diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,38621 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.11582422095049567,
+  "eval_steps": 500,
+  "global_step": 5500,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 2.1058949263726486e-05,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.0002,
+      "loss": 1.7518,
+      "step": 1
+    },
+    {
+      "epoch": 4.211789852745297e-05,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0002,
+      "loss": 1.732,
+      "step": 2
+    },
+    {
+      "epoch": 6.317684779117946e-05,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0002,
+      "loss": 1.7501,
+      "step": 3
+    },
+    {
+      "epoch": 8.423579705490595e-05,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0002,
+      "loss": 1.7667,
+      "step": 4
+    },
+    {
+      "epoch": 0.00010529474631863244,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.0002,
+      "loss": 1.7663,
+      "step": 5
+    },
+    {
+      "epoch": 0.0001263536955823589,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0002,
+      "loss": 1.7459,
+      "step": 6
+    },
+    {
+      "epoch": 0.00014741264484608541,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.0002,
+      "loss": 1.7606,
+      "step": 7
+    },
+    {
+      "epoch": 0.0001684715941098119,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.0002,
+      "loss": 1.7556,
+      "step": 8
+    },
+    {
+      "epoch": 0.00018953054337353837,
+      "grad_norm": 0.375,
+      "learning_rate": 0.0002,
+      "loss": 1.7616,
+      "step": 9
+    },
+    {
+      "epoch": 0.00021058949263726487,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.0002,
+      "loss": 1.7136,
+      "step": 10
+    },
+    {
+      "epoch": 0.00023164844190099135,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.0002,
+      "loss": 1.741,
+      "step": 11
+    },
+    {
+      "epoch": 0.0002527073911647178,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.0002,
+      "loss": 1.7446,
+      "step": 12
+    },
+    {
+      "epoch": 0.0002737663404284443,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.0002,
+      "loss": 1.7817,
+      "step": 13
+    },
+    {
+      "epoch": 0.00029482528969217083,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.0002,
+      "loss": 1.752,
+      "step": 14
+    },
+    {
+      "epoch": 0.0003158842389558973,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.0002,
+      "loss": 1.7788,
+      "step": 15
+    },
+    {
+      "epoch": 0.0003369431882196238,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.0002,
+      "loss": 1.7478,
+      "step": 16
+    },
+    {
+      "epoch": 0.0003580021374833503,
+      "grad_norm": 0.375,
+      "learning_rate": 0.0002,
+      "loss": 1.744,
+      "step": 17
+    },
+    {
+      "epoch": 0.00037906108674707673,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.0002,
+      "loss": 1.7614,
+      "step": 18
+    },
+    {
+      "epoch": 0.00040012003601080324,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.0002,
+      "loss": 1.7266,
+      "step": 19
+    },
+    {
+      "epoch": 0.00042117898527452974,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.0002,
+      "loss": 1.756,
+      "step": 20
+    },
+    {
+      "epoch": 0.0004422379345382562,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.0002,
+      "loss": 1.7207,
+      "step": 21
+    },
+    {
+      "epoch": 0.0004632968838019827,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.0002,
+      "loss": 1.7491,
+      "step": 22
+    },
+    {
+      "epoch": 0.0004843558330657092,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.0002,
+      "loss": 1.7283,
+      "step": 23
+    },
+    {
+      "epoch": 0.0005054147823294356,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.0002,
+      "loss": 1.7468,
+      "step": 24
+    },
+    {
+      "epoch": 0.0005264737315931622,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.0002,
+      "loss": 1.7473,
+      "step": 25
+    },
+    {
+      "epoch": 0.0005475326808568887,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.0002,
+      "loss": 1.7557,
+      "step": 26
+    },
+    {
+      "epoch": 0.0005685916301206151,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.0002,
+      "loss": 1.726,
+      "step": 27
+    },
+    {
+      "epoch": 0.0005896505793843417,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.0002,
+      "loss": 1.7517,
+      "step": 28
+    },
+    {
+      "epoch": 0.0006107095286480681,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0002,
+      "loss": 1.7738,
+      "step": 29
+    },
+    {
+      "epoch": 0.0006317684779117946,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0002,
+      "loss": 1.7384,
+      "step": 30
+    },
+    {
+      "epoch": 0.0006528274271755211,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0002,
+      "loss": 1.7413,
+      "step": 31
+    },
+    {
+      "epoch": 0.0006738863764392476,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.0002,
+      "loss": 1.7713,
+      "step": 32
+    },
+    {
+      "epoch": 0.000694945325702974,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0002,
+      "loss": 1.7332,
+      "step": 33
+    },
+    {
+      "epoch": 0.0007160042749667006,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0002,
+      "loss": 1.7555,
+      "step": 34
+    },
+    {
+      "epoch": 0.000737063224230427,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.0002,
+      "loss": 1.7619,
+      "step": 35
+    },
+    {
+      "epoch": 0.0007581221734941535,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0002,
+      "loss": 1.7695,
+      "step": 36
+    },
+    {
+      "epoch": 0.00077918112275788,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0002,
+      "loss": 1.7644,
+      "step": 37
+    },
+    {
+      "epoch": 0.0008002400720216065,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0002,
+      "loss": 1.7493,
+      "step": 38
+    },
+    {
+      "epoch": 0.0008212990212853329,
+      "grad_norm": 0.83203125,
+      "learning_rate": 0.0002,
+      "loss": 1.7452,
+      "step": 39
+    },
+    {
+      "epoch": 0.0008423579705490595,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0002,
+      "loss": 1.6903,
+      "step": 40
+    },
+    {
+      "epoch": 0.0008634169198127859,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0002,
+      "loss": 1.754,
+      "step": 41
+    },
+    {
+      "epoch": 0.0008844758690765124,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0002,
+      "loss": 1.7764,
+      "step": 42
+    },
+    {
+      "epoch": 0.0009055348183402389,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0002,
+      "loss": 1.7231,
+      "step": 43
+    },
+    {
+      "epoch": 0.0009265937676039654,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0002,
+      "loss": 1.7465,
+      "step": 44
+    },
+    {
+      "epoch": 0.0009476527168676918,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0002,
+      "loss": 1.7356,
+      "step": 45
+    },
+    {
+      "epoch": 0.0009687116661314184,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0002,
+      "loss": 1.7407,
+      "step": 46
+    },
+    {
+      "epoch": 0.0009897706153951448,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.0002,
+      "loss": 1.7647,
+      "step": 47
+    },
+    {
+      "epoch": 0.0010108295646588713,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0002,
+      "loss": 1.7212,
+      "step": 48
+    },
+    {
+      "epoch": 0.0010318885139225977,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0002,
+      "loss": 1.76,
+      "step": 49
+    },
+    {
+      "epoch": 0.0010529474631863244,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.0002,
+      "loss": 1.7608,
+      "step": 50
+    },
+    {
+      "epoch": 0.0010740064124500509,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.0002,
+      "loss": 1.7617,
+      "step": 51
+    },
+    {
+      "epoch": 0.0010950653617137773,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.0002,
+      "loss": 1.772,
+      "step": 52
+    },
+    {
+      "epoch": 0.0011161243109775038,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.0002,
+      "loss": 1.7483,
+      "step": 53
+    },
+    {
+      "epoch": 0.0011371832602412302,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.0002,
+      "loss": 1.7029,
+      "step": 54
+    },
+    {
+      "epoch": 0.0011582422095049567,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.0002,
+      "loss": 1.7188,
+      "step": 55
+    },
+    {
+      "epoch": 0.0011793011587686833,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.0002,
+      "loss": 1.7435,
+      "step": 56
+    },
+    {
+      "epoch": 0.0012003601080324098,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0002,
+      "loss": 1.7238,
+      "step": 57
+    },
+    {
+      "epoch": 0.0012214190572961362,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.0002,
+      "loss": 1.7492,
+      "step": 58
+    },
+    {
+      "epoch": 0.0012424780065598627,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.0002,
+      "loss": 1.7301,
+      "step": 59
+    },
+    {
+      "epoch": 0.0012635369558235891,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.0002,
+      "loss": 1.7522,
+      "step": 60
+    },
+    {
+      "epoch": 0.0012845959050873156,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.0002,
+      "loss": 1.7609,
+      "step": 61
+    },
+    {
+      "epoch": 0.0013056548543510422,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.0002,
+      "loss": 1.7026,
+      "step": 62
+    },
+    {
+      "epoch": 0.0013267138036147687,
+      "grad_norm": 0.2109375,
+      "learning_rate": 0.0002,
+      "loss": 1.697,
+      "step": 63
+    },
+    {
+      "epoch": 0.0013477727528784951,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.0002,
+      "loss": 1.7366,
+      "step": 64
+    },
+    {
+      "epoch": 0.0013688317021422216,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.0002,
+      "loss": 1.6928,
+      "step": 65
+    },
+    {
+      "epoch": 0.001389890651405948,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.0002,
+      "loss": 1.7447,
+      "step": 66
+    },
+    {
+      "epoch": 0.0014109496006696747,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.0002,
+      "loss": 1.7125,
+      "step": 67
+    },
+    {
+      "epoch": 0.0014320085499334011,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.0002,
+      "loss": 1.7203,
+      "step": 68
+    },
+    {
+      "epoch": 0.0014530674991971276,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.0002,
+      "loss": 1.7666,
+      "step": 69
+    },
+    {
+      "epoch": 0.001474126448460854,
+      "grad_norm": 0.208984375,
+      "learning_rate": 0.0002,
+      "loss": 1.7205,
+      "step": 70
+    },
+    {
+      "epoch": 0.0014951853977245805,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.0002,
+      "loss": 1.726,
+      "step": 71
+    },
+    {
+      "epoch": 0.001516244346988307,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.0002,
+      "loss": 1.7006,
+      "step": 72
+    },
+    {
+      "epoch": 0.0015373032962520336,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.0002,
+      "loss": 1.7336,
+      "step": 73
+    },
+    {
+      "epoch": 0.00155836224551576,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.0002,
+      "loss": 1.7516,
+      "step": 74
+    },
+    {
+      "epoch": 0.0015794211947794865,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.0002,
+      "loss": 1.7275,
+      "step": 75
+    },
+    {
+      "epoch": 0.001600480144043213,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.0002,
+      "loss": 1.7405,
+      "step": 76
+    },
+    {
+      "epoch": 0.0016215390933069394,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.0002,
+      "loss": 1.727,
+      "step": 77
+    },
+    {
+      "epoch": 0.0016425980425706658,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.0002,
+      "loss": 1.7659,
+      "step": 78
+    },
+    {
+      "epoch": 0.0016636569918343925,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.0002,
+      "loss": 1.7078,
+      "step": 79
+    },
+    {
+      "epoch": 0.001684715941098119,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.0002,
+      "loss": 1.7204,
+      "step": 80
+    },
+    {
+      "epoch": 0.0017057748903618454,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.0002,
+      "loss": 1.7338,
+      "step": 81
+    },
+    {
+      "epoch": 0.0017268338396255719,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.0002,
+      "loss": 1.7066,
+      "step": 82
+    },
+    {
+      "epoch": 0.0017478927888892983,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.0002,
+      "loss": 1.722,
+      "step": 83
+    },
+    {
+      "epoch": 0.0017689517381530248,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.0002,
+      "loss": 1.7529,
+      "step": 84
+    },
+    {
+      "epoch": 0.0017900106874167514,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.0002,
+      "loss": 1.7395,
+      "step": 85
+    },
+    {
+      "epoch": 0.0018110696366804779,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.0002,
+      "loss": 1.7222,
+      "step": 86
+    },
+    {
+      "epoch": 0.0018321285859442043,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.0002,
+      "loss": 1.7122,
+      "step": 87
+    },
+    {
+      "epoch": 0.0018531875352079308,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.0002,
+      "loss": 1.7129,
+      "step": 88
+    },
+    {
+      "epoch": 0.0018742464844716572,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.0002,
+      "loss": 1.7576,
+      "step": 89
+    },
+    {
+      "epoch": 0.0018953054337353837,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.0002,
+      "loss": 1.7223,
+      "step": 90
+    },
+    {
+      "epoch": 0.0019163643829991103,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0002,
+      "loss": 1.7014,
+      "step": 91
+    },
+    {
+      "epoch": 0.0019374233322628368,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.0002,
+      "loss": 1.7119,
+      "step": 92
+    },
+    {
+      "epoch": 0.001958482281526563,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.0002,
+      "loss": 1.7084,
+      "step": 93
+    },
+    {
+      "epoch": 0.0019795412307902897,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0002,
+      "loss": 1.709,
+      "step": 94
+    },
+    {
+      "epoch": 0.0020006001800540163,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.0002,
+      "loss": 1.7377,
+      "step": 95
+    },
+    {
+      "epoch": 0.0020216591293177426,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.0002,
+      "loss": 1.7108,
+      "step": 96
+    },
+    {
+      "epoch": 0.0020427180785814692,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0002,
+      "loss": 1.7172,
+      "step": 97
+    },
+    {
+      "epoch": 0.0020637770278451955,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.0002,
+      "loss": 1.694,
+      "step": 98
+    },
+    {
+      "epoch": 0.002084835977108922,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.0002,
+      "loss": 1.7164,
+      "step": 99
+    },
+    {
+      "epoch": 0.002105894926372649,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.0002,
+      "loss": 1.705,
+      "step": 100
+    },
+    {
+      "epoch": 0.002126953875636375,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.0002,
+      "loss": 1.7086,
+      "step": 101
+    },
+    {
+      "epoch": 0.0021480128249001017,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.0002,
+      "loss": 1.7554,
+      "step": 102
+    },
+    {
+      "epoch": 0.002169071774163828,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0002,
+      "loss": 1.7301,
+      "step": 103
+    },
+    {
+      "epoch": 0.0021901307234275546,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.0002,
+      "loss": 1.6998,
+      "step": 104
+    },
+    {
+      "epoch": 0.0022111896726912813,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.0002,
+      "loss": 1.7334,
+      "step": 105
+    },
+    {
+      "epoch": 0.0022322486219550075,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.0002,
+      "loss": 1.7066,
+      "step": 106
+    },
+    {
+      "epoch": 0.002253307571218734,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.0002,
+      "loss": 1.7335,
+      "step": 107
+    },
+    {
+      "epoch": 0.0022743665204824604,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.0002,
+      "loss": 1.7545,
+      "step": 108
+    },
+    {
+      "epoch": 0.002295425469746187,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.0002,
+      "loss": 1.7595,
+      "step": 109
+    },
+    {
+      "epoch": 0.0023164844190099133,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.0002,
+      "loss": 1.7129,
+      "step": 110
+    },
+    {
+      "epoch": 0.00233754336827364,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.0002,
+      "loss": 1.7182,
+      "step": 111
+    },
+    {
+      "epoch": 0.0023586023175373666,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.0002,
+      "loss": 1.705,
+      "step": 112
+    },
+    {
+      "epoch": 0.002379661266801093,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.0002,
+      "loss": 1.7472,
+      "step": 113
+    },
+    {
+      "epoch": 0.0024007202160648195,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.0002,
+      "loss": 1.7069,
+      "step": 114
+    },
+    {
+      "epoch": 0.0024217791653285458,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.0002,
+      "loss": 1.7329,
+      "step": 115
+    },
+    {
+      "epoch": 0.0024428381145922724,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0002,
+      "loss": 1.7351,
+      "step": 116
+    },
+    {
+      "epoch": 0.002463897063855999,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.0002,
+      "loss": 1.7262,
+      "step": 117
+    },
+    {
+      "epoch": 0.0024849560131197253,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.0002,
+      "loss": 1.7034,
+      "step": 118
+    },
+    {
+      "epoch": 0.002506014962383452,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0002,
+      "loss": 1.722,
+      "step": 119
+    },
+    {
+      "epoch": 0.0025270739116471782,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0002,
+      "loss": 1.7172,
+      "step": 120
+    },
+    {
+      "epoch": 0.002548132860910905,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0002,
+      "loss": 1.7305,
+      "step": 121
+    },
+    {
+      "epoch": 0.002569191810174631,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0002,
+      "loss": 1.7188,
+      "step": 122
+    },
+    {
+      "epoch": 0.002590250759438358,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0002,
+      "loss": 1.7248,
+      "step": 123
+    },
+    {
+      "epoch": 0.0026113097087020845,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.0002,
+      "loss": 1.6953,
+      "step": 124
+    },
+    {
+      "epoch": 0.0026323686579658107,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0002,
+      "loss": 1.7362,
+      "step": 125
+    },
+    {
+      "epoch": 0.0026534276072295374,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0002,
+      "loss": 1.7352,
+      "step": 126
+    },
+    {
+      "epoch": 0.0026744865564932636,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0002,
+      "loss": 1.7389,
+      "step": 127
+    },
+    {
+      "epoch": 0.0026955455057569903,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0002,
+      "loss": 1.7134,
+      "step": 128
+    },
+    {
+      "epoch": 0.002716604455020717,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0002,
+      "loss": 1.6862,
+      "step": 129
+    },
+    {
+      "epoch": 0.002737663404284443,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0002,
+      "loss": 1.7311,
+      "step": 130
+    },
+    {
+      "epoch": 0.00275872235354817,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0002,
+      "loss": 1.7399,
+      "step": 131
+    },
+    {
+      "epoch": 0.002779781302811896,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0002,
+      "loss": 1.7003,
+      "step": 132
+    },
+    {
+      "epoch": 0.0028008402520756227,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0002,
+      "loss": 1.7573,
+      "step": 133
+    },
+    {
+      "epoch": 0.0028218992013393494,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002,
+      "loss": 1.6674,
+      "step": 134
+    },
+    {
+      "epoch": 0.0028429581506030756,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0002,
+      "loss": 1.7218,
+      "step": 135
+    },
+    {
+      "epoch": 0.0028640170998668023,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0002,
+      "loss": 1.7181,
+      "step": 136
+    },
+    {
+      "epoch": 0.0028850760491305285,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0002,
+      "loss": 1.7341,
+      "step": 137
+    },
+    {
+      "epoch": 0.002906134998394255,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0002,
+      "loss": 1.7303,
+      "step": 138
+    },
+    {
+      "epoch": 0.0029271939476579814,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0002,
+      "loss": 1.6965,
+      "step": 139
+    },
+    {
+      "epoch": 0.002948252896921708,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0002,
+      "loss": 1.7085,
+      "step": 140
+    },
+    {
+      "epoch": 0.0029693118461854347,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002,
+      "loss": 1.7004,
+      "step": 141
+    },
+    {
+      "epoch": 0.002990370795449161,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0002,
+      "loss": 1.7088,
+      "step": 142
+    },
+    {
+      "epoch": 0.0030114297447128876,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.0002,
+      "loss": 1.7003,
+      "step": 143
+    },
+    {
+      "epoch": 0.003032488693976614,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002,
+      "loss": 1.7173,
+      "step": 144
+    },
+    {
+      "epoch": 0.0030535476432403405,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0002,
+      "loss": 1.7199,
+      "step": 145
+    },
+    {
+      "epoch": 0.003074606592504067,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0002,
+      "loss": 1.7208,
+      "step": 146
+    },
+    {
+      "epoch": 0.0030956655417677934,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0002,
+      "loss": 1.7088,
+      "step": 147
+    },
+    {
+      "epoch": 0.00311672449103152,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0002,
+      "loss": 1.7232,
+      "step": 148
+    },
+    {
+      "epoch": 0.0031377834402952463,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.0002,
+      "loss": 1.698,
+      "step": 149
+    },
+    {
+      "epoch": 0.003158842389558973,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0002,
+      "loss": 1.7043,
+      "step": 150
+    },
+    {
+      "epoch": 0.0031799013388226992,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0002,
+      "loss": 1.7165,
+      "step": 151
+    },
+    {
+      "epoch": 0.003200960288086426,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0002,
+      "loss": 1.7049,
+      "step": 152
+    },
+    {
+      "epoch": 0.0032220192373501526,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0002,
+      "loss": 1.7164,
+      "step": 153
+    },
+    {
+      "epoch": 0.003243078186613879,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0002,
+      "loss": 1.7111,
+      "step": 154
+    },
+    {
+      "epoch": 0.0032641371358776055,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0002,
+      "loss": 1.7018,
+      "step": 155
+    },
+    {
+      "epoch": 0.0032851960851413317,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0002,
+      "loss": 1.714,
+      "step": 156
+    },
+    {
+      "epoch": 0.0033062550344050584,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0002,
+      "loss": 1.7111,
+      "step": 157
+    },
+    {
+      "epoch": 0.003327313983668785,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002,
+      "loss": 1.6922,
+      "step": 158
+    },
+    {
+      "epoch": 0.0033483729329325113,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0002,
+      "loss": 1.7051,
+      "step": 159
+    },
+    {
+      "epoch": 0.003369431882196238,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0002,
+      "loss": 1.6804,
+      "step": 160
+    },
+    {
+      "epoch": 0.003390490831459964,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0002,
+      "loss": 1.699,
+      "step": 161
+    },
+    {
+      "epoch": 0.003411549780723691,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.7237,
+      "step": 162
+    },
+    {
+      "epoch": 0.0034326087299874175,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0002,
+      "loss": 1.6972,
+      "step": 163
+    },
+    {
+      "epoch": 0.0034536676792511437,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0002,
+      "loss": 1.7049,
+      "step": 164
+    },
+    {
+      "epoch": 0.0034747266285148704,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002,
+      "loss": 1.731,
+      "step": 165
+    },
+    {
+      "epoch": 0.0034957855777785966,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0002,
+      "loss": 1.6945,
+      "step": 166
+    },
+    {
+      "epoch": 0.0035168445270423233,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0002,
+      "loss": 1.7267,
+      "step": 167
+    },
+    {
+      "epoch": 0.0035379034763060495,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0002,
+      "loss": 1.7396,
+      "step": 168
+    },
+    {
+      "epoch": 0.003558962425569776,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.7107,
+      "step": 169
+    },
+    {
+      "epoch": 0.003580021374833503,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.0002,
+      "loss": 1.7223,
+      "step": 170
+    },
+    {
+      "epoch": 0.003601080324097229,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.7131,
+      "step": 171
+    },
+    {
+      "epoch": 0.0036221392733609557,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0002,
+      "loss": 1.6982,
+      "step": 172
+    },
+    {
+      "epoch": 0.003643198222624682,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.7305,
+      "step": 173
+    },
+    {
+      "epoch": 0.0036642571718884086,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0002,
+      "loss": 1.7435,
+      "step": 174
+    },
+    {
+      "epoch": 0.0036853161211521353,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.6927,
+      "step": 175
+    },
+    {
+      "epoch": 0.0037063750704158615,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0002,
+      "loss": 1.6945,
+      "step": 176
+    },
+    {
+      "epoch": 0.003727434019679588,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0002,
+      "loss": 1.7027,
+      "step": 177
+    },
+    {
+      "epoch": 0.0037484929689433144,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.7237,
+      "step": 178
+    },
+    {
+      "epoch": 0.003769551918207041,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0002,
+      "loss": 1.7273,
+      "step": 179
+    },
+    {
+      "epoch": 0.0037906108674707673,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0002,
+      "loss": 1.6934,
+      "step": 180
+    },
+    {
+      "epoch": 0.003811669816734494,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0002,
+      "loss": 1.6898,
+      "step": 181
+    },
+    {
+      "epoch": 0.0038327287659982207,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0002,
+      "loss": 1.6878,
+      "step": 182
+    },
+    {
+      "epoch": 0.003853787715261947,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0002,
+      "loss": 1.6875,
+      "step": 183
+    },
+    {
+      "epoch": 0.0038748466645256736,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.6636,
+      "step": 184
+    },
+    {
+      "epoch": 0.0038959056137894,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002,
+      "loss": 1.7399,
+      "step": 185
+    },
+    {
+      "epoch": 0.003916964563053126,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0002,
+      "loss": 1.7184,
+      "step": 186
+    },
+    {
+      "epoch": 0.003938023512316853,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002,
+      "loss": 1.6886,
+      "step": 187
+    },
+    {
+      "epoch": 0.003959082461580579,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.0002,
+      "loss": 1.7033,
+      "step": 188
+    },
+    {
+      "epoch": 0.003980141410844306,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002,
+      "loss": 1.7147,
+      "step": 189
+    },
+    {
+      "epoch": 0.004001200360108033,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.6534,
+      "step": 190
+    },
+    {
+      "epoch": 0.0040222593093717585,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.7233,
+      "step": 191
+    },
+    {
+      "epoch": 0.004043318258635485,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.7084,
+      "step": 192
+    },
+    {
+      "epoch": 0.004064377207899212,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.7158,
+      "step": 193
+    },
+    {
+      "epoch": 0.0040854361571629385,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.7134,
+      "step": 194
+    },
+    {
+      "epoch": 0.004106495106426665,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002,
+      "loss": 1.7054,
+      "step": 195
+    },
+    {
+      "epoch": 0.004127554055690391,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0002,
+      "loss": 1.6769,
+      "step": 196
+    },
+    {
+      "epoch": 0.004148613004954118,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0002,
+      "loss": 1.7065,
+      "step": 197
+    },
+    {
+      "epoch": 0.004169671954217844,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0002,
+      "loss": 1.7115,
+      "step": 198
+    },
+    {
+      "epoch": 0.004190730903481571,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.6873,
+      "step": 199
+    },
+    {
+      "epoch": 0.004211789852745298,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.7159,
+      "step": 200
+    },
+    {
+      "epoch": 0.004232848802009023,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002,
+      "loss": 1.714,
+      "step": 201
+    },
+    {
+      "epoch": 0.00425390775127275,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0002,
+      "loss": 1.7203,
+      "step": 202
+    },
+    {
+      "epoch": 0.004274966700536477,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0002,
+      "loss": 1.6976,
+      "step": 203
+    },
+    {
+      "epoch": 0.004296025649800203,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.6981,
+      "step": 204
+    },
+    {
+      "epoch": 0.00431708459906393,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.7151,
+      "step": 205
+    },
+    {
+      "epoch": 0.004338143548327656,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0002,
+      "loss": 1.7174,
+      "step": 206
+    },
+    {
+      "epoch": 0.0043592024975913825,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.7091,
+      "step": 207
+    },
+    {
+      "epoch": 0.004380261446855109,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.6981,
+      "step": 208
+    },
+    {
+      "epoch": 0.004401320396118836,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0002,
+      "loss": 1.6876,
+      "step": 209
+    },
+    {
+      "epoch": 0.0044223793453825625,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0002,
+      "loss": 1.7105,
+      "step": 210
+    },
+    {
+      "epoch": 0.004443438294646288,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.7091,
+      "step": 211
+    },
+    {
+      "epoch": 0.004464497243910015,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0002,
+      "loss": 1.7066,
+      "step": 212
+    },
+    {
+      "epoch": 0.004485556193173742,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.7054,
+      "step": 213
+    },
+    {
+      "epoch": 0.004506615142437468,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.761,
+      "step": 214
+    },
+    {
+      "epoch": 0.004527674091701194,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.7352,
+      "step": 215
+    },
+    {
+      "epoch": 0.004548733040964921,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.6842,
+      "step": 216
+    },
+    {
+      "epoch": 0.0045697919902286475,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6548,
+      "step": 217
+    },
+    {
+      "epoch": 0.004590850939492374,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0002,
+      "loss": 1.7116,
+      "step": 218
+    },
+    {
+      "epoch": 0.004611909888756101,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.6706,
+      "step": 219
+    },
+    {
+      "epoch": 0.004632968838019827,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0002,
+      "loss": 1.697,
+      "step": 220
+    },
+    {
+      "epoch": 0.004654027787283553,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0002,
+      "loss": 1.7024,
+      "step": 221
+    },
+    {
+      "epoch": 0.00467508673654728,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.7163,
+      "step": 222
+    },
+    {
+      "epoch": 0.004696145685811007,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.0002,
+      "loss": 1.7081,
+      "step": 223
+    },
+    {
+      "epoch": 0.004717204635074733,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6975,
+      "step": 224
+    },
+    {
+      "epoch": 0.004738263584338459,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.6649,
+      "step": 225
+    },
+    {
+      "epoch": 0.004759322533602186,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0002,
+      "loss": 1.7132,
+      "step": 226
+    },
+    {
+      "epoch": 0.004780381482865912,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.7132,
+      "step": 227
+    },
+    {
+      "epoch": 0.004801440432129639,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.7111,
+      "step": 228
+    },
+    {
+      "epoch": 0.004822499381393366,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.6466,
+      "step": 229
+    },
+    {
+      "epoch": 0.0048435583306570915,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0002,
+      "loss": 1.6868,
+      "step": 230
+    },
+    {
+      "epoch": 0.004864617279920818,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0002,
+      "loss": 1.6956,
+      "step": 231
+    },
+    {
+      "epoch": 0.004885676229184545,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.6505,
+      "step": 232
+    },
+    {
+      "epoch": 0.0049067351784482715,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.667,
+      "step": 233
+    },
+    {
+      "epoch": 0.004927794127711998,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.6902,
+      "step": 234
+    },
+    {
+      "epoch": 0.004948853076975724,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.7194,
+      "step": 235
+    },
+    {
+      "epoch": 0.004969912026239451,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.6822,
+      "step": 236
+    },
+    {
+      "epoch": 0.004990970975503177,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0002,
+      "loss": 1.7176,
+      "step": 237
+    },
+    {
+      "epoch": 0.005012029924766904,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.6858,
+      "step": 238
+    },
+    {
+      "epoch": 0.005033088874030631,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.6671,
+      "step": 239
+    },
+    {
+      "epoch": 0.0050541478232943565,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0002,
+      "loss": 1.6725,
+      "step": 240
+    },
+    {
+      "epoch": 0.005075206772558083,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.703,
+      "step": 241
+    },
+    {
+      "epoch": 0.00509626572182181,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.6822,
+      "step": 242
+    },
+    {
+      "epoch": 0.0051173246710855365,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.693,
+      "step": 243
+    },
+    {
+      "epoch": 0.005138383620349262,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0002,
+      "loss": 1.7021,
+      "step": 244
+    },
+    {
+      "epoch": 0.005159442569612989,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.6968,
+      "step": 245
+    },
+    {
+      "epoch": 0.005180501518876716,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0002,
+      "loss": 1.6963,
+      "step": 246
+    },
+    {
+      "epoch": 0.005201560468140442,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.7204,
+      "step": 247
+    },
+    {
+      "epoch": 0.005222619417404169,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.6709,
+      "step": 248
+    },
+    {
+      "epoch": 0.005243678366667895,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6899,
+      "step": 249
+    },
+    {
+      "epoch": 0.005264737315931621,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0002,
+      "loss": 1.6948,
+      "step": 250
+    },
+    {
+      "epoch": 0.005285796265195348,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0002,
+      "loss": 1.6926,
+      "step": 251
+    },
+    {
+      "epoch": 0.005306855214459075,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.6856,
+      "step": 252
+    },
+    {
+      "epoch": 0.005327914163722801,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.7411,
+      "step": 253
+    },
+    {
+      "epoch": 0.005348973112986527,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.6778,
+      "step": 254
+    },
+    {
+      "epoch": 0.005370032062250254,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.6848,
+      "step": 255
+    },
+    {
+      "epoch": 0.0053910910115139805,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.0002,
+      "loss": 1.6836,
+      "step": 256
+    },
+    {
+      "epoch": 0.005412149960777707,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.6927,
+      "step": 257
+    },
+    {
+      "epoch": 0.005433208910041434,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0002,
+      "loss": 1.7086,
+      "step": 258
+    },
+    {
+      "epoch": 0.00545426785930516,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.684,
+      "step": 259
+    },
+    {
+      "epoch": 0.005475326808568886,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0002,
+      "loss": 1.6969,
+      "step": 260
+    },
+    {
+      "epoch": 0.005496385757832613,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.7123,
+      "step": 261
+    },
+    {
+      "epoch": 0.00551744470709634,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0002,
+      "loss": 1.7236,
+      "step": 262
+    },
+    {
+      "epoch": 0.005538503656360066,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.6874,
+      "step": 263
+    },
+    {
+      "epoch": 0.005559562605623792,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.6924,
+      "step": 264
+    },
+    {
+      "epoch": 0.005580621554887519,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0002,
+      "loss": 1.6545,
+      "step": 265
+    },
+    {
+      "epoch": 0.005601680504151245,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.7146,
+      "step": 266
+    },
+    {
+      "epoch": 0.005622739453414972,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.7107,
+      "step": 267
+    },
+    {
+      "epoch": 0.005643798402678699,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0002,
+      "loss": 1.6644,
+      "step": 268
+    },
+    {
+      "epoch": 0.0056648573519424246,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.7131,
+      "step": 269
+    },
+    {
+      "epoch": 0.005685916301206151,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.6551,
+      "step": 270
+    },
+    {
+      "epoch": 0.005706975250469878,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6973,
+      "step": 271
+    },
+    {
+      "epoch": 0.0057280341997336046,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.7111,
+      "step": 272
+    },
+    {
+      "epoch": 0.00574909314899733,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.6686,
+      "step": 273
+    },
+    {
+      "epoch": 0.005770152098261057,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.6846,
+      "step": 274
+    },
+    {
+      "epoch": 0.005791211047524784,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.7044,
+      "step": 275
+    },
+    {
+      "epoch": 0.00581226999678851,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.6833,
+      "step": 276
+    },
+    {
+      "epoch": 0.005833328946052237,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.7105,
+      "step": 277
+    },
+    {
+      "epoch": 0.005854387895315963,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0002,
+      "loss": 1.6638,
+      "step": 278
+    },
+    {
+      "epoch": 0.0058754468445796895,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6967,
+      "step": 279
+    },
+    {
+      "epoch": 0.005896505793843416,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.6958,
+      "step": 280
+    },
+    {
+      "epoch": 0.005917564743107143,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.7417,
+      "step": 281
+    },
+    {
+      "epoch": 0.0059386236923708695,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.6871,
+      "step": 282
+    },
+    {
+      "epoch": 0.005959682641634595,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.7185,
+      "step": 283
+    },
+    {
+      "epoch": 0.005980741590898322,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0002,
+      "loss": 1.6819,
+      "step": 284
+    },
+    {
+      "epoch": 0.006001800540162049,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.6993,
+      "step": 285
+    },
+    {
+      "epoch": 0.006022859489425775,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.6664,
+      "step": 286
+    },
+    {
+      "epoch": 0.006043918438689502,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.7015,
+      "step": 287
+    },
+    {
+      "epoch": 0.006064977387953228,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0002,
+      "loss": 1.6846,
+      "step": 288
+    },
+    {
+      "epoch": 0.006086036337216954,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.7087,
+      "step": 289
+    },
+    {
+      "epoch": 0.006107095286480681,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.6947,
+      "step": 290
+    },
+    {
+      "epoch": 0.006128154235744408,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6634,
+      "step": 291
+    },
+    {
+      "epoch": 0.006149213185008134,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6947,
+      "step": 292
+    },
+    {
+      "epoch": 0.00617027213427186,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6972,
+      "step": 293
+    },
+    {
+      "epoch": 0.006191331083535587,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.693,
+      "step": 294
+    },
+    {
+      "epoch": 0.0062123900327993135,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.717,
+      "step": 295
+    },
+    {
+      "epoch": 0.00623344898206304,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6691,
+      "step": 296
+    },
+    {
+      "epoch": 0.006254507931326767,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.7448,
+      "step": 297
+    },
+    {
+      "epoch": 0.006275566880590493,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0002,
+      "loss": 1.7048,
+      "step": 298
+    },
+    {
+      "epoch": 0.006296625829854219,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.7012,
+      "step": 299
+    },
+    {
+      "epoch": 0.006317684779117946,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0002,
+      "loss": 1.7019,
+      "step": 300
+    },
+    {
+      "epoch": 0.006338743728381673,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6798,
+      "step": 301
+    },
+    {
+      "epoch": 0.0063598026776453985,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0002,
+      "loss": 1.6943,
+      "step": 302
+    },
+    {
+      "epoch": 0.006380861626909125,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.6669,
+      "step": 303
+    },
+    {
+      "epoch": 0.006401920576172852,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0002,
+      "loss": 1.6784,
+      "step": 304
+    },
+    {
+      "epoch": 0.0064229795254365785,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0002,
+      "loss": 1.6824,
+      "step": 305
+    },
+    {
+      "epoch": 0.006444038474700305,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.6717,
+      "step": 306
+    },
+    {
+      "epoch": 0.006465097423964031,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.7211,
+      "step": 307
+    },
+    {
+      "epoch": 0.006486156373227758,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.7237,
+      "step": 308
+    },
+    {
+      "epoch": 0.006507215322491484,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002,
+      "loss": 1.7176,
+      "step": 309
+    },
+    {
+      "epoch": 0.006528274271755211,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0002,
+      "loss": 1.6662,
+      "step": 310
+    },
+    {
+      "epoch": 0.006549333221018938,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.6955,
+      "step": 311
+    },
+    {
+      "epoch": 0.006570392170282663,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.6601,
+      "step": 312
+    },
+    {
+      "epoch": 0.00659145111954639,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6981,
+      "step": 313
+    },
+    {
+      "epoch": 0.006612510068810117,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002,
+      "loss": 1.7239,
+      "step": 314
+    },
+    {
+      "epoch": 0.006633569018073843,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0002,
+      "loss": 1.6826,
+      "step": 315
+    },
+    {
+      "epoch": 0.00665462796733757,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0002,
+      "loss": 1.6939,
+      "step": 316
+    },
+    {
+      "epoch": 0.006675686916601296,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.7061,
+      "step": 317
+    },
+    {
+      "epoch": 0.0066967458658650225,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0002,
+      "loss": 1.7104,
+      "step": 318
+    },
+    {
+      "epoch": 0.006717804815128749,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6612,
+      "step": 319
+    },
+    {
+      "epoch": 0.006738863764392476,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0002,
+      "loss": 1.7063,
+      "step": 320
+    },
+    {
+      "epoch": 0.0067599227136562025,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.6798,
+      "step": 321
+    },
+    {
+      "epoch": 0.006780981662919928,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.6625,
+      "step": 322
+    },
+    {
+      "epoch": 0.006802040612183655,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.7182,
+      "step": 323
+    },
+    {
+      "epoch": 0.006823099561447382,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.6498,
+      "step": 324
+    },
+    {
+      "epoch": 0.006844158510711108,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0002,
+      "loss": 1.6928,
+      "step": 325
+    },
+    {
+      "epoch": 0.006865217459974835,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.7006,
+      "step": 326
+    },
+    {
+      "epoch": 0.006886276409238561,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.6825,
+      "step": 327
+    },
+    {
+      "epoch": 0.0069073353585022874,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.686,
+      "step": 328
+    },
+    {
+      "epoch": 0.006928394307766014,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0002,
+      "loss": 1.6604,
+      "step": 329
+    },
+    {
+      "epoch": 0.006949453257029741,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0002,
+      "loss": 1.695,
+      "step": 330
+    },
+    {
+      "epoch": 0.006970512206293467,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.6883,
+      "step": 331
+    },
+    {
+      "epoch": 0.006991571155557193,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0002,
+      "loss": 1.6402,
+      "step": 332
+    },
+    {
+      "epoch": 0.00701263010482092,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.701,
+      "step": 333
+    },
+    {
+      "epoch": 0.007033689054084647,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0002,
+      "loss": 1.6617,
+      "step": 334
+    },
+    {
+      "epoch": 0.007054748003348373,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.7031,
+      "step": 335
+    },
+    {
+      "epoch": 0.007075806952612099,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.6707,
+      "step": 336
+    },
+    {
+      "epoch": 0.007096865901875826,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0002,
+      "loss": 1.6682,
+      "step": 337
+    },
+    {
+      "epoch": 0.007117924851139552,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0002,
+      "loss": 1.6762,
+      "step": 338
+    },
+    {
+      "epoch": 0.007138983800403279,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.7078,
+      "step": 339
+    },
+    {
+      "epoch": 0.007160042749667006,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.6818,
+      "step": 340
+    },
+    {
+      "epoch": 0.0071811016989307315,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0002,
+      "loss": 1.635,
+      "step": 341
+    },
+    {
+      "epoch": 0.007202160648194458,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.0002,
+      "loss": 1.6842,
+      "step": 342
+    },
+    {
+      "epoch": 0.007223219597458185,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0002,
+      "loss": 1.6659,
+      "step": 343
+    },
+    {
+      "epoch": 0.0072442785467219115,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.7083,
+      "step": 344
+    },
+    {
+      "epoch": 0.007265337495985638,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0002,
+      "loss": 1.6705,
+      "step": 345
+    },
+    {
+      "epoch": 0.007286396445249364,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.649,
+      "step": 346
+    },
+    {
+      "epoch": 0.007307455394513091,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6753,
+      "step": 347
+    },
+    {
+      "epoch": 0.007328514343776817,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.0002,
+      "loss": 1.6722,
+      "step": 348
+    },
+    {
+      "epoch": 0.007349573293040544,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.7328,
+      "step": 349
+    },
+    {
+      "epoch": 0.007370632242304271,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0002,
+      "loss": 1.6997,
+      "step": 350
+    },
+    {
+      "epoch": 0.007391691191567996,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6963,
+      "step": 351
+    },
+    {
+      "epoch": 0.007412750140831723,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.6563,
+      "step": 352
+    },
+    {
+      "epoch": 0.00743380909009545,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.7243,
+      "step": 353
+    },
+    {
+      "epoch": 0.007454868039359176,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.6871,
+      "step": 354
+    },
+    {
+      "epoch": 0.007475926988622903,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0002,
+      "loss": 1.6619,
+      "step": 355
+    },
+    {
+      "epoch": 0.007496985937886629,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0002,
+      "loss": 1.6938,
+      "step": 356
+    },
+    {
+      "epoch": 0.0075180448871503555,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.6559,
+      "step": 357
+    },
+    {
+      "epoch": 0.007539103836414082,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0002,
+      "loss": 1.7006,
+      "step": 358
+    },
+    {
+      "epoch": 0.007560162785677809,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0002,
+      "loss": 1.6549,
+      "step": 359
+    },
+    {
+      "epoch": 0.007581221734941535,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0002,
+      "loss": 1.6891,
+      "step": 360
+    },
+    {
+      "epoch": 0.007602280684205261,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002,
+      "loss": 1.687,
+      "step": 361
+    },
+    {
+      "epoch": 0.007623339633468988,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0002,
+      "loss": 1.6972,
+      "step": 362
+    },
+    {
+      "epoch": 0.007644398582732715,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0002,
+      "loss": 1.6662,
+      "step": 363
+    },
+    {
+      "epoch": 0.007665457531996441,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0002,
+      "loss": 1.6789,
+      "step": 364
+    },
+    {
+      "epoch": 0.007686516481260167,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.6769,
+      "step": 365
+    },
+    {
+      "epoch": 0.007707575430523894,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0002,
+      "loss": 1.7075,
+      "step": 366
+    },
+    {
+      "epoch": 0.0077286343797876205,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0002,
+      "loss": 1.6437,
+      "step": 367
+    },
+    {
+      "epoch": 0.007749693329051347,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.6559,
+      "step": 368
+    },
+    {
+      "epoch": 0.007770752278315074,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.702,
+      "step": 369
+    },
+    {
+      "epoch": 0.0077918112275788,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.7217,
+      "step": 370
+    },
+    {
+      "epoch": 0.007812870176842526,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.6892,
+      "step": 371
+    },
+    {
+      "epoch": 0.007833929126106252,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.6711,
+      "step": 372
+    },
+    {
+      "epoch": 0.00785498807536998,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.6981,
+      "step": 373
+    },
+    {
+      "epoch": 0.007876047024633705,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.7112,
+      "step": 374
+    },
+    {
+      "epoch": 0.007897105973897433,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.6832,
+      "step": 375
+    },
+    {
+      "epoch": 0.007918164923161159,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.6634,
+      "step": 376
+    },
+    {
+      "epoch": 0.007939223872424885,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.6878,
+      "step": 377
+    },
+    {
+      "epoch": 0.007960282821688612,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0002,
+      "loss": 1.6416,
+      "step": 378
+    },
+    {
+      "epoch": 0.007981341770952338,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.6833,
+      "step": 379
+    },
+    {
+      "epoch": 0.008002400720216065,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.6796,
+      "step": 380
+    },
+    {
+      "epoch": 0.008023459669479791,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0002,
+      "loss": 1.6958,
+      "step": 381
+    },
+    {
+      "epoch": 0.008044518618743517,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.7005,
+      "step": 382
+    },
+    {
+      "epoch": 0.008065577568007245,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.6801,
+      "step": 383
+    },
+    {
+      "epoch": 0.00808663651727097,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0002,
+      "loss": 1.694,
+      "step": 384
+    },
+    {
+      "epoch": 0.008107695466534698,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0002,
+      "loss": 1.659,
+      "step": 385
+    },
+    {
+      "epoch": 0.008128754415798424,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6733,
+      "step": 386
+    },
+    {
+      "epoch": 0.00814981336506215,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0002,
+      "loss": 1.7008,
+      "step": 387
+    },
+    {
+      "epoch": 0.008170872314325877,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0002,
+      "loss": 1.6701,
+      "step": 388
+    },
+    {
+      "epoch": 0.008191931263589603,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0002,
+      "loss": 1.6823,
+      "step": 389
+    },
+    {
+      "epoch": 0.00821299021285333,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.0002,
+      "loss": 1.6684,
+      "step": 390
+    },
+    {
+      "epoch": 0.008234049162117056,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.6669,
+      "step": 391
+    },
+    {
+      "epoch": 0.008255108111380782,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.6896,
+      "step": 392
+    },
+    {
+      "epoch": 0.00827616706064451,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0002,
+      "loss": 1.7016,
+      "step": 393
+    },
+    {
+      "epoch": 0.008297226009908235,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0002,
+      "loss": 1.6645,
+      "step": 394
+    },
+    {
+      "epoch": 0.008318284959171963,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.6648,
+      "step": 395
+    },
+    {
+      "epoch": 0.008339343908435689,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.0002,
+      "loss": 1.689,
+      "step": 396
+    },
+    {
+      "epoch": 0.008360402857699414,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.6276,
+      "step": 397
+    },
+    {
+      "epoch": 0.008381461806963142,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.6383,
+      "step": 398
+    },
+    {
+      "epoch": 0.008402520756226868,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0002,
+      "loss": 1.6336,
+      "step": 399
+    },
+    {
+      "epoch": 0.008423579705490595,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.6381,
+      "step": 400
+    },
+    {
+      "epoch": 0.008444638654754321,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.6355,
+      "step": 401
+    },
+    {
+      "epoch": 0.008465697604018047,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0002,
+      "loss": 1.6611,
+      "step": 402
+    },
+    {
+      "epoch": 0.008486756553281774,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.6604,
+      "step": 403
+    },
+    {
+      "epoch": 0.0085078155025455,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6407,
+      "step": 404
+    },
+    {
+      "epoch": 0.008528874451809228,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.6809,
+      "step": 405
+    },
+    {
+      "epoch": 0.008549933401072954,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6693,
+      "step": 406
+    },
+    {
+      "epoch": 0.00857099235033668,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0002,
+      "loss": 1.6621,
+      "step": 407
+    },
+    {
+      "epoch": 0.008592051299600407,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0002,
+      "loss": 1.679,
+      "step": 408
+    },
+    {
+      "epoch": 0.008613110248864133,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0002,
+      "loss": 1.6417,
+      "step": 409
+    },
+    {
+      "epoch": 0.00863416919812786,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0002,
+      "loss": 1.6702,
+      "step": 410
+    },
+    {
+      "epoch": 0.008655228147391586,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0002,
+      "loss": 1.6776,
+      "step": 411
+    },
+    {
+      "epoch": 0.008676287096655312,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.668,
+      "step": 412
+    },
+    {
+      "epoch": 0.00869734604591904,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0002,
+      "loss": 1.6659,
+      "step": 413
+    },
+    {
+      "epoch": 0.008718404995182765,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0002,
+      "loss": 1.6605,
+      "step": 414
+    },
+    {
+      "epoch": 0.008739463944446493,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.6779,
+      "step": 415
+    },
+    {
+      "epoch": 0.008760522893710218,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.6728,
+      "step": 416
+    },
+    {
+      "epoch": 0.008781581842973944,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0002,
+      "loss": 1.6723,
+      "step": 417
+    },
+    {
+      "epoch": 0.008802640792237672,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.686,
+      "step": 418
+    },
+    {
+      "epoch": 0.008823699741501398,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.6508,
+      "step": 419
+    },
+    {
+      "epoch": 0.008844758690765125,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6437,
+      "step": 420
+    },
+    {
+      "epoch": 0.008865817640028851,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0002,
+      "loss": 1.6623,
+      "step": 421
+    },
+    {
+      "epoch": 0.008886876589292577,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.6476,
+      "step": 422
+    },
+    {
+      "epoch": 0.008907935538556304,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6803,
+      "step": 423
+    },
+    {
+      "epoch": 0.00892899448782003,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.6654,
+      "step": 424
+    },
+    {
+      "epoch": 0.008950053437083758,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0002,
+      "loss": 1.7083,
+      "step": 425
+    },
+    {
+      "epoch": 0.008971112386347483,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0002,
+      "loss": 1.6945,
+      "step": 426
+    },
+    {
+      "epoch": 0.00899217133561121,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.6757,
+      "step": 427
+    },
+    {
+      "epoch": 0.009013230284874937,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.6935,
+      "step": 428
+    },
+    {
+      "epoch": 0.009034289234138662,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6196,
+      "step": 429
+    },
+    {
+      "epoch": 0.009055348183402388,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.6461,
+      "step": 430
+    },
+    {
+      "epoch": 0.009076407132666116,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0002,
+      "loss": 1.6708,
+      "step": 431
+    },
+    {
+      "epoch": 0.009097466081929842,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.664,
+      "step": 432
+    },
+    {
+      "epoch": 0.00911852503119357,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.6939,
+      "step": 433
+    },
+    {
+      "epoch": 0.009139583980457295,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6849,
+      "step": 434
+    },
+    {
+      "epoch": 0.00916064292972102,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0002,
+      "loss": 1.6325,
+      "step": 435
+    },
+    {
+      "epoch": 0.009181701878984748,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6732,
+      "step": 436
+    },
+    {
+      "epoch": 0.009202760828248474,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.6798,
+      "step": 437
+    },
+    {
+      "epoch": 0.009223819777512202,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0002,
+      "loss": 1.6865,
+      "step": 438
+    },
+    {
+      "epoch": 0.009244878726775927,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6791,
+      "step": 439
+    },
+    {
+      "epoch": 0.009265937676039653,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.661,
+      "step": 440
+    },
+    {
+      "epoch": 0.00928699662530338,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.6631,
+      "step": 441
+    },
+    {
+      "epoch": 0.009308055574567107,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.668,
+      "step": 442
+    },
+    {
+      "epoch": 0.009329114523830834,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0002,
+      "loss": 1.6636,
+      "step": 443
+    },
+    {
+      "epoch": 0.00935017347309456,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.6786,
+      "step": 444
+    },
+    {
+      "epoch": 0.009371232422358286,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.6953,
+      "step": 445
+    },
+    {
+      "epoch": 0.009392291371622013,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.6817,
+      "step": 446
+    },
+    {
+      "epoch": 0.009413350320885739,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.6796,
+      "step": 447
+    },
+    {
+      "epoch": 0.009434409270149467,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.6598,
+      "step": 448
+    },
+    {
+      "epoch": 0.009455468219413192,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6756,
+      "step": 449
+    },
+    {
+      "epoch": 0.009476527168676918,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0002,
+      "loss": 1.6549,
+      "step": 450
+    },
+    {
+      "epoch": 0.009497586117940646,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.6718,
+      "step": 451
+    },
+    {
+      "epoch": 0.009518645067204371,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.6319,
+      "step": 452
+    },
+    {
+      "epoch": 0.009539704016468099,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6852,
+      "step": 453
+    },
+    {
+      "epoch": 0.009560762965731825,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.6703,
+      "step": 454
+    },
+    {
+      "epoch": 0.00958182191499555,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0002,
+      "loss": 1.6476,
+      "step": 455
+    },
+    {
+      "epoch": 0.009602880864259278,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.6494,
+      "step": 456
+    },
+    {
+      "epoch": 0.009623939813523004,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.704,
+      "step": 457
+    },
+    {
+      "epoch": 0.009644998762786731,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6249,
+      "step": 458
+    },
+    {
+      "epoch": 0.009666057712050457,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0002,
+      "loss": 1.647,
+      "step": 459
+    },
+    {
+      "epoch": 0.009687116661314183,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0002,
+      "loss": 1.679,
+      "step": 460
+    },
+    {
+      "epoch": 0.00970817561057791,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.6862,
+      "step": 461
+    },
+    {
+      "epoch": 0.009729234559841636,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.6769,
+      "step": 462
+    },
+    {
+      "epoch": 0.009750293509105364,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.6558,
+      "step": 463
+    },
+    {
+      "epoch": 0.00977135245836909,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0002,
+      "loss": 1.6806,
+      "step": 464
+    },
+    {
+      "epoch": 0.009792411407632816,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.656,
+      "step": 465
+    },
+    {
+      "epoch": 0.009813470356896543,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.6627,
+      "step": 466
+    },
+    {
+      "epoch": 0.009834529306160269,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.6666,
+      "step": 467
+    },
+    {
+      "epoch": 0.009855588255423996,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.6836,
+      "step": 468
+    },
+    {
+      "epoch": 0.009876647204687722,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6361,
+      "step": 469
+    },
+    {
+      "epoch": 0.009897706153951448,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0002,
+      "loss": 1.644,
+      "step": 470
+    },
+    {
+      "epoch": 0.009918765103215176,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.647,
+      "step": 471
+    },
+    {
+      "epoch": 0.009939824052478901,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.6547,
+      "step": 472
+    },
+    {
+      "epoch": 0.009960883001742629,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0002,
+      "loss": 1.6954,
+      "step": 473
+    },
+    {
+      "epoch": 0.009981941951006355,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0002,
+      "loss": 1.6745,
+      "step": 474
+    },
+    {
+      "epoch": 0.01000300090027008,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0002,
+      "loss": 1.6825,
+      "step": 475
+    },
+    {
+      "epoch": 0.010024059849533808,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.6629,
+      "step": 476
+    },
+    {
+      "epoch": 0.010045118798797534,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0002,
+      "loss": 1.6731,
+      "step": 477
+    },
+    {
+      "epoch": 0.010066177748061261,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6636,
+      "step": 478
+    },
+    {
+      "epoch": 0.010087236697324987,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.6697,
+      "step": 479
+    },
+    {
+      "epoch": 0.010108295646588713,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002,
+      "loss": 1.6577,
+      "step": 480
+    },
+    {
+      "epoch": 0.01012935459585244,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.0002,
+      "loss": 1.6549,
+      "step": 481
+    },
+    {
+      "epoch": 0.010150413545116166,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0002,
+      "loss": 1.6758,
+      "step": 482
+    },
+    {
+      "epoch": 0.010171472494379894,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.6576,
+      "step": 483
+    },
+    {
+      "epoch": 0.01019253144364362,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.654,
+      "step": 484
+    },
+    {
+      "epoch": 0.010213590392907345,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6547,
+      "step": 485
+    },
+    {
+      "epoch": 0.010234649342171073,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.6847,
+      "step": 486
+    },
+    {
+      "epoch": 0.010255708291434799,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6548,
+      "step": 487
+    },
+    {
+      "epoch": 0.010276767240698524,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0002,
+      "loss": 1.638,
+      "step": 488
+    },
+    {
+      "epoch": 0.010297826189962252,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.7198,
+      "step": 489
+    },
+    {
+      "epoch": 0.010318885139225978,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0002,
+      "loss": 1.6823,
+      "step": 490
+    },
+    {
+      "epoch": 0.010339944088489705,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6795,
+      "step": 491
+    },
+    {
+      "epoch": 0.010361003037753431,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.6807,
+      "step": 492
+    },
+    {
+      "epoch": 0.010382061987017157,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.6428,
+      "step": 493
+    },
+    {
+      "epoch": 0.010403120936280884,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.6855,
+      "step": 494
+    },
+    {
+      "epoch": 0.01042417988554461,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0002,
+      "loss": 1.658,
+      "step": 495
+    },
+    {
+      "epoch": 0.010445238834808338,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0002,
+      "loss": 1.69,
+      "step": 496
+    },
+    {
+      "epoch": 0.010466297784072064,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.6481,
+      "step": 497
+    },
+    {
+      "epoch": 0.01048735673333579,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.6509,
+      "step": 498
+    },
+    {
+      "epoch": 0.010508415682599517,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.6285,
+      "step": 499
+    },
+    {
+      "epoch": 0.010529474631863243,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0002,
+      "loss": 1.6663,
+      "step": 500
+    },
+    {
+      "epoch": 0.010529474631863243,
+      "eval_loss": 1.6845688819885254,
+      "eval_runtime": 901.4075,
+      "eval_samples_per_second": 68.559,
+      "eval_steps_per_second": 2.143,
+      "step": 500
+    },
+    {
+      "epoch": 0.01055053358112697,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0002,
+      "loss": 1.6882,
+      "step": 501
+    },
+    {
+      "epoch": 0.010571592530390696,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0002,
+      "loss": 1.6395,
+      "step": 502
+    },
+    {
+      "epoch": 0.010592651479654422,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6245,
+      "step": 503
+    },
+    {
+      "epoch": 0.01061371042891815,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.6856,
+      "step": 504
+    },
+    {
+      "epoch": 0.010634769378181875,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0002,
+      "loss": 1.6788,
+      "step": 505
+    },
+    {
+      "epoch": 0.010655828327445603,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.6188,
+      "step": 506
+    },
+    {
+      "epoch": 0.010676887276709329,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.6678,
+      "step": 507
+    },
+    {
+      "epoch": 0.010697946225973054,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002,
+      "loss": 1.6698,
+      "step": 508
+    },
+    {
+      "epoch": 0.010719005175236782,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.6639,
+      "step": 509
+    },
+    {
+      "epoch": 0.010740064124500508,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.6506,
+      "step": 510
+    },
+    {
+      "epoch": 0.010761123073764235,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.6731,
+      "step": 511
+    },
+    {
+      "epoch": 0.010782182023027961,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0002,
+      "loss": 1.6258,
+      "step": 512
+    },
+    {
+      "epoch": 0.010803240972291687,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0002,
+      "loss": 1.6701,
+      "step": 513
+    },
+    {
+      "epoch": 0.010824299921555414,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002,
+      "loss": 1.6559,
+      "step": 514
+    },
+    {
+      "epoch": 0.01084535887081914,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0002,
+      "loss": 1.6539,
+      "step": 515
+    },
+    {
+      "epoch": 0.010866417820082868,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.7107,
+      "step": 516
+    },
+    {
+      "epoch": 0.010887476769346593,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0002,
+      "loss": 1.6344,
+      "step": 517
+    },
+    {
+      "epoch": 0.01090853571861032,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0002,
+      "loss": 1.6496,
+      "step": 518
+    },
+    {
+      "epoch": 0.010929594667874047,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0002,
+      "loss": 1.6438,
+      "step": 519
+    },
+    {
+      "epoch": 0.010950653617137773,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0002,
+      "loss": 1.7041,
+      "step": 520
+    },
+    {
+      "epoch": 0.0109717125664015,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.6439,
+      "step": 521
+    },
+    {
+      "epoch": 0.010992771515665226,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0002,
+      "loss": 1.6754,
+      "step": 522
+    },
+    {
+      "epoch": 0.011013830464928952,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0002,
+      "loss": 1.6861,
+      "step": 523
+    },
+    {
+      "epoch": 0.01103488941419268,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6511,
+      "step": 524
+    },
+    {
+      "epoch": 0.011055948363456405,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0002,
+      "loss": 1.6635,
+      "step": 525
+    },
+    {
+      "epoch": 0.011077007312720133,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0002,
+      "loss": 1.6327,
+      "step": 526
+    },
+    {
+      "epoch": 0.011098066261983858,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0002,
+      "loss": 1.6757,
+      "step": 527
+    },
+    {
+      "epoch": 0.011119125211247584,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.6672,
+      "step": 528
+    },
+    {
+      "epoch": 0.011140184160511312,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0002,
+      "loss": 1.6608,
+      "step": 529
+    },
+    {
+      "epoch": 0.011161243109775038,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0002,
+      "loss": 1.6848,
+      "step": 530
+    },
+    {
+      "epoch": 0.011182302059038765,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0002,
+      "loss": 1.6913,
+      "step": 531
+    },
+    {
+      "epoch": 0.01120336100830249,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0002,
+      "loss": 1.6567,
+      "step": 532
+    },
+    {
+      "epoch": 0.011224419957566217,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.6535,
+      "step": 533
+    },
+    {
+      "epoch": 0.011245478906829944,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0002,
+      "loss": 1.6928,
+      "step": 534
+    },
+    {
+      "epoch": 0.01126653785609367,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.6937,
+      "step": 535
+    },
+    {
+      "epoch": 0.011287596805357398,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.6642,
+      "step": 536
+    },
+    {
+      "epoch": 0.011308655754621123,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0002,
+      "loss": 1.6671,
+      "step": 537
+    },
+    {
+      "epoch": 0.011329714703884849,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0002,
+      "loss": 1.6381,
+      "step": 538
+    },
+    {
+      "epoch": 0.011350773653148577,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.6785,
+      "step": 539
+    },
+    {
+      "epoch": 0.011371832602412302,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6847,
+      "step": 540
+    },
+    {
+      "epoch": 0.01139289155167603,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0002,
+      "loss": 1.6246,
+      "step": 541
+    },
+    {
+      "epoch": 0.011413950500939756,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.6677,
+      "step": 542
+    },
+    {
+      "epoch": 0.011435009450203482,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.6887,
+      "step": 543
+    },
+    {
+      "epoch": 0.011456068399467209,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0002,
+      "loss": 1.6619,
+      "step": 544
+    },
+    {
+      "epoch": 0.011477127348730935,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.6387,
+      "step": 545
+    },
+    {
+      "epoch": 0.01149818629799466,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6443,
+      "step": 546
+    },
+    {
+      "epoch": 0.011519245247258388,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0002,
+      "loss": 1.6393,
+      "step": 547
+    },
+    {
+      "epoch": 0.011540304196522114,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.6595,
+      "step": 548
+    },
+    {
+      "epoch": 0.011561363145785842,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.6662,
+      "step": 549
+    },
+    {
+      "epoch": 0.011582422095049567,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.6484,
+      "step": 550
+    },
+    {
+      "epoch": 0.011603481044313293,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0002,
+      "loss": 1.6609,
+      "step": 551
+    },
+    {
+      "epoch": 0.01162453999357702,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.6948,
+      "step": 552
+    },
+    {
+      "epoch": 0.011645598942840747,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0002,
+      "loss": 1.6618,
+      "step": 553
+    },
+    {
+      "epoch": 0.011666657892104474,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.6587,
+      "step": 554
+    },
+    {
+      "epoch": 0.0116877168413682,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0002,
+      "loss": 1.6424,
+      "step": 555
+    },
+    {
+      "epoch": 0.011708775790631926,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.6329,
+      "step": 556
+    },
+    {
+      "epoch": 0.011729834739895653,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.6556,
+      "step": 557
+    },
+    {
+      "epoch": 0.011750893689159379,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0002,
+      "loss": 1.6084,
+      "step": 558
+    },
+    {
+      "epoch": 0.011771952638423107,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0002,
+      "loss": 1.6447,
+      "step": 559
+    },
+    {
+      "epoch": 0.011793011587686832,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.6314,
+      "step": 560
+    },
+    {
+      "epoch": 0.011814070536950558,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0002,
+      "loss": 1.6335,
+      "step": 561
+    },
+    {
+      "epoch": 0.011835129486214286,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0002,
+      "loss": 1.6567,
+      "step": 562
+    },
+    {
+      "epoch": 0.011856188435478011,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6627,
+      "step": 563
+    },
+    {
+      "epoch": 0.011877247384741739,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.6662,
+      "step": 564
+    },
+    {
+      "epoch": 0.011898306334005465,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.6493,
+      "step": 565
+    },
+    {
+      "epoch": 0.01191936528326919,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.6504,
+      "step": 566
+    },
+    {
+      "epoch": 0.011940424232532918,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.6583,
+      "step": 567
+    },
+    {
+      "epoch": 0.011961483181796644,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0002,
+      "loss": 1.6457,
+      "step": 568
+    },
+    {
+      "epoch": 0.011982542131060371,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0002,
+      "loss": 1.6701,
+      "step": 569
+    },
+    {
+      "epoch": 0.012003601080324097,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0002,
+      "loss": 1.6665,
+      "step": 570
+    },
+    {
+      "epoch": 0.012024660029587823,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.6699,
+      "step": 571
+    },
+    {
+      "epoch": 0.01204571897885155,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.6524,
+      "step": 572
+    },
+    {
+      "epoch": 0.012066777928115276,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0002,
+      "loss": 1.6298,
+      "step": 573
+    },
+    {
+      "epoch": 0.012087836877379004,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0002,
+      "loss": 1.6595,
+      "step": 574
+    },
+    {
+      "epoch": 0.01210889582664273,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0002,
+      "loss": 1.6377,
+      "step": 575
+    },
+    {
+      "epoch": 0.012129954775906455,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.6637,
+      "step": 576
+    },
+    {
+      "epoch": 0.012151013725170183,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0002,
+      "loss": 1.7005,
+      "step": 577
+    },
+    {
+      "epoch": 0.012172072674433909,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.6646,
+      "step": 578
+    },
+    {
+      "epoch": 0.012193131623697636,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0002,
+      "loss": 1.6449,
+      "step": 579
+    },
+    {
+      "epoch": 0.012214190572961362,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.6782,
+      "step": 580
+    },
+    {
+      "epoch": 0.012235249522225088,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0002,
+      "loss": 1.6523,
+      "step": 581
+    },
+    {
+      "epoch": 0.012256308471488815,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0002,
+      "loss": 1.6457,
+      "step": 582
+    },
+    {
+      "epoch": 0.012277367420752541,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.6428,
+      "step": 583
+    },
+    {
+      "epoch": 0.012298426370016269,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0002,
+      "loss": 1.6509,
+      "step": 584
+    },
+    {
+      "epoch": 0.012319485319279995,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0002,
+      "loss": 1.6971,
+      "step": 585
+    },
+    {
+      "epoch": 0.01234054426854372,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.635,
+      "step": 586
+    },
+    {
+      "epoch": 0.012361603217807448,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0002,
+      "loss": 1.6193,
+      "step": 587
+    },
+    {
+      "epoch": 0.012382662167071174,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6599,
+      "step": 588
+    },
+    {
+      "epoch": 0.012403721116334901,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.6789,
+      "step": 589
+    },
+    {
+      "epoch": 0.012424780065598627,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.6879,
+      "step": 590
+    },
+    {
+      "epoch": 0.012445839014862353,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0002,
+      "loss": 1.6231,
+      "step": 591
+    },
+    {
+      "epoch": 0.01246689796412608,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.6422,
+      "step": 592
+    },
+    {
+      "epoch": 0.012487956913389806,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.654,
+      "step": 593
+    },
+    {
+      "epoch": 0.012509015862653534,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.621,
+      "step": 594
+    },
+    {
+      "epoch": 0.01253007481191726,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.6217,
+      "step": 595
+    },
+    {
+      "epoch": 0.012551133761180985,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.6868,
+      "step": 596
+    },
+    {
+      "epoch": 0.012572192710444713,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0002,
+      "loss": 1.6763,
+      "step": 597
+    },
+    {
+      "epoch": 0.012593251659708439,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.6488,
+      "step": 598
+    },
+    {
+      "epoch": 0.012614310608972164,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.6702,
+      "step": 599
+    },
+    {
+      "epoch": 0.012635369558235892,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.6282,
+      "step": 600
+    },
+    {
+      "epoch": 0.012656428507499618,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.6202,
+      "step": 601
+    },
+    {
+      "epoch": 0.012677487456763345,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.6627,
+      "step": 602
+    },
+    {
+      "epoch": 0.012698546406027071,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.6607,
+      "step": 603
+    },
+    {
+      "epoch": 0.012719605355290797,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0002,
+      "loss": 1.6873,
+      "step": 604
+    },
+    {
+      "epoch": 0.012740664304554524,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0002,
+      "loss": 1.6425,
+      "step": 605
+    },
+    {
+      "epoch": 0.01276172325381825,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.6564,
+      "step": 606
+    },
+    {
+      "epoch": 0.012782782203081978,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.6118,
+      "step": 607
+    },
+    {
+      "epoch": 0.012803841152345704,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0002,
+      "loss": 1.6689,
+      "step": 608
+    },
+    {
+      "epoch": 0.01282490010160943,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0002,
+      "loss": 1.6296,
+      "step": 609
+    },
+    {
+      "epoch": 0.012845959050873157,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0002,
+      "loss": 1.6626,
+      "step": 610
+    },
+    {
+      "epoch": 0.012867018000136883,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0002,
+      "loss": 1.6505,
+      "step": 611
+    },
+    {
+      "epoch": 0.01288807694940061,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0002,
+      "loss": 1.6708,
+      "step": 612
+    },
+    {
+      "epoch": 0.012909135898664336,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0002,
+      "loss": 1.6498,
+      "step": 613
+    },
+    {
+      "epoch": 0.012930194847928062,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0002,
+      "loss": 1.6399,
+      "step": 614
+    },
+    {
+      "epoch": 0.01295125379719179,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.6354,
+      "step": 615
+    },
+    {
+      "epoch": 0.012972312746455515,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0002,
+      "loss": 1.6614,
+      "step": 616
+    },
+    {
+      "epoch": 0.012993371695719243,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.6468,
+      "step": 617
+    },
+    {
+      "epoch": 0.013014430644982969,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.5986,
+      "step": 618
+    },
+    {
+      "epoch": 0.013035489594246694,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0002,
+      "loss": 1.662,
+      "step": 619
+    },
+    {
+      "epoch": 0.013056548543510422,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002,
+      "loss": 1.6803,
+      "step": 620
+    },
+    {
+      "epoch": 0.013077607492774148,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0002,
+      "loss": 1.6713,
+      "step": 621
+    },
+    {
+      "epoch": 0.013098666442037875,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.6621,
+      "step": 622
+    },
+    {
+      "epoch": 0.013119725391301601,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.6544,
+      "step": 623
+    },
+    {
+      "epoch": 0.013140784340565327,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0002,
+      "loss": 1.6452,
+      "step": 624
+    },
+    {
+      "epoch": 0.013161843289829054,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.6553,
+      "step": 625
+    },
+    {
+      "epoch": 0.01318290223909278,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.6345,
+      "step": 626
+    },
+    {
+      "epoch": 0.013203961188356508,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0002,
+      "loss": 1.6224,
+      "step": 627
+    },
+    {
+      "epoch": 0.013225020137620233,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.6556,
+      "step": 628
+    },
+    {
+      "epoch": 0.01324607908688396,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.6781,
+      "step": 629
+    },
+    {
+      "epoch": 0.013267138036147687,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0002,
+      "loss": 1.6771,
+      "step": 630
+    },
+    {
+      "epoch": 0.013288196985411413,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.6476,
+      "step": 631
+    },
+    {
+      "epoch": 0.01330925593467514,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0002,
+      "loss": 1.6193,
+      "step": 632
+    },
+    {
+      "epoch": 0.013330314883938866,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0002,
+      "loss": 1.6583,
+      "step": 633
+    },
+    {
+      "epoch": 0.013351373833202592,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.6644,
+      "step": 634
+    },
+    {
+      "epoch": 0.01337243278246632,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.6457,
+      "step": 635
+    },
+    {
+      "epoch": 0.013393491731730045,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0002,
+      "loss": 1.6568,
+      "step": 636
+    },
+    {
+      "epoch": 0.013414550680993773,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6341,
+      "step": 637
+    },
+    {
+      "epoch": 0.013435609630257498,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.6321,
+      "step": 638
+    },
+    {
+      "epoch": 0.013456668579521224,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.6557,
+      "step": 639
+    },
+    {
+      "epoch": 0.013477727528784952,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0002,
+      "loss": 1.6666,
+      "step": 640
+    },
+    {
+      "epoch": 0.013498786478048677,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6195,
+      "step": 641
+    },
+    {
+      "epoch": 0.013519845427312405,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0002,
+      "loss": 1.6456,
+      "step": 642
+    },
+    {
+      "epoch": 0.01354090437657613,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.6543,
+      "step": 643
+    },
+    {
+      "epoch": 0.013561963325839857,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6581,
+      "step": 644
+    },
+    {
+      "epoch": 0.013583022275103584,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002,
+      "loss": 1.6332,
+      "step": 645
+    },
+    {
+      "epoch": 0.01360408122436731,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0002,
+      "loss": 1.6455,
+      "step": 646
+    },
+    {
+      "epoch": 0.013625140173631037,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.6543,
+      "step": 647
+    },
+    {
+      "epoch": 0.013646199122894763,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0002,
+      "loss": 1.644,
+      "step": 648
+    },
+    {
+      "epoch": 0.013667258072158489,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0002,
+      "loss": 1.6882,
+      "step": 649
+    },
+    {
+      "epoch": 0.013688317021422217,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0002,
+      "loss": 1.6404,
+      "step": 650
+    },
+    {
+      "epoch": 0.013709375970685942,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.6521,
+      "step": 651
+    },
+    {
+      "epoch": 0.01373043491994967,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.6522,
+      "step": 652
+    },
+    {
+      "epoch": 0.013751493869213396,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0002,
+      "loss": 1.6337,
+      "step": 653
+    },
+    {
+      "epoch": 0.013772552818477122,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0002,
+      "loss": 1.6688,
+      "step": 654
+    },
+    {
+      "epoch": 0.013793611767740849,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.655,
+      "step": 655
+    },
+    {
+      "epoch": 0.013814670717004575,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.6473,
+      "step": 656
+    },
+    {
+      "epoch": 0.0138357296662683,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0002,
+      "loss": 1.6272,
+      "step": 657
+    },
+    {
+      "epoch": 0.013856788615532028,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6573,
+      "step": 658
+    },
+    {
+      "epoch": 0.013877847564795754,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.6141,
+      "step": 659
+    },
+    {
+      "epoch": 0.013898906514059482,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6223,
+      "step": 660
+    },
+    {
+      "epoch": 0.013919965463323207,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.6754,
+      "step": 661
+    },
+    {
+      "epoch": 0.013941024412586933,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.6506,
+      "step": 662
+    },
+    {
+      "epoch": 0.01396208336185066,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.6826,
+      "step": 663
+    },
+    {
+      "epoch": 0.013983142311114386,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.6228,
+      "step": 664
+    },
+    {
+      "epoch": 0.014004201260378114,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0002,
+      "loss": 1.6452,
+      "step": 665
+    },
+    {
+      "epoch": 0.01402526020964184,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6496,
+      "step": 666
+    },
+    {
+      "epoch": 0.014046319158905566,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0002,
+      "loss": 1.6473,
+      "step": 667
+    },
+    {
+      "epoch": 0.014067378108169293,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0002,
+      "loss": 1.6585,
+      "step": 668
+    },
+    {
+      "epoch": 0.014088437057433019,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0002,
+      "loss": 1.6528,
+      "step": 669
+    },
+    {
+      "epoch": 0.014109496006696746,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6477,
+      "step": 670
+    },
+    {
+      "epoch": 0.014130554955960472,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.6507,
+      "step": 671
+    },
+    {
+      "epoch": 0.014151613905224198,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.6641,
+      "step": 672
+    },
+    {
+      "epoch": 0.014172672854487926,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.67,
+      "step": 673
+    },
+    {
