diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,21462 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.0,
+  "eval_steps": 500,
+  "global_step": 3060,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "grad_norm": 15.222698623854631,
+      "learning_rate": 5.434782608695653e-07,
+      "loss": 1.3618,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 15.441764980759185,
+      "learning_rate": 1.0869565217391306e-06,
+      "loss": 1.6375,
+      "step": 2
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 52.36910019147159,
+      "learning_rate": 1.6304347826086957e-06,
+      "loss": 1.5114,
+      "step": 3
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 79.14883281224802,
+      "learning_rate": 2.173913043478261e-06,
+      "loss": 1.4802,
+      "step": 4
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 24.183275549739147,
+      "learning_rate": 2.7173913043478263e-06,
+      "loss": 1.633,
+      "step": 5
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.501537134978104,
+      "learning_rate": 3.2608695652173914e-06,
+      "loss": 1.3411,
+      "step": 6
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.795534698444224,
+      "learning_rate": 3.804347826086957e-06,
+      "loss": 1.3528,
+      "step": 7
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.0583519491438245,
+      "learning_rate": 4.347826086956522e-06,
+      "loss": 1.2969,
+      "step": 8
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.788290915563179,
+      "learning_rate": 4.891304347826087e-06,
+      "loss": 1.2831,
+      "step": 9
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.2075536163674627,
+      "learning_rate": 5.4347826086956525e-06,
+      "loss": 1.4016,
+      "step": 10
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 292.69404423078487,
+      "learning_rate": 5.978260869565218e-06,
+      "loss": 1.2624,
+      "step": 11
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.4924855041222993,
+      "learning_rate": 6.521739130434783e-06,
+      "loss": 1.2177,
+      "step": 12
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.9260952512869878,
+      "learning_rate": 7.065217391304347e-06,
+      "loss": 1.3548,
+      "step": 13
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.6252583655256587,
+      "learning_rate": 7.608695652173914e-06,
+      "loss": 1.1306,
+      "step": 14
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.4581566669462607,
+      "learning_rate": 8.15217391304348e-06,
+      "loss": 1.1593,
+      "step": 15
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.075170120782009,
+      "learning_rate": 8.695652173913044e-06,
+      "loss": 1.2969,
+      "step": 16
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.8073357482535246,
+      "learning_rate": 9.239130434782608e-06,
+      "loss": 1.0924,
+      "step": 17
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6493661976101195,
+      "learning_rate": 9.782608695652175e-06,
+      "loss": 1.0853,
+      "step": 18
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.593199599682604,
+      "learning_rate": 1.032608695652174e-05,
+      "loss": 1.051,
+      "step": 19
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3487932209095463,
+      "learning_rate": 1.0869565217391305e-05,
+      "loss": 1.2454,
+      "step": 20
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.461986960782079,
+      "learning_rate": 1.141304347826087e-05,
+      "loss": 1.0397,
+      "step": 21
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.449986905945311,
+      "learning_rate": 1.1956521739130435e-05,
+      "loss": 1.0583,
+      "step": 22
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.2290206143746267,
+      "learning_rate": 1.25e-05,
+      "loss": 1.2072,
+      "step": 23
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.1801703758534974,
+      "learning_rate": 1.3043478260869566e-05,
+      "loss": 1.0254,
+      "step": 24
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 13.150239286733365,
+      "learning_rate": 1.3586956521739131e-05,
+      "loss": 1.0013,
+      "step": 25
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9258807620747714,
+      "learning_rate": 1.4130434782608694e-05,
+      "loss": 0.991,
+      "step": 26
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.410781838262468,
+      "learning_rate": 1.4673913043478263e-05,
+      "loss": 1.208,
+      "step": 27
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.183741569116391,
+      "learning_rate": 1.5217391304347828e-05,
+      "loss": 0.9846,
+      "step": 28
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3240750984972838,
+      "learning_rate": 1.5760869565217393e-05,
+      "loss": 0.95,
+      "step": 29
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9221095027527438,
+      "learning_rate": 1.630434782608696e-05,
+      "loss": 0.9441,
+      "step": 30
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.4135259985039434,
+      "learning_rate": 1.6847826086956524e-05,
+      "loss": 1.1474,
+      "step": 31
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7424946413917688,
+      "learning_rate": 1.739130434782609e-05,
+      "loss": 0.9314,
+      "step": 32
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7290844736214271,
+      "learning_rate": 1.793478260869565e-05,
+      "loss": 0.9352,
+      "step": 33
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6744894943069046,
+      "learning_rate": 1.8478260869565216e-05,
+      "loss": 1.1446,
+      "step": 34
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6319386135304128,
+      "learning_rate": 1.9021739130434784e-05,
+      "loss": 0.9333,
+      "step": 35
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6450219018775063,
+      "learning_rate": 1.956521739130435e-05,
+      "loss": 0.9154,
+      "step": 36
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7523492996985188,
+      "learning_rate": 2.0108695652173915e-05,
+      "loss": 1.1138,
+      "step": 37
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5917251920016904,
+      "learning_rate": 2.065217391304348e-05,
+      "loss": 0.9082,
+      "step": 38
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.0752507950227674,
+      "learning_rate": 2.1195652173913045e-05,
+      "loss": 0.94,
+      "step": 39
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.47915742627259,
+      "learning_rate": 2.173913043478261e-05,
+      "loss": 0.8736,
+      "step": 40
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.4570020964130346,
+      "learning_rate": 2.2282608695652175e-05,
+      "loss": 0.8659,
+      "step": 41
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5748572712987965,
+      "learning_rate": 2.282608695652174e-05,
+      "loss": 1.1299,
+      "step": 42
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.4696994787517195,
+      "learning_rate": 2.3369565217391306e-05,
+      "loss": 0.8834,
+      "step": 43
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5277484092558196,
+      "learning_rate": 2.391304347826087e-05,
+      "loss": 0.8929,
+      "step": 44
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.55082383618551,
+      "learning_rate": 2.4456521739130436e-05,
+      "loss": 1.0904,
+      "step": 45
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.4840026113699543,
+      "learning_rate": 2.5e-05,
+      "loss": 0.854,
+      "step": 46
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.39115686214600043,
+      "learning_rate": 2.554347826086957e-05,
+      "loss": 0.84,
+      "step": 47
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.44221721745566117,
+      "learning_rate": 2.608695652173913e-05,
+      "loss": 1.0949,
+      "step": 48
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.40880305123941396,
+      "learning_rate": 2.66304347826087e-05,
+      "loss": 0.8591,
+      "step": 49
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.4009917924961596,
+      "learning_rate": 2.7173913043478262e-05,
+      "loss": 0.8908,
+      "step": 50
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.36794737384781123,
+      "learning_rate": 2.7717391304347827e-05,
+      "loss": 0.8302,
+      "step": 51
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.36879903751240933,
+      "learning_rate": 2.826086956521739e-05,
+      "loss": 0.8436,
+      "step": 52
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.38452605707078225,
+      "learning_rate": 2.8804347826086957e-05,
+      "loss": 1.0787,
+      "step": 53
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.40856109847585587,
+      "learning_rate": 2.9347826086956526e-05,
+      "loss": 0.8594,
+      "step": 54
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.3743439125442982,
+      "learning_rate": 2.9891304347826088e-05,
+      "loss": 0.8614,
+      "step": 55
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.39606007585239084,
+      "learning_rate": 3.0434782608695656e-05,
+      "loss": 1.0686,
+      "step": 56
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.3549431649470009,
+      "learning_rate": 3.0978260869565215e-05,
+      "loss": 0.834,
+      "step": 57
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.33596662451382026,
+      "learning_rate": 3.152173913043479e-05,
+      "loss": 0.8311,
+      "step": 58
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.3720965839273748,
+      "learning_rate": 3.2065217391304345e-05,
+      "loss": 1.0408,
+      "step": 59
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.3530884191531684,
+      "learning_rate": 3.260869565217392e-05,
+      "loss": 0.8113,
+      "step": 60
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.37252406291024454,
+      "learning_rate": 3.3152173913043475e-05,
+      "loss": 0.8743,
+      "step": 61
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.34827116980591866,
+      "learning_rate": 3.369565217391305e-05,
+      "loss": 0.8245,
+      "step": 62
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.40041470793341244,
+      "learning_rate": 3.423913043478261e-05,
+      "loss": 1.066,
+      "step": 63
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.36608898247106253,
+      "learning_rate": 3.478260869565218e-05,
+      "loss": 0.8195,
+      "step": 64
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.3642854690890474,
+      "learning_rate": 3.532608695652174e-05,
+      "loss": 0.8066,
+      "step": 65
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.4996455325604121,
+      "learning_rate": 3.58695652173913e-05,
+      "loss": 1.0554,
+      "step": 66
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.37335479755717277,
+      "learning_rate": 3.641304347826087e-05,
+      "loss": 0.8028,
+      "step": 67
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.37562369757216285,
+      "learning_rate": 3.695652173913043e-05,
+      "loss": 0.7926,
+      "step": 68
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.36563187961943366,
+      "learning_rate": 3.7500000000000003e-05,
+      "loss": 0.8083,
+      "step": 69
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.3974357942487217,
+      "learning_rate": 3.804347826086957e-05,
+      "loss": 1.0303,
+      "step": 70
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.37072319735647064,
+      "learning_rate": 3.8586956521739134e-05,
+      "loss": 0.8009,
+      "step": 71
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.4147074701690486,
+      "learning_rate": 3.91304347826087e-05,
+      "loss": 0.8234,
+      "step": 72
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.39829632861187725,
+      "learning_rate": 3.9673913043478264e-05,
+      "loss": 0.8009,
+      "step": 73
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.42681594778769705,
+      "learning_rate": 4.021739130434783e-05,
+      "loss": 1.0047,
+      "step": 74
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.4699272036789851,
+      "learning_rate": 4.076086956521739e-05,
+      "loss": 0.7888,
+      "step": 75
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5019551320116233,
+      "learning_rate": 4.130434782608696e-05,
+      "loss": 0.7924,
+      "step": 76
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.48661523390511685,
+      "learning_rate": 4.1847826086956525e-05,
+      "loss": 1.0729,
+      "step": 77
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.4640674557689061,
+      "learning_rate": 4.239130434782609e-05,
+      "loss": 0.7691,
+      "step": 78
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.4640340491796376,
+      "learning_rate": 4.2934782608695655e-05,
+      "loss": 0.7881,
+      "step": 79
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5099749485019944,
+      "learning_rate": 4.347826086956522e-05,
+      "loss": 1.0024,
+      "step": 80
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.4601209096365815,
+      "learning_rate": 4.4021739130434786e-05,
+      "loss": 0.7623,
+      "step": 81
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5679665903354001,
+      "learning_rate": 4.456521739130435e-05,
+      "loss": 0.7996,
+      "step": 82
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5110745811690282,
+      "learning_rate": 4.5108695652173916e-05,
+      "loss": 0.8116,
+      "step": 83
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.44603534899570146,
+      "learning_rate": 4.565217391304348e-05,
+      "loss": 0.7637,
+      "step": 84
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.48187254639707855,
+      "learning_rate": 4.6195652173913046e-05,
+      "loss": 1.0057,
+      "step": 85
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.49632442454765957,
+      "learning_rate": 4.673913043478261e-05,
+      "loss": 0.7761,
+      "step": 86
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.51283948645882,
+      "learning_rate": 4.7282608695652177e-05,
+      "loss": 0.7723,
+      "step": 87
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5040374723519099,
+      "learning_rate": 4.782608695652174e-05,
+      "loss": 1.0125,
+      "step": 88
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6424505816581453,
+      "learning_rate": 4.836956521739131e-05,
+      "loss": 0.7821,
+      "step": 89
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8952380849131504,
+      "learning_rate": 4.891304347826087e-05,
+      "loss": 0.786,
+      "step": 90
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9071716847965982,
+      "learning_rate": 4.945652173913044e-05,
+      "loss": 1.0205,
+      "step": 91
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6394496217131463,
+      "learning_rate": 5e-05,
+      "loss": 0.7865,
+      "step": 92
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6430017414643429,
+      "learning_rate": 4.999998600447087e-05,
+      "loss": 0.7767,
+      "step": 93
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7163800789469034,
+      "learning_rate": 4.999994401789913e-05,
+      "loss": 0.8076,
+      "step": 94
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.649129937462473,
+      "learning_rate": 4.9999874040331803e-05,
+      "loss": 0.7852,
+      "step": 95
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5725005822442988,
+      "learning_rate": 4.9999776071847235e-05,
+      "loss": 1.0133,
+      "step": 96
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6437637771603318,
+      "learning_rate": 4.999965011255512e-05,
+      "loss": 0.7835,
+      "step": 97
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6756152038997708,
+      "learning_rate": 4.9999496162596475e-05,
+      "loss": 0.78,
+      "step": 98
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6457462371257101,
+      "learning_rate": 4.999931422214368e-05,
+      "loss": 0.8314,
+      "step": 99
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6356704939082438,
+      "learning_rate": 4.999910429140045e-05,
+      "loss": 0.9231,
+      "step": 100
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6788057103679841,
+      "learning_rate": 4.9998866370601804e-05,
+      "loss": 0.7779,
+      "step": 101
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5520979287668127,
+      "learning_rate": 4.9998600460014154e-05,
+      "loss": 1.0115,
+      "step": 102
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6256903741791827,
+      "learning_rate": 4.9998306559935224e-05,
+      "loss": 0.7621,
+      "step": 103
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7294199624973984,
+      "learning_rate": 4.9997984670694065e-05,
+      "loss": 0.7604,
+      "step": 104
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.4576508070967194,
+      "learning_rate": 4.9997634792651084e-05,
+      "loss": 0.8004,
+      "step": 105
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5454888114392468,
+      "learning_rate": 4.9997256926198025e-05,
+      "loss": 1.001,
+      "step": 106
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.4802088536214176,
+      "learning_rate": 4.999685107175794e-05,
+      "loss": 0.7888,
+      "step": 107
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5526430942139623,
+      "learning_rate": 4.9996417229785274e-05,
+      "loss": 0.7261,
+      "step": 108
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.38792234854392654,
+      "learning_rate": 4.999595540076575e-05,
+      "loss": 0.9922,
+      "step": 109
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6055331379079447,
+      "learning_rate": 4.999546558521646e-05,
+      "loss": 0.5851,
+      "step": 110
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5158780093685665,
+      "learning_rate": 4.9994947783685826e-05,
+      "loss": 1.043,
+      "step": 111
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5557440556659233,
+      "learning_rate": 4.999440199675359e-05,
+      "loss": 0.7595,
+      "step": 112
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6541675572511069,
+      "learning_rate": 4.9993828225030856e-05,
+      "loss": 0.7426,
+      "step": 113
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.48516518176340506,
+      "learning_rate": 4.999322646916003e-05,
+      "loss": 0.9139,
+      "step": 114
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6227813906892624,
+      "learning_rate": 4.999259672981486e-05,
+      "loss": 0.7568,
+      "step": 115
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.4822101602003976,
+      "learning_rate": 4.999193900770043e-05,
+      "loss": 0.8239,
+      "step": 116
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5496537105402288,
+      "learning_rate": 4.999125330355316e-05,
+      "loss": 1.0059,
+      "step": 117
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.3978993826972184,
+      "learning_rate": 4.9990539618140794e-05,
+      "loss": 0.7701,
+      "step": 118
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5486285780784311,
+      "learning_rate": 4.99897979522624e-05,
+      "loss": 0.747,
+      "step": 119
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.4105849981925598,
+      "learning_rate": 4.9989028306748376e-05,
+      "loss": 0.987,
+      "step": 120
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5407481109428779,
+      "learning_rate": 4.9988230682460455e-05,
+      "loss": 0.5671,
+      "step": 121
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.46192105495930946,
+      "learning_rate": 4.998740508029169e-05,
+      "loss": 1.0014,
+      "step": 122
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.515768812632998,
+      "learning_rate": 4.9986551501166463e-05,
+      "loss": 0.752,
+      "step": 123
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.4489119192762821,
+      "learning_rate": 4.998566994604047e-05,
+      "loss": 0.7593,
+      "step": 124
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.4948643606593063,
+      "learning_rate": 4.9984760415900746e-05,
+      "loss": 0.9888,
+      "step": 125
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.4144264963514692,
+      "learning_rate": 4.998382291176563e-05,
+      "loss": 0.7317,
+      "step": 126
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.644150545172914,
+      "learning_rate": 4.9982857434684794e-05,
+      "loss": 0.7506,
+      "step": 127
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5743062921983124,
+      "learning_rate": 4.9981863985739236e-05,
+      "loss": 0.9872,
+      "step": 128
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.4702554412193669,
+      "learning_rate": 4.9980842566041255e-05,
+      "loss": 0.741,
+      "step": 129
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7178943449630688,
+      "learning_rate": 4.997979317673447e-05,
+      "loss": 0.7542,
+      "step": 130
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6146409881065324,
+      "learning_rate": 4.997871581899384e-05,
+      "loss": 1.0151,
+      "step": 131
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5786533119465405,
+      "learning_rate": 4.99776104940256e-05,
+      "loss": 0.797,
+      "step": 132
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8724720245588223,
+      "learning_rate": 4.997647720306733e-05,
+      "loss": 0.774,
+      "step": 133
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.525014784938383,
+      "learning_rate": 4.997531594738791e-05,
+      "loss": 0.7247,
+      "step": 134
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.47961931659849577,
+      "learning_rate": 4.997412672828753e-05,
+      "loss": 0.9808,
+      "step": 135
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5520771032628174,
+      "learning_rate": 4.997290954709769e-05,
+      "loss": 0.73,
+      "step": 136
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5452349936125844,
+      "learning_rate": 4.9971664405181195e-05,
+      "loss": 0.7384,
+      "step": 137
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.5483944106882118,
+      "learning_rate": 4.997039130393216e-05,
+      "loss": 0.7648,
+      "step": 138
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.47411282320932635,
+      "learning_rate": 4.996909024477601e-05,
+      "loss": 0.972,
+      "step": 139
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4432103501065713,
+      "learning_rate": 4.996776122916945e-05,
+      "loss": 0.7259,
+      "step": 140
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.5169005172216407,
+      "learning_rate": 4.996640425860052e-05,
+      "loss": 0.7665,
+      "step": 141
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.38832723066296326,
+      "learning_rate": 4.996501933458852e-05,
+      "loss": 0.964,
+      "step": 142
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.5057742363345261,
+      "learning_rate": 4.9963606458684096e-05,
+      "loss": 0.8326,
+      "step": 143
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4645815993603433,
+      "learning_rate": 4.9962165632469146e-05,
+      "loss": 0.7146,
+      "step": 144
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4264564064658369,
+      "learning_rate": 4.996069685755689e-05,
+      "loss": 0.7335,
+      "step": 145
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4652161623420289,
+      "learning_rate": 4.995920013559182e-05,
+      "loss": 0.9866,
+      "step": 146
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4276246440384927,
+      "learning_rate": 4.9957675468249736e-05,
+      "loss": 0.7441,
+      "step": 147
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.48883644289079486,
+      "learning_rate": 4.995612285723772e-05,
+      "loss": 0.7315,
+      "step": 148
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.5741526280573931,
+      "learning_rate": 4.995454230429414e-05,
+      "loss": 1.0002,
+      "step": 149
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.48833509869946146,
+      "learning_rate": 4.9952933811188645e-05,
+      "loss": 0.7647,
+      "step": 150
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.48237145756013533,
+      "learning_rate": 4.995129737972218e-05,
+      "loss": 0.7362,
+      "step": 151
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.38961943725695874,
+      "learning_rate": 4.994963301172696e-05,
+      "loss": 0.993,
+      "step": 152
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.49984464434990367,
+      "learning_rate": 4.994794070906648e-05,
+      "loss": 0.7243,
+      "step": 153
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.5081943829830544,
+      "learning_rate": 4.9946220473635516e-05,
+      "loss": 0.7944,
+      "step": 154
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7278779921107277,
+      "learning_rate": 4.9944472307360114e-05,
+      "loss": 0.7259,
+      "step": 155
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.708817365357396,
+      "learning_rate": 4.99426962121976e-05,
+      "loss": 0.7373,
+      "step": 156
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.5001170508609586,
+      "learning_rate": 4.9940892190136555e-05,
+      "loss": 0.9945,
+      "step": 157
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.5224475375305859,
+      "learning_rate": 4.993906024319685e-05,
+      "loss": 0.7324,
+      "step": 158
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7291291279241043,
+      "learning_rate": 4.9937200373429606e-05,
+      "loss": 0.7198,
+      "step": 159
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.5235328982523992,
+      "learning_rate": 4.993531258291721e-05,
+      "loss": 1.0035,
+      "step": 160
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.561889500283033,
+      "learning_rate": 4.993339687377331e-05,
+      "loss": 0.739,
+      "step": 161
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4946852683190996,
+      "learning_rate": 4.993145324814282e-05,
+      "loss": 0.7288,
+      "step": 162
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4712789709491518,
+      "learning_rate": 4.9929481708201914e-05,
+      "loss": 0.9815,
+      "step": 163
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.46136505491866114,
+      "learning_rate": 4.9927482256158e-05,
+      "loss": 0.716,
+      "step": 164
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.5198368728871099,
+      "learning_rate": 4.992545489424974e-05,
+      "loss": 0.7605,
+      "step": 165
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6005999581282977,
+      "learning_rate": 4.992339962474707e-05,
+      "loss": 0.7062,
+      "step": 166
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.48131486398597517,
+      "learning_rate": 4.9921316449951155e-05,
+      "loss": 0.7287,
+      "step": 167
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.37638386829187126,
+      "learning_rate": 4.99192053721944e-05,
+      "loss": 0.9983,
+      "step": 168
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5086724257469257,
+      "learning_rate": 4.991706639384046e-05,
+      "loss": 0.7235,
+      "step": 169
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5012591833990734,
+      "learning_rate": 4.991489951728423e-05,
+      "loss": 0.75,
+      "step": 170
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.4123801940912291,
+      "learning_rate": 4.991270474495182e-05,
+      "loss": 1.0152,
+      "step": 171
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.4436187959955827,
+      "learning_rate": 4.9910482079300614e-05,
+      "loss": 0.7274,
+      "step": 172
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.49163105586351025,
+      "learning_rate": 4.990823152281918e-05,
+      "loss": 0.7114,
+      "step": 173
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.3687771471038135,
+      "learning_rate": 4.990595307802736e-05,
+      "loss": 1.0287,
+      "step": 174
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.4235884728473828,
+      "learning_rate": 4.990364674747617e-05,
+      "loss": 0.7188,
+      "step": 175
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.4768674468642702,
+      "learning_rate": 4.99013125337479e-05,
+      "loss": 0.7617,
+      "step": 176
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.38591093400253124,
+      "learning_rate": 4.989895043945603e-05,
+      "loss": 0.7138,
+      "step": 177
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.40554878364799335,
+      "learning_rate": 4.989656046724524e-05,
+      "loss": 0.9793,
+      "step": 178
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.4491881260557013,
+      "learning_rate": 4.989414261979146e-05,
+      "loss": 0.7413,
+      "step": 179
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.394028146757425,
+      "learning_rate": 4.9891696899801824e-05,
+      "loss": 0.7099,
+      "step": 180
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.47213498681139704,
+      "learning_rate": 4.988922331001464e-05,
+      "loss": 0.7243,
+      "step": 181
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.38367475094736436,
+      "learning_rate": 4.988672185319947e-05,
+      "loss": 0.9915,
+      "step": 182
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.38478024145859013,
+      "learning_rate": 4.9884192532157025e-05,
+      "loss": 0.7367,
+      "step": 183
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.41608182380561226,
+      "learning_rate": 4.988163534971926e-05,
+      "loss": 0.7479,
+      "step": 184
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.3723464578340394,
+      "learning_rate": 4.98790503087493e-05,
+      "loss": 0.9858,
+      "step": 185
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.4109881393086165,
+      "learning_rate": 4.9876437412141455e-05,
+      "loss": 0.7185,
+      "step": 186
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.46678252566194167,
+      "learning_rate": 4.987379666282125e-05,
+      "loss": 0.7729,
+      "step": 187
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.4194302994883171,
+      "learning_rate": 4.987112806374536e-05,
+      "loss": 0.713,
+      "step": 188
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.43663901336787003,
+      "learning_rate": 4.986843161790169e-05,
+      "loss": 0.9747,
+      "step": 189
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.37437894981166436,
+      "learning_rate": 4.986570732830928e-05,
+      "loss": 0.7383,
+      "step": 190
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.430038039238556,
+      "learning_rate": 4.986295519801834e-05,
+      "loss": 0.7137,
+      "step": 191
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.39203866985681707,
+      "learning_rate": 4.986017523011031e-05,
+      "loss": 0.9734,
+      "step": 192
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.4148604572475499,
+      "learning_rate": 4.985736742769773e-05,
+      "loss": 0.7101,
+      "step": 193
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.47611011058079505,
+      "learning_rate": 4.9854531793924354e-05,
+      "loss": 0.7073,
+      "step": 194
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.42276417682247536,
+      "learning_rate": 4.985166833196506e-05,
+      "loss": 0.9686,
+      "step": 195
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.40779712881977187,
+      "learning_rate": 4.9848777045025906e-05,
+      "loss": 0.7186,
+      "step": 196
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6096990905995647,
+      "learning_rate": 4.9845857936344114e-05,
+      "loss": 0.7408,
+      "step": 197
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.49516833938133936,
+      "learning_rate": 4.9842911009188024e-05,
+      "loss": 0.7363,
+      "step": 198
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.40496580832539364,
+      "learning_rate": 4.9839936266857156e-05,
+      "loss": 0.7125,
+      "step": 199
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.41390815507326556,
+      "learning_rate": 4.983693371268214e-05,
+      "loss": 0.9519,
+      "step": 200
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.4550836198890942,
+      "learning_rate": 4.983390335002477e-05,
+      "loss": 0.7109,
+      "step": 201
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.44374668945283674,
+      "learning_rate": 4.9830845182277974e-05,
+      "loss": 0.7079,
+      "step": 202
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.3501967204626946,
+      "learning_rate": 4.9827759212865795e-05,
+      "loss": 0.9979,
+      "step": 203
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7251275511471813,
+      "learning_rate": 4.982464544524343e-05,
+      "loss": 0.7343,
+      "step": 204
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.44255373460318553,
+      "learning_rate": 4.982150388289718e-05,
+      "loss": 0.6992,
+      "step": 205
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.3858233284170097,
+      "learning_rate": 4.9818334529344455e-05,
+      "loss": 0.9686,
+      "step": 206
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.42399659582625276,
+      "learning_rate": 4.981513738813382e-05,
+      "loss": 0.7254,
+      "step": 207
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.46090222223060706,
+      "learning_rate": 4.9811912462844914e-05,
+      "loss": 0.6993,
+      "step": 208
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.37798152754335634,
+      "learning_rate": 4.980865975708851e-05,
+      "loss": 0.7388,
+      "step": 209
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.41715440893810246,
+      "learning_rate": 4.980537927450647e-05,
+      "loss": 0.7055,
+      "step": 210
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.4142253379343246,
+      "learning_rate": 4.9802071018771764e-05,
+      "loss": 0.951,
+      "step": 211
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.3885917816246488,
+      "learning_rate": 4.9798734993588446e-05,
+      "loss": 0.7229,
+      "step": 212
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.42470786773266117,
+      "learning_rate": 4.9795371202691686e-05,
+      "loss": 0.6996,
+      "step": 213
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.43268214894608603,
+      "learning_rate": 4.979197964984771e-05,
+      "loss": 0.9572,
+      "step": 214
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.40923823827132655,
+      "learning_rate": 4.978856033885386e-05,
+      "loss": 0.6961,
+      "step": 215
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5857706119032127,
+      "learning_rate": 4.978511327353853e-05,
+      "loss": 0.7249,
+      "step": 216
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.4749764698493411,
+      "learning_rate": 4.978163845776121e-05,
+      "loss": 0.9539,
+      "step": 217
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.53393189947027,
+      "learning_rate": 4.977813589541244e-05,
+      "loss": 0.6965,
+      "step": 218
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5988262650471734,
+      "learning_rate": 4.977460559041385e-05,
+      "loss": 0.7262,
+      "step": 219
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.3717157873229088,
+      "learning_rate": 4.9771047546718104e-05,
+      "loss": 0.7405,
+      "step": 220
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.41701300043880013,
+      "learning_rate": 4.976746176830894e-05,
+      "loss": 0.9535,
+      "step": 221
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.3973368940198127,
+      "learning_rate": 4.9763848259201166e-05,
+      "loss": 0.7354,
+      "step": 222
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.4376329320075737,
+      "learning_rate": 4.9760207023440604e-05,
+      "loss": 0.7189,
+      "step": 223
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.4969947504226051,
+      "learning_rate": 4.975653806510413e-05,
+      "loss": 0.7385,
+      "step": 224
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.363488628392514,
+      "learning_rate": 4.9752841388299684e-05,
+      "loss": 0.9764,
+      "step": 225
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.4623741396278026,
+      "learning_rate": 4.974911699716621e-05,
+      "loss": 0.7564,
+      "step": 226
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.4221570000508353,
+      "learning_rate": 4.9745364895873695e-05,
+      "loss": 0.6712,
+      "step": 227
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.35675941714574255,
+      "learning_rate": 4.9741585088623144e-05,
+      "loss": 0.9413,
+      "step": 228
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.3967835606598053,
+      "learning_rate": 4.9737777579646595e-05,
+      "loss": 0.7089,
+      "step": 229
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.4321313839529362,
+      "learning_rate": 4.9733942373207104e-05,
+      "loss": 0.7151,
+      "step": 230
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.4433219881142451,
+      "learning_rate": 4.973007947359872e-05,
+      "loss": 0.8069,
+      "step": 231
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.496052266649088,
+      "learning_rate": 4.972618888514652e-05,
+      "loss": 0.9655,
+      "step": 232
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.38784085774304267,
+      "learning_rate": 4.9722270612206555e-05,
+      "loss": 0.7161,
+      "step": 233
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5078988669434336,
+      "learning_rate": 4.97183246591659e-05,
+      "loss": 0.6946,
+      "step": 234
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.49764363243525056,
+      "learning_rate": 4.971435103044261e-05,
+      "loss": 0.9599,
+      "step": 235
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.48477036858646966,
+      "learning_rate": 4.9710349730485725e-05,
+      "loss": 0.7103,
+      "step": 236
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.510847219060672,
+      "learning_rate": 4.970632076377528e-05,
+      "loss": 0.7527,
+      "step": 237
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5080988263981355,
+      "learning_rate": 4.970226413482226e-05,
+      "loss": 0.7265,
+      "step": 238
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.4681822587304612,
+      "learning_rate": 4.969817984816865e-05,
+      "loss": 0.7173,
+      "step": 239
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.34825254208229506,
+      "learning_rate": 4.969406790838739e-05,
+      "loss": 0.9774,
+      "step": 240
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5286088990470657,
+      "learning_rate": 4.968992832008238e-05,
+      "loss": 0.7041,
+      "step": 241
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.4744262788372777,
+      "learning_rate": 4.9685761087888475e-05,
+      "loss": 0.7386,
+      "step": 242
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.40330579811511136,
+      "learning_rate": 4.968156621647149e-05,
+      "loss": 0.96,
+      "step": 243
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.41250937747319844,
+      "learning_rate": 4.967734371052817e-05,
+      "loss": 0.6912,
+      "step": 244
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.3859025396840167,
+      "learning_rate": 4.9673093574786236e-05,
+      "loss": 0.6963,
+      "step": 245
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.3914926826478865,
+      "learning_rate": 4.966881581400429e-05,
+      "loss": 0.9797,
+      "step": 246
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.40988442876449765,
+      "learning_rate": 4.966451043297191e-05,
+      "loss": 0.7075,
+      "step": 247
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.408534610101867,
+      "learning_rate": 4.966017743650957e-05,
+      "loss": 0.764,
+      "step": 248
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.38916017957326016,
+      "learning_rate": 4.9655816829468706e-05,
+      "loss": 0.6925,
+      "step": 249
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.4015029790994531,
+      "learning_rate": 4.96514286167316e-05,
+      "loss": 0.6993,
+      "step": 250
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.3585056042676868,
+      "learning_rate": 4.9647012803211515e-05,
+      "loss": 0.9231,
+      "step": 251
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.40045842956952415,
+      "learning_rate": 4.964256939385255e-05,
+      "loss": 0.7139,
+      "step": 252
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5217862669222747,
+      "learning_rate": 4.963809839362976e-05,
+      "loss": 0.7071,
+      "step": 253
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.47159794648210673,
+      "learning_rate": 4.9633599807549065e-05,
+      "loss": 0.9668,
+      "step": 254
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.394144164401167,
+      "learning_rate": 4.962907364064725e-05,
+      "loss": 0.7269,
+      "step": 255
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5606342840299178,
+      "learning_rate": 4.9624519897992026e-05,
+      "loss": 0.6973,
+      "step": 256
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.40735393209845816,
+      "learning_rate": 4.9619938584681954e-05,
+      "loss": 0.9701,
+      "step": 257
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.4740593819723568,
+      "learning_rate": 4.961532970584645e-05,
+      "loss": 0.7226,
+      "step": 258
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6150000473210486,
+      "learning_rate": 4.9610693266645826e-05,
+      "loss": 0.7271,
+      "step": 259
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.40110304541754327,
+      "learning_rate": 4.9606029272271224e-05,
+      "loss": 0.698,
+      "step": 260
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.4568781098986371,
+      "learning_rate": 4.9601337727944665e-05,
+      "loss": 0.9586,
+      "step": 261
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.4385825324479858,
+      "learning_rate": 4.959661863891898e-05,
+      "loss": 0.732,
+      "step": 262
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.48149372844973326,
+      "learning_rate": 4.959187201047788e-05,
+      "loss": 0.7037,
+      "step": 263
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.43228780031169345,
+      "learning_rate": 4.958709784793587e-05,
+      "loss": 0.7186,
+      "step": 264
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.38297946768422614,
+      "learning_rate": 4.958229615663833e-05,
+      "loss": 0.9663,
+      "step": 265
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.4210106428452665,
+      "learning_rate": 4.957746694196142e-05,
+      "loss": 0.6782,
+      "step": 266
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.46677900180367604,
+      "learning_rate": 4.9572610209312135e-05,
+      "loss": 0.6885,
+      "step": 267
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.36526389344566385,
+      "learning_rate": 4.9567725964128284e-05,
+      "loss": 0.9452,
+      "step": 268
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.4233679858939789,
+      "learning_rate": 4.956281421187847e-05,
+      "loss": 0.7074,
+      "step": 269
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.43600904534549273,
+      "learning_rate": 4.95578749580621e-05,
+      "loss": 0.7413,
+      "step": 270
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.3731134751261524,
+      "learning_rate": 4.955290820820937e-05,
+      "loss": 0.6969,
+      "step": 271
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.4005931844490536,
+      "learning_rate": 4.954791396788127e-05,
+      "loss": 0.9783,
+      "step": 272
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.4717655903240338,
+      "learning_rate": 4.9542892242669547e-05,
+      "loss": 0.7117,
+      "step": 273
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.3820392664778176,
+      "learning_rate": 4.9537843038196764e-05,
+      "loss": 0.6921,
+      "step": 274
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.39594042900675397,
+      "learning_rate": 4.9532766360116195e-05,
+      "loss": 0.9506,
+      "step": 275
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.37234658754518213,
+      "learning_rate": 4.9527662214111924e-05,
+      "loss": 0.7264,
+      "step": 276
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.42464470635694057,
+      "learning_rate": 4.9522530605898755e-05,
+      "loss": 0.6818,
+      "step": 277
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.3859778178946074,
+      "learning_rate": 4.951737154122227e-05,
+      "loss": 0.9591,
+      "step": 278
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.3637562995627391,
+      "learning_rate": 4.951218502585876e-05,
+      "loss": 0.6645,
+      "step": 279
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5028396993631085,
+      "learning_rate": 4.950697106561527e-05,
+      "loss": 0.6886,
+      "step": 280
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5629953663863988,
+      "learning_rate": 4.950172966632958e-05,
+      "loss": 0.7378,
+      "step": 281
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.40221976772437057,
+      "learning_rate": 4.9496460833870184e-05,
+      "loss": 0.6938,
+      "step": 282
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.418363143128043,
+      "learning_rate": 4.949116457413627e-05,
+      "loss": 0.967,
+      "step": 283
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.4063011947317089,
+      "learning_rate": 4.948584089305778e-05,
+      "loss": 0.7015,
+      "step": 284
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.4820621130481272,
+      "learning_rate": 4.948048979659531e-05,
+      "loss": 0.6819,
+      "step": 285
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.37135415190858523,
+      "learning_rate": 4.947511129074018e-05,
+      "loss": 0.9488,
+      "step": 286
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.41465448489348883,
+      "learning_rate": 4.9469705381514406e-05,
+      "loss": 0.6891,
+      "step": 287
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5197800219742525,
+      "learning_rate": 4.9464272074970666e-05,
+      "loss": 0.6892,
+      "step": 288
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.34942997823106053,
+      "learning_rate": 4.9458811377192315e-05,
+      "loss": 0.9631,
+      "step": 289
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.4920775803088157,
+      "learning_rate": 4.9453323294293384e-05,
+      "loss": 0.716,
+      "step": 290
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.45423562369815385,
+      "learning_rate": 4.944780783241856e-05,
+      "loss": 0.6752,
+      "step": 291
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.44019844136885045,
+      "learning_rate": 4.9442264997743196e-05,
+      "loss": 0.7546,
+      "step": 292
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.4361419629115155,
+      "learning_rate": 4.9436694796473284e-05,
+      "loss": 0.7006,
+      "step": 293
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.37351564023925105,
+      "learning_rate": 4.943109723484544e-05,
+      "loss": 0.9444,
+      "step": 294
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.41573861046707145,
+      "learning_rate": 4.9425472319126954e-05,
+      "loss": 0.702,
+      "step": 295
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.4293534326291938,
+      "learning_rate": 4.9419820055615694e-05,
+      "loss": 0.6949,
+      "step": 296
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.3956404839189456,
+      "learning_rate": 4.94141404506402e-05,
+      "loss": 0.9778,
+      "step": 297
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.44162788180557705,
+      "learning_rate": 4.940843351055959e-05,
+      "loss": 0.6758,
+      "step": 298
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.419336594929885,
+      "learning_rate": 4.9402699241763584e-05,
+      "loss": 0.6831,
+      "step": 299
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.3522734321392301,
+      "learning_rate": 4.9396937650672525e-05,
+      "loss": 0.9427,
+      "step": 300
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.3723975140730089,
+      "learning_rate": 4.939114874373733e-05,
+      "loss": 0.696,
+      "step": 301
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.3987586160693117,
+      "learning_rate": 4.93853325274395e-05,
+      "loss": 0.703,
+      "step": 302
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.39649598321268764,
+      "learning_rate": 4.9379489008291125e-05,
+      "loss": 0.7308,
+      "step": 303
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.3602783352075075,
+      "learning_rate": 4.937361819283485e-05,
+      "loss": 0.9371,
+      "step": 304
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.3913433171149988,
+      "learning_rate": 4.936772008764389e-05,
+      "loss": 0.7123,
+      "step": 305
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.3815346180246776,
+      "learning_rate": 4.936179469932202e-05,
+      "loss": 0.6868,
+      "step": 306
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.36693133701616176,
+      "learning_rate": 4.935584203450354e-05,
+      "loss": 0.6945,
+      "step": 307
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.4064289190516712,
+      "learning_rate": 4.9349862099853315e-05,
+      "loss": 0.9413,
+      "step": 308
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.3720597638734654,
+      "learning_rate": 4.9343854902066746e-05,
+      "loss": 0.6922,
+      "step": 309
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.4085425283775875,
+      "learning_rate": 4.933782044786972e-05,
+      "loss": 0.7158,
+      "step": 310
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.36171211639032247,
+      "learning_rate": 4.933175874401868e-05,
+      "loss": 0.9504,
+      "step": 311
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.39368619829871776,
+      "learning_rate": 4.9325669797300576e-05,
+      "loss": 0.6785,
+      "step": 312
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.44056732410356314,
+      "learning_rate": 4.931955361453284e-05,
+      "loss": 0.6826,
+      "step": 313
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.4210424532903724,
+      "learning_rate": 4.9313410202563405e-05,
+      "loss": 0.7487,
+      "step": 314
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.3436944000167263,
+      "learning_rate": 4.93072395682707e-05,
+      "loss": 0.9311,
+      "step": 315
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.3764767410289898,
+      "learning_rate": 4.9301041718563636e-05,
+      "loss": 0.7003,
+      "step": 316
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5983682584702216,
+      "learning_rate": 4.929481666038158e-05,
+      "loss": 0.7378,
+      "step": 317
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.3819423225989554,
+      "learning_rate": 4.928856440069436e-05,
+      "loss": 0.9654,
+      "step": 318
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.35612527589805715,
+      "learning_rate": 4.9282284946502286e-05,
+      "loss": 0.6958,
+      "step": 319
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.4090608944209128,
+      "learning_rate": 4.9275978304836106e-05,
+      "loss": 0.6795,
+      "step": 320
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.3912420916469698,
+      "learning_rate": 4.926964448275699e-05,
+      "loss": 0.9356,
+      "step": 321
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.39226157839155057,
+      "learning_rate": 4.926328348735655e-05,
+      "loss": 0.69,
+      "step": 322
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.43450878909500334,
+      "learning_rate": 4.9256895325756836e-05,
+      "loss": 0.6957,
+      "step": 323
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.3932728297270571,
+      "learning_rate": 4.92504800051103e-05,
+      "loss": 0.7265,
+      "step": 324
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4469217023179367,
+      "learning_rate": 4.9244037532599815e-05,
+      "loss": 0.7616,
+      "step": 325
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.37872408385481127,
+      "learning_rate": 4.923756791543863e-05,
+      "loss": 0.9264,
+      "step": 326
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4060385517725805,
+      "learning_rate": 4.923107116087041e-05,
+      "loss": 0.6953,
+      "step": 327
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.496374714044711,
+      "learning_rate": 4.92245472761692e-05,
+      "loss": 0.6961,
+      "step": 328
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.33891038193950784,
+      "learning_rate": 4.9217996268639424e-05,
+      "loss": 0.9781,
+      "step": 329
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5176872867997517,
+      "learning_rate": 4.9211418145615837e-05,
+      "loss": 0.6731,
+      "step": 330
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6502990898125993,
+      "learning_rate": 4.9204812914463614e-05,
+      "loss": 0.6862,
+      "step": 331
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.3993647811280849,
+      "learning_rate": 4.919818058257825e-05,
+      "loss": 0.9134,
+      "step": 332
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.512734643363362,
+      "learning_rate": 4.919152115738557e-05,
+      "loss": 0.6625,
+      "step": 333
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6036722356990293,
+      "learning_rate": 4.918483464634176e-05,
+      "loss": 0.6941,
+      "step": 334
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4066731277083294,
+      "learning_rate": 4.9178121056933306e-05,
+      "loss": 0.9255,
+      "step": 335
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5600025386652908,
+      "learning_rate": 4.917138039667705e-05,
+      "loss": 0.5084,
+      "step": 336
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5085975555778626,
+      "learning_rate": 4.9164612673120095e-05,
+      "loss": 0.955,
+      "step": 337
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4882498539166704,
+      "learning_rate": 4.9157817893839885e-05,
+      "loss": 0.6682,
+      "step": 338
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6602828805538752,
+      "learning_rate": 4.9150996066444144e-05,
+      "loss": 0.7233,
+      "step": 339
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.35117243608425364,
+      "learning_rate": 4.9144147198570874e-05,
+      "loss": 0.9438,
+      "step": 340
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5877744102637726,
+      "learning_rate": 4.913727129788836e-05,
+      "loss": 0.6924,
+      "step": 341
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.45114157400514526,
+      "learning_rate": 4.913036837209515e-05,
+      "loss": 0.7089,
+      "step": 342
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4124527558286833,
+      "learning_rate": 4.9123438428920044e-05,
+      "loss": 0.9697,
+      "step": 343
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4224824565998749,
+      "learning_rate": 4.911648147612212e-05,
+      "loss": 0.6845,
+      "step": 344
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4333033110797867,
+      "learning_rate": 4.9109497521490653e-05,
+      "loss": 0.6751,
+      "step": 345
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.3669560095825111,
+      "learning_rate": 4.91024865728452e-05,
+      "loss": 0.8946,
+      "step": 346
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4691968393183868,
+      "learning_rate": 4.909544863803549e-05,
+      "loss": 0.7217,
+      "step": 347
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4161491264293211,
+      "learning_rate": 4.908838372494151e-05,
+      "loss": 0.699,
+      "step": 348
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.38705185387066104,
+      "learning_rate": 4.908129184147343e-05,
+      "loss": 0.681,
+      "step": 349
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4114063112161177,
+      "learning_rate": 4.907417299557162e-05,
+      "loss": 0.6975,
+      "step": 350
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.3346498648551156,
+      "learning_rate": 4.906702719520665e-05,
+      "loss": 0.8969,
+      "step": 351
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.41782166665005777,
+      "learning_rate": 4.905985444837925e-05,
+      "loss": 0.7195,
+      "step": 352
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.35993416170833864,
+      "learning_rate": 4.9052654763120334e-05,
+      "loss": 0.6797,
+      "step": 353
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.3966186856178466,
+      "learning_rate": 4.904542814749098e-05,
+      "loss": 0.6728,
+      "step": 354
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.3637639846454043,
+      "learning_rate": 4.903817460958241e-05,
+      "loss": 0.8394,
+      "step": 355
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.3546350878006171,
+      "learning_rate": 4.903089415751599e-05,
+      "loss": 0.7036,
+      "step": 356
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.3441071184214713,
+      "learning_rate": 4.902358679944322e-05,
+      "loss": 0.9004,
+      "step": 357
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.38019824110885203,
+      "learning_rate": 4.901625254354574e-05,
+      "loss": 0.7413,
+      "step": 358
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.33723273570268897,
+      "learning_rate": 4.9008891398035276e-05,
+      "loss": 0.7024,
+      "step": 359
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.32943861757542053,
+      "learning_rate": 4.900150337115369e-05,
+      "loss": 0.6769,
+      "step": 360
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.33624657302478417,
+      "learning_rate": 4.8994088471172925e-05,
+      "loss": 0.9548,
+      "step": 361
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.3791941017472018,
+      "learning_rate": 4.898664670639502e-05,
+      "loss": 0.6776,
+      "step": 362
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.365141439695109,
+      "learning_rate": 4.897917808515209e-05,
+      "loss": 0.7159,
+      "step": 363
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.3226186758105058,
+      "learning_rate": 4.897168261580631e-05,
+      "loss": 0.6767,
+      "step": 364
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.3615933825770506,
+      "learning_rate": 4.896416030674994e-05,
+      "loss": 0.6719,
+      "step": 365
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.31638139259805487,
+      "learning_rate": 4.8956611166405255e-05,
+      "loss": 0.9511,
+      "step": 366
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.37228926686190805,
+      "learning_rate": 4.8949035203224614e-05,
+      "loss": 0.6902,
+      "step": 367
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.38789369478387903,
+      "learning_rate": 4.894143242569038e-05,
+      "loss": 0.7065,
+      "step": 368
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.35939739587997144,
+      "learning_rate": 4.8933802842314934e-05,
+      "loss": 0.9512,
+      "step": 369
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.39141910952593323,
+      "learning_rate": 4.89261464616407e-05,
+      "loss": 0.6959,
+      "step": 370
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.38233936024848536,
+      "learning_rate": 4.8918463292240056e-05,
+      "loss": 0.6759,
+      "step": 371
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.35947185377081875,
+      "learning_rate": 4.891075334271543e-05,
+      "loss": 0.9409,
+      "step": 372
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.3817088449206171,
+      "learning_rate": 4.8903016621699205e-05,
+      "loss": 0.6791,
+      "step": 373
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.3504951506087064,
+      "learning_rate": 4.8895253137853735e-05,
+      "loss": 0.7312,
+      "step": 374
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.36444983596888947,
+      "learning_rate": 4.888746289987134e-05,
+      "loss": 0.6768,
+      "step": 375
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.4276700974361382,
+      "learning_rate": 4.887964591647431e-05,
+      "loss": 0.6717,
+      "step": 376
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.36131744127739734,
+      "learning_rate": 4.887180219641486e-05,
+      "loss": 0.9479,
+      "step": 377
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.3564044593489454,
+      "learning_rate": 4.8863931748475165e-05,
+      "loss": 0.6815,
+      "step": 378
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.3524266547952009,
+      "learning_rate": 4.88560345814673e-05,
+      "loss": 0.673,
+      "step": 379
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.438753119098464,
+      "learning_rate": 4.884811070423328e-05,
+      "loss": 0.9889,
+      "step": 380
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.32373024386955457,
+      "learning_rate": 4.8840160125644996e-05,
+      "loss": 0.7016,
+      "step": 381
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.3440571761597369,
+      "learning_rate": 4.883218285460427e-05,
+      "loss": 0.6534,
+      "step": 382
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.31636847831136944,
+      "learning_rate": 4.8824178900042775e-05,
+      "loss": 0.9457,
+      "step": 383
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.3747310805732178,
+      "learning_rate": 4.881614827092209e-05,
+      "loss": 0.6744,
+      "step": 384
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.3965951147181902,
+      "learning_rate": 4.8808090976233644e-05,
+      "loss": 0.7179,
+      "step": 385
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.31670021332219195,
+      "learning_rate": 4.880000702499873e-05,
