": 1.4609375,
+      "learning_rate": 3.2911505259510255e-07,
+      "loss": 0.3341,
+      "step": 21600
+    },
+    {
+      "epoch": 0.9485273938768566,
+      "grad_norm": 1.421875,
+      "learning_rate": 3.2799711810312107e-07,
+      "loss": 0.3411,
+      "step": 21602
+    },
+    {
+      "epoch": 0.9486152123560601,
+      "grad_norm": 1.5625,
+      "learning_rate": 3.268810729990329e-07,
+      "loss": 0.3265,
+      "step": 21604
+    },
+    {
+      "epoch": 0.9487030308352635,
+      "grad_norm": 1.453125,
+      "learning_rate": 3.2576691736831144e-07,
+      "loss": 0.3259,
+      "step": 21606
+    },
+    {
+      "epoch": 0.948790849314467,
+      "grad_norm": 1.53125,
+      "learning_rate": 3.246546512962689e-07,
+      "loss": 0.3057,
+      "step": 21608
+    },
+    {
+      "epoch": 0.9488786677936705,
+      "grad_norm": 1.4453125,
+      "learning_rate": 3.2354427486809e-07,
+      "loss": 0.3533,
+      "step": 21610
+    },
+    {
+      "epoch": 0.9489664862728739,
+      "grad_norm": 1.5234375,
+      "learning_rate": 3.224357881688067e-07,
+      "loss": 0.3012,
+      "step": 21612
+    },
+    {
+      "epoch": 0.9490543047520774,
+      "grad_norm": 1.453125,
+      "learning_rate": 3.2132919128330664e-07,
+      "loss": 0.3049,
+      "step": 21614
+    },
+    {
+      "epoch": 0.9491421232312809,
+      "grad_norm": 1.46875,
+      "learning_rate": 3.202244842963331e-07,
+      "loss": 0.2996,
+      "step": 21616
+    },
+    {
+      "epoch": 0.9492299417104845,
+      "grad_norm": 1.5703125,
+      "learning_rate": 3.191216672924824e-07,
+      "loss": 0.2839,
+      "step": 21618
+    },
+    {
+      "epoch": 0.9493177601896879,
+      "grad_norm": 1.4375,
+      "learning_rate": 3.180207403562119e-07,
+      "loss": 0.3177,
+      "step": 21620
+    },
+    {
+      "epoch": 0.9494055786688914,
+      "grad_norm": 1.53125,
+      "learning_rate": 3.1692170357183193e-07,
+      "loss": 0.3231,
+      "step": 21622
+    },
+    {
+      "epoch": 0.9494933971480949,
+      "grad_norm": 1.421875,
+      "learning_rate": 3.15824557023503e-07,
+      "loss": 0.3116,
+      "step": 21624
+    },
+    {
+      "epoch": 0.9495812156272984,
+      "grad_norm": 1.546875,
+      "learning_rate": 3.1472930079524674e-07,
+      "loss": 0.2935,
+      "step": 21626
+    },
+    {
+      "epoch": 0.9496690341065018,
+      "grad_norm": 1.5234375,
+      "learning_rate": 3.136359349709378e-07,
+      "loss": 0.3204,
+      "step": 21628
+    },
+    {
+      "epoch": 0.9497568525857053,
+      "grad_norm": 1.546875,
+      "learning_rate": 3.1254445963430914e-07,
+      "loss": 0.31,
+      "step": 21630
+    },
+    {
+      "epoch": 0.9498446710649089,
+      "grad_norm": 1.5,
+      "learning_rate": 3.114548748689411e-07,
+      "loss": 0.3214,
+      "step": 21632
+    },
+    {
+      "epoch": 0.9499324895441124,
+      "grad_norm": 1.4453125,
+      "learning_rate": 3.1036718075827806e-07,
+      "loss": 0.3303,
+      "step": 21634
+    },
+    {
+      "epoch": 0.9500203080233158,
+      "grad_norm": 1.4453125,
+      "learning_rate": 3.092813773856118e-07,
+      "loss": 0.3243,
+      "step": 21636
+    },
+    {
+      "epoch": 0.9501081265025193,
+      "grad_norm": 1.5078125,
+      "learning_rate": 3.0819746483410075e-07,
+      "loss": 0.317,
+      "step": 21638
+    },
+    {
+      "epoch": 0.9501959449817228,
+      "grad_norm": 1.5078125,
+      "learning_rate": 3.0711544318674514e-07,
+      "loss": 0.3354,
+      "step": 21640
+    },
+    {
+      "epoch": 0.9502837634609262,
+      "grad_norm": 1.46875,
+      "learning_rate": 3.060353125264065e-07,
+      "loss": 0.2811,
+      "step": 21642
+    },
+    {
+      "epoch": 0.9503715819401297,
+      "grad_norm": 1.4375,
+      "learning_rate": 3.049570729358076e-07,
+      "loss": 0.3008,
+      "step": 21644
+    },
+    {
+      "epoch": 0.9504594004193332,
+      "grad_norm": 1.453125,
+      "learning_rate": 3.0388072449751e-07,
+      "loss": 0.2968,
+      "step": 21646
+    },
+    {
+      "epoch": 0.9505472188985368,
+      "grad_norm": 1.515625,
+      "learning_rate": 3.028062672939508e-07,
+      "loss": 0.3307,
+      "step": 21648
+    },
+    {
+      "epoch": 0.9506350373777402,
+      "grad_norm": 1.484375,
+      "learning_rate": 3.017337014074084e-07,