+      "epoch": 0.014193731803751651,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.6261,
+      "step": 674
+    },
+    {
+      "epoch": 0.014214790753015379,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.6615,
+      "step": 675
+    },
+    {
+      "epoch": 0.014235849702279105,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.0002,
+      "loss": 1.6664,
+      "step": 676
+    },
+    {
+      "epoch": 0.01425690865154283,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0002,
+      "loss": 1.6183,
+      "step": 677
+    },
+    {
+      "epoch": 0.014277967600806558,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.6347,
+      "step": 678
+    },
+    {
+      "epoch": 0.014299026550070284,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0002,
+      "loss": 1.6616,
+      "step": 679
+    },
+    {
+      "epoch": 0.014320085499334011,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6433,
+      "step": 680
+    },
+    {
+      "epoch": 0.014341144448597737,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.6371,
+      "step": 681
+    },
+    {
+      "epoch": 0.014362203397861463,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6353,
+      "step": 682
+    },
+    {
+      "epoch": 0.01438326234712519,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.6687,
+      "step": 683
+    },
+    {
+      "epoch": 0.014404321296388916,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0002,
+      "loss": 1.6445,
+      "step": 684
+    },
+    {
+      "epoch": 0.014425380245652644,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6346,
+      "step": 685
+    },
+    {
+      "epoch": 0.01444643919491637,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6891,
+      "step": 686
+    },
+    {
+      "epoch": 0.014467498144180095,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6539,
+      "step": 687
+    },
+    {
+      "epoch": 0.014488557093443823,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.6643,
+      "step": 688
+    },
+    {
+      "epoch": 0.014509616042707549,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0002,
+      "loss": 1.6504,
+      "step": 689
+    },
+    {
+      "epoch": 0.014530674991971276,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6665,
+      "step": 690
+    },
+    {
+      "epoch": 0.014551733941235002,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.6852,
+      "step": 691
+    },
+    {
+      "epoch": 0.014572792890498728,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.6548,
+      "step": 692
+    },
+    {
+      "epoch": 0.014593851839762455,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6446,
+      "step": 693
+    },
+    {
+      "epoch": 0.014614910789026181,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0002,
+      "loss": 1.6912,
+      "step": 694
+    },
+    {
+      "epoch": 0.014635969738289909,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6635,
+      "step": 695
+    },
+    {
+      "epoch": 0.014657028687553635,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.6431,
+      "step": 696
+    },
+    {
+      "epoch": 0.01467808763681736,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.6485,
+      "step": 697
+    },
+    {
+      "epoch": 0.014699146586081088,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0002,
+      "loss": 1.6323,
+      "step": 698
+    },
+    {
+      "epoch": 0.014720205535344814,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.5934,
+      "step": 699
+    },
+    {
+      "epoch": 0.014741264484608541,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0002,
+      "loss": 1.6785,
+      "step": 700
+    },
+    {
+      "epoch": 0.014762323433872267,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.6674,
+      "step": 701
+    },
+    {
+      "epoch": 0.014783382383135993,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0002,
+      "loss": 1.6426,
+      "step": 702
+    },
+    {
+      "epoch": 0.01480444133239972,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6188,
+      "step": 703
+    },
+    {
+      "epoch": 0.014825500281663446,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6374,
+      "step": 704
+    },
+    {
+      "epoch": 0.014846559230927174,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0002,
+      "loss": 1.648,
+      "step": 705
+    },
+    {
+      "epoch": 0.0148676181801909,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.6413,
+      "step": 706
+    },
+    {
+      "epoch": 0.014888677129454625,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0002,
+      "loss": 1.6218,
+      "step": 707
+    },
+    {
+      "epoch": 0.014909736078718353,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.6394,
+      "step": 708
+    },
+    {
+      "epoch": 0.014930795027982079,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0002,
+      "loss": 1.6801,
+      "step": 709
+    },
+    {
+      "epoch": 0.014951853977245806,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.6215,
+      "step": 710
+    },
+    {
+      "epoch": 0.014972912926509532,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6423,
+      "step": 711
+    },
+    {
+      "epoch": 0.014993971875773258,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.6765,
+      "step": 712
+    },
+    {
+      "epoch": 0.015015030825036985,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0002,
+      "loss": 1.6326,
+      "step": 713
+    },
+    {
+      "epoch": 0.015036089774300711,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.66,
+      "step": 714
+    },
+    {
+      "epoch": 0.015057148723564437,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.6558,
+      "step": 715
+    },
+    {
+      "epoch": 0.015078207672828164,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0002,
+      "loss": 1.6495,
+      "step": 716
+    },
+    {
+      "epoch": 0.01509926662209189,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6677,
+      "step": 717
+    },
+    {
+      "epoch": 0.015120325571355618,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.6524,
+      "step": 718
+    },
+    {
+      "epoch": 0.015141384520619344,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.6446,
+      "step": 719
+    },
+    {
+      "epoch": 0.01516244346988307,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0002,
+      "loss": 1.6619,
+      "step": 720
+    },
+    {
+      "epoch": 0.015183502419146797,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0002,
+      "loss": 1.626,
+      "step": 721
+    },
+    {
+      "epoch": 0.015204561368410523,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.6303,
+      "step": 722
+    },
+    {
+      "epoch": 0.01522562031767425,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.6285,
+      "step": 723
+    },
+    {
+      "epoch": 0.015246679266937976,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.6365,
+      "step": 724
+    },
+    {
+      "epoch": 0.015267738216201702,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0002,
+      "loss": 1.6346,
+      "step": 725
+    },
+    {
+      "epoch": 0.01528879716546543,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.6301,
+      "step": 726
+    },
+    {
+      "epoch": 0.015309856114729155,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.6371,
+      "step": 727
+    },
+    {
+      "epoch": 0.015330915063992883,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.6369,
+      "step": 728
+    },
+    {
+      "epoch": 0.015351974013256608,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0002,
+      "loss": 1.6331,
+      "step": 729
+    },
+    {
+      "epoch": 0.015373032962520334,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6505,
+      "step": 730
+    },
+    {
+      "epoch": 0.015394091911784062,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0002,
+      "loss": 1.6174,
+      "step": 731
+    },
+    {
+      "epoch": 0.015415150861047788,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0002,
+      "loss": 1.6263,
+      "step": 732
+    },
+    {
+      "epoch": 0.015436209810311515,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.6613,
+      "step": 733
+    },
+    {
+      "epoch": 0.015457268759575241,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.6725,
+      "step": 734
+    },
+    {
+      "epoch": 0.015478327708838967,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0002,
+      "loss": 1.634,
+      "step": 735
+    },
+    {
+      "epoch": 0.015499386658102694,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.6206,
+      "step": 736
+    },
+    {
+      "epoch": 0.01552044560736642,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0002,
+      "loss": 1.6131,
+      "step": 737
+    },
+    {
+      "epoch": 0.015541504556630148,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0002,
+      "loss": 1.6379,
+      "step": 738
+    },
+    {
+      "epoch": 0.015562563505893873,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0002,
+      "loss": 1.6567,
+      "step": 739
+    },
+    {
+      "epoch": 0.0155836224551576,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.6372,
+      "step": 740
+    },
+    {
+      "epoch": 0.015604681404421327,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6601,
+      "step": 741
+    },
+    {
+      "epoch": 0.015625740353685053,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.633,
+      "step": 742
+    },
+    {
+      "epoch": 0.01564679930294878,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.6288,
+      "step": 743
+    },
+    {
+      "epoch": 0.015667858252212504,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.6437,
+      "step": 744
+    },
+    {
+      "epoch": 0.01568891720147623,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.6546,
+      "step": 745
+    },
+    {
+      "epoch": 0.01570997615073996,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.6488,
+      "step": 746
+    },
+    {
+      "epoch": 0.015731035100003687,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0002,
+      "loss": 1.6262,
+      "step": 747
+    },
+    {
+      "epoch": 0.01575209404926741,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6411,
+      "step": 748
+    },
+    {
+      "epoch": 0.01577315299853114,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6126,
+      "step": 749
+    },
+    {
+      "epoch": 0.015794211947794866,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0002,
+      "loss": 1.6531,
+      "step": 750
+    },
+    {
+      "epoch": 0.01581527089705859,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.6314,
+      "step": 751
+    },
+    {
+      "epoch": 0.015836329846322317,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0002,
+      "loss": 1.6277,
+      "step": 752
+    },
+    {
+      "epoch": 0.015857388795586045,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.6466,
+      "step": 753
+    },
+    {
+      "epoch": 0.01587844774484977,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.652,
+      "step": 754
+    },
+    {
+      "epoch": 0.015899506694113497,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.6087,
+      "step": 755
+    },
+    {
+      "epoch": 0.015920565643377224,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0002,
+      "loss": 1.6437,
+      "step": 756
+    },
+    {
+      "epoch": 0.01594162459264095,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0002,
+      "loss": 1.6394,
+      "step": 757
+    },
+    {
+      "epoch": 0.015962683541904676,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.6047,
+      "step": 758
+    },
+    {
+      "epoch": 0.015983742491168403,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0002,
+      "loss": 1.6286,
+      "step": 759
+    },
+    {
+      "epoch": 0.01600480144043213,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6467,
+      "step": 760
+    },
+    {
+      "epoch": 0.016025860389695855,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0002,
+      "loss": 1.6179,
+      "step": 761
+    },
+    {
+      "epoch": 0.016046919338959582,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0002,
+      "loss": 1.6751,
+      "step": 762
+    },
+    {
+      "epoch": 0.01606797828822331,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0002,
+      "loss": 1.6604,
+      "step": 763
+    },
+    {
+      "epoch": 0.016089037237487034,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0002,
+      "loss": 1.6213,
+      "step": 764
+    },
+    {
+      "epoch": 0.01611009618675076,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.654,
+      "step": 765
+    },
+    {
+      "epoch": 0.01613115513601449,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.6117,
+      "step": 766
+    },
+    {
+      "epoch": 0.016152214085278217,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.6573,
+      "step": 767
+    },
+    {
+      "epoch": 0.01617327303454194,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6697,
+      "step": 768
+    },
+    {
+      "epoch": 0.016194331983805668,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0002,
+      "loss": 1.6277,
+      "step": 769
+    },
+    {
+      "epoch": 0.016215390933069396,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.6315,
+      "step": 770
+    },
+    {
+      "epoch": 0.01623644988233312,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6593,
+      "step": 771
+    },
+    {
+      "epoch": 0.016257508831596847,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.648,
+      "step": 772
+    },
+    {
+      "epoch": 0.016278567780860575,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0002,
+      "loss": 1.6388,
+      "step": 773
+    },
+    {
+      "epoch": 0.0162996267301243,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0002,
+      "loss": 1.6772,
+      "step": 774
+    },
+    {
+      "epoch": 0.016320685679388026,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.6438,
+      "step": 775
+    },
+    {
+      "epoch": 0.016341744628651754,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0002,
+      "loss": 1.6444,
+      "step": 776
+    },
+    {
+      "epoch": 0.01636280357791548,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0002,
+      "loss": 1.6093,
+      "step": 777
+    },
+    {
+      "epoch": 0.016383862527179206,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6548,
+      "step": 778
+    },
+    {
+      "epoch": 0.016404921476442933,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0002,
+      "loss": 1.6248,
+      "step": 779
+    },
+    {
+      "epoch": 0.01642598042570666,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0002,
+      "loss": 1.6233,
+      "step": 780
+    },
+    {
+      "epoch": 0.016447039374970385,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.6477,
+      "step": 781
+    },
+    {
+      "epoch": 0.016468098324234112,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6413,
+      "step": 782
+    },
+    {
+      "epoch": 0.01648915727349784,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0002,
+      "loss": 1.6376,
+      "step": 783
+    },
+    {
+      "epoch": 0.016510216222761564,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0002,
+      "loss": 1.6424,
+      "step": 784
+    },
+    {
+      "epoch": 0.01653127517202529,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0002,
+      "loss": 1.6407,
+      "step": 785
+    },
+    {
+      "epoch": 0.01655233412128902,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.6373,
+      "step": 786
+    },
+    {
+      "epoch": 0.016573393070552746,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0002,
+      "loss": 1.6551,
+      "step": 787
+    },
+    {
+      "epoch": 0.01659445201981647,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0002,
+      "loss": 1.6339,
+      "step": 788
+    },
+    {
+      "epoch": 0.016615510969080198,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0002,
+      "loss": 1.6303,
+      "step": 789
+    },
+    {
+      "epoch": 0.016636569918343926,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.6764,
+      "step": 790
+    },
+    {
+      "epoch": 0.01665762886760765,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.6175,
+      "step": 791
+    },
+    {
+      "epoch": 0.016678687816871377,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.64,
+      "step": 792
+    },
+    {
+      "epoch": 0.016699746766135105,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.635,
+      "step": 793
+    },
+    {
+      "epoch": 0.01672080571539883,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.6122,
+      "step": 794
+    },
+    {
+      "epoch": 0.016741864664662556,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.6562,
+      "step": 795
+    },
+    {
+      "epoch": 0.016762923613926284,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.6398,
+      "step": 796
+    },
+    {
+      "epoch": 0.01678398256319001,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.6267,
+      "step": 797
+    },
+    {
+      "epoch": 0.016805041512453735,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0002,
+      "loss": 1.6705,
+      "step": 798
+    },
+    {
+      "epoch": 0.016826100461717463,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0002,
+      "loss": 1.6333,
+      "step": 799
+    },
+    {
+      "epoch": 0.01684715941098119,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.6618,
+      "step": 800
+    },
+    {
+      "epoch": 0.016868218360244915,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0002,
+      "loss": 1.6206,
+      "step": 801
+    },
+    {
+      "epoch": 0.016889277309508642,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0002,
+      "loss": 1.6114,
+      "step": 802
+    },
+    {
+      "epoch": 0.01691033625877237,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6424,
+      "step": 803
+    },
+    {
+      "epoch": 0.016931395208036094,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.6494,
+      "step": 804
+    },
+    {
+      "epoch": 0.01695245415729982,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0002,
+      "loss": 1.6269,
+      "step": 805
+    },
+    {
+      "epoch": 0.01697351310656355,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0002,
+      "loss": 1.609,
+      "step": 806
+    },
+    {
+      "epoch": 0.016994572055827273,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0002,
+      "loss": 1.645,
+      "step": 807
+    },
+    {
+      "epoch": 0.017015631005091,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.668,
+      "step": 808
+    },
+    {
+      "epoch": 0.017036689954354728,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.6493,
+      "step": 809
+    },
+    {
+      "epoch": 0.017057748903618455,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.6446,
+      "step": 810
+    },
+    {
+      "epoch": 0.01707880785288218,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0002,
+      "loss": 1.663,
+      "step": 811
+    },
+    {
+      "epoch": 0.017099866802145907,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0002,
+      "loss": 1.6333,
+      "step": 812
+    },
+    {
+      "epoch": 0.017120925751409635,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.6601,
+      "step": 813
+    },
+    {
+      "epoch": 0.01714198470067336,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0002,
+      "loss": 1.6445,
+      "step": 814
+    },
+    {
+      "epoch": 0.017163043649937086,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.6298,
+      "step": 815
+    },
+    {
+      "epoch": 0.017184102599200814,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6612,
+      "step": 816
+    },
+    {
+      "epoch": 0.017205161548464538,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0002,
+      "loss": 1.6217,
+      "step": 817
+    },
+    {
+      "epoch": 0.017226220497728265,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0002,
+      "loss": 1.6361,
+      "step": 818
+    },
+    {
+      "epoch": 0.017247279446991993,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6586,
+      "step": 819
+    },
+    {
+      "epoch": 0.01726833839625572,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0002,
+      "loss": 1.6218,
+      "step": 820
+    },
+    {
+      "epoch": 0.017289397345519444,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0002,
+      "loss": 1.6224,
+      "step": 821
+    },
+    {
+      "epoch": 0.017310456294783172,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.625,
+      "step": 822
+    },
+    {
+      "epoch": 0.0173315152440469,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0002,
+      "loss": 1.6813,
+      "step": 823
+    },
+    {
+      "epoch": 0.017352574193310624,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0002,
+      "loss": 1.6438,
+      "step": 824
+    },
+    {
+      "epoch": 0.01737363314257435,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6271,
+      "step": 825
+    },
+    {
+      "epoch": 0.01739469209183808,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0002,
+      "loss": 1.6339,
+      "step": 826
+    },
+    {
+      "epoch": 0.017415751041101803,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0002,
+      "loss": 1.6257,
+      "step": 827
+    },
+    {
+      "epoch": 0.01743680999036553,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.6781,
+      "step": 828
+    },
+    {
+      "epoch": 0.017457868939629258,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0002,
+      "loss": 1.6028,
+      "step": 829
+    },
+    {
+      "epoch": 0.017478927888892985,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.6335,
+      "step": 830
+    },
+    {
+      "epoch": 0.01749998683815671,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.6194,
+      "step": 831
+    },
+    {
+      "epoch": 0.017521045787420437,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.6902,
+      "step": 832
+    },
+    {
+      "epoch": 0.017542104736684164,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.6452,
+      "step": 833
+    },
+    {
+      "epoch": 0.01756316368594789,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.6163,
+      "step": 834
+    },
+    {
+      "epoch": 0.017584222635211616,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.6649,
+      "step": 835
+    },
+    {
+      "epoch": 0.017605281584475344,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.6409,
+      "step": 836
+    },
+    {
+      "epoch": 0.017626340533739068,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6236,
+      "step": 837
+    },
+    {
+      "epoch": 0.017647399483002795,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.6355,
+      "step": 838
+    },
+    {
+      "epoch": 0.017668458432266523,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.6378,
+      "step": 839
+    },
+    {
+      "epoch": 0.01768951738153025,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0002,
+      "loss": 1.6293,
+      "step": 840
+    },
+    {
+      "epoch": 0.017710576330793974,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.6538,
+      "step": 841
+    },
+    {
+      "epoch": 0.017731635280057702,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.6452,
+      "step": 842
+    },
+    {
+      "epoch": 0.01775269422932143,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0002,
+      "loss": 1.6397,
+      "step": 843
+    },
+    {
+      "epoch": 0.017773753178585153,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.5787,
+      "step": 844
+    },
+    {
+      "epoch": 0.01779481212784888,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0002,
+      "loss": 1.6356,
+      "step": 845
+    },
+    {
+      "epoch": 0.01781587107711261,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6498,
+      "step": 846
+    },
+    {
+      "epoch": 0.017836930026376333,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.6298,
+      "step": 847
+    },
+    {
+      "epoch": 0.01785798897564006,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002,
+      "loss": 1.623,
+      "step": 848
+    },
+    {
+      "epoch": 0.017879047924903788,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0002,
+      "loss": 1.649,
+      "step": 849
+    },
+    {
+      "epoch": 0.017900106874167515,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0002,
+      "loss": 1.6529,
+      "step": 850
+    },
+    {
+      "epoch": 0.01792116582343124,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.6536,
+      "step": 851
+    },
+    {
+      "epoch": 0.017942224772694967,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.6239,
+      "step": 852
+    },
+    {
+      "epoch": 0.017963283721958694,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0002,
+      "loss": 1.6501,
+      "step": 853
+    },
+    {
+      "epoch": 0.01798434267122242,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0002,
+      "loss": 1.6017,
+      "step": 854
+    },
+    {
+      "epoch": 0.018005401620486146,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.6279,
+      "step": 855
+    },
+    {
+      "epoch": 0.018026460569749873,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.6455,
+      "step": 856
+    },
+    {
+      "epoch": 0.018047519519013597,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0002,
+      "loss": 1.6619,
+      "step": 857
+    },
+    {
+      "epoch": 0.018068578468277325,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.6174,
+      "step": 858
+    },
+    {
+      "epoch": 0.018089637417541053,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0002,
+      "loss": 1.6331,
+      "step": 859
+    },
+    {
+      "epoch": 0.018110696366804777,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.6318,
+      "step": 860
+    },
+    {
+      "epoch": 0.018131755316068504,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.6585,
+      "step": 861
+    },
+    {
+      "epoch": 0.01815281426533223,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002,
+      "loss": 1.6266,
+      "step": 862
+    },
+    {
+      "epoch": 0.01817387321459596,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0002,
+      "loss": 1.6365,
+      "step": 863
+    },
+    {
+      "epoch": 0.018194932163859683,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.6406,
+      "step": 864
+    },
+    {
+      "epoch": 0.01821599111312341,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0002,
+      "loss": 1.6525,
+      "step": 865
+    },
+    {
+      "epoch": 0.01823705006238714,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0002,
+      "loss": 1.6476,
+      "step": 866
+    },
+    {
+      "epoch": 0.018258109011650862,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6244,
+      "step": 867
+    },
+    {
+      "epoch": 0.01827916796091459,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.5978,
+      "step": 868
+    },
+    {
+      "epoch": 0.018300226910178317,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.6055,
+      "step": 869
+    },
+    {
+      "epoch": 0.01832128585944204,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0002,
+      "loss": 1.6666,
+      "step": 870
+    },
+    {
+      "epoch": 0.01834234480870577,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.5938,
+      "step": 871
+    },
+    {
+      "epoch": 0.018363403757969497,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6593,
+      "step": 872
+    },
+    {
+      "epoch": 0.018384462707233224,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0002,
+      "loss": 1.648,
+      "step": 873
+    },
+    {
+      "epoch": 0.018405521656496948,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.6886,
+      "step": 874
+    },
+    {
+      "epoch": 0.018426580605760676,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.6267,
+      "step": 875
+    },
+    {
+      "epoch": 0.018447639555024403,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.6445,
+      "step": 876
+    },
+    {
+      "epoch": 0.018468698504288127,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.6372,
+      "step": 877
+    },
+    {
+      "epoch": 0.018489757453551855,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.6797,
+      "step": 878
+    },
+    {
+      "epoch": 0.018510816402815582,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.6379,
+      "step": 879
+    },
+    {
+      "epoch": 0.018531875352079306,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.6374,
+      "step": 880
+    },
+    {
+      "epoch": 0.018552934301343034,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0002,
+      "loss": 1.6227,
+      "step": 881
+    },
+    {
+      "epoch": 0.01857399325060676,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0002,
+      "loss": 1.6266,
+      "step": 882
+    },
+    {
+      "epoch": 0.01859505219987049,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.63,
+      "step": 883
+    },
+    {
+      "epoch": 0.018616111149134213,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.6311,
+      "step": 884
+    },
+    {
+      "epoch": 0.01863717009839794,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.6426,
+      "step": 885
+    },
+    {
+      "epoch": 0.018658229047661668,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0002,
+      "loss": 1.6624,
+      "step": 886
+    },
+    {
+      "epoch": 0.018679287996925392,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.6369,
+      "step": 887
+    },
+    {
+      "epoch": 0.01870034694618912,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6102,
+      "step": 888
+    },
+    {
+      "epoch": 0.018721405895452847,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0002,
+      "loss": 1.6295,
+      "step": 889
+    },
+    {
+      "epoch": 0.01874246484471657,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.0002,
+      "loss": 1.6711,
+      "step": 890
+    },
+    {
+      "epoch": 0.0187635237939803,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.628,
+      "step": 891
+    },
+    {
+      "epoch": 0.018784582743244026,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0002,
+      "loss": 1.6459,
+      "step": 892
+    },
+    {
+      "epoch": 0.018805641692507754,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0002,
+      "loss": 1.6375,
+      "step": 893
+    },
+    {
+      "epoch": 0.018826700641771478,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0002,
+      "loss": 1.6038,
+      "step": 894
+    },
+    {
+      "epoch": 0.018847759591035206,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0002,
+      "loss": 1.6387,
+      "step": 895
+    },
+    {
+      "epoch": 0.018868818540298933,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0002,
+      "loss": 1.6274,
+      "step": 896
+    },
+    {
+      "epoch": 0.018889877489562657,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0002,
+      "loss": 1.6084,
+      "step": 897
+    },
+    {
+      "epoch": 0.018910936438826385,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.607,
+      "step": 898
+    },
+    {
+      "epoch": 0.018931995388090112,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0002,
+      "loss": 1.6388,
+      "step": 899
+    },
+    {
+      "epoch": 0.018953054337353836,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0002,
+      "loss": 1.6062,
+      "step": 900
+    },
+    {
+      "epoch": 0.018974113286617564,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6045,
+      "step": 901
+    },
+    {
+      "epoch": 0.01899517223588129,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.5999,
+      "step": 902
+    },
+    {
+      "epoch": 0.01901623118514502,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0002,
+      "loss": 1.6442,
+      "step": 903
+    },
+    {
+      "epoch": 0.019037290134408743,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.648,
+      "step": 904
+    },
+    {
+      "epoch": 0.01905834908367247,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.6321,
+      "step": 905
+    },
+    {
+      "epoch": 0.019079408032936198,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002,
+      "loss": 1.621,
+      "step": 906
+    },
+    {
+      "epoch": 0.019100466982199922,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6163,
+      "step": 907
+    },
+    {
+      "epoch": 0.01912152593146365,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6196,
+      "step": 908
+    },
+    {
+      "epoch": 0.019142584880727377,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.629,
+      "step": 909
+    },
+    {
+      "epoch": 0.0191636438299911,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.665,
+      "step": 910
+    },
+    {
+      "epoch": 0.01918470277925483,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.6315,
+      "step": 911
+    },
+    {
+      "epoch": 0.019205761728518556,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.6545,
+      "step": 912
+    },
+    {
+      "epoch": 0.01922682067778228,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.6372,
+      "step": 913
+    },
+    {
+      "epoch": 0.019247879627046008,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0002,
+      "loss": 1.6359,
+      "step": 914
+    },
+    {
+      "epoch": 0.019268938576309735,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.6585,
+      "step": 915
+    },
+    {
+      "epoch": 0.019289997525573463,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.624,
+      "step": 916
+    },
+    {
+      "epoch": 0.019311056474837187,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.6265,
+      "step": 917
+    },
+    {
+      "epoch": 0.019332115424100915,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0002,
+      "loss": 1.6249,
+      "step": 918
+    },
+    {
+      "epoch": 0.019353174373364642,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.6165,
+      "step": 919
+    },
+    {
+      "epoch": 0.019374233322628366,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0002,
+      "loss": 1.611,
+      "step": 920
+    },
+    {
+      "epoch": 0.019395292271892094,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0002,
+      "loss": 1.6338,
+      "step": 921
+    },
+    {
+      "epoch": 0.01941635122115582,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6137,
+      "step": 922
+    },
+    {
+      "epoch": 0.019437410170419545,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6039,
+      "step": 923
+    },
+    {
+      "epoch": 0.019458469119683273,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0002,
+      "loss": 1.6348,
+      "step": 924
+    },
+    {
+      "epoch": 0.019479528068947,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0002,
+      "loss": 1.6254,
+      "step": 925
+    },
+    {
+      "epoch": 0.019500587018210728,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.6149,
+      "step": 926
+    },
+    {
+      "epoch": 0.019521645967474452,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0002,
+      "loss": 1.6296,
+      "step": 927
+    },
+    {
+      "epoch": 0.01954270491673818,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.0002,
+      "loss": 1.6434,
+      "step": 928
+    },
+    {
+      "epoch": 0.019563763866001907,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0002,
+      "loss": 1.6116,
+      "step": 929
+    },
+    {
+      "epoch": 0.01958482281526563,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6404,
+      "step": 930
+    },
+    {
+      "epoch": 0.01960588176452936,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.659,
+      "step": 931
+    },
+    {
+      "epoch": 0.019626940713793086,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0002,
+      "loss": 1.6163,
+      "step": 932
+    },
+    {
+      "epoch": 0.01964799966305681,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0002,
+      "loss": 1.6564,
+      "step": 933
+    },
+    {
+      "epoch": 0.019669058612320538,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.6478,
+      "step": 934
+    },
+    {
+      "epoch": 0.019690117561584265,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.618,
+      "step": 935
+    },
+    {
+      "epoch": 0.019711176510847993,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.6503,
+      "step": 936
+    },
+    {
+      "epoch": 0.019732235460111717,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.6146,
+      "step": 937
+    },
+    {
+      "epoch": 0.019753294409375444,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.6315,
+      "step": 938
+    },
+    {
+      "epoch": 0.019774353358639172,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.6167,
+      "step": 939
+    },
+    {
+      "epoch": 0.019795412307902896,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5994,
+      "step": 940
+    },
+    {
+      "epoch": 0.019816471257166623,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0002,
+      "loss": 1.6217,
+      "step": 941
+    },
+    {
+      "epoch": 0.01983753020643035,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0002,
+      "loss": 1.6347,
+      "step": 942
+    },
+    {
+      "epoch": 0.019858589155694075,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.6399,
+      "step": 943
+    },
+    {
+      "epoch": 0.019879648104957803,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.6524,
+      "step": 944
+    },
+    {
+      "epoch": 0.01990070705422153,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0002,
+      "loss": 1.6498,
+      "step": 945
+    },
+    {
+      "epoch": 0.019921766003485258,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6563,
+      "step": 946
+    },
+    {
+      "epoch": 0.019942824952748982,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.6275,
+      "step": 947
+    },
+    {
+      "epoch": 0.01996388390201271,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.6546,
+      "step": 948
+    },
+    {
+      "epoch": 0.019984942851276437,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0002,
+      "loss": 1.6079,
+      "step": 949
+    },
+    {
+      "epoch": 0.02000600180054016,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.6183,
+      "step": 950
+    },
+    {
+      "epoch": 0.02002706074980389,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0002,
+      "loss": 1.642,
+      "step": 951
+    },
+    {
+      "epoch": 0.020048119699067616,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6106,
+      "step": 952
+    },
+    {
+      "epoch": 0.02006917864833134,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.6926,
+      "step": 953
+    },
+    {
+      "epoch": 0.020090237597595068,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.644,
+      "step": 954
+    },
+    {
+      "epoch": 0.020111296546858795,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6402,
+      "step": 955
+    },
+    {
+      "epoch": 0.020132355496122523,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0002,
+      "loss": 1.5973,
+      "step": 956
+    },
+    {
+      "epoch": 0.020153414445386247,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6506,
+      "step": 957
+    },
+    {
+      "epoch": 0.020174473394649974,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.6483,
+      "step": 958
+    },
+    {
+      "epoch": 0.020195532343913702,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.6261,
+      "step": 959
+    },
+    {
+      "epoch": 0.020216591293177426,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.6237,
+      "step": 960
+    },
+    {
+      "epoch": 0.020237650242441153,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.6382,
+      "step": 961
+    },
+    {
+      "epoch": 0.02025870919170488,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.6411,
+      "step": 962
+    },
+    {
+      "epoch": 0.020279768140968605,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.6045,
+      "step": 963
+    },
+    {
+      "epoch": 0.020300827090232332,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0002,
+      "loss": 1.6084,
+      "step": 964
+    },
+    {
+      "epoch": 0.02032188603949606,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6348,
+      "step": 965
+    },
+    {
+      "epoch": 0.020342944988759788,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0002,
+      "loss": 1.6363,
+      "step": 966
+    },
+    {
+      "epoch": 0.02036400393802351,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0002,
+      "loss": 1.6232,
+      "step": 967
+    },
+    {
+      "epoch": 0.02038506288728724,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0002,
+      "loss": 1.6178,
+      "step": 968
+    },
+    {
+      "epoch": 0.020406121836550967,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.6237,
+      "step": 969
+    },
+    {
+      "epoch": 0.02042718078581469,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.6666,
+      "step": 970
+    },
+    {
+      "epoch": 0.02044823973507842,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.614,
+      "step": 971
+    },
+    {
+      "epoch": 0.020469298684342146,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.6207,
+      "step": 972
+    },
+    {
+      "epoch": 0.02049035763360587,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.6401,
+      "step": 973
+    },
+    {
+      "epoch": 0.020511416582869597,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.6113,
+      "step": 974
+    },
+    {
+      "epoch": 0.020532475532133325,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.6171,
+      "step": 975
+    },
+    {
+      "epoch": 0.02055353448139705,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.6284,
+      "step": 976
+    },
+    {
+      "epoch": 0.020574593430660777,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0002,
+      "loss": 1.627,
+      "step": 977
+    },
+    {
+      "epoch": 0.020595652379924504,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0002,
+      "loss": 1.634,
+      "step": 978
+    },
+    {
+      "epoch": 0.02061671132918823,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.6277,
+      "step": 979
+    },
+    {
+      "epoch": 0.020637770278451956,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.6225,
+      "step": 980
+    },
+    {
+      "epoch": 0.020658829227715683,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.5944,
+      "step": 981
+    },
+    {
+      "epoch": 0.02067988817697941,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0002,
+      "loss": 1.6136,
+      "step": 982
+    },
+    {
+      "epoch": 0.020700947126243135,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0002,
+      "loss": 1.6257,
+      "step": 983
+    },
+    {
+      "epoch": 0.020722006075506862,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.6533,
+      "step": 984
+    },
+    {
+      "epoch": 0.02074306502477059,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0002,
+      "loss": 1.6294,
+      "step": 985
+    },
+    {
+      "epoch": 0.020764123974034314,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0002,
+      "loss": 1.6164,
+      "step": 986
+    },
+    {
+      "epoch": 0.02078518292329804,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.6284,
+      "step": 987
+    },
+    {
+      "epoch": 0.02080624187256177,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0002,
+      "loss": 1.6287,
+      "step": 988
+    },
+    {
+      "epoch": 0.020827300821825497,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0002,
+      "loss": 1.6626,
+      "step": 989
+    },
+    {
+      "epoch": 0.02084835977108922,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0002,
+      "loss": 1.6201,
+      "step": 990
+    },
+    {
+      "epoch": 0.020869418720352948,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.6195,
+      "step": 991
+    },
+    {
+      "epoch": 0.020890477669616676,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0002,
+      "loss": 1.6268,
+      "step": 992
+    },
+    {
+      "epoch": 0.0209115366188804,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.6077,
+      "step": 993
+    },
+    {
+      "epoch": 0.020932595568144127,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0002,
+      "loss": 1.6278,
+      "step": 994
+    },
+    {
+      "epoch": 0.020953654517407855,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.6191,
+      "step": 995
+    },
+    {
+      "epoch": 0.02097471346667158,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.612,
+      "step": 996
+    },
+    {
+      "epoch": 0.020995772415935306,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.6431,
+      "step": 997
+    },
+    {
+      "epoch": 0.021016831365199034,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0002,
+      "loss": 1.6593,
+      "step": 998
+    },
+    {
+      "epoch": 0.02103789031446276,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.5955,
+      "step": 999
+    },
+    {
+      "epoch": 0.021058949263726486,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.6325,
+      "step": 1000
+    },
+    {
+      "epoch": 0.021058949263726486,
+      "eval_loss": 1.6914067268371582,
+      "eval_runtime": 898.0675,
+      "eval_samples_per_second": 68.814,
+      "eval_steps_per_second": 2.151,
+      "step": 1000
+    },
+    {
+      "epoch": 0.021080008212990213,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6336,
+      "step": 1001
+    },
+    {
+      "epoch": 0.02110106716225394,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0002,
+      "loss": 1.6411,
+      "step": 1002
+    },
+    {
+      "epoch": 0.021122126111517665,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0002,
+      "loss": 1.6576,
+      "step": 1003
+    },
+    {
+      "epoch": 0.021143185060781392,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0002,
+      "loss": 1.6168,
+      "step": 1004
+    },
+    {
+      "epoch": 0.02116424401004512,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.6233,
+      "step": 1005
+    },
+    {
+      "epoch": 0.021185302959308844,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0002,
+      "loss": 1.6156,
+      "step": 1006
+    },
+    {
+      "epoch": 0.02120636190857257,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.6707,
+      "step": 1007
+    },
+    {
+      "epoch": 0.0212274208578363,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6217,
+      "step": 1008
+    },
+    {
+      "epoch": 0.021248479807100026,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.6352,
+      "step": 1009
+    },
+    {
+      "epoch": 0.02126953875636375,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.6402,
+      "step": 1010
+    },
+    {
+      "epoch": 0.021290597705627478,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0002,
+      "loss": 1.6176,
+      "step": 1011
+    },
+    {
+      "epoch": 0.021311656654891206,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.622,
+      "step": 1012
+    },
+    {
+      "epoch": 0.02133271560415493,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.6245,
+      "step": 1013
+    },
+    {
+      "epoch": 0.021353774553418657,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.6714,
+      "step": 1014
+    },
+    {
+      "epoch": 0.021374833502682385,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.6911,
+      "step": 1015
+    },
+    {
+      "epoch": 0.02139589245194611,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0002,
+      "loss": 1.6083,
+      "step": 1016
+    },
+    {
+      "epoch": 0.021416951401209836,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6309,
+      "step": 1017
+    },
+    {
+      "epoch": 0.021438010350473564,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.6159,
+      "step": 1018
+    },
+    {
+      "epoch": 0.02145906929973729,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.6106,
+      "step": 1019
+    },
+    {
+      "epoch": 0.021480128249001015,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0002,
+      "loss": 1.6292,
+      "step": 1020
+    },
+    {
+      "epoch": 0.021501187198264743,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0002,
+      "loss": 1.6053,
+      "step": 1021
+    },
+    {
+      "epoch": 0.02152224614752847,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.6189,
+      "step": 1022
+    },
+    {
+      "epoch": 0.021543305096792194,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0002,
+      "loss": 1.6064,
+      "step": 1023
+    },
+    {
+      "epoch": 0.021564364046055922,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.6111,
+      "step": 1024
+    },
+    {
+      "epoch": 0.02158542299531965,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0002,
+      "loss": 1.6346,
+      "step": 1025
+    },
+    {
+      "epoch": 0.021606481944583374,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0002,
+      "loss": 1.6228,
+      "step": 1026
+    },
+    {
+      "epoch": 0.0216275408938471,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6298,
+      "step": 1027
+    },
+    {
+      "epoch": 0.02164859984311083,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0002,
+      "loss": 1.6337,
+      "step": 1028
+    },
+    {
+      "epoch": 0.021669658792374553,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.0002,
+      "loss": 1.6006,
+      "step": 1029
+    },
+    {
+      "epoch": 0.02169071774163828,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0002,
+      "loss": 1.6476,
+      "step": 1030
+    },
+    {
+      "epoch": 0.021711776690902008,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0002,
+      "loss": 1.5851,
+      "step": 1031
+    },
+    {
+      "epoch": 0.021732835640165735,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.6288,
+      "step": 1032
+    },
+    {
+      "epoch": 0.02175389458942946,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.5907,
+      "step": 1033
+    },
+    {
+      "epoch": 0.021774953538693187,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5862,
+      "step": 1034
+    },
+    {
+      "epoch": 0.021796012487956914,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0002,
+      "loss": 1.6431,
+      "step": 1035
+    },
+    {
+      "epoch": 0.02181707143722064,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.6375,
+      "step": 1036
+    },
+    {
+      "epoch": 0.021838130386484366,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.6364,
+      "step": 1037
+    },
+    {
+      "epoch": 0.021859189335748094,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.6236,
+      "step": 1038
+    },
+    {
+      "epoch": 0.021880248285011818,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0002,
+      "loss": 1.6356,
+      "step": 1039
+    },
+    {
+      "epoch": 0.021901307234275545,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.5813,
+      "step": 1040
+    },
+    {
+      "epoch": 0.021922366183539273,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.6238,
+      "step": 1041
+    },
+    {
+      "epoch": 0.021943425132803,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.6377,
+      "step": 1042
+    },
+    {
+      "epoch": 0.021964484082066724,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.632,
+      "step": 1043
+    },
+    {
+      "epoch": 0.021985543031330452,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0002,
+      "loss": 1.6066,
+      "step": 1044
+    },
+    {
+      "epoch": 0.02200660198059418,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0002,
+      "loss": 1.6219,
+      "step": 1045
+    },
+    {
+      "epoch": 0.022027660929857903,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.6233,
+      "step": 1046
+    },
+    {
+      "epoch": 0.02204871987912163,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0002,
+      "loss": 1.6028,
+      "step": 1047
+    },
+    {
+      "epoch": 0.02206977882838536,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0002,
+      "loss": 1.62,
+      "step": 1048
+    },
+    {
+      "epoch": 0.022090837777649083,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.6118,
+      "step": 1049
+    },
+    {
+      "epoch": 0.02211189672691281,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.6278,
+      "step": 1050
+    },
+    {
+      "epoch": 0.022132955676176538,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0002,
+      "loss": 1.637,
+      "step": 1051
+    },
+    {
+      "epoch": 0.022154014625440265,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.6312,
+      "step": 1052
+    },
+    {
+      "epoch": 0.02217507357470399,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.61,
+      "step": 1053
+    },
+    {
+      "epoch": 0.022196132523967717,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.6094,
+      "step": 1054
+    },
+    {
+      "epoch": 0.022217191473231444,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0002,
+      "loss": 1.6337,
+      "step": 1055
+    },
+    {
+      "epoch": 0.02223825042249517,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0002,
+      "loss": 1.6495,
+      "step": 1056
+    },
+    {
+      "epoch": 0.022259309371758896,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6151,
+      "step": 1057
+    },
+    {
+      "epoch": 0.022280368321022623,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.6204,
+      "step": 1058
+    },
+    {
+      "epoch": 0.022301427270286348,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.5997,
+      "step": 1059
+    },
+    {
+      "epoch": 0.022322486219550075,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0002,
+      "loss": 1.6116,
+      "step": 1060
+    },
+    {
+      "epoch": 0.022343545168813803,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0002,
+      "loss": 1.6374,
+      "step": 1061
+    },
+    {
+      "epoch": 0.02236460411807753,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6257,
+      "step": 1062
+    },
+    {
+      "epoch": 0.022385663067341254,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6663,
+      "step": 1063
+    },
+    {
+      "epoch": 0.02240672201660498,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.6162,
+      "step": 1064
+    },
+    {
+      "epoch": 0.02242778096586871,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0002,
+      "loss": 1.6099,
+      "step": 1065
+    },
+    {
+      "epoch": 0.022448839915132433,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0002,
+      "loss": 1.6203,
+      "step": 1066
+    },
+    {
+      "epoch": 0.02246989886439616,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0002,
+      "loss": 1.662,
+      "step": 1067
+    },
+    {
+      "epoch": 0.02249095781365989,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0002,
+      "loss": 1.619,
+      "step": 1068
+    },
+    {
+      "epoch": 0.022512016762923612,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.6197,
+      "step": 1069
+    },
+    {
+      "epoch": 0.02253307571218734,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0002,
+      "loss": 1.6199,
+      "step": 1070
+    },
+    {
+      "epoch": 0.022554134661451068,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0002,
+      "loss": 1.6183,
+      "step": 1071
+    },
+    {
+      "epoch": 0.022575193610714795,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6073,
+      "step": 1072
+    },
+    {
+      "epoch": 0.02259625255997852,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6481,
+      "step": 1073
+    },
+    {
+      "epoch": 0.022617311509242247,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.6485,
+      "step": 1074
+    },
+    {
+      "epoch": 0.022638370458505974,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5991,
+      "step": 1075
+    },
+    {
+      "epoch": 0.022659429407769698,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6267,
+      "step": 1076
+    },
+    {
+      "epoch": 0.022680488357033426,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.6284,
+      "step": 1077
+    },
+    {
+      "epoch": 0.022701547306297153,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.6177,
+      "step": 1078
+    },
+    {
+      "epoch": 0.022722606255560877,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.6306,
+      "step": 1079
+    },
+    {
+      "epoch": 0.022743665204824605,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.5567,
+      "step": 1080
+    },
+    {
+      "epoch": 0.022764724154088332,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.6132,
+      "step": 1081
+    },
+    {
+      "epoch": 0.02278578310335206,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6195,
+      "step": 1082
+    },
+    {
+      "epoch": 0.022806842052615784,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.6194,
+      "step": 1083
+    },
+    {
+      "epoch": 0.02282790100187951,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0002,
+      "loss": 1.6262,
+      "step": 1084
+    },
+    {
+      "epoch": 0.02284895995114324,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.6283,
+      "step": 1085
+    },
+    {
+      "epoch": 0.022870018900406963,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002,
+      "loss": 1.6024,
+      "step": 1086
+    },
+    {
+      "epoch": 0.02289107784967069,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.6494,
+      "step": 1087
+    },
+    {
+      "epoch": 0.022912136798934418,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.619,
+      "step": 1088
+    },
+    {
+      "epoch": 0.022933195748198142,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0002,
+      "loss": 1.625,
+      "step": 1089
+    },
+    {
+      "epoch": 0.02295425469746187,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.5949,
+      "step": 1090
+    },
+    {
+      "epoch": 0.022975313646725597,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0002,
+      "loss": 1.6353,
+      "step": 1091
+    },
+    {
+      "epoch": 0.02299637259598932,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0002,
+      "loss": 1.6138,
+      "step": 1092
+    },
+    {
+      "epoch": 0.02301743154525305,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0002,
+      "loss": 1.6221,
+      "step": 1093
+    },
+    {
+      "epoch": 0.023038490494516776,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0002,
+      "loss": 1.6223,
+      "step": 1094
+    },
+    {
+      "epoch": 0.023059549443780504,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0002,
+      "loss": 1.6076,
+      "step": 1095
+    },
+    {
+      "epoch": 0.023080608393044228,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.6176,
+      "step": 1096
+    },
+    {
+      "epoch": 0.023101667342307956,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0002,
+      "loss": 1.6344,
+      "step": 1097
+    },
+    {
+      "epoch": 0.023122726291571683,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0002,
+      "loss": 1.6199,
+      "step": 1098
+    },
+    {
+      "epoch": 0.023143785240835407,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002,
+      "loss": 1.5975,
+      "step": 1099
+    },
+    {
+      "epoch": 0.023164844190099135,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.5918,
+      "step": 1100
+    },
+    {
+      "epoch": 0.023185903139362862,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.6169,
+      "step": 1101
+    },
+    {
+      "epoch": 0.023206962088626586,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.6261,
+      "step": 1102
+    },
+    {
+      "epoch": 0.023228021037890314,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.623,
+      "step": 1103
+    },
+    {
+      "epoch": 0.02324907998715404,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.6211,
+      "step": 1104
+    },
+    {
+      "epoch": 0.02327013893641777,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.6389,
+      "step": 1105
+    },
+    {
+      "epoch": 0.023291197885681493,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0002,
+      "loss": 1.5897,
+      "step": 1106
+    },
+    {
+      "epoch": 0.02331225683494522,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.6222,
+      "step": 1107
+    },
+    {
+      "epoch": 0.023333315784208948,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.6309,
+      "step": 1108
+    },
+    {
+      "epoch": 0.023354374733472672,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0002,
+      "loss": 1.617,
+      "step": 1109
+    },
+    {
+      "epoch": 0.0233754336827364,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.6224,
+      "step": 1110
+    },
+    {
+      "epoch": 0.023396492632000127,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0002,
+      "loss": 1.6505,
+      "step": 1111
+    },
+    {
+      "epoch": 0.02341755158126385,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6487,
+      "step": 1112
+    },
+    {
+      "epoch": 0.02343861053052758,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0002,
+      "loss": 1.6249,
+      "step": 1113
+    },
+    {
+      "epoch": 0.023459669479791306,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0002,
+      "loss": 1.6512,
+      "step": 1114
+    },
+    {
+      "epoch": 0.023480728429055034,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.6008,
+      "step": 1115
+    },
+    {
+      "epoch": 0.023501787378318758,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.6267,
+      "step": 1116
+    },
+    {
+      "epoch": 0.023522846327582485,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.6223,
+      "step": 1117
+    },
+    {
+      "epoch": 0.023543905276846213,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.6225,
+      "step": 1118
+    },
+    {
+      "epoch": 0.023564964226109937,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0002,
+      "loss": 1.6139,
+      "step": 1119
+    },
+    {
+      "epoch": 0.023586023175373665,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.6082,
+      "step": 1120
+    },
+    {
+      "epoch": 0.023607082124637392,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0002,
+      "loss": 1.6414,
+      "step": 1121
+    },
+    {
+      "epoch": 0.023628141073901116,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0002,
+      "loss": 1.6348,
+      "step": 1122
+    },
+    {
+      "epoch": 0.023649200023164844,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.6217,
+      "step": 1123
+    },
+    {
+      "epoch": 0.02367025897242857,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.5911,
+      "step": 1124
+    },
+    {
+      "epoch": 0.0236913179216923,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0002,
+      "loss": 1.6044,
+      "step": 1125
+    },
+    {
+      "epoch": 0.023712376870956023,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.6305,
+      "step": 1126
+    },
+    {
+      "epoch": 0.02373343582021975,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.6247,
+      "step": 1127
+    },
+    {
+      "epoch": 0.023754494769483478,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.5967,
+      "step": 1128
+    },
+    {
+      "epoch": 0.023775553718747202,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.6199,
+      "step": 1129
+    },
+    {
+      "epoch": 0.02379661266801093,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002,
+      "loss": 1.6135,
+      "step": 1130
+    },
+    {
+      "epoch": 0.023817671617274657,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6403,
+      "step": 1131
+    },
+    {
+      "epoch": 0.02383873056653838,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.624,
+      "step": 1132
+    },
+    {
+      "epoch": 0.02385978951580211,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0002,
+      "loss": 1.6105,
+      "step": 1133
+    },
+    {
+      "epoch": 0.023880848465065836,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.6437,
+      "step": 1134
+    },
+    {
+      "epoch": 0.023901907414329564,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0002,
+      "loss": 1.5916,
+      "step": 1135
+    },
+    {
+      "epoch": 0.023922966363593288,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.6119,
+      "step": 1136
+    },
+    {
+      "epoch": 0.023944025312857015,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0002,
+      "loss": 1.6003,
+      "step": 1137
+    },
+    {
+      "epoch": 0.023965084262120743,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0002,
+      "loss": 1.6303,
+      "step": 1138
+    },
+    {
+      "epoch": 0.023986143211384467,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0002,
+      "loss": 1.6294,
+      "step": 1139
+    },
+    {
+      "epoch": 0.024007202160648194,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0002,
+      "loss": 1.6217,
+      "step": 1140
+    },
+    {
+      "epoch": 0.024028261109911922,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0002,
+      "loss": 1.6383,
+      "step": 1141
+    },
+    {
+      "epoch": 0.024049320059175646,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.6308,
+      "step": 1142
+    },
+    {
+      "epoch": 0.024070379008439374,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0002,
+      "loss": 1.6204,
+      "step": 1143
+    },
+    {
+      "epoch": 0.0240914379577031,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6054,
+      "step": 1144
+    },
+    {
+      "epoch": 0.024112496906966825,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.6421,
+      "step": 1145
+    },
+    {
+      "epoch": 0.024133555856230553,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.6038,
+      "step": 1146
+    },
+    {
+      "epoch": 0.02415461480549428,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6592,
+      "step": 1147
+    },
+    {
+      "epoch": 0.024175673754758008,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.6276,
+      "step": 1148
+    },
+    {
+      "epoch": 0.024196732704021732,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0002,
+      "loss": 1.5886,
+      "step": 1149
+    },
+    {
+      "epoch": 0.02421779165328546,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0002,
+      "loss": 1.6242,
+      "step": 1150
+    },
+    {
+      "epoch": 0.024238850602549187,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0002,
+      "loss": 1.638,
+      "step": 1151
+    },
+    {
+      "epoch": 0.02425990955181291,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.599,
+      "step": 1152
+    },
+    {
+      "epoch": 0.02428096850107664,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0002,
+      "loss": 1.605,
+      "step": 1153
+    },
+    {
+      "epoch": 0.024302027450340366,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.5756,
+      "step": 1154
+    },
+    {
+      "epoch": 0.02432308639960409,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0002,
+      "loss": 1.633,
+      "step": 1155
+    },
+    {
+      "epoch": 0.024344145348867818,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.6105,
+      "step": 1156
+    },
+    {
+      "epoch": 0.024365204298131545,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.6181,
+      "step": 1157
+    },
+    {
+      "epoch": 0.024386263247395273,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.587,
+      "step": 1158
+    },
+    {
+      "epoch": 0.024407322196658997,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.5906,
+      "step": 1159
+    },
+    {
+      "epoch": 0.024428381145922724,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.6298,
+      "step": 1160
+    },
+    {
+      "epoch": 0.024449440095186452,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.5965,
+      "step": 1161
+    },
+    {
+      "epoch": 0.024470499044450176,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.608,
+      "step": 1162
+    },
+    {
+      "epoch": 0.024491557993713903,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0002,
+      "loss": 1.6367,
+      "step": 1163
+    },
+    {
+      "epoch": 0.02451261694297763,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0002,
+      "loss": 1.6158,
+      "step": 1164
+    },
+    {
+      "epoch": 0.024533675892241355,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.6029,
+      "step": 1165
+    },
+    {
+      "epoch": 0.024554734841505083,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6462,
+      "step": 1166
+    },
+    {
+      "epoch": 0.02457579379076881,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0002,
+      "loss": 1.6016,
+      "step": 1167
+    },
+    {
+      "epoch": 0.024596852740032538,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002,
+      "loss": 1.6033,
+      "step": 1168
+    },
+    {
+      "epoch": 0.02461791168929626,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0002,
+      "loss": 1.6193,
+      "step": 1169
+    },
+    {
+      "epoch": 0.02463897063855999,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6367,
+      "step": 1170
+    },
+    {
+      "epoch": 0.024660029587823717,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0002,
+      "loss": 1.6041,
+      "step": 1171
+    },
+    {
+      "epoch": 0.02468108853708744,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0002,
+      "loss": 1.6419,
+      "step": 1172
+    },
+    {
+      "epoch": 0.02470214748635117,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.5985,
+      "step": 1173
+    },
+    {
+      "epoch": 0.024723206435614896,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6452,
+      "step": 1174
+    },
+    {
+      "epoch": 0.02474426538487862,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0002,
+      "loss": 1.6152,
+      "step": 1175
+    },
+    {
+      "epoch": 0.024765324334142347,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5864,
+      "step": 1176
+    },
+    {
+      "epoch": 0.024786383283406075,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0002,
+      "loss": 1.6626,
+      "step": 1177
+    },
+    {
+      "epoch": 0.024807442232669803,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0002,
+      "loss": 1.6221,
+      "step": 1178
+    },
+    {
+      "epoch": 0.024828501181933527,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.636,
+      "step": 1179
+    },
+    {
+      "epoch": 0.024849560131197254,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0002,
+      "loss": 1.6143,
+      "step": 1180
+    },
+    {
+      "epoch": 0.02487061908046098,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0002,
+      "loss": 1.5912,
+      "step": 1181
+    },
+    {
+      "epoch": 0.024891678029724706,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.6055,
+      "step": 1182
+    },
+    {
+      "epoch": 0.024912736978988433,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0002,
+      "loss": 1.6034,
+      "step": 1183
+    },
+    {
+      "epoch": 0.02493379592825216,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.6055,
+      "step": 1184
+    },
+    {
+      "epoch": 0.024954854877515885,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0002,
+      "loss": 1.624,
+      "step": 1185
+    },
+    {
+      "epoch": 0.024975913826779612,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0002,
+      "loss": 1.6475,
+      "step": 1186
+    },
+    {
+      "epoch": 0.02499697277604334,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0002,
+      "loss": 1.6437,
+      "step": 1187
+    },
+    {
+      "epoch": 0.025018031725307067,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0002,
+      "loss": 1.6337,
+      "step": 1188
+    },
+    {
+      "epoch": 0.02503909067457079,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.6257,
+      "step": 1189
+    },
+    {
+      "epoch": 0.02506014962383452,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0002,
+      "loss": 1.6502,
+      "step": 1190
+    },
+    {
+      "epoch": 0.025081208573098247,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0002,
+      "loss": 1.597,
+      "step": 1191
+    },
+    {
+      "epoch": 0.02510226752236197,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0002,
+      "loss": 1.6389,
+      "step": 1192
+    },
+    {
+      "epoch": 0.025123326471625698,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002,
+      "loss": 1.6071,
+      "step": 1193
+    },
+    {
+      "epoch": 0.025144385420889426,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0002,
+      "loss": 1.6186,
+      "step": 1194
+    },
+    {
+      "epoch": 0.02516544437015315,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.568,
+      "step": 1195
+    },
+    {
+      "epoch": 0.025186503319416877,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.6266,
+      "step": 1196
+    },
+    {
+      "epoch": 0.025207562268680605,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.635,
+      "step": 1197
+    },
+    {
+      "epoch": 0.02522862121794433,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0002,
+      "loss": 1.6003,
+      "step": 1198
+    },
+    {
+      "epoch": 0.025249680167208056,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6364,
+      "step": 1199
+    },
+    {
+      "epoch": 0.025270739116471784,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.5939,
+      "step": 1200
+    },
+    {
+      "epoch": 0.02529179806573551,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.6229,
+      "step": 1201
+    },
+    {
+      "epoch": 0.025312857014999236,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.583,
+      "step": 1202
+    },
+    {
+      "epoch": 0.025333915964262963,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.642,
+      "step": 1203
+    },
+    {
+      "epoch": 0.02535497491352669,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0002,
+      "loss": 1.581,
+      "step": 1204
+    },
+    {
+      "epoch": 0.025376033862790415,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0002,
+      "loss": 1.594,
+      "step": 1205
+    },
+    {
+      "epoch": 0.025397092812054142,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.6383,
+      "step": 1206
+    },
+    {
+      "epoch": 0.02541815176131787,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.5929,
+      "step": 1207
+    },
+    {
+      "epoch": 0.025439210710581594,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0002,
+      "loss": 1.636,
+      "step": 1208
+    },
+    {
+      "epoch": 0.02546026965984532,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.6064,
+      "step": 1209
+    },
+    {
+      "epoch": 0.02548132860910905,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.619,
+      "step": 1210
+    },
+    {
+      "epoch": 0.025502387558372776,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.6399,
+      "step": 1211
+    },
+    {
+      "epoch": 0.0255234465076365,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0002,
+      "loss": 1.6246,
+      "step": 1212
+    },
+    {
+      "epoch": 0.025544505456900228,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.6223,
+      "step": 1213
+    },
+    {
+      "epoch": 0.025565564406163956,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.6077,
+      "step": 1214
+    },
+    {
+      "epoch": 0.02558662335542768,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.6345,
+      "step": 1215
+    },
+    {
+      "epoch": 0.025607682304691407,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.6388,
+      "step": 1216
+    },
+    {
+      "epoch": 0.025628741253955135,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0002,
+      "loss": 1.5735,
+      "step": 1217
+    },
+    {
+      "epoch": 0.02564980020321886,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.6032,
+      "step": 1218
+    },
+    {
+      "epoch": 0.025670859152482586,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.6221,
+      "step": 1219
+    },
+    {
+      "epoch": 0.025691918101746314,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.6351,
+      "step": 1220
+    },
+    {
+      "epoch": 0.02571297705101004,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.6198,
+      "step": 1221
+    },
+    {
+      "epoch": 0.025734036000273765,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.6674,
+      "step": 1222
+    },
+    {
+      "epoch": 0.025755094949537493,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0002,
+      "loss": 1.6306,
+      "step": 1223
+    },
+    {
+      "epoch": 0.02577615389880122,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.5934,
+      "step": 1224
+    },
+    {
+      "epoch": 0.025797212848064945,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.6132,
+      "step": 1225
+    },
+    {
+      "epoch": 0.025818271797328672,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0002,
+      "loss": 1.5858,
+      "step": 1226
+    },
+    {
+      "epoch": 0.0258393307465924,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0002,
+      "loss": 1.597,
+      "step": 1227
+    },
+    {
+      "epoch": 0.025860389695856124,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.5878,
+      "step": 1228
+    },
+    {
+      "epoch": 0.02588144864511985,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.6001,
+      "step": 1229
+    },
+    {
+      "epoch": 0.02590250759438358,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0002,
+      "loss": 1.6302,
+      "step": 1230
+    },
+    {
+      "epoch": 0.025923566543647306,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0002,
+      "loss": 1.6049,
+      "step": 1231
+    },
+    {
+      "epoch": 0.02594462549291103,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0002,
+      "loss": 1.6234,
+      "step": 1232
+    },
+    {
+      "epoch": 0.025965684442174758,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.5932,
+      "step": 1233
+    },
+    {
+      "epoch": 0.025986743391438485,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0002,
+      "loss": 1.6165,
+      "step": 1234
+    },
+    {
+      "epoch": 0.02600780234070221,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0002,
+      "loss": 1.5812,
+      "step": 1235
+    },
+    {
+      "epoch": 0.026028861289965937,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.5991,
+      "step": 1236
+    },
+    {
+      "epoch": 0.026049920239229665,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.5866,
+      "step": 1237
+    },
+    {
+      "epoch": 0.02607097918849339,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0002,
+      "loss": 1.6271,
+      "step": 1238
+    },
+    {
+      "epoch": 0.026092038137757116,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6186,
+      "step": 1239
+    },
+    {
+      "epoch": 0.026113097087020844,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0002,
+      "loss": 1.5955,
+      "step": 1240
+    },
+    {
+      "epoch": 0.02613415603628457,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6197,
+      "step": 1241
+    },
+    {
+      "epoch": 0.026155214985548295,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.6009,
+      "step": 1242
+    },
+    {
+      "epoch": 0.026176273934812023,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0002,
+      "loss": 1.6218,
+      "step": 1243
+    },
+    {
+      "epoch": 0.02619733288407575,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.5983,
+      "step": 1244
+    },
+    {
+      "epoch": 0.026218391833339474,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0002,
+      "loss": 1.5949,
+      "step": 1245
+    },
+    {
+      "epoch": 0.026239450782603202,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0002,
+      "loss": 1.592,
+      "step": 1246
+    },
+    {
+      "epoch": 0.02626050973186693,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.642,
+      "step": 1247
+    },
+    {
+      "epoch": 0.026281568681130654,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.5977,
+      "step": 1248
+    },
+    {
+      "epoch": 0.02630262763039438,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0002,
+      "loss": 1.6097,
+      "step": 1249
+    },
+    {
+      "epoch": 0.02632368657965811,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0002,
+      "loss": 1.5939,
+      "step": 1250
+    },
+    {
+      "epoch": 0.026344745528921836,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0002,
+      "loss": 1.6219,
+      "step": 1251
+    },
+    {
+      "epoch": 0.02636580447818556,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.6273,
+      "step": 1252
+    },
+    {
+      "epoch": 0.026386863427449288,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.6068,
+      "step": 1253
+    },
+    {
+      "epoch": 0.026407922376713015,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.5923,
+      "step": 1254
+    },
+    {
+      "epoch": 0.02642898132597674,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0002,
+      "loss": 1.635,
+      "step": 1255
+    },
+    {
+      "epoch": 0.026450040275240467,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.0002,
+      "loss": 1.62,
+      "step": 1256
+    },
+    {
+      "epoch": 0.026471099224504194,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5957,
+      "step": 1257
+    },
+    {
+      "epoch": 0.02649215817376792,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0002,
+      "loss": 1.6102,
+      "step": 1258
+    },
+    {
+      "epoch": 0.026513217123031646,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0002,
+      "loss": 1.5821,
+      "step": 1259
+    },
+    {
+      "epoch": 0.026534276072295374,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0002,
+      "loss": 1.605,
+      "step": 1260
+    },
+    {
+      "epoch": 0.026555335021559098,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.6108,
+      "step": 1261
+    },
+    {
+      "epoch": 0.026576393970822825,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0002,
+      "loss": 1.6175,
+      "step": 1262
+    },
+    {
+      "epoch": 0.026597452920086553,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.6737,
+      "step": 1263
+    },
+    {
+      "epoch": 0.02661851186935028,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0002,
+      "loss": 1.6139,
+      "step": 1264
+    },
+    {
+      "epoch": 0.026639570818614004,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.6389,
+      "step": 1265
+    },
+    {
+      "epoch": 0.026660629767877732,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.6079,
+      "step": 1266
+    },
+    {
+      "epoch": 0.02668168871714146,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.5814,
+      "step": 1267
+    },
+    {
+      "epoch": 0.026702747666405183,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.622,
+      "step": 1268
+    },
+    {
+      "epoch": 0.02672380661566891,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0002,
+      "loss": 1.6231,
+      "step": 1269
+    },
+    {
+      "epoch": 0.02674486556493264,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0002,
+      "loss": 1.6189,
+      "step": 1270
+    },
+    {
+      "epoch": 0.026765924514196363,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.6342,
+      "step": 1271
+    },
+    {
+      "epoch": 0.02678698346346009,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.6329,
+      "step": 1272
+    },
+    {
+      "epoch": 0.026808042412723818,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0002,
+      "loss": 1.6079,
+      "step": 1273
+    },
+    {
+      "epoch": 0.026829101361987545,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.6079,
+      "step": 1274
+    },
+    {
+      "epoch": 0.02685016031125127,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.6479,
+      "step": 1275
+    },
+    {
+      "epoch": 0.026871219260514997,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0002,
+      "loss": 1.6052,
+      "step": 1276
+    },
+    {
+      "epoch": 0.026892278209778724,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0002,
+      "loss": 1.6283,
+      "step": 1277
+    },
+    {
+      "epoch": 0.02691333715904245,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.5903,
+      "step": 1278
+    },
+    {
+      "epoch": 0.026934396108306176,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.6379,
+      "step": 1279
+    },
+    {
+      "epoch": 0.026955455057569903,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.6033,
+      "step": 1280
+    },
+    {
+      "epoch": 0.026976514006833627,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.5896,
+      "step": 1281
+    },
+    {
+      "epoch": 0.026997572956097355,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.5876,
+      "step": 1282
+    },
+    {
+      "epoch": 0.027018631905361083,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0002,
+      "loss": 1.6024,
+      "step": 1283
+    },
+    {
+      "epoch": 0.02703969085462481,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6562,
+      "step": 1284
+    },
+    {
+      "epoch": 0.027060749803888534,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.5927,
+      "step": 1285
+    },
+    {
+      "epoch": 0.02708180875315226,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6031,
+      "step": 1286
+    },
+    {
+      "epoch": 0.02710286770241599,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.5868,
+      "step": 1287
+    },
+    {
+      "epoch": 0.027123926651679713,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6199,
+      "step": 1288
+    },
+    {
+      "epoch": 0.02714498560094344,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0002,
+      "loss": 1.5953,
+      "step": 1289
+    },
+    {
+      "epoch": 0.02716604455020717,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.5842,
+      "step": 1290
+    },
+    {
+      "epoch": 0.027187103499470892,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.6007,
+      "step": 1291
+    },
+    {
+      "epoch": 0.02720816244873462,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.624,
+      "step": 1292
+    },
+    {
+      "epoch": 0.027229221397998347,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.6411,
+      "step": 1293
+    },
+    {
+      "epoch": 0.027250280347262075,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0002,
+      "loss": 1.634,
+      "step": 1294
+    },
+    {
+      "epoch": 0.0272713392965258,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.0002,
+      "loss": 1.6284,
+      "step": 1295
+    },
+    {
+      "epoch": 0.027292398245789527,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0002,
+      "loss": 1.5982,
+      "step": 1296
+    },
+    {
+      "epoch": 0.027313457195053254,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.6197,
+      "step": 1297
+    },
+    {
+      "epoch": 0.027334516144316978,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.6282,
+      "step": 1298
+    },
+    {
+      "epoch": 0.027355575093580706,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002,
+      "loss": 1.5901,
+      "step": 1299
+    },
+    {
+      "epoch": 0.027376634042844433,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.6165,
+      "step": 1300
+    },
+    {
+      "epoch": 0.027397692992108157,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.0002,
+      "loss": 1.6153,
+      "step": 1301
+    },
+    {
+      "epoch": 0.027418751941371885,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.5959,
+      "step": 1302
+    },
+    {
+      "epoch": 0.027439810890635612,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6139,
+      "step": 1303
+    },
+    {
+      "epoch": 0.02746086983989934,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6059,
+      "step": 1304
+    },
+    {
+      "epoch": 0.027481928789163064,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0002,
+      "loss": 1.5903,
+      "step": 1305
+    },
+    {
+      "epoch": 0.02750298773842679,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.623,
+      "step": 1306
+    },
+    {
+      "epoch": 0.02752404668769052,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0002,
+      "loss": 1.6088,
+      "step": 1307
+    },
+    {
+      "epoch": 0.027545105636954243,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0002,
+      "loss": 1.6319,
+      "step": 1308
+    },
+    {
+      "epoch": 0.02756616458621797,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.6179,
+      "step": 1309
+    },
+    {
+      "epoch": 0.027587223535481698,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.618,
+      "step": 1310
+    },
+    {
+      "epoch": 0.027608282484745422,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.6211,
+      "step": 1311
+    },
+    {
+      "epoch": 0.02762934143400915,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.6086,
+      "step": 1312
+    },
+    {
+      "epoch": 0.027650400383272877,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.6581,
+      "step": 1313
+    },
+    {
+      "epoch": 0.0276714593325366,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.578,
+      "step": 1314
+    },
+    {
+      "epoch": 0.02769251828180033,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0002,
+      "loss": 1.6041,
+      "step": 1315
+    },
+    {
+      "epoch": 0.027713577231064056,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.5705,
+      "step": 1316
+    },
+    {
+      "epoch": 0.027734636180327784,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.6264,
+      "step": 1317
+    },
+    {
+      "epoch": 0.027755695129591508,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0002,
+      "loss": 1.6015,
+      "step": 1318
+    },
+    {
+      "epoch": 0.027776754078855236,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0002,
+      "loss": 1.642,
+      "step": 1319
+    },
+    {
+      "epoch": 0.027797813028118963,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.6443,
+      "step": 1320
+    },
+    {
+      "epoch": 0.027818871977382687,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0002,
+      "loss": 1.6177,
+      "step": 1321
+    },
+    {
+      "epoch": 0.027839930926646415,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.6113,
+      "step": 1322
+    },
+    {
+      "epoch": 0.027860989875910142,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0002,
+      "loss": 1.5953,
+      "step": 1323
+    },
+    {
+      "epoch": 0.027882048825173866,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0002,
+      "loss": 1.613,
+      "step": 1324
+    },
+    {
+      "epoch": 0.027903107774437594,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0002,
+      "loss": 1.6342,
+      "step": 1325
+    },
+    {
+      "epoch": 0.02792416672370132,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0002,
+      "loss": 1.6219,
+      "step": 1326
+    },
+    {
+      "epoch": 0.02794522567296505,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6043,
+      "step": 1327
+    },
+    {
+      "epoch": 0.027966284622228773,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0002,
+      "loss": 1.6305,
+      "step": 1328
+    },
+    {
+      "epoch": 0.0279873435714925,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.6052,
+      "step": 1329
+    },
+    {
+      "epoch": 0.028008402520756228,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.6224,
+      "step": 1330
+    },
+    {
+      "epoch": 0.028029461470019952,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0002,
+      "loss": 1.6226,
+      "step": 1331
+    },
+    {
+      "epoch": 0.02805052041928368,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0002,
+      "loss": 1.6226,
+      "step": 1332
+    },
+    {
+      "epoch": 0.028071579368547407,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0002,
+      "loss": 1.6256,
+      "step": 1333
+    },
+    {
+      "epoch": 0.02809263831781113,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.0002,
+      "loss": 1.6391,
+      "step": 1334
+    },
+    {
+      "epoch": 0.02811369726707486,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0002,
+      "loss": 1.6069,
+      "step": 1335
+    },
+    {
+      "epoch": 0.028134756216338586,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6228,
+      "step": 1336
+    },
+    {
+      "epoch": 0.028155815165602314,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.5944,
+      "step": 1337
+    },
+    {
+      "epoch": 0.028176874114866038,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.6236,
+      "step": 1338
+    },
+    {
+      "epoch": 0.028197933064129765,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5985,
+      "step": 1339
+    },
+    {
+      "epoch": 0.028218992013393493,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.6352,
+      "step": 1340
+    },
+    {
+      "epoch": 0.028240050962657217,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.6386,
+      "step": 1341
+    },
+    {
+      "epoch": 0.028261109911920945,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0002,
+      "loss": 1.6032,
+      "step": 1342
+    },
+    {
+      "epoch": 0.028282168861184672,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0002,
+      "loss": 1.6072,
+      "step": 1343
+    },
+    {
+      "epoch": 0.028303227810448396,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0002,
+      "loss": 1.632,
+      "step": 1344
+    },
+    {
+      "epoch": 0.028324286759712124,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.6416,
+      "step": 1345
+    },
+    {
+      "epoch": 0.02834534570897585,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.623,
+      "step": 1346
+    },
+    {
+      "epoch": 0.02836640465823958,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0002,
+      "loss": 1.6257,
+      "step": 1347
+    },
+    {
+      "epoch": 0.028387463607503303,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5952,
+      "step": 1348
+    },
+    {
+      "epoch": 0.02840852255676703,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.6117,
+      "step": 1349
+    },
+    {
+      "epoch": 0.028429581506030758,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0002,
+      "loss": 1.6349,
+      "step": 1350
+    },
+    {
+      "epoch": 0.028450640455294482,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0002,
+      "loss": 1.6076,
+      "step": 1351
+    },
+    {
+      "epoch": 0.02847169940455821,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0002,
+      "loss": 1.627,
+      "step": 1352
+    },
+    {
+      "epoch": 0.028492758353821937,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.6404,
+      "step": 1353
+    },
+    {
+      "epoch": 0.02851381730308566,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.5884,
+      "step": 1354
+    },
+    {
+      "epoch": 0.02853487625234939,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.5958,
+      "step": 1355
+    },
+    {
+      "epoch": 0.028555935201613116,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6203,
+      "step": 1356
+    },
+    {
+      "epoch": 0.028576994150876844,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.5905,
+      "step": 1357
+    },
+    {
+      "epoch": 0.028598053100140568,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0002,
+      "loss": 1.6162,
+      "step": 1358
+    },
+    {
+      "epoch": 0.028619112049404295,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002,
+      "loss": 1.6322,
+      "step": 1359
+    },
+    {
+      "epoch": 0.028640170998668023,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.626,
+      "step": 1360
+    },
+    {
+      "epoch": 0.028661229947931747,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.594,
+      "step": 1361
+    },
+    {
+      "epoch": 0.028682288897195474,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.565,
+      "step": 1362
+    },
+    {
+      "epoch": 0.028703347846459202,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6193,
+      "step": 1363
+    },
+    {
+      "epoch": 0.028724406795722926,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.622,
+      "step": 1364
+    },
+    {