+      "loss": 0.6565,
+      "step": 386
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.4414802500936794,
+      "learning_rate": 4.879189642626847e-05,
+      "loss": 0.9676,
+      "step": 387
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.3369190065135386,
+      "learning_rate": 4.8783759189123836e-05,
+      "loss": 0.6938,
+      "step": 388
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.40631959829842923,
+      "learning_rate": 4.877559532267564e-05,
+      "loss": 0.6675,
+      "step": 389
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.34923431261921034,
+      "learning_rate": 4.876740483606447e-05,
+      "loss": 0.9538,
+      "step": 390
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.3846382777544546,
+      "learning_rate": 4.8759187738460755e-05,
+      "loss": 0.6849,
+      "step": 391
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.46183191400776913,
+      "learning_rate": 4.87509440390647e-05,
+      "loss": 0.6705,
+      "step": 392
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.3618141609824665,
+      "learning_rate": 4.8742673747106305e-05,
+      "loss": 0.6761,
+      "step": 393
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.34485414540664877,
+      "learning_rate": 4.8734376871845334e-05,
+      "loss": 0.954,
+      "step": 394
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.34621162346947904,
+      "learning_rate": 4.872605342257132e-05,
+      "loss": 0.6914,
+      "step": 395
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.3809402893286568,
+      "learning_rate": 4.871770340860355e-05,
+      "loss": 0.7209,
+      "step": 396
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.35349857587666933,
+      "learning_rate": 4.8709326839291055e-05,
+      "loss": 0.6789,
+      "step": 397
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.34895781172678647,
+      "learning_rate": 4.8700923724012595e-05,
+      "loss": 0.8995,
+      "step": 398
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.3612585906684026,
+      "learning_rate": 4.869249407217665e-05,
+      "loss": 0.6727,
+      "step": 399
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.4015498937918312,
+      "learning_rate": 4.868403789322143e-05,
+      "loss": 0.6727,
+      "step": 400
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.33253208087975766,
+      "learning_rate": 4.867555519661481e-05,
+      "loss": 0.8947,
+      "step": 401
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.40846401771270946,
+      "learning_rate": 4.866704599185439e-05,
+      "loss": 0.7207,
+      "step": 402
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.357657448356932,
+      "learning_rate": 4.865851028846743e-05,
+      "loss": 0.6745,
+      "step": 403
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.32866174397425924,
+      "learning_rate": 4.864994809601087e-05,
+      "loss": 0.9732,
+      "step": 404
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.36128724072146323,
+      "learning_rate": 4.86413594240713e-05,
+      "loss": 0.6952,
+      "step": 405
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.34601075876548615,
+      "learning_rate": 4.8632744282264954e-05,
+      "loss": 0.681,
+      "step": 406
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.3401426939957497,
+      "learning_rate": 4.862410268023771e-05,
+      "loss": 0.7444,
+      "step": 407
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.38496026288298374,
+      "learning_rate": 4.86154346276651e-05,
+      "loss": 0.6557,
+      "step": 408
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.36810203481362075,
+      "learning_rate": 4.860674013425221e-05,
+      "loss": 0.949,
+      "step": 409
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.37542407791176724,
+      "learning_rate": 4.859801920973376e-05,
+      "loss": 0.6577,
+      "step": 410
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.40960165147705013,
+      "learning_rate": 4.858927186387408e-05,
+      "loss": 0.6617,
+      "step": 411
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.3160094008486573,
+      "learning_rate": 4.858049810646709e-05,
+      "loss": 0.9165,
+      "step": 412
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.40077011845305077,
+      "learning_rate": 4.857169794733622e-05,
+      "loss": 0.6657,
+      "step": 413
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.3679520982329935,
+      "learning_rate": 4.856287139633452e-05,
+      "loss": 0.675,
+      "step": 414
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.31063984217675017,
+      "learning_rate": 4.855401846334458e-05,
+      "loss": 0.9602,
+      "step": 415
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.3703206024524539,
+      "learning_rate": 4.85451391582785e-05,
+      "loss": 0.6589,
+      "step": 416
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.3747382345373809,
+      "learning_rate": 4.8536233491077924e-05,
+      "loss": 0.6724,
+      "step": 417
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.320837420499059,
+      "learning_rate": 4.852730147171403e-05,
+      "loss": 0.7207,
+      "step": 418
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.358587448708474,
+      "learning_rate": 4.8518343110187484e-05,
+      "loss": 0.6642,
+      "step": 419
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.31995404146302603,
+      "learning_rate": 4.850935841652843e-05,
+      "loss": 0.9235,
+      "step": 420
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.36889164097816285,
+      "learning_rate": 4.850034740079653e-05,
+      "loss": 0.6645,
+      "step": 421
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.35448732515796894,
+      "learning_rate": 4.8491310073080886e-05,
+      "loss": 0.6959,
+      "step": 422
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.3038542590009034,
+      "learning_rate": 4.848224644350008e-05,
+      "loss": 0.9146,
+      "step": 423
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.43445793757418394,
+      "learning_rate": 4.847315652220214e-05,
+      "loss": 0.6849,
+      "step": 424
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.3825865565242113,
+      "learning_rate": 4.8464040319364515e-05,
+      "loss": 0.6791,
+      "step": 425
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.34081121550691085,
+      "learning_rate": 4.8454897845194094e-05,
+      "loss": 0.9216,
+      "step": 426
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.3714975065927011,
+      "learning_rate": 4.844572910992717e-05,
+      "loss": 0.6768,
+      "step": 427
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.3233399327847615,
+      "learning_rate": 4.843653412382948e-05,
+      "loss": 0.6767,
+      "step": 428
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.3466856076463206,
+      "learning_rate": 4.842731289719608e-05,
+      "loss": 0.732,
+      "step": 429
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.3321046638785346,
+      "learning_rate": 4.841806544035147e-05,
+      "loss": 0.9456,
+      "step": 430
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.3603454776438944,
+      "learning_rate": 4.840879176364948e-05,
+      "loss": 0.6881,
+      "step": 431
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.3030006846493396,
+      "learning_rate": 4.8399491877473324e-05,
+      "loss": 0.6649,
+      "step": 432
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.3263262247125177,
+      "learning_rate": 4.8390165792235534e-05,
+      "loss": 0.958,
+      "step": 433
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.31619690043527676,
+      "learning_rate": 4.838081351837801e-05,
+      "loss": 0.6782,
+      "step": 434
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.29577598342260747,
+      "learning_rate": 4.8371435066371935e-05,
+      "loss": 0.6717,
+      "step": 435
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.3235814366357723,
+      "learning_rate": 4.836203044671783e-05,
+      "loss": 0.7026,
+      "step": 436
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.3078864429255855,
+      "learning_rate": 4.8352599669945496e-05,
+      "loss": 0.9037,
+      "step": 437
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.3179255982198936,
+      "learning_rate": 4.8343142746614037e-05,
+      "loss": 0.6624,
+      "step": 438
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.32233746096922317,
+      "learning_rate": 4.8333659687311836e-05,
+      "loss": 0.6766,
+      "step": 439
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.33070516269785855,
+      "learning_rate": 4.832415050265651e-05,
+      "loss": 0.7164,
+      "step": 440
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.3272747348492407,
+      "learning_rate": 4.831461520329494e-05,
+      "loss": 0.9121,
+      "step": 441
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.3347555827443948,
+      "learning_rate": 4.830505379990328e-05,
+      "loss": 0.6788,
+      "step": 442
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.3234555046564182,
+      "learning_rate": 4.829546630318685e-05,
+      "loss": 0.6883,
+      "step": 443
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.340951664052272,
+      "learning_rate": 4.828585272388023e-05,
+      "loss": 0.9369,
+      "step": 444
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.320643712079137,
+      "learning_rate": 4.8276213072747194e-05,
+      "loss": 0.6664,
+      "step": 445
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.41799252485236044,
+      "learning_rate": 4.8266547360580704e-05,
+      "loss": 0.6871,
+      "step": 446
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.3470353775868447,
+      "learning_rate": 4.8256855598202896e-05,
+      "loss": 0.9426,
+      "step": 447
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.373009610723888,
+      "learning_rate": 4.824713779646507e-05,
+      "loss": 0.6745,
+      "step": 448
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.48654109569431053,
+      "learning_rate": 4.8237393966247706e-05,
+      "loss": 0.6621,
+      "step": 449
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.36298062673259435,
+      "learning_rate": 4.8227624118460395e-05,
+      "loss": 0.9384,
+      "step": 450
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.4724799851091046,
+      "learning_rate": 4.8217828264041875e-05,
+      "loss": 0.5024,
+      "step": 451
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5212547783812888,
+      "learning_rate": 4.8208006413960004e-05,
+      "loss": 0.9405,
+      "step": 452
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.4300466335095121,
+      "learning_rate": 4.819815857921174e-05,
+      "loss": 0.6852,
+      "step": 453
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5749572971510649,
+      "learning_rate": 4.818828477082313e-05,
+      "loss": 0.6552,
+      "step": 454
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.344822693159411,
+      "learning_rate": 4.8178384999849315e-05,
+      "loss": 0.9225,
+      "step": 455
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5392025988007628,
+      "learning_rate": 4.816845927737449e-05,
+      "loss": 0.6783,
+      "step": 456
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.4516627207524827,
+      "learning_rate": 4.8158507614511916e-05,
+      "loss": 0.6661,
+      "step": 457
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.39407844163388384,
+      "learning_rate": 4.8148530022403896e-05,
+      "loss": 0.9501,
+      "step": 458
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.390486283525756,
+      "learning_rate": 4.813852651222177e-05,
+      "loss": 0.6702,
+      "step": 459
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.4474833397662032,
+      "learning_rate": 4.812849709516589e-05,
+      "loss": 0.6657,
+      "step": 460
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.3298706375912751,
+      "learning_rate": 4.811844178246561e-05,
+      "loss": 0.9062,
+      "step": 461
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.4981766891905329,
+      "learning_rate": 4.810836058537929e-05,
+      "loss": 0.4837,
+      "step": 462
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.39405682567545824,
+      "learning_rate": 4.809825351519426e-05,
+      "loss": 0.934,
+      "step": 463
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.4654717951181095,
+      "learning_rate": 4.808812058322683e-05,
+      "loss": 0.6719,
+      "step": 464
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.4779732256270801,
+      "learning_rate": 4.8077961800822254e-05,
+      "loss": 0.6828,
+      "step": 465
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.35222455360689353,
+      "learning_rate": 4.806777717935474e-05,
+      "loss": 0.9635,
+      "step": 466
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.41466500911713183,
+      "learning_rate": 4.805756673022741e-05,
+      "loss": 0.6715,
+      "step": 467
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.4305882019211664,
+      "learning_rate": 4.8047330464872334e-05,
+      "loss": 0.6716,
+      "step": 468
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.34156228315764015,
+      "learning_rate": 4.803706839475044e-05,
+      "loss": 0.9614,
+      "step": 469
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.4211565296608561,
+      "learning_rate": 4.802678053135161e-05,
+      "loss": 0.6785,
+      "step": 470
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.3740690523552254,
+      "learning_rate": 4.8016466886194544e-05,
+      "loss": 0.6715,
+      "step": 471
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.35564693316565527,
+      "learning_rate": 4.800612747082685e-05,
+      "loss": 0.9311,
+      "step": 472
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.3939873810206657,
+      "learning_rate": 4.799576229682498e-05,
+      "loss": 0.7665,
+      "step": 473
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.3743206881686847,
+      "learning_rate": 4.7985371375794194e-05,
+      "loss": 0.669,
+      "step": 474
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.3756618255160079,
+      "learning_rate": 4.797495471936864e-05,
+      "loss": 0.6523,
+      "step": 475
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.3557424400923363,
+      "learning_rate": 4.796451233921122e-05,
+      "loss": 0.9052,
+      "step": 476
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.3767425738951787,
+      "learning_rate": 4.795404424701369e-05,
+      "loss": 0.6726,
+      "step": 477
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.36956247181845064,
+      "learning_rate": 4.794355045449656e-05,
+      "loss": 0.6679,
+      "step": 478
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.635700188627281,
+      "learning_rate": 4.793303097340911e-05,
+      "loss": 0.7078,
+      "step": 479
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.3840906473080569,
+      "learning_rate": 4.792248581552942e-05,
+      "loss": 0.6553,
+      "step": 480
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.3255816462631777,
+      "learning_rate": 4.7911914992664267e-05,
+      "loss": 0.9154,
+      "step": 481
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.4121676496198656,
+      "learning_rate": 4.790131851664922e-05,
+      "loss": 0.6388,
+      "step": 482
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.3836409831544727,
+      "learning_rate": 4.789069639934853e-05,
+      "loss": 0.9416,
+      "step": 483
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.3962819084992907,
+      "learning_rate": 4.788004865265516e-05,
+      "loss": 0.7318,
+      "step": 484
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.3926965060829225,
+      "learning_rate": 4.78693752884908e-05,
+      "loss": 0.6733,
+      "step": 485
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.4763781959889562,
+      "learning_rate": 4.785867631880577e-05,
+      "loss": 0.6603,
+      "step": 486
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.3488008926751173,
+      "learning_rate": 4.7847951755579124e-05,
+      "loss": 0.9174,
+      "step": 487
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4316376226734586,
+      "learning_rate": 4.783720161081852e-05,
+      "loss": 0.6998,
+      "step": 488
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.3819953874121938,
+      "learning_rate": 4.782642589656026e-05,
+      "loss": 0.6587,
+      "step": 489
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.3802923167861821,
+      "learning_rate": 4.781562462486931e-05,
+      "loss": 0.7326,
+      "step": 490
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.3241706354973059,
+      "learning_rate": 4.780479780783923e-05,
+      "loss": 0.6573,
+      "step": 491
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.4007030695019951,
+      "learning_rate": 4.779394545759218e-05,
+      "loss": 0.946,
+      "step": 492
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.33211644827456427,
+      "learning_rate": 4.7783067586278906e-05,
+      "loss": 0.6537,
+      "step": 493
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.4044302455825178,
+      "learning_rate": 4.7772164206078737e-05,
+      "loss": 0.6728,
+      "step": 494
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.35069817451286067,
+      "learning_rate": 4.7761235329199547e-05,
+      "loss": 0.9112,
+      "step": 495
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.3554589578154093,
+      "learning_rate": 4.77502809678778e-05,
+      "loss": 0.6606,
+      "step": 496
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.3371680286635473,
+      "learning_rate": 4.773930113437843e-05,
+      "loss": 0.6622,
+      "step": 497
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.3387317576841068,
+      "learning_rate": 4.772829584099494e-05,
+      "loss": 0.9066,
+      "step": 498
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.3459103516892969,
+      "learning_rate": 4.7717265100049325e-05,
+      "loss": 0.6476,
+      "step": 499
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.3407936670473931,
+      "learning_rate": 4.770620892389205e-05,
+      "loss": 0.6732,
+      "step": 500
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.3324175416063468,
+      "learning_rate": 4.76951273249021e-05,
+      "loss": 0.6997,
+      "step": 501
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.3173105828792188,
+      "learning_rate": 4.768402031548689e-05,
+      "loss": 0.932,
+      "step": 502
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.3198446968159297,
+      "learning_rate": 4.7672887908082305e-05,
+      "loss": 0.6968,
+      "step": 503
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.29640403640165225,
+      "learning_rate": 4.766173011515265e-05,
+      "loss": 0.6565,
+      "step": 504
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.31262917365811665,
+      "learning_rate": 4.765054694919068e-05,
+      "loss": 0.6697,
+      "step": 505
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.29627220345019945,
+      "learning_rate": 4.763933842271752e-05,
+      "loss": 0.907,
+      "step": 506
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.31790662391498836,
+      "learning_rate": 4.7628104548282716e-05,
+      "loss": 0.6649,
+      "step": 507
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.2855030908860598,
+      "learning_rate": 4.76168453384642e-05,
+      "loss": 0.6524,
+      "step": 508
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.4151986793754677,
+      "learning_rate": 4.760556080586825e-05,
+      "loss": 0.9104,
+      "step": 509
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.36675382838199155,
+      "learning_rate": 4.7594250963129515e-05,
+      "loss": 0.6406,
+      "step": 510
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.2808948978254953,
+      "learning_rate": 4.758291582291096e-05,
+      "loss": 0.6709,
+      "step": 511
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.31360712258197015,
+      "learning_rate": 4.75715553979039e-05,
+      "loss": 0.7409,
+      "step": 512
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.3308358344100906,
+      "learning_rate": 4.756016970082795e-05,
+      "loss": 0.9004,
+      "step": 513
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.3507496224877065,
+      "learning_rate": 4.7548758744431006e-05,
+      "loss": 0.6772,
+      "step": 514
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.28013832079985235,
+      "learning_rate": 4.753732254148926e-05,
+      "loss": 0.6532,
+      "step": 515
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.3475145589793015,
+      "learning_rate": 4.752586110480718e-05,
+      "loss": 0.9068,
+      "step": 516
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.3108021191099516,
+      "learning_rate": 4.751437444721747e-05,
+      "loss": 0.6573,
+      "step": 517
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.3329315812957667,
+      "learning_rate": 4.750286258158107e-05,
+      "loss": 0.6537,
+      "step": 518
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.2994465718387245,
+      "learning_rate": 4.749132552078717e-05,
+      "loss": 0.6514,
+      "step": 519
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.3205722679750019,
+      "learning_rate": 4.7479763277753126e-05,
+      "loss": 0.9214,
+      "step": 520
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.31896829171522906,
+      "learning_rate": 4.746817586542454e-05,
+      "loss": 0.6575,
+      "step": 521
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5302186926950957,
+      "learning_rate": 4.7456563296775145e-05,
+      "loss": 0.6755,
+      "step": 522
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.33173406373290226,
+      "learning_rate": 4.744492558480689e-05,
+      "loss": 0.6946,
+      "step": 523
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.33418202586142454,
+      "learning_rate": 4.7433262742549835e-05,
+      "loss": 0.964,
+      "step": 524
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.32783102808525766,
+      "learning_rate": 4.7421574783062196e-05,
+      "loss": 0.6546,
+      "step": 525
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.2966340272178718,
+      "learning_rate": 4.74098617194303e-05,
+      "loss": 0.6562,
+      "step": 526
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.37051902180804985,
+      "learning_rate": 4.73981235647686e-05,
+      "loss": 0.9251,
+      "step": 527
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.31336040180088987,
+      "learning_rate": 4.7386360332219625e-05,
+      "loss": 0.6862,
+      "step": 528
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.31386857036494387,
+      "learning_rate": 4.737457203495399e-05,
+      "loss": 0.6462,
+      "step": 529
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.30375082277998305,
+      "learning_rate": 4.736275868617037e-05,
+      "loss": 0.9039,
+      "step": 530
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.3217339696757782,
+      "learning_rate": 4.7350920299095495e-05,
+      "loss": 0.6615,
+      "step": 531
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.3478794124810264,
+      "learning_rate": 4.733905688698411e-05,
+      "loss": 0.6682,
+      "step": 532
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.2992928067660319,
+      "learning_rate": 4.7327168463119013e-05,
+      "loss": 0.8856,
+      "step": 533
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.33275034631183875,
+      "learning_rate": 4.731525504081098e-05,
+      "loss": 0.4906,
+      "step": 534
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.3339426347460363,
+      "learning_rate": 4.7303316633398765e-05,
+      "loss": 0.9271,
+      "step": 535
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.32976631990936206,
+      "learning_rate": 4.729135325424914e-05,
+      "loss": 0.6689,
+      "step": 536
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.4293446398606489,
+      "learning_rate": 4.72793649167568e-05,
+      "loss": 0.6658,
+      "step": 537
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.39038302515236806,
+      "learning_rate": 4.726735163434439e-05,
+      "loss": 0.9131,
+      "step": 538
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.38803532556889225,
+      "learning_rate": 4.7255313420462497e-05,
+      "loss": 0.6679,
+      "step": 539
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.49535922408076877,
+      "learning_rate": 4.724325028858961e-05,
+      "loss": 0.6849,
+      "step": 540
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.34325370048986376,
+      "learning_rate": 4.723116225223213e-05,
+      "loss": 0.8963,
+      "step": 541
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.46314093999761485,
+      "learning_rate": 4.721904932492432e-05,
+      "loss": 0.6692,
+      "step": 542
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.4874326887317448,
+      "learning_rate": 4.7206911520228345e-05,
+      "loss": 0.6572,
+      "step": 543
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.3088033241805757,
+      "learning_rate": 4.719474885173419e-05,
+      "loss": 0.9449,
+      "step": 544
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.47800778657496296,
+      "learning_rate": 4.71825613330597e-05,
+      "loss": 0.4856,
+      "step": 545
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.35220817701950125,
+      "learning_rate": 4.717034897785054e-05,
+      "loss": 0.9061,
+      "step": 546
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.4152789986020536,
+      "learning_rate": 4.7158111799780166e-05,
+      "loss": 0.666,
+      "step": 547
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.40471621804674995,
+      "learning_rate": 4.7145849812549856e-05,
+      "loss": 0.6808,
+      "step": 548
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.3605488775657854,
+      "learning_rate": 4.713356302988864e-05,
+      "loss": 0.9034,
+      "step": 549
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.3817034322677417,
+      "learning_rate": 4.712125146555333e-05,
+      "loss": 0.6664,
+      "step": 550
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.2990799514797713,
+      "learning_rate": 4.7108915133328454e-05,
+      "loss": 0.6789,
+      "step": 551
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.3390775662216277,
+      "learning_rate": 4.709655404702632e-05,
+      "loss": 0.9088,
+      "step": 552
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.30465898864834257,
+      "learning_rate": 4.708416822048691e-05,
+      "loss": 0.6558,
+      "step": 553
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.3446538315522148,
+      "learning_rate": 4.707175766757791e-05,
+      "loss": 0.6406,
+      "step": 554
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.29762160791541975,
+      "learning_rate": 4.7059322402194715e-05,
+      "loss": 0.9438,
+      "step": 555
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.3504637472556722,
+      "learning_rate": 4.7046862438260364e-05,
+      "loss": 0.7209,
+      "step": 556
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.2977042157683329,
+      "learning_rate": 4.703437778972557e-05,
+      "loss": 0.6639,
+      "step": 557
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5377599233809938,
+      "learning_rate": 4.702186847056867e-05,
+      "loss": 0.6734,
+      "step": 558
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.326006725298342,
+      "learning_rate": 4.700933449479562e-05,
+      "loss": 0.9412,
+      "step": 559
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.34926849437089846,
+      "learning_rate": 4.699677587644e-05,
+      "loss": 0.6611,
+      "step": 560
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.3547308283754457,
+      "learning_rate": 4.698419262956297e-05,
+      "loss": 0.646,
+      "step": 561
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.3026727611129689,
+      "learning_rate": 4.6971584768253265e-05,
+      "loss": 0.6682,
+      "step": 562
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.3417898970534142,
+      "learning_rate": 4.695895230662717e-05,
+      "loss": 0.9247,
+      "step": 563
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.33570583216358385,
+      "learning_rate": 4.694629525882853e-05,
+      "loss": 0.6512,
+      "step": 564
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.43013390678989066,
+      "learning_rate": 4.693361363902872e-05,
+      "loss": 0.674,
+      "step": 565
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.3217783929134509,
+      "learning_rate": 4.692090746142661e-05,
+      "loss": 0.9221,
+      "step": 566
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.32885176489537793,
+      "learning_rate": 4.6908176740248575e-05,
+      "loss": 0.7028,
+      "step": 567
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.38141790723882935,
+      "learning_rate": 4.6895421489748476e-05,
+      "loss": 0.6386,
+      "step": 568
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.3344781107300549,
+      "learning_rate": 4.688264172420762e-05,
+      "loss": 0.6618,
+      "step": 569
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5167489738446536,
+      "learning_rate": 4.686983745793479e-05,
+      "loss": 0.9048,
+      "step": 570
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.33759821238851806,
+      "learning_rate": 4.685700870526616e-05,
+      "loss": 0.657,
+      "step": 571
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.31489325465336987,
+      "learning_rate": 4.6844155480565354e-05,
+      "loss": 0.6673,
+      "step": 572
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.33109146237199377,
+      "learning_rate": 4.68312777982234e-05,
+      "loss": 0.937,
+      "step": 573
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.3275648284959282,
+      "learning_rate": 4.681837567265869e-05,
+      "loss": 0.6707,
+      "step": 574
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.3726303500905275,
+      "learning_rate": 4.680544911831698e-05,
+      "loss": 0.6774,
+      "step": 575
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.30707913394442704,
+      "learning_rate": 4.67924981496714e-05,
+      "loss": 0.9061,
+      "step": 576
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.3516493027031166,
+      "learning_rate": 4.677952278122239e-05,
+      "loss": 0.6596,
+      "step": 577
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.36475496640511446,
+      "learning_rate": 4.6766523027497735e-05,
+      "loss": 0.6961,
+      "step": 578
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.2961623397073101,
+      "learning_rate": 4.67534989030525e-05,
+      "loss": 0.6541,
+      "step": 579
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.3348011764372994,
+      "learning_rate": 4.6740450422469054e-05,
+      "loss": 0.6364,
+      "step": 580
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.32002026867259675,
+      "learning_rate": 4.672737760035702e-05,
+      "loss": 0.9258,
+      "step": 581
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.3256635232500437,
+      "learning_rate": 4.671428045135329e-05,
+      "loss": 0.6339,
+      "step": 582
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.4311227436398567,
+      "learning_rate": 4.670115899012198e-05,
+      "loss": 0.6615,
+      "step": 583
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.29481175023536366,
+      "learning_rate": 4.668801323135443e-05,
+      "loss": 0.9068,
+      "step": 584
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.41078957106497044,
+      "learning_rate": 4.66748431897692e-05,
+      "loss": 0.659,
+      "step": 585
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.3380489099911994,
+      "learning_rate": 4.666164888011203e-05,
+      "loss": 0.6699,
+      "step": 586
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.3179794666557753,
+      "learning_rate": 4.6648430317155814e-05,
+      "loss": 0.9393,
+      "step": 587
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.35425585150773603,
+      "learning_rate": 4.663518751570061e-05,
+      "loss": 0.6226,
+      "step": 588
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.3042187811057487,
+      "learning_rate": 4.6621920490573636e-05,
+      "loss": 0.6999,
+      "step": 589
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.3157076453303041,
+      "learning_rate": 4.6608629256629196e-05,
+      "loss": 0.6468,
+      "step": 590
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.36531318819188197,
+      "learning_rate": 4.6595313828748745e-05,
+      "loss": 0.6682,
+      "step": 591
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.2889434147603217,
+      "learning_rate": 4.658197422184077e-05,
+      "loss": 0.9204,
+      "step": 592
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.32399033414040584,
+      "learning_rate": 4.656861045084088e-05,
+      "loss": 0.649,
+      "step": 593
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.327648690711374,
+      "learning_rate": 4.6555222530711716e-05,
+      "loss": 0.7045,
+      "step": 594
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.2953845437189068,
+      "learning_rate": 4.654181047644295e-05,
+      "loss": 0.642,
+      "step": 595
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.32528894756638915,
+      "learning_rate": 4.6528374303051294e-05,
+      "loss": 0.8043,
+      "step": 596
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.29661883925018595,
+      "learning_rate": 4.651491402558045e-05,
+      "loss": 0.6687,
+      "step": 597
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.31406023057137744,
+      "learning_rate": 4.6501429659101125e-05,
+      "loss": 0.9057,
+      "step": 598
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.34648344018856064,
+      "learning_rate": 4.648792121871098e-05,
+      "loss": 0.6694,
+      "step": 599
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.3369859182122131,
+      "learning_rate": 4.647438871953463e-05,
+      "loss": 0.6911,
+      "step": 600
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.3364073772249518,
+      "learning_rate": 4.646083217672365e-05,
+      "loss": 0.6522,
+      "step": 601
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.31959556275112755,
+      "learning_rate": 4.64472516054565e-05,
+      "loss": 0.8928,
+      "step": 602
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.31053111409306994,
+      "learning_rate": 4.6433647020938584e-05,
+      "loss": 0.6467,
+      "step": 603
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.34020833483180507,
+      "learning_rate": 4.642001843840215e-05,
+      "loss": 0.6332,
+      "step": 604
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.3031294862347552,
+      "learning_rate": 4.640636587310635e-05,
+      "loss": 0.7481,
+      "step": 605
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.33213720755615606,
+      "learning_rate": 4.639268934033717e-05,
+      "loss": 0.7,
+      "step": 606
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.31498546372878194,
+      "learning_rate": 4.637898885540743e-05,
+      "loss": 0.9094,
+      "step": 607
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.37686265185362844,
+      "learning_rate": 4.636526443365678e-05,
+      "loss": 0.6502,
+      "step": 608
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.35091043701323993,
+      "learning_rate": 4.6351516090451655e-05,
+      "loss": 0.6515,
+      "step": 609
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5050947157601271,
+      "learning_rate": 4.63377438411853e-05,
+      "loss": 0.8168,
+      "step": 610
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.3059293399443996,
+      "learning_rate": 4.632394770127768e-05,
+      "loss": 0.7063,
+      "step": 611
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.3016355124596301,
+      "learning_rate": 4.6310127686175566e-05,
+      "loss": 0.6545,
+      "step": 612
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.33879250418031326,
+      "learning_rate": 4.6296283811352416e-05,
+      "loss": 0.9009,
+      "step": 613
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.3012611857088822,
+      "learning_rate": 4.6282416092308424e-05,
+      "loss": 0.6602,
+      "step": 614
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.3519326932811355,
+      "learning_rate": 4.626852454457047e-05,
+      "loss": 0.6497,
+      "step": 615
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.29166985223058367,
+      "learning_rate": 4.625460918369212e-05,
+      "loss": 0.684,
+      "step": 616
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.34309614517136916,
+      "learning_rate": 4.62406700252536e-05,
+      "loss": 0.6461,
+      "step": 617
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.3307243833176125,
+      "learning_rate": 4.6226707084861786e-05,
+      "loss": 0.9435,
+      "step": 618
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.3260253961541448,
+      "learning_rate": 4.621272037815018e-05,
+      "loss": 0.6513,
+      "step": 619
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.38169717684960736,
+      "learning_rate": 4.619870992077888e-05,
+      "loss": 0.6626,
+      "step": 620
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.32455463694883074,
+      "learning_rate": 4.61846757284346e-05,
+      "loss": 0.9323,
+      "step": 621
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.3480962014148446,
+      "learning_rate": 4.6170617816830606e-05,
+      "loss": 0.6585,
+      "step": 622
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.32272972467132516,
+      "learning_rate": 4.6156536201706744e-05,
+      "loss": 0.6388,
+      "step": 623
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.3202985095637413,
+      "learning_rate": 4.614243089882937e-05,
+      "loss": 0.8924,
+      "step": 624
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.32461592987355214,
+      "learning_rate": 4.612830192399138e-05,
+      "loss": 0.6472,
+      "step": 625
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.48902372046999837,
+      "learning_rate": 4.611414929301218e-05,
+      "loss": 0.6507,
+      "step": 626
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.01538053173734,
+      "learning_rate": 4.609997302173765e-05,
+      "loss": 0.7185,
+      "step": 627
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.30645165206776165,
+      "learning_rate": 4.6085773126040154e-05,
+      "loss": 0.9037,
+      "step": 628
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.3490608945713199,
+      "learning_rate": 4.6071549621818476e-05,
+      "loss": 0.6686,
+      "step": 629
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.3289520007002725,
+      "learning_rate": 4.6057302524997874e-05,
+      "loss": 0.6763,
+      "step": 630
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7463011506014906,
+      "learning_rate": 4.604303185152999e-05,
+      "loss": 0.6572,
+      "step": 631
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.34566080742912003,
+      "learning_rate": 4.602873761739288e-05,
+      "loss": 0.9248,
+      "step": 632
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.3341367257186404,
+      "learning_rate": 4.601441983859096e-05,
+      "loss": 0.653,
+      "step": 633
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5898158013566411,
+      "learning_rate": 4.6000078531155035e-05,
+      "loss": 0.6378,
+      "step": 634
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.30343389917387803,
+      "learning_rate": 4.5985713711142243e-05,
+      "loss": 0.9305,
+      "step": 635
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.4037040488203365,
+      "learning_rate": 4.5971325394636025e-05,
+      "loss": 0.6539,
+      "step": 636
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.2768122385229475,
+      "learning_rate": 4.595691359774617e-05,
+      "loss": 0.6906,
+      "step": 637
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.3141716992644719,
+      "learning_rate": 4.594247833660873e-05,
+      "loss": 0.7076,
+      "step": 638
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.316569731123052,
+      "learning_rate": 4.592801962738602e-05,
+      "loss": 0.9055,
+      "step": 639
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.29353079630761053,
+      "learning_rate": 4.591353748626665e-05,
+      "loss": 0.6881,
+      "step": 640
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.2913916563184274,
+      "learning_rate": 4.589903192946541e-05,
+      "loss": 0.6309,
+      "step": 641
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.30635558874031305,
+      "learning_rate": 4.5884502973223356e-05,
+      "loss": 0.9149,
+      "step": 642
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.33613483933063454,
+      "learning_rate": 4.5869950633807706e-05,
+      "loss": 0.663,
+      "step": 643
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.3153264284393153,
+      "learning_rate": 4.5855374927511896e-05,
+      "loss": 0.6573,
+      "step": 644
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.29513948627674863,
+      "learning_rate": 4.5840775870655485e-05,
+      "loss": 0.8995,
+      "step": 645
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.37221478327334895,
+      "learning_rate": 4.5826153479584197e-05,
+      "loss": 0.6725,
+      "step": 646
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.36005614737511404,
+      "learning_rate": 4.581150777066989e-05,
+      "loss": 0.6447,
+      "step": 647
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.2987705043486789,
+      "learning_rate": 4.5796838760310515e-05,
+      "loss": 0.6723,
+      "step": 648
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9809921519798263,
+      "learning_rate": 4.578214646493012e-05,
+      "loss": 0.7245,
+      "step": 649
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.3143798435071705,
+      "learning_rate": 4.5767430900978816e-05,
+      "loss": 0.9388,
+      "step": 650
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.3321378175936754,
+      "learning_rate": 4.575269208493277e-05,
+      "loss": 0.6522,
+      "step": 651
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.37001797284526683,
+      "learning_rate": 4.5737930033294185e-05,
+      "loss": 0.6654,
+      "step": 652
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.35881027649132724,
+      "learning_rate": 4.572314476259129e-05,
+      "loss": 0.9414,
+      "step": 653
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.34728787464908817,
+      "learning_rate": 4.570833628937828e-05,
+      "loss": 0.6571,
+      "step": 654
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.34440061384999915,
+      "learning_rate": 4.569350463023536e-05,
+      "loss": 0.6476,
+      "step": 655
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.29279584572107575,
+      "learning_rate": 4.567864980176869e-05,
+      "loss": 0.935,
+      "step": 656
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.4012078729105803,
+      "learning_rate": 4.566377182061035e-05,
+      "loss": 0.6318,
+      "step": 657
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.3192894718371079,
+      "learning_rate": 4.564887070341837e-05,
+      "loss": 0.6457,
+      "step": 658
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.4096636409008902,
+      "learning_rate": 4.563394646687666e-05,
+      "loss": 0.8848,
+      "step": 659
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.4712840233484657,
+      "learning_rate": 4.561899912769505e-05,
+      "loss": 0.4719,
+      "step": 660
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.307205116776837,
+      "learning_rate": 4.5604028702609184e-05,
+      "loss": 0.9232,
+      "step": 661
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.34023789310338737,
+      "learning_rate": 4.55890352083806e-05,
+      "loss": 0.6551,
+      "step": 662
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.36553689034130127,
+      "learning_rate": 4.5574018661796644e-05,
+      "loss": 0.6682,
+      "step": 663
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.2974953068094857,
+      "learning_rate": 4.555897907967048e-05,
+      "loss": 0.9301,
+      "step": 664
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.4376611330672298,
+      "learning_rate": 4.554391647884106e-05,
+      "loss": 0.658,
+      "step": 665
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.3063130777731919,
+      "learning_rate": 4.552883087617311e-05,
+      "loss": 0.6522,
+      "step": 666
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.2970964580146152,
+      "learning_rate": 4.5513722288557106e-05,
+      "loss": 0.9137,
+      "step": 667
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.30860868356281057,
+      "learning_rate": 4.549859073290926e-05,
+      "loss": 0.6416,
+      "step": 668
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.32139863878580943,
+      "learning_rate": 4.5483436226171516e-05,
+      "loss": 0.6428,
+      "step": 669
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.2898056047494988,
+      "learning_rate": 4.546825878531149e-05,
+      "loss": 0.9218,
+      "step": 670
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.3274884821502218,
+      "learning_rate": 4.545305842732249e-05,
+      "loss": 0.6967,
+      "step": 671
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.30852185196493015,
+      "learning_rate": 4.543783516922348e-05,
+      "loss": 0.6715,
+      "step": 672
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.2854307495437414,
+      "learning_rate": 4.542258902805906e-05,
+      "loss": 0.6579,
+      "step": 673
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.32059210463319526,
+      "learning_rate": 4.540732002089947e-05,
+      "loss": 0.6647,
+      "step": 674
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.29861537327235294,
+      "learning_rate": 4.539202816484052e-05,
+      "loss": 0.9219,
+      "step": 675
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.32718437937597916,
+      "learning_rate": 4.537671347700362e-05,
+      "loss": 0.6448,
+      "step": 676
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.3166932309935094,
+      "learning_rate": 4.536137597453575e-05,
+      "loss": 0.6728,
+      "step": 677
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.2725269590215528,
+      "learning_rate": 4.534601567460942e-05,
+      "loss": 0.9317,
+      "step": 678
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.30968667804741445,
+      "learning_rate": 4.533063259442268e-05,
+      "loss": 0.6412,
+      "step": 679
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.3018409416866063,
+      "learning_rate": 4.531522675119908e-05,
+      "loss": 0.682,
+      "step": 680
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.2884255423865988,
+      "learning_rate": 4.529979816218765e-05,
+      "loss": 0.9502,
+      "step": 681
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.33396847415896047,
+      "learning_rate": 4.528434684466289e-05,
+      "loss": 0.6851,
+      "step": 682
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.3107059790216546,
+      "learning_rate": 4.526887281592475e-05,
+      "loss": 0.6648,
+      "step": 683
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.32745062166833383,
+      "learning_rate": 4.52533760932986e-05,
+      "loss": 0.6353,
+      "step": 684
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.3147935222568101,
+      "learning_rate": 4.5237856694135236e-05,
+      "loss": 0.9315,
+      "step": 685
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.3456677540936494,
+      "learning_rate": 4.5222314635810836e-05,
+      "loss": 0.6233,
+      "step": 686
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.3603705816765344,
+      "learning_rate": 4.5206749935726935e-05,
+      "loss": 0.6476,
+      "step": 687
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.2839353405873735,
+      "learning_rate": 4.5191162611310445e-05,
+      "loss": 0.9346,
+      "step": 688
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.3543412675254845,
+      "learning_rate": 4.5175552680013585e-05,
+      "loss": 0.6552,
+      "step": 689
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.3032500071467932,
+      "learning_rate": 4.5159920159313896e-05,
+      "loss": 0.6549,
+      "step": 690
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.2979705512056598,
+      "learning_rate": 4.514426506671421e-05,
+      "loss": 0.6408,
+      "step": 691
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.29844547645749,
+      "learning_rate": 4.512858741974263e-05,
+      "loss": 0.9084,
+      "step": 692
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.30217101999341345,
+      "learning_rate": 4.511288723595252e-05,
+      "loss": 0.6893,
+      "step": 693
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.3824270002226234,
+      "learning_rate": 4.509716453292246e-05,
+      "loss": 0.6514,
+      "step": 694
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.3048980830132978,
+      "learning_rate": 4.508141932825626e-05,
+      "loss": 0.6126,
+      "step": 695
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.279576685170488,
+      "learning_rate": 4.506565163958292e-05,
+      "loss": 0.8939,
+      "step": 696
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.30792775298299013,
+      "learning_rate": 4.5049861484556606e-05,
+      "loss": 0.641,
+      "step": 697
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.2797563094395619,
+      "learning_rate": 4.503404888085665e-05,
+      "loss": 0.6476,
+      "step": 698
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.2873342640141601,
+      "learning_rate": 4.5018213846187504e-05,
+      "loss": 0.9173,
+      "step": 699
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.278862314501891,
+      "learning_rate": 4.500235639827874e-05,
+      "loss": 0.661,
+      "step": 700
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.2813828622445945,
+      "learning_rate": 4.498647655488505e-05,
+      "loss": 0.652,
+      "step": 701
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.27919906116595744,
+      "learning_rate": 4.497057433378616e-05,
+      "loss": 0.8883,
+      "step": 702
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.28744133238549324,
+      "learning_rate": 4.495464975278688e-05,
+      "loss": 0.6269,
+      "step": 703
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.2896757269125007,
+      "learning_rate": 4.493870282971704e-05,
+      "loss": 0.7153,
+      "step": 704
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.28452408608253865,
+      "learning_rate": 4.492273358243149e-05,
+      "loss": 0.6363,
+      "step": 705
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.27832274874399776,
+      "learning_rate": 4.490674202881007e-05,
+      "loss": 0.6379,
+      "step": 706
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.2988520487549258,
+      "learning_rate": 4.48907281867576e-05,
+      "loss": 0.9249,
+      "step": 707
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.2831753446414981,
+      "learning_rate": 4.487469207420387e-05,
+      "loss": 0.647,
+      "step": 708
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.3283928934215106,
+      "learning_rate": 4.485863370910358e-05,
+      "loss": 0.6433,
+      "step": 709
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.32709040540926027,
+      "learning_rate": 4.484255310943636e-05,
+      "loss": 0.9146,
+      "step": 710
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.3262050606713422,
+      "learning_rate": 4.482645029320672e-05,
+      "loss": 0.6307,
+      "step": 711
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.39598315623359825,
+      "learning_rate": 4.481032527844407e-05,
+      "loss": 0.6526,
+      "step": 712
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.28462225534427327,
+      "learning_rate": 4.479417808320264e-05,
+      "loss": 0.9035,
+      "step": 713
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.382218160611698,
+      "learning_rate": 4.477800872556154e-05,
+      "loss": 0.6546,
+      "step": 714
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.31075841101697466,
+      "learning_rate": 4.4761817223624635e-05,
+      "loss": 0.7001,
+      "step": 715
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.37090572745822253,
+      "learning_rate": 4.4745603595520644e-05,
+      "loss": 0.6446,
+      "step": 716
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.3955117775065539,
+      "learning_rate": 4.472936785940302e-05,
+      "loss": 0.6424,
+      "step": 717
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.31391056086721913,
+      "learning_rate": 4.4713110033449966e-05,
+      "loss": 0.911,
+      "step": 718
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.3242028110838875,
+      "learning_rate": 4.4696830135864455e-05,
+      "loss": 0.6352,
+      "step": 719
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.33196453224614814,
+      "learning_rate": 4.468052818487414e-05,
+      "loss": 0.7391,
+      "step": 720
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.3408294520954801,
+      "learning_rate": 4.4664204198731374e-05,
+      "loss": 0.6557,
+      "step": 721
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.3027198738509431,
+      "learning_rate": 4.4647858195713185e-05,
+      "loss": 0.8995,
+      "step": 722
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.3416791409270043,
+      "learning_rate": 4.463149019412126e-05,
+      "loss": 0.6306,
+      "step": 723
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.3188277027831064,
+      "learning_rate": 4.461510021228189e-05,
+      "loss": 0.9314,
+      "step": 724
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.35770850910477875,
+      "learning_rate": 4.459868826854599e-05,
+      "loss": 0.6424,
+      "step": 725
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.3164638386826528,
+      "learning_rate": 4.458225438128909e-05,
+      "loss": 0.7233,
+      "step": 726
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6534510884991952,
+      "learning_rate": 4.4565798568911245e-05,
+      "loss": 0.6484,
+      "step": 727
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.3155338537032338,
+      "learning_rate": 4.4549320849837086e-05,
+      "loss": 0.9039,
+      "step": 728
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.33398210175212023,
+      "learning_rate": 4.4532821242515764e-05,
+      "loss": 0.6443,
+      "step": 729
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.3240078836058026,
+      "learning_rate": 4.451629976542094e-05,
+      "loss": 0.6477,
+      "step": 730
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.31457755671361143,
+      "learning_rate": 4.4499756437050756e-05,
+      "loss": 0.6942,
+      "step": 731
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.35033180289475785,
+      "learning_rate": 4.448319127592782e-05,
+      "loss": 0.649,
+      "step": 732
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.2977218616946143,
+      "learning_rate": 4.44666043005992e-05,
+      "loss": 0.9001,
+      "step": 733
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.35183561123523716,
+      "learning_rate": 4.444999552963636e-05,
+      "loss": 0.6499,
+      "step": 734
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.3518815842754916,
+      "learning_rate": 4.443336498163519e-05,
+      "loss": 0.6467,
+      "step": 735
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.30086135546287707,
+      "learning_rate": 4.4416712675215964e-05,
+      "loss": 0.8947,
+      "step": 736
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.38676053760598983,
+      "learning_rate": 4.44000386290233e-05,
+      "loss": 0.6479,
+      "step": 737
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.44330719558000614,
+      "learning_rate": 4.4383342861726164e-05,
+      "loss": 0.6689,
+      "step": 738
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.3164863683157211,
+      "learning_rate": 4.4366625392017844e-05,
+      "loss": 0.9202,
+      "step": 739
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.3096574610507007,
+      "learning_rate": 4.434988623861593e-05,
+      "loss": 0.6639,
+      "step": 740
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.3157270994171228,
+      "learning_rate": 4.43331254202623e-05,
+      "loss": 0.6408,
+      "step": 741
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.31827600904007775,
+      "learning_rate": 4.431634295572305e-05,
+      "loss": 0.7175,
+      "step": 742
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.30971968670079375,
+      "learning_rate": 4.429953886378855e-05,
+      "loss": 0.6384,
+      "step": 743
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.35128248723050753,
+      "learning_rate": 4.4282713163273384e-05,
+      "loss": 0.9039,
+      "step": 744
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.33468162573923793,
+      "learning_rate": 4.426586587301629e-05,
+      "loss": 0.6346,
+      "step": 745
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.3035527780645913,
+      "learning_rate": 4.4248997011880235e-05,
+      "loss": 0.6578,
+      "step": 746
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.28692093614621295,
+      "learning_rate": 4.42321065987523e-05,
+      "loss": 0.9107,
+      "step": 747
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.3591240496702593,
+      "learning_rate": 4.42151946525437e-05,
+      "loss": 0.6474,
+      "step": 748
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.26820878856585756,
+      "learning_rate": 4.419826119218977e-05,
+      "loss": 0.6395,
+      "step": 749
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.2942573130999204,
+      "learning_rate": 4.418130623664994e-05,
+      "loss": 0.8959,
+      "step": 750
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.3587087849393897,
+      "learning_rate": 4.416432980490768e-05,
+      "loss": 0.6543,
+      "step": 751
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.3040216671216807,
+      "learning_rate": 4.414733191597053e-05,
+      "loss": 0.6378,
+      "step": 752
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.3216285937899872,
+      "learning_rate": 4.413031258887005e-05,
+      "loss": 0.7029,
+      "step": 753
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.3049854615720848,
+      "learning_rate": 4.411327184266179e-05,
+      "loss": 0.9385,
+      "step": 754
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.3139787813727069,
+      "learning_rate": 4.409620969642529e-05,
+      "loss": 0.6557,
+      "step": 755
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.28929506651649756,
+      "learning_rate": 4.4079126169264065e-05,
+      "loss": 0.659,
+      "step": 756
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.29889234058089903,
+      "learning_rate": 4.4062021280305533e-05,
+      "loss": 0.9167,
+      "step": 757
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.49058067897223523,
+      "learning_rate": 4.4044895048701076e-05,
+      "loss": 0.6888,
+      "step": 758
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.3069612228014401,
+      "learning_rate": 4.4027747493625927e-05,
+      "loss": 0.6271,
+      "step": 759
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.2857612071589618,
+      "learning_rate": 4.4010578634279225e-05,
+      "loss": 0.6345,
+      "step": 760
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.31270075716164014,
+      "learning_rate": 4.399338848988395e-05,
+      "loss": 0.9123,
+      "step": 761
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.29355436285821274,
+      "learning_rate": 4.397617707968692e-05,
+      "loss": 0.6517,
+      "step": 762
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.31199407034527804,
+      "learning_rate": 4.395894442295876e-05,
+      "loss": 0.6775,
+      "step": 763
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.3037629354769261,
+      "learning_rate": 4.3941690538993864e-05,
+      "loss": 0.6976,
+      "step": 764
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.29305447138299306,
+      "learning_rate": 4.392441544711043e-05,
+      "loss": 0.8616,
+      "step": 765
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.2818773216872479,
+      "learning_rate": 4.390711916665037e-05,
+      "loss": 0.6607,
+      "step": 766
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.2896024796818521,
+      "learning_rate": 4.3889801716979336e-05,
+      "loss": 0.6525,
+      "step": 767
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.28433515880163146,
+      "learning_rate": 4.387246311748669e-05,
+      "loss": 0.8986,
+      "step": 768
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.32828393565752334,
+      "learning_rate": 4.3855103387585446e-05,
+      "loss": 0.6877,
+      "step": 769
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.3237286803262996,
+      "learning_rate": 4.383772254671229e-05,
+      "loss": 0.6571,
+      "step": 770
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.28751627453135453,
+      "learning_rate": 4.382032061432756e-05,