+      "loss": 0.3145,
+      "step": 21650
+    },
+    {
+      "epoch": 0.9507228558569437,
+      "grad_norm": 1.4296875,
+      "learning_rate": 3.0066302692001724e-07,
+      "loss": 0.3367,
+      "step": 21652
+    },
+    {
+      "epoch": 0.9508106743361472,
+      "grad_norm": 1.5078125,
+      "learning_rate": 2.995942439137728e-07,
+      "loss": 0.3307,
+      "step": 21654
+    },
+    {
+      "epoch": 0.9508984928153507,
+      "grad_norm": 1.5,
+      "learning_rate": 2.9852735247052346e-07,
+      "loss": 0.3209,
+      "step": 21656
+    },
+    {
+      "epoch": 0.9509863112945541,
+      "grad_norm": 1.4609375,
+      "learning_rate": 2.9746235267197053e-07,
+      "loss": 0.3235,
+      "step": 21658
+    },
+    {
+      "epoch": 0.9510741297737576,
+      "grad_norm": 1.421875,
+      "learning_rate": 2.9639924459967105e-07,
+      "loss": 0.3129,
+      "step": 21660
+    },
+    {
+      "epoch": 0.9511619482529611,
+      "grad_norm": 1.4921875,
+      "learning_rate": 2.9533802833504043e-07,
+      "loss": 0.3338,
+      "step": 21662
+    },
+    {
+      "epoch": 0.9512497667321647,
+      "grad_norm": 1.46875,
+      "learning_rate": 2.94278703959347e-07,
+      "loss": 0.3112,
+      "step": 21664
+    },
+    {
+      "epoch": 0.9513375852113681,
+      "grad_norm": 1.4375,
+      "learning_rate": 2.932212715537092e-07,
+      "loss": 0.311,
+      "step": 21666
+    },
+    {
+      "epoch": 0.9514254036905716,
+      "grad_norm": 1.3828125,
+      "learning_rate": 2.9216573119911217e-07,
+      "loss": 0.3196,
+      "step": 21668
+    },
+    {
+      "epoch": 0.9515132221697751,
+      "grad_norm": 1.5078125,
+      "learning_rate": 2.9111208297638303e-07,
+      "loss": 0.2995,
+      "step": 21670
+    },
+    {
+      "epoch": 0.9516010406489785,
+      "grad_norm": 1.5390625,
+      "learning_rate": 2.9006032696621833e-07,
+      "loss": 0.3024,
+      "step": 21672
+    },
+    {
+      "epoch": 0.951688859128182,
+      "grad_norm": 1.5078125,
+      "learning_rate": 2.890104632491536e-07,
+      "loss": 0.3242,
+      "step": 21674
+    },
+    {
+      "epoch": 0.9517766776073855,
+      "grad_norm": 1.4140625,
+      "learning_rate": 2.879624919055912e-07,
+      "loss": 0.3083,
+      "step": 21676
+    },
+    {
+      "epoch": 0.9518644960865891,
+      "grad_norm": 1.640625,
+      "learning_rate": 2.869164130157864e-07,
+      "loss": 0.3083,
+      "step": 21678
+    },
+    {
+      "epoch": 0.9519523145657925,
+      "grad_norm": 1.4609375,
+      "learning_rate": 2.858722266598474e-07,
+      "loss": 0.287,
+      "step": 21680
+    },
+    {
+      "epoch": 0.952040133044996,
+      "grad_norm": 1.3984375,
+      "learning_rate": 2.8482993291773506e-07,
+      "loss": 0.3142,
+      "step": 21682
+    },
+    {
+      "epoch": 0.9521279515241995,
+      "grad_norm": 1.5546875,
+      "learning_rate": 2.8378953186927457e-07,
+      "loss": 0.3063,
+      "step": 21684
+    },
+    {
+      "epoch": 0.952215770003403,
+      "grad_norm": 1.4765625,
+      "learning_rate": 2.827510235941355e-07,
+      "loss": 0.3172,
+      "step": 21686
+    },
+    {
+      "epoch": 0.9523035884826064,
+      "grad_norm": 1.484375,
+      "learning_rate": 2.817144081718459e-07,
+      "loss": 0.3375,
+      "step": 21688
+    },
+    {
+      "epoch": 0.9523914069618099,
+      "grad_norm": 1.5078125,
+      "learning_rate": 2.8067968568179505e-07,
+      "loss": 0.283,
+      "step": 21690
+    },
+    {
+      "epoch": 0.9524792254410134,
+      "grad_norm": 1.453125,
+      "learning_rate": 2.7964685620321953e-07,
+      "loss": 0.3213,
+      "step": 21692
+    },
+    {
+      "epoch": 0.952567043920217,
+      "grad_norm": 1.5234375,
+      "learning_rate": 2.786159198152116e-07,
+      "loss": 0.3321,
+      "step": 21694
+    },
+    {
+      "epoch": 0.9526548623994204,
+      "grad_norm": 1.46875,
+      "learning_rate": 2.775868765967221e-07,
+      "loss": 0.3309,
+      "step": 21696
+    },
+    {
+      "epoch": 0.9527426808786239,
+      "grad_norm": 1.4296875,
+      "learning_rate": 2.7655972662655736e-07,
+      "loss": 0.301,
+      "step": 21698
+    },
+    {
+      "epoch": 0.9528304993578274,
+      "grad_norm": 1.4296875,
+      "learning_rate": 2.755344699833767e-07,
+      "loss": 0.3061,