+      "epoch": 0.028745465744986654,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.5932,
+      "step": 1365
+    },
+    {
+      "epoch": 0.02876652469425038,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.6126,
+      "step": 1366
+    },
+    {
+      "epoch": 0.02878758364351411,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0002,
+      "loss": 1.6159,
+      "step": 1367
+    },
+    {
+      "epoch": 0.028808642592777833,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.6521,
+      "step": 1368
+    },
+    {
+      "epoch": 0.02882970154204156,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0002,
+      "loss": 1.6255,
+      "step": 1369
+    },
+    {
+      "epoch": 0.028850760491305288,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0002,
+      "loss": 1.5906,
+      "step": 1370
+    },
+    {
+      "epoch": 0.028871819440569012,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0002,
+      "loss": 1.6229,
+      "step": 1371
+    },
+    {
+      "epoch": 0.02889287838983274,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.6187,
+      "step": 1372
+    },
+    {
+      "epoch": 0.028913937339096467,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.0002,
+      "loss": 1.5872,
+      "step": 1373
+    },
+    {
+      "epoch": 0.02893499628836019,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0002,
+      "loss": 1.6163,
+      "step": 1374
+    },
+    {
+      "epoch": 0.02895605523762392,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.6029,
+      "step": 1375
+    },
+    {
+      "epoch": 0.028977114186887646,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0002,
+      "loss": 1.6159,
+      "step": 1376
+    },
+    {
+      "epoch": 0.02899817313615137,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.6296,
+      "step": 1377
+    },
+    {
+      "epoch": 0.029019232085415098,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.6194,
+      "step": 1378
+    },
+    {
+      "epoch": 0.029040291034678825,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.6062,
+      "step": 1379
+    },
+    {
+      "epoch": 0.029061349983942553,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0002,
+      "loss": 1.6184,
+      "step": 1380
+    },
+    {
+      "epoch": 0.029082408933206277,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0002,
+      "loss": 1.6045,
+      "step": 1381
+    },
+    {
+      "epoch": 0.029103467882470004,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.6155,
+      "step": 1382
+    },
+    {
+      "epoch": 0.029124526831733732,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0002,
+      "loss": 1.615,
+      "step": 1383
+    },
+    {
+      "epoch": 0.029145585780997456,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0002,
+      "loss": 1.618,
+      "step": 1384
+    },
+    {
+      "epoch": 0.029166644730261183,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.5576,
+      "step": 1385
+    },
+    {
+      "epoch": 0.02918770367952491,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.6689,
+      "step": 1386
+    },
+    {
+      "epoch": 0.029208762628788635,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6087,
+      "step": 1387
+    },
+    {
+      "epoch": 0.029229821578052363,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.6104,
+      "step": 1388
+    },
+    {
+      "epoch": 0.02925088052731609,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0002,
+      "loss": 1.6187,
+      "step": 1389
+    },
+    {
+      "epoch": 0.029271939476579818,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0002,
+      "loss": 1.5843,
+      "step": 1390
+    },
+    {
+      "epoch": 0.02929299842584354,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.6471,
+      "step": 1391
+    },
+    {
+      "epoch": 0.02931405737510727,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0002,
+      "loss": 1.616,
+      "step": 1392
+    },
+    {
+      "epoch": 0.029335116324370997,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.6148,
+      "step": 1393
+    },
+    {
+      "epoch": 0.02935617527363472,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.6599,
+      "step": 1394
+    },
+    {
+      "epoch": 0.02937723422289845,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0002,
+      "loss": 1.6011,
+      "step": 1395
+    },
+    {
+      "epoch": 0.029398293172162176,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0002,
+      "loss": 1.6202,
+      "step": 1396
+    },
+    {
+      "epoch": 0.0294193521214259,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0002,
+      "loss": 1.6336,
+      "step": 1397
+    },
+    {
+      "epoch": 0.029440411070689627,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0002,
+      "loss": 1.6078,
+      "step": 1398
+    },
+    {
+      "epoch": 0.029461470019953355,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0002,
+      "loss": 1.6393,
+      "step": 1399
+    },
+    {
+      "epoch": 0.029482528969217082,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.6385,
+      "step": 1400
+    },
+    {
+      "epoch": 0.029503587918480807,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.6185,
+      "step": 1401
+    },
+    {
+      "epoch": 0.029524646867744534,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.636,
+      "step": 1402
+    },
+    {
+      "epoch": 0.02954570581700826,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0002,
+      "loss": 1.5728,
+      "step": 1403
+    },
+    {
+      "epoch": 0.029566764766271986,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.6191,
+      "step": 1404
+    },
+    {
+      "epoch": 0.029587823715535713,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0002,
+      "loss": 1.6167,
+      "step": 1405
+    },
+    {
+      "epoch": 0.02960888266479944,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.619,
+      "step": 1406
+    },
+    {
+      "epoch": 0.029629941614063165,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.6011,
+      "step": 1407
+    },
+    {
+      "epoch": 0.029651000563326892,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0002,
+      "loss": 1.5973,
+      "step": 1408
+    },
+    {
+      "epoch": 0.02967205951259062,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.6077,
+      "step": 1409
+    },
+    {
+      "epoch": 0.029693118461854347,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0002,
+      "loss": 1.6238,
+      "step": 1410
+    },
+    {
+      "epoch": 0.02971417741111807,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0002,
+      "loss": 1.6273,
+      "step": 1411
+    },
+    {
+      "epoch": 0.0297352363603818,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.6019,
+      "step": 1412
+    },
+    {
+      "epoch": 0.029756295309645527,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0002,
+      "loss": 1.5945,
+      "step": 1413
+    },
+    {
+      "epoch": 0.02977735425890925,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.637,
+      "step": 1414
+    },
+    {
+      "epoch": 0.029798413208172978,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0002,
+      "loss": 1.5983,
+      "step": 1415
+    },
+    {
+      "epoch": 0.029819472157436706,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.6058,
+      "step": 1416
+    },
+    {
+      "epoch": 0.02984053110670043,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.6207,
+      "step": 1417
+    },
+    {
+      "epoch": 0.029861590055964157,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.614,
+      "step": 1418
+    },
+    {
+      "epoch": 0.029882649005227885,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.6096,
+      "step": 1419
+    },
+    {
+      "epoch": 0.029903707954491612,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.5942,
+      "step": 1420
+    },
+    {
+      "epoch": 0.029924766903755336,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.6093,
+      "step": 1421
+    },
+    {
+      "epoch": 0.029945825853019064,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0002,
+      "loss": 1.6512,
+      "step": 1422
+    },
+    {
+      "epoch": 0.02996688480228279,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.6429,
+      "step": 1423
+    },
+    {
+      "epoch": 0.029987943751546516,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0002,
+      "loss": 1.5726,
+      "step": 1424
+    },
+    {
+      "epoch": 0.030009002700810243,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0002,
+      "loss": 1.5582,
+      "step": 1425
+    },
+    {
+      "epoch": 0.03003006165007397,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.6285,
+      "step": 1426
+    },
+    {
+      "epoch": 0.030051120599337695,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6457,
+      "step": 1427
+    },
+    {
+      "epoch": 0.030072179548601422,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.6293,
+      "step": 1428
+    },
+    {
+      "epoch": 0.03009323849786515,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6092,
+      "step": 1429
+    },
+    {
+      "epoch": 0.030114297447128874,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.6335,
+      "step": 1430
+    },
+    {
+      "epoch": 0.0301353563963926,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.5601,
+      "step": 1431
+    },
+    {
+      "epoch": 0.03015641534565633,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.634,
+      "step": 1432
+    },
+    {
+      "epoch": 0.030177474294920056,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.5892,
+      "step": 1433
+    },
+    {
+      "epoch": 0.03019853324418378,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.6294,
+      "step": 1434
+    },
+    {
+      "epoch": 0.030219592193447508,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.6103,
+      "step": 1435
+    },
+    {
+      "epoch": 0.030240651142711236,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6137,
+      "step": 1436
+    },
+    {
+      "epoch": 0.03026171009197496,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.6093,
+      "step": 1437
+    },
+    {
+      "epoch": 0.030282769041238687,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.5959,
+      "step": 1438
+    },
+    {
+      "epoch": 0.030303827990502415,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.6133,
+      "step": 1439
+    },
+    {
+      "epoch": 0.03032488693976614,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.5965,
+      "step": 1440
+    },
+    {
+      "epoch": 0.030345945889029866,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.651,
+      "step": 1441
+    },
+    {
+      "epoch": 0.030367004838293594,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0002,
+      "loss": 1.6389,
+      "step": 1442
+    },
+    {
+      "epoch": 0.03038806378755732,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.6395,
+      "step": 1443
+    },
+    {
+      "epoch": 0.030409122736821045,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.5959,
+      "step": 1444
+    },
+    {
+      "epoch": 0.030430181686084773,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.5967,
+      "step": 1445
+    },
+    {
+      "epoch": 0.0304512406353485,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.6281,
+      "step": 1446
+    },
+    {
+      "epoch": 0.030472299584612225,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.624,
+      "step": 1447
+    },
+    {
+      "epoch": 0.030493358533875952,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0002,
+      "loss": 1.6063,
+      "step": 1448
+    },
+    {
+      "epoch": 0.03051441748313968,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.6102,
+      "step": 1449
+    },
+    {
+      "epoch": 0.030535476432403404,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.6239,
+      "step": 1450
+    },
+    {
+      "epoch": 0.03055653538166713,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.6007,
+      "step": 1451
+    },
+    {
+      "epoch": 0.03057759433093086,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.6162,
+      "step": 1452
+    },
+    {
+      "epoch": 0.030598653280194586,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6271,
+      "step": 1453
+    },
+    {
+      "epoch": 0.03061971222945831,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.6057,
+      "step": 1454
+    },
+    {
+      "epoch": 0.030640771178722038,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.6431,
+      "step": 1455
+    },
+    {
+      "epoch": 0.030661830127985765,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.6073,
+      "step": 1456
+    },
+    {
+      "epoch": 0.03068288907724949,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.6251,
+      "step": 1457
+    },
+    {
+      "epoch": 0.030703948026513217,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.6312,
+      "step": 1458
+    },
+    {
+      "epoch": 0.030725006975776945,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0002,
+      "loss": 1.6331,
+      "step": 1459
+    },
+    {
+      "epoch": 0.03074606592504067,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6186,
+      "step": 1460
+    },
+    {
+      "epoch": 0.030767124874304396,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.5831,
+      "step": 1461
+    },
+    {
+      "epoch": 0.030788183823568124,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.6044,
+      "step": 1462
+    },
+    {
+      "epoch": 0.03080924277283185,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.5956,
+      "step": 1463
+    },
+    {
+      "epoch": 0.030830301722095575,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.0002,
+      "loss": 1.58,
+      "step": 1464
+    },
+    {
+      "epoch": 0.030851360671359303,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0002,
+      "loss": 1.5907,
+      "step": 1465
+    },
+    {
+      "epoch": 0.03087241962062303,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0002,
+      "loss": 1.5986,
+      "step": 1466
+    },
+    {
+      "epoch": 0.030893478569886754,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.6344,
+      "step": 1467
+    },
+    {
+      "epoch": 0.030914537519150482,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5903,
+      "step": 1468
+    },
+    {
+      "epoch": 0.03093559646841421,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.6097,
+      "step": 1469
+    },
+    {
+      "epoch": 0.030956655417677933,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5944,
+      "step": 1470
+    },
+    {
+      "epoch": 0.03097771436694166,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.6254,
+      "step": 1471
+    },
+    {
+      "epoch": 0.03099877331620539,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0002,
+      "loss": 1.5912,
+      "step": 1472
+    },
+    {
+      "epoch": 0.031019832265469116,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.6188,
+      "step": 1473
+    },
+    {
+      "epoch": 0.03104089121473284,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.633,
+      "step": 1474
+    },
+    {
+      "epoch": 0.031061950163996568,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002,
+      "loss": 1.5669,
+      "step": 1475
+    },
+    {
+      "epoch": 0.031083009113260295,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.6276,
+      "step": 1476
+    },
+    {
+      "epoch": 0.03110406806252402,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.6197,
+      "step": 1477
+    },
+    {
+      "epoch": 0.031125127011787747,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.5737,
+      "step": 1478
+    },
+    {
+      "epoch": 0.031146185961051474,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.6145,
+      "step": 1479
+    },
+    {
+      "epoch": 0.0311672449103152,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6141,
+      "step": 1480
+    },
+    {
+      "epoch": 0.031188303859578926,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.6139,
+      "step": 1481
+    },
+    {
+      "epoch": 0.031209362808842653,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0002,
+      "loss": 1.631,
+      "step": 1482
+    },
+    {
+      "epoch": 0.031230421758106378,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0002,
+      "loss": 1.5815,
+      "step": 1483
+    },
+    {
+      "epoch": 0.031251480707370105,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0002,
+      "loss": 1.5871,
+      "step": 1484
+    },
+    {
+      "epoch": 0.03127253965663383,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.6116,
+      "step": 1485
+    },
+    {
+      "epoch": 0.03129359860589756,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.5951,
+      "step": 1486
+    },
+    {
+      "epoch": 0.03131465755516129,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0002,
+      "loss": 1.6043,
+      "step": 1487
+    },
+    {
+      "epoch": 0.03133571650442501,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0002,
+      "loss": 1.6482,
+      "step": 1488
+    },
+    {
+      "epoch": 0.031356775453688736,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.6271,
+      "step": 1489
+    },
+    {
+      "epoch": 0.03137783440295246,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0002,
+      "loss": 1.5916,
+      "step": 1490
+    },
+    {
+      "epoch": 0.03139889335221619,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0002,
+      "loss": 1.605,
+      "step": 1491
+    },
+    {
+      "epoch": 0.03141995230147992,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.6277,
+      "step": 1492
+    },
+    {
+      "epoch": 0.031441011250743646,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.629,
+      "step": 1493
+    },
+    {
+      "epoch": 0.031462070200007373,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0002,
+      "loss": 1.6005,
+      "step": 1494
+    },
+    {
+      "epoch": 0.031483129149271094,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0002,
+      "loss": 1.6365,
+      "step": 1495
+    },
+    {
+      "epoch": 0.03150418809853482,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.59,
+      "step": 1496
+    },
+    {
+      "epoch": 0.03152524704779855,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0002,
+      "loss": 1.6138,
+      "step": 1497
+    },
+    {
+      "epoch": 0.03154630599706228,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.6167,
+      "step": 1498
+    },
+    {
+      "epoch": 0.031567364946326004,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.6102,
+      "step": 1499
+    },
+    {
+      "epoch": 0.03158842389558973,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.6836,
+      "step": 1500
+    },
+    {
+      "epoch": 0.03158842389558973,
+      "eval_loss": 1.7313117980957031,
+      "eval_runtime": 898.7214,
+      "eval_samples_per_second": 68.764,
+      "eval_steps_per_second": 2.15,
+      "step": 1500
+    },
+    {
+      "epoch": 0.03160948284485346,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5911,
+      "step": 1501
+    },
+    {
+      "epoch": 0.03163054179411718,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.6335,
+      "step": 1502
+    },
+    {
+      "epoch": 0.03165160074338091,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0002,
+      "loss": 1.6045,
+      "step": 1503
+    },
+    {
+      "epoch": 0.031672659692644635,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5905,
+      "step": 1504
+    },
+    {
+      "epoch": 0.03169371864190836,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.5942,
+      "step": 1505
+    },
+    {
+      "epoch": 0.03171477759117209,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.6136,
+      "step": 1506
+    },
+    {
+      "epoch": 0.03173583654043582,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.6084,
+      "step": 1507
+    },
+    {
+      "epoch": 0.03175689548969954,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002,
+      "loss": 1.5827,
+      "step": 1508
+    },
+    {
+      "epoch": 0.031777954438963266,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0002,
+      "loss": 1.6446,
+      "step": 1509
+    },
+    {
+      "epoch": 0.03179901338822699,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0002,
+      "loss": 1.6009,
+      "step": 1510
+    },
+    {
+      "epoch": 0.03182007233749072,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.5915,
+      "step": 1511
+    },
+    {
+      "epoch": 0.03184113128675445,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6117,
+      "step": 1512
+    },
+    {
+      "epoch": 0.031862190236018176,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.6005,
+      "step": 1513
+    },
+    {
+      "epoch": 0.0318832491852819,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0002,
+      "loss": 1.626,
+      "step": 1514
+    },
+    {
+      "epoch": 0.031904308134545624,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0002,
+      "loss": 1.6183,
+      "step": 1515
+    },
+    {
+      "epoch": 0.03192536708380935,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002,
+      "loss": 1.5971,
+      "step": 1516
+    },
+    {
+      "epoch": 0.03194642603307308,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6177,
+      "step": 1517
+    },
+    {
+      "epoch": 0.031967484982336807,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0002,
+      "loss": 1.615,
+      "step": 1518
+    },
+    {
+      "epoch": 0.031988543931600534,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0002,
+      "loss": 1.5933,
+      "step": 1519
+    },
+    {
+      "epoch": 0.03200960288086426,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.5807,
+      "step": 1520
+    },
+    {
+      "epoch": 0.03203066183012799,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0002,
+      "loss": 1.6462,
+      "step": 1521
+    },
+    {
+      "epoch": 0.03205172077939171,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6035,
+      "step": 1522
+    },
+    {
+      "epoch": 0.03207277972865544,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.5976,
+      "step": 1523
+    },
+    {
+      "epoch": 0.032093838677919165,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.5999,
+      "step": 1524
+    },
+    {
+      "epoch": 0.03211489762718289,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0002,
+      "loss": 1.6359,
+      "step": 1525
+    },
+    {
+      "epoch": 0.03213595657644662,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0002,
+      "loss": 1.6127,
+      "step": 1526
+    },
+    {
+      "epoch": 0.03215701552571035,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.6089,
+      "step": 1527
+    },
+    {
+      "epoch": 0.03217807447497407,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.6326,
+      "step": 1528
+    },
+    {
+      "epoch": 0.032199133424237796,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.5891,
+      "step": 1529
+    },
+    {
+      "epoch": 0.03222019237350152,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5946,
+      "step": 1530
+    },
+    {
+      "epoch": 0.03224125132276525,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0002,
+      "loss": 1.6025,
+      "step": 1531
+    },
+    {
+      "epoch": 0.03226231027202898,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0002,
+      "loss": 1.6143,
+      "step": 1532
+    },
+    {
+      "epoch": 0.032283369221292706,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.6418,
+      "step": 1533
+    },
+    {
+      "epoch": 0.03230442817055643,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6017,
+      "step": 1534
+    },
+    {
+      "epoch": 0.032325487119820154,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.6064,
+      "step": 1535
+    },
+    {
+      "epoch": 0.03234654606908388,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0002,
+      "loss": 1.6064,
+      "step": 1536
+    },
+    {
+      "epoch": 0.03236760501834761,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6277,
+      "step": 1537
+    },
+    {
+      "epoch": 0.032388663967611336,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.5879,
+      "step": 1538
+    },
+    {
+      "epoch": 0.032409722916875064,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002,
+      "loss": 1.6506,
+      "step": 1539
+    },
+    {
+      "epoch": 0.03243078186613879,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6257,
+      "step": 1540
+    },
+    {
+      "epoch": 0.03245184081540251,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.621,
+      "step": 1541
+    },
+    {
+      "epoch": 0.03247289976466624,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.6056,
+      "step": 1542
+    },
+    {
+      "epoch": 0.03249395871392997,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.6221,
+      "step": 1543
+    },
+    {
+      "epoch": 0.032515017663193695,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0002,
+      "loss": 1.6464,
+      "step": 1544
+    },
+    {
+      "epoch": 0.03253607661245742,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.6132,
+      "step": 1545
+    },
+    {
+      "epoch": 0.03255713556172115,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.628,
+      "step": 1546
+    },
+    {
+      "epoch": 0.03257819451098488,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0002,
+      "loss": 1.6171,
+      "step": 1547
+    },
+    {
+      "epoch": 0.0325992534602486,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.5599,
+      "step": 1548
+    },
+    {
+      "epoch": 0.032620312409512325,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.574,
+      "step": 1549
+    },
+    {
+      "epoch": 0.03264137135877605,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0002,
+      "loss": 1.5811,
+      "step": 1550
+    },
+    {
+      "epoch": 0.03266243030803978,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.5824,
+      "step": 1551
+    },
+    {
+      "epoch": 0.03268348925730351,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.6265,
+      "step": 1552
+    },
+    {
+      "epoch": 0.032704548206567235,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.6438,
+      "step": 1553
+    },
+    {
+      "epoch": 0.03272560715583096,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.6078,
+      "step": 1554
+    },
+    {
+      "epoch": 0.032746666105094684,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.6268,
+      "step": 1555
+    },
+    {
+      "epoch": 0.03276772505435841,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0002,
+      "loss": 1.5757,
+      "step": 1556
+    },
+    {
+      "epoch": 0.03278878400362214,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0002,
+      "loss": 1.6165,
+      "step": 1557
+    },
+    {
+      "epoch": 0.032809842952885866,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6161,
+      "step": 1558
+    },
+    {
+      "epoch": 0.032830901902149594,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.6304,
+      "step": 1559
+    },
+    {
+      "epoch": 0.03285196085141332,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0002,
+      "loss": 1.6509,
+      "step": 1560
+    },
+    {
+      "epoch": 0.03287301980067704,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.5895,
+      "step": 1561
+    },
+    {
+      "epoch": 0.03289407874994077,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6204,
+      "step": 1562
+    },
+    {
+      "epoch": 0.0329151376992045,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.5967,
+      "step": 1563
+    },
+    {
+      "epoch": 0.032936196648468224,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0002,
+      "loss": 1.627,
+      "step": 1564
+    },
+    {
+      "epoch": 0.03295725559773195,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.6286,
+      "step": 1565
+    },
+    {
+      "epoch": 0.03297831454699568,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0002,
+      "loss": 1.588,
+      "step": 1566
+    },
+    {
+      "epoch": 0.03299937349625941,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0002,
+      "loss": 1.6567,
+      "step": 1567
+    },
+    {
+      "epoch": 0.03302043244552313,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0002,
+      "loss": 1.6112,
+      "step": 1568
+    },
+    {
+      "epoch": 0.033041491394786855,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0002,
+      "loss": 1.5737,
+      "step": 1569
+    },
+    {
+      "epoch": 0.03306255034405058,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.606,
+      "step": 1570
+    },
+    {
+      "epoch": 0.03308360929331431,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0002,
+      "loss": 1.6129,
+      "step": 1571
+    },
+    {
+      "epoch": 0.03310466824257804,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0002,
+      "loss": 1.5904,
+      "step": 1572
+    },
+    {
+      "epoch": 0.033125727191841765,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.6233,
+      "step": 1573
+    },
+    {
+      "epoch": 0.03314678614110549,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.6132,
+      "step": 1574
+    },
+    {
+      "epoch": 0.03316784509036921,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0002,
+      "loss": 1.5877,
+      "step": 1575
+    },
+    {
+      "epoch": 0.03318890403963294,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.5909,
+      "step": 1576
+    },
+    {
+      "epoch": 0.03320996298889667,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.6259,
+      "step": 1577
+    },
+    {
+      "epoch": 0.033231021938160396,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002,
+      "loss": 1.6395,
+      "step": 1578
+    },
+    {
+      "epoch": 0.033252080887424124,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0002,
+      "loss": 1.6033,
+      "step": 1579
+    },
+    {
+      "epoch": 0.03327313983668785,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.5902,
+      "step": 1580
+    },
+    {
+      "epoch": 0.03329419878595157,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.6249,
+      "step": 1581
+    },
+    {
+      "epoch": 0.0333152577352153,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.6291,
+      "step": 1582
+    },
+    {
+      "epoch": 0.03333631668447903,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0002,
+      "loss": 1.6002,
+      "step": 1583
+    },
+    {
+      "epoch": 0.033357375633742754,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5868,
+      "step": 1584
+    },
+    {
+      "epoch": 0.03337843458300648,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.5746,
+      "step": 1585
+    },
+    {
+      "epoch": 0.03339949353227021,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.6275,
+      "step": 1586
+    },
+    {
+      "epoch": 0.03342055248153394,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0002,
+      "loss": 1.5987,
+      "step": 1587
+    },
+    {
+      "epoch": 0.03344161143079766,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.6521,
+      "step": 1588
+    },
+    {
+      "epoch": 0.033462670380061385,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0002,
+      "loss": 1.5996,
+      "step": 1589
+    },
+    {
+      "epoch": 0.03348372932932511,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0002,
+      "loss": 1.6361,
+      "step": 1590
+    },
+    {
+      "epoch": 0.03350478827858884,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.6203,
+      "step": 1591
+    },
+    {
+      "epoch": 0.03352584722785257,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0002,
+      "loss": 1.5911,
+      "step": 1592
+    },
+    {
+      "epoch": 0.033546906177116295,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.0002,
+      "loss": 1.629,
+      "step": 1593
+    },
+    {
+      "epoch": 0.03356796512638002,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.6434,
+      "step": 1594
+    },
+    {
+      "epoch": 0.03358902407564374,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.6314,
+      "step": 1595
+    },
+    {
+      "epoch": 0.03361008302490747,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.6035,
+      "step": 1596
+    },
+    {
+      "epoch": 0.0336311419741712,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002,
+      "loss": 1.6554,
+      "step": 1597
+    },
+    {
+      "epoch": 0.033652200923434926,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.6018,
+      "step": 1598
+    },
+    {
+      "epoch": 0.03367325987269865,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.5738,
+      "step": 1599
+    },
+    {
+      "epoch": 0.03369431882196238,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.5935,
+      "step": 1600
+    },
+    {
+      "epoch": 0.0337153777712261,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.6285,
+      "step": 1601
+    },
+    {
+      "epoch": 0.03373643672048983,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.6414,
+      "step": 1602
+    },
+    {
+      "epoch": 0.03375749566975356,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6071,
+      "step": 1603
+    },
+    {
+      "epoch": 0.033778554619017284,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6192,
+      "step": 1604
+    },
+    {
+      "epoch": 0.03379961356828101,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0002,
+      "loss": 1.6178,
+      "step": 1605
+    },
+    {
+      "epoch": 0.03382067251754474,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.6154,
+      "step": 1606
+    },
+    {
+      "epoch": 0.03384173146680847,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.6021,
+      "step": 1607
+    },
+    {
+      "epoch": 0.03386279041607219,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.6044,
+      "step": 1608
+    },
+    {
+      "epoch": 0.033883849365335915,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6268,
+      "step": 1609
+    },
+    {
+      "epoch": 0.03390490831459964,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.6323,
+      "step": 1610
+    },
+    {
+      "epoch": 0.03392596726386337,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0002,
+      "loss": 1.6262,
+      "step": 1611
+    },
+    {
+      "epoch": 0.0339470262131271,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6007,
+      "step": 1612
+    },
+    {
+      "epoch": 0.033968085162390825,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.5836,
+      "step": 1613
+    },
+    {
+      "epoch": 0.033989144111654546,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6018,
+      "step": 1614
+    },
+    {
+      "epoch": 0.03401020306091827,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.614,
+      "step": 1615
+    },
+    {
+      "epoch": 0.034031262010182,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0002,
+      "loss": 1.6194,
+      "step": 1616
+    },
+    {
+      "epoch": 0.03405232095944573,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0002,
+      "loss": 1.589,
+      "step": 1617
+    },
+    {
+      "epoch": 0.034073379908709456,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.6081,
+      "step": 1618
+    },
+    {
+      "epoch": 0.03409443885797318,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5886,
+      "step": 1619
+    },
+    {
+      "epoch": 0.03411549780723691,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.6132,
+      "step": 1620
+    },
+    {
+      "epoch": 0.03413655675650063,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.6266,
+      "step": 1621
+    },
+    {
+      "epoch": 0.03415761570576436,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.6185,
+      "step": 1622
+    },
+    {
+      "epoch": 0.034178674655028086,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.5941,
+      "step": 1623
+    },
+    {
+      "epoch": 0.034199733604291814,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.6127,
+      "step": 1624
+    },
+    {
+      "epoch": 0.03422079255355554,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0002,
+      "loss": 1.6169,
+      "step": 1625
+    },
+    {
+      "epoch": 0.03424185150281927,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.6155,
+      "step": 1626
+    },
+    {
+      "epoch": 0.034262910452083,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0002,
+      "loss": 1.5958,
+      "step": 1627
+    },
+    {
+      "epoch": 0.03428396940134672,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.6116,
+      "step": 1628
+    },
+    {
+      "epoch": 0.034305028350610445,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.628,
+      "step": 1629
+    },
+    {
+      "epoch": 0.03432608729987417,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0002,
+      "loss": 1.5775,
+      "step": 1630
+    },
+    {
+      "epoch": 0.0343471462491379,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0002,
+      "loss": 1.5662,
+      "step": 1631
+    },
+    {
+      "epoch": 0.03436820519840163,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0002,
+      "loss": 1.5872,
+      "step": 1632
+    },
+    {
+      "epoch": 0.034389264147665355,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0002,
+      "loss": 1.6273,
+      "step": 1633
+    },
+    {
+      "epoch": 0.034410323096929075,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5463,
+      "step": 1634
+    },
+    {
+      "epoch": 0.0344313820461928,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.605,
+      "step": 1635
+    },
+    {
+      "epoch": 0.03445244099545653,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0002,
+      "loss": 1.6065,
+      "step": 1636
+    },
+    {
+      "epoch": 0.03447349994472026,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0002,
+      "loss": 1.6039,
+      "step": 1637
+    },
+    {
+      "epoch": 0.034494558893983986,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.6063,
+      "step": 1638
+    },
+    {
+      "epoch": 0.03451561784324771,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0002,
+      "loss": 1.6061,
+      "step": 1639
+    },
+    {
+      "epoch": 0.03453667679251144,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0002,
+      "loss": 1.6303,
+      "step": 1640
+    },
+    {
+      "epoch": 0.03455773574177516,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.6254,
+      "step": 1641
+    },
+    {
+      "epoch": 0.03457879469103889,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0002,
+      "loss": 1.5806,
+      "step": 1642
+    },
+    {
+      "epoch": 0.034599853640302616,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0002,
+      "loss": 1.6324,
+      "step": 1643
+    },
+    {
+      "epoch": 0.034620912589566344,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0002,
+      "loss": 1.6107,
+      "step": 1644
+    },
+    {
+      "epoch": 0.03464197153883007,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6329,
+      "step": 1645
+    },
+    {
+      "epoch": 0.0346630304880938,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.5787,
+      "step": 1646
+    },
+    {
+      "epoch": 0.034684089437357526,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0002,
+      "loss": 1.6358,
+      "step": 1647
+    },
+    {
+      "epoch": 0.03470514838662125,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.5805,
+      "step": 1648
+    },
+    {
+      "epoch": 0.034726207335884975,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0002,
+      "loss": 1.6357,
+      "step": 1649
+    },
+    {
+      "epoch": 0.0347472662851487,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.0002,
+      "loss": 1.6247,
+      "step": 1650
+    },
+    {
+      "epoch": 0.03476832523441243,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.5739,
+      "step": 1651
+    },
+    {
+      "epoch": 0.03478938418367616,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.611,
+      "step": 1652
+    },
+    {
+      "epoch": 0.034810443132939885,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.5943,
+      "step": 1653
+    },
+    {
+      "epoch": 0.034831502082203605,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0002,
+      "loss": 1.6355,
+      "step": 1654
+    },
+    {
+      "epoch": 0.03485256103146733,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6046,
+      "step": 1655
+    },
+    {
+      "epoch": 0.03487361998073106,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.5793,
+      "step": 1656
+    },
+    {
+      "epoch": 0.03489467892999479,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0002,
+      "loss": 1.5959,
+      "step": 1657
+    },
+    {
+      "epoch": 0.034915737879258515,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.6051,
+      "step": 1658
+    },
+    {
+      "epoch": 0.03493679682852224,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.5824,
+      "step": 1659
+    },
+    {
+      "epoch": 0.03495785577778597,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6045,
+      "step": 1660
+    },
+    {
+      "epoch": 0.03497891472704969,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.6156,
+      "step": 1661
+    },
+    {
+      "epoch": 0.03499997367631342,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6329,
+      "step": 1662
+    },
+    {
+      "epoch": 0.035021032625577146,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.6073,
+      "step": 1663
+    },
+    {
+      "epoch": 0.035042091574840874,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.664,
+      "step": 1664
+    },
+    {
+      "epoch": 0.0350631505241046,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.6037,
+      "step": 1665
+    },
+    {
+      "epoch": 0.03508420947336833,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.5835,
+      "step": 1666
+    },
+    {
+      "epoch": 0.03510526842263205,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.5762,
+      "step": 1667
+    },
+    {
+      "epoch": 0.03512632737189578,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.6209,
+      "step": 1668
+    },
+    {
+      "epoch": 0.035147386321159504,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5672,
+      "step": 1669
+    },
+    {
+      "epoch": 0.03516844527042323,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.6377,
+      "step": 1670
+    },
+    {
+      "epoch": 0.03518950421968696,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0002,
+      "loss": 1.61,
+      "step": 1671
+    },
+    {
+      "epoch": 0.03521056316895069,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.6655,
+      "step": 1672
+    },
+    {
+      "epoch": 0.035231622118214415,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.6076,
+      "step": 1673
+    },
+    {
+      "epoch": 0.035252681067478135,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.6174,
+      "step": 1674
+    },
+    {
+      "epoch": 0.03527374001674186,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6268,
+      "step": 1675
+    },
+    {
+      "epoch": 0.03529479896600559,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.6115,
+      "step": 1676
+    },
+    {
+      "epoch": 0.03531585791526932,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0002,
+      "loss": 1.6236,
+      "step": 1677
+    },
+    {
+      "epoch": 0.035336916864533045,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.592,
+      "step": 1678
+    },
+    {
+      "epoch": 0.03535797581379677,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0002,
+      "loss": 1.5799,
+      "step": 1679
+    },
+    {
+      "epoch": 0.0353790347630605,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6257,
+      "step": 1680
+    },
+    {
+      "epoch": 0.03540009371232422,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0002,
+      "loss": 1.5978,
+      "step": 1681
+    },
+    {
+      "epoch": 0.03542115266158795,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.575,
+      "step": 1682
+    },
+    {
+      "epoch": 0.035442211610851676,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.6092,
+      "step": 1683
+    },
+    {
+      "epoch": 0.035463270560115404,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.6042,
+      "step": 1684
+    },
+    {
+      "epoch": 0.03548432950937913,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.6263,
+      "step": 1685
+    },
+    {
+      "epoch": 0.03550538845864286,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.6079,
+      "step": 1686
+    },
+    {
+      "epoch": 0.03552644740790658,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.5731,
+      "step": 1687
+    },
+    {
+      "epoch": 0.03554750635717031,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0002,
+      "loss": 1.5906,
+      "step": 1688
+    },
+    {
+      "epoch": 0.035568565306434034,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0002,
+      "loss": 1.5599,
+      "step": 1689
+    },
+    {
+      "epoch": 0.03558962425569776,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.6101,
+      "step": 1690
+    },
+    {
+      "epoch": 0.03561068320496149,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.5947,
+      "step": 1691
+    },
+    {
+      "epoch": 0.03563174215422522,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.0002,
+      "loss": 1.5805,
+      "step": 1692
+    },
+    {
+      "epoch": 0.035652801103488944,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.6101,
+      "step": 1693
+    },
+    {
+      "epoch": 0.035673860052752665,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.6348,
+      "step": 1694
+    },
+    {
+      "epoch": 0.03569491900201639,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.5941,
+      "step": 1695
+    },
+    {
+      "epoch": 0.03571597795128012,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0002,
+      "loss": 1.5894,
+      "step": 1696
+    },
+    {
+      "epoch": 0.03573703690054385,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0002,
+      "loss": 1.5782,
+      "step": 1697
+    },
+    {
+      "epoch": 0.035758095849807575,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0002,
+      "loss": 1.6318,
+      "step": 1698
+    },
+    {
+      "epoch": 0.0357791547990713,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0002,
+      "loss": 1.5965,
+      "step": 1699
+    },
+    {
+      "epoch": 0.03580021374833503,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0002,
+      "loss": 1.5629,
+      "step": 1700
+    },
+    {
+      "epoch": 0.03582127269759875,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0002,
+      "loss": 1.5906,
+      "step": 1701
+    },
+    {
+      "epoch": 0.03584233164686248,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.5796,
+      "step": 1702
+    },
+    {
+      "epoch": 0.035863390596126206,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.5959,
+      "step": 1703
+    },
+    {
+      "epoch": 0.03588444954538993,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6076,
+      "step": 1704
+    },
+    {
+      "epoch": 0.03590550849465366,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0002,
+      "loss": 1.5982,
+      "step": 1705
+    },
+    {
+      "epoch": 0.03592656744391739,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.6165,
+      "step": 1706
+    },
+    {
+      "epoch": 0.03594762639318111,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0002,
+      "loss": 1.6494,
+      "step": 1707
+    },
+    {
+      "epoch": 0.03596868534244484,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.5864,
+      "step": 1708
+    },
+    {
+      "epoch": 0.035989744291708564,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.6305,
+      "step": 1709
+    },
+    {
+      "epoch": 0.03601080324097229,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.5995,
+      "step": 1710
+    },
+    {
+      "epoch": 0.03603186219023602,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0002,
+      "loss": 1.6088,
+      "step": 1711
+    },
+    {
+      "epoch": 0.03605292113949975,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0002,
+      "loss": 1.6147,
+      "step": 1712
+    },
+    {
+      "epoch": 0.036073980088763474,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0002,
+      "loss": 1.6155,
+      "step": 1713
+    },
+    {
+      "epoch": 0.036095039038027195,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0002,
+      "loss": 1.6134,
+      "step": 1714
+    },
+    {
+      "epoch": 0.03611609798729092,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0002,
+      "loss": 1.5811,
+      "step": 1715
+    },
+    {
+      "epoch": 0.03613715693655465,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.5833,
+      "step": 1716
+    },
+    {
+      "epoch": 0.03615821588581838,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.6356,
+      "step": 1717
+    },
+    {
+      "epoch": 0.036179274835082105,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002,
+      "loss": 1.5993,
+      "step": 1718
+    },
+    {
+      "epoch": 0.03620033378434583,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0002,
+      "loss": 1.6024,
+      "step": 1719
+    },
+    {
+      "epoch": 0.03622139273360955,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0002,
+      "loss": 1.6354,
+      "step": 1720
+    },
+    {
+      "epoch": 0.03624245168287328,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.6091,
+      "step": 1721
+    },
+    {
+      "epoch": 0.03626351063213701,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0002,
+      "loss": 1.5913,
+      "step": 1722
+    },
+    {
+      "epoch": 0.036284569581400736,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5904,
+      "step": 1723
+    },
+    {
+      "epoch": 0.03630562853066446,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0002,
+      "loss": 1.6041,
+      "step": 1724
+    },
+    {
+      "epoch": 0.03632668747992819,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.6058,
+      "step": 1725
+    },
+    {
+      "epoch": 0.03634774642919192,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0002,
+      "loss": 1.5993,
+      "step": 1726
+    },
+    {
+      "epoch": 0.03636880537845564,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0002,
+      "loss": 1.6136,
+      "step": 1727
+    },
+    {
+      "epoch": 0.036389864327719366,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.6139,
+      "step": 1728
+    },
+    {
+      "epoch": 0.036410923276983094,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5759,
+      "step": 1729
+    },
+    {
+      "epoch": 0.03643198222624682,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0002,
+      "loss": 1.592,
+      "step": 1730
+    },
+    {
+      "epoch": 0.03645304117551055,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.5542,
+      "step": 1731
+    },
+    {
+      "epoch": 0.03647410012477428,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.6252,
+      "step": 1732
+    },
+    {
+      "epoch": 0.036495159074038004,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.5766,
+      "step": 1733
+    },
+    {
+      "epoch": 0.036516218023301725,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.6156,
+      "step": 1734
+    },
+    {
+      "epoch": 0.03653727697256545,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0002,
+      "loss": 1.6018,
+      "step": 1735
+    },
+    {
+      "epoch": 0.03655833592182918,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.6105,
+      "step": 1736
+    },
+    {
+      "epoch": 0.03657939487109291,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.613,
+      "step": 1737
+    },
+    {
+      "epoch": 0.036600453820356635,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.5816,
+      "step": 1738
+    },
+    {
+      "epoch": 0.03662151276962036,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6076,
+      "step": 1739
+    },
+    {
+      "epoch": 0.03664257171888408,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.6547,
+      "step": 1740
+    },
+    {
+      "epoch": 0.03666363066814781,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.6218,
+      "step": 1741
+    },
+    {
+      "epoch": 0.03668468961741154,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.6065,
+      "step": 1742
+    },
+    {
+      "epoch": 0.036705748566675266,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0002,
+      "loss": 1.6045,
+      "step": 1743
+    },
+    {
+      "epoch": 0.03672680751593899,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0002,
+      "loss": 1.5764,
+      "step": 1744
+    },
+    {
+      "epoch": 0.03674786646520272,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.6176,
+      "step": 1745
+    },
+    {
+      "epoch": 0.03676892541446645,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.6423,
+      "step": 1746
+    },
+    {
+      "epoch": 0.03678998436373017,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.6186,
+      "step": 1747
+    },
+    {
+      "epoch": 0.036811043312993896,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0002,
+      "loss": 1.6053,
+      "step": 1748
+    },
+    {
+      "epoch": 0.036832102262257624,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.6318,
+      "step": 1749
+    },
+    {
+      "epoch": 0.03685316121152135,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.6391,
+      "step": 1750
+    },
+    {
+      "epoch": 0.03687422016078508,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5968,
+      "step": 1751
+    },
+    {
+      "epoch": 0.036895279110048806,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0002,
+      "loss": 1.6101,
+      "step": 1752
+    },
+    {
+      "epoch": 0.036916338059312534,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.6168,
+      "step": 1753
+    },
+    {
+      "epoch": 0.036937397008576255,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5969,
+      "step": 1754
+    },
+    {
+      "epoch": 0.03695845595783998,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5838,
+      "step": 1755
+    },
+    {
+      "epoch": 0.03697951490710371,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.5844,
+      "step": 1756
+    },
+    {
+      "epoch": 0.03700057385636744,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0002,
+      "loss": 1.5927,
+      "step": 1757
+    },
+    {
+      "epoch": 0.037021632805631165,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.6227,
+      "step": 1758
+    },
+    {
+      "epoch": 0.03704269175489489,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0002,
+      "loss": 1.6242,
+      "step": 1759
+    },
+    {
+      "epoch": 0.03706375070415861,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.0002,
+      "loss": 1.5955,
+      "step": 1760
+    },
+    {
+      "epoch": 0.03708480965342234,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6174,
+      "step": 1761
+    },
+    {
+      "epoch": 0.03710586860268607,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.0002,
+      "loss": 1.6206,
+      "step": 1762
+    },
+    {
+      "epoch": 0.037126927551949795,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0002,
+      "loss": 1.6082,
+      "step": 1763
+    },
+    {
+      "epoch": 0.03714798650121352,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.0002,
+      "loss": 1.579,
+      "step": 1764
+    },
+    {
+      "epoch": 0.03716904545047725,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0002,
+      "loss": 1.5848,
+      "step": 1765
+    },
+    {
+      "epoch": 0.03719010439974098,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0002,
+      "loss": 1.6277,
+      "step": 1766
+    },
+    {
+      "epoch": 0.0372111633490047,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6189,
+      "step": 1767
+    },
+    {
+      "epoch": 0.037232222298268426,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5865,
+      "step": 1768
+    },
+    {
+      "epoch": 0.037253281247532154,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0002,
+      "loss": 1.6111,
+      "step": 1769
+    },
+    {
+      "epoch": 0.03727434019679588,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.5833,
+      "step": 1770
+    },
+    {
+      "epoch": 0.03729539914605961,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.6287,
+      "step": 1771
+    },
+    {
+      "epoch": 0.037316458095323336,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.5966,
+      "step": 1772
+    },
+    {
+      "epoch": 0.03733751704458706,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6174,
+      "step": 1773
+    },
+    {
+      "epoch": 0.037358575993850784,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.6008,
+      "step": 1774
+    },
+    {
+      "epoch": 0.03737963494311451,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0002,
+      "loss": 1.6294,
+      "step": 1775
+    },
+    {
+      "epoch": 0.03740069389237824,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0002,
+      "loss": 1.5896,
+      "step": 1776
+    },
+    {
+      "epoch": 0.03742175284164197,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.6171,
+      "step": 1777
+    },
+    {
+      "epoch": 0.037442811790905695,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.5954,
+      "step": 1778
+    },
+    {
+      "epoch": 0.03746387074016942,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0002,
+      "loss": 1.6256,
+      "step": 1779
+    },
+    {
+      "epoch": 0.03748492968943314,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.5872,
+      "step": 1780
+    },
+    {
+      "epoch": 0.03750598863869687,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.5856,
+      "step": 1781
+    },
+    {
+      "epoch": 0.0375270475879606,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0002,
+      "loss": 1.5942,
+      "step": 1782
+    },
+    {
+      "epoch": 0.037548106537224325,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0002,
+      "loss": 1.5825,
+      "step": 1783
+    },
+    {
+      "epoch": 0.03756916548648805,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0002,
+      "loss": 1.5898,
+      "step": 1784
+    },
+    {
+      "epoch": 0.03759022443575178,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0002,
+      "loss": 1.6136,
+      "step": 1785
+    },
+    {
+      "epoch": 0.03761128338501551,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0002,
+      "loss": 1.5853,
+      "step": 1786
+    },
+    {
+      "epoch": 0.03763234233427923,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0002,
+      "loss": 1.6042,
+      "step": 1787
+    },
+    {
+      "epoch": 0.037653401283542956,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0002,
+      "loss": 1.5865,
+      "step": 1788
+    },
+    {
+      "epoch": 0.037674460232806684,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.0002,
+      "loss": 1.5853,
+      "step": 1789
+    },
+    {
+      "epoch": 0.03769551918207041,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0002,
+      "loss": 1.5975,
+      "step": 1790
+    },
+    {
+      "epoch": 0.03771657813133414,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.6065,
+      "step": 1791
+    },
+    {
+      "epoch": 0.037737637080597866,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0002,
+      "loss": 1.608,
+      "step": 1792
+    },
+    {
+      "epoch": 0.03775869602986159,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5896,
+      "step": 1793
+    },
+    {
+      "epoch": 0.037779754979125314,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0002,
+      "loss": 1.6268,
+      "step": 1794
+    },
+    {
+      "epoch": 0.03780081392838904,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0002,
+      "loss": 1.5722,
+      "step": 1795
+    },
+    {
+      "epoch": 0.03782187287765277,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6395,
+      "step": 1796
+    },
+    {
+      "epoch": 0.0378429318269165,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0002,
+      "loss": 1.595,
+      "step": 1797
+    },
+    {
+      "epoch": 0.037863990776180224,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0002,
+      "loss": 1.6087,
+      "step": 1798
+    },
+    {
+      "epoch": 0.03788504972544395,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.6041,
+      "step": 1799
+    },
+    {
+      "epoch": 0.03790610867470767,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.6147,
+      "step": 1800
+    },
+    {
+      "epoch": 0.0379271676239714,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0002,
+      "loss": 1.5663,
+      "step": 1801
+    },
+    {
+      "epoch": 0.03794822657323513,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0002,
+      "loss": 1.5874,
+      "step": 1802
+    },
+    {
+      "epoch": 0.037969285522498855,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.5665,
+      "step": 1803
+    },
+    {
+      "epoch": 0.03799034447176258,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6276,
+      "step": 1804
+    },
+    {
+      "epoch": 0.03801140342102631,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.6161,
+      "step": 1805
+    },
+    {
+      "epoch": 0.03803246237029004,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.6328,
+      "step": 1806
+    },
+    {
+      "epoch": 0.03805352131955376,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.6246,
+      "step": 1807
+    },
+    {
+      "epoch": 0.038074580268817486,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5932,
+      "step": 1808
+    },
+    {
+      "epoch": 0.03809563921808121,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.6207,
+      "step": 1809
+    },
+    {
+      "epoch": 0.03811669816734494,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6203,
+      "step": 1810
+    },
+    {
+      "epoch": 0.03813775711660867,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.6118,
+      "step": 1811
+    },
+    {
+      "epoch": 0.038158816065872396,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.6038,
+      "step": 1812
+    },
+    {
+      "epoch": 0.03817987501513612,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0002,
+      "loss": 1.5791,
+      "step": 1813
+    },
+    {
+      "epoch": 0.038200933964399844,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002,
+      "loss": 1.6181,
+      "step": 1814
+    },
+    {
+      "epoch": 0.03822199291366357,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5981,
+      "step": 1815
+    },
+    {
+      "epoch": 0.0382430518629273,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.6115,
+      "step": 1816
+    },
+    {
+      "epoch": 0.03826411081219103,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0002,
+      "loss": 1.6006,
+      "step": 1817
+    },
+    {
+      "epoch": 0.038285169761454754,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.6372,
+      "step": 1818
+    },
+    {
+      "epoch": 0.03830622871071848,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.6259,
+      "step": 1819
+    },
+    {
+      "epoch": 0.0383272876599822,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0002,
+      "loss": 1.5941,
+      "step": 1820
+    },
+    {
+      "epoch": 0.03834834660924593,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0002,
+      "loss": 1.6278,
+      "step": 1821
+    },
+    {
+      "epoch": 0.03836940555850966,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.5952,
+      "step": 1822
+    },
+    {
+      "epoch": 0.038390464507773385,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0002,
+      "loss": 1.5975,
+      "step": 1823
+    },
+    {
+      "epoch": 0.03841152345703711,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0002,
+      "loss": 1.5658,
+      "step": 1824
+    },
+    {
+      "epoch": 0.03843258240630084,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0002,
+      "loss": 1.6069,
+      "step": 1825
+    },
+    {
+      "epoch": 0.03845364135556456,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6112,
+      "step": 1826
+    },
+    {
+      "epoch": 0.03847470030482829,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.6003,
+      "step": 1827
+    },
+    {
+      "epoch": 0.038495759254092016,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.6071,
+      "step": 1828
+    },
+    {
+      "epoch": 0.03851681820335574,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.5912,
+      "step": 1829
+    },
+    {
+      "epoch": 0.03853787715261947,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.5888,
+      "step": 1830
+    },
+    {
+      "epoch": 0.0385589361018832,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0002,
+      "loss": 1.5842,
+      "step": 1831
+    },
+    {
+      "epoch": 0.038579995051146926,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6217,
+      "step": 1832
+    },
+    {
+      "epoch": 0.038601054000410646,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0002,
+      "loss": 1.574,
+      "step": 1833
+    },
+    {
+      "epoch": 0.038622112949674374,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6071,
+      "step": 1834
+    },
+    {
+      "epoch": 0.0386431718989381,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.5787,
+      "step": 1835
+    },
+    {
+      "epoch": 0.03866423084820183,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0002,
+      "loss": 1.6114,
+      "step": 1836
+    },
+    {
+      "epoch": 0.03868528979746556,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.5771,
+      "step": 1837
+    },
+    {
+      "epoch": 0.038706348746729284,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.5815,
+      "step": 1838
+    },
+    {
+      "epoch": 0.03872740769599301,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.5942,
+      "step": 1839
+    },
+    {
+      "epoch": 0.03874846664525673,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.5875,
+      "step": 1840
+    },
+    {
+      "epoch": 0.03876952559452046,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5964,
+      "step": 1841
+    },
+    {
+      "epoch": 0.03879058454378419,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0002,
+      "loss": 1.5991,
+      "step": 1842
+    },
+    {
+      "epoch": 0.038811643493047915,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0002,
+      "loss": 1.582,
+      "step": 1843
+    },
+    {
+      "epoch": 0.03883270244231164,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5934,
+      "step": 1844
+    },
+    {
+      "epoch": 0.03885376139157537,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6071,
+      "step": 1845
+    },
+    {
+      "epoch": 0.03887482034083909,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.606,
+      "step": 1846
+    },
+    {
+      "epoch": 0.03889587929010282,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.5832,
+      "step": 1847
+    },
+    {
+      "epoch": 0.038916938239366546,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.6305,
+      "step": 1848
+    },
+    {
+      "epoch": 0.03893799718863027,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0002,
+      "loss": 1.5599,
+      "step": 1849
+    },
+    {
+      "epoch": 0.038959056137894,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.5752,
+      "step": 1850
+    },
+    {
+      "epoch": 0.03898011508715773,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0002,
+      "loss": 1.5825,
+      "step": 1851
+    },
+    {
+      "epoch": 0.039001174036421456,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.6291,
+      "step": 1852
+    },
+    {
+      "epoch": 0.039022232985685176,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.6348,
+      "step": 1853
+    },
+    {
+      "epoch": 0.039043291934948904,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.5951,
+      "step": 1854
+    },
+    {
+      "epoch": 0.03906435088421263,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0002,
+      "loss": 1.6028,
+      "step": 1855
+    },
+    {
+      "epoch": 0.03908540983347636,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.6309,
+      "step": 1856
+    },
+    {
+      "epoch": 0.039106468782740086,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.5804,
+      "step": 1857
+    },
+    {
+      "epoch": 0.039127527732003814,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0002,
+      "loss": 1.6422,
+      "step": 1858
+    },
+    {
+      "epoch": 0.03914858668126754,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6407,
+      "step": 1859
+    },
+    {
+      "epoch": 0.03916964563053126,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.5876,
+      "step": 1860
+    },
+    {
+      "epoch": 0.03919070457979499,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.6223,
+      "step": 1861
+    },
+    {
+      "epoch": 0.03921176352905872,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.5936,
+      "step": 1862
+    },
+    {
+      "epoch": 0.039232822478322445,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0002,
+      "loss": 1.5957,
+      "step": 1863
+    },
+    {
+      "epoch": 0.03925388142758617,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.0002,
+      "loss": 1.5882,
+      "step": 1864
+    },
+    {
+      "epoch": 0.0392749403768499,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.6099,
+      "step": 1865
+    },
+    {
+      "epoch": 0.03929599932611362,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0002,
+      "loss": 1.5659,
+      "step": 1866
+    },
+    {
+      "epoch": 0.03931705827537735,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.6221,
+      "step": 1867
+    },
+    {
+      "epoch": 0.039338117224641075,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0002,
+      "loss": 1.5818,
+      "step": 1868
+    },
+    {
+      "epoch": 0.0393591761739048,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6202,
+      "step": 1869
+    },
+    {
+      "epoch": 0.03938023512316853,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0002,
+      "loss": 1.58,
+      "step": 1870
+    },
+    {
+      "epoch": 0.03940129407243226,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0002,
+      "loss": 1.5926,
+      "step": 1871
+    },
+    {
+      "epoch": 0.039422353021695986,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.6076,
+      "step": 1872
+    },
+    {
+      "epoch": 0.039443411970959706,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0002,
+      "loss": 1.5995,
+      "step": 1873
+    },
+    {
+      "epoch": 0.039464470920223434,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6201,
+      "step": 1874
+    },
+    {
+      "epoch": 0.03948552986948716,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0002,
+      "loss": 1.6056,
+      "step": 1875
+    },
+    {
+      "epoch": 0.03950658881875089,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0002,
+      "loss": 1.5909,
+      "step": 1876
+    },
+    {
+      "epoch": 0.039527647768014616,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.6122,
+      "step": 1877
+    },
+    {
+      "epoch": 0.039548706717278344,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0002,
+      "loss": 1.5896,
+      "step": 1878
+    },
+    {
+      "epoch": 0.03956976566654207,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.6212,
+      "step": 1879
+    },
+    {
+      "epoch": 0.03959082461580579,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.5957,
+      "step": 1880
+    },
+    {
+      "epoch": 0.03961188356506952,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0002,
+      "loss": 1.5789,
+      "step": 1881
+    },
+    {
+      "epoch": 0.03963294251433325,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0002,
+      "loss": 1.6168,
+      "step": 1882
+    },
+    {
+      "epoch": 0.039654001463596975,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.5993,
+      "step": 1883
+    },
+    {
+      "epoch": 0.0396750604128607,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0002,
+      "loss": 1.5937,
+      "step": 1884
+    },
+    {
+      "epoch": 0.03969611936212443,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.5667,
+      "step": 1885
+    },
+    {
+      "epoch": 0.03971717831138815,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0002,
+      "loss": 1.6094,
+      "step": 1886
+    },
+    {
+      "epoch": 0.03973823726065188,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.6117,
+      "step": 1887
+    },
+    {
+      "epoch": 0.039759296209915605,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.5944,
+      "step": 1888
+    },
+    {
+      "epoch": 0.03978035515917933,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.6253,
+      "step": 1889
+    },
+    {
+      "epoch": 0.03980141410844306,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.0002,
+      "loss": 1.6042,
+      "step": 1890
+    },
+    {
+      "epoch": 0.03982247305770679,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.6088,
+      "step": 1891
+    },
+    {
+      "epoch": 0.039843532006970515,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0002,
+      "loss": 1.6036,
+      "step": 1892
+    },
+    {
+      "epoch": 0.039864590956234236,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0002,
+      "loss": 1.5838,
+      "step": 1893
+    },
+    {
+      "epoch": 0.039885649905497964,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6039,
+      "step": 1894
+    },
+    {
+      "epoch": 0.03990670885476169,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.6171,
+      "step": 1895
+    },
+    {
+      "epoch": 0.03992776780402542,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5895,
+      "step": 1896
+    },
+    {
+      "epoch": 0.039948826753289146,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.6371,
+      "step": 1897
+    },
+    {
+      "epoch": 0.039969885702552874,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.6087,
+      "step": 1898
+    },
+    {
+      "epoch": 0.039990944651816594,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0002,
+      "loss": 1.5916,
+      "step": 1899
+    },
+    {
+      "epoch": 0.04001200360108032,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.5966,
+      "step": 1900
+    },
+    {
+      "epoch": 0.04003306255034405,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002,
+      "loss": 1.6115,
+      "step": 1901
+    },
+    {
+      "epoch": 0.04005412149960778,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.6112,
+      "step": 1902
+    },
+    {
+      "epoch": 0.040075180448871504,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.5854,
+      "step": 1903
+    },
+    {
+      "epoch": 0.04009623939813523,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.5728,
+      "step": 1904
+    },
+    {
+      "epoch": 0.04011729834739896,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.6113,
+      "step": 1905
+    },
+    {
+      "epoch": 0.04013835729666268,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0002,
+      "loss": 1.6155,
+      "step": 1906
+    },
+    {
+      "epoch": 0.04015941624592641,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0002,
+      "loss": 1.6017,
+      "step": 1907
+    },
+    {
+      "epoch": 0.040180475195190135,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0002,
+      "loss": 1.5607,
+      "step": 1908
+    },
+    {
+      "epoch": 0.04020153414445386,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.5974,
+      "step": 1909
+    },
+    {
+      "epoch": 0.04022259309371759,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0002,
+      "loss": 1.5876,
+      "step": 1910
+    },
+    {
+      "epoch": 0.04024365204298132,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0002,
+      "loss": 1.6018,
+      "step": 1911
+    },
+    {
+      "epoch": 0.040264710992245045,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6007,
+      "step": 1912
+    },
+    {
+      "epoch": 0.040285769941508766,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0002,
+      "loss": 1.593,
+      "step": 1913
+    },
+    {
+      "epoch": 0.04030682889077249,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.6258,
+      "step": 1914
+    },
+    {
+      "epoch": 0.04032788784003622,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0002,
+      "loss": 1.5898,
+      "step": 1915
+    },
+    {
+      "epoch": 0.04034894678929995,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.607,
+      "step": 1916
+    },
+    {
+      "epoch": 0.040370005738563676,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.6196,
+      "step": 1917
+    },
+    {
+      "epoch": 0.040391064687827404,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0002,
+      "loss": 1.6182,
+      "step": 1918
+    },
+    {
+      "epoch": 0.040412123637091124,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.6237,
+      "step": 1919
+    },
+    {
+      "epoch": 0.04043318258635485,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.6046,
+      "step": 1920
+    },
+    {
+      "epoch": 0.04045424153561858,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.565,
+      "step": 1921
+    },
+    {
+      "epoch": 0.04047530048488231,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.5871,
+      "step": 1922
+    },
+    {
+      "epoch": 0.040496359434146034,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0002,
+      "loss": 1.6143,
+      "step": 1923
+    },
+    {
+      "epoch": 0.04051741838340976,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0002,
+      "loss": 1.5711,
+      "step": 1924
+    },
+    {
+      "epoch": 0.04053847733267349,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0002,
+      "loss": 1.6167,
+      "step": 1925
+    },
+    {
+      "epoch": 0.04055953628193721,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0002,
+      "loss": 1.5939,
+      "step": 1926
+    },
+    {
+      "epoch": 0.04058059523120094,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.5911,
+      "step": 1927
+    },
+    {
+      "epoch": 0.040601654180464665,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.6079,
+      "step": 1928
+    },
+    {
+      "epoch": 0.04062271312972839,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5844,
+      "step": 1929
+    },
+    {
+      "epoch": 0.04064377207899212,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.6206,
+      "step": 1930
+    },
+    {
+      "epoch": 0.04066483102825585,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.5936,
+      "step": 1931
+    },
+    {
+      "epoch": 0.040685889977519575,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.5745,
+      "step": 1932
+    },
+    {
+      "epoch": 0.040706948926783296,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0002,
+      "loss": 1.5739,
+      "step": 1933
+    },
+    {
+      "epoch": 0.04072800787604702,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.5967,
+      "step": 1934
+    },
+    {
+      "epoch": 0.04074906682531075,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002,
+      "loss": 1.5997,
+      "step": 1935
+    },
+    {
+      "epoch": 0.04077012577457448,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.576,
+      "step": 1936
+    },
+    {
+      "epoch": 0.040791184723838206,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.6067,
+      "step": 1937
+    },
+    {
+      "epoch": 0.04081224367310193,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002,
+      "loss": 1.5923,
+      "step": 1938
+    },
+    {
+      "epoch": 0.040833302622365654,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.5872,
+      "step": 1939
+    },
+    {
+      "epoch": 0.04085436157162938,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.5733,
+      "step": 1940
+    },
+    {
+      "epoch": 0.04087542052089311,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0002,
+      "loss": 1.6116,
+      "step": 1941
+    },
+    {
+      "epoch": 0.04089647947015684,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6228,
+      "step": 1942
+    },
+    {
+      "epoch": 0.040917538419420564,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5874,
+      "step": 1943
+    },
+    {
+      "epoch": 0.04093859736868429,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0002,
+      "loss": 1.6016,
+      "step": 1944
+    },
+    {
+      "epoch": 0.04095965631794802,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.587,
+      "step": 1945
+    },
+    {
+      "epoch": 0.04098071526721174,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5799,
+      "step": 1946
+    },
+    {
+      "epoch": 0.04100177421647547,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.5829,
+      "step": 1947
+    },
+    {
+      "epoch": 0.041022833165739195,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6216,
+      "step": 1948
+    },
+    {
+      "epoch": 0.04104389211500292,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0002,
+      "loss": 1.6079,
+      "step": 1949
+    },
+    {
+      "epoch": 0.04106495106426665,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0002,
+      "loss": 1.5933,
+      "step": 1950
+    },
+    {
+      "epoch": 0.04108601001353038,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0002,
+      "loss": 1.5821,
+      "step": 1951
+    },
+    {
+      "epoch": 0.0411070689627941,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.584,
+      "step": 1952
+    },
+    {
+      "epoch": 0.041128127912057826,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0002,
+      "loss": 1.5867,
+      "step": 1953
+    },
+    {
+      "epoch": 0.04114918686132155,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0002,
+      "loss": 1.617,
+      "step": 1954
+    },
+    {
+      "epoch": 0.04117024581058528,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.6158,
+      "step": 1955
+    },
+    {
+      "epoch": 0.04119130475984901,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0002,
+      "loss": 1.5998,
+      "step": 1956
+    },
+    {
+      "epoch": 0.041212363709112736,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0002,
+      "loss": 1.6092,
+      "step": 1957
+    },
+    {
+      "epoch": 0.04123342265837646,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0002,
+      "loss": 1.6213,
+      "step": 1958
+    },
+    {
+      "epoch": 0.041254481607640184,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0002,
+      "loss": 1.5695,
+      "step": 1959
+    },
+    {
+      "epoch": 0.04127554055690391,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0002,
+      "loss": 1.6084,
+      "step": 1960
+    },
+    {
+      "epoch": 0.04129659950616764,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0002,
+      "loss": 1.5859,
+      "step": 1961
+    },
+    {
+      "epoch": 0.041317658455431366,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.6044,
+      "step": 1962
+    },
+    {
+      "epoch": 0.041338717404695094,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0002,
+      "loss": 1.6034,
+      "step": 1963
+    },
+    {
+      "epoch": 0.04135977635395882,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.5724,
+      "step": 1964
+    },
+    {
+      "epoch": 0.04138083530322255,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0002,
+      "loss": 1.5941,
+      "step": 1965
+    },
+    {
+      "epoch": 0.04140189425248627,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.5908,
+      "step": 1966
+    },
+    {
+      "epoch": 0.04142295320175,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.6162,
+      "step": 1967
+    },
+    {
+      "epoch": 0.041444012151013725,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.6222,
+      "step": 1968
+    },
+    {
+      "epoch": 0.04146507110027745,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0002,
+      "loss": 1.6013,
+      "step": 1969
+    },
+    {
+      "epoch": 0.04148613004954118,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.5867,
+      "step": 1970
+    },
+    {
+      "epoch": 0.04150718899880491,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.6077,
+      "step": 1971
+    },
+    {
+      "epoch": 0.04152824794806863,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0002,
+      "loss": 1.5653,
+      "step": 1972
+    },
+    {
+      "epoch": 0.041549306897332355,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6137,
+      "step": 1973
+    },
+    {
+      "epoch": 0.04157036584659608,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0002,
+      "loss": 1.6051,
+      "step": 1974
+    },
+    {
+      "epoch": 0.04159142479585981,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.5789,
+      "step": 1975
+    },
+    {
+      "epoch": 0.04161248374512354,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6153,
+      "step": 1976
+    },
+    {
+      "epoch": 0.041633542694387266,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.5798,
+      "step": 1977
+    },
+    {
+      "epoch": 0.04165460164365099,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.6105,
+      "step": 1978
+    },
+    {
+      "epoch": 0.041675660592914714,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.5493,
+      "step": 1979
+    },
+    {
+      "epoch": 0.04169671954217844,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0002,
+      "loss": 1.5896,
+      "step": 1980
+    },
+    {
+      "epoch": 0.04171777849144217,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.5841,
+      "step": 1981
+    },
+    {
+      "epoch": 0.041738837440705896,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.5943,
+      "step": 1982
+    },
+    {
+      "epoch": 0.041759896389969624,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0002,
+      "loss": 1.612,
+      "step": 1983
+    },
+    {
+      "epoch": 0.04178095533923335,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.6343,
+      "step": 1984
+    },
+    {
+      "epoch": 0.04180201428849708,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6086,
+      "step": 1985
+    },
+    {
+      "epoch": 0.0418230732377608,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.6427,
+      "step": 1986
+    },
+    {
+      "epoch": 0.04184413218702453,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0002,
+      "loss": 1.5919,
+      "step": 1987
+    },
+    {
+      "epoch": 0.041865191136288255,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.601,
+      "step": 1988
+    },
+    {
+      "epoch": 0.04188625008555198,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002,
+      "loss": 1.5932,
+      "step": 1989
+    },
+    {
+      "epoch": 0.04190730903481571,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.5902,
+      "step": 1990
+    },
+    {
+      "epoch": 0.04192836798407944,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.0002,
+      "loss": 1.6053,
+      "step": 1991
+    },
+    {
+      "epoch": 0.04194942693334316,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.5728,
+      "step": 1992
+    },
+    {
+      "epoch": 0.041970485882606885,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0002,
+      "loss": 1.5954,
+      "step": 1993
+    },
+    {
+      "epoch": 0.04199154483187061,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5832,
+      "step": 1994
+    },
+    {
+      "epoch": 0.04201260378113434,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.5696,
+      "step": 1995
+    },
+    {
+      "epoch": 0.04203366273039807,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0002,
+      "loss": 1.6157,
+      "step": 1996
+    },
+    {
+      "epoch": 0.042054721679661795,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0002,
+      "loss": 1.5868,
+      "step": 1997
+    },
+    {
+      "epoch": 0.04207578062892552,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.6228,
+      "step": 1998
+    },
+    {
+      "epoch": 0.042096839578189243,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.5957,
+      "step": 1999
+    },
+    {
+      "epoch": 0.04211789852745297,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.5743,
+      "step": 2000
+    },
+    {
+      "epoch": 0.04211789852745297,
+      "eval_loss": 1.7890796661376953,
+      "eval_runtime": 898.0231,
+      "eval_samples_per_second": 68.818,
+      "eval_steps_per_second": 2.151,
+      "step": 2000
+    },
+    {
+      "epoch": 0.0421389574767167,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5621,
+      "step": 2001
+    },
+    {
+      "epoch": 0.042160016425980426,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.5761,
+      "step": 2002
+    },
+    {
+      "epoch": 0.042181075375244154,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0002,
+      "loss": 1.6679,
+      "step": 2003
+    },
+    {
+      "epoch": 0.04220213432450788,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.5733,
+      "step": 2004
+    },
+    {
+      "epoch": 0.0422231932737716,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.6052,
+      "step": 2005
+    },
+    {
+      "epoch": 0.04224425222303533,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.5922,
+      "step": 2006
+    },
+    {
+      "epoch": 0.04226531117229906,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0002,
+      "loss": 1.6025,
+      "step": 2007
+    },
+    {
+      "epoch": 0.042286370121562784,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.5753,
+      "step": 2008
+    },
+    {
+      "epoch": 0.04230742907082651,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0002,
+      "loss": 1.574,
+      "step": 2009
+    },
+    {
+      "epoch": 0.04232848802009024,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0002,
+      "loss": 1.6063,
+      "step": 2010
+    },
+    {
+      "epoch": 0.04234954696935397,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0002,
+      "loss": 1.5711,
+      "step": 2011
+    },
+    {
+      "epoch": 0.04237060591861769,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.5817,
+      "step": 2012
+    },
+    {
+      "epoch": 0.042391664867881415,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0002,
+      "loss": 1.5955,
+      "step": 2013
+    },
+    {
+      "epoch": 0.04241272381714514,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.59,
+      "step": 2014
+    },
+    {
+      "epoch": 0.04243378276640887,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002,
+      "loss": 1.6189,
+      "step": 2015
+    },
+    {
+      "epoch": 0.0424548417156726,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5851,
+      "step": 2016
+    },
+    {
+      "epoch": 0.042475900664936325,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0002,
+      "loss": 1.6274,
+      "step": 2017
+    },
+    {
+      "epoch": 0.04249695961420005,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.6306,
+      "step": 2018
+    },
+    {
+      "epoch": 0.04251801856346377,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0002,
+      "loss": 1.5685,
+      "step": 2019
+    },
+    {
+      "epoch": 0.0425390775127275,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.6,
+      "step": 2020
+    },
+    {
+      "epoch": 0.04256013646199123,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.5981,
+      "step": 2021
+    },
+    {
+      "epoch": 0.042581195411254956,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002,
+      "loss": 1.6069,
+      "step": 2022
+    },
+    {
+      "epoch": 0.042602254360518683,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.5855,
+      "step": 2023
+    },
+    {
+      "epoch": 0.04262331330978241,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.6244,
+      "step": 2024
+    },
+    {
+      "epoch": 0.04264437225904613,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.611,
+      "step": 2025
+    },
+    {
+      "epoch": 0.04266543120830986,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.5772,
+      "step": 2026
+    },
+    {
+      "epoch": 0.04268649015757359,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.5933,
+      "step": 2027
+    },
+    {
+      "epoch": 0.042707549106837314,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.5734,
+      "step": 2028
+    },
+    {
+      "epoch": 0.04272860805610104,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0002,
+      "loss": 1.5894,
+      "step": 2029
+    },
+    {
+      "epoch": 0.04274966700536477,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5752,
+      "step": 2030
+    },
+    {
+      "epoch": 0.0427707259546285,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0002,
+      "loss": 1.6598,
+      "step": 2031
+    },
+    {
+      "epoch": 0.04279178490389222,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0002,
+      "loss": 1.5786,
+      "step": 2032
+    },
+    {
+      "epoch": 0.042812843853155945,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.5795,
+      "step": 2033
+    },
+    {
+      "epoch": 0.04283390280241967,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6147,
+      "step": 2034
+    },
+    {
+      "epoch": 0.0428549617516834,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6133,
+      "step": 2035
+    },
+    {
+      "epoch": 0.04287602070094713,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.5375,
+      "step": 2036
+    },
+    {
+      "epoch": 0.042897079650210855,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5853,
+      "step": 2037
+    },
+    {
+      "epoch": 0.04291813859947458,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.0002,
+      "loss": 1.5774,
+      "step": 2038
+    },
+    {
+      "epoch": 0.0429391975487383,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0002,
+      "loss": 1.626,
+      "step": 2039
+    },
+    {
+      "epoch": 0.04296025649800203,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0002,
+      "loss": 1.595,
+      "step": 2040
+    },
+    {
+      "epoch": 0.04298131544726576,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0002,
+      "loss": 1.5977,
+      "step": 2041
+    },
+    {
+      "epoch": 0.043002374396529486,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.5909,
+      "step": 2042
+    },
+    {
+      "epoch": 0.04302343334579321,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.6905,
+      "step": 2043
+    },
+    {
+      "epoch": 0.04304449229505694,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0002,
+      "loss": 1.5851,
+      "step": 2044
+    },
+    {
+      "epoch": 0.04306555124432066,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6042,
+      "step": 2045
+    },
+    {
+      "epoch": 0.04308661019358439,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.5979,
+      "step": 2046
+    },
+    {
+      "epoch": 0.043107669142848117,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.5956,
+      "step": 2047
+    },
+    {
+      "epoch": 0.043128728092111844,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.6312,
+      "step": 2048
+    },
+    {
+      "epoch": 0.04314978704137557,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.561,
+      "step": 2049
+    },
+    {
+      "epoch": 0.0431708459906393,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.5861,
+      "step": 2050
+    },
+    {
+      "epoch": 0.04319190493990303,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0002,
+      "loss": 1.5792,
+      "step": 2051
+    },
+    {
+      "epoch": 0.04321296388916675,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0002,
+      "loss": 1.5846,
+      "step": 2052
+    },
+    {
+      "epoch": 0.043234022838430475,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0002,
+      "loss": 1.5737,
+      "step": 2053
+    },
+    {
+      "epoch": 0.0432550817876942,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6068,
+      "step": 2054
+    },
+    {
+      "epoch": 0.04327614073695793,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.5881,
+      "step": 2055
+    },
+    {
+      "epoch": 0.04329719968622166,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.6127,