+      "loss": 0.9213,
+      "step": 771
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.3071881844879142,
+      "learning_rate": 4.3802897609915187e-05,
+      "loss": 0.6446,
+      "step": 772
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.33115113773842403,
+      "learning_rate": 4.3785453552982716e-05,
+      "loss": 0.6269,
+      "step": 773
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.3073648768904819,
+      "learning_rate": 4.3767988463061236e-05,
+      "loss": 0.6226,
+      "step": 774
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.30229432615624036,
+      "learning_rate": 4.375050235970541e-05,
+      "loss": 0.6833,
+      "step": 775
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.28294002219439357,
+      "learning_rate": 4.373299526249341e-05,
+      "loss": 0.9122,
+      "step": 776
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.36555295785431086,
+      "learning_rate": 4.3715467191026944e-05,
+      "loss": 0.6508,
+      "step": 777
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.2771607997383557,
+      "learning_rate": 4.369791816493116e-05,
+      "loss": 0.6466,
+      "step": 778
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.29973503034360066,
+      "learning_rate": 4.36803482038547e-05,
+      "loss": 0.9096,
+      "step": 779
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.28953286723177346,
+      "learning_rate": 4.366275732746965e-05,
+      "loss": 0.6669,
+      "step": 780
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.3165549067708345,
+      "learning_rate": 4.364514555547147e-05,
+      "loss": 0.6451,
+      "step": 781
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.2768199920043759,
+      "learning_rate": 4.3627512907579074e-05,
+      "loss": 0.8975,
+      "step": 782
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.3143692427833259,
+      "learning_rate": 4.36098594035347e-05,
+      "loss": 0.6664,
+      "step": 783
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.27616089474502936,
+      "learning_rate": 4.3592185063103966e-05,
+      "loss": 0.6228,
+      "step": 784
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.28990912398388813,
+      "learning_rate": 4.3574489906075824e-05,
+      "loss": 0.9115,
+      "step": 785
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.2886873984831601,
+      "learning_rate": 4.3556773952262495e-05,
+      "loss": 0.4488,
+      "step": 786
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.2662379863616495,
+      "learning_rate": 4.3539037221499535e-05,
+      "loss": 0.8943,
+      "step": 787
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.2827536368849426,
+      "learning_rate": 4.3521279733645724e-05,
+      "loss": 0.654,
+      "step": 788
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.2813937139926161,
+      "learning_rate": 4.350350150858311e-05,
+      "loss": 0.646,
+      "step": 789
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.266218278510744,
+      "learning_rate": 4.3485702566216936e-05,
+      "loss": 0.9032,
+      "step": 790
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.3033756771955927,
+      "learning_rate": 4.346788292647565e-05,
+      "loss": 0.654,
+      "step": 791
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.2745915195704939,
+      "learning_rate": 4.3450042609310874e-05,
+      "loss": 0.6328,
+      "step": 792
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.27457333077892904,
+      "learning_rate": 4.343218163469739e-05,
+      "loss": 0.8959,
+      "step": 793
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.2912126320140852,
+      "learning_rate": 4.34143000226331e-05,
+      "loss": 0.6243,
+      "step": 794
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.2716940139847096,
+      "learning_rate": 4.3396397793139014e-05,
+      "loss": 0.626,
+      "step": 795
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.2856018306717644,
+      "learning_rate": 4.337847496625922e-05,
+      "loss": 0.915,
+      "step": 796
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.3116113812866413,
+      "learning_rate": 4.3360531562060887e-05,
+      "loss": 0.7095,
+      "step": 797
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.28477603527665063,
+      "learning_rate": 4.334256760063419e-05,
+      "loss": 0.669,
+      "step": 798
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.27247895156340146,
+      "learning_rate": 4.3324583102092355e-05,
+      "loss": 0.6285,
+      "step": 799
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.28966245215194303,
+      "learning_rate": 4.330657808657158e-05,
+      "loss": 0.6402,
+      "step": 800
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.28334842862638054,
+      "learning_rate": 4.328855257423106e-05,
+      "loss": 0.9458,
+      "step": 801
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.28481888298497565,
+      "learning_rate": 4.327050658525289e-05,
+      "loss": 0.629,
+      "step": 802
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.3116094000639633,
+      "learning_rate": 4.3252440139842153e-05,
+      "loss": 0.6513,
+      "step": 803
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.2770231642566091,
+      "learning_rate": 4.323435325822679e-05,
+      "loss": 0.8954,
+      "step": 804
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.32262707238918076,
+      "learning_rate": 4.3216245960657644e-05,
+      "loss": 0.6286,
+      "step": 805
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.2925298211029747,
+      "learning_rate": 4.3198118267408415e-05,
+      "loss": 0.6436,
+      "step": 806
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.2744375017134578,
+      "learning_rate": 4.3179970198775635e-05,
+      "loss": 0.9056,
+      "step": 807
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.31568568031294175,
+      "learning_rate": 4.3161801775078644e-05,
+      "loss": 0.7204,
+      "step": 808
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.2697671068716749,
+      "learning_rate": 4.314361301665958e-05,
+      "loss": 0.6277,
+      "step": 809
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.2771970006539298,
+      "learning_rate": 4.3125403943883345e-05,
+      "loss": 0.6309,
+      "step": 810
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.3173417575979328,
+      "learning_rate": 4.3107174577137596e-05,
+      "loss": 0.9081,
+      "step": 811
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.27074247293814335,
+      "learning_rate": 4.3088924936832696e-05,
+      "loss": 0.639,
+      "step": 812
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.3062360841613198,
+      "learning_rate": 4.307065504340172e-05,
+      "loss": 0.6346,
+      "step": 813
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.293171767403898,
+      "learning_rate": 4.305236491730041e-05,
+      "loss": 0.9093,
+      "step": 814
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.2913984411633076,
+      "learning_rate": 4.303405457900717e-05,
+      "loss": 0.6294,
+      "step": 815
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.3015665523788799,
+      "learning_rate": 4.301572404902302e-05,
+      "loss": 0.6533,
+      "step": 816
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.2753762506330584,
+      "learning_rate": 4.299737334787161e-05,
+      "loss": 0.6349,
+      "step": 817
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.39341984010903425,
+      "learning_rate": 4.297900249609915e-05,
+      "loss": 0.8895,
+      "step": 818
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.3028264926068928,
+      "learning_rate": 4.296061151427444e-05,
+      "loss": 0.692,
+      "step": 819
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.2613296309229208,
+      "learning_rate": 4.294220042298878e-05,
+      "loss": 0.6555,
+      "step": 820
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.2789031645813559,
+      "learning_rate": 4.292376924285603e-05,
+      "loss": 0.626,
+      "step": 821
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.2670138274709757,
+      "learning_rate": 4.2905317994512494e-05,
+      "loss": 0.9118,
+      "step": 822
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.2881585908346101,
+      "learning_rate": 4.288684669861699e-05,
+      "loss": 0.6496,
+      "step": 823
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.2701369463903409,
+      "learning_rate": 4.286835537585076e-05,
+      "loss": 0.6417,
+      "step": 824
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.29244381270033754,
+      "learning_rate": 4.2849844046917465e-05,
+      "loss": 0.913,
+      "step": 825
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.2877172085383411,
+      "learning_rate": 4.283131273254318e-05,
+      "loss": 0.6421,
+      "step": 826
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.2777100694228787,
+      "learning_rate": 4.2812761453476345e-05,
+      "loss": 0.6456,
+      "step": 827
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.26780844896318323,
+      "learning_rate": 4.279419023048775e-05,
+      "loss": 0.8916,
+      "step": 828
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.3178671287833813,
+      "learning_rate": 4.277559908437054e-05,
+      "loss": 0.6194,
+      "step": 829
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.2990230586453986,
+      "learning_rate": 4.2756988035940134e-05,
+      "loss": 0.711,
+      "step": 830
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.2934188808856755,
+      "learning_rate": 4.2738357106034254e-05,
+      "loss": 0.6312,
+      "step": 831
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.31560944367814736,
+      "learning_rate": 4.271970631551289e-05,
+      "loss": 0.612,
+      "step": 832
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.27774745732844874,
+      "learning_rate": 4.2701035685258225e-05,
+      "loss": 0.8887,
+      "step": 833
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.3215264855986915,
+      "learning_rate": 4.2682345236174727e-05,
+      "loss": 0.6453,
+      "step": 834
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.3055033620225006,
+      "learning_rate": 4.2663634989188985e-05,
+      "loss": 0.6501,
+      "step": 835
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.31508172510387933,
+      "learning_rate": 4.264490496524979e-05,
+      "loss": 0.6946,
+      "step": 836
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.31357216780701624,
+      "learning_rate": 4.262615518532808e-05,
+      "loss": 0.7801,
+      "step": 837
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.2946109380430888,
+      "learning_rate": 4.260738567041689e-05,
+      "loss": 0.6241,
+      "step": 838
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.29988748387966085,
+      "learning_rate": 4.258859644153137e-05,
+      "loss": 0.8855,
+      "step": 839
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.29249889022574044,
+      "learning_rate": 4.256978751970874e-05,
+      "loss": 0.6412,
+      "step": 840
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.2770050600336557,
+      "learning_rate": 4.2550958926008244e-05,
+      "loss": 0.7164,
+      "step": 841
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.298311016429622,
+      "learning_rate": 4.2532110681511196e-05,
+      "loss": 0.6183,
+      "step": 842
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.26096309171140347,
+      "learning_rate": 4.251324280732088e-05,
+      "loss": 0.619,
+      "step": 843
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.27155774678655853,
+      "learning_rate": 4.249435532456257e-05,
+      "loss": 0.9032,
+      "step": 844
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.27688686532562384,
+      "learning_rate": 4.2475448254383484e-05,
+      "loss": 0.6398,
+      "step": 845
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.2723006542901508,
+      "learning_rate": 4.2456521617952785e-05,
+      "loss": 0.6412,
+      "step": 846
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.2815844589511535,
+      "learning_rate": 4.243757543646153e-05,
+      "loss": 0.6927,
+      "step": 847
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.28779054334493304,
+      "learning_rate": 4.2418609731122674e-05,
+      "loss": 0.8987,
+      "step": 848
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.27291445608685627,
+      "learning_rate": 4.239962452317101e-05,
+      "loss": 0.6228,
+      "step": 849
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.2607000170802453,
+      "learning_rate": 4.23806198338632e-05,
+      "loss": 0.6368,
+      "step": 850
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.267578532878797,
+      "learning_rate": 4.2361595684477686e-05,
+      "loss": 0.8171,
+      "step": 851
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.2766150664801454,
+      "learning_rate": 4.2342552096314706e-05,
+      "loss": 0.6969,
+      "step": 852
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.2635606262077664,
+      "learning_rate": 4.2323489090696275e-05,
+      "loss": 0.6306,
+      "step": 853
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.34059443252787175,
+      "learning_rate": 4.230440668896615e-05,
+      "loss": 0.8759,
+      "step": 854
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.27838172231333663,
+      "learning_rate": 4.2285304912489774e-05,
+      "loss": 0.6242,
+      "step": 855
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.2852531967111486,
+      "learning_rate": 4.226618378265432e-05,
+      "loss": 0.6528,
+      "step": 856
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.28695504017087786,
+      "learning_rate": 4.224704332086862e-05,
+      "loss": 0.9,
+      "step": 857
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.3265947492644721,
+      "learning_rate": 4.222788354856313e-05,
+      "loss": 0.4526,
+      "step": 858
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.29414938340525787,
+      "learning_rate": 4.2208704487189945e-05,
+      "loss": 0.8996,
+      "step": 859
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.32393524858943384,
+      "learning_rate": 4.2189506158222756e-05,
+      "loss": 0.6244,
+      "step": 860
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.30681611689948673,
+      "learning_rate": 4.217028858315683e-05,
+      "loss": 0.6116,
+      "step": 861
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.28417195702258724,
+      "learning_rate": 4.2151051783508976e-05,
+      "loss": 0.9112,
+      "step": 862
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.2998235057644029,
+      "learning_rate": 4.213179578081753e-05,
+      "loss": 0.6247,
+      "step": 863
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.29463186980636585,
+      "learning_rate": 4.211252059664232e-05,
+      "loss": 0.6179,
+      "step": 864
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.2811721411012102,
+      "learning_rate": 4.209322625256466e-05,
+      "loss": 0.9614,
+      "step": 865
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.3012768150774497,
+      "learning_rate": 4.2073912770187316e-05,
+      "loss": 0.6577,
+      "step": 866
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.28563526175734566,
+      "learning_rate": 4.20545801711345e-05,
+      "loss": 0.6296,
+      "step": 867
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.31816640608614744,
+      "learning_rate": 4.203522847705177e-05,
+      "loss": 0.8721,
+      "step": 868
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.2934663580926699,
+      "learning_rate": 4.201585770960612e-05,
+      "loss": 0.4424,
+      "step": 869
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.302915272499066,
+      "learning_rate": 4.199646789048589e-05,
+      "loss": 0.9491,
+      "step": 870
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.3233878042363829,
+      "learning_rate": 4.197705904140074e-05,
+      "loss": 0.6273,
+      "step": 871
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.3752437679838047,
+      "learning_rate": 4.195763118408162e-05,
+      "loss": 0.6465,
+      "step": 872
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.28002464108566355,
+      "learning_rate": 4.193818434028081e-05,
+      "loss": 0.887,
+      "step": 873
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.3396522589226574,
+      "learning_rate": 4.19187185317718e-05,
+      "loss": 0.6535,
+      "step": 874
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.32297902576707516,
+      "learning_rate": 4.189923378034935e-05,
+      "loss": 0.6337,
+      "step": 875
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.29729774941077053,
+      "learning_rate": 4.187973010782938e-05,
+      "loss": 0.8846,
+      "step": 876
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.3367309642345147,
+      "learning_rate": 4.186020753604906e-05,
+      "loss": 0.6342,
+      "step": 877
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.26608697573237233,
+      "learning_rate": 4.184066608686668e-05,
+      "loss": 0.6195,
+      "step": 878
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.2945362807268212,
+      "learning_rate": 4.1821105782161654e-05,
+      "loss": 0.9228,
+      "step": 879
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.2935183594054431,
+      "learning_rate": 4.1801526643834555e-05,
+      "loss": 0.7154,
+      "step": 880
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.2709913950874,
+      "learning_rate": 4.1781928693807e-05,
+      "loss": 0.6392,
+      "step": 881
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.30218767012083375,
+      "learning_rate": 4.176231195402169e-05,
+      "loss": 0.621,
+      "step": 882
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.28784550406392484,
+      "learning_rate": 4.174267644644235e-05,
+      "loss": 0.9062,
+      "step": 883
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.3013171497762678,
+      "learning_rate": 4.172302219305373e-05,
+      "loss": 0.6454,
+      "step": 884
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.2959060338756323,
+      "learning_rate": 4.170334921586157e-05,
+      "loss": 0.6395,
+      "step": 885
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.2846225104713399,
+      "learning_rate": 4.168365753689255e-05,
+      "loss": 0.6075,
+      "step": 886
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.277165989873569,
+      "learning_rate": 4.166394717819432e-05,
+      "loss": 0.9143,
+      "step": 887
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.3061407286775032,
+      "learning_rate": 4.1644218161835436e-05,
+      "loss": 0.6357,
+      "step": 888
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.29181362836139924,
+      "learning_rate": 4.1624470509905336e-05,
+      "loss": 0.6318,
+      "step": 889
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.28001982758351734,
+      "learning_rate": 4.160470424451431e-05,
+      "loss": 0.9521,
+      "step": 890
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.30585584065613275,
+      "learning_rate": 4.1584919387793544e-05,
+      "loss": 0.6732,
+      "step": 891
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.26966291222695804,
+      "learning_rate": 4.1565115961894966e-05,
+      "loss": 0.6103,
+      "step": 892
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.3003464047988866,
+      "learning_rate": 4.1545293988991344e-05,
+      "loss": 0.6497,
+      "step": 893
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.2793032971425848,
+      "learning_rate": 4.152545349127619e-05,
+      "loss": 0.8707,
+      "step": 894
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.3182596245521193,
+      "learning_rate": 4.150559449096377e-05,
+      "loss": 0.6707,
+      "step": 895
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.31860952042091956,
+      "learning_rate": 4.148571701028907e-05,
+      "loss": 0.6435,
+      "step": 896
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.29344954455898564,
+      "learning_rate": 4.146582107150775e-05,
+      "loss": 0.8823,
+      "step": 897
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.2930067650545813,
+      "learning_rate": 4.144590669689614e-05,
+      "loss": 0.6542,
+      "step": 898
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.32253651066036054,
+      "learning_rate": 4.1425973908751236e-05,
+      "loss": 0.6325,
+      "step": 899
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.27366463126445695,
+      "learning_rate": 4.140602272939061e-05,
+      "loss": 0.8946,
+      "step": 900
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.32394356480243836,
+      "learning_rate": 4.138605318115246e-05,
+      "loss": 0.6328,
+      "step": 901
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.2935931936574356,
+      "learning_rate": 4.136606528639553e-05,
+      "loss": 0.6761,
+      "step": 902
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.29697703216788623,
+      "learning_rate": 4.1346059067499115e-05,
+      "loss": 0.6244,
+      "step": 903
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.29923428007364516,
+      "learning_rate": 4.132603454686303e-05,
+      "loss": 0.6237,
+      "step": 904
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.27923801773591134,
+      "learning_rate": 4.130599174690757e-05,
+      "loss": 0.8918,
+      "step": 905
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.3092738862965256,
+      "learning_rate": 4.1285930690073506e-05,
+      "loss": 0.6607,
+      "step": 906
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.2940958410194855,
+      "learning_rate": 4.1265851398822034e-05,
+      "loss": 0.6361,
+      "step": 907
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.26633418330304687,
+      "learning_rate": 4.12457538956348e-05,
+      "loss": 0.8954,
+      "step": 908
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.3099666778524849,
+      "learning_rate": 4.1225638203013805e-05,
+      "loss": 0.6263,
+      "step": 909
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.2690849153347127,
+      "learning_rate": 4.120550434348143e-05,
+      "loss": 0.6024,
+      "step": 910
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.28812232075502275,
+      "learning_rate": 4.1185352339580404e-05,
+      "loss": 0.8967,
+      "step": 911
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.3118759106403073,
+      "learning_rate": 4.116518221387375e-05,
+      "loss": 0.619,
+      "step": 912
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.30091884247605605,
+      "learning_rate": 4.1144993988944814e-05,
+      "loss": 0.6973,
+      "step": 913
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.3156007286804202,
+      "learning_rate": 4.1124787687397175e-05,
+      "loss": 0.6289,
+      "step": 914
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.2804046728732873,
+      "learning_rate": 4.1104563331854666e-05,
+      "loss": 0.6322,
+      "step": 915
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.27134321244443327,
+      "learning_rate": 4.108432094496133e-05,
+      "loss": 0.8968,
+      "step": 916
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.28309218340123,
+      "learning_rate": 4.10640605493814e-05,
+      "loss": 0.6625,
+      "step": 917
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.30825631741143183,
+      "learning_rate": 4.104378216779927e-05,
+      "loss": 0.6412,
+      "step": 918
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.27618640458509175,
+      "learning_rate": 4.102348582291949e-05,
+      "loss": 0.9052,
+      "step": 919
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.31184717029531034,
+      "learning_rate": 4.1003171537466686e-05,
+      "loss": 0.6238,
+      "step": 920
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.2906385356683692,
+      "learning_rate": 4.098283933418561e-05,
+      "loss": 0.6287,
+      "step": 921
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.2643828388194801,
+      "learning_rate": 4.096248923584104e-05,
+      "loss": 0.8773,
+      "step": 922
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.3006896311461778,
+      "learning_rate": 4.094212126521781e-05,
+      "loss": 0.6121,
+      "step": 923
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.282209425287015,
+      "learning_rate": 4.092173544512078e-05,
+      "loss": 0.6741,
+      "step": 924
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.3333524166420371,
+      "learning_rate": 4.0901331798374766e-05,
+      "loss": 0.6234,
+      "step": 925
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.29196997888138543,
+      "learning_rate": 4.088091034782455e-05,
+      "loss": 0.9058,
+      "step": 926
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.32936250748340906,
+      "learning_rate": 4.086047111633485e-05,
+      "loss": 0.6396,
+      "step": 927
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.32613380756427646,
+      "learning_rate": 4.08400141267903e-05,
+      "loss": 0.6415,
+      "step": 928
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.280842320473299,
+      "learning_rate": 4.081953940209543e-05,
+      "loss": 0.6264,
+      "step": 929
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.29055710520755074,
+      "learning_rate": 4.079904696517458e-05,
+      "loss": 0.918,
+      "step": 930
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.30052536726598034,
+      "learning_rate": 4.0778536838971966e-05,
+      "loss": 0.6205,
+      "step": 931
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.3178220760892087,
+      "learning_rate": 4.075800904645158e-05,
+      "loss": 0.6545,
+      "step": 932
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.2814829452215227,
+      "learning_rate": 4.0737463610597225e-05,
+      "loss": 0.8762,
+      "step": 933
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.36619782643125975,
+      "learning_rate": 4.071690055441243e-05,
+      "loss": 0.6472,
+      "step": 934
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.29527824791107543,
+      "learning_rate": 4.069631990092046e-05,
+      "loss": 0.6751,
+      "step": 935
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.34721951641651516,
+      "learning_rate": 4.067572167316429e-05,
+      "loss": 0.6386,
+      "step": 936
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.33250687865719997,
+      "learning_rate": 4.065510589420657e-05,
+      "loss": 0.9204,
+      "step": 937
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.3765632124280686,
+      "learning_rate": 4.0634472587129604e-05,
+      "loss": 0.6401,
+      "step": 938
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.34976981436846105,
+      "learning_rate": 4.06138217750353e-05,
+      "loss": 0.6314,
+      "step": 939
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.32480660597342725,
+      "learning_rate": 4.05931534810452e-05,
+      "loss": 0.9067,
+      "step": 940
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.38186569881302285,
+      "learning_rate": 4.0572467728300376e-05,
+      "loss": 0.6583,
+      "step": 941
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.28834562721926477,
+      "learning_rate": 4.05517645399615e-05,
+      "loss": 0.6131,
+      "step": 942
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.288732328420397,
+      "learning_rate": 4.0531043939208726e-05,
+      "loss": 0.9267,
+      "step": 943
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.3016930351521152,
+      "learning_rate": 4.0510305949241714e-05,
+      "loss": 0.6239,
+      "step": 944
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.27986802031942387,
+      "learning_rate": 4.04895505932796e-05,
+      "loss": 0.6482,
+      "step": 945
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.28955065643121264,
+      "learning_rate": 4.046877789456095e-05,
+      "loss": 0.7131,
+      "step": 946
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.31847310232136994,
+      "learning_rate": 4.0447987876343764e-05,
+      "loss": 0.6188,
+      "step": 947
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.2920082305212223,
+      "learning_rate": 4.042718056190543e-05,
+      "loss": 0.9,
+      "step": 948
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.3200494962372411,
+      "learning_rate": 4.040635597454269e-05,
+      "loss": 0.6003,
+      "step": 949
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.2927645721072571,
+      "learning_rate": 4.0385514137571635e-05,
+      "loss": 0.6336,
+      "step": 950
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.28499617061396115,
+      "learning_rate": 4.036465507432767e-05,
+      "loss": 0.8911,
+      "step": 951
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.30855011772074536,
+      "learning_rate": 4.034377880816549e-05,
+      "loss": 0.6436,
+      "step": 952
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.24675672089746345,
+      "learning_rate": 4.032288536245904e-05,
+      "loss": 0.6032,
+      "step": 953
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.27847658196950265,
+      "learning_rate": 4.030197476060151e-05,
+      "loss": 0.8895,
+      "step": 954
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.2906208833219937,
+      "learning_rate": 4.028104702600529e-05,
+      "loss": 0.6393,
+      "step": 955
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.27945097510887307,
+      "learning_rate": 4.026010218210196e-05,
+      "loss": 0.6222,
+      "step": 956
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.29200987003626194,
+      "learning_rate": 4.023914025234226e-05,
+      "loss": 0.6844,
+      "step": 957
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.2907610468694776,
+      "learning_rate": 4.021816126019603e-05,
+      "loss": 0.6089,
+      "step": 958
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.2704168899213435,
+      "learning_rate": 4.0197165229152275e-05,
+      "loss": 0.8767,
+      "step": 959
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.299357952253091,
+      "learning_rate": 4.017615218271901e-05,
+      "loss": 0.6386,
+      "step": 960
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.26664106908173324,
+      "learning_rate": 4.015512214442335e-05,
+      "loss": 0.6562,
+      "step": 961
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.49441402942498397,
+      "learning_rate": 4.01340751378114e-05,
+      "loss": 0.6738,
+      "step": 962
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.2704118135493093,
+      "learning_rate": 4.011301118644829e-05,
+      "loss": 0.8876,
+      "step": 963
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.2929330328249647,
+      "learning_rate": 4.009193031391812e-05,
+      "loss": 0.6074,
+      "step": 964
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.29798787710083346,
+      "learning_rate": 4.007083254382391e-05,
+      "loss": 0.8825,
+      "step": 965
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.31353105980401064,
+      "learning_rate": 4.004971789978762e-05,
+      "loss": 0.6293,
+      "step": 966
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.2941872852761109,
+      "learning_rate": 4.0028586405450094e-05,
+      "loss": 0.6357,
+      "step": 967
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.3108048011590293,
+      "learning_rate": 4.000743808447107e-05,
+      "loss": 0.6776,
+      "step": 968
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.29049427417129275,
+      "learning_rate": 3.9986272960529084e-05,
+      "loss": 0.9088,
+      "step": 969
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.3026734473387737,
+      "learning_rate": 3.9965091057321515e-05,
+      "loss": 0.6616,
+      "step": 970
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.3037304886137329,
+      "learning_rate": 3.994389239856451e-05,
+      "loss": 0.6432,
+      "step": 971
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.26820619500282944,
+      "learning_rate": 3.992267700799299e-05,
+      "loss": 0.6436,
+      "step": 972
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.33832396282871163,
+      "learning_rate": 3.99014449093606e-05,
+      "loss": 0.6822,
+      "step": 973
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.2852625256949599,
+      "learning_rate": 3.988019612643969e-05,
+      "loss": 0.8898,
+      "step": 974
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.3252603742106901,
+      "learning_rate": 3.9858930683021315e-05,
+      "loss": 0.6149,
+      "step": 975
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.33873023656486045,
+      "learning_rate": 3.9837648602915156e-05,
+      "loss": 0.6571,
+      "step": 976
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.2935583966875909,
+      "learning_rate": 3.981634990994953e-05,
+      "loss": 0.92,
+      "step": 977
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.27302395683441083,
+      "learning_rate": 3.979503462797136e-05,
+      "loss": 0.6018,
+      "step": 978
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.29799768909862184,
+      "learning_rate": 3.977370278084613e-05,
+      "loss": 0.7148,
+      "step": 979
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.27521829244656487,
+      "learning_rate": 3.975235439245789e-05,
+      "loss": 0.9103,
+      "step": 980
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.286134675543587,
+      "learning_rate": 3.973098948670918e-05,
+      "loss": 0.6655,
+      "step": 981
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.27014162090000265,
+      "learning_rate": 3.970960808752108e-05,
+      "loss": 0.6488,
+      "step": 982
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.269348532936235,
+      "learning_rate": 3.968821021883309e-05,
+      "loss": 0.8769,
+      "step": 983
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.30278884862551336,
+      "learning_rate": 3.966679590460317e-05,
+      "loss": 0.4498,
+      "step": 984
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.2665441880356117,
+      "learning_rate": 3.96453651688077e-05,
+      "loss": 0.8824,
+      "step": 985
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.2874581913311213,
+      "learning_rate": 3.9623918035441445e-05,
+      "loss": 0.6208,
+      "step": 986
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.25581010880299376,
+      "learning_rate": 3.960245452851751e-05,
+      "loss": 0.6288,
+      "step": 987
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.28839495220387285,
+      "learning_rate": 3.958097467206734e-05,
+      "loss": 0.8705,
+      "step": 988
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.26463768253506037,
+      "learning_rate": 3.955947849014071e-05,
+      "loss": 0.6276,
+      "step": 989
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.26558055109254314,
+      "learning_rate": 3.953796600680565e-05,
+      "loss": 0.6087,
+      "step": 990
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.2679618851929887,
+      "learning_rate": 3.9516437246148447e-05,
+      "loss": 0.9202,
+      "step": 991
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.2729949370156611,
+      "learning_rate": 3.9494892232273604e-05,
+      "loss": 0.6323,
+      "step": 992
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.2472410796743674,
+      "learning_rate": 3.947333098930384e-05,
+      "loss": 0.6244,
+      "step": 993
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.26840892759333546,
+      "learning_rate": 3.945175354138003e-05,
+      "loss": 0.8801,
+      "step": 994
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.2682282603636198,
+      "learning_rate": 3.9430159912661204e-05,
+      "loss": 0.6666,
+      "step": 995
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.2540930403886213,
+      "learning_rate": 3.940855012732449e-05,
+      "loss": 0.6494,
+      "step": 996
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.259072983754302,
+      "learning_rate": 3.9386924209565146e-05,
+      "loss": 0.6345,
+      "step": 997
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.2463748880943405,
+      "learning_rate": 3.936528218359643e-05,
+      "loss": 0.6181,
+      "step": 998
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.3154241907212907,
+      "learning_rate": 3.93436240736497e-05,
+      "loss": 0.8841,
+      "step": 999
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.27210415082303807,
+      "learning_rate": 3.9321949903974264e-05,
+      "loss": 0.6275,
+      "step": 1000
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.27100550739610185,
+      "learning_rate": 3.9300259698837474e-05,
+      "loss": 0.6125,
+      "step": 1001
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.26579127963283805,
+      "learning_rate": 3.927855348252457e-05,
+      "loss": 0.8828,
+      "step": 1002
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.2812036063477121,
+      "learning_rate": 3.9256831279338765e-05,
+      "loss": 0.6073,
+      "step": 1003
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.27815319491019713,
+      "learning_rate": 3.923509311360116e-05,
+      "loss": 0.6487,
+      "step": 1004
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.2774837143869838,
+      "learning_rate": 3.921333900965072e-05,
+      "loss": 0.8579,
+      "step": 1005
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.29485068166495326,
+      "learning_rate": 3.919156899184426e-05,
+      "loss": 0.7002,
+      "step": 1006
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.26537513589238315,
+      "learning_rate": 3.916978308455641e-05,
+      "loss": 0.6437,
+      "step": 1007
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.3026306120538417,
+      "learning_rate": 3.914798131217962e-05,
+      "loss": 0.6304,
+      "step": 1008
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.2819193192052739,
+      "learning_rate": 3.9126163699124044e-05,
+      "loss": 0.8978,
+      "step": 1009
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.30325979906447476,
+      "learning_rate": 3.9104330269817615e-05,
+      "loss": 0.6196,
+      "step": 1010
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.27333938671225994,
+      "learning_rate": 3.9082481048705965e-05,
+      "loss": 0.6229,
+      "step": 1011
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.2862867837273008,
+      "learning_rate": 3.906061606025242e-05,
+      "loss": 0.8599,
+      "step": 1012
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.3000791923821104,
+      "learning_rate": 3.903873532893794e-05,
+      "loss": 0.6377,
+      "step": 1013
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.2820552117839618,
+      "learning_rate": 3.90168388792611e-05,
+      "loss": 0.6419,
+      "step": 1014
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.26113321709201986,
+      "learning_rate": 3.8994926735738116e-05,
+      "loss": 0.6482,
+      "step": 1015
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.29127762894051384,
+      "learning_rate": 3.8972998922902734e-05,
+      "loss": 0.8759,
+      "step": 1016
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.28580395178798634,
+      "learning_rate": 3.895105546530628e-05,
+      "loss": 0.6625,
+      "step": 1017
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.3153189700658617,
+      "learning_rate": 3.892909638751756e-05,
+      "loss": 0.6208,
+      "step": 1018
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.2747122188753921,
+      "learning_rate": 3.8907121714122896e-05,
+      "loss": 0.603,
+      "step": 1019
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.27394292265451936,
+      "learning_rate": 3.888513146972606e-05,
+      "loss": 0.8966,
+      "step": 1020
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.27671563167978014,
+      "learning_rate": 3.8863125678948264e-05,
+      "loss": 0.6337,
+      "step": 1021
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.24449382863127112,
+      "learning_rate": 3.8841104366428114e-05,
+      "loss": 0.5955,
+      "step": 1022
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.2784631248975209,
+      "learning_rate": 3.881906755682162e-05,
+      "loss": 0.9171,
+      "step": 1023
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.264730423617762,
+      "learning_rate": 3.879701527480211e-05,
+      "loss": 0.6261,
+      "step": 1024
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.28280595161346234,
+      "learning_rate": 3.8774947545060256e-05,
+      "loss": 0.6266,
+      "step": 1025
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.25279513967620343,
+      "learning_rate": 3.8752864392304036e-05,
+      "loss": 0.8847,
+      "step": 1026
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.2999851031502291,
+      "learning_rate": 3.8730765841258666e-05,
+      "loss": 0.6346,
+      "step": 1027
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.26695060195149956,
+      "learning_rate": 3.870865191666663e-05,
+      "loss": 0.7,
+      "step": 1028
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.2613585441350884,
+      "learning_rate": 3.8686522643287605e-05,
+      "loss": 0.6141,
+      "step": 1029
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.2672105675195364,
+      "learning_rate": 3.866437804589846e-05,
+      "loss": 0.6345,
+      "step": 1030
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.2642551915431863,
+      "learning_rate": 3.864221814929325e-05,
+      "loss": 0.8986,
+      "step": 1031
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.2731144403873441,
+      "learning_rate": 3.8620042978283096e-05,
+      "loss": 0.6074,
+      "step": 1032
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.2607879412410738,
+      "learning_rate": 3.859785255769628e-05,
+      "loss": 0.6316,
+      "step": 1033
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.2787630880458566,
+      "learning_rate": 3.8575646912378134e-05,
+      "loss": 0.9331,
+      "step": 1034
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.2719003438096046,
+      "learning_rate": 3.855342606719103e-05,
+      "loss": 0.6234,
+      "step": 1035
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.24973763037457455,
+      "learning_rate": 3.853119004701439e-05,
+      "loss": 0.6172,
+      "step": 1036
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.27401926315285363,
+      "learning_rate": 3.850893887674457e-05,
+      "loss": 0.8715,
+      "step": 1037
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.26141890291386766,
+      "learning_rate": 3.848667258129495e-05,
+      "loss": 0.6155,
+      "step": 1038
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.2754741644323808,
+      "learning_rate": 3.8464391185595794e-05,
+      "loss": 0.7,
+      "step": 1039
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.25339727092947495,
+      "learning_rate": 3.844209471459431e-05,
+      "loss": 0.6457,
+      "step": 1040
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.27407541926326584,
+      "learning_rate": 3.841978319325458e-05,
+      "loss": 0.6158,
+      "step": 1041
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.2566686433241358,
+      "learning_rate": 3.839745664655751e-05,
+      "loss": 0.8954,
+      "step": 1042
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.2716260618084611,
+      "learning_rate": 3.8375115099500844e-05,
+      "loss": 0.6113,
+      "step": 1043
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.2851699490203774,
+      "learning_rate": 3.835275857709915e-05,
+      "loss": 0.647,
+      "step": 1044
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.2715899284133272,
+      "learning_rate": 3.8330387104383705e-05,
+      "loss": 0.9489,
+      "step": 1045
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.2831573042961197,
+      "learning_rate": 3.830800070640258e-05,
+      "loss": 0.6311,
+      "step": 1046
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.2609686348614478,
+      "learning_rate": 3.828559940822052e-05,
+      "loss": 0.6342,
+      "step": 1047
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.293247874146037,
+      "learning_rate": 3.8263183234918974e-05,
+      "loss": 0.8855,
+      "step": 1048
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.27622752312323523,
+      "learning_rate": 3.824075221159604e-05,
+      "loss": 0.6387,
+      "step": 1049
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.2676765098012523,
+      "learning_rate": 3.821830636336644e-05,
+      "loss": 0.674,
+      "step": 1050
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.27212260909426994,
+      "learning_rate": 3.819584571536148e-05,
+      "loss": 0.6296,
+      "step": 1051
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.27549447263356563,
+      "learning_rate": 3.817337029272908e-05,
+      "loss": 0.9117,
+      "step": 1052
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.27300399919751595,
+      "learning_rate": 3.815088012063365e-05,
+      "loss": 0.6181,
+      "step": 1053
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.2773644733657228,
+      "learning_rate": 3.812837522425615e-05,
+      "loss": 0.627,
+      "step": 1054
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.2685478189871759,
+      "learning_rate": 3.810585562879402e-05,
+      "loss": 0.6396,
+      "step": 1055
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.3031867350271663,
+      "learning_rate": 3.808332135946115e-05,
+      "loss": 0.9331,
+      "step": 1056
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.26929050413535455,
+      "learning_rate": 3.806077244148786e-05,
+      "loss": 0.6546,
+      "step": 1057
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.26931163177634376,
+      "learning_rate": 3.803820890012085e-05,
+      "loss": 0.6486,
+      "step": 1058
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.28066540452947414,
+      "learning_rate": 3.801563076062326e-05,
+      "loss": 0.8708,
+      "step": 1059
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.28970507725310735,
+      "learning_rate": 3.79930380482745e-05,
+      "loss": 0.6295,
+      "step": 1060
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.2624585064049828,
+      "learning_rate": 3.7970430788370326e-05,
+      "loss": 0.7005,
+      "step": 1061
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.30136716440736233,
+      "learning_rate": 3.79478090062228e-05,
+      "loss": 0.6276,
+      "step": 1062
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.26272088717224346,
+      "learning_rate": 3.7925172727160225e-05,
+      "loss": 0.896,
+      "step": 1063
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.2954386190682708,
+      "learning_rate": 3.790252197652713e-05,
+      "loss": 0.6139,
+      "step": 1064
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.26621740023472573,
+      "learning_rate": 3.787985677968424e-05,
+      "loss": 0.6507,
+      "step": 1065
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.3040052208089204,
+      "learning_rate": 3.7857177162008507e-05,
+      "loss": 0.8814,
+      "step": 1066
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.28695572591852614,
+      "learning_rate": 3.7834483148892954e-05,
+      "loss": 0.6282,
+      "step": 1067
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.2501959679494841,
+      "learning_rate": 3.781177476574678e-05,
+      "loss": 0.6274,
+      "step": 1068
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.26327778063679164,
+      "learning_rate": 3.778905203799523e-05,
+      "loss": 0.8973,
+      "step": 1069
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.25389719296274377,
+      "learning_rate": 3.7766314991079665e-05,
+      "loss": 0.6263,
+      "step": 1070
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.24879537293469256,
+      "learning_rate": 3.7743563650457425e-05,
+      "loss": 0.6302,
+      "step": 1071
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.2702931490050958,
+      "learning_rate": 3.772079804160187e-05,
+      "loss": 0.6854,
+      "step": 1072
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.2461039179023537,
+      "learning_rate": 3.769801819000235e-05,
+      "loss": 0.6391,
+      "step": 1073
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.27980703367564563,
+      "learning_rate": 3.767522412116414e-05,
+      "loss": 0.8954,
+      "step": 1074
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.25829141649923326,
+      "learning_rate": 3.765241586060846e-05,
+      "loss": 0.6078,
+      "step": 1075
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.2697240956728446,
+      "learning_rate": 3.762959343387239e-05,
+      "loss": 0.6353,
+      "step": 1076
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.254335308890293,
+      "learning_rate": 3.760675686650889e-05,
+      "loss": 0.8954,
+      "step": 1077
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.2776851159099169,
+      "learning_rate": 3.758390618408675e-05,
+      "loss": 0.6313,
+      "step": 1078
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.2511199134239234,
+      "learning_rate": 3.756104141219056e-05,
+      "loss": 0.6138,
+      "step": 1079
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.26608518051157043,
+      "learning_rate": 3.753816257642068e-05,
+      "loss": 0.8503,
+      "step": 1080
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.2788315360822804,
+      "learning_rate": 3.7515269702393234e-05,
+      "loss": 0.6056,
+      "step": 1081
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.2617230501715106,
+      "learning_rate": 3.749236281574005e-05,
+      "loss": 0.615,
+      "step": 1082
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.26397470824969627,
+      "learning_rate": 3.746944194210864e-05,
+      "loss": 0.6704,
+      "step": 1083
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.27017891959184265,
+      "learning_rate": 3.744650710716219e-05,
+      "loss": 0.6495,
+      "step": 1084
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.2784747062958372,
+      "learning_rate": 3.742355833657951e-05,
+      "loss": 0.9195,
+      "step": 1085
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.2721241753299792,
+      "learning_rate": 3.7400595656055026e-05,
+      "loss": 0.6444,
+      "step": 1086
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.2690406076679362,
+      "learning_rate": 3.737761909129871e-05,
+      "loss": 0.6377,
+      "step": 1087
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.26189796056935244,
+      "learning_rate": 3.735462866803611e-05,
+      "loss": 0.9071,
+      "step": 1088
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.28544399796816755,
+      "learning_rate": 3.733162441200826e-05,
+      "loss": 0.684,
+      "step": 1089
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.24260006017894953,
+      "learning_rate": 3.730860634897172e-05,
+      "loss": 0.6546,
+      "step": 1090
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.26950813597662,
+      "learning_rate": 3.728557450469848e-05,
+      "loss": 0.622,
+      "step": 1091
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.2455304111370655,
+      "learning_rate": 3.7262528904975965e-05,
+      "loss": 0.7883,
+      "step": 1092
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.27475680086465476,
+      "learning_rate": 3.7239469575607e-05,
+      "loss": 0.6273,
+      "step": 1093
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.28318728012594396,
+      "learning_rate": 3.721639654240979e-05,
+      "loss": 0.6853,
+      "step": 1094
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.2973788091306982,
+      "learning_rate": 3.719330983121789e-05,
+      "loss": 0.9142,
+      "step": 1095
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.25851447816222484,
+      "learning_rate": 3.7170209467880146e-05,
+      "loss": 0.6434,
+      "step": 1096
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.2593170281090378,
+      "learning_rate": 3.71470954782607e-05,
+      "loss": 0.6195,
+      "step": 1097
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.26587737121103333,
+      "learning_rate": 3.7116125738064426e-05,
+      "loss": 0.8267,
+      "step": 1098
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.2733498612225663,
+      "learning_rate": 3.7092972177631e-05,
+      "loss": 0.6033,
+      "step": 1099
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.2649283234271871,
+      "learning_rate": 3.706980506826863e-05,
+      "loss": 0.6112,
+      "step": 1100
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.2771571890150801,
+      "learning_rate": 3.704662443593366e-05,
+      "loss": 0.8544,
+      "step": 1101
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.2517161484595993,
+      "learning_rate": 3.702343030659762e-05,
+      "loss": 0.608,
+      "step": 1102
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.2817660198091642,
+      "learning_rate": 3.7000222706247125e-05,
+      "loss": 0.5807,
+      "step": 1103
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.27626390549824775,
+      "learning_rate": 3.69770016608839e-05,
+      "loss": 0.8555,
+      "step": 1104
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.27212845824729404,
+      "learning_rate": 3.6953767196524744e-05,
+      "loss": 0.5907,
+      "step": 1105
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.29006225286914405,
+      "learning_rate": 3.693051933920147e-05,
+      "loss": 0.6075,
+      "step": 1106
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.250464439690243,
+      "learning_rate": 3.690725811496092e-05,
+      "loss": 0.6274,
+      "step": 1107
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.26460909325030646,
+      "learning_rate": 3.6883983549864876e-05,
+      "loss": 0.5984,
+      "step": 1108
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.2550606968279735,
+      "learning_rate": 3.686069566999011e-05,
+      "loss": 0.8287,
+      "step": 1109
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.2555774000933953,
+      "learning_rate": 3.6837394501428276e-05,
+      "loss": 0.5921,
+      "step": 1110
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.2517239216712275,
+      "learning_rate": 3.681408007028594e-05,
+      "loss": 0.5932,
+      "step": 1111
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.2604459432401807,
+      "learning_rate": 3.679075240268452e-05,
+      "loss": 0.8356,
+      "step": 1112
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.27217971062026886,
+      "learning_rate": 3.676741152476025e-05,
+      "loss": 0.6027,
+      "step": 1113
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.23813118811053668,
+      "learning_rate": 3.67440574626642e-05,
+      "loss": 0.5777,
+      "step": 1114
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.23493557642748994,
+      "learning_rate": 3.672069024256217e-05,
+      "loss": 0.8472,
+      "step": 1115
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.25079287759077656,
+      "learning_rate": 3.669730989063474e-05,
+      "loss": 0.609,
+      "step": 1116
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.22232243334897078,
+      "learning_rate": 3.667391643307716e-05,
+      "loss": 0.6031,
+      "step": 1117
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.2595351188737807,
+      "learning_rate": 3.665050989609941e-05,
+      "loss": 0.6188,
+      "step": 1118
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.2527907357186759,
+      "learning_rate": 3.66270903059261e-05,
+      "loss": 0.8548,
+      "step": 1119
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.26423707283727377,
+      "learning_rate": 3.6603657688796465e-05,
+      "loss": 0.5891,
+      "step": 1120
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6578030020669737,
+      "learning_rate": 3.6580212070964324e-05,
+      "loss": 0.5957,
+      "step": 1121
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.2524032244436048,
+      "learning_rate": 3.655675347869809e-05,
+      "loss": 0.6006,
+      "step": 1122
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.25801620213663423,
+      "learning_rate": 3.653328193828069e-05,
+      "loss": 0.8711,
+      "step": 1123
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.2545369512887966,
+      "learning_rate": 3.650979747600957e-05,
+      "loss": 0.6125,
+      "step": 1124
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.2557935866127448,
+      "learning_rate": 3.6486300118196655e-05,
+      "loss": 0.5975,
+      "step": 1125
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.2525195703672497,
+      "learning_rate": 3.6462789891168306e-05,
+      "loss": 0.8526,
+      "step": 1126
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.2871110986291175,
+      "learning_rate": 3.64392668212653e-05,
+      "loss": 0.5953,
+      "step": 1127
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.2673544885161775,
+      "learning_rate": 3.6415730934842827e-05,
+      "loss": 0.6156,
+      "step": 1128
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.2713865547313045,
+      "learning_rate": 3.6392182258270426e-05,
+      "loss": 0.6377,
+      "step": 1129
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.262321598505739,
+      "learning_rate": 3.636862081793194e-05,
+      "loss": 0.8206,
+      "step": 1130
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.2741568095799229,
+      "learning_rate": 3.6345046640225555e-05,
+      "loss": 0.606,
+      "step": 1131
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.24296988349477555,
+      "learning_rate": 3.6321459751563703e-05,
+      "loss": 0.5927,
+      "step": 1132
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.254940947697632,
+      "learning_rate": 3.629786017837305e-05,
+      "loss": 0.8421,
+      "step": 1133
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.2784221743562254,
+      "learning_rate": 3.6274247947094495e-05,
+      "loss": 0.5937,
+      "step": 1134
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.26699297871266237,
+      "learning_rate": 3.625062308418312e-05,
+      "loss": 0.5783,
+      "step": 1135
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.2710772059920261,
+      "learning_rate": 3.622698561610812e-05,
+      "loss": 0.8168,
+      "step": 1136
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.2835926634272541,
+      "learning_rate": 3.620333556935288e-05,
+      "loss": 0.604,
+      "step": 1137
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.2858911353963848,
+      "learning_rate": 3.617967297041481e-05,
+      "loss": 0.6073,
+      "step": 1138
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.261080627050715,
+      "learning_rate": 3.615599784580542e-05,
+      "loss": 0.6039,
+      "step": 1139
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.2638478460019322,
+      "learning_rate": 3.613231022205025e-05,
+      "loss": 0.6455,
+      "step": 1140
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.26126900756082744,
+      "learning_rate": 3.610861012568884e-05,
+      "loss": 0.8512,
+      "step": 1141
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.26956961366065907,
+      "learning_rate": 3.608489758327472e-05,
+      "loss": 0.5836,
+      "step": 1142
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.2585151350082767,
+      "learning_rate": 3.606117262137532e-05,
+      "loss": 0.5842,
+      "step": 1143
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.25260422267553234,
+      "learning_rate": 3.603743526657204e-05,
+      "loss": 0.8589,
+      "step": 1144
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.28988164550103573,
+      "learning_rate": 3.601368554546012e-05,
+      "loss": 0.5913,
+      "step": 1145
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.26010154949032405,
+      "learning_rate": 3.5989923484648693e-05,
+      "loss": 0.5824,
+      "step": 1146
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.2721898617995486,
+      "learning_rate": 3.5966149110760686e-05,
+      "loss": 0.8287,
+      "step": 1147
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.2795682170558887,
+      "learning_rate": 3.594236245043284e-05,
+      "loss": 0.5851,
+      "step": 1148
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.2420093991480192,
+      "learning_rate": 3.5918563530315664e-05,
+      "loss": 0.5871,
+      "step": 1149
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.25315822462343957,
+      "learning_rate": 3.5894752377073375e-05,
+      "loss": 0.8579,
+      "step": 1150
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.2903949085249403,
+      "learning_rate": 3.587092901738394e-05,
+      "loss": 0.4212,
+      "step": 1151
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.2607625322539113,
+      "learning_rate": 3.5847093477938956e-05,
+      "loss": 0.8628,
+      "step": 1152
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.30046151712122454,
+      "learning_rate": 3.58232457854437e-05,
+      "loss": 0.581,
+      "step": 1153
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.29459523870810356,
+      "learning_rate": 3.579938596661706e-05,
+      "loss": 0.5989,
+      "step": 1154
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.28095541839248905,
+      "learning_rate": 3.577551404819148e-05,
+      "loss": 0.8847,