+      "step": 21700
+    },
+    {
+      "epoch": 0.9529183178370308,
+      "grad_norm": 1.5390625,
+      "learning_rate": 2.7451110674569237e-07,
+      "loss": 0.3429,
+      "step": 21702
+    },
+    {
+      "epoch": 0.9530061363162343,
+      "grad_norm": 1.4453125,
+      "learning_rate": 2.7348963699187214e-07,
+      "loss": 0.3205,
+      "step": 21704
+    },
+    {
+      "epoch": 0.9530939547954378,
+      "grad_norm": 1.4921875,
+      "learning_rate": 2.7247006080014513e-07,
+      "loss": 0.324,
+      "step": 21706
+    },
+    {
+      "epoch": 0.9531817732746413,
+      "grad_norm": 1.53125,
+      "learning_rate": 2.714523782485878e-07,
+      "loss": 0.3208,
+      "step": 21708
+    },
+    {
+      "epoch": 0.9532695917538448,
+      "grad_norm": 1.4921875,
+      "learning_rate": 2.7043658941513783e-07,
+      "loss": 0.3202,
+      "step": 21710
+    },
+    {
+      "epoch": 0.9533574102330483,
+      "grad_norm": 1.5390625,
+      "learning_rate": 2.6942269437758015e-07,
+      "loss": 0.2927,
+      "step": 21712
+    },
+    {
+      "epoch": 0.9534452287122518,
+      "grad_norm": 1.4765625,
+      "learning_rate": 2.6841069321355827e-07,
+      "loss": 0.337,
+      "step": 21714
+    },
+    {
+      "epoch": 0.9535330471914553,
+      "grad_norm": 1.46875,
+      "learning_rate": 2.674005860005768e-07,
+      "loss": 0.2998,
+      "step": 21716
+    },
+    {
+      "epoch": 0.9536208656706587,
+      "grad_norm": 1.515625,
+      "learning_rate": 2.6639237281598783e-07,
+      "loss": 0.2922,
+      "step": 21718
+    },
+    {
+      "epoch": 0.9537086841498622,
+      "grad_norm": 1.40625,
+      "learning_rate": 2.6538605373699897e-07,
+      "loss": 0.2955,
+      "step": 21720
+    },
+    {
+      "epoch": 0.9537965026290657,
+      "grad_norm": 1.46875,
+      "learning_rate": 2.6438162884067365e-07,
+      "loss": 0.3054,
+      "step": 21722
+    },
+    {
+      "epoch": 0.9538843211082693,
+      "grad_norm": 1.53125,
+      "learning_rate": 2.6337909820393634e-07,
+      "loss": 0.326,
+      "step": 21724
+    },
+    {
+      "epoch": 0.9539721395874727,
+      "grad_norm": 1.546875,
+      "learning_rate": 2.623784619035535e-07,
+      "loss": 0.2911,
+      "step": 21726
+    },
+    {
+      "epoch": 0.9540599580666762,
+      "grad_norm": 1.46875,
+      "learning_rate": 2.613797200161611e-07,
+      "loss": 0.3035,
+      "step": 21728
+    },
+    {
+      "epoch": 0.9541477765458797,
+      "grad_norm": 1.4921875,
+      "learning_rate": 2.6038287261823944e-07,
+      "loss": 0.3146,
+      "step": 21730
+    },
+    {
+      "epoch": 0.9542355950250831,
+      "grad_norm": 1.4375,
+      "learning_rate": 2.593879197861249e-07,
+      "loss": 0.298,
+      "step": 21732
+    },
+    {
+      "epoch": 0.9543234135042866,
+      "grad_norm": 1.484375,
+      "learning_rate": 2.5839486159601746e-07,
+      "loss": 0.2863,
+      "step": 21734
+    },
+    {
+      "epoch": 0.9544112319834901,
+      "grad_norm": 1.4609375,
+      "learning_rate": 2.5740369812396193e-07,
+      "loss": 0.2981,
+      "step": 21736
+    },
+    {
+      "epoch": 0.9544990504626936,
+      "grad_norm": 1.4453125,
+      "learning_rate": 2.5641442944586144e-07,
+      "loss": 0.2854,
+      "step": 21738
+    },
+    {
+      "epoch": 0.9545868689418971,
+      "grad_norm": 1.515625,
+      "learning_rate": 2.554270556374777e-07,
+      "loss": 0.2901,
+      "step": 21740
+    },
+    {
+      "epoch": 0.9546746874211006,
+      "grad_norm": 1.453125,
+      "learning_rate": 2.544415767744196e-07,
+      "loss": 0.3035,
+      "step": 21742
+    },
+    {
+      "epoch": 0.9547625059003041,
+      "grad_norm": 1.4765625,
+      "learning_rate": 2.5345799293215734e-07,
+      "loss": 0.3136,
+      "step": 21744
+    },
+    {
+      "epoch": 0.9548503243795076,
+      "grad_norm": 1.546875,
+      "learning_rate": 2.5247630418601673e-07,
+      "loss": 0.3432,
+      "step": 21746
+    },
+    {
+      "epoch": 0.954938142858711,
+      "grad_norm": 1.46875,
+      "learning_rate": 2.5149651061117105e-07,
+      "loss": 0.3208,
+      "step": 21748
+    },
+    {
+      "epoch": 0.9550259613379145,
+      "grad_norm": 1.515625,
+      "learning_rate": 2.505186122826547e-07,
+      "loss": 0.3199,
+      "step": 21750
+    },
+    {