+      "step": 2056
+    },
+    {
+      "epoch": 0.043318258635485385,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.5723,
+      "step": 2057
+    },
+    {
+      "epoch": 0.043339317584749106,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.6329,
+      "step": 2058
+    },
+    {
+      "epoch": 0.04336037653401283,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0002,
+      "loss": 1.6023,
+      "step": 2059
+    },
+    {
+      "epoch": 0.04338143548327656,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002,
+      "loss": 1.5703,
+      "step": 2060
+    },
+    {
+      "epoch": 0.04340249443254029,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.552,
+      "step": 2061
+    },
+    {
+      "epoch": 0.043423553381804016,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.6157,
+      "step": 2062
+    },
+    {
+      "epoch": 0.04344461233106774,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0002,
+      "loss": 1.6175,
+      "step": 2063
+    },
+    {
+      "epoch": 0.04346567128033147,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.0002,
+      "loss": 1.5873,
+      "step": 2064
+    },
+    {
+      "epoch": 0.04348673022959519,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.5981,
+      "step": 2065
+    },
+    {
+      "epoch": 0.04350778917885892,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0002,
+      "loss": 1.618,
+      "step": 2066
+    },
+    {
+      "epoch": 0.043528848128122646,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0002,
+      "loss": 1.5708,
+      "step": 2067
+    },
+    {
+      "epoch": 0.043549907077386374,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.59,
+      "step": 2068
+    },
+    {
+      "epoch": 0.0435709660266501,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.6023,
+      "step": 2069
+    },
+    {
+      "epoch": 0.04359202497591383,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0002,
+      "loss": 1.5976,
+      "step": 2070
+    },
+    {
+      "epoch": 0.043613083925177556,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002,
+      "loss": 1.5735,
+      "step": 2071
+    },
+    {
+      "epoch": 0.04363414287444128,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0002,
+      "loss": 1.6009,
+      "step": 2072
+    },
+    {
+      "epoch": 0.043655201823705005,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.6065,
+      "step": 2073
+    },
+    {
+      "epoch": 0.04367626077296873,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.6232,
+      "step": 2074
+    },
+    {
+      "epoch": 0.04369731972223246,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.5797,
+      "step": 2075
+    },
+    {
+      "epoch": 0.04371837867149619,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.5828,
+      "step": 2076
+    },
+    {
+      "epoch": 0.043739437620759915,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0002,
+      "loss": 1.5933,
+      "step": 2077
+    },
+    {
+      "epoch": 0.043760496570023635,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5795,
+      "step": 2078
+    },
+    {
+      "epoch": 0.04378155551928736,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.6197,
+      "step": 2079
+    },
+    {
+      "epoch": 0.04380261446855109,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.6049,
+      "step": 2080
+    },
+    {
+      "epoch": 0.04382367341781482,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.6032,
+      "step": 2081
+    },
+    {
+      "epoch": 0.043844732367078545,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.6002,
+      "step": 2082
+    },
+    {
+      "epoch": 0.04386579131634227,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0002,
+      "loss": 1.6031,
+      "step": 2083
+    },
+    {
+      "epoch": 0.043886850265606,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0002,
+      "loss": 1.5946,
+      "step": 2084
+    },
+    {
+      "epoch": 0.04390790921486972,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.5992,
+      "step": 2085
+    },
+    {
+      "epoch": 0.04392896816413345,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.5925,
+      "step": 2086
+    },
+    {
+      "epoch": 0.043950027113397176,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.5633,
+      "step": 2087
+    },
+    {
+      "epoch": 0.043971086062660904,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0002,
+      "loss": 1.5859,
+      "step": 2088
+    },
+    {
+      "epoch": 0.04399214501192463,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0002,
+      "loss": 1.5983,
+      "step": 2089
+    },
+    {
+      "epoch": 0.04401320396118836,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.6131,
+      "step": 2090
+    },
+    {
+      "epoch": 0.044034262910452086,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.5873,
+      "step": 2091
+    },
+    {
+      "epoch": 0.04405532185971581,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0002,
+      "loss": 1.6104,
+      "step": 2092
+    },
+    {
+      "epoch": 0.044076380808979534,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0002,
+      "loss": 1.6049,
+      "step": 2093
+    },
+    {
+      "epoch": 0.04409743975824326,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6067,
+      "step": 2094
+    },
+    {
+      "epoch": 0.04411849870750699,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5931,
+      "step": 2095
+    },
+    {
+      "epoch": 0.04413955765677072,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.5595,
+      "step": 2096
+    },
+    {
+      "epoch": 0.044160616606034445,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.5983,
+      "step": 2097
+    },
+    {
+      "epoch": 0.044181675555298165,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5902,
+      "step": 2098
+    },
+    {
+      "epoch": 0.04420273450456189,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0002,
+      "loss": 1.5605,
+      "step": 2099
+    },
+    {
+      "epoch": 0.04422379345382562,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.6205,
+      "step": 2100
+    },
+    {
+      "epoch": 0.04424485240308935,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.5896,
+      "step": 2101
+    },
+    {
+      "epoch": 0.044265911352353075,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0002,
+      "loss": 1.5778,
+      "step": 2102
+    },
+    {
+      "epoch": 0.0442869703016168,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.5794,
+      "step": 2103
+    },
+    {
+      "epoch": 0.04430802925088053,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.5911,
+      "step": 2104
+    },
+    {
+      "epoch": 0.04432908820014425,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0002,
+      "loss": 1.5852,
+      "step": 2105
+    },
+    {
+      "epoch": 0.04435014714940798,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.5999,
+      "step": 2106
+    },
+    {
+      "epoch": 0.044371206098671706,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.615,
+      "step": 2107
+    },
+    {
+      "epoch": 0.044392265047935434,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0002,
+      "loss": 1.6074,
+      "step": 2108
+    },
+    {
+      "epoch": 0.04441332399719916,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.5793,
+      "step": 2109
+    },
+    {
+      "epoch": 0.04443438294646289,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0002,
+      "loss": 1.575,
+      "step": 2110
+    },
+    {
+      "epoch": 0.04445544189572661,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.6551,
+      "step": 2111
+    },
+    {
+      "epoch": 0.04447650084499034,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.6081,
+      "step": 2112
+    },
+    {
+      "epoch": 0.044497559794254064,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0002,
+      "loss": 1.6188,
+      "step": 2113
+    },
+    {
+      "epoch": 0.04451861874351779,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0002,
+      "loss": 1.6099,
+      "step": 2114
+    },
+    {
+      "epoch": 0.04453967769278152,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.5805,
+      "step": 2115
+    },
+    {
+      "epoch": 0.04456073664204525,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0002,
+      "loss": 1.6216,
+      "step": 2116
+    },
+    {
+      "epoch": 0.044581795591308974,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0002,
+      "loss": 1.5917,
+      "step": 2117
+    },
+    {
+      "epoch": 0.044602854540572695,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.593,
+      "step": 2118
+    },
+    {
+      "epoch": 0.04462391348983642,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.0002,
+      "loss": 1.6089,
+      "step": 2119
+    },
+    {
+      "epoch": 0.04464497243910015,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.627,
+      "step": 2120
+    },
+    {
+      "epoch": 0.04466603138836388,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6215,
+      "step": 2121
+    },
+    {
+      "epoch": 0.044687090337627605,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.6516,
+      "step": 2122
+    },
+    {
+      "epoch": 0.04470814928689133,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6005,
+      "step": 2123
+    },
+    {
+      "epoch": 0.04472920823615506,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.6134,
+      "step": 2124
+    },
+    {
+      "epoch": 0.04475026718541878,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.6151,
+      "step": 2125
+    },
+    {
+      "epoch": 0.04477132613468251,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6272,
+      "step": 2126
+    },
+    {
+      "epoch": 0.044792385083946236,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.5782,
+      "step": 2127
+    },
+    {
+      "epoch": 0.04481344403320996,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0002,
+      "loss": 1.5806,
+      "step": 2128
+    },
+    {
+      "epoch": 0.04483450298247369,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0002,
+      "loss": 1.5827,
+      "step": 2129
+    },
+    {
+      "epoch": 0.04485556193173742,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.61,
+      "step": 2130
+    },
+    {
+      "epoch": 0.04487662088100114,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0002,
+      "loss": 1.6023,
+      "step": 2131
+    },
+    {
+      "epoch": 0.04489767983026487,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0002,
+      "loss": 1.6042,
+      "step": 2132
+    },
+    {
+      "epoch": 0.044918738779528594,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0002,
+      "loss": 1.5673,
+      "step": 2133
+    },
+    {
+      "epoch": 0.04493979772879232,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0002,
+      "loss": 1.5994,
+      "step": 2134
+    },
+    {
+      "epoch": 0.04496085667805605,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0002,
+      "loss": 1.5844,
+      "step": 2135
+    },
+    {
+      "epoch": 0.04498191562731978,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.579,
+      "step": 2136
+    },
+    {
+      "epoch": 0.045002974576583504,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0002,
+      "loss": 1.6008,
+      "step": 2137
+    },
+    {
+      "epoch": 0.045024033525847225,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.603,
+      "step": 2138
+    },
+    {
+      "epoch": 0.04504509247511095,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5966,
+      "step": 2139
+    },
+    {
+      "epoch": 0.04506615142437468,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6106,
+      "step": 2140
+    },
+    {
+      "epoch": 0.04508721037363841,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.5794,
+      "step": 2141
+    },
+    {
+      "epoch": 0.045108269322902135,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0002,
+      "loss": 1.5868,
+      "step": 2142
+    },
+    {
+      "epoch": 0.04512932827216586,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.6446,
+      "step": 2143
+    },
+    {
+      "epoch": 0.04515038722142959,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.6136,
+      "step": 2144
+    },
+    {
+      "epoch": 0.04517144617069331,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0002,
+      "loss": 1.6259,
+      "step": 2145
+    },
+    {
+      "epoch": 0.04519250511995704,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.6116,
+      "step": 2146
+    },
+    {
+      "epoch": 0.045213564069220766,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0002,
+      "loss": 1.5756,
+      "step": 2147
+    },
+    {
+      "epoch": 0.04523462301848449,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.6134,
+      "step": 2148
+    },
+    {
+      "epoch": 0.04525568196774822,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5883,
+      "step": 2149
+    },
+    {
+      "epoch": 0.04527674091701195,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5584,
+      "step": 2150
+    },
+    {
+      "epoch": 0.04529779986627567,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.5829,
+      "step": 2151
+    },
+    {
+      "epoch": 0.045318858815539396,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.5994,
+      "step": 2152
+    },
+    {
+      "epoch": 0.045339917764803124,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.6224,
+      "step": 2153
+    },
+    {
+      "epoch": 0.04536097671406685,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.5991,
+      "step": 2154
+    },
+    {
+      "epoch": 0.04538203566333058,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0002,
+      "loss": 1.6016,
+      "step": 2155
+    },
+    {
+      "epoch": 0.04540309461259431,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.5804,
+      "step": 2156
+    },
+    {
+      "epoch": 0.045424153561858034,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.5816,
+      "step": 2157
+    },
+    {
+      "epoch": 0.045445212511121755,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.6024,
+      "step": 2158
+    },
+    {
+      "epoch": 0.04546627146038548,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0002,
+      "loss": 1.5725,
+      "step": 2159
+    },
+    {
+      "epoch": 0.04548733040964921,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5904,
+      "step": 2160
+    },
+    {
+      "epoch": 0.04550838935891294,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0002,
+      "loss": 1.5804,
+      "step": 2161
+    },
+    {
+      "epoch": 0.045529448308176665,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.5867,
+      "step": 2162
+    },
+    {
+      "epoch": 0.04555050725744039,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.0002,
+      "loss": 1.6093,
+      "step": 2163
+    },
+    {
+      "epoch": 0.04557156620670412,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.5837,
+      "step": 2164
+    },
+    {
+      "epoch": 0.04559262515596784,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.5825,
+      "step": 2165
+    },
+    {
+      "epoch": 0.04561368410523157,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0002,
+      "loss": 1.6057,
+      "step": 2166
+    },
+    {
+      "epoch": 0.045634743054495296,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.559,
+      "step": 2167
+    },
+    {
+      "epoch": 0.04565580200375902,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.5964,
+      "step": 2168
+    },
+    {
+      "epoch": 0.04567686095302275,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.6379,
+      "step": 2169
+    },
+    {
+      "epoch": 0.04569791990228648,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.6101,
+      "step": 2170
+    },
+    {
+      "epoch": 0.0457189788515502,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5986,
+      "step": 2171
+    },
+    {
+      "epoch": 0.045740037800813926,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0002,
+      "loss": 1.5735,
+      "step": 2172
+    },
+    {
+      "epoch": 0.045761096750077654,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.5861,
+      "step": 2173
+    },
+    {
+      "epoch": 0.04578215569934138,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0002,
+      "loss": 1.5571,
+      "step": 2174
+    },
+    {
+      "epoch": 0.04580321464860511,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0002,
+      "loss": 1.6198,
+      "step": 2175
+    },
+    {
+      "epoch": 0.045824273597868836,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.5953,
+      "step": 2176
+    },
+    {
+      "epoch": 0.045845332547132564,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.6039,
+      "step": 2177
+    },
+    {
+      "epoch": 0.045866391496396285,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.6041,
+      "step": 2178
+    },
+    {
+      "epoch": 0.04588745044566001,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.6119,
+      "step": 2179
+    },
+    {
+      "epoch": 0.04590850939492374,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5567,
+      "step": 2180
+    },
+    {
+      "epoch": 0.04592956834418747,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.5945,
+      "step": 2181
+    },
+    {
+      "epoch": 0.045950627293451195,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0002,
+      "loss": 1.5728,
+      "step": 2182
+    },
+    {
+      "epoch": 0.04597168624271492,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0002,
+      "loss": 1.6189,
+      "step": 2183
+    },
+    {
+      "epoch": 0.04599274519197864,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.6273,
+      "step": 2184
+    },
+    {
+      "epoch": 0.04601380414124237,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0002,
+      "loss": 1.5951,
+      "step": 2185
+    },
+    {
+      "epoch": 0.0460348630905061,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.5727,
+      "step": 2186
+    },
+    {
+      "epoch": 0.046055922039769825,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5695,
+      "step": 2187
+    },
+    {
+      "epoch": 0.04607698098903355,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0002,
+      "loss": 1.5949,
+      "step": 2188
+    },
+    {
+      "epoch": 0.04609803993829728,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5592,
+      "step": 2189
+    },
+    {
+      "epoch": 0.04611909888756101,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.0002,
+      "loss": 1.5942,
+      "step": 2190
+    },
+    {
+      "epoch": 0.04614015783682473,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6061,
+      "step": 2191
+    },
+    {
+      "epoch": 0.046161216786088456,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.0002,
+      "loss": 1.5963,
+      "step": 2192
+    },
+    {
+      "epoch": 0.046182275735352184,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.0002,
+      "loss": 1.5956,
+      "step": 2193
+    },
+    {
+      "epoch": 0.04620333468461591,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0002,
+      "loss": 1.6178,
+      "step": 2194
+    },
+    {
+      "epoch": 0.04622439363387964,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.5679,
+      "step": 2195
+    },
+    {
+      "epoch": 0.046245452583143366,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5601,
+      "step": 2196
+    },
+    {
+      "epoch": 0.046266511532407094,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0002,
+      "loss": 1.584,
+      "step": 2197
+    },
+    {
+      "epoch": 0.046287570481670814,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.5911,
+      "step": 2198
+    },
+    {
+      "epoch": 0.04630862943093454,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0002,
+      "loss": 1.5917,
+      "step": 2199
+    },
+    {
+      "epoch": 0.04632968838019827,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0002,
+      "loss": 1.6175,
+      "step": 2200
+    },
+    {
+      "epoch": 0.046350747329462,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.6204,
+      "step": 2201
+    },
+    {
+      "epoch": 0.046371806278725725,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.6096,
+      "step": 2202
+    },
+    {
+      "epoch": 0.04639286522798945,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6021,
+      "step": 2203
+    },
+    {
+      "epoch": 0.04641392417725317,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.0002,
+      "loss": 1.6191,
+      "step": 2204
+    },
+    {
+      "epoch": 0.0464349831265169,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5732,
+      "step": 2205
+    },
+    {
+      "epoch": 0.04645604207578063,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.6043,
+      "step": 2206
+    },
+    {
+      "epoch": 0.046477101025044355,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002,
+      "loss": 1.6177,
+      "step": 2207
+    },
+    {
+      "epoch": 0.04649815997430808,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6073,
+      "step": 2208
+    },
+    {
+      "epoch": 0.04651921892357181,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.5933,
+      "step": 2209
+    },
+    {
+      "epoch": 0.04654027787283554,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.0002,
+      "loss": 1.5746,
+      "step": 2210
+    },
+    {
+      "epoch": 0.04656133682209926,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0002,
+      "loss": 1.5902,
+      "step": 2211
+    },
+    {
+      "epoch": 0.046582395771362986,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.6061,
+      "step": 2212
+    },
+    {
+      "epoch": 0.046603454720626714,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.5973,
+      "step": 2213
+    },
+    {
+      "epoch": 0.04662451366989044,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0002,
+      "loss": 1.6374,
+      "step": 2214
+    },
+    {
+      "epoch": 0.04664557261915417,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5846,
+      "step": 2215
+    },
+    {
+      "epoch": 0.046666631568417896,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.5722,
+      "step": 2216
+    },
+    {
+      "epoch": 0.046687690517681624,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0002,
+      "loss": 1.6209,
+      "step": 2217
+    },
+    {
+      "epoch": 0.046708749466945344,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.5989,
+      "step": 2218
+    },
+    {
+      "epoch": 0.04672980841620907,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.5628,
+      "step": 2219
+    },
+    {
+      "epoch": 0.0467508673654728,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.0002,
+      "loss": 1.628,
+      "step": 2220
+    },
+    {
+      "epoch": 0.04677192631473653,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.579,
+      "step": 2221
+    },
+    {
+      "epoch": 0.046792985264000254,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0002,
+      "loss": 1.5852,
+      "step": 2222
+    },
+    {
+      "epoch": 0.04681404421326398,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.5906,
+      "step": 2223
+    },
+    {
+      "epoch": 0.0468351031625277,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0002,
+      "loss": 1.5768,
+      "step": 2224
+    },
+    {
+      "epoch": 0.04685616211179143,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.5892,
+      "step": 2225
+    },
+    {
+      "epoch": 0.04687722106105516,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0002,
+      "loss": 1.5862,
+      "step": 2226
+    },
+    {
+      "epoch": 0.046898280010318885,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0002,
+      "loss": 1.6256,
+      "step": 2227
+    },
+    {
+      "epoch": 0.04691933895958261,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.5428,
+      "step": 2228
+    },
+    {
+      "epoch": 0.04694039790884634,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.5973,
+      "step": 2229
+    },
+    {
+      "epoch": 0.04696145685811007,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.6115,
+      "step": 2230
+    },
+    {
+      "epoch": 0.04698251580737379,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5907,
+      "step": 2231
+    },
+    {
+      "epoch": 0.047003574756637516,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.5975,
+      "step": 2232
+    },
+    {
+      "epoch": 0.04702463370590124,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.5617,
+      "step": 2233
+    },
+    {
+      "epoch": 0.04704569265516497,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5845,
+      "step": 2234
+    },
+    {
+      "epoch": 0.0470667516044287,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5938,
+      "step": 2235
+    },
+    {
+      "epoch": 0.047087810553692426,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.617,
+      "step": 2236
+    },
+    {
+      "epoch": 0.04710886950295615,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5549,
+      "step": 2237
+    },
+    {
+      "epoch": 0.047129928452219874,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.5941,
+      "step": 2238
+    },
+    {
+      "epoch": 0.0471509874014836,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0002,
+      "loss": 1.6139,
+      "step": 2239
+    },
+    {
+      "epoch": 0.04717204635074733,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0002,
+      "loss": 1.5802,
+      "step": 2240
+    },
+    {
+      "epoch": 0.04719310530001106,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0002,
+      "loss": 1.5661,
+      "step": 2241
+    },
+    {
+      "epoch": 0.047214164249274784,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0002,
+      "loss": 1.5734,
+      "step": 2242
+    },
+    {
+      "epoch": 0.04723522319853851,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5963,
+      "step": 2243
+    },
+    {
+      "epoch": 0.04725628214780223,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002,
+      "loss": 1.6141,
+      "step": 2244
+    },
+    {
+      "epoch": 0.04727734109706596,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0002,
+      "loss": 1.5899,
+      "step": 2245
+    },
+    {
+      "epoch": 0.04729840004632969,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.5898,
+      "step": 2246
+    },
+    {
+      "epoch": 0.047319458995593415,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0002,
+      "loss": 1.6051,
+      "step": 2247
+    },
+    {
+      "epoch": 0.04734051794485714,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0002,
+      "loss": 1.5814,
+      "step": 2248
+    },
+    {
+      "epoch": 0.04736157689412087,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.604,
+      "step": 2249
+    },
+    {
+      "epoch": 0.0473826358433846,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.5657,
+      "step": 2250
+    },
+    {
+      "epoch": 0.04740369479264832,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.643,
+      "step": 2251
+    },
+    {
+      "epoch": 0.047424753741912046,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0002,
+      "loss": 1.5994,
+      "step": 2252
+    },
+    {
+      "epoch": 0.04744581269117577,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0002,
+      "loss": 1.5882,
+      "step": 2253
+    },
+    {
+      "epoch": 0.0474668716404395,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.605,
+      "step": 2254
+    },
+    {
+      "epoch": 0.04748793058970323,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.593,
+      "step": 2255
+    },
+    {
+      "epoch": 0.047508989538966956,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6179,
+      "step": 2256
+    },
+    {
+      "epoch": 0.047530048488230676,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0002,
+      "loss": 1.6007,
+      "step": 2257
+    },
+    {
+      "epoch": 0.047551107437494404,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0002,
+      "loss": 1.587,
+      "step": 2258
+    },
+    {
+      "epoch": 0.04757216638675813,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.6399,
+      "step": 2259
+    },
+    {
+      "epoch": 0.04759322533602186,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.5978,
+      "step": 2260
+    },
+    {
+      "epoch": 0.04761428428528559,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0002,
+      "loss": 1.5981,
+      "step": 2261
+    },
+    {
+      "epoch": 0.047635343234549314,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.6017,
+      "step": 2262
+    },
+    {
+      "epoch": 0.04765640218381304,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0002,
+      "loss": 1.5893,
+      "step": 2263
+    },
+    {
+      "epoch": 0.04767746113307676,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0002,
+      "loss": 1.597,
+      "step": 2264
+    },
+    {
+      "epoch": 0.04769852008234049,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.0002,
+      "loss": 1.6333,
+      "step": 2265
+    },
+    {
+      "epoch": 0.04771957903160422,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.0002,
+      "loss": 1.6663,
+      "step": 2266
+    },
+    {
+      "epoch": 0.047740637980867945,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.616,
+      "step": 2267
+    },
+    {
+      "epoch": 0.04776169693013167,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002,
+      "loss": 1.5909,
+      "step": 2268
+    },
+    {
+      "epoch": 0.0477827558793954,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.5946,
+      "step": 2269
+    },
+    {
+      "epoch": 0.04780381482865913,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5683,
+      "step": 2270
+    },
+    {
+      "epoch": 0.04782487377792285,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0002,
+      "loss": 1.5998,
+      "step": 2271
+    },
+    {
+      "epoch": 0.047845932727186576,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.5991,
+      "step": 2272
+    },
+    {
+      "epoch": 0.0478669916764503,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0002,
+      "loss": 1.5833,
+      "step": 2273
+    },
+    {
+      "epoch": 0.04788805062571403,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.6072,
+      "step": 2274
+    },
+    {
+      "epoch": 0.04790910957497776,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0002,
+      "loss": 1.6102,
+      "step": 2275
+    },
+    {
+      "epoch": 0.047930168524241486,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6282,
+      "step": 2276
+    },
+    {
+      "epoch": 0.047951227473505206,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.6009,
+      "step": 2277
+    },
+    {
+      "epoch": 0.047972286422768934,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.6353,
+      "step": 2278
+    },
+    {
+      "epoch": 0.04799334537203266,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0002,
+      "loss": 1.6021,
+      "step": 2279
+    },
+    {
+      "epoch": 0.04801440432129639,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.603,
+      "step": 2280
+    },
+    {
+      "epoch": 0.048035463270560116,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.5933,
+      "step": 2281
+    },
+    {
+      "epoch": 0.048056522219823844,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0002,
+      "loss": 1.6063,
+      "step": 2282
+    },
+    {
+      "epoch": 0.04807758116908757,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0002,
+      "loss": 1.6145,
+      "step": 2283
+    },
+    {
+      "epoch": 0.04809864011835129,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.0002,
+      "loss": 1.5623,
+      "step": 2284
+    },
+    {
+      "epoch": 0.04811969906761502,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002,
+      "loss": 1.5876,
+      "step": 2285
+    },
+    {
+      "epoch": 0.04814075801687875,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.6138,
+      "step": 2286
+    },
+    {
+      "epoch": 0.048161816966142475,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.6106,
+      "step": 2287
+    },
+    {
+      "epoch": 0.0481828759154062,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0002,
+      "loss": 1.5788,
+      "step": 2288
+    },
+    {
+      "epoch": 0.04820393486466993,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0002,
+      "loss": 1.624,
+      "step": 2289
+    },
+    {
+      "epoch": 0.04822499381393365,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0002,
+      "loss": 1.5955,
+      "step": 2290
+    },
+    {
+      "epoch": 0.04824605276319738,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0002,
+      "loss": 1.5946,
+      "step": 2291
+    },
+    {
+      "epoch": 0.048267111712461105,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.6034,
+      "step": 2292
+    },
+    {
+      "epoch": 0.04828817066172483,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.6028,
+      "step": 2293
+    },
+    {
+      "epoch": 0.04830922961098856,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0002,
+      "loss": 1.5989,
+      "step": 2294
+    },
+    {
+      "epoch": 0.04833028856025229,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0002,
+      "loss": 1.5621,
+      "step": 2295
+    },
+    {
+      "epoch": 0.048351347509516016,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0002,
+      "loss": 1.6268,
+      "step": 2296
+    },
+    {
+      "epoch": 0.048372406458779736,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.5719,
+      "step": 2297
+    },
+    {
+      "epoch": 0.048393465408043464,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.6334,
+      "step": 2298
+    },
+    {
+      "epoch": 0.04841452435730719,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0002,
+      "loss": 1.5648,
+      "step": 2299
+    },
+    {
+      "epoch": 0.04843558330657092,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5861,
+      "step": 2300
+    },
+    {
+      "epoch": 0.048456642255834646,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.5743,
+      "step": 2301
+    },
+    {
+      "epoch": 0.048477701205098374,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002,
+      "loss": 1.5762,
+      "step": 2302
+    },
+    {
+      "epoch": 0.0484987601543621,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.6071,
+      "step": 2303
+    },
+    {
+      "epoch": 0.04851981910362582,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.6033,
+      "step": 2304
+    },
+    {
+      "epoch": 0.04854087805288955,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.5778,
+      "step": 2305
+    },
+    {
+      "epoch": 0.04856193700215328,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.5825,
+      "step": 2306
+    },
+    {
+      "epoch": 0.048582995951417005,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.6034,
+      "step": 2307
+    },
+    {
+      "epoch": 0.04860405490068073,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0002,
+      "loss": 1.5886,
+      "step": 2308
+    },
+    {
+      "epoch": 0.04862511384994446,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.5816,
+      "step": 2309
+    },
+    {
+      "epoch": 0.04864617279920818,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.5981,
+      "step": 2310
+    },
+    {
+      "epoch": 0.04866723174847191,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0002,
+      "loss": 1.5786,
+      "step": 2311
+    },
+    {
+      "epoch": 0.048688290697735635,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0002,
+      "loss": 1.5972,
+      "step": 2312
+    },
+    {
+      "epoch": 0.04870934964699936,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0002,
+      "loss": 1.611,
+      "step": 2313
+    },
+    {
+      "epoch": 0.04873040859626309,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.5743,
+      "step": 2314
+    },
+    {
+      "epoch": 0.04875146754552682,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0002,
+      "loss": 1.606,
+      "step": 2315
+    },
+    {
+      "epoch": 0.048772526494790545,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.5898,
+      "step": 2316
+    },
+    {
+      "epoch": 0.048793585444054266,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0002,
+      "loss": 1.5734,
+      "step": 2317
+    },
+    {
+      "epoch": 0.048814644393317994,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0002,
+      "loss": 1.6076,
+      "step": 2318
+    },
+    {
+      "epoch": 0.04883570334258172,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5997,
+      "step": 2319
+    },
+    {
+      "epoch": 0.04885676229184545,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.6098,
+      "step": 2320
+    },
+    {
+      "epoch": 0.048877821241109176,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.5645,
+      "step": 2321
+    },
+    {
+      "epoch": 0.048898880190372904,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0002,
+      "loss": 1.5676,
+      "step": 2322
+    },
+    {
+      "epoch": 0.04891993913963663,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.5777,
+      "step": 2323
+    },
+    {
+      "epoch": 0.04894099808890035,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0002,
+      "loss": 1.598,
+      "step": 2324
+    },
+    {
+      "epoch": 0.04896205703816408,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0002,
+      "loss": 1.6044,
+      "step": 2325
+    },
+    {
+      "epoch": 0.04898311598742781,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002,
+      "loss": 1.6068,
+      "step": 2326
+    },
+    {
+      "epoch": 0.049004174936691534,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.5947,
+      "step": 2327
+    },
+    {
+      "epoch": 0.04902523388595526,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.0002,
+      "loss": 1.5867,
+      "step": 2328
+    },
+    {
+      "epoch": 0.04904629283521899,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5654,
+      "step": 2329
+    },
+    {
+      "epoch": 0.04906735178448271,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.5914,
+      "step": 2330
+    },
+    {
+      "epoch": 0.04908841073374644,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.586,
+      "step": 2331
+    },
+    {
+      "epoch": 0.049109469683010165,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.5908,
+      "step": 2332
+    },
+    {
+      "epoch": 0.04913052863227389,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.5924,
+      "step": 2333
+    },
+    {
+      "epoch": 0.04915158758153762,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.6278,
+      "step": 2334
+    },
+    {
+      "epoch": 0.04917264653080135,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.5881,
+      "step": 2335
+    },
+    {
+      "epoch": 0.049193705480065075,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.5798,
+      "step": 2336
+    },
+    {
+      "epoch": 0.049214764429328796,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.5777,
+      "step": 2337
+    },
+    {
+      "epoch": 0.04923582337859252,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.5701,
+      "step": 2338
+    },
+    {
+      "epoch": 0.04925688232785625,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0002,
+      "loss": 1.5824,
+      "step": 2339
+    },
+    {
+      "epoch": 0.04927794127711998,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.5982,
+      "step": 2340
+    },
+    {
+      "epoch": 0.049299000226383706,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5909,
+      "step": 2341
+    },
+    {
+      "epoch": 0.049320059175647434,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.6241,
+      "step": 2342
+    },
+    {
+      "epoch": 0.049341118124911154,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0002,
+      "loss": 1.5857,
+      "step": 2343
+    },
+    {
+      "epoch": 0.04936217707417488,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.5957,
+      "step": 2344
+    },
+    {
+      "epoch": 0.04938323602343861,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5719,
+      "step": 2345
+    },
+    {
+      "epoch": 0.04940429497270234,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.5909,
+      "step": 2346
+    },
+    {
+      "epoch": 0.049425353921966064,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0002,
+      "loss": 1.6035,
+      "step": 2347
+    },
+    {
+      "epoch": 0.04944641287122979,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.611,
+      "step": 2348
+    },
+    {
+      "epoch": 0.04946747182049352,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.5928,
+      "step": 2349
+    },
+    {
+      "epoch": 0.04948853076975724,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.6077,
+      "step": 2350
+    },
+    {
+      "epoch": 0.04950958971902097,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0002,
+      "loss": 1.583,
+      "step": 2351
+    },
+    {
+      "epoch": 0.049530648668284695,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0002,
+      "loss": 1.5872,
+      "step": 2352
+    },
+    {
+      "epoch": 0.04955170761754842,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0002,
+      "loss": 1.61,
+      "step": 2353
+    },
+    {
+      "epoch": 0.04957276656681215,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0002,
+      "loss": 1.5635,
+      "step": 2354
+    },
+    {
+      "epoch": 0.04959382551607588,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.6209,
+      "step": 2355
+    },
+    {
+      "epoch": 0.049614884465339605,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0002,
+      "loss": 1.5828,
+      "step": 2356
+    },
+    {
+      "epoch": 0.049635943414603326,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.5779,
+      "step": 2357
+    },
+    {
+      "epoch": 0.04965700236386705,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.5767,
+      "step": 2358
+    },
+    {
+      "epoch": 0.04967806131313078,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5973,
+      "step": 2359
+    },
+    {
+      "epoch": 0.04969912026239451,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.6,
+      "step": 2360
+    },
+    {
+      "epoch": 0.049720179211658236,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.5986,
+      "step": 2361
+    },
+    {
+      "epoch": 0.04974123816092196,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0002,
+      "loss": 1.5897,
+      "step": 2362
+    },
+    {
+      "epoch": 0.049762297110185684,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5719,
+      "step": 2363
+    },
+    {
+      "epoch": 0.04978335605944941,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5908,
+      "step": 2364
+    },
+    {
+      "epoch": 0.04980441500871314,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.5645,
+      "step": 2365
+    },
+    {
+      "epoch": 0.04982547395797687,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0002,
+      "loss": 1.5835,
+      "step": 2366
+    },
+    {
+      "epoch": 0.049846532907240594,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0002,
+      "loss": 1.6007,
+      "step": 2367
+    },
+    {
+      "epoch": 0.04986759185650432,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.6236,
+      "step": 2368
+    },
+    {
+      "epoch": 0.04988865080576805,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.6026,
+      "step": 2369
+    },
+    {
+      "epoch": 0.04990970975503177,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.0002,
+      "loss": 1.6107,
+      "step": 2370
+    },
+    {
+      "epoch": 0.0499307687042955,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.5673,
+      "step": 2371
+    },
+    {
+      "epoch": 0.049951827653559225,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.6112,
+      "step": 2372
+    },
+    {
+      "epoch": 0.04997288660282295,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.6088,
+      "step": 2373
+    },
+    {
+      "epoch": 0.04999394555208668,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.6032,
+      "step": 2374
+    },
+    {
+      "epoch": 0.05001500450135041,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.59,
+      "step": 2375
+    },
+    {
+      "epoch": 0.050036063450614135,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.61,
+      "step": 2376
+    },
+    {
+      "epoch": 0.050057122399877856,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.5624,
+      "step": 2377
+    },
+    {
+      "epoch": 0.05007818134914158,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.5492,
+      "step": 2378
+    },
+    {
+      "epoch": 0.05009924029840531,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0002,
+      "loss": 1.6157,
+      "step": 2379
+    },
+    {
+      "epoch": 0.05012029924766904,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0002,
+      "loss": 1.5921,
+      "step": 2380
+    },
+    {
+      "epoch": 0.050141358196932766,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5712,
+      "step": 2381
+    },
+    {
+      "epoch": 0.05016241714619649,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5943,
+      "step": 2382
+    },
+    {
+      "epoch": 0.050183476095460214,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.6012,
+      "step": 2383
+    },
+    {
+      "epoch": 0.05020453504472394,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.5972,
+      "step": 2384
+    },
+    {
+      "epoch": 0.05022559399398767,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0002,
+      "loss": 1.6081,
+      "step": 2385
+    },
+    {
+      "epoch": 0.050246652943251396,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.6204,
+      "step": 2386
+    },
+    {
+      "epoch": 0.050267711892515124,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0002,
+      "loss": 1.5849,
+      "step": 2387
+    },
+    {
+      "epoch": 0.05028877084177885,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.5819,
+      "step": 2388
+    },
+    {
+      "epoch": 0.05030982979104258,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.6225,
+      "step": 2389
+    },
+    {
+      "epoch": 0.0503308887403063,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.605,
+      "step": 2390
+    },
+    {
+      "epoch": 0.05035194768957003,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0002,
+      "loss": 1.6118,
+      "step": 2391
+    },
+    {
+      "epoch": 0.050373006638833755,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5658,
+      "step": 2392
+    },
+    {
+      "epoch": 0.05039406558809748,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0002,
+      "loss": 1.609,
+      "step": 2393
+    },
+    {
+      "epoch": 0.05041512453736121,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002,
+      "loss": 1.6149,
+      "step": 2394
+    },
+    {
+      "epoch": 0.05043618348662494,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.5982,
+      "step": 2395
+    },
+    {
+      "epoch": 0.05045724243588866,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.0002,
+      "loss": 1.5839,
+      "step": 2396
+    },
+    {
+      "epoch": 0.050478301385152385,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.5866,
+      "step": 2397
+    },
+    {
+      "epoch": 0.05049936033441611,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.6133,
+      "step": 2398
+    },
+    {
+      "epoch": 0.05052041928367984,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0002,
+      "loss": 1.5834,
+      "step": 2399
+    },
+    {
+      "epoch": 0.05054147823294357,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.5836,
+      "step": 2400
+    },
+    {
+      "epoch": 0.050562537182207296,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5626,
+      "step": 2401
+    },
+    {
+      "epoch": 0.05058359613147102,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0002,
+      "loss": 1.6125,
+      "step": 2402
+    },
+    {
+      "epoch": 0.050604655080734744,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0002,
+      "loss": 1.6058,
+      "step": 2403
+    },
+    {
+      "epoch": 0.05062571402999847,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0002,
+      "loss": 1.5926,
+      "step": 2404
+    },
+    {
+      "epoch": 0.0506467729792622,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0002,
+      "loss": 1.5939,
+      "step": 2405
+    },
+    {
+      "epoch": 0.050667831928525926,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.6173,
+      "step": 2406
+    },
+    {
+      "epoch": 0.050688890877789654,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0002,
+      "loss": 1.5952,
+      "step": 2407
+    },
+    {
+      "epoch": 0.05070994982705338,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0002,
+      "loss": 1.6027,
+      "step": 2408
+    },
+    {
+      "epoch": 0.05073100877631711,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0002,
+      "loss": 1.5826,
+      "step": 2409
+    },
+    {
+      "epoch": 0.05075206772558083,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.6061,
+      "step": 2410
+    },
+    {
+      "epoch": 0.05077312667484456,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.5703,
+      "step": 2411
+    },
+    {
+      "epoch": 0.050794185624108285,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0002,
+      "loss": 1.6009,
+      "step": 2412
+    },
+    {
+      "epoch": 0.05081524457337201,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.5704,
+      "step": 2413
+    },
+    {
+      "epoch": 0.05083630352263574,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0002,
+      "loss": 1.5692,
+      "step": 2414
+    },
+    {
+      "epoch": 0.05085736247189947,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.5654,
+      "step": 2415
+    },
+    {
+      "epoch": 0.05087842142116319,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0002,
+      "loss": 1.612,
+      "step": 2416
+    },
+    {
+      "epoch": 0.050899480370426915,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0002,
+      "loss": 1.5802,
+      "step": 2417
+    },
+    {
+      "epoch": 0.05092053931969064,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.5807,
+      "step": 2418
+    },
+    {
+      "epoch": 0.05094159826895437,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5972,
+      "step": 2419
+    },
+    {
+      "epoch": 0.0509626572182181,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0002,
+      "loss": 1.5843,
+      "step": 2420
+    },
+    {
+      "epoch": 0.050983716167481825,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.622,
+      "step": 2421
+    },
+    {
+      "epoch": 0.05100477511674555,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0002,
+      "loss": 1.5887,
+      "step": 2422
+    },
+    {
+      "epoch": 0.051025834066009274,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0002,
+      "loss": 1.5709,
+      "step": 2423
+    },
+    {
+      "epoch": 0.051046893015273,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6231,
+      "step": 2424
+    },
+    {
+      "epoch": 0.05106795196453673,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0002,
+      "loss": 1.6185,
+      "step": 2425
+    },
+    {
+      "epoch": 0.051089010913800456,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.5905,
+      "step": 2426
+    },
+    {
+      "epoch": 0.051110069863064184,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5775,
+      "step": 2427
+    },
+    {
+      "epoch": 0.05113112881232791,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.5758,
+      "step": 2428
+    },
+    {
+      "epoch": 0.05115218776159164,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0002,
+      "loss": 1.6167,
+      "step": 2429
+    },
+    {
+      "epoch": 0.05117324671085536,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.5977,
+      "step": 2430
+    },
+    {
+      "epoch": 0.05119430566011909,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.0002,
+      "loss": 1.5738,
+      "step": 2431
+    },
+    {
+      "epoch": 0.051215364609382814,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.5768,
+      "step": 2432
+    },
+    {
+      "epoch": 0.05123642355864654,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.5878,
+      "step": 2433
+    },
+    {
+      "epoch": 0.05125748250791027,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002,
+      "loss": 1.5849,
+      "step": 2434
+    },
+    {
+      "epoch": 0.051278541457174,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0002,
+      "loss": 1.5851,
+      "step": 2435
+    },
+    {
+      "epoch": 0.05129960040643772,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0002,
+      "loss": 1.5922,
+      "step": 2436
+    },
+    {
+      "epoch": 0.051320659355701445,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.0002,
+      "loss": 1.5886,
+      "step": 2437
+    },
+    {
+      "epoch": 0.05134171830496517,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.594,
+      "step": 2438
+    },
+    {
+      "epoch": 0.0513627772542289,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0002,
+      "loss": 1.5773,
+      "step": 2439
+    },
+    {
+      "epoch": 0.05138383620349263,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0002,
+      "loss": 1.5868,
+      "step": 2440
+    },
+    {
+      "epoch": 0.051404895152756355,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.6003,
+      "step": 2441
+    },
+    {
+      "epoch": 0.05142595410202008,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0002,
+      "loss": 1.6046,
+      "step": 2442
+    },
+    {
+      "epoch": 0.0514470130512838,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.0002,
+      "loss": 1.5878,
+      "step": 2443
+    },
+    {
+      "epoch": 0.05146807200054753,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0002,
+      "loss": 1.6082,
+      "step": 2444
+    },
+    {
+      "epoch": 0.05148913094981126,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0002,
+      "loss": 1.5925,
+      "step": 2445
+    },
+    {
+      "epoch": 0.051510189899074986,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0002,
+      "loss": 1.5907,
+      "step": 2446
+    },
+    {
+      "epoch": 0.051531248848338713,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.5636,
+      "step": 2447
+    },
+    {
+      "epoch": 0.05155230779760244,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0002,
+      "loss": 1.6255,
+      "step": 2448
+    },
+    {
+      "epoch": 0.05157336674686617,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.5846,
+      "step": 2449
+    },
+    {
+      "epoch": 0.05159442569612989,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.6109,
+      "step": 2450
+    },
+    {
+      "epoch": 0.05161548464539362,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0002,
+      "loss": 1.5711,
+      "step": 2451
+    },
+    {
+      "epoch": 0.051636543594657344,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.5943,
+      "step": 2452
+    },
+    {
+      "epoch": 0.05165760254392107,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.6017,
+      "step": 2453
+    },
+    {
+      "epoch": 0.0516786614931848,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.617,
+      "step": 2454
+    },
+    {
+      "epoch": 0.05169972044244853,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.6096,
+      "step": 2455
+    },
+    {
+      "epoch": 0.05172077939171225,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.5665,
+      "step": 2456
+    },
+    {
+      "epoch": 0.051741838340975975,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0002,
+      "loss": 1.5819,
+      "step": 2457
+    },
+    {
+      "epoch": 0.0517628972902397,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002,
+      "loss": 1.6206,
+      "step": 2458
+    },
+    {
+      "epoch": 0.05178395623950343,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6381,
+      "step": 2459
+    },
+    {
+      "epoch": 0.05180501518876716,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0002,
+      "loss": 1.5907,
+      "step": 2460
+    },
+    {
+      "epoch": 0.051826074138030885,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.5836,
+      "step": 2461
+    },
+    {
+      "epoch": 0.05184713308729461,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.598,
+      "step": 2462
+    },
+    {
+      "epoch": 0.05186819203655833,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.627,
+      "step": 2463
+    },
+    {
+      "epoch": 0.05188925098582206,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.5617,
+      "step": 2464
+    },
+    {
+      "epoch": 0.05191030993508579,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.5799,
+      "step": 2465
+    },
+    {
+      "epoch": 0.051931368884349516,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.5963,
+      "step": 2466
+    },
+    {
+      "epoch": 0.05195242783361324,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.5858,
+      "step": 2467
+    },
+    {
+      "epoch": 0.05197348678287697,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0002,
+      "loss": 1.6237,
+      "step": 2468
+    },
+    {
+      "epoch": 0.05199454573214069,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.5768,
+      "step": 2469
+    },
+    {
+      "epoch": 0.05201560468140442,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.5735,
+      "step": 2470
+    },
+    {
+      "epoch": 0.05203666363066815,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.5988,
+      "step": 2471
+    },
+    {
+      "epoch": 0.052057722579931874,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0002,
+      "loss": 1.5947,
+      "step": 2472
+    },
+    {
+      "epoch": 0.0520787815291956,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.5686,
+      "step": 2473
+    },
+    {
+      "epoch": 0.05209984047845933,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0002,
+      "loss": 1.5656,
+      "step": 2474
+    },
+    {
+      "epoch": 0.05212089942772306,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0002,
+      "loss": 1.6348,
+      "step": 2475
+    },
+    {
+      "epoch": 0.05214195837698678,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6371,
+      "step": 2476
+    },
+    {
+      "epoch": 0.052163017326250505,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0002,
+      "loss": 1.5803,
+      "step": 2477
+    },
+    {
+      "epoch": 0.05218407627551423,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.6156,
+      "step": 2478
+    },
+    {
+      "epoch": 0.05220513522477796,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0002,
+      "loss": 1.5873,
+      "step": 2479
+    },
+    {
+      "epoch": 0.05222619417404169,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0002,
+      "loss": 1.6313,
+      "step": 2480
+    },
+    {
+      "epoch": 0.052247253123305415,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0002,
+      "loss": 1.5947,
+      "step": 2481
+    },
+    {
+      "epoch": 0.05226831207256914,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.0002,
+      "loss": 1.6069,
+      "step": 2482
+    },
+    {
+      "epoch": 0.05228937102183286,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0002,
+      "loss": 1.5696,
+      "step": 2483
+    },
+    {
+      "epoch": 0.05231042997109659,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002,
+      "loss": 1.5729,
+      "step": 2484
+    },
+    {
+      "epoch": 0.05233148892036032,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.5934,
+      "step": 2485
+    },
+    {
+      "epoch": 0.052352547869624046,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002,
+      "loss": 1.5961,
+      "step": 2486
+    },
+    {
+      "epoch": 0.05237360681888777,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.5966,
+      "step": 2487
+    },
+    {
+      "epoch": 0.0523946657681515,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.6011,
+      "step": 2488
+    },
+    {
+      "epoch": 0.05241572471741522,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0002,
+      "loss": 1.612,
+      "step": 2489
+    },
+    {
+      "epoch": 0.05243678366667895,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.5847,
+      "step": 2490
+    },
+    {
+      "epoch": 0.052457842615942676,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0002,
+      "loss": 1.6032,
+      "step": 2491
+    },
+    {
+      "epoch": 0.052478901565206404,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0002,
+      "loss": 1.602,
+      "step": 2492
+    },
+    {
+      "epoch": 0.05249996051447013,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0002,
+      "loss": 1.6155,
+      "step": 2493
+    },
+    {
+      "epoch": 0.05252101946373386,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.5852,
+      "step": 2494
+    },
+    {
+      "epoch": 0.052542078412997587,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5672,
+      "step": 2495
+    },
+    {
+      "epoch": 0.05256313736226131,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5793,
+      "step": 2496
+    },
+    {
+      "epoch": 0.052584196311525035,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.5786,
+      "step": 2497
+    },
+    {
+      "epoch": 0.05260525526078876,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5817,
+      "step": 2498
+    },
+    {
+      "epoch": 0.05262631421005249,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5944,
+      "step": 2499
+    },
+    {
+      "epoch": 0.05264737315931622,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5953,
+      "step": 2500
+    },
+    {
+      "epoch": 0.05264737315931622,
+      "eval_loss": 1.8555322885513306,
+      "eval_runtime": 897.3744,
+      "eval_samples_per_second": 68.868,
+      "eval_steps_per_second": 2.153,
+      "step": 2500
+    },
+    {
+      "epoch": 0.052668432108579945,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.5904,
+      "step": 2501
+    },
+    {
+      "epoch": 0.05268949105784367,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5793,
+      "step": 2502
+    },
+    {
+      "epoch": 0.05271055000710739,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0002,
+      "loss": 1.5756,
+      "step": 2503
+    },
+    {
+      "epoch": 0.05273160895637112,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0002,
+      "loss": 1.6115,
+      "step": 2504
+    },
+    {
+      "epoch": 0.05275266790563485,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.5798,
+      "step": 2505
+    },
+    {
+      "epoch": 0.052773726854898576,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0002,
+      "loss": 1.6013,
+      "step": 2506
+    },
+    {
+      "epoch": 0.0527947858041623,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0002,
+      "loss": 1.5795,
+      "step": 2507
+    },
+    {
+      "epoch": 0.05281584475342603,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6114,
+      "step": 2508
+    },
+    {
+      "epoch": 0.05283690370268975,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.5896,
+      "step": 2509
+    },
+    {
+      "epoch": 0.05285796265195348,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.597,
+      "step": 2510
+    },
+    {
+      "epoch": 0.052879021601217206,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.605,
+      "step": 2511
+    },
+    {
+      "epoch": 0.052900080550480934,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0002,
+      "loss": 1.5908,
+      "step": 2512
+    },
+    {
+      "epoch": 0.05292113949974466,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.6022,
+      "step": 2513
+    },
+    {
+      "epoch": 0.05294219844900839,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.5697,
+      "step": 2514
+    },
+    {
+      "epoch": 0.052963257398272116,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.6134,
+      "step": 2515
+    },
+    {
+      "epoch": 0.05298431634753584,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.5768,
+      "step": 2516
+    },
+    {
+      "epoch": 0.053005375296799564,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5885,
+      "step": 2517
+    },
+    {
+      "epoch": 0.05302643424606329,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.585,
+      "step": 2518
+    },
+    {
+      "epoch": 0.05304749319532702,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0002,
+      "loss": 1.583,
+      "step": 2519
+    },
+    {
+      "epoch": 0.05306855214459075,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.6008,
+      "step": 2520
+    },
+    {
+      "epoch": 0.053089611093854475,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.6221,
+      "step": 2521
+    },
+    {
+      "epoch": 0.053110670043118195,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.592,
+      "step": 2522
+    },
+    {
+      "epoch": 0.05313172899238192,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0002,
+      "loss": 1.5909,
+      "step": 2523
+    },
+    {
+      "epoch": 0.05315278794164565,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.5942,
+      "step": 2524
+    },
+    {
+      "epoch": 0.05317384689090938,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6052,
+      "step": 2525
+    },
+    {
+      "epoch": 0.053194905840173105,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.5789,
+      "step": 2526
+    },
+    {
+      "epoch": 0.05321596478943683,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5898,
+      "step": 2527
+    },
+    {
+      "epoch": 0.05323702373870056,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.5964,
+      "step": 2528
+    },
+    {
+      "epoch": 0.05325808268796428,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.6376,
+      "step": 2529
+    },
+    {
+      "epoch": 0.05327914163722801,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0002,
+      "loss": 1.602,
+      "step": 2530
+    },
+    {
+      "epoch": 0.053300200586491736,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.5657,
+      "step": 2531
+    },
+    {
+      "epoch": 0.053321259535755464,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.5909,
+      "step": 2532
+    },
+    {
+      "epoch": 0.05334231848501919,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.5756,
+      "step": 2533
+    },
+    {
+      "epoch": 0.05336337743428292,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.6181,
+      "step": 2534
+    },
+    {
+      "epoch": 0.053384436383546646,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.593,
+      "step": 2535
+    },
+    {
+      "epoch": 0.05340549533281037,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.0002,
+      "loss": 1.5748,
+      "step": 2536
+    },
+    {
+      "epoch": 0.053426554282074094,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0002,
+      "loss": 1.6143,
+      "step": 2537
+    },
+    {
+      "epoch": 0.05344761323133782,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.6145,
+      "step": 2538
+    },
+    {
+      "epoch": 0.05346867218060155,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.5718,
+      "step": 2539
+    },
+    {
+      "epoch": 0.05348973112986528,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0002,
+      "loss": 1.6014,
+      "step": 2540
+    },
+    {
+      "epoch": 0.053510790079129004,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.5941,
+      "step": 2541
+    },
+    {
+      "epoch": 0.053531849028392725,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.5955,
+      "step": 2542
+    },
+    {
+      "epoch": 0.05355290797765645,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0002,
+      "loss": 1.595,
+      "step": 2543
+    },
+    {
+      "epoch": 0.05357396692692018,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.6078,
+      "step": 2544
+    },
+    {
+      "epoch": 0.05359502587618391,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.5888,
+      "step": 2545
+    },
+    {
+      "epoch": 0.053616084825447635,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6112,
+      "step": 2546
+    },
+    {
+      "epoch": 0.05363714377471136,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.5546,
+      "step": 2547
+    },
+    {
+      "epoch": 0.05365820272397509,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.6221,
+      "step": 2548
+    },
+    {
+      "epoch": 0.05367926167323881,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.5854,
+      "step": 2549
+    },
+    {
+      "epoch": 0.05370032062250254,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.5973,
+      "step": 2550
+    },
+    {
+      "epoch": 0.053721379571766266,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.5784,
+      "step": 2551
+    },
+    {
+      "epoch": 0.053742438521029993,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002,
+      "loss": 1.6059,
+      "step": 2552
+    },
+    {
+      "epoch": 0.05376349747029372,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.0002,
+      "loss": 1.5758,
+      "step": 2553
+    },
+    {
+      "epoch": 0.05378455641955745,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.6042,
+      "step": 2554
+    },
+    {
+      "epoch": 0.053805615368821176,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.5824,
+      "step": 2555
+    },
+    {
+      "epoch": 0.0538266743180849,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5912,
+      "step": 2556
+    },
+    {
+      "epoch": 0.053847733267348624,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0002,
+      "loss": 1.5921,
+      "step": 2557
+    },
+    {
+      "epoch": 0.05386879221661235,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6186,
+      "step": 2558
+    },
+    {
+      "epoch": 0.05388985116587608,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.5744,
+      "step": 2559
+    },
+    {
+      "epoch": 0.05391091011513981,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002,
+      "loss": 1.6282,
+      "step": 2560
+    },
+    {
+      "epoch": 0.053931969064403534,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.5777,
+      "step": 2561
+    },
+    {
+      "epoch": 0.053953028013667255,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.5955,
+      "step": 2562
+    },
+    {
+      "epoch": 0.05397408696293098,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.6096,
+      "step": 2563
+    },
+    {
+      "epoch": 0.05399514591219471,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.5787,
+      "step": 2564
+    },
+    {
+      "epoch": 0.05401620486145844,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5965,
+      "step": 2565
+    },
+    {
+      "epoch": 0.054037263810722165,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6102,
+      "step": 2566
+    },
+    {
+      "epoch": 0.05405832275998589,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.5921,
+      "step": 2567
+    },
+    {
+      "epoch": 0.05407938170924962,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.5937,
+      "step": 2568
+    },
+    {
+      "epoch": 0.05410044065851334,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.5991,
+      "step": 2569
+    },
+    {
+      "epoch": 0.05412149960777707,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.5935,
+      "step": 2570
+    },
+    {
+      "epoch": 0.054142558557040796,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0002,
+      "loss": 1.5601,
+      "step": 2571
+    },
+    {
+      "epoch": 0.05416361750630452,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.585,
+      "step": 2572
+    },
+    {
+      "epoch": 0.05418467645556825,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0002,
+      "loss": 1.5844,
+      "step": 2573
+    },
+    {
+      "epoch": 0.05420573540483198,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.5671,
+      "step": 2574
+    },
+    {
+      "epoch": 0.0542267943540957,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0002,
+      "loss": 1.597,
+      "step": 2575
+    },
+    {
+      "epoch": 0.054247853303359427,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0002,
+      "loss": 1.5861,
+      "step": 2576
+    },
+    {
+      "epoch": 0.054268912252623154,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0002,
+      "loss": 1.6102,
+      "step": 2577
+    },
+    {
+      "epoch": 0.05428997120188688,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.5528,
+      "step": 2578
+    },
+    {
+      "epoch": 0.05431103015115061,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0002,
+      "loss": 1.6068,
+      "step": 2579
+    },
+    {
+      "epoch": 0.05433208910041434,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.6016,
+      "step": 2580
+    },
+    {
+      "epoch": 0.054353148049678064,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.607,
+      "step": 2581
+    },
+    {
+      "epoch": 0.054374206998941785,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.6063,
+      "step": 2582
+    },
+    {
+      "epoch": 0.05439526594820551,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.6144,
+      "step": 2583
+    },
+    {
+      "epoch": 0.05441632489746924,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0002,
+      "loss": 1.5942,
+      "step": 2584
+    },
+    {
+      "epoch": 0.05443738384673297,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.0002,
+      "loss": 1.6068,
+      "step": 2585
+    },
+    {
+      "epoch": 0.054458442795996695,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0002,
+      "loss": 1.6001,
+      "step": 2586
+    },
+    {
+      "epoch": 0.05447950174526042,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0002,
+      "loss": 1.5979,
+      "step": 2587
+    },
+    {
+      "epoch": 0.05450056069452415,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.5743,
+      "step": 2588
+    },
+    {
+      "epoch": 0.05452161964378787,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0002,
+      "loss": 1.5878,
+      "step": 2589
+    },
+    {
+      "epoch": 0.0545426785930516,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0002,
+      "loss": 1.6118,
+      "step": 2590
+    },
+    {
+      "epoch": 0.054563737542315326,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0002,
+      "loss": 1.5729,
+      "step": 2591
+    },
+    {
+      "epoch": 0.05458479649157905,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6074,
+      "step": 2592
+    },
+    {
+      "epoch": 0.05460585544084278,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5974,
+      "step": 2593
+    },
+    {
+      "epoch": 0.05462691439010651,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0002,
+      "loss": 1.5633,
+      "step": 2594
+    },
+    {
+      "epoch": 0.05464797333937023,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.6041,
+      "step": 2595
+    },
+    {
+      "epoch": 0.054669032288633956,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.5818,
+      "step": 2596
+    },
+    {
+      "epoch": 0.054690091237897684,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.5868,
+      "step": 2597
+    },
+    {
+      "epoch": 0.05471115018716141,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0002,
+      "loss": 1.5836,
+      "step": 2598
+    },
+    {
+      "epoch": 0.05473220913642514,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.5778,
+      "step": 2599
+    },
+    {
+      "epoch": 0.054753268085688866,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002,
+      "loss": 1.5863,
+      "step": 2600
+    },
+    {
+      "epoch": 0.054774327034952594,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5708,
+      "step": 2601
+    },
+    {
+      "epoch": 0.054795385984216315,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.5663,
+      "step": 2602
+    },
+    {
+      "epoch": 0.05481644493348004,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.5629,
+      "step": 2603
+    },
+    {
+      "epoch": 0.05483750388274377,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.5663,
+      "step": 2604
+    },
+    {
+      "epoch": 0.0548585628320075,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.6129,
+      "step": 2605
+    },
+    {
+      "epoch": 0.054879621781271225,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0002,
+      "loss": 1.5985,
+      "step": 2606
+    },
+    {
+      "epoch": 0.05490068073053495,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5558,
+      "step": 2607
+    },
+    {
+      "epoch": 0.05492173967979868,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0002,
+      "loss": 1.5482,
+      "step": 2608
+    },
+    {
+      "epoch": 0.0549427986290624,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0002,
+      "loss": 1.5628,
+      "step": 2609
+    },
+    {
+      "epoch": 0.05496385757832613,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0002,
+      "loss": 1.5968,
+      "step": 2610
+    },
+    {
+      "epoch": 0.054984916527589855,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5704,
+      "step": 2611
+    },
+    {
+      "epoch": 0.05500597547685358,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0002,
+      "loss": 1.6135,
+      "step": 2612
+    },
+    {
+      "epoch": 0.05502703442611731,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0002,
+      "loss": 1.5626,
+      "step": 2613
+    },
+    {
+      "epoch": 0.05504809337538104,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.5811,
+      "step": 2614
+    },
+    {
+      "epoch": 0.05506915232464476,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0002,
+      "loss": 1.5719,
+      "step": 2615
+    },
+    {
+      "epoch": 0.055090211273908486,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0002,
+      "loss": 1.554,
+      "step": 2616
+    },
+    {
+      "epoch": 0.055111270223172214,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.6253,
+      "step": 2617
+    },
+    {
+      "epoch": 0.05513232917243594,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.615,
+      "step": 2618
+    },
+    {
+      "epoch": 0.05515338812169967,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.624,
+      "step": 2619
+    },
+    {
+      "epoch": 0.055174447070963396,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0002,
+      "loss": 1.6247,
+      "step": 2620
+    },
+    {
+      "epoch": 0.055195506020227124,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.5748,
+      "step": 2621
+    },
+    {
+      "epoch": 0.055216564969490844,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.591,
+      "step": 2622
+    },
+    {
+      "epoch": 0.05523762391875457,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0002,
+      "loss": 1.6238,
+      "step": 2623
+    },
+    {
+      "epoch": 0.0552586828680183,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0002,
+      "loss": 1.5523,
+      "step": 2624
+    },
+    {
+      "epoch": 0.05527974181728203,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6159,
+      "step": 2625
+    },
+    {
+      "epoch": 0.055300800766545755,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0002,
+      "loss": 1.5625,
+      "step": 2626
+    },
+    {
+      "epoch": 0.05532185971580948,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.5852,
+      "step": 2627
+    },
+    {
+      "epoch": 0.0553429186650732,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0002,
+      "loss": 1.5943,
+      "step": 2628
+    },
+    {
+      "epoch": 0.05536397761433693,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.5604,
+      "step": 2629
+    },
+    {
+      "epoch": 0.05538503656360066,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6234,
+      "step": 2630
+    },
+    {
+      "epoch": 0.055406095512864385,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5855,
+      "step": 2631
+    },
+    {
+      "epoch": 0.05542715446212811,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0002,
+      "loss": 1.6307,
+      "step": 2632
+    },
+    {
+      "epoch": 0.05544821341139184,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.5798,
+      "step": 2633
+    },
+    {
+      "epoch": 0.05546927236065557,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002,
+      "loss": 1.5895,
+      "step": 2634
+    },
+    {
+      "epoch": 0.05549033130991929,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5597,
+      "step": 2635
+    },
+    {
+      "epoch": 0.055511390259183016,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.6119,
+      "step": 2636
+    },
+    {
+      "epoch": 0.055532449208446744,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5777,
+      "step": 2637
+    },
+    {
+      "epoch": 0.05555350815771047,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0002,
+      "loss": 1.5953,
+      "step": 2638
+    },
+    {
+      "epoch": 0.0555745671069742,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.6017,
+      "step": 2639
+    },
+    {
+      "epoch": 0.055595626056237926,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0002,
+      "loss": 1.5907,
+      "step": 2640
+    },
+    {
+      "epoch": 0.055616685005501654,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0002,
+      "loss": 1.6218,
+      "step": 2641
+    },
+    {
+      "epoch": 0.055637743954765374,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.5843,
+      "step": 2642
+    },
+    {
+      "epoch": 0.0556588029040291,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6108,
+      "step": 2643
+    },
+    {
+      "epoch": 0.05567986185329283,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.6051,
+      "step": 2644
+    },
+    {
+      "epoch": 0.05570092080255656,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.5759,
+      "step": 2645
+    },
+    {
+      "epoch": 0.055721979751820284,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0002,
+      "loss": 1.5936,
+      "step": 2646
+    },
+    {
+      "epoch": 0.05574303870108401,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.5875,
+      "step": 2647
+    },
+    {
+      "epoch": 0.05576409765034773,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0002,
+      "loss": 1.5902,
+      "step": 2648
+    },
+    {
+      "epoch": 0.05578515659961146,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.5842,
+      "step": 2649
+    },
+    {
+      "epoch": 0.05580621554887519,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0002,
+      "loss": 1.6006,
+      "step": 2650
+    },
+    {
+      "epoch": 0.055827274498138915,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0002,
+      "loss": 1.5868,
+      "step": 2651
+    },
+    {
+      "epoch": 0.05584833344740264,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0002,
+      "loss": 1.6069,
+      "step": 2652
+    },
+    {
+      "epoch": 0.05586939239666637,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5873,
+      "step": 2653
+    },
+    {
+      "epoch": 0.0558904513459301,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0002,
+      "loss": 1.6088,
+      "step": 2654
+    },
+    {
+      "epoch": 0.05591151029519382,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.5855,
+      "step": 2655
+    },
+    {
+      "epoch": 0.055932569244457546,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.6026,
+      "step": 2656
+    },
+    {
+      "epoch": 0.05595362819372127,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0002,
+      "loss": 1.5953,
+      "step": 2657
+    },
+    {
+      "epoch": 0.055974687142985,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.6051,
+      "step": 2658
+    },
+    {
+      "epoch": 0.05599574609224873,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6029,
+      "step": 2659
+    },
+    {
+      "epoch": 0.056016805041512456,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.5797,
+      "step": 2660
+    },
+    {
+      "epoch": 0.056037863990776184,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0002,
+      "loss": 1.5865,
+      "step": 2661
+    },
+    {
+      "epoch": 0.056058922940039904,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.5665,
+      "step": 2662
+    },
+    {
+      "epoch": 0.05607998188930363,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0002,
+      "loss": 1.5664,
+      "step": 2663
+    },
+    {
+      "epoch": 0.05610104083856736,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.5957,
+      "step": 2664
+    },
+    {
+      "epoch": 0.05612209978783109,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.6215,
+      "step": 2665
+    },
+    {
+      "epoch": 0.056143158737094814,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0002,
+      "loss": 1.581,
+      "step": 2666
+    },
+    {
+      "epoch": 0.05616421768635854,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.6117,
+      "step": 2667
+    },
+    {
+      "epoch": 0.05618527663562226,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0002,
+      "loss": 1.6226,
+      "step": 2668
+    },
+    {
+      "epoch": 0.05620633558488599,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.58,
+      "step": 2669
+    },
+    {
+      "epoch": 0.05622739453414972,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.5784,
+      "step": 2670
+    },
+    {
+      "epoch": 0.056248453483413445,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.5885,
+      "step": 2671
+    },
+    {
+      "epoch": 0.05626951243267717,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.6313,
+      "step": 2672
+    },
+    {
+      "epoch": 0.0562905713819409,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.5783,
+      "step": 2673
+    },
+    {
+      "epoch": 0.05631163033120463,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0002,
+      "loss": 1.5676,
+      "step": 2674
+    },
+    {
+      "epoch": 0.05633268928046835,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.5918,
+      "step": 2675
+    },
+    {
+      "epoch": 0.056353748229732076,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5929,
+      "step": 2676
+    },
+    {
+      "epoch": 0.0563748071789958,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0002,
+      "loss": 1.583,
+      "step": 2677
+    },
+    {
+      "epoch": 0.05639586612825953,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0002,
+      "loss": 1.6212,
+      "step": 2678
+    },
+    {
+      "epoch": 0.05641692507752326,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.574,
+      "step": 2679
+    },
+    {
+      "epoch": 0.056437984026786986,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.6188,
+      "step": 2680
+    },
+    {
+      "epoch": 0.056459042976050706,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0002,
+      "loss": 1.6065,
+      "step": 2681
+    },
+    {
+      "epoch": 0.056480101925314434,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.605,
+      "step": 2682
+    },
+    {
+      "epoch": 0.05650116087457816,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0002,
+      "loss": 1.5809,
+      "step": 2683
+    },
+    {
+      "epoch": 0.05652221982384189,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0002,
+      "loss": 1.5884,
+      "step": 2684
+    },
+    {
+      "epoch": 0.05654327877310562,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.5698,
+      "step": 2685
+    },
+    {
+      "epoch": 0.056564337722369344,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.5956,
+      "step": 2686
+    },
+    {
+      "epoch": 0.05658539667163307,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.593,
+      "step": 2687
+    },
+    {
+      "epoch": 0.05660645562089679,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.577,
+      "step": 2688
+    },
+    {
+      "epoch": 0.05662751457016052,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.5796,
+      "step": 2689
+    },
+    {
+      "epoch": 0.05664857351942425,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002,
+      "loss": 1.5872,
+      "step": 2690
+    },
+    {
+      "epoch": 0.056669632468687975,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.6305,
+      "step": 2691
+    },
+    {
+      "epoch": 0.0566906914179517,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.6102,
+      "step": 2692
+    },
+    {
+      "epoch": 0.05671175036721543,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.5731,
+      "step": 2693
+    },
+    {
+      "epoch": 0.05673280931647916,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5589,
+      "step": 2694
+    },
+    {
+      "epoch": 0.05675386826574288,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.5873,
+      "step": 2695
+    },
+    {
+      "epoch": 0.056774927215006606,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.6103,
+      "step": 2696
+    },
+    {
+      "epoch": 0.05679598616427033,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0002,
+      "loss": 1.578,
+      "step": 2697
+    },
+    {
+      "epoch": 0.05681704511353406,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0002,
+      "loss": 1.6029,
+      "step": 2698
+    },
+    {
+      "epoch": 0.05683810406279779,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.608,
+      "step": 2699
+    },
+    {
+      "epoch": 0.056859163012061516,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0002,
+      "loss": 1.5909,
+      "step": 2700
+    },
+    {
+      "epoch": 0.056880221961325236,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5855,
+      "step": 2701
+    },
+    {
+      "epoch": 0.056901280910588964,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0002,
+      "loss": 1.5619,
+      "step": 2702
+    },
+    {
+      "epoch": 0.05692233985985269,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002,
+      "loss": 1.5706,
+      "step": 2703
+    },
+    {
+      "epoch": 0.05694339880911642,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0002,
+      "loss": 1.562,
+      "step": 2704
+    },
+    {
+      "epoch": 0.056964457758380146,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002,
+      "loss": 1.5656,
+      "step": 2705
+    },
+    {
+      "epoch": 0.056985516707643874,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.5647,
+      "step": 2706
+    },
+    {
+      "epoch": 0.0570065756569076,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0002,
+      "loss": 1.5954,
+      "step": 2707
+    },
+    {
+      "epoch": 0.05702763460617132,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002,
+      "loss": 1.5997,
+      "step": 2708
+    },
+    {
+      "epoch": 0.05704869355543505,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0002,
+      "loss": 1.5598,
+      "step": 2709
+    },
+    {
+      "epoch": 0.05706975250469878,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0002,
+      "loss": 1.5839,
+      "step": 2710
+    },
+    {
+      "epoch": 0.057090811453962505,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.5533,
+      "step": 2711
+    },
+    {
+      "epoch": 0.05711187040322623,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5614,
+      "step": 2712
+    },
+    {
+      "epoch": 0.05713292935248996,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.5905,
+      "step": 2713
+    },
+    {
+      "epoch": 0.05715398830175369,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5641,
+      "step": 2714
+    },
+    {
+      "epoch": 0.05717504725101741,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.6035,
+      "step": 2715
+    },
+    {
+      "epoch": 0.057196106200281135,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.6472,
+      "step": 2716
+    },
+    {
+      "epoch": 0.05721716514954486,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002,
+      "loss": 1.6103,
+      "step": 2717
+    },
+    {
+      "epoch": 0.05723822409880859,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5716,
+      "step": 2718
+    },
+    {
+      "epoch": 0.05725928304807232,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5923,
+      "step": 2719
+    },
+    {
+      "epoch": 0.057280341997336046,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.5762,
+      "step": 2720
+    },
+    {
+      "epoch": 0.057301400946599766,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0002,
+      "loss": 1.6077,
+      "step": 2721
+    },
+    {
+      "epoch": 0.057322459895863494,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0002,
+      "loss": 1.6077,
+      "step": 2722
+    },
+    {
+      "epoch": 0.05734351884512722,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.6006,
+      "step": 2723
+    },
+    {
+      "epoch": 0.05736457779439095,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.5734,
+      "step": 2724
+    },
+    {
+      "epoch": 0.057385636743654676,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002,
+      "loss": 1.6142,
+      "step": 2725
+    },
+    {
+      "epoch": 0.057406695692918404,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.5973,
+      "step": 2726
+    },
+    {
+      "epoch": 0.05742775464218213,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0002,
+      "loss": 1.6103,
+      "step": 2727
+    },
+    {
+      "epoch": 0.05744881359144585,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0002,
+      "loss": 1.6247,
+      "step": 2728
+    },
+    {
+      "epoch": 0.05746987254070958,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.5799,
+      "step": 2729
+    },
+    {
+      "epoch": 0.05749093148997331,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5964,
+      "step": 2730
+    },
+    {
+      "epoch": 0.057511990439237035,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0002,
+      "loss": 1.6099,
+      "step": 2731
+    },
+    {
+      "epoch": 0.05753304938850076,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0002,
+      "loss": 1.5822,
+      "step": 2732
+    },
+    {
+      "epoch": 0.05755410833776449,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0002,
+      "loss": 1.5873,
+      "step": 2733
+    },
+    {
+      "epoch": 0.05757516728702822,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0002,
+      "loss": 1.6042,
+      "step": 2734
+    },
+    {
+      "epoch": 0.05759622623629194,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.6014,
+      "step": 2735
+    },
+    {
+      "epoch": 0.057617285185555665,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0002,
+      "loss": 1.5787,
+      "step": 2736
+    },
+    {
+      "epoch": 0.05763834413481939,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.567,
+      "step": 2737
+    },
+    {
+      "epoch": 0.05765940308408312,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.6316,
+      "step": 2738
+    },
+    {
+      "epoch": 0.05768046203334685,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5763,
+      "step": 2739
+    },
+    {
+      "epoch": 0.057701520982610575,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6204,
+      "step": 2740
+    },
+    {
+      "epoch": 0.057722579931874296,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.5699,
+      "step": 2741
+    },
+    {
+      "epoch": 0.057743638881138024,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.6197,
+      "step": 2742
+    },
+    {
+      "epoch": 0.05776469783040175,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5903,
+      "step": 2743
+    },
+    {
+      "epoch": 0.05778575677966548,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6286,
+      "step": 2744
+    },
+    {
+      "epoch": 0.057806815728929206,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.6015,
+      "step": 2745
+    },
+    {
+      "epoch": 0.057827874678192934,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.5798,
+      "step": 2746
+    },
+    {
+      "epoch": 0.05784893362745666,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.5844,
+      "step": 2747
+    },
+    {
+      "epoch": 0.05786999257672038,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.6236,
+      "step": 2748
+    },
+    {