+      "step": 1155
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.3048988243172019,
+      "learning_rate": 3.575163005691302e-05,
+      "loss": 0.6076,
+      "step": 1156
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.24954300841245164,
+      "learning_rate": 3.572773401954119e-05,
+      "loss": 0.5867,
+      "step": 1157
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.26526092950846053,
+      "learning_rate": 3.570382596284907e-05,
+      "loss": 0.8205,
+      "step": 1158
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.27951351478421604,
+      "learning_rate": 3.5679905913623165e-05,
+      "loss": 0.5849,
+      "step": 1159
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.2310969804801054,
+      "learning_rate": 3.5655973898663405e-05,
+      "loss": 0.5883,
+      "step": 1160
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.26637821772775994,
+      "learning_rate": 3.563202994478317e-05,
+      "loss": 0.8566,
+      "step": 1161
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.2730903526268172,
+      "learning_rate": 3.56080740788092e-05,
+      "loss": 0.6112,
+      "step": 1162
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.2459292003056661,
+      "learning_rate": 3.5584106327581536e-05,
+      "loss": 0.5843,
+      "step": 1163
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.2549694428683813,
+      "learning_rate": 3.55601267179536e-05,
+      "loss": 0.6118,
+      "step": 1164
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.2695402529501881,
+      "learning_rate": 3.553613527679208e-05,
+      "loss": 0.8384,
+      "step": 1165
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.26533339165598285,
+      "learning_rate": 3.551213203097689e-05,
+      "loss": 0.5868,
+      "step": 1166
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.24747942435623238,
+      "learning_rate": 3.5488117007401214e-05,
+      "loss": 0.5719,
+      "step": 1167
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.2594866831017697,
+      "learning_rate": 3.5464090232971395e-05,
+      "loss": 0.5934,
+      "step": 1168
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.24734600487823027,
+      "learning_rate": 3.5440051734606954e-05,
+      "loss": 0.8636,
+      "step": 1169
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.24938536422029978,
+      "learning_rate": 3.541600153924058e-05,
+      "loss": 0.5863,
+      "step": 1170
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.24951372610272732,
+      "learning_rate": 3.539193967381801e-05,
+      "loss": 0.6055,
+      "step": 1171
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.25477451748449,
+      "learning_rate": 3.5367866165298084e-05,
+      "loss": 0.8249,
+      "step": 1172
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.2670922624041818,
+      "learning_rate": 3.534378104065271e-05,
+      "loss": 0.6216,
+      "step": 1173
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.22741615645779528,
+      "learning_rate": 3.5319684326866765e-05,
+      "loss": 0.5776,
+      "step": 1174
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.25295588437004674,
+      "learning_rate": 3.5295576050938145e-05,
+      "loss": 0.5965,
+      "step": 1175
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.25564562321336154,
+      "learning_rate": 3.527145623987769e-05,
+      "loss": 0.8316,
+      "step": 1176
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.2317886943112256,
+      "learning_rate": 3.524732492070915e-05,
+      "loss": 0.5938,
+      "step": 1177
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.24126394375779175,
+      "learning_rate": 3.5223182120469196e-05,
+      "loss": 0.5782,
+      "step": 1178
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.261126482640371,
+      "learning_rate": 3.5199027866207334e-05,
+      "loss": 0.8327,
+      "step": 1179
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.3692268978973825,
+      "learning_rate": 3.517486218498592e-05,
+      "loss": 0.5841,
+      "step": 1180
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.24407688295516927,
+      "learning_rate": 3.515068510388012e-05,
+      "loss": 0.5964,
+      "step": 1181
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.26377947440227456,
+      "learning_rate": 3.512649664997786e-05,
+      "loss": 0.8249,
+      "step": 1182
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.26948379484508833,
+      "learning_rate": 3.510229685037979e-05,
+      "loss": 0.5936,
+      "step": 1183
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.26438703051254603,
+      "learning_rate": 3.507808573219931e-05,
+      "loss": 0.6317,
+      "step": 1184
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.2715708010219487,
+      "learning_rate": 3.505386332256248e-05,
+      "loss": 0.5866,
+      "step": 1185
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.2523541398186074,
+      "learning_rate": 3.5029629648608e-05,
+      "loss": 0.6149,
+      "step": 1186
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.2561336338935547,
+      "learning_rate": 3.5005384737487225e-05,
+      "loss": 0.8841,
+      "step": 1187
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.2533047740412533,
+      "learning_rate": 3.4981128616364056e-05,
+      "loss": 0.5778,
+      "step": 1188
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.26031121076819735,
+      "learning_rate": 3.4956861312414995e-05,
+      "loss": 0.6122,
+      "step": 1189
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.26439316536209917,
+      "learning_rate": 3.493258285282905e-05,
+      "loss": 0.6039,
+      "step": 1190
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.25692826722702294,
+      "learning_rate": 3.4908293264807734e-05,
+      "loss": 0.8296,
+      "step": 1191
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.2980205327363792,
+      "learning_rate": 3.488399257556502e-05,
+      "loss": 0.5728,
+      "step": 1192
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.28761454683041693,
+      "learning_rate": 3.485968081232733e-05,
+      "loss": 0.8253,
+      "step": 1193
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.33170181279109534,
+      "learning_rate": 3.4835358002333495e-05,
+      "loss": 0.6053,
+      "step": 1194
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.31063040129587055,
+      "learning_rate": 3.48110241728347e-05,
+      "loss": 0.6551,
+      "step": 1195
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.48991408727527513,
+      "learning_rate": 3.4786679351094526e-05,
+      "loss": 0.5829,
+      "step": 1196
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.3035110971723903,
+      "learning_rate": 3.47623235643888e-05,
+      "loss": 0.6019,
+      "step": 1197
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.2763238137664252,
+      "learning_rate": 3.473795684000569e-05,
+      "loss": 0.8216,
+      "step": 1198
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.30325332235021213,
+      "learning_rate": 3.471357920524559e-05,
+      "loss": 0.6007,
+      "step": 1199
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.29449029411956185,
+      "learning_rate": 3.468919068742113e-05,
+      "loss": 0.6649,
+      "step": 1200
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.2883862579177778,
+      "learning_rate": 3.466479131385714e-05,
+      "loss": 0.5835,
+      "step": 1201
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.2616610328060092,
+      "learning_rate": 3.464038111189059e-05,
+      "loss": 0.8461,
+      "step": 1202
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.31478395266635717,
+      "learning_rate": 3.4615960108870594e-05,
+      "loss": 0.6227,
+      "step": 1203
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.29971611726202524,
+      "learning_rate": 3.459152833215839e-05,
+      "loss": 0.5914,
+      "step": 1204
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.2562415859248358,
+      "learning_rate": 3.456708580912725e-05,
+      "loss": 0.82,
+      "step": 1205
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.30585382044616116,
+      "learning_rate": 3.45426325671625e-05,
+      "loss": 0.5659,
+      "step": 1206
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.23523487909550297,
+      "learning_rate": 3.451816863366148e-05,
+      "loss": 0.5933,
+      "step": 1207
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.25883857105298935,
+      "learning_rate": 3.449369403603351e-05,
+      "loss": 0.8556,
+      "step": 1208
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.27114277455001456,
+      "learning_rate": 3.4469208801699846e-05,
+      "loss": 0.5766,
+      "step": 1209
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.2714164891537529,
+      "learning_rate": 3.444471295809368e-05,
+      "loss": 0.5995,
+      "step": 1210
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.27145450377543995,
+      "learning_rate": 3.442020653266007e-05,
+      "loss": 0.6332,
+      "step": 1211
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.2755910479204361,
+      "learning_rate": 3.4395689552855955e-05,
+      "loss": 0.5889,
+      "step": 1212
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.2777496193946996,
+      "learning_rate": 3.437116204615007e-05,
+      "loss": 0.8269,
+      "step": 1213
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.3121370077048329,
+      "learning_rate": 3.434662404002296e-05,
+      "loss": 0.5924,
+      "step": 1214
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.2724229634035726,
+      "learning_rate": 3.4322075561966936e-05,
+      "loss": 0.5999,
+      "step": 1215
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.31302112978462626,
+      "learning_rate": 3.4297516639486035e-05,
+      "loss": 0.8254,
+      "step": 1216
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.3156255719880212,
+      "learning_rate": 3.427294730009599e-05,
+      "loss": 0.6044,
+      "step": 1217
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.3674525764820428,
+      "learning_rate": 3.4248367571324237e-05,
+      "loss": 0.6024,
+      "step": 1218
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.29282289559179325,
+      "learning_rate": 3.4223777480709805e-05,
+      "loss": 0.843,
+      "step": 1219
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.2592997096518948,
+      "learning_rate": 3.4199177055803364e-05,
+      "loss": 0.5978,
+      "step": 1220
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.249114943965817,
+      "learning_rate": 3.4174566324167156e-05,
+      "loss": 0.6012,
+      "step": 1221
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.2840186301347517,
+      "learning_rate": 3.4149945313374977e-05,
+      "loss": 0.631,
+      "step": 1222
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.24440103230360824,
+      "learning_rate": 3.4125314051012114e-05,
+      "loss": 0.595,
+      "step": 1223
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.2574360809974802,
+      "learning_rate": 3.410067256467538e-05,
+      "loss": 0.8615,
+      "step": 1224
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.2725932492540637,
+      "learning_rate": 3.4076020881973e-05,
+      "loss": 0.5841,
+      "step": 1225
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.2727301341151183,
+      "learning_rate": 3.4051359030524654e-05,
+      "loss": 0.5814,
+      "step": 1226
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.2433578767540718,
+      "learning_rate": 3.402668703796141e-05,
+      "loss": 0.8234,
+      "step": 1227
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.30609464432394173,
+      "learning_rate": 3.400200493192568e-05,
+      "loss": 0.5895,
+      "step": 1228
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.23619517332544013,
+      "learning_rate": 3.397731274007123e-05,
+      "loss": 0.588,
+      "step": 1229
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.2721818898937165,
+      "learning_rate": 3.395261049006311e-05,
+      "loss": 0.8527,
+      "step": 1230
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.2567232118729849,
+      "learning_rate": 3.392789820957764e-05,
+      "loss": 0.6048,
+      "step": 1231
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.24296432298047038,
+      "learning_rate": 3.390317592630239e-05,
+      "loss": 0.5995,
+      "step": 1232
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.27797171539914295,
+      "learning_rate": 3.387844366793614e-05,
+      "loss": 0.645,
+      "step": 1233
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.27686102796070533,
+      "learning_rate": 3.385370146218883e-05,
+      "loss": 0.8197,
+      "step": 1234
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.2493323868205391,
+      "learning_rate": 3.382894933678154e-05,
+      "loss": 0.5764,
+      "step": 1235
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.2330995924287119,
+      "learning_rate": 3.3804187319446484e-05,
+      "loss": 0.6009,
+      "step": 1236
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.2619922751981524,
+      "learning_rate": 3.3779415437926955e-05,
+      "loss": 0.5822,
+      "step": 1237
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.2642636741702953,
+      "learning_rate": 3.3754633719977294e-05,
+      "loss": 0.8398,
+      "step": 1238
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.28361869837050313,
+      "learning_rate": 3.372984219336284e-05,
+      "loss": 0.593,
+      "step": 1239
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.25525222542697507,
+      "learning_rate": 3.3705040885859975e-05,
+      "loss": 0.5921,
+      "step": 1240
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.25452859350277207,
+      "learning_rate": 3.368022982525599e-05,
+      "loss": 0.8176,
+      "step": 1241
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.2791743551289705,
+      "learning_rate": 3.365540903934912e-05,
+      "loss": 0.5791,
+      "step": 1242
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.24854830609762285,
+      "learning_rate": 3.36305785559485e-05,
+      "loss": 0.5962,
+      "step": 1243
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.27756335783773806,
+      "learning_rate": 3.3605738402874125e-05,
+      "loss": 0.6356,
+      "step": 1244
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.24891139347547742,
+      "learning_rate": 3.358088860795684e-05,
+      "loss": 0.8261,
+      "step": 1245
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.269715221449274,
+      "learning_rate": 3.355602919903826e-05,
+      "loss": 0.6068,
+      "step": 1246
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.24282076775670564,
+      "learning_rate": 3.3531160203970805e-05,
+      "loss": 0.5788,
+      "step": 1247
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.261447048857834,
+      "learning_rate": 3.350628165061762e-05,
+      "loss": 0.8238,
+      "step": 1248
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.2663573626183426,
+      "learning_rate": 3.3481393566852556e-05,
+      "loss": 0.5898,
+      "step": 1249
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6064309166040533,
+      "learning_rate": 3.345649598056013e-05,
+      "loss": 0.6113,
+      "step": 1250
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.27147051204275274,
+      "learning_rate": 3.343158891963555e-05,
+      "loss": 0.8576,
+      "step": 1251
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.25732926311420806,
+      "learning_rate": 3.34066724119846e-05,
+      "loss": 0.5901,
+      "step": 1252
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.2453449628203063,
+      "learning_rate": 3.338174648552365e-05,
+      "loss": 0.58,
+      "step": 1253
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.2606721600660097,
+      "learning_rate": 3.335681116817963e-05,
+      "loss": 0.8408,
+      "step": 1254
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.27879272472663214,
+      "learning_rate": 3.333186648789001e-05,
+      "loss": 0.4229,
+      "step": 1255
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.26032012985761416,
+      "learning_rate": 3.3306912472602715e-05,
+      "loss": 0.8165,
+      "step": 1256
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.2828902159760864,
+      "learning_rate": 3.328194915027617e-05,
+      "loss": 0.587,
+      "step": 1257
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.27324952333261454,
+      "learning_rate": 3.3256976548879184e-05,
+      "loss": 0.5898,
+      "step": 1258
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.2550079254475527,
+      "learning_rate": 3.3231994696390984e-05,
+      "loss": 0.8626,
+      "step": 1259
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.2697919741815741,
+      "learning_rate": 3.320700362080118e-05,
+      "loss": 0.5996,
+      "step": 1260
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.24228473788549146,
+      "learning_rate": 3.318200335010967e-05,
+      "loss": 0.5912,
+      "step": 1261
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5241459420225462,
+      "learning_rate": 3.31569939123267e-05,
+      "loss": 0.8487,
+      "step": 1262
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.28286588405840957,
+      "learning_rate": 3.3131975335472764e-05,
+      "loss": 0.5999,
+      "step": 1263
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.2525734855722254,
+      "learning_rate": 3.310694764757859e-05,
+      "loss": 0.6051,
+      "step": 1264
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.24524273567696298,
+      "learning_rate": 3.308191087668513e-05,
+      "loss": 0.8216,
+      "step": 1265
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.30619080482346756,
+      "learning_rate": 3.305686505084351e-05,
+      "loss": 0.4162,
+      "step": 1266
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.24730217232494356,
+      "learning_rate": 3.303181019811498e-05,
+      "loss": 0.8368,
+      "step": 1267
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.2862246582310058,
+      "learning_rate": 3.300674634657094e-05,
+      "loss": 0.6048,
+      "step": 1268
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.28828480119349825,
+      "learning_rate": 3.2981673524292844e-05,
+      "loss": 0.5934,
+      "step": 1269
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.2519984133314176,
+      "learning_rate": 3.2956591759372204e-05,
+      "loss": 0.8095,
+      "step": 1270
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.3024101091481878,
+      "learning_rate": 3.293150107991055e-05,
+      "loss": 0.606,
+      "step": 1271
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.2565913554763378,
+      "learning_rate": 3.2906401514019406e-05,
+      "loss": 0.5975,
+      "step": 1272
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.2565291084607752,
+      "learning_rate": 3.2881293089820255e-05,
+      "loss": 0.8406,
+      "step": 1273
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.28934880189955003,
+      "learning_rate": 3.2856175835444483e-05,
+      "loss": 0.592,
+      "step": 1274
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.2500303210194966,
+      "learning_rate": 3.28310497790334e-05,
+      "loss": 0.6093,
+      "step": 1275
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.250674387753426,
+      "learning_rate": 3.280591494873815e-05,
+      "loss": 0.8421,
+      "step": 1276
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.27432185693176125,
+      "learning_rate": 3.278077137271972e-05,
+      "loss": 0.6276,
+      "step": 1277
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.2366460865145804,
+      "learning_rate": 3.275561907914891e-05,
+      "loss": 0.6116,
+      "step": 1278
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.2593996223274248,
+      "learning_rate": 3.273045809620626e-05,
+      "loss": 0.5929,
+      "step": 1279
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.2488235492609129,
+      "learning_rate": 3.270528845208207e-05,
+      "loss": 0.8505,
+      "step": 1280
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.23552889704550162,
+      "learning_rate": 3.268011017497632e-05,
+      "loss": 0.5871,
+      "step": 1281
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.24202393489917381,
+      "learning_rate": 3.265492329309867e-05,
+      "loss": 0.587,
+      "step": 1282
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.2439242859652751,
+      "learning_rate": 3.2629727834668446e-05,
+      "loss": 0.59,
+      "step": 1283
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.2452497284717802,
+      "learning_rate": 3.2604523827914545e-05,
+      "loss": 0.8126,
+      "step": 1284
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.28724445431957374,
+      "learning_rate": 3.257931130107546e-05,
+      "loss": 0.5964,
+      "step": 1285
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.24548326552330194,
+      "learning_rate": 3.255409028239924e-05,
+      "loss": 0.5894,
+      "step": 1286
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.275164146873826,
+      "learning_rate": 3.252886080014342e-05,
+      "loss": 0.8154,
+      "step": 1287
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.2811966460533248,
+      "learning_rate": 3.250362288257506e-05,
+      "loss": 0.6148,
+      "step": 1288
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.2297652759665439,
+      "learning_rate": 3.247837655797061e-05,
+      "loss": 0.5843,
+      "step": 1289
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.24629386044147836,
+      "learning_rate": 3.2453121854616006e-05,
+      "loss": 0.5965,
+      "step": 1290
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.2601470683315707,
+      "learning_rate": 3.242785880080653e-05,
+      "loss": 0.8183,
+      "step": 1291
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.2643370632780166,
+      "learning_rate": 3.240258742484683e-05,
+      "loss": 0.5892,
+      "step": 1292
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.25082501704911675,
+      "learning_rate": 3.2377307755050866e-05,
+      "loss": 0.5994,
+      "step": 1293
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.24569835086526984,
+      "learning_rate": 3.235201981974193e-05,
+      "loss": 0.6087,
+      "step": 1294
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.2670853838718378,
+      "learning_rate": 3.2326723647252515e-05,
+      "loss": 0.7487,
+      "step": 1295
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.23229936768126075,
+      "learning_rate": 3.2301419265924395e-05,
+      "loss": 0.5846,
+      "step": 1296
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.26187634781075936,
+      "learning_rate": 3.227610670410853e-05,
+      "loss": 0.8503,
+      "step": 1297
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.24965315021731857,
+      "learning_rate": 3.2250785990165e-05,
+      "loss": 0.5942,
+      "step": 1298
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.335106657591346,
+      "learning_rate": 3.222545715246309e-05,
+      "loss": 0.6344,
+      "step": 1299
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.2664604154818425,
+      "learning_rate": 3.2200120219381145e-05,
+      "loss": 0.6086,
+      "step": 1300
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.296600591925262,
+      "learning_rate": 3.217477521930658e-05,
+      "loss": 0.6131,
+      "step": 1301
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.2510187265953851,
+      "learning_rate": 3.2149422180635855e-05,
+      "loss": 0.819,
+      "step": 1302
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.28566157649993634,
+      "learning_rate": 3.212406113177445e-05,
+      "loss": 0.5993,
+      "step": 1303
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.24263443141089605,
+      "learning_rate": 3.209869210113679e-05,
+      "loss": 0.5746,
+      "step": 1304
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.27054949308864373,
+      "learning_rate": 3.207331511714626e-05,
+      "loss": 0.6102,
+      "step": 1305
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.2608754602656662,
+      "learning_rate": 3.204793020823517e-05,
+      "loss": 0.8442,
+      "step": 1306
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.36341692431345834,
+      "learning_rate": 3.202253740284468e-05,
+      "loss": 0.5831,
+      "step": 1307
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.2517115226308986,
+      "learning_rate": 3.199713672942481e-05,
+      "loss": 0.5847,
+      "step": 1308
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.27863177627717717,
+      "learning_rate": 3.1971728216434415e-05,
+      "loss": 0.6039,
+      "step": 1309
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.2762493196398508,
+      "learning_rate": 3.19463118923411e-05,
+      "loss": 0.8307,
+      "step": 1310
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.23602379810276974,
+      "learning_rate": 3.1920887785621235e-05,
+      "loss": 0.591,
+      "step": 1311
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.25768685309479655,
+      "learning_rate": 3.189545592475992e-05,
+      "loss": 0.5821,
+      "step": 1312
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.2479726486697962,
+      "learning_rate": 3.187001633825093e-05,
+      "loss": 0.846,
+      "step": 1313
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.2561662218588898,
+      "learning_rate": 3.184456905459669e-05,
+      "loss": 0.5876,
+      "step": 1314
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.22440409106046544,
+      "learning_rate": 3.181911410230828e-05,
+      "loss": 0.5829,
+      "step": 1315
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.2637266032989342,
+      "learning_rate": 3.179365150990533e-05,
+      "loss": 0.6254,
+      "step": 1316
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.237933997349468,
+      "learning_rate": 3.1768181305916066e-05,
+      "loss": 0.8236,
+      "step": 1317
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.29994447677180613,
+      "learning_rate": 3.1742703518877217e-05,
+      "loss": 0.62,
+      "step": 1318
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.23959204166666126,
+      "learning_rate": 3.1717218177334016e-05,
+      "loss": 0.5932,
+      "step": 1319
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.2786227343920628,
+      "learning_rate": 3.169172530984016e-05,
+      "loss": 0.8191,
+      "step": 1320
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.26351467748321683,
+      "learning_rate": 3.1666224944957794e-05,
+      "loss": 0.5859,
+      "step": 1321
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.2558552141997149,
+      "learning_rate": 3.164071711125744e-05,
+      "loss": 0.5939,
+      "step": 1322
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.24341306374189361,
+      "learning_rate": 3.161520183731799e-05,
+      "loss": 0.8192,
+      "step": 1323
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.28266765015435275,
+      "learning_rate": 3.158967915172669e-05,
+      "loss": 0.5807,
+      "step": 1324
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.24142812853472076,
+      "learning_rate": 3.156414908307908e-05,
+      "loss": 0.5811,
+      "step": 1325
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.27241300449600037,
+      "learning_rate": 3.153861165997895e-05,
+      "loss": 0.5816,
+      "step": 1326
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.2592009350090651,
+      "learning_rate": 3.151306691103837e-05,
+      "loss": 0.6195,
+      "step": 1327
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.2675657668318975,
+      "learning_rate": 3.14875148648776e-05,
+      "loss": 0.832,
+      "step": 1328
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.2550539362529064,
+      "learning_rate": 3.146195555012506e-05,
+      "loss": 0.6011,
+      "step": 1329
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.26631279719398915,
+      "learning_rate": 3.1436388995417334e-05,
+      "loss": 0.5867,
+      "step": 1330
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.26206767824048033,
+      "learning_rate": 3.1410815229399116e-05,
+      "loss": 0.8327,
+      "step": 1331
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.30219804729612026,
+      "learning_rate": 3.138523428072316e-05,
+      "loss": 0.5938,
+      "step": 1332
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.2488208638395867,
+      "learning_rate": 3.135964617805032e-05,
+      "loss": 0.5755,
+      "step": 1333
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.25770648814813935,
+      "learning_rate": 3.1334050950049384e-05,
+      "loss": 0.8474,
+      "step": 1334
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.2567848576785906,
+      "learning_rate": 3.130844862539718e-05,
+      "loss": 0.5828,
+      "step": 1335
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.24401574453395092,
+      "learning_rate": 3.1282839232778496e-05,
+      "loss": 0.6023,
+      "step": 1336
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.26219524782416925,
+      "learning_rate": 3.1257222800885985e-05,
+      "loss": 0.8682,
+      "step": 1337
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.2546318668675058,
+      "learning_rate": 3.123159935842024e-05,
+      "loss": 0.4087,
+      "step": 1338
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.2506168755295733,
+      "learning_rate": 3.120596893408968e-05,
+      "loss": 0.8175,
+      "step": 1339
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.24554512411018148,
+      "learning_rate": 3.118033155661055e-05,
+      "loss": 0.5968,
+      "step": 1340
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.2493881717792444,
+      "learning_rate": 3.1154687254706906e-05,
+      "loss": 0.5904,
+      "step": 1341
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.2440722991262341,
+      "learning_rate": 3.1129036057110524e-05,
+      "loss": 0.8144,
+      "step": 1342
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.26068799650483687,
+      "learning_rate": 3.1103377992560934e-05,
+      "loss": 0.5703,
+      "step": 1343
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.2515130911479186,
+      "learning_rate": 3.107771308980535e-05,
+      "loss": 0.5865,
+      "step": 1344
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.24574314455938087,
+      "learning_rate": 3.1052041377598674e-05,
+      "loss": 0.8106,
+      "step": 1345
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.2484162373682119,
+      "learning_rate": 3.102636288470338e-05,
+      "loss": 0.5822,
+      "step": 1346
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.26923230201768616,
+      "learning_rate": 3.10006776398896e-05,
+      "loss": 0.6067,
+      "step": 1347
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.23959788656138792,
+      "learning_rate": 3.097498567193499e-05,
+      "loss": 0.8395,
+      "step": 1348
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.2861592945638938,
+      "learning_rate": 3.0949287009624744e-05,
+      "loss": 0.6256,
+      "step": 1349
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.23761761718184476,
+      "learning_rate": 3.092358168175158e-05,
+      "loss": 0.5966,
+      "step": 1350
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.2538586853561143,
+      "learning_rate": 3.089786971711566e-05,
+      "loss": 0.5813,
+      "step": 1351
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.2457841016149223,
+      "learning_rate": 3.0872151144524595e-05,
+      "loss": 0.8419,
+      "step": 1352
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.3113892756742568,
+      "learning_rate": 3.084642599279339e-05,
+      "loss": 0.596,
+      "step": 1353
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.25507163046601045,
+      "learning_rate": 3.082069429074443e-05,
+      "loss": 0.5817,
+      "step": 1354
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.24861840928570283,
+      "learning_rate": 3.079495606720743e-05,
+      "loss": 0.583,
+      "step": 1355
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.24788133274605365,
+      "learning_rate": 3.076921135101943e-05,
+      "loss": 0.8374,
+      "step": 1356
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.2711726040034529,
+      "learning_rate": 3.074346017102472e-05,
+      "loss": 0.5788,
+      "step": 1357
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.26477803661428484,
+      "learning_rate": 3.071770255607484e-05,
+      "loss": 0.5985,
+      "step": 1358
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.24800552416915647,
+      "learning_rate": 3.069193853502855e-05,
+      "loss": 0.8338,
+      "step": 1359
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.0628487361818373,
+      "learning_rate": 3.0666168136751776e-05,
+      "loss": 0.6084,
+      "step": 1360
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.23716151426728968,
+      "learning_rate": 3.06403913901176e-05,
+      "loss": 0.5803,
+      "step": 1361
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.23693324973012142,
+      "learning_rate": 3.06146083240062e-05,
+      "loss": 0.5796,
+      "step": 1362
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.24117551809579935,
+      "learning_rate": 3.058881896730485e-05,
+      "loss": 0.8233,
+      "step": 1363
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.2480031237499639,
+      "learning_rate": 3.056302334890786e-05,
+      "loss": 0.5982,
+      "step": 1364
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.22800940319358612,
+      "learning_rate": 3.0537221497716574e-05,
+      "loss": 0.5684,
+      "step": 1365
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.26205916468410523,
+      "learning_rate": 3.0511413442639296e-05,
+      "loss": 0.8274,
+      "step": 1366
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.23812580388490734,
+      "learning_rate": 3.0485599212591303e-05,
+      "loss": 0.5713,
+      "step": 1367
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.261233553678737,
+      "learning_rate": 3.045977883649477e-05,
+      "loss": 0.593,
+      "step": 1368
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.3145372520754924,
+      "learning_rate": 3.043395234327876e-05,
+      "loss": 0.8139,
+      "step": 1369
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.2444442613972144,
+      "learning_rate": 3.0408119761879217e-05,
+      "loss": 0.5821,
+      "step": 1370
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.2481923399216713,
+      "learning_rate": 3.0382281121238872e-05,
+      "loss": 0.6378,
+      "step": 1371
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.24882834388392955,
+      "learning_rate": 3.0356436450307256e-05,
+      "loss": 0.6083,
+      "step": 1372
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.22758019662439571,
+      "learning_rate": 3.0330585778040677e-05,
+      "loss": 0.5997,
+      "step": 1373
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.2604797201761227,
+      "learning_rate": 3.0304729133402127e-05,
+      "loss": 0.7913,
+      "step": 1374
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.23955657710950687,
+      "learning_rate": 3.0278866545361322e-05,
+      "loss": 0.6098,
+      "step": 1375
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.2625918273091092,
+      "learning_rate": 3.0252998042894627e-05,
+      "loss": 0.5921,
+      "step": 1376
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.23577644512407545,
+      "learning_rate": 3.022712365498503e-05,
+      "loss": 0.8205,
+      "step": 1377
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.24137572550933925,
+      "learning_rate": 3.020124341062212e-05,
+      "loss": 0.5863,
+      "step": 1378
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.21978070342100228,
+      "learning_rate": 3.0175357338802033e-05,
+      "loss": 0.5976,
+      "step": 1379
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.270744730213721,
+      "learning_rate": 3.014946546852746e-05,
+      "loss": 0.8029,
+      "step": 1380
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.2549840252690061,
+      "learning_rate": 3.012356782880757e-05,
+      "loss": 0.5743,
+      "step": 1381
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.24381132102792744,
+      "learning_rate": 3.0097664448657987e-05,
+      "loss": 0.6346,
+      "step": 1382
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.23871511045601684,
+      "learning_rate": 3.0071755357100782e-05,
+      "loss": 0.5849,
+      "step": 1383
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.26487077214651944,
+      "learning_rate": 3.0045840583164436e-05,
+      "loss": 0.5793,
+      "step": 1384
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8238346083146554,
+      "learning_rate": 3.0019920155883764e-05,
+      "loss": 0.8659,
+      "step": 1385
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.242730436782778,
+      "learning_rate": 2.999399410429994e-05,
+      "loss": 0.5889,
+      "step": 1386
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.2413725252176032,
+      "learning_rate": 2.996806245746044e-05,
+      "loss": 0.57,
+      "step": 1387
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.289983092352179,
+      "learning_rate": 2.9942125244418995e-05,
+      "loss": 0.8493,
+      "step": 1388
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.27326504794863493,
+      "learning_rate": 2.9916182494235574e-05,
+      "loss": 0.6141,
+      "step": 1389
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.22414163487220173,
+      "learning_rate": 2.9890234235976366e-05,
+      "loss": 0.5792,
+      "step": 1390
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.2548741953239817,
+      "learning_rate": 2.9864280498713714e-05,
+      "loss": 0.8005,
+      "step": 1391
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.24096640880833956,
+      "learning_rate": 2.9838321311526107e-05,
+      "loss": 0.5907,
+      "step": 1392
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.26186083450719433,
+      "learning_rate": 2.9812356703498135e-05,
+      "loss": 0.6363,
+      "step": 1393
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.24091710001311029,
+      "learning_rate": 2.978638670372047e-05,
+      "loss": 0.5813,
+      "step": 1394
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.2528645028488262,
+      "learning_rate": 2.976041134128982e-05,
+      "loss": 0.8267,
+      "step": 1395
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.24290734360768043,
+      "learning_rate": 2.9734430645308886e-05,
+      "loss": 0.5986,
+      "step": 1396
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.22851511593239773,
+      "learning_rate": 2.9708444644886385e-05,
+      "loss": 0.5978,
+      "step": 1397
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.23107857518722813,
+      "learning_rate": 2.9682453369136942e-05,
+      "loss": 0.5808,
+      "step": 1398
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.26599493335347,
+      "learning_rate": 2.9656456847181096e-05,
+      "loss": 0.8252,
+      "step": 1399
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.24509306777012657,
+      "learning_rate": 2.9630455108145278e-05,
+      "loss": 0.6101,
+      "step": 1400
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.23804925430561533,
+      "learning_rate": 2.960444818116176e-05,
+      "loss": 0.5836,
+      "step": 1401
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.2559025349595327,
+      "learning_rate": 2.957843609536861e-05,
+      "loss": 0.8257,
+      "step": 1402
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.25496663191072977,
+      "learning_rate": 2.95524188799097e-05,
+      "loss": 0.5924,
+      "step": 1403
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.24473571207620853,
+      "learning_rate": 2.952639656393464e-05,
+      "loss": 0.6219,
+      "step": 1404
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.2522598841212925,
+      "learning_rate": 2.950036917659874e-05,
+      "loss": 0.5874,
+      "step": 1405
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.25849464563261676,
+      "learning_rate": 2.947433674706302e-05,
+      "loss": 0.8206,
+      "step": 1406
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.2513268307480838,
+      "learning_rate": 2.944829930449412e-05,
+      "loss": 0.5738,
+      "step": 1407
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.266990526769919,
+      "learning_rate": 2.9422256878064325e-05,
+      "loss": 0.5901,
+      "step": 1408
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.24674618416884175,
+      "learning_rate": 2.9396209496951482e-05,
+      "loss": 0.6405,
+      "step": 1409
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.25268053146307773,
+      "learning_rate": 2.9370157190338997e-05,
+      "loss": 0.5789,
+      "step": 1410
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.2455349577550988,
+      "learning_rate": 2.9344099987415797e-05,
+      "loss": 0.7999,
+      "step": 1411
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.2697266113128308,
+      "learning_rate": 2.9318037917376297e-05,
+      "loss": 0.567,
+      "step": 1412
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.23022985618760966,
+      "learning_rate": 2.9291971009420342e-05,
+      "loss": 0.7388,
+      "step": 1413
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.2543749181295915,
+      "learning_rate": 2.9265899292753236e-05,
+      "loss": 0.5934,
+      "step": 1414
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.25860775979907497,
+      "learning_rate": 2.923982279658564e-05,
+      "loss": 0.6476,
+      "step": 1415
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.2575365961652106,
+      "learning_rate": 2.9213741550133584e-05,
+      "loss": 0.5746,
+      "step": 1416
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.24402319536446837,
+      "learning_rate": 2.918765558261841e-05,
+      "loss": 0.8558,
+      "step": 1417
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.26981602558215734,
+      "learning_rate": 2.9161564923266767e-05,
+      "loss": 0.5819,
+      "step": 1418
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.23350137747523206,
+      "learning_rate": 2.913546960131054e-05,
+      "loss": 0.5913,
+      "step": 1419
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.2586511635883862,
+      "learning_rate": 2.9109369645986856e-05,
+      "loss": 0.6386,
+      "step": 1420
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.24351486044456278,
+      "learning_rate": 2.908326508653802e-05,
+      "loss": 0.8392,
+      "step": 1421
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.2618279184903256,
+      "learning_rate": 2.9057155952211502e-05,
+      "loss": 0.5875,
+      "step": 1422
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.2345655248196274,
+      "learning_rate": 2.9031042272259895e-05,
+      "loss": 0.5785,
+      "step": 1423
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.2475409555345084,
+      "learning_rate": 2.900492407594088e-05,
+      "loss": 0.5775,
+      "step": 1424
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.24928467026504914,
+      "learning_rate": 2.8978801392517213e-05,
+      "loss": 0.8518,
+      "step": 1425
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.27899365523249,
+      "learning_rate": 2.8952674251256674e-05,
+      "loss": 0.5865,
+      "step": 1426
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.2480676689670008,
+      "learning_rate": 2.8926542681432012e-05,
+      "loss": 0.6024,
+      "step": 1427
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.2501143947570872,
+      "learning_rate": 2.890040671232097e-05,
+      "loss": 0.8387,
+      "step": 1428
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.2528517474491162,
+      "learning_rate": 2.8874266373206216e-05,
+      "loss": 0.5754,
+      "step": 1429
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.22220329222261997,
+      "learning_rate": 2.8848121693375285e-05,
+      "loss": 0.5846,
+      "step": 1430
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.25966914012108605,
+      "learning_rate": 2.8821972702120616e-05,
+      "loss": 0.6096,
+      "step": 1431
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.24655780668573513,
+      "learning_rate": 2.879581942873945e-05,
+      "loss": 0.8289,
+      "step": 1432
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.2760566796693583,
+      "learning_rate": 2.876966190253383e-05,
+      "loss": 0.5886,
+      "step": 1433
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.22056050422309378,
+      "learning_rate": 2.8743500152810575e-05,
+      "loss": 0.5787,
+      "step": 1434
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.26464766095815007,
+      "learning_rate": 2.8717334208881232e-05,
+      "loss": 0.8334,
+      "step": 1435
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.2475325730618213,
+      "learning_rate": 2.8691164100062034e-05,
+      "loss": 0.612,
+      "step": 1436
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.25174526588503615,
+      "learning_rate": 2.86649898556739e-05,
+      "loss": 0.5842,
+      "step": 1437
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.2674082659285835,
+      "learning_rate": 2.8638811505042362e-05,
+      "loss": 0.8443,
+      "step": 1438
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.25815921036744416,
+      "learning_rate": 2.8612629077497576e-05,
+      "loss": 0.5735,
+      "step": 1439
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.24700159033239455,
+      "learning_rate": 2.8586442602374246e-05,
+      "loss": 0.5632,
+      "step": 1440
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.23141027443426213,
+      "learning_rate": 2.856025210901162e-05,
+      "loss": 0.8251,
+      "step": 1441
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.25988756210003294,
+      "learning_rate": 2.853405762675345e-05,
+      "loss": 0.4078,
+      "step": 1442
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.25221097094079276,
+      "learning_rate": 2.8507859184947957e-05,
+      "loss": 0.8572,
+      "step": 1443
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.24286462270162792,
+      "learning_rate": 2.848165681294778e-05,
+      "loss": 0.575,
+      "step": 1444
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.262204224831166,
+      "learning_rate": 2.8455450540109985e-05,
+      "loss": 0.576,
+      "step": 1445
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.23788348282582872,
+      "learning_rate": 2.8429240395796007e-05,
+      "loss": 0.8259,
+      "step": 1446
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.24491610488491553,
+      "learning_rate": 2.8403026409371603e-05,
+      "loss": 0.582,
+      "step": 1447
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.23295379832901864,
+      "learning_rate": 2.837680861020684e-05,
+      "loss": 0.5785,
+      "step": 1448
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.24074529060202424,
+      "learning_rate": 2.8350587027676067e-05,
+      "loss": 0.853,
+      "step": 1449
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.23150664379175395,
+      "learning_rate": 2.8324361691157858e-05,
+      "loss": 0.5872,
+      "step": 1450
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.26040172050672294,
+      "learning_rate": 2.8298132630035005e-05,
+      "loss": 0.5923,
+      "step": 1451
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.24662879632111842,
+      "learning_rate": 2.827189987369446e-05,
+      "loss": 0.8263,
+      "step": 1452
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.2517099308928183,
+      "learning_rate": 2.8245663451527337e-05,
+      "loss": 0.4021,
+      "step": 1453
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.2417612140944877,
+      "learning_rate": 2.821942339292884e-05,
+      "loss": 0.8216,
+      "step": 1454
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.24231813262992719,
+      "learning_rate": 2.8193179727298236e-05,
+      "loss": 0.5661,
+      "step": 1455
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.2419835281214996,
+      "learning_rate": 2.8166932484038855e-05,
+      "loss": 0.5962,
+      "step": 1456
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.2688714481871249,
+      "learning_rate": 2.8140681692558035e-05,
+      "loss": 0.8332,
+      "step": 1457
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.2604151543453605,
+      "learning_rate": 2.8114427382267074e-05,
+      "loss": 0.5813,
+      "step": 1458
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.2383528014954793,
+      "learning_rate": 2.808816958258122e-05,
+      "loss": 0.596,
+      "step": 1459
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.24016570799469208,
+      "learning_rate": 2.8061908322919644e-05,
+      "loss": 0.8074,
+      "step": 1460
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.3445329375719745,
+      "learning_rate": 2.8035643632705362e-05,
+      "loss": 0.6054,
+      "step": 1461
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.2362405789900997,
+      "learning_rate": 2.8009375541365263e-05,
+      "loss": 0.5743,
+      "step": 1462
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.25018162905642927,
+      "learning_rate": 2.7983104078330037e-05,
+      "loss": 0.8126,
+      "step": 1463
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.25754685755656676,
+      "learning_rate": 2.7956829273034148e-05,
+      "loss": 0.6414,
+      "step": 1464
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.2414445011751455,
+      "learning_rate": 2.7930551154915806e-05,
+      "loss": 0.5836,
+      "step": 1465
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.23111998944764212,
+      "learning_rate": 2.790426975341694e-05,
+      "loss": 0.5763,
+      "step": 1466
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.25460290331907937,
+      "learning_rate": 2.787798509798314e-05,
+      "loss": 0.807,
+      "step": 1467
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.26011046589779535,
+      "learning_rate": 2.785169721806367e-05,
+      "loss": 0.599,
+      "step": 1468
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.23696852538051932,
+      "learning_rate": 2.7825406143111376e-05,
+      "loss": 0.6008,
+      "step": 1469
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.23665843702000758,
+      "learning_rate": 2.7799111902582696e-05,
+      "loss": 0.58,
+      "step": 1470
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.2578925117864392,
+      "learning_rate": 2.7772814525937635e-05,
+      "loss": 0.797,
+      "step": 1471
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.2549372646134835,
+      "learning_rate": 2.7746514042639677e-05,
+      "loss": 0.5983,
+      "step": 1472
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.2512044729844455,
+      "learning_rate": 2.7720210482155802e-05,
+      "loss": 0.5789,
+      "step": 1473
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.23895422244720388,
+      "learning_rate": 2.7693903873956446e-05,
+      "loss": 0.8293,
+      "step": 1474
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.2604876838450578,
+      "learning_rate": 2.766759424751545e-05,
+      "loss": 0.6423,
+      "step": 1475
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.2289905980863152,
+      "learning_rate": 2.7641281632310035e-05,
+      "loss": 0.6087,
+      "step": 1476
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.23263734914603434,
+      "learning_rate": 2.7614966057820786e-05,
+      "loss": 0.5843,
+      "step": 1477
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.24047808204393162,
+      "learning_rate": 2.7588647553531576e-05,
+      "loss": 0.7912,
+      "step": 1478
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.24503410916118237,
+      "learning_rate": 2.7562326148929585e-05,
+      "loss": 0.5833,
+      "step": 1479
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.21711259137769587,
+      "learning_rate": 2.753600187350524e-05,
+      "loss": 0.5954,
+      "step": 1480
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.25654554930174805,
+      "learning_rate": 2.750967475675216e-05,
+      "loss": 0.8274,
+      "step": 1481
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.23177275304405545,
+      "learning_rate": 2.7483344828167194e-05,
+      "loss": 0.5876,
+      "step": 1482
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.23268517752176712,
+      "learning_rate": 2.7457012117250286e-05,
+      "loss": 0.5768,
+      "step": 1483
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.22998586683633482,
+      "learning_rate": 2.743067665350453e-05,
+      "loss": 0.7905,
+      "step": 1484
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.2664297320610862,
+      "learning_rate": 2.740433846643612e-05,
+      "loss": 0.5882,
+      "step": 1485
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.24707313824036906,
+      "learning_rate": 2.737799758555425e-05,
+      "loss": 0.6315,
+      "step": 1486
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.2798932130685161,
+      "learning_rate": 2.7351654040371185e-05,
+      "loss": 0.596,
+      "step": 1487
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.22207372307648304,
+      "learning_rate": 2.7325307860402143e-05,
+      "loss": 0.5841,
+      "step": 1488
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.2351391062540594,
+      "learning_rate": 2.72989590751653e-05,
+      "loss": 0.799,
+      "step": 1489
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.23486377862646352,
+      "learning_rate": 2.7272607714181758e-05,
+      "loss": 0.5828,
+      "step": 1490
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.22851380109372932,
+      "learning_rate": 2.7246253806975507e-05,
+      "loss": 0.5765,
+      "step": 1491
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.24369953653562934,
+      "learning_rate": 2.7219897383073373e-05,
+      "loss": 0.8534,
+      "step": 1492
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.24772867016221342,
+      "learning_rate": 2.719353847200502e-05,
+      "loss": 0.5691,
+      "step": 1493
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.2230790652807608,
+      "learning_rate": 2.7167177103302888e-05,
+      "loss": 0.5556,
+      "step": 1494
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.24499496896885858,
+      "learning_rate": 2.714081330650217e-05,
+      "loss": 0.8243,
+      "step": 1495
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.226540053380452,
+      "learning_rate": 2.7114447111140793e-05,
+      "loss": 0.5779,
+      "step": 1496
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.2501417044560644,
+      "learning_rate": 2.7088078546759353e-05,
+      "loss": 0.6279,
+      "step": 1497
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.24109143736241334,
+      "learning_rate": 2.7061707642901112e-05,
+      "loss": 0.5813,
+      "step": 1498
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.22362097131165068,
+      "learning_rate": 2.7035334429111958e-05,
+      "loss": 0.5814,
+      "step": 1499
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.2604369682178177,
+      "learning_rate": 2.7008958934940344e-05,
+      "loss": 0.8614,
+      "step": 1500
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.2435605358100275,
+      "learning_rate": 2.6982581189937295e-05,
+      "loss": 0.5996,
+      "step": 1501
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.2342650354742389,
+      "learning_rate": 2.6956201223656364e-05,
+      "loss": 0.5705,
+      "step": 1502
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.24770541490106163,
+      "learning_rate": 2.6929819065653577e-05,
+      "loss": 0.8206,
+      "step": 1503
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.24081858708419723,
+      "learning_rate": 2.6903434745487426e-05,
+      "loss": 0.6036,
+      "step": 1504
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.23801890990234456,
+      "learning_rate": 2.6877048292718827e-05,
+      "loss": 0.581,
+      "step": 1505
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.4639921693725214,
+      "learning_rate": 2.6850659736911073e-05,
+      "loss": 0.8389,
+      "step": 1506
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.24117465504792418,
+      "learning_rate": 2.6824269107629822e-05,
+      "loss": 0.5712,
+      "step": 1507
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.267266972833994,
+      "learning_rate": 2.6797876434443064e-05,
+      "loss": 0.6325,
+      "step": 1508
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.2468179954411601,
+      "learning_rate": 2.6771481746921057e-05,
+      "loss": 0.5752,
+      "step": 1509
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.23851336025723593,
+      "learning_rate": 2.6745085074636333e-05,
+      "loss": 0.8499,
+      "step": 1510
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.28279175217515035,
+      "learning_rate": 2.671868644716364e-05,
+      "loss": 0.5893,
+      "step": 1511
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.25585652233880213,
+      "learning_rate": 2.6692285894079923e-05,
+      "loss": 0.6143,
+      "step": 1512
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.26265852673737056,
+      "learning_rate": 2.6665883444964278e-05,
+      "loss": 0.5897,
+      "step": 1513
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.125015006242018,
+      "learning_rate": 2.663947912939793e-05,
+      "loss": 0.6405,
+      "step": 1514
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.25839861092984523,
+      "learning_rate": 2.6613072976964194e-05,
+      "loss": 0.8047,
+      "step": 1515
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.24972994386032038,
+      "learning_rate": 2.6586665017248435e-05,
+      "loss": 0.5791,
+      "step": 1516
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.26401820086654304,
+      "learning_rate": 2.656025527983806e-05,
+      "loss": 0.8311,
+      "step": 1517
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.23576468919266735,
+      "learning_rate": 2.653384379432245e-05,
+      "loss": 0.6091,
+      "step": 1518
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.26451596971334135,
+      "learning_rate": 2.6507430590292963e-05,
+      "loss": 0.6289,
+      "step": 1519
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.2587652625110017,
+      "learning_rate": 2.648101569734286e-05,
+      "loss": 0.5848,
+      "step": 1520
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.2525333119725884,
+      "learning_rate": 2.645459914506731e-05,
+      "loss": 0.8399,
+      "step": 1521
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.29401379164275,
+      "learning_rate": 2.6428180963063343e-05,
+      "loss": 0.5828,
+      "step": 1522
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.4295140444031216,
+      "learning_rate": 2.6401761180929797e-05,
+      "loss": 0.6033,
+      "step": 1523
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.3393247615273886,
+      "learning_rate": 2.6375339828267327e-05,
+      "loss": 0.8315,
+      "step": 1524
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.267497955559874,
+      "learning_rate": 2.634891693467833e-05,
+      "loss": 0.4066,
+      "step": 1525
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.2739690395766233,
+      "learning_rate": 2.6322492529766934e-05,
+      "loss": 0.8361,
+      "step": 1526
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.0122365902189245,
+      "learning_rate": 2.629606664313896e-05,
+      "loss": 0.5869,
+      "step": 1527
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.25688616518334967,
+      "learning_rate": 2.6269639304401893e-05,
+      "loss": 0.5911,
+      "step": 1528
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.2671461527353263,
+      "learning_rate": 2.624321054316484e-05,
+      "loss": 0.8433,
+      "step": 1529
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.25511807491052746,
+      "learning_rate": 2.621678038903852e-05,
+      "loss": 0.742,
+      "step": 1530
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.3309031938108321,
+      "learning_rate": 2.6190348871635172e-05,
+      "loss": 0.4131,
+      "step": 1531
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.25221475675430816,
+      "learning_rate": 2.6163916020568595e-05,
+      "loss": 0.8296,
+      "step": 1532
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.28181004340543514,
+      "learning_rate": 2.6137481865454087e-05,
+      "loss": 0.6017,
+      "step": 1533
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.22444048761550092,
+      "learning_rate": 2.6111046435908383e-05,
+      "loss": 0.5785,
+      "step": 1534
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.2516951931487376,
+      "learning_rate": 2.6084609761549656e-05,
+      "loss": 0.8392,
+      "step": 1535
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.2680089033798968,
+      "learning_rate": 2.6058171871997476e-05,
+      "loss": 0.6521,
+      "step": 1536
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.25449846451966557,
+      "learning_rate": 2.6031732796872775e-05,
+      "loss": 0.5851,