+      "epoch": 0.955113779817118,
+      "grad_norm": 1.53125,
+      "learning_rate": 2.495426092753578e-07,
+      "loss": 0.32,
+      "step": 21752
+    },
+    {
+      "epoch": 0.9552015982963215,
+      "grad_norm": 1.4921875,
+      "learning_rate": 2.485685016640177e-07,
+      "loss": 0.3055,
+      "step": 21754
+    },
+    {
+      "epoch": 0.955289416775525,
+      "grad_norm": 1.421875,
+      "learning_rate": 2.4759628952323867e-07,
+      "loss": 0.2988,
+      "step": 21756
+    },
+    {
+      "epoch": 0.9553772352547285,
+      "grad_norm": 1.484375,
+      "learning_rate": 2.466259729274667e-07,
+      "loss": 0.3319,
+      "step": 21758
+    },
+    {
+      "epoch": 0.955465053733932,
+      "grad_norm": 1.4375,
+      "learning_rate": 2.456575519510118e-07,
+      "loss": 0.3322,
+      "step": 21760
+    },
+    {
+      "epoch": 0.9555528722131355,
+      "grad_norm": 1.421875,
+      "learning_rate": 2.446910266680369e-07,
+      "loss": 0.3233,
+      "step": 21762
+    },
+    {
+      "epoch": 0.9556406906923389,
+      "grad_norm": 1.421875,
+      "learning_rate": 2.4372639715255776e-07,
+      "loss": 0.3157,
+      "step": 21764
+    },
+    {
+      "epoch": 0.9557285091715424,
+      "grad_norm": 1.5625,
+      "learning_rate": 2.4276366347844305e-07,
+      "loss": 0.2984,
+      "step": 21766
+    },
+    {
+      "epoch": 0.9558163276507459,
+      "grad_norm": 1.484375,
+      "learning_rate": 2.4180282571942546e-07,
+      "loss": 0.3368,
+      "step": 21768
+    },
+    {
+      "epoch": 0.9559041461299493,
+      "grad_norm": 1.4296875,
+      "learning_rate": 2.4084388394907954e-07,
+      "loss": 0.3091,
+      "step": 21770
+    },
+    {
+      "epoch": 0.9559919646091529,
+      "grad_norm": 1.6796875,
+      "learning_rate": 2.398868382408437e-07,
+      "loss": 0.3215,
+      "step": 21772
+    },
+    {
+      "epoch": 0.9560797830883564,
+      "grad_norm": 1.484375,
+      "learning_rate": 2.389316886680121e-07,
+      "loss": 0.315,
+      "step": 21774
+    },
+    {
+      "epoch": 0.9561676015675599,
+      "grad_norm": 1.4609375,
+      "learning_rate": 2.3797843530372344e-07,
+      "loss": 0.3483,
+      "step": 21776
+    },
+    {
+      "epoch": 0.9562554200467633,
+      "grad_norm": 1.5,
+      "learning_rate": 2.370270782209888e-07,
+      "loss": 0.3088,
+      "step": 21778
+    },
+    {
+      "epoch": 0.9563432385259668,
+      "grad_norm": 1.4765625,
+      "learning_rate": 2.360776174926499e-07,
+      "loss": 0.3143,
+      "step": 21780
+    },
+    {
+      "epoch": 0.9564310570051703,
+      "grad_norm": 1.4375,
+      "learning_rate": 2.3513005319142634e-07,
+      "loss": 0.3293,
+      "step": 21782
+    },
+    {
+      "epoch": 0.9565188754843738,
+      "grad_norm": 1.4765625,
+      "learning_rate": 2.3418438538987952e-07,
+      "loss": 0.3208,
+      "step": 21784
+    },
+    {
+      "epoch": 0.9566066939635773,
+      "grad_norm": 1.4609375,
+      "learning_rate": 2.332406141604293e-07,
+      "loss": 0.3239,
+      "step": 21786
+    },
+    {
+      "epoch": 0.9566945124427808,
+      "grad_norm": 1.546875,
+      "learning_rate": 2.3229873957534841e-07,
+      "loss": 0.3204,
+      "step": 21788
+    },
+    {
+      "epoch": 0.9567823309219843,
+      "grad_norm": 1.46875,
+      "learning_rate": 2.3135876170676806e-07,
+      "loss": 0.3359,
+      "step": 21790
+    },
+    {
+      "epoch": 0.9568701494011878,
+      "grad_norm": 1.5390625,
+      "learning_rate": 2.3042068062667232e-07,
+      "loss": 0.3261,
+      "step": 21792
+    },
+    {
+      "epoch": 0.9569579678803912,
+      "grad_norm": 1.53125,
+      "learning_rate": 2.294844964068954e-07,
+      "loss": 0.3228,
+      "step": 21794
+    },
+    {
+      "epoch": 0.9570457863595947,
+      "grad_norm": 1.484375,
+      "learning_rate": 2.2855020911913826e-07,
+      "loss": 0.3152,
+      "step": 21796
+    },
+    {
+      "epoch": 0.9571336048387982,
+      "grad_norm": 1.4921875,
+      "learning_rate": 2.2761781883494094e-07,
+      "loss": 0.316,
+      "step": 21798
+    },
+    {
+      "epoch": 0.9572214233180016,
+      "grad_norm": 1.46875,
+      "learning_rate": 2.2668732562571016e-07,
+      "loss": 0.283,
+      "step": 21800
+    },
+    {
+      "epoch": 0.9573092417972052,