+      "epoch": 0.05789105152598411,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.6141,
+      "step": 2749
+    },
+    {
+      "epoch": 0.05791211047524784,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5956,
+      "step": 2750
+    },
+    {
+      "epoch": 0.057933169424511564,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.613,
+      "step": 2751
+    },
+    {
+      "epoch": 0.05795422837377529,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.5997,
+      "step": 2752
+    },
+    {
+      "epoch": 0.05797528732303902,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.599,
+      "step": 2753
+    },
+    {
+      "epoch": 0.05799634627230274,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5759,
+      "step": 2754
+    },
+    {
+      "epoch": 0.05801740522156647,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.5712,
+      "step": 2755
+    },
+    {
+      "epoch": 0.058038464170830195,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.588,
+      "step": 2756
+    },
+    {
+      "epoch": 0.05805952312009392,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.0002,
+      "loss": 1.5824,
+      "step": 2757
+    },
+    {
+      "epoch": 0.05808058206935765,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0002,
+      "loss": 1.5989,
+      "step": 2758
+    },
+    {
+      "epoch": 0.05810164101862138,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0002,
+      "loss": 1.6239,
+      "step": 2759
+    },
+    {
+      "epoch": 0.058122699967885105,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.6158,
+      "step": 2760
+    },
+    {
+      "epoch": 0.058143758917148826,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0002,
+      "loss": 1.5622,
+      "step": 2761
+    },
+    {
+      "epoch": 0.05816481786641255,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.6141,
+      "step": 2762
+    },
+    {
+      "epoch": 0.05818587681567628,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002,
+      "loss": 1.6099,
+      "step": 2763
+    },
+    {
+      "epoch": 0.05820693576494001,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0002,
+      "loss": 1.6027,
+      "step": 2764
+    },
+    {
+      "epoch": 0.058227994714203736,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0002,
+      "loss": 1.5761,
+      "step": 2765
+    },
+    {
+      "epoch": 0.058249053663467464,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.5925,
+      "step": 2766
+    },
+    {
+      "epoch": 0.05827011261273119,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0002,
+      "loss": 1.6139,
+      "step": 2767
+    },
+    {
+      "epoch": 0.05829117156199491,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0002,
+      "loss": 1.6219,
+      "step": 2768
+    },
+    {
+      "epoch": 0.05831223051125864,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.5672,
+      "step": 2769
+    },
+    {
+      "epoch": 0.05833328946052237,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6296,
+      "step": 2770
+    },
+    {
+      "epoch": 0.058354348409786094,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.5672,
+      "step": 2771
+    },
+    {
+      "epoch": 0.05837540735904982,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.6268,
+      "step": 2772
+    },
+    {
+      "epoch": 0.05839646630831355,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.5972,
+      "step": 2773
+    },
+    {
+      "epoch": 0.05841752525757727,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.0002,
+      "loss": 1.5838,
+      "step": 2774
+    },
+    {
+      "epoch": 0.058438584206841,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.5931,
+      "step": 2775
+    },
+    {
+      "epoch": 0.058459643156104725,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.5864,
+      "step": 2776
+    },
+    {
+      "epoch": 0.05848070210536845,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.5737,
+      "step": 2777
+    },
+    {
+      "epoch": 0.05850176105463218,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0002,
+      "loss": 1.6105,
+      "step": 2778
+    },
+    {
+      "epoch": 0.05852282000389591,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0002,
+      "loss": 1.6129,
+      "step": 2779
+    },
+    {
+      "epoch": 0.058543878953159635,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.5703,
+      "step": 2780
+    },
+    {
+      "epoch": 0.058564937902423356,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0002,
+      "loss": 1.6097,
+      "step": 2781
+    },
+    {
+      "epoch": 0.05858599685168708,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5902,
+      "step": 2782
+    },
+    {
+      "epoch": 0.05860705580095081,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.6116,
+      "step": 2783
+    },
+    {
+      "epoch": 0.05862811475021454,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6121,
+      "step": 2784
+    },
+    {
+      "epoch": 0.058649173699478266,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.5928,
+      "step": 2785
+    },
+    {
+      "epoch": 0.05867023264874199,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.5852,
+      "step": 2786
+    },
+    {
+      "epoch": 0.05869129159800572,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.602,
+      "step": 2787
+    },
+    {
+      "epoch": 0.05871235054726944,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0002,
+      "loss": 1.6231,
+      "step": 2788
+    },
+    {
+      "epoch": 0.05873340949653317,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0002,
+      "loss": 1.5872,
+      "step": 2789
+    },
+    {
+      "epoch": 0.0587544684457969,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.604,
+      "step": 2790
+    },
+    {
+      "epoch": 0.058775527395060624,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.5861,
+      "step": 2791
+    },
+    {
+      "epoch": 0.05879658634432435,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6121,
+      "step": 2792
+    },
+    {
+      "epoch": 0.05881764529358808,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5855,
+      "step": 2793
+    },
+    {
+      "epoch": 0.0588387042428518,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.6039,
+      "step": 2794
+    },
+    {
+      "epoch": 0.05885976319211553,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0002,
+      "loss": 1.5732,
+      "step": 2795
+    },
+    {
+      "epoch": 0.058880822141379255,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0002,
+      "loss": 1.5777,
+      "step": 2796
+    },
+    {
+      "epoch": 0.05890188109064298,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0002,
+      "loss": 1.5859,
+      "step": 2797
+    },
+    {
+      "epoch": 0.05892294003990671,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.5689,
+      "step": 2798
+    },
+    {
+      "epoch": 0.05894399898917044,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.6183,
+      "step": 2799
+    },
+    {
+      "epoch": 0.058965057938434165,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5746,
+      "step": 2800
+    },
+    {
+      "epoch": 0.058986116887697886,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.0002,
+      "loss": 1.636,
+      "step": 2801
+    },
+    {
+      "epoch": 0.05900717583696161,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.5808,
+      "step": 2802
+    },
+    {
+      "epoch": 0.05902823478622534,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5998,
+      "step": 2803
+    },
+    {
+      "epoch": 0.05904929373548907,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0002,
+      "loss": 1.5915,
+      "step": 2804
+    },
+    {
+      "epoch": 0.059070352684752796,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.5949,
+      "step": 2805
+    },
+    {
+      "epoch": 0.05909141163401652,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.6258,
+      "step": 2806
+    },
+    {
+      "epoch": 0.059112470583280244,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.5923,
+      "step": 2807
+    },
+    {
+      "epoch": 0.05913352953254397,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0002,
+      "loss": 1.5979,
+      "step": 2808
+    },
+    {
+      "epoch": 0.0591545884818077,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002,
+      "loss": 1.5856,
+      "step": 2809
+    },
+    {
+      "epoch": 0.059175647431071426,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.595,
+      "step": 2810
+    },
+    {
+      "epoch": 0.059196706380335154,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.5934,
+      "step": 2811
+    },
+    {
+      "epoch": 0.05921776532959888,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.6112,
+      "step": 2812
+    },
+    {
+      "epoch": 0.05923882427886261,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.6359,
+      "step": 2813
+    },
+    {
+      "epoch": 0.05925988322812633,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0002,
+      "loss": 1.6083,
+      "step": 2814
+    },
+    {
+      "epoch": 0.05928094217739006,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.5733,
+      "step": 2815
+    },
+    {
+      "epoch": 0.059302001126653785,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0002,
+      "loss": 1.585,
+      "step": 2816
+    },
+    {
+      "epoch": 0.05932306007591751,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0002,
+      "loss": 1.5912,
+      "step": 2817
+    },
+    {
+      "epoch": 0.05934411902518124,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0002,
+      "loss": 1.6004,
+      "step": 2818
+    },
+    {
+      "epoch": 0.05936517797444497,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0002,
+      "loss": 1.5871,
+      "step": 2819
+    },
+    {
+      "epoch": 0.059386236923708695,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0002,
+      "loss": 1.5781,
+      "step": 2820
+    },
+    {
+      "epoch": 0.059407295872972415,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.6671,
+      "step": 2821
+    },
+    {
+      "epoch": 0.05942835482223614,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.5751,
+      "step": 2822
+    },
+    {
+      "epoch": 0.05944941377149987,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0002,
+      "loss": 1.6136,
+      "step": 2823
+    },
+    {
+      "epoch": 0.0594704727207636,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5822,
+      "step": 2824
+    },
+    {
+      "epoch": 0.059491531670027326,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0002,
+      "loss": 1.5868,
+      "step": 2825
+    },
+    {
+      "epoch": 0.05951259061929105,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5675,
+      "step": 2826
+    },
+    {
+      "epoch": 0.059533649568554774,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.6064,
+      "step": 2827
+    },
+    {
+      "epoch": 0.0595547085178185,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.5617,
+      "step": 2828
+    },
+    {
+      "epoch": 0.05957576746708223,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5764,
+      "step": 2829
+    },
+    {
+      "epoch": 0.059596826416345956,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.5262,
+      "step": 2830
+    },
+    {
+      "epoch": 0.059617885365609684,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0002,
+      "loss": 1.5867,
+      "step": 2831
+    },
+    {
+      "epoch": 0.05963894431487341,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0002,
+      "loss": 1.5631,
+      "step": 2832
+    },
+    {
+      "epoch": 0.05966000326413714,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0002,
+      "loss": 1.5727,
+      "step": 2833
+    },
+    {
+      "epoch": 0.05968106221340086,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.6295,
+      "step": 2834
+    },
+    {
+      "epoch": 0.05970212116266459,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0002,
+      "loss": 1.5926,
+      "step": 2835
+    },
+    {
+      "epoch": 0.059723180111928315,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0002,
+      "loss": 1.5513,
+      "step": 2836
+    },
+    {
+      "epoch": 0.05974423906119204,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002,
+      "loss": 1.588,
+      "step": 2837
+    },
+    {
+      "epoch": 0.05976529801045577,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.5922,
+      "step": 2838
+    },
+    {
+      "epoch": 0.0597863569597195,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.6122,
+      "step": 2839
+    },
+    {
+      "epoch": 0.059807415908983225,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0002,
+      "loss": 1.5758,
+      "step": 2840
+    },
+    {
+      "epoch": 0.059828474858246945,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0002,
+      "loss": 1.6076,
+      "step": 2841
+    },
+    {
+      "epoch": 0.05984953380751067,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.5862,
+      "step": 2842
+    },
+    {
+      "epoch": 0.0598705927567744,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0002,
+      "loss": 1.5776,
+      "step": 2843
+    },
+    {
+      "epoch": 0.05989165170603813,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0002,
+      "loss": 1.5941,
+      "step": 2844
+    },
+    {
+      "epoch": 0.059912710655301855,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0002,
+      "loss": 1.5857,
+      "step": 2845
+    },
+    {
+      "epoch": 0.05993376960456558,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0002,
+      "loss": 1.6107,
+      "step": 2846
+    },
+    {
+      "epoch": 0.059954828553829304,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0002,
+      "loss": 1.5781,
+      "step": 2847
+    },
+    {
+      "epoch": 0.05997588750309303,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.6101,
+      "step": 2848
+    },
+    {
+      "epoch": 0.05999694645235676,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002,
+      "loss": 1.5965,
+      "step": 2849
+    },
+    {
+      "epoch": 0.060018005401620486,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.6008,
+      "step": 2850
+    },
+    {
+      "epoch": 0.060039064350884214,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.5428,
+      "step": 2851
+    },
+    {
+      "epoch": 0.06006012330014794,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.5873,
+      "step": 2852
+    },
+    {
+      "epoch": 0.06008118224941167,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.578,
+      "step": 2853
+    },
+    {
+      "epoch": 0.06010224119867539,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0002,
+      "loss": 1.6043,
+      "step": 2854
+    },
+    {
+      "epoch": 0.06012330014793912,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002,
+      "loss": 1.5796,
+      "step": 2855
+    },
+    {
+      "epoch": 0.060144359097202844,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0002,
+      "loss": 1.6123,
+      "step": 2856
+    },
+    {
+      "epoch": 0.06016541804646657,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.5845,
+      "step": 2857
+    },
+    {
+      "epoch": 0.0601864769957303,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0002,
+      "loss": 1.5982,
+      "step": 2858
+    },
+    {
+      "epoch": 0.06020753594499403,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5843,
+      "step": 2859
+    },
+    {
+      "epoch": 0.06022859489425775,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.5734,
+      "step": 2860
+    },
+    {
+      "epoch": 0.060249653843521475,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0002,
+      "loss": 1.5915,
+      "step": 2861
+    },
+    {
+      "epoch": 0.0602707127927852,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.5905,
+      "step": 2862
+    },
+    {
+      "epoch": 0.06029177174204893,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.0002,
+      "loss": 1.6011,
+      "step": 2863
+    },
+    {
+      "epoch": 0.06031283069131266,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0002,
+      "loss": 1.6006,
+      "step": 2864
+    },
+    {
+      "epoch": 0.060333889640576385,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002,
+      "loss": 1.6002,
+      "step": 2865
+    },
+    {
+      "epoch": 0.06035494858984011,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.6319,
+      "step": 2866
+    },
+    {
+      "epoch": 0.06037600753910383,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0002,
+      "loss": 1.5671,
+      "step": 2867
+    },
+    {
+      "epoch": 0.06039706648836756,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.6143,
+      "step": 2868
+    },
+    {
+      "epoch": 0.06041812543763129,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5957,
+      "step": 2869
+    },
+    {
+      "epoch": 0.060439184386895016,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0002,
+      "loss": 1.6095,
+      "step": 2870
+    },
+    {
+      "epoch": 0.060460243336158744,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.6156,
+      "step": 2871
+    },
+    {
+      "epoch": 0.06048130228542247,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0002,
+      "loss": 1.6086,
+      "step": 2872
+    },
+    {
+      "epoch": 0.0605023612346862,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0002,
+      "loss": 1.5702,
+      "step": 2873
+    },
+    {
+      "epoch": 0.06052342018394992,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5553,
+      "step": 2874
+    },
+    {
+      "epoch": 0.06054447913321365,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6046,
+      "step": 2875
+    },
+    {
+      "epoch": 0.060565538082477374,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.5751,
+      "step": 2876
+    },
+    {
+      "epoch": 0.0605865970317411,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.5865,
+      "step": 2877
+    },
+    {
+      "epoch": 0.06060765598100483,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.6128,
+      "step": 2878
+    },
+    {
+      "epoch": 0.06062871493026856,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0002,
+      "loss": 1.5626,
+      "step": 2879
+    },
+    {
+      "epoch": 0.06064977387953228,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.6007,
+      "step": 2880
+    },
+    {
+      "epoch": 0.060670832828796005,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0002,
+      "loss": 1.5916,
+      "step": 2881
+    },
+    {
+      "epoch": 0.06069189177805973,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.5888,
+      "step": 2882
+    },
+    {
+      "epoch": 0.06071295072732346,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0002,
+      "loss": 1.5954,
+      "step": 2883
+    },
+    {
+      "epoch": 0.06073400967658719,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5966,
+      "step": 2884
+    },
+    {
+      "epoch": 0.060755068625850915,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.6092,
+      "step": 2885
+    },
+    {
+      "epoch": 0.06077612757511464,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.0002,
+      "loss": 1.561,
+      "step": 2886
+    },
+    {
+      "epoch": 0.06079718652437836,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.581,
+      "step": 2887
+    },
+    {
+      "epoch": 0.06081824547364209,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5738,
+      "step": 2888
+    },
+    {
+      "epoch": 0.06083930442290582,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0002,
+      "loss": 1.5896,
+      "step": 2889
+    },
+    {
+      "epoch": 0.060860363372169546,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0002,
+      "loss": 1.5899,
+      "step": 2890
+    },
+    {
+      "epoch": 0.06088142232143327,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.561,
+      "step": 2891
+    },
+    {
+      "epoch": 0.060902481270697,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.5747,
+      "step": 2892
+    },
+    {
+      "epoch": 0.06092354021996073,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.6181,
+      "step": 2893
+    },
+    {
+      "epoch": 0.06094459916922445,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5812,
+      "step": 2894
+    },
+    {
+      "epoch": 0.06096565811848818,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.5968,
+      "step": 2895
+    },
+    {
+      "epoch": 0.060986717067751904,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0002,
+      "loss": 1.5755,
+      "step": 2896
+    },
+    {
+      "epoch": 0.06100777601701563,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.6082,
+      "step": 2897
+    },
+    {
+      "epoch": 0.06102883496627936,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.5916,
+      "step": 2898
+    },
+    {
+      "epoch": 0.06104989391554309,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.6061,
+      "step": 2899
+    },
+    {
+      "epoch": 0.06107095286480681,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0002,
+      "loss": 1.5612,
+      "step": 2900
+    },
+    {
+      "epoch": 0.061092011814070535,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.5545,
+      "step": 2901
+    },
+    {
+      "epoch": 0.06111307076333426,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0002,
+      "loss": 1.6013,
+      "step": 2902
+    },
+    {
+      "epoch": 0.06113412971259799,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.5885,
+      "step": 2903
+    },
+    {
+      "epoch": 0.06115518866186172,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0002,
+      "loss": 1.5584,
+      "step": 2904
+    },
+    {
+      "epoch": 0.061176247611125445,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0002,
+      "loss": 1.5952,
+      "step": 2905
+    },
+    {
+      "epoch": 0.06119730656038917,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.6214,
+      "step": 2906
+    },
+    {
+      "epoch": 0.06121836550965289,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.5832,
+      "step": 2907
+    },
+    {
+      "epoch": 0.06123942445891662,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0002,
+      "loss": 1.5581,
+      "step": 2908
+    },
+    {
+      "epoch": 0.06126048340818035,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0002,
+      "loss": 1.5858,
+      "step": 2909
+    },
+    {
+      "epoch": 0.061281542357444076,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.5741,
+      "step": 2910
+    },
+    {
+      "epoch": 0.0613026013067078,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0002,
+      "loss": 1.5896,
+      "step": 2911
+    },
+    {
+      "epoch": 0.06132366025597153,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.6074,
+      "step": 2912
+    },
+    {
+      "epoch": 0.06134471920523525,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.5767,
+      "step": 2913
+    },
+    {
+      "epoch": 0.06136577815449898,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.5643,
+      "step": 2914
+    },
+    {
+      "epoch": 0.061386837103762706,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.5622,
+      "step": 2915
+    },
+    {
+      "epoch": 0.061407896053026434,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.6049,
+      "step": 2916
+    },
+    {
+      "epoch": 0.06142895500229016,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5876,
+      "step": 2917
+    },
+    {
+      "epoch": 0.06145001395155389,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.5791,
+      "step": 2918
+    },
+    {
+      "epoch": 0.06147107290081762,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5735,
+      "step": 2919
+    },
+    {
+      "epoch": 0.06149213185008134,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.6117,
+      "step": 2920
+    },
+    {
+      "epoch": 0.061513190799345065,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0002,
+      "loss": 1.5737,
+      "step": 2921
+    },
+    {
+      "epoch": 0.06153424974860879,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.5987,
+      "step": 2922
+    },
+    {
+      "epoch": 0.06155530869787252,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.5796,
+      "step": 2923
+    },
+    {
+      "epoch": 0.06157636764713625,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0002,
+      "loss": 1.6035,
+      "step": 2924
+    },
+    {
+      "epoch": 0.061597426596399975,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0002,
+      "loss": 1.5918,
+      "step": 2925
+    },
+    {
+      "epoch": 0.0616184855456637,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.5885,
+      "step": 2926
+    },
+    {
+      "epoch": 0.06163954449492742,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.5628,
+      "step": 2927
+    },
+    {
+      "epoch": 0.06166060344419115,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.5971,
+      "step": 2928
+    },
+    {
+      "epoch": 0.06168166239345488,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.5758,
+      "step": 2929
+    },
+    {
+      "epoch": 0.061702721342718606,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0002,
+      "loss": 1.5956,
+      "step": 2930
+    },
+    {
+      "epoch": 0.06172378029198233,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.5761,
+      "step": 2931
+    },
+    {
+      "epoch": 0.06174483924124606,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5888,
+      "step": 2932
+    },
+    {
+      "epoch": 0.06176589819050978,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0002,
+      "loss": 1.5697,
+      "step": 2933
+    },
+    {
+      "epoch": 0.06178695713977351,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0002,
+      "loss": 1.5948,
+      "step": 2934
+    },
+    {
+      "epoch": 0.061808016089037236,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.5668,
+      "step": 2935
+    },
+    {
+      "epoch": 0.061829075038300964,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0002,
+      "loss": 1.5874,
+      "step": 2936
+    },
+    {
+      "epoch": 0.06185013398756469,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0002,
+      "loss": 1.6189,
+      "step": 2937
+    },
+    {
+      "epoch": 0.06187119293682842,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.593,
+      "step": 2938
+    },
+    {
+      "epoch": 0.061892251886092146,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6187,
+      "step": 2939
+    },
+    {
+      "epoch": 0.06191331083535587,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.6067,
+      "step": 2940
+    },
+    {
+      "epoch": 0.061934369784619595,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.5938,
+      "step": 2941
+    },
+    {
+      "epoch": 0.06195542873388332,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0002,
+      "loss": 1.6212,
+      "step": 2942
+    },
+    {
+      "epoch": 0.06197648768314705,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0002,
+      "loss": 1.5653,
+      "step": 2943
+    },
+    {
+      "epoch": 0.06199754663241078,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.6083,
+      "step": 2944
+    },
+    {
+      "epoch": 0.062018605581674505,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.5916,
+      "step": 2945
+    },
+    {
+      "epoch": 0.06203966453093823,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.6082,
+      "step": 2946
+    },
+    {
+      "epoch": 0.06206072348020195,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0002,
+      "loss": 1.6118,
+      "step": 2947
+    },
+    {
+      "epoch": 0.06208178242946568,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.5965,
+      "step": 2948
+    },
+    {
+      "epoch": 0.06210284137872941,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0002,
+      "loss": 1.5617,
+      "step": 2949
+    },
+    {
+      "epoch": 0.062123900327993135,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.5769,
+      "step": 2950
+    },
+    {
+      "epoch": 0.06214495927725686,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0002,
+      "loss": 1.5891,
+      "step": 2951
+    },
+    {
+      "epoch": 0.06216601822652059,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0002,
+      "loss": 1.5869,
+      "step": 2952
+    },
+    {
+      "epoch": 0.06218707717578431,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0002,
+      "loss": 1.5884,
+      "step": 2953
+    },
+    {
+      "epoch": 0.06220813612504804,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0002,
+      "loss": 1.6096,
+      "step": 2954
+    },
+    {
+      "epoch": 0.062229195074311766,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.6331,
+      "step": 2955
+    },
+    {
+      "epoch": 0.062250254023575494,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0002,
+      "loss": 1.574,
+      "step": 2956
+    },
+    {
+      "epoch": 0.06227131297283922,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0002,
+      "loss": 1.6117,
+      "step": 2957
+    },
+    {
+      "epoch": 0.06229237192210295,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5479,
+      "step": 2958
+    },
+    {
+      "epoch": 0.062313430871366676,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.6285,
+      "step": 2959
+    },
+    {
+      "epoch": 0.0623344898206304,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0002,
+      "loss": 1.5694,
+      "step": 2960
+    },
+    {
+      "epoch": 0.062355548769894124,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.5903,
+      "step": 2961
+    },
+    {
+      "epoch": 0.06237660771915785,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0002,
+      "loss": 1.5592,
+      "step": 2962
+    },
+    {
+      "epoch": 0.06239766666842158,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0002,
+      "loss": 1.5959,
+      "step": 2963
+    },
+    {
+      "epoch": 0.06241872561768531,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0002,
+      "loss": 1.568,
+      "step": 2964
+    },
+    {
+      "epoch": 0.062439784566949035,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0002,
+      "loss": 1.5848,
+      "step": 2965
+    },
+    {
+      "epoch": 0.062460843516212755,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5861,
+      "step": 2966
+    },
+    {
+      "epoch": 0.06248190246547648,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6007,
+      "step": 2967
+    },
+    {
+      "epoch": 0.06250296141474021,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0002,
+      "loss": 1.5867,
+      "step": 2968
+    },
+    {
+      "epoch": 0.06252402036400394,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.5906,
+      "step": 2969
+    },
+    {
+      "epoch": 0.06254507931326767,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.0002,
+      "loss": 1.6045,
+      "step": 2970
+    },
+    {
+      "epoch": 0.06256613826253139,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.626,
+      "step": 2971
+    },
+    {
+      "epoch": 0.06258719721179512,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6292,
+      "step": 2972
+    },
+    {
+      "epoch": 0.06260825616105885,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.608,
+      "step": 2973
+    },
+    {
+      "epoch": 0.06262931511032258,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0002,
+      "loss": 1.6152,
+      "step": 2974
+    },
+    {
+      "epoch": 0.0626503740595863,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.5555,
+      "step": 2975
+    },
+    {
+      "epoch": 0.06267143300885002,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0002,
+      "loss": 1.5855,
+      "step": 2976
+    },
+    {
+      "epoch": 0.06269249195811374,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5993,
+      "step": 2977
+    },
+    {
+      "epoch": 0.06271355090737747,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.5872,
+      "step": 2978
+    },
+    {
+      "epoch": 0.0627346098566412,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.637,
+      "step": 2979
+    },
+    {
+      "epoch": 0.06275566880590493,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.5912,
+      "step": 2980
+    },
+    {
+      "epoch": 0.06277672775516865,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0002,
+      "loss": 1.5965,
+      "step": 2981
+    },
+    {
+      "epoch": 0.06279778670443238,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.5999,
+      "step": 2982
+    },
+    {
+      "epoch": 0.06281884565369611,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0002,
+      "loss": 1.5987,
+      "step": 2983
+    },
+    {
+      "epoch": 0.06283990460295984,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.6009,
+      "step": 2984
+    },
+    {
+      "epoch": 0.06286096355222356,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.5902,
+      "step": 2985
+    },
+    {
+      "epoch": 0.06288202250148729,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.6002,
+      "step": 2986
+    },
+    {
+      "epoch": 0.06290308145075102,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0002,
+      "loss": 1.6044,
+      "step": 2987
+    },
+    {
+      "epoch": 0.06292414040001475,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.6099,
+      "step": 2988
+    },
+    {
+      "epoch": 0.06294519934927847,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002,
+      "loss": 1.5466,
+      "step": 2989
+    },
+    {
+      "epoch": 0.06296625829854219,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.5766,
+      "step": 2990
+    },
+    {
+      "epoch": 0.06298731724780592,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0002,
+      "loss": 1.572,
+      "step": 2991
+    },
+    {
+      "epoch": 0.06300837619706964,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6028,
+      "step": 2992
+    },
+    {
+      "epoch": 0.06302943514633337,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.6016,
+      "step": 2993
+    },
+    {
+      "epoch": 0.0630504940955971,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.5996,
+      "step": 2994
+    },
+    {
+      "epoch": 0.06307155304486083,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.0002,
+      "loss": 1.5902,
+      "step": 2995
+    },
+    {
+      "epoch": 0.06309261199412455,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.596,
+      "step": 2996
+    },
+    {
+      "epoch": 0.06311367094338828,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.5985,
+      "step": 2997
+    },
+    {
+      "epoch": 0.06313472989265201,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5796,
+      "step": 2998
+    },
+    {
+      "epoch": 0.06315578884191574,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0002,
+      "loss": 1.6203,
+      "step": 2999
+    },
+    {
+      "epoch": 0.06317684779117946,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.0002,
+      "loss": 1.6133,
+      "step": 3000
+    },
+    {
+      "epoch": 0.06317684779117946,
+      "eval_loss": 1.9333325624465942,
+      "eval_runtime": 897.2967,
+      "eval_samples_per_second": 68.874,
+      "eval_steps_per_second": 2.153,
+      "step": 3000
+    },
+    {
+      "epoch": 0.06319790674044319,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.5906,
+      "step": 3001
+    },
+    {
+      "epoch": 0.06321896568970692,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0002,
+      "loss": 1.5969,
+      "step": 3002
+    },
+    {
+      "epoch": 0.06324002463897063,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.0002,
+      "loss": 1.5957,
+      "step": 3003
+    },
+    {
+      "epoch": 0.06326108358823436,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5964,
+      "step": 3004
+    },
+    {
+      "epoch": 0.06328214253749809,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.5886,
+      "step": 3005
+    },
+    {
+      "epoch": 0.06330320148676181,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5531,
+      "step": 3006
+    },
+    {
+      "epoch": 0.06332426043602554,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.5673,
+      "step": 3007
+    },
+    {
+      "epoch": 0.06334531938528927,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.6109,
+      "step": 3008
+    },
+    {
+      "epoch": 0.063366378334553,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.5875,
+      "step": 3009
+    },
+    {
+      "epoch": 0.06338743728381672,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.5615,
+      "step": 3010
+    },
+    {
+      "epoch": 0.06340849623308045,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0002,
+      "loss": 1.5999,
+      "step": 3011
+    },
+    {
+      "epoch": 0.06342955518234418,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0002,
+      "loss": 1.6145,
+      "step": 3012
+    },
+    {
+      "epoch": 0.06345061413160791,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0002,
+      "loss": 1.6371,
+      "step": 3013
+    },
+    {
+      "epoch": 0.06347167308087164,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0002,
+      "loss": 1.6008,
+      "step": 3014
+    },
+    {
+      "epoch": 0.06349273203013536,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.5836,
+      "step": 3015
+    },
+    {
+      "epoch": 0.06351379097939908,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0002,
+      "loss": 1.5772,
+      "step": 3016
+    },
+    {
+      "epoch": 0.0635348499286628,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6209,
+      "step": 3017
+    },
+    {
+      "epoch": 0.06355590887792653,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.6118,
+      "step": 3018
+    },
+    {
+      "epoch": 0.06357696782719026,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5861,
+      "step": 3019
+    },
+    {
+      "epoch": 0.06359802677645399,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.593,
+      "step": 3020
+    },
+    {
+      "epoch": 0.06361908572571771,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.5714,
+      "step": 3021
+    },
+    {
+      "epoch": 0.06364014467498144,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.5833,
+      "step": 3022
+    },
+    {
+      "epoch": 0.06366120362424517,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.6076,
+      "step": 3023
+    },
+    {
+      "epoch": 0.0636822625735089,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0002,
+      "loss": 1.5769,
+      "step": 3024
+    },
+    {
+      "epoch": 0.06370332152277262,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0002,
+      "loss": 1.5562,
+      "step": 3025
+    },
+    {
+      "epoch": 0.06372438047203635,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002,
+      "loss": 1.5904,
+      "step": 3026
+    },
+    {
+      "epoch": 0.06374543942130008,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0002,
+      "loss": 1.5971,
+      "step": 3027
+    },
+    {
+      "epoch": 0.0637664983705638,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.5802,
+      "step": 3028
+    },
+    {
+      "epoch": 0.06378755731982752,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.6065,
+      "step": 3029
+    },
+    {
+      "epoch": 0.06380861626909125,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.0002,
+      "loss": 1.5643,
+      "step": 3030
+    },
+    {
+      "epoch": 0.06382967521835498,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.5886,
+      "step": 3031
+    },
+    {
+      "epoch": 0.0638507341676187,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.6016,
+      "step": 3032
+    },
+    {
+      "epoch": 0.06387179311688243,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.5743,
+      "step": 3033
+    },
+    {
+      "epoch": 0.06389285206614616,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.5864,
+      "step": 3034
+    },
+    {
+      "epoch": 0.06391391101540989,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.5741,
+      "step": 3035
+    },
+    {
+      "epoch": 0.06393496996467361,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.6237,
+      "step": 3036
+    },
+    {
+      "epoch": 0.06395602891393734,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.5835,
+      "step": 3037
+    },
+    {
+      "epoch": 0.06397708786320107,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0002,
+      "loss": 1.58,
+      "step": 3038
+    },
+    {
+      "epoch": 0.0639981468124648,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.587,
+      "step": 3039
+    },
+    {
+      "epoch": 0.06401920576172852,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0002,
+      "loss": 1.5824,
+      "step": 3040
+    },
+    {
+      "epoch": 0.06404026471099225,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.5974,
+      "step": 3041
+    },
+    {
+      "epoch": 0.06406132366025598,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.5969,
+      "step": 3042
+    },
+    {
+      "epoch": 0.06408238260951969,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0002,
+      "loss": 1.5584,
+      "step": 3043
+    },
+    {
+      "epoch": 0.06410344155878342,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0002,
+      "loss": 1.6098,
+      "step": 3044
+    },
+    {
+      "epoch": 0.06412450050804715,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.5999,
+      "step": 3045
+    },
+    {
+      "epoch": 0.06414555945731087,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.6047,
+      "step": 3046
+    },
+    {
+      "epoch": 0.0641666184065746,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0002,
+      "loss": 1.5648,
+      "step": 3047
+    },
+    {
+      "epoch": 0.06418767735583833,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.5723,
+      "step": 3048
+    },
+    {
+      "epoch": 0.06420873630510206,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002,
+      "loss": 1.6345,
+      "step": 3049
+    },
+    {
+      "epoch": 0.06422979525436578,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.0002,
+      "loss": 1.6192,
+      "step": 3050
+    },
+    {
+      "epoch": 0.06425085420362951,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0002,
+      "loss": 1.5963,
+      "step": 3051
+    },
+    {
+      "epoch": 0.06427191315289324,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0002,
+      "loss": 1.6091,
+      "step": 3052
+    },
+    {
+      "epoch": 0.06429297210215697,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5706,
+      "step": 3053
+    },
+    {
+      "epoch": 0.0643140310514207,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.5992,
+      "step": 3054
+    },
+    {
+      "epoch": 0.06433509000068442,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.6289,
+      "step": 3055
+    },
+    {
+      "epoch": 0.06435614894994814,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5666,
+      "step": 3056
+    },
+    {
+      "epoch": 0.06437720789921186,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.5682,
+      "step": 3057
+    },
+    {
+      "epoch": 0.06439826684847559,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.6141,
+      "step": 3058
+    },
+    {
+      "epoch": 0.06441932579773932,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.6054,
+      "step": 3059
+    },
+    {
+      "epoch": 0.06444038474700305,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.5862,
+      "step": 3060
+    },
+    {
+      "epoch": 0.06446144369626677,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.5831,
+      "step": 3061
+    },
+    {
+      "epoch": 0.0644825026455305,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0002,
+      "loss": 1.6014,
+      "step": 3062
+    },
+    {
+      "epoch": 0.06450356159479423,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0002,
+      "loss": 1.5685,
+      "step": 3063
+    },
+    {
+      "epoch": 0.06452462054405796,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002,
+      "loss": 1.6278,
+      "step": 3064
+    },
+    {
+      "epoch": 0.06454567949332168,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.6106,
+      "step": 3065
+    },
+    {
+      "epoch": 0.06456673844258541,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0002,
+      "loss": 1.5928,
+      "step": 3066
+    },
+    {
+      "epoch": 0.06458779739184914,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.595,
+      "step": 3067
+    },
+    {
+      "epoch": 0.06460885634111287,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0002,
+      "loss": 1.5907,
+      "step": 3068
+    },
+    {
+      "epoch": 0.06462991529037658,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.59,
+      "step": 3069
+    },
+    {
+      "epoch": 0.06465097423964031,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.5732,
+      "step": 3070
+    },
+    {
+      "epoch": 0.06467203318890404,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.5765,
+      "step": 3071
+    },
+    {
+      "epoch": 0.06469309213816776,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.578,
+      "step": 3072
+    },
+    {
+      "epoch": 0.06471415108743149,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0002,
+      "loss": 1.6227,
+      "step": 3073
+    },
+    {
+      "epoch": 0.06473521003669522,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.6012,
+      "step": 3074
+    },
+    {
+      "epoch": 0.06475626898595895,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.0002,
+      "loss": 1.6134,
+      "step": 3075
+    },
+    {
+      "epoch": 0.06477732793522267,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0002,
+      "loss": 1.5859,
+      "step": 3076
+    },
+    {
+      "epoch": 0.0647983868844864,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0002,
+      "loss": 1.5393,
+      "step": 3077
+    },
+    {
+      "epoch": 0.06481944583375013,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.6228,
+      "step": 3078
+    },
+    {
+      "epoch": 0.06484050478301386,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0002,
+      "loss": 1.567,
+      "step": 3079
+    },
+    {
+      "epoch": 0.06486156373227758,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.5892,
+      "step": 3080
+    },
+    {
+      "epoch": 0.06488262268154131,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.607,
+      "step": 3081
+    },
+    {
+      "epoch": 0.06490368163080502,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.5929,
+      "step": 3082
+    },
+    {
+      "epoch": 0.06492474058006875,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.6184,
+      "step": 3083
+    },
+    {
+      "epoch": 0.06494579952933248,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.5705,
+      "step": 3084
+    },
+    {
+      "epoch": 0.0649668584785962,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.5949,
+      "step": 3085
+    },
+    {
+      "epoch": 0.06498791742785993,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0002,
+      "loss": 1.6382,
+      "step": 3086
+    },
+    {
+      "epoch": 0.06500897637712366,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.6078,
+      "step": 3087
+    },
+    {
+      "epoch": 0.06503003532638739,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.591,
+      "step": 3088
+    },
+    {
+      "epoch": 0.06505109427565112,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0002,
+      "loss": 1.6111,
+      "step": 3089
+    },
+    {
+      "epoch": 0.06507215322491484,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0002,
+      "loss": 1.583,
+      "step": 3090
+    },
+    {
+      "epoch": 0.06509321217417857,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0002,
+      "loss": 1.5775,
+      "step": 3091
+    },
+    {
+      "epoch": 0.0651142711234423,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0002,
+      "loss": 1.5945,
+      "step": 3092
+    },
+    {
+      "epoch": 0.06513533007270603,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.5741,
+      "step": 3093
+    },
+    {
+      "epoch": 0.06515638902196975,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0002,
+      "loss": 1.6234,
+      "step": 3094
+    },
+    {
+      "epoch": 0.06517744797123348,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.5584,
+      "step": 3095
+    },
+    {
+      "epoch": 0.0651985069204972,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.5895,
+      "step": 3096
+    },
+    {
+      "epoch": 0.06521956586976092,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.6119,
+      "step": 3097
+    },
+    {
+      "epoch": 0.06524062481902465,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002,
+      "loss": 1.5983,
+      "step": 3098
+    },
+    {
+      "epoch": 0.06526168376828838,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6006,
+      "step": 3099
+    },
+    {
+      "epoch": 0.0652827427175521,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.5793,
+      "step": 3100
+    },
+    {
+      "epoch": 0.06530380166681583,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6044,
+      "step": 3101
+    },
+    {
+      "epoch": 0.06532486061607956,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0002,
+      "loss": 1.5612,
+      "step": 3102
+    },
+    {
+      "epoch": 0.06534591956534329,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.6095,
+      "step": 3103
+    },
+    {
+      "epoch": 0.06536697851460702,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0002,
+      "loss": 1.5979,
+      "step": 3104
+    },
+    {
+      "epoch": 0.06538803746387074,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.5933,
+      "step": 3105
+    },
+    {
+      "epoch": 0.06540909641313447,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0002,
+      "loss": 1.5868,
+      "step": 3106
+    },
+    {
+      "epoch": 0.0654301553623982,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.5832,
+      "step": 3107
+    },
+    {
+      "epoch": 0.06545121431166193,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.5897,
+      "step": 3108
+    },
+    {
+      "epoch": 0.06547227326092564,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6334,
+      "step": 3109
+    },
+    {
+      "epoch": 0.06549333221018937,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0002,
+      "loss": 1.5847,
+      "step": 3110
+    },
+    {
+      "epoch": 0.0655143911594531,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.6056,
+      "step": 3111
+    },
+    {
+      "epoch": 0.06553545010871682,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.5805,
+      "step": 3112
+    },
+    {
+      "epoch": 0.06555650905798055,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.5843,
+      "step": 3113
+    },
+    {
+      "epoch": 0.06557756800724428,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.6168,
+      "step": 3114
+    },
+    {
+      "epoch": 0.065598626956508,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0002,
+      "loss": 1.606,
+      "step": 3115
+    },
+    {
+      "epoch": 0.06561968590577173,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6039,
+      "step": 3116
+    },
+    {
+      "epoch": 0.06564074485503546,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.5845,
+      "step": 3117
+    },
+    {
+      "epoch": 0.06566180380429919,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002,
+      "loss": 1.5681,
+      "step": 3118
+    },
+    {
+      "epoch": 0.06568286275356292,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.5683,
+      "step": 3119
+    },
+    {
+      "epoch": 0.06570392170282664,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.0002,
+      "loss": 1.5582,
+      "step": 3120
+    },
+    {
+      "epoch": 0.06572498065209037,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0002,
+      "loss": 1.6254,
+      "step": 3121
+    },
+    {
+      "epoch": 0.06574603960135408,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.5739,
+      "step": 3122
+    },
+    {
+      "epoch": 0.06576709855061781,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5972,
+      "step": 3123
+    },
+    {
+      "epoch": 0.06578815749988154,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.5846,
+      "step": 3124
+    },
+    {
+      "epoch": 0.06580921644914527,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.5959,
+      "step": 3125
+    },
+    {
+      "epoch": 0.065830275398409,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5985,
+      "step": 3126
+    },
+    {
+      "epoch": 0.06585133434767272,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.5916,
+      "step": 3127
+    },
+    {
+      "epoch": 0.06587239329693645,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5687,
+      "step": 3128
+    },
+    {
+      "epoch": 0.06589345224620018,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0002,
+      "loss": 1.5815,
+      "step": 3129
+    },
+    {
+      "epoch": 0.0659145111954639,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0002,
+      "loss": 1.5674,
+      "step": 3130
+    },
+    {
+      "epoch": 0.06593557014472763,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0002,
+      "loss": 1.573,
+      "step": 3131
+    },
+    {
+      "epoch": 0.06595662909399136,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.5584,
+      "step": 3132
+    },
+    {
+      "epoch": 0.06597768804325509,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5609,
+      "step": 3133
+    },
+    {
+      "epoch": 0.06599874699251881,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6015,
+      "step": 3134
+    },
+    {
+      "epoch": 0.06601980594178254,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.5862,
+      "step": 3135
+    },
+    {
+      "epoch": 0.06604086489104626,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.5978,
+      "step": 3136
+    },
+    {
+      "epoch": 0.06606192384030998,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.5902,
+      "step": 3137
+    },
+    {
+      "epoch": 0.06608298278957371,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0002,
+      "loss": 1.6151,
+      "step": 3138
+    },
+    {
+      "epoch": 0.06610404173883744,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0002,
+      "loss": 1.5702,
+      "step": 3139
+    },
+    {
+      "epoch": 0.06612510068810117,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0002,
+      "loss": 1.5842,
+      "step": 3140
+    },
+    {
+      "epoch": 0.06614615963736489,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.5535,
+      "step": 3141
+    },
+    {
+      "epoch": 0.06616721858662862,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0002,
+      "loss": 1.5695,
+      "step": 3142
+    },
+    {
+      "epoch": 0.06618827753589235,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.5851,
+      "step": 3143
+    },
+    {
+      "epoch": 0.06620933648515608,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0002,
+      "loss": 1.5686,
+      "step": 3144
+    },
+    {
+      "epoch": 0.0662303954344198,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.5837,
+      "step": 3145
+    },
+    {
+      "epoch": 0.06625145438368353,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.6189,
+      "step": 3146
+    },
+    {
+      "epoch": 0.06627251333294726,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.5518,
+      "step": 3147
+    },
+    {
+      "epoch": 0.06629357228221099,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.5905,
+      "step": 3148
+    },
+    {
+      "epoch": 0.0663146312314747,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.5973,
+      "step": 3149
+    },
+    {
+      "epoch": 0.06633569018073843,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.6209,
+      "step": 3150
+    },
+    {
+      "epoch": 0.06635674913000215,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5701,
+      "step": 3151
+    },
+    {
+      "epoch": 0.06637780807926588,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.0002,
+      "loss": 1.578,
+      "step": 3152
+    },
+    {
+      "epoch": 0.06639886702852961,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.5621,
+      "step": 3153
+    },
+    {
+      "epoch": 0.06641992597779334,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0002,
+      "loss": 1.5551,
+      "step": 3154
+    },
+    {
+      "epoch": 0.06644098492705706,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5914,
+      "step": 3155
+    },
+    {
+      "epoch": 0.06646204387632079,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0002,
+      "loss": 1.5788,
+      "step": 3156
+    },
+    {
+      "epoch": 0.06648310282558452,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.5987,
+      "step": 3157
+    },
+    {
+      "epoch": 0.06650416177484825,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0002,
+      "loss": 1.5284,
+      "step": 3158
+    },
+    {
+      "epoch": 0.06652522072411197,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0002,
+      "loss": 1.5909,
+      "step": 3159
+    },
+    {
+      "epoch": 0.0665462796733757,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.5987,
+      "step": 3160
+    },
+    {
+      "epoch": 0.06656733862263943,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0002,
+      "loss": 1.5567,
+      "step": 3161
+    },
+    {
+      "epoch": 0.06658839757190314,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5952,
+      "step": 3162
+    },
+    {
+      "epoch": 0.06660945652116687,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0002,
+      "loss": 1.634,
+      "step": 3163
+    },
+    {
+      "epoch": 0.0666305154704306,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0002,
+      "loss": 1.5741,
+      "step": 3164
+    },
+    {
+      "epoch": 0.06665157441969433,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0002,
+      "loss": 1.5805,
+      "step": 3165
+    },
+    {
+      "epoch": 0.06667263336895805,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.5575,
+      "step": 3166
+    },
+    {
+      "epoch": 0.06669369231822178,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0002,
+      "loss": 1.5991,
+      "step": 3167
+    },
+    {
+      "epoch": 0.06671475126748551,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0002,
+      "loss": 1.5642,
+      "step": 3168
+    },
+    {
+      "epoch": 0.06673581021674924,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.5953,
+      "step": 3169
+    },
+    {
+      "epoch": 0.06675686916601296,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.5967,
+      "step": 3170
+    },
+    {
+      "epoch": 0.06677792811527669,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.6008,
+      "step": 3171
+    },
+    {
+      "epoch": 0.06679898706454042,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.6075,
+      "step": 3172
+    },
+    {
+      "epoch": 0.06682004601380415,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0002,
+      "loss": 1.5334,
+      "step": 3173
+    },
+    {
+      "epoch": 0.06684110496306787,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.5771,
+      "step": 3174
+    },
+    {
+      "epoch": 0.06686216391233159,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.5455,
+      "step": 3175
+    },
+    {
+      "epoch": 0.06688322286159532,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.5859,
+      "step": 3176
+    },
+    {
+      "epoch": 0.06690428181085904,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.5846,
+      "step": 3177
+    },
+    {
+      "epoch": 0.06692534076012277,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.6238,
+      "step": 3178
+    },
+    {
+      "epoch": 0.0669463997093865,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0002,
+      "loss": 1.6122,
+      "step": 3179
+    },
+    {
+      "epoch": 0.06696745865865023,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0002,
+      "loss": 1.5924,
+      "step": 3180
+    },
+    {
+      "epoch": 0.06698851760791395,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.5964,
+      "step": 3181
+    },
+    {
+      "epoch": 0.06700957655717768,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.5815,
+      "step": 3182
+    },
+    {
+      "epoch": 0.06703063550644141,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0002,
+      "loss": 1.6021,
+      "step": 3183
+    },
+    {
+      "epoch": 0.06705169445570514,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.596,
+      "step": 3184
+    },
+    {
+      "epoch": 0.06707275340496886,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0002,
+      "loss": 1.5652,
+      "step": 3185
+    },
+    {
+      "epoch": 0.06709381235423259,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5659,
+      "step": 3186
+    },
+    {
+      "epoch": 0.06711487130349632,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.6112,
+      "step": 3187
+    },
+    {
+      "epoch": 0.06713593025276005,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.5256,
+      "step": 3188
+    },
+    {
+      "epoch": 0.06715698920202376,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6053,
+      "step": 3189
+    },
+    {
+      "epoch": 0.06717804815128749,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.5716,
+      "step": 3190
+    },
+    {
+      "epoch": 0.06719910710055121,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.6144,
+      "step": 3191
+    },
+    {
+      "epoch": 0.06722016604981494,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5854,
+      "step": 3192
+    },
+    {
+      "epoch": 0.06724122499907867,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.5918,
+      "step": 3193
+    },
+    {
+      "epoch": 0.0672622839483424,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0002,
+      "loss": 1.5865,
+      "step": 3194
+    },
+    {
+      "epoch": 0.06728334289760612,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002,
+      "loss": 1.5762,
+      "step": 3195
+    },
+    {
+      "epoch": 0.06730440184686985,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0002,
+      "loss": 1.5988,
+      "step": 3196
+    },
+    {
+      "epoch": 0.06732546079613358,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.628,
+      "step": 3197
+    },
+    {
+      "epoch": 0.0673465197453973,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0002,
+      "loss": 1.5663,
+      "step": 3198
+    },
+    {
+      "epoch": 0.06736757869466103,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5896,
+      "step": 3199
+    },
+    {
+      "epoch": 0.06738863764392476,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.0002,
+      "loss": 1.6017,
+      "step": 3200
+    },
+    {
+      "epoch": 0.06740969659318849,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.5523,
+      "step": 3201
+    },
+    {
+      "epoch": 0.0674307555424522,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0002,
+      "loss": 1.5864,
+      "step": 3202
+    },
+    {
+      "epoch": 0.06745181449171593,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.6228,
+      "step": 3203
+    },
+    {
+      "epoch": 0.06747287344097966,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0002,
+      "loss": 1.5868,
+      "step": 3204
+    },
+    {
+      "epoch": 0.06749393239024339,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0002,
+      "loss": 1.6038,
+      "step": 3205
+    },
+    {
+      "epoch": 0.06751499133950711,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.565,
+      "step": 3206
+    },
+    {
+      "epoch": 0.06753605028877084,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0002,
+      "loss": 1.6388,
+      "step": 3207
+    },
+    {
+      "epoch": 0.06755710923803457,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0002,
+      "loss": 1.5923,
+      "step": 3208
+    },
+    {
+      "epoch": 0.0675781681872983,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6203,
+      "step": 3209
+    },
+    {
+      "epoch": 0.06759922713656202,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.6211,
+      "step": 3210
+    },
+    {
+      "epoch": 0.06762028608582575,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.0002,
+      "loss": 1.6208,
+      "step": 3211
+    },
+    {
+      "epoch": 0.06764134503508948,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0002,
+      "loss": 1.6139,
+      "step": 3212
+    },
+    {
+      "epoch": 0.0676624039843532,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0002,
+      "loss": 1.6047,
+      "step": 3213
+    },
+    {
+      "epoch": 0.06768346293361693,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.5747,
+      "step": 3214
+    },
+    {
+      "epoch": 0.06770452188288065,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0002,
+      "loss": 1.6023,
+      "step": 3215
+    },
+    {
+      "epoch": 0.06772558083214437,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0002,
+      "loss": 1.576,
+      "step": 3216
+    },
+    {
+      "epoch": 0.0677466397814081,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.6063,
+      "step": 3217
+    },
+    {
+      "epoch": 0.06776769873067183,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5833,
+      "step": 3218
+    },
+    {
+      "epoch": 0.06778875767993556,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0002,
+      "loss": 1.5709,
+      "step": 3219
+    },
+    {
+      "epoch": 0.06780981662919928,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5727,
+      "step": 3220
+    },
+    {
+      "epoch": 0.06783087557846301,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002,
+      "loss": 1.6391,
+      "step": 3221
+    },
+    {
+      "epoch": 0.06785193452772674,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.5911,
+      "step": 3222
+    },
+    {
+      "epoch": 0.06787299347699047,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.5796,
+      "step": 3223
+    },
+    {
+      "epoch": 0.0678940524262542,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.542,
+      "step": 3224
+    },
+    {
+      "epoch": 0.06791511137551792,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5873,
+      "step": 3225
+    },
+    {
+      "epoch": 0.06793617032478165,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.6205,
+      "step": 3226
+    },
+    {
+      "epoch": 0.06795722927404538,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0002,
+      "loss": 1.5749,
+      "step": 3227
+    },
+    {
+      "epoch": 0.06797828822330909,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002,
+      "loss": 1.5544,
+      "step": 3228
+    },
+    {
+      "epoch": 0.06799934717257282,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5965,
+      "step": 3229
+    },
+    {
+      "epoch": 0.06802040612183655,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002,
+      "loss": 1.5878,
+      "step": 3230
+    },
+    {
+      "epoch": 0.06804146507110027,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0002,
+      "loss": 1.5614,
+      "step": 3231
+    },
+    {
+      "epoch": 0.068062524020364,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.5762,
+      "step": 3232
+    },
+    {
+      "epoch": 0.06808358296962773,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0002,
+      "loss": 1.6164,
+      "step": 3233
+    },
+    {
+      "epoch": 0.06810464191889146,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.5852,
+      "step": 3234
+    },
+    {
+      "epoch": 0.06812570086815518,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0002,
+      "loss": 1.5969,
+      "step": 3235
+    },
+    {
+      "epoch": 0.06814675981741891,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.5763,
+      "step": 3236
+    },
+    {
+      "epoch": 0.06816781876668264,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.5924,
+      "step": 3237
+    },
+    {
+      "epoch": 0.06818887771594637,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002,
+      "loss": 1.5841,
+      "step": 3238
+    },
+    {
+      "epoch": 0.0682099366652101,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0002,
+      "loss": 1.5696,
+      "step": 3239
+    },
+    {
+      "epoch": 0.06823099561447382,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.5658,
+      "step": 3240
+    },
+    {
+      "epoch": 0.06825205456373755,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.5994,
+      "step": 3241
+    },
+    {
+      "epoch": 0.06827311351300126,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.6002,
+      "step": 3242
+    },
+    {
+      "epoch": 0.06829417246226499,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.5845,
+      "step": 3243
+    },
+    {
+      "epoch": 0.06831523141152872,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002,
+      "loss": 1.6075,
+      "step": 3244
+    },
+    {
+      "epoch": 0.06833629036079245,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0002,
+      "loss": 1.5567,
+      "step": 3245
+    },
+    {
+      "epoch": 0.06835734931005617,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.623,
+      "step": 3246
+    },
+    {
+      "epoch": 0.0683784082593199,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5745,
+      "step": 3247
+    },
+    {
+      "epoch": 0.06839946720858363,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0002,
+      "loss": 1.5872,
+      "step": 3248
+    },
+    {
+      "epoch": 0.06842052615784736,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.6086,
+      "step": 3249
+    },
+    {
+      "epoch": 0.06844158510711108,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0002,
+      "loss": 1.576,
+      "step": 3250
+    },
+    {
+      "epoch": 0.06846264405637481,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.5546,
+      "step": 3251
+    },
+    {
+      "epoch": 0.06848370300563854,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0002,
+      "loss": 1.593,
+      "step": 3252
+    },
+    {
+      "epoch": 0.06850476195490227,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.6151,
+      "step": 3253
+    },
+    {
+      "epoch": 0.068525820904166,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0002,
+      "loss": 1.5743,
+      "step": 3254
+    },
+    {
+      "epoch": 0.0685468798534297,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0002,
+      "loss": 1.5772,
+      "step": 3255
+    },
+    {
+      "epoch": 0.06856793880269343,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0002,
+      "loss": 1.5424,
+      "step": 3256
+    },
+    {
+      "epoch": 0.06858899775195716,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.5404,
+      "step": 3257
+    },
+    {
+      "epoch": 0.06861005670122089,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.6115,
+      "step": 3258
+    },
+    {
+      "epoch": 0.06863111565048462,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002,
+      "loss": 1.5763,
+      "step": 3259
+    },
+    {
+      "epoch": 0.06865217459974834,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0002,
+      "loss": 1.5986,
+      "step": 3260
+    },
+    {
+      "epoch": 0.06867323354901207,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.5683,
+      "step": 3261
+    },
+    {
+      "epoch": 0.0686942924982758,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0002,
+      "loss": 1.601,
+      "step": 3262
+    },
+    {
+      "epoch": 0.06871535144753953,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0002,
+      "loss": 1.5778,
+      "step": 3263
+    },
+    {
+      "epoch": 0.06873641039680325,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.5788,
+      "step": 3264
+    },
+    {
+      "epoch": 0.06875746934606698,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.6188,
+      "step": 3265
+    },
+    {
+      "epoch": 0.06877852829533071,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.5628,
+      "step": 3266
+    },
+    {
+      "epoch": 0.06879958724459444,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0002,
+      "loss": 1.5833,
+      "step": 3267
+    },
+    {
+      "epoch": 0.06882064619385815,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5544,
+      "step": 3268
+    },
+    {
+      "epoch": 0.06884170514312188,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0002,
+      "loss": 1.5928,
+      "step": 3269
+    },
+    {
+      "epoch": 0.0688627640923856,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.5971,
+      "step": 3270
+    },
+    {
+      "epoch": 0.06888382304164933,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.5722,
+      "step": 3271
+    },
+    {
+      "epoch": 0.06890488199091306,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.5935,
+      "step": 3272
+    },
+    {
+      "epoch": 0.06892594094017679,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.5724,
+      "step": 3273
+    },
+    {
+      "epoch": 0.06894699988944052,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.5803,
+      "step": 3274
+    },
+    {
+      "epoch": 0.06896805883870424,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.5884,
+      "step": 3275
+    },
+    {
+      "epoch": 0.06898911778796797,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.6164,
+      "step": 3276
+    },
+    {
+      "epoch": 0.0690101767372317,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6018,
+      "step": 3277
+    },
+    {
+      "epoch": 0.06903123568649543,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0002,
+      "loss": 1.5726,
+      "step": 3278
+    },
+    {
+      "epoch": 0.06905229463575915,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.0002,
+      "loss": 1.5867,
+      "step": 3279
+    },
+    {
+      "epoch": 0.06907335358502288,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002,
+      "loss": 1.5925,
+      "step": 3280
+    },
+    {
+      "epoch": 0.0690944125342866,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.5906,
+      "step": 3281
+    },
+    {
+      "epoch": 0.06911547148355032,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0002,
+      "loss": 1.6063,
+      "step": 3282
+    },
+    {
+      "epoch": 0.06913653043281405,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.6078,
+      "step": 3283
+    },
+    {
+      "epoch": 0.06915758938207778,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002,
+      "loss": 1.5895,
+      "step": 3284
+    },
+    {
+      "epoch": 0.0691786483313415,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5782,
+      "step": 3285
+    },
+    {
+      "epoch": 0.06919970728060523,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6235,
+      "step": 3286
+    },
+    {
+      "epoch": 0.06922076622986896,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.568,
+      "step": 3287
+    },
+    {
+      "epoch": 0.06924182517913269,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.611,
+      "step": 3288
+    },
+    {
+      "epoch": 0.06926288412839642,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0002,
+      "loss": 1.5989,
+      "step": 3289
+    },
+    {
+      "epoch": 0.06928394307766014,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.5946,
+      "step": 3290
+    },
+    {
+      "epoch": 0.06930500202692387,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.5604,
+      "step": 3291
+    },
+    {
+      "epoch": 0.0693260609761876,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.5668,
+      "step": 3292
+    },
+    {
+      "epoch": 0.06934711992545133,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.5919,
+      "step": 3293
+    },
+    {
+      "epoch": 0.06936817887471505,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.6296,
+      "step": 3294
+    },
+    {
+      "epoch": 0.06938923782397877,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0002,
+      "loss": 1.5508,
+      "step": 3295
+    },
+    {
+      "epoch": 0.0694102967732425,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.5985,
+      "step": 3296
+    },
+    {
+      "epoch": 0.06943135572250622,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0002,
+      "loss": 1.6122,
+      "step": 3297
+    },
+    {
+      "epoch": 0.06945241467176995,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.6016,
+      "step": 3298
+    },
+    {
+      "epoch": 0.06947347362103368,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.0002,
+      "loss": 1.561,
+      "step": 3299
+    },
+    {
+      "epoch": 0.0694945325702974,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0002,
+      "loss": 1.5835,
+      "step": 3300
+    },
+    {
+      "epoch": 0.06951559151956113,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.5935,
+      "step": 3301
+    },
+    {
+      "epoch": 0.06953665046882486,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0002,
+      "loss": 1.6008,
+      "step": 3302
+    },
+    {
+      "epoch": 0.06955770941808859,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.5795,
+      "step": 3303
+    },
+    {
+      "epoch": 0.06957876836735231,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0002,
+      "loss": 1.5838,
+      "step": 3304
+    },
+    {
+      "epoch": 0.06959982731661604,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.6013,
+      "step": 3305
+    },
+    {
+      "epoch": 0.06962088626587977,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.6131,
+      "step": 3306
+    },
+    {
+      "epoch": 0.0696419452151435,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5872,
+      "step": 3307
+    },
+    {
+      "epoch": 0.06966300416440721,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0002,
+      "loss": 1.6032,
+      "step": 3308
+    },
+    {
+      "epoch": 0.06968406311367094,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0002,
+      "loss": 1.5984,
+      "step": 3309
+    },
+    {
+      "epoch": 0.06970512206293467,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.5868,
+      "step": 3310
+    },
+    {
+      "epoch": 0.0697261810121984,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0002,
+      "loss": 1.5732,
+      "step": 3311
+    },
+    {
+      "epoch": 0.06974723996146212,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0002,
+      "loss": 1.6163,
+      "step": 3312
+    },
+    {
+      "epoch": 0.06976829891072585,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.578,
+      "step": 3313
+    },
+    {
+      "epoch": 0.06978935785998958,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.5921,
+      "step": 3314
+    },
+    {
+      "epoch": 0.0698104168092533,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.5945,
+      "step": 3315
+    },
+    {
+      "epoch": 0.06983147575851703,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.6089,
+      "step": 3316
+    },
+    {
+      "epoch": 0.06985253470778076,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.5871,
+      "step": 3317
+    },
+    {
+      "epoch": 0.06987359365704449,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.565,
+      "step": 3318
+    },
+    {
+      "epoch": 0.06989465260630821,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0002,
+      "loss": 1.6194,
+      "step": 3319
+    },
+    {
+      "epoch": 0.06991571155557194,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.5778,
+      "step": 3320
+    },
+    {
+      "epoch": 0.06993677050483565,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0002,
+      "loss": 1.5738,
+      "step": 3321
+    },
+    {
+      "epoch": 0.06995782945409938,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.627,
+      "step": 3322
+    },
+    {
+      "epoch": 0.06997888840336311,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.5512,
+      "step": 3323
+    },
+    {
+      "epoch": 0.06999994735262684,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5877,
+      "step": 3324
+    },
+    {
+      "epoch": 0.07002100630189056,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002,
+      "loss": 1.5496,
+      "step": 3325
+    },
+    {
+      "epoch": 0.07004206525115429,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.5985,
+      "step": 3326
+    },
+    {
+      "epoch": 0.07006312420041802,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0002,
+      "loss": 1.5819,
+      "step": 3327
+    },
+    {
+      "epoch": 0.07008418314968175,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0002,
+      "loss": 1.6326,
+      "step": 3328
+    },
+    {
+      "epoch": 0.07010524209894547,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.5975,
+      "step": 3329
+    },
+    {
+      "epoch": 0.0701263010482092,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0002,
+      "loss": 1.5582,
+      "step": 3330
+    },
+    {
+      "epoch": 0.07014735999747293,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0002,
+      "loss": 1.621,
+      "step": 3331
+    },
+    {
+      "epoch": 0.07016841894673666,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0002,
+      "loss": 1.5931,
+      "step": 3332
+    },
+    {
+      "epoch": 0.07018947789600039,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0002,
+      "loss": 1.5806,
+      "step": 3333
+    },
+    {
+      "epoch": 0.0702105368452641,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0002,
+      "loss": 1.5703,
+      "step": 3334
+    },
+    {
+      "epoch": 0.07023159579452783,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0002,
+      "loss": 1.5241,
+      "step": 3335
+    },
+    {
+      "epoch": 0.07025265474379155,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.5873,
+      "step": 3336
+    },
+    {
+      "epoch": 0.07027371369305528,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0002,
+      "loss": 1.6145,
+      "step": 3337
+    },
+    {
+      "epoch": 0.07029477264231901,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.5453,
+      "step": 3338
+    },
+    {
+      "epoch": 0.07031583159158274,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.6414,
+      "step": 3339
+    },
+    {
+      "epoch": 0.07033689054084646,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.6024,
+      "step": 3340
+    },
+    {
+      "epoch": 0.07035794949011019,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.5954,
+      "step": 3341
+    },
+    {
+      "epoch": 0.07037900843937392,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0002,
+      "loss": 1.6043,
+      "step": 3342
+    },
+    {
+      "epoch": 0.07040006738863765,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.5852,
+      "step": 3343
+    },
+    {
+      "epoch": 0.07042112633790137,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0002,
+      "loss": 1.5989,
+      "step": 3344
+    },
+    {
+      "epoch": 0.0704421852871651,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.5929,
+      "step": 3345
+    },
+    {
+      "epoch": 0.07046324423642883,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.591,
+      "step": 3346
+    },
+    {
+      "epoch": 0.07048430318569256,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.5675,
+      "step": 3347
+    },
+    {
+      "epoch": 0.07050536213495627,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.0002,
+      "loss": 1.6384,
+      "step": 3348
+    },
+    {
+      "epoch": 0.07052642108422,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6158,
+      "step": 3349
+    },
+    {
+      "epoch": 0.07054748003348373,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.5776,
+      "step": 3350
+    },
+    {
+      "epoch": 0.07056853898274745,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0002,
+      "loss": 1.5895,
+      "step": 3351
+    },
+    {
+      "epoch": 0.07058959793201118,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.5845,
+      "step": 3352
+    },
+    {
+      "epoch": 0.07061065688127491,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0002,
+      "loss": 1.5991,
+      "step": 3353
+    },
+    {
+      "epoch": 0.07063171583053864,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0002,
+      "loss": 1.5643,
+      "step": 3354
+    },
+    {
+      "epoch": 0.07065277477980236,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0002,
+      "loss": 1.5937,
+      "step": 3355
+    },
+    {
+      "epoch": 0.07067383372906609,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.6088,
+      "step": 3356
+    },
+    {
+      "epoch": 0.07069489267832982,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5473,
+      "step": 3357
+    },
+    {
+      "epoch": 0.07071595162759355,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.5495,
+      "step": 3358
+    },
+    {
+      "epoch": 0.07073701057685727,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002,
+      "loss": 1.5824,
+      "step": 3359
+    },
+    {
+      "epoch": 0.070758069526121,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0002,
+      "loss": 1.5883,
+      "step": 3360
+    },
+    {
+      "epoch": 0.07077912847538471,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0002,
+      "loss": 1.6201,
+      "step": 3361
+    },
+    {
+      "epoch": 0.07080018742464844,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.6182,
+      "step": 3362
+    },
+    {
+      "epoch": 0.07082124637391217,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.6024,
+      "step": 3363
+    },
+    {
+      "epoch": 0.0708423053231759,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.5826,
+      "step": 3364
+    },
+    {
+      "epoch": 0.07086336427243962,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.6057,
+      "step": 3365
+    },
+    {
+      "epoch": 0.07088442322170335,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.5789,
+      "step": 3366
+    },
+    {
+      "epoch": 0.07090548217096708,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.605,
+      "step": 3367
+    },
+    {
+      "epoch": 0.07092654112023081,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.6236,
+      "step": 3368
+    },
+    {
+      "epoch": 0.07094760006949453,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0002,
+      "loss": 1.58,
+      "step": 3369
+    },
+    {
+      "epoch": 0.07096865901875826,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0002,
+      "loss": 1.6024,
+      "step": 3370
+    },
+    {
+      "epoch": 0.07098971796802199,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0002,
+      "loss": 1.55,
+      "step": 3371
+    },
+    {
+      "epoch": 0.07101077691728572,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.5934,
+      "step": 3372
+    },
+    {
+      "epoch": 0.07103183586654944,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002,
+      "loss": 1.5746,
+      "step": 3373
+    },
+    {
+      "epoch": 0.07105289481581316,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.5806,
+      "step": 3374
+    },
+    {
+      "epoch": 0.07107395376507689,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.5752,
+      "step": 3375
+    },
+    {
+      "epoch": 0.07109501271434061,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0002,
+      "loss": 1.5422,
+      "step": 3376
+    },
+    {
+      "epoch": 0.07111607166360434,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0002,
+      "loss": 1.5962,
+      "step": 3377
+    },
+    {
+      "epoch": 0.07113713061286807,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.558,
+      "step": 3378
+    },
+    {
+      "epoch": 0.0711581895621318,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5894,
+      "step": 3379
+    },
+    {
+      "epoch": 0.07117924851139552,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002,
+      "loss": 1.5858,
+      "step": 3380
+    },
+    {
+      "epoch": 0.07120030746065925,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0002,
+      "loss": 1.5549,
+      "step": 3381
+    },
+    {
+      "epoch": 0.07122136640992298,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0002,
+      "loss": 1.5861,
+      "step": 3382
+    },
+    {
+      "epoch": 0.0712424253591867,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0002,
+      "loss": 1.5841,
+      "step": 3383
+    },
+    {
+      "epoch": 0.07126348430845043,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.5844,
+      "step": 3384
+    },
+    {
+      "epoch": 0.07128454325771416,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.5963,
+      "step": 3385
+    },
+    {
+      "epoch": 0.07130560220697789,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0002,
+      "loss": 1.5755,
+      "step": 3386
+    },
+    {
+      "epoch": 0.0713266611562416,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6009,
+      "step": 3387
+    },
+    {
+      "epoch": 0.07134772010550533,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6011,
+      "step": 3388
+    },
+    {
+      "epoch": 0.07136877905476906,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0002,
+      "loss": 1.5814,
+      "step": 3389
+    },
+    {
+      "epoch": 0.07138983800403279,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0002,
+      "loss": 1.6063,
+      "step": 3390
+    },
+    {
+      "epoch": 0.07141089695329651,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0002,
+      "loss": 1.5543,
+      "step": 3391
+    },
+    {
+      "epoch": 0.07143195590256024,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.5871,
+      "step": 3392
+    },
+    {
+      "epoch": 0.07145301485182397,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.6066,
+      "step": 3393
+    },
+    {
+      "epoch": 0.0714740738010877,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0002,
+      "loss": 1.6039,
+      "step": 3394
+    },
+    {
+      "epoch": 0.07149513275035142,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0002,
+      "loss": 1.609,
+      "step": 3395
+    },
+    {
+      "epoch": 0.07151619169961515,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.535,
+      "step": 3396
+    },
+    {
+      "epoch": 0.07153725064887888,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6119,
+      "step": 3397
+    },
+    {
+      "epoch": 0.0715583095981426,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.5568,
+      "step": 3398
+    },
+    {
+      "epoch": 0.07157936854740633,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.6563,
+      "step": 3399
+    },
+    {
+      "epoch": 0.07160042749667006,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0002,
+      "loss": 1.5814,
+      "step": 3400
+    },
+    {
+      "epoch": 0.07162148644593377,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002,
+      "loss": 1.5728,
+      "step": 3401
+    },
+    {
+      "epoch": 0.0716425453951975,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5521,
+      "step": 3402
+    },
+    {
+      "epoch": 0.07166360434446123,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.5606,
+      "step": 3403
+    },
+    {
+      "epoch": 0.07168466329372496,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0002,
+      "loss": 1.5802,
+      "step": 3404
+    },
+    {
+      "epoch": 0.07170572224298868,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.6006,
+      "step": 3405
+    },
+    {
+      "epoch": 0.07172678119225241,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.5956,
+      "step": 3406
+    },
+    {
+      "epoch": 0.07174784014151614,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.6143,
+      "step": 3407
+    },
+    {
+      "epoch": 0.07176889909077987,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0002,
+      "loss": 1.5521,
+      "step": 3408
+    },
+    {
+      "epoch": 0.0717899580400436,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.5941,
+      "step": 3409
+    },
+    {
+      "epoch": 0.07181101698930732,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5813,
+      "step": 3410
+    },
+    {
+      "epoch": 0.07183207593857105,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.5846,
+      "step": 3411
+    },
+    {
+      "epoch": 0.07185313488783478,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0002,
+      "loss": 1.5972,
+      "step": 3412
+    },
+    {
+      "epoch": 0.0718741938370985,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0002,
+      "loss": 1.6199,
+      "step": 3413
+    },
+    {
+      "epoch": 0.07189525278636222,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.6075,
+      "step": 3414
+    },
+    {
+      "epoch": 0.07191631173562595,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0002,
+      "loss": 1.5843,
+      "step": 3415
+    },
+    {
+      "epoch": 0.07193737068488967,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.5417,
+      "step": 3416
+    },
+    {
+      "epoch": 0.0719584296341534,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.5617,
+      "step": 3417
+    },
+    {
+      "epoch": 0.07197948858341713,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.5716,
+      "step": 3418
+    },
+    {
+      "epoch": 0.07200054753268086,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0002,
+      "loss": 1.6202,
+      "step": 3419
+    },
+    {
+      "epoch": 0.07202160648194458,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.6093,
+      "step": 3420
+    },
+    {
+      "epoch": 0.07204266543120831,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5577,
+      "step": 3421
+    },
+    {
+      "epoch": 0.07206372438047204,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0002,
+      "loss": 1.619,
+      "step": 3422
+    },
+    {
+      "epoch": 0.07208478332973577,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.5855,
+      "step": 3423
+    },
+    {
+      "epoch": 0.0721058422789995,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.5796,
+      "step": 3424
+    },
+    {
+      "epoch": 0.07212690122826322,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0002,
+      "loss": 1.5688,
+      "step": 3425
+    },
+    {
+      "epoch": 0.07214796017752695,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0002,
+      "loss": 1.5791,
+      "step": 3426
+    },
+    {
+      "epoch": 0.07216901912679066,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.5866,
+      "step": 3427
+    },
+    {
+      "epoch": 0.07219007807605439,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0002,
+      "loss": 1.5786,
+      "step": 3428
+    },
+    {
+      "epoch": 0.07221113702531812,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5435,
+      "step": 3429
+    },
+    {
+      "epoch": 0.07223219597458184,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.587,
+      "step": 3430
+    },
+    {
+      "epoch": 0.07225325492384557,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.5762,
+      "step": 3431
+    },
+    {
+      "epoch": 0.0722743138731093,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.634,
+      "step": 3432
+    },
+    {
+      "epoch": 0.07229537282237303,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.5788,
+      "step": 3433
+    },
+    {
+      "epoch": 0.07231643177163675,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0002,
+      "loss": 1.5792,
+      "step": 3434
+    },
+    {
+      "epoch": 0.07233749072090048,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.5701,
+      "step": 3435
+    },
+    {
+      "epoch": 0.07235854967016421,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0002,
+      "loss": 1.6185,
+      "step": 3436
+    },
+    {
+      "epoch": 0.07237960861942794,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0002,
+      "loss": 1.5845,
+      "step": 3437
+    },
+    {
+      "epoch": 0.07240066756869167,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.5587,
+      "step": 3438
+    },
+    {
+      "epoch": 0.07242172651795539,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0002,
+      "loss": 1.6045,
+      "step": 3439
+    },
+    {
+      "epoch": 0.0724427854672191,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.6058,
+      "step": 3440
+    },
+    {