+      "step": 1537
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.25210639937671914,
+      "learning_rate": 2.6005292565797817e-05,
+      "loss": 0.5859,
+      "step": 1538
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.23972664665147195,
+      "learning_rate": 2.597885120839614e-05,
+      "loss": 0.8236,
+      "step": 1539
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.24369561427730424,
+      "learning_rate": 2.5952408754292585e-05,
+      "loss": 0.588,
+      "step": 1540
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.22108160227791998,
+      "learning_rate": 2.5925965233113174e-05,
+      "loss": 0.5806,
+      "step": 1541
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.26224420593865544,
+      "learning_rate": 2.5899520674485156e-05,
+      "loss": 0.5823,
+      "step": 1542
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.2396616235219977,
+      "learning_rate": 2.587307510803693e-05,
+      "loss": 0.8297,
+      "step": 1543
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.27861904759114176,
+      "learning_rate": 2.584662856339804e-05,
+      "loss": 0.5673,
+      "step": 1544
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.22566264280043985,
+      "learning_rate": 2.582018107019909e-05,
+      "loss": 0.6102,
+      "step": 1545
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.2517792997902646,
+      "learning_rate": 2.579373265807179e-05,
+      "loss": 0.8342,
+      "step": 1546
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.26035775760002017,
+      "learning_rate": 2.5767283356648857e-05,
+      "loss": 0.6122,
+      "step": 1547
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.22696582283516864,
+      "learning_rate": 2.5740833195563996e-05,
+      "loss": 0.5731,
+      "step": 1548
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.2606042106158623,
+      "learning_rate": 2.5714382204451896e-05,
+      "loss": 0.5907,
+      "step": 1549
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.2491066057448401,
+      "learning_rate": 2.568793041294817e-05,
+      "loss": 0.7842,
+      "step": 1550
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.23760446452632053,
+      "learning_rate": 2.5661477850689313e-05,
+      "loss": 0.577,
+      "step": 1551
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.24305652532104985,
+      "learning_rate": 2.5635024547312704e-05,
+      "loss": 0.5908,
+      "step": 1552
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.24948654267400694,
+      "learning_rate": 2.5608570532456537e-05,
+      "loss": 0.8566,
+      "step": 1553
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.2417524436508635,
+      "learning_rate": 2.558211583575981e-05,
+      "loss": 0.5941,
+      "step": 1554
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.2237527131603693,
+      "learning_rate": 2.5555660486862294e-05,
+      "loss": 0.5819,
+      "step": 1555
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.2527475782262144,
+      "learning_rate": 2.5529204515404465e-05,
+      "loss": 0.8452,
+      "step": 1556
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.24108792766301773,
+      "learning_rate": 2.5502747951027516e-05,
+      "loss": 0.5713,
+      "step": 1557
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.23324943911092794,
+      "learning_rate": 2.5476290823373316e-05,
+      "loss": 0.6359,
+      "step": 1558
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.24869184709492206,
+      "learning_rate": 2.544983316208433e-05,
+      "loss": 0.5818,
+      "step": 1559
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.2326667111470398,
+      "learning_rate": 2.5423374996803646e-05,
+      "loss": 0.5817,
+      "step": 1560
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.22933790108588345,
+      "learning_rate": 2.5396916357174916e-05,
+      "loss": 0.8363,
+      "step": 1561
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.25930985588171146,
+      "learning_rate": 2.537045727284232e-05,
+      "loss": 0.5763,
+      "step": 1562
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.2535019158343706,
+      "learning_rate": 2.5343997773450523e-05,
+      "loss": 0.5681,
+      "step": 1563
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.2590834401234739,
+      "learning_rate": 2.5317537888644688e-05,
+      "loss": 0.8263,
+      "step": 1564
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.23895924005143385,
+      "learning_rate": 2.5291077648070366e-05,
+      "loss": 0.5723,
+      "step": 1565
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.2655861057341571,
+      "learning_rate": 2.5264617081373542e-05,
+      "loss": 0.5855,
+      "step": 1566
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.23436987041984159,
+      "learning_rate": 2.5238156218200554e-05,
+      "loss": 0.848,
+      "step": 1567
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.2761274901928886,
+      "learning_rate": 2.521169508819807e-05,
+      "loss": 0.57,
+      "step": 1568
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.23944309601799826,
+      "learning_rate": 2.5185233721013057e-05,
+      "loss": 0.6275,
+      "step": 1569
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.25685577406515814,
+      "learning_rate": 2.5158772146292748e-05,
+      "loss": 0.5698,
+      "step": 1570
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.22935713665549906,
+      "learning_rate": 2.5132310393684618e-05,
+      "loss": 0.5647,
+      "step": 1571
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.23715500810233797,
+      "learning_rate": 2.510584849283632e-05,
+      "loss": 0.8318,
+      "step": 1572
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.2725600995956339,
+      "learning_rate": 2.50793864733957e-05,
+      "loss": 0.582,
+      "step": 1573
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.2338654199695956,
+      "learning_rate": 2.5052924365010716e-05,
+      "loss": 0.5873,
+      "step": 1574
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.25086574824206015,
+      "learning_rate": 2.502646219732945e-05,
+      "loss": 0.8189,
+      "step": 1575
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.24551119117299947,
+      "learning_rate": 2.5e-05,
+      "loss": 0.5904,
+      "step": 1576
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.22056563013598254,
+      "learning_rate": 2.497353780267056e-05,
+      "loss": 0.5799,
+      "step": 1577
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.37159224855594175,
+      "learning_rate": 2.494707563498928e-05,
+      "loss": 0.8607,
+      "step": 1578
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.24791740446051322,
+      "learning_rate": 2.4920613526604312e-05,
+      "loss": 0.5778,
+      "step": 1579
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.2340016462209478,
+      "learning_rate": 2.489415150716369e-05,
+      "loss": 0.6354,
+      "step": 1580
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.25425929238029915,
+      "learning_rate": 2.4867689606315388e-05,
+      "loss": 0.5918,
+      "step": 1581
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.23838333790175992,
+      "learning_rate": 2.4841227853707254e-05,
+      "loss": 0.8228,
+      "step": 1582
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.25556365177926,
+      "learning_rate": 2.481476627898695e-05,
+      "loss": 0.5675,
+      "step": 1583
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.22562787759100933,
+      "learning_rate": 2.4788304911801934e-05,
+      "loss": 0.5762,
+      "step": 1584
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.23225610423010656,
+      "learning_rate": 2.4761843781799455e-05,
+      "loss": 0.579,
+      "step": 1585
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.2519045506099703,
+      "learning_rate": 2.4735382918626463e-05,
+      "loss": 0.816,
+      "step": 1586
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.2277877204925039,
+      "learning_rate": 2.470892235192964e-05,
+      "loss": 0.5892,
+      "step": 1587
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.23402377136072905,
+      "learning_rate": 2.468246211135532e-05,
+      "loss": 0.5867,
+      "step": 1588
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.24097998860146763,
+      "learning_rate": 2.4656002226549476e-05,
+      "loss": 0.8475,
+      "step": 1589
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.23989202508309385,
+      "learning_rate": 2.4629542727157684e-05,
+      "loss": 0.5676,
+      "step": 1590
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.22776987457320128,
+      "learning_rate": 2.460308364282509e-05,
+      "loss": 0.6169,
+      "step": 1591
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.22987391289209225,
+      "learning_rate": 2.4576625003196356e-05,
+      "loss": 0.5805,
+      "step": 1592
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.23634011079277953,
+      "learning_rate": 2.4550166837915676e-05,
+      "loss": 0.8328,
+      "step": 1593
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.22903304130928,
+      "learning_rate": 2.452370917662669e-05,
+      "loss": 0.6095,
+      "step": 1594
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.21903738973159545,
+      "learning_rate": 2.449725204897248e-05,
+      "loss": 0.5682,
+      "step": 1595
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.23760221172252471,
+      "learning_rate": 2.4470795484595548e-05,
+      "loss": 0.8188,
+      "step": 1596
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.24025676866502915,
+      "learning_rate": 2.444433951313772e-05,
+      "loss": 0.5771,
+      "step": 1597
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.22297403852655182,
+      "learning_rate": 2.441788416424019e-05,
+      "loss": 0.5808,
+      "step": 1598
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.24493948434349685,
+      "learning_rate": 2.439142946754347e-05,
+      "loss": 0.8069,
+      "step": 1599
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.2598455358434989,
+      "learning_rate": 2.43649754526873e-05,
+      "loss": 0.5737,
+      "step": 1600
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.23872338665485227,
+      "learning_rate": 2.433852214931069e-05,
+      "loss": 0.5905,
+      "step": 1601
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.2757569270928004,
+      "learning_rate": 2.4312069587051836e-05,
+      "loss": 0.6061,
+      "step": 1602
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.24809431322424808,
+      "learning_rate": 2.4285617795548106e-05,
+      "loss": 0.5912,
+      "step": 1603
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.2393887702773814,
+      "learning_rate": 2.4259166804436006e-05,
+      "loss": 0.8066,
+      "step": 1604
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.2349622385077667,
+      "learning_rate": 2.423271664335115e-05,
+      "loss": 0.5852,
+      "step": 1605
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.3151210172392955,
+      "learning_rate": 2.420626734192821e-05,
+      "loss": 0.5962,
+      "step": 1606
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.24492950158202648,
+      "learning_rate": 2.417981892980092e-05,
+      "loss": 0.8289,
+      "step": 1607
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.2561274526099224,
+      "learning_rate": 2.4153371436601973e-05,
+      "loss": 0.5838,
+      "step": 1608
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.21866004821628512,
+      "learning_rate": 2.4126924891963075e-05,
+      "loss": 0.583,
+      "step": 1609
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.24455133942169,
+      "learning_rate": 2.410047932551485e-05,
+      "loss": 0.824,
+      "step": 1610
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.24897537544312418,
+      "learning_rate": 2.407403476688683e-05,
+      "loss": 0.5866,
+      "step": 1611
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.21460993352013089,
+      "learning_rate": 2.4047591245707418e-05,
+      "loss": 0.5814,
+      "step": 1612
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.24159849006764905,
+      "learning_rate": 2.402114879160386e-05,
+      "loss": 0.6148,
+      "step": 1613
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.23519280767719403,
+      "learning_rate": 2.399470743420219e-05,
+      "loss": 0.5825,
+      "step": 1614
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.24188772700811959,
+      "learning_rate": 2.3968267203127227e-05,
+      "loss": 0.833,
+      "step": 1615
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.2422817891586685,
+      "learning_rate": 2.3941828128002526e-05,
+      "loss": 0.5989,
+      "step": 1616
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.22532918241424887,
+      "learning_rate": 2.3915390238450346e-05,
+      "loss": 0.5855,
+      "step": 1617
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.2357600500941346,
+      "learning_rate": 2.388895356409162e-05,
+      "loss": 0.8176,
+      "step": 1618
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.23265968547894642,
+      "learning_rate": 2.3862518134545922e-05,
+      "loss": 0.5627,
+      "step": 1619
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.21359935308971753,
+      "learning_rate": 2.3836083979431407e-05,
+      "loss": 0.5706,
+      "step": 1620
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.24150647732116276,
+      "learning_rate": 2.3809651128364834e-05,
+      "loss": 0.8338,
+      "step": 1621
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.2313220790192701,
+      "learning_rate": 2.3783219610961485e-05,
+      "loss": 0.579,
+      "step": 1622
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.21513306566409116,
+      "learning_rate": 2.3756789456835158e-05,
+      "loss": 0.5783,
+      "step": 1623
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.24215821856799488,
+      "learning_rate": 2.373036069559811e-05,
+      "loss": 0.6199,
+      "step": 1624
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.23244924070832532,
+      "learning_rate": 2.3703933356861048e-05,
+      "loss": 0.8156,
+      "step": 1625
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.23614676088967992,
+      "learning_rate": 2.3677507470233072e-05,
+      "loss": 0.5908,
+      "step": 1626
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.2256859966862592,
+      "learning_rate": 2.3651083065321675e-05,
+      "loss": 0.5977,
+      "step": 1627
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.23462520918391655,
+      "learning_rate": 2.3624660171732676e-05,
+      "loss": 0.8535,
+      "step": 1628
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.2293876057314008,
+      "learning_rate": 2.3598238819070202e-05,
+      "loss": 0.5734,
+      "step": 1629
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.24260856402680941,
+      "learning_rate": 2.357181903693667e-05,
+      "loss": 0.628,
+      "step": 1630
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.2177687371871527,
+      "learning_rate": 2.3545400854932698e-05,
+      "loss": 0.5726,
+      "step": 1631
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.22727676053397036,
+      "learning_rate": 2.3518984302657146e-05,
+      "loss": 0.5978,
+      "step": 1632
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.2325110153754206,
+      "learning_rate": 2.3492569409707043e-05,
+      "loss": 0.8198,
+      "step": 1633
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.24180609066055297,
+      "learning_rate": 2.346615620567755e-05,
+      "loss": 0.5424,
+      "step": 1634
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.2431148556011808,
+      "learning_rate": 2.3439744720161942e-05,
+      "loss": 0.6079,
+      "step": 1635
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.24804914421978969,
+      "learning_rate": 2.341333498275157e-05,
+      "loss": 0.8067,
+      "step": 1636
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.230563899450472,
+      "learning_rate": 2.338692702303581e-05,
+      "loss": 0.5791,
+      "step": 1637
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.22555724536748886,
+      "learning_rate": 2.3360520870602072e-05,
+      "loss": 0.59,
+      "step": 1638
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.24144628386979378,
+      "learning_rate": 2.3334116555035724e-05,
+      "loss": 0.8185,
+      "step": 1639
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.262927508455714,
+      "learning_rate": 2.330771410592008e-05,
+      "loss": 0.5873,
+      "step": 1640
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.24015395091742592,
+      "learning_rate": 2.328131355283637e-05,
+      "loss": 0.6337,
+      "step": 1641
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.2549047554688796,
+      "learning_rate": 2.3254914925363676e-05,
+      "loss": 0.584,
+      "step": 1642
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.21646470165136894,
+      "learning_rate": 2.3228518253078952e-05,
+      "loss": 0.5796,
+      "step": 1643
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.24414162353438182,
+      "learning_rate": 2.3202123565556942e-05,
+      "loss": 0.7967,
+      "step": 1644
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.24603177254231343,
+      "learning_rate": 2.317573089237018e-05,
+      "loss": 0.6001,
+      "step": 1645
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.21995188119777967,
+      "learning_rate": 2.314934026308893e-05,
+      "loss": 0.5785,
+      "step": 1646
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.23835568829947798,
+      "learning_rate": 2.312295170728118e-05,
+      "loss": 0.8402,
+      "step": 1647
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.25193538814199384,
+      "learning_rate": 2.3096565254512577e-05,
+      "loss": 0.5802,
+      "step": 1648
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.2298579388987882,
+      "learning_rate": 2.3070180934346425e-05,
+      "loss": 0.5647,
+      "step": 1649
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.2506770686483704,
+      "learning_rate": 2.304379877634364e-05,
+      "loss": 0.7875,
+      "step": 1650
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.240009369021274,
+      "learning_rate": 2.3017418810062704e-05,
+      "loss": 0.5998,
+      "step": 1651
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.27136098146729903,
+      "learning_rate": 2.2991041065059672e-05,
+      "loss": 0.6533,
+      "step": 1652
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.24210344180458224,
+      "learning_rate": 2.296466557088805e-05,
+      "loss": 0.5525,
+      "step": 1653
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.2462146270097717,
+      "learning_rate": 2.293829235709889e-05,
+      "loss": 0.8171,
+      "step": 1654
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.2663913812049236,
+      "learning_rate": 2.291192145324065e-05,
+      "loss": 0.5769,
+      "step": 1655
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.2285759401404454,
+      "learning_rate": 2.288555288885921e-05,
+      "loss": 0.5726,
+      "step": 1656
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.2868941842160905,
+      "learning_rate": 2.2859186693497832e-05,
+      "loss": 0.5786,
+      "step": 1657
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.24542483205314183,
+      "learning_rate": 2.283282289669712e-05,
+      "loss": 0.8214,
+      "step": 1658
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.2979377980702921,
+      "learning_rate": 2.280646152799499e-05,
+      "loss": 0.5889,
+      "step": 1659
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.24139138229340756,
+      "learning_rate": 2.2780102616926633e-05,
+      "loss": 0.5712,
+      "step": 1660
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.2635411907685015,
+      "learning_rate": 2.27537461930245e-05,
+      "loss": 0.7972,
+      "step": 1661
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.24404282991378143,
+      "learning_rate": 2.272739228581824e-05,
+      "loss": 0.5838,
+      "step": 1662
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.264913320522561,
+      "learning_rate": 2.2701040924834703e-05,
+      "loss": 0.6486,
+      "step": 1663
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.302738913342269,
+      "learning_rate": 2.267469213959787e-05,
+      "loss": 0.5771,
+      "step": 1664
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.2598581069186348,
+      "learning_rate": 2.2648345959628825e-05,
+      "loss": 0.8526,
+      "step": 1665
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.25734677294602676,
+      "learning_rate": 2.2622002414445757e-05,
+      "loss": 0.5855,
+      "step": 1666
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.23392588254306784,
+      "learning_rate": 2.259566153356389e-05,
+      "loss": 0.6065,
+      "step": 1667
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.2546115188033964,
+      "learning_rate": 2.256932334649547e-05,
+      "loss": 0.8402,
+      "step": 1668
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.25043981306095103,
+      "learning_rate": 2.2542987882749727e-05,
+      "loss": 0.5883,
+      "step": 1669
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.22083442493514793,
+      "learning_rate": 2.2516655171832818e-05,
+      "loss": 0.578,
+      "step": 1670
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.2598295267384029,
+      "learning_rate": 2.249032524324784e-05,
+      "loss": 0.8063,
+      "step": 1671
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.24803377020464426,
+      "learning_rate": 2.2463998126494764e-05,
+      "loss": 0.587,
+      "step": 1672
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.23459922023999316,
+      "learning_rate": 2.2437673851070414e-05,
+      "loss": 0.584,
+      "step": 1673
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.24453172803458004,
+      "learning_rate": 2.2411352446468426e-05,
+      "loss": 0.6161,
+      "step": 1674
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.2368240547622479,
+      "learning_rate": 2.2385033942179227e-05,
+      "loss": 0.586,
+      "step": 1675
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.24536405916460782,
+      "learning_rate": 2.235871836768997e-05,
+      "loss": 0.8275,
+      "step": 1676
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.27566212501708587,
+      "learning_rate": 2.2332405752484558e-05,
+      "loss": 0.5942,
+      "step": 1677
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.20730894779421719,
+      "learning_rate": 2.230609612604356e-05,
+      "loss": 0.5788,
+      "step": 1678
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.2526886593475554,
+      "learning_rate": 2.2279789517844204e-05,
+      "loss": 0.8242,
+      "step": 1679
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.24852515091790178,
+      "learning_rate": 2.2253485957360336e-05,
+      "loss": 0.5835,
+      "step": 1680
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.2131471495019849,
+      "learning_rate": 2.2227185474062377e-05,
+      "loss": 0.562,
+      "step": 1681
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.2582442261639104,
+      "learning_rate": 2.2200888097417307e-05,
+      "loss": 0.8226,
+      "step": 1682
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.23668694820986383,
+      "learning_rate": 2.217459385688863e-05,
+      "loss": 0.5859,
+      "step": 1683
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.22570855103072726,
+      "learning_rate": 2.2148302781936336e-05,
+      "loss": 0.5574,
+      "step": 1684
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.24676587210181739,
+      "learning_rate": 2.212201490201686e-05,
+      "loss": 0.6184,
+      "step": 1685
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.24950452936946094,
+      "learning_rate": 2.2095730246583073e-05,
+      "loss": 0.5887,
+      "step": 1686
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.23347224837432473,
+      "learning_rate": 2.20694488450842e-05,
+      "loss": 0.799,
+      "step": 1687
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.2498474047511383,
+      "learning_rate": 2.2043170726965858e-05,
+      "loss": 0.5896,
+      "step": 1688
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.24916141155457647,
+      "learning_rate": 2.201689592166997e-05,
+      "loss": 0.5868,
+      "step": 1689
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.24634192629731463,
+      "learning_rate": 2.199062445863474e-05,
+      "loss": 0.8434,
+      "step": 1690
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.2519734283979241,
+      "learning_rate": 2.196435636729464e-05,
+      "loss": 0.5563,
+      "step": 1691
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.22641599186671288,
+      "learning_rate": 2.1938091677080365e-05,
+      "loss": 0.5609,
+      "step": 1692
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.24420419599034618,
+      "learning_rate": 2.1911830417418783e-05,
+      "loss": 0.8035,
+      "step": 1693
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.280669770851636,
+      "learning_rate": 2.188557261773293e-05,
+      "loss": 0.5844,
+      "step": 1694
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.21471714685955978,
+      "learning_rate": 2.1859318307441967e-05,
+      "loss": 0.5596,
+      "step": 1695
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.24946708216229266,
+      "learning_rate": 2.1833067515961144e-05,
+      "loss": 0.6003,
+      "step": 1696
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.24843618777688933,
+      "learning_rate": 2.1806820272701774e-05,
+      "loss": 0.8217,
+      "step": 1697
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.23100922944110547,
+      "learning_rate": 2.1780576607071173e-05,
+      "loss": 0.5943,
+      "step": 1698
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.22707899854554794,
+      "learning_rate": 2.1754336548472666e-05,
+      "loss": 0.5749,
+      "step": 1699
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.2305226763167618,
+      "learning_rate": 2.172810012630554e-05,
+      "loss": 0.5756,
+      "step": 1700
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.2395766976739678,
+      "learning_rate": 2.1701867369965e-05,
+      "loss": 0.818,
+      "step": 1701
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.2595228121132451,
+      "learning_rate": 2.1675638308842145e-05,
+      "loss": 0.5953,
+      "step": 1702
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.22357686142376598,
+      "learning_rate": 2.1649412972323942e-05,
+      "loss": 0.5658,
+      "step": 1703
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.23830638799187712,
+      "learning_rate": 2.1623191389793167e-05,
+      "loss": 0.8248,
+      "step": 1704
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.265522095885245,
+      "learning_rate": 2.1596973590628406e-05,
+      "loss": 0.5755,
+      "step": 1705
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.20765765615399415,
+      "learning_rate": 2.1570759604203995e-05,
+      "loss": 0.5707,
+      "step": 1706
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.2632316742428065,
+      "learning_rate": 2.154454945989001e-05,
+      "loss": 0.6223,
+      "step": 1707
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.24667322885763088,
+      "learning_rate": 2.151834318705222e-05,
+      "loss": 0.8268,
+      "step": 1708
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.23299162348729047,
+      "learning_rate": 2.1492140815052053e-05,
+      "loss": 0.6036,
+      "step": 1709
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.24334774171099263,
+      "learning_rate": 2.1465942373246554e-05,
+      "loss": 0.5901,
+      "step": 1710
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.24358718302063342,
+      "learning_rate": 2.1439747890988383e-05,
+      "loss": 0.8306,
+      "step": 1711
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.24907152423132897,
+      "learning_rate": 2.141355739762576e-05,
+      "loss": 0.5714,
+      "step": 1712
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.22788502651288692,
+      "learning_rate": 2.1387370922502427e-05,
+      "loss": 0.5658,
+      "step": 1713
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.24517856756918932,
+      "learning_rate": 2.1361188494957647e-05,
+      "loss": 0.7943,
+      "step": 1714
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.2256781659030662,
+      "learning_rate": 2.1335010144326112e-05,
+      "loss": 0.5934,
+      "step": 1715
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.22605807481269982,
+      "learning_rate": 2.1308835899937972e-05,
+      "loss": 0.5976,
+      "step": 1716
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.25498312519512356,
+      "learning_rate": 2.128266579111877e-05,
+      "loss": 0.568,
+      "step": 1717
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.23917040050390717,
+      "learning_rate": 2.1256499847189428e-05,
+      "loss": 0.6237,
+      "step": 1718
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.24217391444571365,
+      "learning_rate": 2.1230338097466167e-05,
+      "loss": 0.8095,
+      "step": 1719
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.23448441373290987,
+      "learning_rate": 2.1204180571260563e-05,
+      "loss": 0.5738,
+      "step": 1720
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.22563783864442993,
+      "learning_rate": 2.117802729787939e-05,
+      "loss": 0.5797,
+      "step": 1721
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.22198216280791302,
+      "learning_rate": 2.1151878306624717e-05,
+      "loss": 0.8045,
+      "step": 1722
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.2633197665010454,
+      "learning_rate": 2.1125733626793793e-05,
+      "loss": 0.5785,
+      "step": 1723
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.21331457349614494,
+      "learning_rate": 2.109959328767903e-05,
+      "loss": 0.5764,
+      "step": 1724
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.2386856292308648,
+      "learning_rate": 2.1073457318567997e-05,
+      "loss": 0.824,
+      "step": 1725
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.2441428996551371,
+      "learning_rate": 2.1047325748743338e-05,
+      "loss": 0.5751,
+      "step": 1726
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.21573391513286572,
+      "learning_rate": 2.102119860748279e-05,
+      "loss": 0.56,
+      "step": 1727
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.2465276354071746,
+      "learning_rate": 2.0995075924059125e-05,
+      "loss": 0.8174,
+      "step": 1728
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.24424991972658314,
+      "learning_rate": 2.096895772774011e-05,
+      "loss": 0.3962,
+      "step": 1729
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.24247934297151524,
+      "learning_rate": 2.09428440477885e-05,
+      "loss": 0.8485,
+      "step": 1730
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.2456679133613001,
+      "learning_rate": 2.091673491346199e-05,
+      "loss": 0.5726,
+      "step": 1731
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.23469772040288783,
+      "learning_rate": 2.0890630354013153e-05,
+      "loss": 0.5683,
+      "step": 1732
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.245864797225916,
+      "learning_rate": 2.0864530398689462e-05,
+      "loss": 0.8217,
+      "step": 1733
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.2435582882063739,
+      "learning_rate": 2.0838435076733236e-05,
+      "loss": 0.6146,
+      "step": 1734
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.2395614636624905,
+      "learning_rate": 2.0812344417381595e-05,
+      "loss": 0.5592,
+      "step": 1735
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.22621964954148585,
+      "learning_rate": 2.078625844986642e-05,
+      "loss": 0.561,
+      "step": 1736
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.2298106325581369,
+      "learning_rate": 2.0760177203414368e-05,
+      "loss": 0.7199,
+      "step": 1737
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.23600778217369495,
+      "learning_rate": 2.073410070724677e-05,
+      "loss": 0.5724,
+      "step": 1738
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.23420933584231493,
+      "learning_rate": 2.070802899057966e-05,
+      "loss": 0.8343,
+      "step": 1739
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.24924430311701648,
+      "learning_rate": 2.068196208262371e-05,
+      "loss": 0.6248,
+      "step": 1740
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.22554602731148918,
+      "learning_rate": 2.0655900012584202e-05,
+      "loss": 0.5752,
+      "step": 1741
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.23283234357386054,
+      "learning_rate": 2.0629842809661012e-05,
+      "loss": 0.5836,
+      "step": 1742
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.23288467409295677,
+      "learning_rate": 2.0603790503048527e-05,
+      "loss": 0.8104,
+      "step": 1743
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.2388052220590123,
+      "learning_rate": 2.0577743121935684e-05,
+      "loss": 0.5929,
+      "step": 1744
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.2261349941744784,
+      "learning_rate": 2.0551700695505883e-05,
+      "loss": 0.6153,
+      "step": 1745
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.22563922770024172,
+      "learning_rate": 2.052566325293699e-05,
+      "loss": 0.577,
+      "step": 1746
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.21610185020940678,
+      "learning_rate": 2.0499630823401265e-05,
+      "loss": 0.5824,
+      "step": 1747
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.2362540730414215,
+      "learning_rate": 2.0473603436065373e-05,
+      "loss": 0.8246,
+      "step": 1748
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.23005636416740788,
+      "learning_rate": 2.0447581120090307e-05,
+      "loss": 0.5734,
+      "step": 1749
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.2148959574034927,
+      "learning_rate": 2.0421563904631398e-05,
+      "loss": 0.5873,
+      "step": 1750
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.24730244055601047,
+      "learning_rate": 2.0395551818838247e-05,
+      "loss": 0.8298,
+      "step": 1751
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.22292475208015353,
+      "learning_rate": 2.036954489185472e-05,
+      "loss": 0.5767,
+      "step": 1752
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.23103504166185754,
+      "learning_rate": 2.0343543152818903e-05,
+      "loss": 0.5626,
+      "step": 1753
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.25096956669524717,
+      "learning_rate": 2.031754663086307e-05,
+      "loss": 0.8211,
+      "step": 1754
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.2318304658708562,
+      "learning_rate": 2.0291555355113624e-05,
+      "loss": 0.5758,
+      "step": 1755
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.30531101267094934,
+      "learning_rate": 2.026556935469112e-05,
+      "loss": 0.6161,
+      "step": 1756
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.23280093744988156,
+      "learning_rate": 2.0239588658710187e-05,
+      "loss": 0.582,
+      "step": 1757
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.22656858077381242,
+      "learning_rate": 2.0213613296279533e-05,
+      "loss": 0.5766,
+      "step": 1758
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.23205752788745168,
+      "learning_rate": 2.0187643296501878e-05,
+      "loss": 0.8206,
+      "step": 1759
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.23235173638027568,
+      "learning_rate": 2.0161678688473902e-05,
+      "loss": 0.5709,
+      "step": 1760
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.21731100557963798,
+      "learning_rate": 2.0135719501286292e-05,
+      "loss": 0.5712,
+      "step": 1761
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.25402664760278465,
+      "learning_rate": 2.0109765764023636e-05,
+      "loss": 0.7901,
+      "step": 1762
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.2256493259506461,
+      "learning_rate": 2.0083817505764425e-05,
+      "loss": 0.5673,
+      "step": 1763
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.2155998514149159,
+      "learning_rate": 2.0057874755581008e-05,
+      "loss": 0.5893,
+      "step": 1764
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.23308361558888221,
+      "learning_rate": 2.003193754253957e-05,
+      "loss": 0.8166,
+      "step": 1765
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.22726874907572892,
+      "learning_rate": 2.0006005895700067e-05,
+      "loss": 0.5812,
+      "step": 1766
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.22256711425421516,
+      "learning_rate": 1.9980079844116238e-05,
+      "loss": 0.6248,
+      "step": 1767
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.22140437217081324,
+      "learning_rate": 1.995415941683557e-05,
+      "loss": 0.5726,
+      "step": 1768
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.231419527938471,
+      "learning_rate": 1.9928244642899217e-05,
+      "loss": 0.7993,
+      "step": 1769
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.25299659689506,
+      "learning_rate": 1.9902335551342026e-05,
+      "loss": 0.5672,
+      "step": 1770
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.2035576083475669,
+      "learning_rate": 1.9876432171192442e-05,
+      "loss": 0.5797,
+      "step": 1771
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.23822602122798173,
+      "learning_rate": 1.9850534531472546e-05,
+      "loss": 0.5631,
+      "step": 1772
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.2382217063774644,
+      "learning_rate": 1.982464266119797e-05,
+      "loss": 0.8405,
+      "step": 1773
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.2142034807969817,
+      "learning_rate": 1.979875658937789e-05,
+      "loss": 0.5681,
+      "step": 1774
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.23015639873980806,
+      "learning_rate": 1.9772876345014974e-05,
+      "loss": 0.5701,
+      "step": 1775
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.22979896826526489,
+      "learning_rate": 1.9747001957105382e-05,
+      "loss": 0.8177,
+      "step": 1776
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.22185112103550447,
+      "learning_rate": 1.9721133454638684e-05,
+      "loss": 0.5782,
+      "step": 1777
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.231020108769617,
+      "learning_rate": 1.969527086659788e-05,
+      "loss": 0.5968,
+      "step": 1778
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.2327645388444387,
+      "learning_rate": 1.9669414221959332e-05,
+      "loss": 0.5819,
+      "step": 1779
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.22551361670785167,
+      "learning_rate": 1.9643563549692743e-05,
+      "loss": 0.8103,
+      "step": 1780
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.23500225916578896,
+      "learning_rate": 1.961771887876113e-05,
+      "loss": 0.5752,
+      "step": 1781
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.20806700900195296,
+      "learning_rate": 1.9591880238120792e-05,
+      "loss": 0.5491,
+      "step": 1782
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.23936571508262725,
+      "learning_rate": 1.9566047656721243e-05,
+      "loss": 0.8226,
+      "step": 1783
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.24147160575602133,
+      "learning_rate": 1.954022116350524e-05,
+      "loss": 0.59,
+      "step": 1784
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.2341536283399213,
+      "learning_rate": 1.9514400787408703e-05,
+      "loss": 0.5637,
+      "step": 1785
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.24019681804405252,
+      "learning_rate": 1.9488586557360703e-05,
+      "loss": 0.7891,
+      "step": 1786
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.23056809485133034,
+      "learning_rate": 1.946277850228343e-05,
+      "loss": 0.5764,
+      "step": 1787
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.2174799683542844,
+      "learning_rate": 1.9436976651092144e-05,
+      "loss": 0.5676,
+      "step": 1788
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.24435162187545903,
+      "learning_rate": 1.941118103269516e-05,
+      "loss": 0.6277,
+      "step": 1789
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.2443091722766792,
+      "learning_rate": 1.9385391675993805e-05,
+      "loss": 0.5824,
+      "step": 1790
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.23451638753783074,
+      "learning_rate": 1.9359608609882403e-05,
+      "loss": 0.7954,
+      "step": 1791
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.2264494561620157,
+      "learning_rate": 1.9333831863248227e-05,
+      "loss": 0.5804,
+      "step": 1792
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.23084771692726086,
+      "learning_rate": 1.930806146497146e-05,
+      "loss": 0.5801,
+      "step": 1793
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.23178815677020337,
+      "learning_rate": 1.928229744392517e-05,
+      "loss": 0.825,
+      "step": 1794
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.23151597385908043,
+      "learning_rate": 1.925653982897529e-05,
+      "loss": 0.5842,
+      "step": 1795
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.2094239608849007,
+      "learning_rate": 1.9230788648980574e-05,
+      "loss": 0.5698,
+      "step": 1796
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.23104966109045097,
+      "learning_rate": 1.9205043932792564e-05,
+      "loss": 0.8168,
+      "step": 1797
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.23937582983623681,
+      "learning_rate": 1.917930570925557e-05,
+      "loss": 0.5881,
+      "step": 1798
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.21487123800131888,
+      "learning_rate": 1.9153574007206615e-05,
+      "loss": 0.5746,
+      "step": 1799
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.24356741917021435,
+      "learning_rate": 1.912784885547541e-05,
+      "loss": 0.6203,
+      "step": 1800
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.217036672270838,
+      "learning_rate": 1.9102130282884342e-05,
+      "loss": 0.582,
+      "step": 1801
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.23631732605092537,
+      "learning_rate": 1.9076418318248424e-05,
+      "loss": 0.8471,
+      "step": 1802
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.2326674842433094,
+      "learning_rate": 1.9050712990375255e-05,
+      "loss": 0.5516,
+      "step": 1803
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.21906550528731397,
+      "learning_rate": 1.9025014328065022e-05,
+      "loss": 0.5618,
+      "step": 1804
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.23131298619116344,
+      "learning_rate": 1.8999322360110406e-05,
+      "loss": 0.7844,
+      "step": 1805
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.2548359119468388,
+      "learning_rate": 1.897363711529662e-05,
+      "loss": 0.5798,
+      "step": 1806
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.22923432508577474,
+      "learning_rate": 1.8947958622401328e-05,
+      "loss": 0.5793,
+      "step": 1807
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.23867795675845793,
+      "learning_rate": 1.892228691019464e-05,
+      "loss": 0.8433,
+      "step": 1808
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.2426429648400533,
+      "learning_rate": 1.8896622007439065e-05,
+      "loss": 0.5845,
+      "step": 1809
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.22300492814466868,
+      "learning_rate": 1.8870963942889485e-05,
+      "loss": 0.5633,
+      "step": 1810
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.23882520978468674,
+      "learning_rate": 1.8845312745293103e-05,
+      "loss": 0.6458,
+      "step": 1811
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.24908139952488695,
+      "learning_rate": 1.881966844338945e-05,
+      "loss": 0.8248,
+      "step": 1812
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.2200426750700071,
+      "learning_rate": 1.8794031065910322e-05,
+      "loss": 0.6119,
+      "step": 1813
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.22342438748732268,
+      "learning_rate": 1.876840064157976e-05,
+      "loss": 0.5794,
+      "step": 1814
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.2188015409807883,
+      "learning_rate": 1.8742777199114027e-05,
+      "loss": 0.5633,
+      "step": 1815
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.25014575835224473,
+      "learning_rate": 1.871716076722152e-05,
+      "loss": 0.8145,
+      "step": 1816
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.25283128064485527,
+      "learning_rate": 1.8691551374602823e-05,
+      "loss": 0.5661,
+      "step": 1817
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.22374235385173688,
+      "learning_rate": 1.8665949049950625e-05,
+      "loss": 0.5618,
+      "step": 1818
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.2378077528791449,
+      "learning_rate": 1.864035382194969e-05,
+      "loss": 0.8299,
+      "step": 1819
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.22990439102627336,
+      "learning_rate": 1.8614765719276832e-05,
+      "loss": 0.5841,
+      "step": 1820
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.23037099933172184,
+      "learning_rate": 1.8589184770600893e-05,
+      "loss": 0.5697,
+      "step": 1821
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.2307605521892731,
+      "learning_rate": 1.856361100458267e-05,
+      "loss": 0.615,
+      "step": 1822
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.2327048667062388,
+      "learning_rate": 1.8538044449874948e-05,
+      "loss": 0.8221,
+      "step": 1823
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.254608480089308,
+      "learning_rate": 1.8512485135122404e-05,
+      "loss": 0.5549,
+      "step": 1824
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.22951820801501566,
+      "learning_rate": 1.848693308896163e-05,
+      "loss": 0.594,
+      "step": 1825
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.24536392571181348,
+      "learning_rate": 1.8461388340021048e-05,
+      "loss": 0.7882,
+      "step": 1826
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.241142075372955,
+      "learning_rate": 1.8435850916920932e-05,
+      "loss": 0.558,
+      "step": 1827
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.23062132788595896,
+      "learning_rate": 1.8410320848273315e-05,
+      "loss": 0.5687,
+      "step": 1828
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.24140879760231693,
+      "learning_rate": 1.838479816268201e-05,
+      "loss": 0.8304,
+      "step": 1829
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.23191828977100115,
+      "learning_rate": 1.8359282888742564e-05,
+      "loss": 0.5691,
+      "step": 1830
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.23471449394167265,
+      "learning_rate": 1.8333775055042205e-05,
+      "loss": 0.5684,
+      "step": 1831
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.2402344014117845,
+      "learning_rate": 1.830827469015984e-05,
+      "loss": 0.8163,
+      "step": 1832
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.23015181317204053,
+      "learning_rate": 1.828278182266599e-05,
+      "loss": 0.3939,
+      "step": 1833
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.2338386914160794,
+      "learning_rate": 1.825729648112279e-05,
+      "loss": 0.8362,
+      "step": 1834
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.23369287653253834,
+      "learning_rate": 1.8231818694083937e-05,
+      "loss": 0.585,
+      "step": 1835
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.21420080931792182,
+      "learning_rate": 1.820634849009467e-05,
+      "loss": 0.5787,
+      "step": 1836
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.23455180837467554,
+      "learning_rate": 1.818088589769172e-05,
+      "loss": 0.8067,
+      "step": 1837
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.23984782125689239,
+      "learning_rate": 1.8155430945403316e-05,
+      "loss": 0.5699,
+      "step": 1838
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.2319707529713084,
+      "learning_rate": 1.812998366174908e-05,
+      "loss": 0.6045,
+      "step": 1839
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.2191019501651141,
+      "learning_rate": 1.8104544075240086e-05,
+      "loss": 0.7119,
+      "step": 1840
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.22715066005047116,
+      "learning_rate": 1.8079112214378768e-05,
+      "loss": 0.5566,
+      "step": 1841
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.2156728848443151,
+      "learning_rate": 1.8053688107658908e-05,
+      "loss": 0.5672,
+      "step": 1842
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.23573655329297394,
+      "learning_rate": 1.8028271783565587e-05,
+      "loss": 0.8097,
+      "step": 1843
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.25198821866965854,
+      "learning_rate": 1.8002863270575187e-05,
+      "loss": 0.3919,
+      "step": 1844
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.2419103431089826,
+      "learning_rate": 1.7977462597155326e-05,
+      "loss": 0.841,
+      "step": 1845
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.2410031338886957,
+      "learning_rate": 1.7952069791764836e-05,
+      "loss": 0.5586,
+      "step": 1846
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.22029715317375168,
+      "learning_rate": 1.792668488285374e-05,
+      "loss": 0.5606,
+      "step": 1847
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.28577863027359685,
+      "learning_rate": 1.7901307898863212e-05,
+      "loss": 0.7868,
+      "step": 1848
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.23721650686372459,
+      "learning_rate": 1.7875938868225563e-05,
+      "loss": 0.5799,
+      "step": 1849
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.23677459838615414,
+      "learning_rate": 1.785057781936415e-05,
+      "loss": 0.6078,
+      "step": 1850
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.22338269948252773,
+      "learning_rate": 1.7825224780693427e-05,
+      "loss": 0.5704,
+      "step": 1851
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.2318844501007659,
+      "learning_rate": 1.779987978061886e-05,
+      "loss": 0.8245,
+      "step": 1852
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.24619892483820724,
+      "learning_rate": 1.777454284753691e-05,
+      "loss": 0.5681,
+      "step": 1853
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.22006892455312926,
+      "learning_rate": 1.7749214009835e-05,
+      "loss": 0.7378,
+      "step": 1854
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.2510718938892098,
+      "learning_rate": 1.7723893295891488e-05,
+      "loss": 0.604,
+      "step": 1855
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.238294152834878,
+      "learning_rate": 1.769858073407561e-05,
+      "loss": 0.5793,
+      "step": 1856
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.23394097290507823,
+      "learning_rate": 1.7673276352747488e-05,
+      "loss": 0.5721,
+      "step": 1857
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.2341722690536227,
+      "learning_rate": 1.7647980180258077e-05,
+      "loss": 0.8179,
+      "step": 1858
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.220649592444817,
+      "learning_rate": 1.762269224494913e-05,
+      "loss": 0.5969,
+      "step": 1859
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.22404355793257172,
+      "learning_rate": 1.7597412575153184e-05,
+      "loss": 0.5838,
+      "step": 1860
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.24271191470711811,
+      "learning_rate": 1.7572141199193475e-05,
+      "loss": 0.6369,
+      "step": 1861
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.2224256184115036,
+      "learning_rate": 1.7546878145383996e-05,
+      "loss": 0.5782,
+      "step": 1862
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.24254618604623293,
+      "learning_rate": 1.752162344202939e-05,
+      "loss": 0.839,
+      "step": 1863
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.23021959480035237,
+      "learning_rate": 1.749637711742495e-05,
+      "loss": 0.5579,
+      "step": 1864
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.21519673294920505,
+      "learning_rate": 1.747113919985658e-05,
+      "loss": 0.5939,
+      "step": 1865
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.2761953428233501,
+      "learning_rate": 1.744590971760077e-05,
+      "loss": 0.8156,
+      "step": 1866
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.22337343927257428,
+      "learning_rate": 1.7420688698924542e-05,
+      "loss": 0.5818,
+      "step": 1867
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.22043125015474171,
+      "learning_rate": 1.739547617208546e-05,
+      "loss": 0.5823,
+      "step": 1868
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.25181597582226223,
+      "learning_rate": 1.7370272165331557e-05,
+      "loss": 0.8303,
+      "step": 1869
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.2176649344450055,
+      "learning_rate": 1.734507670690133e-05,
+      "loss": 0.5552,
+      "step": 1870
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.21854846081640533,
+      "learning_rate": 1.7319889825023684e-05,
+      "loss": 0.5703,
+      "step": 1871
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8236369850356001,
+      "learning_rate": 1.7294711547917942e-05,
+      "loss": 0.6382,
+      "step": 1872
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.22280324409461122,
+      "learning_rate": 1.7269541903793747e-05,
+      "loss": 0.5536,
+      "step": 1873
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.2486954696305139,
+      "learning_rate": 1.7244380920851095e-05,
+      "loss": 0.8457,
+      "step": 1874
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.2310436325092519,
+      "learning_rate": 1.721922862728028e-05,
+      "loss": 0.5796,
+      "step": 1875
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.21895360171646658,
+      "learning_rate": 1.719408505126186e-05,
+      "loss": 0.5544,
+      "step": 1876
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.2536684202857725,
+      "learning_rate": 1.7168950220966614e-05,
+      "loss": 0.8241,
+      "step": 1877
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.2126917992576458,
+      "learning_rate": 1.714382416455553e-05,
+      "loss": 0.5583,
+      "step": 1878
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.23171760483840734,
+      "learning_rate": 1.7118706910179754e-05,
+      "loss": 0.5878,
+      "step": 1879
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.23260047092637146,
+      "learning_rate": 1.7093598485980593e-05,
+      "loss": 0.8326,
+      "step": 1880
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.2452958882355137,
+      "learning_rate": 1.706849892008945e-05,
+      "loss": 0.5814,
+      "step": 1881
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.222744931096726,
+      "learning_rate": 1.70434082406278e-05,
+      "loss": 0.551,
+      "step": 1882
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.25306713960003174,
+      "learning_rate": 1.7018326475707165e-05,
+      "loss": 0.6045,
+      "step": 1883
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.25018456003661094,
+      "learning_rate": 1.6993253653429063e-05,
+      "loss": 0.8229,
+      "step": 1884
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.26787082614360563,
+      "learning_rate": 1.6968189801885022e-05,
+      "loss": 0.5746,
+      "step": 1885
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.22603779033409413,
+      "learning_rate": 1.6943134949156498e-05,
+      "loss": 0.5652,
+      "step": 1886
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.24762782292462504,
+      "learning_rate": 1.6918089123314875e-05,
+      "loss": 0.5643,
+      "step": 1887
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.2535610572091777,
+      "learning_rate": 1.689305235242142e-05,
+      "loss": 0.8486,
+      "step": 1888
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.2276977756565606,
+      "learning_rate": 1.6868024664527248e-05,
+      "loss": 0.5775,
+      "step": 1889
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.22691449070691974,
+      "learning_rate": 1.6843006087673308e-05,
+      "loss": 0.5686,
+      "step": 1890
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.2365509995497305,
+      "learning_rate": 1.681799664989033e-05,
+      "loss": 0.8161,
+      "step": 1891
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.2350762322276925,
+      "learning_rate": 1.6792996379198828e-05,
+      "loss": 0.5873,
+      "step": 1892
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.21161273443782583,
+      "learning_rate": 1.6768005303609016e-05,
+      "loss": 0.5621,
+      "step": 1893
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8681307773717777,
+      "learning_rate": 1.6743023451120832e-05,
+      "loss": 0.6703,
+      "step": 1894
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.23487653564824093,
+      "learning_rate": 1.6718050849723844e-05,
+      "loss": 0.8432,
+      "step": 1895
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.213122326598215,
+      "learning_rate": 1.6693087527397288e-05,
+      "loss": 0.58,
+      "step": 1896
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.22249702691644369,
+      "learning_rate": 1.666813351211e-05,
+      "loss": 0.5763,
+      "step": 1897
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.2266353963454185,
+      "learning_rate": 1.6643188831820375e-05,
+      "loss": 0.7983,
+      "step": 1898
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.22083128218482373,
+      "learning_rate": 1.661825351447636e-05,
+      "loss": 0.555,
+      "step": 1899
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.2163951013195406,
+      "learning_rate": 1.6593327588015415e-05,
+      "loss": 0.583,
+      "step": 1900
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.23093232532716912,
+      "learning_rate": 1.6568411080364455e-05,
+      "loss": 0.8092,
+      "step": 1901
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.2555315956386295,
+      "learning_rate": 1.654350401943987e-05,
+      "loss": 0.5755,
+      "step": 1902
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.21173269565678976,
+      "learning_rate": 1.6518606433147453e-05,
+      "loss": 0.563,
+      "step": 1903
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.21529216785838604,
+      "learning_rate": 1.6493718349382382e-05,
+      "loss": 0.559,
+      "step": 1904
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.23139122831850242,
+      "learning_rate": 1.64688397960292e-05,
+      "loss": 0.6077,
+      "step": 1905
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.233926395624559,
+      "learning_rate": 1.6443970800961743e-05,
+      "loss": 0.7948,
+      "step": 1906
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.22155460149332531,
+      "learning_rate": 1.641911139204317e-05,
+      "loss": 0.5557,
+      "step": 1907
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.21500767436780052,
+      "learning_rate": 1.6394261597125878e-05,
+      "loss": 0.5796,
+      "step": 1908
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.24258027328394904,
+      "learning_rate": 1.6369421444051507e-05,
+      "loss": 0.8241,
+      "step": 1909
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.22823111726608364,
+      "learning_rate": 1.634459096065089e-05,
+      "loss": 0.5849,
+      "step": 1910
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.22402758377523915,
+      "learning_rate": 1.6319770174744022e-05,
+      "loss": 0.5631,
+      "step": 1911
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.2304388120067028,
+      "learning_rate": 1.6294959114140034e-05,
+      "loss": 0.7903,
+      "step": 1912
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.2126423711148029,
+      "learning_rate": 1.627015780663716e-05,
+      "loss": 0.5688,
+      "step": 1913
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.2196039502113604,
+      "learning_rate": 1.6245366280022718e-05,
+      "loss": 0.5627,
+      "step": 1914
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.24675819927998954,
+      "learning_rate": 1.622058456207305e-05,
+      "loss": 0.8303,
+      "step": 1915
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.22808504253369,
+      "learning_rate": 1.6195812680553515e-05,
+      "loss": 0.3813,
+      "step": 1916
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.2266049988742338,
+      "learning_rate": 1.6171050663218475e-05,
+      "loss": 0.8209,
+      "step": 1917
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.21389209082523458,
+      "learning_rate": 1.6146298537811182e-05,
+      "loss": 0.5641,
+      "step": 1918
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.21128389451560498,