+      "grad_norm": 1.484375,
+      "learning_rate": 2.2575872956270283e-07,
+      "loss": 0.3198,
+      "step": 21802
+    },
+    {
+      "epoch": 0.9573970602764087,
+      "grad_norm": 1.453125,
+      "learning_rate": 2.248320307170315e-07,
+      "loss": 0.3325,
+      "step": 21804
+    },
+    {
+      "epoch": 0.9574848787556122,
+      "grad_norm": 1.4296875,
+      "learning_rate": 2.2390722915966167e-07,
+      "loss": 0.3085,
+      "step": 21806
+    },
+    {
+      "epoch": 0.9575726972348156,
+      "grad_norm": 1.53125,
+      "learning_rate": 2.2298432496141441e-07,
+      "loss": 0.346,
+      "step": 21808
+    },
+    {
+      "epoch": 0.9576605157140191,
+      "grad_norm": 1.5,
+      "learning_rate": 2.2206331819296934e-07,
+      "loss": 0.3222,
+      "step": 21810
+    },
+    {
+      "epoch": 0.9577483341932226,
+      "grad_norm": 1.46875,
+      "learning_rate": 2.2114420892485333e-07,
+      "loss": 0.3223,
+      "step": 21812
+    },
+    {
+      "epoch": 0.957836152672426,
+      "grad_norm": 1.515625,
+      "learning_rate": 2.2022699722745454e-07,
+      "loss": 0.3287,
+      "step": 21814
+    },
+    {
+      "epoch": 0.9579239711516295,
+      "grad_norm": 1.515625,
+      "learning_rate": 2.1931168317101125e-07,
+      "loss": 0.3536,
+      "step": 21816
+    },
+    {
+      "epoch": 0.9580117896308331,
+      "grad_norm": 1.4140625,
+      "learning_rate": 2.1839826682562015e-07,
+      "loss": 0.3204,
+      "step": 21818
+    },
+    {
+      "epoch": 0.9580996081100366,
+      "grad_norm": 1.5,
+      "learning_rate": 2.1748674826123084e-07,
+      "loss": 0.3263,
+      "step": 21820
+    },
+    {
+      "epoch": 0.95818742658924,
+      "grad_norm": 1.546875,
+      "learning_rate": 2.165771275476458e-07,
+      "loss": 0.3258,
+      "step": 21822
+    },
+    {
+      "epoch": 0.9582752450684435,
+      "grad_norm": 1.5078125,
+      "learning_rate": 2.1566940475452602e-07,
+      "loss": 0.3304,
+      "step": 21824
+    },
+    {
+      "epoch": 0.958363063547647,
+      "grad_norm": 1.3984375,
+      "learning_rate": 2.147635799513853e-07,
+      "loss": 0.304,
+      "step": 21826
+    },
+    {
+      "epoch": 0.9584508820268505,
+      "grad_norm": 1.453125,
+      "learning_rate": 2.1385965320759038e-07,
+      "loss": 0.3063,
+      "step": 21828
+    },
+    {
+      "epoch": 0.9585387005060539,
+      "grad_norm": 1.5,
+      "learning_rate": 2.1295762459236368e-07,
+      "loss": 0.3192,
+      "step": 21830
+    },
+    {
+      "epoch": 0.9586265189852575,
+      "grad_norm": 1.4453125,
+      "learning_rate": 2.1205749417478604e-07,
+      "loss": 0.3364,
+      "step": 21832
+    },
+    {
+      "epoch": 0.958714337464461,
+      "grad_norm": 1.453125,
+      "learning_rate": 2.1115926202378565e-07,
+      "loss": 0.3031,
+      "step": 21834
+    },
+    {
+      "epoch": 0.9588021559436645,
+      "grad_norm": 1.546875,
+      "learning_rate": 2.1026292820815195e-07,
+      "loss": 0.3354,
+      "step": 21836
+    },
+    {
+      "epoch": 0.9588899744228679,
+      "grad_norm": 1.453125,
+      "learning_rate": 2.0936849279652727e-07,
+      "loss": 0.2824,
+      "step": 21838
+    },
+    {
+      "epoch": 0.9589777929020714,
+      "grad_norm": 1.4375,
+      "learning_rate": 2.0847595585740676e-07,
+      "loss": 0.3315,
+      "step": 21840
+    },
+    {
+      "epoch": 0.9590656113812749,
+      "grad_norm": 1.5078125,
+      "learning_rate": 2.075853174591358e-07,
+      "loss": 0.288,
+      "step": 21842
+    },
+    {
+      "epoch": 0.9591534298604784,
+      "grad_norm": 1.4765625,
+      "learning_rate": 2.0669657766992923e-07,
+      "loss": 0.317,
+      "step": 21844
+    },
+    {
+      "epoch": 0.9592412483396818,
+      "grad_norm": 1.3984375,
+      "learning_rate": 2.058097365578382e-07,
+      "loss": 0.2892,
+      "step": 21846
+    },
+    {
+      "epoch": 0.9593290668188854,
+      "grad_norm": 1.4765625,
+      "learning_rate": 2.0492479419078336e-07,
+      "loss": 0.3132,
+      "step": 21848
+    },
+    {
+      "epoch": 0.9594168852980889,
+      "grad_norm": 1.5390625,
+      "learning_rate": 2.0404175063653275e-07,
+      "loss": 0.3096,
+      "step": 21850
+    },
+    {
+      "epoch": 0.9595047037772924,
+      "grad_norm": 1.4453125,