+      "epoch": 0.07246384441648283,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.5625,
+      "step": 3441
+    },
+    {
+      "epoch": 0.07248490336574656,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.5778,
+      "step": 3442
+    },
+    {
+      "epoch": 0.07250596231501029,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0002,
+      "loss": 1.5831,
+      "step": 3443
+    },
+    {
+      "epoch": 0.07252702126427402,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0002,
+      "loss": 1.5807,
+      "step": 3444
+    },
+    {
+      "epoch": 0.07254808021353774,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0002,
+      "loss": 1.5553,
+      "step": 3445
+    },
+    {
+      "epoch": 0.07256913916280147,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.0002,
+      "loss": 1.5624,
+      "step": 3446
+    },
+    {
+      "epoch": 0.0725901981120652,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0002,
+      "loss": 1.5435,
+      "step": 3447
+    },
+    {
+      "epoch": 0.07261125706132893,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5991,
+      "step": 3448
+    },
+    {
+      "epoch": 0.07263231601059265,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.6157,
+      "step": 3449
+    },
+    {
+      "epoch": 0.07265337495985638,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.5991,
+      "step": 3450
+    },
+    {
+      "epoch": 0.07267443390912011,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.5931,
+      "step": 3451
+    },
+    {
+      "epoch": 0.07269549285838384,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.6002,
+      "step": 3452
+    },
+    {
+      "epoch": 0.07271655180764756,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.5955,
+      "step": 3453
+    },
+    {
+      "epoch": 0.07273761075691128,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0002,
+      "loss": 1.5912,
+      "step": 3454
+    },
+    {
+      "epoch": 0.072758669706175,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0002,
+      "loss": 1.5777,
+      "step": 3455
+    },
+    {
+      "epoch": 0.07277972865543873,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5973,
+      "step": 3456
+    },
+    {
+      "epoch": 0.07280078760470246,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.5901,
+      "step": 3457
+    },
+    {
+      "epoch": 0.07282184655396619,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6113,
+      "step": 3458
+    },
+    {
+      "epoch": 0.07284290550322992,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0002,
+      "loss": 1.5924,
+      "step": 3459
+    },
+    {
+      "epoch": 0.07286396445249364,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0002,
+      "loss": 1.6108,
+      "step": 3460
+    },
+    {
+      "epoch": 0.07288502340175737,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.5556,
+      "step": 3461
+    },
+    {
+      "epoch": 0.0729060823510211,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002,
+      "loss": 1.581,
+      "step": 3462
+    },
+    {
+      "epoch": 0.07292714130028483,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.6052,
+      "step": 3463
+    },
+    {
+      "epoch": 0.07294820024954855,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0002,
+      "loss": 1.6085,
+      "step": 3464
+    },
+    {
+      "epoch": 0.07296925919881228,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002,
+      "loss": 1.5569,
+      "step": 3465
+    },
+    {
+      "epoch": 0.07299031814807601,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0002,
+      "loss": 1.615,
+      "step": 3466
+    },
+    {
+      "epoch": 0.07301137709733972,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.6093,
+      "step": 3467
+    },
+    {
+      "epoch": 0.07303243604660345,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.586,
+      "step": 3468
+    },
+    {
+      "epoch": 0.07305349499586718,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.5704,
+      "step": 3469
+    },
+    {
+      "epoch": 0.0730745539451309,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0002,
+      "loss": 1.6023,
+      "step": 3470
+    },
+    {
+      "epoch": 0.07309561289439463,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5889,
+      "step": 3471
+    },
+    {
+      "epoch": 0.07311667184365836,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.5746,
+      "step": 3472
+    },
+    {
+      "epoch": 0.07313773079292209,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0002,
+      "loss": 1.6073,
+      "step": 3473
+    },
+    {
+      "epoch": 0.07315878974218581,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6115,
+      "step": 3474
+    },
+    {
+      "epoch": 0.07317984869144954,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.5391,
+      "step": 3475
+    },
+    {
+      "epoch": 0.07320090764071327,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5845,
+      "step": 3476
+    },
+    {
+      "epoch": 0.073221966589977,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.6046,
+      "step": 3477
+    },
+    {
+      "epoch": 0.07324302553924072,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0002,
+      "loss": 1.5583,
+      "step": 3478
+    },
+    {
+      "epoch": 0.07326408448850445,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0002,
+      "loss": 1.636,
+      "step": 3479
+    },
+    {
+      "epoch": 0.07328514343776817,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.5836,
+      "step": 3480
+    },
+    {
+      "epoch": 0.0733062023870319,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0002,
+      "loss": 1.594,
+      "step": 3481
+    },
+    {
+      "epoch": 0.07332726133629562,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002,
+      "loss": 1.5718,
+      "step": 3482
+    },
+    {
+      "epoch": 0.07334832028555935,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.5682,
+      "step": 3483
+    },
+    {
+      "epoch": 0.07336937923482308,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.5765,
+      "step": 3484
+    },
+    {
+      "epoch": 0.0733904381840868,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.6066,
+      "step": 3485
+    },
+    {
+      "epoch": 0.07341149713335053,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.5663,
+      "step": 3486
+    },
+    {
+      "epoch": 0.07343255608261426,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0002,
+      "loss": 1.5665,
+      "step": 3487
+    },
+    {
+      "epoch": 0.07345361503187799,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.6001,
+      "step": 3488
+    },
+    {
+      "epoch": 0.07347467398114171,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.6255,
+      "step": 3489
+    },
+    {
+      "epoch": 0.07349573293040544,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.5969,
+      "step": 3490
+    },
+    {
+      "epoch": 0.07351679187966917,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.5601,
+      "step": 3491
+    },
+    {
+      "epoch": 0.0735378508289329,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0002,
+      "loss": 1.5614,
+      "step": 3492
+    },
+    {
+      "epoch": 0.07355890977819661,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5852,
+      "step": 3493
+    },
+    {
+      "epoch": 0.07357996872746034,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.5937,
+      "step": 3494
+    },
+    {
+      "epoch": 0.07360102767672407,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.0002,
+      "loss": 1.596,
+      "step": 3495
+    },
+    {
+      "epoch": 0.07362208662598779,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.6275,
+      "step": 3496
+    },
+    {
+      "epoch": 0.07364314557525152,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.5641,
+      "step": 3497
+    },
+    {
+      "epoch": 0.07366420452451525,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.6325,
+      "step": 3498
+    },
+    {
+      "epoch": 0.07368526347377898,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0002,
+      "loss": 1.5639,
+      "step": 3499
+    },
+    {
+      "epoch": 0.0737063224230427,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0002,
+      "loss": 1.5621,
+      "step": 3500
+    },
+    {
+      "epoch": 0.0737063224230427,
+      "eval_loss": 2.018354892730713,
+      "eval_runtime": 897.5659,
+      "eval_samples_per_second": 68.853,
+      "eval_steps_per_second": 2.152,
+      "step": 3500
+    },
+    {
+      "epoch": 0.07372738137230643,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6097,
+      "step": 3501
+    },
+    {
+      "epoch": 0.07374844032157016,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.587,
+      "step": 3502
+    },
+    {
+      "epoch": 0.07376949927083389,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.6112,
+      "step": 3503
+    },
+    {
+      "epoch": 0.07379055822009761,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.5799,
+      "step": 3504
+    },
+    {
+      "epoch": 0.07381161716936134,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0002,
+      "loss": 1.6066,
+      "step": 3505
+    },
+    {
+      "epoch": 0.07383267611862507,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002,
+      "loss": 1.5433,
+      "step": 3506
+    },
+    {
+      "epoch": 0.07385373506788878,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.5748,
+      "step": 3507
+    },
+    {
+      "epoch": 0.07387479401715251,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.0002,
+      "loss": 1.6106,
+      "step": 3508
+    },
+    {
+      "epoch": 0.07389585296641624,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.6073,
+      "step": 3509
+    },
+    {
+      "epoch": 0.07391691191567996,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.5724,
+      "step": 3510
+    },
+    {
+      "epoch": 0.07393797086494369,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5596,
+      "step": 3511
+    },
+    {
+      "epoch": 0.07395902981420742,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0002,
+      "loss": 1.5845,
+      "step": 3512
+    },
+    {
+      "epoch": 0.07398008876347115,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0002,
+      "loss": 1.6012,
+      "step": 3513
+    },
+    {
+      "epoch": 0.07400114771273487,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0002,
+      "loss": 1.573,
+      "step": 3514
+    },
+    {
+      "epoch": 0.0740222066619986,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6054,
+      "step": 3515
+    },
+    {
+      "epoch": 0.07404326561126233,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.5982,
+      "step": 3516
+    },
+    {
+      "epoch": 0.07406432456052606,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.6318,
+      "step": 3517
+    },
+    {
+      "epoch": 0.07408538350978978,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0002,
+      "loss": 1.6141,
+      "step": 3518
+    },
+    {
+      "epoch": 0.07410644245905351,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0002,
+      "loss": 1.6044,
+      "step": 3519
+    },
+    {
+      "epoch": 0.07412750140831723,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.5955,
+      "step": 3520
+    },
+    {
+      "epoch": 0.07414856035758095,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0002,
+      "loss": 1.6132,
+      "step": 3521
+    },
+    {
+      "epoch": 0.07416961930684468,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.634,
+      "step": 3522
+    },
+    {
+      "epoch": 0.07419067825610841,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0002,
+      "loss": 1.6095,
+      "step": 3523
+    },
+    {
+      "epoch": 0.07421173720537214,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.6186,
+      "step": 3524
+    },
+    {
+      "epoch": 0.07423279615463586,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0002,
+      "loss": 1.5757,
+      "step": 3525
+    },
+    {
+      "epoch": 0.07425385510389959,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0002,
+      "loss": 1.5993,
+      "step": 3526
+    },
+    {
+      "epoch": 0.07427491405316332,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0002,
+      "loss": 1.5753,
+      "step": 3527
+    },
+    {
+      "epoch": 0.07429597300242705,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.6104,
+      "step": 3528
+    },
+    {
+      "epoch": 0.07431703195169077,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.5885,
+      "step": 3529
+    },
+    {
+      "epoch": 0.0743380909009545,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.6155,
+      "step": 3530
+    },
+    {
+      "epoch": 0.07435914985021823,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0002,
+      "loss": 1.5737,
+      "step": 3531
+    },
+    {
+      "epoch": 0.07438020879948196,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0002,
+      "loss": 1.5877,
+      "step": 3532
+    },
+    {
+      "epoch": 0.07440126774874567,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0002,
+      "loss": 1.581,
+      "step": 3533
+    },
+    {
+      "epoch": 0.0744223266980094,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0002,
+      "loss": 1.6153,
+      "step": 3534
+    },
+    {
+      "epoch": 0.07444338564727312,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0002,
+      "loss": 1.5776,
+      "step": 3535
+    },
+    {
+      "epoch": 0.07446444459653685,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0002,
+      "loss": 1.6121,
+      "step": 3536
+    },
+    {
+      "epoch": 0.07448550354580058,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002,
+      "loss": 1.5517,
+      "step": 3537
+    },
+    {
+      "epoch": 0.07450656249506431,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.5706,
+      "step": 3538
+    },
+    {
+      "epoch": 0.07452762144432803,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0002,
+      "loss": 1.5925,
+      "step": 3539
+    },
+    {
+      "epoch": 0.07454868039359176,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.5817,
+      "step": 3540
+    },
+    {
+      "epoch": 0.07456973934285549,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0002,
+      "loss": 1.5599,
+      "step": 3541
+    },
+    {
+      "epoch": 0.07459079829211922,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0002,
+      "loss": 1.6037,
+      "step": 3542
+    },
+    {
+      "epoch": 0.07461185724138295,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0002,
+      "loss": 1.6374,
+      "step": 3543
+    },
+    {
+      "epoch": 0.07463291619064667,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.6131,
+      "step": 3544
+    },
+    {
+      "epoch": 0.0746539751399104,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.5945,
+      "step": 3545
+    },
+    {
+      "epoch": 0.07467503408917411,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0002,
+      "loss": 1.5538,
+      "step": 3546
+    },
+    {
+      "epoch": 0.07469609303843784,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0002,
+      "loss": 1.6129,
+      "step": 3547
+    },
+    {
+      "epoch": 0.07471715198770157,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0002,
+      "loss": 1.5956,
+      "step": 3548
+    },
+    {
+      "epoch": 0.0747382109369653,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.5839,
+      "step": 3549
+    },
+    {
+      "epoch": 0.07475926988622902,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5767,
+      "step": 3550
+    },
+    {
+      "epoch": 0.07478032883549275,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.5561,
+      "step": 3551
+    },
+    {
+      "epoch": 0.07480138778475648,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.5828,
+      "step": 3552
+    },
+    {
+      "epoch": 0.0748224467340202,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0002,
+      "loss": 1.5926,
+      "step": 3553
+    },
+    {
+      "epoch": 0.07484350568328393,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.5643,
+      "step": 3554
+    },
+    {
+      "epoch": 0.07486456463254766,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002,
+      "loss": 1.5759,
+      "step": 3555
+    },
+    {
+      "epoch": 0.07488562358181139,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.5893,
+      "step": 3556
+    },
+    {
+      "epoch": 0.07490668253107512,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5945,
+      "step": 3557
+    },
+    {
+      "epoch": 0.07492774148033884,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0002,
+      "loss": 1.5925,
+      "step": 3558
+    },
+    {
+      "epoch": 0.07494880042960257,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.5613,
+      "step": 3559
+    },
+    {
+      "epoch": 0.07496985937886629,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.5672,
+      "step": 3560
+    },
+    {
+      "epoch": 0.07499091832813001,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0002,
+      "loss": 1.5701,
+      "step": 3561
+    },
+    {
+      "epoch": 0.07501197727739374,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0002,
+      "loss": 1.6199,
+      "step": 3562
+    },
+    {
+      "epoch": 0.07503303622665747,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.6131,
+      "step": 3563
+    },
+    {
+      "epoch": 0.0750540951759212,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.0002,
+      "loss": 1.5747,
+      "step": 3564
+    },
+    {
+      "epoch": 0.07507515412518492,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5847,
+      "step": 3565
+    },
+    {
+      "epoch": 0.07509621307444865,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.6031,
+      "step": 3566
+    },
+    {
+      "epoch": 0.07511727202371238,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.5539,
+      "step": 3567
+    },
+    {
+      "epoch": 0.0751383309729761,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5912,
+      "step": 3568
+    },
+    {
+      "epoch": 0.07515938992223983,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.5981,
+      "step": 3569
+    },
+    {
+      "epoch": 0.07518044887150356,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.5391,
+      "step": 3570
+    },
+    {
+      "epoch": 0.07520150782076729,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.5792,
+      "step": 3571
+    },
+    {
+      "epoch": 0.07522256677003102,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0002,
+      "loss": 1.5544,
+      "step": 3572
+    },
+    {
+      "epoch": 0.07524362571929473,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0002,
+      "loss": 1.5911,
+      "step": 3573
+    },
+    {
+      "epoch": 0.07526468466855846,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002,
+      "loss": 1.6121,
+      "step": 3574
+    },
+    {
+      "epoch": 0.07528574361782218,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0002,
+      "loss": 1.5803,
+      "step": 3575
+    },
+    {
+      "epoch": 0.07530680256708591,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5749,
+      "step": 3576
+    },
+    {
+      "epoch": 0.07532786151634964,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0002,
+      "loss": 1.5404,
+      "step": 3577
+    },
+    {
+      "epoch": 0.07534892046561337,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.536,
+      "step": 3578
+    },
+    {
+      "epoch": 0.0753699794148771,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5996,
+      "step": 3579
+    },
+    {
+      "epoch": 0.07539103836414082,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5857,
+      "step": 3580
+    },
+    {
+      "epoch": 0.07541209731340455,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.5584,
+      "step": 3581
+    },
+    {
+      "epoch": 0.07543315626266828,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.5525,
+      "step": 3582
+    },
+    {
+      "epoch": 0.075454215211932,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.5753,
+      "step": 3583
+    },
+    {
+      "epoch": 0.07547527416119573,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0002,
+      "loss": 1.6083,
+      "step": 3584
+    },
+    {
+      "epoch": 0.07549633311045946,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002,
+      "loss": 1.6094,
+      "step": 3585
+    },
+    {
+      "epoch": 0.07551739205972317,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.5954,
+      "step": 3586
+    },
+    {
+      "epoch": 0.0755384510089869,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0002,
+      "loss": 1.5987,
+      "step": 3587
+    },
+    {
+      "epoch": 0.07555950995825063,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0002,
+      "loss": 1.5873,
+      "step": 3588
+    },
+    {
+      "epoch": 0.07558056890751436,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.6274,
+      "step": 3589
+    },
+    {
+      "epoch": 0.07560162785677808,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.551,
+      "step": 3590
+    },
+    {
+      "epoch": 0.07562268680604181,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002,
+      "loss": 1.6023,
+      "step": 3591
+    },
+    {
+      "epoch": 0.07564374575530554,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.5997,
+      "step": 3592
+    },
+    {
+      "epoch": 0.07566480470456927,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.0002,
+      "loss": 1.5825,
+      "step": 3593
+    },
+    {
+      "epoch": 0.075685863653833,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.0002,
+      "loss": 1.5779,
+      "step": 3594
+    },
+    {
+      "epoch": 0.07570692260309672,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.5499,
+      "step": 3595
+    },
+    {
+      "epoch": 0.07572798155236045,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.6006,
+      "step": 3596
+    },
+    {
+      "epoch": 0.07574904050162418,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5951,
+      "step": 3597
+    },
+    {
+      "epoch": 0.0757700994508879,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5702,
+      "step": 3598
+    },
+    {
+      "epoch": 0.07579115840015162,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.5896,
+      "step": 3599
+    },
+    {
+      "epoch": 0.07581221734941535,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.5758,
+      "step": 3600
+    },
+    {
+      "epoch": 0.07583327629867907,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.5685,
+      "step": 3601
+    },
+    {
+      "epoch": 0.0758543352479428,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0002,
+      "loss": 1.6113,
+      "step": 3602
+    },
+    {
+      "epoch": 0.07587539419720653,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5567,
+      "step": 3603
+    },
+    {
+      "epoch": 0.07589645314647026,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0002,
+      "loss": 1.5755,
+      "step": 3604
+    },
+    {
+      "epoch": 0.07591751209573398,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5833,
+      "step": 3605
+    },
+    {
+      "epoch": 0.07593857104499771,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.5932,
+      "step": 3606
+    },
+    {
+      "epoch": 0.07595962999426144,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.5823,
+      "step": 3607
+    },
+    {
+      "epoch": 0.07598068894352517,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.576,
+      "step": 3608
+    },
+    {
+      "epoch": 0.07600174789278889,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0002,
+      "loss": 1.5927,
+      "step": 3609
+    },
+    {
+      "epoch": 0.07602280684205262,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5696,
+      "step": 3610
+    },
+    {
+      "epoch": 0.07604386579131635,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.6064,
+      "step": 3611
+    },
+    {
+      "epoch": 0.07606492474058008,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002,
+      "loss": 1.6075,
+      "step": 3612
+    },
+    {
+      "epoch": 0.07608598368984379,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0002,
+      "loss": 1.5595,
+      "step": 3613
+    },
+    {
+      "epoch": 0.07610704263910752,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.5595,
+      "step": 3614
+    },
+    {
+      "epoch": 0.07612810158837124,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.5954,
+      "step": 3615
+    },
+    {
+      "epoch": 0.07614916053763497,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.6064,
+      "step": 3616
+    },
+    {
+      "epoch": 0.0761702194868987,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.5926,
+      "step": 3617
+    },
+    {
+      "epoch": 0.07619127843616243,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5642,
+      "step": 3618
+    },
+    {
+      "epoch": 0.07621233738542615,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6129,
+      "step": 3619
+    },
+    {
+      "epoch": 0.07623339633468988,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002,
+      "loss": 1.5795,
+      "step": 3620
+    },
+    {
+      "epoch": 0.07625445528395361,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.5812,
+      "step": 3621
+    },
+    {
+      "epoch": 0.07627551423321734,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.5679,
+      "step": 3622
+    },
+    {
+      "epoch": 0.07629657318248106,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.5766,
+      "step": 3623
+    },
+    {
+      "epoch": 0.07631763213174479,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0002,
+      "loss": 1.608,
+      "step": 3624
+    },
+    {
+      "epoch": 0.07633869108100852,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0002,
+      "loss": 1.591,
+      "step": 3625
+    },
+    {
+      "epoch": 0.07635975003027223,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0002,
+      "loss": 1.5778,
+      "step": 3626
+    },
+    {
+      "epoch": 0.07638080897953596,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.5805,
+      "step": 3627
+    },
+    {
+      "epoch": 0.07640186792879969,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.5558,
+      "step": 3628
+    },
+    {
+      "epoch": 0.07642292687806342,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0002,
+      "loss": 1.605,
+      "step": 3629
+    },
+    {
+      "epoch": 0.07644398582732714,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.5889,
+      "step": 3630
+    },
+    {
+      "epoch": 0.07646504477659087,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.601,
+      "step": 3631
+    },
+    {
+      "epoch": 0.0764861037258546,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0002,
+      "loss": 1.5688,
+      "step": 3632
+    },
+    {
+      "epoch": 0.07650716267511833,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.5878,
+      "step": 3633
+    },
+    {
+      "epoch": 0.07652822162438205,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.5862,
+      "step": 3634
+    },
+    {
+      "epoch": 0.07654928057364578,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5576,
+      "step": 3635
+    },
+    {
+      "epoch": 0.07657033952290951,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.5689,
+      "step": 3636
+    },
+    {
+      "epoch": 0.07659139847217324,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.5765,
+      "step": 3637
+    },
+    {
+      "epoch": 0.07661245742143696,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0002,
+      "loss": 1.596,
+      "step": 3638
+    },
+    {
+      "epoch": 0.07663351637070068,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6129,
+      "step": 3639
+    },
+    {
+      "epoch": 0.0766545753199644,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.5989,
+      "step": 3640
+    },
+    {
+      "epoch": 0.07667563426922813,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0002,
+      "loss": 1.6102,
+      "step": 3641
+    },
+    {
+      "epoch": 0.07669669321849186,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.5825,
+      "step": 3642
+    },
+    {
+      "epoch": 0.07671775216775559,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.5988,
+      "step": 3643
+    },
+    {
+      "epoch": 0.07673881111701931,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.6051,
+      "step": 3644
+    },
+    {
+      "epoch": 0.07675987006628304,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.5876,
+      "step": 3645
+    },
+    {
+      "epoch": 0.07678092901554677,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.5593,
+      "step": 3646
+    },
+    {
+      "epoch": 0.0768019879648105,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0002,
+      "loss": 1.5714,
+      "step": 3647
+    },
+    {
+      "epoch": 0.07682304691407423,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6031,
+      "step": 3648
+    },
+    {
+      "epoch": 0.07684410586333795,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.5732,
+      "step": 3649
+    },
+    {
+      "epoch": 0.07686516481260168,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0002,
+      "loss": 1.5712,
+      "step": 3650
+    },
+    {
+      "epoch": 0.07688622376186541,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.5885,
+      "step": 3651
+    },
+    {
+      "epoch": 0.07690728271112912,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0002,
+      "loss": 1.633,
+      "step": 3652
+    },
+    {
+      "epoch": 0.07692834166039285,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5984,
+      "step": 3653
+    },
+    {
+      "epoch": 0.07694940060965658,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0002,
+      "loss": 1.6018,
+      "step": 3654
+    },
+    {
+      "epoch": 0.0769704595589203,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6076,
+      "step": 3655
+    },
+    {
+      "epoch": 0.07699151850818403,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5681,
+      "step": 3656
+    },
+    {
+      "epoch": 0.07701257745744776,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0002,
+      "loss": 1.59,
+      "step": 3657
+    },
+    {
+      "epoch": 0.07703363640671149,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.575,
+      "step": 3658
+    },
+    {
+      "epoch": 0.07705469535597521,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.5718,
+      "step": 3659
+    },
+    {
+      "epoch": 0.07707575430523894,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.5835,
+      "step": 3660
+    },
+    {
+      "epoch": 0.07709681325450267,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0002,
+      "loss": 1.5651,
+      "step": 3661
+    },
+    {
+      "epoch": 0.0771178722037664,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5599,
+      "step": 3662
+    },
+    {
+      "epoch": 0.07713893115303012,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.5873,
+      "step": 3663
+    },
+    {
+      "epoch": 0.07715999010229385,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0002,
+      "loss": 1.6038,
+      "step": 3664
+    },
+    {
+      "epoch": 0.07718104905155758,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0002,
+      "loss": 1.5761,
+      "step": 3665
+    },
+    {
+      "epoch": 0.07720210800082129,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0002,
+      "loss": 1.6053,
+      "step": 3666
+    },
+    {
+      "epoch": 0.07722316695008502,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.5485,
+      "step": 3667
+    },
+    {
+      "epoch": 0.07724422589934875,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5739,
+      "step": 3668
+    },
+    {
+      "epoch": 0.07726528484861248,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5713,
+      "step": 3669
+    },
+    {
+      "epoch": 0.0772863437978762,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6199,
+      "step": 3670
+    },
+    {
+      "epoch": 0.07730740274713993,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.6088,
+      "step": 3671
+    },
+    {
+      "epoch": 0.07732846169640366,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0002,
+      "loss": 1.5761,
+      "step": 3672
+    },
+    {
+      "epoch": 0.07734952064566739,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.6056,
+      "step": 3673
+    },
+    {
+      "epoch": 0.07737057959493111,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002,
+      "loss": 1.5823,
+      "step": 3674
+    },
+    {
+      "epoch": 0.07739163854419484,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0002,
+      "loss": 1.5945,
+      "step": 3675
+    },
+    {
+      "epoch": 0.07741269749345857,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.0002,
+      "loss": 1.6409,
+      "step": 3676
+    },
+    {
+      "epoch": 0.0774337564427223,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0002,
+      "loss": 1.6021,
+      "step": 3677
+    },
+    {
+      "epoch": 0.07745481539198602,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0002,
+      "loss": 1.5784,
+      "step": 3678
+    },
+    {
+      "epoch": 0.07747587434124974,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.5877,
+      "step": 3679
+    },
+    {
+      "epoch": 0.07749693329051346,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0002,
+      "loss": 1.5781,
+      "step": 3680
+    },
+    {
+      "epoch": 0.07751799223977719,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.5993,
+      "step": 3681
+    },
+    {
+      "epoch": 0.07753905118904092,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0002,
+      "loss": 1.5975,
+      "step": 3682
+    },
+    {
+      "epoch": 0.07756011013830465,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.6273,
+      "step": 3683
+    },
+    {
+      "epoch": 0.07758116908756837,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0002,
+      "loss": 1.5863,
+      "step": 3684
+    },
+    {
+      "epoch": 0.0776022280368321,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0002,
+      "loss": 1.5913,
+      "step": 3685
+    },
+    {
+      "epoch": 0.07762328698609583,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0002,
+      "loss": 1.5759,
+      "step": 3686
+    },
+    {
+      "epoch": 0.07764434593535956,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.5976,
+      "step": 3687
+    },
+    {
+      "epoch": 0.07766540488462328,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.591,
+      "step": 3688
+    },
+    {
+      "epoch": 0.07768646383388701,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002,
+      "loss": 1.5528,
+      "step": 3689
+    },
+    {
+      "epoch": 0.07770752278315074,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002,
+      "loss": 1.5809,
+      "step": 3690
+    },
+    {
+      "epoch": 0.07772858173241447,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.5863,
+      "step": 3691
+    },
+    {
+      "epoch": 0.07774964068167818,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0002,
+      "loss": 1.5802,
+      "step": 3692
+    },
+    {
+      "epoch": 0.07777069963094191,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.6001,
+      "step": 3693
+    },
+    {
+      "epoch": 0.07779175858020564,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.6108,
+      "step": 3694
+    },
+    {
+      "epoch": 0.07781281752946936,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5979,
+      "step": 3695
+    },
+    {
+      "epoch": 0.07783387647873309,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6082,
+      "step": 3696
+    },
+    {
+      "epoch": 0.07785493542799682,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.6079,
+      "step": 3697
+    },
+    {
+      "epoch": 0.07787599437726055,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.0002,
+      "loss": 1.5785,
+      "step": 3698
+    },
+    {
+      "epoch": 0.07789705332652427,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.5612,
+      "step": 3699
+    },
+    {
+      "epoch": 0.077918112275788,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5838,
+      "step": 3700
+    },
+    {
+      "epoch": 0.07793917122505173,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0002,
+      "loss": 1.5769,
+      "step": 3701
+    },
+    {
+      "epoch": 0.07796023017431546,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.6104,
+      "step": 3702
+    },
+    {
+      "epoch": 0.07798128912357918,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0002,
+      "loss": 1.576,
+      "step": 3703
+    },
+    {
+      "epoch": 0.07800234807284291,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.6191,
+      "step": 3704
+    },
+    {
+      "epoch": 0.07802340702210664,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.5742,
+      "step": 3705
+    },
+    {
+      "epoch": 0.07804446597137035,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.5669,
+      "step": 3706
+    },
+    {
+      "epoch": 0.07806552492063408,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0002,
+      "loss": 1.5614,
+      "step": 3707
+    },
+    {
+      "epoch": 0.07808658386989781,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.6113,
+      "step": 3708
+    },
+    {
+      "epoch": 0.07810764281916154,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0002,
+      "loss": 1.6016,
+      "step": 3709
+    },
+    {
+      "epoch": 0.07812870176842526,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.0002,
+      "loss": 1.5706,
+      "step": 3710
+    },
+    {
+      "epoch": 0.07814976071768899,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.5678,
+      "step": 3711
+    },
+    {
+      "epoch": 0.07817081966695272,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.5952,
+      "step": 3712
+    },
+    {
+      "epoch": 0.07819187861621645,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.6148,
+      "step": 3713
+    },
+    {
+      "epoch": 0.07821293756548017,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.5946,
+      "step": 3714
+    },
+    {
+      "epoch": 0.0782339965147439,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5859,
+      "step": 3715
+    },
+    {
+      "epoch": 0.07825505546400763,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.578,
+      "step": 3716
+    },
+    {
+      "epoch": 0.07827611441327136,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0002,
+      "loss": 1.5699,
+      "step": 3717
+    },
+    {
+      "epoch": 0.07829717336253508,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0002,
+      "loss": 1.5702,
+      "step": 3718
+    },
+    {
+      "epoch": 0.0783182323117988,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0002,
+      "loss": 1.5995,
+      "step": 3719
+    },
+    {
+      "epoch": 0.07833929126106252,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.5918,
+      "step": 3720
+    },
+    {
+      "epoch": 0.07836035021032625,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0002,
+      "loss": 1.5779,
+      "step": 3721
+    },
+    {
+      "epoch": 0.07838140915958998,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0002,
+      "loss": 1.5815,
+      "step": 3722
+    },
+    {
+      "epoch": 0.0784024681088537,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0002,
+      "loss": 1.5629,
+      "step": 3723
+    },
+    {
+      "epoch": 0.07842352705811743,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.616,
+      "step": 3724
+    },
+    {
+      "epoch": 0.07844458600738116,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.6054,
+      "step": 3725
+    },
+    {
+      "epoch": 0.07846564495664489,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.5746,
+      "step": 3726
+    },
+    {
+      "epoch": 0.07848670390590862,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5866,
+      "step": 3727
+    },
+    {
+      "epoch": 0.07850776285517234,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0002,
+      "loss": 1.5859,
+      "step": 3728
+    },
+    {
+      "epoch": 0.07852882180443607,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0002,
+      "loss": 1.5515,
+      "step": 3729
+    },
+    {
+      "epoch": 0.0785498807536998,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0002,
+      "loss": 1.585,
+      "step": 3730
+    },
+    {
+      "epoch": 0.07857093970296353,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.5717,
+      "step": 3731
+    },
+    {
+      "epoch": 0.07859199865222724,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002,
+      "loss": 1.5535,
+      "step": 3732
+    },
+    {
+      "epoch": 0.07861305760149097,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.596,
+      "step": 3733
+    },
+    {
+      "epoch": 0.0786341165507547,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.0002,
+      "loss": 1.58,
+      "step": 3734
+    },
+    {
+      "epoch": 0.07865517550001842,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0002,
+      "loss": 1.5906,
+      "step": 3735
+    },
+    {
+      "epoch": 0.07867623444928215,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0002,
+      "loss": 1.6056,
+      "step": 3736
+    },
+    {
+      "epoch": 0.07869729339854588,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5606,
+      "step": 3737
+    },
+    {
+      "epoch": 0.0787183523478096,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.5969,
+      "step": 3738
+    },
+    {
+      "epoch": 0.07873941129707333,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.5736,
+      "step": 3739
+    },
+    {
+      "epoch": 0.07876047024633706,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0002,
+      "loss": 1.5978,
+      "step": 3740
+    },
+    {
+      "epoch": 0.07878152919560079,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0002,
+      "loss": 1.6203,
+      "step": 3741
+    },
+    {
+      "epoch": 0.07880258814486452,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.5674,
+      "step": 3742
+    },
+    {
+      "epoch": 0.07882364709412824,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.576,
+      "step": 3743
+    },
+    {
+      "epoch": 0.07884470604339197,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0002,
+      "loss": 1.5857,
+      "step": 3744
+    },
+    {
+      "epoch": 0.07886576499265568,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.5984,
+      "step": 3745
+    },
+    {
+      "epoch": 0.07888682394191941,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0002,
+      "loss": 1.5794,
+      "step": 3746
+    },
+    {
+      "epoch": 0.07890788289118314,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0002,
+      "loss": 1.5576,
+      "step": 3747
+    },
+    {
+      "epoch": 0.07892894184044687,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5756,
+      "step": 3748
+    },
+    {
+      "epoch": 0.0789500007897106,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0002,
+      "loss": 1.5734,
+      "step": 3749
+    },
+    {
+      "epoch": 0.07897105973897432,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.5882,
+      "step": 3750
+    },
+    {
+      "epoch": 0.07899211868823805,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0002,
+      "loss": 1.5962,
+      "step": 3751
+    },
+    {
+      "epoch": 0.07901317763750178,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0002,
+      "loss": 1.5565,
+      "step": 3752
+    },
+    {
+      "epoch": 0.0790342365867655,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.5731,
+      "step": 3753
+    },
+    {
+      "epoch": 0.07905529553602923,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0002,
+      "loss": 1.5806,
+      "step": 3754
+    },
+    {
+      "epoch": 0.07907635448529296,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.5807,
+      "step": 3755
+    },
+    {
+      "epoch": 0.07909741343455669,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.5608,
+      "step": 3756
+    },
+    {
+      "epoch": 0.07911847238382042,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.5828,
+      "step": 3757
+    },
+    {
+      "epoch": 0.07913953133308414,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.6233,
+      "step": 3758
+    },
+    {
+      "epoch": 0.07916059028234786,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0002,
+      "loss": 1.57,
+      "step": 3759
+    },
+    {
+      "epoch": 0.07918164923161158,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.573,
+      "step": 3760
+    },
+    {
+      "epoch": 0.07920270818087531,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002,
+      "loss": 1.6179,
+      "step": 3761
+    },
+    {
+      "epoch": 0.07922376713013904,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0002,
+      "loss": 1.6117,
+      "step": 3762
+    },
+    {
+      "epoch": 0.07924482607940277,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.5659,
+      "step": 3763
+    },
+    {
+      "epoch": 0.0792658850286665,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.5852,
+      "step": 3764
+    },
+    {
+      "epoch": 0.07928694397793022,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.6018,
+      "step": 3765
+    },
+    {
+      "epoch": 0.07930800292719395,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0002,
+      "loss": 1.5907,
+      "step": 3766
+    },
+    {
+      "epoch": 0.07932906187645768,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5704,
+      "step": 3767
+    },
+    {
+      "epoch": 0.0793501208257214,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0002,
+      "loss": 1.6199,
+      "step": 3768
+    },
+    {
+      "epoch": 0.07937117977498513,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0002,
+      "loss": 1.5887,
+      "step": 3769
+    },
+    {
+      "epoch": 0.07939223872424886,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5995,
+      "step": 3770
+    },
+    {
+      "epoch": 0.07941329767351259,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.5907,
+      "step": 3771
+    },
+    {
+      "epoch": 0.0794343566227763,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0002,
+      "loss": 1.5937,
+      "step": 3772
+    },
+    {
+      "epoch": 0.07945541557204003,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0002,
+      "loss": 1.6053,
+      "step": 3773
+    },
+    {
+      "epoch": 0.07947647452130376,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0002,
+      "loss": 1.5964,
+      "step": 3774
+    },
+    {
+      "epoch": 0.07949753347056748,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.0002,
+      "loss": 1.5861,
+      "step": 3775
+    },
+    {
+      "epoch": 0.07951859241983121,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.6002,
+      "step": 3776
+    },
+    {
+      "epoch": 0.07953965136909494,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0002,
+      "loss": 1.6099,
+      "step": 3777
+    },
+    {
+      "epoch": 0.07956071031835867,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0002,
+      "loss": 1.6108,
+      "step": 3778
+    },
+    {
+      "epoch": 0.0795817692676224,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.5484,
+      "step": 3779
+    },
+    {
+      "epoch": 0.07960282821688612,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0002,
+      "loss": 1.5812,
+      "step": 3780
+    },
+    {
+      "epoch": 0.07962388716614985,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.5864,
+      "step": 3781
+    },
+    {
+      "epoch": 0.07964494611541358,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5866,
+      "step": 3782
+    },
+    {
+      "epoch": 0.0796660050646773,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0002,
+      "loss": 1.5994,
+      "step": 3783
+    },
+    {
+      "epoch": 0.07968706401394103,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.6268,
+      "step": 3784
+    },
+    {
+      "epoch": 0.07970812296320474,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.5903,
+      "step": 3785
+    },
+    {
+      "epoch": 0.07972918191246847,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.5424,
+      "step": 3786
+    },
+    {
+      "epoch": 0.0797502408617322,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.5644,
+      "step": 3787
+    },
+    {
+      "epoch": 0.07977129981099593,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.5977,
+      "step": 3788
+    },
+    {
+      "epoch": 0.07979235876025965,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.5727,
+      "step": 3789
+    },
+    {
+      "epoch": 0.07981341770952338,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0002,
+      "loss": 1.5634,
+      "step": 3790
+    },
+    {
+      "epoch": 0.07983447665878711,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.5651,
+      "step": 3791
+    },
+    {
+      "epoch": 0.07985553560805084,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.6226,
+      "step": 3792
+    },
+    {
+      "epoch": 0.07987659455731456,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5542,
+      "step": 3793
+    },
+    {
+      "epoch": 0.07989765350657829,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0002,
+      "loss": 1.5956,
+      "step": 3794
+    },
+    {
+      "epoch": 0.07991871245584202,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.5789,
+      "step": 3795
+    },
+    {
+      "epoch": 0.07993977140510575,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.5623,
+      "step": 3796
+    },
+    {
+      "epoch": 0.07996083035436947,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.6089,
+      "step": 3797
+    },
+    {
+      "epoch": 0.07998188930363319,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.5873,
+      "step": 3798
+    },
+    {
+      "epoch": 0.08000294825289692,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.5929,
+      "step": 3799
+    },
+    {
+      "epoch": 0.08002400720216064,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.5595,
+      "step": 3800
+    },
+    {
+      "epoch": 0.08004506615142437,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0002,
+      "loss": 1.586,
+      "step": 3801
+    },
+    {
+      "epoch": 0.0800661251006881,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0002,
+      "loss": 1.5884,
+      "step": 3802
+    },
+    {
+      "epoch": 0.08008718404995183,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.5775,
+      "step": 3803
+    },
+    {
+      "epoch": 0.08010824299921555,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002,
+      "loss": 1.5825,
+      "step": 3804
+    },
+    {
+      "epoch": 0.08012930194847928,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.0002,
+      "loss": 1.5689,
+      "step": 3805
+    },
+    {
+      "epoch": 0.08015036089774301,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002,
+      "loss": 1.6086,
+      "step": 3806
+    },
+    {
+      "epoch": 0.08017141984700674,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0002,
+      "loss": 1.5863,
+      "step": 3807
+    },
+    {
+      "epoch": 0.08019247879627046,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.5761,
+      "step": 3808
+    },
+    {
+      "epoch": 0.08021353774553419,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.5858,
+      "step": 3809
+    },
+    {
+      "epoch": 0.08023459669479792,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.5851,
+      "step": 3810
+    },
+    {
+      "epoch": 0.08025565564406165,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0002,
+      "loss": 1.5996,
+      "step": 3811
+    },
+    {
+      "epoch": 0.08027671459332536,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5653,
+      "step": 3812
+    },
+    {
+      "epoch": 0.08029777354258909,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0002,
+      "loss": 1.6083,
+      "step": 3813
+    },
+    {
+      "epoch": 0.08031883249185282,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.5893,
+      "step": 3814
+    },
+    {
+      "epoch": 0.08033989144111654,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5803,
+      "step": 3815
+    },
+    {
+      "epoch": 0.08036095039038027,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.594,
+      "step": 3816
+    },
+    {
+      "epoch": 0.080382009339644,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.5744,
+      "step": 3817
+    },
+    {
+      "epoch": 0.08040306828890773,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0002,
+      "loss": 1.5767,
+      "step": 3818
+    },
+    {
+      "epoch": 0.08042412723817145,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.0002,
+      "loss": 1.5805,
+      "step": 3819
+    },
+    {
+      "epoch": 0.08044518618743518,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002,
+      "loss": 1.5754,
+      "step": 3820
+    },
+    {
+      "epoch": 0.08046624513669891,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0002,
+      "loss": 1.5991,
+      "step": 3821
+    },
+    {
+      "epoch": 0.08048730408596264,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6183,
+      "step": 3822
+    },
+    {
+      "epoch": 0.08050836303522636,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002,
+      "loss": 1.5916,
+      "step": 3823
+    },
+    {
+      "epoch": 0.08052942198449009,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0002,
+      "loss": 1.6063,
+      "step": 3824
+    },
+    {
+      "epoch": 0.0805504809337538,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.5825,
+      "step": 3825
+    },
+    {
+      "epoch": 0.08057153988301753,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.5571,
+      "step": 3826
+    },
+    {
+      "epoch": 0.08059259883228126,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0002,
+      "loss": 1.6108,
+      "step": 3827
+    },
+    {
+      "epoch": 0.08061365778154499,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0002,
+      "loss": 1.5882,
+      "step": 3828
+    },
+    {
+      "epoch": 0.08063471673080871,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.5555,
+      "step": 3829
+    },
+    {
+      "epoch": 0.08065577568007244,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.6218,
+      "step": 3830
+    },
+    {
+      "epoch": 0.08067683462933617,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6215,
+      "step": 3831
+    },
+    {
+      "epoch": 0.0806978935785999,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.6023,
+      "step": 3832
+    },
+    {
+      "epoch": 0.08071895252786362,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0002,
+      "loss": 1.5751,
+      "step": 3833
+    },
+    {
+      "epoch": 0.08074001147712735,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0002,
+      "loss": 1.5808,
+      "step": 3834
+    },
+    {
+      "epoch": 0.08076107042639108,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.0002,
+      "loss": 1.5703,
+      "step": 3835
+    },
+    {
+      "epoch": 0.08078212937565481,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.5615,
+      "step": 3836
+    },
+    {
+      "epoch": 0.08080318832491853,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0002,
+      "loss": 1.6016,
+      "step": 3837
+    },
+    {
+      "epoch": 0.08082424727418225,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.5814,
+      "step": 3838
+    },
+    {
+      "epoch": 0.08084530622344598,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0002,
+      "loss": 1.5893,
+      "step": 3839
+    },
+    {
+      "epoch": 0.0808663651727097,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.5915,
+      "step": 3840
+    },
+    {
+      "epoch": 0.08088742412197343,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.5764,
+      "step": 3841
+    },
+    {
+      "epoch": 0.08090848307123716,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5761,
+      "step": 3842
+    },
+    {
+      "epoch": 0.08092954202050089,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5667,
+      "step": 3843
+    },
+    {
+      "epoch": 0.08095060096976461,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0002,
+      "loss": 1.6,
+      "step": 3844
+    },
+    {
+      "epoch": 0.08097165991902834,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.0002,
+      "loss": 1.6131,
+      "step": 3845
+    },
+    {
+      "epoch": 0.08099271886829207,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0002,
+      "loss": 1.5961,
+      "step": 3846
+    },
+    {
+      "epoch": 0.0810137778175558,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0002,
+      "loss": 1.5825,
+      "step": 3847
+    },
+    {
+      "epoch": 0.08103483676681952,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.0002,
+      "loss": 1.5858,
+      "step": 3848
+    },
+    {
+      "epoch": 0.08105589571608325,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.5835,
+      "step": 3849
+    },
+    {
+      "epoch": 0.08107695466534698,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.5482,
+      "step": 3850
+    },
+    {
+      "epoch": 0.08109801361461069,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0002,
+      "loss": 1.6136,
+      "step": 3851
+    },
+    {
+      "epoch": 0.08111907256387442,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.584,
+      "step": 3852
+    },
+    {
+      "epoch": 0.08114013151313815,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.5347,
+      "step": 3853
+    },
+    {
+      "epoch": 0.08116119046240187,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.5785,
+      "step": 3854
+    },
+    {
+      "epoch": 0.0811822494116656,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.5818,
+      "step": 3855
+    },
+    {
+      "epoch": 0.08120330836092933,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0002,
+      "loss": 1.572,
+      "step": 3856
+    },
+    {
+      "epoch": 0.08122436731019306,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0002,
+      "loss": 1.5732,
+      "step": 3857
+    },
+    {
+      "epoch": 0.08124542625945678,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.5929,
+      "step": 3858
+    },
+    {
+      "epoch": 0.08126648520872051,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.587,
+      "step": 3859
+    },
+    {
+      "epoch": 0.08128754415798424,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.5869,
+      "step": 3860
+    },
+    {
+      "epoch": 0.08130860310724797,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0002,
+      "loss": 1.5925,
+      "step": 3861
+    },
+    {
+      "epoch": 0.0813296620565117,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.5722,
+      "step": 3862
+    },
+    {
+      "epoch": 0.08135072100577542,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.5834,
+      "step": 3863
+    },
+    {
+      "epoch": 0.08137177995503915,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.6138,
+      "step": 3864
+    },
+    {
+      "epoch": 0.08139283890430286,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0002,
+      "loss": 1.5931,
+      "step": 3865
+    },
+    {
+      "epoch": 0.08141389785356659,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.5812,
+      "step": 3866
+    },
+    {
+      "epoch": 0.08143495680283032,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0002,
+      "loss": 1.5822,
+      "step": 3867
+    },
+    {
+      "epoch": 0.08145601575209405,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.6299,
+      "step": 3868
+    },
+    {
+      "epoch": 0.08147707470135777,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0002,
+      "loss": 1.6222,
+      "step": 3869
+    },
+    {
+      "epoch": 0.0814981336506215,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0002,
+      "loss": 1.616,
+      "step": 3870
+    },
+    {
+      "epoch": 0.08151919259988523,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.5648,
+      "step": 3871
+    },
+    {
+      "epoch": 0.08154025154914896,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.587,
+      "step": 3872
+    },
+    {
+      "epoch": 0.08156131049841268,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.591,
+      "step": 3873
+    },
+    {
+      "epoch": 0.08158236944767641,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0002,
+      "loss": 1.545,
+      "step": 3874
+    },
+    {
+      "epoch": 0.08160342839694014,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0002,
+      "loss": 1.6478,
+      "step": 3875
+    },
+    {
+      "epoch": 0.08162448734620387,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.574,
+      "step": 3876
+    },
+    {
+      "epoch": 0.0816455462954676,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.5823,
+      "step": 3877
+    },
+    {
+      "epoch": 0.08166660524473131,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5827,
+      "step": 3878
+    },
+    {
+      "epoch": 0.08168766419399504,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.5786,
+      "step": 3879
+    },
+    {
+      "epoch": 0.08170872314325876,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0002,
+      "loss": 1.5983,
+      "step": 3880
+    },
+    {
+      "epoch": 0.08172978209252249,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5559,
+      "step": 3881
+    },
+    {
+      "epoch": 0.08175084104178622,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5897,
+      "step": 3882
+    },
+    {
+      "epoch": 0.08177189999104995,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.5838,
+      "step": 3883
+    },
+    {
+      "epoch": 0.08179295894031367,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0002,
+      "loss": 1.5518,
+      "step": 3884
+    },
+    {
+      "epoch": 0.0818140178895774,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0002,
+      "loss": 1.5711,
+      "step": 3885
+    },
+    {
+      "epoch": 0.08183507683884113,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.5927,
+      "step": 3886
+    },
+    {
+      "epoch": 0.08185613578810486,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.5575,
+      "step": 3887
+    },
+    {
+      "epoch": 0.08187719473736858,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.5799,
+      "step": 3888
+    },
+    {
+      "epoch": 0.08189825368663231,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.554,
+      "step": 3889
+    },
+    {
+      "epoch": 0.08191931263589604,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.6094,
+      "step": 3890
+    },
+    {
+      "epoch": 0.08194037158515975,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.5863,
+      "step": 3891
+    },
+    {
+      "epoch": 0.08196143053442348,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5628,
+      "step": 3892
+    },
+    {
+      "epoch": 0.08198248948368721,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.6101,
+      "step": 3893
+    },
+    {
+      "epoch": 0.08200354843295093,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0002,
+      "loss": 1.5377,
+      "step": 3894
+    },
+    {
+      "epoch": 0.08202460738221466,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.5739,
+      "step": 3895
+    },
+    {
+      "epoch": 0.08204566633147839,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.6073,
+      "step": 3896
+    },
+    {
+      "epoch": 0.08206672528074212,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0002,
+      "loss": 1.594,
+      "step": 3897
+    },
+    {
+      "epoch": 0.08208778423000584,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.5833,
+      "step": 3898
+    },
+    {
+      "epoch": 0.08210884317926957,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0002,
+      "loss": 1.6135,
+      "step": 3899
+    },
+    {
+      "epoch": 0.0821299021285333,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0002,
+      "loss": 1.5964,
+      "step": 3900
+    },
+    {
+      "epoch": 0.08215096107779703,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.5941,
+      "step": 3901
+    },
+    {
+      "epoch": 0.08217202002706075,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0002,
+      "loss": 1.6095,
+      "step": 3902
+    },
+    {
+      "epoch": 0.08219307897632448,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0002,
+      "loss": 1.6087,
+      "step": 3903
+    },
+    {
+      "epoch": 0.0822141379255882,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.5702,
+      "step": 3904
+    },
+    {
+      "epoch": 0.08223519687485192,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.5922,
+      "step": 3905
+    },
+    {
+      "epoch": 0.08225625582411565,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0002,
+      "loss": 1.5749,
+      "step": 3906
+    },
+    {
+      "epoch": 0.08227731477337938,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.5844,
+      "step": 3907
+    },
+    {
+      "epoch": 0.0822983737226431,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.5683,
+      "step": 3908
+    },
+    {
+      "epoch": 0.08231943267190683,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0002,
+      "loss": 1.5987,
+      "step": 3909
+    },
+    {
+      "epoch": 0.08234049162117056,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0002,
+      "loss": 1.6063,
+      "step": 3910
+    },
+    {
+      "epoch": 0.08236155057043429,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.572,
+      "step": 3911
+    },
+    {
+      "epoch": 0.08238260951969802,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.5661,
+      "step": 3912
+    },
+    {
+      "epoch": 0.08240366846896174,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6054,
+      "step": 3913
+    },
+    {
+      "epoch": 0.08242472741822547,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0002,
+      "loss": 1.5797,
+      "step": 3914
+    },
+    {
+      "epoch": 0.0824457863674892,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.5874,
+      "step": 3915
+    },
+    {
+      "epoch": 0.08246684531675293,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.5916,
+      "step": 3916
+    },
+    {
+      "epoch": 0.08248790426601665,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0002,
+      "loss": 1.6077,
+      "step": 3917
+    },
+    {
+      "epoch": 0.08250896321528037,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.593,
+      "step": 3918
+    },
+    {
+      "epoch": 0.0825300221645441,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.5967,
+      "step": 3919
+    },
+    {
+      "epoch": 0.08255108111380782,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.6174,
+      "step": 3920
+    },
+    {
+      "epoch": 0.08257214006307155,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0002,
+      "loss": 1.5941,
+      "step": 3921
+    },
+    {
+      "epoch": 0.08259319901233528,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0002,
+      "loss": 1.5828,
+      "step": 3922
+    },
+    {
+      "epoch": 0.082614257961599,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002,
+      "loss": 1.5557,
+      "step": 3923
+    },
+    {
+      "epoch": 0.08263531691086273,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.6207,
+      "step": 3924
+    },
+    {
+      "epoch": 0.08265637586012646,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.5856,
+      "step": 3925
+    },
+    {
+      "epoch": 0.08267743480939019,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.5842,
+      "step": 3926
+    },
+    {
+      "epoch": 0.08269849375865392,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.5838,
+      "step": 3927
+    },
+    {
+      "epoch": 0.08271955270791764,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.5953,
+      "step": 3928
+    },
+    {
+      "epoch": 0.08274061165718137,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.593,
+      "step": 3929
+    },
+    {
+      "epoch": 0.0827616706064451,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0002,
+      "loss": 1.5751,
+      "step": 3930
+    },
+    {
+      "epoch": 0.08278272955570881,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0002,
+      "loss": 1.5749,
+      "step": 3931
+    },
+    {
+      "epoch": 0.08280378850497254,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.0002,
+      "loss": 1.5896,
+      "step": 3932
+    },
+    {
+      "epoch": 0.08282484745423627,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.6058,
+      "step": 3933
+    },
+    {
+      "epoch": 0.0828459064035,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0002,
+      "loss": 1.6298,
+      "step": 3934
+    },
+    {
+      "epoch": 0.08286696535276372,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.5974,
+      "step": 3935
+    },
+    {
+      "epoch": 0.08288802430202745,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0002,
+      "loss": 1.5773,
+      "step": 3936
+    },
+    {
+      "epoch": 0.08290908325129118,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.5909,
+      "step": 3937
+    },
+    {
+      "epoch": 0.0829301422005549,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.5679,
+      "step": 3938
+    },
+    {
+      "epoch": 0.08295120114981863,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.6073,
+      "step": 3939
+    },
+    {
+      "epoch": 0.08297226009908236,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.5631,
+      "step": 3940
+    },
+    {
+      "epoch": 0.08299331904834609,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.5829,
+      "step": 3941
+    },
+    {
+      "epoch": 0.08301437799760981,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.5922,
+      "step": 3942
+    },
+    {
+      "epoch": 0.08303543694687354,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002,
+      "loss": 1.5625,
+      "step": 3943
+    },
+    {
+      "epoch": 0.08305649589613726,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.6263,
+      "step": 3944
+    },
+    {
+      "epoch": 0.08307755484540098,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.6159,
+      "step": 3945
+    },
+    {
+      "epoch": 0.08309861379466471,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.5987,
+      "step": 3946
+    },
+    {
+      "epoch": 0.08311967274392844,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.0002,
+      "loss": 1.5685,
+      "step": 3947
+    },
+    {
+      "epoch": 0.08314073169319217,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0002,
+      "loss": 1.5815,
+      "step": 3948
+    },
+    {
+      "epoch": 0.0831617906424559,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0002,
+      "loss": 1.5767,
+      "step": 3949
+    },
+    {
+      "epoch": 0.08318284959171962,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002,
+      "loss": 1.6038,
+      "step": 3950
+    },
+    {
+      "epoch": 0.08320390854098335,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.5815,
+      "step": 3951
+    },
+    {
+      "epoch": 0.08322496749024708,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0002,
+      "loss": 1.5727,
+      "step": 3952
+    },
+    {
+      "epoch": 0.0832460264395108,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.5769,
+      "step": 3953
+    },
+    {
+      "epoch": 0.08326708538877453,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.6272,
+      "step": 3954
+    },
+    {
+      "epoch": 0.08328814433803826,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.5731,
+      "step": 3955
+    },
+    {
+      "epoch": 0.08330920328730199,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.5644,
+      "step": 3956
+    },
+    {
+      "epoch": 0.0833302622365657,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.5596,
+      "step": 3957
+    },
+    {
+      "epoch": 0.08335132118582943,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0002,
+      "loss": 1.6111,
+      "step": 3958
+    },
+    {
+      "epoch": 0.08337238013509315,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002,
+      "loss": 1.6061,
+      "step": 3959
+    },
+    {
+      "epoch": 0.08339343908435688,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0002,
+      "loss": 1.581,
+      "step": 3960
+    },
+    {
+      "epoch": 0.08341449803362061,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0002,
+      "loss": 1.6083,
+      "step": 3961
+    },
+    {
+      "epoch": 0.08343555698288434,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.5434,
+      "step": 3962
+    },
+    {
+      "epoch": 0.08345661593214806,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.5932,
+      "step": 3963
+    },
+    {
+      "epoch": 0.08347767488141179,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002,
+      "loss": 1.5912,
+      "step": 3964
+    },
+    {
+      "epoch": 0.08349873383067552,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.5998,
+      "step": 3965
+    },
+    {
+      "epoch": 0.08351979277993925,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5314,
+      "step": 3966
+    },
+    {
+      "epoch": 0.08354085172920298,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0002,
+      "loss": 1.6056,
+      "step": 3967
+    },
+    {
+      "epoch": 0.0835619106784667,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.5697,
+      "step": 3968
+    },
+    {
+      "epoch": 0.08358296962773043,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.5678,
+      "step": 3969
+    },
+    {
+      "epoch": 0.08360402857699416,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.6399,
+      "step": 3970
+    },
+    {
+      "epoch": 0.08362508752625787,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.5924,
+      "step": 3971
+    },
+    {
+      "epoch": 0.0836461464755216,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.6072,
+      "step": 3972
+    },
+    {
+      "epoch": 0.08366720542478533,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.0002,
+      "loss": 1.6014,
+      "step": 3973
+    },
+    {
+      "epoch": 0.08368826437404905,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.5933,
+      "step": 3974
+    },
+    {
+      "epoch": 0.08370932332331278,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.5721,
+      "step": 3975
+    },
+    {
+      "epoch": 0.08373038227257651,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.5739,
+      "step": 3976
+    },
+    {
+      "epoch": 0.08375144122184024,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002,
+      "loss": 1.5979,
+      "step": 3977
+    },
+    {
+      "epoch": 0.08377250017110396,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.6014,
+      "step": 3978
+    },
+    {
+      "epoch": 0.08379355912036769,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0002,
+      "loss": 1.5868,
+      "step": 3979
+    },
+    {
+      "epoch": 0.08381461806963142,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002,
+      "loss": 1.5782,
+      "step": 3980
+    },
+    {
+      "epoch": 0.08383567701889515,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.5918,
+      "step": 3981
+    },
+    {
+      "epoch": 0.08385673596815887,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002,
+      "loss": 1.5816,
+      "step": 3982
+    },
+    {
+      "epoch": 0.0838777949174226,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002,
+      "loss": 1.5691,
+      "step": 3983
+    },
+    {
+      "epoch": 0.08389885386668632,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.5709,
+      "step": 3984
+    },
+    {
+      "epoch": 0.08391991281595004,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.6117,
+      "step": 3985
+    },
+    {
+      "epoch": 0.08394097176521377,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.5888,
+      "step": 3986
+    },
+    {
+      "epoch": 0.0839620307144775,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.5504,
+      "step": 3987
+    },
+    {
+      "epoch": 0.08398308966374123,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0002,
+      "loss": 1.5408,
+      "step": 3988
+    },
+    {
+      "epoch": 0.08400414861300495,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0002,
+      "loss": 1.5982,
+      "step": 3989
+    },
+    {
+      "epoch": 0.08402520756226868,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.5785,
+      "step": 3990
+    },
+    {
+      "epoch": 0.08404626651153241,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0002,
+      "loss": 1.5939,
+      "step": 3991
+    },
+    {
+      "epoch": 0.08406732546079614,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0002,
+      "loss": 1.5608,
+      "step": 3992
+    },
+    {
+      "epoch": 0.08408838441005986,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.0002,
+      "loss": 1.5855,
+      "step": 3993
+    },
+    {
+      "epoch": 0.08410944335932359,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5583,
+      "step": 3994
+    },
+    {
+      "epoch": 0.08413050230858732,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0002,
+      "loss": 1.5972,
+      "step": 3995
+    },
+    {
+      "epoch": 0.08415156125785105,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.0002,
+      "loss": 1.5906,
+      "step": 3996
+    },
+    {
+      "epoch": 0.08417262020711476,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002,
+      "loss": 1.5982,
+      "step": 3997
+    },
+    {
+      "epoch": 0.08419367915637849,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.5777,
+      "step": 3998
+    },
+    {
+      "epoch": 0.08421473810564221,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.5935,
+      "step": 3999
+    },
+    {
+      "epoch": 0.08423579705490594,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0002,
+      "loss": 1.56,
+      "step": 4000
+    },
+    {
+      "epoch": 0.08423579705490594,
+      "eval_loss": 2.1100332736968994,
+      "eval_runtime": 990.4764,
+      "eval_samples_per_second": 62.394,
+      "eval_steps_per_second": 1.951,
+      "step": 4000
+    },
+    {
+      "epoch": 0.08425685600416967,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.5891,
+      "step": 4001
+    },
+    {
+      "epoch": 0.0842779149534334,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0002,
+      "loss": 1.6125,
+      "step": 4002
+    },
+    {
+      "epoch": 0.08429897390269712,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.5656,
+      "step": 4003
+    },
+    {
+      "epoch": 0.08432003285196085,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.5996,
+      "step": 4004
+    },
+    {
+      "epoch": 0.08434109180122458,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0002,
+      "loss": 1.5828,
+      "step": 4005
+    },
+    {
+      "epoch": 0.08436215075048831,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0002,
+      "loss": 1.5553,
+      "step": 4006
+    },
+    {
+      "epoch": 0.08438320969975203,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.5919,
+      "step": 4007
+    },
+    {
+      "epoch": 0.08440426864901576,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0002,
+      "loss": 1.5849,
+      "step": 4008
+    },
+    {
+      "epoch": 0.08442532759827949,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0002,
+      "loss": 1.6342,
+      "step": 4009
+    },
+    {
+      "epoch": 0.0844463865475432,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002,
+      "loss": 1.5912,
+      "step": 4010
+    },
+    {
+      "epoch": 0.08446744549680693,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.589,
+      "step": 4011
+    },
+    {
+      "epoch": 0.08448850444607066,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0002,
+      "loss": 1.5954,
+      "step": 4012
+    },
+    {
+      "epoch": 0.08450956339533439,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0002,
+      "loss": 1.5923,
+      "step": 4013
+    },
+    {
+      "epoch": 0.08453062234459811,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.6099,
+      "step": 4014
+    },
+    {
+      "epoch": 0.08455168129386184,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.0002,
+      "loss": 1.6005,
+      "step": 4015
+    },
+    {
+      "epoch": 0.08457274024312557,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.5652,
+      "step": 4016
+    },
+    {
+      "epoch": 0.0845937991923893,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0002,
+      "loss": 1.6411,
+      "step": 4017
+    },
+    {
+      "epoch": 0.08461485814165302,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5379,
+      "step": 4018
+    },
+    {
+      "epoch": 0.08463591709091675,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.5945,
+      "step": 4019
+    },
+    {
+      "epoch": 0.08465697604018048,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6066,
+      "step": 4020
+    },
+    {
+      "epoch": 0.0846780349894442,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0002,
+      "loss": 1.5596,
+      "step": 4021
+    },
+    {
+      "epoch": 0.08469909393870793,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0002,
+      "loss": 1.6042,
+      "step": 4022
+    },
+    {
+      "epoch": 0.08472015288797166,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6043,
+      "step": 4023
+    },
+    {
+      "epoch": 0.08474121183723538,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.5601,
+      "step": 4024
+    },
+    {
+      "epoch": 0.0847622707864991,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.6272,
+      "step": 4025
+    },
+    {
+      "epoch": 0.08478332973576283,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.582,
+      "step": 4026
+    },
+    {
+      "epoch": 0.08480438868502656,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.5457,
+      "step": 4027
+    },
+    {
+      "epoch": 0.08482544763429029,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.5613,
+      "step": 4028
+    },
+    {
+      "epoch": 0.08484650658355401,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.6101,
+      "step": 4029
+    },
+    {
+      "epoch": 0.08486756553281774,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002,
+      "loss": 1.5534,
+      "step": 4030
+    },
+    {
+      "epoch": 0.08488862448208147,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.5849,
+      "step": 4031
+    },
+    {
+      "epoch": 0.0849096834313452,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.0002,
+      "loss": 1.5762,
+      "step": 4032
+    },
+    {
+      "epoch": 0.08493074238060892,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0002,
+      "loss": 1.5405,
+      "step": 4033
+    },
+    {
+      "epoch": 0.08495180132987265,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0002,
+      "loss": 1.588,
+      "step": 4034
+    },
+    {
+      "epoch": 0.08497286027913638,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0002,
+      "loss": 1.6134,
+      "step": 4035
+    },
+    {
+      "epoch": 0.0849939192284001,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.0002,
+      "loss": 1.5778,
+      "step": 4036
+    },
+    {
+      "epoch": 0.08501497817766382,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.5787,
+      "step": 4037
+    },
+    {
+      "epoch": 0.08503603712692755,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.6093,
+      "step": 4038
+    },
+    {
+      "epoch": 0.08505709607619127,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0002,
+      "loss": 1.6064,
+      "step": 4039
+    },
+    {
+      "epoch": 0.085078155025455,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.607,
+      "step": 4040
+    },
+    {
+      "epoch": 0.08509921397471873,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.5795,
+      "step": 4041
+    },
+    {
+      "epoch": 0.08512027292398246,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0002,
+      "loss": 1.6022,
+      "step": 4042
+    },
+    {
+      "epoch": 0.08514133187324618,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.5921,
+      "step": 4043
+    },
+    {
+      "epoch": 0.08516239082250991,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.6174,
+      "step": 4044
+    },
+    {
+      "epoch": 0.08518344977177364,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6026,
+      "step": 4045
+    },
+    {
+      "epoch": 0.08520450872103737,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002,
+      "loss": 1.5504,
+      "step": 4046
+    },
+    {
+      "epoch": 0.0852255676703011,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.6043,
+      "step": 4047
+    },
+    {
+      "epoch": 0.08524662661956482,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.5691,
+      "step": 4048
+    },
+    {
+      "epoch": 0.08526768556882855,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0002,
+      "loss": 1.5868,
+      "step": 4049
+    },
+    {
+      "epoch": 0.08528874451809226,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.554,
+      "step": 4050
+    },
+    {
+      "epoch": 0.08530980346735599,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.5782,
+      "step": 4051
+    },
+    {
+      "epoch": 0.08533086241661972,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002,
+      "loss": 1.6496,
+      "step": 4052
+    },
+    {
+      "epoch": 0.08535192136588345,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.5904,
+      "step": 4053
+    },
+    {
+      "epoch": 0.08537298031514717,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.5592,
+      "step": 4054
+    },
+    {
+      "epoch": 0.0853940392644109,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.5766,
+      "step": 4055
+    },
+    {
+      "epoch": 0.08541509821367463,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.5647,
+      "step": 4056
+    },
+    {
+      "epoch": 0.08543615716293836,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.5597,
+      "step": 4057
+    },
+    {
+      "epoch": 0.08545721611220208,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0002,
+      "loss": 1.5685,
+      "step": 4058
+    },
+    {
+      "epoch": 0.08547827506146581,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0002,
+      "loss": 1.6284,
+      "step": 4059
+    },
+    {
+      "epoch": 0.08549933401072954,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0002,
+      "loss": 1.5666,
+      "step": 4060
+    },
+    {
+      "epoch": 0.08552039295999327,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0002,
+      "loss": 1.5728,
+      "step": 4061
+    },
+    {
+      "epoch": 0.085541451909257,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0002,
+      "loss": 1.5866,
+      "step": 4062
+    },
+    {
+      "epoch": 0.08556251085852071,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0002,
+      "loss": 1.5571,
+      "step": 4063
+    },
+    {
+      "epoch": 0.08558356980778443,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0002,
+      "loss": 1.5601,
+      "step": 4064
+    },
+    {
+      "epoch": 0.08560462875704816,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002,
+      "loss": 1.6075,
+      "step": 4065
+    },
+    {
+      "epoch": 0.08562568770631189,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5675,
+      "step": 4066
+    },
+    {
+      "epoch": 0.08564674665557562,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.5699,
+      "step": 4067
+    },
+    {
+      "epoch": 0.08566780560483934,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0002,
+      "loss": 1.554,
+      "step": 4068
+    },
+    {
+      "epoch": 0.08568886455410307,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.5507,
+      "step": 4069
+    },
+    {
+      "epoch": 0.0857099235033668,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0002,
+      "loss": 1.5855,
+      "step": 4070
+    },
+    {
+      "epoch": 0.08573098245263053,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.5741,
+      "step": 4071
+    },
+    {
+      "epoch": 0.08575204140189426,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0002,
+      "loss": 1.5787,
+      "step": 4072
+    },
+    {
+      "epoch": 0.08577310035115798,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.5488,
+      "step": 4073
+    },
+    {
+      "epoch": 0.08579415930042171,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.5575,
+      "step": 4074
+    },
+    {
+      "epoch": 0.08581521824968544,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0002,
+      "loss": 1.5774,
+      "step": 4075
+    },
+    {
+      "epoch": 0.08583627719894917,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.5579,
+      "step": 4076
+    },
+    {
+      "epoch": 0.08585733614821288,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.5807,
+      "step": 4077
+    },
+    {
+      "epoch": 0.0858783950974766,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.5799,
+      "step": 4078
+    },
+    {
+      "epoch": 0.08589945404674033,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5634,
+      "step": 4079
+    },
+    {
+      "epoch": 0.08592051299600406,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0002,
+      "loss": 1.6134,
+      "step": 4080
+    },
+    {
+      "epoch": 0.08594157194526779,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.5569,
+      "step": 4081
+    },
+    {
+      "epoch": 0.08596263089453152,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5557,
+      "step": 4082
+    },
+    {
+      "epoch": 0.08598368984379524,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002,
+      "loss": 1.6015,
+      "step": 4083
+    },
+    {
+      "epoch": 0.08600474879305897,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.5807,
+      "step": 4084
+    },
+    {
+      "epoch": 0.0860258077423227,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.5863,
+      "step": 4085
+    },
+    {
+      "epoch": 0.08604686669158643,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0002,
+      "loss": 1.5584,
+      "step": 4086
+    },
+    {
+      "epoch": 0.08606792564085015,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.638,
+      "step": 4087
+    },
+    {
+      "epoch": 0.08608898459011388,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0002,
+      "loss": 1.5633,
+      "step": 4088
+    },
+    {
+      "epoch": 0.08611004353937761,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.5691,
+      "step": 4089
+    },
+    {
+      "epoch": 0.08613110248864132,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0002,
+      "loss": 1.5981,
+      "step": 4090
+    },
+    {
+      "epoch": 0.08615216143790505,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.0002,
+      "loss": 1.5972,
+      "step": 4091
+    },
+    {
+      "epoch": 0.08617322038716878,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5719,
+      "step": 4092
+    },
+    {
+      "epoch": 0.0861942793364325,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002,
+      "loss": 1.5868,
+      "step": 4093
+    },
+    {
+      "epoch": 0.08621533828569623,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.6196,
+      "step": 4094
+    },
+    {
+      "epoch": 0.08623639723495996,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0002,
+      "loss": 1.5609,
+      "step": 4095
+    },
+    {
+      "epoch": 0.08625745618422369,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.5594,
+      "step": 4096
+    },
+    {
+      "epoch": 0.08627851513348742,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0002,
+      "loss": 1.5484,
+      "step": 4097
+    },
+    {
+      "epoch": 0.08629957408275114,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.6014,
+      "step": 4098
+    },
+    {
+      "epoch": 0.08632063303201487,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5937,
+      "step": 4099
+    },
+    {
+      "epoch": 0.0863416919812786,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.5857,
+      "step": 4100
+    },
+    {
+      "epoch": 0.08636275093054233,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0002,
+      "loss": 1.5569,
+      "step": 4101
+    },
+    {
+      "epoch": 0.08638380987980605,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0002,
+      "loss": 1.5532,
+      "step": 4102
+    },
+    {
+      "epoch": 0.08640486882906977,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.5848,
+      "step": 4103
+    },
+    {
+      "epoch": 0.0864259277783335,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.6018,
+      "step": 4104
+    },
+    {
+      "epoch": 0.08644698672759722,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.5992,
+      "step": 4105
+    },
+    {
+      "epoch": 0.08646804567686095,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.5558,
+      "step": 4106
+    },
+    {
+      "epoch": 0.08648910462612468,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0002,
+      "loss": 1.6024,
+      "step": 4107
+    },
+    {
+      "epoch": 0.0865101635753884,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0002,
+      "loss": 1.5963,
+      "step": 4108
+    },
+    {
+      "epoch": 0.08653122252465213,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.6019,
+      "step": 4109
+    },
+    {
+      "epoch": 0.08655228147391586,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0002,
+      "loss": 1.5948,
+      "step": 4110
+    },
+    {
+      "epoch": 0.08657334042317959,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.5713,
+      "step": 4111
+    },
+    {
+      "epoch": 0.08659439937244331,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.5913,
+      "step": 4112
+    },
+    {
+      "epoch": 0.08661545832170704,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0002,
+      "loss": 1.5665,
+      "step": 4113
+    },
+    {
+      "epoch": 0.08663651727097077,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0002,
+      "loss": 1.5586,
+      "step": 4114
+    },
+    {
+      "epoch": 0.0866575762202345,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5852,
+      "step": 4115
+    },
+    {
+      "epoch": 0.08667863516949821,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.5899,
+      "step": 4116
+    },
+    {
+      "epoch": 0.08669969411876194,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6012,
+      "step": 4117
+    },
+    {
+      "epoch": 0.08672075306802567,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.5713,
+      "step": 4118
+    },
+    {
+      "epoch": 0.0867418120172894,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5905,
+      "step": 4119
+    },
+    {
+      "epoch": 0.08676287096655312,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.6097,
+      "step": 4120
+    },
+    {
+      "epoch": 0.08678392991581685,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.6431,
+      "step": 4121
+    },
+    {
+      "epoch": 0.08680498886508058,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.5615,
+      "step": 4122
+    },
+    {
+      "epoch": 0.0868260478143443,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6194,
+      "step": 4123
+    },
+    {
+      "epoch": 0.08684710676360803,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.587,
+      "step": 4124
+    },
+    {
+      "epoch": 0.08686816571287176,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0002,
+      "loss": 1.5808,
+      "step": 4125
+    },
+    {
+      "epoch": 0.08688922466213549,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.5967,
+      "step": 4126
+    },
+    {
+      "epoch": 0.08691028361139921,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.6032,
+      "step": 4127
+    },
+    {
+      "epoch": 0.08693134256066294,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0002,
+      "loss": 1.5887,
+      "step": 4128
+    },
+    {
+      "epoch": 0.08695240150992667,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.6248,
+      "step": 4129
+    },
+    {
+      "epoch": 0.08697346045919038,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.6105,
+      "step": 4130
+    },
+    {
+      "epoch": 0.08699451940845411,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0002,