+      "learning_rate": 1.6121556332063862e-05,
+      "loss": 0.5811,
+      "step": 1919
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.2348380870398043,
+      "learning_rate": 1.609682407369761e-05,
+      "loss": 0.792,
+      "step": 1920
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.2269815849577964,
+      "learning_rate": 1.607210179042237e-05,
+      "loss": 0.5684,
+      "step": 1921
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.20563857474930514,
+      "learning_rate": 1.6047389509936906e-05,
+      "loss": 0.5641,
+      "step": 1922
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.23649215199278523,
+      "learning_rate": 1.6022687259928783e-05,
+      "loss": 0.8119,
+      "step": 1923
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.21736339716334646,
+      "learning_rate": 1.5997995068074328e-05,
+      "loss": 0.5393,
+      "step": 1924
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.2139904026074026,
+      "learning_rate": 1.5973312962038596e-05,
+      "loss": 0.5653,
+      "step": 1925
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.23317726630601016,
+      "learning_rate": 1.5948640969475346e-05,
+      "loss": 0.8202,
+      "step": 1926
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.23400876747760127,
+      "learning_rate": 1.5923979118027e-05,
+      "loss": 0.6038,
+      "step": 1927
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.23211330466270205,
+      "learning_rate": 1.589932743532463e-05,
+      "loss": 0.5704,
+      "step": 1928
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.21218308425747628,
+      "learning_rate": 1.587468594898789e-05,
+      "loss": 0.5738,
+      "step": 1929
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.2366195460371928,
+      "learning_rate": 1.5850054686625033e-05,
+      "loss": 0.8292,
+      "step": 1930
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.2397109690537746,
+      "learning_rate": 1.5825433675832846e-05,
+      "loss": 0.5894,
+      "step": 1931
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.22480597816326736,
+      "learning_rate": 1.580082294419664e-05,
+      "loss": 0.5906,
+      "step": 1932
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.21408663108527978,
+      "learning_rate": 1.5776222519290208e-05,
+      "loss": 0.5834,
+      "step": 1933
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.23524169481898544,
+      "learning_rate": 1.5751632428675773e-05,
+      "loss": 0.8011,
+      "step": 1934
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.22814744279071666,
+      "learning_rate": 1.572705269990401e-05,
+      "loss": 0.5561,
+      "step": 1935
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.21937817289821274,
+      "learning_rate": 1.5702483360513974e-05,
+      "loss": 0.5853,
+      "step": 1936
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.2332389391810761,
+      "learning_rate": 1.567792443803307e-05,
+      "loss": 0.7785,
+      "step": 1937
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.24365802878116943,
+      "learning_rate": 1.565337595997704e-05,
+      "loss": 0.6236,
+      "step": 1938
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.23828800548059767,
+      "learning_rate": 1.562883795384994e-05,
+      "loss": 0.5709,
+      "step": 1939
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.21019496940509202,
+      "learning_rate": 1.560431044714405e-05,
+      "loss": 0.5805,
+      "step": 1940
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.23177995261026213,
+      "learning_rate": 1.5579793467339933e-05,
+      "loss": 0.8164,
+      "step": 1941
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.24403436125691574,
+      "learning_rate": 1.5555287041906326e-05,
+      "loss": 0.5453,
+      "step": 1942
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.2139720585291596,
+      "learning_rate": 1.553079119830015e-05,
+      "loss": 0.569,
+      "step": 1943
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.23571531766628218,
+      "learning_rate": 1.5506305963966487e-05,
+      "loss": 0.7886,
+      "step": 1944
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.23759761876895405,
+      "learning_rate": 1.5481831366338527e-05,
+      "loss": 0.5804,
+      "step": 1945
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.2295006244628689,
+      "learning_rate": 1.545736743283751e-05,
+      "loss": 0.5709,
+      "step": 1946
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.2391450476415523,
+      "learning_rate": 1.5432914190872757e-05,
+      "loss": 0.8389,
+      "step": 1947
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.22268346427599575,
+      "learning_rate": 1.5408471667841613e-05,
+      "loss": 0.5665,
+      "step": 1948
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.24362764612511228,
+      "learning_rate": 1.5384039891129405e-05,
+      "loss": 0.6334,
+      "step": 1949
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.20943810777046717,
+      "learning_rate": 1.5359618888109424e-05,
+      "loss": 0.57,
+      "step": 1950
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.2056843858477842,
+      "learning_rate": 1.5335208686142878e-05,
+      "loss": 0.568,
+      "step": 1951
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.23265393126168968,
+      "learning_rate": 1.5310809312578875e-05,
+      "loss": 0.8283,
+      "step": 1952
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.21941672023235806,
+      "learning_rate": 1.528642079475442e-05,
+      "loss": 0.5764,
+      "step": 1953
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.2342267881739272,
+      "learning_rate": 1.5262043159994317e-05,
+      "loss": 0.5983,
+      "step": 1954
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.20427360813234927,
+      "learning_rate": 1.5237676435611204e-05,
+      "loss": 0.5731,
+      "step": 1955
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.2440685372090404,
+      "learning_rate": 1.5213320648905488e-05,
+      "loss": 0.8039,
+      "step": 1956
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.20479417654240065,
+      "learning_rate": 1.5188975827165298e-05,
+      "loss": 0.5597,
+      "step": 1957
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.23681088478567833,
+      "learning_rate": 1.516464199766651e-05,
+      "loss": 0.8562,
+      "step": 1958
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.2382598937589917,
+      "learning_rate": 1.5140319187672669e-05,
+      "loss": 0.5587,
+      "step": 1959
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.23288191267523498,
+      "learning_rate": 1.511600742443498e-05,
+      "loss": 0.5943,
+      "step": 1960
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.22757236495750371,
+      "learning_rate": 1.5091706735192268e-05,
+      "loss": 0.5669,
+      "step": 1961
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.204979951425395,
+      "learning_rate": 1.5067417147170954e-05,
+      "loss": 0.5477,
+      "step": 1962
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.22980517091840566,
+      "learning_rate": 1.5043138687585006e-05,
+      "loss": 0.7984,
+      "step": 1963
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.22889544972031195,
+      "learning_rate": 1.5018871383635947e-05,
+      "loss": 0.565,
+      "step": 1964
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.24272296287280956,
+      "learning_rate": 1.4994615262512784e-05,
+      "loss": 0.5951,
+      "step": 1965
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.23229194859519958,
+      "learning_rate": 1.4970370351392002e-05,
+      "loss": 0.5602,
+      "step": 1966
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.24731816589404523,
+      "learning_rate": 1.4946136677437536e-05,
+      "loss": 0.7942,
+      "step": 1967
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.22310661756097755,
+      "learning_rate": 1.49219142678007e-05,
+      "loss": 0.5722,
+      "step": 1968
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.214879451285188,
+      "learning_rate": 1.4897703149620218e-05,
+      "loss": 0.5607,
+      "step": 1969
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.23805215366847501,
+      "learning_rate": 1.487350335002215e-05,
+      "loss": 0.7991,
+      "step": 1970
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.23344422438661325,
+      "learning_rate": 1.4849314896119879e-05,
+      "loss": 0.5576,
+      "step": 1971
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.20695547514639728,
+      "learning_rate": 1.4825137815014073e-05,
+      "loss": 0.5553,
+      "step": 1972
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.2313972133670885,
+      "learning_rate": 1.480097213379267e-05,
+      "loss": 0.809,
+      "step": 1973
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.23633262862531138,
+      "learning_rate": 1.477681787953081e-05,
+      "loss": 0.5679,
+      "step": 1974
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.20478893857312802,
+      "learning_rate": 1.4752675079290851e-05,
+      "loss": 0.55,
+      "step": 1975
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.2357168115371893,
+      "learning_rate": 1.4728543760122316e-05,
+      "loss": 0.629,
+      "step": 1976
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.2215792744612389,
+      "learning_rate": 1.4704423949061853e-05,
+      "loss": 0.5761,
+      "step": 1977
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.981555534644714,
+      "learning_rate": 1.4680315673133243e-05,
+      "loss": 0.8773,
+      "step": 1978
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.21984015120188868,
+      "learning_rate": 1.4656218959347302e-05,
+      "loss": 0.5759,
+      "step": 1979
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.21468524960406762,
+      "learning_rate": 1.463213383470192e-05,
+      "loss": 0.5695,
+      "step": 1980
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.23740466817422073,
+      "learning_rate": 1.4608060326181999e-05,
+      "loss": 0.8376,
+      "step": 1981
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.23467993721056218,
+      "learning_rate": 1.4583998460759424e-05,
+      "loss": 0.5589,
+      "step": 1982
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.20858887925857916,
+      "learning_rate": 1.4559948265393036e-05,
+      "loss": 0.5644,
+      "step": 1983
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.24039381959253642,
+      "learning_rate": 1.4535909767028616e-05,
+      "loss": 0.7543,
+      "step": 1984
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.21856638648262458,
+      "learning_rate": 1.4511882992598796e-05,
+      "loss": 0.5724,
+      "step": 1985
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.19635586708988303,
+      "learning_rate": 1.448786796902311e-05,
+      "loss": 0.5721,
+      "step": 1986
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.23339686347264654,
+      "learning_rate": 1.4463864723207927e-05,
+      "loss": 0.6007,
+      "step": 1987
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.21930058627416704,
+      "learning_rate": 1.4439873282046396e-05,
+      "loss": 0.597,
+      "step": 1988
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.23338553253327782,
+      "learning_rate": 1.4415893672418462e-05,
+      "loss": 0.8157,
+      "step": 1989
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.2637362796669564,
+      "learning_rate": 1.4391925921190816e-05,
+      "loss": 0.5667,
+      "step": 1990
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.22338329454786793,
+      "learning_rate": 1.4367970055216828e-05,
+      "loss": 0.5726,
+      "step": 1991
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.23405214778510988,
+      "learning_rate": 1.4344026101336594e-05,
+      "loss": 0.815,
+      "step": 1992
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.246012268004688,
+      "learning_rate": 1.4320094086376844e-05,
+      "loss": 0.5719,
+      "step": 1993
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.2081582601173589,
+      "learning_rate": 1.429617403715093e-05,
+      "loss": 0.5877,
+      "step": 1994
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.23942801548167816,
+      "learning_rate": 1.4272265980458815e-05,
+      "loss": 0.7958,
+      "step": 1995
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.23808249690279606,
+      "learning_rate": 1.4248369943086998e-05,
+      "loss": 0.5736,
+      "step": 1996
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.21640653519962674,
+      "learning_rate": 1.4224485951808525e-05,
+      "loss": 0.5816,
+      "step": 1997
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.2435077779956155,
+      "learning_rate": 1.4200614033382953e-05,
+      "loss": 0.6286,
+      "step": 1998
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.23260126072665832,
+      "learning_rate": 1.4176754214556304e-05,
+      "loss": 0.8413,
+      "step": 1999
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.21700528191758708,
+      "learning_rate": 1.4152906522061048e-05,
+      "loss": 0.5683,
+      "step": 2000
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.20933402757939315,
+      "learning_rate": 1.4129070982616066e-05,
+      "loss": 0.5977,
+      "step": 2001
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.21195269562378416,
+      "learning_rate": 1.4105247622926625e-05,
+      "loss": 0.5772,
+      "step": 2002
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.23392127576716573,
+      "learning_rate": 1.4081436469684339e-05,
+      "loss": 0.833,
+      "step": 2003
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.21904002956531132,
+      "learning_rate": 1.4057637549567153e-05,
+      "loss": 0.5732,
+      "step": 2004
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.21904754001762705,
+      "learning_rate": 1.4033850889239308e-05,
+      "loss": 0.5717,
+      "step": 2005
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.2224924649514495,
+      "learning_rate": 1.4010076515351306e-05,
+      "loss": 0.8031,
+      "step": 2006
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.21868757813064432,
+      "learning_rate": 1.3986314454539884e-05,
+      "loss": 0.5782,
+      "step": 2007
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.21566359567902468,
+      "learning_rate": 1.3962564733427969e-05,
+      "loss": 0.5827,
+      "step": 2008
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.24297503555832045,
+      "learning_rate": 1.3938827378624681e-05,
+      "loss": 0.5984,
+      "step": 2009
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.23462978200514717,
+      "learning_rate": 1.3915102416725287e-05,
+      "loss": 0.8445,
+      "step": 2010
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.2367025934930114,
+      "learning_rate": 1.3891389874311156e-05,
+      "loss": 0.5612,
+      "step": 2011
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.21826679136901064,
+      "learning_rate": 1.3867689777949754e-05,
+      "loss": 0.5909,
+      "step": 2012
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.2593336649715566,
+      "learning_rate": 1.3844002154194588e-05,
+      "loss": 0.7932,
+      "step": 2013
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.21067326550984192,
+      "learning_rate": 1.38203270295852e-05,
+      "loss": 0.5527,
+      "step": 2014
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.2147540906453524,
+      "learning_rate": 1.3796664430647127e-05,
+      "loss": 0.5665,
+      "step": 2015
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.22972534041577566,
+      "learning_rate": 1.3773014383891874e-05,
+      "loss": 0.8265,
+      "step": 2016
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.22455950523273155,
+      "learning_rate": 1.3749376915816886e-05,
+      "loss": 0.5853,
+      "step": 2017
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.2065258160532227,
+      "learning_rate": 1.372575205290551e-05,
+      "loss": 0.5657,
+      "step": 2018
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.2285162382536817,
+      "learning_rate": 1.3702139821626959e-05,
+      "loss": 0.7998,
+      "step": 2019
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.23205093084931352,
+      "learning_rate": 1.367854024843631e-05,
+      "loss": 0.3965,
+      "step": 2020
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.22913462272296078,
+      "learning_rate": 1.3654953359774453e-05,
+      "loss": 0.7897,
+      "step": 2021
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.2418067703262328,
+      "learning_rate": 1.3631379182068066e-05,
+      "loss": 0.5779,
+      "step": 2022
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.21484293062502047,
+      "learning_rate": 1.3607817741729584e-05,
+      "loss": 0.5876,
+      "step": 2023
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.23781508322934491,
+      "learning_rate": 1.3584269065157174e-05,
+      "loss": 0.8284,
+      "step": 2024
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.23806136793938512,
+      "learning_rate": 1.3560733178734703e-05,
+      "loss": 0.5542,
+      "step": 2025
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.2180612367633781,
+      "learning_rate": 1.35372101088317e-05,
+      "loss": 0.5754,
+      "step": 2026
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.24768258242241262,
+      "learning_rate": 1.3513699881803346e-05,
+      "loss": 0.8248,
+      "step": 2027
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.2246069481692399,
+      "learning_rate": 1.3490202523990425e-05,
+      "loss": 0.5574,
+      "step": 2028
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.21205710007592557,
+      "learning_rate": 1.3466718061719317e-05,
+      "loss": 0.597,
+      "step": 2029
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.24011580926444134,
+      "learning_rate": 1.3443246521301917e-05,
+      "loss": 0.8229,
+      "step": 2030
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.23085396346988582,
+      "learning_rate": 1.3419787929035682e-05,
+      "loss": 0.3898,
+      "step": 2031
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.23767771964981563,
+      "learning_rate": 1.3396342311203541e-05,
+      "loss": 0.8238,
+      "step": 2032
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.22442143815506765,
+      "learning_rate": 1.3372909694073898e-05,
+      "loss": 0.5709,
+      "step": 2033
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.21398470273739778,
+      "learning_rate": 1.3349490103900582e-05,
+      "loss": 0.5618,
+      "step": 2034
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.23790198852595695,
+      "learning_rate": 1.3326083566922842e-05,
+      "loss": 0.8134,
+      "step": 2035
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.2309702023019256,
+      "learning_rate": 1.330269010936527e-05,
+      "loss": 0.5554,
+      "step": 2036
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.20874770009764093,
+      "learning_rate": 1.3279309757437832e-05,
+      "loss": 0.5681,
+      "step": 2037
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.24581020103497594,
+      "learning_rate": 1.3255942537335805e-05,
+      "loss": 0.8384,
+      "step": 2038
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.244087650429581,
+      "learning_rate": 1.323258847523975e-05,
+      "loss": 0.5661,
+      "step": 2039
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.2116662550285409,
+      "learning_rate": 1.3209247597315497e-05,
+      "loss": 0.5697,
+      "step": 2040
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.2455011343689653,
+      "learning_rate": 1.3185919929714072e-05,
+      "loss": 0.8121,
+      "step": 2041
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.24820037058707645,
+      "learning_rate": 1.3162605498571735e-05,
+      "loss": 0.6255,
+      "step": 2042
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.21976620734091176,
+      "learning_rate": 1.31393043300099e-05,
+      "loss": 0.5851,
+      "step": 2043
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.22824664393618063,
+      "learning_rate": 1.3116016450135128e-05,
+      "loss": 0.5765,
+      "step": 2044
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.23692859817292014,
+      "learning_rate": 1.3092741885039087e-05,
+      "loss": 0.8277,
+      "step": 2045
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.22235765331263907,
+      "learning_rate": 1.3069480660798528e-05,
+      "loss": 0.568,
+      "step": 2046
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.21633494865849026,
+      "learning_rate": 1.3046232803475255e-05,
+      "loss": 0.5635,
+      "step": 2047
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.215311715137634,
+      "learning_rate": 1.30229983391161e-05,
+      "loss": 0.5681,
+      "step": 2048
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.2317272977314365,
+      "learning_rate": 1.2999777293752877e-05,
+      "loss": 0.7945,
+      "step": 2049
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.22156600419559513,
+      "learning_rate": 1.2976569693402383e-05,
+      "loss": 0.5786,
+      "step": 2050
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.20735501430002837,
+      "learning_rate": 1.2953375564066345e-05,
+      "loss": 0.5844,
+      "step": 2051
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6586321037230484,
+      "learning_rate": 1.2930194931731382e-05,
+      "loss": 0.8157,
+      "step": 2052
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.2309215924503925,
+      "learning_rate": 1.2907027822369005e-05,
+      "loss": 0.6274,
+      "step": 2053
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.21138017741746376,
+      "learning_rate": 1.288387426193558e-05,
+      "loss": 0.5533,
+      "step": 2054
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.22130125480522259,
+      "learning_rate": 1.2860734276372283e-05,
+      "loss": 0.5478,
+      "step": 2055
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.23765823704432676,
+      "learning_rate": 1.283760789160508e-05,
+      "loss": 0.8003,
+      "step": 2056
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.2631589549209998,
+      "learning_rate": 1.2814495133544712e-05,
+      "loss": 0.5689,
+      "step": 2057
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.2224617534910379,
+      "learning_rate": 1.2791396028086628e-05,
+      "loss": 0.5792,
+      "step": 2058
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.2341753645743883,
+      "learning_rate": 1.2768310601110994e-05,
+      "loss": 0.6148,
+      "step": 2059
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.21973777685016155,
+      "learning_rate": 1.2745238878482662e-05,
+      "loss": 0.7264,
+      "step": 2060
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.20778925298276546,
+      "learning_rate": 1.2722180886051115e-05,
+      "loss": 0.5661,
+      "step": 2061
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.23349370146341958,
+      "learning_rate": 1.2699136649650456e-05,
+      "loss": 0.834,
+      "step": 2062
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5010866997423191,
+      "learning_rate": 1.2676106195099388e-05,
+      "loss": 0.5816,
+      "step": 2063
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.23060768709668503,
+      "learning_rate": 1.2653089548201142e-05,
+      "loss": 0.612,
+      "step": 2064
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.22387050246376378,
+      "learning_rate": 1.263008673474351e-05,
+      "loss": 0.5457,
+      "step": 2065
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.20956259166467364,
+      "learning_rate": 1.2607097780498772e-05,
+      "loss": 0.5683,
+      "step": 2066
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.23459792746278915,
+      "learning_rate": 1.2584122711223681e-05,
+      "loss": 0.8017,
+      "step": 2067
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.20952060919344756,
+      "learning_rate": 1.2561161552659434e-05,
+      "loss": 0.5701,
+      "step": 2068
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.20831664371865746,
+      "learning_rate": 1.2538214330531644e-05,
+      "loss": 0.583,
+      "step": 2069
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.24187781211411413,
+      "learning_rate": 1.2515281070550311e-05,
+      "loss": 0.613,
+      "step": 2070
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.23094307133639028,
+      "learning_rate": 1.2492361798409786e-05,
+      "loss": 0.8295,
+      "step": 2071
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.20921183555586145,
+      "learning_rate": 1.2469456539788748e-05,
+      "loss": 0.5639,
+      "step": 2072
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.2211287781165504,
+      "learning_rate": 1.2446565320350185e-05,
+      "loss": 0.5671,
+      "step": 2073
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.20539992519038697,
+      "learning_rate": 1.2423688165741354e-05,
+      "loss": 0.5695,
+      "step": 2074
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.24202830831518868,
+      "learning_rate": 1.2400825101593733e-05,
+      "loss": 0.8145,
+      "step": 2075
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.4103736847809047,
+      "learning_rate": 1.2377976153523033e-05,
+      "loss": 0.5674,
+      "step": 2076
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.21430151105756617,
+      "learning_rate": 1.2355141347129154e-05,
+      "loss": 0.561,
+      "step": 2077
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.248705393188941,
+      "learning_rate": 1.2332320707996133e-05,
+      "loss": 0.8043,
+      "step": 2078
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.22430186150534162,
+      "learning_rate": 1.2309514261692153e-05,
+      "loss": 0.5729,
+      "step": 2079
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.19701503483393776,
+      "learning_rate": 1.2286722033769493e-05,
+      "loss": 0.5928,
+      "step": 2080
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.2504339337036128,
+      "learning_rate": 1.2263944049764475e-05,
+      "loss": 0.6117,
+      "step": 2081
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.2399711940597156,
+      "learning_rate": 1.2241180335197494e-05,
+      "loss": 0.8198,
+      "step": 2082
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.2110209070942086,
+      "learning_rate": 1.2218430915572943e-05,
+      "loss": 0.5633,
+      "step": 2083
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.2344915749194354,
+      "learning_rate": 1.2195695816379205e-05,
+      "loss": 0.5727,
+      "step": 2084
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.24364484993953214,
+      "learning_rate": 1.217297506308862e-05,
+      "loss": 0.8177,
+      "step": 2085
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.2168735293703769,
+      "learning_rate": 1.215026868115744e-05,
+      "loss": 0.571,
+      "step": 2086
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.2085602085938762,
+      "learning_rate": 1.2127576696025828e-05,
+      "loss": 0.5716,
+      "step": 2087
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.23107574126783129,
+      "learning_rate": 1.2104899133117816e-05,
+      "loss": 0.8027,
+      "step": 2088
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.22483740676296984,
+      "learning_rate": 1.2082236017841275e-05,
+      "loss": 0.5579,
+      "step": 2089
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.21462766222837168,
+      "learning_rate": 1.2059587375587894e-05,
+      "loss": 0.5548,
+      "step": 2090
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.19708705163700876,
+      "learning_rate": 1.2036953231733139e-05,
+      "loss": 0.5693,
+      "step": 2091
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.23140456847479104,
+      "learning_rate": 1.2014333611636238e-05,
+      "loss": 0.5892,
+      "step": 2092
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.2385445698151682,
+      "learning_rate": 1.199172854064014e-05,
+      "loss": 0.8307,
+      "step": 2093
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.20017021632024143,
+      "learning_rate": 1.1969138044071501e-05,
+      "loss": 0.5546,
+      "step": 2094
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.20585447993869643,
+      "learning_rate": 1.1946562147240641e-05,
+      "loss": 0.571,
+      "step": 2095
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.23490236813745785,
+      "learning_rate": 1.1924000875441538e-05,
+      "loss": 0.817,
+      "step": 2096
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.22177528169084573,
+      "learning_rate": 1.1901454253951749e-05,
+      "loss": 0.5753,
+      "step": 2097
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.20392393693074426,
+      "learning_rate": 1.1878922308032457e-05,
+      "loss": 0.5407,
+      "step": 2098
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.24548262894075704,
+      "learning_rate": 1.1856405062928374e-05,
+      "loss": 0.8437,
+      "step": 2099
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.2260156517859528,
+      "learning_rate": 1.1833902543867759e-05,
+      "loss": 0.562,
+      "step": 2100
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.21132553594624126,
+      "learning_rate": 1.1811414776062366e-05,
+      "loss": 0.5624,
+      "step": 2101
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.23630418031217737,
+      "learning_rate": 1.1788941784707428e-05,
+      "loss": 0.8217,
+      "step": 2102
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.21882655696255734,
+      "learning_rate": 1.1766483594981597e-05,
+      "loss": 0.3931,
+      "step": 2103
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.22296377310427232,
+      "learning_rate": 1.174404023204697e-05,
+      "loss": 0.8396,
+      "step": 2104
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.21257504724764137,
+      "learning_rate": 1.1721611721049022e-05,
+      "loss": 0.5671,
+      "step": 2105
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.4588268411248062,
+      "learning_rate": 1.1699198087116589e-05,
+      "loss": 0.5897,
+      "step": 2106
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.23122147981853014,
+      "learning_rate": 1.1676799355361833e-05,
+      "loss": 0.7988,
+      "step": 2107
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.21631218602336463,
+      "learning_rate": 1.1654415550880243e-05,
+      "loss": 0.552,
+      "step": 2108
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.20689033069856633,
+      "learning_rate": 1.1632046698750537e-05,
+      "loss": 0.5658,
+      "step": 2109
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.23806717936486213,
+      "learning_rate": 1.160969282403472e-05,
+      "loss": 0.8126,
+      "step": 2110
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.21813194038573183,
+      "learning_rate": 1.158735395177801e-05,
+      "loss": 0.5676,
+      "step": 2111
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.21070719383436617,
+      "learning_rate": 1.1565030107008803e-05,
+      "loss": 0.5593,
+      "step": 2112
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.23079403203503424,
+      "learning_rate": 1.1542721314738674e-05,
+      "loss": 0.8082,
+      "step": 2113
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.23375148114875904,
+      "learning_rate": 1.152042759996232e-05,
+      "loss": 0.5814,
+      "step": 2114
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.2176310224174072,
+      "learning_rate": 1.149814898765755e-05,
+      "loss": 0.5809,
+      "step": 2115
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.19848297225983025,
+      "learning_rate": 1.1475885502785258e-05,
+      "loss": 0.5513,
+      "step": 2116
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.23153044046900434,
+      "learning_rate": 1.145363717028938e-05,
+      "loss": 0.8033,
+      "step": 2117
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.2259780808717556,
+      "learning_rate": 1.143140401509688e-05,
+      "loss": 0.5573,
+      "step": 2118
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.21203248955746265,
+      "learning_rate": 1.140918606211773e-05,
+      "loss": 0.5779,
+      "step": 2119
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.19923486492131537,
+      "learning_rate": 1.1386983336244832e-05,
+      "loss": 0.5398,
+      "step": 2120
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.23265152007478943,
+      "learning_rate": 1.1364795862354067e-05,
+      "loss": 0.8228,
+      "step": 2121
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.22738343321995458,
+      "learning_rate": 1.1342623665304209e-05,
+      "loss": 0.5455,
+      "step": 2122
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.1985475397371029,
+      "learning_rate": 1.132046676993692e-05,
+      "loss": 0.5664,
+      "step": 2123
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.2309641709691795,
+      "learning_rate": 1.1298325201076718e-05,
+      "loss": 0.8065,
+      "step": 2124
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.24389420368490947,
+      "learning_rate": 1.1276198983530958e-05,
+      "loss": 0.6007,
+      "step": 2125
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.21248529082634696,
+      "learning_rate": 1.1254088142089768e-05,
+      "loss": 0.5456,
+      "step": 2126
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.2104636212485788,
+      "learning_rate": 1.1231992701526078e-05,
+      "loss": 0.5543,
+      "step": 2127
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.232849744106806,
+      "learning_rate": 1.1209912686595548e-05,
+      "loss": 0.8278,
+      "step": 2128
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.22347447371725385,
+      "learning_rate": 1.1187848122036563e-05,
+      "loss": 0.561,
+      "step": 2129
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.2188846675645258,
+      "learning_rate": 1.1165799032570199e-05,
+      "loss": 0.5859,
+      "step": 2130
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.22265710287169807,
+      "learning_rate": 1.1143765442900173e-05,
+      "loss": 0.7846,
+      "step": 2131
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.22580972474658698,
+      "learning_rate": 1.1121747377712858e-05,
+      "loss": 0.5639,
+      "step": 2132
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.2257830262385762,
+      "learning_rate": 1.1099744861677219e-05,
+      "loss": 0.5756,
+      "step": 2133
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.22392751724465396,
+      "learning_rate": 1.1077757919444825e-05,
+      "loss": 0.8426,
+      "step": 2134
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.2100182729278371,
+      "learning_rate": 1.1055786575649771e-05,
+      "loss": 0.5715,
+      "step": 2135
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.2447507250764129,
+      "learning_rate": 1.1033830854908691e-05,
+      "loss": 0.6043,
+      "step": 2136
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.2269207520773483,
+      "learning_rate": 1.1011890781820688e-05,
+      "loss": 0.5605,
+      "step": 2137
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.20763983298065594,
+      "learning_rate": 1.0989966380967368e-05,
+      "loss": 0.5856,
+      "step": 2138
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.2487022503054038,
+      "learning_rate": 1.0968057676912751e-05,
+      "loss": 0.7867,
+      "step": 2139
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.21306270180543788,
+      "learning_rate": 1.0946164694203293e-05,
+      "loss": 0.5722,
+      "step": 2140
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.2258414562146985,
+      "learning_rate": 1.0924287457367826e-05,
+      "loss": 0.5632,
+      "step": 2141
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.23686938684159534,
+      "learning_rate": 1.0902425990917529e-05,
+      "loss": 0.813,
+      "step": 2142
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.21021430254205892,
+      "learning_rate": 1.088058031934592e-05,
+      "loss": 0.5586,
+      "step": 2143
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.21144059449876923,
+      "learning_rate": 1.0858750467128833e-05,
+      "loss": 0.5701,
+      "step": 2144
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.2205615316381155,
+      "learning_rate": 1.0836936458724362e-05,
+      "loss": 0.8105,
+      "step": 2145
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.20733048982873842,
+      "learning_rate": 1.0815138318572857e-05,
+      "loss": 0.5547,
+      "step": 2146
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.22551813614931807,
+      "learning_rate": 1.0793356071096893e-05,
+      "loss": 0.5855,
+      "step": 2147
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.21966504235647344,
+      "learning_rate": 1.077158974070122e-05,
+      "loss": 0.565,
+      "step": 2148
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.20278126421431406,
+      "learning_rate": 1.0749839351772775e-05,
+      "loss": 0.5617,
+      "step": 2149
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.22488947976708365,
+      "learning_rate": 1.0728104928680624e-05,
+      "loss": 0.852,
+      "step": 2150
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.2182229956693299,
+      "learning_rate": 1.0706386495775955e-05,
+      "loss": 0.5509,
+      "step": 2151
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.2075363474186994,
+      "learning_rate": 1.0684684077392027e-05,
+      "loss": 0.5661,
+      "step": 2152
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.22513141769125344,
+      "learning_rate": 1.0662997697844165e-05,
+      "loss": 0.8164,
+      "step": 2153
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.23288836052276912,
+      "learning_rate": 1.0641327381429722e-05,
+      "loss": 0.5684,
+      "step": 2154
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.21683913652338935,
+      "learning_rate": 1.0619673152428055e-05,
+      "loss": 0.5701,
+      "step": 2155
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.22603828966618092,
+      "learning_rate": 1.0598035035100498e-05,
+      "loss": 0.8081,
+      "step": 2156
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.2690130577276569,
+      "learning_rate": 1.0576413053690327e-05,
+      "loss": 0.5635,
+      "step": 2157
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.23900647845749037,
+      "learning_rate": 1.0554807232422756e-05,
+      "loss": 0.6034,
+      "step": 2158
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.21312572794176632,
+      "learning_rate": 1.0533217595504858e-05,
+      "loss": 0.5695,
+      "step": 2159
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.22915783865396788,
+      "learning_rate": 1.0511644167125611e-05,
+      "loss": 0.8136,
+      "step": 2160
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.21588504573565284,
+      "learning_rate": 1.0490086971455816e-05,
+      "loss": 0.5773,
+      "step": 2161
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.20703517232943563,
+      "learning_rate": 1.0468546032648086e-05,
+      "loss": 0.5764,
+      "step": 2162
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.21238507943066823,
+      "learning_rate": 1.0447021374836824e-05,
+      "loss": 0.567,
+      "step": 2163
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.22914550979320777,
+      "learning_rate": 1.0425513022138203e-05,
+      "loss": 0.8294,
+      "step": 2164
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.213677014249282,
+      "learning_rate": 1.0404020998650094e-05,
+      "loss": 0.5749,
+      "step": 2165
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.21020618914675637,
+      "learning_rate": 1.0382545328452107e-05,
+      "loss": 0.5714,
+      "step": 2166
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.22315493903911657,
+      "learning_rate": 1.0361086035605511e-05,
+      "loss": 0.8047,
+      "step": 2167
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.20653650099377044,
+      "learning_rate": 1.0339643144153235e-05,
+      "loss": 0.564,
+      "step": 2168
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.2226514328468365,
+      "learning_rate": 1.0318216678119832e-05,
+      "loss": 0.5962,
+      "step": 2169
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.19928144408195772,
+      "learning_rate": 1.0296806661511455e-05,
+      "loss": 0.5666,
+      "step": 2170
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.22361088233824178,
+      "learning_rate": 1.0275413118315799e-05,
+      "loss": 0.7978,
+      "step": 2171
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.21674009668977773,
+      "learning_rate": 1.0254036072502141e-05,
+      "loss": 0.5562,
+      "step": 2172
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.1985261973796685,
+      "learning_rate": 1.0232675548021251e-05,
+      "loss": 0.5531,
+      "step": 2173
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.2318296457562149,
+      "learning_rate": 1.0211331568805399e-05,
+      "loss": 0.6211,
+      "step": 2174
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.20477613252990048,
+      "learning_rate": 1.0190004158768313e-05,
+      "loss": 0.5619,
+      "step": 2175
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.22734884405860678,
+      "learning_rate": 1.0168693341805157e-05,
+      "loss": 0.8137,
+      "step": 2176
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.2238055833287567,
+      "learning_rate": 1.0147399141792505e-05,
+      "loss": 0.5819,
+      "step": 2177
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.20414670941109042,
+      "learning_rate": 1.0126121582588316e-05,
+      "loss": 0.7099,
+      "step": 2178
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.20328541552857082,
+      "learning_rate": 1.0104860688031898e-05,
+      "loss": 0.5733,
+      "step": 2179
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.2383260563396624,
+      "learning_rate": 1.0083616481943895e-05,
+      "loss": 0.6034,
+      "step": 2180
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.21442939265096303,
+      "learning_rate": 1.0062388988126257e-05,
+      "loss": 0.5809,
+      "step": 2181
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.23275292651599677,
+      "learning_rate": 1.0041178230362187e-05,
+      "loss": 0.8384,
+      "step": 2182
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.23171870519808957,
+      "learning_rate": 1.0019984232416163e-05,
+      "loss": 0.5663,
+      "step": 2183
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.21030908703513795,
+      "learning_rate": 9.998807018033871e-06,
+      "loss": 0.5577,
+      "step": 2184
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.23474302729985244,
+      "learning_rate": 9.977646610942202e-06,
+      "loss": 0.642,
+      "step": 2185
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.24089368929915178,
+      "learning_rate": 9.956503034849216e-06,
+      "loss": 0.8236,
+      "step": 2186
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.22324232749213033,
+      "learning_rate": 9.935376313444094e-06,
+      "loss": 0.568,
+      "step": 2187
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.20484180763964863,
+      "learning_rate": 9.914266470397163e-06,
+      "loss": 0.5856,
+      "step": 2188
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.2294277839143203,
+      "learning_rate": 9.893173529359823e-06,
+      "loss": 0.5781,
+      "step": 2189
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.23118865282695203,
+      "learning_rate": 9.872097513964537e-06,
+      "loss": 0.8088,
+      "step": 2190
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.2076065398853168,
+      "learning_rate": 9.851038447824815e-06,
+      "loss": 0.5556,
+      "step": 2191
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.21888598824823913,
+      "learning_rate": 9.829996354535172e-06,
+      "loss": 0.564,
+      "step": 2192
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.23479488883925742,
+      "learning_rate": 9.808971257671096e-06,
+      "loss": 0.8244,
+      "step": 2193
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.20726235211742472,
+      "learning_rate": 9.787963180789039e-06,
+      "loss": 0.5465,
+      "step": 2194
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.20431132650010572,
+      "learning_rate": 9.766972147426396e-06,
+      "loss": 0.5723,
+      "step": 2195
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.24072062369132713,
+      "learning_rate": 9.745998181101449e-06,
+      "loss": 0.5901,
+      "step": 2196
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.223667544898505,
+      "learning_rate": 9.725041305313368e-06,
+      "loss": 0.809,
+      "step": 2197
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.2117486515671313,
+      "learning_rate": 9.704101543542174e-06,
+      "loss": 0.5522,
+      "step": 2198
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.20296570507105613,
+      "learning_rate": 9.683178919248712e-06,
+      "loss": 0.5543,
+      "step": 2199
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.22845370017070013,
+      "learning_rate": 9.662273455874623e-06,
+      "loss": 0.8021,
+      "step": 2200
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.20992433438553193,
+      "learning_rate": 9.641385176842327e-06,
+      "loss": 0.5552,
+      "step": 2201
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.2107429018146333,
+      "learning_rate": 9.620514105554986e-06,
+      "loss": 0.5537,
+      "step": 2202
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.2287409897838188,
+      "learning_rate": 9.599660265396496e-06,
+      "loss": 0.8281,
+      "step": 2203
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.20665662093640141,
+      "learning_rate": 9.578823679731413e-06,
+      "loss": 0.5545,
+      "step": 2204
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.200201667341098,
+      "learning_rate": 9.558004371904994e-06,
+      "loss": 0.5543,
+      "step": 2205
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.21849550323315686,
+      "learning_rate": 9.53720236524313e-06,
+      "loss": 0.8255,
+      "step": 2206
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.20842756834689566,
+      "learning_rate": 9.516417683052328e-06,
+      "loss": 0.381,
+      "step": 2207
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.22516164844641864,
+      "learning_rate": 9.495650348619675e-06,
+      "loss": 0.8045,
+      "step": 2208
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.21064843810951858,
+      "learning_rate": 9.474900385212842e-06,
+      "loss": 0.5659,
+      "step": 2209
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.20027040775621968,
+      "learning_rate": 9.454167816080007e-06,
+      "loss": 0.5731,
+      "step": 2210
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.23267654031451854,
+      "learning_rate": 9.433452664449882e-06,
+      "loss": 0.8094,
+      "step": 2211
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.2131669454888573,
+      "learning_rate": 9.412754953531663e-06,
+      "loss": 0.5832,
+      "step": 2212
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.20568856682475242,
+      "learning_rate": 9.392074706515003e-06,
+      "loss": 0.5737,
+      "step": 2213
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.2224327495780646,
+      "learning_rate": 9.371411946569983e-06,
+      "loss": 0.7948,
+      "step": 2214
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.20041772290590437,
+      "learning_rate": 9.350766696847108e-06,
+      "loss": 0.5414,
+      "step": 2215
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.2006317365896454,
+      "learning_rate": 9.330138980477233e-06,
+      "loss": 0.5432,
+      "step": 2216
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.23231718524004,
+      "learning_rate": 9.309528820571597e-06,
+      "loss": 0.7835,
+      "step": 2217
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.20416707684373298,
+      "learning_rate": 9.288936240221763e-06,
+      "loss": 0.3823,
+      "step": 2218
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.22023833395389542,
+      "learning_rate": 9.268361262499592e-06,
+      "loss": 0.8208,
+      "step": 2219
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.20315193502726037,
+      "learning_rate": 9.247803910457226e-06,
+      "loss": 0.5603,
+      "step": 2220
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.20994807878951918,
+      "learning_rate": 9.22726420712706e-06,
+      "loss": 0.5588,
+      "step": 2221
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.22126822199055687,
+      "learning_rate": 9.206742175521713e-06,
+      "loss": 0.8013,
+      "step": 2222
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.20547424430656977,
+      "learning_rate": 9.186237838634007e-06,
+      "loss": 0.5645,
+      "step": 2223
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.20667816780796477,
+      "learning_rate": 9.165751219436942e-06,
+      "loss": 0.5506,
+      "step": 2224
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.22192476981384132,
+      "learning_rate": 9.145282340883659e-06,
+      "loss": 0.7862,
+      "step": 2225
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.20983881079516198,
+      "learning_rate": 9.124831225907435e-06,
+      "loss": 0.5646,
+      "step": 2226
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.2055371674911453,
+      "learning_rate": 9.104397897421623e-06,
+      "loss": 0.5602,
+      "step": 2227
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.2207476332993922,
+      "learning_rate": 9.083982378319667e-06,
+      "loss": 0.8072,
+      "step": 2228
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.23516563113871472,
+      "learning_rate": 9.063584691475052e-06,
+      "loss": 0.6128,
+      "step": 2229
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.2070892886381353,
+      "learning_rate": 9.043204859741289e-06,
+      "loss": 0.5421,
+      "step": 2230
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.195035591759891,
+      "learning_rate": 9.022842905951879e-06,
+      "loss": 0.5611,
+      "step": 2231
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.23195847689893795,
+      "learning_rate": 9.002498852920282e-06,
+      "loss": 0.8087,
+      "step": 2232
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.20457157591354555,
+      "learning_rate": 8.982172723439919e-06,
+      "loss": 0.5684,
+      "step": 2233
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.19813841536189222,
+      "learning_rate": 8.96186454028412e-06,
+      "loss": 0.5839,
+      "step": 2234
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.2146268211806279,
+      "learning_rate": 8.941574326206118e-06,
+      "loss": 0.5564,
+      "step": 2235
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.2206185672366353,
+      "learning_rate": 8.921302103939002e-06,
+      "loss": 0.8103,
+      "step": 2236
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.19337361920603038,
+      "learning_rate": 8.901047896195719e-06,
+      "loss": 0.5442,
+      "step": 2237
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.3504768365985834,
+      "learning_rate": 8.880811725668999e-06,
+      "loss": 0.5627,
+      "step": 2238
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.22759728752379035,
+      "learning_rate": 8.8605936150314e-06,
+      "loss": 0.8195,
+      "step": 2239
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.24226608742050193,
+      "learning_rate": 8.840393586935228e-06,
+      "loss": 0.5941,
+      "step": 2240
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.20445194447038967,
+      "learning_rate": 8.820211664012532e-06,
+      "loss": 0.5654,
+      "step": 2241
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.20063532117490854,
+      "learning_rate": 8.800047868875078e-06,
+      "loss": 0.5534,
+      "step": 2242
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.2279409261353751,
+      "learning_rate": 8.779902224114319e-06,
+      "loss": 0.7928,
+      "step": 2243
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.21031293410798663,
+      "learning_rate": 8.759774752301375e-06,
+      "loss": 0.5706,
+      "step": 2244
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.1957522501901852,
+      "learning_rate": 8.739665475987005e-06,
+      "loss": 0.5593,
+      "step": 2245
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.2351095664927906,
+      "learning_rate": 8.719574417701576e-06,
+      "loss": 0.8258,
+      "step": 2246
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.20560378520842068,
+      "learning_rate": 8.69950159995505e-06,
+      "loss": 0.5463,
+      "step": 2247
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.2030879340684866,
+      "learning_rate": 8.679447045236962e-06,
+      "loss": 0.5548,
+      "step": 2248
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.23176696063586205,
+      "learning_rate": 8.659410776016354e-06,
+      "loss": 0.842,
+      "step": 2249
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.2115635637644575,
+      "learning_rate": 8.639392814741812e-06,
+      "loss": 0.5748,
+      "step": 2250
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.22528490438696516,
+      "learning_rate": 8.619393183841393e-06,
+      "loss": 0.6231,
+      "step": 2251
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.20902206715138075,
+      "learning_rate": 8.599411905722626e-06,
+      "loss": 0.5612,
+      "step": 2252
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.19964674371041577,
+      "learning_rate": 8.579449002772471e-06,
+      "loss": 0.5536,
+      "step": 2253
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.22361227560534688,
+      "learning_rate": 8.559504497357315e-06,
+      "loss": 0.8096,
+      "step": 2254
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.1996517123315839,
+      "learning_rate": 8.539578411822901e-06,
+      "loss": 0.5563,
+      "step": 2255
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.19865783386298666,
+      "learning_rate": 8.519670768494367e-06,
+      "loss": 0.5547,
+      "step": 2256
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.23016090055019617,
+      "learning_rate": 8.49978158967617e-06,
+      "loss": 0.829,
+      "step": 2257
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.21076407101465316,
+      "learning_rate": 8.479910897652087e-06,
+      "loss": 0.5659,
+      "step": 2258
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.2036522751800532,
+      "learning_rate": 8.46005871468519e-06,
+      "loss": 0.5982,
+      "step": 2259
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.22159048085798558,
+      "learning_rate": 8.440225063017787e-06,
+      "loss": 0.8239,
+      "step": 2260
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.2033839269483129,
+      "learning_rate": 8.420409964871449e-06,
+      "loss": 0.5533,
+      "step": 2261
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.22616164413297254,
+      "learning_rate": 8.400613442446948e-06,
+      "loss": 0.6006,
+      "step": 2262
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.23637963157582723,
+      "learning_rate": 8.380835517924249e-06,
+      "loss": 0.5568,
+      "step": 2263
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.1993621688071713,
+      "learning_rate": 8.361076213462479e-06,
+      "loss": 0.5572,
+      "step": 2264
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.3305931505799798,
+      "learning_rate": 8.341335551199902e-06,
+      "loss": 0.8048,
+      "step": 2265
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.22062132290875175,
+      "learning_rate": 8.321613553253895e-06,
+      "loss": 0.5764,
+      "step": 2266
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.21177488535809755,
+      "learning_rate": 8.301910241720926e-06,
+      "loss": 0.566,
+      "step": 2267
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.22513556461099157,
+      "learning_rate": 8.282225638676522e-06,
+      "loss": 0.7899,
+      "step": 2268
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.20346305331473616,
+      "learning_rate": 8.262559766175254e-06,
+      "loss": 0.5562,
+      "step": 2269
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.19871702347897566,
+      "learning_rate": 8.2429126462507e-06,
+      "loss": 0.5556,
+      "step": 2270
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.2199647040209428,
+      "learning_rate": 8.223284300915449e-06,
+      "loss": 0.8019,
+      "step": 2271
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.39784315052425784,
+      "learning_rate": 8.203674752161017e-06,
+      "loss": 0.5561,
+      "step": 2272
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.21815278973492164,
+      "learning_rate": 8.184084021957894e-06,
+      "loss": 0.6099,
+      "step": 2273
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.20349715636119245,
+      "learning_rate": 8.16451213225547e-06,
+      "loss": 0.5527,
+      "step": 2274
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.22008580451590065,
+      "learning_rate": 8.144959104982033e-06,
+      "loss": 0.8127,
+      "step": 2275
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.1976559036159888,
+      "learning_rate": 8.125424962044742e-06,
+      "loss": 0.5395,
+      "step": 2276
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.19833829000122674,
+      "learning_rate": 8.105909725329577e-06,
+      "loss": 0.5895,
+      "step": 2277
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.19613962677092922,
+      "learning_rate": 8.086413416701355e-06,
+      "loss": 0.5742,
+      "step": 2278
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.21924654864083684,
+      "learning_rate": 8.06693605800368e-06,
+      "loss": 0.6183,
+      "step": 2279
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.22149824587431471,
+      "learning_rate": 8.047477671058926e-06,
+      "loss": 0.8286,
+      "step": 2280
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.18813288831950709,
+      "learning_rate": 8.028038277668209e-06,
+      "loss": 0.5289,
+      "step": 2281
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.22244098913830374,
+      "learning_rate": 8.008617899611375e-06,
+      "loss": 0.7629,
+      "step": 2282
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.20569273069050859,
+      "learning_rate": 7.989216558646942e-06,
+      "loss": 0.5671,
+      "step": 2283
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.2205398855916073,
+      "learning_rate": 7.969834276512118e-06,
+      "loss": 0.6302,
+      "step": 2284
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.45743624355001694,
+      "learning_rate": 7.950471074922753e-06,
+      "loss": 0.5727,
+      "step": 2285
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.2211885571923775,
+      "learning_rate": 7.931126975573316e-06,
+      "loss": 0.8097,
+      "step": 2286
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.26664341568459454,
+      "learning_rate": 7.911802000136884e-06,
+      "loss": 0.6081,
+      "step": 2287
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.19655665916994886,
+      "learning_rate": 7.892496170265093e-06,
+      "loss": 0.5562,
+      "step": 2288
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.23799427886564065,
+      "learning_rate": 7.873209507588138e-06,
+      "loss": 0.8175,
+      "step": 2289
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.2007719806541293,
+      "learning_rate": 7.853942033714736e-06,
+      "loss": 0.3852,
+      "step": 2290
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.21885282869480996,
+      "learning_rate": 7.83469377023211e-06,
+      "loss": 0.8152,
+      "step": 2291
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.20101687271526436,
+      "learning_rate": 7.81546473870595e-06,
+      "loss": 0.5575,
+      "step": 2292
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.21087728743740966,
+      "learning_rate": 7.796254960680416e-06,
+      "loss": 0.5539,
+      "step": 2293
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.22710401408869788,
+      "learning_rate": 7.777064457678063e-06,
+      "loss": 0.814,
+      "step": 2294
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.20776790342666587,
+      "learning_rate": 7.757893251199885e-06,
+      "loss": 0.7081,
+      "step": 2295
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.2028112951905871,
+      "learning_rate": 7.73874136272524e-06,
+      "loss": 0.3874,
+      "step": 2296
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.2381435367658184,
+      "learning_rate": 7.719608813711848e-06,
+      "loss": 0.8167,
+      "step": 2297
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.22000684494755549,
+      "learning_rate": 7.700495625595755e-06,
+      "loss": 0.5883,
+      "step": 2298
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.20625307908728027,
+      "learning_rate": 7.68140181979133e-06,
+      "loss": 0.5666,
+      "step": 2299
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.22332727079743364,
+      "learning_rate": 7.662327417691193e-06,
+      "loss": 0.7888,