+      "learning_rate": 2.0316060596270726e-07,
+      "loss": 0.3124,
+      "step": 21852
+    },
+    {
+      "epoch": 0.9595925222564958,
+      "grad_norm": 1.5859375,
+      "learning_rate": 2.0228136023678623e-07,
+      "loss": 0.3051,
+      "step": 21854
+    },
+    {
+      "epoch": 0.9596803407356993,
+      "grad_norm": 1.4765625,
+      "learning_rate": 2.0140401352610195e-07,
+      "loss": 0.2949,
+      "step": 21856
+    },
+    {
+      "epoch": 0.9597681592149028,
+      "grad_norm": 1.46875,
+      "learning_rate": 2.0052856589784507e-07,
+      "loss": 0.3101,
+      "step": 21858
+    },
+    {
+      "epoch": 0.9598559776941062,
+      "grad_norm": 1.4453125,
+      "learning_rate": 1.9965501741905645e-07,
+      "loss": 0.302,
+      "step": 21860
+    },
+    {
+      "epoch": 0.9599437961733097,
+      "grad_norm": 1.5078125,
+      "learning_rate": 1.9878336815662978e-07,
+      "loss": 0.2926,
+      "step": 21862
+    },
+    {
+      "epoch": 0.9600316146525133,
+      "grad_norm": 1.5078125,
+      "learning_rate": 1.9791361817732002e-07,
+      "loss": 0.3298,
+      "step": 21864
+    },
+    {
+      "epoch": 0.9601194331317168,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.9704576754772663e-07,
+      "loss": 0.3247,
+      "step": 21866
+    },
+    {
+      "epoch": 0.9602072516109202,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.961798163343187e-07,
+      "loss": 0.2966,
+      "step": 21868
+    },
+    {
+      "epoch": 0.9602950700901237,
+      "grad_norm": 1.515625,
+      "learning_rate": 1.953157646034043e-07,
+      "loss": 0.3175,
+      "step": 21870
+    },
+    {
+      "epoch": 0.9603828885693272,
+      "grad_norm": 1.453125,
+      "learning_rate": 1.9445361242115545e-07,
+      "loss": 0.3292,
+      "step": 21872
+    },
+    {
+      "epoch": 0.9604707070485307,
+      "grad_norm": 1.5078125,
+      "learning_rate": 1.935933598535944e-07,
+      "loss": 0.3042,
+      "step": 21874
+    },
+    {
+      "epoch": 0.9605585255277341,
+      "grad_norm": 1.4296875,
+      "learning_rate": 1.9273500696659896e-07,
+      "loss": 0.3043,
+      "step": 21876
+    },
+    {
+      "epoch": 0.9606463440069377,
+      "grad_norm": 1.4765625,
+      "learning_rate": 1.9187855382590547e-07,
+      "loss": 0.3103,
+      "step": 21878
+    },
+    {
+      "epoch": 0.9607341624861412,
+      "grad_norm": 1.4375,
+      "learning_rate": 1.9102400049710035e-07,
+      "loss": 0.3412,
+      "step": 21880
+    },
+    {
+      "epoch": 0.9608219809653447,
+      "grad_norm": 1.4765625,
+      "learning_rate": 1.901713470456229e-07,
+      "loss": 0.3092,
+      "step": 21882
+    },
+    {
+      "epoch": 0.9609097994445481,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.893205935367709e-07,
+      "loss": 0.3328,
+      "step": 21884
+    },
+    {
+      "epoch": 0.9609976179237516,
+      "grad_norm": 1.4140625,
+      "learning_rate": 1.88471740035695e-07,
+      "loss": 0.3059,
+      "step": 21886
+    },
+    {
+      "epoch": 0.9610854364029551,
+      "grad_norm": 1.5078125,
+      "learning_rate": 1.8762478660740156e-07,
+      "loss": 0.3126,
+      "step": 21888
+    },
+    {
+      "epoch": 0.9611732548821585,
+      "grad_norm": 1.4140625,
+      "learning_rate": 1.8677973331674982e-07,
+      "loss": 0.3218,
+      "step": 21890
+    },
+    {
+      "epoch": 0.961261073361362,
+      "grad_norm": 1.3984375,
+      "learning_rate": 1.8593658022845462e-07,
+      "loss": 0.2991,
+      "step": 21892
+    },
+    {
+      "epoch": 0.9613488918405656,
+      "grad_norm": 1.453125,
+      "learning_rate": 1.8509532740708102e-07,
+      "loss": 0.3121,
+      "step": 21894
+    },
+    {
+      "epoch": 0.9614367103197691,
+      "grad_norm": 1.5078125,
+      "learning_rate": 1.842559749170608e-07,
+      "loss": 0.2948,
+      "step": 21896
+    },
+    {
+      "epoch": 0.9615245287989725,
+      "grad_norm": 1.5,
+      "learning_rate": 1.834185228226648e-07,
+      "loss": 0.3117,
+      "step": 21898
+    },
+    {
+      "epoch": 0.961612347278176,
+      "grad_norm": 1.5,
+      "learning_rate": 1.8258297118802502e-07,
+      "loss": 0.3318,
+      "step": 21900
+    },
+    {
+      "epoch": 0.9617001657573795,
+      "grad_norm": 1.453125,
+      "learning_rate": 1.8174932007713476e-07,
+      "loss": 0.3209,