+      "loss": 1.6043,
+      "step": 4131
+    },
+    {
+      "epoch": 0.08701557835771784,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0002,
+      "loss": 1.6094,
+      "step": 4132
+    },
+    {
+      "epoch": 0.08703663730698157,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.6092,
+      "step": 4133
+    },
+    {
+      "epoch": 0.08705769625624529,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6055,
+      "step": 4134
+    },
+    {
+      "epoch": 0.08707875520550902,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.5837,
+      "step": 4135
+    },
+    {
+      "epoch": 0.08709981415477275,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.6092,
+      "step": 4136
+    },
+    {
+      "epoch": 0.08712087310403648,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0002,
+      "loss": 1.5562,
+      "step": 4137
+    },
+    {
+      "epoch": 0.0871419320533002,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.5651,
+      "step": 4138
+    },
+    {
+      "epoch": 0.08716299100256393,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.5981,
+      "step": 4139
+    },
+    {
+      "epoch": 0.08718404995182766,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.5675,
+      "step": 4140
+    },
+    {
+      "epoch": 0.08720510890109139,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0002,
+      "loss": 1.5645,
+      "step": 4141
+    },
+    {
+      "epoch": 0.08722616785035511,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.5881,
+      "step": 4142
+    },
+    {
+      "epoch": 0.08724722679961883,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.5876,
+      "step": 4143
+    },
+    {
+      "epoch": 0.08726828574888255,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.591,
+      "step": 4144
+    },
+    {
+      "epoch": 0.08728934469814628,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0002,
+      "loss": 1.5777,
+      "step": 4145
+    },
+    {
+      "epoch": 0.08731040364741001,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0002,
+      "loss": 1.5809,
+      "step": 4146
+    },
+    {
+      "epoch": 0.08733146259667374,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.5889,
+      "step": 4147
+    },
+    {
+      "epoch": 0.08735252154593746,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.5937,
+      "step": 4148
+    },
+    {
+      "epoch": 0.08737358049520119,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.5889,
+      "step": 4149
+    },
+    {
+      "epoch": 0.08739463944446492,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.5858,
+      "step": 4150
+    },
+    {
+      "epoch": 0.08741569839372865,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.5533,
+      "step": 4151
+    },
+    {
+      "epoch": 0.08743675734299237,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0002,
+      "loss": 1.5734,
+      "step": 4152
+    },
+    {
+      "epoch": 0.0874578162922561,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0002,
+      "loss": 1.5629,
+      "step": 4153
+    },
+    {
+      "epoch": 0.08747887524151983,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5914,
+      "step": 4154
+    },
+    {
+      "epoch": 0.08749993419078356,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.594,
+      "step": 4155
+    },
+    {
+      "epoch": 0.08752099314004727,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.5676,
+      "step": 4156
+    },
+    {
+      "epoch": 0.087542052089311,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0002,
+      "loss": 1.5723,
+      "step": 4157
+    },
+    {
+      "epoch": 0.08756311103857473,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0002,
+      "loss": 1.5533,
+      "step": 4158
+    },
+    {
+      "epoch": 0.08758416998783845,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.5767,
+      "step": 4159
+    },
+    {
+      "epoch": 0.08760522893710218,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5894,
+      "step": 4160
+    },
+    {
+      "epoch": 0.08762628788636591,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.0002,
+      "loss": 1.6121,
+      "step": 4161
+    },
+    {
+      "epoch": 0.08764734683562964,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.5989,
+      "step": 4162
+    },
+    {
+      "epoch": 0.08766840578489336,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.6408,
+      "step": 4163
+    },
+    {
+      "epoch": 0.08768946473415709,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0002,
+      "loss": 1.6001,
+      "step": 4164
+    },
+    {
+      "epoch": 0.08771052368342082,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.5846,
+      "step": 4165
+    },
+    {
+      "epoch": 0.08773158263268455,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002,
+      "loss": 1.5863,
+      "step": 4166
+    },
+    {
+      "epoch": 0.08775264158194827,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.6249,
+      "step": 4167
+    },
+    {
+      "epoch": 0.087773700531212,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.5793,
+      "step": 4168
+    },
+    {
+      "epoch": 0.08779475948047571,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0002,
+      "loss": 1.656,
+      "step": 4169
+    },
+    {
+      "epoch": 0.08781581842973944,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0002,
+      "loss": 1.6073,
+      "step": 4170
+    },
+    {
+      "epoch": 0.08783687737900317,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5695,
+      "step": 4171
+    },
+    {
+      "epoch": 0.0878579363282669,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5621,
+      "step": 4172
+    },
+    {
+      "epoch": 0.08787899527753062,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.5981,
+      "step": 4173
+    },
+    {
+      "epoch": 0.08790005422679435,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6283,
+      "step": 4174
+    },
+    {
+      "epoch": 0.08792111317605808,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.0002,
+      "loss": 1.5937,
+      "step": 4175
+    },
+    {
+      "epoch": 0.08794217212532181,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.5893,
+      "step": 4176
+    },
+    {
+      "epoch": 0.08796323107458554,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.5719,
+      "step": 4177
+    },
+    {
+      "epoch": 0.08798429002384926,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5876,
+      "step": 4178
+    },
+    {
+      "epoch": 0.08800534897311299,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0002,
+      "loss": 1.5818,
+      "step": 4179
+    },
+    {
+      "epoch": 0.08802640792237672,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.6221,
+      "step": 4180
+    },
+    {
+      "epoch": 0.08804746687164045,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0002,
+      "loss": 1.5782,
+      "step": 4181
+    },
+    {
+      "epoch": 0.08806852582090417,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5731,
+      "step": 4182
+    },
+    {
+      "epoch": 0.08808958477016789,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0002,
+      "loss": 1.5683,
+      "step": 4183
+    },
+    {
+      "epoch": 0.08811064371943161,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.5691,
+      "step": 4184
+    },
+    {
+      "epoch": 0.08813170266869534,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0002,
+      "loss": 1.5968,
+      "step": 4185
+    },
+    {
+      "epoch": 0.08815276161795907,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002,
+      "loss": 1.5764,
+      "step": 4186
+    },
+    {
+      "epoch": 0.0881738205672228,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0002,
+      "loss": 1.6485,
+      "step": 4187
+    },
+    {
+      "epoch": 0.08819487951648652,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0002,
+      "loss": 1.5936,
+      "step": 4188
+    },
+    {
+      "epoch": 0.08821593846575025,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.5935,
+      "step": 4189
+    },
+    {
+      "epoch": 0.08823699741501398,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5761,
+      "step": 4190
+    },
+    {
+      "epoch": 0.0882580563642777,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.5715,
+      "step": 4191
+    },
+    {
+      "epoch": 0.08827911531354143,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.5753,
+      "step": 4192
+    },
+    {
+      "epoch": 0.08830017426280516,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.5899,
+      "step": 4193
+    },
+    {
+      "epoch": 0.08832123321206889,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.0002,
+      "loss": 1.592,
+      "step": 4194
+    },
+    {
+      "epoch": 0.08834229216133262,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0002,
+      "loss": 1.5935,
+      "step": 4195
+    },
+    {
+      "epoch": 0.08836335111059633,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002,
+      "loss": 1.6238,
+      "step": 4196
+    },
+    {
+      "epoch": 0.08838441005986006,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.5699,
+      "step": 4197
+    },
+    {
+      "epoch": 0.08840546900912379,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.5642,
+      "step": 4198
+    },
+    {
+      "epoch": 0.08842652795838751,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.5702,
+      "step": 4199
+    },
+    {
+      "epoch": 0.08844758690765124,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.5811,
+      "step": 4200
+    },
+    {
+      "epoch": 0.08846864585691497,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0002,
+      "loss": 1.5978,
+      "step": 4201
+    },
+    {
+      "epoch": 0.0884897048061787,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.5812,
+      "step": 4202
+    },
+    {
+      "epoch": 0.08851076375544242,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.0002,
+      "loss": 1.6292,
+      "step": 4203
+    },
+    {
+      "epoch": 0.08853182270470615,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.5625,
+      "step": 4204
+    },
+    {
+      "epoch": 0.08855288165396988,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0002,
+      "loss": 1.5882,
+      "step": 4205
+    },
+    {
+      "epoch": 0.0885739406032336,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.5826,
+      "step": 4206
+    },
+    {
+      "epoch": 0.08859499955249733,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0002,
+      "loss": 1.6048,
+      "step": 4207
+    },
+    {
+      "epoch": 0.08861605850176106,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0002,
+      "loss": 1.6066,
+      "step": 4208
+    },
+    {
+      "epoch": 0.08863711745102477,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.6181,
+      "step": 4209
+    },
+    {
+      "epoch": 0.0886581764002885,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0002,
+      "loss": 1.576,
+      "step": 4210
+    },
+    {
+      "epoch": 0.08867923534955223,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.5765,
+      "step": 4211
+    },
+    {
+      "epoch": 0.08870029429881596,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.5811,
+      "step": 4212
+    },
+    {
+      "epoch": 0.08872135324807968,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5938,
+      "step": 4213
+    },
+    {
+      "epoch": 0.08874241219734341,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.6075,
+      "step": 4214
+    },
+    {
+      "epoch": 0.08876347114660714,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5544,
+      "step": 4215
+    },
+    {
+      "epoch": 0.08878453009587087,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.6046,
+      "step": 4216
+    },
+    {
+      "epoch": 0.0888055890451346,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0002,
+      "loss": 1.5895,
+      "step": 4217
+    },
+    {
+      "epoch": 0.08882664799439832,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.5939,
+      "step": 4218
+    },
+    {
+      "epoch": 0.08884770694366205,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.0002,
+      "loss": 1.6365,
+      "step": 4219
+    },
+    {
+      "epoch": 0.08886876589292578,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0002,
+      "loss": 1.5596,
+      "step": 4220
+    },
+    {
+      "epoch": 0.0888898248421895,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.5874,
+      "step": 4221
+    },
+    {
+      "epoch": 0.08891088379145322,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0002,
+      "loss": 1.5398,
+      "step": 4222
+    },
+    {
+      "epoch": 0.08893194274071695,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.5826,
+      "step": 4223
+    },
+    {
+      "epoch": 0.08895300168998067,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.5677,
+      "step": 4224
+    },
+    {
+      "epoch": 0.0889740606392444,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.5407,
+      "step": 4225
+    },
+    {
+      "epoch": 0.08899511958850813,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0002,
+      "loss": 1.5561,
+      "step": 4226
+    },
+    {
+      "epoch": 0.08901617853777186,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5918,
+      "step": 4227
+    },
+    {
+      "epoch": 0.08903723748703558,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0002,
+      "loss": 1.5475,
+      "step": 4228
+    },
+    {
+      "epoch": 0.08905829643629931,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0002,
+      "loss": 1.5749,
+      "step": 4229
+    },
+    {
+      "epoch": 0.08907935538556304,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.5699,
+      "step": 4230
+    },
+    {
+      "epoch": 0.08910041433482677,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.5815,
+      "step": 4231
+    },
+    {
+      "epoch": 0.0891214732840905,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6148,
+      "step": 4232
+    },
+    {
+      "epoch": 0.08914253223335422,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.6187,
+      "step": 4233
+    },
+    {
+      "epoch": 0.08916359118261795,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.6143,
+      "step": 4234
+    },
+    {
+      "epoch": 0.08918465013188168,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.5916,
+      "step": 4235
+    },
+    {
+      "epoch": 0.08920570908114539,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0002,
+      "loss": 1.5795,
+      "step": 4236
+    },
+    {
+      "epoch": 0.08922676803040912,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0002,
+      "loss": 1.6065,
+      "step": 4237
+    },
+    {
+      "epoch": 0.08924782697967285,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.5985,
+      "step": 4238
+    },
+    {
+      "epoch": 0.08926888592893657,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.5928,
+      "step": 4239
+    },
+    {
+      "epoch": 0.0892899448782003,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.6073,
+      "step": 4240
+    },
+    {
+      "epoch": 0.08931100382746403,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.61,
+      "step": 4241
+    },
+    {
+      "epoch": 0.08933206277672776,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.6209,
+      "step": 4242
+    },
+    {
+      "epoch": 0.08935312172599148,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0002,
+      "loss": 1.5842,
+      "step": 4243
+    },
+    {
+      "epoch": 0.08937418067525521,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0002,
+      "loss": 1.5733,
+      "step": 4244
+    },
+    {
+      "epoch": 0.08939523962451894,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002,
+      "loss": 1.5937,
+      "step": 4245
+    },
+    {
+      "epoch": 0.08941629857378267,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.578,
+      "step": 4246
+    },
+    {
+      "epoch": 0.08943735752304639,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0002,
+      "loss": 1.5456,
+      "step": 4247
+    },
+    {
+      "epoch": 0.08945841647231012,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.5927,
+      "step": 4248
+    },
+    {
+      "epoch": 0.08947947542157383,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0002,
+      "loss": 1.6015,
+      "step": 4249
+    },
+    {
+      "epoch": 0.08950053437083756,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.0002,
+      "loss": 1.6103,
+      "step": 4250
+    },
+    {
+      "epoch": 0.08952159332010129,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0002,
+      "loss": 1.5965,
+      "step": 4251
+    },
+    {
+      "epoch": 0.08954265226936502,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.5916,
+      "step": 4252
+    },
+    {
+      "epoch": 0.08956371121862874,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0002,
+      "loss": 1.6145,
+      "step": 4253
+    },
+    {
+      "epoch": 0.08958477016789247,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5745,
+      "step": 4254
+    },
+    {
+      "epoch": 0.0896058291171562,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0002,
+      "loss": 1.5524,
+      "step": 4255
+    },
+    {
+      "epoch": 0.08962688806641993,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002,
+      "loss": 1.5961,
+      "step": 4256
+    },
+    {
+      "epoch": 0.08964794701568365,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5805,
+      "step": 4257
+    },
+    {
+      "epoch": 0.08966900596494738,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0002,
+      "loss": 1.571,
+      "step": 4258
+    },
+    {
+      "epoch": 0.08969006491421111,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.5863,
+      "step": 4259
+    },
+    {
+      "epoch": 0.08971112386347484,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6026,
+      "step": 4260
+    },
+    {
+      "epoch": 0.08973218281273856,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5858,
+      "step": 4261
+    },
+    {
+      "epoch": 0.08975324176200228,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.5867,
+      "step": 4262
+    },
+    {
+      "epoch": 0.089774300711266,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.5746,
+      "step": 4263
+    },
+    {
+      "epoch": 0.08979535966052973,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.0002,
+      "loss": 1.5785,
+      "step": 4264
+    },
+    {
+      "epoch": 0.08981641860979346,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0002,
+      "loss": 1.5737,
+      "step": 4265
+    },
+    {
+      "epoch": 0.08983747755905719,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.5686,
+      "step": 4266
+    },
+    {
+      "epoch": 0.08985853650832092,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0002,
+      "loss": 1.5777,
+      "step": 4267
+    },
+    {
+      "epoch": 0.08987959545758464,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.6159,
+      "step": 4268
+    },
+    {
+      "epoch": 0.08990065440684837,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6163,
+      "step": 4269
+    },
+    {
+      "epoch": 0.0899217133561121,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.5991,
+      "step": 4270
+    },
+    {
+      "epoch": 0.08994277230537583,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0002,
+      "loss": 1.5892,
+      "step": 4271
+    },
+    {
+      "epoch": 0.08996383125463955,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0002,
+      "loss": 1.5886,
+      "step": 4272
+    },
+    {
+      "epoch": 0.08998489020390328,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5704,
+      "step": 4273
+    },
+    {
+      "epoch": 0.09000594915316701,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.5332,
+      "step": 4274
+    },
+    {
+      "epoch": 0.09002700810243074,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.582,
+      "step": 4275
+    },
+    {
+      "epoch": 0.09004806705169445,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0002,
+      "loss": 1.5712,
+      "step": 4276
+    },
+    {
+      "epoch": 0.09006912600095818,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.5871,
+      "step": 4277
+    },
+    {
+      "epoch": 0.0900901849502219,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.5751,
+      "step": 4278
+    },
+    {
+      "epoch": 0.09011124389948563,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0002,
+      "loss": 1.5682,
+      "step": 4279
+    },
+    {
+      "epoch": 0.09013230284874936,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.6082,
+      "step": 4280
+    },
+    {
+      "epoch": 0.09015336179801309,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.5954,
+      "step": 4281
+    },
+    {
+      "epoch": 0.09017442074727682,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0002,
+      "loss": 1.5668,
+      "step": 4282
+    },
+    {
+      "epoch": 0.09019547969654054,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5889,
+      "step": 4283
+    },
+    {
+      "epoch": 0.09021653864580427,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0002,
+      "loss": 1.5681,
+      "step": 4284
+    },
+    {
+      "epoch": 0.090237597595068,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0002,
+      "loss": 1.5682,
+      "step": 4285
+    },
+    {
+      "epoch": 0.09025865654433173,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0002,
+      "loss": 1.5689,
+      "step": 4286
+    },
+    {
+      "epoch": 0.09027971549359545,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5682,
+      "step": 4287
+    },
+    {
+      "epoch": 0.09030077444285918,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0002,
+      "loss": 1.5518,
+      "step": 4288
+    },
+    {
+      "epoch": 0.0903218333921229,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.5745,
+      "step": 4289
+    },
+    {
+      "epoch": 0.09034289234138662,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0002,
+      "loss": 1.5594,
+      "step": 4290
+    },
+    {
+      "epoch": 0.09036395129065035,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0002,
+      "loss": 1.5884,
+      "step": 4291
+    },
+    {
+      "epoch": 0.09038501023991408,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.6053,
+      "step": 4292
+    },
+    {
+      "epoch": 0.0904060691891778,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.5704,
+      "step": 4293
+    },
+    {
+      "epoch": 0.09042712813844153,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.5701,
+      "step": 4294
+    },
+    {
+      "epoch": 0.09044818708770526,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.5498,
+      "step": 4295
+    },
+    {
+      "epoch": 0.09046924603696899,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5899,
+      "step": 4296
+    },
+    {
+      "epoch": 0.09049030498623271,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0002,
+      "loss": 1.5929,
+      "step": 4297
+    },
+    {
+      "epoch": 0.09051136393549644,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.5682,
+      "step": 4298
+    },
+    {
+      "epoch": 0.09053242288476017,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.6108,
+      "step": 4299
+    },
+    {
+      "epoch": 0.0905534818340239,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.5793,
+      "step": 4300
+    },
+    {
+      "epoch": 0.09057454078328762,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5885,
+      "step": 4301
+    },
+    {
+      "epoch": 0.09059559973255134,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.5738,
+      "step": 4302
+    },
+    {
+      "epoch": 0.09061665868181507,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0002,
+      "loss": 1.6013,
+      "step": 4303
+    },
+    {
+      "epoch": 0.09063771763107879,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0002,
+      "loss": 1.5859,
+      "step": 4304
+    },
+    {
+      "epoch": 0.09065877658034252,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0002,
+      "loss": 1.5937,
+      "step": 4305
+    },
+    {
+      "epoch": 0.09067983552960625,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5848,
+      "step": 4306
+    },
+    {
+      "epoch": 0.09070089447886998,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0002,
+      "loss": 1.5971,
+      "step": 4307
+    },
+    {
+      "epoch": 0.0907219534281337,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.5663,
+      "step": 4308
+    },
+    {
+      "epoch": 0.09074301237739743,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.6073,
+      "step": 4309
+    },
+    {
+      "epoch": 0.09076407132666116,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.5471,
+      "step": 4310
+    },
+    {
+      "epoch": 0.09078513027592489,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002,
+      "loss": 1.5635,
+      "step": 4311
+    },
+    {
+      "epoch": 0.09080618922518861,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5767,
+      "step": 4312
+    },
+    {
+      "epoch": 0.09082724817445234,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002,
+      "loss": 1.5722,
+      "step": 4313
+    },
+    {
+      "epoch": 0.09084830712371607,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.5631,
+      "step": 4314
+    },
+    {
+      "epoch": 0.09086936607297978,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.579,
+      "step": 4315
+    },
+    {
+      "epoch": 0.09089042502224351,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0002,
+      "loss": 1.5923,
+      "step": 4316
+    },
+    {
+      "epoch": 0.09091148397150724,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.5608,
+      "step": 4317
+    },
+    {
+      "epoch": 0.09093254292077096,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.5991,
+      "step": 4318
+    },
+    {
+      "epoch": 0.09095360187003469,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002,
+      "loss": 1.5869,
+      "step": 4319
+    },
+    {
+      "epoch": 0.09097466081929842,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0002,
+      "loss": 1.579,
+      "step": 4320
+    },
+    {
+      "epoch": 0.09099571976856215,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0002,
+      "loss": 1.5718,
+      "step": 4321
+    },
+    {
+      "epoch": 0.09101677871782587,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6072,
+      "step": 4322
+    },
+    {
+      "epoch": 0.0910378376670896,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.5565,
+      "step": 4323
+    },
+    {
+      "epoch": 0.09105889661635333,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5873,
+      "step": 4324
+    },
+    {
+      "epoch": 0.09107995556561706,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0002,
+      "loss": 1.579,
+      "step": 4325
+    },
+    {
+      "epoch": 0.09110101451488078,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002,
+      "loss": 1.5642,
+      "step": 4326
+    },
+    {
+      "epoch": 0.09112207346414451,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0002,
+      "loss": 1.5861,
+      "step": 4327
+    },
+    {
+      "epoch": 0.09114313241340824,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0002,
+      "loss": 1.6046,
+      "step": 4328
+    },
+    {
+      "epoch": 0.09116419136267195,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.5692,
+      "step": 4329
+    },
+    {
+      "epoch": 0.09118525031193568,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.5746,
+      "step": 4330
+    },
+    {
+      "epoch": 0.09120630926119941,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0002,
+      "loss": 1.6105,
+      "step": 4331
+    },
+    {
+      "epoch": 0.09122736821046314,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0002,
+      "loss": 1.5596,
+      "step": 4332
+    },
+    {
+      "epoch": 0.09124842715972686,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.5994,
+      "step": 4333
+    },
+    {
+      "epoch": 0.09126948610899059,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.5531,
+      "step": 4334
+    },
+    {
+      "epoch": 0.09129054505825432,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.5756,
+      "step": 4335
+    },
+    {
+      "epoch": 0.09131160400751805,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.5302,
+      "step": 4336
+    },
+    {
+      "epoch": 0.09133266295678177,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5969,
+      "step": 4337
+    },
+    {
+      "epoch": 0.0913537219060455,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.561,
+      "step": 4338
+    },
+    {
+      "epoch": 0.09137478085530923,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002,
+      "loss": 1.6308,
+      "step": 4339
+    },
+    {
+      "epoch": 0.09139583980457296,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.5895,
+      "step": 4340
+    },
+    {
+      "epoch": 0.09141689875383668,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.5925,
+      "step": 4341
+    },
+    {
+      "epoch": 0.0914379577031004,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002,
+      "loss": 1.604,
+      "step": 4342
+    },
+    {
+      "epoch": 0.09145901665236413,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.5561,
+      "step": 4343
+    },
+    {
+      "epoch": 0.09148007560162785,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0002,
+      "loss": 1.5743,
+      "step": 4344
+    },
+    {
+      "epoch": 0.09150113455089158,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.5661,
+      "step": 4345
+    },
+    {
+      "epoch": 0.09152219350015531,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0002,
+      "loss": 1.5908,
+      "step": 4346
+    },
+    {
+      "epoch": 0.09154325244941904,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.6173,
+      "step": 4347
+    },
+    {
+      "epoch": 0.09156431139868276,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.6097,
+      "step": 4348
+    },
+    {
+      "epoch": 0.09158537034794649,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0002,
+      "loss": 1.5715,
+      "step": 4349
+    },
+    {
+      "epoch": 0.09160642929721022,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.607,
+      "step": 4350
+    },
+    {
+      "epoch": 0.09162748824647395,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.5826,
+      "step": 4351
+    },
+    {
+      "epoch": 0.09164854719573767,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0002,
+      "loss": 1.5455,
+      "step": 4352
+    },
+    {
+      "epoch": 0.0916696061450014,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0002,
+      "loss": 1.6097,
+      "step": 4353
+    },
+    {
+      "epoch": 0.09169066509426513,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0002,
+      "loss": 1.5569,
+      "step": 4354
+    },
+    {
+      "epoch": 0.09171172404352884,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0002,
+      "loss": 1.552,
+      "step": 4355
+    },
+    {
+      "epoch": 0.09173278299279257,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.0002,
+      "loss": 1.621,
+      "step": 4356
+    },
+    {
+      "epoch": 0.0917538419420563,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.5796,
+      "step": 4357
+    },
+    {
+      "epoch": 0.09177490089132002,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0002,
+      "loss": 1.576,
+      "step": 4358
+    },
+    {
+      "epoch": 0.09179595984058375,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.5621,
+      "step": 4359
+    },
+    {
+      "epoch": 0.09181701878984748,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5843,
+      "step": 4360
+    },
+    {
+      "epoch": 0.0918380777391112,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.5881,
+      "step": 4361
+    },
+    {
+      "epoch": 0.09185913668837493,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0002,
+      "loss": 1.5735,
+      "step": 4362
+    },
+    {
+      "epoch": 0.09188019563763866,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0002,
+      "loss": 1.6037,
+      "step": 4363
+    },
+    {
+      "epoch": 0.09190125458690239,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5835,
+      "step": 4364
+    },
+    {
+      "epoch": 0.09192231353616612,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0002,
+      "loss": 1.644,
+      "step": 4365
+    },
+    {
+      "epoch": 0.09194337248542984,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5896,
+      "step": 4366
+    },
+    {
+      "epoch": 0.09196443143469357,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.6055,
+      "step": 4367
+    },
+    {
+      "epoch": 0.09198549038395729,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.5924,
+      "step": 4368
+    },
+    {
+      "epoch": 0.09200654933322101,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.5692,
+      "step": 4369
+    },
+    {
+      "epoch": 0.09202760828248474,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0002,
+      "loss": 1.5646,
+      "step": 4370
+    },
+    {
+      "epoch": 0.09204866723174847,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0002,
+      "loss": 1.5821,
+      "step": 4371
+    },
+    {
+      "epoch": 0.0920697261810122,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.5779,
+      "step": 4372
+    },
+    {
+      "epoch": 0.09209078513027592,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.5941,
+      "step": 4373
+    },
+    {
+      "epoch": 0.09211184407953965,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.5565,
+      "step": 4374
+    },
+    {
+      "epoch": 0.09213290302880338,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0002,
+      "loss": 1.6021,
+      "step": 4375
+    },
+    {
+      "epoch": 0.0921539619780671,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0002,
+      "loss": 1.5647,
+      "step": 4376
+    },
+    {
+      "epoch": 0.09217502092733083,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.5942,
+      "step": 4377
+    },
+    {
+      "epoch": 0.09219607987659456,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0002,
+      "loss": 1.572,
+      "step": 4378
+    },
+    {
+      "epoch": 0.09221713882585829,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0002,
+      "loss": 1.5381,
+      "step": 4379
+    },
+    {
+      "epoch": 0.09223819777512202,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0002,
+      "loss": 1.5875,
+      "step": 4380
+    },
+    {
+      "epoch": 0.09225925672438574,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.6024,
+      "step": 4381
+    },
+    {
+      "epoch": 0.09228031567364946,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.0002,
+      "loss": 1.5973,
+      "step": 4382
+    },
+    {
+      "epoch": 0.09230137462291318,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0002,
+      "loss": 1.5902,
+      "step": 4383
+    },
+    {
+      "epoch": 0.09232243357217691,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0002,
+      "loss": 1.6084,
+      "step": 4384
+    },
+    {
+      "epoch": 0.09234349252144064,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0002,
+      "loss": 1.5993,
+      "step": 4385
+    },
+    {
+      "epoch": 0.09236455147070437,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.624,
+      "step": 4386
+    },
+    {
+      "epoch": 0.0923856104199681,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0002,
+      "loss": 1.5944,
+      "step": 4387
+    },
+    {
+      "epoch": 0.09240666936923182,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.5446,
+      "step": 4388
+    },
+    {
+      "epoch": 0.09242772831849555,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0002,
+      "loss": 1.5563,
+      "step": 4389
+    },
+    {
+      "epoch": 0.09244878726775928,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.5849,
+      "step": 4390
+    },
+    {
+      "epoch": 0.092469846217023,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.0002,
+      "loss": 1.5658,
+      "step": 4391
+    },
+    {
+      "epoch": 0.09249090516628673,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0002,
+      "loss": 1.592,
+      "step": 4392
+    },
+    {
+      "epoch": 0.09251196411555046,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0002,
+      "loss": 1.5988,
+      "step": 4393
+    },
+    {
+      "epoch": 0.09253302306481419,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5676,
+      "step": 4394
+    },
+    {
+      "epoch": 0.0925540820140779,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0002,
+      "loss": 1.6044,
+      "step": 4395
+    },
+    {
+      "epoch": 0.09257514096334163,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.6029,
+      "step": 4396
+    },
+    {
+      "epoch": 0.09259619991260536,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5774,
+      "step": 4397
+    },
+    {
+      "epoch": 0.09261725886186908,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.5541,
+      "step": 4398
+    },
+    {
+      "epoch": 0.09263831781113281,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0002,
+      "loss": 1.5892,
+      "step": 4399
+    },
+    {
+      "epoch": 0.09265937676039654,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.5816,
+      "step": 4400
+    },
+    {
+      "epoch": 0.09268043570966027,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0002,
+      "loss": 1.571,
+      "step": 4401
+    },
+    {
+      "epoch": 0.092701494658924,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6011,
+      "step": 4402
+    },
+    {
+      "epoch": 0.09272255360818772,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.607,
+      "step": 4403
+    },
+    {
+      "epoch": 0.09274361255745145,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0002,
+      "loss": 1.5817,
+      "step": 4404
+    },
+    {
+      "epoch": 0.09276467150671518,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0002,
+      "loss": 1.6085,
+      "step": 4405
+    },
+    {
+      "epoch": 0.0927857304559789,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0002,
+      "loss": 1.6069,
+      "step": 4406
+    },
+    {
+      "epoch": 0.09280678940524263,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.5601,
+      "step": 4407
+    },
+    {
+      "epoch": 0.09282784835450635,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.6021,
+      "step": 4408
+    },
+    {
+      "epoch": 0.09284890730377007,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.557,
+      "step": 4409
+    },
+    {
+      "epoch": 0.0928699662530338,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.595,
+      "step": 4410
+    },
+    {
+      "epoch": 0.09289102520229753,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.5878,
+      "step": 4411
+    },
+    {
+      "epoch": 0.09291208415156126,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.577,
+      "step": 4412
+    },
+    {
+      "epoch": 0.09293314310082498,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6027,
+      "step": 4413
+    },
+    {
+      "epoch": 0.09295420205008871,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.5565,
+      "step": 4414
+    },
+    {
+      "epoch": 0.09297526099935244,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.5787,
+      "step": 4415
+    },
+    {
+      "epoch": 0.09299631994861617,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.5773,
+      "step": 4416
+    },
+    {
+      "epoch": 0.0930173788978799,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.6056,
+      "step": 4417
+    },
+    {
+      "epoch": 0.09303843784714362,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0002,
+      "loss": 1.5709,
+      "step": 4418
+    },
+    {
+      "epoch": 0.09305949679640735,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0002,
+      "loss": 1.5695,
+      "step": 4419
+    },
+    {
+      "epoch": 0.09308055574567108,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0002,
+      "loss": 1.5839,
+      "step": 4420
+    },
+    {
+      "epoch": 0.09310161469493479,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0002,
+      "loss": 1.5561,
+      "step": 4421
+    },
+    {
+      "epoch": 0.09312267364419852,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5706,
+      "step": 4422
+    },
+    {
+      "epoch": 0.09314373259346224,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002,
+      "loss": 1.5913,
+      "step": 4423
+    },
+    {
+      "epoch": 0.09316479154272597,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.5888,
+      "step": 4424
+    },
+    {
+      "epoch": 0.0931858504919897,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.5916,
+      "step": 4425
+    },
+    {
+      "epoch": 0.09320690944125343,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0002,
+      "loss": 1.5884,
+      "step": 4426
+    },
+    {
+      "epoch": 0.09322796839051715,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0002,
+      "loss": 1.6063,
+      "step": 4427
+    },
+    {
+      "epoch": 0.09324902733978088,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.5809,
+      "step": 4428
+    },
+    {
+      "epoch": 0.09327008628904461,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.0002,
+      "loss": 1.5798,
+      "step": 4429
+    },
+    {
+      "epoch": 0.09329114523830834,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.6048,
+      "step": 4430
+    },
+    {
+      "epoch": 0.09331220418757206,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0002,
+      "loss": 1.6299,
+      "step": 4431
+    },
+    {
+      "epoch": 0.09333326313683579,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0002,
+      "loss": 1.5517,
+      "step": 4432
+    },
+    {
+      "epoch": 0.09335432208609952,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.6036,
+      "step": 4433
+    },
+    {
+      "epoch": 0.09337538103536325,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0002,
+      "loss": 1.5896,
+      "step": 4434
+    },
+    {
+      "epoch": 0.09339643998462696,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002,
+      "loss": 1.5919,
+      "step": 4435
+    },
+    {
+      "epoch": 0.09341749893389069,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.6042,
+      "step": 4436
+    },
+    {
+      "epoch": 0.09343855788315442,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5872,
+      "step": 4437
+    },
+    {
+      "epoch": 0.09345961683241814,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.6116,
+      "step": 4438
+    },
+    {
+      "epoch": 0.09348067578168187,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0002,
+      "loss": 1.5594,
+      "step": 4439
+    },
+    {
+      "epoch": 0.0935017347309456,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.5887,
+      "step": 4440
+    },
+    {
+      "epoch": 0.09352279368020933,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.5727,
+      "step": 4441
+    },
+    {
+      "epoch": 0.09354385262947305,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0002,
+      "loss": 1.5568,
+      "step": 4442
+    },
+    {
+      "epoch": 0.09356491157873678,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.6175,
+      "step": 4443
+    },
+    {
+      "epoch": 0.09358597052800051,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0002,
+      "loss": 1.5872,
+      "step": 4444
+    },
+    {
+      "epoch": 0.09360702947726424,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002,
+      "loss": 1.5557,
+      "step": 4445
+    },
+    {
+      "epoch": 0.09362808842652796,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0002,
+      "loss": 1.596,
+      "step": 4446
+    },
+    {
+      "epoch": 0.09364914737579169,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5563,
+      "step": 4447
+    },
+    {
+      "epoch": 0.0936702063250554,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0002,
+      "loss": 1.5726,
+      "step": 4448
+    },
+    {
+      "epoch": 0.09369126527431913,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0002,
+      "loss": 1.6028,
+      "step": 4449
+    },
+    {
+      "epoch": 0.09371232422358286,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.6025,
+      "step": 4450
+    },
+    {
+      "epoch": 0.09373338317284659,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6078,
+      "step": 4451
+    },
+    {
+      "epoch": 0.09375444212211032,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.5658,
+      "step": 4452
+    },
+    {
+      "epoch": 0.09377550107137404,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.5285,
+      "step": 4453
+    },
+    {
+      "epoch": 0.09379656002063777,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.6155,
+      "step": 4454
+    },
+    {
+      "epoch": 0.0938176189699015,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.5906,
+      "step": 4455
+    },
+    {
+      "epoch": 0.09383867791916523,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.598,
+      "step": 4456
+    },
+    {
+      "epoch": 0.09385973686842895,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.5892,
+      "step": 4457
+    },
+    {
+      "epoch": 0.09388079581769268,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0002,
+      "loss": 1.5531,
+      "step": 4458
+    },
+    {
+      "epoch": 0.09390185476695641,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.5717,
+      "step": 4459
+    },
+    {
+      "epoch": 0.09392291371622014,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5693,
+      "step": 4460
+    },
+    {
+      "epoch": 0.09394397266548385,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.5939,
+      "step": 4461
+    },
+    {
+      "epoch": 0.09396503161474758,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.6006,
+      "step": 4462
+    },
+    {
+      "epoch": 0.0939860905640113,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.5753,
+      "step": 4463
+    },
+    {
+      "epoch": 0.09400714951327503,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0002,
+      "loss": 1.6176,
+      "step": 4464
+    },
+    {
+      "epoch": 0.09402820846253876,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.5771,
+      "step": 4465
+    },
+    {
+      "epoch": 0.09404926741180249,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.5715,
+      "step": 4466
+    },
+    {
+      "epoch": 0.09407032636106621,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0002,
+      "loss": 1.5499,
+      "step": 4467
+    },
+    {
+      "epoch": 0.09409138531032994,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0002,
+      "loss": 1.6075,
+      "step": 4468
+    },
+    {
+      "epoch": 0.09411244425959367,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.5719,
+      "step": 4469
+    },
+    {
+      "epoch": 0.0941335032088574,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0002,
+      "loss": 1.5692,
+      "step": 4470
+    },
+    {
+      "epoch": 0.09415456215812112,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.5776,
+      "step": 4471
+    },
+    {
+      "epoch": 0.09417562110738485,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5693,
+      "step": 4472
+    },
+    {
+      "epoch": 0.09419668005664858,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.5748,
+      "step": 4473
+    },
+    {
+      "epoch": 0.0942177390059123,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.5788,
+      "step": 4474
+    },
+    {
+      "epoch": 0.09423879795517602,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.6038,
+      "step": 4475
+    },
+    {
+      "epoch": 0.09425985690443975,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0002,
+      "loss": 1.5678,
+      "step": 4476
+    },
+    {
+      "epoch": 0.09428091585370348,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.5706,
+      "step": 4477
+    },
+    {
+      "epoch": 0.0943019748029672,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002,
+      "loss": 1.593,
+      "step": 4478
+    },
+    {
+      "epoch": 0.09432303375223093,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.6094,
+      "step": 4479
+    },
+    {
+      "epoch": 0.09434409270149466,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002,
+      "loss": 1.5652,
+      "step": 4480
+    },
+    {
+      "epoch": 0.09436515165075839,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.5767,
+      "step": 4481
+    },
+    {
+      "epoch": 0.09438621060002211,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0002,
+      "loss": 1.6065,
+      "step": 4482
+    },
+    {
+      "epoch": 0.09440726954928584,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0002,
+      "loss": 1.6009,
+      "step": 4483
+    },
+    {
+      "epoch": 0.09442832849854957,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5912,
+      "step": 4484
+    },
+    {
+      "epoch": 0.0944493874478133,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0002,
+      "loss": 1.5867,
+      "step": 4485
+    },
+    {
+      "epoch": 0.09447044639707702,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002,
+      "loss": 1.6074,
+      "step": 4486
+    },
+    {
+      "epoch": 0.09449150534634075,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002,
+      "loss": 1.5674,
+      "step": 4487
+    },
+    {
+      "epoch": 0.09451256429560446,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.5793,
+      "step": 4488
+    },
+    {
+      "epoch": 0.09453362324486819,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002,
+      "loss": 1.5848,
+      "step": 4489
+    },
+    {
+      "epoch": 0.09455468219413192,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6006,
+      "step": 4490
+    },
+    {
+      "epoch": 0.09457574114339565,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002,
+      "loss": 1.597,
+      "step": 4491
+    },
+    {
+      "epoch": 0.09459680009265937,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.0002,
+      "loss": 1.5679,
+      "step": 4492
+    },
+    {
+      "epoch": 0.0946178590419231,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.5814,
+      "step": 4493
+    },
+    {
+      "epoch": 0.09463891799118683,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0002,
+      "loss": 1.5845,
+      "step": 4494
+    },
+    {
+      "epoch": 0.09465997694045056,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.5778,
+      "step": 4495
+    },
+    {
+      "epoch": 0.09468103588971429,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.5665,
+      "step": 4496
+    },
+    {
+      "epoch": 0.09470209483897801,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.0002,
+      "loss": 1.5804,
+      "step": 4497
+    },
+    {
+      "epoch": 0.09472315378824174,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002,
+      "loss": 1.596,
+      "step": 4498
+    },
+    {
+      "epoch": 0.09474421273750547,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.6045,
+      "step": 4499
+    },
+    {
+      "epoch": 0.0947652716867692,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.6209,
+      "step": 4500
+    },
+    {
+      "epoch": 0.0947652716867692,
+      "eval_loss": 2.20310378074646,
+      "eval_runtime": 897.7013,
+      "eval_samples_per_second": 68.842,
+      "eval_steps_per_second": 2.152,
+      "step": 4500
+    },
+    {
+      "epoch": 0.09478633063603291,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0002,
+      "loss": 1.5777,
+      "step": 4501
+    },
+    {
+      "epoch": 0.09480738958529664,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.5885,
+      "step": 4502
+    },
+    {
+      "epoch": 0.09482844853456036,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0002,
+      "loss": 1.5977,
+      "step": 4503
+    },
+    {
+      "epoch": 0.09484950748382409,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0002,
+      "loss": 1.6173,
+      "step": 4504
+    },
+    {
+      "epoch": 0.09487056643308782,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.568,
+      "step": 4505
+    },
+    {
+      "epoch": 0.09489162538235155,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5889,
+      "step": 4506
+    },
+    {
+      "epoch": 0.09491268433161527,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5593,
+      "step": 4507
+    },
+    {
+      "epoch": 0.094933743280879,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0002,
+      "loss": 1.6217,
+      "step": 4508
+    },
+    {
+      "epoch": 0.09495480223014273,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.5943,
+      "step": 4509
+    },
+    {
+      "epoch": 0.09497586117940646,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0002,
+      "loss": 1.5811,
+      "step": 4510
+    },
+    {
+      "epoch": 0.09499692012867018,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.5914,
+      "step": 4511
+    },
+    {
+      "epoch": 0.09501797907793391,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0002,
+      "loss": 1.5892,
+      "step": 4512
+    },
+    {
+      "epoch": 0.09503903802719764,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0002,
+      "loss": 1.5772,
+      "step": 4513
+    },
+    {
+      "epoch": 0.09506009697646135,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.5777,
+      "step": 4514
+    },
+    {
+      "epoch": 0.09508115592572508,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.5935,
+      "step": 4515
+    },
+    {
+      "epoch": 0.09510221487498881,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0002,
+      "loss": 1.6134,
+      "step": 4516
+    },
+    {
+      "epoch": 0.09512327382425254,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6021,
+      "step": 4517
+    },
+    {
+      "epoch": 0.09514433277351626,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.5804,
+      "step": 4518
+    },
+    {
+      "epoch": 0.09516539172277999,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.5752,
+      "step": 4519
+    },
+    {
+      "epoch": 0.09518645067204372,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.0002,
+      "loss": 1.5396,
+      "step": 4520
+    },
+    {
+      "epoch": 0.09520750962130745,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5962,
+      "step": 4521
+    },
+    {
+      "epoch": 0.09522856857057117,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5922,
+      "step": 4522
+    },
+    {
+      "epoch": 0.0952496275198349,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.5322,
+      "step": 4523
+    },
+    {
+      "epoch": 0.09527068646909863,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.5903,
+      "step": 4524
+    },
+    {
+      "epoch": 0.09529174541836236,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.6082,
+      "step": 4525
+    },
+    {
+      "epoch": 0.09531280436762608,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.5842,
+      "step": 4526
+    },
+    {
+      "epoch": 0.0953338633168898,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.5684,
+      "step": 4527
+    },
+    {
+      "epoch": 0.09535492226615352,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0002,
+      "loss": 1.6035,
+      "step": 4528
+    },
+    {
+      "epoch": 0.09537598121541725,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0002,
+      "loss": 1.5392,
+      "step": 4529
+    },
+    {
+      "epoch": 0.09539704016468098,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0002,
+      "loss": 1.6136,
+      "step": 4530
+    },
+    {
+      "epoch": 0.09541809911394471,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.548,
+      "step": 4531
+    },
+    {
+      "epoch": 0.09543915806320843,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5683,
+      "step": 4532
+    },
+    {
+      "epoch": 0.09546021701247216,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.5754,
+      "step": 4533
+    },
+    {
+      "epoch": 0.09548127596173589,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6265,
+      "step": 4534
+    },
+    {
+      "epoch": 0.09550233491099962,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.6023,
+      "step": 4535
+    },
+    {
+      "epoch": 0.09552339386026334,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002,
+      "loss": 1.5859,
+      "step": 4536
+    },
+    {
+      "epoch": 0.09554445280952707,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.564,
+      "step": 4537
+    },
+    {
+      "epoch": 0.0955655117587908,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0002,
+      "loss": 1.5721,
+      "step": 4538
+    },
+    {
+      "epoch": 0.09558657070805453,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0002,
+      "loss": 1.5735,
+      "step": 4539
+    },
+    {
+      "epoch": 0.09560762965731825,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.5704,
+      "step": 4540
+    },
+    {
+      "epoch": 0.09562868860658197,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0002,
+      "loss": 1.5604,
+      "step": 4541
+    },
+    {
+      "epoch": 0.0956497475558457,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0002,
+      "loss": 1.6378,
+      "step": 4542
+    },
+    {
+      "epoch": 0.09567080650510942,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0002,
+      "loss": 1.5357,
+      "step": 4543
+    },
+    {
+      "epoch": 0.09569186545437315,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.5763,
+      "step": 4544
+    },
+    {
+      "epoch": 0.09571292440363688,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.5945,
+      "step": 4545
+    },
+    {
+      "epoch": 0.0957339833529006,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0002,
+      "loss": 1.5659,
+      "step": 4546
+    },
+    {
+      "epoch": 0.09575504230216433,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.6143,
+      "step": 4547
+    },
+    {
+      "epoch": 0.09577610125142806,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.5555,
+      "step": 4548
+    },
+    {
+      "epoch": 0.09579716020069179,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5852,
+      "step": 4549
+    },
+    {
+      "epoch": 0.09581821914995552,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0002,
+      "loss": 1.6237,
+      "step": 4550
+    },
+    {
+      "epoch": 0.09583927809921924,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.58,
+      "step": 4551
+    },
+    {
+      "epoch": 0.09586033704848297,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0002,
+      "loss": 1.5872,
+      "step": 4552
+    },
+    {
+      "epoch": 0.0958813959977467,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0002,
+      "loss": 1.5679,
+      "step": 4553
+    },
+    {
+      "epoch": 0.09590245494701041,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.5715,
+      "step": 4554
+    },
+    {
+      "epoch": 0.09592351389627414,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0002,
+      "loss": 1.5677,
+      "step": 4555
+    },
+    {
+      "epoch": 0.09594457284553787,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.5921,
+      "step": 4556
+    },
+    {
+      "epoch": 0.0959656317948016,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.5783,
+      "step": 4557
+    },
+    {
+      "epoch": 0.09598669074406532,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5865,
+      "step": 4558
+    },
+    {
+      "epoch": 0.09600774969332905,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0002,
+      "loss": 1.5759,
+      "step": 4559
+    },
+    {
+      "epoch": 0.09602880864259278,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5873,
+      "step": 4560
+    },
+    {
+      "epoch": 0.0960498675918565,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.0002,
+      "loss": 1.576,
+      "step": 4561
+    },
+    {
+      "epoch": 0.09607092654112023,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.582,
+      "step": 4562
+    },
+    {
+      "epoch": 0.09609198549038396,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0002,
+      "loss": 1.5849,
+      "step": 4563
+    },
+    {
+      "epoch": 0.09611304443964769,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.0002,
+      "loss": 1.5598,
+      "step": 4564
+    },
+    {
+      "epoch": 0.09613410338891142,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.5787,
+      "step": 4565
+    },
+    {
+      "epoch": 0.09615516233817514,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.577,
+      "step": 4566
+    },
+    {
+      "epoch": 0.09617622128743886,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.5941,
+      "step": 4567
+    },
+    {
+      "epoch": 0.09619728023670258,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0002,
+      "loss": 1.5258,
+      "step": 4568
+    },
+    {
+      "epoch": 0.09621833918596631,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.5818,
+      "step": 4569
+    },
+    {
+      "epoch": 0.09623939813523004,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5556,
+      "step": 4570
+    },
+    {
+      "epoch": 0.09626045708449377,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.5896,
+      "step": 4571
+    },
+    {
+      "epoch": 0.0962815160337575,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.5773,
+      "step": 4572
+    },
+    {
+      "epoch": 0.09630257498302122,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.5954,
+      "step": 4573
+    },
+    {
+      "epoch": 0.09632363393228495,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.5748,
+      "step": 4574
+    },
+    {
+      "epoch": 0.09634469288154868,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0002,
+      "loss": 1.6214,
+      "step": 4575
+    },
+    {
+      "epoch": 0.0963657518308124,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002,
+      "loss": 1.5914,
+      "step": 4576
+    },
+    {
+      "epoch": 0.09638681078007613,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.5951,
+      "step": 4577
+    },
+    {
+      "epoch": 0.09640786972933986,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.5502,
+      "step": 4578
+    },
+    {
+      "epoch": 0.09642892867860359,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5459,
+      "step": 4579
+    },
+    {
+      "epoch": 0.0964499876278673,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0002,
+      "loss": 1.5479,
+      "step": 4580
+    },
+    {
+      "epoch": 0.09647104657713103,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.5754,
+      "step": 4581
+    },
+    {
+      "epoch": 0.09649210552639476,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.6126,
+      "step": 4582
+    },
+    {
+      "epoch": 0.09651316447565848,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.5932,
+      "step": 4583
+    },
+    {
+      "epoch": 0.09653422342492221,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.5848,
+      "step": 4584
+    },
+    {
+      "epoch": 0.09655528237418594,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0002,
+      "loss": 1.6002,
+      "step": 4585
+    },
+    {
+      "epoch": 0.09657634132344967,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.5953,
+      "step": 4586
+    },
+    {
+      "epoch": 0.0965974002727134,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.6005,
+      "step": 4587
+    },
+    {
+      "epoch": 0.09661845922197712,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5955,
+      "step": 4588
+    },
+    {
+      "epoch": 0.09663951817124085,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0002,
+      "loss": 1.6052,
+      "step": 4589
+    },
+    {
+      "epoch": 0.09666057712050458,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0002,
+      "loss": 1.6168,
+      "step": 4590
+    },
+    {
+      "epoch": 0.0966816360697683,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5765,
+      "step": 4591
+    },
+    {
+      "epoch": 0.09670269501903203,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.5675,
+      "step": 4592
+    },
+    {
+      "epoch": 0.09672375396829576,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0002,
+      "loss": 1.5601,
+      "step": 4593
+    },
+    {
+      "epoch": 0.09674481291755947,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.6295,
+      "step": 4594
+    },
+    {
+      "epoch": 0.0967658718668232,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.5569,
+      "step": 4595
+    },
+    {
+      "epoch": 0.09678693081608693,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.5529,
+      "step": 4596
+    },
+    {
+      "epoch": 0.09680798976535065,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.5818,
+      "step": 4597
+    },
+    {
+      "epoch": 0.09682904871461438,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0002,
+      "loss": 1.594,
+      "step": 4598
+    },
+    {
+      "epoch": 0.09685010766387811,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0002,
+      "loss": 1.6204,
+      "step": 4599
+    },
+    {
+      "epoch": 0.09687116661314184,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.5707,
+      "step": 4600
+    },
+    {
+      "epoch": 0.09689222556240557,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0002,
+      "loss": 1.5689,
+      "step": 4601
+    },
+    {
+      "epoch": 0.09691328451166929,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6196,
+      "step": 4602
+    },
+    {
+      "epoch": 0.09693434346093302,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0002,
+      "loss": 1.5776,
+      "step": 4603
+    },
+    {
+      "epoch": 0.09695540241019675,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0002,
+      "loss": 1.5882,
+      "step": 4604
+    },
+    {
+      "epoch": 0.09697646135946048,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.0002,
+      "loss": 1.5801,
+      "step": 4605
+    },
+    {
+      "epoch": 0.0969975203087242,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.6349,
+      "step": 4606
+    },
+    {
+      "epoch": 0.09701857925798792,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0002,
+      "loss": 1.616,
+      "step": 4607
+    },
+    {
+      "epoch": 0.09703963820725164,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0002,
+      "loss": 1.5849,
+      "step": 4608
+    },
+    {
+      "epoch": 0.09706069715651537,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.6066,
+      "step": 4609
+    },
+    {
+      "epoch": 0.0970817561057791,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5787,
+      "step": 4610
+    },
+    {
+      "epoch": 0.09710281505504283,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.5864,
+      "step": 4611
+    },
+    {
+      "epoch": 0.09712387400430655,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0002,
+      "loss": 1.5608,
+      "step": 4612
+    },
+    {
+      "epoch": 0.09714493295357028,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0002,
+      "loss": 1.5856,
+      "step": 4613
+    },
+    {
+      "epoch": 0.09716599190283401,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.5886,
+      "step": 4614
+    },
+    {
+      "epoch": 0.09718705085209774,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.6034,
+      "step": 4615
+    },
+    {
+      "epoch": 0.09720810980136146,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5875,
+      "step": 4616
+    },
+    {
+      "epoch": 0.09722916875062519,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0002,
+      "loss": 1.582,
+      "step": 4617
+    },
+    {
+      "epoch": 0.09725022769988892,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5759,
+      "step": 4618
+    },
+    {
+      "epoch": 0.09727128664915265,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5954,
+      "step": 4619
+    },
+    {
+      "epoch": 0.09729234559841636,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0002,
+      "loss": 1.6123,
+      "step": 4620
+    },
+    {
+      "epoch": 0.09731340454768009,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0002,
+      "loss": 1.5896,
+      "step": 4621
+    },
+    {
+      "epoch": 0.09733446349694382,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.5558,
+      "step": 4622
+    },
+    {
+      "epoch": 0.09735552244620754,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002,
+      "loss": 1.5939,
+      "step": 4623
+    },
+    {
+      "epoch": 0.09737658139547127,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0002,
+      "loss": 1.5532,
+      "step": 4624
+    },
+    {
+      "epoch": 0.097397640344735,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.573,
+      "step": 4625
+    },
+    {
+      "epoch": 0.09741869929399873,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.0002,
+      "loss": 1.6013,
+      "step": 4626
+    },
+    {
+      "epoch": 0.09743975824326245,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0002,
+      "loss": 1.5944,
+      "step": 4627
+    },
+    {
+      "epoch": 0.09746081719252618,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0002,
+      "loss": 1.6027,
+      "step": 4628
+    },
+    {
+      "epoch": 0.09748187614178991,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.5818,
+      "step": 4629
+    },
+    {
+      "epoch": 0.09750293509105364,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5649,
+      "step": 4630
+    },
+    {
+      "epoch": 0.09752399404031736,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.5585,
+      "step": 4631
+    },
+    {
+      "epoch": 0.09754505298958109,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.6072,
+      "step": 4632
+    },
+    {
+      "epoch": 0.0975661119388448,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5901,
+      "step": 4633
+    },
+    {
+      "epoch": 0.09758717088810853,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.6071,
+      "step": 4634
+    },
+    {
+      "epoch": 0.09760822983737226,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0002,
+      "loss": 1.5717,
+      "step": 4635
+    },
+    {
+      "epoch": 0.09762928878663599,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0002,
+      "loss": 1.5918,
+      "step": 4636
+    },
+    {
+      "epoch": 0.09765034773589971,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0002,
+      "loss": 1.609,
+      "step": 4637
+    },
+    {
+      "epoch": 0.09767140668516344,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5811,
+      "step": 4638
+    },
+    {
+      "epoch": 0.09769246563442717,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.5918,
+      "step": 4639
+    },
+    {
+      "epoch": 0.0977135245836909,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0002,
+      "loss": 1.5991,
+      "step": 4640
+    },
+    {
+      "epoch": 0.09773458353295462,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.5984,
+      "step": 4641
+    },
+    {
+      "epoch": 0.09775564248221835,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.6066,
+      "step": 4642
+    },
+    {
+      "epoch": 0.09777670143148208,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0002,
+      "loss": 1.5762,
+      "step": 4643
+    },
+    {
+      "epoch": 0.09779776038074581,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.5615,
+      "step": 4644
+    },
+    {
+      "epoch": 0.09781881933000953,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0002,
+      "loss": 1.6052,
+      "step": 4645
+    },
+    {
+      "epoch": 0.09783987827927326,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.5764,
+      "step": 4646
+    },
+    {
+      "epoch": 0.09786093722853698,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.5461,
+      "step": 4647
+    },
+    {
+      "epoch": 0.0978819961778007,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.5702,
+      "step": 4648
+    },
+    {
+      "epoch": 0.09790305512706443,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.5613,
+      "step": 4649
+    },
+    {
+      "epoch": 0.09792411407632816,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.5715,
+      "step": 4650
+    },
+    {
+      "epoch": 0.09794517302559189,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.5733,
+      "step": 4651
+    },
+    {
+      "epoch": 0.09796623197485561,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.5928,
+      "step": 4652
+    },
+    {
+      "epoch": 0.09798729092411934,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0002,
+      "loss": 1.5814,
+      "step": 4653
+    },
+    {
+      "epoch": 0.09800834987338307,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6052,
+      "step": 4654
+    },
+    {
+      "epoch": 0.0980294088226468,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.5894,
+      "step": 4655
+    },
+    {
+      "epoch": 0.09805046777191052,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0002,
+      "loss": 1.5726,
+      "step": 4656
+    },
+    {
+      "epoch": 0.09807152672117425,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0002,
+      "loss": 1.5926,
+      "step": 4657
+    },
+    {
+      "epoch": 0.09809258567043798,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0002,
+      "loss": 1.5716,
+      "step": 4658
+    },
+    {
+      "epoch": 0.0981136446197017,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0002,
+      "loss": 1.5963,
+      "step": 4659
+    },
+    {
+      "epoch": 0.09813470356896542,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.585,
+      "step": 4660
+    },
+    {
+      "epoch": 0.09815576251822915,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0002,
+      "loss": 1.5869,
+      "step": 4661
+    },
+    {
+      "epoch": 0.09817682146749288,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.5761,
+      "step": 4662
+    },
+    {
+      "epoch": 0.0981978804167566,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0002,
+      "loss": 1.5731,
+      "step": 4663
+    },
+    {
+      "epoch": 0.09821893936602033,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.0002,
+      "loss": 1.57,
+      "step": 4664
+    },
+    {
+      "epoch": 0.09823999831528406,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0002,
+      "loss": 1.6285,
+      "step": 4665
+    },
+    {
+      "epoch": 0.09826105726454779,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0002,
+      "loss": 1.5885,
+      "step": 4666
+    },
+    {
+      "epoch": 0.09828211621381151,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.571,
+      "step": 4667
+    },
+    {
+      "epoch": 0.09830317516307524,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.6143,
+      "step": 4668
+    },
+    {
+      "epoch": 0.09832423411233897,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.5926,
+      "step": 4669
+    },
+    {
+      "epoch": 0.0983452930616027,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5924,
+      "step": 4670
+    },
+    {
+      "epoch": 0.09836635201086642,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5833,
+      "step": 4671
+    },
+    {
+      "epoch": 0.09838741096013015,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0002,
+      "loss": 1.5759,
+      "step": 4672
+    },
+    {
+      "epoch": 0.09840846990939386,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.6126,
+      "step": 4673
+    },
+    {
+      "epoch": 0.09842952885865759,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.5952,
+      "step": 4674
+    },
+    {
+      "epoch": 0.09845058780792132,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.6463,
+      "step": 4675
+    },
+    {
+      "epoch": 0.09847164675718505,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0002,
+      "loss": 1.5535,
+      "step": 4676
+    },
+    {
+      "epoch": 0.09849270570644877,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5926,
+      "step": 4677
+    },
+    {
+      "epoch": 0.0985137646557125,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.6102,
+      "step": 4678
+    },
+    {
+      "epoch": 0.09853482360497623,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.5682,
+      "step": 4679
+    },
+    {
+      "epoch": 0.09855588255423996,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.5942,
+      "step": 4680
+    },
+    {
+      "epoch": 0.09857694150350368,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.5953,
+      "step": 4681
+    },
+    {
+      "epoch": 0.09859800045276741,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5776,
+      "step": 4682
+    },
+    {
+      "epoch": 0.09861905940203114,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.5425,
+      "step": 4683
+    },
+    {
+      "epoch": 0.09864011835129487,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.0002,
+      "loss": 1.57,
+      "step": 4684
+    },
+    {
+      "epoch": 0.0986611773005586,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0002,
+      "loss": 1.5671,
+      "step": 4685
+    },
+    {
+      "epoch": 0.09868223624982231,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0002,
+      "loss": 1.5968,
+      "step": 4686
+    },
+    {
+      "epoch": 0.09870329519908604,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.0002,
+      "loss": 1.5621,
+      "step": 4687
+    },
+    {
+      "epoch": 0.09872435414834976,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0002,
+      "loss": 1.5489,
+      "step": 4688
+    },
+    {
+      "epoch": 0.09874541309761349,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0002,
+      "loss": 1.6083,
+      "step": 4689
+    },
+    {
+      "epoch": 0.09876647204687722,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0002,
+      "loss": 1.6168,
+      "step": 4690
+    },
+    {
+      "epoch": 0.09878753099614095,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0002,
+      "loss": 1.5597,
+      "step": 4691
+    },
+    {
+      "epoch": 0.09880858994540467,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0002,
+      "loss": 1.5965,
+      "step": 4692
+    },
+    {
+      "epoch": 0.0988296488946684,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0002,
+      "loss": 1.5732,
+      "step": 4693
+    },
+    {
+      "epoch": 0.09885070784393213,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0002,
+      "loss": 1.6013,
+      "step": 4694
+    },
+    {
+      "epoch": 0.09887176679319586,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.5932,
+      "step": 4695
+    },
+    {
+      "epoch": 0.09889282574245958,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.5682,
+      "step": 4696
+    },
+    {
+      "epoch": 0.09891388469172331,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0002,
+      "loss": 1.5736,
+      "step": 4697
+    },
+    {
+      "epoch": 0.09893494364098704,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0002,
+      "loss": 1.5798,
+      "step": 4698
+    },
+    {
+      "epoch": 0.09895600259025077,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.5542,
+      "step": 4699
+    },
+    {
+      "epoch": 0.09897706153951448,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.5595,
+      "step": 4700
+    },
+    {
+      "epoch": 0.09899812048877821,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.5806,
+      "step": 4701
+    },
+    {
+      "epoch": 0.09901917943804193,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0002,
+      "loss": 1.5723,
+      "step": 4702
+    },
+    {
+      "epoch": 0.09904023838730566,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.5936,
+      "step": 4703
+    },
+    {
+      "epoch": 0.09906129733656939,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5798,
+      "step": 4704
+    },
+    {
+      "epoch": 0.09908235628583312,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.5809,
+      "step": 4705
+    },
+    {
+      "epoch": 0.09910341523509685,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0002,
+      "loss": 1.6059,
+      "step": 4706
+    },
+    {
+      "epoch": 0.09912447418436057,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0002,
+      "loss": 1.5966,
+      "step": 4707
+    },
+    {
+      "epoch": 0.0991455331336243,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.6326,
+      "step": 4708
+    },
+    {
+      "epoch": 0.09916659208288803,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.6041,
+      "step": 4709
+    },
+    {
+      "epoch": 0.09918765103215176,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0002,
+      "loss": 1.6137,
+      "step": 4710
+    },
+    {
+      "epoch": 0.09920870998141548,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.6007,
+      "step": 4711
+    },
+    {
+      "epoch": 0.09922976893067921,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.5889,
+      "step": 4712
+    },
+    {
+      "epoch": 0.09925082787994292,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5812,
+      "step": 4713
+    },
+    {
+      "epoch": 0.09927188682920665,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.5933,
+      "step": 4714
+    },
+    {
+      "epoch": 0.09929294577847038,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0002,
+      "loss": 1.571,
+      "step": 4715
+    },
+    {
+      "epoch": 0.0993140047277341,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.6076,
+      "step": 4716
+    },
+    {
+      "epoch": 0.09933506367699783,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6282,
+      "step": 4717
+    },
+    {
+      "epoch": 0.09935612262626156,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0002,
+      "loss": 1.5573,
+      "step": 4718
+    },
+    {
+      "epoch": 0.09937718157552529,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0002,
+      "loss": 1.543,
+      "step": 4719
+    },
+    {
+      "epoch": 0.09939824052478902,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6088,
+      "step": 4720
+    },
+    {
+      "epoch": 0.09941929947405274,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5742,
+      "step": 4721
+    },
+    {
+      "epoch": 0.09944035842331647,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0002,
+      "loss": 1.5938,
+      "step": 4722
+    },
+    {
+      "epoch": 0.0994614173725802,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.5726,
+      "step": 4723
+    },
+    {
+      "epoch": 0.09948247632184393,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0002,
+      "loss": 1.6096,
+      "step": 4724
+    },
+    {
+      "epoch": 0.09950353527110765,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002,
+      "loss": 1.5797,
+      "step": 4725
+    },
+    {
+      "epoch": 0.09952459422037137,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5553,
+      "step": 4726
+    },
+    {
+      "epoch": 0.0995456531696351,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0002,
+      "loss": 1.575,
+      "step": 4727
+    },
+    {
+      "epoch": 0.09956671211889882,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.5962,
+      "step": 4728
+    },
+    {
+      "epoch": 0.09958777106816255,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.5888,
+      "step": 4729
+    },
+    {
+      "epoch": 0.09960883001742628,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0002,
+      "loss": 1.5727,
+      "step": 4730
+    },
+    {
+      "epoch": 0.09962988896669,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0002,
+      "loss": 1.5721,
+      "step": 4731
+    },
+    {
+      "epoch": 0.09965094791595373,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5733,
+      "step": 4732
+    },
+    {
+      "epoch": 0.09967200686521746,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0002,
+      "loss": 1.5696,
+      "step": 4733
+    },
+    {
+      "epoch": 0.09969306581448119,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.5922,
+      "step": 4734
+    },
+    {
+      "epoch": 0.09971412476374492,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0002,
+      "loss": 1.5753,
+      "step": 4735
+    },
+    {
+      "epoch": 0.09973518371300864,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5872,
+      "step": 4736
+    },
+    {
+      "epoch": 0.09975624266227237,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002,
+      "loss": 1.6019,
+      "step": 4737
+    },
+    {
+      "epoch": 0.0997773016115361,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.582,
+      "step": 4738
+    },
+    {
+      "epoch": 0.09979836056079981,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.5436,
+      "step": 4739
+    },
+    {
+      "epoch": 0.09981941951006354,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0002,
+      "loss": 1.6001,
+      "step": 4740
+    },
+    {
+      "epoch": 0.09984047845932727,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0002,
+      "loss": 1.5918,
+      "step": 4741
+    },
+    {
+      "epoch": 0.099861537408591,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.6088,
+      "step": 4742
+    },
+    {
+      "epoch": 0.09988259635785472,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0002,
+      "loss": 1.5957,
+      "step": 4743
+    },
+    {
+      "epoch": 0.09990365530711845,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0002,
+      "loss": 1.5778,
+      "step": 4744
+    },
+    {
+      "epoch": 0.09992471425638218,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0002,
+      "loss": 1.5532,
+      "step": 4745
+    },
+    {
+      "epoch": 0.0999457732056459,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.5815,
+      "step": 4746
+    },
+    {
+      "epoch": 0.09996683215490963,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5765,
+      "step": 4747
+    },
+    {
+      "epoch": 0.09998789110417336,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.5895,
+      "step": 4748
+    },
+    {
+      "epoch": 0.10000895005343709,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0002,
+      "loss": 1.6022,
+      "step": 4749
+    },
+    {
+      "epoch": 0.10003000900270081,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0002,
+      "loss": 1.5659,
+      "step": 4750
+    },
+    {
+      "epoch": 0.10005106795196454,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5666,
+      "step": 4751
+    },
+    {
+      "epoch": 0.10007212690122827,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.5604,
+      "step": 4752
+    },
+    {
+      "epoch": 0.10009318585049198,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0002,
+      "loss": 1.5659,
+      "step": 4753
+    },
+    {
+      "epoch": 0.10011424479975571,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.564,
+      "step": 4754
+    },
+    {
+      "epoch": 0.10013530374901944,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0002,
+      "loss": 1.6102,
+      "step": 4755
+    },
+    {
+      "epoch": 0.10015636269828317,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.591,
+      "step": 4756
+    },
+    {
+      "epoch": 0.1001774216475469,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0002,
+      "loss": 1.5757,
+      "step": 4757
+    },
+    {
+      "epoch": 0.10019848059681062,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.0002,
+      "loss": 1.5806,
+      "step": 4758
+    },
+    {
+      "epoch": 0.10021953954607435,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.5774,
+      "step": 4759
+    },
+    {
+      "epoch": 0.10024059849533808,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0002,
+      "loss": 1.5845,
+      "step": 4760
+    },
+    {
+      "epoch": 0.1002616574446018,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5893,
+      "step": 4761
+    },
+    {
+      "epoch": 0.10028271639386553,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5514,
+      "step": 4762
+    },
+    {
+      "epoch": 0.10030377534312926,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0002,
+      "loss": 1.5684,
+      "step": 4763
+    },
+    {
+      "epoch": 0.10032483429239299,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.5916,
+      "step": 4764
+    },
+    {
+      "epoch": 0.10034589324165671,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0002,
+      "loss": 1.5913,
+      "step": 4765
+    },
+    {
+      "epoch": 0.10036695219092043,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5664,
+      "step": 4766
+    },
+    {
+      "epoch": 0.10038801114018416,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.5855,
+      "step": 4767
+    },
+    {
+      "epoch": 0.10040907008944788,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.6303,
+      "step": 4768
+    },
+    {
+      "epoch": 0.10043012903871161,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.5444,
+      "step": 4769
+    },
+    {
+      "epoch": 0.10045118798797534,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.5704,
+      "step": 4770
+    },
+    {
+      "epoch": 0.10047224693723907,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5639,
+      "step": 4771
+    },
+    {
+      "epoch": 0.10049330588650279,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.5744,
+      "step": 4772
+    },
+    {
+      "epoch": 0.10051436483576652,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.0002,
+      "loss": 1.5783,
+      "step": 4773
+    },
+    {
+      "epoch": 0.10053542378503025,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0002,
+      "loss": 1.5711,
+      "step": 4774
+    },
+    {
+      "epoch": 0.10055648273429398,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.0002,
+      "loss": 1.6102,
+      "step": 4775
+    },
+    {
+      "epoch": 0.1005775416835577,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.5853,
+      "step": 4776
+    },
+    {
+      "epoch": 0.10059860063282143,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.5531,
+      "step": 4777
+    },
+    {
+      "epoch": 0.10061965958208516,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.5953,
+      "step": 4778
+    },
+    {
+      "epoch": 0.10064071853134887,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.0002,
+      "loss": 1.6002,
+      "step": 4779
+    },
+    {
+      "epoch": 0.1006617774806126,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0002,
+      "loss": 1.5454,
+      "step": 4780
+    },
+    {
+      "epoch": 0.10068283642987633,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.5686,
+      "step": 4781
+    },
+    {
+      "epoch": 0.10070389537914005,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.5857,
+      "step": 4782
+    },
+    {
+      "epoch": 0.10072495432840378,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.5818,
+      "step": 4783
+    },
+    {
+      "epoch": 0.10074601327766751,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.6072,
+      "step": 4784
+    },
+    {
+      "epoch": 0.10076707222693124,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.5546,
+      "step": 4785
+    },
+    {
+      "epoch": 0.10078813117619496,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.5741,
+      "step": 4786
+    },
+    {
+      "epoch": 0.10080919012545869,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.6379,
+      "step": 4787
+    },
+    {
+      "epoch": 0.10083024907472242,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0002,
+      "loss": 1.6138,
+      "step": 4788
+    },
+    {
+      "epoch": 0.10085130802398615,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5782,
+      "step": 4789
+    },
+    {
+      "epoch": 0.10087236697324987,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0002,
+      "loss": 1.5807,
+      "step": 4790
+    },
+    {
+      "epoch": 0.1008934259225136,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002,
+      "loss": 1.5546,
+      "step": 4791
+    },
+    {
+      "epoch": 0.10091448487177732,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.0002,
+      "loss": 1.5877,
+      "step": 4792
+    },
+    {
+      "epoch": 0.10093554382104104,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0002,
+      "loss": 1.5676,
+      "step": 4793
+    },
+    {
+      "epoch": 0.10095660277030477,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5973,
+      "step": 4794
+    },
+    {
+      "epoch": 0.1009776617195685,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.5917,
+      "step": 4795
+    },
+    {
+      "epoch": 0.10099872066883223,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0002,
+      "loss": 1.5671,
+      "step": 4796
+    },
+    {
+      "epoch": 0.10101977961809595,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.0002,
+      "loss": 1.6046,
+      "step": 4797
+    },
+    {
+      "epoch": 0.10104083856735968,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.5591,
+      "step": 4798
+    },
+    {
+      "epoch": 0.10106189751662341,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0002,
+      "loss": 1.5986,
+      "step": 4799
+    },
+    {
+      "epoch": 0.10108295646588714,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0002,
+      "loss": 1.6103,
+      "step": 4800
+    },
+    {
+      "epoch": 0.10110401541515086,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0002,
+      "loss": 1.5506,
+      "step": 4801
+    },
+    {
+      "epoch": 0.10112507436441459,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0002,
+      "loss": 1.5817,
+      "step": 4802
+    },
+    {
+      "epoch": 0.10114613331367832,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.5941,
+      "step": 4803
+    },
+    {
+      "epoch": 0.10116719226294205,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.5832,
+      "step": 4804
+    },
+    {
+      "epoch": 0.10118825121220577,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.58,
+      "step": 4805
+    },
+    {
+      "epoch": 0.10120931016146949,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.5767,
+      "step": 4806
+    },
+    {
+      "epoch": 0.10123036911073321,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0002,
+      "loss": 1.5936,
+      "step": 4807
+    },
+    {
+      "epoch": 0.10125142805999694,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0002,
+      "loss": 1.606,
+      "step": 4808
+    },
+    {
+      "epoch": 0.10127248700926067,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0002,
+      "loss": 1.5909,
+      "step": 4809
+    },
+    {
+      "epoch": 0.1012935459585244,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0002,
+      "loss": 1.5674,
+      "step": 4810
+    },
+    {
+      "epoch": 0.10131460490778812,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.6087,
+      "step": 4811
+    },
+    {
+      "epoch": 0.10133566385705185,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0002,
+      "loss": 1.622,
+      "step": 4812
+    },
+    {
+      "epoch": 0.10135672280631558,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.0002,
+      "loss": 1.575,
+      "step": 4813
+    },
+    {
+      "epoch": 0.10137778175557931,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.5611,
+      "step": 4814
+    },
+    {
+      "epoch": 0.10139884070484304,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.6018,
+      "step": 4815
+    },
+    {
+      "epoch": 0.10141989965410676,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.0002,
+      "loss": 1.5743,
+      "step": 4816
+    },
+    {
+      "epoch": 0.10144095860337049,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.5478,
+      "step": 4817
+    },
+    {
+      "epoch": 0.10146201755263422,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.5565,
+      "step": 4818
+    },
+    {
+      "epoch": 0.10148307650189793,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.5672,
+      "step": 4819
+    },
+    {
+      "epoch": 0.10150413545116166,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.5675,
+      "step": 4820
+    },
+    {
+      "epoch": 0.10152519440042539,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.5764,
+      "step": 4821
+    },
+    {
+      "epoch": 0.10154625334968911,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0002,