+      "step": 2300
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.22845388486035165,
+      "learning_rate": 7.643272440666257e-06,
+      "loss": 0.6047,
+      "step": 2301
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.20850681832861434,
+      "learning_rate": 7.624236910065663e-06,
+      "loss": 0.5498,
+      "step": 2302
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.20811125095501307,
+      "learning_rate": 7.6052208472167544e-06,
+      "loss": 0.5493,
+      "step": 2303
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.23196299339664514,
+      "learning_rate": 7.586224273425082e-06,
+      "loss": 0.7936,
+      "step": 2304
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.20632382265016322,
+      "learning_rate": 7.567247209974335e-06,
+      "loss": 0.5937,
+      "step": 2305
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.1977479742047526,
+      "learning_rate": 7.548289678126358e-06,
+      "loss": 0.5597,
+      "step": 2306
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.20148645285216316,
+      "learning_rate": 7.529351699121118e-06,
+      "loss": 0.5688,
+      "step": 2307
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.2288206985933724,
+      "learning_rate": 7.5104332941766665e-06,
+      "loss": 0.8055,
+      "step": 2308
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.1991511186655231,
+      "learning_rate": 7.491534484489115e-06,
+      "loss": 0.5718,
+      "step": 2309
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.19779127854278442,
+      "learning_rate": 7.472655291232664e-06,
+      "loss": 0.5534,
+      "step": 2310
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.23746896250440458,
+      "learning_rate": 7.453795735559471e-06,
+      "loss": 0.785,
+      "step": 2311
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.21974411666116261,
+      "learning_rate": 7.434955838599738e-06,
+      "loss": 0.6109,
+      "step": 2312
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.20325409175032683,
+      "learning_rate": 7.416135621461626e-06,
+      "loss": 0.5492,
+      "step": 2313
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.22431868158450186,
+      "learning_rate": 7.397335105231246e-06,
+      "loss": 0.582,
+      "step": 2314
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.49998885994953396,
+      "learning_rate": 7.378554310972643e-06,
+      "loss": 0.8476,
+      "step": 2315
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.20038403990459783,
+      "learning_rate": 7.359793259727768e-06,
+      "loss": 0.5471,
+      "step": 2316
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.2039422755170208,
+      "learning_rate": 7.34105197251643e-06,
+      "loss": 0.5389,
+      "step": 2317
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.23108319285005025,
+      "learning_rate": 7.3223304703363135e-06,
+      "loss": 0.8205,
+      "step": 2318
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.1952000257053664,
+      "learning_rate": 7.303628774162938e-06,
+      "loss": 0.5593,
+      "step": 2319
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.19868639023280654,
+      "learning_rate": 7.2849469049496216e-06,
+      "loss": 0.5614,
+      "step": 2320
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.234866718909701,
+      "learning_rate": 7.266284883627483e-06,
+      "loss": 0.8082,
+      "step": 2321
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.2037474706508701,
+      "learning_rate": 7.247642731105381e-06,
+      "loss": 0.5508,
+      "step": 2322
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.23535328029255748,
+      "learning_rate": 7.2290204682699315e-06,
+      "loss": 0.6088,
+      "step": 2323
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.20506520621957483,
+      "learning_rate": 7.210418115985457e-06,
+      "loss": 0.5542,
+      "step": 2324
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.20318100076030807,
+      "learning_rate": 7.191835695093982e-06,
+      "loss": 0.5583,
+      "step": 2325
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.22837424884281815,
+      "learning_rate": 7.173273226415189e-06,
+      "loss": 0.8171,
+      "step": 2326
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.20516637232167648,
+      "learning_rate": 7.154730730746423e-06,
+      "loss": 0.5766,
+      "step": 2327
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.19840771833284182,
+      "learning_rate": 7.136208228862615e-06,
+      "loss": 0.559,
+      "step": 2328
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.2404496031673577,
+      "learning_rate": 7.1177057415163394e-06,
+      "loss": 0.8246,
+      "step": 2329
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.20184902870592433,
+      "learning_rate": 7.099223289437723e-06,
+      "loss": 0.5686,
+      "step": 2330
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.19573148989413472,
+      "learning_rate": 7.080760893334443e-06,
+      "loss": 0.5715,
+      "step": 2331
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.22894931902342505,
+      "learning_rate": 7.062318573891716e-06,
+      "loss": 0.8224,
+      "step": 2332
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.1986668645357486,
+      "learning_rate": 7.043896351772267e-06,
+      "loss": 0.5658,
+      "step": 2333
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.22506968326974916,
+      "learning_rate": 7.0254942476162824e-06,
+      "loss": 0.5947,
+      "step": 2334
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.20371833344247306,
+      "learning_rate": 7.007112282041428e-06,
+      "loss": 0.5562,
+      "step": 2335
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.19633554900585043,
+      "learning_rate": 6.988750475642805e-06,
+      "loss": 0.5723,
+      "step": 2336
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.22798329108224288,
+      "learning_rate": 6.970408848992921e-06,
+      "loss": 0.8238,
+      "step": 2337
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.19844875125172545,
+      "learning_rate": 6.9520874226416895e-06,
+      "loss": 0.5471,
+      "step": 2338
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.20406658442556777,
+      "learning_rate": 6.933786217116364e-06,
+      "loss": 0.5526,
+      "step": 2339
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.23368723575633538,
+      "learning_rate": 6.915505252921569e-06,
+      "loss": 0.8192,
+      "step": 2340
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.20180478981577601,
+      "learning_rate": 6.897244550539236e-06,
+      "loss": 0.5662,
+      "step": 2341
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.19319194969652248,
+      "learning_rate": 6.879004130428604e-06,
+      "loss": 0.5729,
+      "step": 2342
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.24788483046064147,
+      "learning_rate": 6.860784013026181e-06,
+      "loss": 0.8286,
+      "step": 2343
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.20637157631206327,
+      "learning_rate": 6.842584218745738e-06,
+      "loss": 0.5595,
+      "step": 2344
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.22693955557128836,
+      "learning_rate": 6.824404767978257e-06,
+      "loss": 0.6061,
+      "step": 2345
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.19901237709462646,
+      "learning_rate": 6.806245681091944e-06,
+      "loss": 0.5492,
+      "step": 2346
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.226171664069894,
+      "learning_rate": 6.788106978432185e-06,
+      "loss": 0.8162,
+      "step": 2347
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.1980023142364903,
+      "learning_rate": 6.769988680321526e-06,
+      "loss": 0.5693,
+      "step": 2348
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.20061991091661494,
+      "learning_rate": 6.751890807059652e-06,
+      "loss": 0.5384,
+      "step": 2349
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.19177116383866338,
+      "learning_rate": 6.733813378923367e-06,
+      "loss": 0.5652,
+      "step": 2350
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.22867974402647365,
+      "learning_rate": 6.715756416166566e-06,
+      "loss": 0.8098,
+      "step": 2351
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.20697509817193424,
+      "learning_rate": 6.697719939020211e-06,
+      "loss": 0.5689,
+      "step": 2352
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.21225209144174656,
+      "learning_rate": 6.679703967692322e-06,
+      "loss": 0.5493,
+      "step": 2353
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.22734891741382426,
+      "learning_rate": 6.661708522367932e-06,
+      "loss": 0.7935,
+      "step": 2354
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.2002981195022134,
+      "learning_rate": 6.643733623209092e-06,
+      "loss": 0.5669,
+      "step": 2355
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.21458445046825397,
+      "learning_rate": 6.625779290354811e-06,
+      "loss": 0.5976,
+      "step": 2356
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.19938200634180656,
+      "learning_rate": 6.607845543921073e-06,
+      "loss": 0.5709,
+      "step": 2357
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.2209417440705201,
+      "learning_rate": 6.589932404000795e-06,
+      "loss": 0.8009,
+      "step": 2358
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.19653175673095016,
+      "learning_rate": 6.572039890663803e-06,
+      "loss": 0.5701,
+      "step": 2359
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.20801630183021688,
+      "learning_rate": 6.5541680239568165e-06,
+      "loss": 0.5689,
+      "step": 2360
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.2171439073725004,
+      "learning_rate": 6.536316823903421e-06,
+      "loss": 0.8135,
+      "step": 2361
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.1934048649610309,
+      "learning_rate": 6.5184863105040405e-06,
+      "loss": 0.5431,
+      "step": 2362
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.19967753737014546,
+      "learning_rate": 6.500676503735925e-06,
+      "loss": 0.5611,
+      "step": 2363
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.22807893469122856,
+      "learning_rate": 6.482887423553135e-06,
+      "loss": 0.8184,
+      "step": 2364
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5108071256055979,
+      "learning_rate": 6.465119089886501e-06,
+      "loss": 0.5765,
+      "step": 2365
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.19547331305582477,
+      "learning_rate": 6.447371522643614e-06,
+      "loss": 0.5703,
+      "step": 2366
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.21839238552513912,
+      "learning_rate": 6.429644741708779e-06,
+      "loss": 0.604,
+      "step": 2367
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.20565807139704237,
+      "learning_rate": 6.411938766943035e-06,
+      "loss": 0.5554,
+      "step": 2368
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.2247809170168605,
+      "learning_rate": 6.394253618184101e-06,
+      "loss": 0.7595,
+      "step": 2369
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.18848273978198646,
+      "learning_rate": 6.3765893152463675e-06,
+      "loss": 0.5605,
+      "step": 2370
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.19632418897737422,
+      "learning_rate": 6.358945877920861e-06,
+      "loss": 0.5596,
+      "step": 2371
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.21795679976777052,
+      "learning_rate": 6.341323325975238e-06,
+      "loss": 0.7926,
+      "step": 2372
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.2545378769310287,
+      "learning_rate": 6.323721679153749e-06,
+      "loss": 0.5599,
+      "step": 2373
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.1958196270919068,
+      "learning_rate": 6.3061409571772254e-06,
+      "loss": 0.5721,
+      "step": 2374
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.22418373232740083,
+      "learning_rate": 6.288581179743053e-06,
+      "loss": 0.8078,
+      "step": 2375
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.20097885302267765,
+      "learning_rate": 6.2710423665251565e-06,
+      "loss": 0.5782,
+      "step": 2376
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.2005304674262192,
+      "learning_rate": 6.253524537173966e-06,
+      "loss": 0.5735,
+      "step": 2377
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.23053135621562157,
+      "learning_rate": 6.236027711316411e-06,
+      "loss": 0.6056,
+      "step": 2378
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.19715553981909956,
+      "learning_rate": 6.218551908555869e-06,
+      "loss": 0.5647,
+      "step": 2379
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.23029952344606786,
+      "learning_rate": 6.201097148472182e-06,
+      "loss": 0.8022,
+      "step": 2380
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.2082034898728833,
+      "learning_rate": 6.183663450621607e-06,
+      "loss": 0.5581,
+      "step": 2381
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.1922894311835887,
+      "learning_rate": 6.166250834536808e-06,
+      "loss": 0.5664,
+      "step": 2382
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.21732085565190773,
+      "learning_rate": 6.148859319726832e-06,
+      "loss": 0.7927,
+      "step": 2383
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.20071233529190236,
+      "learning_rate": 6.13148892567707e-06,
+      "loss": 0.5662,
+      "step": 2384
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.2049892262546387,
+      "learning_rate": 6.114139671849259e-06,
+      "loss": 0.5724,
+      "step": 2385
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.21917035650576877,
+      "learning_rate": 6.096811577681452e-06,
+      "loss": 0.7859,
+      "step": 2386
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.25056408068981784,
+      "learning_rate": 6.079504662587993e-06,
+      "loss": 0.553,
+      "step": 2387
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.19850872826354896,
+      "learning_rate": 6.062218945959497e-06,
+      "loss": 0.5717,
+      "step": 2388
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.22109844357983716,
+      "learning_rate": 6.044954447162837e-06,
+      "loss": 0.6189,
+      "step": 2389
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.22449321210533416,
+      "learning_rate": 6.027711185541085e-06,
+      "loss": 0.8048,
+      "step": 2390
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.20621868304122637,
+      "learning_rate": 6.010489180413547e-06,
+      "loss": 0.5412,
+      "step": 2391
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.19762111913717234,
+      "learning_rate": 5.99328845107571e-06,
+      "loss": 0.5794,
+      "step": 2392
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.23489333705484114,
+      "learning_rate": 5.976109016799211e-06,
+      "loss": 0.8136,
+      "step": 2393
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.2037526299695046,
+      "learning_rate": 5.958950896831841e-06,
+      "loss": 0.5515,
+      "step": 2394
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.2296736431792935,
+      "learning_rate": 5.941814110397503e-06,
+      "loss": 0.6002,
+      "step": 2395
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.19614196104991974,
+      "learning_rate": 5.9246986766962e-06,
+      "loss": 0.56,
+      "step": 2396
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.19946897580652684,
+      "learning_rate": 5.907604614904011e-06,
+      "loss": 0.5784,
+      "step": 2397
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.22746017407986815,
+      "learning_rate": 5.890531944173072e-06,
+      "loss": 0.8048,
+      "step": 2398
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.2046463269156806,
+      "learning_rate": 5.873480683631546e-06,
+      "loss": 0.5382,
+      "step": 2399
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.23000073357592896,
+      "learning_rate": 5.856450852383627e-06,
+      "loss": 0.5953,
+      "step": 2400
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.22922409308369568,
+      "learning_rate": 5.839442469509468e-06,
+      "loss": 0.8504,
+      "step": 2401
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.20755520288699028,
+      "learning_rate": 5.822455554065217e-06,
+      "loss": 0.5538,
+      "step": 2402
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.21847510671644055,
+      "learning_rate": 5.805490125082963e-06,
+      "loss": 0.5557,
+      "step": 2403
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.2198301599393458,
+      "learning_rate": 5.788546201570716e-06,
+      "loss": 0.8134,
+      "step": 2404
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.2020221879376674,
+      "learning_rate": 5.771623802512402e-06,
+      "loss": 0.571,
+      "step": 2405
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.23057217989187104,
+      "learning_rate": 5.754722946867827e-06,
+      "loss": 0.6048,
+      "step": 2406
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.20514867468563472,
+      "learning_rate": 5.737843653572647e-06,
+      "loss": 0.5527,
+      "step": 2407
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.20425397108468035,
+      "learning_rate": 5.720985941538376e-06,
+      "loss": 0.5758,
+      "step": 2408
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.22274362867210923,
+      "learning_rate": 5.704149829652342e-06,
+      "loss": 0.8164,
+      "step": 2409
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.20984319495299547,
+      "learning_rate": 5.6873353367776745e-06,
+      "loss": 0.5402,
+      "step": 2410
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.21446984164135152,
+      "learning_rate": 5.670542481753288e-06,
+      "loss": 0.5728,
+      "step": 2411
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.23450544913184945,
+      "learning_rate": 5.65377128339383e-06,
+      "loss": 0.7982,
+      "step": 2412
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.21300079798129307,
+      "learning_rate": 5.637021760489708e-06,
+      "loss": 0.5662,
+      "step": 2413
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.2014132853155784,
+      "learning_rate": 5.6202939318070355e-06,
+      "loss": 0.5613,
+      "step": 2414
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.22096258309273378,
+      "learning_rate": 5.603587816087619e-06,
+      "loss": 0.7888,
+      "step": 2415
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.19722422416069055,
+      "learning_rate": 5.586903432048943e-06,
+      "loss": 0.5664,
+      "step": 2416
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.2207930313453233,
+      "learning_rate": 5.570240798384135e-06,
+      "loss": 0.596,
+      "step": 2417
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.19807171073142818,
+      "learning_rate": 5.553599933761963e-06,
+      "loss": 0.5601,
+      "step": 2418
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.2293254631375211,
+      "learning_rate": 5.536980856826799e-06,
+      "loss": 0.7869,
+      "step": 2419
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.19725466847918663,
+      "learning_rate": 5.520383586198605e-06,
+      "loss": 0.5687,
+      "step": 2420
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.21338819793349714,
+      "learning_rate": 5.503808140472913e-06,
+      "loss": 0.5689,
+      "step": 2421
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.2095384595617453,
+      "learning_rate": 5.48725453822081e-06,
+      "loss": 0.5653,
+      "step": 2422
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.21988469695838186,
+      "learning_rate": 5.4707227979888834e-06,
+      "loss": 0.8057,
+      "step": 2423
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.19903093838648062,
+      "learning_rate": 5.454212938299255e-06,
+      "loss": 0.5587,
+      "step": 2424
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.2001616662385984,
+      "learning_rate": 5.437724977649522e-06,
+      "loss": 0.5447,
+      "step": 2425
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.23097933439842136,
+      "learning_rate": 5.421258934512741e-06,
+      "loss": 0.8017,
+      "step": 2426
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.1974580016105894,
+      "learning_rate": 5.404814827337418e-06,
+      "loss": 0.561,
+      "step": 2427
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.22454803686976266,
+      "learning_rate": 5.388392674547488e-06,
+      "loss": 0.6627,
+      "step": 2428
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.1942071637022234,
+      "learning_rate": 5.371992494542263e-06,
+      "loss": 0.5771,
+      "step": 2429
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.2245350585387767,
+      "learning_rate": 5.355614305696468e-06,
+      "loss": 0.7899,
+      "step": 2430
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.20170960230556165,
+      "learning_rate": 5.339258126360169e-06,
+      "loss": 0.5424,
+      "step": 2431
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.19303805481708192,
+      "learning_rate": 5.322923974858784e-06,
+      "loss": 0.5489,
+      "step": 2432
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.23105859058784034,
+      "learning_rate": 5.306611869493044e-06,
+      "loss": 0.8198,
+      "step": 2433
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.1970413573654372,
+      "learning_rate": 5.290321828538988e-06,
+      "loss": 0.5463,
+      "step": 2434
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.19542300070357776,
+      "learning_rate": 5.274053870247914e-06,
+      "loss": 0.5468,
+      "step": 2435
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.2371995457234757,
+      "learning_rate": 5.257808012846402e-06,
+      "loss": 0.7997,
+      "step": 2436
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.1970900389667059,
+      "learning_rate": 5.241584274536259e-06,
+      "loss": 0.5593,
+      "step": 2437
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.2008622073178912,
+      "learning_rate": 5.225382673494514e-06,
+      "loss": 0.5699,
+      "step": 2438
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.214934234273939,
+      "learning_rate": 5.209203227873388e-06,
+      "loss": 0.6036,
+      "step": 2439
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.19482910144803964,
+      "learning_rate": 5.193045955800283e-06,
+      "loss": 0.5744,
+      "step": 2440
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.23578639050646755,
+      "learning_rate": 5.176910875377758e-06,
+      "loss": 0.7927,
+      "step": 2441
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.1991657634658738,
+      "learning_rate": 5.160798004683506e-06,
+      "loss": 0.5446,
+      "step": 2442
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.19433665313059084,
+      "learning_rate": 5.14470736177034e-06,
+      "loss": 0.5539,
+      "step": 2443
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.22718152175238573,
+      "learning_rate": 5.128638964666166e-06,
+      "loss": 0.7931,
+      "step": 2444
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.20311056648144807,
+      "learning_rate": 5.112592831373974e-06,
+      "loss": 0.5725,
+      "step": 2445
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.19410752027948577,
+      "learning_rate": 5.096568979871788e-06,
+      "loss": 0.5642,
+      "step": 2446
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.22116739920132403,
+      "learning_rate": 5.080567428112689e-06,
+      "loss": 0.7971,
+      "step": 2447
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.20256840096418977,
+      "learning_rate": 5.064588194024769e-06,
+      "loss": 0.558,
+      "step": 2448
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.1921589475836885,
+      "learning_rate": 5.048631295511111e-06,
+      "loss": 0.562,
+      "step": 2449
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.2144669968281014,
+      "learning_rate": 5.032696750449775e-06,
+      "loss": 0.6235,
+      "step": 2450
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.19610717400290142,
+      "learning_rate": 5.016784576693781e-06,
+      "loss": 0.5744,
+      "step": 2451
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.21910675535408722,
+      "learning_rate": 5.00089479207107e-06,
+      "loss": 0.7956,
+      "step": 2452
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.1974812534011117,
+      "learning_rate": 4.985027414384513e-06,
+      "loss": 0.5446,
+      "step": 2453
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.19334359504046716,
+      "learning_rate": 4.969182461411873e-06,
+      "loss": 0.5631,
+      "step": 2454
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.23468910829597006,
+      "learning_rate": 4.953359950905784e-06,
+      "loss": 0.8233,
+      "step": 2455
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.1955507349616059,
+      "learning_rate": 4.937559900593747e-06,
+      "loss": 0.5773,
+      "step": 2456
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.19842389540204114,
+      "learning_rate": 4.921782328178074e-06,
+      "loss": 0.5523,
+      "step": 2457
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.22540868486229937,
+      "learning_rate": 4.906027251335918e-06,
+      "loss": 0.8122,
+      "step": 2458
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.19563875266149003,
+      "learning_rate": 4.8902946877192185e-06,
+      "loss": 0.5629,
+      "step": 2459
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.19056436784540504,
+      "learning_rate": 4.874584654954692e-06,
+      "loss": 0.5558,
+      "step": 2460
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.22087744704745157,
+      "learning_rate": 4.8588971706438084e-06,
+      "loss": 0.6055,
+      "step": 2461
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.23536472941498854,
+      "learning_rate": 4.8432322523627784e-06,
+      "loss": 0.8216,
+      "step": 2462
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.2046483635165671,
+      "learning_rate": 4.827589917662531e-06,
+      "loss": 0.56,
+      "step": 2463
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.19836321639674004,
+      "learning_rate": 4.811970184068684e-06,
+      "loss": 0.5527,
+      "step": 2464
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.19377606892165772,
+      "learning_rate": 4.796373069081547e-06,
+      "loss": 0.5494,
+      "step": 2465
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.2236611068695306,
+      "learning_rate": 4.780798590176072e-06,
+      "loss": 0.7922,
+      "step": 2466
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.20190977658535006,
+      "learning_rate": 4.76524676480187e-06,
+      "loss": 0.5494,
+      "step": 2467
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.19599612119355392,
+      "learning_rate": 4.749717610383142e-06,
+      "loss": 0.5673,
+      "step": 2468
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.2221650150359786,
+      "learning_rate": 4.734211144318712e-06,
+      "loss": 0.7839,
+      "step": 2469
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.19613157443009704,
+      "learning_rate": 4.718727383981977e-06,
+      "loss": 0.5519,
+      "step": 2470
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.19575657158051363,
+      "learning_rate": 4.703266346720897e-06,
+      "loss": 0.5512,
+      "step": 2471
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.21647052587734847,
+      "learning_rate": 4.687828049857967e-06,
+      "loss": 0.6189,
+      "step": 2472
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.48380022181838905,
+      "learning_rate": 4.672412510690216e-06,
+      "loss": 0.7818,
+      "step": 2473
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.20271039187673381,
+      "learning_rate": 4.6570197464891555e-06,
+      "loss": 0.5736,
+      "step": 2474
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.19598913423677913,
+      "learning_rate": 4.641649774500795e-06,
+      "loss": 0.5751,
+      "step": 2475
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.21964446424268294,
+      "learning_rate": 4.626302611945604e-06,
+      "loss": 0.8275,
+      "step": 2476
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.20179609083477718,
+      "learning_rate": 4.610978276018496e-06,
+      "loss": 0.5761,
+      "step": 2477
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.1864869948987877,
+      "learning_rate": 4.595676783888814e-06,
+      "loss": 0.5558,
+      "step": 2478
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.22089695296755849,
+      "learning_rate": 4.580398152700305e-06,
+      "loss": 0.8144,
+      "step": 2479
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.19123261087502705,
+      "learning_rate": 4.56514239957109e-06,
+      "loss": 0.5457,
+      "step": 2480
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.19032855921483166,
+      "learning_rate": 4.549909541593675e-06,
+      "loss": 0.5562,
+      "step": 2481
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.21367686781594245,
+      "learning_rate": 4.534699595834905e-06,
+      "loss": 0.5608,
+      "step": 2482
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.22816945182825332,
+      "learning_rate": 4.519512579335955e-06,
+      "loss": 0.6111,
+      "step": 2483
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.216514126336908,
+      "learning_rate": 4.504348509112319e-06,
+      "loss": 0.788,
+      "step": 2484
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.2011552925772886,
+      "learning_rate": 4.489207402153767e-06,
+      "loss": 0.5523,
+      "step": 2485
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.19370893291637659,
+      "learning_rate": 4.474089275424351e-06,
+      "loss": 0.5634,
+      "step": 2486
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.2145691132227128,
+      "learning_rate": 4.458994145862377e-06,
+      "loss": 0.7989,
+      "step": 2487
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.20489012703875392,
+      "learning_rate": 4.443922030380379e-06,
+      "loss": 0.5526,
+      "step": 2488
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.1890535847724818,
+      "learning_rate": 4.428872945865112e-06,
+      "loss": 0.5737,
+      "step": 2489
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.22641023272811212,
+      "learning_rate": 4.413846909177524e-06,
+      "loss": 0.8291,
+      "step": 2490
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.19394101556989884,
+      "learning_rate": 4.398843937152736e-06,
+      "loss": 0.5628,
+      "step": 2491
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.19506804152888915,
+      "learning_rate": 4.38386404660003e-06,
+      "loss": 0.5761,
+      "step": 2492
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.215030539499329,
+      "learning_rate": 4.368907254302837e-06,
+      "loss": 0.788,
+      "step": 2493
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.18741944600457267,
+      "learning_rate": 4.353973577018694e-06,
+      "loss": 0.3826,
+      "step": 2494
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.22780967873320337,
+      "learning_rate": 4.339063031479254e-06,
+      "loss": 0.8125,
+      "step": 2495
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.19145020826231685,
+      "learning_rate": 4.324175634390249e-06,
+      "loss": 0.5549,
+      "step": 2496
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.18398145896908863,
+      "learning_rate": 4.3093114024314625e-06,
+      "loss": 0.5522,
+      "step": 2497
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.2193991863243341,
+      "learning_rate": 4.294470352256738e-06,
+      "loss": 0.8291,
+      "step": 2498
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.21347066004792303,
+      "learning_rate": 4.279652500493944e-06,
+      "loss": 0.5886,
+      "step": 2499
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.19510652422654756,
+      "learning_rate": 4.264857863744956e-06,
+      "loss": 0.5691,
+      "step": 2500
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.20072747440452138,
+      "learning_rate": 4.250086458585648e-06,
+      "loss": 0.5478,
+      "step": 2501
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.19604970809685177,
+      "learning_rate": 4.2353383015658416e-06,
+      "loss": 0.7439,
+      "step": 2502
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.1915655384997651,
+      "learning_rate": 4.220613409209334e-06,
+      "loss": 0.5588,
+      "step": 2503
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.22078019096832704,
+      "learning_rate": 4.205911798013843e-06,
+      "loss": 0.8118,
+      "step": 2504
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.24601699853072276,
+      "learning_rate": 4.191233484451021e-06,
+      "loss": 0.5968,
+      "step": 2505
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.18723603299238195,
+      "learning_rate": 4.1765784849664005e-06,
+      "loss": 0.5487,
+      "step": 2506
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.19431050504092343,
+      "learning_rate": 4.161946815979403e-06,
+      "loss": 0.5687,
+      "step": 2507
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.22303898989145768,
+      "learning_rate": 4.1473384938832934e-06,
+      "loss": 0.8244,
+      "step": 2508
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.19589710735446317,
+      "learning_rate": 4.132753535045197e-06,
+      "loss": 0.5514,
+      "step": 2509
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.2191972114306376,
+      "learning_rate": 4.118191955806058e-06,
+      "loss": 0.5876,
+      "step": 2510
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.19445444437933687,
+      "learning_rate": 4.103653772480626e-06,
+      "loss": 0.5623,
+      "step": 2511
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.18980990834008107,
+      "learning_rate": 4.089139001357439e-06,
+      "loss": 0.5767,
+      "step": 2512
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.22739756624448734,
+      "learning_rate": 4.074647658698796e-06,
+      "loss": 0.8022,
+      "step": 2513
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.19736890349171343,
+      "learning_rate": 4.060179760740751e-06,
+      "loss": 0.5755,
+      "step": 2514
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.19319463258721828,
+      "learning_rate": 4.0457353236931e-06,
+      "loss": 0.5674,
+      "step": 2515
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.2395255459373776,
+      "learning_rate": 4.0313143637393416e-06,
+      "loss": 0.8554,
+      "step": 2516
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.19597421210217533,
+      "learning_rate": 4.016916897036677e-06,
+      "loss": 0.5556,
+      "step": 2517
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.20155262868091806,
+      "learning_rate": 4.0025429397159896e-06,
+      "loss": 0.5623,
+      "step": 2518
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.2315582506847439,
+      "learning_rate": 3.988192507881805e-06,
+      "loss": 0.7937,
+      "step": 2519
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.19167948350044023,
+      "learning_rate": 3.973865617612308e-06,
+      "loss": 0.5457,
+      "step": 2520
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.21949845757416664,
+      "learning_rate": 3.9595622849593005e-06,
+      "loss": 0.608,
+      "step": 2521
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.19845282922453092,
+      "learning_rate": 3.945282525948199e-06,
+      "loss": 0.5592,
+      "step": 2522
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.327345479280492,
+      "learning_rate": 3.931026356577996e-06,
+      "loss": 0.5896,
+      "step": 2523
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.21953202362005317,
+      "learning_rate": 3.9167937928212635e-06,
+      "loss": 0.7593,
+      "step": 2524
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.18894099582057805,
+      "learning_rate": 3.902584850624119e-06,
+      "loss": 0.5614,
+      "step": 2525
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.1945062108344911,
+      "learning_rate": 3.888399545906221e-06,
+      "loss": 0.5625,
+      "step": 2526
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.22081375728771463,
+      "learning_rate": 3.87423789456074e-06,
+      "loss": 0.8115,
+      "step": 2527
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.19064509194968957,
+      "learning_rate": 3.860099912454346e-06,
+      "loss": 0.5652,
+      "step": 2528
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.20214620169425432,
+      "learning_rate": 3.845985615427197e-06,
+      "loss": 0.5595,
+      "step": 2529
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.24485348630574832,
+      "learning_rate": 3.831895019292897e-06,
+      "loss": 0.8106,
+      "step": 2530
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.19515011725682008,
+      "learning_rate": 3.817828139838514e-06,
+      "loss": 0.57,
+      "step": 2531
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.21914203518876338,
+      "learning_rate": 3.803784992824533e-06,
+      "loss": 0.5979,
+      "step": 2532
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.18791834004884161,
+      "learning_rate": 3.789765593984854e-06,
+      "loss": 0.5382,
+      "step": 2533
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.21699417191744455,
+      "learning_rate": 3.7757699590267704e-06,
+      "loss": 0.7985,
+      "step": 2534
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.19389076125358548,
+      "learning_rate": 3.7617981036309537e-06,
+      "loss": 0.5531,
+      "step": 2535
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.18649709793770364,
+      "learning_rate": 3.7478500434514147e-06,
+      "loss": 0.5432,
+      "step": 2536
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.19887840856231775,
+      "learning_rate": 3.7339257941155253e-06,
+      "loss": 0.5534,
+      "step": 2537
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.22353055841160666,
+      "learning_rate": 3.72002537122397e-06,
+      "loss": 0.8017,
+      "step": 2538
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.1861843329572144,
+      "learning_rate": 3.7061487903507412e-06,
+      "loss": 0.5393,
+      "step": 2539
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.22432979253811217,
+      "learning_rate": 3.692296067043116e-06,
+      "loss": 0.5565,
+      "step": 2540
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.22303386339856202,
+      "learning_rate": 3.6784672168216517e-06,
+      "loss": 0.7788,
+      "step": 2541
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.0955175675361795,
+      "learning_rate": 3.6646622551801345e-06,
+      "loss": 0.5686,
+      "step": 2542
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.22273743974527407,
+      "learning_rate": 3.6508811975856123e-06,
+      "loss": 0.6171,
+      "step": 2543
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.19035285629037502,
+      "learning_rate": 3.6371240594783335e-06,
+      "loss": 0.5579,
+      "step": 2544
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.21914858918129004,
+      "learning_rate": 3.623390856271755e-06,
+      "loss": 0.7896,
+      "step": 2545
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.19105988900010107,
+      "learning_rate": 3.609681603352519e-06,
+      "loss": 0.568,
+      "step": 2546
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.1906990288620984,
+      "learning_rate": 3.595996316080427e-06,
+      "loss": 0.5837,
+      "step": 2547
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.22366029717536412,
+      "learning_rate": 3.5823350097884346e-06,
+      "loss": 0.826,
+      "step": 2548
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.19844876015380775,
+      "learning_rate": 3.568697699782625e-06,
+      "loss": 0.5706,
+      "step": 2549
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.18785294613329298,
+      "learning_rate": 3.555084401342204e-06,
+      "loss": 0.554,
+      "step": 2550
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.21995917928287506,
+      "learning_rate": 3.5414951297194633e-06,
+      "loss": 0.7869,
+      "step": 2551
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.19367122620291935,
+      "learning_rate": 3.52792990013979e-06,
+      "loss": 0.5691,
+      "step": 2552
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.1954694571062416,
+      "learning_rate": 3.5143887278016126e-06,
+      "loss": 0.5761,
+      "step": 2553
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.21974066129841655,
+      "learning_rate": 3.5008716278764287e-06,
+      "loss": 0.6174,
+      "step": 2554
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.1913481052251383,
+      "learning_rate": 3.4873786155087516e-06,
+      "loss": 0.5614,
+      "step": 2555
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.22510452743975196,
+      "learning_rate": 3.4739097058161114e-06,
+      "loss": 0.8149,
+      "step": 2556
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.19006428795577066,
+      "learning_rate": 3.460464913889042e-06,
+      "loss": 0.5828,
+      "step": 2557
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.18834616583801078,
+      "learning_rate": 3.4470442547910293e-06,
+      "loss": 0.5727,
+      "step": 2558
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.22768128168433593,
+      "learning_rate": 3.433647743558552e-06,
+      "loss": 0.8178,
+      "step": 2559
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.1950549604483384,
+      "learning_rate": 3.420275395201014e-06,
+      "loss": 0.5589,
+      "step": 2560
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.18734918115744523,
+      "learning_rate": 3.4069272247007593e-06,
+      "loss": 0.5439,
+      "step": 2561
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.227979810522179,
+      "learning_rate": 3.393603247013033e-06,
+      "loss": 0.7899,
+      "step": 2562
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.19470895157277993,
+      "learning_rate": 3.3803034770659825e-06,
+      "loss": 0.5578,
+      "step": 2563
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.19184114485475345,
+      "learning_rate": 3.3670279297606243e-06,
+      "loss": 0.5884,
+      "step": 2564
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.21832656754879773,
+      "learning_rate": 3.3537766199708416e-06,
+      "loss": 0.6106,
+      "step": 2565
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.19833530694919116,
+      "learning_rate": 3.340549562543363e-06,
+      "loss": 0.5633,
+      "step": 2566
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.2267000701439619,
+      "learning_rate": 3.3273467722977432e-06,
+      "loss": 0.7875,
+      "step": 2567
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.19757290822953258,
+      "learning_rate": 3.314168264026346e-06,
+      "loss": 0.5598,
+      "step": 2568
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.19456676469673997,
+      "learning_rate": 3.301014052494336e-06,
+      "loss": 0.5588,
+      "step": 2569
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.2253266795472492,
+      "learning_rate": 3.2878841524396465e-06,
+      "loss": 0.8244,
+      "step": 2570
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.1956145493007036,
+      "learning_rate": 3.2747785785729818e-06,
+      "loss": 0.5349,
+      "step": 2571
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.19325862187975662,
+      "learning_rate": 3.261697345577783e-06,
+      "loss": 0.5608,
+      "step": 2572
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.22551153697755488,
+      "learning_rate": 3.2486404681102227e-06,
+      "loss": 0.8049,
+      "step": 2573
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.2121337267113259,
+      "learning_rate": 3.2356079607991972e-06,
+      "loss": 0.5515,
+      "step": 2574
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.19115267637885583,
+      "learning_rate": 3.2225998382462735e-06,
+      "loss": 0.5662,
+      "step": 2575
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.21632420224628054,
+      "learning_rate": 3.2096161150257153e-06,
+      "loss": 0.5981,
+      "step": 2576
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.2324232878798505,
+      "learning_rate": 3.19665680568445e-06,
+      "loss": 0.8101,
+      "step": 2577
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.1872838662118444,
+      "learning_rate": 3.1837219247420475e-06,
+      "loss": 0.5583,
+      "step": 2578
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.18642055662327994,
+      "learning_rate": 3.1708114866907055e-06,
+      "loss": 0.5422,
+      "step": 2579
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.19495835154477914,
+      "learning_rate": 3.157925505995249e-06,
+      "loss": 0.5649,
+      "step": 2580
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.2189211482541116,
+      "learning_rate": 3.1450639970930782e-06,
+      "loss": 0.771,
+      "step": 2581
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.20520184314462941,
+      "learning_rate": 3.1322269743941952e-06,
+      "loss": 0.5481,
+      "step": 2582
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.3437158874538998,
+      "learning_rate": 3.119414452281158e-06,
+      "loss": 0.5638,
+      "step": 2583
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.2197934848990998,
+      "learning_rate": 3.1066264451090815e-06,
+      "loss": 0.8108,
+      "step": 2584
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.19354454433439908,
+      "learning_rate": 3.0938629672056047e-06,
+      "loss": 0.5683,
+      "step": 2585
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.19070108123344862,
+      "learning_rate": 3.081124032870897e-06,
+      "loss": 0.5731,
+      "step": 2586
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.21186073089997828,
+      "learning_rate": 3.0684096563776103e-06,
+      "loss": 0.5728,
+      "step": 2587
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.22366620177262744,
+      "learning_rate": 3.055719851970898e-06,
+      "loss": 0.8223,
+      "step": 2588
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.19246565096064314,
+      "learning_rate": 3.043054633868381e-06,
+      "loss": 0.5512,
+      "step": 2589
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.18920611107501165,
+      "learning_rate": 3.0304140162601245e-06,
+      "loss": 0.5463,
+      "step": 2590
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.22250748187827635,
+      "learning_rate": 3.0177980133086453e-06,
+      "loss": 0.815,
+      "step": 2591
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.18937888102192593,
+      "learning_rate": 3.005206639148869e-06,
+      "loss": 0.5514,
+      "step": 2592
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.19169485953381835,
+      "learning_rate": 2.992639907888134e-06,
+      "loss": 0.5493,
+      "step": 2593
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.22734745738005424,
+      "learning_rate": 2.9800978336061687e-06,
+      "loss": 0.8098,
+      "step": 2594
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.18694698591091702,
+      "learning_rate": 2.9675804303550742e-06,
+      "loss": 0.5658,
+      "step": 2595
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.19773140945842094,
+      "learning_rate": 2.95508771215931e-06,
+      "loss": 0.5536,
+      "step": 2596
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.21949875665683913,
+      "learning_rate": 2.942619693015683e-06,
+      "loss": 0.7997,
+      "step": 2597
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.18782410025826793,
+      "learning_rate": 2.9301763868933157e-06,
+      "loss": 0.3723,
+      "step": 2598
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.21593545717718743,
+      "learning_rate": 2.9177578077336565e-06,
+      "loss": 0.8036,
+      "step": 2599
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.19961337008371907,
+      "learning_rate": 2.9053639694504392e-06,
+      "loss": 0.553,
+      "step": 2600
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.1936619180727206,
+      "learning_rate": 2.892994885929687e-06,
+      "loss": 0.5626,
+      "step": 2601
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.3559110827577502,
+      "learning_rate": 2.880650571029683e-06,
+      "loss": 0.8023,
+      "step": 2602
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.19179245752157584,
+      "learning_rate": 2.8683310385809543e-06,
+      "loss": 0.5439,
+      "step": 2603
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.22123226063334794,
+      "learning_rate": 2.8560363023862695e-06,
+      "loss": 0.5772,
+      "step": 2604
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.19055773332938916,
+      "learning_rate": 2.8437663762206163e-06,
+      "loss": 0.7067,
+      "step": 2605
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.19339986717458554,
+      "learning_rate": 2.831521273831178e-06,
+      "loss": 0.5467,
+      "step": 2606
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.1959624836129281,
+      "learning_rate": 2.8193010089373335e-06,
+      "loss": 0.5415,
+      "step": 2607
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.22398871526016348,
+      "learning_rate": 2.807105595230633e-06,
+      "loss": 0.8165,
+      "step": 2608
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.19030904310669886,
+      "learning_rate": 2.7949350463747693e-06,
+      "loss": 0.3831,
+      "step": 2609
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.2197624834720522,
+      "learning_rate": 2.7827893760055983e-06,
+      "loss": 0.8004,
+      "step": 2610
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.18665184252167874,
+      "learning_rate": 2.770668597731088e-06,
+      "loss": 0.5561,
+      "step": 2611
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.19141234958905906,
+      "learning_rate": 2.75857272513132e-06,
+      "loss": 0.548,
+      "step": 2612
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.2187016366997306,
+      "learning_rate": 2.7465017717584753e-06,
+      "loss": 0.8004,
+      "step": 2613
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.1920987524330266,
+      "learning_rate": 2.7344557511368135e-06,
+      "loss": 0.5508,
+      "step": 2614
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.22345559915216814,
+      "learning_rate": 2.7224346767626587e-06,
+      "loss": 0.5961,
+      "step": 2615
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.18934350580478412,
+      "learning_rate": 2.710438562104384e-06,
+      "loss": 0.5817,
+      "step": 2616
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.2162277958476069,
+      "learning_rate": 2.6984674206023993e-06,
+      "loss": 0.8002,
+      "step": 2617
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.192653228267827,
+      "learning_rate": 2.686521265669137e-06,
+      "loss": 0.5742,
+      "step": 2618
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.19658184354824615,
+      "learning_rate": 2.674600110689038e-06,
+      "loss": 0.7271,
+      "step": 2619
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.21707149561301212,
+      "learning_rate": 2.662703969018515e-06,
+      "loss": 0.5703,
+      "step": 2620
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.18445155836054786,
+      "learning_rate": 2.6508328539859766e-06,
+      "loss": 0.5589,
+      "step": 2621
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.18532059409623952,
+      "learning_rate": 2.6389867788917806e-06,
+      "loss": 0.5648,
+      "step": 2622
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.21929080774702614,
+      "learning_rate": 2.6271657570082346e-06,
+      "loss": 0.8188,
+      "step": 2623
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.2009131896787724,
+      "learning_rate": 2.615369801579576e-06,
+      "loss": 0.5451,
+      "step": 2624
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.18037920546413114,
+      "learning_rate": 2.6035989258219622e-06,
+      "loss": 0.5549,
+      "step": 2625
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.21884080614986076,
+      "learning_rate": 2.5918531429234368e-06,
+      "loss": 0.6052,
+      "step": 2626
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.18613162809126743,
+      "learning_rate": 2.5801324660439414e-06,
+      "loss": 0.5398,
+      "step": 2627
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.21627185549305025,
+      "learning_rate": 2.568436908315286e-06,
+      "loss": 0.8024,
+      "step": 2628
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.18623007873994724,
+      "learning_rate": 2.5567664828411414e-06,
+      "loss": 0.5586,
+      "step": 2629
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.18662601917021648,
+      "learning_rate": 2.5451212026970205e-06,
+      "loss": 0.5608,
+      "step": 2630
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.22768645725258746,
+      "learning_rate": 2.533501080930245e-06,
+      "loss": 0.8255,
+      "step": 2631
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.1845160309365859,
+      "learning_rate": 2.5219061305599755e-06,
+      "loss": 0.5591,
+      "step": 2632
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.22288827140623776,
+      "learning_rate": 2.510336364577154e-06,
+      "loss": 0.5618,
+      "step": 2633
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.23986213986310015,
+      "learning_rate": 2.4987917959445134e-06,
+      "loss": 0.8098,
+      "step": 2634
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.18528903418924814,
+      "learning_rate": 2.4872724375965523e-06,
+      "loss": 0.5564,
+      "step": 2635
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.18496387064126657,
+      "learning_rate": 2.475778302439524e-06,
+      "loss": 0.5443,
+      "step": 2636
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.22442498336238612,
+      "learning_rate": 2.4643094033514198e-06,
+      "loss": 0.5754,
+      "step": 2637
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.2530350342252588,
+      "learning_rate": 2.4528657531819656e-06,
+      "loss": 0.5592,
+      "step": 2638
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.22096191591175204,
+      "learning_rate": 2.441447364752586e-06,
+      "loss": 0.8305,
+      "step": 2639
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.19028871026339605,
+      "learning_rate": 2.430054250856412e-06,
+      "loss": 0.5612,
+      "step": 2640
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.1820282534912051,
+      "learning_rate": 2.4186864242582486e-06,
+      "loss": 0.5511,
+      "step": 2641
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.2231754153556352,
+      "learning_rate": 2.407343897694586e-06,
+      "loss": 0.8255,
+      "step": 2642
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.19307912986996192,
+      "learning_rate": 2.3960266838735395e-06,
+      "loss": 0.5541,
+      "step": 2643
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.18936619984135455,
+      "learning_rate": 2.384734795474891e-06,
+      "loss": 0.5571,
+      "step": 2644
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.22467620109331715,
+      "learning_rate": 2.3734682451500347e-06,
+      "loss": 0.7751,
+      "step": 2645
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.18921028386348226,
+      "learning_rate": 2.362227045521978e-06,
+      "loss": 0.5564,
+      "step": 2646
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.18723789526906043,
+      "learning_rate": 2.351011209185336e-06,
+      "loss": 0.5562,
+      "step": 2647
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.21969092811302268,
+      "learning_rate": 2.3398207487062814e-06,
+      "loss": 0.5912,
+      "step": 2648
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.22844569637615098,
+      "learning_rate": 2.328655676622582e-06,
+      "loss": 0.7927,
+      "step": 2649
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.19570540288231073,
+      "learning_rate": 2.3175160054435486e-06,
+      "loss": 0.5709,
+      "step": 2650
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.18548356589174064,
+      "learning_rate": 2.306401747650033e-06,
+      "loss": 0.5566,
+      "step": 2651
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.19128562268918237,
+      "learning_rate": 2.295312915694417e-06,
+      "loss": 0.552,
+      "step": 2652
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.22670449804443132,
+      "learning_rate": 2.2842495220005975e-06,
+      "loss": 0.814,
+      "step": 2653
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.188680445907785,
+      "learning_rate": 2.2732115789639603e-06,
+      "loss": 0.5597,
+      "step": 2654
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.19048328957066973,
+      "learning_rate": 2.262199098951384e-06,
+      "loss": 0.5601,
+      "step": 2655
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.21725844351685053,
+      "learning_rate": 2.2512120943012195e-06,
+      "loss": 0.8043,
+      "step": 2656
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.18750193906418236,
+      "learning_rate": 2.2402505773232745e-06,
+      "loss": 0.539,
+      "step": 2657
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.1935027260797383,
+      "learning_rate": 2.2293145602987954e-06,
+      "loss": 0.5642,
+      "step": 2658
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.2674026076227791,
+      "learning_rate": 2.2184040554804665e-06,
+      "loss": 0.5852,
+      "step": 2659
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.22798141636767943,
+      "learning_rate": 2.207519075092379e-06,
+      "loss": 0.8028,
+      "step": 2660
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.19494296619660706,
+      "learning_rate": 2.1966596313300365e-06,
+      "loss": 0.5452,
+      "step": 2661
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.19120882905054729,
+      "learning_rate": 2.1858257363603224e-06,
+      "loss": 0.5475,
+      "step": 2662
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.23094538007096263,
+      "learning_rate": 2.1750174023214996e-06,
+      "loss": 0.8049,
+      "step": 2663
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.18678283598585535,
+      "learning_rate": 2.1642346413232012e-06,
+      "loss": 0.5699,
+      "step": 2664
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.213587029056211,
+      "learning_rate": 2.153477465446385e-06,
+      "loss": 0.5508,
+      "step": 2665
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.21958481834184473,
+      "learning_rate": 2.1427458867433663e-06,
+      "loss": 0.7846,
+      "step": 2666
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.18774270751726677,
+      "learning_rate": 2.1320399172377744e-06,
+      "loss": 0.5779,
+      "step": 2667
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.18624109749008047,
+      "learning_rate": 2.1213595689245386e-06,
+      "loss": 0.5739,
+      "step": 2668
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.18053582975468893,
+      "learning_rate": 2.1107048537698933e-06,
+      "loss": 0.5468,
+      "step": 2669
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.21729233135510262,
+      "learning_rate": 2.10007578371135e-06,
+      "loss": 0.5762,
+      "step": 2670
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.22281581928779165,
+      "learning_rate": 2.08947237065768e-06,
+      "loss": 0.7908,
+      "step": 2671
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.19423402960968952,
+      "learning_rate": 2.0788946264889196e-06,
+      "loss": 0.5764,
+      "step": 2672
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.18796036225087367,
+      "learning_rate": 2.0683425630563384e-06,
+      "loss": 0.5622,
+      "step": 2673
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.22963946207541464,
+      "learning_rate": 2.057816192182438e-06,
+      "loss": 0.7976,
+      "step": 2674
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.1860393785329912,
+      "learning_rate": 2.0473155256609366e-06,
+      "loss": 0.5497,
+      "step": 2675
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.17945212794639626,
+      "learning_rate": 2.036840575256735e-06,
+      "loss": 0.542,
+      "step": 2676
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.22249281273000424,
+      "learning_rate": 2.0263913527059475e-06,
+      "loss": 0.8202,
+      "step": 2677
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.23096172912427654,
+      "learning_rate": 2.015967869715846e-06,
+      "loss": 0.5579,
+      "step": 2678
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.1861195863900527,
+      "learning_rate": 2.0055701379648712e-06,
+      "loss": 0.5396,
+      "step": 2679
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.32987267262032594,
+      "learning_rate": 1.9951981691026024e-06,
+      "loss": 0.8119,
+      "step": 2680
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.1896059657743882,
+      "learning_rate": 1.984851974749777e-06,
+      "loss": 0.3715,
+      "step": 2681
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.21770706667993892,
+      "learning_rate": 1.9745315664982276e-06,