+      "step": 21902
+    },
+    {
+      "epoch": 0.961787984236583,
+      "grad_norm": 1.515625,
+      "learning_rate": 1.80917569553829e-07,
+      "loss": 0.3037,
+      "step": 21904
+    },
+    {
+      "epoch": 0.9618758027157864,
+      "grad_norm": 1.4296875,
+      "learning_rate": 1.8008771968180403e-07,
+      "loss": 0.2937,
+      "step": 21906
+    },
+    {
+      "epoch": 0.9619636211949899,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.7925977052461186e-07,
+      "loss": 0.3217,
+      "step": 21908
+    },
+    {
+      "epoch": 0.9620514396741935,
+      "grad_norm": 1.4921875,
+      "learning_rate": 1.7843372214565723e-07,
+      "loss": 0.3387,
+      "step": 21910
+    },
+    {
+      "epoch": 0.962139258153397,
+      "grad_norm": 1.46875,
+      "learning_rate": 1.7760957460819793e-07,
+      "loss": 0.3066,
+      "step": 21912
+    },
+    {
+      "epoch": 0.9622270766326004,
+      "grad_norm": 1.5078125,
+      "learning_rate": 1.7678732797534735e-07,
+      "loss": 0.3127,
+      "step": 21914
+    },
+    {
+      "epoch": 0.9623148951118039,
+      "grad_norm": 1.515625,
+      "learning_rate": 1.7596698231007459e-07,
+      "loss": 0.3274,
+      "step": 21916
+    },
+    {
+      "epoch": 0.9624027135910074,
+      "grad_norm": 1.4765625,
+      "learning_rate": 1.7514853767519878e-07,
+      "loss": 0.3352,
+      "step": 21918
+    },
+    {
+      "epoch": 0.9624905320702108,
+      "grad_norm": 1.5,
+      "learning_rate": 1.7433199413340317e-07,
+      "loss": 0.2812,
+      "step": 21920
+    },
+    {
+      "epoch": 0.9625783505494143,
+      "grad_norm": 1.4140625,
+      "learning_rate": 1.7351735174721274e-07,
+      "loss": 0.286,
+      "step": 21922
+    },
+    {
+      "epoch": 0.9626661690286179,
+      "grad_norm": 1.4375,
+      "learning_rate": 1.7270461057901367e-07,
+      "loss": 0.3296,
+      "step": 21924
+    },
+    {
+      "epoch": 0.9627539875078214,
+      "grad_norm": 1.4140625,
+      "learning_rate": 1.7189377069104784e-07,
+      "loss": 0.306,
+      "step": 21926
+    },
+    {
+      "epoch": 0.9628418059870248,
+      "grad_norm": 1.4765625,
+      "learning_rate": 1.7108483214540726e-07,
+      "loss": 0.3067,
+      "step": 21928
+    },
+    {
+      "epoch": 0.9629296244662283,
+      "grad_norm": 1.46875,
+      "learning_rate": 1.702777950040424e-07,
+      "loss": 0.3172,
+      "step": 21930
+    },
+    {
+      "epoch": 0.9630174429454318,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.6947265932875655e-07,
+      "loss": 0.3347,
+      "step": 21932
+    },
+    {
+      "epoch": 0.9631052614246353,
+      "grad_norm": 1.4375,
+      "learning_rate": 1.6866942518120877e-07,
+      "loss": 0.2929,
+      "step": 21934
+    },
+    {
+      "epoch": 0.9631930799038387,
+      "grad_norm": 1.3984375,
+      "learning_rate": 1.6786809262290816e-07,
+      "loss": 0.3185,
+      "step": 21936
+    },
+    {
+      "epoch": 0.9632808983830422,
+      "grad_norm": 1.4609375,
+      "learning_rate": 1.6706866171521952e-07,
+      "loss": 0.3116,
+      "step": 21938
+    },
+    {
+      "epoch": 0.9633687168622458,
+      "grad_norm": 1.484375,
+      "learning_rate": 1.662711325193689e-07,
+      "loss": 0.307,
+      "step": 21940
+    },
+    {
+      "epoch": 0.9634565353414493,
+      "grad_norm": 1.515625,
+      "learning_rate": 1.6547550509642406e-07,
+      "loss": 0.3371,
+      "step": 21942
+    },
+    {
+      "epoch": 0.9635443538206527,
+      "grad_norm": 1.421875,
+      "learning_rate": 1.6468177950731967e-07,
+      "loss": 0.3152,
+      "step": 21944
+    },
+    {
+      "epoch": 0.9636321722998562,
+      "grad_norm": 1.421875,
+      "learning_rate": 1.638899558128404e-07,
+      "loss": 0.317,
+      "step": 21946
+    },
+    {
+      "epoch": 0.9637199907790597,
+      "grad_norm": 1.4140625,
+      "learning_rate": 1.631000340736183e-07,
+      "loss": 0.3107,
+      "step": 21948
+    },
+    {
+      "epoch": 0.9638078092582631,
+      "grad_norm": 1.484375,
+      "learning_rate": 1.623120143501522e-07,
+      "loss": 0.3131,
+      "step": 21950
+    },
+    {
+      "epoch": 0.9638956277374666,
+      "grad_norm": 1.46875,
+      "learning_rate": 1.6152589670278552e-07,
+      "loss": 0.3149,
+      "step": 21952
+    },
+    {