+      "loss": 1.5769,
+      "step": 4822
+    },
+    {
+      "epoch": 0.10156731229895284,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.5642,
+      "step": 4823
+    },
+    {
+      "epoch": 0.10158837124821657,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0002,
+      "loss": 1.5429,
+      "step": 4824
+    },
+    {
+      "epoch": 0.1016094301974803,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.583,
+      "step": 4825
+    },
+    {
+      "epoch": 0.10163048914674402,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5974,
+      "step": 4826
+    },
+    {
+      "epoch": 0.10165154809600775,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.594,
+      "step": 4827
+    },
+    {
+      "epoch": 0.10167260704527148,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0002,
+      "loss": 1.5982,
+      "step": 4828
+    },
+    {
+      "epoch": 0.1016936659945352,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0002,
+      "loss": 1.5589,
+      "step": 4829
+    },
+    {
+      "epoch": 0.10171472494379893,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0002,
+      "loss": 1.5943,
+      "step": 4830
+    },
+    {
+      "epoch": 0.10173578389306266,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.0002,
+      "loss": 1.5532,
+      "step": 4831
+    },
+    {
+      "epoch": 0.10175684284232638,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6086,
+      "step": 4832
+    },
+    {
+      "epoch": 0.1017779017915901,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.5628,
+      "step": 4833
+    },
+    {
+      "epoch": 0.10179896074085383,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.592,
+      "step": 4834
+    },
+    {
+      "epoch": 0.10182001969011756,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.5747,
+      "step": 4835
+    },
+    {
+      "epoch": 0.10184107863938129,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0002,
+      "loss": 1.5881,
+      "step": 4836
+    },
+    {
+      "epoch": 0.10186213758864501,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.6124,
+      "step": 4837
+    },
+    {
+      "epoch": 0.10188319653790874,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.5608,
+      "step": 4838
+    },
+    {
+      "epoch": 0.10190425548717247,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0002,
+      "loss": 1.5702,
+      "step": 4839
+    },
+    {
+      "epoch": 0.1019253144364362,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0002,
+      "loss": 1.5581,
+      "step": 4840
+    },
+    {
+      "epoch": 0.10194637338569992,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.5612,
+      "step": 4841
+    },
+    {
+      "epoch": 0.10196743233496365,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0002,
+      "loss": 1.5977,
+      "step": 4842
+    },
+    {
+      "epoch": 0.10198849128422738,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0002,
+      "loss": 1.5574,
+      "step": 4843
+    },
+    {
+      "epoch": 0.1020095502334911,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.603,
+      "step": 4844
+    },
+    {
+      "epoch": 0.10203060918275483,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0002,
+      "loss": 1.5731,
+      "step": 4845
+    },
+    {
+      "epoch": 0.10205166813201855,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.6153,
+      "step": 4846
+    },
+    {
+      "epoch": 0.10207272708128227,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0002,
+      "loss": 1.5863,
+      "step": 4847
+    },
+    {
+      "epoch": 0.102093786030546,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0002,
+      "loss": 1.5806,
+      "step": 4848
+    },
+    {
+      "epoch": 0.10211484497980973,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.5815,
+      "step": 4849
+    },
+    {
+      "epoch": 0.10213590392907346,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5704,
+      "step": 4850
+    },
+    {
+      "epoch": 0.10215696287833718,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.5774,
+      "step": 4851
+    },
+    {
+      "epoch": 0.10217802182760091,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.5596,
+      "step": 4852
+    },
+    {
+      "epoch": 0.10219908077686464,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5873,
+      "step": 4853
+    },
+    {
+      "epoch": 0.10222013972612837,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5946,
+      "step": 4854
+    },
+    {
+      "epoch": 0.1022411986753921,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.5588,
+      "step": 4855
+    },
+    {
+      "epoch": 0.10226225762465582,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0002,
+      "loss": 1.5951,
+      "step": 4856
+    },
+    {
+      "epoch": 0.10228331657391955,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0002,
+      "loss": 1.597,
+      "step": 4857
+    },
+    {
+      "epoch": 0.10230437552318328,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0002,
+      "loss": 1.578,
+      "step": 4858
+    },
+    {
+      "epoch": 0.10232543447244699,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0002,
+      "loss": 1.567,
+      "step": 4859
+    },
+    {
+      "epoch": 0.10234649342171072,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0002,
+      "loss": 1.5895,
+      "step": 4860
+    },
+    {
+      "epoch": 0.10236755237097445,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0002,
+      "loss": 1.5981,
+      "step": 4861
+    },
+    {
+      "epoch": 0.10238861132023817,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0002,
+      "loss": 1.609,
+      "step": 4862
+    },
+    {
+      "epoch": 0.1024096702695019,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0002,
+      "loss": 1.5454,
+      "step": 4863
+    },
+    {
+      "epoch": 0.10243072921876563,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.5611,
+      "step": 4864
+    },
+    {
+      "epoch": 0.10245178816802936,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0002,
+      "loss": 1.6253,
+      "step": 4865
+    },
+    {
+      "epoch": 0.10247284711729308,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0002,
+      "loss": 1.5831,
+      "step": 4866
+    },
+    {
+      "epoch": 0.10249390606655681,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0002,
+      "loss": 1.5754,
+      "step": 4867
+    },
+    {
+      "epoch": 0.10251496501582054,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002,
+      "loss": 1.5888,
+      "step": 4868
+    },
+    {
+      "epoch": 0.10253602396508427,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.5743,
+      "step": 4869
+    },
+    {
+      "epoch": 0.102557082914348,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.5945,
+      "step": 4870
+    },
+    {
+      "epoch": 0.10257814186361172,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.0002,
+      "loss": 1.5677,
+      "step": 4871
+    },
+    {
+      "epoch": 0.10259920081287544,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.6344,
+      "step": 4872
+    },
+    {
+      "epoch": 0.10262025976213916,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.6317,
+      "step": 4873
+    },
+    {
+      "epoch": 0.10264131871140289,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.5644,
+      "step": 4874
+    },
+    {
+      "epoch": 0.10266237766066662,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.6169,
+      "step": 4875
+    },
+    {
+      "epoch": 0.10268343660993035,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0002,
+      "loss": 1.5953,
+      "step": 4876
+    },
+    {
+      "epoch": 0.10270449555919407,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.5414,
+      "step": 4877
+    },
+    {
+      "epoch": 0.1027255545084578,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.5563,
+      "step": 4878
+    },
+    {
+      "epoch": 0.10274661345772153,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0002,
+      "loss": 1.5852,
+      "step": 4879
+    },
+    {
+      "epoch": 0.10276767240698526,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0002,
+      "loss": 1.5693,
+      "step": 4880
+    },
+    {
+      "epoch": 0.10278873135624898,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0002,
+      "loss": 1.5896,
+      "step": 4881
+    },
+    {
+      "epoch": 0.10280979030551271,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.6075,
+      "step": 4882
+    },
+    {
+      "epoch": 0.10283084925477644,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0002,
+      "loss": 1.5907,
+      "step": 4883
+    },
+    {
+      "epoch": 0.10285190820404017,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.5803,
+      "step": 4884
+    },
+    {
+      "epoch": 0.10287296715330388,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5848,
+      "step": 4885
+    },
+    {
+      "epoch": 0.1028940261025676,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0002,
+      "loss": 1.5559,
+      "step": 4886
+    },
+    {
+      "epoch": 0.10291508505183133,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0002,
+      "loss": 1.5638,
+      "step": 4887
+    },
+    {
+      "epoch": 0.10293614400109506,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.5466,
+      "step": 4888
+    },
+    {
+      "epoch": 0.10295720295035879,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.0002,
+      "loss": 1.6031,
+      "step": 4889
+    },
+    {
+      "epoch": 0.10297826189962252,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5845,
+      "step": 4890
+    },
+    {
+      "epoch": 0.10299932084888624,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.602,
+      "step": 4891
+    },
+    {
+      "epoch": 0.10302037979814997,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.5532,
+      "step": 4892
+    },
+    {
+      "epoch": 0.1030414387474137,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0002,
+      "loss": 1.5861,
+      "step": 4893
+    },
+    {
+      "epoch": 0.10306249769667743,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0002,
+      "loss": 1.6094,
+      "step": 4894
+    },
+    {
+      "epoch": 0.10308355664594115,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0002,
+      "loss": 1.5706,
+      "step": 4895
+    },
+    {
+      "epoch": 0.10310461559520488,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.0002,
+      "loss": 1.6062,
+      "step": 4896
+    },
+    {
+      "epoch": 0.10312567454446861,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.558,
+      "step": 4897
+    },
+    {
+      "epoch": 0.10314673349373234,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.6091,
+      "step": 4898
+    },
+    {
+      "epoch": 0.10316779244299605,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.0002,
+      "loss": 1.5385,
+      "step": 4899
+    },
+    {
+      "epoch": 0.10318885139225978,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0002,
+      "loss": 1.5819,
+      "step": 4900
+    },
+    {
+      "epoch": 0.1032099103415235,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0002,
+      "loss": 1.5955,
+      "step": 4901
+    },
+    {
+      "epoch": 0.10323096929078723,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.0002,
+      "loss": 1.5784,
+      "step": 4902
+    },
+    {
+      "epoch": 0.10325202824005096,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5531,
+      "step": 4903
+    },
+    {
+      "epoch": 0.10327308718931469,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002,
+      "loss": 1.5678,
+      "step": 4904
+    },
+    {
+      "epoch": 0.10329414613857842,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0002,
+      "loss": 1.585,
+      "step": 4905
+    },
+    {
+      "epoch": 0.10331520508784214,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.5931,
+      "step": 4906
+    },
+    {
+      "epoch": 0.10333626403710587,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.5806,
+      "step": 4907
+    },
+    {
+      "epoch": 0.1033573229863696,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0002,
+      "loss": 1.6001,
+      "step": 4908
+    },
+    {
+      "epoch": 0.10337838193563333,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.5692,
+      "step": 4909
+    },
+    {
+      "epoch": 0.10339944088489705,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5904,
+      "step": 4910
+    },
+    {
+      "epoch": 0.10342049983416078,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0002,
+      "loss": 1.594,
+      "step": 4911
+    },
+    {
+      "epoch": 0.1034415587834245,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.5801,
+      "step": 4912
+    },
+    {
+      "epoch": 0.10346261773268822,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0002,
+      "loss": 1.5826,
+      "step": 4913
+    },
+    {
+      "epoch": 0.10348367668195195,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0002,
+      "loss": 1.569,
+      "step": 4914
+    },
+    {
+      "epoch": 0.10350473563121568,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.5578,
+      "step": 4915
+    },
+    {
+      "epoch": 0.1035257945804794,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5837,
+      "step": 4916
+    },
+    {
+      "epoch": 0.10354685352974313,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.5951,
+      "step": 4917
+    },
+    {
+      "epoch": 0.10356791247900686,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0002,
+      "loss": 1.5623,
+      "step": 4918
+    },
+    {
+      "epoch": 0.10358897142827059,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0002,
+      "loss": 1.5646,
+      "step": 4919
+    },
+    {
+      "epoch": 0.10361003037753432,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.5646,
+      "step": 4920
+    },
+    {
+      "epoch": 0.10363108932679804,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.596,
+      "step": 4921
+    },
+    {
+      "epoch": 0.10365214827606177,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.5915,
+      "step": 4922
+    },
+    {
+      "epoch": 0.1036732072253255,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.5872,
+      "step": 4923
+    },
+    {
+      "epoch": 0.10369426617458923,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.5703,
+      "step": 4924
+    },
+    {
+      "epoch": 0.10371532512385294,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.5795,
+      "step": 4925
+    },
+    {
+      "epoch": 0.10373638407311667,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.5715,
+      "step": 4926
+    },
+    {
+      "epoch": 0.1037574430223804,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0002,
+      "loss": 1.5886,
+      "step": 4927
+    },
+    {
+      "epoch": 0.10377850197164412,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0002,
+      "loss": 1.6063,
+      "step": 4928
+    },
+    {
+      "epoch": 0.10379956092090785,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0002,
+      "loss": 1.5746,
+      "step": 4929
+    },
+    {
+      "epoch": 0.10382061987017158,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.5695,
+      "step": 4930
+    },
+    {
+      "epoch": 0.1038416788194353,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.5428,
+      "step": 4931
+    },
+    {
+      "epoch": 0.10386273776869903,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.5612,
+      "step": 4932
+    },
+    {
+      "epoch": 0.10388379671796276,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0002,
+      "loss": 1.5604,
+      "step": 4933
+    },
+    {
+      "epoch": 0.10390485566722649,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.5352,
+      "step": 4934
+    },
+    {
+      "epoch": 0.10392591461649021,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0002,
+      "loss": 1.5679,
+      "step": 4935
+    },
+    {
+      "epoch": 0.10394697356575394,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.0002,
+      "loss": 1.5609,
+      "step": 4936
+    },
+    {
+      "epoch": 0.10396803251501767,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.5896,
+      "step": 4937
+    },
+    {
+      "epoch": 0.10398909146428138,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.0002,
+      "loss": 1.589,
+      "step": 4938
+    },
+    {
+      "epoch": 0.10401015041354511,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0002,
+      "loss": 1.5784,
+      "step": 4939
+    },
+    {
+      "epoch": 0.10403120936280884,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002,
+      "loss": 1.6108,
+      "step": 4940
+    },
+    {
+      "epoch": 0.10405226831207257,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.5645,
+      "step": 4941
+    },
+    {
+      "epoch": 0.1040733272613363,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0002,
+      "loss": 1.5961,
+      "step": 4942
+    },
+    {
+      "epoch": 0.10409438621060002,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0002,
+      "loss": 1.6042,
+      "step": 4943
+    },
+    {
+      "epoch": 0.10411544515986375,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.5785,
+      "step": 4944
+    },
+    {
+      "epoch": 0.10413650410912748,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.5824,
+      "step": 4945
+    },
+    {
+      "epoch": 0.1041575630583912,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.5756,
+      "step": 4946
+    },
+    {
+      "epoch": 0.10417862200765493,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0002,
+      "loss": 1.592,
+      "step": 4947
+    },
+    {
+      "epoch": 0.10419968095691866,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0002,
+      "loss": 1.6292,
+      "step": 4948
+    },
+    {
+      "epoch": 0.10422073990618239,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.6209,
+      "step": 4949
+    },
+    {
+      "epoch": 0.10424179885544611,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.5519,
+      "step": 4950
+    },
+    {
+      "epoch": 0.10426285780470984,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0002,
+      "loss": 1.5642,
+      "step": 4951
+    },
+    {
+      "epoch": 0.10428391675397355,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0002,
+      "loss": 1.593,
+      "step": 4952
+    },
+    {
+      "epoch": 0.10430497570323728,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.5932,
+      "step": 4953
+    },
+    {
+      "epoch": 0.10432603465250101,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0002,
+      "loss": 1.5937,
+      "step": 4954
+    },
+    {
+      "epoch": 0.10434709360176474,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.5626,
+      "step": 4955
+    },
+    {
+      "epoch": 0.10436815255102846,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.5514,
+      "step": 4956
+    },
+    {
+      "epoch": 0.10438921150029219,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0002,
+      "loss": 1.5662,
+      "step": 4957
+    },
+    {
+      "epoch": 0.10441027044955592,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0002,
+      "loss": 1.6007,
+      "step": 4958
+    },
+    {
+      "epoch": 0.10443132939881965,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.5734,
+      "step": 4959
+    },
+    {
+      "epoch": 0.10445238834808337,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0002,
+      "loss": 1.5737,
+      "step": 4960
+    },
+    {
+      "epoch": 0.1044734472973471,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0002,
+      "loss": 1.5901,
+      "step": 4961
+    },
+    {
+      "epoch": 0.10449450624661083,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5684,
+      "step": 4962
+    },
+    {
+      "epoch": 0.10451556519587456,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.5872,
+      "step": 4963
+    },
+    {
+      "epoch": 0.10453662414513828,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0002,
+      "loss": 1.607,
+      "step": 4964
+    },
+    {
+      "epoch": 0.104557683094402,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0002,
+      "loss": 1.5951,
+      "step": 4965
+    },
+    {
+      "epoch": 0.10457874204366573,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0002,
+      "loss": 1.6145,
+      "step": 4966
+    },
+    {
+      "epoch": 0.10459980099292945,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.6245,
+      "step": 4967
+    },
+    {
+      "epoch": 0.10462085994219318,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.0002,
+      "loss": 1.5643,
+      "step": 4968
+    },
+    {
+      "epoch": 0.10464191889145691,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5588,
+      "step": 4969
+    },
+    {
+      "epoch": 0.10466297784072064,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.6144,
+      "step": 4970
+    },
+    {
+      "epoch": 0.10468403678998436,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5883,
+      "step": 4971
+    },
+    {
+      "epoch": 0.10470509573924809,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0002,
+      "loss": 1.6145,
+      "step": 4972
+    },
+    {
+      "epoch": 0.10472615468851182,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0002,
+      "loss": 1.577,
+      "step": 4973
+    },
+    {
+      "epoch": 0.10474721363777555,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.5911,
+      "step": 4974
+    },
+    {
+      "epoch": 0.10476827258703927,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.6318,
+      "step": 4975
+    },
+    {
+      "epoch": 0.104789331536303,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0002,
+      "loss": 1.5619,
+      "step": 4976
+    },
+    {
+      "epoch": 0.10481039048556673,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0002,
+      "loss": 1.5573,
+      "step": 4977
+    },
+    {
+      "epoch": 0.10483144943483044,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0002,
+      "loss": 1.5739,
+      "step": 4978
+    },
+    {
+      "epoch": 0.10485250838409417,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0002,
+      "loss": 1.5615,
+      "step": 4979
+    },
+    {
+      "epoch": 0.1048735673333579,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002,
+      "loss": 1.5528,
+      "step": 4980
+    },
+    {
+      "epoch": 0.10489462628262163,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.5809,
+      "step": 4981
+    },
+    {
+      "epoch": 0.10491568523188535,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.563,
+      "step": 4982
+    },
+    {
+      "epoch": 0.10493674418114908,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0002,
+      "loss": 1.5721,
+      "step": 4983
+    },
+    {
+      "epoch": 0.10495780313041281,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6004,
+      "step": 4984
+    },
+    {
+      "epoch": 0.10497886207967654,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0002,
+      "loss": 1.619,
+      "step": 4985
+    },
+    {
+      "epoch": 0.10499992102894026,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.5889,
+      "step": 4986
+    },
+    {
+      "epoch": 0.10502097997820399,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0002,
+      "loss": 1.5847,
+      "step": 4987
+    },
+    {
+      "epoch": 0.10504203892746772,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0002,
+      "loss": 1.5467,
+      "step": 4988
+    },
+    {
+      "epoch": 0.10506309787673145,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002,
+      "loss": 1.566,
+      "step": 4989
+    },
+    {
+      "epoch": 0.10508415682599517,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0002,
+      "loss": 1.547,
+      "step": 4990
+    },
+    {
+      "epoch": 0.10510521577525889,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.5939,
+      "step": 4991
+    },
+    {
+      "epoch": 0.10512627472452261,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.5902,
+      "step": 4992
+    },
+    {
+      "epoch": 0.10514733367378634,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0002,
+      "loss": 1.6007,
+      "step": 4993
+    },
+    {
+      "epoch": 0.10516839262305007,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0002,
+      "loss": 1.5982,
+      "step": 4994
+    },
+    {
+      "epoch": 0.1051894515723138,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0002,
+      "loss": 1.5839,
+      "step": 4995
+    },
+    {
+      "epoch": 0.10521051052157752,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.589,
+      "step": 4996
+    },
+    {
+      "epoch": 0.10523156947084125,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.5983,
+      "step": 4997
+    },
+    {
+      "epoch": 0.10525262842010498,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.5936,
+      "step": 4998
+    },
+    {
+      "epoch": 0.10527368736936871,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.5952,
+      "step": 4999
+    },
+    {
+      "epoch": 0.10529474631863243,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.5795,
+      "step": 5000
+    },
+    {
+      "epoch": 0.10529474631863243,
+      "eval_loss": 2.2983527183532715,
+      "eval_runtime": 908.9901,
+      "eval_samples_per_second": 67.988,
+      "eval_steps_per_second": 2.125,
+      "step": 5000
+    },
+    {
+      "epoch": 0.10531580526789616,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0002,
+      "loss": 1.5877,
+      "step": 5001
+    },
+    {
+      "epoch": 0.10533686421715989,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.5572,
+      "step": 5002
+    },
+    {
+      "epoch": 0.10535792316642362,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.5884,
+      "step": 5003
+    },
+    {
+      "epoch": 0.10537898211568734,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.5688,
+      "step": 5004
+    },
+    {
+      "epoch": 0.10540004106495106,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.552,
+      "step": 5005
+    },
+    {
+      "epoch": 0.10542110001421479,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5849,
+      "step": 5006
+    },
+    {
+      "epoch": 0.10544215896347851,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0002,
+      "loss": 1.6067,
+      "step": 5007
+    },
+    {
+      "epoch": 0.10546321791274224,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0002,
+      "loss": 1.5991,
+      "step": 5008
+    },
+    {
+      "epoch": 0.10548427686200597,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.5874,
+      "step": 5009
+    },
+    {
+      "epoch": 0.1055053358112697,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.6083,
+      "step": 5010
+    },
+    {
+      "epoch": 0.10552639476053342,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0002,
+      "loss": 1.6053,
+      "step": 5011
+    },
+    {
+      "epoch": 0.10554745370979715,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.0002,
+      "loss": 1.5833,
+      "step": 5012
+    },
+    {
+      "epoch": 0.10556851265906088,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0002,
+      "loss": 1.5442,
+      "step": 5013
+    },
+    {
+      "epoch": 0.1055895716083246,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5651,
+      "step": 5014
+    },
+    {
+      "epoch": 0.10561063055758833,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.5951,
+      "step": 5015
+    },
+    {
+      "epoch": 0.10563168950685206,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0002,
+      "loss": 1.5547,
+      "step": 5016
+    },
+    {
+      "epoch": 0.10565274845611579,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.5629,
+      "step": 5017
+    },
+    {
+      "epoch": 0.1056738074053795,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0002,
+      "loss": 1.5629,
+      "step": 5018
+    },
+    {
+      "epoch": 0.10569486635464323,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0002,
+      "loss": 1.584,
+      "step": 5019
+    },
+    {
+      "epoch": 0.10571592530390696,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5783,
+      "step": 5020
+    },
+    {
+      "epoch": 0.10573698425317068,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5965,
+      "step": 5021
+    },
+    {
+      "epoch": 0.10575804320243441,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.5732,
+      "step": 5022
+    },
+    {
+      "epoch": 0.10577910215169814,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0002,
+      "loss": 1.6001,
+      "step": 5023
+    },
+    {
+      "epoch": 0.10580016110096187,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.6072,
+      "step": 5024
+    },
+    {
+      "epoch": 0.1058212200502256,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002,
+      "loss": 1.5861,
+      "step": 5025
+    },
+    {
+      "epoch": 0.10584227899948932,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.574,
+      "step": 5026
+    },
+    {
+      "epoch": 0.10586333794875305,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0002,
+      "loss": 1.5816,
+      "step": 5027
+    },
+    {
+      "epoch": 0.10588439689801678,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0002,
+      "loss": 1.597,
+      "step": 5028
+    },
+    {
+      "epoch": 0.1059054558472805,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5803,
+      "step": 5029
+    },
+    {
+      "epoch": 0.10592651479654423,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.6053,
+      "step": 5030
+    },
+    {
+      "epoch": 0.10594757374580795,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.5483,
+      "step": 5031
+    },
+    {
+      "epoch": 0.10596863269507167,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.0002,
+      "loss": 1.5434,
+      "step": 5032
+    },
+    {
+      "epoch": 0.1059896916443354,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0002,
+      "loss": 1.5883,
+      "step": 5033
+    },
+    {
+      "epoch": 0.10601075059359913,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.592,
+      "step": 5034
+    },
+    {
+      "epoch": 0.10603180954286286,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0002,
+      "loss": 1.5701,
+      "step": 5035
+    },
+    {
+      "epoch": 0.10605286849212658,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0002,
+      "loss": 1.5742,
+      "step": 5036
+    },
+    {
+      "epoch": 0.10607392744139031,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.5961,
+      "step": 5037
+    },
+    {
+      "epoch": 0.10609498639065404,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.0002,
+      "loss": 1.5982,
+      "step": 5038
+    },
+    {
+      "epoch": 0.10611604533991777,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.5774,
+      "step": 5039
+    },
+    {
+      "epoch": 0.1061371042891815,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0002,
+      "loss": 1.556,
+      "step": 5040
+    },
+    {
+      "epoch": 0.10615816323844522,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.57,
+      "step": 5041
+    },
+    {
+      "epoch": 0.10617922218770895,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0002,
+      "loss": 1.599,
+      "step": 5042
+    },
+    {
+      "epoch": 0.10620028113697268,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.5853,
+      "step": 5043
+    },
+    {
+      "epoch": 0.10622134008623639,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.5436,
+      "step": 5044
+    },
+    {
+      "epoch": 0.10624239903550012,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0002,
+      "loss": 1.5619,
+      "step": 5045
+    },
+    {
+      "epoch": 0.10626345798476385,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0002,
+      "loss": 1.584,
+      "step": 5046
+    },
+    {
+      "epoch": 0.10628451693402757,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6059,
+      "step": 5047
+    },
+    {
+      "epoch": 0.1063055758832913,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0002,
+      "loss": 1.6124,
+      "step": 5048
+    },
+    {
+      "epoch": 0.10632663483255503,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002,
+      "loss": 1.5814,
+      "step": 5049
+    },
+    {
+      "epoch": 0.10634769378181876,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0002,
+      "loss": 1.5583,
+      "step": 5050
+    },
+    {
+      "epoch": 0.10636875273108248,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.5633,
+      "step": 5051
+    },
+    {
+      "epoch": 0.10638981168034621,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0002,
+      "loss": 1.5957,
+      "step": 5052
+    },
+    {
+      "epoch": 0.10641087062960994,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5357,
+      "step": 5053
+    },
+    {
+      "epoch": 0.10643192957887367,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.5512,
+      "step": 5054
+    },
+    {
+      "epoch": 0.1064529885281374,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.5994,
+      "step": 5055
+    },
+    {
+      "epoch": 0.10647404747740112,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.5752,
+      "step": 5056
+    },
+    {
+      "epoch": 0.10649510642666485,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.6052,
+      "step": 5057
+    },
+    {
+      "epoch": 0.10651616537592856,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.5898,
+      "step": 5058
+    },
+    {
+      "epoch": 0.10653722432519229,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.5859,
+      "step": 5059
+    },
+    {
+      "epoch": 0.10655828327445602,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0002,
+      "loss": 1.568,
+      "step": 5060
+    },
+    {
+      "epoch": 0.10657934222371974,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0002,
+      "loss": 1.594,
+      "step": 5061
+    },
+    {
+      "epoch": 0.10660040117298347,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.5896,
+      "step": 5062
+    },
+    {
+      "epoch": 0.1066214601222472,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002,
+      "loss": 1.5987,
+      "step": 5063
+    },
+    {
+      "epoch": 0.10664251907151093,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0002,
+      "loss": 1.5697,
+      "step": 5064
+    },
+    {
+      "epoch": 0.10666357802077465,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002,
+      "loss": 1.568,
+      "step": 5065
+    },
+    {
+      "epoch": 0.10668463697003838,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.5795,
+      "step": 5066
+    },
+    {
+      "epoch": 0.10670569591930211,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.5785,
+      "step": 5067
+    },
+    {
+      "epoch": 0.10672675486856584,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.543,
+      "step": 5068
+    },
+    {
+      "epoch": 0.10674781381782956,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0002,
+      "loss": 1.5794,
+      "step": 5069
+    },
+    {
+      "epoch": 0.10676887276709329,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0002,
+      "loss": 1.5456,
+      "step": 5070
+    },
+    {
+      "epoch": 0.106789931716357,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002,
+      "loss": 1.572,
+      "step": 5071
+    },
+    {
+      "epoch": 0.10681099066562073,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.5465,
+      "step": 5072
+    },
+    {
+      "epoch": 0.10683204961488446,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0002,
+      "loss": 1.5574,
+      "step": 5073
+    },
+    {
+      "epoch": 0.10685310856414819,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.0002,
+      "loss": 1.5703,
+      "step": 5074
+    },
+    {
+      "epoch": 0.10687416751341192,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0002,
+      "loss": 1.5909,
+      "step": 5075
+    },
+    {
+      "epoch": 0.10689522646267564,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0002,
+      "loss": 1.5884,
+      "step": 5076
+    },
+    {
+      "epoch": 0.10691628541193937,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.5698,
+      "step": 5077
+    },
+    {
+      "epoch": 0.1069373443612031,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0002,
+      "loss": 1.5978,
+      "step": 5078
+    },
+    {
+      "epoch": 0.10695840331046683,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5992,
+      "step": 5079
+    },
+    {
+      "epoch": 0.10697946225973055,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.5883,
+      "step": 5080
+    },
+    {
+      "epoch": 0.10700052120899428,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.5926,
+      "step": 5081
+    },
+    {
+      "epoch": 0.10702158015825801,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.5791,
+      "step": 5082
+    },
+    {
+      "epoch": 0.10704263910752174,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.5717,
+      "step": 5083
+    },
+    {
+      "epoch": 0.10706369805678545,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.5728,
+      "step": 5084
+    },
+    {
+      "epoch": 0.10708475700604918,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.5631,
+      "step": 5085
+    },
+    {
+      "epoch": 0.1071058159553129,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0002,
+      "loss": 1.6017,
+      "step": 5086
+    },
+    {
+      "epoch": 0.10712687490457663,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0002,
+      "loss": 1.5889,
+      "step": 5087
+    },
+    {
+      "epoch": 0.10714793385384036,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0002,
+      "loss": 1.5647,
+      "step": 5088
+    },
+    {
+      "epoch": 0.10716899280310409,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.5824,
+      "step": 5089
+    },
+    {
+      "epoch": 0.10719005175236782,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0002,
+      "loss": 1.5736,
+      "step": 5090
+    },
+    {
+      "epoch": 0.10721111070163154,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0002,
+      "loss": 1.6236,
+      "step": 5091
+    },
+    {
+      "epoch": 0.10723216965089527,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.5461,
+      "step": 5092
+    },
+    {
+      "epoch": 0.107253228600159,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.5999,
+      "step": 5093
+    },
+    {
+      "epoch": 0.10727428754942273,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.576,
+      "step": 5094
+    },
+    {
+      "epoch": 0.10729534649868645,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.5385,
+      "step": 5095
+    },
+    {
+      "epoch": 0.10731640544795018,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.5842,
+      "step": 5096
+    },
+    {
+      "epoch": 0.1073374643972139,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0002,
+      "loss": 1.5614,
+      "step": 5097
+    },
+    {
+      "epoch": 0.10735852334647762,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.6028,
+      "step": 5098
+    },
+    {
+      "epoch": 0.10737958229574135,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0002,
+      "loss": 1.5764,
+      "step": 5099
+    },
+    {
+      "epoch": 0.10740064124500508,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5836,
+      "step": 5100
+    },
+    {
+      "epoch": 0.1074217001942688,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5847,
+      "step": 5101
+    },
+    {
+      "epoch": 0.10744275914353253,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.5996,
+      "step": 5102
+    },
+    {
+      "epoch": 0.10746381809279626,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002,
+      "loss": 1.6194,
+      "step": 5103
+    },
+    {
+      "epoch": 0.10748487704205999,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.5639,
+      "step": 5104
+    },
+    {
+      "epoch": 0.10750593599132371,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0002,
+      "loss": 1.5961,
+      "step": 5105
+    },
+    {
+      "epoch": 0.10752699494058744,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5944,
+      "step": 5106
+    },
+    {
+      "epoch": 0.10754805388985117,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.5796,
+      "step": 5107
+    },
+    {
+      "epoch": 0.1075691128391149,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.5355,
+      "step": 5108
+    },
+    {
+      "epoch": 0.10759017178837862,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.0002,
+      "loss": 1.5677,
+      "step": 5109
+    },
+    {
+      "epoch": 0.10761123073764235,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.5873,
+      "step": 5110
+    },
+    {
+      "epoch": 0.10763228968690607,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0002,
+      "loss": 1.5757,
+      "step": 5111
+    },
+    {
+      "epoch": 0.1076533486361698,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0002,
+      "loss": 1.5909,
+      "step": 5112
+    },
+    {
+      "epoch": 0.10767440758543352,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0002,
+      "loss": 1.5969,
+      "step": 5113
+    },
+    {
+      "epoch": 0.10769546653469725,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0002,
+      "loss": 1.6183,
+      "step": 5114
+    },
+    {
+      "epoch": 0.10771652548396098,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0002,
+      "loss": 1.5635,
+      "step": 5115
+    },
+    {
+      "epoch": 0.1077375844332247,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0002,
+      "loss": 1.5932,
+      "step": 5116
+    },
+    {
+      "epoch": 0.10775864338248843,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0002,
+      "loss": 1.5818,
+      "step": 5117
+    },
+    {
+      "epoch": 0.10777970233175216,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5889,
+      "step": 5118
+    },
+    {
+      "epoch": 0.10780076128101589,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5724,
+      "step": 5119
+    },
+    {
+      "epoch": 0.10782182023027961,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0002,
+      "loss": 1.5715,
+      "step": 5120
+    },
+    {
+      "epoch": 0.10784287917954334,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5324,
+      "step": 5121
+    },
+    {
+      "epoch": 0.10786393812880707,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.5722,
+      "step": 5122
+    },
+    {
+      "epoch": 0.1078849970780708,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.5473,
+      "step": 5123
+    },
+    {
+      "epoch": 0.10790605602733451,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.5598,
+      "step": 5124
+    },
+    {
+      "epoch": 0.10792711497659824,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.5657,
+      "step": 5125
+    },
+    {
+      "epoch": 0.10794817392586196,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5816,
+      "step": 5126
+    },
+    {
+      "epoch": 0.10796923287512569,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.577,
+      "step": 5127
+    },
+    {
+      "epoch": 0.10799029182438942,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0002,
+      "loss": 1.5868,
+      "step": 5128
+    },
+    {
+      "epoch": 0.10801135077365315,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0002,
+      "loss": 1.5805,
+      "step": 5129
+    },
+    {
+      "epoch": 0.10803240972291688,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.587,
+      "step": 5130
+    },
+    {
+      "epoch": 0.1080534686721806,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0002,
+      "loss": 1.5661,
+      "step": 5131
+    },
+    {
+      "epoch": 0.10807452762144433,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0002,
+      "loss": 1.6286,
+      "step": 5132
+    },
+    {
+      "epoch": 0.10809558657070806,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.0002,
+      "loss": 1.599,
+      "step": 5133
+    },
+    {
+      "epoch": 0.10811664551997179,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.5453,
+      "step": 5134
+    },
+    {
+      "epoch": 0.10813770446923551,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0002,
+      "loss": 1.5705,
+      "step": 5135
+    },
+    {
+      "epoch": 0.10815876341849924,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.5694,
+      "step": 5136
+    },
+    {
+      "epoch": 0.10817982236776295,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.6127,
+      "step": 5137
+    },
+    {
+      "epoch": 0.10820088131702668,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0002,
+      "loss": 1.553,
+      "step": 5138
+    },
+    {
+      "epoch": 0.10822194026629041,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0002,
+      "loss": 1.5719,
+      "step": 5139
+    },
+    {
+      "epoch": 0.10824299921555414,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.5743,
+      "step": 5140
+    },
+    {
+      "epoch": 0.10826405816481786,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0002,
+      "loss": 1.5776,
+      "step": 5141
+    },
+    {
+      "epoch": 0.10828511711408159,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.5593,
+      "step": 5142
+    },
+    {
+      "epoch": 0.10830617606334532,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0002,
+      "loss": 1.5555,
+      "step": 5143
+    },
+    {
+      "epoch": 0.10832723501260905,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.5755,
+      "step": 5144
+    },
+    {
+      "epoch": 0.10834829396187277,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0002,
+      "loss": 1.5793,
+      "step": 5145
+    },
+    {
+      "epoch": 0.1083693529111365,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0002,
+      "loss": 1.548,
+      "step": 5146
+    },
+    {
+      "epoch": 0.10839041186040023,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.5994,
+      "step": 5147
+    },
+    {
+      "epoch": 0.10841147080966396,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.5818,
+      "step": 5148
+    },
+    {
+      "epoch": 0.10843252975892768,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.5822,
+      "step": 5149
+    },
+    {
+      "epoch": 0.1084535887081914,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.0002,
+      "loss": 1.5955,
+      "step": 5150
+    },
+    {
+      "epoch": 0.10847464765745513,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.0002,
+      "loss": 1.5635,
+      "step": 5151
+    },
+    {
+      "epoch": 0.10849570660671885,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0002,
+      "loss": 1.5854,
+      "step": 5152
+    },
+    {
+      "epoch": 0.10851676555598258,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.5592,
+      "step": 5153
+    },
+    {
+      "epoch": 0.10853782450524631,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0002,
+      "loss": 1.5739,
+      "step": 5154
+    },
+    {
+      "epoch": 0.10855888345451004,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.578,
+      "step": 5155
+    },
+    {
+      "epoch": 0.10857994240377376,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0002,
+      "loss": 1.5922,
+      "step": 5156
+    },
+    {
+      "epoch": 0.10860100135303749,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.6204,
+      "step": 5157
+    },
+    {
+      "epoch": 0.10862206030230122,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.5832,
+      "step": 5158
+    },
+    {
+      "epoch": 0.10864311925156495,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.5465,
+      "step": 5159
+    },
+    {
+      "epoch": 0.10866417820082867,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.5998,
+      "step": 5160
+    },
+    {
+      "epoch": 0.1086852371500924,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5541,
+      "step": 5161
+    },
+    {
+      "epoch": 0.10870629609935613,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0002,
+      "loss": 1.5642,
+      "step": 5162
+    },
+    {
+      "epoch": 0.10872735504861986,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5657,
+      "step": 5163
+    },
+    {
+      "epoch": 0.10874841399788357,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5815,
+      "step": 5164
+    },
+    {
+      "epoch": 0.1087694729471473,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.59,
+      "step": 5165
+    },
+    {
+      "epoch": 0.10879053189641102,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5766,
+      "step": 5166
+    },
+    {
+      "epoch": 0.10881159084567475,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.5954,
+      "step": 5167
+    },
+    {
+      "epoch": 0.10883264979493848,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0002,
+      "loss": 1.5988,
+      "step": 5168
+    },
+    {
+      "epoch": 0.10885370874420221,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.5494,
+      "step": 5169
+    },
+    {
+      "epoch": 0.10887476769346593,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.5965,
+      "step": 5170
+    },
+    {
+      "epoch": 0.10889582664272966,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.6017,
+      "step": 5171
+    },
+    {
+      "epoch": 0.10891688559199339,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.5797,
+      "step": 5172
+    },
+    {
+      "epoch": 0.10893794454125712,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0002,
+      "loss": 1.5661,
+      "step": 5173
+    },
+    {
+      "epoch": 0.10895900349052084,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.0002,
+      "loss": 1.611,
+      "step": 5174
+    },
+    {
+      "epoch": 0.10898006243978457,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0002,
+      "loss": 1.6259,
+      "step": 5175
+    },
+    {
+      "epoch": 0.1090011213890483,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0002,
+      "loss": 1.5817,
+      "step": 5176
+    },
+    {
+      "epoch": 0.10902218033831201,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5861,
+      "step": 5177
+    },
+    {
+      "epoch": 0.10904323928757574,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.0002,
+      "loss": 1.552,
+      "step": 5178
+    },
+    {
+      "epoch": 0.10906429823683947,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002,
+      "loss": 1.5803,
+      "step": 5179
+    },
+    {
+      "epoch": 0.1090853571861032,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5778,
+      "step": 5180
+    },
+    {
+      "epoch": 0.10910641613536692,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002,
+      "loss": 1.5706,
+      "step": 5181
+    },
+    {
+      "epoch": 0.10912747508463065,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0002,
+      "loss": 1.5718,
+      "step": 5182
+    },
+    {
+      "epoch": 0.10914853403389438,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.5881,
+      "step": 5183
+    },
+    {
+      "epoch": 0.1091695929831581,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0002,
+      "loss": 1.5666,
+      "step": 5184
+    },
+    {
+      "epoch": 0.10919065193242183,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0002,
+      "loss": 1.5943,
+      "step": 5185
+    },
+    {
+      "epoch": 0.10921171088168556,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6,
+      "step": 5186
+    },
+    {
+      "epoch": 0.10923276983094929,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5662,
+      "step": 5187
+    },
+    {
+      "epoch": 0.10925382878021302,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0002,
+      "loss": 1.5766,
+      "step": 5188
+    },
+    {
+      "epoch": 0.10927488772947674,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0002,
+      "loss": 1.5836,
+      "step": 5189
+    },
+    {
+      "epoch": 0.10929594667874046,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0002,
+      "loss": 1.5714,
+      "step": 5190
+    },
+    {
+      "epoch": 0.10931700562800419,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0002,
+      "loss": 1.5867,
+      "step": 5191
+    },
+    {
+      "epoch": 0.10933806457726791,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.5771,
+      "step": 5192
+    },
+    {
+      "epoch": 0.10935912352653164,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0002,
+      "loss": 1.5627,
+      "step": 5193
+    },
+    {
+      "epoch": 0.10938018247579537,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0002,
+      "loss": 1.5891,
+      "step": 5194
+    },
+    {
+      "epoch": 0.1094012414250591,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.5917,
+      "step": 5195
+    },
+    {
+      "epoch": 0.10942230037432282,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0002,
+      "loss": 1.5678,
+      "step": 5196
+    },
+    {
+      "epoch": 0.10944335932358655,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.5946,
+      "step": 5197
+    },
+    {
+      "epoch": 0.10946441827285028,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.6273,
+      "step": 5198
+    },
+    {
+      "epoch": 0.109485477222114,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0002,
+      "loss": 1.572,
+      "step": 5199
+    },
+    {
+      "epoch": 0.10950653617137773,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002,
+      "loss": 1.5797,
+      "step": 5200
+    },
+    {
+      "epoch": 0.10952759512064146,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0002,
+      "loss": 1.5816,
+      "step": 5201
+    },
+    {
+      "epoch": 0.10954865406990519,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0002,
+      "loss": 1.5849,
+      "step": 5202
+    },
+    {
+      "epoch": 0.1095697130191689,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0002,
+      "loss": 1.5933,
+      "step": 5203
+    },
+    {
+      "epoch": 0.10959077196843263,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0002,
+      "loss": 1.5954,
+      "step": 5204
+    },
+    {
+      "epoch": 0.10961183091769636,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.5984,
+      "step": 5205
+    },
+    {
+      "epoch": 0.10963288986696008,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.5773,
+      "step": 5206
+    },
+    {
+      "epoch": 0.10965394881622381,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0002,
+      "loss": 1.5774,
+      "step": 5207
+    },
+    {
+      "epoch": 0.10967500776548754,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002,
+      "loss": 1.5877,
+      "step": 5208
+    },
+    {
+      "epoch": 0.10969606671475127,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.5948,
+      "step": 5209
+    },
+    {
+      "epoch": 0.109717125664015,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.63,
+      "step": 5210
+    },
+    {
+      "epoch": 0.10973818461327872,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0002,
+      "loss": 1.6111,
+      "step": 5211
+    },
+    {
+      "epoch": 0.10975924356254245,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0002,
+      "loss": 1.5785,
+      "step": 5212
+    },
+    {
+      "epoch": 0.10978030251180618,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.0002,
+      "loss": 1.6114,
+      "step": 5213
+    },
+    {
+      "epoch": 0.1098013614610699,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.5499,
+      "step": 5214
+    },
+    {
+      "epoch": 0.10982242041033363,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5681,
+      "step": 5215
+    },
+    {
+      "epoch": 0.10984347935959736,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002,
+      "loss": 1.5947,
+      "step": 5216
+    },
+    {
+      "epoch": 0.10986453830886107,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.581,
+      "step": 5217
+    },
+    {
+      "epoch": 0.1098855972581248,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.5667,
+      "step": 5218
+    },
+    {
+      "epoch": 0.10990665620738853,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0002,
+      "loss": 1.5796,
+      "step": 5219
+    },
+    {
+      "epoch": 0.10992771515665226,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.5794,
+      "step": 5220
+    },
+    {
+      "epoch": 0.10994877410591598,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.5737,
+      "step": 5221
+    },
+    {
+      "epoch": 0.10996983305517971,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6114,
+      "step": 5222
+    },
+    {
+      "epoch": 0.10999089200444344,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.5648,
+      "step": 5223
+    },
+    {
+      "epoch": 0.11001195095370717,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.5917,
+      "step": 5224
+    },
+    {
+      "epoch": 0.1100330099029709,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0002,
+      "loss": 1.5727,
+      "step": 5225
+    },
+    {
+      "epoch": 0.11005406885223462,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5657,
+      "step": 5226
+    },
+    {
+      "epoch": 0.11007512780149835,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.5701,
+      "step": 5227
+    },
+    {
+      "epoch": 0.11009618675076208,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.589,
+      "step": 5228
+    },
+    {
+      "epoch": 0.1101172457000258,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5689,
+      "step": 5229
+    },
+    {
+      "epoch": 0.11013830464928952,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0002,
+      "loss": 1.5683,
+      "step": 5230
+    },
+    {
+      "epoch": 0.11015936359855324,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0002,
+      "loss": 1.6193,
+      "step": 5231
+    },
+    {
+      "epoch": 0.11018042254781697,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0002,
+      "loss": 1.5946,
+      "step": 5232
+    },
+    {
+      "epoch": 0.1102014814970807,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.5773,
+      "step": 5233
+    },
+    {
+      "epoch": 0.11022254044634443,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.5677,
+      "step": 5234
+    },
+    {
+      "epoch": 0.11024359939560816,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0002,
+      "loss": 1.5716,
+      "step": 5235
+    },
+    {
+      "epoch": 0.11026465834487188,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0002,
+      "loss": 1.5938,
+      "step": 5236
+    },
+    {
+      "epoch": 0.11028571729413561,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5791,
+      "step": 5237
+    },
+    {
+      "epoch": 0.11030677624339934,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5624,
+      "step": 5238
+    },
+    {
+      "epoch": 0.11032783519266307,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0002,
+      "loss": 1.6113,
+      "step": 5239
+    },
+    {
+      "epoch": 0.11034889414192679,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.6147,
+      "step": 5240
+    },
+    {
+      "epoch": 0.11036995309119052,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0002,
+      "loss": 1.6148,
+      "step": 5241
+    },
+    {
+      "epoch": 0.11039101204045425,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.5825,
+      "step": 5242
+    },
+    {
+      "epoch": 0.11041207098971796,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.5944,
+      "step": 5243
+    },
+    {
+      "epoch": 0.11043312993898169,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.5752,
+      "step": 5244
+    },
+    {
+      "epoch": 0.11045418888824542,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0002,
+      "loss": 1.5438,
+      "step": 5245
+    },
+    {
+      "epoch": 0.11047524783750914,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.5494,
+      "step": 5246
+    },
+    {
+      "epoch": 0.11049630678677287,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0002,
+      "loss": 1.5801,
+      "step": 5247
+    },
+    {
+      "epoch": 0.1105173657360366,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.559,
+      "step": 5248
+    },
+    {
+      "epoch": 0.11053842468530033,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002,
+      "loss": 1.5978,
+      "step": 5249
+    },
+    {
+      "epoch": 0.11055948363456405,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0002,
+      "loss": 1.55,
+      "step": 5250
+    },
+    {
+      "epoch": 0.11058054258382778,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0002,
+      "loss": 1.5341,
+      "step": 5251
+    },
+    {
+      "epoch": 0.11060160153309151,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0002,
+      "loss": 1.5734,
+      "step": 5252
+    },
+    {
+      "epoch": 0.11062266048235524,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0002,
+      "loss": 1.5993,
+      "step": 5253
+    },
+    {
+      "epoch": 0.11064371943161896,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0002,
+      "loss": 1.5763,
+      "step": 5254
+    },
+    {
+      "epoch": 0.11066477838088269,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0002,
+      "loss": 1.5968,
+      "step": 5255
+    },
+    {
+      "epoch": 0.1106858373301464,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0002,
+      "loss": 1.5602,
+      "step": 5256
+    },
+    {
+      "epoch": 0.11070689627941013,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.5588,
+      "step": 5257
+    },
+    {
+      "epoch": 0.11072795522867386,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5588,
+      "step": 5258
+    },
+    {
+      "epoch": 0.11074901417793759,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6229,
+      "step": 5259
+    },
+    {
+      "epoch": 0.11077007312720132,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.5807,
+      "step": 5260
+    },
+    {
+      "epoch": 0.11079113207646504,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.5508,
+      "step": 5261
+    },
+    {
+      "epoch": 0.11081219102572877,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.5502,
+      "step": 5262
+    },
+    {
+      "epoch": 0.1108332499749925,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0002,
+      "loss": 1.5654,
+      "step": 5263
+    },
+    {
+      "epoch": 0.11085430892425623,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0002,
+      "loss": 1.5988,
+      "step": 5264
+    },
+    {
+      "epoch": 0.11087536787351995,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.5903,
+      "step": 5265
+    },
+    {
+      "epoch": 0.11089642682278368,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0002,
+      "loss": 1.5992,
+      "step": 5266
+    },
+    {
+      "epoch": 0.11091748577204741,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.61,
+      "step": 5267
+    },
+    {
+      "epoch": 0.11093854472131114,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.0002,
+      "loss": 1.598,
+      "step": 5268
+    },
+    {
+      "epoch": 0.11095960367057486,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6084,
+      "step": 5269
+    },
+    {
+      "epoch": 0.11098066261983858,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.6052,
+      "step": 5270
+    },
+    {
+      "epoch": 0.1110017215691023,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0002,
+      "loss": 1.5933,
+      "step": 5271
+    },
+    {
+      "epoch": 0.11102278051836603,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0002,
+      "loss": 1.5867,
+      "step": 5272
+    },
+    {
+      "epoch": 0.11104383946762976,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.5912,
+      "step": 5273
+    },
+    {
+      "epoch": 0.11106489841689349,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0002,
+      "loss": 1.5827,
+      "step": 5274
+    },
+    {
+      "epoch": 0.11108595736615721,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.0002,
+      "loss": 1.5917,
+      "step": 5275
+    },
+    {
+      "epoch": 0.11110701631542094,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0002,
+      "loss": 1.5671,
+      "step": 5276
+    },
+    {
+      "epoch": 0.11112807526468467,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.0002,
+      "loss": 1.5716,
+      "step": 5277
+    },
+    {
+      "epoch": 0.1111491342139484,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.5992,
+      "step": 5278
+    },
+    {
+      "epoch": 0.11117019316321212,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.5805,
+      "step": 5279
+    },
+    {
+      "epoch": 0.11119125211247585,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.5933,
+      "step": 5280
+    },
+    {
+      "epoch": 0.11121231106173958,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0002,
+      "loss": 1.5723,
+      "step": 5281
+    },
+    {
+      "epoch": 0.11123337001100331,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.6037,
+      "step": 5282
+    },
+    {
+      "epoch": 0.11125442896026702,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.5609,
+      "step": 5283
+    },
+    {
+      "epoch": 0.11127548790953075,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0002,
+      "loss": 1.563,
+      "step": 5284
+    },
+    {
+      "epoch": 0.11129654685879448,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.5795,
+      "step": 5285
+    },
+    {
+      "epoch": 0.1113176058080582,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.5911,
+      "step": 5286
+    },
+    {
+      "epoch": 0.11133866475732193,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.5958,
+      "step": 5287
+    },
+    {
+      "epoch": 0.11135972370658566,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.5768,
+      "step": 5288
+    },
+    {
+      "epoch": 0.11138078265584939,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.5656,
+      "step": 5289
+    },
+    {
+      "epoch": 0.11140184160511311,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5752,
+      "step": 5290
+    },
+    {
+      "epoch": 0.11142290055437684,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.0002,
+      "loss": 1.5812,
+      "step": 5291
+    },
+    {
+      "epoch": 0.11144395950364057,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.0002,
+      "loss": 1.5645,
+      "step": 5292
+    },
+    {
+      "epoch": 0.1114650184529043,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.572,
+      "step": 5293
+    },
+    {
+      "epoch": 0.11148607740216802,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.0002,
+      "loss": 1.5825,
+      "step": 5294
+    },
+    {
+      "epoch": 0.11150713635143175,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.6001,
+      "step": 5295
+    },
+    {
+      "epoch": 0.11152819530069547,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0002,
+      "loss": 1.5634,
+      "step": 5296
+    },
+    {
+      "epoch": 0.11154925424995919,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.5846,
+      "step": 5297
+    },
+    {
+      "epoch": 0.11157031319922292,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0002,
+      "loss": 1.5638,
+      "step": 5298
+    },
+    {
+      "epoch": 0.11159137214848665,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0002,
+      "loss": 1.5392,
+      "step": 5299
+    },
+    {
+      "epoch": 0.11161243109775038,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0002,
+      "loss": 1.5668,
+      "step": 5300
+    },
+    {
+      "epoch": 0.1116334900470141,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0002,
+      "loss": 1.5751,
+      "step": 5301
+    },
+    {
+      "epoch": 0.11165454899627783,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.6205,
+      "step": 5302
+    },
+    {
+      "epoch": 0.11167560794554156,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.569,
+      "step": 5303
+    },
+    {
+      "epoch": 0.11169666689480529,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0002,
+      "loss": 1.5713,
+      "step": 5304
+    },
+    {
+      "epoch": 0.11171772584406901,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.6086,
+      "step": 5305
+    },
+    {
+      "epoch": 0.11173878479333274,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.6032,
+      "step": 5306
+    },
+    {
+      "epoch": 0.11175984374259647,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.5887,
+      "step": 5307
+    },
+    {
+      "epoch": 0.1117809026918602,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.5904,
+      "step": 5308
+    },
+    {
+      "epoch": 0.11180196164112391,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0002,
+      "loss": 1.6022,
+      "step": 5309
+    },
+    {
+      "epoch": 0.11182302059038764,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0002,
+      "loss": 1.6052,
+      "step": 5310
+    },
+    {
+      "epoch": 0.11184407953965136,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0002,
+      "loss": 1.5822,
+      "step": 5311
+    },
+    {
+      "epoch": 0.11186513848891509,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5844,
+      "step": 5312
+    },
+    {
+      "epoch": 0.11188619743817882,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0002,
+      "loss": 1.6124,
+      "step": 5313
+    },
+    {
+      "epoch": 0.11190725638744255,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.5607,
+      "step": 5314
+    },
+    {
+      "epoch": 0.11192831533670627,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.5705,
+      "step": 5315
+    },
+    {
+      "epoch": 0.11194937428597,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5617,
+      "step": 5316
+    },
+    {
+      "epoch": 0.11197043323523373,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.6148,
+      "step": 5317
+    },
+    {
+      "epoch": 0.11199149218449746,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0002,
+      "loss": 1.5705,
+      "step": 5318
+    },
+    {
+      "epoch": 0.11201255113376118,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5304,
+      "step": 5319
+    },
+    {
+      "epoch": 0.11203361008302491,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0002,
+      "loss": 1.5561,
+      "step": 5320
+    },
+    {
+      "epoch": 0.11205466903228864,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0002,
+      "loss": 1.5424,
+      "step": 5321
+    },
+    {
+      "epoch": 0.11207572798155237,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.0002,
+      "loss": 1.5863,
+      "step": 5322
+    },
+    {
+      "epoch": 0.11209678693081608,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.6091,
+      "step": 5323
+    },
+    {
+      "epoch": 0.11211784588007981,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0002,
+      "loss": 1.5968,
+      "step": 5324
+    },
+    {
+      "epoch": 0.11213890482934354,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.5568,
+      "step": 5325
+    },
+    {
+      "epoch": 0.11215996377860726,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.5975,
+      "step": 5326
+    },
+    {
+      "epoch": 0.11218102272787099,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.5702,
+      "step": 5327
+    },
+    {
+      "epoch": 0.11220208167713472,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.588,
+      "step": 5328
+    },
+    {
+      "epoch": 0.11222314062639845,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0002,
+      "loss": 1.5807,
+      "step": 5329
+    },
+    {
+      "epoch": 0.11224419957566217,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.5695,
+      "step": 5330
+    },
+    {
+      "epoch": 0.1122652585249259,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0002,
+      "loss": 1.5798,
+      "step": 5331
+    },
+    {
+      "epoch": 0.11228631747418963,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0002,
+      "loss": 1.595,
+      "step": 5332
+    },
+    {
+      "epoch": 0.11230737642345336,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.595,
+      "step": 5333
+    },
+    {
+      "epoch": 0.11232843537271708,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5856,
+      "step": 5334
+    },
+    {
+      "epoch": 0.11234949432198081,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002,
+      "loss": 1.603,
+      "step": 5335
+    },
+    {
+      "epoch": 0.11237055327124452,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.6207,
+      "step": 5336
+    },
+    {
+      "epoch": 0.11239161222050825,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5916,
+      "step": 5337
+    },
+    {
+      "epoch": 0.11241267116977198,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.5937,
+      "step": 5338
+    },
+    {
+      "epoch": 0.11243373011903571,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0002,
+      "loss": 1.5823,
+      "step": 5339
+    },
+    {
+      "epoch": 0.11245478906829943,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.5692,
+      "step": 5340
+    },
+    {
+      "epoch": 0.11247584801756316,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.6058,
+      "step": 5341
+    },
+    {
+      "epoch": 0.11249690696682689,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.5844,
+      "step": 5342
+    },
+    {
+      "epoch": 0.11251796591609062,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5735,
+      "step": 5343
+    },
+    {
+      "epoch": 0.11253902486535435,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0002,
+      "loss": 1.6018,
+      "step": 5344
+    },
+    {
+      "epoch": 0.11256008381461807,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0002,
+      "loss": 1.5645,
+      "step": 5345
+    },
+    {
+      "epoch": 0.1125811427638818,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.5649,
+      "step": 5346
+    },
+    {
+      "epoch": 0.11260220171314553,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0002,
+      "loss": 1.6068,
+      "step": 5347
+    },
+    {
+      "epoch": 0.11262326066240926,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.5468,
+      "step": 5348
+    },
+    {
+      "epoch": 0.11264431961167297,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5454,
+      "step": 5349
+    },
+    {
+      "epoch": 0.1126653785609367,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0002,
+      "loss": 1.5777,
+      "step": 5350
+    },
+    {
+      "epoch": 0.11268643751020042,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0002,
+      "loss": 1.5436,
+      "step": 5351
+    },
+    {
+      "epoch": 0.11270749645946415,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.5846,
+      "step": 5352
+    },
+    {
+      "epoch": 0.11272855540872788,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0002,
+      "loss": 1.5699,
+      "step": 5353
+    },
+    {
+      "epoch": 0.1127496143579916,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.0002,
+      "loss": 1.6036,
+      "step": 5354
+    },
+    {
+      "epoch": 0.11277067330725533,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5748,
+      "step": 5355
+    },
+    {
+      "epoch": 0.11279173225651906,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0002,
+      "loss": 1.5741,
+      "step": 5356
+    },
+    {
+      "epoch": 0.11281279120578279,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.6056,
+      "step": 5357
+    },
+    {
+      "epoch": 0.11283385015504652,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0002,
+      "loss": 1.5673,
+      "step": 5358
+    },
+    {
+      "epoch": 0.11285490910431024,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0002,
+      "loss": 1.6194,
+      "step": 5359
+    },
+    {
+      "epoch": 0.11287596805357397,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0002,
+      "loss": 1.5633,
+      "step": 5360
+    },
+    {
+      "epoch": 0.1128970270028377,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.0002,
+      "loss": 1.5882,
+      "step": 5361
+    },
+    {
+      "epoch": 0.11291808595210141,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0002,
+      "loss": 1.5791,
+      "step": 5362
+    },
+    {
+      "epoch": 0.11293914490136514,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002,
+      "loss": 1.5739,
+      "step": 5363
+    },
+    {
+      "epoch": 0.11296020385062887,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.5563,
+      "step": 5364
+    },
+    {
+      "epoch": 0.1129812627998926,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0002,
+      "loss": 1.5849,
+      "step": 5365
+    },
+    {
+      "epoch": 0.11300232174915632,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.5887,
+      "step": 5366
+    },
+    {
+      "epoch": 0.11302338069842005,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.551,
+      "step": 5367
+    },
+    {
+      "epoch": 0.11304443964768378,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.5716,
+      "step": 5368
+    },
+    {
+      "epoch": 0.1130654985969475,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0002,
+      "loss": 1.5929,
+      "step": 5369
+    },
+    {
+      "epoch": 0.11308655754621123,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.5775,
+      "step": 5370
+    },
+    {
+      "epoch": 0.11310761649547496,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002,
+      "loss": 1.596,
+      "step": 5371
+    },
+    {
+      "epoch": 0.11312867544473869,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002,
+      "loss": 1.6188,
+      "step": 5372
+    },
+    {
+      "epoch": 0.11314973439400242,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.6115,
+      "step": 5373
+    },
+    {
+      "epoch": 0.11317079334326614,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.6004,
+      "step": 5374
+    },
+    {
+      "epoch": 0.11319185229252987,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0002,
+      "loss": 1.5701,
+      "step": 5375
+    },
+    {
+      "epoch": 0.11321291124179358,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.6098,
+      "step": 5376
+    },
+    {
+      "epoch": 0.11323397019105731,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.5986,
+      "step": 5377
+    },
+    {
+      "epoch": 0.11325502914032104,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0002,
+      "loss": 1.6023,
+      "step": 5378
+    },
+    {
+      "epoch": 0.11327608808958477,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0002,
+      "loss": 1.5505,
+      "step": 5379
+    },
+    {
+      "epoch": 0.1132971470388485,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0002,
+      "loss": 1.5682,
+      "step": 5380
+    },
+    {
+      "epoch": 0.11331820598811222,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.6159,
+      "step": 5381
+    },
+    {
+      "epoch": 0.11333926493737595,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0002,
+      "loss": 1.5783,
+      "step": 5382
+    },
+    {
+      "epoch": 0.11336032388663968,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0002,
+      "loss": 1.5449,
+      "step": 5383
+    },
+    {
+      "epoch": 0.1133813828359034,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0002,
+      "loss": 1.5492,
+      "step": 5384
+    },
+    {
+      "epoch": 0.11340244178516713,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.5663,
+      "step": 5385
+    },
+    {
+      "epoch": 0.11342350073443086,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.5579,
+      "step": 5386
+    },
+    {
+      "epoch": 0.11344455968369459,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0002,
+      "loss": 1.5808,
+      "step": 5387
+    },
+    {
+      "epoch": 0.11346561863295831,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.589,
+      "step": 5388
+    },
+    {
+      "epoch": 0.11348667758222203,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.5899,
+      "step": 5389
+    },
+    {
+      "epoch": 0.11350773653148576,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.5711,
+      "step": 5390
+    },
+    {
+      "epoch": 0.11352879548074948,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.5641,
+      "step": 5391
+    },
+    {
+      "epoch": 0.11354985443001321,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0002,
+      "loss": 1.5746,
+      "step": 5392
+    },
+    {
+      "epoch": 0.11357091337927694,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0002,
+      "loss": 1.5703,
+      "step": 5393
+    },
+    {
+      "epoch": 0.11359197232854067,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.5952,
+      "step": 5394
+    },
+    {
+      "epoch": 0.1136130312778044,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5547,
+      "step": 5395
+    },
+    {
+      "epoch": 0.11363409022706812,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0002,
+      "loss": 1.5511,
+      "step": 5396
+    },
+    {
+      "epoch": 0.11365514917633185,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.0002,
+      "loss": 1.5833,
+      "step": 5397
+    },
+    {
+      "epoch": 0.11367620812559558,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0002,
+      "loss": 1.5566,
+      "step": 5398
+    },
+    {
+      "epoch": 0.1136972670748593,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.5658,
+      "step": 5399
+    },
+    {
+      "epoch": 0.11371832602412303,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5584,
+      "step": 5400
+    },
+    {
+      "epoch": 0.11373938497338676,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0002,
+      "loss": 1.5737,
+      "step": 5401
+    },
+    {
+      "epoch": 0.11376044392265047,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.6198,
+      "step": 5402
+    },
+    {
+      "epoch": 0.1137815028719142,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002,
+      "loss": 1.5585,
+      "step": 5403
+    },
+    {
+      "epoch": 0.11380256182117793,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0002,
+      "loss": 1.5867,
+      "step": 5404
+    },
+    {
+      "epoch": 0.11382362077044166,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.572,
+      "step": 5405
+    },
+    {
+      "epoch": 0.11384467971970538,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.5758,
+      "step": 5406
+    },
+    {
+      "epoch": 0.11386573866896911,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002,
+      "loss": 1.5436,
+      "step": 5407
+    },
+    {
+      "epoch": 0.11388679761823284,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0002,
+      "loss": 1.5863,
+      "step": 5408
+    },
+    {
+      "epoch": 0.11390785656749657,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.5911,
+      "step": 5409
+    },
+    {
+      "epoch": 0.11392891551676029,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.5873,
+      "step": 5410
+    },
+    {
+      "epoch": 0.11394997446602402,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.5915,
+      "step": 5411
+    },
+    {
+      "epoch": 0.11397103341528775,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.5908,
+      "step": 5412
+    },
+    {
+      "epoch": 0.11399209236455148,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.5507,
+      "step": 5413
+    },
+    {
+      "epoch": 0.1140131513138152,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0002,
+      "loss": 1.5821,
+      "step": 5414
+    },
+    {
+      "epoch": 0.11403421026307893,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.6239,
+      "step": 5415
+    },
+    {
+      "epoch": 0.11405526921234264,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0002,
+      "loss": 1.6023,
+      "step": 5416
+    },
+    {
+      "epoch": 0.11407632816160637,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.0002,
+      "loss": 1.5531,
+      "step": 5417
+    },
+    {
+      "epoch": 0.1140973871108701,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.573,
+      "step": 5418
+    },
+    {
+      "epoch": 0.11411844606013383,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0002,
+      "loss": 1.5759,
+      "step": 5419
+    },
+    {
+      "epoch": 0.11413950500939755,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0002,
+      "loss": 1.5825,
+      "step": 5420
+    },
+    {
+      "epoch": 0.11416056395866128,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002,
+      "loss": 1.5533,
+      "step": 5421
+    },
+    {
+      "epoch": 0.11418162290792501,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0002,
+      "loss": 1.589,
+      "step": 5422
+    },
+    {
+      "epoch": 0.11420268185718874,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0002,
+      "loss": 1.5615,
+      "step": 5423
+    },
+    {
+      "epoch": 0.11422374080645246,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0002,
+      "loss": 1.6026,
+      "step": 5424
+    },
+    {
+      "epoch": 0.11424479975571619,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0002,
+      "loss": 1.5761,
+      "step": 5425
+    },
+    {
+      "epoch": 0.11426585870497992,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0002,
+      "loss": 1.5704,
+      "step": 5426
+    },
+    {
+      "epoch": 0.11428691765424365,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0002,
+      "loss": 1.5956,
+      "step": 5427
+    },
+    {
+      "epoch": 0.11430797660350737,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.548,
+      "step": 5428
+    },
+    {
+      "epoch": 0.11432903555277109,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0002,
+      "loss": 1.6072,
+      "step": 5429
+    },
+    {
+      "epoch": 0.11435009450203482,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002,
+      "loss": 1.5926,
+      "step": 5430
+    },
+    {
+      "epoch": 0.11437115345129854,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.5584,
+      "step": 5431
+    },
+    {
+      "epoch": 0.11439221240056227,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0002,
+      "loss": 1.5815,
+      "step": 5432
+    },
+    {
+      "epoch": 0.114413271349826,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.0002,
+      "loss": 1.5428,
+      "step": 5433
+    },
+    {
+      "epoch": 0.11443433029908973,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0002,
+      "loss": 1.5716,
+      "step": 5434
+    },
+    {
+      "epoch": 0.11445538924835345,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0002,
+      "loss": 1.5553,
+      "step": 5435
+    },
+    {
+      "epoch": 0.11447644819761718,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.5829,
+      "step": 5436
+    },
+    {
+      "epoch": 0.11449750714688091,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5924,
+      "step": 5437
+    },
+    {
+      "epoch": 0.11451856609614464,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5867,
+      "step": 5438
+    },
+    {
+      "epoch": 0.11453962504540836,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.5859,
+      "step": 5439
+    },
+    {
+      "epoch": 0.11456068399467209,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.5278,
+      "step": 5440
+    },
+    {
+      "epoch": 0.11458174294393582,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.6249,
+      "step": 5441
+    },
+    {
+      "epoch": 0.11460280189319953,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0002,
+      "loss": 1.5847,
+      "step": 5442
+    },
+    {
+      "epoch": 0.11462386084246326,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0002,
+      "loss": 1.6403,
+      "step": 5443
+    },
+    {
+      "epoch": 0.11464491979172699,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.5911,
+      "step": 5444
+    },
+    {
+      "epoch": 0.11466597874099071,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.6062,
+      "step": 5445
+    },
+    {
+      "epoch": 0.11468703769025444,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.5879,
+      "step": 5446
+    },
+    {
+      "epoch": 0.11470809663951817,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0002,
+      "loss": 1.5824,
+      "step": 5447
+    },
+    {
+      "epoch": 0.1147291555887819,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.594,
+      "step": 5448
+    },
+    {
+      "epoch": 0.11475021453804563,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.0002,
+      "loss": 1.5621,
+      "step": 5449
+    },
+    {
+      "epoch": 0.11477127348730935,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0002,
+      "loss": 1.551,
+      "step": 5450
+    },
+    {
+      "epoch": 0.11479233243657308,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.6232,
+      "step": 5451
+    },
+    {
+      "epoch": 0.11481339138583681,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0002,
+      "loss": 1.5959,
+      "step": 5452
+    },
+    {
+      "epoch": 0.11483445033510054,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.6236,
+      "step": 5453
+    },
+    {
+      "epoch": 0.11485550928436426,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0002,
+      "loss": 1.5791,
+      "step": 5454
+    },
+    {
+      "epoch": 0.11487656823362798,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002,
+      "loss": 1.5871,
+      "step": 5455
+    },
+    {
+      "epoch": 0.1148976271828917,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5993,
+      "step": 5456
+    },
+    {
+      "epoch": 0.11491868613215543,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.0002,
+      "loss": 1.5406,
+      "step": 5457
+    },
+    {
+      "epoch": 0.11493974508141916,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0002,
+      "loss": 1.5492,
+      "step": 5458
+    },
+    {
+      "epoch": 0.11496080403068289,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.5629,
+      "step": 5459
+    },
+    {
+      "epoch": 0.11498186297994661,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002,
+      "loss": 1.6103,
+      "step": 5460
+    },
+    {
+      "epoch": 0.11500292192921034,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.5556,
+      "step": 5461
+    },
+    {
+      "epoch": 0.11502398087847407,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.5877,
+      "step": 5462
+    },
+    {
+      "epoch": 0.1150450398277378,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0002,
+      "loss": 1.5568,
+      "step": 5463
+    },
+    {
+      "epoch": 0.11506609877700152,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0002,
+      "loss": 1.6126,
+      "step": 5464
+    },
+    {
+      "epoch": 0.11508715772626525,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0002,
+      "loss": 1.5748,
+      "step": 5465
+    },
+    {
+      "epoch": 0.11510821667552898,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.6023,
+      "step": 5466
+    },
+    {
+      "epoch": 0.1151292756247927,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002,
+      "loss": 1.6044,
+      "step": 5467
+    },
+    {
+      "epoch": 0.11515033457405643,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.5485,
+      "step": 5468
+    },
+    {
+      "epoch": 0.11517139352332015,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.0002,
+      "loss": 1.5993,
+      "step": 5469
+    },
+    {
+      "epoch": 0.11519245247258388,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.0002,
+      "loss": 1.5913,
+      "step": 5470
+    },
+    {
+      "epoch": 0.1152135114218476,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0002,
+      "loss": 1.5698,
+      "step": 5471
+    },
+    {
+      "epoch": 0.11523457037111133,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002,
+      "loss": 1.5892,
+      "step": 5472
+    },
+    {
+      "epoch": 0.11525562932037506,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.592,
+      "step": 5473
+    },
+    {
+      "epoch": 0.11527668826963879,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0002,
+      "loss": 1.578,
+      "step": 5474
+    },
+    {
+      "epoch": 0.11529774721890251,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.5853,
+      "step": 5475
+    },
+    {
+      "epoch": 0.11531880616816624,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0002,
+      "loss": 1.6153,
+      "step": 5476
+    },
+    {
+      "epoch": 0.11533986511742997,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002,
+      "loss": 1.56,
+      "step": 5477
+    },
+    {
+      "epoch": 0.1153609240666937,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0002,
+      "loss": 1.5879,
+      "step": 5478
+    },
+    {
+      "epoch": 0.11538198301595742,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0002,
+      "loss": 1.5511,
+      "step": 5479
+    },
+    {
+      "epoch": 0.11540304196522115,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002,
+      "loss": 1.5615,
+      "step": 5480
+    },
+    {
+      "epoch": 0.11542410091448488,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0002,
+      "loss": 1.5635,
+      "step": 5481
+    },
+    {
+      "epoch": 0.11544515986374859,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002,
+      "loss": 1.5894,
+      "step": 5482
+    },
+    {
+      "epoch": 0.11546621881301232,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002,
+      "loss": 1.5455,
+      "step": 5483
+    },
+    {
+      "epoch": 0.11548727776227605,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.6208,
+      "step": 5484
+    },
+    {
+      "epoch": 0.11550833671153977,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0002,
+      "loss": 1.5363,
+      "step": 5485
+    },
+    {
+      "epoch": 0.1155293956608035,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002,
+      "loss": 1.6207,
+      "step": 5486
+    },
+    {
+      "epoch": 0.11555045461006723,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.6014,
+      "step": 5487
+    },
+    {
+      "epoch": 0.11557151355933096,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0002,
+      "loss": 1.6124,
+      "step": 5488
+    },
+    {
+      "epoch": 0.11559257250859468,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002,
+      "loss": 1.61,
+      "step": 5489
+    },
+    {
+      "epoch": 0.11561363145785841,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0002,
+      "loss": 1.5642,
+      "step": 5490
+    },
+    {
+      "epoch": 0.11563469040712214,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0002,
+      "loss": 1.5958,
+      "step": 5491
+    },
+    {
+      "epoch": 0.11565574935638587,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0002,
+      "loss": 1.5623,
+      "step": 5492
+    },
+    {
+      "epoch": 0.1156768083056496,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0002,
+      "loss": 1.6008,
+      "step": 5493
+    },
+    {
+      "epoch": 0.11569786725491332,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0002,
+      "loss": 1.5397,
+      "step": 5494
+    },
+    {
+      "epoch": 0.11571892620417704,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.592,
+      "step": 5495
+    },
+    {
+      "epoch": 0.11573998515344076,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002,
+      "loss": 1.5626,
+      "step": 5496
+    },
+    {
+      "epoch": 0.11576104410270449,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0002,
+      "loss": 1.5662,
+      "step": 5497
+    },
+    {
+      "epoch": 0.11578210305196822,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.0002,
+      "loss": 1.5576,
+      "step": 5498
+    },
+    {
+      "epoch": 0.11580316200123195,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002,
+      "loss": 1.5579,
+      "step": 5499
+    },
+    {
+      "epoch": 0.11582422095049567,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002,
+      "loss": 1.5457,
+      "step": 5500
+    },
+    {
+      "epoch": 0.11582422095049567,
+      "eval_loss": 2.3854761123657227,
+      "eval_runtime": 1003.3844,
+      "eval_samples_per_second": 61.592,
+      "eval_steps_per_second": 1.925,
+      "step": 5500
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 47485,
+  "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": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 2.7218310856704e+18,
+  "train_batch_size": 32,
+  "trial_name": null,
+  "trial_params": null
+}