+      "loss": 0.7892,
+      "step": 2682
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.1855037756986657,
+      "learning_rate": 1.9642369559109116e-06,
+      "loss": 0.5396,
+      "step": 2683
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.1925565837706489,
+      "learning_rate": 1.953968154521876e-06,
+      "loss": 0.5589,
+      "step": 2684
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.2222701710880819,
+      "learning_rate": 1.943725173836261e-06,
+      "loss": 0.8054,
+      "step": 2685
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.1875676402180658,
+      "learning_rate": 1.9335080253302632e-06,
+      "loss": 0.5474,
+      "step": 2686
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.181026406593221,
+      "learning_rate": 1.9233167204511543e-06,
+      "loss": 0.5553,
+      "step": 2687
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.22302317727790688,
+      "learning_rate": 1.9131512706172276e-06,
+      "loss": 0.8174,
+      "step": 2688
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.20649374935329393,
+      "learning_rate": 1.9030116872178316e-06,
+      "loss": 0.5537,
+      "step": 2689
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.18979244115982863,
+      "learning_rate": 1.892897981613323e-06,
+      "loss": 0.553,
+      "step": 2690
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.22801021503586136,
+      "learning_rate": 1.8828101651350627e-06,
+      "loss": 0.8227,
+      "step": 2691
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.24826598377009992,
+      "learning_rate": 1.8727482490854153e-06,
+      "loss": 0.6013,
+      "step": 2692
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.18590724404930142,
+      "learning_rate": 1.862712244737716e-06,
+      "loss": 0.5598,
+      "step": 2693
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.19834639076970828,
+      "learning_rate": 1.852702163336273e-06,
+      "loss": 0.5598,
+      "step": 2694
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.2164079392553112,
+      "learning_rate": 1.8427180160963537e-06,
+      "loss": 0.8006,
+      "step": 2695
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.18759723763930874,
+      "learning_rate": 1.8327598142041658e-06,
+      "loss": 0.567,
+      "step": 2696
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.20342443356229312,
+      "learning_rate": 1.8228275688168455e-06,
+      "loss": 0.5704,
+      "step": 2697
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.1868204097027238,
+      "learning_rate": 1.8129212910624554e-06,
+      "loss": 0.5648,
+      "step": 2698
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.21331880665356995,
+      "learning_rate": 1.8030409920399505e-06,
+      "loss": 0.8344,
+      "step": 2699
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.19536428586506427,
+      "learning_rate": 1.7931866828191923e-06,
+      "loss": 0.5463,
+      "step": 2700
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.18882890976076033,
+      "learning_rate": 1.7833583744409217e-06,
+      "loss": 0.5545,
+      "step": 2701
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.22101934322498212,
+      "learning_rate": 1.7735560779167381e-06,
+      "loss": 0.8062,
+      "step": 2702
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.2097407795515249,
+      "learning_rate": 1.7637798042291126e-06,
+      "loss": 0.576,
+      "step": 2703
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.18401279733122355,
+      "learning_rate": 1.7540295643313497e-06,
+      "loss": 0.5532,
+      "step": 2704
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.18292392157095363,
+      "learning_rate": 1.7443053691475835e-06,
+      "loss": 0.5485,
+      "step": 2705
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.22672893265407584,
+      "learning_rate": 1.7346072295727768e-06,
+      "loss": 0.8086,
+      "step": 2706
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.19645708324418024,
+      "learning_rate": 1.724935156472693e-06,
+      "loss": 0.5534,
+      "step": 2707
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.18960397574278148,
+      "learning_rate": 1.7152891606838944e-06,
+      "loss": 0.5459,
+      "step": 2708
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.2286764842050619,
+      "learning_rate": 1.7056692530137308e-06,
+      "loss": 0.796,
+      "step": 2709
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.18735657694258312,
+      "learning_rate": 1.6960754442403054e-06,
+      "loss": 0.5463,
+      "step": 2710
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.19378676990380947,
+      "learning_rate": 1.6865077451125005e-06,
+      "loss": 0.567,
+      "step": 2711
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.2175915353956136,
+      "learning_rate": 1.6769661663499331e-06,
+      "loss": 0.8084,
+      "step": 2712
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.18664355783297873,
+      "learning_rate": 1.66745071864296e-06,
+      "loss": 0.5525,
+      "step": 2713
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.21114707385607992,
+      "learning_rate": 1.6579614126526616e-06,
+      "loss": 0.5794,
+      "step": 2714
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.1809376329038445,
+      "learning_rate": 1.6484982590108277e-06,
+      "loss": 0.5453,
+      "step": 2715
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.18875443494808544,
+      "learning_rate": 1.639061268319944e-06,
+      "loss": 0.5832,
+      "step": 2716
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.22712459247139294,
+      "learning_rate": 1.6296504511531836e-06,
+      "loss": 0.8162,
+      "step": 2717
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.18380367769445763,
+      "learning_rate": 1.6202658180544005e-06,
+      "loss": 0.548,
+      "step": 2718
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.22017601854055455,
+      "learning_rate": 1.6109073795381097e-06,
+      "loss": 0.6094,
+      "step": 2719
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.1904710905479985,
+      "learning_rate": 1.6015751460894762e-06,
+      "loss": 0.5658,
+      "step": 2720
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.22956466798869277,
+      "learning_rate": 1.592269128164306e-06,
+      "loss": 0.797,
+      "step": 2721
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.1761445213663517,
+      "learning_rate": 1.5829893361890308e-06,
+      "loss": 0.5293,
+      "step": 2722
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.2244273949275136,
+      "learning_rate": 1.5737357805607005e-06,
+      "loss": 0.8079,
+      "step": 2723
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.18173620324384784,
+      "learning_rate": 1.5645084716469777e-06,
+      "loss": 0.5496,
+      "step": 2724
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.2102317128868273,
+      "learning_rate": 1.5553074197861034e-06,
+      "loss": 0.6196,
+      "step": 2725
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.18522281311766572,
+      "learning_rate": 1.546132635286912e-06,
+      "loss": 0.5688,
+      "step": 2726
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.18451115590671952,
+      "learning_rate": 1.5369841284288006e-06,
+      "loss": 0.5546,
+      "step": 2727
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.2174799747524659,
+      "learning_rate": 1.5278619094617308e-06,
+      "loss": 0.8144,
+      "step": 2728
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.18736318188932052,
+      "learning_rate": 1.5187659886062078e-06,
+      "loss": 0.5387,
+      "step": 2729
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.22164581091022562,
+      "learning_rate": 1.509696376053274e-06,
+      "loss": 0.5924,
+      "step": 2730
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.18650616555236035,
+      "learning_rate": 1.5006530819644925e-06,
+      "loss": 0.5671,
+      "step": 2731
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.23167947268334643,
+      "learning_rate": 1.4916361164719506e-06,
+      "loss": 0.8017,
+      "step": 2732
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.1857567510617165,
+      "learning_rate": 1.4826454896782194e-06,
+      "loss": 0.5725,
+      "step": 2733
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.18659436143220143,
+      "learning_rate": 1.4736812116563748e-06,
+      "loss": 0.5758,
+      "step": 2734
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.217935686172602,
+      "learning_rate": 1.4647432924499632e-06,
+      "loss": 0.8114,
+      "step": 2735
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.19009450115075152,
+      "learning_rate": 1.4558317420730022e-06,
+      "loss": 0.5886,
+      "step": 2736
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.19216224442866583,
+      "learning_rate": 1.4469465705099717e-06,
+      "loss": 0.5595,
+      "step": 2737
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.21388304616189913,
+      "learning_rate": 1.4380877877157834e-06,
+      "loss": 0.816,
+      "step": 2738
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.18577649407382676,
+      "learning_rate": 1.4292554036157925e-06,
+      "loss": 0.5606,
+      "step": 2739
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.18707255236178327,
+      "learning_rate": 1.4204494281057744e-06,
+      "loss": 0.5788,
+      "step": 2740
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.21617770765201214,
+      "learning_rate": 1.4116698710519183e-06,
+      "loss": 0.5639,
+      "step": 2741
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.1815913382061466,
+      "learning_rate": 1.4029167422908107e-06,
+      "loss": 0.5481,
+      "step": 2742
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.21517812405843303,
+      "learning_rate": 1.3941900516294348e-06,
+      "loss": 0.8011,
+      "step": 2743
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.1846555392261744,
+      "learning_rate": 1.3854898088451445e-06,
+      "loss": 0.5641,
+      "step": 2744
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.18769550986991115,
+      "learning_rate": 1.3768160236856675e-06,
+      "loss": 0.5737,
+      "step": 2745
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.22320985483139327,
+      "learning_rate": 1.3681687058690856e-06,
+      "loss": 0.7764,
+      "step": 2746
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.18949034681014426,
+      "learning_rate": 1.359547865083824e-06,
+      "loss": 0.5572,
+      "step": 2747
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.18375068382906912,
+      "learning_rate": 1.350953510988648e-06,
+      "loss": 0.5429,
+      "step": 2748
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.2317843169714528,
+      "learning_rate": 1.3423856532126523e-06,
+      "loss": 0.7978,
+      "step": 2749
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.1898334135517858,
+      "learning_rate": 1.3338443013552277e-06,
+      "loss": 0.5237,
+      "step": 2750
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.18730705475685225,
+      "learning_rate": 1.3253294649860831e-06,
+      "loss": 0.5716,
+      "step": 2751
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.21660884387330287,
+      "learning_rate": 1.3168411536452152e-06,
+      "loss": 0.5941,
+      "step": 2752
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.1818647696622766,
+      "learning_rate": 1.3083793768429026e-06,
+      "loss": 0.5591,
+      "step": 2753
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.2290034905001648,
+      "learning_rate": 1.2999441440596977e-06,
+      "loss": 0.814,
+      "step": 2754
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.18401326551812883,
+      "learning_rate": 1.291535464746399e-06,
+      "loss": 0.5583,
+      "step": 2755
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.21168168661644168,
+      "learning_rate": 1.2831533483240731e-06,
+      "loss": 0.5509,
+      "step": 2756
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.2215920784947388,
+      "learning_rate": 1.2747978041840164e-06,
+      "loss": 0.8079,
+      "step": 2757
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.18696534857872069,
+      "learning_rate": 1.2664688416877541e-06,
+      "loss": 0.5644,
+      "step": 2758
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.1805109432991267,
+      "learning_rate": 1.2581664701670298e-06,
+      "loss": 0.5409,
+      "step": 2759
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.2299196223201972,
+      "learning_rate": 1.2498906989238002e-06,
+      "loss": 0.8505,
+      "step": 2760
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.1854896212558077,
+      "learning_rate": 1.2416415372302038e-06,
+      "loss": 0.5424,
+      "step": 2761
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.17853476785341316,
+      "learning_rate": 1.2334189943285833e-06,
+      "loss": 0.5493,
+      "step": 2762
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.21770446861442172,
+      "learning_rate": 1.2252230794314507e-06,
+      "loss": 0.6133,
+      "step": 2763
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.2136364998441464,
+      "learning_rate": 1.2170538017214794e-06,
+      "loss": 0.7909,
+      "step": 2764
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.18712147033224544,
+      "learning_rate": 1.2089111703515067e-06,
+      "loss": 0.5791,
+      "step": 2765
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.19052641153939487,
+      "learning_rate": 1.2007951944445122e-06,
+      "loss": 0.5469,
+      "step": 2766
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.18382103230663718,
+      "learning_rate": 1.1927058830936089e-06,
+      "loss": 0.5522,
+      "step": 2767
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.22443694490625377,
+      "learning_rate": 1.184643245362038e-06,
+      "loss": 0.8096,
+      "step": 2768
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.1822196077535541,
+      "learning_rate": 1.1766072902831522e-06,
+      "loss": 0.5622,
+      "step": 2769
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.18329971563408237,
+      "learning_rate": 1.1685980268604097e-06,
+      "loss": 0.5547,
+      "step": 2770
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.22024315048575277,
+      "learning_rate": 1.1606154640673717e-06,
+      "loss": 0.813,
+      "step": 2771
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.18948460544443424,
+      "learning_rate": 1.1526596108476701e-06,
+      "loss": 0.5592,
+      "step": 2772
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.18741084145725928,
+      "learning_rate": 1.144730476115019e-06,
+      "loss": 0.5616,
+      "step": 2773
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.21104166449895506,
+      "learning_rate": 1.1368280687531985e-06,
+      "loss": 0.5924,
+      "step": 2774
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.22563317114292905,
+      "learning_rate": 1.1289523976160398e-06,
+      "loss": 0.8193,
+      "step": 2775
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.1830049186791831,
+      "learning_rate": 1.1211034715274238e-06,
+      "loss": 0.5538,
+      "step": 2776
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.18726517884758362,
+      "learning_rate": 1.113281299281263e-06,
+      "loss": 0.5703,
+      "step": 2777
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.22593210759689702,
+      "learning_rate": 1.1054858896414915e-06,
+      "loss": 0.8188,
+      "step": 2778
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.18435110021959894,
+      "learning_rate": 1.0977172513420646e-06,
+      "loss": 0.5479,
+      "step": 2779
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.1951602337262968,
+      "learning_rate": 1.0899753930869394e-06,
+      "loss": 0.5466,
+      "step": 2780
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.2169407982702769,
+      "learning_rate": 1.0822603235500744e-06,
+      "loss": 0.7673,
+      "step": 2781
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.181274692063615,
+      "learning_rate": 1.0745720513754082e-06,
+      "loss": 0.5493,
+      "step": 2782
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.1942333075729655,
+      "learning_rate": 1.066910585176853e-06,
+      "loss": 0.5447,
+      "step": 2783
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.22137065912123038,
+      "learning_rate": 1.0592759335382952e-06,
+      "loss": 0.8389,
+      "step": 2784
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.18751140559065932,
+      "learning_rate": 1.051668105013573e-06,
+      "loss": 0.39,
+      "step": 2785
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.221758573389987,
+      "learning_rate": 1.044087108126479e-06,
+      "loss": 0.7818,
+      "step": 2786
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.18790897019160413,
+      "learning_rate": 1.036532951370736e-06,
+      "loss": 0.571,
+      "step": 2787
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.18373205428792574,
+      "learning_rate": 1.0290056432099958e-06,
+      "loss": 0.5572,
+      "step": 2788
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.2241400997711331,
+      "learning_rate": 1.0215051920778346e-06,
+      "loss": 0.8171,
+      "step": 2789
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.18698516045926944,
+      "learning_rate": 1.0140316063777354e-06,
+      "loss": 0.5612,
+      "step": 2790
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.19329893631231385,
+      "learning_rate": 1.0065848944830786e-06,
+      "loss": 0.5679,
+      "step": 2791
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.22003253680093582,
+      "learning_rate": 9.99165064737137e-07,
+      "loss": 0.8062,
+      "step": 2792
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.18923389625828896,
+      "learning_rate": 9.917721254530722e-07,
+      "loss": 0.5865,
+      "step": 2793
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.18388652453907103,
+      "learning_rate": 9.844060849138997e-07,
+      "loss": 0.5518,
+      "step": 2794
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.2180912599445104,
+      "learning_rate": 9.770669513725128e-07,
+      "loss": 0.809,
+      "step": 2795
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.18895971966853234,
+      "learning_rate": 9.69754733051656e-07,
+      "loss": 0.3764,
+      "step": 2796
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.21480497302904172,
+      "learning_rate": 9.624694381439147e-07,
+      "loss": 0.7874,
+      "step": 2797
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.18405213788143235,
+      "learning_rate": 9.552110748117126e-07,
+      "loss": 0.5599,
+      "step": 2798
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.18043510878047675,
+      "learning_rate": 9.479796511872974e-07,
+      "loss": 0.5608,
+      "step": 2799
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.21986557594525166,
+      "learning_rate": 9.407751753727323e-07,
+      "loss": 0.8007,
+      "step": 2800
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.18423846862577428,
+      "learning_rate": 9.335976554398912e-07,
+      "loss": 0.5751,
+      "step": 2801
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.1912129788999045,
+      "learning_rate": 9.264470994304441e-07,
+      "loss": 0.5526,
+      "step": 2802
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.21772406367606562,
+      "learning_rate": 9.19323515355855e-07,
+      "loss": 0.8145,
+      "step": 2803
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.18316560747942798,
+      "learning_rate": 9.122269111973614e-07,
+      "loss": 0.5614,
+      "step": 2804
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.18989162568609833,
+      "learning_rate": 9.051572949059839e-07,
+      "loss": 0.5441,
+      "step": 2805
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.2242908739563131,
+      "learning_rate": 8.981146744024921e-07,
+      "loss": 0.7629,
+      "step": 2806
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.22120897589671715,
+      "learning_rate": 8.910990575774159e-07,
+      "loss": 0.5888,
+      "step": 2807
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.18550390152322763,
+      "learning_rate": 8.841104522910343e-07,
+      "loss": 0.5632,
+      "step": 2808
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.18598246514287667,
+      "learning_rate": 8.77148866373359e-07,
+      "loss": 0.5462,
+      "step": 2809
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.22543479839454159,
+      "learning_rate": 8.702143076241287e-07,
+      "loss": 0.7826,
+      "step": 2810
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.1830381036588974,
+      "learning_rate": 8.633067838128034e-07,
+      "loss": 0.5406,
+      "step": 2811
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.19045755960087926,
+      "learning_rate": 8.564263026785479e-07,
+      "loss": 0.5659,
+      "step": 2812
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.18244825782484442,
+      "learning_rate": 8.495728719302348e-07,
+      "loss": 0.5455,
+      "step": 2813
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.22356515864997656,
+      "learning_rate": 8.427464992464273e-07,
+      "loss": 0.8438,
+      "step": 2814
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.18160179118686584,
+      "learning_rate": 8.359471922753715e-07,
+      "loss": 0.5631,
+      "step": 2815
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.1835505546723046,
+      "learning_rate": 8.291749586349957e-07,
+      "loss": 0.566,
+      "step": 2816
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.22382856344916377,
+      "learning_rate": 8.224298059128777e-07,
+      "loss": 0.7873,
+      "step": 2817
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.2328937244054098,
+      "learning_rate": 8.157117416662724e-07,
+      "loss": 0.5922,
+      "step": 2818
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.19134088171657768,
+      "learning_rate": 8.090207734220811e-07,
+      "loss": 0.54,
+      "step": 2819
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.18105850602261336,
+      "learning_rate": 8.023569086768406e-07,
+      "loss": 0.5378,
+      "step": 2820
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.21451638460411535,
+      "learning_rate": 7.957201548967258e-07,
+      "loss": 0.7833,
+      "step": 2821
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.1820335687118869,
+      "learning_rate": 7.891105195175358e-07,
+      "loss": 0.5567,
+      "step": 2822
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.1797094405178482,
+      "learning_rate": 7.82528009944683e-07,
+      "loss": 0.5565,
+      "step": 2823
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.21275851502695084,
+      "learning_rate": 7.759726335531958e-07,
+      "loss": 0.578,
+      "step": 2824
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.1841704043213033,
+      "learning_rate": 7.694443976876908e-07,
+      "loss": 0.7121,
+      "step": 2825
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.18971953121409812,
+      "learning_rate": 7.629433096623895e-07,
+      "loss": 0.5677,
+      "step": 2826
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.220746760108773,
+      "learning_rate": 7.564693767610903e-07,
+      "loss": 0.8013,
+      "step": 2827
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.1817034181827822,
+      "learning_rate": 7.500226062371662e-07,
+      "loss": 0.5534,
+      "step": 2828
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.21414751074835145,
+      "learning_rate": 7.43603005313559e-07,
+      "loss": 0.5963,
+      "step": 2829
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.17955333769281137,
+      "learning_rate": 7.372105811827679e-07,
+      "loss": 0.5614,
+      "step": 2830
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.179199804465308,
+      "learning_rate": 7.3084534100685e-07,
+      "loss": 0.5562,
+      "step": 2831
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.21879293671204503,
+      "learning_rate": 7.245072919173979e-07,
+      "loss": 0.781,
+      "step": 2832
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.18449616859727316,
+      "learning_rate": 7.181964410155423e-07,
+      "loss": 0.5555,
+      "step": 2833
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.1908742701242775,
+      "learning_rate": 7.11912795371944e-07,
+      "loss": 0.561,
+      "step": 2834
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.2103936308886009,
+      "learning_rate": 7.05656362026777e-07,
+      "loss": 0.613,
+      "step": 2835
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.21164593752323774,
+      "learning_rate": 6.994271479897314e-07,
+      "loss": 0.7673,
+      "step": 2836
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.18293385513191457,
+      "learning_rate": 6.932251602399991e-07,
+      "loss": 0.5581,
+      "step": 2837
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.18539215694285777,
+      "learning_rate": 6.870504057262695e-07,
+      "loss": 0.5526,
+      "step": 2838
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.1784151138659564,
+      "learning_rate": 6.809028913667165e-07,
+      "loss": 0.5271,
+      "step": 2839
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.2246974776221031,
+      "learning_rate": 6.747826240489974e-07,
+      "loss": 0.808,
+      "step": 2840
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.17500156983267598,
+      "learning_rate": 6.68689610630241e-07,
+      "loss": 0.5477,
+      "step": 2841
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.18473460077243853,
+      "learning_rate": 6.62623857937042e-07,
+      "loss": 0.5814,
+      "step": 2842
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.21959494705883004,
+      "learning_rate": 6.565853727654503e-07,
+      "loss": 0.7919,
+      "step": 2843
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.19130102867994564,
+      "learning_rate": 6.505741618809652e-07,
+      "loss": 0.5606,
+      "step": 2844
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.19146688890569988,
+      "learning_rate": 6.445902320185299e-07,
+      "loss": 0.5519,
+      "step": 2845
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.21277002569156214,
+      "learning_rate": 6.386335898825202e-07,
+      "loss": 0.5825,
+      "step": 2846
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.21650725555242786,
+      "learning_rate": 6.327042421467394e-07,
+      "loss": 0.763,
+      "step": 2847
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.18272598276814775,
+      "learning_rate": 6.268021954544096e-07,
+      "loss": 0.5509,
+      "step": 2848
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.1820467788018498,
+      "learning_rate": 6.209274564181716e-07,
+      "loss": 0.5805,
+      "step": 2849
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.21863990639762787,
+      "learning_rate": 6.150800316200605e-07,
+      "loss": 0.7994,
+      "step": 2850
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.18649075407459303,
+      "learning_rate": 6.092599276115107e-07,
+      "loss": 0.5508,
+      "step": 2851
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.17943273407165075,
+      "learning_rate": 6.034671509133533e-07,
+      "loss": 0.5395,
+      "step": 2852
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.2195883333263948,
+      "learning_rate": 5.977017080157965e-07,
+      "loss": 0.7943,
+      "step": 2853
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.17605798480329862,
+      "learning_rate": 5.919636053784233e-07,
+      "loss": 0.5316,
+      "step": 2854
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.19079189848979644,
+      "learning_rate": 5.862528494301883e-07,
+      "loss": 0.5813,
+      "step": 2855
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.18014446536295453,
+      "learning_rate": 5.805694465694012e-07,
+      "loss": 0.5518,
+      "step": 2856
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.2138360289525479,
+      "learning_rate": 5.749134031637349e-07,
+      "loss": 0.5888,
+      "step": 2857
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.3367883119677382,
+      "learning_rate": 5.69284725550201e-07,
+      "loss": 0.8084,
+      "step": 2858
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.1802339275691405,
+      "learning_rate": 5.636834200351493e-07,
+      "loss": 0.5662,
+      "step": 2859
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.26731480718098616,
+      "learning_rate": 5.581094928942682e-07,
+      "loss": 0.555,
+      "step": 2860
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.26418521426947805,
+      "learning_rate": 5.525629503725705e-07,
+      "loss": 0.8028,
+      "step": 2861
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.18387844260692288,
+      "learning_rate": 5.470437986843824e-07,
+      "loss": 0.5421,
+      "step": 2862
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.18106662296944584,
+      "learning_rate": 5.415520440133437e-07,
+      "loss": 0.5465,
+      "step": 2863
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.2196067633670654,
+      "learning_rate": 5.360876925123992e-07,
+      "loss": 0.822,
+      "step": 2864
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.17879030574553978,
+      "learning_rate": 5.306507503037961e-07,
+      "loss": 0.5387,
+      "step": 2865
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.1942320051460634,
+      "learning_rate": 5.252412234790644e-07,
+      "loss": 0.5666,
+      "step": 2866
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.21699901497270033,
+      "learning_rate": 5.198591180990253e-07,
+      "loss": 0.8091,
+      "step": 2867
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.18048639592947135,
+      "learning_rate": 5.145044401937665e-07,
+      "loss": 0.3731,
+      "step": 2868
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.2199827326638059,
+      "learning_rate": 5.091771957626557e-07,
+      "loss": 0.8108,
+      "step": 2869
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.1836358966517926,
+      "learning_rate": 5.038773907743238e-07,
+      "loss": 0.5544,
+      "step": 2870
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.18329070485219945,
+      "learning_rate": 4.986050311666518e-07,
+      "loss": 0.5821,
+      "step": 2871
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.21310567240630432,
+      "learning_rate": 4.93360122846781e-07,
+      "loss": 0.7927,
+      "step": 2872
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.18524116563899448,
+      "learning_rate": 4.881426716910831e-07,
+      "loss": 0.5581,
+      "step": 2873
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.17829501116652116,
+      "learning_rate": 4.829526835451792e-07,
+      "loss": 0.5576,
+      "step": 2874
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.22007704917014162,
+      "learning_rate": 4.777901642239124e-07,
+      "loss": 0.7948,
+      "step": 2875
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.31802658144646595,
+      "learning_rate": 4.726551195113588e-07,
+      "loss": 0.5466,
+      "step": 2876
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.18164513138567995,
+      "learning_rate": 4.6754755516080497e-07,
+      "loss": 0.5613,
+      "step": 2877
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.21564872717599803,
+      "learning_rate": 4.6246747689474847e-07,
+      "loss": 0.7946,
+      "step": 2878
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.21767538381322019,
+      "learning_rate": 4.574148904048947e-07,
+      "loss": 0.6129,
+      "step": 2879
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.17793860239750894,
+      "learning_rate": 4.52389801352146e-07,
+      "loss": 0.5556,
+      "step": 2880
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.18175843055527507,
+      "learning_rate": 4.473922153665988e-07,
+      "loss": 0.5355,
+      "step": 2881
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.22446658836277464,
+      "learning_rate": 4.4242213804752685e-07,
+      "loss": 0.818,
+      "step": 2882
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.18014427338061503,
+      "learning_rate": 4.374795749633981e-07,
+      "loss": 0.5598,
+      "step": 2883
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.18923768060115384,
+      "learning_rate": 4.325645316518384e-07,
+      "loss": 0.5701,
+      "step": 2884
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.18462021958998662,
+      "learning_rate": 4.2767701361964843e-07,
+      "loss": 0.564,
+      "step": 2885
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.2286095034628065,
+      "learning_rate": 4.228170263427894e-07,
+      "loss": 0.7704,
+      "step": 2886
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.18429549673661172,
+      "learning_rate": 4.17984575266378e-07,
+      "loss": 0.5544,
+      "step": 2887
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.17989872185911274,
+      "learning_rate": 4.1317966580467495e-07,
+      "loss": 0.5498,
+      "step": 2888
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.21905818407070252,
+      "learning_rate": 4.0840230334108775e-07,
+      "loss": 0.8062,
+      "step": 2889
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.21231090014975584,
+      "learning_rate": 4.0365249322815715e-07,
+      "loss": 0.5661,
+      "step": 2890
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.18370846868724616,
+      "learning_rate": 3.9893024078755393e-07,
+      "loss": 0.563,
+      "step": 2891
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.18473284524437425,
+      "learning_rate": 3.9423555131007925e-07,
+      "loss": 0.5496,
+      "step": 2892
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.21532837697366616,
+      "learning_rate": 3.8956843005564793e-07,
+      "loss": 0.8002,
+      "step": 2893
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.18277856755544825,
+      "learning_rate": 3.8492888225328817e-07,
+      "loss": 0.537,
+      "step": 2894
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.18237770017722474,
+      "learning_rate": 3.803169131011336e-07,
+      "loss": 0.5406,
+      "step": 2895
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.21743524588657775,
+      "learning_rate": 3.757325277664231e-07,
+      "loss": 0.8106,
+      "step": 2896
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.18259397535370797,
+      "learning_rate": 3.711757313854841e-07,
+      "loss": 0.5607,
+      "step": 2897
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.18429489739916607,
+      "learning_rate": 3.666465290637411e-07,
+      "loss": 0.54,
+      "step": 2898
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.22511842077846816,
+      "learning_rate": 3.6214492587569316e-07,
+      "loss": 0.8374,
+      "step": 2899
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.18389411509580178,
+      "learning_rate": 3.576709268649309e-07,
+      "loss": 0.5703,
+      "step": 2900
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.21285632810951138,
+      "learning_rate": 3.5322453704410286e-07,
+      "loss": 0.5851,
+      "step": 2901
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.17930172388743737,
+      "learning_rate": 3.488057613949325e-07,
+      "loss": 0.5672,
+      "step": 2902
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.19341618944748504,
+      "learning_rate": 3.444146048682012e-07,
+      "loss": 0.5501,
+      "step": 2903
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.21515361064013058,
+      "learning_rate": 3.4005107238374854e-07,
+      "loss": 0.7566,
+      "step": 2904
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.18822396652180348,
+      "learning_rate": 3.3571516883046386e-07,
+      "loss": 0.5626,
+      "step": 2905
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.18311314926845912,
+      "learning_rate": 3.3140689906628054e-07,
+      "loss": 0.5306,
+      "step": 2906
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.21747048825103263,
+      "learning_rate": 3.2712626791816524e-07,
+      "loss": 0.8055,
+      "step": 2907
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.17933915885151824,
+      "learning_rate": 3.2287328018212883e-07,
+      "loss": 0.5464,
+      "step": 2908
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.1805761056323616,
+      "learning_rate": 3.186479406232068e-07,
+      "loss": 0.5559,
+      "step": 2909
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.23046675980417322,
+      "learning_rate": 3.14450253975454e-07,
+      "loss": 0.8034,
+      "step": 2910
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.18716346978488074,
+      "learning_rate": 3.1028022494194454e-07,
+      "loss": 0.5393,
+      "step": 2911
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.20925650717778274,
+      "learning_rate": 3.061378581947716e-07,
+      "loss": 0.564,
+      "step": 2912
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.18075645512581315,
+      "learning_rate": 3.020231583750255e-07,
+      "loss": 0.5431,
+      "step": 2913
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.28051814757126914,
+      "learning_rate": 2.9793613009280466e-07,
+      "loss": 0.5533,
+      "step": 2914
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.21649372410485054,
+      "learning_rate": 2.938767779272017e-07,
+      "loss": 0.8103,
+      "step": 2915
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.19000885122562405,
+      "learning_rate": 2.898451064263036e-07,
+      "loss": 0.5509,
+      "step": 2916
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.1831720060325967,
+      "learning_rate": 2.8584112010718314e-07,
+      "loss": 0.5455,
+      "step": 2917
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.22204365350171565,
+      "learning_rate": 2.818648234558935e-07,
+      "loss": 0.835,
+      "step": 2918
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.18207133074318627,
+      "learning_rate": 2.779162209274655e-07,
+      "loss": 0.5586,
+      "step": 2919
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.1812308631692644,
+      "learning_rate": 2.739953169458992e-07,
+      "loss": 0.5605,
+      "step": 2920
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.22582572835777132,
+      "learning_rate": 2.7010211590416946e-07,
+      "loss": 0.7635,
+      "step": 2921
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.18651202429360425,
+      "learning_rate": 2.66236622164201e-07,
+      "loss": 0.5725,
+      "step": 2922
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.21596160364844,
+      "learning_rate": 2.62398840056885e-07,
+      "loss": 0.6034,
+      "step": 2923
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.18353758531944053,
+      "learning_rate": 2.5858877388205695e-07,
+      "loss": 0.5659,
+      "step": 2924
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.22652559765790123,
+      "learning_rate": 2.5480642790850493e-07,
+      "loss": 0.7769,
+      "step": 2925
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.21353457514806692,
+      "learning_rate": 2.5105180637395864e-07,
+      "loss": 0.568,
+      "step": 2926
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.1793619775302095,
+      "learning_rate": 2.473249134850808e-07,
+      "loss": 0.5491,
+      "step": 2927
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.187860079694905,
+      "learning_rate": 2.436257534174785e-07,
+      "loss": 0.5794,
+      "step": 2928
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.21885155198204423,
+      "learning_rate": 2.399543303156726e-07,
+      "loss": 0.8076,
+      "step": 2929
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.1879189469980571,
+      "learning_rate": 2.363106482931171e-07,
+      "loss": 0.5701,
+      "step": 2930
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.1844868714376107,
+      "learning_rate": 2.3269471143218256e-07,
+      "loss": 0.5633,
+      "step": 2931
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.2204277158901944,
+      "learning_rate": 2.2910652378415055e-07,
+      "loss": 0.7789,
+      "step": 2932
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.1876647239626229,
+      "learning_rate": 2.255460893692163e-07,
+      "loss": 0.5677,
+      "step": 2933
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.21038024752174844,
+      "learning_rate": 2.2201341217648331e-07,
+      "loss": 0.5913,
+      "step": 2934
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.1786816654033123,
+      "learning_rate": 2.185084961639494e-07,
+      "loss": 0.5463,
+      "step": 2935
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.21556264669646846,
+      "learning_rate": 2.150313452585123e-07,
+      "loss": 0.827,
+      "step": 2936
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.18464415391413702,
+      "learning_rate": 2.1158196335595847e-07,
+      "loss": 0.5727,
+      "step": 2937
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.1797963011345438,
+      "learning_rate": 2.081603543209687e-07,
+      "loss": 0.5549,
+      "step": 2938
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.22888943171657974,
+      "learning_rate": 2.0476652198710144e-07,
+      "loss": 0.612,
+      "step": 2939
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.18248936282642464,
+      "learning_rate": 2.0140047015679843e-07,
+      "loss": 0.5487,
+      "step": 2940
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.2291274577732883,
+      "learning_rate": 1.9806220260137064e-07,
+      "loss": 0.8473,
+      "step": 2941
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.18006721160962627,
+      "learning_rate": 1.9475172306100408e-07,
+      "loss": 0.5612,
+      "step": 2942
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.1842870975383117,
+      "learning_rate": 1.9146903524474845e-07,
+      "loss": 0.7149,
+      "step": 2943
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.18734484043410826,
+      "learning_rate": 1.8821414283052007e-07,
+      "loss": 0.5705,
+      "step": 2944
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.2212412106877855,
+      "learning_rate": 1.8498704946508793e-07,
+      "loss": 0.6503,
+      "step": 2945
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.18283335720038996,
+      "learning_rate": 1.817877587640793e-07,
+      "loss": 0.5442,
+      "step": 2946
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.220490782938471,
+      "learning_rate": 1.7861627431196849e-07,
+      "loss": 0.8147,
+      "step": 2947
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.17605840094200773,
+      "learning_rate": 1.7547259966207708e-07,
+      "loss": 0.5361,
+      "step": 2948
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.18180081953793115,
+      "learning_rate": 1.7235673833656819e-07,
+      "loss": 0.5553,
+      "step": 2949
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.21480046904858568,
+      "learning_rate": 1.6926869382644372e-07,
+      "loss": 0.5849,
+      "step": 2950
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.2278186392132417,
+      "learning_rate": 1.662084695915417e-07,
+      "loss": 0.8038,
+      "step": 2951
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.18299749105970423,
+      "learning_rate": 1.631760690605222e-07,
+      "loss": 0.5609,
+      "step": 2952
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.17988103746303633,
+      "learning_rate": 1.601714956308814e-07,
+      "loss": 0.5433,
+      "step": 2953
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.1787660817206703,
+      "learning_rate": 1.571947526689349e-07,
+      "loss": 0.5524,
+      "step": 2954
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.21693671649795568,
+      "learning_rate": 1.5424584350981487e-07,
+      "loss": 0.7963,
+      "step": 2955
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.18224914540128717,
+      "learning_rate": 1.5132477145747003e-07,
+      "loss": 0.5442,
+      "step": 2956
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.2085284639477484,
+      "learning_rate": 1.4843153978466306e-07,
+      "loss": 0.5504,
+      "step": 2957
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.21459808712805292,
+      "learning_rate": 1.455661517329593e-07,
+      "loss": 0.8148,
+      "step": 2958
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.18922451417254293,
+      "learning_rate": 1.4272861051272958e-07,
+      "loss": 0.5529,
+      "step": 2959
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.18255165900859302,
+      "learning_rate": 1.399189193031475e-07,
+      "loss": 0.5662,
+      "step": 2960
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.20940323092868626,
+      "learning_rate": 1.3713708125218383e-07,
+      "loss": 0.5906,
+      "step": 2961
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.22020606826952158,
+      "learning_rate": 1.343830994765982e-07,
+      "loss": 0.8212,
+      "step": 2962
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.17586561568986286,
+      "learning_rate": 1.3165697706194746e-07,
+      "loss": 0.5345,
+      "step": 2963
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.18205795864987376,
+      "learning_rate": 1.289587170625689e-07,
+      "loss": 0.559,
+      "step": 2964
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.21517509435454615,
+      "learning_rate": 1.2628832250158594e-07,
+      "loss": 0.8113,
+      "step": 2965
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.18204330081425465,
+      "learning_rate": 1.23645796370897e-07,
+      "loss": 0.5642,
+      "step": 2966
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.1835898291438336,
+      "learning_rate": 1.2103114163118378e-07,
+      "loss": 0.5418,
+      "step": 2967
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.2213142180120702,
+      "learning_rate": 1.1844436121189739e-07,
+      "loss": 0.8151,
+      "step": 2968
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.1858007184625043,
+      "learning_rate": 1.1588545801125838e-07,
+      "loss": 0.5782,
+      "step": 2969
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.3471877393891899,
+      "learning_rate": 1.133544348962512e-07,
+      "loss": 0.5615,
+      "step": 2970
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.21363117644591517,
+      "learning_rate": 1.1085129470262689e-07,
+      "loss": 0.8067,
+      "step": 2971
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.17983493244787724,
+      "learning_rate": 1.0837604023490045e-07,
+      "loss": 0.3781,
+      "step": 2972
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.21274299385893355,
+      "learning_rate": 1.0592867426633402e-07,
+      "loss": 0.7951,
+      "step": 2973
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.18344436399900854,
+      "learning_rate": 1.0350919953895366e-07,
+      "loss": 0.542,
+      "step": 2974
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.18388972009621896,
+      "learning_rate": 1.011176187635271e-07,
+      "loss": 0.5727,
+      "step": 2975
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.21416750988079356,
+      "learning_rate": 9.87539346195776e-08,
+      "loss": 0.7896,
+      "step": 2976
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.17599970922672367,
+      "learning_rate": 9.641814975536734e-08,
+      "loss": 0.5538,
+      "step": 2977
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.18620757234883817,
+      "learning_rate": 9.411026678790291e-08,
+      "loss": 0.5665,
+      "step": 2978
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.219315766832973,
+      "learning_rate": 9.183028830293261e-08,
+      "loss": 0.7925,
+      "step": 2979
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.19453753858518777,
+      "learning_rate": 8.957821685493251e-08,
+      "loss": 0.5674,
+      "step": 2980
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.18218155415702708,
+      "learning_rate": 8.735405496711758e-08,
+      "loss": 0.5749,
+      "step": 2981
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.21877112264664517,
+      "learning_rate": 8.515780513143334e-08,
+      "loss": 0.7792,
+      "step": 2982
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.18030983035469966,
+      "learning_rate": 8.298946980855315e-08,
+      "loss": 0.3789,
+      "step": 2983
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.21777807160391482,
+      "learning_rate": 8.084905142786703e-08,
+      "loss": 0.7768,
+      "step": 2984
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.18383188893511418,
+      "learning_rate": 7.873655238750111e-08,
+      "loss": 0.5349,
+      "step": 2985
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.18387859522589065,
+      "learning_rate": 7.665197505428711e-08,
+      "loss": 0.547,
+      "step": 2986
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.2151114500309983,
+      "learning_rate": 7.459532176378181e-08,
+      "loss": 0.8109,
+      "step": 2987
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.18369010679164222,
+      "learning_rate": 7.25665948202503e-08,
+      "loss": 0.5626,
+      "step": 2988
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.1833098448769581,
+      "learning_rate": 7.056579649667716e-08,
+      "loss": 0.5515,
+      "step": 2989
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.21974073155383017,
+      "learning_rate": 6.859292903474702e-08,
+      "loss": 0.7991,
+      "step": 2990
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.17850120825949012,
+      "learning_rate": 6.664799464485283e-08,
+      "loss": 0.562,
+      "step": 2991
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.17953284498748617,
+      "learning_rate": 6.473099550609874e-08,
+      "loss": 0.5391,
+      "step": 2992
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.26728785169215774,
+      "learning_rate": 6.284193376627779e-08,
+      "loss": 0.804,
+      "step": 2993
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.22636486955316565,
+      "learning_rate": 6.098081154189695e-08,
+      "loss": 0.6109,
+      "step": 2994
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.17824914244279016,
+      "learning_rate": 5.9147630918143794e-08,
+      "loss": 0.5377,
+      "step": 2995
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.18554514166588904,
+      "learning_rate": 5.7342393948914276e-08,
+      "loss": 0.5412,
+      "step": 2996
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.21696237871323829,
+      "learning_rate": 5.5565102656787714e-08,
+      "loss": 0.8194,
+      "step": 2997
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.18318204694595264,
+      "learning_rate": 5.381575903303515e-08,
+      "loss": 0.544,
+      "step": 2998
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.18197141109052437,
+      "learning_rate": 5.209436503761655e-08,
+      "loss": 0.5362,
+      "step": 2999
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.18452936137417378,
+      "learning_rate": 5.040092259917528e-08,
+      "loss": 0.5563,
+      "step": 3000
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.21650427605893252,
+      "learning_rate": 4.873543361504362e-08,
+      "loss": 0.8174,
+      "step": 3001
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.1812306019174369,
+      "learning_rate": 4.709789995122615e-08,
+      "loss": 0.5402,
+      "step": 3002
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.17140905150496832,
+      "learning_rate": 4.548832344241083e-08,
+      "loss": 0.5368,
+      "step": 3003
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.21885359085546408,
+      "learning_rate": 4.390670589196622e-08,
+      "loss": 0.8172,
+      "step": 3004
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.21404823076072146,
+      "learning_rate": 4.235304907192483e-08,
+      "loss": 0.5792,
+      "step": 3005
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.18023763523717168,
+      "learning_rate": 4.082735472300536e-08,
+      "loss": 0.5673,
+      "step": 3006
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.18559562811418479,
+      "learning_rate": 3.9329624554584884e-08,
+      "loss": 0.5463,
+      "step": 3007
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.2171120048979534,
+      "learning_rate": 3.785986024472388e-08,
+      "loss": 0.812,
+      "step": 3008
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.18219977873845303,
+      "learning_rate": 3.641806344013288e-08,
+      "loss": 0.5546,
+      "step": 3009
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.17895273235008802,
+      "learning_rate": 3.500423575620304e-08,
+      "loss": 0.5449,
+      "step": 3010
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.2253687054909109,
+      "learning_rate": 3.361837877698115e-08,
+      "loss": 0.8064,
+      "step": 3011
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.2195462843042891,
+      "learning_rate": 3.2260494055175175e-08,
+      "loss": 0.5596,
+      "step": 3012
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.1779024355379711,
+      "learning_rate": 3.093058311215702e-08,
+      "loss": 0.5497,
+      "step": 3013
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.2210531335113059,
+      "learning_rate": 2.9628647437954236e-08,
+      "loss": 0.8031,
+      "step": 3014
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.18535198171936534,
+      "learning_rate": 2.8354688491250004e-08,
+      "loss": 0.5554,
+      "step": 3015
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.21820138087603475,
+      "learning_rate": 2.710870769938867e-08,
+      "loss": 0.6154,
+      "step": 3016
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.17899239473459988,
+      "learning_rate": 2.589070645835634e-08,
+      "loss": 0.5585,
+      "step": 3017
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.18725270682353604,
+      "learning_rate": 2.4700686132803076e-08,
+      "loss": 0.5439,
+      "step": 3018
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.22340025818441706,
+      "learning_rate": 2.353864805602346e-08,
+      "loss": 0.8111,
+      "step": 3019
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.181855104109582,
+      "learning_rate": 2.240459352995661e-08,
+      "loss": 0.5556,
+      "step": 3020
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.18729224310491296,
+      "learning_rate": 2.1298523825200036e-08,
+      "loss": 0.5759,
+      "step": 3021
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.2272613190094978,
+      "learning_rate": 2.022044018099023e-08,
+      "loss": 0.8199,
+      "step": 3022
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.18122608615002866,
+      "learning_rate": 1.9170343805205437e-08,
+      "loss": 0.5453,
+      "step": 3023
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.18412508217628834,
+      "learning_rate": 1.81482358743712e-08,
+      "loss": 0.5483,
+      "step": 3024
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.22432894954297303,
+      "learning_rate": 1.715411753365481e-08,
+      "loss": 0.8037,
+      "step": 3025
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.18236255791401626,
+      "learning_rate": 1.6187989896862544e-08,
+      "loss": 0.5659,
+      "step": 3026
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.22272841765803572,
+      "learning_rate": 1.5249854046442414e-08,
+      "loss": 0.6007,
+      "step": 3027
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.18005912708984861,
+      "learning_rate": 1.4339711033478642e-08,
+      "loss": 0.5446,
+      "step": 3028
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.1803930789372506,
+      "learning_rate": 1.3457561877691648e-08,
+      "loss": 0.5639,
+      "step": 3029
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.21897426773846082,
+      "learning_rate": 1.2603407567440827e-08,
+      "loss": 0.7806,
+      "step": 3030
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.1806773448294582,
+      "learning_rate": 1.1777249059716222e-08,
+      "loss": 0.5688,
+      "step": 3031
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.18010283778647984,
+      "learning_rate": 1.0979087280141298e-08,
+      "loss": 0.5451,
+      "step": 3032
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.2176594946494113,
+      "learning_rate": 1.0208923122978498e-08,
+      "loss": 0.8055,
+      "step": 3033
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.19369287801231944,
+      "learning_rate": 9.466757451112584e-09,
+      "loss": 0.542,
+      "step": 3034
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.17629826743359345,
+      "learning_rate": 8.752591096070073e-09,
+      "loss": 0.549,
+      "step": 3035
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.22407893843526247,
+      "learning_rate": 8.066424857994249e-09,
+      "loss": 0.8223,
+      "step": 3036
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.18768267329368238,
+      "learning_rate": 7.40825950566737e-09,
+      "loss": 0.5728,
+      "step": 3037
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.21441028689403113,
+      "learning_rate": 6.7780957764940204e-09,
+      "loss": 0.5816,
+      "step": 3038
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.1869463075859334,
+      "learning_rate": 6.175934376509429e-09,
+      "loss": 0.5515,
+      "step": 3039
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.22469243941033126,
+      "learning_rate": 5.601775980371149e-09,
+      "loss": 0.7901,
+      "step": 3040
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.21304994873276983,
+      "learning_rate": 5.05562123136738e-09,
+      "loss": 0.5533,
+      "step": 3041
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.18201810343970956,
+      "learning_rate": 4.53747074140587e-09,
+      "loss": 0.5578,
+      "step": 3042
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.18408385517449014,
+      "learning_rate": 4.047325091022236e-09,
+      "loss": 0.5603,
+      "step": 3043
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.21103240011478852,
+      "learning_rate": 3.5851848293744215e-09,
+      "loss": 0.5694,
+      "step": 3044
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.2116850448249744,
+      "learning_rate": 3.151050474242689e-09,
+      "loss": 0.779,
+      "step": 3045
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.16480706560416483,
+      "learning_rate": 2.7449225120268484e-09,
+      "loss": 0.5376,
+      "step": 3046
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.21848314783904926,
+      "learning_rate": 2.366801397754581e-09,
+      "loss": 0.7959,
+      "step": 3047
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.1897769200650689,
+      "learning_rate": 2.0166875550731157e-09,
+      "loss": 0.574,
+      "step": 3048
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.21204817330138198,
+      "learning_rate": 1.6945813762436757e-09,
+      "loss": 0.5828,
+      "step": 3049
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.18178245555275138,
+      "learning_rate": 1.4004832221581332e-09,
+      "loss": 0.5639,
+      "step": 3050
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.21518131890975364,
+      "learning_rate": 1.1343934223195796e-09,
+      "loss": 0.7901,
+      "step": 3051
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.19258546745093913,
+      "learning_rate": 8.963122748589791e-10,
+      "loss": 0.5573,
+      "step": 3052
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.1811540367740528,
+      "learning_rate": 6.862400465157403e-10,
+      "loss": 0.5468,
+      "step": 3053
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.23945740266900611,
+      "learning_rate": 5.041769726571444e-10,
+      "loss": 0.8097,
+      "step": 3054
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.17677975886995786,
+      "learning_rate": 3.5012325726446747e-10,
+      "loss": 0.3701,
+      "step": 3055
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.21742577040129474,
+      "learning_rate": 2.2407907294408336e-10,
+      "loss": 0.7878,
+      "step": 3056
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.1909886791089104,
+      "learning_rate": 1.2604456090803406e-10,
+      "loss": 0.5539,
+      "step": 3057
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.18008866249305067,
+      "learning_rate": 5.6019830999010094e-11,
+      "loss": 0.5652,
+      "step": 3058
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.21602136512655468,
+      "learning_rate": 1.4004961673697026e-11,
+      "loss": 0.7949,
+      "step": 3059
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.1863156555436779,
+      "learning_rate": 0.0,
+      "loss": 0.7074,
+      "step": 3060
+    },
+    {
+      "epoch": 1.0,
+      "step": 3060,
+      "total_flos": 0.0,
+      "train_loss": 0.08551377936515933,
+      "train_runtime": 10116.1848,
+      "train_samples_per_second": 309.938,
+      "train_steps_per_second": 0.302
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 3060,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 100,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": false,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 0.0,
+  "train_batch_size": 4,
+  "trial_name": null,
+  "trial_params": null
+}