+      "epoch": 0.9639834462166701,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.6074168119172006e-07,
+      "loss": 0.3243,
+      "step": 21954
+    },
+    {
+      "epoch": 0.9640712646958737,
+      "grad_norm": 1.3984375,
+      "learning_rate": 1.5995936787700782e-07,
+      "loss": 0.3016,
+      "step": 21956
+    },
+    {
+      "epoch": 0.9641590831750771,
+      "grad_norm": 1.4140625,
+      "learning_rate": 1.5917895681856475e-07,
+      "loss": 0.3022,
+      "step": 21958
+    },
+    {
+      "epoch": 0.9642469016542806,
+      "grad_norm": 1.4921875,
+      "learning_rate": 1.5840044807615138e-07,
+      "loss": 0.3238,
+      "step": 21960
+    },
+    {
+      "epoch": 0.9643347201334841,
+      "grad_norm": 1.453125,
+      "learning_rate": 1.5762384170938947e-07,
+      "loss": 0.3044,
+      "step": 21962
+    },
+    {
+      "epoch": 0.9644225386126876,
+      "grad_norm": 1.4765625,
+      "learning_rate": 1.5684913777774536e-07,
+      "loss": 0.3587,
+      "step": 21964
+    },
+    {
+      "epoch": 0.964510357091891,
+      "grad_norm": 1.46875,
+      "learning_rate": 1.560763363405493e-07,
+      "loss": 0.306,
+      "step": 21966
+    },
+    {
+      "epoch": 0.9645981755710945,
+      "grad_norm": 1.421875,
+      "learning_rate": 1.5530543745698457e-07,
+      "loss": 0.3222,
+      "step": 21968
+    },
+    {
+      "epoch": 0.964685994050298,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.5453644118608447e-07,
+      "loss": 0.3017,
+      "step": 21970
+    },
+    {
+      "epoch": 0.9647738125295016,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.5376934758674077e-07,
+      "loss": 0.3307,
+      "step": 21972
+    },
+    {
+      "epoch": 0.964861631008705,
+      "grad_norm": 1.453125,
+      "learning_rate": 1.530041567176954e-07,
+      "loss": 0.3025,
+      "step": 21974
+    },
+    {
+      "epoch": 0.9649494494879085,
+      "grad_norm": 1.4609375,
+      "learning_rate": 1.5224086863754594e-07,
+      "loss": 0.3043,
+      "step": 21976
+    },
+    {
+      "epoch": 0.965037267967112,
+      "grad_norm": 1.421875,
+      "learning_rate": 1.5147948340475115e-07,
+      "loss": 0.3172,
+      "step": 21978
+    },
+    {
+      "epoch": 0.9651250864463154,
+      "grad_norm": 1.484375,
+      "learning_rate": 1.5072000107761164e-07,
+      "loss": 0.3221,
+      "step": 21980
+    },
+    {
+      "epoch": 0.9652129049255189,
+      "grad_norm": 1.515625,
+      "learning_rate": 1.4996242171429197e-07,
+      "loss": 0.3245,
+      "step": 21982
+    },
+    {
+      "epoch": 0.9653007234047224,
+      "grad_norm": 1.3984375,
+      "learning_rate": 1.4920674537280688e-07,
+      "loss": 0.3153,
+      "step": 21984
+    },
+    {
+      "epoch": 0.965388541883926,
+      "grad_norm": 1.4609375,
+      "learning_rate": 1.4845297211102393e-07,
+      "loss": 0.3186,
+      "step": 21986
+    },
+    {
+      "epoch": 0.9654763603631294,
+      "grad_norm": 1.328125,
+      "learning_rate": 1.4770110198667197e-07,
+      "loss": 0.3236,
+      "step": 21988
+    },
+    {
+      "epoch": 0.9655641788423329,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.4695113505732715e-07,
+      "loss": 0.3182,
+      "step": 21990
+    },
+    {
+      "epoch": 0.9656519973215364,
+      "grad_norm": 1.5,
+      "learning_rate": 1.4620307138042412e-07,
+      "loss": 0.303,
+      "step": 21992
+    },
+    {
+      "epoch": 0.9657398158007399,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.4545691101324476e-07,
+      "loss": 0.337,
+      "step": 21994
+    },
+    {
+      "epoch": 0.9658276342799433,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.447126540129351e-07,
+      "loss": 0.3146,
+      "step": 21996
+    },
+    {
+      "epoch": 0.9659154527591468,
+      "grad_norm": 1.4453125,
+      "learning_rate": 1.439703004364884e-07,
+      "loss": 0.3166,
+      "step": 21998
+    },
+    {
+      "epoch": 0.9660032712383503,
+      "grad_norm": 1.421875,
+      "learning_rate": 1.4322985034075366e-07,
+      "loss": 0.2911,
+      "step": 22000
+    }
+  ],
+  "logging_steps": 2,
+  "max_steps": 22774,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 1000,
+  "total_flos": 1.498954685106157e+20,
+  "train_batch_size": 16,
+  "trial_name": null,
+  "trial_params": null
+}